From 38fe80f45ef37283434ad5d03e7fdbff02b6c8a9 Mon Sep 17 00:00:00 2001
From: "Documenter.jl" <documenter@juliadocs.github.io>
Date: Sat, 23 Sep 2023 05:10:02 +0000
Subject: [PATCH] build based on a2e6a02

---
 previews/PR3525/.documenter-siteinfo.json     |    1 +
 previews/PR3525/JuMP.pdf                      |  Bin 0 -> 4687304 bytes
 .../PR3525/api/JuMP.Containers/index.html     |  192 ++
 previews/PR3525/api/JuMP/index.html           | 2074 ++++++++++++++
 previews/PR3525/assets/case9mod.png           |  Bin 0 -> 234660 bytes
 previews/PR3525/assets/documenter.js          |  883 ++++++
 previews/PR3525/assets/extra_styles.css       |    4 +
 previews/PR3525/assets/full_sudoku.png        |  Bin 0 -> 119423 bytes
 .../assets/logo-dark-with-text-background.svg |   48 +
 .../PR3525/assets/logo-dark-with-text.svg     |   47 +
 .../PR3525/assets/logo-dark-without-text.svg  |   46 +
 previews/PR3525/assets/logo-dark.svg          |   47 +
 .../assets/logo-with-text-background.svg      |   48 +
 previews/PR3525/assets/logo-with-text.pdf     |  Bin 0 -> 2157 bytes
 previews/PR3525/assets/logo-with-text.svg     |   47 +
 previews/PR3525/assets/logo-without-text.svg  |   46 +
 previews/PR3525/assets/logo.svg               |   47 +
 previews/PR3525/assets/logo_constructor.jl    |  395 +++
 previews/PR3525/assets/n_queens4.png          |  Bin 0 -> 8151 bytes
 previews/PR3525/assets/numfocus-logo.png      |  Bin 0 -> 4883 bytes
 previews/PR3525/assets/partial_sudoku.png     |  Bin 0 -> 71711 bytes
 previews/PR3525/assets/power_systems.png      |  Bin 0 -> 23020 bytes
 previews/PR3525/assets/space_shuttle.png      |  Bin 0 -> 292247 bytes
 .../PR3525/assets/themes/documenter-dark.css  |    7 +
 .../PR3525/assets/themes/documenter-light.css |    9 +
 previews/PR3525/assets/themeswap.js           |   84 +
 previews/PR3525/assets/warner.js              |   52 +
 .../algebraic_modeling_languages/index.html   |  129 +
 previews/PR3525/changelog/index.html          |   15 +
 .../PR3525/developers/checklists/index.html   |   65 +
 .../PR3525/developers/contributing/index.html |   28 +
 .../custom_solver_binaries/index.html         |   93 +
 .../PR3525/developers/extensions/index.html   |  304 ++
 previews/PR3525/developers/roadmap/index.html |    6 +
 previews/PR3525/developers/style/index.html   |  185 ++
 .../extensions/DimensionalData/index.html     |   42 +
 .../PR3525/extensions/introduction/index.html |    6 +
 previews/PR3525/index.html                    |   13 +
 previews/PR3525/installation/index.html       |   24 +
 previews/PR3525/manual/callbacks/index.html   |   87 +
 previews/PR3525/manual/complex/index.html     |  151 +
 previews/PR3525/manual/constraints/index.html |  737 +++++
 previews/PR3525/manual/containers/index.html  |  227 ++
 previews/PR3525/manual/expressions/index.html |  230 ++
 previews/PR3525/manual/models/index.html      |  237 ++
 previews/PR3525/manual/nlp/index.html         |  347 +++
 previews/PR3525/manual/nonlinear/index.html   |  199 ++
 previews/PR3525/manual/objective/index.html   |  168 ++
 previews/PR3525/manual/solutions/index.html   |  374 +++
 previews/PR3525/manual/variables/index.html   |  637 +++++
 previews/PR3525/moi/assets/latency.png        |  Bin 0 -> 25175 bytes
 .../PR3525/moi/background/duality/index.html  |   84 +
 .../infeasibility_certificates/index.html     |   32 +
 .../moi/background/motivation/index.html      |    6 +
 .../background/naming_conventions/index.html  |    6 +
 previews/PR3525/moi/changelog/index.html      |   34 +
 .../moi/developer/checklists/index.html       |  116 +
 previews/PR3525/moi/index.html                |   13 +
 .../PR3525/moi/manual/constraints/index.html  |   26 +
 previews/PR3525/moi/manual/models/index.html  |    6 +
 .../PR3525/moi/manual/modification/index.html |  129 +
 .../PR3525/moi/manual/solutions/index.html    |   39 +
 .../moi/manual/standard_form/index.html       |   10 +
 .../PR3525/moi/manual/variables/index.html    |   17 +
 .../PR3525/moi/reference/callbacks/index.html |   36 +
 .../moi/reference/constraints/index.html      |   16 +
 .../PR3525/moi/reference/errors/index.html    |   52 +
 .../PR3525/moi/reference/models/index.html    |  136 +
 .../moi/reference/modification/index.html     |   31 +
 .../PR3525/moi/reference/nonlinear/index.html |   82 +
 .../moi/reference/standard_form/index.html    |  993 +++++++
 .../PR3525/moi/reference/variables/index.html |   28 +
 previews/PR3525/moi/release_notes/index.html  |   34 +
 .../submodules/Benchmarks/overview/index.html |   24 +
 .../Benchmarks/reference/index.html           |   21 +
 .../Bridges/list_of_bridges/index.html        |  114 +
 .../submodules/Bridges/overview/index.html    |   57 +
 .../submodules/Bridges/reference/index.html   |  199 ++
 .../FileFormats/overview/index.html           |  141 +
 .../FileFormats/reference/index.html          |   22 +
 .../submodules/Nonlinear/overview/index.html  |  187 ++
 .../submodules/Nonlinear/reference/index.html |  151 +
 .../moi/submodules/Test/overview/index.html   |  170 ++
 .../moi/submodules/Test/reference/index.html  |   54 +
 .../submodules/Utilities/overview/index.html  |  254 ++
 .../submodules/Utilities/reference/index.html |  268 ++
 .../tutorials/bridging_constraint/index.html  |  106 +
 .../PR3525/moi/tutorials/example/index.html   |   49 +
 .../moi/tutorials/implementing/index.html     |  118 +
 .../PR3525/moi/tutorials/latency/index.html   |  133 +
 .../manipulating_expressions/index.html       |   26 +
 .../moi/tutorials/mathprogbase/index.html     |   58 +
 previews/PR3525/packages/Alpine/index.html    |   49 +
 .../PR3525/packages/AmplNLWriter/index.html   |   15 +
 previews/PR3525/packages/BARON/index.html     |    9 +
 .../PR3525/packages/BilevelJuMP/index.html    |   37 +
 previews/PR3525/packages/CDCS/index.html      |   30 +
 previews/PR3525/packages/CDDLib/index.html    |    9 +
 previews/PR3525/packages/COPT/index.html      |   42 +
 previews/PR3525/packages/COSMO/index.html     |   37 +
 previews/PR3525/packages/CPLEX/index.html     |  164 ++
 previews/PR3525/packages/CSDP/index.html      |   13 +
 previews/PR3525/packages/Cbc/index.html       |   12 +
 previews/PR3525/packages/Clarabel/index.html  |   26 +
 previews/PR3525/packages/Clp/index.html       |   10 +
 previews/PR3525/packages/DAQP/index.html      |    8 +
 previews/PR3525/packages/DiffOpt/index.html   |   23 +
 .../PR3525/packages/Dualization/index.html    |   13 +
 previews/PR3525/packages/ECOS/index.html      |    9 +
 previews/PR3525/packages/GAMS/index.html      |   25 +
 previews/PR3525/packages/GLPK/index.html      |   39 +
 previews/PR3525/packages/Gurobi/index.html    |   97 +
 previews/PR3525/packages/HiGHS/index.html     |   10 +
 previews/PR3525/packages/Hypatia/index.html   |   45 +
 .../PR3525/packages/InfiniteOpt/index.html    |   15 +
 previews/PR3525/packages/Ipopt/index.html     |  118 +
 previews/PR3525/packages/Juniper/index.html   |   36 +
 previews/PR3525/packages/KNITRO/index.html    |   13 +
 previews/PR3525/packages/Loraine/index.html   |   32 +
 previews/PR3525/packages/MadNLP/index.html    |   48 +
 previews/PR3525/packages/MiniZinc/index.html  |   57 +
 .../PR3525/packages/MosekTools/index.html     |   10 +
 .../MultiObjectiveAlgorithms/index.html       |   12 +
 .../PR3525/packages/NEOSServer/index.html     |   31 +
 previews/PR3525/packages/NLopt/index.html     |    9 +
 previews/PR3525/packages/OSQP/index.html      |    9 +
 .../PR3525/packages/PATHSolver/index.html     |  168 ++
 previews/PR3525/packages/Pajarito/index.html  |   30 +
 .../ParametricOptInterface/index.html         |   16 +
 previews/PR3525/packages/Pavito/index.html    |   16 +
 previews/PR3525/packages/Plasmo/index.html    |   49 +
 previews/PR3525/packages/PolyJuMP/index.html  |    7 +
 previews/PR3525/packages/ProxSDP/index.html   |   59 +
 previews/PR3525/packages/SCIP/index.html      |   13 +
 previews/PR3525/packages/SCS/index.html       |   71 +
 previews/PR3525/packages/SDDP/index.html      |    6 +
 previews/PR3525/packages/SDPA/index.html      |   15 +
 previews/PR3525/packages/SDPNAL/index.html    |   21 +
 previews/PR3525/packages/SDPT3/index.html     |   32 +
 previews/PR3525/packages/SeDuMi/index.html    |   20 +
 .../PR3525/packages/SumOfSquares/index.html   |    7 +
 previews/PR3525/packages/Tulip/index.html     |   31 +
 previews/PR3525/packages/Xpress/index.html    |   11 +
 previews/PR3525/packages/solvers/index.html   |    6 +
 previews/PR3525/release_notes/index.html      |   15 +
 previews/PR3525/search_index.js               |    3 +
 previews/PR3525/should_i_use/index.html       |    6 +
 previews/PR3525/siteinfo.js                   |    1 +
 .../algorithms/benders_decomposition.jl       |  379 +++
 .../benders_decomposition/index.html          |  189 ++
 .../cutting_stock_column_generation.jl        |  386 +++
 .../703571d3.svg                              |  350 +++
 .../fa9bdba4.svg                              |  516 ++++
 .../index.html                                |  290 ++
 .../algorithms/parallelism/index.html         |  129 +
 .../algorithms/tsp_lazy_constraints.jl        |  277 ++
 .../tsp_lazy_constraints/43baf431.svg         |  122 +
 .../tsp_lazy_constraints/8946bc6a.svg         |  122 +
 .../tsp_lazy_constraints/index.html           |   94 +
 .../applications/optimal_power_flow.jl        |  470 ++++
 .../optimal_power_flow/index.html             |  232 ++
 .../tutorials/applications/power_systems.jl   |  559 ++++
 .../applications/power_systems/6fee809a.svg   |   43 +
 .../applications/power_systems/9f136f96.svg   |   80 +
 .../applications/power_systems/ab8a43ee.svg   |   78 +
 .../applications/power_systems/index.html     |  394 +++
 .../applications/two_stage_stochastic.jl      |  239 ++
 .../two_stage_stochastic/01f8597b.svg         |  108 +
 .../two_stage_stochastic/27244c58.svg         |  100 +
 .../two_stage_stochastic/a5d2410e.svg         |   76 +
 .../two_stage_stochastic/ddf1a4d6.svg         |   80 +
 .../two_stage_stochastic/index.html           |  140 +
 .../PR3525/tutorials/applications/web_app.jl  |  165 ++
 .../tutorials/applications/web_app/index.html |   91 +
 .../tutorials/conic/arbitrary_precision.jl    |  156 ++
 .../conic/arbitrary_precision/index.html      |   80 +
 .../PR3525/tutorials/conic/dualization.jl     |  255 ++
 .../tutorials/conic/dualization/index.html    |  213 ++
 .../PR3525/tutorials/conic/ellipse_approx.jl  |  262 ++
 .../conic/ellipse_approx/07c030f6.svg         | 2450 ++++++++++++++++
 .../conic/ellipse_approx/e71dc9be.svg         | 2451 +++++++++++++++++
 .../tutorials/conic/ellipse_approx/index.html |  341 +++
 .../tutorials/conic/experiment_design.jl      |  218 ++
 .../conic/experiment_design/index.html        |   98 +
 .../tutorials/conic/introduction/index.html   |    9 +
 .../tutorials/conic/logistic_regression.jl    |  275 ++
 .../conic/logistic_regression/index.html      |  107 +
 .../PR3525/tutorials/conic/min_ellipse.jl     |  151 +
 .../tutorials/conic/min_ellipse/4146ff43.svg  |   53 +
 .../tutorials/conic/min_ellipse/804a7acd.svg  |   50 +
 .../tutorials/conic/min_ellipse/index.html    |   95 +
 .../tutorials/conic/quantum_discrimination.jl |  147 +
 .../conic/quantum_discrimination/index.html   |   73 +
 .../PR3525/tutorials/conic/simple_examples.jl |  422 +++
 .../conic/simple_examples/ee174d1a.svg        |   49 +
 .../conic/simple_examples/index.html          |  242 ++
 .../PR3525/tutorials/conic/start_values.jl    |   88 +
 .../tutorials/conic/start_values/index.html   |   83 +
 .../PR3525/tutorials/conic/tips_and_tricks.jl |  418 +++
 .../conic/tips_and_tricks/index.html          |  162 ++
 .../tutorials/getting_started/data/Cereal.txt |   24 +
 .../getting_started/data/SalesData.xlsx       |  Bin 0 -> 7463 bytes
 .../tutorials/getting_started/data/Soccer.txt |   21 +
 .../getting_started/data/StarWars.csv         |   21 +
 .../data/passport-index-matrix.csv            |  200 ++
 .../tutorials/getting_started/debugging.jl    |  355 +++
 .../getting_started/debugging/index.html      |   23 +
 .../design_patterns_for_larger_models.jl      |  688 +++++
 .../index.html                                |  517 ++++
 .../getting_started_with_JuMP.jl              |  509 ++++
 .../getting_started_with_JuMP/index.html      |  224 ++
 .../getting_started_with_data_and_plotting.jl |  389 +++
 .../0c9607c5.svg                              |   59 +
 .../2af4f5d4.svg                              |   73 +
 .../f8d9d8f6.svg                              |   62 +
 .../index.html                                |  238 ++
 .../getting_started_with_julia.jl             |  722 +++++
 .../getting_started_with_julia/index.html     |  183 ++
 .../getting_started_with_sets_and_indexing.jl |  293 ++
 .../index.html                                |  222 ++
 .../getting_started/introduction/index.html   |    6 +
 .../getting_started/performance_tips.jl       |  193 ++
 .../performance_tips/index.html               |   40 +
 previews/PR3525/tutorials/linear/callbacks.jl |  185 ++
 .../tutorials/linear/callbacks/index.html     |  152 +
 previews/PR3525/tutorials/linear/cannery.jl   |  130 +
 .../tutorials/linear/cannery/index.html       |   89 +
 .../PR3525/tutorials/linear/commodity_nz.db   |  Bin 0 -> 45056 bytes
 .../linear/constraint_programming.jl          |  190 ++
 .../linear/constraint_programming/index.html  |   92 +
 previews/PR3525/tutorials/linear/diet.jl      |  194 ++
 .../PR3525/tutorials/linear/diet/index.html   |  144 +
 .../tutorials/linear/facility_location.jl     |  300 ++
 .../linear/facility_location/49a24bc3.svg     |   69 +
 .../linear/facility_location/a3695eca.svg     |   69 +
 .../linear/facility_location/a9295f9b.svg     |   57 +
 .../linear/facility_location/fb5e71cf.svg     |   61 +
 .../linear/facility_location/index.html       |  170 ++
 .../tutorials/linear/factory_schedule.jl      |  327 +++
 .../linear/factory_schedule/34a51857.svg      |  160 ++
 .../linear/factory_schedule/4024f2dd.svg      |   74 +
 .../linear/factory_schedule/c5e9cd8e.svg      |  162 ++
 .../linear/factory_schedule/index.html        |  165 ++
 .../linear/factory_schedule_demand.txt        |   13 +
 .../linear/factory_schedule_factories.txt     |   25 +
 previews/PR3525/tutorials/linear/finance.jl   |  146 +
 .../tutorials/linear/finance/index.html       |   54 +
 .../tutorials/linear/geographic_clustering.jl |  176 ++
 .../linear/geographic_clustering/index.html   |  137 +
 .../tutorials/linear/introduction/index.html  |   10 +
 previews/PR3525/tutorials/linear/knapsack.jl  |  154 ++
 .../tutorials/linear/knapsack/index.html      |   86 +
 .../PR3525/tutorials/linear/lp_sensitivity.jl |  155 ++
 .../linear/lp_sensitivity/index.html          |   73 +
 .../PR3525/tutorials/linear/mip_duality.jl    |  124 +
 .../tutorials/linear/mip_duality/index.html   |  112 +
 previews/PR3525/tutorials/linear/multi.jl     |  193 ++
 previews/PR3525/tutorials/linear/multi.sqlite |  Bin 0 -> 45056 bytes
 .../PR3525/tutorials/linear/multi/index.html  |  171 ++
 .../linear/multi_commodity_network.jl         |  218 ++
 .../linear/multi_commodity_network/index.html |   96 +
 .../linear/multi_objective_examples.jl        |  120 +
 .../multi_objective_examples/index.html       |  142 +
 .../linear/multi_objective_knapsack.jl        |  185 ++
 .../multi_objective_knapsack/0a06bb1f.svg     |   55 +
 .../multi_objective_knapsack/144638a3.svg     |   53 +
 .../multi_objective_knapsack/index.html       |   94 +
 previews/PR3525/tutorials/linear/n-queens.jl  |   72 +
 .../tutorials/linear/n-queens/index.html      |   25 +
 .../PR3525/tutorials/linear/network_flows.jl  |  168 ++
 .../tutorials/linear/network_flows/index.html |   91 +
 previews/PR3525/tutorials/linear/sudoku.jl    |  212 ++
 .../PR3525/tutorials/linear/sudoku/index.html |   94 +
 .../tutorials/linear/tips_and_tricks.jl       |  330 +++
 .../linear/tips_and_tricks/index.html         |   67 +
 previews/PR3525/tutorials/linear/transp.jl    |  142 +
 previews/PR3525/tutorials/linear/transp.txt   |    5 +
 .../PR3525/tutorials/linear/transp/index.html |   76 +
 .../tutorials/nonlinear/complementarity.jl    |  208 ++
 .../nonlinear/complementarity/index.html      |  140 +
 .../nonlinear/introduction/index.html         |   10 +
 .../tutorials/nonlinear/nested_problems.jl    |  235 ++
 .../nonlinear/nested_problems/index.html      |  130 +
 .../PR3525/tutorials/nonlinear/portfolio.jl   |  255 ++
 .../nonlinear/portfolio/319d794d.svg          |  581 ++++
 .../tutorials/nonlinear/portfolio/index.html  |  136 +
 .../tutorials/nonlinear/querying_hessians.jl  |  257 ++
 .../nonlinear/querying_hessians/index.html    |  139 +
 .../tutorials/nonlinear/rocket_control.jl     |  155 ++
 .../nonlinear/rocket_control/2e6d6b3d.svg     |  130 +
 .../nonlinear/rocket_control/index.html       |   67 +
 .../tutorials/nonlinear/simple_examples.jl    |  168 ++
 .../nonlinear/simple_examples/index.html      |  183 ++
 .../space_shuttle_reentry_trajectory.jl       |  402 +++
 .../0d7a8519.svg                              |   48 +
 .../78651b37.svg                              |  182 ++
 .../fa7e5a76.svg                              |   97 +
 .../index.html                                |  270 ++
 .../tutorials/nonlinear/tips_and_tricks.jl    |  129 +
 .../nonlinear/tips_and_tricks/index.html      |   70 +
 .../nonlinear/user_defined_hessians.jl        |   75 +
 .../user_defined_hessians/index.html          |   42 +
 302 files changed, 44666 insertions(+)
 create mode 100644 previews/PR3525/.documenter-siteinfo.json
 create mode 100644 previews/PR3525/JuMP.pdf
 create mode 100644 previews/PR3525/api/JuMP.Containers/index.html
 create mode 100644 previews/PR3525/api/JuMP/index.html
 create mode 100644 previews/PR3525/assets/case9mod.png
 create mode 100644 previews/PR3525/assets/documenter.js
 create mode 100644 previews/PR3525/assets/extra_styles.css
 create mode 100644 previews/PR3525/assets/full_sudoku.png
 create mode 100644 previews/PR3525/assets/logo-dark-with-text-background.svg
 create mode 100644 previews/PR3525/assets/logo-dark-with-text.svg
 create mode 100644 previews/PR3525/assets/logo-dark-without-text.svg
 create mode 100644 previews/PR3525/assets/logo-dark.svg
 create mode 100644 previews/PR3525/assets/logo-with-text-background.svg
 create mode 100644 previews/PR3525/assets/logo-with-text.pdf
 create mode 100644 previews/PR3525/assets/logo-with-text.svg
 create mode 100644 previews/PR3525/assets/logo-without-text.svg
 create mode 100644 previews/PR3525/assets/logo.svg
 create mode 100644 previews/PR3525/assets/logo_constructor.jl
 create mode 100644 previews/PR3525/assets/n_queens4.png
 create mode 100644 previews/PR3525/assets/numfocus-logo.png
 create mode 100644 previews/PR3525/assets/partial_sudoku.png
 create mode 100644 previews/PR3525/assets/power_systems.png
 create mode 100644 previews/PR3525/assets/space_shuttle.png
 create mode 100644 previews/PR3525/assets/themes/documenter-dark.css
 create mode 100644 previews/PR3525/assets/themes/documenter-light.css
 create mode 100644 previews/PR3525/assets/themeswap.js
 create mode 100644 previews/PR3525/assets/warner.js
 create mode 100644 previews/PR3525/background/algebraic_modeling_languages/index.html
 create mode 100644 previews/PR3525/changelog/index.html
 create mode 100644 previews/PR3525/developers/checklists/index.html
 create mode 100644 previews/PR3525/developers/contributing/index.html
 create mode 100644 previews/PR3525/developers/custom_solver_binaries/index.html
 create mode 100644 previews/PR3525/developers/extensions/index.html
 create mode 100644 previews/PR3525/developers/roadmap/index.html
 create mode 100644 previews/PR3525/developers/style/index.html
 create mode 100644 previews/PR3525/extensions/DimensionalData/index.html
 create mode 100644 previews/PR3525/extensions/introduction/index.html
 create mode 100644 previews/PR3525/index.html
 create mode 100644 previews/PR3525/installation/index.html
 create mode 100644 previews/PR3525/manual/callbacks/index.html
 create mode 100644 previews/PR3525/manual/complex/index.html
 create mode 100644 previews/PR3525/manual/constraints/index.html
 create mode 100644 previews/PR3525/manual/containers/index.html
 create mode 100644 previews/PR3525/manual/expressions/index.html
 create mode 100644 previews/PR3525/manual/models/index.html
 create mode 100644 previews/PR3525/manual/nlp/index.html
 create mode 100644 previews/PR3525/manual/nonlinear/index.html
 create mode 100644 previews/PR3525/manual/objective/index.html
 create mode 100644 previews/PR3525/manual/solutions/index.html
 create mode 100644 previews/PR3525/manual/variables/index.html
 create mode 100755 previews/PR3525/moi/assets/latency.png
 create mode 100644 previews/PR3525/moi/background/duality/index.html
 create mode 100644 previews/PR3525/moi/background/infeasibility_certificates/index.html
 create mode 100644 previews/PR3525/moi/background/motivation/index.html
 create mode 100644 previews/PR3525/moi/background/naming_conventions/index.html
 create mode 100644 previews/PR3525/moi/changelog/index.html
 create mode 100644 previews/PR3525/moi/developer/checklists/index.html
 create mode 100644 previews/PR3525/moi/index.html
 create mode 100644 previews/PR3525/moi/manual/constraints/index.html
 create mode 100644 previews/PR3525/moi/manual/models/index.html
 create mode 100644 previews/PR3525/moi/manual/modification/index.html
 create mode 100644 previews/PR3525/moi/manual/solutions/index.html
 create mode 100644 previews/PR3525/moi/manual/standard_form/index.html
 create mode 100644 previews/PR3525/moi/manual/variables/index.html
 create mode 100644 previews/PR3525/moi/reference/callbacks/index.html
 create mode 100644 previews/PR3525/moi/reference/constraints/index.html
 create mode 100644 previews/PR3525/moi/reference/errors/index.html
 create mode 100644 previews/PR3525/moi/reference/models/index.html
 create mode 100644 previews/PR3525/moi/reference/modification/index.html
 create mode 100644 previews/PR3525/moi/reference/nonlinear/index.html
 create mode 100644 previews/PR3525/moi/reference/standard_form/index.html
 create mode 100644 previews/PR3525/moi/reference/variables/index.html
 create mode 100644 previews/PR3525/moi/release_notes/index.html
 create mode 100644 previews/PR3525/moi/submodules/Benchmarks/overview/index.html
 create mode 100644 previews/PR3525/moi/submodules/Benchmarks/reference/index.html
 create mode 100644 previews/PR3525/moi/submodules/Bridges/list_of_bridges/index.html
 create mode 100644 previews/PR3525/moi/submodules/Bridges/overview/index.html
 create mode 100644 previews/PR3525/moi/submodules/Bridges/reference/index.html
 create mode 100644 previews/PR3525/moi/submodules/FileFormats/overview/index.html
 create mode 100644 previews/PR3525/moi/submodules/FileFormats/reference/index.html
 create mode 100644 previews/PR3525/moi/submodules/Nonlinear/overview/index.html
 create mode 100644 previews/PR3525/moi/submodules/Nonlinear/reference/index.html
 create mode 100644 previews/PR3525/moi/submodules/Test/overview/index.html
 create mode 100644 previews/PR3525/moi/submodules/Test/reference/index.html
 create mode 100644 previews/PR3525/moi/submodules/Utilities/overview/index.html
 create mode 100644 previews/PR3525/moi/submodules/Utilities/reference/index.html
 create mode 100644 previews/PR3525/moi/tutorials/bridging_constraint/index.html
 create mode 100644 previews/PR3525/moi/tutorials/example/index.html
 create mode 100644 previews/PR3525/moi/tutorials/implementing/index.html
 create mode 100644 previews/PR3525/moi/tutorials/latency/index.html
 create mode 100644 previews/PR3525/moi/tutorials/manipulating_expressions/index.html
 create mode 100644 previews/PR3525/moi/tutorials/mathprogbase/index.html
 create mode 100644 previews/PR3525/packages/Alpine/index.html
 create mode 100644 previews/PR3525/packages/AmplNLWriter/index.html
 create mode 100644 previews/PR3525/packages/BARON/index.html
 create mode 100644 previews/PR3525/packages/BilevelJuMP/index.html
 create mode 100644 previews/PR3525/packages/CDCS/index.html
 create mode 100644 previews/PR3525/packages/CDDLib/index.html
 create mode 100644 previews/PR3525/packages/COPT/index.html
 create mode 100644 previews/PR3525/packages/COSMO/index.html
 create mode 100644 previews/PR3525/packages/CPLEX/index.html
 create mode 100644 previews/PR3525/packages/CSDP/index.html
 create mode 100644 previews/PR3525/packages/Cbc/index.html
 create mode 100644 previews/PR3525/packages/Clarabel/index.html
 create mode 100644 previews/PR3525/packages/Clp/index.html
 create mode 100644 previews/PR3525/packages/DAQP/index.html
 create mode 100644 previews/PR3525/packages/DiffOpt/index.html
 create mode 100644 previews/PR3525/packages/Dualization/index.html
 create mode 100644 previews/PR3525/packages/ECOS/index.html
 create mode 100644 previews/PR3525/packages/GAMS/index.html
 create mode 100644 previews/PR3525/packages/GLPK/index.html
 create mode 100644 previews/PR3525/packages/Gurobi/index.html
 create mode 100644 previews/PR3525/packages/HiGHS/index.html
 create mode 100644 previews/PR3525/packages/Hypatia/index.html
 create mode 100644 previews/PR3525/packages/InfiniteOpt/index.html
 create mode 100644 previews/PR3525/packages/Ipopt/index.html
 create mode 100644 previews/PR3525/packages/Juniper/index.html
 create mode 100644 previews/PR3525/packages/KNITRO/index.html
 create mode 100644 previews/PR3525/packages/Loraine/index.html
 create mode 100644 previews/PR3525/packages/MadNLP/index.html
 create mode 100644 previews/PR3525/packages/MiniZinc/index.html
 create mode 100644 previews/PR3525/packages/MosekTools/index.html
 create mode 100644 previews/PR3525/packages/MultiObjectiveAlgorithms/index.html
 create mode 100644 previews/PR3525/packages/NEOSServer/index.html
 create mode 100644 previews/PR3525/packages/NLopt/index.html
 create mode 100644 previews/PR3525/packages/OSQP/index.html
 create mode 100644 previews/PR3525/packages/PATHSolver/index.html
 create mode 100644 previews/PR3525/packages/Pajarito/index.html
 create mode 100644 previews/PR3525/packages/ParametricOptInterface/index.html
 create mode 100644 previews/PR3525/packages/Pavito/index.html
 create mode 100644 previews/PR3525/packages/Plasmo/index.html
 create mode 100644 previews/PR3525/packages/PolyJuMP/index.html
 create mode 100644 previews/PR3525/packages/ProxSDP/index.html
 create mode 100644 previews/PR3525/packages/SCIP/index.html
 create mode 100644 previews/PR3525/packages/SCS/index.html
 create mode 100644 previews/PR3525/packages/SDDP/index.html
 create mode 100644 previews/PR3525/packages/SDPA/index.html
 create mode 100644 previews/PR3525/packages/SDPNAL/index.html
 create mode 100644 previews/PR3525/packages/SDPT3/index.html
 create mode 100644 previews/PR3525/packages/SeDuMi/index.html
 create mode 100644 previews/PR3525/packages/SumOfSquares/index.html
 create mode 100644 previews/PR3525/packages/Tulip/index.html
 create mode 100644 previews/PR3525/packages/Xpress/index.html
 create mode 100644 previews/PR3525/packages/solvers/index.html
 create mode 100644 previews/PR3525/release_notes/index.html
 create mode 100644 previews/PR3525/search_index.js
 create mode 100644 previews/PR3525/should_i_use/index.html
 create mode 100644 previews/PR3525/siteinfo.js
 create mode 100644 previews/PR3525/tutorials/algorithms/benders_decomposition.jl
 create mode 100644 previews/PR3525/tutorials/algorithms/benders_decomposition/index.html
 create mode 100644 previews/PR3525/tutorials/algorithms/cutting_stock_column_generation.jl
 create mode 100644 previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/703571d3.svg
 create mode 100644 previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/fa9bdba4.svg
 create mode 100644 previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/index.html
 create mode 100644 previews/PR3525/tutorials/algorithms/parallelism/index.html
 create mode 100644 previews/PR3525/tutorials/algorithms/tsp_lazy_constraints.jl
 create mode 100644 previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/43baf431.svg
 create mode 100644 previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/8946bc6a.svg
 create mode 100644 previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/index.html
 create mode 100644 previews/PR3525/tutorials/applications/optimal_power_flow.jl
 create mode 100644 previews/PR3525/tutorials/applications/optimal_power_flow/index.html
 create mode 100644 previews/PR3525/tutorials/applications/power_systems.jl
 create mode 100644 previews/PR3525/tutorials/applications/power_systems/6fee809a.svg
 create mode 100644 previews/PR3525/tutorials/applications/power_systems/9f136f96.svg
 create mode 100644 previews/PR3525/tutorials/applications/power_systems/ab8a43ee.svg
 create mode 100644 previews/PR3525/tutorials/applications/power_systems/index.html
 create mode 100644 previews/PR3525/tutorials/applications/two_stage_stochastic.jl
 create mode 100644 previews/PR3525/tutorials/applications/two_stage_stochastic/01f8597b.svg
 create mode 100644 previews/PR3525/tutorials/applications/two_stage_stochastic/27244c58.svg
 create mode 100644 previews/PR3525/tutorials/applications/two_stage_stochastic/a5d2410e.svg
 create mode 100644 previews/PR3525/tutorials/applications/two_stage_stochastic/ddf1a4d6.svg
 create mode 100644 previews/PR3525/tutorials/applications/two_stage_stochastic/index.html
 create mode 100644 previews/PR3525/tutorials/applications/web_app.jl
 create mode 100644 previews/PR3525/tutorials/applications/web_app/index.html
 create mode 100644 previews/PR3525/tutorials/conic/arbitrary_precision.jl
 create mode 100644 previews/PR3525/tutorials/conic/arbitrary_precision/index.html
 create mode 100644 previews/PR3525/tutorials/conic/dualization.jl
 create mode 100644 previews/PR3525/tutorials/conic/dualization/index.html
 create mode 100644 previews/PR3525/tutorials/conic/ellipse_approx.jl
 create mode 100644 previews/PR3525/tutorials/conic/ellipse_approx/07c030f6.svg
 create mode 100644 previews/PR3525/tutorials/conic/ellipse_approx/e71dc9be.svg
 create mode 100644 previews/PR3525/tutorials/conic/ellipse_approx/index.html
 create mode 100644 previews/PR3525/tutorials/conic/experiment_design.jl
 create mode 100644 previews/PR3525/tutorials/conic/experiment_design/index.html
 create mode 100644 previews/PR3525/tutorials/conic/introduction/index.html
 create mode 100644 previews/PR3525/tutorials/conic/logistic_regression.jl
 create mode 100644 previews/PR3525/tutorials/conic/logistic_regression/index.html
 create mode 100644 previews/PR3525/tutorials/conic/min_ellipse.jl
 create mode 100644 previews/PR3525/tutorials/conic/min_ellipse/4146ff43.svg
 create mode 100644 previews/PR3525/tutorials/conic/min_ellipse/804a7acd.svg
 create mode 100644 previews/PR3525/tutorials/conic/min_ellipse/index.html
 create mode 100644 previews/PR3525/tutorials/conic/quantum_discrimination.jl
 create mode 100644 previews/PR3525/tutorials/conic/quantum_discrimination/index.html
 create mode 100644 previews/PR3525/tutorials/conic/simple_examples.jl
 create mode 100644 previews/PR3525/tutorials/conic/simple_examples/ee174d1a.svg
 create mode 100644 previews/PR3525/tutorials/conic/simple_examples/index.html
 create mode 100644 previews/PR3525/tutorials/conic/start_values.jl
 create mode 100644 previews/PR3525/tutorials/conic/start_values/index.html
 create mode 100644 previews/PR3525/tutorials/conic/tips_and_tricks.jl
 create mode 100644 previews/PR3525/tutorials/conic/tips_and_tricks/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/data/Cereal.txt
 create mode 100644 previews/PR3525/tutorials/getting_started/data/SalesData.xlsx
 create mode 100644 previews/PR3525/tutorials/getting_started/data/Soccer.txt
 create mode 100644 previews/PR3525/tutorials/getting_started/data/StarWars.csv
 create mode 100644 previews/PR3525/tutorials/getting_started/data/passport-index-matrix.csv
 create mode 100644 previews/PR3525/tutorials/getting_started/debugging.jl
 create mode 100644 previews/PR3525/tutorials/getting_started/debugging/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models.jl
 create mode 100644 previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_JuMP.jl
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_JuMP/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting.jl
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/0c9607c5.svg
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/2af4f5d4.svg
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/f8d9d8f6.svg
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_julia.jl
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_julia/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing.jl
 create mode 100644 previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/introduction/index.html
 create mode 100644 previews/PR3525/tutorials/getting_started/performance_tips.jl
 create mode 100644 previews/PR3525/tutorials/getting_started/performance_tips/index.html
 create mode 100644 previews/PR3525/tutorials/linear/callbacks.jl
 create mode 100644 previews/PR3525/tutorials/linear/callbacks/index.html
 create mode 100644 previews/PR3525/tutorials/linear/cannery.jl
 create mode 100644 previews/PR3525/tutorials/linear/cannery/index.html
 create mode 100644 previews/PR3525/tutorials/linear/commodity_nz.db
 create mode 100644 previews/PR3525/tutorials/linear/constraint_programming.jl
 create mode 100644 previews/PR3525/tutorials/linear/constraint_programming/index.html
 create mode 100644 previews/PR3525/tutorials/linear/diet.jl
 create mode 100644 previews/PR3525/tutorials/linear/diet/index.html
 create mode 100644 previews/PR3525/tutorials/linear/facility_location.jl
 create mode 100644 previews/PR3525/tutorials/linear/facility_location/49a24bc3.svg
 create mode 100644 previews/PR3525/tutorials/linear/facility_location/a3695eca.svg
 create mode 100644 previews/PR3525/tutorials/linear/facility_location/a9295f9b.svg
 create mode 100644 previews/PR3525/tutorials/linear/facility_location/fb5e71cf.svg
 create mode 100644 previews/PR3525/tutorials/linear/facility_location/index.html
 create mode 100644 previews/PR3525/tutorials/linear/factory_schedule.jl
 create mode 100644 previews/PR3525/tutorials/linear/factory_schedule/34a51857.svg
 create mode 100644 previews/PR3525/tutorials/linear/factory_schedule/4024f2dd.svg
 create mode 100644 previews/PR3525/tutorials/linear/factory_schedule/c5e9cd8e.svg
 create mode 100644 previews/PR3525/tutorials/linear/factory_schedule/index.html
 create mode 100644 previews/PR3525/tutorials/linear/factory_schedule_demand.txt
 create mode 100644 previews/PR3525/tutorials/linear/factory_schedule_factories.txt
 create mode 100644 previews/PR3525/tutorials/linear/finance.jl
 create mode 100644 previews/PR3525/tutorials/linear/finance/index.html
 create mode 100644 previews/PR3525/tutorials/linear/geographic_clustering.jl
 create mode 100644 previews/PR3525/tutorials/linear/geographic_clustering/index.html
 create mode 100644 previews/PR3525/tutorials/linear/introduction/index.html
 create mode 100644 previews/PR3525/tutorials/linear/knapsack.jl
 create mode 100644 previews/PR3525/tutorials/linear/knapsack/index.html
 create mode 100644 previews/PR3525/tutorials/linear/lp_sensitivity.jl
 create mode 100644 previews/PR3525/tutorials/linear/lp_sensitivity/index.html
 create mode 100644 previews/PR3525/tutorials/linear/mip_duality.jl
 create mode 100644 previews/PR3525/tutorials/linear/mip_duality/index.html
 create mode 100644 previews/PR3525/tutorials/linear/multi.jl
 create mode 100644 previews/PR3525/tutorials/linear/multi.sqlite
 create mode 100644 previews/PR3525/tutorials/linear/multi/index.html
 create mode 100644 previews/PR3525/tutorials/linear/multi_commodity_network.jl
 create mode 100644 previews/PR3525/tutorials/linear/multi_commodity_network/index.html
 create mode 100644 previews/PR3525/tutorials/linear/multi_objective_examples.jl
 create mode 100644 previews/PR3525/tutorials/linear/multi_objective_examples/index.html
 create mode 100644 previews/PR3525/tutorials/linear/multi_objective_knapsack.jl
 create mode 100644 previews/PR3525/tutorials/linear/multi_objective_knapsack/0a06bb1f.svg
 create mode 100644 previews/PR3525/tutorials/linear/multi_objective_knapsack/144638a3.svg
 create mode 100644 previews/PR3525/tutorials/linear/multi_objective_knapsack/index.html
 create mode 100644 previews/PR3525/tutorials/linear/n-queens.jl
 create mode 100644 previews/PR3525/tutorials/linear/n-queens/index.html
 create mode 100644 previews/PR3525/tutorials/linear/network_flows.jl
 create mode 100644 previews/PR3525/tutorials/linear/network_flows/index.html
 create mode 100644 previews/PR3525/tutorials/linear/sudoku.jl
 create mode 100644 previews/PR3525/tutorials/linear/sudoku/index.html
 create mode 100644 previews/PR3525/tutorials/linear/tips_and_tricks.jl
 create mode 100644 previews/PR3525/tutorials/linear/tips_and_tricks/index.html
 create mode 100644 previews/PR3525/tutorials/linear/transp.jl
 create mode 100644 previews/PR3525/tutorials/linear/transp.txt
 create mode 100644 previews/PR3525/tutorials/linear/transp/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/complementarity.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/complementarity/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/introduction/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/nested_problems.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/nested_problems/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/portfolio.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/portfolio/319d794d.svg
 create mode 100644 previews/PR3525/tutorials/nonlinear/portfolio/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/querying_hessians.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/querying_hessians/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/rocket_control.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/rocket_control/2e6d6b3d.svg
 create mode 100644 previews/PR3525/tutorials/nonlinear/rocket_control/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/simple_examples.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/simple_examples/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/0d7a8519.svg
 create mode 100644 previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/78651b37.svg
 create mode 100644 previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/fa7e5a76.svg
 create mode 100644 previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/tips_and_tricks.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/tips_and_tricks/index.html
 create mode 100644 previews/PR3525/tutorials/nonlinear/user_defined_hessians.jl
 create mode 100644 previews/PR3525/tutorials/nonlinear/user_defined_hessians/index.html

diff --git a/previews/PR3525/.documenter-siteinfo.json b/previews/PR3525/.documenter-siteinfo.json
new file mode 100644
index 00000000000..b3d6e02484b
--- /dev/null
+++ b/previews/PR3525/.documenter-siteinfo.json
@@ -0,0 +1 @@
+{"documenter":{"julia_version":"1.9.3","generation_timestamp":"2023-09-23T04:49:51","documenter_version":"1.0.0"}}
\ No newline at end of file
diff --git a/previews/PR3525/JuMP.pdf b/previews/PR3525/JuMP.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..044e761ca8b16b0c693cfaa9221d435947f523ca
GIT binary patch
literal 4687304
zcmb4q1#BM8vZa|ZW@g6EpY52L*^W78W@ctPW@e6=nb|QjGcz;OW}n`>yW0C!`|otC
zHJYiGdZwjws!ml?D2RwL02tZeD5mEJmf_flnTc%;EZ}%}n3Ub^jG5#OEPgmyGszJH
zn3<Wx%pDw^h&fm|nPiM@Or6Y#0jw<FiT~sJ`_9VA*nvsRO5e#?#Msc*$e520&e6%i
zSl=4XZFNCo&0&ol#d}F@8*n0H;Md|nnIKneKI@XsUO9IXp=<Q3cHx-42!F|Ut3K0$
z1_ok)wr-9OrUSIybV`F>U=)TXfVzV^m?@YxDE`?zuLE`TmylHEuDkz_JbM-Dyg<xg
zRIyS;{on^T!k`3I^g@_f80|2>oW=Ygz=cRD0FzOO1mRqg0)Q=k7VM9Qb3d%kF{F-W
z_eT$$nTuNAe{fhyU7*hC0z}%w4U)0PIS5&d;2cuGsRm01n{$~Co&o<nNflO&6a?EE
z4AdY71)TuZ5cC*52&&2trV$CO3eK<r21>kI40J}r35j?N)YOj`*^c7e-4>Np0_5`B
zV*p})P-lDq$*2^9$ozJ7D|kqOE7>39G67zgo?nm#5M_v5v?4`_$I7un+<`kN!rb2=
z9gY=f^QDiK;>>E+TZk14SjiwYBEiTZ^<@+lj+C$w@xohz%`?qnt55#y1i1x?44#JA
zWlj^681W#6$_!>W6yzqrLZzbqf`!rYSN;)z3CUhTjx{~{tr^ZhgH%nBp<I|Ot42)G
zA}(=TLy-9lf7vck3Ufz*<qSMT82B^(8f8N1mb&CHF+@=0%#6C#o)|Y@*b}iABhfJg
zJJ4y;jazAYvh@&kc3x3TkawxY&x;YoB=$gRT@__oT2v!&Kek8uN00O@UY}cj(Cs`O
zlQ@%3W>!w8MyGC8=QsINi@LPRWL;i0dEG92-qz{Oog=1##4V-zd=trukaXIvikxTz
z$FrM>qW*d0n;N_(p2EZAbXfkZMxUn%`KqBaS?4=H!q1((*3;`7Xtt&K$d&nE)_B>)
zc2^&xA=^@S@rfmOSJB<e@01-T9nJXb<UVZ=um0_u=bK%-n;EIW6#N@5m971$6Ai2K
zcc{nY0arNV8&#o*cj_0L*$!8*p>4+!ZT2bAADZ*_2Iqys(x-VSvQOk`1iMcBShs{$
zyE{jzkxTRMeO122hdj1Z6puaznF*=uAqBMG4`$zaOmWucGwWm5rmyyp*4COUhkjGZ
z&TxyDlQW$%A-A~k^hkGga;o63X-)GUxK?|Ow{mh=Z!rhq45=TK%y(J}V2a>gW)ei-
zb%oD8pXR*3a_BVM3`!nWv`kSgzENInI89wcy`*QTEafwgJ0C^oJ>(>J+3tE-(Jlck
z7&*6J{}9i5yikfce8LcWJ&|4$q<jv6pn9+e!P~HuUzsKcxIk8OLYBXW0JE~M+?Ofk
zRy7ccCzCklvY?UMf>~l}s@VvPvKZ?&s$6d-k6({(ExAs2Q-aoY2POiGCKVc-&uir{
z!;KAWFt===z8XVWi5fnTA$X%K$J_*_Ehl?bFwOanood|%N|-oxt6Ff3=AW=DS4LHo
zfJ^rG87<RHp<OhShB#FQz%`wS79^FIfK{kuM(QhrJ<l-ATAKu;sa3ef%gTeMBLcG#
z)gj5)vzA?*n}gIH<Nn4Ng)wWJ9qFw~{<<^!F)UrV7HP_R>BZ*V>1ORUp_$oPxpG=$
z9b`y;W4^L|=R=U`4p-p(8@XJ+ac2D!6{ulozzd3HQ*&#peE2CjJ1}Z#<CL7MEpHpE
z_KAHW2QXi?alp#>@&rqCcR3{^UcM6w)ez};%c#1y_XT*ICU~Q-!i82{knPy88IzW8
zrH{!3t9Qdnn`MpRrz7-w;z;MBNI=mUIvd|KXnJCFu(O^x&xV>m?J|~nIA3pD>|`5b
ziIB{qQgI~owOd__M_5J#)tUKV<gr@x9z^F9S=(N)M4MZ(`P0T%%%kU01i!hqnLfYz
zm2h!oEzd{LICb1>*3K_D(S}*Dv+z(NfkYn~I@|x?C1d$@Y<Aps0Z4fHHXrn0`EG7>
zB<vH@Yj<qh)kbD&dNcT-dme)|BJzc|<^2YLGqy4MPjvo!_dgMzg_->yU_S@&U)aya
z3HV3Y->I=?gZKS^V*fS3J*;1Q)1O!%1H#(BNcOD3dnCeYtVwf&*p~5f>B$Z}j?hs;
zGFp*%cXs_?d?zz)C-$TeRq383jPQ$#rei?(R~69BBF`zN7;V(q%|jm}+ncJw-tDnZ
zqvI;dPEhW3)4=1M8-|EjMH1D^ODH>$kN~mwxm*p20A6u4qH1Znl-ns&tQafHKDH|A
zxPmp|xvd^_&KV^uFn&{F=7nQSpf08!!AD8Z^$!I@BCgCay(;R06X+EgD@F!KSx3%2
zJFQEPp)h(lvUVnp_~48^1W-E>7r<$2K<2^E?nXdv^sGajD?XHK=!avXJj2(Cd<??G
z7X;An<)d~Yb<<NC{DX9E$#bt4H=6%D7aKEF`BzSg8*+A{noEAJJS2$PwoEVzhEyK6
zx{&|(4NinW4g!X<p3LZo9|{!<VTEX#;zFQu8MTJfGbZ_Z8W#cyV7@hQ%6<YD6y$-x
z3qVCrE)psM_Pjp+P2X@B<ZI*~J13kuf8Jms){<X+7P835N_x}xphW$oVuA3G>tf>d
z{(^xq)Aq-yW_S|ppv*nFwSG7wK%{bIy=jgOd2|GQugqXkTx3I9W;0_}JnRM~QDF=H
z9iw9or~EvD>?oVUo?OWG<9j=6Eb+)1T%F>cTsC!^hIBT}`kvgzI67vt-$Z|>ek}X%
zD{^BufU?im#39nUppZM@3noe!Q!?rB@Oyvuzka;FSl;ZLxbTfHm7_V%1{X8w%4K_j
zz`d_S(;Sp9%NW*6OlaV8$4v0owz_kbN?sU~Jg~CYFINnX)TRxYNOa0OowKtO6(mbO
zg_E`y2H0HsbVpAZljuJUkxpn*Xzf9-wd-2X4pU#!esKD<SygoJRdR)EP|gsYysAl#
zAoV<7a*?bf>K2TwU=LOJq$&GI)`_Q0Q74mhc1m^H0;XRcoPgs^R82LLOp=4AGsoty
zCA>5=ufG_1q$eNwiX**6wU%b?^qGroYN%ndDQt=w<?-l#1TVV6p0hJ-vqUi*dxW7`
z9|zPQy=P1l*dS?xyz$D8xf>@3UB-6gZ`#LXiE8K3pnq!qRE{plTD)soY|qhAdaDz3
zO=msg;iBVfDRS4}{xHlCGePKDBMsg!dFGZKt;+P@T9nloJbyFlk&W|nXzSYU9Y<c5
zKL16B{uMWW5P463l1?FXNr4iuKZGyggf%jZm&%jVQ`h0Y^g11DVblpcC3&mUam$Q4
zSsPzUI_O<nO;#2oOF>MLp_eJGPhuMorf*&A)9R0=JKc_{)o%ULZ+94Xd5B)EiKn5P
znM|T-cg>*@9z*Z{j^Nu*xK=VWEPE9*DI2{qjkLC9Taq%7F=8o)Hi~-LU?p<hoQNI0
zqOJ=x;;&J#8#lSr*4)Ht97%4IqUgDc{$Bi2HeRi@+ET<x$+X|m6~ZF?E;Ew;!<7kb
zFgbZ^O>=U?bi$$8BVl-J+S*A+x`@2URQYRx@FOqoLha3AERs;<YT%xa-`FMnyR5$@
zZp<GVc9~^w{cEDjBki;W-{1C$IM1nxvGgnojR%~7_7zUyyUgtJC&4Xi-)!V*QHM}x
zV0c-iyGhPlcH180!!q6`cdLWGcDtL|+<uz(4F)f?k&OgxYyPk)t1dxRcU`x!VvbpX
zveG9_R1N+B-e7^Tf!7ja%IP(muByxlE~fHkz^lyzBgu7kOWvD?T-CyB+i8OT&4^d-
z=lE7+<}7^KP7fF_bC=7XIOlrgMdS3_hjK2Yg_hp6*|oOWxw^-CF-&W9!8^h)PLd;w
z0lKb9++%`%=hv>rC%yvpEtzI;!{%nzw5`%O3qEabqDu4VpR7^KBMGAn`^lT7NpE8A
z8g9Cd!sZ8)bW~&V9hMHJd*JzKxES5zmKIOtdsiB2tj#$$J$Jnh_1)y4TAf#kDyrHT
zUPDTT&n2&^^VQGN`(p+ix+v?dOlD?HQC2L<+O~R9v$s(EO8^E4W6n>kANlw=P>{>c
zaaKp}KP_|b-Y3kEW+Otu+Y3sqUIfDOK<(ocyj$ryU-@Bw9%gc|G7%fHCwQ527+1wO
z7A}v+5co!B0gndrjPEQ+nK0f*G6BUhypZc(pyF?fT>s8btjyp42|uy@i=TiTtpA9g
z4%B6B|MFAkRrNLi6gNXx*ukH;1P@CTxpHmrcmzv>C<<=Pft;e`?(9u0ctl~Wkgtvl
zf;W|0hryI^3lGV3Ns6kVMhb3YH$R94SB)<5M@#%)e-<|=F37UI)s!u}7E_HZ?8uxd
zOvtbit83s&EIjw7-JgFX8UcS4(;or05RCs@&}?{)Icfr8xxPAhKg<A0r86`AyCxGl
z>U(Su>aw5dQXS}&PCe)pD3HF_QEwh%%Gz(IVOptHKS&G#8!4nd%2~$@ym~PZ6m3~S
z2#9&cfDpRiULycC=k2!xve_)e%~21Ut^Id%c%TpTdw_bZ>^+Wn%micX@hRXf>eMEU
zm?1;-cE+sowt_V~Dh-v)m<o~>*vWd>`uMvji7Yy$Aq7~Je(V7jlr13O**{@~w`Ur7
zB2?mu{W+Lf{d&1bsB7KC>Uy>A=!c;>OG%DRC8Y&XP=e|}WhU$O)5zGjIrSio+$7Tx
z1>$mpdT^O*d)@s-NRL6U+DNFwa!@0(>Qw2eIQO9h1r2C^T{Q`WDhCK!HT2PtV+#rv
zhar+>`eUZT^|yxP<{B=-QWF526{a~|kRh|+tzzwZYf$xSJbQv|U|>wN$53)}Q}DsM
zIuP<msE-?s=)Z;Z!uieO8T9vU=PFFEB}WnS0HqYB(OA~~4X%j80s}UGc7adTq%T!Z
zZDoBmwCH5&hGUhVxbf=pWYRbn>2$fzb+iCBcc!)^Xtr)^_Vq{m!$O_kpTsnAZ27<L
zXf{{IGlqO^-CNdVd{*uarU?*y|G4>zhL+<?2yt+tj}HEXj5*nE-QMvJ6>Z3NYl(T{
z6N?lUswJTIy*zoq84&i`_Nfk=|MkrHEr*m(#JbGh?fG)7Se4VUIj*by*u$%3NzsdS
zEO-t7vcRuvFo@gE(kPbl#wEVsXom2frt#agM#0rI`51>JY)0Q=>NNqS&v2^@NOs63
zXdc?z5w=TK$}eTC5Qu42b`gh8#?a0PnXAa(MpV<rY`^3b!;@Z51AvobM!PhH(%;IY
z!We=4QSFA=R}C73#yUloVVMd=2n;GsZ(23RtsRH~g`Q$sch_C@w^zt;E2W~pVSIs(
ziv1$u#WQ8`(&d>`N~*>btP|t-$+XWryt+Tto@AaGbeBqqeJD1DuEb-`PU;fRY>_qI
z+OjFyrjN8m+9*30_K822lCIk_k>{|D2*?WWpS4t-d8h#<SU22o>Qq}M#r+o0R*9ja
zV*L%{{Q(Z(=cRkAX4+<HpE=eIjI4HzdnaKC9G|(f!H1VDCWq8iPH5+)ss`<d>r9>x
z*KbMdikyIE_E&GjTwJj2;@@w*^92*PtXV_4u7u>lfwy6d?lq2;)u{N-7Xm459eWWg
z0t{`85AIli9u6k9a;b&4QEh%{PhL#_{L7zL^IsIQ*4|YehDuv4G=^$CJcFKvSo6kz
zI7eGKCuWMh4mN1Uf*D5m=f`wkP*z9teJy%~AeS+r@uGANqIL@2^SjyZ633ekD6Qfg
zYZzU%%WW+xH7_5aKFhY79#x8pr_>)>8I#J<FZU$*djlX;D0cuU4%t^}rVwi07A}s4
zmVuXU^)1ouSG5U{Wli`W6Aa=@$%eFB&{Xaj0NM_-Y_q-P1g|z+Y;qON@HgrKZ6q&E
z>gTah8UV=Qb6i3BLrlgUrGfTiBfc)kvGJdKZ}Kz@^iiiYr11Q9LEkCl14+59w)e^J
z^d*cB-G>QRaPN^^YgXzCO?6L4oNPZg#*d_c=?YZQIj6r?ARRE|)4Jh_+755*R_))H
zfsM9gDQs<TrM1ggMvysta9g2<EUzFfUtlWJWxD?k=<LiK|9^na2K;9M$AX5m<6p6<
z>r8Ds72hDxje$8ql4`DDf<$(XGE2uR&X`tpPoj)4I!QOjb_Ek1A~u;|MGh%Hne_Rh
zS)QJTN>XB{Osd3>iGsH7xk*Se(3YxU?~d51r?pRMVoQ@FG=`;5dc9b}?E+fP9n~tB
zEeR%j39PyfDIj>;zaXom;ZPS00=id1QG?z|Ot?k>D5fk2I*0;62R-QqlN|)^CaQNp
z4@4vci|yQ%7C^!n2F-^H!I4t$Z4gvZK{O>xY$85{0h{UJWJ=UefFWeILZivEBI|)=
z_rvDQ3*-FZ*YoLTJqo_*PV9+1NjU892cE1c#cQ_-kC+h*j{%Xo>sOcrtyd$ohYRt%
zL`@9MvD||m|BY~nA|4maOE@gh#h(}?4INPeY^C2Y$a$3;Gm-h594W-41d1sUPN6-w
zkf~k>@t?_Kv1yc^XcS&yJ9Yt*SVELtQywiLnB1ERbPJgpdJs-Gs%2DFD9nJGa}_Y@
zaul3^xyk@r?0H-}l2q#`05}-u*ommRX;!a+$h3!BwqDbN9S2DOd(X-rNn#2cQ!F(r
ztf4?)0=_9xxd22~vc5p>8SWCQDZ;KmSnggVnVfMnCn$PepZ1?T11jdAeYP-SNu7RI
z)oi9-O_f&A+89<kL@T^q0e?h+PZUxKOuawR#A0F)zkYd+P>l*rcSO@c5G8^_BuYZs
z@_2i^KR&0dw6EO2-<WB{s3xr3m#~Px-qf20ic_WI^Er_gdX0sqBh&F^N|3#Hws^A~
z^wUdWd8PiD<EqlN+M2K6Kj`1GU!XoUx9PTG<RNsLWy9IndZ~%_iyEA0T7~-HVQc|p
zTx1L?t~Atq>W1FeG~>MG<Y((U?Q>;Uv4A~U%OrztZ~|gn8`Ek|?L$156&UxJ<{leD
z8I7HMquQqOv+W1nQ(taSoVTxXCLU#fyO|O^MDD#RCZ2Y(kgU(9hRTFWZI;>VU}8@g
z)bXF3LODAPs8P)vdvWM4MEs_;72_{*?hY4m(e&(hUOB<{R5qTT+VhSGYdaBC!0C%n
z@*}-hDQNA8YKzvk%eFBJT}ZPv<}Y4hmeT(6iJ2$bAP<0Dq+I>^WH;$}XgvRHj^|ix
zHB@@YW$dQ#2q2XFU>L39iotC@?XEY_;*>TquS{;5zBI58Hdw)9v(%fIjk%dl?>Zrg
zU0T(h`JUmRF!eY*U~Dm9%;`1OfXEJb&kMEEDYYd%Bxifi51ptaH6E2U_jbj-^LQh>
zc&umLw${%|YL1dOk9a|exXZ8HBjj_3e8>FVBqcm8kwX@-gzAgYoW5v^W!&Vvs$}YE
z#aG2#!W(mmTgYIrGN-?0?wUloY>qqe%Ei}Zs*qYl@A7ylxw>_WQ{}u(38`t~_(@;h
zp(cBCkAl2}2(x#;Qcb|`v6e+I?|BBZ=%Jr|I7mWQUN*XzRs8UfM<qOXQh&)Z5i5C`
z$rD*0cC+9zKoDNFChgL8oiN|gWIo>BCfQ|~BK5-hx#&^LjXTfAtW$*6Bg(_OK?h)b
z%2F?_f$oY}sX_ZJXku5~<HBPMIiNZ=Up#bCm7h~hJMXGLPjOz&x@Fbgbcc#WiUob?
z`lKcGnzS%g#tu?FIp&pYeLR_bdf=anvx_^?xq@5=FfIkRFw7Yc5i~J9XRChJPMcTP
zrFdT$xvv6;E6EmlHk{gszCIM4@f^OpDR;?;??|d)X*}8~Z@~<GE#(w>7c<Q8k=+yE
z&Ai{r>yuwDgCc7TbSVrl&B=VIr`F#Reo$VfGu3I^B=dzn*4Da?9)e^?bdnI=>1-Wd
zd`{A}x7i)N$@vCX6>r<#`GzSRLr?uaF%2#BtXF@syQKaSE#v;0+BuUoYaiiRXt_go
znINLhQ`oD%%9dxpm#t&rm3DNtG`au`Ub4CHNZt%Vg5P2F-=Q0-yDhD+pd@y5htza_
zimX1iHbkea;oNq2_N!|1sob_(kmbH@G@{ylZmz+|Fe97vP4dI_-%n_l+t2iGCOdll
zaI^POq0H=Pc?DCUf5ZBBj^SYD{0AJv$^5UNf$gsf_D>QT6Y67sMJ@D=DK$GBrn2h~
zkN|Y+xgwj4McG-Gy(s#zI@C3?<G6X>En!HKfgQsFY^XMr&~%b1@653=X-x}Qie&pf
z0>S`hiktZ$4gX>ip&j005%mZvvUFM=iW#uI7E&>L9tJ?ND<(FQVdeH8$#JZn-|Ma*
z5=mbMFov-v6r$oRPS{*fh6Uu2KfWaM3`tZq`yjxpoPs1}8c_xdLX04dCIIT78g$##
z#6M$x`=9efn*{xUY6STIV3MZm!&y33cS=02a0zlu)koy%>AdjUjt3hubAoVUha!L$
zkcL&e_HPtW-4w(ePS=lwtjp^HH|p0m)dy+Dtjm02;ISl#&B{4oMV;sF?d?{G;}}QB
z*Kv9#Yf(IwAceq%_Jtqu8p<crI4S|Bo1O+0qQ0`s4Wx<$wHFFxJ4yrI0}1+iK9D@%
z;Sr#3gN}N8t#(`t{mz3f@ShXV+c^BBMgdR=QgCzx#N>hap=nCbCwdBSh(>Ozv?nwv
z$JnXhP=&qx&`^bdd`CS6u9hW#!IP}*o?HXHA%DREa**C!gH1T*5{`RkJq7!y-oFFK
ztiK=y&O10%QexfTDU#1C{(^<?RLmvlpPbAk4X#Ys%D6!Mzk`meo&ruavel2lv%kj9
zJ_xZp8I%5k2lP<F%zDIN2o|Rgpn>Tg-hPFx$QD=Z6OA#B67yg-0#ns!dilj+$tFwn
zB6JpQSWMa7o}X>LTl~JF*_B826SBViJ5KrXI$h`4S@6GafJZ$aW8W>e#=YxrC%$*J
zB3xB@Iqqe=UE4f8WJ;ap@V}VgL5+LW!zR35>wC|tumRIdxaZB`mZ1rj-r!^c)71p~
z{V};AG2apvL9f}5b!T74o}Vvls0#EWCn8^fVPQ5tX4I}}NE=bF*fJXhSBGiBRej@p
z)R`P-V>!A3I<{VpO<xy`<1wr2>=iC)3$}qB@^3xd1-{<(2p(d1i-<`1&B<}~AySbh
zTpzf^ues`^hG}h|v{CKtdP<QG?nMTy<PO=QUk+`nGDbz)uKr<j-s{aL9+}12Yd=h=
zO<Z`jZ4S#}P18!$dYd#9+!CO)5hh+wjT7)au+^hRyqxN6DZ5@s5F=!X4P&!aokPoI
zR-erCUr^|y;i)N`$#1ekmKLaSOcH6iJMdwe1v9eru>?g}J=xpL@O~Y#M;(dQxTM2X
z)lb2nf9++Q&BMNcNwUm(KOTM=yf!1wxx25G6>M7T*UL7gbkzzDD0kCBH8C<>xpO_S
z`!SArAUF!K>4{QSH*gtJwo;$h<q+sTR8t2nPMM!pwcENxvfr^Us<Xq(jm*8o=-00m
ztax}U2b05Bx#2raYb5<zFZwA&IDthIH8<Z&rXkE5vf9f<tIA$Kv~sQBtkyNT2?>g<
z|BE8L6RB<e&%Ls96l05t@}GS2{7fUJv0xK~bkSByTy)6#YuB_g<rvl<RVZ({TgGkf
zjOBbRXpd|smyFs}BV7$#a!GoL3`{v^_vE)E8&L)_iz#6Y%g-Fg8+>sI*OVJ2uGhX0
zIo7jMra!1-ilLQrUQV_fVl(z0J1Yq(%8?u)LwWsCbX@dwInoGJz{1~aZuvF}U!8CU
zX-jcM0v6*(a%)QFCFUG3kA*KI$GVhGl>Ks84j&_r4*Xi^3M!RsIR4N*uvB_Ee0Ug&
zaPp~|d&KB4ZI~$2TWDh^mT*ZAG0i&8dyId3xo<!i?e1N1(cC64g;hOKZNsjYV5>Xz
zZu?;~*DpSnzh^6@nMPc#<s;pQtwo`1N!&YF5p`5@8WpEE!DXsY{b%%?C1S0&^{vNg
zVY)nN2ikPZ8JtqL+7}D%z+A8uvkafxl>p_G!Hca)*D9uK3(|tn(=_yoGBDvo)`O0h
z+o$*-rURl3chu7PgngsN(8tG2??#Lh%@(xSoH0y1J+FYCcTn35c)1kUvmb2t#Y?=-
z7QasX!?%Ls9elD*i08%Gf?1tM@RjT26C4Fk(d6H$1i<|Le-ohoGXc)R`kx^Q@&6+M
z9<11I)6a~s{(<rCPwgcQ28Jwo29ni@p}ki6)t%gAMIi?7lxsv3Gej1A4GHrE`qf%}
zXO}JFZ@~hfXZ8gXh1x^^BcjuzfyUMUj(CM7;v?7)fNhSB!++SUTUKr)saii)PYO@5
z3#LkraT*?v`hcrAn4qBJ8A;WK|7wNvAi&utrh^@WtF0Vh+ql7HEwf@5aUHCU*)&MR
zphB>c1m3v*>=99OXgJVv_Ri*-16=wu&?}>4eBt<dD3ZSV8dWLm*fJ&LAKOf@?IciW
znnO^5H&u#l%-2EUsN@=p8<DaRwuv-QeRmB`E@LO~-~2uR3-EtY@qg71u(JdI5#3&7
zj@aC>qxfC9!*%ye#n;<`f`K>TeF?;f+K5(Pl%{lj<t2)Qv(k)2B>tgsQ>Ylm<YeWa
z9E>}-yvkbt@QT}cc>cT{OS_OW2TdM6-~3vo`HH^xO=9h{*sOH~OWNrCs(;&4k(yRw
zc@}=@myl%7s^9#2`CN-Y?^*1VgS!nC9o*>d>V7@F;t##_7Oy~%d6W~Y3GnnlHjH2Z
z>?%t;okkDsHwI+QaL_PG<Y7eLW<T<yTWsW*<lizQoKAkANQCZW?p~XOf@!8?QLlbl
zrl^`VX(03ajfeCzH(dr1VMERCGou9Eqzs$ulNabe3nTUm)}1r$^hS8zLp$fcjg3NZ
z@W_4guOYrH>AO*X?0kX9Ai?**nrbhUu6|8RpqY9-wXd)Kv=0Rvim_L<H<nD5)5g@Y
z-wl77e^{d{DQReP+yw7Hk~Vg;HJS{Tx8ILO+5uAjR4r>KyxM8mrludv{Wc<lW|37B
zPj`c-s4Jg%Q(d*F>_H9}?H|MXMtRK0{mn2yXwjeD-JBGf!8I>;FAEEo$zcJQMDr+7
z<DtEe&<k@=24EX!%$+&Qo7wPe#ASiSB`9iAB5Ksn09b5yQmQyFj4#eL9<}C2wUE-a
z$tP*87;P#mF>fPc_4VMou$~ylhYzle^D{`A@Jf~h2Wkqp7UWHmV9uqLo=z3h+=9Z@
zm*U#ge==m242`s5)_#k%6Ep5Lid0ubwo!8GJz3{wB=E~BN7E+Z)~O})-x-Yp)r+*0
zRy=>SuN2zB@Wws&(z0}9iaF{E@68$ShEwsza2@Ca%u+*GM;H~(nJtwYl?Kh18H6^Q
zhzHY|u*p(;cZ^nj>zr{49d*tq*Se3;8fj>8&LSVbMdM{UO2c+is1FrbG!HW})$%Jb
z8M}<hDA!=+i$3_1Q8j=5E+GAf=9RPIm>@-)oqylFb8pW=B{DtOp=~#UX!|S#yX(BF
z_S)AZLJ>s8?1X!ov~3vJf6dq48{o<*aA>|g4Kv%8m^bm5!j%wLIp@(N0eg+v5=3`E
z0HDkT;44@wXo*_7h+|<xEGV(!^Hp|YM_KuO<M3CHOJs=>&~UTcd>SvY`572US6Yqc
z985`6#`QyL*5vE=njs!Apn;Ig7$RMh%-DTUcrd;zD=#<da=QejQ1-nUOwz&hr&Kf&
z7B`VzC_#*I^f{ci6OCxb?Nty8$Ni@5g$EIni?L^nM9A;W@NO#ic<b%Z{^$V;>5&zs
z>5zfnt`J^OKW`KW)*Mtw%ox-i%V_aw;kB=(f3MJ;82BuZn|exohdcoOz=s`tn^t$3
zt`53*GV_x(I{BG!8!Vk9XLMEVWKaVpn8Z69dOir^Kb_gd1I(dhE%iC{?;BCqS9e*)
zymX`uc*tP$Q%Z%S@4B`)l7^_J&pbXOti#RQ>vXpi-st<gry$lVIu3H|AW|CoC5tMr
zXmOfSDnGA1)>IAHbDM7MA1E;lh#Q%!@`t^^L;=HaF<z0ar_)_D8^<*X91U(Au||0=
zjo5m7Ow9YcPSx@sG)J<~N|13M5Mk6zAw}+CZR^fgeEkLJ@E}RQEg=pdR;*!e|Dr`n
z9{m}Xdsy&!s2%;3{yKnx+bnymzQ_g%S*1l;0EY&Dp7ulmbRe&$E0Cr<uCG9Ka0rca
znb7&bJSTM%HiszZb4M|01f4BjJMv7?A7p;P50`Ig{d|uy=8n3l??e+=h@b1~=SgB9
z#sO|1AZhYwMxTyH^u`(Ny^w}L6lWKdgl+xW-N!)G!^Nwf-;}BJ!zd^Qv0ifzQ0y`}
zo=M;=JzlK;_<Tu1EFm~Fegs!otCFIv7mh57vGmOnPD4WOc{N6^B8h;#D~CK%w0tp7
zH4l<H@ef>O1zej?+bu%{W%%g#{iXJ$W<_2@e^ThRO@2nURbQgGQ|!1w=g^yk2$ku=
zW5hDNO947ZzeWC)g2Zp(R*<neVAN*$Bt@O5siJQ>Ff$6z9u12rHe6JR<thmpq|#Kj
z0NnS@j-1<R^3Ug<%EVxcAfari0qEuw(8dk9cs4%biiITYfKiIK<SRO4g#i|t*%R};
znF}u62FLD7@TCiBHKp&$mu+Ns>1GK*89QB?xk1AW0`u}NGO7@7OG`4c^F1PO%l(+F
zDSc$B{9g9m10n}Oit5q1plsEi9LL#nGU-1bjH8dEz_;hCErRhjH+4V3lZ0JM{yQ@J
z{=dbQ|MCTZ|Ew79B5TZMj~zYa=`WcHsOwDbK*AslTkr^^h`NY2UzBBZedVJMWJqYb
z(7W%ZIpo<JjFDP<zqnW7dpy4hAaCi&5ixaVLgJ&(45RaVy*ukYLRx$o&-2mT<YJR4
z)$|YXhw^VyAH*y5Ui6;lonLW%(C^=1G@!3$LmP5L`}({;3Vrw`+RdjEQG<G#`R`s=
zdl&jVKG`zaQwE?JQ&C}{aLc8_){_CbS@5E(!C1Q{68sW)8@$71=LFb4Oo%S%IF|e6
z2xj=6__FeffX6pUvN4bJ98;!pXw9|-#2?%x{O42Kx~^*YNA{^4u)c>V)!e*+;+Ivt
zoybRL?JLs*Dc|@?K!nLn7od%Y)M#>V(0pE1-fzz;JB5!=)(rHJ-zI=SU+JH*ahVF9
zs=vl*gb%4pjutjpaBPVEhQviv6GKfiXZ&1Uea=)J@?a!1WVegklK2TNaP%y=n!rO6
znRP5@Tk<9KOj_`Teyars=X1PI>ZG*rqs6mx7V6_AhFA)-4p_Vr6km)MK7mS`?W*pB
zRhUEAdmZ3XCbn?WV$Mexr(;Y9;1tx%8%w87_GD}VM3U|}ZYbqaFqnu{WID(i321-}
zVQX=_@peV&5`sb&*Ujluui$~Y_j+Re<}0bJG&={`oKwl@Qq`(l3~7Y*lst>OwU}mX
z14vVbnEdrI&I5jSD~r^T0<edInt*<3a7|t9MaHvOWfAHU$|XgQ1}Q{2adu%*PE=*~
zym;8M)wv&Oa)4?9(!9g@5_92`D0XGQ4$>eWwRLmev;t`?o1zjd#j3++rAb<5ux{rF
zIWS<nugrH!h8b{&Rr&t3pi-S_N_7k<-`oxm%_vn$hnd5q&MLXZZ24I#f|5Y}of}Xb
zQ?KuPyUqaB-Xa1ACz4mnCwX)k;tC@RE??}eEF}dttx8kBJ%ak1hxkhTQg6oEJTIKM
z>~vR~$46w?1%&|utc^ia*tkqyiG2Sa!tL}OfTwD`i~Y^hhfRrU|5tSiL-Zaa0186x
zoQ`2!j^+>X4>ywZ)3g`D0LT)?5RSXtAo<C4Yy)bO<sG-kN;WP_GoW0+H|6DhbFFK1
zAw|YnyGJN=t*JSP1zM=n@1HL!@Wl!(@S(PYk>;~uhC8H-m-HT4=yU~E`5Js4qCbo5
ziCcpwSBdIf$BvO@(oOO;kZ+24gsMLlt_`LQ^1`j>bV1^jl5$$gRr}1LBeJF5C=3&*
z%&G%6?S-=d81*cpKU4}(lbR>MG}O$)%lAi)Iy(7#&AJX>kA!hSY<~u>HPFbNw=Q7a
z-dW5%o4fgN{?;k)1~trj*zb`gB%wCVs7LvuH3KWbW#8*yQ5`Xiktdb%+b}ZQ`_oW|
z8hGmq%)KC^6p>LsF(l104PxX#@P)2}yJT1MvI(0h!mX)pp2C3YB5H(V5dFncBn1B>
zYS5zQa*`Sknixui%?o<(42%|guOTB(NS0LqGD$|^*Ka(ZI6oGh`T(7~mNtj#IoiBq
zLSO5m%JiN*lGPFm3xZf^U8c}6=;^@=KYwz~rLw|rdTtQIbbmHu7|p&@t%9p-Q3CTc
zxu(hoA`Z}BqkKblMem0mUlD62b5R>ZPJ+~>lMH5Je8!{-*4!YOmLUjaeC%5Su7AB^
zSI)f9B}}1pkbaLaCY*?A3X^}6Fu3i-svRIL#awW=@?>6J?NwIQU5L@H^zZ8N@Z5*s
zpb}f4HSSXLcjol3qM6E9v>@uiCQDqxO=9zYb`U12jNP3OXVoGP{Dn64zS}^b9dMJq
zYhUVQ)>v1OsU*HXqgm)OA`5E>lhh_eV+o~?gH9_h=DR7fr7mZUzSEZN6lxejX#Mm?
z8R!R46imeBbK7Kf0Th*A%@bekCCweHd+`&=dlK2c?NXwD3uz=}$lwuyTWgQcIS_(^
zv}tYHr@X=cR^^Fx#n^xuiN<L3K*Gh<ToTt+UU8&ZCm<onAzR1%$4Zb8Ds|$Fpbq@N
z&ds%YwDI0gK#YTYI7GY{dK8p<VJd}ixtKu;Mz{qPLo-#};fMpQDWBL{%_S`>N{Rc=
zy4LxyAuEuBWFWlyekEz{Qs4r7{?RL;=9Ieb@gq+T(TkMQV9J<yi;n=C#&qVz4O84G
z=a9QMHq6S@KW`GbqM3MjTn-;^CthK*fXSooOf=Qs%-iTTnh>%ZOY8;k6;sE0zSpWG
z=(kjtnE%mHLj7v>8Rr-fsRHh7?QUZGxZk;Q5rY0b(yt7I<^as?TF{Xn9avv{k`UsA
zxL=qMLBK4vwMY?VX2JQ0G^LqYndv2dxOrgh)O2T(npkU~RIU>Kc6SF*m20sW0M!2u
z3_ZJeL?|F(Ffcy1ZfpwpDO+x`K=Ts5JklcVKae;ns4=D@1uf&5A!`x=RCC(-I_VIn
zaP#t~9@?@?Dl&ljxb(G6-?w4g!jBua!t%@8grWT&v(|{NFQ4sriMy$6q>i*@`>}+9
zGum$1OIIR*l4gy--0<ChR`l#gNc2zqA<-`4bNQOE*QpyA9)_8Yf#WC18;3H2RB?}B
zhvr=~M(j8-c2wz3+FE4nI^wquHYrRwb*tLC;~&3@+UZqiyteA_^xy|Q04GR8yn}5W
z?XlOIxZyxy<)*V9n!-FGGm<Tcvg_3<@~qK_v1>&ac`n%u>XM424`r$>W{rSYYu99t
z+Fcz_V-*lvC)BFgtM-&J8s}T0J694{dxJE=H{5cNzA|<4%M9_m)x71PAr|QUMH7%C
zgGAtQhgWb!tvn6+T)=UK_{aPxopCEE%o=2)%d0YpAc#A)&&gRneZ-0q<-L;N+G(4%
zPk`K1=WD^!4Anr`4u%5llvg`u0=8c8xbQ@4e9Bel6dbytbE*4cf#0~)`lGF>iCNf)
zX75n2IX3=WkBaBsSvjQvXV|@lT4(#OVF9%C<fYm=6fA8&75UCHZ=8q=hm(bMAAkkb
zi-yt4d3>p{(6{u)q8Mk<*@do(V?*?D2A5TBAbtd=6J=MQ9li@J84FUuB6TZ0IxKsR
z`6cWX4VYC3>H)1}#LS!b7hLq?%6b$)PJL0GZ}BnTS%ADNShlMYBX8>zZfK>Ffh@E(
ziW42K(Xn;OCE%z0T?$d&p5v9$_qLWz?ZZ4uvqW@;I|Z^G6t%3<B2T_^@^uC*R(K&#
zx&<aL>%dr4eTKpVV^iOE63AyMl<agx@9Yp7Q;mp0-){SXSZy6fnYV^_rpEWZx1lj#
z=>7?!|Hh?b1~C6Ged7Oi_JQ*|;J;D@EWm$Q|5~ZJWsBd8>hq-AJ5+$|i60Ln|62iT
zC2H<p#7zI)jP7iQRF|OlHj5++{;2!vdq)N@UGeYP5U7qJ?YOP??QuI3myx&A4l?<f
z;Ii%!<Jql9={I5fKOSh;En!Fj@c7O>jETA6_HoW^#0E4|!{S%VhJ`UR_VWh-&MHah
z70Ky?S`O4f;<$t|4w{y6E~?rw(Iawuibl?bq)fl@O%>yTz~IqQ{hu?>)?Sncae{d0
z%-_ZfOO6BJF^qqdLSc-<Q-+3+J_`@yptIQgmV3kjsimz&z%7N%s%2*@b)1vI)31!p
zH`<m|+MWJpY@C%#KwCkrj#<KDCeVT+=}t2ejJ^hQYGi)*2hzNa5r?+uu|hA<z!Qq%
zt?2SpyC8oza0EM}FK;-N)L$%W8DkM++A}eqI;mE$tFkA1q3M9Fz2S!e1EZqNRGGBm
zL$O$_lG!PFO6pqucD&JfeHlm6M#sJqqV=C;&kacG1wbjMARt;;#oA9&uJ^2j>ky(1
zn+tOT<lQl|JYT`DGj`U`JY4=l!$GLcSOkHrEL>=pJdU&<BFIo!aV)+mpNqMS6>L$p
z#{Y!vq%T)h4Gcz)to{`7Z=O(h?a@6Yc&+2yVbYomQ2qLduvl`4yxBqbXJT*|1ABUM
zRmwuj{{d?l5tsRkbX}F1zO?~^*~S?5_jy?xoY;WM>_qTR?{q{JNIP38Dhf{_yDMpn
z6o#TsZdy7#|Jz~4z`$<k#M#8|xVLk6+vcX#;*ZE5&#!B?Io=PHCcR(pgPvc%b=W=J
z%TtncY#toG9_kaAZ(t6oI_V!Aic>OP9*=PB9?&hgguDU3W}AqIx~yOHrc!4w$h=-Y
zQqLg|SME-{Xe*ObU3x{rlF`&37dyLap%vAH7FSX7I==jq@~fk+$P{19x<~?7Trvkk
z3+<UNtyxc&oDpB2G>T)Wr%%8P=*26-;MPy$7yI_MjNO|_-Q$C~@#)Y|yoZw~T$S`=
znAevM8{UYQF`4}u_^X@%q!?i${OyYlOZhtYbHyXzxZw0%DBQ>Ow1DUH$yf6*(rK;u
z$46c}z5?=8_WVHYQwLv#r8*VjNorob`-v~%VDrSeemU6?cLtT#Q`&*}`PqP#LV2i-
zlfvLfDQ~+KXRvGa`tG}lC#2c?%|kVjT13j2>tFX}!lPZ&_KK}_n9KUc!nym!$+fd>
z>j3ZdcxYiQ<KS-MNIK?c?@XFyOW!MZ#eAf{(<0J~fbC6;@250cPU*J91P9Vmj*V@#
z-1ul}Fc3@pL_Znau{onXB4+F?DS{hShp~A%sS&<??TX*lw`G~JnNHjjl?3T}>h~eu
zz}2&6N5`6$Zbef{M2hKA#8UYbFPlPGw@P64LHAF@h^xi32_%U}i&m@BwA&??`VZn2
zKWk$4VRmZvk~hHzc49ni2ey5S;f2_85;ftf5Ur=-^sozPqd>htvp^&8CjVvPwcO>N
z*pYv0r6jr&+{>(8EzkFNs#l*L^*&daQ{P=})m*KM-xqeLBD(2yv}P|4F|;lmOj9LL
z6J#WTi^@d^f^jTzG{xluu9twX_aAqYOch7SGeuk!6PoZyyhbMYqe1xRVUI%x8!pt^
zEg6U5g)#_t`n!@PgmH~1903jmG-Rm-;d^LZIK<`w-8JGSWK@m3GYY#DG8Q7B4$w*_
z*Xv72H_WkaMUva6C8I^*WsC=0dDs^yG2p*|<RPI@LYWJz7Nd<+uDB9h{@9~G(Tftq
zD+9kFOC2HE6H4G0n+rC;Fay!YX?mxmPNtLX<&uR`6@LCkda|=ouHG&kbekO$Tg82$
zX__U(K*8r_%iv~k#Xd2v?Q2xbhkD1%1RcS$@;XOB-b+L(Ee@||P)J45?v0eu8I>$V
zzOPM<>#*1ghM1uU#0cVoqdJHdPGZ!L&qa!kodg}%ljn2|LuPFrCdS;`oXvB(31jCS
z5ijFSmQ>f0Fc*@-u#XI#NAuFNir^Rkn;qy0!{_KW{Ij@}y$x>oQ<*!VqH~r;9G@PP
zLib#f5&V}Y3Yqo%<p;tiCcJalk^SHDmPE-)N7$Ia)y^EHSfm>k#y*H0SkrW}cU5v!
zP@9;@x0_6BJ)<9Z<rmcb)wF`dUPx4AYXU%I*n)2X76M$99=dT%*-Z_sEOq8ZC27*C
zgx7e$Z-di$yv#M`)Jp;6MiY2?7lMvhCn!YN%V1iO1+8r5wP;>prNW-u(P%vMZc#L#
zX$IzEDVFlr{{B&L(a1KVrq+Wg;XRvwze<ZtP-H0uTgR1;Yg8gt`mc$7lsZc9&hE(f
zi)Iuw&v@iW%b+{psYcP$YZv&C)PxsHSLFINjFOC_R~~wi+-nb6KRjgh0pBEgBGv04
zB4;Gl<==(q!j#r#XSeNK5U$`UhCcROD=oAWe>X+ny-=a}HkiE*-Ve`Wcw_7ueHK9p
zae=hspBz%5&O!Q<WI|O-Fj2Zx_!UKCl5ver%~IXOqu}?;0mPXZ%My(U&isN^byk0s
z?W?$EW-DPM$J4~tSM30&s}Iu}-GsI$qk8{;bUNnMOtZ?K|FaLz#Qr9s5|#_eMBv16
zVA(es9*Ir(j|bx+|K8|B{=H=)a^g9#?^_M8#JWa3my-}u4$mpv{6-QkgD{5(b-%bz
zciRW0wPXJX7lK#xp9v!_aV6+LAkRjTWK*Ms(Rif*B`s&-9XS)&MHSJ%&LfWOPlq@S
zC3~1u;E(kXFfRX~97%W$+7>95>xP|&CfpIy^cHpz>_fo9DnQMfvjcnZH{~F+kL{on
zlb)BCM7b`cd+YguubETl!<?jIg8!ZEH{;bP-!=YGg<N(VgJo;(I<e?Z)j2q_qnHPN
z;|0WNfZ0UJZ2VuF8YMj>34F&h*0MF=>CLg{r5rdJ!4Q{9Od{u-WQ>N~D`ja05E;Q4
zUqH5DOe(=AbBKDhb<X4O3Towa)l@!+8P2y%@@S7{6L=9Kog?ek#V=a^GJH{g$r#1G
z5i<DJx_%5*H3m4G;|ENfVybj8JUQ7`wtu3F-lK}9kaqTolR3Sx?nWMC@j4vEhZbsd
zXpb<#YCFuAZYppV=f82_SpLdy|2n+-Uz!{X>p!dKxX2oD*keZtx$=SQ^aH53)hf5D
zQ_$>{jsgyT(k|^MN^gFl#6rNJM6mb8n?wc<Hi*Pc^d5ENZOF!6ey+@H#H92OjWxT1
za*MLE47d{cUZsBct@`3L+Dfg{42GdwJ>6VgPJNeZEl{?5Gk7fgtIAoyb9MwIC$4>L
zOVFh6RL5-TmT==x;5Qy}OZ9X$2-Ln!`NVg>4aKAk!1eyB3z3BUlimNbx2G$}yVl;*
z?Etp{h-ANQ#oB`l<?>y5@9~;86?ZGx-f|tQ)O!UU^x)B37v~IdTL?9l1?Y!Ztt6li
zU;T;xc2o5kGQEclJYeP+H*AJv9)WH2OakeLWF{eCy&~HHa$s`ABz`~y3tt}lG(!ux
z<5up^69k(IS_bHGYf8ZG8DGTHT?7Qq(Z8Xd6Nt<`2BbQ8o*ZPMGV`9{=!Fz)5XAtK
z=cq8y8%fd)Of1BY25revdR@YTC=)^%1quTAe)*5sF9Ui4N4dMfc+|Lh|Aa*k0cljg
zhg--qlai#Mhv$1@yEs>Rdq<{)a7=f5RQVLSNxu2Ku^Ruv$O<>jM`->r=x3*q%u)~6
zDB$@rd$zm*t<H*B51{_J$(uby5BP0(Ba=FEk*X*Mkt90)2wMLEg5o(oRvcj;HgL36
zaof7$&nt$?*WjP5m}XCqEYFuc?bK`OmZh*3otg#q&A%dSnS?k5XuwAXdWzf86O=i)
zFLUyKaaSTK8Wb=-96iWO>eEQTSon+C_=78X$HR4Qs2^<E-o!f_x~20`?t<7MZH~1z
z5)}kM%;NZ3!{4JlNXtI3pFxrk50Z#sHR;pgFgX~+_%s}8zw#vLc3kuxRA!2a{f(BG
zTM>HyS4@go+DbGI7$`$|s$kD}=%YQZZdDyNotj2SG@e1jh}>G}JLjC?P&w$^WAv-?
z0-QfuyXp8E_G76ee=v?+JQIxl7rFH@cv)E{iCPeC7NKtkN(YUG7DM+-H@+Y&t(NXc
zQ(p?F^Fgtp-=EaUFGE8RE2SKPBFXq@|BhRXp|mbm5@F@md?vHT)|2Y-sd%(8F2bh_
za>I5>mL!qUx&YAwCbdURi=v_J6MM=7(hsgFWdpg;?N0^k9Q;7geD%1}5?m7~3i;O;
z(c;v#h-{PL!svwhClcS}aR7T7dSDZOZH2nW27LZv<6?<I4yy5Yp-FY9U_A4CUJA6l
z+d4z~D6l+%Z$I-5%F9IH8*VCQRw#TKj2_qG6cecp!-rGh`<(b|ew0+@J7v4Tc6e1=
zJ*%PZp_1>VgGTQWvN&el#f=3V=l!|EAp{+SGN*rIdJa#}u$qm4D)|LP-G@%AhQIWp
z&0lk8XZx#b!-|RRcYEQ`<fBo-3ZWmlW=az6e8n?>370&}7;i^v{FwRun;k57TYiPs
z5WRyYx^$s}>__7(t3)Q65k0rD_(7Vf-|PfP^HZ5Wf$#58u4LsZ)f?qZ?1jp2ze@oG
z{zK&}kK-83V$`9u26+~2Cb;t}GW|grkq%mN;3@9uE*|P*diG$gDp^tD^P+1gl#wLL
zr?mRmFn1TRRju~f<dd=jXxu}}%QlG9T5AdxnnSzwR57<jQf48EszCrnQK_m}&DnF-
z=RwF+Rhp;y0>1kv<Q81N!dT`C%Op6x;5V7U1Ru|(3iaK^5bA7#)QAZGa*U}~W{C5B
z7S3CUR!e={IXf?lQl83y5mDE9pKln60d|^dg5ILXm{>$=J3qBRB>GxPASOR10Ao+m
zINJv<GY-*7_I<(QjOBZdk`+bi`|J=beR);Bl*ANT)esOI2GE|2ldfq|_d(j)*()<;
z8F-YF4|T9-6D=B+IE)ik?_i=r&P{m?yuZHM0TbM-LA<{uk{x<Fej^_4F0f9~tTTgR
zX+Mq>$XZAgztu|OEfg%MGn12Q41?aeuy5hkHDGaPdcl4)&&TvPb=Onm%8Z}IdbtRc
zrlUVYVnDcQG%iMPL_UF4UAEGgKIdPfj60X9hRXDow5Qb)gioCi%*cXC+hBvLsM0sq
zf;Pul6jHk6^8IXdN+88i8<e9MEi1G=?zAqDFV*mV=9kZDs$OCa2N|v7OKUqThdxka
z-)OVrW&16Bw``~Vz0}WMgfu|~msz)PNA0vxmg(i}e)W{{Vg(-}Zlnx8wYtqt8<mvv
zSUsp^c10xuHP`e*GlzshtOjiWX5!a2SNnC9Is=+po8+rMyjTIvnZ~gn-p_c2Axq{U
z_N!Gcg!wjp<rUbr-)}esM~^`erm8Ku6#-|T-|cU^{9;urC$Kv2B34=ZGadQ9@$!1g
z$XxWS8$PylhH*(n&(?)wuLmusv5y%QaXxUo7>h^ikHl%(Y%v(6r*ERJASIc{Ul>4L
zm^yFtmA|<(Ck7lQkIom#jggJs7fFsrm<u}9=|L^NZE0?V#AY`M-;NJ>?1ew9B^b&8
z4BV2MrqNwBvbrQQp4Xkz)EDhr_;uFi49Ag*{>l*W-d586K~795e}vMC#;YpeUu=~Z
z^E!6Ya1An%nd7vz#7}vY)_~{;ta%U5V-*3fh5R?n{Eq|to1~rmP6k6FbP%Vi6lA|)
zmeLjol72eY6iGh%!%9*wI-BQgwS4pXc|Jre8^j=4!|F}sF2s}h==JL_lE0nJKdPW$
zLo20}3PY*Wv;GD*SQG^8bz&fuH?keRE&q^T)@f8-nr>Yo)m*#Nh_vsEcV;sviuR{j
zvI3IiF2Xm8RDQ@yWTOZaD6cd=CWjS~D@v)?g$>RcwaIl+Jz2A~5f0w&;sy_ROASd<
z{Te#9SEyW^xEp&bNO{Kq=DtzlC{)5?DHua9!qTF$n&31RG6`NQ1l3kNe|9@C>4p6{
zMJ$=C_r7NEF-zT=IXa}RV*DSBy<>D`;kGRr+jdg1ZQHKcUa?)VZ5x%OV%xTD+jjD5
z?{@b)``xqOy=Q%^tv2RZtxf!yv(M4{=%cml>fBcK4|^PyR7)8w8b6JRJ4)R_8Mlcx
zk=0EHBT>L|GVSj*k2xwfe~4Qf%t{cuZ?Eh|Q~ad&NucL*GM|rr-_7Z3LWXZpnSRxP
z|AdLS{>{+iUzmu6iR*tRy2&1oA))xoQ+}fP_?!7TwMeCqwVUxOrzl+)SSPBodwh#g
z!cfAT+!fVq&eq(>Hru(>2>49FXIy=4&2CON&kwIRv+T$t)1~dT$9B6;@C#mcN3HP`
zt^7qqU`_8Yfal>HWqK-bTf2#^q)**xfk57l!lkj5V`!(B`~7B)GX_y;Os#u-GIKi|
zu#2z8*reyLuX{Qzpe#zOBo)bvcSu;Qwaglhqi^+4w<+xHSO~?4C8A#pgS}e|FFb*X
zY#KGE{JU2?9ia11=;r!AJ_?29V`1JMMZc|&t`j47!BAV-CVd=(!!K^=foTN7GYQeV
zKi>V#_7$GA4Tf)Na=XIm1f}HvcrO72PodBd-8J&SG)SkQ4Lpa82L=uIfl)<hdK!LY
zMD!gXtYlY@7$k7?KU~s2Wavo?LGQ~n*0qf!6~9Z-4_X9_8s_m~%<Us7mF^L>5UZT-
z5gT&8#q4?!e;Zc$kzeIIbNW=}WmDitjB#nji!s#Ix150{zB~6oPCauVX28nVH2hZZ
zFX_~0w|H;`Kl;l#mZ^kGqDP5ZR40eqau4FCx~*ogmO@d|PMn+bibN``*BVK-R756c
z)z+mFFyI;3N4_Bq(9j-{Xq^F@o07Yt5k#kaZedwVL^^_~sLd+hb90*b4>9H+q0t&V
z@%3I}F=gguD#CJec*UR+GeHGwexCdK=R!|4t^=);<tk%I3)X;pQNJbqC6q~+w>y1C
zJ;EERAEX=46pk>VMB*nurSe&U>$4=vZYYe2s`p@GZ%Y&owSFn-e>3vmGTCIAsu=1W
zfA?OePGw8cT>Dk|$w~)|l*NUY6QQcN=RAWPvq-d-;>Qh}#K_z9phzk%FUsgaVzdw4
zS_7+@SwfyW7?fTb{GfP>$5Ahf3Vj?1hcyb>5#nZ-xA;1Q2>{78^8`p3z0pdXS(-7|
zcNi~`8(H|FDH13BVJ<oj6Z?gs6SS8;H)>Rp+oP^Bs=$i+(P-4wAx)}*f^EcfqmML7
zOKbszI!`K70hTg>`XDv(dy7~=|GgnPq;;zDGGk#u&MpBMLj{HE{LDEkydWoG#}zmC
z2R4QWY7^I+6MD!-@zf*@HJr_haG>ZSNzXjV?bujX)jLqh_9ptn8Ol&JvD$#+IS;3H
zhf_U1d3zP55@z`>L`F^v*b%u~kXE%ne2wy7v~$*~m?}w+llfPqrdhD>W$6&d*hR)?
z->AU%u__{hUfbcc_<(G{9~GmOiFC8l5#uL}YD4lHjg#-W_;;nW<!S+I>jUOF)GK(O
zV#tvTn5a~P-W0y>sZ6Qr;u-5t=|A*a&%Svw-rSY41%~a?peUdkWH$8V_PI2M@6O>*
zU^yS62`uqvFvLR0Np*~Qptq_(xLGUK5K&G>D!OASV&KaG5kle1!0qC9UZ_I0f{=TS
za$q8fTIkH;wD77G2{GtS_CxN~4wg0vjR?ui?!A!+A`eQ**Jl3X{xW8+p(Um}%&byX
zeoR5qx?(g`bXmLv!%$JM)L2<&f}HCNb;Lie9VxV?MC8VyZ(&Y(jB99z7Rtecy;{0m
zg$!G0o;uMo{M3SKO~Dr?=;VK7AMG!jex4X<By@XViNDu?B@xNe0zyigICQb(6CRO$
z`oq_qQPjQ->Eg4sXW1Um!!PosUVu@cN)*YcPffW!RdG|aaOZg2a=c}^ScB_|p~=dt
zvZ~p$4nkZd*C6rtiDBj5(uElNYUx#^W`L!Lm?ltIU18j-Z}?3SYHe)ALyu?%nt<ys
zYw7%cY%DCw&?QjRuHWgXmhjAs(QhifNSoVi55g|B23l*kPz4|%K@NB;FYko2pmQgv
zXdyeC;ERY_Uw;~cVua;SoS~$pFCcs139h4(_JH|Y<F-&`W^_EGE{0C6dj>J237Ma%
z{zNCUedang@IV|d@hx$d0lMtN25oJ{uNv-c>@-B6T#>A(k{|xf&-SWLPxXT)6K0&x
zh?+~pAnmNl!C`^s*PoJ4x>im6g5!g&2bC$eI2${>oe*RdmqYI4btDZC6|E&i>f*7O
zlHa-pl6rKQ@SVKT7_Y0ruvVp;8y5&NKRed)ir|nds<fTT=NG7v%CZxES#2V<1SSXQ
z4bf^^B$v~xb7x_~@S91<(UI$_^fcE|A02kNl&aj}rk6Np)difKxM8ZxeIQ1fO5|FR
zM9tnGe~#EY49zEDJo2x^Zal-l>TMKK0Jg_JeXFuR4&qpGe}G4bLjQy-5pls)d5QCW
z#q%Gr*8i&2YhZRl#4m(4=(P!D#z&?pa2VAcNZv(aioFO%(Xh3AQ53yL!GaVwY3l#h
zIGx}L69ZVLNt}eb>Rs@?9~5?wUZzudFLn4}*eZNDn1zkGJ`;{dX#8=7clt>{AOg3a
z2XohMZmxOt5;hx0ottQVq-L>b;~bT^?X1kN$-aQ5_)BI`s4ezZEj#TUw?O`7A5b+9
z1116_b_?9iY;GGAg>GGRp24c#-m;v_RJnWwHk>Dx4z0_hS1q>gG%^A8uAv*55dF@5
zwF=F`d+9W%4nFx>#7+Din}?lwkeCipP`~*xY&o8-jWQ=DVFrjFt!#WR9i?pkGk-(f
z=6v51LM1-IEaF7{41E|%oGV<hw*x&)8u_~#>NtnJI*5_*dI!(55ogfE^%Tpg5y2>p
zalmAKUcnS8u~-N*H>CO|yJ{-{kP<4%(4UvQpI<WJX=|IyQ|Uc(X`v93p{PbC&71w_
zWFrCzbdqT>SVe_BCT8u@=lChn7F$wSOgHxc)Z7s#DJR^`r^oP_aVQ%uqSs|W&Hhg8
z5JJ*{)Wt)&$QPSEZPrxNLq3OcnU9-H$U`>vw{}$ZiJC^n0+Ty|+SfxdMD$RUT&B6D
z!Ks;n%@+&qR_s{RWf*JlSyjqjWGzflkSPT)oD=d<#y$J^zP9tz#lELTLOK$}dRI#U
z+wQ8~!*20>Vy^3#xtT?-RP3o5n%v#KG)m5;AmP{wDJxA`Fu)yBYq|8XOc{XT)}aOK
z5-3m~{FGT{rHWBNy6!jbBykOLQ7o`l>xrh#9)YliQKaRmzxz!{X!i}gjo9P(AHWq8
zC-Z+5oc{k9L#DsH;{Qt-vNQkBED0L5<?M8u(Kp`J?N@DIMnvxLP@q5#3guxN{E&am
zF5L*b1hJbFVc)^VeZ5cJoM9n=+!lZz1`kMVa`aE{a%G)oS;12s=xC43Y0MDWENM`(
zTX~vWEJfuhEA?!{H&PzVRMrp~MrrmdB<SY_Q4~!L`zt%w{0641pAuf;J-=ZqM+GBn
zM-da^fH-G`9Q9a|LQ0>ufL3k@6tRi}RnSJ797EwYV5>I~YMW{!Fp1LS5FWkflF64e
zgryBtP8uQ4Uyx^iC^N^f*BInPGuDMIs%L54Z|^tD4$fePtTzAS1XWb5SQIy@2na9I
z6TWov`vvdPpt;OgW43dGlR?^)r)@5_Ld{yZ7-$o;#NExLY@RX_In|)|68$3|9zkQs
z&*JR98l5Him_utud4!48e2}QXL}>kjtXP)L=>y`Zsc=`5agjUeQ?Q@TVS!YNaraMc
zyjod}N-Y#;d4d)r>n^x7&V<R=V!$K!_gWx)<~Dvl;7QpLMt4Pqy)Jdg&IG#CQI;%q
z=Fz>%$Xg0f8oZ?q4w*440g$N=jz`sw0B1f5gH2hW)X^cPidqXN=7O92fXHA!(uqi7
zY;z1Z5{qqR51LP8+#o`3;27ON3mE=c`wFC$K)w8G?KQ`>2|hAvMz{ht0R#;r!Cj;|
zZZXDe#dYS$J6@W+)x)lCFE@bC-i(f`cGU6E8f~_W0%;w?hpC;D3Cy_fkSlfR^mrTU
zcvo~+TG*W|(3XO!@8#fepNrU>T77kORh_yl@V!5#;QMuZwY_`y_}P!xjA)07_M@}?
z<Mys(oWt(^cK3b#dAYguz+7iujszf@A=V2~zGkHGc6Fx;^l?7!_H1{>upP4PJ&@|t
z(H)61e%(FL&3Di!c8(O7U^HyMsAgp6A*9EVfEiAv1StqJ`iHF$xjd&i`XIrmH^*o?
z*^z`KrtPfNmQ;bLe+zZ6#(1HAaCSXl{mCKKUn)ptya!<2m%&<Gw{wNt`YReeQ&B5=
zp_$ooz*avlHu`|PR>-pEfVkQD6s4ZF#c542X7-Yof@vPU3~myV3OA8>YIVD&`+Irv
zuiT`63ED2NFHI(fcUtCrCX=GwhRN<XJ>Q)P#zZ^yKCroNIjrNVBt09>6fhoXyo4i_
zhO;s42nUSSC<^?ob}0tBb_Y~LRh4I#l7<Wj_7aZCAh$8Aq^`b;S7DX(b?v@RGJnks
zvV9CftL8sa5r)l%4GtRnZh!%t@=DJxhurWbPR@(8I|xIuov3Y7p)Ms&WC$^cStv7N
z5<I1_>?iP7anrxHeNd3#`LGHYP%%fU!_AcCiTgWhQ;(8PjU=>DVI%4n9=TOcu}qdj
zn#xF)Bul|ZNff5u$5H!O#t30)4mp%z*CZ&T6z`jEFg$aGBS{d3%<MQ|1QEs}7zen)
ze9w5;F|dW914Bd+)J;U&loygN#>cV%IG%ggeW(ggz2{1Hlebz%UH=hbt++=LVU%Bv
ztu`C`*y+N%h7sY%o#4bRU2%`KE{*;T(wTkC*SMD9RO<m=$Ce@+DjvU+Ryeo+$Y)?~
z@L=NJ1r2T3H)y11nnj%xbKw2W7&WB>IP8fNuvd1U)}g=c@9j6|4kgsUPCK@RtnnN?
zLvxTA#}Z*b>H`nfo3Vc(hi&VBy)}1-f+~0{qFniWotC)Wm)gG`eU-e2E)&{rZHM%Q
z_8%@zqAR)y#l-r)9KLOypHDur9~AB#er}?_Fn9#Vd1k;dtUEmSAnN;gcs$yFe?4A4
zULGzsGcefM`nGnxTeIqIe%xa;xGk@@pG)X5uLQk(J?(!^BVOAH5G-R>yp3NX3T%0K
ze~;sli8V2jJuXgyc1c~W<{3K+5U3j`ehsdBa19Mjwbu>{KMLRXl>jDSH(~c{851z;
zFb7~`c=0?q@9YnDNBZN3Ntmnq2w+!X_k#C=&w`J@9|Jt{fAz-KlK>>9VzThLcpP01
zE=E@WnZS_v?~6W02d|aWF@5B^2`bEpCz@F`Q$1xx&Nv?Uhr-IAHJIH$l@mbG9%*M`
z7V#kd-i6jQ1@>F?(hN$-Nz^e@3#s1Qlp9(-L$HsIL6i~mEUBj397n#?oaJc{yH2xH
zGUre>I<<BaK*(<rI2bz4`0rv4eq+RlAn!wHiu3W8G9nXxd_|nXVt@CO2Fh;I*&$%|
zw)g|b2J1cm{e37YPcu@y5Td+#-|akXb*=6imt<h*qMDI;$ZH?VG6Y+4yz6=b6qme%
z2xx~-y`-=(Y`u)`l!H6~p3B};GNz(?KA=poyEaURo!Ta_Pi@W7yx-7M(>gK;3bi4T
zuUT?hqS(!Eh&3#>VuWMNO9pAB?;Y`=S#1_;q?5M2Yo1>Sr$VZDBGPD7#LNDmYQ^o1
zf^!Nx@Gc+~SvvSdV+s9T9#o8LFJDC#Xi1MP@aX{kRUULuuD<U)%gZvuf77+iEX&rq
z+4y5xmL={vy9zs2HLT-iX@7$a^nJx5(|_{r;+MpSuUV#&<s1tZKv6yC4lsp%>h^}s
zIR&|0(6g$Hw3g!zFiAajn63d+*kqYY?fhEZd8B{Bv3n}<EtaVr=aGXCQn5H`xMp@7
zDZy))K3(_0q<dajOQYA)qVZmee%dXJ52+x?{FWlT{ApjiC=RNgzI-;%&+9N-XU&Gh
zvQmLB&h_SNR_L$$^lq;4()XMp@5LII3SgYeV+#p<nvewj<qbh0mv8@&D4)ezbIpp}
ztCcRS_B-KN9%sa*JNd|PF+Ewl#H~72?YKZB7httyhO_dn$3n*xC`z=>c}eki)9o$n
zHSFl*H%?(Y#YV$)BROASAw;~*;f4~)r<wx1v&z~UvebBKiTGP(70yYTDF0A0pbF<S
zw_n|u7_yb;eo?Yp>3TzxW`r=h+KGR+!drg(soiSE>pZlb-Fnl*8^Kp@-%I^rK%b=Z
z(s?D17!;OZ^&#)s$*4Bor|?L~0cl_z`D7xsVVNBFlp}&b8O9ueq^<W^6_vDx?wuZ&
zteQjCokyGi)<6(wxbC=4>83*T5H~IMpz~%jne0{G7sy(cl+cN4Kl`xl_|U05o&mmn
zF>L1MRd;AyrmJDf?cy78{Fd>|$CQh}cI(@ci{NXk0q<4%(NIIRxu#58V{x+f?G&41
zxGdB7Z+Ld3;o?-$%C&^o%H`;Ea5OR*pGm?YVH>l7r-}dH*DBt>EgNkn3HHD1DrN(}
zm8UMVbI+{0WtGPEDdNAvF~g^;Mm!hM3;{q23A>s%#56J_)i>-!=`qw>6N8-FJQIm_
zAGiCM(??RVoeLf=v1^xin56K<h6Am3r=A-g+jQ!zJpORq;`!krvlF>onDJffzJ@FL
znN8Fmug5y0QW4&N{C!V{ALV)AJH8jmERxHLCqUO4_}0y2g=z60pehUNzj>|vs}9J-
z#{5641L};%Y_KDBPOCp4!c7)Mz|NQp(zU}MU`^1CuyGwGd<z5>v?MCqSpu?7h}tD^
zHRsANRh1rYj{OO`y0Zngx^%$))vVxP@C4NO&1~r=%}e>07QMK+UEXabDbi7<ZrM!R
zO#0L=*%r>(5|SCa^f0FkzI}KfX0Tg86o;0meDz1E$;6;U@Leal+|uvv9@K=APs&bH
z5xl@M#DKeCUz?1XSUAnLni;krE^%Xrx1|VsLeeRgdf9d@Fklej?Sw6kna@E@oL>dB
zL%Rjt?|YGtc{r)Rx93NxxBZg$l`9OmTmIt_bV{iMGaJacLD}a*6VsWBHuD<nMmYGr
z@ww(LtYDYE3YeM!?%_KJ?2vKgKUENPI0vLchJa7OvG7W9kl_gS45S0>e*!Zi2>11a
z9hc){B@2Wg=;@9~{$&r%B6v*n9J^t(`Zs6Z9ECc~6v&kO+?78q&<v6DS^1`lAIcsx
zdL@rp0w1ZS2>LW1cmt#LKQsfIl@h({0rR!6+y@bgMVX`Wv10tc9U^VX&GTjk@W%)v
zmjxwpC<9BQhcNpXIu2(i0rIhv2tfLuN~h7!C!Z--Rt0Qpu3wWXZj&C-7OCrLuJq7=
zGFaNEo;}845c$l4RUa@Ngui9|8*+l5COc2%{5`A`0q168JUuyG!kp(ghDEud=%9V3
z+j^s5G+d#{7imIblgI<8NbNNCx`?7QOs;?Uj4Rpw&>PA9#^@^+A`K5gR3w*REK-&y
z{LiVA%1W;lJt3rzEGSncL!eA_DNl(D`c6_lA_X{hppi62lv;j>XPS-RgX9h%tL@<A
zK&H-#rgddA<C0EnWGTw3oBOgN&Y}|5r!X%6+~~b0$#d}`&{9V6>)OOgfMul+iPdD*
zW~Q2j*Yc~CPr*z$NYB3g@qR#{v@96e#CTCOj$Pmu+z7>^Lf{fS7Cf&Y0^{#@rCsh+
zr^eX*mGwJQ#~3zk<i$uFM^KI!jHoSc^LN-}Z$wi(&Np|g027a}=n5OJ0;Z?%-);f6
zf-950VnGo3i+F;Vp`b#fbcHZ1f^znxWC-72Ju`$EBpVgJ?#-(YPA#EH2;)G!nJ%jU
zn0J+v2@Izx@{1cg%`kqnDQOdoF|cL#j7|K<l?Mma6t)=X$XS#z)vB_b3xs$%p+`m0
zLusUpJSF7^X(UPTc<MJO+wHa*As(Dis<e*j1`aixnO_8nhSHcXr^`rK$VrQZT&)d(
zw@XFnj8T+&;xR}vG#e6OL^MY1ZO>vgp^lP#f=J-(455C^qNB`4{>6C8VsN^AMM;XA
z8`l(yft!l7LCmo~s)W+y64{J(gLYB+u%_;B(MTCQ3bV9y%FRKMKnv1)W5Rra4YyX|
z_=m7K+}JU&25K}6)9|7lNRft<I{P{7d}j-(TQJeHkPugxHFFfo8?0Klkh7H7a49%=
z>7O%L_A;<&={U&&U^&r*9d?6t(ooD$F&vApn|dR7>44|s_jjGA2&Zy#-hrc_$!y$u
zlDWAD1MTzNVjl!MPVzb9cr}sPFqd%=Qw}~A*U$X5)RFhn7P&qwL6W2I83NO#LncbL
znS9BMUl+oVPL7HmU8{#HWcvZ}$@FFVfBki-gsOSWIw3-`1bNywK|gEWUti3_Qe3SB
z$dj71Dia7Zf0E|1f(y@({!)Ja@DNZl`BC0ghQl;w#{R4-EZrx&Izg0JT_)Ta$I5kI
zhO6v7IlKT0uh6hs`Htl;*h0DPW0EPD4eBwM;a{8%kaFJaeo8r*NThr;apT*L5XIJ8
z8s_rws2V2{C(EAAIaLj(Iekd1K9gqa*+Gt4Kaa@62r2rhT3Llg?9uGru95Xuq|X+3
zsdrOfw-)#eGCcfwn(@Ub7exIik;*Rt5Y-LaIkKiGxg=s4e``}y?MFad4ym$4PX^N{
zReiHX4AO)Z+q{L7K@6wkBgMn~i2JL@U^&!f$+3!XhtN4+>xV~*D|$+6h#&$odXTD!
z*#$@%5668}S<3LqBAIXKXcsF&l@CeGUw+$xWzKHFsM&+%xBk(nIpo7+Evhe|NBYQw
zGN#AD?yVTds6#eeF4Vqbe?%deBG&;MH#&<$8@!K>SyaL<q+D7o=s4b~ZS%$+(KBY5
zvwW&V93`ofhrq%0&@`A^sI``8Rr;FIQdh7VelG7EDZFVbz$FR~0?|Dt26Or-pW=Lz
zSt8A=g3w<43?y0yFnXQ&%oXKiY4e4nyrQQ)OATV9SRN}OHtoM3-!L4Jbr3awmoM0J
z+y}65jBq`|vKc3z@mA-!sc9CKVU*SBQY;NFY15R|7A9Fb$@IdlIgPtJ$rc;YQq?JH
z32mNl=&}9nl@}uIY@IaFLeg#2jINxd6dmV1^4<$ZM|sR^Kl15{nzWu997ZESFXCqp
zE~^P^H^-x=Uh@ZNA1;*Dx~_PcNl&^<+q^R)h*+7JFw@t4H285dS;mywBd`5^wp_>@
ztAn(82IbMOFWX_6CoE1K?i~Z|^})~6?TVx?0U+?|ox3XGIov$`aS@|Ik}JNi`GM<4
z-&n~>^TTvM3{E#J<S*Twv`Hs6(TSkd*&rgTK-`hA(ZYo1!)n(T-+Nj45y=sL+eOSt
ze~-2WXm9(6J4U7a)@ohpo!@)Q8@ht(jv+m_Nr#%4kq_m_D~kgRFL<YM+tnq6)vi80
zD;BY{DRTJ4k(T2O<_Qi$^szOyW0*oN&{+?7Bv!AlxJjf<()?na6U#szCo<NO!@$K;
z<#28Les{N$qkmL!OS4W_jmm6Lx3pnvvrC#Q^YNkT=0w2}I<kwPs}NQKQ(Ao9Mjg+c
ze2oqlf!BY0`1XBwe}`nkgmeB+sO3L&Su!zm{twy9%xT+wPPn0K@4sD^t#2X`C`e$E
zQLcV`2#&ZHF2}5?-rxPwb!h8#dkx_{i9p;fE*C0EiMdifxvI0D7oyLXhwh%Q=eowj
z22yor^53p=xkz6nXkU5xF-2g7rfzSGpDWP$QAk>^a&OU}Ry<c$^}b@|=x@*Qs37k(
zm)~5UpTx?7z!4Ry*+37Oi@)2yKhh6pueU!YMye(!7z1K60py<(q*wZ6+PnIeyMcwd
zVYzs#hJ@RI#8J=$BP44~P=b%8P@9)>1-Z|@&@%KI0h1_@-ET{Fm+sMBlq4}+`d&hE
z0tV1~{>!4(;N^F9I<AXXh0Fze@(%L&>9`n)<PG?1K^VfoZEG*}%YDTVAr^Xas^@BP
zW7AlE@02z7V?LWh?;5%tVqj$lbb6d!6lf>{11QC~QqWyHPGUW{jjp>;g4ij)U@yv8
zpLt~ecHnsoBcpO4Gx45{9OIM-Y#LQiTILz>9nx3Ps7OKhI7kA9!`i67JZw61o(W&<
zEq47(T-=?hZ>ps?bp|pQUQOiiE3>UMuv9D5%8h%=ZE>sh%T!S(A!~S^EMM-b2S>4T
zenTnC`Oj^#>f<_M$-rF?$5jZvbj7^jQsN`=frpuu&=B24+^3Bth4yB>GvgJG(v_*6
z3K1!Uiv7U`9F%SbU&Aqd=fE}{%ptvBf#h6)(NldIhkaP~9{1~plqtxVnaM|Giu9m(
zsFR_S!OXSbmq*X=bU}4{r9-jB{WFL``=4dSRG_RwxpMOx5L+Ou#z0DHLbzh}(vv1P
z+JA*9I8%RY>V`By(sXeW=Qs~Zb)`0F$|zr%4^?m(^-ss~P@#@|2B3g#_)_Ie>hWg6
z1;doOcMKzDTTf+)%!2RZLs%ET9Jw#8zP%fBZI|fPtpq&j$^lX7wGYN8Od>e<_1NZx
zNh2F+8O#bIZCZ9<N3(aGci`TB7=p6*=W9!h)mt<fz|opPcGlni8NN(xc8Z@x!F@oC
z+MGv&E$~qQD|*CC&U8{Zte?|*nK(zLc@QJ}nX;c|<fz<WDXL>#0}Kp%Y@7s*hZhvC
z^Pk)hmEUrY)@J96j*Iy^yA%HPw#YT!lfDh_WoC>K3MN<Lk>E8>stp^$!~iddVY170
zr*U5k@Cid%*IDu0$P4PQD4`76;^m&zuF`9cd+j({jR9)eP)$BrtA2>>Lp~QB)W{F%
zWmO1E{SPN+U9-eR#xRJA%HxXtuylNzu+#m{6L5W$c#t67sPCG^yy)j2doh{6eEuM3
z<xp@FVD@D^*hd}3Mvv+xC$ApSfCkCzW+vUT<sx=A*z<O$yT5iYv?UdWgnPWDNqeYG
z+bzG35o_$f-M&e04fqLB2~k!UiLq2rYTnO!?dEIiu;?YsszaF9@Fvcu3OTt27H5MR
zTBA42p8m9*28<}fJPkLjCM%GMcX|y*Qh{<95G+RZj;;;bEYX{f6BShn59g{*S_zlH
zE#$d-<p+AKev`|XvY|?r9XE$RyXg_|%r1PMynE;6rTF9EOsgoeIC~{-X$k^ZDV3(}
zP}So)+g#ga)$=N))0|<V2rKa#PtO3$uzxL;$lq@=#+FEhc6lFH9W3PZ+K<U2K^J!T
z5n-+F5;9vf*!bXD3);uyI9ny_Tc4KJ{2lMtC+Hj9=GHWQb5oIz@$8n~7dNEQde=7Z
z@G3?9^}m|09^Nk>?gXUWGXH~-G7~e`zr76pA9#!ZYIR~~;rt)HDNbtZMrJo7cRg3n
znEom(ob{3nK$6_!rfV?R(`nH4s&zrIF<)UjN0t2c&Yf*v!`dqhW)8ves<<cUygxC1
zm`IgX^`s6jgY&}rvXoLS=b&^r-})mJ=^mI^>!v0p2LQ-*PEWn4N#A4ehrNPp-5=h*
z^k(A&S6kf@MxkK0YXtfo38;vqZY^pjSmwXV{wylYE%F?xPoZ`L4snTkisMk`J7Am0
zoAinP8#HFLAbCWhqP|OJlIFynWT%A|vdL7i`r3rcL1Aft*|O(G4|u=ec9K#ZU}Y!7
zaYA4S<Dt;W_Zjmil4>2Z$QkWLF40L@by;?jofd)?x>Yzl&OXN{4a_nU@g2ZRdye3y
zs8G;{<ex=Oo>Uy|8((BM9qr??VY}qcVs%+O1+*lY^z{r%lm6Uhrau9A^Y<UQ>Prpj
zt=}Xc6qKTGV?#+szj+lF4Y6*YltNE)gq(vu1=gBQdXZsAZnjnbcC!rxME(>wi2p6W
z$pl4xP@QjDD&ykeURh4_<EO&rniqTC-<C{EDeo68CJa}!TYW3Jz%{QSiv^}!DICX#
zdDe#$fW^veb<u1{EKHp74qHQrRV}MuOAh8;qsb6P5_kLC8G5=mHldg)`;Tfoo}QK8
zb3<Nd5=Jk2wF(5+lO<T-@vj<K0*1a`(jcq}tLlE@XE1{xtjw7To!38>UBcRO7d?0_
zIFa0z6Zh<D_EoldSal0y*s`h1s2RFm_8vZcw{jKqH#ax+<?VdFh6lSpUx(X$OXioK
z5e2dcZDA<$JNQ0tA4=vqIGnv+uRg9E18qj&0iu~^0w|o%t@{t=%M)wcIbG67-$xIZ
zpU=-Fsi$s)H*>MiZt0}+e+i3REa&}=e0|QvRs*3%hm<~^2b{2~Bpw9B7&G>h5FFF^
zCXr)7X95qjW3~$*gsQO-<}(CckXfs7P#CDgsXeFTlv}sZD*uz;9&p=k<5<w*g_`fK
zR~TV>U~6W!c6HjgWK(L*PIe9_EBkPL;Fka=C&fH38=gl3s4W=?4};xbu2#T>$+)Ws
z-e3jV(=|{v@mOH3mtfB`DOS=zp1^-iam0X+!M2!i(ny#l8I8LAeV?n>pHuHeiUor)
zft~y}AyuHaVkwQGaC7D*Q-aImHX~Gqf*h8_dVcl0p7_(!G@3MfqQkLlxRvej<>Px2
zjZ2@VLwYHm;ADa1bA{o%6OG>4JK`sr>;@Vzri{BTv&nC^g2HVP^!D|G4jc4%%2k*D
zBI;+pXf<j@jxuql!;%>c@8)-iRB>xc*p)OGmkG%od<m6x+wosRn8xp5i(O-_UUMqK
zcg?HnCz}jHRdeXu5D5)TN8Y`hAxZ88xzu=s8Ak77Mc7#JhrFW^PTWW;A+dleL=yLN
zZu%Lk@~rHLOVHEj{b-OR=;6)tEEw81N!3R;flnajG+hwr(&U8rK)I2K6cpXa+ET5^
zKNHh%xdF--VQH_3H12Ow=S&Fyu`s4xOyz&0b#HD%?{S+5*x@_Zc8@50ff3O|VpMl$
zu8cmNS)<s9pooZF;D#7?<v2HPyFouNZttlTbM$g?fWiY$alujpI+%AYZWK%KqPv;q
zwGWb+hne9Mj{AEw3m@+IsRjdVK&Donvoz0xU<O6n-O(sFR_dO|Z>1fTCVbDlpCS;U
zeV`Jv9yd@I^8(16{1#qB^T>OU)8ML{Di^P$fmfG(Lj<L_LD){ArM+gqo+O35+G}bh
zsCk*-aQnnvb^8n{c6H2o9wmcAnFv9L;cyw9CfxO09gGoG-9d2r#HUqR6?o_U;A{UL
ztPiodjME0HzzhOgErs7<DffYIrpa~kNNL~L)!OggRcJqK!4;3`E=!<*J+R9qL_vz@
zLeP|BVwq$}XtjPKjo&K+|Eap{>uzYiC+>CgwA%Om7dJaHF?CUUuJ7CF<wdgm_QtpK
z_s!AgWxJEnb(Va!9m7gjyGm^JU-ISdGIRX$Fzdwq=IrYV<h9x?pe;50W^9}}XQDTi
zp_{kUr^WaA$g7pB)93Z<;qx_hd_8t)|M5~_+OM-c>zbaBH*}*U?0Smu%Gbx;?aSVz
z@N-swO5nr&{qd8*DcjoYqnEkbj_{_~cp4c0{C$^FMfL@gqh4h<FgrP7^+tHmI8d#y
z3ZgosBBUy$Qd~Q_3rB{TZnS}PK0)?>3zo@R=|^3ps$?8>6V4NKQV-fl|C#C_y-%2s
zVP+Wp{{*x$b+n_*WZltRNSq+dKT|-+PxN#_2xB+np$_zahTYFhxrewzVCG#2w<C{s
zLy;rPdgKA)+~D9^E53@FH$$~Rh+__AxInl<g5v4`t%14LAp4_Pqni_~@Hy7>Li)8g
ziXf61<~n&B8J25&%NZO?4fVkEVnG%0Rng2}*Mh(rueVApnQ`~QAfDHN)>A`J_+?zp
zkXW(gP+itT<aT1c>WCqoJYiD@6cT4uOsj$UZ|%_p7@Y8m=2k+1boz!w5OEHDytAMS
z<;A{47}yA2eprjmtdiGo7qUbIOY9#q{nh)+w{)w%JvSq!#`QOq3F99H0hIpq4fU>0
zEmfB&TAJuWP(%+gnciw;{b*?K+CtMpWzM!(J1c_LP_+m+b_vx3#mwd*Zb|R9v>y~j
zsvRw9=AKE31pN8DiWi#S)sCPzE!E#XAE8duC%9>siG6e}%?h<~uMD|z1;!zK#Ht>I
z-i=G{5<%Rsc?t?2zN{%bJ7kQdT~98w@o%XigW)H`bHIh)_k+9-_Z*q`kytV^keZk`
z@#%&*jx?0AU#hO=Jnk7_-1FmQQ1V~&9G(E++xA=LyAT=t$SBj(_}LwM{54>9gu?AK
z{V>H7I>Q9!7B<6NP=3x5ODnva#Ks`(e1iXgA5S^;VRnSW<@B7yN3u3(=lnt@g_V4s
zK5H4zmKXo712igD`@(5&guC1;OmXFW#lg}!qx%l0k;KjtTM<Ah8;0-lXAhbtO?7c{
zLi=O_8L3QP|ADrp<SeN^d2)i}+y@GBQG`_8vHIQL(fXMBc_w||9n&otIqfxi;jCHw
zBm#C8<Qy@!5%UbcODnrEm7Sf2pI_u~kWc7W*PXHmOWlH>IIop1XoFAX-5s)L9;^@y
zYb*i|FKV}>OVLx^QnrlOp{Y3F&iXJ|q@J<-PI;Gy>Ca>97;LDSAynB>GnK*0h+U<?
zPxsv!Qn+1fnqL1N&F7#2r(jG3xjr~7<V-i}-=~Xh;}R=4`X2sVY!|OU%Qh&C9}XZ#
zhm#+6VC5jdi4WiEUgrB8gP}`bseApm-WY@*RUgJjzf5OLO=n!l23?BQU0jwHHFLSE
zo!hJa^{sSnryFe|T~Dx<sr$FUM#@&kQ9EHhVI$?BiIf9Y@~JwA1Ira}@G?O1R;%he
z-2BwZ<jqA-*@u39Iv*QZB<jG8D`NTmz_Q?*<R;Mk8}^=M<KjQn133S?nwg31e?im#
zn**b23=cw|Ae`uXRv%{1?KUoS5fQQE!uk)bKP|rFA*3i&7qvroxTqceJ4>_0scVn*
zm<D386Mk<Xwz!tMQjDv;3Gt1BpFR7szqzGZ*+?AilJ}+PGD+rNDxTr7C<kf3{Zag-
zAjtitYe7vdk@N&{6>+@MW#%=!Gu5_VF>AZGr!C_&vVI~|!;-uLaT7&e30a5kTU~Y`
z1#y7?hiheJ{10}CoPXuq|NQ)4+$JwIW$Z{e;JV(`CkSiQ3by<V7hKte)_+E9L=Z#(
zWrA!zc${oAJmw9=1FT6jX+2LM>T`C(k|g7EC1bMm$hr8eO+NYfZTL6V8L8+h?ziL=
za>;}3NWZFoWA?6&N+D_Ol`o$TXD99aW|ioON>bZDyyf4i-n46k<`ho<R8ZJf?Uff~
z9AbcOT7un|L-i8=Q3ln3!64XsN)U{zPQi`drs@HsyuUwhhGyHdU)KNd(e~;@5>>~3
zgnYr?P245@a&uXpapkkkyGJNh@<ThJ)3mQ1yfOVIT9qK>fmIn6(*I#0v{TMZTAbxs
zwfT3VEGuilU%O~<NrilV(?ZAslX1aAgn>=Ev}S-U(_U1vQ8<X37{pcaf@ds*bi0OZ
zhk@3H=`Os<yPSDZC{?B<NpMWB{I_g6Et*DP2!=eBGc8rYkng-c=S@E$LvEMHmnO8y
zI}dzvBC>%h+`N@H*bk8~hc}(vXvQ^?b`&K`0Qr6pI|?L{F$u3iW&@w}de5I+R5ln?
zI3VLC>64v#!3@86b9gDElBHLECnIgu%mPZGaY+R!=|j2L5e1!7Cf$$p5aMNq`J?#;
zj|#O$1TP#E6J$L6FYUTAVTETchjA3>tT=<DALXL4Q#HbeZ%_Gs;D8wozZ}2YoTs9x
zC1-M&`AkX5NK&LyI-9)e6~h^Hs!6HJ8YMZczxH%GDyrugu+Q(u%VW<0I*1pk0r(A!
zOLE%y_QqMZy)l29&@Nm!ql*pH*;bgPkHQMlQ+lkB0)Fz#TT6*G+o)gNngvplRj_ej
z9pF7?{K!H>+GJZG!9O!+2K%AWM<PMl9A)%gxuz>pH?e7As>1@VP`N*n-fHaqt1K+@
z%BJ%=%yW;j6d4W}5ZvOI37&7NpV?NfO>@6pjk@W*ees7$4e#pIAQk<B4q(L`8(v>n
z74C#ZOdqqi@Yc(DF&<&(B0$ykxXwx_E<wZtMoEqI4s>9Ikgtli<E33h#cfpVhf*|A
zLM%hWTZCd0I{C^KJgu07F1Y{=*C_TNDIuQ}$LMf4iHN8^Z5@`DdD5t>wT2XYf^~sS
zVPWuld~4;3m*)}QgD^u>4RMK0MDlb=>eM&}WlvrwiLjBkK$<a>m3mm>@lm)kJ&Edh
z^!i-7O}~3FW0I-}9r?i=lz@sX65G6_3=p$ZOi}}OufjEVy-7??dl@dJ7oEP`Lc!Wy
zxpY5~@RyEQIPB^tAEL>R(z>Ty<1$WRgVN(Xm9T}ouY~olGuw8}vQJ_hy8?C<F-v2-
z8@P$^@UBVg{?lYfDz^-4X?A6$!u^L|_v#KV(8ei6b4wV4<w=yE|NfUbj$av$X_Tww
zaK~!%s^-&%qrud#Gaprx8QN8Zcuq5%TR4w}+8nZQl5vDZf)aE-7kvev`Kj!Nt>RSU
z*|V`^iJfX$nrF{p;*OZ6gLc}xb5AaUpf01N#+GY>bAhyBt3Cd!#*IfTr8=3Sx@rTd
zMES+Y^Apv*L{g%iP+b8@Xqe1$%5K~hYEHCWf9MiZj4zP*0ip489hPOPNEPdp_XOuG
z+{$PIjyuKd9>!xHmm1ybT%gFkyo<#*s%05v_fVucprfHTseMhbrRF#2aeJztIFiZI
zNHE^B&u`&S_iy2U?%zt8p6kNM4M?^6eJgw%(?ShC(FE7Qmie*neLtPm1h6NrCNN8L
z?SY9Kb!^LX%~=_6l~qo)tti+xjV>3^-pd+McjT=w?JBa(S<j;mQHy|(4|xqCY@CmR
zQyMF#_N=cY;Oj^F(dikyZb(ue@$LZ>ZaXd+7A^~~Z#FhmbE~-u4BJ}}0Uw7xix=hw
zY9l3G)vve)l4L5Qm8do@qnlHI>UA*oSF=vb)tB2kIF{8FR3|ic#K@FIEYrR{{fBif
z`M?Qyifk!w(2h#ivKHN%mfe~FtbD_TA2I44^%v!e2(;5U_xK_N3!dO9rW9Xw3iD@7
z#1&D1&InUG_DGd6Qp8D21ec2n8K&!V*GYHU3Ugr+1lmEaDi765$t?9!ElN>|qE4WV
z&4OE{wZmDQAc@3DiG97R=yVC}bFF3mN9}!i4_Oy4tYd3^$TV-qi%tDA7f{tsIys#M
zRMxg8?#!R1<LoM&+-2G}oVebZ@E*D4-q?mzCpRv?7Xc-!>OqXY>of%ysdk$bH5%t7
z6UJ4TIuVOL@*VyC)hZKd<i=&zLoT%s-;)#p3fm4l=T=_%=O*@Yz>y(_a_YPs#2P$q
zh5J(wGBr+wvNU$b+RhbSw%heE=$8KIJVl4FlD8brjc0$#`E{8=SR%6{*_<A%*xc#(
zS$rgN4i9FBhpe{HWw^zGhY*UF@Sunr$A8V|W8XPO8qV4<raPrUOjFRRIq|rD4KW0r
zwUE+wii4Lzx?kY)uVeCY8(GSi63l>s5<ah89U;u4u<rD%%^_|nc2vfvCH}Y!C5ziN
ztZe>Skjvcj^Si3I#&9V}!%8rynS*HR2D!#w2(E_&aW@Y*1b=?a7Z|0;ap*rhv{^X+
z_sW{a-vP>|#?A~sU5uRnF;7|C&e7(dKmVw^5iv0+s)%XOD;fe!<y~xyOdbD~k+L<j
z6SXjQ{ySm$pJUII8H9!H+_nGVP{j7Pdy9#aQ<p))(ay#GUvrxOxxw(0LB-L~*2&(`
z(bU%1gF%YO+0n(6LF6ZcsHv-kv8l3zFh4(ovZ<4ui=(lr(?53pkMmXjmp++^nS<m1
zj4ZNjQsiPx+2DeYHO1Hk!%&ayz$GQ&{C*hu5r71hLBfMJO#>*Nw@M%y+iX|SDW16a
z<w?!HyvaV|xc1L~zTi1h-@bkKbKIWmJAQm#6LB3@{8am(ko;gjcmcee0x%9=Xndo#
z2d9<1nm-dO&m=2FM-#M`UZ|uQN~5teNwaTBzOcFxzoAhA?o4bHDw~@B2&>pp(3aGC
z4#^}mCKu&g>UT@T=UCb*CA}!qecd&rj}c|(4Cz?&yfn3Fjd9!FqXSG^gVtHfeC1zE
zvoH0V3#s+FW|gF6ik|C#wOlScecf)HX2W;U?x6d!hzB`VuDK@DwcS-={kEiYsNLFK
zD>k`PeX`?|>xEdEeWj*~UMJ0$9uw=YFIZr`_jKiKBD`svl<Kn4+jUTk(OVB4_u4$x
z_Z2wRoxL8tR$gDA9s(?;q$aN|Pq=h0zYSd-a$V>x65Lf;+PUUhR->yo0s_Z7cDwm7
zg-Uir3|gk$DZ3I{D7zdxTwe+zL~iN0Vd$LS@x~Q1O_c$qG!|5;`}?jh#WfE{N}{VN
z>+>>XLZSi2jye8dV>uQ=1+5oB$2}Dz7HtPfc2p#l9k6j~0b*!!Wa>IB(Pe)ei@<f6
z@AyS<@@?fEiHMHDBAHhl>uh6wPoni=tL4u?w>yxFa>OE~86c{9_E=4laZ$|A_53#J
zPZkPL0f|XJYv_~4K`=KYmRc7@NABpi!KD(TraJp-FRKm_%I9|qkIuMCjJ))$Oe4`e
z!vjcmXJyNqna6cJn2j^5-N3|x@(&Xo#41+wTIXye-DScLh0g6W?lx*}@fBDIRj!t|
zsi~J`FAObO8tG`;1_E<e@FipaSUFg+Chj$iHH~0FsGmXAm>wulx~)ODkYswS+aT~`
z>-BQxcxMz!P3FjI0^%Gi6E|4l!RmtR<Sx%Oq|}7^eZIz3Bc;hwsmeRdJE^CQ6>{Fd
zUhW(XESv6{k{%K-s8ZNz%fpbMPtn5*VMZvZ{=;VRrYI1Jto-T7&zdZK?5{mxmM8Lr
zStU`h$`8Wd)ZQd1Xq`W~R@9`jx5^vx)j{^47qf!ecBK_o299E+Bj0J#aH4-oEVX;C
zi;MBwxTwKxs!*azPKvLP)3=@dsRd<Q3_r!1lSysJSPedyocc8f<TXuS)g+Y=KyWI1
zKBi?8p&1F>@4zWf&Tgd(p5Zw3r}vwRI<Ze;^ww*GpkKSf4ZS6#VpvJ(Eu&Y8RvyTB
z`1g0kAI`i3ftcg82;ugPoSUmUj6a(Sb*poDf2h{yb-iMbt45E0Vy=e?BeQznn5%FN
zJ$10bm{H!kQd%-4#pPQF9oO|D4?R$f(@^iFMFN0%;d(drvT)0XRwuGKo1cRN!~4tg
zSAypp*~EoO!<f#W9w1e3UronPE0??s(yC}XUT_Pz{UWBCtjV25dBo+iM6$ktzdMN2
z*fy`r$V&1WbvhOyIqmp!eN5fjlwU`_ey?))IdZ*xx?e2s19wbeE+ACJLRM6(vgg!u
zpu35B-+R3nE{35wfK&EZmfJQ{=VY?^-*Dsm6jIHY`Kp;#2l5QRZ%}~C89kQbnCwrm
z6tOi#3QVMt_2OGyNEx2C*cA5I1M-96evhv^sGI&;%S@Jl)B2fVH&YMXi%1C0?Pgrn
zmRKC@E{Zi?UjLb=Ws&;Y<Az7L>)ZQA=SBDhu7@VB{vYo5tW5uQzyDvdIu<6T|CtSN
zrIt)AUMqapboHN&IzjQJcVIY()6;bJxXi#-a4s<M3%n&ESErF}L^36`B4wB6Wvjf+
zdFpUts=Iv=vgYc{4SxQP4gs$gtafnZY#7g9-*r^(M7zFZhBNJSlw=)B>+j8%b_&Xt
zh`E?A8dTB#3Vz)rUhT&cg~StzH=?D-!v?B3z9E3JDHa5NoT0mIrg=}F2`89nCvP)@
zBB|NmvZDgupPvkfDCyq6eACkJ=U}P!pI+Dd8E|DkT9Q{0m4V=!KkaDxg|T0nrI0(|
z1%<JW3%7lO!wJ9cAIEbAzD1xlp0;^aZPm<6lsCE-b9=)KOrzGPUvOQcq|W8SNuwb*
z!Q=n>PGHRS9AEaH0kyY}i2F_zO0AgtOX^!EF)|&5+kmn!xR0P&G+aKj5L)Xz(nZJ)
zh<_!Q?yP;FE;XyZ0={mxZKx~WmY;4+f4u~0xoo`y{WMw_>TO1xT}xI=uA%mWW}0mJ
zaV)C5)>fR1_S*@vP)84E0VHM)C+=@we-nUxE!xDEMD~gutN|yOeWC047L-r6t`7>k
z6qUYoQu~H5TTUh4wLIi~JGXnct-9un<h{z7Aep_7yPNL~NUJGb#X$!%H72s&f%(nm
zi3yhfNcOh0X?U!8=&@$CB>Gl4*z-c!yG<L_j}#&1qKO<w9Txr%kSjnE9CBq*v<A3K
z3LHp;eUE~*rH+BqadZrrXl^=IEfY?+nz2C1rfkwJzp`|PR0oWaL=iZfth62zkN4FI
zQ$~iwgc&t+Y1syt_MGDAG9G?qBb$Zh$FwZx{4Oom4w(&V7L=00I+}7@yknQ7%FfD~
z?472S$xzn?2kGVNn2z4HylP@F{`8~jb2!Z|D~y{(0P<qT{d&EF!p8tibyI}))15o<
zu{=$z5Kx&JKQ`D-$j|tIbdfb(dR<nediWFnL^H@6?=b}%kF~G1!NiUpS&H1Ug>=i*
z%2jPmOaMLH{EsWK>3HRx`x*B2Fd+gM1p({YAsj=X0}4KZah2$l2lRCD=}E06<(R+I
zmN~MAib?|`z<G0Djn@Qz96B}IY)rGj!cb1PYNCxAFS`cUEIj0Ai1fiMABw(DrX@h|
za_vjGRL|l$F%O#_IPa)#>YE37Jq1~@NpxyIMSMBY@j8?04Ti=%vhjcX!Ywfbc^+YI
z!I&p<nALhyWj-U~atsO3l4Fx8cuC8?C_O6vP0a($9<$<ic5{)o=l)PAY<`H1SgI^O
z$oA2A$q_)*r>1sLNxRMkdkTxKm{MRLZ?+N-#&Z^)&Ya<*ChVF<EN6+ZV1}uQ0(v^I
zaWeogy&vdI<G=Mdc~=`&>{cpiNjMqF<TmZWG#`X6YO&XRzR4GVF~8r6lSpvz3&+PB
z2mUkm63n%N8~qf=3?$BDSN@|2k@sl66mF+phGzsPM34j2@Jab_?kEY|&c^kTc$4v+
zF_^UnA1=E5N2r$?Gp5-gqp$*h#2Z~%pFW5CAJdLpD#Ad95&dS~+RLMcya~U1CH`sa
zE)tqQ2?8eF$qqmUb)3-?3SS|=GHt<4i;Y!G3$xPbmgX&al0=!8N0^r<rpHW=(8JAm
zosiY{$s!sUg}fZak<XeYD$Y~rLUGl(7PIjsC}S?LD;-dePcv|*jyB-LEDe87Lqhln
zC;kT3!Yfn}MR$As<>h+PM~Rr}?M2j30<ii-(uMpo;%P}44{=Gdd9@5_gP`Nsm=>qg
z;dJcQjNHG_s(X~)pV6dS=yUDXVh-0Tc5l7PL=I5QYVAtL#=g`BzwLubs1U<nc`N21
zyX8e2_kzpOGiL>b<l~tEl|Vc>IoQEtgb@&#5?vUxP@`@#^`nJiKVPUMxAaO*2bk(f
zhvB{$qYEE<9f)-oJktv0$+z<UbX*~p!ZZQ#PljF7sd`R&#t^)K*Sp#=h>7NWV~&+A
zucA_ua{TqmXRBZ^xTw1$WueCX=$I@xvwhHxA+|P{mTB60iVh!Ctc-LiuE}IlguZsc
zT8+)inwVo!J@=cQ2V4=P&cy??mxWs3T9S+lJbGJIz65Th^hg&zwBOQZF<{#2-5D9g
z7ZQEFF0)7Iqp=;=G}3s)+yJE)JF_VJ*FDIk$13Ek8dm-O0Y*?JWq{PP@C)PnFCx4f
zSFH`L3+`tzXN;{HpI}olaUtdDCe8i!kl!*C9M?{;+b+^3Esj^aX~Fr+nyD#k%~9fa
z3o>J^tmWk>R_=7acp6efaBjuSRF4O?R{OXBUmuK2<;2yD*vYk$6E~wCNcO`ALG3aM
zvjlxLx*TUuX<iC}28siSMS9H$Cn+34t%RGKTdWgb1a`{&Q{^)C)jx)C)fB6;5v9RG
z%5*uBN{6XaT`p8porIFk!{afJF*!q3YnN^bN-FA8m>&n+<H{>mI1q;;B$fw4WN?m&
z8K(CJL6vh79M9;9Xit7%leTIs9)VmcB+a99=yJtMsgX)e)ytO4@|IP~9?)C6HU>zQ
zFN}&3%)qIRlpS|8QcNjhHLCMfE%aZNP=T_;2JcX^mB3w3tTWzx26#X5&mBJfx;KX~
z09}?Rpc_rC{4{5O4tC*@%W!fZY*g?w4Uuj*pC%-4hoDX}K~-?Qf5Y7|?m$LTy#~v5
z_<vY?r{GMZaBDC&JL(u6+jctaj`_t#2VZR4wr$(CZQC{{bN>I#oSHc`buOlA-@f~5
z*SnvEz1HKCchpvum82eUH0M3=q>O#t`PYRjgQz)GKHQzQY7f~-8%ZZXnCNvD2MX8H
zQCvw}LTus=a@A-zAbaqEe-1=0!5BQ)5cAVJzUtWBLiUI9?pl3foWUa5QsZ%av}gUm
zU|9^!Ww>jqPn6AGQWGEw#KKhk$sa$@pP9w{*4_Yh|AlF?EYvj|&((UkwM<tbtMN5p
zSH?7V65jx?!?;k4iD+;G8)fgVWF+8@M3DsFpPLtknT+MZa%4H^FXeRy=dfc*pcRRY
zn>dz@nKo1{woy>)h$6e`6SnpWS;<;LguLL(rjkXnJI7CMDll}sIeo4{_&N6<I^d@z
zpszaml9Lf|(9IEgo7brYE31-LPAG#-ZwbsYGz<ejv;qTchioAQPaJy@mO=|(r8ayn
z)lTW(R~v1wHLMg*rJ%YJUZYChIA>yoOzrq)LT@Q&;LJQTh)UhYx{43)$w#@O|CV4j
zV4SV0h-IlQD=IQ$9&~2uNHdfsoM7VyI5;l;>A}%rtElvAyb%(l%%Y0_P2+O;mch}!
zRrVy856dILV+z<!U~E9tm))j@bU=CiNSM}{it`j@D|DuLn;C({_hgVW^~m2zG;Jns
zb7&~<K$CCXSg<`-9m~W9=SZ-{=7mghqxWcFBMG<^uDI-GUv(CjcbhW2rLH{n)}k~;
zQjSy^DB5V)O@Xu+0*W-_92hlIgtTJ|rNZ24@U|x|cL4=$C;2}8aB;>O2(f?$H&6jq
zOFa`1MQFIVguKuUvioDmT~r1+*Q(KJvL@qk-nB(~T!%Jivtl4vdnW&w{YGhSQ{A=W
zyZB|gHa7`9S2hG`@coskyPRVP=Q$+UcpH1inm=BrF2?$bQfR#_o4d2`H(Pod6CnJW
z<kCR+meO{irOBndH*H~raP?lj5rnDzwmQ{zxeGvZn2U?odk;$#&e>-Lr^jLuM&DIO
zUbj#@C-;nOJX()R`F4nlBY(Qjg>uJ47W-R;Oga^px1UqZv;`%;Otx%VIaWFxTVfC1
zglN$|^Y@MMJccd!+1v-{lvU$$-}ht?KxH?n^w*a_WScL%kK#Wt*4@_4GJPJhIQk%+
z^pZ4K%HG$&)NtTjON=|$#j<=ztyMSajH0c~|2BTy0So83ZjPo)>C(XoK*+Q3dY_YR
z?ypQDRofjT^Lh@!wezCc>u#7HR)zw%R$e3|*KM!&TS;?fN<Q)3nnMV6DinlsLR57S
zUdk@F7_8%b8Ur~kF4SJR9&bpcs%rZV(a8)mcSPFc{OYyX?yqc?)dc5#I+=HxD5WYh
zy9`@8s1phzN@wpoNYfkgme|!+Et!2v!CE`fzT5pHcGqgAvI)PW^Ruanmqk$2Al5lJ
z`-vv8#e^kgF}f(pY2W=Cu8z_3?Kg~<))J3dh9BqAGcL{i9d9tI9c`-^{@(!&*;b2#
z^%w72(b3S441sLa0H%WQqYQ(Mj(0lo$YTydo;~Ls=Y33r24=Bm35L9#a);xrF5;$V
z0mRYR-v17IF#o?BNSN5Zon8MEeW(I368W8W*zrbl4_>j;2@(jxa3N&-XImqB4rHVR
zmGsL?qNY?L(RAdpXS~M}KpIx1U13;uawRFPgM9S1Gd9smH92x-TKJ|D7YGXJ)!>`d
zBP)N#rN*}*P$&>7l>(B*;-0Rs*h|rC=L_s=2W0352BKCZH67chc=<3%-<1bN`8XYL
zVF1Q7Qv8_dS&j7yCPofPsOY^mz4$1hcs#qdkiQ7(m0^}!MxI&S2#dDo?=SYQ50a2s
zw7$6B7cqHf7X4YpuL^a?%Y`jg56N(z9la+r;%~iI-v+kZlvpFB4L&>Rfs@U@%p&fT
zWBO8|XV(Kq6We+a`erAjwir0F3VC@9l$LA5Wvw2GRm%|}w|yBNaD-RR6J9(R?dnTV
zZOKA(T^I$5L&fOtpXEG>nJW1K%VQb}(^#}VSw6y=(6O533W5Z)09Y~gO5U<^3l<yF
zCauhWuna&%6Olqa1Wjm3MU&Vn`up;mmdW4i7+n|39W##>q@^T%7=X%geU<J8kGKG!
z{T#{d`#h^;y{y*Ci<yO3;#6LMJVN8SfXNl~n=9LpVX4OwPA^5>UDHSC-YTMObt};|
zh0mwv=6BHzdNQ%Nj83_*`6$EuJ!@p(FY~gHAd-lq-?OnlD+q?&-FjvM93&9+bR)~J
zrDYvSUnu)n?M=i~!4}vNB%SSJ3Fr!x&TdnwLmbZWUZnZbEcBiVfiA{`BA@<MHJNz7
zkWJY;x(m5dnvWpQC`qI23Y~o>(h*X;dWFm^psu62<}ExVKs!ZLj9-ys*B*zjba*PW
zjS<b9`^nzdj0VZoW?K)izAeu6QpI#DXCN@QlpRv{o#c*H#exgz4Ki<+y?L$;iK<Ba
z!`KkXYx`TpTt=!f%VW8g(mtKQpQUN@r#I^<d&bH73pdBq0{_S<|H}D@-TzE0d^6h<
zywd1DsU4v-i=s0+#8aPSc6)0m1>Y=92(~DCP@$V^9n7g_t5PBn#y=`e%`nS2`Ww}y
zy+9gH6pug`_^>hbT(Mt`%~s#>4}S*P0Ydd^5w*8bQM_74eH=5PqrSJl1-}-~*tp!C
zdhB)U27RR<u~eH!jv_`>F0tyxYi^eQ3{XPwwctq}<*Fh2NMdM`lVY4-e<<o9lK40F
zk#vM(@uc3+V?XsQtbZ0J?~)T(?$w^kCqC@9Q)dcKsM=`iW}PKZ=fA8UhZpEswpjU$
zx92xep}tbfyj%*z1b8(RE*6s0U7pvN-6WfI)6>Xb6s1nVRB)=BbEBJb`9aI$&N&?z
zV%?WJY@WDhJx-F3gO)B431w)PQ!gCr>D<$u(zSMKZn&Pu2H8Oir)H(_&6Mco|0MgB
z$Zd%&DfTC7A%nxIEDs#(Sw^gN%7M`9^ku9Axzn60lJUr$It*<!?$`~dutqkoiC+>b
zOWZ4s%xze+pmBzREGgz*TX(KXGjMb8@_B(gN%J!WBV|Q>^IwOH`!Si3Ve`{Ks{_`r
zy%u%8QR*fA5{LMbTP&U{61WLWp7GRjDXb74NdfzY0Lh^pUJUBo`qOzG4iU3D>`?KM
z`s?gczub?1SwkBP12=bjEpskNCpXTSYd7obuEMyz_blReX=B@$Gwb=3U>RN)mlBi0
zb?@7n1NFSM{9^BNSH%<O`Zx_l_W`(41_9I;DctkacyFS@W}?*eUal6KrjD?itLzun
zeC+uuLeYbqxp|Efw|vXmDIX`A07}?{1@FO)PFV`mLSE8L`A`YW@yjU@q@sSV)Mg_i
z!z*e<F0E-4w`oj+Jfn5r9P^DrneE)VG346=`N)c#!Lp)5{G=ms&I<OqW@!QoeIs*j
zRt(@KbSLNlng%1uMr*T-f9*}3^U-s3xi9L?->&1~sesSgPF*2jdwq%-{)w>q5V`xR
zmM8(tuSs%;_RE|mHf|taiM~s=ywwTH+1qT`v}C+r_7F4Xcpa1gFFQObIBF&&526tb
z>w_z0IkMuecbC?%Wxbu_hqZq5#!N-^dMg5U-UZiiS-c6h5oJ_J!v&esHY=-4k|#|<
zBr{hi4wakXocTp7J8(U++^Sg}vhkyGSnK$5s-m*(32wUR6{Pawd_9%k?&0tYcJnIe
z3l!=d{{6qZi<th4SK<E@HW>b&vWKMQx5UAqYG!C}O2ov%$RJ^4W@2hj#LCJbY-VpK
zXJjh^w6q3V8ClsAaWaSiEr7O))_Mj;|F>?zpeU_pXa22Nu(1AD#e#*E`F|I1{(tKh
zs}~w8ig-iVTPH1Qx-A19bugG}m$(qYibT<Z!nOYX^e8ZTq=Hby5Vc0cfLg-4SS9V)
zqIjL#=w%zS5+wzfYY{6RWt&SSvM&kGE&HqUbJ#YpfrQn{%8Hbe%&MX;9WRroRg0P*
zH2-UTHHk{eB&@m3r*J(it{E%KNZ0BgEq(2=?A1>};Y?x}Ls`<<Ezh_fIrDcp6f;`^
zx$XQMCnJ_-xV1<n5=dv;mZa=@y}A=(E$qRK?n2J3=B=s%OGf)42|tSp%aVX6ZG@+?
z{BY{uAaPvTJT5<)Fz3wH02gKdej#mav^zOIQLnMC_s1sb0R}|pM5JoS;=RFUQ7S!0
z{cd0Vfwae4K;u2geT_`Mf4?A{%gw9TWY)8OJD+w8xQ4mHkD1r{%J8j1Q#dQa1$5{P
zDVtpU3Ub5BeTel><8nGlW#I{a8Dkxp`aE1pH+j=5*a2DX1>8TIlXMxcc}Ko$XR`4%
zR4t7m$_u%p8y~0Zp&^m3pQUt@1pYw*1*;!JLQt;n3McrF*DC{rwXwH?aJE`3z=Ya?
zS3}5w)sBnpt*#F+5J*u_hshKtbjs{a$}z-WFzvE(==B2*mY>?X%7aBNvPPzw_nI#K
znwQ(D)V~xvoVC1!)lA{dJ^#o{*_-0ZLdH04qKUMBOdF=I&l6C={RGR0n20DMDxReO
z*K~Ep!*By3#T-wY?de%6nzSOH4Khj%pNq)2UNxxRmrX=g;4g1P3h5q`l?JP+^_hXI
zUek!k)oQM&bjJm9@EZm&aEBOc__##CTQg$&n^G_hHr3#$omGg9U+D52kva-5K=uvy
z+KonVr;8uXE91QXN@D=v^Pv|$juev%f>;jp64cB{ol(MAL}XLGQ}y6BXs41-Ag0xK
zaY9aBKXe5T>>GSn&9;4Q-*)$u|4rFljxy1(A()y2C$6`k{JhsAR{poGinE;`Zp*ad
z**162L`2#%gKf~0YxaF1WzyuqNIqNe@F$Cwr>Ww?#;O9n+*k6c)MJCt)@AXobQjMC
z)bDs7`O1_4?a;a~Q65Ije(im%5|+E-ZOh-3re!?!y!6@v%~AdG&8h%vZU%5Htgu=n
ztbmT!S~gW;`pLIey}Uu|F$~CTOe#21m~@Issr60^TIhg^)ZKzmcYy(p*tEza0bW29
z60Eg=lC;JzOezXl)QodpV*CipofdR#I2w(5;mdMyA75HY0ep-12emZ&Wy<`2r-T&T
zp4PRSJXDgeGlG3tQ;xp6=#;}0PzY-sgFr60oDDf#tVD=D*!VUX9N~I1XJ%2~CS`}9
zV6gLekfuf|1$1jZt@wbmv4NGKFtExdb1CbMw1m;LK-*cHHhlyfNVwHr>WQ2F?@fuC
z(fZNw#r?s`%2DYTFM|Xjeqhln!3u(IzA*%*<y5jo0k~+<@k(!>Mk=qK7Iw_#1TJkv
zP?+u6c;U)(Q)%;xQU@d3p33i?(J8xkW;nKvR?maw7-`?QvZ`%d6s;}-qwG=8<&6K>
zS-bTZ;3L69#`E>TH5rpO7?bYl4{7yg1qt<3g#x@IG~Ddd;s_DpKx}L_w%spga=anI
z9r`V!oA}q`G+L2-!OFo6L4Wk{y^H)N=Fkp8zT2ogy|P#SGO9(}G>E5UEX1{0T*85z
z$=()y*kvPdh4C8a`1pYi89#9A&byORAo3d&(*DyU@_q2>RB9qfPp%u1+*E}H><3QC
z`2(v$$lb7f_3z`+<NVG$DHJa41_rq2)vc(eFT-loS?gxX{qLuzcBOB?I4ww68)zZo
zcn1|=hjiG#YkStkmAwOk3<o6at~Ojgm+H2+TyNVZ-+nztiVJIbL6-%!$bTL`B8pZw
z7{Ijqo`PW@3!Iki54RZM>OHM!<)_4_kbQ+yD?jzxKk9q}MZ}4RDzVfdh`G`8<CQY0
z4v@IScWT-U7nF;i$;-*%BX?dXJ~E4?qvjSJB_m(OJ}D@J%ZGd97*qQCv<1-rJ`GmY
z=TlsS3G0S{L8>cZRbrRS(@_!M#Gr;}>mQIlA0BlrTz;Sa1MKh+_`s^$zqLbXsgfg^
z;|gBamU)GaN+U4{O)A;jr5<$XVw1o9!Fb&v+I4=8cnP3)I*~Xdr9gY!9k>8NAgSm?
zw%o<I9UXnB;QX=Jh;_pte2{^ms)kZ=wwX<iZq-UJn-G?K5VwsN{w5zhWa2-P_@&~Y
zZWrT{A+6s`^H&Cf<pf=}uCFU71y{ZJ*YAs1GD=c$ma~c`Q)lBgLEz>qIz@lBrMOsA
zom&<EmW<M}m7j02Qu_t?DMeCy7n>_g_-=1`f}eMSVdz6}Ege*Gq#$NIeO~DZ)R0R|
zX67D%`(T6emAfP_VX=c2C)QF$y;heO1TZ6TkZu5XZmh57KB4Ra#UsET&d*t#R`wHR
zaN%pUhC`(m%9<dxRSzp*z5Y2Yz{z@=%VKkltw)s4(lt5E6(V>Y+@>#(F*@RH1^|RO
z)X<D{$JeK0cw9Yx8$4m46G247EV9k@4H3fZP*d}9eslM@>7JDRBj;!aEW3A#bW^8e
z*Slg4Gedk<ygu-!GB=lo9iZr(xYGQ`xL0L!b^W=iuE%J9*A2<<VQKRSA$-jgllm;w
z?BIJLr8-(sT+trsNl>e(@BE>BFba2w<kK4VaY3svAkmN2ejZm%I@)>!yjGQ50vlVX
zV|ObyHWA?Z3B;KJe1F_=hI0i!wtp#3;4f;)biPU61aDzxmQs22wVDLY(*$aLc;a#g
zQ%Zt()J!LG031B7=b)L-CZrzj2r3GQCwA=nq4}XPA4-tg*WwaVb0Gl+asX30<j2Z}
zc+%CGiBr8^)B##bR+V@dnO;&(H3$4f0cZNtd)f?_<tZRjQfqud-;j|@z)qlw(kML+
zU%|S&B7|>rxNJ5|#?$z4{eV<?|9`A@F&`?Q4ODRm>^$@Naf?u?drq4%SwW+t5Pm=>
z1@VRB<^(wn|L-4o>lPtvhJEn*watj;=Lq6}=mdpXf60e-n50>^j8v-@8NVtc?KRd3
z7W;qK05`|}d;`EP1`bCg&}7?h;4B2mS?>L<Am303QI@tYadvhB7={iOde1R(BbGfO
z73Bwp%!OJ(1ZGrl;9e%o<&daZ9yFx}fJ_z{0=APHTH0u3Xb3%*q_;>|4XiZPH{d^v
zTE5Tj6t^0G%)oOq2N92Ic?S~W=isBX@mbh!2fm3*l7_prhbo|;t8Mg#(RM}8{$69a
zF0r*MYVPbL<!GepL;IPeUp&aJB4-f}3Fj6(<4$rDQ3TUBGgo*vzTwR&(_tUydoXS>
zdU(;E8xk(_?}>2EdJ;}WQ!K*>vIcj`!W6ztOlxmQ_nORJA3I~mUQ~@yJhOF(<NM~|
zf>W>p&G6`G8JB0=Ft8Tsa2!I0EJtxABaY#sX>f2wQUe$vCeqtfm10tMPaNuzgX~h`
z^%sz$hm(fe{)?5V`$<_X45SP#5dJe`HxXC)eL=Kk>t=)S(}?UXS~9d|v>UR>a0^u?
z%uw0zv{5$M1LOsvg-V;lg%1N&BEj;(YlumMF%o9<lmLYf(~I~S++c9lb!Pn|f5_o%
z%{lO;x|}oeKpt%~Kq^Tz^#^2V;H+a2^9{jzHv~7Phn96oI7`o@(f2cj&>&k1=2FLf
z?qqLf1{LkEh4kLQ{Rd%c4a*I|AzSiPDF>%1O`BAS>jo#0a=giMHe9mrqh9wbowQR&
zl!bfLLibPoYE%Hp8pJoMY{GhHEseEgauSkx4e_3TrT*rE%Y=hM9^7-Cwr^pFgXe<G
z3~FZUa>1dXsjq2b7FLz29{S~OM<615gCR1pUSbrVYaZ;wP&;3i#c0nG?^N*73g^UF
z(WkA$hxGDX`BHcbn&$uEXox79e&qK(Eceg0nU+SU{plj-d<~ILD*i(*yYJCyQNDuG
znG2T9^9b|+xevDuJP#y-w39;cmk`K_VJ$|0?+;>l09W*JH6=p8tb~F}<t#@yTn@~<
zgf7oKa2_Xwv++%8bhwuMgVHY}MpQ@bl370cD@SQ5XjY;*(Hz%haE%7O<%Kb<8mlTV
zCYx?y%+5aFT@T}slM^m=(7=6_X=Gd_aOzp@h?A#g0|2Wf)D9-1Tlgln4UFcea{K{@
zoZ0<%?u$)q(iT0u-REEa)KeXND%5#(CL~<J!F?|A1F?L!z91oh{Q~T#krC=*n(H1t
zElzlq=F;!=&Kig*aSe~ZUcuG5j4a~$Kldo7WA+yCYbTQ*7VT<XA10i`YP|$=@|K?2
zF>BIN6LTiED3n>+bp65#hI*z&I6@cK!EU_B3><QhTLMCmc0;Zf8I!X&wgy{CE|nK%
zPbPw<noR<Ne24F8@z)!M(!PA)URALe;F6nvdid~5%C|G%IKt5~v#>WGLIXyMZv@mH
z5dWlbHG8X22>f&I<uU)O=NMy*2$}nDy5k-heO+HW73usrv15IvPf`dXKb_ae)b~x*
z)`SpP0I0f{4ME+5;k&XRzzRuF`(F!I=M%nsw(IV{1BCy9(t!H^;QrH}_Wv&+6t~rL
z`A^dtGvogh0xr2JDT?EOTr+UM67z}_dm}65^Ij@I2%Y7ENvlR<HifzB+Gx!y&{P7e
zl|rlL<{t6j8?`noHcE*S>B*)QXrdmJ<<ZYha~EEPn<OwGnoRqC`R}i_MDa?%I6t0e
zjdNL>98PAuGqa?;(08lh{P`-{k5^N2d)z-t^Z05Sx;WF>wX;IB$#*7=WZ+lCMcmeW
zs;x1zhyZ^foYR$f`CH*WWw1Z1dqFFrF+&J<CX4qsCZ^oPj0$*1N0%W?-y85UG_`O<
zCefmP<O7eMbl1<8&)FPvWsqw{rkDLm58@y>WZ#U3c*3I)-Wjzz2ORJht>n({?b#ry
zyqJgA44V`M@d>_bUTq2P0g0QsiPFh+2JP(Kc>1a09Viqms-69joBSsQgoe$T#2-jB
zrC?CtKf<Lc-hT5m%i>oY!i=@7Mf^1S==lJ$bUKt>6@nvaE!CX>S#%E;vCg>6gtO5m
zyvHlb-#?XncvAewsI-s!hS`W*EuvL=oG&fO-?fJ#HAx&Ly~nz;d^)2BCEU#$sHjkq
zoc4{d(O-Wzg&R&8@GTKxICL6=k+CCbg-n9uCLn^H`Q8)$YbGvkOZ-PX+{eA6l5h&e
z^VX(Z8a7|%JVlV-rH|9~*jahYgr?aoI{{h30-aY5uMk(ylCpcZnjdlKYs2UF8NJj%
zoJj`>EOsL;oZO){w)~iRhBfw8RI?=ZPL2Bo)89m$qCX&w3*SS{`*(0BGN<Z^AY@0@
z8P5<Wyoq{vz=siDs98o&3#7_C6CCEJJ0dNF(f*t8r<{+SPS-}TFA>{g2gk@?B2JsX
zx>`gp-wH^xs6M*$h#|8{Y)7yUCYY(%7b;}6ieGW1WPB<dkvhjzE6C!5n_J0}kC1n&
zkB^TG6O8&ypFZX`j956W_pliLN|gHUUr1j&Hn20Ww?^Myb{;g)I8S22N3_T$^m*N9
zO0aC<uhqz6X%^IOnpH;YI&$y&P0zzzECl+<&`^bFmo{$}S_t2=W1duR(v#PrV%P)7
zlX1P9s(;ZM7@UswoT)GfG&K-k`ZY-Vrvz(7DABzjmwS*v=A_@2!49B?#@*Jw2hQ`t
z(cul(N??>jYlUFnN2Wy`YZxyOW4P5JCweKPA!bAISM4PKkFZpm6-Bt&>yQ!<?S+nE
z*x47TI0qU8FnR!tE=MM26%Yv2cRIkoxELytcK!o~^wAtEKpz*Y+2O#MB`}2Y@wc^a
zNCcwMDT*0P(kbUN-9Y9-9-_7_`44Qs`-voTC_kNx)N&w~rZSj*uu_CZ%dEyS5O{7S
zpp8r7-M}pdGIY6YeSjD7Rt%Or7e=^-qe1KGyE5MJlhCEMklXznW@|@Fl^w?1z#w(0
zsMZqPbKpG(Z0)UFm&&+CM{;He5BfViT91L9VS7kdss>ZQ2>C+}vC#?%UR%NMnZNr-
z|5=&Iq`Dh!jQ`)0)=H;CXwmayVYUhQXJTKvazSgDsX}@g`i1q3XI8LehwsN!^0U2U
zDGz?jZ2?31W-k)w8b4TIZy)ru2wL#B#;{n+UhpM5qzP{w%k{*qKB92u9nf2Ac1o@4
zLIlU)?7<7cs@Ghju4<pMqjDUD{8HTq57^Fe3O?OK(jimB#?%Nqsxr<;J05mkpY2k^
zK1*(7+<6Z#1w_Xg=z%FC#D^39c9EN#^V8tkx4o6Sr?<$7t%xNK9`w^9*$gneVBi~w
zWHC$kAV)P#={2mHw#YHy2eNH`$%P%d)PU^sSlD|s@9h<$+L>8TL!P_556D46(2FY&
zFrC${m_jqBTaL|?!j1mFbaCY4)kdueKKlUc0<)G4PHP)TJrL<0($q!{=<&7~yP>HP
zhAOiX6P-)U=A$e7l#<F)&PGJ$$#Wit$d8t5vEXXAFM^nzl;BBU8sgl_E+g5glZ`IQ
zq7G)8fr|7kw}(T`5K<={)6F%BiH;PvIl621_L=lNIaUQKZ0qf`Un|+Xse1<Wl*8T~
z8tF=rHd0eM!@!FklXBhk(Zi-MfAgun3+x)i>q`Q?rlb)T)QRYu!eVo<5q?tKi<g(`
z-9I2zLA~R=3Jo4u`y&GG(4|<+zF@T@yJ1w=jeN{qm~xf1>He{-!yA|LEId!(QxrZ^
z-5pn6eK3W^{`51~S47P?LJ(_k0UPkiL`nSa&xm5vKkkqNivKXpP!_{iKfF$rfB*T`
z`aU3j@EG={hlFR-3YIs&N=pTXtxob|#`AoX!wf@kIzJkT3!BdlwR^vFV7SK)pR?WE
z)M590WE|y*Y6ZZG^*ofrYV$!e)T{v{##F=vcfm=TlMl)%tElxM^2GpcBD(aP-Ec~z
zm<A>}3p=C8<8SR!KjnJhdP8*GKn38fMn+=RX=!P2jWO{EncK>d`8p>Uyl~y0c|K_4
zgPx2y%Pz<lka*yELP^CF?R>3pQR`PL+eXVDC(Xi6cYf!#8z`KcjbP#WZ}5}MIS_3o
z(43~yGQ&~-Mh8GEj|+A@7Sx+r9fC^E;fcwS=As|?4kDZ$=?*+)|3w9a`@GAE#PnY>
zx`bJmRXxLuTHwJYpFYPq)DhPo&buaSX*7sSd1$?~MC_!TI3&FB`X`KjZ!JM7j{-Ks
zhA~1Q^$)d9$|6jrOSUDR)eycYL9n+8!={Iam>g9<iZH@oYXL*Z0C!uBS^5hyk3iY@
zzZ*qZ{|h($pUH=eZ2vQh=79QGB`F2M#(yLqg8KQys?$AGwBS>XnL5fhj!>t)f06Kp
zGogQ^{vORO0;>AltPL68pG?K^f4qy%Jq<lE2y8HBhASV7To+w=rT=x!o9hyT#^B3r
zhca?-eONkas$1M{a3nn0s~1b-C;q^n<IwI@$C`ulOWNxC687)#J=Ro16+;+>T!PUc
z36JIv$!;r<HT`V^1U^ghvtI2`TvzZ*Ynr8ff)a_9^A!A=fl#~IX_U6~GysLB4rv^z
z*W%wg(%Jdjn<lJ9)BSTNKjBZ7=}hLQ%qSoK$HNhWJ9FXF-~@$9spzS>a!FsgifiVi
zlRN+En5tbRj*X#rEsLEQGvwj@VsnLkn6Va5o|NF^bTc?%JnT;Qh8Sk|B;*Xz{tp}1
zMpgO@VC{_z%4FSK<@<;9WuZgi3P@FcR7a)m5{qBLtLtGHZGwLj(@J0Nclgkov`E9^
zZA5jlos5~m)r-_zhu!$do9k5lOzI1Q;>Y=R15-eOP-G@K&70YbH@=J|KPtoEQ4?YP
z?xF|n-P#AQc{<(N1GPa-WgAwTGFB%bR^@WTyo>|!hWGiJy7D1{iUTpI2b|RfOUaAM
zQ;=qCtBG2q=x%EmaD2qEL!_uVX3G@3(NAVzH41wC-G*YCaOj>AE3ul)KXNpiG%3Rv
zCXaUTm8e!BpS==u;mL$K`N650+&Ppv$H=r-57<9yt$dzeT$IQ;_Micz<tOYVY|8Sh
z-@-@Q$k$qIvuuv*kw($+Bd2%7kVbw-sq2s|&6+I{Y>{Po@<uP%1#r)n#Z(1U;#O^t
zgh`NI4u!iiHm@jb7$AGZf{W4(C*}pas4E`2{!mr#a6H2RG{hH*dPe`?!fs*<D_nkU
z9G{0OXKv0ohmtbcVz~7*)vRjF7FK7gEL86SkRiDkOv9==Z*gfT4MpU4tAvqX-9{O~
zX1BWl67?Qv!?4r}t?h5vO=$V%z3@0GUZmE0s-fGXEB7y#4dkEY5+BwiN=(gg|80qV
zS{5o#qi*ma>V2|T{WDejh-}aq{(X?r6^2N=?XAGiV8t_6!gD8dvr9>B!>Kw5-P@WO
zFl3s-I*$2^!Sjb=wJ3(ZyZqSRssE0@=So-TgoYC9S`aML+$5Ehl%#e7XV8hJn=cAn
zY}Jtt@a~ca@`sji^4y_;UYVHr$Vd}Kfb<T`fh9e3F&1n@=4p)tmw+5AUk@glfo!$@
z7p_@~bC$QUBNyi%qp@v+q)(i%mDn3|#h)dd%Gz$Q|Jq+fr<-y$p1QmKEipQDr-jsY
zXqzqtZj<J@{SC6QNIbn~Xlrh3!7m>(wdH5Y6YCI&6gE-5^w$pW)pKA{k2zViYYQIK
z`mLLfL@G0-ei?iv`pcVHGhs7gQ16Zyr)MV`#Qfd+E`I%_stkoFw7$>WrUH-Jv8s$>
z$niiD6Oz9!;g8f}{|7|e=u0lPT?YB0f)MXlP(=L<b#_+6cNecdWOLxrE#2CQsS*fF
zRzIE1=8MaYjZB?pU?7#s?@7BX$iQ-aE?5IbYvE{T>}6ITuMw%XrlN}8SsX5H7JQmH
z4U1(TZr|lJKJ}gCN@Ax4eZpIfUc8Sa`}*_XaOm7_xC%6I4u`Fbw{LsVrOOq))bE@d
zM=a%+2U;I9B=!3p!x%uyIB7S?s}Yb=kTOsqg5aq=#Om=qq;oK+O%ONXdUbrBS=2hQ
z>kk5Z0M%O8h_5vE)@om6N1{G3Bb|PLkp7zvsI}%Snun9J3rdmhduHNa8pZoubTz^&
zw|7IPowtK}$H$DTX&q(K(YtJTmhOA7NfjH7`{&O{V&cp?UYd%wYiC*E&hMJ=<Hhc2
zH*lALmXRM7;qjZisazzQ&9eA1lW2dq*-x&Qa)1+Y^?*iChLE{${P;=vljECrvBtbR
zKGbOPlh{*wJi%*QW|!exm39Q~<-c(TRnBPGm3I(w*+!0<;2ZyvcPt%%82@9TY=&hC
zft23qd-u4B{+IrFz8xh#h_K&%?{L8M24i07l`=v)LLPw$3130`gv|Jk#Am(GXPuGv
zzSh*7!HrqO+IOj4d6bQR3RcHMQjI;@AS?_jHoeldKaRwwhS1m-6s&7@1rel1X(%wb
z_~|$c(M-;TP&<tmiNi$%?_h`ha^o)QN%jsQaZ_Sv7#8jzQdIpfq7&Q=qn^!#LnX<q
zcQf*tTh#S*&MN``jh@#c*ExN6;4(ev<ke8I3@6v8DI5X2xgZmQ!=3~jJB?t|b3$9y
z1Yap##oLJ+!k2Bk+rmanZbYJ)M*NXa(-YK!3s(E~WP!;_qw<ov?19Xo{ta#GQbK1T
zc>rZ}oC+Tvz9~i=6DY&jTx^pM|7y_^Sv(MBQeP}kX8RQ`GFg2r>sY-^DKkvS0kN>m
zH0gBG99rX&#fWXKkJWr@fKg{{o`poH4O!1%&8XPzD0I4PEJigmrJ~ZhpSD6FGS_Us
zd_$d`^h$e~p0xkxYjh}urdd*hZE%D^2W62fjoAAnw;BE#Nhv?jdvOO|)d2Ii1$*mz
zQZwlerqbYc=QY!A;ndCYV8w2|aJD%<LXM`;W}H~$=~^?-muf_4#y_TW>Qvtc`J_X3
zjng$#EDw)bSY?Rxh?4@;B)|RA=?u5B5w0KQP18f!!1Z*63gHj}EXq419jB{z!p##H
zaE|s4<a&0d+73F`mVqzY7Qdz>56YhfV_G><TIf-_Ln#YKPYhomqu{y0|J_)@{$IL3
z{&TEgXJ-5#*{>SZN32*{P&?nK?!gYNbwEVC7e3hpBYl8El_O^KE?**{e=x!<8r2^K
z_=js-TjpOEgh$7`7;bNm#xCv-qS|Vbh;kK$bojsQ<gbN$b;x#doRYHRuf??ZzgpkV
zq@^x)?EIn<W`90IJPY(TwRPfP==s}T%P%eIV!Sh3w$FT*IKLugc^gbPIy@q79Tx3W
zMAF=mM9vXjl$5vqLj<NXkzXZMeto_Vg;~(<JP|r!CAkAAgjul&%Rx?Yb*`KQ%55nz
zOaf%*_JuQl-Bwm{8?v86$*$FcX{R~I^l}pNxxbM<9C%f)yiN#wx#!U3VMqnJ$o?4c
zAi^^-D+!Qt%{017poiaW5Uf#58GP-6R37D#Bg7B4zkN-`P}^MQuC<Qu+KLf}wmYwg
zfO#?dJ^98EMKg_-v{ayzf)pZ4gMXm-Nw}!wE?soi&dgD@Q>nh=sj)G)88XY>S6C)I
zy5W|GV0_Ave*&uGuG*K69agwjTzeX5j(#zIxYrOZpQm8TatneLOm?uf8ko7M!k!8Y
zuiFcW?O{<p%&kWq1vMfM$=A#~C*N?bpCR!!2+s?f!mRnNuz9RVK~Ulg!|plvD`-Ta
zZPFFBuVO`VhTZ)uwS2{hk~qRd3scvYQ&qJC6TTd13{ORK9-SSI&zpU@pVE;V6kzOQ
z^c#~jk`m5Ix~3*gL?eunD70ztV)0N3cPPQ>9}eea7pqRhAd)9!;e31d0QKB$mzZMS
zhGCw-KK37AboQ4aOs;;Du99qnn|=Xr1umBUW*a<v*7b#@JmE{ghXx3Mhs^heWrK_)
z)7y5Mgp6$}-Tii4RuiJC?$XVZS!Q#^9PQq_k`1YO*#Kn4Zrc~x?OeoR*b18-L26dH
zf}VKWhK=O#NqP9y0k>@-c>q0qwNxXF#1Fdyw66UsdtlH=_Bb75e>s&G4BBGd2yz{L
zt}#-;s*EL}#q4nD6FoZvIvsvT(wjisFGO3i3mxx0TVNOa05DC=G;~+@9St1%=U==+
z5({w*Au9?tvw@Xf+S=e{@f0_tyW4|w*0$XfeGHyRJeX}Dwl4S-_o$d$v5{QJ3<o@o
zi!r4X0XFN!aIXk<hb`7>imk5{^w88YxF9Z15axIDR89eRf-<Jo>mc4&VEZ;VPF(Z6
z<AStGr+nLfNPI^`8G2u7-Y<BQg#fG_@KKN*eBxCc<;u=@I9rUFRAi)+1Uuf8-s7yJ
zxGHi4Wu<(>nvc;ef@*c7j)iAwVMkuzyDNydCdDW)k$!LA?>li3=HKEa$kysF&JZ3>
zE=9x>Y$qn#5coKbnahWZIk)sGtvm1w62oH!wddc(Z01(gZlkwg@UhZ759oD7!R<}I
zkqREoq3!Ea?RwAW7ibIABz~xvpQwtqMaU-$LW<Fi+;J@bf+0RSJzl5@P{Td4$GWQ)
z1dcSFg#43zO<qP5G?<(0mAcRdIIlZZZ^D~I$ph9*<XXaO6F=?HZdV4g=u1j*@5XvX
zHlSw`JQFs0lmE3LCt~K5qhDMApw8Oj=W%_y%NnpD*=RU9=M7(_YV$s#2dy<lwX4hV
zvlb!3F`%A=VsQ)<JWnx2W%dklJp;%I?>!&nx>i`(48I-K2XA`-*xMV!BS1!|c%q!q
zA$L8pJj4a!_KZ-cz|y{B5Qk>z0~KbPo&-pdx}oSwyo9{|&zME|A(p|6dS>oKr}h@d
zdCx7#Y%m)6X7qNzYd(^&nJ-427Mf`A?v)alTHw56Fa&$13ktFLjmBH`4KoG_ei;gb
zsi6){jsD<O`va9ltUQEtBbc+{#Tw<ob7Pfsg=%?mjW*#lU~8QBZnia2(7k&&7L~Ns
z;>a~m!m&TB5rt%Xp?@rd@E>#y!-)&#o|Q89zkRiwca6}6=Wey~6i1$<3ESo6i2d=Q
zcH}+A%scWi4<KDap&cRdDPXY2k))#-JB94T?$=xD1?;RD<s8g{M4(GnRM288tui57
z5)L_K;u4}HEt?*&BfaoL^Wo`7Nj<<gN~tkR91LIa#P{!d7C={lgo|zJ^rv)e7@p+Q
zhr*@o#l)RN%bad7Nv-zuqGNpIExASB^c*%t!BJ{?*Q9lAc<&-f+8)kBJ7}*4a)9qE
z<|rYX32J#t56+1t<{@pP)3&1uK)kRdG`efUjm2AF&_AQSt6!3@hP#lEO{-(7(#<8h
z6gHVh@jFjz$=L8w?Lhy8Zx+|>5kSV*PG;YC_^=8^%pn!1zAZrHQIK~%IL>o2e_HYP
z>_W+QlnQmhDq(dH9_>@_1Nh{!(8*(+$;Ib$hbT&5AMMi@6(r4V&`sVhU^@Ge2&HZS
zT+d0*n&h;}4iHII{xRE}mG82wCwXUs@nGa>FD3lv&hI^(M<Va^#E}cbB#0o{g||Ci
zgh1GrS^I+&xeqA-Iy7q?{!iv~X@mi?KpElfh@^f7>XIGvd0??4il$>Hn^Lih+p)5B
zO&IHnWbdj^t^3m2s|UXD@n)5r+c;HnSnxPc$xaE*pl4a-^dL+Tk{f{>`_-wXyQ`+r
zA*#q^QRG{nWyRblV?=tS8%}`i@DCo=1olkjri7c}yS%1<;?l@>FOhn)I0A<i=0lzY
zomYs9j*LrAMZwe<OAdcyt@zScagX@YUzX7tDiW+G@9d^p=D;DJ>~)RYj2+Z4>Rm-=
zCw!KpHdoy_$`_f5cf}$xEsBtgCyCBZbw-I+KphVvjLITnUx+4=xNLCmsyQ;mS5gZW
z!JF4|4XSf!&*z$FAK<-H>k>3NuH;HS9H*8Kl8a3g8UKK5;l}*!9n<C}NM@vo_$wP)
zoQ*&3oU!^_w|{8EGbAes!gH|=7?<<wk?SM;o@jf_a-_kRgz?y;rC@xzP^@*_Wer+v
z#nS}-NbU7XuecQl@Hc{TnGJcOx-nxrx%+6WGG29V0#95kc?1@wed0JUoL!8nvA_vW
z)HN5|bek@^drKu^St#<|F*HqsnaA>;<*B5l6sq+*WzBb<{CPM^@AO2ok=oOIdmE{r
z9nbK_FEUR9!nt%QQf1vf4g|TdI9cKju(wOSB#s==aiK3P5Np&#W+BTu#Nz%J?(pqj
zhgRR}ulH#20p&b>hx(x?4|Ra(nLt{zL&^1_^8T>z0VBKhXhkmghZ`k}qHLkFGaX%M
zzY9ax<1n!-Q2(ICrW)%d&8Mb<QEu@=@wS6{VM|hv#zb6Pf+q4(F?CGiGC|Dq2+08D
zf($rkW0`^WywNJ!as>@r%y)3_iKV=#q8BGIc;_hH@G(?V#_<Q2S>Nws5$)L4%TZ%1
zpUX-G53l>`33t&bXx;OA?j=`j$eT?zRBJ}g#)|(5&wq7LQLLu@%7MWAgda<GRu&h;
zJVKZXzIJ!%`rEf9)TV`nI@TsD%a$WHlF1$z&q^|I?FaeM+_yFJb98BHL-sKs5Ye-h
zZoh(Y^^=IB52}L79={fdB~vnZrIW?*1ubIz!|1=6L|OhPEp%KQod1_z#>MnME1{cE
zx3ohO|M59ll?;S-I#@}`%xqE$r3T{^AN59#D_uuDl^TY}jHwUR?jEV?DiYUTiHxaM
z{j*R2hC`Qkl9_o~lywO>YgZNn97zK7u{lLMdG!wTqkU92qY-@J_Ki@K`>z-!s0S$h
z%FCIM^1QdyL!TIX5pEAs-nosjcgl38{Hg^HuoJ_z7ZFjCVquWwW0*+zNJSS%z=&zS
zABpA4@Ba`ItIxHJHOZ&s1aRp?%cI-IU=%22x*UX4P5IiMJrE7YFmy%yGy-tx&1vqF
zkMH3T{o2v#@M}%J3KOj+bqvyBJ_j2y@_&&uuIYI{Zx4^vuS-l|f+L~99Z(uisnAab
z{lQH2NK0c<#A#R~YCJHKXrgW7n9HMdFNSbrtF&2oOG|Y{8^W57)<*eDWi#yMfJFrC
z5aL+Ke@e~1epDQgL8l^RoBUVi<@|uvPNheU;w~IcB{Hh!O)fH}LfI$oJnEWO%?^1=
zBH(z`_1V&Q6vAYL6FG>BYt$k*Y%tW5@u&Zo#ou7`vDfn4IwOxNlAv1tOd<5eptS+R
z&_p#diWP7bnMTt~TCsy;C6ZXb?jp4`s?z=JRwMERwf1zq<Nbqm?4Wqx01s;<>wSHm
z_XGt%R>$aIyo;pKS!jNFElVzL-Dob@;sgRfjm_-_&o~Tq?8+K~%Xx<ITZW-5!%0x2
zcl{lbQVW`5X+i%*AG057C;1bHUe+iPR@CEVEMK1rmOd^HMzmfXK5g^u^X>G3u&&jW
zHG_TZ)bPQIYAaG~ZcFJwghim@2#7$k*<SGp;435OsySn*I^DTNd@TOO%iCN$Fg0v=
zVY95QF$63VhYQP65@4@NI}K^=>XsFoP}&o}>=Aiiin8!!{Y|}+Kfr}Lw>nt=hx>_g
zLpW9EN^_Sndf~TKlKobmy+MQMHRyqTpe%$bsK6!}-#yU=#J9JCUsMZ)^Z5Hb{DN3t
zf4#O~)L68;DDH}QJ9eN=Ms_3d^p?OGF?t)6Tq7ueE<%%w@Lhs3ews1cexyI~xZKdl
zh+Z77!=!|;HrMT=KdBrNF_8ejUBe?CzTq>UtpAvILvWCdCVSbywmg^z<$132II3Jy
zY>EyfrQjC_aYx?132~IXx|>JLCx)#c#&kUv<Z;@}+vp*;2Z1yrUhD%SzFc>SLV;Jw
z1Ggm%=^(QyEboE({sL8h2Pi=S^Pq;5VE}Qxtko+L)ytRfqZ{3@8cn(9z|7M`;1(ar
zOMsRjC+p2M4gGWTA~y3C94GNWa7gg_k0-SBELG5j;9SudaTo;^MsR(!DAcv#rp@U(
z#dJ2pwD8dgQK--asx3D~*B{p?K|ci{6BKz^N5RZ^VpmnGJ^mOd<@iKq8NAFLS4KZ}
zo1;ZHUdMSa_2-741U%vc#dH_38^0#A4;RFwq_T0Bx$T)?UYfr*Qkt}xRmjerp2tQ)
zrY{(=ifZ8JxJ211Mv|{Lr_#cQM*LUjvG*xe&ey}tq_?2D!0`1Mr4EFKBm}T&MGOJQ
zXqEY_1}EMP*?%sMju<NABPT8oE6u!~&3&U^im{$r6~~Y5>J<8#Pd%!=II15=1yJ?H
zv^NJkeW%Ww2(qUKX{XO4dJnDqI-MQ*F?&5hIQm6nuC1C9d^(;184Rfl?rIE+(x5K)
z?fm$No~Dyw1z@MUloqGXj6)26b{a94E-|@GpR9nBp~=eXVAhw@W{3Qk?fUiRhwXWg
zJyl=-gv_MYSAES1waRfB%)>gVEHUL9?E&;qjrejr))(GuK0xRk{C>yiMR2o9-D{Ob
zKg-O@zz4h=mY);20jY?TMV=jAjhd5~Tgu8V?|S^n84-c-8{-1+uiHWo&f3WRJ6XZf
zR{}3cIk1=F*bSXI@5mxvEQdX-+PG^5yo}FhH@}Q(N^6k`H^b)`f(qrc6IOK~ihfaD
z_}*7tW5aVxh7G_X3THT6saLB;?Pp(o)w_C@n-pCUYW}{U;Y=<;5<gG_mIWw3GA#cd
z1|9K+LC?q+@{f<}D(-WA@kN{7IZ~b4pBfw|6dS8s;Ni0%Ul>|2bhK9YDh-_^975(c
zqf9lNu2m7igt6;v3w%)WBKTlW4G{?=RJyHj<yXsGK;!JQX8Z)QXhf3d351#3&!y%m
z$$l!Bqm+ZB&H&D$%5F><3H}P8Cv@e{<16}!RQNrk_7di(lH<dAg6un4xhb6bfd`ap
zQFj<v%jAo&tK&o@3SH~3XCgmg(Ji=RI@e2_3o$dwh^MYYI36|CMQb-_h56euTlljC
z%Sk<A_PQ-|xyzuLKsv>%i^I{IxrcJhK1N1fiLjcW&-rM~5DZDbE3bHAdCn4of{n#f
zhm|)wDP22aZCn`s@)?T|PV1JhS>Fd!rKrc|Cpz=;NVJ=sYoBm6%%?sCPW|+BdAa<B
z-~*T<x6O)Iz@pZhL!{Od!J-zrjFsA7H~xgDCgvc*@`INlej>z@HSI&X1ltgvDv*Hb
z&hw8W@2Bz+hw_G8JO|r|OG)H??o6p7$FC#mGbniZJG>;h3Za7AZwC@=JMN@<YG40z
zR~NIPGZT=wV8O(hjNn;@EH>unTK#k3&d<osS|!Q^+KaVp)b_q6=pDSWB*qd%RKeM#
zJEz}}ej*EB+IYTv@jMECbC&Idj@m<-`)Sr=#Xd9t^eEO3G@#%PeRxaxm2MdZ0$^j*
zcuhjT!IxMj4va?r*yu1}6d?4V{$zG;D(+4(`b8N>{~;_mKBy5dzhqL8GkKJKFnf2#
zLRIrfiP33Etd5_^dvW^;pF|_;b|ab$FOL<R?+#8vwFyRYa9k;r-{vo&r+{^Rq#<aL
zHGL>~>>s{p!pPe4!<*+oskS6$IR?`h8X>ik{3>myRLNjNHzr>%D*c%5IHK9FJD=h}
zevp*#QO=C%=bo3d)6MOiF`N@B9y(){*xcNA1|uB>zIp1{WZIH77<yWDnW6>h-n!O_
z<lT`euQnFTR8-V=HhQT+Dl7<u5@dz`im8dP$$%re-$I+sU1cf)Cd~Nh@ycEigG}b>
zh}qD;Fu|+*yhfEYc|DL*?y`@UI}?wE!iW3qqqqnoWWtZK)=0bf&R(Rdzt*-Tq!zt3
z?Ah#VCfdHlM3XBT<AzE82<~r{>&k&@+GsXw+sJ4ju2D`oL|f-vH5b0V?kje&NfL{4
zN5VYK#{?s|q*$`6?6WvaTbD_5*!<9wW41z52aDYOq(CUiDknES>5U)UJaoVdO^Lff
zQ`w~1$20nS!OQ(y=3i8BPLCtN<Ai*3+`f%*o29txV8)ge*y1Lw6Vu^0SqM+b`7g=j
z@`Dcc+7>}>3O~c(bqPS)@P}?8YmxS3;uH$Y-4o0K-9&wxqX@{O4E>jPDnAJ>EU6J5
zPHZV6k5%6N!*22kJ8KkBnb=DYQ@o7N_Vr7(h<6%s9<C7nUx0In`g&qzJ7^hZUUP-J
z2hUIwWL1&l_Nvg>vwN6BPJmDsc0GyvsTfq;-0iDY_suX*!tG!n6KW6~T+H1kz)67o
zSAn!4#(tu_91-`ti~J<`&T7GK%J?7|<r}4kCFQ8|wF$043HZT+ifp;Tum<typbW}E
zbSmDMr3r|n*aI{p(y$A?%ZcYDq&!96%w68;$nnRU`a7=V{>fF@EQgN$<#Q4b48;>M
z!edIpLp~aRM*5ESDnX>`k(?@jt9et4LIB&rfs!e#mOMDw>Q(1Hg=8;7q)h5fk!xqM
z_9$8Z=fB>i{mgOSQ3@BgblzT;N9$vTD$UsQ3i6!Yr`sQ_asi`mJCsa*TU(hFafMVP
ziuM?Fot)gBo9T;t1E1W#GD&_ULYBmE98nR6XPoswr=C=&c$Q9;O_=7Le}UM<kNmf+
z@_(J>|2M15#L4(SE4*3wCJ29VTwUtq2%uC@%141B<ki$6O|N0LcswJCdT$W0ixLKX
ze#FKW$B-Mi&GrV-XJ>>50%Oo6aUZ0yAKH03&-pX(ON--zOc%yfA@6&pzjRGmoH0P{
zt+TZ^=f55Ymj%xx;3I4zRr0&g_&#=egh*hn>$k5_P>&~+olVS8)A{B?!Omo$qugFb
z+iY~UnYDj#z9~6L@BE|lxS8ZSd~$yN8jIplRDQL8lq9<-`njIUPn=6$B1(989+FGJ
zcl}#-*E=6KGX$I&NGi!IBII-1**=AWt>eh&m6HHRM_6m$RF2oa1ZqMP*V%dU$354C
zKj+-ovr_7yOi3u)I+h{GVuDt*R6(zZ>clT9;*pI9z4hwcy1M$x0<~r(_w9C*Ko~iQ
zBc@U+A}V|?-!H`}5dB55@5eCN*N}c5!oqW1{-GLes@XJid7q&ee5m~ihz}J3kCfG7
zxOE-!zca<!QPcc{VUY#G{e&wgFW1cwXy-WA>7@W@&6_`m_mf`JC(*vTTf8&^bPoss
zyZA27&)Pt>;)A+jwRgy<1vG<%YfWP2N^7(+q!_y&+_iCBy!Zq-V#Qk)eaCwmtW8v=
zwfn+0cGM$5C{JWJhRjjQ(IxyIe3z!BMCEct-Iz`^P&^Z-utw}>x{Y8i=bD~+=YMmc
zB#@Hl?}mfe*~;yXP>M!dkj=Rst1l^$?9=DCpoOlW;z!HchR!Z!qHXke(sZ|0LR8U1
z4+L?;Aqs}GaMCe`PMO4!hU@o#TQ~l8OrN$NM!BdcWHv1+P;2GV9R&U#jJ<P=XJNOt
zS+?yib=fw%Y}>YNb=kIU+g-M+er4PC)OlyVWHRR@b9SD;R<g5`C)vruy{>zfnsPB~
zOCmTZtK=RLNh3A9D~MM`A_f%wwrjQNF~|qPDoJ0f5d57k?BcKqI5EaS7S~6}|L`@X
z5-a!UpKN@K<dVPa3Bcs<kkb)XxoSwrv840g9Bwa}!wvFg`Q{j%CNv5>R3x7262v;$
zpxmTbS+A^$DzSY?^TN{`?SiID9y9F1L1Sx)vr4^=e+bhFHkX0Iba|2l6e;x@@2YR!
zbbRAqhPO!f1gYwH4lOtf4W~Bc!5<cUMJWxFp2N=0@1)u=i&r&%H5D$p<c@_h6%FfX
z>=m+WQ9h?(JjL#j&pZVILG_$2{@f9~a*g*-&v#;v_Oc^hmUDCTOye&4j#ws;M(O2W
zbSkWp(K_08jbfV>_1(}oB`K;08&0#!eKso+e$B&68CvrL4$WJ-%WkG+wyld>X(t}_
zcXSfPCG#SlX9+F=o!a@|=U73M7Gg&PMMJUb7|XQOYCa2Rs^B&lpTfg?LV+{sMmgQ1
zRcbcVK~Kvun%Zrm4AKu#-{DqZ@edJKF>}0*qprIABZ$Ypn-+JCI!Z4<L-$BX8@Z3L
zCZ036o74NAc$^hTs08+O?4sg9rE+V}vJ_?>-+oqjE!G~Fi+kj+?<YFb?3b-Tb-Q~m
zebvBTs+d?#4@wS}^r-Bt>b&6-Ux4D5((pcw;N5b4-=7{sS8@eW0l9FNXo`N4Ya-pr
zYO#PKfs^fXuUT=A70OvuRzGEad1A+KAh3Fyt_plWS-(Or_qFfwri76Y%m;Ql4st3n
zC`^#8l(Bw<joORPHM>Yfy2x^Qc)y)=@!lQvE-Xgx-PtyHQk}0tXRYOPvE~vat`n>s
z%r50@FKcC>=nh$BW0iG#Q8f6Tay65PBr~R~m(P01p=)Y{Rg;7m=;*OxJ8QN{>4H{E
zSc-fctq0vR{i;De^(71WqcRFo&I$&Zt3uf_Ek@3_uTQ1>Hq?2$7X}w1-pEPN9)ylo
zM|Zd?fr{c<y=%OheLzZ{^@@{h>|q3D)-tiRV6eKcLbkX_Yp6y$eI`6eQhWEdew+#4
zIovLY$+u}JnA)Yka;ApgYlV6@8C{A5k%g2NY(??Kj;@pCrwgsW_4=dBgL^`yRu7uT
z2`%DA)@jYnzPj;xCTa>-Y#Klw+4zoODWS^cd^1GXSZ*SRYhquRT|*qSxM?wvoEe=O
z`dW70seSY<(4s_@>)wHG=$)rusBay$D&IO`fxd>=$NvGSdfJYJnmtz2^0C9jRm=Q9
zGm?wc+-1R9J<Tj|k{1yGqw2=tQqx-<f$YsW&h`7>46%|X_5_XV0vjzPH7mf$#0%*9
zIN?!~zyi55Mf501etxvp60CCNQ`l`=3Z!e^ZH>~>Qd<k;f~<TsR4}d5gurvNxU#`v
z9_Y(cZ|FzoWoIvoF(Gi-MOW<%)a_h~{%QNd`$VGLwHmt$<e|7N1AV97GRp8z;|#KA
zVaG>nrf7DR#wonv@>8Y+W0|X+C&jk6qxxZB(~jtpluI`OP^n`n_Ty5`B-pm5qEXSJ
z)L~}E<k+@yFcMg7c$@P%I%;5akm`-m$)Wu{bTRlgaKPf??zZT5An!{XW8%X3Rxzt{
z^@W>XEO7Qew8~gG*?yLX{~uZU{~6jcbN=tzVa+%bwxsPp-m?n)vzpz%{O8o&YHbKx
zv0pF_kg(Fn47x^b5ney#Q79EEDtwLOfGq((+&+oh>V%1B5js1_@ZRI+LHb%cx^bvu
z;hTx~R+0Xw;O}HOV$W9h@{o9j$M%m$3CYXiCNuY{r0G7?oolGxD`yrqkl?%0-tKR3
zWUu_fW&_xIC&5*H+&x%Ag&X%jFu1E~-ZU{LSlPSg%%5;}_0y|DFt)ZAAnP3H56fF@
zzjqT0x@81Pn0)=XEOYf2YaAA&dCGx$ue3u1SOCMqK|CZwGb{<O_8n8#j_^o=rU-t*
zRJ!pf>&t+SWe6N1bwUXoC&lT*dwGuzsB=fIiY^9LL-D+LRBfcvfL%z$(E2l3u4@f9
zV2aR6!yz)*!gQ{{NoH`r9d{U=NBiPgI?--%gWd4q^h3GX<HwVgR#5yd7X@|nqo5gp
zHbvqK+ynopONmXfIb)c9#-lR$AbA2Bm6~}Xd{F&>6|d1Kb!70zUkS8g8Whoxu?W?o
zZ*R$pGoi*Q(yC?0g`>u%rj@4BJCENyd7d{UwVGCqlg(ROCh5blR)E(i7<jzlQrnWz
z+TvToh?eo;o~$G>lCi&8F4()Ml9Tyt{Sah<(?7t%P`%uu8k}a%!6wPy)XEIpf8?@e
zAgj;`Su8ZOs+TO4vQX^h7xhq2=@LMYCJOy4I1jN+rb9%y@^ej9#2_KqkgY&H*dPAX
zK4wp*B9y&Ty85pJ8h=yn7m5>YpgApEr#>eWWyJlFw|ITw`Ka$hu#;MYFg8?lp72Xy
zqDczy>+`)~D<eaQo83+I^Z9-{AK6a!H~F5<bF3#sPTso<uZ39D?5z9fbA$3#3~W>&
zme{9FYmnE^$)~jj{WbAl)Wjgz^FsbRiZqL+A;ek3!r*Ws(0K^!F!NEYcZ#d$AVjtb
z07l4(sHpFRy1k8W$YT>#?@?$kDn-@XvExe%UQeEK_A{SxEnc&OMmq>8xR#>Ings69
zZz*8_Y<YD9X)>BXx0@(uFTQf0Lg1S^z6g+>X-bRMe!3{uHVarA6ev9Fe90|7r*zkV
zxRN3{9v+GN2aqE$x(dZ6)$Z5x2zkvrLgm(~-H?+q0P6h15iB0+TxM8AVk>1q#6Ixo
zRGv*Ch_U)g$aJjQN(gA(M+1v9xz@<+Vq$YZUXp{Od3_+QTLd#uLh?jE6UmhrF%Xqy
zLd>ZkI4AYD<jEk7oCZ$TK;G9zpduk6fm;^F>%}D+)UDy$V&L*VWn2ri_h!VD87W7d
zH!<_=MvJLz2N>GqODFftRc{rw2Y|0~eHwo>Lcb4HCxoB?Ru~(a#^{fnp9uNdsokmZ
zN^M1AFFgs?a5M=lp0NZnL$Ed7`NI$gZIh!@vogmlR#$tlO%Ytlhf3oWL0jp?o$4J}
zJ*;th2ELx`_z8JU`)>lTh%vW+MkPZu;M>1RsgMPaP>JR@)ei|lUOkeZWR5F|16|CU
zVuWl~73XyItu>ddTInW>Y7|O>FzDJKUjsZLhy&Br4CW<Rw#gMpXw5Ty<3LD=0#d1A
z9?r|XP*qU(vt}EA{1T)Q5~9`5GHoMjuzPvby2$ho7}_g3?RU?0b{iv%Gvl-FJUZQg
zj6YYjI~{lKNwK=gDtY88y@rZKfWSzc?GQHPeIVD2(w;lY-=@RjgumO)=$72RGR``0
z5-aS8vL(I*eVgEk%ah@>hK_!^LGk$oG1cbnv9Jlj!~QzU7s`-gW~|Az@P6kWEHGN=
zr`WuppTCPb84ns|(+VXU=WqbwY`SIN23@%mMIF9zny$kzV)-4cF%tiDhOa;^f9zm4
z)VPBehlXi#*dbKO0w9WlC{5jSfum>LaFIws;zDLB?QoDTEC##IkmX97yLV5K_Q<2g
zOq^~I?;vyq{x$7YyHLWappC_gMmN>B&`e*HVtCuw!dtekxK-H)q)R6vuoL6m>U{Au
z7Y(Pv;~AZXr(}o530~H~T1lv0ATaXACScv%nQB}}7_Zrl7<Zu%?8C7sOp}wO4?Q77
z^F2R_rJE4fNcc1xUi3dYixAiIN>cq>)FM^-<(m8#c_Zhzs*#;{R$jj?KT#Cq{Ien6
zQ)Fa;bBL}Y{jG--bBshi@kn)BS-6Y6e4DH2>>4LombX5}CKAKbJi*1JR#$Z@L_6e|
zlZ(=O(3O?7=KR{p4sgh2IJ72%2{g-9b>N`pn^$uL`BoxD?bw-N+M#X9JDLD2_>Wb}
z=w(4!HHTZNFdTU?CATBC9Y=F%q3-X>_4LL~W?RYIlnm=~+#h&WT8-KoT)V8ojA;dK
zk3+o2?qm1Gt!Z{X8R3=(JvS>REcuLF3oCkItH5P#SXS9v`9`PU6BCoatW)d=cr$`f
zh$x4yL>w?euG-xx1NT@2SQim@6{K$Vo#iM@ty=i>FdCT_m)q%oU%vH_r8(*nnK8as
zKwmrNMA9_3kBa+G@-ORbnO}URRwox5I+=zF*KI(zqv@#*pJLad>BntE6L|Kul=E6S
zOKYEVC?xpZCN+O5|DJ5G`0hh;X=j#c)T>FON<c(q6BFh!jW6mWvP<}dScBDtczfh}
zot@^4bAYi?+9w=WvK8$DN3Kiwy&_e*sF%}s%JhfBShS56KU*)`BXQBj2Cm{|Hm<%$
zc|t=nU5}EciN{!Lrbb4o7hu`rH+HG!l+aT-&P8*k`IekCHNB#nZ?`UthvgfoTjgou
zRmCWFuk?|_-_<z4MSx=3{eFbkbnJqBy4uBhz197Ok3Wi(Ui-kIjUWHUZ>Mv;*9z6#
z%6a#;dl21pilo>|Sodds?Y{lZSKd#f;w}a<hFUC^NX|#8aMKmFS|;1gjS($P{Lu5*
zwmI(Y$}qfl>*Ho2<9wIg?#F6@)elwuG8ce#EA}I<?uY<E=C$6=v5lvf2meRCV(X67
z;QRi4+{k&h^;M=o`*jNccfB{acY6GMsg}^uI=xy=pK<piwrJ0unU2JOgw%C5uB@g6
zhi@k-NuX*_W+JuGFRdyC8{(!{Y6y0s$z0;qR7csM=G`k~OB>fs4iG5=X6gsMU3;uo
zuBXzj2-Ja1Q;bneG(WJKXY0p!vm(L+J0P?#H3}?I4)EoHd>M{L1~om?i-$>&o1%}X
zorYqyg%z((a11njCb#ZHpjVdstYg^5ZxER`g4_QgPh(~H?|B+C1Lyy;4Wcy_^B=6T
z(ju(1hTkHhKAdaGC1YGB{B|H7#t&8bLga_4d>)!W5w2KX(J>NaY?N2psf;3`a3B`(
z!%VP2o(SzMopH0vY1Y^{6Xa_4{^}6-Cc3FbTQxUuP^?j%IJ}QNthEb~$KaQHVH7ft
z{uSEG>pM8uokHLxenE6N*v-uJ)VsqS{tNe=kK>&!R%Pb)-o94Z(%t@w7C91B84nG5
zSoJYXj;~k2g3Ei~K8IuNaMv#P+mYOy_nLn@`a|o0-@ViO)#3R)5TBo>{5f0Q<)-e#
zOJb^W@x>0gvT>LwTqS5o&`*Da{rltT@yCa496j`t+7&%g1ATNS_1Ehh7hXxScF(a{
z=s^)!=gwrCJzA}12-t-@#PVPIiEIS~Gsqdu-jU@x9J?gLpl0o6w|b-WJXRF~;vs<M
z(fc$mteQ<kQ^~lfcZ9OJwfDbU6J_tHUtiiCdu4t!ZhEPmv7{<>%Ul<G-Sht2%N|Ci
zipiDnvQl(w0JP=2=z9^WerW%0)s`(lnd*jQp+ktTNOQklva^npg~2(UjCiFyh0LLs
z^>+MVovL-YBn&lJyT2Z7T3JCtQReRrU%ywu{hr-2#0&0)_EgDQ2=u+J0^8M}!5`a8
zxs8(DnRE$$^-NxLfGnHJQQDd+KLPoB8-wm$!PR;NpDryli)x!BN=Rim*bO8L^hKVQ
zMu5YM&hMKSAmjl!Vz@w@FA#Tmrsf?;A~>XUfde^f&|tQG%WK$2@%;r3?pc8o&=E6+
zl8*>`%dP$6`j`@I!r5u0VwTX0cK`7Z=tWSDY!0V`VV-2wSzeL*nCg3d86>kNZ}>AE
zTam!F2rjEb&`grWchYs(TdPRYFm|<5F<qhJgsr*IVJ2wa1w1-M(r)|~^+8La&3(yD
zbZq6ZE<f=vGG6`!W1CBCGx?7(o_ILz2LlT*2P6M7p46|_OJmnO(HFyHkCKB^RkWt3
zCE<Kvh<xIrtFG5Y>p$R~FlM6R&Yr1i<l;()gz#mRE=`J3Zwxnjk6q@m7az{Mqj}w#
zd7?4%o-*V<3N9Hx7v6I6XW^|W(gmIulS8uWP%h&7$V4xasRHeH!6Er$yD_79J8r}S
zi|bB<px)n!?UWweA?9bAymn&mApO}(_9#CZK+;Ms361+C$`r76sM-u%tM%z2RBKr;
zsb_5SZ#18L1?jKFpehFb$CvoIz&FN+G1@Uj8L^24(KZ6TH2|GW>>+nUq5i(?P%Yvh
zs_o`=CI<?TYWxs^-hsU&Eta6W@&^A`k9YSwhvIJ)fmU=MOj1a)hH6*ZTnZFKEbf{C
zO*qXT8(6qJd1fGzeym+|lJ+=_9}_F6b#2IP9xu*fb0j<lLR<<a!V~GdNjrxz)yoHr
z%J51`zj@+L42VB-gjEkDBAk%W*d|c`-iI7JZkU7c7YD|i)Ccn0g=Tx-K7+D?5$pnY
z_6lexf^OYnSX`LY^oFI){u4N1`{Lm2j2fZ#a?6faZS>w>NN~Y<gDS^T@EX=lzxE@O
zOY{Whr0x`!_)xvfunYsm!YU#G?p~#1Bxw!5S(&J?Y`G?vIzOC0D@U@QuTL!hJ_Pn`
zUwQ!XM*8OE>o8GsKM#QNvDGHZHb?*~RTU1eQqD-4E4zlP3P5G>)jU<7vp2Sl*;8c<
zp~x^<$bhC!d7V*)!s&4%QFQFk3WZjRa;%{XnQRs@rP^Jt5jpcdK5LTovjnd)m<&dQ
z&Q;`znG#iM7(X+x!(M4x|8k$l#rPQj&`dd>;vE>oX>hK=cW{M{CehFNm^R_2u2rGF
z*-dJayjcYg2wFm4(nu5#Z4z*pTg~~U$4hRQwk@)L1}9x;OOj^e*|+EiQ!9JUYoX4}
zm~0&}Mr^mUl7>-2gtpxD!a_&YTlH0YcgN}Hxuc+9MU%r==rWRGQRD=J3ETOqdP`vG
z!(DBNJCD$@(d6@rcN(0Mf^ax=3GK<U-eUOE2ebtVkI3V}?1my531pAl`*Stlfcn~6
zoYxjc44H@-ovrfnOxcN*%`Arc8)gBkGv!y&RPQp*jHtE_ozwV$Z;kMnO{#gq7eQq%
z%l&#%U_McS<-L+fL9G^N)t1N&WEIG2ca7qo141R4M`CMK5T7;9KBjx0EsA9{bVKs0
z8}n1tR|;&pU@h{M6nibV6X~x}1k2$&4WwOax-J%@rnQkiis7)FQ&gb(hR0jN(8F2B
zrzNv={+n<92AGHk<ZWsEIkwKdgM(t(>c*JiWi`EF-dKT9E%(mOY+ycw^FMQA{=($*
z9G9HzXCrx>)v#L)v(;}*%UmsR+GJSB(Ymu`^&}ziaTcIN-ZG&xx0$UBPA!y40h%Pe
z_ihaO15*iSdqCwen(*d5*^8PQFtzM8rS&8*6bDxAn}$w~D%q1orGEoqP((H5Y?jP=
zvh-+r$ldw3Us|g2LxZi_h)QvVhJda#mLuGCawWlcIT4!n$n0rB1MDe2KX+5sSqXPc
zN*I`LJ|%V`dI=<!u`YgEVUdlJiTZFQZ#r@kMmj9qgHlI&HdNWUr(rbe#k#0g#PmGA
zC2lri5^E+HD!+MD8|RJK$>K6{(#5`#A{AO)X?Isn)mW8I#GV)x0pq(nZ}fu3_P%74
z$KsQpmWc%S$|YiQd|%^u9LMQOiznaukSij%mV$ZZv_$2oqYAfj@<vI;H00Q1|1jC)
z=!DNX&L|H=*<y(*i%j5zq3Gj?;ZJ@~0gQCXk{bLo0;^>IS@gy}{h^%(y13g|YE}#%
zqHWBAxMZ>SV4q;>(e7AlQ-@}F>O~kyYiNQq!{Qt5b8WrI0iL-m1+Rk8bW0K7$zJJD
zGr>7^JaUE`yxDh@s_m35pyhWbcQp-_JQj<l??0>LLwECqxLWmMD<!(?n6rlM_8gm%
zLNjRgGq0=^9snyi<JRg89f6!EWfWk_bcl7oR^5|Omr2T7_%QWw)#~p=3o`4zOZlsv
zFDmf@l2`hTW-kRL+W7p;8RlQh#i65N&k$@oiA7Qg`q!*ntKSD5awYp7rHVrJJEFNz
z>F%<1V74<Xp}0x_)^IARx2HT)<3|sw{35H^dGJzjCy0hlbSIA5w8oPvj;f%BOgICR
z;{Zh$r{=Q~8*09%^kvn4&WiRWZV0ne{?omQx?|R)&_=EddcLBCXP%mly?OfL1^=}~
zt7ovV`C^`aG>x(+be09O#@(i(D+^WKS%yEi1J&J8xm&$dg*N*{D*x`nj!m_e#ME6I
z$(`gc*!W+U$!kSbw!AnminPtp#q>~{ZI+|f8|-IEyMvS(53-t&NDeH|wS|f~ls`8F
z*>q}A9;GdF@9fh73%Uhbl}V}RUZn+E9dksMLh5*A!3$ch3&#|)@$Mb>jP~cpi&e%D
zJy&1(h>Ex4#8F8Rre;^Q(K&EcrgY)BP5TU7D5jg%(wj#h#}=JZZ6s_i><$UobiA_g
z@gVsQLC2Z$Z=2>+(iW>BAT14hWcMn=(-aYg73u>pT@*<#5WWty0olfCPm}*>kC5R;
z%MQ(D5^o<d>z;=`^~_ybbftpkdy}R-#1n=IrW`=U=fgnbq6BFuq@WG1x?pdFlO;=?
z%1!9)O)6m%=N5QHa_Ip&X=~`sMRO%Oe=lp1yPBwGgvD`wxD@5rGHS?jyBvL%g1X_u
zI>!*h+&{oKEyR^T#~VtMZ8_c%<1Cq9`wh#C4$MfEbv!luNGU{B6dU2<y0FQ`ddc%7
zwPIx-1wL&!@ft6>&o`<Yhbv5ePX+bR82E+h=AzI1J^;V({u%9VU|~1PPl3~a_E|bA
zL*LAUr(EoVx9r9Wr0P+_WJ$Ub+euAWjco>j?L9>2f_3`L{7Wnm*&Ak(Qq_P&vBqGS
zXT?{hjioR8nW<*JFkb0tpBV;Dy9xX77<ayhi)PH{BN@f}bwg%3w0q@>PKWvF$t$uI
zXOMwJeIi<bT@9B<$~Ts_O#6mHt&H}INrNmF_j7e(N0PRfMk4pfs5)PrE7FQYAe4du
z_{AooAxh;lsS}bFUg)wjBXMn|UxgRT#=|xJ17%GGJ{<QE${`Voe_SOfJ58QW<am3q
ztIVl9PTthnl(=;tUkdLl;xbt2^#4$3u(AD@qt^eYA<V-5zuV&3)RJ-{VMq8e*?joL
zARM<-DY>MQ@~eEDwzH?$QIg7QjgvI>^BwkPj<`JKkC(X|Stx<WBAS3d;SDg@;a#o>
z9sd@--nFm$TpPzOJV$faS3OX7+ylM%+kD0lFjS?1$7jQbW1OOfXUs!hlIFtxiJe@o
zY4^&G1=uW<|Kk?X&5M{P<bygvCzqFauL}J~)wARfz3MSSI@Y(qZEPB$UnRkbA78gZ
zJbrgQ0NhlxNT9a2=gXPAQMwywIiGM`1Sltr=uI4zA%HKuf$>4Tph6u;_~U2Wa4$sc
zD}|HD_S(ePZUOb14Efdo_k4j}nrPmardP&jqU~S6h#0i5`)yq-8ca5jzb?w-A2jcT
z;L|&}YUASj0YVE157(L3>(GdquLLViOM%V%eoZ63v0URY6rzbXsxZeDF5|ICg2JMn
zSDnIeCyfN;e8x4phMb7MdQ{!~e#x$L-~o!UYhx;(bH>4<c#FT3xEfU|gC8v44~<$T
z+UljNL)w^OOfoQUhWA|LhrV92r>`}nY8n@tT3xLvoUE+PWj@%rBPx}yJC0X6`G1DW
z<jHSC1(?I_hpqH(y&qJm+oy+ek!W;nhkgzxso)Z35nuGF-*n$S@2?eN1tIfAcz`z{
zp|*PY=iv-{g@F^BhY0sJvO`ZqGS0H9wld|^(5?f>d$(ATwTmU5b9wZfmOE%RrIcxd
z^qv_jmF0;89izgSXO~J=Z0sJ9Rz)O|ON%77Wv#a+!HPqA&HhjJ?+yZ@ER^NBVFkk3
zwJBPY%|D=25SZ|P`Ep-2YdTAK>CA&%|4|qNkuC6XLO8FZ437$p7jg6<h|fO}wm5Vf
z$MAGPafODtQRpWw&whBTSHTUGX|6VCrd3a}bgJ^2p`*#AJ(np&<MfzXj$8KIb(Hg(
zz5YrBUHS`p9&u;yMVo{!*bzfW&4$B$fRB!u5ryA9K1zsFD~;G`SG>8^Q1m;$S8rTM
zGNRG`1Y?Gbv@aM^R8HH{%={NE8Hb>(@j~fX*@!ie$reb#4T7x`o^E^0Ja@_!a|_yt
zL~P`(sCrp3ouE~aCaO!cqW?3m(&YSkQ)pt6Rha5iT>)p4I~inp9t2zBGjSI5X`&1>
z*cY74J7H3x1jilP@S;9}1{Jql_w+(1b=Kd<CFC;n5}-kmRwh9Wuz@PgDDd5kt`QS6
z79|~MN4;Ivx?@TNBPm0_L8)IkP)b4bzpKY_9_KTKuTwvF6A<a_+!K`0Y{5MhK_GT#
zv&#Ere##@AU1b9{X8rQuh`wxvN;E00_0B<_g`!t2IXq|(?q3<3DtTpNl{T_r2V@WP
zfq6P|Xq}{lr35K(mfNZ~>*{iZ|0pQ6sY!T<mSo9AXBr9U!F3H-E&yC0U;7s!+Uc#v
z*Ll$TEKexFA~ZYF&ZhsO>)u^#<0}5NN~4I()`?X>QnCGpT46jo)+lr#_Sc94Lq^Lv
z?JLoU0lDpB<~H1(m~INlTMh){3*^-k*84qFmuz8b*p+i18q}4uOJ2w$_NHnoIG~TB
zxxUbTFwV>+8&#@HVPWTlC#%gRc`dG8#eOC|L#ekh!963^>}3>#_!qnyom8l>z_wDx
zp<ff|1b2f_RRvO}qi`{U<t(z!0f&Q>g-*f`0$jX%YOvTad!X6TK}k40#Wc=p?ieR>
zXGM7&7tO0;2#AtnOY5+{Vd5krrZT0yVnRhb4tq0sPqSRm1b4caF-sM8_fSkZsf@+6
zPjYqIj3TE$OvVJfY0(ngf0+x9%&b?E5SV=t4I|Ljcy;e?F=_hbeDbeSWR7IErg-Ux
zz&u<2okhsf!9hS)H($EIsF-8?`tW<&#OkNQzdzuKVvcxJfLHA!9s+N!i0|MXVW5N^
z3Q6iDyNj|(F<!aH47!aSHrw}OWxh&~1E_zfSIfX5sx>(Qw=U4V9DCs{m!q4yhj8%6
z2J2T8U#QlZd4^B{7a@kt+}U1uiMj<#_{SZtft~Z)rW|;X+>$bQN#)Y)aVxSB4;QLF
zTw%iY&Ca)fq*}$YPlJ7>$5RYxi&K%Wyc-7!G5BWxL}QxE!TyX#hVG6xP$V8`S~+3{
z?sMe9XM%v)H4)O%gXawsbLY%M;?1V0))mV54p;1eV(^zBID`Imf4mhC`52SQ#soGn
zR4b&N{dxKgeB}7=m(LZh>}Vu{W<~VEOHx!6UPvoZ4tJXVd)NS_Wbt^vHglw_Y(v;n
zbya$Uo<svkw{RJ`YMl3<Csal>&t3aeqo&La$pIT3ZRzHGEM<!T?9fhbnk^`3`7s8x
z!eO^V<!1!68j0fmk*_dfa%U}vmdMhiDFjD+h(Y{j95}ziY{7B(Y5H+gYmUcITal~R
zN36pocw3_$vZY0C$yCd;yrPgnu>m@{whhY!q>xIS+y~2gyXb87wjsjVP~I9Yr$^kC
zNM2W?<l6&)m#oZ~n5Rxp<3fx<7twi>q&s-BcWFv*=~-VBi&?qLg>ytx=_R&HyUlnT
zNxM9xsTFZo6#LdO<+JGYT~?Er`~ul-Noz%PnW7#`kXuD{p{In+#>VF@zIXshFd2i6
zRydmqF1s0^9M!@SKGI>xW!Nqp^|i0*cN=CSXi?1hy!`8Y;OxfCCvx4~zr8ra2LsDY
zNyghn%yOgf{Z)L8RrZ&Rn4xX*-2+rdoctXY&P3V3R#<e-A|hFSgz7fm0|j%@)aIzv
zxZ<eM1Z5<tFK((AG;#s)zPLGY+Jo{)@|upRzs#tk@^`_lZQanLO-XdX*k;vUzm#*!
zEkNpxaW)!Agz4WGnXxe6i|80}o=EWnIyP?HtXgi+8j@vYfJ%>k{2c{d7n0!61_j{l
z@qi&=Sa<bTloILns<EfbSPl9!?+=)(5B8Acx<|^3{X*2$u_u?gRHx>KN|Vg*mf-&O
z?+mn$dT5PAx}yrhLy6@$hn<^iS;<{U8d<ehp|-xdeh*ZUZ8dt(WScQ<V~;_;I*8ot
ztNr`2BWGBGDz;Z+kGi{9EF|5B<UpwXBX-hh$w=m2r4Dh&G|pBsHv6}|4$hNpoIF=F
z-y`h&hROL~Po;@G1xq(DypXU_;-O~tm?R^oJbSAHY|h#hV^95qQi114PmPGTd-uq#
z3S7?!rnz((eMfrQv$7kym}Z2N>}FzTl2YMA%O(9*w+>WaEB!^LEoRI%<#EjALVB_<
z1x0g@g!H`+pm(LZ2`FLl&cjt>u#n|Fs2L@OYz~fYK9#%9F%n(PX?I?`0l$ww>jwc8
zsIysTlQnvpa`6pZAGZYiKjeHYZ2vFMzW>0=<qR#AoNef32^kp}elX&WPR@j!oSgq>
zUxt~9jqQK6)jg%P?zldR?6ab_<8YtXzd>JKT|F<^V`D5;pEXWKrv^i60F4300A?=r
z{>aID;D1L;4ZNYOZBj0X;E5osSK<YM0!WK86Vl3r4*v$MO5;cbPBcN0zD&By7%3k$
zjls=i-8)w^)9+7_-KSjD^T2BVg1QtTYs<^z5sly&7f4qwn#+`?z*QNRdfR3AM<s@G
z^%!_%Ka0&n=}MJ2fSuq>mdug{n(F99&!C0h>yqvq@WV1QsZ-D!O$yugOr{)?_?06@
zDdXf0OF6x(Q_>Wb#z`vW1eQoC9IM|pngs6&s>TT)$MquZVJHi^fh_bP?cef}BxtG$
zK}_W@NvmSY;v1yTP~k$DBQ)Bizy|6?$P+V!CFPB6%fwUo8+xA|3=%y2h|wKxFr05N
zUwvK)`-%zF9GeiE_5hfJ6(?~I7e3s*zAmsHrjQm5r}Sj4@9$EUX`o3J5h&J;VRqDZ
z$E58M7;+4%6$Bo~*ku5NyJP7NP~4xRsECBgEcjq3_iXkEQVv+m7*h90!cZolgXEZ>
zcW{w^BkBRXaha~0IgnBkwdPPY)rp|?{_QbQP>L&Rg!$^<CiS0Dq|>+~|IQ3S+Ut(-
zPB^I}RJQ-USBmv-S1;|acZ^*d@<DIRq>Tj{;w>9Un4g0`S?gadp?@TUiTkAjWK~f{
zjG6X)zLG>p&S;bnZEk7N)-=WW<ar};TxZEIMDQLRA%pr5_>cmHVBn;R7eK$m(_|zX
z;*-;4%3so)tE0kk7%~Hg5PUxv{hV59?h8vS-aia^&ri>qENdD+j$NB6)a*Ftf&5bW
zXY>w6?1ql^TsW_{adxpGfx@>16U-dQSZj*SUaoX}lLS9SzN&l{j>y*#@og&rf5ZDG
zbA1PTB^n0_CkLl?t!d(nxMs3MtYIgjl4G+!*HHskq>G@F^%0=KYX*?s^c9$yewzt~
zx<Tgl&%{C;I;ifvte;~(e1}%{=(+k$-%nKZYR;^7b&vUJtf<_cJrq>)r&QAMYdtN&
zzsrLj_PeL)G6?%3aopd`5n^@OqD)45Jl;76)x1@Q8OwrOfJc@uP>If+fhBIq9;0C0
ztCe0he8cUGX@QZn&z^s*_<f)z5-EjOX{<bU+93#J4Gx-XIKkb4idz2hcR`$^tTsrx
zaU*H`Xscdd8fr*OO)tQ0_~<%$cU&@dbb38#@pGNS^hpn&f|Oe?8Nri7r#Q$N(a6bY
ztC`>Xd@IepyI0}vrFd_BeH~l)knvdW?KG>t-)s2SB%{aXBL=IU1x0oSc(l}Y%|wso
z^VFcL*gJk=%Sd9%g@vlWHw@usl(eypTE1qp^2++!vC-ayw$|QY50uH}$N`H7O9x8^
z%MNzt{xGF(BYh$LGj&~3%*w=0>nO??LoQ<hKS9iS>V_J*X{Oq!$q*wX<r&H_%w@Hs
z0pNGM3xzCj5I0f}BCX*~cEO|6=HWx(Plg&SpvFMAQcK-W2vW}QP<QMLZrZ7<ih*#U
z=T*M1EcmXrtT=C>s}NZ_91(GIR+O1GhJb*f&<{?P?lX%9Ukz`svfKi-&?AmNq;I)f
z!<4W=Q~eO-OB`SzU_(GS_T^bT)bUzqP|=`yt8|>3=2mU?SUb?sn>=IcwZJsdVNLxO
zTBgiI;Og*JNV=%##eN6sXAGIgF9#zn1WHvlH>QZLe~7h@8uA_kYVnBm&}dheD^#Ig
z6jW{k0DTWNi{ckOkYc{n^=J^#jmIVx^QW{X%ES)COTaJb0R@$nv0Acn(-s~5vP*T#
z%w4F<mor?%6Gu~z@gzmC-r^zH&{5b{mTDPNZ(h)s2>FnsJ!5N_D{o*d`+^LhRfsC^
zibm(8ZADo|aKu^8G+a`fs-imoD6os?>+H!WtZ36lXWumbr<^k^3X%K^@OzAN-#df9
zCGtth*TK>8OXc?Xdc~k=8epg5|8;ITHBWW12`Y-iAge{x%dZGcy35}IG?_c^3wv1|
zRR$6CS%CxR`G-}Un^c(fMClS!aSB5b8SCd9JW$Evj<7^k)*IV*R_?*U#g1Qfh9nvK
z`l)0+rO`YABJzu4VUyv%;uHCeb%K4B2XXG*>u7hiROeuduspv<LKh&<3eF1Jv*d?7
zqMOPgrPW)p(a+xdlSoE0J7_U<zG&TJn^ONtR`0Bj8w~{{Qgnz8l?R$Bu3ZCQ<v3O&
z2YAx9PwB4VR}V2h{T1|Mu%F}NVX})G5wR(wOn}D>`6*8ttL`=UU>s%=<fg>JHt6q>
zFLh(a#GX4A>Mofr)}5)#?2L0EU%08L5WZU^hm%DXG@}rvEyHEEWp1DI<Hlxc>FX;F
ze}Ru*#4c6X*oI~L1Oo{rP_lr}9fI%Rw#BcZ8EE_BgD`RAle#hWv9hX(I~6Ko;w37O
z#KH~76Gg*G&-PG?HQI<O)8+E{7_y7_J(la}xYZM=*8qk*5mIg(bS8Vu`H^s&a=+eP
z`i~1+D!WT!7Wwo^RScVs-^J*ky&_nxUQ_E$qkkb=Ij6)KM$4kI5v8~Wp3oTl=TEbS
z1y4K!{W++LWeEMa2}f}(p+%geQ{_8(`#}`y(}{!$go$|xd5QW7`iVOUJBe{DkwqG$
zQ^Xb+5{=>+BJ(r}rt$xGh{+;FqER}-WS%C`G@WIlNP}$ZKOxBsmHDa!(`1&)B8}fu
zR2HfdjsFQHnr5?{7io}Baak-&G|FbU%vb$aD6?pdbc)P^O`=gELuTGO!E|4=*mGB2
zm&mMNe{vA=x7gW;H({~`t4$(4bT1!GM$LHanaNVpJ~s_MF^4+?Y~2_TN?vp7J`=TP
ze?!mu!IG78b!B{ab5ZvqgT7K0+YFih`*O^fo^H4M{hKw9pDZ)iv2crl^CHg;<m5R2
zT^}TpT}6qmXmBG4?6dloymTjiSpbJWlVirQkNHI9<446ml%#L(hqsF*u(P_f5gp^E
zwo#67E+LyNZx)wl_Dnzpr?r~@tp9BN`$QR|`L+`*eLM%u)^peOiK#{}*XMi1Kp<a^
z_yY+_`0%Pf!0|Y!3DuT7=jb<Ctl51$3m4{$AZlB{Ft7i*jo0)M>S`fA1|Y-D`KjG#
zrXV$8;mKrO3&5OHsIS@V;_L0`&6x6M8VsW?*ntokOYCy&TP+H;>`(I5Y1lSH5*#-b
zJbuf&MNFIe;*#~dmA7V*xa<uC2Jp-+FCIm1>*J<MyE=N_?LkvUG<kw-hr%dMaoWMN
zx1v!qqhan-O{5AV_4%#*{xVIs(2w*fnn^$;=sa*!?yfx5XrGJ$5G}m}5qS-~LHDZP
zpOfI-6(V)lN172Iz&o~^V5$AOq?6CorP){j5;WmU<6`L4)gi(r4{59`=hVO7c+Vbs
ze6FE0RmejmYt8dse5qA@&H3wjQjSOUlUR0|PFJ=JCBF1DnUyPE+1JbzKh7f%AXe_Q
zQ8Q6OS-Qjpt;fW+3oH4-zr=+gU4XqrLC_4)(Ds`8NGpi|DaqLUQzoLc=;p(EefQcx
zXNR~xB_(}G^{aD-Lo^gts|vKT4<~C)>4}ry=kC1egXAPxp?}Y+E?kxzS-_Ehty3#X
zcAF`M8+ucK&;`M(qA!-dhXcgh`BMCPtss)E(^E`qCu`M=HbP~A=rQi$`tSws#~1Gg
z;*Cb7%T(@0Hs;HduiU4^iTwv{_NUQEegWc|G<c#$XK9GW&+*~HIw62{Hy&bh9D!ix
zE__Prv4?}^y>k1`am#go9i-4%h<z3FBzR59*CR05j>iY0kEMH7cKw0-%|CTLK2z$t
z(gT+$Wy*FThmWUAr)IoHN`t${({r`Dej{pX>n|q}!~<xzKyT^6z11I{er|=~`AnGy
zs>v9g9rRcmvz7y0wsUt|K=bvx+)%|N{rRB_AKyF28)(SS+zj!gXlvnIc1CvQn~CUo
zmU_3Ya=R;v@n+=obC2=bvR3V~ia6K^e;ULe9|%4v8sdR=oQ<$_Q7@}fnsX!z5!51B
zL&iJua_rp9EOg|uU@}*2CPy!-w!Li|Seo<-bVtDOAA3HofMI?w;eegowZU%$i^@rv
z#`_OT6_@*#u8G`j+IWyiiLs%PUOmNnWITD0=2az&7emf+{^+^R5cbt{-Skwi2B&}Q
z{nNlMet(D2mUeiVLF-!#<c|4!t1>M)i}QuLpg8gq0>HgyCIxJQGlK6rvi)dOMSKA@
z5fY)wXcSFu(teVpSB%h;!_w<&7af%q@l<{qG`iWcl*Te2%s@kwq1V%MN%jm+tL$ii
znywfMvEeXL1X3g|dEqFaq(e)7#=8W89*cEPwUH0V821b1|Djf5`_V@KxBfW$PXYCx
z=l{oY>8C#)tJMj=HB-ysr;pbCOF-ZG;vaB_6k}&WB(cD}NXd!O+|T`>{3gI(b8@ty
zzk;_n)Cx@yk09n(#A*LfKe#(QeR+G>wZ1_4<4kpO`1oG16Y?XO4i>#m`7hcPZLQse
z9=~KcFOS%dWP12VGVK*C=jP5v1rh=RQ;aavPgTU<&kY_c1`xjaC8g&xe0)>oV#^VH
zQ;+Z$pT%_mxL&2VuU38Qga^OtmW(uOa?j}})6LB_P7deGBm_Vm&hG`GGYF#}P;6GH
z6|!l9&Hl^K%%HK|7f+~Lz4B*=&Bbtcs_}Na*#zlbMlEK3dCF6uL^x=qkJ3zm^x_D)
zKI)1morWo%)Rj<OG$C@<7$sKFmIAxP_<a{LGl?i(=r2?}&I6Tz9zY+r+PtLa(kTMm
zj2X6~72oWn;-02pcVu5Ic7%L<veC8p#WC$L+%2_woi}T^8K1Issd>W^a|jErfO_0s
zeP1Mz7=DWwF><&`Q@NxM7oW1^Yd9g=AyF)oQ<IsvER?x<)Krj?XKyMwc$A0EOW?dL
zPO4-Gh&Ne!p*ozY$ga{_SsBj;NNzX?OVCD1d$HCPOmD$=ueGHM`K-8tN<wXz`Km1a
zEk^Hc{6hy4J9m|2E|aYT#05Cb@NAJKzXKY_Tg#a_$N07LlC?7M)-B|)_an}$&avn3
zdnVykyIdZfvt65>x}?WN#7a-i(=ksEIo`6FQXsL&{kHVE<12e7GyUz&A;ucHH%z6z
z#y}AS9h=SH%NMP%ElRF@H__T!>PMfqM=!N}pNK{#()wXHHYH?Xwm=-BM&<q@4e>?$
z4LpGO{dW5)xBXQE=KM~dFg^rxsT?t(^$KkO{Qxq$XZCq%U=8m?9p5)eqh34<oe$|^
z4Zod2UKu1mEVBHDs+CwbXF0`{Sp`g51g(yG>rc|6nsH5TM_TiSrcp8s@im}esrLc>
zS&sSjFZ7#C@#Hb~@bvxeq1h50gPKqx$z<nXcU!trQXM@laLY3hnm7bZO1C)y1Zz1&
zmCnv;aX`!tUK~pVD9R*-O)D{7Gi_rL%xDg3@lUSNJ2)>~gOC)nBDr>;uYLhhUU)^t
zX2I^)=epMAGpMJJL9R)b*#q_R>mjyl0z?0fIyQ>1cDCT2r0gOMkTLl$fNN3uDElow
z8ytv|uMU=!NXs$Z#>yrPaM4xPW`5k9iL1!!frz~fBBmlJieQ07TM7;mR#hE|FsqC$
zj1EPgPo(|0=0}Q-MhI{^n~R#}2^7+0!J)y>gGRl~`5-gP3fv${(Sywa3`<9+O%_|n
zFk*CNS*uKz(#$}omjw$R++PF>dr>$v_y)$lFbU;dO{c!~rW5fLL#G{f;sc4g>EV6|
z;o7Vs(N>4uDZLJ0_1L2pjX8`9o`tz~KG%t6Mb@m)NCUjf5*&@3B0|UrhdXITI>g$D
zxy4m1SQ0o5XK;oOUCp!Sncz<GunH>pD_CB4*T&)Wf)W{OrTAq6;xZk^T;`P=C>zBD
zd7JKriK+sKyJ!(L_Y>88T|H!uMPzT4SvbC0Rf`TLh+rmDG`eH)5!);L`Ue??sR1_Y
zBY5mFcTFo1F6w`1o$ckUji}^+&85juS3h;{6MjrtveaXw;N`1$LdYi-Nq=FegBOY@
z`0~MUf7Mcdf3B7v7`jC@b=(3sd)hpizRbVDUkC^4K-7en{@Qd+Uwncw!fX#Z0<x}m
zj`X8MXPE*BnT{=jo%X#@jvF?FV!wBrw!<9>`H_}ibSOgGh4A5tvdwGrQ>9mK1_=Pq
zaWPC?F_?Z{qoCIn{g7}teGeQ?m}w9nJT(Hd{{Z_8Ac$8gCGQ5<LFL<7=NRy@#Sj_I
zYF5)c<uicdrf#8_jvSt?X_J-{zk^|0Rf_si8Z9aKAa4v@>?1B#S|=}s^8vy&YIi4G
zt$qTNsYR@uY+=|hWg%Ki(o}cH%Iwp=r$9qGCDU=4)21d_n-&rcv$*j(MLk`35a$p`
zs*+$<HJg|UAF-R7nVgCgn$qr65fADpwL;VF2{)z)Y8-NPOX$yM6gkR3-MS(u2dbER
z+ef>#oN@5^f&l0u&0{R5_wudy%cZPh^KoUIhC|NijHKL^Jo^yeeU->|-J)L?Y9<D5
z`a@Tb^noI7&@_BX25?MFRrwSIBEZxZ@Z7$Ro~A-*Y|TbRj%*N!AU+z7+)@ciOpbI+
zTs_>DD8}1!8kC^A{F9)tm{5(Q<WEUw%e)|j8UlI0$Yx!2cIlh%goXCm=^-<x7H=yY
zw(u4X7^(8-MT2t)6_(TRgUrT~$P;1Sqv&p?I+&JML2Vs2s3W+!@X&X3HfYr52#S3z
zd@znR{WC)iv7|~_V#@y#3jH+%LdsmqC(2$%bIaaxc;!q)Ot7fozW8%BzHRGD2^_3U
z(hFc*S*&<c>NdBpf!-Kk012XKy(^>T%7Kr}8EP2qKFOMbYOjdEW#g>}pq;;<=v42-
zncX^~3NEA#o8jDbDJvQ@{k7?t3GKnwQ-vT&H@rN%A?*x~LUrt(H0R@AF9=bKU?UaW
z$JFj$%mNPmJBuhKA?C7=CN@WGu6<tqi?gS8Y64eM6uEl6s>cxbz`H+t7I=wDaYG&$
zz=dR4--_F*Hn1Mug&FImCKe)GZtMHUbq<-?OP<^Cum$2Mm<@OO+_MMs#6^4ntQq>D
z9P$mbCHSj9Kv&rEQrc5KPtK8fP?SsMnl$TpZX=+>8b>VO+|~u~=bm6}V%;i!BgIK;
zLRL?%+j?r^!=<WYO|>se)-ASm#K}S=ZntBaj&#<atr^Gbs4!r%9lPM%c2m1U2{iX?
z;Sv%CsH1R8QXAjY6&T%UcUQDqkVz)6FJGVw7b4-9JIezMC`kz;^O|Sr9Oli^+SQ$i
zv>5Ux?3j_wmOZob?><1Mol*~?a{mq3o0Z*&$m<MKtcl$B%&doy!wpFFaD0|wfY-h8
zXOd>t^$D`2s1Sj+MDh~3BhYUZ`;m>v3wVv8o$SKk2yN(N)_mCX_}c~r5Z@oIGwH^^
zwj)qeD=6`XNLPx3yur@8K`ImDJ1<_cPIU5GgnkNJM%)?mce5X#b_dkUMac>chu)Y}
z#TksJE_5lXUEXmV&q%F~;AT<-Tb2pCC2{JuexNJ6H*U9z?zB($In1zVvOXzt%xm?^
zE9@PRp>v_oUJBbOTCnI`uFFtS8v>;P=oAg^jCOrG#YN!HBb66Bdc>Ew?zNOWyqAUW
zK5nlRJPdT|EZf9bZ9coV3Ym78FLh$DSO@rSdf#KP!rw|rMA;4>2vkt7Fk`R;k95px
zF69tx?ZtFFr~*nEykdW|2BRl0+D-S~=*vbIO4}J{Np_aF(hQ!w#G~gpza35pEmr=1
z_EdV9wSYr*w4civmw!|Z*EN|g-`YfcB<y$Oyt1E!!lq%-(Fye!1p}fm$VK2)JH*0&
z{mewI%vqj%d1447(0HfBv~H%a4URm=-ENmwrI?92Uaj6kS**d3d6LUw+36qj<Cgb6
zZVI(2Ikk{m-&>LMQh(TT02F4S36{zrpA@49FfqlTKk{}!{V3A^q5X0LFv~JPt$eBW
zRV$`yo^@a45{ouF1X_jie&m^LpYG+{&30gT<@5ZzI@>%48U_2Kvx2_yHELXQ)P5cg
zGU<rtwkm6__4!Nr^Uvo0kR7u#u>QB~n3<XVe>W{y(vpb9>xA!_sV(DYXuwc}oi#7&
z<TWB>hwK681pASx=LB`73V-=hglkq-V8-HindNQjP|=bO9g-BsPppX0{mPln&DrY4
zQO~+5%hKiZ(LnoAkMt%inmag1LR&4~ovz=l1SJ9HkEMUb&11cByj`B+#?FX(oaL8=
zDvp>zQa$y~a*O<E)CsbA59;4oePd=cOaB%+LFUjP%sR+tF|?ea^1gnEmL8r&v(o(=
zGN%d|>GgTLPc;7dYJpxLP<8R9;V2hUlcE7dqEznvZF&#seb^rm$*+Ig@pxe${lEF<
z$e)NE{l)#&zJOYF*u*rxkeGM&um8MEl$N4LD@N{2-o36b)WSME8Ax34N*Dx=9QI0C
zVvNuBiolrl&!G+EJ-WU|bM?GGuh9FweYpKLU-dIoYxkcm3)Jg56#KJVd~fg?lG`nB
z-mH-X)ubiB0u>&>l@!ptdw&N8m^(Y^T-GzB9@|Vj=<FAT-p6;Tf_0V&`O9HsLwo)?
z&yG<6KnX9E{L#2K-mJce?@yGyVDW!Cok|vdwe$FB<nT2p&uN~QMLq)veYB2&TZ$^c
zD8x$UBwARo#ojUdFv`Kmk=-=Ce60#GntqH4=5;v~@7w6RG!NUGoV`=YGAbwKGsd4c
zdPqXU)hx>;BS1i-8wZ9BwsX*=*rnP&zdoWze2Nu*bzJkzG|r0x0S?!M2q`N$)b7NM
z`?LtcJ5#r;`$|?bC6<*{DX;QcYb-!CulDET6JI@TqkOuwuntI(cM%NG8rXvK1oB)}
zptx<@Z}Mo@0&0W^e?gnMQXe??jh$7TjEU+=*sM{CJO`~QNI3RW6b1=R+LigC?RTox
zwk(sEQsH6WHu-8&c1R9d6}`sJs16e%fr59<hL$PQ*Q?A_N2sx<PoIgG%b@7p4Jg}4
z4aF&YKWLDsq;b&Aht*&voL?S-=J#xC=4SI|MWEfz7lo;h;2?lGM5h);7|`+NR#eLE
zCv%R}9~VlTyV<#S-%D@fu_PI?<~{^8m!5zU0FT2oyo-KC|L$P(l>{0=l_CRg2FRI=
zqs_q7^cYYc&n=QW#F~}#J}q3QN2=xgI>Xv|UC+|7r|nRFQ{_UxcdVdo8BOwgHsFXw
z5aB;hlReuQag@rhSRbT~p@wqj%%HiSN94I=SAw2IzTReiUc~E9@v^KT>n!!7HDdpZ
zG=u+_LUw5ZZaRZjIT#_Nxg*qLD7Q|QIS%_ESxacH9KQURfNLF_Uk2k}br5)U)p5q7
z8o?~Ijm=@d8#77=v%!Sorha*`T1yv~Y3krKM&Ow>xm5WygvShcy|p*EB^YhsBC^Is
z*Q-<{Dvkh0?_T`{#26!y1%<8<n{4H$v2H{uez#z4;#4J-=0x=}O=tS?`6H3jx;2d#
zhzDx{7H)7!7*nlKW13S8+|j`Hb*)&yj(Ch~4kZ~$I+GJ8RUu;b-~`yZ;f3%-)mQo^
ziaC9^VL7hke7T}iYY;9y6&yuW#cD}MK0{lwl<Q!p#^_0oSaMej`gr0ow5P~WEYwxE
zuv=B&Q_W6I5M~0<`JW{`s7Nqq0wpe%yg?^sa~l_Ev||K>ZqslJD+HMo*5}6bZJw4M
zUPgA8BNmK%*;}q`hlA(+EbB7kF4n3Pq>?w*Py&<CeDJRSi?Mf#(j;oPbknwNqtdo*
z+qR8L+qP}%%dE6*SK78u?e2pfWB+G#$Hlr|7ZEY%dgq!?#Xgad%y8zZ{BU=g3+AHM
zv<GybkTAyY^G^|E^V|m-ahhUOX28-sWnTTDQS1O#aAueIO_{RdZA;<a0K9%uq!5GW
z-whF_Jk95^HLl*_JkmH7WI3B~)S|uKDwZSI_;ta|t#0RVn`?LY;n5rfb9?d~tWe?e
zZd2B)7d3;L97ofz_<0m3-Wsfo;ha`JwWNXhzlB0wJyxJ#q5Iu&k=d(W<>GRqB7^1A
zmrirmmko4E;&pY;os;6XbKS*N^fdwKUZ6b5tA4d<;vMSaG#e#5eb{zXQJ?bR%P<xz
zvZMV(D>ITP(kF;;g=J=0OhAGfsxp4N9C#=vQE^DR&BJRJNln7Mx28AEhADXkv%N(#
zf5joC(W$5NHMb&Z{?;ZS(6k&5IdOzCn;iaKAx6)Pd6(Nf3IZy!FG4xyrWj3a5}Ti9
zE`riCeL-)kHN}9XT$#NXOC!sw<*SHPb8<-|tzgjU&5v0YEv(4NCzt+k#Meh5AY$a4
z{{+2*wnPf+S#AlU2~D|(#4Nx(Lti|?t-~D%Ry%%da)u@Mq7fv}tV*wwdyWWdw5*fD
zba63>ns~j-ONkysCfl>dl%;RM*IpNwJKxIq_5crwv1g*R81B{CxaOwZCf{D2R7X=X
z6uaqdX_h-=9;UjB@g_|5BybbikIDTY5wl-ch?9vQSJPkCP_3L{ywRh#%t%FNx2fzN
z8WwHP<6ghRqgHj^Ot)5GG^1LuL$j8?tP@KoOc15LH(1C<2Y26yMQ6Jj$M2fjTl%Q2
zBU1gW!DSgZH&AsTzDOgJu}o-8qo5${yssbLr8S!=JF2J&&Zh2l0qMyicdoxllBTMB
zujt55c1KG0Q#a`CE(cTXQO~axVPLoi-yrbr;~Z=`e!2f{1u<ySJKaZy`+y0~pHS#c
z)u`gYk<tDLZS?W2zSHK6RO=$=nomX;dLpwQ{wQj_XfFfJOiz1soK^lz0u>BfD5tRB
z)0`}fMd5T!%1Ofg_j%d^AV~HUC*(DVm;E^gSoyo<?=!m~&dHFdBZ~`>rw@SV;m|}m
zL3}KDlnR`M(?yEV!{!}tV_TTq_QbNsmZrch>BA3}s5}a@yw5M7e9b)ZFE!};X~asq
zq<|Q<R_>H@a*#5g9P>@4Y)M|m;<i-HB(ZfaC_g11^kJLa2YTOkxiq>38~bF)ANz$*
zxct~0U2GMj@jhIIUX_&euoqjk`8I$*Wz{7xJJ<{0z^GK?IM4#^pm|=mVvC10BbG5R
zX)cXA{|Xz#S;Urit5A7$!F9Zj%2xdr-(xL(=T<XTX9nm?HAsV9nQKif#wrN{R(T9x
zJU3}rThJ_Ld*ET_(C;rCUEk-3Ori7jwAp|Hk9FHUN~?jua1Z=G-x>J;zTWR{Vj1G}
zXpLWQn<~DE4hE;g)W{4lpU?3}lZ&HXE8i4K13L@aL&SO1Qb}RMRHUrxJ=iSOolrno
z99|LyDUs-(+elXC==)4QZw!U~g{2a|^BCi`=8Kmpgq4BCfLLCxIfzwSe*eh^10lA{
z60t4}b_j3D6);AvQHmrrAH7oh3^;h)Q^G!TMIN3-E00#Tc8lP;X+bFAq=3^}`R?~F
zn;wpqjrx&kfu~1}(rC~Iw*;?li<W7Bq={n+D@`9E{0Q>pZ>RT<<FmoDb{|S;*YS}S
z+pzEBt*hLCs{Uk6j9L<<jW3OEkOMAtU-`?*Xwt?R!6wBopV%s_#V1E+&{pY~aJ?Rr
zdxA88LvY3%e^CiGVRk}J@#Jee;JG_7gGyq+7V~~of6*K3fw(-o^5vAuap8}M{!8mc
z1TpP;;U^reat%s(-|7i~FrvAE&K8e@BD;&JOVgUJ-g!NtgQzt1`!9b~d+}jyibC=~
z2SfWe0MJp&YFdmZF`1sbKWFGH`=#CQVFj-(*<9qd6@1XFzLxH*=V8h(c(=A_PiE)!
zd7eXVvR%hHM89^;U&10E>Vh*7nO$eu`48@}Dr4lpqP0uS;IlHOgtJSmp-805dSz7>
z!R}`WraOVb_WLLyf2LVhE-EVIEeJ&Bl~ra<Uje1Pp<12kRNatQkEaJ%V^_DYS=%tf
zGDEe*{0ze577rf5KI_m2xmo#+`PWp$x=qp7bfA`(^D~A&N1#{H34~#u_b$lPVA-^i
zFc*U>ZRi(p;DhL0?jJ2JO4hY}ID#@Y|A0SQh2)s0{y@%o=H5Wu3ZxtjF6vAk-rI=H
z?^k2)$A7Qp-1F|(+iIoxVq6&*9FCA{p!&Zh36zx5db1dtMm?Zk+5i>$J*H)Y@rnSa
zvJz*hr#g~ND^+gs?uK(cW=aRi&9ZfeLpzll{D7|!F!?hqOba7`5GWN1YL(73mgmQ`
z8>Fou)=Tr(O~V)=hldR2$_H$w$C3SF#m~PV72)u1dbbzME5x;6=`44?bk2reV9FS0
zdwSCnhi1Ng9{u%J@EjFh=Ch>jw7NLyI)F)19^;EQvd6z&le7dEYq205*U}M+5mCu1
zLQRy9!E?`a+@#a?+^4e+WyZ$G?E%kgt&fqQW2*y0fhhPheS5PK@_s_FF1E%kCnDaq
zPDxNEqTmHeGkl2;WNzq#;+n}#2=xcbe4#O-wJk|mVtyyo2PP(aVUoKohy{@aM@3pe
z$(@YS3J)(|Df`VHzV6yxM8;Z~Ki2uXm|WGbu_ULmY~{J6UA|vw`&On$yK`LyWbA1Z
z<}M8cEF-hmY36tw`!8LkxEa^#FR{q)dV~y(3cvc?UE=^^zDYB;kVtFJ<?A>~39Ui7
z8|@$8?EP;4=<y-V%GP|!tO3;z)TSR`xL*fX|H*wZ{dcezI~Uvk98D1YZ)Ov*YU5G3
zOh|wijZ;YN;-spUK&ySGJ3)KU-cY<CBZ_dRG3T+Ha%tt()Da4kgk^8<dUm^+9vq9j
zveCsyM>xx*X|VC^#c1F>$4|t#en|@8HLO25zm6YfxClG>oJCHA|E9h2xF}5G5wU<~
z05?oQ3%x?-f!Zr|$V7FQJxN$iUiyKjrMPcrK=nin6%_c-7H^sRx{b0hm5|IR;{7s^
z&!<>*dH2j};}Rvb6=FP@d|YI_8&`ckw57_NB2+Pg{T(o}j>gmojZ1ZG>CbexH+nG{
z{-Du7_d%~}G3OnWG+hcGY3ojn`LHs&4$^1<+U9s~Nm8tuQ80kcnGNk{e3JtCC^u)!
zRQwXm83@nC<mAREh!d=-tMv539)z>JEH-TgFCuPSYGXF$tv0T>!KqGD-J5wfEISLR
zLNk$-9bbps=P!%#tgbRdM!9JzY=Q|jm7&3Er6K1k;yKN@f$^S{iPWXbx7*hvy;uO$
za?vs$)*@{x!@^kg4c{yoCDW2#P>{=ny3LO$n??)Q+gIN{efXYrX{htT`!%zG-Z+<7
z9%pn+P?ep(mH9sSdbm@*9MDrDz2zuXSV;SAaqIk?udg>_8(T?Duh!)4>OuE9M%?13
z&Q;i}5+w`^+dyCII%R=ST(6>pfG7GKR)-=?#>ijO#o!kXPXfYxiH-IeA4gNkW0tZQ
zq2Ak5mL@eboGaDRH1v*mqs(((rrnRgzX&WEW7^oy=Zq_r0(8&Wv`2l+)avasg^I%x
zl^#@A^lRhqH0;l>=AJ&51zf+s{gCf?59H5JC$FPa)vl{VH1Ko|TCH?kn=|~*KCJDF
z9TZ^ByiNUDhjgA=NFpw_0;<++N!(yX*t$xW09S6}=R8Y{?Rf^%K8WiSojm1N-p(6V
zANoJbdCvjB8o+7m|M4Z@_-{uLO#jd{|3?ElwRix+03%$;3xB8q&}u-V3bb26k>u1m
z60K~{PgEg&(cn>%=KHnmEm)9@M0LLQ>TD?pEk?YfCF0*4#XRW{UPevECUr}O4*W<}
zi91)8B4uSJf+dmw(|60B_6r!d6l;th#mcBWE{Ztm@8FR!*2B^zL?&|B?K1N5W8aT-
zfZL4)8&plGh&!P3nE4E}=``&R^xhiqOsx;7ZUl$Tf7<W=3Ksr<8k|@;nf{OV3&0t-
z$LqeUt3M@VBzyr!3XwD!7qmz4g>{C4TjN?2N)z!H5NnE}`?W~A-T*(wH)vj4lS>+3
zAI{6aCFRTS*Y$BQJS>8}=L_}y`go*xD?ace4S@{c=~)_(^nChyy#S<a$XMn7os~X=
z`viZ*+dCuJ!NxTThd~VZc(f3`42@d}!#6$&YUpPlijSZoT7cz6=Ou)0e0ZfucF%1w
z150lB=KzV=!3!g^Ks3SEQvEz35i_v`gF=7D=vyq^PLc7NNT2vgF06qz@Ss1YnK}-<
zii*k!Zi^vFp(n}?8=PVfU0F$VNc>XX++}R>EBF0Z@6g6VxUd}vGk_%^l1rd?^t1we
zb!No~W_EzvxwHdyg@Mn*+jzf}8b4)qJl<<cAe9~(u<Ob20m&y3LW8lpL$x$uqGM(8
zTlwRu1xGG?<8CaK@EOsfcDMKnkN`Gh(_av!Uk)%7UdVtv(rt5E{vv@%58}`vZGDm&
z{S1k^@NdV#_i<5KH>1o13x1kf&WKLr=^1d$0>jzL!=?!>XT`2ee5nvGzE&!mx-$97
zIS^EG7hNmz-OxLfxy(TJVF_!i#xp7%cGKDomv6T$ufw@!1>1J2T+g8Ssdy^toC+1S
zM{^luvP3o@eOgTRoC{SKb*awbE4pMTAx>g<(Re;t<ggQPn)3<}=a;zRWtOo?*{hY`
z7Hbxx!zc)HtU5{()+=_;%YGnlS~!;6vQsj7Rwk#~ur4gj&8&J(atr$w8*jNQiB<fC
zPeD;gy~k_Bs>6;;52E?hkVzS(^3g6uHda7-TYGv<v@XE`oD1>NFU{Vco4*tK=tj`f
zy!NXn*2S(!(jUTG4X%pag0i|wy*uPzOpywTX3I-SaEVNTYFgt;ER&K=hhzTq4%*F5
zvF|X}KbFZF{!3SV*iXGrqMJZjx^drQ1(z;Rv*(PT#}g0RHqd&)BR<<rhtEPX%_d9F
zQ{#>m;M990Ba<C%cJF%e%Z@PKEN#Y=+Ss!#nO<3!ibPlH5=+Ngw4{UmpqvabN3{mv
z`D=IigIF?U3h?Cf5AGE1q)sN|b#wQxL<(i180(2RG_=n$hwB^EVP(0`mlGO8UC<aG
z`b(Ew(i$sx_$kW*>8v%1NlFqsAlMi}xKD`blv<z!ZA=Db;G|w-Lhuxmrc`a)sNy31
zcAg5OXL)%mVJ<+nP|=s_Dj--GQn*jCFij)&xl35azKE*bxl!d<7yrysOtroyE^K;p
zL^9lQ76O&Oe#qzBeypQx6|S2GyRp$5xlzf0dBPje*gAmC(HKxY;e^~}BmiutuWg2;
z>EHn_WglhYNqSTf0^<b2tXb%4*N`Q{^tBgx*hPF|j*s7|Z~5Lmx2}k*L))T_`ZsTk
z3@}Ht4DXp1516TL+2o&nV5rU0jJaXcE0bmW!p%uvvM8i$wPII?S{d1XVGqVub0Z(e
zt5I;pK!n0Un7S)!hjNG#Ztn%1{;kYHG(PDwq<(UoOr0EDlz(4YqvCM(UV*r4Odh3}
z-c=9Zq^TjsR_E?-mTyG6qtV#7gT0#!U$dm?rtjm5Ad-qxm+@s3@?`{+f=~9;5w8RP
zgf02ot-GVC`JZx1_QF0``^dH{bJr$wp-=jbAm+%%Wt72RgMu2Ra4RWzgCd1v*;|2V
zV2^3v18?gnH8Z;-&8Ggxda1w7)o+O(pyF`Zl_2PX|IIqc%O18QoT|<+EiPo2+Pm(5
zuSd_tww76LzeE~~Y|Z2F4_n28oAr8;)w>{Ve~^$$5~?0`ISZaWfn`<YZxVx9pwDVF
zLQm)*$W@r6!%4Ft*<;3H1qW8ng(OX1j9JspTpqMNGOKtZlKO~>`T*pr>PeeE*Hmq)
z5gQvE=A@eurX)_1-T5eJbIH9gQ)^rDLSM^6gNozMf^IQs$ObPG&x)vCyA>yd0(<(L
zX|^{nEd0TaPC&J4ek$2pQ24f}8{1x95xBAo=hX{Ak+Z7n#BhR=zn%g;;|R6W)EQ&H
zG|ih{N$Wb~hQdng61;uu>qYEcT5tH~9}d9UVo?LKOH~$rQ>WlSlP+df2&<sOGA-tt
zUE#S(elyG@1?PF@+B{GP-q>hOO6rEqOQOnqQK>g~_ahiOhu1yos#E+_C%xy73;#^>
zR=e4Gjj_%dxjJ#66#8O&siA?`OVYO2*6XwS%B<GkVB%<aHuZPC0Fm-KJ^f<#>zhjV
zDv#6$xBv0OyIU*osl9I5%*#R7=t;)79T25jig}nHT%?5JDUI3eH7d<%gYiZdfUYp`
zr0!};iJ$CcFt($u0dis}gSNWJ@ek5D%}H8dX)ladiPmlaR6nhAx8aWrc*Fb|+<tde
zG^L&Ev9l%L<sLJ|sDxD0V_~HQrm6tO3c`VpIZ&T{Q>H!bSbZ=ac*D?mDa^i*0z{18
z*w;_(T;$+3%Xee7tGs9w4H$PH_<s883+#N^Ssa!(x^+}?b<!9xyk)y`bg6F1p_I!-
zE5_PKE>L<PGxR{6$If#DcrHNl1?VXecPe_g`2+esUVzyj_o%|a;w+V2M}Oj){;Trt
zuS>s9gUG#*<7BtM_tVp4KL3v%G|-2`4h<sgSE2OXt9~ChWf9!pyWLx#keNlBTku4(
z!=1swDw&4gK$`<62Q$`GJn)a%eos33p=c<?k&K4L_j54p>>#$0o?}>>H1i)Ln32`;
zMPo~AW4p{4A=3uHc?I{YpQC+?>Vfa0PA^9t9}(JaJ0A%;rxmf2$n>H%9@=UkJ=k4J
zWm?+Iz*Tpqd{*S(4e_+qx;t$Z;CQ9rW@l7Es_gl&$~Mf4Fv;0zK$u1a<lZwE8jp<Q
zFtmPwdeXob#4Dq=MKG;8Y+C{W-*Juxz~`Vo>by7iErZ<o0)fOGDNghCtL~Bb?Thw#
z>o&_;CYGhbzteeELx9Hoc(HiY>OAxs;_KDJln1yzV9ci83TTOc6Vk4_I+k`2Ba`8o
zbZBptJZ9}vkYmW&RqkK96AYn<;CPR+zIfYddRTF=G(=0UNEed@!osFBVP2==!ZoG}
z%vxi=FRKR>uRB~Zq8E{+^eAW1M#A_VY3PiRx|NyT{CYZapZ=iB?BuQBt}3ii0)GbR
z%}Cl<&0Xu#=+ua*JJ$f;mdw+6igeWm9Pa|)lPkNtFLE0Evvk7P#O~Rxt{*qqx{X5}
za0+%mQDYtmai(4-1e9vB5&~2(DN7yv0WGvwviHK2e839HJ93=oaedL^d`(N1kF)nb
zIN}YCo~yLs$&4T{A*|_I7=Oe}*s|Sn*1rUb(13vdgZyJ>|6e@c|KHsmCPpUK|L^h+
z^Z!}9=d_NE!{!+JSDt~8Vc#Xjt>aF-G4h`aUd*^F@g4U~TO>wiQkFDHda3nwj4Owm
zPSV7-hZf0Bs)UU6!`qom9*=7q@d28=TTxB&TS^pWC1vD76gzxm?!q4;+5Jx90gC>4
zNSK?@s90YTwD?(Jk`{EtkmVr-?az77J&?OOW@I7MfX~F&460m7Y@St$L~RsIJ5(#(
z&lU28Vg43{62zp61S)&hcSbx%^54uPN?D7T5F(6muh^+n<p(Ue6@NgLD1=xBjBSfb
zA4NMwZRBTbBwHFoCOj4Ui{gQ#I>xyE!ZJVsV`15!WwJ#kCKB4ia}!UHk3_sqQI2>I
zScn++ki*Q;X(XaVrfwA=QW76N>7*Kh4Xq2^#-Z|Cpj^-whD+jMHs;FY+}>p><V0oY
zN#;bl9C_{IQC4~-a^6_+_msR9GxSLJs6K|SMjFK95q)Xz?@;WVt<8`lamJ2)kU1)_
z1pgY;w#KraL$*SaV&qS#fm8j5;=Ale*7Q9AUz5}|+0FQaB&mYuHrY93Ksg5)4l?hf
zP(peNe8p9kzM%|?zXY1_npUqCb%6fD`j=CH*DxqhAL<t<AL-L!$&gY^!NH<nN#rgq
ziBu`xbd@5RmINv#t6Vf{kmP~}WkfwwBAL1SS4zfw&0J6YE~haG9(4;THtY?F6d^cD
z11Ws^ma+~`4w{p0%XtY(NSY}LTP?ao7rh)FC_RW3sdiG=<3-yeYt7~Yi7sN&NWEzU
zesJ)DCU^@_>hpVlyg*rb0NsOYA6+%OR6x~vbC17RSYV%ge(CHY=oZky#iJR&odNVL
zcDQ!|X4so#L1WC~!T`F=*;BA<llLad;9yaY<crVqVq3+8Ytr}qJ=fRIcaL5AbpE2N
zddo?5yRA0;LfJIKgYz>#{mF9E*WTVB<ZC-Q|5%6-;Q1Bpgf4`QA1hYr)n9))`qFc~
zOd|Zx=+11o5E(K=4e%K`s_7^@Ns17Y$<#Aa_3;aN^|=<eqNzNiluB3sVr0zipPg6~
zt*1SF?|gML9gT-a&-?UbWMdWC{u=4gOsLf?$Yl#|+e$CN&M9-R!7})6nk?$cS@pBN
z?}r-?FFj27oz|1f9zD1anpQ}%Wsxg#m<r90Vlr;Nz7FB+<gZ1GWNZJ-)j@D;`iFn5
zhrG$>VFbp|TnrM0zJ1zsFEjU(k7Pdi%VIk45@%$KP|PK5a|^hO_IM8qS1?CMoG~!8
zIJA;hWro2%F&m6GhbvWHZf?VAUQ*&=hC#mFR3oRw-~=HW_`Ujs;eS7h;8p{7XBP|4
zD>8%A)0LS+kt#5Yf*glCBJ~9!s~^g=Th+XkH3%u<#n*^4WDqfF!7a46GHfGHlpW%m
zXVTP?8_~52%`4zF9DVk1R0Rfs&@bc6Jp&qZ2=Tw+C%s~p@4~P}-$?$5Z{RLYf!53e
zZLY~ilT+_tK}MJXjjjrvhe8@T^UFW5pDv8|KQPtJGLexE=$OB@jt~b8O<<iAl&2G#
zKcO&Y7dAD=2!A8!Q1T>^a~N|(ad2hCB;bES4iC$dc6F_AEQlOZ>ueq$+h#IpCV6L(
zwqzdFOl=l=Q;yMX+F0ToXF79?B|&zRuJduP6VmZk;v<a{7zuRq2xhDfyL#oN!n1*m
zQLYzd{4h<>ZDvqQhpB#f=K<$F07oBQ^<*_VFDqJ(WnueDZmOSPF1A$ev}$O<>%i+#
zHNQFc;@Gy)Tg<!?xH1?7VrXB=g2p`~N<gg`nGG7^&QL90gTqewN{YL2^tV|z$Qngf
zOf>axPNI1hw(}&PRpOvCNK3z^TBsVGiNTDGy|=eZC6Iv(kyt!@_GILa+L~89BKau+
zEI@11caSU%QuMzqpNg=4@kh!sDt>{kFTsa`nRXN_At}r>cUeF0np>QeOmC>Mbkb~`
z{ewk}nAXJwDiK2=*L5>(V+G=1X6LfK+a~7K9S`4Q^hiLJ%K;UnpVbHaNbHLWhn6|B
z<@d$GCp{KqX*`)`$~}lGb%RMMM*pdkJ$!$@3xB2_CmNO952hU3?|+9*UF$eWd_$!H
zC3!8$4o9SJ%agx<R`b3;mGaICur%0BSb;{dchzb0T4z_CUrZdH^u3bHcH>JR0F2E>
zU6!AVpjDzXI|=oUys-7khNxetE4@#5q2P80+|sNVL)+2~+e2+4*OB^4guw54N<_pZ
z;1_Zi9b9tH)cwCp@4+R{%zs)65AaA%RyVCn#e2zl)#|WEC1aLiXbvERpha7=L%$!+
z!Rc|C-yDoia8EV7)#Z%-URY_T$YoETVBeWUH_z7A>R|B;kTv#d;WAhu{rM18xkG%f
zb1aQ&*o)!<M&D5Q<DbM|D_{Mnbnllu+W#2|{%ihgZ70}BS$l)eFi+S>a>*o#O|~57
zJ9(%4WIc(Qx?f=NrTs(!+oPJN)~JywQO`q1foK}1Sj#7|^<;GNnM>XJ3jrd?97*NL
zJrB{92Fe;){9IIKwg)-zyudpUF1ofA2bc}xZb8cpdN4J0k@pKh;lZnQ>Vj{qy|&t#
zX=$9uUX?~EtPZUNQamke#Mnd4V?#>3HY8TaJDin9oN+Mmn7o_jR}{I{UWWRjo;Pd2
z!{&>m7gF8V^PpfR2{jtbSc;wl0aYCB)pGwa7|XB3+9>Dxs^5jWXh*kQwTb6AgwCSi
z`7JWOkjpd-lfIt44Li3jvLr65{mQCoR9JQbCCZ6Hz^iN{UEDfyI-19ts`kp7#{hUh
zxYllQ+SzTX=tnIIjC3i8iB=#-s-s|7JS=tobzjC%D(fW1^esBtSMVtiO-3cqDkm1v
zvIA+snc_upX~@YrvQ^ZC(S=74Rp{h&w)<5CrNc9#Lf(}CKPR@5!HQ%j+0A@Xw;H41
zOVIP_<J;mv@hIsW&UMB!fE(v1WAR&m?z`pZ{?!spwcje5VonGO1CAL=S4>wb3ER#i
z>gn(6>Fc2@oTjD}8N{j~;19RQ3)pk|s6=s#pyw1nxe0CyWAR1N4POBtFaMfA4j$q6
z)7>OfH>b3NyB&K(<v#k$jeC4X8)iJh=o;8rf`X<~EJp8cf)W6-CD~%s1Wm&U*s?CX
z3MtLeL!jE0**THAH1Nthk(xFG>D|A9ARmwa+bzamPIV=PPt)1tK96o7`s<{ajvnt9
zJ_77-Y#Pr&4lJwvx)jFqA7XK)JZk?~MKAEu)IsEHBFTR4LJ>J!Q9W=Y$_V@EYX*f~
zJdDL2+k4Kz-9*t{Jk{voD!L7UE|2LB(fBH-q?u*nd05vK8R9Ym4Y`_`s_MtnyF1gQ
zRTFt@MgIMq3t}2Vta1@+P~tSN<x4gj*|8h@bok)&$*_ytL4Q@A&z4jZ*9@ip0h7jn
zZsJZ5qnm*wcDDlQ7u^oo{5g!61am?3MW=uG(qwrQdxPN;lXyRL7T4w8hRWyp%u#q^
zZeawP52;vP$ocx;p~ZvU(sSd$+?iFGc*^va_rsLlE!*!0Q$kn~0A&6FF0hc3F@hMu
ze?xeC*hDw}+2*~n0k5(>-`03sq$Kx@iW>C4$*I^%Bf>!d9PRRjJg2HsDUtFv6m%wK
zyJ^NFh~<@IG4|OB>`kbqiiPB8Y^fXNlIWT#b1|;XL6Dw^;r5#{wKscx+08;>F)$O5
zPWV~8&YeKr^y^Ry=9o<a>ZT&PA8ZSOfVVpc=v=PveTeZ<c<yM?;dh62X2v<?t<ZdT
ztl9(*E9x=8PB?Xs(btIWDQ6)+Q>gIPHky9C{+L5Y!3Wm+O0iph+{%_wi~N1OxJ_SS
z7zfVdgdoM}nh^ac)J<B|R>EQ1fJ*-57Z>@>uO6HFO5YncDQBBiDl>_SxH<gBb>m{t
z$85?r(s#;U(g6$I%d~D`C<81zHyRPx+#HDO#0I7PFvbTNA7!{5&(Ac~T`mwwF|A@I
zI>j%|gY({n_6?)m+Z*jp&E_`eWWCAT)R`v)o|7zu@KKu>B_q*2`o4uu?A^|1es>(i
zrNKO$>C6&BcJSYf`u?RZeZh&ejr^v6J+}x-UWu;Fh-&W!zT&yd`6OwebjW^1B$NuB
z?Aefd1V0OsU->O4G@un9OlR+w%_TZnjFMm7oPXB*`gI?j)f;9P!i47^GqZ?nzPYy)
zVY~0()i%6dw^1eTxKh&GV}~3Hu-d4xe`t=s+0`{7b1)|3;5~DV(;~yp+n3p}c9vaW
z=)=Qr#zj?Mh$AVw?E?Gz=q_4zo#eA)|0JdPY|x4|g8v^Z5C<F6e>?GKVPyP2b2<M)
zhYon%$p1ozrve4+>H82+!r~_Jfm^nx7>BopLo@vd_xcELAJj==O_J`jB-lbEhOE}?
z>Bfy|^;J}jz6L}0&Yu2!zMM9iTj#N^)?Qz2|J97!{;wH#dcQ=sHf{1^_J|=~m?}Eo
z;xiCQ`(Si$m!NNYG5EilacvXt&-#DIe8oTR@JW40ekr`tA9Qn$46A2*cf5LJt@Cny
zt(j{e_;5oyK}6n|dA?s<<Rb>P-R;?a?rWgtEbI6*AgJe5UjwglboBPC!PfzClIEO=
z+z>55I0>ASw!J=M5Y_s{I;><Y3m@rh1pE<*p~()@V}_wUDGfzB>3Kpou<VOa!@Z#Q
z-y~YPhh#SWm`+a@A)nt+^JGhc`lTG6Pulo5Cm&U>PBE8C3>e6uLUsulaAziPWY^9k
z0=UpL$x}r`rRsiUO_u|_5@VjXCft=soDX}IF<uudV_R~Zw6VC<xVGp4(z`dh*j^KP
ziy_64xyAxzTPhB9&hfR7*Bz7Ixlrsc)azho)&{XqN|^UWtIPSbZP}?t`{iuKr{e$u
z00Qyf9K<y!pZ8BEdoSSSK3WSoV>0+XTHk<@V>Rnfp*Aj9d3AOY3q->2x3A6o-fs#G
zyyyBgHM#~G(!@-GGo|8Q>Vq`GM_Y|8Q9+!QKz1;U+iFp27dJz@*DgFkF5PX<m3PvA
z*A{rGh9PV^t#6D+n}BR+SckB6Xlq}BIn=ApQ6f+OM=Ww~?KR?m#v(sEU5Awd3=nqH
z1}*lh8vkeYF9l5L1CWkxDy?MragI|PD|HJD3>jJ-Sh#cZ;>62lSDj}{fm67@n`QZv
zvPQMfFM8YEQs@&JmXtE-jk(^zaHIM)<<O;qX~FG~Wnu$GQF12^myWNmC(785xMKY^
zkGvBh8bDCvWbiT-l+bWz5~dhp%MpBq&zT+o_^d#F<6kb4d#u$a11SBYhtc`ImtbRr
ze|HMb^%6qiikn4z@lOVkQetSG?3z=g8RAkmV?T$0%9l8sb6EzJ!PgNbvnJq3J$RU7
zJ(&D?V~b*Dv)P+@Y8aOM$I#orOn#Y>aNKutrcdr$5duUg$ehdV3$%Fgxhx~*f}F_`
zZWkrky+jM{yYX>*OjUqU#}5>YmEcLf-6h$oJ(F@a6Ljih1~2V!%C{eRTnzYS205t`
z?3vR5Tb*W&gtaC?eZF`ydJYv$lv@NW%?(`U<bwOsd7n9aE>0%2$YEBuy%KHR_u5K6
z1zBsucNAX7-3d6j2h%CA!PC>7w_KjqcD?oE1{ez8?pqRl{FTf^y<YwuX5M1%pLExC
z-!duo|J!=zSmv-&FmBk28mV%_4@qv+Dt>cCKE4pSXFAAMvcZ&Zr7#;+v%?;q`ckdN
zSSQioj9KrIfg8pGuPbj~5Uphd)hC#Q?)}VoEs5$WuRG;VJWp9{_Lgzz9xVVJo*~sH
z2~rOpE5515v)*|*FbepSuwyAA#IR61!x0+27Dolct}%&+;i^q(&rz+w4Nv)7;rB(O
zjlKo<1Io4i*^oWiMy(;w&B_Js^zu90R-xCB3(N;_hv`a%3i>9Z9E$domc<I!xFOuG
zSMGtVgIe>|{*qTrEKzQTn(POVa+e9r;xYGWe~cY-xjp4;jwt9fPf!%Dtclx%LKjVk
zoi$7N-#V9;W%~4tC0p#!sb~BRfgHdViVON)b?Sk>A!&c%p;)LAMqwQeB@LB)>I3@H
zicxaVj$!bpe6@Q#3};kDS1|KqqOC#0Jk)S~`X2QB*3II2P;2!(^vWnEm%A!#m<)f$
zQG6f$E<*mO*7yXmpaK=&H#5oRYTXg+(MymZKTPy*UkgtpDQ{E}bJn<(ZATJ2qX^m*
z1FOBdqe+>>U@tYn;`jnXO|^k-+7@iml-y@xet@6&K@x7CcyXjv%sp*4$IgHSU5~r<
zQxO*l%T$Nk)G8jNWjlzfXKBG&IYEqvqY0c>+%yx}aOD9@gtc*g&Rotb%X-6zS5&eQ
z((4$V6k_l1!NV5uvvQu&VElyZ5kaz9iMOT)tk3P5Z!mW)nX^$_LIbm)%)S9us6Ghj
zVygsq_zit_c&RH7w1x<huo&hszgEhk8%5k&^~y`m2SZ*85ym+sEQ`#1%1aU$VGpH2
zvM*^?H~Jik@CZDntp@3ogB0UX_ZDSzjf|hLAXCX0oRye^_B+nM>miY#`z_=^xJ>-`
zJ@*a9tR779I4tFP&i*)L+yk4;70Qw!nGwn|FlsQGtw~A^RI^m`%`|2O<ZL=J>0)-%
zdN`@E4w-XZFb=dOpNMA4qz0g76}70(GiWq<Iky{6R`b&%KZ{`2f{^9`av*9vuDUSx
zvIS{7<ow+HRu5xHS(nC{xXD#9A+@&Jo0s@fYD+S;yQ!*b{RLIKS*ptr>bwiq_L-yd
zQ)fv6eu1YuirbG6^X=Fg=%cj74fBJXQF|%Sj_E1m_KAr(qN1F1BUO1mD9{jB7A5n0
z(R0P5>Y3Rx`EZ!2KmZ9WJ-N<dMGI*x@wfmIrDx2Qby+<7tNPFbbd6Tg;z+iq?mJ0j
zRx|a`f}+33#OmaS-FFd9PzFi`uCjU>mUIcf)hp~wgB%VIe0X2M1AOJSqnz(?!T`RW
zF1>>uT3x_P`cs`(X@NE)hcJKfT>|1f3+LKJ>u~_ZtK<cl+FrT1Up+z;v3S_1Ae)~{
zW3v%cVe=t1&k|fI{Z`Xi<0!5RMWt-cQsSm9`cc2aCRj$;0Y!A9!$%@9t(Uo?|MEv0
z-gs&-s$tBmJgfLX=%U2ExDw&i^R(8m_;(rCCmslNGSqR<Ul(Rn$3^Uuip$XYzeLzd
ze4v5RG;MhyX44Er&(FCOo4=2CqRh0cEwxqhs><@Z`0HZj>qJ?@RrFLx8?x`avC`|W
zFhaTSK*UMOe84dRb`_>{PkS=tlBwgc#=7@sQkU-J|1egdCsdu>9Uf_3H^^AzE7v&>
zkgF6;&tl7`)z0;VY@yqfvp$CjUA{Nkjsvje1tq9L2{N4nf>43fGvZ9kZOhO|J`C9>
zA>28f@?EyPhq&l&jCtsc_}UqUkC4KoN+G+oOeZf=KAWC;5@;=3vS4HNjCa+S)aA>`
zOcyVca}L?)2Lrmc%O@n`bxYEVMc1gy6_$b2y%%KpZ8JL1$d{O*YeUmqMk@~z6q*)~
zLVE5Kg)$!6z=m*v=ZM@YTU|=#{(k3dbU|Ln4aedarVwvHE{2lnu&)oYflu+-s|ZHU
z3<cq81js}RjqoTu$oAalP*t-Xf10l6{#L;NrN)w<ej%a#tinAM>nfo>Xloq4Tu!4Q
z_K_v+e4y7gfBkbo4=~H#j3isMeZVR8OlP&OVK?idx~5sBtU&6e7AF`qUsY@6NKq_}
z9(ZzP$^hBW^;g*v>0eU1KZWn7Jo?MT!kVOO6PU>SrOUIpemHaV_iUVrb@aw74@V(V
z_Wr{QPr%4p=~5ak8pkTvyn4q^N{uuDgET(ba9D@a=;OFH@&)(MlM(Y5>(f-6xuwI&
z3n@O$R+OM;)bz|HRV!oj>P+)Q{#u#<t~#|`c!vXHO5Nhp<<U~<g_1GN-mLi6JfEQ!
zTUnTsU!sqWoPF;jk<fisI{!1~1<qVu8%^-B@|#oZQKsI~!l2#P7m`j^Lht5q^hvI%
z4;=?*PQi`e9!$Tlv_r`Ib^yNZ73kD?k36(}wag!9-p$Ppq*OY%VA_~y4U885p8z1v
zW{8^+wLtp^wY^3ry_9BZ>f-xGCYoLaY;<=PhG%T5)Dhn2Vbsd!Z)&363)&(M=Z)gN
zy3S%A?CscXneg8DRh-%~9E`3B>tESzPL3hW%d08Y|8So5M6Hz+tC1hSzoX5wH;ef<
zG`oM}E;nAtE}tfKIJ-_;Ncv^=*Ac7_Z!hv=0er<0EX-uYJM_i4W4)Ho3Be}gdk4FV
z*iy$r^RV0Tya27XUVFVvhgZf8UQ2Yt4}?qpr%eCJ39$cnC&0|Y@_#4b{x<=*|H{IU
zw%=dZ&mqEv(_wJ@VI&$~0lG>1?I0vCbT7}<^ZB!fp1!}bouqUBI%gj(hheiT>GgUe
zMBvYxyOZB*z)FyFE6&B)^W&%ZD70WeZZ($`tRk&hi*WOU=O2)ULdQqxttef4=JeQC
z-tP5QmRgP-C7$B_cC0x3HhRAep!(1~&OfW&Z|@ixR?0T-cnt`$)z9&_pX;BOJz*49
zeAL4?`1bsnO{O!w{rDJuRnwk>I(h2nMgwb}bM9FKo}i$FgTXBuj)a0^*!?3a3Qrt0
zgz@MWGRMjvH@BIF-7*5iM>8{U!ffmX-*e11q+xswD++;<U&6!yG>eU)f7%bi@Lks4
zg9L_~ncO>P|2o3Bh~U<TFe}<saMdCpSs442w!<#T`6&aZ^>hC)irOqjyD#x4lYx7o
zoX6YDtIypN(KkYKLZz&0h7)~e+zEZ-lgC*o5Au#lA4(ofbBxw$*aC_af$wJ_0=opq
zgGc@Qjz(ar``1OpQH?A{$fIH+Pa1gSV~e2*ZY~CaQy9^7teSiQ(*q414CcD+qZ?E|
zv|zW#?!H*fis|oO8mUJJ;=63P&E{i@vyHg;w7%!(Qa@KzOPJYNE!=%L^6cJpt#WfV
zQKEQ?*p0ASIVbIsB**dh#J-_UDM;7W!W+!`6-mgOkVJn9{J%7VfonCNK}J@)BeeNX
zX?}uBLu}sBF9}+xjZ<+;1+qJoDGcdNAW<|Feaw#UTP}u&>O*8ss`pen%8&6n)eAeJ
z2)cLi_<cq#MC#Ov12%eUnRnrbUoF1BZb>oqquXKGxuaj-nZCSz0^WCxVPPyEmkGZ=
zPpJI`YeOiO!v&*FIX^#E_ioB}%^UnI<+6xt$tN#jg{4q<31p!_R8#Lts<BN<$e$GM
zCkG`@;?dvu6ed`O!1p05)D)x=Q-r%SHOOx4){>G+;XW-c0hopZ1AWF?cGa|-VxetA
zCTt=W(7&hz_Xa^y$$@ukRI3eOy47-kaaq<%v-hpncwuC_TS=FI0_?=UP*T}iusNqH
zY_H_{W-fz@XYE~z+QIlo$~anrESx&=44Z?CA-UgqKJPfG#gI16gB!{CRX3BU3cq8W
z*6N8yplvXCDpy03hEhV)CZRxO!?<HUW4H_wswt%BQbWDcQ#uDy*zB>mN(omgu|XQy
zx&D&mt0b-16M&I|{PHSz)nda)(C4%|Q6rDlmQ0d`E$q_pyCgenB4uDrnOi#Z`9e3F
zrdFFjGIA%xnDCaj5t$cpkC?{c{BjoS`!w7G)HU%WxzR)7tPWCtDp(ttswWKn7#&8n
zSvr?c(0PV+ioQm)Sr(fV0U(U_?V9b>EG>J(n#blo!KBjrjKAXTbX(CpW$RYif|OBO
zG@p)#fIkr^5ru!}64P{EA99h!G*_2tt~90&oNn1z?T=f{kv~@e5ZjW|hnGYA5dnsk
zXiBgM_Uo0PMnK2w-VNw(0?)EO9Noh>^3QEW#5Z~ef%MO3*`iHxWlv4nUiL?L`se<I
zQuPeaVXtk-Hi55Xa?6mH2j)+*;2+TS^A<*JnJL_PKK716c}3q`>vQ(&Dp^rz?i98h
zuK0hTAUS7`;fmr9u-2@rWix(s+nlXh1O@~P?n&ux*XGt}kg+!*uH6Yn{t*@k+E)9t
zVS4SxCgZAk_|86g^R$H)#dLXw@|AIpIv6kA?eT6N%fctU=l?;LwA4;gKO#ml-U;%<
z&g{92mVP<1oNye5Ac_9Vio>rm=`uhPM5dp{0_mPk#(75eFA~5?mpxAvdp7A8BMt~-
zMcp+*JF?)?m9BtMzqwo%dq*!UC%>!*#5#nsmq$XWeUoj3CGHfFn_-AgV~L_l^5|yG
zN_qury(nzJlKIc#y%lldl*SqD_=#+`whUx|*C8F}q&$Gw8>{-iuDiIVZRmJt94tRy
zdOhP#!dQ<Tk021$(_%Z$Jm44{nMLSvo!1Uay5?0SAaWim5P@Qt!b!2Cc1bVzt2wH3
zo=|`|3{kSSjQv4UOLx1<(c^li47GCa_Fk`(E3k>&o~m_>i15t)>eebxjYP0M9--f{
zLepQhMXOtz;zOsu0nzpxeG6-^C-ka2pihMwna00?WbQhjlo-pFYa>G2=k)=wG;LP9
zq#Fehh=J(wFmXs|^(oWOqA!-JN_!5U6a&FW^R0Y9aI;H@*a%z?8RM6gOV$oDB<z8`
zg~3_yUSAq&?H9X71CCDI>hZ2BFO9=p`13&dxuK8Fno4oK;4F|)57%H5!1MsMle3;x
zFMAqe4k5lW{s)(}*JKt{_f5M5!6+*qUgd5+TWuZ6EHjGF2cS#`9x}ex-z4!DSVU0a
z)~Lbwzc(uWd4+;le|{_xuQy@EYy_J@a4T9#rGU_`*aU~lO->yLQ-#XJ5%LcmDIO=~
zVHcj8o#MDD<OuNex24+#|I5@mLr<fiVml3lpx@F0w%|HF5Cf2NW6<f;lzpe=ZSow(
zWxOixweeeMZ)1(Dn)gmZPbfGzI(IE7tS7Ap47OI;5SO!-4|tn1>MkN}Gn-x_MSm9G
z2qN}H7hIv!ZOw$%57u%O>K_DUq(KLRAcu)<>zz$!IP0xBRIv<r+rn@tS3;8}C=bIF
z4r*=?D$Z%>tQ>#)TE*TpMa^chTrKh?d16QUOZWf|WhBEkEgkfdFSi~XO7BLy)1xYi
zn<PsKI8=MIU0j~Pc)WQ^b&?$hc#M+kB&LUrdM%f=4SAV!nof}ICEiv%=?w~O9NlU|
zH)2{}^0R&9TG#j2bTb9*(d!7a(?7#_SZnnLNd+e7-_=Jc2<26JX{J&`O!m43P~h#^
zJVxZuuXhX)h)4zB+y(wPTY7_vppruUjHmy3Pb6+Y<5wv&w&nSqBQj)Xw3Ac<Bc!=x
zKsRBvs*rF5r^HV;g<a&OnKj`vq2J9Cg+3u5RI~zPv8^HFy5-hi%R|O3oKhbz^9T{=
zA(<i^lirG>as5zQ>DCkUFg$?AuBYf0PA3s~+d(A}`0XSG5;1peM5}l0ox^WXq+Ggq
z&VX^&QQd?%#;OOCJ+2r}J07xsaEvX<Shfwx8Bp*iYd0nEi?9K@@Xvn~TmMUI?Ee!y
z=J=<;{>Ssb5nER}lW``U@Y|~zTZB}xR7wm5hC}WeYzcaZ-k7GytP8nf9toLr_aB91
z=T2^f?&~u|LS;bdcKVS;l2Ijbrd)p6hmUt>k8fw?9I?r$JAo>TC-tavpi)0VDDzq4
z|HTS;|1^1Dj82(YxDaJXQMU2sdcQo|#>->#v<4Oh=H%rGS^FNF-3GpdZ5LSN8M0`M
zk5<jr?s(;LVuvg@!0|Y$D0RW&0^$o_+?&n+c_s70DV|y>Ea+7;-w$G?5dUQuVSJjc
z^mw0$(urdzZ6$+J$Z&Nxq!3>Uct`SCOy;nUZ?(gxcb0RuKhvZ}+Bs9w_v)FWZBsX4
zSTvV#XQGfvJU+_E4_EUShXLih2R(=OHbAKc+ZW!I#29dtOp|+mRLY~tHPf5WA50YX
zi(=DV@p{2)>Q!M-@ie!W#?MnaldBeulfrD<sF=t10X6MJ{W1O7*pJO0!jL$Qe{)F@
z?p4E%r*pf@SILc^vV$c|B0G26YcEkaQ){)nDMiqPRiZp|Ic3exD{E}zjZ_`(Vy*He
z^-eoY`6{uVXo}M<%XB%Kz>uYn`ZXogFM|y(9^L9R%PVc8ahX%vzrl?*Gwf`8($4E0
zLp4&Ks<BEkE5Pa?*>+pDhU=|$yg<G)*6eLId0*2sqqSLg(sgd2B{ST|{6eZCtP|>K
z%YHh-U+V)>E2U^>cVk^Hx^uZ~U-(hOr#g1WV^cH*c+9Kk>ND?`t)*RNP<I(E&3Cgu
zbGdj!<skH8kheM#g@Rp%Q!%@H9!O)?_k1L!gv(vGlL5E#X2S9xlac_#eUSJ4;PmzR
z`h1)mNE=DqQdcnW=N*Qu5VF)hRELpp{(4_MyI3qwq@`r1CUYDAiOKQ*Vuw%3=L(0E
zLQP%4VVT?`vU)~aPsLow<em@3Og8vKSKI`jhNq;=W+)zaAx@P3mzSP#Lg%~GYKlt0
zgKG;J{E`#oulGp#lnH)J?NR0^o457PLI;bqfc-o=CorpjCB8KRIks?(y}7flNrB7?
zQ$G~c_wN;K&Tq&XmY4+VCUpk7n!{Eg%n^ML&2djO>2T}PKUy+QIa^ID8zkMTz_@V2
z!F49N);siEjtthoUS$v7hXda|Ty5+FG91q|TH@eO%2knH^ReibMoi&u^*bpe$s*yQ
ztV|45f7xjW|E1%HzE~{Tj4LaQ#gaF(O*B$f^XQ!RthBRvldhZ49_9gnF6+@55#Ma3
z+Sg|ri>n$%Ypbp52@Cv`hLrhVSY8@HDf$UWu%I^#5%iHrzqf|@2j(&n-b+rDAJ6(&
z;<RlwFeVjwgfhD!T9@I4=-DAg$3w+efB=UJcIwgDwA&E(Bqf&uFmeFVTZ-dv1Bxh4
z!g4fsDwH$H@70ts5Y{gxU1tpH%y|(LKy2deS$Tl($9p(K0I<;ZH+Hr%FmXzx(j#QR
zvZ;jG)?;p8vV=6YBm}ahNi10~P*z<vKJDJ*V~DD%Qo|kolTQi6Z-Vh=ofFlzj<p83
zxx^^AM3drs&s`tm)GCC;kUHZbheX(yScA%dGxWX@jUNDOseV9qUa!1DzGn5vzm9{*
zm#8F?Qt=by)fQ=aM0t_=bzx<7$<P)a^I9+of<~dMlg%s`W^(H+;v?WFIPa?BWiCu<
zjvHYT{ZJ8NwRwp5FoLj7lu~j?qp=vAP*OlbG>8p{O^4t2eej20u|8q<%hU;W3Wnw7
zqisZ~3h}+P88!Xta!N*a`btlcTO>l(Y5+lL$(g?!Je*~Rf(oU4-l|LXT6uLUFi}jW
zIJQabVoD&|O02!~+F0dQPs=s4%Y+U$98{q28}oH91Bt9*A`K^BoHMWyh>g`<MWcpj
zMw2cUL`zAvn#!|$)2}8DB?hvPEr(2`iDY~<DspNQn2jJShidp@v>JuUFP_h!d`9Vk
zgGV7k7U#7MHy<@Ao45g)nq-rL`BHy9#crIm*T}$zhEM5<yfcdUfM9!dn|sc#7uZ?J
z%4rANz?1HTRPgoA?~NeaXj+D`K7|MK9M=5q8ZU6+SoLUj;D6yxgM!MGr$tqW;Nv8h
zl^~%Qm?$p@Mw_RG;=gtgCYYY3vN#&+|B!ha*MoJcOETdiqdI%g=yn)WPzr&CpL7AZ
zz9eb6+5*myypF6Km#&py5V~V!hYiKPkP-(WrHYJ>U@kloQHw}?mNQDdJ2Q!QDN5qH
zB7AT<U6}wyb36c)wNc{CH16ul0Z>JyrhDPyQFAKIMWDd&^xC=Q3n_=w1G>j5;^QWl
zgh?0PBL;GI%+~JjoYTLS?$(PR*Q-(PXxDuFinOGoF6jsH@$)dtQtFw00euRz6&2qB
zrZKpj_Qk$0HsQ`egLC=742$IX)PTA|8Rd=9dVYG)W%YPnKhRO0rjb8V{+gp=T%<9z
z4p^%r_>H(+^q;R^k~d>_ai61^hKH8+v-B~v@p%m~&#e+b8*Cq;BMt_*G@g5U(dBi=
z_*w$%v2ZQw%C&)#g6KE{W=K-l+Qo<CKr`ZP(5hVsq9lt{SGDv+DCmDjx?!3<QHbW6
zDz|BjAc)n@;e@SoX`zNLaSP*q3?PW8)7>4JSrF2`cc6)eLn~8p6bs$Xf2aaYsgKXX
z|4KwX8qeuQ5~i*U7K(F^s;wGJZYS92tZ=t`2umLSf|8PwTj@g)Qsy3EW(l5u#Up&M
zOONty>Bi5vX~Qa8$wmTaviif>ro)5HHJ)v>p;$ev7rF7iwk@8i;)lQ4)`hVPvrJc_
zfDXaVWx)a8MbG(S0hB-bXTt1AJY>>ib_R0Jjs~zW;e_6-86@0I$&#`)-#O3gfu?|o
z|9k)U%NIS}h<}>`w$)F?@|;o@Y_4#S+UPt2a@#tv3SBVu=>S3oiV|<J>Wv7fe!4{F
zTcpA8<xZ98+FN;eDn8gV#~-+QBnHTvYFd(qKWd%cw(9L}LA^#c%`6HUrBesj-Dql`
zD~MGACq}=xWkT021p;AnE8KnZ^9$OAmAsnJ*lQISJ|Ys>xZ|Nd%Fi$F(OiO=S1*5e
ziD*?#8pa`K+}FH?@!kYG8k&f4LM7~7V=*;={?UnGIr_6<(2*1=Jj%v?;?Lx6oA<6C
zL(IL5>yO9uOJeHJ!^Akl%4CXO(R#m<7A-DiyI#^M0=)L?jy$YL)rMA3kDpeD(o8!j
z1b>|hL^MPFqUZU(9oE{8Ck0}Eq@lQZdgH6Qh^_G5F*Dlk6<=N8KB@28b$y$(X$KRG
zJ-Pq?#n?NAX%el=x@Ft8tIM`++qTUv+qSJPyUVt1|6R6st-a6PoX>OCU0!GA7$Y(=
z;{DjIm6MKjs%hKcFdKR%CT-eBJ+H3*sYmn{fvTSGDHBxe&UCE&VbD8f$_Y>TRcs^X
z>ULz{Sgrd83$>A9e*S#$b_<g@A8ixPgadh>-S^gae&HQ>k~RHJj%2}Kr60I|y?X!(
zVh1cTnz$I2Wp}FrBGhYyvTM@K<>Czev3>U*nUCMGd*^v+?HSlgCI7x#pYh>8eDu8^
z1>FIIt)>T!oU+1Uz;qtwbB3P$egljs4DLtH-xpgALseNCB)^=~<ZM8~H-9Od%RBw7
zlw(7F%An&Ppz55FdFpRpA2-N9oo+zXo?e&Zc6i0Fet4YDPWxAOt*1~*XiNEqp!sGb
zcNVkCSi&bmDb$5;DxhD{`E}#5zl6Mtq4wFdDYUu%L!~`~$%LuVV3L%In0j<JQ@0K)
zTSL!)X~lN<`WpgAR@Lyo$zqQGL?AJ9F#T`S2s-t+9q|6o^%(+=Xo;PmVSk=Fz|M4{
zy3zeK!Q|c&esyk!e|;*WxM9bRnh@e<v-joPT9QSTNJjlM>=)al<Ow~>XS)y`q!r{n
zK@rMl4JZpRlfHsV#BC@biE4t*@8i!Me5mIa+=^smD$(@f{CzsJ2a`b*_yV<o5ESG^
z{skO8Ujez#e&(Me-0I;S8duBl>3-x=We2Ogi{s-zy2?wy2wv~M-aO9+tV>~tL>V5=
zdgA8m3tao}#tLf%gE+q3nP6_VfE4n|OELs!e!Yx7xIJ5uX_qIll6^Bdza4I}Hh-kH
z;E-9VY1>T6Ruf4rdpEK}p&OXhUSLwgk5}%Vkz86PAsbN~hI1fVP+Lm@OvIzYSS=!w
z3fUG!*NL`+Q|?kK)%_RiDa5UanJ(buVc^{}gvM8~18UnaR{F={kI+%Y<t`PnJXRKF
zbzIH56`P@kL5fu=43rd<7Al&k3V&~5JX-3H%j=WmD{Fg=>sN&__ymceRW#f=9dO0h
z?Du&b$n>I5uFQs)pFwh`q;386jE)0@98)9IGUr+Y63KH~$}n8USQ##gX5iK*7JijE
z+1T!{T~Ge{c<Pu@FL7-dZuRnBUwpBgtSgsf9zi#x)j!kpV!%5KjRUPGXMZIc!Wvwf
zsm~Bc(x<QJhuz$ZogeTlc>xk&(#=2$0e|mvWK!IM!{olx`66R7r?VIsW(3XX2X}xL
z<SLn>jsrfUAm~Ba8wir{6&M!WEO~~Fdh$(x69=mu+3^6A5<w9CeJ4G-4ih4JCQL-e
zfXQ7U<5Eh52CBGq;06N{<#UsbCV7{=+>qi)o{ooIFb<A1d6e*1Vd!a;`d0p+VsH8(
zpPGG<=)$(PxN8B3n8x{e-(a`*>+V<siFI$+3GB5=5VT)G;6*7Ch^Fr$H@{X8hRhrU
znM!Fs`3a<^5W(yxSHXPADh96Dj8psQV+$|r6T1d4uQW6qXBJdfd4ojVs-Ac4v#Z%r
zr9xIli+fA0y#MDmaU@;o$0`1QxGAp7UimHswS1v@fleLDOD;DMuU?h8>SoqL^pp*y
z2n#nbiRkWy4cLv&qyViu?AT^CMU09vZXNX2W>=7OiXt_tl(LDRNJUYg&`M2`ZMygn
zxmuPxK58wBX02As<#ReA?-g$B&k4kmODRa+{AV$ln0Q==w6>YAhLg;Lp=X_V`L<vx
zFL%Cyls!B~Ud^1IewSxmFh%o))$NzLR(@7Pew{i}D$0m%aRHOQdsLz}=OS8DQX@Kc
zyGJ^=d!(jRiZ*i^ntNYL!!Q*X<|3T4snL9V)7aq%6=jacVMUH+X$lh~Ws=9UeLBa9
zvNF^~3H^|fX6)czy#gYa*mAwCY4642-Ah&fn|a=mJC1yc?k+=V-YoT7`d8Zf22=l~
z%DmHM5r$3NLz=Qy6!yDn_LI&Y*Q5ix`^;VTPi^emF3aOk?xJ$Dw63D5g{2w3Sx(Jh
zD*bnc6I#@$()DC||C~LDf5}Q3ab_v~cw{<_J=uifT1>W3!kE-V2Zvvya9Mp&-fWXu
z^HukJo=kPZonM3xfGFG?h9(>8Dw&H0`+9imr~{^ICZ62!h*4=%LS?6-rAo?&9T4hs
zvRqUhV}g0TWi%sGh?H@vj689VPOS4i@CH{a8-AB@;v6Zx&(YZ&%Ix5(8{JcS<dNmJ
zn#X+!YD380GHW!tkA*|DT5LC>tzu!Xuk~HYUGMAIlLr!1O{%Alxu=FrR--&c>!LsV
z8-P||dpw{<+o@R{6i@xx0EFukA?{31IHEM7M{3JJMGq!7lDjGMo5AbWm-b4^4A_~`
zk6FBQ&PCDNAoivv(TX(1My<v|g|Fq4emUR4EABZBpsidWevdloHZc&|RjU2BY+<YF
z>@+zjDjIcu;W@q8RF^&LB&NaFOF+n`@})mf<=sCB4sHrM33$N`l6@~k-CHi7w5=`u
zLN!|XLN~Xy#K&rPhUv+_MtL5=G8NDm*~l=S6V=~)8A_*QLqE%>YSG#n?cS@fSwy<2
zCSfmWnReY{UyD}CyYnU5rY^435``$b4-w@jeM4{*i|=0Gjd*8H%D4v+<w5KagAm8N
z^Shgi0fgSQBdnY0j^m+jZ1XZ=SMYr1ELhdYvWo9SwP(RYF~&1;du~<^tsfffzK>0m
zY*)O;s-q23r!;X(VH@3cmMFoprw=Z!hBQGL=+JD&8nT746D6EXq@cit2X6s;xj11(
z)BA^<+hHV1>;i}ozS|=RjXCNHWB&d|f{?17il}>*yk#Lt)CX)9N*?~I0!Q;Id;CU^
zDgogXLG%zwlwOG6s|z8KxdstS3zFA^<#Auk4aKiEZ0nD^qF+MRLz>Zx=1$u&ldzO1
z$0WuVJ1L!J>=gZ4=3>Lnc|6QP<{k@gv*+E;2JNrj@}#X9Nvu0)w}NT*#nu!qz}u)y
zoT!Ni!ya2Ku5Zl)u62qW8G&lpyGs?opNBIWo7DA=nktZ>*%UOgz_1Z4+#G@K>N$eR
zNjO5b6k)4t13lf#js})&MPO83`+394PDX^_F>20g@`)Eq(HtjdVcrx7A6P0-Rb>lB
zb~^TMT%eqH06nE(oRl&8cxO5r6;dw(#VLX@m7QP7#6d<Ykj7x3$B=<wr1j6mV5DMR
z&!?a^Z{38Xf(3mT7pp3JnT@A19vWWkPA69$wv-wnD)tHpq>I*+M^~P8G~Je~I^ys`
z$)FvOE2E&pkjtZ>%aBW>E>ZCn@e%~r_3;u>>43w+XZ*)o&TGCFCN))~r6<1@r@B@K
zz)9-{)?wIf`>{xgc3+rWaSz|fQ0bGR0JgLfXlHv7XIx6Va^7TS(YbN_DZ9aeY#h_A
znkP^WcGK2bho9P?4^gG~1amIFI*S!wj4i)BtBH=0e`t6XbPV?&Lx%Dl>TwH!6QbdC
zm~{EZC=zFC<MFO1cs*Bf4|gg7Kd?>JZL_QGLShxXnd!_KwO6&m`H`xtp*-rxd~avC
zl}oaU2ksV9jx+%0MDE`<%TwjrABDk#o!as8r{d{nxQ~fXBT7?Os)76AO{_9#U!pR9
zpDdpK0qIFBhGBnrM=s(<X0y)POoxf_irJs4ww<^O;u~MRiQ6M5_Z;69_kL&#yL!mR
z$tOE;t|4(g^9RePX^Q4tLO_|x?UIRCLjEjL3;XLJ`C;^Pk^Ud-GS}>BqS93TE~Nom
z7RO0#BSqnbYL?E83U#$DyMs8+8^y(uug<YxWyx65LbR1ReTTjYEM%kR#KKK@O|Oko
zq}`t<Uc}i|l`^L{I$htZ6bfEyu8%ju&DViot;k|RW8TVo2STi~Lw54{Krv@BHx%Qc
z2hioC%TdAEs0SG5uaM|Zth_`jkVeJs_BJp{Y07@K0^@rXlKx}IWks!@^4qz~+#KPl
z!xpVzeKg-a43m&5Jfb?Ez_>m?`M^h-%IR~1n(~!}>`RH~&cjrkgbS2n_xYz1IbENa
zXP|mMc)VimN#37>@G}9#?&Bl8__eZzNlm)jyv^ACPbZ`M($W3Px4ra)vO?`CA>PRl
zz}Qzg#kPQOwLAB&R$KLBY4eZ~3Bt7Gcl0+nD+CSBf77`?h>8CI;d8S7f1r*3ANzM%
zXF3kI1GyJK{RSRpew*KC*nkr|`v1rinpl5a8Zg(ViR<%8w1z3xYgZ&|TatlUlh#ZV
ze_{NsdXpSK2auH`hMy{gd(>~p>Vv`R%jNgV1|pJuK}Lqo$P);1!Y`IzFYXeQsv;Q7
zzA!6W$1=p{gZ6HgL&5{w0Hp!(@_B+*ekJC;Q#!`0eurEq<dw4Xr*`TNmd|Q6`rW-b
zrPcASJnWD`8{U@Ye|?=ZKuRqpkaRvU^~CP(1aN=&V^JOx!I(WfyuK6jG2(<ESUW3w
zE)5WVXY}In$0^$wpk${6sLKcV=7uHxP0}tFZIM2$xm<!lakHa1Et)%5QmXb!TsRP&
zrtMYagGenf%qWrKyZO82q!EPw*9X2~qdKuY!u=n*t2T{qX0=aMQZ44pg;Jw$hhC8j
zNRD>QvKxwxokPey<B~Xh`o`)5ZZi<!?8O*7%YfAEcjHlAN$Ql1eej&6EZKUIla364
z0OCA7_Fqw`6=e*fhUS_W;GA4rgG_5F&bL>8*OseCwZSH<f%R1jRd+!PA%g0YNNMu9
zZL??;#W9+x3Lhmn!n$zMWo4RueKoDT9FlU`&b_V#4M~Cy+>Zx>$|~fKqmjIZ6j-X_
zv%j?lOo-EM(Vy*x%2afWgKYN4EP+BfkCpF@1_)b=9(GM)5iIRgyNfju+(G|UP7pA>
z+Smtuz79T&jD~XUpb-<})WeDUU;Azlpe{9!O1LK-)(2a<sZj&iP5jn4!DZ=m#SOFB
z-ilBM^KYR;n(7qs3-d4XWK6$#J8EqXp>0S&lCTe<LlufYiEl&1>d&X~`~8vHJJ6uN
zpWlNm#fCC!u06wf3=05YSGN$-5c{RL678U=eX!ugb+a!ZVGX^FqM?(kfZRZKVXI}>
z&0|nkB#M{IN?r0fPUJ(rR6Dz!;_N3jjktkLrS+S<5Bqh<JfK^MN77$(?6pa4vvd%@
z_t+6u(L@#Ak@GZ8E0g~^+8e2|76Y~m#pvFFIJ!haR*`k#!=~AV0N1GM?~KQtYr3O`
z<=?WESlNbZr@Xy3pl5MT4dEW&Z*Rxxvwws7Aim}4kTOGsE%3{(kE;g>QiOm$)O$_Z
z!{f;0spFOq_@Yr3C<Bk&TlVf<8KJHV>@21KmM%o#_ERmh&Rq4CN09~4SCUGC*id^s
z*U}vcP4gXLFDsQeNWrwo8T{dElQvka{+v<!vn)_Pn0|iq$C}0ZW6&cV$U!VKLTrz%
z8Kze3%-ap7Q4)84vLxhj!MvzN5gHTE8>ARz2gJXs2w>V#lF5i!_$#eX?1MioU#X9A
zwFpI8J0v#HDD)=54n?w7tmXc6cDHg{!}hVzi@}*d&?+W#`@wJ=vY_VQiJfLFpn`j~
z?6NdoVS^b@E6V1C>X#kIm6u&Gys#4#VYzvPSsmiNsX(um5YT}3Jp_1cAS><2o_V6!
z2!@>4vfL6B8B5v-%7>{hyIjHOex=NjYMk-X#}Gx<cqO|b{o@H^TAoQl%|bF*y;kob
z5%Y3Fyo(fp7<{dkbs932IgFjaRv5tI9Ifu8vEwsQV)*mPq_nSRNiU{sMWY0AC86Xw
zUEM_ivE^zjnJhE!;YRyS>WHxOx9qW#-r!rJ3lFSm!Jst6kQ^7Ks&gFwy=Ib%kAiu8
zlOhv2b6W6ALoR`r<O|aiaXhBBRbi2SOw9&Rd{Nk@n{P>~N)_@2Ou0kiX)8Bj>tAJt
zdcIn#uaDj~^Q7$!5gYK<2`1guHF1#O(aJd5nbI8oBaC|9kp?W%Gw!&#Q;?JP%n$OE
z_=;tvE+=_<nFm3w5!ideO#L)2`=W{%%l#jnww^EOg6AXsWF1&}x@92edk?2kn=)r*
z1lwHoVZT|j!GR$N?Hmn!?A4^wnW{^+THBI2Iw{q<Od`n_=c|x}9}A>;VLVhS&87-A
z*W=f65As>V;n-zq2j4b{N|fu~S>L`il204g;KXGU=jd)djQqv^xJ;$*R{;#bjni;U
z8Zh3PK){!j$+I1J$5?C<XV<`%Mbhr4@))H7J)wd1@`_cE{k-9ve*4<jYW5jg<X)%L
zZnN8zo2pxg)@s4p&%$>c(Fd)I)>e@_X`Ud8gtktp?>x1m+MCVRRfA4-75_Ku5J6Eh
zW}$Bnu_1@lLQSkgb%sJzR7lgZ1Qm|AS!Uji&UBfTy5eW;OOT-Qbhr}RV*OEip_)n*
z@<@V>@rhFA<x}Y5R8VD#ptjckC-@27e{~BXQ~-`>wJ{jya;A9q)$3nDV!P`6_uk@2
z;c8r*qibW?Y|cfhOS=AmT)DUF4s%?n!sEjgC=h;a0=I5oiA*@qrap*@K61Cc!j3c?
zQ!%~R8R^|hMp1!%0GQ%_w4j4qb?!RGU}7JIx6P2=)Oxs<E-+eTT(lA$8KdhUNT)^|
zE%zA(9U5Xvi>Iw2ucPj~bz(e^w5&?;Sw~yxlXF+L5ol3@^tb5R+y=DSwO0tQbhl@X
z<9!&R_eI}nGXl3O5L0t^s9rhfyb~OtfF%ziruIsl8?~5!Y*&N_wY0Ys;I^aX6<U~~
zefH*lphX)k<=strTO$ke^2z@A3mi%9BKJ0udwB)?SsY~J;Mp+P2xNnMivq$Ng1Zz2
z%OY!|0mBu}mF^Pjtiy>s(FRfuIetGNB1rz$u<%|)F#<O-0%1gVT(o@99{~UGaOZvg
zO^s?-19W{gAX>c_vicS)%es0pl#1&1slL*2Zcn$KN9hg@h`C?3#stJv2mX=vB?Z*v
zV605l-2X9eiF#EIz8y9xRBV|r0;9a6Zcs#vLB0q@lVlDmn3kx7`drC#)n|C6=Ru%?
zWQ(%WCJ=&$%-n&!w6vZp@%3pG1Z-|&ak(g7*36j~=13D%ixJe)<c@XR*eqsi?JQCi
zUpBh9C34d|u2vK?t59?lrut>#I9On86sW=vu>zhsE!JGFMPcidVr`RXRZS4aT@<fr
zh!vXpw4<!`D0*?K4aj>hdN+Qq!e1dSWQ<Pbr?bluPjq@(`rl-U``PWx%;YD-^y!s8
z)uH>^4K3O3L6=E3)goMUe;LS3`>~o!#4X65L+y8SjQD7}7%OkG3Yr4!0B}m3m(>fN
zr7gC<&@jN<10ZAK&9|n~Z-^6=C<upB2V;4|XuzNS{%DC0x}Sz`IvAJYChvStl}Jfh
zDq#8}tujA!o{G2q5ZjFjeC+2J<aUR#vRQ!zM7nz*J1Kb{1j|J@WJ*JBlI-gh)QQT1
z!|*-C+Y-?iloT51e%v^WF_EF$h%)lyZIeSQ3-40W#4wXcp1qyd!))B1z%(dG2=Q3m
z8xU>nkZ8_UYgbF0C6$oAa-$h1%B0^yKUs5<Bw+W-5B&9f{XM#)I3~__p_w{^`m>bD
zc+gE?w+GY^r;uz(j-oZjtVMVYE8xXRC+&vdv%>0T4LLyP`>VPqTh^BR#|clR2Meb8
z9Q(8x*{C}%->%r_eF~wiap09DJ+cx<)}}kKffQd8ti59^%)NP_z+)f<R!ru{o?}};
zMPaD3NSDkwwDpJ}7LM7j<xzrAx}{Gqv)ZR^ofbo3io;<kD5%X&PixJp(pz5FufWnJ
zO2_?Y9EfG3y2Ws3c`^j&ISKq~?OV<9mpe~am)wgGC0MW9oHkL*;xgvJ)31~w0^~>T
zK<;Zs-FF@}_>12}-0Ya2ri09mQzfJZPx~%jMVXE0ggSPou*&}b=3QB=4%%B#UGh;3
zm+|M8nwdZMR$Hm5=b6qWpg_mx65zkVeOn14|Etbs|4*IG`oHy)wPhUe{zqqPxY}zG
z6>0gRB0<c@kb`!373w62{NH+8WA+bHziRBDsmr1S+XxeR?!t;)%fVlbOox6Kqr<n8
zr+3TGbH)6YMDxbl>$MZ8H{peljGep76KJ3H7lT0BUw+9-Jbj7om=x{9<6DouF1{@r
zs2>~~_@3i6yyCZn|25(sLHjrKwcxd7Ou2fxx93|f)n@Q2Vr=K#+M<sLTHk>gLV(Zf
zWUeV2KksG(m{GBZw+!TItf00SBtGKG-Zqh0a1NdG9~g!b+gmMuQVx+u!(i4hpS@tN
z<DaAm_k|cvIYCAItw{7S(@6pOJa;J**)w@Ik4-4c@`)+_K~4!VbRI0^QMpud4KN>h
zwN&FlzlKjSzw{rgR5NfDjYnii2nsLh0cI%|Q;Is~-{vu7@?;`QZS*RQS@p(>2Gk=9
z+U*uF7C8ljHaQ8E&+--GQhtG@fQmuX4(Z$W62BJcN0vXzUZy%|b)&zF`KUap+$D=`
zU^jvD)x6wuMStm~rHkVvU8aDTq9cjJV(LOeo#d3sBmBY9f7Tw?4}UOC8Wkk`{rb!d
z@?HTLXTP>h-%#}_gSKy;NT!zCv=dP8Y^rsx#;wFa+>BFiQ6fmR3sIPhPp6X>ba!8|
zqgffNeTV{2dLArkBCOFe-ZV#|<(KOgiR_AFA4;u>3^Upmx3Trs4ZeYQkG0b-q4MRn
z)Zsl_rzB*diaLYBJb$wC_V(_0Hyef-g+-mZJ08XFVI0$fY{{SkIXhkr_+@a-KdXxE
zS(?B}_TA+C`nbv5@-f9lAw9%%j<g2R4H`*gi9tzvk`Xm)4lWI)%1G=VC>JG8Lf%zm
zxy~APMDv~0E&?01`O_aXC(lj_#%pPN9MIx>UQuv31qh0sTSz0dzH(FcRd#R$oo2rp
z$GH-y`yR3z5Q?hW@HDxBWfgtO_=czDF^mSr0>6e&te#YM3kV@T%L{`P;NN&nq>5+G
zAi3BHca+Me;*2;XuEHcZ{pp8AbQ%@VP<oK}Y%_VeeD$bIXn;@}I>U)eXSSl*Q<j-1
zE9E79kgCwz{fpDG7?C2P@NuH4qbG9Kf@Z;#vu#aa<1nk`O->FTf4oPHQoaGAUaB)w
zUI*K{C|Yj)sjr;&D}hR2C+v_aobl!DdC&<q0)nEKG)@66wF76=Li*In_reb}P9Y^w
z8VwF+Zt=k%R@vfTF7>9|gq!M+n=VwflEW&Th?=0T&58Yr>~Bn%m8{pMENXPI2ij@-
zaWNE^dBZiy@xhgprr*U+W7L%FHB^2(T#p%N?;8!Ul|G8f*P{b!Qy}K#^&H*0nXmZf
ztERt9m`so{Ziak#*A!RaxG>?@sc*8#>HO~(UBbHyb*MDD75E5V?KCtDqPsK$iA2M8
zkv-sNfKK=Doji&x<24SiM9)F4W_s%TeCi@7rQf{D<EGWFZH0W;+_9hrWFBYS{y6~A
zU`^{)N;0o@_;=n3ONNZ%l0^&i#NNXqQ?047->gbM!DI|X_GX<-lm?DieN6`83(!?^
zpv)nczF#M0fm20uL0KRJVcbiq9}-e~Kkx7U{8o6K+!fku_*T>yf+=LQ2|b1spebbI
z_IjO}Mav;>50!?#WNgoA5Hbp9ZeaJ-x{A32X@=JWcv#0acWPAO<{(X|j1Cmo&@dwn
zuS|UOKF6w6O_0zk4HZxp4VauGkESa<NNRlRb{<p&(0dr9cg@<W0aD88h0F)$_jl9O
zgs=*>*0fq#3%`*R`_QUI<L_<CTUOiLO8GcNh{ROzZZwY7-RS9ZcS@nH&|yj|qu5)n
z*Nhg>b#VM)aE<1*sa!$Bs2bdwLN~7&1U$Dm@E$9>EVwfkTNqd9o;Mc$$y7(C{I*?_
z-Q4PVpl{bhaoo5&hOe8lvI&zDO*lR+MWqI~@20$eSO)_^Mdx+dCU7yihRtPU8P|lZ
zRT@MgY;|EBthwth4>s`Ay|-vC2l8JaZ*i=|m5x?!Hrd&%8P)13hkM~%<Vp0hfYX0=
zplMo_FeZw(EJVos=HuC86-$Y^uh7>?(`@^^?FKn9Nit~z-M(JuU){dti);Z#evmf)
z96FymO%0H5V~Y*xR9#z?f((w}NDlo|v=McR!KHJ%Wy~^q|99cr<=CQ{@7S4ee$KkZ
z1{bYX*Pfm`yCLWEs_Wr7DoRy`w8U>T$d$}<kiSn-(w|o+E-u#w{o|=hQH;(s<h!d%
z(TMlA{j#m+&&|CfYIw8X%;ia87`-UfR6X-D+vb5d6uk|z%%ReQ8m;6QAr*%%Ltjb?
z*lu)I;)X&8=zlfT;_et()ss;;?iylqX8FlSzf|p#9G8sL9wV_Ig$lGyoQ!9&x=Ep%
zpd2Inn(84UmJd#BRqvwN=R9%DxM(Z&|9t~;ORlRg6%K*qHwKrEq#?LdVkOL@oz)@n
z8tl}RVZV%&Pa98VKwMBFZy9q!J=Q&7nTawq9~~cs4epN6?2<383CP{dLKQEcM%d(K
zE0vS4opQAHwpNp<O>CtuYudK)_?;t_Q|>BjX_5Oa=l;8onEUKGkSi=ZvY^MC!JMoa
z(zzSGY4GvV69VIQs(Sy1mNi65EBP-2&F%%-uRfx^kL;6bnloz-D>w(4@rw-$Qt)7A
zvB<c@W|li9gVh{^6J~Y3Ey-h75ey4l8_H(3&sC$xf||im^;y;<E6vO+y-J0bvGp3f
z%23?u%<g2c2^jpY1BZtY;|?f>uF!!iog~GqTh4+vKGU9=#a?sQs-&0B;ox>jC!{%c
zajb1?Y&}zzW2(jDoN=Z)teQd3H7;LWaBP1nlq86w9*GS^ywaZG<P6=I=O#)68->d<
z-76hJI>eiXfZpP8>NUtQts;vu%GkqmO~G;okAH^#X+UeuoxXlc>6kt;py`+da(6DU
z5f}$m-xF1|jAJCCg}6*drlFC@`y4bz#1BwTX;`paSsAjbYOR-s;o>}8<ZPhP=<`0f
zBfp?C&I>S4;{)EPF_~R?yV->I4`vZbzit*=kh`;Oy$B#S;3IGs9q)N*MFyUvR?0AW
z>vS))Wm>#)bnCR)n+75A^cwCck--1Hs*K&i>yi9Xa#Hl+%m1}^tQlewjZX!AxJHrR
zBiktr=<-P09UkCLl-*ipd-|ujUc*@->RhqYRpcHT;O}<fNG)C6rH@@)`;At(34ZE2
z8uS#fCe3^H-?SUoe|n@$9IXGlaY(~jTn?rja95wS`klX!Y{VkD2^g{Z2z)?FWcKJK
z(M{VnNZ!Ar6;Uc}uKpy_<?Z6)qdE8&J>I#^Pu3&OIw0mhW26@p1skTdF~H7+l>IHR
zn(HD5qH-|udT;&cnBVa9&N~;Xtu3XT^C)TC<kya92C)P)e8)msc%#fi)Vaac|69ya
z($P#=qLi-@@{mj2W!I1`z73|04MG94x6t$nxNL)*Hnwbed&tzkuU~Bo4<Ixz>|UO*
zZ2I`TB$g`xH!Ie1R73uE_nCh@Ie5*o?48ym^M3ycV2=Ihw~iIGUf$v1s6baY>(T2j
z_sF8LUS@bm(B3cQl_HJ{@gVz^)S1xBl-8#A%0oS!!e$B`tNC>H`7{qG@r^x{-S|Bl
zo0Mbzbc_bc8y;%S3|`hjN#L<5%FYrxKG98<NWnbA%AEPCq1|*!nlS3Y7wq<A4N<me
zHPc90_UG4LHB?5X+#+k$qrGb;?v$3DMa>-&=5<ly5gH1L`cQ{n%#=#8!j6>E8532r
z&{;5bm0N2C+o!zxl*vCz<M|qQ{d|RoT;seZ;e`cR60Ki%jrf3+k>75xJpMlB#2u#-
zx6=}^Qfc=n=M#2kCwQ~9&aM=h3LGBZVR*+;dxkLS&0BvQBSlP4C!xu6`Mt+Z`34S_
zaDt*uz(7OU;MCt<(-o{}?fMUOVo3T&=>+_6RMyM{x-(R2G_Z^xaT88;8*3Kw7(h$J
z&48<CyNr!6$wGO0{HDf#U*=tQ)9PmjTyu-83>moOlt9k<(<7wK+$d{rq0Y9WoLGT(
zqy2ny1IzP>NqN!(&r{)aD+04#QU&DDm*I$)fYiLTx3E@bevpB2TcgJkL}P;{OXM{O
zSkug31(IM&g+?;GCa>d*zrlr99B`v9C?tb!3(jdtr``r0ObY&~dn!=Jmxpl{slp)l
zsTY3Sz&qxIc3IEBq|0nJpOVmaF7k{g*8unI+oT@mQRcL|0i8idl|P-KcMYemE$wSr
zPD=q@VG|Lm>*ZTO-b#s54<arLCRZLpp$Sa)xH(3%guyV8H2(J*B-t@%x#jrSd@T?X
zH4}wS&%C^!@L2UrA(Wb?Ev#g0@K=04RRp@1sVg>)xhwRM$;41)&53NytJE%0(>h4f
zy9@-waiY9{ib+C}m}Ypjcz}2r*_<*}UcI&@erbLcJ1{(nN*MvRDDP?(7%!^zHZ;7`
zsb60r-I+B5o4h+xBjD+&Q*tkTwZRvBbGDeWBuy=*mHCus#jiHeiwurgjka;e`HO1~
zSC8jR-?f0lPk(a-wT8)QN6i6u&6?NUGqs!g-rWJbycn%Cvm=NSNlx~IkQ@(sm95)N
zZ09dUvVIl(VIk!4zb6zXl%@J>s;AN@-MzW?qQkN`qY_H2GpHMuaDFr2!OP$OjJi;8
zOA(q4!blVU%8Rdb*`Qh%SVPuA_yQfY<^fezw@s;To7=_?9-omiOL9&To!<ReKP(M1
z-DL~b59>Kbb2L!*;vEo%J?c>RSn_1;P5)~1=$D-MhtDRd<$TO)Gd*^4t5}WFv919e
zdU5GZpJfMyX%Dr?ahhj!eP~r_^sRr2cvYN)Z%0vYU#?W(sTk3rJ84ORxjNWExB5|g
z`VWyYeGD)dFQ1%&5&4Sx<p#;9^SsE?uB5dLB-X``)0;I${FSp>vg;~O-|bC{4Ya$k
zZ8mjRW7l0DJT>|wd~NFC54gfvruWkUWV5QJZ3Q~vfzMJNF0&Cf=(`oodEC64&Ha8&
z|A=oD|N5O49U^HSaOxC3?2laKX)dWuu~ISMnP2qs)UK}NT0x>PzB}O~m}rDPfz}VB
z_T#@}G@{#%^TuVTzIiPP6r-9H4Jfi0J<3<`WaYe8M>3FtiV7V^Zb&Z6Q&rW*S(mCZ
z-xkQIe4KQ(N`eV-RcK-Tdj9+eO#oAo<-s)CcErz>A64q&UPb3j8;}HN+sRA%DrPMl
zDfWe~Lti;VqK@+{xpW0M69Q-ETS0b+T+l<K-|jO5SP%}!wd%|tgQq!m2Ba!A+_{x7
zPrZe>{?T#*5|dlV?g{a;FY;FWbSS`5!Ns+Rn!bMNFIL9ESDW+Snf?+8=M|m0#-a9T
z10mLoT7o$)kx5Ze?r!yPE>i~G_V4n{`VQ9UKDw9zS9Aqo9mDCg4?S$Jh}ex($i8yF
zqzxHSPM#D9pqYrT9o*@2s}m37;R}6|;<~DrA*K?kC-iVKNL(3!R4+^16Z)P+n_hWC
zNy-NK9v6lqm8k1G1g1vRb9rTxAX~ih)YMiDv*%qDdLGqjp4bm{Y$y45!ojr^pbwPn
z^<U3NuK)Bln3(@}eck0BZv&SD>BrmnM~E~y3P%`7yrg^p)KmIH9xsH#c}_r+^FG_X
zzF{Iu?@IcmC9eazV%>4%xN{G$;q0At^x)qVu+{Cw$jG&?#Kp(wpM}Pk@K8wxubq>l
zJb=T-=b!1{o2;z!w!0`GX$AMm^LBL})m5vMXHJEqWZmXgaPaYdw{@A)Lwk?ts&=cT
zdw5VG$D;d@M|C)p*(I*WlXZ?K3IlV#d*$!-;Y1pZMLN6kyP##Ni5N!m9rzNq8P&QI
z#q4Iv#W#rErFcp|{q_Dd>Y8z(;4$JOr0&^Ew`V#h(vZ~U_RhcE=ehm0({w#haYv5u
zuk8USqcm1MrJMq&{H}jCza4*!YV?D$q;`^kqWo@NnbA?X@fp$h+|RI}%)(SMz@VuT
z^juadgo6~3G9EA5aH3XoR^{h(utqhSP}1a{Bva+Ds!7ph3>Glz<KBuX>;sL|c(M`O
zqw<+Kc^F0aEukXTyKQ@$(yJJ2GuNRp0mB`(iO{&yN=tTT;y<~$9QuAGtrw<QPfM`-
zM}<(o6Lr2Mg{Q>5i8XIa=o`ws0aYH;np)4nE{V3}Q1}DCXl(E9B)hO&I<!fL7uq>D
z8Y~!@@%X^ih}}cbW%3F!YRH_}D1*g)+G5itA6}{ZIvBgz!RCX2ZoZA7ipq5X`(ebK
zioEu%w1Y??Zq{3&1B}a1hC77LYvYgz;w}at$dhmfA_KPxEc!s?qfYq%lQ=NUxT1{e
z4fL%YB9g=;{tW<#bFaF1+LIc&1P26pj_cro*UPl;phac!PxO%QrU@sauw*s-pgym<
zlJM+U@1r`5Nt3*r$plN~#Hl3a8f6456J(EW&{H0nOO^O;m@smCUGn+`DI^HoK^wu|
z7g)@7K-m=@P32KnxVbt=HOa4=Pv*6IentyUw;+5!3<H;{SNQ<nJu;F**nI8uX6oDq
zJ1@qU8Km0doy*ypSd%UHMXEQDFPzs|co8Sq5eoya%fd=c#VV~N$mx;ac^f_G)JBm<
z=CzNGtzAE`;zn_@F~iRf7z7YnrMDa#JJR>0*{7>gZA@@E&`qnJGy0BaFMTstbG8o!
z?h2@@YlcB??^TDQ(g7-ag+v&EfWDu(N_}I(hx_e}RUNUQUeLN*8WRXNkfC!@f-Tgn
zftkovC<fm_6rYO2OqntL7iuIL3NumeW`L9Jp!<lL?T4@Ya?sOVT-MAscjFwDQUa7c
z949Uo9g)PBw<x0DA1?+BJhO3r=W3Y%RQbmSe<-V#>O18Cdob{?fP%oA=!l^-Y@|K{
zQ_TF*+Z%4!z2GhD&=~XI0z+)M9?URWlJlsZvx~Tq`Qryyq8s3g)C0?&tRew5IW{)x
zv#ozqrwLq#B2`1VeA_cjY>dLMX~!77Ze@kd12zK6TBql%0ujQYiAaPsX)DMc+#OK{
zd1oaeTJ{VYxI<(ZyYFh&g|QTh#I&+!g6iwDcR3(h2=J{$Pkj`%p#0&|#cmb|J-I8s
z4wG)~&Mn4ibz%^!^^qp8M4hdB==7zA`p1#UCcK?g%#bVIGAw<gSn!>xqFJ(+TxumI
zAAr7OR^#>G0i(gJwN{NiW{~`fzNP;5YsUd|v$1#vE3>iH9Tl!96O*&t&4|Y5V8J(T
z%)ow_7)>rZUTQX0WRFm1X(PJZ0xR4K6;3&Rem_;(vRc>g!kwg~p$w9g1>yG;(q4Ji
z+s_npRN$b0oS%HdM%tpuwQjC#1r^N+AnXobzHbkMqXw0v`FUSav{b`Tu{1B|GQBNM
zh%!&FIoG*pSs8AqEe)!x74BBny?k<ep>BJJ&A`&C69w1e0;BUpeWM)wc23f^^zujW
z8;Zxo$~DE^yQlsM9o9NznKZb!^vcvWhjg{4YPalP&(FN6>YqQmkz6*ST7Lf=p<ZpW
zyIgLRHx_0*ca}6j5vgCR(N;;qF$FX});(xNXrOurgj(Kbe=KxLRZ7Fis{}P8JOV>y
zkGMDPPy&r{EK5zfAi57}*Hlqmran+r$wuAo{p(gn>shpV$75C2qPn>-Z+Zb!ViL*~
zrB?M%9MAxBu7%6o&NbXl%f@tlJrB?XPd)hmjchXhPh_)AZVho(J%MCPI*qr!fN9n(
zs*qvlr7D9|Gsz4{Vr|<aN@z9^aj&kg52U>>SG-fT*>{d~+H0IE){$?#ekn-!GK`xu
ztK+bZAkYJUEE(Lhs&MpdxAqAAo%udbLfy#p&1|S9+-ubh5wO9C1NOKL3vUqj4sX4m
zf=f+4mgSjev(HLJmnweQN@aJ>+62KZgRYnpNN<gh;vAQ;@48%Iw`^Zl_@2BW1WhL6
z8Cl-|iAxqTk8n5PAmu7K_dfSY2JGefMVnd!KPERs4En`f|1nfJN7E^@GcOer&P{ls
ztMlM0aA<Bxn>TyNEwPs+7B$oL+3_gNeVm)lqTy9liz^}Q6T~OLw$M4Y)Q>Jhk=4rP
z-%}QKik9a+ORHsWbegof(E;98vNco5BojkteCH(TTk&vTY3gZr!dM3+!XO6*3Q_qm
z97Iq;59D+y6(4_C&)|!Nz2e8A8z*VK?WDz3nheha>Ty}>nMJ+2-$Ke*l{UdOH4APH
zIa$Bq2ED(PpY?QsoyLIQnLaU`bG>3fr#jeaJT(}UcS7iW>m|lVc6YoCdevB|nqG(a
z53ea1VNo99h1K{(oo@LtOQ-s<*0^r@Rz2AT|C`h=?_xt|lD@pQuqdK_xrW0W_GZ_8
z>>E3mx5as7=WfYW{^p;<0=*}fvVSuB?w;i8IQ)0t^p{v{esl#lZ*g4y)yfDh*O5Sv
zmiOy_ktUps|2am8nT7d(*KS?LnY72{xT&jOBV-(0y#dA&r~5&P>%i<r5x^8nAxh1c
zgn#*4q~Z0tw=|lz2*HmnIFIjl8<gbm%Q<`h-4wXo!Hdo9hAQaJ%j1`aF_8SICqv-Q
z=vE%UX5;hw(Ix51JRkpVO-flI1PB5kKf=O!ia}-+Ig6GCbK9a$1Wum6ggV}NzXNxp
zcII&uitY3E&!6G23A#J#=UPTHcXWaZ0-rozj?YJPr%Yowc^ts&%yOppg|K4<1P>{1
z$}l)u+Oes`@f~@hht@%351x7QWN7V<yjEs5|8haJOmq^3E_nZjxPk=JA^{s?jK^N+
zL~2eVIiLV7!M4#jTN@DvQ0Y2WuMzbC!S}>(u5nWL(XZ*Wi)+3PFy)IUgpZ)46%J#c
zjibTDedL`^EhKoUnvV+|<nyX?nRb5)2Z5*0$#_TeSC=wc_fx-mnx|r{)H{|cO5h>T
zsWlxzF9tLF<W@9~+gg@~atEVI3)=*-2qZX1WJ=7<!+ND0y=Ux)hkc=X1FYs^-}Q>A
znW5@2GC1xwRx|}`SF1gBEd`){PFdlg8XI`F+LbGo(KEVcZq@c3#}Gp;e{;HFi%0ej
zDr9^w>Njrz-X5O~KRqbiAmp=)vQq{gZ=+b0Fi+x8iH+V*cE2q@tU_}Rs?Z_wbHcIn
z&2lTOS8AURszj{?u0Z*ZGh2|gacolC2r)|smc-E^6qF+M-0$h2A6<ZA0vUs}`r{{e
z&i-l&VffA-KNI)2=ItFtW?uJ)H2=aPoD<E}fQ7MR#~zX)U`8-3dvh2m#I<rV>@6e>
z#G#f_;2E3Se?T$HO-&K~qH+0M;^=?G$6BJ-m6%4KrvG;`3_EWbGTi*^0qW1#HmZx-
z7spTbp=R7KuzC8vk=s}vb-?)Xu`xVQ2;*<`SBobciYAMmD|j(!!+b==FRq|^RC!;4
z=_cgBNZ_)+<z85?K><V57D`RG<cbXvB_0l69}l<Trv|BgW&!i64-{0!=z|aBd_|nD
zjAz^er>BCId9cYS$j34Csb}CheXx^#_svZkhgV2h^Ruit^R0|j?9mp4Cz1QzLM%g^
zNJGpM?|$}>*wvlGvMNHsc~H{&GV$$s8uO%n8y0y{)&BYJgyq}eQ)-SgUO6a1;;@2r
zxBVMzbu_9X@#4FVt#IH*{o>}*4`(;>aO_KILULN!Y?6c{%9S@7z`9f?iDA|R!kTcu
z!(V*x1o}gNE8|hqz<?BG0nOhu1yxOL4(-0Zzsl%8Lu>Rn^rGhQl+GiYj5t9CP9_kA
zut`usRYObX%`~l(zT!#9%I7c8+qAL)^<DDI!gEK~bxE&<W3qta|7LvypTf#j8U%~P
zaL_Xvi2N06HTQ7Gwj3%<4~8xwnew|>%DEQ;QG2@_P90LmrZ9`4bc(q?@%`fMsBpYR
zcw&(yhD(P=p>72>XSSS!nvY5$(WQpjx$G$t!7p6<c-aiQb{^sKXFp>zbYUN5XD@2W
zM2<t?YRUoR+R4sE(zF#sR*~Qz`^~&O*Xn9ov*Z@z;YOCIg=UeiD#hj$CdU9y2$WkO
zJ2r}$I9Zbp4D8&L_uW$w0h~A*DdeSDfOfx;kA!>?YmPOnb#zmj7;?WQE^6i^91N~L
z8`nHEJ}MZnK_P#p4T!ghWxXX*aj|3s-Qm05lsVx#k`rx)nlh3NnnJ}W&@F2>xntt&
zEKTwXmloT{hM-HkI<bSOxcoUe6L7;Lt>RneS>v3MEi<W9+AXVJT=8i9W-?e8hsqji
zT1^_z8Zyh;Q~V#ov&~jFk{FqP$n5F=a3&)#G;_autz4vGj#sC}Nu*X@puV;gs1J^6
zD^>&3fVT*g`D@GbTxWH=(4x7dMeRT#*A%`8jcelY+Syp}pfPc}#>=OT)nTH;b8(p?
zx1|?YF%*s}^wTgo?}k;BoyD)uPdC{_(9Er~i3jXnP5{hg^fFJY2mh_su)_`@>7t%=
zsIiIgvm3dDiTwj_nb@gG?ilxW)?Od2yYjEFhY(&I`Wb^Ar`FS-Jor**ms)f0nkHC=
zDl9Ri4LgMs|CmbejhwvX(_~f3zewqGQE&Hypd=z$8QHBTH<^#v{JI9{FoTNTq_^qM
zsix}fl^u6pgQ65*TNtRk@l@@7fS|DEcGaYm5Ljpfk)|9D`}XmLM8Wzd%vVgg>i!{k
zPNk6R=zqeQ<iy0>toP`vXt{V)F60C|*KiZ686q1Gr`e+WQ<AbQwagSP!RDc)&e&xH
z34e)|e2%-JhM1rtvKJyu8D$zLQ@_`8D{Wc1iQ~C{xoh{J4SVfYqr87sYRo-a-KeSH
zQsR0CzFq60w3{K1%Q3iG+0^AdanMl;untX-PqXH%wW$yuR|wD2Xe%Vk<%r$iJ>Xte
zmzv8XR<zs9Q4-?*2-(WZ!G4Wm0l^olXy7C*2x1%vKuLh>WU%hro^$k$7!tj|x+q{3
z<k=5F$FeS?uSi_gDJ65TodNp|s1fBs;!ab6-NyHi3)hI%&Sy@Y_RW!(ZqUYhn(BXe
zs3K}+JEYnBx5*l|&JOdDDdHHe+=+xPSKdgqVYJ++Fu%dh)na3(EJoK?eVsvOEfcHf
zBa)3ryunVgMjv!fJt}75anAaq&Q(FMV-)%0nq&^VJ8G@b5?5~oCe)i_NXB&1fJsH$
zv$AO*kZoTyc=Mfer@AN%I#vH@%h{8`@~*lSc6L<=X(?Z#yRL?))-rH)RHQ3Qzo+7+
zywmqPPLmfw{QVM^kfuBFL$H&n;jEUBC!46-1~1*f$7bLPS%Ef}J&jRd*AVOUFk7D6
z9^>h$=$c=CO7BvCIPTB5&bUhGki@jN^aLH!SlR_GVR?LJ^r06e08FF%fD^1UIUG7E
zh34DbBd<=<_|nHnPz@U)p3-5WZ?8l~M7hVEz!HHJaMPD~)kW_y?mHU=k0`h5$(yha
zNcf20#4Zhtg(Boxs$rdj1sbgF*lD9n7e9`*8$s$4$ebL%9vRM6@hTJ7P$I;}q|wfh
zqpyajOsN*vu3i{-diW@P{3AV%j=Ee-zxrj6d#|U;#R@6Jo_Y{m#Trih4HVo4k?`N-
z71Mu7wQzF$@8s3Cj!ZmW2Yl~yeT4vHqy7>o798_p9(zJIW-mC`ubBTCT<GpHwu|^!
zP?A(--C4ESwazxTs7Dh?LLC*az?|PJ@B8NH{p#xNq2%2-^}E6O?ZFJ?CFszg|Id7%
zX{e!i<oV+BxkG%8OmIHthviJ_K>YA7!r%RKA?R>EuvVzs2g4Sq&-{Czcd{vytAG&)
z<eg9^_U5d$!LZ##W?rpF%}SYQzS`yAK|mkv<$nX-0Ff3<oc9zTa0Z`?fbvC#a(yb$
zBxo+bFMpUzZRoO`9J1aA#?QC-^T~4)g=X1dC&@QZ-?vA=Qp-n5%LSKpDlRDFV%1)S
zH6Qk2CQN;j`lHVb@bMDnM}a%*g<w=>%r(9IFXbh?RY>nB!$r$7P=g1`Mjjg>cUtn;
zyXTOhoY;+Z7Cc7T=t+1Kfp*C(O(3>qfy7WPnMsgJNIaE_XW3LpYfINGRadDd$2I2Q
zB2LF-lf7Cc83_9jlNk)?l4%*kAe6;xszk|26mlL){l1~s-3?*MNGpI<nBl)fn%NY3
z(bsb!Zco|!ppx>2>Ajo|uga|(*ryw(WcU*ilj0TEqZG?zJhE)W#*{V+C6H7BFS_}h
zh@$e9g+>{ZaW2ckCqR~`s*^i5S$vP*m6;UE=h_94RK_tUW?jBZb%g^egV~nLNl^@R
zla)c9r+>a1dI#cCUa{JPIuJsv;zi@JLB8tI;PEos5uDXyGEeLl!1lgwu?Q|nI-hF3
zij1ZrQZ)8gpanfcq<BN#mP&d@lxQirOirHUK)ZU~JTcO@{gFUD-8qL9)#MGoPekq&
zF4&&V^1H)e<}m(?&Wi(Mn3Ipf_qGb0lIr#Ie!-kmXI|D(ZZT_~Xj%f*jJE$<Lme<_
z#;JRTQEO^_@)q~WF%eh5Izx+!Y(elPRpM9F5)YO4l=oH;LAaY(HmsBEkwD@_O=%%b
zhdIs=5mQM`Ijn_QN`q{al0y2ncM?iQYVacrXANZ~YqIo?j=~0G{HD8<LPH(p%~a6e
z@Vpq47wiYqc)&Q$gWS`Ow{u?X3{};5U6qv6IGV;_#r?i5JWvLbIbCqjYRa9;5fVQ-
zG%*M^y|Z{T8YvtiAKCCh5~xx&<3luwmrSYX#B*-Yvt-0N9RyC_l4hVC9Pw;6CZj;D
z<=0#cj2;g6U8mTpMacUU6}i3HF6Znc_G4-Xhxj$WyN#d^JNM9)0mZ);qo|%lpb@%^
zFP>rZni;|H2-A=TiUhS61}JCm=j(cdue(=^e#f-{VGUUzMssj8zu9=pJ^9IV9si~F
z-w9i2An5~@^IceGrr6@M`Q^zYIEqxV%hHE!b7z_2L5*=S*x~To5uS@F<%mbtCV2pz
zO-(EsRI;bY@Ef#>qd)t|ql*<cSp1hPZo^97Ir0W&i=#CPj0<7(ag+l)dw<FOV9G*=
zz5mWrvyJaD*3fd_Vmxn1+t4~(Mlmg0hN5tr&kroAn$t3?$dGcd5z`e3%7wPn`H3Os
z)~qR`Qq~~NbJ0-DQx-Wz)ohpSzrEi-^Ip-$z0LVjaw7`U#N3*D_ko+?J)<r}L6fT&
z%CH$<$@h$hCt=1ctJhI3$jbOslLo2y=isCz*&?d7$7|4%6$otU6(;acu|_er6W9ET
z(xgt43fsz#GJ7W@Qs+3;{ac`cIP6eKo1@53w8Gp$GtZV}`klWx=VZ*l<X!2104QEF
z|JWFSEP2du-}E=<tA?;~`!QVbt(SE@rb$oIAke_t*r^PW+*PoC_082!3S}%ab2R>f
zk!>EyNyEYxZ_-?Pp$N|KG{L#_k;Z*=!`)4D*J%NiZmWC9Z`gfjooqYv`dWe<pgY#6
zqnH?lnlN-sq<5$X<emp3fv>JS$n^+}>F=1*A>An_MurTcYbUG<Qg*>j)>?JgVO!iC
zpt={J(mOSOX%O&bvi~LZufEmc@fv;mTaE<?V0~B}#1fQH`jrlvvBDgTZ6qKdpo>5i
zuNJB5plWAhc-i`DQiMahW@2auD^(kqvu-F>O4<O0oV}-UYgv0+|H{3=8Kym%gN+cD
zo;gTap+39ULV>jdQX<8W@uK&ewNy-2x^7o;HBjU8<ieZ_ME%%qL5aq2`B97q*pjcb
z<%e@KG8?GPaEUV3KnMe<aNpCugr5TZ-=#;=U@N8In4DKK#0)HV((6q^RkNzXS%)^5
zXol#vEo2ZF3Aq~m*4TDFDeZc8xAW%Qcf)c4;1rXEts!J3`dJ1csgG!K5{bV0A8LBU
zEmKrD5RCu6TnXA^R*i%suPilJ4?T00Gn5?;-E7e1g4ZMQM7<_J&hgH^A(~Y7yNumq
zcgIksoN813ye-YIj6vAFpN=yhr&**OGV9KSnod{Ej*aLr=&r4QF@WLiPsb14@ghL(
zgZ=i>tbsf!EYK$?2Ggh$<}WpZI89x_&eMRzl&zDUV?{;`;1->wIbk(y&+B)ctPM~U
zo!zIcy$U)D1*Py}E0XAO5Pry_PG=0G*`gOk1hN-?2!W-vJCGJtpqVV_PsO0Gu2Ewz
zMX5@cKcHiF)nxIUT2xDSamf@}wITVl;T?H{wQ)weod?yI;!AC&dvTFAkMK7(eM*zz
z)1r-1-1{A=CkED@#OtgF4@b28?pU)Y_Wv;UPC>E;%C_!o+s16$cK2-Cwr$(CZQHhO
z+qTWq`^3EwH||~!>#6>iim0fEKQl*;{6?}~G}T+J>)EnsADRxuPkVCJi6R~AgbMv+
zkp^o!CQzr14XF`_9N_~ev!Oa7U3sD;$%O~fmD1^E<(0@aFsA@BIRCSv{c<-)OWfoD
zT~iI(y6B^Y>MiKYHy{|?`1!@2QW!>HFLaB^xM7z|OE94M3gMnj{$_8c;}^6Zn11n!
zZYqM}oZATcS})x?3vAyh;BX-6(skuS<%|<7{tUCng?NS8CF2FJ4Q(VWr*KsB-LUt-
zB~Rdtbs|fY2px3@u%MNK&=#evhg)Cm{$Q<0vTuo!2bOcW>}0CHhDsI>w{Ed`$L6v}
zwquE;dR;BY{Hs9)VO>7b_|s=+9>SQP+N55G9DYR%xT^;B+4;-^MnA=p?3~@|gaj}2
zW3Tu&yjYrsHH$WuQ)d_U`AG<<*b(xjxLI-#b#-rI(FJ5zyWzz+xdEEKH;&93A3NVV
z52?SbDKwM^(PNuzIrPLlL9^OK%COm5KI;aN>lVo_V8da1p5)UfUH+Ky_2+reRH>X=
zHkq02)rY-YDg!lVhqgK*6k|uNS%yR|$`s8PIBQOkc<`p2%apQGi8SiKMt;vDTML=1
zILc-9;V5j1PFM<{*<{Ud_6l56_&3uX=Yoa@cvoh;thHwTM^uVKg!86+ZblAqs+#ya
z7iTG>#54Ry6F#v{70qp=nr7Y5lCw>9vqxph+=RJJaYHLo8*+g3j<~ckV>xRA3Jm&v
zEyua><!)&Ze-(L{GHigu+!79`P1efd045JN@So*STqjA!Y%$i}X%rJSKz_%@f3MIb
z8Arj4nbGqwCT!eAA6a{k^V=uTB?XwUsf5R6(FWI>L$S*i`%i2`o?C4vN>#V++|M+G
zmdvJlKprXM^jE`9JhmJOGj7RL$HNO$1sLW>)2ISy$sbCp0%I>Z^P4mJr-{ekr~)Rm
zD;kmgu76Okb?L>$hJOQPQ#Ym`6kmHhG`H9Ek>3Yh$!hiKN4z5Wb51u~yIC_Y*e4Ww
z`8mUnjU)JC#({fbBj!%|qA&{P?;g5taoMXLTC#4eh}H1>%n{3%9(e9FvR{M@Wv>2R
zLsgeJp;C(IgQ3UAvr|_81FgjVB#WkldE&q3AeqFTwc8^R5wUjE%Y<mLv1y-W?vWIF
zLClz!Qk0-vX}L6YV($9E6@4gT`A<#wublb+KjEZj`%P2%ZwP)}zw6fxHu&xnl~cID
zqgZSnUjp8%MUdxaA?pZk>Spez+OMCGzw(8|FvJ@KN!W&*$jq592MUL`4zU9*B?}u1
zB@rwENH~3EdsicW90le{Mb#VG*#77-L)-KO_wY!RQE8F3|C-7dj!s)jY&_kTcg?_N
zA$unp7~x-h#$L9dR%hIP(z@}tBW7A#%@<(Nd`QP%_a9Hf+U*+`*pbNp%%f<3dw(2=
z8PgYnvwTPDeAd+1$oz2I<-s@6@j(B4x`(qyg64H$@~Zib!}k3A_2&8-8`FkMc$|Fo
zvw!_^(52jI&<-H#61PA=a|aO{X9;NHJ9V{KtZ++06q9$zi+D5;g@Qzt_ilGpEixri
z1E&n1gJqM4C%ISVL7a#{(SH+<QY;8KSxhv<Y{?6)v@L8^efzf3074avt>)-68XH5K
z<tAx+ASe~LcN>@Il$zxI7UiN*K9={O?hOh@jQ%Ez6Z>i9Nh)mMr>?5J?1-Q!Jxz)v
zO9kmOYe0fDXV!AA5F<yHPE%sm%Sn$!Or(!iMj{Y09HQNA&!z(p*#qZRqdiq$O1M}^
z5aC-9a5smOL+Okq!G_Prq@%G8K0S!Hz=JHmhj-O#DpFR@tw_+vn^x+4&XF-g4d7W7
zaFT3M6{*)sfy}C{J<H{cZc_deD3WwmI9pw<#<onpsbkdWqG%Ki>w@G!tt<@ILXi{~
zoLw{Ug<V61FuY2Gdv-ZA-szv%9-U3paDgwdvQoDp)??QhvIvz%I~F~o%zy2-Pcg4W
zc~G9O^cnvml3r3TULEg)k&}}?CvDMOf5|f@`Y#!KhfS|hL+m~1V$MB?_VS;chF+#l
z+3*Ev`ZrQpFaD5FR;`N6;c`+*>6x2hBI$6qYoqP{lT&u2amVGM7boU)$5!#8KP*rE
zh9UCSYT8B3heyr49SylP6N&<w7{+9IINTa`8+u}`uniN0h^n7v;Ccz6hqrXy4rJuM
z@I-%A?^`S*j`JkWH28(D!bYih(Z`9&On&(YmOi8Dl`x++F3dlzS-9+WV39IIqhylg
zKoynr2aed!+)wUoCmRNlvcbdZ_fJ3$r7*G%+61S<GzNskH?H^4yIdIxmFywy$tQw4
z1eehv2ND%r#*Lx2moh?GS&De-N3}0ZBgt>AvC`4zLl?>6Cqf$v2?+`7sZ->TbttSs
zW0<i<Y?fI?^sYDfRbM7?lXf5l@PgL71I_P7><q*%t1$9B*&6XWqtY&hb9H$;pA(%w
zK*Hq7v*lH1zQklv=(V57PBc<W2I|!E3RtEjMcErc+VJXeZTmiVz3w$_)@0{5@YSNY
z-OUf3g^$|7>e=)DxNoa->EieDZ5*!9c$F_GF&~*?Ll7W-3T@LKe}TFm`^7dUasc`E
z`pZ`3og~a9dWs`el&#Ndh)C1_z?WRp$gmAxT~Cz@Vtxp_7}5@tr+?iU`2SJ38?IfU
zuukc)DLExM1py4#El+6P9gcLZ-=d~dzmpdxmr=puK3k$>oxTahL=nt&OWZK=IhGp^
zhc_R4-VnwYtZp;ofFF+ZSC(E48RQ_AEeoG<%9LQI%&gEVg-f4f+GVeO;0?1@94D~y
zz;n&tn8`~Y!#FwkqM)k<(O4}q1cS{)Xm#krO+92AN+vQ1gv*5Qb&sCFh`Z`$iGz{m
zy}oi<9o5rt*t!yvq$C8iBG2<LJJ6WaIM7T#F1GwL=_;EYH807u=X9W$oXz^+cHWMA
z?|D4l4V?69#v_5gNX)&!bUDZhgG!C-jJyT&ir)}joHqr=*6-m^c8DzH5HUe!)n*ec
zvU;72K_);LDgUJI3^w#Ro**#~v=QyBu=ojDb&OO*9*sew%CSZd>L+%<HyuTeHIZ?x
zxqF+DVD9hJ?^J{Jz;VtV#JOPUw^Rc^&Tx6Jm8l@<A+zueenG0X1z;u|hbAIhxcZ&w
zHfEwby3;C3b@JGJ!G3l~LzQ@=**q^Dohr2oO4g=6eQKq^k5yaHN%~Te7@ugD9&Yn{
z4XXAJjq2RF*)k-g5<wAy!f|w=<vljAjA;qq2A1k2Qon&17qZpUrz7~zvAXnn9OCWD
zHyORig`aAbJZd%`FiX4a#Xur*9%vWQLm#ZYuyj&37u%PHUi`+W;)wk3DPqben`79+
zYA_8w1z5gB*5?s>wj4XWR=7Lx!47~H;3Iz<;ZBQv_|++5fg_lfe#Zu*D^`2l&t9Zm
zT@aU%XYiFZUv?wfa7Xz3-$S!O{pRQ!@Y2XJ+Xd!7G8k%mTr8=<@0ewI3{o>c%^-f1
z-)k>$%8A^Km7lH`xRr*385U2E05ZCnv$-C=C0JQ<{CVwm<WF&}$g-*zwnVSf<SYmB
zk+Xj_!>_J;i~f{iT6Ba-JW+C!`PmYoj)JpPc#54w52s{ksDEp&CI`CvCPLrS(ZAav
zi`RzgCG4K#jdhzzf8lD6fENA7Li``pFh+K^|EdrVVNOJEvfOX$<l&;H3!&=t=GJjr
zwfJpfPg>Q5^s0KmJ?kf6-~Ug7S$KY-cbSRvA1MjqdTpv@iNt{yKG+-=+L{hH(x`y{
zq38}=v;dq275?Bsm`?5f30$J<{Ko!%qbjO*?3xXdH<wB&@Q~>-{weHR0LKIG{T{;~
zyaV&TdDhKw`;7C(?~ZWzeSI?Z4}Ao#U1oXX)KQ8MW(|Y_Kx>73^=SXUyfCZJvmBnT
zYww*WezgMN(!Glrt-yl+Da6QK(8jL^I|D6SDL7`t#Af2(CWp5pFL?o<<LWT+jTBYP
z#`#LIh343mF#IDbY89_?;sW;jVKnow@vz8}hM%_gwV?>aRWWILgIpTu+>t+s9Z2AH
zoCi~`d19u+f)SaoVV)SF78!cSsVh^nm>2_}4-W1(4_@gOsw_P*!uUG(4C4#mW8=W|
zZn>4RC{lfO9OzcrDm&waMBadygfVv<CYVZT!*JTEyOr7u?ofQZw@asCP6e+7Ymrh+
zG**B2MP17%lGg}Qr~_!#-1{f62q-xCRDno{j_zCy)Vk7Xha`3>hGlc1;IUdz2aTRO
zxmfOH%~FK8mxU<Vj@`+y|E9{Q&=Hp!tQBWA;G-PS(N<0w?lM}LGl*R9*Gwt%H~S)8
zK%^<;5KpooR#9cxa`LDp>$A&LB5a&$8EK1fMi9=Qw>Gfo79xoA%`kf&TDcVr^y`kQ
zdxW4g)}n6@>0yX$1S#*ihcL^Fx)AOGnrXpt)@6^k4y&~B9|RKTEF0*>*n<44P_I??
zRdSdd%Py8T5LEdlra&64F?l7u<eO^z<H{HmK4$?CST^!%2P8}?J97!}34CBFMD?{S
z6{{J7K`6d%S>lPS#l>RnIgyW2z@V(E!I)|d6eHe>O|F7gZHdU#&z$ksrkHM^RE}*C
zrC=>u&TNlMIntR^Shl|+oz{0FS`N^SyG37f!aI8Nc+%{b`vBi{d;<jY$yGQP$O~o@
z+!G93EI1pBM63Z|%aXtJh=#G{^}6$7^Yi+S1)lzFf-PRHLWaZ`kn>pzY}6+V^Y0ad
zT-nVA4BoGv5PKH#00o8&iE{U%R$!dXZ(jrG3n7m$SEML?Y;3QCtrp&tHz*|3E|8%M
zm%7$oNz4!*@|RGizt?sbcbua3ODMZ?zz}cvX1wf%>7cDtGDwOL=$s4<DC!}u^5?Zx
z)MZMi#xnv*0`e82?2Fq`^=rP576^JuZib_f!9qH%makwy6D4#3K&bZ5^QEx(>-j$#
z6EKDrA&BF+HB%8p86v>4_42Lj&lBm)hjarEYY)Z@(rmD>K?rH1D4)29CV8$xct>L9
z{VhW~QORY5!>H6pm8|-@-d{w5HzQNk@p|s4ey=uld_`+a^|lGfGXtpt@O1>y={z!h
zVhLJ?+f(`L)GmKbUOurwIBaHf_w4y)hFO);zevIj@)$+H4w4IX<Rr-pIzrNu!5A4J
zKF35m10%uK;&U^_rQ_`DN!=sZ6SOqI!;2P4oZ@_Q9U5%{G}_jOshO~C)|X{{OduNd
z3v&d?myV5XmL)$@KQqK<2EjCD?vqa^SAil%Cxji?E7|G2&O+zW@tL@<tkPQ^6fZ^B
z+{q(F*Nl@k@$D{(qU>cl3=<pHqOCx(rFV>zSn{sr#6agt+BdYUii2cVKyF1=D6En(
zvy2scmP>g7NYjBo3Tt^?DBpj>VXq=@!O^6ypP9JYHEu@ihv|;5hyz&Cn<W$ie;<ni
zhE3^=sf5uRcB0iN)!jV$Kz}dNPdI}pRM**<F&2t!oGm>pt-#<^I9(rjv@#kCS;?i$
zI9-Z-LUU+d9&0bgnziZ?uNORW$hS7`<azTnrq#7!1tnW{3XX)HppIB<uvs_>UWBH|
z;}0iz?jW%@Xg=g!9szJ-vK&dr<$as-a*I^cOeOKE4pfBMmT3DQ1A3J><PYJo`z1xF
znj)}pUUl|O0nOO+&ABEABXOw|-)-%DA_-Y3^K^v1+wEB8MR11NV9n4rE_7v{lCsZE
z9Xp~u;Ty6N6b)I*L?t9J+C3`6nSk?@Xq}?$C;m<$gK$}^B$0sW`fJZ8Z4_6WBJcK<
z+sm7VFe+K`o{~467J<3_<G(3o%A_vt+TgoLbvejaNr*`1ljik3<L2mCNgHWaM+aFc
zDrVhKF2mrUU@Q=SNMMkPsN<a`ZqpQaM6P}ulT+%OGJf&9aP0t+$BWe$4{=lW+oWw1
zpAzz^Im8T_Hy68a=z9Z&Ni?;0z8aa+o@K`RYGn|`K=J6ikp(9ws=!S+MYlKIni$c3
zsO86VI;=&drGIC2Z1#DOLr$q|v6@~Om%B0ac+~5o-%<(p6%lV7kq2fYblN;yZ2qc^
z@jS)jn>cZ*GTD;74m|cY;UsnWK37Qf_OB5^XZS`k*v+maAClR~nQD}X|2C_UI-4u*
zaAgqM+g@rAO~j@gnkW^PSDi}4PnRi<1U4qczn?7QRq{fS-xxk%j}Elri%Lu2NJwE1
z!*S#E_m<QA(35lAqq9#kO%+p%!|JGyn<aUY^jD+@DFH60e@=RS($kc5++s%=>O2W=
z5RIo0X_Do$yLg^=!{#M4oV~(t+-|%lhs_WA#(2!DoY^Z+;K-#?BDnq(%c?b9bYrG;
z!W62=r+gx<I&@GND<M^2r8@YxBm-$ASeO5(7`aT1MU7>geTwyHmV>-Z8e+$w;N)_f
zmpQN~8Bd<`!;$C2mC2}}i%7vl4Qb6;_`!Gsc^MAGg5G10(BzmXA|=^;n-01vc}usP
z#15+BRc=kGcacVOJ|f$^3;3*=F8bn_Wd>OM-jscD(byn~R_X~_^|8h<c)_J5g}*#w
z?3@+0R1-mh3dvYa%BXUxE`;3?abG>_<QLyEv96x?*YlkDN%N&zofaD&DhdVnp=`-R
zL;LO`xxf2Bh?B|~<Yw*;GO@htEhImB)a!2z?AQSh(o^WpA>~{m`03P12S?yWjEfWa
z*YS)>%t0*W3@kp4ZCfAa>ZIo<9d@zY6$hu&AnPgyVDkdjpix2^f&aoW$~Am-=8nT%
zD}tD)#!Hkt%29@%#J*n?<h7Jct&UoG021V+qxnzTEhu~2z0F)wWvmQpPP(jt#Y{JK
z({RItcUNJ_FEE$m=%W=mF5JCCCtS)0h*uxF{y+6<j12!lG@<+NdbK6>vFJk<=&lu&
zS1@%#xqQ@ISzZ;Vl|OApvn_Zfc#I|NhTlK%<HeOzP3uv3U@98PPl9sDZ=x;O-q7~1
z?VR41-I+}onYX3b(6~EbLVx4ElL|vn!AcTq)nRb|h5UP#_b<8-O^K7YXvg02c#jA{
ztK#a)I}MW$K{HYG{7L%Q;#=eH@yYL@cpjps;b}U+qS=lab+vWRnd_U&m`)4LjjcY|
zn~B+Rj!(fCs{QqKnrYnNcKA47<--@&io^MQZ4SW}B+cT02JxXo4Ak_5+;r?{thzCX
z82}ICwtqYsm}Yw#bb*SwCoK~wy9<$0o%gNdIbNnenwusaQxU6olm1PUt-7CBs}K^r
z3CxqSM37l9RiBIX5w@!!CAyA8mIYoIE1r&Oz7YX73*=kEpU0up%A00jBdjbyI1M5h
zdH^TV5XgM=H{?83)p4X<!wg&?^4-sot3DSaPC=7icSa4H5+AucXs;ENl<D^BYMsf&
zTbJVW>#&gEV+Vljm>UrTxPN^0Z$ik;6T)6HgQHdM_m($h>NX8rSa1y#Q5s2;Qgm^5
zS?`xNGUK7<GcD^tAsVyE$;$v66>M;s04rS^-e_<-E~%oMPH}gCWNQYW!5zVbFzi&u
z7GwFsd|Ho#Mh5J}rOM?+;kgh+YzJDHCaA0B{kLJ5dTICT?ZG*vdJj(Gzx97^%5VUp
zlyWYN<n;)N(`G6u^eh9ZI%`PMnW-=O7l`M7v20f(AbDMZ(bbQikLXLSdjO}_cbrsn
zie;Cj1FVnNt})^a8UheVTq38c*$8ov?GkorJ0~K+q(Oj{ua|zBTHpoP85Ze2`j625
z>2M=Sp^!*UlM#=e%jD!i6xL?WiVGS1Ye3#v9)q?sQy}dgOLVE@p@e9`u-X_F(c<r&
zyIU{hjQ~m`s<EO994%LIQxVbb1jg4c2&l=E{IX#vu<Ol4Wte?5`b7Q7vslVyA9c+i
zd5^>cWpCq9wG28B#4K5&n(s-A!XzLLZFYXkH?O%G66z=qQ^=?_CTmYdf^J-<>yoT&
zcf7N?<Swp#t{B#bvSU(upFxG~h(>#ikyZHL{wbfSMa(|Nq_=^EImW0w)`;dRY2=Ol
zz1T&!(78<Mly+a*q<V3g4e7R0jk9Dd+BMmHyKX0$bHbjWCVIo|#UxXlWqX9)0Ec?C
z_M1fFj4I%h1r@k$_9xK^E4v}*-xTxwGFsD&H|aG@O!M0z=?s(JhkaMW5H8-HX3Kk>
zdu=O@6f~f*1Q?@){j|MwyA-#W4;7L_kzBJpj>0H?b^u13p-T57A=Mey$Q9TH+&Uh+
z=f*NFSJ1Vs$`iiOBDVHI0@f1PbAhlBX1tr3%d}|EIgj@^eJ9`IjnH@l$=X69&%5+?
ztn{ZB17Z)|xHFkXEsYFI827zvIIh&3_0gXxs&TF?;Nr!4Ay17sWyAs4YJtazFtn8*
z^)5U04`4>R_*6eiHe{{(P=Ra%{d-QDJK(>sRfkCmL~c10NoSuGbW1D+NSBX~p`B!j
z&;ik$xt~K_o`7>$ag@}MD?sw%e7ubLS}~@~M;i@LD%RWt{;}FRbS)BxR~$tIJ39Kq
zhStysG$otE#_owg+5`#MT($&%Hgo*B6BrzoX3WS%rwZ4b`pk&H7&x!kJ<$@2`W&hb
zXF_OwRW}@ev|jyri*YuC_bT`*`&WRhlLnLmNOqDh%$37<&2(DQVfyEhZ`BXM21ryG
zG!kL@NkVd9-4<%oD6*!gDEmp!o-}$<<PxC1sP}_PA;e|7SJGD8rhQj3U`=VsJiacg
z(#maCQodpd5X_v6C#u4V2%x*3G%FJVp8Tt&on{g+b7N_mXvL&LnTRG2jqq9a?vTbL
z;E@NwY>`K_$e@&oHeQ`r4mVajJtkY|<p|$fgrrRK|LF3O$xeYPCFpKBIx#Lz8CdVL
z0LIr*LRrn7k7pKJx3|BgXE<X4kc0bA5<yzP(1hAn3iY<Dku6#g;sFB;D+0bNLn$q^
z5GYRNuk2{ubx32P{C(#aBs=G3^wKD;OMQ?$kw+OmbQj2~BWge6Xg}j@PrJ}GQX*OU
z)|WUVL_UnGDEP#o^RFljAe#GZtoW`bOr{`-+x_9x7t5eTeuoJ&3TU;{DolE5YZr~E
z*U~B!_)csWWt;$YibJ*Y(ke`_>$|DcwL+rTl6}egX^Q3fqr(schB}#Jjz!ewzkvZi
zT`C>a>c0{7*Z`pU5rdLEnm*53Sw|dE)t^HeL7%0F|7k;^i#$ZGUk*d5D~=>}TdUU>
z?jJX6+t&FAOkm`>ZcOpejw7dDKls_>R06XasX%P_y%_&5=FTMV7|FmJYC?vz?xj_e
zyd$L@Iq!;J$f`};c}Ju=yPfFiPA8x82k^$L!s$O&*Z*z1rf2ytsghUfoH4&EiQkLq
zN3i6**i3u~bWqC?Slca-jlb0#L}l`t-#&H%;fn$8s_{os4ndZ*Len32ahhshrG`&O
zwr>AC+btB-_TVeaw|i82X^DLN5Bd7X0m0{GyM8LaImF9CbD;hSi&wqXeQw;bW^XgG
zWdm>l+PXJ@_J2CK+sN>aF#U}0z;6dJvqZ{g>J8k8V#p}tYMhVXOIk?f{u!NldVKl7
z1xs6u?N@lk=pOH}z6tSxE0RC0nuV<J@y6+gD&R9Mp(p`<J6~xz|Ja*}Hysea4}S+T
zdb$u%iu9CHeu}T62!r=f1{o`6L5V}-=q#@?=aC4jENOcfSf;lG0tt$yQr91d7a8tC
zFy6GK|CBZu6Iv6@s)}kB`C8cOEA2D|$)h13lju@cotnh=5=?U^)Jg*nOmXIgfTBKd
zM~Yf2N)_q$PHU!)`y&qo3&q?t32a(wX2s~HRa3VBSm{%0r%3H1`0Mqc5t?Z`-R!Ng
zho^Q6iWf!?<m)uDvR@q5v7$SP!?2)~#b*@n&wes+gGa;ABB$+is)3i`)XV%UZ@i}4
z1VnNq0$!#<2hgxK02$YWwFvh52`ofQ^_~~`dFp?TblfbSD{pYj;;lx;hWsfR+z%Pm
zg}Mx9Nw#H-WnLDT96$lS!KfC(^T(Vi;mEmM9I7#9zb<QP%2su~p-)cbO2lvj>nyow
zb8!}8idnC5>(-$H%j+|W9qlax-@h|wTU59=R(OEwmN^p-W1GS|;-rygw);@d_ASNk
z>U@$gdQKFvYA)7F(2~+1Xs((1$ceyCY-WC?h;70s`=nG1y*gGa+B&$vBfDT95x9Hg
zj%(oVlS_mSux#X(W8n6kOP~w{>B6oMBbq6MC8IsK+TK5om=V2iPE?((Ami|s^dN@$
zJVGoMkr1>+oKN$?D{yu$m#!E>+=!tZmJj#L?j5*FmNf9}K6161lz23hCtCJFCG~(;
zCq5CsE7EE^eR~raj9|qh-{ZREK>gA>wgQIprbW-?+Ch_biQ@9RwIO-6>yY85%m{o1
zq6RckK+$o+t`OVPS*-i&l_pp3jf-uf1M;w|s6k3vkX(~uyZYOs%$f$-m@${Fc%$pP
zOAoF98GJxa*%so-U{Redddy;06>+VvlHHLAj3<@E(SCJxDac`dlN=?|;B#9)<R~u@
z%4$<Pi<6;UdZ)O!;vbm;a}|W8F6JfP#w{e`P`{zZL`^ATjCdeEsbK^XYGSCE1ZfT`
zGArqoyP^r3no$I=uxDE#`*aH(Y8O`B9(brrtWyLB>COb;mfXT1SyNWCo}RUR8}U%F
zQO6Voi>Stke)fum24j?)#fAK!Ferrv8v1qx1Ji5haZwl2_y-zmP5gxSR7EaV8dw=@
z;OdB1M)I+8!BjMN1s>Z*{qq8e$2j5$o+oe3>>!;|nNjve$G^y%zk4oEU{;P?$7?nJ
zkdbjuhTab*GyS}zE&5^{`^7P3*S2#SmHk_IBR&>jw-0ROtz~YW$Z4Lp!wU{ByLd`m
zp~27Oi7m%316SBEJE>Lv?J-yH`H#{j8o<?%TLcNM&RKvAp}AM>DrOocp}`hu;$zfQ
zQEjC+4><_-yE999fpL{O(oQ#uFH_s93Oc`lHf>2`I>2@O^4Z>L2>xm^h0^aygumVc
zkA%QCXy@*(a&ps`4a%Yk3-c*qN|Dkb#BN)hYxEI{K38M=i}3mx{sZO;-lyJMBE=<6
z`RX;`b0c>R)Myx5q~3?JA4<De;Mf^MzK53nR>_pSeU8nE*9SLy%o25C2PMUF?7ko)
zzXHo`ivthyAUKWRw~+RNq~YAwqzvjW_rrm*seT|fZ+mAzsj0ATaG9eOIuZVw-`Y?0
zPLX1Lc&oPCF1WIs`Qt$6*29g?hB$u!cG=fUkm3Km69A%-iwgNHc=!8P?DzREh$HJ%
zdILiF%lT*gn!`;i%(_19BjLm$t)2&8L55V`8)wX_WFnSYND7?&sm^}PAbDnZQ3d)L
z#i_}XV~s$|nWFM?Zw0v|aQU7@xoScWs5x;qB*d&9_=dx$q`XwH2i!(UGPxuaYSa9*
z6YkwZRZhM5uVCt?Jjf|*f<Hp8>|6>Lu+6^}O0|8iaAsewX0Q9(^-G*)E3wE0uc{m^
zDZ00Ys<@wLqNR5_H9p{!iRvH!Q{DP4QU7a|&cgQJO@&*(Q{g7t|1%X*ix;Bi%JC|>
zto(7v$h8Szhl$U$PyPJSE1iP6TG#rE2ev{W5rZto9ar!4(zURKoAEVr{H9y?KmqT%
z;N^MYYh(HiJQWqq0hgN@bbTQ5@)EH;eAjdW)zq9!JJSQ?v+z8|zl+mL{bz$TCiea2
z+^YxVMgF7bp7obhA1PBs^KyIUB3xY6LJQP>x4GaU3Eu0~XZ`c`>QHdO)EG|lC8=}2
z^D%PV6P1a~t{9s4@iIcP#S1kKx&RPMaMF>Z(_xd*Lu5%S`H*_g_Y3puLZa3w*D5gX
zFh2<;aR!+h9-eM1wQN_q47X87nn8y1tn-Sk(&e437aQ#X?W+p<{V}o)E2bpGsN{po
zBF%(v1W}Ua8KNFMz^aQoU?@}2X^*)+{7cw`gJ_*LIhoRNyfq*?dh#t^hia1Q!VPqw
z$n9iC7}S^cF#3vfU}4~i`p&2-Ut~<-18ULpR>CJ=_Dq<glp@Uz$caaE7R!cSbjT9M
zK66Lk5Zd^(W?`*8>+kSlLYZg9JIPjsZej?Nht;|Ts1vEi#~-TJc9%j-wd^4=RA%${
zM}b%C>y6U*B4hfNEx*^m(H(I(OEPXx?TZtcI(AKHSLPP&9X3_IUtZjUK-Uwdj{?Bj
zd{QQE5V7pHkm?-flgG^vyP2fbF%n=D=Cr;O%?^$b0U}3Sp?qJEJh+UEoMsiP>|L*4
z`LE1e{C+OGzLf2Pc-5y!ex7*>hov6R(4i2NqLaU$RVagRoBe0pK$7EFgSni5(cA@{
zY1V4bqO;UT)H7g^C5YVh#*GZ}sL5;<w~F_IQAi4Jkr>nZY!kL^Ci2&s*<zh^eNVyl
zNoRD0le9-(hvlfUM!G!}ORhNvNFyl#VNozc%2OWFC&P`&m>11$?_xU&4fG_p(+DPr
z3D=Fx#n&2YE;{VoYV!*cRzWN}or6F{Cm3*rmYMpgs!gj7wbseQnEetKMJaJyDwT@d
zb1^uE2YI*+SYxLq3708Gt#hjY^lGb*Ka;f!Nm|Sso~ZQeWCaICZB}vVRDL8%s3g|Z
zENYAA@Y!zP(rFgU$|_TELP(`;Gfhvy#L{?Qm7BGhvvjPU!fkzgWn|TDVWNYs+)4(D
zzKmPIN)w7|kKM#|PSI#3aiUIlv^I)Gr)qr8wUDWJg-^T3S7pSX_AHK0>Rb^RX!Ru<
z+exVK&5e(&(tZe8_q!{alz{Zx$%ai&3_?iRILi@7{8XdcE{1rfVs4NE_AT8<edWHj
z#YZKf?BTVs?(F%kS=_~=DYYT9N3hTE&gy>LgO<^Qic-}=HeRjYrX?gx@-*9V#VJ7G
zVXw|$`m)*`82SH_ZY35r>o(bXAo?WD%f$VuanHmMhSe&gno=y%l(_laX-#UV1A<j$
z+cq{OS5C2#oSGZ3vFBC`syhMb3k|_z5T{{6dG6+91-rKWL?PKe8j(>n+*6(EFCdRn
zncXkH9KEw+Rva*NE#jlSGu0Pj8=pCiIJr72!5gsc(ci~bx03m(!O>e$8LA}vur1j~
zg~^>^Tg%1B8`WhnlN?d5oIVn7B*W|E*AsE7=t}Erf1F54N)a{k<O6VA477%mE9X(E
z)toX_869v^5b0=KwJW@y1G*FP&lStUV_`K+aC}?5&2S$6s>7O5f*Av)jU$Jcr*6r*
zD1piH7^peU)~DqzOJ-irh)^q;krXEZm6+eid!SscD-?dhiD)KKvQ{QxQ7c<t10-b#
z7`Y9gBxQ&ld3|Pr2@SiU?9IW27bZU8AXDRK*l(r!HEr_Y#Vn8OLZEFNswG8on6bwt
zG;isMVzA6cRL<h)K_sH!{>8ATm`{(1Cfff#7Etwbl*v}d6^0m8OY-<JAPJx%50-JA
zG;Gb@H?-<z@P2q-w+)8|#hEhY9}kag_U%M0kcz_{GF2cafhFhXOr<I6GlDh&BE@0c
zS}|mkwvr#}A{dZDU|je1q^OPlW=2H9!nFG5C+7=A67VX45j0INKApW{f+9+bgZ`qc
z94EhQkVJgwq$-5>FH|Wo@zEcxi4ZcKYm|M4tTd7t(N0Jh;BlOzhoC@oa6_jBLG=CW
z8j0SIjWbovZ*Lxou|UKQ&X7%jt3*fK_F4rYv)PiUVSlexFjIE=31Aa-1^H=J&fd#F
z@Ui=#lL=F3o9Ff{(4<1^Y517^1FF|)@*fEIgE6744KOL_le8+=GW^aBZpIHliP|yt
zf2?x<JMxv0j-Kwn+EZU?h}jtaUTu7+bpF$gLNy19<i9lc@ZID%)rcL)jl0G!4w(Fn
zA;J|WXzHr!#7l|vvsXG3Lw*X|FOQ#i5u*9G!|m1faW>SuFH}2zaWv5)v<DRSlQbx7
zf>x7gD@?89P37YiE6?f4y$4mu1geAc56rVA#iIc5S#~vl@n}0h{0sN2t%vM|_p9c$
zdcJ{a_?JfN@%qTF(2Qn-gXwftSMm}|Ms``Td*4Fq)1A-=O`b1-yRh+j^)|6(<MNQ7
z8OJZt<@sfog5xXK58TCz%$Iib(nGs*{maL^&|Idv+PT?DDLbMCwvYv{aGYKTP$ovO
zP|b|!FBAb$eoEyn7{c#_#Fq!J=^v^P&8gX96tOx}OYMziQ+hKo(z7+?1NCtayN~6(
z1&lPg-c3kEW%on#+`IJEC|shuxOV5ue_cfpI@~zFkCzFG_Z`Ihnec?89(5@)@3IYj
z5_BR)t;s#jl$D-AK6=<7cZZ*@P{481xG!ne*kq7}COBtKZY5NPRlEp|b+3L_(1y5k
zkkb|mwgdDmt>`njDWnmdBi+2G&TfubNPFjt=f5cSwu}?JP;*k{V!u^1tZyVCR@9Jm
z<HHO%nP^`p9J4#LNfR%f(Zt9)`Xa9JKlB(XQEnwIWhCLr@LV{(QZhFPih9_clmwhD
z30e%|ARsqm&rWuumIxTh8lppv(_=Fn+XW3tBYbmZ0OhN#kh4m_y9+OW5<{vHOMCUE
zY_Bpnwarsc$nBQz2@uiD%6RIKFI&qYe|7juL;0+O6e1rA$p6P56kne~4oiV-b@Xtt
zL`I$%c3+sE&Op!q#`cm-Ta&|@FN4r;uHC_H(@5a|=Iho344_<f?3D3fjuDt*9n|5O
z0@!5wS_N2aT)dF!j5K;7MvB*S%2X;NHbhcGQNb_;xnez!o(;(DWRWFDHa%a+lazvA
z6(>vrIT_(a<E?o-m_X=%Q!Q|Ug0lcg%%qAMV|3zFoRDDR(x|S|2zn<&!C5oDfbp<3
zE+F>}lCWtmL7D&VY&{p|K$cd@xRvcd*f5$}r3D@&GVBgD^P*c8WeX{j)_fYOm{0{{
zFwBKwrfV5V9P=0VBBfAev`}Inc{?Tra8JpTRl80M<VEQ&ABP8)!z|F{hJV_d%pSf8
zT)qpf4cW_xmEx($LDx$78c0?^f)?=9rSBM4*CPTKLM$&3)5eVL@w~oZSEayOWjW+J
zo*xN5DZ8uz(}X#BF)4cO>7gK|!oc~Vnn|OSI{3Y{m?ZBRtTczfs>eM1Z)$pZgL24&
zeU8#)^n%-<K~4A!_KfdcJQO2^Nwo>Di&6(la`$}`2P_KpG20_Gh7x0ob>$_;<q#p^
z%)|a`=54JEH6TXq^y!hJnEH;mHa$b{!}Pv5*s!m6Dhoaclde0IC|1lj0#6#6xYt>y
ziSMP5Nx6`%8|<2z8f*w#H5s>Sw_gNC<$R0#>k-x)Leh7bZZ2Z<DRp#g=2+06E6yiV
z>A_xFS81bL(^Hwv-V>2b1vhSIWD0=N>2iIQ>YufH5^2k>?J*8b&Q?6dMMhmEoLsp4
zk}vv-uQfhP1=9q8+b4KfQ;MAecHv^e^pJ4&BD4G)ll5m{MAeaSHsQR1@N_PSH~R2d
zbg(I|N|tlB#}%9?->uXPgX8c?dRAq9MNr19zfeA-2*Lr&m9uytL3YobTcAhKeWG&D
zHc;yOIVqOWbNm^h<*)xFtWn2zan~(wdI=%)DxqlH5MBO#U3|3s^ZF8kz<$12L=kYx
zs|we9<I{n;FRRk=F1mp=Nknjim%|L15SAMOdFS|rqH10Lp;sObu6|ls#01Ip$?xc6
z=i5~f$<DRV$3Yxg7Jv9!XH~8~M@~B%7k*DJon3uI8F?suWbS&u*+dcfGHCl~-t;0-
zta6w(NP%yxdyU57=|wFU_q`W>8nsVw2IR(v7KTVVl@oSW|9EPyMjz&cp5iN!Rog4!
z(9mLRalAMtDI7v>Evx@!qwdmGp=hWt0|SXR0->AnqRI)Kq6Ti9)ciBubisadc+o!H
zB&|WM)VSs|N5CgiXa6x6D3;E<-prUG%<vLOZ<y@Aw(w~u0c6`cDKaRjyy&LBe3;TD
zBc{4uXGeFSDD>0yMBW~T?2pzj`u!4LEt+r0@saV)V(Od_yW%S9nP1k~IHS)7Q4wPd
zhd6~PPzxzaG-Ky!4#kgnG_z^mR`|jB&PMix<>NkRd?%#+^JZ{^hyiS77X$?MOx+(5
z1Y(J7?7r*}!(+2c!V%aekn<9n<kOr4QmtiH*^GxTtqVGew;k`9pJQ8B$zE&he(B8?
zSUpbJBlvFzjE=ax&Ks(uxO3f(@QxgUyzXPY-xZIy<MT-9{wvWfG)EhqJ1i~JsG7gM
zM2u0k5@Ine6=9q+3>#oHHTc}JlMqC{tA}Rk91g~Takk8-PZTP|zIl<FJ|>ueSbnun
za&5SHaz97=hqAeth;@2g*)|m8Kc_`$<N$xvxyhM|)1ofh7PL3C@92iVBt?bz(+!p)
z#3P@>hl}ND^6jv?_KKA{qHqXX$cl!K+TwU>heO5L75-v9%F6;u%c_iE%K&$IxImc`
zvR?R+!$IXf&Hk7Ljl40P`Zr{U#47&^vh5*<VTLLe84ht<5!rJVvv^5W<G3o#o{`+f
zf>2~D<RT4D9l1j&>M7YccxJBkxJInF4ME79W=sCK=4iEeOl9A%c1m>bI}xOn@sw|N
z;yA_A`5N8~K!R-8t0r3_q6oST1zbHUn(9Td5-3nr?_fKn>l8Gt{IN-q$I%mZ>k@KT
zbtVqhs7|uWi7o=WtDl=Jn6K8C9F?}raV<YGo>o01^3zDMw}3ifsS(K0s$|2f#$Y{9
zdJ-~qE=1}5f?*Qb9+CIRWt?a)iN3njYC+|I7#pd>Dv6L0>XR9fcv^)d`Ce>q23)d;
z*c}5_FFPh^9T3HS`SOtilEDBenFP}&OeY^+S;+<v3}wt4L$b^&=ml`NX-VEFi~Tp#
zrC{ws4!5h~(37`$T?f*1%5@{%mUT_-)!XT)r<=X<cUvjF<<;xR`3muP_Lv6MW`|%|
zeF+P4=Z1n<BacYmD9d*4dK=XPNT{|yLw&HY@vwuecr(Ce`PE&(%bL(Db->={M-=xv
zj_1=y6nFoTzs%L(J=UX#-D_3qCGSkYRpp%yBc;?v3fJC7TEQ`ls5%b2qi@%4XG-!U
zVEk2`VsLh5(c;~gA8OyRRleQHf)S*mrFC#g>`4Pnhn*<sB%%&Z3+4v36PjqKyZ?(i
z>`zSb&S2WSn9)@3nToL0HCmCTL#*8t{)--jA_u^n<>XA#s5cB}lX_K<A|)nfdSc@4
zODYmm53I7N>6U+a1fa;AV^{<inpp?ch}-c^vZt}jxAQ}qZT8HiB<lr3By+yZ(y<eA
zIkdVEdNtSjY*_gf)R|f+AzHS@CaH#^Go7@;MTwa1mXw-Y(QOyy8Z=t99c))OA;0H2
zYEH3@9jTAdS1^~mIkj89{e-bc5K+7FsEqbTR5HWKG7Q0fdk3YU=HfbNOm)gouGnJh
zwR@Q015KUy1SL$PKLo-!<r4H(y8a&5$QbbpSW?M|b04)1I<2G3{oK~2HLur4byWPX
z&V*`bY211nn@*46FIBuO-(PvkDMXzSmM3X9=BNE(fq=3hUSv%9nW<2k==6gXY`2`H
z^w^W$8ckCBRjvQj;P)g09UOt3d0;7A!pOE6QvJ0tf&{D7ZB_3;O#%Fy2-t<P#3xKp
z%3(1_J|X>M^Y@@16?gRn3N0SiD>9TXZXU_RrXwaFhGYNqm+`24o?rQ_NM3EI1XG{R
z7#nT<T-UvO!F{#DT$8wEn{fFKv&PLbL;`NW18Nva0Ak8KzmC9qNwK9mpi}g3qvff_
zQ52Kar>wyacl<YDCC209Uw_LpJ8MmVHSWO@`EA!Z4m+1rW&I*R5$Bb0`@&Z2^`Rrb
zgYh@i6WPkpf0_rFnf{mR#lS%S->F_*SYp<Mthd)Hk6@E~@|%D7d=KluMyZ2s_@w!a
zr5#2;e_{#L3>zxYtghHalUIk<RpgTgV)$TpvZrxnbhe?PwSg--ad^7nAaO-KRS?m#
z{{CM;j^`)#H-~81-|p#GK=P=*l#v;LpY?+VIRJd*YBDF!=Ul|ku<VZldoX{Wz8(s0
z2@g+oCyf4*8H@&QcL(%3i<sDLf&N^0<+XzyTpunmM1k<r`H{yEACE8B_nR;%5He&$
z%<s3i#Q!@4aGRnDMw-7q-Q0ei$wW6JB(#=vs<D5*)7K`s%v&Uus+?suA$3uNjR*I&
zkx9BMK0@jgHIk8c?m;~Pm4Z~ICMNNJGa*(}inhuDgef;LQq;-Tm3r}VtLv1iWi68|
zc6Wj_jrNHAJ;N}jglM4c3?OC~&2F?_a47_a__coD^9b0s(#(i71xNEDhKJMGzy#R=
zh`~KkSD4h#myHk+l1&B4>MTT%Zw-~3Byt#JRd)JlFgMACYEeoL3C0IN$0pg-$ae$4
z1AI!p)fD=HxdxFXW||QNncJgY)Evx|Ve3FcK~fxT+^vlnZd}6NFIYi?28QQY>i>zV
zz2j^0YXaPqe7h&Uv5qt8vR#Ir2Gye@wi4Pd$DI4pwoxup<~ZUNR_zmPMOcNZq6z_Z
z<5aT>8ljq^8&grNH6-Z-sDJ?U?~zY0cKn@rZ00|4aF^Bd3<RaW^k95vg&is-fmRYT
zmIzduj5Gh>rdOzoe`hxun;4x`6NGr{(=S$7nfpCJGcquB8%-u8y+q}!L^K2HZFMu(
z6bq$gXLCG@@nIBLj5&1p6DCS$W(mXYYO2ijKDA-P0G!lgc}}pkG;*Z8V`I?^__yWF
zs4)YkW$1k#!Yj^&hc2Qkp?Xs{5`ou-3+yB~7{P0EM4K_69Yas~)*WwW*s-Z~NG3Bp
z!a9<Z+O=D!(Hz-cYc>sQMn9%yfMPoUMmpXJ@yTbEfk=^GBF9c_X`YY0v;3(WV!$lY
zHeL4J?5`Q(fw?*$Hs7v3zUHJ`C;KIrp!Pr!0wWaa)E&``32`FuZ!)&gMfQ|UuQbvZ
zeTznw1kwq;#%9Eh&$Z53zCgpRArpXyHLQ8F7@@ylDJ`gDsMwtL;QkF9rsxW(x6@S^
zzAC#UA0BkL(l@7&TwO59e2T#4Y0JoOp&BtPdf|ubcN;1rLfI}KQbU|(av)pFj%csX
zn^KQvQ(2Mn!88$tb4Z(pa~*!dWR`~_oYE|a2wD*(t;~{mw|ChSiVM8|-Zn$6QiZy+
zpipP-w(sHmIvxhqn{g}(n01~(9HjGB29_!l_3+_YXN>O+3w4Tca4{lMOPaQLJ$`^K
zdk*%^SLjLePXQ9g()-Fh0Rt3w)5E#Yd9O`ZdM#qSlSlz4k(>`|HkiqzwAi0)nhu-f
z903(ZpT*v2`rDci=oPebI)~U>X9BbTS_M>S6U(zUC8(sA0@)LN<(MISok_G$uM==s
z9l#c>5JL)c_V-xSeS*)!b9!o!0FK&>E&uRBp(FHd{tl3jeH|=GrQTO(MAL1$j8d|7
zwtxg9C*fe`T!ID?Wj#o7*`N|ljC;%eZxgj#TR9-i)ok;X1EJ*cWPE2Jp;`kk&8+64
z44lnVafyXZ&{FWkdoNS{@za?w-w<2Dp$O+J#1;jpB}ioUnou^)CBrJi6p^F@vzk4=
zux50elLqY@I=9+Ood}o7+&={bf9GQ}?db8-VSeHj`J8liDXGYs0t)d^9US#_Sm$q6
zhefFqiLcTa&DePG8oPd}0Sn0nObgNxZ?UySgH88^3a?&K`Oo4Ny)WTz?-8MzzVYkH
zY_MGHgQZ~duGzmKaZm7xa@w4ciK|FLGkG`cD1`#D(ks?~{jYhGVfonATHL_LrwDM(
z59ng+WI}j&)XlLpinvKJqJ*EFXBz%Q-;s!P4TvouSnn#X&^(V8VLSq%4)BP)vPsu3
zGRfk%7l$?L^;)8IF-k}^f>PG5Sp^l*s=}MfPr6So1@PpMz}H)d=XTiCCpN?dM6qaF
zu(ko6rB4_Sv{aMGVR?jgZ0g!HA1Cn?1HniQoUsb7igIIJxUX^9DIE5fv~AB*r*JdB
zE~skty1ZD`y?X^MHi)iP+qT<|wRX+In<ad%-D#OUl4<g;_!>tfFSVFuq`h5UX&07^
zPDoWagp`*RN336td#muc&3=su1o)i2rxi~v$8<77m&A2gx70X4gpA#)7?pd;y`h%g
zak@BP38!<Uqg$A0D?2}DH_tbgodQoAPc^ji@c!C)-i}AStL}1g66Y>Um-On@9RSS8
zbs)#&I@VZpEockJJp5`CyRDBv*HT?ig^}-13HXNMoRn+oEYWa(bA>ZZQsI1or3CCq
zZK`Bw3uK-T$j0vG4ch8)K|)hkOi0vGohb)pY4%WTB{z*e{eonbuSqT#UHav9+XZfE
z_?jIRG7Dub6l5gmn;HG4s=E9=!_!8T7{`9He#&g+@ai=+vRus#2tGI>)9Px`tJA1o
z|4^%K-D!2}G`HHZeT%1-(L2iNm`@ysro96&G7gW9G_UC4M9)t;_e9REz@~pTZ(wA)
zW*(%HA#HKZvBeZb>;Xl^<7Gnk80E{c-FZ~;yp>c$j)&jTHs?=695IbF$8q1we1|ra
z7mkg8c+Tss7FDFcdL;;_c#FZJOYr#^(|<YnU5mr&F)F%u<Xj!j+fndo<@U-#)EPkJ
zQpP6%@_i(0Wn_=nKs;Ovg!CJe|M6_5rKh;y^L#s*SuG|Xp}#(sIdb^}{gM9fEaOv*
zdtauayY5;gXP?&0+(B;y!ztW$A(}NZKhddpG)X;MaIC1fx2SMspK`*j|Lr;fYdB2f
z-xMBNrnRds6)aoRcx%;eA(s6$$6nz?#(GHWT+zb!eGlQ44~X$UbxkU-OWNe+AAo%Z
zQK|pX_pmblf4IepZnj3WGWzBUj#jkNc=U91zuBku4vu*2Y>c#$M%E^drg#i=bpM@3
zx}?5gjU|HU^`i1<;OzJKCEo=O&2~p&7yHNm=uZbFIG|9p8Ob3uA1w9BjYB#4Jhnn&
zX;|kUT+5ZmMaQiQY1S+;Q@!|zehs}wl*2cK#wa|0)MXQrcw78iO6pD-Npu9}2UlwO
z{hCe$_Chqg+qSyk6l@lP<lVl#mCscYq6tM!8*oTY#JlD`$cncHWb9JK#@l{O0Z9;9
zd;#ImeswKFL)oJUa4Ri+C&KRA!FP9kDA$zGECYQr1Wr<|=YbYWqUHEn!AUUGjE0~{
z^uUGo%gi9g<fjM(#Y0-%h6bY(;1ky=)IO>c7$1Nkt<g$8hJGQHW$e;H3)?V41H6FR
zWSE;w@I<4kQAo>vnjrdZL2oF7PRjOPsqG_v1pU2Luu#nWehZAduxKO0%Xmv)Xu1z%
z*Ar--`ATlZhe1DaU<}3nY%$;|K74c}$=wz@I&`+lTN40j-Em(E?Fk6vk8t>PcvX*7
zVt&Ibi1T1{$Ri<L171GiWh3eUdu-Rgyh$x2p{BQJr9}BlK?Jhfe8G3QY~_Ycq1lX)
zl_>P7{Oq!s#)b@J(Zp((DO4AcyFwJ}0%M|;+-kRSG#0DuoNySnCi9uZfw4u1X8!bJ
zwIdK?=~Nymvh-ub$rqw{2~bIVM$JG}qw3lNI-}Ihh~(jwyk78$lEjdK>Y~JugG!3|
zyrm$oJ)Kj4+lG+m{qE1!bh-9!SRj_iHge>%C7m#2yS%jBAJSdjf>!K`FB(rdKi_ao
zSw#4G0eADgPfz9?UG8u90Me^O+?wBPB};MySuSir?|Eq~5fN{?s$utjvL^w}Rj5t1
z7b8AC=02UKU{=JF?t7JG8l!9ng3RF966^<D^QFo6e-MWI$6Q_C<g;kg_k@<T%O6%u
z&b6M(x)koWavmW2hfy3$H_EM3G*5*>-8HB)mh@GrSG)pJysLy0`d+3=Bv29K@X0bM
zJ$rgncIz$HYR=o$Y!a9Ebu^QghtgBDTS-2Rhip_XgKe?;Z9;4D_k+nYbgZh#<b&S3
z)gsG<<YaZqau%_kfN*vrDbblMX{Dg$33m)^<p%&us-Vf9(5oWXwYp@{Dl(E;KaJ#n
z^0cj5aGMi~mC?>yT5MitSaH{#yFXLpduoak`W$6rj2h4@<Qz0)GWtx*z5zO1GCd33
zX@t|_*0G-aNz&#jv$eY39A~?2P#dz{-TxBl5d;d>J5oi-^O*w0IkIzU&1`IFs5yQ;
zAK3pN#=apu({S6eVkZ^bwr$(CZQDl0)*o9H+qP}nX4mbfM}5!WeA9RS?X~yPYD@*U
zYq)fi4Fvmw)St@coJC_E+I6c%Wn1g)7>S7huO*A))I^q4?3&bs+?pqkc|VP<QA1{B
z{)YRW-nnyG!lRpo?i$v4v6iK|4VpGGfZ)iBY*%*W!&!wbeA&v{r=RZByJ1d^L~y{J
zZ9`HFVdoRlRgVfP!)%HdY28=o`(MwMvX%WTQ`N>#@@zZ8ZgJmM3fn+pk~til@j>(J
zW0Ix@-*IEMw8C@Z{+A0*pZc-pdk*}r8+0KT@(gAIy;Lm_8q-8HU92t4uCQGBiSG6M
zv|q?bKH-1ifBhTy1R?F=5N%?nEfXqD?-uWD0+ZqcbpxZLkg!lGKvGCnK}HByWT#}6
zL-pqsubS+53xf6|EDBQf#}l0019ce(h&jZ+^!TPReZrvr$)e0BP~s5ngyIF3XF!R#
zTJwV`+yp|nj0sR1uY)P}`3fMrev3W=5JOA#e)SrHS~SK)B;TKsY#Z%!!Iz0m9h*xt
zvWi$vf+uOQKB!@?$hNg&FGl1If85h;Cy{I_Z$gFUJ`3R+17^v;I3F1|k>5*W={T6Y
z`PTOIiGA20`D33q;hH=RoQ|$`sR*UQy+;~Ej};iW7`?4blvbl<oVth_Kpv8=BF7!6
z`hnISEP~FKfiE0XMuB09OVLeLCpO0LlQ^}NDiQ#Tn055QqxL6LMRv6iFam{|OLAdb
z`!e6@`<D(|O<s8}aovxoN_ogna(thEQ1ZUIv%Q{7owzAyYw6J^2VayCeOL45M9JUy
zB1(;B|L#N@b!xG&Ck6QiDyuPD@Kan`6j&*C0pP4uYt2>;hNhl8STR)>Bz0+P9U0n(
z>py3wOSTpXzWga(YLPEm(XH=CM!ZJA_R`N!9dLtW^vzvx-SBsq?o?Ob1MM}duMQ?U
z_1Tj@UYYpAcpmtGdpV^`;DI?tG5gNq_E`8N@fXwqy5YdLKFYKwU9*Rag~fMG*UiO8
ztChu?S4Br13fANKOZ&4{uixuHr!Rq-8_on+i&o25G8wUy6Ywj?t9vF6&nj4$XOE^K
zZ%UtckV$Vu{lhlIlPfnzS_wDx-)xO6KFQNQ{xPp0h<#QnX^_j~RrswagiJ3wQ9O#R
zFF@i))t3PvEI&23t+zPtfwVsocLvt$rmJk(pftGm<pKsw-pmR9?3`RFd>{Qd%*Y)&
z?#y#Qj9UIoL*`T6BTo2nYvN=(bOUI5-1NFpr3Y1CkM&*i6L=|l&)<(yIUr+%a>x>5
z{#?qvYrH!({zhK^)WSXDyzvp-up1(aVqR%-Bbpz?3!!?`Q6nS;8vs~BQk+|P@~%%m
zuQ$E3BYbIke|axY`rt@$2d+i-90ek;b$9E_qh%KvBDdCS9<+e)D(;|#SDLq;;B4r0
zD8HGOzUMnz2MDES`nHC2u>XJ@IN1J!)0cyt<-cN3|HtWz-FC;PKgZ9&vV$P-TP=|m
zq;2h%c9f98=%x?;UzMXDzi54wu1NzcmUTSr?1B}WX4OkHZQ4z=4j*4m4^LMIRtG3z
z4)nkIV5@j5)bCGf_^%X*g0xd{A`LL1kDz3Qu!2cPUCihJ_KLfA9Xlf?nx1bUQhwlE
zM>WI0_u-)O9q>T!p#G+-hkRc~&n)z<OVDA=5RTdo&%g|tMMwdtJ=p2~Jdz~BR3ZWR
zA8))q?oBM<%M&Y3TkTEWKb<J!wTM9>KNZOQW?Mg+@u%6z<@=+sp@+1*=r2d-@24M@
zFf+PL5?DR&eZtjD1=ZOc{UVwESy8{#oi-1Y{|t>@Hv){@cpl;7{>KV>PU?3&Hen{h
zuQ^o-_;lOA6eLh`dQ_`{mQSs*yo4G27fVRB*E}{Bl2yqQ-J1dKK%=47z<UnAfM5km
z8$xqA*eGLWj(2spGcmrKgW8noOh%i6%Sv_)An^PpvH0_^TH!@04L~zTzbRMQMHGqB
zDzs>ULa)WLzv+!#$ki6gmZ)KX{MNQZAeokuM=Q6_bj5k^de`Lk07>@I14TKyPTnTh
z(voKth37MO`8$4k_tA#6gS(LAZWdU$fzSZ&LG|oF&Mn>?v1vQq>qkf$U{W&9*tD~b
zpb)VVW**;h@*sU+2k|e{K$pTjhuS(+={$HP^MP!yqZb>j^}wsS!{sA~+Boz{X_b=)
zKck`EZ4_)D3xT&E4y9#CnpZGtd4zDxDLd3Gvl^I5bssffOp3IYhqc?%c}xd*GdI*c
zk(+s#Z+6I;elOKAg<+{if(uU)!AMYYF2f&=5#CtUcp?jSRkwZuMsfikt_E`^WhZ!h
z0>LzqTW#C8$t$PQ-{46y9te}rc<h4ag%M1fje;&4wl5C>PZ@?cWY<=yWnIauTn=z!
z?7lPQCA=pN!jV;s;?@YPt%tu%_;DsYJo)A(6T78S&6`?Ck%cr!Uv3|hnR!nb8$;Q&
zfN}b&)}(UMc2HV5HS{)o77mUP2Cj1rI18qhccZvhgYu^X-HMAZ%PsMCN<p_>S;DtO
z<h_!5(4elydPyW)=P7_ay$ltRmhIFSnJ_CUjzo#5oz>={WJeMYKR=vQ)bOQ%k`ysr
zhk>`Z+3kxK=30d(yr4XI)LbGCkUTxEUOVU5#{reei076|px#!@yohUJIN>e;KU~1l
zUTQ2D*-F)3+h&rw45+@QgXSo3VVv<$@t&=dJE{&lbs#r19X3yEp{jL+D7Etql@`GZ
z)2&rNs-`hN3q@8JZ>x~CyIBuep*ofl1gdz1lx%}%${UOr%Osr<yb2N#JfXKk=npjf
z7UW2uR{T07OXZuG_9hN(M6TK!T*!n>%k|^2Xkn389}{g$-hEAjY-7+8QSD*dNp9_`
z<3rx5C;epZ{w-gO+t4RzU&l^mQmDAyUY9&(eyHW=`B}DCHWp_cWX!zC?|4T(uhtLH
zF`FqcGyC-ObQHmTE8=R+mxCB%XWV*l7m7+Z1j`wF;nDaUH_$)z3q_<z7OdaB90UpC
zWDYl)Cpr#G@3~F^(%6RU1~WKbp~zhFvc#gulQpsF4sj~#Tpo+`g|x_wSr+!!OXgdT
zZO!jC65;F1)QDp)chwuUqBT%cVBG7DQSv1n2GX%vEueIu1IcbAyUtse?S;5XFzE}M
zg%|rQN%7~25v;lY27Z&e*2(#@KDO;2cHoMnAN@KoL`du{Y_JWWb(p^8SCZ^!dkh<h
z2zt;B!{-gq4IqnWHqDeMXmFMo9h|@kg?7D4%DqhuRc>^B)=ym2ciQ5QY{m^`yIDnE
z;i)#B+K?xh`=&Fxv#SJ|rpx>8Oo6wlx;04{8M37?xayV#V08OpuwjeeQCcZ`WnYG2
z9GS&%5X<2*5e>rZg~NQNj<zjUsbLVH1lXC-)=nDTt?(|@ZNDfk#Y*)(Tt#1pM4E4N
z5TSrKe)+3<9r`KvZ>$cJ7;Op^3M_AC#N`+*j7ncug#hie$>~Rn(}-Tv6Sx8=LHP|n
zhC(Zm(;{_Zojv?1LB62mcG|zX5~H`WSYP;qgQYu48<IGwz=@6sjFmZgTeu1HN?lt!
zN;>1Y0{kfzEKHH68ZFh5)9;*;BA&!fQ`Me&X#0pRRJ4$kgclVWshSKAhJy*B`U(b)
zmc|yh#>JJ7Ewo<Hd<V)_eF99gWJ@YwzJ=|gPk%EU)ps+`IcOc+>IPH**uqirQZW4C
zkKD(ak7&w~pu7{&zUD2s1agx*7v{eAB+xr(GwGSqYlT-N;oTRxwyqKBLFqfwsNJDx
zs1MN{VB<KzC}|>%>cNPPt&VmkZZwOe;@!WM=*y9ZF4PM3<fK}#Dqq@U%c+oD^2&52
zzo`vA@tyqg@u?chGTt1Ur)`t6A-;z$WB+P#z+fNflI#xYN_Z1)y}`;ioVr?l`^oj7
zB|(9YdNzb$A#zYD#G}aiRlHyFts&seM<*I8P?@*tKH;Cr1j`2Gzb4fQ22%_VgfYF)
z;qUwNys^?l4K(qcQyNqod@fqt`dJmn*UMD6bYI8^@vAiXAv$Pt#3K*noc+iEnAQeF
zKaCv5vyF+aeh%N0<;b3wAfzffIvmcE$hHoi6^Ch%P3q%oH#lUeT{OQuKLNmmqLYPh
z$Xh)g_qIR7f~xk_&Eu<`%SWi5^TOAgjKSKf?odyt#eGnLqy|>VVZNa-&Z!tNn1Ui;
z|00IjWvc-5yO>qL5JiycbxD9d(=udER`^#m!R+s?kvm*I*v9_ROgxyh#r=ZYnzAyX
zYc=@Vo}TYP*B!$pakpBBTxUuPLBiao7+}b$Iw{aCN`42fe*jx%Z65BJ#TDJDZL=Nx
zSKU@=Xuk59mmbuI3p;Z%(sUblIwN>O3e<RT&BbDR9Wtdu`*?Y#r9OE2!knnnQXpW%
z<%(md{!Th;EK9jFbONUm&X&!7J`^jKLex~2CYLxYl{jpbocQs&iLSe5AIdqx#$0-;
zPe3qEm<HA5<GXkY3)>rPKW079SX+wnG1PVdaYGl4I=Kkp)1+E{Tynk-5|{i0w(VR(
zr(U+g3AWAkFn3p<(9VO9y)!<YV5t#Ao3Z$usX-)=q(P`K%-?e#E~iFOZ)d(wSAd56
z_jP3wa&40PYom7Ua!YNkda07A^$uP#zZXvYMbKsRFmCHoc}-ile{U?Ov-C3fdggVK
z*<@vqj*H+MlhLB;8XR%sY0hBMnvX2@5eiIDX{qF?DbKLG3bPJXwXiEQ6yc=ijL;|+
zycOiz$-Mjk%T30SC(0{9*PPb7rt7cbXvoXx0RaF<pgIuWZg`)|elw?Yy(7=G`y_{O
z)|~k?$2vf0dns4C4Q<m8dfvZLFp`DaC2ydjX-H9_-bZ&qwdvW%LvjrM)P4OBK_k7n
zqUeq{@O}K1P0-SId-@Lv3867oK=VKpMBN`X5(;Gqn}Ky6>po#-@~#Ko&_d8aK>vF;
z<|5Sic1;(NLe6B^d8}%OX^M*hZ-sW%q-IO4P>FdqbQHYAcZm<4Du+LL>6Lbw!^MFs
z>a!QoxT;(A+5w4ges2R*iZS(ceX1#3)Iqrx1`@H=>YHx(dd@WJmZQuC`D)$c-vvN{
zy`>wK+;l59<MhjmcMV_t?A-eIikw)XyfQMNBzt28)V^=$-!DM7y%u$O@U%pKa;YZ{
zqzt|NVqAc<`m4_3J5=+ItF*1kOdn*_nmmQ!gJ^0#Un+;4q&pa1O}#juZR*IAuAwlW
z9*ezjQ&ey)MS3=HtR-med4q-h>6;ylKRen6%V!#<Y40l@r;cQ=r`gNv@IT>j_W$iO
zW8z@o{O@qMmUJBP|IMGi#kcsgjnK<r;5`g-m^!o>lpS=L`yBr-`Az|eQjscUVvAW2
z{<>KeNmR+lr${RIn`3O@rDY?og&x6JUQnboi0|w5t7tbcq(`bh&%HD+`Bp5)@6+>r
zL{sW&&p9|QWgg=d^c8Qfr=t@a)i5jwv6$|ruY%_%YGy$6jd{9fz&{0W#7Y?V&P058
z6I$N^>U_QE8jwM_DkJ#TIdTNx|9&P@jG;$Bmse~>>C_zC4Mr7+Heu*Z0>h-aJY+J>
zdTZ`Yd3coe57p1G*F0Kmzw{`gy~)09wzkeow3Nn3X*p?(K~u}jNOggz?2NZc1NkUJ
z(vsF$1(j#<xgSeu7}rtD$@4SFIWYF@s_nT7J`K=0f(?wFR$-F$?n<tON0%l^R;?O8
zFfTa$Euyd&3_Ys|Rk&cb)CIK(&+(x4s`OYz(VsWkn?Jtci4&*WO5|r#NfL#9E~UgY
zF7d%<(ZUloVtDZ)Re4G^Db}`|0YA9s7kI^6qu6L>{cDJ~fsB3F`p@kg`$8Krxgg+d
z_M+6{C@LFOoR;f;G{2A>>&Pkdn6a=@j2P41#O!aiEXTP571tlJu55_vr<m|*O+LUP
z#Xe^eCwt{5SEgr1DHa@aM$3Ti9nZO4Mf-CHmwf8)4iu&i*+=u{hbyDQHbAM7-~Oph
zt!_g~&;$xiLve5CpEj$VnM77<ul8o3LXp$G?<HOvn)Rniu^W@D@y*q45@lV}=}Gq9
z<tkb8<ecS8OQcV%={+6Vzlz<6f7sY3d~3%_Jy_R`8GD!nYPJS#NIp-uIV(j|80ecI
zBME^t{rgn7T%m<pQ<YUo^jx3WMu<NwOol}F^LX|xVVkpp)g(h^v%6;DHlZ5a3;?UX
zqnahcMk-@|WTaXuJ==o|<M%kqta0V0q%C346q!d#Xc-50ImJco+_*T(3UbbaeR`is
zQX<xFN$yAB2L!kwmX1|CZ*GjwlST?j<sR}z_nv4iS*iReS?ALz#6G-9$L5D^f08&n
zMdvmOJs{Gi+_$wvx}1gW8Wt?h<=i|svaxH#MpdgqHv?r}=FIq~lCL&ncXL5n&1K#p
zH!JHXuGgn0e$4{COH76Z{A=@xlwB?zg!^8vkxFx0C*9UO=Knjzb_v^~+kT~Msir{y
zHymiReLRcxgz`k}n8`uaTA%UKEe_+!B^S>eS2y`v?67fGGn+R+xuh(%C&CIx498Kn
zwL08yeu|lAj<V0PYCjz0Gug<7@QDERQiQM+Xq0<}hC(;S7zyKqdSnUD4%RSBuIJDf
zU~cu;s}!;2RK8>p57BzuaIN_`=xI4H!6eLX)qqwbe*s)X=^Xi1R;(^~YnOz*8B*Xf
z@qSd!J;MQ=betl6=pQvOsY7W7K^I5Z`@Kd@y$J5F&fjhP{T>ikf64qxVP&*eO5@+0
z-QOTQLxKy^h`95h)*@QJK&UL%+<@owMfPF;k}-FufZH0)9-^ec?ECGE=O#ZyMxT6h
z6+EqIDFk_Cv`9`#<5S>x{J4T&I&^Bk!)<CSdhaHT`J~5TezNZ5gF(H(DbC5?D7tP(
z2~|GdiQU)5*Z^@o2MG<xeHW)b`C%N_c9PD}AM({X!S1Qf$)C}!5vt55y(X<(FyiwI
zQLBpU_kKWZWMa316^`{f!j0SudSyxl8u;D55gi*NRnrjU6J{;NyGe4#0w|EfwOqZ6
zB9+e>uvQ)KHPd5X{NsvCZ*bX5xLkv>G=9T@<{u*-t4syQHzTvWM^3%RK{hH+;Jr9e
z&@t$NQ*_c+^7fRY3i1(}$Et|2MExyD+{MkT*@SZ2D-vNF4__Gm!zZarymJo~=hxT}
z#+iTc2&g}IJ%o|uC_Txca6iFiI!f%2-Ssmgx<SlL72mRuZ~Q4?V;1IhHcua@V(`f6
ziIj1*%YN4iBi1|o<2m^o2K&mi$?OwnL6%2^!FJ1a@Z-v+{nP#1TD+ZVWS+?sMJ<a^
zf6D`H%{mO0qW9%f<w2sF%VtDedC{OzKL$%8ocJJr9nW!5Zcrv;A0C(y%0WbFQ@P^m
z{&K0R$>|`BX38ewNT*(F8~A}|#gKH)o6ByxI#^b*_TFt1Dt*Rs3l*}~W$i_A4};Jn
zZBNge^?LHkvR*+oDx!-R3QI_u`nl1V%&e`S{)THc^le2En&UM(=8qP0V6K9}ToWNK
zMfvSX+mx<9T;J8)aU<IMLo|wgh?HZh0-nyg2LurRWaW8E-{8?~tC3PdAX`owqLsg=
zNYna;1-Z-W<yQkkIZ$iUASX}XZs!FgLfw#~H8~DOBijawq42`6>m@jULxB;3JNV$r
zly4b*NE&UP5^|&#!TQAT&ps0Hjcc{bsv5!W@K$OKg-e&6=iENZFIkTD$R6ToWi5>)
zHl>6iPglHJL?)LJMJv_#;VvUeRDaRsOsitBi$rRz%9)3Vbr=aax<=iSiD<5Ikk84i
zT&$*Q3nt^e<#OC6zz2La>nB`cqWXqg?|eNC3L_hQ>JoC{&UU?wz}rG-BQp^x_f1cd
zk#~KYy!iz5DjQoP?0XlxAC7&~b?P=&6n7T!!st$PasHeX#j*@pw2|8C>sl&eZL)`$
z6`~DC>wRn7RS`H)iCY6;eTrHG<+-+Kr@FamPtzD;c=paJzxGPDs<V3_nHUqiu5SJi
zG6AQDKy$PDs->K*Eqh}-7k)q~{E8NqRWCxOPBH4R0qEF&r&NE(uk*|Np!v6~h|Jz&
zE&t9vh;=f<t~FLJnfv?2PU+6hH=nX7Oq$TR&czY`ZMDZa!##GlUNQ+r-XmX2V+N4W
zXxwzJOR^B=7BIvG{<GisW4$~OkaiT>e1IeK&}P;YJePV;V~@NLSOYV5kX&lmHws~R
zLVl@vh`M8mggCQ$iRQ>|(t-{2(pzvXlq67GgT#~ko*~8_j|L1mJMt7(&W?G1nq@N>
zb42lHSii5S{*^%P2!<`}sxbY-4LV1z8lym>qLpJU^qOx<8LEG2(f2;Px{I>O--9@2
znk%L-JXpf3Gev;`O?9;?g@S4ySyyLCUiVZbp0wtG_U{RrWoC-uomfFWHc5DUNG}Bs
z_P?EaBi94*3*`edV(48wCFVaU5gYNpMl;?<Tig{gIuPUS-;G>v%q3w}KtjhmF-?L{
zh34X=NjZ23Lxu?FMJHy3_&AiXJ^H4YhI}M%-`p}%@fbn?q^rsmUtB0NGqPThn&TUZ
zSzmlv#wsd8AUr=mM;H`l`3ey|3?niq2u?9kK=2G-|L(<#Jfn;Ym-<P@WCf%nG&D}^
zGa7rr>k4lfJcT$KsZO6va2-&9o)TvRQ^S%^K^(c($MIW(ReMcx`s}5LTy2yoE$abS
z+@5r5dcJ4tc8137+-Bl%jMd6l(+)>FhHi~56XV#geZdxVr*hYJzqY;Nb}>l^juOBs
zNP??PO}Dh5u!*8oiNkEPf)<EgCVGkwpkJ<Ln7qoNjM!!I%T{~xV&(5`?;5wgKEV@0
z_Xb1FJolG>b`wR|d6H}r)s%ZT*U%P<*u&J*{=VnzQd638@eoMis)6&?XNO+Y+<D@w
zgLu3B+R|4=J`(o57ErWFh;6N6Y;BAaL+>`K&T9Yl)%%?6<h1SkClzJ;Rm$tD0EE8>
zOmS$ZM%K&(k!O0th5&*b2QD}ZvraY6E4D!a1{qPfTxJ%-!}MiS8bSrW$OqS;!#e`U
zX^$3(J;Q&3(f=H-|I7Bu`Co@2Nylxl{qDZ0oplU0!kmsDfN1}Rq}@RtmksPK<wTQ|
z;it;KUT~F2NR{eMw70>$;z^aW$<4d>gzG=H$F3K)_MWZ}qUUo^V|$HW-m5|U2;b02
z*s;!O0vMp%SMaye#3|=)U3aCW%Te@G0k}JsFL}L=2)k$jTfs}e-)|PYJ3t=N-|^qz
z_jL3A<}VH6vGIscH#VF*?#8BiYgK|k>~`<E>FfJ<B#p(y(F^3DZ|BV!W6x^YUBYVS
zH%^gRsOb~y%O{MDpj)3rplTV~%K9R_=9e;ZNEfDyy=5K)5TcD(payiQ6ZQFc2=;l$
zO)bMf@5)q*e#jEv6h$HW@2CqEEz0YqDPEJsL*D;tFRFV9{z47+Zm1%+$nUC<RO<}N
zw<ad_9IA!~`e-bflw)+Uf<#%8Yx@x<qOFu5$*zuE4NWj}NJSeAbEdGo_r>q1@^iaM
zHyKjz)$Lm$8B;eOw02oef0Qx*BRTE1oDR9%uZ<0FpavhWE7U|SM=@fAlAdXz%Q;P4
zsv_G|kfjvF9NT#W8}jf8F0*Hiq~n;0GR;b&uZBWA;;?1$d=d2~C85?%q^OVXNTLo=
zpIHbb?P_pICI@WGTrXZ{3QE3iSO&3vh4yAen#@eA-=?2$M6if#J8D7;T=SH9NL7q^
zM}Y#wZ3p)SInlFu5_;2ph5m-yki1#}&P|N_)HvtI2K*jGEG)a<k7@_V^q;A9_upxf
zW3xd$zn;N&&)20kKsPMyLJz3<cvFA5x_(Kuu82{RC^_?&aO3ZBmz_9afx1n^rT(+$
zgUyY`0nE7;si2{cQwdo8=O;5px4?4?sUSf>+Fu~xXsO#Zd?gv7X;KObVE~;QBP@HC
zopWcXY+wZD;KZ^^cR6O<kCeuWa6PH=0_P}1GV9DYx1!YMRdyag8Fj31gG~CH<tfhX
zSN;jM%~|e!FEYSbv=l!a=sd3rLHZ!S*ky(J!QolF7gIl9`Dhlw2%9Kt&pH>34G+`J
ztIsb&hcn!Q1Z4IkOk(J`?->e}YZHZ)aV?LT+qRePa)3E;i|ksdHQYUMVDl&5t_-*X
z;ve{R*@SMGwn93D25Sugfv-dXa9j{75@F$o??!9*PjWr2@`i7$p{*mh0DUlL+RtUW
zz;^LS@L9?Td#SXiQ?OdXT6Or;G;>q|n+=nCmwGN(2YNOgzo%$H*7p)O7rWwRUTIOG
zls$QXVeuM^v2q0MavJK<FqNmty{uOtHH_68_n5?>Rm@STm3$~`g4s8xF5v0{nX_gD
ztEdUo3&_!M0P&zt!Yudd-hql>dMlnCTq^~f=OuSn96fjeSripW`Eh2jlN)S8#Ua}P
z)my@khs$v2!0lEaGf*<2JF05s^`}|~LkruJx01+QcH!Np7{`JfoA+!ob+<BDBi!KX
zH!W7*sXA*%)J7LZp;z)O*@MbN^`jcDJNR&NqEiuSD9_$G8MDkRv1Su~5Y-G$2ysG@
zFgz^Bqo<NR+e;=8=o?<7X_rJO^#}Ofpp+L6ql_(-s8}>%q8j>h$z~%N_)a+DhNa*@
z|FDiF?bW4DfJ^1Qoz9BA#@n+Fu^thKn;JhnJUQAm?oI+Ugi9v^MFk^}tz$R#aTN#S
z%R4hlXTm=RQ%g#V4C$)v<MEZCE4O{4l%(O_5?Zo^I*-fEX~?}IjO)pr2BE9TnyXGE
z;xUH}r2Q6|`T(PULbhFRDxHGKdypaQB>%vk{~-+b6m}wF4bxGabqZq5uXCI}p{Rk=
zvKA^RUl_a2iTBzMhKr1i?XsHKtVL){Tr`F2->XmlEUwlMeMo9h*q+{Sfl&}ZZdTp#
z)?1V4LZo2PnE?e`2ZjD4vq~Q%wD94jU?IB&4_!?Nv$Vs}w=$S+9nj}+E|cAgiP4!Z
z=4L-DsiS}tnrOAZ3fC&YBblhN2Ryag?}vy!-UyUh7q|_M?IC%P)4dnwI>30t4FA>@
z--Im&+?GiA$NkwW()8svXbf{&cDiO3s-Ls4+^u<Yvpzq@*B{HhO9zR7bBc3XaNV*g
zsHIjd%^C&|)PjAAu$3zOb{?l)$ao}l_6@OLzIaO@>JE4b(sM@yl5tu)J3%QHX!jYr
zJWuAcXx}Yw!J#O{*0FpCG~|~g1%JTr;v{Ja?3IBtQ|Oxv*$c;3q1RjLtqJU>?8ls~
z(Kwfpcsfc!)AoyZPZmL{5uOZj9+OcB$?&L-D^_TGX`w4`iD)E{;!W7>-`xNuGoN2d
zVvii_shcnFaKEsIS$;+EEa&B3o9)#R2!~JLyTrud-8#BFDi)=9u$I#wp7c+_m;PZ~
zZpMcXgvH-SaZqpTY|ofsC%Y4C;g5LSLndOk=|fpqSTo`fg^hv(Vt<@74wVM*%|hd2
zRAY*TxBRv3;Kqd{`sq>KVfxHLg|^dDR6WNI3S6h1oTIYME6ZV$`y20psb>-ZIJtYT
zv`FlRIp?fh5GU_=&nn2aaI7(jQf)p`Y)um?xmQ^I%Uccc{xj7(>1VmP3DCd!24SSU
z`;R@hod5d}#>oCJ<M97}|D<d-E~ZZOVm5{@rXr@s_9mu$e9+D=PNs&o&>rhY|AH$H
z*nW4tsGmY)4p6B%0%3vMxc~F?GX3-PD#vFm;4uC8L%J&^cQ#hb03Q-a7mONJ#|;%6
zJ=!KNoRkiHER-e{j6qHqIC%YV1^y6x(M#-<c{2SHqE6rLk?-k}h+Nl)@J~pVdYgDP
zzqsl2Rx|KLxTXEoVJ`e{em}T3fP5u>;D6!vw=fKcsNs62-{Md{OqZpJ;r_X?gaw1%
z{RrCl`u?6snlzHwbMb&}xx2dhI;$;JqyVExE<B67k>3NM{5`&3)QE^>_<nct_;j|C
zV3QNGwCLyOlNFXquUU)MsaL!_$=nt6I9?)YO|J5q{D|c;c1fR{_?aX3eV^E{$!m?t
zG8W%m4@LW$_#V!8*fh9TGvR%PNOAY~S=|=)f+r0Su4@&%!nN}Hlzv?5TYPj<B#fc1
zb&FDjap|P13e#2hR5GVZpRvLTG1m~J-x*pdrTr)n_%LpXNt*;-VTRxvr-(jzL_La0
zn&s+I^{`3*_vKeLnXFhm6Kj90W``-+_|1roKfbn!G6Sl)=Tz8=u6kO`0jXQ4)vSIh
zz0y=>KG6&%!W2m&K;?|klKoT>uYP&_<Iaz^?*6$sNlh#MuP96&@CCtv@vmf`o;_%r
ziI#i4&0<gur+W(xxu6j5P8F0QA63QI@F5{I*qW#zS(@v?dVG{<`iSCsjjSl;0%QpX
z7`Y4(@(ycAE_&#+bfzh9&X2_T-)rTU;P|n<CR*uyI2sw6<spC~CvPA~rWF^~=R}h(
zWZp?KZj?+F(}%FUX^N|;3a0C|W1U<<0m-<JwSjp}0KZ1XshdRNVCU}d7^UL9OHu}H
z<0$FXId=mfIFQ<T=0raYGCgzJ+!c9pa{B7INjVJNEnW?h^S8xYd_%U2HB~^-Oo2zy
z>W5Zxecof#4q|R}Ma+m3`VvR<2>8@XdTdqQ{9~-9a<_5RcYIZ7D8Al8u^k!(Q|0t%
za65WY&8sUZ`1<y8S!k1;wDt_;@})JnQ_HHWb|El~dLOv1ToDB(@hU3k8cU<{U(V)6
z-qqIi4ta9hMoSUm2S7b7ClZo_@J1|(dB%x+3#BG0$->WM85U?6ZJ%e}s5<(-&FK(R
zr`ch}u=wSqQUD`u#w0ahtaA4**1D(5a2EQIHo+y&Lv1g7q!C6;#o$r7)aDEXJ8sC7
zeWlt7hZ=R`%<bQE%SgCOB@I-6n;W5bUE6^5)^q04ska>zIml4!2A|1QW;Xa0bBIfu
zVEI01b@)kMr8vtBK{}S%D4lh;`o~DU<@Pr8X~;0`U*kNF7SvZp>Ht(PyAUMcXP|&J
zeX3wJu)jG~2Ih@g)VJ?3z2pcLJESYNky4!{Oh#VEYcCu>O-M}duBs0<_=qo{T=o}6
zYZwjX12sbBN*J-$)pWk&b_Md)1=iXB0;zMQZqfk;zSN<EFwD)I=VkT!W2AkXSh%L)
zu02gR<6!j;IIkzZX^&Z2X0mpVeFdykq#E>5%Xz+hwf`EvYs1Z|H1%WUROA|~KjHe;
z$Rgxzx98`>rQlcJNII*2wo$&cU}iafW7BjSg-Sg~iA$p`!iz5Yjm)`s+7pHqNb3r@
z-V2IMld9BLO8~TG&xCX0MouZB>nIzI6QahE>1Y%zD<*~W?wh6mii{~^ry;(=<L=_V
z-*ttj2?x|0Lg8(bz}+c%Oc;;cS?RHvGandt{kk!VBp)moP^A_BrZn!y^mTIt)KPIw
z_SY9xD}WZ7rX5v;tXHZ|IQUm@?gYi`CKD_q#tP%&vZmFIXa8Nq64GnXzHR5B!-r$7
znjGOwYYZt1)w}nVlXg6DX)l9%3_Op)rPO=g`CvqG8F%~Pg#|-K6aGL5K4q8)rr=W0
zF5lZmD9(!!a55)&*R-UXw(3x(QkFwr?zQss)AEYlT@R{KNr#D_M2^NF**bj}OA{95
zwkKd8Z*eklV&utwMr$TcaW0Phm5S3gAlQ%RLjCg=sUGW?Xhtn@>Zcg#z?(gWmMJQ{
z3G(9^Ps1x+k@}huobc%Fe-W$#eV?XEdZO3p_vW*Vm499iWZyVE)g4VU+Y~aIf*g~t
zk&L|UEhwF&&^r1tYR;DD2X8j^Yuao;cTFH+W>rvOx8YOd*_@X59<?e{S&Z|b@|o@M
z#hBq)&DwQs0Bo=$hhOA((G~}BFO@%E5+XQyW>2mgPouOe1qRln@|X3<ukp9jH>cap
zfdH_+>gjOyZ*{`20kHOeT4Vn?1IeOTIAbXpzTU<=I$gNr3G;@Yh{u@IZ+o_A;)x{;
zylXt7Z*W7#RJZ@I)Bex;`v32=OiT>_T`SVi|3o?czlj)T=16V;%M?8#V8?chcK93P
z?$7`t(s}ge#1U1rVrk3Dn@J#U5k+%|3QfPOAXA<m8L*9i4M4EZEy|>E&a6t$<(ChE
z!1;+RoI4#uVkEZo`sVt&PjxoDZ5t7Hq$A}+dx2N?^(G$6hsX~-eIBNQ{<^t$_Le@j
zz5WTH4mCH7fvYBauG{02T5eJ6&MAh8&Rwn`2%Js3w)6M>e(P<(%=SdK8z}yHzJ9xX
z$!8y@Ffy@te|-9lM<FT0#V85EUv{=}vv4vHu4lDWe8lqe@yz5(i{GG)+gH7bQH^d#
zisy=PkO3Q)sJ$my19k!pPK{BOYiAFO0Kd3}>;)??;+QNgmq;`gVY@ce?;^*=duN!s
z&It98HKab9*N-ZgmnUjWxTKpIIE>Obb;YBPztGuKgO;%=eRMHTl}1C^M@lFhX_IdQ
z9C31C4A)y(t(r<{d6AiZ`9`kuWdLRh^J*aYECc)XL10d;;e1hUYf%Rmn&N85XPe3a
zlJhu}GH2QCSW{wCyLh1%;z_Np_4M2M&1ynznXY=db9xMktZ~3?PiF7H<nCtn{myp{
zB3M`GxAXRk8&rm`?_y0TeVnH(#!#|SmrIpm=1-_YJ}o1u2m(5=d>0=WI*#x>|5OqH
zjTVfKXD0~8Bh`=@O?c5*EBiF@l1NZD9TNtQgdDoPU|)2ilcP{Av|H&#i(V(2YKrT4
zh~UpoZ8!lS|Avn~4Z54^r+_p%)r>m4PH7l{8DTBV{ku(|LPbVmPMWmMs#X9?aT@-a
z&%~Mzg3Okdd6h;}>2887k0AboK~0Ido-E{U2brrIU}O&&zkPPLL&?UBr1kcj7KKZ=
zYEVD@bZ<(Z9^n7}h~01Q^+xV8q}N+Rep6C=Mx6k=Ct~-mgvFv<)F?3PDUdG|Z5n{?
zLTP-GD!(9TsGbmVlX*;^zcx^Ah|G3yz}NWKdTzcsr5v=psg2@mo-MU)N~^%um1<8r
zK5-<h7DHbww03U!*!wI>(O)KfW?@hGxrVJ@EZ|VGj{Q~pfetu>CPMvSZ~hRjVO_n)
zBl@a!dsRyt8;p6H0_p>&AzS5FNR_plpPgkDoou`+NpgwP)kd60egg_Es|xD^pGcL7
z-p23p<eV%Vo*n=9Oo(oS)sN#$(5?7*8B69Zi!;L{*m|N_Bk`J}wau<ax?V*l0D=H&
zi|$-w^BfnU!keL0NpehwWz~m6w{(PM)s|)4qXUb#5M@Y2(Rn0sOd&>~Z;8y2U!c3_
zH2P~WO9np*o5m`bPcopn8gU)F89hVtytsm8d`3u-z?hJMq-|aOu;?NESXdPB=Zv^&
zUUAzHgn72h6;mnh9x574RGFQshzfHedocd|6Ro@6)Gd;q>>6w);xfaat=Y9|tcw{C
zyNQ(O2v_OgUb(g5hHAC5PpR#?ZC?Aaq}p)xc^-p5e%=V{*oRvz=l(Z);l(r6^bfRl
zqi3xP;}GIxQ!1=<89o(mSC*zIEs>L>BctQQdortpaW@h>eB$`sDnWUM?S8ZB02H#Z
z1ScYtL0T$30UD^$UcxT7B#0`f?2*|y_{1L!lItS7*mxCiOa6#<LTv-6_Y8gMt$yHV
zp*Ti%Fd=~fCzWG2#XU74hG0AF2ReI?dR~p9`U{M@5~FFekrkwwYTMmath9~dQb*zk
z@Q4XFRZh^{al05|->vs(qq?QE<3nhRCIx3BaeJXg7K(objk(-FbMpbNg=S~sYb<eA
zwm}PSPwGe>NtEiy9#cb<{y&_w9j4vB3%{?K<<<o<(go()YX|A61zf?rU*)UiJ#}x6
zLH>}bl_mc)SG8s@C@iSYNRt1NWUOoHO{H)sO8YSmr0p^idsU~a<E~amTg_>NOYX+V
zk_4@Fhvq`rQZ9W#0)W@uSl`-(__c0gkfy7;gk47$Sp|UL15@Uz;g7M2G4f!i+Wv3`
zx-f+poMcu`;}+L5NY|<XLA|5v_#E#cUZ%}H%mSxQ8&v5ZH6p?z$vrO*ez|#lo9A#?
zNx)0jE!JWGc4Va-`qU4A<T}>FML?Hwr<sw_A!Vw)0nI(GPMo=GfgpdF+fkyeIw*=`
zkV_^4(nJS&t~&wdmk!>!8fgfUlU9Co<Y)-8B=l0os}W{Nq*y9@F!7rMW<TOy&A7P-
zmBzg1B~S>+Ud2=srEInIOF>O)$`mGVrq%SIebr9&|9p!6>I<Nl3%@kTvVlioyW&Pb
z5ylk_hB$Bxi$7A$(8fc_;pk1sBNB?%GfQVJzk()2N%z6weAt%lqGC*C?efyD6;l-S
zQ@wdrvP&?y8jl_>0f(78*s%8+o1ckU?2?pnm{KTtlpJ?3+Y4NgPy%{_?7MB(tXODw
zUYIPW>U71k;moO28r4rJ<;ppXBPkb2@7V6BN0IvI3tTawyc*Ds>x=)89p1vOFAN(U
zt3e2*MySjNm&N@=Zph`y_2@G04eYD@XsA($N=U<PPKDL>KMinL8y{?<&88aSHZSMX
zZ8^iy;edh$#_T^|0S_%+z5#=lE;kPf>V`*2<?7hcg3GJB^aZ1EDe;3)IE7uy+Z$%|
z0XKgB6)gCSo$bNO;K3&5#&k+SCqb8f=;UyT+A6T;v+1&FJ{Z~|t5b%P+gvADYcPQK
zEaw(E8;yJ;J1h?T7^Xu{m>Y3!2!WpiQIHj8IXQB2B*C8(%VNKv;*HbuotPIM*d;O@
za_Ynz_mGHLl(Mqxcnd2ut@CTGbhkB$gKix>{d$KWgh)_(<GZ4v6H{)_PZEfF8G04h
zIgzIu#MXc5bQxatJhkzu1^(laXufS^iP=~TUXVXm_%fX)jz1JL;g1_(Rx8RH9pJAx
zW4RU`A6185q%f0ny<|AQ5LEmeR8v8SI_dGG$*g{$+YeBJk;UeJf;s=8{E>;7^S@J~
zZ~ZHOBxZ;Gm*q8yAEA~k!W+<Ab|nk^*c`nXg%_3A!F!3eb(Qbmv-rWt+>FdEyYXv!
z(t;$aXjC4B)90P%?#!^kf3veA&5CJ~2!oN=FEc>zAJ34D#VQ9LU<tKwuyR(aJsn!I
zlj|4dOnGL0x;>BD!j44q^veopbsGb$`-iTJU3<2r{3ozxcjt|9xaE9y;PpcQvqfxM
z?Yv(>-Z;7R*CP7!1_0lG^Oj_Iw1F*ZZgtW*ePa$TiignZXRxikw6d^r>XKNt{JD+v
zvDlAd0ow?-t;f#3)8jSwTJ>T2C6sACuD?1!MAR@_aZMt0=OkB3oNMpu3z%L|hiq1h
z(Xr&zoOJw*o6=zW{Vp(dseis6+PJI-L$#5{IC17ZTyD==j5ySUU{OLhA#82}R&icW
zyD|GF$w%%3>hs9ZquqPcC#$-lz87}=?CfH=M*%6FMD`b}G^6j2>wDuHgOV(g!?2~q
zXE@LKU+cH@&H+cHWa~exo)GaB3812VD62%L9~2fE!@@dBK*8tz{gaeba(-dNLklp%
zJ-#*wWM)PkCjt10T4N;VJp@)O_kZv5Xwra7W>FFP+XD;{5zeIWsyuHWd~M*qyI=6G
zb|7zR#}9zMY}!-V@9FNca7V$FEk+PU38DnB^78L0-o0-`5+ItSvomCfIpU(OuWzu=
z$(#;%z@sqCmCm63NN&`wMZJ<AONOTFb93co)as)*MZ3F5{6xD7XQ9pf7$sw>!B?8}
zHfhgN3@EH~$U*5EP_38#oz6alsud(tOXyc%Il`n+V3H*9-GH*qm)TmLC*Jl_^HVbL
zDc+%(K2$eErz^Ua_$rQCA0yq=P6fa^FS|mYkKRSh)5%W1X724-iDh^u@N5$DUK?b5
zyF@`j^;XSza<~dsq%X|3KbZO*;so4>l(ppM$?zxvc=(StAHN8^vrbs%143kZ{|M!v
zdp=mDtK}+!+`Zj5F^<Bn9IQ_YXYK=L_UJB#m1z?9=bmORAX`djg<uJF`=`9=hmO4F
zqA*ntCI>J2uTPCvxqqnV&3MZ&Zb{`_w~o5Nqs(4x|J`m-H!6MeYVb0N6QUL78Ox7C
zC^G(%av_HM*=s<DQcA!C`m?2i;^qV&LuNatl?%aQr`WpP3KKb&d8?5jUKI^g;F#Pb
zE#E)v9cdAq6*nMXX-`HW;RZfdf%M2~vlLxm_WbVOIy$kTTkFeoART~SR>%tk^TY%-
z|4ZT5T(Tutud*s(3ZO*AOOPTfuc4B4lM^K5PY$lk4u@qYfiVME?JLMMldwkkC22wr
zFRAt;9L921?ad+*Na$21b+cc4G@_){muH(!r%)W=l*&<DoddD&=2kC@!UpZ0I|+Aa
za9N+NY8^Zconkp%8W4B0ddsjuaj(Nq<{9p~o<&HBZ?6TWS(i{q6ocHfL~`-T<g+6Z
z{0yDdQnD-{;-nBeVf9+btIY${`;NEp0sLRLk1z~a&!sm~rDkNmM?ZRo#f!|N)nxVO
znL|yL0<!`*=K~NHCA`D94g-t#EY~RvF(evgFVJ_+BWy`Um^reSzSNY+EHar8a*Zsf
z9(*manblpR2$T&6_&upCtNe@ZJd^z&d!9Mnixqohv*Z_!h3BcP427aX#rx$Jb`sUj
zPGQA_D%Uni;GBmkbDH*H*XIAHD&7)BM!T9s(t?r%q@B6}HN#n&*ximuD?`X`eC6-I
zs`%ilMWq(Ru-IWj5aZ5Bp?SMShpvA>b=bE=5#e9W1R+DbOn$B%3UCe-PkDTlU`KkH
zOqcQr8<Qj>Xw<H)XD;c_x#esKen}XBVpTTfa$s%!mx08Dn-;b)UBeu(hf2LD9QE5e
zO%OXOx%G8}G+|9Km>>{H=)r07vvv;h7)+TZ2CBohtub`t7{3I$oLQp}eTF-G4xk^c
z4Cy*^o~obctz33<7)_!Oi3|eeIigxB*)n#VJlma{ePR6ilwu5wvj9_^bv}7Pale(5
zz6g(NEL&bGyxp4(X|n#0s5j}2s@6I8)*wU;f3O>+O{rFOh1x?YQahzj)|y%UVO4@|
z3co}feWY{lsry==rxwqYmKDOBm{D<``I5s?wr=z<!W%Ly-ps^Addx;4-l_#f{sPuQ
z-@cf{+*{(Y1k0$3Fh#%Ol=yryC6Cd{5o<IWG@O5QrD-VxqFv?P3S>R9C_bvFV-Tp3
znGQ<^$YEJ0asj2FxCt@3h&4zaDMd|LOUMwaIx*T0-_w>-ugijnj5$h2kt=9$=tQE)
z8fw=O$0vtF{0Fi(!#p<-4gv1fnH>%*M%9&$A$zO2ahc~sC0)V{HT6=3uJ+S7YPpq8
znbT=*!0T*#PXD<?N>X>G*$8NuZW&nLl9jHhDqN0rKJZ=qB;C3ehXCbYFCLy!JaG82
z^c9SVp>Dgthp`6j&5BdxbDMj?DeZ;JwtBNYJuR0crCG_PKcehWeANT~MW<wxmjPao
zT|#rVc^<ehW1(%fls|f<Y~)a6BD}@#%pT=~_)5OwLpbD}ar~am0)d)sMDlb7lnZ!Q
zV6~(j%&|@6jOlE``J(uv=n#~I$`w5qn0ll!hcEQ5nL(Kkzw$-~o@{0x>g*N+>Rj+p
zN@%fD9Y%I`bP))PjUXE^rSH3v&Y>+VGUlDwvM_rlPhVPh7p_^=8Wc2=TZ-$_!>)-c
z1cX&q`e6vvC`LHWuCEBxET%|$j@{uvhEB3pD}7D~erKqVCeiCuER{h9;rW1y(L7yI
z8CRF-GfFP(+Gzi+-_X%{5$dEYffhtFaq$+A?{;7%L6j&3<IF0jcGit07-p0mWt^-J
zA#n7!*k4l@PxxE(qIL?T#(;7gb_x^=&ty>(fk!;lFm<k~08(W-y!OLw^^3|3tOgJk
zj(nh-w7x#YQ_Y=^rr+<cEwk*CR6g96YB~nUE-qzzdj7Y0XU${>(vFx~w#A;dIz>T{
z+NlRC0>;a*i*%-dX#CkbbZ}(;c3gPTqe2oNy_ut`e_*K*5r<=j3?`{gg>X2h&b&*c
zS_dhy%e*T!mZq<R9AXFY5;TC<Kt1ygrO}Tdy`p<`f&R1^!Haw;6_jfXWGrnZ?QDJ2
zS6?t8b-AO`NNz9gMRG>LTY^4Xmag0}*qu>;6mrl%;0P{X^$1@?n$C=!*Ki*706F$A
zR{kfP#{3@$8df&e|B9el`iIm05j6jB+9?7}8uP#&2qf&FoN`_ZofEYeRlE1P@XHTA
zmyc9Y3P<c>k$3#j?4rYQ)`~cl*E`$U**_}n?Dj2T@-W%lfrHmKN5F6P9!d+GRvunJ
zGwAH><8dTs+2V+6%(Rg#abnD!-kr~n=#~k@1a$Xx3^U^4X5!`=^0{V`e*$;k{tgpe
z+%|J>|AizYF812&?9`;hnM~?e2F%F;K)wgpg;BWD;9$iA`{(`W>~_YGWn4(X!s7kn
zgYeq`62&vSN=;7T_3q{I>Gp%7sk-<m;j_={{mvFktQ$om7)_(8X%e?856qfb6B}g+
zz*KcCXpJU3Rj30Thxq$97coD-mqv<3BS(aIBu;%Y1*2rDbo7A+o$ju(%))1a%$e9=
zGECPRHIl$MRedWsJ}w=4*PT-^b1jGus&4%_M-O>>*P!_vue~t3CSwTvRs>C2BiD|H
z<h`H=rE2CAJvz{{Z)21mjTfdma=IwDa}>uSlYx#PS-y(6W@4J-FfFh9v@xu0>J0M&
zy<^Tt5Ji0&%hLGhQF(zE<9qSjLq8EH`NNilqSlp4qpK*s;LKphAcb=zPL|Cju{kcc
zOHt4%YZWNMP%5@GTCghG7ep#evl?fEvD_1|y|bWP%PTM0SXi>L*^$qMJ(6u3g{@?V
zv*gyt77RABB+G2B+$m}I8cgRDOazt1f@eM&e@}kX=8eK<jD8qp8z#YY&V?q>-%RQ8
zXW+-Ll<@5OZFq#g)hWx_aR36<O*W*cEG#f-|2MlLHRYVGbl{SubC#&Jy^;{+A_wp(
z-4g5QbsxnuQ`8u&45cRXsl-W#yF*V)gpM|36Y67ySg+GI_f7Y&A#)+jH`oQl<qf4B
zHK~WlW4c9sbtk=uXcKMXp6?*i4^a8Ne%x1;-Le@sTe=Bg3ddf*`>5b($yOv^8kH2S
zk31Olv34`v8PJF5ziKNCe;va}NA$UxG<UT}Gfsp{r}q9C%Xl)Iu(iC0)?B(Po`$Ca
zv#t?P*NnA}s;`n`RDVyd%JEWRZx?4_-9kIgxN%A7+{uWIi|}5EEA?P&s+n8P8aO7$
zv@cXnxru2)9N1`4(*3nIpaioF_lWZD&3Y0BH67F+i27Ceo}X+Sjy}4uB(br$76Xd2
zp4vl^Va!^if>3G5+w%eM)&X9xt6a2^3O7JhX<c0=+9IM)HGEQ`eDxtoUo%Baq-rZ!
z{Klg`lbNF3_*^yGdR$z;cfLNLgr22Fiv?n<uV0Yi<57N%<`Sk~kES8d<WmC6I9?FL
zqBbosYqv6=$ka%)35C&d+E_l?O*WY&*gR=v+QVECYQuP$J!Z%hj|RchzKOpMG-qav
zX^OC2#zp&*xYrL&fp!W5lS|2yB|#(zFnO}+eY&c!RPFkGT#!BufBbA|4!<b%^87{5
zLEp#}^X`=YAW83|`r*5Tw>s!|KC%SuZf8;uUl2<}GL!l`<%~UchkHWHVv5`H^DmF&
zHI>Q3aEW!Y(G%57=}aX)U+`T!5<Y|gzK7!+ly*w%!#$r1L!#i9TM<|AAa40I!5j1v
zrWwDuE88ptir6_Me>`6v$Pvz$$)liF{=ETT3|}6UQ_GL@d%$4|!Zq)ku+JT`X&17&
zfD2Xp&Nyblaty7Wekj=X6ov{9Rz07$&uYb`qe#R$)gN!P4Y}Pbnfy*n&B{NtFwd*7
zOcLn$jlOo&qcj$=8e5IXJ}{D?8>nEg;`<##d$?y!%2YvjLz3xFkq411W|B`a<-F==
z3S64`UQ($=t2pzbe!fThFWyT?mPoC0R?2hLD${g9XGhOZG^X=ay}PYS!$XVU&RW?z
zfXppOmmtzyW;xqfJ;E%{*tAk>EkJa#lm_<q)gl6TbQ*p4y;RM^*gcKYGt%TM!;E%#
z;OYNi?46=x3ma|0*tTukw(Vrcwr$(CZF|SIZF|Q~=iJ8aaqr)wPd%)v^;i$JzFCvH
zLwM4JI%NN(Ft?lw$&w-V4+Llv!2{h2G6bsUZr`MQ;@gR*aY*zEGKx&tP3C;dY&!bj
zCUh$52dzC#+U7QQih!|1&H3Y5f#ZoTeVJBZu!hgS2Q{1kOx%&pYGo29`QeWJXf`U-
z(sUs3o$gq6N}j)=bEQ)$|I<;!%KX189~=z-JxbtMTiTA~e}rPn`SojlLowz<wq3AC
zY(X8MoHSq=;|4g!Ny6X05)qzis%}q%8GnWZIy(O=7(*iJTY0kh5+QHbbxPl*7b_(n
z>p`5I-RqqN!n1d$XK&tY>L|oeHS~Dw^FB#@x-jqrlXjv**^cA=Y|s7yPRjF$Fh?}>
z-ER@_m3P*$3H>4Q1#gf33Ny-K-O>;?IGfP8&i$lOAq~!(v&gon{_)rM?IAV^qfXD2
zcvB72G&{Gq_trNn|8Nt<>y|^gKkJ+ynQf5Z%(^D1YeDEY96fiO=I8tKTvqoN3Jmx=
zhWNEOgN5kzJg3)^%pa5UEB`hXlQJj+1ezc!x7kcfOMB$_R<A54H5Ssu+vYPMCcvVu
zq?$NIe}eleHG*B%!?##n+oBQQ#%RBM3IUT&W{<sC$mkz+^lZ@`weI!H7OL2n&xTKy
zYhEN>P#}N1_ULsY73ktX04?jql<Wf2c|d=4%w9Vqbxlov3w6_uC1|0?NxQ$iNh^`H
ztp&)8tN^w0m~2Q(fsZI)P_-oKI9WeRP<kH`=SJU$T)bA2<>hJJEz1pew5Qcg#7t0p
zq=iu#xB^4de0vc%_TRJqr7W%_;Ht77Ci~Sr(~&;V*m8JhzewAZ(?5=NxnvA?7M8El
zfqW{G{5C6a!I-B@VAsaAXw8MV?G4nHnvm9-AQw4LK!jrw?Hp(Qr+vods`xMP7&9_`
zWyPFXUKU|JjsPOEVp6ru1gJPAXNe;HB@AF-$Go0op6Tjh1rIdMzP`W}-5QBfRB;8?
z$xxh8RaQv5Se*TAi7XQ*XX_2nl~(UT3LwyAqZ#eGu;W1k>d~cG0!*2HGCu53SMajN
z{3f3s_<I4G_A3S{PYr*hHOONwi%DwaS2By2n-yD~v`NGJQPNE4T2{HdGHtGht7ntE
z9?@FHTyF5aco0d8U%mOOkOBbBij0}RIk)ww?B)l!YL&E?LW?G^<}2!>*p7&dql@pI
zM#__o>N#U!BGHuu>m>l}>;~Bx=5|drgPy^n8S-uRgPLO|C@COZ<8eN1-djIdlt|$S
zg@zeIeqiJ)Z?yQ7J5f0xfc6oet5-rnszc|K8tQ%44eQ=XD?ruxDj}Q6hx0SrU7aGa
zgqoR%hgs{)u->MAh(7*Io?<zKw8#iNGb*HZA6AXHjnbrN1azGixQ|A|o4fapv?jvR
zX0B%1lvFUp4WBZK`;6?_H&7)@Xm1l_BScjy?8!+*W>B$(PdzYk%G85=MRmK-(%QAl
z#4o}%5x$ihFGyssui6NLGqJ}&0zTgwcZhInVU6XO-VcxzA*!=S<W;^LrD7DP{t+Ks
zDQm$qi*PiD%4D;>(B2Zu>dQ{Aq2S*WwbEBuf73f^v9kC>+Qn-8Pr=m~^z19>7>Pf@
zuZX?R7qG@~o*m5OPd~s_Cf^cMtZPCyy9s;5i&4zRNxVm-sZc4h7v4C>uiz2=%n7kO
z(6XeUvV_aJ)XgeB7>kZ=Mj9JN^sn+QB=BCSX0*C6XJ7I_=)`fQObHDa*J6l`D!8O4
zN4oa+Xt0UEV>P0rNxkxwE~i8~mQh^oTbEF^P9@<k-a{l+k<FMr|5nlu)Gk>TWTJVU
zF?J>iA%J|92~643<IvtXLiY!oHb5TAyhoV>DA2?jBhVK|8(3GOvJ1kFHYE_DQaW^l
zZZ4^@5BetqwwYX7QUHu2WfjJ;)DOx+fy^SC5w?qX1b}8(66HZ79~>0&7dioaj35Yl
z<Jo?|@e&`!V=ff(^d^wU92ha-BW6zcCda2mWyFMg$Dfr&nQl2owLPj2%;-^8A6SH&
ztiQi409H$q%w^qIl`h<STow-n{#0skjo8ND3$tm5!?s6cOIt({wCuDp5a69)sYxae
zu?Exu53)qonq;6tIl8dcXdbyYK`WA@#AK~vGF_wL{yOl4<YtbfWI7h!!l6{$vf5Jf
zM278`1NzeVlkus736$M1<Jj!y5>*7eeqG;Z4A(JXd6GiM!`5Y=${w#0>pn-BC()yx
z%FM4i4fm7j|CC*u#sh(_dN;EMML!yvw<?sv03QF-U>Zgl#WA=_=`136Vo`$0I+QmP
z?KtJZ`-`=*t7{JP3(D59)_g@9yQd@nMCL9zjtc+bNm8jP^eZ-T)9PYoAel95*I&pG
z-NgPhrjST0w2m>OKEn~6T0GN9g0tb!2@KAC0}JB43P~g#LUB()($c90acgn;6F9EZ
zbcSmEVAkIn2Qteb9A6YoDaU&Jtl4KsOa*<OmNo54ivVVUstmbj7r@ZZ+EkDFLm$sz
z!mBK_g9wRD@jotgga-L4mqTk_u#jEAJfx(>v24-IwSCG;SSLp^4_SK^mj#F>IKuMp
z9&6zX8n<Ws7;J$bm1-wt2E<a9r3I;9K1Re+6krYaKOoqOIs<Ii{5aU0_<OUWZV2h?
zS_s>Oo$x61yJqK~29#KOaJ{3`yKf0q*!?d;+ga(=F-y#Ot3l^X6ek1TKDEG$5C6u-
zP7NrjDVzKJd5v>sDoe5vfk2@9X*b$`#dGgQY+?iK+3OSKKa`|LJcS;91Ft$pT3NKP
z7W~JlI9|l?gfA}6GN{^lj?$dxvd!LgyItA)LMsNQj>&@ziy~a}L(OEWv8Oh|Q_Sc%
z3qmT5O0hV_$$TsPFNvxKh`@B56O~vbf$Gw|s&sMd^|}jJ^@W8aNh-6fPNvo+8p8e5
zkuKBMSvz9mY`LCj3grnySIQP7mNgaCgmDaEc#~sj^mE$RNU8I^sKOf3f7?J5A0`uO
zArd*YXY(WJ=g#JE-S{L_cTx~o;Kyi@!84zus^{O2WOyWcT5*w$?6mKY5?@kw@;Zyc
zxJF@&&x3qru${h6q!4;w_ZI>-#NDy+%)D3+xy}0N2f4+=I!Cm0>G4%M38zeBs4?y%
zGtl&H`joV0qk&bm&u9-xF50ITkbuB7J!Z;0<>an;#QLE^GkCKFr|yE)G1#?XyRM?}
z<Ylc}Vh=WDQRb()9XO?}jWnalES8WhZWIJkZeU9j?aW@^o;<AwEWuhvXM@Q=ax*%l
z^CM^}*VE@>n}SW58=CrBDOFu_K{<=MLAXbZS;|1HK>szAPJ6FOot5AR+<rp3$d3(E
zlj&ek)(O%f(9q=med^&ya_nzY?8Y|J3DRNVs6u1@rVG$HxV!|Chio!&q0Kw3GlVqB
zU_!l9cmU2KJfgi^sn0iCnn6=c>r<}?$1+V1Kv%Dx5%||`DYyU;O1B8lFq;M)M30tH
zZ*UP+&K1EApRTw(IlwWQofY91MJYL`!tlaWlDi?EuEamcD`S!GzLWe8cw*l}nUnKQ
zrHoac%UEJCZp%PQq0-V=Lzx01vQ#*po|!%D+BSuM1iTTRZ~>V#QNM;-TSH4??aH=#
z8HtIp_WZQ=fyM2w5+2oYG8E$~K)tDq>Ixbe_@v^*yQ>eb!+5QOt~8U1Ld^z%49c#=
zV+R~50Q8WpeVme-pl-gm_vA`{O$p_+A^ov)QlHr%_@p1a`(DbwO1|XLju|s0<H6$t
zCtT4+dkFdk$&Jts|C~a)Yer$J_zJC-%2RQQ%G%wUhoH6zufby~Kr}kZn+QkqZ}fMf
zkgw%0K6gh^oABGiF$0%}z4*xEZ8J?C)47q`bE)#MJnb}nbGb<c2$lRKn=zWNq>IcA
z$q<!R9C>!@>VYJPcegO=r=G^+%gTR3&}@wVJA!85VE%8BHN(HL@<TS5-r2f4e2p3j
zMS6V%f}{+Tvzn+5R9@8WT+g%bpWiP`<Aq~}V+|jRELlpy$G?!UxSN#i>*49)_wK36
zqPfqak^SqP?L$uF7rf7(PXa{{>wo;7ef(%jHh=Zpk1AG)q@Ucy-F-Tb0ve_E7YC5j
zqY>2bn|XK!IpM7Pdxtxq<319qh^)SL$E&nLUX_dO(%ikE#T3Gq^X>lgBza;au`m0I
zWqVlH_EP4THzJ(SG(g4t^){=VP5>&zncY_;a9M`$C%c>Hi#@(FO~Oy?{JLky|D79}
zV5eED+o)5*s5N_JDTx7{#Hd@&kaASnk~~Dd?9Fw(U<W%$X=kV7Hm`<JLspcdFy*^L
zdefIk?v>4?DLApy|Fc>Zq#DsK5P^omt8A>DZ&ip<F5SvAon{Q3xVnf=Y+z-g;UdX%
z#jziTMLdp|PhJ#FaA`Y{7_!KBAlH;&hn)y89y^vOK+Ez<1^EmY^XAQWEwleZ>ZCA2
zQ3HEeyy)pDDkVtdeDgY^6<H;#iE4h{I6&ev;=^!}q;9ZxmMuPbC5>rPNAu9A3EX3R
zW3xE2brVUHsWvf|suo#YjMQzhJYm%uV8w?y3F5Y4F#nK!z<5sHp-$axgMIkMGD;#`
zX<y{LD8JQX??8tIGtU`4GWLOFXtgLGKIPrvUB`SudrqF)Pu9Ed@{FSlQVN0d@HX#U
z_V#j;pmJR_2cf1w`p8sdi!5%yrnF>1KH&x-Khfo5O(;I<F{PL{S}0qELS9Lp-88(B
zv725#IM6U-w$cN`pX&8QuH9Wb@#n-^!>!8Iic6k#IF9&tXN~<egKgBrsH9%QNpb9w
zH8*bMZ++VsaO>jy69@kWwAIv@{8q^W4J|c1sghz^K;>uE7IawYhc2yixV<)0`3~Zj
zjUEag4YK3FcqZ<$(uLky)MF?^a)5cy@`EqhX`~?fa(6`}njW}~PNXyc%%wL(EXNyy
ze`0PExlr$IyP=FP_20NWxETA|lZz(`08GLX35JIw>G9N}Kz)Z+%(^*~iD2{41Xf)^
zXgy`g!0WdG=Qu}MQxt^GkqhB<waBm0#WmLI0A4Jd?sAdl9iG#(LPLbBL*rrE0=zM@
zKQzSkxE6n6C&d%*T9>$X%CX93J>YpwxA~CDqg}Z;tKBWa^2_6t6zSGZr&N#TnMB9+
zJICAK7g$YL{Rf!{<%NLMGE|zwDrXH9vD0*;1nK98+_CB)je<Y%{DHc`L>g2FRjW(4
zkAk09%3Cgyo6(G^)Re!Ag=Hu@sUYgx6hsRqNO*_1m6^|LBGavs0r0A@|HwCl%073<
z0%}eG5QBR3{zHMPKNMyxEQ^tKPQhTX#Ath~w*XB!=mz?6!jz0)>8gyeRavRUu*+?r
zWI=(F1SnvDRw>Nqc9P&hF{<sSQ-pf-k#FNr!28FZoS~KhbMQWwC}0}6?bZs8LK$E>
zk7pk}^uh|xMx9x6BH+LUx3rdWJsOo6a3G#9`|H5;{&ipy`}i}60d(yMV4f=YSbM=#
zr7@h?_lw{Rsag^1(@HT)GhS7wz$(Xu)lE<rXcYnQ2dbx)&L4PSQdtA-m;mn3d_k;Q
z?<{4)SuQ>-NNQbp0mLhzaU|RAZ$5bz4K@`xohjkP`iq@3Y&i`brX(`;N11KUQKaiE
zKer1QKv$w3hkRJ+f!3<hKPZFZpfNI^C|A@*X<dx2ZiMi93*8GTPnJ_-6d&R+tvbeB
zhT=WZY3@0xH?Ijc(pli>(_5~M?LjG+oAHHV9cXq&`jT4;j3`cbF$AsBm95B_^Cv(;
z<Jpt4EuC5(4$q4555Yhhy6WhlWuxVL_BM#)**s#wX(AiDTL?%X+U7Znq2Z21a0`?u
zKnXJs61nLYXqY})U1n-G(uuj#*9wSL?U@gujy382py&s<q8+&bK!?Jox&fqEhP3~)
zKH9XoW4wyIAgk97Ly&XVQHaX<X2o#GSxzbun%LI%ficcMPg;Hl>jGMi$?q=(>}2(u
zrJYXh6A9LK-EFy~=#7Vd&dcpEITrLcuAu+^7}0#blA#u_`<!GgUUVJ(mHQCe#iqvo
zszCP7DT`4&v<EyQ*HuRO%F>bA9wAOlH=d$#gVk<BwNVa<GwYd0MVpB;6tx;U4-6=#
zGS?3GPJ<LCfUJ|cVCCwb)@UIe+TNiiT`XiPhPfMk(kEj^EN)7kD{1K_0yy%MgYM7~
z7Aqoq@sLwwk9$;67*5NN=+x_!x_{yu(b;nfocTIV;kxo-IDuPo8{TYYem_`R%VjC<
z`qVngXH*8F0ifSw>{4tC&qIeWDB*l3DX14AZP1ar=k(xi+(x=NT%cq{`y8$FLJUUO
z@pS3(cdZW4%YDC9Ll=m!!uD)6H-H!G6QFP6u!6eHEe$}(+ALAD{16T?eJh#G)MQ$W
zlP>dHhz6ITy)^e;efC@=4n%^YZcsyV&0a}+NLd6-#4Tj>91n+;2qb3fa{|P6T}uZ9
zi)r{It%8?`=uWpBC$3Kas7@(u4FIJeuJycq)U#au{48t?wrN#YsVPNO6r3$Yoz&b|
zK1o|Kuv1mFvY~ZkwW4+B?JpCAfv=}BX;DBzUwl|ovrIQT94^8)mMq<mg4%u9oZGe=
zhdXT+3b86_m|mCVlj^ru;jO>pQTlG&KKKcmYLXkQu-ZEjOiRk8w_+yQ(8r<I9mf3x
zOa#nr@GI;%1&$7JPq0ONq`T8Lt@}%R^!EZ&MZ3A?{{WZ}H2<f)bWXPaU{+>k`0oLZ
zUv=pJ?y+$=ox@Tr)Rh<%_}eXDi}^@9PC0f%icJ6wGByieQmEqDvDHe=JXWyLDn&%e
z$ETz+Iamko@%pxZe7JupdDstW*H<w(N)Miop7<SDENBG>C!hfv7PkxM=bNac>8X1r
zNZC^Q>%^DOJ9#bdUxdj=^gk>T4u8md>){UcIe4pkg#7nM=kTChmRaXBmm;fZQJ&i6
z*u2!qoY}|e^Kuj)pTw>xZJi6HnQaos15*^fMkMc;WoY!3Lntp4FA?gTkp(--*kW3`
ziEfTwE2^x!x0?_T<DP!#ewx8=cIl61LWQ{I1O-<q>ggtt<hp3aDSA#Odg|6mo5lA7
z6V~yN^1gX2R4$XvPJe}N+gbkSz~|(tYG?-%>2O!+Euk-Y7kO8>ev!q&=<IggOI(;=
z4A>$?G;`SBuV)0ke6WHZE(%ZVsr{F~_`V*m%JL~T^dZHhlla8&e}6pB#CjXDz$I<z
zYFJAhoZ^o@-0Q)D>2+i6s!S8%B%z_O@;*EZ&zS>9Ag!F`?}j2hb#dXAr&x=V=d%M&
z>w`u*=BMscP-Spd_$?Nv$J(eU?%^1HE0IvHMk1i?7^RY0W$0{79Fr<={?%l1Trg>I
zQ_W7&;W60^0}kwhU88&>e+gKHcW8U_V*UlSQnM^jSglWe1pp^i(~k!nAnQ#7Sae6h
zQ@``z!-888&mm}74HbJ*5tKsBx1hOrWvz=L%(6)sKx*<u|2jrYqFJ1Ch2e>U(J<Y>
zdRF^**o4+)I#nAX{RlIvkUx&O<y>vXut%3yA3;Wu2zdPc!TVr*sR=pON5jbxob514
z>22v@-#K*=QyC~eEgkM4vuX<;GX_}7#9_$fWKMje&ro`{c2CFu6b|K9dH2>&Kp2T@
z+Wd>ci7=ji{P1~7CtBS5Cx<A9UM;N{EiF{Dp#F>~7n>qZ7a_Sa9?mPQp1VS_+tI_s
z&%eKVXW`vQFi}r$1QHyW9+?CLDOxi7PjTH?xo5<BK}Ualiz=b9&%DFjAt7v4J9YW)
z`@wFP{IZ+K;6Bm$L1Nw%Tk$is!G;pVeYJJLainKRP(1`j`V}(gL1yJmsAS_%=QNH2
z7TrU7$jqA>#dS{O-}Zg!Ttxy~^>9pGWY*(N^eVAQe9M^5S530)F#12I^^(Y#R++2g
z({Rg)>FBKKT5n_@&O(jZyAcgWaaog^yNQ<%l`F_`>=lY#AzV~L96@=>T(-BQVxVrk
z*VFbuR&272N}^XV%obq6L6M<SIJy~ozGd?QhkWE&cF8Br`Y#aiOHqvFp?KE{m_mmD
zz|#?4%eO~LOeR(s!S+Zk@bnmFdRq(A3d5Tl&V-fW1P4-)OXXQ<RRY@n1OwZW6?B6r
z>pInVPrC4S0vj{qa#3z+p2G~fAhmz~Z0+>!DNO?k050#KTdo%JWfXUVm2I0d=H<Mq
zifeiE40DX)ere2DT;j`7T?g@Gpt9r`^Wfp#NQb6Jn<Ij;Gtz2VxtA%bMg=DnGRp$A
zxlg9+j6m>t{)XH~IkKUOdT;U@vdl_3@5w*`Y>%I7CTd?g!t)SiVARHLoe_oz{ciyW
z-5o_NNPQK8L1mXN!l+%sYs&2Aq6BHXitJ!RXLP43!XcFjM*d@6SYQ&M{wco4v?dUe
zV6L*X>L@b>2u&Ftf0uN_aeG_th|F<RU1GXbbQNyI76MDQvfT23$TL~F#_uB%V_Bsx
z)~%9scWFTi>x*DawZWPQ2fMo*2oWa*_H+oUW;An|Z>7hBd#R31RCPK^f1At#wKbu5
zC`0aLCUM%5P~5RV*l8>;0_O1a9jnTz4C`i0H;|qc96ud!#bvp{i(bq*9ZN4LcTze1
znYi`{^ul7Tx{{!F_CD6x^O-ad1UqN%N$r8ELU|{%_UX_x>r*Ao_6{Abb+03-I0W7A
zfT4eh|GAOt^GMub?mrDBt~bx*U54Brc<NWLy*6EjZKvEi2~sufNmLJkTS6C!3BsE@
znj0egG{jZ{0q;s|-C5GsA@5Bx&l44qa2tD>&_U-G&fT1oZEge|6`9y}jaDWq)9uGx
z`PFo-Mo<sJqpQdrgvYE^G;FZx?oWq0Jt2AD4<YL5Z%|UB$#Ivc7n5Z0hZW_v9^RAG
zziB0-vYvBnnGNuX4gwu-)U?lsNmKIig1H5^)-m>+$?>i;KSvjpcZ>&K^N<0~?W=wK
zEYAfkC~-cGy4NU`_6n*Pl}&1=5KfkiK~FDIY--B7a?|<0?8{XaWKw5K)OxKcCkK+q
z+$Fq1ujlZOtDOT6Q!5k!*s$vEnQ%uk?jc15A_hthFBam$;hfFzo^PI|e?Rys6lgs<
zu{voHk`?H@(<%iD2zhTc^IcJ3w{#GllassO@^fx4xg#0b1A2+s8{k+Y_Z*TgOpN|C
zqny1u%1gVT$hT=%mMm#n)(dfeq-G{qwZ$%k7>Tg|*Jq_97lu35tm{79S0vv;yhj+<
z8Fm0tq`$y=sC8xo<!0=+3Wy*@0lcq)k>;ySIc|SY=jNfF6`XE-pS=gO>D}1e_OQEd
zd{rJE{?D7PM|`hV`rk!&+%=^Q;$WQD-kCQg*HSb1UguXg+dwQ}-89r42v4x$X!5dX
ztruditBL@4ZXz%7`PZz3o+4+!^{#a{0xxlkL9xAXur=Rd*|Xlgr|6fAse$%x;M;l?
zGUgpqWSiU5X>)EPn^}Tg$J=ILtC2a(TV~1721oE6?>y$iYj#3=@Ew={<{h)f807(q
z7H~UwA<Oi8z9a{41k*GdaQZU>RukS?E-fvCDZG;veWcJ?_gUsRHLkBg$GBPUB7Uv{
zLBCVz+t|zB$_(5Vc^&t>R50BGVmpfTRI7tn`s^9ga-^&4NcW_UF$XUQx6JOZWG$iY
z&1GEm6`g;ND7!PkoWR0m#O>2!f@MDvZ{0Cc)o!0Y*VJy?uEG~PZsboti@n^%AHa)V
z_i|sZ5`TQS6?RX7a1GgaI|4elLp<{tfcO#Ix)91R7|CFUzUYQ3l5th>_`Jy1eR}<P
zKGC0qj|Be{d-ESk0U0?s|9fES`u}2YPSs!Gqp2bb@%jh^$TDc=w9%dZr`1M^*Jg5N
z@>!;B7!QX$QRcVV#5WIT-RIi)#p30@_2W3YvuOZPW4XuY+@G7oFYriAI0qB~tibDI
z<?EZKgvq<-uB1fdX!71|)NO;GkOL2J3#i_&5o_R8_@>Q^Vb<d_;TGY2fSIj1lVc$C
zR*~tNmhWXD<v3x*Qu!$B3*YYVjd5-$%1&_I{>rzf&*oQ;6}gE56shXVY1^N(IS8Pf
zi)M7V`j2n*7n2ul>Xr+ltM|BDy{^(@C)X-q*D97!^CL9^zy-Gv_kaTGj+C{M*9l}J
zH;Z+do7RB{XGfs*E4BAYwuKf{tZR{X2kdk6;8t6roJAmeu|b`9Thf*UDU5RFzQJ6f
z3rlq*B_|=v6$ym}uFnNC+kHHx7S_+gp1@aU-w`;{gxQ**7o${RNi{6gV}<yMG-$53
zHrUL{Xo7zw5_Pj{#YqgihfXDFj`GP`W{RXwnQ=WM=6T}_uW)5Ym*Q#Y76`Fj=(P0X
zamDb7ltyB-CLSJ~f4fP<|F|6qNL)+%KT&d$aPfi*3MNU6G=jCl2iBImFWtW_9*fp^
z&?r@@?oZ#r)0Qt(BkRf*OTuCEV*qOj;TMlRe2<UC`#Z}>i{d-2Ip1l%g)?Kdz@X(R
z>ALGWU&Cw+Kl_U~3vq8sf%NDUss%<(h#Xjih!FqIb5kdn9u`6j$!HXQsLC{sZrf(>
zSxL8P6IX99T?-=FUbu<})_HB;bf;_1-#eIjp_4YxgsQ@g)*t$JIx&rj#<s<cTjS|9
zlq<yNlpbDo>#7KKuKOsWZO0{FbWdvSq=74^E_`r2brb&kR;M;x2N32Z-zmPk7{MUW
z-#Ku{Nh`GI<&C8Y=ea#Wok+cAB7pXgbCatgS3o-Lv_N(j-hiy<U<|Cq7F3T;7Z*6H
zB_uG^kZ&XsG4X;C!~_H(z$-*8GQGE(KwR`nUgwk2=RqHh!fIJXbfRBbGw<@h<LHl&
zsiHO6(5(o(mq3qBb`d`kLD7N-B{n%q9)U|xw%!;0K$*Ou@vJO*L?BeNh)>9<p{0^O
zC=)zs$v?fK4v&F?;4scnF-b8qn6pBnC0@$Vh%rNUxK?nC9+oJ>jE$jXntj(f)g0;a
z5vryCz>`N8u`~i9f7?%1=rltOt0vaca(Dwt8tqWxZ=K;inr=AI&GRw2r%0)}K2D#f
zPahurst$c#A}><G*}E~gqWZ?G2j}&?S9@F>%UGiuK0nK5(1FFuo(W+17S`tq`;CZ*
z4?c=rN~P~ebOD#<sSQ55f?XPT5;p|UOY3`5JMhZbk!D&k^uSJ{G583#nWgW@S@6p2
z&Rt`c+NEPbPS}|jiW~Iw{x&H9n!DwUR)9ziQWzoiE~^Yct1KcM@D$I2<jX!D+v{4<
zS7QylX+&P-ufYRuIw-Ur>}#y=p$T_^TL=OTdAZN%75t5^?y?CYz9(a-yr?KF%QdvG
zUNXhKRnReAFUWk+u+k6quH$N6%PkjgpaDB#aeD3PE3~((fAv9|A0x`4-RNsf4AKu?
zOu?hv&a)zeZpO`SjQvEOQ@O3kji|(ri?Ea&V8o@@cyZ1FY27wOCX19w0P?qpVOWh}
zP><GdM2icBPOIou7^RM`%=IFNvC$?=d7qT*z|VDJAImzK=_`w_JeAdo@mP%iGu2I-
zKff8AJ^Y1i8R#tTiua-XRkEgAfW{K=O!z$H{zlx#TZrE22l*tp*E`>dw8g$~1Mm`7
zFM|f)g^($!UStQ0r!q9`jTFMFPn{H}cF+wj{7tEX48L;S={ar~Y5y6H7OiC(E<oM1
z2a#4tM2|yahQi-+X28wlnb%Nsq&ns7HjkDpl|M9&l`y$y&ZlZi?4%1W4lL1!O}UQ8
z&_<^jE2hYK)4;!fwql=bg@g>=h8A0S5>7RR3_hxn9=AS_3t1~4!kWr_%~YjmCBV2y
zYs6vbOIQc;hq5Y#+yG2RAG6nmpQBe*4tnz;N(JRZol<xfp3s!J8nz!SU0oBUwG3UA
zc-su4C(l3(l3O1njB%saB;A955F!}6KDcmx$1L54e-Pu2(}*mz;{(RY4XZVHfugr*
zOjC;4CuHxvgkxjkD*O>=l%B8lucMQiJ0vnU8tOn*-6jfVEotlW(qFU)=0{A%yrw$U
z$id9N7=i#!i+*<@Li^{!XRsVt-_SR3o=#DnEJ(<Qz8ElpzqY=iUT8lF0HJ`LmLh5j
zUjh0Dclv9QTOM5yBM^I9yB-SkL(zKA$G`n=%qg2uAOHB1ECXbkzgyDG{ly5*X^|ah
z90<#5A~;f8555FtSH%A7nMMtBs!TnvBe9Wva5!JZVI!e!WI*)5CkG8f-ZO$3g?G}N
zm<B(+;XqzSdkSpBH{mj=g*ck+yU<r-K>z@bGp&WScH5AWQ_Yi0_bhqzwe?-Lo~s_%
z_cg0$I^?JGh99S-NoRUHU1mDz>0~x{2TzhJC-k3=Y!0UXAh`UJBLCy{|5IPGt;KCe
z!VYuuMP15+UWQDn1aI8n-(v<Yk=Cags3J(G8hP}CUXV*{VQSV5|LTu&iEBmDE24zo
zk+`V$tGRf)`+B|JS2{HQ9@(rtK3V>H65r5?;IOi20vNE1o4bp4;FKx*v&?=9QZ;Wz
zZ|ncqarv$6d4S!2{juvr!Ef?!J-mUOM{nfcp?7rE4F*d?n7;*N6VA==49e=MW?E6D
zb8Ww7ZC}rL%@{>fBgMN7BB%UVI`0Llf-z23$$(FX%*RLiMa|gn1qC&d;$~u#<G9S%
zEEK=9auj}}y%8)Vt8i6e!-{{L6Y<w5C&MoByM|3OqMjb6?$Y})xh@zi>R2Eeyc1!)
z6KY5Cqq7WBR}<RJ8E(Yzy(g$-H&%P!$0=K>vlz}mDrP?v++ev8N_$AInYm3XWIyuf
zLBHo^{1aP7Rgi407Sx?P_G-Xk#+tpDd68IbgtC!xkV&2g3fww0<0d=LCR0>yq#p=;
zc~Q-3`kuwIu=P$(HLA4EE2y2KiSJ`={c4KxAu=OTkhpnK(Wi-O_`PM@0rGm?A&1Jl
z=L(&INz6gfWaO!_K!P9yIu8q7##$%Mm=hV@!$=R=$UQ3vf6EDn3zQLBQ8g!N5NTNq
zc85Kxy^w_-o2|ho3QSdLt@C6xQOj*0_N6jo9^GRWV|vAj!Ys40VR=!fTfx|6vRQ{*
zChbdj^N|zYuyxJq2<oY`@U>}}Nb1XDOQ*+t-SHT=1-z{LaY|y|zLi|>PVg3LuP@HG
zmxtG@S%YeEGeK+v|F<s$eLQl`3CnddgYWyx?D^*4Eb)9cHmTVa<qu(oqc{7#G#^0E
z5kcArf{P%WHw=Q@kt7By_}l6T(z*~-XE9Ak!xZ2nNPxQWRt&>o_Ef4)!l@-z2UU(6
z?njRx6?eNYffjkUmI5>%LXpgLFGhK=1asq1oiC*KeXT*g<{L8Wg;-{h5+!}N3pDJJ
zei1Z<CHdTX0@2<RiX!9)Idr6+UA}p+*$X!1!n=bkfm$nRCVwuT9zA>4ezNKFl&;w?
z8u^p{n0f7bweVhB>P+(5lyAPm>Y!UKQURk$PCck#`;DyWMoc?J&3LO|Z@dF4Wsh=B
z=XeOX4sIvV%?aa!9V$?x4#J`X7v4x0uU>{J8YmCpq1*DDYH}<O`&?En#G^aOWxtHQ
zEmJ_27Y%{-;T_SyTSUQ|usk589P@=F)YT+Zui955W$1W*Cn1|D^>8_rMao3Ty9OI^
zwv9CJLvj#okh2*-)~!wCgdg#m?>OOxG#|oN_^}ZVk7s5eO!QzIv0~jE%J^Jq_Sj>`
zk2`*RAyB|Or9k7)8*;+{%*+HiO8qF@qS1%D4hAL!e}1NNo>Rjg3A5Qp6fjBy>XOVj
zY4{P@)_qN)09$Y8o{-Uh!_(#e@AXruUHvt#B6s{@7T}8H39EdjNY*ArT%S$=ie$z!
zFKpTB81z2F9I~?$0cxn0p-1T6G26KnvuX&W)kn`}c;i~f&!7#3qKh+?MinnZEin6q
z;a#oqP))YBWWRqN;jU3r%?dyeMruS=T@8++XOGyU8YkNLsEDuFS>QhUZ5lic4tNz?
zIxJv=TMFFI>72pR-s_w}p<DQzL8-gP8G|ynRVvKkUK-Hq`eZCx*29S|><LjV&U!}w
z&VJ%;*xI85+`J+<)$<&X-6H@DYQJ$#tj@9sI9DMGx3{6VZw|TT`6CU*-A0lT9oDkN
zTms=TC=zT*jp7XOA654t8R<6Udm$4}An)T-SNE5XdlU6#?1M$x!?~Qk+9n@`I=3I{
zLHkJ2(H?v$I3+B|y7ePBb$4(Gjx77rmqBxc2uen{2$-D1@s`ySGf8FrS;fnpG-K2X
z-!mWBRN-q)8GJttS9$Fo&{SiXI9O!GO6RLFz)?+Ki1(nRu2~Y=Mwi1gDmWpAepwTA
z?_~@xgP6Y}H1$N)ENl!(IHl{#<~e<J*Jwy>xr`xCJCR+G5yLxSErp^+zE`KAe1q|r
z)m8(ChWtMm0iec#yKlIrI>*r~ePBI5u%QP}259MtEFGo+UZx*63SoQ@ljOt5fz4YW
zlVs5S<}To5O>mUW7m(x?o$Uh%3d`L?3)fVn$8)-&=emawE@4uWDqE^DcvtueR{R5}
zwu;|>Fq`y?6=veHZY3NpXGO)jiksK{t;mZPs&?+hGmEz1bU)E>|5o3cKDW8MJUX9m
zO=Y>imSM^&SK(ijO~1;wI5ZxY?<LO3*SqbyId#skP3pJVSu!v8Y&ZFQykw{`qH1dG
z2ei-9#=3)(jy8}cWjN3}N%^gv8e-*peL;C_Ef1W#@fT>{E*&N<<{6QfiWWz@;6>6p
zm$j6YB+;bKpRYDukTqP8>(e?VZ4|}W0oyj}b=o1#LQxMJU!b>t7wA?;d0B3IuxRgQ
zCS`P*A22Ih6PHT*KWiUBk3h9{b-TH<QWs+N{k%wbRXEA1uiCy>v4@|k*&j;sx88Rv
zcThNMGna+xEN8DHr>Fw7;;QOung-s?l7C`d(9%+9#>MrhF@LKcu<^>TUqWZgXuj1T
z>C_agy&0*Rq&8ZK6*pb5sa=tzeG6=EExkmA?Kzb!3yz<7`|P8ZCNV>AwGQEAu}_V<
zI*omn|In~LxX3qhud1<gGkZH1g$1n70J^9--i9=rk~}BYZioSuZCQ-3Vl2-34{|nT
zL?OJ1^v!d;K@hJ}c$49RRuer5=+l7Rvzz@YEH^pSu>S#JaIpTr6qf(*txvZ9XX}%N
z`M*~cv#h0Ur`>_*J6pGR*ec*LhM71>i3Zr1b)sagKr9!~8i44ITXQ&+0Q1Hn`{VPs
zfJFN_!A4T8bvo#p^Yo4>dy*ZUYeh%LzQNQX;)=c_^RCTU+!#NStJF|W0Ae$1@BT{J
zS+oi<Js6VUj<@&|BDozy`XgKqr9SXW!LqI%xPtb#gnHv1^??~1IuKrSsr;-E#7KMt
zyg~1ZA~>9ugfp_$C<V&YCLq#cX#TkzP^E#_(m6ShRsm9_@rg=Ja`_n*hb5^-XIRnV
zQni{WqX9fLo?@*WN($Q493ocONCLUsl9>Rk)EuXLyedbDnS#3n%wbEe0?}!c5*5aQ
zjw1}rRVVoxccm&hW`GJRB8Vy<<&)AlmuQp2kRobx6-LTNOolFi1?slGQE|AWP$5jL
zyiKXaZJE?YFG2^-X1APfbOay$tQEr#Qp<EPhp8N8WA)6@_TvDeBb}83V}cdpdGDfe
zrq_^{)rQhcU8vcOp&F^l0ZN&AiDT~ws;MVPNs1|JJ{U9IP{lzMOC6F~-){{#khU_0
z--YRr6Fkg->@+*uo*R)ZF=NPJH*tg0(K)dwW>CN1RV)E{UyPALat(NelC`yCt;jSB
zdPSt*rkuk$rl2NKQ1lRmPl>Xn4p4=g)4I)(n=`u9k(<7{N2q)a7!A6vfmgqwC7BKV
zQZkA8H_GEA9zv?d&C*MzFioUyk2NKU`j<s)6Lti?r>*WlpYq>)9KWml?(gX7%~$8g
zDE|*AILhuPmwKL+%VE^wV$f*ZE!H?j#U$~cmxm)~rfq(19;^<weiyGWAO6f(+-ba@
z`xd|6UHoTabTj%~*%97ncQBh#fVSp8xfyy#41RqcMe9yp3?AYffOvTPH_xR>MPG_P
zt(#u&)n3f+8?&6g4|z?9u6%d_1rGdJFQ=i@CA9xMo%uS_@nUAo6pZS0V|MazzFFSh
zS6pcjwbZ%#J}FHNRk9iHjd^mX&=%I8<967#nv7VsvXC7F=K{?OIoge@X^%Sd$QECU
zEH;^OoasB$wrAb@4B-0?6@OAyy=6`YmeDa}4$h|4kdNT#dEA$-^*%Q?1OH}6^Sj8?
zrSV_LlO5Xr{rI>{dwv)U>*eSEaa#51=HTW0wEtJdpXvP#yr!49oJk|9mNy-*PjbhC
zA^p6rU*}I6quuE2$-(zLO5W4p%5?tBhx-v%ww<;XGomL~n)J_1S-%Z;w9ge@e--E(
z0nzfVUr2CFSCZ}2ZH5$N?TZb!Cok6<%O`WgdW|C*%iC_;)Wpu8=ahN)BBXV%>F*i2
zvczLIH{R3N`0MMohabna_^ln^-w$@V`<puSx@X({w~g2;cmy-}CSc=#C@{kY7!WG%
z(GY}4yvvSrybc=Pf4%rTn9p-N@n^(i9P)S!ZmE+gJs?dY+0#%X*&EI_$}!O#q!gN}
z8JPwMn7S{#1Zjh#vdT$4aF;62xh$8xL$n$Js5SjlX!@kw1&dM3TMASKfOKISlYLIm
zUcFnri5RAY?OLFp&ll*}2FgrXpneeP#{dfj&LI>dhMYX(yZpLey`Rfois-ki2EPY)
zXWrB1<{sJ8r))pTP46u|45|1Cpo<Fm1kquFX$>eUir~Y{^!vour4&H)1rG%JGDODj
zHSyzH2<1m^JlF$09qvDezhjGq$F*0|5L_)Md%PvK1`zHmv_l40`$Q*5G5W65_a+wn
zLA8>hf@&H^B%<$K(%EY;B2+e!x1PQerXbBbXU4`D?(S&?Y#=*r{Wp-dY@jUT7iATy
zOnel&kVI3wfJSDH7^rto;A>WvIJ_uo){5SQu}9p%gJp+H1~KwChXDSVPa68)0PH#i
z2WT-Ghf1kIysMoPU@4yj=&gWQP%M)OQOr^)_MQ~{Lb;3IkYhnfzQi{e)!*q2e56^$
ztnH-|)PLG%OuFX_8=$o2+HLxBR{~uk{hN`b0Iq;-DhMnJk)zAylhKXbkPZkGSRu?Q
zs0|j%c!5Mg$(n)SdsfGk@r>k4Lc(}bwJ1@>a?u$Ph)b@0z?6xTYlS5KE1HeE1^5_D
zjgMUVM2?V>M99u+0E%N0#_KaRhY#EZE9{I0k-*5YoiUKe^cnWK+O{&#2luj?2#o|6
zaMjuhRwkE31=VmX1<<<wTcE6k7Yv7i^(<QW)cx=%kjb@EJ#5p~Cz#YlfvvfSY<DYI
zC@B7kVb8K9rNI|cjV|L<(i?#A(N|lJGm*_wTFMjJdC)24>ao-${UJ^;l*I!n>zRh!
zdqAAJ4|ILft)?U*{uuL)O1k^T9KI)+`yBQRBp(9hxb2;wjqzrHv<WSaKCNbb@O?IY
zpo~*ME%aeB1!`7{*FX;-o&a+Pd;MA~N2D6S3s!{%1pbUv;pGa1DN6op`&b(a?Yx!W
ze!#Hs_jp#Ac7QAvlxp|g&n>=B*%~79C%Ao5Ak`?zz@VYPR%oyB8n<2Uv5f=K35X0+
zWqtmNo`l=Gf=kVv=z?<&BBZdO3Xy!R+7Mt))Y+89DoNWM1T#|89^?&KN$x}Y(sUX|
zmtb2H!|Z|@Omy*u=o~5iJOJ#C{m}mG@CLmH!t;?@-6j>GEvwxTlUTHsmqjNgd3`aL
zZj0^<9;1HiZ;o3dZ;dpZyF~3}(uRn~i$oJ;qRH|uf>T*nt{D&Z;Gz}L!Hp~;>Rqi?
zFSwwXq=6e^&vyXv<-a9fSSs#fLIrWjSe6r(HD3nl5v&MT7`cf1n4D&)J{;&Ps7=}$
zm#L`6LJS3HKi9mjoAumn4oWe>VOLVhT8B}|JlaIX3i23%z!fG-g7L}u-4HqwZUx{d
zahPHVE}?1h{(WvVhvHawAH{lOc?$9=)MQltfoE!Dy|V8n=k~VjuNSPD6l2O=7NF{+
zcB!sB0%)mM<EZExbO^S`2eqG*_8g1bxFN@YX}tWA3{bj8B!JA0@3tGZTA?A@E407w
z16+c48|5AdG#7*UFCCU7N#kt3?E8%{A6zC?y|FZ?!%ZJmVZ+>F{=fsoD;k@cI#T6!
zh)(f;2mvFpTO2ZLkMPo{n4BxpwZA=n#$e{8!(V==TXvUcT7JIg*G1P$R~Wm87iDe#
z)ryc=mE=|3kjFch3GK>Wt6LB>y!pCG%bdDFr+ur`eQOK5IhC#76r$M_p_OPi^mf{?
z$Y*|s_pd@nVBP*JDgjLMRk`{>H;+k3ZWNDf9+7}<(V?qr|AwZM!XEk$K#%jc`{sY`
z{bS_h{O^Ie|LNVex1*h#k1Q+)XMkWBmw|d*F|-wx7u{UM>-_mqxUkT4#UaNthL<z7
zjw*irln{mcfpdI4K7al9cpU9n|6fd!_a`#|fBYw8q8%32Z<9U5;OOdp9QrSlbUBum
zij+V<>Bs$DI>;P&$nJj&fRkGjdiQPRYYVThd;jrYOcG6X!+vdq-P)a7u?%<=J+X_)
zo2*p}<>U7LY@A-6@y>tAKm3F0Kf52Ji+=K7r7)C$mamtW%I$TCAY7~zzGoJ%&o6hg
zi*~&86H@nykAjY`&$n~iZ@gB^(kqp}3Q(bHe6=Mn#vy2^xMbD4tki+e1+!i??lhOe
z-1WT12>^${9xA5%4Z|V58F-Wq>))s)Csk|y`v~m$F-_%E<H}>HX|Da{=uG3fy^utd
zx@H5HLZaC)GvXk6koO_ivinC?AMu^;$eoEz%cpCSlsUn}W8w=V$I}-@!(&Aa{>9G3
zoX{v(ljXZ3xuv_tKZ;ZVNM3M)&-GpAn-y&I>rerd6zLW($!@Y7(fEoZ<2k0Q%d2CM
z#^iLpuFFtX4XP#NZK<r$K!U{bqQtugHqHm>D}+YX9|XJXiiK7g6B4v2N?8?a>MD;s
zX`Dm6Y1Kt8<C^b@-v4^285iaxtPXM+>KM-}m-HTV3!luq6gQypd0IB5Q|wfcC(Hp|
zOCiV5dRV0s2?Nf0jWS8fKs3MhSk@i3Gsw6tu@{aCnNRk3S}7Z0t8gI$Y0}&=qUNc7
zWXBD76(DN}t}PQ)7!{=V(s3Ty!Y(m5NO1<Yjg~)36o7(+yv{oQKs+sj9JUBtn<ekS
zOK^68e9c*HOG`ntrlc>W^rZ<k6g_ulpvOn`%jX8|KBC;f`d;TSfb25A1_V2u%=hyj
zce1~hExCgIpy2;;J<6(2Ansqa-pvb(fqK8-*b0eRW5#HE(b`UUZ&&g(VK3>OZXaWO
zYyeAB%3}*9?S~Xo#Ei>OB%dFolB4F3Tnwiy#Lt{wldu{URCtM~1CK;mRfeKiuS>2|
zSF<-7raDs_mf&1<hJa*OH8wpBr|CgXwCrJy%6ca$Sq&WAB6Q$1MRB!#(lnvSOrs5r
zdi0TIN6!@Us_NNm?75L2kIV_n+T$a25~i8!6LxG=6EpXcmg`Nm<Bo(@>t!thtb&Zb
zpmf8unOe%U0(rDMR7&N$nPAOqLK~G%5m(1ZYr$|QN|o|QXHkUp3<Jr&4yR&K=sk~)
zj7!Bzqs8L=R56D7hvHOt!NsEaj3l6-iEaqK<0>7a!$0<6nNhWEc$7&Bwl$jv*gc<y
zTSQgnl^CpGQhdC?0j*kyczbM+&h!xigRvt!;;hI>R#qQbm;^E(5Bg{F^Ggk!#zdTR
zIHty_tBPR&x>?8}x@*9+oBZKL8yEg&ryyZ~O7%Kk*@1!kE#s)ff0DF9xMA$3QxYA~
z_Pmtbg7q?S7%&xJ(XzKE)w<&aO^&bg<GTF@qz|{{5+MDVR1*d^5)a&dbha&$WGBO`
zS+!K>Max9WCP-DTtGci~=()7?ZlTn>P&HY!U-3u3%2~%vtE8y4pQ>Bh`)|*&G69Qr
z(II%};w&vngww+mRMUYUwCFEm=!9uT+EcqHQ<RhA5eSnzS+$cU@G|1iDnyLZZ@6h%
zcK0`Wft}41YGBY|F6)a;b!V`tv3Ch?@P|{AAhl{XHsdet>=WDhgeSk;9@7^r0QE09
z=L*;zi!?f6t<*A|?0_>ap3CP{zn$;rAr?Buj#>D^MIGa5(I%(f61PP}@yJW+p?f+T
zlbT1XB<hY4mZn$?)HU^o$P#wIql>^$&AS+Q{?1CsDrms}=3L}8cslv;fuyb$H&O;x
zco)vo2AFPg`#IPi=lKHc1QmP<G&OgjvVpj+ncU~IwU*Zn&fa}R>MzNf)F`6Zie?~n
z_Lj9?EvJ<ffP-S_aDKK7vs}h5hsWHth1#;*c`Fn^?L!PkeTTkCZyh+k5W)oHg69D6
zD5M8x9~_0A?R^kKxZT}Rs8PpB{^fwR9=Ab;mAL0g)#bgz>236Rcyz6iol6e)DA#_z
z`mEf#(Be5h(<0yui#a~5Id?4}mrZsptq_G-$jG{GtuTdmoX3$VM96-y>a;82EjW)B
zWK}^~J9x6i;PJ#1#SMT72uzSE$TSlku_1nTS6cS>*rDt4CIe^-v+`eB>4Z4;&NgX1
z_t7|s_w79TSTsc}L6)S)SYBd<@Z$B{-CH(NDiq`M$gUVWr^*4$kHwn;I09%eUuV+Z
zBaAYO*0nuq?ZXxM27$_c?;bl|RY>6qnZYub_H4=lywQgk071^1ekbVV84xR6ZaH%6
z%8w3Nlp%HmLYcCB2p@{UD%K9@aj2}gqE9DAq;YH#$T@{9#6k3Mj0r36U3@Z)=@LmM
z6YQ9?>V^TYTsAWnrDBjg-&r$b5)aAs1qZSK`DNFOk<0kS;&~H`($?|p+R{`V1)R)K
zCNjc_3p#8+PA5qHHESD;#kQ8r%gTV2Ti?&OvS8b0UEP9q*w!Re{$$ik9xN9-lLb0S
z9oumz^ah|Z-gw`;eRSB$@$Cvv4(&Kx{aJ5LMu+U`KfD<sT(5tO&0L$BNTL^HVgDGJ
zvE%T*>_)TxfCD+M;{WfUfQjMPO8dWp0w!j*|6V=*wYGHJQ7htaphv+Ln?Slr0u8^f
zGaw`P9Mmxq-6_h4$$AhR|G>!CCz3vbsAR<nJG=p0)-B4oN27X_9QM}t(*5P_{P^p<
zD@!HE0Ifg#=4%zs+<#GT&so{X3?MyquIFd|>jU{zXz@uX?LbGW6UL|QF(NGviDnw0
z6WIRE9Od<G)?!OXz)$Mo*I?^ARmb>U3&$M&gGX^VJSGq9Vqjj%C}sA3bHDh(pXaW0
z*oD2}#bsWEF@NDg02NtIH15yk`=XQZsI=-%|2g7L>5`DWVt@giomO3{$h@ba1^bVv
zTa<_WxB{HA5p)1oqM24^K`j;2P<wHwI%f5#77YkSS%FzdFzOF|eXJXkcAXp<{+dvy
zxX~R}TNLND+mxSot#3~c|JP|nE7ACHv%>bQAI}>gX@Dd(7eZ#i+;-ky=f8Yk3Q4kj
z{g`E}|KRU$W_LR;X3~znW0N4;SHg*nemm#m8%m5PI<0}9q_Ey#yG3bxT7kr7#xmH3
z0FA+O)%AHgAC8@N4pOTMJCLkug=$+~BocEgORke+=(2HwA%=Wp<&K^X+9t)Pd#Wbb
z<+-nO?p-JS0@aGR^}u|EsK|=VoA*+3*c4O(F>e!$BJ<%CmZlO0mr(RxltuIl*SY$}
zX%WMZg053R7NhpeH>YhuaHEu`ZO08rY0jEsh?qKOSh`%QdCT(4rf*1?Yk8Nwjo+{F
z=mEbBY+D~(l%Oa{gcmUY^;#cN0v$Fj{j6Gh^a@a#CYl{=xb2c<467bCZ)%dU4DY+W
zCc@T}bNX06>rnlYx9`aB@V6;nUl{e~1hJ&>@(CpSSx3gKqBDwuy>N+~!bf~Qx6(4t
zJ*7zpBR{DaMQz>w(R?X!;x6@Qda8n&Xh^+hR&qa%^H*M|Kes?aQuY8aeWQUc<|mt7
zB??G#`qQ|hnEtjaGeKWmC!(4Uhm){N98^u*X_o}&$Xv_forouyL=4=tr|%Bll6Z9d
zH89g5L6^xc20@2bo(3N0v~rTVRRPPk_}!8v3i`De8WW*QPp~qLR*iiyheE$Fn_z~7
zQy)@7<}v6P$A}>2CE*5H3m5+%jJ;!!Cc)M<+P0@{+qP}nwr$(C&1u`y?w+>IXWF*D
zKJUG8<3@bv$NQ%uswyiZvm&Z;@0Dw>Rrf5AaI~nbk?1$5FapGE0f+WgQXPkq9-gEV
z@G`b*QdhI|fC5uwIV$s&p0Nc`wgqU@A19BS$&_Z-;iQBRk%`D5*V@8giP!cB6i&iX
zI@3}kreMlPJu~+2Xe1i#BL3N;?QE3g*&^afPma+1+&KCsSqEw@YLj_RuDWQ}3f}8|
z#Z0)6j58t1a4B+?_Lv+dKzMeG4A?aJJ{d?@+Qc&H1OJKu^H&|GUF+A~dtM!#)8*co
zEacK|qG!#3FqkUSbvCvvUjgbkPs}!sF>53M4Pxs-M0vYQ(p~xsp97j^hZ4-4&=fbd
z?k<jrq)VqzJCZGIl8(eJ=4XU#+S?B<Fg0?!Ds+k(=00_-A#_T;@y@C(epaalY-Meq
z5OG$jwiK6zUCkmm8W=(77X6n&^WCtt{=geWSwrdbOj56I#HqBt*b_uqLxneXX&LOA
z#O*OXH)tu(@+tKHJ47zk{U{%wovk}WC~`-m&Sh>72e?>j37?T0$6%w$`;;ohJ3G_N
zz?nyeV~=^?;g2&8nztHV3x+RcYtiVC*aGDC<+{XUND_-mqU{II?{znrAl7V8paYBU
zj8r=gJN>Ia&$|m!cm&1z==#_K6Rd#H*sxIc7$QI94jF-onyjwqsP5R+>1CAm<bRbY
zjlOdp%oRRpDg{-C`fM`&Iq<E$Ae=%w-&k6pTVbZk$XzezCT9cpk`LoSKm^;7wq?9-
zKJas(JkZF!CVqxorA%_oO%BJF3MGkJ$`yr;qsgD^NUoclV^D9`UdpA~94q|FIO{KC
ziD3e%x3VZpot`W@_U$t)G@TM>Z2f&|mPeQoWoh2kEEXb&W*0Q`_U|><DlYpJP)q=;
zMAm>ISFtk1+%(t-p{6<E`A#3zLKGrUEL5g41!ilb$O5^;QCY%s>?8$mH8VnXZ-n+?
zJ`~8na*Dg6>g?4`J6nt9yImzF(<!0awiM8Or&4z&Y7z$Ohm^P*=mbBq3W{oyVxY2=
zGZ<yQS~-N~xnd)NDxW1>S+xU9(l(5kPGw__u)OKV)}lDGGY?7zRH%M&2~@n7NhIA$
z%o4sF?PS>o+t3{p>V3FRn2DtGt&PJ|Cue(JGqKbD0<A<M#x`53VB0yaa-Ilymr^0j
zy9V}>Z)B$Uox2f^dUnQzMTIc&XvXjMNV0l+D7#ED{j@lTOzDGRhMszs7oz<ZN8yjI
z*7l`)Hp=chw^ZStbkWK|28x2CTr54I&UlHRoHU{9>bp0X=rr<3I^I#-h0MGYAj}e~
zyLq?OlP|OvJ%#McauS%mhpfs4GQu{J#eu9cS~)ZNOtz2;X}!Y|we$y2wS2)B#cCG4
z@rq!&5c<B7J#DINffaTno`Ppdy9(FXI58+o?T1KHI6t~0mO~N008Lm%sHGS)$grpr
z*CuE=*>3mttWzq}>}-ev-R3);LmjMB-uVHc$`4ivCQ%ovT%rBWl?r>Q@(4!vCD7BV
z%)KvNXy0OrhpUMlK$bC|e7~y=T$9zNNt%kSct7ax`HM6`-=_oqP08ow5sh&?cCJEF
zgdg~>Lv(2p%T7LJJb!LJ5EF8`m2r<wOqeMc<hmcR8Jf1`RVj?FyX>ns`{QqXTOLVV
zl?=XGptf?A5U{lMsNBOi;<vQLpSAci5t7ReOV+;h-z}od@TbfAdscDj9;;BLY^S=P
zeFNctp)zliuZfzGQ#KSM%%p;y8~8t-`%G1_sGPwFs)9BfO<OKibaBgkdCc!fj?-)7
zv$T&2lJ5s9G}BtMo_cPs8Ag}#+uIpPH!)G&bpHKH9gcVi8=yMi|4=KAK9}+{5_9Wh
zn5I~og=sc$X~4v9#=XK&IaJy7#F5X_$;_UF{<O8W!kz>?HCH)Rb>RMp++xTstk+QW
z;TK}+6pN3NyR3XeXuNO5yVJG`vt4gCy2m<4q`D8D4a0IDG0Ok7ZtewlG0orCGUC^K
zeKq6t7wMc!Fu})IJoGl&qK(}Qb82V&>S6q9D%^E5@^kDkmVbrQsH4n0Q;_EV24z)8
zMEGxU7}I}=!<ae#chuY;wPF{XPUOCOjp=!AC~=M*wP@;M+b(L3Zr9DD#7bj;VMSAp
z$ld)tFsLDrBc(E_#=5B5QX-#V|J=(QyrBFpIhWstZ2|kWhqQvcr$~H7MZ>TCus;FG
zgxV}z#71ax3IgQ?SEZ&gv*-kLWy&v{zR#Z7?>b`mbWo_p?>ZG!Z2{xg+kETYTVKcu
z(bvWV7**`8j<-GswpV#wIvvd-GJqDk4}z=jhlhx`Y4e_4ECDa0;AZO?i1QYp2ICN8
zs@6dODF0qP(RN@>xIVckcIl~Q%qT+$dU&`x?_#W+H|~8Ao4BT)A7)*ga~#c+4c;pn
zVz$ad$>E!tKVla4uD!)9H2cmyY1rr(O-Z@I-jv^2#y>p;n-wM=qbP}U%{)^f0sr$a
zlh~YR>r_F4Cip>dh}&DMm&aSH`+IX-2s<rH-hu$z1K2uAN=+Gjg?#_l^9Q$kucK_X
zRN>;OkJ<5c4ZLHSd|9C(B&#AP3@9)UY(A|rututI(dR$=i{0yNOmLYHtE7NymnD?N
zg;?DE`}J3P9{&uzvXcB`1Lgh_!OzKcZJ~6DzqCBP);4fSZmezUKHh=9^NL8cvki@K
zDu!wvO~_-7G09Ns&c<v*5BX*#pct#6rLzjS4CP39&9@oh(*vY_$oX&eu`DdUmNnN8
zpLJ=t{l~}&UrN-<R?9khbQV@XYvx)+2_h%BL8B*@EU%q9dxN~_dC4BKENLF=EKO?*
z!;{Q(4D;F`!7U9Xs7<5hSp<3X>{2t>DP5X<w?dAbxPO+phI6!_>1#3yH0m);lO&3d
z-4ZU3rMVeY`-v&W*1Rwm68}8z?WZy}preJZysqo0iLmRSLxwbmRckGq`TJLD4UVj6
zQ>f~qr?;|r`D>^r|Am=EALoNb0Gg$_`NZtNu?ro&cXfme<@rs)ri7YW>YmClh7l-a
zt?g%!%MNL{6*J(oMQobGMrmL^^kfS}z*f;KcA}=6h8}C?yT=_kw6Fntxt-q&5a|o@
z#Y+$K^!a=tMj!2Lk8jXFE4y8G;UVh7pG{~Gx~Cb@n0?grg@4jVGC*;|TrO*TIgY+9
z(?fE=)61J?p^liGCFyl30(kItrRuK8yuPV+#yV4^O?!)4qEkk_w*d+(gp4^j<^6Rr
ztma1H<e3sM!iw7P%^gi_Cm`C7aayQam?T%FYfMI()g8A$YMHhNgJ+#ytbd5LO1v10
ztv6|}pYCLz^ESs#GRM!Kb_f0~U<!4g$x5?U8apVQimW<;i<T+X-hCZ;1An1F+eEHi
zri{>16Sd=x8LPb!PaKpGY<Snh+DCKs&)5_z0G`BGM*}@z9^Rc}i4_ihz)x@|_(BCj
z7iKRd!jU;_t04rkalKJOnrC|1m;wrxFILiZgM#1=<pFY0v5CV3Yg{#Wnowr7?eozy
z?*YE=Z|?vB62c0H?^rpe?mbQz|2L*@R&mc$-^@P1u6|-8ZU|=~dsNzQ!J{IfQ%&gE
zMRAk(Wahpi|AeK$%DHJd0BZ#vwmL9;)juRIb?c<(UNZDLbmWQhYVi-W&ZE9tink2?
zTZ@mD{vWmfx6PJUZaaR925g(f(5HMRjl+v4?FKJ2PH6-)$HR4DWqB}+7CL=s1esA_
zYa0i$rU69|Y&MhDl+wcvQ>*0@VXF&z(i>`Am?~b0H(&v{wqWHDR2xA-0UBE<CQ&LM
zkEffv&sT}7XZT)zw%M#%x<=tslM0>GCB__*hS2rV2Kup1FDt$D`ru?2eBIL8nff5$
zKYsbxZsy(g7{mMp7jVQzf$(E_Ovb8*V57E%8hMU{6jbbl>A8Sjx=g+1oVFv7)iqlR
zP$Y|fir46z%j7T!tHea~rk+5x`f7Ba@!;w0--D^l#cwA?>t<VO)E=80?D$WAaCzJ4
z&5o0oPrJ=B8k4HtWEpR1)^(SK3S0}B3AC)}Bjnvk>Nbk(d}K|e{W~mG?gaaBr&yo(
zQ7?56%B^!KW|7Re$fthl;o&z*N?L5D*bcWh@oOBozth^L%(cqc_4PL2XO!Pb%A9f2
ze5I{r+)Oj(Ph0toL>I2?3%=?GCzzd9_e_JrvE;w_A7}2812$`4!NCkT6-^z&Dvkvi
zws8Dbb&^MyA2#c_HyZS?D#i}dQkuT0anLf$%iKLXI3}K*!0K>GmKU~z<y>4vsWGaj
zv}Z1DeOj_E89U8N?6T3)T!gt-RVEu!;+a#}+2U3DZ*xgqHXB+ea~zAGEv9s0>DfxL
zdeJ!M7Im6EiHbv|OtxP>%aj&ZpkjQjFxg%mOEv!NHL8ToY0Zq84N}EhLDmidE((#C
zY^j`F(t9w>M`837em^;u;RShJowtRb=;61&Zku@EZsW5l#t(XdoxJ&Rp(C1+0;WLW
zzffIVh0>_UR+7&?5&A@=EwH34x);Dna$$a@U=XR|{OY$njM&<|({?DMJ;Ppp^ndnE
z>8_k0?pErn;Y|~#2cd_Xo~*pRIz161+JJsI3;TFK8QBI_6-PH?l!?12*i0(nbzsTH
zN1dcKj^1GOd{zwQZoh*2p_dEqeqKdd@h^wzW%O^gFjXc-%2+5FKng89ujn+_M~aO6
zd!r9XwbfTA=rSBR_nF9PZd0ihH>!T7#~6@?{p-+}Cro+gDK*pJ*loDyiDdT5&4I|s
zAW|FWT?J{6|L0{sd~~J%WG2Mx-2}<=-vRKX4aqaau|}V#ATkFK$ailHB$%h_Ox4c!
z=m@AVdQ~?3di&2kNrBd;{@WZ#(j*rS3hVD(4nz>xX55a?a0YzNz(+KpeltiQFScl#
z+On?Iv$<oY!_ct%&GNHDERL0fzs2WkiTv-S83Ph2z4+a)|GG_{J^6XA15tZ@P6uv9
z35|NMbNY6Ap}d8R92@w)3H3(#)ZUYU8;m978R4(g0L@d+;5_b|pc=4;?zhKRGZ9>P
zI4*s_8|Bga+F#>VvSakQvj|$uOIRl>`n+@$`2&Wkvv@-jt%rY$r*1^czX*J%Y}fu<
zxWxQ_)1sNUIJy2;wCH8+&iDs*l)nI~qiNWS8;c}0-asMHMHkxNS#cBSI9P2YkWcTB
zIz3d<XEIHV2usP7)%O0i)#<tZc}EWo75A%)?Fw-zXz>Hpr{@{aFF{8JiM*DUO+Z7I
znSK5_U-!g6g-p4xf>f<1!*{KREP45CocX}KAdcQ`P)7gR%2N2*%KG5o7x6^j`F}{!
zx<91ornfSs|CbakQ1JEscukSDoH&%@h$VlxJ+``dtsqktdT0N3+FrpJj~(i6&Ab2&
z>;85KIL)Xa5U)<dIIa5pxIJf${m7g-3!I^jBg)P~Bi+QNFzK1wS7N654cvn`nU^02
z7jCjCZK{G)BlBm~meqK8QG%IJ$Qyz`D->GNlVI9eZ;2cOI?<o`DAYVGM%U9Dg3PJI
z_fUqSgk#G;-4fh-Q+EEP!@UY;>fS$n1FEIIsg$e0_mpS5MAxM)*{Yl%IqK3jblv~`
z;m#Nd2&TGi)Z>W~lE<Ii0rHwAlW9AaI<-gAid6E{EJ&#I?iX2J+yjI#1CYc-YOa(L
zMG9ItJa0j*z!kt+(9;T1_@?XcGA6J%e9AH4Q<o0L#OTO+rSpd(TnyULg>+dG7Gb7J
z;R9g+5brnl*g|{ZIw(aKuf}&L1d@a1;aBNX08-f>2WXLtBuoE(pcqE>BKvWjP4tW`
zh9w_?_Ye@Bmo`GP>$ro;C3GE<@*z}=MV%RzWK!$`K^wln(_~US*gi$kglqOF92tLR
z*2Nv*_yHB4nGG)hQ|CH?1{+!GuXpTJiRdfqL8_%(YQt?}o8KO#%Ho^=j)PmGXy2zd
z$wFfaFW4MGHvzE+O47=>Jx-9fIcg%lZD2n*$MlJid_u-(T?-Nc{fM!kngi2Z-@jv3
zPU6&qp2%Pj)_1v8bzoH0s4M&UcCa4$16(~K@HarV)Cb8HM3cnQ*F=&RLt!E4G^hB{
z`kHsJ(e8LovQT`zP)J_=R+%7Ft&ZHZ?#<nm+$g@ZE%WCaoDAvn^t<@s3P*M*EnMjQ
zk0WeIlTRP@0&wbCJwQ4fsi>b-gd&{<T~ES6I*Ju0l{8~Ee;F!sOe#aryoEOF=XMf?
zHdHPWKUyhy9pNc)OO%f@+4MZXTp~zE?x?W=qSl37h;eBAfq$S+#JP4j4C774JDk|_
z2m-lCXd6m{P((8~Ww+8+hu+M<6h$f`j5){B4qKE?L#F0)j7{U!k+)a)@LX%1Y2>Br
zvRjrx<VH^WG$M@1Ag;K#pQ$XCLQuvi<D^XHA(hTlwhkJg^d-@SWON+$Vny!|q=Z@6
zcTApy#XsgU<L>=PwMbzas0x6Hb}S|i?+?NgSSb=?$VZtV6G+=@?L0%$PZVt;EHY&e
zv#Yy87u38K=K~F=ES)}buia1zd-!Y<cKMTFDgVyVb#g(m7-DCFZ$E#3NO-0h3_fab
zCJ^!j9CzPMn$hPqIvg+bCwFXt!fe*vwai<|>p{0G2igX_BG)fD|C%ki^s>}PWSJS2
zZO^oN&#KnPo6A_}IZbl$!|>hfExL|Q)yXb!Sg{CbtV+WWf9)IbFX8_(|K6+4Yjfr9
zW2h)YL*HwVk2%pbajJGE{^H+7CxLjs1X;F{um6;EGLS~y=NfjzR!eZhl+k0Ce(owE
zH^H&dRP^mvO@U7l$-JzbN(IoyF4Y;fq3pP_Xxh=?Y-ln|Fd&*KamGQm)j>utU4Aef
zN&h=G8<5^co~Ejyy6qVr5qH^GNEWf~RGLnZ-g(fDiFCn^`%Z(*$ix~Fa<u1Pr5Lnx
z`ZSVp=PdZQ>jV3){ExG=lsn(ae_Nfd&wRpGuaV4V0!vn6n>=c0Q8EdIV%mc9n%1!y
zPYN2b?fb5f`=icAxQF~z+X+b*UzcuId)y+gs(rt0+lF3eaopM?tKXn<%bGJsy|VON
zx=u#iq$BIIO$aI9x(%uCPoeH##Ws~`y-)2Px?7)qUTv7`Pw6M&b(;pMCUmDc;<}8r
zhi>s$dOa&44W%9rcD1oJRi>`}T^;b#Fjq$v&5mj@y|q^2Hms^uX|Bb7rgpi_)&*i(
zMG{a{za{WRnPi#|(N4(B$LHm#vMa$N;N>flyvd@gm@dyhPuy|plF!|rpgYit)(;AD
zyVdG$RXZeljS7Wp(YY~9ioRH0dSLM|rU3j6Rm$u~|87p2fhGRLzfJYFX+`Tu(f8=^
zZ4CQ0rR8Q|jTiLke9tW6I;@!vw?gITq&;f<2{spD=5mABMc5mE`owmd2b_-c^5@GK
z{?!BCW@{S*PO-(vz6Z^*Xy1I(;{<lQ{mDjOrp{^2mCL(#=f6qlT0DMXqQJZhNNBny
zGI+Iz`}XZep6*C=Hnd72e9?PQ&4qSuXExU$U_bCv(pMwH9CryAW?Qp2`_Vv4p}(_n
zWWG7|h_nW~ze&1H!icHMboN1Vlp^il^Lkt5O}*V3BT(82&UoasUf>~0NeI&Do3Y`|
zW6TlC#A#scbvzCNZw7$`*|ql>8CjG*)y-oz&bID(I}p%a#WZhE-g5b(C?nakt%oOz
zbc8x>^Q#aK#T9|VdUEzSyTzw1K%&>X1g~zgb+YTdc-7_0>0Q`-W{sU|*??>1ex~@^
zsMf_TmMOGP{d4?!@qQy9O4GjnuOY$zkH;))W^dtYNyx(ZzpRvM)sb?b`9V;Ad;~H*
zJaMmJ4KWi<&xu^7j<z9Oq?v^Rdqhr>KR!@1h+|}A=+^ni>D*FAHeQy8*abfB2M~Aj
z<$v_xx+wh|r*Za=_TR4|fe|@B&~WXwT^#=-?Y#e<Fx)VkD+62b%a;`(JvcV#fv9nR
zrv{Bk0=9*Ga9p!U950X#xDs0fFn_^mAQy~q(eJccg~sguGP^%maw(z5i*=RPDGl&{
zy?mcl=!v{t9C;Bj)Rwl_VAia_niIxij<|ihL!<BcU1daccqaiG1Cf6^UL0?XrIQcj
zGn#tMDALV+pm{PvLBZXi+~&!lgv`mqn}P<b=W^B&d9cHr#+FP5Z_1>0P@>pqng0j!
zj-;8^N%Tn4V2)4<T@GNilt`(eBCg1Hgyx&n1;}UdEc^&!`AH-lTZ5*C2@HoGQw;j}
zib9oPPr~rXQZb5e!4>Am-`FO^w#$^=f}DOrl)awolT{n^(A*cXJc1Z=`d5#I-kL6n
z?$>H1of9%F1Jf<djRUA>-gRk=fLr5ZhRKjEp&4h7jg33@8F&}@hvqQu30lOPmW5XI
zc&<|pS3a+%g=s7`zATK_MSL)-IC36xFXwsb1@l`+J0AuP0C)2u#d?WS$mkw@lG{g^
zXR`!crOEQ|uOv_6*#ZOw6gklt(RbzRt~TMOo2q(e@q7+X-d`6;@FT9{136_HG@_GB
z#Fv5Zl{<Xwa8@7x=4BQ&b@dK+#%xMysYm7*S^3d?FM*YrF?hWN7b1iZ&;9ZUt#)87
z_2w0EF6Uom6+oZWHdXM5G3fI3Rq@lV#{?_NiY;ZHEpgH2Kn+T1pI^#!N{B5Imek@8
zMAQ<{$dVtyNAZ^|pNig0=bqchQ|fseT^8f$A`Qtx*A2^-@rW&_1b^ktd_D9tn8ED9
z!OD9r$}t6V;m^=!9&jtXu4dG9nBa!^q*$<iqR`TIpHZ*sGhYhNvNZwWcjO*q;Z=+?
zXlOv=8^>N(e7sf{hSQL{VS;Z5O9z^i=|s-&C2r|E6@0YV!x;cg!ZCW(y23%L&a>^D
z{VIMFMoS6d?j{FCvp!Eb7<kR_oiQO<qEY)<3t1M>7rhoOui7lo_arbJKFH$rdRYQJ
zHt5l$4=F=7l=*tpIAVDF+DIe+WdzC{ytdSF!D!<oS_T_a7Zy=9&0DUDWvAmr(u*7(
zU3F3pVmLiS$6q56$Vn%ah^5MDu>JjASS?a>7;}qvKy1(@@!JgFhh^hPDLi&z3C~IH
z6-I1;B}pL5{ZBH&JZuD2v3XwuH?niysnVbNCHhk9aRsyJEmifFn5!G_ec+#(TGK}a
z8k?;F>OH$-J{2{WjFC7CZsGvo3Knw^@7P6B@e*AsX>Gr$Xf_<F7HRbCd+h9PU|XuV
zo%FjxRm(fw(Mg#Z<5Ath6InN$j9fuSW4@?8jeA_3y0wGH#`z3BKQZUN_FS9N-Lbaa
zCjO(2U$bfVzp6+h)!#m7EzaUJ_~vBV*dL|p)VVv&4Ry2$?t;_*c>KjM1&DfD(75FQ
zA#ix*iviPD`T!vyV_0fkR>z=tQbp;j_MsOm`0nbEQx|19+bb>|Wk^X{MMm_~+0dF1
zWhpx^y+@@+2%%`m)*~}gm5GV80}In}Bqg88T*+!7E3#YKY~WhjOs#`*Q_tkfv<Dv`
z2}`9pNG>j#^VAi_7C3)f(@9(!9D{vY50g2fm*R5mpN!IC)-7?$80{T^L(sFnR4Iwz
z5b4p!pjC(EC{Gw>t$M(eb(jiuPf)6;wTayBlgA(-D36`U<JW~4i^MjQ$|a(mLb>|R
zLfmx2{dOF#S&oZ1`nyAw`PA5w{dv=aml$RO!s}s+T<%w2fTLXQSEs!h7CW#+N^+T*
z3*K8Xf%$On9A7&-$M@pSWB|L(E21H7G7*!TYk|dT=kEStDXYGb*@&E6FgGeY@Q0iS
zs@H|Do{h?qcG1Tq%RmRmTEqyC4hvnAnIbINcAmw`u(!ULf{yz4XE|TFQBHm(pRII}
zm38JXeU39$2hLCgTcxOTO&i*fNl)FdUCUCDuCqda0&}eP2He89JS`9vAS5I5^@afn
zsYy61rxUJ4ZZbgPs1Gz&Sa?DNDl-6Qal4+34hzwL;k`8^fp+3<j|4-m)T)>ESNMi+
zRtu0pVn?j+GP&<>C*dAn9RFUwfVbHTb#g{npyzf1h1($YM3f+TFd(vlbAoTqkFSS}
zB<uZ1mcJ$o=-&hFrSn&R8IA{N9UbxQ>rS<`g##DhITP5XBh)J#a>JKOYL-e0s;FRN
zW0E$ci)4%zk}@sGIcwA{#I(7|b<E0nt>Q63`pvnI&xA~8J5x{cy$DupjwRPuZ1j{5
zbT^qi&o&mqRf5?Q0NVa%*==t<1B?)Gi;o0gb~CBFYwY)?>6dREnXrZfg=)H+nmV5=
zRRH16!IZF~fX!z(+Vw=hvL?t2c;D}l*5b}H7g-9rO-DpEu*FL#^%>_l3tm440X8O?
zit_7_)xBP$L%t#mg_-iJQE`i%xhsXIT>Q#*ZYMnrf9{e=t8Py}xVKaYG#L)zf;;H+
zdSQd$R(L|{-{rPo>P%!W77K0Ns*-`2s6xNKNog;y630nfAR=)v5`D4gL!;yr1>)7q
zh|T4J{gl2+&ANG9wGK1GFu4>D!dVj%*uV>39Py<ES!VjSW+B1&>mHI3H+!r?5TV!3
zIzEB;NmC(|Gv)X#@7`CmA#{APK%G<qzziBsep$ybTqxSz?#o5H5DZ8tPKTA{q1fj)
zyv3l9U)5RRz}NY2*(OG6>rAz9we|(1r1=^b_Own-re8qTh=mjQQ1|r24Hqs+YX4#j
zIJ;7GS5(B4xlfO))`3jNxqyJ6O;{p!sHr9mH=Giqnp@svV~rsvMyVfA8Bj|+N69S%
zLLY1$?MV$ODiwAWN1-EowBkeD#Rx>ysyiWC_;}>-f8pW%;Bv<S#{bR<|EfH{8SY06
zh`R|0`3*@C7mAj$SE>PZA7TouOPXRw8Vdu*9R4nKaTF^<Ue_a!?n{%IbhV5`QRoU1
z8sxmbl&bk};h;k8w?h=zC4q?niBFpAuw+3_S1X?Bl!J&j)-AthahH>BgRG3*V1k%Y
zs=*|U-LZ+6+iekJvo8HE`!L1g1cEuq3~gSBMu>oI);W3xQ6Etu|JuO<%Xq5k#W;kJ
z!EBekm8!KmS2Zp1crN=KvJ0aD|2AABYV*ke$HLjL5<D}9+)GEbL*^VPav(rqy=8&M
zJgqc>VohM`?F><&s@|A8zkejZe~(VsQou8)rcIyx>iK$G6Z&dfJK_8~T#pbV>iXFC
zx1R9<q|cO8<kur=wYYYZg!}vPu<uTyK)$@pyVRY|{3DW*e=tb*aX{uQ&YehCY27>N
zD}3n!UAFt3!`_y?AMf-tyq{paZO1pMAJ9GhS{gsngvZo3V*$re#y~Yf0jWT!?k=_<
zr9hn8|8$~rCu4rmy|B|QbusloqUDx%p<XVKj>_|OEFy?Mut^hk@1wm9tZ}JrJSe&&
zowtixG8T5MnT;z}x3i438`b}0-VH_uRyhv|O+B9L@KkGd$^ei0#&3CkIFImSPAkR7
z$kosJ8sC?{EGxyQ7Vl&L4As}(Ktb3siC&&?Nk^y`&zqwAaw0XpE70_*@n*nBB4R?^
z#}hKfw2-<3(tAJEei@Qws5xh?TiCo+&w@o|s_LK+QE+wrBE?ACp~q{VDt~psyHlp}
zaNrM4I9=wMTiq(V7Ucl!3#(>5`$$v}NUGPtDCO&=>yZdbS>*06L$ILNKO}KlUqSuh
z)|*CuPkG7nfc_pRF0i)UD`O1l5sIhb+kiix;$1xf+3*jU$T0B>&pQ-X6uh|^=QHv9
zk;XUm`%!ho3GhXS$tU@V@Cgl%dij_BP2}xd<Aw6`QB@^n?s|2X@X>7j6XQQR+CF(U
zT-1W)R3CNs9jLW{2LS&(QB^QppMm4~QQ^f~P<MRDLRJ9hvZS8EG&@JjAvg=U?+~!G
z!pqyMl)%%kcm=`rr;xw$lvU3JxzqOThlFN5!t5UbS~nF+|1FbX`7cCyPS*b=C+<fk
zY4gXC?7Cr&0BV@X;sppD!ac<^Xlvt&ZoJCQ*$QE%ETDh5f^@RYbOR=u7;=+YEdGS=
zJ}G&4RS1Crakc@`Apskb1~*7;_&fvnCH%<YKj`A`A9N8@zaPE;zf3h;VD{rjZt^7I
z*k-KX)w5*q7H$tYa2O{B=|tf8{R!xVX*OW2AL%NLjXgG-b2$85G7CXnAbu(BIBUhY
z5a@&8VgLSo3^Hp3=qqgEI$|}VpFmq^+*?Q=zCXYRibFAEPX5-8O7V36y<j*a!hkwT
zSQ`eT{^k7`OwJ_EaS|fCfPZQqvF|mB!sa?W;9$w&k<0nT<|U5cxA+<~7PQSuSHGwx
zg0v@qWjzVZih=a)T*~%nn$l>?teP;aBHi|Z4f4Wf0kb%(p)rAKK{hAU;$VF>kS}P4
zB&io@-GWlL`7sJ|;)p5ji)41>a6c|-{@j*jk*vQM&@*}Toyu>-7Iuqx;llr2!OEeG
zusS9xq4Fk{PmusQ9y@8`_<n0*Y80M-CjzQ6Yoy9|f;>J1u6~|3<d{iNzwhV#6s2X3
zpbx)elZ6)Y3#w-CFS>@j<JL`O;=<}tll8yi#nRhirWteCn7Jb_h0|}&lC6;opxd|M
zmsBTqSqqb<p=9(^S-?Yz?Lo<wSm%83cC|hek+f0V29+gpSgExzt-W3<79KARiiNod
ztp-ztMM(x{4p&fJ`IA@r1aPr4ssUDKkFSrRTaCjA?IbM4WABV`Flj{kitw9uyT}Qw
z5i@+!jQ#oyAo^o3RAfuazYNUMP5R44%CMQg>s5bV;vY5cWho+p#_zXMerZ?xwEmpg
zsk;=8A%m-O`^18GUhuU^0l6cc<7<G6n19Abf!+}XfJ<+2=qp9JRqN{q!8roBFYPQn
z=*t=F6cCJSuP&andm4it5I+n_<I-3QNmT_+Um}t5qMh0H)g*JI<Z;MCr<K*Wf5@aY
z*SnUT(mpT0vtrX&)RxY({B^2d8_T_=1Ke4eq2=TaS89#umaEMdWYnO-3_ESrF0@+p
zeaIe_ht(Lr%ck*!J>OpiJK07<ITU-`%*xzfldg!4hV5jd+hAXjWEQ}&;Z*AgXMLt<
zx{h$@g-K~@*Dr9Hu;*qQN6|ttSG_%cp3lX8yfIG-t7BeWXID^hh3qK*-X`O?{V1bU
z%8g?jX_sJ7m^6t6x~j>hU@fOr;I2VC=fqKw5pyma_s}{Uhnx7Le&odN95q(#VyJKP
z2#05S#@pUFUdzSqD{@r2Z4c;Jv-|oD#lb<Hss}79Q^+cf#Si&8O1>Zm^Up<IfI1>o
zVO2(o#1=swTZ)W@u7xa$C%rLbIYkyNHYO^{?n)rc1E9V&uQx~Hfz$e4i+3WPL9dz{
zk5of3r*NDu_inmTTqt&SKuBWs=01^_Qg=>Mkbwgs`YU+`{&T1mmHn<vN)!Hl&kn9O
z#K3$CEEpCD(W$Eq61-mcwt3*ai?7muLW6ZuO(69zkIj;VZUQxLDSG^f_=5O+x0Wt4
zUW;R)2MS&-5Om!IpnZb<tT+dNrYDK^>yW=IQTJMr&Q*y6m}7Eh=Fn`t6>^{VlBD}d
z*CsE2l7IJLi`+*bfu@N^zhp+fiKl)U#=^b8ZqGioH)jlOy$%h3n@;V@8fEj$o6zP~
zV$e7a?!y*dL2tkjysnIG2*wPfkZanrI+4=>aZFSoEWRR$;<p3NDY~yMSpL?u_Ab4A
z<W~M4fTp)jH)fpOiz=LR%C;ld`n(Uv$#2b_gvGR2eN^6fF@gFgmE#dkM=lIpz(Xaj
zuH0q_jZ;%PWv+Ja1O(84){%N{Z5qn-2_CR0%7V2y#GOVcWOEW#xVw6xzN0h!$kHAq
z|6%CyXsO=zKp8RJbHqY70W3Ny22Xv?vnVXJDO%XG1S~aBPAixnSJKZ1Kw=>d+vHBr
zQy6R1dl2x8;X$UaSV?&(aI?ZlQFJIlBEL7VsNiS*Nw8?Id8Po6lldJS#)tc!V%V7p
zFA6!!nV>8CFxXz)`f<EW(KRXOvM|*z_1|-|htU^Bwq1EoFsP1nZSBTk{^+Zsh<fFH
zRe$tkj7`)mnvQNEjll8I!49K9jxKB~Jc+K<$8!BTuO(>U=q=;T``4aBgo@JmyEo*7
zfYkUcW_thL9#5XrHJkKQP@SC$G=a=4Q{tUn%|(NH(`OxSLSosaX6&9dj-idIY%qq*
zDci?ah?6kS-CUaX8kTTzFtfKlC?bNGlc=~y7j8KD;UYS?g{xW^$Z!wWV@x~C(B*gL
z9(3Z{Et!1AWs(aPQrmIB>>;EW0)r?*hgBnj2})vQaGbmg0x=pw+q;N__hkQJ|H})d
z%F|l{P>O}L5-_Ogg(-I3JZ|d5qhKxZ@Tys{GJ^K+Q_Z<pt^1dj@z_Hc^&U#|46{=6
zv--F8Zs?>Yoje<GinBZa*Y)@RM)H1S3K{=bn|mYLgUOrhaJ|<W{~(7@NCgvFczBx6
zO+Col+kB>ume@O4BTg1}87PxdMW>3%bQVKsb%=%blECl{&#rtby4xc%o+AwO>D+61
zfpe}tbs+DArzkAF{#BX)9n-pA3v@9gXyZ_<#^_eIwmLWZK08-bAforQ0C|DjJOm*d
zzXKlf3kLYgpC~#-d@<e~eb3&|ar>W>PopjmDYt*On2u5cA93t0u0GEqa%TZMM@)P!
zz7Go^wh|#o*1B#J2;d*UwiMQK<cWZzb`=D;?c)2>p){G+`F@ZWIItw?pHePxb-JLw
z9@kubrDhbAHNyA3w>!L%u&+{&0|j{W$N2Jr?}Pv~<dI|pGk7PPBjcyWH9+i13APus
zhzCW->VnXo0f$E=vn)<Plre1;S`7OWlkFll74%3t%RH6%j;Xxi7fd7*t_6<S_u_aK
z(vl|yz~PU>&@mhk>4bZRe0D$Dt~t~N13|7k<Ox5uewr?7uQd_4pRfPk#Wbts8uush
zfU>xs4G~x~{kG@Bc@j~Cxa`s`37$2^P~D_1F4J=VQsS+TngJHl2tH4oQ$TzDO9OLp
zp*&6MkosqaSB<Ibk)B#UEk!b%N&YDXeUib+0Cm&SOS+2h@$xXWt(a8`vso}hoE4Rv
zq}(7wLYB_EU08v9)al{Ys=>-O<H!Wl|Ij9BgJFy~C`C5|1fy(5TFZFXLQ6+JzR$xR
zK2P?$1%P5EYh{L?l7A;Ffu~h*>s|M*$eh9(?=Q{QC{7la=JU8xf&7RWebOeH13j7)
zU7JYS5v$9<BT7dSA)KeBm=IbJ7Y*IH7}>aosxh^GMo6M+?20#A^-Oi&z-t2<f@4l?
z!VAsN*V3TDI<3yis_u#pU_IwsG}1MDje5;o)o#Ml@Cd&x40dY0eCwQtO?&bT*W3x~
zjp4sWYjK9N=%qA*oZx$3ZlvQpZsj8I(Twv_dqgpKdu-%fMMvvIAs=g1In;^-m9=P%
z6&$XYv1_cYlfew!yA0LWI2?YX4e~f}z0_(*ph|@Pj*-{1cUW!^i%$S1vs}+U#PPzR
zbor1zV?)*)f|)1=Dj7(C6RjYdU56(vQ|d}fDA?^jn{aR_*<sH{Y&9*Zc_42^-ynBt
zu(q9DN`z!Sd`n$IK4?H&Lr9fBo6eVAo(Q*A98E>Dw7pa1yxq%mf3(qB1)M<`Wmzl~
z+qaZ9zgSchDdt=nLTCjA%dqLE(b%U*QbS=?k@;`YDcPc*Ojw|JAq$l#5eaQ;7ZO-_
zSkW1zA(lWXNX^j_n2U|;-0zbvc!iyNosgxs%G1s#G2_}3rRRw0R$)QkcJUl$oT4*?
z|CzMH$B5GPj&nYOH^s0D(=Ax&j=|wd*B7^@*p|@P8mDpxogcucYq}~-MVY6Hu%IH_
zd#qaW|HO9JA*nwsR7Ihq@<^8gv19aU_l9qVr?f2#=$>J(^rSUI@7{*!_Bdbtt@Bnm
z7IFbm!+BRfc?FMF^7+Pgkv<+IJY(;4p(Cb?HFabk+^sWW$QRI#n_+${I#2YBg`~jJ
zGIXdAmSF)vIH%3&`Hi5GurMBGY9NUZ;?|&4k(?J{rJn@)S0BM)ge!<EdU%t+lRJ1=
zBYBm}cliYf!9mFp_W9r@t!0?YmrWl|&M;ye&K30iDY}&>#UHdFLzlKA=4@feK&`3-
zn))WI=(*@gu%04*865-#2F{<Ly*)Uu8Pd_GMoRchZp*(p@|=}LXraLouAL)@j9g)o
zG-}g@I!4aRkNOTip5Jb>X&z34?(+tNy7ve~=S?mpc*)9#o=^uTx%iDuy<l|ewCTKH
zv5;3)J&(QYjjL)po#USS;BjK~D$7c?qdoefHwFHX>=t!h{`K}6XEFC*>}jB<xa}1*
ziSg7`dDNc03Jy|}(jQ2&KjBioN<oYn>M}bqyj1G5hB0K%V1(qqv$I-We@u!#k=V;e
zrEJ4OoM{df!Qo}Nj$J$Ta~v*KTZ4e1U5Z2QXtj!|3$;cJMd~{ZRv{pWcUw`)w=oT4
zo88eu;JpC;dN`PoD(2@1Lj%}kJ633*{lFpHUF?j3LBY5oK|%hBY;Z#maPH?D1A>T3
zub6P9B$n;bgW)PT?{0cygkL|2%?1pT@z;0_Qjad;puHLn)C-X(9%#@rjIU<cUJG#W
ze>&UyxfBB2%$yM-++_g&pH@g6MD9PWT#_LMs3CB^Jk}a)KtW-s!FtGWc4-f~Ktxfe
z&Q<r~!L=BjM$pI1?wvB%o#K6c!w6i|^AVIXA5pBhx|~lb!&GShsdDEM%wjI|-)$jQ
z5LL{+`Hs{RcsX@3ATY$W`Cc4>V&@=a$st_GB2d9A)JjE`vbN_g;U;|0;x=XDa%SNQ
zZRp58)#frmE+-3>Y`q3*JCk|!rt7lt7pCLZ#vWd)O6SA!^d$2V(Ep0ZbRyc)m8=il
z)s?6}x`n?pWdYq*8hjb6fNdSn0yCg@)oGY>S#Hz<)L%Dse}mGPGx-0v^_unn&X<{)
zSpRn#g_k<L$y@AjSJzakqhVITq=L@Q%`6@Op4ko==e$GesO>*uTcimDHwrXXWJ(A1
z=E|iV%k*L(dyq&#(D@gA?oIEG&oAE3S7JGTVImCGp5Mwq{fS<Gk#eB8qYEa6Up?wP
zpXy7CeLRl`i2ukG&i+UXEAaL%?{z>r!Va9)u_3()96djQxfZ+&-2NKsW*iMykLI&|
z!KXT$ID2PtGwZBqC07Lcg7<v;c)OK9u>uk>y7BezeLuJ7Mh2x!b=`K@y&8fo5wfh{
z78B>R>b?$eHMsIWlR{lk?zo8)5SX4Ll^LSKvGjkV?cRMombj|zJ^zZ=>wok$eYSXp
zWvN7V7ZxANR(iGxHLD2q;7{R85xfjz^%sbWOHM*LTB!;~#jh|espB%#NfC`C(J{eM
zsmx-Bnf=pT2do2<6UAx^9&QIQ?6Jb{OrmVBa@xM%cXIrYZrZ}lb9$Prw{L<k_R3M-
z&xkGZ&a#<1+t2hEmpc5MAu3_wjeSL7AfE`H#xQs>8$cSv_>!0fQG_p2(=|(%NP!`G
zxr-Vk@_mE1Lt#QtX=b|DJEX&=qQn{e8^*rY=@_g5#~w}jZKqQeO^;D765~&C(QoUZ
z_aMYbGY-^zSvXfOd(kGZNi@nh)bu5xm@!PE#)t*fd+J0n5tem0i5sj%P${g++Cbxb
zlCZS}p8C|en5M?+o?`cUV*I}C@f@PsN^vg)(6^Opp&(oxNgE<R0+A4mh1m>qn<H_!
zY_mE;AbCy`gdssHVC@;>o79ZTogks+S~NIeQx8uhh#u9!Gc~vigrAmrV?cb+{rStf
zK|!g)L0ThwPCO8oP3nw}RA^japsx9P?DWQl$rwAa|MR$Q^xrW<LoU#pmgqr*f;D;@
zaL{O_M@*=3n7VCZu#Q<W_84s;Ae9?#2oTbuI)gC72npMa*5Dwi!E;PFIG7T3qP0V$
z9I-mVLY0f3Hg=&zntuY_x&N4J(O_1hcH4`!XFQL(`goZ|vyA<NOBHO+o#wr2*)-Q$
zWn$zX$ueO~B%Gt8U{QF#XwzGPfuIM=4ng7G8H85uh#rg=-1*OVa)@V`Ff|z>=~J+H
z9c-UO@7#a-R*YpG6BLx!^sg8CO_I#<KZ7jff&gh5!fuBdL`e9bnUrX4l}ryLW=;TL
zz<4~^FN@U)=cxYk-xA+vtu_8VM#Mg)Jt$bV<^~5b31zWN48haR{-0L`L@{^BW|qE9
z+NPEcqSc!ZL#vivPF7n<@2Z!$n4C{@XN7^#^ZrsMffl)_V<&9HfP0NC@R3H1-O2Fe
zAZ=6S$SX=MG~a4mLOwGX;2VxNKJ?ki80hV|p!>JwzO#kB8Y((OSo-A(vSK&HSr|LX
zS~yHa0GOp%TfEaB7451w)=ZTb>-rrHx<ER{o8aHk;!TAsNWcjm*C{N_lFY*J&b=O2
z4T0go?1Q&Is9HpI-EGlmb^hd<GFRlG-r4NEo=Ulq2@Jnst?a8JRw-9Vw)TKvxR$BP
za!~<VJR?(<(Szy~c|=b%ErOhY@<<VI15U%qv_uwFnQ2MckX{Oe6d_MT6>H)J=^5&K
zSQ~X5%McA)8LC^2u$<m&3VX&~r;4xfAXh_+v09OKs(1W~Q0+^?ijZ}dHEgs=WHZ(r
z9vbKRE_lLt0%D3mGn(H|L6(csJOvf4Niyu)uOG|<^WD_cjavPjtE__g+!n`dTQPn6
z=HJ)h`>OrI-F-y!OSnl%!JLM(b4k%!SUi8gmX?4;xjKv6I6=~BrP+9@i;6HG5$Y0d
zPwgMCt8_ESYy|K1**vymuLRCvyo^pCq@%6JEiF5cp1ec(m@TwQL*_#!E`v@+G{cZo
z#+I}l58n3m$P+xmO+e^fa++xjwugNq$|3l#9OY=mbu%$!C<5FsQ`ML}(~z9~>-1Nb
z3m_k+Z!$9}lq=tzYMfM1eaGBvWo_m7CMLA}JG0@$gsFJ3rwpy~6VigFKg^Ng`y%F1
z5_Gh;-53lOez_9+Gb3=qO#BgH`?Bf0W?IB1#P`3tv`x)W9qRXmk1F^Hs-k0z9E{s0
zayBwj9UpC3uTwg1qu2i#s(afro2)i>AfRprx&TyVPJX=FU7Oae<R}|>>ZC0Eh{UzY
zd*@T+^g+F+CFn}bvMI1xioX)oPY3M{pWG*dJ0{rVD@Vw9Del*u?{+xstB(xHUA!jB
zb7BoSsU#(ihkGb=!C+>8&Z8H_Sihb)u+sM9Y%fu5{9<}@{)kVsn1S;DV@J183;y`#
zz{;^_ZnG&bb`l%f`dDMoGgF;((FR@BH>0&0<YFP{<D)PhOo-j#MYunG!^Dg)g5bYz
z>bz_kY-3VAY<`4m)O$rMIZ(8>4~YLUR`2M@>b+90_-5GsXFJwVj9~><9Rkx&9V*vB
z`6Hl*kg27VQ`hE&e*Q0T738!qnb9*kYZZmtlg4$!LP=<h>c<LkGhmPN@UbHjVH+;*
zL`!w9k#HPWRT^^YwY7~$LtTW|vOo+RFFI?NX)h6drz;P(20V8~O7%Arl@vEbuH!4>
z$6vQ)5czvr>wk&ATV)W>b;^D|_e%uOb!m8S8#d?5)PV3t)!NM%5`V%cWY<UJ0HB>N
zEQOx$J!LDQy>e6=!1?QdDR=e$9xGn?bfW=R`wUyqY306hrpVJaY}H?v9AcJth7w$N
z<;k~U0SD!}Xg~CtX={%8ka#tQsWU~Z^M~OET338gXMZGUV_b-k;Hw=#HRkE7P><<y
z<oqx<M%YmB#snEEEcyMbQ=VK$Rkb4pj@`lI3Q^AUxh-6cBK==;A)fG%xXw9*9;^*z
zucnd@3}+_d?Ee<Au>BVsOLkVa{}saUpQ`t*4&=V)hB+q}%<K`4q(Q>_8_czwluo)Q
zB+f>c8?f~mqSx2gmvqY3(v=4=DJ0M)ZB(sV>Xi@!rQrw%Py+rue0_O>{4#7>m+V?h
z{XT?H4BYQu5+(B?WX9qPH+S40u2Y?i9r%t^f6M-u-EQvJU)z&%JrK4aZEtIe6@B~p
zJ6w-%oqQ#9sCZ(GuFY8+BjWf=Ouqe{+!WW-d}vVzhY7V{_7Ld*@I26p&9^W?8(y1i
z?oYw?pmRKBt~lusl;Jy!_d@LeQPDl#-R~Xpi4(VT&XJsz4s#Macv&H%-sf;|n8ei3
zPL(QQTjsA<q|A+~Sb}*J9#dLysUgiZpHyLB>i<dk*LCGKRF^ZSxy_cr!U?LO{A(wP
zRh8s$aZrnzz2#IKRP~JGSc>v~fm*<1q-Z1b8#4jw`W>cd2HRTg+j<VyS=!y$Gq|+a
zG0B)pa+OLerb{@C1UvlB^7;F6{j)anmgZ>S4fGZE!;=4ymLhweKi-@dL^M>)@+~(~
za1U8KnY8445(}Ye!;8b7<W^0Mq^!;Bm*82HqY-QbC{?ATCkaboGfMKA4QBMpP>s)f
zq_eQk1p*Bcu<(o;>nDRU<*#S8#I92XQyc5IJdU7lSkNq4*&Sk^v*O@dWa-TgI4jWI
zlxKK`FJj{hr;8)LGU+spBUCq^cby!B$80KDLan4r1a*>sFIfM&V(ntkPC*ZMqVTbC
z&ZjDgW<!GsV`jyy_|dC3EF{hRVFP}gTJV2BGX#iQoNnJFtUb2M4@X=N69jkKBe^qb
zBAqC<{eY>yNYlgqDC^?FGP6i3^c0N=ESGq{NLMC?ZhEdx>$^yal%;NDWmhYU7!(L6
z4ni-zyHX)S$VQdO1<q}wsK+eUa&#YK6Q!Hu-Yr{Gv5senWZI{0tiPVU76z6?nC`6)
z!j>bVKx3N>;PdU&$WQcE-d~1nX*rujHFM7CJAj(ed0ooYHsS2*U6rV-oTj1X$>3mY
zupw7D8r2xuvRyTNE4&eFj6`}@826XoD_6bCk$vbj8uV2g^i)f`6_X;scN9zKWu-s)
zgQRfAViVO;sy!X%GpRSbd>y!zt3QbUZM+w5S8MHvrv?5BUMdTsT)VsCbkbD7d0UMJ
zR?*e@&bC63Sy?k8k;&-Q-cc5f2>6>G7bz1wRknHsa?eXU8t$EH8ok9+C4pt$nO`N*
zP$NO0IBUejzC;49a?Vr5Q3=TjV(3wh-&~LHBFGmpvZk0YK)>8)#^Xs9=7+N<s8;L1
zJZl#S;JLj2=g8maEJQo+a5BNjFq2bMbQhU)Q~{AfH!vC6b`XFZAg4|KGR0zLs!(^S
zn(l-y8yT7v=iTS)4{jSN+0WX~C(xhn^x37f44SU`w!a=ZSRxmS$;aSL1)6KB;`_E=
z&@a`BY6($eR(75)$P9lC#0^F(uQN<=74VJ<IS4Q>bwbR<3BQ>F3g?GUs!?GThNp?0
znX3%{B}gE;0`^4q1^$@F|8cpwy6gAjEO38wd0j+P3(Zwf7`*x+IqeM#-4Cz;C#>xS
z-+R&(HWoGL$gDb?ywBJ#4IQbrT*dB>5PvCF5pGGE_qY3}9+{9m`4OaVGn3*+z@Z+0
z2A8xyDe4KMdbf5Hd26QX3Ag&v={i!`4_DDJeCo1N$v!k%jiM0RJJ?Sahk_EG#=2r{
zh0?eY?x&_6ErGIVp*?ojRzJ;IvS++@qg@!R2sY8ImL)SKd9?th9VTp5UF&vf@h~?L
zoqi%A>w|4dh<Z<Ec1aH3zH6X=-(dw`GAE=hJh^P%F^GUfL_?l|IozT{%L=$E^H^^}
zAHb@Ko6mTM;(e-*hbsTzjn$NH+GgG7zgCBGU(%gJ2zv;!Dur^Eby7leDnHOS9q*M1
zieuhrEEC6-sYmQ=U6~`XcetkT<Xhee`l-fx?~VjhR*i`=(<c%)o;PehEVE@h*-Nvp
z$fSwbZT`Wp&Ki|YPciCckhz?zQS2v_!<+I9j`f^>3B7Uc5Io(?5^$I$rVn3R5LY^p
z^#&GpsPaRLvQC)DsBc}EO+@F2k<)ygGhU{%%*;?Oqkg*-8YYdSJLkLd+VjV(%T018
zN{kG{;Pf`~ZYKzqN*xz}lrPEISz+a2*HX1Qhrq!x>XZyVNP-ntIm;uPC%vL{il=+Y
z_o_nlzcID^)0(JO?pClz8?;F$fBdO~k|zaByu--Sir^;=Y!HdyB%sB-=ZtWa_{BqM
zsmAD!YP-7BO|~l?qOUME^;MwCYD2r|M0=X0Rg2^6pmm=EPPnx_*pC8^Et;mq3!I2V
zrm6Z_y=S{t3})J71w|WS-g7sH)ac01nEwxD?+_$fuq_Rrwr$(CZQHhO<Fswtwr!ra
zZQJht`@V1Z?%@5?wJRzr_MlcqWv<Msn^9+C^shS$6+sYO2=Imjh9JZz^+}3Crg}yC
z>=fD5iDE1OQK`Jc_Uyw~h6nU$okH6a#ug2y2(iQiF}8hU<mwSK@4Kz9%c$7aheX$~
zbn9o@;L}SAm!*FmImO0{hNZV|Etb-)%yF>FHolDr--G!ngQiu5G*)L)x2n6Eu_CMU
zRKQa)iU+2?U6oP+m-Q`U2}mOs^v|sp12q;mJ(eyk3Rp#u<!D7ZGQF(W%G1P=>E{mK
z{JMl9JZix^WgB+}PdQ0GZi(K;wq{Sve<Oxe@R;OuUgqUrp*L2_e;sXQ<uc#p4WR7I
zSe+iR)!89NS$I}cX#<~H1AQPAx=Y;H!LG?`bAjBl*`@=`1nt4Ey}Y`^&a|sf$z7!N
zgc-0G2kAK@wnj<CWSfG)WaI87$<}e2zsS}Z$h42FHSMu3yt9L?F1rG1VY6i-;ccRO
zzSo*&Qy@M{3=fBiT^&QB^SXSad2j$A)FUlep3xzGsANl*{07b*?~2#d9m`vL`q;g_
zDdY5`8dlh|NgIEl_u9m<DiHlqZw|4-9`oTJTMp*-6>8QVgjZTDnQuxaTkT!UK4V5p
z#F|^{Nr}@HA9+W01O45CZlGAv*7gm)TMl96`MW4WwIvPB)`@$bvPZX(s&Vvy5~Ye}
z0{(%+YL*uXyxk|;oRf&vu}6o;@mGBR?fB>*ptM_Ql?m2#$atQ!@GvU*R3UR4KO<^O
z%<Ln*BV;U-kkdd5=pbLj60@wz&nS(kGK#r>S(q3(tkgP91AOt_KErgDvn2dMjGC#l
zg*+GrABklyMD6DWA+I=ZX+i~q;yGU~>=$DKoqXy<)KpIghEcH$Z@FF7`UlBQH#~`=
zpJT!W)oR18=FnYb1)v*c%z%Y*WKUTnBKOrrf*V>^<gX%*#}T^?<nP+2uDGfcYrnBU
zuOMcFN2*Au)^TqpJ-kS$in&*Op-)<vhJ2n87d}R4P;ntdN;v$u%%*m6c5R5r#c_~u
zcH!5Y+!8zl^H}i>mkujXav;^%bEr%2<Pw(!pSbdmE$s6KP&yPub0v|;wB|=d^B%f1
z%?_{Vi*CPoa9v9ed2)@*$Iy}_uoUZWakH~)xDW23#{r#iT*ax4AHE|&aJC<!QM&FA
zRnQh5u<MH-K-EjA&wqueZ2#MsiS_@7G1Gqx0WQ=Y@&1EQ1N!la){OM+T#s8;g?5Lg
z`;m;<LO;Kh>))<RJ&B1!(Px{6NF$$&QTXP)Um0%q9?m{J-){+*^wSsghi~6~{&@G_
z(b-!wd)EE=CyuVW6QYO5fd<NF_DTMWPh*E~U~}Kk1Of+f{b~Nl!Knf~`NiB^A*GG3
z{shY9xiOC@RzmPl`@khVB__uPchs#eXo?{B&0)*+c7HqWEBrrvS|4s!P-CAoP>TB6
z-;SbqOed~_G$Wac3wg5gsuQYUkOM~tYwBW5!?Zp#RT|$zT9L}W<5a|4Yts*NAo!Tn
zL;(9{bSTLOF;C(6grDpewRkn=l_n#ip*zw-n4c2=%Xy*n+(TD3NY<dUxM1IFjtq^b
zud5v;M=(7m<gAYw3@dytJ8B<7X~LpT8DZ+?1iGC$J+LmoWpIi4q<AR9S6JG#Yt9{J
zd^4My>KE|6L6mypmd_&5h$@rf+5jKO8dvkhI*lN}g$P;0_-1Q~Sh_e>_sc$})UHEv
zdvZW7-TlEEMK|7-N1s1%dbwD*ID084S7bCYczt#)K$8Sa6i0!6n1}ZJw0ZS?pI3@@
zn0CN#3BdMu;QbW9$w;YnkbolH9Wmbm2gEK`8VfMfZSN||7ig9vV9h9iDS8_g^#HXL
zjLsE%A!wBb5KxXtXQuJXEx513UW4-s<d>OTvE+?~s&`6!TL!8-(FE6@c+G)~n0zyt
z&hz+$!`8`ZED^y?b9FV;oK8n%Tw{$VS9LMq4c#%@49;m|B%IRoI~dfvJEB|F@MQbs
za`u^gxZOLsHFL~?8GzOK3r;0%Z5d6E2mCNC)({*ddgO3a`#-pqLhzAOL%Df{XG`S-
zjI>F3AKhH^m|S#nK@d^<Btf-jF8lsG#gP<5#MQ{4GspJd)}i+SCzjtuLgv^LOO3lD
z%O?5HXx|-yJwD@vURe`4-k_5-kJ|Wc7HYeNb`e|nUL3!4a~et~Gj0yG?<O-Asav21
zJsWGgy|j8!bTf#WNBc&Tv*LN#eNjWYw8{x$+7#Dl<6H{`Y^C)eGHTk=v+SH13uA2~
zjk4ox(Zg;9C8uMdUq#4>T|VG3Ni#6j@^aIfX-|c>Qgu;ntN?ZopD^r&-5@Qf6A@y2
zy$<sNw$YHj%CU4!CtYDLSoh`A=-{W#ShP!sk{hCy6zIUSAz11r)w9AgYf(xJ2iR{B
z0(3;C%BB|ngWeCz5o+?&v;6_Ky<Q%1Q|{1JZOkXF7pC2STI!mpb+_$(U}$RQe`+-C
zBoEIzDyK`t&nVl%%QdN-Qo(L|dpG{+NrhGMX!_|b#}<RQc_drOv;f=STAzPVqI`0g
z)Di!b)}<`zkFN49G}p_CN3$x$(|9NHr$A!~1V#fY8)LyD6~(<tSg83#gm(fu4REPV
z_~-Dc4dQW^6s0C~K(W!#$9ru0yL1^%$UgO|vi}><%E#3Ei;z?gO^Rd!k|{ZEsr;xA
zBw(PLDEWhi09wq$1ZW#X@4Wa-04*6+tR;Wj?oLPJ4K^DZg!R~^9nNb``PYk3`FI=D
zKMuIYV?j$7t0bE%gpXX>eIj$?08WyewB2iF!jC09I}2bqws!_r+yFfz(C-`1uMN_V
z-8KUz(J4Uq8jz2?#sR{GEdUQyE%PgQ(?DVcg`B4*7V#R1=~pq@-;2Z||1(+E@9<@_
zI_ywociohm04sF}%B()w;gr+4Vig-%pN|{4Hq~wQ;lL8^<B>WR;nLvMipOJt0NVs8
z`1Il5%?p=*K_R~)&dY2B5cE?<%s2rkQytBuO!5oHA4XsJd_JI1g9c;$6x66$%>RaW
zbw~5De7j<<{b2Yh%`2e~^l_ukTF%B;Skwz)reXEKkX3N-D~=~<`6*VCi5whvLc?O#
zEi%Z+vykA18QRJ`d{to~shLFJE1CZFMM!2(@!fxAj_{o?{m$;#FYARXy%{`{TC89t
zMP8hSC>L_MsvF9lqpShbcMvIK&dew>kkkT^pxWN|Y?{@a%)|m&-yx?Kr5DlA#wc%#
zZ8S+&k|J@l6x0kCw3M9b@ezQ_u3AmVMovYIJzxgH(nc<D6@y8=+1cK2YoV2O#^o1Z
z48_1LLOgvYNEpZ{*}4?vc9rYCh-RnJX8au>$PbQOLIlXVo|La=ns|~2{%Fv`XsdIe
za3e2_eU*ej|2TMW2Nfko=nGZK;oR8|(L_gLI)9ZJz4F@}FsO{i@;q|N1qVM(vN=f6
zS>?aDe}unxJIkHGPt`)qLvV85H>6G;;YhXR0WUO*SQs!vPc+l4fkvERow_W{O09<s
zKe+f{>FE$rN_LIco8Kgna=#P>xgu<i>uzRMydvZlU_n*mX*TgL=q`=tHgj%NMu@7Z
z^9&Z3L@vN$YqV4E&dZ0;n8QKvH-+LwE9W3R25bi2LNEJ4U$<pQ=T%-~rlCvA=_+PA
zrkD;r;!EDynyUJfN0Hj4*us{<cuAH(v%C~vGM^Ub&GGW7E^<*E0u7{JYROLz;00k#
z1nW;WxLBus>wv*mE<0_XcVFHdaIZOLpF6abwO6v<?(dzfit5kXK&Bd+t;+bQZs`29
zjjFTVKZld-Ze`(Mdqz(4T<B@fa$3_<X}03w(8iU(RAns9ZkMQ<iV!omTM|+3dJ2&R
zp4_dcCt3$N=PF18dN3qXJka0>=-K&Rqjw-8$o)L~(k3*Squ)_Ko~aF$ydk1jq^;aV
zDJ&I`Mu(!Z)%VZFm>cc2=V=kce>3Satx+}xjlvLEbPQiuvPe^8*l8}q-MUC7vo1Bw
z!ltnEBqP!Lqj796A|<9^P91b#Rq8Q~7$-AOTddFxe+m#7Nq#<=wp>{yQb+o+@K1zg
z$s<`_&u_n>0Nk1)bWhZ*DJVOesJf&8wFi^|8KZhVSgeM0gYr7Z+Yy*!pnD7|;E|n^
zsPg}!@o;VfN=2dqcn2JN%;GZ;*Byc|rwnB*(d4m`^OOi1dCCzsV{C&vH$CU3cwucy
zZSuMCwsjcp8lmqyy&wni>l`Z0Jt0^t$H7qAQsExT%#`9@75M;{{ZMiO!UbO4#j@i5
zr&m#mO;Xeb0@ub=^yM)#*pdrL{ZyT^N|^H4&Gz>Mx^lh+G8S)+tbWkSkZm%7Z1kz-
z!e=aTH0_j}WnJlkdx)h9tG$Tl0t3}L`WY4RRyngBy4yB-7?u9(pA@FdytB6~<eIx;
zhx%yrO(_Nw_DfQc240mdU6gZ##=*_Sv|A*T++UI+c^?k}H{NO<*F%ViWk`<?A~|TR
zkzilAb?^V&eG-)O|9oQX{=5JyR+1XhdBGa#AJB^`W06B&*G1?#ojPit!`MLf&!%5H
zodOCH6)#OVC|&<g;{gS8DW$6T16kQ+ahyV8Wjpad4TZyO_cQyj<g2~8(QzZ&5%1Ve
z$Myumzb$f6mmbDe_4I)^KbEnTwxO8mi%VzHrLXC^k)&gr?7`rg^}9081o7I~yvnD9
zq55U?(O{dDMd0NV4r>dU<iBB7*8go{#lXt(e-lm*YmWa4x8S=ks9u2r=1xacS&A#-
zFCCKYB#w{oDZ+ii;qFL!b8X>+$m56tlTNXT6^U2f;U+JF=eJsA-SFT_R|Y%Lil!^|
z`!lbftiZg59@xx#POVP>l7vomd);z#n^<G`$iCy2!#?l4abL}2d+~v70lYzWUn9xl
zo@d^A_yWBqu4cS}+sS6!+s_rD=uXBPcs#v3)Y)>hJmm}l0=;AObop?cO8~rs=I#bB
z_Hu<}f*>f3tY1c7{%+y}2lqSgM%~wc+^M}h-CEVtt1nZIVfguVF=8Z{CQftkdNRDd
zMP7d&X6CjrjLd#zX85KaYl>iX_IB>9_29yJ|G{;?%4${vf>YJFT60$>Z+#B{!$sW<
z0HPAwT)+$HEU;#vl(~cThU%(T@$Nyg!&*(qJo2$T?8evuv?Bzu_QTsgM9I&+bsgxA
ze=+@ChJYfx8qUTk!u#(&Cc2C5Kx@|~iqRa`v~|~~ZYwagdEI@}BF8@|+jepx!loT#
z8_;G-)ZewBT7kocv}=Z{<;c{#)njAUD@@0I!#s&7>u9TuruBE1D=)ZuxSPH^*}MU*
zO<34m(e2DTnp@ZI(ACh^l?g1a$xepN)fl^g)+?SN-Yy&E?RJgr=#Cbg-(wnv)hV?Z
zP?Y6;1+g_>th@f6lEEXx$DGIv2-(!4bz)uB9jYptJtI4i+A0f@K-PRR9@D~Zq;*J(
z3JgjiK!E|;^dzsExju;6EXq2?HWQ~oTnD4(VJt*AGf>Sv{O8UanAv-Pya`6AL>gNf
zo=z1?sE4NaG`N=CVxtv^dt52YC=qYHzRzmkJX)|F&@i1ZqgJzh@Ps5b)t4egS8xXk
z=1RzSTWUs#$_Mj_h+OK%`5NIowe=3!_IcIm>5ZeUPP(^Pj`!RNBl+PeEQwcJ)Dt5(
zt`^occj0{I!;Fz*D3W(wUQQMx+}L0@fo4jMNuzM~H`rtQWNY*;uu7QMyUKS;l_{*z
zn2@d7v@FcF%v4!b>oD*ous8<WfnMKsVQRtV{+{avYS56c&4N+JkB26Qx!$5)Tqo?r
zHsmG46Ba^M=n4Eb$iu(+Da%^XI&P<7&&~=vJ+E)Xi`lk0T+~<*#j9y%k9mfW$n*_*
zrRI6_*s7&Kb6;}8zugLTIm2Ar-CA{rwR2*j&Je_;(2f1XVzv=jVQq^YRYH=mTue|8
zTL^Vp3001|T4wvHU4`zmUBy+X>%uPFNMNJ_wH)}M+#EZX*0(w}*_c^MX5Pu~Xi2l5
z%xgvWdY_JT+wvC6V8n0h9zW!Pa`Hh{VG{RZIWJ8=Y4|LWp)i0Ob4eb^`hh%5)55CN
z9eMlYQ~d4yuDY`3bNSE(9;Z{ZE!i^_wpqR9zSt>LO|5OvjZR_aIYczT!Nsl+0>je-
zi7tL0Mi*ms7yRbYzkwh8E*=-2C35BJ&@>Z_Sz-RtKK8kWbuNrroDlovE`SuX*eLur
zC~MXC$8<~mF?i#S=B@>suV06$VOe;C5V`TXWlh!2T?)jElT^R@`KJ&V6OovmD&Wx`
zqRLZ?E+9D|mo5OFgU{(X9?_f~HL>pL4sk5QUM=0II6zeS-Z4<0>@6{sj5J0l)uFvW
z1CY(AYYK-DY~Z;jNWVF}MTFe6e_u`#yicM;wWS`zl89Gqu(Z-s0`?=6ivkjF-<qwW
zrV8pSU{xtZ>dgw#F7sv0?wAsO&Ze{KXedx|ZD|Hd?k>;-rzYo|N}J-x9XDCfu^T#v
zRzl9uM#^=s0sUA~!_o!Q4Bp2SX`L}Mf952_d-!NhQe&n~)qRgqDZr~;`COExu@W0%
z2l;K@m-4}7+F1wd(#n@K`k;88<B;VpkmAWGG*H<?_nkfU2e17~rJ0<rOz296o_NRG
zPT_Y1@4jS7V<2Hc8<rgsa0a8Oo&~V5(of-NSwQzd=5{|6Q);B12bMwK7`ZY=G0=(@
zhdPLGL>R?m{(R4eUYQEl2C$jOkWp5WyUi7)i{oRX9A8XqdB!n+b*Om8yHu!#2`^V_
z5XDusw3!Oc<0yn4gT=m)3Jr*oU$2$6`55J^0g>Ht0f%Xg#i}adcs4<)Jp*RCfgnaP
zE7QKT=O?_w-U^DH4ehHl^QnTM5j(HtGsW@|LtqJ=jQ-b9r{tFoS8EZ`7*SIjv1e3>
ziK?RcnW0bP5!N;GF}6P<Ii3?84x=)%n42arZ9Q*lBZF{HfVV}pOYaL_!F!6=S@PRm
z9pd_LosNzEf4|N!G5+5O#AdC|gbP;as~4(Api}|~|JZb#nbpjvY>8CkPU(uSj7>s|
z`wsv;L>j4*p=QWhlL+7dA$VZ^D(r8}!`r8$`@8!s6#Ke?_y*6{S10*T!FxTD_PJU8
zWJJ)#+dG@rBNa)TSEoHVNr*nY$(fkmvqzAm1KbF$Bu=0DGQl4wUpu*puH$dJ&+7jt
zcS7=7y~U<j7$3M0*3r7QpaF%9@T2m3e>w<7OCOg1tgg6uiB0v>Jr0Z$HR#6i{dh*y
z#U}@lg^Lh3KML{tLI3uCb0Ag_m)aXZ`uXbMUr;B?iedJ9Q-SS$n8a&ia%cNQ>-5Ds
zi67h<oZ%LTtV3c%1Qp%lBcukS>3T@N6b|fz3jdGXDGP%p%1c^BV$VXB^*lqV>OUhw
zSUN^H3&8&~BBZf`Yz0^uF1!qAoSlFBuTpMPoV4}q5v!d}S57Y~O4H1YM`xa@<4}S0
zr{oc$kfN!#ar*5uQ*lR}gIf;?C+{(OQ%+Zol8Lfb;Hoq>J{W`ZL;^@Db+$s!DRSa>
zL}-*PcyE@EyYeQnL2c5paQ7harAFGQUjm|XDN#+EtN$ZJD>%uQnRY6eHjQ)$%J4QP
zIXG3V8{|R0H`-2+st6a#j=WpP-DqCRquMOSZu6Ym*Cl82oZIe+`%dPV!uJve><Ff$
zUEX#v_Bz;_9Q=1?h<A(_A5x+)P=AdXB&yAa-mV6#n7<1Eqk+a@LAFPMiHkCErdxVD
zkaHtEyCophZl_}lo;5|(Ymb=Q%R)Tva0&{)2#@ElzE&#hzUkBDadSX`;{X9}P{6Wa
zfbLKpBS$8~>2FcsRt*~`2*6Io!~o^5bq%Z0?9_!3^cdD2-iKvwAHU=?p##ca?<nix
zyxd<wB`t?~s6PL6<9LY*^d_&@?98u_G1EK7h`#P}#l0~iV8;yZ-NQy-f4|%l8h8Zq
z&#$`3B$wmVv^n5U@!-X;Pzt=x7bUPYdpR%jcegayS|fUXM0Y>>>8|xDc<07-|3C9<
zF3`aKfYF-6`lUbH|JgrgIB9U^5YshrHAR=e#sm7PNbT4o3nrcRP$-pDPuA%Eq8#|t
zIikHH!6px}>YbtjoFfQ6pr{J99r0i4X4*Gza|k*-KJ5pkT8;@B*NJ2fUd+C_2)W5|
z;aC)Rde)h|(m!NHxbR;1)*^Q^YX+s5U`b(Ni3mi3=<+BhRbv@i>hQk0hCYSOUW4cx
zO}}@}8g}5HaGaet%*94C+tD@c<Dq|!0qKNV&{Q|%F$frV^8pm!@0=gazgI~EU~;Y*
zoQ6PlX##ha!j>TgI;0^f4Wv-w2YKQu30u(0KjsO8cO-#si}F7zePZx(oP||Kgz6E6
z*4vql)ApmFY(YaKTBf7wIcoiRs}d5In@x+C$O16n2c_<~_um@P<!Y->f_pg4eONSi
z+$FLR5U^bMJ<$)Q7Qb=91G$7#GEh<vTfdGqCX14xB)Rp8$WNmFEIhLrK}&9cA_5H>
z;ei4n<mf0oXhddEJ7NY@<b=$P+m5}e0OkZ+IJuT4P@p)*a`rvab{{7&@8Qp~K8M+&
zHw6=38L=nr^mWux3ghtqP&HHmX|BFy+>q8E=4`KTNY$;UN<CJ#fbTUe>}Ps2`L8l@
z-;CT0=z^oYDcpjja_{5HRIcVjp(@+3DzjJI#^LGUSvQ%e$$D&_8S^GAdrGld98E{k
z7D~L0JAS2qsfb37SR>FCnPpMSYge5~r+P^vhK(sy?i~l$MwqA`yU3kY#4Nq!Gz2*1
zdkDF4fLCbe4C;NnaSiU!>bpKTM^#09TIMdqQXj>Zc}D}$^f*B&PL~R&B26X2XQ)lC
zB`W{<U^0kd0xLMFDG(NEE<>?IV}2#w_bj!<AK49dbtPbq0qcTIQ*gw(Q5~af)2j^A
zBrXPakKPm(G-rxH(>rN<C4K&%)li9%j?kKy5tz^Nsk0rI4O(0ey<dtfvw`j-ctEKB
zb@McVPc4(ySO1XRpqBb1i;2E9euLt3rQHtIN`#)%hL6JXAc0d7*JGDkm_WAa7LDpw
z?H@3_;>d9BJz*CKHQUa_&!tmdq?QDsK;}vdc7A0z!yf_EvkwVJhoGwLb}-Z=lPnwt
z3R@(WPAQbUo#pCavs)@HJhFbN9&sZ4n$qAW6ZBb#kKAy_^LV(An!z>d#dZ0V=k<-o
zzBP=Kg#?OA6^Hv}w78Jn_@Fmc&9x@jDZB5AN9VLe!%jW0;$M!KTunuWC3iB^+?ZnN
z;kf<eM;s5CPTmaTcppS6?75TJ&Bo>xx4p0z?O?+&;3r7KQ2y~7V*n{ov2WT)KHB`W
zG3Cpa2}vuIr~O+UR-+YMrfW4yk|fHO#hu~gZFFo{m4q&)pMxq9piZfYOao3AW=!9j
z4Yf5R?DpwzUmAYfG@$0PJ}eix^QVp5w>c-xaS3mCa#=4W#W|_?m>C$w`mlze)Ox6U
zUb>Q5mU<*K562SjEjvyTD_dQ69)xj=h<5fhkbt#MBC0CyrK%o_$k!fs7~CO*U?7pN
zN8Nge_D6`YQeHivRK05MATWW{JOL|?V0A;4)m1%}kgY2Wz?08g+c#iuYe}A~=Epso
zJHV}d&nAm1Wg}-?26$X36Cs@-9~h;EZ2#$6Bf7ZC*R|>wOUl$JKMqA21(lj49mlK&
zI?oR|j+?&aSmq7=Pww((8g>1c*0EPM`*TxnKyda0cpzYw`PCm$?REq6lfFPGu0F)r
z${8hr0gB%kWYS1Ewj7wR<S93088onbAZgv9nn)40DO&RS+>v}v_Ln5}cU8)>75l!1
zXDJJSrPM&GKO}&6Kj23EX*rkD<5D|{r6^v*Mqp-O65#unmhPz|M{#D{S&B3&6-AT;
zn&%cw(H7I0_;!@fW63I=uYTx%))Difmz13cHNq<pi^T#8Qftp@IG#p1`?5t3)Il3t
zqReT$+}X88M9Z<92$MqA@Ea!GZTU?ppiOR&uau{=!G_1Y2~Phhc!!ix0#~7VDTjnk
zrg?U&cqEv&UPFB=lUFS;i7p+i3r=}$Yw_VZYgH=utmS>d-JYY-{#S_gUkES`Hs=46
zMr&AGD)F!d<~Nz-^q@m=kfh(u#Hzwmx8|N^d{dj3q@Hd~$XmkK_fJHTcmz#2jf{8N
zm8}GDfE@lG-?$V#-_F~~ShyKHIK7?X4)x=vI{UZRC&F9cQ9qE;zoL3o@c+ae-d=s*
z-mp&eZ*MSukWUo%bAMSdzNdvkWY}7Gcj2&7u}WYazvz>zMEafmC%RXPQ|#z^=fxnZ
zUmh&smj~N1J=4FFv_g@9f7Q~?^?iFi3lL*se+cQB7yG!lI@sHRt5z5hAF1yA{Q3aO
zmRF=d+a>%Zh_rowPL6&Dwwn~C#<KXw+88v|;viz!DU+_#i<FRzA*1EZ8$^kI1IB<@
zGHP#I9i6=yLnx3Ff8M@YrJN3H`KDdX$BW#4uRSXD`z2BO85^%=Z>-!EbW-T1gdl6%
zQGemew0c8=IWaND7pdpdHrs%B`69Y@c%NP^PH_C*QwXwVmh{<fj6XRZoyu09rvA>-
z9A5iVe}bX%jFtexHJPT+qzpTnp`t>Fwe&GlYQ$TgO!A#@kC+F0nPy)SG-_psBpvk=
zCYeYS`4R0agp+_Lk1Y2St_o>M=&u(c%A|(VpB?qLU?fKQIf_FUZX2>Byed-FN3%BW
z6TF)frN=~~<QWv4nX>xi6ue!T%39EM<`6QVEvXkKN}w}{|G9S1XAj<F7g4i35Rm<R
zLiK(E&3px|C}w=fwcUH=1nYgfykVJ~!nDIWF^_|2m)DV|>CtC!;SEm{Ha$jB@zENV
zn)-i}9(AZ|59+o$s#U`G^B4a0Hw6^NaP%3gkc$mbvIvo|uMoM7TCNxEpO@{GA;DZC
zccwiu)IQMe5*<ySs@Nm~a&0xi3R4gsd~&8W+(qBTM4&`wdx-Ev?o-i;J`{4nP{6ST
zoFi3r+h~eOepa^t|Kd=082{s>5`eONMcaSeR^I5%yI6EWV}0Ff0EBg7AN7W8-2?_I
zxz#7?*pHQ#ao8R{zSUu;Z%Cu+cYSFt41yoHFHGL`4)FoX|FGiIw`tRdN}pkH%bma<
zB62B}j_D?6t8-bMwRU`rqwKF5sF9Aw)5RA&<+VDvZXanY`DeNfz8ctf3mrsy)fL`0
z6|X<8wd^FVkDEe3j0Z>M?USp90TW;$*K^WLa9xV%kp<U-qByj$Q5oroJwYn--O*V#
zx1|;$Cx^KXNxMjc!430`JGSUCQ*1#zV~-_9R(N1?z?IJ;?KIdMh1kPDR`ON;yu`2r
zF8A!kio#>QmIv`rRMMz{LY(<dGl8(QmwV79z<g0D?APY1bnvTT_{-6e^M}}_qm=Vb
z%wVL<EA7hXz4~7i?Uv4Wj(UZ=o7&b4CTS|BH3!Y|38pa@rFoJWT!)TPIEx10t&N~t
z133X@Pq9PGko3PckR2tafxWbGcVC|sTNec*Qn!*mHk3g2Qb=(TKgitp2Ev<IGoQ!h
zHo!nq&ULOtEv<Z!y&ml~ZLe|jgCH?ogt=5iO#_`W_tab|YYP;)gM``zI)YpR%@mXA
zm*JbnlDN!+*eS7E2{W`CRqb&rbofO4AS;H|Tc%sgNSK9~4CnEesU-9i^`a-9d=8Ou
ziYz3zO8!xpfiA&^w$r~s<H$awU8dHgz+~l2m<HIncQ95tqjv>iH7m|+T%2a*PSP?6
z^wS9k$}K%;DoM%E2s9L$R}t*EL@Jt*8XeZo>iY3S3DloNH5r8OX~rtgtKK`u3U6cF
ze&H`taBC)6;|vPLS#d5=8U`&i<FFpa5mEuoGcVP7)eoxnabBtG`XD8!n_O!D?A_A5
zM~^1BbBV&y$=bsY7#otEBelVpp;q*qTw>4(Mc+?TKij7z2_}T)_UW2yNh9k*9xZG9
zQksYAf)Sap23Ey;O!5Z^7h2IKgE12w=D??;hCPD2z^T2-Tq=lAcIV_b`8-nDWbs=8
z?_pecgp#!^iZZXyK675-_j#|+!iRW;n05lT1N&xW#aj{eK9X3)<hyNZOF*&k+@e=k
z*k{n;Qvj}a-7tXyDGt0wjHexogbL8b)!9?jU*CCj5|#_FaKL1oHwUm5DEEl<7WgcZ
zmBsAWYj1LDLAe7umD9?%IGj_@H;<_eUoDHW^7J4Xs}Hr|qy7Dqewx-VG-9fHR<&$M
zo^e#oC!<oXmK>gj#|m^~3&S@c<$NF8%}dl>aWg1Vt^}Dz8&IbXUP<^)_Kd@G?>|%{
z2B1lw1P4sravLU`WzG;@K$x%7j!KpQ47iYI6=$o9x53TpvE#Pssm~_r5;4IEmUwqq
zKYE6cbqdqj>~Hs7AG7B*xK5*y11U92xL9k{rNO4KBR?E^Z%Rl#nS|CqLuC>-jt0vu
zw#7M#exvD09JnYKrM-#}S78h1jyWt)kWSsT5?U{&t7(8yf7Dr(O-@;XR^(unD9b@?
z)e0K&4oeE-T09&Muv!p)R>qL<+{~rVLhl}lg?#Qb#n~_Y5-MQU1bNh<B|7xhn*?KO
zV;=Pds9qNjWuS)mq51<hvvFC>fqQCrE#&pQJ#mWK1Zlob2+Qqf1f4f5#3?80^j`?d
z`Hc`rJ;i&F>pMncy?t7r12>$3M4!x6A)G>f43j<$4AEhn!2p7e6<JPu9l&||N4k=2
z=Kz+PZs&k;B*y6orwBWRi0MsScn)VS8h2Z!DHPxWn2X^25z}6kZ6<Sl-GTk-z+=?G
z2KKj*=z)Vo;06S3fWbrGAe!+209UdUVKyCcz+=ndrSZT)Oi{qWZMazgf%7a=IuY<s
zAtVk&?f-c8NVAw%s$ie)NkH{P{!PH5-yL9{hwkT}(F;D~r1b&<jT{e)B*3N?g*6iU
z1i&t_+zRgZ?1B6{ln1yT(}!P<y<!LqDgkGrG~R&+>M4$ph2;R&v8XjSERX?%&CJt#
z5Mn<|YQEg9z-CXMhPTn&6Yy7{sU2?s;7i%7jgDLVCLMD?z+sa{eQBjG7KqeuQy=;W
ziiG|Z>U1>k>6)CDpR~!MvoL?hDAHGA3yB8OMTgs1&3XUjQP{#_-NRC)X?%h-HD^QE
zOS7oOGQOi?_!u#F&#+SYUv=}DxRpD>k>)A!eyN>H6;{z(Ka>IsZ6j|BAX9E=jj=dQ
z*Lp(1mf?9QK}5>B<>bq^YHD9J7ioq_Wn@k`<FkDw6g}~*BK!8aC9zFYM@FTbwF^Lf
z`?R-UWo3NVZ<Y2tNEIy^@6U6Qsv6PNf|kydH`(l9^^l8aTl>3u<psBqVx?6nL%hos
zx>MfL66Z2Ub<J|YlB(K8G?*#`C^mVH4CnL^3*ws-PKcuaV1|;T9~ZQVpH|dRpy)ZH
zJu<hho&PN)3ev-1G|8x1Y?`6c>Vw3q(K_5hgoX0%#rKx>S2?x5;e=M)h<V>oLaSoT
zgFnBQH$_T<TyZU;+QvUe;sy^iTq9eOnuXq}r3S-Bxm`dK@v4(2vcmMVqlZ~(y80{O
z#O3Y=DdyM%w6kkiUoNkLgl#ZXKWll#w6_#W|F+yJmA)uF&Q5Kb2nVFmq--uCy2&`a
zM$GP*>fw8ZI4^J=0FIr@W2Z@mk5lacsHtI{TUpLl%!b(M>X6GqCn+V-wM_tY!gMY`
zAz~*n{V-!&7Gu2hjlTwTshxNs!bh-u@a3BT$cV{WFl%B*uJWsshp_(e;hQy$HAwA5
z<idn*eQ1R78vq7$v7LB9!UwSaaps!<@QBe|yO7}{S9x`*!#Ml|32j@MO|ttjdGHdO
z2*`xtUUKST*w?<uh`}36cG&~JD-_zcRvTsa5(F%x`lf3C6j9|=t^hvAwIctVSR6e`
zbiRM(On(?NMPn^iZX=ZK9On_;a^e)rlq6k95`e-|rE8W~qe-{*b^9@y?t4+`vNBA^
z>0XS5Fn7{IGsfr&8%J!L4edKo4x)-x{_qO#+1NSz-;<DQ^?oYf&TsLQlkHd@vbeat
zdw;mU;#HzfSo}Av&(89{o>Q3_SXlp0Sbs@d!;ZKW!FQ$hR16}djXMIZ3%<Y!ptrh_
z9~YYsy}@l|?3%P^_w!Tv&N5+vX)6l@B-}r~WX-t#Ri$2?x?!jk4k>qAd$YUK{i#U5
z;h%-U(c|?O@h!;cPM}f#Tp~&!21$uqZt0D#6D^GI^({~qRWM%U4$k)y+lnN66+xYF
zSf&+J;Q3C;%d=?OO8Ga>H_Btox;iCGen|06KtoUc`IEGcW|ajEJV?OI{`K*P+m7}v
zGxT&<e(htapQpFm-J}Q~c_DkEw(LwU-6;rpe^Mfn`*Bc)@0W$A-(5?orqD64;7Igb
zM}V~b5CW+Yx6&h}V2QAC3V<qAukRPV3&~DT`Wv!%e_2k#;Z&WJO?aW;nM?B?1Rr>c
zXc@q89)dLgoWtiIK-ib^OWvhHMqfP7Rg8F7X#(i*hLH$OLn<t-frVG_WPvdeIgcH1
zfV2l6ljcEDY;8<nvk}8ya`}FD!%_;1KXO1Ill*G~u&(($ilW(zoC3i+0%q<nbr%>7
zn@U96kBBpJY%3uP)^qAYJc@jS{)955U0WCcH>MNdBBRu7JR$(2O2`SMN^-Xb2*MQ9
ziGnZ9<M#jj`v=+OCm8I_Kp>$XzN`2x`ir(9la{5O)Bz>A5#{e)=3775#|Mi-jnV&c
zzZD)g|I!X{LRIFDRINbu?3nN(pnqJ0NEZnSBaZPhUTRoj--tK-3jr@MH}Df_+{GK$
z{*kh*M)N!iP&FCVl(yO8jqQW1n;^zn|CQFU#)|qd)rghDrAkj}IMELK)L51;#qfPs
z6*EtHjL_&Mg6iRBLQa)-SX#-cwjKs5oo4=;c%MzehpWxu*3*1WR+iq_ntLOcsK(lU
z=758&W9l8Wil2k{JX)JoBCTE>$wU~pv7~5g!pt~TH=g>}QJMHFbKX9Y!<<R5dpaLq
zA2&*8t~-p}C+06DsQF1<exM@!EsqV0Dl(4xY@F_tL%;&sw5Zg$>=p^}FBS1{2I`KG
zGQJd`W|D`;owkgxgnB6$PYQ1UJ0_a^;ooig>q@Psqa#H@2U9l4j*g7Uc}v8;OO4im
zmBb$PmnEu@Ud+C(4W4Kw()^|$04DrTcNh?5I6F8nCld4vH4(q~czUucFgQg3M$;7#
zzqo`G>^kW=KD=T+4sa_pu(%UCc%i+NS%`mLb@1aPXD1kVWF$745Wo6V)EdwWE<Sd~
zt4VO5%#i>(vL;IT6*zPRwmMA;Ni+0cM0N5baA-(3zxj3qp!_0H&V6>DJ{}@YcA=C6
zRp9ESu_1X;Ff3NUp!%jOtp0w$n&HV1#6?KxRT;iw*up>9Q}My^Pk?Q3sMdc#;KQ2Z
zq$FQnZONMQ0Sr+2^P}}hwKaC64LpWX(0F8dzBqyn8E<yS&rr<UEpSv{g?xvHEB$=j
ze^O8IW}$L;KSfpEfZ64M*PCX}6d13wUKUCpp$|={?z4E*z$vdoXS4eu+YKn(b7=om
zXUngv6z7&D^?_xPI~~9$*DEu)$)lLyhV5%-yniU8q1H&EIl!mGcoI^U)9@I6Y)0Uo
zDe;|Dtezh@-}!d9XUZ>$YF6yK6JNr7Q|D8n|Nbz?U~Afhwm3<C2vX%P5j00R3!d$E
zN+#HBKx~4m9gsL}MLUy!?M0Yww1S%Hg0z%XE=OMWry~(04<lec+Ut0XS#d=Lg!xzq
zIlX*-vRAzNm268p;wPGUGcfDYE122fkyaS<kntQ>sDIPTJ!^Dq;{Qh4PO8$pB<ak1
zhPVjVO|~ixD`IavaksT*0_0=V7y#ja!IlCFCL--Y!h02TctDrCf)WOUB7s}r)oo-v
zlb0FjSlxeg;_PmJM-P2vsu$$R#DIo=0SDFXH<tmX>r26~uSOo8{v~*-L90>I<j{8g
z5H&!mMnr;lF)@itC?TYIGuooc*zd6;5lU(aAvLcW0|=?>ho(m`LkWumy3gNKDYSo+
zaW9a$J&2yKRU59I1y=~G$6LV}?l|`knl&kGwQGtnzoazW76i>P!Ig6Fm?F7$X<h27
z6~dA-`ukMB_dei%bKl62#2BkrbXVFp&~s;{%u1rXqld<?(ZKt_t=GWYND{t$GP$TK
zr90qqM2g*zxmFJE%n0~IFk6GFnqO<b@1Njb;K#@Z(jB!UXp?V?XA1e&MQHx14aN8#
zY?>f;+6>usCU{bZF>$cplr-f;Bhgv~;r0NnL~D*P<blllE*j!Nwm}bm2Ef}>(=>ei
zjL}RFW+ZV9HeY7Mw}zjhx_bbSG2M?MV5cHuP?8=PhES}S@k(QWwSi~K+=T|~OfxvP
zrTcahRdL%yG^&2XtKisP8W?P+`Ql2@GVOgk^!Ex-q4hVyIf0Yi8F-Fo2DK-ZpFM>7
zptWW|I`h0s`lL0u^3^#L$glsM*i%_kcJq24p?r@SlB9C5cq76~+|eU~J>EJgH9du@
zYs<tV1(Ce#m6bjlFd@CPOs(CPrPLG{3)#}ArR6Go=AH)mdSaubltz_h9|<<V&g-oW
z6sWUk1EkdirY^-o7Z!9sswx}0^&rW5w%M}V<`o5ow?enr8PmqCEN$m{4m#pH6~FTA
zl2&<!aV=b2rr7s-!I~vHl)knas~AWTBy+9R3Rb{)I2rbSRlsez%5yVFbava2Ame>}
zIV5B0JvVktln!`SR(C|Dxd&dKy|K+P%ah<wlc=M}hBHe6;qV=jv-3#wJkDLVk@UfX
zLcro#RwXTmy=|OaeJ3m?g&ikv9yl}SHx7aCitLngIDmO0oye88;DL5PjVuD_u3+!D
z`O<<I$Ky$YlQ*UX_(CiDpv3u|8OPgg$BD&dRk?n&mr6w6ih8TIQ1H{8HrnOJH>hl0
z1T`n*nvQd_!E!Q!v(<{u`CLc_+>>a@Tr*P9u|1XCT|(h#?hMZi@ArZnngZj*!2GIx
zKi<43#o*b-+wq#e%WcLRJd>3>``!rPFg|}|&VCLSlu2>k7gkdAa{t0SFMCT{-P~K)
zVS(N%yjT<7QH2)Xlc1@fA8hZ<yz+9W6^i&n=gr$AQO~C-<lWKPjQo?+S*T#SUvaTh
zaGg*y{y>PtZ7BGI1mlf2Ay?w;1e0=`10og&(<xq+ya<BlRwB<S&MXODa1+kLGA&)*
zt4GvbJ8Mwl#7o=dQgl2Ii~0RbvvG4bBJcJdbJr%FMGGy-n3GmjQ6;NcM>G8}LFSyz
zW7F7`y1&|CNdN`OG!6R<bp~gI0mj@={}^N_x!}4}zYz+D^j(F1`H6^FDNanL8_m@9
z<g_kx?BrbY08Za!=V6{wLMA;H0v~1u%9y}jAif^BBmKWL+&h9#%tMoHu~Xan@gom`
zsrqc<0A`zrKgK<;(v&y9mSpC6s=)keib0+P0OT^El4eQ#tTrBSx>qKVJy-;8d!qCE
zzPbu(d%Tz_C~d`k!MM8Mwh;dUHyId4%dY4z&BrcjFG<rgs_XDzW*{$DLA!hjRzW&8
zWV3D8TKk}BfnB-M2r!U&^yOnh4(S)?y-E%OPN4XwNSqK|^K^k~f;}Sjry-v1;;Q(s
z3g})5<)!dy9dc@x#A-4I9J--<C7-cb2mF|&t!}MESBPX+{h<?ATTTcYCf<ymwZD#S
zh2~R>pw6c+b>Y%#b4tZ+Vg}g2UYv;8tPML>y*q5;56(FdD=x%6Qgxbsx;i4K@vwf9
zwS%eBwpcaV>!^Me##AU*kDE|MaxO`5rE)IGdiXvO*OLVn+)D4tVmA0b=L~0-4Kkm$
zdW$?Ef8?a0<2RW=%>F@HK;%%-2!9Yc7W+$(c!fM45(EhAsW>NNrDqZRvGw0`oO5AB
z|Czp9)yWE<l+69~5G*PpD-d%a{vk9;a~{K;Jx!da4iYIBYC^g=1EH8IuxO&03Y>@j
zv;W!i$#rgV9Y$9w18y2*qlo|s#}@j)bV`Zd<g@jC_ZdsIxvlNn(ZQO^=lynPKN?bF
zG-y7mo;g4zGpdJoJ*IfxcMD!UY*D%q^X=*qJ#u(oO5b;TaYBbzV^8Gs_QBVSZ7c{G
zYP;Y|`dmyeo(1eFRz#KhQY;#uq3YJ7*;yLRl8(8+UOs0DZI!i{65h8R#y!hnTCmb~
zaFl<_*@0ssICu6p2;oX;ll~&~<L+O%sut}{Z4FKD=POlAtlszM#?1`gPv4&2$jM1;
zX!|Vl8fdc|MwZn)tR-G>R`>biX=QU0W3U!n;ZWE!1>9H;?jr9Ig@=nWK0g@Obe7b|
zAK@`mOk+TDOB2y1IRE1e>?`pFZL1FtaLeb>)Lh?B4-;!y#ytc16@Nz!0oZd)_H~Ie
z*_CC+BxiOQW}=P~3M*Vt6_{@Dk|s=Gy@hCHTGbO?A2h26T_OaqL2C7Atjv6o6x$mq
zo{llgM|73m9-C!mqS;zL`s7rRa4I5`I%vxb1a)*%>uL)%sn*?7#-2w?61!@X@Hqw^
zoh>`fYIaX&vOWbd<g}ovm>h!)QTgWe{ZkK9^O{8;7*0>%^tnQu!XFW~_RN?SY@Wm9
z<@NfQSzV2_X!}N3$pF^1ux_7ghpgEw;G3Sr-3&c3h?3cgiQ`M<P2h!Ibo5y!3E98&
ze?3Y6uFzA0D8oOFUe+t-iRgs&^P<q4ug8|DLswqx#mwetx=auk{22Ywg+0f}4z@&B
zW_lk5t!OpUYV1ElgxcysmuEWi_o(}8DGE!(Rv3622yrco0K9t~;kYu&2`lkeHMR^B
zL<%J8oOyRdyUjMJ^b)(T{4@=Y0{4BwJYq|CiVK^pw2xB9`tsoL%$BJ_J9`dZ(z$~L
zZ-*uI<uOYQ{^CgMp<E>y4@kC*i{r$b#Ij)uQcAf(JE}rAbti#CYsQYkp@huZ<l-_)
z{(vHz)TN#OYb?J*PS3U^lQLoH7P(;fEhZp`j$<TJ(WD5946MAB(urV+2R)URt-+1B
z#yD*5auP?P-R-GdjH7VuW%tdnq~F^V@JnJ0kVZGVH0-%Y1U!WGgmf<nbl}QM+xLXI
zG5FCzNNefo@-B{MdM^JD)hW6pmKm#Q>;fOw?sD>_!&Ta>==9nFJCXvY?KCbOh|IG)
zg-yfg$q30$r}KXx==W&Ds0_QMy(}fpG!Ay3{&QP{`2qKPzl`Lo<JphJGrJDJ!#DNX
zfip;=R5ODY_exDQ&lFcPi&NLRErpvC?y4V)Q`YmWtJVIm%uaG?e4L<~3YUFq+#3u{
zjvo&5D>wzYSH~hKhAD7?)e`!^`&5N2WV>WWDYXK{>SU-i&RjK{a9CL?2gbt99>=a*
z&^8rKN~4%I*(4}c(^eL5^QeGWN>o-~H*aA7u_=!5H<2!=v&?pU2Zhta07CAbz)@f{
z__0@Okl|Z<1{e9y?9aF02R-p`q8|o8_o6h0$BpT!YZie{mDt8V2#1H&j^MT_>WGTW
zM{Xtsvz!zzV2;wNuq{?LjOU}S5tV1(GuA=bZL7rIZIf-6cIVTB@O_wME@Tou&Ueh@
zdD9tvo=ESUhmFLsNs;8o7ru6rPuBnT62b9bSVwkdw*S*hL^tN7Eq41|P3<Y3d}i!E
z6g((KG&88<!$2#3XkS?NU!4C;-T6is$5S_~@epaL0~Ds4CJuY{ir0H<4c*`Ey*&(u
zR)!2$Z;O|v%;=$o(*b{{0)ka*WF=KDIlg@wy=^4IPwbyy@Zkq7zdF2b9JF%qSe`kA
z3?Wd+#ljkA`IWp}6JH@s_3pMt-RxtciP>cA?cV^eWo7wV&L$70PYVR$-)ZQ|`u%*a
zfC(6Tw_*0aE&-c8MS))Ei>L>Ow6WnM|8-6S#~|Du86op5JaAe+K=rF)s0^OS4P)RI
zBQF)JL)ZbP8eo3ZI_kcU5_iB3kUv2KPYn`EKLC#)0w<mMbBL;avF?OBegVl`NQs<j
zKn8xoSc)gX=sw;_&_RC08T%}YZJ8#adBxrpByw8Tl>Xg@Nj;il?vw9^BCF|;kAvOs
zO?!)vEPML~!KQYy`lwq@76=1;CTL|cjB(G~=9Tm=I+3S#VH3~D*jwGkj6tT!>SgTp
z8^ErK;+fV$Lc43{K*MT0p{Xa<b~qth)AJ-#Y94fEfF)6>qKiW!Zs&ZQK&&I}*lC;<
zA>5EHxVR_fL8+NQ;_jIg9_w-){7hV4q_KCnPklI=G&xZ<eHu*@PhzgJS3;&ouNb_|
z38~TIA>vt^{)AyFmtf#|wt{rTtoYI@Gl(RkR?x-Ib+RLop$Q+Oa$yjsIubY8X~55D
z@sKdRG6xH|#=&Eepuc(4+;{0L_yz=3xYeR<FWsB}cR#FF*zAaIkX5Ss!2x1f8?8uJ
z=T-y_67E4)WOP1}4slac?itpJMh?CfSOxi_QuCTa_}U6wGo0Y08U~T9!slsKg(l9s
z*R(WwG-v!s(Ydnq*cRrcvZnFVsj;xCd$V#gMf=FLm8{flG)IR1taG)I1AyBh7ma3!
z!RNFqj!Y|J^*o}Q@aE!zTtu#Vu<YQ;2a5!wq}?c~@GUex<xulQ@>BdGVv9_o^rg>-
zN5eRfHCzz*ua<T&0JZGl78_@}m^wBYbi3Xh*43m_8=It)XKxPm7da9mU}Gh|m__76
zw2G?={D?|)56%wr5>xGpyi#3E_*$Q6E6L7fizz6*3lm9NDLy36yNJk`w&FUbs=VZ?
zhNT5`Ri4#+JkmwlAEQ2jZ$(!a5j8QsIkV>3eck9Y`R>7B5&1Mv&(%}1AViCrsR>Z(
z>+!s3)k+jowsR>zv2Q6r!#BF{gMWmLc$YJBPGz`W_4`m)X!VvzO#Joh_>1>os<;nq
zy*^d%<@`0e>+ymvCfK>V$idKb;k3F(%4M`w1EC0%6iak_7XVk{Hn%j~d`M>E+e@KR
zY%}aOX83D))Uo7$t0Hpm982idiM9mn&aT^<OKtG!Tb&#9Xr3A=HTsfP@VCyBcJ`h`
z6G^IAZKb?BZ^?JgJF@6tZI9$zE3=;&zUCR__ji*;AQP#sqXa4+%wHa2ln&O8RC{I{
z3N<g%WRB<CHS~GAt{S5my1v&sZe=c0buM?xJ@V93Y}Vi&D89~x<XWhlz+GK9?bFTD
zgmnE*AyoL)eS=(x0wotE@-N#g6Ixx`&sC2{8&*Z(r?<3WrdiDPrIPP-^G^Ui-J^?j
z@1-&ot`np5Ojd7?>F-&h*CM?gG2adao|!yBqun)%MY}?H2WI4}a7=~wX2{a!cgwMy
z#i^kM$EKN5IWb7n$1YB)lg0lz1?l#Dy1)7e+~TD+f!&LN*`@r@Shy!o=u3tLNJydq
zDbK#5aXjDd@$K|{A5HQdMofNVl}3QXMJh1)9*RYdCpe`k_>`J;dYaJdj<JB-7}VqH
zc=SP-7~s4%aQiz}3a2;B-8dl6J*VZ>4{O$}ce2-8_k}>hA{gS(eE6LF1*vQiDJU;M
z$R$M9#KLp%(LljF2<1yHZ#&nU99%itLvmPBKT0SFhh|s*%ydwxLBDA_%&Hc`^&>BZ
zQVSFszlmn_yvU3AV)_-f92UjF9|`7BO_>DE>4d<m2FQm@ys-BSh6bo&mrGizB3+>U
z^+aP)Ms#BOyKzIuXTHX!{TT0KMjxwqh7zTN3qYejR&-T97z4@h^zPxAKGm_%Z2E-k
zbT2zh-5vnLZ$BvqB}9?<D**WT<#ve^s-AE@(X&Aggcfb6^&#<|3-IFgA1nH8xz2o~
zm2~CajnX<4^;%x9&xu{8ZtYX{Q~*`bnI#$G@n!DuJqH`0H8m_P`^EN>InG)#9APt<
zjHQ`5jio(~s3l8#8Ws~P!sndhH0FI5CH`<Z*He#j9>;t1oY#$<*gKwwFEJE6orhd{
z6FF7IF;i{9aTcTVs?)&$Z_c&LmLdznC{9{U@*Z)MS^-r~2(LBf<R9|PM5WexCxxnN
zQlNi^xNzK*&wBmqI0mYNCF2{nPj9FRq`1IQDC9uPvs67ISwyKqNpL<;^t21o)6afz
zs-$ytGNILI_!<wFs2r9H$OirY*n8`syq>LVFi0Rl9w4}d1cF1b;1)t~cXxMpOK^90
z4esv2-QC^Y9p;eyl6$M(cV_N4?>F<;R8bVt=X9U$z4uys?SA^;I8<-)>n;}7=It*&
zHlB}>ymru5=*A^Zc}JaIV|%J#U4F69$#yqCQPqG)65>Ro(zGz;P-cBw|J8JsAa|C+
zr%~I7N9|}Nj{l6A)n>)vFzCwgNrAH4rn30XDXBOqP(4Giy*Q)|dAIv0;e%AyIC!}I
zT0o+a#x7CGv0NYClX4S>GgeyoU>bNwtgfUwGBe*1uWLrz={s>9&&t>l^C&L(b0AP%
zU8_WK(8vTri?xl|7}^@Y2iI+x4Mfzdy~WL0YM;Z0=4OAlL^kTeWE21wLr{V7aWsv?
z#Us5`h~-W9r_W!j?u9;C6BMw8e`HXos`oY6z;9H|g2jeWgQ}boR*?wn7}oDZnaBVc
zXq6Hf_(&GfUqqypEfXhKi<hqr9YG23E})~|DsLrtv?#MQgh5KX2C+q+Q<YX>uM~8I
z^Ue;4hW!|~4_48$>fhAZxHw>f^t)+DSrI?Nn=XiEts&WwB^+Q~PG5DXaL4@ELm}L$
zgMjc^hl}bAWUqc}+Kv;FkPYkxGf0yS8abXi85;@e&RzUzI)~5d-WFTqVK+{Hun2LK
zW2Am%9kf-$TGRemGC#Z7?;nsqA)iCaA<;D0x{xH7AWfZLc5o|!{)6-#H~OW9mHg|C
z;}YdD>f0g@?2Y|_wFLpk<zy(S%L46X9~a<bG>Vg!hmO>IkwmGr?-LB`E*rLTDg<Xf
zbF!-F?8P^$?6^0px!#!bbZ&|duMHA*h~nYewTASjjI5vrcG3>-tBV;JituG&6_Geh
z)=S|w+twCn5)Uu*ZJ65DvBehMdPN8`@+~HH#8TuS^=P3X8;NvRBV5U|-2#X24o&wx
zDAa3M;u-Pm)8siR%ndyD#NWHrydB5##HGQiO}2d-nV^+;#BiW>hvrw4@MebOo|2Ee
zvAv+x%0~mPeug*KHOTCzW@+;s1F5cBz8RZvweA>AeSA#Ror_yt<mPq94<{iZ+tTea
zO5!_$t|Sj;LEas3MfUx|>A6R=b|j&aJYS#DJAcy)mx1hr{umNBW`oaf-wFx6lt-?y
zxIi1i4DPbe64ZmZiLI6Zza5;6jTSF9ZZ-cdD>Q~1)_RC4Szi0D+9V=qONN>~h?X%V
zZpA?NdSOFRZcBv)E^VPF7kaBPLRb2&DsMzdkAeErvU!G1YtbfrtM!_TmMJec^g$sm
za=}$1&K+UtJlv+(PE5O#((s|{!TitK@`GUrA<PA%m$Zlb)Vh^-^KI*ll?2p|N`bKK
zpKD|CHzk`=lr!S)YV&a4C8J+aXY|f*xnCMNQc!|#BOPy~Dq=QE>^vgWw9=^mH)NZJ
zn&H1G`_cVp2fYs{120-(dh>h?8=gK$p<oq4AH#@s0c<mRW**84Lz{7iyKK7u?dFl$
z32At=cI}!r!#Ayf^lK8=iwO_uzMZiT+0_WyR|uSq`iTYRnH0iDoTC0_{Eun^A3uFk
z$nqR9bbQjey$}$jzGpsx&Ki&BjCNyv{Lz*n8*D`KQDXO4U&8CQ%X#T-q(kq;5b$A&
z#^XTvETYMh6+7eS_pJ;RdThi`YU*=2JZ`s{9l5!NI$%lKBfEMRpX}Z|v`zU~NOtPK
zFHGQYb^zYFLY{^z-219{;Jk}7I|SjVQ`Sf=91jWQaZSz5Io>;5txUdAyHrrim!*`f
zr;PXQcXXC{K}{aC*8-0v{I9Ru(6#FY9?w_DH3LW-zUvz!$y@uTMV-TQRATkj5Nnr)
z3$Rm&M!sb94@yLadGC9$*N(a*TFStf+L*C3M2ZIc(la@MPEY|ke&U{re|IJO$@%uW
z8ljoICFFroT4x6K-NwQ>p~@RIR!b(A!6|QIpI$LRd+*n$7?1{isJ$MQ%Wdc1)V)U8
z`fJavNG`x}!9O8%p;o79?xd2EX}zMH935TSswxRIJj5T&1raEF$)1+4gr`u6dG#F|
z6bs3??y8Nob?vOEM3Z5br1?kWYM5Owg-b_$)=|*H?uo`X@v43+66RkW=od9=+f<3@
zX#qt)H2j(z^(n7$f;lE!cQMO%j*K!g{ggerxLYHZJVrV+{N2z{nD``=Zyv?Mhn1Tv
zAF7*LvHS74P1ee@oX@Zy=-9Cey>zwBl_WBHDBtK5sT%4%j0rp2ld*K4dlhZh=E}*t
zsu=FIK!W4YqrEdU#ykrz0=2EQG$+LO5R$SFYC2XcRT#har7>5j(2<Y9Cp_{xsq(k(
zxQt|moZ}WJrSydzFRP(Y_9Y*WE*xGfmse^jijov0iuwwp8>UqY@)YnnU!T7eGQ>+k
zh`(jE3aG`)ROX!YTMLq2x+s`aJ+%Tpu8NjqSt>LD#T~#B-+RWjGrP$K6S0+JNcHg|
zk~huY*|_mEV6CiN!~bH&b<HB$;A5PUMbF|G=>TcJ)6ES<KIES31FYidZMMW#DR1-p
z3g0ICw@LG=@s&0X+9z=VGGarw6nilZy9$C5KfleO`>j%CxD^fAX5@?t%f*()+$P`A
zVP?q(x*SH-5t9m)F5fCE8Oo7HC7$xDf<jOi3Ns_xu_z?Tvul^Qs^!58vCBTw2T4*%
z3<zWU89uUKOq5xjOxvR?St~gsLA1K(ThD3sM^~<uF-qWTKuxi!d_hr^tIFLwI2{{#
zP+TU{mI_1%KFOo!1lRtl_i$=ha2%&!@!q)`17<Jog0aw`4YQOpg}FzT(G1Bgp-%c7
z_moYzgU^TU>eF`hKzDApk<BZ|1tz-#%?z(C%s5=5$x@z*YAag>vTX6bPnDZn%OV~(
zR>VPEgSrZXWg2ykjKOU_7APXM^jPl&Xxg#Tk!*7kv+>>pA-)J{m!3vNrWxxh;{S^M
zwgyWnAn#aNsdiiG<<(T-cr<vMaIc$%yGjq087l{EZxAI2l3_~srVI^K8PEcc=g_NF
z63;4i)6C>ul!*&LUUH_D0mdQ(M`+-nkLFdnn;vz>TVqSNJ1pPw{z44{f<P?d#HI_w
zUf;sYGltk^VELVoz$Ym>Y73r2-&$BToEvNt5Ogx!yD?okL#L>Nd_9@Qpt!k@6n1T9
zVa0<P2-27p7sF+!;<u1qCam=%j$ye}yBIgET7_bgCyMj5#6xtl)UQft%=UPXFDmlc
z2^O0L!SHM3Qe#^R8-rM!ITf)L8@?EEAKVSk4(orOZuqumts5x5smXtpMJK|tZu|I!
zWo+F`%ASt}ThZ8h115<1vY~K9D6nlsp?kg{`*qA4gzpRf4B5U!PU4$&dP=uZiuwZO
zb_4rs1W7XNYQ#ts&5rKYxEeO5S!5sg_f=(kXQJ4apLC&EY_+GLh6xHyykLG_^R4kp
zRA8Y@Qw3{9w?v!M&wc1if~Vv~U(jjO!1BS7oH4>grtq<b%Rg$1U5+#_zm3KD;1l|S
z*5PKZhb~`oiK$yB{YzCRQ}Fti1JmsVNa#Tjd=ZV^kDeyW5bw{6Jsh1s3*M<P`f_Bk
zfNfZi6S@)L2yRjwHj6`E7uvoEM83!D#8^<GGm^%z>UWD?=Y8FoK(A7Y$`VV(Ugah_
ztDjkCqo?(xXprqt#yTa)+SlZP6;`IDJtu4Ca^yv?78+D{b@BZHOg2M4;|6o0%uIh$
zV>uLDL5`V$JFs-KxVdKs&-rMJ^l(YotutQjiwQ5~-uBkcQNd7*U%W1fl7Mn&lg6Nx
zW-z3TnKa_w+xKX7hNtReX78gDw`x7Fdo*N3Z)5d2L!A_JHCLlGroUub6=}w2@``Ga
zloeS=mLp7HP|7flsOm?k)LFGva8LZit=*9S;IprNvftyA`(lDPVnM=UE&`@vFbGdf
zToNmnWFFK(ikV2)qtm+4vphl&I;dG5R9R>$2<|-e)IU563DK#VnofCKV$^c&E}(n<
zDM+7JWc>>w-1NhX`Bn_f?+iI<NZ;rs3KCFUV|-!QQ0`x#mk;Jo1cSXn={VfFpoqt1
zOoga<wX8_)S1OzDc}}(mMlGaf@cQR!aCF6KeV#zMo^r22ZFW4Do5}S)`95Y$7cOqm
zT11)`?}Gc#fC`DN*)^K{Kxs%38W=j1c!^0ZqXU(>(GO9(5zY$CZb8li0sj(oRB=R-
zkl?c>UJ^zzx@LEOftm103e`_c*(qfu9_98hDTrWSgM4Ecb`jYIW@ak%-WWPep>cCt
zF81vwGEm(4Mpnx)y-Uqj7%;r7eeteU>CUn}6YeO{9(nt7>}AW_mM_eHZ!#wqM@ib&
z&DTW|bgg=W%Z*QQIlMTiG8}p&J_|??r(Mp;V}5Rd9k=2LXL#{dRA15#9Pta22hC;i
z1z4%T>v2S}s5M4H3+XEN@LHH}EoI!kreI9KX?ZG`DN$G%=WFR+LJYQqG~-0aAp-M(
z9Mu*|<^6kx>p?}G%NapVsU0IfHN~0XzFw;JQaeNPIM=lWS^W@Je=VE3YAy{$q`dk{
zIkS}z$HK<haFy10zoEd}1&I?U+MW?Pbmk`l78C$2t~G^ePWC+>wv-tKXG3Ymi0Cls
z(eA>6;L2&KTI2ElnOcf<#{-`-qYb*3gt_a65mhtZXXt|w*R6idx6MjPcvWSwGELO^
zy`hKtQNz03mS>PPyVvxmz*^8qD^a32KcCjqOR)*JE+nA}%|4cQ2IflR+G8{A;3Vj{
zMkLPWO%Rm19Ws6m8MMWnbM!xZ2q^aaFF~=I)e|ldJTul<9`Ua_to4x5P9S;tl^A)^
za4X!i5nbx)M<S3tP50Hbet@9`gmWoa`Nm}==+f0}ihUua>j4q5@IAUf;tk~yE#<VJ
z(2c*rE){UYwg1Eu@adz|j72ex_F@k-UZEBaAtgqM@se<@ivf70HTh%MqpN48IEhRK
z;|Al4_GFG7)?2>R^XX=9wpGYflrS8Ga+#b#1~$C#kF~rTB>9*#je|BuGnO^l4UIn~
z!cCk|_;0i~U{-UJB@@P=c~`<K>FR=1)*yqeG-(LmzIzic)Q0t1fY^?AM<})wsukg?
zl(2o>pk>$sm%EqFOukMI%H1=H`ch3&z(t)w&&Qp0`vROpz96rS?~`<$dC;O(&OSLj
zDpvQ-VWQ`**jQz@uCn&TXZ*fkOBO%-1^Jo$PXBe4Ak_Lt@b!NBZ1_Yb0=W4ALzJ#l
zYFj+^#7lm?g{vd?ikSBuOEu{fj`d^T8Q2GECNSP+(K~R0k~cTL;F^412p&>NtK1dr
zwo#<#tAdjpN~>!koLBUMuG!9D%$_jhV0?=eN^YI;F_>NDb2kgvS+2It;x}=oZmmNe
zdYn$WF}MYvpu`*>T0%b$dM>jKC=%<nzH3YrBF=)<ymoYJ>UJI^pNmnJ3FUBqUQMGD
zky~fZ;Ncs%(gRE?LTdNi@)AV5grFKF3&cW~`B0CWYm3&MPE8-=+}&U!laJnrHOv&&
zi2BZVFsqsgxNyr>4T<IUabWhP1BV^D+23vTc7Mt(bmA15iGQo@9|!9=ws+cyT2ggz
z(dx`9U`rd$sf$|TzdP;s%&O02_URc7)Y=}r(&DHW#d-koT*-_W`8wR|@`0ys630}W
zpeWZS@m4XK)4b}COcZiA#U#6nsCz}DBCw^ZzrQtX2y;@`rtq${#l!Z8M>nq~<A<|_
zMrBdL;Wj4llD0X>>^HlHGwrWd!~+P+$Mnoy7CGxuM%n1+a;ZeTDVuZ<8Q%p=qOpAT
z(q&sk{)Cc<cQr6qIEquKZ(>BE*_=(<BV%sG)&W0gsOhVUS%>n$fK>cSyF6@FM%|OU
z0eyzriWIBs1!X=b-WwzzopZ%Qv2c{G)ySJAGxI?%b{DI3cznT%+|eR=A%SFP9ATpg
z`=kI_zreZfj>0YeCT?e?&*_Pu=f(zjS7k|225+2@Kpr0$W`n@umkEgwa+tG?EK~Iv
za=mHQN%-mnrykJz{gC_(yuAXkK9PJ!BIKZywNSR6?KcA<NWFFb1iC?9+72g`5XQuH
zM8a*7wt#hoC8<HHLcDe^FrO^*E@*{(ou9t$H^DuPEn&(P@X0+jJ2I%EApBiJ6L!|O
z7d05%{ncoRQnqR+1?!(vzk<m0wt{?gxn^!k=6GMB1tMlrCOZm>Y*3qbc!5zF90iLP
zIcd1ZL$Loq#Bo$S(YmZ@@QWwwneKOaJHF?R6XFhM!#~sA?1-s~+AJsPiOr3{`(j}v
z=hB%vmI(Jt`5EsI=1}DVTC(Y#)x_J>?ORzy*3Z;L$-Q@}q<a2A<Y@k-%n#saT%}ln
z(onFZaTS&Nqa@9Gx+X^0{0pH9eV<{v>Kr>~x?2>fFdnXG#Yn_8Iloy(7Kw{^hl<JK
zC{1Yi80w`jo{KZ9`1tP@$ep1HO_R9?5i_TMAJOj7?IL5L{oy4z>rv+!SoZbFtY;by
zliVvU*@@w_w6im^JBb=uRmEEavV%A3Oc(x&0*sGna0wo#;ky>Y_l)XRThEUH#UF2o
z3iRM>d$s)JCG*r>wnAlbHq8Vlnm}ypUsK}ey4&8o&N|1nbB@)YeO#G_^!|uzPGv^e
zMOUrZi)%`~`7YPOL&Y1mBL_R=j7XB@rSUzMFl`}MxlYa|WBd$$Za>uz_7tgkF^Eln
zPAOXqE`@Rjb1s!LgP-j`;nBM!l`mo)a?FaF`^}uGn&4gMJwaF+7Mr-UgL~;{{__um
z6rP+>;~}ZMl3pPX`)BKr69$7EneeecNoEhHxyH{x>3uQU2JrMd)SERU1yZW^6Y?4j
z@I^Lrmb>1^(9E67y}B}f`K_L>v?fDmAx`AKTg*A3AE7;#3<Mp0jpWSh1gp%*LKd%h
zXpI3UwO8<gT0iOKeiw<G9H(|z=(fMvaD>2O?bZr8dzN8(GQ=;vAHTs#Q&#on2|_e{
zbL+qA4K&Qm|90%Vw1c@0m6)cXl(h+!=sOx}YARj>ODpSl%uKXY!aAmU*81;gsTuyW
zEyx3kCT75Sn~%wAEg@3~E0*0WN=O*pD)D<@l!()$!;{o-3bR(u5*6MRO+rsDDJz0i
z>y;TQ9ezjLiSO-fnGzi~y_6+Hc`<pIS$RlbuS5lTz3P=f#3$S86P7fFv6AhVeU;Bt
zTGlC#%Jiy*tnL`xQ;tSE?ll?GYb2Rd(N~fJ(KL{l>oBOO3Lll_e?q0bitNKrk`Jqv
zAVDCr??#_+C$)m~M<#(KE2S_jP1=VcgyADciS6;%_Tj9G@R>3u!TJ`bveFjb(u;+L
z5Kbz2Ow5`CmSs#lh`I5VkIgeHkjhH5R+QRy0|&L&{j_yUoj+I;On=kaT;BaXVQe}|
zCL|gu9BC;@M`*Zn+VLmCnGf>Hw)hBN;<TYV%QeXzv8O{R(P0)X-)q3<bSCq_^bc|P
zL6y9LR@!u5#n^ynijyKiN&Z1Gt7UNrhgooX3XMW8rqe7wD#ajA0J&j8gXavb=7&<V
zWL1jDaKQeW;|rt=1xgFLf4RCSo}>dKDPG|=`g<w|2Y+Y|^ce1(A&JZvsV&+w4*g`I
z5zH@`#UpFKlF2JgUZnaDel}QM^+J(8R?c~qcMU=G`nxuKD#wSBNVv^1$|KFn#IIYC
zGMTF(#^27;U5HRwi}A72sbsl7b!_vXFEehIS>+^@U@8z0;6ujzq@OG+i<NjSiUiZ5
zQA?uEg&IR3AC4S@oK9vzB1RbRoc3;-n@$`N2dB~gLQsa2{!G`wl09y+_kC-K%!TP_
z)rDo6G2{K**5y`&?dYU`onQ4py30kLBRMrAAtjES8R7A3CQ=BDXo{cWP$e%M%5@<S
zNpz;*H6y_aQ{*wHFOYo)&0j$A@lDk0(6Q&vjqNTZR8QR9S~=NTLoVcBJ=GVb9q0s!
zP|P}x6n&U5M*`m89iOmKp=vY3_v2AvsECM@Aw`TQ=3C;Ma`v`6p$5{vEK}yN`^^e%
z7Kgp1ryHs&Olx;~g0j|CT@FqsPjYuyeoIk^4hSY@Ov;*tAp>TEpRO;6b@Fjj?YeHX
zITlaWgG1~ZLdL%sPnq)gxBLvg`as)Izttm&#OqXsk`xpbD(omauTf*8d8*i7pAum;
z{vILBucA=GuF1;?%6mzxi0fM5k=g*)Ri^|EQCSY-PU2RcI>hGF1+EAIj!L5{MZBoE
zGV5c-IkaZc=-Y2h^(u|#a>R`2g@KP2mDf3WcnKQ?s?jq23Kv(qF}vAPyHH&h73@3*
zQ{OY+F;On~2B{L0lD6!s&nsLga@`ADSe-THF6Q95i$#Aq2X@~|)^l4t^^J@~zXnOT
zv+YdVypFGf5)i`PjJ45zo8|c~a{m$1F2o2KTf%9)_8}1jIAY+pY;cJ)MEh!uMkT*P
zb#}C#J^fkUGJ2oP0-HdN*s|n)@3pgbeW%PO_SW}d2iqUZNKNori(mI)w{jxxA@HG9
z<?ACa8e4})2u!{eFUg{S*y1n}(xb9nt#EW9cb-+HhlQMP4t5K3<8;=a+nZ3!-#yKH
zFt<v}9?0IE(qUIpbt&NdxM5QBiesdC0Lh2yP|pL*kSIHgZ+^Qc(z;^p!q{w9*MUB+
zIIpkHq!=w`E8n)vcyH$6<OXlL=O)ioZY!c=|K5-$pL6Y`1>1bR0%Dg<i-2k7vq?qg
zTY;#RhY{Ok=elHcaw^8_yjg?`p*UBEkxzK}J3c-1#*dbh_ocCn5i@y?{70pg@4}*!
z6Gq;@62xg(Wormi@WF62u$Mu9pf{?Ha5cdRceSfw*olEQFIAs)+0tD*fvX=QH{OUZ
z#q6-4%3yNh_qI92q}$(CxjM^%g+6dZ+*ZaDWl}W4ZmJEcN~NFVo#%tX^Mtc?3WlBM
zw5i>vk8eFPNVy~4OFBCfg`l#-z;WVYp8T173pMD2Vl`yfk!YY&$L!*ga40z}0psvt
zfcLPHUx^q~ie+P7XkhR5o9XTMLudZEZ!pW;q}mi$k~D_?mV<q>Um2TH2jX`GpKR&)
zOp;Uk8T$J2RHd7xY-kiFCMx2F>J`gWx*$Ur;rw=Si{OnhS$&sqszRo`ieK5|IoY4P
za=tBmt;@LVGL3w!1XqntI~t(j<hJUGaoe@~*=L#<j)lY<I@k@9oxci#i4@#-kVEMa
z$7oLfwo0VU#3=?gqrT7UJx7ZOPxKF#9-W54ZApvkDG3JxN(KahZyj6?VxNs^=Aph;
z%x)ez2|xHR`#!8AJ-i1AUk-i`OZoYo3En{yHA44Z&uSbKEGC?x6TwCujvJrhax9)W
zB%%cSC#fED`g_`ztPwnREwghk1o)77>tHXbQ+6goJN_b(2IBK74AUfU2EL{=*Qsf2
zvoWW)vxY)limX>2jCgMvw5!U@W|hga1}Y^NmEUiPM9i9$-kDbuT^pGuHVm-vxjC(^
zt4y}6b1p@V@s12cjp_67Eiw-f9c0cks|+QaUWefZiWW&I)z)(@1<Y?ye=rhq2=s0;
zw>g8Hx&l!K&wneJ(q*RTKA2jpT;TFShLJqY2<H}zPJvBzFS)mDbvjt+ZtEGG`Knw-
z<`co3T4|WymdaykXesI{m$uOt6qJU|%E<whx@>*nl%1lUodf+<rGF2ivJY}usCW3y
zzGzc(Z?T9t9``Clu&wy+)W*70qoa6_6LcM0j_Vfn`B*<hA_t3!3C7PI;ZZ;D0?}`_
zaNM%Q({x_f#%kWYvN%jDt{v~&pvUcQG&<4u?ikvV@-b4aAeiDdh*l>Ho}e4P#M}N?
z?m<gK^It#WMDyuCJuynYCwPttvGEQ#F>0Wso9yE^jnsn*k2odwCiu-yvFqa)){jrr
zGRhyq%Jwfup%rO)I3}CK5t^aco#*z|7}bK&po_u3Gq&vwfn9rSstvR4=pVo6Rd_zh
zySYjcyGv^bqOvO5XY1;x*etuks|Ij;DD|s#B<%;Qn?t+hDxHV#DesoFLt+$SWbGZq
zx+~6XS<|qG$Q${%zI8*`xivr5j(EWrrboy*oIk0mG9D!)lV*F+g88^DGMs`@l3>G~
z<*}9D-Yu<cZ90XppewvUi?a>RLBiCJCIAic=xj)##reJtQH)q=feY(vCto)KN}X!l
zjAv_eU#sb10{6lc`mJb$^Kr`|9t<_#XTVWdAAd&s7&hsc63PqNW~u39juNSJ3G2R(
zhN6l3&dS3bv3SdyBE$QwOAz*HZ`gs{+?FtV8BJQl5gp&HeLV6xOyLuSsN%U$Vnu#O
zlBfkZRW`H6K{`T)?A1)&*L?n$JN6L7UR7ud+k+zw8-?PoZ#3!X_^d_0D-GxG1Ric@
z&dYPcq^rFv)SnH-_swtyg+z%3JifZA@WoG|52*Q&0?LgL3ljb2&&xRKgYPd@TExNZ
zeTg*Qp8BHk>+n_3sdJt)o|>fkK*AE^zMjFGz6eg&o3#Ou+EWCHQnZU&$%XWgV?6HV
z_dp${eTP_ZuOaY$4Y5<-MO_~NuEA3w-S^A#r$2yUI2u1VYoL>@X*C=1@`%8G=@Ckb
z`quCYDp7*;GrY5q>u%rx!O6>xzz+fOb!uf)K6Q?hY2t{J2H`3P2GTlMo2LbSgtdwk
z&gx~`H@&OAJAtvJ+EOlMVAXL)Gt3oqJLlq;H~2@9-_TEW<+FlR<bU><Tgh5se2%O6
zEI8hhh%CCf)?8i@R--1bm`&ajOI9~*znGVoWW|39ijUOwfBX{dm`*RFqqUv?@^k53
zW{k9l2uo65qpOJ1j~T0T%o1hnxa6ic4U=}v8*G>Ll1aW>c9zQ_ICe#BL-=$4Mt0v9
z65{a%1I=fyhWXhIgKPaoXj#m-I?GXn<;5?6CntSXv=mi9v6rSFXA?$(loWfAIk(PQ
z$YDVds9W{fM;CVw)Bh-&MnO}7(8euqREtMI)`+)X5FMqSB{DX&aHpVHptd*ng5@DC
zD6wE0%85&JX>w{7_S6uqqjk@mVp}dF9CrT%R!~ZsEpmU2WB(zO>t*BTe8R2DBt%F_
zB=n@vwK}h%#%g5R0!0yoBRp{{AwhOWxd4%T0(1meMO`SDiyM&icoUy;{;g2bR3TA(
zcI5=9-&`bnZM@o<lB)?rNbrJ*xjuSwK8C1{jc4>n)xxJquc`4o?>n7h6)};#jfKeb
z+{w!gcIYA5BO`hW*4H5&uYD$1mdtTsZFi;Xz}qQcLg!e_hUgND>c=Ln55gia=D*Ad
zR_j&BEU#^(N^FY8ypglO$z<(#Tz6tgBBANL^zCtcH@6)x^DQiP%#3Ye`tzsWSOj{8
z7(w#XpG{cklwb7)^VWP|!2D$U^+FtSZ9bbaX*L_=L2O3RY`=NcDfGi4iZp7`o^@(+
zr+Nd<r#GIb_Z4rgy<I22o!q_i{epsz^U7g?l1tkHrum2RFpizhT<&Ib;N;(?(?d9>
zc#E)0g~W>O7$yTbH3vH=9RA_ET%=-ny^GDcTK%x2I`rNN9-kUlstPAd#V_Btxxqch
zbQ^al)uiB&x+?Xo>{X)n-8fwYq5_Ouxb`&(qm@ob>O;r<2P`OvuM30*(eA<Q>?xOV
z<q~m+4Q3$4zoyu<%(!!YW&r*+mGJyG^IC?Nd3^kY^;bBy?~!XKr9W`i{iLVk%v|)n
z)t)48S`^2Ho#;C8)^(jX1c$HXjpq)EdaUhRm?1oR$~i2Q{BOfM?Y~OdzfwziAh*=f
zvZj)<(X@U}H3dvG^mM4?477p7k^#sz`E?BR^sV200)G5rU~MI?W65o1Vs2)tV`}}5
ziHh6I*vwMOTtiET>Z^{eftHRW-<LmWDF7SPzt7YKQdVPY9ZM=+V-0JauR2<0+BzH@
zFjm%<IvOT0_Ok~I(9R-PPhBTkAHE2AQlx+J{m>o!fgx&IrqW;5Pl2TXKT9`{H*#9U
z*xz3qzA`;ro|91j05x5^7lrFgZ1*Xu?LmbxM=inPsAuotczbPg{d}{@>hOkbsJ3?P
zMC;3|_nwIV3*WpFzJ2=^8;em`C&0|i+<Uzs>51<ZIX|!5fpW7~+>9c=Ufe*E;j4M5
zQk+t~<a!sQ<~ov>V12a6u5)in`X%g3@59|y6palYoBgrd)0G<x$*G<yiuhM!P^pA@
zs0I&`|MEM~>N^l^0m&C6oOUkY%Nv!4#NC6**Mq6MB>feSIU=+LA~dEfO4=(Bn!4-L
z!xf=WjLCEf*IaU0@Sip%;t^3%6PJyPok8`{2i2L%N6bW*1H591ziuR~;xC@*$uw<u
z{K_TevjNZLm>ZXytRbC#$F9w<e#Wz2@t_=~##Rtr;5-GS-kZ4^dvq28-CBg^3L*KE
zEVk6!9mUfdX74Qf8wcDL{o6(AwzgG%LFGR@U?Ys@<M%X^#>dBVa&n&52?b-U4rb|b
zwB;o{@tNd5Ib_>C(GpbUuwLtKmBL0-Om%#0YHFf495<ghVPayEmzQ6D*XECp($bPH
zpr{yc)K|r;>XEggBnYo+w7-~pw2Z`ZdpWq)8I-_kS5&l=vAVW4Ffd^8dk)nlh$txf
z?Ro>vHeK)L%$j#hVgJtL6BCR*C+6Q@|J#WLR$%pa?_4;}4<Cq$%@*o_gE7@Fb|+$}
z)H=Cvw0D7pXOjPd<9xn7(mkSe(Tmxjq0sDZ<-=%<9Wz2cRZ%HtZEgL2VKy)@u%Lhf
z=+0jp6566<8n=I7;KFdJCp9|b(@neElB7CUqH5pwsGtGU`5K9m>an4rV#P|s@DDDJ
zH(S8qp4WL*y_B`a@xl;7O2|M%^RZfS!ss$g&S&tY6YJAi%9C=9xyE#f#-GNi3{2bW
zBhtI|)ST{Ok2gzXGIher%JZBxvm1ReZr3Y5c~m|KcpQ3zI1@1hyn#Qta58_HRb@+y
z<;GIzkFSl>EvPf7c~djxdPBff<KxoY+}tz6=jZ4CsN|jP?SP3sKfSMq;)c4Zsp$@_
zPG%gs+U9aXCGM9mUmj;fc9)u;3KYuQy$FS(t#`)-9+du$Ejke)Ub>hxX#PeM8Uk;-
zS;Ky{-OtT7A{dX;R7Zy}$Hvwc1P|ZwcTB@AEp6U=PM40Q9&HSAG=<R-5oaN0QTh4#
zDJg0}SoEKYfLRR%%!CVP{0%g)OfruHf?21WGP=9Dmfl49vW2&d)np_jIo3O)QrV(1
zlz#sHFmQ0jGi6z^H><5XW7&XNlRo=ei>9^s2R62bbO{*<Cn|$wLxX#+ov{NQC4@1$
zoa2ee+^197g<7j@;W%2i`x76-p6?LQL^&JFEziFBdzDUDNWw!ysdw^wY(-6@MEaNV
zjP5ktQX+`O9`BUs9ZuH|7wVfIZ~wM0Ws=8emV||cB`41}nkriPGp~feD`wIDn9Q@+
z{fB|!X|e@@^PXD13mFB)(*utDm&qODX?F$VvO8VwC&wh$CDLeh{^M9&IQM!tbW|`v
zbor<0|3X)PV-Vmfh62tmF4&Bw#zsbSrCMFSzP|V5#)}OzLX@vxzh>h99n(@Zag-Y!
zFN&KV9XEQT0K<^Z5r0oa#H;0{P_ApUJ5IW<@Rq~*pP76IBJ1S3kXKZEygwVd!Np;*
z-rl=E`g^O7EEu@BE~o1PZf<TyMtgvZ-Ge4`rLtx6WnJzr*{!z`|1nUe9BOD67ng&B
z19F+X3iFlL`z3b!<K@XhCBOmL^ac}8H~M^he46(i7ajK~t*p)zhgMfsEQ)K8(b4V1
zkkVf9z>osZu(~Y=#`6XlSxQO@Rr~`pFlG1*KGP({&91h#wxvcG*w{uI8Yg37oQ;}E
z5*fTM$1TuQccy*Ol+90%ciVbM3E>{L{^FizT!YZec;ji-dy^)eftbbe8k_xbDs|Q?
z781W!3K2SLC6&o6$lpJlDuvBqw7Z*xf<joi$zrV&z!6d^PUq_q=ep<V0|Er4Y6He^
z!{x$}_Vy?(Ev>4m0vsG2TsB|UWF(#E^fZGnBt9;#C#K%^z))RXywY$&Fp4Y)kCTI#
z82aBX&ASuR)!ihQMcne(+1YV(V*j^!mKSgN9k~7<n#SKK3P9QK$PeI{@Bh%={!ecD
z3~N9z_=8ga0IAI{=Y9s#A0Ph$=`+Cn2GY2?U#|Nc@1<gXg^}lx{)+CuM*6?uroi()
ze{}<$K12I6y4#TcMs#1`_=bNF|5yM2H2ueAp4<Kz@?XaQF8>{TpWFWNEm|C5At7vf
zL+QCH6QwF+X2<jG>+9?Lz4i6={$%!ghqFy~$8%qQ|JFZ?iMpn)shQr~{M6zBO+Y~K
z(yFMa=w>R8+3H6|#@a{*A3Bw)v9@+(Vq)O$&;~eU8`s5rjRkjjCIDwa5iPqD`95A=
zJ!(}(Q>KsiH-Co`jw38477`r%mf141y1KfvdUbUbGB_?S?lcC7zXwyrcz?#9^8uQj
zjV&-J2*B+SJkFw%LOQ*{?VX+6re{ENC6_6!s3=gXE;=i(Fqva0uR+GcOUd;We<ob*
z?O!$O!2LfE5@uy(k*4ws{^3xnZZ0NFfq={AO6P`mraTp?HGdq0E-x?NplZI^OnA%g
z*szZkPj6J^d~=$kSL<*_1bqM)z?MfMf+a@|cP`*q+S-|&Dnf|}s!h%ckCwx!Tw0xh
z|L&p7g-X?<V`CeAX=!OtP*AI%WR*D_&-KmCOBx#VhLQ>Rf}Inus=sJytpRcW?*jIk
zn*(bTMJ`tayk``^EeZ+>04vuw#(s!MJ>XKau*^<QP7V%=CQ^zfGM%7KdH<RB-Dr=~
zqx;*7J;1}U?%Z!<YHMpXlOiJ{<<wCB&ZHGU1<ik>;Qt4X_CG_5|Mwj2nFRb#y!Su!
z&VO^{|Kf@NH!hn)2m<&MoAb4}R+{`9T-*niCnsCm%V<g^Z#c}+e7RzSu^(BPnd@rA
zY;4Z%ild{We^-^#Vz|G5e|sbY3pNCol}KO|07OR9dHu<Pnb9nvYO|%qEFp9!C#Uxr
zoh;?L1D1AnDGwBwn3#VZlYtcj6EnBC_=)8SI4?;pk;ngO?|f(MA9Pj5{$ot?pFlsd
zv-!3<rl;-qCJRRlYwb_c)%$vTcikEXV`F21-nu$FH|>TeCTsxKNvGS7xVQW#PIVkE
z-`|{N#qgO|>5rs8?Z8-BS~fH^a6Fuk-7mp_Kxp7lFJInVUpD~GxHXjGHWT&AIVmww
zkyB_WnVl|x_2lFv&i@q;i6p;(urSIG@>G6FNlAeIgZpQ$^MLWRZz5uETueQwA!rpT
z*S<+a2Z$*rXR~^xr?+?0W`wqM&%gi>V$=Xv!viATF8$rr5vTd`&ls1fv9YnGCC+^X
zz`a_ySb0cVZ{9t7J(}TUv06%K=&@N5*FLxhGy+<-nSz3XTm5MwvlWZ&e)n4XKLBCq
z=kY|l`pBu-?n@*RufIPIj1?elQB+H{R(wsNVPPx#`)n?^*8i|+3vnSF6uadV*IR?1
z-@p!LE6*G`iqD##F6Yd`KLA8!U{^uA8y_DZVAbn3%F4>d#>VlCW~SQO<5@y6a_4^|
zb-;8>7y%zlw;{05bvAo$%j%+{qPu5dYMdmG39+#`yOxLZwLs8WcOxVHyM%!tn2z2U
zs;8aol+DdeLvEu#gDN4x73SwhB=*P_O&SfK{9_#INWc!>goLB%QV8gO5K;ZLwfMRZ
zo@+^U^^=+!^)IA)hF5Vr-1&OD;@dNg{{DVo@K_AS*9ghP9)FN5wTSh)Z21qEN@sc~
zV*ly<PudDV7+RbM5I$R|1nupw2ND>6=am3rlg)9_(Ur?NzlaD3=<5hT|JOezS{j|M
zfH4Bb&p=Ci+Uf-h4Z^`-GFLBDth_v!tF}E*6cQ2w*dTtr*<W+y*-uSKQ2PA&TW2dU
ziI>|WYZi-h`Y~iOf{1?41OD|u4j>|8Vxe&Bo70VZV!tRdnQ$r~M!PK3+5Ez69{;`;
z;5mWF-F>tY;=v?eA|N1u1ln{{t9%KIhJl6E8)GzCP;0X%`<plL94$}e%VAd@s_%}E
zR><J3)q&c0gUK6Od_4%qxWaS?;{E8F06N}%rF~+>Xfi8Zs@dWCaLYb;$FxAJ(~B~L
zZL#*to$@k&{E#iT2q$_XE9`bXrP>yIK!BLF+w*uHL#@S{cVS^66Z!bm6eA-el~Pqf
z&3vW+vPB+Alhf&X4^{JhSvN=2>t8%idSj3I?EIF>$$z=rjmY!>X6GT)qNJUL<Y^QD
zAtJG4HtX$SOgi2D#m4#?j^N)3k#v^S5!h(O7#65$MoaO%%U2L6S{u5O_?OM3k~KoR
z;EVVUo0pujYPGhuKC`J`X$l|?3=b!Ht9qnZnH$s@u=AZD1C;DZhXIIH`ua?6kIoh|
zv$M0uge|`anPecyKmTC*%%gXaJFA$AUuZPb0y+?zL8V+n>OE>p%bN{#Y3az~VzaMb
zJ!{ykHiegBAb*)CKMf0so=|s{tjc1wmDx<+)|ybvN7O}xR^XD<NvL1XO;H4xl)sau
zqa({R=lYf9g>Noz4kW0ksKCR)0U*3i2%z-k*;(5kdYEGm4FLf`DByUtU<ahhMV7~R
zN6m~341+yAzefx(x16y3{r&CjZN8AVq(55#^#`!oleMmZJ|rY0GH4#I?^?pbaDRH_
zk^#cF<DIjy^3Y6ygBWr>FRqhoY-nie7^t(W3-)L*nLU3cwMHvogIF~2=bvGa@EPl8
z>zee!&E+`q=;elSDzG}Cjq-g*3G&Fz&26pM>IybsBGE?aD^_a`{%gvj_VzVlQ||`s
z>h@?g_9hV*Y{c7JTG9Y>23SK_SXh|{pjv4P050SI*c9}eVx&&8FmbAft(Ob7aqXO?
z@VYSbWaG|Po2kXo>R>UNvT1jRlF8%&G;(fWpinwjvZbYkr}^d|TpCw|mMf4~xb%1d
z4=n`$;Of6bkz^~r{_{HoKqyDc#|xDhKYYlPXZS<<6!(?^cMrmi7vc&T$1T`h-06>_
zt2<0h%*@OL)P_YMzs6wvSL4*~Ow97U8oTHpiILHIf@5xshL?}SVQVludM8bW|L_p;
zIUE#b=XL)2w|J8_2BSZWQh+27eG!4o82aj9?rLr6<=>cQkyU~V2opcf?f$iv($Fs%
zSyKo<eUi>$biKb$8iM}QJXI1`TFlh6$lUyFbD(1p>rc+~f2qj<!OgJTLlYg)S@3v6
z-ag!HrI^iC9q81z1>>@M18JwhXy){^{GXTz2tq)rTpOk6>G1<__n|GDBE5lldXw3T
zldCJW8uPTQtc~eXt+V2C!wE82*Z_b*24m2CDFF06uRjzKt=Plm_T1pX&S79Kx30nN
zNEM*;L8$E@O)aeq$2LhZF|ii>zu3?KLgTG6o(X@Vpd}?0NHhkz?NtL%aipv3Us++S
zhk=0sc+35kBS2{>f#71us8FKL1!R({wuAsv1s3t>k^)eVy2%bd5E1E;;-jSTzvsc(
zgMoR@keQhY2)My`Mo-hf;cHs?2A$sSp&w}o>+0$Pe3$cPJ>q_;&T>OAS29bQ(QGi0
zS*70AW#;0O;kZ5!x|EW2jusnhJCcB?tIXqxk3yR1i#(T@i1QK!DL!C~<RywjEIROW
z0=IGGlv*<eUJN2kwc&&zK-H-l;5G4(K#O%YLXW5ALmbF@4s0I<k^GNqNJ0WmYjSAc
zqN6q+01M%IKEl_k88r5Dtt+H2rm?XRm~_G;s_5uw4Gj&ZoNoM8zMMzkoe`wWJf@2c
zxq%3NZO?LvLY;N)2LS+pi-z!<DKuv*4BR;#l8}i-0bbN_wdi!QH+3{wI7be`wGihE
zxVNO9pUv`%lOm^!VW7vZF}bsVR{DCv(;nQQ-eYNH1T9r<`{Pn1QK^F6?hpr1`CeO7
z)6mfD-T-ME+n`gtzOZnhVI@HIhlj;2ysvk=#LUIf13dY@%9fhFA7~_fMg=o2<qh5>
z256rWo=BB;yWJ_&M$YJH9kzwTG0}usrT{B$K-mi4RBr)ZtXJwQ)K@jiD3=W4zFBH*
zR|FjzBfUg-xIRwvtIR;T7I&$S!NR;9=}KAqm8C{k<SGpeZZJchJ;qI5bq@~e29Q*C
z-<)lUEFEUoF2wwpV=(;woZ9kis#{uGBHjDP*109@Rt-uHr(h9`w0A13<w_YN5u(=G
z9>g1zAI5_QLu`yUy?$M?yxdE_M`@fiFyAn+pTP3T<!(Gz3NBxhV)Vk}dlv80{XAQu
zV{fAfah0!o!2QLP8vTA^)=P`sxyE_uvuqbS^yfq$cCMusObAZNnS@f%weIRBiLTV9
z-h2hTEw9FZ9$;mGUmIJDG6{1>%bavLW(-w_PF6btjFEKEXDaoIOHBw2=BY;Q6h>Ry
zVH)-3=nXIZ?MAwQ6ru&(f|{7PtACbDG5KbGrpwdU&JjUf9#tz0P@0B1#ma2V3=U?E
zm$S71O=u<K0}AkfvrZ3CQ<mn{X0w}nwP!Ximp~lY_4ELEIfKzW&sp&sE;mI6<Mw9-
z-=DA{1rAtk36|K%NbTL-texdDCt1<t-{Gus9srd2FD?~dMFmyO#dAUa;})lp_gdq)
z6s-0adTYcxhb7@ltI2>6k0V{;0scc3!aIK4iq>d+G1IgSNy`ORt9YrJvmn??OUeiB
zM=<4X-6v-AW%Q&)e5UKNuWwn5jn^a%`xlAblGGA4k4IKnK!<>@*D)1*v3pRSYiS$L
zjmqVw!>-;8^8wL33l+Uq)U2#?omm2LV(uvqq+4OG%{&9MHl))jWky%YWP_@s4N78e
z{%*pkKC*Xhpykw493TU3YGS`dC_LC!SSpV?s)+Nt4IpdKs3g>`4^x6RcudN+o1K};
zXE3(C?ne4?xMce9#Zd<zn3VU-7&r|V10UDuA5*0&d5!?zLRl6owJ;(YxLd1Dp}o?D
z7>Z`f;fIc*DKKQM5dwQXv|ym8YH2c&oO{;_WLD<;?YTcDFMpVth?pcU-KncfO1=TD
zriwCM_n;$Zg|X7(P&|yB>hv{}|4h*|ymPm+r7gF4R*wfY2&;+>T(0izW({pW*2!dc
zuP$7;cyK)CF;-BV&0N+6zYTfI-U_sIh3;~pSHUWk(bAjRyF1vWH-Z^$aSWz1FG(f<
zy0Qh4(=ZAfC!RIG$c!(flVWrmOk&mBzbWc0&i_(PYBX607#VPr5FBRB@rtc8);j{N
zyUks^-m{~zbS*}M(b6qff3wV~`Xj&(4G&5zjTX4#Z8vMH_wdTM7MeJO7w#3BT^pI~
zC)tcK+)AE&M*tt3JxZ;a^+&VG+=NKsm~z=9>?s{nx`4e-y#j=qO$Jf1<e+jkse5ud
zXDfQ_?CouRAX6^Hl?8Gz#{4b$+j6|DjjJr3t8;IYk`kLbmpcYv*Z@jMHf1kKJ%YbU
zZYsRWQNasI{syq02M`prS>g2{lfg7mPZRJ#R2fTGIjr(*!?P=ZN-C6al8}-b0N%07
zL|?Qvuid?+22@b?&e2+0!l?^@I?bKSeQlJKmeG`MTh}$(Vm_y}%Wh{!2j-MrOr3D_
zqGjl=lK_2EN1DxsZe!aoYFW9|BGx71xN@c)?S?Lo!5MrN(Zb~&g&#0(aZWOEu=nJ8
zyyv?Nqgx4Mpf@x)IN0B>9k8s<MR?>u`#ydsVn^!d5O4v`_<}}MRFn+lkkj~um<#>o
zxU<>}W1Wa<y(@RiV=BNu>Z%OQW`6=MkfqbcyJWW53XIt;rOgLH&R|h#O~8T(L<zuw
zA3(rcapxQL%v83BbU_0m*bGHSBd6UX8Nv(bn>}aRxBcY%TRqB6&NJ<kGVzadV|!|`
z`qLFt=^I?eNI3Ub4%J65+Aax_n`ZhL<pSd`aaauu-#<JTiBvFU=WOq_*Jlr#kDiVP
zqVnM5aoBAGa}zi@?~l1}RMm`F<cEO=8lBqe$=+eVLOaCVoCJBDU(z$lr{kbBr(a+*
zvR3Qg$-2mi9es4VIwWs+F4WY~78KyHf8rQay1jK*+5<7gasqM^5j@s24RcfBHF3T?
zT2lMQ*-@rJ5QKNKuG1Y<k1`rzZ&C7<Sw8Z4=KP+$!%u}^QEc;q&Y+WmV5mq;)t=B*
z;!KO&25aK5aq*?I_Ubw*Wqo!jkFZK3?2d=!Z-rH;%mW~wR6xyb<=qkS*(nawvCvx8
zk>WHnE5ns)siVsGp+zJu=RT0ilQuUbM`-kqtwyaUN{-n#zqH(v(4v`287G6=qoKx>
z{^^G#rz^c(G`TPU7I$s1Bl8Z36l#Sg2Qslhs7B+kA-=%E1L&nLU`ze+V^L_)&HhaJ
zj|&{3pxK-CeXW~^GCZ?9(>*|er~F}Rf$DJTEP;(~Fl#V@Jtcj*U$tG23M>$yAyIq>
z`Ad0Rxp>D-TG3h<8>kLhte(etoxEwDo}ONW@R_nbF0nLZTR)>}N0{K@Kv`3tLVyxn
z1L<?9RU{J(p#uf1XV^`xfdzQ`<5gP*1YYFPg-v!cUaL=8K*Rn`y4~Tt#_yr}S(+c*
z!lN!G6wn}uu(7B$*duYED6jXN@x<r><9i-Xd&?sMuTfL7BTtoUiUT%-@$kYDCFs$_
zJ0w}$cH4Wa>aIaaHu&v7V>bYSaOqXCF3P<<pxB!q0V2B0-NF+d5n)7cJ%(4Y+LQmv
z<E;QPexS($$eXX^ml9r*$z0<!l-#T8LC?W9o5{5XaMWvY+b|%>muQOEV`0plDUsM(
zd?;3J>_vOl*;?LgFGnbi*4{-74JgN6L$=IJ1;$%nB{3j#{8PH7cn<=^hmY`zGpQCv
zu&n4;3`qbVFWzfbx=#H^3ICr-?oaOkrG_sUN1@ho11RLF)wIUZ>48s#$Hn=1@T6pB
ze)y*!G+xDOQxiQsJyX-et*xt8R3HZ@l*PEE=;`TUl8@vgB_MXzF<sLiE%CcPB3bIu
zeC_<ZkN`_PGc&W=yijlF=H=nuJXK4ELow~UvwmXhWBGu_{=k&>(`SPN?$~%tB0C%H
zmwcFNyEy@>G8iKQs>OhMR?LCf(&9b<sO&+raO3uE<{1neD=RC<T)(Z17P|TFu)fLV
z4xpq}K$!>NpT+johODghAY2wW$7nxfVh)N!K;~v2pAn8b3Ho&I1P|R2PU>22T^%WE
zhQxygkWsWYK`<UkDJeQSI)Hl<{9AQOR5W;!B_~-liN)n|Mvv;QsJ#5FJB%QPy{|I}
zyF7rEl$3OaB_Yo0UI<>fa3xj#bYR}0PDwDk+7~|{Gb_ul908!&wy%a!#n)Z`rLaE$
z32&15@Y0f6;>k&k{>VDO++^67Cd`MY%lm*8uxARxXZzT@md^H3F$V<G4ru!y)uL1Q
z&zO$5aT*#Mfvh>mgC`o`EwYA;Ulb9aAX{FZM9qTSmh>?GPCP97$Me%uT6+3IsqEmI
z5~nNEQfhqsFF-1z+S1mR0SIp%Q1E3yfuX$IsRMS#yyPbV)c7nsHnx3Lp6=d!s>Pno
zlG*4**e|xplw(qBy(1Z`_$-y70qQ0mi#5OM3Qz`o|NcEtB4z<3!u%uMKWbR&AhZ0p
z|G)0_KP{#`7pVSUu6t<$karAhFIiq*26m)087Bc|Lm^+H4r~i4{`faP-c9tly}gx{
zm34G<1SB2T6&Ro~AR=P1+7<v7R9}BberWJ#@zsTeh40?tNtavEl-n4b+wxiDWo<Ar
z(bLP^Pfbr>>wp~X?IEF{dTZupX1qK-|4NNi{gp;j{l<xcL7XJdiP_MR@&uH~<srh@
zqtYWSv{!ie2f_~@y4C#4Ti<hg|LZd0F4c$3*DGR3kxv768AVoF3pU4qgs#&Y3250-
z&G&#_%xbfHUPf+VZVt$u%Tq&T_9v@+9ZA45a7cb70rnIL3JUfD90hQGN=lK~qHp+D
zr|I?8&2@$4DCb*NprURDgK6nRnZ{c;X|4<tkZO%N*Fp<W6SO&;ztNG?G&5TQDi2i2
z9ZO3oz(#`&)R_jtnq(t7DQt9D)|fPcW`03IfcCbxG~0;Jk6Dj%S@aPK2nq{FHlx(R
zzh#T<;vtpP3~=6HthYSl7<I29cCs5zx%H5nusI#xK^HiA&b`rY*V303Cp)@$Ub+Fy
z<5Y){u^JG;{a^v903MqG4g~7t@{jDt2MW+c3?C(zxi4JiQ?K@42E+h))M{#C_D}@o
zRRD-xOt(Ll1`tmv<cgGm@Y>;3Q^Qnh4&>pBi;Jf)tagV5l3Be#v3#afiw_+=W`Exp
zCzHR2!t!{TD`PO@)7xXD`r0Wcm**m}IO@feNKN%l=0kD9biJKfi6Si>T}NvGI#piE
z+sDVdBS7>~s<$<fm7OFnf`KMYcZOcVOj>GTP`P_%7^%GWYV<%!IZ_Uo^?~V59vH{-
z`?YkdW|f-XMQA`pm3-sx%>01ehK9!vySlN}vkkAH)HRZ{)$@D<ppKilj&h@6<Xb5u
zg^aBS?4$DF#7;;~E*i7>&74*|@Cj@Gg_QrPKmA7~<L3eVPa6K(ngtle|DLiw|D(39
zt}BbFCM9&pu5FQNAuA4xGsd4<e%KqyukYl$MpWE}VMveDD)5zVH_O7x?UsLUSNm0(
zi1frCYl_QXG8;h#lX;5PiB+5f0ROj;0Z6~;*Z7C3NBmZ199zJaHRNJ2VDHB>i}_XS
z2P7F~CA-kghqRSci@6qJvE*mU_2*wrv7ZX}QzwR)F|L2M&z6Sr!u+liI*WhJ_XiT2
zsS>M??6+Ilub@6)FdVUJ4OV-dul-%XS=kQ&4;4xQ+pHS=0nb*Dh-G%KA++*~>_6x4
z05t|M2-rvn1r1G};X945Sx=J64=CkG`)fSZ(+D_Bq2b}_-zZO0p6-c(?J{6IKr(5!
z()#ix3Qz**1XO2>5__OBZl<gIBXG$CP}CwVhE)N97vOI%Q7kMi{r&wVKZgNy;?)?c
zczahDF*UU~7!Odv`ed^!{Srk1Nb-O!OwSGhe3f+L+#l8-MXoGIl1ib7YwwI^{bO~%
zWBH%&gdYG65Kwqk6(i6n6(h;z28EE2KxBW%-{Fr6l=e}?+ksI$2PB{%0sytt=djsI
zBP9g|(o|odcH!mKu~$qrFa6B}*qb!d=wuhmKuAJDB8-&gTcrbRasjdzna^Q1dy_!%
zObCnzC?M2mb%9=@yaqiBeEH!1_t$G72@EEV{|9mJ85PyGb&FyIMI;D<2!$eu5+n$S
zL=gccBRMNUG6+acYLTNL2$G{@keop>l2ZYagXEku+)=pC-sjzRUc0y5_kH*6KigJX
zRjbw-bIdV%@1xJ7EdarfL;tgm^EiZz&*>0Kl=UgTlS7%K`!6~>J2{nqn~ju;FOOC_
z!-sSBOf&`4K~=rG)<E%)g+&(oWPc%Rb#)akTchON3mRo$yjCEq;8)s8qjPNjAqcs+
ztQPSuFshd&)2UFgvs;O(v#K}V=QJ6uKr=`4){mE3pz3ozIkMh5=m%X6O2Na&rzhC)
zH#8*k=jJ<N1yW5tac;9Ww8Lf1cHUWHq};~dq6AVg)pWSRQN>Bw+{%ibn@HS`>~UOH
z*2>frweweC4UQ+cvLC-oY#gs~7sNL;HQj5*#6<vdnhfbQ+F{)j*a4Ep0gcxHb=|#t
z7fnCzEDzTi%kwbW0S^fPB?Y@t=<YIbAo-Pza9cp6y2OvX2%A?XsDqLuqpxG5n|KH$
z2sCGaNj1lTC5C?sRVuAQCN~q~L^UYieos{%ZL5H!s5VmEDC{B*sh+L(Wg=Rj!+=a{
z*6_RWP-v(@c9iI@mzuf}V`P6T=-v!b?@|GL?{Aoz0%=k~ULN5IQM0j;5!9~v&g$iB
z^*2)=r0&RqdL<dS3V(k|*v3K4lH${TzH{B<9F2lPAKWa};i?>+=6MtRU?lAJ@2ssu
zM&;fBdkhgdaB}q_{19|>I1~dNv<?obe9r2tw{f!pLe2IwaXecWQ|5tv-!_040h>Qa
z`9PSW>@+E-GHy0<@C6Ow`fu=l6(sv|{`~pbrn`7oIXO8&nt1GbRU(w>_3BvFQ`<;;
zo-l83??Ywt>;vx<Y6=t2;j-5?ZUfF3jAVp6afi)y_qKO-oFjHRA~G_TfSNQ*{eH*F
z7sQm<EIs6e=V-aPxmj3Vy7L}FK$}@ix8MuPUiEAAc2M%scoBI1p7Ka_YpZzv2{aKx
zJ@~}67tu`Al<UNp6q!IOUwz^cv^xHwd8F8;BW3s7_<JZ=P;DMH>ma;KZQ}qL0p$Z=
zuD@y4c>(zkV0cLW$*u^;J-?XR&K{si+29(J0TS|-vFwyB<jv<_yZ|DWbz2PpwNa^Z
zfJsq4N>wRb&QvSS8?EBBoL>Yj=kD4h5EntjZ*+7#k0zjZNZsRpJN6`kw#TD}#_Vcv
zWMszpn2KpRs@P&i$d}Z!^h>|kfX44W4QC=@H*<JUAL}4ZG=|)vq;x?ZY%g`MQ$97w
z7A&z^nxC7y-xm9u&%yy36m<CRKc0lOmG2zy@sQUn2{8Q+Pq<HPqx(zkH)msy7yxh;
z^aivw-*`}dC_gYT5HaI;PaE_!wQ6W#&4foeh>SF^8Wtmgw;<;Q!QAT_Rx-eXlN{tJ
zuR1a?Fj&sUUd-0mTpCo^O6+f_i_CyDGtr_DQEhlq<@xrzdb!<7*@j=yF9>ado6qS9
zlYVQyOyQMRyuxZWicGEc-jz!m*;gFq&%Tqq{8Y1;L%wI-K1c|D<HQr`Po2PbWy^4I
zaor>(g`c6&{K$*L)SxTj=Qr{2vuDrNN^wGLJ7}gI-}{oEpD$6?o_kBcNuvk^Ngz~G
zKT?5`BR4xcK)iE3U9D7ps0R`dh=E1K7>XO+aBkLq=d@XkDWdoLT{kS&e4<CkMM6wW
zBXNmQtt4SNrk@{H!4b#G_%k{>Iuyl|d}oInH*cmLf4gf^#*seXl?Zx7mBYis+UuRi
z4K{KO5)<{<-BZYIAhta=_)(+hL+F*dzKP$=qb*zP&lVtg#LB3-t*oe+9R04b*nTIo
zf3)sK96RwI<j6W-=hp#n-!93>XR6+~b}c?D%Xw>F&iDa5ke4Oz^icwon%37Fyw)$6
zjl>S<oS!Y(s=|5d3qK*Xg0Qq8v=A+zl^C)y8%g4v$1C;WFd-(c-6g6KVQbjmoO@_C
z@odRjN_BN>%Ms-5a#y_s@Uu0rM%6Z6CW;VkJ)}4G<5<NTd}_H(yN%T;y~Wyje01PX
z$|mBz9RRebrvRbZZJ?igR8bFs+l?B^(%^&NG*scpwX5Xd;IO!~RA@Z-E!&Ud1HsM%
z2Im&xcn>5};Fyw5KK*-sK5G4%N;bB~T&msE0|gC7RKB6Y-Wpo@X-h$T;O*(@DZ$O4
z8VoE5n!2x`Z%eJU&1|w;@-NDVeSBTVEK@p3f_!fmo|)O~14e5L3zZ6oVhN7WurS*1
zhN#rw#j%eUu;GDU-=c?Y@4iMw+4f1NeC^<%JxD@R86~56;b8i=C&!Mc0MGfX63zfn
zL8L`l(qDbvo33<6V|#TKuk?yCk$4XU@eyp*u&QCX*OLEiY}6Ymvn`Q(wpEn;N^tRN
zTqA`3GyFZ%S!IM7>jN@kE?17L>5AEL^DFN)-$Is@c>l7{M6L7rfw#xCe#VO3YG~*W
z*98Nhyx?uKw@l&~C%UOOn>z$tTC;ac7+ad%)_hzeETuHLw7Yx(lG^a7G(|lvRI+nD
zeB-9t@5wyA_lQ83EC=c5=SQj~e?^_)k=k#6zT=MWZkfsr53MaIIy{E8f92>*h^@9h
z8kOBb4S1XzhK<bIzNdl0ncn34{47wnOZD~j6UTSk&ey3yBxA>V-JNNRl;J5CiBR`E
z2!yQ=c2LO(j@VMc+RSXo1t4{J`6KpdLtX;7+J5r$A2k&{AEZv<BDnEx3cd^A?>zRH
zv@ko*_PYAKS$!<$!8MiJt9wJ`_6i53F7ED<?V_Og+h0d-5Q4Q<oySRX*o(4Riue|C
zhtqYhV~b=wvFA<uZc(J^oybf+(5N6IBTLU-1&t*1AK|R}fkl&bwY7RDN3g;~&zw1<
zNwEhl8{0_l7-+T3@lm3@m;M(cCeLnodV52skugtgn$@;ad4eDzNmhv#6vVpy?DemF
z!w`V1LH%IxQ^Waa>i2J1h_5ZlXXUmV-xS4<TZ$C4PmYfQsmiBk$_@?=yn{8ie-5uV
z*R!c7iNSCSp%EJMU<!i^HTVg8P#uAz;}JQRRr2*rrCcF;*_m}$w3GzI^bQV(m9XkO
z!JHSB_DcxHVlJ~m@nsC`^nA0R_x19vUKQxFVCR-+fK8$D!-o&^GQt!y<vI~DB3mB}
zvpf_Z!V)+MPfX(qjfhA$ljC@fbvr?i#b))D*>;kh${1wCorONhh~DC&qU2sscRCy_
z6;3sSUTXyk5r_*{8Pl@0dO)b;Tm*%LK<tw?LaS6XH#r%fltj}T#E12>s%qSuu|mpv
z72Ys`>4`mYd9cuwk&ncFu8SJ9V0-(%>p3RFoCSopT|QMe?u<G!)&}sA%0=XV4GROs
z*h>$2NKZ<MDupJVM8Ex;@pY|W)3bQ@ZU`2HD>m`jbLSunpv%B!g~YWTex-g`=-zCQ
zFXVyxel+S}QR{&}e|4b0pI)Vao|e|&XCSqv&K}h!WFV+i@JEUC+@_|Wa5B0NLq)Cu
z(4F3HgHpP7YGbC|yBRGj+MEse@S#eQ?`V@_C3m>4TEP9PXY<*+XQKBtFfLe3$Vyz7
z8Lk3-YqkSa2dQiT6W^~u>qE>@mnj@pKOBB2_P13H!h8B0zkXn#1fB~NEbr7KmVaW`
z1%kHU1DltZ7s|~LzGwaLw0`}1EXIHq#=Iiuv6HQ-T3vk`90g>pIdJQ&WQ3}z@}W>H
zDDoeRQeGu^igiEd$_<tT)nap4JhKojz{5|uYB+A8=Pin);1;kO4>V%8wzSlS3GES(
zlBSNM#WZzO=%)GP)>@?@NAj_Qw5{hD0VYH}q0VnUd<3_R6i4E|qQwVg5S085k>1_g
zOKBs6$99slAZ;xbajB^<Wk2(lr0Q+-6?7IfI#UwTASWt`AE<sJ9;tNJKx2M3yQ{Bl
z$7A@BwDRd7w?8Pn3bF;ShYe78rBsxHYVy#T!r?0ts=0}Es1Ex%A$+}V!MF~cT1g%x
z?S1ScZXTYxYA8{Y)o+*`?OQ^$=<3dRcwq4da~Q%1#!aSIPn<9X?}~*=rOKwD@8lG$
zW|&99J6SR0+~ruQBNY#4C(;d)mX@|&8PU0k54WrA3OT1NNe}ccTdtraAjU2)E9)VS
z;&mJt9#*rvZr0TY6%jG#y=UPpI$qG~!GzPU24KDNP+k1|OjG#{2_m+)=H@I+OnBJH
zk01Nd%Px-wSmR@$yrH?+9jiLGHAP#uF6=~3>AZf&1@Glk(2Y7CAKLD%11=4>QKQ0P
z58}+}A-q-ODda<H2@FtFLxZ)oxrx4AWMnN_t`IOke?pKKn2sU{38}<DFFr~99|qr&
zXHL!>*$q|wr};pZf3u$X%U<Ro3mkiS`rrRzjt`|G`gyPjzhXgQ75B;&){?d4u$O{e
zZACecOdSRX1e!b}ku;ugw8RlBDegx?VA^u%zS81Q<DDn<#ns@$m>+MBao=&@Xs9XG
z^+qau@NplXoT*yHX+80sekzQ7B=p$O&7ZL*%U>Be+5VL;IY0k!?T}|R*ble&eCF$-
z+R#a)@aOsHCP!!1<?NoyRaAog+>L0F$ez`?(mXL?Q;LS(Dhu%y=KNQb&a=CN4biwI
zSK`&9c@f_x4yyE}&Rgwnn(takNpRgt>2!pHSllY|tuTTGNUpqp%9^sARXI1#2-b-F
zbT?~OO;h#)4Q>*7x@FZwU!_n!hbtH8E)m2~&edD!RwW=cq$q9wk~l~6`34&q1zxyS
zz?Q&V0pqCcT`}q7$Hcek$YkV>#nVNYzPyZ2x<l>DD979>ik(s=no|Aw*^j;%hhV=(
z_W}HA`6808cyCmsJInB{Y-;#a>Xkgs^Y^<1J!aytZZOAxjKg}e^x_QRjWc4^*t;wx
zQqdS6eo5xv8<wcM?yaZYaHfyuXLi>IBWaangBn$}efF~=7ss(Ev`DDh){WxUuNJOO
z=F8N|y*ne-=b<w;+bD`3WRG*P@Rg86tZtQ$40j)nuKq6XTy{9Qo9nnaGCdK)E^XjG
z|3`wNTYek&mDhFV?!LwU@rI}Z^Af|^52&_X{2IeEp?6Ih&yTI8UbbGf84hFXT%PSQ
zC=_PDmOdiW`nrtme4agL9=`RvB0m{764DLQ2$K(2d~$l%?#N%rBhtQ-V0ebMK41RH
zG90(fU;Jd_ICdR#o3&?<(t-2xO{)2LnUuYV=Hbc4+*eOh=Q2f}YGZ9@-wQurp4lAV
z&XCBTN@6UeC+3P(PRJpBYErA|<H-Cjfmka{+vi%Z<CB61{&hXy{N}>~g=5cs{4vC?
zrB+v_jkv%Wn0{UT_K!EovZBhOg+|3X(n*3o7KklMm;L~zRH4nA(k-nULqk+{3<9y=
zhRz3oSr;p<OUL_X3s1~%-%DGm`w|-3*LG02?IG8xe~9ALy~Us37ZMlgg(vm{{~W_Z
zlIs|?qxKkeIYk)-mR(OVZ*>seN9{HRMdu9M@O7FjpnjdRp!E8Qts&RUxwHTs6Co?Z
z)O%)U9^EeDd{3?TA!Uf}hl?Q@|E=<mcQT`D<Cq>yzS*&SxaqySCN0QC5HR?|-?}NT
zl)@<vX?wD&nsAjTzs1u--owL=()_{lt+>zWUxl6}trx#|HjaAc66`U19shBzFCq4k
zDj!i^O9W?upWRoVv4ens4ch%_&Ks3y?z;O#lm_6_k{_0xT>duQIvM_6+jUjJ^XCjn
z)>)CeQuhQ8OB%R6$(dbO1jO!QV#?)Q-zU^xAd630G-d4^@nEzvI1;ZWpfCye&b^5(
z>gA0raAmaW{Do_>d->)rrPbi#aH9xiyVA~e>^lj2DE$G^L#3zD1V^HZa$gFA%xk|R
zLmxb?Z~YP3oWyVRHIyT0u-;$2%AcB9aziQPM;0!VfM-a29@Wb(kK0@itxW7npSH}u
z=Zgp@5H1P`pSWlxMR#Q<*J5#?rL<7$d|Urd{H8e$&Xm?Yw<Ilf!B`a|YUOk5_K{1L
zH!)q_T}VuyXhH;0)`Ui`T(+bA+>=Bc5lcNdL;vnf=#}3=x|JD(xKBlIdt12@HwF&6
zs-*GmSi2BR)`kYSN!*kYc}y2$uXFFYt?$OMVX8jqci&&}zY!wm`5G>zp~T)&e#536
z?(YlDIA$-aHi|_C3NZXkH6k`pUntHU;>TiN`?<revfxw@qTO!TdR<RPFCl8<qPmZa
z>V`y*7DsR4#&X?Gihc*xyYw#x1ZC8-{qb`Ss1aeErGv;3OWj(D+<V5qJYv7lPo2lU
zL@Qsdm;GEVWVXBlKOj=UHz=!mfH`c1lEtPJc`nv${Cxj|!dIJ@&wb_P%J;cTPB<Ij
z_~52qS0n#^d4I!=ZKswpXHgyT`3OacFBjumev)z;%}mtdlh-NfS>5lSrcrC+w79mE
zkX~LYS7&D*@8!$Frh;X@w{gVLFd#<nB=Y^Sbf8$WyX)?VaqZ6C3kz>zW(nVRj1c(7
z4awC?ou3*aKQo#2?73?6k3d4&1D#v%&A4W%rfW@C8eJ~Ef<qw#o76Aj44+|rTs9J-
z<kIhPF@$HLNpP7n4U3A$IKH9Qc(aV1OSbcx6hG^=O{0X=VoWQk<@!g9%nM8*ALKNR
zwwTNb)^?v<pS)V_>HGG(-}Jk}ezIEERuc=|;-}@2)I&El9veCJ>#)VnYJ_+QeIiVw
z;^v&n#+@FS#Zw5i!Rzj+(Y}K({@|Yfnz}@`G>O@RbGw%}C(0tiEz(QxO+R+MoYxn8
zshiw;gr8m3;)UWs#skxfo{Nl|$IqxSS&L?uo_HKFIeaCY)qA1IIeM1nga=19<zx7|
ztZq3y2@U6->Dwt1qSqpPH!m4bCOOpFcj8)nU9-3mGqand^=+Efw{u=W$CA`*!oTtA
z#~-yN-<B^HM0{yl_RH^mhggg27R&sVqp3}t!?o>hjFd}wC50t(aiEd+THNJrlPf{#
z>Y>dTY7x|XeqmmgQ=c`Jg;ks+x7_Yy&fqmLP(^Sy_cmQ&_<gSRdd#67d*311ohl~o
zU|Bm)8y&j`lkOpB+wI$2Wcb^k7Pxg@GYnoW_~iZM8Shr}cdcAjz3PHJg^S+HR^QbF
zw7Z!<-Jdb~{naz;_K>f|lMm&A{49-@I8tjiSJD&S#ayJn%P8&l%*<cKh*^xjKX@w#
zyBB-Cm}os7`%dhV7x8)N((=JqiOa!D${c;Q6-2G7w;fz}T8^JfidDRQbKWMt?@DrG
z$N8r>=iYn~6)YmYar==QM+l^&+RKA8hdB!bysR`IC+cOr{VrOSh|K7R7C2+K7+`J?
zhTlIwS+&KBHKkd*_6*-gcxFI5-$%RYwut=Q=;8g7#@q5@H!e2!R2U~jzQ80pDp&NB
zinV%(KfhNk^8A80bLr)NMyhirGY?$D<EG0f*o30ygS|3i+yu6KX{5c|KVONkU-S!7
zw<~=9^w2A;4Aa^dF==sSwbBET5!@{!G|>1|Ytzf!hH+QXReOd-HeSAc+|nn{<#t6G
zjz9?4Ydu2^BD}k$ByXCYe$Ar_8JQ^F*M1Wb@l}zJ^_fm-)QHVnJwu9PoVV3@LE&-w
zcd&gw1X*qS*}od<mp#v6Ii9qrPq@@)(Ime>?qtQ@bB&Ly<(9VKnhoir{`E@+#C<U`
z{kSDdJ!fq%on(aej8wLDp>m&mAH2jeGR!?-c9TwmQcyK0Y5zE<`l%U}N}1Vw*V$ab
z99Qa<An!Y&dR+XzBed}~M{Sn1V>2vyjTt#sWRZyzc##c3!XuvoeZG%o|1@Y2?`3Mf
zeLH<!l9!i#^s_G0#20;mU%%Wt4jidXgB~3@zqi0KA4t4dMQyzPBRePSgU6COnN70_
zJw^RKZOJB^+H|GRJs!Qf8WM!!&Se(H#G7RV!8lA9);|<?kFsiJdN%WZc6BZ(D1ChW
zLw}|Vo9<}MG_#GQ7LUm9{pXl8Dn0_Ido;zYrD`?1D?5xe>Cj*VYU^}UhUQ<fi|cSQ
zylcRhmC>HH()YOT^+H)<j!V}5)?t~M!Vzy}g+{w5@ocI_;gIR8-qp3#QPG7`zHJJu
zuLK|Cm?P5}1*@r9)2oAQl;hU-R*wsj>vyLWGUh{;1=CD1)v;fCOf@|kJzBfq9(~-&
ze^^6$(X-Mga>>TG^~vuA8}m{r>#INMU{|$@Hhgh0e<*Q8)R2{t=9Rep&dmJLW7#&*
zq66cx&N2$^=a{$6G_Qwf>d>ELe=8ULidXSdx<oXTqwHWNq}_RUQ#XD$+c_~co#lcy
z$r0mJ($wp7wgnh_g);Q^+V~_}i!BeY_Iun?NX9MrC|PsHHh@$#mhLD_<Ob&l^77@^
zudhVY_FgTCqZKePqh~mBT*3~Q&A`Sl@eFWElyAzNqv6wt7UAnoI#EkomvO#%HQK$#
zJx<Qt@0ZtY#18_ENL&w5WwtVja<=j}hWH6GpY*;PP}@aq26s7?Y01^ecrMoPOGPon
zlPjAVx8x_>v}TwnFu=}F_GG8i=&UhHS7qr+ri>EuZi;tl|EBlI(B|-XE|Gi3HQH`e
zZ6VHuaH}Yf)Al7-&6C%A<uR(|z01T?f)W$jflcM~_-*MQw29p%BDgc%&0bMh(ILz4
zQx3%X7`{mr!7AgWAEEsAvQ*Y&Mzkq5L28sXWl<(l{EBP%ukh?x=?h4%oUg=mIW{vP
zDLg$!b>Gxz_@qW^jE^-eW2s88wJ4>yZ}D-_vQ=O_R=ARN?K*W!$wJ#@6#i4bCOevU
zNinr&%0+&;@25w4e>MzHQN(5<r7^lBc!!e3>-fRR%=9YXEoU9;|I{Ke|J%`f|J@cU
z?Ck7+&h-0Vv|dY*mm<V<C>mB?QRcM0QP3XUQG0Q>wx9IidQJjC3fBFDiVy}vDymYW
z7hkA+`T{*Z)$bF0*4A{}(T(C<nH&1bOxjF8y3)`v(^5P$V!v#0JhS5%Wzoj7)60PI
z=Rc2YKA>MR9xMP>wRn&QJ&pyi223Z$;5hG15$BI%QaHRVlmyJJaNsSBrsL}5gRb*B
z%orEMW8vWVI_&{&>$uVSNO@*#>=jSQ(lKKZ3>}8q0ouZP;YZ~iP(wsTMS=B(m2<N5
z$+2BBp!?PbVbDdJ7psCYvSyf<8R!tZiD>8J8%LT8eOb{kN(H?)LOxzRnpbTB1hb9}
zhX8GsP%6Su#bm6>aEZPXz#Ce-*H8A^ohco?xz>S!nDayrN;_qL`h=H14y53UaWku4
z$B(=x`&ekJmFL9hS?b?EKU@569M1^=Gz86%443w0$F@iFIpg+8sbQBZz>EX=MxoI!
zO?fStDf4sZ1X;q~0Qto918_6oGu#4vfk1;r7x2S@bY<fS;t4E(k-8<op@MqHeZ+BR
zIFu@t>*Wmi34q7J&x7Apz^UkG18B#cE_47ReR*eZuxMl9;uf|mL=S1}hdfXtle{k-
z=Q2&#4?hRwRdP2Ct_{)P!U*;=Z=Qq2+@0@ApzqA;qnAsSO_<W2iL@T+44l_tGa0fO
z`|$y!BQW@<6+kJ-Y5|knemF<N)}zBU!H1xM*a`#C#=M1s>DBvRRshbqO~#(Ra9OY7
znJ4DTr2($I>YXy&i+D-gyIG!$2*Nk?S`=j-!12O#VF|b`xLzj8Y43Z>1tTy1-0RO}
zC*$oK1vbY(gs;UieKZxK(L)L-$jNlLDH-n4Nco@>$+(k^lz=oJA7A+*;KEBAD8l*v
z-_0SCh{E<hy`GvHbR<}F*L!2G-J(L^(!L-Rr*PUG2gDH2qS<3rokifS$ckfiVSYv*
zSPZ+k&IX{%^81}n4m7NB)gcOing$5-csAaDu|bmcxJxEgAuF>EsFU&qQ?f}GWJ0_<
zkmZD*-8b)V`hohK48pj~cH=67o5Zmc0#0Nh#x>(HOo;$Do~D#L4X!_#zw;;x)Bxtc
zo(WUuM?_9XTMGFj$<pz>yJJX*IZpKxGz?B^Fg6#OHN6k4wRp2evALkd$B!QuIScGp
zfnNjK3Z;^+oZsZsy9L;$=ok7<P57<Z1FL?M*2W=#-1vFFg2ABeWiO>cUdR1eHW0fQ
zsY_LOZ(8O}D=(%bMR`q^FW$6+X`1Dm${7s4sTWzl$Oe#cyuH=6mJ$_L-JNPM3AAi|
z1Us8uB*0Pm3G)mP4}g&^^rU<yGiKD!E!e(_yu1!Es&FSc6!pC?Q_Z4m1z^Cfif<Ts
zns#@LKLX>bDA~u})udNyNC`GQZl$U0c?HuR_;+M6w)gYwr34d&^IL%7fVsjUjb~}M
z&*)S1(9R9#muY<frdgcMd7}1Gw-(xHK)HYZ;cSi5-c$&T5R(8YUEqD0J~bajm!^^6
zpR%J2W+6a-wH|`n2lQ43t|+_SH<sG2B?3U|^d{|Yc1r?2Hfu9Q;e6Q1Em#*5-02p|
zZX8w~EQXy-Td_NVo$HS=q=MT-KF>5Lxp420`sxA8@(gCa%)On(b^}<W0GqQh!@#aP
zCXvRQ>X_F-`94+x3fj+<moDpO3~KrFOdls-Nz;WJ)O`5ZdRdiW+NKr4OQ_pqn(#HK
z4;Q(zrUC!!x#CU8(kp;o9MjLQ^^!4s8_Ta$gGgCa`}n*aw=J4aV^VzKKESeo-{-5-
zfStfH0v<~Y<*+hjxryF|aE67k{62N>=?xK1N$xclN86)fa6CA`m&RPiB*77?0x^X5
z^itetN7EzUVKQ7)hDu++A&6eP9CW59{rOD3plPIOy02`*=}Hd`UpV5mO6xhV|D^GY
z^}11Zg@ozH;oLo^*m?!h!BGXTP6w1Cq@Ikrp!Pyai5pZe0&CYU+jY-cztn%ubG;$A
z=bD9$6pkSyZboQ=mUASrhC`=1!BZX65D%F4Mot%k-vjXAz12D686Gleq{M^KneB1g
zW~yTee(nbC(j)sDFfiE{zs;7%VYL`)n|YU3HhD}@;7055{x|D{EqeL%(5)BZ3fo_G
zqs+>3l9FZw%06KbBw(Vfx3;!WGC^MfoR>?I<q=5Cw!*Y0<ls&|nkb%$>LzRLd-8Jf
z>e6=z$adho5)iUKQgt)ayYIc&ix-iI;x3+xNyRL3I<kwG^&Rd#5tj`i1+-ILG}U=F
zR3lGM+7Kt7_hp8;CGcpM$;vx{4IKe_28i*Vp`U1MNJt3VQ|37ujpvU@aVz%c(|W7#
za_7&4@oi1)KLi+lJ*%3Z;a8j>_QO|s>aD{_2<HrCb=x?SJ`sbMwI;f(<TVuUGD7xd
zDsD_>4ymfvqt9_({JAZ_6tM+%MG#PnC>oQ@{rE*XMpv;JJK}U;d|%QugS#)f>7Hd<
z{`<`Yg;^NCof{7&X2mUzPw%$yRb}1xFyJO>b?a?vwwotPewsHx8uXQ#fvuk>Sw!Zc
zfnt()F34J26^1z_@UN1xHYU%5R3pp=Hqs1tK)Mqi4+zOmCFACrO$r4!0qopUC~fxb
z#b+S_BX`WL3h>>DWP1Iv>BWUC1$Orfn;*kEo{MC~NG7Pgh;P=T72it9NT)&1fG{Q+
z7lNXhO$~g6mJmG+(#FY#g^xd?r{Lc}l3!v>92QZL+KRdUiBb*$q{DzMD%{*giY|*G
z)4J2C9QYWEAN(I~<u2qVb8KBEP5=5Z$&Y6$qj024%+czGyoxSpW=JCp>{H^Cg;DPS
zbyt3pna7@1l0J=>7nQixDcnt~zta<y)}Cn4`F3QSKO?p$i|2VbDnbUGJ=ipmtx#~J
zb#7J+%z~D%7KtZ-`{jXJNyN?67w(wis%i0t&KD9th`^4VBW=5ymQ~Re4PH#ePDdq2
ztFD&>tZ`G{EoGTDsl1-$Fed!Gft;p%(xAk&bgQ@H$CaGTq@j0(;<^&}^Qe~Sfl`zT
zyD)YzuKaoZPnr@}!vq9!-Vbm?w)Q*6J1x?z)BHtniXU8wTEB8yrXDFx<9#m0%okj2
zPzHfYIHlla>Toq<W<I{;*w5F$OfNocy;TSzu3oaKD92es9PmaenafPO-*eK}!PRxn
z2mT@GX;Rty(tb7c)FbDSDKyh0-;?V$Kj|3cqjxx8?JfGHtG&|C@)Pov?$jwtrYwiX
z2b-y~81v?H+E~lI31F`{*yCr`@1n97rUatEy3A-$amAHb{cVdGS%=4ElpK(_iL`01
zQ^NV7cEO#hS4mNORiGCyUh$l_PNBxWPCvUE!(ONrKf7VTSUQQifFW(mWT~B;yoit$
zyDy5`vZQd(FS?zgr2B*n!CG_vAceIC>K5`8Otw`sKA$&yjaCYMKWEwUwFY)Xt_efx
zW_M1tPv_IgE*(a;NN0h5Exf<8zoSb8INl5ymx0lIMaXtb0=V`@q&;-K4r!lPR>y!7
z&16WlF-uS2tO>CD4kj&reG0;Y8w}#LL&$+4(w!GmevB@(=ki6&T~Oto#h9*cZ%9uD
zaJ%Q!F5)7Ltb5heXtXN6=72|lJ0>|M15Q^VDy0Z$&(w0@OSX~j(^#JRVi!zVSJKXe
zrBJ}X8GgLL@ubq}NK;*FU@lM?Z|)~=DJI*Y^<rpN*}B&>OhIsl^(h!G<Qf`Ht-b<^
zl|0ft8FqZPFNND#pp=?#w;r*j2}^o5&!16NLb;#t#z?P$Y>8Fag&vpl>dfUWqx|k-
z4TlSFas9o+HZ(vr?=*<hy}{bRGsKhZ;<-t8?hT*iPK=h0SkEKVk;75j4pm&ngHS#q
zyBs$clXCCvQkrOoE{8lQRU9DEf9Z(V1^Iyd(9*BmhD;2LFny4Jg*ibYV9Rh|!cAE*
zmZ0YxkTNyC$QWSv7Ruf?{mY7&by^9Co?8Ri36Dv1)FcTkaKrBW02bEU;F*zc>T(mZ
z_r0jVt27s~kZ_^NDk3`h<QDVkmXzOl(YQI$sqBcLtl?-A3H-3QfeOdcZy>BT8BHgR
zHrPigXyg;OXLyMEu*x7E8!n0X8jvIM3W{7w=R1Wd*tBje$rF6ApXOrH`NZdZtoMZ(
z@lJKxK(69(zHX8El$UBcwH(v5o{!~v+c<w|?~<0-QU=;D0~82JNl8bO;knWBUyGfc
zxBU<>?)UwToxOc&gZJLkR?yfiB47f=j&1t6M_J0%;Pu;G%`+I9ll)t0t+q3p(lOkg
zJxJ2<S;-wy%GN2@WyHHHwyfOmR8emg5UdJwDn}}bnFoEzM8RF3F1~w{)VmH8HwfbV
zj)TE;;>`p3PC9>(wRML-GG)j8+ATKSbj`w~jWjs4x8pl!7rO)a_1)Qu;Y~l@+}W{S
zX*a|cvWxW5`U$d$U;Y#ILr9@2v>l#vqXTQ&$p$009>rC?$gOvl(&?Ywq~ZBaCyXNt
z;>#Yho^`x#L6lvC42a@lPr^cNbZ&mSEwDJHU3Su|VyZtt_1*IECm*dVFM-c~G-4y*
z)N>Gq|LLgg{aVby89WOy^_$iGQhKZzlH|TPa?;9TRZhaVCC;XrQ0Z?Ui3*5Xy!K2_
zZ3@B@FHIptg}k--4t?9QsdigI+68pwE@>QP64p2%G=0n8l9O#3!AgV>FH<DfpR9V`
z&daFy{mISh5`z1X@s|58V${|nX)P@hkH|H;$N3+g&1rK`($lcVK$H=!p_mv<S&b!X
zGL&-fcHQgM?a4Hj7vSn4$Hn=21})ejGP!*1;Kl>j@03q@-sXR|H1AjsD1J5RckWCx
za$upA`nq2ol(N_6Y*rC=9kc2J7(4Dcy6u_nn=i9>#xgt@8AGcgQL$b<NP`!?8iYQ6
z-rKEe8k)2GuOD=GHT5M+Pi8kDa-++7#}q7`)+PiSe)3zK7tq2vSO_%)FJJxzJePez
z(z#IVSDQJV)=ya#{mL<0d!bSBEKfyoOc$lqAXy?OQF&b9Y-VJ!Y!kT!0rwy7&6@D&
z+1(|*KbkEOw-z%FVi?j-utRoTQQnKA91_5#bq<;t&{^KqdFzfDd~=y*KmS}<e85RH
z66WO#ce++Q_Fa@Kpwt&(&#1KX>>O=+$@x%WV04Yna++}5<<d2njv?3Z_&mk-mdVt(
z{hmc&qm!Vf7`I+qRPxbMAN9!o0cEuFconoIvy!fgVnPO_EP6hRHikzjUtb1pwbm-y
zbwqgNxu^c>@VoO$2q%@fp}X}`N2q;=s`O*qb!|t^g|~V9b>EJvRbLuXAVEX_%PF9;
zirz{Ak@u-dj%vGg=E=dwi~a-4%Nm=dD!8Pcj7mAR<t>^MXu(2{BO_Qd%1sSE`g-5=
zVVK`#i;p9n%<_Cx;NCZHQ{?=574A-~uVJMa>8))9;k>5cRyosw^(Gmjr!CX(tJlzR
z##@=VlQ!hEkIJT_gzyd*c;Brzu8!9%Y219LVc`vZ;h-?5G&8t-Sj4<I;adK{4$dmi
zDPDS|vgPqYUu((NsK3QLlts)|J_{hjU|-p*s*mu22HHA)C6vtiPvY$xh-o$Km)Kg3
zI^$f~XxtD%jp_K&k<hfy&3QSd;kUS1MHsO(L_i;7M9~-O9RO+W+u7EOInif)rf(Zx
zJ@BNxb9Q02ikxRwUH_YDO13d;Ya)riQwxazZUv}A{qkR)gR!AUZV1*S!9|5HBRRv*
zJJw^YZfyDGWfhN}t*%~l+ot=mbjNPtWp;Ad16tgYM_GGuUOb!_nk!Y{Z>HV+rS<nl
zoS=6Q3i8Mo>`h(P(eR?J9l=28={wSVZ#?`4Y6d$Ei*H7d(5)BvExi3{qHLZ9a8aK<
z;Lmu@oW$Nz_?}!L5SN@YPcd6lnU4AgbjoZktbL}UuFTwT^xmq5=t(@h0dnctC<JsJ
zHEuZF0_LBs6fF}#Y^tx8zPwxoKR>kQJ7iVz$L89w%>`|sZErXhHC(-8F<l?4PzBt-
ziE``o$zrn!Dq=I2mUPlRZytBt^Sp_d!z&h3?(wusJZ}!~E(qijv)w-SsYV-BMzEyf
z7|$nwNWyFM9Ol(Vu6do5Ex5hYPok^Vezo#E$Jy?jg-(}EIMN+0<4w$g%n1IA;XA&b
zm>|ARzH$OhIjp8Q5)Bw{-g_XH@Y8g3uI2X7j~_pv|DE5+1T#3((TZLs^mDCf)k#!7
z<7l;192h2K5fDZ3Zf6X1FZl`VB_qmp#k%c?Et>qFEdkdKvJLO3Znc82Bc@K4!D6Nj
z<zWw<?VB?jg0w^#1^Fdp1L@N~--SND`F#ILPcy1CIW~_Uy!5nBD79q~Jy%g+sVCQm
z7NuT*X=&5t!?z6<co^39d|v^pvN%d2Pn0S&8HRMwA@RB*Pp3jO8RH#AZ(Rox=KSoq
z4t(z~+w4#+^qo)7!chjZlZK-=EU=gQj32G~hO^e@wqYoTcIyVHU(RepqDZf!A(PCz
zUYkxV%87t2g|3j!>acP1*>9&G=ix4BF7xZidJ<5lFI<eTjP}W?^YN#LTc}GtsAenm
zLOu75_<}M%HjX#z%kTc{aRVLv{QOEe+S#M11ke@Ub<$d{QgM!^><Qx#kysd8STL+;
zdcC4K46c4^90#0fXHZzoo3B7F-n#ro!(vNdh_0c$l)Cj-baKYGkpVT^4zj$w(bl`L
zCP_mVPC{)uu4F9^K3)1Ik^Z=Bk;9z}T<~6WS5&L3t8d?A)6>$oC>@m)+8T=CKX;Iy
zK5Z_)P$M+SZ((tzp<J)tk6g5|2BPn`2Q6(EwhxM@TVR{<?jR=Ce2run>)dBaUU0ig
zo7{enHmbTd+@NPi%%gqe<M-mtLE!}usHeORnw%#rFM1<nCxe9(;N&&tNBn$&tu-*f
ziRS5E0SpQ_Qf^70P>&wH_qq+6cyF^1t+=Bi4<YJ>kYgrP+RDt-jv$AuHY0oSGo9UM
zGYmEAe4@5wn)#Difw;etAzAp&QK$$>tM!Rr!nEF0yul4F0<VI6p=-*o#fuIE%k9eB
z$eoR<ITbUcL+$ueDYq^UrQO@d$w)HfW`oHJ%Y#<WW@z~3u>0+weT#@wa@S-3v~O>B
zuS+EX#tt;jkr{JjRQ3ul#ZWI5UP5)eu%J9*c$g6-Bmj6ucCYG9p6N%#Z1J*6>eQ)c
z{I+yB;fM#WhF^m!_&Ybc9U92(H=p^2vYzjb=^{iHR5(gyxuJ67_A}_%)>Gxu@~(m)
zi~XiD*S9a$qgML~q>P3yGIHm+yY8^R!Ev3W5k@US29SN-68=bXQeOB`*P>ZKMGT^A
zUcWmzuHtaLIeA(BJf3jJaVU$9$QHrmqVeJ6h3UX^4Re6_dKe}Tj)K`Lv|7nC#b7qG
z)!Q;&8!O=6#lbuSt%VdOD1(RR1@#m=VF(Ks7gs}L^m)S~1_qur(~d&+XKvf|k#c*B
z8u!xbp^aY>AI@GNtGo|5ve{=ykMv48&uYFxPsE^{cQ5rC9v%)LZD_3+fMyLoHy#=f
zu!?sF`~^Fo{;!CJ9>5P!WPuEH9e}`M5?l-nf}0=_zX`6f2`@f_L0KqZNF+{dBp4XC
zKA3^V8kV8+S{*U#1q=+TrvSo%a#Od`8y^GXiDnE`Ly&+$Iu-HARGrmZE>m@{j#$I!
z;1TG%DoG`Gc|NA$aY5<z^v=*{g9QA0f?=KH4Xp|kOasgS0at?S9rWOE1S?7tpiK$@
z4-+bxcPRXb5Apd)YW+Zp*NK%)1!hCz0$fsFxa2QrsHmtUG6E;9F%j3%_AxM$BJ2N6
z?CZK8UpT>|y@zKn9%sAQ4>TZL{SPE%&Lr-RWuIKPq7*Kd5ydn33|)XZlu0&tSQOmP
z?@v@cL;IY8$l0JTa~RZeFyH_f8e`a<JeyCevG}XE5<M`lAz(=<19}8HIW($y6!>p{
z_wb)TuiZjWIbh(v^s>Rjo#3y=Jbm+j+l#BW5DA3v>9t4~T*P@1eE3T4=euM4G5@z-
z6oI?*=i3tq&;0A#|0`b9#wOsk`SkKUw+H5j)hc8V<rqIV7W_!`&s}=|*K3ILf4?@y
zf6L+YUZ24|JC4M>FUXDcugA;(1Nj8*3EZ8ti1XE$r_W~f79RTJYmx6UFHw-5`M>qN
z{w*-zX8r&0y#5J{|M<Loa?fL8y1Tjk6~5e9o=wR8P)E$GTR-JGvQ)`OKTgr^Vv_OL
z1;MR5jqvAG2&3bdAwgRQC3iI}#Y>$IM!8fTPBB=)u-%{E=YhHQ>f_EB&`J)M<#wCj
zJD-U5ZpV+T;Kuz4Quvs=Hn?a^YVO>~TbmBH-M7I5AA%?$tkB<3b$+D^ykqT9v%fdT
z?D^Tws{HHXU{RJ~W~r4<hnIR7PEeO(QefK{PDl8|vxo0Qho$E@=T{|*8GEA4MLXRz
z!zPMM>B@x6zBJW*Wk`_!e8sc%qbjro_Gr=8HNmDY>B<|(6be^&cbPw{m>X+I((TU=
zMhB(UKR5CJWyvv*+pa=BTa&%<uNbUy76;G9j?T%$DEdW(ZQD`GU=i6hFxkX7T|E$;
z%mFsv4cjaHQS<X*@9Ii|Kbb~p!i?L)zyH3<UMg$ub77I8LGqEwQ$&`-$T8{>tiC_#
z=F$if$`;=CirKQ##i9*YT5crVgYk0^<YKF!CSHv(4X&5Ky~?LvqHGNJ=Mu#aWRT{^
zomVzCHq#Dus#@&>JiB_IZ`U;VG$0|Joq@y%kHu>$P2PTnanjBLX}B)e7uuady2beN
zcn);`e}9PPY7B?@@8?K)#wl9)id~s@Uy?`CI?=BNaBx9jau&U0Uuj<a`_F&!#p!xH
z%~lxyq_F>;7hjc4!qkH*l;5y<{=JMnebEJ7d(<$vl|sD^Xej988~(NK?;?O^{dSQ#
zF_=N60J@s-e<h1+G*H2ZahOs5D={$P=RhFj>32uLi>Vg^X(Yk%KjMOOE)4pu+EnLb
zRo}J(hnlZK9@O*x_Z{>wMb=`kUBCWq`B{W=qWgm?*zVve9APrElm0Vo%W&?+Mdyd$
z6PdQ=B8so-+HbZz^qBPa7Z(>7uxtJRp+vzs?;pwM%qx-6Rdb}|rQx(k4ISpSyJ8+r
z%zobHOkP4`MfVaL%{pBg0Kx}Bd{)&kc?de;`@Tn~xzN>nFB3eVtp<d)8h&tXL~{ru
zU{wdOT?Z5>N=Zq*eam$E3{>C1sS5(e=uD|;^23c`@PLqhW^#0-@SSfQQQFaJ4eoR-
zj5V=3r+7<#VBtMpI&4BcklIcG2hr19Rd2m|Yh2(?|9zDV`5eCu)$Yubt*p^yF<Wow
zNMGRc$jn@5FX%o#>3DzSWa|~hWqtMUv`X`O6((^TN?(qhh!d~1=T7b7Yt@6-Y){fV
z@dkkxgRQgO&Zif78XS_<uYgg#Pl{{V*oLw{z&7U0P|T*ZL{_x9UtHWV?LYf(pZnc>
zs)^Ooo>OOMKQ*MJq8O@>EBWeXygeHWLdIN^f9KfRzB6Ohg24uCEpq9qtwI@A%R@H^
z2qwalLeZ&Xu}k3!ny-Y0>2RIU%=aYgDE3`@o>IO6+uzYC>5w2XDL)sr@SD`3pKUKS
zt-bJu7ay0wp?mIh+s*^U5tOB?VGm>G$lycSD-LT@eXZ-!3`)~7F+L4xq9%Mj*24Qq
zDis~6M+}+7iGS_SlHA%b_BxkVaNNM(_EW)eJ-7@AU?@h^7rQdFQDHdBmb>fCzT3Ky
z9A*=VeRi|4B@6>u%15qWJ<-n|o#w``t=Ob2glC$hk<T0;40o3UsYRE!6aT3@CdkDH
zD<|fQe~nlD7Izcg(26|W_#n3u@S6Do(F#_Y_HY$Y?yJMB0&qlDD<kBeX`GmB4ZP|F
z$Pk+k6n4Ox^55PC6SbdUCDE%EeShwnIwJXK$B2EulLhCqjgI7=gSEK5g8s|-lo`r$
zwq)`;!R?RKJa*|Pq90{W91`k^49<Kw(R4g58&0nQ*9U)+cePc8TOdQ>^P8qgyM?sD
z%l3?oTiSwswHg_(BZd9*>*)CG8$$Mh`;NMVwaK8^mZf<vdPtDV+Wug42^+ml@T1dF
zG-geN?MAn}38c`ef&-C9`=So0(h`le3AF-=E4ZJ~zYlt8$o(gF8UJ<#ZVriQMfDLC
z#w7|NMe~P_A-M!wsx?mLHXZQ;baT=4UrL|3`f3=Lj2gQB*;pV!+A?W3X-qJ7m*=+1
z6)$OYOP2qh-^<_d)13+{-e0o9;hL69o_I7^cp8Fd5iaNPcz<0EDsx;R);DQlQ}83m
zj1@MK+)-SiMT~{*&zw+mtlIZoOFRkdcGj>x!9drgP`G@;s7?~~SNG5Rx*4*#zYM1d
zEvNGRKsoxRiCt4Dh|AozFp=!DADs%G+<sA0nID??<o-YJ6h5%|nRM9QJ8yUJL~ms&
zWqy0TIUwm-@cd5^l(pqaH4#UP+@!d|;MYRaAO=Cd%2z^uq=KhQ9j?~}pLd{}n0`N9
z$z4H1uUo%9{js?e>V177t_x=yuDt$C(}_<SDj^|~SARH#?sS!m@jp)<zgCQyRz7d=
zs5yR{D@t7n_%pY^s<8>RT^CnZ2GzXT;79g5H#t~gY(2Hk%^wj`_N&qP`P-)a!3IGZ
z)t)@v(z(dea^HWwO<^~J+m^+8ZQcJu*^-@OV%ns`3QK$8{AQSy1<yMGXgxpQ^(W_B
zD_yDRj;0+`6>PStU|q<<ekL4Uazu1M?514#Eht_}|F0W^S6Q9(4i$ztIppzRJ&xoH
zdt9n?5K=DMCpKXBGLSgQR^hTkfytf6v#_HqX*k}osF--+U{PYOU~Z{o<g7V;s7fOM
zJ%bI{l=Ao?c;-X*fV#1q?SS&PsuqtIBo(U+`#~E)qu;fZu)m1N754Et-A_T)7O0gd
zNwx$f%`TN@{)%S%-4H0*R||zL!bby!d!T`8W!?xktuQNAg?PT}w)*(p?5CA(@km4j
zX;=&0!qUpKL#^p($gGNzWGn`%^ST=Qm6sbLo2AP3_qT5TX;R)}f~+Ns*U=7E!!?0j
zy$pxTL6&ykPb!32rr)(iSPd7x6)bg2cse>H=#ZMfZJ+w$%OUmk2kPdPeb>=z$R<k>
z#2)Tg2*=G%u+6sm-U)p%Y$Py(MGimj83C^)Y2o@J(?P);@1gh|{=S6K?vTt#X5BM8
zYcZzF)yw(W&{IM$&dgp(`2qa^cgMcM<+Nc1-W8`^n97s>{;CQXhmvU4TBNMgQyhv6
z#qpfr&->iVP3kRoESw~x$;$V)C|EVcQr}0NaPx7*4$}=&ET^8X?-f8}e}hL!1Jafe
z+ML=}qMK~THSdB+OY0+ZonBQlnr6e~e#)MxTA3&hv03@7@CcDcvU=8<fHfYvnq*Q(
zVy<9{)I}X#oTgR3ZCLe^C@QV=i1%b!vKpJ>YaOHY>xSTHgjz`c>~1RJ2!uSQ&dzVE
zF(GSs-7!R~XAor+kRQ==^QFg^5IzS8S*mF|IcC9TGv~V9_eVFp<ic6B*vGF~paVnU
zafj`TmIOM<>sfe@GJH?pUbr||m<6XOWH$al!esQFdLPM7?S8(qhc$SkL#~t41NE(|
zkIg1rO!qL7Flkg<R8m>s%0H2CgFR0b^rhl3PY+G{!pZ5kr^}|r8m_(;cC$Yy1V43}
z@oO8yj}4C}GP8A;jAwtD*cGa;z@5W_;t*6|Fa)bu@@@`k%f)TeeqV*>2isv$4N{$r
z%NZCi7s(c-0`p*ns`Wx=qel~nSpU56?PswsA)czC4kswX6o7*l_$|)8<|d)w*|i!y
z{yE_Ey9nP)GHSwX`{sg~)ir3sE0u@qA?TZw(hm(ycc5Dq8B&!*Z&PDDNx&I`-)<Fh
ztp^F`WV8fU7C~a~RmxyR4U}&!V^q_`enbb1jU<QVe2^u<TPOjXoSe}31hm{7d?YK}
zaO4iskkx29B+JD#K;#$pmoOH`Z+lg@pr<<vEqc5rV859p?=dG`F&QSDxKA6%TBO-q
zZW0_DCMVlZ&__OjmCJDzYG%Fjr|AH_YoIG(ScP1&z)xyx?;adj&vv{9;_P&Hf?dz$
z{9+a0=DLlaZbSSL(aK_BVcnDYCKgEb0<H|sjbS)iF)xO*Cfw26F}nABo{S)4)sq6P
zlVl{PPU~av*95p>rZqf?XBj5cb~ZQj_`Lq=VW~)PX$OJaDA#Ge-f;NXV)}P`w8k~C
zdB6WML9xt+`LzCr^m=>TR_Yv2NOQjPiBrQNV7BNXZ86|~VIJCkgNxT@HTD_-0gK*l
z@_@L8(dbsIxQ4Shw7uL|KA3C>hb@%C(!-jX8c;21)g$*4fU{95w_`(yVIU4oX4=H|
zL~IK7f1peY(+)_pL_UbCFK*s4<rP4GfCuK}EdjSHou9sg)diRr!|dZUn4fqP(M^w!
zGlMlbPs!_=i{hZ<L*UkfEI^NfQFVZAGx_M-kbbnIepOsMz$_k^x0&sw#3DJ-vzgQC
z1}_4s?knfbLb1{!0Jw6gdnGRtlY$alIW{x8`u7ZTSC-aHJ0jqbLGTr0;tUjnp#*}I
z)b3>DEH3XesOP8U7KoBCg)u{HN-n}(U%NF{g?SF)-~kJ?KB>Q30@q4qI8{agf@6aJ
zxe)?w)(*z}B!q+sbD-gHB)MS2blX;Hx6gFqms34#ztefvA9CRPsQ^bnxYSPbZ<LC{
z$p*c7Sy_ygr9scRZGd98YgO+_W->01;zbSWFi4rB!tw!>fE9|xV{^2>1p>%xi%s>P
z?)^YsYC*bz!4_1M%Om9~I%os}+Uag2E;gjD4y@B^%Iwp3@C{w)1t$iy5A&Ng@IC>Q
z*7NT;g3b+S9bl!^9)u(%>3*SCYH0y~h;s0$_$NA7s(^u<Y58gaL*Q_pR33LS-)E{9
z=6mZpr_l{@{`Bm246k5U!VAa(nRCE3C?U!~DJtXvbCf^O-5wfXp?ZZIHX6<}dMno+
zZKV}f3GQ+K{sB2^CM<ku+B@i@D%d1`7u`F(Q;ArY29a^#L=B3^0T2hk)iL1C$D{d6
zkZb`JymMH7+BTf_cJ*waa06fT-{!?s_a!CU(V!n%@(5vE^Y?Yw;Es6h9*+Ur!^)Hn
zD+3%D`1WVx4>r@T{e!bz`;VC0YOD;f3M|%V;hCQzJ)fj_W>;1;8r_<wnn1b~>-_QW
zdz*~NyM{nKKugU|oGS7EQ%-<A<4-4iy8Zl>DE@h|3bO^ouEd9}NzqQUxRBp0tN0*Z
z|GJ&BxUf$QThg4i=&(IwT|yy&yi&LSf4qWw;h%6geWa&N-DzQe?w^YLzvD&tw!cxz
z)BJggQm)>T`WJNh-}d4P1GonofbT-p#|vOk@&R)a9|9np&2ZZu>BgXx^BZ#M>8g5<
zc|s1<6FHxgL%fhL*irgoPGZaGwym{l%m=)QS0$g}lP}cudzuL8S@?8;*{o`G=juvw
zlW6S1_FaBi?wxqaM3~i4mNW9|^gYFwH$?yZvs28;%CXd6IHc|*<W((S)(of7V1e&x
zzF8eR+IR)oU6oxqPtS}x>Mirxh874!{Z~W;1m@Hg(z()feWj&j2(eDRRY*6>3kOk<
z6EB|v)zHgF{b}n(Be<*|KwE#r8&_x3!HXDz3yWQsJRr)NSIGX*{@HcpCO0=6rhssa
z03+C~*|Ltui1ZN?SjF$EZ{V^Z&euc9VuyPSn$3zmk)n%ayUn*g!x+-xO*OksFYpim
z4p{v{Q}JE({YuQy6UjD8o#n#TOXyq+h#T552?`gqWD^PoHheO&i0?f8B{B;BeP52W
znAo-?qsyo#6kfC(y$jG7nee}O^=*E5aAvQe<aI{p*^b+%dGiPI<RjoE!FkN!y5q7`
z^^_>e%$IA0BYAdm^Z}IbPL9dInLGd}qjokMAQDHW2Ft%LfRg65&3Cc<fy~zp&_ZUA
z?PAo+xZwUlc&gl&s{$tPj_^k4?TpQi8C?)brtPo&3*Du%!F6+U`;vTUG=I_2#Q!p1
z(c;>^NVEfWpYd+s68iZ8uJ{x{CoF92a3+0&y_xwbI<-Wdq&}zPJze$(f1$sgf1^L>
z3h1xT{y#^F|A5f{1~{MY%l|uo)&EDl>wkXx82`YBzp+TXdw+odNKEvLm)em4dqs2D
z8a&ys!cnn~*jg(C?*hm$ry$!a%C#}jbw|_7KhPP&=qu$?D=$pCYLf;%JZ`z(U$x5*
zByK&joy};v^9Z{yZ=sj3-HO6;cK<Hq$5Zs032;c9h0moPY99+v#P{wPen2xJXt)p!
z)$$kDC%t-{gqft)9$I<+-`}I%rqaST*Tl!ou`gTEB?o)LpZw-og+vPbi7$=*Vw>`>
zq}qM`yyP-ti$)O^z!QG}E)wqKf~AMAl1(cD)$`TYGCP^IDkdgrB<*Jn{h)e-Z$67~
zU_!70re-u;vIT9R_npzw1~est2Dy2k<JPmm*P#nD^k&C^CP+>xia%t)d)N{FA_MZy
z{R;=g`x{yOKj#|$Ly`H<-f@ao{HZ+ui(iE0|1W6hPgMM=6#l1B`R{x24{7svH2wLa
z|B@Gfef;Z@PW-`df#LZV+4G$1zi?@f@mISGoi+vkXOD*D(XUG96R=TfzwSc~lyL&S
z=1-7pqOIQsZhxUyD+!7+rib&vCcxTB53>eJMEFB56i@ILJ|H&v?Y=`74dtYkN<$!=
z9ivuk?rP<M12RyP!2*UQJW=zRHb+R&1Gm}H)<iJpl81~15iKZFj$l>>=6Ez<48Uc+
z%6il>#Ml>;g7foTada1lTMq{lASrBr2LNdt`r-9_Ps+~XFC^wEa|ejQB`CG%q+;H0
z>4FeHFB%O%!JE+hD7dkJjh)faFStrQhX9bjL+Pfw8xE!?Jl7;`4%Zr@Z6-W*Z&^`^
zenC%F#F1+XfWrohZkrfaEy32$=3qKuu39&!TWxJ^>r&8|`-H&S7;K(RLHD(el%Q$@
zIA17#_5GQ1ZofjoNMzrp8cwSs21Xm9*#NU0`Ywt&+CKv&dG}nbJg8w;1Hn7_65dVl
z(s~Tq3XXgckUBZuLxBH_77)Q;IIn+94^|+cHoI~69L5vMagfh!ZfvOJzVZDARC_p-
zp`bJfesc!neQgcYv|vI4r-jWeE%AWm5$Sm}9S&8=AbcDE=;CMou;*(b;4KE9Z3_s0
zwBz<yw~}pl==C{)a{w9|4n^&jvi=QLZw6_QXiW$5=Ae{$^DPYKrYsHW0i5u>TXF|_
zJGkJlY=*sXMMlwCO;9JGQCsws1f5_Lw05Amo+x_w-D(_;$^wuTZQHZG2Tk+g@6VEX
z%@!+qTxfYK`0eBq4_59OPAdYgri8l$<4Di}!=@iMH^BvJw{+Fwb@2Tm!NdCms&nY1
zyIMfaTpSY8C(2kZ3!wvH*Co|y0c!L>{A6l4>@|VBwk1m;D~j|xupkC45Kx^`op=NU
zFN@xegTE29Zoe8DbfKPjQ;{<~I9LisI^YCE2dV>Wnzu>t0MG+VfHBBA^|k<!uM4)b
zV4Qi7WwivyvxV)=2k!0du|ewzR^l+@Rr+p@0Oy{pg29ueBm@Ehg-`BuCHN)5*w1dR
zv(^zt6U;<Jg9-saM0f}K+&tD-A7f!{z>O>W`am|h1E?Sex(Q*Gygmm2G4do?^WauP
zd*}BdcYo$G<>x_CGf@v#S=KA0tg|pW!Q4$+t=NhJq#rb(HCHmxwnbp{-X(t()&<*3
zFm)_kLV(mFp}y~p9bEjpR;a$9H8B4zs|};0&bzDQtT7!=Rscf|Tix7zXY(D%p5>zv
zVG75dp+sF6jq&H%R@~$(#aQ!z>$u;}W(MO_uviou#6W(%)yB2b!7Cl>AQ3=W_dnQs
z>!7aNuU%9{Ktw?WL|Q;fQo0e4lt$?i>6Da41w}#x>5@)Cx=Tdq?vU=5ZaC|!KKShO
z{@yuzpE>8u{^QIz41?deKX<IP?zOIUU9j3vgGk5lz~OI7hL$*$O6hGD9BEeNM`<lR
zx&{Z|ys;R<nhA;EwuYcXY&`;rT$pR$A|Gro)D@S5g;WU4*t!9z1YLTS5g!3_0<|n7
zZXRf+xcgbF@&HoyeJ!wouxS*&9VnWcUM(j5n4tBp?hvTa+!xbRPY%aUP1*pK2OB>V
z8Bk^VN0$Q0%4rJNLeVZ=>UcrnesVNxunY%q9;g!w5vik0UexCYJ1eF{{(zx-9r-Bc
zvRxhPlWmgVIoJrvSpZvG3=g4}2d%&BDeZ7s&4V_x8r}jZ(SV)kwC!||^8m5zrrvbL
z`u_f6uv;l%2>6q;yVT(cG&|n!{%*>epnUoa4i{uv<lWV6s(}9z=sHc$zlNu6HBL)I
zuuO&Y!o8f`9V7^=A?A!U8Y)Q78&Q*zE`t<GB<<ZE#|}A>cYQ!kJYx^m1wmUk`DplL
zqnXhl0>+6$rxH-)Q;G_`l)&$$0LKL`%|CofB55Hf(Pv&<y*KGg$YtFD!|f<zngu<2
zgh`YTx4WCkAN6y07tkq$(qT+0#TGH|-@m8CAmPuHNwn*nf-}{VD26#ib+;v4B;Nd_
zhd6${kg^PRD*$>}pgrtVDw35(`JHGxoP}9ny8@d53WQL9jfT!NrEjvu7m<-8kSS%N
zZym3>O$r_ozB_>{k7kw{+tMONyw$s261XBM!MQOmGvb{$ngDW0BcaD?K+R?92JNA}
z!iftvP+*o=P$hbhn;=Db;A6lI)k}?~TYTUhdS47lRc|!W&K8VAX1n)eia<Jw%cM8e
z9*X1@JU%B|=%#G71q$3uf+5t47vaFX-SPHDh5pb%DBXjEIrrxr^PaRoyAnhT4EUFX
z1;7n&*R+J(fHJiWv|P60wn7g?va!wGi<2}&03#YI{Pbo&yK)DHxBv{)Pi6t5MgxRj
zaF>JjK`kIlz4UWFaF*Gi1yMg<p#&`FK(%}If>1EEywKG&1E9}?bd{oUno-BRY|IJ(
zHeoIMnwy&s)cSxx3w!_%P|!oqN?IV%m^E@PG%>9-;D=1NMk!-KOKT|}kEs-Wx82+O
zGOXxpKT27Ey?3p0R||X%Acjlrn@0oXgefFGOauI?<C|smu|cUUUoNC#^xC2a2@|g^
zE``!48o&Ww%uzK*Uk`>&=TE>PgB*9ISo@kC?IctBWu2qFb(g2Kl+;>qEI!>5fXpqM
zV)=(i<yIrjcvSi8(1&870@>nTS{s)&5Q^Yit2$f-Tsq-qP274jHo@%Dr$zArLzyV{
zRyD2T0=MG>W4)CUZ9y_aXpY#q$$}b$0}U+3xJm+j;hk|>tKct#qhX1khW{JP^*{c8
zjRqXkAZ%NLAc|862WS*iINi)|o;x=#c!QW&tH7ueIC%&8Z6M(UmsTjP<hdFEW$e?9
ze10e}zn>uDf*FMbO~O9fS0Fglh291;@^j}(vU}3x$-%+waKUDBoemU(=;3KA+SRL2
zD-eTKb*xv8g2t%Q2=iq=6*yI}W%6Faw^lLhHC)?oIN1LP#bpO4Cy+j7fHs`IwH}ag
zQ*D-t<sWehB?b@bcRc$@KAr+*$r(J15pSB9JG6xIUOIOU^CPT|5`UimD-%}W+hpG4
z@hb-|dKS<r%@$~tm2NbJU}<w>ql*xLbLmZPEW^NgxxGBteP9DA2FOfxeE@F+TVNCN
zmd8g&%fy5{_GVz#S&w>bfhibz+ny3{@9u{A7g8>eGgZ3*#O+x-F#L&vg@|M~H#uRg
z9?U{pN{tqQzzCAQ1}QvHoq=X)DAgkP`4HVvUqK2Is<*Ekd)(X*s>KpPl)X7J83?*b
zaGi52EA%sid}(t<&<s1(3NS#h%MXCR*Xi+l6URzx7-zp&oWP(0d~<9X@coLK8>sC#
zf<t2Kgw7bi5vX6swNH@@$$)%eF_~NUgQxcoG3~)SDyab;r=T1a!oKXVLjSkWuB*vx
z=YS}T_dPWKQ6g`1=fNgby~z1<<26jS^KxkOkT;l}9*04(*?^@53_Uo3?X-QaW+<(g
z96)QNmL)l05TGPww>u7)()8Sq$M0A}Z5=vJ6e|=mW?P`c0Q17RFCO2b$)W*n3+YD)
zFr6Ub*oyF-1|U{@m5@^o3mV_LTi^+exhn}E6Szw3vYcBF2GPPj5wMIryhhbpJuSIy
zk0u4ropWFcSM&tDzdN*9ONP=Jx3&k8*8>+Mgj)>dgJS-mpD14$JmLXvz#M|7WBIED
z>$lIH`zF{z^A*h5u-G#wNM}O80=sg!!1(E`*;tikMqV3==$fE}obY$Z%x=y>N7U-n
z6qXGpJH#Llyl<G{!xx{sLclJCA^G7>BLEt+BEYJ26CQH$$U`*%qFw}zG1wq5S0lhL
z2R=1xt!<bu%}XIkTpq$qgfE2-7AOjIB;MN{hv>E7Cqx%;%7H#=@1;3}k-N|aV}A=%
zZS2K0==QM}d3LmCWjc%KzXd-4P4VR?DfeGqfM^yvmj()q6~PV%LCXup!<!2m!1r0e
zAXS-&x^1YV69=jb29U%IMSQ9i>br|TeTPUo0;1&b2za+4xhT4w3Cn8JgaC75LXe5*
z)}y>5c6Ez2kATbiPV0qp=e#`u5DNR~E`)*rmI%2BT;L`JaL58~T|z@sg@X%;BM5w(
z3mN`LyK?UNw7V6OUFLrxaA0P?&2-yR*W@xUFXJDMJ7x|p*8eGvyGk{lpgYA^+t!q&
zj8GHAuZ`D&YDB(Gd4__)K641!u)-vI<~R4TbK0#20x5=;JFiAZ7Ysbk?M+B5c9T%K
z9xQZ2r@KIvDz76>g9E4Q{;}TPYoJHj6c(eh^t0}Zu5#f!cIv|?)Mj*v@9W!`*+2dA
zyzEFvdi?sE*jqOHHHm@i{Trj&emr68990C4RAUZ~i(Qu(PpO)0*LQg3QQOHY#o~3y
zv{yHV!(s&%$m$yMH`Dfp>sS;8QzU0F)+b-aVuj2_uJNLCO+=rbw)7LUUm|<YOW|nR
zjJ*FaUesSq)s=zOGt|YMmfWLbVYVtH`(20soKSxV$Hq)ySFOrulOl(kwm}=~;x6~t
zYnx7&$dDoGiJSPvMN>s4^)q(DH_}zbLfak>Ng~%$WZ_myMLsMT%|j_PY8C2Lb0z<=
z&tBZJ`@!gOM(x3!K||B-0)?$dN!PE?^^LylI=?tH#kP)_(xjlQuoZUAo3)XVIWAk=
z!ZgH<!t=+*$b8=2017%a$*q<-KWUtpFNzidSF5$ET@KbhSeb@+&u+D6$4(=wCDJuN
ziT3G>IXT*0*p+oBN#8Lbyb_gb?{8#KCz@|guK514CJWkiZZssD8YR!BueXCtA$pu*
z<=sza5i)$#;}(5=tE|!KWVGb_{OiWg5lgidFVA=WV8Tmud&kD4^)mkT;^*%tJgkrH
z7Q+&RN_FM4+us@tJA{wi_sK@Cj_u#SDo?MBZ?7CSSe8{&7*K^v7+ut4GCVuPRAi>K
z<UjCH*^iGW!qkpvl1a^>l-)xkX61uhndEeiUA2LG*gMjC)V@v3OFN%$C?8*2AF?A&
z^NhZOiSuHwX}WB5QT==f+Pf9EA8d+U#i-%*wgbd0t|?kQTS`Jo0frK3rHL05B3E7#
zX2f}L!i}*z)qTA=7LSwtP4{V6M>cbG%sJ<T)98~LEmOwU+mA%WDDlHDD7i{Fb&y}k
z)yKM3R)Ui!o^pqPQDjkF)XNjElR3PgXvc}vU5n)F$9|%YnE1@e8>buAv&e(DEF_&g
zvNglC2^P<z&upb}mWB$1`)NKS%_45~kh#DtqOLiYCCnrEbzrU$W1Z6T3r=68ciBR7
z^q|gT^{X$uafba2y#syR>fR35zqJW`^5DoFrSWpdCG0#gXN(S}w0S(E>EOE1_Ms2z
z6u3QkOpQ$YLVbegDmj+#`bychk!CO$*;&&lZ=<0;vwT~Disw_pL26*Aqx*u##VRKe
zDfo7(;BzlLX0bi%Td_>bL$>zsZ+OKb)fT;Mx`kBNj&y$nDI)i79N%@dnXlM6R`pQ{
zO0yEz=7KvP+%oi9<j>g^a$TUPk-bm#kx@jG&fhDxd~@0WPkx}6n`>#Vo^-%Qo~`}q
z9j(fwU`^4~*s6*TakD7}<syV`enyA&O#8e9IiU|?y7h&^X(jDn+J2<F>w1~!il>Np
zi0EWuo+B2A+LiYV!I?6I<TK3OhOOJ#bc^4n<|RD>uhRs3R$)ucOX0K)-?EYveI+uP
z(?WhX*g?M1eA_Rch+0q%WkFSkOon`lQ=lT`GSk-2TQv^|bP|ZA%#=~T*h({VFljck
zz4RtV2~tJHddxJz$K<$j*SDC^*YKQfM4RUJ$x!ckO%0S5ij^BuTUI*E@{R)g6varC
z47}1VO3Q3MgM_KYDoRpa$`WFlt*;H!9h)>vCKzSSFT3u(uz9$F_RMTsh;ioez>d(h
zr74jFdd0+hpJ)T_UsqD9+`gZIg`r;}Gr?*lL3?WhZ`I;tNJ%Wo!}S&qjn}8v_~(;i
z36urX-kYK3M@ElyVrTv2zHmy#R!B?lzg`$P3aHd6oGhrYpa1&!p@9#os~jrDKHe${
zYY&Z$*^J;lXV%i`!c7`;@kZw?=bYKyuDmIYY%&L;yN@CYo5PU_3);Us#5p9KXQw1H
z`|>7VQM){lDceqs#H^JlLRHr0s^l$iET;ZA?4o$jK$`-?-l%8-!33@-Vs7f>xW_{%
z5*>O3d2Q47Uxbja2j`Q#jYmdX)*7DHJDRFOr7m)Cq$To<x4ROMgPhGJT3l8C(7SGC
zQoJQy!&=#!UOj;2LNkiCm*)p6`Wk(un#UDV$l=Xh8s!cLLn6YR3C&L;DC0Y&$pT{c
zlS#%1qZ52IYA4mSFC+1gG>v#xAFX7!ylCcVFP74&&VAiR!{dnWu%zl3XYQeH9<^I+
z_O&C=n8|g-a`^U#4eVG$(Jmw*zovv?lVO#hu2oCRVb7JKckb@PC`H5U&kD?0<VP)s
z9V&Z^8k2t}n0pLm7Mst@4A*%Ei?^0(q7Jjax^7+mohj#3&#+>oP|y^|WH<8*N${p0
z%hDsiX9cFUP8ehHrrq4^Tdq9pvpV&&;)b5@<QNHL2HhFoj2bs;MDhKI*^;v{BuMqU
z*MWI{`+h&l`H=hx@qQG+yZ8<kV+P^XGk1Dvzuml4#$n7+GkYa@+^|KC?v8R-*I;ly
z<#8%&pZ$CDjK;8dk)uHLkyT~pi|G%{k}6*{mh~wywpBaHJj)7Xq7CA~UGVkwY*8Bf
zkREbvI<!U03B7O1QF6t4hZINNy0OPu;m40QK_cv_B(B?IGrlR98_zy^k!Hk~(egBf
zcDt~YTc}9WctjD1sJ9($2TE`9Zp`0Qi2{;L!QogB37ZEa@kj!MM&Y9M#<ml20#`bc
zv<L|~sjaxa{cOAH@Q~l!0E6Hs;Y&TFV|9*k^V_ejhD56J>7!mOmsK#IpIL8pXlM8y
zWwn#lx~CJeZgua#gMDf+%t^(_z?6fSb8JpZr)Mf9B4lK53YF+$wK<0ddi5Q5+v`tv
zrf~-e-aDDslxoN(*EY=C>gf+@YJ2$=T6-e*Q$M@1r*v|K=?B-jT;)fsA_N&RB--c1
z=pR?qXFsX6(i8I`K_<xJQai9Jq8~7NByK>OwSZxzf&OSw@SSD-eHJG}oYxVpbFbnI
z?M9N#MhrirlTmLhmvIsa9wyyH`<5o9gZ%I_lEd<C@2v~b3}Ur?%v3+G7&K&l<^R-4
zlfD#2A&RA3uXkG(tDiepU9LMf;*cPm8^>1j3Zs5)(R8$s=Y34^s(QO|oPM+KW>*Py
zcCGs|*mM?0`NfQ*dGF;Ma~>xohhX9bKXRKU64Gg1voL2HMY_ATpxUBNToPVAo5M7l
zQ#V88X-8V|aAvhFR1)i}$K{`NJt#BA_l}?W8wP&1d16KxPN!!zsJdFP8$~CRnzSH+
z{<ZCW<rvec#i#DD+3B*|*4B4c*~6l4TQhv)#Z+0|_8N@pc3C1cX!2Fuw0(GSTAU?D
z@t&2~MPq;7rD&-<GgaUHBZnAXx4XABxE~iF-?_9*OnI#RJbW&Jqi&CG)|(~p`_ZTE
z%&Sf}bZCeC*)Aw|5jyJn-pFDbY~tu&DQ)GhraxfB;BHGW9(>VgSTUyElB25-x^*wr
z`yJ+GKDTEnh7FX2SMtsEoY>HbM2jo>TJ}3t@~UcItMIGhr@T1*h*gQLh~<WDJoSL_
zP}n?_Oq!-o*v2L;U}V>7UF8~+u;@Vp-i+<yUfli8yCk@YVXs#?K0ai>q|sDQ9Gk!T
zXgDdp$|YKcc*!`-g?=wso|ejL%=~lRUDk=jmXL=ExYbdmZHDEBpEo<_%fGg)&Yato
z+Ux2w$dB1*6W%Ahkg58%)Q(z!WP&7YoLqJWwM}T%?r!$;S4zu0#AWt*yobC?{q)=p
zPh%}Qq?8*@NQg(#%KuZ`%<^Bkq3_+}xc5I5H-A@=4jN{^Qh5m6&~(B9=L8IRKdM?z
zwM_T*6-&vO2R+>F&^XvVjKgeBNYZo;iy$^@x`AAQ`obY8X5S)uzkuiCa}t@C(s`=X
zjy5L-moN#d8Yc{}gWj55(Q9j;Z64(!S?8LpZF~Jf_N!eTwU^g@JeEt(-(p~TOUwyi
zw0FOpbPGTA4BU#S3fi(*Hn>x{O4m&K`dDv-szu7X_X5tg)t=Kyg$>fXto13!-FAyR
z)97#Yu!4&HzaOnv9B%YR2&^yXTd@b5ovaw}FPY_ud^<9C8#_Kw2ykia!^KQwM!#_z
zU7T4Ig>C(o>@oEvY>6)Mckdshk`s63M<&XcQG7W+*s@O1^J3D|YA{jjMwsr?c>8Z$
zqF7IJ_-6u3X5~M8xLBV05hwA&C9dQvlim(vxc6lpAdH^8-}?6F$~SGHm>8Teou0u-
z#|P3ClWv#rsR)EjunNX?!(KM=HOmpH(W|1zj8yrCC8+6O4u+~Q4CoU$Un9t~e9d>e
z34Pax2xsNkje16`m5y`Ig1*}-#zXd|eD$-p!-Ot5%Zi^8dLcd*c<-*Aqj^66^JJ~Z
z`#XvubR_||nH2fE*m1|Fd+12oJ|qtD>FlC2?!+2je<fzc`ncs4>J2Fd4TouIVg(=M
zXeHi?cSiLghYxpoHSJ_V(AV^%JkH%-PbQEboDB#HnIFzt^3-^=TF#BnE%5w8AI+pF
zk;L85ChXIC^X0qj3`Jkt*c-ndmr&YXO_N#|8=w1%b<$t3dz1cVjhwDJ{mtu3HhYe0
znQ5g-yGFU5x=4pbYIMt_v@>(Yxe1gQ=7lO3>h5qD<?b0hWH2u*v38HWiLUQUxlxAg
zYj1qc`0jU`HA_X^PNoN?g5DyPx~ct-ylHpW6m<)zqUKRZ4OY7O3vhBucZ(Nr4c{Bk
zaTqa~miw1xxjkN8>t>W`8%?Mn4BEyX%II&|=P5u}SFh}eKVh3<CiZ{QC%VvA$UlC6
z#J~Jv@Pm5-C(L6ns`H#&x7sc$VbI<;5VO2MpfVGe%(mh!>iwZgo#&@beN|f#4o1J*
z<5AYI)t>rVW?t<oHy4iHm@$57;|C3@X((dVEo3~cZPZyR?OI=7A}7(<^Yma8@YC!j
zdhGeCd$yH3j%IZ$i&X#6v9d?KDG=?W{?PxuU#+mp9YeVxLC(*W=E;MGuc+zB7Css*
zEh${l3TvO~NI0$3_>fsOHIhrVM?qrR{NPhW(lz<vNA0t@d`pc_Ki~6ws#NEs&m`_<
zLTd@2k-;Zk?eq8{yY%*&Hci`Rr<ePl9QXJk$%v!z<ehq1%^c&@vGr|-lr}LNMs@Dj
z<2<V{DQ9jPJEpr6pZsR%Ko>dng|hM9B3;|6hPvN%wbC!23}<MnpRTF4=4I}(Zwz&r
z<`Yl^GRYX$wi(yzv}P4w>idy^^6)forRl45`1VCMR`aWBo!V}f$VpQBm`rj*6;HLY
z4b|guu#<z4eclaRuxaadSD>H7FC&uBFOBz;`*sucvEoxkq@C`cr>*7+LR>16$wf-_
zrU4($1=1CcX09nH$*iBANAh2E`Hu$fza|Fmu`zM|PZ_vE)kgm*_G^`gv|C8C0U|eX
zzg*M#H0&`Q8fPge<Ixh0U(<3*qVQcBn?l{5!&>g)tU~iAU3xRFqg&HIx1TxkxRjc(
za8Af`-5RaJ^r5_S{P+V3J%$X;lS2H_QLkzlg;&XQLhl0=Oq~c_ot>r}dYZZ~i;zpc
z6Enol7MOG7xa-`sMCC;Laf;7h%v@t)k4K@dbD>*`h%n2xjR_fNetl}pvYJFsDX~I(
z45M<7Tjfa-aaARW8?mP%X<I)gK0}l{vG`j~qm3v^(4+LPZLanbSj^l1l*4^ftNxAZ
zgPX^EdJ<euUNJP71l*9&7M{oN!o!oVef#Pm`K%=76EEiK=8{je!jayz5pUjct_-S>
zU&7M}ndT)^WT_gczsCATNAPiGOLaJU*A*c=kyz2LHckb$rVkE<G~SpOI<`LK6pM`M
zL<QA{<IXxMQ=Ov;?&729Q8fr%wUbg?U$gl{^>En8gFH>KZa^=3N#>5@vkac%%&>qm
z)5`T*TqQ~u@ZT9CGam7@QoFv??U&@zs|w%p4wg<3qzaH%Q&l5~&rILb?R-LN++ds6
zBPFZl>zOYtioE`1QywGHLvhLFal8$SlljO6tMlU*4#*_!tkmPg+rp}D#{ICnj4EV2
z_KC5br`+$Hb-V>GQN|<+4{E;EE2eTq)(DMTN6b-d3)s1;^O+5|C6%3CCuX2l-fn*D
z`EIIMXt^kenpOvI!?#xKUFs2gJ{Ca$=0xBfwgnZOeRF#jZR#Iaan4aQ4QNR^8c>ZC
zr(=BODjssJZ0{bQWtdEFK0i2N`Ywk!cIF|Ct|=~s4DG~u%$+xM2L(lfU7w{rtw^5G
z%LiPIolZ}0xxHk${wVv9w~^}HSCsZdZ;{&N>fK}}K_kg=b%l!MhffNbMz_kJ-b)%(
z)7bkt*pjL*ImZ+GhU@$2)RQ}A&Xa8n<gc{$<tG_rZAuvBRgiCLjADE<RwvNP!^iY|
zQAdY<sq9fMUZnW-XGPT4z8CGUM-0m(KWQj^EIVH_A0P1S%BK;INooh$f^=kJ4V%I1
z6BX@VKDQEv6bFJZsGCdOx}!fcp_*s77jSLcwMSEa2(Y<&Lq4?hjT+^yeqOF-%*=2^
zn*B$8#_;a?ggg^0-}DYi!RxV27H+O9Zt}gSHt+bjpIlRTDN9kQB;<Un;nbt3HrxHp
znnojIdXcoeEwM5E<b@|9ZNmC@J0zZu8@|o=3Bb9eY9u_o7^^Mt#$vj?)LSmnZ}zb)
z-gH7|L`?XklM4Mu+Wgc(ERCCXvbk|%`NfhK7nGQE)M<liH59%RSKgsITs%KG<#2VE
zBX_CU@=YK&e+YR>T_j)9vq(MVec4CG_qx));f38z(GuU$%dug~xS*{_K|nrEF=zYZ
zkv6Sfzl!`v1*3-%c-&%dI!Y=v3JlwC`9z7X0Ms@i<ZD;Ps626iw2V3`WA~UT_4(*8
zrMf2%gjIzkW%wxdJFs0+)Y-NaLrRT5&2=dxV+WgV+k6nNbfLqx>iUUoTPCO1XoprQ
z>qREBLfPY2lDtRt^#;xZY%?c@;;#zcWFw2R-zcic<%?1sXc=<Ld3v+8>XY@$C=Ivu
z&~R8Ob}vtRA@$ECgl>J67)Gl#=wV0KduT5m$#<*ZT>5%V2iC`oU|P<Kn|TWtPGcRE
zuKchXzV{PVPceH~jr-aMkAbYI&6kfdUSm|H{ve7lbV*1W_9y#Przx>4^%F%)>*AN*
zwzf&GU?!3(?oU-#97u5_ylj(gi&<CVsKreBwP;M}o|BkG^rxK8nd)?p;v};am7Hj(
zJTEw~?LOSGR*bvRq1->=JdZtHPd+cgd@r5j{EDwp)ltS~JkdxQ<Fb*(=kM12?8#P~
zt&gpSsxy6-jL@V@jhpi#HVV&cR3xb<JKa?~x1C<f@o~E?>|unxg!i@bw@p^LwV|}1
zk##wro8I#t_PN?Z<JO5Q<tS5l!`9xY=`HV=g4)P?iz4HX8C+?V8oH`W!8Dp3?8+Gg
zr?z{TALhx${*%jK=KPyW{(qYcn3<L1car-5iVS!J?a^cX;ktDXpG%it`A~Cuh9D&w
z%zpV4z-f9n0^cA|*dTp6*xOi7CN848oX+L~z1KGsVUc&(bLXw@V|x)Q-2d{RCa|;0
zdEaH#?jqarYu?&ZHHLnJ{Pv&A&asEZv9&Lc`?W!e2DGRCyC2hNphb^`kI$|1<ptox
zEq{~*62a@gW!|`UZG3h0q!v>ED(*=DvB~?w4<Dn;Z`=`A0|4dT4WL*9f?UlVgr?SJ
z;}>VpNUnt978DYpTdCpqi=kBD1XUa}bMvgsOyF5usW+oh{0^0&=hY;@MNv>tKp+z6
ziux2cYfSaW>CMl>@;@-I?<}<?3{i=MgDw)mEPg(bv#2PFNa}oDNOW{`>WmI_NP%=7
z3>$b<=r?Ns4lia7x>QqZYt;vCWMpI;KGYvq%ICS=GM(K+a!+ZWt~o&6lNZUi;5Dz!
zyFq_VHC+n}1`&==)RKO?`vUYAL9Csgj!p(-%&HKIt6DDW-^I-7@<j&YI6uBo_WI0y
z8f)j9_wHDmD>V|KmUmuezjc2FW0SC?;2Ts76>RgHpqKn(0;MRCm{>iOgS6;jjAei^
z2Eo>X2ZTxUoJRbrfNiL(bcL=2Q0jaP1wSbKg6?qF4Xc2JsW?>%dU|uHYyn+IC4gj}
zveya{CQ*PaaJ`@n258*UUq_|k{;hTUfAiO;^@2YB2odlC<T`&3Y$AlltTDPD(Vx%l
zx_OHSL%|jywVOoGWRDm#2+WGB>nwAMcOv9y1B$*3Acl$%D>i?bJB1h;yn5)u|1KZ=
zTPOI};J*pJ`ZM;o9v}c6Q1k-vz-Yhc55^7ui+H$Bs{jyx{=EXXL4o+sgd^Vj@Ec;b
z5bq6r`_FrSPy7vdXf-m#RNXmSU2%{YHunBs`oLdP`uns0mw)}=aPt43-7W~7aab6{
zmYn>xFo+%bXHmcWSs28Ygx&aCI2ykGUu(e23MymJGr!iI+79S7Xka>9pAHx6on=)j
zs#L%Qf_|V2)OA6E_-rAK_!l}8>w!V}ay|uo_3stOR|9tR3a4$8!szhu0caqBimtPR
zLjzrQz%Cdckam0nC^(P}11jEM2kiHWrwg2_0J8~TFgfj=)sf-RQU2Svzl@BGba&@7
z?LJ!>VvCWPo1LY^0Gx)jMHQ5w!RS;V7=K3A0La8aPro86pCmHemNKEEnu=flo}~JT
zj)aU%)sxTnhI3Psk}jcQk7Ry`jEoEnd{WUaKt(l2!kLB*Qpn+SK;k-FS2qhVy#R!~
zZ<CWF>Cp;bLzZ4MfH6KXBAUEl0>sRh|BKlSZ*lnv;6_K3P_Vwg&E>(9f0@hE{QbD-
zowvG+^l}mPjd@a{dR#m_>Y-0{buZf>Xb;ptJ5f!m)LI{?&K$LacSRh1{dPrYxYvB~
zMf4g#7Z`X1!e)STupD#;$puJB1qq?LulHnC4`vVEqu7jua%H|ho8<5GKYWl2;0Pxt
zr%-D7$^3SHMcV#MRp^w90DvR*EHTK?0xZW7@LZ)>pkxo71AonG?j0T;9%|||zz}Y@
zD=CGlfu6BZ1TlYS%rbBpK;3<a@5}<kO^6L#Mn{Ja{ZYO7?C+M4laQ3`TCimfL73#<
z8<EA9iHB#nuaBLVm++5eyoTDkIp+Zc3kc;;Cntoh{$(XGA}78e`i!77&c(^Ox3_o2
zi3wqRepwZ1e!$^nZNG&vz?lD-l;8h_Fjl{}U!=*68#nj?a|4)%?bQ+WECj3-I@Ny5
znuAw}bip9xgpH_UX*mY`Ubwiph0L6sCGhfpebGCBqxEKhf@*(4LINx5)a0Zsx2B27
zUBnmvax++Mz)tkN9PopU8}R9@s549f+d#km^_ToQ!i9a6?<e7f`Qt}r)F_F|`epuq
z`L>KeQXXz@5KV9108AkpDx&8fxZ_${q!8}pmnBV!iHo~UOZy49ssIO6VA7ib;@U7<
z0c?Q%q1mX1Z~0?=k)m(-<FjP|D7Wk|Hz#L)xOi<U`wqff|F+2~!GL}CeDelQ%?j|n
z^Sd26+~Y&Y!2fetb)P<+dC^|H0a&hk_wM~`3P4pEIU6+V+KFRiy#Db39lS(-K0bMQ
zd4HcXLqbBryEF_)n*sa(ZC)STGLTh#R!l`W;@^{{*qW)DZ~NX~3MGtQ{mt2DJwkw9
z8=S44Jqzb~Zjqz~I2apf@0)axJ98SW1)y_^8-!SzKas&+;AdyRxw+7`18c%;3*OnB
zQN{w65fq_k)%Ot!tpK8Hcz)^cgFXUzbMq%p<bUo`GJx`V_X033_?`{LGblLt&M3~2
zu*t&0LLNUiH~ueCc|@#%M#aL?-PCj!FyUhdumQb2z+C>FCmw*t+S=Ma3QJOk(=u<<
z&NnwPAwT{8_1O}<2~A2$0_17qayH;m0d?g#@fueW;C0E8A$xlcXAdwN05&8pE=~!+
zh9KO~-OU0v5@0po*YE#*{I<W27xBhFyTGFf_8{;DBg|ACCd%2^8x5iT_2$i+Ky#T6
z8yo=8fKRb7H#avl><4nKv#+L%HT>vU*WNB$Hwt)8fXO#Ue*5+f;J)elXRk?dO>5Qz
zb>HK>9{{WeYVc3n0HuNLX^S|&_$3HVPRBwo%VCR3pY7P)Pb0pEq|;-szLWl5xIM->
zKwtJO*iy+v2j~mRrFfbg1NXK5F@(`p9bbsxsH92H;Fs+x;~iJCpJ+O7$NviGYTUW3
zqlbrvL?1k8*Z`VMO%076!k+*+#zqA<^VZ|dnvP>Xu9VJr*~_utR>9@2<MDR?7{uDD
zC7?7L6r_kSv;{^_jEx6jn#kPgem@iZ6mEtgurM?CYydRv`1lwBhQ4)6g;4Cm@4bsL
z87D|VMMcHToY#;|`|{UzfwN_V4QCt7J{96s5D<Yi>!``hq$408*yj4RBN6L@uw`IL
zS5{VBp?wbVo&Wq5ra&cKfdUsd3{Yh#ytmHA6C?|uJ0jxiXMgw|dLfSUZ~FybCCvjE
zsHy{<zklEd8;r+gos^%S|M~vE{tO{=QC=QCi-Dh)ml&Xlz|gGD%;-OTiiM5sdG;(D
zz`%b552GIs+?9YqRHXHSMhH;TJ{v@L8^Gc1xJLk)rsw!ubwhax3k$1s^~RrRigW=u
zKFuhomaBv!At51XObbr{k*48=`)e=41d#&N{t-Js;W~$g)K0=A5~{1KlarH;ei?Vf
zaWWc;fPMs^!8<K`Zhin@>@-;)(C`MlZh8X=`k#-n>Ieu53d#rYBe0)@hlfv=xStL4
zSLFX|KygoI0fCj9n|triS%B!LV;YNuWNl_9*;(aZQ*;%d6*NpEfllDTg9p<dzr!O~
zDb^J|Jv|i_l_Ti;J8i3-_~mh6N%7+lvZr4VW*VZ}$vDR<#vw4$XjqiLoDvvSs}2{!
z3gCj?iSGb>A+)o*0uUSEU6Y#$zfbWuK^t*#@wcC^{r$4WI}{WmH_3h*75E8(G>jNA
zXipfl_D`{wR#sMG{)B(uvEP%e4SE#b`$`OO-2`t53aw{vPvJ9(ktqhuv&#|$b9%=>
z(f#+4Il){{B9?L^OSwPsy}y6mFOP+I<_SH0{SZLgf>qu};JUR4u_^ux1hMTbtM(0m
z<8#aXp)ObfSWYlkbiaHNyczNS0<d#ES?bCL?tqcsgq}A$yTY8DG=RawCGR9ju^HG<
z!E64y6Cu_KUm11{7qrd;3?C85#c<mJ&T|M}61yeC58^NIT-z(d|7;<vjvqG&2x#Ac
zzq@=kK|kPP+YR0Y_QH!7=l<@6W`GR>IS4(v`%)Zqgh&yTkG;>%wp3FadY2&XTNEQ*
zTazcl2EanCy_ajl#qFzjSXMcTyBrp`wBQph!}S#u6g-hO3Cjm>^0Be8k(1Xt#8yi+
zeZ2e%JTd5M1m(Z|$~phv#e=U&_5hT+ETzvskUSUV_$CYy9p?LO)&BUQ@#`}+<#j9C
zPmQHfpk-VUdUVuwbrA@A)1{S_*29Df#-_6$CJhMsV-=NfwNRGWQ+)=|33l9edxJ?b
z3fTHm;j<?emVr$b6&0msZ*4sUs=k3g=wpNa*)WL>a2Nsko5?^W?}?F;oD7mLcxQ9?
z#RlYi2neEp69fYkH-?6WpueP}R$==6ZwHAuoVS6~)D(Cq;9M!^=}Cb8uo@=%pVL1X
z*O+=~4EOulMk)i`=kdMRpTGA7VrzIsb6V-bt^mCh6?j~joU<L-dI#*;4(Mm_g#(cC
z*yEwh@AwX}K@c4H8lX7@$Xk&5fgt|d_wUj%+=bA|0QA6rH+;-5;~NFw-)Z$~AiO|F
z_wDI<bT&6wG71VGWKx5;bgH%iL>jXF`x5i}{V9Bl090;VW*j>JXW;7rOu4^fHN?1A
zY{4MIwFH`B64{R$8pM^A`=PT4cmL^Mlb{A6^~jGOdOL>t`j3If*3{Hg6?#9MTwHk1
zCWz|}3yY$I6CmHA$Ab-$N6-V&_%M!>_|n-bM0@;WlFv-yfBNNjDgtpVAvhUVl(4X{
zkdkTv1O4kya7E@cDsopghdv&};~}CzLRMi^$k~9ZGKcxlFS`#O`J3S8(f&AaF~E(l
zetFppF<1Y%C&SiA5Nn@aT;$ek2mtFQ)BUfN7Q5EN_a7`3A~gUj2LB>Wn3<WOErQi$
z?<e%qX=ed;{h^f9m9x2MoC7Q`Y@_0W0$Y1~dn+q>T96d)hUo2VP9t3^Dk?lHAas2o
zB4Ug|%xx>4+6~^~?+bQ#5+JSt#CzZjg7y?6B{kL811$d8o}eV$e%=c`OQG$8g(m>E
z{`()079(_aK7<=&<L>|*!Hul1PjUFy>=tZ7hs*Hb;OjSUJ|pa41?)^v2GV=CGbsPB
zT?JR9?Ck6VBxVZ9U?7nQcnvRW{B-VYNu@yVnh&IqK&S<DfJ8+_Z+5T0|7Vqr5TP8*
zI%0voOwZ0*4CIhfQya~9#6t$~V1F()KQHg^At42`4XqRM0wlK2p2-Q2WGbe%!P1<)
zK$;(>Axkd)D<J+MeYe?P`MwoGXSn)xkBn3{Xd|CJKS~I|xtlIN10%NIw>DJx(BL&(
zmj7<~i6i*#$q@*M(1}rwG}-<cqZg}5j}d_M_m$j0POPo1ZE$e#>_aJ4XYzORK`sR{
zdI&iFKrW(>at!Q{j*bq*xjsn$3cCJ_`-=wZz0AM<5RobR{qOIJ+#yfIKu338$|_ca
zyV73G-|glCw(xWJ0`UC}LO;*$KlerWfRvb)mKMl9yeHsT9pf55b7fcWF0M5a0v+R+
zBj?#_3j3CqJBy1WZVZ1yy6&3!d1#+8%~mU1gbXfXM8A(yaZ+;f{vPaZJ(zw-WWDv6
zUC;qtBJergV9Q|UWS#%qTH*gF1@M9$wT+DCe_xXkAM@4&B?LOxz)FRO{|f$opS-!l
z1nAfa2hziT4<aXr>30GK(L5FS?|_E<lJzlS|NP3Y<=Z`jGavZ=*8}`IiD%pE?|f25
zzz$_|d+@Jq1AON_)w4R#hk@uVH8pMCKHI)#?@wVy0Xcv3>`%N{1O%qW#;ViGale9~
zU%S@_!j}cxGSJQYdp_0SjwtGPOFrR&H$3UtEO*OPwfnLgb;JSv5;GnlyO~O$z|Awu
z5Kl2SCvv#80OZLCMwUs3znkW7aQBa$foJ63&eX2vUtT+ll8rLbp=)FH-+u@^!Q0FM
zD`L0MC33BxsOWFF6AR%Ey-H3TL>aI{H#RqUfJTSZGCUxF31Sk6%-w$0pkBXz{r8H&
z&pI4D2Z#ybW(kB5V8;=+DylNa`}e;<YD8FA7~#v~g<eBD!ru_*AsIP20jp75_rg$-
zIRO^dX8Wz)P@q$4hFjj*Y~j1Xvdk<j&|HA<#}@j3MXX~X59QW^9Cfx*SMR_@6Z<?Q
z#1J|0N4t*Dd#tWL49#o2CP<i=@1dWl^zeA^-~hb)2(?dv(eNNEA>GoWDd>}ga0DJ_
z`QyFos62M=Q{Fx&W7)IL)kBpfr;c{ZN<T0=J5%eS&)4FX3L@<BATWrSp4QtWbTBb7
z5fy#2&VLFvlZyG_@#q8Sd?O-`p(?8dw@_MA0{yF#_l6-)1B^}CRk3g1zO5khF+60T
zWFg@N-LVN&216qI6lRj;uFY1eyJcqCrc>F!$_)ATnZyus80rMNg+tV07%Cwv%hCRl
z_E(Kz;l~Peyx?s;r#Y&O;d3qm*8r*=jz<Q+0wa;Vsx&jkV&<2*!+cZ3xVZIyVx_t$
zNM`kAJYIrEpo3xhv*Rw!2syHxf`Uvt0Z@24Tg}-Z$gRcp`buG*z4A}W8}{RWI6G&y
z=yz!T|Ns22Gw?SEK&1VD|2sRIe{Zn=-8AnN5Msv<*{}9(g%o{1SepqQm98AZ-c5Aj
znyZ$5w52~r&$ybpb^4TX*2^K;bq;PDUom+hOZK#Q^*jHw_P$f0PAZw0-}$AzYEEjg
zw}l=-7CoYgr@KRQ8%?eW35S)3Ri9Uwux>e*;%`H`NpbIGYI1T&Q7+r-gx$!KN%ucz
z^5i{6la=mek=@h%SkAp2WlARdL#Ni*o*HI*9I5#EMDnvO^Z)<y|M3g}eHt_mBIGIS
ztE11pTN<+lvf6{4o$@cxF))gXi)F>d|2ilVRP_QQ1W8)q^3oE)ty}+u7o0|v7|;Y)
z4DzLLt6OYFsF*=H@ytXROpb@pD*wv-W*jin(Pct!==-5P5L<;lH?!A=R<^cgaH&~J
z5DzgNp^Mt09$6fjJT%;IdU1pP@aB?mLpAf`SWAe1Cs0=>CZ0Ub(y|hu^}Xq;zxTb|
znc(n|^XQ7V55sVCmQVRS^Nuc~Q=yYW{Z~v4|7%<_K8LD}om$HxmR-VHrLg-AP0Cf-
z7uAkL+*3=QXFq<2NJY{geSyC5CeG3mbP;G>Xc4jOYtB`w?E4-B!7U#Ss^v%-1%_6I
zgS_khO7R#d(k#mpWJBXHkdSq+8A2=<SQi9I3LwK^?V^KJK_1kjSqMw~)aoogii9O<
z`a7}^?m=Q0SP#rTikw2R_iJtKyWRa*=vg(EMk6951izIWnkp611@5XEQ8&+T1Yd}H
z75@%8TvUOL0mMQn4xya>c7GZe1t24^y|Ix2eRdFU<mKlNQo(<<+-1i40hxAaEe6Wu
z=F!`Ks_(iDSObC{C~y_Yp+!V?9RjYvzxWE`+=w}TsO%twP<uAu{sQ@R9lCZGU}%F6
zZY|z`3)&*p-n0aQFWSou8<8i0)EB!CDdaSvo%z%F0i^$7!2F$>jmHg@R*%A^AQ6$?
zz3@QT{yXdkc$?<4%4|^H)^<1lczB1OzY1amlfmBJZ75mQZM2LzEoGga>{jn9LEB+0
z^a)POouZ<lp}{7iDXD@ngX{x`4=E5r6he33_`&JXoB(u&x+j!>I1F^2N@L<fl^yln
zdD=R>n-94>Kffz>3<xff6-?v?Y=tak0?5Q^(vJc8Cp>o^)V>PgIs~lKBwW_tsQH<b
z2!rR~a`j|o_x?6fuvO!HkK~{fN}4%2lAo_{5MJ^GrsCz))0X5(xa;;Vehd$%-d4bI
z@$;~zf<Pabx&4+=3q?cJT_2X$!o7Lri^E++N}$oEFC%452MR|P7N1W#KaVJlNb*!X
zh6L{&#d9<T71#7M7L}~q7um>mDNUgmNJvDKbxf4xa;UUIu_AbN&OBgXK*jOH0ndXs
zj?k|SN|z?G{IMP|EgQ^%J_Ioy%&h58B_*4LiGnW>aX!0`vFt+fGl07(t6*1nizpcB
z_E@}-Ur!CR;TM?px51?agq2M!INgr+Kq;=`=T~%a-cW5i0;&jzMyAW*;6XcOlG*Ls
zp|r{}<_KahLqkJ5FVJv-_{x_~MO_{97E>Htv^OM;f0X1-$Z~zJfyVNN>!YPMhN^o&
zwY|v&OoXfsKXtM&gXz_k=s`<(3J59!W4fVA2}!98O5>bBb_+|(Q{ce(#7Is?w%C_N
zo9(1J20BXVC0DfHkXIhdV>Fr1P=dxY5D0!CCM#)Z0-g^<SM^OHVd0s#H>C6(L8(x7
z$yMA+?hQFkU|B;v?f2HXIA34i0;fXW9j#yu$Zcdp*DR!0^gQ5V{OM{6MB|5oN^oMO
zFbuY7<q@SHWohWIo?*LzgR>13%uVqODh-1Ow#-N5b~ZMLz-ZYUBmO*uk1?AQiR;ER
zUe)o=a6A8@*_?CW`HW&z{o;lu2ss;TZZo3xPHz(-+LgD?mwKCR{;+As3-o{{^Mpro
zi_~7<Ks7b6y#|KU_>#UhII;@Z3Wf&9fpX^2_RMSsjuCi9ifmsfkO_ICai(FQaRr45
zPeRD0fFVf|kbm25c6od}&P|lkG(#yXgZm>lPOy#8pn_V$mQ*PRB*^V7=rAr5nk?QK
z!;!c;COC+*7!cYmgqLz|w-}>oPzP?My`fzy(;|9%mI()kho6cV9ajqbfyFONYk93=
z(|>Wa>t>SPduuuTufR~#<1mA51e7FsM~zgmX(3%|e7w9DkyWAAQz_b5y%tws)!PK+
zXu8`s>jcls93*iEae-_u6q#nW#vuy<=?A!1=P88J)6=t|>d!%5Nrb0H4xRbahx`k6
zw;QWMG?UivLBDn_*Oh6PLSR-fA*+B}F4ta&hxJ_^NX)Ta<HAFb!@)re^IpCg!Q8w7
z-NDmJqSJAYv?;6ez$%HTtE=0wr*Wk{mLv1f1PSjUmW>O_RL+aiq^&hTM<XF~OHa#>
zJY^!z>D@!%q&Tc2u%t$XG;^9#lNAHj#T7~tm!B_L`BZP+)BR=`>uL`v0)PZH`|3mO
zbYrOhBr=*a?f9G~<(rIqP=*18rk<+c34|1pC6S-Me`wGV^6n-GRr^q)@GB&%!!ui3
z=h-6lc3<RQmwzWzr=IWtZ6N^DPR_y_x}po2d3iNty!1$!nH4A{^KN_jL{q{=M*>(g
zKL3C;Q*5ZmWbjT~$m`dyqrV<!c_@@a@4abna4tgr1za6-de16?06x+KJZW!V^Fk~k
z=|%}NUhobluxQ!|J%;ojNXaygUgN)#Vdw?~IzwJ@cdU7RJ**!E-+e4uZtw$YUf^ES
z*Tz8tU`vcX)whm~q^Amd60TS<%~QY(!DUTGg%cb^y<oYRglN<l)AaVR_NNj%dCKrv
z%&$$5Z7ryQrUN!*24RmDc>-K8qU?hpF9AFu8Zk7Dq^wXzh3tu_h=>GIAd6Onk`V?a
zC1u78nC`~r`kzF>qiL@DU}!Q%)3|OC64K3`TP#OFnXG>Tr9;q3(S1h9=rbRcHFKRt
zvYAd7j6vxvE3l8xq=)+WC2x;l1bKs}_@M;of`R<6;#?!>-7_;<fUcVvxwFEKN91ch
zdelN;89qZUe3F?s!s3iRW}Z;%rL~KtAyQQOmorE52{$7>Oc*lBwlMLSr@^HH$4!VR
zw|*1$${jiXXvFDc9gOoozp!~gq*ZYfRbS@LRl%s~rr^2cFs-iVa-vz4mwTQ(eSpDk
zIjV7}a1i3O_R!{M<?d+EBB)}E-KA!9XKMjifowEqZk&g;cu&UThea2!2U#(ZkqJ4s
zW`w<|@&H)cwVr}ly5{|#fa1`b02ZY@t%FgxNiwsrB<G5c1=Sf<U`Ph_xM7bbNnml|
z^}+X!fMv`Qpay#*ek$O8QeoArIh0oss=r!fGbd9hs#e^xO2Tc^!a)+;ty5@tQns7x
zm@7;VCHjL9Wi9z?XEGkiP~^VU`8ZbYLb`fGs5fkEY=C=s`uMOhX8H2WILIOus2d}A
zxf2r;ow}t~1s6}`Q1v4-F^C;+b6NHfV&f%00b$KggI*KU!<N+{{drDQ8XuE(s?rE?
zm|K;H#j)hpER8TAL#tM4$N*@C4~>b}xr_`AQ~66mpjcSxWVhYIhPu@IDfRf}JQNJ?
zqC)MAR?J!ub!K#0<j(Vi80qUzpXWdg3S2Vo`IzYF9x3fH!N!hw;qS;?e4#s&tE-=n
zFRIE|LFuozzGbM`QaeQoZ^h<$-F%!a;^Oc99L`3_lZ;p_V~nPcf&GCxAnBwHv>Ya}
zLBP~~06aAUrhRk~oDKdT{GaUEg~dx`&AiTbWmpBmFYtRobqp(NlzUKScMJAFIJ+N8
zb+nh1GF9XcMeZ&f4axc@IX7$ovDC)|j}1?KPV0^)zsJWIud6%|&<}7U$$1J(h>8Ge
zvV`OaBwWZvfYc=Az#*Pn0Gs2r9oINez9hQMN3_>K2DOQ%29jtq?hzEm7#TKq9V3iL
z)Y$?WI;LMU-f#Za(cJt{GJSMa3jljifLzW61+2#zk0dk|AKS=Xbc6?kaDCb+4A(cB
z^j4sU_2*HjQA5bo;262XgB|Gs`KRFP2-*mUd8}KgQIq2(79jmXQydATC<hL}bb`nA
z1R+m*5J7%#v_U>=^a$<p<;xftef%3l5Q6aQrs_k0WSvp{Y(nT5v!c;V$d&t1jNOv^
zEl~^t^Jd_ENW;8_Z>^xH2t}wvVB~sh3#31tZlyYqXc?LbnekLTm;<^6eSB)99*c5d
zb{=%jbJt5q;-}QxGw%N!W0YJ?3FsoAg};7X@5LgWvi&Re?Vy;%N=<tR!g3YbP`4P@
z$nr_KLVFr(flfkipT><{A>$@4E)^CBS7P_*YC}s)@M>15e&5c>^D$Qty$uWuJX{kQ
z7NwN5rlpw&1KKOer5^B9#a;wV^ZoR`X0#VYFYl0(3-3e?kTNtB2pI$p6+<l~lQBgw
z1_Qw-qE;478qA=n#$&5*Dc&B&a(-nc5W$YxZHSjn$ORFXhaRwJE0`v7jsuqq?Pz0*
zvJY*2sXa4eX!=;thptak?tlic_WXodeQdYJ=g!3y+|-oTt25(P@ccn?-<wm?`$FP1
z5XKDF)Vv&XzN=ScF>*Kctu_AbuHH=5O>Z|ean7Rre3ll0GBPZCu0*UKURUroI>zon
zL?6@wRa}roq{ecw7a=9L8HOuFH3V?5U-`jhPFQjpwh?l0aCjU)#*e@N-lr0v)q)aX
z)eaT&e~Z2Kqks%6%P=}Mme3`d)d~<f|BRe)Y;G^yfZ%)rFvDGuOQcDhT{%yLp*You
zIcuz`kA@0h+oMB_40QVADM+zXQc*2wZ>_9LJbc*AfRq{wg@Dr7SMHUY@fZlg7WTI0
zX6K&71{B#PHF>y%NY&QBJqk86%~)&<u-2IZ^U^{7DCE7NO#VqO{*_E>eooHMjSZ(G
z5@Xc@qrgv}KCNm^D-DHw6Zrz{c>{xkjR(LfhmdkV<>uTc1mJ`_P@D9|aB^|AMzDw-
zDnVV(!=o0wIb2Xl#~v$xAcOBciSS@HrBzJvJ3+>Ud>kD;J=~Tus$77T1@h1wa5~`f
za<VgQX)OT8z`19P3NB<R13hrbfy>eN@83bGMabqEgb7fW*jvtT2X>teKqezFF@Wh9
zB9ZQWF6UBESZMeB`52I$K>Y|j%nU!<WuT^J&@L+}kphN$FnXd^$P$2VmK_n)d<jHs
z!mhiZe4YVkvFtVws53l4l;$7{Q@yJBNvKF8N9U!zGKBkZ1l;CHCMtmTWH0&i)fmR~
z^mOZ3D1Ch_5~Z{~C|lqGHfY9cOWKqQl<~~QA$ZBN<B+7Jp(XVU_{uZCxL7t>1LZ;?
z*XM8#BlC?^8G(%h_pG(fM6ZKBuz)k_P{Hu;@r~nx`qZoU8-QzSs)GvM(;7?j!90;F
zisD{vRlwA(2$daR6jk6J)4$d%EiJQ-ww5d3x&fC|g<Gg3r(6+`R;)ZZSc=pi>Rp20
z^`AVMkh={eNA@*5@b=^%?A+XC<HbOhJ*~djm3x4B7pw9WhEM)JKTAHyahH}#k+Pu{
z`dZ$ADLO6p0B4}Q_WJUDz$$m~CxCvy1HrlY*dw~|w{I^Ea`N(4j2G*P(u_6jMr|-K
zFaU6jeqIM4YQTcET@2j4@js5>avABdsGA8}Ks=DsVXE#b4^KHnEo!Q&$f&5MONj{y
z7h{C9J}}bLOIxhY%_;v>XLb#8oB=RQX4dtd>5srgbD101yWAnpc6BMYTQV>;{q#A_
z*8{PsUV1P$AD+}SY)QLEK|$)$V^}8{35ibaF{l6nPZ8Q{xPO#6ZW@@GWlG&}2ht-i
z{tmcZKMbZu5D&xSHP->&TC>P3;sDBCmcZx)7aWlOvLF9)9*~4kcrxh^f}KVDwW})+
zvJ(KC#J&o-0Ww)S;2(6w{dE|TP|@AA3Lua^7zTc~2TC`O^zA3ifCg(a(Y_WU41vX`
z2rL6aLZYy+a3~L$H$g{gC^#SB3<*gGN?^3M-9Ln{?V>^=C?+PRH&k~LyhD$dh*iKf
zp#}kQxc*tZoId2C2LvD8UZ9NYRZL=D2RI8~kB&njKhRhQsHG}^r&d`uwtnw;Rqf3v
zD|$Rs#h(rsi@e+&o;Qf;zb9g#4p%y6^A@k`2pIo{<J3D2xOT!>gjE<IfY-Fp1GkSV
zu&M#PVgLvJ@P9I04c)D$=XF$a!~au%1uOI4{)#iM9;W|Et{(MfYdWeMBTKZ4sNe2?
zXe1Zvyq;C1?EkZkWr(X{!*9)C_3LQMnyio@jvTgTmik6Q?~U)9(-V4>ucx=nKXl8=
zX0OSfI&^cYPKc|9kDi8O3j`b~N-(gpswR7OnmC>EA0LoW+&Zb)Kuzw66bN;Hek2`c
zm_^txr;*SzGnjCBJA0wDl6v#Q;@F=0Ys2_=SCc9Gawc{MSTr-;RGm?q*|I=Fc4NY3
zf8}_NDk3zny1GiBu$#*ELZdY7&4Wz|&N<D@oRd1tA&V0Yr}Nk-PZX$M7jgF-Z*y=R
zN5vlVT3vs>Uahj}oR;ND6?wvQytELO>@mmteomK+yyxAlkN_U$g^=bn)Pevj(#EmX
z4XH>eW|Vup-`?B?B(4ZEj@Qs$-S;mf7H^6>tUsslK9PSnU|aJ@JRSE#C&tO-Q>(PE
zdmhNqzUFjQt@OI7*mXtM3^EuawYo~!3)wfRSiQSLcy9}45TkwJ`&J!h!0=ev^f8No
zX!(M~Tzz~mTlU@0lpUVNxsi`(*rVlxto$9Xl1C+Mf6d+3XEy2dMya8euET%X$9LZ^
zaHSCq-OQQlHEYShXiI<|?T1L*^b(DU89V?!h4qf-H+CtZr68?n&${BO?Gj&Aqk7?-
zwVBvC&{rd4dUa?;+V{!|e^=3|(O91u<*lgn!tDl`IL}$JI+dEm@Re^x12VTazs1SD
zFkHXmC%dxOY~RPiBu_}a67(hF=TV6Kb1QXPBE3j~my&ceYGPzLJrW@o1s75{3om`Y
zpUauXFEXfgFqJBvPsL=@KR=6q3+1_dEsd*?vZpl>=lzoNk>V<%lt|J@acG{*Cv<Uz
zIUk4CNznsU@v7&;6%QV_a{J_4e1EM!aAi=#)9YN_g>k(Q6un|}>-h~k^t=TRzh%dp
zZ(Gml+P{fwo6NI%RgocvnJZ+w?K^$X+)*W8r2A?I>r>WZ@vdtU&f%Zca)UE%3z6t*
z(Bis}P%pn|-3wq7Y$qeqpmmYBF4W<f9r3x3ABD-{d>`$XfqtaQ3sP)f&`xgX45O;#
zDjgzgb9_<8i%;f9*IxA#E_K?=L?Ntcxg3y$tNk?XCNA-wum7|aTU-_XJm1LyhI@MG
zaOOpeq6PoU&q@}zn=}Q!C?#~vw*{$Xrx8`7TU=%*bFbF2RnhS`3oIabky(6O;OmpI
zb>piI#rV@z_XDUVl(Q;VpGrle7`fxS3zoG0BrSYPdSNMQ%f=d)R2i*L{nEXsSBjst
z)vH=+D3fD;-ci1it<E;EVw6K9`2cAx3o~1|CZl-#ko!xRuz8Q6)j-bE2fh9zQ_o9#
z)phO1ylZn(erBBuQX(nLE>>r{ey1|2=JIW;z|O1TH#zZCzwfZ03v%Tu>u!*|Fm2$1
zV%oMbs_@Yv>isR<q!*@b4H9I8xEG|JJmOGO5%NI!G0dRYJkEZ8+4Px*de_J6K|<*v
z>G_qr=M7LF=(H;$35~Ro8eXMPPaI0IeOT*6_F`!ML(Qtix47clQ6V)&=OVls-?`Me
zns=Zz-{|P)@0j-_(>~89-hWA+X}t}-x#K>M*N4J;tTu%=#h%wyeZCmJLy<sThM^f*
z+B}UiZu(=}*-98iC8|X8;R{q_{Pkja!+WYcPY8=_2DtO^$JxZ2y;Qts0{m;fG<qET
z@b=c^SfKtm>!3^GBAL;1hp5c*-1s<&=^M-soLWgtC~e&)$&Y!wd|FvMOo%XZL_1&e
zVVB6q&b$r|{}?=wWib3QnEKimQdP9A+qL8io*_HDamGAeqogLB6z}|*V~E=(6rcIM
z^`OhfB3%;8=T#SLKu&D!ySLilt@pOz(|zJ3$N6_Twl|{!ISp-B2q)36j76KFwx?k&
zq9?a5(!abMXqaK@$0)M%^~0Ol@C!y9R}y%%f_BoX(2{m-R`k$pWsH=&T?4Sm>Z|&q
zu&{0>4{^<#lKvR*aU8mwy{_|PtP{tz!Ff_svnF$aAZJrPBfT(L*7-^{R$R*QZJr1I
z?}(@fQ!8DW3|LS#2^ZQeW0^O(Jmm93d)8ykzZE^dGgp?N=~RC~1D|={sYf_rg0rPh
zia6;FUGyt2uC;I$LLxSs7uYI^vD!;`#?$h33B2}n)j_4$7tBrXj|8?L3-834q+lor
zPY?7j<%gTR2->l~%$+<y+TvRu^W`Gj`gRYJYMlF4i-Rvk+uAo>iY?B86>8-6XRcZI
z+D<r!Z;XDV&Wf|;L48MKB5r89l~-36a>H}_fd~Hb2}X(iC5DxUrn!d%d6COx%Elw4
zbU79Gd?*Gyu3UTR>4}B5Bv@>t(#EWD89xF4VjnlYr_t&)(^kz`JE9B261+K>bnXNh
zs=N(Ha>lTHvYV>sqPTuEFAsd*zcH?*HsX%t#79?%+);*sarE<>1FD|)*S*L}k7bPH
z8<H`Ow1S0-Z&M=|1}-Rmiz3(E!`Jl?(|W!fP{KdY`*S0|rf<voT|L9q<Gh}7_hc`3
zB8{Z3w*P~%cMQ%X?7D?x+qP}ncJAbkZD*2+GchK%jfrhfY}-yIwv#XK`A(fb@AKpN
z*;noA-c`N3dTp%*$AMq0_>SOU+qH?x7UIDfxP;GvYTO_&>h|F^Or6Vi95nlKPtNqZ
z#mH>$<90m2w0pm?R`#g~gOQ`lS(S=#B0t-2P>;D)_R<{ITE4d>)===nmcihGHfH?j
z&GMkUXDPWT;qX%>(qf+yP5sfZ3E5!!>Itg3m*DtJ^;*r$*Bax&oQilK^BJSt&*rnd
z#f1xj--?r1G8Fm`W~o>&Q^-Co`^Q7>X{^X>`|_=eAx9Pmm7XDukru)C7-Y=u@WX&b
zN7{whTs9qTT!)U|`-8oW+-KWm+O-RFxH5Z&+M=JHtuB`9=2ErncRz@cDdo`CSisQp
z474liCR$p%rz0-N5|eUK42f^6SvKyN2#XVy3F)}}?ORhQXt|-EAKf#}2lUinn5kWr
z@#BLbq!9U-`{#<xU5cp>@#WsiM6c_J|C2f7df1v@Sf?f!7`GXWVe0Zg06g&eP<NzI
zW7h{;s&c;SJSX7!IqHg3v=g`)yK=J26f?{A!?x)QhEeyExK$Q`01987U|+95H>Zi?
zUmLB`m_^l8i}qyJOfsBdISb*v))c#MZMLk2kK8|h!dH5uH~eP*`F?>j60!apPzLsi
z`yoT7#IERJ>1ojXwJ72H)YaV8;DyM8f1*!<1c#<RK*Q&YSLybFA}6VOf)<fKwU3>9
zO<zM#C0j0m8bg`maK3cjzqK$6u0H<MKu)4LE>gef7%5Mwe`NaDv#${vt}j$`4~#6Q
zc)iKe2VT#K@AY9iACcIt<gH^S2fuZmi1LzSWaA3_IyiTK-~Lv=B@(d-i18a?6<iyv
zOD{tS40y@f5v^NxJ&O)=4GfE&envt&ua{XTuAA-GCnz*O0mPanpa+FHy`V(~jSZtM
zlyI)?Elgs3k%&K`1H{29H}5?Jj2W|*{{5>>J(*UT!MIe#`dW9XuvwvcU_L@QZb+Ol
zDPEO7twQ>H`|dqj&BwZ`NkBX)@d5EwvC@Ix=%4syFV(T{pSfo0=QrcMxo*5wpCJAv
zJoO4u&&zf9E%$xugg@*7U)Oviq!~|L<S*B77!r?hwOX%OhZ_S6er>~fVli9P{<tk3
z!;dmDQO@6n{rh;bPB8RcSe3B={y(By?*EN$d0DvrJHm;K(FF1qRFSI>%=d4;QVA-w
z@Dn2lSw(s=?sZr76AImS52U5Ro~r?XNWdT+R_W`EVm%fEavNE`vVYF~>2b?hOG}Th
zwm{vj8IIESCl790{$bCo%j|4KwCYsGe9Kj-;dT=gwZGPjcvV%haLSc|(CPai9<wq|
zo5{ord||}2;h3)w{*ysY&uD4H(;h!TP5V8>@Ku>PUxTk>75I4K1dc9?=qJ<h_l)ap
z$v6j*uiwYx<93U+*ntMi?g37c6yL|^$HnsY{i&hLU1}nL>-n3cuyA{SUkb8d-e2j<
z=k1NI*a*|eF3C&Ix>>Y44}r*BtZ3SBgwJDo;kr<-J6YPyZz%mav@44=P`phNEMnsa
zd@z)%=bf4Z>VVWP`95WM>Ybq-PY<FVfNU$qarkqd{N^W3jj9(%r68Wa;n*>U(93=!
zf$IhqW&<Xhl;;oIR&#Sq-ju8Kr@tAhDoklXOB2IV=4MJuF&~W<)?^`0>3Ug}@n=xe
zhAd`LZd-a1tW4R+&nH;W9!~IPK;liXXj^f#&!OLPBN=wWj~>>+<r6m3bCt%DY6bao
z<p9uH6U-k^;D5#=D=QBRQ!ta8{mzNY|EkPZ6yVBM=l3QYVhTnw%-a}}>xe&;Cg)mB
zdK=(#3{H%)y%Zfpu%|%JD|V{qJv+$|s$i9dAxr7}RH)MG3yGkuVreR>3@c2t3Belr
z926jTUqTe760aUlRw5{L7FTMnDg5o)9#@Ka>cV!5igT88`GDK*u-_-y-YeIhcG8a7
zZiM+)6>XfX&T9A*bUH#*<Me>W=f!V)PQF%h9qJywUcuTSkzqT7Axt7F^QS&3qtx*7
zl7qJ_zFI~G!5y4+ez;E>K?54O;G?5B67Ro1=+^3X163=K(DDiLK)tF_;W`@5Ff$aN
zy#yBha{ZlQ0r87pkGBf|Tzl%APgZ3(QP`BizeH8qcPl+rs?s#?LUhY_*rh<n@q^;g
zgQn2~^L{fWHeABXf!y3<nLe5zbsCb_>$JjHBGedo$;t-PFS1nG6AOd}<nfe%v5%=w
zH1mGm5n8~7gK%M7KKIq6wWjcpE}+@P9bd{Mo)wyM_jcg9{x6Tz{2HOTxokjx-B)^Y
z7ZY%ePF!SDZovwyAkRjRY->EPj_@d_-S#uOV$Tr><45Wex1$(fpWv&X&va8;*dJ(T
zt&myge^SUih1L9AKXaQSz#VVBZ`Egk;U}oCQ~Z`By#E~rvy}0S+h(uLVyP6dNYbN?
zsU!M-kUo@edZlwKV25Y)CXDb~B@LI4dTOPNCYff4-n@O+J|l@y<H;Mx#1iK!Sz>`}
zB+tl+8frq<JW$t^Elra9ib~7#b_HFMx5#PjZC!*M+tY1Al$qU5gI#)6s$5Df`|Sy_
z@el>)$PFEUx}unV9J>d7ZBKWsU;4z+n7^|M>e|6>PK|Z~G_M;JCcOkHQCfH(t`Qz;
z=~HP2qOD4Tfr2;5cF`0h4@(5cr-82fIDmRits{+n{an|GEOiNF>~nD3;kKxbwy3)-
zF8}f<TfoT?TX#ZVd~jtB^7<z@CRmi|uf1KgW!ngU`?U*Tf{vY<w=YeTgB*QMoX#3;
z6OPO%fdzt=3wx3iCqY_<tZqNdo*z}u&sB(K#d?{<l?gKNF{bmS&eg!fuBJb|nq?c2
zZGWp}T4s_~!mOC$>C<uLjPM8uQ`1RjctTqX`6XM@xP4s#7B>MJYd`>?xc=PWLG3gr
zS#E1QKDvM<bu4)TAzSN}MxCy%7KsB1wAHB5ZMu^Z0i&G#PWIQ^4wfF5!#sbty#N?0
zQ~0ep)0hDE;kZygOGN#qwH~WOXJI;HpWI&PEiJkL6R$-#zBrwd3OlSDI0Iz+!QZT6
zdxps@N`~5vwfR?8XDNVEK}F7?9A9Yo7=czXS+Fz2sN!aO5nv;fJ_{-k6)urzE*R3Z
zcUq=kyGc_01Ss2fu7{>I$EZqZjuM+7QNt4KjG!W2`CBv3VLk36fr82d2TIG33*FG=
zp`#dug5j7dq%p^a^rX?%YlDknjt9y~9g<BRA~juQGf=T<GQK#|mXoWh#1lqtx};_r
z;$4^X_*-^GYpXR8Q3;e-auE{p*PNBgmZC;!t#!hfSfkyDPPR>#rbh#BavE_$M(KEq
z&!|q|r>q0w(R$9m3lCbf#x^|r_D~y5FFKVmFrfX`F}@frWutX4yqkP!7EiE@^4mjw
zv0}ApS%DrvWUmlJ+LJEEoGnEP7#{$Xd;y@Stn-0L)4r99_S%<A7`Q1Ytfrah-<F7D
z%wRMN`GC)7JUHelKGY8GvGM7e;@XiK3p9vWljZ`!`i6HNEBkCtwcitPY@J!<u&la%
z=q)irS%f8rnf*ZviT%)3M|L!ez7bjgr=taFw2uYzql3eue79|;W73v`=CeUt@er3s
zxTE^9fU5v)p%#8LCuWx=lcxJ5#*$-w0w>le#uBP|G@|Gpks9`yXkK&UvX(S?8X7Gf
zTu@hqndcH+y2*Y*TzHOxx?y&qNEHH!H<e4l(q4~Xor)ty|FWz}16MZ^9~D?lasc2N
z(#qnUNlC<igI<0iJ+brR-2*R(^}PUvS7&!B)=M04;oUQiZ6HU8u1Z@4777_sWp>#V
zToDsQ0L^U79rQuNwbF(}<cZr+t%pD9=LZtthT|Aqr&gr>DCzd~bbH*yRz{1jnDA@W
zpJ$oATJ6LHIQN%-cZZlFG$PTT)?O#ZIh_#y<xOkPA0PFY)Ihs>zz6+9Lb5A#!V%y`
z*`4^bVOwu;OPjVB=wM87ckGl$C~;KhZ#<wbnrhP=$42tc@@l-u2fGQ&^9%z$qW`?g
zqdT-@b15>u_bV;y)Mv@O$tV+73_%c?euPIu^!~@m_L}Q$>FDR;N-;8>uaJecz>z?c
z!W6@EYTirN9erAi7FF^ktN1vTUXyR7VS+Tw3&*$(ZgXuC$=q8%nR?;*N=IB4A$eN(
zE7o05|ENd$UyG7a#ou+-bsful*Wfw~v}$|}^*RQO3nmx+MA%b7dU)-*<&*;0Jb8ey
zZugvB0LBX&yNBjVxmFh=zdm>IQBeQG^bU9*w$je6YVu^cjVMi>?hZI)ECns$kdeHI
zwlb=^TW3su&OK{3=zxnoaDyqk9%{dXj-<`U#VT>N|6V^rKCb4*^qsR*Yicy8RR^4w
zKAM{aAUFt$Lv3=Z<p1$@9|nT2bD_NJO^(x)IN~2VYh_H2LrSy2u^x8m(qgq*DyS+r
z7-AiGyH)n_b~ZilbVw}xGr?%y1-T2CU3@VIi;<FFNYR2I%%{!0xxG9t3b;LnlUj@O
z!m)Ww=1y(ncv+^KE)!y~?I?eyZfVfH&26=J^tYU}EPG+Vy+WUPT&_&x#F^X}64R%3
zy&-&<5Mue0Du*DQd(Wxau9N{O`l1@Hnu3Ll;O0ljN?;peg79XuZ$jLO`zsq%Jd-`b
z>WmTy%Fn~Zn%KmB8vCPCGJyRW`)>NLZh<?*f~?p)Vg(K@nf)?jt4sydY(fpC`NTVa
z&n*1AQcJpwnkiui*0_sf;cbf**00pzHdg<nNACG$tz%^pYlfIR|4zZ$pyBcTtsUb-
z^p`}ByfjmJc>x`po}j*Q(@i53<QMNRS%k{Q1vt70Q7h8F^RE7$QTVQ@G$Sj@8L#iP
z^BfPrk}mMR3h_;nyaxPt0*E=-X8W@P)N?>VC-~^Nsi7rWJ@Fd`MSyyrJB671=;a-S
z0nrT~OfKPC>91=gRet_>5r&ewX>kwF#xoQi?<VDd-rL<>q`&$^uL8lT(j7>Bkmj5p
zGBLF+A;$@1){B+Hkg*8`9S640@50;T6`21Cr*r;q=1VpfcFz9_r(fuGB%E;}Ts(af
z!;Pbm^bSuDPFH4cP87-4>r-hOC_2P*ynOt^kd3HMZT30bkO~19phA4J7tmzX5`7Zf
zUvJLee4J)-iwCLa^zB}M4uj0)9Wq;aT6sEwjN##I2_p;p&*4?+Nw>+^njW~^u8wCT
zPeXO8{1O1=5#SG$`AS&d1U)ZR@K50Xu7eH_mU5}KKJz*s1+nK+U7iHb@kSDmjJOHB
z-+q21PR(nN*%8v7@~5V&G6@yU9>mc6M4KX5S8%{imys$`|I|B_SC(W_dSaV@Ce9uD
zB(IrEurkMxDa2rMM%epn&J`0&u2IgQNiT^T8xvxR7gp+w_H(R$jUi~VYY4Sq>YI$h
z3D&W{XeIZQ>A^>Liqo~-j(4bmtbkpKbQ)3WrPYehmZ+PMRkKe@d;g+i(Eq#s?w=c^
zSo!Px`s}!LMFcA8hUb@N-^y<8c9?LEpSDYOroUQPbINxV&aknkLj;ix;4kE3$}Ybi
z=L&}nCK6kQ+El{8%3tnrVNYf>wvsTHJ+Oik^bH|&(g1UrXxZ)<eySd-#bd~%8x@d!
z`KfK?Ax4-yvYxh@bASdGF&HN-@kBNXe-(sn7hQPA|Ai2K&!iESwowtq_*}kD{H2Z~
zfGwM2S#U3wM4~Ka9<3FsI4h69IQs17j(OWH0h-&TFMtVt_RiBbsSAr%2*e7L8g`dT
z$?&uzi_om2B>=-H#pb3WAUEV`#V}VgjU7-apwNlzisHqYa?k}6le|lk{H{BIkb6ev
zQo1iXbW6aICPVY4u}z5qcLzw5j#kdhGvX(xG8E1X_dtgtqw!!7mL9InSmgL$c=mYr
zjaBLzwIw8t04oy6?g8sw&Weqkf4OfwC%Cd2`Cm@hogx#lt*s$1AiJLzt@YVT){mC*
zKXMN~R#pa7P^?1g9uOcYF=Y~yiBihzK=g5XaLPyWdF{;pP$0*r?w6{#RQ`sLJ<Qru
zPYYC;HPks9NDkndbz^0QU89<dy@<9G`gbwwzhD*s;Fi>$>YcCuL~#3neP{I`=mb;5
z4_c1t32K27cogMj_sqHYbzH@;&-K_j_{{H7(d(+jYgp$>cy3mVY-~Dg8@8RV3z>ab
z)E}6AWn@XYpHcNuP3RBpi0NUMR8upF!naZfIB9w8uy`hH*PMx`!Kdo+J~hHN)R}dR
z(&zs~+M>H=8R$v~+^h>C)G^Om#reOM0a{`jb^LC%;A#-MZ~w<oaVNkdzEVJ2x7hs9
z&~W;YjSz&Bu3nelkiH*j2M=W5p(^F@i=cOk2p*F0DO%8p^!u0&M39c93flb<IO0iJ
zF%S)EXTAce2bnhy=W`a2-Gx;TaCX86Um2?fZ-o<igsXe#oeG7&dDdTV25DwH2dM^&
zs}x)IGO}C_fk}SQ^)~vSN6~>G(t>w`);2tdNzq$2KlTDb341j`$mmQJ!Hq1xB)345
zUH_jcDN&-eU~L1e`0dz{vn}%fc|3UD9|P1DqLtMM*6-7*O|0G!1pZ$YAol>(G=aD`
zepa#sIFSG6m2~7N2Z+17NmOS&<klLvF0MO>7N`|u`f95G#&!fNL6V9c!vFP)H)nAW
zs&`5$3N_ny?f*CR2C3%9`RHHWJ&#3ZY`9v~46ma>o^N9tD}B7EWHJUqidbrboXUFd
z)q8XG#Kh!Ty%ww_#s3hhdFSFz)j9jsB$dC!6?GiG%KX+^21k~wJN1=2<aydMc4lrc
zt5QK+Lah?RlFGgS9P)DN0ElN<a1sbDdl4`%x+*l#F5(@x*HoE1xZ>6mDr$@-^mE+Z
z`gr0sXTjCHju2c``Sp?nKuiXe6RQB`-Dss&DR>(nppM*pKg6r8te%gP$2=c`I;nzF
zG%770;_TG4yR2p0$HwF(_<c8ND?r-%&;4-e(`9E`2jHSz&|ugAA>z^UWTFI9+)Y@?
z#paS|xBM@K2<k`*hNkC(1Z4Xj4CiBy;S67p>4*9LDI-`t^7@l|QlmH+?Y4%*Lm?Ll
zj>|hC))E2$(N!-v@^;Ipi7a`!6Swl_T0FS}pGoHnb2>`dBwiissT6j_in`eH+73&1
zNz((&*K0}bYIEnp7dTzQ%et-N%ng;%@LoBA?xW5-oA$W<=k-d0lS321ox#`cx*ezq
zRGTK{$pq6OD;I;w&|}`m&`J$<gjLyp3DiABIY@h=b_3An7;!%4Vjf0Hz^J?qpt-D%
zzX8yxJpq_v^Srk~<_WR^L^VU9pqg1K#tOemky$UrG6HO^FvCh$05@oW9FY~<up8+*
zA|`~|4&2KJb6u1citME7t7ud%yIW`XCW$-~a*>Mphn_I_ABr0{{qD8<TUCo;JU^nz
zwd$8Y#yY?vS?PL@UUKH&xN%BV{EL13zb2}!5Lh2PqlRYaD7crk-wA)Ic^X7INZXT6
zZqaTk=r)SU|C~OGOaUP-B5f;tidH<_5YQPFw>zy~)tilj+C@a3OQ13@nWvJJFV5ku
zb|lIx?yy`~)|iodgHBhEdyqVHNQyq?u!wy|$KQ{o(P(*dk36|)*A)DzTv$VhZ6<0>
zR7gk+Kc26u{FxkGK%~C3mgr?*N$W1<Taj^@qim??u172&p@@XR$~EP#S3WplNOA4%
zw!!$c>h;6<3z@V<IN?76i2uMW%f<5F0f^0vNtb;tgkM*Fa9zCx5TMtVRQu@^nPwJ_
z_O9!z>g${Gt|tbsuaJ^s%}JHCtR+87pn;K+;=R;?AmM4d>qeoao11r+snUCW;wbm-
z9-n-9TP&{;QawM|JS3cg`|rlDUxskb?px=BqDkla&)jAj&V^B+2m@Xb_dbIx74{v+
zeS4&Tst;|S;m&v6<L7p}?Zf0&OOMt+{cufk=!&_|NrZ^<+PLZw@Lx+ef%=E5JcT^n
z;qNXlJ^<C*+fU5g@goQ}AdpZ6C2;NOqh|N)ayckmwlN^C`u@(Eiv5>r`V_C=8^zrn
zHv4-SxA3V^r1A4lg4cvl{BX3ep0I7rZeB`al~Md-35x?&ktnTQRGQ7PgJ?6Q%6r}e
z@oyj5BV4_(D-WJkdl;7s?zMsNN3&9FP3VahMZsh^F;}(aV)798?dXiKbfc42rvaxd
z)=-CMiwGHRRV+qO8ZW*$wa`@lV<a!O^~Cx&S{(tR4b_-X62DOSSxKPJ`|XG*h*dgP
zEqk=>{Tp&H->M+xn-f0((>Ru{c`J-+OTiue<;G*#aau8pW*Lls;%1FZ1>MMk5cX1x
z<_2)6hsB3#@kATSF(ITvNPdbL!DQ8qLz=+R7W8ElL!fEJFe1%YzIjrixlo+{l_-E5
z<}kgjV~Hb=Z9-W4_V1dv`@tB6V|0GwB&p97h{wX5Wl}2H-=3%;E*H6b8(&>&SVgsD
z2MkFYpRGk^b?jXSx>yufQtU_Z&O{Ej_y_^it(UsRBwc*Y#U;r+j<;@$oryN5j#TN>
zE#T?Qc%-L<%y>jt?iS7;PlS2XnP+Lt40R?8fmd+Ic>CtW1Vxg#Vcg)VeSOZ&J4yVe
zNaFaZ2LN#!ACv*D1QK{w1(WSLyTb4ci{b^vrVGzH^%q;LK*QypcD7RLQ|p)IGiX<s
z9ajUMiaJ6G7xjoPls1dGKly*br9n-R${17A65m5h?+t2B^A+Hia*2hy{qeE3-yeT0
z<RbDU9HJvIH{+JfBD_ka1QE9g1FU>#;bot1E`}qasT&Eh-PML5moR!*C1YD?YSw7$
z0E_j;FIbGaFg8y^*$Hgy1Oe60jOw?Y*!7`Vz0X9<8w@7J7S^N4*vxm0=Q4>R40-%Y
z>5Y@3Z41Y06Va^2@7V=4&*H0EWLGDm>WE2LOJSY7+(Z9{Uye`}O%D&=*3Sd}<q3D@
zauF?bMX`Geq^~z|b`TmI7F+E^G^>Y<Rx(3dirSC)@Mxu_PwWb|s{PzbU8^u#+#iVp
zRwKGxglhf*>UZjwdL@mX*ySu1hD9p~Pi~3hu`s$k4ObEPZKS|8>Z<1#kt!vhNWhY>
z$c+E9+i3LUq5LuXvS6ngk~DiVm^~-A6o%4OZQ8Gq?oPUBH|wc}fva$$(W7>sK{k~M
zKQFc8yUNzX7{6O<5Z$B)TyN2MzDUu4fdp&JwgJNFSnPG?vuHkrplm8NT?EYZ%XSLl
zc4cTZNMfo9mBRbVc!H*Q+IzQ9updD;pX1qEi8b09knRm>Z$gKg?%-5!x7l))|CMt|
zy9Ag2O|tl<7Tw{lPme$Yy(@&!*qGo{*8RN-f(gMTy(|RN`09Fh(AMO*Tx~h%H^JHv
z`?2)Pb)m&^*+DcUYP7B%85g||zguI!T7UTt!EQpe79JcQ+#|){`<9#Z3SvmXzG)E1
zLap+gm;diO*dc^C^H<uh-mHgX<(BFf>dg{y$~WAu2EM?dkM7nXV;F~1J=(7~>8g<S
z^Ey3}_7@4C&o<d=bp-YwS*=1t1H(~k=9_I>iljyYq&<~~jT*w;6cZ=_Q>9eCukJW7
zVpH(^_ZKX9z57g8>%JWqlk6_#6imhBpLv$4{64>D?~Ycz@hx>86MY|uhg4??XdP@b
z^C5bhl`cK0dm9g+{bJGWu<F#<EoRa-O6@}{AyMYOn(dd6YHroS(H4YM&F{^2bLY8j
zTPnpSbX1rxB_<-}63!kb*Q&}?)OsW?lg`L(B&9FgeH+|=(-lQQRnL?~Nr?Au8MZhs
zIeforQn4B+|AjTAezy#ZDvPYJeQ@pVvM2-?vYU3q1`BVxuaMmNV5YQAivnLS^GTf^
zIXUd(FOaecQO*DKOX6br-@PZc|B~#yOLx@e;D6K{zH?d&7%-Bu^6KhaKFv1D8^&by
zX89u)*U>BJjm9>{@>2_>db0?y0BRI^U1Wb`uV<I1XZNRm`njDXOD7ADH(L<T|4C=t
ztp6vSB|~sD3^P>L!KWWCBq-u6AW*9A%6jpM-7$w4A^dgMMGmu1eCzrX%6n`-_W)rR
z=rJ79!4>1Q%cpkK&_A!R$%A`J00~Ee0Xp5&_2E(4fJ0)bon^Rn*}Z8GX}pi=f#>(-
zg-wJKDCP)dq+)RJ_IZ50ys3;HA*Mr5Tzznrim6!BIMcW^whEci^H~xme6SnKh9fQf
zjvwZS*x!WbN$`p5hx+Oj(FrYTpgZ0vMx<TmF6N_wLqvJNNu4D#L!FC6H2nmC&gyR=
z5sfi8bm+DajfJoc<<M~zvnQQ{qa9MU#uFyhgk-d!T+W^`VNN=SW8@_v;ImR`7m5;P
zufaqJ$AI||!iiH4c8Czwt`YUC_3X-^XR$hJ3Ir_(MCCL71&F_+7lg4DIiN~i8Z>^%
ze5o99jAFCm8+AND*q4K@-<VjX5>Nt&A5#kfWBNxp<W)qavZ9?=U~x$X-Qg8fqW%yt
z!-n>wa5U%yh`^)fDWO<t05YorXqZ9<o6sz!P41j01FesIiP0ZN(cSxkr3)E$;S!Id
zK&L$j2J^C);h>maT7+;sU~eh7bWE-@6u~L_M@w)JO|kQ#j#2tc$n{cNkjR%9fRL=2
zoyNoiS%9g^K=%IC1c~VM8N9nSms6f7Irb+^4W5@_p+DDuvYJwt)D{ZEl8Ktg<{3mG
zapiIvkl7BBm(OO-jSYlp)26pZN}GP*#9DOIdl=oP`tFFxj*S@pkih+>MS`Lhj$lPF
zuujm0EZYT-)Te}Dv$+pBr<|&$P~B%JZ6|Hnq}H7?>>_;^`obJ1ut=DH?0G^@{0Wyh
zt8=&lh_;fC>>~A)u$E9M36P2~FuGMs@>SC2*7)n)Hk4Y4MNy;d-~A<MclrKqew5o-
zKeCRy<Ed^zQz$!%y^63@v#dOhBRuA=(=NcsZ;rd&mv`5>7v0EEeNR2r(cX}yh2~Rx
zlnudE$>S=wm~IsbR@7bDpTwWECWgzu8Eq`mhA)hCoUCVq>c<%fFjdr9e#fYgmVaSB
z8_&;(WhLW#trEgx8ym;&tUpS0+kmU(X=bSQ^%lGJ9aQui%F$EYBe66~CUVTFl>@_E
zNT)E$5fH6hHYx|IOvx2)9ej!7#Q$JEs^JX|$J0n?+aWbjOJ;>4tU05@NvZ2>{fbkF
z(9S^44TFnZfwh0Y#cIw!@>#2^@Ig0&T%pyZx{pMt+;EN9rTHFs%{Ucb$r2SlWMDgn
zb}mvjB63^4NYF%KFY0w0s~nYSb<j$G&`5BT^ez1rdYfI^Mk86X49R&hr=bF-0mXN?
zMiCEUyK$6W_kLe5aZ>WSz@Q3AYd`SIZH!rVJblaZp!;A&Kh7q_ikOyQa3MC^MTvm2
z{ppNLmeyM6aPLdp9Wk*h5@^kJxx7HPMp;I35CY>qCO9jfX1P=quU0w&vnxJ2M;&Z{
zh4hp4l|^k5O*5{ZRqq5miG6I%4e;6u3FX>|H*To)Lp=kxW~Xmd^9ghy(8cThOE>O9
zO>?H;pyVxYP=~i$6S_Wo5e_2F$B7sw*pvHj6<^#%CxC@zwBxa_*W++@qX2w$%2Lu#
zMyMh7og|{ZzngHx9hPq7<beK-q_7Gvz+o}a?%0hO9G%ak%e^Tr{!sGre@?mrBb0M$
zk*RVPc-C~ANY*{8jCKnA&%gu7i@9~RK3kB?8jLVId3gJUegQSSx~31C5l7rF>Hny>
z#QLfj7xvy?2sgv4fI=m1pMxU=fWZsC#$NbnU9*hpKK6k(eRD<xw|Ros>nGre{gLX=
zJ;nKs4DCdUIJdX_(OIN}0KqX98}rgQFnRH@RK`(`DOzvYb<;f7<fn0|1B{{^#pi#g
zA0FHmb-!X&Y;4R2HsWN+<soBS*%>CMSG~<0Qu3pjb8P+ix{@trH*(Tb^}}L2>ap>S
zQP4aev+){?B#Gnsxh%NZ?f8n8Rih@?8T*cvQ3#Szs%l&9UmT=bIARYIV@yULDOa0)
zT5{zK=Sl;`gnmnP%~O-_W<--7weZ8X_O0@mIOA&CD_Z5J7S@}62ic-Fp-Fq3orLE{
z_qB76k=e9LttJ3|JwL~Ysv@>6SF@d6V%`yJY>tYX@x^;Y*xEyHVV5|e3KpvuB$P6;
zUSsq076oTNmgKK8-m~>=yIvCdzJ%R1$1+jk`Y6@$?X@|zQN7Ju(teeBsvBIBB*twY
zSo-ylwHH5-2jxwI>r<_tG0U7KfMa>dw=P3W!DT^>dk}a+dyX=13DJR&EzHP;xTDd6
ztkd3XXX32_h9^?E@hQYR%K^OgyXAl<8=?($T8ivy6GQ)*jEK?JDAlVpIRDXIpLZO+
zh5v}sZSVt()oGHr2?IC?Y2$PQp1DbSI}vO=;GH;1zZ2B$rM^4*i&tzEWJMgkUWkpM
z>?Xjt1n!hoE>PchWmt~xHH4@}bxsfxy@gPR-WWP0tZEp@u7aUQCT02_jX|IaylT1g
zZxlEBjKJ>7idRbYO15k)U7d+k8Z`D|r8tRu@)$E37r4ej%8%uXiur1PhdFGH#=}nB
zoijy3B}8R%%v}AKPn}&j7_y4!D1nIFV^lcbnNw1~xv-6bWqz2v-HXYUt7gG8l+=HQ
zqd+#;M5IMtMV#*(e{n2<q69Dgk96uQ&6Pv2r9?uwaN2Y8HA@eJHn|Amnr?yj;K0A#
zMLdt_G6J*yn-l2Fz)zSUda*HQ9by^X2jjF@#8vPSJrjaEQRwOtbwgyPThgijvs3(f
zv$_X|#z-lSed;{bdg?9PU)_)-H)!3&0Lr{941m@w&JWIP_2>^+pDq56aOZ!I+HkP_
zm!hwWdfyp%D{A)}?K9XA*k*JOT`FC|@K$-3B3sJ>ky_%!5liBj1Cr+l2ObEl7>4ZV
z?fDVOIGAYO?DLGjl?ZYVg4grS?)~X*m{w6=g;8|a?tE(BHTb0e_lUM7^tbM-RjBfI
zXpfLJVvI2LXWUZw`^~%mj-$7S1bS=eE}AJ)Y>q;?Kg0A#Uu&t_S9*`4SIX8+*@7%$
z?q_PsMc@AX0-s0qsbDb{LW|YO^<8N$(GOgy55JfDnZ&D30Eor&`^D}u^kLSGk`2e#
zZDDu5VX1q8%Ry*w^ozz_jqus-)%!uID1DUibIrnA%0E2JD#C6gf6r&0?oaJg4?7s(
ze!JV4H5LCammIof^p@}!6HfoIe)8E7J(I_k`<2u$Pcntw4_a*0yG%9BPy6&lqnQ$^
zSm;cgZjC{S^r`{xH@)&}0^rWo@;f6Tuy+mh5YURFc8pz~$+mv4R-QSTQY?c_W&lra
zHXwij2kRQQ<$<@2r2rfvNkEr*Vf!1mEjITz7rBW>@MwsjRV-)Fdgej#tQGFAx|T^}
zNwu2E7r&2kzg9HPxQM4;#tbE-7K%(N1MXcH8-p|Ym)4RK`4sK+%xdg_W-YFyPz`FZ
zbHw<wuZbDEr1dN>&?6Q{q(+m!WVHd-7^PGelZ$c|B{4h>Hgu8h9YKw=-ZuM&o{FLq
zY9$XA;oKTF%DIED(M!1rVV`i(D6#lW$7IW1s{=u8C{@xMQDzy$B1<ePsKG!oB^T02
z(rX&_4@gT9%LeRr_}ES4S$i0r@E=U^D!PJH@HurX-$!km?jx5cj!|G}lE%DZGDJSg
zb=h^s8*AzP?xSky+U<O*^vIwRh;%bUa2&oUncZ(=G%p;QN!G#i1w}`g-oz3`!o@?1
z8nK`t2{Pu+@(TL&Y1x=ahNe?D1+;jxvT>857jg(@Gyf69$dMWs4xH1kBx&{4M30GU
zjO4quiYy9)ai{Z<oin%h;kvC!yjHGA1pYIrd>*-XI1HEh+WJFs%2Fv&Lfemfj%d}K
z7{(R8fEr2f5ZT!zAt4N`_R17_VU07HjynT|Qb=7pDgwb+dv)t&MOW5n^Ma~kd4>)S
zddm&AJcfz{_Xx)3Y;Pj(I@^tUmQNS-9%=Nvs2QITo?XhE5+UD@m5&Vh46%B|KV<E8
zpQiycwL1X;k^s7o3*!FN;1bV@s#OAnerf*U1S_gr4%0}eqg)f+e70aX_TVa&6u7Tg
z<`9p)<zeBT54Ina{y89%XUT{z>rXS816oZo471aC@7JMFg?m!dIzJk)sM#STdw(`P
zoIO-JSK>Bq<BV#o)9&ZhblGv<wRC3?<ye{T@d}PU&2qb@HYo+}ufB7y4Y@@#2drGT
zSa+vlM^F_PUSPv|)23kB0x6n8%6dpPL)xe{GtKN3oEHcFY-*sJR_ykkv2ae51Z`c<
zrMV$IY;`u{VAc$uWv<BLvi`LmFW-Hs^X;fe7^0~iiEbDVchE6=x9%!t@l-A#(2nMD
z_LsoY+zt(>Mt2c9Kpq-8t>A*-HZ@#Qbg$0AayS)bc_E!I%2lV;2zk)MG-0f|mvfw6
z1g~JMdf=t|WesPcO+@{41YAkehmf-Ed^*=LPoxpA$a=D^@NwPXo^N4wFnU7Wwq6k#
z%e*cz-j>@M`tw3)l6d1AYOmtycf;7<o!%Yo-0)oT%Q2yPnX67c|C+m!+gmpOKEpQ5
zP$Vb*^u$K~r)u6&+DR1_L)`(l(e`K{CA^(S+RsL#4x%Xk3dnWtXLtB!N7ibcTKx-d
z?b<W)Q^u9*M}F;EUu3U%A|v2#)0yoyNW97ZqaDuSuFI2AGIV=C2S@@~@b3@ieh?4A
zA=h%p(`_nWk>!|e$idu(<cidRaZUpewkeVs!7dOveENXPD@c^z!jYGZUjP14xvQ0U
zs?GMqVvB6Zf_QgfUt`Jp*Btu3jV$v<#wR`O`r5HZ!IoXlt*ztTT{PN9@#eW4Wp^SX
zC@oAwdbaQ$Wm>P}G(aD2@FTA3V$yC1OtvI6vziR=&1Sa9)Kz4zP`9c#;C~Jp*m_Uy
z=%3(hsL}jWT4JK2UUk&&p!d4pdknU02&Q#wE+MfRHsLEJiTbK8A*rQ54~K@UTs(aw
zfDjhoN~-kavkT%w1W!0V+Wk2pdTpVF6o|ZhK*EE}?~0N)ki+GIGBcDFR2%2NW1e`<
z@YNe=I3g@U-w|W`0zJ_FS@I5ka5@m&$_Y|;QO)qy4F3Y%DNt#V|F<G$a@_lnxHkJR
zovl&M^x5=?7=zxL-d3b=drTNa*-~?C$Z#1iqt&v8xM^KH6SkFJvrlyc4-9@&@P}l(
zAT%hX>1a_G*dJRGdA3+gi2COrebq-K8gUJd1^b%`D1rCu4__S=q^-#m5b;gNvM{U$
za$@1mw6Fnk8DlvV5N2EFvSQaL?GzA|vM@<cNM54<Lc)t7Acl^TVPPyqOwwMh*C(rY
z!XK(v5k#*BQCa;E+&9UI{?zkEchemskpKDj+IDTRA*5W3`k_$gjHLS<cYw;+3w@Rp
z0wt73uT*}S49nc7GE`vb83j04dC6kQ5->6KR#~iLNFSA<i=YL2L*OKB<@-gc62Pk1
z9=zh=%V!jaQA3I0?8X3#?}j>!##EnUD26<c_nx!@EO`BW9blCVyX$uyf%N*h+w?7=
z*f;l^)LRr`6@b8tU4tz?e7lMvzH7rOp^31{$eM&+%R~;DpM91(L~dorw-oIK#OUjY
zAT#<Yn6?5rliD{>&AYQ{hDuB5v+WN#Z#6NmC?baXgbkTE4LnNBokFJ}ldNnn6^Mv7
zg1aJ5Q+PzAGcP<=|5N3vuw`LuN^v3OQzv`I4Vj|t^c!qkyrXt#gSjV$S^1nt=`UU7
z&UnM@(MQ)cCdTqvyEGjd>12)4yurrw-b+$8*G>7@ZHijb!VZyi6aqW4GbMp&Q?ybs
zj~<qs)ujE3DE|51sNd7euPOPhW3$n<&8e8#x_(i7Ge1-1W0Ip|Hmg_J9PBjg@V1pl
z>-6WJVpV9SdPk>5Z&+4K3pWv&!#ab-8s;Jkp_b;62(qaN+{O?YvZ-HF2=_Cf(+U3K
z3{y<2Tw}J^2C}=$p7Po?jMy5)Qp7J&?oq-T*tP6~X_Hi#XAyChG{0<>aLxCk3y<dR
z!^No8j%~_+YFBiYbo=u>c=Q1mXjVh^S`7o@)sG(id>4g6>F+UlX@*<m;^^puS&E@z
z>G;FV6#pJ{JuI4ouggar>-Y|KyGJE%@K)_WR*|`Kx0$ZA40roizU*-N^8vnBJl4_N
zuZ{5N-BQXNo3@R$h>t&4>+YsgG;yq9m8W{+Y6I&_B#Alo6P~>KvH2(6P?YG#v`lW^
zX+YmhoI0$*jWI&m%{n#$&jXnvNam%AQ+6Q?tTXgJjxl*a>FS@*gr{f_ORk9Rfk}MQ
zNs4GMsz(QCd`^+&;Q{J1stO^htdInZMaQ0e>bpvC%N<p8C(ea%Lq4C5kfKcfTgoRS
z7NdK|Xk1}32}lpMJF6X-(?WzU1xfWLtF#%86nk(cIEa9LLe%|ft6vF_aXfTkZlWmr
z?jwn5<lR-dL>A$!`+BK3k8C#Qger-GXTGG8;e|!{zaL#Bb%>Q9=mo3#hX>rA_3&}Z
zuV?=tOwGS&nJm~*(lt|GK@vY6yPa4YrQlR@NyQKo5izQ5kUH<tIRA+`^5AYE7zm#X
zbuDZc^Ub2&I5XhIW2`&=T_@qM50N2KP=R)4l{O4W>zJirVqu-7**{jZpoFjw)Z)tE
z&%usam<Q<w^Bp^7KB{wgPg_1<Sm|n3NORO|q^hsujFK?s6X1q|I#dI-LyV*&7?~+(
z11D|nr#ZeKV%+xXMEf}5dwE(=DTb?>Q=fdOH_Ey`U~Iiry+N^P)rnBVcRZXK%^rva
zY(3{dS((3&lm0&9h(>)t!Vt`L^ccKC#ep!4O~-bRH5x}PbE!t1<)FGLShU;x9=1-<
zhq>JO={iQQmX*pdk3~x#V7Wg>dfp*RA_YfZ9s^}Xz!>4NKy&V4$I&L@K6zPoR6BXe
zJ7E13eze3^FnHD<utVMCIXcC<htxlA>@K8La_sqft+wMiE$XB@69j)dGUgYh0o(GT
z2bc<9H|;>cw@!hwx88W74<d|}tZ6;NlwG3ZU=g2_3+F8(M*f+|n+Gy&6}RK858kXT
z0L291mgmJl#XReB?K{sP1;>u)Uee_x>Mg3IUFXMUJkZi*1Sz`UaD}8@=3UZ(#g?(s
zT#xMlv0^Hyi>=dl@OvT(qy;!hm3`W97rdBk)zJ5+vnQ2>tcpxk8GK@{{~xWp;h;wJ
z^NIlq278Ik54PHQ4Zg|E#*YreKa^BF@b&3rMxA=!eQoR1$cPdRvb2Pwp;0{31lmX%
z<Mgb?hc8sg$}~A%2QQ`+0NM2E5gO517Y2dzc6Dj~5f&g`;~Jx(PUJ$3ZO36fRQO^z
zI{uq^Hi5pi?`A2Xm|OPuVxmTtx6Pbvu5Q0f)_Y`TC+yD{VlI&jVoYHwx`H+Zod}6i
ziF+sJDhqWSG;g$WKn1Qxw2V$5ivK25{xQEO%n|}^;Hp$;`tN*(CW;Ipz8?>h3@$EK
zd$&S1ZE|+3KN_j6t?ZFMu{esw8=RS<0bsLIg^_Dp&cqYO%HRa}$Q%b0p1b~vYEo4{
z)7F4m(+U?e0WcvCKeF6cU0xf1r}nB35HoBuiqq&can_xk{>g%c=(r%+-?T6p=F-XH
z`Q0d^%=u*=9o@2{k-EPk0o86^A6SxxY!W^C`;7Hpnv+XUQVfodbni*4W;`LaZrO@>
zDeYc7cM%22`OowEy~c68b`-Qe$<svi;a5dZD6*>4CZ7ka;H7hV<3gW+&7ZYZqzm30
z<>M((X0l|s&4L;tZ;s0ugoeQHRX}>GA07&!ZtjzQgspS2K?^=8joP8VF-lk1$izqZ
zO~~~wmrW+0@U`A|w|jhE=8t=D2Wvj4H5b8YYeVZEtKKD`Fl*C4n%><brE$`gQhwU8
zBOtH<==Actd9Zwlh{B^ef=(O%X}@wN*JmJ9s$`@dYR4;M+CnR;7LzvdB&U3KH;0gG
znqrIH%SgQ@>(E}6<i7j)(tjxAKK`s4bNq1icxDsuVbE)uy5X?K1j{>gJO=lsgx51q
zob!0}#GhI!_|Y7xk5sM~VNeQDIXLIusjfMwE;~Z5Al<9Jbhon3bEGpD)q&-#+Z3Y1
zk+76(xFgODU`lUH3+|0injT##apHFMc>TI-x!rn~x&IL($K^@*)j!8^kZC4yuv+VL
zOQKcNbXG05xid_be;l^AhPlES2iW;mKNx=CK|J4guqc|8mgi4%_RBUf7n3-&+u)N2
z*rbjz!Yldy(%3^$F7&XGQ*4t1Cu@xNGn7xDO0#TQXdaS(%DAI;iM1tgBrE?{da^br
z?b>E|y1nGSTNZy^!g%4x^^Z2!o;6N;+p}f0S^DhuVU!}DqLEUSVu9dm&Bis`NGN#(
z52~L=1Bf7jHQ_SeyO(&>49fdkxjD0lZ@>q`<>oJFNqDOJmw}zqjHIix*Fg^-qw~ws
zUsl|b+V%~h)_MSy1`}g4?4xJdyKS0Qg_b0gk(<%e7KSXGu^ifO%E`8WUv{hdoHI9p
zBS4c11(+pUf?`B!s3xHv&Wg?}xiag$(9*l#s#NZu$VT)D>i?-G!p;7_TR^NF-2c^B
zV;A1213mZcWz7t7nsGb>(=Sqyxd!lql-?F2rJ&I)7n#qmI5MD~x2pWKCv=-RQ_~fr
z=j)ejj-OwfU-$cc{^TfZ?9RgJT`c-cXi*P|_3Yj#C``rf{mb#`6E|&g@pQ6~lw8&6
z{z*qEb&esY2Onr7n3JCy)ZPpHRR{YDw~&7!^+1=$@StX<QtPvz=upGFBVxPDk2zj+
zQV>xHfv@AcRP0f^B+FubYfv1eC0tRKw8shp$HjRaSm}w_QgXzhnRygHaSmLqKavGw
zdizSD)HCuI%4$}PTXXYJWlH|&DG%A(d>b*3S-H;ism^4=nmqiNtmHN03fOBjwy&Sj
zLc}3L;w>t(iSDyMOY=n)rgeqFoD?#`|5^#qP451v&tzy&NcCiXOrQy0Xk(5mA>D8f
zcmuWmcbh%!K6lafaSHx13Xd@LkF%yOl8gQm=4#C;KjxZ^=-J(mb14e&P_CE|b;O?{
zoKYMitFu|wh{hR%$TlW%Bly)!0x6?z{lR02Z-ciQoL5BhGJ$e%IK@cWJDEBII0m!0
zRYRbN5aF)dt;(1Rk+jezv0q$XHy1{zAfBv4oyK(4!6zWp46y}%3x-&g<n%QJ3R!8C
zPS}if$ZHnIr4LGlHWf16o9)dWEh%(l1RrP`Q7^9!qIJ6+_gGR{Rn0U*4l^SJjD?eY
zfCU|0^Yiq!h1a}-jI?B_qBStZC$S+KSZjr1e#PeTz2%UnfiYV-QqjO(4F$KF)YqrM
zK3V`Ua{nM+7;s^vAwVKl9H%TV>azb9a^Q}l!k|sU46pl#B@7u<r*9V$Gp767olQ0U
zdXrE6y4C^?%I<zcmVs6T>myiSmu(GJPwXUZG>0s+-GtOw!%PJmP#amRb?j~a)~pge
z=ABnlo+ex1YP@e14613WMjV%3e`h<W>LZ^*!7ou;AY^82ntQ46=j$T$`EP`<6-#D(
z1$Un=F}-UP-O)ef48&nN+Q8^;bYfxUuP*=Y_a>BD*bMiJ(;ols0mMR4s~Rhy$4qI@
z*U#7e!)06>!!oLlq?EZglmeP$4-V=DjZ%?6|2oKZ3V93~uhY}nAeujAN$s@L%eikO
zaT^lOlH>2lX!Nq0^}a%fZ=lO3yK+gvD)*6T0%zfQjScq`CJm&IRf_m;O`|;50l6PT
zDHAy=A(KMnHN58~CONsPJLO4VX@SZXE7p4m&&nnq+NU^xIe6*|35EMc?=`5^wklKg
z?xP1RCl@bcTjh1}AHOZNBrD1!q{7(sTgDE=G8QyxVb?ML>A1ai`o^p@p9hs&p6%eS
zSup*UmZMEl?Go7WW&E*F?$rasn#2=Wb2|zAdM}@O+|582G2qN>6SfpK|G@Z%*`7dG
zYxV|L#SF?j@Dt;nR7bgOgUPi`>|ihcYP;---27Z8VctSZMM~$OO0=r1dd9#<z|EwB
zqk$OVlu?Ko)8<f^`fp9Ds1`z_7+*<_e3dd5Ky)+54n498VreauO!PpAcl8j;jt=Pa
zE4jW!xvB&R;cas|uqwveq=aV$yZkJP$JK>ECorE1AqjY~u+HR>^w)iIzC0|O$mb#>
zI6oYoN2BwXt4_fQ-M3G>KMA`v1F~MJRtDwIc>PFW&xyG!fE=V!F3wV9FBLTvpDpH!
zWU`W6X4*u8)NxI^+~v@Gc%l&0j>oM;gyug?X&C7Hx87t|*?^N~$6{MaF=%u3rbjMm
z125f1q$EER*48bkN_I6@CCFJW{b3@IZAS#Ka{V1+UwF?nm~?TA*$6Z1>}5UQXr9!{
z&uz07V(`|)RKFPx0+5F4P_-#`uJ`S##k1zL%5OilUA~+hOm?3h>E5)pG+9Ug!e0{<
z+MLAm{-Z#j8^4Z7w7qI|jTf`BbF&KmeL=r<l1UY%>jH(3s4tXi1shV5^`NcB?-}o`
z2iM4usH*d({<}_RhO~1r#wOod_MzC({xOSpiM{R?V^)`u*UTisJDums;I#q1Be~>B
znA;n0j819b&fT1ELX3H!mwMBE$O;XS)3auJr-j*h%{P(I#_mt~hEDqY0-&ZH<(eQ;
zZqD+32DA=^oFCJzGk68VZQdG@eI1v8Ubbd9tctU3%=BFf(5w6;TV#M`c@wYP2?kh@
zKG+_s+WBsLp_)4sc`bGmHpO%!4q~m6dKRATf5sWc+#v!~ix%0X#@aiAjPQd+`L`o|
zGPmyR_PZAHVq+pi)J1FF1*pKxA?Z4VsT0J8gmCtWM=x$_Z}F!S0yi=#X^Sjb`org1
z>^kor?Of%|!B)q?Zs*q9Rq^Zf_0=s?%{_Cw6t_Gc^M0V@cM+Nr3l9Zhq&O&{T)#e>
z708$dV`SD@dp)9Pz>~>vi-%kVSe3<Uk3^XcsH1p}v5m95PDv?VWV5XKge-W1JvXT_
zi?;!#+ykEpQ*04lCS+-u0~_vk*g`hZP<D;ExT3pgYzEp@KCpSre0Cui@85)ygre-R
z@E|qOrem?z!joOcwcro~G1-C*tn}S|<%|g4`p5@M{K0FBv82nApA5nR9gTeg>-CrX
zfBxNZ;xp|eRQ9q%LJuaswvnQRlbV`N<CfZ%!>5bDs$YDIzF#vUj*F~#^!s#Hf<1fN
z56D&@2-eSt#uj~Pn)S($Q>0(4R-Dc8(ZgPY7pL40cHHMoXG_7u=!iRT$iL@NB$}J?
z_*-vIjDdF25V!IWX01IOC|e$cxAbTKiZN&LdO0bR@*ghUHsw}y>2xOw7@Z)@-g4Bj
zv$HqBBr<sGU`-N{=Z&)^&s8$a7*UoeQrI%|Q0<E@$lXQ_d`kZWHPq`;k<jZ-4eVP#
zGD5gNh~c4Z)w!_WW4G^r0xUfLVP=Js_rKhi*5!9S;6j^vpvRF5FzUoZ_O>aPDom~*
z*j$CHwwVqXe{|SGoEqJn?ior183E<lFlX5dmk_>oE9{uvnK`|aC?zH~*dp%sMt7MT
z0Gapk`r=^q5xHS5^ic0B!X=lhJ^dKJzmcHTdiLy#Ez*E_w|aNDhC8<GLx?f*2&=nm
zNSTQ48908OSfGvR*-LJ-jFIZ-7>k{`>^MfSWK?4RS?yf3H0NwI5?^*#b0sV|!xMeA
z)q8)G*GMr!KfRl*=ESGU4Sm3ny>%(PlO&f!dsF7qk8A_?r>xV%le`iGM3v3-@Yqf{
z=HeyYwN!v1xRqK-pp~Ebvbtmc7svw4&crJ1oQTF4kC1heM5_oTlbVGM-l9J*yjSF0
zmi37Mql}IonbwHj_-G~#WPNp&1@@CaLCAYqW-`Rnz1k#=^IbcG+atSzn)wdwvH+8H
z#onJEhcRU1wd_^i|GK9~B-!bNfF_t)^&Ev733;iS7C!IW>J+GnWB_<@%!qFlyHJV*
z7F>bzo2Sy81{uTwwne_*jm}gJh-8<3-ORX<n8WF-p{S)>8}+~CM3Yo?G0daTk9KHe
zl>F>}Lb49aEe5`lura~#EUM%jW@_xG&_H`v{eO(TLzJk|wyc@9ZQHhO+qP}nHh0>#
zbEj?FwpsgDjcUAiPvea7cmK{>of)xWexY?2RgP!XTbjoJ0=|6F(8~Y`GTlTbqs0<^
zP*OyV-qvlM<=Np?bpEAMY>LfKl(`dsQ5dZlf>CwLNQ@qI7mC{BbCPZdPY0L6KxOUO
zOb^dO2f7ILggnJaC-<UMZGyZIR!nCXakQTxD2;g662h{u3Bvap#>xLVR@~_55;G8}
zj0pJ6`u(ewSsybR7#+nmZ$i|BB|4{D3{4cnAR~>};LqcF2uD<Uq$@2(B^c6eJ}zMz
zsbaHC(+cqmD1_<zN;}|4ef6el0K~z_L?Z<ELBlttsfc=41yMiy<q2!)YL#<H8ZTM8
z#JQGUAy7uX48vr~W`WUUD94wIX#w0U2k13IS6WGD-kyeY`d%6Wt9Zc!KYNUXtdpn7
zINL@fNKdIrT8QtCbOTwY%IQZ&<;=3|-$?;%L+7C88)PU-W==1o)>4J_)|ESyekNs~
zkU29mEJelVQ(!m)wv`vkgPPVdXN2B<wodM}CTSW)2c2kO7}A<E&U#>y>>l!#-62dl
z2~M&eBCRzk#S;{-PJ!Yq8;Zq0y&=7b;?FhBm6o)<2NvcK){(#rR2^%_E|f~B#gWV$
zBy_1!vgD;Y3@KMiwjV46V_a@ktn+EOYM^sa`7k`X`=LZ5Jj{{gP_3b~^}Qc20%GkH
zU^H~xo%U$&GFa;-AS+Qss>)9ZfIKuN))i=HAr3cpF|uEwjv8Pf()CbPe1Wb=fHL2|
z87gld0zLuR!fE;@ka-z)wDZlRA~7{sd5{w8T{R<Ihg^!9w_>+efu>1v$7GY5t|XEH
zTIUBcVs2@jq1MUsD$Okwq5vIB4Mzzyv)P0JxEcT+T(8J_Yw*{ma4ZZlJfmwcuP?@8
zf98dvS%B@(mdz%JxQ5aypk818NpvN?bA%_4n4I1!Hp{JqxL>bG6r*}|GE8HDJi)@B
zMXPa4y7}6gy%Q>oomC64Eo@`t?4#}`%bf)3m3i+gH=A}|b<ZujBCHr+7nB_&e3j(;
zEXq$3<VHn`1ngdVY;S~slyT5*W2>D@0~^LSY)r#8Ux^i~n*UZ^opDqt27{IX$;znX
z-wq%#V3b%X)M52n5GQ+@!g9hB<0)DLS~VG_ft>D63My7`44K3Jp&c#;pR!Zp+{!S4
z_eoAS3Nr)kN6vmtj-3?}#uf^TQdotg#Fksd2BDKxCm}?eej}>M2Ly#&^*pO2|Dq8E
zX^o|ae#F2Tovkbqj=>9Jw4t`K?qK9#f5R{mO0IgSQW$3IIFg1u#G=H$eC>t**HRmp
z-8T!}Y*Z550oV$ffprr8hZ4h5y;@)3bWyenNq&Bw?lWIA@84}NVq1U$>cII=Y=ADM
zW^H4IWF1Ap=eO_0G`(h(vt_E+?%`vYQkrZ#))fnBN$?)~8Q=V(LMiXtjC{?BCeF4>
zPUy0i>6)1D7FL1n+gVPW)gmWwh?@uF$KCMP@L!rJ;whhY!}`TreY`DTn3;B(w~)xc
z8W}%7&L2;AvuV~mg#{r*kY=&t{-2+ZOIz=pcq1B71j<|EOSnp6h@R)((mH~2&Xbg@
zqeQ(Nx=zj&>-A~c&NN5_SXI1BnATu|mw2?gwq9`0X&H!HS>>D$8Y~}0rRb{$TFmw*
z+xl!k#8kP24d&Cy3k<hEr(A39q!Twnm0i;lafNr^bJFAP9%rQJJmNjAxqMd(2zx6v
z_lD_dj>PlNhPV0bG|0BJ&4ak*nefm>Nedm)ABR2Voxyr6&W$1lCsv&E6T&BZ!A(NK
zY(?&NO*adR28cDb)SPJ9(7do<;tl=cG4MbuvIBP$QJ0k!n8Qq3gG4%pskSk)t#ULc
z5~kfe!Mo~P;L^_3XSlQReGBLCK5xiyXziNsb@(5olZA`=i}W*@U@?rPULz8rmc@TF
z<Z90v6R)R$LY_~V-J90RF;@y3XHj}~g(h3NbaYlhb9LJFd}cB+b}70>;A`X)7J7NF
zGf`a;#n(O<NVsRi=gjGe#U&6@I&FR8@E(V#(I*maay|hRbq<vp?PDu8Q!HsXIajni
zBRTgxt^&uT=he-e>o}ezeQwd%h?>0NfNk!m9G`K5VLx)!+ybYyt1F;%wqB<-u2wqY
zLEGBdhiTs#H55=L*Qn6a-(^nqXkIRP`UfetTE~s+9kwBnu}47B_I=tTs2s@Ax?J%L
z3I|@h;|74(^+uvpY3M{x@9=QkF1h;!CH?JuVBp&Go1DR$h6wWob%kOtW;TTQsCKCO
zxoe$H6uRSp9lLvx`?eo(_P`0h@q+Owb|-D`gNb_Uk2Md>WZW+K`bEN?JYmR`5GR(f
z$V`2m-KVx|9s!^c7mgV<AjZH`NzsNeouk^$O}o1o)@wcikZ%@F1QQ%2%UOjbZbFUf
zCV)vSB(ip+?A2beA7ww7u#q+yON(;qC!C0SP>d%;5cLDsEjqrb;0qWC+i-zEoo^u=
zTKf-JdrkuL;q}|z*Xn)&&~cqs7~3(V+AE=`4Np6S=XsoW*^L{tPXW=Wjw+4o71AK6
z)Eg}H?m8yv-ToaFkWeib7&L>3goc6P$?GtoTdA|~QQTRd{wp9E6G@dZLt++05n6`M
z>4>8ZK*K@JXHCS_icdfULP%ur1gX0jp{FWc<YgmkD`xO^Cmh2bA3Po~AUKg~u8Dqh
zo0nUPSl2RcLB9|b4;UEPy%jox4m8GrJs*KxGuz(=?OQ-Rk7FSpS?5nXw9Suqy}x{$
z+TUL<dj^yC;{f#0>5~r|zSR2U)cP2kF6FHcla*sN7E(jp$D6;MZ@F{Y-_ilubqTP3
zktpY#eIk|d_el*0s6CIrwmQ<H|8e5{-?1SkCdU8LBGVpEI^=-sIaNOcPZCAgLsQ&`
z*vwQ#^v%xec73YwuB2BEKcD>i0nlmQOi5kA$!QdY1i}-jmfHSaKEL^#JU%?0WG|kn
zG8o9dITQ!)E_kQ^S0rz93Ya2%#^>kO=exYY@uTx5tVR4V`uN%}?(HQC;RAYy8L(JI
zo^Uy9x}D$eOS`K74$(;V>cM%rfU4I^KJjvNe<`r-(YxS{1m%!sp5yoXb}Wj{*c>I-
zJ7M^9`F(iIPw!lq12+M3@$>dh1RYYqfRU!^iS~T!@5-<D!&t@@Mgt&Rn%m?CglRF7
z$aVThNOK}Va~i*hwSi%1aIS5to2a+u_wiMK`)aCg7rb|#F_@vu<_RS1$BGE`;LAv-
zet?`#GuwqWt22TQ&z!z7HiW^SNPB?Fd%jYleJ|d#X{z`+6F-m^cbBP8O-=`S7<7+~
z!h}|}U$;Y;BJ<%hi&XBEurBh+4K~$Uu-_96wH~+x8Yr3B7=()a2(|DQmE=LBR4Vg#
z2-KfkyQ8D9zg#2I*^2%B+s(^W%PkxAyXRDQBA$|3jDu1heVv~eADif%A#oPPGb(Q^
z)&;|U(jw6n?F40$=?|Xd!X!CYE8%Xog^tL~?6_5RBf@AS*}6<a0g?FvH7t*&zp!3i
z&M@aP>_hLQRHD%2Nz+4_6eH(fC7*Y+=PYNYEELTQetu`({kJr4=Dk$Fu6|P^>v5<C
zjc)3EBhnz1?PN(cs=ndFKK?6L8Ij{dA^9w|dv-<&Sw~0LS=R#z{wdHP)N(YJ*(Qi*
z0MX!y573zpTU9!dA{UJU1NwMrAsY!t!$L5$z!SpC#@TBie2rD42|7)TbJ{le;%e)#
z^%e>^pOAS+R%wFKxj&iakOnbA7mFcMMgZi+r=UN*#g@Pn5ONkjFI_i3N5lGWM)x7V
zvgtop`9)p+?!PHK#ex~vS`glKM9Lk)8t!7EaxhHmfg!^gts{R)*qlz*G7Y-r(^Rpp
zb8fbEpi>OpCw_zr&tv-)@b7k<1QHwqIpFZB#F!u$7(?;0wwYM!8{*t=1?Mzo>T>W;
z{aM9mAE>d0uTXYJdU@V$nP;?)ZulPZCVidw#hE=eA<p&Yu_KaH4fY5mDC@m1lTo{~
zH;(s49Z)C&)%U6LV*#$yiUSdnfXv~9AUI8PkDg+;>cw_79@sN<m$i)>S;yOIN&iBi
z>L^!=6%yiCV37V!wss+2Z5OcgpBUJ&;DNO6Z*jh~aPITsJaENy3w*>4Gtbor&*yy9
z=ubRrF)a(lL>((rwg5N*X08lI=yd2!{MKc9X(2f&l?`=kCb?6#2~6>VsJClx3y5BB
zJ~(P_T7Y)7EY|}uOd#NzfaU~h$tjDe%{j?vETi0NimI&-G2pd9;iTTOVWXkNlEVG$
zV(rvKhJ~?G6R!ocOtoY`dvql)ieIizEb$wS;j#|sTNFEsrkYXvbscxd<7dlF9KsF6
zw<{eP87eiXG55qY{&`HJQ;BcEu2k<D>XPS$K3UayCA)H50$)CUTX|JvKR4vpgXMk8
zVZkd3H%P&D{hHX3LnNTP3k|B#Gw`l5ULUk+vw@w`nSqKQm<o0hwnsKEuivfmP|eqS
zbYyp)A$(4()bE*jtZYUW@;56XXMH>G?V1zvVP~Qpsv&yJ)OuBf+N|_0w3^C0@Y@kx
zW5^_AsgUMrj6tECbwx79?yas{!pq4^7%|Cu?2S-H+%794FD_~{voK)|k?0v!BleF?
zMe9CG>>1vxo*}0$m_QpBrgmb@yat4KeVfb2WK%mCS|iv{Ct(YLF|x{|o%z{lGfe;T
z>>c@=!_VzjYN#gw8=SB~(&i<Bc<s6ZpeH_paEsgH;Z)@^X;DHlt)i2ZET!{eUMi90
zC_%AXcl<ikPg!fAJIlhb-UXN-k;rca?BG=UmHfoDi_KZG+sMo2Dj}=@ht7e)C$_bU
zZ)&V~cK!HS1fjDB{+pR`&kz}#6UogAUIJg9+qfqt9D=lWORk?U-})l$icgE&5#T&~
zIi%*7FWYv?Uf+VZ6vq;HIf!9L;jWT>)ZZ-Y*oA4gM-c1v=|b@5A&wVLoFv?LMqv|s
znVt;$)wiml73D)6P_E7-fXqz)R9Uv!EqXaXI_q^F2PFNF%jV1|J_>{z4`~YzsybUe
z)Q;hAi&6S5BX$|xr=UduwXM%kaDPidci{LMDa_xyB@YgItzFY0CafwN=|<J~9^bu2
z{R^<Hg@;!cgskGlhE9(4^;=;TDG5C$Hfws0ZHEwtOJ8ED%=r<zDZhE_ge;tve7vJP
znYL4Pt+|(i+Qh(uL4{K|#hG@)ECJEz<X5S|A;WE%Kh!-bRJ*XaOeqO?Ejt{8oAJ`X
z_PM~njM{Qa_Y5e#98-v4ybDmEX9if%GMEx2rA^eb=F|-Y`DIwtQ+GiWAF&L#VTXKE
z2+mrL5K{;pIxK&yf<U`%@029$7vqFzQ!J|}?H99PY_H$RB}`#k1v~Yhz={b=@J0!+
zXjbd@@AsX3MVmUoH4V987h~p50?{Ji`_IW)|LjcO#%3PckzfIQ=J??R*SASswi|cb
zmvOeS{D~L$U`K(C(4CQt3tUTFaMK$v9-!S8dDT&KtVtROqGD4RjxE1Q>kXt5Ma|>z
zzJ2#)+RoFTvK%Mmgilapn0gXV{r2=oPusCiZ!Su*Gv4mg&jXVqD?=W=LIVLU8h`sE
za-sXhF$td>GYdNt{n)7v;f1BO$e$zN$ru`y-FcF(vVb9)4%>`+wjIC{5SF8W1Kd16
z0YmKK0-SShTKzGyS;9VkY4as9x$~?wOHhc-PSl}&0G7l>jx<V8L2Q4&@9%L{u5T}{
z4EJeC@(f())Hd{*7{jU00WyY>k_k5we(|eUH;Xi++ta%}$il!R<!_btO%P}soB=_#
zHJjTS*LXVm3;5%pBK4ofo`0Si1A)Df6%-E-{r`{MX5sknu-n7h((y;_u)V8wclZ{u
z6pHi(3KHtFs{r2Fvb$*B5M-I-hFoKIaPOamcqyjolWrZG62NwwjWUV(6nrLWSAH&y
zpH3Woczn97ZgL*La;|l|+y&rCKf%Rf)?DBO)nPZ!HJ{fMC5<jQ=K_>1rGMvJ%9=QN
zE$^0q%R%>^ju4D~Pan2&`|+fIp>qlOqHo*Rtd!x=>yS;Ioj%r-)#JRl!V`fIMDNOW
z_5C@=h(^k)?wy%5!h<XtU>sSBiifCkBRdNf?~*oQNG7DqHH<@s@lr1-#{(lbctA-(
z_nAux^3ivcfVSR9jWGp^DFuOMLdHPkL5+Gk(eD4O(SJV<*>y-W<ZQl9^C-w1FYI7y
z=4B@b5i3jcaDh?8<LpMK>^tYlAll^2dVDZ-dVBJIIksm4n}#xPx3vk|h{FnbF9)Iu
zZr;rLyX?a2kBVrwyb1_NODDhufJvSio4249U!hv!k{W47!>szy&W8GzSrSnE{=^i)
zY59FGL<dm#V8A0a6ggfHoYO!*Jv^D1g0R5eJ|eS#Dk0}r8fVrj+`43%Xv0k>CyPcv
zEh)#QEbM<U%gPm!P8&=&P-tlx#&jhfbL~+)ymu(H#1e%rluod|OrUKZcc_<7pw_f-
z^poN#XDMTT30^&)O5*-A<rMs?MWmn`y{koJ0x6TjiuZYR<r828HGmV3?E5BfZlCVZ
zFRFuYf3iD!mkmu`n?9|V7C;B{o>-7tfok%IX}C#&DAmX|lNq=ml%hu^tB<Cz4M}E_
zOF~hnGH$M<Kni2X4)od!D(`>zX%Jf41rFK;ZUbUx;^A~zn!;@h{9T3Eh}&3*r7-Gm
z7gEI~t99^1)Mk@=+IZ|%_kH6Xxf#4@=eq8d+y&IIcCP7RG+l%h$7+9!SR@|MD`8GP
zjuUDubm5|@@KWDuebuQV_P(aqY`jQaiBWGBRFRD49;{<#f^tIRc)&4bOJfSb5dLD^
z&VW#4=|QE9s9b%%FQ&18*iq<PrL>P>C%F$`LoS2gjVLMWSTvZ%+H7*Ikc2*)iR{(F
zmkRfp?8?S}u8)pU{YoRjXlbFsrHFg*V=2G}`h*ZvL<H=Kfe4}lUy-+DqAST*@m#fN
z<)RB_-cvMw5OkoZsIH5lSSg^A9}uSzR3XNmsysK08;kj$iZwOyEO<B;!4E19HHdgR
zgMgc?@b=4soWjIfw}m=b98AFD;|~_j3rTv!Yx`vkNzEughFaw#z$N~{QN7+3f3J2W
zTyYymLm-A#(DP?8V#IS1$?3M;u0OjL-NK24L(0$2$tO14JYBgeKpR-683N*nilEqn
z#F3h%=?CFz6%VDIA_B_h)D;i_E38?E1SH5ee_{c$EF3c%C{G!!7B*&HeY8LTblP<z
zXCd(%4?wnPZ@hLG&omJFb7Nd1Y{GpC9pOQau+0~7hERme9608DKOoAUgxgV^JLqa*
zbK(UEYfj1e$X&Qk0p{oi6w7>TDn_#QUY;&^$)&^g*cT8zo#%|NQ(ZP^IWc+n6XmVI
z&F0({5G1+dfR7-vW%r6qa98@;L4(Q3M05i9yhhazM3;fP&(~j2qC&X^RK%%s==*ib
zry&hV^fLZQ#E?y@As}WB-T}|F5kcv$9N6(99q!(~%C&bLLl5w$F@6giAvDFIABe~r
z_r^>P9U<Tf`Be}=VT>=(4uN8Gq_h)#dRGtw9jA0C`pqWI7!Mz{F?|ahzv7%w?r_n}
z&7;@xZ`i$7_^cZz+TH!S_DxVclvF?t#2SnFVCU7C;c`)+nWIDchfdZks<q-K%{Gt~
zUDv9w!Xp6yIZ)SIXy2c^Edq^aOH!)r4t;@Pv)xX3Dmkb-J1kT=IPa>jfVltn#FD?x
z`3y1$QXbs<CA@o=_nJrEjsZ1V8~XyF%*s09Y0g3$l?QUE=WUxRo)Z4QUvhD;_$(~Q
zS*8LEqr`?TQw^uw!q5X0){}35LmGv$DM@^cB{-RNs*tid^#p(#XScv1xh@Mz6U(EE
zRsdD7gG9Cec-HKCqm;#`|4{aIu`#aET?0iiea`tRq);eaiYqQ%%e@qF*ru7|kwdIk
z-2@?n(9X$*`sZiMlc?4i&9(7Gtk1jy5$!-i{P5y8;gL1meBydg3qjy38aMtCLW@BY
zR3%-VB|q%An_tUbC|uh>Q!lNM_1&JJ*S}}3Z@pO)aiAx+jib$F<EEf}eViGbuOE(;
z7Gh52hHm5e=Q~X~CS~#K=uVi;XYTe;8cyTffw>CP`q!R5EPQl?pBog~WV72)Co1~x
zdb!qMt5?WjeuBC&??Mo*^)8*tb!38BOiZ)fnSXZrYe2!a-vxNawk(!|Z7z0O&U#Ch
z5<q=u$O6XqFJrfkK{AvC*0;XZDC90M2XMc8`-%7DFyMR|8K|09IkA>N_ltT(o?1h-
zOZ{#VnGds4<5VmxYBVKUMQFiReC;g9$wapl8F2Fca;Z5~2~NvorcPDaBUIV-=~8zn
zeAMo=2kyK*dm(p_qB|&4#14`ezln>xT}!9OY!a>u3xhGTLMw#9G{$r_%#29E$oe;o
z9NwMp^7sB8Z|(}rS^q}rnuZHyTD0vx)F=*Wh>&P!-BN-B9Q%kvhuQ$Sg&~R8mtRD;
z>}$IVh{^@fC@g+3w}J)cp$R=YHLlaS9RX&Moe(Nj*YNDjwkCF5giA1~vK{CaEA7O#
zJ@NIS+8@sEIIDYI=5OzNyT55Eu&Q;e4NuB&XkbSx*(rhsB3vS1n{-HO2NGh8mV<~p
z`<Jz(wy{we_K3<NdF_+Tsp8g5Owq<}ZHY8z0R7`tZTW$RCsr>hsiqv{3*1^73P&jv
z6l5db)b;hGzJ`$s%A7({j0+f_2mL`#Ey_LWy8U7*b~)X->RQk082Nw)4s!F2jhIW&
z?+PoS*;J<AT3rrM;2>+!0vv01ss<k{tmpjAi8hq7hsua?P9}^=9yRi?0WBG7YCD24
zmE99N!YijbVu9iSz*~OYP8PP5Z$H29x2<)xlc^47Yaci<=U;T9{1d&mOP?Mio(gMj
zN_2>HJ?x?2rWm)KNeSA-DN`x$^COwWnVwI!9;Ucz7r(-p_uV4zn4IZ`47cZ~NX>vz
z7=#JUBv2u4apoNDhh~b0?x{C2`W0U;GsU$Ou_gek^WA_1=QqeA&6a@aVvHwkditGC
zvkt?LW}`=JWQSeVf`=ukh*I!3WR7#rF<9m3iREir?MB_KyBg0zITT53pgzuhP~zpp
zVnXcYg>G}qzvP#7llt@*P$a-K?mtX`|3d=!zb8Ou4n~guE*pk38Hdwx&8M%(&(OF7
zZ@?hozG+Orj@=8&NfWcixhCl5Ja!wNBoV1p(do@2iDvQIEcJ&X%Dz#}ZkN5o*N4O3
z)8p0X3Qfd?>g;;5O?Chn@|OxSc*j<dazA#X_iO9tfu!W+j`Q!RlqHN0&=>ChCNDo0
zs$p0VVo4r;u&U<|WPZ5$alXS3Xs@bY+G+^T#bivvFI~d6GlM+tKg*FBB%6>TpnTy`
zzwhf^@PKirr@fujuqEwZz6`$_K*J?0$}sQ&LZ8VVN*Il;QBEqoPyP*1C3nI<9o8Ok
zlA9I^&b26}mzNkt9>l5eMzNRGM<>^CfA1w!;y})|xQ_ZwWQhELOM4$>OL&uz|C#%$
zBmkVuP6^|*DBM23c;^Rzs5al{=S_foNiDMc%SY%ts4Ik3E_3R|mlM6O9*mWQ&^W6Q
z=(8Jh*FsduE1Z7*D<s9Uq7V97xQKUK+Xm&iy*B4w(|n~h&+LeGQCIUQy<F|fh$x3U
z^ScAw3fD0^RW77pWO*XTl^1@m`A>_gAP6qJMbDv$acD#0V$EH~FfW%muNHEoekAgY
zCkt=vLrO((QTo1mY0e-0I1Zk8vhG|OTtIjC>y+lfy{m3@i?<`!itE5v^#Yi7@|^5K
z4lk<vPu{~5=$Xq5DKKUt`dD+75LsSgAG&K#qXi;GFF6@vp8&x*uoZWUF=u*-#beE<
z&kvM1V5kpGPWR)a7BsT+(ptDo9QYGb@@n-e=-MLjGqEhih{F59iwLqrkeaWM;>tO-
zDEcOm-P<#-%SEp|stpf1B4_?|o~CF4PzA#o2cE%sV4UvSBgtX?gbEgaEr&UR>h&Lq
z;@y&iQj#1VC#9jl0zCf2T96fe4uj4ZIhCZ<e{BOY1;zE-99*3a<LzuNK}CEiFyu*7
z@UtVlu5N-o5l(ZqyXBg;L_`Oml05s9S}zL1#K=jlxGx?hEQeIgOyaA|n1cn!%dXl3
znfY=fnx$t@`u3292|aONFwZ`&tLt#7+_JHx$vkLI>CFy=xbGyb?UrpmNY>OR9^&Rp
zogU&YMqu$XbcxZIUTjhuO=n|{(}^DqQC0X|D*UeaZqlwMB~{>}hVdoU0|XPADfLnC
zj!H+~s2jNok<A^MNX1_?{51!7D#1QPGgK;CJGLOR;+te$Wq!_fN}*$xNQ@e5HL2Zk
zE5#RBbk19s)m?Bo6TIhZEyH`{^IuCf9))<0ho-d1uq#)f`D-8Zrz^Ir1KF0&WbkGy
z8qa2m6MR1F`e7+TV12&d?{}YIH>;yHZw7lKSt#2tw>2=(rhEK%8o_OQe-pu><*xsB
z`TPC8l-2c70#5#<mQ74UU8_WG&VDQgL!ZW~ZN`6a18|h&?50Q<^bTk*2~pM4+ZtlO
zb@FQ98n~mcpT+Y)B`SxkMbffKWnesrY!tw=;xG-dNqv6r1&65a{>iB*O!POUXlMHk
zd2Xiqox=agE*-4wvxnbp4VRZtJ?(*KAQiLKi31mEb`w=4?MGybQf`GMO^ks}EKdS*
zSe>Wfw#Qxqs$Ib1FhmigdaDYwGcQGSv%><?00|5`^Ty*Z`0JvEVJUMFA`OKSb~j^&
zrYf;5wg2<Fn%{oIUHviEOPwQL5?kMo5D#Ynnk<+~10=Fn)sJC*Xkr%}<sG6`b0dfd
zonGU}Yx81h+`c5t`=adInT;d`39jyzjpU3oOc79lNHid57Z7J)U|a>~3{1C`jP0;*
z*(S!uFPxkTSR%BeZ_ugg8m;xz3f-iu+%*qW^RB9c-C3nrnW!gV!H=uGqX9Bg9xUOf
zRU$t{FcmiK=sX8J;5JqD8EPZ&Y~27;c$8eKx=LSnkil$RYc^8Z(T#OxZNShXeB)xA
z2ndSnKK35yt@i4JTKPCg@hK&w30|rA`Mc=~v<&L-M%`e81hq@|!HLx9hsc~UP{lyY
z?@rj)J#F|?i1~oqPQ}!B3i^%9EKU-t73s;Asbm%rLUqM_+y>{cb<BA_p2A+RJc%Q2
zYK-mNTsF&Fv{FAr(Bl!kf|6I|XFlvApMIj~gsV1N_7*vcT+$NG<uT#XW{j16TN~?d
zbbnJ4o8tkJBM!XTGehZOgGU|mY!<L*jaMdChy!G)mg6q$Xu@S>PF9D_m_bw%r>?2X
zz|sFLbo#(^x(6wzu~4BI#gbzX#iPYi1flfugeY{ut|e}^^S3A7GyaraeMg9WirT#}
z`{#o0tLTeb^a0c4=Y!+>qX#f?vj}{z)dSnJ+^dVLE7IY26W@Qcm)9pNkg&2KC(1hU
z*RheTF+6d0Yr4PwAaNLVcv@ob75bbPCy<)4J_j*}0J?QtustxwZnwz=sxbu*a;=Tt
z-?iEz#IS^BP1>QpOg9b+u7*!}Vb43K2Yuoab7Kgu1;ruTiJj7z*5mlzUpCVcT}Gfq
zrodY5m&~=Fy<uV&ft4$^q|t3gQ@eOB>M%+lFYubsL>sl*iBj;~A>FX`zJl9d6k}Rf
zlo7cHqB_W<Yw5*;r6fA5p_&*@Hkd7ir~EKO#{Hgvu|Aj=9ad)Y1PxhrPhTFZ(6fqp
z`lwPzf$>p<jQtMv7y7zXT-AH~l?9<B`(GF}M+kW@&(K0^Svd(7+Rm=!<fJw4sbyT1
zm;8w9JKW{ck|vy`UE(tMs1%X_yD<HgPB`j}FN!K+mzZM;xf&ZqRSX)k_PW{1!C*v0
zmBd-bvROb&LSHhuuLFU~*~?$dIX$xEVItIgdoUq?RP&)A$yy-ss(gOIE|7c4P3kW1
zHGA4}L}*yqofcl%dlH2?xf^ado)kK5q0Kna^yqv5a{Zn-L{ie#_X5FfBm9gjE+a-u
zYE(o{`8|4)G;XWEkt24TakVu+&8>{&Izs)R*64e95P>R{802P3jA~SAI+|u1`)&0|
z0j_#aMAO-$^khX1N=)ofTE3x9PV$t3yyhkLBsR-LU5g!Wsx8;Cn%%&Y>UV(?@AWCB
zZm>6J@T~yFbAluFbSoiZ{hsIzcD(K4CWX(QvRyDOENKlvrq-lJdY&%Vz9OYYpIB2|
zDYdDru>_vA6);5=#fDn%rdslQq=8=dRb_f}f)g2hnKjrJTkLYZLSxvs>qL}{CJ|Ma
z7<`AYr?pB0{%v(rp&Mnx<;Vywo$bI;t&=9utD)VDLZy-r;nFEkzP$`N{2L1$H{S^2
zdb3-vSk;?$fO1#bMp`d`LhgAag;3CFq_g{2&7DFjUUbte+lD@&VI)sfu|9oZ1=KbP
zMsFz_E3$ksAoZgQ6dQWO8hef4=8bxX-m>ig7#G;$=#FZ4`@uyizE%r_@nMP=G2}%O
zV_o_5W_@EE3$yOG%P+YRAv=ua))yVrCt5EeeYzi78eEE-&|Tctu8^I%$rCCfE9?8t
z&79E!kpL?fa>g39l8f34k#G0x1=te^W(uaKct|=_P#Lr%rxkIw4RMpAfi3tcB}GNo
zb%UG1&IHpA;ycIuviUYvJQd7_&t2zodTUj&)7TX~q`^-v6vn9d#mJ>k;M9P;qvN>h
zHJOD`b{n%0jqWsys7Kl+6^ZS6Cb=OW5I5^$qx^6O&8PE{H&VGS01K5@u&ih08`t)>
zq77JF9n)i_Zs3Kmtm#KexhS(*-YVYgxN3Y`-YWc>X=#LOFUZId-$&i%sN?bEM`t<`
zt3Gj*#Sjd-VWgV%FeVdOMQU(;r))6mcWzAVDv>MR*K}!D&HUO5r3ZZ)E|=(QMu<7~
zd3{cpEo-NvsyY_m61%tZLhssW$`-Cemk`Uma%Vf%@5f^e8-_6(z=J!h0q&BJ)97_%
z`=TO9jO9-G;GRGe#W`Inhmk=2pkGp55pp)Ey`pnD_VG_{aeIlEjJx_Vbo#OndZ$X}
z2*aZ8BHFR#;%MTs^aXNP2tquN1@E)!i}M?IOQLNR7++K}FX4Et#zQk>n{MyqD4sfN
zNS|G;)p8<sW8f)vMe`}~PjfhhbT5GQt~UQ%fxdN|Iu+zyQ)wvY*`z%-%KGZlgv9al
z!)ATDG|6<IU0Hue%Z@xY;a|OB22jK;YX+J_EdBfePt&U)^1C;l_>!-MN#}FKcD<Vw
zf&#7?aEmO$m@QWF!GSh?%Xd+&)4U&zkdYm&fyB_9xbvc{fXLWOGuR!Zvs3e;JPwdN
z6&GKT<dZXQ6$+_ZP86S7gsp<ptzP5O<sNB`x{3;78Q7C1IXN20z!jJlz%oaORTnLY
zCYo@Ecp7kdm72%8o&5;+o9Eg~cvgYYOj1gV*=!@b0MB4H=I6&~Loiz(--hZ5mh>DV
z#kXo77f7f&w?9W>`P*g>$5c^l9Dq!%!t?HdbaJ#6_#&k0^H3Xzpr4&xS25tWf8ZT9
z8LM20E><b1Q9~uov@{&x-{|m@9?-|KtW(}$lMp$m?YAnTOHHDVbBB!e8E>=cy0O9N
zWnVkCHhbqK@!@x&Fp%=1UkB|ZmJ;iVgC+FtO<odU7S9#14VEmAa;Pea-Nhws|DooR
zxE@{kB(;y%Xt;YW4UV=9eJwqnJ?-fTMW{*ct_57LmglMcs_RY^q8dvMH>dCTo?j{v
z{9Vg0W=$&ytO`}LI1A7<H?eJ9&@H{9N&2GyynEA$tI2v2etP?LwK0bWq4lTu6;5AT
zOb$xysBRdoh3?uM-N6N5OFyjO*GM_vh<=>s66DocY9jXWsmuCJ624Ej%X^6pf1WYF
zl()Px#d$M2)3kL(XXEAwT3DW_*3KKp{fWrSeSpUD$`cgm5VJVEboBD#osnR_$Rcf%
zOkgZ)&(>Fa)8jOwv>Ww7ld(`%JG+C`@eB64{k(Ef3r*JeypH&L^PM`J7hbiRW~B(T
zcWm*OW;#Kzoh!O~X2CD9$m(+_`e;9{T&qXF1p>B@&&N>K?s{2(I%dbU6VUwx<{+DU
z`^Lg7x<i@1r_y-x1okYy@s{D?U8)b&%^^0MD)_ev;PWE(0~=h>!<?-D;kyoasJ8=B
z`S%ha7?rOavS5a;{mZq0vzzTDB9FP1t`9}@SEXx4oWtNK?rMR3N5VM&`w8g_FkNn!
z0OS`?k2&??Kk4Rw68V4WX2$=L^xC668Hd{eyZx%Z1m7%n3qEI2{JCLE(1q!NZi0wW
z-&7yWgZBQdfF!D@q`GZoOa{EPMTh!t=S&n&sd1~<*6H)<?&Ezwo!z*N;E%GPeSV)E
z6raRfU((#!Hd+eOPWkm?{58G2k`??cHIJH<@gC%zdx-w)mxT%l6ez6}FF#ln(>Jsq
z{Oklj;Ww<G>QCBd$kqk4=$L<c;{PaTUl-f}CXj4#1^Rc-pUYEzehHnC(OG!rO+D<y
z>VLA_<82izB)?VYo8TWG#P)Yvs!0LtXR{b$2Mj~#pY+}Rfeq>VdfZ&)_kGVsn!D!D
zRIyXDC{x<<TT(w2Ng*t~=I#|Q#F^@%Of(R6dz`yKLTHX#x-<2z=jy+rYh1bQV1>l4
zS*dPP92^1&VF9$Z;yFUBXqkS^WoN7XKrxja67!oZGT8ls`qr-d%EJHIktkf<Z40m4
z{98xyKWae$sfZ2E&o^S_Mp6~B3J3!|*~XelusEUGUP)4r%YANQRXSEHfQlAES{$*s
zQInnnyQZV1-nESkqbhbTYvXpyk09K#W==@#pI(@oh0a%EQw&|GLAvqxe-Hfb6|9rL
zYKwa;@1Ll7+$@48FwG99(%u-}rj1>ah~ehCWrXHl$ZRaz!14CBzJgEw?qt-v_=ERZ
z>DNO8dB-@j4xMD7b1l)&VVF#eh~v@H3WB|tF0z+b6RHngI}-HSu)b{pNlOQP%IaSM
zC*DpWOCk>FB%x$J$@li4ZLaD`nvzvE^D{qO35cB#%Dxm)c{^k7*2vxhxx>O!+hUTn
zwJTZd%8@o&(rqgB9YXBoLlispZDFh`P7m}poVdKzP=BgO$8I8q$jsFwVXx~pCZtVM
zufGHBYnXwCQNJudsl<_&dkPlRE215`+Pj8+a18p1l85zuTdvE%3mC>?-DbH|IOTkm
z)ILiVE%G^>_Y%4eFgz~3L8rKuNo6)X{1w)7)Nm4*4bbI&aH8_;+FSZDL9XumZf7?;
zVC=Hm4_>eC7+~q^C-xk`9OuwavI1<rj4SV|_PZv=B!V0dBbUReK@#3hdlZogHHwr^
z#)+-iT2gYp2=D!f$LCB4$mi^Ss&uKd%(dnt>s>_GAiu5=dCB5kWaP^JH6e~GM$>7m
zzZ9-}IVq(~2UrO>w*a(M<?=Zs77f6J8BrpMD%!fe-)Idv%-M~IpIU~lBmK^GMKk(w
z%8Ejx^+5##S5hJsJZqfQZQ?FwD!)R%dQ#B#$SdQU<@blF4(-_!`&I6Q(%1=oQCW29
z29?u>`@>e_FZ9iB<Eh;BoELpR4jF5~OdL0$)ME0Vcor8f1sS4?yP0+5`m0kpR=y;a
zd1rH67D+VtWAWSVm^qSw^=<uy+RfLn%e=u1y;sz>W~sh7hta<D0*#W0hS7No*4nvE
zXl=tsL5r@&(d3~)>#faQQ#{^J2vf&Vay-fpt<`NN>EIPR&Q+G$BjYv>jg`w?V)`w@
zGCfJbgQI2DHWr0+pLB!qklVQv{FcX6B6fvNTm|J%^q94$^VJN^&ri+9zz_J|tP`s*
z_s#%5@ut+LItyc8S=X(T_%rB{tVm*!Azdn1+=QAf2tXdKO!sQJ*%8LiN}s<_Qrm^^
zT||P5q*=5&pN-O_ixaUu{ul`wt~83*BS)p)5z}&9H>(p|g8ByRq3E96@i#jum6I$;
z@1|gX#)a|jQi|u?a3QaTCxIrMNyS>(P37Jfu#z%NecJ$e8SFdE@!Rp=rIya2SOlz3
z7ssJ$o!~nHARajyHNZTX2IXhtFDyqCLc#|vW{H*fAJQMh_)hSGKn4H>t|6&r)_t0+
zJ%^#ApI7hxfKDm}-0iY6Jr2t`5Am6?^G7ou-i9*UZ{H>!{Ux2Uj75#CKFD*&uLuVc
zisa4{ZSID<&5%`CF(bdtTipkapYt1l-_aKg+nMDs!m9XG7rV?v^Epkl2^#4L<9x3k
zwxYV74)*t9d$CNIt)J<))X&g|E7Du%na;Bi4e2AVh0dL9nqBu;tXuhG+jlVtmyQ7*
zIt*I-?*oxj2T>j&$oko4BqfrW1f0!AyQ2YDnb(K$P}|78i+46!YdR~qL{}24woB~S
z_-3*h8UO0Au5N5(Sv?OU_+Y{EW^@=qWY$wdGTir4tW6?nP?9znZs;eG17NBySeGA;
z#HQ8=p=`<4F+cxJ>|pj?dZi;_m4w8k`=}0V#~Dj`M48Y-r7ZJ)#m4Bn@CZ(+*5X5q
zYsyT4E(gs=%`xK9zwjhufv}q;s6yCgNGh%oSVMi!<}sawTZ3|Qj)`*TJUH-yZPA$8
zwBTtv2v%D)(!3eoI?6)?={H%?ZwU(Y>U-)*Zu!v(wX1<kYugNfyU`Gu6?}a>u<J`y
zrJ76!%$&}4bnvBOm~3?xooTdEy!P*b^S1Er?)A5A;wA;~*I#}V&~?inrV4<2h&;YD
zb_MRG$`cJexrydK>*1ZH5QkYexfwz9bq#o9JI!-dL2F99k=nsoTJBCa!(`gp@Jg9=
zgM6bzc=ueI={oAL&4?G~8s+gpg*MuU7dP}vmjfO7c>Uv@9P;AFvG1+>&^#<|XrccP
z(FylrRclyp^pB3}a|n7(V2XWjJbik5)W-GhL&969sW}1Z`U%Z6@_01|0hwQ#n6b)v
z_*a86M#tnAxx@c{PH&o6ab)Xj>eA2EmxFpixSMBs?BCMGie)gG+Iwi%tNCFE3^X%c
zmNBMN2nmNrQ~D?tXh-*|%Z+2UPn+;$i83{pz2ns7pWH47PXCp0h`%psyMPWjFdu5h
zjlxHh&y=14cFoH2GSYz->0Y@x-%-ls%cjMr9CQjOH*CqgS+clC`5To%JQ&RklE-p@
zcrbG92B&(WOlC=jv8822>ku!-JCni8L|QF({if*ZJkF(&CxX_l%5u*TTiPKK&x;nh
zYelAu%JAH(nOyj!rFfb^MhAXP1~KT<rN+GzLK=bE=M~|EPc7Pj*hqPY7jpMeVb#qf
zbtKx4TOeNN6r^~7b9f)+@+Jrtt7l<{@r)VUW^Ef2P^L%R+GJV|I?;2-leu~bC(W8F
ztR#lO0I#4vOIrQ5a?7P8S2O$HuJ!CFoaoGtVrvNTx-a)93@cYlm}IZ9DlQNzaci{l
zD3_bo{AShS&@6t;>r&_=Io*#&rzR!^r|G;=pq%eZ@ILPG<<+^zdfH>H^9^@JgIwmo
zA(4Xh8|tvup1OwbELF!<*o%6t_f;YGJ|RgjD|)L)>%(*es*VPmO}Z?eMh$iNmUORf
z&C!gz=2Ndo#2znR=H`~)*NgXzk<I{}W@=$hF#Px%9nnV5DneEbuHBZ-9*stTUD{m^
zcMH&>t@-DF1Y<Zn2hMMRyu!}NmwBq@uiS27?JrZZx!n<5P#D8Jn;GqIRZA~#yi*6Y
zekq$|$U&V$zYor%SL8fOe(7J`S-z}}f7?H~y)lOHJ2X*i+t>3u7K?;J07~iA=*6t+
zc@Jy_Szz2@-H$o)_R!(HWhv_`hctD|4+wLDt<;WdKf*VXUd4OH(Owl-=~OT(C51~@
z1SIyuoqE2r325vukq+KJ8;b^m?up=|EyKX3eq?5U6no3K+zW<L>Mnc=ZrI6~*w`%L
zv1dWw!eXges|1(Q0P&TjANjs+r0IW&v+zH)XO=ZVu_-mm!Ra?<a?pPs=4IR6)eCd9
zP=C{AE)f~iy3LX<daxvdeCVcfwYtUGwoxD$LCdc=U2fZ-jxln?JoPh2!w7_&xF<Cy
z;n=u8bo@9xt~!kEPAq~03gnnuXg#&*fVQAzkiAaSd%_R&;N#0e1*r&d?Yh2hv6qf*
z*%g@8yYMdK%?-e3<ZtzMI^CaXE6Rj@G5}O9x>T4aTOJ%hxNX^0DzU*;^6~x@>m}jg
zy5CMNqL0eWoo8|GWBnoAKi>yXyoHj7ALQOUzuPg$A3(N^2B!5p>w3L@y3x!iZKB;K
zW;mGb@o$(@_rVCvPqs4;uFZjsYx@TizRm<O`8zM}7qkf!H0<OCeXk+m{53{DgN}EQ
zgouC4PerpF0%vx6o$<V9-JXU3RyQ9*gNti@Q)|!$!z`4=Z|2MGAhC>%c36SdRrSly
zE$KG?pX3J9|9+-pVr6Cc@1E)Y4}`-oECY;2BdUR&=N)avBEdLi_|Y%C@>aBoyR_aN
z+`s<DMh$N3LY{Beq%wW{nOpci9oD>@JFpzCo?lNOFT!(d5?C!~CxAgKj-FrNk8epS
z({txic&Q3jUmRZV57yauES_lqEI<!G7YN56Y0uld0lw6C^pDUtx>`rarBX~fKJu}X
z_46_~UCyj?95Dz1g#E9t-Pd29F`*H9^}~3JdN|O5N|vE*7(q5MsiCi!W@sdn?jv*K
zMYy(h<5)zr{on-XvIQAc&j44t8dgi$Z1loP9;>OQC?C<u!r^B}u&_?D8#97gY3pg{
zJkiT#<#kHK%@y^r-YX1q9q2S$bU5~&XIy6w+~Ng~l4by8IW<qiV5)8|=_X!n>ivlI
zLz{(`DF0EygE*!G8#>@{3zmg)y+?=&OC9~wiNUIgTkBl?W5pj-xS~2@$)LbrV+;wh
z`}T*e<Hf?$c`!nm#>Cu9ShW%uH7)EM=S+N#r(UY>MbkJskIYkqLQ!eb$B3;?GicdM
zBLAI@xYVXttuW5hkz#u{?cG01=$65#wb~X3-=v$eNiHp7@YyR8Y8F6K3P+m_u-c?N
zd)<|m>83Byiuz>I+=CL5C|VtuY-7SPDKJ7=pv0cGE8(G&weMh{A25hc_c-ZO?i=R~
zgxP6tCs<!pc{Gg!xhm)hDV_p*8W>u^4t|8&2}u`igA2xN7Q%4GsHbfZV&Eup2Z(No
zsR8T_B2Lx|wM1dU<Q7558{3y4qwD8)HF2clp)x9_6S&U6wXsa<g^gE<+#eiS%>WbK
zy$kNdG8>Nz1aAk`6L!xC;|W>s;?IY6Wu3O0tV|bLJG7kL(tpIWyGI+s5*;;poyj}+
zO-7M?BmP1vdlIsz>!91p<-gHi_@*$6JkF0crimr6%3}+e^d-vMY^Dp{lq8*UGRKb#
z1!Ge!tf}?5e`E}fI1}C{JRJ$PcGupE5x6YdPK`C2hAk&kV6L_>*_jw}eHX2V3|$E0
zgQGB3Fx=*HxZBNCP?qbVXtp&qdI=wLKSu(m6FMzS89YGdrT#1!G$ePlcXTah{2jxw
zV&MULT_eqEKgO#rK^EA(#0y)e91+tFDi*;D3jyLaWt9aK<zo;Ndkvtfs{kQ$3u855
zKrpXBLNGo1E14|bT6ou{%0Rm|p;78gF}tB0BlNFB`LB$a^I&0Ei-XS#NTZ7vYlB+v
zmWC)9Hd-KwpBWXE9lG4+6FjH6*W02A)B78w(`n~Tj#LVEX<DP>jgm_CZQnV`8148T
zw(Rg;65xj&l0_9F9A$*5)5}*58?B*?%V`<u^Y|&e{J&zzk4>*EQ?c>FvbVYU#D>H{
zOw-WS(%c4Vtd+$!FC;4)bINA_XWz3#$z<bXf!v<Nkga4ptaH$dIS1JsXl_YazeN|B
z@pyyvm6H^wxxQgh_|Kyb1?A1VrTA(VaVOsM-qQ9-YgyKjCG)`oEdg!*V(B!VoP%>m
zj{|U#V^R5xzkjp?@quVgsv~{n1s}W4kQLt#kda3+cj>pg*sLit#;8YQYdYk`f@{#e
z`wc+7Y+5@jP*s1iY%(SxjC&c-(2$1dRXhD&Y!$lGdNkn#utVlOP8Hc_(y|tCYVt;^
z`w$!^ziF4_Slt|Ws&1aKFfLfxWi1>Ps*LcN#w6x}inG7=la;q^__HobFTh^2Qw|1^
zXr>4`Ypc4Xdu=8RgzwGLefCZi3K<$#aeU?3kxY0o258G77oZZ%T(aVDn}k!`aV?%n
z)wXtgyB@-L=xj!>bmdlYv20UiS!ob$mcrg1FQ0C&zxVTpxhZTz(te)$`tT2+i$!$3
zU5Rv&DVF~Qape5@zSx3kKgdFhS>Gqe7W@2jOB3-+vT#Vog|w`v2%r;42zV^z2SkuT
zdmDpRNtwk4R5oPC7oZSI>Rqhj0o>zMC7*IX)P?*o-Unc!ggEN0fU1vLpE*%h23R!d
zHI8g#P{}i`4jT3h_+xUF2*yz5uy06Er7-uGrd(>G;$5x;;`-{oBceEoyHT&~pnpo2
zU=y4NZ&5xW8dj{3Wur7#Q_0y$$#;@wjz29x=~W!?3sNbZK=7U*55z224Ua}~zcAcY
zJ6A!p0^+3K>MNVj1!?EpI^1elft-K3$^Kolkqv%OW#R%mt1#_x?yvw$=;99TlbUUT
zOs#*3Z#d!&Yvt4UvekBS$)m{l>a+B4XXMBD1*=_FbCsA{U3aTKkL21agM33QG;;qY
zRH=2x=~k2`zH&6pnWOd^Tm_1fL@0yoPfS#DvJ&&7l=6Y{ge-UqGxH-rTSfiRljP9i
zO}@IVs5!2d!4x&9z=4csl6?=w+jo?G3PS|H_T#8-(Zy~$`b~LX(U))GaqG_`FMrXm
zYRRatJ8XZH!2B-Dr!ty>L??959Voc|vR6<<e)t(gk`naL<tiKz6ZeM?`EeeTLAQE{
zAT+v~=x5OTzj6xivl*&AO;Jkfijlrfuv@P_0lNrm>f`p*RmzD=8~WDL#j#f*lFLpW
zD;DQS^4J2yb?fV@lMc75BN>GH%r`smj39fRgsa|05Uy0u(Qk6o<~zkpaf^x|lK@pY
zS@OLSB&|=U9zDyGOFa5&zS9e?Ewgpsop?r?R2i%;d)9$|DEu$J-6&nP-c&u>ax<td
z-tLT(1xXG;xosHChQ;nhfKMssqW)&p^k4x`qq@RZpsDXvV?S7f-nGFAc8=!0_8492
zg(D_l8}+XTiHec1SR%x_U+fjz^-9!6A*IdT7d@HUtOJGmgCz`ALl(W35uEG>H0?Nh
zih>6ru#A$0+LBPGbC}1_T8>(oZUW@dc-N`SrOZ46|5gLwsP49~4g+ITs`u|0FW7l<
znhX1>E%JyHF;<}i`SA@wWeeMD_XSR&YIHRvYc-}K{Tt6`sFAdD6FnaG-KOEVc>Egy
zwEdFL3{sJv8GK&}9ig<vGItB*G~{J^oBH;HOGRK%v7Pmf&DGjr!Rw?ZJN(Ba{A&@l
zy6bdlm`%Jc-Gh4J@KAPYt1=&%X7|4d1|g*RjEVm$3yR1_yb9Tl*VA1~V?If?SeEK8
z@OFy}w}fWiQo`;W4FoEZp_mC*V<bA$VnFXX+{PjNgXP(iav0t$50#vB?WDB}x90Li
zBZJ>xtMWeS2x=F)PR)ZoXIVvL-`+55?34Yz@>%xe&aD|LY`71amSaq8jA3^zm)a6@
zqsiWB8$TDvi`*ry4zw{dgS!06)Zp0z_=Cl}aLw3VM@OiOICkB-o2JXdUbLT0q`emH
z*BJU#;&bmWmKLmc@4Y{-EB8;P*F^r&^3OjWFdR?x{08IA$vuz1yEBDj_Dd%mlikki
zmh_|6f8g()J`MiU49W7p+fA8SIsdz;*EIIHoleJnZQWT>tN;$(Fsvm+;zm;7R?b`6
z5d*`)xFL=SS-8o`r7}shNuvGvDi*F$!S~aej+OrU!_!~MJiZ-X-`>xg;r{VxgVfcs
zS$%n^zJ!04HOJZ2I5laGV%hAjoIj^@1+HJ-9k3EsFn+v0xX(tocM;Hx0>_`Fzdhio
zC;SWl4`c5XoLSVa(Z;rI+qP}n`jU=qb<!PmY}>YN+qUiG<gb0|)T!G4#lBco>vCPK
zdfqvocZ}yR;a1kk?j6;_>p7fk3H_|y`E5_j_npbs>ni@K2ZA{S7<aYgux{V4{Y)}8
zi<C^rKmhH;DP8$(u%burbm-SmCa_{Xv|9k{q)0_Yv36V*;h1d%=u+b<RHMB24r<n!
z)mH7KbjuO``jlN`L2ELH3871#w!mm|rz4%}NEsXLELs0kf(5*T$695TbE?R8Cs^Z<
zzG=oyO-_Qu8J4`>@4KvFB-oiq=hzpJP<LxR-890J1J_oNN--u20qxSiQKc{XT=>11
zI3ieFO5D6EjdzWgt5y*G8w#Ufw<3QtFj!7tjEm|RP|QpdAveZXS#OhGl8N)Mch)1}
z0A8{xaIw9oFil3%0-~Ah8Q3X$Hhm1&k@xk#RwdZuKz+uN3%IZNY=n|po%-ZbK{gA!
z4nLv)(w#G+Q*zD|N+x)g6}xn}E;kLBpM#u&4{#LkEii>yHocqThQ@Bm$rAUvk-ADV
zmh($B2zgbG1tp4{QdFk4C7q;H8XR^XeRQL>+C3Gmwo7bxG-<<m0lb|$n<C%ss%!p=
zzf-teSj#u1iL3`W(Sz?$VgVEc^YQhgzfSjyK4+$&EBAsJs&uQ(7&Vwn76KWT3j!Cj
z>{k&?^Bi<0i!t;mQ)!ptv{lFnuVA^Wu-wt6Oj>MLoWm7@nU<WwQrwdsO3R=FscXoO
zfvZw$_(@l%;xxo?tFsD|ERlinSW`pU($K15^VXD8V4YL5VEMSG6pj$j8MJY8kwDLb
zfyA1O?#M3RZ1Y1URU+#nP}L`p)QhEHAP++9Fav~W8RDtwYKEH>@u3?`*+c6_WSFsB
zD-AWvLia3cr+!DY{<(`X=d&Tf#qhEIbrHMcX?C?4b7Znr+S>IC49?gYD2ar$gymRR
zA*-E^Z9>Ast<@ehEg^s6IKEJu(?s{Tv^FE#i#6(NNA)j+0TcQ9;>F*wUJfkgD3^ix
z@gVzI`%6}FPd&3^Q1qKLd{M;uD|*khllQtojqf4)PdWOFYxsxLOt|~v4{tQY;v}xr
zWo&Eu;?c<ghEjs{WZK68$eqX;2hCu-8>>21yI+^=KAZ_l!X*QHv2>?+_i|bn_v=Dv
zC*H^STVg=+Y9g1tNh?J=G?CP{F|f+1AuRLtt#5_qu=WfsG~i%wB|8Z9&Uy2yKtXS{
zVKd?epvPdax-t%&D}a=zeN)W4L3X*!Po-Gv(IDHDeqRjs3rQ-ZD5#k{2-evj<ioC)
zC?CicOwG|yW;vdROu*mYORY%`em-`1sD8S5{U#9@(Ql&6@Bx_0VZn2pQw+>e*KwRT
zyl%27#O!J35f`_@urKoUp@A`GcC2SDO|v`$MeEab@t0eI^Jl5qY23^rt=<T$=h3*&
zQ0B|TA%?Quv=*Ym1r?cbw^?uKV})#JPB_*Csa<7FS;hHWECr3CURs`jnaTI7vRvs`
z<uXoq?6_MNo%!`+NYiOp{+$)X(i5JiFw*a2<LOid`nid=Ilf0E^Y*SH>4|X5mRZc3
zOyj@br1hEV_8g{xpIkvKz4D!!3RsviXP->%m;wgm?eggW$BK)+lw?n0MAe5gs^sz2
zm`#zRAPa6-jCO*h+t9~Y@_MXG%!vC)oC~bc_*&h&g^FP)<Hv<Yu>?!Fm$F5;lH2Ty
z;3|Rgi?Pmk^Gy^fIdZjeJzY9%w}-|)DW0K&40P9d)pPw;HefL&$U&!+jRKY9J-lZB
zwL#W;88grE4zuHv_gp<qTd>Ji5>N9_zubDMz$fE^lXUd@ykV#<MKbMbrlhyUTPD&;
zf2G9nKz7gzlD5sr&cxvdG>aE0wm%cbg=I@3brKd7J`Gn8>J%FNRkUmdq!5*SO;=z`
zlTMYX8p2wLfztf0SLjAnqa@j%6*%{7#}+RK`(Gr%d-;1H<u<lDYxPX46FX{}Z0f_$
zdIE?-f2mW@9F%D(tZDbyc|5biAH{r9(>j;v37yhPf=d#@D(9Ja-4S~;y=U=l$sT%3
zZKxRwk${bD3Y9McMr@lM^YSRvYT7I~(q5?zSP?qcfM)Hlv=e^sq_0Tk(?9R>wowx>
zI8$z`x;u1kNs@*E3iY}Euw+=|C+1;iZY}2}@N}Ce!9oP1@G#imFXM80(c`#!S?Q#*
z0zyh@8#%&)VLhr+!O_8G#yiGT&ewGW8(J{9L+2wjQVjr9m~uJBCs^i7rL!I@U57H;
zirRK&XZqAR9#f~QYAvFUfa6%)Ed=M!wEbCSBEh4gfBI^kDxwPtZSgt<sFGuNPoENy
zr5Y|Byja62RMV!Hjyt99nl7^*J9AmbqIMDp4j|V-3kHKQ2J)JEfMIy~{WhB&uP=l8
zNMe21(Q@2S%~}UKzPXC1vrDa6y&KGYB2>EN#;b~@N~S{`H)QXasi(Ao?k3fufD9Rt
z<JbGV(hM(jqlwe;=Jsrb6Bxm)I>CXIB{8uF!5=_WLa#s8kQXFDGtyAoj!J^}^Jcf-
z%Uc-ZHf`YeS7y-|Dp@mj|EaF0`VcH_Bk!lyHmMFP_HKcg7-N_soAFi<fOfqnc^=;?
zGE%-wzIm2#-U<y#2+*`bWfrZ7x(@9VnIom!H0!k!<5Lr~I6b1p&NZjI7P?fVsb{EJ
zxEx90U9IZ4D5&6WsOX*@UbV!1tm=i-Q2XK{08}xnZY*&ewbxK7N|TttU-Xj5Fq?m2
z3-27co`W%>z!YwZWI*##p1EAAND6n)6XLyykE8UEG0aKDY#c$_nMz?RN319GXTqs0
zyeD+z^O|wnY8SEPg48=H>Pi(2YAN%|wT6UD`bC2w+fJ>{+ADLjPiT%&6eBM~C0MG;
zWFEp}qZk#mn8<CK>D?s*Pe$ut-PLBI*RD9oNm%RvP}nE2*h#YcFMbGK{TU<Oq=H>-
zQqy$CpwgQ-{X`}WsC*QBeYR%dlEqE%tn&|Ee%Z)g@|ryX$%y!IXNC3?u-A9a8y(jI
zW<V~ZP<wYCBakRdnyJAe>UMfL0(}Ic&Q2;T=lv*vxy`b62MZB?t9C409CRrHbhx}d
zYdBNwyCJ#u-0IM#uBZ@Ir~LIIE(=#?U1n4GKA&h}l07Nd0D?l1K39B%=WjLcQ8f$+
zqY|UN<0HnVi|WJWFNM?{CRZld-?vB6#=WLFe^JK=SXB+ZRGCe|hh?Ovpd$$}>GCC#
zd-N!XLt<j)adDQjlT?aP`e;^bAHY=<tp20&lk|Rjy)LeH&py(BUEG%=In^txaA-dU
zBa;_Ck2K^!@ijvK%1Z^elG14YO2%f1WWrb53Dxx5(zZ`vs2K0O!j9gapgF4^4NIbX
z-MWcaHEu-#?KgzO!yS2@V)u3FV*a!-iR4w3S4@ac$w-@UGH}C_sH>1{9UGD9fqv3S
z@h97pp9Ip^Pqv#f3Hd1m>gE;3-w#=N5gr?`h_MMz*Tc*5XQ+i;MHQ!xHUQf4)2|1M
z0a#ZSUAx>QFW~B&XZ8|0xfoFCqCn+|5zFB4ez(aIO}_rZGsSokv;Nq<#v$G^^p+mI
z9e@5zK&48jEmh1n{*wdO1G4DIN$D6-!Yds_Xr$Xx6Es^g&#Owc?<`M!VXH&+c%0Y#
z!>MP=COWB^O34~mV_3&h;WN6UMcYK>OU{SyXf?=K7}h?UHU`i40SGM`XX;j8Vb@QS
z?^dJkmhb7#)&MQRiPH?E-Ar_fuLNg%tJ7yBltn_bK&<_1cxcr!*6onuzsvNEk$jbd
zzGQ&9fdgdB6p>X@;%=<ntXZN>fbM@ZJVRfmS!Gp_%okQJ-e|YRP{q)WMoUdPKB=KP
zl=PVxd7EAAh_mQ+rR`|RI9pjrlpo^Ux!oJ#-k7ghO)qS8Ij}_Ot#h7?-?mUV=c;N6
zdEKGTh3}VK<91nY*nYCy+4s{NW^B&uwa3^tt-;DnEo~cP<Tq^1sg=JXjJa=|x^fQ6
zGl;W>&nxAVUt^Y)5?fua!T7>5w8yji@kMuRw=>k8hWW{L>RQI&%KHP(({f5g4ls|_
zr=OW&;R=KlvPqoOd9nPXgM<T~Gx)B%x)^EaZq{HZ^9VFQ&|^Hiz2R~ej5P)^Ki$8v
zdNPRDUa{0;YRe<&pwNx93Z8+Ti8fkxrmx)@dAyN$%A~U(3{HLjO43izKSUE3QgO!O
zl9fgO#vOr|mRHk?dncgmdf)uA#+lh{HNIqQ_W{Wdg}336HxPz#!NJ4$Z1ay)ClP89
z#_J(=KA^t?T3vjTCXk7luyyBn4yZxFFI`0CzmD>Ket&)*d!^ONe}bjA)UatE_m7xb
ze~%?~!gNgc^Wzoq-1@jQX6k!-4;Br8jYtPBi*EWND$ySHxS1%Ayew{PT(%c7tiQq!
zVp^GO8!uUR9p0s!5GDFHZp(Er)--sVcyU-X!kLDCzIyjvhZN4KxCbv$@!D=$A<Af&
z&jU2!xq|upd~Kb7WU|N*a;k599EnxsW6j?3Cu@*+yRC)VLKkz)sL+cv-yv%GiAejI
zr*V{(fsNHFY$?A723ZZjFI{6y99VWrpZWI|Ec)`HX}=P<Svo6g6(+(XATsf;uMgfq
z%T8GVlOp$1rAu#FzmbIClB7Mkt#Y2G{3L<0*5}l}-OZ`)zs#1*%>TdHQq9xBoKexl
zM%BfRQGtk=iHT9d+R534h?|T3|64L=Vfw$(92a$T?QvVt{idslW|A+1xes=IYoY$h
zpV|Foz!{*f(d}m%nj6T{nk&hl<@tL4M38JM9H}aF+-C#M%Sh!Wn?B^lj3cg0R0?XS
z>$CYNubd%f+68}=!!k|)6_(MO?l23v&k_{hWnqnl`t9k0(mRalu`dQf7*5nMm)f^5
zc1t1w9eEh|8;t7?P1-8o(@U<Pau*rGR6lD_Oi3UHJX;e&7Hi^|z8YrKRgG^6M~PbX
zWlh$oR_87iQKTFzbuarLwGWYo1}_|I0yT<R<1h{as#bamq98{t7y>EZ1sF^aWFqwk
zn5<kN;~iMqp!_E<T>^+KP(3y}YhuW(v?;t|-@h@^JSK`7Fr?*`4v8L3%M}r(_}e7*
zJBL4ro+0QIUVfN6_~stA0kqV*c(OWm6_6B-3G9eEJuSAm|3Vs9W00>i#?b<0nmD5$
zJ-fndKXI5U<g+S~YbE~j!BV&i!7l~qDCj1!1qK<X!1SfUrGW{=Yk|mJ(`AU<ggQh&
zhI#^*j^L1k6xIEpT+jJa;%;buV`AloP)guX(;G#Opg~_pS&Tsux*CnaWUu53H9=up
z00E5J_YyGzJ%xdr7f5>o&XS>eSZ2cAb8HEK*&s3RP;xr<s>dRbR>g_LrQqN_bOsi2
zq{U2Jh5uTu{CfeVA>x735ij?|026O8c*crkkQG-Jn-(H_p@zt$o<PA04xlROz%e0{
zh?fLYf@2EOkDq%RR1`$FkIwB7b?IjGd;c^H?<6AKALD}dwHO~gI5?DL`n%`KYQ2{4
z$z2v=?5Es8+g@d38T2;X?Q`YFeKuBw(&q2DGQmUVIKo_#Hx_{hLI9Pw2+9;H7W)PX
z7I2gh5SGjs;cmyuQ`WY>x0&9*C;HfWen=&tH%(jNV<4QPtvT5SZkb>Z{JfZ_-IQ-L
zYf0w!2w<-o&rY8n?#Q#$q~Pg&!5~FGHh3K0OaZBCV+{?X-VuB`qFpf-ePquUdffhU
zd@9Ac{nxB2(Nq*>3j(3F_vp_Ge`+^huqm(Fkv5s;M^ewS=s0znqxcJRQZDbM)p}Sx
z{aBT9@Q0^02;dW%MXeA#i;j&i43f@|&0@`yW+6eRAdU4_G(eLaua7Q%_vUY>GuTuR
zI6Je@z(fD``A+)&a+r0!AB+)a*7+b7q^ES{W@uO8_)&bHjMYo4e_*R@y(TPFDhB&u
zxc?&Y+|0~ii)GH*QRqHYs1yCCaFtGk?+VwLd=>m@fNbAnEZ^RvgIjM6d4F-@@pox`
zi_`7jQjDj6IzSzZL^@*V#8E+BnWE>BHpQ8~i^KL^yyno4O@MfX&MtMmjQFVO&G?`1
zr~}|Mp<wYU^|Bcg&u*nD2G@-68$NEfYpv}v!4$F(ub4o%AE%!lCnKr%-I787ebthe
zyH$8}wDlSYzWm8bbe&6T{h<Tg%3i!Vc;g!Rz#YCo@#iTPhA&uD_EO6(Am<QXSse|J
zD@lyYdia7r(7*REaK{{H)#e<;<s)c1Y|t4|jDN?Kx$c%ee+Zd@vltH(M!YqK0tj5N
zj~~PGBQNyW3CEq#h9K(b*gn?|CSk0b5GKk|UtVjvoen1=Uau3%jceatn=cwH*Wb@i
zcq}#~N3e5Vllw8TgphTRV!t9y<T>PD-((eF*Hvo=IxfgXbl1MXjk4=~=?JGQ#s~Jr
z#YzKq`9hB3kuS-!Rk^8*sraQ=yqxr>7}Cv|88IU~=iMM5ZS{NwjP4oBezAsw@T-6=
zWBvM@(@chLbG4tr;EC<xHFb0<9p$@GMPaPO4sD%0<QEJuUq3;zJu26=H3bbv%bOqb
z4I80<$<s6hk}uplBhF?@TMRbkaE|$GJfC*WzKFd1i|JnG!E9$SUT{@s-tc@d6lK&d
z=;8Y9kWTdqHVsCOPiVYGlsUJOKyB^ALu~*p=dboL=OaXnDrhBXhKH@SsC+lYi;5KL
z&K3T}yY^9Z`Ngujx#?8wx@LW!zQGw0`yD$dRmey`1iAbUn)+rsOEBVl@g#XoOY#Wz
zsPc4tMH}=B%2{+pm>)!&yj5JtZn9Zgh|e(ABl+OplzpX*uWh-buAxRUea!FUowDA;
zM3Q7pe_Be;iomDjdR;ep-Q;+2M*~;<W@;<_wJIFU{WDunzi7^WT_6E1pcgr!RQvWB
zx7$aR0fhMwV*LjfhT%oR!Qmo(g|Jbx$i!&I+KK~QDnO8$0TvrIGn_Th`5@FkLp9%T
zPHJO@tM4`wfzp`>yash{fs~}=Md3~S9eMpmNPTAqW%(n><rAX@cm_DOYS*|3bWaxz
zpKl4Dj>+TzgE-E2T4#+(io?6IH2S3`YC;AkF88nw`Bb3#h$=2XnylxnJf%O5euDEH
zC^H-%SVB22I2_Z(3^QLb$g^~pWo?03q0@$u5c~qd5NCE0ib|L2uW~J~yIGy!5?7+#
zUL$+Jp@EVFj1q9Wiqjkx3dh|QS|>bFmQ}!OS!zca<2iTnMWl%z4FaY7=$~Dqt%drl
z23zA*>&3<VD2{Xu`h{S-C!T)&)Oucr!%mMIUI&~)@>PAVQ5@^a&tEXM{)kP=Mva6j
zM)pmQjM4X(w`FQ?=)#<?jort1F4sk^9y$w3li!a*cLxNkOEznSv6`BicRs5emabJf
z`xXruBGoMQ`8VixkFn1y$Kzgl@@LwOT;g_InSte%l7|NkeoQ0oT%da`61dL!QKYoF
zqYide<D(5()rCb~O#vJ3iEiZ2JNLjO&=20Ct&1VNg`>LK=Y7ty?bp9SIS%*n>+sI;
zvhhYe+?3a>p2u?W?micfNi&}*mTAaK$y6VB9oEbWoOzjWrl!4;2cb8^?{~jnD1KYD
z&lfWp*^51Bb>_76H=PhGgfVQ{Y9@PgRo;dpes8LQw7YV<u%2fn_EafoeS@kQnQi@-
z9K`b9Z0lT1|F@pTAZ}|kDJREuHI*enMOgp-EeIHJ<yjK2B?r$v56e<C#mcB#w;$oP
z<WlZjg9S%P(f|qJwr}iYaW|#F=ZA}@_nV74llD<Vjg!|q0OciU->}zl=42G6HxlLN
z$m{eO^>}W_xfha46NWHi8{hBjUXmPd3W<wk;58o{`AG2i`H60VdOG`7HN@RG3O=7=
zs%F<aq+(hl6UW8cv8WM=2x>#W<et#Jo9|W*!g4F5de`*h@$Kp1x;U1~jve$@v7n$}
zC}s|rs=^+?=y&_+cz02EmLGoa(j-@vv_D&R*9Sj5X}5WoSG3RJkk#g`x`P>q+>l36
zU-^Kh$F3e}p~E2LE9eL~H4lxTex_C^6drkjWOy1S70*2kbk<sSxEp7Xm1g)FOa3C1
z@*P4p&pH-Gq~1jglu$0!PM{O$pX}*BB2jgrEJbsA^m7|E6dlXXM0<E>e+*Hiaf*ZO
za$fmLK=wAj#8}#qWh{_G-Q;p3QTMh=?j`%pmAa23+onn;Ur~n#WmCxO?)m_J<t)su
z7NL!zc{--(W*C!yEa=DlcJ9NY6Etne&N$X;fX9WjZECFf1t_jdSrM>hGox7cl00!8
zW~35nY|Qepf(F!*u=I<@Su&9#cjR$I1)L9YX(RlC6^N>>*Ki;{WSF_?K3i+728cZJ
z%Hn5CEB-QwK!k<JW`~lYj&z8<9PA;n2m}WPrLf&E*(|4b2$iE}n)!X4ifa9pHS+yq
z)}r5i5Qdmpw4HmKUX`yEzyeoMj^RM9Lz_;&K^_WE?jT;AzH?l3<XEb$X;Yet(M8oL
zY$QUF1ISWEqRjPu>r@aDJ79umS^uYCid;borblbXFYkF84I6IYFb_}WHaQswyK7w5
zqnPV#lyN;T)wh(7Bc1Fc8im%=P^wVjP|sDfydPc3qiKkzgEE69S#|z*7`+Irj_3}G
zRw|s5Z0j<i+aOsyTCa%O)NzIwJmTJ0l`l54L2(JgKb=`DD$9A&uE?ru%I=b!HHTpA
z5=$QiAm{be`I%H6BQ^i8yRU`IfXEVXt;w-eG{GUGR=7BleqsBLR~iXJwfz^9+6$Er
zih|5NRY3V7zeq9pLn-OvN(9`?A^F$nj?vbk785VK%^boxHYEEshpAU0O8bbt{^6{d
zk~tb*lb3uI=2%I6$7BIY|0r>`p<%n~xSI8qrZrC63l>d72EwRS?^T8tp=W&v3I4s-
zH=NRM7!y=TJ@bhcuCi)sZ~`G}L}Q1;1<wUNZ-)AOV!gG$MsTn2_7+J`j+tD~cz{g#
z&GU^9iD1m-aUR0BJhZfsHX+9-*J#<$U#u7P6%^_CNRy|i6%-|p>I*y4cCiB{mj}xF
z1;8{IJx+4ZET6a8BTcp;S~ZpQbuhBTCz_wf{Z|4EVi$gcf=`^cEg~g;qyEpFw`tsv
zur>o>{F7kU+BYo0HZZf0*(Rb`Aq{}gc|1@f<QZ1<I*LH5zrn}T$-gxIu-b)$x0-ue
zst?<4##9$kksf%eE&Yu;mn$hfcGf7e0GnjvFO9BC3t|A*2~C!#laOs@+d~Pvx*QFF
zNXQ}dX8Fw;q4G&Uk39+8ALn5tYZ1^=&9OUtDKHGIVm}LwL0ai=c5WRY<y<F|uXS#H
z&{sSV^kHCGo$OJYRfXxcv3&(hnxfR$1SM^LX*NZD3a6b$SGXuUg@uUh!f?8a!(V77
z9kP({FCOgrvlM7V;rM{>V5ISJmYdZO^|$)5`S6NO7}<i7)d|P*96U9iO%x6@f66bZ
z!c>%qL3X8eRoK*EOg(wB9V2|{IJgs*^15PsFC|~^S*`a+KV;W<YHNw^6DHZG?Lm5w
zWVuG2kAQY7-T`~NeHM|*x|V;&$z7wK^T>zMk3i&HNHp6TVV7<hyqfK@gBaHZXmF@C
z;OffR2+CyWOr-}<#!PXseuv7uK9j~Au63{F-qR2c=%_9IHS<HO0SD3Azcg-xl0*k~
z++UAWO7;2@&=c}iMj6hFvb6Q6C!lt9Z?MOWyH~7?t=xeoXt;&&qUUgOd*IqML)F1n
zG*HTx`l=mw(xT4e{xlFj{r#ZaTZsB-$|Lw>5FS75m0`&K6A2mQI6ML_??^@JT?7aq
z*2LgiHMIcX^;_1yiJ(Aip>;VcsEx+lD?#JWV7Z5^wK9r}{4&4zT{<?c7Ap9+cubqk
zzxLqNasIS+<!~a&R>;J>RHMOe5d;XIX(@950Md*2vPQ`MP~z76X|q-hSffs2E9B#g
zqw@#vU3JB4Yb&vi+p}C=VN0Xi=J9W>`<)XUp=cqT(lu44>C1iOyCC}2P<e%qWI<e`
z>zF%tX{z3SCgGUypm?D&Q{9~i+J}I9#`oV^h7N3*hCv(t{Mp8I(onWeDrmK3p`tmb
zs<mtJl4i_518P5F)9G%hp)Bj^&*a!Nf#+Y&!CiDM8O~u+itDt8i4U5Hz6SyrRzVMD
z4_@C=Ji|Su465nxdr3~znCi~fpl;%rGYIk%e`Np(O^T&>kDZSb5}0x(zeTHjB75|t
z$NJoAN>E*|ZX}e`fdjkjN3m@nF<7j0UbxitW=tAda|;8#dRp=z3J5*!3lLv%!KooQ
z={EzwsEt^Pf276oc8`^tPa_|d-7Z~z$dXyq->SRi!%fC4ISs=d42&!Ep=VX%uoEJ8
zuT$bKI6lp1%NtB|YmywjziO*PjYJ}hm2E=K_L`1DfvaiUpql+e1vqdExn>Qdc6;qK
zGc=dxL-FU`9>DVL0j~JiowhSu$53URHcr2d!UW{44d`lkw=h=S$cgcpFWH_`M@lk%
zxSS~NjIMK-xqE-WRr}uUkWgON5_EA0F@Hqk^%@_7tN|<M3JU+24ee!Ifz2JIHV}Z&
z(^6x^i&@BLTW!sTL`GNQuPvPq=fS5A*%_TedVHl9$y%k9)&U<2(TAvmA8cyr4EQE$
z`ee!BagBGke0Jf&<;xgygLVg|cu~$~wS<0DsWQt($`Nh-WL-5JeyqD?Et3g~YjB(~
zRixd&2v5-bmbr|j3Arw%tGV^O4WFfS>3Psq>~8(*)v2*b_)TVQ(dF)qZ{N?Oe|~Vb
zxjOfA7-fo^b!y0MTvLP%3w|7$_t@JR;p@P%V)%)fG<+@d0Y*|>Juzm%?THEn><13s
z-(vt4Vti3w_S^<o*EA(LG(2v}(5sQI{9MI&mBOW^*UtN8HA3b6<yGnF0p;M$gUQM^
znrkH$AO~1dpO#tZ;L=!?2wvvD8}fP-w32W#Tx?W_n7lK*o=)@*^qW)t`q64y)x-**
zzqVHosZLYCI89<FzvtE4sSheS%n-(>LV$W(9ya-<%i^bBiI-=eFH9z}AD<d>A)9%q
zK=8}07o=CZj%W0znDx^<U`2)BA@X$z8YeO^o+-ijMJseP&t4c@DV77;;wFxGF@BV$
z&4_B{kRH~+g>Qo07&2>~qgIraGgh2@G9Y<(>SJ9Yq~bz^Z{nCYMzIywvc{E2P@5vA
z_?MA%W(?<r43c?D|DqlMnaHi-AdUu^sxOrMg(7Y9N)1O^YM+%CGR+OoawUtdF61IJ
z{z!lL)PlU)Ny@!SBla{^xwP|7$gk~2TG<n{y9xpth7<Os_R`mnhDDkIbGOJQAJAc_
z<7~ohmu$xJH(SiKi$8bYbKvLfTu!kMipRHOBHR+WJT2O5Zp@sbjitPNU-wF1KEB>>
zgh^K9$Nx1w|L>(r7FKqa|69vhXEd6&8R2KC@&!@{ncRk|hiT@?4swJ+_&TT}D7m7}
z<Mms-BoGHs^UiVA7cV33ctkvY_wf8bs=AI3mt_y9K3s>hhudZ7lb{0ke+^cX|D&qA
zoj!f8PZ*m!mtjm$0eEr0U7gPB8L)e-{YO>j=?Z21t=o0QU^MOd9VI}l2m9PoW3!Kd
z*(Eu4wD-`eq8aVR6BY<Ea~hB=DB$NAs~wqrHFP~{gb$OaZbtO?cL`CZ8Cj<kLdX2L
zWSJyR?GRLiI_ZcCRvd_Z;L#!b$kxmx?}_e0-_WU?<ZN~5Blfd!D~1Q<X0G)6m^~&{
zh+-GxoW^9LcdmzQM_+S+c<oBg$j$Bj&3_Qz-;DVnQp_N98DAgA<a$Cs$Vpca^HAlV
z$o{S1-}%84H(3-XGRJ>9`JZND_V2{)I!%T+TOxYu_Ih9`7Pzd6r68YJyMXODiaY{)
zX-gOg!rK}dWb)T$Z0ag2##0#3><^69YmM+vkyRa)cpPfl)mwG9_!u3?{a^e)m-)+R
z!NPM`hRaC8iuc86Pz~cABt1WZuAg#ZfoEkLB<^rmwqQ58pS)Ji_D$9+FUNQ8Wg%L?
zxaL|{G1mTyWvW`#Ia(q0R$rqLWW7c9dh!pL8+GiI(v18MUM!?<kb@Fmqweda7GZxW
z&5HYI>c~R8D1y!1(Z#F*olRa-Df^hB#!!dnX9HtQMn9UpqVghDf(EJj%sFyir_eQp
z=PGWbz+$t<ujV-mgr=@q5m1co4um+LK>yefuUf&i?l;wZ|I!OZZu=ZNr?I+AQv+Cs
zajDytVG)C<FQ2G##dqBGHbKmZHp?6ou&|-c>|1A05qX$|>uh(&dMHZ@a0H<8izcvw
zk(vG1?cF^Y*ov7*G${xbY3^5C>Ev#By-LzZ(iaJfbCt4E1Wf!sXJ9I!v&vwHrcg6*
z1z|UBXJv;ev|1&Z8WxDbf#f(?^z7Km3tZhks+<X<N0((?3(A*|M&a<P8bVUueN}Du
zaqrc@NU8`BmPQHk;n^`>=S`^Np^3U`IJ+n!I>B&oi)F&51Tg}cu0}Zw|MJ7l6_O@v
z4gBC_gphUprCF<8WkGAH2Bf=#f&E-XbExdN`-6Qq<hBL)&!H#TU*$sZ8tOftFoYU(
z6Dg8+RSW!RG-01ocl3p#g2hh0(5sr~BEinLl@)9xlEGxA1qe_u-89*U<G!uVV;Lse
zVRE7zzd7#pY4W>EaBZiq)W6C;<J#xTgcn`l|48hTeyFe?jWLw)Oll(P*)p5Na;0b@
z<E|ACX>X#M(bGT0;abVbyJ=}sT&libR^>hog2ILQHd(^eT9rD%@#iP@&DIL$nTs_n
z^rSG_;zImA%0Z9&;pIbw{0sZg99vbbSgnmkw&@wz#!Iu|&52`RAYZR2-S-p<RDr%u
zzDA7RIX^7~aG#7F{hsg=H89r|#tv8U?pvJj18$pbx|bfTIA3ea1HAYH!W7WwD-wK;
zzNW|pIG@cc%(Ni8@P!>ss$JPsMyQ#u0knDwji5K%Vu}eU^yG2MSCH~}8^n=J4Wrw>
zUDF4ygz^N*XLL@@IkK|cUM5sDLl0QxROIhTlVwWxU%jq(zIb!jn08ZPWg^$1iSZF&
zt6Jc-v?|*C%@4eOSL3z~$XlziP7sum^V;&(L(2@w{bm&jIJ%r{tDiM$M`0GtI@EzX
zh|@u@6&HDr9o9PM1OuYb^=5BP!71`Y4u$x<1~w@HI2?^RK!ZB%V~<m)4S&I#w(&PR
z<jPfXR_^hbz!YZX)npJmtG$c&@vK_Xrx^~3`*mG{E!_JF<!lL;xgs9IM4Eev&lhr)
zVr&{#(~ENCIn9?;Z8=3{!|hY@*(XO%_-oC!c()H+Jbm-zmBEfXF3jOk=@~?xo5&W6
z%B%W44`0Uv4d62mC-T!S-Tc=yy=gqf>PAb&nYSO<8EBvzy!q!KJZE+Qxdr;iB}uVW
z2F>PTRf2M7=WX6)iNl4CQoQ?Fkn^1?21@CM6Y120IXx27o(Z>NPKfadgA9r)zu4lD
zq!;L}*|{gS1a#3pp3y@s5R@dSu5v+u2||3iPTn?=FPMsHPe(NvY3{l)SJxeb0bS{K
z1zIGN@LcTta~n~N$Km(VLAhYIL)hpDHjcm3S>M(&)uhQL_BCSotN!Q&GFk96BVJQz
zsd5%XQkKwy&y54F!iyGp0<)(JlU(njrr~y;Eq<-Z*%+6M8FbzLt9_{gzil~hO8?mq
zq@HdtaP=I7ie3MZY}xJo2wNjK>K3WS?IlPsK|PM|1hR%%4h!iJ=(Fm(N}QC%7JyIn
z@`h2?xt@HI2&Lrby4jeDjIUe6V5V3nPe0UftD*h}Z?+82ruc7_=yBl})R<pjghRw>
z*4dWIo4rZJcQrWX3xT@h^P#Ze>sh5Q#>RKOiIC~K^t5HB9w2z?hzABqoNB5!A9?N)
z;ajUzoWaT<i4s(c80=`6;Ps4#1YR?_$Rp1-{o^Jk-pBX(lv(u-U_gAlRZ_vwf0E-c
zzUj=<IitJQe*VA}!2A@WH<RW+e^Bp+#BIZcScou2Z6`^&JVX_L;A-wpAc*+Ay0$X_
z!rpq#w<8jj->H)3GhlGIHO1n{E+Ia%gsaNuB@m#YIBQSfK@oV{3PQPQ@OSje8XYwm
zZTtc_qw2gAkgsR9#ueTFL4)Pz%5b6f3^%IYwV&Bql{!7K=vQZBMS7OMpIA`RXp5Tm
zyD=o`>Eb&z_}ZL_rN!NQE-CJgzg@<LTJj*XWDRE`@kd~9R;4$bPv8BodUW?zNkF<W
z!Ryb?hS5#4h$HDmu;*+UpY_HuL-XaR@;ZqUy-WJPyEIrNwL-ukcmR4Lu1Z8DOQkU=
zTcUZzz-iWsgiif4(n8g+Lv9Tlsw_a`&clFzgjQL3DvPu9RbbsL)x7fJw$%M_{6=7Y
zRNI1mvZO0D?`{(h$SoxLbps)GQNzk6yF9vScR#JS6Qyogs|lJ$*VU@c-@mG7ruK6f
zqY<wtxP>-gXle@eck1`-+IO{4cQy#pi7VrM@4{~_B!x9$#TW!T7#ZFd$)JdDo69LK
ztSzhJ5?WmIR{4@>!R;Tl2dKQ6K+JzjU!4CfeQ~k>pVHSLPHVI&XGM1P4Wd~2I}ikf
z?X^dsi?$qQ0Zg14b=)_jh;+e~rQkGH`rhkIZ+gOrq~FVV58}4L)HS1Fl?)-?y#$x1
z$M-#vQBZ&J4wKbnFLZAOO4xHZVt1Dl13jOm2ivhs@BU76n7#9dG&CYDSR|C=$EY>(
za<;=3eyH!@BjyW+Bg55yR?N^Dz7+mV&Xz92=<{fuW<(=_8DSp1tiMk);;mv?Hz}~L
z=vqPe@Bo3~8sh|%0{@0+8NxF!P<iK~-Dm8?iSmw&R1EBMq`xs{FlR#od+9pZ=K1O(
z5m;AI&MkC}Wzdv>GkL(4%)oaCg<ht@{QL`;y!TjG=W7GuJY}dFOBhXMi~&qR{jqE%
zi}o1?1XlmReGl61w5(!L!)9?PXlvpe8Ao9v#n0eGL<An7YwyWoV4eDTf8Ax+9WIGb
z)3SYz;q^V2Ms74j1>J_Qk$ri!NV_lRb(5`1`3`1Rh#x|diy~2Y=V2pgh2I~LW22mY
zMLzsH>f*rR%XMoJ9z*#(FR<Yh6>qfN=SltkI07$aV-5S8l~`6xDlNfEK3E@#0+fju
zqj9UiTFpg3V`+1i>1v+}dVNpo9ENO-c;weEvQLf}TF1$I3uoUFwq)MUnhIW}=T4c@
zk{6(kq{dYmDfTZJKDS8>&qh$%5vjZ3EI?@sTIHk#Au6h&TuJp-O_?~0nFUWB_pty5
zcb38vM8fcvJsXE^#lvBs+P=TL`8nC2Vt-bq9;o=>+P`}Mx)MwkNs~-dj>iAxIi^3G
zu{+Y})@m=ZKkLKz<;QrA&_CkN7YZ7tnxYGkbBBRNkgONbI!!?)PZHOr_%tXll%CD}
zV_0QCM{I*|+=@+1p)Ji9iei?K80|7ANz#q`x|Pwj#A191rG59P4R6`tYNA(3!DA9K
z<{mv_-KLnaAGK@zR)8kA&SZjSBAuP^o10tOK>7i`Ib2!bT_Egx5B{{h`pdhGQ<yS}
zU-21q;9!}+Gj4Huyi)w2wFI($Zu-!(L#H?67V!oKj~j?_L0C)-8o+ty8^-5Vp|Tlu
z2#9#18-ZRcS|DOr_qoaxv>hisG`VXI%Mr<mmx)#i)$tC$lk?`hm0dDg$n_)Z*p$NM
zmU?%P!Q>)KgakR&Kab1uWF)L(P_hCstaT&xt|%WKvs-8<<8AlmxvC{v+*rD^3SpFs
zBl+65H`mjo=id=WHS7!QEVuNON3;uD-sslkyHFnrP<WJzXz`mWuduw4Q*wQ%z;oPp
z_gl)sKyx<EJ7>)rqTf(;xoym>^T}hhvB%nZ?Tz*8!Lf6*WT<J-zsAx-kSzI*xzw5L
z@srRI0JKlB9B+WR>ifqMJmd_typbNKBieGHd_6fz1VKs}n}{)D2VTTc3qaqaA{3iS
zk0pb#Q8*Kc!Htv?byC{N3?1NC2Y$Z(vLVwu>yL4pCphm4cIp>LM`ega67(v#;;9le
zCf8x27Qjo?gg)tft2)a3Lp!!zijY*+q|aWi@%iS1Y$@ZLccHLIn}T{qmwtwzKE=|Z
zZcqqaRmc-QptFch3oaTOpjeLlRyv4&D21Nj<~pQd<JvXjKbnZMvALAN^3a5#ZItAt
z74oj(7FMK@?tWd0ZAmk{X7sy79eKig##QFhqA7cp1Z{)E<D_CH;Y>hVq%ZII-@?6^
zSDBATTrm|=`d{SejRx@TQ-B?0!n~l)X{-)2U_k2>8D`3}+Fz;=67Qe2Wp>IXBGP0g
z48*b^*}Nu^Dk=f578D;Nq%3FMh_VFT;=7yl3>On5yGSX)y39U>e)?Pw&5D%e(?5N-
zJ$#ptfURVLR?0D3>rW~J$-$qt5=v1+g$2P2gb`-(Y;-6-kldI@j*)cUIndCu4KBNE
ziU~<eY*1(ND`m<brU|cFpQzn5mqGhN+xLtxwn291x+K}N^Ym#E1?WokY^ES&kThrJ
z{XyV`Xw}rFl}A$XI=v?xs+6EbP$YJ;ed9f945&E(tap6?L#J8RvPo5$g2Bb!-(^yn
zfsC=|+lN@Ia|xZb7R%^$WZ6j6v#IkcpO__^f3kfE1x7j>Y)<{KC48)Ndv4l4k`!*#
zqD{*N8!vv}X_cpO1;_D7M^cG@Q0Ax!rg~N(2B2$iJen2hz$yiQ#9v)$s)(9OnY4#d
z_gh?oUW)?mLrwz1ni?F$OIq(J4&GlEEzo*p!xsX<wPmKf>2VC1ujcYM3WxRoWZZ`J
zG$iaqM~E?am6c!yZ2SgAEU=w@G!1uptWdnjydkLKhd^kXT_!**i<m8A@E;{zmaw{P
z|AupQj9}x=*&LS(ZI;vS&P-K#62cZLDp%fY9RGGn4$i>ozl*3s(qv*o8OMi8*z&I8
zvy#R>x1)0cW0-~0WYzD1Tn-eq-BB`FpYSN6E@RbasXAZ(0~`6yBtqXhLwWL;22M=r
zrGNP%71tZd>CQU2dYjUj+Qd8RhV4~N8}$^I;aP`y7<ipy(Zy?Rr>$Gz&qevR*{b5W
zUm1#lP$l6fx)e{Y+rqN&T#4eGsb!)<<Hq<x5q{E~&r;29IaRL~+~dFCYvw#747IQw
zyNuQGHj}QW`9_uo194@0Z+$kZ^kvXgGCYyaRMDBm!=;kE?RkO$Gx>hMQfngPA6Bvy
zI0K!1ed5lJpe_y1QmeJCz3?vI+LpXHJ_h5O@q+7fcfN)Y8Cq`YZAt%gOSL9}tIqld
zX9DRXR0DH2VW>~N9!2DreXzV0@o(fkEu%UgTUea*Z!uIGd6YB&oG9i<Izvo@X0%*x
zYD><X3PGdYJK7a?WF4S<T*X=k&&#LulkR(cv14>O4Zy<OyF9j>hHc%nhO7%u-uxVW
zh1@omt+$jMWbMLrEuEysb1M&;&IqpP($dh!A?@$dlq1=E_p4=jAhsyr9O`}nFWKFI
z$9KM@kV57_v+?82$=>E*boTb%Z|_2ldoq`PmP(~tJCa#vWH2fh#piq|fP_@^c-<Z{
zf>rHqc{5Bz+G0bm*6QI<X-v40-u0DG-yl4SO-6jdVzPDis&%44XOcJlt{JYbUK}n)
z9zB>(!=#?6uLT_kb>j3Ntj#-*R6SY+ZCFPWNm7qIq}o5xTmDNX@U@vT*;wiHA8BAi
z9igra?PPvf)3<#O9mcBu_3iuR=b_+@Q{wr5i)TMc?0*r*m^l8Ahhw-ur0(Y1F2h;D
z8nkUhquP-P8QLCB?lx=zESUnl0wOnruTQmDy1J38tYaP!4gP+L^G_0O(q#JxJplnc
zLBS4hb~k7e9`uL1*Ymuc(2(!M5aH{joP-+*`fPu?Za(Q!{vSN|^be=|1N;SV_cV(z
z>MxUMZ@9{IOfOA_f5~-!{hgxg7c3uoS3f(jq;D4Tjq~q|tYFWx8P|YRVy%!Ip!U1z
zZojWnVGRxeAHS4QhyW3aPM_D?n|BaYK?cvKU^EKx=8x6`<6O88-T%PGfQTdSXBqu_
z2Hd}2pC0WE{Zv7cJ~4~O2LD4R9*?<WC)?GZv3WJ<lB6$inPQM3+Zkb)?Sae*T$+7m
z91mUuc4%LL?A|v*;n1mkm&p(L16+)JBd~Bo+QA_+(seJKS7JKy!A5d}D^d!$rL8Dd
ze?k9#aZiIO|7zs((I*g4|HE_b8$iu$v%hmH+TbReC!>YlQ<I`*Jwl7@3Ktj>2U63;
zI`(Q8tP_A!35@W=uI(-UU3n`<8fpLkl8L!vu1Adb&?V9Wd3NJ;K%~Yb*c8<Zb;&ja
z{a=-reS_5rSL=H`)z+ZSKkiO|F|5Fa%-!7_=QGbQcv`(!-;%-YT`1mBcUSJp9r=mg
z8<=!pRAVyB6)|N&_9EnuK%|k>aZ!M<5IWMtYEWmx!EtxE!*{d>cZesr;6xB4^eYsj
zvNSEUpcI}GE#v`YlfFw>Z?*%RSao+aAqiuakiFAh8G-wFmPKUKT2U5OSVpYT9BH??
zcTX!Ea}g;NRfnnUFlf||4N#`7<hlP2iD<(lDyr>mk+r|{QuzKLM9G}fTEhtkJQax-
zqia$PTpY-_i6AJ<Td*Gpvd6n|)ei|t{9(2Bt4XJa&jvf|U>Y-^Bs(V}v-$p3A(PSz
zFgKJmy_el5LSb@LPsl<pE#HYG)d##;Ai64)O{pBslyS^$l>0<vj@lA3zCdIyVPB1*
zU&($(2C+2p<jcda)td0mg!>s@AO{F{E|kPAk_hryQehWSrZj0C%H>j?Ep^1V)s>}8
z=w<XN2foPg2_Bz~6X*~TSJ_b27q}g9bL6<V>?cT5V!#{#CMn|mc29NAM)SmaTNBfC
zCh$nDvx-wdHi*JIY{;+pctnn2MnLGD5%@VQckP9mqrnqOCT8J~W{d>Y9!EKN2JGF1
z!{bs6eK^&}kW)bnmq&=ZTiSS3Yeadm)pKR;>0Cr$6Jk^%6loim4rG6MpWrpVpp=iJ
zJJ&a;*Zi?&jj4D@mQ+mH-{LSL`BiLi+9+A7`PCUYe`c{4dr46;^cRTOqrTBRC{NEQ
zR2uy^!KNt;8Ko$xu!}}X!IKR}<t8}0Y$m$56+OWBwx$Btoi4CC6iPi?6bNE;{0yFP
z+I*f;63a8nti}YIfWo@q`zC`nl@(iY7z($J%h{je^$e^r0_L73lDuswjszOcd$mmK
z$uw@^a-9Y<Zeb&Ly_KJ%C8?TjHb6FSg-$D+t=5h^7f43Qp;y5AcyhY)rcr736g?Qr
zE3HjVd@4wU-1<-^7?vGP0Geq(@nkl#B2`2cQ-@Z1vC8<-Rg{TXv{}b$-!pDRxI6?H
z#)>^sjp6Nmz{wNVoV^|aI84rKB8H81BDYznn`*cz$Av?RiZ*DrlB_?+YQJmtra9}@
zYD%LfWR`-8sNEL6rD|4DPOgdx(X&z_rq2l#h&J2`aV{${LC#8<3td8WqMNOcGqKYF
ze2T4MVtBk6dOvz3MY)j<6X#Jip`rnfNf-Hf?9{HD=QqWl5mp*>!(0hPLFt@X6VQm>
zCQJTUJ}-I!fge^R>Q~nteaS^HmlMMP=Any}XxT$wFUD?+KZ7Bmz@!mCFW@1}+cBB*
zpf04;_~6(pC50>{=YinH0X5{)Y@oj!2$yuz)XD6~fP@KA;@Tj8EySYd-|ixqT|)&*
zMTc6WmN?!k(_Q#*l<5ueS;ucXUFT5~KtVS~<lhVWJNa*OM>tYC9a>oAjTY>-r1U_|
zRk8&ASrY-hm}P9f^9<tBmMF}~eqPq;s=^O-k4{V&q6v>r4uh39-NS7V`|z4Y{ZZ{!
zyN+2X^f}!fd6}1%q<dprdnc0B=704`7&3Z|eYhOAHh(B{X}NV<FQFBnsYE>FvrXq@
zH|Yqj??aH^xD~N1xA&n*3;4&0{^u2SYyZcgLCEOY7s8qG6!Lx>N2y0)c^A@t%b}QZ
zGc=#YG_`dPR-@g~lcU{*xl-9ulzBA{PoVrGxMT4gaBVWEbPp-1`Ng#EMWo=0QhU80
z)s}^u7;<az+XaymM+T$Bf->BjJCaM8nb1q+VQuwJPZUxekMv-3#Qb)^H_es<f%(9(
znX7g5Rw8o#c4X|9;*BkF$oD%o`+9YhAl0L5Pt^lnVsVG*?*6w*4vak0A2G|NcXor3
zbq^lu$Qh+*-7v9Euz}BG`qE%B-aXYPw|I(EMWpR(Fj^!Vn0wm}en{h6l;ua3Hw5}r
zu*asz;3TW`N56eOflELc#U}Kgr#v=K8vk5L5xgcH;Lx#}OA=pGH*UEYnGpP=l4@&K
z?uI0UX_~=}W7fg%x}n)|jv8jiekcfUxqL{LLr#JlPfyshHC-Us@~yX$_?R^Z;A@ub
z2iI*;n9AERBjn{WT0!*de9cIxnN~mj0`UB*`yj$|_3JXkFiKK^o2#qg{f}zmIn+bY
z3fKFRMXP_dy>Kl!`l;Az^v}^mknH?Hvs(oe4YQ4iQ=uXIHC|vxpsqAy1}trl{kghU
z)u}evGZ#8|F!D~cqOmpSdfD#kIfOj{a`%;n6-0deeZ7w&#&c9%A|hmX)kWT}9DX9|
zbC%+j(G5nBlwgWXMK*CAZ72XnhaPRg5A~q)Z*5Zd{KdpzthMd=fy0&I8eiJN6f|JY
zhYk8&i*kMRU?UT(ZuEr2XN&8S4JA)X=wJ5-O|pN_cy`jpXY)sD@t2iV&FbZ~$Ac#X
z6BjQCM8B&*CI}rSJ<0_T%nCwZs@mRQBN0P^oTsfiH#9mOO)IvoK<g&=PxZJ(9sz^-
zQ4ITZs(SmHTRK_r@S#1)*_sab)M8GUZn+ev_KI?YCJ{p=WlIEd--{`mVp<nlkW}%M
zD7&jwNL3(!HGnK|^d7$LEd8OLZd-2MI?J`LFFF0H`B+=@bHH{)Z}{%AAk&x|2RYIc
zkc2-_1=O=^-kmR$-I{ite#+bL{$m@h^^Ywx_9j~`eMmg=<+Zc)7@{ya{36NfME#p7
zXfS?1*^s%rDvs>hLTRiIzgLGBWz^Y4Ejc~d1Kw>mgQGWBnQmp$p61FQc=*K7w()c>
zd%1h|V%|Yk^|<Y$E{}fbc*4As3A&!h^29<t(;9SAik}ILe>hd!COkbmZ@VOoyp(*_
zv=qx&5o&wZ60UfU{%EZuiu5x67uFizq8qxhqkoS7b|@qdbZJ)@Ke98|wh52Vbs#Cj
zcO1GCH;&^fwXf!k_GhESU9CH8{GmEG&rb5q(wt`^3kSV#hW`GeB&$imNb{4a1Haxq
zx=W!Fs}h~$8ZpH;IXbg_J-Z?FjC%?W=YC_&{N7V~dkhhR*@eGs$zR@=c)fW+qIYOG
z44cx<=LwC1_~umhagfp-^PjZ?aO)yc#<9zNvf@X96^uwOC!;y^$cmYUUyiu1(@#m~
zFnR%`SvJus!gW(&nNWA!TY5M;CQW;p|Eo<;IkJy?Hl_C66qrv{#WCq_C9+8te1XM!
zY32{_?WTnf6eTbss3ZwqDGC3h#O$J5X=(Ri4#}5o$G>tlRnuz=)nMe#x(m0H)pFVv
zcDGh$7!BAcW=PRJL*-}9d(zRj1erWua&>g0>FW!yrIOPFyKhd=6BB}q{kiaw<LDof
z#^-_XbIFFTfvnF2CvA1H_FoOFaDCFE$KJzV=Vf43H^-<xoj#jWAsw-F?m7?zGg{eX
zu;2@a)mpTfetYl$<He;h957v3+{d&gL!cYNf02#K2X+n}r;~L_qEGQDEYCBv<p!*c
zoQ}=p%u$~138Org_Kl0mhimC$-;g%s*M$FVhTvxTUo!*?7uWx76{k^m(&<NY-E~a;
z0WwJ}6(p!$sy3)lTz=Hl*(yJ2?N-dV9+o-Y@f9GLBPCO3uFcKmMg~J9074|?7gtP3
z>G}5L^5yY%sGlB>O3+(6Kb%dOgB$l1R1kroh#;f|UmTNNyq#63mLy>LhV|onjLF~5
z>bgCQfKa3lr~>NnafC2=Uv}E=4De;RWPE_XHSif4Eazx+xOI<<WK<AToJ-74no6Yt
zo>Cqjzb5P!7&RGUygpGhFmljfTs6u>CX29}sb-X>&HH~CJEs`YqGesTZLPL#+pBHc
zwr$(CZQHhO+uf^;+vnsy+~n?u{WO!2IWv<=Mn;Xls=hxdq$!{DoK%~@JLCupuU5#i
zEL=!f^#JE0+&$J6Xq&+Z6sVP{<QvLqd(`u36D>WJI(TE$XLn$W5XzEt3ii-QCo*e^
zNQB$`0YjHtG<pkHy_ZhKgrc6Uh4;Fk>wWjqPqJs?E7svcjLsnOM#(DA0Jyb&+{X|l
z)(bPdruOfwcMJVFZ9*n*5aFNHk9d0b>V`XEVPEM6DS?_N%I}cYnZt`P3L1dHaZ0kW
z;EAwo657@k33kkf9_JPd%*GP3D0&B9>r2R!?mgWR%LbliHe+JCW7-q4h^)$$IJ^aV
ztL73*9yP;^`59j2x_`xNCHHePWR}<^R2MXSe`6JQ*7y8N&YYpY<L;Y@N)l7)Mnyah
zy!ly^v>ehU&%JIHyu?$vY923ZTE!zQHvY|OVR1Y$qxIWjxA!<ld_}c++K9kx=#Unb
zP7TTvZq_rdKU_>-v_}t7e;i9<20CK2mR(*3DLmPXwJSvE8NM8@w(j@V;>rH_VkVZx
zi;V;?;_>Akz!nOa)yf3L&?{trzu%tEmik5;K=Uhzd5GwJ{_XvKx5upUqH{EfOX@8y
z7TP|xjZ0dZhMb2`V)3<2Lo7+38Z0nE21!dd8sRezNFa)uP$RXL9x9$U%qoYS`+3m1
zF??{`TT1)OK4@ET{z*o#p#8?%p3JBMGT?Yd+%>&Wio#0Y1l727Gx1_uiBFY*EkBbt
zr|(dV-7>!4;#bKn=MGRPBtAYQp$uFI@@eU^iSugKSN55BKc56Ox9|UYcIJh`8Ts1h
zn7dV(sU49s`Na-O$uVtuDD+j7#4!q4O6XUTNb`zrN2uaJ#B23-oz;*bS0;BS-;mGH
zyp1A>JlirP5N4Jo?q*wZvT8nXjVzc6i|KhY2C)DUNN<rERGkp1m`F&Cy0yAHd{*_9
z9Y;T~2qW&8Tm00I(F9=-SV-?q8*<}Hs{sND?6OqC{x~^~-Q*36O_il}n&P5GFzj+_
z4LLpcDRUVr;MqKB&ImA(YozLGOi>J<U^gvoAz?<DQ7PWERn3xItBTF#p;&7esJX0+
z=ye!D`)oJg4MUFDH+$8Nja5TQN?$c|Nrjdvm{KgqqTgK|4tWE|H<x08hfY!*gbOaV
z`x-Vz1(uEwr0BffDA%=X{fnkA_roEs#V`{9IUiZ^Sc$dDrktE@3bv2QMB#eK{Ufs>
zPt+%IqgaoP(^)D})!bhA&&A4XwF7%V&bzg9L2xwipggjr1~93o*+a{J1*(q8d|W(b
zue$i}R_&bb3-IP8FjEq^s~GhjJkV~<jKbf`wsO?Ui*fGY#e$P~0f#4=zieV%cfOaG
z9Yr)%@Ji`vVBxLc5|2twH<X7BNwKp!l{il7TNv&?a-Jwp>Hpk}jRMeU)9RgF8uQ^l
zka<)e0#i&P^%b?H6XTv5doo7J2&I+rz<5%*HSKIL!k52CuOgXHa3KVvWz^CjwQwz@
z&$(${KUOe9-n>NH<XIorL^3g2fEOo1TDNJN@zrD4|3c{_bPzDb<To>}^CD@iK)*+6
zYlc&s20PlAdD-ce?|}x-=N;{|9h_YSV<V^-oAW^NvQ`+(my8Oy6PK}LJOQZ&P@?&p
zU>n1BrubsdbRWDd14%C)6=h@GcXxEw=~xpgaBx{CZ<7}mwId~s(!j#(V>MNDAAm+5
z!X9301|VuJF@^iT&`ywc_V!k|=zwVK!X>CysN8AzK^tBM9lxPpO|(?ZYk-(O(^jfM
zp?V}~#cm01%l8Rxd8{_>^hqqExej&xp0O`D!A#R${nJmhcvN@a(|{h5?mT6~m&gAO
zEl?LUEZ`n58+1^|`r-W5J@r4aPmYoNUo1mb78M9nHau0t#i+z#H04`G#3~}->UJpn
zC8nQpYX6|cULF?@ur7HcJbKlM()^*x)LT)2q>1%YAB_UDFoY{to-at;uI8FCWg;>V
z1nI8Inat?Fnp~jJ^?nBD;qa!*c|lAO)a1$eb<hr3c(BKlOuHb<U2=c##!tX)VUCO&
z+x1B;{h6EAEw|q?f9GQPQePfh{E_*IISpq6Nem_KCzqOA1(`%9k-&>s69>77%et~b
z#PSsBk7dTwMwR?VvrEdH`MX-8g%BZuVhyRvf{MgE9FR3ZWx_e->XHgyefAf*4hj~y
z;41#_*k@rI;1sBhq01>B>fCn9E<ehn!M>glMWLqfK)GN}jHtp^KZ+-t_0D!H7VDyr
zNxx6RG?;MO^OeA^`QuH`c3nE7zp!i5^nKiByd7mlX<`sMi|b~0r!U);%02s(?}FN*
z1eaUqX0PkZb;xu=N9iEhI*i8HN!*@J=I>I<v{DWMH?!$BmBAsiv;m^4^r9h8T*xVB
zxY9a)*KJtcnjI)(cy$8*rgL{?5_!*MhAE|e)3)l6%Yr35?=>k<xnSP$Xj9<$2cXxQ
z_L#9dR{Rs0pjNRj$$*OJPX>e|j(%l}b$JT1J0dhmbD<yVMrMLni<@{tle@|FL;HGm
zJ;(Vp{D34M_6uO@8_ia3vf`AJ61I3?xRZt(YLuuXxKc&&V#;8>kmHnfI(gZ}e6#KQ
z)69o#^Va~a;z%)WvTj@IGMM6VW0!buMO0imogBh}P|=wd9ab&Utp_H^33RI$w}o*!
zfi_!DBytj0h26VgcY{jz`kKKJPYP8fu=6p-4G9w%ritm|fJ1LWX9Fzgg+ohw&ES$Q
z>}majqyMdnND*Fo39?KSA{i9i@X=K~AdR-ZcPcCJW#^bVTg@k7rQHcslUby(Zl`^B
zAFwM<tB29vN*>FqB4S9>KF3*7?P_jA6X(+5Us&PRt2ffqCmh&+jC0$^oZ+q_&n`Ik
z&N3du!HT&Y@rnPMgRHw`^6>@{Rls6STr4kuB!d4$)0CqfJN`iNCs07kkQ6E8{M|M2
z2x3?)><E@$84k>q$XKVRJO(A_d-ik>>AXO&hn7rlCngQcOaM(J!Ex4+Wr{V|K^cRd
zI#9HN=Ck#LVk8T8Xyy8?E7zzc)x%1*W92W_yCj53I;vtu8*5q?RJHKKZoca}qQwcW
zI2*PfnImM?`FODS>js;_5p^miS%`~jg*p@dWI^;|Lfd{o?8!jYu^;-5q9!}9zU``3
zK6!Zozy&+CYGs_`vB4oe5r!MyvqM)Q0yyL9>hE5Y$iXhamuxWJTg)}|1UmkarQHB9
zdv8U@CACyF@;J{8mZ~)-V$_6NHZj8>!QYF&B}KVe#dpuWrJE0X@Py>r6(_fahRNo7
z&kk~MVidef3uH<)kMsESgmkC-aU?kS`Y|L}n18{a!spzZ<7m|PFM%(<#y#=o$i*YZ
zPR^NlyT`}baa|OG;6<FO>ZKaxV6msbbUWqEl$l+5-bk|4;j>Iri%3Ut+x6t%h&#4e
zY6xZ3jqw!~N7v=mDePyvbc?w?e(-~QaM36$iM_aSF?#=uu$wm%DLJFm))CE3Q@7_}
z?klTXMX-tEyvaCk*!wN8I>X;tLHX9Ef`0<tN4uJka_&0Ud?IcYpc5m-2sJEq%|J0#
zcdLNX^~iI0DGq<bD~x{s=bZdMp?&{LCz0V_UH_2vidT1kM<cHg^&eUtM4S;BsD~w2
ztFWy1Rk5cT-Pe~w1F>Y&p{a?7Sk@Yncsxoj<86;q*T??(!~4@dsj_MMqMq=}yD9jq
z-yxe_m${QEU|RfKx6d`N?~xYzr`BB|dnytinlD$^#bP5!C=tLdKvqt50R4AX&L(nN
ztobKuEkaH(2kRD#eGqIfk&$=9!y$mS8n$^g6aqkfAlKLX`ysOIWF9kCCHl-XGgVNT
z^hj|$VKDvj^%FKZ%<A_z{mig0;%I1Qp!m1hU|V4sGp!KOyv89Je)tb^9#@o0i2J!{
z9703~ae);9;;k8C*hnnN(TT493CrP)Zzu;QQWO6sOKi-X*CTvxrEIvkI}d6O$ao4z
zx`Jw%S9#rd4%NP^Ux|)MJ3+Q#hU5gSPfqSJ2NT%H0y#>$e}>>VOZBx!g8<^5Vy)6t
zW}fQ)1Mu6+Z){%FHu>rGy<Rp~OYiu_^%lZq+)wyjem>{F-ky+yHEw&$bhpr}v6=y5
zE|s$9`;ri#p`NYWwyYbY;6KQzl;&>(WcMuMDn;x|qNrbQsDx&zgBK@P`E)rQPERdd
zEheK<P=LKK)pFMY2A<fn!O;a@<TAM+{tZq)iuYg)_C>7_WJu5Jby5M6VW{yPj7i$o
zHlP$>g60%z@3$Ibj3J83#<|4rFA^f+z64R8Sm~s)MGgZzZYfa5qD2aK>5-%DdupOB
zQ&W(>?g>Ghgb6<4st;{#m9HkU!R3+2ci2s%b}kF%Y@|xc$aDF7df7#!g6QT7IZ@c`
z96%2yHKI%eL<8@q$rEi2)wAFdX*z8$s(gg@Lmb(2TgQPI`^6n28$T)RMjz_RDEL-Y
z`|del`|_GQ_tj8ZrZuN*DAUEme%Jlt%a|;uG}kza(;iX$_hL6CH5kvgAqL@)Q(Z{>
zhv>}bRR|g4dI+PZvpwd=Rm_jm4g+^rk_nQ;Ew$QY+v>M<5T(fZX!SU$hsyfY(0~2f
z%RH0Gx71}D<7#&Ae@SzKWl?w>;;sCvu^Ig89-A?|VX7a99z3x-P$k1^MXM1ZI8cQ5
z6<US(9Z*!2mjoo|g#ZE8D&p9W{+mW4T$JHRFo~oRfN7v4^Yi%%JvVbmf&ENS(FeWM
zX5zNR6yJdw0kjMGh}F$QN6bXRy}vi@sFbF)6*I{6dp<L9$^HeM84bZ#4}M7qpH3M`
zh~M{tFcnWl#4pxP@^Zz`B_fSf*b<nm$+}8Y<)y?X(AZ}sMKDfyiak;j#VJ<|bmt@w
z>0Mn=J_MTgBHW>)b=pwYb1K=1+cU<lfjtrmBhfCi3fL!6z}}ojxGW?hs5O96p6zzt
zg1WhsNJP;<egxT%m<MMee)6c9M^t&`s6*lx3+ZQ1J;Z89SrVi#@IePbVg*U|f8wb0
z*69A3J(Y0E$oa3sfVo6atRo9zFGLY)Ah#!ke{v>6q?GI09oS{dK;B2xKb0+&q$84$
zOd*6!UlDSK_LA{a#%{8y-17cRbnf2o$Fj2@ss4=Q>q2Ahd@X+=In03e!68=zpf`HK
zJNmhkrHR_E1>*3Vonwq*yE26(Cjpks<(>2wU~)OFf!t(#a)#*;F49~29h{TMVrEqs
z3%Ea<wb#jKUcNtj3DBmBydc=k9vqM$H6o;(6j<Ft-sQT{<M@Nk#S?s)RQ=&rdo;(Q
zn0wri0nIhnGv<8erD9B+Y>!GwyxR@`SXgW=YCipRPtf9#yy>tDnt-BLIt&m(w{2tO
zYG-8YY&bWV<$?&5bJ0awJuyF=iimx;gQh?W!BF9{*iL$$Zh3qDxLrVD;St_LS(JG@
z=~Pyu^sgOOkUg)L;s_q;M00K2rK5huf^ne7W<P*z4NyD(M64RnvkrYMl^=;v>1awB
zUx<UyzU#Tij67)<zmyR8ma-bC=EjJop&=;SY?>$E@Q|_sF;d0AzXMWinQ3H>DI}s%
z8gZAN<o`zlu-pXj?xOT|>R`*mVYBMG^N}(HqLqtJ96=PbWHV>JduvSzu@&qsqzoR1
z&p_m8(NkVm0*TtZC$r33jHN&|r5D<hBx~Sb9`n^f$s1CaClVK2vv%Y11p0A+?ho}i
z49}>Mxj!W+ccxeJv1~mn!hZ6x6LaQnm8+09-Qd*}<7OmWl@J0jCC{%25tzxHrgs9g
z$A~A!Dw5j7^4=rET3VbiwUHmoYAkB=bt~}tF~L>-EI|*(M6C3D3N0#|08wDweNspz
zOizV*DLhP=EYICh!nK>O$Z)D&2W*6K$>}n1NX?hy%~7R_<isyZu)BVX#GMFqcIJ+z
zLgM$&AQjBxR-9h>G|ppQ)IRcvwV`sJ1BOPh2TonkPEX%xjjISKtv(cGtB{+b_mMM2
zdqCSaj!iDCtdI4ZU+A=L=6TeF2@q*{zc|J*P}Un{zI;tQyr;%hrTkT#0Gl~NOaGOt
z;4<a`jc7MUhaH$-omP&nS=01Y1E4&R=C9D6_ux==34;L24in$f=F^~+N(z@i-E)$F
zZQJ;fBxTANfkyR{OCuO@>n5T9WD`){6-o$5P=PlXWvV#jiR|z?@${Lj23)^=O)>Xe
zJ;0$uX`F+zx1g9$Cw)RHA$5BrR(B+a4wqYo<p;M|T~k7h6@By@KRw4+k|Y{@Vk}Ii
zSey4ld|UXI8|udq(L97C{N+&~aJqcRglU3h=TYH+4GQ#%C3;o5J$=eaP&gVKtgLh2
zjlGO!9`mam<;%_f0PH9llM(V9oR!S-z*(q6I0hVE3<_o<SC13>FSJ*?&g*Q})cf^D
z2R<(^`4`sLy6@GtzIq8sQ3QPdssn#g+}5}p54CE7Kt<sglK=X<Y1z9=>{OrJ0GQGd
zuur2DlJPwmpaZ+r)i8xK;@ofkCI%GsLn|+SVe7<YEi;ID*BuQA$j*lv_3$Y0S{P*K
z^xJncrik}2Mv5*fbg3PhKUz5;=*xq2syE{U6{aaaM+@m8h%19<U3K6f`g0HmOo||0
znJfoDFpDJ2bjL$5^HxbfNh9AsQFGi`$9tLVydhlv6Ir=6nFqQ=H5nb~MZ81>!&iRi
z+eDz5?zqnEvq=PrFWe*#-^SLA0|6tssh)~kf(c<9Y4i2;K#SP)-g3R|fPZ!{_kSZK
z);W-~YDFFDuw(5(JR0J1rsFx$?pl*DuM7jsxH)P+2Z`ZK9F$Zb*Vu`r3ufBv5@YV_
zV16O3ftW8V02%S4>!!fMiV!#)%6C-^9=1~j5NMj~%e=AM$aj>`$<b&vDd*W~3QEGM
z_1)2|6nyNfF0{1XDWWZwNjGabRMF}<2>F7~jEntG-MCEjO#i=B%gn+2Uv=Y7YmVD6
zv%&RzP+Wle2>)g%A%LydxcIHazs~H&r0r9_X<8@$LYHWlSCqGLkLGGFEKr4gipXgh
zbHB4*pJGcVPZ!};i^(RjA*f1c`AGfw>^`a3d70_e3DV>Z-amQ2-ly7`U)%RY9cxg0
zsCT%!1}51^uwxKjg~C#8RiH6{GJIO;2DnS_ZQW3xu(}P;%OQ!Vx4|dw43-^>X=!4b
zRFFaa8=-$5Hcw+^xx@R@*_<R(?>9$J7kzU?^7$db$Zv0VbI-t|b(0el=VHvOOM1Uq
zzV4qE#ag6B?s34Lz;D0b9Wir0v+Vw1+aHerIJH5==v|JZ&!T*x*e~Ze@{0r7&xe8K
z9bwu7iO+JH`?hh`i>Z5(K1DI6i3#i8M>Lvc2`|3HGGiUUQpOZ|W@P&mG61okhPAes
zV_U?hEyoj-h#!5D;m`935HmU|5H!uiOoh9?I~^pmHR<EV<L$(8SxpXDbUn24cl!~~
z1IaytVJnL91P&pKu>v~rQ0D^#qil<*KM0=(Ax=4R#i%gx766%z$8!W7JLQ8qY5{LC
zOir^GB21n^i<|H@kRhL~bGmo*h`Inm?PXy^azn6$H%N7m(KxA7Y%o?^`OA<*F2^%Y
zqbR{5qF<}zu}Gz1umm8c>~;5eF=b6*LafCaeGyM<GS~WmW-<L(P4UQD?gi>0jr9B5
zBk}lWZwu+g=j~V1GXjcO9^)pH4>mbxOQ^g;G`=}EtAPlo#UJ!&m`Q`2(Y=*V*-Q_+
z0>k?1Q1FOyNt}a+E(3&n=knK$s@s<VY#EE%sc02GMzzHGC$D%VuD0vfYHXIFSSuNr
z0O}VKvH`GJOGI3W(NJWsS*_&QPCdbu^fNLAGSmcaz6w(*yM$b4PwH|++9++AnX@Ng
z@H9?#|MQ)tOzq=;bxF2N42*4Bv2qybcu%%EYfP@O7hzC1PW2&M$>89_$j7j{8tfNK
z**T2JCp#<1`q&a!XY-h#lAye?+B;-nkbz6Tp*#KjZ7y=>QqX+_celiIbFcavx4;y8
z!an6NM}ShwCHAb0E1)r0l}+H@M`C2lh0wN0WH$(K6+NQjb)UB<9c!0GCV~Iz1*oN3
zs1__weXIySGQxRcw%7)XRXW!^jcR2PBli&48Z^7PRY@!hp;K_HJWQS12;O(S!M=T=
zPel!vzpjzx*c<+*W{yN%9hpx|ze>6W(MFcU91E1CFfHIdPND<z9LK()Ofwb*e8N&D
z#mVxEom0=<9ma(M2mEd-qXGUrD`Io=ZbAvgLa>ugx*MxLbpjcvwxFGMg^&3qii@ez
zcdw?<xZKwWsrg}lzzgp7j|Jt~d=iS#<Ld6~lkDfcD7F;Ez3<@MtuZd|m*;yh6g5{6
z4xx0Axql_ifY*Q(YG8m^O9DW!gN1L_kad>16B_!JRig0X9sL3<XaS360ry`5d%X>%
z+t7$oC7D*<k>%Gaw8#fcn)Eyzkq$MZKqbXIQ^w*ojsfU$I%P$#1|4;tj=TD?(c1V*
zJ5+tx$Wu2<H}OTC%FX4}DFrzEMxr3e%oWs9V+;GT`M;*~>Vv1s^@RpcPYQj3%G5ZC
zXy%4zlao3`t}L1Rk!7Z1<U@t20WWP!f>N(t!@Cz}^o*A3Ppg~?NAcG(H)!VVC}%w-
z-On243!2cjMzyJ;WE~_z+lv3pldsA!*)YNh)>o@cD(S<h5d)Wuea9sYsx;B2=mR8N
z2T1VtC>^q!*9?$-!XS(rLbF+<{ST)lZ<uwsJo>8D2;;#o0UhOIEN8^PYDRt~Y>98l
zZSoiMZ<nK$7>Z*&8n$5;r1X(6<&}zC`E=RG79u7vsR{4+u?!|dtyX_r#5Zb7E+ol7
zhLl&rcn0Cttz5J>W?L@CNc0J>=&1n)=u%<c>5ZJRg#6iFGBX~jgBJGev59Lja~c^u
z<bXUJyF|miYI(_PLu1(Y>;+yG4ud)STg8oK1F7#7Rrm)*UFhq+L&ba@zKm@8=b>lJ
z4Ji88igu&3s`I4W<GLpnw#LCvgZRq3rs^oc&fO67E8&prZB$Hx4V}ENjH3j|<QSJY
z>EV)&lsFGAeXvb=xZO`HT(fs3z)H(w{9#za>7SM$0hnz`w2YIm;2FVk;a_boAS>G6
z4aE-uc0S6ylJHH8KYiKSGH<2eQEMpMOZL5OrA0q5_ok9_ov|sujyn~=L;}((yiWlE
zfEELROL0}}*ab}N7ml}hNZtPo^F7F~gqmtQ)=raks<dA$QPSwHrls{q{k<yur(^{S
zaZeiJEnXDybzdm7$Prg_#5Olm1UMF2lY7;vGzP-iDLpCOZc@0gu#_(-Wy*r9(#Opq
z%ID*0^dtl&l3&82Btv1rlLiDd>UbM;!D~q`&76F}AtY1O>?4TJ-{9rw^vyoV-#`J%
zXGU?R3?f)z$>e#|J)(T8oc$)Yd2K|$6d#VDd3-yV>m(MI{2X*~rggE_xaGPje15pa
zo~z@?lThT+8?JlH1k<z-)4*r<PT{df)NB-^Oy(q7W5B$9n!L45-6lOYj7P=h5JPPq
zvPa&AgNA*JqRqQ)51j^sfUEzHg<vVD&)24UTTFGV${v=4Rw-betHo9`z}B^5@qpcS
zD&ul@oJj?e9n6%<tcquiYGY0+_03vj(Of>6R#8_jq3RT5Yol$9t^Ck_o)rNJ;UVOo
z^`wt|x1_k4K)A*#95E@H-gvT{4P_$R;M%nkOsryc!BBhQ@E{vAPZ|-)9&sH9Lzz&M
zzg0dtywL0Qtw2#ItVx-k@;jF7m|K1pUI!%3<ZR38e^vk;6L*3a_(vtxWVq%Q1~Brq
z8D_Q?Z%@Z)M5VG6by1N<wiAPk1;CBFYBWp7Yb1OAl}2uK@H-R>yjfFG-~rQQKPQ#P
zBtN8;%yS*L&V)@YngG|z61Zw<mCu-NImoC}UKda9Sivv|P&k%}X8KgWU(RFX%)z^z
zWM^HPwmYwu6q;+9ejJ-i{1u%}nh7-jNOFX&&V?NFD)c;5Xa=w4ytMl#H*bGrSnCb`
z*`a%T>UJNN^SXAnb&Fa9a-*SvCS0!7R6blWRk}?ZL@^04Bo^Cy(}gNIIjAXZT}Y09
ze5cJ(TCQmIQ#xodPr*$?(tJfZauLz<lpInYF`v?Hp)7j(puN(9V}Xj89``aECl3WI
zS}1B1^Hn%N`aEGdvONW~(y+bI$EEg%A?-twCq+03dh+_`o{s_D!|zP?x}>}DiO!Zg
z_HQ*F47+E*U+oKd(NH^ICfz$*ZW?vJos&OblgD7O&CU*HT?Nsf@B;e>Q2(K9&qV(p
zCbQV+|I1tu>vt}m?Pf=JAI~kKOxsVt&!rXkh&BEHrh;6=sh{6+NX;tkXq!A~ft_2*
zNReC=2`8a^aG!UFo5zpWp_xkHlUZO?aJjvB0C;wv(Ac(^-gN*|V(eZ%e4h^guSVGW
zKN=w$B%CV14j{XCBh<h<6HhA{y{LxoFjbJLek`o3w3hzh3xs-Z6}QUSz=4FNvLC>!
z3wy55*XMaOGDem=a!<(Q>+b03X09!fyoO{+7T@Rd7Y(-@Qkv5NzyLkX)PAimo{#Ux
zC`8L~I`9z&o5*F1DNsKVR`k78h-kR!1gTZ}fbsFHt}k{77vGN>Vlo&*0Aj)(&bD#0
z0!%r^pI&4`VZ9e06d3y1>tC2GE~p`T8ss^qvQR#M^9cMwfn3_eA-FRC_QKC6V|S_l
zv%Q8R$)`KNQYc^6-R%wjH{TYwE=r7xRDlpvNXAZBMG21n#-f|*)-RRn7((8Q`Bkz2
ztGCkb3t~cQ_H2SqDDyrYJ0ID4JDX%f3^2)9Vk01oq$K06!J61O|0ND$c${)c!B~6#
zG4@Ns<Q^^!5y^u3k1NWwIM|A;OBFrO?~T2{hm3?-96;cWnKE~(O*32!=qtR<)fICq
zc%{tMVu<+6Y6xFTSo!*ul}9qsF6SIFFAww6<=w>1t_Yqm+$V(tp;0U+992Mg;AN69
zf$=P@p;3+3Hk)w`Ny_hI=x_Clc`g6AU+qSwQ=aJwl~%;)?*@A9A-}e!6|x&n3V{Uq
zND~QwsJ9^IqdJIs2WM(8O~MRV#b7w>Cc_^EjyOL6cw>3+u)#^4vb!fv_nX)BU=xIP
zqEyu8XYgg@_3oCWa+~f-K>mD9LWm&TfHtucC3tf;hDwRh;qS~-s@RJw+!q1Q$v2y=
zVxTFtoN~v<L4-*xM7-E`0wGFWI=sL(FVq6UH5jVQxJaOi{n(M6kAZyy&)d{o3~L{P
zyMvl`W_1)YG{#1C9K){ui<mmZ@VD%(JhImD)teM!xJuCj0^j#78Sgt7-gjVFVDKkK
zvNv`0ZqOI*2PXX-N-s155kQi(O$%rW{)}Oc9(fH)P;xdd^FzRD2nHuv&H{tlqG6S2
ztN02lU@!p!n<s@@gj(tD{h+Nxjf6F;a#Aj}!F9g=NV)6{+ej>7u=z@PIOvg!_`I5c
zy5=aSa3{qZ13HzNy(-9cVQ*qbMmBN}_v9|(h@El$Tvd{m9m=T!9k!W9hUMW^@X7KW
zzqa4rJE{s_WV%%Ijl!dx8aG=wF%i(dD;0eDsztbNQ|OMqGDX-Ag)w~QRZ=LATZF%C
zUX<-Zvt~A|kDDZ5k{6MCrUDgFjr$n$LzvJmO`~Z!kW{F^HV*q$WaaN{*P~+9Dwp~X
zeWZ$-EjB6HO{AJF!%GF(+gPgOyU?g_rHUNKryt^)cLMs+nTy1MeoVC(P^{HOBq3mj
zkJOLkSH6&yx?z#r7Vn7n3tw}fXnWI$7dJbm8)@q%tbrycRv#3~KJfWf8k?x_{Egpu
zbF0Wqaw)2s*BY_s(eFTjTCjtYR3mZwtR)ZauQ~kD-@pmi1K)2H$whwA&^(v&0F2`v
z71q=pCNqlwN&#q0k}lqprvy93{Tg$mK_$UFk0}~<ZyhJ*+UQQyecjt<vTiT7Wa0;3
zC%hwJX0YU(<pjxrG*dUwN5^6AVl&VW+}<q4)t8+x&Sc`G`toyzZvPl(dv96oeDcz1
z`&0KK!nc4#KSRZjwq+$wW^yGfF$mv;cQvya^xyteU1%lM%C+`_ZL!i+GgX!`jeTm9
z<757*(i;(t+{GG%OPtv&JMyZNt$5kij1HO*@L|~ObTJ5y#-XktD<Nj2)ZumhKW|Z4
z3~7`G>bcQnHPcfaCoG^pLeYdR-oz@Rj+9L!$Vq0aXuvXcUNl%sgx*NjMKtIQ&P6)$
zY_GV^S}e-N%<lpm5W$o}FKirAGfQ<8b6CH;#cV9#Y6s~U632%v`_C=1KTV|erYh-G
zF{hDL(nhbzlkJ^ovoQt78tw*I)C8_~62imHu7o@Gs!mDgKASt{ov=0HDH5i3+_O;i
zV;1^x!$iqc2NYfXWmL|v(80HkHEi3}Kj|*flCz1GQzE%Xih9~um*=jPHrCBq!W|5p
z4!}gPpIr@-jDS8{$=hV}X_9s^N7qk!xfRW_r5Yd9mX+QYRj@NwbxV8xN>V0NNh{%h
zcxi&KnhHX$h7!}HXV9tWkC(<0je<M>b~X?!r#Y-CZCW;W{+n~u5b4TnwtA9USiO3Y
z;zk&9Z5YTvcFmVro)Cv|6r2g-2-sIp)zP)cZaP^*VM-`vcm39*bton?rF>#-(XmkB
z>U`28K7=FH_J~|3-WWAV^^qjuElHgmcB24klyUkh(^6$E46&h_qmqJI26<pU&X0Hw
zF>os@9~y9<A3(;&t9XPC7Zo-YJ=*JrvcE#2802>FS*vdahb#W9a5_YCFl)EbAF;78
zmLslYmHUvg#dSnJ$;z@5V-xY-EaGHzDi>j2T0s9$MDpVx*i=$H^@*&s-C6`LsM!Q~
z(}54us2{Z8=qD<#$+N>A4W`w%G48!N7+-T>>balkf@+o!eHt12>m$?4=KgdkB7Y}+
zcBX`v%IXonWPnA$ZOJ`SVnc<J(ZR(hYZ5bX7I5l(B9`!>zydRrx^yzKAU6s?17|vA
zE@gcLuKk9udPY8+(Ka~hl`kJfv=1!v@-)sKl579ZVE2o2#w|?5i%yBjA9t*>r((U7
zJ33cW+3$lg9R;Y-<#wwQr6BbACm(@F#u942R!24LOFv4)OJrtnh~qYZl6X-?wk-K^
zP(Warv}(V#lx))oJtzs8H9{k04Ta<H!Y~=NM{)FiK0xACa(Q&=6;&@J6Ri!`M@OOJ
zeDdoDt4O0OPoFb#V*8fk`NgTz6};!7+_n#Jp;m+FS&nz{eUZ;~Hd>#J5H{4`fwA<L
z;{9&G_!Zdu)6&C+a96r@4xChLC&C-L^yotW9>BxvM0*YnJOJ*~fxpYNQ5#kZ^cW={
zOJy;Za9wpaqLj%rnigOH)#g--DST(rONpwFymc9^18f5j&e+GrWi%*(B_#~E04J<K
zU>0<phF6X{__7J`aktairev&p47<~tz9szr?p%Vsr;ukLj0X=Iy7(X_X)DpT>*p{L
zwK(J-l&<(y`YST*?WRg8pH~2T;-P)8iyJmdEe9dsfOH)}U+l+5#(n25FLgBcbvRHa
zbSJQ=w5$)LMpq7$d(Hk^Z}%GgpvVf<C>#Ty&2`)Qmgi41%Yq({9ZCq;BD8MD@KPdS
zPB{^w?5dBQ+P$59L<A+Wz(#7{)5CR~4-XFu`=()hcny{2PlJ!002vfaDM|NLhJd(O
zyP#3EJL9Xdw;}nKm;acEa65mGG^xDa_!G+!R{^~q!XrzlLym~yHWwR|36|rq28-gZ
zT^T58u%P@XjMNi&7Qn_z3X=v#h9`Dc)gmz3Z@*Iv?+s{!LP!F&!mUY>kQC2B876DV
z3M`+n3=z@|G!^UkRz>P3FB8t-oF?hRrH34vv6NCOPHq>1vj~kxU4>nkkpGn3GW-Xn
zll8yk4qRi2$5^)7@0QMi@=hNp5zisum-oWBr53q{sLEVPe-G@uzEjKu?rhNz(Ra!9
zhEuO_$4|@l=DePNq+WhZy1JIjj8%c5l)mW7@Z+(6fy&VRGMoUIkYsLmW%z!{12A7T
zPl{QR;(SG(lB(X{g#8N;d29fubY*?9y+yZfV2>FiyJt|}nYn@A)sRGUyWnH*Q4h_6
zf5eColS2Vif4F|0z6YUYjqFaMdjF-~pa1^TmhRUWTx?`M$LIU`fR7wfO2Zy#FpIf#
z_s8-5_!y69Da9Bq%=ovm%#{L2#!DG+llGKLILVAy3VM1WqbjOODE4+*`y^#0CsdbW
zpVg~M>hoymcf)~LMITMX+B2g0>hW#Rb|?EQ!GfXlTyCZa6aGmj{G80)_2q%J{eJV-
z&*L9NuqT$k6Tsb)UH*obMd)M$x$o>eRFpzdVS4AGM?}`nac|MImcdSdCT+(ixYsQ3
zkm%r7HLFS7C@eMv!fg&kZVj5cNEA>iS*yz_{(y{53ZbCknZoluDD_<&#WSmOdVUf`
zf3L0@l>OoPn!G%P(vzgV_ai}e(gv844dA#$T7w!`98cbaHy|<@m>i`JBuRAe2<kKo
z>;t)<hnQ|sR6pB;4I$%tC>bck@mPc)wU1=YkQUzVwg&|d85G0PPU<O9C*`?(g$_B{
zP27h}t<o{NB%#)FivPr|hf;;0U8Fm;g`G{QrZY#39bzd|vf?Rpp*chj=+`*I-ASlX
z<#58(CiYT8p&}$T7Yoj%HqcMU{{s>Mr06VAnJ|47sWnTTNKD2@!9HY}QCW!$3o!XB
zLP=u0TjD^v6RWg)wjtMbadi&my3XooGD`7T8i=?YxjO|^zKY1>p^ZnBjem$+<{C|F
zF$R{5aXR8eiZ^B5991gy0B3snRcBo`fkw(4ky|HFgcre6v0exMGfCu{sx{^?yG%>S
zr2%s&!Is^C4DlN-_i!+F&~?CC51rS{$D~n&$dzs}Gy~a(foY~Ox+Qs|x1nOjGQu26
zG!vPbYp3<O$^YLa%fY4;*#dPg6IASy|6Qm|o8S&LF-|WodLW+;wbv|Eryt}?YQ6!h
zr;&frTXv{mPA!W??GL%obVN$bitwzri)k06?RGRsdyI}#5ArIBQlkZ5EQ@nV(K@3R
z3~F|tzFS5MGZ`Wr)e9hV=1V+Pdj~*XjNmECig&FHZpD4k<dSJ0K;E2lx^@d>xkzR2
zZgkEgN4Eo%wZI{dlw$w<CAEPe<<4=UtH{x=I<@m}tnuF6d;;#qw!#;QfSo9If=f_6
z@V`I)J{5sx?umTd*utXIC}x5@6uR@1j5IQOJHay({0<A$5bN!X+-1#i$tluAK0Dp(
zYTfPB2&n?VCX0f|)*Ap!`o5Drz(?Pj3BjSrkcJ&Iz#~X8uVYl^N9{FB)Hw%HxXbkk
zaf)w6DR^Vnp@x%4<p=JFBdU$dEO^Xa7rfA0Q8VNjj-P%5Y-Dz<&w)2$9c+S}n07j6
z1=A>q^FPtGWL7NB<4vmlpI{2V=p8#GG*#N>U5_(f&yAJ$_@g9T%+sC))1P92S5OrJ
z9vxdw5LhJ!8RUKqsTV(u@SZQ1MA_X(!e^V4RF&I|aWOur<#;-)Uj~TDNT_B<@bMu`
zcwcTpa~|$zf~@2skXyr38%hik?CX7ONl3>kx%YDD$A&MkGlIN)7HPnVQUh({*V3~p
zSWLv>Pdl)X5>c9*CeM)Mh?}P?6HUGvyCWIj#W35ap)zKB3NSSut}tWXl*9ieMlLTn
zmxUb=SC2DyAXim$mXtutRZ5@6I?4R(cJd@VWNSzlAwGt#umR!|5k`*LN~q2b%%Bpv
z0nL3aD-mRnywS_@j_eL)wY7SOZ<LUY5tMUa<DZ0DtFX@XFDZzA6;uDqh#Vbh3Hfl6
zLbzJD643%7!PlPAclRkQ1bHHp6BFvw#XCf|Znm&U!hyYj?`AE5v%mqw%Rhi2#-3))
zIt(HF<gn#W(DOW;C$mAKaLO2MyCs{EsNc!$8=on`0j718Byd)d29Br%l&NB8B14ZY
zC1O=^D#_$<>_!!I^vIf8pphtI_yt*iksbK9aTE${DdYm_a%H{|E{1$+l1y)`bKkl@
zj%FxfJES@AjB8p#-c&X=Ko7#CP*?IDz)j6N$D`*l8DApTgbj{Y=A@mJ6K7;MrTy4d
zvn=F#o1|LksaDOY!Uf6Fdh2*9wq4lCz?gf5RW~Nj>a1XaO@lz!eynU4_9T0eZa=k@
z16z&U3RmO(QV5dhW4NC}ti%R<k!Zt)tkDL@w(}x+QsoD;aM7WBI79v*(a~Xhn^Y7q
zpMN=o8cxI8a29PHV0p8vJ>iyo>jW2&%BV@=R+5uHDgD`DP+N~NdB7=+;c7nc(yd*H
zvAFM7yX+3e9g=*&ow@O3?nu7KVsj*9rY>nbE@sVTNs$TFKOZ5ZiN!~FeTf}n+*oa+
zEo}jainx;8#pT&2V<Jo!nEkS;;phSf?*bQk<+VoJxG2|ZjU?k4_&3u+FK)cedFLiv
zjx^}V^AFo{ixqlsfVsykxosCDe@T=LQ!i-adZ(eX%459_PZcXj6iuh1Uls*khliK}
z;;oxk=atftg0+hIKu*96<5%BEmV8v8mtMp5!liOTVp-q0TpoP5JqM1J0~f1R-;|YI
zL^95jxybeI6G{w66VZ%a?evbqQ7MK5ebJFe+WilO8U-v-EeD~eQXP3ZqD`Tm=`81`
z`#^F8B@4qrH@Vhd=N1f8uNb>oV^`VV2P&+igOH0z+v1LUbt5pN4)PT(<zLGEA}hFT
z2n55_5|ns#j8F12$H3u-ZH?!!iTc7nlfi+kQXDKbC`h?eN{{GG$PDeY(J*E+Yh|k`
zX9^Ee$W*}wF?@2BCo~2vrbUnmtKWbMn$@N_IPWO(Rq)9iVc-I`REAX7(#<^5y@C_W
zK{*4QEw?Rrb@MeLNorOeR52=UbnS;J-=q-J*jCG-;~)++)47uWrVMY5OuBi^trl&_
zqj$k=nyT1ya24M9J-AFllO*NUpGu8F`$!bSut=IHPKu@M{66!6b=e*FF8TEurPQ(G
z16%*~t<0*y4kc>d)?LLt%MljqZB^hTC6%$sJ($0(Vq_#npEIjp6DJNN#@Yp(dl>E4
zFKkSM&oNrD{PPQKbZrKPMIo|7rzX4LskI?=Rvqy5(E0jtxLIYh!Ej)rhM=#u!0(EN
zABz^|lU;4iK%KeU09yQ~7SMBZS}hweKH<sl*rgny&KC8Rb%tE+ZP%_NsF>_wX#O~=
zYvR;_?awSWh5rX;xYfm01DC`^aarRhQk1lbd&2gsJRP6wMGiK0*%sO+TcNg}5~rt9
zDP@IMmVV3UOM0hP*i^>TdsoMmC@{!5rj8?iJ2r?fjXM8yJgT)NW}+E4+>(15Z7I^X
z%jM1=*~6U;wnzHZwtHd>9MKb`Qmi`$#rYGU12g?xJ#}L&lfKv;$<yaqrJ2u0T|yDZ
zYMvZ(B<dpBcBU#ow5=N30Jk<N;S0V4=0@~Crn3LPP|Lu<_TMoYwlySUH(3#SZfoc8
zAO<lUUIEa+tW34>TT48l9V{}nGX6TA@7~4{QiLb`t2?%y3slvJz=k{hicgwu6=Ch=
z&eqA@(M__cvI9ZOT&;z@10VMla45_l0fH#ZbN|}$@<~C0{gL@xRG=K@o4k$5b#g21
zvxnb9^Bd%<BzWR0xqSvW;rPe*h&<3rKN6yX?74P_OR?BraT>#^absQ$hvH`%`nW%N
zIZBqELJ-Wh&%%B2qob4i?Njkbcn2zY>GS)G!fOsH$YG5gWrU6I+p356O9Pok5o6Li
z4--Lwl}|8KV6v208Jt?#9A=)sVhpA(FZy0ESXR7(T6~%j{|HYqWPqzSJ8V-O_dRRl
zT5mhJVtEQ=N$Hs4nYOzLZH6r*ORlSADcm${(IQ8tvFu+k7HuMMGpfcqBmLc(22g^t
zV9&9%8LslT5Fo4u@4r`85x@QnM$OqLN_XxwM7+@<_DQkO7e%sC*seMV5I0T5R6J65
zf%BkbLmfApe*@e>xl-I?^ccZ>>v~``Q-dFosfsHBXThyCIl>=C6wG_q8PoZ!10RUB
z0;c}5D$iU*fwf4B$)_LfEh1`M2}YKKm+US&akz+e)l70;HY#-qA_(irBO~|X?C9wx
zPf2@8cnR~yqAWb~)_dZffDTsxSApcEZ;S{;Ycmb?GVG-*6C)IGu7;qQLxB<klf=c~
zZ&{=AB%y-FK_*1AChj>m!Nd?=y72lrZ99%wK?h9d!QHj65{mQJKFD-h{%A|V3KnGH
z&@R_MR5Z9_P=6~(&k~90XaufXLe*BPPNmq0y=Vr&hBb7ZV;s1ziCXuR`x}dqLauo=
zV=NalTO7^1=|7?=oM+ELWT_!AN!*kIF4MF2nSyP2|CBm@@(Ls@%0i4tJv{zd|Ln)(
zb1*&Dew4j>@0KFU?9JE}gskUZs}D-6s}Bg~{j17K&TYDvVj=3E^d}3JRZ(sH55|7m
zLng^2)IX)gY%0IDEFo4WxHKS-@2rSivvb}z%3su-hDZdJkw^V;@ij0jSzX#f43DRT
z47wz6qUZ2ax@VdWO!%<sS#W!m*t}$}1J&7sw#y8zWDJUNM5<E(j+)v0`t*YKLI#q|
zq4r5r@{k0$L4!!8CU)8L5^7s;9Bef-!x?-;j^uWdou~>f_)3rtv-7eGa|goU?D1t4
z2U-Q?jQ<=RI_<dDFN1zuScHl{nmTD7>VTW|OQl0Z27egR>Er}}<0fNhs);=rF4-T8
z!BM<mDk`n(wM`C9t>|X~svGr|fA2I!U+nmw3P5Rmw`gW2z?Q>%#0g7of0Jpu8WZ_=
z?P!uR34a;rzqE(KCEk(H^0N6EDEMbi{w<xLqheT!!O$6i3kl-ozsnHKKvgHo^icOl
zE|hV|AT?n#wEV+kD$#wQ*ZY3c4Q<j@CV=RTA+j|+s8~j|Y%(Ezs9i6t^Z|T!C}Aru
z*q7v7h=^p^WfDI;P>$2@32<ymCzh-(DHSgkD|MBQ+o+umJu2a3`~Si=l-U>eU%qbw
zEt6N?-?pD~8N7!VQL+7FNJg5atAo|=whUTqSiG6G5IxL>(ZQDw^FkZ+7e}rjdDQ3y
zJ$w0{<n-}S_HqpDnZlE^IASD&Vgd0c+-rnhX6vlGI+F%0Sg8AR!BVzpUr4+ic$qI~
zB{wAuvK@!Fsarx?{YfH<>dM##6C42I1Wj4VG~lJR;4}8B=YrTNHat*cAM@1B1j+TL
zCV?2u<->g-C<R8ICfa`BlF6Os?OsGlSdQ65NvbIFlYIdw%gJl`NtVN4%ZRZB&$0!^
z<4&NY7_sj{;y~IPNP=MNiZK0E2lrIm7BaTn(Q@*jH}_rv+yYpyd@I>P$d;>I2NEMT
zWC3L<bI|t0tAmBz8osomD}5T66S$*hURdQ}XY#kqnl0X&WE#!i>w6lo94#ckqLQvE
zb{P$)-<}$Dwx5iHEmzWInJvGc1HH`H0e?F?#zGgEcpE(hd0k`(?0ttq^$(S{ow@7M
zha_hF*jXz^@p(a5m4ScA#zZCKZ0*SO0iU;NKS_!U>|$;hTdRpxIMEYX9*tasy!Y7C
zE*zMswtONbc+vD+u7DF9l7S{(<^j#$8ZI$buJ^zp*e48dK>8tSr?}tKo%dKnLNP~f
zoprQUv8$0UwW{_4S^H(Japb)oPP`18<jO3?9jW4aX|@3p_B1?yynx5>wKqJ0`b?0h
zI1ZNpHh4&P0n9|gwgF)@#nwhvzAhzdIxp8jzr6NKuL3qJN-b^26l^<z8of(43+n6O
zTFyGJ<FvE_W#R11+aqMdiW#|EYejTsx#u<`CzN@Uvyw2TlpP^Spw~}bl&Q<ux6Yk>
zOi>Gw&D_(FHc}GBwU(^V@L&$40EG1AqVP&<<i_qvk*K>3XRD0Sc{_9BJf8)UGmDZL
zo*1+@JDr#Z%fb;kZpv$}!mk~QD{NK9yDGB(yyk7p7vi6_NID6?uj7qi5x4hHZ9ZiE
zYy#|sd{hrBSusH9=F3Wr)a}jXuW4S|wnACzfU>?0->eRY#phU~*CRV?I_Xzag3>nI
zOtU$%^_n5e^t!Nd(^d74FrwV{eYG6#Os$sLl;!*Y!PlW?{ih6o@jpyhGcf*lrP^z(
zNrxlWo15A>c$mL_0C>tZH5boOrIf#Yd&LeDHV)(136tO70kZP&tr`xlaOK!W6(mSQ
zSorvt`!53Uy1BHrvb)`wm{|`naG|ZcU~zvmz5`N1Q^3LVs{Hq_H}4)BSImw(#)LJn
zh*A?h>4<W9G42?Fk0ATb6N$neXOFwN{Jm)(b&Y~vX>087%cYsLx@6<tj;0<VaI&YI
zVhBOvmi1(MyS#sjpfT`?oDNf4OZjq7g47O^98F&TZ7O;RfF<4k8fOq=|IYN@Z7#yY
z3I<mt$GSeeT!Iv-at1W1i2M=w{*1)^CZg+29VdO=_cgz0W!J9k(|<L572mtexyKE<
z-_SULJR~4DmG=WTKB$~>%NtvXe}b!rY_|$nK;iY<1e>0cxWN-*^7J=?U;MqG!)An|
zKzxqQNy)_9hnn+7ZK6%*^$j?;l?Bx5t;%ge8bCK1j1ezsfon!<&H~I%e}Mg%9+Kxp
zx`_=965={6yVk#wBgKmKc<_6SJ1$_|>p0C2vEhm9LM1^+cad*;$=ZJ^m=?2n4V;N$
z9K%lj)6g@QkDnD59x@>8G<xOyXug145BCB%$w8=s8c~sD5A=xSjqFPUd#;*e3lUyG
zdI(?-9YEMJ5XeFP8y-{TY@7oUPYC?w@ovPzdurA&W9{Br4~cfPkQL|OYlL90`-N1?
zLV63yJyc}X2in=J5@cd~L_ugI2l$1E)n@Z7rNe>DYw+ADuob2rwTpXm9V9@2)sJyh
zoXf7SjdDw(skuNw&iS-(F1=J*6iu;f-<Zy8%Irl^Y4tY!CXs=WD{2ihELB(0sfQSW
z(PdJ#06{Sru{o?+SxIr%Zk3>=AY9%-R8gkA@fo0(g?F3zcRh?_RJfkNplO0=M}6oF
zm9t-6(gr~Tx07Ohf;Y_#i^w4&G6B3jDu+~??roc$Yx;=R!s(D5`9g|qf@eV6U-@g3
z<D(G;*6hibimKw2$?!^S_Fr@XFLeT}Ja8lt6MTb-cy*rf%v$i30U}Q!*#Sv~>s}=E
z-D2Km(lASMnfx@G<Uj!&7OC|=xJa(48dwH11%bux;l7CGv@9ziq}K{XmJi0w!#3^G
z^nhQI69coN9aJ=j#H0NdtJyQd-Uj&mRkfR%1@lKX^RIjpXc8*%`59>DQZ;pE=rZ{$
z;9xOtW{q-(m`G04@%%bZ{C;JEu^p<gvW;&uxe#vr<hv-49G4VIio?W_GGd?+p`-NJ
z|C*E|>7ihN=}!M=s-C7-+MbFT0y*tzi)5^&2%V*3Bg;QUT^os!V@`n&*|NrB^TwGk
zV^2x3Nd#j(1FV_jK9#Pt-J>AN{<V0NLjM-iS&rTQrd*c2(|lrApMAWIN(Xfl*P=r!
zH9FUm5@VB2+fDIg#L%3Q>B#X+@WPL=<>5+6&^~*sc)-#ZU>_8>#;*SQ$`MgDiPA)F
z->B^&&~vOtuq41e)dZJD7Qe9Iij4Hsm_r3A9e$OJSUZszk?bF&Dq~B^@UqnWE3J2p
zFzC%MfTi-Ft#(hY5;j<tIIM-w_ltKb=kv9>9*V=`2$bfI)1jLqp<!TxLM)9_6VdVT
z4A6sC$?Nde*$eu5YyD%g9J3jHUr1iQh!S_QEswrwh@3*4$Cm_nV;Tg1{2}}#ZvQa=
zQez!|(mBXKD_<rDV(C^VP)8Z-eqUt<rBQpNV$rG;B*7Uh$0dXN{Js{7qJW{KSaup?
z7m!h1*ibAC$V$TU)W*7Dvr|OLoi}hR>->7STM33bIj%@Wi7Kld?(~t-TV0iGP-4Kc
z(SYZlHqcQWr!+xr*`gpz3Dp!(s?&g0v7iB~^YHO_`x5|Y0$>owTqQ_aQuLIH6=(hN
zy?Zh{y#!K~$d<R3?w~u=7iMyTE10;hDb>y8M&eA*Da|QN#Ywrm{h<dHv+d?vXHSJ4
zE--oI|6%MOf<p<nwqZE7?PSNco$T1QZQHhO+qP}n*|BYXdCy<<Po6V4lOA;UU{!V9
zy{=m8f??P}&6MheGW|=OOkW^o-W8|&{Q+vO+>~AmRfZ*N&O)fNIcN&<EszHF=)*@y
zGnJcxT>sk}Tw~5XZJ3Ax@(mgSDBse>ewtyi#h)@sT6$9SGszOm6P~!gJ%E9-WZ4`M
zbrCvCQibS(X3}M|jS;ES-3cm6UAT0}-e_0)+2QA451_91v?JK^?ygjsUC6pkL&ou&
zP97fI$bamhQdYiXNbhr+6_$Jq(Vv~cZxpw%Jo3jgqP<M0q(wOrxc>E4N4xdf#rgZT
zBf-&rJRy;|aO60GUUuk=u<IQs-YiZ>T_xRj_iuNkhf(vNes4<&DIr{nO8ouLFUU$6
z<0b7I9Ed{XyEW5YQTx7)oK=XL)PIJrC^KB+)4%!0h3UnaDK_%3qpTm?_?BS=1~7T~
zI(KQB`}ZhQu-8Tfk23T1N%jbRn>`$`{~kt7vT`8aFcY%$EcpNY&cUXw6iI$hwWFNP
z=MLtjOa}ZxivUwS4g>yDS!}WM2^`yb?C+h%aq}rdGu8yp0X3H6|D5zl6FDp1r21<x
zd(8yA{)`Lrg&G^ukA5mR+8mVB?J53=jyvQW$K9gK4V0&%ulA;{u~`ei$Ptm3W;2U}
zBh)X|?z3^v;+Q7hrF*A^<msQhFkR80ar<p;T?mo<?CQmnXcmLOJ<~|SU<@ygU@}T;
zn}Fle<|nn2C|Dks#Kg|2G~WxY6)R@H$trBDz5_U15ia8Ot0lXgKjS{$AHyfK#vU{@
z^Pl)s$TC4r|7N8ZwAw5m(*pV_TNEu8J#p$KvoJZdH=YkB8h$i1*?r3WtBaGSOVDk8
z?mUPYUvj@p63d=@uqz|zM|~ceOna?{ljXC{e*neJA_x8}GGY5Kf&v!y|4C5LtvQ}>
zz>3tps(K3!C1M~d6iP_!Jc-y7U1|fD%WbGz0)H_1`E3*s9=u0dQN&v|M(8he{Z=Oj
z=L_rb`gV1H_q-^b*&nIbU%7uU3wRqm)Jt%e-xvsvATxM-=lyyWnKZF~6TnAQyz=_s
zhJR}zdxHhk0^kK?=kWqo_LaW-f!s#7^&O#_=ejhysF<qZ)*+j8HGV;t+;P9Q;Q1E?
zz%!8J>-%mp(}>}|`L?DB02}-ynjdTGLwG~Hc0q$`U;fqH(mdV~nK2ZMgLy1Hn40V>
zHD>|dy(z}$t-6^^_<j-lZ&J<HfTN{X?Y(ENXnnkBy<sdf*!J!{MzGHmO@b8PHW0PC
zJSdr97}GlOH%jyr4uSU<?9)4HZlQB@mYKz}XhTjl>dIF2iTFi<X|V?nRq|t&8;H03
zqtn9kl!kwm;z~G`x4a~4>)1wbOv5761!MKlFv&~lhhSlABPnMv_+U~^rpw#}Bz;S&
zcD;9_DKVtfQE~yzgGIyeEH!c~t}$nPW3+U7ym2DtQE<W>QvyUgr4q~NS+@dSblns`
zX18(_j{vLE`Jc-E%QVuk)$~9N8Y`uAHVL3{W)mgu&<GrDKn<d{`xX#eHMkuraDSTs
z%K>fJ2PB7>55UfyYTUv{6PwXni$vjb)m)?Sf5g*TrG|~245_C^?#vEo%Q4m*QIRZ8
zv-$R;mupc;XZ9I}o#KeN!|d8V*WVw{ch9@6Jt^b|L81GxH@1&kW(b^R0v`lc7~kJ7
zw<nu?9t9kO(Baxa#e8lrFS>tU4pnJQ<O^0ML$7(nl?WJ6ruO>u7^q}=({Vds(*AkK
ztO-S3VCoN=NBR_<*>a1ulv*xKu-P1JG%b|`qHL_+Ef{*y&@ImO*TJE1DnOs;^UOBx
z!<5tfh;L+Zzcr6<s1HyWSD@TFbIfYeWi5Ndcgr~*TFXPh_&g?7$n3J0x$^@N)q%JO
zv4PGBj?TTvaj7UFvqsv5s!c;d5OBX=;C5Mr3V?XWrlO>@ce7X372nPu_a8a0SFlFp
zY2Qsx+CLfTmnJ1?slaY_W4j=}<Cua9v+jAETdD}Y3g^=+3IuDs;{8s00~7X0WmCF_
zK)O`u^IIFlG6h@QvlOn7VX2fo5&u|GG{2x;hw+mW8cpS2S#T$-b*xl0W@9!=6~?|V
zGvr%rDH9WxON<_*xI!FdS(F1+nI(A_Vdnm^jA;!*`P9dg!v-p_rI<9HRb+&qE9KS_
zsmrC5eeU_X?@-Z9jv#!Wi;07(lz6?#A(W^i7=es6!u|&j`+iM)8}u9ARND3gA@C{W
zc=^}W9;+bFYt`vn%X(n}(((&;rify7VCYI=NS?sUxYPlbw+-(kn6f*I^VyoE(*hiF
zkw+dq4EXQd%PHYrm!fQRTcX%XTu753)+yzxDGs$FE28OEzLbH&z`rvA+J@}KZ6*s{
zcUW`5<OW@n!S9Ep`5uW5biJdTdtg2H=?5mQ@Ph-?xG0`r#&1x9_Chakv1<<;fN>Gg
zd_gx_vForMz=TdbJ_KTH)F72RfZvCoYH?P5ksUy-0r9Xaz`(a4QCspGBV%86v%_P!
zPq<Osiv6YZ4KXShl^Ri9IouN#Q)n3%<fJI=IL%R7<RyufTK9@iIK&zDilg%q^Gaxo
zHCNcB7|l~Y)X;LF>iuV>OZ18}P~^D!O0yG=_U2$x+vMQ&<)nNIk|Ms4+{;R6CKlX&
zax5O@&rI7sXc?eW+KAVW%vKax>2F7@THche3;httPm%qp&l+l9P`AYlD!8xcezh?T
zM@XqEF_<jd&;tZ^1Sq>$s+6&vsS5&Wd!EOvlpf;Vi+J#P@G-iLY8SEK5h`Kt(}3bl
z0<!$7S0wI;tm+}SLqU!E>0xR6`l9ZzA$v&#61%NBLW8OHz!@MinUhK^TfkK02OVN;
z|1_>Zgg+2@JSlI$R$z;rzBFElR#9pS>==A3;F{W0xkL7<ua_?mkQ*^jELg}_f`OLG
z-M>l4C~oei*#vyDtQMg0_udj$Iyx0etNEnJQL`Z|6!Ka>)iYqPb5y&O>j>M3hCQ24
z?}$4~$Kn$sbjtjG=9Wmq!pu4fyIK&WDw%not5==g`jjG(m1buo?uOFK0x%yAq%Js0
zjBpC=WYCZB1T|Q6OAO@OZZ)PBMB*E!MfgnBH2!W0qc@&nhv$wmkYTu*Cu09Uj;B35
z?kRD2(-$cx97WWEJ33Ed83A+A9ncQl<s=pLN{HC^b0^Ms@QqJ@i&D=|s)i5U;MhLj
zAfI`vPpA#QC6abl<^hOaU>(j_OZ>PBl8UIm1h>pTnHk~%eIe~kqxmnhN(%|LY(}*Q
z)eQX$l&F#wD1P)mr^XviSQhb?@Wj#P6l5MpohkYtA<qi%N#&P$n2+<04kKfT1s3&w
z?0B<@t?`vk!AUK9lWrGA>Ur#|s7{=)!_8!)!X{S>v}Q@%R~Gl?m-P0h2K%AX_0|Sq
z0%g{<K7B(_;U0&WLDj*ER~oF9@TB1(DqLlYZhKV$BUyGx{R^DUP1EWH#~PY~HtPXF
zbz)Tb(%fvN!BWC%&^#RvV(B*O)$0sT2c+u^&V2#LSC$2sfg)=fB+8Bnmgq6*UZT$6
zMl^SlC(~<W2CbY0Yxph(8GE<ge_q;pTrT&r=h#EN8*r%w;bMpECzR?z%#oQ-=)6m}
z;2&CvS{nrqx^BlEME7x3T+oA<j_juO)A5h3MNKT@#Ju9CK(+!imF9>T`KC5wrbM6(
zb%+_Wqm}mc<MEBHMGbfa(^tc)Xmkz+3IFo~B@{AWum*f$8+<9UdNXYDik}DHAF<>9
zCWN5qoQloQ$yl)udy*S7&B_$z>2_mem1Coo<5p1Wa~HU^EheX;iE*BZcM10e10t6E
zoxD<R)JB^J6tB0oVSU`Ft5%xurm7P$(p`H3rV#<h&kgovA1;KA=fu`0K2?7~B$C;W
z(M8J}(vU8sK}zPBo}l%dEwuB8U`-z;8Fa>G5Dv2_$4OQ7za{oHUSeC)0{XlmGfF(I
zK|1f_e|}!{7FBsq5^<&5zrq6wlW*5BMRQy^&G)@eIgQDxQ|fr4Xk3)1?>>(C$kFH*
zS(sh^gcg@@_%%NCL#C^lcfP7mj)ehj4k><5=zDEk{X1%bCVL#U*LpbuR5VwWU)NhV
zw<?=wf2N1`;^^~wnhd`ZWlMHcezidN_1v*ZPyO^3UY_8=m%f8=)ZTK|X*o>%4rHp5
z`~g3cAcXyIIFISS(5V>zSDc46o^ZkXD;CJdcY^`ER@Sa**q$w)pb_6y;4to-;*z3$
z_yA7WBa~>cYW=|NAaMk&NA@Eq^(E!_dUf#decLa1JPcynv-9}z1j>ke#3zQ;VB!GK
zN0qwSE!ov67LxQOd;eWP3H!$E@bFCUM*_)G`H29)>1KtreUCj{13V|M=AJ?R|8rb=
z$i|NMmMjBfVpKRwi9qtU_27DYe;x_bqYzCmn)13qP5;U<?^ii{(V1EdNPRFOF~mM>
z+~v;BiiL|zw-J-u$8zWu$$UUb*T$->`McsxhV{cW!Zc2)<%Abfe$(Dg)$_~J)8s-s
z4+*!S&RnVv^%Xg`%q$P7XH1zg(m+nmEh@c?EXTcJFod)^r{^K*4Hz|}g27H660;GP
z$+0U8y`9H2M(wkzM8DvIFCup^=v}|1GZXAT4h|1y2<NE`%rE(2GtwoL<QCA*l5aNE
z4xl#?=5;62H7;k5s8m=dT6z3vJdW0C1n62Q>)9yfJU!1l(NxRBB(jk0JEos8j39=n
zsIcnTT0pBe2)a_KnpCc~EnL>!9xs@OsINg>jQZVmXrEO@hA%kI*0#%(Anq-Gwb2eb
zu6q*8n#)A(2k&c)3^P2R=p17X9&;rme+Uk<iiFE%m^$6NhA$GXja|OJ+P^+MeI4~e
zr$&N-u=9Q0U6II`feN~3B`pk?GktxmKE6!O5~-n;q9i;+Apgv7_Wno>j?sB3f{LdQ
zSF@Oyh!c!=JsM>pn$=tMJFCxFGpm0hHNXKq0^Lj1uGd-`X#yXYPFy6=HsFr1@LwEJ
zWbZiTsEKJ<RiR0odrV|_51(kxYhbli{biIO*XCK$8ll$3b=UkRNtqS#IjfSg+#iHM
z*=f=C(Zc*}19n{t45YL{ER9#HZ10nKTO`4jK62xW2R-3<v{ceS5a_TMh0$$F$2Q1k
z{2o#E+FaJ+h*Nq3xMUGALCGfLTAY3uf8d;q8*A>dK-<k@nQ7;qso^k6T2*J!twwVb
z+H8}5U?WfV?-Ehl0cq7VYcQK0V1MWm4-l(a`!&>?X8}5WZ$whjvdrn1&fX@L0H=$3
zgOI*p6DkpJm7|6AHQne!ChZFa30eR5s0r<PIuVF8-w^tpBg7%efxL4*UfdI~oo?;;
z6S38W53RBSaiL`FT=Zwl4PK&PGz!5Y#(*<J&q<#<L|bR@!Nf!&G{L4S!MD>gA|8U)
zE}DDiRz~cl2888WZq2yu<g&WJar7Tebz!49zEb#<ND5mSD+-K2Cp>EC&ZSk0pn)Z*
z+%3c9F@Ye|06T8RHwm)|5f4_38>kE5cvbs91C(((JvhQfhN=)g>y$wC@8)#kCcTxc
zgZ;`TD)I6Zd+4ZKJrwGMv_;I#`zZzFQ-`B=H5)Fc!I&&kT>s&6dL>{BldryUE?xo?
zseEGzn$gaeB2Xic{p8-j#7r$l8MQ(*gujZ}N$am4g=i-fo~%ShYxLZ*?Pbsh=QOg_
zJv%oQ0EVQ@1&w&NgzWxx{JbUMci(Oy(spv?2aCa`l#qW_azOO<krkQ%@5zeIB2dC`
zUXT@-jl(gjb21BH`EAAKLvSl(H9FpR&=5zU&YsAM4S*l973%dzU!;f07g6UV9A#m)
z3=l!9JPRyw|M;wFM9>(KYO{qRsnKJy9F8VKd^Rz<<P2Y~5<Qr<{SB7Ld*r&dOFRPo
z-AFN4Id~N^d!&9fqG#bYBuAr@tc{$y;~@v4RD5_f$y59KtpMJykj%^&sa4~>Y@G(k
z)g)uuiudkf`ARY>B{wgj<dqyiV)jP+J)w-K=0=;0bxFaMyG~j|PmV+2G_KdDSG#i(
zgBFB&?Vqh!bn?rd=2Q+Jh^52Qqs2lp<jW`jvj?09LdczYc<fP#6UpvLLYVqBv0Kb#
zenX~->A3A}der0*mHA2ciC#X#?Q?n|>*L2v=>=1hL0O&m&gy4H5CQG59{eJ(8HzeQ
zAh+rKodba|4M;CP(fpt;nhW<)NZP1^4cUSwbU(D8S(VdIJNRysC7p&O^X@K{81bg~
zW?rMk9LSpjbCFxQ3Oe!LWj1yhRV^SV{PkP?v@o6)jolVaQtTF*89eD%%)MCYZ21^l
zAJ;FLR~~?m{MQlo2SfntLkqGH{ggB*>AJ}h&J|R%4ZKky{XK&U>ML}Na3bsMrE+nC
z&?t@Xh+0dG>O-!tcHsU~|7Xpcd$GEsy`iv8blJ-uE45YpT|fL1BI6#@cHhwo6KDYy
z>z%#b>ukF^==>fmhsW=Rwq8fwLR>7khko+h@R^1)1<aPE?U(XL1b#x^EH|4Hm=Cao
zxImB>@rMx^g^SLaXdS+=^>0fvp<LXPQA3~|nj-FRIQ*4P#_xDI)Wq;~f+bg|5%SQ;
z-o<<8@55l*)tr=gWfvZax1Y&XFwhtM^zQ`fZ{tM#)X7{`mfV%$_BJzL)F$3ELD7fY
zoCjOj=0B%~ZzU@)g8!%LCB{1-qm!~zsdi=xQS)Q-W-`<b8M7bWk6h?shAnRG*|I!$
z9-TJex=Ku$>gjM(ru?dho?Bb-bVgM|0ayb};?~(MGUS_a^SLY>T*yzGz2Ur&AR%=|
zJ6c~zPx6V*76+5`!%<CJt+AzcQNnE9v8?m!KHas<;&2sk4|KO;5JCcdF~_3$;GJGL
zNXZ{p!H#HjCv}!h|6XdETuD?)Rb*46@UH!E-u5i_&I^}q9$Z5C-qS4QKWS*qT6)of
z!*7K4J!zUwcv$$T2p(7watic_KLNTO9N~E?fI-W~mA$~$laHx3*Vb3h%)d_NbDgJk
zvlcLb;X<t_P%D{I{c~6Q8d5t5ZZdO{_wi2+SwUB}A<B@VfZW7&*sw$*%)66#U@a)J
zxGufdVhPvFMmo*=14-j9Dtd&H73-iqG$m=TYoQ@>E!W0!ylg6~GQtXyp_xLjP4<4n
zeZLEvWVc8@3b}ol;X#OI@}{|bezuqDF7bNw2Y8Inq~gEgqyNj%VWj^bIXcrC?J-Pj
zFxM}?9?0@S=U`}HE?>34MUr*eag<}Zl*0Gl!sOwcrXwfRaRYdoMDpBx3R&@?y>R!p
zi<2KGld-17@G@?|aPYai@ceAV-ho8|mT<r3$DpCJzd?8165%PlH;?nb_y}LLmq&a0
z4jeSc1Rxp!$JY+)-zQtoO^lwd`*(N^!p$%XZOjJq;P7sE0|)2RMp<ED`plUapc%{B
z!Od@`t{9r(@j54snf3lU0&1Xr?O6Pun5cM#MP89%`67Ag&a^(&4oeyLY*KaSPwCI|
zR!*+yCEt#P6niqPa4sYg?-`UrUS>i`CG!o5KZkGTCyYd2_^*-NYXk0DyH^G7#W!an
zI#R{R98nk0p&POHj~NyYB%Cd)g7K9*pgX!LUrfIC)OV7G_c*4NsyUddZ99axbixBN
zy;HT4D!OW$1#x(h&576XBu}`i)Vj+{;ZMNBj+x5ambRfDe78`E-s@wWY}e=wvQQgJ
zZ-FQiYPM$<fkT76hvgTpKGxrnW<i5?@`RHrjD^Be9zuD|zt*3M#J*a~MJqr~<c%mF
z5-WPk1K6<!_4*5bm^{26dH4mjtNIzPCK9&J*P16E)3mH-IQ|EEb$Zc~syr*UA_m1}
zq7k>&xUzDP^hWUXAaNm2_q5aRR>1>^VKI%3@FnD8l|I~_YQvhiAZlE-GK>;2uGA3y
zNvv_3k^<x%5|Gy`N&;Y;w<M#6+Phx9lB`Y&W)PVf<}O{t=wMW4cTz<isfPMUp7^50
zx{(Am5dS<<DQhNX=sWkqO}ka0K#MR#L=n!keVl6bd18<2Dzs{^)7#GU9}mV#3=y6)
z^>uMpL5KvUcUEy+v%G5$WGY#7Toeq0@Ad0X%%nF3S}-qN=g|Ujf32=Z%e9|mdZhHe
zK8-N7U0gSxo4e`iShS@NVSBL5D&WTc<_^(mF0213?q04KyyQ(s>#>!jyRfT)PI|%1
z)Em0ZK3sgVF<X&xbz)njGcc>$etNA8myXaY<otJ3diE>Ko#dOGG7ah<#Q)9h)35gP
zLWbY)yuY-2eVHA6N!2wesInQ*z*O}7EYPKxg(vEDNXI1}_HSYpR7D3TpkROOHxnkY
zgbj3-x}KieZzC=%uaPq$48zY(bv#?M3s&g9oE$Kkts1=;&`yRC0N8_tyfT>=iFl=<
zqL|2A^6^?kjoL`Oi8nap!u6uZ3mplFau34`t2pYQ<xJFEi;_?wwd%*djKZ<qU7Kl<
zn3oF?46yslMK-6rMXr{-=sjJkzR6k9>_uV?Poy0#c@K^HDTUiZ6H%vzhu)Y!t`r(C
z$=XB^;PgO?p)s#shbG9~uwyb>fJs>a$ysT_j!wpXAoU$(B66-<oNk|r&EhDjMbl&q
zgX`uMT{xNkLCS@2qnDb5|LhBO@v<`*d;6Z0SO_^FUyy^bwVq@KiBpy9>b@|};3m*8
zt90<jGdLx4U8BSue@?d&Pk72Aa8*Sle1ds|)-}2KFdizir-W+RQHJoorMLmTYB0#I
zPcbTm@;pF(6I2r5Qvi1|uIDNj_0+3#3vJO|LUUT2H0Qz2LL4q-PCgW7a+cFKL3q`M
z-BCkRpqU+2>~Ct3d0OJ9nBl-Z)2Z~GO(^2ui}W)n2G2DCeTBXe9#!G|ks3E6>h`Xh
z5kkcRv}=9E@x5oadWeju2vl;*%3|F%-(l|{KTn3HkmUkPFQlMMCE?9u#hrDXMh}op
ztntS-t?dG<i_L_iG}yT=Z;W%Aw?{cu8~oC{N!_8V=uG9NT&Bh^8Gb;*G5Ed&4jNUm
z>~!15nfW=|s9S@N$}o5^17}L)NS>43RXGClpcL1o7R51{<tm`&soaL=$Bm>fI9P<Y
zZMbW}Htt09yxq`yK!CVfi!G9MU_C{n^MoS#wwyWQnhln$%sjPQI`O^SW#@=0;uVQM
zsfgQ4*b5lRs)9!rydJ+iHK{(|{>nc&Ce+4zB&-}%)Kb;r9G=mER!cJBPWf*?_&F77
zX92yUV7m_NQ4*7^P75mD{a!p}pNPN+0mBUKd=fxPF2IaL_uqNXB627=#H@892BuFL
zlp9&>50vtCNP6B!zz-OS#(gr0secA3umR+%`+V1~%wYB5NaphYLK^My(qDX@w`66?
zc&}?)-s1S|k-c3TNi}O-8OZkobjC;|QA2Ct9n1TK6l)#YmC<dKju2HzY8>9Y(kSmM
z*TQnkLpx^_mbs$Q8Jhy%Jf8z<NZ~&GBmpX@zPj-FEV|Dg$(rZD&``DvdBr*C)`DM;
z72evfQosnGr?RAtQGeqWha@oIEKiohllJ%286+#6h|z|-v}UX$G&&HFT^%Td$4SjV
z2rt*)_5&aUpqpMni_t((lo;UPfu<kFWNS;JK4$vcKQml1@w#?g0A?T!j<FxO3^uz?
zT9+ma`!x)^nx%F#5J%sc@p4%8&Btb7Y}0Q$26eesNm2Lx&fgqgWEum3joGX>%WL^^
z&z>W;mIF-n!@NWej)lWBMKNe7<z$*#1r~fvJfz1~6T9)=m7v|vw<8snqqa=nzA;;H
zIO%S#iht8dD@zf4HV=AcHbq*elU;Q;pq0|$<-XwVv{OIfAM2i(dw#QhKH)LJtLXu2
zUQY~Am#J0ut+zA2xEp)?oK_3*V4NUsnP%;Fyvku%pi1?1?tX%s?_rEhUN7`$RtEKY
z0G7GScbfhxysTJSD>hQzIOf7wltTBlyq}LbeZ6|U!S^-OPX0HP!Tf)xFbwQ$9RJg{
zZCX<@_J9?p=S1}tzDlK_480DbZhQ*mv^ugKl?Qb@=XvhyCsd)qNb`}%r9jH_mR+7a
zqUiON+|%gAm8?@&CiP|Uge^K7d50W)`{PAs9zNllQao_QNrpnd`FzLMsn<6iiCP=I
zLr$Xl%;9l;aXX98{3u*-K1|_67EZb7C&Sm()VSB^ooZY8lV;{OQkN78osVqh`0_C|
zo{PP}362QdIodMo$L;5dxDiVM1e<RJ`{()H=lx_Txq+Nn@8u*`5l6H-=y^`vI0&}l
z<4yJD>k<2ev+9UBy8QFRl(<Z!$GFi$*d%!cOEFzpyn)4K1uXwSwopAf9!XrC7#IVc
zFPBdpGdfK~=@aa3lf{RM*~$CrW&{SMmPa^4G}FlH2lL>oS*}>uOd6OOYt?7c%Se4y
zXeM=_TP#WC7C9!>-*EpifucagCe6?^Ly0?)cTF-^MXCAQNgI6Qm1bO{P3lS30pysl
zwbi;sbOc&DGxvpj+;!UWeW&8^6uve<CeOt<d5BjV^<<se8kCc7%*69;!AB)UR-o;i
z;c>I1)@U|J3fTjCZ&ioYpLpq(%v>{CstGh2h}o_k^)1UoeF<eQf`7%73E>?D(bc=c
zIiABuH-*Dfs(`2REQEBSDj}*N^HtO&xdKt)>_Ib5GQS!zSd<NU%~Zf`rlWhoY7rHR
z>D6#)3gy^EJ-chof^~oZWd`FfYw0BVmIMV)G~}CWqI6CRMY_irjSwxJc#8m7UY1nz
zs00nU7kF$6M@-~w5AJijKxamyhx!wTaaNtk1Lp05X2xSlfvXl==l~~6;$ND{y)Fnn
z_U4@RK(HM&8gn3sNRKrcwsH!$P}K2iVABqBA!+l)*&#pN@e(9wx!bt5VTc^N$Wfkq
zg)WN-cvj#kHH%2q%nKgK##q1Fhp!2J=?y4wV6?c^P{nyY5aggnLr|*Zh3IxW+mf;P
zes<Ofm=R|c<?4&cQY>ouXs=9B_*BK(;?ji76w5dVprxqHF6yJwre07E#9*Iny>!cQ
zZ}?Fn5W2B?n>8zP38bKuMsH0K3-u8*Y>s6UEGU|;##bJX{bEnSd}h7i(xT-a1&5g*
z@p9&+GLWwz33a9&dGZU$>IMOtVVCr(@K`1Ba6<OwHi;Ff7bS7hkx><K81kY3VC95q
zvsI198Q<-yq|@d&uv|HcdH=Y6=BlJ9!@@N^Eo>zT+7rX%1p5zNr3vfIJBe|geUoBt
zYE^TT_M?ONG@}o9Kg-9?_P|qP<7x?TeXmXl()#=%1b49;k;V9G+O)nZ7j{LPG33qw
zF~vM89wzy}xBf<e6Xvo2_+`Ms0lq-5ds#Oi`0V4I2Z$5xSxwTj=xqED0c5HWvi!d(
zhj17JCQTE~QZd#917+|8h(%lLZJM=e*ZhB)A@HLFi2pz<iPDSl6i&zMO-O?VqAZ}|
z6a<RP!kjk^p)TTq4umbzjY=n>nq;XZ>v_(mYTF5Yd5Fr*g*9Y-Y{(oTWTfknr!g6F
z(9E$0O0J`@EQ-!4%t&iVb*IR<(`vIKzsd7Yx8Fx)CfD;DkJ?;Oxvp~0lk3}EHeO+@
z{1ww}+*++hVpuk37tmElE*E)!e-rZ2Nu<|j?<rn^GWK`WnVwkJPRD8Iq1eaIwMp*g
z2%sk65Vy<f)N)Vw({1NKJBji(5$X@ipC_{Lpp5kCr?mINAAP?RZ0T}#l(qLf2NL7o
zIMAL>(inS{um+B($Y^-0hPkYyA7#&#Ej`+hXHwI?&ABztDhHMD0N|!c0|9!()&dKN
zck`sPX+%70+!i0Cl9xbA){2m02auN>JMQcI<11^KSBxxEPR5M87j1XIp*C0&=-Bfz
z<oLGj2WS`U<5s*5Fo3LcfSdt4;x5|QZ^CTB+J>_^4OYlIXXUr-^xL{%N1Q|l`-flG
zQh;H?^0xd_0;gYJQ-I~d^**Ch0&mHaZ|3rj^s-ts;d-n#Qv&UK=N+b`>NXqz<XDGW
za7CvZ1qS<rmM-cHqzb9JMZCJMFb6lmo96MxmFbL8HUK>8bmp3^8xYA99)P`OD^%A6
z1rS{Gy1!3T(~pgjFB!x)0^F+w#AaQ82vX?$?0XJm{qI518Pi{g+;MqOXiRU_f(0#l
zs+K>xV3eosw@7SxfXu=K1;933pNudn0??uMAnBFlxxlCNH~ZvfFdg6mcy&=f+rVRL
z^rVcV0jlsmi`Excc@Msfy$0F?>XQTjcUp$kqR!x=hKJkm4X06rGyp41YWlef0=HOR
zXr#;gVHc(k%|m8FR$FcIn1tUnHtm43DdY^eTWscsP$HG1R_invRHQ%6J{H)BSr(hG
zNpe8j`ohpNQzvHtv>2GJp|O^~X|nOEvA<JhXkdaL>N2`edJjXWB7zC%`9Hw@x??k&
z!Ts<{a$g2c$+B3MZam>lVev~+9S7P*szs(9X$~FUs={r9RXl!dr#V5dsJ~rdI(vI^
zzd_5d!>Rsjlm5S>00uVZ|4pv5tTFy;y^hfRLG=hcSf{3}<rf0DxeXVm8Sv|@1e(q<
z{paB))DE3QUPt4cfMcvdyj-p7Xqm{I!wdWH_F@0i=k+Xt8ABg43TvkqZjY<z2jmQ!
z1s;swENK7h<MtymN#f(;k}d&hUU$T|^?PQN4q}B5um^PSbr>P&M7DhwN1*rcgzg(I
zpRUVLurlU1QIALBxUfG}zJu{w+>l%h;D>eX0RN^J$0#QRMJBX%pY`YI+2{3xJ~xTf
zKO%&$7hjO;4@8s2T%RBphR-L97tbeWs>W!-H*Wv_pzp_>bz1Xh%Jg{=Tjf%QyMf9P
z%WcZa1|*F3c-0}9unl8*ZVg%ol9)&6tyX|XazbFfv7J-`G5M(yZ$U*OTCGL$goxQl
zyVA{=O6hwO*nUwu-}H|5p|jp|s#K%Fjak7Kx=-a6+BIkE$#<TwS>nm<H^Y#uTULWw
z$9v7>qN1V*qzZJRBwty=_8G<GfG$@iMlljOJ_~~cO&U6>$nxPf$r%!(h;}3I#;ZRd
zc-cIMxw*Csan&z<%%68?lDJ!l*Eko1uSVDpEpPeL+I_|XLn?42wrr%RSb5OxufLAi
zKjwCG0GwVk%t0Ur;`F%!9S7=i6ynSK2SeL1T4mp0w)$}K^keLu=-$D_RQgxz)hTA$
ze0O{TmH9uNK~V(EEY<9#%Hr|Gnt0ewpmH#882-a$%R@eoj)_yH@KIjiF!IohEdlw{
zMCqlX8e%Hc7+c6GTFnZ8W938Db(&Uu$sJ&wBT%Jo3{-3(wsXOk4;U2Q8k`S#coGQ&
zI%e5=p5g1S&t1G?g8Ku$w;QYA)dZb$IkSj|uHsU0IfU`05X5S`uuC7Otym`R1TGS0
z18A^|Pmf{`NqQHM*~Jz`cs3!sw$fK4k6JK{+3!#bzCwvX4!n@T@<i^Z{tj)uF-3wp
zZ158@7(5=gGt`A7I-*pl$AYwuEtXQP3JC=m>5svyPWq%Ah@O-o*No`D#80TTrE|4M
zuILDmZeV2symVD1%2~Egm5zp*_OyUh9>!K<S`zB{-SL^oGg0CclzXS0WtaisELB>0
zs4SY#Jr^Y<H2-$LGNVNl==gRt4yC*Nu0=Zj%8V+e<aRWw?yIV+E&EJqshA)T<?NEl
zHp<3PEOrN>Np{z6eJeDAZT_4UNw$TzRSRb?7%7Oz?S1F-`jz+j#<%Ju5}MMeor0ER
zCn^%LSR|{9+O=9ot?j*XeNx2$L<xsj^wf1HBj&O>6CT-P<fkG^z=mWQ>;<e~Fs?XV
z3ZQe;4yAuRO!$6Rhyy*NX+A+-g!eB%mB5V10*(}k2u_A>D<G4}BsJjqbdA_R?(N!C
z^JC|$`*rFCu~<R%1xnXb_Ks_L$WRt6YM~q6K*H+zYt<kV66)K)yebPz%#Gd8L}hN2
z)QUt!Rl-w-#Z{eiJ}|scLDmmKcU%|@DOPOSDT=`rpk1o4WI1~SL#&O+_mV!ur=ns1
z)>Y8Sdds*F3Q6Np^NT=K)g;CxVH&C{3J=;k*Hyn3`U0l?vo8(~&dHXyd(KTG{J}uY
zbrfJO70R{CKp7!QWcKA!6Ex?hpShH-STU(=>fMh#%o%0SlhV=7SpxE;{<boXhb2|T
zYm#dmJ*J{9z8{uMQ3MZ0Z-iW2<VgR?mIE56+o-m7ttd%GPGL98+~e^D$>%!6!E(?n
zi2pcv5zWx7$AtA{mzM6>S-8Y&U(Yuwa1gANnI3YPtvft2kLsKwy%4b>zNqAiF4xsL
zc__`|NVdw%lnUu++UBi~V`^VTi+y6tzc91;7cFXU;HcvV+xgX<joV3`Qs@0unuQkP
zdC?1Gz(Hfku3HJn7HcV{R^_P{Uy@8Us80gh<jox2lW=e7LermrIGq?SyQ4qZtXq!+
zXCYLUrFAFwoJGJ{L&oQRQEm2@0B6%hwy`F}Q6YeGFQrE|6;FKgtwPG-5Z{wLOo5&4
z`s!xjA|OgA4$GiPcpWKO0!p0gI+fxqsHsk>VmhH83lQt4*a@^SIN^~`N_CDJCEB!`
zvDic}JGO)(XM{*<;%D*pshk$wQ4E=F#7t%zUo@;kE8SnHrF8142?j}*7C6N^$+BgS
zg4CV@cVEb}qfmaP1<JGSGU2L^IBEGun;l5dzvAB!uVwRc@iAN+NEwJlD-o$bkU%bw
z_%~h*{Y`eQH6NpsOk5=j#dSvVHm|qiU!N~{ei?viP}`tt_e*suj5}acM@NAX2-c<>
z_szBw`kMC~UVXUuBSYs3T~4YCe8vyZSSFX}f$qibZ{IPKm)4g-zFfNV!~fPBS^kTV
z<$v{`?$(%&!RGk?`A=7cy`#_3w_FtgPtkOGV^8VFZyUj#_0L@D>-vg|E0rr9QFeB4
z!M16WD58kVABu`Qyz-9i|29ak*W04$J>j~&<=andfKP#czHw`(mwSLgi}c*RFulBD
z<v2bwuM7$n(fF~mfW22|Wq$}j^`QICPGJ$g4BxhM`*={^p?!eAM7+71JB0SnIP6=<
zYVk#`Cy0t)Lf8SGL)_l4-ckdpEyNIUc@|WEp3mNHPx4hKMCh6t23J#ozx0?PaF-C0
z?Cf7Jb_ZJrQ*+EM>jBId&a#@3BPonxM4V`$r1^{^F_)$(Kl4UN|A8N|#j5-0*I9nO
ztiI%1zI~j?J5>h9<e$LdpE1>oT<!A<otfU3Ss5D2lP17gTivav5mwGNa;IYM4vI-S
z9Xk|=n<%^`fs_OJ3H34tHUbK}Qu}pDm9JTiKo@$WF0u`#eM}n{02q4INYG|Zc<fi@
zjORsk<fo0Ew3J(wA<{xTl&3{*IJ44ZE_q3qNNDq+RW>>+8}65KD;X0#crv7+Bs?}f
z%A!4y(Ug6MmXfP_D{|?3>4B=Z9|9ZO&`3!+EpM(HDaMQ&3G*$Cju`MZ8y8rO7@@yU
zn<i>UV4i*$BCe3ia&bb!Lp#R%$g867Y^EjhXsYY~Q8{wI_d?HU5zVkq(a6Kf%t|#t
z*<)kgo%}N=0pn5u?=Jxz^@;<WKFF>kHVEpaqjx4iE++r9x_fv__j!j-TvGPbEA*}q
z8b27jkB?%+h5_Eh!|3u_{fTABL5}#B+Da7dm$KF9%NL1F(s+ogpG_lU-o&%pEEGVY
zy1y{+<nP}(B26>t&~Bs)Ad!zP8S$<bqt*nI8*vjWA>CD&VcDvq`OGNz{#udVl{y?f
z6Zysx#M84$mw?TmtPD{hB`tAxxID;SZw{0z@!544mtv39T^xsoVV%ynkI@VrZC`2#
z6_q^M#Re*WtY=EgNEqR1pWM$}_=TtCEEy#gq76SrDgpvTv_f%4L=&AbWsNV`Dlohp
zY(XDlu^9_h2g4YJc$$QJH-pNpxJBdu`JJqFBd9h5X;*>&mT`~w5qMGl@;6u*#$1lH
zk`g0dFj$KK^^Ykflgq(a%mWJzwAf9VGbl@Cn29`5B*b7egyA4<>)$El(Xa@{S_nU#
zi#^kK<vAew9$D}nlUshj?MH<Tb9(8??*^K$uFG*1kEt~!othMO5hAU|xn%ZQKT{2Z
z6vg`3BJwnPiJL)F5yBd~8^75&;F3et{l-r9Epp6v@IJ16Cfp<9%R(}FM5_&D9Xrqa
zCdmVBQnF*MLGoSs5Uqkuq~0>=dVI$>>8r!WdLdbs6#;|?)+nzDh2onaj6}Z0!bC07
z^-)F$Nc$RNR62aLE3GqaMWzlgDl7hpFU5L>VZ4X9Xj|@KN{A}7rco{Hk*W@kMl0Sl
zrGgB`6dL<u?h&sT$%U4s@B~xLl3fRd57LR2Z`jB#K{$PLV9`_@IX8vcU_2cBr!)qI
z&Mx~3SK09qB`&HZu4>~dcxjCI!1AZ(XWqblG6)i&AkxK_o`031^eg+Y>|mwXR9NG-
zx^`UE%?#jbJ`}*VJkVhd#5HWxT-6n3p;nL`L$uz7E1gH1bDL)Q1DT$>D0mD_tYPS)
z5Z<*Dox+=P>QwoNR$Z|Z8-xad;`d#mC)*9yr%}o6BMYD2PAXvv)f?3hHtVcTAyI}q
zqNe;tI8Qd&u9GeKOa=zPwSiSH4Q;T+$ePVF&nwwwIV{O4BBbE&j{Fs~e+hYruhG^^
zO#mkfJQ7k$U6s1cJ_`%3Zuj|id%vzO_yqJ?e~>6tNYi)$jXlR}Ar6%DUhv<tq0mZ6
zo=XSG?{(rN)UqoM7o4r$IC0FNA)8>WAK)!Sh|!`HtbdHaa>d+}>Z>ILr7Wpp-wLgb
z=0^k)Oq7v4h-i!Hx($4<g=@u^kF&+=S8oRQH|<^(tSqzdg^-L)aB#nA*3rgVnVX)E
zO7g}8_#3qgL@(`%o0nTzDfa`I%&3Lh@luJUWiYLqe9X-sk4X=QaoWOrKc1;vPMwBR
z?7@ZC!YKa<@jEDyuS}!i<@WX@`79b_yx|Ds*J`riB&!l0b`S+qVWYX4KyxTQ%)p{p
zrOY`~e3eGR46d+{vLB}ABT`(r;3R)IGz!?AqipHL`kckO_?5j)d0K~NDAy6ttEh0f
z)o3NaO3dZHuwqCMJa<nXPcJjT(k`2Q=(K}q<HO$(wtrf6ZF4EF8Yx9cV6i0DS8pw=
z1*^MqR>ZMxuRD+)_@V+`uzEW_P2VbR3_~B4^~A5Pya$LfNOvo|tAwY#ky1Rh@v*MN
zB}>T4*12aXNrpIoT?U1)A-{GAcrpVXIW=Zk&3qyx(yT$@cT7h(>nb&otTE(qQJEJ$
zGIWabth`*%o>$-U&~+RcplmDVW7CkXASbU=NIUe!!76FkXv#@&N%iLAzeQ(bj|)RE
z8K#1cok)~S^z8ErfxcH|Ob3>F=q#hOlxPH|{@v$GKg2DSA9f`W@z2Y<&M!!jD3G29
z!mxhUeXt+Eh*I%m$-FL5P7f$;5OJ41)<U_(0uQ*Sy9$uiAT6Z|^Zd&PA?$DA+!BM;
zig5bovdg~QtojUocIcN=ti5HWiia$dW&TaNsu&?nV#0;*yK<N940PbZcivnW!bqg}
zZyTI2GBj`Xk8omlDuHk?IHvq3gH$=xHy<k~fMLELOxL1h>5>utnNug3ir|raCf1Ok
zWAs!txcqTtx802F^9K0`$D}$`t^<Ji%f!PCOK1{+P`<GHYVdaaO4jJ#8<dz&_JuOp
z{M?Jdrg)yZ1((pXV&^mZ?1dmFnuvd1H(KpHq2s__RDg4}$NF1*Ra`CC8BHueJStGb
zSt661E%{l;`wM@`tWkE)B|qkz$r6n<>P8OA)OU_@HqA43*U?c(RphRW$IqO>{!Llp
z%R`hxG$+WlpDdJD?n5Ky^LfL;78^h+I7oCpvcTqOHVqT>NijH_mGiRtxbCBz<Cln)
z;WZX1Wz>!zFQw&R5NSscx9}`#fx>eF{IgizRFnNQaURAJXJb2V9AdRwrsS5akiJN<
ztxh(Q`DnM)(Y(aX%&Eyu9<hX0Z2~grCbir3v>e8SsV?EOU_rEG|5CK6_8Y3Z{F3cy
zUWo4Su`DLq$$r-HJz=EO9|hVDIfj56$z0ZKRYg=ixIay!se)lnp523~7)neD?##p!
zugHH?(AY;_vXQy*3q8`1;6izu`g&me>sgyl1j@S&U_H!ug?MP}((8r`$r5(@^QJNQ
z@?Wvw|CJ8?E;AVZ$7RN3%mpjV_6gM|a3PgM0bU;ge}Xua)2e}su&S`Fs_Xp6PozT2
zrJ56W3`uNhWE8R{p?jI31n2YEaqh3L3C_$-s4bf{78!Sd;`GN0*l9pfFTq*S@CYD8
zWxo5*{O28Ial<3W;;2|Ti#Lw<i__4-e=?(I0C04({MWvdYC4J5R+qj7MZ;DN@X=>C
zT7n~YH7DJi-k)r=IV{ey!qNOZ2G3qkK5D_!^~4aZKC$3GSD(IaZ~Oy+4uXS@Yw<(f
zcFYhqD#4nO6bv37ZjL|p*=(F|jd8Pqw*gped`7_5({~j$c;qySo5R#KP=)!^TUa4*
z>57{aDC7IEfFfHy@M5oQNBcOp+bw^Z0F6w^{rWSTfi8zKg&J@;FGh%ElT?srcNg4~
zeHPt=mSTJkNVGdg9!A+mu1K<{=3EUsCy~_ml5i>Lzt5Obv-A6Uw*$PASdS=LhXlcj
zVClg;&w;18XW#-D20}*)&@;f;Fb5r1)zdI!e@bDQxR_?>EE@LONk|&$uL}VQ#n{Zf
zagAREu$-rirriF8wpYg}`XfWdd4joNlqwKJWgm~0u!sG@5TRBN({*vrHtJ{BZ^CY{
zyAG+N!8lPvT8Rx=1=89}F_jC5t%k9!Ru<exDmLQmDeh#*Y2y(os-~ZTKhzqN%<OFS
z)Y)vI+)hvZ$X;7FyJ=#@;KTP5^hf9kP-{OSF{o!ZotRSiw8(1XH7{kAPrx9iTNEnB
z#*garU%gMzqeFMG5w*}S&Y=4j8-79_spo#f3hWZ3QYs}BbONWqeQpW$@@B4wQM_fy
z1sm1%6lCb^S!YlMp^w#yv(Od6fe|z)4%g59Yz-^D<`J;x;+3EKg>(20)XbZa7YuE}
zSdcKSH><hy8#McbqMK@1PzCaXPzjcCfFf0gwulmlv~nt#(Xhicm*j3`zM+VEe}JQm
z7?goQX>9QiuMx~6q+Dmt6v%+9WMN@h3cNX_T!o+mvWG0KqKTgkx6YwtfpOMDabl_K
z1y5?4r$s6nZ}z2GkT0EvJ~>!td}(U~Nf9$_I(-7n8MuVeiz;oRE`=$X{RAb2k>#cv
zT+uO>MseFaj4$gsIj{SZ_?PdSo8q%zXX&`;qd|Oo%2rqg4@B0lMDMi69r`j<6>DF}
z>8~*}m&Ro&iJ9@T6PVvF%CWPND+0a5{2@x+$ih+|qEBvAr3jS#Z{--{qRMVYdafxd
z*kc5aBre)USpF(}^o~Z6_1suVien|G;mAR@HX;9RU{c?f_cElNu*J)f`pmxU%<NOf
z1Lxtn3D33N(YK9lfwNH2fHN7uW#Le6E;O)w@lZ~E-+PTe(J%##nPbO;?<~k&Cr-xA
zr)Bvt&O%@=mtg!T1RbTl%RZPu6dD#0$19t_R8T9fl1>lxp{tJ)?*!c0q?>$|hlrgf
z+##fl&_1xZJs(mtWTBYjqI+;4)SX)C$ZgBH0;iF+Gw|{`6Q_Tr0tbKILpkQaQz!XA
zZPHwV1<2{_J?m*Fu%vy|K{+N=r_`C=5?TWsyuNv*AH4Y&D!=Bea<Z>Dln`n=*Cds@
z-|(bmcv7n#xt%6UO;^(}B>S&rm0QWN=}`{029(2Caws|zfW0b|9CyFySzOoiCm%4Z
zmOD&sOx3|gf}nkk{1!0vTfpG|D?n}%v*Pr5L~dg4_q)SP^VsA~*u(s%Tveq+%F%hN
z>WRDgYSkn5o)cw`7-uf!83eXRxQwj)DhpsPbBC!+oZcjbry8x+pYH3jN#`S`QAM)J
zGrudUgzkEs$o{7Yxdf54VGeg%RNj@2WbGOv&M~EkkT#Fz{rH8vZbXqITV`&q3HahX
z41Tr=AMZbk0w5>ua!hY>8_);aAOo6|+TC`AqLE?Dh(_)0)wu`@I1}8p?a}}x?;BG_
z`Q}^AOvtveql@PHmvtngiV8z?h005Y-HMtS!bRIDkm-bNHAux}W-tDvo4t?tT`8*^
zC&GEs(X2|sUM<H%Jw7Dj%TX1UU2wKv-0BCpHv4+aH1C0Rl+=^j?Yk&vX2FZG{H@2Q
znw2iw8}YD@Zps_avJxgQ6?(3YhhK$keV)Et8o5h(VE8A!r9+8Odnv%7rT(p7vzx%F
zGVFdQtOuxL^+@fTIPFkPr$SG6$2CBAXTGBMuR?R6-8PMK@qnrv$39v9=K~cu(;^;j
ze)5@%I~B`K7$f6fhS!Z<^)E2eH-#>hdYnnP8xh40YyuQYgnpB8aEVJ(oV5Alp1eLt
zMKJ9GKX*m4oq&h~Xz&r~#lhgiT~9u-k7*~tq8{9Qo4n*I5iWm@k?~MYLePIN(NuUS
zRG4P7Fu84Thj{)ei+Mf&#_1or_=)gNc77@lL$NNwYS9841;-auBh_ybED|XRV*ftp
zE`Rv)e%xdHDdxTc4AtkdE@6HYgjGYC*ef7z0zwwspO=ef6?E@3d)xQ5XgK(7NBBFU
z3w!^qIsf)ydIDPm3n*@GIw@ltQztV5MivgH|E)RydXC|=BKFMIo^DwYNZx$`fCWxP
zxq!Tqn%aiI`{QtcTO&|~etw~h6RI1z&OY+$0~TJJBtCsss8#9ke(~O&EnYv?eb~u+
zI80&hA3okXBi#lJ{sau>u_Ph~sFRe0j-I9RoGc97fbdPXE8IHT-RxVyxk|XD{m~N&
z8_ZrLoZ<WD>5AYI`!H*VdZ(>*@UQGYwj)W?{pH?;w764FGNMpGDv^u(tD(6rILioj
zsD~aubsvUolm<YiBT~9}uJ&3~{Yj~b@KTyQ-ajXg&M)mq<Ut${cW~c7P5kqmlU+{n
zhjhL7XKZe7$Wki(vOkiF_*B@u-kINUFgA$KQDg&9R(Lgt$B9P$;KZ9^7^oPl`)9;0
z;wAyc`yCTrK#lBwL_1I%bN~+9n)l$LIQ%l0EduP(s_5b(3FZKRfC(5NwGa{9rNCbS
z^`yFb4_^*GA!;6B5I`Jo<|bj|+Cn4yjwoegUpRI4vm+LXx6oBmyJB?VTvMQhEQ<69
zb8d7M=g5uGR`wKPCl65cN~K8?M@V>c!v#qrlXsMbk;h52YcD`!UsXn4AI>O7be*#v
zVfX>Bjj}@jwH7!ipetKSt?Y`B&Vz`9^YtKdVyl2if#NC3yn%LEbkWU1oe<WxwTM4Z
zSJ=Z9vFYD;ax4qu&Q=~TX4pTxm(j)XDaG6PL;4#*brK~o+Cq|hb;Lwoq-pR^CuAoZ
za4+Fr3D=BXKm$4px~0?~Q2B{VkS2LqGAAcJP*PAPF4g*xf6asiL`$+iQ47iMey~1E
zgmM1cx?Mt9hRJZNT%|w7CUwbPq@TAPm(F6Mc#Paqg7MJW6i<Q87<#jtm&>c)%eph%
zE|flfp2h!uC!(F7lq9CE|4QinYzvp%G-+Y{L&bVeK;c4#++jq7o^a6OkT1#V@Tb2!
z8Cts`9=iWnh1N;@xIt;iH8kIW+f4x1_rNPpP41h7ER`&A%|$KJXxq8LGPL<liJb_R
zS4m4`!B>j@Vlk=Y8p#>>aFKow?i*0jv>7?Rxqe4xSN~t=&*%4(t?u_94C)WtET|Jj
zc1Xy$D+WF+=;Pq^-P?I48Y{U(J-}xZST#Kfhdz0ARq!4uE*e~Vwwx%R$U?ivCk=@P
zcFey1p$R{je292GT<Xi7`F=vBgE;02g`rY0(A5!9HC>hA(bvLRW6G-c)us>5E8ars
z=;jsST+ebCt?CB^W@y%6^1v{Zr5a<g{^PtqndpTe;9N@+<jbva5@0U#dQ@vBI9BNM
zG-K80S|VXP#t>$xP{w6iqj#uWFdYM8!Daej2^Aq`Y~`2)a~<mgSKpIFH5XhI!86%$
zlqA_)z;kgbLK<h}@a^X1o<WO~RvIBASaSQ{Ti|F#6)Zq~*AGHN;~v}pA7k$nBucbw
zYq#yywr$(CZQJ(hUTxd9ZQHhO+ugVCiTES_xMx4?ih7za6;Us9Wac+UvJCOsv4RO#
z1Ri>Cpc{4mA?k&7E5<Y^3Q~2Jua#Iir+1xF)X@&t@@rL6G_Wa@tGdPAj&<Bzy(d*$
ze3BU9>ar^mqb`za%BrrOLpdjwii!r6w49to1_Y<KTFM<ISN3DpuHHYuZkSaMb4mS9
zg$F=OG+8Sh_e2%ak0lw$Hw%`ATA6b?wkp?Pkz9dMmsDa@%u<x;xQ=u#w3z0{!Qptt
zuEA_41iMq=q%dh3YM*DjNOE0vwKJx&bA&ka%&PGe@f2TKj|w1r7@a>_dYVo48p8Rm
z^|Fp82AdQh)$qxSP3%h$N@9J<uy|`o6w@4}J&H^I<B0qAvM4u@^&M^fMDpR0yNl0`
z>Upv*G=-=g<(;^liIBX06E@)OFs~w|w!TcinnQVY(D%v{cG#3SIFj_~ZUm}4nR{G2
z4)PcwBS1LEhW1!pwdY^rd)2oQ-YuiwC}dhc^}(q*I?e!cp{c~FXfRr&LWpy|fasCQ
zpk0-J+DL-IJP5nSIHqu$L83J1_9U#8E%tc)loB8D&=R|;MrC(2(n*WklvIw5YRKdm
z1ulEUEaM{s{U~}h*orA9{9ruiG_(Pu2L>dzK^yN=hjC}cdH(zX-z0tY_$X03saT40
z#!form=Ee<B{k~eVL&k#(T2Q)iS42AgK9abh_Y1wexdDL#tmH`A9+-|V?Aq!t~?=Z
zwv1EJAR0LiS8y-kF4bTb8oz>o(z9lP^8EIJZQdFR1zTi(k@c))xvJ*<M}m8u>wYO%
zndk+W5V<C;P@%%jsxIMFX@_)+H{W*2$)Ju}ydYaysyXtxSxA`_GE7=5QO!_x^ieT+
zMsZo1?j)^Sk;5#JZF}lO<YD}%TVn2ven6JLf>;;n26~$@$S^GYO|WdxHNIe!D)|l|
z1i!#CIak+lQR>VhSj=I=t>s0?Za1|`Oro`|j*>o=DWOMsp=~`kZnHdr>3r=xdZ}Do
zq7*foTC2|H^L|@WyG;{i%8*t{&Jiz-%TQH+N&YUq)FB*wP+4<EPH1}K!U2BV51=<B
zJjqrd-v&BOm<l}P(K}De9&y=4#6*TBKfRd3V>-(E&f!(*uPPc8I|n8Cobs2chv3Sj
zX&@1fds)twUr=an@>_)@RW_GpMTm(_>H}0XYi5u<wr}UE(9S~OszNtwG<yRe<=jo(
zQPb=3iQEMO6Xr1t%*d;>m%gK^gKnwcfk7<}ZcTdxS|xZq){Pfvb2?!4?yeiN&`2#P
z99YI?!hU>j7rLp!7{+vNr0clNkOmtUr%+lwM|@13r3;lG?l-%cX1E)(_sH!`0N)wQ
z!uMC@?41|ZqXcvayA^AGO-@&Iw<7by^Hr<qrIR0w2Pt6@)a1o&boVK`s&H}wzp|XB
zVo`PO+)aqhb(%QH3+Q*Y{3UufB5Ip{u`MTC-Rcxrk$ua^z4AOTE}03{%|(>izFh0C
zt@7r?KYK5Sga``SM6OD2X8;CgAZ3r90!=?BZj&}?O^~8>c}ovpa}ZJ*te6(ZYu<+P
zD+Yv721-ZI%kznA<k(mDvwTA3r>M}yTwJzoJ!njquB5NM`WrL_KYcRXN=|OTXW*TK
zx4H!X7Z7k=l-Jd4r5$TTs!yU6hI%n5)a2}=i$i-pz(17MU1{!Jzep^a@;jPj_g8b?
ztvYErrt*3E^EaGq<@LyIv7=!bbDcRP)HPp^=L3#;T@!jF^O33<Hx8Q_Lj46HY)6=1
z!aCZRXc&>9G1eX=qj$*styX6{+6KCii`n+Uo0lErY(j(U?wnv-0(G+#wWnzmX!^3Y
zG&nDVO;};QsM!K;wgPV`(PX~7WUSO#y1u}0MM>HUr5&N6mJ?6EWLphQ_Xv$U@R1YB
zqZ3eldT{c+aU;$vS%q8HvkHaT$=x}3Fc@eWhf^Pf{e(7t;ynEiYC)e6P1T0mxYgAs
z_V*FAu?)p}W6#1QZd#M4ITowZnJ4kn9|0<9y-n$RFU2padJ4|_yqPI?)xm}0`Pn%J
z9C#Vl#y0IXjGIH>XT(@<m`}c(9!a|YAU7GA|9{MkvWLA1ot&YilCuq+EItE0J)Nk9
zqmwf}2RqyUCq)@JSpPdIx};%acQ6Y3ovp`!SA)lyjsbU4#169Bait4y^H-!EbRA4!
zSH_uxE!t#s+~wou!~T%{nWAlVoU{r}JtV%AAjJDM)`;t*fN`7yO<Dhwq{dp7u$HhM
zftzYZ;TmLUEDS-VQG}!HFd|xs!xK?l!)ObUnA$k?P}e^(!dG=UzH%sbbleco&YcQX
zqi`SOK!;YBo(d}hD^f)f0n2{)7+arc_pghFYM-JiFZP6$=pzt}Lb4a4?Q4BD^-OW+
z-<aJArAt0yNWnqRxg@XG(7%JC^0Y}$gv%4UkGaxd8~r@+6v&Ww@uvdg;hh7fwMZVc
z@W6%p{Qb~*r6i>ycKU=G;(64OY>X1Kit^i<kSgJ^>izLZh1%C~_ee4)^(xu<+*7S9
zA#`5iLz>B(C|GO!lxqXPcqF^V_H=3a;8+r@!ow6fBpR!^h6&+=;Wj_PLH{-d!oziH
zQc#rkiNu!4lcNrBa}~Ygu$_9XVrz9rMfyGGJmP9~6K@fl$1#Qp47xHSK%_ARwMR8j
zsEKeO>YxG`#c>rA-Le$^hWiB=TB{d=aDNb*h}SzgRtl24j+Mz1%8HAKYZG(JNLT~d
z5pdn<#@d3I{Oc6cLJ%PosS5@X-X{bV*YF0+C)5f<h#U}(qbB}~nYU^PgG3eMmSCKO
z@SIwI8CoSdJWfd-WGoZMO+3cjH)fph3M5Vfnfm8oEQ64NnC=2_I!uU@lS;&9Q9^Ht
z>aSWr51M(^pPL}!{<pc79o3xghSCbR2ag4n61N<#w)Y8V--7a+^yhNh85rg%kEfFJ
zk@F9x4ZY08lr0P>nE~E<L>CGONzHQt`$h84hx4rd`;sO2sgh#Pn(nYtOJZ~krZq?V
z;rpBDyBibP!(18!v9WU)$6Ho&K~@T_tV%pi1Nkf5^?!@T&#&HXe{hGRz{LaRij|{Q
z1n@EI!A4*+)%S}mm{x9CdNw%?tBb1Rf{8l~W8nng<^fiF>K$qWSPgAVwn9oJ*q<RU
zF!td2VLid={qw~iQ8I4}ixh46?-l5f8${7(Rn^UM^teE1fjnzXqn=e$xi28vo2Sk3
z_4T)e)--C}iuJkQ$%>l(+$hZHQfDdq{fzL<DFw>z%AQudZCSg1|3fu|Cx!cWb(23D
z{M)(btFb^_18Dc0RbG0Ts<e<pOPJ79$V)gsD7-j#t=9H?8=>2RFWD-rNV>A~Hb2I(
z_KxivDBK<gpUq7AR;7()w!(QqJ2gt4l7wZkQF8qPU1y@bJcpbRaoBrpeP3lwDRFUA
z_3{46dsp^5n`WDTcW~4hxbaY%=_rB$``Raas&BKRyD;HX8m;PCR$KTbu2bT#{ijhI
zxJEu`wg+G1JpG`hlfR6WCZYc@RgEUyq@~_K{@?dt!ehlWS4R{iQ8Ups9Laf<G#G8_
z@}+;noh%g)=1fw-y?L=G#W%dC^p?#rfSM9k$Y^-_V@{Q+m@bcBc^`}GuPekbFliA6
z^0fixyFcL&dOS5}-azjV@~7Hqz$lRDU=uugXk+6~Nv3;r0DDaufc7`Uhc$U+SwNTR
zq;w5Giq^l{AuRNcN;fLO*n?2F7{wvdFl+m%?anx`PQ-rn?NF>v#Mh%te<lc0*i5n6
zHy9+jwF}~q(nMwiHerHdlgReS=`-nP<53$x{6QVh+2rMN@V}A%<v7e?IOnF|B9A0d
z2Hz<5PtGfA0BhaoVM39E3g~BDmkdZZT9V!{)G-IZ9}3Ue_y_~UBK(}=5<~{_DAn<*
zNLz1nsGH?(7fTbFAe*08hhlF(Q!_d(>fd?g)g@aqZaGgTPFi1!%W7AT#oZQ1HpGAm
zD?jOgvSK}^O*o(b0@s3@3bsk3oCdz5UEOpRWXT##Ty)&FoPB-jE=g_F^i<TI96fn@
z)F1sdbUzmx(HbPYbXw3_rc`*eY&qysBkwR+BV+uR^KP-bXt28d*O1`lGt7}KAAl_r
zbn;l+6iP+R<^_``QNM6PI?Qj`DBx5J{I20QJga|ry3%NkrP?iOsM^ZExpSzK6DBeu
zL^iy$ms*JOE`%_6XBS!8I|wJ(^+(l&8H<Cyvbrp3{|<Fv$aAOZp}AdnJVjL5l09^c
zRcbEM)J8;1yuWNY*HtzvWPH*hcNaBr;t*q(QC#Ok#!AJ*2c@h1=_MU)SOYwy4M8bY
zt5csskz?u)?iVr}j7{H%OBW2th9j@^KwoC3tUSXQ?VjFqqJoo?N%R5({j*dPNVrpZ
zjfec!{LEI@TvnT1RX@jyEBQc>5L^dUVAwRV+FrE;4s%po<!4mdg>UE@O3MYM^UUGA
zCN@h_GL;BNrY_{Zrf^$-=QkzyVY>y(nf)2hXn1j*7Tr=-`z_k5&u#y$Z@<4!ug*=W
zZh(flKF>KBIu{#kjto2J#4LP$8^}^fupSt^EJl3~uLm<qmsHRVZ;yLorTDku`5xLg
z^d3tQ;{GY_=iLm2yQgEwmh|YWDiCU}udW=gfBOc=>%J4#>j$g$=V#*yp@%ne{V9Xw
zlPb$TG{h!xDz)mt$tDrcqJUgd<4xW@`uEI3xwbzKQK&JnDx~vc8IQdCr~BpgwFh%2
z^XNZ_0w(tV01L6PvHVv(sIFg-mWTtc>(}i9E?E?%2|xt6k!GuJyHL8JFB^Ex`(*I`
z1206ANFGFZIa!$<=oA(!9)3Y&;_-e!EZ5Vi?bqqugXMKk=zstIc<_?YkA~z&83vD&
z*{Rr%;qrcWa;9fS(KIlF!lx=}bbD}j<KMIQl%W8qM(&-zc)@0;qUl$5^Ht$^(*7OX
zqvDzJ{Q9_fgu?rkMKMspb|S0erG3^VgS<%Nqx<u9fJax4D-?b@pL4eIx$_Z|^3y#T
z;wTD5^K{1N*IkIVOl}asyG9<sSkmLm_4R!}lV*89;vn?~c>lP9S8_I1qf(`iuUX^-
z5U80Xth15eq>%bpUb@sR0qHL9H3E-2n~&6(4kv+qub0KZBQP8lkz~%C+C~hw3kA*(
zygJdl_rX(IG&v1<^e8Q(mNo}Hivm?dmiQpgGz@t$uv3*5CN&R{4SY+Vdl&fQ?<dQp
z3stcx%>XKqpANj6uH9*Aex$r}gCHxw|M<Jk06`BTOBZDwtf(RHw7+<^mf{v5aEP?}
zZ}w@_^}41Kd7;z#=rlF-(j`vx!uX<{iqH{tj3W4}e~(W>`UKj(MjDlai^j%|bmAxW
zmed;7tce?^#}{4CM1_tPWDRFgj)0WGykZ?V<tPYqC?Fd4w=3cwq366KtYaxm8r*~D
z%;TkAkD{HGO#BqiX)?{zRPKN7n2S}(SlJxqLU}8t#M$L@%M$TX(ZD<!?~-d{hoJK(
zUBn({NB#9I$&c1+E{P21Tj2f@e^^;<*9AYwXl%IQnN*eAMcnfBs$yq?y_pMvjOc>I
ze1V;tAD6QyvOCc{ug5B`m)fh%`y{7+%|n&gwlzD8X?$aKW@~45OSaCs`>BWRanv)X
z(d?*FYPN@vd-SGJ45vH3%bA_rtDvv5Qyq$rXHIq6NoU(e3_}8xk83{~kts<^DtTY5
zITS*n-sjj_XH~i@Ec59&#JQy%^m5&?na`7*KUs#U&tx8rGzcUE@rNYs`zWw$h~ee9
zg^qtf-J8V2Lpod#rtMs%*tdg3RchCkkp5AJE5`>=d{s>zWtAMgZ!h7ut<<t7t65dr
ze+TYcau+svZmaB<`2rbz<-VJfB-)t|vAdpP#WaG@(A3c49>bhzft1J3Jjjka<e6R6
zDF3mLx8<4WvZJ1)!myiYi(M8eE@_4wXHM3IykNx+iIN5?9}>|jR$EKvp`au5@+b-C
z5_8_lsS3-l(4}y)rE8BSa*5rP?063qdtddWXCP9`9aMy7BGpUENKHH-pmqGLe;y@o
z44?9P2_vcpn$Nyq$+aMC;X?H^nQ<mx32OH1^|1}AcyBM}$e%eE(h<w7KXlV<sWa5d
zIC{x6y-RKmX0*xovhmib$)j<*$j`at;%KNrq?UhzC~7u^%WeeYP6nqa^G$v&>9SEN
zkh!ruwdWejN~q!PGhMXh5GC5oa&e$at|xC=;s(}$4K$HVE|<d_)UoK2deD{|rI*-i
z7%ffm)B``UZ93+tlhlio&=J8McNFHCDKbh@R<LPM(dwhN6;QE!8|=<%ePbJWx}tXV
zwJfKpa8@yGwLaZK$F$d~VuM5WD4)Z=YK#%G0WdJuc1`2%P~fC&HK*8tpocrp2CdAN
zeP~ay@g}HTSlFK>Ii!W(Fe=-=HJ)yEI(`Ys6ha33^4{O9J(#X8{pV(|4O*46b)09y
zLSbbz<*rQS%DGSQ=a#q9sf%ffFNC>r=>1Il;*;V7#p*m=N|dhJE=IwkPI)@G$my`H
z{UyGo{adpsWhb%J`U2<zaU(sMWCCu(7<V7F0r?3#r8#-eINb6vW9Z+(9sM*V4fd)B
zdP!{&{k1!K3b+lIQ5vUgJtck?oWR{JZHtwwf@Y@jVrMZ~)88ps;7XR6KhuJMI1lEP
zPM7SY`U@m2T>vBoV}Z4x)ts)9vHe8u+2(xahsfrLE}3pe#MEeInZ_3du--r!x0PXC
zl{^59CUn}Td{QP8&xh(qxr<bPTjtrmvjw{=HPHOP#?_H3GvV>IbPzomq|Ih@nZ|ql
zRei>xrQ)(Bv({65cYcyVNAUCATqQ-Q_gWm#i%|Ts@l7OP+JP&-i39PT8bZvG4w$C{
z1{9VoIWF1NuBK<1RgciGCKEj2ro5u-2-#ey9W2HoyNNs@y5uOB(vvoLbk-2v^UL?i
zsJJGRvgCK!mn>ZnRyQY{;B1DXZtW}Gz3nsVgH1Jnw=igyEqRtp%_wW;bp5TJ9BeL(
zDdfD&8m%#W>P02wESnwdt*_IdgG>odtzJ}B74$*W6*(?{pn||dyQR6+)LU*hNtM&x
zlzvt{Xs)eY&AbGlg;^V-@nk>(QEa!BWC6T%hj{X=vL%O-2#C+MEOQPTYb2`G^%rKq
zcE>Jx&AZ6DObAu*d!IiDR?g%j0jDDDwsIWTT;sd~s!SFqR9g1QU4uL6Uo=cmOpqH=
zft1u#T6=3_(2_oVxu7T#6`B1r<4p4yodtp>1mMcN;2cMZE#e}^+qAZoR7+d#e&?vS
z4pt4Cr}tubQcC_g$yR!?qWp`UvhsTU+~#a*7=cT}Gs@~K<Uzsyp=wYOIr_bc^KBmQ
z7jzw*^2<=ad%!TqN|14?d4Wo4AtGzjG^>>GL|-)x153XmZyEYpU;Wx_RscByetNU+
zZM76`V#va43(O1>G9QT|__+7j7M_xQ^0j>&hv~<SvY#w<yPDX3lO;Dr`RH<%E!45i
zq9v;slEL<#L?1-mASO!gZkHQkJxkm_`9b_1nMDiCyxeoI%(^vdPrQcB2?my5me&sA
zb_`z0Ab6A#*noDA7yXcTr*iE13Q}(&#*Oq>%UFjW{uQn63y}VTwFp5hurueQ*AQ0E
zExnkm!CjUGw(pB(!ZOaNt&cusVDpjtREit;o^cE_+K4@Iz2n|lT>|DsO9BBWN-)pG
z;sR{Rq=_CvP@|4bYdsl7TPsD2(eO`=2WkKZ<_`{%=46ga&@9>EGC22R=OAP`v$b~F
zWA7hTK2BE?B_g|foSQ6C?n|<H?@xQq0a?Ud-^o}%=d%J4g$UR~w`&fy2QmllPsrnH
zI*mqXpwB+odgonhQgfqBG})2|d$VMi$$Ae6_(c)L>)$$SpjCTORa;%d5)^@o<8%K)
zX+xjWIek4?*kG%y=OEl2e88>CRl#<AG8vjDYcW(Cplv!jDcc{5i(tmCNJZe#%+cC~
zo#cElOdFwKms%F}=c6<SY)yE6j?Wx6p5E>ngmx?07R+k?iEYpLF9<o`g0QPVBuRhX
zr!Yqe^({%;u*6hNO>=Ho7b>AfG=8M<rx4>%+yLrWTMK5Q{Y0^|ZArBum_~E4tBD}Y
z{bo(e?QVu|u|kbxLvyXaW%%r_1%b#FFpsrYM-Cw5Ke#Ig#k1<+WK)l8E&?fohaM}y
zt=8!wn-7GoRyX2wYxCK(N!T{|KjE2s5o4k*3OfTi`Zyryw!hWpumR@#_`fDd(<5=w
z_l2O~MauKa!m#SU93*E#Y`ZmpXjp6lw8XfzNtO`*#NLf?p3Fnb3A@pfIU~=_i2VdW
z8)Te-f#;i^+1w@VwcZGq0S>vwC3s;Ei9QQHLyqLT?@K?ZA?sIh9f2C&8N89*L#$b`
zpMRW8*;USp#)SBPVH-Bxsz!yhGN7po#+|VdX8iz$Px}2wit2YJ`~MDVnHcH+y8z}U
ztchr>wmV+EIX-&k2s@yF*f?v9(JAPi0BlszOYBSh4vxbYr6IGpEt)9iKfst}YsPiu
z>NojE_eOCOCnXCjCGcoAsB{6KHOVzQXuW7iKLLXRW@zyEnZZXFHAi;f6iEX!kbI(&
zRd)w>ZoPJ1-pdH~V7m~#cWYR=Z*)gpd;wgEZ)Trho?=$qQSHnF3pXHZxLv%?r{f3G
zgThGO-tg(XBRzg!7YWpbzdZd{Va?mw?^pS<P5Z4M1*tl`Kc0l?BEj+kH3N`nY5l&9
z-o77Az}A}`aYy*+@C&uS3n2$a?SdKrV;kl<tiY!Ogk3X++Q^x^C^JRUVeb1VKk?nD
zzeL}wh8zmlGACN`PHp9@(5BBZ$_E<+0U*l-%)#0$XxZ&Kc)29VguRgD9Ekf+zyO>1
z!deXuTe?f)1>Db$c?fI`OZgG;O2(d@f108G^$Dd*d@(}~@*ot1@>W7{*!;<llGm&v
z=h%@8QviB0CV?u%zIU*!L{Y5<u@-Y!r$htsRn@Li-b#K;^54K`&VJ>Beua2S0<zs9
z-95;@byVpc8M8yY4O?_WxD7)<g6tqV{0(+hTjVA@lpSQz+onELaK~z)SVwcL`KMwz
zv?jeI=^uqA`XR~%Sv@V7@lPfBCPw43u#GU9i_4r0CVQe0xQb9ER>ie*x;*)txEqE@
z`>hcvcM~QpcD)}OQDVp9Gl<lhDhHl0zacYE_j-g5L2ioF_v3@e73fz6m(P<+@Kb-4
z4O^i1eer4MZ{82f42g{n5><d{KYw~GEJOCU379uh!SK*eIzquR54xVWqxYEtG+>)R
zY$mQ5@-Bpdl7<ZxVyYbfb6jW4Pz|u1`Iv{yS%<3=$J;Z4HWo8-yZE&Ut6!K|s&W3j
zu36MFx9*!cP!0x#zeF)Dma;IWV#IUzL_{_@ffHG?6*bqcrvMWr?%5t!N{27@ODql6
zrXn5G1WiQotRw(lfCY*hcp=rS2CX#a@<N_a;0Q>-*5~4(AS;OhiSDP^C%WhaC)U!8
zG!2vGOF5Vho2~T5{q05$-9V<i3cB|VPA#3-tXM@!|FfNvW}81h7&k7vmQF;2yFb4E
zEXDMSo0S0SyRWVAYK$zNyg4dy*f&Dh|57YY-hyi6Pa#5R5&i5L{TTi^f=LwcyMCT$
z4c{g8)5_dnb3~T3i=}v~$LF%5t4S2<34!HbCzLMv?Tj@wD}yptD!QJlOdfNhcgvL8
zl4OpTzt&JkrIi7DKF+y05{P1qs9ieR1a9%}5qnJh?ki=qme{9c6Kg6|2f9bJ-cr41
z46*jfWrXw001@w){psf(V0T$Rxe<z}+R4IMl3q%?R5gM;ay*F6dpZ9U8BT+G4yUbz
z0wdU((F&PPZuSOm9e)E#L0`r+X=NXC3C2ruW9CP@BN)nku`q7V93P(J7qP*;;;A7o
z+XI@Z5rNvUV2TgbOR%{9kk$<LUn_-ANw|O^XJvTANTNz2CekEQ+!9PMyoDCCf}^~H
z_!TG5@sEAD#b_?_i%nz`qL9d^1P59Ko#elr^^)d(Eo6BnWc_(lGxI9ECsFDZneKWM
zM~gP2WeR(}lb>x@#?We*IpxaD;*<8*jV{vp6Bst<ik{otb~Qopd+8sO)5f!)2{Wl}
z%>CC33K>E!3g5l;MEZ4>==iEO)|y$OmgLNrnao)I$;+f2T{1aM=8vAkX}}R?zjKho
zsrrt0(zBy7hiZ0B&k9nG>sZ~26w1^KPYr`?4pDBVR$43#K4odwVs~jrX3ZP^YzH%T
zr?3fkvT0=<b{1ckXk)y88R!18;#z3fDB*;|1r+XE-U9*ZfptN{nFgF3gWIX)3A7Sg
zaM98!4oz-QZ^%aXT}I0EW@)TJ(^PF`-ep`L7G@r6$-D(3GZMP(WD@qVUh9Aj;DrLu
zz0SWQbj?B=S)5cc+ImHN<u7qLnG=J=IBF}KWQ!i{7<)}+f{SSGWDABg16|HO@iohk
zQOC7xpXL^V4N@=6qn;~y9`$^`LK&Muf6$pVPL}ocu`G7Gz4Yy1BsG|Za+O4`?mvMC
z=F~F+;NFff6XRJ^@b082)ODkA(9(U~6l3`Ys?Z90L(ukaod7FWNQJJlP=@#|7Qr^n
zAmlw;Na>%Z5TlyS$iZ=17)FF$Y4^gRhO|irqiLh!lYqu9uW4D>9j@3U`b3m?Qzea%
zf@|z2O1S@4>?=;3JCI@>tFxel$GjpV1zP}WNLYDpePrrD8ihIyfQrGXJYsB)QD|KZ
zx<_t~?Kr{mcshM<Iw~}OjrOe%{{VZJ+RvD|pbvxT%e7Ejmj0cr#}~I#Hd<@?dA2%o
z3lM?~iyQVxYHKnx{R1x_{I$qf{z=&vQ8If=jF9(w6^>x_2!c;KGt1FwJI$hRrW%cE
zptlT9zrJ}|^n7i6D3|mEzoA?r5*GB`P_|nTt(zZQ5qoE1LdC0L#WSO|)(Y&ng-}>U
z@p4;K5s?nvw{+cZPmM%szJ!cz=*R!p)`(~O535|r&M5@whoP-k@qjIkt|y1AsD+K-
z11znAryFvAsd`~#9%JL|HPI;p;JOvq>fe&>5~^n*da%avuBJO91dGPi-_yG`Q4R|9
zE0>k*?qpSAT^!c=Qi?v@iUF-S72l2{Y}$`mQSMLCyrjOXRiP-Kb<iBmSMNQFH`}|K
zfq!+x>*H=!==?(|^<p%>yyzAbj~%LgKV^<pv;FVcc)Yr8q<=CdFQcd<mO}MRy0nho
zzsK20-_3!vfD3V4I~yo%A5N;v**$Pt^ic42Ln7fJ4f;-asP>_VTt*DFhEU2?t2DFP
zOxJkw&QRzV8pu_#9=hH$oPH+Jio2taJf?&jkx4SreivGl1664$Hmg-Gl&5@YYOl(|
zP!_h6$(3x%BVrpzXVfR>S81vnsdiYVWN|EUiBi>%@3||Mp{io@Vn)1`63D-cB=5(7
z9vwWy{{uM!Gji?tbdwlBN7wpOrVqMXW{<-1O_cNcxVOXC{d2T&gKxj)XHJ=pHF=L8
ze5B?peDbf+DEOs;+vpt`#fFA3xW=>&{ZqKzSOcy3uieC-q(T0TUQxe!b73^wM8k-L
zxB^ex4S!e_YzRZFs*kV3K$MDZk^Hhe0XdMO$|5msJ+(O^pI4`>@QUv?ZXdnuZscms
z5^fntM&M|;9~%c0df%5wfu8r+XF|itV@kFB5uqK7qsL|e5^Z~QIxG=3f!^7m=6bDy
zwrKkH3xxi1gmd|?>=CzD`nhl-o;)PheNL+05>{&DYg4$FogLpK7e4zXv*)Zm?!bBU
z+4dt13YbOXfF$fJ*bLKJ>6(TL3=mU8AqPB<5=-^A02zM!2|8#YiiPLRu#bDoU;v1S
z?sHb5<I;L2Ao@iyy%qF2r@7pD{3`k5tFw<KE(qs}zwx;a4XE-->A8&OvlSgvo{Azn
z=Bf-+a!m-Ts&xppld7A1N!NYl9m@nQ(_**|mC0*vZHQCIlc{2>gGobZ;(HTARpv~<
zxGe3IfE`aLiWjo?-Bs9}$wrhZGo6wUTp`%G?^M1!Ht%1NU$YmW+0>7{E8Q>zTrWi)
zy4zpgPo9yEr9i1WQ^?%DJYZ!grV&J4el4<2j)IE?V6$YY9hO~Y((?9T7q&7I@4*%Z
zlAR`+?P;UMM|u<NRl)!>JRqg4hz%j9XTGDH%plH1=JV+?)`sE4=-azhs)TxZR(4iG
z1C71ZwZ_{mY)#9(ln@eio2$X1qiQyiLz-HMkEru)A#6`jH37_+I4#-y<0)7X^d^e*
zG=@AYe=RQWl`{zBWcD4kYi1dgtmA<oY@~t=S>88JDpp3BTOrHx?rPK*RJy*E3)3ji
zxLT&=ZzD71#O9{}<TsN~s_%kp6KQF0%IdUGJ_MDtIAXOp#`ZVX@7dK>&ujfmcA7{1
zCjUTH=c$*g87&eEJk6_RsW^qr2jPB9Q7MQWAbjdSHIrDqYy*gw=TiFOo#VpQYFBB!
zb;y>kEmC!^(mYttZ&W=t-b@emX)j}TDR=$ty=hhETjxf;VQzs=xYGUbpt>>Mw#yyn
z!(V$#R;f2<cv&dXB9*4KI*h09fa&_|fR^2cZhP-B_WxVsYDdckO4E3rqAge1;Uty=
zm_KXYB&M(5j73%|2-nZNHu~X4!&8Jf`-rE}9@G1&x#5<-A`ED1ihIdY&0O?x!(S7a
zc+nZH8#B?@^{1<7fqm|x*=<;3_W-vm1NQU`t7+FS8lqGGKeYat>6!m`RYU)K_&?7t
z4`A(V;z%cIZQyJoY+__*Y{JV6<>c&WVqgR1zCMICW{cByS6y?8hY+G->;MQtFyD~#
zchmY6^^l(4@I)Wplr-eiH~eFM%WjE1I7c%^2GwdueC%k0Fa;jICI`=#gR)C~4}zLD
zr-!zej`#;;EHD@q8b31N=pp6^Pfjs^Z-&aRF;Vhv<ad^v^Y~fXdk3?N=zk(sK``kz
zc5}s)TEhL)D+h7YkBKEJlW8z;gUHCE^|Y8<aFiAq26OwyQ+n(DeR9&nH!3|I*;1!@
z`ySA8%rl0J%`!k7EJWLL_j!2pp1Q-}Y7Ff<sE$y)^^eV9;P#c?$Lr^9FNOuP6qG`{
z`_mnQt`R5?n|7RG<mvvTC%Fb5&#bh|9=q1QkU==mo`3i$SYDkuX><}P&qMR;?xe*8
z^KG#Xx>H_sDpuZ>2`!CsIl+RpPO$W4<)#XT3E0%!D_|)eB~F!@Oj<2-4!_dD1$ApW
z4RM}hTKeOBqm5B8{>H5}Kj`7%PzzL0lKjThoPR?r0VP-!X8lj-e#xUta+?l5QK1u1
z+H=tcD5D@JQLMv#7Z66?Uk;ksFypqP_WO0G>zqy$koI1vXiiRP2tt_Y`!_@%`?5zy
z?*Z-*97>#Vb8W;;S{@?qq_<}ZF1_(zWUL$T<nl*#HP4wEiN`1<$ufo%gaklemn7!$
zl*Cl5Mvq9B%uP))SGi*A>1=DwFHeVVJd#t0K7w^H%7Z0qd)_>+Nw3OrhN7CGtT3Lv
zR#+r~S7B5ml)#pip(+UNL_a)L<1;pQD28UH94zWsmSNqNV_z|P=FLl`<5wl4p#9KZ
zl#W@3NKpKFKKSI!$K)~S_LcY$g(2fph3SvOw;VR<$ig?}&8UbpTCf#3)X9%(IiWr;
zgaM4AAD%`yoJdp2)Zew-8HL7D(J?JxA_AX}e61SiP!iJsV|AjRI`^<-uu5rD?VzNd
zZh#6-T^?}}A$75PT4WCRDM=d<RSJ=-v9TF63&VS73lvS$DYxz`)z^JeGYR$Z84YZD
z6xdI!%5Cjkf*aTl!CetoTRv^yR%L6NXJHg{LMO-cy!6mx+$Hx*$wqw8w+Xp~>9g5Q
zaNJu1KAg+sL))NXyvQx9Z`N=+Ot*53;O0r*(hkn3tsru-y*GF_={>OA+GtT9S_`h!
z9_$Yj{otrrnx2-Yt<@}0^CBx#xfZH0#B3Tfa&g_IG*Mfj9(O3*|1>Dq;BTOtJ9nNV
z41%$c{uJQoz!2p<>K~p^3W9YKb&pXA*>OpC6kjBSk(jkDOPmOXG@Sdu%!*ziGaH7W
zXBXO@5NpF_K|Nlx3<8I*<vt=f7zp0<74{B}BjB6vG-7a6E_<y_VPxn^%ZH2UCj|rk
zZqfhsDa<c>1N#J5W`Le!%Y+LT$nhX&#$TCsmOdEN&mt{M2|Xy@Ui(&Z)B`W}E51!r
z;b?}he<*FU7{Q!>e`yG@2kopxj%S_8^aDxV3>I~V^@9*aA5}xAf7CH&3VW@*1X@n|
zTs1WQ6G${3+cx&?uIO%*SKbe{y(>^a@e~0}9+plze-~%u+gajW+(JfzQtqbnAZ&?g
zlVdi7n|i6PgIkkcUWss5X!x`7s=$%qWZ@sy{{Z)r*|MZfPx+6kljcKtVpo6$fi=7X
zjwol^>(D|?xeC6;#)->!2{MMtyf0)v+0I7QSxX`sLJ*7ZhK;4FI#9fSUlNKc#uZkW
zuG57obi)Q%IsG^PRulgwT#(b%s`e`D{*f>w-s^d589T;%B5}-h?d;=$S0jl!<K*uC
zBOgc%ST>;5)9|{yE-s;cR&yIx$HAn!Nmrnx<?pp1swCB^DeybZ6oCi<YHt?DJ;+f%
zwuk1D<6$aBV&v+wc3Wg=b@{(NY5dcKnkUIBD%uOV%8FuMsh-Xa#j?_<T3WMup3Ns&
zFBYG=fQk!%-Nd9Mr*dFHqooBo3SeZrvrV~Zse|IuX_UPtj}|DI=qRlz<;``dXJT$O
zJ)+vI2BhJ*sDbIvgcUKaXJ6HH$XVG7thPph+?Ko+;i0hYtAbFZ+v^gww!;M4d^IfE
zqe-kesX^Ded^a-V1hnK&+^!WdE7=N;i1YfiU(5HNhd`nJ4}%x{B0hx9Ev_Fm_uT-J
z856(}2kDnv*Z@#$BfKh<z6?Ocs~-RRUdm37&(}tZ-kNoa-4CSF8YpC;y!WTL9*2Tk
zCe!+n*Y~tR%&JUXnb-728?q@A*~D}?btSM~GWb+9404<8qZj&9B_L*<7B)@}?4Jp+
zE%`nI{>lQAo|`ZIzH%Exi*oaSc7T$)Sl##U^_1SQIp1Ds;}zds+}=94yr$Lh0$9cL
zmWZ`%pLP)ynUevqAS~0#1VejkUiCp#+m6a>7*}Uif#i`)m-WnGWI&v)yPaA^ZIRB-
zDFneP_=$P<ng1gq$BC{=rDNd@k%OO7Rgf9o{)BAY==mPH?6@z5T&L}F15xIUQOBH$
z;}7d&0AsaXU6RPqHAux$s5I2D!H&ynp|S$*(eTmYNQO0!h;EKVVDw+}7jHnBp)-0F
zMk5Zo+EL;?*SAUsC<Nr1mY&xGv(F;L{1h*#{X5s`<jzRR089}qYHr<f=I)m$l$F@B
z;(&u9Dwi5JJa>mEc&Es5!EHBfD|^sWm2mcT0<<tGreTjBF}1kib(nnEat96825*y+
zdGM4Da~z=TQetD?JvsJLh$s|l@_UtT=L)0zir@T$XZL&vRqHB0uZ`#BY*gK9qM+i|
z0#lA5##p+O?4Swf9iwyDu^WJ*RV%;zT#cVZ;Hnb*7$RChh#%x@T!OhT>HJ;(?`O$e
zDsaU;4T1I%sz#BytNHm-7Luv?=w*68jN|%5=NPHDae7{0KAot&G%zb7*U;_?-=d25
zI#XpzOoR<pu{U}>2>->8s!*QN*TT<{gyqUph{6RD)v2wi>EbDz%WZ_T)KAB64%1)_
zzxeAJif2R{0&<Rc2a6K#WJputnG){?WMh&0Lo_~#Ilx@z-nFVEKoYncaE!8%9FewA
z1Jp^yPQ<ZQsdzK|@XArHjwepdD*ontHb4yU$`+#Y*sJM0e;WU!B|5HN)CQELIv%9C
zfth-GvC{KBMzW)%u41`_`t8bjT(E@BQ-vD&W6PCZ>dg=|y=2<lry~|ID_%x4GMzPA
z`azTF*slw?qV#G@+X(R;lBr4NbF4yV4^Fc*6pSCso08m(jba`mi)ROh>B>8>m_fe)
z57Pi!QtXF0(vzt0bq{c7DudnpLrZX>^WFEOv)b4{X8b_i0vV>vC8!DbDg5d8hu5&{
z9KJU=@{kRV_oM1`v6f`;Vkt^mS<!NBLr`X2vW_Hk98o)TkEv3`%pPvi$=;HoEH^I%
zL^Zs@gm~bosf1fm+aZrr44>1gSS`LonrV2j??QbjPJINKb8n*5+xLB;dvIGA4i=xo
zPR=Jgn<k+cFG-3`=U7)a!U$HBbO=s5$DPkG0+A`1C|g(ma0Poe(9E3Dtq9U0&bTW~
zLo~iH;hLSo`&go3avhY18@LHkpK0e&YRy@cm}wBYlZe(~H+T-sEFe*+ziq-6VdG%U
z)FcM9%Rp&kLl)E&X*lSHVn3g#90lj}1;|ijBY}3ol0?R=t*g0RziViv+3!cBSEW5(
z=0Pp4FljPuguL`CG1V&BqFrWh1nuc^M%w(-R1#dNi&d(8&U(jotNGJDOS+tLC<MVi
z(!@TWGU!<%kg)Qc+^BTuMdWYVXgekp3+MpE2(&&a1YmhkVQ5)wJJoq6i*)2dom(Qq
zrH}cdIR0tNUz=#HUXyaxjb%EGYhqQ_W}a~?;f31&6Cf*dS`iqm(Nf(Z)*uxxlGi$z
zVUGEuy(d#@FRLzwrL3K7qAM)(WhFW+_R(dl)B(choM~^)Hb1MzXe4E%xXX4n;*eI{
zdYU2Mqp|j&dHF34o^wIsy6!@oQp>S88frPrVs@GKx5ToqjAVst7iXobo1nU;!)coh
z?gX862hTF=;0G{0@-Fm0RRRqEd%nTI!p!<#RRUe;W40Hpx7Rhh@Tq&T*tGom%+pdp
zkI5mM;ke<s67UjpFP^72ZowJklCJxMp@0(0_Us>`6nE5#ts9qktY|I6P;|fnEBkvA
z05s(7frPn2dw_*Z__(@5b}5n;Vk_CS6^RVa9`AUkVI5U}`O*0i{I{01^gk==uibWO
zp}$#p{;s&kn3YPQXaA6mw?6!4@l^8|#1n^~EpQ^3^={q3_t+{ch>V@S((w(qWQ0t`
zOqHVN$FU0Ey|(XxTJfjnf|`>laNRw~-8Hu1jj@`BKObUI-aTl*#NG@wE5z9(IfXK9
z1P}&l?P(_lr_|eAu@%KOHnlj%MNwOn8x$W6+4!r3tH!0+7lww@$|R0Pvnot47MgHe
z(1h^?#=b_?J0WG_QdqGhR2X<sWrZMaP#@w0h{O@n>=lnbHcRX@mN?A6+sk@QQ{+$F
z6J6vzaQ%o)V#nsM>n5>`Vy}T)#8BO~7rSNRFDG5)<$eg3=hmg!PkFa=t<|bqGHL~3
zGh{@IzPB=KO~%uwRNi(-n*w04D=Mv=IMp!DQ}_gUjfG{81|rT@G6j13i9t-ZS%#ZX
zqn6vsE%y^ju*aech<4?=4t8jju=CpQtRY0D+Dc?#6v}NFGDT`~TBT2g#<Y!zESC4F
z=DZ+));wbNtO^0EJ<E@m#|tzNH!0=k>%A!7&Zh^_+IfY0lrKP6vZ`7Zi1`Y6Wz+(@
z(C`@^d^>k9kBRHp9C~$-Uy~^ek@F0Ls&rIGHN>aC^)Kgyx*f}@nAa0}h|v-8eNelr
ztbjwVe0*ySAX!ddvr^y1Cgjr(N#VrKZ*;4Ea00QAi<MNgVp~w`O9J<0@_--ZuENg6
zIF?o%bP$qo&=%$xAB@@}t)Q+v?j;o-sVu~jbkSD9M}RefF>fWq)+<zcdj~$kF3)6K
z)xx`Rni!cUSH0DI{2IZ#lf=0k+#7c(JU-dE6@^?_ea9}v!QvQNnWi~`hg211<pd`H
zi@wlUa4rHs*b4}bK~}=M+um-mdB}7R>=<ZfUF{2qKgN9V?w~v&A@q(=0vDGjIkp%}
zUjHV8vr5)NuZ<!cX}XC%ZT9|1J0PBU<MfMv$?FRz#tpXx(eR>Sq>CchmGtL?6dl&s
z2REAw;tjXQUVG|s-dr|YY{NEGCvXbvFre2ce{5XSyQp|V(5kM9utfnKq>7Mbdp_Ka
zbzg|Ocb#q;EdQX6^b7gi6!46$;#+KDdc__i0`wGuV}#6$dJSXak>k3FRU52fT$o+I
z&30R5_3-g|CCFi&B~vp5PLi6^bOkAhAIHjCfRnlk*Pr&Q+YIv<G}A8J^Plm*dCsK#
z73G-Kn4o4J@b%rAG4z-|dyF@loe%%8AIpnGe-L;eAjF&w!3#jR5$-1iA_YVK2?I&a
zjT2>tVgtBIt<1xN?rM8Ic{@uP;TuKXf725!D$IuoL?5*BCCPK?)gO}ba6@iFCB#uF
zV2m!V=hyTQEXBgwRj{TmB1MP6c?`;nlXX_Tuyv9&nF&o2054~x9|EiT)b-R)vNViy
zwluuOQljvBOL|_>`5agK(N(Zm(KCD?W_=kdKW+URYo|hrjq|l!3B|nX4%IYWk78_I
z{5;ognRNQ5Tn~5kWWz6$mB{C+j#F6D7tBY1Y%6mDb@o_|pRYopEbkM23+#&SOd<}g
zNC_18oClH<4J%)nLBsv!=S}mGZLy;BGU?pRAHp0VYyIY>#rwBVGRmL{80|7Zn@V)*
zAoZQXOzWs2S<eY`@K)FP-gu}v!Sv;(8O&d}9Mn$BsIn{p#d&sqbqod^EH0qBAW2jv
zP;V2X)0%9cs;Lgv)dVx|A)mHH!jd^f2R711FB@oa{bi%w$}`ZwkX(6zmfp+UNt&yK
z)YdjQYQfB$IzevO`V&qXKj>$?#Ns<zgk>M6<fVU)Jn9={XUzK+gzvsZ2QOC{`mdB)
zIbE@9Drn;k<I)ZZ=lM%&pi>%O{@h8o#Ur!qcXb`B9Qu%+RevegBzRTQ=ar%f&EFE^
zNytyP_7+<0>>psJfMVhQbgwY}@AMHy29Ey@E^We`2*+moy;n{PeAGoKlXMMu=((Oj
z9GQ7R;;{ws`$)$K5N=OOh?5rMH5V3)Zu45a+EvP5l&Dhp(Aqw4CGWqE4k~V{hv_7t
zjj|dhM)*)g^oYMy<{|B-3`rnp&K7niW@Z!vl92RH1u3G6C*~STIyiVN9vcPU7HaVP
z4x_37Vopu5uns^tE3xz}3ZQKY?}H%)wwQkj9Y7zBlaoflc@wpN0{H%!i~)U+<g3UA
z^8Y0E&cbXNk$Ojo#86bcqX*TGALDnhlH?@?@S5}rXNNtd+^~qLf8w1_^msq{jV-l9
z2@9~L$DBpYv97~Ksth^)deY`|6piR|cB*$DUN!Pkps67$-N$1X1eOkhP)!D-4rkTo
z#cZ@cHuDBiR|!{WN$~ebvUU$x5m1by)O5}_50csnepF~~1i3}MxWQ~(949-&7(G;(
zZ=z-xVZ9_x_j+NW;s{i@S-XzMu7N_MYWV~R=jjX31c4}tptnSgLe(xsp!8YTdFmsG
z27wZI7E8|7^mG~^aDh7rTCM~F1hO-MD&WUv76HG>*X{-ox)<ev6O@R^;0MD9o5lKV
zU)}@hvg1qO8#oyN+$pUg3iSQ?NdHVD@NEwS30kleI09uiJyP1+o)OTlzeC-xqWpE^
zqQx`TEu3BKyqpOZLhw7sAw=Sywg(MDN;#g@dasQbGQUww)=dt;YBbbhihV>f7)C|0
z@K&Cu2x_#Lhr?$_M&;#SP`cU|$#!f47X7*1f~)yyFVG4XR78Z>inF<P0!-t|0RucF
z!%5a7-o-JA6NzOjfB_2NG%hn<oK83_BEj5?EhF^6+7+w`dVG&9%a_S?!+)_}ayQg@
zA$+yTVT=$Ao`rs2IPqbYg)9P-*r`($lLB|~T~{js-vi!OJ;`j$1ue%t?eQ5G#5a9}
zmC7KTQ&s`VZy!+P&&>Fn%Z%&_9@}D<-P^YE!CAB(4D^({aI?crVf8STr00j{_}v%9
zb;7pvcGJ$7=D+YxpysLwkR&GNT$2>q;jDgrftv&YsQ~<R)cpeffMA2{MpM)XZb8JR
zWUzZ0ZLrP{hhRJFIzUems^EB2>&<DJN;7H{ye23M`+(Oz_dsxA&5BjEV5_K9mNRCj
zm9cC?V@9)LTb3-v;^UoRsa8#?GzGB<n24#`L}lfoUViVi#o-Dqze2G+*0lO3aqBV&
zhi25}ylE?~gIdKyydzaZC)+?$X2VF6Vr{YU@(MYFOmi+;x$=s9XAuMNh4gO?wAyn@
z^D`fh1_njq&Hf+Z$h_HUtGcQ=IWN|LL?M`@Vqg|V$;r196}wa0iFH@UGh_B9>Tu!h
z0bjPnrtKd?NkuZTtK)~t3@0zmY97gjH;HTM8kv^@Xw>ANVkS$jN@9x{RWDOd?-jnQ
z^O8g#Ck=}7T`QF`sg7sI`TI15-&(KQRylRViAi2wqLoccKef5H;aklSAS`iEwCAzU
z^(qj$IZE@5f@-fwAz{2$>#n@I=${Vmq5-ZlBwqQwuda}S;;YnGzirya%{{B6J`2%O
zP&y&->g;yQKv+h8MSs^;q$n#G0R=$_=T)1XrVbNsPdA^%0z9sE;8wGyRe<fXraA|U
zuTiIkqi~jzePTZ;5r)0``Yd}Kg&_CJo^sfSx*4Vk>`9ui<aanqx$rdApVTN{x^tC)
z$Ra@>oGg^(;K|@&_%#rj)-Vp}E{(l4`i@BT-!$D>7{8TViPsxBIk0hE5zRu=?Se5m
zFUC23*wB^&=|diJWZKs#Vkh<0phe`hLN8>zr2;6?-2Ewy!m6A)i0X=<n*5mJ<V)^b
zgm_IJNhw8q9bK#}@YswNyydrSxKvXZV4nBlb*af}=1tbTHEzTznQgvZq93n&bFPy8
zu5=zF#SwpNq_e9wr<5o!e3Vtr#)UAg<Lrna9;X#rZZ_Z%F{acFC)Dka(1ucV=C%BS
znnP1^%|FjV3X5+A3>!<gd&^52(cQ+CQgu&XNOZAVlzf&C(FJjN5dUibS~~jKSlKn=
zF+xN<3v`m=x|p*FE_q*p!RM@3fkByAIiq*>U`6^>olsF4209Q`6J^(2l=T5;!z)6I
z(A@jqm>(ylVqR`7YE8V*&SCPlkBxcMu2X4IEvxo9*EVy2Sw{Os(orP{M;BU3$^doR
za)`h7??!Nx5AO3<oD`QIK+z4QS>a(Rd6h~no;o9@A75-%ORDG$FO4($*<opE?VR_;
z{CR9o!K50L=99vS(2<G$%NoL(R*m~nO}Z=|L}h|)U<s_#7PXk<+ke{LEnhqT-EFa{
zdX`xdwVx1No9xe2qpHj^%%E|wWK-9`p-;I~W%H<Exx3EmN@K(pc}lr4)rK3(mR?I<
z$Xyx1)@^USQAwKPV!ZryZd(Y%LYI<?V4`btzp3S+>odqiFbReQ<7c~TsTOPFe|ERI
z*2S!${E}u_RD=U%+(|+-;Dn{IEC6QI=kWEJTUULe-AjMs`x(zYd~69Wh;F16lz}6X
zEWrOnxwBX44U0SqMmUtZ?HlpT0@XIW>*&7jX@4(S%=00{<_pNS7rbLzFY&pLHU7|^
zF1A4Dp5hUW!Q*;2Y|l_ekR%?B(*<jak)=0L(EP-bf~!Dkns9t6?rKx>(L$mu=oSy#
zgwiEly2>>L*0NOs+oq-P{_A!s5(Sp5znN$#+yAS`s_*7}Rq_L9PjLXm2Bv)EMrOhy
zP*xoy+ha@77NuO~n`)xR+^?_<E8wmnc=NWNLLQSS$F|U!2SoDd{+{N<R!&YaT6whS
zdwP1xGupV4=(gEn);_>ww9b0q)buPIwlu+p?C(P8|Md)XdOHasKP=l!At4FnR)rv`
zK%cXY`{9>dR%$sQcNHd07IHY>gc(Kl`MbJ35$^SCMC}JyHDWgIKM9=Q8u|YP0~y)b
z|2u)xr6pxY)P~sop>}GHH)(P=g;4__(F{IH1C56&?x<;u>a+@24;lOU`9BdZEt*$j
zrf`l@TgF`#uPAHDA$sulw`bR%t}g=>tl<9_4}`PJf%k_e{ss+W%i?hd*ki`t_T~CK
zl8`*QbsCO|Q?&YY|HQqujETVniU#-uwfBAmn}1dN)V}3Nbnwu{%AipHB=>e4mvjC<
zjJ;EkDB%*V*|u%l+HKpmZQHhO+r8ViZQHhe_r=US+%pj;q8>8-sK<(`RhhZg=Uu3P
z#0MwFsdsk3)0edu=jZjgm_vek@f@axt@_NM`j!trgB`Gth*ze@rieHZnh&Rl@daY@
zyzl1S(Dk`kc8S&~+Q!b9NQKuuP$fXD&JxJz>~gI8VWXQQz#S{7B_1`xV6KY1fXw=+
zJZ(A;vx29>WRGWKVN3vZvng#H1Yp-26agm6!P;<E-o%5NVy$>yk`DT$vAF%)2^o!q
z_2kCT-$#)JU<~#0L>wZX4*)UzzFQgRIH3Lp_W={l;2;4icjLifnW2rUx@<m_jHaKE
z$s*WpnbX)bSk0<Qy|0kz;7liRk}a)f+ATC*n~bv3NqSr;G9#Xo^I&f5s4;1ocXI^r
zj535i@thWdOw9%=Nnv4Xn;|FzJzTCZ%)q1)EzHFi!o<Ya-IPyg5pCfJsF0G(LP(CJ
zFxeAnv(m0sP+S*rv^>&Xs*i_Sl|8CEX$i|rvK=CEnF$y4{N-|~f=iV$N8*wqSrq@|
z$N_axt+Wa;N-8&%U`%XU0y7QFR#_j}-tgENaP;@QLYG>Zd|_%yP$g9gn2bJ6%$_Cj
z0qh*$QKZ%V(2tUry0^_H5`<)yh5Eh*QI%#ZH~t#AL_td@uW}^oeM{%j!<;j{N)>a$
z4PUQI_t|{6R#W@>FMFTopMnjeZlsZloMtne@3192WRc{*`B!g-{2!~)k>9P|dsKz~
zW+??QJ%2MG)hUOWdcGuCOYPLGwax7|-IonDHJ4meFI5v3|5!`8=Tsnc(?R3La5ERn
z4;A^7Hg736smCw8xnXL`h|FEois3Dj5(kx*%<;nF@FZ<E?`Z^Q4vI*y>TBrCWsfCs
zBpYh-m(-Kcf{f5U&blGdwVoR8vvvw7%0NrQ6|6<d%g$+47#G`rDp_EBhm|C*GMgNG
zDjvau-o%(-2^OB1S+Wtg@rUvtYKuNV(B)T%{sq$<p^8Im??_d{Lbe)6qzM?V2Fra3
z5qiA8ub-y(b*<P7G#>7I6|eJHxG_OUJm72)X<uuX*WJiW<tsCPe_VW6tzn|~>j<bX
zgp8itwS=o6;GQf{qEBi$mTF`Rwt5MLsyy3_3GYZDXVR~O^?EeMn0^E)ad6^EG|G;d
zF7C$cMakGx^%u_AO|=;veSs+aShT_}18bJ$fRmuhvGmg<u`@!~q~AAN{5#@#xZpUY
z+RFu25`%x%FlCc36X$sfF_Y5fWwhcJma|H<A{=Jlr(%~yB|@Tqyc`LCW=kL?{in=E
zNZMc9F!n%GUM*VQTFHEwT@p38T+fDkBYRFwX@@iVv8vf3*2tk!Bjx^BBcWa}Uh}Y$
zY((a6QT1%GP4(~9GJ?opIwr$sC^VgM1oUZtg^BqUb#me7roP+qWq<hErUl%Eg^b=&
z;jaIU&#{?{jfqYfXNntYVPCjxK<gGV%L){<Oh?0E7=Dv5R8c!chBlF&`kn<-*?G4Z
zriyg)4l+WHJuRxl)AhA<QL{ZSU8LFY(6+pmlqcg$g&Rc^O!itL1_sefS2}1e*0u!b
z(H<}^!i4!{;9ymhB5Ho)M{oeH2|tK^W9UT+y^d5y;^CCjij|AqfEb}rcEt+lwO;~T
z2zz-**YXK@!xXhh(0;rtq@fkN790|eT<o}7=mBY;W|ujL7t=+_PEd2LNLIh@s27G=
z5N{Qss+kt{-9VN7@Jf_8*b(A1CyaT!FrJ;wxPC$p16)i$mlGj(&5T7zX@Jo9tM3|K
zoCZMEqhV*O!qe@Vi@t9d;U`*&U9fXcidW#EQI5gNr0etHfPUWlgCvSFD829K$shVb
zfwi;6(!GY%bA(`39-s?b-EH2Ucps}{3)EDh#joxa+SaO2%dnVuC_i!7V|kbOtZlbT
zjalLXpj%Cjm09UE&qV_BRov?zX%*Uz<jnI2i;w3A*Vi}^DnsHx_8nZ`_iH*H2CU$b
zQsw<cbNJrh@4NS#S?1*yP<j)p)N^>>$rPuyN93FZkqEkk84a#BkiVDBC;2hxoG-ag
z6Aufru#WRuii>cdVHug8bSRGz*<{GuXrHB1W3W;>LM>^z<}g{eiCiv68#5@VpeCFe
z(!0%CPV{^d>*A~->|0@dvCq-ohKaVeuD6oZ$H1*0hQEiKX`1qAl$RNUGU6@h-E{mx
z$(WH-UzP#pZ|Yt;Y3G(~tHiBt%QC4tq-iNyb+5^rr-wz$l|XIKm%0X&Hk<fQ+Lc|=
zXg2I?iA5vUgby7GDr~V17l^R;e2jRf8a+TvR~c}XlIMCs@DnB47F>ka03m7PQn#bW
zbFIywnF3J98)&^%--<I3TpQ9H<JO{$s@9@%?!A-n#5mH>Krwteo4YmnMUrzgy6^*U
zrfzA>4RAd6vN5d{$^jl_Hy$OOM@hr!fcE_zvW{lf)k?PhZ;~+T#M0(I+Pn1ZFFG}z
zaM!0IF1Vp~qw|O`+G0&mldUnfu=7~)a@FeNp0()j8Kh$pBpptN)e+@O_B=7^`L+k>
znldTAX1pd9WrXN-3Inw|nSg$bMM%T3wh}Ggo|_iyJv@mENNQd0&Pg^f@vhKSO|%o0
zc7uX3=o*A!%nEV>(vNy&NLS$4?Wlg$AKIAvFjv$<*x|mI;p8i9Hy27|!R{%JmquHE
zEx3g>WIPJZPj~h$?=`fBpjHdis>okqAAz*G&bVx(Yc+L<((8E^WyP>_P2Vp(I32c}
zI`CPc25JA+A2#Pc%ld|_hM-RNJ*=}r)m5`)$+53`HNB&}XYk{$0k;UratsRSP}Lyb
z<FO;mc=q@}Q{>cj!b7m?a27!U=c@>q@I&uFY?yWae(*C7Q1RW02>R>^_*2DS{+x!%
zkcH1=Gdk~S%RhdjXC4f8XcNFFVwZkFq{L{Z&{!hY;L$kzPPP~`h6{t@yr|eFTIWQi
z`ZHjN(*3R<6*6Kbx8Cj-TBE?tbCvtMQgE@n{(jxn!NYfUMc7DI_pG1GUOQoQkLLKd
z!=C8e94WnYBxUXGvcbR#9?;647i^!(3X?f9J;AyeYEdHcR6A+ND8_gSVPt%%P&LwG
zEx2XCXSsN|gSh?P2cT;m=g5P(#A}+Hy`IJ6ASo5+4+riR1wsjoMTB|sm*3p-sqKEc
zKvNs&XH(d{kz7MbEAiSg2Q-o6j#An$k_*Uc`*!am_N{2wK)Z%Su?luXNUfa|8i|$-
zCev?Tl(jOPQu3kQTAO8zFokBYRs{>3-0&4g(Wkv!8nt`ZSnj%4U=ugmG3W)vYCQz3
zoRy408o*gMuxet!BVR_2CaXGK7CrB@EKNWnr?wC7<MGWbTJCM_&|czHxL9T(=-_en
zRBq|b^Vo`&3wOqyd~{gQ#VijSEbB6^&XFfUN^7rU5+sCiY#JKcj2aG<d^OOTXqFES
zRIGZw92`Bj7)sG4^(kXSI~a{K$`ZnG5_muK-^>rOXSE86%B0(<?SHo{DH!x)L^}b$
zCO7`PhQ|mHpbSY|J#hxcrphi#>w6rcZ6-R|*oWd%(d>|5ZDvek87TjVL5SXsJb;ft
z=jm#q3tUI?{nb=2BG2$jI;A{rT)ns-a0g&)^4E8Vu6JZ%ZNbQ8Bb$72T%&Zd7E6fD
zbVtrI2J{KxLV)dcPb$~d+qwCHZk9^I`cIPl&;PyE#>B|L_Fv7inz1?)H(CGo*6CZI
zgB#Vmd)l%PpUdVtpr3M%rXY7lM6qYo^YKNI1I81PPN<!^gVe)cL3cvm?O*nVxw(a%
zcCwKt2N|daAPzpJ0iFSu_4ZlFSM>uh#4vFCVe|6Z)3JQRJtbumFU-#Ursi6@#pzps
z&0+W5?1<kz$~|!{xP?CgX=h+FP<%jsG{Ec4@f2%h`(~P2BHu4y<d61#K7IYfmXSVv
zSrTJ9q`W_WNgSvl!AqKTEX-riD_eqx?nHprhI2`5Nfs7K2i`gL3B7A5metAMA>=px
z3;O370kJ?r)0%~MCeqn2K{jB%xf`DC6DN<(7)#+y&Bz7*0E!cUvy;sS8_h=3T=_tj
zWr!ewUxKNge7I`>IG7y+sTsy$0b(Mq5Ac0{cLACYGM&j+p$#|D9v6_Bi8lgS@766Q
zB4HkBR`vypcfbq{AQbImtSQ3m`FOqdv(wv!Hq!l8H<bVY?u8PlV!U8V?ZRA7-I;}9
zqP8HpQD{Ruf^XP25maQRR)Kc0UeQhNlaSWe)II55saq>B=WodItO_>fqf%dD@{5Hm
z(K@h0b}#{-OlHP96%1txh98MHpema%ZN+g;1}o*4$xh!AdDwHOmBn(Y(JDh+<_|^0
z*`A$+E#x2-I|lW06$_t18x(W2(y{k&Pcs-`!v_+6zaiSW8cS%!0_KZ$7R6g#F(AJL
z;118^$^s4<wX*bOliw&)iHAAj^^Hjm|266-QOciSc>z82g^74znAu%!(Pfb{6;n;S
zW)<6_B1%2;+zLQ}H)zZl4sLRQ<Z3QhzoRZIk|?dsJ#~RzFTmrvjupsA;(Pxd=i70y
zll#Ymlg>B}#9SIYwR8IkBH5Xb1f(dRKYQ&BrFZLnR=?!g0%}KI)%<f0@1GPWF>r$c
zo*+^{z({GK`TqEQITzzy?y454D892}fYh}fRX1=3OzYtncE5EK8yGN`v@Yj#Ap6%!
zKom;=u?-{Ko=X|Nh|4%+bc#PEJ0zOWmxH8noXZu8=u27uRmN3ciGq@t*S_R8^_tsA
ze4orP?p;0zm#Q2)T0zoSwmX?I@sc6Hy&y`@rW}QjU$%V9uzZ4{=I^Nbf&zqP`X5b6
z5lW97{vG2c{_`3-QLWK8V`r5w*;2+%C6<Ibee)OIoKqRMt?Q8x5){A|HadLR?2h%e
zIl!dKa85cNFIm=VSkd7MFh!o1nQ~9$E7QZpXJ=F<=ax|v>gq?+GRmyd;@Y6XHs?Rz
z#(g|K-O{^QmE1>l`L`%e$pEiU%S1l?w9MMSKp#Lb>lebefAvb~DZc-_sLhE@7XLy1
zFwN)zn%MFpa9qdnAd@wU_?t9rVfxs8ExTx_=alLCLi?}KWXwEZp<Vh{&hu%Xu(U4D
z{;Sy1IHsqcq87DDe-Hl8Q7bx!474)Gbo;euJBp|yP~NK9p<i@{=9$g=mQ2fb`LgU1
z=`b@?zx`9IlC+x8woxh~gzWeqXNWgn&~PTF#r`4LEK1x46!BP@+iL21p@Fi63zy~x
zW$995)5INV;|=Ex<BR_BMd{nRiK<SV--?SA{fSK{#}8%2nqk#}^#PQ2lB7Q>+Xmj4
z*ejKy8gHC}=8P+dQJMr?$T*%`Rvk&jEZe{#r0xd`CVx^U+L$CS&DO#y%a9ctDG{ZF
zhm6<RRpyYY0)tPxa+=j6*RnLb?sO|SV~TyH)Q!ya9Y(VENz^wh)k+ho-E<acG<uJb
z{w5w5j<Yp2^y8W<--p$ULNSoFte-DSIrS>dPokHIpN&avr<f8o>ql;)a*coE<93o*
ze0K=vaw~G&v6EktxJ`*0In^R(Zdh6y8*4YznmhJVKC+>W1qiNi*4eb%HvEGzw~NML
zEP5-2k@Oxsv1&}e&;4m`-s1i1(pqbSg`;-J4pq4na#hk?Xi|)u{wZ0VRG*8ZKm%%x
ztp@zJh7$#AJ<0_y*mWvaOvsWKNGHYWeT5J4njvr6ffTj<+c=v!P>9>)apL9F@W*cn
zTTwuNV<Xp4Rd!Za636RDc;no1^nP~qs<Hsqn^z3TQ}8)kh1(qB-LK3;s2!C)a~5NS
zH5T@W=9|w*su%}K2k&@P&igHix2J5phxHAy)R)@$?e6uildT4IC^dSLSd=VUTW|`z
zALpF25LbbMguDpA`%fe^$D;<u@lKTaICl2MJzYnfT+BFKuwiE0Hhg>Ms`iwze&*mi
zY>zI#5u^x9*6Lp(Y)!C{wFjUEK&p?UbmvYg4N2H(Z!mP$_{vu7l4a-04&n-m;}?t)
z%f6a%r^gPpqj4UjN5O|9@dPVw0B55EWMc|0%3)i#L=(DuRnkTe(8)x=%ktD#dScA{
z>vf3kcu}*ciyq<kaPQ865n@LCZJt;L(*qsY0{W!g7nj<xxWXQmn~)>3fPX%)Pa2~_
z9dwYiD{qk%PwVKaiY_T@X{J{+n#3!IWUsj(Ki*WXKWXO%ja{lw`LId{Axrw<RQPgc
zr&j(nx)^aAob$O-i7Ub_E*4fuNzw&vq|9wt!zSbLVo`0Bm=}$;&~!0?f^FQ1&Cu8N
z7HU;7&j`f?tQB<|9Tgy*7p9_KI+G`x=uugs85ORAM|hYm-?57Q1KLuR)r9K?wW~_Q
z&ee1Q_a{J^+-{IGZWNXsTC*GepFgo1in0!(jET$yawgm<nA2lnQId4G7@Dza0roPq
zQDMNiE%qKK(vUeoqXwIZf|XgEo&>G<L}eS5qi^n|m~xJ#L&aIRQ?aX;$jjZe!&udR
ziS1wT;ZR-y|4E3jGW_4YL8kvQ-Fa9`Chn0H?q9m|UA+jSr-<SbI0FPjQbrfQ-DZwf
zYuR>7nh*S@)aRqrc!TwNY;U4Ie8y0NFi$bXeL?a3^lgl74{x^*uTOitMfL*>T-NL^
zRy?}qPhhIxWDKx;YX1Az%GX!nAI<l!@lkUb5`6OBZtutQj6NyYcyj;6MrMRdy(6d8
z*SQs`Ysqbrfa9QtlRw=BcZpUO@9Gs6I@D?t{7{71zVBc65@x9UJqCB*_wP@=W%Wp^
zL58@Qetuq&^e_QLaWRjImRWqhHojgTVhM8+D-)x753iTm;$|sg=psK~QMNzC^nS?`
zDBh1|l^-?EMOnYBzsYbbAYLU&`ba@xr2QQNsNQ3SUWn2H7RWDAYvEIc4Za#w=BVy8
ztnC@#4$L;RPRFJPx?>~_8lZ@|+cX>r?EVRMP8Pi!e2rcDZl0+EF?1-7wAfO|#I-|;
zDSQuHi@aG*pPTmdj2*g`)X&B3p|2~=V5;R?6>G2MgsNAQ`8WgJuu}&|vcZiDOrQ+`
z8Yc`sO^p+0F_!ovt%a@g6B^#}J0ld+$ju&M8)5r3<uL*tGVqF7;^zR!-Pds4@P}IS
z*j*1J+~Rh^pux!9+yDEBOB8MvenD$R!JcJ)^LFl&q32I`%tQY+%G^uj8pfbVHEfvS
zTRM%kzT#D|gh58)ljp&hRWu(D_Ffv!n&F!#EomKx|8cxoUyrjvt(Rl~yi*Pxgd?2m
zsXDK<*^Z-)*0VOy@i_<Ba=ZR`%TmkW4_8G+cUJO8bRfCAORAr3eOYGh0(a}5I!w2O
z+>NydWdt}WYHj%=!w}Mz$XlP4S0=kd-7Hzz0q2=isbUhDhVgVK`IQaZsiGT5_~Rez
zR3tC1Je(dFaq++7F2|RbXBMf4k<!OFaU68&sO9!s2}?{;)k#eGV8O&JenU_y<8sti
z@zM;ef&99Tc~o;~r_zI($R9K*ISm6aVv5gNmz)Zf4m?6jX|ey_mrQ!n&1O0kI`~Q&
zt-M(0ZC?_i<&wS^^@L(8oB=PqRS=Qz>>A8s&7Gtjf%!&yIKR$^@+lC(^lAC0V1V7x
zDE)0|;x>)5@bx;_k}$lqZAO%*S@#Jok?S&~Y3++%haddfgK`Kc`$JIruCMYWM3k;c
z^`Y<eWXa>~_v>fJA6=W1;^E|B5iH@9&AGR^5{R1-3Fv7#fGzFA|D0CsWa5~bO6`$|
zt)X)mw<aMLn5%AYJ}6JeM9v<ykl4o;V1CD7Gi(j=91qQ08yGP$EDjU#R$2V3#(1fy
z%OaY~lkmHI#0PG^m^or4h;IAH+Hh?nVUurxh$#a!yp7>;O>Zp)Ja6J_1!QW6MIzoz
zUL@mLA+6kP!m5Hf>!7pwZB=#q87A&zOoHjdqmyqmkpKJ-Z`Hzf=$76k=(*R!5jYsn
z5A-qAnk8_es^B@BG>ei(CedD+5d|fPA*vbD@pK-wdIu+2m*w$x=klGnX@-{3Gq=yV
zl5)-+nYe(H*Gw7{jxM@3JB~a#K;lHHUpo>G#F{C`wef>n_@qneBt&!AEk}P>CxZ!F
zEb#xp+Cm=iGq2vV4`Yc@W+YGoKKU6Ni;|_!yN->qM&EN|7bT;|kkmFZVz~=xMLOZc
zJk;X0B$r@^Xt8HLOY$NQsflM!X^|&6DSz&e&@NUf&~r&>YA%yjU`VwCOH3M@5xw`2
z$drQ48ynjaT4Yehkm?dzj8NE>8s(^O2(T)bcFk$CutT9X0zv|4VGNQ@0z&pt$Ii<^
zCopBbArMdtE4?NHlhGa`7YSzbbKPO0QR~OpWxxX#l@B$(r9y&3EHd+64V;Px+&x0=
zGT{TE3zKq1&HlT;^wa>JTLZ{AH4|%x$%+4->&_N!hbd>MNvQ)~Z^|~_%DXF~B%T5L
zA`bt~QL>&>WS#Z5mkQ8i83;pau8z7!n2XJJNfGUJ0X(~(Ojry+`u%wYyuucpt`-sZ
znPIy&0sA)9rQ4a1aUO>2ECEaaJ?zOSrb{}=CXP#~L@=1M5Lrv;^^1{E&$Rrz7LOj&
z$NQ6!Bv``q_h_^K1m}!;V_$1rj1tX39&8ISwmE<=!QP7c-RzVXxlBW!UrNDvvEWZ)
zum@jBT!T19u~FHIaH4}i+i%4m1M#YH*sQNk5o=*GMrJxVc#f9X^#)3q!qX1iPf$G!
zJD<Hwsb{fQ=&NEfy^8Uq%ooY&RS6lllG3cDEeA*Ets!b5a@{C?cr4EF{uaL2am;ZR
zHmc7Fo7B!}qOb&dEs!7(E8n?w%1{F3>g=4EH;sFp^6pcc#@mkRp2;kqRlWP|mZi>x
z(r&>f8<03HPZN;tyZgI*C4H2R!Q5Kyhp|J|>#CzoM(-W-l)*E#>VxRw;1>Z^2U!gk
zwHyDfL8`F%I`!d1qEkF$RP0x)_X*epFZE-F57a905%WI2F{-HvnODD$=2O!u9MN7w
z)THz=qiDkqZ#&nF?sW>W{<^i`E~XnY#SX0@Ry@4ngN8M(9d3bXsKA^~52i5`nUpZN
zo5G{Su0MGVJr@l5GnkX}0}P|f@1Nb1Y@J(d_|W`0X>MAdMc5P_do_+aeC#*k6cztQ
zA<UZsc)K>@<k#9eFJVR;l;SJmrjE$Owz=YGU4O_y<B$#~wY3|X;^3<&FN%XcauxsF
z_;%S5_SWyr(07t%!)I6S$eA1-(Qz~kF#vJ3uhQPoNnI7dU>;1dz_+c!GUl9S-MHMM
z)gxkJfyPNR@Mq_CVdL}|Q^_VR7xRQqSvXaMy(2LGxbc6<xM~S=R4zuA_@ktMj?<do
z(zNa^mm7Dx&#g%+Gp>fNJVW!erh&J@ZW{DWhC%>g-u5IQEOg)LqX_lJ^*I5S#FKEb
zyWP^c5T#F?b_V_ZV}z?}xvr{P<*s~(pyi=BWfB(9aqWNyLWiLgoQel?^CII@#jJfB
z#QyeL?9|DO{#yY-zbWmt21cz`5JWT`$5=Dy48pk;N#kbt_&#Jc>DxdZ`ZJa@i8dZj
z3LAStPW>F~m&#H3NMak;^9z;{1daS3CDecD<*@x*R{n1sirre1iI=Q!+plV8;B-Rq
zAowaewHL0fb}FvSi=E1<84|2c=t(lZe0!gS<B29-Xbm(GKmouAApDVXe$w6_KMz08
z4~rs|;|x*yLr>42AW!~_^!6PU4om>jQt3Uu(!E`dbS${H&J;5x#e;KxQfAyfN%|F_
zbEtjiSqgibdS~p5PT|)BUL`j9{>~7d4`{jKubG+-oGi^tWD6x;{A~NzPnWB+^xP2u
z&Y>q?@6QjtX$lZXv5*0ac6_>Dz@z~Z)MRskm~7wgdFQWbekf2GOlB*;MZ34S5|iu+
zOPY)8?*Cv56J>sjgLxJ8&aNNSEZ_TCPA79Rzd1i5Pae=6pg=iRF>3*Ev6!^sZAJ@x
zk@-%6;y$8#r~wraR3g6$V(IaFO2rT&098?8F*om4sI%rin@Ou-Vpt}g-f+o6%)eNp
z;nu`%VPuk{ZH{i?J%E%%MZfSxT5moK+XHoqFyTKzG63oP5<{>`3n<8M>^_Uy#fTUv
z&=TjT&axA#E3Ho*FDHo7hl$bzms^_v#-3Z)NTP?3b|l1tN$2*sV)%TwO>u*5)_b(1
zz~FLNj*84C)+3u<rT&s^qMXL3z|r2Q?&1YhlYeaG#MCGl+C|`-#AW4`*(j%F)8Pl}
z0UHsd#qUYdL770*N|Qi#c->P#xBId8kRU~K^q=77N;A%JZY9<r&kg3;2%dvB#>fSr
zA`9q3FZb_4KNC)Xw_&_j+riaz{N@082)UYZ2Mkt7llx27TRFHrellv8WrN}W^|~0y
ziSl!U8PcNn+Hehbsob@)zMp@A4+iD6!!T@MBy)inIs2OuK@F?tr8Cw}?s`hf61{@}
zmItYjs~v$=+u>IZ5A6CZX%Oe(<2><XTJNfi^dFMT!=9$1`#hG2cNfH=3hWsHw&Uof
z1^(KJAdm2N<E5GblScy%JB31jIJtb-6t8l<v(B{YOqEr=7LMIuAKQE%VIgyHp6M{~
z-b=pD+>3akfX~giAF~r+^H*Uga)X<jWE_`*at@v^mha_t^>?Euale|bdC(_JDWgYc
zY3P$jhLvpdD}+)Ua1cmJ4yr`dh=hV#MdR*wU<>;K$BTjP^eMcl?@LHPK}}cF&NssQ
zS>5#BpvdjY2XL9%h7GR=Mli!-JLK!nm^MgNZ;Km7H7ZjRHC%A6Yx5ld@7D)56?K(X
zm$a2}IwDhZ>1NdnufT%lE=V&-%oaeAjCu5`hTSNA{xc^wBtfVeBn?oK+zl+Af7&;z
za~A2QM~{2!aT>{9Xmeth-!Pg7;-f08N=Fg>Xs@Mjswi+V7V$@lr{?kfriXI<e(9PN
z?)Zn!&sFUSi7+;^Mp^(HS-qdXOZzzOLVn=qe%##1@AB#3=n40I{56X8pLCENUBrqF
zq%K7`Vc(g~1K4kFFD5us-z@T1c>z^yV*@7SD#P|HN5OSMF^)#zLsOwKhr4_Knd?Yj
z!M&sGj}9dLO5QBfERM*X{hMKSO_}XATPgu=7wPg@pQOz!%#4j^ua1p02hF-DDm#5u
zs9H?0N^CZFPYUijN#);!u=Tb#K|v#8Wv2rwzLNkm#_nX?(8kO6(aPfXIu*y3v^JiR
z=X%d^o>yo-IF3XS7)z*A0kW1owlqdI!oBcZlAcX<U`7LUa>THtE?sVpp7$3O2p7#A
zSnMX<6`sEFpKV2G&CBjkG^~U2!XUSHQW7LR;X$IovES0c){fOPy98FjLYGaHM@;$L
zA1`%y?TU}hglB3owfg9rX6LZjmaKl3-mAg<c`(quDaqt4I#o)bOZi{cFB;^IEa4a3
zr5K0PGwL22)Qk$u%r{?_QI9s^N}~&B-9ZhDpd`}Fl8UEGA_hgCAd-U9c{Y~5njH39
zRRs9d<DEw{7bvmwG`-Kpg3h^P$|qVq5#DW~_6_mS4=HZsOm`7is$nQ3QMa3I;z8tq
z^?g@bk9NhwZDsAi40v5#hHu|3OJ50#oO-tn@KR;5H`VUEMF^oH<ilENt(jw1v_@e<
z{XA$)HUaB3up1Q-`&LaiZGKvt<cBOz&-036`>SgDE)2K!LNGo>ptoH4%LB?}vWHIR
z(^73vRc#tK+%EL13;~ap&~5yxOhZl(44)}?;LbrvQr&gtR|Ba1gG#0FJ@C*tj@kO}
z3X(6Eb0HRE*`x=DUaOEVF9Q&`i}}V;$IfEDWm>B>2;`nRsQ@)u2v(6@9uSbNGSZ8&
zVZU76(MH7vL9|H=9q;X3r<Md3#VG$Y#bGsG-UUWp1zSpTX;m+@uDUW`xJSypW8mt=
zhIcJ^UU$U1cf<*30neNX(iueiL82X9W<Jd02|q5csm7m>aXM;I{o;-7c2&6fJ{%Et
zA^J&&Z{Fo%pvmwFGr>JLT}r4$*U<3YOGFg9O5%w3ksfb70HvV;{^d^$N2aKc(y8U;
zu~zm)*RVC*kDCfUsa+rXxVcyV3E<~#CEw&8ugcY@n)vn(QQbhmb@l|7>HT+p2NsLs
z+v9mpww{jKQQe#2L^3sd<{dGdM-~E9QWO(9*^~y%4M7FXh{nu4*l<Z1C7kHwm5FId
z#tFyJ5z@*K6t*%1!sIVBaBrm#VkV2op&?7RGLt{huuLddOe`ibH8JXo6Auqhi5qP5
z@*Rx_xetrYK(hM;!dV>UjMs2k9ksX~l-zGCmRtH{wfRs+6-kMl4Ky4AkX#*HQV@>)
z`E(f28A8>j$3Rs^crgU<1Gy6_)apxv1wi(6qj5}kDn+OZOeSVuL^N=ex8eo0!1PD(
zV&qi;4;u(lXK*!XE%IK*kFHj|%-g?dAFjgfCZU)0qy00Xf6@}VOK3NLz7(b{hQk6#
zwVHL*D;?1i(iUdAJwO*d^b}={6kFk;=VBq(V<WTuQg8vNMmt^V9ZP%Ld{>dRj$3ZX
zVXcTqQ&wHvW)GAX?BwiBEGGOK0N7dJ;QU(~Agdwoc|HUtgAsEATUt5j=(9#Jee>ID
zNHbbAyln0dD)EztA)csVm`itJFBfk61lfw(3T0$!nsBwi6>jA-nWP2TDko0h%QwrV
zRBDu;YHQu#mEB<KMoI8I7+tS^;c!bt))EdPV-q8uV(!2FYe*N*DEm!(u>J%Vq(}Jj
z4CuXn@ZJn*T5Y8B1qyitj+W^Qt|XrJ^PA@{*xVh-ta};+spsyW=ph+<$d%}~kUYRb
z&0b&&_aL!p7&}MXiO+khxW&@V`*9M{jV=R3E*SdMVR=%uy}3-`jo}BmVO0c~tL?2M
zD|^I;vc>x4Fb8Ewl}2lTP|z1@FVN)hL?$S>{Np`uZEK88hv~BKJTVS%P!-w&9koTP
zZU`~IxSlUmSNV`Bsc%*(PUR=F7LitWu!sS}TqVt4b3yLe`U<Q9dx0*fUtxB*_pvsB
z#Dj*O!QBu^m4pS10H;f>uZzDGc#SzK&bc{?CFcUK1?-J3=U3QqpY$T?6;{<{+#xQ?
zbhQ+B@3&ye1rQ)W0vSoM?zVSmC@?W}gKQ~dpLe!RyjNl=B{W|P-ap`y?mlQRHeX;=
z|5(0Yssg{<YnIAqwihVX;~3T1Jwz28K<#$Eez?pynZVkx`}uZ%!OFet9Q`Mm%l4l#
zxEcS;k)@_YoKZV$&+5NU;pv=Z!Up1?e^m!}9{>l5Qfog>paf^y6QR@iE_}0+=R>2_
zIoOu#`MJWDux~L5{4ej=;4FL`J-i;BB!2cC7!ICp53WC+_y;}-tfrGazyK9{mp9*+
zYcdk858ioEiE<=w?7vnXYuPLm_cXu<^xo6DMffBAlSkoc{AtPS#+GpZ9QgS-2am%3
zzcJs3dZvX=sgx0a(4G6o_i-AyhHyO5+8eX)v>#6|_t&cd+878qO6DN`F8%><4L?O3
zc-%nfVBRmMx7XXIbh6bXu!WSS)l@!hO$a==RgnC>gla0GD{MrW1YjtNzPQFmqTeJ@
zsD0#0z{75y!@R-KEuR}?HhKVio@zo<U{^M;7A9)#IRAx*qC#PLL?nZE45y%#cRNTs
zQZx)PLEChuLSor6RcE5M622sTcGDx3rf3BN7XBr@(~tUlGFtha*kqZ)oNpT)ev94$
zOcg7XDzxJ-Y+hPv+JRLlO{vg52f9m&-=gOFZldCapG2fV@Ou^zvPi}%iB>hca(S==
z_HFdbLeCw7AK#4kd~cxzVN(BrU&9@1^wk<-yepdsgH@M&lT}ry&yLLoEVthFNJ^xu
zAoXhb+5p@6in=IjXYQgpOP)`2Za+03-BeYf_5~&osPXOTx<=mZ3b7<@Q$P%uV8YH;
ziI_0LhAqhl#1BLG-dCz)(^!+9oz~uoFRmKk_XR4a2sP0AvGo0=XDc=W<6K-)mH(|m
z9^^pAU&TDQ*Nl|{hPHJgvQQkIg{;|4fqx5l?j}{eL@>*JOg8ZDjf}~HgxxzB#D&n|
z5T>|$hOw)*MO^O={ZbkQ<Ly<lGLZlB2v|a1H7+)*T2!KuQyca#)!MP>e3pW9DSQ4U
zs0euxK=vS^n+FHnt<33Fh?7zjv0kqwD$);+`Xn5-%de6?Zyfxx@<lm?CgU`FbM)6o
z;kANut8KN+?AGnj(9T2~YLAnei<>x7KcUjhE1Ymv^#U3osPeZZX`!}8Y%q9%>PN@7
zZB(q6ha2nS$2kWN9dW-h)<AA(cJ6!Tk)x5Tg+R=){k?U`0akT)cFHlTW(dZUEKs6+
zaz9vf%=M)g=cHsv`nq=Wcq8Zem&sVQ2mg;V*xPDyw9gl&XGH1}^tUWWJec-c9Q6fa
z5;&5dRF!N~0s|F<<NQ)$YaH`j>cgiJg(EOy{021&p|O)_<LJAk4Bs(89fQoQ=ZReb
z81uUBV6GC3NDBvS(${BZ*tO<QzzLwwVpAwfdQ;FQ3r~;A9=7L_D8Tg-ys9ybUTnD6
zMv-r*1)GQuvmBdoNhT7_4}k{i@s@5>JJ-K6DK=qJ>rlmS0)+&95%Zcds7c$257S)R
zamji3r_R77D~T=4w1{eu$J91%a(^z)zzyUfd04ZG$MH81D?_zeT&^99wOQpW<xgKo
zK;BfCGu9%T9W03z7tT~}AK848!|?}}e(D?t{&Q*0A2m&`L!p)S;x4T`wV#!2H#}<J
zmLM8P!5oHWQ-merm9iW4AlIx;Fs0{;3QsDe)>p}SE~gjeX8Ct<yDVA4l9){6g+|f&
zJSIOCR}G3}XOMQfH%G>wy%5M)|LF-Kih7IXK;^4%GFX-Nrn<2XRiH{0AhvEfp@~OS
z7amo7E<zYxb&@Pm>fG6gl+B`ZxCYd8I5DgC!s!#rbJ)`MWY5(Q9(ik^NG>eHAOUnr
zygHU|{|jX?OANCzJnx$3XOruUT2>oNRO5DVSYIbouJd{i!AvP%=tf^I49Y6@lRVDX
zpFbBUn5vy!G35tZDTTN+W_H3Mm+ro6xFU`tKBYUaGzYaM%KU-zZdjJ-UGcly$-A$%
zQ!4LP)&VBg>m$vV_rtgvDps9kLJ?d^GA&+LzVz*XBfqvvI!1wFld-0{$uLaa*ry6*
z9x}!-KQe0yTNrWb+@lT?3VO<1yuxtro76v*^Hyfmc_1XKvMLishTEW>NTlz~ks%q|
zoYk46EJIn(n%OJM-|mzFWbMkj$~`Uk&VV&~?61eO@w0HQ$0q9`;5(TtEY!3MI!byM
zOp=FVYt!Vqf54Yx6X?FmOJ2wp#}8Hwd$ot{x+aixtpVLhlmtX|qgVAx!h16pxcgaa
zOb<}_Y8)!2xH#OdFe-;ax~z))KpO`X56wk@`(tgsq(x-K^=A--tA^0WHm;jO+mgks
z#b7;_rf^yK>J*NjtK2g@cB0ZOLeS`$(amNS$!w3=G+$`EmmoOKX7@JG%Qs+p@a<E&
z%JSqUqSb|1ZDuKgH>=J5fr$cRj->b0xhKlq$DFE*Gz%>T%s~bbD3xjKm94um^@ml?
zeUem-kPp)1Ua|sQz(JdJw*0qBr!_L#q34F}O|`CTS149ob}Oc1(V*#Ms!lPv&X^}T
zeFqkN#}Ytc|FY*r$#~{_)09R(r}o&yr+)@bk_%C%649<n^tlrnGOQ0NvIPEQ4r#ht
z<EkVo<HZUC?#$M#5&_I?b@T`%YEFb&P&zxG-n7x!u)MZnyYIl3xGe@fkhsWWSAeVN
zII+;eUTb0K&Ir^L@0o0(QH&SrU3wf27^9SJkAz962g|pL*94>1nOT5Iwk%G5S)f17
zJD?jEkGKiuuJ24XL03}GzMVtIo|!w3eE3%N%>t9NSA)S&+f!1<Wm|eemvsPAZW7kH
zSzRv9TuN76PLTUfe@L)}DIdB7z!Ds@Qn1Bzlv;ArCEF<vJ~X_5N!22kK1Q^tc(-y_
znbLLBRVavWy4_m*tq&+QjYX*fRSNE6dE(vkL_0%w8D(0t8%SA6Heuy#Z;ay~K|K5H
zsmwvsk)KrJ>s}J##F9N=nFM})#yEHe-NlZ(osj4v!g@13bj20N6|%-|8n{GUoA3#Y
z^zFF1LfA`jJMl-LN-=uWNQ6t?Tly(!_mXv0T6FleM;Amm+vv`<I{#Q9tZNuzpIR#I
zyH@~T*y|>shI3OUfsgr~dP3W6Lab(D9$tfp%+d&nJ6z|0BL5-wz@xMs{9!usOXo%b
zQh~syoE1lu(hpx%JG3)#gR`{*<>sfZzDctglu)yeXWyn@y@9m3X8`_upZ|w(4qzj)
zOh9B09`kD!&Z84-QrI>HhK{Mm?3?T<0P(V9f$=Kjsy-Xe;CAJ+C}`=Pq_{5v-V%l;
zQ=>Or6o*~N++IJP9<DCYOh0Zu8K8*+D<xH{HcE(W(4W5mEW|=ssp#j->-lbqvZN%@
z>SZ)%PKFl(x6+S1wx8H^v1Q9M$>3vf?hxeahl0tn3S8>|^)duLMx^H{x<7BAW-%}3
z<d=)$y%!=>by4<zp5mTU^=dsCl%x<dSb1XTtnIvIqd!Zx1tRwv?6|K@72;uB&;+Zd
zois3hesugci6c!;w@q+N{h^>4-u(Bwdt8=PaBqqTX&@nTfk%iS@1GMApGNKT!`%77
zkh^ftR6motO}3p^u_6xc;fNbF((=@7#Hos^yO_(Q>5oJ5`4p#fUE#hO6e(ZZee1Wr
znpZf!(XkHc32)a*|D+3hJwh$E69jCQ*Q546{QxV6waDHTz;|7Xn9Kkc%J$PyMNix;
zk3{a%?`+iIjr^}MK=P?t{vV_+^FOl8|7)1}|7Xy#|33^mCRVopP6jV)+SnmWApK*|
z&EvoF7MJN&?#g;)H^g6wOiIxKo&$)Cu?yXeyr>^S2Ke>LJ-gdSBOk7v3`s_{m?|!+
z=;+YPscf$C3ql&{AdT<+;=%JGYsMe<?u}PTt0M}MiIYqOBWqe2PC(ljmIlrp9pIHy
zG_m2`1|o76k0qmWYQma&Biw_ukT@^EDK8CB%ti^{UC>bPc@;gB31BEzYxWhX(GZas
z2ugo}rb>qXPH2)?3xT7DsJus=@gqt_;VDlo1peIsDMg`}t?C0=q581`!nc6MRNp~T
z9^tz4NO>3Uvwz1`g$r<z)zk;iO=$wUg;f%WcdX`dm7}U>R75gfYXGT2y6OUl8YxZg
zwGZs7m-din59zXiirlr3pC>ErOB-VhQAspvXNN3X3zye!^JW0y;yOezR+n!%1|F6Q
zfEs8|SBn?`O;M8_5XtIQ86X`sWy9h{&@e*|VVe^%0>_&ofDjg^0n}dk>kXY};0I_D
zaw-!5gpXKfic7)!hbbRtGxaYj{@dA@K}vc9O?4vCC1eoDG`Vc3FoOmcw(!`EBx@vp
zf*aT(NZmW`7jy6wIIhr?x3RFY_W_7nIHwjAl*lE-W;lVNKLn6S=yDMVl`$*Qyv0uZ
zU|B$SUN$o0Nl>tgQE+F_fFmWf|6^PYvc0*97;&~~pq>zzH_^WOTaE|=CM!ZAB}Wb=
z{6swqh~5CEe;^j6u@WJIVv`3?DYsp=kql)Pp8z){$okGE&az~7ivbU@vsilUUt9s)
zea-<l;j{3uq4%A-UNrh%>4^h<?piE`FV$sqlke1!^2Y<dVTXGRc>CDLgQhDsyXM9P
zr@1zYN7I^AFxj{eQdiF&wQQYUJNN#wpnDaEJC}D~=JJtYjr)tI^xyA?dJD2eKeXZL
z>6kA?9k$h5sr5S1VVSyfzpvLazwg&OoyV{7{R+-PKW@*auwT~Y<m##WlIkU#jT!Nh
z^ZiU+xXifK+s4Gf2CUbMsL9GML!TaQuY2JijxAZaq`>GasvHrOc=6EA#<TG#^cY@G
zZ-hKNy9$e@#793fg`_57aa*(b9rw0fY&q~@2^Xc6*~%zK_U)ZgR`J30evk@fI2O$2
zogA6hU1j<!_lMKzOEoP^%jjsySsTlgXanjbUOpc04nJ3K?}f}KYlaTIm$%02>&l?C
z_K_)X6)}1RcsH}vkLM~jzv8OZ<NHh6KhL6Abb!pAKvCjJ2dbew?my1I_xt<%>jL+d
z+~#;{Mf3TXDTk&;oj{o4$znB$#vlfTi7+-zvWgn5DGm4+4~2NzDQbWGwv%1zDA`t#
z6Y>B!IYj_jjTx{hHniDPc+OzUhE3L4g6@wyEf%Y?9tzWwpTjEgo=3hPbePj=J}!`z
z2jZdi?|WD6{7C=^n=tg?!%p3p%#Y^IF!c%367_<VxDv4_>}Mz=N0<Lr&XRTeQf@$L
zWj|r5JJM`SXyN$U`tn)ntIw|-9h76S#XEqk6=K@dCl`(Bua~|A+@l56zXZJXH^Bp-
zXl7}*;qy?*UR3(2vf4!@k&6vaJ~%n~UM|Ucrja&bv)y%54(_@}EIMlHtbRLg5E=SD
zS9f7N`Bw3s7CrMv6J|Jh>m5qjE1<1<_BV-KYj6|C%?X74yAZgDY=Q~>O^t!vT>~1y
z`Z^tvKEc&p!UJIq9}6!A<!Zz1%Xi)sRi8~1Vx#0b{IPXjLtA|;^L#2XqUk4QAB>nc
zSm00^{xGBj{c??<@k)&9ZwXh_?COIDS>T(o()O;4DlMtp8ZKpB83BKGf5UfE;=pHN
z7?3ajgbDE2ypRu8QpO);?=q;Bk`NP<UI{wvB1<URMi)L&UHx#|R(@Oz{_U`UKcp)`
zC)w!+-sWxMA`m4pzqO|uib!%RY@=0PAobJ5fuQkI{nG8OjwC-w6PK+B_6gMo+R!n=
zX-X##Ew>{D%ZvMn6>jWXE`@KR0l+YYm30Pnq!oRFzN`5QR^}2-U;!(On<UNp6;2c8
zD>vBYR@_C4q}3o5L3=KPr1kJPgw~l#RS~m*!ReS0M$IlUxXqCr=R~Z@^VSDuz#NyZ
z@r|dvCh-e30x*&e6KWD;sji)z=ZeHUF5&#Tft8{4XznN>MwCi;UJ~y>r%_f2e#f>$
zmCZ;SEX}k_IoXQQ(?OUyg0KZo-vZ6s-j1DEXty0t&I#`(luPF0GPtC&$nq$rSV{bZ
z8~Z+I<O<=EYmPw+e9r=hFw;D@4b<T0HKR!l8!vrCjKl4Ya_M2ar4B6!Txt*kV3Iw{
zF)fx`m46-~#mP^<l;@8}L7)Ww(uP}(ux}myo&goBALpJ2MsWZosfkPsgDCVwX<Y||
zg<uSk8(Hv=gvbVWmSI-kPWi10jzSikA&_^R<3PubZYPEVPu{CnjMi=&lJol+oTEGO
z!;Q90aMkpQ<;rcn&D%29R8+-aUy<)2TvOu6A7oEel`3SVc``-+2kK)$;sUk(v5amC
z;<BX30^Rd$Ufu7~+!PlazS&IpDx6L%lTHk)dmV>aOj?jkustY49>zmIj>!BP?6JU+
z#tuv^2-1%8P{;QmZc_K#R#7d@#g0WVt23wrmVtyPt^ttCeL<SrLxlGgjqC1}6F1#0
zzOP!4ao+>$1D8Ois7-68l_q<~@9*8ZICyv0$t6*4y2EL763o8tJOD>0GVUL&6>rGk
zYpw)$%OutCo&DPGq)@vY`wh5xZ1E8cwQyNz%#k-2%VD6^>Vk+Su(ZWH;1pGR&Z1&5
zULqp*InC@$=qBBuMbWr$2&=OQN4gZ5OZ&pc)AoH^r286}PZ`yfcFZgu_b6c$QzwER
z*oAspA2A#w@1*4)QoT%<-9FruNamKWxq3|94lUZrb5-cC{4~jap+>&<k}J6;(HF7T
z?)2}RTr)qTjwxF%L)V;P^bqF4D)5#n@TFzo1>E@>Ay;$?PqYw#+M%?y9v66g)VK!K
zp#s&;DL`4yiPqo+03Z(%P#d*lUK>}6oa7^FmmG0US=i(ra-J)B&c23YXM9NgNoYF-
z<t80cA8?tV5j?~$ys??boN7rA$Mb}&=mOM%Lpo!z41K%P`6B){k^1HVK<fp`dC_Ze
z*S%d#`+*~2BvQFRT+oPA`;_HH+=bO1K*C9ck4sR90W6=w?{_+WQ%~1deE1b7-(SWl
zjQTdLSF%OjU(3fT4&YmHz-fp~N!f2oAi_NkBHQe3)Z%ZWQu3gU*Fnc7dp03vWxrd5
zGuA=x9tk9MS6o?P!rz6n&us43W<JxNcKL~0JGXzR`D>^bA*&{>Tn&y!-<aI2^_oAn
z_LGNTX>AG`VT+t?f?tF|5yOw_3{N)M&HB%uu^$Cub3evr_Yeo7V(~xlPJy33o`9@$
zYxeUor>x$8OHy-|?T)H_Y0aM9+Va^JEPq=wVnRx&^`I<~A@EL!EiQaXM0*C;eFw94
zS}L<#I~G)2z?k?<RPxT(Ri42w+m)B;!H!S(pyv5sleK(D8s1;S7VhJC6c`3?s6I59
z+i+5!xmKE0p6FR#GhF|Jyk%i${Qu;ww27^mvpE41+kY8&Q}ZuUnY#V{e+Aw!;A<E=
zjAH<itYhi><1pP5_8G(Vlicc@;4<;Yd;iZ^<*U>#oQ=daW0m?Bss-=&mb>H8&c(~^
zv?%2$1GJEU<L67FGtjc$z{UKLVTk@Fl*!B2<23Rafkmf6F$X$Qp3okxkD<XLlvGoQ
z57MFid5Zm0y)%wQFT~Rj7ZY25|0jTMj!?D*o)V30JQ?QZ2zv|ZxqNnfU!U*u*fQGa
zC(9xX1(h?>v&cKDLM=p);@h)ct-vdm={}>+s;NO>_m4wt@5rLad_oe6kzU4qnE=uU
zF8F|*?8DbP$a^1urwI65XkF`_auy}>4Vi%1Me)pW`Y<ulmCn8<BpEIwian3q@w*ww
zO^zS9r$<4x(vq_NxLnp%?Tf4_z@-o^I)y+J-<^cA50v80#7eLcTH}hDH=GyhgY?rV
z6tYKBh0HzjZE76o-ed~`{=`h!ChvB0bsQ(GCQDk7VB?0OtDo1fCjXHP<^{uEer5(=
zNHcRA?FSy(46T1(TEz>8-xy_AO65e0T2HLsM!6zKX~#!0tb|mI`b0zx1d?;G>!f9*
zD48}EG#demUjOV-Wzc^X=1#-48~+Q+3P7^>vS2uPg`uu;O~R7O&<!V8W>m~=L1l(;
zEo!X^$=_^?2zlb2KTFn?`yP}<CTu3gv&zIi>q9O~UqHKYU}&5F76}hnF~G{1TaOv9
zxwxA2MLD^8!+#gjJ0*IlU#sKGthLuN&x02=hzg^X%Lp3|1vlHI8WSUA=t?u)4s`eQ
zUs=cKC0OsEn4dFx$tLyeq$n2Ev$=|WOe%2IvXPeGujQSjWo3tbqE3yUfXYPa{~!aQ
zHPD6*YXeo_sq0o>J2Mgi)BDa~62_eTw)(^cL`rn#*B@CKZgu}=s-3+<jfxX}Bdv)Q
zqAE4*+KwL^tiEIF_lZn~eeS*_U`)?B%^>;6!fRy^*C*p*E)u@)@*-IUwq_Hp+y6zv
zv=l4^ZR`qP(rOV0q}RA7%Yca1Zvm(srcwmC{z955?&F0!Yd^4<W`8o?`OR@`57UWZ
z4yC;NAf{li%wn0Dlo$u4RMnVctH2qD6z)BU9MKRo)b2fs9N}mH!Meb>i}RtGQE=t4
ze-HD?NfKntkM0s`taDaG!0E%AUDdIWyvuAOp60MwNOYa_UG^Ozq^xx+Vq>w11s%Uw
z>N~@KX%&JK&NPVF9xb$pFLjlAk@@!kQDU^ff2xl;sdKEI<l%3mXV;`{`TJ*i!)QLF
zHfcz_?@sWu#;8$k6B1e&X$sTXT6Sh8hU7B&F6(L9bxcQMgRWM$DcDX&Qg_@GK|N*l
z68*8Nm>9AhKHf*1`p$FJ@sO>Hue6MOA<f_^GaemC*HnLBD$SAcG?NP4o8VTe%89Q;
zaWByfEH$*O6hGI_3KcP@^oW3z2}N6&3eBbaj%41*WK^6ZE2tf0EPK-wM7wv^4muqy
z5(gDH!yRjCgw1sHq_VWcPZCpb&$1Fr;h^KfJI;CM;5denpWI4J6q39-G?y_ot7t7u
zCm8?hbiyj9ljCkVz`HT52xR!x(A+|q2Q{vk5nxeV3fFp8B`BdQ$u^uuiPZN$7(0jP
z%o;9AC#e_}JE_>VZQHhOn-#vXZQHhO+qS!U&};SJ8~pP-y!Wns&ffbul?f3$`kJ)B
zyetv6*QULVWQd|Fayox(I=~T4tw<2JtCt6>6BSA4>Fz9$4VQ-)+nck?*@gnweQxfQ
zk?1I3a6HCi)l><T4s*j)hFFQJM)Gcy-48xo=lN)3<Qc%>+?@9eR!u_H-eQaI8E!Gt
z^Ffu!D#mZPUm>?1=%$CMS42&Xx0EvEQP|2myltasIO_4W<NDG__gWR<k3CzKE!Qe_
z^a@DP<r8zuQl!AlR3|Se*H~zI9N?R#d03{)ZeU?ikNs5^BWuHJD`5yLmx!5Z2T`gY
zAJ)Gmt|$&w+RORn-~YK`kulxq7ks`rzFnQ$UtTm)0$9>bM|L|mpD-j8D4Ob^B#ebC
zexFuvo-ZkgmK!zAftNv6(K^#YNeBWeFsN<P%6cp2M#QRfkM}3Bb`$jXP74apnDm%?
zK!~rMKhh>_A<+)52s0DwzLYx5{MIKYm`pG+cENeRvTKwqBtD0#;LU-?qT!RvumnP+
z*<jAe-E#czSSs$}{~};1e)AnU=GD%wZ+59~UHW?yHxqhYtJS?ll@z@q<z@Qt^m_gr
z7f7qnAVoL0tt~;V@r(uLCysd37U=_s*VUGacM*(*X=S~4-`T`;iX|LYA&|J~#)pqO
zI0qh*_9)D>;!44>4H_92x|~_$+MX}2)pZe@3?xT*h`U+>o2B)`Y{s`doLuX}aA8J@
z4xydl%CbDJGI7gu1rH7FuIXx7{oN%Q3E$mp&~sJH@w@D<)?#uw+D8BG68srl%$1&F
z&g{9fI-z|4H^pmKe7K(y5V9ZQ|C{9YtGws-Kp_aythl*DEID+#OxDo;ti`IlK!xAe
ztFbPWqAlN;YMtFU=KM7N`h;M0f;5G9NPIygTPlBHbSSLqm27q`n%Vh%82=tM%Xb5_
zetZO$UTJ!@(YozlFtk=oA`fOPQB3-TS7;K6ScIWE^W)ekjXJIN!A<H3KYd{3g<i&A
zZ7e_(elFv5_`Ay%Tm7oi7<!Wm@tu$ROqEUWv_~5!$AU*3^}G+bJ`&16Y9e{dKi}G>
ztH8Z>ae~Z5VV8M{h9pE!jZ#<iOH!n7f@YVf>x!XM3KH|rkT66(s-Sm!bPq&ra<{@_
zwn7UOXHS%vvkUb~yp9_w>67lmXXB~z#=x27F>q8x@t@-h(YcwU%G(03!t;cEZMX8o
zm+&WgRy4-7k+>nFM$ep%a3?ILH(9YYqg2!M8(N<ii@4uPe3CsN<Gjjwdhjj3bg^RR
z=6$eEhuzUhK4Tb>JSxgPisqfZQPwGr9sbhX6folaVL>1AC3=DJjJG{B23Q+gy=`3?
zuMSgewZ5^Rb(073cDh>fB5c-csR=TE9Bht0J2k`}l+2FfBCk7xWm?jKv#5DY2c?e&
zT=#E^SFFn?JpB8m*e3(m$f_=PK<@DdV!e8A=d(@KIjzfioQIti6+$Sx<*(@mDw>Y?
zoJQQ2+vpyY+^w2}%R0ZuBZsC0%ga8p>**yVMY!0-n{v%s@kv*Rl4h8SyE-+u!e&v=
zsfu+w;?j&f8AB9|aQQ%Id^Ld()dZi_L&7a68^%`9E@%srV@{9o(iKp~k<f>z-|#P4
zw`waR{j#by!gbOC4x~`qD!Z}`(wW$TMtL9fRB}`l?QRQ|;Hd?^SA|Oe<&mzAK4&YK
zk(?DmF^tiLIc~)?YgtD4oGC&3J$CtyjN$iZPVY{lZAS|ezs)rKSQK2m1_DGCq&1*<
z%l2pqd*~a@xNAi1<}&FZ(70JEtY&+(cN|T=Ev+sf9rjN!j(kb6ls1O7Nf$zZ1%|~o
zr6n1S{}dNbCXz*&STEc5MWORJ#R9|b_!4P33GkL+Q*534MiEhY_Y;@;v68i!t;w3b
zg+mJA2!#fQ#_^0#<8#=`w&Ky9|MA&7VrHB)JDY#ID+n2RYCa~~6Z(*j6FCM|tz#o?
z#zy4Hw{%3P;tHJ)Rfq#-oAUD!SNWZcm*is9>T~tsb+dA-8xu6=(At%V@<kw)N8J)1
zbITL_3q2u_2mBW+BO=a%Axc^#S=@uvuyo*gBahyo+>6=LA$Kj!{sKv`3&GAdA?g*y
zZl_}Gao>lZ>u3&LjoS}I(FZQxe+uXS04e>iWvMF73CD|8#H|mi*I&axr?%YX+T{S}
z8oQ-SfIAyY1u;{`rif|6&P-nr0u(Rtn)YnHI6^e<>Bw2=<Lyaq7#>|pY>M7?1Vl|B
zhFw@JA9xT1mS6O+3|SDHC<o)f$jn{~TUFJ7OW%+t^4|R`->#BujWLKQhBwpk<M~?Q
z-A!w?yX|=E=Rbs!8G<&N-1HmNPz$Hzfg{G5cUsDYNXZVLug{!HNY%^@9^H=*W>Vcc
zd5nYoHsuuK=flZ_UC+gyp8M-XF6t2dCmsA(S|~QENazpjkNexkAHii%k#!@*G(zIp
z#5&p1A`bnpZSzf2AToKS`4YY^O(ZqDlDSV^yaSf*WHw}95YQ;~jPIhySL25(=EV{N
zlU5-63%C+YqokV$nio}*eYj^AXYq*%`e}W}091#-vC>{!c+!g?L!#H7-NwKs;iG&G
zXse3Op~4sx;aK$qD*C8WlgpVUS^r!X+B3TYxLpvqKL^=&9^sv+*fMrAe>k?iLDybL
zheNA+x}v%856pA=d}e}(ben4B<|sA)Ffes$^KURcM~>n(nuX&}cj)r!k$w{_VH_EM
zYOQh6?&wuLQJ`Q*>=EyUUO>Cmd-U9|Ai*IJutPvn{G$+%P1^$WVq5~5=MQ-p#)P$D
zNML%&OHU1dNxK3ODWF^`h)zp7pB|kcIFuO>eEob0Lj}Snj!B+T$ftm##PUo0q7*uZ
zRwok29L5%BJv`Iu?6Uo3($${|p11)#4lulk{S}D-#`y+Sh=hSmzLxF%H03m}dRdxG
zu@^u*n_9PzlaE*j9iOqgA>dqgYQt|VY|*v`*a?6<Z{QnGu2dyZq~jwA(~6`x3&}5_
zpigNlXj2&Bo{#-~?38Z{KF^V=Ott|elE9dbMQ(tN5}_c_QgAs;E|iy&Lxi~Fct<@}
z9OWpJwm6*`m5e{ejpg9P*s*U6M7h(woZ5sNO>8)~shAzXr4Lp!h>I3Y=ooNzM1}@X
zx@vJ`ATuJ*Gi}Vh5ZVpqV6Sh?m-2Ez#lTC7V58!^0ZJfGcO3!#<<RAwM&I_C)>U}g
z5<j5o@4z8o2h0@BRd2^KO(n%v(RS(oOZMoHJm1uShQ3eM*R{?H?8GaQn4v=Ig5$52
zv<<_Fqo;{|2u1pue`hub(sCbyf`4^@!xpfR{Po}fwa9=V+PxMbJdIqN*5X(q??b^L
ze^c5~WzknFctQnzyjj;D2lM0x*NM_@Sd>evS+;#2-lMM6QhGFowq*>MP8?pLF7FO2
z>?`b>Q>cKR<eMhRG9pG|)65~c{$_2~9giNKv7GM88E>lgAQKXog^}3njm#W1gm5fA
zIrvm*Qa^huk!e`JFR#~M;BzuuG)+MPY33*T+)DYsnV|ZAUD~0^OD~PMadfC6TUYEA
zQZ3&&?3)>AMrB9Zs0?8OHgbBtaoV~3ei$5JRijR4X{OK9-lQ*>OSj;T?UF*Hn-I3f
zi-gx7O%yYVr#x#W9)q7!-^u8lT+Q00m7mEIyv8gpM$$Mo?K{G39MdV9HC2~L2tqi0
z37$CNg$}+${Zm&@EuTT!$7IO8eP<@VX`8#ee_(T2%OPX7XgFmkPGtD03uMh2+%Ghb
zE{yd7-=~&t4P0?Q6JCo}vWAGMRgcK%fqYuy$teU)x$3z?uq&)W8+3`N!10+pj;Sfl
zdM6SEdC$6!Cf}q?>My<16pXgC`$`@3vf-eD`z0Cj)gG`vgM+X9w7$%Tr!O9*wmVK3
z-}i=)EQB-T+e_2x3_rwPy8cUEjP>OvfRDlSN)?>R#INEI>JK}W*CuMq&2u_P@G)^<
z8^zXW0wx%&rs(J=Z;{kc_;$=Boo5)XK5eN^Cn9A%b)XQ)z-c$w&Mx9VQAucaT{Z0t
zZ@D-V0d=KQuz(A}hX?Dtf|p=R^5UT5?nPc3{62EKjE9!jYdmN#8#J^d9U}XS#ZwDc
zAS(Ho!W^F<Tny3l!y(Q2G{N&jMQ`2iU#KhHIdiElZ{ZVV>}R5}AFAceHN^XX;F^wz
z*Cu<<3o@K=DJl+_^r-K_P<cifPXu_viht15kzZ0)x{bpwcw7m<Mncs1(dP?)RP6JJ
z<-DW#SF1eYBu;F7twqFn+{Pd%J$2106Qlc~sbnNK&up>op;1@Y10;a^MG<g<$AbLH
z&6HRPJGa;UUZ$naB^K5r!l3#pdHZ=fprjQgZ>W}jHIyU;<w2Rk;eV`OWt%I@2M96W
zv0vQNKgxdcDgZlXZ`NykTFNB0sI=G1qdygmUONacq21AN@``WxR}C0DogEMqZgB>q
zjN@inof&B)a9z?&EQWhDbh(6(kVn*eRm-!0GlV|;DRSVfYwUVG3!oGkgG*bKyCb;K
z({9X4$GYU_cGdb(fgvqAs`XiTTeYF}Ie>gcEx0O;+gqOB!FP2_7cOJ6Yz@)LcPwu)
zqigbTt`b9hFrA&(93Gyg7HaBnaPDPQvZ3ExSKg2*xUT?f&h%@tABQ*znKFTG-b1tJ
z-?^xsaog^f(ZyG#zqJdcWhKkhjh%y?*s493NrIv^Mhb%9TA9G{OP%#xIjF3D`rF)?
zyDN{Qz;|G_VFiA-BznGK6*1R~-LtO4G?$mlnF)j}`KAB9)&gl>G_iHs9qcRWOB2$L
z*#hZK(#wPPBD33;Y+&KV7dp18e<^yBz02Tb9Fauwaq3ClTl)3<M+S={7Trm;7ngtV
z-3LpDt-7YjmBq_?`QtBmo;Ya__NvegOe2qi6@(J+JwIPBIQ!ekCI1hv-3(t<yG!AH
zjqN_qbj)KEc|1!kkV|ZaCIV$}F3Y^GL2d`=3nm7nU4NQE?KG~yx{s_UG1>9h98%9s
zt{^?>%uH_m#A??3DD1H2B9R-0m&hWh%W4h2z8XQ+?BtLipbp!>FSbt2WiOuCG^bw8
zrAg?@XHgC*KCW-y^KAz3tZ$5&;fYH`IeGfb;i<r|lpE&YEPdQ{Q)s8)yqK(o(nTH*
z37g>)shh|@g&tFITDD`zbp|i*OtsD{Hm7Fz_5*w*d+7FG5{lzL`sU0`jQ>+vYUm$}
z@{kqjAB)o54qHFWAfB2RXu!Y220m_M`W@}Cv182SBu@MOrRBAlys0#q2DEl_3rVB$
zUA+>Wg7=H{?(AaV=xc9;t&DVlIuLQ;qXv2Qm+Vi#f=qM*u)b#h<G0WITf9@=S^I1e
zdkTP7XotpUe*qmTSrh0B#Qoh6<@MRlS?7#n&^1rb0-GFPCor2kjBK8#ST&PYhNdaP
zS*1JM*W>3bfDF;sTZl^|blSXaxQ8_oF(u>&eR_`c98gLnr^V4O#f^U_=$98lIE@+|
zMx57Pe1b6i$DkN7O%;rZ&g=9m^GFn+vSy<>&-kmpWGJV_i(<Mmpc{QZ;kH4*2nm98
z3vo|2Q87Nzoks49F1}5LQ%Wx7uMdb~@hB0Ysk-TMf%<GmA+L8#X$khQFBMUehkK;A
z^^C~AD&mfAd#~OeH@8CCz2P=YN_KLJZjl52^qrw>kwIujzS#8`=w~$)4DQIWor(3<
z{`Re1%BIKkFTt6&HJwji1wjv<g$O+H)pgk&#_^QF=!`9zP0+@gq>9<*>B>q+Fx_k-
zkUUg7#MQi`d3>ey=hNSf+E@dPvpzMwZgWMsvF$?}8#Ht&EzI9uwX_nhJFY?*A^@%n
zJh>b?4;EpckQDQb)P-&Cx7eDknw6KZ9rSUL<-diE3q3u2uMm<T398b8*PNN(uQ!hu
z1AJcPRZuPb^bRr=zNWDLQNZOAzZkEL8NgvkB&Dg&RlmRPHWs(uo}2DzSQsFAO!~}<
z+ptJP93P{PB55Bc&Z)PivWap`Z1@Z{*I=G;7)uTVd-gIHqAxa8jeFY+@#}!9mEh$h
z3U^itYns<-=)TsW<q9fien!#rSSQn#QRf1~Cm!Zgsi(1-2r~a*)Gi7h*{KXVj==1F
z*65$tv#$xCwGqhqXtT0H;osoY0!xxLAUygI)Z-iNo3Eb`9+)g7%gYP;Hjk7I9H!%H
zQ0XwoiU@Apj%`HHn5&#DKPx~bNp>()ZxcX<t<CXYYiwCwlSnfOm2;LZGk@Hb8}C)8
z<T<jRn88?qP8@<X122=uNnEwd4&k1ujQlOd469Gg&wxJdB59$kKQ>Lk%{DX|O-Niy
zO)#~UfDmN@YsSMqbiFkR2$=t?E0zV4=7&)42A!Lh;JHCzUktZZaEpi`H|*$ovOzl%
zp!km@+JQqei#@!9NRC^TksMR3>?A%c(arjov}y38D*$M3E{(x{02Pv6+P%tl1N_QD
zHE#c<As4j3yYv?d+Mfs7bd-If8-E$B_Ba?^rqgzbRYxlRoLTQ`%TndW7z+L~mPB^s
zMo&PbDUymf_He_akH`gwnKo!T-%o$Z&+I@>>HlpGGBJPY4(~1e6#Kog(BoT63^*vt
zGD$`3f*bJ07t0qSjj1A$;}Xeq<_yYCQ!CD!wv5dKoL3jNv<G@YB?)CO<Sz2@d4rBE
zmUOt(Se(XL*#1%}rdFGbAKxkJw2L?AvE%yXsMg`IyXyUdl$rp;28vMlGoRsUp*tyb
zlEgwchM}AgcqD-x_;xcyy5f;$;huIkpv08wNb}{np-QYDqpBe0Ix-0}5uBEo2d@i*
zHvl!WDtgG|NNBW962eH*=^;7rbrMXw>H7y|R`lRHjbb~8EGJ9Q3e`wKoIu)L*O|e2
zvCOHOxKp*y(&|n<49YM)pSt+ChU!YEpo$_v`#z2es|^saTKO-e#sFi5f++^Ah&seC
zxMmo|V2aH#0I3nGiQAwoV|ipc<~OY>Sh1o4RnsJiP&VS<1NH1-Ii|FFz=Hv18_KvO
zyOYSk6!HBXi>HX*v^}JN&T(_!f_GeP-^dJ4mv@yE0RSOpAHz^GCby;`WI=RK52TH(
zQ93@@eub+Z3$`=HsVj13xe#1PWWQk^(|vK%njCnAI}Q;fAJ69V()ETFDVO@IxC|l@
z!^0+{^giv@@0j`pOpOBLKqVI_LP-ioZNFCA*>T4{`2qwlATjOw`U=bq8Zv{w{8;#o
zcIrK0EBxDump^y}=oUwqyd@vChu?U}>w0*PzGs6SApikBE9e(Qapt)L#t?)T1IP3S
z5Hj}H4Ml2Pm;w0{NJ0|18Qbr$H{4zQd_7(Whyx6)hF$}bn2qQG178UC(_M&}00FJy
zzG~WdQu!^oJ_S_>NPTCI6Mp|t3uhRsHY%hYxJd&j(b2*335attw}e$S#(QG}uH;q~
z#&atruHVAqwXl}DG*igtFBooFsOcl5RmWrNMT;67mhTuZYf4f|L}lsxA0^jy8&9%B
zHX+SxzPD$<ZCR!lq|bvIy*Y00Qa7d#_)U5;D2^YUT<NWfA+iltS=mJiAlXjR77`px
zb-I=>#ju&D2Dq2R;uRQy7hXp)04>Ms4dS)K6!8QWz0$sD&L^7hBvfgBxk0-lt&j_?
zw~Y>lL-@irWMLH0eANp4bI!L7E)owF`*cy12Fr!5q5i8aRKGYe8c|IgHpmWce7by|
z)H%FHNVP7n6NeOakIp%*Xq9Xtr(Cqut*|f@6jkb7_VzVmSP|9tQQd_YA7TR%k-&l`
z#~ol$fib`}#m@Sk+z4v>vW3Sci!Jx~ETZP+zPT7r+l?%g&ZdvmjrYUGb4<-Y>KS{r
zZ?MZ&X&8{LG*A_$4O(E>Rm+6gB~XCeL7UcTG=elGd-E^iI3xoaUOxzNaz(QOUZlQF
z8ACEN#%qimMQ4=3j>9LE$?AK8Pm!a6194acw9(PoOym_91BtDrw5NfCPk4La`||6;
z^6(wm6ttOAygw&i{tzj!hZ3lW0^OQWKC)l8>?a%H^7hlcnabFAW;cnfIU0Q2(>Fj6
zQpb;lUl6xvuq|du@bC@1{Dbn$1v7zz@}v2iW~%V94_v*#1wlLz5Hzdx#*gv)dk(G;
z^rfd@GXcP6HF@5Rp6;0|z~GFYolf-f-zYiUE#B_hfryNe;is{nJO5su6g&6&T3ktQ
z!#4zW`D5q?rQ^SDaDv2NjIrH8OHK_hV)tHjo?l&Aj)u6)vb8Nof3-=h{!KEH%bht1
zQbg^3A0v2Aw%hP^Q*#bmf9`CovP5F4|JXN%8i0Cmyr5^Nup}>+W#gve${kzw5vl+m
zgN~G#nlK$aor<ScB_4TT<v0{CpY(z5W%g;6Sk95<>uks-U7PxoVp9~KVd6vSEvMCf
zZ!QJ?jTPF1u@5)JLz%aNezYFZl<wShDfkx!n`eCG%att3`A;?X{{<!gpOehQ#PC0<
zxu$>B_seFutr<0Pt`5qS6#PE5rK+;?>oeCRw>q<vd&;_oF`dDTPv71!a>;yM>!*A%
zED6v6p?x6v0*{xQ7&pEy9lo7zUSii!4KPvX96winHuf(tS@foJSfB-@_L-Y4yB&7)
zbY8W0{1e6bk=eGQ0?!}9zIkN+AK*?u2FUL>2p?N#ynQDFZC_S#`A0<O6QpbL7aE?S
zvw|}Tn|y6)yr=JvlNmeiH2NhqY~w<e<&`ce5it=aa=sd&A3Td_V=zrWlFhJNJPbeI
zC%H#A`I5Q9Sil)KH*TyhnqQ}%03?9Usq8F)<Q-Ac`0ey2=>^fu>BGYrt86}1i_edO
zO+BQb!ngCP3h!8M#!DiWd&j1hVzW)`F<&ILFmGaJ<1zD@=|;?&IY-Z^#Ou;F1Xmy<
z&4sn{$9k(te>2i|>n?uyLBKSX;<5;_+BC{cx1tuk{<p!Va99mIjLMEfGE{FCr-+0L
z9ZElg?V&P&{)=*i=*p3(8_tVpa`)?&=pEW=z2h(6C-d9Q_uKOhX6iU+vAo;Q=l8K)
zE-0ym>r@)S*Y|tD$HNt#QrQ?D$3kdzz@@*#_SX>}hRz$woUDJ0Rn;OS=WZ~F02%JW
zxM|Rltyw7`h>=)Ew&?!CoeHWx`*NY5@k;wgd3BhOqbzzt{I?!1R4C%j6VVZmB}QX+
zVG%wgyVAED{LCqI-^0_MJ1Vv_EJ8^`A&usGS9mq}aci0cMw>%f!-%=X!91U{>;T%_
zZhD%*!gL(Jo#PJ~K12JxRGqJQhIo0Iu>ZRy5#xpF!UKb{Sn-WPhV+dQc1;DpLiql&
zG(=WwDw)1cAh8XdpWZh71(`Vq*yQe>`jQN!eyQ2(D@JjfyKNq5WNqCu&A+h*r)~s=
zM4bHTw=cGs*5Sw#bTF&h-^uUNa?&<^z{FYjTOh8g2=rLuT1QIJOHmpYuh3r0fl>lC
z1sVLMQ5kW9WX|chq80e=bGjXS0(eDt6jT`0TOm+>rlIO+XNLuRUl$r2lS%bS-(fmR
zgp+(+ibOSPLr9`V$X0qov;sI<&7`2@wDg}L@QxI<AoN4;C{=0ZVOPA#u{a592uG2!
z6{C-s)~dqTMozj!9Bp>8iItF4Pc4l<xkM-<<^Spl<6mYLe{+eh-i<Zs48+x~Q>W)W
zz-VF;AJpmYTXsBN=-S_yc&M|kzR~}DWtZ%2`L*luxMBy=<eh;1dfQLe-iA6Rj}a*g
z0u_;fMVvE$`gjx+qCjfvZC3|Dxm3Q+)&Y^0{}U<;@b60tos}jSO^MDW<ob05-+I*R
z9L=HP%FQSHG%a$Ha_C738o?a7^U*;4q?uG6R!@|<6IP2{)TfR`YSCP*M_v2Qm4-nJ
zlfGAOYKsK<6e!x|Q%_n%Hl>e^M*LgiK3%s3-y!ozA>N3yvER!QF|}WtvFcRpGuUH=
zxw^Vx-X}maag2rL2+PxFAH8t1Ww}1muZQSfGE7q#F3l-5s6;X59&ad<q5B^e$%ojF
z2Iyn&hOZb=9gn~~_)qlfMyyFHwT{PNIFA_j-qZzk9u1T7T-A6{-0otc&vR{1o?!~S
zin3&SSIL1`zxK;E*Z@ot22V4B#lI87h~p;F3pM*PB4C6Z*r<d!VTEIbSyW6&GmM22
zyc9X6VmA`ilxOT*=Z72hX_!v{9hu5j>Z)lkqV2>$3IS|U1wo52P!|Et##;hX*~E|5
zTEndBe4c#)_2WMFY3NIm*ExX-s~QardvL2Id$R-y)HeDUgv~<5ds&QFcFCfv-`%n}
zAZa=?^m6#W89_iz_iyzxo*T=LwdIgQIU}eieF3Gd(UwHvlAH!^0HH15I$5kl=YZ@I
zt-pixa~JEvkSoW*<d*a#fm*`m_&{?#UTz;*J^Y3x4=F@3oPHe%)m(h=Wm{H0J3Y9E
zriUNb%sV4WfDq20#w3N7l25(t<Dc@ipD;TPv+o%eWb=$KJ({G-4tJoIgunr9dRb6y
zX`B~F{(xcG-{fLANB*XIIz>PhMzTMOzvE;X3?5+6TN5d^DMEyPu0F3&crl71$!-HB
zSzFK%T8au=%luo9y-SXd<_Qz(-?E@d_gq|J7ZEJ^RH$ionVbolPeB&yb@xH83Zp@`
zeJT}xvx{TNxxa8)rK*o*eu9wJg~Ur^rP}riaRjo^cV=(ASQc*;!&NF%`|1WIp}`7F
z)~*k@(_pY{)CV9AAHK3+Q2d-LEk7^x9@GvV(2%oeH3R^n${wxc1*d2CV2hd{mjY!P
z;ILQ2Q#HH5vG-B|mS{*X$Hz46EE^30$b-t4EbIX$<EcVrLNKgTtk(JiLb*P0%$xLZ
zBa$U(=9)6<Vxp}W06mas`*e$=@@*BVM=X7d6~ABT*aJ6K_H~K$)32|)Qn%oh={D=V
z<Lse=ybDhD>Vw_1Fm$SwW-L-VXbY|}wUvoO$ncy6CRIuQmKYiaVmow^i=GMV(;ym=
zo7AmKIKu5ri9%l2M67@?dz)nv$Z-$gVx(6SqeOlE;vmD3_%p5;V*_ZsSH4Hbk!nKe
z`1z>!j1r%nIh4&)`Eeh4*L6L-n##b<?K0Tqmy=PA3HU9S^D|r0zAv^pp-J_Sq!H8A
zr|aS~*j;>f)iwNf5adV)WDDEihlm*)&JQkcu7)m2CF8%cnu1r-r)`$!)gU4s@19<Q
zT5hQ1#7O;SrPsCLS?L6fw2M}pWI%8U^xh?(@jF?GALCC)p4ZLeGZD)5?|28VPGv74
zm97=6vw(Rmuk_8zn=!|R?(EbVI)sIDXc53-|2Y~TltS~V^tNW3X*63Le$n2i<19k~
z7_QImgkF#&OX4u)oX(x106I<U*@+l=e@#<pbWXS3R%pdqC0r}JnZ2({>h3;ZPaaj0
zsj~*#7ggai-(1}H-45a$yK=;KH9Y@SWtWE4^?m0)Vno(CLqT0^MbVQ&FQ^Jx8cwC?
za7gx&YGR^P-G0a_@^iaOI(0sY(g;0b9z3y%wC0zE_Y7GlxCD$eoJi?TTKDNC5mIsJ
z$1QT2q;FGiK%DClwT%=BpW4T+HLEwS9PYu>+Fq5$CACim&JTRrB|-jw2rU-+|A_1{
zv;J>FtLy(dkL<s<PN+VA?I95Nim8mN#MlAKx2Li$+dY?fTG*V>7c$<yfOGQ6#{j4m
z==IhS-~mKHpub7v{IU<v4$ht)&hT!q<W$)~Dr9r`WP90&{Q?jFhRUMxW5AWn<do>?
z6)K0us`1K8R=w4JZr!C#&tT!i0)2zo`!+)wd~)Hbb;Da|f9m04lOW(l-e~XRUg_PF
zkA>~U50kc~!aIFmJ*1GO$pG;RIr)Bgzuzd?fl`q5?w!5l$nAk!BCMAoBS4Nce?8qS
z#FqG}B8oJk(0U)t#KudL5l=7=->!kD^?Z}c`Q1+Gxz|2lKlZRcnqkm}kompOoas^P
zHJ5VfFBUNf6D?xeh@^*j<qcE?w|~0&vMQQ<#w*|tBV&4X9y_(7P18ofjv~hr52L<8
z#Z9=?oM8#Xlf7vmh5Q`(wSK~D9f~K|r>q2f{Thd~n%5P7)ax|?vm@-=!HaG|UnWPp
zRtHeHL-a(If_54UqVI@}vZjhT7nBJ4qlAVXVVAH`AfO!OLh@JOfLKBz*|7C@V#lDS
zB7)U^8tIEt$#0}fGrDi=ia;FS1i7)nQr%k8^bQEzS`y8+pAQkd2ci2(-9;CVf%qm;
zd3!*<1>E9_E=BLgjG^DBEr84O>Ls|p4ez+L_8{(W$yf4MCdzB#g4PiV?B0a2M@5#c
zBpvcV)J-E?oG}Ed+XcI`p->@&nT($Y;e(?v7w&hm|4V-T-9*`v>aO6BBky&{ZZ#eK
z5?yrFY6(m?Ch&*IBW==mQd7JjbyI=CB=|UWfHGj?ZE+w=i7XQvqj~4mv|U;$Hav_Q
z2jHwcXhX|UO8JaIQNE67!BI2GYLyL37mQkaRexM*^o<o-?#H<s3WcaNfvniD@aS^c
zD3P@*=-7mkADq9XaTPcKBmW`=jgu2Vo}mR(ekIzG1a7P|UxOcoo=rTDH1aRDXQi11
zlbz?dfe7=eqV>S-*fUp=Uz3@VZeLIRe$na(BrKla!$6sXA`2UD-xU-hO1OPzjKr3T
z{aUWhEj4nnhYeqc4iox<<v14ed15BY&VlNSRRC<HbZwzk0fPlzN?w%&yNO(8nz}`*
zFHkoy1o*bmG}ANJBX24(V?$Cu%9a*4^<+|@Z0!YD*6wBGs2)dTOA1T9p^lu=?){rv
zBs7b)+24AI&jPT7iilWTW8iF}dK33jtQ8;yDHk>f>8O78!6#%8!j(mq&~^F<*D=md
zR{(J5Dj7BHHtMZ#k4mDi>7tH@h3W<@K?k>I1s^O@mj?JKnO6}!Q)6+px1L#x=LLWt
zLUQlhmY#2J7qxD)mwQKEvr%>k&3(`E`+}u3^Fc*g(?3J5w@Md<@)l@xn8_|G=ug5Z
zI5}PEbtLP>gc=5|_Ln4y$Nj`R7|OCka$tmH@{CD>nHY9)QdR5i@>1PNfCnC*lp3a8
zef~UeVS7=vZa-KiO_X6lJ|ccGij+B7cUoYtFZ_|PNt<b}<sdCl3Vc6d9#J(co<!jQ
z<~bon8x0d0VhdHaqVS&xWT=~o#OahdB~>UzOCry4EQ6}7+!w6R+`NRI-E|UBiHG|t
zt(i%W&Lt~e80&T3*<8@qvuEJKhpY_<4krrTyUd*i$Zg#h#LCzCa=#M)XiW(slyW0a
z&BScn27sdra6+0Z70GoGU0@h}#bgE@&B|<Kz*y&S)j@sN?ndm8O;i(EUM4TNQF4Ei
zeh9nS*Q#3ZjUCCgEq_Oy)2iJtH23ciRd)pL7&Qt1WiX@Ustt1oI9RJalk|fdQS%lO
zDo72Q#=Ibe{U`iIM4P-t*WzQ~>;|nLRPCt!qFa2y7*w=jHNk4t`Qo!(C%%{!2>r&5
zWYj1~P_+XH17U3;b4#j0bzEO5i2YyoYgIqcw4)}8P7r(jxY<oIW^9Yzch-3R6SiM!
z_6inLyx{0n!N0g7%dCcvY`~Fo7}}0Qw0G)m^&Z$;5NEbrYj&X=e0afS!L-dgnmwY<
zEG8r7W>El{8i&wpSB&Oh`jb6Zp$GMdhh6%W2GW6>DUlcdsEQY)f$CS^w30F$B#k)a
zEL?=qwhD>}%`-_CB$Wsn0p#@4k3U%9m9;>|f@ePDo#5rrK3_{kfyt-X3VzYCCscDI
zGtj!4d~!5ybNMNX$c}mR)mN|L&CwJr|32Kxlo<S|cuoOvtLiir_}SUT&0scapXm*#
zPSn=l`)kjXPH94akCuAjq;W>=fvxVq$*eq|n;bDo@5(Tl)zl|Fs_XANLXhc^Nf}8{
zL}&*?dl3xN8U2{r-oK3Q5Xp-|4)Qx2WYSFeNP3o(Q>zS!Sa?d3g({57E(2VwxB44E
ziTLFZVr~@5BEzM;orPCnf#Wx)n=^3ao4FcXTiEwV+27Tm8Wuz6u(S7eHOhVB5eHy&
z%_?-$zse?CuT1_X2~P92pE!WPot?32c{?RZGb`htUDZm`^7MM0O4_IkoRU1Gb4sa^
z7PsJaj{RJ#O@&k)N6bk{m&4QHwn7QBeepabIdLENI)*}DRwZJL;bkOVfD?ZI-e`pN
zkhcrg2^)bg$E(`8&PhRcc&>&j-a$Kl3j9~^cb@4`bAA8K6G-Gh)u;gf2y!^++rI}^
zc<$vf!$BRi7rpPxmkSE`Z6dp{u(>MMzF`;CygF5fBJXix#O(?wD1NgL<E~=?>Brif
zZW^pS*$V1-io9b+H(P1*)`Xr3c_1I4oY09rsNBV<YcfS_y)D)b*)C;Ga=Q#29TS2I
zVZYh!biNHy@%#H6&GMT&JO=B)CYqpAc+3``OvI;d-7!7+mApcT(;$tP_~H4Ah-({d
zK<YZHzoUJ4>{HO0!v8Pb;Px`@#!P!qq1R<kQKtbj36WuC?jFyLbkvMz)Lhg=P_EM;
z=IC<+-BN|Cb`9^ZvZX8_E=lzzZAlp&hLq%~BgGBAN!{PDhGb{OIgSrzGIYQAW)h|l
z1duc2VG2T3`>8jA6}hKBRA5JJpOa;0ij0Lytn4nqgId>Nyn2T4JFMgOU3caV>Qhey
zJ2QM^zC|ylz`6P-B<Z;^TCBw`!Tz9X>5|`mT`D!d;{=tORI(u2(*Lz`whA3NNSWF{
zg1;QViMdd(-fllT!7wH6>6_w)T5;Y@5n<{tZ+`dP#8nla209o!In^i5cOlCB?W9bk
ziym#iNt6-B1C7@rGLj$%i5;ojM`*&8lSea%R)gWJ!yG<h^;TNa-xN~coZ(H?Fj&Ef
z!9gIabfP=9aJNB#ZQ1<uYsR>0^gpHEf1D#`W#jlCcWPq*BDUXMEB|nkk)*vm`T2yz
zE@Oz@GICouylb|4#qbAz{dRV{56i93asiYHP$2O5kpBOa@5P`TUynAgo^Lyp2je6q
zz5Dy8D@c3a**}<M>CzYwJ<aap+t=d-_DSiPQ(o+$Hsm|TyUm$aR}XY8Gsr2Q|BD(?
zUu@6ZiR24G49rYYT|i(y{Ct*EE77JF*1vU{MImc^wKU$t_r?j`Kb9o@eMPkKpvv*e
z16hQg1UGrsNfAXGvF81O4*9#@y7x+FQ5|XQ)61v22%y3$kKY;fm@HA8S)7rs$@$y%
zm2dkbq`3st;yI%PgF@{$h|7<4eoFPJI}5kb!DoP6)Z?y{a0tg$z?~i+{Ev@!_op9p
ziZW~=@7MR&&`umkYOOoSROW0+&kwhc*V~ME6U!eMV{UD$BF+O~Su_$kP2hxZ4a`4i
zGtLP8#-p9Ir%n+)LHsAgQN)f%Gdq4+p;UU<PgeRd84jX?p+u9|u$a(k%ddm6xB74i
zhJ-BClQ-rD8eD05qF`o16p^fbpVvoFa|~vr4Y2=ieJH4tUZGg*kv#1*%o9e7vhWVt
z-PEKX{Qfvz;rN)or1N$|iydYV;kO@|_4N9AV)K0uvi%Y~3~ng#>+iksrC&qBeEgC9
zAvw^yi$A4q0HFL9|2x)b`sele-knbn{n?_~Or0oANk4UgQELwa0ZLn-hiC#hFR9=u
z>R3T>VZYdAfI|sCvS3L9DXH*Y_BvC$*~d@Bh6nN#5W@g05bIB+>+>6Erc=Yr7gRsm
zOXhe23dstSTy0WqdNQZC0YE|Y2Q*t{lAK8vubDiJl~g6uZ<BF)>Np=^W9WF^F5!8r
zolo|VC7mEeM7N2$Mxc1lKq@#nW6<Fp-)tSq9N8CDGK6QFr0o<fGr^jQ(L!Rd;`dX%
zbUITNUa`=`Z0f(~<W|R}PIH5n5!UGg1%W}Qf_?)Ub)8l)I6y62=4Jq9A7`S18pW9e
zd8YZSA)x^@3K_8t_ru3iRK2!J+JsKRC|^nFpo4EjP()&SIIruGSg0+N>MG3{baRD5
z(qxc8ZD=8R)&;ffu2KC+At)hBKa4zYXsGe#U@2M_MWvO4A<<ENHb*Diytm8v2W_pW
zoE*taRl=wuJ;;S=)i8=>s_DHQWpZk|j4PZDIyhC-&MsS2Di@I^MX1$h7+DKP2w7t3
zz#a|d8tMKLa0RAs9tRZnB<#E>NtpxdK;<-Q)6r~|{6|#zP|ZQ5Sw^GBWEK;(t2NSU
zflp2?qA;gVk}bY$0yav4$q&v?+@8VKQ4|ED+u!Ssi`+TVl+$_uYPD-JVoBmT62y5I
z3b_jht9%7O)OJuBD*W>1P`edsK@;Kv&f9iIqIe=KdyE8>MlQj_`&^;0;%=;bGob}m
zNO@|!1WrM#p`{oAU?6%30|mtc?}<ahJKu>CXb*lSm)g>C?~=X_o7kD{$#lDNJm3O-
z-7CXh;{K(pJ4@$Ned2C0CF4YXyqvzy7$(UuRI+~9B{a&p5(px^#2LKnX)pWpa-qQ~
znNQY0eWrm~Oree|da^a!(GmwMgp9QQ;o%s`!I5S{LO`32^_qSV^Z}LhGk}b@l~ltk
z=-_ROdK~O5OT?omb&=XHVM;7beyKacaTDAUR_O>F1C@E(5cdp&k|<N^k|@>9Rc9m;
z@tdQQChf&W`vsjH^w)tXEjU++_=YeW&=6-km0pQRTzr13T<9D?DZ(%pOBEHbplune
zpPTh9YohZiJeKBX?sbf-a?}e9Lu1;cYcAZ}2nySTj}x@9lOlPOp{s6ElOgSgwP2g8
z)bmM_C-3~Y=$_UpzcC~PpMn&jSv-b<bu!Os{dh5@G4n9;9o-Zm5f4kvX!!{wX>i*+
zl52)^Zt7bXq<llv2Idqy{btR7%#I@PFN6Ml_Pk~aQL%Ks<$UNH^;h;hN{lz1O)V9L
zQ#M(OPBWyI<8gMaKH8%vZOkO%ff0uk)QGvlw%k8cEcsSZ%Y;!e4(qHsGz-B}soxSB
z$#VaZB!ySRN~3z&GhZr@@wZa@IlSbjU3^P9rJ@-QSml7CO|Ho(6vqdBd&myN0?nu5
z_inHkz}1Haaw^bOhNh5w0~Av$oon(A90{f74~RP|<OpohyP$%0!CxaZ$`u#IyAQO5
zRnqN|^uc#7K>pkGwUUJ^c2$9*jsMXY5jPag{cK>zbt5BgWD*UmX@$MbAR|i+td7f{
zH`ryClUhz)v7~8>E7@FlqWS>lcRO#SdEXk!ZMEBBY9$%b><;Bc7PEj(9Hk?y*esq_
zM-`tVepi^u0YN)|s8C>ZIS^zXl2gE2Y`f5y@&L@<z4TzyE;7NIMXmeFiubRzV#lls
zkc`_up$5NntA)mYZl`YgoSAyW>sZvkGN)#8f+UHGGw{xvC9@k>$-8M=o^Cl8CS*V+
zxS&_=V2cJ9v7a4qQs6pmq9C&CKWDLu)1mXsV{{(U#%-*^a&hW+2tS{9u^qfg57f-f
z$L}hZidl24I!X1(8p8-y{g($J4>)GArj(WXnVu$nu@x-_EXHv=uA9!{2n7k3>;zMi
zWV#ff-uVyGug5k$nv_t(2sY$HV)iDSl=G7KcBxgfXK0lq={A(OR^E_S9@UZtNNmGO
z8F9PcZ=^V*x&{kQYWuto4^FZW@InoL8KUkdpWs&#FEoYM;~A8_xAJ8B$3UjU<errX
zsb31ek|SXQO<+uLLl{t`R*Mv7!iA8F(&l?jlL0fW$-~9D=DVtQ!mZ9SP-LT27Vw+e
z4?x}2t7_ju82O@j(mK+GPUw)`y$~4jz|MeQA_!H9jbz0ZYrP&ZD0|3mo=G*~m_^c_
zcsJt0cchfV$uxI05IbpSyJYqbwH{DRKxG=KK393_jB73{Yw@<|ELL4m9(1dX-SaL&
z{RK;BWzKqDZ1n%^GqP6MfOvG^XZ3GpD0lxoyPT6higpTXcB(mMqn2U_SlNax3CVtQ
zoYrM?)?SB0k%_{|4$5bkG>Gm~aRLA~z}vCj78JOtufQWYMImMc6s(S&Im8|;#H$7>
zQT3BRR|W=O!um!;)&nD%u6z$CPj#U`WXHHw+WzR$WC5Z_9Jq9cftjTP(>pf}H)yNo
zvzlg))%BdAutw&;nlv3ZUI@n@pHj~*sNX`;Chr<4SenT5%{5Ub7(T^b%BN}WG&IG-
zXnUHDh5eNCh+|4^8=0YtQupY2kBa)Tf9X{G%9}bfx|f`rJ>^&lF>C3#-7Ji)$;<e-
zJFN~{Pp)6$mqrw-x{3UUoc2Wc7X`R$&kvO^@-gd6RtTPUgOP!|sGmIo<3-Fp>FCqe
z-Z>LxsN9@dXH^?N-A>PMXly!8y&k`p5_1Y^xiX!34X$^ICw>x94^qnZxaeh=n!+F9
z_VU6~@vXcoNfblhaP3pegTM_DDT_WmZwHIDs({15vNBl}!|GH@c>@zYS=yfrL(|6a
z`~cV>G)g}CVEy%JH$Be)mn+m(Dz5HUj<KR@l5WxI=74TXtZR8rEGwWt{q4}amD9ZH
z2MpnoCGuY)mf=5|5*)1m8xwsBdm`qb^=?~l?_kn^nY#Y99}9f*`WXZmJQemG0e@}5
znzaO+)oc8u<Q3&FfZU@#4CfP|KIB=aOZj)9M|S`5Vc}t8;iRKKAuAJ{O3tn)%ik8#
z6GGr0Y(&F<jGVozE9cu6O%ThIX8|TeK2kTio9lfsMGhLL3g{io{>>P%|DCJrpD{e&
z4ym>x3Y)GAxYj<vJ(}@I%1y~j8Yb;b)w{l**wD)sHqG-^Z7=#4^+>2=M|4TMHp|0Q
z)~Px)j&SHse?}EVmJv?p7h_g0tT=s8@yx4Ii{P>uXm8Zba13)rb(f>Meae4q(^uNl
zYW?d0@3{c1tJfRswGu_AO{O_mC1{{c5^zwT_FMCBu@1umnsa^^?B*8j;0p&66A=R@
z4+=P;1gR}b@XOn^{*-beasu!I41<6vh%qUGX`m@7)3b|^a1mJ*IwiDvqjqUAy=|&{
zt>fWe$OcV%D8BMC%JhSnM?xioTsH$jQQ7`6!5#k%>(7wnTDLf6Qws=@JXq%nD<jh8
z-X$u^E<zmR0jX4&uP3!&hbKwaoB9UYW1@!V1zrK(?*2_IFW`)5TJ3no3%QPj9C=BD
zWt<2C4$QqsZ$L&a`ke^gex|ns2?{RL*kifbzN)G)!Re%|iZQ#2P(itH0RGX`Nw>Vc
zR`>#Ho$Ig@-Tv%~v{gt2r5(9k`6Q`=iYB}!D^@3|9m^ctk*uNOI=fKLFpS50koW3)
zmR*Bc0%QVA7{H1ACTrx}9GbE(WppLICviC;{^Ji-+U1XQ&A?cY_N<n%?-ZcTOvflb
z2RGgQxqOp?_~EF%9~WKB+pFm?L&9ny1WHD}H&1|P;@o}NI9i{KYAdwzzyVvGwe5Y3
zz*dv>PyJ<MdDEqHk&c(mC2L1RX_1y!WP;Tg7pnMHYmVa|Bv7t{d)^Dj;++&O+h@f1
z<dwh7Cu><~ngd^y9d&B6tgD39USQmZsI_iQ-SSKOsaI$(+`54Z4I(w}UH>4qDlC&L
zT=T44pzC_3Jdu)a_sECoGfcP~FQZh6?S&01mq_-?v4oDcOdrgFl^n2%57=%njeKmn
zP*{HCCp!_mS#E468s~_53HFnD)Zp{`A$HMdk3+#;`0ng<to(bMRs(>vOSS}B>H@;C
z+;#1uB}R#8a}xvtEnf;X%V()t{zV}fBzq=Pk<S-LTMgy}>dG6<1Rgs39@A*$p3XC+
z4ajorjtRzc=L&!odIz_t9?u-iRbF<m@?YdjjBUqe<>Qc{Fj-l?RLs@~N@<zt+A16@
zDKn%7nxlY7;x`h}iG*iwK6S-XbHSt{|10fUq5{-rr4iCO*^Y?IfEzg0WJy{RjvF*`
znAy5zB^NaX%gRf)Ywq(l2`{U~@{{7!vOcx03u+zf{IKf+8R>k6oV1nCFHN7P`0#*j
zxmrUzK7d$r6LvvIo9!|a-iCSU@SeA18=mfc7M0xazw>S+X0F@8kHp@en_r$jUhgma
zK!oiP-Zy6-A4;tO(4-<m@s7ko++TMy3wCmTS9QsezDR_`QG3zdLznS7Z^_u`x^a@h
z44O<y^6e<k{8o_p%h*~JuG`FHzEalX2z#?e7qDy*<~f%->2OQ!ry6h+SP(@8R@}mD
zijh^$u?nUIuv5vmI%#~duF=l4KhpD<{nHUu7$zcIu85HAZN+&<bmZxYt|NK44&dV%
zP!R!5yF{)D+z#YpQttM@i`H2P7L~cxn(cPDC9RCNXQnOx=Fnv=ApA{KNknnO7!9!%
z0!s1MknilSeq<x825bsuY(h7Ywp=>SRR$MBE4b(=?(aWFb9>a&TGTEzkiz5(rk=eG
z+~6^V>w?8Q=1(RBjm8e>jGE*-;2d6z2!V??F}I79P(xfMF0Rn3Us>Yr<2Dw~XW88C
z`HXU60FrNE4JSF&SupKk_a{;vG|Q|<UPLWRVLco(v0<yB0%wZ!-A7~itJmk$V+1O;
z87#};s+cH32)V1xa+`lmY3nW7ncAYNlAchS3g<Fg2F7&Xh}fJPliDRM;PfL&@#80Y
zS-FIhNR0$`x{b_hhu$kHWSs~$E~Kp1pAsZM!-=pBwRz(KDjQ{vKmF0tae55Qg)stS
z&cKN-SMpvJLW|S)_wgPA=Vj*kXQaw$j`D34t)_LWR9{W2OB-$Amo`Dv{?ga_Y8#=+
zZ=5@k7_9UUXcs*05peWCz?TJ$1sO13sx!CFg6ILZ-t=v&+$S}QkdWi<F*%Ij!f5HC
z7ZMUK0DGqCtPP`V*jl-Zo{9!^B4L~#S37>|1ha9Sp7+h%Y)AXWuEvW``8pEDiJWXI
zvJP;OJd=5PWha`rk($KJ5~x!PjZ?lR*gz<>)>dN8mWtagus*LdG3H7?8Wb|5x$xu3
z`{d+HV>Fy<N2(mLX8lt$#1|GS)GSIF6r1L_wJg!9o;8XDE1s!T_1iq`d64(>3?8wu
z41m+dGm#a<FCzvr<Jk1`qU6-NKTX6=h>l9JGIFM;Q{xN(c8vhBc~zjuSOHj?QQ3Ra
zr$g?r)iF(>2wc#4|4eu>$=n8E7vc>&I)bGHkv9aT<bDV7F3mW)jsXAgo3?VZYVP@H
zC8SEZ?G$}lw>=r1sqBPtA_b1>eT6oU)k~QhImGt>|KL6GmzsXInH}gXFB2NU4G~#-
z7+ZLk*7<QSCl}wqIxKB|fXbrlK;ZoKnCSIipY9-C_$7c~UNo(`M>Yw#HQ0LV(<>12
z5yHF#A&m<q@23qJZ(%l+i4z0FiP=VFkUlp!tnjt*Ya%j`eVk(}n|d8T{yBkeX$+(W
z`o@vZq?Eh<$)r$c>n(NrzYJa;dYo?XXYtaRuIQoA%akxh@X40W>NdG#a8BdJGiC`y
z<TzNQ9rlK`F~p%OE`Cus#zp9{0Qw=XxJ|oV8bzBfNdZZKEBu-TM!LgwoNd`|eg3bt
z2BmR+3O5scuJ$J2K_-jKBjaV6Wq<`yk;XP%V#hT`pa}c!$ET3^CIb|*k|p?)&j>N7
z$#-{wZ6}PzzwOQbfSizP;=QpuZaDUhpcJk%;7<vGq!cAQ=ko}nR)__)0mo=e#kj5-
zmZ1c%@f6`yHTS-YR$P#73)mZH^YOkQL8gP-WV+BoG=6Dl!g(n}LNuahGaP2aHiIdH
z$Grrs<)v8A7xGC|D=r!IV9bR$m-z(EsDPHOz_y!pTD=bCIL2r;O-XQ40I?hE%3Dlf
zu^c~E#Dkldq$DTiuu{IHE$pE{v{JBu)f;d0@Yyn!@@c^=fw8=yb>B2KCTB4pxTCZm
z>|rXMA^wOr;;*;&yM)HUIWl_J8Dks^$?~w~M+0?L$2cV5k(m)ek6@5>;7(KN-O&RG
zeyPvp6FnrWj_NCyt#ycH7LJy!(d}{e;o;t$pEdGOt{OP1YIxmMg9QNJ+va6VKw;p*
z_;FkD{8Jn0-h4Bz5t0n@(J~*#=oHw&30b&Mm)3lX3lG?S`!A7^r1`Vol+k8jo@<p(
zwCq8b{e{XDCUv;Diz)*TbUd+f#xfqGG`7(WkCm)Xny_>8kNN+Hv3HCSCD_`w_t>^=
z+qP}nwv9cuZQHgzv&Xh=esl8X`}Le3C+T!LsdQI&bvm_b^}4THqU0;%`_WLGaj)NW
z<bxG$+<Mzv*Mjo<j+|04pRyiJD_yu6DYlMD`Nq)QSbhEPk|AJ?Pk7ZDvsw8^<u2^E
z4u4nxKr8*KZk6cJPi&Rk34r><-^+J<ep>yd%^WJ8iBn5knAq7#D~p&X#SKQ<?=M4j
ziSfC>1Re#Q*<OsDy}`*&A~k9H>AI&NB92f*xClW$`go*L8`e;k`$w>0ESeVQ^s!-j
zDaGymaEQ+P=pQPuqO&+tCH2n`ve-!H$dwTh+&upX*+P}DCh?xqzTk!p*T|JiIc_t_
zGV7uu-d+*O)#MaJGcx6F{|jfDAFV~9@W$h;4Ca}mpNm+cBx5sYMrsmqSS`sFGwKB%
zcM36|`loYy^%sehZ$0)BPV)a2;4}VT(<25J2KN6cOZ`uCJm!!U`Zr7c3Z{vGX2YNF
zUn8sL-%fJ2p)Z>#O?xJEoUj9DGHyka(ki=Po~+X%tq_a+MxI*pUFQ6>xcBM$x+kd3
zaRkF&wcVqF&(`D}m=d|H49l+(cz)IW?!AOr-PFjuoFSf(=tL*x<VkqD0Cb4ecQ(#2
z^o?dg=ZL4+``+GSN7FAZ)c&BolW#xO-Na1AQ!VAl(Q5yG|9qDBfZPPhnlm!^x^c0w
zS3Bq~cv<B*qoTJ1cL$0@ey^TS8gf{9)2p@v%NIM`62y^U701^z3L8Y#Yb?zJUb3!f
z1e(T3g-2->RNzh~P(kzukP3>M8i|>frKia1j`!2uvw;)PQg;!hrJb$V=<ADw?~BnF
z(wa5uF-57GjtIW=ge52VE&@2~j<`V5(CTgFG(qxdXs82}EkjxyNg`2$7Xdj8zO?lc
zInm0r0X0XIQe$}L<u3k*d6NBLC0L3_$SZo%@*z(FbfAwrF3el@I6u`mK{T7I#Nmx%
z3xiGf!qezLuc_P*fQre!7TxnTONWwcDp>1Ft3!$!VGZb32Q>5E(_Mzrru5Y^mIWZ$
zWZfMHo$hVW2@h(<f#BT-=QIQ37{6`QGtMDta3F018Vd(1KcLh5YiZ=~4!+3N6RwH0
z{LDJbc`6N_HKUF4mj}|nHLBVz%0n`=S^ldBQp?0wx(|=HE51{9^W+zUNDFHFJR52N
zQIPc^_=nH2YB=8v75ll6r6WWED6x|WAS#%)VqrPSOaXsc8~pan^di<<HM#DhZ8i<2
zF$B-i%qH!&B0;o~rcHGd@5u=)h%Ij)o~+adEa}{!28qKHx{h}lTi=8kWbY@pt*34r
zQ8ho+UeR?vf>o&U*uant5O-pAwX#5VJH(B@^vFra^m@Yiro>ILxAp1_-t<Q8u@k5i
z0P}I+sp8uX6R!V`m~qFUSWZYw-NmUC3-}ueA3CC&bi3GU+tD$x24+uUcA3pwUW4;5
zLn+55WO`Nrm~QJ(?}t~V^<xH<O_FV*JfvUA2MXy1^P=*yXxhUhWYT5j0&6;9HydF>
z)It+_oixfbMta!J%S)V?xQnpbio~r!Z~k?{(uI82uf`yK127oh%YFl>h8UD9h}Pa5
zzQQa&j;WA7Fw+uBR~LslQ&=%iGv48Hk-wx4yJp{_(YU0kd*S)b3zno!w37VxRkg-Z
zBCRmF!Em&cG&RZdsF2SSPSQTL%rK*E`Z{K)#ruo^DQHuKtKH9U8c9y1qd=|I8tfds
zrqh4TtU*-(*10&61s^k5jF-wCz{Lt&*pB#;EQ5rqvmr)3o8Rq6w$;*Pbo|XEVKBZI
zaXP{-`+WEOF!+?c%FRG=vK|dsP(>`&`dnB81*!(>A|P*UK@IVg?TN3#S-5S3J?=qX
z!{*8pGz9bARnAmu>QF$ldS{MQb(*qGo{WLjwxDIFFq!9aR~{d-MjmOPK|zF}Fa^Ak
z$~X{<3|I|G>?D$85n7MW+$Bd5gjfvS+63WsE!piJmjD(P>f3Y#ZkG<Wp)<1~5=TK9
z*fGVe=~;X<@7P6lc^W-Xhm<XFlYFQAO<4^eIm{`izo(9{mR)Br3kW^yN2C@$Zn#xS
zAKx}?D+8QCQ$T$xTHp5_R0f@}yJxRP$)x{Iwto<JMOVyR76b^#rGN=cRNmne!0*XW
zj3JSc$NZZ_R?3(+2MPHV_Z=iJsx7qSv-s<3#A?~j2Z|N**v?w@c^m=N{-tabLAFF4
zCrI~khY=qWEcyYdt;ZX*N;%*$5WD->b)7iUZx)<9;`1XPGOV9LQm<3N+>}+X6SlLN
z@Rd;rA2jqKv&S!Ng5cW22M$=B`H+w!z>R#$84w!yptNLD6Es+l&)UcFhHr#QNEEv0
zee`DhUL|vUlI%Hz@v<J*Y_ZwWc484d^+qjhy=i%tX}rR_g7ZmO9k2Q@@a>g)^I*%1
zQ!Bc^y9f?oO|L0I(1fXBHGMaFI_4rQEfJH=RNp~o!y8Wh0RVW3Mf|UG2mAl!&#^N8
zZ^@x6tnnDC(JNlP{7pmrTHShes#|=0(;VZz)>S7aZ%ISFd)PaYUc829-KESTU<gO^
z<|UnBLkF_Rgwh?^`^T@zjnT;ys`EsoZGWZGSUov@I%1!|qOj#GZ2vm&(pbCF++Oi=
zY8^Xg3krlY$H%S3tt>tQcT9jssNK(Ti`_f5j_os^A_oUu94rF#4&-i+XghlE@j_yq
z6j9C8-E;5m9KW~MJ&|J?aRh3=qm9><qk|(nx&s)ENI_1&_sLc?-+TmMPZvNc!1cu&
zKd!I$`xAzx9Btfi?S<c2;}B4Sc?p;Y{7~($cE(tt#YgCv=_{8_A3O!~*Rn9IL=kfu
zD=~9G1zzuuD<;p|-A4;r5MU5qEvVibW?bn^9e$wm0Y=&#Bk93-8v@bc)91LPX<mG1
zJ&`FSxGk8gq~~7IR@ByqZ&ni`ar^;FQr76Gzr1NmdItj{9R^HT6YgPb_{gqAY^c%z
zIZG(@bVu2(lZENL2)?;x0|yP<UU&pDNGW;D60K}(5-U&NeefUO9#D=_s&GkhzIl{0
zP^!%ZW??FLvW8sd^2BHyEK?gsS4u4B{sUK=>ObIOZdXP8rDkq4u~+LvJPuUewuLwv
zSr`bD7c=Qnd0fsf9!}DsXiYeoMj?URatxy&f(JpdlOFH9KncXbx5jaNKZJ6A#i@LA
z`mAX`Q9`}{Ob*X(N^@7YYyUhCA%I2+BKe{InM_3N4})$Aphpj0GT;l3PAfN7L+ww9
zgfxYAn=h?LA(qOEx@EL}uo$+hGj6DkDS^7@*>JG&NHSCp$a!boL0*l>Rl$4A$i3fa
zWkhp+8^MI%8-1~0HPUD<h_}TLN~Q=aZyi8sm;}9}^;m;4ER3id&Fj1IkBD+%fQ-8%
zqm#BrJuws)DiO7;3VpIN^o+WVBjmn{BrhW-o|4DSugxB{G3=MtiPVbzH(;qI1TogA
zCv}MtGC?;e2q`sY8KXbGa-@+_cCF7nnbv4%V2UDO|0bafnCm)DXtjNz^<50vqQIf7
zh>VoMiJ|sV*g*O3l(Qnl))L`lj6kGt;eLIkL-A;kW3PTHGOC)itFS^a*2B9H>JVaw
zDn-P=MHdV)mW=yf3i-w%d6rJmj#r5YEB?LaN(;{x*XI{<r)81uWT(!n?}Fq>Tytne
zOV!f2R{p)#2q|=qsx@Kr4wkBe^#OH@kBGKzqr*QA>@wt4{G-#Tx#igMPQs(pth$L9
zGm)-J>U8W@R!^tvv5$T0*xpfSy59sHt$5?a2kkxC!<W^3C~+;x@M<O+=B=O%P(0XC
zh^U1lL$b6I!Fn!`Bery84F1lTIlq+vuOTzKnQ=PK3v{Y$AsWB>%_tS%R3Tav_#wf_
zYxRAX(|wqbybL!h{79DD;d~JfYM|wMx7f!kyt0>^?C=hf;w5N30uw9v9tQ&XohGoF
zPI*QA&Z=;!JA$3XiX^tXI}}w?{?sf1_h05DbjjOkGS_dy4?>-RYb&Tij+wE37So&Z
z`!TD($d>Lr;JsW>ouV^?AG-Oa3<K?CBwXw8_a&rU=n*&5%u7U+YvQreQhm`$99=;u
z6Lra8-iRhmX!v_AH|8XRdzr#`D(4-%D$#2#Oi`~>#gz38SZ8kTxm1sZms#CUPN!kA
z{;EDCJd%mi^ZIJL+Am(o%RY#_#cnLo>BrRll#94nTSR!T-_5EIt#^C$-Kwwr5bbcs
z65jHv{-Xl%8YI!S=_<u^O=X_K8(IttnA5N~HX)`-3t$#E5zx<ptLV@O>SSAk5ipZ~
z`P3*;P6fa@#EC2Vf4=}08T0AB+tNc<%TFy#j?`VXWLXL__0SL3k&Kv*jtoplJy{RS
zecjxoL4SeuTA`32odR9OJA()ilP!V2{%$TDg>MU#7HiEok|lHUx0hS-#*&a1iuIVm
zQxYS)<FT2Dv`vSE)>4u7Qwt1Sc11D)*3ZD26rFO8rN=;PgxU$oGIA=<>F<33n04Zr
z<2m{X*H?iR;v3m`LNp9huj0_n@@1q+ZCoFUBMV>g<7l~b0Zy@UJ@RknpV##DodVWF
z(5K%;ZEz^_6(;EXfp866pmEh>e9Xnw4QbAT?ek-saQfRB1gaW*RxFXZddmm)*M#Ak
z=mk9jqRJ!t+MQL}yh`Yrm*&+{h{ks6BPecuX-rhmWE!d0^Bg1)rH;Xj0G1SSnkLRN
zevf^tIt%@*>jrl$irkGffYnUrr<bMbQxoS116G<0xT%xTk9^W)CZJ_~qwTNApk?sE
zy!^gQV!r>5(3;nG5FCYlS17CF4^x>7qk|b6mM4)4Z@aU1PA7RuambD#6T1*@tGs1B
z=$dS^6?E*A%b2gxERvR+{fD`sj2ru44S{(qy*)PUK-zVTE8{t3OHtmrXm;;Bvhr8`
zn}#^@eWpvThMJ9SYbM|M@8AEvp1S&X7SoIISDP=H+Hgugq1b8`!O!vj&0K~oe57Vu
z+3W|c2~2SRbBX8{8|aOk3g=`LJ*h~3bq(R@7PN}G!ld5rZGYFt8K@|mR)co;Wc_*j
z+Jim)&RcjLWB)<)(UYjq%HBA6pyNiaeF50a()Wv9!1;k~{PdlEbe3lz2Vi{;+p2EQ
zD{5e(f#gv|G361i`x5ZN4#MfVD(mmZ0c(UOH&OiNQ<suo04H(LlUg0yT+0Y_7;yZg
z5AM)a(x+0&3^BQ6Ka6dt?@tU771mTT2Re8{TF%Tskm4}}4zR+}gss45McXmn-qBhU
z!EWw^P3SZeyV$m*bX8?PA^k`i#SHg>S5ol5m_~9-lFZAhdjpUeDw9WNBj8i===pZ*
ztV|R^I3`zab8ZS-_1NQ?$#1i!v@NJW0k2MJ_vnI9fl~G~G{E12L&^6yLHRt0l<&i7
zb9+L*4yi++Ycf_<?JGdcQFa7&^O_m<Gpeg*M)-ldx2qK?mucxWHmXNO$vmfJ!?dj#
zG;LJ|${Uhz`>txfIlPz!Q)xA{X@JSLeN%qNG{mXz2x-eu?#Fy?oo9df@{>v%5^sOT
zX^BcX!x;@bkDieDt9gkuV-4{JFULY1pk$Sru!rSLT5OMrD|uGKhOdMP=CH)-ebtw|
zO^CS;>o}vhFr^sl<p}clkX&Zg#xwd_KVhl(k95rMbJ5KsFef+9Jn)L_bFWi|%9(nz
zLzIuCn$hgY6seXAdaD6IMf5Z;S5e_G)2cb)=Q`oqL6C+CpBysuGZ6<pWTxk0nOmC~
zc#T`!=JJ^aZweafs%oRHrr);}tMh0u;cw5?2yyQ9*$-DkDHr2)Uh=WuOstE^r3&BN
z__`|JoZQ;64JW6ac2REq;tDpKL}bcLDfv4^h<2^W-1EH3oA<SIdH)#D#6y@tjha1@
z^^GMQ&FbguYcx(l7)R5T`)K|@rPb>=M$w<D`!B-E>#Bntn2&JwI-jAdefu~&DnB{r
zJHD2ePc8?t;AlTt=K+6D;dDRmAGA0+htmHRj50C(-(ZxP<$qIFTxm_lP_-ibN?4s;
z8RGU#?qu}w`4^I*-O;>UL)GfDWv&Ug4_}nL6TgZy&w;@kPA$qsP$Nq#Dq%ISIX`{!
z24Cz_XItsqXhCJ7fTu%ecfk7a$iG1%;F7VK|0rOG&e{#m;w3SLtbou7Q2dtNd|Kbq
z%H3c?xBU48wf9^B)c)YoRr|yd<l3QYeL^JGh4$n4&};h&uO`=p)S|>WjBwYH?e+e?
z6+J*6d8ecSw@iLSSryt}73^ea_A(q$3SdG=ErOs#Sl<V8;Br9Fn%JCej=TPAnKOYD
zM65E*HKV5a!XfdJ5DEHHAQFGP(Vnx;GcD)56twIlfZPL$9IgTj?)KsPc=i1DJWIhX
zhzM`_y3xw^612<TW&lKP*pB)AYV-PaRclm^U?G__V%7(Sp9JgI20ERHS15$RA66hK
zvBr)p|H#Nd4k1jRf)fF@OAM6=<u5E04i4cSXF5&GqL2_<ydcGlP6f}odiM2LQ^the
z+m4DLtek^*_FeRsMb6$(?>c2>IzmApIQZ^W;6n6(L>t|FfIo(W{blq_4b#iGfa>*4
zXOns#I#`M2>yF7h=TjZtXaSMt%DwAQp2$0WU<IP&)w5dN(;-WAWt4>>Wr{sG64_Z5
zLl$Gtxgd(I4&3(O8UuMc$nBj^H~Cni7}6|DoQEMWQwIw}Q5pO=&rw$nWX}ZaOd@fk
zcsBOJf}MBX$&*$j0t5%7BOB_CiPKbRMg*XbOZjDuoT!RmBtjy|z?=_&Vu~1nqz|1h
zr+=0y9tyf2f+L;pM=(t9ZWOPMq1E&?)$W}pRw&F5ny;O`g#aDIvmX#BQ3pvHOth+A
z!5V>}T_pSRfbg9;5ws3Y4{Z{IK{NJf8N!zW7s!Ctk+4T!ItGY#c;Qrb<y5w0$n=ML
z#1b#p$=lF#iLt~Bt<;+irFRl>tJupIs>zTG2)M!D;-8r|R>%Dt{S7y}Tr$SMBYrhu
z)EenPNu2`(vGDzEwu;d^$2h${DQwPR)YG#kO?i&+@yny)qn;{J_~l%1<*SZxAVR!r
z_{TG=S@DosS~LWRA>?wkN4J4B{0TygCLk|R7ga}iQ~dR_WZvu@+iWJ|Tsc&1fzs}U
zf)2}?3k2T}x5Tl*&>T}2)i>;sBJ;_G6M=J#d8ju9sjd9U3upUK)aOwQf)Zk7GxNcV
zn*LOI_xr5h8?7tCgATf@f=w!)&kBoVz4;BkN-X_|_MFv%7-Xs3Q1_F2MSLk^JuR9=
zMe+fAH>z0S<UvP}kOrAe=&bScuJAYyHlKaAx?`u+QOe2O3e+@@xwGGRM383=;dx4&
zrr^>E_5iosyRhp)x^Tg;mgzdoNoaaoKT8BJ>y^YXL~4kO>cx=%a+(V25X1M7T%KVA
zvx&-zy0%&8dV^?EhyJ(wz1D;`K@}Yt-``CrinX7_oxO26{jtM)KhBT#N;W=~n3YV<
zbK2~;a=>o4q>jl#3`@UoO*kLwxVsfo@;G5dXXzL!`|1Rv(61nKBc-k(FI8yMn60W}
z!uU1V?jxK4P1{FtG-4&mN26lpWvU%>^FUfRHM5NM=(Esj{JJtDYmgN2x!85fzu;X1
zvY8FJ_4BsWY}Ud&5IiHo?4e1=21qvab*X6;Zc%2^EokdBkqiUyuVk)9xTy2307pZL
zgU{Cd(QX~T_b9k$Y!16TmE+v#E&cZ$msLTVG8P2_^{-4#!z=$e^-uFr9+OLim5vuU
zcXN&3;r!Rc@gr>c><Y1nA99J(;c^pwqqc7SZibZ_RZ#_~Bh4oJjV)<tQzL4iBY4l$
z&#-1VJmeSH`EP|y{qY6<rq{l>p%_jPqg8VyUj>VGdSC2drsc2_OTZm7Y6Q~qE1K^T
zr^sZv;Qm`6+zBnDH1nes<IS{=2p%mpzyxW=!^x8q2t!)gieBshvs_z_bX;tox2^(|
zFvL@E$-uwu)1l8zFWt7?7R8BI$I;7Zj!ny9Ga-5vVTYWfk)s;8c2mTvfdnI-GN&M_
zA;m1y7bRMW(f8v5;81tKD3ZAcQ998$;TX;k@LVU5+7iLhyG?1!)8j&FY98!Xi{y<d
zvjCMpFQX2$(4f~6GFTS9^^QC-ZbRlZ#0phi1(k8b6WfYQ-7Oy=7QhtYid|`DIHdjS
zui9wjJSgU(vsk#vfO5*u(C`+Iz?K10Iz>vIkX~5;3oH-s1Gt_z9ueZa<atU|YIaDs
z=PVMd|AhDDfSFXbr9WbumDZYEl3wzhxKPpdZAZs_jw|d~>ikqM%o!@}?`CvK+0H&J
z+cosmp7cM_FZQ(riTeUws9<qsD4sNIdw)rk<^pM+F7+VP^Ap3a(N;H(>ZI4vC3UCW
zU6+TImMX|@M<77+?7ZQ!ZLV@wK(uPHdH48jNrD+lY0}MZo75MkDf<7`3d@5?E+dE<
z&eeBW>={}iX0NkPuSE@jbXp6sW4dwtYlK7lHWnluM=bdk5%N#tSPf?f7&&X&DF>9R
z2S^1v2_?16uXCP2GX{!{eBd(x*Wyy9=S#FCq;bck`b@7+M+M1)V_WVM%C$DdBua}|
z3AT8iV8lJdH|b`b=u$GDM64iu<TD0UB2zpwj?*k-TRi;AT4SV<YExD$%bk5GOq$=8
z_8@h7`2DV&W4r6D^JIkf`Hn+`R-0L#1b2cKx(KxX`cs)R1yu4K1MN^61CbsC*dwVj
z@G8|g@@HMhCb<pP!!oV*QN{$M@7HZEg9#U%g{Yy+|1h|}4AueELQo&VyZ5Ur$}-%?
z(A<Q#(Ma@d0Qm{XPlscV{MIS@`%1VZ*uRPV(D>dQZhpkC@~eZ9IzkVu!)xKl--N1<
zrTBIF6j2Ou3r>})0Xdj?Sn?rb!B_Ij#RRBZ7*5K{^WHk)IXs|lJ8|YRG**p+*;$iW
zcb2vSWF86+3IK0@A%V*}5H-2JQE7og9E8kF5)IgTv)ZEtNOfnZ^=R5Ir-Cs$g4YAn
zLa<nx{v2wrjHbT8RG2u`>bT9C0g<7u4l&#yApuuBY$TeP`<7-X)xcD%9?=Nhf5L{g
z0D~HP2RL(k;_pW~$B5`)>gDM!G+@QS<dvVGjP62b`7^T2K*;!w9rN-M=ENpuR^;&V
zwy`(TEaHSb%evwSVTBJepmz-7HqBXTpa>ar;3z$JxEZZVaJ$O=wj!nV6QFCN;WnZP
zI#sxvz^Opi(FiyB;kmqcv+5%bS>8#{pC~~1h|nfgtXQbrtmWT*?a;cI+x@oykNLky
zXW2QJ|EE7y-M_>+BDVh!=N{*s=m!ZS>4f3&Cky`oJlARL1Y!pgGo`-P0ibrS>h%u4
z6zVk`t_gLLuxL>GKGC+kJ-X~K!okhf!tLk~%l?N^B3-LTng^Ek4OoiTobew<Ny$p<
z*$L2z;Np`snj<ZVPAIQdN992xuoRJh4*+&9bReEz{)P1HucYE(VwGQDp5%0#f;;{?
zPSl};IeLk7Vc0o4wzup1wQmts)VEncRNKUtL_uU9&7y?S;P^Zy#N^Zz`PITJ<xrXm
z+l*qC1N~>57}k2bxnPSjJqMl@n12_J&LmjYX~v=mPvDdyyfHW&oxYR@`ulnC_Uhpm
zvUaiyJ|0}RXlDoa4Irt%5Fw$tjwb8(#o*)d`XSD;H-%A3A^pbpcvf4xcsnU4k&R83
z*0wC^LJt6EWex%zGDe>-WnU&+wxdH$t%%SBH{h37U>s<aJFN0uL`tqdl8g37dlZWk
zlzk&#%|SNspSHW~+gBFm4E!Tcmgab08%Ak-E5@?;g@s1_^TNah2}#e;>?a3r>DP!3
zro=|TBiUuc$5H8s_qg}i+~~iy`!NtPK|Nq&>~>9OK<4lOCMHUlZ#D-qV1nYUEG|be
zIxp^cm;9~$ag&GASyoI>{sh{$cKi%`@S47x{ma^qW{QnsduHrNRuZ$Il8wa%t3Mq=
zPI}Hf#z#|u=sL0WO;s~BWJ+ZhYjs)0vrxp@&yIu<{~?s^oUO$&t8#LEaYcDkRo<S>
zp5bjbU(GcG$CXm039S-NkzpF@5Nb(~pP_R!za+LLFhMbk*(Q?w<z!iasqq-bkq<_b
zM06MyhRNcz6oAh9H{*k@)gV3}OuZygC^z4!h6Ug$h`^+~dm|BaQNwjb;lAB<hr<s;
zKf&6AkaxI9+*cj`zC>9HX|R35-YW3WO}^c=frHEf_hJ`RkRna#z!VYJg>hB@r*NOe
z6cc`mW4_&Wfy3-WD8PHSkyj3^FaRqP)w5klq$r6A54-G@+T4papXt^UL>{8g{}2r?
zG67T6fi_Q7^`*Mw@o?=XUkr>;^(9)Ss1YK%6Mg=)>_^qXR#+ECt4MiqEh)GywLj5h
z#T8g}q-|u2mc+j{{)QPG<(vQFpA<h<090SUdoxN#H)+gV+#;F@Ik6k%3<)Q(dY&Z~
zyw2H4SIbLUyp<Xg6Aj@m&zUpD7_)i_YCJ<>PO%s9o>7vsQam9?)#?peOxx{0$Hz`W
zJHcIhOCr!}p$i%2AZ7xR`J5TS!c9_K0=D4UPhrEd)qFsR3L7c6t{A?VU3~Afnt$^U
ztTM$Yr)&R;_}r{cjG?HpFc?7D_w}jyi|CJu--&voAaMG-1N3ZK=vKJcRX19AOG{>+
zIPC<gVX%<8du{qoNOQF@{?^0QwA=*KZWtPIY|A=kHh*>=&Jt)Kx{6X2Bx7sAN+}w$
zQek?al7S=EoDp5jlroxw*)4PA5T{sgJk=m>6V9mg?_VU(5XL;xkWXl!dyCxOSgsS-
zwbEGHV=#-X%)OI@m;8L<uE(!gH)B?SXTR@grF_+VbBKMc8Ub%^-i1;^2`S`$@m=`I
zvxERg$o{StZFKR;Eu>8?lFro&E&?11g9<fFL`X=AenS<1<jxHcb~5}n*yHP8r%EoD
zVMK(-g_;JvTMDsHRc}1zVu{5A-ay_V6k*-HrQI<lrPBA)<p;Ce_s!?2f9>xjLu&=a
zTK!||qdsIdnHFenQ%kabSXv?R?N6z{PUzGLpi2l6>t`EzxdXc>?-(b`91Z<kihbBd
z?nXbJbFz!~D)yYy6cenn>s_Ma^h`J%P;Sfup3zHyZ~<ZtWvqMMGMX$Oxl)TKd}`Z*
zm*Rz5wKyJkG<*iB6mXktUUp9{kqRoJC$xby7J<Zg=F$`v4Y?wVB7%$ARmi;yZw1R0
z%EP~2yCiEuMnE<7#5&!ccWoz@=^Bt6+p1NFaExZeqiu#nA;(FiPGA<kX3?(yFbikN
z&XTpM53+rv1Wf11lQA*kV|l+dAz??bF1bPDG#Cvkk}HHk5-(>Z#5&v1=*2Z0XsM-8
z%WETe6cx7BT=r|CT87IGI^TSbKBt(&L`jI?qn$m|T|Q!b+jhLj0qi)X9w?}GFR5jm
z7;*0WIHVE>7!(}Co3;IJ=*e-spN1SE;b9Gq!m=6Tl+zC9vVSdR(#jmgJN?#04RkBw
z3*_d#Yxz5357G~56ugONd4@g8G@zOAn(_9!`i{V^&GN_-v~I!PzZ*aKx2p}wZ;3o^
zC?vPoPY=_Itv91AjO&fpy{@(7_iJ66ncRf8k$DZ+Tv^I{MiihVpEHcImc2AlQl0Hr
zPZQLqU|O4LB`(Z>WO5HO7)DiOcn6ubu-^tuKDa3tAalCxGab(o28Ahkjr<1(w2;5Q
z8;0^T_Tb?fuzXt^zS)UZ7%j*`Mweb=lqYH*I7w%<&lHBOOO{L&RyCz#l%>tBL#n;S
z<_)MWN9l4GpqtqTb9>iq17DlkfkBv+zoI5fE1BVuleI`xz&D&ocrX`7F_$}+C1u7^
zA;)hC<%@?uMES*ES^<y+nP7NoW9>4*e~W+PzM^cvK`={-E=oUbkc9H7k!*wQiw%J$
z7JX*sml^>G^Ba938@G_Q_HmPp`|7od$=4!4&Gm(?cPl5WUG`>`72KXFrANQC+cpA4
zj7OfOd$0JfKlDux*H`qCkWqA<E9RFT6z+u*4Mw+4HQ~at`0jC&T_g1%GSU%_^cpqC
zC78^{@|~4{=VF6fl#CaSG5cRHXYA2m)Ei#f!l*%XDh6wu;t%+RO-+ZqI#N`BYl!(J
zYhLsvXMHdP9&h5%*N9$oWq{`iNjgi!Zm&LUpiy4HeApLlH<Y7FhMmgd&}i6;@Jwmp
zvQ)B;Q0P1>)@7MzUn>XaBskg3AGm*44hNiPUqVk^tY%W2jp`WYDPzBxGUHJasS06W
z!Ladz+>I3ca&rJ^yCx2g<j4eZ=iqFD*exizN^VQVEM4o^pLJ6OBn*<B;exxX4vffZ
z4A8xkU;9gNn#~jI3Qtm3mok`?v||(nw{vY}VzHf<iot>ftnoQ$JK7doveci3rs8@J
zxf(0Vvg2Lam>ZlsQ-3Gjw3369yH**CKq9?D9GN@XSIQNtHe=BVhm~&>3+Vf#0+%9-
zpS4G&MDj5N`xLgloJpEA;kQ(diec(rTWl>hnyp7$TU<6RF9P<H`bJj`^q+;+al;F_
zu9acpX>%<f1DP>o2L|;Iw?8Q?3HPWVp4RRu@F+;{mswYub<9xx3zHgPP)VhQ4PiBH
z3^IlCOCz_6;{{>DkivYkx#24(2asi0KIf2FJh+k79A@L6dy<h$2Yzl@jm6k0*2BGI
zBtz)NsfpI*%o4yKE<>ecXUg?V4V}g5qVn#jU(!A4xn5$CNegaDJv+L?fQ{f^$gA54
z#Wa5d>T5R;`sOLMV?58g|BSeW%2oH0AKjT|5XW^(`JQqGZadIS+8v2wc7K|Jy4V){
zFxH6&f*qv<=x9U+GQ3jT_NY@Y{<|8E-4E0_<r$1ARQof<&d|VA>08010TFvjK9ZDj
z{ENwrOIk3LGb#N8NSU5w)}_ZTO3uL=um8f5shQ|D%Hzs6J)7;nOt$X5yWPFSNV@s-
zD&YJhy!brlJQsOdJiApNWs|T=mB4=wu)q&@ceBwG=k2d(jCQ(n&A1{iqUZ_;h*s24
z#=M}m-SWMmbPu5eyu8KW=M}fyBa)f3?p~(yQNaOKZ<7dnN!?rLb1e_MW$H6ccyr9|
zC?hvwx<63KYN;Sq%SApgP4tPuHhAa5D>yOAb@u)CwC5N3CWQZAV{?xGlYC`jWu^b0
zV)N!-nFv<<9k1T}C<AJN1ZwV`@K}^J26!D9oi}*gw85TnRuI{kZJl_XUhaNlq;jPM
zOacIN!YZ|D{kC!P>%Gg~<Y;H<tef3z9_nNksw!=6E?NK_`!^akShQ{x{FJ{sZ!fi<
z@2)ncH_x7w1FSz?WG+t6#Q{L@AO>JBgafxa$-pO5kL;^<kq1J}Of;3bw}kgSdal3U
zSR;E6^b9j=b)*sC2_*aX(F5L%`}V*_<6NZKws;}=mm)c7v`!RgNio$URgn$>>l1Iy
z{j>EZX_KNfRwFu5`U`yi&CergP@iF1>#j2pV1mV5q(Y#D3F4R&0K*1;iUh3|e0!`2
z0V<Vd+(RbN?S~uP#~3qV<2k8un^z&{dk-DHW1<7Ae_QC)3*VJdeTqjwu*En#rO{-Y
z4K*x!YtQig1}~~vI!eRv{e8`Dy(<))y)ux6Cil)&0%)1rgiDdn8VQU|K)Y589HPB^
zEBkk2tdOT^Q*X<c4IX|xJr*$$j8jEg9BY5y-Il1(PBN1Haa9Qv^l6g*+-;m}IR^RG
zqD6}h&f9K3tGTZ*SwpQ5C`fTWz`ILcSr&wi<~nGEr9%ZWOhQ18rnYG12Kpb&?Cc~M
z%+^}Lb*4(K6c~H-(IlHB0~fpv_PM%jD~_ld$@LF6JJ^BMk5rxD)0_0;Z_wa~gFJf}
zVMjGI$iE8H9!`m(kTb|=Y?j2sL6qeu$_au#SZ(H$%D3`K9VhvbusQyOZ#Jj-#laiu
z@-7Mc>B^FjjOEG5#)>o)V^)ujxF#ynkcvQOuAqQ@f~QCRutQr3{^nugn>24=A!$;4
z<U<~M8_Z_@vPxck-iRgCvRESyc)E=IgP9Ra3xu6<gD@F8aEait0kr^c7x;f~rNMXN
zSxrD;QaG5*`H!da0MeO)Ko3z&?SUo?L1D&Pf|gHLg0myj^5WsO|CyU^R{5obk2IMZ
zuXc6gkz)#qFACh}4<=7E<KKXuTc@3!A$@2GCNc|Hk&msgff?aUQDTE+E4QTtC_#z|
z=oZ>PkXH^U#NF&*GjIObvOUn0>P3^`kva2x?Vl+e{B9a5dlsr6pD`zjzPz|ejBpeT
zF(*?>;wcWwp^0_(7^-+8St0nC1}Hgk$rJLW++x7wx-*p(8e++EnERGzPV9^O6L1~^
zsqo=1qd@^}a};s%X!RFaZC&6-PZlIuCs*bk+{DZ<1_slkj~q5TyvWDOgixVMIy!8#
zIwLjriB1nos9A)$+?cBAZDa`bIpp2(oNTK!Ph2Xy?y`<0+o)-TvBA^{g-rFO+_sCy
z1PKD^>DTP!93F>)w#w!v#bYB%zK*nYAs%m9B?2Zq(isVRV-;^{&wCHUzR68dM#LZY
zjxP_6qp7h{T^j9!c~`%pTd%)^(to~bvNQ_m-$%n^wBJF<dcICpPWZgvB<O)Z)Y2x<
zpxrgyTf2&Cd9yNtktSO|ag@UClN-U0Ir92Fr#v}+sg~+sJ=AUi`~d6m^L7SH2NYTc
z+ByLO^!BXM)Zk<C`2<z!Vo*JH&UNz51>s&5#=&+5$<SCmgx&#=dV`U^bV>;4H<Mcr
zjxQ8WDp*F;vJS9D=sks%s+N(JfKFIYfsQTE2G#&mC^~x)tz}wjqG!Grc|&;?$h)w2
zb+aJ86|3htli+J^7z51{kS1kaF9&L`b5|&06$ys=FNXbnBMc~_C@ATje+Msqxe6H9
zRDwRlHx}7JacMY)h@O4i07&cz%Mot5>o%T0+}%?%+kYZ90xmfbQyh^bsw=hr%=>79
zb6Dtwxftb^qz8!6!!k{g8|yF%7_8I1P6SgC(94No>El-C(eKr_HDZ^vQTI#N<gjw)
z>}TORImgxQDHc{zerv2e;Yuxy#gj*;nD^yFl|5_n!!Y=ZiD?O1MTVueiCWak$>=i7
zy$P3>8LXwuw{1LSN;+HL6hfR^&{gnmxE#TlFdo?Z+{@#9hfvScxjS<x=Svg-l8Qc$
zvT&<NT93F;HbUCOU&s^F?n;<y)LfwSoKH=Uepb9Am@c#H)RjZth%W>?kB1SuMHHt>
z;`(S8<5Dz%s9`q?r#@ji9&Bst1l<-`4y7APaFut;A7SOf&L>yYYr|=R!HxwxDNTTG
z1(cy?E<{%)n(8oFHW!Xd)$b>rH=RwOxc?(26Ilqoy%WCbE|CyX?2#~2{$k!&EwB#M
z8owAL%m)_x<R38uubDs+Ygm(RU7IVFoYb|=*sDF<`gMYSVk=GDKvd1yt12l|ak@JY
z@RYNdW;_UhoN+?LFrRp6&;i90v0>z~X~I*z(QG2S9!gejtMQ=xX0u8zZ&Rz)p`Qcx
zD6W+kj!}d6F#)SyVG=IN@<iFr1!HvS`FFf8v-J0sR+fhAAmJ*$b|ORRTq<dTUSr{W
z5MfCue;_##&U%tf?9{M-8lm7^;kKpVeN;u#Fj+PRh>HSJ0<=XjiDgTiVuCNlRg+f~
zdyK=dpbUHFDDNpl3A=f+3t(-vu#AX!d2UeY3IBeep?GIXdLq#f#KR}~WK}k7L0e^o
z3fr%Wy^&P~M)7J7ip8xrEsP8f7Dn#c?{EP6+#_Q+i_@{GJACt)2X_~}!g%(>Tn#^>
z)Qhm^tNfZnY-=CrJ-lt*uLZ}#t6Juw?2_XjXRmt%@n1uey%IvF)+(LPsA|UcW2s_2
z8n=rmxa|hbX=;U%YjzJO*#gzg(Hdxxs%or;dw%o*^bwM#qN+Vg^kAPnP&^&FZb}M>
z=MQ$Pmv@EG0-$1>G95+C&?tWN=o59G#zJ{gATqAwbJAA|%;3s#^8(-wrs4!g((r`U
z6JvRn{Pr{h3DB{`0(pwD^-U>6OzD}k?jGXyIL>)u@n(pjqiossQIFmdewA*YgQ!}U
zD)5VR1b926kmgp`s-a3*3_8<jL?#fTy}=M_PtXv8zRL1n0|;rVl|X1oQSK(URaD~|
zBN0jC@LmXvP=XuIcTgNE^gleZ#nyN(Rpida=2zxVBje>6m5$Qmu;Y1E`2bGbD~1>;
z{ZVGHFpd2WTX~!T_<34!bU}=1hJsrof5H^R!ciylpu=ate(i-p`OE4+v)7FMH(>i9
z<2$ve$%d>mX~Z<={{c+1hC|6Vf~={T`pa`#6ZGutbVI`bdef!eeF386ErXvnnPvs5
z>vt#x3M>hA?|}m(P@L$_%Mo|=fBiKLiAQw^xCYYz^u9X8N&yQoV+3Z3Fxylh=4nhb
zV3F^YU7|IS+8Rr?H<g|oj;B({=(oZCLx+CT!bnb7U@bkz19{I~b{c|P8Z#Xfkt08K
zE(>H!L0HfVrhKO*PZ>0_&Cpe?sn(5*I#}2tK?h<QII&Zfiags^U3N+m^3HlmQm@*y
z#1nf!o`yPUcVLTnV0R!eFCAg3$V4$#WU3e^F<pdTJ3oabkNSJ?VIwlYWg7Id3wvZ>
zD}f{gW^=`wEro;?*UnrIfYF2s!3|XSFr-AY*e=b~&|$y&3@U(qFd(am!%iv^c|58h
z4TF(eHzyPgtBa|mOZ)qZC~~7*SMV{Nc|%Mowis;Z-i#SI06S$?Yw}sekD;4?=k}%F
z9F;s)oErK$4S8ceEB*!nF*6hggwl1Q(^NfjV0g2bJ+`3srI^+2lnbu_dyGq*H#!n?
z2{z1F-UEaY;LRKTsrLYI`UC8w8ZyawVwMUABv-j34OppiOkkJjkuAT*?}4@g$JB22
zQzI>8^NfBStWL)6QWRMnAd_6yIzffSB#I2bUbPlF?~y(EP|+jBSQ!+|muG8o{H>7l
z^{Hj%+-=&;oV@kv->xnixavNQ!2G0ts1jx!s{?lV*kzFocHP|xH;dhBpVKHwUQBOo
z+f&0ct?EP7Y`QuIWl0(5I=U3|IkXXGM~ca=3K*WPDHAQzHkp&fO{LA;DUs7=icK0)
zuBDz1UQaXE1)G~FpV;l!v2;?CiN(;#H|QUV?;U8Z3dR#JNroSRoW|N#Uo9wThc>0>
zZU&&tQ{qHS&}z-+kZ8~ZZXYdE>a~h{1u+#RlvYK!fgDH6B8=<C_Ubf9hdm`>c$N<H
z{m#|CEAXL{*yW!3<u3@5(J9xQdCIJzGC09uEqnO9|DGgvkjz(>6MLALp(-ix21WW?
zV{$fh|GD$oo-c|f{xcN-W`!bLX<o4)omywQh|R;RS(h@3b~tGtE})^ddm3XkdAIG}
z#=aD?%Z!IhNW#1dNj_1Cx7}t4O_vvgvBT@M!{gog@*W3|iX$-lzIQ~!EkYG)-!))6
zyy*S3@#y(js+IIh!}^zuRk(4GCa#eS?$N6*MT3-rHVK~7l$;Q#SYuEcT`?!|>r`^@
zawh8gD((p0FMSGcT^`?6&b7D+Kyw`U{Gx$8%m#oUaZ2Mm;XG_BfxrprQBsLP;2gq~
z>4=s_Y)m_bEPf!D=z_6P!ttpsIpJG;In3S|M;`t|C^HRSO!4>;@;zot;-lkP4^#lG
zDwGvQ!b4so{;a+p`<#@g-CQD$IcI0fWlYn5@-(zBxj&e~v$SJ$b3!O+$XX;Jccg&+
zBR|mnTd`BImm`syLjW@8v=495+hP^uyL4%ISP=>^WA9+!{1?Mz%qg9?_3bb$cEt!K
zG~1qK!RFkLXu4s<!6q0_qVdDzGkLC<8*`6jUWU8AS)f=7YSWh*2BP`45!xffsiUQG
zu7vUqzO9)`{lLlwH<%S`vB&|1Nj3j)@afN}(D0P5R4JPf+}UeaM$CJ2X7y?H^#`$4
zt0q!Z=6B=@7&RKF!sic#ZirX+Qz}Z*&yyOmoX1r46?;92?Pd)M&W?OQLMeSc9{TAl
zhk5Beh3kZloBbD;<9%<J@xsu5TbZ)_7k~jhJ<I>JGX3=_O59+D?zyR*!&49a1Ho(R
z%5B7TH*>h`Y?q$1PMUf=r~Uo{#rsH4qk-IQ3D?l&?LQy}|Eh0b^3t8Sb+LGMbv7t+
zHUlZ9U+?v?4B#gI4Llwl$_a%Z6?lCm_VOem$?To|x}ZoooO*N<bK~r;y6X;ntM@0&
ztdbzA_txduSNveV%fLE`-<#kC+bH&)&Lin?@>bj^QQH;>@@3G@E;rX$72p*%@uRf0
zv^Q6hBuNd>=j9`|bgl*k{HRXIELh~+_j~PEDQyxrCzUb~V)yyRNwlC&guRacaADW*
z5yZv!pns(EB8UBMq%P*_=hAx*Xcd?pr%-3u_r=rSAhee~Ybb_L)D^dGho<mm3>xC_
zQXGh!BUwK`g<Y2}VH9376*UbdQE!cy6Jo}+m_xU1KYc7}Z8!uf*!AYj%Z-Hr9r+_(
z9VdFKblSgJI9o0jhCn%cbi+C3Sh&efE15n4L$zpeVSSnVk!9A&OeFr7X<&@qNuA0H
z<1kD^;o$Zw!E9$Hs)TQ?n92$7x!0vo^*Psn7V6UYkFcO5-z$cX`!YBuuD2^7;-M7p
zkEN%`uE`DVk`3G0Jk=5DRUa%9C4!Br4~Qa72umAMktURcQ@A2afhSWz2jaZxv>Ln!
zU6wKzK}Av=?lC^gzqkcc@p8lk(|>CCAY@s~Tmli~{Cn6%Etra2!?q|3P9#~=mbHBt
z=)r~AyN6urEMTRTp^J76b4ved*a`cs{QTxc@cJW50W!!2R=Q)HeFOVjJ3?jOAw1+7
zkeGGyVJ*(7I(Wv4<EjQ>*7)17E^PH2u*fNEHHDp3O37})WX7pFuvYa9nMJCY;WNjq
zN`7=a$E{Y_9!;krT3}kO)>kxS`YezbCGVOuO9szA?uIh@g<WL}C%aSg4Ozv>kl~~M
zA5-c4ib{DN8aE#J*o<af0VNW_Mz{<~C>wnU*$P(0R;h5i|BQ)rktv?I0z^V#$ig!B
z_OxsNWKyaq`LDu#yXqb?i4h^gCl468=pS~HGE8x0Xv6(zv1{^;P`M8w(+59=K-{8}
z{4&Y4Z=y;DC;5hoZomQvX`_;4Q8zNE1+3%;F5l9dDeO+o@1w9Pr2FhP>7pAQ?`9ES
zLWeG%j@Diut%Fl>$aPXu!KaH$JqwJn8x&^?mC=In*>ZOF{F(qotN9lVi9ZV@RMo9d
z6L@U)L#cu$xhAXSE&cs_2~^B#n6NdsJmuAiRMkBBJ1%p|f1@sgN(lEz>C!E%&bx7D
zBw@Gz(|H>BIIn#H>dcJOtxWq1t2kTNmxY=ih^qLZF`l{r1e_&=styd5O=vS~7h9zt
zp6U^j#2T3DUv34u{&!l5Howzh&CYkLI!uEqpQ0tB#_*2w6ns^!6~!&ZlzHVPbc(`Y
zTt6oHQJ@f=#ehGf(~u%)`F{wTgL{<k5wr$SO|d$40@^{=9RuWtiUy7z4nsy8yNab#
zBi8B-iI!4UKcu*1uXjZ`nluD^3TL12I9GCnRc?6EC?~NZiVx+>7gnZD<-)+VKR?*T
zulAx)Sb&n8!;xuKvrfy@aM7qVtodqc%Q4Kt7+x7xhJrItV3*_6LSo;n9elAaqwH#j
z^%{#c$KK*Vl$I?u;8P$3Y5!3cEx%&6I6pH`D&w$s>Y64++6|CC4M{bY?H!!F)R@2B
zbDj@yZ`7GlsS0T**Bs4{fYZA(y|0V2Roi9XVuM2d{qu0aMdIIX&A^Fyq>pHw{+9u!
zaZWTQUvM6v2Gw@>zT_MEK%M+wD(5f|3kTpmW3Q$=nVt8b!FniOYaI+~CC#20OQ*iM
z2kR?yElF-j{RA=Qp8deUA!v~Dyqg{r?Pa{1-ktOqp2Dumq;ej>9tT-3DtW9AP8Roe
z*tzHWr>>FvpbQc8$>Br+)WAG%SSRLHvgg>PYZ*8Afp+^lj4Gpa@h!nGtPwTkb#MPD
zi*DZxZP%t<keencyS!7=X1tPM72Ob0f+T0{#lU^|3hbS7)(O^A{myqt)W_xgAA0mh
zp5Z-gTD^259Ce=TYVIdNTUtSuk{n=6`rr4juC7I)wd1;(Ij*e!d=|6>({#3wA~;~W
zLtPi7uD+pVu^=?VRa=$!5UYPyn^sQz6gm17c^p(t9X{K!?GY_Isw`_+i}3pFLG|N1
zefCW?nxwNK;sg8Bl4XpLX`A-c{G~lM&*|}LuJk^1#7Biq!z={`+h&{5#G0j8cXMgm
zLu5}Qi=O&kr{kLKcXNr2N*z)eDk>mwrPI?uIbrTg$4G%E&=ljlNU-_?bjw6)ms?t8
zsWZ?I3Ft&ek|*IaBsu^nBfWTQ2gySG*a1LBGO^PG$xt4U7-oT6?j@rRt)^Z*U9=5u
zNt*0n#fmqLL0``}yGPrws0{kJKE;Z~GH-eB>nEnWDtkybW^$LEulSu<Yq1<TBr?qL
z_L!zD3O)qC!p2M~y|f*)nZ;nzdN;6Wq=}Y)@nM~k|HiqD3HoRSD>=osGXjVQ)T)}&
z0*G1FVtk2#B!p|@+Q#yf+$luWOz{ae7E*9qXGBOL*TDfF@Jw)qEvYob1{-ONar$zC
z$FAdCkrQ?ZZpevM$#9RlLcT<4#yWo}9>`U92!40&_z&d7G{(4p&VJ@C8DsBfuHW={
z`nB`_TZr(BU#G{nGqi-{=BE4q&2Jdk*#4&wp*dqbhKLQC|K<zj38;Er&=LR>CMnv%
zza8XED^4l!FoEMRX$P)3MVVysDCSt+#?68VkK!-maB44V&*!7#oBNxiQXRUf0(!gm
zXFE9M?xUK$XSoA|Abs`F{hiO-Q}nOxO#2*A7L`bB{!I$+-Iof46btYd+0aKNz~05q
zX}e;3_>*9#BHN??qd298&|gQpa@0~WB^evT-bdLyziq#d*XMym$QRyT->%OG#iaSZ
zHs?sz1c$z(mm9j??~zo#vfGV;bVt`WHs!DDN~mM_)ZSD*U0+;8#;kmyb7U*Q-BSXM
z0C9qOfdv=|+peBoNalo!IlsH+aFyNtNMqVSD{4SJk+<cWg~msjz8CY>ZK16>l?@OG
z|5Q0wf~DafFAufN3nL?>9ToHaNx?;%p`r%0#TV42mmRzExB7B611m!KDNX#0M+i@2
z!eSI9<q$F~m&>Ra3$o)i4aaVCOjXRxtEvW{`i9l>ucS8hTUx5x%xDk>#9(NA`70EQ
z%EN3zV47<e^Ih4>k0=N@_BZGQ1iYlj1^L=64v2I(2X@G~E1*yc6VkO&NK@p?(Ntrw
zIM_n+T5{pnb2WGsKIKE=V-Qi7h>%1Noyl5gehY*p?mqbxjw?uL>^}Liq~t_{rh^sw
zTV`l?OwdOI@xijHzBwRtMTWRWo<f}7(0sYUzET5yyik5oD$@W1P<n9%`v2ErAUjCk
zyzFLRUNcx}g$JxAZLh+=`hh>kA^I*2Iv=AKR97m{rlcD$pxuE#2SJ)=O+5g^uRbU*
z_Z+_U$^!n9B=%_sKf<ub8mkV5?>#sCWDI}x$`by>4F0jSROj<c)+}d@ZG+~Y6RlSG
ze-~`a0t_f^U4S;c3gS0ju*G%2@#p6(Yg@zpPj78Gpa60Kt28OQu?B_$dG6A-xb-5q
zKIjt+oc=PQ^Ao!U<YrT(u)_tic>eP2pRQS?YIYAm$nmN>tNR9Z$$RAeNvZ9DOyRuo
z0{}*5UXoOUxUk&8g(PM4{hZr^0blJQ%s>dpyNkXWKp4V#;sgOIjorn`25`W+_oy4$
zf;@H}e?qZ9JD!pRm?Mx0)q{}w<;^=21q{`dcJDYyyu!P02(I^F6Z!pxL<5=eLF38g
z!*>t0VYv`2Vcj!=c&7UwIn;;YwZhPW3l1Bct1>R(xiBo@-;sX@O6ZgqoCv*RnE*6^
z++0emVjNrZ_bk&9fbY-4TRyx1b9HD#1%4T_iZ$8!3HiZ1l-QUTQkukN^~|1yn45ds
z0v*o{l{j9OEC1%eAatZZtJYerh)d_~O1!<iM-Mfz0t7ZH)!ASOq~sDO#A(n4$@!r5
z-~Qu%W++J!7>#d0F@;ibSr8YbLFOi90@}WFm%Rrl#O?ndcW)V0Rok`;i@{Q)1VkD|
zX=#uK5eWh5?(XhJTnI=CETp@;q#FdJySux)_r<O9yzki0kLP>${;|jBSc99JHP@V1
zoN=7Tan89y6W!^&N!*p#C_9gQWIb`34<Yp{5Ro?4Gh<FI)DIMC5sV?6TBJ}fo}gq`
z>n*ak2Uy;MzFpM}kb`>pM&<Qa(6^Eb@k#yKqy~H=TJHVwJ5usz?7s9^u=H}oIUzPW
zx)i;0VXzH|vG7{l50;c&EyMF*0YmTHeebs4*~H`M6MKm;_q6$yeglIbzvqv*9C{zF
zron>L1NW#$eL{FRQa$c%*el6fK5aq_^`sA^e1zxH()+_vItY3{Bz@@I)yAR95@s*O
z6|Ru8)5fR7zWbHJ#jqowQ2#;bG@ah%LE8PzRUx}(x6m%{b0fZL1PtQ$A_ho>YJ88-
zdOFTVbIB~QY9-#qNPTXR9(ZmP8v0P`v98h_t!t78SBi%<-517RZ=MqCvln14Q&`|#
zGYEx;Utw$e_@KFN%7xglAWpBScdf>2E$Mz+AZOFBHSBovQ{$o=>#oaClOGx*-=o_0
zU#)C-o`APsw*C}Oi@hN0gItF_T!eIUSxwF~51-Zm9@S*B7gXpWVX5;wd=TfV8l1iN
zw7wX00=v&sBo0t9u%34yAR_kvXg}16SHB|@lg^TW7|8ksLhhZ22qB4%$`+CH<j1xW
z!k*RSn&CNnPRXk$M50lsb4JhDb}+f2`ss3!-~C`FF{#HIqka=+AlZS-wR-=Fsr6!%
z>8$QJ6W=fMGJ)kqd+P$p9=4fGUXu!n<vss;dkqx*B8j0vi;Ar6sL(!)`hDCO)3Rqw
z19VyvufH@g#wbagu$eA6*TGlvrx(4#M#+8D$rM^)9%cTq(9_rZ)TAnLWGV{L-i8R(
z=*u{aZ$5}T@{M`Mblp^bf$E#7=1TeJ{#rhsxz*xdt6(E`vu<IBax`>=v9#6b3RCp&
zYS0X_daVsp=BX4T5zKbKn`H6rdUlFM&fDA4^HtK)=yB%1Q+aPgI_RaPZhMx-%r-_W
zh>qf#@k*!<ZO>IeART{KdOGFI$RJEO*8ndjH+gJ@l3uhVNyMn|kjgT9!F}P$czWp=
zVMF@v3N;6BX%db=M-k0Y9KYK5Y00Wbgvc_HNoI<jp3%sekE5re`70Tv7e8iNaZ33G
zsuf;3U*jl!W2%^XP98fN4HtqiRZ!e!U+rs<{-xAFUP|cs$E2Ladi<CAe#5epl?|nm
zFN-rsg=wXeG_A>X%(+?H?dzWmJ-p*<tTXzgbG*tLqf^m1MNO)_e@M|-cjDx}fQ8q0
zC5Pg-gI(2h#+avl!pa1XQ<N@J$Oln=s^^UF2!F&JQ>`$qqnbB8hZ8ib-O+AY9{x$_
zqo$lm9HA-u6f-p!zv1Z>z0O)nm1->SdH$W2Sdqdap|O@yg}jkvS$45%BZB^|G^JP}
z`HzW=tdx(Unn;h)y)$i3sZ|q#pZX3ot+jDgJ$C!Xs75p#6~JCAu9JYl$3@uO>Vg4t
z(jWDxaZ;Tji<Vu)St+3~q0}8coMY_VqBeYT1a)Y6l8Eaqv2j8eVaO%$Z{f_pd+s>P
zoBs&T+&r>z;my{Rnglm3&W8bXeMWH=hU&U0v;(SN20;t3#-(c%;Z~Tjgzc%sRs+FA
zuH=VTP0ZxQRTnlvVQ$q`ZZ*!=cdT}a<QESY_q%-?2wz>vph04)?NdDN*_>XS?~P}{
z#%A{e0)x=*5C+EGtJxf97KhU#yGR`vXEu|Lyms$Ka+7PUs><!^M19SL<5s?;_PnO=
z_1D10h%OX!3z1pn#VOb8i(}tpaAUs8S-9%Ocz1W$b<GY^1eS;I(Oz7wS|18L>{C1q
z#f*~L)#?1+ayt#4VH2^%*Zfr!*R$ok$GE4V@;JnW(rkJrjgrJc-Ai-I7~foR<P8&H
zm<n1!eDI@2K`HPGx;8jQZ_Y1YrYrm+LC|6Qh4}(kJL(kcd1^N{n<m4<$&*&BnnkB(
z*#yG{EW<BFJa_05dxTNNF&ho}v>s^U>0^>uq+GOi81jEpcp75$t{ge!EGX9Vp3N8L
zyHBBcJ;nlTyF-29uPP=@&wlK^qa5!|(>3T<d^A7zWNe|O>b~DI_zP*RLgf!ig{jXU
zIys3T5D38LN4A7sc&AgT?aD}!IqMuyeW*vs>wGd+hT${$8NJ!-?uRLT$gew=aNliv
z{6tul7ry>bcuM)o+Dde7xPDb<T<+le0<Lys5MJhkV5U~qOSXBYhpH-ZdIJ)C`)h*V
zf1TeUck((o@hOr&JI9l8WRQu;el<o*|Dm2w#%s*4V^<VgLm=rcQ74kvly|36l#G1Z
zdsr-72G`!p(>j`>yv$mpJ*GnwlJN}JmCmvJs6Ny$O!x^W$>!90kKk85it&wVJww0U
zjU5=5e}qX(QNb#UnKCpkbMmAp?0Ju>W6|*A7f<nf2kt+^ZzMp+)6gBA@V+-VQ^ao0
zSS#2WYI9~Iq|x+IvMT_FhNRV+-atRFy5-2s6sdz>fUUaz?x?+<pDNm26?u@3L?-M<
z`_G8=SNaqBwX+y@x^&yt^&BhRSg@{yn7G1Fn74-Nhp#L#Mb?{XO_{Iu12DAH)-DFA
z>!=?(v;Cs2{gC%%Fh($|ktH54JjlU>P$mGgEhGuQu?B5xWMjFxGGM^#Y5ILWWCvCE
z=fshj=;tpHw3BwWdX2yGIF#Hi9vQ~)?j37lNzC}lcWywL+GXH0vG|RGzV0yr>F1q-
zs7vpL@aLZo0x3iuFODMgYpO|w+a6s?FX1gdlAY=Aclcb{d-s*~w^`HXE?v1z1v6VJ
z+X|mIa5W!u!WqQwwN~k)i6smb>{Uz{lD<nH{PD<ys9=blQat%dZ`OQuk(0qt6ve3M
z2K|l<Z@9O1F6HTEqo*+H8%Ky0zVf{Iq;#k4VW9hIc`bt3be%_^$LkSt+-@c2qI#Pg
zBV$CL_cEcMqXlZoui5N1g^Q8i5|ks$-2LPbvoIkY)T6&7dMb-(u5KTb<f|AOWSZBw
zRw0+TvgPV!TC(eN)|X-U!nyTDO@GyTkL?S@(hoR^CJ)&nD7fH^dL@(cdo!>}i-Llz
z3S_?qNEg-{Xe8H=8BQipSo2n1zcU+U?d(1jU3`T>0QwI55!Q7l5ky<JJS4c=*sWTo
z2VX+Tf5_%SEmRb~e5qk}9waN!>?s+rD)pq1oAb@>O6oVFJ;{CQ{6Qe?O*HYmAkBU&
z-{7Qssj{6HD|2dXqkhZb-eUv#Q{A<+bFW9}KEX@KXY}u0v<b(16^Ih*mkvvutxD`|
zDjgd0)K|ui<mIua-sft3jHk8w3RUOmW0o8N$&}~x!2Azsf^%|Ct>FA)##Yt@|5>jO
z>M`ak{&5zO+Aw~!i)Ya!biYc-wpvE+wDK4ddyghchG0y1z94cHX{%hHgM$UFAicA1
zfWv6~F?+`+=Vb04ndtPTX*=8)CUX39AviX!)jK{K<K}moWSw{6Fw(YN-7P+qe7Q%q
zBk_rQ_TrK9TmGyFerqffCl)@*5OaxWJyDb?qy7UnlRaC9BE9nc5k_hrePdMS?2Hfe
zGV0h)*@zTIrZ7=eVP1vfgy`a5VGW0ZatOm7)vmY&gJ#=We9rHbaI&8@cj2K(qxbu^
z$!e&tX%R)MH6dOp#)&sszk)u+?D(-LN57qEZ~C~Vsvi2?zbUT3r@zn~d>0q3CfTxw
zSL?3d3*kp<<Gb;qcayYRUm)eAAYFVfUD{YZ#f9qgO_Wjnq=s#Vn+FyWe;nYK2VXQu
zGJ3q_{XKv{Yv$XtgUlb@BgWww)o9yH_$_HsQmYJjz30!Xtnuc9?*>gV>>bEcz>-=o
zuIL<Fh&?G$MavC570|9DS*Mz(lq7n8oR)t+aIeu@i)O`5b?FY}?R0=cHh2Xs5!}|-
zG(D&<-l(TLDPKZ8B=Pb#V{J;?47ca5W5N$Lr;Gi`1aES0_zi(Wo|VMrJkc)YcORse
z=XH6jzhbV)K4H>_2e_q(PYspd*Lq)f=C{8I=KxP$#VQ<)l#kT+8k1UyCSQ7MG9)ud
zp!*K?<F3_~(zkMByMAUZMb4XvI~pYIUXcZIHO$!gaKVAtl*gKStm9r>azoNxVH|^Z
zyZ0$QQa_6qh1Ow=hY#vaepOnBtbJij8QgpNonxB0bc+8ZQ1@sI{>g!|t{LgaY+5>I
znKeYFYL+CDFg+oBE8Jr4%haQ|wCtEU{SZf$9_w(=qumlyG{=lz?y#o7Zvy2{Nw-U?
zdu2GDLd^vZOP8x$`gr~BczsX}aTU|*nH_oxt%PYxj{8VypfDPF4n2nckdA{#ZR_5u
zq2y}zR!3oOE0K`FVKY7|{B$fJi&oL>*!<&}!yD{F4hE?@6n9=uZVgTHCHHt7XkBr1
zpVP45Dl(#y#oMQEN459xx4sA`QXuqXHae|*UvX6oarm_Tif#JP<CR<grg4t9dPXkC
zcZ%S5lrJ^S^(d_$A6_dYj9v{^yep+7kXXU&i+smW)m<(j_cZc@HD3aDPRr9Xk8A;)
z7sYDjGKEdnnEfaX$wV`OyRVfJA9ly@B^B5E$@FM)K;-j_%WYI2>8v-r{>DJhMic!2
z*+q<BfK(8#L6N`IF_GoqJ?SF1o;Pohz@$DkTnFf%jbjTwTxiC7safF5oGci<a4K!g
zOb#!VfvvcN6lZE$)ap;XU7q8Q8MTr7g~V;DI;V5gJo&LjU6>R^KkC<=eW_3ReA*CA
zoMeP|Hf(;Mo@wDMUA?f5=4ycWpyQsRa(s8r><?RXi;GxiOFKuxnelLq`23vFDBQsC
zo5XQt!Cu(}BV|C=m$I4SC!8erAQL5)r`!o7kjKT3ooCFUpM-qhVbesVUiW*}>-Ra?
zBbW9VvKSS-eub9Op<TpoSJnS3kR{2Qc-$8Dn)T(GKuD!^DOyTVqP~awjv&6+B^0wq
z5?jo0O!O*K_y{9iQTuT)O(_)?_m<ALqkM7Z-As>{{JpFAD4Bw9dZlst81|G5_3^z)
z;ftz&1-;gH?bTvbo}Yi@IB{38SIhJ8k=?ZoMlZzE{zDoaR(Q|ulCM?kES@$xZu@l{
z?cHu!3&B~;k|1fcEZaR2ByQERJDE}y_7g8B`^iX!e>r5Z6-kJ;P2HW6l97<sxzlA5
zLZck4vmF;lbA6A^g0A!5VOi$?{ic38R{H-;y{lGi4W0l`C+r5_p@WG2sy@ZgFw^n7
zZK&CYrKP>ruP$FvZXPgY9{T=6PD-U!&w*g#9n|(lqOV&5q!r*>jJHt;yM^p9@e)EO
z&ySH$J!fjkP7)Kkh(1e<R9|n@a3H)uBJ6h7e*v3o+`A5NTU~{|YoYwsjT*-#@2_zg
z5%{bo&_KmGH`gD-8>Q7wbT?ox&>?iDaYslCU8Tme3Ujl6__!N|$c*O`^X&9ae{6L%
z_r-pXsUZ7f*@&(2ZZC6Q%KMAjUaFse)SMs69be9te%}yRLsn30z)B3N`2oMouEmjV
zPMRTH^<HD0sFglH7Kv3{#6c&tBtDcX%RBUa_EXJJBU$owqRX1SyvL-@9BJzsIVGzV
zTr^hV@IP!q@8QaM&9id1VvtEaT-FX=<6($6TOJO~E#Es&iw#+%)K?UhWaMXzC44&T
z@Zsy+3WC&y{gI~Z?vVR3F_l+#N<JY@Wk=#8PWNHM#L&{J^?CY2raRRp<XO*^3SQ}C
z#W1&05W9y~q>yzDw^MN1r;fC(6^Vr}xcA~3L}kB!`)-;yZ=pUIJoSk*&i`!SgX#hL
ziac!CA^vMw)waF&a#j<yeEpkO2h2O3I67KBa7aJ$dTS|7Rq6;PCAEV*eg=u46B2Us
zaE(46dwCK#PUQ=GI3{xN<*_B~SNIxHA}RRZ^m}drf;i5R=%J3|G)YYAo7%@GY&1!{
zGccB*-HU_zho|A$m@`$1)Q(TTE~rTa_l7?1O7(zC{HpEV7=R@!38xS#S+2@O%v1KY
zI+4>n%J@a}!3J{DvOg-N-WRT@kCsG+@q+`eG4fMNNV|ANWf~njOzJaw3}+M#h@yl=
z@im%s^ax>)-3n2oIi`Z?(@^*f5=DbxXO$dca|@QBefYN;j1Q&JU8wGSv8WTMWV_oW
z8Z-G^#8>ahaDu1rx5TGw#NLv<OZjgL)dh<&tgZ|#pU@Fu2Q)O47{lOgNRCFWy}$F&
zNsCLPyk~-Z<I=Eu*%|y&r?k>{qfzmU+?sygvf{g94gsrnPWGYO&)7CC?mJlb^tFX;
zcnBZdi5#RGf57aB55B1mB_%V&3a(5BQZUu3D(WDdMA=BC!}+DH^MKaq-O}UN?|88_
z4u~DaTfHct8+WAFxug{a6$?ELAL%K{E#<$X`}LT&He*jbs6|FWB%`x1&s(HA?9F0s
zgV8%4WSxNwRe!V<)u{(P6l?v+(fP;56P9oz=8Bfp6Kus<%*F=~MaUo%&7~9H&AFDN
zSYtZNq<jP6>-VLSvEqs=kGnkJKaN-`jYXKNe_<t3Lfo*)t8a10rxXf;k6vdFFP(qL
zoccq4F8Oo1LN&We9Ex`8z*Do3BHh&8HQFJz1|{5`DdWYDSaqd3AIlTMyA^-Uj>|n{
zfTtOJ1QhcT=@z*`PVhCEr8zF?lZH=apDU?F9}VVQs|#IjUrdNux%muqU6$n)c<XG)
zksVwpYcHQ%V+VUWxh~#qrBUNMCY`?M@C~(_thHF}ZsWIhvvcz(pF_}S!O?6`t}v*0
zFK3O$p+qgQN&aoGE~eWQ0e4YTY=D3!L<j0<xT+c%t!Xab^t1dG7RpPbNPHN%u!#y*
z49wEh&+thTMSiH^jx@b<bZ|60^)kFmBDU?q1HDuRXym%ssbR<IBE_1co40uP<*P(^
z$-K)Tbh#peH+7L3<<-jsD>#AU!kSusbPB`+L0Cpm*0bm3U2qOY0~6!OREPpUaQqiM
zQ>I2%>a~rw#bEQ)MTB!+$Io|Hr5qq=D$SrJ#kU=n*>-73%(h+5MJ&u6p5k4=6x!e*
zo`YEGoBf>E5V1MP<^!D&s-THd!q6OLM%saHBUu*$w)CiiMH0P!=9WFb7@AeZSEbTH
zRGODBS6oSyq-A{yt7+mfO1Xr@(4JTZ%PpOpr$q~dzMw_ZGvVB>uWVy#*3ZROq!w1;
ziG8aayh`vZmd?eXPKJV4jU@M^tgDKve);5$-YHvv0^I%FSFG%EgE4WzwdHIt<ddJ5
zhrcr>_wH|#LhC6(`J&6@CA|urAbTGxyrbOwjCwl9N@~r^LaH!QT;z=cgnZ<fscFVt
ziQLl;k~BjxI6;NQ8pA@hjp<ilZx;R<8QF5-4bh}!k6qcDu%BPx9@{mJDJwo08yO#7
znowy$d5}vMY8x^_rXcuwkE^Mi5zTL98os@LnUH2x%fDyRvzv2=3Ewcpi1=ZPvQej>
zMhdoTm-tL1gQ%La8AS|=k>1BucvwT$ga*x5|0i|YyL$npvRz^7HlAQXr=}iB5wnTQ
zPCt{vC7}!-%Z|lABW&hp#QDKg`jg)pr|P)0h2Q!iavp-`<f}JApf_&f5niM3vyB*H
z=XB3gtzvPi%)P@8Teza*ch<{(QD)NCKOT_b9~0J0QlnSUI=Bv$|JAFVT#K%$5sv@3
zf<}``L>ASEKl$T!U;rbBYSS=Hie)}$)PUnaiI$3YU*Q6|5|&r;D(~i)0VCWm6_*IC
zWJ7Hhx|y@2=+nI}I;%lw?w-winP&=p3YXiA_MM4+T#?u=r;GO2C=^;X=Kn6Gv(Wx;
zY8xE`{eO6?h(c%ZO+(H(`N`cSwEJ~&5glFG(UT%p!j+{Wq9!52c5Fa)to`)97mbR%
zK#uiSA^F3)WB2=TV-aS}5hz|;r{>2V+KC?#d!v@~;1}FO#LVfjM&94SMeW%4J-z5+
zC0tlrAX-LU^x5NVjoA)-_~F%CSGWnnm+o!ctv12Sw7(9wWkyx+wJ$EV@?PKH2M@|y
zKp?`Ni<fKuBy)Wkwj2nNd!9W>7>APQ^~5ht+a-YHy%~PN>Hgx*sV{?DYhz%uLPzVx
zuy0K|%Seau6`dO%mkHfM#A`mZ6b>El2gHvREe3;JpLq)INW8AKiH67a`5ef*@?x5h
zwTyuC_~G(H=lnWVKlhIqC8%#nq7JIz>hC^|Hu;s0{<wGaPOHbKp%&@o517_yPDc-N
z;vv}gctoh@QQmVDO=BwP3HC@F$<hA!f}HfAJ8@ui!{9yShq{|~SR$XFQ$bYD<JktT
zHJtv+{D>sU2PisNUK;v8Im}gJ)-`?B?y{3a@#dB|++#0nTCi)L8d0kwNUJL;NL$Z8
z*C!quv>mD2>J6S|-}1j4pO4K8WY4#8Ug9)Wf2NFa9N?to2In-=jQn+?%xb2g*nKkL
zf!Cm|VoMjRl6yYGQS%jslR6GhlZ?1PoJ2_W#n}_tE_HXqCC@5%CcbtJa3l25v$<$D
z(*SN|&K3pdPtHrH8uuzq_o+AhstamAT^?i&F?;EA#yV`j$!}jGz_JCytK}-4tf`3@
zJ+ofA{NWyxa;M9w1sq&hGt7gZGEi?+;8Y)rgsym?H<ckoJ>O!S@_icHZN3rVs(*Y>
zsoqyzVlsSYGJRUUWFq`1z;3>=L3#T~ZTSJWAkS;u0gP*Yq?W=>k_q-Hf0W_w<3syA
z+s~3rYjV}awYtArWS{BS@xJBDMG@2@X?o2P{4=iOxVBS{XFNT|wy#g>{TIp(5q1Of
zcZk?R>MmI1b>fPReM5rV!=;T{l?;AJcc{Lmp6y;n8SuMan>RO%co~|Cob8fJ>CKWE
z;NxlpXSXX?^>{w@YA!ZYFSpNDF_x<e8Jkc@s}a=Qtd8>cwzCr(rILJUMAho|WDrxu
zB%$}Cy@YQtq%4X@p4jOp_wY)E`)o~+GewzZZ#K)UV`_@=qt%8!oWTea662IX(!MNq
ztL&PLz)KBx4y(uqgnGi2dnDD=k{>DiS5|_c{J8?p5c-&jL?vO(6N-nXHHc=j=DAO%
z%&b_#mRg0@%AGN~V!xjT<9TJ2NO87H+f!yw|5$3Hk2Er~&**z0|L|GOj8*}z1mfG%
zmCY{xR9hYDCmymPv9u}#?<NVM{hgYpSVGf=i6{H6_OHknE`kXiza3+CN;-XjSCH&@
z=CdyBo~&2%(95y#4OKG5s{#hmbW;Pv2`;LoF`Dn~PvHpDx|Eal63DIOhji$yvn~d7
z2J{tXn<2sprpV4?VNvBroMjP48|C%;p0eo1`u($K&UaU(KG86Oo6i@~D_Ml;3`V5M
z2c@Q?P9u%ngxB!Ff0Tu*wK5}wk8!O>^(p!|{PK2R4r0Vrx<s0`WqviIaaf9!>(7Jb
zevS{SA-)Vi>|#2MV1iv!<l{;1ikDN4l_v9s8Cz=%&e`L^AX#j~3FE|YG`Pu<MMgwD
zXEA+5KQmwJWZ9TA+{4_n%m}ATT9g*MGnw&IKjTi^#kSS%GtAM2!taN95Zo8Xf$L#4
zV}w&KocE~|i;rmR9PE$Bg=x312it1BT@+1NuOPE>Zp!~I>VkHL+a3pL2Xh@7c`;2x
z9W84bDH~1eoB#h{qM@flBWIv(t^fQDEiH|pj)9)O^>aq>>pKH$D{&o5UNaMOGgBQ?
z>*p*qyk^E`mQv;#S~@iEb!-i^bR-4d{q3EgV`OCc{rq?uegk7`9ZMR1V-0Ja_c~f;
z+B#fZ2v*jXIvOSj_G6oq5P6K@sTK>`^iK~ZACObXYt}#ij$8S{d!(1l)?bsasy8bL
z3tg#4-K+p&#z;q@89AY!&7;OsJ*1v30afSW`OGiwzHrIiF<*7Ql{u@b5IVZO9nrGf
zVG+)tr)$UMq^oCleG!8C?%@lMKR>*6trputeZK_Vb%LV4^Ui1*AFmY9AR;0f!3Bqi
z^we`-oj@<&K78?08-HV@Kr4>RB{(?v_nTA=)16Vw#~Y*btE;U_9^iPt7%vpUt3~?b
z?P=z0=bafI3za(OjZAnX91L9Cb~qvrRLx|OXqMi|$;rOHzSdT93W~LYcIT_p<-OXH
zk`iz>v5y~RD2j@T*lpHcg+8T_4relwl^CnAS(lZU508tBD=oFt)lF4&TkeXXR;l=v
zxv{aK+Z7!V6%`c{^77Dbdn!kz@~AgSfK}vXIIu5|pT%;!sUoztwjv`VPrIJyOx=~s
zQ>Bv2ov4RG9wQ*|`{NAeC^4tFUYzW1j+chHFQf>Ez5k%_HlzF3FVtTH++i=Kij9V0
zd;Yl0OC1qKI_;PY3`J4QW|{5p`T5}?kaqi>VKp}!3yYu!YePAd$_SsWmb(_Qp1gbL
zz_mJ3Fz<)Wa6qBg9Xpw;!k!Wx9IUCKVR3$BLe9?4zCBtbkP+zZjYX&FAsI`r;}f4M
zPWI5#%L^>2l$mt;a(q_SdRGj4=pOra6@&FkPoi;7Hc$LUW(bqf(8QqW76yMKen(eV
zP1fqg(YTrQ&g!Gh(X?OzCM}UIf++|x-pq?L`Z$ffy**Wy&ka6FNm}pTeOA59*Jy0^
z$4%=R8yUf4w-q5k(TwohhP8otml*W(legzv)O~{E;o)IrWu4P1((Tlq=4z^+cDq*5
z1;tB8u6zN}^XKtoxX!10I@lbhjxew?v1~S(Z!24={5;vo0z)n?oRRFr!^wKAW2>Q^
ziuWDQkaN-V^G8UPyxR6EE)_~XcJ*224(-f_FdSW;AFIv4SDD`@<gkDKjnnoyCJqjp
z!=ARyP-myf@z!MR)mhdIu?KnR`cRHhPrQ4DUv%f$;rjCIY;PjpV}}K-8sYSHb--kF
zAY`{~fOjr0E_@0*3z0%8<%X*77>#`Yz86o6vlY(~{o^u0qWYPKbq_61y~-LqwE@0u
z$q4CjrvKHl+to4{mFNJtVE@i6`h8!^#$?%-c~%v+n>c7_dm|bK_Vy>?bXo%nl$2Qu
z3)&7GE68)0G^!IdE>0@J3whesqhNa*D@6|RU=KPvI|oq&e)SZZjF+%@LuuWi;Ab}}
zz3qB`T42xqNFbje*XcsYcPb@{MFzbtZzuw@rns|nl_a-*&-25vjOaGtev!;(SL0^2
zNO&ybL)lK_9a__b2S-P;+uhjo+TPTax98>W1V?SIu9n-Z55;_8bvnwHoKE2NLMIg!
zoITncx3#rZbKY$_Tp!jM!NSB$eNC&??02*K$;lg2)#F1$M+d9@LN3zMu(Awg`ls6G
zz;pA|xQokGDs7vZn!ZB!@?fEDD=RCWNVq;mPDdL-CE8VAp)3#4!3bDRoY8P_aMDAF
zg};*>(g(1+UX)2*frtD>F&Yr9AQEuMySx-T8adgW*CCUub-e`cokdw?yBT>_lkjd9
z*xS{%1p!8KNy?OBNLbkMR^>KD*4PgNsVAj)5DV0QUOh1A&FcL}uM^k=h!cWeKT2u@
z=re<v3|FThkeGuLeZ#1PCh_|NM=*S=+BT4@GCn-~?DsoZ-Stmi^1oKAiB>tbTw0EE
z!$i=ydvobW$aBv{W}bte^*neUkMMr-;JvzC_50um67&2Kd=9H6kxC*`78VvVvh<a9
zx4l+U7D4lG;Bz${$hm5@Q)4B;a@7e<T&#{Irue9S(1k_2{P=kF)?ngZ{;siNBUX8E
zQkn$l_1SQ3m0^{J-;$~cHvU)^H*#>;bp64>L7z?=n*-yUsfCrPeCt(unoTYdQPHVF
zo%Y8DY)njt{bJlrJZxZv1XZ)4i^dz7CjwYznH$V|Us%ZWlI!Z~YLz#Ny1CZ-QraK0
zI-ehHW+|0YBlq!u`Fw^tIvCS{MX}zN>A(CUzyuuL5PS(=e@Dj#c@CKot+luvvEknS
zzKs&2F()(galy0;Md;oF{Y2O2&v)^XZ+7J;BR&Ss#^$EXKtn^rP`1L$FFNht4}ply
z!ZOm>n55a{Q)#y?i@9@j6wBq}IG8D8Vqzi?fafq%2Up(!;y)RgATfyFtg*4Ng~=ei
z@2$7yD3xMDsvlde4UkVFm02#ft>@KJx3@cjNS(rlMM$Vtr2CPOkdTIk=Fltj^71lG
zB!Vau*tyHuK~U&$mRxsJQ;Ng>Qrm~^Fg4fR!^6-{D=VuIe-O^ayS(}YMk!d|g1qug
z7FMjIqhqtulOUNQG?*n9N2_u7_qu3CBKwH1eai%SqK%|8j(fVfnS_;9p^c7??(Vn3
zPf%#%Og&WR4sbZC`V`<NeIFgl-WZRVU!ENnehQbVb#=kMBXk;NR+Ax<^?k!9%?|>S
z8Kdb03%R$#T+rN1lkc+!YJ44!pHY`pRHUY+3Mt%oUe75_W@T+EudFQ9?|t_=WLD$t
z+q%9K;iqH3h_3WzKMFK@p|~AS+NDIt#8Bn9oy>X(Q6?T0XtgAB=yrr}O;)zYwVz1H
zet}xw*!xU;F7(BV7eb3*VmUcEGLS$5&P<q6iE+8(p<&AFRFPN?1%x>mVwo+M4<fFV
zq-5-Zx~i(`YF|o!Umt8VJ)Nwn9;{(2s-((?5R#8I7kk1@^<b*xkK2Qb^txr+q|!w3
zIURpUe&<O<yd--0Qb+iwz#xx~F<Fqs90(Q3;yH@NAhc747P|g6l=~1~US1gpNJ{C_
zZ9Utea;g&eFt|ubWJJV?9|j?kv9uZuPTSR_UzR$fe7PE}_b?zjCnSgTcv8A@+42c2
zWJXz1X#rb8u*XGnGqa6ep<kR1R$jDIE=DluwXKLDsJUI*0P_l?Qo_W=rNnq+I3Q>v
z-XQ~b39KJh^Uzjk5rzBW)9=*@kB1h<&<{!FvRx^hz)vu+fynsA^}@c9Ay^O=7iU(6
zuKfiVsYKh8Cr^B!VB0@Ny<t*5p~1(aRj+IK@;J0D5TK$}WU%StWzrNWwJd@+5##G4
zHnP>81YQv7(9qC+(vXs-=~;_k$;r~&ZA~O6Cj-N-+3#X+y9-tttlv{R)9IR=@e1iT
zrW5wtQ`MSkDMF!O)*fW={&=jG*H@R7wwq@&P!wv_DtQb@Qexu$o_i4FY?Vq`OxFr%
zoZD6T>mrb1sFX?uvK7!Ft2eRAVHR02jB@*MOnm-hYz-6%A7(2MNPrJ=Skp5#=hMY7
z2nqpA2u_e`>v-BNj>RHTGWLy0<Qf0?+F+JgEJp^c_WH^Rj&JhmJ41l5AV@f`hR8zo
zx}t+>{r&u;r;()X2IiW4!N4f4i7~U(>|Q#FiG$3n1YqHP9UZ-{SWYGUinbw&D2gZi
zwP$OTi{&qKtk(uUyqN(DBmDif1clDxg9i`V8u~oWFw+j?9C)Q<iF0P2<KKi=njRj7
zsn*dH<?=oL7{h_=rg{n?^KW2Q$id4SH(?kY^pi)*mX@2_?dsyR`6_l(CQH^L|0cwv
zLRy7a)EyliBj{!MUER#ZSLt=)esB)ws#H4clX#*1Ssn@Gu@N@xd4+gUs%fs%R}g?Y
zW)>EtK}bO+{^R|hkxfVP6BLSc!MQ&Q3C-!x{T^NFlDfEf4_GJCkCBYVBeVe|e?B0I
z44n8or5rUS<@&|xJ_tn>wHY3N^8WjMWYMTMIUtq@ogNOW$#Gu<{Og^zR7WSL^Yim#
zvW~30nzLqHGxN6=|4bMS-NGE$1T{5vL13XqBfPM1yKorw?9CQ~7u^F0NN;&p_Hn^r
zeiYm!e^JJQ1DkwT^OacwyFW9N>$;lf&BZrCimzb$@Zp2+Gn(c<i@3*ue{p#{UG1W3
zlOx@f9-?;1kN5ib`tG5z3I1MDiyjz$M|*n|heMmkqagkmfPdTT8~!=NE0D5QS65|+
z79A18jjL93T(`rhj4~>b3S*uTE`mi`3mioI?fDC*?E2d?7o(-v%~7&uHJQsZPzXv#
zN5|SgMt?ldJu$KFNG7AU)3~Isf4mb01*Rue0&;}=+Ozh5Qt-{3$sxo+K|u)#33wm>
z<ep@_jnSfI0B0+#SLc?O(~cG1{ms;ocIE&d78De;f2Tr+Bqb%WMzWk(E`fv_<`ccJ
zu#iX-9T5R=jG)LU&tb(1Kum2g@{p*gcDSbcFsf41NqS~xx2A^7ku*t69m<hPFbXt0
z7`R+ZLjw^hsn<OpD2<ZoRF#3&x;`QaPW4l@^0lm+d<etG<92nnwmvAcn5U4*JUFhs
zjsY_RZ-ET$4YhrWx0KCcSY@!(_ns57zrUYEsIxoQ^!4jkJfT0!{|>LpZd-M8%%-|3
z6UWAIEc*fE@r`|3e5>*RgZRe>z|YTrLD8ShVtq3E$3<QN1oDqZJo<ZhA}0i7{>tDg
zSOl95#UC5WgigC0Pv)aVz4(HiItkF4{nkYJ%{aj<?2*5(aF78se{#NE+s^a*nm6$0
z_v;4!{2tQZ-T8Z^e={&}=na<opTEAZD3@C<YHB2Z`a5NP!`-8!Es1b%?}xYRfo{D+
zBlh)nh&Z<Syk5xFPRF7mB7c_*kFgk!!+r;ZD-gbu($bbgrigCebH=t+v52YEmI2|(
zF%$}q<m2z(^v7H97$uLkCdZ47<ls_V&W{3BX42wZ3x`|v;#^V(^z`)XHb#1fhK9Pj
zsM**m$V^&M&30z$57!1m4h0pCFM7(Y^mJqxbUS`^c6PS5202>f*lv!2u>7yn$KGTa
zVBJ)TKkPT|wgeI$@9dOak`?B`&UWWp>{fbsj>Z?;!(yUA0T~QC^YcS=7YR$tlFOr5
z#AP<^h+#hl@q?a$AtWul!tVzG9^UILv@vC6<+-`JpF-T++&tz}RSsffmVZ~Pey%N3
zCM$L>J}fLuqZ*XkSedYwL`3|4SP8H+(HM$!a*Q?ulZ9VFoK8mrxKGzuSJu`ljE1RG
zF1pya5%Jl6kc~SpbzmDAt>>~S6lgZ7Y?2g#A`8Ud3c-X~WL1y?v4?L;4Iu)qP>xJ&
zjlBqNw;ENeJ+x9~2QAhWnUquw$h2{PuOPB)W}i9mcgN0FVbkul9RCIN(yeOYp?N7!
zTKYZ*cieBFGN5OdPZhz&-u>$NS<%+E#8N`~b;xc^VFy6Lpk7FVfjS-@5epRq0~Qj(
zU^tMj+Znl9jW}c_&JRbQ6@V?5tBj3`Dh@OGPMU#8z|7eO3Lvo<HktUz1x!#u!&*ML
zy4C<&2bC1(JkFHUHs>;hyUkrp%=3kXLBq3yRWLa8iPWkmEl^K<<NlJOf@Y`;&enLT
zSYG+biT%7k#OYu|u8<(kvRT%&&tAeIq6NA~G`lm~m^irFpBDATWDL|)86qD(JgFL3
z|MsVnJtMVz6Vfy>IjIxlIZSbgCn78yqQJ^WR<u5xcf7w(+4jat1-71;YO5j*F1HRI
z`|VJfIsF_P#q#FQ%Fiy_o6ykF+rwyRNVA-dw^sitNVA=y{9mII5K!|99s(W@4;Eu)
zYMSfHq6wfKzzVk(=llp?L{KnYJmIrtAsV#AbTV@Q1&2|8qRb*w0H7XFsR7=j&ix+r
z^XCIHMrXI4@$qwjXeHWsJrS`PbP@l0^UVHNjsXB~SZz1t)zs9;qd6QTGs1Q!zoL^S
zqY@I<f`JVT3;-l}etL?8`$j}T9V8Kd940hB-~u{EMowo31}R(=6u=r~Ek&iKDL2MT
z^TwiI<2`#uiriMaR!s*82*aNEdSLx2FWA{D0dV@81(CVpUXhT%uW17t5OOg$&!1B_
zHa6~0{g@@4(Fd{uxl}4y+(yVR0w;i*vt+WCwzp~0GbpQEFDpLBqb2MBczu1$f`TA9
zE@_t23i2GS#^=IJQWryl|0ulu4TiTN=l|rFNV^>X9RLAkrWgHMKK~BHVh0-^kIT6%
zh#-#B@u2RXK;-A^TSUFRCS#^wAQqr%#%Q)nf;>0Fz(#i7UoC5v=g8~hLeETd=(h~H
z@$-RCP<FemgY2ORth8dqoMlhz^)H}g_d{l5D`#4bQTM~7dfj##zyD7H^#Ay?|2CV6
zJ0SyR&-%Ls8ZNHUi1NF4?|O%KdIbpi1O%R0%(7Z7VG(RCZ_fZ9_MNH+?)A)N78u9;
zO@gVfpR=K*q->w4f>lSk6^Pk(6>o0DASoSWU6)!u&BS!zJoKGP<cqKbF6lUnJT}HU
zT0TI>4_OqPth7V9+4V=rm`Tvu%cH2!;vBkIsM~r$Qm|L4hA%$5O&)%J9Tt`23F9cJ
zfF9fR4Cd^uCsg|NYK|5T0o2TX;}ZXNLcdexf6an_q3qvq{5Q?~PDUNAt>u^3x4cKX
z1H^2i><K335+I&*QExxf$v#r3BMd??sD@}1f9Nk>qx|*C_!9gZ43Qz3qN%BA@Ah{9
zAA9OS<^c3%l0e|?o<%O%^n$gdQLV!IYxX^i_#jQj2L^V701K**Uzdr0Y+K(DSvFg~
zcjHlg{WVcUL`0Zhh|zHFABR*g(3`BVp^QVtyP(lDT9$50+@QpiZQ=stte%d}U$1qD
zz(3f@94b4v)#4D!X_v8Y_8t6l_7j-4wstg&1raAF{_R+j@sz6_i__B5&Pw8aecJ&9
z1JV8uK1GEvZO`m2wl~!PhPfkxVR^rsTP<Im2SnJErA8F)+vGk_0;ONEva!LZRsVz<
zuf%OJ_g+|7xRLb9?TUTJyNL^kmjQUJdn1mWZry)FTx1y(Omr@0c64T<+)C07&=R*>
zU{cF{GG>&=q~CL6j)orr^arJpvWf}-c7G5ODx`Q*9nh{RDJh|$p(D)pJ2L_J>|>*&
z)q8D}78Vv0q<=R$(oWnGU}Cs?@X{HQO;b}-{rwD#j0d1@a*KR>JK>M4|9hB8U5dx+
z1ax-$6WAW0=po!MZjXtyHU|h8umEZWnqOdmmy^7FbZdPJ6=b*b_r&LL*h`*mRg2~F
zRHOepCYE$g2&HVcJ)mo<+kXCh?;7|wHhvtc2jB{Dn3#WeN<0gB3V>cg4Lm$Nzyw*n
zCw=tj5rFG}tNzzMWJ346nVc+NQ&A1)D0%(GbLlb=z(LqzHcw5t0@%)OV+rsRC=s|n
zKL0b5U*|VyRfV6wIgSMTcKr)(w!|?s48#lb*}J!MS;9Zq-v?Dop(>CzC?_W-5dN#h
zY^aNh21G<CXlf>^UILal`3h9Dw+n`2y9271xyf$sYh(M0kTXELgelK^?A<6Ne=jf-
z@@BlRkDi2tKhG85>N#~qMa5PdkRpF8C%1bQhY<ugE(1`>8+L+(=<Mtq`Nl*sG0(vV
zI)^jN_s2qZXlVR>9O=)pB$D{wzI_Xb6QNLw#kDm@kST+=)WHRu4%c)}ax8Ag=tx=$
z7+L0N=l<#GT0ldAoLCJ6k`J|zC15Ss*w|p<0ii}tPJa6l8V*2hGh_p5nkpAX|Ca=<
z@?bm<R8vdq)=d2|SR*_<JRXvPB;?jd4_wH~+B#c2;T=e*gSad=l8p6=x?DJ*KFR>Z
zeJi89e|T72S~`sXW2)RL4UC~Jb0H!&BqS-E@(?H{T#g5~)-|IAP7(z#CirAHA4oL7
zBC3w3oPL8{0|NuFOYF`k`G`Y+_Xdo6=0G4WbFNlPK;bmCS`83*qG7OPO;YH4mDmAx
zx&HqC<<6*MP)kQeM6^nYsOUC$JlY&94v|SoPA)4e``Oe4%)Pt2TMU)a+4MHdWdhz@
z5GCe5JbYW-M}8T}tH7Zm!@*2aRu-e7*E~Yx?dm-Z4ewavd|fotCAl3BRsbtZwwNK6
zR$?+Pn;>o+d=|^;wC0a+h-YnMGYQH_NdkcA!E*Z+b^?B+prm9gGvq0y?1W26nbR3o
zD^HfCZAM0hV`pb9XJxRL7;&H>UzGDz$d@dT$JUwb4%cLa*Z24>dpfH#XRn-dTzQ=Y
z#9F}3h+LsX1!|SHIjUB`;A6bR7+$iDWLViXBFkDI(g#Cz+7#Q@`Ky;wMItm`L9(ij
zj#>!avWF*t(DgdU*w5kwFFK2*Qks4u4j2aW)ID9W^8X!MBeAcjzlR9x>FVNesr$G#
zoA2*foQsJH0CuKaDI4G*N@{Gd+g}m@VU}Nphxf)f9Px5M0T(w4YzaBxV6DZxr)0VD
zC_Q!rg%0)`WIPsHGBPq!QoVt634j}iaZ^`W$!HvPWWkN48^HhW-MtG08G7W<`uS9W
z(J~7yOH0f8qNeb9o=T97s+Sfd)|5~E+29`gx&#llX}!&3b%4_v*o~jPGDaV$F<A~u
zBKe7NzOW_ZGG1ch*tt1U&}ctjPp#sFW7}gt&ttIz6f(z7K0ZEdGY%HBDGqSPN2WR^
zT7Heg>>EjKGR}y{%eZhRV>b9Yewd3kV_+W-B7Fk`nWyaY1-e~5@fp|b#mkj=nJw3B
zbH|3T^sq>~b@zlS<K?BQpg)0-0;5(r<32k~Nn^pn!O1*-nrdol3RrE?8#pgL1FFe1
zC*n8U+_i4kS6f?KfR1CVDoI)fK_T+LYf?C6)NH_1GF>(JS#t&niSC^~b5;3aR7x<r
z?Rp02ktVL{rRmS>iQTgm`VX8)weY}$dGXR!lfKVhZS;_9AN(wC)-MpP9{#vfu`-Y$
zl?9moB-ZWWGZ!E_(Cjx!E}jp?T06D6t*QS@tfIuW1<*#!4z<*NS95aQ*dsaUr2rc%
z>jMa6y+-at9<n`Mds87yxfWTkN@924`pcCaumoIg*Q#H^p`kLbk%jaIui0GAeiRoM
zH+&4Gkhv)tTq}60Z7B%I!+007vs9nEoSj3txVWaB_x;ZT(*Dt?e#alaTD2dpmuCS1
z0f~NyEk0%N4(aWqT0l*%si~1<0cl5a8qiobk`+(a6v&|<_JHjC_|c;$X|M-C9T){*
z9<X1nYNKOg2thY$l{7dcG%bHmRb#rwWpD}95>G+FphR|^dEER9dwhCYO_rFcZwR>W
zZ7{y}J?8`Z6?x+2P9qAR8ZSUb@uDCqr7p*7gOhYdGKEM)G1@J6;iv2YM1@WwlCD}^
zaZ(8cUz;z(d1}KhR|sFiXw<H1&l91a2TBH%jI<GPv9T4B_>p?e_m?{PCT`1X9x3F%
zx#E`hh|gd>0b&i`v;UTy!DV&%9{xQzeksqFgoMP=!J&2a0N@%R$bi}v6xd%@;$0Kr
z{!6B9i>L*9NSy0ghAN=(+|0ng6FA%3#7M;&9vf?_t806-p-}BO`*tmmkUQWu+iF07
ze+!3FJVi%OlB+cte`9e|eB9=WF;};IrBYK6m7_ob0kAdDPo$)zz(`gX76y_(pidG0
zP3WEMWO^mY_-rea<YU<iMgPKr&ygd4Ew6-xga87T-D;@=gnUrVoQU=RYp&#w+xY}X
zMNvZQ-7e+{0bX#sKA#43X^!jDH@E9jf)A==P{Mr&d3&@_=fML?N=kAkN5`v+{qEaw
zN806s2nGHDodXC<;$jbEGNpHqk6Zr2Yk&gx7z^_APxjhr4%g7p(6pt~&$HZaPcUZn
zPwrQ-v@BGujy6kC-DeK?;}G#**~zLwhXM#5C92hasDQ}<iRd<Y7GGWgr?Uq|_RE(q
z1M%5yFVAeSae=PY5459KuRM-#r@D{4w6t`zHYmJ(u02*_Vr1WTTayp0Ew6OPa-OfI
z#qkRW<N+3Zp*6%UlK+nh_x!}afvA30BD17+w^v32-3ItpR{MaK`iroUZy^25%nT5R
zZrwRD6$)xgJ3Bkz?{1(<mXz#!5_4Pp8BW#vJf)1}{4b_yNBiQ1#xnXX%lK;><@RKT
zef8a3=N%x&_VpElB>IxykJkBLrKI5?5394Tt}bY}aCdiSG937JBWB!=_~Xy|(kVw_
zPj>)R+D_G)p_rJM<>h~vN3a*EGx2e8)5#w~21Y@_gN+u;Y6bsy`Fb$$VEeUOP}$SZ
z#Xt+WoxttT=Hi#|e-suLx?zC+3X6Quqj1D_i+2Yyp|i8IoL3aLJMJ9t7th`f@Bh{>
zzFX>ia(QR?TyC7!b_HS}Z~wDDU@*?D_^-{3Xw+Lraz;q^B~uUy|FPj4(pHTE!pV`<
zp<pwB1*SpKVh>8U8!`F!{Qj?0mw#RNHdg%4e_{Lx(Ye<KD=wtg)AMVcs(%3S-&6-m
z&_bvm9dcZ#&FKqX&zblFFxzGIWYTz!VqkoMH&MuqmEcE7Nkq{eAToj^rtM{PIrOa|
z+-ECEpKPIi7Wb8kR~hWKl<^o#pwv$%xx$M(u>^V=5*g=#C)X#k=xb>i`Gs_3KfTUX
zZh}DK8f|*J614n#mCtiumQ$bM;bpHV8q`N>Ygy-88|3=_%Q*fn<bU#)nq=g*q=ByT
zT)oNC8)PCH%YpsJ427+80V56k4v8I$fM83xV9yBY-pw%qSakpN4gebxHU%1Bc23T4
zwnEb8dLn)WAJ+Vb-rby9D;11?PLsc|G*#^c97IDy<JKNY!f%2g_r*>l==ixU!t@PU
zLD>lK`o9?PlRf2h4E+O&6#g||oy&@>HGo9r<>f*B+~|n}JpFpx?Hc#iYp{YCimFAq
z&Q@MF>VUp0ptr~-yn7%Z&=MYwRqP66Q{bwS(}{e(Ab>wjDF+h%gZuXZmhOmVl`2)0
z&C1RMVlarwoIrL9YB*RMGy-Y~aNpZ;Hw%KMtcZxwO##qoMW;-{$Y_1CGYgn>KxLN}
z6g+WB-UCIJBp1l_11Nw#21UTy9)QOcZn^B0rjr%mS~D~1lYu3z2s$!_mI{O3PhKXo
z9@eIc#nm|M>)!5Ryac{;+uhO8(VOOx#V|G4=(lgE4kSVvfC4nDKKw4}05lAGoi;Lf
zm2%6b!9f|9Dts36x4KW+1|LRNAoKI_wHk&NGQR5kki<870-A5+m6YrTTS4&K;!?tk
znYaPZvIpcxt$mrwLP9>OR(;1ull<7NJlK=#Or%;S^MJY@9XEHgU7P3>L>|hTekZEw
z*a>75I|8mwzXkQQu1Nr|w<JdOkre@DvChP-(}n>6Jhk@1Z<r3UssD1kSg{<^v7ULQ
zD8FTX$@cVuPB5Ux3nz*kjli)vPG*hB71GuWu~18{VWTaO-<gA<GHUDWW>Zy_O_5bq
zRnJz`n;L+s*Rr#x3z8L9T9+uGO+d{oxiVdQ4Iy-nNB(&~uRl$UvIyveVa1!1m2#-C
z2GAvNIFzwG19)z#EEzpO3{d)G>dY^GSeu(GT$KPiG?|<Hj`8kJdY|1&hXlH&j|+v7
zK`0R2h9>~;;?GRr7Z5OHBnqF*kB@@0)%`7Z68sJ3^<CA9-FAZ(;A!FE*g%ImEGrIs
z;2i*ne+k)bhFU828*Kr8<oAb=@hh>+ILgeX0ot9Lp9d${=q-xVfcUc!+%!%ZiK?;y
zAYZ248vI$a(-#M2s@6?yqz%X>FO)!?4|uMG(oYw8)z#H(09T0UYzzT)#o;i0NG@HR
zXFy8t0Z)@G?|q<qTSRkI4k`!;2so61Ir>|u{K&LBTv_z;+a;qiYqyce_5H)6yJUoX
z|4yRCTOB0{<ZB}hjf5%kGti5KOvt5-0GxB}IW++HfRlMftt@^mRE|f1jKj5qhz|;*
z5sydc((TEt2ESbz1nV8vWk4%~6tAG5P!UQm{iC3O)nd+r9v2I%Ouv_iyd#R42K2mS
za5x^MgT@TR*3eNvtMG{Pe_nggm&RUZKHC85dMR>Y-o6Pfv*Fv+_4?@5t5-i;TGVQs
zPl2vrR<+9y2ye++&>nL`b1&}f>;Mp)lA7uYG=2b~+dk28a;j{M1LO=Sj)gq_K!Tjt
z*%PEmvGMnLCeO7&%|xn4VUVs?>nbU3N(|_O@7ac+frWk?bhH`ddgJ~Rr*HiG#P3Pk
z-Me=Id2w-h*#bJpV$<bv6p7iz$sbb#awJc++7VFnt{1xkhh~ef2SEE)1|h3u@>bd^
zpcK=`J+?LgkiBQKw^=%!8TEoPOvWFJu3@MA_WXaf=6}nN{!4Pa4K@G&`{jSGrrw&z
zzt-vhk(2#J?*Csh?thY%{~tK!f4=CkQ8!dF2$jF(9Du}LuRiG|`4~9>lw3gIfC4Gs
z8@aiqrPO5nmR8MwzYHiheL$`NWL0yvmIaRYBj5yIjQ*R}@3AffRR6PQVO3Qe<Z5N+
z^=)l!-%xHrT@DG5q(NF=%dV_E2Pe8!t@Me2j@R++1@FM<u&R>@fDM5vdq2OT=T+_b
zG$z4l)TTh`GBY}-P_DTDspr!SToYHhY~}=z9IBTLK~-1knnjeGAS138;Zv5MEi0Ql
ztehpALs22~hJRZ-tu>cnBF-|`XyAv|6S+S6e6jU-S3c#k%s?~Hhj|bQ`0=7bixQhm
znGLCO1R0772SDaDwx=~kar#z-O>JEA2iX|__zcLEbtbtiS*q!?9Dj@Z>s3RAqA_fM
z&w4qOBN<AR%TqBl*Ez-KR%b7hVOqwWC6hN;ra%gCpO0M($T%Qpr#;KEq~>!*gxNU{
zHYe!z64IKQs(RvEA0L=vkBnpCTW757g1(y-Cb?om86%}`BD?%l5kBRSOy?9FCd2y}
zMxSIl!Ky$IvEsV`tGy3$r^`n0>FFsT2*MA^_9PvqsvOvhqX6^O(%fuV1_f&!%2Vyh
zUS>E*fpl!z5`YgjcgP$9dP*uxC)tqf%mE_@h<_UMdmvUnyDwb`h!+R!Y}o$Zo;J+x
zU`4|AG3d@m{{Z--(9l6(7Rj$0LBC@h=(+soo&FD?Ti4fD7{vdO5GiM84tCJ#?wvtQ
zOsoT{M}W#U5dQY}i?xCXJwHDWwks+!GP8Q>5;W(6#)Hr)lgWy~9R?jmAx+IW&|$zd
zS`12k=fgocOH?L@=`aqsQnP95YIdLkeP<(vP`-G<lLtDk_OCCNxdC>g66*+_s(&dE
z0MuHs(ml{;Y#AHGdkW|c@Yo9>(B1|tA(GL6-hM|N0`UTcntnDEXuxjZ{EO3IPb<sI
z4Gy>FZUh8C-GkC~Y<;XXgmkg)0X%%o<<Yp!+Cc9}jvBbq;r_nSXyGDIHNYwofCgrF
zQ6OXI7>^c;ii+mu=2jenV&B5r8jsC7JulA;v`h(`Laq<<+(OBv_}Sz$B#DdnfJOjd
zm)US`42O73Obp0jUNpqSd-Ro{O`y&0CM9ooGEWc4LmqG)hUx*H9Q6KWR6}oi@*EG>
zzIb^Vb@Q;=YR1i$7VDGT^cezX1Q>9N^X?oUsJuyCo@qYvv3|k8Af+0$%0}iTDar(z
zGA@dC#v|^(#}*6iLDt3O=P#~6*45JkppbGA05iZ{gLX<JO`-SNSS+CR@aE0{A~9ap
z48!6Zb#AD6-J}}aIl!L`Zrlhs!s7rK!r$LN`pxm?xE;7tA=s6!leUkBHS7i*wQ*k_
zgdmbJ?%yA<6897a(73C!Q$`o4225Zf(I$B~A&2oWtJS>PD+D@XBCvic`ReYHUG425
zT4~M~;!fqxr-e``bawdF_k&l&#MM`4>-6zG$<UjnwX-`uK1M-8$^qnywfHHH1C<VN
z1HdhBPE}W0n*((say2X#3Yawuz!K?!#bvf$5zk>`vE3M{t9$EAP96j;2FNExSzKH^
zQ#u1+ZvDI{V5NZjlaP?WSr7Aj2?Q6=K~uo}xYl0T3E6eG=^0=TfRHB1w!XPp?Rx1*
zWkX9vl>;_|(0PXkuouA2!FWTBlE<uWdPgb@MZ#${dg6J2+69Uk0C<KLOuQ670@eZa
z>LtT~eEMQJdSk=b|B#N7QWzcsblQxuJ6rC7_G&y?EbOO3r!jU@y^;_eR1Ls|UL{%s
zLU0<ifl%3)0}ke<kyhcpHMqm0D2oK2!)6Tz#JtQ4&>aD8ig4LvEdy?if)cpD+zkcR
zFCro`|G&6<3#cl$E^JgqMWj?p5JW*5DWyYE8l^*!l5Xh^1py@l=>`cwx{;QY?(Xhx
z$vd~f`TjA^fB)~+xOdz$93F=3{qA?KcdfbReCG2!tIj8=P`NilnMMUBe$~O)$u6tU
zLM%Bf(Ot=DvgeRzfvmycz9j;>22g2{`KoOiECd-*qDWw4a(FtgkG}<)aexy|dID!?
zNC<^KWn&ZSEucK(-_>ufgjgfm0u3r}S~41#uCUk?fpDYU7QHe(od<ZOk7X7VYcS`D
zVO22cf|MXy4l|KF(@O;Esnw;W!v>0&QdnjH){?{r+gVRkU3$yI#RXu=Ol4)Ilao^;
z_X9Ex2pR!%iV(jbPfffdDDhP_PgfGA{#+XvRP+Eo<8R;B1+5lK&VA_43n#AYyn59E
zJdRhsW`H`rr=n_TZjOYg0<X;5o2Yz*>a1>)D2Bj|d!}q)Sl22}RRWX);#o{5YoOxP
zTZ=0*lY*I936wKpEG)nd<KNx}iuMA}0??ScPXR59AQZ8cm5!_Eo`jH7OeEuu3BauC
z&r?!CT-ARhP5=JWG?}oK)h_fS@((zJAYzJ$*zGPatgMVAisF?W!1u7nn}sc-ZL|^8
zLt1Xd%y{%F4`EF^k^6#F%2raZBhC$;dr{HPBJ(+mB_N(Bsw9huirVkW0g5bFY@tVZ
z@7*;ZPxeEh4i$Dvg$IeSm>6z?-QKFvJg}=rL4*;rNm`Ke9iHDQDY7|KZaFz@xNt8(
z3<PqFx2LC>wY9u+noe^lBP<U{t2-?v-gtR?=jG-GG|H>kl0oUPH^QyZkSrOGOZx(v
zbC(ppkB+JwSGGVut*x<9<ZB>E5t1?LYHHAKGJJ=QRwI0~zR@;TtS48fu47!adNKs<
zXY4g<xireS1@@cQufvRovV3~#hwqQtTEO9KZEay{yly=0O_r?g>B$EoEZ_;^A|h8&
zP^e^60ymr3K+ObIb*x(9mZS{wjCfUbH6|fj76M1kQq0%y<Un@2_N3n2O5&=~VBP}o
zd9cf2wlgll=oVvqNlFr(L1Z0|hD)q7k1|cZa?CF;r=3Xf%gM=wF=+wZ|FxNCp;#Bz
zbx$}nQ-oRo$7G{~QHri`1ucskw9)r_;GxYV4OtQm=j!+d2Az*5z-LK$v`L4tIAWm)
zWI<?h@QffhL#fP3;xWPDuPcs&U13)QE@~)jnxCJz0mc2isQWuRjFYXH)4)yW6j+@0
zLf_y-$<hQON%1KSGt_?K))IS8l8NLjEPvlli4XlSH#fK3+%4E!cL=eg6A)C3k&2!5
z_4)p6Qj8l<u5`G&o}TPgAR!?^r!uK|ks}mjUogxN=XWpx9VS2J1ZrFy8=Lcq!%uTJ
z`Ck*qj3XJZ3|bV>>($dMf~_enjrMvjWIKS(EEcwKjZ_jl?eogv;NS!W23kzlUq`%4
zYLc+je$N=O?FCV5$%V?RI_^MuSiHzxixBQ<0G|tbHU5u6-rhyr7}e(@0CIrr3WaV{
z#vK^be_t!Veo>RKS)q=(g#T~qNjBzx)RTW4q8V8pvj10>inWTO=2M_qpOc@*d`Bj<
z&K$7ey;8%OTWD-PG|*L00*6SHZLi@uh$P%=3Qk+#{lxEbXMKYA2HKjFRq5fjg`M-k
z3Q6{ZW@?5w`_qo|mL><Nkp%gRh-1f#_k35EP~q5d^EhPo5dYgxEAE%|OenaJU5%xG
zQ=QAZzKv(s+Ku^%&YLFht!td1L>uF_(4K!X8+SNm{s)Jz=<jidCx#B^D57O_XL>H<
z3kS{YY*9~U2#Mv{j;`}7-Z>a6@7eF2^1J+;0OeJ=LZr9)$ETe1kLbK}ylMJAE^1MA
z7KxCLO)_)fJvY>!h^8~I-}Jy8_1!Y^+mS5Dea0R?6F9=G|0RKz?e?vjc8tWpjY?V{
zI&4jQix~r{FL#Y}?-EN>OI@ap60jY^>N9QWL~V-wA*+OOz2|+qEQZU2s^pn0P4T-o
zL#-s<+$|1n{F3rqq1+H#BI@$vl#3|D=Y}HWwOAQQ=P@p*5bT?sW(x&A{jOpiy5T@p
z@Az&vzAfF#W<wmuj6b4taFJSS{8RHf0a6tEipEF<53}m)HStq#rZ;9EQDjE;d@07>
z-U!W$Lu!fa9?7gICKj(KRPy-QenBDBAYDf5)wrhB@<aJ&D4(8l;c10pR|clvyb-ZE
z*VmSz_min!bN=SI?icLB_|l|%0j;CteaAU3gq>FEa;^=@d(x;hSYYZ@dAxl<+Iq9d
z)rd_z40RsebuWR_kL;#|VZmxFHkMlMBTd}ncP!013w!Up@UL2<aaA}?jl^{S40~Tj
zbb_8htaW?zc;tA3$oMT|I&yS&L32m3i5uhEMe|;=0&Y|?B2w#;m|7qI4l~mrvRaRr
zf>~SU*f%I-!YXzzU$VbZot5rIRjB!x8#~MMLpUVaqM6}U<!;!bPh?kZ$K52gWn_IM
zJ1pB6Rl9r|Wa>8=oet-p-{ZS>+lSKa2J1Zr(;FF2wz0P?clH?qSk<!>M=ud3Xh+6r
z5)%sTJ27G~^!qTR;aZ=+{Q2gUA}q8#D&99&J+?05?v4L?={bDeVH9Krv>+QwjMN$B
z?{7uPzYU@wXW$p-<YG49vewn=+h$I^5o?!gU3Nv{u<Noqd3%xtmNA5eN~C+0l_b&v
zvL~@DlfN$ZOle@ZvQp>^uRL474_jKW^?76*xdKPEb{D;#wIADj`rq)}^C#(I-<ZA{
zLcV<I8K=B$wO%~yuNUIu6*!taZbjoU%YH$>dI60j!{G4y9l{5=w~NA*=+|yw`194>
z*Qqa27`8IpHJsTosv6EN;*J@UDEg9n7wecYS;oaTabIS!ESs+{q{ZPU72z%V>Z{8Z
z1GL>dUHuFKJs!RJTC~D*>QpUaTC`u$?^}^&xQdj^L^PF7+UccUsK-dLcdT@H&Y1iu
zg^GTsI~YIX<zjY+|82S+(jkJ%qQ2Ta{#<t^{DelkEDfrt8yHD1shzq_8!mY!HZ)zS
z*m8*23`{tRD-3W}8>s2o85;ja_cbm+zvziQ4;QkfD&Fu>N^M_BQ7ma-z6R@D{lmhc
z<4A|v-3W)UX1^s~6&h!hpe4TSl@lgC>MEApmSDq%*FlDm`d{k1hVm-xn2`*}`}inu
zy6fY`2^E6RaSpOgcxw+oipP;&ao8lCI4Uu;B_qBQK(m=7;-OHAbpidyHHK4jV_x1{
zH+9J(vV9&mW=?CO@30A?-QZJXxO|tZYT$-EWn%JYX4g6<y^z=X1$vQQyDe$AJa@)*
zA3U2Ya@Y9yv$tYFZAGG>TmPE<`KCfYN{Wrmkd$Nmd$Haqh3^hhD1A)(OkxzY9vfV?
zyXVy`JSY@?t}}+=bq`Zyh}c^p<R*74`L4b~y@}8d*D?rbM(I5K>fd2Yo9(0S-H!~)
zoDpY|Bdyc&6A*})`IP;$^kw^%zL<t1t%ZK(-V**6M{GLln_MMgEge%6(uOKbTO^#W
z;wbU%O=v8h?aO;l=6*bpW^}&Dgj0Rads)81UwOwShH>thTIjhRrs&G6WK+kl=qNU_
z9uMqTC*r4t9|q?|bfmDiCXPgE>o4Xk1tL9HHjJqjyn5NSWO~Ada5poRRS11<pc}Qh
z<8#Spl>oa3F3d@OvX<03^)k~^8+PpmC1f~XsvCyA$y??p(<UC7iD3$tcVVfg^s1!{
zPY-n#vE_X1C@9)_^C2>lrT@vfyGq7b`#&4;zc_Zi9e<><Uu6Gp+ZX%4SgtU${FetD
z;5U67Vp(6!Y{#UnzRT`#&i(w*E##K;l~I>SI01>ybGmbE&5uwDn@qHz(7wHq#A9o=
zR1RYdncZP9O2vwpZV*k(=3<h2$ZMo__k@Z&dL5JF=ic=&l$K(s?n+UMwtcuwn%KL|
zcCfPG<3WKQedkU2ISldE$L}iI&A&L-n=B}~J-_L9o1gPXUVL@*=X>12chZ||^w6*1
ze)PzskFoaWr(E-;FNwA!N@I}1bS_<WT7Kv1x4VR?Nax|aER>d~yZ=0b*vZzidh1b%
zDoy`>s-GQoUNnI^tFG)znrm6QvH{x7Uv)3Z9+7O&o7UFiboy*)ja|!izsWBY{!OGw
ze?oI!uVJc=>K=xp^Ht@UL<x$fhzHIdNd4ce=_IaKDNR23L@h&0*VHeBBQJ!hsUA$1
z$@}TX@%W|7Iy-R!me(%6Jl+nZn*6kWOLc!2|Mh{RvF4R#VV7p99`pN2IMV*^(@)3L
zBpb+gh9^+`hjB8sp79rxnrl)73ZlJz#kh7aie#y$@CBAu1iPgENIAc%$W_ikd1Z_N
zW8r8^Gz%vE;_9p!c@Y$o<Ij}EO!_fy&l=z5NOcFkOBHzUhf}5Vg^}Q8h_YO~Ve`tS
zmK2+5+19&qgSH-7tMKheuL0`gx&{K;B$UfRwh`m3Kh-JYn`+-HJ$gXIY5lAYr_Rf*
zd_=)KT_7Sp?+00F!{s>~1#*nAkrMO9`3n!^15{qTd^UP$uz6xwM{t+RUj2#jP`_@~
zw{Wr4MwWzk)e~`ZyqBmV&3Z<vE5DjeWHEFhC%qfaWifaxW~gLbLe292W+untOZnQj
zSFRpN`^2rs#wy9ZWG{M<*ZX~$O^6qrcS(v$1ig2dDlkXo+cmj&QqL1OS^8S?6cVU;
zvOL;Zx#m0{U%b_2?%=psp3Je&xQBvTWzg()F3}17i9Npn7Vpr+oN9Gp=tk0up{fQX
z6zX*T$QPLRgkSrAKbKN`9JH;CU4ex9>dN!ut*yiM^dlBrVl|g@58`WBW9enj<+oRk
z)%6{h+B~1&ZnphQfopB1m^bn^ObqpkvHJjnRRv}G1)Tf6#0TG|sT?e5G^jUy)dsJ{
zWgpyk6OvCM++GoRCbb)VZ^jg@J@pld#EVNKTltol!cncf&iZInpVnkh2-!&r_1fIX
z)z(9ZXga0Bvgigc`oChO4oAxmyeqI$hgNCQkBcljCeOqdCoeT=bD~u7EV1*Q%&h$r
z>uC3vw@HuVTKHb(a7%p)9jLyD{WeZa2eW;7)UGF7$T-8)xW(`@bFt7)sAO)yNPe-1
zy!2H+^9*928Yxt%344BZ{CB<j&)CbV&9KYm6#Eb3KgZ)_anG>w@tQ>IUsz_$vFCc1
z&9{s(wRMYu<5gbRRz028Xf)CZt4h$rsv57ZS4yqaO&Q4B?z3g{gKt6uJBte9MkDPg
ztyWxKCCp*5$7o|c2>kfua`F4kCVBQF;fNR%Hu61Zta$@6QjZFcv52Bt=9*|@alW)<
zC3J)LGmmY1kOes7zV|ytm^|{l#qO0zkfxD1bURD1gy8mL%5J^MyC_|>n*zuNBOw)>
zczXNMY>#{D<I`U2UMyR`I;xp!FF8CP{jQgK=e_;z-nhFk0VTGszV%l$-;3(@@yv6U
z=`Bu%2j-l3@*5(8p9r{DUZ+)aNo+E4Q@nhB+f9{P>QLH<sP2fq`;|iU;5V%Wy2zZ)
z@1LuZC39W`XopeUS9gDzJb80%C$Il;AT?WN%}=A=i9iDJLh8M>+NTLAS?g-SinI7c
zW%be>b_5gN=%>x7yr^gldMS<GftH`shOI5SuGEW_xaWtX(<Wx)XE(4h4NJ5|ZaK*!
zMUXT3bf*P!G9C=+KjLku2Y1j|{1&q3LaR>u1v92f{o=?s(tKJ~!R%k4EB$69dg}7*
zc*`i4O|Crh>;7kk_N`4K8%>65Oms@^t2kbGr~Vri4D#B#x6J$Q1TJ0gHXdVcnP1qr
zz@zP`G5XcXIxqt}t=>ugQ%cY|EzSgGEgrEK_w>*?)f{e@>{or@Vh($UoK<^ka&+g*
zltfZ;?zn(p+s~92D2IY5hH8&qd-LO-bnFt}I?el6%T6|?e`=-wr=4aVFfji%4*Op^
z&7fF3+dDno;M?W9p#XnAyr3aT4<4+=m67<YijKpVQofnkEte83#V=|g{L)^uqbH+-
z1V<HH)<7LK`<)2M-5Z}0PS#$pE=2x({-mqonKx%!U(t`EEz`MC+qUE8ImVc<hGtjf
z|DnQ+J^&GPBz-I3bKh8@@VxnQczD<jPX~(BVT9R?_x+QTlVKFZPb>iAp`xNLK{p#p
z_qsA(@So}T4-K^eo*m$qw)XZqnwlmaP``fNKl(Wr19-geD1uKhWBE<W8t3-y+r-3e
zMNr^MeJZkAQ3N$#T3Q+ahtamPVIXM>92S^00IC_3)~&6rUby$En~Z_s1|WYsG)w{a
z$Hv9g9|VFIP@ZF#7X!q3fYu;+)Y9C%y|?$sdX-QIZc>PU7T`P3ioBGR{PYY3@QS6`
zS#lbVAvA)A^Pq3Pj%do?eHX>!sHg{=Spi<&x7rT|#=)Yb(N099(Sz6t(IUXY!ZJeW
zQxCTnNbiI6&dkE%=;Uw>G+@M_;ybOYtu^I<vEr(X<hJi^Zl+~rl`q+VFAA_>IA-?t
zQr~NUgBuTkF9GCf?5l1pKQn>Uv9biLDS@F%=>8843=r`+yeR4iek$;;QfXOf!OVa>
zojWfFIue}v;#O8H%qe%EU){Z57zR*blKh+K;Nl$@HeYdmpFylt&o%&V0AlU}vcFgu
z6aQR=rZz>e-Y9n@k7IFvl4-jA-YR{?QO*khbq4^e0e*rYTWEMVH6C*-FyL|liqtBy
z^}rz7+}l%43W7O6k}v2TD}cjShpT@y^el=aKSezCz~~Y6UZQd80B&$^<d3AdItorK
zLT)<*5^K<tk`L_Fs8j#2SU_H8CtWU~K6(0-8L##Sd;`F%EP?bfouq0($Yr%0+RVy)
zS^{l!5ZwjN%r+Q7=j_w$B&g-7+6cWT$PDylctLmn+{gJ=80G;hLBE}ZR1y9C%Y^wk
zITKt`8u+BrkAQp{g_){U?x0|Q1K~=tyVNIs$jw_cKRFp^`*!;}hDwRmb2@!@o-oC$
z<<O7zzTf@cbz2tc!#-F^1VKN4MWCZ)V89k&na0}nb(0NIS2bYfz|)-`4cG-&3I&fR
zM95nFH6X#0phJ)WWq<^zY4}-y5RPoGgr>@CfRKUhAOai|AD^9#jUxNKj67(yD{i^P
zKzqP=s9*(Fk*27<y?xkS>Xp~!KegfIfgw+!F9FnKXe65WL~m?tfOg{=G6Gu%kSNg+
z{tt-UCuO72QUxorAOFK=j#a3VTfq85s8$=In#Tio;R8NG2STB(rG*GEFT<^2^EPN3
z$(oOJcgw8oU37cmW8k8fEcY<R1E|7SG+OdN3QUuDQFrTuYj=G=Fb9QHnP<oEFf$bZ
zgRN5@`juZlWnTh9Te&Cv&d2*c4BfUYVa*kNO+#p^CS*>IlL8aeex6GF$&;$PvJ_-w
zmyo|+M_x$f6UP80z;I6w^R%8A49(N-h=bZnUi}*01L}?d*BHm$0kGsKT0&z1v@b(C
z3D``e@QDc7u?8yh<Q>N#xPw2p@%ta)H-G;zww!1<TSIWp*&jgV5XY(ddZ7dAN`$^z
zw;tmo@uSho{BrT?%9(PTM-SSEhyBY%(Ap6Yd|Rh;UWG>_T$v!v6|lL_&g}*VZ0b#Y
zA3+Y%CI8_($BU%Z2eQM9MRv>`V^X#0PAeSfXuHN-Cga}&)PaL-3*=uI2N)(Y@%G9=
z%Ud?xAJz@Hl$^bmE3BdpaGXyOQ*f(E&bJ;YJ?jZ7G$*qWuYu|0>iTQGxwvvr<@>=B
zpPQZ4_mp+QSD*+-zb7m#3|h+Lq4+=x344lgY;0^@Vn(o{nEtF4;V%zZW%RIFlt58&
z;G5r9QIEv<l9*^~XQxkRwKnPqqnpD7Z6lyS7aH`nrm`UnBkBN1hD$)hc>m|mfuSJ?
zdP6|QM@2)ku(CohPP{xl1C7YVBIj3DR@T-|ft5>B5upT5C?GZ?sD%(5zKed1<Q9wJ
z{L&JQHG<0aaIXQx_*$6f&!0z_S!HW@bk0vrJun)4sAp|rvJbNP<0krw?(~;^VCMqj
zlB*`*EQVV;O$iixCM4N!r6v<DN?M%>!e-dOV}s@)jeQApil7&gz){rbe6mGMn1EPL
zH%KiIq+Ue0^tvzxPF?qDpmS?tZjQyP=?I-H;OfeFKVW2p@8*qd106lkIg~pdHr^Bc
zc0CKhY8lW5MG{zd$Tfx#fbG}G{7*G#;A)-b0QCl^QnW}M6k{Y@pgo~nf^7%7y#AIG
zuj~EpURZ@zm}VKYgU#lTj*ftj3r(|z%bg(1$!LDzHVZ1krUPgtJbV5e(NS4~^_v4Q
zzq3{uaB@02I;2EF>HNa`@1q{X@MII2LbN7GzL+vFGt26Q6!io1J2TCT3Kf^uF9@P@
zqve!WuH&H<@U=fLzQTaUbS?xW2#}4n{rxXyc<@<`m4Kyur1}Y|p~BM@M2jUkrqE>w
zrdQ2KYwDII&uTn??|!itqPV-88$TZ(B?ippMjv^WD<~)fpz=wpAC}0HCg2zTGq{9?
z(2T9CgxjL=f$-)Z7G75gxOE~oCFMMLDarMM-wT^gE1|-5PVi{Bksjpj?d`#C1JZGb
zTrIVFU0<m7J}f`a>m2>CS(V0?{Okr9)6x;l^1vNXLzTfa%h?1vbT4#yPL!d?1-tXR
zC$+p36cjCz7u%sb0Mq*l5XUz>I`@GNIC<ySV!!#>6`1o?Z#PW?<KjqERmXl<WWgo{
z4X1LkY#3~7D=Ru%ih5J6FJDSUs6ayMYZg=gZZ$DPg}Uf?jH4eIW!Il~?O#FGs%>k7
zzOk~%6nI#f`Vm0r67g>5MOfmvV-lCcH33Tp@*>w4SA<X5vtwL3rEeeuSt~y@hO7qD
z^HDO@jg}}*`7e4L#B@PVJ>1=a4G#L=#~I1P6fjpdH#YD&tqh?F57Z3GLO$ERwbw-6
zK0YhmUjqp3z!(C0>AqGH0tMiybF$qfj_7}Io&Wwj^4U&e6!%@-Lm`N(;NMm4<PH1y
z;SFeo!5WBoT8`+ElM(?m@iWvmV)v^lpye!2?yVj$yHe`Qb0G9yvX|Ac|4y*p#p}MB
zPgr%^V<F9w)8jImt_R{NseB(pN<wVxV{K<9uSoEv^Gl+nZK$sg<+3#cX$e>jP^Zk!
z&56e6X=-az-vY%I73&V~hWrL)3{{ZFhvsFl;UcF6wxKm}^*bru8(WzLm7sYjCFQS8
z0qGaEfYy}wqhsvf!F?Kq<7RJSL{wB+d5AIn&bUJm*`KJ0gD`!m!$oje9%>_39IlpJ
z-H$eEK@Ivva#~2Oi6u<yyOVHXlsN5W#YVMpCp3H>Ffp-3{EezLmBl+xpQsGR`Wj~{
zJSlcop1X_T`39UCKx8**_|I&C1#|@!6&i9J_Pewcf9|WMpAZcL;{=j1NY5z5BI{9V
ze~&vOH8M1DW+x~8V`3^0OxB)MgK|$S*w}tgkV=H3*wNwGsa!9#fZ!Rlvj2VO4dk_v
z5{iII=os$OeXOXc*nEHvyWp>Iu8xK<4CA?TN9|Zuq1>S;KcCt8@!LPYyN6-5I;^lY
z<tTzMh65G{=$(J<al)54`|oTA%#lv+6h6G7oIC|Ckd;77d|2U-pb0r?Bdg5WOWoLS
zcCdDibcv6NuDzFMS4B|iVDq^=|DVGQ<i6wHml``|^^$D=yh2SPUgv2E+UkgN&I#^-
zb9L;s6$EcT`jiIrdyyi)!+}4={5Yrir2LUzM#W525vL_#sohQrc$*CG15Fnb6H`I>
z`SY1~f5$xt!G_9Fk&)>JV~v(yPA|W{(g#U)$`78NcEAl~`~2ro3e1eZIgb<<A75Qn
zrD8MT239s^PyP;gBcjun31PF+<r33o6Z+J<$=J6Pj~<;?J!$}QOH=s<gklrK+HK)&
zS&D-wZe?mZ@#{t8<aK%eMPWrgw{Z>3Q4J6qM!cm7$e4j%Vs^-$1UWz$r*+1t%t1@z
zATFdLFe@D9qO74mhXX!S0hJKKOrIdASaA91_pYRMnFn`PvrdAqW63SCQ5Kr_swx1f
z6LwSfh(E(EUQ^k383R^+-yVtJgu^dYfMd|jSl5fl_%DBJJy3opD?oO;0cqiH!r48x
zUpW=xk(u}*(x~6k01Z?~Wu68!5*k9dCuHmf<4=S+AmX_mn6CqGrbFB|4*rvk0_FTK
zC4pnmd;>4eJAywqZ346Q_wAqk^gpI&yl3gz9RUf_U!UiQgUskc@b_;2zw{#thWf?d
z2`7jpuIN9E%+>;iL==bhO206DQ!H->Vr<^K-sJUh=PCfj_>a*d0ohZrneuN3`Gq+J
zf4F_{MiXS^lt)W_Ha11{JU;HfX@_v>zx+cO2?WKl3LWuzzy7h37it<rr~)qLby(2w
zDHjo=^yfMKdBd-}`X#}DQ#E;gaNs7u;Wj7NS+-&qZel+#KQHHxM-$w{bhZ}w)8Wc}
z42seZJXDvj9UHE?`kVMin%v>{BN(3PKC5S^*Y#uH7eoEO1wemq2=4FSK6E3m+XiIj
z3U*DfL4zjjg$f2a1;tWkCX||R0rhvLy=?ZohPUjk1^4{SR0)R(2KxKq8wQ8C5<6~x
zN_RR)(a2XrMHPkb>dl~tytBhP`)gQdmSPQdt~qYYjtOK!$WgzD7$Eq=llZCpb7^^*
z*~}k@BY2~RS4c?T!Q}FA5roY@JjVV>@aJ%SB`JA5^RHwh_WBo~7b6H?lJRvw3ppbV
z>l$9~o*sir=kmpiz*zeI#o@E$OrTCU0;(}0l7c(p^`SJCvIC&r(}F~G$`~L8Fge2~
zCe9Q}YCsfF@<7IHALtRKFWb&4yn=`)ziws+djjDi2uwNfLfPHjC6Q?`0y#%yR1`o2
z>re;fzaBg3SewzCPEjh#QEGyc?{Rk?gm(D>S;!3_u?0?z8@R^7qJv1m3`85)?0-9F
z!NX6huV1F6pb%6*i1I^1<Qt5DHbLenAcruh>`epx1#Y796;9`qQtey_bWW$od*K@^
zu+3bE{OAg!5#$k?hG0(s?}tV{gOFKY1}aUEUABN0jQfc!A{hUgVvZC%JmsLz;_!j$
zUOmj@s}iIZJ|8|v_JWRa6%4b$Z5CJ_Pv<ySEeHd7+q%2AfxO<xf87M~SFNH~K;8+Y
z3D=!*#pL;xzMGJkC^<Ez4fRKB^nqFXVBPM90hoCXYOKDh^Tv1LwKH!Y%$UB8;7y(0
z;pTX;`5w4l-2oz#v57&?8%SHjnVaZ!-<%NA+Sa^9LJpt5udN;I-Y!r404ztHbFSdx
z;N<IdX+w?5u>z?iP{sUb^CiwFgAy_P6#PKGN_+zYB3=_nia7duOc?;gVCyQqPuX@3
zwuDBrp-@aI(k<RuSL#sAgC)_?vHP^?p_n|{e8I1(-4O&d08ZvAfcv5i46mPQUKV`-
z-Wa^TBQAHT1Q_&z!|5MgDa50@$t{8BNW@`rM$vVc#D@WfrwU{pk&`|s{AzQyw=HHS
z!Rr9xYTg)DU>H>Q3A2|v0rm(BqrwBa*=HrtV@$%1btXTJF$BIS!l4M%6eB?CO2Ur^
zznv`Y7J^H$etv#XvQe#oAO>n7rEeymH&s4oW@fhPD2q6W31Do4cOPZR4ci&T-_Z;H
z7H5FM0w)D1aX82Fe2?{Fn6yNIi^n|lt*EHs=~R!rDmogP42y*Jr$J(Gg#yC?V1+Py
zH)epD5h*7zSjZZF&C7gQm;tjRC-vm`5c%T8KnHwW<fe`*VNE}l?j!hV4WtklDo(|j
z05r&Xoo0JsSCc3Z45X9mfGrd3I0b5V#D=1FX%3McpRzVmSb-EY;pgF%2FU$D`Ey!{
zOg#VW7I82b0<Spgp9a@}d@@j3PsN~?_qZi!nVM66Cs}sYOEf=xcNc@q_a+vWhSA*Y
zbjP=5Q4{>aG^zXbIe8svlL)CrXtchC8D<Wp(b1pidFmX-Y!I9~XMVt_{yyDk#$PZH
z-^Y1f&5^4!LhwZ2U{37{Mx!*_setQteCMlM5siqk`@ctQkQ@NOW%KE+->VG%?%}in
zP9Lx9<)Y%^;3IzO|M&&MqvhqxAq3|p=8@{$t<JL<(X%oR88VG*N*)k^|B=*^S}<^P
zkJQvm!xKAu+;u;Tbzuqu=j$I2v`CEAv5E=^FMwZiIG=FE;QdvSYM3K<Smc-}7(65Z
zn1$Jig5$+qwtaB6k~KA$?}>FTqK~#b92^|{9XJV|KqSH0^BW&0CGovrZjOq7vC%oZ
zz%qSN`wE7DfIzq$6YGP*j%szISAaeI8pmiAWUY*Z%s#YR7zL@zYIicd4v;ZBd_|B~
zs5yU!eFB-zt}d`0wVSF#KO2vDvvn||-P+ylExkVvXsPQ{2={;XiOJYSBBfHB?0&o(
zH%KtsfVbAv0C-b=-m~zBD1Ubei4ly(gtn29GN`hixLO^;%maF2(~R^#4jz{y^70;r
z9l#n+?zZZ2I5TpYRQn1w%(XSAlb<?KyC5yq0~?sXo77}%@fa~0_=KRK6oJV03mI5V
zK{N%h0S55wF)Uh{0QD5=2No6<bybj|Lv2VzL<G1ng5kz<X62FsRxXn@A7EX9?G==w
zfL9Ch@e#6_jKB3fQ`G7<pD77`P3hl*L<U8rkEiEgj+R%NFCjbA;@z`X$t1c$kbe%f
zOE9eh$r5ZY={x_3LpT4!%*FGUxuYmXD10_qh31)jR>BcO4#@P3jFL&F0IVL}=KeGG
z1nI88w}l9IwnxdNx&pWIkqn`Z&K$5ai8>o$?E=IM)PJ8}Qvg;|vK(^aP$(A{3<%&Z
z|M^w`Gd$@(1_X2gXM8F@2Obs_{>ao<|32#R>KxOlI-uVor+`v^yc$%!^W%y`Wpch|
zswtN&atnTEV%2Tnl?y~rO9Fa`ioh!UwbpLp;Am-s)Bcj;s#l_qSCJbt;<5xz7vBR6
zwDb`27+V_~Y#bbz+$krPl8!{#r+06?v`38Z!}aOe+3?U%1aCcw8&H(NwwOm}Sw*rs
zGGMIKqfuaDmc$KA;X%yXT)#3o#0-Gn!Va%vWK;(0H43o8LN8a)#Q*bn_6}Wv(QajF
zc??j*!~2juqyu>p@Rj|@zh*)yqm{s82e+3)z|H_Wf%$C-5ny_C)d`Hy&X!3JXL9K+
zN>YHCR|r^*5w6y-MGcqQZvC&OPo=*eJlFcCXOk@14jBNv;g>Yf?Eo`T5Xa|yboDJC
zOEXf2)6kn7vn|)Na!e^L4Zp4`r!IRX$v+-oHNLq``|Cl)T7kv6(g^2caSgGkVsvx`
zA24HqEmRD#Ac;SLrH$N<fqcb^;7ob9-iFFoQs9kNq-T;AM1n2m??9Ct1^&Fn2YKy3
zLZvfzlpZ`lMe>-M?}&%mG3DnD8Fv}oA|96vvEyc~Jf=*6HFduD0kdas)BEF;WSQSJ
zjszzVtM@P$Gebx*xWIWsZ`_1(tk7zI{ngoULA(P??d|yv@L~PWJWHm~l_LfZ9RmZ(
zy-y9};3M67cs6(KkS%%W2LUezuIMWG^#8AhmX{2u&=IokziKGN@T7EB!omm({6{pO
z;;fA5S3tJ6*$-xa4J|FtpFda5(fp8^xds*ge?<P+_t$RT+=s^1Qr>G&Vt^4gJ^e5L
zF+~0;QHwBAs;;TIM@1!^q#WvJHGy;%|7sB`0{IjfJ@6XfSb-UYh1?d&ekK)8Ufn|m
zF20_Q4ux+Z{0Ge#a0o;&dy9(^9g%;G2zD^E!Tx%olYcHCj=p?<5-fYpZqR3XZ5|vW
z${}UP9|B8dE0C5VFP^R5SjIE!hV|=!{RcFR10flL1NcX74Cus+lso>1_L#5~E(7!Y
z`Lk!xgmZ9oG%_^Q905N;a5p%MtSQ!6kkG&kMMOC0r*+;SCSIGH>jS%(v&U?S;O=S|
z8<#`U6txRDQhgc&qDAnhfPi6xVnMbP;X*HF4|?Rh)LGZ~vrkAQfAnaqw^z>nLX=h5
z-SF^m7~_A2bz}__;}%PERh2p<prHQ{1~bN@&)(}$Bw%1-&K>`?kdP^J1w0iLD?LAX
z!VuY{w)XGs4iSMG#yX&9;RvT$Fs*`l{xLvDnB6#k@4r$Cp)<{W#C$wW_csoz4Vc4O
z5Yp0J@7`Vf`#Rn!2jx6Kyg^=bwl6#sLk1iPzTQrKeLe8VPXQn2PB^<XBp@%+6`2D(
z&eVSrI~c12xfCb=>>dyOD<a1KevhYC5_N<+ZGHsL1wsxdi8qPOQAMXLOlm!G;#|Su
zrJ}+aWDH_re_c!nzPN&w?ng1>{6!Y3J+j!=TwH;LaX_5mRZtXxvo`=;THH)bH~u^f
z^j*siu@R4q{kvtuA#E&5&|G7Af(+bdvaautJl9?d!wipzn46h7tK0}&T_>wRXvK^N
z6OT^13IIzeB7cu9{0`{wo#qemg6Lm}{)eCI+4vw*FmMH!+Lk9IBC6#DaG3&3rT@G=
zrKb{VfSnVBiSO?VZdz}j&5Qr5KkV~$!rJzm<eJ=LI$J<J1kTRRvwvfaZXVdxssI3P
zJ?3bAdg6HYq&1qLh=9!;gi_;h5gzX+|NP1(9$^4|c6T_DVD=(>IRK2mz<w~e`0ck`
z=vIJj4rtn=;PegJLm+}9RMt?$(3ae%rtTaKYCb^#u_YWJ9@>@F<N;*b>F@O`uZo$N
zz^ZlyKYnLu;7GCh(J4aN1&t1|7@jgH3m}jl3_0Eg3rc=3Trj5w98Vau(Yv-NXlQ}T
zC{ICr1>2LSw|C$OoLJ)F=BDTM+vXUy$`_xm2Z>lp7y`(3y(rKNO4j53Dc_(S=qqG^
zy;<%IRLBUCmE#CP4#Y|61pRv`GK7ozL4CY%1Wki}K>ESaZpN9|YG~RS)ZTaoGq7#A
z`QlZ|XJQ0%O(Kp}SyXEv_1v+u%$mLq)rRT-d#qwhNTZ`^YKKByo4+7!fr>BUrKf3q
zNv>qR`;V4(7T_JQdwtC@e|{YKg8@}E)n5=SOLIm)_SX(q<xF2c)hXj4qP6A2U;TA;
z+?Sbx$f!-6z7pdN9LYzWj=$VzZ{Gnin|v>%f;4<^9LkekgbD@LW&njaPZff9|CA2m
z0yQ3-Owb9iWX^B_1S_yO;M3!uFFmxe1-k&alOXtLcyK&K7ElobjJVLo`?5sp<4qzW
z<&h+-tpdkJN4>$Eq<kNv2GYh{)G|axeQIy@nH4$Rh&qI8640xi`Ph8r2GVj+Fh;B7
zZ=2qZ$hTAQrnIdAym4vky&Rb@Wt+~C%jNmDN+^%!9vqmrcY@LH$yz=tH;1liVX3r)
zgjy&@V{;QU`0VxJ;1Fb^1<fzGSXfgtGnSB%L@h0{*&OYgK6voJXP+t45b|9G<6D_t
zJp7qnRx8X;h<&$f4e0xaI5&%R-OA_)dNo3MS~Z4?Wg<F1Qko}b-OgN^dLNTfLE9E<
zqa4`BPyB;7Y&$GGTt76pvJ_+_98Y7^SmEFXf2io!!CBu7L2U@Xo(5C-Z05WFGVCiV
zDw>T?083{;FnXNCCA3`t#@*f6=y1x+l0imjC*`5@*;orJ9i5oxpYub`5ph|}`^|qz
zYf3YMmQ0R+bTlOO(V?Nqc?J~}Ak2^g|E8I;Di3zbPes!&Uzd^50FMEhbl{<Pb_?Px
zkZfC;nK3algX3LD5M(4Y#Ojc0mE|BDNVSUhpyO$8V^azbBAf>7<9QzxNvrT4FQ#S&
z?m#!cDdSgIQb4(KMPg9P9*lCe<iYS+tMs4J;sDxxR$t@-A5AtjHFXuhW~Isi5km@p
zk4i>wP5MwA1I$zCONxOAG4T5C-~Ho1cH-YC+W)z~vz9T+f6hi=$Z{!jFNK^=N@jN3
z<NR^$j-6oJ2h-<ngZ|0<S6hva9|#Z16BBX%TAd;}E9aTpc&mK;HsKkRkN)?cA<mrT
zp#NpjeYSV~Q`7t(HN5#p27mVG{;MBee8Rv0xfI{)?3|om`8?da?)X+E9k<6)d>xsp
zG6=AKxk&wY-|0WLrN8m^Z~pMV_bXVVaPC8bk85GByqB43Bv(Yx0lFGnGA8m5Z+s1n
zwM5^od8K21R8*SZSH<CcL7HV6qC(o}xUeqeoo=K>n8tv7j6|?7#bxENup=dC`1^%D
zrF4TkGcO#uR&mqWU$1>%io@9})_u{&x@}ft86x7ybCy1y6+8d-gYOzR@*^3o5Yljd
z3p`a{BeDN`ngHQLO=Lu{i&Gw>+H#`a(^}qM6J`Eug=5+ttAe+(yXm~WLHS~Th8<Z@
z<qWF$xACd0ux?KsrJ?6Ll35iM+PD7m2!+swD?tNS+uS_d+4$uMPjpGDqU5h<?FZf#
zWH2`g5YZt@W%y%H_RnnizjbQ;cNhD&0sOD#;{TuC`2WVrpN?S%6t+m?2gcvIRhh&M
zO-fMQ+0F#XL$)Oaz>NSh!efDe09swY2Yg>9DK<71;B25?AT-`^)Cbfje@csM$UYQs
zcAODFFW@*QuJ8zS@wvI?Dk`_n0`VtJC>%i09PV4*KQMqu{m-7ZfIoQpBE;P09Y22P
z07{!^T%^}U4lIqi3U$@U0^blw(|E9ki;a>_X3F#QPw9OZoqTO;dWlG1x}T(y`0o1#
z_Yxi0+8hLPsY9AUaIuiu-sgU!QaoWtSsp2nSh{*Y7rPpaY^o~h5B0Ol>iQ6=TJEje
zHj|#@ZtcG(Wzkpo#M~QlF~3{dr{}gAA9qjx7Y$8j>ot$={wKuDJj?Hag~Ngot+l+q
zn_)@Ki5&d%yws1u-fSb=Ip97rvLI`KTcU;rGah*9LLWo~aXN;4jl*dsG_RgUH`S(c
z07|c1?!eB>voZjNsbCwO{E6UeHZ8mLYGgJ$8|?4^1X3=6QT!+TF|afUB?mn!qd(+7
zjg^%X(AmhjjfaOQ_EzEF?+OF~p~T+K4kYB+kcmQ@XBGo25!4Oga>J1F0#5af=C5**
ziv;=yWAKR}?F9~}ESKO3Y`fsi6I*JVGAZ;HSa5)X>EB0K9Cpe>rsljd$SjL%18qUr
z)P;Q@bCj1>Mp#vSBnoK#+C^903U*K+Q>MO^tO4HwtP=B4$3SKpDWOMCOUp-y<Q*9T
zr8F0?oZxu_f3)Ig)9XMV0bQ1q*h9_^WzNj_cotATabj8kpA8EQjRX>C&?Tqs4natp
z9X+v#h|<e1T(SmN2lzU5b#-=jcA)f>C9(vnW<M0Etx*<YGpLz?SFSNF*N>Avm4H+`
zTAy6=$%G6cP^u7)3WH$H4;~Af;6L3U0=YckQKmoIH^KF~J4JR0?gG=Z0wWGMM~kVF
z(I!CGpjiU)Y`|!8Z|+Z+onWymmW}~WB@cK^kcr7|Oi9K%gY-UV2e^_09UYQHmLot<
z0=g>XRU8=5LXZ*(#=&<1FCa~lqwU@oGBOq?N45%|D24_x^iN#G$=!iF_%cuT`sz(Z
zI3$2cznfC)7`%S~lv#8t&Vi-HVY!%`lLx1-0GkEO%#(X6QLxD26vZ!ziM2m|h^orY
z2!q$RB~Z|Cs4ltz4LA8K2NA-nawcw82544h&<v%fl>WiHlBeGzVj6Q(1`Ztq5^ic5
z9xkq2ij43<ilSH0TqK+?5DvBp2O<G;ahu(Cz=Ft$lS2lEZ5Mb{f>SsjA0K4nv~aAF
zBXgsM1RIc>7T_S62W?#DaAYM8I6cZ-CTubIy!m~3FhA9kC=jI_piF=QOn$U?LJb4P
zxDpS7!^6U&*i4@U1vOSbzuXB-)UTWAi$AyKnz~ikKj>Q`7)PU{qreXd+5uP)4qHrK
z%l86<%AW@ZAiq5X7R^<3bTC@af)DsC03X<v(D)d9T`BT8pg~Iojsn2R!L>kR$Sm6(
z1e`l)2T@tr=>f7qog|8M8BKSr{J7fb{i2>PoZ`{!59A2`r|chc=%kX<>Ow<90WF|f
zVhRQV*9qWq)TxqoUuS7ta&vMtixa-U(l>=ZgEAek2BeGFfUN{QIY6|a9;+v{0QU8J
zjZp9%<rflq-}#_oe-c&J$(=7H;BFr1hpA5tT!2uKLEj2-OVjXEQt#rS{s0Krbq{oa
zunJu)E^}!GA6OjVI8mHK_gKhB>Z>AAARL;s4xB!j+X?V)IS*_O43$!wj<GRkI1JB3
zBMaa%=vf{0Rh;ZV-$*81QTDRbPR}>B$_x17FbnaNBO$l{y8kX^-?GES#Rcq78*Bj;
zeg$*?*6Y&G<pe;bURhqQ`$$B?%ma%viqpDRZfc<gnk+M)@vdLL8zn6_3g}YeF$}#K
z90^w`=pGA&=%W?-K7<>BR8(T3mW|D~wEB1Q1WP@sb4yFNj^z_HhZ(;Ec^A%Du?9##
z<!Fkx0lpSNQV_6RpVjI}yFdTB0hT5yX{*psBz8W^_GSV9K3B!IaX>!7li@xyEO9v^
zXGOY(wzfVeH2t<i2%(zm2&(nf(NEW>@|prysePc4+ie_}#1Z@SZ(#eMiL=F7fHngL
zGT=`zjnHs!UhIPT1*-+-w_wsS1njSpk!C+f-@q|TbT12@mCOLFLQYOjeG=ga!Bw0;
z+HqF_jtu~pacHXwmz2Ro>7<jPQdQdt4TB>v%F4JAT5-n5$iOz^hY$s|sU86YS$ZKj
z*sn0b6@C;K6_s}S{I$FOa(E0oWxQ(+YrfM<B8bt|)>!}-fR!)Wl`HKIYZlL*HKWp(
zGJp%75#qEXK_3FLNa1*JafQ=dcT4Qi2_D`A${LiiM1+KzNN|)+i|<cZ=~7>UA#8}%
zZmh4jvweduCnSFSyW~qiOqCpXJq5RCt4PQNhh4HII^1$lQHj>g*KD{!guuFK)f)0P
z04ivFjSt77NpIP|gjEP7AK(i=fj0${gkKo8-hma@$rxdrWe^2SdToJojUFL3mWGA~
z7_1b6<Dj$iDa<(oCyro*#()7?>(`Nj;(3`0xdLBa12)oWwGOcj#PtM#kj#1l5F2pX
zbF>}7Hovg&_0<N@`TSO39z30ytX9G6IdN!(<JUg-#^fW8z}?4y9@HF&4<I<Kw10Me
zHkQCYR0(XxdOJ8NVAj(aI#h7w!239eW1CXTbix=u5-=}zU?9MTF5u{0*ZZ_^j{;m4
zglaj#X)g-4J~%|Hskzw@UeAF+r<f<y+ZM%{<Os0q($#PRWJ@PE1;vM3KXM#YWb_8|
z^sxTCnev0T_d__@adVTJmbTl}1nfO>PR348<PVEsqk?Hw#|u+t(S!l2RXB@a%J-2R
zojzx1WF+kpv?Du}5M88BCh#T9`PE&7r>e{2>*)&W2()(IsCfsKhg@94y}eejGXMnv
zsLGf?88$a>s;{T?a-ED?*56$oV2TIY;T6P`vt1;_K7og=rLV6q6$#^i3V~~Ik-MrA
z8tuTT1MX^l51_{;e&8=59gfgQu$kOpXAdiJ`JXKoxD&*Bif=hZO506v`ByP0%Rl;@
zzg`AxOpO1fmqCO7Fbi5)oBTX-^R;`GiH*#%WzR|)3!lx-Zc=@4{D8SZRzd7XLWR=Z
zDK*&2J97WSg~x`NZxpRmIR_kVI4Zc0&APsrSCO*ylsPWEF%~+wD2#XC3ght$^pT;$
z)h>Igpv0+X9-pKPI0%n-w|&~y9@;3Lx4681)QbOp>9vK@JkR!0r@5t3+#3tZVsjB|
z3!4~uVGi2#yo@B#zBj#9KYwlw{h$ad{63T4dc{AbQB?fqq=As}Q*1vcjiVJWG#e3}
zPqT4!?d|+O^qRWrq{aMdj;LyMaG1M9$0u<}+!D?|Eq<6lgB5+a6mmi+Z2e_@li`P`
zFuhAoWC+S}#?)%Zj?%RUGgXXft4}`TvCO-h1*66-Yf_uvV&ivgdO0$p7iUzz``#l}
zKEQCtki;lHopQ_e$%8fOh1G|h3W0NS@><H>K^t`14PbRc#q3|%%lSc{`+@B3`52G9
zKC}Gx58VN64QZJMYh@(;^WQg?kT~sa?GH*U4O}mHuQ;9VV!yr#+Qy!@IJ`rvhcP1|
zNviBUrhXKKpL$8el(6q93|rvx-b~yrLU)@J9reCzwlj6u_fgrF$c^ff8-EMC$OGbT
zTt2@H1;?Sbz7CwMNGV3-Inh=zJ&#Pf%g+zxU(y9x))F?>kL1wsz0Jw?_s`9t5q*2r
z$32y>xG<h;O$5_^y2sO@;QAYeS7J)(L4M@}!`FL66p`u8H-c=9UqxNQ4qWp@E7lI@
zv6}IDRd)|*J+fElN=!A%1E09Xz~|j;@p<f7YP}Di`fQF-lhl+N-5<kJ@}0NBxf@rM
zFB#e8a5?QKWB1YMJk5k~aOOO<<Q7t^cl1k($zpXoZ8j2Y6xN6OF1T-e1axTgzuZp<
zT)B{E#qN$O^!QE9+X)w>iDccEwU?+gy_LO*nLM_q8x86{-s#Iia=p<z&uEp#cF1;#
zHqIaAohAB=T*r^xEUrgMDI`b=-i9=<kPZ%H^?9*vg(FfU-V6JoxkfMQG``M1XnH9A
zwV)qegP&PC1yw@gdvH2WXm6$B`GSFtCo1%Bnr!I>myBNo3fz1r7oF^@Wh3+M0n2h=
zYBu#sD^;ZaLtMV$svn*~aSK!-wd!OBcEtxeI240d5??C38lXY9l-4|lR(1XH#&^o1
zY9jfez|CM+G5Mxp6{#-i&AThq8&OKO$61m1<~#M2sl&mVY$MGsqbN^rR?Vae&sU1c
zqc`NSu33oKzbZ9I+dq}5Cva)Ms@N1jo{_3W5#r3AJ%$wGbz8F2eWOa0b@!)?Qt|R$
zzuGjn(k;H;nie#-n_kHxwi%S?4%|ET3_D(x;-0r{`J87WVes*Zn%W!|t9XS@V`yko
z)W>^|9D_%QEw)JS_f}e;CM1qg-x@gXImY!0N3V)*zJ}r|al7wIlh}vHC|@Z!{T>rt
zywvY>V^3J%vnY>2UxPnCl38&oI(LJ}dtw7w)Y*+h>8}A&hc*Q4A;gl|Q~LVpTMa!j
zVFL-RUz9FMuQlC$ramXFt;NiM%q%{fe{=oY7gky!d_{vjRR;ss4~JD${+CvK-(mR<
zaSGI2Mqh|8SS^m-Z;$)Q)y_3HYdlTc^NpJ+cz2x%zf9Sq@_y^X{+{z*UOfVl`QdUE
zYl6XFXQZq7G#_R};@1sY-Heb={oKIH+!Soj=A21nYS!pv)!bmZsoRu{vWdnPU0&vS
zskLqXCaItX`O`(in3Baf??i7A5urX|8lBs&BSjx3rBqkfbj%YQsIW8l;wir`-yiUr
z$_Vz~%hM7*He69$R(~WSE;TXhe2z)x#W!<1B?Fl%N8yE(Q=x(|C0iq%2tzDM$Bm#*
zH9r-56ec1KqV2vVVdFI2k#LJY87UO>7Lr=lB6-LoFh87T^4a>EwXqDFn@5rL+Mw}B
znSJ{8DijM}%hkpcLC!+tyUbbK_r~h9#=bmhyHrg~5YDG0gA>+r;o%zcW}Uix1L3M&
zGOOz~<onTEFA3U&q(n^|m_K?nyrM!bm`%QILMO{>QSY>CeXH8Iz<9T5Iuj@FR?{dW
zO(a*gbAeVI0cCS20j>Z(xs^Da7mg|VI=oZwVomz4ne$Z2eJONx4CaiN{3&t)%L&Gg
z!h?G_OKTe_EJ9n6RlP&bi5GDcmOqz&vDD(SZm8RcC&bBA##7mEmUFR;BJ?*%r)vAL
zRKs>{bCfU{Yt3dVEfMFHu7u^+VXvctQ*)v8u90k0_QmdP24hjIN~MMD9@!y=t6ON=
zwu<#n`7R|G-zYTEWqgZ<RqCF@N^<d#lrk4E3*#xR6RK1<i>K_iw~uiWj&^3P_Ej%<
zb5whCk1J|dd#j@naBrpLi?5@YK5^>|!t*)y%qPH}Qv33%LhWk8Fk|V}3NE=MtVv&q
zg?Hnw%$^YhQD!<<TLuL1??}=x$UTlWG8<r>sdGNxcV)8jU$)wZtZe^mwZEjW;_CX+
z7Dn`9cNiHMK(cCPZgGd5je-83ZTP?QWEfE#GQP!n&2dkDJ=^h0NFlyIzuKJ;4)pf8
zw^0}L&`JGfp2yXwT-qcsIVsnnZ0OZapU0)pBQ|U^^O#<5#&7jAz>AX?D7xmlsWLDw
zbI<mkmJ#;bphVgWdEUC@@eUorVwdP#<TXp@HTmD}wm+d3pva69QF<3EMc1gHpn;WO
z$ml`k?~46mlw9a*-X$&WhIjbn?<5kgP}x26dqtz*Etu6{=1ubXK6y5&M&SGN?e3p)
z1Kc#5BZ{`a^9jt}2_sbItGW@D??+2+Hoe<^S<B;tQWzy$vL6->ef*0w&txHq4`R13
zW#4OWP|Ufn%xD*#^4YmnJsHWRPUqRHx5ejUb+^8c1r}b^OPLGAEHCwGl%|xDix$1*
zWUI|M%E+ZIddu$K`DiYIR@$aF`8=rG)9F|1n(A*-_tT<YO<>@+ZSpI9b2;cvJ+Fv~
z>3Q^;H}Y>EDc$g-pt8C2i0|3;3kfdq$VlE;(odvbB<FpS4pqE-nU+EFKpp#P?YSJ!
z8skzQKlRWmk!+jy3?hx^&94asdwtGwc}^=OVTjsnh|QUH&R=?2`4iVYYIDun%J)m5
zIMN#r_M2%jNb4w<Z_J(xC;aN|ahoyZb^@EN9#Tf6LY?cUj=)c^1Rg4+`O5LX9Dj<<
zM=|jP#VDHp0$Ybp7}DEoUz<O@;K$(i^o#39JE)_KtW1%8c5kW@&96x4eQTs9%bNi2
zfN~T5aI>3{!XLXvt37(UG%m*Ts}aj(o}37NFj6x%w{&vKsc`miVV0ywd&rmcfL19T
z#dGl@lAM7HHb*6MDzgy-fuvb#atg)N*2eVwl-diLfg}vs9rgBtodikagZ(AT!{ub6
zy!@OlcURLcbx)*kRU*0#_mEBAS&nPl)(gDgvVWR0@y37V{BSl^$pBJ80O1_wz1N;7
zhlf304K_A1g+F?{l8}3v4vM4Fmu8ZDnb90xR<C|ZkJyze$Xj@Vxzx1LqxnKrY)@G4
zAfD$I`fcA#-&)*)NhNES6RhobxM2$x<~7=0VI*~rz1KGN1r26<_&CVziNDhcHsjXJ
z>I?e|hI7A}XZd+MtLJ=yDOW!yS8ns$Tb!=;t>;DxO)uZ~&9F~e`H~u4^f`AiuP0?o
z5xbfDs+p%^L6QAAJ$7-IAQx=`3d2|%lM)SKZd#q~k!wwaa?5R~R}1Z_42Y)+M@T}d
zznwdFXYgNG^R#tU&bI4*IJNLN_qpxbw{Ay2abxoPB0U{FdNkq{zTucDc3jbsJpoML
z*uOfZM$3zTT-lz;x?S+2mHi&evb>>XjlN`&Zg3-s;^P>x6v3<2#IkL5^gnu}cLD@e
z1fy!m)yx}&&34-k4_Q?r<G*ofI^~VHNVqx85>UCxmk2DT@VkqCKIV6)r*6jo*eN4e
z;ZCy1L$p6?y1Y!s`9|OVgP`YWK5yAfNiM5H$b<K^@r1kwH5Cy{FI;lWGms1e0}TD^
zimvM5#iQT2)u`T$N?{*W|Mm6=8jqpi)aASDeG}$Ce=aBwIh7tIri^|Zw3bh@V=GS{
zeYTj}F#q)U6-n6y(wmI27*eKP+{cQ2&C`~b-<Wf{KIbTISuxBU+|Yc;R*}{!_?$ab
zrkb~_J3ySM`kirbuaLQNx<WXq;f;PbIfF^Q1oX^!e5FpF^aq?D&l4;x*t^xt^^B`)
z6+YMdCV_5DU_`eaXB3>vGF%mTTA(#rhx+{3ZBq`_XGx+wo|)84`#y5lnzjB*xfnyw
z#651>vXRT8YjV?Nm)wRZe+;k|o$_ko84}l(KL3Jh5wXD4#uVjies{%x^TL$x@S_$f
zmG=U!=N_rFM%?0Pv%Pg|P}J`A<)?WLc8=^0@-fO$YijCNgI*s#=AECq@C+lF&qc6y
zS(k!Vd`a6xMb$ZZ>5JWHUcknH?7~hg(WB6esmdP>`zx9zri}WdR=L(M=Asi)nTHy^
zhXeMRA94o|qV#Hce$27jRWUQ8@DHJwXExu&XkeLE`R?)U`S*Yl^!@KcZiR0(uC?`E
zA~JoreOEsi6Mg*Nb;sH)TW+J9Jt4(q#*Z}A6hEEA#WoJT@#2RKePN7hav_IxWR!go
z|DeU{3<sg2)z2I%k7Q@Axb%;LfyV57POKP-2~pXN_L(S|=AG`teF^Ofj*Pmk2S;}g
zGDq-D9xX_Zez8)|C9&uqHeE##;@gd;$76L*aQ3OnLr;t^N_^-s^7>mTa^gL4k2n5U
zKNfT?t5opon=radCW_NgbD6EWTU&6qIBbJ|e$0H9tkv&WnDFf@g~qAcwLXrUe5?MP
z85}ojh^N)EkX|%wSBkuR8|3`v`Fc-G(i*R`g{{-wF&zyhcNFY5H39Tbx(+QKH@rrA
zuDG$FCOH8j-<gD|@)f*LA|{d32%N-po4j6Dn&PR5Ac~|96=Z5L*TqiQ?-mhV2|1LN
z#TZfx4RKsK!pl<QB>R`mn1$hAJp9<09{yK#Z@NN*@jX_wvOW23%op))gXiKu?RE;7
zZGA7i<MI7{R5e?}GmM|demAJTcK4VfPo39K@rt9|{1Bm8R!V1Q+dsm!!b3zvD@pb`
zPB~fy*A<E2l$GE^C4I+DoJfV}(cqZO{DSi_w;UcD%&mXe=bRW?%ZwDRf3@BedMtbK
z>5-s=vFWf(IaB5aF3N3UFH#St$xUW+J%<Zs?k{FYTJB^Pjht?7cZubqY?eC@wxpC0
zV(S)iRjib+$~-LcE+iHTZcw$>mAiYP!gRgM_qATpq4{2a{{g+iqDXHs0dC1TSv;4<
z%Bv|DM#y|PCbq;VQSW80JY#eT5xmE(J`*7EhVuP=O^0i$V_9TZofHP$m;}9)h;8%9
zYdso~1pQkGb8uH8?}xrtGnAG>+i;oZqP-LsB}Ye5ag>Xl!_<r9lHtP0IL^s=ROs-u
zPTaYy{Am3sRg!JxwR-J+iQb6F;KG9sLJHq)oA6G=8k3eKF9mh*Ueiv$!-+LmBVeOq
zr54ZBn)LR-Em?v0f`LavD`f}?Q^BRn{_~G-+>-4*WV*qhTF#gK^((LHJ-4||iq;T?
zZQ-{4`)O=Tds=%r%eV1mi4E@sC!~%Ou8-lS2)zApvl=PKC(!i49dceWL5pzn?#fYt
zJnDpV826u1tNRrvelTW1)yjG)L@9djity-zV>FC_bRj*S#h^CrjRbcxJZ5iW{%@yF
zPE)p-Y)h|fMEzbSeMd$!TW`EL<T9Mpa!1E%BDkCVT}tM8G8!YYY~d?(Kfj?-%MN7}
zoBHA*x!{l3e>>;bWk~V3Kax>syv&e-rJeI>lNI-A;Iog1;Wui9O||Dlwf{HH-YO`r
zFX|q|8*SVrKyZiP?(QD6ad&rV+yVp%F2OA%c<^9>;O-XOCAiJ~egAJ>rlxAD=B;0v
zd+2-axqGj*_F9XD!DrbJPIcA3Sj!e4ldH8N8|j>wBRHGM9^xcfo}h0`aW!T<GF!tW
z+*<CIABPr<6tYN?8g*OQn&*t-z5Vq*IBY@adw4!F!(?e4)ZpH0JoWHOP;~oKm|l{%
z+OLi<nN^Y1Y^fVp(S@T8GwHn7sk7$P#Wi%oC7#^->D9#e_zv;!pbm$Fpum)Ic=2xz
zD`m|4S@o>IxQ@l$X6B^Y%jbC(A+V6N?M@Y6G_~uP#W$E<mTy=S9&<aFK2>yH66zg=
zQNcNU2TUHQ<CRY+iFW_!%Evka9xU&wf@7Tygijn~)kGItM})_8@l+DpzeE3}%cz7b
z4G0$a<V{8`c9zv`wxr;;{q!fX3HSR%`{#lplh|!;AerYb0MsFRYW$9typ8axW`XtV
zWB|)$MC%gTu1==sG9O0slNl%{h~{C^lO%;;T>aT{RE$&xBHv;tQ%L{M0O3`C-f{Zt
zt+h%O;Vr}bQT<%8`J)b%KWumN%{@dsS*rG$Uw4p(K$QsV1Hz$%O~~K8pS3s%LS(c7
zd|65Fce$&v?9aWH=MmF48*f+RT1wE|1<wmC=Xo(p2ER~!miWd7)hngy)BaY`gb5<!
zaT&L0kz#e9Z=UZSODRG!m#f0@uKOJ6xs7Ag;#sT`(QdwAJs1D+FwLfN`|HM67r699
zktx1gf8%5`Ooqjwy9pM}tWmECAt}eZPPp4l%jV|++D0H4x?RsU@0xegWWaX5K_HVN
zp!|aw5a2G8GY$!A?!=qmt#BK)5g}!}IvL^d_byH7i0*Z1GoP<wTxIPV({_7NC#Iz+
zwA$g_pm;LzerZT3p;CLNIM`w~`m=_4VKx2-+^zR`?iec88sef(-3{4M%n^Hlo=s)|
zddoJ&u^Wj${!I0<k>AJnXYxOTad4F%lAeVYPlVPHP5B_t8;^O~t=2v#Pdbgn!qUPP
z;s;L+i3`5gLmo4YGpE+WIvZR*PTggqpG5{BCazs_OR^ZIdj6ODQx*eIhfdpv%DWEt
zcV!??7o_u0ukyuX%}$xW4TFmz@8&yQQ@G~x8A$~aRibh*=4+pul5681{b#vk2c7B?
z7r&&Pda+9y<CF#})>zirWg0e<8;$eX%wLc8trO|9hjg<OL)E14u|HVW=ddL=^becJ
zIhNET<>qUZ?(uUv#l-21r6A`C4!$ipl^l-=bQX}See;f)<zIN>eGeG7<N|(VP|igy
zq|a;_YSCJ`n4+D2$E{t%B@M06Hp7ogAgNF6yoiTp{*lV9R-CQ1EG{r<b0Xfm${13m
zmmXHBGaq9co;aOXRC)vE2J}DWWj!V5d+^WUEB8Gl$zLf{Pq?GGrxwifo~^tpdrzlR
z)1vr`D4Y5!KKSE^!=^Q;yUg=!6d3$eq2ca&Y3`WBe>A8t`pf*&SFdYT>@a&<FKyy^
zA+847k39U2c_(r36h<2XQ<<~bFzkadANxq>W+nH@kP%lyBO?Wq<azZ={`eA;%k)ul
z^?q?c)^e!T<by(Cp84=zqG9TV**uM4{2pebHS451na20RjkCYppIX;JM<Q`TpHF$s
z9UrUbUFk2_e}4Wv|Nru~<q`Ps|7d;nf!^23!joOg%iPn?)rwug$;`%zUC++a)0UDK
z=!WI3>}+g3DY=1<l6IaR>Q?SjE>5m4&Q{Kzl>F>cE{-nlTCTtp1iQ4Ax1EKRrkv#e
z-8S>^eBl2-@b&+P3)D#-$X9#d@$GUd;QWoL_?5-Cv)D*of(02FHO{jndNSagmIeWi
zhDsa<t`b51eMt)3VKOj7)Oly=qt$HnbXNJ+fRJ*mBfmYm1U-4$3A0-2_#otAExMog
z;W|h`)4^aSd%m%Zr`As8|L;dB4uW`m5)ylBYXWTS%l3B+OiY5;5nf(icBKFNH*Yh)
z18g{8sjwN(-ZeQh^R$NP>*uGE&3$;pKDg!DnE1cfe~MoMg#P)zB!GDxC*1#Ck%>O|
zKlhTTs3>5B1yI5+H&6ec6?dwWCB()SQ37gPVCt|A_$C1!-g64>5j+}0&RfHo6cA3C
z@@(f&<#KFmq0sYDMd3qBCOfQC4S}i+E@_P#tlFgJGIbb4*!x|3oyA4L`T6;~=V*a+
zl^5Y;)=`)9;^3ZUn3jqPdQJ_=OcbRb2p`?*I|xOSwM#M1_1-FMc4g|O$IXe-b?``)
zo&spIIs6txh+C4+GzZ{fDMKIQ_sp>o!X)AzAN~3GS_qS&YS~>~uUhl+|7rCBKPZiF
zK}Y-hfPoZXnG^YD*U+^Y^kNSK9#3LxW>l>nkT9$<e01ugtQYmm>p;^Qbw!@fHNbDY
z-rE#Y8_STgjbqIap~Zl4SRugDu{p<RMno7MKF!GBH>Ufew6jbs?p`-~kqRmxT|a--
zm?+LF(@#Q7ST<kExg>aJgd2s@kbG78B~{7eJ6BkuVg^7kFvLl2Z~iBl{buqK?lCzu
zv=7XQynKCI1Jus|J3I6#Z`j-mTD5peYL+=3+VR4%zLQE~g5Ck^(Wi9k;~Rbx%aE+~
zj`Y{TS^kmgw_RQ6TZcXi#dEaQCzEq!Mnru>Z%_ZZ$Q+2FxUSZW=QF?4U$oQ1Z8AgD
zVS#xQrY2L-j*XYE`LqUqr;@EW?3G-%FVZ5jvky2jNHyDYXw(JK2EIF^+|h)g9WrAN
zHIY4E0~54z5Xh#ul6%`~7@rQ)955COSe##Yf>jHe8;d)Tv<D>NQ!#E=;GWCo<$}&P
zR6*;Za~~6}mB_=|MP3TDn6gk?4HaCvXMFZ4-8dFzs~NfBAx=*};*>DzV7i7juN`<D
z3td975dt=hjA%loKBnJ~N*^_&P=XZFk=q_9Vk;@R5u;?(sHK}ovX5LFxA^D%68ALI
zr{(>+t23YpO#t@=C^(#fR2U!%3hCz8(o&W`Fg4}fAA>Df^oV2)j2{DPWRoA3!+_5i
zKcIaANYHW**TZer%1VL4s!SZydXGfLp7VfX$<~QE+s(1xRp(O$&M2#=19pV>#(~IM
z*v?N`6gTwf;0*o<?KY_6*6T=AxX55nuXd9;Vx9Ih@8ly3n1RWiN+46Prm$EtnXKgC
z=uV`UF_cS5@8jm~zPt)ClO&Vw*0Zz$8m&6x2D(m$oxpEpTx0}RwMCex3f>YEtwaGV
zBmyA0m6le+l?Hj6u?~s9)$4jHd6PkXj+wd2lxH6SSmU|>HpW-V%`t{OC7+-`q*lWf
z8QwhnJdVegy?-Ax&mf1qG~jvPA;OQK@Z251@66#Qv_<^oY;7b7yo_Yf-3`4DBc$1n
z!@||sbIlbiwfWwt<(mRe=GbC^{Z~G>D<^qFQei$cjg@AH)srBODiP<*E<D*m;C`VQ
z7JNj`*SUKH8nH1%3%Jc}htcJlXCDguMt`3ogO#&BC8tBA-%>|vd^CLo@yuBG@u#zi
zg^0{Cc63bCgA}5JUBvG+MX|{`Y46Cw#!%?w`TK{3Uy^Kc?t4TD_qVN8Hqla>%fY9`
zm!^8UqXZpkn=eT8%(;Y6?W5E$Y2%Xk+d_&8l;j(#e<V~qE~XKm{$wmjJmB3Zbw9)G
zKt^M$IcD7=E^jcEIkr3=dmsOaxR>+SAj>*Iz1JPbh)zCaQ@JJQ>7;jTQ0#~xCDD(q
zN^c@85<rZ*$Kw-9oIDFpnw?*C>i0U&kYeYI>d9z!jrt$pt8R*<E%S<eqoU>nCV9~s
z4{3A=AGg&d*|tc`kpJ;(|B*=axY)x`lY%P2{H?iksTJ4mx%HXDYd$LoA|6~2d-*#z
zkJ__cZB}M>cy`EPPZ~7R!o-b<1oGoKojPU<uXS1uH;$2qo3}BAP!6r-b!|y()~tzb
z<6zV5Z+n<xT~TdBiFDd%oV&zyG4X$BLX9U<Q#9ZaFvDglu@;&?m9xo;DHb0omjxj;
z;jw~v#(QDjF4lx@V1MR47O1S6KuAyZS%Z%0CH!PvDvyma5PS)ZI)aR9ywDSt%Zxh<
z_OoH2P))Fyr|9NiiY14nKS1>K3hK<v%GlT;8D$XR9YZx5u&D)H#7Vh&-LaH}!!vuA
zcVja12C7*V5=n}VBt{5ux6K97((g93hqr<A!oh=sO4Ml%HCgQ+f%3|R`U-d?r^K|`
zmpct=!3VxC+kk)WVEiM*e6P%ruuR}jQ?4KJYm>4}8&O#Rrnwohed0!s2M$RtC<yR(
z@&sdTk(0Wmu>%{Hf!S&rqn@na0j|Ensh()o6e^=agOV^?u~iILa$+Mnr(_ySr1ncB
z)*{FN{2D2sudV;ur<^<?G0!fllZ@f$7DX8&@D&VaWE<qzMR)h!r6ly`BZ<aQ=erTQ
zzU-J^@}0f@6gVKp_NP-b=_;5d@T2Z(u4X-JC~D-K!@{(HcnNVzAiahBc*+$~*1jmP
zLHdnjR`g?Lyhh-{an<aXIc}Xwya^R38`I?npzcNQ`=jRE)n~{Yolftc8wy~4u5}-3
z8cFktM>C6w4j-R^UT@wRDP0M@!PHC))*zKP@tJw*1~|WTvO?F!xwXMj?7XR-VI6C=
zKV6}?Ij|krRlro~uKO|)^7PcAQ!CQ^EIlNCtSoG|@X@*V<%8rT0yB-{O|M5Y7vzX~
zl1{3oFJlnLjd%l2!?N$HZnv;ZfZ>R+`|dUqW-VXf$ag^qT-fKNHI`ldv8OxOK)h>)
zs%K(#iRYyHy6)Eji}U*-i|>du;6ca6P2JE{JUy55IXBFXM~S|!pg3KU7JCSh!vbc6
zB+_u_{jZi*S6elWf~5xi6N-)zkS?@P^Q2d<q&R1`)y%v^60|QnrAwLQPEnV1Rm8vW
z-;Yva(!qK4p%mu}^~~IL{<t?K<{Fu?qgJ*C=JRnT-K>$xG4Mo$_`E#~mvIOxQKbp0
z3Sh)S8KcSDr~kmtHxIloc_6rl>N%B_6?X^=4#|Hz^tKXV(ZJ3cq9`%9^+}jG`PlIb
zH(m_po5V%&Uup6+CL16|Wh94!qRyqPRyv(Um##6|%~k4;1f8qv$rs|awVr&Tc7Z5;
z5~y&d(+}dfS#J-!=qWYc&~r`flN};x`Ja1qpG916Tr~(ZKvhs1y!XYIH`)fa#?i*r
zC@&Fo!3z-(a68{9_eek%sikqdqZH7iGt!gAbh_Eh5>dK1*7&4=b=b$;zg;%%7jtfR
zzd1maUag=ksTUTZ3`zVw(@6HL<nDUJ7hhc+ifGguj1mQs$o~^X(=>L9L-zPA?Al@@
z%A*-mDBS1%(5u}~G5cdwT|nQQ(C8E!t-fjvWhtF*^J`RBGFEdnbDo8CHqN*Y4|IpD
zRw=X15$TVFeVwe47<Iq##WI3guFOBHlu?ex7u022anj6pWyKsloIUM2O=OmEAVT)l
zv=ms!Hbzc5mx-TfFvpb{%Uy7<X1K-TyxrB9!rmX=&Tv{1(Y|({)iiIVwe)U6n)Hmm
zv&v)ZnFRX{bD{y~dqLJx@Fg{{iaCQ73k2o~L@M042^Er5xKW5dLHugO0%ETNZt)Z`
zJHI$rKQdnB8q}@hu${(IfeRLY06Kd*<(`Z}w%WL*?n_kfTFBom>(d3><G<dX@B6}y
zm)@Ockx(Mp(@uGkx!9E1*jM3n5ZqE$f(pE8x(f1wz&l)J6?L>BLcVUQep*_8i=?<G
ze$*H@AeX|Wh79i(T;x+Zo|7q0Y}4|6w-kv<EZ(-b8%k!P8b*JQ#irSXa9)1e%Sna1
zSY=6hZE7p<8v#rq5IEi^0+Y6qF<2+?`3|neCO3o)z0zsqUs-koDJ;8NtCL%`LRFN;
zN~;H~l@F+if=3(tgNsQM9A-3_))4SPj~sR#?u)*W%2VN)K+DHDZN%ZW)pLGLJi+@3
z%rcwUb*_Ko+!apY`=@l-s|tfeY)qQxSsFj=n81Il^yD-k13;;O0M5=@);IM++Rd22
zRiS90fQW1<IOV(s*76cw1b=w52u74ppQg4Hl(sT`mu9=7F7Uh7F>wh-^9OqRh}QmD
z=oB&@h%+`~+>E-_icJAE4Bw!1LpK6ri|VG(Kcfe0KP1JSqCcjB9z*TRaK#e(U*Dpe
zXxjkwnBOy2foSSKzZ!tK!k|eD&IJM($^uwJv6n7vLmZIfeV)kLr{D-`N86UM_vf&i
z#+P?vzqE|C%x0VW`9!v!3+JPwM<h_<Nt+M27nkYY>WSrgs$qD@gc)4yi#@)`<3JJ5
z>OEDy&r=^ZQB??F{IHpFCGQ|US`vhd&r?Pokw+p<nmsGIl%}F0G|^M&jX}Xoh4PQ|
zPTo?zL`41sM%3eGmnr!V0_ZI_D3TTbYy>uj#i2_0LFt=Low4vjgm7ynq=zCHm>Hno
zMeF{(Zb`7pVx&VWl>K~2Hus$mm2NA2H{Ve;8MjUr<f-g)Xz=LV61jvC=rQ!$(4`(g
z)Xu>_Y%bwQSzcI-&}Wx7X_06?CnWO#_XS&2WdPve?q~)`mia5N)-o<xqyOO^wCKp$
zR1$bYw;+znadqI|{gocwODn4|EUfL)G|@Co<d*ojfwTlQ6J*`F;!+Is<3HmJpXZl*
zHz}=PwJ?i-tv}?!OCp~w6vhX)GS|e;|7F}vk^$ddhZw)c@kFzoR(}WrUoCOV(mO8o
zV$<tHC_oy(ks<x4Ku0|9hvC#IZs-<6tvAvwjT7R%Ce_Y1f&kvzPk;XRl}&uYfgqP<
zq#BkN8@qp%4@{+a5$||;meryLw{B)aVSLuM0xug9+a9U9fx*N^fdqQ5`vo~i-1!5;
zR09HAlS&o?efjP`yKFhdjorSha7+1d{4K(?VQ7$`*kb;GM*pr<MLF_ZB;N5BQ18kT
zvuW9UO`f0BE$rA(VMKXpXqtCVh9z^fw|~02Zi5jyHx8lV6eKL{OSix<lXuz|VQ5$Z
z`_kKMInQ|I9jgl_P@d2qzQ3~hR)hhC+S@g0Cw~e!augPXn)kMZDeBZ#$KTG7M9I&b
zU6UnJ3dmLCTGc^~eaHhR%r8Ngnl`Pc#W)b`zbG24!!&(Qzfyj_{9`5D#6h{^E~}bx
zzwp~H3}p{|>l}s=DG=Pj+R1^Lq@t-jUCu4L7K6_aIjlQx&JH5h%v>|E9Cr$A=DOzK
z*Z19ZPQ?7p=71Sf)p8d#g?vgt@a${Q@V3dt8E~%FYJ_S<t)n<#BSxPhL9Rg-IDy0Z
z-FDLa0MFKQ@B>jZ2DB%o3Zs=Hu^CQ7Co#f7db6)vxbeRux6;J-Wljytwf%`LXVu#S
zp>6K@z&t1QpWr5IkNv?^4wC+@^eR2?1Dx*k&*V-vsDPpvI(Ytpo=`t@v}-C;L*E-m
z@d(4r-ae<2aSR&iP|f})E^Noh-<Z&eY<91J>K{v)bnZHRsYr4WbQ@&@H<v$EAu1j<
zB~HhajzK%_eE2E)+kN>iB`&}GQuI6J(_?t=&EEr}2S<L%$BD@$&83n{<D8)j?XGxy
z_v!^bMY$G=yqJvt0=>^R&0LeyHBth5U?DoRV|fCk|B6Q4-P<1Vij-ydAt_cvZcdQc
zMuxv9=~MHe0`e&w{veY>#CEHgd~OzRu2^PZ&n~;7AGgVw`XMmwno8pJ%{BsuKNZ>Y
zUUj6GS5K1F+D+02K{)r$sl6}nfFn1Yvbn=1lmW&xY@aSS9ksBg2tS~lHd`dXVuFL8
z!Tz|d8(~K^K)Fcy)eWuW36_ZhudH-RKp(j@x6_4<E$qq&$6-9v#1JJH>MKP=B0E$}
zo8OsI;X{*U8PS2yXeeJyy{y=;aa&K@`UXd)U8xNX{5(vSE9PVDtxLr)ahW3U#Aghr
zuOqLyl183uJ7R&-^VpbcHMO#e_O*GeQm3~)7vhQ~HyK(9a>IkyUrDe6;dQEOtK45+
z0-J^V6X9!O!*m)4Lmj7dpfe5?F5KBi?I`|4I{b@BaoSq_{t%*+g1cdt4eMC1>1-wq
zTBS6UnrN!tT1N2l($ZLgSO_q*2~2(XJ(qp|e%VUzxZEg55|qR1<o~?4J<u)c4~&5N
z0e*ZYAul(;w1PYTCvSG-$X|dTpq7iMDVb6Ct+*vZy%P@NlFl(fnD7@&HTxCYW-le;
zO$+A*UH(Rb=AH`HZ*BYY9X9kvL`Wb;G*d^KGlbalVtDIt90s99&XfuRIgsA;PXc;&
z<<kaX<l(l_Adn_DT8GVd(DCJRyQj1izP0ZbcB+CbWuVSspyZN*O%_*Cr=sAFkhhB4
z<kFG=kcJL8WC2DV_cxkp!wB7rz~ubxSr83@2w;s0WJ`zL$rQ_lc?SeM{Z4rwA0N+&
zzS12cHZT%)cLNN|dwNm^G5V4Td-nWZ!6mXKITj>PhHvxw$lBBi8Y$}d<6x|a3o|iy
zCbY`@m#Jv<jcL|-8)f@rwpZU@$h4@CH?O?Jp}mKVt|e1RbqjEqX{drbudCuZL)ZqR
zaSp4eSMjOndIVw??k%uM?P9xV^k=2RO_jNZIhsMsfl>4E(yUw-F;uk^h#-(704u1u
zUwl_6od)b#clY*0_aJ~FsEeqDot?s^^QITzWB{m#eU5Nwk+*AG)<??p?og0HD&xjY
z4Tu?AX)DZMxCnL*{v>{jplLewoMdOEd{;Bkj^uL<2eArs%d(<2zch2@cuRG8S;Gs}
zak@lzau35#xS$X|X#8H}_Z!)WU4FYT@O^?Y^Y?6c(P_Qg>P%T|MvO$M#)mCNG~(2%
zSHd2Kl&jr{2wie*?T<{m#JK%f?p~aqp@5Mp;QQ;FTa>=N`5%M{*m?s~7(&d)8rwNQ
z5@(*!U3JDUvDeE_M@L71*#Mw6RyX_WF(eTn#P_#BFVq-?f%kQ6m%mj6=c1%JNabA2
zRT=L1!!$MZgEdZxNqhT)L4dOea7O(HLqLrJqteoYd$8DA{KrV(WDd!=;^J1<uZ6bM
zBd~H*`W}pX0thgSWBJz0=?Q*)-p4)-qxp=!FY}&9$hNFOa;FySi7NsM*blM3y8ONL
zg>G#E6DJS{ho}1jK>AMxj-J@pasgEGKfZp2`hN~*!N(DO9R(mY6{;WKzo%?(dVE*d
z<0Jcbiz?XquEc0A^Igp;A>5x&efEcjcLI28n=!}3U38%~v6kgK$k)j7odga$dco|Z
zNFbq9CCyK3F#2<^M`h3CSbC+||NaXEZA>CY|4U{NbNd}~ue(n@iV3&g;2Q}dXo%0>
zqw<XeHVjUBDmoWj<x<o&t{rQtx}qOlg?oK<zq-IZ;N`3O<Pwl-6w==d;H=#*r!H%2
z-Q9O~z!L#jT9lapRy-5Ge;;JJsB~P7Fwe4Urce$@uw_22x^XaZxvh}^O?B|2y^kRp
zN2j8%=E~t6sx#RI;_-(!=XK*382QeOW3mG07Z@Z^WTG*4wRpao+~Ytc8s&ITP;YO|
z@q!%%kE9y2TLfPkGM`woB-MR>`b`z)>{nF(Wf1$~C%vrC%3P903EDL@C<cAF=yso7
z)v)gi!OI-3#%aYlb!YG5RMqkX=l&=KanJSkngL$W@GiRA+Ujn_Mr`FKZ^x!@K;kD&
z0Vp69DzGGTy;p;H5IC2MTN+7;51~|W%A}n$ID*SCtBL#Oz5&Z@=;Ne)`SQNS%oR(g
z0<Me#E@1o9HYIg<I<MZ`nsI{}eAr*lBoZrQg^48<b!2Kt{Dc}63S8g>5*v-dp=Pnx
zzlsTU`RxDRA-*`rx20|BrQDKi9lT#?KdQLZS5+}FGRB541H*Hbgd!rJW~0^dO2rBA
zZya9vMGMXk4*?@bV7Sa1nA&0*v}~9&Qw6M%0ppf%D75mo8zB7$9JVR$JF>EtfbBLp
zHz&ijp1%VL3cB85Cxr3rayWfR^G&u}qEc{b$?(a4#jbzWb1*w@!Ure&*S7!Y3)sac
zA}e9BtP86F?aZ7$pUM`K#AulVHV)HOOa{>yVIE0rGWzP{1M)R{kQ0MVk@(EjejM=3
zf@G7o@3^jYWW-2a^w3e{KZ5u_2zuIAHs>NUc%a&v8Uj%>U?l%jG4#g|z$ULqO*oJc
zHBwMe5HQ3z1Q29~Xj(PlPx7j&`J5Jma_@o~T{c8xccxUnDV74x7yW<*Y5_ag{ptQ)
zFb9622lS=6=#}y=&auE7Hfol)U?v0-j-p=>#M+tB0Li=3joVr15=4oGUJBLS;?2X?
zI1LCypGn88lm`R`2{=1*W99~nfixJ>B!N@2%F}04L@DND%%5l))rEX&qsReODt}FR
zpe!|mHW|(sF(ln6w9>mZ#%GbuF1zr@m@OWE3l4`Gn*a11JiWz8-X;ZV4NioWe_)7!
zdX~DkGhDypX3|D(s+dlT(<0d@=cOa+9|Zk5+5J%6ZydfVMYc}0fc5{OLu`-AoY7kC
zH~78}n%K-98z^6GG~V9DjC(F;$3SGvFF(8nh8rYD#mz1q?Xq$c#{X_8c}7zsd;#iF
zDzcIj)34oI*P*qsYxP|Y9#bHQ)12mtnGDCW*}{-be_*oEDsni#z+fmfOG(o8N56Ef
zbtH@C8qatpFLXqGoq6bvy$4}MrW=5;UX{^h52?2F+Lz!z8xvEV6Bgu17c&93mT(cm
zh}0M*Nn9OE=L6JUUQfk|NdGdK&$~<0B5Vcjja)CP6|PB04J8Eb?qWpCLGHMsG(dl;
z8tBkl1<K7I+)=^>2K+RKMELhKa6ckIp-$fsiRmTG<4ui>JtxtRtVS;{4sLJD{7z!R
zbD!V~k%i>Rbs=L9Jpmb5+(j*eGyVyx+R-uz^y$K>0<x6ykzlrw@^*bmy4}1W#X1vK
zP*z{BK*Pe$E}T#zK6$>ME%Wcs=fKfEmLaHX!xMhvw@Y7r=CA*w?~oT;fbTEFHCb<t
zq?4jV{AY1a&i7}^*+2TarfQK9+@HuJ=0*gX@do{CxE`yx#wmO!G@~AsY55U7n{iD1
zQQ3KbQj1><yK-F^o$*A>PHkdNNFHpSmt@;8l^iT-?4|e3pg3}EQ=rxQX8p^7rTVzX
zhY>zD#H?k^{MqIg@WkqcX1%COP1kxFn#1=xY6kkf#&_u8s{S7V8{~+7^a;4b4K3S>
zU*uYli@lF^ccr}01Z#(EJFAMwo+Fl$X*=uCpiuoJfebTc2hz$2{T+Jg5T37lA|>9K
z<B7%HMx`_dpWJW2sZed2`0LM6|8UQI+?vKwSx=BAdS3$6)-GW)q`T^>ye<=@U>wkZ
z^69U(>uN1}y{O2YZ-!i^)KP0v$6#n8n1f&b<7CZWRM_A6Yq>Y{CAF@fu)pW!1;E4E
zQiv8mQAoNVKVfQ=-k<D8NWyj4#CD<1{`JT<jJ54bEdSs@L06(9|DcJJIb(7mUbnt#
z=HW_IW=u%V1h`u@ckIEOw0ZO8@V$Iy{hu`m-77k7?h@6D!egB8-v1J0r!>FLvmue%
zhK$OhISc$2@hW}P4a#zz!IB^OlcSnj=yzYZMxOg#tnY%wV<}<24DS4+SDpwRwvqqW
z4{!sx);PHcz$WD_7~mKEml^++<$u^9032cgu^XE3bt9krYf&z{e-=qTHO*QGv|Yz1
zIu%ERYRe<f+eG+D|Ip$pN$S(CIJNOt^W(p<vHp|RmsFF08roF*^PoQ|s@>upKB~j~
zxAK?Ue2CLql6bG0X`Jf^oiajnuz76sC70gMmC6Co#9D7nvN}LSu)jyieJxH;g+$)h
z0c6i^;fxBq6nt_3&0u#M8Di!}M`=`(6H{7VbXHj8!rdC|@rNl+sz=m*8?0N?gX;fa
zp;4}%b0mmdcG`KhHa6-kt4djvOqr}-1?~{6a{J;rsC+ribgFWwr*V)#mn32={z7ZG
ziHb36Z^{2P)ma9C?WXW3SI9=xTStHO3o@XiJB8<Y$?#9&Wx4l{eO{8TYT5qn_dg4A
zqo%;Lmvg$uup2Zs{?<{sXkiOJY+_$TY0<U{vRKhro_nmR9NpW{ZWvxd=bYCQChJ@_
zEE*{e5xPlH8ROG71mK`h=@91a+jo!UB*He#Q+q=dOPJ^FpLp|U<);)6;Uj;`{gJSC
zzvz9rPdt`5b9^wH9v}CH*nmo?AByjidTpa0pLIXWgzk!?Nuj<ZNzUHfn7EaeSK%Uo
zl*s#-T*4A8SI&fOI}S0Q%$<e(wbv)buaS=$7QfZNemf0q!dD{q0v>Iy*%#cnC{K`t
zF}ClcqP0p$tC5airuvONB@NAv{db-k!#@&8?m+5-&C&$roLlCou%^}&+}Spj^(Wy_
zi_Oe`A8$k+XJF|cBD3AYfAl!<E&n)zxYjR!XzmB^%uLr!Ml4rPddV!07ZC35P6>zm
zCdpm}UZ7;UQOO0G3DSVgHC8T2LVoU(D2IvMe@Scj?*$)>XyG7xDN7rl9={6gL`qP6
zJX7$OHq^qm8?l0si7JGHi~Xq~+83_PQ_2(U(<i)(wyLt#vYelMn>TAv3yP+c0S^8}
zqTGBR*d5)UXZJOVT5^ko9t%wl&tHhL<&t}56jKq*Zb*lu%f+ZT5rB=LNyu^;gY@fL
z@mJ|S44|F)O)>{!;QnqHMVW?k#z=NL>z-^!2Fy#o*iQ?GzTVY!C*z@M{<0mjkrl}c
z+&m3sm>N4jf=}YC{>Y0)8Mi|JWu5s_CAzowwpFrz+CJ#xKq-<Q#B@}ot7{f>?7ckR
z2{mdrMqfj}Il=gOM6bjRqvW{@H^=pz^>(ZnhZ9)Z-e>z4dLmwJD(&!79snDVfCBs<
zMPFS7#oy>}4Wc*T+y2679?_08Cx}*RhIWAc$c3+oGdFLH%YrP&6ZMJ6EN%2?&s#vm
z-05=!<Q&i-&g<f@ACr=}KT#?;hF&DGqEBd+-{^Z#po2BHJ_(9+)_n5D+SWg@3|G2a
zzWGqB?M$;=QC{SVm*P4Bk&g$S#O*?(I$wub>CjgdjNG%;IN%;kK3C%lsP~V6l{v6s
zRIaL3JGrsVk4mX0W-YI=h{$$y8~dH%vV=8O7xUY9@%~M+@U@JMChTF;{E6EV%Idjb
zxm%Em`&H<gOfyxl<EusUf(xEHg4!H$IpFlCS4}@e#C^@O)y}`<Fzd{KqHj}Lxg<O*
zm%`1H&l7g!IiADTTvX02cx*F4xFdUgGKG%ClMu-UFhpnR3wH;qQU846x(Mr)&@jA%
z?(U1L9-geTkRM={Sl-;I3?%^@5pP#um08ZrS69}M`Lnb5Hf4SYzLHpV7%@n${tG)w
zFy}Ch*8aX#n{AaV;{=h`oP)q`2Jq{Xsg|D+gb~^J1X`DWhoHZO2Am)8MLWuzFAoOB
z6q^Wn<wJRz!;nKIzH;5j6Yr>A2}CtKVbYx$O=%hhLu8S62)}mco!6DGku{rU|B<*v
zZVL?QVF!Y7K(Tc!+>mD3J+h`%+gn*#73lwf6g8D;y)o2kboD^}qCu9+ZKkS(#<3RM
zy)rdO?+jPN(SaAC$G6UuR1J50{cJh&>a0&fax$%JYD`K*){i~=mIF_@PmHdV-G<^l
z;tMCk{NGW?DY87YM`j8h@K6~!rz->flEBfT=3A%5j-`{7q$S`<!AeV2=JG;9m8dG?
zuOZmb)wLJ+O$H7Yvyky4<{(}NNmLH`m?G+%5q?$=Y3C<War+zUGT13uFIpSiu{yH|
zUwMw;)Y-CU651;<0+jmwq1K9u4|b$<dcV5z9p-DLp6u|?sqwhx*%$&S-Bk1X{jpiV
znz@~P1qbpoqa7Ou;rouXK;^)W<iuZ9Ijvz8o8S^lsXm|7NJfVAe{&X6dWwO)bxr%Q
zy7x(52{|A66Ur{QP-DFi$?ODK8eEv2p|d`yMYESVTFLxGTtnjPARLB6r90-{+LpHz
z0Fkb+wJY+&`tBX5R?+i5Up)qZ@}h?%SZ44hnrL{;zetkfa4|7b6-kZ|;EbrM;IrfE
zGg?J*7&!>=17Q`04Ps1O*lBOSR*$^m#jU@BJ(QXjqEoHeelyYCqe5717ZTaL?7Zq_
z!K#aa?g{tsUU^~yhO#vrhc2*f4)S^7_2fS@<d54twu`B=LD>5Tnh7A<qyI|C-!jSk
z@{Rq)RDTpOh&4jUCstIrhU-;9HOFOTigM{*?$~MM9R^@96dK@7jBpQ+)_PQB5b_Ex
zf!GDY)M$51vDvNLz&J<*oZm4ehTfC)yN%ej!2~wzj+nRT-=_1y*Vw<-HN4EJ)PLJo
zH8hlfTk|{fA9t2rJ~H#o&}aX(*aTdkDdP9ooxIV994UG2vT-f;Dl2vA3R}4mf)s?}
zl<B4?OQHdZAdwGxgb?Bwee%u^VxTM^7UrL;wybFZ&5faf<x-(H7w5$5e{Los7lK;$
zjCWdYe{DjpBtW`!@)8}}U-zEF6&DvgTFPzCI|H~ei@W|r<EN|Na{O8KDdu6U=a>I}
zFy11q7CYFg<q%CsMww-4yB3^VP?Ay%ybtg|W6Dl7-vpso>_?fw54{GV_-ub1`G5XK
zlGVZ+)HZN|t7(1O+<w8!#PeO-Zy(IinMQ>|v&vVO&Be!0cWqAmwXV%fCsv>Cmd3}q
zF{xh@J$^Vc1hn8$4OWqv`jYjh-dvcEp#)Q#vzU`H+R+hP@QYE@ij`y)gN7wEM5q`@
zr`1{u`s~T0tR5l4pi@U<mBiX!FE-AzM620?LNZ=m&$Ed4y1F?tmdyzDOnMzsfulMx
zBNJ2sM%h|b5CG(v&WkGK4%MmU^|6+d*vMQ5C^Tnh)oV%|e?z71LWH7`lAIA+2Vf<~
zaNmEckmQEMWHb9tasF4$AFXl4c@Mf{O>98vU${#OF5*pVOuoybwNtrGu8qCjb83ak
zDEpi0Q6ptgh0hEc1E>diLK4aC@Of>|<7jIhg*6W9x2Ix^bM44IfX$1QlAm)*M)=lp
zcZ3{~k1lCoyjJr~4OmI!&sa$#Ghf-bzTH><3x4_kT0Np8m$AqQYA<;PCwgKF{ROX@
znUq)ersLv-ZOCS4jIod^49oTt$_hUV<m0cWht$${xuH{d!Wo;J(pvSRXyK^OpQJ{6
z^Bl1?4^q&fP@S`>(gksxjt5;xyd1;H@yCI!+0#XdYsyp_1k}z#%bA|+>k__&-oKPq
z2d5|zg&I&TEE6vj_9@`v5viq2#^T#dKUX9GR@aS#mrh2{2D68nAUL3~h+y%BB5GT#
zGZ-eywp_Am^_E?#UKBtGMZDLd7J`75hU<jd%(b41gXazBH<_$cFgC_Q5xvZ>^pPrj
zW?KsBAFICWHy_lyz};HrqPR?;?~7}`=`D2ol=!B31$y%eT!WL3ef;c6dy@aN8ygx>
z_a}@&^)WnXssy`pFrKZOd0*X8@T^jSqTsz^^2+ZV_sz0xR<z%7LWF<-!^sU1qJ{N?
zf@VQ~lFG`cwpyv7B;m);`txZ~04vt^eAL3x0sd)|!^nY?*%Ousy)RpSLOgB4PdSgs
zn#tLxvl|BBu6nYXTBqDAp5M_`0Sl(`39wmijzBrGMlMWL>*jbha7K`<u=Hr)&uT5X
z+T6=z*Y?!{U{;=gCjFkHj10XWn3Dk=$^*prYX^r25JMZGBm)RhQi+iI+Wsvo)-+*J
zfD{l-^WxR1D0ytZkf|AmXOg5Ub4s%EEt3}SZle9^T{k)V5~2pLAF(CzL}{8Vm*W%S
zVq%KA+xzpODhI!ka@Q1#*8Rvr%C|C89nt(PQnXAm%mYBn+h@RLiV%3Mfpay(#^@-e
zY(nBCI}536{v&}V@U(`%Ov$dx4|do+N|~ZB#=fziqim?9Bk0GekH{6Q7yCZkprPn|
z%Yd@efi&;}03xMA6;`L;>hnt7LOAUF$B;;=U>rOv$(Fx9pF2Ey9=RW;rDvJL7oaA8
z*nIT7VD>4kw1t0vI|XpQ#(@BBxC7>ls{H)yL{htPK`K|iOd14<wz*1$j5g=U5eY9a
z9Z;2ZvoP*gXyG<L03>?CZWM3cy!*w#4yDz3H{VBO=Pvi}m_N-Q14}5;Ta=Jnk@dGH
z&TQI~G-FnNSxnq#KH31sUdB}I9*FWBH4gj*Me!h6w{D|BZhcmIpFg!s?jkQ?3xN)E
z4uC=^Ky!^ZS&H@U@`*VmGPT`ndGcJn$Ox4N(P-0+WPzLl_XGU`Dk@HeajpBfHl8~q
zJ%U0+s~Fy@QBx!zdu0~91E4pDZ9@&hRp=&|8Hkn|3=zG(_MaBb4U;)lPx0^S)CK#+
zmH=xDSXj@b_h~chwIGLrs##~oc@xxhsPC`}i#O&l%=`NR{eJ%N%!!zh7^RdeS{hlo
z3|SR7zy)J!b~X)8!f5rIp{0_d;=GvItk^r&<_c}VUu7fcZVfQv0Py>FA^H9)ol>1H
zj*d0jqRag;W?$0N=YX)bw_Nv&@^@ZnS7DV>Lvnj3E6AF#gEL&e`l#V#9k}rtN!aV*
z$t?0Q)jXgTU#E@dw%Y!2p#!`+^XJ<z$Hb(bFM>ksg2i#Esn+oD{x-_tL9Vca(WzdN
zZf~hv1Bgr@1*>lVmV{g-t{!AnID!X8MrVkm6`<j&fhG-&Wp4v^uOJ}*M8ND<1Swx|
z5IP`*eF5y65#P+o#=p)2)<3xqKz5iSkbaq)`+ENNs>Xn)>2iB_r}eGUw0yllB%nDZ
zWvV9{FxlyPJgTN%zU~g)-aOK2x3CDUG$+{6t~K_>f#gXUYn`ax_t|2YeX`x457ld7
zK-Of4;15(E-9m$xE`uAF`1p<TUM@KmvE%t_`>SV6J~pyur#}x8lu+%#+Z*XBkO61R
z)~W-bS&YC{m^f!sCG!U^{`K#X<MX9Q^s~RA(n*YD;-p@6GyNY%qI+v|G^<;rRS^-2
zcYbw6;q+SY!3HcNbrjjq@FB80j1P!?+3-KE4NG<f48>AXQg*H(-G6#J+tnT);6A_F
zR-NaHd5QH1(jyXPJ9Q)Om*p?sb`?pYUb&k4dA99Ss8!_o+BB)?xuLm&H3|NG%?}4T
ze*RRPnfX1T>=_Oa<h&3QdxB}*N?Iltz#~?S|0zHGb16}9Kq$K4Bc3dQ==+$T4!+Mm
zC5jc7*_+P@CHC+NJu2!0+~6d@fa~FE7f7KgG{NjxT3vmfR$?zY`08D_-316{F{RT0
z%=}aL0AQQ6y12-I6AL(OPpzydiDKTilW_}LxvgRACy9x+gp3T4!pyfZl<K-<iIgBs
z6__o31aZAN%6*JpoA6c_;+$fs_|1%u1j6-MsrL{+b5zLd#>JkxY>Xttp0cx%+gI0|
zPruD}j@~gcZy-8m0Q3u9n|?MVQ|ZIGODODogIWa1UyT0pNRVQ1fLDR28bGexksEgd
zI2i%Xrj_^OeNu@~AQS^&g+yLQ0jDe=P(r2?z>u5j>wTYJ0d68MJzb_Kd7BbnQnVY0
z1sgd0<!1;vMkNvjSQR3q6G~YrKTt^;ZraCMA6B4_C#L%stJ1|Cwj7&I<(Klf4sg+)
zHztWb!aY?4C!7Qs3p1`})j!+qL`5VhndZ;lp_IZ!dC0V8G2IY2g11Y@^+(z0#?%l@
ze32*Ck%b5XyhXIvfNo0tT3vj1bK~{4rR8-6Fgkw01B==LPB&U*%0(vHdV1Xe1@TEu
zDknSp%(r4!>BY|R`S@?i+uhr%sg4dlK8hw8nQ!Pl@D$&&zI~Mw7p2P)w(bnYbo8rj
ziSPD*W|YU+c4Ok7$Jp^F@mk6yafo4{q-)#bbzdbk3h(VLyQn$wpn$11a60?tpR7F5
zWZWIl(vda4>fO2T!{iVTR2odcsHacgULo)pUw0GzFx(>)ru5w->9}tlWDdVU5LQ+}
zJ|0J90s%sAx_g%=m6BV${}MU?C-RKI)^}0c5Fl&B!aG;&TShL|hYyB8WMU7Htd|Gm
z^KwOs9UUhCq)~nS#ltSdSN~)$F&Z|5%jRl5V~`C3)kzu+IOZrpitR{oA0oPFCiwG_
zAcz%fVHhR1^>$97R34lQ?Ko;aXd;Bw)v0Yh^^MMV$T{7X^l*Lz5b;ZRhH9IL2zuVv
z)PBxZvJ&g3))!ycHGAQSs?TI<A7k`P=Y1XHW>0$jA;uaB286X06-0~2lGsi_WLvwB
z*iX>R)D*EV-s1E$;2T%*Q?LoZ?}9{bem-o8l>c|=m@`@d)<!E50at{fWXft2#vNFe
z-l}U4NTH?88=`Ou_CjLCqu$J$=-A`Oc_dt$Y)FqTuC^1kpm7Q1O4Kj@$a0~zif#=`
zimH>`MX~X5wxS|b?DBidxuH8o?q0BiRZ(LSl9Ms5f%1q^quJ$Y!Ys}tQwLQ`23U$2
z297F7RxA#2P{8Dn{4xR&RNd4KL_1#sHiMdGn__PdYR1OK6z3`V9}wGtXh^`^>}-s0
zqtQ>YGx>jB18OfW*X;2eYinzOhvX>`M)jQQUjMTycsZ>AHsgxyf9aG<CjA1R8~Cg9
zJ&DsHyg&8A3kaJ4onjsti)F;ZZgXc{C#XDAQ>v+Rd;q~gar%mKyx(v)7Wq$I6WhFa
z<RWyeH=1folw~2+O)I_8n3ZS7%e0}X=RpBLzGH`b0<6E}Cn(c7xVVG_GQ2LB$zTC%
z)~x&&BBs{))@?B2$onL)OBp@$!Fv=vv@}<62S(pMfR3fosZzb!BLQ2#z^(cFEaq%*
zO)SgLkR4uhuywAdQz9Ia8yvmno}-)<_bH4c59i@Ti5I-(U2Hl-J4TdE35(`lQvnhf
zr`Q{i6>HARagD~j+OwDoPkpXhvM+Snvt_;?Wi&A6DRWaVm0Tx!Ansv7nMj%CwQ3jT
zDz-(QG&2!jH-@&WKWR*^?_YgoL<j4x$1%CVWQEd&3=wN!sI2JuVp4cc1^9=bJ2I(2
zaT>6W@`g;cd~nR~A7S)=Pym=W*F;Pg3lC}}K1m{qj{J>mE@AENrfCi|trO6NQoHVV
zK;)h1SD~fizK=uIwY!uJ77OJILN2aR!V*{u<@p(x{XGJU>ALyu$R)y#bCIMcw%J$k
zRX^Ju4n;($6i~5O!@Vlx%y|ec7N&b7G?!H(BO)SVnGFzxNaCLp(zqRdfUm4|4&fwW
zxh~$(fGVGGS?rIKgMTRp8+4j~C-Cmu-0WAgX=#GB=oKkworxxg(S2H0&<Dg8OU|8n
z?_f2F+gsWe+zI}H*uqv^xXX3m_Y2d;(6T_91mEU_I;MsTnwdL-7uaN!(3}RSPBGS^
z8~_<6LY{_|f4tpPeBbnhIpgDKZc_{aJ0L%)oGj5rStl(f$m$I{ohPLM#ok(WP&vwr
zZH?RLelw7;cRC=;y?Ghg2AD@o6pgmR(&qO7-KVwZ!6tlVjVc};11G>kV9vV^`=>tf
zcNYbGE>xQX|NezD5@vCACC5bH+BdzKIrsErm5@mOFIi&4-FUBz`gHsIfxF#x(BT#(
z<382+8{MN}mVn#VT!!y78BV%=J=nH^Qyj(l^d|$nK&$)1cs~LpUhjYQ!FOk3it_MY
zpo0CS|0_}`h^Id@`Xn+noicc6?(Eq>ech5gw}xlPw*A*YWaL<P%)4m|0M5cIRp;Hu
zdRg1P*6OFm7BNNm!#T8#$>n|aOG$0(kQqkSQp$yrQ}a8HcDN)fi1>Fe(6g;#z}&WW
zS9W=<rv<@O=iwRo2Ir8L@G>S<|1iM3FG*c9Um3pwR~R#%5v=D1BO!s%*q#<1;gk{;
zH+@IcRyYt<W(e!FtJKuw57EdT;q#=6>e*CdcOBi7h*dVE86}Mdv7QB^R80_~_MQ8#
zy|psa2CFD*bqg4@Vby5>R4a^(o4a+#6ZZ?(;5BX2+Q*@(0@!=E{C($7aYP4|hPhor
za#L9RAX4azhl<2V$&RxCt+VadQ)(9;-m{63;#z*cV4Mj+Ar<7e$*2)06vU|)#!Aqc
zXIW#0D(r<~VqY=L<l2$)vlA~3(5oN2UHsVfx{mMhZbp}P`paJrvP{I!1IVv*uDRwr
z*iSVaGxeT62RSrS-a>q|dT{#?Q$LRrI<(N68g93p{wbfDPXTYDd2RF@Y|$RR*gW5Q
zU6b%O2nAfqT*Adm4r(u@4XKCqwNNf2vV5o9aMEonkocEf@ti)qeKxBEP|Tgeoa%4u
zZd##;fT}ZpB0#jHnQ`TE%dU;cay!&F`u>1<sy^VuRv@I(6v-7l^7JUR7mDs`AS^|i
z2Y-<B=i&LN$pHnSsJ7qavs0WR|GvL^`oWc&trw2dM&tn-p!%hI<o35bCw8LljlThK
z%vci)+QeVG_ISMcG?!cD#MA%%-smVv%gu;8)=_O6Fd57FPk@qPte9^ucLrTk?fnWp
z^MNPU$@*v;ai$p|pq)ZcMF7P5>Q;^HU+ZT#QS%o3u7^JtKL;+yeWpRU%7f6Z;OLbp
zj@sZTKIZ#wvvO8El}EIGwyej_^<}Y#(*PuW+t`f{=msH>#Z(7CSDFlEWL#1rbb-~M
z^q^@{S$Zh=Hn$rl?5Rueaj$6s*V{!mYG*L_7Jgh}<QG9+Mtx}M-;@S_E<N(1Yr%aP
zS#Oil^LivU>`FDY--RbwW&XOs=V{xow1T8q>|J4x%2v6rw=U<Op6XjGrLUdF(o{EX
za}YHf@}#3bd#e4%Wz-jUSP!=Rfgvc#SWX&8Ja7ft=3H-z66^0PXDGJ}K^ESBRn|!B
z0~&a~^4bom5RoNpSkD<wBRdaLYVMzk5B!tiwhakss$;7warq=N(TkPZLo(M8q0GRW
z6V;SCEuaUfjIzC$pcy<t|I0-{@DjPV(~1Igj@q@f+9JTP0wVw7+yJqdH(gJ7{(FG;
z6-I~tb%6%X_f%S$W?3M%^F19c<%<~tjbmAnp`jzJcr;>rVkG{Jr^ykk&G=+x8!Nl&
zaeE^n4rn-<)Vib@FWpvoR&`Uqe=r*57$vp$HlEBUyNoIHW>tns7bmVL@ALoM^t3<t
zE{WO5^_0xGnDv?~NX*}QKQDvZNC++Q;U|kK<SZZ4QBKyikxP@bRQom|-<{Nfl2bgI
zC!@}ItMN>7XK0|wP=8X7BFYWSE*bDMR4910;m1n5w%u9dIAH}@rx4-9jHlwNP4Ptv
znNKNjeUGonSF}<i+7-K$@B;EnZ=IKI+Ndn2;qg)zz<tj<NI$K^)J#B=s{d?(j(i_J
z;ayN5j*sjLo`0kxFoDhD`ud)uY<v(p-|iVJ-O0_AIip()5ElvXH=hCs2<!Xz%Qri~
z2Ic4im^Y88-!3f#QJ1Fl|LvDFN@?NuMB2$OQA-LhHaP%n_i(S;(Ad!l0v$i5jQt8q
z#;muQXXF?`kJ`#apttNiJ9Mqc*<&<<+-Q@`QH?=y0hGlBHPzPfoJ`oO?i`!d{-S7l
zD_Bz7W$1$F+Vq(|KsRffy1fbN+>dNHY4|th)-rWTmOH@w`$n>2BCTe>j7-%uVk<;F
zGxtCCaMNcdGc#Q`j7c64t!+{FWCs|?uJeY#RDL8uHAp0cTMu=*?DR85`$u)yz1~$_
zkHQaX+hzTpBVm6+Dd`4`3vrvChL@ByqY)mkeBU3HwQ}MV2B-2fH5taG$xg1OA}|}W
zw$1CwaIIWaDfwO*VT`!pPLnV(^|wu&6%s~6UtV5-B(IB`o71zivR6O`(+ao|1IfS>
zSs#XmhNA7rjJhlh-MKh;s(!871N*|vxe_=H16KUxy1S6)%lo*Vnu~cC$w}N+8+d;Z
z@q?^H!YBPgdAz4HPqf?-JxxwqjQ|r>9=akmUum*adRO>Bs+TgG?@8{Lk%tkvXSf)?
zQU`rfkt|W5NC3eUb|6<MB&iMq<d>#VH}?(#^2E`74?CaLWQ7}nfcnn*dRM?#do~^j
zU9sQ!oidvPm~Z}DIASDQ?Yi+*htO1n=d)Kjm6CNdF4sD>RN)qGhyfEtz+4ghK4sY7
z*9|gCj%204fuT`9zg=$Ai_=Ut6dW8|W-#I9QoueIYVz^c+e!n#Kg&!KV@MlDdLQJu
zD3r!n{?O2K8rKyLDL`sz!(jqzR>pg+O}bA`PF^%Em|S06$SWxscKBWb3L9x`bxQRO
zOs$aMmzxD33&REvNP!(SGCm7sikUcl+j~E$xuQfBC+qySnFh4ctsg`4LtaG=Qt1%S
z6e=_OLOqgyOg=<SJBx}(-+)C=UsHfp{`&V%*TO;&m@hpT{vp1;{pvD-qgc5!eP?R{
ze!)e~DE<_6oZLECT-!URnN-d;$jok=I*9zPay_BRv^3_9+QSv_L5~KDq{2BkIhDfo
z00WX@uN%OKWXId<6F}SzlsaVWCH2(xM{BMSwT8k~0cGY;SKI8&2;#v-@4yq|_+$&g
ztTdkNksp8vpAh`%`au^8{+Lp!Lzs5}HcK0^H#XL0D5;h93)Td*;Of9lGx$vXJzJ@`
zk^b*{j+0)0mMxB9SFOwM_lMHZ&veJ3e*k>cHvaOQZMut+?0F*O)dEoP9V}DDWSRyG
z@ZT?7djgqkK=3TiSN?;O9Vb_~9dbNKrBen>)ok;jdGupo@^I_tI(k_HNz4$zYf^Y#
zV`y)0!e_Qqw@4WBIQ#7DX40vUF(0F@(eg`TgfQw8C>n<|pEg*N?t1xKLU&nMB}RIY
zPslGc$E+kt0Yji82v+;-MX7B&!`P^C9x?;!29^cH@{p4Si|}9k7bpdg-IroOnra1d
ziUwJS&t32MsXxpgIOe8W&7Sn-<~9FzO4O3U;JT2r&d)w(J{5(vck`=gKx=MEuKySQ
z|KjPY0-|cWu9U<JBi&unB_Q2McS?tVfJk?DBPb2h-QAti-Q8W%{Xc#e|J7X1oHOU_
zXYaLF{2jp0Q_Op+iR?I#&!QU~FlWt*Yll{#zko3aHUpYwM5XcvE<**3=P3l*lTX$P
zK-MN{N<dcBOD+gVQ|wxZgjTHq(<c<H11wyU*UYnO+&Sqi@A_wo)!BmmL0~QRlMyDo
zr-uje>(ty_f;pkA^sr_l^7cdyK%AlD7rpjq_zJC@{WHWAq6f6;jp+HDgf-rL+NOJy
zdH_Fhi%9<)2B2%)ByUYRagNMAWg}m+pW>@=s5udoP(u~c@a3AReHH)<VaiNj6f)Ry
zY}@CMeMzb$E7)?yND<*iQ%he*9#XX<#2o7a!q`WmghghO$+6azDTjC_BM7_%-w<wZ
z=g-Z-q&44`aF@26I{)$>Wx+Q@b=x0#%N%kt70NxW3NQzJ?JF3|NVjt6HPd%K)$!B1
zUg_VIQd9qfkm7;HEhvlZ8$`J>=&Fh7CyAs*S4f#G2T~&kWp$)Mzd=lT%umx1b<y8{
z+Ki|&kMYsrgAKtf4U&}ylg#k4x#m|_IW~TgK*c*AMvJZ!MdQacn<dZ1)d*W8!Tw;N
zqlDeB$gk`_l;8=#<Hy#vA>an1jrJmDJ)F&RiZvYP8dBD4l}5f{Rb^(i*aZx;?|a<L
z>-<}W(L#(AIjSBVWsIOX_-x`|lAoEUl;$@hd*!LI@XOw>5<Dpe)x=C64j`{E>9@5w
z987}J+TZ^wpuTN)4~F(Nm#s*^Nfd0Pca$W=49`7lxrXbo5Y$&*f>Xk((-vkjmUcnk
zlG^*)zz1Ew@^z>PbUn1eBz14j`*PB@b?7K_ykW^&TRD!*bp?CwT92RY5wUzsp3H9k
z5w4cc8v1r`_~@s-u|vPgc9~RMm$XH^ADPIQ9?rHn-;Ayb49F1~>daisl}P~B^}rpR
z02}+JnFNHgccpUMQ=&`h>aP65r(XM$8q&Lj2Uk5j{dCISs9CKmFX1?^1ii1-O<%KU
z!p$M2i`sZga=3+iWkog-<Xj|j2?|*UF>x+@Q>pl3-=JQ05?-_6n!soN($Mq=CX)C)
zD9Or;<1}$h^r3~Xk$qQ!l{s7cp1OG?bU8s^4ruv2dtbk=<TT@VF@OP4HZgl=o#oqP
zjuaru0wQBipMYDPBhc~#7jMW9|1E)+w2`-UYXo7_oHZD({H4prYi|V7zdqe_(~>+q
zn>ERarOEL9!zbC!a!PP=(DB7*V!4-C8sAcrkDe^?JD-k!&SIbsvP60F;D7MBhE~If
zHC)kxX1)`R-GM6JA@;zY#?d;V8D(q@M(01zKUgubN=js1{#7y8z1tZ&OeCyEVju~$
zLlB@Cdmsau58}n^C}kqfSbzkRuz1smf8vFvzG)t)+V~99G31lrU2$xF=<8{S`Itkc
zhXAmzc{%LmFzgBeP_;rWI4X>VwLUMnY$KG~|CSzD6?}*Bvhd#7|M+v`E24PLE;?LM
zPr;e9%Y|c}LbgR=5|v9J@UAB>lxUFBg7|HchWBQ1(L9N>Qn0X`gH*>Ciub^A+{3QB
zC_6j*4pE!k+gMh%3O73UJx*Dgvk3C2pDqUDjcZ8e!tBI*S5FTg$VtM2TCQ3snfCqj
zat6OVAq2i*53`lP(eS{)Gyb>pw<ZGGycIDK!|Nu;mO*Kr|4tkm$M-B>-~UFgGizQF
z-^&jxp6X3zlW!e7!EjyWV6B$xIukL@rypeC(2jh0@k)}-+aaZa(8UcV=(!bUZz(}b
z1EU5Id;wu5UamTj)TCkAiQIkLp}e%fy7S4>y32Ay^w)5ME@=W3*w2+k_@_}c*XIq{
zhf+njR4h`|?b8`>L*&1i-sAB`Mv&^Ms#g0#l{T9M@ZEI6NxH+JTx7U8){x2k4im|#
z)Qt%Uh<uRmS4QDzJHJC#x!Jf_L#yS)FX;-eQwZ7=04D?|NY*GQDDa(=AYMG@ayc#t
zZ6>H#{?w6MXBnn-I~LUN*O+~&skQ|pquc%c+3x*bS6hcFTkbQZXC;CMwoh7C&^%+j
zk0m{7F5|MCK_YdYdDg)^t1w&Xqg2j7CIB;^@R*PfAQE`S^!BTgo5^2Sp;QniIs$=M
z7G1?$sbFW_nBaryFmU+(_48-Q#%c&BJ3A&0&O3y7WoF8EajXDVMn)mYMXIW)0q-bv
z!cgwZEszrcdNUp|Jah7bZY4Cf4OGJ7>FIju?>86YxxE&9e`9%kAQX%gzrIsW*hH?R
z+Qv^!!!0xBsjmF?!c~DLE_Gds48Xd^$BXzI6toyga4;R8xxVCV`-7b}fEWS}xW?~l
zeG9+v*=0k;t;9v+f6|$neO0~!T-p$=5rkZeFn>=^PmqELV!ALHh})$qyFpM_N5?v_
zhM1XYPm}u(0iod`93;;^;cuPaPxFFE4-h4zzyqxE!Flz%Vplk(LPN-E)DwV+aeQ*(
z+CoO3Mabu3vxuRtS2l|PvLODMwP(A3i#r6#&)Fb7R?zF|S1n*bTI<~MS$&17km!)M
zD#rZNQcsGs!h3o>fW!p7&(h1yy*UYk-xvGp_Sl3;GpICi#C#nGvpPGu?V&-De*j-H
z1hAe%X|qt8eBIKP5t3cI9Yu{6L_`I<1Kt`YxnCif&@bB<pa(=mO!-p9FOvBd`;eoz
z^Q`^#rTMHa+W_SBmLH%wAI}TJSz@Mvl#*+Jds4qvnX&=#3aFh>GdPFd@T)mR{N|H6
zz{8%4m)9HQoB)e_d>27JzNg>LtFDtLuU;l}#5qWAAb|&HEWDov&d<l@r?~geaO%tw
zpgJ~py{hXQp|+*B#<G9Nh=c9lmI)^MQ8HxQ@TK+12gBhjT7T-z(5-J3d~?TS`we8B
zug>phvfs*{Bnloi_T0DSe@vEYMWu@W#%vR{*1BE(hS$c0jO1-TF)3#U?dc0!D^T)G
zN>Zl;DHhN|AU0<Ilq}FPg-_Qs7+6@!i7vy!Vc**?PzecTZBe~ls<^1BO@Z<VrW^#s
zo*r-)z-%#Kha>UP1}SfB?4^63?(fr}Pc%6IN07eWyu*@ZlnO3zUZk{f9()UML?;9M
z)M%_fu>3QKYNt!q6XT_g^0uMaoh*FyVmr_pcvOlEj-5<#LdM~ySJsJSsA{)`wSPA^
zmQ7Hs;bYl)#SaTajSvczD73a{F^foW1|Je5yAH&NO2Y?|Jq7Lvz|JBVLbM3v6UGo?
zE}M?UMII2n`6{lwH81LpJpQlbw6;Ef?7vO3RYMh%mL3ETR)Ibqx7{|_BRv5{M$=@2
z^rZs?8#{L71!@~+){}tAFK$twYP^ZB8BKYsr`gO6*Dps(TWP6w2g2@PX{BoW2Z}}`
zrXd@rp$Xr_BxrUe6kjs`#f5OP-SE?A?P2b=R7GUh;wKhoXpDei%XR=oC~pM~R_e&~
zTvN34gi^BnOv>BO#0wnWHn6q505w}r<&wq){GyP|%#DR=Gr{MpDUkL2E+cvq2Wy1?
z1q;a#<g79>GGb0+hME0|xxTpp2`Y-K1@FKDGs!4jjQn7Ke;%l$uH`>}#>3#6UtSh4
z?;`Z+1yzA_WiuJr#}wMybx9LCxk5S2&rO=KPNfEYKPfvr*R((!a6MOeFi<n0IYJIC
z=X9P}uetr&9TBSzh3&f}p1BrnwrH3hLbgpFy7NQQ6&b99VI#l(Q3f5xgi4nVW}m0b
zH21g)zAB{c@wEBB*QM{PE%&kzP7V&wyNzHdsP;=52}<-xu-RKz?b;nl0qNkxuXC2=
zt?J6klR#DBEBm*fcU)OT<s-;L0d4m)$?IJ|2u-rH^KvJU#X8yA>V}2f{3o8h_Ka^}
zt4>YK1R<qhaDVn;`KU=dMmI&kR0``Vpft<O`-n&MSv(d!kmo<L)WEjs>^WX^s5SRL
z3$1|SWl3qfYMkxejm(4wunK$&fjmKbHcaEWBKzwb3$C4UW86)S7hvXa4J8_v_$6at
zLG$>d*h=l*Z+{S2E&X6^`Te^W54^Uvw&25lcK)~OR6*~Tj&en`O%MbAb`=vT5c8#}
zxw#PRTZfqEUMmBb11CVG_^hpM-`HMW@x{O(`lpCW-*&!zG-(L1BPvpcIwC^y!$g9b
zU0b3%tm7^rO0<eAQK>HTA-sXLwTplX8n485@Jx=$sV52Aqr2TV(mr3}t)3%HV0Zp%
zGpb^pxKQD{i}IoBcs}g|0pd4=pVQvTWWsS<^p&4w3$`gmWyv~zMrGCaZJ`*-if??u
zSNv1pI59dZkE9GDniRa=icj=&<OrPi5m+m-5?%99P;B|X6Tod?0Qt)tz#qqVtVFZ2
z3oKxNn0D>&f0cXy&9UTy#Oq64J&q<*bqDW-0RY4KK9u};5V0U6)^;&~HrNoJDwh>&
zb@+yGq30L~2v*(p&ceD+ob>D8kG}COJ_9`$p-W5*yLDn_gk|>osGICa|HDFlA$?+(
zWThH1&4q$7)vFhLk!fMJkuMC6s`CiWr9vu%>>M1wLBJGH<^YY9XlsUu_~8hL<?PdC
z`m2WryEVw*oK69&dsWo@?CffgJ5;~wc67d>KUPwz+kEqHQw$y^(9u3~h0rrG<-LHc
zvAkEt*zOG=Xd~G{0Zih2kJf^^xAUJcLCH!|GBKOu6}BK03;ph;L2~RGPj6qs<w%I{
z#(8{Rn*4_xH%+PW@99-H7c^I~yuPXJ1*<*Ek6-}01vr%U@mY)sa6Ya>*CdD&z$3=V
zf>oL_su<g#4J@-Qz`P9_<0CNAm9lNvR58H;T3TA#tB@#^;dQk(J$<#E;^_EeTMQo`
zpT>G)V`Dmet;bna6&D@7%jkK|WO;552Os}|cZBv=>(k#M5a2fUA3V<JT6Em!H$@GA
za`<oREAe&|9`fZ@KVLL^gwNkLf)fh<=xEI@vb6q?L!||~*Vxgt5~gMEtBuGt-?Ke9
z-N?(*C{j%!(rRcrMV<U{IE13LQRBx%pYV0b(h|9VAe^z<FXRtd?N}(<fe-TUx$cLH
zl1hGwkO9KgPxZs_yy8!AaUj;oQ;(~`aHSLe)>kec^i14V%Z};kfpY>9ZB64-eK7mu
zmgQ!jDTz{i?fKES<W5EG(!2-iM@d?QbS^f+fx}AVyD%nqckunuxU=@(s~MaXq2V$x
zr(-4&LT3m?$jFk+Tx5_BU+}Q&6JjrMRT;5J6!b&Q%BaxJ6;|N~I$z#Mq-6u`h{x^W
zA%kvL4O1JC3W7iUUoJi9(br!&-`bs@XIASGcPJyDyOIMyOPafC-74hs_UI@O4t$5-
zXW4g9tW&^7;nyb8AgqB=XN%f}=T6G+{13Wj#Z}SzEcXH}EL04fO-dXZ_EAp}GU?C9
zxCYc{;^*A=zbPm`$kBEwqo5_Qx&*$*qGqGk6TguDz~F>L3tbWd#4U7mmpk`Tq7R5w
zZ3WeKa8pgrAe)`yr33emu!^yzHi5xyAGmF_+IOc)jj-w>m;aNn3L&kF>4!!+XKWqV
zNQj3&K7~i8hA%dC;}AeO(|`Xu^x4}F(O;DUJa+<|eoJHVmvBi+|EWoxBowo?O6B2C
zXh*#r{^XC1@x=3XridREm<Ck{r}cPD%!)59EouMn%i$lj_eOm@-bnQNzX9LBbl%IY
z{vShR9s#_^)EU+rheRiO#wR|a$8I=JkLJ+#2!x}u@~FashXakz9V?~`yrgE=U%g^2
zjJlCCqiv%Y)zon&`-LqYuqt{?RVwR@epY^3SCtImdL<ZQr&5Abq3fmZ$f91KQO_`%
z!HM*bs4}w~@8`Yi5Qx?;@s+V6Np{X<dpFx?kX~)5Bu_eafD?rIKzxy#X;!%KcQ>N?
zXPTGfiwcC-iQ3;gUc+&ske)9FQB)M%0nFUbEAJf<#lSl^BWx&F)$`;oGH7<CN_En=
zd6Si<k7^|)Tlh28|NU71rBe*9Pj@0<iUCf8h=>TtqQVBr^I!bl=h?FG(3zJq{rS!K
z&m&<N?%s3E%3(Cbr?!!Nn);>86;mj<mX@O|Wgita+f-(aqpk#9GaMPKEQSg~*0JK$
zZx%<e!1g8D!k`;j>MtR(cxL_nc1Ci0Og!5yk?a`0LZBX_Z9)Iup>YuMz*+$PMOU}W
z=DVD$F=BU*39bTFI?y3L0JaPN{TG{sDie`Ru#*`cp-p?7((<8#PFSp)x@`C6B~pH*
zbTeJyv3-}(9aRNs?sXWxoc@NFx%U}E31L9+eLb~dK8H<xh+=A!0gf-5h}UOue^|&8
zJZ;FjEOLMOQ$|T!m#I*%P$pRq|2m-7YN`CA5V|VM2oG-!^8VaGJBo~g0{B6^yu30p
zGR%4{cc8m1`&=F5OZe^U*ZYU?`!_Otp12j$b@?Va`6jSAHv{g9ru8T}sD$PI&Ie23
z>{npq%hMM8;}}%?9zN#J*$9`2!8%3A({D34lUiiFQv3xL3Mqu+=9;q~NP%6#7Hl}T
zq0FPeTWVmhlX%d9S@hYIO)M{Ef2fOG@bLrui9QXaNVRBA&;10%BDTA(jg9>XLw$VM
zA04=sh5T*s^M19Y>p%z3k4k68iLIdIiOTPpuXWRC+ZH}zu5Qb;_-^sZj55z3wj$5E
z=*gA)WA&#X%T~XR_z#0wG0NU@4oAbTx*OKBp9uUWRF%*|`Mmkx6rvLIy~n(d?5#I&
zCt@`3#c1mfzn1jiGU8&!gkSh}6M{xjR8-`3*(OAqv6@dVmS>CFT`{Ns9Pogl8d^5@
zM#joaEH-UvAGy%8>9FZKA4c9l%vKTcB%}XVwTZUV%E}1pwXgk{`$t$mg&Q)O(isD{
zhgXizR#qpwBMr(@W|@)_6p!m7I^<cu{0|>%zPv9y84J)66JVy^GqQbDEH6hShM7fG
zL`-W%RM3|q7yaIW6Xp-t3p)w=5D*zR@(9AEF#k0x=4E3az_AT~rjVH>qWJGIDjT#8
zdS99sTUgg>ji7Ec{k_sk!kqUwE!R}0QONGZQdCpoy=2vko<$kn9jqajU*71Fuk_9^
zUfl@}^UZUbEV(L%n?BaV{P?i2a;w>H)wEKuqV&LC|9O+pHAZ*eV?>B@aq0QpndHK_
zQ*Zm%F?fJfz5sDKAms?8=}t{efqY3p5P$YS{{a``_1yau^wvNd4m}%6#>#%t^@v(T
z_ORahsqsZG`h&thn0$o~pG)%j;BR(e2FffOPJI@Y8&Ry;Z=h#8-gnrI_$==3KJswg
zkgCqUr5xkdEo=5+gpd+Z1kFQzw<=UDZu`<$Oo7(T9a?X@BV>@h$v_DMu@e;&Q=c8;
z<mf1u$l@FJ@^U}s9R-l_f4hYv@7KpQd~SYLT*&CNxz%?_g>fCI&1h6SXQkb+HNTO~
z_7*s?eUaCcQ-7W5`*>~fuxJO7*p){HIEvw`kUQrSb*a+Wj6H209&_{~A{&oZoHW?w
z-rl0yi*iVP69P!NgZ;UoEX6H9_2sg}Ce1xzQVqfN9btLvP(TxE2@N2Zcj6s?s>L~f
z0w@#6Ap?PY2~km_05;Us(GepbVfgbd$lucOR$Ky#w|Q}SocorbgKbOZB^M~?r4R02
z7kJ-Q(3G9{rWonHgu1fsh+qDd^_F56Fjk0Y{a9h#Sw<rcmm?EH4jrZnhj~D*97iYT
z_KU9dwyL?JnPF*Yc=W}r7uEfZndzC>Xnp$H7q;Q$#@o^Rkw*KlOX2nw8V1ivi_;mG
zlP>-M8%6D1cgozQmA*5&O6^fL3!BedjI$Lv%OlWxNm#d6)5r5r_955(K8DGeVduh3
z6NIt)_?!BIlnAp>f(7S@4=hKNdivx<2L|kUuu+d_=EU*5J(>b|aDz5nE;=JD9V8##
zDGaMvWmh@Ji)%CL+17q!$ja8OkTslGTk&?vgJUK9Hd2J5<k)*g=b9%FiPkV4_E`rm
zS!#Y~Q{?)Z>MrXq@`r~dVrI=TvWQ-Eva>;X!<@zlRX>LvNxpp2D|}L|Zp&bdJ!ul$
z=bfXTqrOhSR@&Lw;jjH5C`dBQw^Wdb``g*6tEAx#abaEUoIc$^2R(}QPv;v)eg^t@
zas;C%e^!!@FgLbV-1~I?`ghv8Hiw>;Et5UIrs&eR$kUVo508j#Oe-6dA1~sexo8wz
z2Y0yXLr-}IC8;JKZ2p)fN4)#=CZs_>xl%9LTj=e7{akl;5Re4=FMiR$?tBwY!Cr;o
zL$E=b#nQw1^|rQ)+6jYPQQeZStsNr{)n2@aM|CW+xy)GL?(>ODZ|+Sz$I6N)B>xx8
zSM(bfw=BV|`=u4SADQbFA$u8wfqc|+hPhlLGN0&tFRe1y*G$k^TJ&<s51uI|Yi8Kc
zj3fM+H#0OD>Hx^ZW`dMRcy-yBTj*<2y0lY0T_IQNxg#qU@}FL)3eOQu(<Li}tR>f*
zo|z^A7qWRs2ZJa^X3w6}kSk2DE-2IZpJEWkSAz_<M<>?E{V-gH*Z)LWF=6c$qeM6(
zEt@4%No48J$!#`S_ea%*?8;I4R63+C^<zV?I5t3z@MS|nSZFBXMJYZ6KKqRFWH~cx
zL!4Ei)^mvCz6w!KE#d-!Z1`$4gsuA{v6ZCu<_R4}_P1I6wa}@7B=NY$rCMJ$#KN{V
zT_z@2{i_RHq)l!b>o?iIEuV{X*pp=`W2G}fB5i`mxEMvWV2JxvVq5D78R+QjTxftH
z#`yHKgNv+=uI?0q{4zT_WcLeBnU-&^S@@!S@irHOidBBVYCVqwQM+CjUqA&}@yno-
zhdur~bD0)h6T67k6zjcpXRmQtk;I|Wg-X%s8S!(+U?UpLF5$+kJWiX@QD2MG7{ov$
zPE0Nhtz9VjaMVM7SLM^R=JNOO!Q0D?_9rJNKJ2DkfX8p1%3;hAgwxA6<795SQs|g)
zend4aim<?5cfjtU5{&Y{ue5;v`QO-9gVEazsd@KH)8JPAjW=GHqM0<nadBL?yE64Z
z4y-?aC=l`q{=Nt=V!|l?O3eN15=MPCm*g)rV8b=PB=h<=yb)Iq^HO!m$`NxWiObE|
zm{=qPX;7F6_*uf6a5WE(haF<vE^hG1>`)t$3sGmPk`^q~hOfKHRtXx>%$In<t@19b
z`)LK*NN^`9O49$5P*y~OvY7gCf!V#CQIIV<KF^h<pP0Y5E0-cf5DUn#Ma6>8jU#=1
zrVm#K26l?%zAyJL-mFhTK_aM2s!PjIFN*_<ch~5<WB~;Y*0z?zR+K^h^&Dk0eH%M;
z6dOn)yB3>DK8Rm_RWP$vtqg|U9ZeF+vt_|xx}Pa}V)q+iTM+z23}g)bS*%Y{V(PO<
zX)5&r#7y0y#8;D0!}A^=<JZ=Fy1LY9_nKO)6+(e`%Ll`ulf~hbbZuLk7eKlA`NeQP
z6GPyYlCH|}N?xE!nDFrdCg5v}-+@J9iF9w2>V|lUDx%sI*<I~ngB<0a>VL0+{^8ix
zA+E7{1GyM+t?<8OsgRJ~+M=STCY~%TNoSeci#QwabR%nh<(NN}sqas$Vqzq6umqxC
z8Iq#mlvBheA0=|li@%WfS!y~Lo+kAN5A9?<0`7e0_WGD0qt*AmV;H>M!aAh|)|r6G
zPyK}n=intk8AjobG$!xq=MJeamaTi4iJ$k-c~4i3%JnN64QI~n6V!F)!}q4C6Lv%<
zlqW(=>_B@fY{^IWFWYEb-Ka>~jQ+lZBrhy!o2Rn0%@@}jq=U^L*E~LOa^4g^Q&CaT
zz!MM<AbuZX;cjr*i0^tdNo-xhnKnqVAi%mdxru{wVRTzZ6Pedtv8vzaV<qfaQp;Ps
z{ay)k8fUNcD>^wAX0$p&NO7gpf7vdpJ=RvUM6dd}1-UnvC8XM;(P;=3qWcM3ILGGr
zJJ-vbCOGxxWY6^m5-QJz>2$a$1yf)xGL^4j#)A8!+ZJ%b0auh!<c@b0{Fu-8Hk-P`
z%5$4B7E2EGW~L6b@~r_92xR2qvR*8GV`{s!<sv$|CN%dP1Mo!WX~jQeMQDZ0(;roo
zlN}vHcS(akBpSZWQ^8t=$%5?<c65jT@&!%uO(28+rIoN{Ld|<bi*qkeE=?VsC*bZa
z2birLzA&byrV>$vNqnw<!DQ)qd#bxh@Ie;jMVH`sP#}E>>@#>}6Ng(Ou$^qhbHwVW
zyi5#6&DyLTs-`Op`(J^iFuk-Mnd>TFKT97Qwe64<DXbBeixfTHi|CBm>AAaID5Cdp
z#KbbZ;!az_78_285r;+obLKsnOkiMpejT0y))VtJ=2MSn?Sdej*m=#<`DC&7{{9}x
ze1L`fYk~LZ$P9clngokNEs+7qnN^^%k|3xSN>X;x5&Xd#no}zO{+`_%d<rLm(_O^u
z3TXCUHl)1GN;2V+#QgZ5#TLyR15)szI+os$BXDfHpTr3p;Yq<7&dRj3i7Y}5BA)Z~
zzx(TgBoD(Y3t3#!t*hx^+iT&Q;UpTWjpceJbeAK19sqUu+n?_?YJEeVJG79xFSE<a
zj@{+nY>xbk=%*Mg#=;*Em5|tf^KnO~$jrk9f5OhHxh2$xn&81&LsZs8%^0}l{gl<z
zD(PItS5Q%o6nzuXJK;apc{JyzkjBz|-jtZXII2*@KJS;E)=v?jIE+K!Wl4&*!FHd%
zNxE9brnY)P6=euJtocSxMBdQQKN;VUdSi=XNUCAN(pB@Vh;;}0OD&8aDnDwxMR0i9
z1Hzxjc7%NQ2!Es);0E2%Ap|C#t@m)6kWb>&#|yck8;<qmU0gFxqOScLYv0a~X!t}B
zlx%f8D{Bkl-arOzoZ=1!3Z=P46ytarr4m#Bo5olv$?qqp6F<kG2$y{P7G-<33ft9%
ze(CyMJ78iDymaQiL=MKjvR!H(GI}5W9rPXmA+>J1-{n1Pxdcse6p7%!`m^6o;(3Ku
zK{)*9kAA%tR`(IHTBWR12b+wc{yWRd`nOz%tIctc<}yRCMV9UgLT?@Q`?MW{oMa7S
zX*l_3iZ4~&UV`GoIpT6%ExP%LHbZ!AHbX;45MoN|mW#rez^}UehxAmsLSLkXVZWtK
zz1=ve_WtZeMChx+o_X$wB(VWHA&}L~q|*@HX+X&1@Mm}!*)Id6&U;;sd<0V}c^hl@
zwt>0S&RILHv5T&5ZMYJdrdJO3e^f)%7VC3Lwf7^~mX+mozxG{lp3%BHDW{_0$+)~e
zhikeeD^So}n@q6UYvF4yq-jS)BuYua>g+GX?3jp6A*-RXqHb)2g!H#N3B8mml2alg
zqH)RLG}Vew8_`#wA#JaNKM5dT*4Aej<?t*XAX^^17Hw_sv17n?iJXE0gn1MHpb)K}
z;NO4ASI(L^aJWJqQ<^in(6(;n&@@5ZSs&N>YG?Ns1VD?EXPNzF;hvwK=JY&dLoT<>
z+(Wa$GS-mNQ7Nv~xZP*&XzLM*S5WW4)6scJn~7f-wRJt^>f2rB-JSleCOq7dyBD`}
zHVZRIxji>MeX<lR)W93YSHB87^^8nEjuzYIL*+-1?^2e6M*NGNW|$?GIT{IT)<E3W
zraK75pZ*s}VD@HW6B-8!DUdeb3J@y=!Dr0e?e&+dt1FO}tf{7U4Hhe_ZJv+g7eEgq
zber)w@fuG2p`p5@<QRln3y+;2i<`4A0>>W!8ID^K5&g~#Sn?1cS<#nw`5tCyhkE%F
z1Bc<coXhCj>-YQ7jCulQ_kh)za@G@u-F(JY75t1-g@`%^ltEeU@N9aSP-D%RYuK*5
z(`Qj?&(Gp)(K$2Iv1mB?s1WIU!lJsL708L0B?#owcDp<!luDZ%TyywH4{FqkvTlUp
zwiRV%klJ=2a=GS$w+1onAQ&04XabU$Rg{(SmtR521dBjq`L2J{RRIc(cjxHn%hTl;
zeE!Z%u?DcJo9pzCi;FAOiw-B|;8eN;#deTby+8tw&U61rO-F~XN8$}P4Q3d%*EvPd
z+_eLw{^M=$jUHsU_FGw3F;9^X1>ZT`%^JxLoU5}A*(dE;s@PBI=bhF}wLRNAxVjYI
z4sT4cv5r4N5ArIF&uccD{lMH)`;9Y0f;khX#6H)-@Ro~+eDr2Z>s2~y5OXq1Ci+!2
zK*rUR3-s)^=RhVVY8e+98EM^m9k0o7dU7I8CY~?orl1Jy-KKu!_fOBw34#>r^--|t
zzQUbe7CnP|E^fkkpQYdCai0`?;7W@*2&~lJT?s0r3q)3d(C$K(4fBBD0bx}|NSIh+
zDbL%-p7aT){u@2{l19yVwI54Ij8%^qM6|OrflbBD%q1OE6bsdhRn-*-yG&+JyXe(r
z6i(rAG1wBFdN6K!NywDu)ipV5js8beddk5GM7zTEf>6`VzAbFM3*0sz%%yLo!&6Dy
z@euOE_mBK~IzbPn2Y%Yb(r+Q3tWrP{0FTs)^BNJ)+uCbIj|>9vu7fFna8VE_<9`19
z3AP4T<$CC(`btWmq*36+Bqk=_U+#I|ESOo&mK0L71Pz67aqWb2UkMCjj~RD;kSFTX
zd`(~2@zoDgQiXV$_ZS<d@>%%s*f>!$9lNM&vuS@_r7yMBQAZd3Xl9x5iq7e7e*5>o
zABoVp;(=CHwaw037U*py#Qey&Y5YBMEpV#oLUM3~;h%q2;o?b&X?dYEy<l{a*A!uv
z5vFolC3pW~>&e)&vpJZ|O-oDrvHFD)J@%`doLpdFAk9`#6u0-w6Nsdrhy+>Y+S;#$
zn$+^CiRtNUlaqcgaf){V<*Hzn?c&1aLm}V+G17LUWh^`&cZ#@fZ#5Mv5sX@ssChA4
zSc<yH!k`Sg%1iB|Xn&|)RPMoUc*^>5<Pf&<ZB<`E#|=X)Z2N4&7+W3sTGnxX@n++`
zqCLWYKjQEtK3J*d<C~fH$`&wlyB}>IIR0J3#(%o6Xb090?l$LO=XIu1d1$NP^~WpB
zzaaEuK;7x-_t>xCW#&MfK$+$vTrLNHQf_YG_f*h;OczS9s*uO*M}A!WYHdwYL#|YO
zeo=gPf1fEvyqv~vuB@j=e8APPuCY_4JZy)GwWDFJA830(^=>M}AYbh1VbCB`RGaFi
z@O?zvN6ad?p07hPf(Pps<t7Ar-OZ>QDk4(DOv)eNr31s7nk<}7W07FFe!uPziUp<<
zNqm@xlEm#zcWVPdffXXSX77J>v9X2J++aJL@CN9~jqeC);as1eYXV~Ax)X?u03!{&
z_T5SS8Uz~C(=lDZ)9ve5nkFJZIl3GZeAhPouI~+mr*$~x<2$&uEsaLc#~V=e1Y9Vk
zxEq{i(uQS+RdsY7my+*sgCF5^WzVZZy2{HLJ~pTU3O?ziX&VPoFjCjpOnh2<N+m5<
z(4CrVe}W#l!wO7{^d-<t%XI$j*{`TSYYo+iN@(9Er5mQtWAKK@Gw?hA>`~szSI_u`
z_6xH@i_?jA&W{i{cyZfv@SGI5-d)e>XVZv%q@&;YIAmI#FPoH=nfcEVZcN~jfsM`X
zd>w?KWIpoq^WUDWy-q0zhW6V1K&x3$$(LWLw@yz;nA6brbOa&qo&MQbSpr~}n|T&d
zkg>b*zbYUC4l6(>5O{dsH--_-T~PJx@=caaeeUGo`1vINd45%<MqxJ)L-_7`rk8(J
z9C2O{P5$WjqRbMj;E8>$iBA<wZ4pd6)8S}QgY$CII`=Y0f|6a=r?)>t+}$jkj(r*%
za%k`%@VOpLYTBO_e<BiUp^Bw?|1xHkiQSb#>Q-F*Nm?!|+{rH@p+p!F8R_}dyb+y(
z6et479Dt#L7Y0uldv|vSu4Q3qNl;HDVc<)aclfEq>)953ws-O47vE9v{)>lW%cb$%
z17X+GI#?XKrTGjCf0IGAn{PRIU5z?v%@M~=DR$eK)6kf<{t|>~<h5w&UOt;+)PD0<
z#|qDOKARh+-;tO&4zKc<eIgml&%Q3zAdbFc<d|yG@c^BXMNx?`wz{EFi}=&T+Itv<
zoMEQ6z3)}HL6Vj~2?P1>npz;g{N)n->EwQYy0Txf)pT8P8+VJ#Im*g+WZ>tAt{`yO
zUV8UYl63@LIaR&5^#$F_bH|J@X+6&O-rQedg$B8|&1eGLO}~xEK?INXwZLaD>G?4-
z7kuXxDaNT<m?6rXiK2Sz;J0r>@D}!vc>mI^$YYGFrvD%a>Ud8f&QrZjbKFrmdCS86
zhoNM@H9;^-->0_Ql*(kh3A0v|q@HG=OSrf?f$+F8;x3KTl|I4W1XrZnKT5QH6Bh^7
zO@OQm%3Uz^Yca2R+I#|?OIh7;3##sVT*Og?w7|YltA6pDUpCd)&3SkQ?^~BfA`g#b
zhX1r>=x@)^a9tQ+>3L7G9B;cHZ#G)R-S~dT6%gA<n8a>jqmG;*p@f6~&7{l0>%Wt5
zyXLhlanNNq5i~hEf8&o9#mtg01ua*}e2{h+k8VjI1J3`yax%O)p+^yPd?Y#qu_8~h
zrxB8F8QxF@!7^P2{~)~UTMU*eX7ohiBPD&R?`82&^xLe+XcmYLC)x-8#m0mlR@@=3
zw5v0h(?^zzEPP}?7&V_4q@Et=7-`lnVQ=Jp?mO@6uS8h>_wdTeDwRYDmVGE<<@>O<
ztN;}ssQI&2>9;O?{O5_Sjft%t@s2-jVz9R=g-t1FWcqI_B=Lw^=y_dFmvXK+qnARF
zYlL4uY&duh$ME+FE8DOzu-|kDSRpgFTIK2uct)(OB)q8T`e}RU2w(ZjJXY%O7Ri=e
zaXb9Rh1BiMtSue|9Ytg%)Kea-M%<dZOL_!`VknN`^n6jbUZT#Nx*%Ri_;2d++pDAY
zc-%yW;a33+hi=9BcY$SoTdF3g@qxL0T#`r->Q%Y5ssomw1c8bfK8i-%r6r|@zwFgG
z<>W?jc-NdMOvYY|@m<&&Cdc#y0vg&k99MT`y&5DU&`N)F<Up<jKlMk%kA!=i&Tp<m
zlv-Mwnf3eJzr6afhzO`z7iW-)zipAkJ87Y(0~JaLr^rX#=E=+TS`W?COR*dVo}QlG
z(}Otofdm)VhZG_Df=~ed(k@(sVaazRSAno5JhFqUP0BIPRhvk!aWh4RS2o^Ie=~Ac
ztooVAos(Zf`)`uYE%zrEg(BV~*?B9~z~Q~RjD2Z)jS5xj6x~+O=0xdU$1ktSJ&YXu
zhs6dwYH!;}m2Kmh|Dh_x&~YE03e~$7Ocgu61^CO6ch33xd;7OosKq(jo0WH;PnyX$
zu3Yub-%A@edYTC6(L&LVc;=fLwr}<y?ft$V2*8DqEB9Q$TMZYR@=+x%YApT=!(vM}
z^rf+XX4T`>%!8d-Yai-`n~NcesyMs5`dmDalL1f5w+USuvzIm8-lQ-HJ;xFouNSkY
zchsBl{rIS(6O<zF=OcZ8Q&K7J`H6|=Zs-44TWWqM{Ydz)kGlc6zHyIwU2P~)ZjbmL
z+ub1Z+hmHJ7LsAfZ@nBw=CKdE6Bm(Z^tT~Q?H&?JDor!YMT-w%$dtZ~%lKGWwhHgc
zD0ou~XU&R7LodKYo)b+oav5WY@QonFh%!wjdR7BP!i|VaGJAqrHj%}0mH`VZ?%!nI
z8-_*qud;J>7%99Q;~+Dic#-u?@Vr@C;?Cc7zkREBh8vB$`dr9FY{vPx5IYsjngV`~
z=Un<(@ll_nvSUu{m(-?`0yC6NM7CMN4QJ&qq=+Ydk)z=_Si{1xe`N8+jpG^$r9TnJ
zC=<AhN<Ue`=P)28GK$dz0?12-UVlLDf#c%W!06~IAUc5)7_3}gK}SC7`madi2EBh;
zyTw&U_<I}1U3?aLa?*eo6x6h}k6hvWGfk4T@2SQqd`#tt^~L6r+zPL0t(6ge4f9K?
z)8YxoCnS&A<koJ>QT;eyN|y8=$pKlT@=u?f63O3XbG5>qvxJ#7y+fyJ9t_9DmhfE=
z3iu3vFB`W?;u&y#w970#>8Rv7B=}6?{^mn972S5)s8R$>It<}v`RCQX@@CLvNm=%`
zaR%JJSei`W&Z&9F)XCN%w0biFhZRzB^7u63uj=z=M2Nu&ty0xn>U@gt1SmM`UjD>r
z{K#|>hh%jjHOyPqkql2~Gy5Alqnu5;tzuz0!1UZ)d>tPV2jIoUSDjh?j}Ty2YjYtb
z_TwXb{!2wpE!KzptBCDQ7;V-v!UWV<XZaDF_EWog$5NU$=*ubWAS@$dkvnp$g_yPV
zdbC^BQ~)!8Z39y3Jh_{9r=GvtdqFXz=x#C!o~G|>0!q#wi79x=H>iUY<xAjs#)kh}
zt)%RYAslM-OKCy~hUi=xprw{EmzyXMRi@7$;R$L2x)v_33-F)?#@AU{<l-F0VuU&k
z1}e5_rg}6xRM&$t@`moby$5%Bi;&9DZi?Q5XbSe1&cv}UyTU{9Mf4rNy)#o!lskM&
z^fJ<KF`VTRF|H0|kaErEZ;cmJg__W#{%h&UT43g>$c4L%A-X7Wk6y}d3rly`_{1qJ
z#;5f2L;n-bl|CNj+frA%F}(8foLma~S+^CgKYN!G>XGis@l3kX`Sk18recK4ISw@3
zv@pdo&NWIn^1LA@Cl2h2&SNay<;`{!GQT0~M%E8R*NhFJCWZ+KUx$!`d6RIA&(pMW
z<<Z1%$ih5l6`|FW8)Ls}3G4U>1Xo`EwQxGwoSQZfsG{)FJ@}SoLH{nG!C(JkgWG>;
zjt*gbdyz3Nw%>o2Gp8L5Df!(Rzh4_#l<j$=9Wbr{q<U>lO<+>+A!+#G;UOG{+3UO$
zURqk($Nv0WcqOGX5d;X?-Az$j_s&;rhHH3b1?oFH&M4LHrFL%(?_BH<(`mXrwu@;`
z640joH*%rJg2^^-_+M>kdFNGxv<UvP-{0GPM0dP(8Yp=srI9G2<w@mUJH;|D?h=6%
zV`L2qG0NB&#f%`r(?B&3W;NPe;);009m^cAcG4?DMuB#9!0qpSz4d<GJAQH)IjyLu
z__*rg|0&ydSQv&sW$08>OLWb<gc9ATrKIxI-pWbvUH+*b*+JXD1G&JB3@duf_OJpq
zjOy?td#jH2Ls|6+!?!MXr<HmLLpwT3m;D@Ca&+|$Y<VdGp;||Ak%$sk^sj79$lw|#
z<)#<2O!5G+-Io;CRb$%GmeI;Ys)#z_0K$x9>e5Eqi>Dtolim^jKV+PpZvlKzRFoFA
zW>~4Y@3kq285QF1w6{{qSUGIWcjm5!54Tl+1#?-FOvih1#G2iUlPtY>!F4J3J0)o~
zkt*4}v5w(fTAr=fbK>XH3{IP*L>Elgc*gR`picf<)de?U6_t@P-RFS;+^FMy`0$*x
z52EaQYypOSI!nD^8q!2Do)z`AU+>9g3lZ$=6yCVJ2T=S?B!PlP9{|6ExPgL0-p2h?
zPp75cM_>sz*=>JPj_;^eI=>pziCy2)gY@Kf9M%_EL#iX)G-oo)ZQXIqv9+`KQ`Dq!
zgfpPGVGn#XXkkPn|C#D%C>qu&{v<>HMuIGahu2a~?LbLMGxC_;wvjSh!V`=6!>3PL
zF?=E7D53@U6%m{0OsO^qNDo+s96xMr6vwGO0f`p~l>E(S$Cugx>FJ+p$@|3vz`*3d
zaiZsJ#U6U7xcuYbP!AW%BIw>H1d{HqHIH`y`CiA*HMe5gC<rL}IF)q03mHXk8rUi0
zw)i_Aj(90F_GPFtDC5oMIL)lMR*8PSjS@kf`g$D<QS$Jpz+cc%p4T0z`QW~fBi|{T
z9FYoLpI}?EI~4vAih45HppnGfvj-PmDO0%+5Xgl}1*cj5jfi&_TN5V$Sj@{5hT44!
z;Yv(PtN&h07o;(_Y({_+7kEN+ztrohy_L9JMcz6pJwTNIOCm{}x|tHdb$`cOc#~IQ
zhb!F=N9B|j*7<z=n6ySsOUwRzu}OwyO+q<5w8Ml@=wp--dLW=DVjQhyxBX1OhO=o5
zPC#T6+Tpgs48R{WK-5GzVhr~14pLG0V{oz0#WgNsI9n6B@7(>oXu2c!2J%aj0rKB}
zQ?s+HjrRHR@#+Q!r0M0j?WbtA|HQbfsn@uAE2?{cBad(&`TF5^(*&c8`5icayWd$F
zr^>IYFPr{v>40e)$ke_cJ`dfphg;ep;Fe9RcyTO3$8UBtFd9M$ez1L`91qV*S($ez
z@iO;aupu}i(!ZCBEAQcn4evb?Z8chpHXv7MQJZ$#x`U-?jPI<H4NL#Ig<sfph}UlW
znUYquJSR1-ZAEj+!^+%$EG<phjy@?ceM#Hlr!!xzJL+i~7jN@q2j4KgWA>XLu0Jj7
z>nOG5T12mfS1OM7F^_MUE2GGTEl7Gtij*s27+z`He}DN$=I9rCcV|C}=yxLoJQ(DO
zeyP~);aXkT<h=-w5_%7+G?ko4=TC#3N`Ex@-M;-l_N8&#=&`hju;uaR<a(OFJ#lN%
zoHM07Hg`R4gJqP7X_hk{7s5#y7kq3>W{~Wd7uRcD4{a4ACwVfoU<XZuNscbnT<e(9
z*r-aW%m*KgEPDCuIV(Y?l^+oRU9gADGh3tl#|$5wPrsx=r%MjW`a95Lu5+wxzaC?u
z{RgA;=?b>&H}+RI-bJi()tTwHMLWmU<$80*;A-QAzE?-SZ93QI5{aH?#E$h?pQLz~
zg<4=87^1r9dSO)ECh+8EO$BRjGtMF=*Oj(KuXq_95kyvz_JNHIqp1}RNjK1j)|ikI
z4c`0iyP*WUY3Co8oW)NjCts`}|Mj1h;(a`8CVAI&_5>Ls(n8YKyP)izj+e?KgVPgJ
z%%GxXRy#rW`SHewMogc;jgOx|A=T&YTc7oHQQ?zL0n1u}VhYq+(NCN)?xS4vh$1LL
zm#H!Bcd>dgI?taLliB)&t8C8e@1;$06R|EBn9V!DW`pi=Pm@z)E0H^=wq<JF+5KrW
z9Slu|=<>+$y#J+m%%)Np<VlfnS=5rXw!n4d`>ctNs88;PL0cX~U}LI!U+q^rCS&$?
z3jZFF2`4W!q`6JK5QJpyAuraAGVCup|76n%GmWxEwTIA^f&)4%=^Y<n87NJx9Ve%&
zhU(`_iH(D!AR4#t;AxmCZ^7Y%=3nVOlCnB6)N(3p7mWgv*rxSqS>1OmB(PrZ#iP2G
z3^FU2*=xSPdV#)`hS1{+uG94b;ZZFgveKf-HhMI_Sp&P&-1A(dKj6xShK7E$7Wf<g
zrFsF{*XPbOBGw=N#yohni(w^zZ{{YaG;wVjuCxPpjX~#8&biIXog*qYTqicwa!T*O
za`C*@WVbgF9PY5PoRTyU{gUF2hn6p5WE4#DdijO5ScDJ=^o~NeqBL7Ks_a<OT*WE6
zbUHZvjKOY0-{YcpeN;3UQmUQrgZd-J?z|={H(%mS5X#-z`$4ebx#1;oUwONBHOdtH
zT-_aZ!RrlB`11jR9-HV1vDUv!c(BwWDQYggyV`d!#bi4)s<Zo{W3WYPC0AFVXZO-2
z5+Wq@s6041<&}``Mil~IAL$RoBhNtD`TX>xXlxVW4^{|Z-vYUqxT4s}D2=3l=&pde
zY}3d!(k<l>AnlhY(l;U4Nd&2t#`vF@{M~mYFe(kMtPv<C<6oq2i0;KoFhdufD%KP(
zXk=}a=E)?;sGt}ytIpOE=FI8nVZ!{q@q-TOrOa^scDXOLY!zuR{}JPbh3%}nhlhvB
zfRb2sny;k%eN(`!BzFkDe}PnpA<VgCt!|=LY4LW1WqTZ@`ohV1*EPCcGapHh0BMka
z;M~~?2ZY*mFpeuIP6rw8v#4LFh@LoHn23`eu_hrQJMyDq_qog2f~cYbL6RcZKKE7~
z7VQ6Sy)s~xsSpyDG^^_9$bz#J<YsBJf<N1M-e-j%0?;Z})R(e(FxC_w+T`;d6z=r0
zSgzX64AYff??Y;+5;)g+rrw~VUl<LaoxIbvQJx-)gJQ75q%&7MHT9h-YHCXU%}W~^
zDkKt^8bsz7GRY@gFna|qniMsHL{NZ-=YbYNO-0pu(MQnQ+UjEuJfpSEpS_R?d?bpj
z1SCS7Y54s}gWK44SzPXPMgE}SzZwjaSdnf^mIjS`o;o7t_6#>YRYPGBXW-m!)H;F_
zQob<-MYTAfw3MUS`8F`fFA3^#0!Oq8FX-pw*Ty~BNb=(7=#YO3cTZ1+gkE36hU#h+
z2CoMdNXm%Usnj^NTncAtYb!rJy=AmNfQSLirCaw6@8Y+hG9*4u{h3oAvHLA3cLZ_Q
zud{dt>z1~f-)-pIHNuampranm9;1}s{fPCyb{=|)I|;pr-*C7IVxG;_s2l9S*0a<Y
zo&DvA<C#Ow%2j%3ZEeltTD8*mR$bBV1<aS*c>&o$WW385XyJeSL>gKnET2<oCZv>X
zW5PCpVcUm(erjpyAJdPo&ugzJCwvb3<EY0FV!peJeiESZaD9Hb=3Au@%_u1NIv7g}
zyz`!c@|+zOb6RtJVj_vlcGJE*T;%ef`mZ0Li~y6&oYSf>==`@?1fF(iueLUrgva~)
zFB{E)xSZ=h<w*ar2_`hZ>&MImSkA5R;NZ9kK+Vb2EL+z-KQHgoaYSs=Pco|2Pm=lx
zggKgNd}5m0Q|fd~bp5h^4qNLq_t)vr?oc?Blcx#+S;fQS@EwU4Um7$+Nd>N~!hc0k
zRe<pISux~15ug|zKXeg-U)&b`X@0)-)OgE1zw+#$X=`J{Vo_z=sDl|ouU^9C{i0->
zs_%8*yP~TXS(cGu&=r6P^X`%q&=8*Y1Ya|!^yiZpu!jI?WZ);&KWo7j*GoWErNs;r
z3G&Jz1j^cU+0-CI7?1Mh&VR-GCK-y=>#po;T<-L`M#%ZtZ9r|pCmcyGBnX?f`^p}!
ztvh^3k~_zX)WuRMsQ{wv+r`$W;aZE4)eTbBANcs6)wQ%h%Rih$Y49kgsv^QMFW+7u
zn!G8etV#UU`jTK-UHU2xYozn|_}Kg9GF_DQKN|SvrlySdGU*wd>p-~!Ur#|!o+tOK
zI-{uQu=RG8Aj}(tH+c#9@YS6HIsu?V=HTAM*rCS#c)9;2MA}mX_3yN!>Y+NQvc)aa
zeZP)y`?~a|y@bu*?sfLLxG39I0r$6t;cno*%d=2d-kRr82D_X_0P=S%5xQ0|>kS|h
zD5d*Z&G$>FPcJNF<juNnbjcj5(9HkwBXhmTGaj0c(Lj?Vr|=izFb9hiSP2}uRDYnF
zQr6HQc~FS;r@{VrQ-LBgJ|W>9UWETBV}YuwYDRwsfvv5r;LDvEQC6$_t(KNn76F*F
z+Pt10*5drvJX<rG399Rx!TkKJJ<QTx4qEc(kusOtv<>p$J|ho`5Hi*;tlg=zViZCC
z=)Uuop>V8~0ad|L5-L$V@jT+un}))!E7-WdYd2i<)zL@^V{bEs#tki(3JPiS{4DK-
zYvh|+u}8ktH)N(_8Z(}pZ8UyEU*-x7ae>uEQ-Hj@Jf0OQNCS_zt$v%vWkd{WOl)k(
ziR-x#aZR9>E0_xPJ+I^TH^OIe^gUuypJ*sN?d>nY$kStV0YtmN;F0?(7@#9_Gc$4g
zPk(U<pTyW_!DYq<k>#wfw{RcP53<r{_^TA1oW{2i3F2+7xVHtNRVLWpvdY*<hOiT{
z(taW`nT=LGt==V+rZF<j`09hxdLe7frP=AmjOHX~f9iBXH#EfJR7_qjP&FJ7kgVIx
zB`!|tOe83nWM{<RJ(5U>MNe9JUPm?Ol@u!yu=lo%@M*67rS?$D*3q*|_Wb51O~CWv
zvKh)xvAE*c5CV^BWoY<gb|EZ094L_}lK4VMufa=C%%dkR-a9%9S<!5Dy9WEcG({6c
z#Q%c-T}RVQ(ZN9?7VFu!eyW)|)v%H$F_}Nz&XqJDtFE%>PJ>?GIIe?HQ{?8)<XD?@
zjbo3}n8=xG7d2Ua^+%%xa|=VF&x!~aBs^mU`xRIle3w}<J9Be$Crc6+7f`Oh@z4<A
zEthZg>jNJjJ4h_0Y%%waze(Yp8VZEz;(*PrzOGK1{!e$X3j$jpVuvk0mvd4JV#w+3
z`8jTc?f$_5o}d@s^^YBd*if3Yag-~`D;H(HiR|_DRsQ$n6fEHm9kk$ehTr|n5xe!c
zIFN%(cs~o<Dfp9Kk%DxJ*PjJa)tq!{A2IHYm=avG85EM6&wav<=bH2J*=x5+cn?pc
z@$zR(;M=!aa<W>;$QdtQ(arLM#sFrxCDe|3q2IB8L-MsDKi=Idlw}!q4*46>wgFIa
z{wtd<lK(>5mHw+ND4^lzC*n9#B#?Ws`I-IrUOqL%i0R`;XHbYYHhv6tpSaMdsi|RO
zV}r8Hk36LDOVQyU79raIG{WU_m~+ojPJmkn^+R>3Aa%u?k)s6l48q6SUDOQae6pgx
z>~#7u1&y1&&FH37r#^sJE5K{(d(&@>>@U>2=O)`|WRXfGhH6>nRL4|ItaC#W*nw<}
zGzCs)dR}rU>#gt~+?9$e!2D`K22E@pg23(3ZVdKX%BBfM+==Giyt_GLAzo&jslei;
zTV@pG2ALZhU*4r9CxgVin~YE778k#{?ektM=jg81K}sC6R?Yt5jNtRFCc|tf#{1pT
z-_3xUu)DjP35NIqnos&Su!w>CwbE&r$~NPRlLM!G@fXnlDp&6u=i9R)i~=3?cTG6a
z5*?D<rOrk2h13!L6}W%(%vVyC<hH*tRK1f@YZjAb5OktLKR}TCqvIZ%EUTEYZ1YJy
zEkRpcJk9B?i!aRy09OVFu3Kto@R0ra#y4fHqk0*nY;&Nw=i^m0(iM5K*0AjJ<)jBQ
z9N!PlS}oPJ-tVStrJ*U?oL4oM`=+4;S!q5Nf3l*wF&R7)$4okF4ReOiyW7#5Xb|p@
z#1hAJJ;fJBThbWoix9A#i~lS_zUi@|<p1$GK>KCk4c<b4?ow*AKj~hQVN^*@y+)c-
z-yqWNMaP+tt-WJ<DUX_W%DNX_BJCFOMA%|SbB+9ZdDJ<Jo~zf&Ayel5J!mw+VsNhH
zb9PY?J~p-!;19eLxa(M2O2H)jKbF41A<zH&du!RXux#62%eHOXuGO7wF4wZ{W!tuG
z*K>WozvmygapAnrdFixcdB0u>=)kZ3*ZwG{m$g1j*7UV~Yda~XFw^{A6H4k&XPT;H
z8KTvPkx&ZmR$qD8N`}luM8<2dworG^F>7r6y+mC5axVoye#&t59(%T9Xo=(Q+}(#}
zCwPxZk;+3$FJ`4xyyy@^EyKHQ$VgH9mz`)21saZxaB(YaqW%ki`S|##4RMJ0`}Z&C
z6_A?(>Ju4&)1WZ1R^$#ezybPDI=@~tNmcK17$L=WsJkOKvzHR#J$(u1b{t#YwU~)i
z``G`%E5?ne=^vS^GBN6vYA$^K)k_HQCQ}}N`8YB7^Gyp#Rsqnl&WVDC1cb{MN?0r*
zKEJ-^rmHlgie$}_JD^>8Z`Wi0U6}UYUSJhE)c@&NPVkeHfBL>Xl_<ps9v(S!w|-tT
zME-NEZZ7uT4v_iL*sXl%80&_M@78pC$f6_De0l}f)^Tf>g7Oub`*19Vytn+9ze-Z;
z>!M;WqAt0+W*(*VtaH@5;k4F@I5!%sMPv`j-+j?&WK7KLY=y}3>cYZ{t!{cu%)Mb&
zjjbN!V0Q}zw2B7c;R{`WFeeGZ6{u!kG$31x>NdMutU$cPv1?~iH=7z6u>_pyF}Dtk
zfdux5jg79-?*=fzHvcPA!CLE1I}Ve2DCg-XX1Ee#?(3W=i5G0;^FPWtDT4|-PbWFZ
zX$TR-)kHn!EPJ&RPJ4W(bWXDscI|hQw${#9*2vFh5u$934Ssg55B$CEgJFtd?WPJ9
zCq!cB?G(hf9y0vBEH2^E&I_Xze@lZ=>b~BI^l>*U2Y4WbI2@h+L;5fG_f^!@n*k;r
z2w2V*d%`?du#e^Y_k5Oc4Ah_^payAkeMdP9^qQUt?orJr!wkc!nh2LYZBd;ua91`|
zZ)h)TUUtfDQAdHPA@VPt<0AHsi{{;i|9A0yGHD~SXUxTNefseiFj-d;-mNxSxt;Hz
zVI~yg6t4LX1|SH~ZqlA#9_TC&JFQZiME=x&cG)F_?ap;XheE=~y+iW>c2j!$_=MSw
zQF4Zkgnai!j$p#6U2U2Hp1UYX?_U$oyT0y^M6~`$-Uk27;6kyOQ=%r4!jgo6PKmSQ
z_ZZxCxTkTd3!a;bO#&ikb-Jf<0jjQ-+y{OKaLoQtW%TZk+gDMUzgQN!z9w4qf*8wr
z$FinG`L_5|n0n&I=Z6Xk+d#L1xRuo@aLf-Wx>?9~zD)8`qt{*9au%+L7S`rGFU8-2
zJCl56^9U&*E!`W5rU7?~6+3L6WHV(|Krb?lO-^mUB5M_KB`S^+Cx@U%1^{i0GY(W|
zqdLOp6B0NC`1&7M4e4p!xbbwHQmtBh-N8ldLm-51DFTYZxog}-STI(UpH`6|;dUaH
zG=OmD;^LY)l!1f<UAwz~NF{5D4R$ZHEY2w#smABD7YJBfQ}KF+@%+<cIY#^1EAUHZ
zV{lw<RKv{6(lZoLG!*)$akz2OEQBIYK{?7)Zfcm#Bq)hgLN%VsS{|(AcVPV><?I3U
ztv~)kK+K=74px;V#L4tUq5CXYvI?PlR1=xsFqQ^b)qSy~{;|6?Y7Hed3KV?4hS||k
zDJ?CpX6%xB;Q8WT2RyWqSS;+;OJDdC0=p+C?RNUYoIYaHG0oJm@u1YI){>G`eyfCR
zC0csPCw9#kZ4t~}5jHGX=JOaRQ4VuaDSu*g;|kN&{-eXRGDOpx{;jEw*)h$kGa4oL
zjXUB%si)FrV%ah8TJnnKfTZ8vco_nMt<81XZj(Hp4tV-mm)E@>Vv8%`>l+0SmbFjy
zf|}?|rtDMf^_rf#f%Sa$#m@b)dxwVrdb|qSKl@ZnOiV1V*soay^WOqAvoR47Udf;g
z4m<pb64Al#bsn~hwSlou@UH)kWbIS+c>h-%1|~L-&od^-N1&3l`{KvpCVnff-5s$r
z6BhP?@kvvlMNRP%4q;fRXp55!XYK?M8#qWd*hH|E_rWg|%LSYdg?o8o&&~7T;T8zy
zMmh(G?{^2UjNIt=B327pBjYDrx<z^!u&uF_Gmrxn@Ac(nAPVQP+<(BdZu(uw#82-5
zgA}#nenBfLDh6^L2+}y3&YyI&2Sg-*@qc`LoS`b1ksHy{BJ28kfO+PBFEVFVD~bzS
z=I^+q#XUMp5^Li1ImAvkt|WcL5gCQqSTA-U8g;*ZR&k!b|FSv{m{Es_p=W?*yP~|i
z^oL0-=R<O*`r7`Pw%=WGb%&0qpa7|%>{}H6y~_Kd^{=vhQmW{^)XBt@#?P|i@WpTt
zJmj~19aq-}rB^LY&EVng#zyu#@ofSIJ6l^@A%V~sr>hLRPK{dQVY6(XM)zyOyF3Q&
z)pRy1U^C~=eMpXbPzVu*DdyQZeecrn+vmmk#aWO;?9?cgY%7PsTmK^CTeK9cbYeZf
zc~iO(6^ewe^OuPJp50bn$k!Y1?~*1w@S(rtjg(#cW@Z5{2@C^@JnZfu(YTU-0RxpI
zysO{B_}IaLjhgyhLxYh^q{pDfwnd3m?UrEAf1Xj3WzppOq!q*5=51{+DQb=STe>|m
zdZIiabn~^Ouv`F)U|$>VKW-**u(9P41^xDduI}!%0g2q~s54mi)BvE;Y^bl-b#8=!
z!=Tk_JR2fF#l##W2<Ftio8!0XACz7A_^Y6)-a9DZh&lpj`2G#*-Q-EpF0@?1pfRXS
zQjQj5zHe@9=>DnH!cV%anJRBe>VXnfPwd2X+N>7gMrt=)lE%n}@3B4}+O$(Pb&nG=
zi#|fS###Ot8CiLMUB$&~Y1RR=!*Y+^%x{QB*_k+W{A!7jG9Kk-tTncS(Rt;+@7N;b
z`RcNjQ&q(P%RGk+hkyV)G*OX}$_fe|yHOljE?FMarO+>jLj8Wg4oPJd6+OrHuC7mj
z9<I^BM1YNy=gEsP$i`i=%t{N8vbNJhl6toC0SiV?{wLPOGrKCQy#V*|W%EGoG(6Ce
zcu@qZaERl4#j)(ewR$0E$z~2X%eJMQ%X}_+3{gpu6T->vWqkoF4p$V3dTgJ|e+Sy&
zY@O*My{RJVF)AkSiHa-#x^CI!8q<{G)IYp^?bm&V@KSe#U2-5b{pZ_NZ?2=%e;TQ9
zmg8#v$H`PC5>8HcdoN)2FP7kJM&M)c$N|Wh+FIsDT<OrRCYM_jc%Hsry<BgiKnPLC
zh12<0NE8c7@4H?*JxOv$Viz^H3>l7mwrS{v<YHw}H0><2HI+5^mEL~Kvx%A8NKKoi
zA6XKxnCad1EO@!SQEl5>1a}nH78VQhHghvOi;JT}l+Zn4_eJKFC@Say^VOJ(rS5#Y
zPzpjoybBL6FE1UPO~H2@vc&?41i(`4db2nD`F6+;^3&DQiXcwvkMFYh!|BkovVzU`
zOG)V#zypLTWUaZ}yK<b5cWiz<3_wWv#Ynl&!_NbNOqNt2gNBhp#={>caJhoD-?$<B
zh=%^D?l#NGT8BRuo9R2z{vDQG3MkzOge;98oJ)`n5e;1C?3H2j(&{}e>WOmpfp?LJ
z`$@uD{%~z+j?D|$t|Km`&-<f37~TF7p2lAjVjwh8*FZh6b0kJ2At6C|MoUWzq*p*a
zGNwt30p|}xBrqrdB3xkiFQDP|@<51~cBrnYae0|D`LPk5ZC0O@iz(}~pM&+|JV9ct
z`Dl!9QF}q)%DB12iA9`7oM@%Cs&mjO&ci~VJT(S}tc%>YZC*!3%aZ75dPlLQVBo@X
zLsrMui2@ndo02hQb$45#%?xQon3+DIM_RLL-{3_Ak17Qrxtn?q+V;c|PxYTp(tjoM
z^}oV{ulII#e5|0Pl)>Y60fY^q$7?&!;hve*{>*0m-fP;oIkB)3qXt$l|A@BO70x{<
zNsNmMQ^v<cikjQ!FyH^S<sD=$g_yM8wC!E$!buFFAeUDmDAU7pDg)F>CaYgDa5<#L
zrj<(h<cDBA-2x~;`h<lQ$LVogVb)~=Z6IK5akR{DE<vsTJ8Y%$CzCF`E~R_Qsz(6m
zI!cs^5Z@_MMu9WPZV)K$&xhGSYS8OZa7G4>24^TD;KTJnGEgILV62ifiTW!0=+57u
zX_o(;>3w;zTluw$ik7JLx3!*B_Of#w(2@babNkEZfo%#v$HmGMe^V?#Ui*=S)uPQW
zHJPvu>AX#(tL3&)Im;k1KCU3HIK`hVp>QitjXX3&9uO}beqaf3OXdhdoIY=-A%GB_
zNORVnQ{1W`H|9CicTuO(PH_W#=-v0aAG5KsQRwsd>@2c&a(Y_f;Yw>8sCN`{S+rzt
zbstASp9Z8=W^4@hm15b@@kyECb|2$=wY_rq<jbk2q*u96$&`}&2B?s!XiU~&OEL>p
zHwKc)mq&j_oks`wBGc@`!;&Hc!psncnHD=IA=WP!91Nih#jHRu%BJ{K;TNRhlqlwq
z+e2om<1<0-!o{60y&43~J2KFKBKuH%3^jS+ZNEk!-Cx!UPwb9|_jK!J{#2k@0jS|X
zc1hpB0OGkoHwTclM090hYWfo&;p<N*zqHQU%1W3r)2rv6lI(5sma0rZrYje_A$2|O
zz_(KsYk^_oa=6lI1v*WDol1s|-^KQ`Nh34AMw8=@wDyhj5eE(#fcqr$yTz6smjO8A
z5KxSDS#a>ccM1SR5XQq}R)ir&sRLtByl)Ih4wU)05wbyBFy3-7wO<g?pv4GpiHsC$
zf)SVk!mV5{)}9alT%HIbQPPpE+}qCKYS&nc>Cz`>xzCc@>|bAL+HrNxRWf&lrPA-$
zJV6!xxJHi)RDzXQZ)z@VCa00tN)9>jpk*vn^{*s)#c7RySSBl02mMKi(pikJyEpDg
zAh=oLfJXe%ACLUw?`Ee?oe>mPio|n^n^$zzR-u2A&KukF&nF;o)9DCKAJ7I%^0__o
z)1W0LCFObr*9HencH^x^_Yde}(KOwq<X4QZbFTjdl+TfN7zQ0D&5xd*_Wpu@v5I1h
zxK6EgOf$h6OpGLhUUBmeShu7|h;Gt@iw#n(9CiA@oVhS}ZbqXR(D8qg6g>F%b_>KY
zre$K%&y|)U@+MHC+5@9KYknqx4-11!x?V2^f&jHfw%><`qN1XU%YCiMD3IJz(Kz_0
z>c3>d7y<^uw{gok+&NC$y-Uns11aT&A7}^Lg<N(mg@9%Wb)hWC0bNz{B0Swf=2^xu
zJMBdlV!~BY<TYt?A`v3t7hgEn8*y9g(EbbM_pdtSBC*j^f#nE;SP=Wa9ImFp#6?6v
z>6!o;nktzTCx%>8{F#n0uisYCW)cYsln@f?1!{S)<4|fEpMt;7^-jN33O!k;O8!Ak
zNm$2hVcCzyy}qw)&6MrZjjvQHt&$Uc9X8?nTnjJUsp09?kCkg}^&4^SP4j~$ISJbx
zGu-psf6yGv##W-+N*e)NntIf_eq(i4IlEQ9=Wdez!@n=)HbR8rozkUwf^e{cR}vUe
z30HNyl-Pi<TLPd(rH`CjQ&LhgGj%*)zDu~svrKh&Z<k~OT9^3zlw~X^cr8Pdqy0my
zs9BpC?li4V7Q$Mi4!MwZ2h~(Lya2*QMC~>a<+;5G;h`OfOXuiWxZh-?Ll0r%1at{`
z2f2YzhD<-1oN|OmTjhHd*Z9Ba-;z3e9tOd=Kp{DXhXd9n2`@esHzz<D39e{7H~Bx8
zb}<lFhEwXW2Hl(Nt_|Cve;h-Axgg8kZ*UEBy7cl|(Ryox&RGkOz1ssDQho|MaR(NW
zo2t~=q`WSS>*%f&3h91hGJ}qkA|VjiP`|(@t_$q1nu6zAgF-rq!L?xFNV$867eSDO
z;0?}js5;gyRx>F33S0>iC?lhIl}3WWYBH2=s+SW^(KZ4t$IjZ{Qw38iP7@8uG-o7u
zA-x8!Omx0XRV`}y=Ic?$oNLz_PSNYVq@zyV#kNvQF7*3X{r(uR<~9j2)($EQqW7kD
z@=i~`<n_$W?;p-@OcbGsM{{mx@;v9%V8dqyI5`Kp4J2)4r^8Kn7_B8tC9nM}4xZLa
zGF$!1`EzY?wjbQSN#~&z(@pznMQGBRram@MevG#-F{~T;Q&qGk(8fT5NuAmj1k=rk
zjthk{fE5{lNNCPX(C}FNaxMuCzMULX|CI+I-j6E-TMH@;NS!rBp?L*UfZf*5y9Ttb
z%#MJ^#^T^p)nF-pWhB`JuKucR7CD8~EuWG2HF-fY<xm}jmT>AjR94M+_wgLm2-dih
zLWUDR0&pD-#>xptUs3)yAc|E^K^58dPYn?>uDO$OdL=CUQX#3v8=|KyUTu==l$I<?
zFzhQ_`}z{31dNFG{JF=3fD}L!@!_87uQwA^Z}ODa>!EiWR2R><T<M&WD^au8xGGOd
zA*HFtjl<;&<t=5EPL`th^}Lzqh-THB=ifI~`WF1}nhri^rnk;?9JWn6NzY0S-+h+*
z(@tX$y1aA2hKNz80|G3Sm5-Y267q-(;02n)h|tE_dSG|*eh>cl<$}m>-CmnUsmLnS
zBAuV=FuiN*axe{`ntLyyfHsb0iotvH-#rb5Xvw;(UiZ=ac*uWo%Rm<$J=BA2S=w^p
zXaWCSd4>hL$IxJraA#U?c3*k%a~fXnziUG%J2(F4zZA^OBH|SCgCUCw0?|n5OX+gW
zEAXHlDE@kH|9~1E5>O`hdywK!2r@7__;h24>iKt$j-&yooat*uL2JH2`|4hjTCiM#
zuVk)r^QMZ4&@Ubwje3%d@bQ9-Em@cX3^a>-FPZn`>&vh-JGX0`o+avv0EpKnl(7o#
zY*XTYcJ@Ld1{^F;4GpC3&ju{76IKt)rM=&p+uWlZVheEp{B+S!Qo;$2T5divHo=P4
z<DSD<R>efk<k`tPxgU!aM6NfnbE*Q`2J*a{@=k41ZwJ5)c;xGbq9)#26zti>E;>vL
zjM|I6Ukr>x0^?INYYxO8GrfMf5d^>+o6O^JP8nL$n%hG9R2R;K?89+{;*5MI<3>hC
zYTu{Aqq|-nA1?y{z<*)fhW+8CrTi>D>&*^8N*ESE8fBK0P{H$i#+Wdt0V`>QkaU~v
z4FtNgBZh&P?mDx{4=*SC;k=5V$5BS>vF&rS*?4>50RZK3_}~J6yn0`9wAvs2Z9du>
z*;>a!m#Xutk3rnPB3>i3tClZjA}&rwKdzvMgxot$)<&PX8^%dDOD$?~C8<fN>EKNT
z+*%7KPW(tgfsMz_Rkt0QNdT!BC3#<-Rn_6UgS0eYgXiYrs?ew-zen1>C7`Rwm!6xN
z%26zdj*bS1A@n=OkHY$TX8_3n@=}0oy)GbFOMc+&KVY0PSDy0c&!5@eH!v|23>P5$
z+ioa=zQKeoxS=w8`?{O$f<Ab79HZi6T#UrX&GC(vEDN6}?&;td=$Ta4#+u@&t<z3%
zEmNJzdOgEIf-n0*iqTT-o58m|UppfMd(Ka!l=-r4?30EVg8kC;&@iA0j<}T5+XAq3
zo&s+$p#e`L;DG`WXnmcfQhv8NTXm;GhH$)^$`z73w+4df01!>RTr9go9uaXzT1V#{
z@H2vVdV70|WHN$n8GpdvZ9P1Q!QB8r4gz#6k4S1pM!q*uKj@7bcdhGD*<GDBII2be
z_|gGp(0XZ1_u<9wRJ>1E46<;OlCy0B9T{;1KbgvT^^6?FUy<i$?}zKuLWE(g;hks>
ze!qO>^?jR{br3u|K-STyz$lm6XxZ#agzr_hhlj85d3o~kenDzYsZC?u)p49m>%xQr
z+%wE}8|?t^PDaXm!yO$JmC5aLGQ|XJ59+Yr25YS`pC;DR+ayx$1i%kQcuX`j@0a5u
zA*lbnutP&f*Ih4OPV!Q>k(c+#@Ho4t4udO4)Rm8AH9gzDu+#Uy3#yeb)p!CT<^Uxp
zF*K_8&A>fq!@T!*Y7lh1`@um)hb=OrIljd!Fq(>1u<9?kP`Vmtf2o&Ob{=cr;QEcG
z0}UW{X&D@uyfE(5uv9VYR*imyOIWv<@FEV392}Sp_rL<Uj`sEqfN>XSo7~(4%w*3(
zgSbDvm;U|A%Ek&Z5*F^;lI70)3HY)*J3E`3n*rgS>5)EK00{CRo=j@W>EoQsePC%~
z`4~GmDlBS<*JB8O@wH<=*-?Nd_Q{uCvUpBUUv7_kezbv%)}PPSGY}^PB#5ex?KHx1
z@}wd^#p`GxM+>W0Fq@=NxfJIEMvtr8{j+!B_v+q}0613pIZtd7gd<C#w`^5^x33<L
z^)8kD{#LWu-re!+B))~AVXaXJuavs8^W9eZVUG1?=lXv<;nVx$CLoA`bOXA7z=51$
zo}m!`ajN(EGP?j|+%GPQaj@HMcmsV;7a!F5M4OhDVHyf6f0_3qq>=Vz3g9Q<+^-jO
zY)`Fn8EzN(vwNgxDMcja?yw&eS$3RLj0qjhd4^Ji`Q<~c0u!dAoy>6gxOK}b9B+A0
zWQ{+zC~wvw0`UC=0LpU^RGYRoGE8&Cx6s&4w+r>pd#4|H$bro;P*5Sj?x;K*z!`c8
zOw96pYe28aMJt^ZY4!Q>4zkKBCM6~H1}MhbkgUT0A{;IWEarhlU<cdV{<b;Ae1Ii$
zGYf>T_IB@F!DuaQq%&CBGI-s&-wS7rZ_I(bIUglrqd7}QMjX)j8(vHcQ68%NP!nC`
zw;~x7s-*3OqbwH=IZ1_m+lyi1I5n@~m+hRlppb!~je_~_DYD5w1E1dzwN^!CWB|QY
zGbI|#%Y+qURqBYo^_yeJaJPPhc?!sv<ONT|*YSM?fq~n=(vmu56wi1Q5RQ60-;f+W
zXw;eJsweVA`aA)Y5mcHxU}yf%9}6s^B91ekFI54U89ZM1@pLvwNJ!{_y}doaWZ?h7
z7_tFGU;_@urKzbeABkS*h>#GXxIEz|3jE|CzSFWy_f5y&KJ8`xhOoCfOXql(0<qlK
ze;W%Hl*a#Nuo=&PV<r;OHULay$F<|U-I>0K;XT-&r2WxnyFHgV#()Af!#7)aeJML$
z@jv#44q*v{n3!l?HsRPdKtl1tz_FUSqjF1jIn)<Z`cEPUb=OfKQ1!hY*WDCO?SQmy
z%Ee&CK!6OD0hY=2BJl%BGC<yHcA+v*UO%7Z0ZBn{04<;*IGD_i!H>)`KKe#X4eXl(
z{O;4686c)<2_Cm%Bd)60kdk@(%H@^mU+XU&M0wEstrw6h)0Y!h&{QB!Z=2=BP2-0;
z3pX}Om!)U;iTinWs$_%eR8}&ttWF5cPr#@k6Q=<@YZM=)HGEE)*!eksgx4vp?d;^0
zPTTDTsKHs6!H7{<=#QuIT&58WIXPjmCR&THAo$+a-p!Tz4X>}{ZXOTxMMXbe*h-fg
z%{CrOJCxTzmX8y)#Z<N`;vrs^>+LS|h?IzViWoYkz+B|##?Ps0P~Mn5neKEa3M-LK
zD=TKV=lK56NBqA_TUtB`aGff;a(T9ZpE45>@m?$jR!ghfng~qd#{}dZT#+tHkdeo{
z_(=Cm{K%%C(m*5Mso!F0$v?jPw5EAH2DpA@NMl9FXm6vQBZF<9+&Ifo))=n`sq<fy
zldPmF3Xg!BrLt2=d?qAfehIx(81q)TApd}nI<hAmkhS~?W_P?V@Q!jGM0~N4$I~lR
z{2Rix{sB&m(b2;N^mmd#Tx)|&b}Rhq!`Y@`a#FY#_sdJ|gU%<5igGqo+UE7os5Y7*
z%KCkqh~m-JHKJHm5(Td*H(=SU?(aYb^4(jJDsvdU@>Q*w9H2mHd(K`0O05a2OCmM7
zaBd!)x%pKaYkm4ymk<Rs&&=~WRiVJQ=`8(7P{E$Q9mD%5!-3%{e#dTf97os?YP(UX
zgjAGaXXv&1TmEL|ab=PJfPirLQ^_-g<!|(sFEC^!3xKas<8j_4;Tk0$V=j-<c%h!g
z-Z(ox>uKyvMpKYFQ=2a=w4V<j-{tk;`t;<tH9^l!-DMp&G5Xa2=hKTlAjy+?+Y+;0
zH8MMJnr936jpN1vd&@&mR&3j~`)9}bA0{i}$lkI%nQex)(PaN`c7-xK%+_&0RU}2R
ztF(}&icxcR_X|$MpX4>sIdEfqJ(tcDF}9Kg%Fz)pG{jfMfu%AdiWqG5>&|v4UW6u`
z{jZA?hS<*$58Mx!XImV)H5D%CsT{K6=G@d&Sf@vkWc{0IL)mvd`&ryTmfhc^<@|G&
z=X-t2i1qpXs=1q1uNL@>*qDK&l(p|1^*8B}WuqPe<@9!2^fR{KoV+tGe^W)RW~Kob
zcp%r73blZOdvBPlX-pyuGr?~r#P}4DQfI2rW$#P}3-#W{u~9VmTC;Mv^^0g%TcMl+
zU$(<-=X89;Q_y@@c%!j`x~dcH=&t{5O^EOfn_^Wz2A@eqf!D%`(9A_S2QWpTan{O7
zfTVuKrE}2|>IFN22U`dT{CA~G%f4FU^Q<M9R(~^%+G;S11nuqr|MZ&L50{6I8J`T8
z8+`{}m)*Zk@nAfg&#`(rRJ!!`RmF5v4SX0ZV_MJ*TD-Ux()4cGR@z%h(&2`%a^-^V
zeSNlEG}v2Iz%6FhwVd!s&f=v&PhjrlYRkznJ(NWY{W1YATh6cIhqK1N#&Q0?KMd!q
zLQm6Q>n&tsGTI8E*|+9pVA=4(>;>WaGp2v>CEG-TnEswYS4WHefL-&=P}VB%E2rHu
z-x$s+g9v^b^%2E#t!|jn5`bDBp@&O5FzUby7tFwZFLK<K3QbJJnjmxbygi<!6BYe}
zal79KDSzMCGiJai0L-ThP>Yo-32A9D23{}b?@nE-LSr35sP4U(tsSs*6S^6j-fv=c
zzSIwer;O9dWC~=-b(6J}v|oxU(UK3E2fXN+FR5+GXGuZ`A(FpRiQ$Q5(uIKT#h5&5
z1T$~v+3^eNstE3ndf8@H04Wg}Sx!S#2CD45YOv?$%Dl&aS3GCrVYh}!T{Sku;P{b%
zMk4HHwTbTl^SE#eJUI98yLFSNe%2uZjZNB4`dRgw0CcI9q?T>wAea(bC~2vpzozc;
ziyBKk14%wLbANX?*PcO*U<6=tx+!J0wXit9SoevG%VCRjV~BUOBoRKv-Q5C&V3uNc
zAUpj(sU{1HZK=)yYb|xV_~unz1xrM~iAtU|Q<g@5?EVn4!Lq=1FnKjm2Lo}1csR=*
z`95lu+OFx&2!9nP;*^<X1EwU(tMTqZ3Dlh%xCfku7m;Cc8dUC%i!A|BYg&6H)^|h^
ztBQg$;D7o_A!aSq+}-_a-nu{WI6-JQ?_29sr_|ul$PiJheq_sU(>wQ?9WbQ5#1^ul
z0;u;8pMZKW6#2#R*bu9Nl*5dA)A{)QX^_ay9tbh$5cCx|-5vC+9h)*O;ljz8U|(NO
z9?p%($@~`>eRcD2gED1|aNk}XrcTbtyxK;W*exbSciW|Rcpt_eb)%4m6rO&e3G;KZ
zjf(#39|PfO*-)WKYA5NS;-db_8pDH#!iSSqs_0d)ioli3pb?j^p??fj-5UBymiN;X
zXyvfm%Ah4)uDF_}=2hCf9)G#An<IQjE4VAJLgi44ak#P@5GayTlaXO`ZQ*rUA|mGn
zRpW~zFwCs(!S<-TXTm%1VgH`LAvsuyAKhB{@TH&S`L_2wW@g`msNLBznl>#>(*KRp
zxHg|7UG;SXA`s*#q(bPOb1CmQBPx8E5~l=?fB+C@e|8MFuvHU9r`5?y{TK(jyS_uc
zd)r!+;ux)xrrdJStG|3N3VqrEYzGvOhcw=_!rs?G>l*`746YpdIlb3Y$&o<#dlIHM
z((6^~mM9IOVgUeu_*6RVq-Eez!@ZfsikS>tR*x_}1P5bIPIU8!!4fC2dxe-)@~vRG
zrW9PGwVzndQJ;w<;pj&^_k@E12)1kUq0+n;fC1<t8tR6tJuoKfnpAv^%bc42rSfv)
z?4!X0O%m@>b_Ix3@^lpSroW<6*o38*)f^g5Ow<=og11IyrZP}pNxEj(q#9%Qm^ps+
zgExpP7@{W?9gduz%P)@eiFQ~Qq2n{rs-CpZDUFL9K`^Y*ghyB%8yAr3A~#_<j&AKR
zeI=B9{Q>m}i90q-uic^zu}bd1^=6UKC>dJuk$|VwxJ$A76%rY?t|`;OHM4B|fJS=c
z2#?TVJ{iCzf!BMZ!yWU=93t^|8fKmXHt>=Q2~H{ft?%0lVwkp(q6H5Gfo4*4;w#D(
zOguE0{{cfU0OUSB4M{1_eS2IPn2bK<>I}$u8ZyM&CG_A@*P7+D8n7JGYOwWR02$^T
zR)kmtPKZ)a!S_)k!%HY6AhW^O^@=`6eY3L}HoD<;^Uz-QO`C;k^~60g>^p{Vr)Bg(
zj{i3F)G7X-kYXVXO&vm#3>g^b_~1*5-EKy;z3NY8sj-aCm!F@3!5U0`0Q1B`s6w#y
z9j3X<4$#F{<<F%w#2mTx4{K?7%>FpL+9kv%gH~YT_t|gdnRbva4(6>m+p%$h`ZH}o
ze?FDwHSwH75W}{!lLc-rRX>P|cd0pEV@+fG&9796hR62Y<yO~1Cq*8%mN~^9BCQo4
zrfuTd^T%)dt7(^0EjXkMjI!+gM(23h>9E!|pqJob(2foNB1MhZdp1;tHa`EhbGL0j
zb+UDz-?bWj)O1qRVWn0<0K)UdO%%`Z>-^Lk>mAAKond=eq5fyD@8~AYHr%WPJMkqf
zC|-Z`N5!Xxs#(k#4uZi-EqTL=b?b7+Xh}ipRmF0{$H;@gDz3%+7aAw<lD0VPnlooP
zT{ceOLfr}E{f{IP8W&P0Bu%Ll>f52Gphyhf$-<vnA-AP0vZSfI*iuYzw$?$jL(F47
z)?-nB|8aPIba9k<dyXIEEH2~`GLurYdP4aNK}p0jb3hr6e|Knm(Sle}>%r}jr@{?(
z4j%1#NWL3NfSTlnVUnZ8>^sp;K-ri@KSCVM-6&sP_6zi=Lj7V$YIkXE`=5$}X`pd;
z#;)QgD+JEOFZpl=Kh>6^Mq5XIK*4Nn)e|t$cd$}+y*F3oN0r9@47(oPxQ*aKs*X%=
z|LUKinX^&lYLm3y%ax^X0~3}@d0Tu7XH#;otT^Vq{e^#?d(Ft-LItgPM0?1%2-gLW
zRr$0le}{s<9Q>F`Z_jK+y2A~e`-xQ|JT<56w)HC}1PV-GtKM=<E*M}t+U=zjw!e+x
zxbFPcQ6AzGaftw#<_yaseVGm2-=yMs04>|+QKRM2MG%LNeVVCa=!hmi%T=5}FYO`c
zoxobJ@qtN6Ws>BnVolL#{@%%Hw?SS5ypJsp=5=mQQn+4U&{D^3+1P9rW)r?BRagK`
z<-wNees7rDER$Vow%Hw)jT`Wl!r&!zwH^EE@LRF>F3c6|tr=Age;b&7?GjS2&#xb1
zECpf(f@P$<TD*Ugig=${H^nj44Gk9lJ)r7b$PcZTr{zY1PKo%FxvJH|q}BDg6teur
z6Dtf{;6+f#3h=Ndf4yE|2NDUKu8s`&#-wC(6g3u_O*@0Zyn+{o*WM}wfRMUaImY;{
zB!{~;mXZ&T90z1{BC(^Q12>CvGE>Y1S|qR(cH|;Vdza(o<S}eUts#q|ejC0FEALDZ
zs2a6>gW!z%Qd)+H*Jb7#=%)6!&s8Akl|242=Ja^^-NWrcY0`INdl|jP07{{ThB_Rg
zCXYDB8C(P$O*oxc1l*c1K>e2hw>U6IBj%Ekk_LXld8O<kKjot`N8AF@OzauMrmo~H
zfOq}SXS44cl>Cr+j7#hsx4Xv0i;mE!ZPky2;dDM<%)<DTl+I%ZkGupE(3;WoS=(A)
z&nS({9a*{MN`&KpR@v9Yl|%oS72bK$Y*JK&kh7dUs8cpPvhh+l1^9~Lax-vB&GwaC
z)dZyg{d;v*&(g4tm}=vRx+YRdsAYN8WY=rM<aF$^<P-zeYQ%U=C_76WJ(afps8+nC
zysh`F!8ph7f-FrRu2dro5Yt3!J1P#2$(Dg$yo%gUk3oG!&iBujm70~L>F8<q2p_?A
z@R-^k&X;^R5=F16Vi$;I#*Ta_KG0eT>3GB?sMLj;8CgQfvJ^cNqy&7L*>V0Z$fsLs
z`UbWPKC(eLF0);>UQ)Bchcajvnc#}4XmLT-6a&wH|D;ve`d(9Ij{K5@@?WXNp6wt%
z*m*3`!-xGHqHLP-Glj8nBTol8E$k1FC8|tn%wwkrKkpl09!%vLk&zp!T-(W(AW0%s
zqy_;_dbX-|eSs&}iFZm^;xcvIzY2Ptbj-Ky#NRd~R+Bd`)Q@>OpWS2m4i|eaX32)&
zEFY{#Lqd&7g=dYwH73dk!-k<V9z8_9cqD!s$);6#qqz{Q8Yb_HkY|?BJE6_!kTR8<
zM3}CWpWDi=u)5|pN;M^1d&kMgz#t+SPGMm7zf0WF0Se@>??23>_}aw7D{~lXqqOg+
z80#7h=NQe)q_n(!&?c%%Vp#^eySA%*Sp39$MYQ-OcX5iB6)zG)C(LsvMx3?L&ntO6
zy2<;n`9;w{9;_BY9V_Way&k)yYKW@5U^u>A!L+08nmC_Ru)<P)_y676UBd(nhPA%J
z6BxKZVTfY;%WGPz|8%<as79#7T7;<j=U2QgsqCiYB8CMkYzwgpTTh+OCYG5s*Sp9d
z`mb9WdG7Heuo4Szeoo0}(EQn%+QYP3h2a4qembxiv&QBP(rk3fEY%fJ6Mi`*T<ETp
zmzQU~Zz|L-%?gjKNrDmdufiMu89uI1Nm(h_A_ru8<FH)r&Z6<JK<278L3;KowG^$C
z*`gI;Bv=*v9_izPz+q7MavLkKZ}QYUN#|oJQJB%(xY*JbY3WSRGXpD?zvm40IGp$r
zl#n-QK}97If`qB5`5;YhSPm*!5a7;aGF-X@zWkP@iY25z){=>)LJd`lwUp)*`Eq&h
z>W-$&*)B|LE|ElSUOwcYnnj^fp#X=$rEoql!=TlTmz#U(PHqY7a9lEJSeybx2^R^^
zO^8WdC>ciXFN9E6O06V8r6U_X{9etX4kb|m`d6H7wm4^RQae6MrOW&N#WKpY^o)|s
z9i$0zZud#0(?FmN$+(oAP=s-d5UTL>aPU}!sN^c)jw3shwsQPgsA=84aAumS*PW1c
zepI17DGnu&R*%^ks<qIcCGgQ!&|6qFUtmn+N0P&JCcV=Qa=BjXxY<qu`oDe)OH06;
zpx5rsRI6f+2PmO~{Hp8U7KM1J0DgbD4zQd!0#mY%2OM0)Hz>rR;o<$mPn<7YG&EA8
zq6a+F-5M;are=o@CBa~Ii2J6PMC{B<o-?kP!nQ2PvJMkzbO|7q!b7XZ>(XETFP7=^
zOrve9m#XM0Y;jSd<fpS8kHK*1EbDci?DeLHG&!Yp;o5G+5R{sXr*Xc}qo%btfB3vP
zvLh9kbN2In6WYm1Nd=_Kxe^!_7bLGGp@2+HG6n{~xaIf1>?SBEi1&9!6M*IbR4}s^
z|Lx`Jn*j3A;z-0S&CP#X0uHyVe7u#F6>o2EfKr&5nVJ17W5k&5=jGz!0)z>FF?o}9
zwqI794z;(1(BblW8W=4ih|7S)S;iKlGZwjY=0ib6ssNk`K6PsS@qQiDQ!RN{tPk=T
zW$8OorBjKYzlN!tI880d97z~>K@@ze^wQMpm$lCseAsL)#SGKN(?=RH+U-Hn6qXq*
z)UlnohG>{Y5>jR*rOY%mEi8psqEI+meDB~iNacdjh{1t@^;F)s2jdl5O{l8?>I;WS
zZ*4gtxXS2yrmnZybtwp#&_bqUs)<v8z=v@EH9Bs}R*jurFWZWcCT5ppp28UYStcBH
zSg%MGq>cRfyj17Q+hyi|OmBVgCT1X{sL&vZuPDkHMJE<J!=?%3z>^feA)HFD<f>GE
ziN(=N@7YJn|8Yb}nQ({~<@+rFjC{QojBTvi)8Elkk3B)-?#du9RozAYvFex5_VZRS
z0uBJEcTWod0;TE$VENaW;rsiQApkZ4@mYPq)-l)3PM>3EPe4Kdput=w|42#u?f@c0
ztx}t)ANc|UPg!wlq1AS%5IZ?ZxpIrArx#Z`GiviCIK#~P;f9LnhkM2PaxHI{qSBnY
zPlAC2yXFQ#qTom-9VCVr{#2>Nci#fGyA{t#tLOnKG{Y<?$JO!b)2v*RX)IZsJra1_
zX?WyCNKIdd@)e!UaSA8AHTj7hVHt68Zb#C~iA+&Vw+~g1N)RFV?m$%8lbHzeP+Di_
z&*BXL!iHfdbclj|$V^H?SI7atPaqU9Cs-dQ8k;TL2^RqYAr^h<x7_i`i7!9*(dnzj
z{>}MXnX#(q*H3AOtU>JaP~cR^WT`sJs;dyA!toNTBYVr@!-vdDEm@POgw6UDm^rOo
zF3?)df-8x7bLP5PyXuoswhZ4zDF~LSc<M2-^^W4)J5is@PQoEN{LgMRjn^zD%R&7q
zeUmy<SKl(-`N>{KyQ$*WZYkn_((;Ip2teb0%1rRn`v;$=v$?rBkD8K+3E<#GK(v)r
z>o6V1W?^A5<fkM*(zmdn^0s`s+WvDB9v%+tlK@2lDr?;~*YnqD0rCVk-?O(KLHg+T
zqpLj#U=_U}Dhgh^it<#qY3~#PdT%xGFfo)IjZm+zhR`b9mWH}SkQptd`7y9yFULLj
zibhnDHV0o9n5}l)3Gkmu<CGdI^~mZWptUKVVFg-~SqTb#N5VCCWju>+f6$o|NP-&(
z5l-qlvb&Gx0|OiF{giUJ?wW0gi2=lfMsM%$JTEgq$SEN4tS#C9nin6BTEhY)yG`W?
z$e#xU1OTy{j_|_(&|X?n(mep|$}A`-=p?`cM~ujB@;Lf8U+U|Oq`~|COVkp-HVW$q
zqt2+cJ$FqB4P)}HtB$Nz7rhXR%=6SG$_*=Sg^^XyX<p|Uh9f!NSEsK%FP5~gcfwm%
zMEtBsC`BlmMoqLYUqgsID88^V%kf0n=b&l8=rRvcG>8n!UBqX4bKD9tPfLX7PXpLD
z0dRncT6r5V|6IGjzrDrbwAUAvB1dGT6Zl@DOikoO=pmd2xOi}|+4-?)sHmQSg%El9
z8QGFh;NX!s9F?`TsPB53ZpT|juPLvO4tzEjKtWP_cd4)cu^wEQI8S-W!Ba2K#}z;5
z<c?`Jniy+Z1-q+czCIV!!ac%vg5CA^_~-1~lQ**@QGVsd{-46AVp$biE;-K!r(&|C
zi;((`iH$<F`M5%?&t2=(KVhb1H2mJ;z$tvUTxaLOI#zZF0*VeUpdzpL<Me9evJf;k
zKLVC%s8(vjfc>}nYZBOlbQxiWVFkSA|4HY7?FSSx>51v-BPg@L%2YTQnC=N+!Lk#g
zFAxAToIv5n&#kVY%jJ5f&7FBlFD2S47nqV&De6V&#T67Lo6M%<)7?wBt&>d$Qw4kM
zfjykf9BV0Bt^VyCZ2q9+XJRe=B)uImk%HV?)H8Y;$J1DousqNXa@xc)L<|!uur#`|
z7aZjNIL%T}PV_xi3=R+!(!gqQW@KPs$VdP4T3G!xc?(#GN150tm)yMuYC1XntgCHm
zZCzPi<smCP2j%wl3E%nYuNQ*;VRx@*Wqt6;n?;R2KG>=1PsJfFX(?zIRu8HhhKX9j
z+i5>^Iyb`mqPrOWt$m+YD~52-U7Y5N;!Au^%o^n{qpGkV#tC9tepD^-Fz0qoX&Y%W
zLmhr!!gVDdo&agnNpRx4KueKZX3UPpwo&z&*8C#{?~T2=wtCBS&~Dhs1#xr8?3s8O
zy2u-myp9o7IEmHc#QDv94c=bZq}Q<MIPpm7s7j?PleBd18%tVPr<WF*_-w7rn@|~5
zd6px4(u;8wZgZu*7+E<)9f*0{ETJpek(g|Biapy|<e7g*$7i%~c4RXZQjr0A`p1FY
zt85L>nh%jBkhWW6w7ghVXK7V~=o`ZQ11DjK!^#CI3Ig(WOI2iEiu|a_UP1ObX#9mj
z&FazqqNgeYxLe^sFSfaHD>_xy!Q6#fE;!Fctxv>0g)_uDvvej0*lBNROrgkcfoi~1
zb%c;zxzfH;Dkwp=`F-gn-mx0iCs6Kp(2}?qO~BfwE_w*!yr`c5^0#tOhY(bTCRo(p
zg1afFkdspG?ZTpYW0_M+g~xVC0S}Y?Hj^2$*Hv3S(_Y_EYmGlDmDJv1Rw)`}@p1eS
zisAT@8dkKe*yfl^;*dxVh|2=Q<Y1D(n9*r6ZTe`cKH-jjR4qw<ssoKxRMlGkW@;uh
zJv%-TOk00(xEL5hfzSCRVCW*f)d25|jhoU(w*M+xBt|7mQG>*9<=0kB??t8m5r>O`
z>M|Em$Ly)|&5YMh<u1oWGOcO;C|*oLowCqoU(Rxt#<s>lTV%c3dZ=R*vBIy$bmH*!
z3oQoOwX&+bn*uZbB4VmR4dLsA0g}Z0W@kYaZ_Ou-@qa$dz%QYEM^$fV;96OsqL7`M
znFVeLKljr^)ZyPKD6c0pLC_8dJ(r93<a|5sT>b)x2|76U2#xD$Qk48fm7*ZXOi$f;
z{V9<yok=-E(=cIUCC98<gP)Bsi57%ttxlkWS(j8AG_65-L^uWT^+{}AZEAJwUFBha
z68*vve|kfJqcpg$t8{k47#e<7sX;7s>*j-4HQ$#MoJuL@`Mfd4Apa(n+y+n?y4fyW
zPpRZa7?We*>iO!}wi#G_#T3$US+pK5wyRckcW|AU@B7Efx|(3m-rjkZ4iu&KCEi;2
zj=9Ha;IQr~?~tGBPIfh>;B|IGP8_}tB|LQ=Fy&yNM2{<ljuI)=jojIB=4y!q3Z<k}
z4dR0*9=NMD&%xaeo!kT&MVYu^QEBx*EaTp)cdy>SWpkfims{bqqTtK4wQHrLjb-K)
zmg_l!Qo>Ma$UF-t7&9SY>O?9mA7vgkm{T@RR=1h)q_dj6A}b+kDFo3aiWaums=d1p
zRI^tl7C32&xVK*(Bpf5*PKE#JHBIU8POCPP>tAhBhw|z`Oy8_ah)+D+#`0)~!+6>S
zK0_-BeLj@K*H`p-Uq|%yuoT?y=wrxq{il{{%KQnQJ7LqAsjK(bTA7g?X5J`1Wm6jd
z6hBoR?Wm70#tR(nO4UtE?iI8R$Pjqeme&n(wYT=_WZ!L0dWaYE*8GYYsF3(_vB`>N
zf2SIW_^7ImlqiksqR-C*zoiE8U>{vJy}CHA$bZMG^pur;b2ppnW2c7VMF;tc-IF03
z7243CDOh@Y#`^e{RoJmLeiRQWu~%7<N;$fYhzgafADEfwRK>mVq5}|7rz9!AhGR?+
zheLcg$dCq(8y2Q7z1r|-p~oz|d$WTS%q`NI{rtxL>edYL&SbcuX?2^aNIBylouri0
zi?=+O;%JVkT0$QMzk!bRcUnvo>jwGqJG}<L9!*Dg#+Z|$AURw!a+AyQ5E2Dp)q|-0
z8aQzl@@*H<1`jv%DmId)*7)*4;3C<mI*pWGnH^ec$^rw=6@3(QX0u^`%<L4()NPE7
z^P1hAP=&bhPW7wi1lZ7Z#>>QwW6hW~{@k;`?_FLc?y2b})T6#iaZOd=XrniCYt^Pk
z#jt5lj;YGQ3=4Ot%t`?TLQf}W<IH+Ucx5?OSL@RIz<-=Qv$Wk&*txo;>NwI|Ddli<
zve%E9TUFi(7KAu5sKaJRguxY}qLxxtmO)D;Atlwzz4YqL$Ve#t+mD9(P+eKxS5w>S
zo_F}mCFPwL=^BrOckT=DEI_1f*30KYnU2~wHGgMG%(#przU)l-o@CUzxHkYxH<yX9
ze8y<SkWeUz^%ExaXlqk1U1<mC{_bV(XNv2|Yo~|I#hV@Y?X#tAARU{6(>tgi&S@GO
zOqkMuyPtqqt6m<~ML#wvMnl^^*tWIrjha$mtCFZfv_UlnnG6a$1dffglFlc~!H%pn
zy3rfzq{Ynzo~xY7aPXMzc%OJXxXs`$W#bvLCMjF-^kZfdpgy>IO*^{hfxtaBre$#$
zpeQxx)Z!Vpp0x-Fw!e!I#74AC%WI}PJUhM=O3is!f;aba4}Hf2kH(9usdW!n8J;P2
z#$zi3I!Ao1E}_jpDw+cAg!&b(-P(-z#e%(yn&n#miF%+q7)-k)F2*(1Hfb{B(9hR%
z-B*`FuhS4)ik#O8iOL~jFC@+9<b%dBEm%G4OW@)v1B9u3K5bj~q(60L!d4E3*0)UR
zq;T=z&d;h5AIR~%KgL`p>8V)dxQNJ&UehDIg7%aW9H%aGD7(<Y`qdah*MJB{=A`58
z<(W&5rE81ute61t>pHt)JNveK*<&P;rx?ww62Br*)%>n)oI3Q<XB4cXBSHZa$*hd#
zht-bseTOuef&dM%2!%PMpLyam!lMGH`~+t{-0B(vSUjG#E-rNGX5)rzfL56{wS97L
zLFq1m0~o|6BY)PHle?I>cod3FPA0cV1Tjn1(;8<dkUh9qh?b<KmTY*g_Y?>k)1)0>
zS+xrjorLi#a*L-jTFCn9{}4UmVQ1EmAq^OxsI@`lQNjw6|AzpakaAG|b!LeLlO;Lc
z7a|b*`snzp6}loNB3(rz5v9P;`e(IRO?9cz+?qFy&Z6PYYE|YU%e~Jqo*Y5ASg3eJ
zk2*IxcW5CtUdh$3!LzuKLX&&xzqy0V2E37|p`bQQSvj!p1v>lp=iW^BUcuc^Xfwmh
z>|z%UHFbD@GzB;Gz}Drgn!EcGk&Pvj2Dhc^=rd_vuhP^dEqlf}@ao%?YOC!`URgF;
z=dj!(@=l?!gdo&L2JF=DWRx|W7oNwuu$0qOG@`ZQ>px+z;gR~LNVwuhThXI~4<uMU
z<O@^0LQ5sT5IexkiQw^46V(|mM>D*<8ZFEq-~?7LT?<3`KVEDQ23c5WgmUz}E(*XE
zUHI9)W})d|>}S_H4VVbE6lFq@?{8ZV8&}&B)Xr@<#VE+77i-^fw~m*DG_ju~xmhV^
zUnKU_VwcHi{%oZmIwo|WLrLCgR%P$`7pEOzR+_*JKoGRou;dYlz7S8Hly)x2)joTi
z9y{^>elzy15|Oak_3k7{9$w~USArE&S0$~RE<N)pZcHRcm5c5wHDYT;o}3<$Vk7#y
z`cu~>8Ep`yJH}}$Sapc3>g;NKk1xwSr77hX-*|B!h}BgQ%ag5=7`{P?qxL=p0SG7I
zBVcXIW;H;mikJ89zIUZNA0whu=)QgCB0F+h`}B%QJXHNAp1Zs}c#;7hH?Ffp-!jd$
z_{65upgpa9uB0c@fYI^1ed5Gg7&~qm$?oE?;@(0_4ie7PO-Z_>@NC^`z+@PN0}l5J
z&@XN48Q^B7Zq^C4v;ER8m(ENX<2<@Vmk7VuT_bt-)`5j}o6O;>sN$rqY*EpYAe+u9
zE-P+p+8lFFQE-6?$D*MIv_rV8&Pw%S4~WKDZuPq-Qo!vbJ#n^_B@#EdxOG#9szG#&
zDKB61@T*8^D;Qo?M$;i$;!pPb+nnbA?WObKOJVunXZ#ZRRj-0jUXcLydIjfJ;^Gi@
z`HRW2bBnXwnj=e7N=H|OJL_*z*JrWQWpJuPt;KVDiR~LM^LpD~6GYS7zTS%zer=mD
z<M;T#3s_v9Q?A#hc9TjA@_)Gr3s{S{DP#DPF!fhKf-(HFL+k`4CLIx#x9$32W(Fro
zWPT4D4<-vnxD~dbGBh;HgejGw;G*qMo$!KNjEEhyVv|RMN+ET}Z(2TGJ#GBzSHa*y
zk_CZs&>BeA440ddj@Y@n754rq5Z)8CM&%>lJ7eg`DJt?@c$P|U_xB~B8bw}9<;Zbx
zcW37mJxFB_n=ldVKNurfjP4|E*l6fjL-pJ3zJ=4^uBmjOQ;BgqtTWIQxx$f2#jsG9
z1|SaoO<yv?>*F5kFB8U)ItX~lgXr<_2s&ztNh+!PM-Z}`KRe81kupCNz7dDfsiG<_
zVDIu?)|Pd6V40Y6k~)PV4LK_@FJaZ9YCw$YDs!4;QirXlSL?uO*(1CpQb^1+=#j8P
z5hZ3UvD#N${z6KhZ8?{JidBv1tHv#^Z)nYDbW*#NDtwo5lo|2aQ}W8U9w0HDYt6p;
z(U<|WpLS|!J}MhpFWH$+4V3wb&Lm`ORMKg3R8PYc1<tXtJxwV#1QA$_Cns?>9l@IE
zF_FhwHK#5Rd(E(h3v)rZO0sheM>Q@j>QG%mc;LF+9U07Fo0BQvFR19rWBf{M2Ol&u
z(kmMhtA9th(tU)hZL`?@Da=X^$kF3%t64)bQTBY6(p+hM*k3^}POp4T#|=l4M9Rz;
z-8h2{wJlkD>XGB+g~3_!GE<$KET@h@OP$RP?d}Ej&TdfC>!fyvL(S1<4!q9TDVJ>V
zq}rmUl39JI2tr(F-@;yb0{zhrNYP?#+o!BPVR#aHS=#=-^K4i*9<qtyDBx+5&73+)
z<<Dq0Qa;o05(ZWAMTjEtyV0qh(`WSG_lpC6Ned4gbw|lJyHoHqo#@3zMNNvQRto(?
z|J(aYkF|zSZr1l1r42jLBBIAD$|IA#FszYd6fR?2HlT7?I;ni;Cnqcb(LEivY1&Qy
zRVBpZ$P%#&Wwa-aFc>N=68p0luE4JE<e@{q(0!=^<=@j|_|cysv8_2|T${*%ZKvtl
z$cS`Lq1|V_{L4(2?`CT*Jws)GXgq4$)<LnWtRUDl&S7g$c;B3UdP`y{Z9?srPwV4`
zA;{E*>EypZHrvg3m&7Kt$!Kk@FRHf?5kGIS_qP`l_%dxMe`?8wso-}*FF7Cz^N2(9
zmUu=ZWG8QCYO$M~SUh@x<~ry^glmXSA@nWQ-aFl}U5^@TP}XS)i-{#+eSocy?f)_M
zj^ULCP1|m4Yhuq#Y}<A+6Wg|J+qNdQJ(<{?*tVU0&+~nI|9PL|*l&I$D{CbScUM<e
zcU7Gi&+%WmqN_fx=nPW09mL@RW|#qnDe4Far7}BiC9#h9$zHB7=QvLWY(#M+T{W`S
zj?O5E504{yQfp>St_<%rMr}S$CcGw(ZxBV<Y32B**)o5&!T;o~R=LvZn=-jCG@PU9
zpE}wm1JN_WI~z^7jN^t3*g)Kb!Jv6ski;(*g;s~ZfzENER?D^98(p4dZ?nr*RhA*}
zhp_v5dxe`$ze#69A5pYLUT*fKzuod*X!umeYkt#)a}uUY7n&_a7&(^Gzci5K3Pa=B
z39n`5&F_qk2IN?eW-o8+dj3i+jMAV%M4K+M;{kmQqx<gSWZH^s-0%K@pLFIHa*s7G
zbhNbhx4Tl!iyOZJ@9*#1+uQX<fus4Ksx3qnql1I{SK9*s{wO^5l7{o(U+pXaATFAm
zoZs8NqZ}(Jl!%*4YzRvV)=3jT`de99+05)>cF<hzK$pT-w!p^AB!1$W9Jj<sPLPoN
zoD7wD?@4?nqo_EuTxZxpHBFaaGL|H*JX}ZDLQcA%F2X1<Q;?5z97^_Ft;3for%=%G
zj09D-05lZTBDoNtv^H5T^+SG6W^=RKtUbAd0u88)hJTh<S1rxW-vN;rK*WQ5R&Xmh
z^0f#ya-+kuaP!M=Ql;?hUrnrL#VTg1(3#@)V&%*D#>@*}%0yA8qDH_e?>w=jW)&ra
zvfJu}D&jJ$<8_My1qv)%!~K*b4fr@$YI9J3z3&xIC->(;B%P`g;-8a|{y6!v5j_<2
zjDNcIMJsym5E+{RBm`U^8;OXIhw9h4@&RxPsG+8YvI;bfLp8)}alr&Zh|8t_{+vF+
zfrSDBZBj!T8at?R(ISFlVuc@yo*w72w3Mfl%qzEBE!kRLFY>IHLw$P$Ir&KWRKDfE
zdt!xwXoF=+ZTBl2q@uc!xiE@gX>=;TIgij^hDte!$8|5Jr5PYKXlqrhS_3FDaL_N+
z=x!c>5=9Q40JkhoHZ8A!77d_GTZW4OA|YY+Bdsk0qdHjjh?ewR1Uj=O@8n3;1y6Ua
zKvy+7h2BWeV#J)!`4QS@k08H|u?=tlg<5U(P&d?0C=!n`(3In`#IiPPAsV<x#MmLC
zeMccWMVLtt7m1RKEf*<%Kw_x}2B2sCsQ78G1L!lBS5&m%P<p@Im6w$f>+reX81)4}
z5)fD)P3Hr7owt3Wf9xQnuRp6X@cCUH);Oh?;z2*2@jjk@sM+m%U{Eo;UHt<Nx&Aab
z=jt=}ge*N11{j~!%Z-Np*Rqi<Rb;f8skSK@E(W&|6I$4))cU<XgV2w(t0ldzPIDF{
zXsm_5BWPO=BL{qqEL)AwJ?4y8jUhYo3-vFu0n9_HOldZ^>jM)(V^h=q;o)KDTDNyc
zOw6S3g~yK{{zwivuo{}0HEMMQg@yi2WdIB36?rO`|FQQD4rqnu^Lp}I&40Y!>H}Yc
z>565y-Eebvw-;tAQ?USY5A<__W)%Q@+7XM>ugcVRlkLYMFqr}aKYfM)l@L%iLv*OX
zK0SsmFsMLAAhY87cyQPNw8H#DURGMU@W{)`0`hSLiU$DS0jTh{wosQC7#cd<+VWdP
z{LlY|04It7GC@g1#Av=u@z0+>zvX}f_Q~^o03Hb(dDiH3c`R0{1KktWyYkV%3t=Pt
zrKF_1zPTBl`dp$d0}=|Yt!cLja;3nbuZXz#Zgc;?zxmI5O1s1dFe5MbmH;s=AOLBX
zCO9MnxDKJg!P^I<s47(Vz<>bqymSB@v1L<}UV3bJxR+K8tG>RThnJU=Pa!ukH}?~G
z-@p~IqZDBd!K?jy1|+$$qkw!D!Nfoo=$QI-*t3gXTBX@G;0u6$8fgiM;X~ub#>Q{5
z8BQnjlutEPRnd6S7y!0=Du>q-NRI66>;R2KK$6E?IS=?dKs^fY1GGeT&8$f8?dh6K
zEb1QsQ4nsG7EtYlFvJnNyV)OugoM0i0;xw(J6cpM6zeNHqj+lY^Xq+kz6}TUe7Q4U
zGg}`7j*0b~$?0%`))Pun(%BLLW8?ie5@9s7jDeTF&wSxfB?^o+n<K}^i*-PQ3f+h0
z3%yQ|`n4m$*-{OW4aIDHP7cyK7a)KX<Ag%z1Bh~kRjl!UzBNR;Q!sQw(y4|C$TLs@
z`N|gC&7I~G7r;vZep1G&w76LCOJiI7&e0K!(1`m8FtmGnL92+~dQ2=#Or!u?@g1O%
zu2&;@czOa15SOpc1W=S*vwbfw9t;#5u5FkY7%_yr5t)%9PSt?`zz)y}Yinz3&2Nu)
zK3z=D$Us|V!L0%^MgTRY3LYC57Y-i2g}cK*q2wM@+Qj7Bcr=4>^GqLr?y><OMF1{=
zOgzTx^Mlu>Q712OdWj+8*oo^JP0X^R+?2s+zVH5Y$#%I`P(s3NxmF*Lv~I0xj|8rU
zW4}E!HUhSw^`dm|{!oc*CI`K!hzJ4#f?Y@cRHb@DR(5u{b!)SKPO;KwB51^f2|F8`
zgUQ%pUl_yQLSJJe%aP@zHbLtsKtiCksRj<8X%;ub_0Z7KLlJP$aB(9tD>RpTeEGk>
zqAjm51oTo_X>S3R4WtJ4I)nZJskU8&b6G7Hfxad5RR$d&pdCmK4rqQAUP71bU?^Jb
zTUpUKlA2tx1<Veh6Rv3i$GVygR)3i$L594@07)J|cZ;*m8puWw6AM_Ws;Z*5%-r1E
zKq28vjK^IOHFMk7^yUQu-P{1GA<P=+Oppt}gbM}>@y+bN-0UIZabwU?lh7dunt?M+
zS=Q-vt7~dvtwSWsNOZsu79%K9Dz~$qD?zxs2WG<h+M1oCGyHTvz!3!+PQt3IFNWjE
zfEJGMcjB0TT6xU?z*64`2N$>hGx59pju;svu$}=1|4)+&xccHz*r*s7B4@dPiWSh?
zt1MkWSFj0a5c_^So65NwqCi($=wTxfa5;^o(AElSrJew01qX))35<X~jqsF|goI2k
zUl*Av3xb7Kn*$`*zxe~e$%7GvK8%cv0M+QF`Fc4u2?+^hWpp19pdC-A#fIRJjq+fX
zlA1b@Qvpz!14K5nvWde0^`GBtjcnstaBy%B4-Z-2rLgZnH35EwrnWZkw{Ia0w?O(E
zAm{$>0A`Fbu#2Z`)b3G4$l&1QOm7LeDhZ_nn2@ZcGJ1M>5fP(+($?76n0;hCWSWWQ
zBaOW{QprBxS=gF_etQ8+h6g+l%k(NU{>Ll*U+47y_v0B3Fwl0!#UtQ40$P7<DhuP|
zGX1%9^z=3BU;vyB$iOS)tsDLR&-2*9AR=O7VG(jUnE|3xFZZYHfINq}`LQkk+2_Y2
zKQJgZDo@)3p#b0y=rel*lr`Vp-uUS4fV>lt&yzmT5~$!eod4|X=HYQe6}cHhz@2xK
z6?gpaxwp~u>JjlduBP}u8|&)<yYl#WYwPTM1Mo-+i-Khv{@~DDlne&H!vK5sdi?^R
zmM3!gdk(pQ*DMA`ClCtJBKsBLpQ~T{hka$(Z*J>&at9G+%hXm<>W4jE!=*}=`u%hg
zfwZ`=fQgCe%vkSgZB2{u(a8jGI12$%x_EAZ0)^f(pW9X5gwK-_N3dMK(!%_Fqh!}i
zI1*YgHz%$NVfV!;mnfju2Fz<9tOBXPD%BYT)`4n3yGpPYn$!K158CJmNUs5=csvKN
zR?#_dT2)k3d|uDl4L-{h3+Z&b*IqcRoft?-4Hka=V5k$`&IKkSa3!;OA_3K?;C|qg
z-wlvV&+`z`$HU76^4d5q0In9B?c?RNhK&f&#0w1+_9qbz5%32*3VT!0!q(y8CE#1n
zLK893&_>6{t=`L3Yjh@nq63b2z#spcXyyBpdw!tcTgv<`y(3d#-|K-%U0_ov9t-4S
z*$zdtw6yYyLPMdw=sV$~4J|C-?rjq*M1_P5{)Hadd@N5u{y&z0`=^c$66TS}H6t)w
zx95P9?2kfZb`+nQ0g|`^okqsSfq{W!WMlvzhltlhr1L8PZ?UyS<~B4i0HTH0$IB^6
zi@##f=zzOICPF&d0uYs7zLb`hLbDNg6|LJsaKo!O1IooUx;-I*zzQjV$jZX9RrU7f
zrGPV?w>uO&5Yp4*n>iKG+vRo*HYyyP4>FJAA`}cPM6kh$2?<-;95+BA3}?w2fLq-5
zJ6)>zyxIn53210wR<xe(0Q{bSek@j)-5ucK^BR2Lu7`ktGq3k}u5`O&%PCBllY^so
zOxO1vvl<1kSpb(KFb0K)3sEptt`BI5nAv|nc{rHB;|ObMVYdS8%^Oe~<WOV+W||!`
zRA8O!rA~}&i{qgj5CSYd*9ejcD7QOMpacWWXiXJ?fcb;(mN|bTR%iiOyv6G{I<m5e
z-ARd$L2-fl10nXab$uOxIUSYZ60_LRWgBhYUSIozWi?d>(KKg{9o)Va0L49eBxvc}
zFl;t!Fl@xxa)(EGot<8%v4D^$%w)Es7eF6x)jJ~3JiJ)%02m`O!z^Yj_&~qmmrn&?
zyh18~kR>n}KtW)|;ib8$R%wXJoB^=ht9mkJR277%)C$#_05AZD>;OmtgF_)~$vjm#
zqgYYs15nhTy`@{j=N`Oa5fRO8+s}I=iHi<bm|STI31DDO!2YJ}5SZ}!AZA!+Q^Ui4
z=P?NkKGlJGY&W;JEgOc$#*pGL@bJ4wdYNA<{r&xG6Nj_mXxSVOWaJ9=ra)g&D@B2>
zw#grGT3PtK9vGh@z?&^G&6_a;>VX(+@$m2lq=D6F*~Zl!0cJk}lR5+e2dshXC^8@b
zOifH7&G63N2}Q~w=ctK<lyv8oYzk1kLt5}arXa=#jVi2}6XfOX0vwN^fWX##xw66(
z;5+Q#f`Wp29{~5<Q5KkIpqJIW0S*Q&Zs+QHctmVwletTy)rx{(z3c}MLUHx+{u)E&
zdpJoz0g}9DR0#eTSTX<wM8pq!0|soFDUfflGJ`>g26}oRtL5e8xnCUZ?fqK-$_9t$
z{QUgNf*?b{KN|{B`$7%kvL65muW@Pb<P=1PZZ6rQ@=pa?P+Y@P<j>HMBb=W}UvqxG
z{ugHyoq(5?joYIcG7hloySVATMORtai_ISSHF!V-Zxi?!-0H?$t`}Oh(5qUsy^O#*
zdrglH_`O!p__ub}AQ_Z?R6;_e)pNQNkTJ+Rm^1O6uVL#2qLIMc5db0*pqVfW+AEuk
zn)($GFETMTb$r0*bby%m7s~evM#B2?Wnc`LpUrH@oR9(uK0q5N+w31uncYQso=lA#
z|A3Mfmzw1cSNbEGWGP@bwEa7u6<86sXt*>8AcBH}?P;zT$8&f)3UPr8gm8ct9~ZX_
z6srJJGBGi+PtP5wlgi?B^a1pF^YVxbdVvK13Yicgky+*=PlS+=5QwsCH&<8jzFCWk
zilTG_lb_)+%rx@+M9<Fd3h-cnjNlhEG_YCN{cU1`wsun^BLo~K5di^WI=X*V7*S2Y
zV%fxY)D0BY<pI#fsi~>l^mJGcBU4jgdbKV>>VfMJa=Rdm>%=u29hm{HXor>|bi2aZ
z%1TIBI6SoIil|XWL4kph5f#A{h&IeBs8y>`_PK$5G63{au%3Qoql>8G^n9R6izZ&P
zei(3u!DpkSg#Fa|XSS>Ul0HBav$qDL_lwi_@89J><3a$xg+A%}+mO%E(Q$`1?UVBr
zi1>PEk0qe}fpz=l`r5u5(u#tTvOLBDG8ky-7D@`L0g3;*cn*vO7*HFu^~4{b{&KQV
z!3HVt6t)@*g0Ceb6J<CAbg<gp9nI9{HaHDUOjrX8tB8JF)xR2!&Tl~UoR)^|W(n*B
zvat^opw0eFZGz2VTR^7bKOIH@3#1kp0xv^1DJdx*yBG|x`u8RUGF5pK!OVbamaiM<
z0+Ot`9io0!&_lrXL7xHOW5e<x{rQ3cm3Xx(kD>yF&;{(8UEaw+rkUb?0s1>Yf!^n%
znF2cp#ynz1aU>=%m7obAME~%!{o5XS*#h({LBT#?==EdW{?9#O;o(3Fa<%X0he+6e
z;tIWPYJ6amYh^V%H)pT>|81=P&tw1ZHVd$)j-5}Cjs-y!IAa?lM<)kked~X>Yz-{n
znAjLuh#83g-Q?k+moc_6buuGnV&h;X{?8Y^n7Ngcu>-xBmA;d)h_Ruqkue`1-2Z(F
zx0PZIX-E7P)b3l=vv@8PG4SYA*@lQB8P#V}(HW8-G4d{F5%>fK-!VVmyxb;0mjCwA
zc&pL5LPgu$-CBa&+>J@*ba%PGxZMs$9wee~#3`4?^5??PlYb=q7Pgd&kx-K@jm?$c
z+>|b-;gf%=P14xiz3Lfiac`a5V+^>8JldXQ9K5D~;@Rchw-2|$$uwbjqUdy&v2|*t
zFC^!~BC8s^-NEtaJ`N?~^Km5`Lo0$|`@rk@qr2(V`TRK1N<2nBJbOGe#eNNGiMCRa
z_%IEc>+?Eve3;o2)j-!tE)f*&d$noXlNIuvC_*YQNUFH$j}z6`bT>7;pqQF^DV!63
zq@unEL=63_fmWM%pvZJ3EB<Sg;UAnrMeGxU@!t#}!q<NeaFA*Asrb43R%mZlN!dl|
zhK!f6B1-MqFbPJDN+p@wNqx1XDNgIJ)zy*3`%ZHkb1q;P8|?1l=hEM#TyE$J15y3&
zIWz)W4$hWG94>y;Q-}swW1EU^S&bAVp~_w7E-s`fqj#4-Zy?~JV-uqN$y3xI=AX|i
zXRh4dLFbPBWZ#L464)A|?D-VT@#!DCWKe_lebL2zan~o4|KxgS727v5K&FpGPrMrU
zL?25;>kDN%=L((?6mVRUw^v{aXWuxJk0gdS$znH7hPQ`MPxjB1!lDfAnPvak$BnF}
za%1uL(b)iO4HW75d!_R|#t=S>oMoVAbYM(;+_>r62)vBN{m*RHl%TrE`+o4T*Al|t
zFI&=VhXxae8Lo+tron5qV(YurVmx18H_crhd1JZw@l7W#>%F9^O<PO^Zt+d8AB%D=
zFg>*nqUFZ}3+i*KsYbY{CX7;bvku2Fi>u<&9dHhbD6ZWsc(QPRRpwGj?h!`xQ12G*
zLx@99!{jB^Oy#g~;z@{N?|^4GeVuGAkI;7BY}{e=wqAe^xrJbbOZlzfd2o(-LM`jP
ztC`3$%I!!S<S*o0<~0P_+W$A@`5b&-!*xh%^N=U`?mkO2|3v(WR+m&w28;(wu>57P
zU??i>By*^QiJw+ple+jjBfLo@2GL4v9hK<MfkAa$>sOgBOV6V->u_hLQa`Thf~F-k
z)$v~ctI)jQqGBv^*j=mSl^0nHWQekO`C_y)rqR~-)l8uPII@7bZfd?VCOaNy17rnS
z_z`~U@bzz?W?MNzvS|<qh`b{m?5-LcKKJIisvx5~c9;!%!QVF%JFCboq-`~DTYF9$
zeO`CzzR-dz!L|zYK4srGSDR!ayQ3P}uCMYxJX35n>s9X-`Z0~@@nUVL^uIhj(tqCW
zXQn^8xgm<ba(NGX3J773LLq&JANgIN6F4ToR9A*9RCglzK>N$!cxm2JufV)K_~&%v
zk1Ha+xyHZEWr9WNIa8Z69~ZBn?fxiG$3B~>a)pYGED|mBni0wij@PdNBnruR_@@1X
znMsyhXiZa)U15*S-`&~oWetMx3J>i<wb2>>)D3bbE%Kn4IHf1+9i-?Sxjc1*OKar&
zU8&*e7jqsYpknGR;AVr)93@P@lt?qF?0NF(%D*vwOBZa+{pL;y!!E%Fiy4}N;8tnK
z<R}j9#7AS?KRD1bFTms~9-PxjHgastDzaAMKSX-!=}coUP0jK%l1`G^mnG8;HM?^^
zz{ud?yh=f(pEz^v5IhZMVqHmwrjO*XEyXAS8)2R^L54;jb=?OAfkUuuwPs^2-dQ}_
z7jn$vEv0BN<=sc7FhO0%bw)-k7iWl$0QHAR8)7N*g%(!{Bu>L6ihZp5;guIVD(D$^
zA=k6*<|Unaz9PFuw6sm0H|lp+d^~R{D|RCB!SP;(KH22J+x*{YmUXq)YQ@2L^)N<A
z$p~Vr*uC<WeGh6g0_5iPydy<cwO_Dexe^TZuw_Z>`+khm)GJK$j;xNKEy0e83*>$6
z3c<<q!*#m$Q9^fKL$;*l>03u#&ZpVK_}Q$fq_Z#)K<8(Bw@6Sb2`0|uGMupFsALoz
zITK)Oi9BE%DMgYGZwz^Cc(n?)4~AknLV7Mm;I~=U??ImKY^xShm}3OivkhxcSq1iy
zw)}%sixbqs!9;U)RZ{oQMep7C0ZNviE)61EMzcYsIfgxHwy^hg`yn&*cAfvZgbDFl
zLB133NaYm?%Di4-_2LD66y*cM2X0DKbLc+t{;RoygqsqACm%j~pK(p{!?F*Q7QGY1
ztsn6&cLeMMG()_^z^E{XC&l~kk|p<0TlhWEOi+7J6fg%Wz1x_-OLNiTX{THX>fnBj
z*b9rJuH;(TQkdn2thV3=c=PdI87{>uNaT*>M_H$m8#+WNs<oA%Gx435T*#Q7#!iHA
z-u%yj*`W|>;E*0zmo+0r!{JKzl}L?5sJQJp7K=W1R*qZ|;n3LOZzdiI>aYGTp5*B*
zL_h829e%tjW?jIgEF5bjEOD&cr$o**kvGsFe2Ta$LS6OP@J-Ay<mXT$(qSTea0P$+
z)0$(H6Am7sCVu0S5s^K~eQp#Zgie8oA*JR>z#?cFbCI|243o|<Z76-WDM<xIJc9=|
z=?9D7ZOA<IH6PCxBq=DzAVxCkj$jb;`b(f7jx(e%R8d27DaYa)b=qR+y%KD4AhD{3
z$He(ITB=F6Z+ny`Ql%C5#hb>Y2Sbw$?eCcqUY`k`vb&wwqxzok?|#2x)c0$HPle2p
zhSwfmuB-TVv)P!Kl4)*v7Ra{TSk#&&_n0Ead_dX15Udy;^Zc^$HzJG`Wp{ox+@{N&
z{IlquySmV5T6E*?d#}YRnps+kfs!P1E;ij>rOf6#+wK~RQ3$nIRXA;XO>Rv)lscwp
z$xBiEjz}a__Np`a)rH>biGp&PAy!%wsi+~Y`tOxFy9eF6Z7VV17J>W8;A<m&PN`G8
zHA*a1;d>^o5vY2AE&U1&>1}=mw$YAzn+2Vx(d9_!O!R(ep|Q5tCd)(1dYExxO=kX|
zB@cn~3PxPptt`o#>a~aMW{*tsxpf&p{>JrEqb<jijxwPBcy(<rB&$(E!=U8DqS~W~
zeK$C6p}aMHp|Y)x^QGm~<Mf)V!=S@y;$+-~SkS(%@SKf)VjjB+|HU#OwG+GtGH!S{
zUkA4HA%~hW$+2dL+S4ot95-Oyv$~^Y;)XyWVhGBl?IzwPiI?3K(~Nc2(#FC8<+rI8
z-TFKWQm<(4$Xzpph%?Q4>mOyAdr>-KeoBL84V&LC@_bpFumt<F#`5aPo~~2R8je0L
z+hX69aN5uGzd`7xj2zPlz9br*<TSENI`iDyn=+?cW-}vgq_93YP1)>SW%c&$XBIcG
zU>k5xYialhm~`1v2bG(Wj-BKAP@)KAL)-gmhFR{o(`-B`mASz>Lo}wY3~_t^@xVJL
zL~)f|&9vq7@?_+tz(FoHSDT}*ACPe`^9dl9o+=Xx!~{!Ai0TMBa#hay=A_`EHQ+zp
zO<uwFjC&qZ>A@SS)O2^;jZIto?eBv>8dR}0!aho>Ll^J!P3^|QjvV!2OD_^8!fs|l
z3Uc8COS#b>@-DSi)u9IQz0>xN6%~mHxoAt;P`$C_#hd7aE8k!TVUhn9$}%zj&rp_$
zf$4u0$}VV1C+@N#bwAcj6PeM$qIjvaX)T;z<osG=JJzHsJ^PlLJuqqf`2ps2B)FPD
zE1fvfSA`|%-wTZbLMh?u>(c&ubNBLm7mM7Ot?C~YdT|>&X%~9XBj#9GIEhz?pK)<(
z`FJX5Y4_&2pSyq?@0;_9>{YzWoY6~$u!-a7^{_bng#3}+y18e6+~#9dAn*o$;V#LG
zjclBBkh>c{Y1+PKlBE^f<Na}06bYjt`u%gXxXbg?>*Z3JI?l1I5qbSmH@9e@KoL8S
zSx(exy~Wqnm*>5qm@9-DJ4BjWoW+04?9VjQeVIZ#lUB-hT@udEl+Xv??)Sf}wEVs}
zZjdUIK`FRL;YaUg5XQYth4FoWteKuVBzb4ht2;>zzP@z!53l$yrYiFvQI3r0KO0Tv
z>KJbggX!5|4d!acOP$8L_%I3S%X_aDqN-DWSV^egR7K#ZG#~uhk<w=<iYtMC`-Du8
zp!!z$e9ylA9{K9CQ#UO8DNWY%&~S9YzfNx;1n&fn7e5WTPd1LaNN_b@7e^q0A~vYE
z0UT3L_Lli^$fs2CDM`D!q+>%G$Ou@~n{22I4wsaWu^B9@s(s|@G2ob!b?4h*l#*d3
zS5oaA%)cuUZhOAwtZvB?QdUFVIz0d|rs-S>9XwKatznH6KP~fD3MlO&Vnz@1Ql4zy
zuL<U9^lQ&+qCUU(DHG%E(Pv8q+gW527~reN5QvN)ln>MhzjJ$Z+NBVdjaf5<j1-(q
zST*s|ET_X#sl86awWj`#w#*sflD5S1KGm{;X~d9-kL((1L?eGEp{C1}Wn+HT87N(L
zs`pdgOgJm;aU7KDZ`H+`DjUT7RWqn&Ecv%v)rwBXXJzT>r|h~3y8NRJk2n+Ol5+V6
zKlIHG)ui2amO!1dnSHt=COVyklf7}u*pVT_DAeYp!3N!6-u_;|Fy-O9+z2Emn?bh?
z{?-g*Vo9+NZLj9UGjUTPy|6A=_tbeTw9DSr_H!LuZ9}<84^ut5h$x>8(F!c~k{O}B
zm1|;aJo88_$^-2AZyn`DwK`JujIUCUF8QIqW8bd-9%JYm;_R2`$8Fe-AzuYRMu0v@
zyOVTRK^ucwSts9T9x9$yb4(oM-dt4koS@0Ewn(`PkC-g;W)8%Okvk`;keXi=s;(s$
z(<kl>9yx?6{S0KU!N60DuY08!6e|rNOzG8xw~9QuPIu(Spg46gLwNF><a0!!&2K94
zqyK@b(3txZNmiW08*lkkP#uKZkF+4W_n?=6CZ<mgJM2v@0mXES#BPG7wWW0W)L#u`
z8BcWP5~8++3Zf%961AATC5KJ^?_gBdhpRLFslTQ|SRy&)SrJMo6__lI57-Iz%)0g%
z!D0_?5@fCQUKaNS1e?9b{+h|+<tU#nM#wv(pIgPH*?k45rUFFCxGfQ{nu11X_)P4S
zq|Ptx`aWn`UvC=i$U(5*akpLG3<uURS09EPDR}leh{Ish-wy3wL1@MCl(WA3<Z}f_
zTjA_bfYOeXZnN!PI&D=js`qLfV%QK>I;%jntdcyBFsk!fryN`!^{Rri)xbn?F06{8
z%826duVX%gruv8*xvY?+BX&-B_+j}Wa~7{cKW$-RcP>=j+KjyI+j#R5!wbA-O`d}{
zy26P;m)TZzly&bsS@kW4R=HF6bQ1!fFniC6+gf@iK0qS#*;$n(PSwAc`^j<nZ*C2=
zg&=4Wdb>A<z)zBPO{N>!%6I+LdfX_;IBThchrZ<kq07$=Mn_sc_*&t~(}z9+L3Q*G
z9TG=Y;6*dDTF${)u#La-71{ruGIA_ljzy(ZHLMF_q(r}pR}BUX?EIVpcWi3;QUx4T
z@Ym3bb11#I)IGRESnLZ!Sqps?<vrB9l!)>vsRi1VMsBtWeUW|S0=zp$GF;@iAP4UE
zd7ap{mKXo&r`=Sa4UHaDv6;GV!p}HGH2k5z`eU{a2cm2=j(XgwoQaHz`D7}5r<B69
z$Wp90?^@+H<YPud%q85Fgbd2k(kP|QQpImu=>#*$0U(s{O?1KJWvnvh+u48pBRfN0
zXe0G<sPuJ0rTRRin0|?-tEAJYq+*w15;rmn8lvuS{EFURT5MP#&l7aN)c9ptA=&FN
znt<HmQuYn>l+z#HI{cfKhJn4B335^#8#5j7t?*T<-K8(MWOT3a>M$OMabJGNc9v-e
zIm*g!1y*z$ElK(Bjd<HATOqyrwCyfgv&L{#d-G(^eHDENU4Y*=Dx?@wp6m-KBLC=P
z1Wz=7x?Or$Jkd@gqFd=g`sVgX-i>3tJNA{3EMp15@%7FL6+-O}xfF7U9omdaJER7C
z^O|v}aZ9?;f%IDF89Of2G&2RhJ8?tlK`<+$54(p9O^B@e@s3detHliAC!R5}Sq9#q
zD?qN+$N;N3Av4^-s6b8~W|#NP9g0RD=E~#oM_aHlyRX#BC@-*1WBR!x{WRzA^ltis
zU6uyP_J9Hah$w}pC+T%%lWeIR^pZt_QIJ(#cX!=Z;rI9919ZM+#E(FVM*iq^9p2Mo
z_I&9tYSm1dRgTC{4S8$~T_xXnvRSjkIy4a>EgiNk`SwR#n7bz_rgTcrYto23r1r2w
zRhEqcyNlz7SxCJC->Rgv%<qdYvoDf|c!hKuQ(Satha3eRT=<;wJOdfI?OgV%f}?+8
zl8H|l5r;df!YLvIL<RPYPA>-;vE8hJo-}BAHjx{A!keW7nF%u`XQ>c(*|o8l6s+%F
zOXBRUh0VcG>I(rb{P6J^6qtnxBZ7rj7oA?`e%CR9r6SdM=JDKf3;l}Iaq)HFO>iC!
zl4jbR^EH5TKV^4f#b+*USh~XvNc+x9N-#wK95Lv4upMfpjC6Y44v1h2ttt4EYojA2
zY~)o6g^`+EJw~=A&BZvZeu5VsR{i>~aSrSMjB^+mSpHXW&bj79LKYiJ_oea`Sjxi5
zqHeNfg^4L|^$qRFh8E}7I{Isp?!EgtL?J&tKY#tMP<(6?0jHK1-4|?a?+T~;<L%@9
z?RUuX8!EM{5|3vM(9P*L*bI6uO?QxelwYZr3#rH&ctb<zL`AHmL^(ase5*sQF#h5o
zpI~lYhp<L(!LiJnp(a}IT7;U)wO-iWj=wJT-r~f}`S4?%y7R}D{MrJ&-`@9c1Lc${
z_MLMuq?7lJPg0&W=y`Z`>kh87B}~*EcGBb4nx3n*jOBh$&4^w!*N+m6G1sN#*q;;N
zOkqSRSpQi{(%OkSW?-VS-#Yzm>OC?)XXZe<;8$O9`1*mSH!Amy12SaPeb<Z-5!sJC
zSi&(Q%XYweg?ifK0PHKfn~haL0eZvmvGVol?$`jop<(<Fw$_u_OJs*Y6bvY3e`Qqd
z``wz8(*gRg3*D!SR5O$6+;-a2G@KQ`!5sU};JWRh?Zx4<Nkj-R5^=SD`v3gBcQX`g
z`@?>JDd)CMBWo~$EVQqvf=fa1)xn8Y+Cjo1`};^)E^a6)bKM-vjk}Q`vG@E9V)fk~
zvz~}2q%dv#p^}!)?BDQ9o^D5tPIdk<cCyh(?V<vknuwvW?+X@%<QJUoUswoBOLs9z
zs8|>hj=dUgW@?$EootPT=3c~SNB_*A&yhHGaKwP4?MUhBmB76ZtEvVaIjMo<=DX$~
zaKkC8FM8v+gu$(rS;7L;wjv7~{~mb=Ce1k#xIy}!(-Zmb=X7X8goU2_>C5!U>W^xl
z#9AF<e2&`ST4fQJnA+;)HenefZZxdIo!;*3-%=xuKj!zUPmMUyMKx!$WwrC7$k9x8
z{?ck3#;**q1_}R?M6U8I*_7q7EW|RT4}a3xe_o#7=2CK24gP(-DwRy>Yn6b+LoeQ@
zQWw$?MR!`n_b9Hl8(dY%HiN|}f1a3NSsfjRF-O>j<Q#$rA>a}|(VY1Rix&<)wz>Mf
zRl=dIHr^1?YBEV)I>p@C#-SUlK{|KAN#83g(=t0HbaCvjuR#9&_ib`4vnoYpu9}8k
z7g<_0;=Fn~d?V0{rU2&-+O2jE>bjRinp$;bc&sZiP6Fh5zZNQyo+YsQi!#-(l&|mP
z@K%fS0g$adP7lQ396K94XGs3@a#E!2+Ru9(kg^9*su6RvP=9~5LVBuIX$1u$7$}Mo
zx#AUR552M*$4z8t*22s^-kzTzMWRqyyWs65jIwl623q}G3lNl<yCgJ3VevwD`4X+%
zl6)URQ?Kp_kL6#o%JYX53gavm)yW+E=BDKT+&v@v=pH5dBuLz@sf}zqN}9>aZi^Zr
z%kB{-N+AZ>w^0$3_LM?Xuk9#}#YlBhYua&Sf_CaM98{TW`%rTpm#k?|t$_D{qNA|~
zVQGdXjFOu8ONXZk{5#00S~?l0g{wWSkPy7m^$iR{5=)6|i%Bu#$~><FCqHa4M{G67
zqXWzAc)NC^dwLovShSIYdxe{k@{YP*i?+891dBN?+khM=y8u-Dnw@a>x=4e1I8Obn
zdNyQQviZ>l#1vv2+pt&@(@N>d3r$;TU>n6XV%NSSysR}uyb=6*JiTj!dQHxc?k^&w
z15Q?FujBLWzTgVI`zelkBYwA+<R;|A(DI>9D%-3mwDP0w*f$*^rq6i=bpv&`W{cB(
zc%pVxD*g_c-QX7%uuzNY45!=G`!DU#aATTVr4ad{&8ixe!`@Y;=1XjN6a`)V5pKTW
z9;4-`*+K}#CRa=W$BSjJ8b^6$x#0C%OaxqynVLegNI7_~9xxmaRpw6g#B+G&q-{<(
zt49PwGlUcvzq>)Dq?pwx6e?docEY*!^R2Nc%joywm0pDF6wn3JC5-R4J_bW-5fLyy
zU9UkHFks!xYvB>P(t#PHKrIdJpC48R=Au{T*&%SG)9MJH)~5Xk&lapU)JsyK$r9qY
zA*gro;(YYLNp}=8HHi-qKr$p66sR|h=hWgJbKA~5<2b`1;jBT(1H-RX(>|fBHH-fx
z2gLJaV1j%xS%gGd>r%(`ae!l>DzEDUsf;FOR;WcgErX?`Bj7xF<293xR<9lMy!<nU
zg16N4rk*P{zna7OGLh{v_n?&i9BJgrd%+{M;4|k)WP7>6k4;OrFh6jkfA6kx+<xp3
z^w~1fG_t7zQFLa6YIzT|BY#~E@zk&S4s<EQ`hr!PVt=`pT>}#qzI%>GeH*TTC`S8J
zY{@zBzJC4S0FQ?9bUSCFW&8H+&rS2js+p1xUm)$lqNZeJA_8g45LE1O*I#E0-yX+H
z-nsYcbB~rO55dg7c^@Z<SJ0dKZh|w3Xzu0Y+fACo{<p5qEFY9nm<<Oe+)h8f3nA4>
z--aysTh})Pf_a|+GUN8-GbidchmXvCt+bx~H*;4KGxkKa)W5-%R!Kc_$gqE?mV&8r
z2(7UNNmc!@v|XZhb8JuKy84V?ymdcn;%QHvUXao6t{cBin3&5Qt8cz%9(k1Fh-Dwp
zPG9C+BnpEKO?(%{1aW20%~Onj1eejj&$783+p8g(Ow#&7oNbmAy!6{C1mO#tVQ8SO
zxHA~VYQE;HAOYy>hq#kin4B&xbRVZT)+zY04|DP$dK;OkRQ=B#X<?#F1mqo^9H;G=
z)bx7=_XDjbqP@dg!~Yr~aQ@E-f$@JCAxvnF#O$#lbUms*L5rc56JWDUkIjQPO3$_s
zmy$E5j*~c!T!zl?h`J<OIpYlXS*Ow}?~x74+HiW|?Oi=@Kc8=xM$QJJZ}<&tpPz!<
z3C!xoIn8DCfx=YixxMneJt#^wI`Pcq7pp`di7(|$9$W{4$&>l{fb#g*;kvyA$9(ZX
zGS~95#mA$>=8DVV8T?M=HC{+gkRqm$dNj*h>&5f&@vtvwf?5QVLjSDt`EvJgH6=_N
zw?N$6e>DWu6X#*V?At5kFia1`_Jq_R^Dn6!+PM?A7B&5$*%B&{gVDeX+7XAPdrnBh
z%m|70_3^^{#ognT^Q4l39gvZrPpnV!VqvozOd4{FSBWc|XAY(|x0<oV{=mU6hWbNS
zIh48+`+KG0y_ql%wo*OG*Gx>^A$K@2h^`!CXt5R8s<o#-jy*mpNyFVWlU0ErRjQ6-
zL4o?xrC*J?tOm^m$97@pbHuCNkJGs046l4xACui9|D+VEe_DN6F`-#z)2Xkj#ZDbU
z{wXv+Z^dd&E0@PSKA9Q0>4!<E-SuStEnhKtL03YC@6XUPkG3rb#`GR?ocYH?2mVJ<
zsID19O}g*y^yMy(!;|Mz^15_j?e_qgv>Xoa$Xa{_2qLDx;X#Um5(F`c%R}j~f)ETX
zBeU~;^G>#H3K1;B)xuQ%O3_gFNeCJ1OGkfhY}rtEBf`t##Qz#e%!Dmpd+kUhT&7#;
z-IaVBb<*s&WUG+e3%<r+crCWcJLc^&yo@eE72`6%g-p1m1P$LCiYuTKS2uSV-Iy>G
zb~zLwCOsU!7%AvfAJ)0SbQO58A;hcK+|dqBG;$SzJr{bAo~+#RJ6q>rGh+Wa+%RuL
z;aG1&*Y1%qJH1};Sic;ng$D}rJ-nJX*t`d2{eb4(LE!G+Q&7vcvFDI?6wLQqadG(q
z_hI1`7)DqehQ<6m0?~}jW?ZKAWRIo&TP)zFO6r*QyL&BqWOS9}UE|{C_i%0t^Ybd0
zJsD}*Do4VdRzjC7q*EoFA(_xyj7+91C-#z2{kLL{Tz3%)d5)g+Lw=VcycDu%8u=Z1
zqTKp$BVkQRKPfmh*f{J~E_c#aMA$+7{3;E_wKDagyEDD~n)s|VP0SaSD0hYTU_R4%
zG2|7Ch?*A=jL?V*EvBChd7bKO7IX{eM_`84hDuw~=?;?_vh`swjpx+8j@#-JZdFf?
zNjwGR;^I0vrkowQZhHY~_C~^aqzu=G7>r=g@_V-SrhkY?i%%w(m{Z)axfN5rVy^9J
zm}^er#es}(ncbAsh{I#^$M-K<B!#Wo^#OH-mPM7AAo(SYg=7V+XjKjEVwm@0OJ%s;
zN=s$`40w;bv0Hqpte~;bQf11!r+Nld$6=}Lpt$GXuN}ZQPs6(P4#Ogrz+h*L{1rc`
z9YBJ*3Z7&g7Lg3I@#;g{eNi8coAZ`R%ug9tETR%!-vayZr}RLX?g>;z42zz;1vaHA
z?1Jb3#mgF_x;9kfH>mYaHilY?BB$^AdeObL5_kpCdiBA}O?&}@>@|phywS^{o*u=$
zN+cB8>w0U!9gX5@<bNJv$=%$)K~h>i-eb+(wn+xIUSGa({7Pq|k`}E-DS-m%F3H1w
zRKQ&DSV^U-ZRJQGI54){-8VV@dGOn>Up7*Nmp6kWlE_k;yV%-(vM$CrRA@x_*|_wC
zrG1x^f99O-)G&1}5A7qdkf@yYmqkyd{TMbo=ZBK<R1y7UaQI{EnX0+HuYCmWaM5^|
zYrdoX=mKji!xRp6t;g+B#7$E`Qo=V4`USlv1G3`UH(t=)F<p?WruubrV^%}5s><hq
zEDfyH#OF7zg$d4lEl`^P0sQOcXUK0siSE%MiILZ7J+}Tz^*>LQjHRoY!E0NYq4OAo
zCK1kuVSlCIY*zA8S*Sj1@2#E8?V0R&$<HapDBdbX90(aYVHo$>o<Ydl_QyDAj_qPL
z77(8=!Ss)gA}j7E3nWX6V$Qp8``M?q_P?+Hh-$J&uibbkNjt;5`@smKMcMkQ$~dfx
zAAFS)=acqZZn6lYYT|B#6si~tZT(kR`YS0vZSeDK7`%rT?$NLK=L+guo<S}e)Hd=5
z`nDr;5uOXrK~bU|gP`%U&f1D>B<dk&jYU4@d!Fjo)bbPw+1|Qnw%e@8tCeB84pv9a
z?X5t&j}1>}iO5w&6s0?w@yw<cOl^jJ<n3=nqZ@1&k37k-r$=s*k0pxsLD?BOR94Yj
zQ`b9m<<xk4%<?PK0{kk7m>g2!C{v35;-J&{pbh2zgQl3&cVCDID|6_Y4-VsPx;(pP
zR}2DOMcwP5J*;&QpqtjKl0%wlp}#oCG!raHl!&jpNUyKT#^Y|CHw#v6o%g@^akcAP
zmJO((?n$D?>J?{TwzvqRGp$UZsB`dEgU?}}eha!wR^wT<&pqH{v@rn}*YJHGW@Ra8
zbK)$IA2sLyRTn>+f>Z;Kp`-fXZH#$jZ;J4^V%K7X_&cdY<g&<wlEQ^{j?RwG_h;Xe
zQgG-Pn?n?Co(iqft=cU4{uLU%bL0sRmR4!IeENDGdr~;1G@p3-Q4m;zl1VmumvFjf
zkqcF#kce4)dnr^$yp0)((u!vcR3DtNRaM^)oz<E?StW1WHF9TpL>IF4>AWcQ+pgi5
zvyD`%ZMLGK%Op3P#_3TaxK!2IzTNG5&FWRV=cL5!daf?C71!WV#CZ?leOaaTe22^v
zI#Cz0ji<x-q|xjGiO_yxr@@izj5YM{HE}25iI)nx?2{EXmd+o@SfjBp|1An({-32q
z%nY2&|EqI~E-mR;eD;6&m@@w9`VC|~hGCgiLt^&J7py%Lx24HG9jp4#xA(92RN=)Z
zzsfJd{d01ZBr!yl_Qa(O`Mfiq?ym3Np6@P1k4G47gA(TsYZCpw;QPQKMK8+8!I&hc
z@qMQGxNm7`Jn;;OJ62<O;lEvMukz$1Urh`5CZfcemO?)G`cFY_M9$O4RCtp?f)l~o
z=Ir2M5<1W$5yA6TNjv4s_MY|ge!M&%qEAVqT+wD+BInBcW*w2zER936dq3Ovf6;&~
z>CDQ~7kWCL@|D+>&!q|$N@tGk>6?lrkeuR3XvFF!8`1UIpX=G$pPE*O<bTKP_~^aJ
zf<!O5oa6WBvkkt0gvr*ll=<$YpSFz@23>OX1Fap@zXDlOOEHbkn*)xW9JOHhN<A<P
zAxv4AA8`sQht-z{eq;fwyg=GD8v_ARvq%R`e0UvGNijY;je8#l6MJbJO48PB+V7>U
z2XB4!GW+<aPm>AQ(dOSBh(s|^`=LgM!5xK~9p00RYb#Ehy=TdObm&NCs$-?V9K<8-
z*ZG)l2_7R|xO|=Uh~bojA}F?v;>pG<?(O0;(O<tq=Z?9>#*niw7YZvG#p-<D-;;j(
zVXOFc)j9Zt&zKPFC4{NiNNI#!sU*l26>GFnosHt!E|W?8pGkA9k(7H{T-6q_vxIg3
zDQBO}n;u^#Lv<_#dlVVZhFe2tv{l;D7qq27=`LPcI+nunBH`9MZz6BzpawzaxI4i8
z3--c6BBQ^FBx99jD5ts<d$a(fGPA3$q2i>HF-mkx!RROUREUl-JGG3iSmoYdTBJiZ
z>Si_g0X1+O3X!>kAU{2D&2*6<yW2K-iW}J1CgaoW4yEa-v-17S)+itG6on2Tpc<M?
zMpI`EVX;i7j-ppHG_uT%{#kR*3WG$x?}FIU_sTh)jV7;Pbva;<p_uYn44C{Ie$Xm}
zLzwJNQD5zpc`x4H`8S5IpLi_<zM?jyEid9n(#E%a>I$~6_-De>;<u%zh2)|0#*E)B
zRG!ULf5orBH!Ece?75}+w~e%0zhJ()kG#K2Xtw=G-6WrZl~^XhpkWKDmzj@16zy=4
zMw;^(zJpISkR5aGtbDQ2*x}kR{hl#2lpJJZEpa_t#lpQY+I9M9z-&sv8(J?%8~r_u
z+iUUX24`fvFgJ6=GtRHeL=Lbge4c>l7>sVdgi`&BN`fa@H(5v1zAUt_uHRH1gRJfL
z#e}uuWLDd;4PwIe5rt1;`FmuG_KAp0bNn;nQj}=jCGMZvS`Nmk8jhj`*Nb4ZWx8+$
z&9Yj3d3^~RDhGe{ksv<92dH0uVY>OtifQ;D`*EMOUQlim5t!qL8t*Ie(Vh~*U;Y(>
zP0&;XPl@e|9n+Nv$}g=Er03U++u*jMLW^(9r<jyTs+Hu+SsF6F?oD}^<fc-zBBpWA
zb+ySzaZO|KjCUQDZ4G)vJsJEeRn6Rpf{+9Kr<`M2eQ{-`eQpQ8EX^kxtfVMGlvTaR
z7U<}?SoE)c!1Tk2eyn-2sc=JZEL)Esal-GqZ-Dvgrqr>=&XX{pTO^}dhAJXWosxBi
zPS76?r?aEUXAY+=+1fM`S5fXaO=N!+_SifD$%Z<IlS=L(y4@oA;i^v(nN@-x532F&
zC$}-ksS#=G;2Y#K>569t*@adVS8uvXj``(-Lka=Of*dSlik?3f-Y~aaMXOQ7Wpd$8
z3Qj{(;V1>=Wj+@Al1jqv9+WS}<#ZE>{G|>UQLtSx>#YYPb;xj3siw}7pI<#`A3}de
zTQ)y5Y@X)XP*H<-=k|xf80HIGVy1Ca8Xc4#?it;)FpuIxGj6@)smUHP?_enE&x$pj
zDLK|7(GyNNEv9A}&Ix6)qHCG28o`w}YQ>e9dlV5$8>5Pi!T6%Hil`M+okw8xaPq7T
z67t<fHOZ#hKC6*l_TQ0IZO7hJO{FcPu<F$ccl_PE)e1HC<io2`x|$GfDlO4OJ(Ilh
zw=)v|mZ%gZy2KXvWQ|J7D)I+$#1h|#&DRtM5uai8sZNA9x8cL|!cL)|R2=b#TF$;A
z&iKN{yhmyF)m^@)9Bpo?qhw#XT=D<B^gvVCf!p2{!1BW$TMlm19s#3ysT2^8dc4)2
z@0kKYJ1On7^^DatFj>zNswIBwOVeR=k>5vR)Yj91PE`7M7`kgehqYbOnx<wdt9Qux
zbt((h^m8Brw{p_t>=)dx%kn=w##v;Tdy)+qyTjc(hdH6NZ)=a;0cW}su?^#GU-@@s
z#66~yG^A<?&4}V9{7`W2biyHskOj&_stxW8b&&h>U@iKV#w;i*eG;x*h;14Dj0ff7
zRgUCKwBf&`cC@i$)|MU_SWpSKco7C}5=)q9b#;)z_a+y$Mh_v*Uh>i;RP#2*ujk7a
zC?V4l{@^@1x)wbCOE6x9m5oThi=UW~+%TVlaeSEGAxwRo`c|1d#ald^Jm~F$;5-3t
z18scwR)Q~Zu4C%0?Dwg))&uq01RXlgx^Qo^Y+KXs1ulrxiB{}`F*>G)Bl9k%phJo7
zh^9fZoxP18d=vBnl~mfbNjRE_!=A^JypSpE&a_Q%@+5q3Bji3hM8MnjTV;6zMg=yR
z8|=sE=BYnbzFnzGk${wtca_>X54@SYid>bvWt58%$xvS-T7=}~1wskzLZRLZNZjJJ
z3ud}6VynrrY+Dy>tI44Nw%WBe)TOHV%IjrC5PYDg;rA9Ph>|^_He3ICX1C%$BTI^G
zxH)(58%V<t&4rH^H8KvEp0a_Kc=t$?IP#1|{16SGqcna96kighK$v`GrmuG~Jacz=
z>6zbI^&5J*4Hm_63s@E%{O;#EK7C$a=2AC%2V^!sLdig&`v_6`<oRdG(%M=Ce;s@J
zJjlSA^0*a)2rA2Cc_m>?=Oa`i>XsxzseVnmmOXW~LacRfN}AXT>&DMHNrlZQ=?7P;
ztTOJtv^Lzf$RVvRulJXi_!>JXj;$-UP3E(5IEyLv&X?USS1SJ_N3gHwsH}}3T$zFP
zib;ZOcD4u~K8}2qG-j)JOI2evcu{PmvZc=KBz;%NFm1SDVK61dq^}s(3=^)n+-vRu
z6Ev>rTl^l`n%hAots|R8@vwnxLALMb4O#5zr%LfOXP`a<;`5`rhQJ$-&^=67A6H0f
zkJ9682XSDjQ#!4UNn(pe?$mTg%=tGm5&vL$<pBeYQ+J)S(seMj031UXm@toOF$TRS
zCT*d17ygDyLF##9Pq1e9MC#%sJWBIT>Kru5HkStJ67Kle!?c)CCaxUaV#PqG#!^)>
zD=4z-B$=B&#2J=CE8Du_(t0uJvYWp|vBvRF7|60V22AYF%|`sflDSQ%<PC8Piv$a9
z+|0NRG}+^)vwMW%i_Xmm9=<{>EjpT<kqNiXhXIgP2$N8&H;CSuwyHYRn7^gU@~t+?
zm52kud8#Sz!$KV0e=P0FJSrf2W)?2~%9=MAVU?}>$v$HH6VVw|T&&w<V}VSU`89ga
z@`L$)or#8ny;W~$-x@aWcbX1*u@uG?vN413PcGwOM@;z541bz}sIi5U<wD+{le8tp
zx^E3^%<m^DGPmpwsVN5YTCS+>-9`8<QMHBe6Xt);;U}h<bv&{w63yqV@g+0|qv};k
z@V<a2v5toXe@lRHZhPZRdZ0=4u>Y7<k(@hnVB{EM1p83u?(V8wRaBX1iryK|fS(S3
z@d}D3@N|qKdAjPah5^aWu(-XLFTY?~FlzpGd}GEsZQ-{U>G=Mn%Ho%9TLlG&N8Dx|
z>$zj`RePW=L$o^VF0pa};+njlV&(Z<Dj>nq&LTjjRRdLo#ZvO{6tZudRm|T!=#;*U
zfEwp##0sMX=4yuiOI*=8h;e1Oe>}#NXA@d@Mz0w*Z@yJ0)nAe}y7YXZkJI^5YVm0K
z2B(<H;<RLzIOoY?t%g)h#=VCh`$?s%tR%fZf82MDY1FtaYe;tlAWYcogn^UWtrxSt
z(Swir;`x0V3+%vs;!}LMg4T)e>+#pLtlIn4er@mhPzuPfMmR}5d+@r2p6QTjd{|P4
zurcxFfA9HQ#?#+ayUl(Cij2e3O>&%fxpSt6k$=Q4()$S>UgxO$U$+?l<)D+DmE(W6
z#n6<rr2)DZPF0<37*Ve6%Xh$dVA60Tk*{cDY(nHE&!ji{x}PpfTiHf7G@1FvquMjI
zD__(qqrdXKv0ZJi&fV`Gk97p}F+}zcZ|^P%PJ_#B2F%Gu2cia;e+`Mf9GKGOOwHg^
zeE~U?ni-njO!2wiSBI4*gV2V(zOfQ`y1u+?n{h-P3Vs}2763gaJ0AJUqj+J7J9u1Z
zUc?qpw~oFt=li_BJM=wA1m%wU*Jev>>1M76Nn;#BBf-b}`v^bC*R*<v=@62%?buJG
z9=`Vmh2bR5h7bCe4>s{xGn3+La^H_&<j?$2{#m{Kvnx~VmoAxs(Eo?Aa|q8Y+S+Yw
z+qPXv#kT#$wr$(CZCe%Fso1v7TmN&L=QK`lHTQm2XUw_g9PbbOPa$+Gl-3BSfvA5M
z`+#@+lxEskESUn=NIyPhLlUb&2Dk=g)Su|VT-zEuwfJ6#8f07_Qr@I;h|XC&dewzr
zqO`R{=g9E{#Amo*@KOIXhgM|tLO-7|wfH`#*mK<F2Ts#^-M{ZIASChYA^@*CgoF}O
zeV|MTPdCsDCi#7eVk`#bGI*YHqMydKgH6&naRFrA!yEVPy(q1Lk$3}EXx6h%w~1pC
zO5$__S1;l67+?`vw=bw+x?@ZuWY;qLo6XwWfOOci4J!R~3fOVfw->f2ddJHh(Qq>E
z@Ih|p>7Bt`#!eFYq8f3WAx^hpHiJxg^^1rIL?vj|Dp9qpws37E_(d1Z8xqqRl5LP!
zYk7}nj)|g(9GZEPqL8R2f*h83?cAp$&4>IlT|pIXn*t6#XrVgf{%hSr2<STGmA~x3
z`goIx#U`Osf3+SV3=rJ`3=3p{3~32@J7nfuv#uk>dsOMIsjH)Yhg2f7c>Y>N)`ENP
zd-f_nlyjEzcPf8Ea1Wz1#l^u`|8qd8SGVs**P{N2O7ZTA>bUum9eB*%neO`fzX`AX
zeX8Mh)E1j*8jZect<)LWya26?BJ`e>lzNJ&cud*+fJW^%cDxh5R<w|~)csEI#6)Ov
zeyf1vRuPrS_BPdR(10+;8-dzfE~B|;0<<}H)4Q9WSK=O!hdKrA;7*^Te>=#TCYz^T
zloi{nT;e5l{|<U<T+%J!US<_~5lvGe3bJ&7z@`VqB{AnvwBZZ+;TUnrOSqJXYF)}M
zG@2LYx7c8CdX{%IQa6u;7~H=1L*WnZ_{DgA$3}}Pc#LQWP+0=}3S(5IYu!M`S}Ks}
ze%*Y&SlMS{tin46#y!<TP6gJl=t438LnbA2Y?W+FnjofcB07v4<2cD$CLneh<YZ+a
zsZt@oeip0=2bBqCaj<Z4w(Z)I16sHw189xXAhH!yNN(VFrDz{#&%m7Dyq@1DhqIfn
zROQ~Qw{8C*UU>j@Nt2~}QZ`VX+pS?Ji6p{8N34{u7i3h_Co)SU#noF>t}UkBF4RUu
zFkd#*WEizMTrSfcY;MCR`m4?L=cA3bAZT>sZEh!$4n?lG)-jo!gOs`6KEy7yI1B;c
zxiWQR6Qyn)Nm&TKP=8~y!ehEBIFi~Z(n_tX?X6n|Iv~|DK0w{E_|Y`la)5^;7N)3<
zxttcz<cCgng;rmBVqkzyCSj^SaOVo-70N1Il|Q6AGbXTDM(9IkE;ql5OhDacc)Aok
zx}r&dK1_o_p%E7T<mzq9CL_OhSHk?g0D%Y@j^kJWL7^gE$3&a1+H6G;h+e>FFw~ep
zJ`Ab|!<mw2MV{02ptnuE-PkNAR;9T!V}T1}je^BBvZy0KNZRpa4qR=j6wMCyVtBZs
z_7=G39p{m;LovR>@cOaOl_W{abL4#HvfQ-62U9*ht{DfbpRPFPrENz{-LC!AUX{|>
zJ}2<JWwD}>DjJ~N*=EdWXXBb;md}))=8Uec-bLd_=7fvo-@vviM_|Q$_4%N1S;+(o
zYn=TAX1NZRI&h>+ZW{k5*he5DBi?Yvr*g>q|5}gcM{i=bBEopj#%?O9g7M=?!%TN8
zrD6%_lSdPm8gtYwp;juKeoDl`2AmetVk2&Va+m64(tEP|o>NoT8`-v7D|%EVK2wL3
zB4b^;>Kg1KsA|&P7#Hr@TL`2PCP9Sj839hyX8AHIg+C1=VX7*skvE!c(jwNQWSGVN
zVprgo!57UMxb_x~^Dn<(pc$^PIM>Kz-Z&#}s4@4u15;3p6-6}h)3*)ZR7R9TeNJ=5
z3_x{?&GBvpPf5U?q|lu`N}{O^1JT6EL!4C$$5>rml9R8eGB0VA)`cFLmvGIMwux4B
zo+;;w?GdQ=XT=Z>Bn@1pfyb?e#HPVCVM`Urj<SbrsUMHpQ5O~q?F1?dH+s9Wx>&ac
z>XM_YCxSM!qf3?|@{&~hk?$oVD`q`#gIcoL$UTWQbf56QgGsX6HeJt0EwnjqG5R0)
zO-G?8er>Hr=T`<*H;T6M1S}M~fG?=_F#YZX^>aq^G_jU34y=kdx`}4tR^*Esff(Zv
zLyr^8pp{bA@iC8fW$e@!X8vImdSQdRxRt^!iVyokn`3Ocw6}B$Oidrwxp|Bu%*<8e
z?LnbngNtX#a~w@xwn`n}1tqlnlUjSl7d%Va5KteHFlL0*8mLJfD-5?fvh&S!dWV!3
zY6xnGKbSNsU~LYd|7b3?Fv%S7l9c}|14u&L>c0xUYIeu1FF;_Q9{m#<%T>!EbHU(~
z0!vXK?|fZbTXwhh2cJW}M1DBsrBv%xl@EhCE`qj%Why4@SoP~rDF?fUeg(d=*~84`
zsGA3ua~w;S$CO{vlj`XbSmc7|LS}+Pv4y}5dw6EHS$rKxY_k&HSvu<!qbpEI9a$dN
z+-%N~ARaznG@O<osA=wTZi&;3ZIH`>Z?YF7rn?2e>1N5Po(jEt4Srw!2bc36i<bvN
zA~u>L)4ANh1#UCBOCGesHP9Wf?t&H)1w0_EfQF#qa!GZ*KLMae7i`BM@9oR=)GdTB
zuD*#^$m9Crk}=!~NTaM;5w7Ouh=lf|;U5t*dMd`7scf}#DFvTerq{6wOEQbI_d$%w
z&fWz`);=&=WWdx&YFNg0#8q1k@h<!(CtI*gjbQ`l1f~+Yu>y-;qGNjxt|Jn2Dzp9i
zmu#-HI6J^>E%M<~ZSssy-OLKNl7uNu$z05LxIpiOpD~*XR5ix+$k0QGLG&{J=&ddF
zX`gYjlMv>{<Z<JgHxk?J4uX|^lly?rUP~Ex|8V&xhtY_{>jl}YY(>$Web2E&lY;Rp
z%jdxY+FM)D>!lkgad|5~<&+xlv|LKjs}G>fT%I|)dAHo({hPj|_Qs+yXk)&g;_z9(
zsnCa+gUnSWZ5`7mEIDIrM*2Wkud%`!9s8*Hl1;~2x%#I1+}}V+#y$DLCSo#D;H)<z
zpBINE`wcO;lwPs5bE~_`5p+{$Dzb1@N|IUg!pKzA<ENMcrRbf<uL8XO#O|k6#>NkI
zdHGB1y)L~-3L*QMFR%Pc2$v{+bP*tncEn)@o^8y@@mRxsRVWme(}pUewIMsw(B4bF
z0$)I~)5GCu>okK(8TETKMq`hn<&?pZ6;u3BnS(u*RhYYgcRTFD8ftdq4D&L3Kam>7
zk6X8HULyMso!u$5X~$3j#UK2OU6<sEiCEuvx0=*b;L`DfS5B``3#FRy?M?_9DL`^o
zxDR&-aSbndrn}0`A919CnLGQw*SJ7sk+!v!E?&hvBP_?Hvh|P;y@~v9#5*U5r~k&p
zY@Gl3`2UObe_MlS0Z7|DwxVx+sGD0bPTf1a3BiG~8qGo)D>hfSS*S@Wlbu2qQ%4@3
zyxdbK7K&&-)0v8qCaQObrJWEY7B2Er{`TqM?%Vo0qtRa&k|=$9e0~_J_uv0j<gPp}
z$tsqtzqqk-_CYDtF))kGFHnK>I`3|CtIo#{V_FX}=oih)Z>eAVjn0orXI7AFmv-R^
zi^7j5E!QiuTBZ@TFK!@P5~R~TdV4m-_jCJvg&)Jf^l37DbFs12cLSjGryWSn_5Jib
zfe~OLfy--08}iTa{TY3H*wSl?FwR6dsJ;D=9xYvn#JaZe{Q=GK9T>1{XskDV&#QZ9
z=MR(n$@~sQ#KPtX_8bI3Zqo`q8;W4($jp(IFn|XC@I~eJD&<r(-JmHNnpm!mV;y9g
zDOXZbbAU*dKr+`ti5fA=1{MR~4VR9_{ezW+1s&NL4iXI<aR-8>YA;KUCtVEX=FM#l
zN+BY(r;Lt|T~lNWKbIJ7KEsF2kR6NFYOakUWs{p)#@2G{B&pzldu{LTJEX)PRh(5Y
z8^<ZFnpr*IbF39_cO`sNp9T8rk5cm*vkOY3f-on&RBB~S5(uqO{5xO=dn%_K|Jl|b
zr)@~VfpjUDXsHp{dc{4>g&8J|P5V0Nyj~i#JYDcP3|DZN%OKEBm|TvUD>mO#-WcE2
z3Mg&TOKcm8;nM^Fx2vRRr+orV&C2<fnn`T~;X>#frn4C;2N1n%Re1$+PDYv0H=|(6
zW227iV)LH>!?s5)FK&3HeYwMI%LD&wkOrY}Mg(jNn!*vRr8@p?H5Jxc#yA1(0ho^W
zBIdqa2Vyx|{I)?-&V(|62(K}Z6bl?D59uKV#d~%H%7t6|;W(8P4Q>T}gt3?xyaBy)
zU2Oq2MpF4>Lj)@6Z%5jyXK*QcUM~(>EP@2oCMX32JPe_LeqQJg!?+a-*2H_%huAn+
z3r}=Fe76I^!Vq5UP5G0R*_p{+%11XA+r{Z03qfp;dy4hKjnh?Lt%yy%zLYJRCgt}3
ze-gnO38F0^mW4!ixzW~Zt|<)x+sQ}$KseiVvjkXWggOv++_yfNiUCpq-H}w#1z-!5
z!NYasrO$o24a9+8{KrOrF_o$WqW-Ck359R;?XukpbSyayq)ku;yRc$gth$U<?d@VK
zz;&q$*qo(2a3i0BeO@1;+=YYfcRzWWK(CSCbNjzLCZY=7_yZX^RTUsez*N<=<ZW*D
z(H;hRD1rplCMYLic>K_56X<kqWJC#%V6}a~fRKNXDgjN%h3nN8;WHXAQ@y&ZWi=Xb
z)lo!fVnGpAJz8;|GXXR8oq2RgIHQe!vcIc%J`{&*&-o!-FAi*b`Kg!*EpBv=hA>~u
zZR$16USGj>ry(0g%3;QU-|^;%6Srg7(dG(`k<O`OG_Bl9J)rE3i;Z}y%~&2+Jb%~p
z=h6s9G`+hsIH=D)`tl>5M_N~BK#)jVhyDxA`&m4w13%kP4~3<XhsT-kn6h07J@3LY
zZ=dH6h`}bOv<ccbO;qq*7i^O4xSO8iuD$*c(_hciQqzH!70O>zxBFj}as5TAOTW;B
z+Zr-Yt{~s7{Qo6Ee#xfik+3j>@I3!E(i1ZhU>i<;xyvXk+EN&7u^53d>~??G8j4~%
zZ9|vtqitR;gfAxpG<CA0n&s}dM6okX-ECIu_k<gtoyOLPvRp<UDpc79jqRq0uZb;r
z(-sGok+?F<lT5bcxZRn+$l#AKqXLWDgFJ-CMGSAT5VVt}n=Nm$yfL#mA5t}NQaVr#
z$<WQ$rh}S=Xs-jK>Q1b;&gNpxSac36WPj*+B%71UOTWzYbN#)7p9=SkA~taVkiADJ
zQ>{Y{vTs~NWH1fQY&;_kj=r!!gf5ZyC#z_DG!u+v8!%w>Ajc_d6<grfd63QWv?MXI
z`VbO7IntonxHCwjNhzPszAiTd;$<Y|`9N5~ixW59zLrF6qHjqgn+e`284@EayvLAN
zyQdv9ddg9?UVoLb{yHD;;|$f6$0Dh51#OdQ+ALg3Y1&O~P)9nsA_vGwGMBmjSf319
zb@Q3{Jr}#xk{je`{7&4S#xu!9UGlsK#udN%O-VNkGMP~2?VWo!<4ZEam>}Y?&$T?1
z$2i@0t!0tz80gsJLn~X*EZn>=Z4vXgT9n#_Yr|(Rh4SKj#!XQSu&zf(lvf0CyfFUQ
z-KwX|byrX&jaX2BP~9@37byBeq6ct{>EgX)ZRZ`ptL&|7Cc=?-_;`mecSQObDO6_j
zSP;}OGwsY*b@z+pA9&hl-Kv1Q=gj#ucv<UHdb8j)OQo9TYr1F=?c)cjq4@6pJ}6XT
z{m%HRqqlt~)D~Y%yYX`=Qvk?xw5fYo@mjo#9WupSIq1wUJ!|!|SVO#uXw))V3p$pH
zo38y_u?;BbYjTa|`ary^<ZD(la(6n*59m{E6)*AL_1EA?)My;wS8YDNTTn~nqyXZ@
zx;6sQtj2}vgE{ulS3YNh!ANI^U+V*tJSyp7iBa<SNF7w>Vd<e*o)Dt>>*ei{dUT6P
z*}zu1Z2T&cz<d#nMOR-$=!N&H&8NX>9kVZ#215NbH`f3nLrQK+CDU?(1RKs6*gcdV
z1exJ_-oWbzc)4kM1v9R&irsWie3*{Qnr`lnq`1WBS9Q+|PX7eg-jtKK=~`ylX&NRL
z-GOgXdZ1B6I)Fol+I(Z=aWIBBXZH1%KAs_p#*JX*fq1YgiqHP1z3*AXE~+Tn!h9Hc
zBvp4^5^Y<2$tPuG+qYVapsB;o_=&R)30-oZQzgb(VwCmCG|#~HWoHH*=(<;HRXCsX
z8xvhF_)y{e{ghU*)9*eD&Y|M)zXM~P5fqy^cnH}aGzg=91QXObRrW79XhVYE$8FR7
zDE<~(>DT%W8&kza_F1y+s_f}s9>0UpqKV-&;Lqf<;z4S^z(yte9q2KVUXU0i6(5Qw
z%x|S&X&-9|Cv>h&1d?+)+_FVmk@Qh&3}R_(wtrf=MisS@Cof`m@pil;k43lvR9T-U
zS%`3@)HMLNt&Egn!v+(^F0>OVq7b77b@ZE5G~5uRa#LaMrDTf@f`wjjcSFI>0=gTw
z*L(evrkRya?m=|eOU;2tH@sz8`;f$7OSQK5qpUD-m4TN&K5IS&t6~cco_HGwe<lGo
zb{ilb7=)KP#0~DzcECw1zDk`SnHKRIl&A2?RsrX%CvhFGCKr##n5u3~wHd#(anN~O
z*3#}a2<s+0v&fJe(DhIGs%z~4bPGq!-#Wy7Ujbr|cjas;!_xRb^Tltqo<8-luy#ZZ
z*opShri`r2t#0~N&Oz_Mwh3Q^Mzj+NDj4d&4aJANo5wWej6kNFtwaZbMLie|DL#N`
z3*nQbrnl-N#nWzj2wxL2<3eN|Tcg>h&(ibiiy7ogM<ewMLtav6kQMRTT=f-?RgS+_
z&1%L;`VHt}g+*X(<dDC>l4<%Q?hSMtpa#HNk;n+DZjewJ^2$xYcqlb=bFI(388Xy{
zS7631;T$bjqii~+bCy{)vKuSts>!p}|9myb5-j0k3Bhk7=-Nk<l$+04wa$3~$5F2d
zY>zzDmXxfPy#6$<I_Tx<J9~&Rp1L+bC;95_Tz?~;@z^^5H}YlY_)qfX<oMsn_wT=M
zFy1KoKbOs8g3upF`Z^9tY=4{t2yZ4d@i3rNc~|SPaQ~r5<bN)kmM4|Jk18*4T&|8-
z3E8!6OB-nzdA^yik4Mv4FTeD7Xfo-X02*99KMpAFf9~pbUl&ftVfv#{7C+<{+v>Q|
zAo6~_BArO>8Qpaa%J+0B59|g1(G$A7ri-xuM$Y5SjdErfm^LGYh>06Y`gj$2TX{np
z9o;Q%fM)DP&1blF+v)i^4?UrM{KS!l(?<6N^#k)AN6#c0g>HZREdD^)OqBFnGzuVz
zE8A7N$f?JUa@`Yw#hwiZLYEu11DUlWoz>%k;q^F<VOtto=Q{4riC&O7*IZ8E-Yo*L
z*+gkhPKEx3O=21n@u$ttD}jiv*KC`n3>S(RyIgWSTYAZse^(59q}&8+I$ME$X;hs*
z4qDQ1gfIyuBv~^wkS4^u;BYXV_XsOY8z)`;2X8T8sQJ7-$%q9E+<q~zo0iB?k#}8)
zu~6~)>cxmgjD_J&!lP7w5&5tU%F!FF()l|dL3PVBgU|{)!S3zNW-3luMuDg_{jY06
zdW-bX?vQ>Qz|hZ=nZ;}fL`;cs<AKEwieXAgq1Z5zWo)$IIz>Sy%w92QtZvk}*dWu*
zSj}2Skbm37ctk~ioaVF)0y_&Umbu^z4^?={1npWcYxQJg2tgbN6;~^b1&{7DHD%UF
z4UU-*Z5yzX9(x<8$v`>CP9l@KaFX`p`?fD<)g#0_#4<^2C@14MudK~+gp^_p<<CCc
zKrJ1354{fA6O*8+X<D7z9Clx_jgu824$rPI&&?lOL7YhpMWCn!Jqutd(Sdb=nw=Dc
z{GHmWydpKn3>;^XlFZ;zA2aMsgCC3)=q!5)it@~PZ9g1eJ*$}77Eb&vV-gZ?o^4&Q
zj8d^rP<)w@sWkR*%p*!|Y<+cxCt470w9v7g)oKjs7_DuF?nm%YhW|jEzf8nmJ%Wks
zIM>+5xDkI}PN_l(`O7$_J@R<Y-y2Vfxcu)k(fu&K<Kxweks%7X5oL!u<8?4|+d&jE
zG#WxJC08*ziePz?WxSgWS0F<GZF``L3{A&)Ux<MkwSRElr06j8j~cPTYU(N^!xN`(
zlGs)efQpX;B<vm?Va>n2LDGveNP*WbEWO_PFM({-h3zBr#ls1);C?8ug}i)p4A9@W
z%Yw!&Uh9S~h+Dd@3klnQhr=z-gMC9Oq9S6C^jq^DcbVyUF{(s9(CTJ*Chf4t1^<5s
zb4(G53-&Ro4I&>Zym?-hy)TM?)0-jJZVONesTWRoQnZgV_{^MWHOJ2G60lDf>7tVc
zCjm$Td*sqFvRo+Z6ejPF?i0Ct{H;bNm*r&o#O1(U-2ZYP3aHAyXK?XVu}4C9N!s@-
zsEA{(RTe$YBiz-6SSBpt18l?Hl>RsdURuSwaQm0g$k8uJ0ml~cox;|DGk(_tmJ255
zrLrgSFT;y|S?Bxe*z^T86=ERLj|f?@<sr%7Nwv2`4ltQ;GuNo11$t~(*w+-=%iR8p
z_oaowJ>Q0~*D{WyFTRpQupmYNY#@7Kws-qv0FNq5Ht$KmH?tij@yWfZWX@XT_@!B(
zq{#(i8@+~pdiUywqjtSTPit6bd>_5(=famNL#78xPQGRE%3v*8Gu<J}pIN$cKA^XU
zQ{Je%Ph!D_MCS4Iz&VWt2RI80o~#}5O*6LSP<POk`0OWNn^}5N4)UbQ&G1-Nh-4<q
zZ<|r+*z)A9w%v!9fh6-TD;{RKwxD8Y5ycJ6yn9#z+AFy@&QdLNx0X@$`w;=sMW(re
zrgt-g2LPLlr$thIb)7w+otPq73~qh`;S~r{5dsiWY|dtLj0UTI=CORcc}s$o4H(j^
z)^-eJl%ib7e@^*iUtYw%D5$s_X8U$s6xaT8+P7S1&hV4FX0oFVFkDPXOFs1s!NHZ@
zlERmVxowri*?aEv^KPVkz~nB;*14?j0BValuI<RCIcXR9ljCnXLKtu7h+=NaRA)$}
zQpH5q&MdouE_=o6UU}G1_olY~k$WfJ=H&S2Il;%J`@|s9&9E(**E^G!mBEiJKm3on
zLkFqnud=LLe}A2VVS$t@#h<VHm$T8VzDJFVFI6<?Ut+%`@!Wt<&=xbNxPIR_NNq+@
z5F@|@!7PA7Q62mFzFD02zLl>376c27Etlx+E)t?Q+|A~dw`;PDOsL@CyZe>FNT#V0
zU@2sf_GbCd<s~HZ=qw^x*{zsJ%{DZh&9Fm%kN1J}_y?bY({O*ha!Bx#@_xx%`3_Bu
zEtsNhv^dYOeD9WI$AEG-n5PgzX5V03f%OyMu0k+kcUW-03Z10%7jMncvu4>p98f6o
zXQpfLuaXKaG70&Qb!HrP<S`bIyqh-HZB4TS1y+D~3XW*s>tPn7{0RIwb!|hZG%vGl
zb%au=FM+IzJo|~tZ?bc|rmXg}&;$j^X}YED16{Wd<xQHPUfFX|Th(u((Xm8Q)vT7|
z0g7`^N-Dgm@z1f=YD2N<lxXiM$s>Ir%Eg=3rACFFm|$v;Rl2BLMsT>u1>3%KunuDg
z=E~?r@zviN9?l!mi(w=}h3!hu3IHaFf(rVlH%+1a(t)Q{ITL8MVWG_L$Ht}E6Y`UM
z1h3XHQeBx-fh*(_#Ytk;nhv$__8@cGTV)k~Om@6yI)jMKihrEp^;EqdSfC&JO~Mtn
zgBdNxS!z$1|M|?#_vIg^7t(&V^_b`WxF{1U|C9cWV{H!Z;Qm8Iv@8R~FEmwFkMfLt
z1sozq2yc-aXGm%m{&ad1reW};`(hqu3(^ENg0}A9xB)7{lA}*5q(*PYfvsrisPebY
zDHD0ElSM)%zC_SnVy6XOKu+2&$%G(RJh?13`4g8le|(3xH9<RV#qpTqcVYCX^|5V9
zjD^AtVq28)K0N9KWr}FZ1^#8BL;c!51A7d66snBI9hn{{=CIhyNjrSw0{HdLXa;=U
zoFr=Lcaxh$<|3zRP<nhS;jxHMUW~zx^GIVZj!2Nh!MM=SdiP<g34?=J-MZ@yIr^Mb
z%*cUlOVT8jC7X7#h|!2k7HY%RX}LDQR6?WpS)GP;z;*%qVY5OPn~B=<vP_A$Ku@8Q
zzlZKE`(3))!!Z)^gUBO!54;kgj6vqkpbm4J(Bm16ad2U~vn;r!NzwX1#yu=|MUQe+
zI<~Qy2FoTlsdIwiwEzt6KyA>d_UQ?!HM5=9x(e;z4kZL=Hm?VHmkvV;x<!}1`F=-9
z%>j4l-WP|1u4oHcnUYCLEt@J*MP;R`p=19oC+$`624tWmb3=APd=z+FAySn2)KLoB
zKBwnirPfslK$@+)g`p5jT3AK?fl%p7Qy0Mw_r4G;qlM27uM#!X4()LmL=(ALa&h<q
zYTcbemHwmyj=`&t$CUa@fNq2^5dZpGLA8o<ocFs7d`E5@L~&1f3O2W~rM6dFj`n^f
zcvC<<vCs8Rb^0Wdr*!?H9L+<8@^y#tUYo8RJp^eOd(Q<;$kkV~R3x{vtebIRb1MPu
z(fGp8aj8QCi&jXD2I@>o2%)91l;UzV=x!QUv!VOKC$;{;xL=PhAH9E4<xNNYRtYre
z`)8sG6V_OCjs)Fj1-Y3>Yywa7wYTxMF@Sn8<Y#Tjvmsv%?T>PT8$P)G@Ywo^Otpz%
zD(XeI{O1?S`7j7B?Z7b3Esf=EqQzdTv9jdm_g=%Xl4SG|E=T>dR81AC%*AvJ$~(<h
z80?#&Lv)F?(w-|iwbqQx4p~Z$^W*b7e%Tl6kX;8(U)%V*VF6yuPiNi__`t`E_J3pV
zf0+Fr9!VAsuK#HgV;V4Nd%=OU^^a*m4Y;!RV4>iKY*F&NwoTQY|314ecA5YB`a;l4
zlo(IGq!Pt!wT{4vMVOt!OH3Rj48q$cn7KjR=^=qv<Ahg(ztMx-g-z@eyq7G3ge8I-
zxc#<yeUp?jd38EQl#Cc7ob1CVd^$t!cYxjX11b}jMq2N?KN(mdZi#w{u_*$dCLC7J
zJwShy??^d}+e;mx=<-HrcjL_S_4wHnJQfGy8@PPE|Gr%j{y>6Y<1fJW{dz`XKnfHn
zgAfL3Cj9>Pd40b<FfY(Z9SG6;{8Gzs7J^6`xt)hd;{PED+j-qb+{Eg^{f?0B3prMD
zzuukQqikh7$>-LeD`5tx=Qrg6036rgq0-RKuewtE;wx}T2;#W<!01TEGVDguY2F!r
z7iqz|JZ1}&-w0P&CU2geY%H;{C#-d6a)tWo+y_p7#Cs(tqz61hMMCPt7|Grc{TqNA
zmjnI$P{!69$dv`eeD#ZmR>Luj1QI5_8S1quTOqS0C^rXO#^;F!e>Qa?56{qcWBfOq
zJwSY>C9n$T6zxf(*>X-sKtNp)hPM1|;J_Tp>%V5Pp&A3{)U4Z3861dUd@P>%7rq{$
zr1rx|ft|zwEA-Ku(4SjCHb8Mye+DUm<HX`!oOanM;0(o}+}pD~l_mlZfoo!c-faaq
z{MDqWRPw-@_eoz>Lr<XOJN6$XU`~IMR&F3nK&@|uke{z;_H>xL_Sb7DT6t)th||k*
z*9r-YjoSYrau21r0^C_4i1_lH<9{Cp3N*?u-liI&6@{7-W0xfO=B89XxG$&vR+@>=
zF_}{ephC4cXAS-_=^oAb17>xCI-q<{@_d~+^{IWmlDrt2RmgL6KYtv3Q&UG>W+rcn
zRw}A@F<bs5iO%ox%(~t%9E71>L(w<E_?M{UbG)()`Uk7URxM3a98BqMujZ|dMxtfk
zanFZVVjLDy1I)t6<*@Ehi?asqZe_<0ZaRZ8!&A4mjI6Y`R1nuVq5|e=?1VT(sJ3!!
z2L0u)@#r=vJw5+PKNN<laBoIvB&arY$=_6Jrx+iakt8E&t6^rTlG)V2&O>e(tdu$^
z9hc+D4P#BC==YJv#OX`YHVz^l%J){6<rqu)=r6PhnH{aKe~p3~Hk&x0s&8tA08!b4
z(+Lmq(lp`3&89J5f$zv`RkP#kyG^P4iqcUePC?+Iid`!{h&`Lh{K=arB`DEq8H9#V
z-WObhc_;aa3`RS3P#{o~)Yvul-Wr#6pHHiXfB@Nbx;+hi`Xx0i4K@a{?bx5He!YPZ
zfAkm5zG&wuqdMuaOSt!48qjSG?c-UK>qBdA?lWw2yN&JI1l>d;_?K6V)jk=L(P%-*
zKV$M{vglD26wAe0Rlr>sysPbV%Ps6;1@%B}44%eG+xL$s*NWmkf)WHFVTjNJS+bXp
zQ+K?XvALN33V7^?a8awcQb+0JvrWr$=Jd)er(%kXCasEbXS(_KR?#cfEfY)jH)Q+1
zV86shJY!+Lh-Wq}4b9T)HLa%n*dmBn2BMv$v(X%|a;E+qP16bV6darCq~<bskOR@q
z=fLUSjhGiy+rZZ{anjneX04{G@mx?{m?o*|&A*m1Ir5-#o>jexW<aJ(y||eX+$p-S
zYjsX7=bJ|sJ4|28eaY*GJVkB7iMg}+eC^26Rx~3a);1gUM<MPzNv#p=WvMtW_sGVr
zs5BmZ^wWKI`Gndx;i=-CYQtvb3fo?gI;s{?R|TtC8r(ICg(tGt;?YJ@fh#H&7@0d1
z<vQTsO$8(i|Lzb%W(~R#QKuIf2BI@5N)BN9qKK0JRXgyTFb$e1B-az5Ck#gV6}t+<
zmY>GNhdlE;wA(D_h(rLR%S9A94d$B>6J1|qt)BTJP8ho*vOpBLbX`+}!6TPoytJg3
zHf=utT1RYdtE3mf(ZFkb3q<>-fa?tDg!EaN6RQR8B24s$^pwX!L5dLUHwq?!^#^5U
zIA%H;>Nw9wQj_SXLacvU^OCBo-NbslVi8=Yu`nZ#=dS8nDjZ>st36N^cqB4D#L}5X
zoUq$mfr2C=MA@}3D&Cy0TF;a9QA6=?yILuaYyAA%!TT5yCNJ0Bh;zJPuIS)=cg4p}
zV5r?-nMIOq^hDy=gonCxJW(~eW)}!%XSy|M1QRroJ3`+_Jvgv9YJoK*2VWr<4Evy>
z>H75s6Hbd-L;;#jR=4uD56{v7AdPTG187B1%oBs{UElot+fAV8c^Jn=ASi<q-<%AJ
zfr^yv`!`!;*}F3vi_1&am_N4LuX`H4p5=Wyc3GHU-e>&aKtP{$qBvJBc4q|7EPk+t
zbV3<u6SDB)W8ZjWaydVi`UeX8>@`?<1kVWq8=b#<tPc~L%TLx^{2(tGE3<ghTy#cT
z0Oa#%>OM5nY6Mpr0uPVX!Zp=#@zZ7LN7p%OVrQO=_DH$tg%orW9;Y0lRFMKx&Hlow
zBR>260F|HhlbtuG96OB>$zNTTZx7dhOK+NKN1;$unfOOF`C>$-iV{8*HFGdVXZmwA
zn0$)N0djtB{^dNcxMyYV7-UxOF2-E)8o}Ux)}q$loDyQ_Lvber8UgHv_#L%CA?)r<
zhAFV((ohwrV!Is1qQJ3nr&fo*BXt@S9JA`~wgCKAgb0ZkZuU09w#H+Biij$aE~q^8
zw+`6|qBO1s1OtO$gVqq!w}=H`^buY)NSBdc|4EZP+{&RWv|mM#9YL?!S=xO~+FKbm
zYS+Wugu0#FY+Kme^!yebH*t2>tt{w$w@oLX@PisP7kYDZuZJ7ixmvw4RE(_oceJYk
z99q9^pIfIqZ2OSR;YYpN%>1o-9*>h+&4|*QEM2@zv(R3wXW4jdt!JCRzuY!hWoI3{
zN~6Y(pJl8X+IX8<=Fhp#4}0eIcC@!;@IcR9yd|$HF#)h;_3X|@Q?mzKIx#QkE2a(v
z4@@)r%zZpeE$irAOz@V)b5<>qR+VX5;#X=mcdogGGvdk=8rR)luq26X4*wOE{!5#Z
zndyH+r5?bz?I8zJ&&NMhI<6n1Vep^#=WPW&OdV<?DzpCo+xaIBi$WEDC#krct3(xz
zdkpF8#+C>3f(UxO+~3?^A8kEuX`nn<y+4_e2+X~c#EO#oQK`)=USC;0KZ{!&Ub}Y(
z0a{UUqc{Fl*#z|9k3^fuIRw1Q=RdKtW;|sJjD5eJr%9psAmno&P;v7t8Drvv<xeKt
z%}J-^^5pxxy?^p)gPm5|;==j}aj*`G9M=(Na7#2p_4;@|KcfhI2M2E;Ypmf0`8zJs
zmU2|ku)lG9Jzovvg(*0PW0}jR*vXKz4~=Es;kDu5q2pPNG5tKjLcy4K_8SgOH?!_(
z!X7<4PJVqZV_&1hvPP2k6Qi$Q^%EN<_tyoQ3X=v}6&JCg5F<jks9_Ukidl!G8}ws<
zwf>pa(3Iw&cOAO&QpHHC4d%G#9wX)ucjCS0t<X`37ME*|49!{^!d?7T)1bcm9@LCc
zP*a~tI5DyQZ;Y^QkyKd*L6oCIy=Pr<3LOhq#n}Ojlv}@=3U^A8f@mV_@8@{p!vep$
zLPNT+YxL|eb9ML?@O^KcQRE}5&U+X562=r+BTCj6Kd-(tHWwFq4Q}r@-zTv1R{4Vh
zHdK0x)7%h`N@6I9zvh9I>Ch#toUJZ_rG5S1Z7U+>XoHcuIzI%R$+#U^&Y@$MGgB7N
zg3^Yz!NL;o>qF04#Zpmxz0cSZ9NgTqG#bCgfft&xHl*-#^;D;MmZU{2A@k_3sqxY_
z_Jmv;LnacbN2#{XGh3^QT50Iz3jcC?lw|ISq+#l-EuwHG2*pZ|#tcJ3{x&SApkLcF
zC=faFSOpZ(cmy^&3@h4{SqnuC`|2JhgBQ@cgfO<xT>(3Pth(pap3NtW%IG?Zn!`%O
z6wqUtw7i(Th-|NTIP`Hp#*Y%ZySRSXu7Eic`b53F#^AAnhk#Gf=Wa~%OFgU%9-pbE
zi;R+ezwxR&)~_}q8a-6OBP3}RtXrj3QhThtZK98;yTyMY-WX)GWy08tCT~w`$%0GV
z0tk;fnSN7TpV&1h;?N$L8l1T~?wPl#?2_`U(puh<IxI?AtOv_uXgf9nwSf|9LQdK<
zeI9+L?c!PurmSuK>A4SdwX9Hc5TwTul}N81$Ww<M&3ERRqM?(adwe*V`(TOUiWzKs
zW)gi|fcU)pYHb=rpFwf+P>^?8Z7O6s&QmEOZ)k7VE@8fioMjS%d9gFYL8?5o`SQU(
zler3)Ac(#a2cX8NKr$p^Ond&-XQ)+;V%aOFH#4^SQWl6a+F_c^Vl*PmOtd{0tkW9W
zXK6CVFxmGTro#pXd{Q)RYUX;SoBUb?H(i^2`eB~5JqGjBLdDE5;a_NDmA(sSu+@)A
z_f!dWW!OL^<s+r7_HF{t7xQ;_wS(CxMmmCM<)T5Am8V?<fS77jF!)b1=sV86VdPcQ
zG|iiV_(aciE6)g;QQ>EQxbkJVIopy%sWNhquzUFR7`n11PCm833u^Yrqh~3B(%Qw6
zz~~NYY4iJ$LUXnEs@mm28lw)HK2PouE|`&KAqRoOny3G(M4vkh1}|go!%|$}9~{4>
z(_>7$=(!BdaA4c>K}Q7|<dL!vw5c=V7W*yVyUrS1VqmWY(2LTq7|yH|xCD&-=$4Ki
z#gz;QAJK1-c{!uA%3E~TLlLiWpA>1Y1oC;G9;BzQJm7au8XC}p$*0p#55A;r1BFbu
z4P&{}QMF9mn2wJ=)s2=z<8-`k#WFbN&OF2sjTO2GTLKVYD0Xj-`5U+IxF!>KM(Z@W
zSl9X|5qy(PCDtJi^k)1JeG?tMf@EJa!6|#P53=&M$>Es(o}$)i60T#Pq{w7Ubf-c;
zBmgQIxWXzqZDdPL5}76qr0?6cE?c;`XcEunPTt5$E-gj3%h_w4VnYCx;0qY0Jx;7P
zp>nxm%jc#`oUOX44xfCnt(q9i=ob@X4$^<=E1Y~7<r+^)E@`qJmwS$ifL&UDKOPnJ
zE2E|C0Qhh~-EvDBTmBg^Fl{Qu%L(7KJI3Jfe9XE~id4eR+_}zv;X{?}EN?OB%6zaF
z<RzGEbyG=@yOQhiCg9uDg|c|$$XQsmv@3B2%*cO;uH_!PG-QDfGue2NJvNr4^khCJ
z>LrR}|3-hjO9r)f_(Ur$+0Yc+wIx{z!HTmyK0@k>idDW=Jmkj_H~MYEn(gv1S3J>N
zDIu(?xH#o9O0{Cxaynbl9{uHk(c0?PwYw<smgD+P+Jz5)88W2eldsJ0&ie;@O(fl>
zI0qT!iOG2$o{Z7u!b~k{qE{rFLkU?g6*Q$}-o}DvRTILmvPQdgFY<h2e<nJlRBd6=
z@y-de=b#<BqfyD2Zlq;n^))q|N9=_&!)qFMLo1+32f1AN-t{ZzE*q%>|IQC>)`=%)
z#=^FFeWMjTK-il9phE!A2Hv&J{p=~Pa3gl(7%NxS;jx0$T_4+So?%B}$u^qOJ)+B2
zGYv0_`;VW*xN(>EyycYt#`$z<5>k*IOINj=eX}|&h8Nyy#=LgM{1R#57@1er>-yTR
zf^%-A^zK<*E*V^P7erRJ7jxNFcf4pGdw=c-3NfjX4d;M+2vtNS)!b{ya$k6+h^W4^
zW<gc?6%o0G%XLjmG-j~f?gTb;%=$V!AA3-i^Aka4nunN<n?v>06!bpvU^U6PT*9-5
z15Go!-km}^ArXZ)U!2f8L**kb!)_MS4uh%vf~$4hu2qs~GL@oKrKi>;yHGdU{rHmb
zuFK7`vDp9i!}27Z_YHSk<|mfRWu-+}`U#s8QP}X`P~_KtDwA26SpK*4k6~@zq{mkH
z>l=0JGc<8ju>QD=telRXml@XU#O|EY6x6PWC{EG)2T<S0LN%0@ItJ?JXSG5JBt!u|
zpaxpTKApp(u(FMnvKxK{y=r?fHJ#pXPvBnSH%5tG^HW3MV8M3pudn;XsN{*;ixEb8
za*{r}pPSd5+F7VrEg(OTc0Mimf-lIv?N`2H*Sn5&j%R^SCGQjVZNZ;JElFQ&TMO&+
zv^wB82=}kIi#T}}d4sF3uZu<gymMn7m@>u@8e(=G-?y)`vb>`sJMWv(D74e9qn(DM
zn*w$ydFQv$qdIxNZKZe+G!|E>n!xZfAs+JSq~*G^91Z+(`g@yc*12j(L*7LCwl|$$
zd8(dZ*}n?J;dph?54VLehjdO?!GEzXq~z{o=DJ+36(hU2KjzepUo<t+@CdcpW8a2)
z%&nBKcbwB5+zhe>sG7GrhoOY14*gIb(Lc%Oegrl>)`2w$jN~gUqkkFY`t}o5?moR&
z)1Ss{n8&x;k_K-~JdDP{35&sb*H7jv+vz1@i~AGzVQTH>&QE4N;{G=MGOvq0b)O8%
zK^ft^R)wi^Rzy1G1~K+-Mh|}`^Dait^yc?g(Q4h+j}^@Mt73DDin%0Fcjl<F^?^P$
zS?r($Tz820Z(}cfI+xvDbvg33upZ+t`@sid5reLKQM9e--%MG_vyAecuM_z*WDa`L
zZc*_E@oa@ZZEV7!XV7Evmwy6c$#2mS0AZk#`O1`o%!Iy}2>JCyE8ry>&K4*J`=%2{
zAwg-yY1m}9i3RzMywSo7;oLCLfCBmw@g*#Z2^fy_$Oz3*k%GB%Ys?J9zOV=({I*2<
zkL7ArlRd_jt!BGotIzV*=h%(>rp|Ww8}QrA9mL%w83Qy*(Lf^MC?if_Xvw6TRhMim
zP#gINd`9=D>#&;@s*O5D3+fFQ=A*4QX-(2|DgqhvF=MDFEP`&_fT74Rt>K_<v-+!X
z_o*<#Kf*CJCZ3RS`=pS!<|6pR!Ys7$*g`l1K|D}%DX;;0ql20dAI3kYp*zE%(X`Xr
zYLDWRT~do8VJU6CpqS4L?%bD@HDf{T@kSWZYw}ZA49l=;-or;O_-RP85>ETyh4qBK
zE5Hl6KQR}}Mjt-3|2}4!#cRgbaP*c*jVk-|ur(G^95Zy+nCE!#X1@SGwq02|;>M|d
zbh%b-Q&9#Ex~eIzomnP4*fA-tyT(X}So`<1I2kJD1|q*yt%`Il3!~SRU2jfMi!#=5
z1ZTE0HlQy?*`Bpm1P8ow+|4=VhC%Ig4QW@qcQe9XNunY;B5j*>pUb1jr!kOoKufA4
zU+4kqS5n|=9N^B)f#uRu@QesAl~XwuIE094)6yRh_Z0!%?Q(3Ke@x<Ss1KUVhu~kv
z^D2%{54tL2g(h$mGO1a412h5z#>mDfOYeWN24XmJgCu9_?1U^A=?*iKcViLrS|kvF
zu8S_OY$PB_+nnLG?g<mXiDsWF2Vu!Du(qT{<AYRR<je!QEMtVmB)O?~M~4R@<AuVr
z1Xus09e9k@U*$xtPuVY-{0ThzTWS$iuI4p})1ofJ9DQ`bh?5IV#Ay^T%WRHIR+<!$
z8D)>6&OgF_NxI1k`Zt^(1`XAbfD2UANI(<m@C2bH*L!ZHfjWC~hZ~wNp^wm>r4;9S
zqg;%;M)9*6jfnLy9tEwctmL0u>sR+v-tpMcf&R2Da&%Aw>@R11Avv8nW|sH+(H+J<
z7HyrnD}yFb2t^!W)_G0hJC7IuRi#<Qoy1~U(Iiy<WLZIa+@4Z2RmidJc^<!Sm$Ul!
z{U96S75M3}1W=Rk(R$`!k$$Sqg|n?^u1t3yRvVV+_?zG_ZXEszTB_UFJkM+$3r{#X
zU~6`tI6=9*nvQJgP}RuBl&&to?y4SB!-ZE;8`@AR>zNSQ;po%aSb8LsS?*2PQvu#=
zScY|6!u6v=@hB4ZV&N#`%gIHz298m>%;N>4MWw;}2L#Sv2Nht6?#k=wJ`Rp}o+e>R
zecCugn-(TZizxe10>sC2x{rBBF?T<)e!^zW8LC#@YHXO0drOF<$*7)lBI>MGDGvD~
zU%(pThs|6{h;Cp@%g}GrgrznD)}jNxV*iL{!em)$6g7Gmj}7*#%d<Qt(8HWi_cH7(
zU<fOqziqKz?K6(2pKtUJ^=&R$Amm7Qw|vfa8@MzIkp_62uKY++c_|@^a#TU{W%!j_
zFy!d8$zqc+9PGFnC~KS1k?SO+3!c`HSK&q)hl#!#h4f#y(cl2DGMEaQjP#mpjJc3*
z>@EZxa!#N%TH*ByR>QGFxg2DTATL>}`*oGbWH}Io6jF4PrrhXJ(ur*v@p}%8cw**X
zip{FADglOKxDvOL0v5ms?{{JHs`n96Sq&?x0M|T?FIFYGjjFUwj(PRSlp0e*pO-qD
z!dVLO%UL|D<%z+b{h~teQb$>_1P<rfEJiLK>6bk!+|E5jbaZyhwmWY=952O=Fi%$&
zBrWt5`GsgpJouYe)AHtWb8QC|h-OuQit3Cb__a&z#L2<UnRbJcSdf;krxe6E<Ma;^
zMofB<e!XxS8BeSMung6-+LJ;Eh>*QY>Jz8~o7mQ^S*l$Oy@)Sdgye)GQ98Y*u+82a
z&e>WOhdLgWe@3K->R-@0h?E4dm?J8gHkT#7D>tmG#_WJ3YW4F(K3e>k<)3BmZza!w
zea5!6lxTTtEpzFvLG>ZPc2XWa+9-O*@@Tto)WnpjPno4xj6uaeX)S7$@ls*K)K<!>
zxl3e|=WW-b<P<;u*Y6A3#pu_@E2yM>P-o<kA#StJuJnK-cl={;TG4@aVR<6s_#5B^
zz&hV(LP?k>kYg5=qJ5)t*$h@DpW}^jBjvF5^1==~^teiW)^GsGSIw|Yh}3W0Oi%nC
zxliut0S2}c*E1iE$)B`iHtfAurMixpcf-TE(T9&r((&${=bSXMxp^7$-6Xkp`H5f}
zq5h;!%m7#E$vZfz$*X^fMTc^EKv!BD$GEh=!AVnO$L`_NAW;g)x<`DQTZ%>H9G@#E
zN^z*8uF|qTpwnWWcwQwZoXYu7{WDOvo=7l0dX5EEb<Pk>1RPE6;Aons<Sz#t3C8+T
z@$?Y~(^4GP_P0Qq>96Wjs!ep0<@Dy?;kwME!j|IdQ)?!57a|tHwj6+Y1nEq9a4=Ez
z9!B?x^GZt-Vepibl%~is@SpGTwCV}D@HECoDOwU|UZ^;YKSgp1gtimKWnnWox;~I%
zn}R}iZl6hWuyZ$iY%JokArG7uS5)d9qq6z4-35;<*rhqQ+{nfC-h{N=_b1L&da9hF
zlq!Wxsw#MwGcnpm-K;*iC@U|EDaJ=um8y$8#^I>rl%*{#_iK)n_pX(eaZttv5(&!4
zk$N#ARu@rV2O3wb7ZMXtM}W^kFhVR^Y4IZ0TUW(bin^zDh+a{(=&;DA4i#v{lic#)
zz=x9_^ii>EdHb^>Y-kkZH%kZB2o3lL);WgJ`lY+#Pd<Gr>uYSu-ripe=oTGlrwd0@
zRPn53kZjC-!^ZJ*8zjM{jU=d05wx^-I`N#RkhvJ5+`2ViLu$8gv_u<}wsUMAbY|tL
z%9zCJcb-UV2m<aO^f4TB2J<_0`}$d6C(pgK8)Pa%C|(K8xfMbcWcW*MN@8w(3$6l+
z+J++$WbVvTz;D0dc}S9rWp{iPzj5M*78(wpG*Ml%nklx(O9oP|NXODs$241)I+q%c
z&2%27YIOfTWLrUfHd|T9UR#kcXGP@S$}-1&1&0Nmh*%gA+iF$#X8`xunEn=$30-#u
z#Vkxd@mD5R*ox}X%O9hMtY2lIwRD&>sWg>Bok3b(IDM=lyR)wpE3GP9&l&~6DLR<p
zwvzhdAZPAI3F;#q56F96oqD(nEo-@kkbvg=Sn!kfle4}-y%%+?<qckzCH_;`=59tZ
zY;qJ&i<WIl{70ORi=Ku#>hO?GfNChWhHlKGtI*p5W#XZJ-}er9C@=Rcb-46;zmLez
zCgH!92mfgrXJuvopABAihwaF{AL@?|nEmz=YUqUi27g>jF;}$`JLsR2jb9SEj2Xvg
z<r7HO$jBQi2)M#rH&9gT>MF-bq*K%WJ*VF<{@WHRn^pf4g&v#yH?6Ku2W=<*Q(5x=
z%7YXXh^3>8wn@fWCCkCQU*Gx3K%WhtT?claZ(tL5P{V%IPxs@6Fb8#?9w#4`*qC{=
zMx@~TamU<scL=X1Y*P25w-QE}o!+u*w_o=cD|vawN(MI{lPhI@{OBfsV14fmt1(c{
z?qBTW`#xzHe|EF@^@@RR7=9vz`$G1l+zTiB82r56C+CC?gT)3#Dq4{m7>0fmTgB8(
z{%YU_qQA8<MvsRVw3cXm<M1<jYSSf9oXHYE1{ubzL46VkOa~J~;hB=iqwZ5rW>(~Y
z6^D!S-LtwsN^1SRn!G(=0&W+d-kYkG&|X`9CK^x0-Liz%V9lQ5+(nG6Xr-taT7v^u
z9&99xpdbhiQgXb5slky?&KRw3CTJCs@@xnPWup^{kx0+dJ>O}~(hLLH)(;e88wAtV
za>8E+y>|#V0r{`ubpPlMmT2y)w0GBogV*hqoE63{=3X+?wj?;O!A9H)i>`^$B@D1O
zww=aoF{@N^k6GKKUT7o5t^uky>!0S5tIG9b6N_xxJ@1zwFSiai$5e>8j)jh3WSlq@
zzlw@P#^uq6*3A&pIT({K@1Dt^_ogzI`xaRGc=&r7g}H4W)Y80L%@m7c;rDtIH%WR~
zE#?mE-RC<L1T+aD77CwAsp4t4K~@d4;Q<eAjJyaLgaP>57RjL)g+8*YL9Pe20Axf2
zESDGUF!{qU5-DU$|HdY+Lx9q7Vz534q%TG22DdA<r0L%e)7;Ecc!G2dlp)Yw-W0>+
z_b;_3ADgI}OZ>E{sQt(f(EuLGqIa;3hWo^WQO|9Xta?Jj7p-PXAm=U!Z@Io$>j|Lc
zitr^Bqxr}Vtu$_+kMM<1P5;PTJ6`zvZTW)ZJ~FsCi*^n8KAL8I(GWLezr}L2Pk!q8
zjiZ0Ky8G~^i(0_T&G)`mBu6xjQiWxl*dAdEA^v7|8zneV0jkp{c-+fX%!_4!{s%hv
zN&9((8%j8G4-`thaE64AD`@->jV;U`a*S-?gIPuwI>cnW>J!LdIeBWFdGoF>s-A@b
z(FwZKs;75GI4v?Vt=re1%?#+tmowEWfii(&^e;?_28Xi5%O4pcouxIgb<cv&X%}5c
zi7*uP%!gwO$JK{E+~P^8OYWv8qyNX)IYn0%Eo(SVC+XO>ZL`B2qhs6b*tU0U+qOEk
zZQIuE^L)?4d0DS(t~u72RW+;r@29QPT+j%j7d%s@Tjvhq73%gR7axL!=S*?OC7MX8
zCU*YM8asXz1g6Dt&O$lt@45%qd+-@8?@i@fORCE_TpTG}3B7(r$&xj8v`xN)9M+0H
znGG^0C|GC5eE)*85c+hpX7yPq@UgDasbGra^RH##j?DSh4Sv?w)Qo|!BQoN6l7VZW
z=z{D1)m%DN)yRWAwaz#@^tbdoL7U2^AYX5ThQ(cvLr^%--57hyVSU6G(`9gI4;Vs|
z?z<`y?`ppsP^csPd!K-*#n@`1MRgm-NuL-j7WTf60NWv;b}BQ<opu0PbvYtgkTO%q
z^N?^7WEPpG$P{1Lb9Zj(1`#mHEfTuji+op=wWFp>w5-}h`jcZWY+MNX>>${Kc0g*-
z<tZgL<^%Ae(VK*>dpxr2<mbx8Pw+Yi%Vu+c__P>SA!2rF+^dB>RXtHS6xd5xwM250
zgwoI}ZD`3oXMpg6P=miC+y2oF$LA3NVYLIjZaKo@>3v1hjk;UNbLlT6sM!J8X|+pv
zD;^sVPTA=IdreS`-&sbpzQM7D0HKI`5-7{jDD1?8eWMzZIL4tnpXT0sNot}E(TU=N
zVP67yFY7mtw}UU|%JTljLqLza*#auC;`D2Og?pNQhX+}p){htIp|j9wyP>cmDg6<g
zHaKC_8JKn-h|qmXzG?dnON-_MulK1}B~b{4pPX4S);fmB$yCXlbreO0M_DKM<^CGo
zm0O_M^365)CS_Dw2O~WOXpjYgFx<yU#Sv!u{LD7=T6K*BOA~p$9llPv8q@P_FY6Vd
zxkxi##X0#QyyuafONv#ZaS}Fm!xl`)f^1XGG&`-v7+SphLtO36Cm(fqD<@>W+JE2*
z{#`|$93ovl`2zu45f;2Lzx#nic2ojs9v~sSUj7><<tUzodSqkm&jza7*TWBf&c(5G
ziH4-$z1;IAPgx7m5;F8owvT_4MJ?9r3uC|pjm95^b&GQ<x_ScG*h+X_+?CN!M3s8(
zVQ=@%9NEe|Z$1w$gc`t$Qa%vAzBUbakSGxQjBM`3)}fw;RF-3KBMl-sv2r&Dr=q6-
zJXF5Gc!tD^kn$9nO|lW66mltk(=LHA5+@b@#U126BPR(R1|{1O2?1Md=lh)fEs)Gy
zgQCT3yFjCgb)Sd1vXqZ!kMoAa;fCsF9;K_pwrV9i_x|8$>9ZWHX{aP?$^*?9H<lKD
zQ?(^D8`bUf!z<_30O6@WhjH5PDJ#_}j@-+vaa<Vq+x?4fhLx7z;Z4&{i??l2b=32%
zUgXDRh8-<~=gB200c}S}?j0MkKnn(Tj-{$}mP3}5)-5i>JO)%f)cZf(Ii#m+Ez#9(
zQTvSJ&mQB|S%dY{oy#AQ(M-yDF5;6))&(Ffkf-f}&BhbxSASpO*YAe|607zFrxBMA
zmX|krw5!`IqqP(rr)_*MS&mE_q7If4yFyL2ikfM3EaQ)w#%4kbsacDwYV4~sfE6PT
zx=u0m2N6UbYO;2P&S_T0Rc51Uv}D`^2{f$eo!vA2Ij6{rX@EsF7o(7|_`f!tS_m{E
z1QhBBG>-5pGFvWo;0D)!<LwI=P_IEWh8Fyg7X9e@AN^~iPK7nh+W^f)$8KfTd;NAv
z7=3GQu$$P<w!1Aru6MY)6a!fkNy&~quLjDdC0q;t-PD+pa(d14qmOu%Ld9Hm{vw(W
zaYr=F!$CA;bnYF~yy`4pvPh9?E6K-&de#&bGN*ioGMSPAN4@HaRYmD+yteiWZaK!H
z&^L3({<b86)#1HHV1lM@B>{bGQ+z+qyYOM%Q}%f>eZ>P@lX;nz3L!&W6pHRz%<hzj
ztcH(vn$OKJ&{p$Qzd3n%Yfn`p51pSV)l2JHpqA0%IH+fl%0YIUqHiEX&e=}FeR>W%
z*@aD@7Sk|F!L(>Lh8V8@@<ezt)hhj;-cEIyE@>>Yhz)1Ai@Q%L9XP{gK_zQoM4%S3
z#{SEwvIQ~RM2NxEEhLK-FG!Z9%`qx5LixvFolD{p1n%7<;O=wp0Z4+xZB^=AE7-bQ
ztO31h3%=tS?J>1y2CU)cZ{&-$9w?fOj1i%3+fF3R5KD|5yDc?(R-hczoj6puZWJ)_
zWMJsB<Nfnihu@A8kR`&)YZ%tQ3p>Q>dg7=*Il3=IRjVaem$HzI-?q0<;T#3it!8?#
z_og+TU&d)0h={dkzQ5S8OqnGKP2FVD)wF-o@tE{vf|%daL65YxMcnX;aA6u?*EF9E
z=~tSCZ&ku&L7nXUd1T3q%$PgU+8`qe7IxpZcL=fD#i@4Pm2R07Q)ATNJ_olU4XRy#
z)vTdT%JdbcAx;B&%M(b8j3ie+JDxX~uXM+PDqm0K?!vQq?^~(n>}D$^uD*TK=4x-z
zkH9ZcW9%tC-@QEQP(S$h#<IV0>0?)35aAHk=G0!2Vfa<a)$z1WTX*n&0o*a~XH6V_
zJX{-vN^zpq#!O<S{Ock5W1m*V>T7^m!*nM4)I`eiEB^N>#_NSw*;|fdZ~Af}IL$O{
zf~B4;F0cN8$rxztND^Lcd}J}+ji2@?DB?&Ger^1?3^&nw@9QjY)8Kcd!Lb1AtN-bW
z;1kCKl416$_rMWBKgJ%MN<@m#j_iB-BtpO3kT13ZYummXXfdfsu(I_w%wh2k{2O@m
zRwn&R$UY4pZBJaxb=;qiX-{kqifY=e1_zrcAr)3eJ!5dxU;E<MRT%H}QX-r)Ws#xD
zpYAqF`>9|@r&e38^KF<R6>&4SkyC#dxi`YgBgR!!CZ&@K%HsjDc~ij|3v3=~O}$g-
z!~wkSYqqaB%Ou>vo`YT(FEfkDUrDo7!=DH@>>1Pt|2`|v?jT*OE;t*L*Q2e8sAfQb
zB9qU6<KkL26{k1r$4B%t-4zk!p)YQ}J;HOuINHU%lF0&YnK`U8eXtUn$Nqob!eZfi
z$G3L(y|SD=<V0J;7LVh9C0CtCnr2g)$z+3hv6VdCwS-d>tY~|$y><4by|r~|qVeUg
z$-6~w(YTfz_HAkLtl*QUNH{i3JDJC)e1&kBQu3A!-?55q%*h*w!5|U4XW|vw$-e)+
zPnMmzLrcHC6x48hGX)#n1q=NJvvtG8*M@CIrpz8Hn>}H*w@CR1QP6{lP%bU$nmmmw
zW!s_wJA2d$-6|($RZ8McTOh=jbvDh!o}_0B>Wff*d|dgzmRbLwV#dnM%=ACa&JJpi
z+g@;>Y`)War0MJH3PdpO<b!q?L66v)cEZ)COVy9+$Z)Xc2!^7SQIsW%Hk+8u`jk?u
z-Yz6oCcK<JqY8e$+<ZLU9R8jiWhn_xDrwo41JkqrK*Pg~7SJG0Ror`cx_r1-n=0x%
z#wOg;ruL%ya(Stmjs<W<A-%|l-4~D}5B2Oh^*tNnU|weo4`H(jpltIE+5x}Q#T~qU
zisSH3Mp}U_1POfKJ<Ws5OYcl%+?c&Aem&nk-Q<M22#dQ>NM(IJUJ`o=1DFmEM(s`g
zdvkiSeLP=0d97elQ2iot3!EWSc6#IFhTMw#)ljt|LPqyV=k0p)Zd-f+%OKGh^8r`}
z<*&mzJvr&L-!HfA7&d9M<wX?kkdu6CSQJs(Q3#!&Dp@|#Hzm&PP&Y7pbEc`BqxcOQ
zjSE4oqZw!{`d-jhAKY?Wj7lQ^L?p8@)Dx+ud()shFb`BeU7=TCS=(KXe%C_YZ!{O2
zvdD2pR%Kg2N~S`Q8esgf!7vuNXW3;g@RWiLxB7VeFn6oftfJ^-wZvgn$TZN#Z!}YV
zmwPC5_?ZJflc>r<?>U>e)cb*{q&1u;a{g)$uGZHpMrlq%6`?MKCw4^<2xH3hn(v^>
z2L_TRN0hW|Jvs_2^noG}PD&toeiKL;Q%<{cGG1}a>21g1T^|L^N*baBIQObQ?ubsj
zx+MJAC<3~;oJug+;@>MVVln<PjXmxBV}hljc4kw1!<Cq=LR@(esNMZTL_*T0RlR|I
zD($8jUshgCE7B5o3F)f`c*aYuq>e~FDvJ9%*vylrs7QB+KU8hPL$mTIbiLv>#=mCJ
za_Hl#)mKPnt~hb{LN*ytf3y*%w$wJ_SZInvK#42G5v(OOzjjT-{i}cYr|rC4pS_{u
z(>>$IgPqqaD#g#;=N^ZX7A8Do4AYz8)0~k74kh()g-WucpxxHARSH};6;U7YeY%2Q
zdLdmB`X{Dp6blWSkN1!LWEi6J{R>b10Xp`gblIvd;P|wue%1Ja2w<!Z{!-x@?C+LI
zY((wBtA!#s)en2;_XRH=_#(!|2lzLIe*}Az;&eix(ju2hN35#=9zovDbK*md0=Q6{
zxjBkrwe{;i|G|+QW?LQ#(#b+)7|jfNbanibK{ust-Ei)Bh@}&s-|iyVW(Fz?^AfO{
z#ru{3n%ke(@23bWZ3bI3YJ+el=o|<^TKXflj*#2oUXfi=e?j3ON+QO#ySER<_mhd6
zML=%hwHP+*NzLK6sg75uA3}0RnIt4eHJB~Dr3IjuuzDT!Wrf3kIEves9c<9=2gub#
z!K?-YUd9EmQ&ifG`Fd589Y&L{66+j)Xsa1<;IgQW>}@Cx;h{mL;Jg9rV^_tFa?@%$
ztf<nevk1vGc&XgZ(@K}SLu&`t$dOcc$H~wZLu=W7oGc$~*%Hy!8L16x#G|y6=g?CH
zJSWVIB^wu@w+S<`Ke=}2H$YQ#V)kj77xEB4{navGkN8OzGHE4dG8~Z*2mLZ8!th$Z
z?qvX7(uTha7Y^EJ8^pMW5oz~w=RHOlX2D9x4BL|DB`MU!Pp3+{7MASwbCJPf(W<K^
zwcmN-fuN}6OCXZr<Plbj>f$p&EJHw{%{WAleyt`Du*Le5iwm^;^qO2~4(py><DDG7
zr;xu>Zfl2yP^b!8da$!1{Ccg;YzVS5PAjsoCEz)DJV8dS6>o<&bUdP_+YM$R4pbZd
z%;MLRfs^scireZ;52Y?F|L41$cdb33kT%3cR*t0v&(FW^%-B)nVb@j%f_?w1)4yX;
z#@^Q&%jV=dSg}kJ-_PMDzg+i&U?^+|M}yRngu)#qaz#op9vbd~f`-Mjk6p^wuYv9}
zq9LpH8lAL8oQpI@s}XxQsx{X0JH4S>T?lFwQRlPI<QL<vaB%T2aO+>)yuFT6(CvN!
z;kx8Z>3v1yr6r;smuyxwQGhYl(_{J|51#Y5{vq&E^}8r>!2H;J65*WG0=W)(wrlG2
zVHR>j$)6Z|{9p687NTxEOZ3Czjg-!ZQM2&J+;h7o4p3|c7>M~P`5}T0l9uVtPxxG+
z1$%otoP=3tkWgA<)pQF$<#Ayuu|y{rCtjngyS^l8Laltvy3Cm|uGYy%BjLoUg@x@l
z3dfM$ZkHtyO=V!F8XEZ(tW9~Fo_SFtb3C;?qwOSpdDpoz>^g@ptKy{CCm%zI`+O#e
z9BG8pQfb?fCLI?2@c^sX^*Zn#@<KDsHRV1WwV}ECC^ki+&``Y9w7%)BE<r|e9^osu
zKbqXM(mbq6HMPX@H@}<SK+0jpdhnc+fOK{_XMKTpJ{g1Vkw;}`%@Hi>5^>CPm=tn9
zbIRlnZ@Fw+!hD#1L)^S%E{*wynw(?U+YPw4jQ)7;M?mDJ<y?oA=!l<)4nyXIy7vMM
z5Y`4&U}=VVkJVB%7PvPtR$HOOIL`)p0sk82za=+ch%VXvQ01rur4-5=vlb{h{Aa=R
z`ihnUJ(z6t(UKVEo0Q9n0CA2pq|zwVWE}aRx%~jiI|^tIohNPEu>n+6%hR`x5y@hj
zX?h49mwh)=gH=>|kC@TLkxk5k>Y-85tAf!@LTaX^;mzQ?EI^563h5*`Q)WGKiLoOx
z%SWc^;}=<CPrA$^6TZ-nh(8=q(YH5>k_yb-W-4yMR{o)68pYgCuLRF5zKb`v6lHU5
zZ}vVQx4d#mqu(q@mqY4>Z|s{;=4T@N4L;_#`UNY}^#nalmw8qWZu(AJ_GfJy#Ax*k
zF!V6xA08@Qj7&quHovi`(hNrqVF6n@)O$x=c<vF!95$yokb17{WRRA$9z0Wd6re9E
z@f+Y*G%S#p_bnBy;3xyoL*C}s7sWxzL~Qn#lsN;O#E@Q=u<*_$L|mhiAr6zfG;F0q
zj^-I$ak+~1MOoevD%d5Zb05=+c(M#k=?vhZ?G-nAL#m;-2g}Nu4C|Pf;ONuQDvz$n
z4KCiOA7dL#7(QDJVKvO~&NCgo-=+G{I1wl-A$z;aeBYVf<Lc98=#8dcj2l5=jk`~P
z7;(o^V=L?U5I<z|5OYfJjxtKd+C(i3Uy<Ux<)}0TtF-<*BY5D3E8*OTzfqcNEHe@^
zB4qR&S^G5rww6;*^|)@*N~FdC;T9&iEcLWq*dRnDoTQmW55-57%AfcyTOJ)2#1%ck
zq%Fh6obazbwYIVp0tKopjkra%E4fH{oP8t3wp}DTyQ$EI;Kn~<`rTIJPS*>izzJSl
zv_z6^xnfY?Rw_&W?cJfnu2F}Pl$ShEGMh=(3oF7c$A%U%+998!AqAG49CwF~R%38g
zS1r?cZJc=O;!itwzlJ?@-?>U=nn26Cys1kEI`-DF%R6Z5<`sz9>ZNLCxWK?dvisVV
z)@VpsJUoH38@5c~d-4Y~$4=2f57|oEhqMir#i|p9Sf0czeM*xqMirnzg97cXr`eOE
zUgpPF^n9O4*{QGNyts=jBx+Z6cv$?kG_R=?yVYe9FrZLK5qW<Z+I-LH%<`9J-rk>;
zUwo$@b-M=FABx4c{@&F5V5V!s#>;~kACK+oFG8~3^y&?f^rL^;S3sxgBdfUPug~BT
zHkdA}ArDsKS!@~&`DyvhGkKV@%LO8vJx$uvrTK1q#1LosOs|O3O*xiB8TJQ&c{4s~
zX3Q4J%jc+r{vHfSu}{`NtY7$C2_Imgc8L_-6{f}cSLLR~SznAiIV=r%`ElVX$|&j9
zI=U8X9PSrk`$HK1%yzGnJmS)clU}3d_zgX*1%|BA;K6iwJS-6Z-bo<5?YUjKodayW
zx7Hyz!)jLQA3E$m@C3U+-nGjXm53zNxC(x4jLLg_v>lqg55f$5Qm<?Vv5p92JCe$e
z&)+wr-Lp>b9X1P`-`rG{BlYZ0(&tU9p1L%%52Lh4g{{rHnrrxkOR>(w2E2Z<T&8co
zk|bS3tu^hFrl-6`b$SUg?YFBX>;ATvPKREN6>T=~e1$oMz6*yG^xRab;mjFFT!U2G
ze$@WbH6`I1%vH@!KJb_UvH~~rY1@Kn^@4lXHI|$r-|t`pyUGeVs(w73KX%N+hBm6a
z2hK%!)Mqv3V}I0Hh{;-0R+%%kF;)pdaME&%>9r93+5HBQX-xcYH3t{#|I_QSva|nh
zCbEm#(spG3$jR6LFnZBK+T9U|LP$iT384Rc%_ncD!@a{19S}yRr`0Y@5l7Pu5*uuY
z1ttJzr>5SOm=z;6YS}*Aqw~)>y4_xHAGQ~xkwRnV^C>y<uzHdo6=V!NJnlt7tY;60
z%k#0ZN|vG1a003Tqnq0o{vAqwek^qR(0)AtKfhbW%a^N{?roRmmGrdDb20J{!f3nf
zc}@y}YF8;QysJMx*qF;`h1Aav^C^S^qt37`KHc804>j`Z#$IT|Z;AJslF}tTpKqtb
zIR~fLFL$AdXcQsyzH?PQcb5Rse<(Y7-_M8R><~YVxFswg03O*C@G=Ef0pDv!FCWfC
zm`^zg7}Y~XGt5)CYmMfcPRP}&NOET$w|SWG-M=3r735bpGC6L3@Kl#rShH=}E79s<
z86WOvc}V}GaURim^LJi}M16TaiTMT|`}RG*k`}gvN?T=^N_!B*;piX}c`dZkK9Ox?
zZPK!T+RO#-S5{3PDxeyQzItl3T=OER=W!l<)ZE2(R<U7q#ob078qv(aS_kzoGV1hc
zeJmu~dVzK9ih2xkm)$^xI=@*%aZDF94&J1TX^7B-3T3WL)yGk(gXZW`>ffB(wK4uA
zlve()>Xy!ub*#H%*AjU}(UL5BNNh6d!$0IF8Q(q~d|-$UBp#@x9-<Z?E~!jxw*UGw
zKK**=`T0bsR>Zz<&5_dk$lRghaSC-q=x-t`V@vWdx2ozDG{nt<uem|}JFJPZfDxup
zTAtls(UI+~m2Ca()Y@T&29IbAaPzb}eU9GP#>jHc2j2JB-9Li_4Dz~;;&^v$z>xmZ
zcb1`U-7ug&DDGATn^-O0aYre09e<cP)gNxV^8;2NvTqn5qC;A4`#m%XW{qxj0N7F2
zF#6ox()tT2VEbZ+C4~?RB!ERFFW<M?;usgxH`P7lbTCtcF-iDZCJgB!V4r%s>1USy
zoUVTGM!=|ZfD~i+L8?BioTQ!Bt2Wn4O1`B=Me3yoS4!eI`vS?*`zVMqXhnB*(6xJN
zGnftt0kN=UQv|ja5=G$wWF8P@_+Jsma>7uA#(Ts7OOBA`RCC=5r``8f;ABVd%lK?*
z)GE_tSHa0&!+?^qu1tE@-VE<=y(>0)tNokj*$E>&tUB*+OGbmj3JFLpU!vA!@rC*n
zP0`PzN(X(dsq(nlRLo>UsIFQxqoGcu-*E}1_x<+kaMU6Zcf;<)h)W~W1gR)<>*@aW
zl*KV@A#D&vvr?iDVnw<9t_k)fAc#Qe9yX-|K%sCYQxY%>mdI@C-(EQ-_qr^Zk_M+Z
z^;V)yO@lTd%-i`5)@>Py0Iv7})RvX6si{7?Ruq1RO~nw!!|k#Emh#MuK<7$GUhSh>
zyOckHXrp~~aK}-B^r!TqxhLO{vH~JDvnY0x!~Edd2G{&}V`XjWnmxB(9d3J)O&CHu
zA1Sz?Q(z>|v;c*5$$ElFD<c|bRkx+Ca)Z?{IjJK#KUDQ9{Pj4=6HHyV&C1xQZr)?t
z_=1Q@g|B90I%{Ru+vqgLh7{-Zxe@taKc|N!8w~^9r37r|3T(9F-mM=F^ByekO%4v6
zwyd~Nl-8?IJPP+dTC<~}d$)7uC{s<3r+fD&o*KSZ75m+DIlE2etB2r>BCN&P=(T!S
zq(DtPDZbBIUnj|~ZQqjN6QFhR*KoWEVwK<JLbCY4q&aNrAn8Ggp9ZiX^M=OQmTp%*
zOas-2Tr7tC<M!1K#uslT`G(fmlCJc;HWl8IByWIwx5p(QJ-5qErT@59#Cc_;<dKL5
z^4>FC08?h1J3mHXt3^!(7Vp9>>?h5If?cj&4u)rHRLXq24G_Ab!l*!}12mF5p1uFX
ziQ#C#WS%)I{D#)K_SEv{Kyl)!_H*fD8whjiXC+EsYF@r2s~PKJ4af21zV8Ik_mJ>X
zugOHsC-9MP9u4>r%ua(3!~#B1;rYG~pPFGcU^f}d=Q=!`XyW}YSHdayg-)gz_Ri|r
zpQDBORc2JDL%CeApR`Rx7qa&CdYPi^bjY}ksT)B%)-(q!S8?uMhYjsW??_B&c(?fr
z;pd^z_S3<H<ORKiZXELC_Upk<HA#MByVxuj`RZ@?;>1oz1?x<ZdEv@d7xB25tqZo+
zM^%cI5O3t{O>(wVAXeKf>zmRzt?Nz#m?I&D2y%ruF?W*oSMnN8;GP<F9?PRD<;-Fe
z{h4_*T{2~}@ju_*!%G@q=%l%(uB%Su8}7KL(=$FoNiCVfldX7r9(S<?7Yn9Sb(z4u
z0j}tR+EI)=e!Fsk_|`XGHnv^uT4dViR9>MXjE>xgX(MA>-fthugoN|%W!nnlu@erg
zuq{})*Y<#(<elMGlN)#wW^B4h0RC^1Ty8P`N%Hbv|9V`7$^7O-t!etzXj(%NVKI2m
zVWn@zR3{SDp3&~flX;3c^NPp#=1KO*i~i!X#@~JutXqpxq(h5*Z+H`>5~vM#G{W@j
z5NBkttQ@DO^C&4P7NNXD41)9-sE6fAM?_9M3y7uSR=30VYJX}$@TIvsv<O*jHniBZ
zFxt#_=tN5P;q+iuSM}oD`losgRim5WE3+<*Iq(={e>li&!ZgKY-7{mMT#ig^cxz)M
zI;i_U&*7@`PM{aUm^hOgAOBq8LI!r*KI5)qs*eCIv(%B7Kebprk?mOuH(Loz=p-@^
z?%VBE{yKb@XuTG!&$pYlZ8(I<2&nc$Bg8w;rAQ-i^gB9q3L2Iax@{4{Gy%DeG>d@3
z+cZTh%-fb5gRWVTh)O0WETyX3d$Pi~h*L*RFW-dXKhyr0D)g>9!R<MT<;c-c&8%?T
zr^8578>QKUEvLg+X|o;F6akrCs|r(zU2X7cb70{y%!<m!cgC)83W-n={-{xv#eehK
zoAL2vX-?@r_Q99!O0cEhAs<D_c`56`oxbZ#4WYI=axgetqCuVV>BFWdx`|Hi&OtTQ
zI9%SCuDtF66@SuydQlsPas+-DbY`c2w7X62Xkyl&EKgzM_m9xB2h*}-SRkkwb0|k!
zQF@yvJ>{l{UDzn$9y<$ggnhE@$AiEn_?#4CWBCVgH}6><Y%4qDIWI`yLEs8syg}w}
z+ET$zR$s5enMWb-yl9h7g)yI232lZs7zrp>w4AdtkNJ~~N#(&2cwD@|aqm9Njw{iB
z2TFo@LE!#7f~tSxV0r^wbP*jOh={vNHKbOxDpgCeK$5eXA)1E;2gO`^9rV3Ak;o@O
zSSK7MO4LEu^L`wxIg0=obfu{sD_ycueHe5p{SMEXAc}_qI%C2EF`#gR7`BMqt{`~q
zzzLtAK!<S-IPQD}_j*-t(7`0&1kU4t=3!jDVzF(ROKMqccHEvNNkZd>g_kHD_0+S7
zuLI^UtH7p@SaK9=>Ut37bdqSX3Qlpdkt3VpH{zp3jwCex2zibg@o-NpLLBK#VZrs0
z9FJ(6#r6FKU1w@*`-XIPVgsX91RGT~r6HG!e~Y_3gH^|}n)nwrhbn=F!#oY)r^i#)
zVCq*B0DZQlbXkz`s#S>Ejr646jg>D&Y7=|#({!yZ^^-hi{1cKnC~6Zg-O(bdjVaNN
zaq98506oIEuN`egqnA`yO1947bj=sts~>Vy;H0;8A4C1gMxzC2tf_oM@iwCcBH~8Q
zcR{Jzlxl_#>Un&5e%e&>MB~AZ!PfS0MGievYjB9U@9sYI(XlBt_ND-=q)|A!5j(M2
z<>x8@>Re>|r21Xkw~8TC*;APfD!KKKJ5LOQXsMohvH(vy5|#rj*+v-Kx?gT|Czy@V
zDW_;sPV0&18WM}89Vim{!Ic~MG}N$9m7k(;`f_1=u!{3(D6Ls#Dn?2K%10HubreOl
zUzot^!5!(<ifYuVySd4qILxia73W=vX}EIdl3Z4d;*n<X#SpDi$uKdrkUMSZ?_-^k
zggrVs7aCQlt7`7&{n^z1W;Rbp)wLWBb=!Pd>}{)ju}4pzn;LAUZ(#D(q>I5Ve_4mg
z@S}>pQ)xX{-*Tk>VdYVMe2Fgw)|nyMh@@NAs6JX!11HtzUGs?6ZLQNTV`!JA|2Asa
z*1?`Qg=yN>sr8c@{AnB1*4j;Hp(H;TE3T;iD4G`3U9}C<HfUTg4K`)_=wKT$Re5zx
zA}W^^ZBm1zM*lESwI81ymp_a1h>zY%QyMh{8?_eM^R3q`r&Sm~ex|B9GcHQ!D8RI!
zoPI7=Rhu*C6_Gm`-13lfC-+PrZCp+bcV8NVlQweb`q@csl1-xvw{8%z%=xV0`+O@*
zEJSOmm{v~TvlDE&z)NLt7Rn)V9P~%`Suuhgyh|s=QjtaMnR#i>AFi9(=vE*q_9>*!
zy6?RrJm>j#3RyZRz;#Stn;h3s&#qR+*sgGUC+he{gjT=ct@Ne5=n_)h+jZL94Cl%_
zzW8+$;OTEix!@SIAis%f#qD(#k8-TzqWAQI^^QI2-H6>t*pecO>2cag;^dkuyJ{8L
zxTMQzar;34f~~Vvw#BpXGJTCE2Rg%-Q>XXUQWvpfwxuFt;ay#R=KAC9w<W=>o56>w
z+{p`2PIHVC^)BSO9kyywuX<_OlKq*<vf|~TJ;X$&jpUO-Fy)_r)5bKZ4z2lB6<Psi
z<43HaN!~~r?c670p2<2*21n62r?c&p8Uy>-cyz$!WBaotsmSej^q;Ct6~y{5R81Q-
zDc`m9qvz3Q_S~~?P*xAYjQ>_Ia{qs&5DOd2|LU|e?s(CRym?Oh0v<3MJyda)T+U`R
zS#`${?!r)$xstIV);W4za!b}XjP%Q)K9>rfIeMs1Sdg0ggL`E6ef;qLup}eD54NOl
z@8)$F<U{CAKJFuLHV_g?yzl1K=;bj3X#UcEBc?<%m~yy{zx}vg0X9tP_YLOe*#u$m
z5nV$o7|O=rQxg-7!lpMOzqOBg*mJ0uz&Jn}E$jHoTD$A<etDmQnU(_L6*~I4e}5>*
zcrO-L4>t3&=i3biQPN)l0h$bL>UFoq7v8t?qbh<bzsP=<^Xt`Hthi2`t5Nu<re}0F
z=i>&Gpt2M93pezu^nPS_dqB`FfWJ>}gi4)x?<-rMJNKs24TU#iW(-d1%K<qn9UE<E
zGX+-3jUN-&{m9RqB6_xc7yBNr$#7NE4I|r~!q>->9R*w>)gk>=c!KQmb&76|hY$WE
zfQO+rp*mb**Y!y5VANO)o|8sZ0<$SF3CaLjE>UZdU&D4WrLYz%r?E{`<Se@DkX06V
z#dBeE-Hos62L0oqU&@&SNr1)*h%%>Ug=4LE&>&?Mi`Npo4oo9d^nVCKD$V0ppwtdc
zX#&UO+6~Kwz|02;MrY895!7SBe+H4slmFlafr1q2lB*5aloZWuZLo3_=!4m<Lfsso
zy2!=NjGmVlNs*^i|2><8%90+0-4!}bL+m{ypic+qg1jw_9SZpvJR0NnoagSg_&mF?
z^}~1D#1~YS8j^6z@}PhA3u{wJ^e&+mMpD9yB?nv%w-||qyD42k`)INalG{6uK<^Eq
zkfH6eVRK{Ev7%cQj{4JSBePj9ms46W^p#7@D9JURB7>ix?&zeIy#eE3RGLR&z57=>
z8uPdnr69t7ePBy?_BS=994;-eoY<&#8yR>McHs_YlIsk4LYWpJbp_%)Weheo_<JXH
zm1;fwPLa(;Xo<Sq4Pk8Uhl#f=2&;?tNdox&_pfKBpS(9p5hB^(j26O1L8vi4J2%s-
ztf}EJH_>@F1*l=~eezf;oNx%g7_#q=^pXE`Ge~tBLqpPdErs70khvhKQlq#d^hZEo
zyWttm*?h>lz#AqJtBzOrt2>MQyDr2JV-BXA_Qy(^K{>XN-7!VYXu0V=F0mx9>O)z3
zms>$#w1eh2U2PH^_XmZQwH$7Axkcxc>F1l}-Xy^rIu^J(P^NNr(ROd%D6fI53-gxK
z9(7M)YE>0<pT?k}z+HC^lv>-+(BbaR#?Q2lpvZR29`P`ciS9{B8t^DCP6Jx~t0l5o
zVzTu5MmhY#p`q{c>P~n3<b6XK?+X6-@dPIC`(1$87;4|*Wk#c_u2D}xPR5mpHxRzK
zltpzm(`jQ^FO9R<pF-T87`ieoP&>LIgss6!vwQY|MG`cmHZOQ%{hb=wf3b8`C`7~K
zgBD{kl#vr@X2MXtT2y5Bw^K~`Y+Bf-UHb)qg+HUs$_W*FVDAV}cNXkivvis4u*Bz=
z*y~7Ij2ZS;9*+nI<E0fWuHMsyCyxdU?!RRPqh(HE0Q$Zrho~y>x}@3~F*vyEe-PbW
zzF*WVb{Dc3>~H2Xj>>m2*o>wvWH<D!ms%*u!I4|JVej5k560Lss8JYMD9ZlpMY%Ko
z-8+lz`NA{J`S^TqC1^GY0Z&?u0%=gSA+Qm0b0KV)eM|Tg$+LCmf<H4D$TT%sRG_^W
ze2lb{(oXm5RN_$xf}^Ax6;+g;j^T+eh)jNi^WYz2VZEXNypN*UZrk$Ip${M}f^1V)
zkN1Qkd2+u0lZep|gVm-^vBG@Vv|zk^YObudKFcIFi8V{k*aVX!De3ZrUG<+=YCZcw
zg&ODhDC?flJBIowir~_Ob8^pdCeX280p4f0-p4_*sF#pkA(3Xpd42U{5@R}0ZZai^
z$s6n}e+cyzT`f47KYF;WVz1zcYvZN90*o~7p%G>7ZHv9fvsd=8U~8(`0;^ux>K}>f
zc4^7zporbIFPt?HxB#XTZL$~NmTWtTH6e}D!6Sj6<J@16J&gsXwV`gn47lX%&IJcd
z5wp?#S>fuR8|1PR_~!Npxf3QgJDxLB=Oc?xguljoiy_~6{B0@sn+rA7TY3v{HhaC3
z_?po2yExD6SD2CN*DY5d+R*ZWX+p$kYJ&te2PhMv&beaIX;xE83wYF#e6ALu3&aq^
z1GTIRe^o}m$D5asgeYUDa~4lNiN9r?nr3uK3;?}#B?+9zWO=1b83F8=$0o?RPhk=}
zwTg?UxhB!|rP_7lVyVkP0Cc1F8=`VoH##a2)`fxkwKLI}q)sEQlFk8*7XxnDVq6zk
z=@BNDXL=&LxxiT|_F-z|o|;m=$LG9&+=KxwL`=Mwd)<xdkENNbl6XT7^dem_6UH_K
z$H3XL1h+n`cA1NkA04IDP>39k8fb9{V*zZ(_g+CdqP8j2TMJe8t=XKoZH65oUtH1S
zRo+}^B9VvwNV;kBX#^O2g*3WhT<sQ?zdrDAE|3HZZYeQo;@n}CUYtp!h3>)^;p8M!
zVR--KI>agq+5(d|2Y{D3yTPgCYY{d`bw$H=h=Mq=&|E%rl=?ZCsKcDwR?W-)n%@<-
z`^4eV9EK|Hj{NDyTJ;*RDhk!PhnpGvb6sBP{`4FUcRqjxIAkp({?2$)gorHB{}xZ{
zf{P3dR>X}R@A9T2I#NaeNHcz*B^eztQBi@Wx#-<|F-2{)QIR?6$sR;4GMcyJ;6F@l
z)ZE_#A3rXW^;aLhV^%Eep2GJeBp)yct+0ju-~eQfUAlS`?jlH#Hk>{uYa*ROpi<}u
z$ocyG18Ch8cE_9poWjgzw_J%N7`ov#<2<*;8APQZjx|qKA1Gg|!0ZtLC&t{t&)yo6
zj-B}2F00}ZfkcN7qmoVYIhW&PN`f`op_?L;eM=)1=Ny`O2xW2{N0%3r*iLLi5A(3y
zvRh>uOAtrTIktG+!C`!E%}Is@tY8o7rH&frwJ_E!<2|B4P5<!7bWHLxAfA;+BBDZd
z?j<^(Y%U4=cc3CHnyKb2-S1I~#Bz2SUVo1d0+V8KlVfiiwes?6Hqc2xOCVuwh2fR?
z{?F~`l+?NXMM_z2=pZA8X1KrWS*o5}elHA5Hl!I(B0Peqny~tMv&p~Er=^x;w0Nnn
zMKUm=$xkl{PWv|(y3ec&sGP``95A&NW7IAe4Hw~6v_wg#z$$}}#xoI(%Id23WUL+O
z+Tk-(#Dc#;fmFL2|232QFa9trT>qQqyGv6l7PsR+o{0~)uS{$yGz`5$VME!>aT(EV
zT?a^5b%DJIcL^Z>R2>e~9HUoDJn~R%2`k2z<7&Xe2k(UWlFj+^tItUPQJC1V)6>WG
z!V~|(9>84w;3zeDrQ-bg^y#t$byLwYYL^he=;ZLEp~Q7V5bMw)G~W(zNr+e4`W>uG
zc+xiZKG)XHsVL-O;o&GLFXKknLDofg*~l<o#C}Nf`I)2l-I^#C{LFX~J8FmjBO&21
z_6}?mkqlM*49SADDxdM149;*1^e?LJS@kbni>W40#NhHhr`KRYH_jDD5p+2(6tgR#
z>zy+g>uHIKzpUo4C5?=$aFMDCrD&_8h_kBGH66hkKvbg3hsWcqmhh-o3joHZltIAo
zX^*ld#@I}c6s%bvGK|5D2k&PjNB8&fvz_nD8z%Y5O3%J3xBpCOK#osxO-N_{tgPUe
zE3M?K5KEX)b}&q*LScRd^m+fX1@s!S>%-((Y%30g+e|yRHIQj*FnS_a!w{Qf(7oyA
zeU?GBj3$!o4?3w&q{`UpOVZSR*$aF3+I44=+^5Z&cZN$&<nLL{bMwgOHKH!9W6-6_
zatJDk(SIIoDUIl8NU$q3eTvrBmCbKcT99ohKP^sz3oclc4^^FylEm$FNi=2rQ_9T;
z-a_!9bS#RXsA0)CagbszM2O2HoEw6Vr}cc3owiRGN{;yF-4)h;2=Vf(FNheBB>}d?
zN2xl-kD3ktTC-B8q(U#}aJ0PL|Gl_h7)eO-NNsxaY9=p*BF#2}NwypWJ2=v>faHh}
zu!`e!nhaI1*)*zlC<{1MDpE2LuePcS+10I3I`6%2YZP4gT+A;@E?(4(KjxPY7cYMb
z2*cH*MLLv;<0+mPN@Q$+Z)6q$2vVW1OQWnM=f6+%-=Xd}%D@Zpa}0j=@;2E_+-i?G
ze{MBVV+>(Y&P4SCdA#Y0%ios_En{Nwl)qEoY>G|<*?xaVEyWW5kZv+=a6X${(x6ja
zFR@)C<Q!V7<eO);VH!mda(Dij)|gI3Wt-Yr)BHQ5Yi{JtAG}GaR;Ebl1`pNwus3hz
zDzNwM$|-Ezt9iaMI+f0!u0r0El(2168DzxWH5fFsANAaVl~TR_AkQ2wxMy+(P9y*h
zo#aHWW@^nlot~tZ*MQj`qIL>>?Zhxk%#<t)>>{wM9uK}!8Teo%WfO6pI{fft@gg36
zv#vz-6FAKS<=_>^+nNnT??d8vJ%6xyb}!UKiKH>1Nx4NF(-vq?VuVD6mI^1XDiT<w
z2+7Zubq~P^c4^QM;HYLCk$%SJcaqF89qOE0pap9WYIQhJ!so*681B=p8ndoaP>o&q
z6VIJz-Eb6!q321%I+eaik{541_Yk23m0_GyXEQiyFf=mi$o(OQLFN(K7Qh=(5vSyT
zQrrfr25@y}f#w+{;>L>w&@7W5-sfGY6*W5aQwHfh5D%bk)u$sF6vT>4V-YpJeca`^
z9zs-S*DgSfFkXfWB47AQyqfyHzIPOC;PthAoqWfz-t&aII|XP|@=AK5B_laq6QuzJ
zU6$;S0?f7x?-r{l_&}<P=0|@$MlcN4Hn=N;c<+Ff^7!`H|4OjAZE;dui(7SJRy66#
zg;#;m5Y(k+-P3#x7W&jtC5s3x9u#P4r8c7FY;OJVYP3yiP)$Ibe2$HTZNOH8Rg!5|
zT4LG?LzP8mjPj7$AQTh@4^r8uws!(Ae{B{9WkPUIxf8?D2cvu8ut}$xzO5{hGw}e=
z8v9MIof))FN?D&YhZ~@KT%S~(AYLC#sB~mLFT$^PMd(P@rewt`qBi(-k7rLJ@JW8F
z;rGv}8;sFI$~f<?5oP?VG_-ft+V2Sk)4w33WRjc#O#Vf?O<EfwJ78-dBJ!?mKSfu_
z-z-DJtV!FERn=bg9uffEzzs*$qmg`m^P`Bl!(nFJUm1L_$f2zUbS@_R5JrT$c_!0Q
zXJoCLd*Kx%M6??;cF0<3it&c3PbG}?3?Iazb081JUVvazG!oMD2|`W(cQcwO^{YHe
zx0Z#ljej1$0liwN@u84<Nyy}QTP6k-6nR<cSfh6cxBTiH1T9%=WUqJor1I@hWS@7-
z-C<)NboJWAOhZT15J~tplQ!E*fMz2~wMJ}jOp-;qjLShqvfKW+?;W~y$DBu!jrfcF
zn}@Q{{N>n<Q-<0P5Zd#r%cSh{D&5nvtX*{13^fH*`@2QjtIICKACf_yyhTPh-j2FS
zde706560rMQx21Z<O(Gfu7YDV9OJoAvPSth2pj3;G(T9svoRU7sdS_&+@#2IUX?|i
zPZM1umVxf~Aani`HCH;;(A9QmBH)YL>>=Xt$-;k-7g5Z?e>b5gt3rq{4_x8yD|ym_
ztYbLp$k=Wf;QKe@bP`SYP-qV|?9|6dd7|p-Zmihu9P)&fB}|pK3EI{Wfg!42qHlN*
zp8KR)*sb1=trdk<acv*JZ($*sKUuaiORbuy*^1rmrajaVPq;A}GgT#>|7^ZW`4Qg8
z8K{S|88hWI_T$3KV=o^Zhzk4Bd$|TPFseR|0S+Zbk2KOT?3Y(S_7VavE`I|L9hH3(
z4qCA7RRT)Nf_SyXZ@1$?3dAqXnmw14zV{OAQ3kWddQGB@tnkPZ-+kzZ%S=~OoZxpa
z3_IVP2l)-r6$a+sW&>Cz;P)iO=yls|Y>d@dh6=QAy_tsR`o~9b{F>D;cy(CdtdhEm
zt&4&&3`)DB1^vbt;EaFK&c?TJx&6+)wek5t)<=d<Iq?^fs{b*RG1%5$^fLV&34d_8
zF#7D+y;HH^5d4O#LWghHdc}VXuZ;?L{z)6bsCN5T5h$R*@ASjB4C`TxXQ*w>iSTSC
ztX85!Pmz=>1G;MY+bKfs-`wBk^4UeBP!%SR-GZ1|u#;`sz0IO1RNkXF3I;SM_|SER
zIGVrs8Lye*^h0z)DW%8J<DS-`!t{qIEIdb#Kp<50ojKTJhTTMt|NZR)_G1^dXIOEX
zUHo!`#ArjknKK9Wmc`SnV6vdm!+%MgIGO%`>cq^!`M;%3i<;80>l{e`h9+j}xakJm
zZ$TlT>UDXA<ZWtO@=Nlmafm{X7IL<U#YQp6#}QXyJ>u{a_xQT8r>u-JdT`dgJUrYz
zJ$fi@J7K^GGFlknPeK;uYuir>Z$XBT+3dE{w*}NA<#I5+S2RhyJHL2OTCB5S0=hxp
z;Pme=u;<=Ay_?$+FFjwizio?&-g2+JCV0^Vfi0B+a@&8piz=*gH;8O~eBWP3G0?ZJ
z$tOmy`EZl!69pYm$ms#$zZs-@erxVc44AAPGdYk9ImAl!`{}{*FP~DBaQVx35O-DZ
zzIH5*<OcagvscO!<sEgLap`_$HX<-l=UGPflET>Zxn)?-Ym$x^tVpk-niol7B!9UE
zKe925Y;oPNs~w39x@z8cn_tAn_b+3_(vdq?TxhB^(5>Vx7OWfisK>)o#7aMKyS&+~
z@vl3Me6%%(8{TQsA?YZwdwZArTyAB5xtHS$`!g+k++I@Cv-%S^q2;m$RKI?i-5;JV
zr<?oF;Z&HdA=XvpLDS}pUq-1Sx5(w;uV61=ieAbX3qhVD`SQ;>t)E3GtTK5{rIKiM
z9V;nMj2P4#2)pRT?U9*1#}5)#mcH<t90N&~fU^h(JWu*vw~y;=3BNEGzBt?;sz?41
z%r&@acG(eY`<LV&CGjbMhbtcOsK0uF9{s9L%@?Q3rh%TK+l1%w*JeVNSU>@=1aJA=
zUs$&@CpgeyS$c7M8QVb74!h8&N`-|mJrUe1?KW)C>~$LSCRHgp=$I!(tU;>Yj#rXQ
z)!I6%-afa$*4-sHRk0lWGY6EwrZoxA)G1Uq(^~p$bPmuA7H`jNLAd;f67&26IZkir
zi5Ff7;vv+cl_rb{`7>t0da;!%u3$bu0<?+>58zh;&gEg&`OGfnCnd>TO5LPi`KqVw
zNcK{b9RDL4p7d9M%$VfCPerUW?~aQxV(i=M<<`~_rkV1{Mg-J2-esxg)A589gZPLp
z@RT{JwL1!4S8Adf+EdMEDDIOrz3DOlIp#{A<U>qxgaN|nwnvq6Uyii1*1?rYdugL-
z2{{=7%{hE)YUF9ReKv#S*0peYX+p(2;y_5LY;JxA?Xc)U6Q0X_D89uYek|hm!{d2}
zBScB8Im-hd*WGo828HeeN~i^e0Na$Pya5(m;E;&4M^LnC#wE5JXQZ71Lp$4F;x<$N
zrizjV+$Z*z=IBxv0^=a_&PX$1gv3zl1E^?zY+N{yo<DK3z^@T~5qo&67`Qpz?VPV{
zF>J9&l3e1khGbD;RB#zBM@>4I^eJQWWL86NCg#0DbxIF<@+<Oe4&zjdq^m{Kgy4-m
zm6pdk9SZjaZJmyrsw&*_YqulCdyLuK*a|kg%53?VtCh~*XxIN9z-czS#dTVCJL^IF
z8(Pdl^yOe15wXkqqiC1q<v%|{lt&2ez^iZCLh`X<NooH)-%~{3FtdX%p>6)zENj9m
zzEN=Z*HM{QT!2L$@ORZ({o^u%uJNj~7F8`Ac34$Sd<wQ(eUbGL>q52`U8nSfZ{QY9
zVAwkp60lT!RVrn7KCe<%rz5BHt055Cp$*YtGG1<IVH4u<P&i1^3Ohe3P;$`O0I95$
z-)l(-t5Ns>dJ1Qx7%*LBlSI7CJBdS6F6gtQ5a{_?nU`-Ls_{$gcA_%xsAsA&FZuRq
zL+Z>PMS^Q>^Wh<&07@Kh)y8+^*T=2Gvl+DKX)$gj1CKZ}amXvoFmcTvUM>J^B5(EP
zs{?9+tm>|t<Q`=R*?0lAWd-i?7wJHp#T1h|%h-9X$|xvN_7saIz^_^IR{&-Zv*(X^
zi6y;ckC}MIM!Di_Eeg%B!lv6@`7I68+}MsiymgeP7EbxLnc}qA4d031$lZ=p&R%bG
z!pl1oCW+Ok)*Bn8aP%gSj5y9>1u3ONd%6t_&Ir@2LV{M~+NwfQ9$OQ!_4yv6)<jVI
zVA>q4B7x1#OCNO7_l+L!Fh;U)M^0hbJ3ivoc!(d3+LR$0)a&D!WvXeW{%2tY<4(4j
zi-%d6R?Yh>_5)#&HVQD}ZW}!NN)@Pgu^-<sr*CK8o%>-G`EJ9=YGSVwVHF;pMaY=7
zoM+XvDG2|5giQ;v%U#myh;4jvZT7)K((3fRkFbiT025LMN#CF7U)!;*LRG&M3E1|)
z^e?wI15LGaEt0$REQX478YH#3!+xeQL=pcoL?pB?(mW~0DEL}>_#kv58a()KfsOgU
z1U5FN{|)pos3B{&{~si++DR*S(C$yO7k?<IJu-$FEzC9>Pnz`1rx`KC%bx4Fe7g8}
zdhCSXMD^A_6ZWPAIWK6NUTp2J?yhd@{CM!}a_&%}lk1+~b9slTjcl5>cz*Rz=`Gij
zEfQ6;W6p7rjU3_(l-pe!?!I=t<3Hl`K;AtikcJ5lJ(7LQhl|`x>PErFgXivn9x`qd
z1&!LO;+o5ANgPOaH;vz)mnm6O#OCPMA2@c8HOnoVdL2AO6<s3olZ4x@b%8&r_7(S3
z<oZ8u&F;J3o0>iDicA3SeeJKevnup@PJqB<prut>BrqVEPuXadmNdi(a4Ke%ybxCy
zLH3)>pF~Ailaip94au8)SS*suhsy)&5x}x!e~C`sLpEqE#V_r^<&*Nj(n1?XD|K0|
zJ~~?N8Zl9#-YE3zCs_sA64h`~?7L?LDba-0Ikc`38~|<{Z>}munc|UpDxZ9iJ6p$C
z;e~pFBZ?%*DO^PQ-(4{+984&3&a`_t+VY%tAvp%U-aA=d!x*=k76g_owFh;PQ0)_9
zo8(3@hLZu+gBajhz4O>5)I}`%bUCzQ*=B`m;Ay_S4D1jcgT67NCSd&BoFYgC=#axX
zRu@er_Ou`_6Z*5=&G)Jrk2$0GTk@P%6Y8ATMmD0>rDVgb_UR_t$I!}{^mUw{>teTo
zzxfimwuPUd3>Q@qqeTZ(Srjkhi8H&|Feb<8v23#+h~<i)^mM}onVf9gfu<|$l~@^C
ziPgVUspb(~nbI1*j!xz?Hj#3NqI@{bl&!cUV&ed{f9g|!AheqdiY|F;VWvXH|MlR(
zRd3h1w^NDjIoCjX$d2I7NL+syTjm7Tki@^;kh~rk9qTysQtKLPq%jsk&mo*l^P#1x
zaCedWkNbMv!51u1GD1p(r6u<&OM0xqBru*&GQ<<m3MV{vWbQj#lEMb9{N~luft6;4
z9<&uf{r&OB6>-Vu_1Q3}^sv#;i~Di^O37bA%LJn4w;OV}jgBF|q0SC@!R;7$l7Y&s
znR9M{AFN4_wL`w45fIKEvPI@r+~oGBON~1Oz#&PfsyWId)5e%}E$qv&5!)#J<0No-
z7uyn9O}do^#AaeKY5DKsD{~5g`>NC4)xLYSi!S7cu}yjjCJT%EY7KKOE3mVZ#<&X)
zX>;gzStmm!z{E!UH^O*%KT7(2Njct9O@zVfa;@`kK#g>Tl4gctlF6{Oh|&rnyW8hH
zg@p`_?|wN@S8Z-3X^~sTrAtN3fN_=58N&VvCOBved4Uz<N)7ek!S5Hw<gv+2SLWG4
zVxV%32EOnR)adNe|1tJX!I?!{*lv=J)v;~cwr$(Ct?t;iZ95&?w(WeePWCxBr~ZFm
z>~%42X3eT~v1*QYjWNgb+Gu;~LJ1}>WfGOe7+?u5MTKq}tx^;DBMQziBe2lr(?t>l
z59u*zPL@)wT^Ex-;1!M}J<du#`PRmBCCv$>mb|TyKLMEXS~tf0-K{*&^&w}iSGaCH
zC5s6%Ng~4-Wl!l3^LW~v#*4oCUs^`exI8Hx8Anqi-(b2`34<(79~!=d_R<1WO&$QL
z@j{U!PUohdrUGGF>VH!8LFB@nV2j=kq`#>QQ^;a(@TU-BtS#^J_tG=E(umUXG&8y)
z49Lj(zzkRCkCI5pi_4*H@JI+mo~ROeDaRzk*10+!N!pP`<@UZ+4&!;V9;vDB$M+*5
zx*B-$M=R&BDV;>xAvRxFlst9In5eAT)ep-|JI8~M+#+zg$~1SQD}NR=lE>GlP&qlE
zY;VuKd`01Dtd?J@E0w%-wUX!mvw%?iZMP{GFU<43T8&wC-;t10QMIiiuiCTAQC?bs
zDg8`6dM22e9-*_N5dTgvl2$#|;?jZSCsoKE`eWS%?lletV=1~CH9Nhy|3*ctw{t}B
zfqy~j2=ZSJa#qg&-$Ab8>0nAPZ)BzHVoNVaz{tQrFJ|fF>_WiF!A>t@YG>|ZLBPbo
z^uKY(m$Yo{vDuNnuJ!-LcK_Z>z)28B3k>X9>;|Hx`>UMi9)|d0nn{t;%x4)b{dVyV
zOFERWQlVbCI_4Y4GJ%^gu*W=L?sg>fG){pYt-c%&smb|Ae)MIZf>zAaShS96kS0j^
zQVOOrNlIE}8Po?X%D7=iSXpUV2{VAeG0224K9w!9laC`FmV_=r^jilI)WFjl$;b&7
zud#NhK>|T(Ho0feF~*b=%!GARL(}JJIe@Mh?=Ud4Q749Pg7SloOlVH@ag48{U^5Do
z8Pq!_07wk+1z~ccLFscwswSamo}qOs9!98lg`y174n~x{Obg#D;zXIwZARlE>HOv7
zp)^Xl>rzkhjR)D(tGd_pj9jyk80W6%2EPoaW)D`)slov3^@45~4XF7ija_v+M=pa1
zsl*6f`8<}Q$e;(k&*Jh>-o|H?SdIM4z}BHf0z84<Js5hhA#arGMWYCmGP1HjozR2(
z*(^?4LJo{h&2TC(u`MS*6=ofXp|3&$lrpJDQdE7tj(auBDV0!h5NC{%L&Qc>&k#%$
znkt1iW=tqzK4p&G^9&~BfS4MB&xo$(52lSNzcCt3q;q|}jGPXb2s7%5S=p-$2(*%p
z5Rs7>+U!0ZK!aff5>YOQe%P$AyveXjrw%m0X*eby?#`%s!#PcdRGkgjxv;Vijxmh_
zmA45@`*?~93T1XcAc4YY5DJLWYxQ|BNJ0##grvr3P?F8$DThyWDbj3H=69{!w3R62
z1AB#+?|jRnOTp&HeX0D_%%3YsZE9|N4hyoYh0-NpNt^*R7Bz@lss!O2{l0p(G$(Wa
zUN2tTY50q$GOy43JTGm3tJvpfM%Txu2i54Y?Ug#i`Q=16vT%YV!tKSzV=phYnd5fB
z48Q6baq^i=d2yy1P!4nVs4r!JNE!%P!J`e6<?*khSAe%TseJemsXQ%cC(5Rh0b)!q
zh8lPUe6758`<p3CyTa0gnU|h?B=B+DoB3mfm!ruEy8U_M)2PyN$R~-pix3WnagzuZ
zh|W7gY^6E$0KOVOyM5+)@ta|oT%lP(W6SsIj{cog`p5O_u9hkK181yfUGt%9c)v+g
zyvV5}d2Y<}@?>p}FJ|{7G`FAkgA^aXkHtpiTCD1$-=bd|-;(*1>ElQfUv3QDd1N^N
z91jTS(-D^W8GJQe!o29LIDeaL$}S-P%5_6VocutO6;Jj&(bCnE*G@e63YyG<1lQDg
zK7ktK<19`uTXNlTbvuweRsCSU=M1S=^*!H@Xi8o#^SA)ZtdteeW8A6epFQSC_}stn
zEyYA{1!jG`<Y0ZFi)a6D@hRHyBb!{E*pU(JCXX0@IWA><S!|7;nlJWLZAyAj94z{^
zta9)71%Nm!`m2`iXmh5C(q9l*^I~Y{hlI{;IoUyI3H$1wK71&WenMsKKBIcr2%iFT
z{40hNKe~luS*WjeBdao~g@r~F6O_9_nT-qGsTk%W5ykRhn$qBDhp}Oi%Z}FfALh#_
zZk(>71_+{vNO1PkY?FkQLGf?kHB^J`pi8W@P>|9t1!OT}87-Z@`t+@RBIujqp-_Z_
zt}bP8jZH(snKgLMrtGT%-e>Xm=pQOE`2r89{+d^_Zp6ODtktKPkDvSC<k5A7=WK;?
z?t+P|HJ-y^mzd`5ET0xsy{)gvR+YSOgI|$0AJJRb=;{=hZW_XGCqeo&1#~jt!};Mq
zBTx|<7wG9XKt+h?*B5W|r7`rrmp2pGNQj$o;kOe(BMVBm(uA9W)L!a@D1~lHL-Z90
z{gN5;?i!&qaJp)Q+Sp#~V4e`c^qCKe@2^`$cHb*9^lTa(q>9cG-k@KCe1Vr6Mm;=b
zL<+|?XZC-!x`In>Cy>@Ge{lq!o<?QiR3(p>-97@XFPl4sE6+SD)T4%)5_nIVQa;X-
zd=N=_TdW_t0A2YE3LmQRt;+N*OKiZ~&<x70L;5MKw|RyT&45E3g@K)whCpwOFVL&h
zN#YrhT^Vl@9;+2x$@}_sOn(TUI(t-E(lVr0rI-u-4v3wy$(|71j_9b}JwS_MTk72C
zr|5u>D!_C~5P$a<aW-c6D8O$z*F%WXh2#>L<+F=eUzXggem26}Ct-^PihgX<jBds)
z2<{XmSRg=u8KLs@e%)KPY#v}7+>#1_PzA~Av1Dvov*qS6k6%#FPP{}oAz2}x`gwyi
zg~_}`K%0OzcQanm*W}h>l2>ub(A{h3M^frG%zfg1fsj;}bSTO{+I!i+K{V`J++}eX
z=G3FP7BjarE$)ttZtnD~UhC4r#>lk{Uyx~=>vFh;vjTLD=YM`Pe@w2`ZaCouh_JN;
zjj+H4gfRbv7er=2vXYSE)}2cLn{EZLOPoxt$>_5+z7^~u$aD?c_4+$4v!EZr%^#J`
z;jd3@Xixbzs9rDI1~xpAJ<Za<@&PaxL<@x*et%6{J}i95Ht(h=7%v27bY0>Dv37zq
z;~Y!IM(acGQFjrU&)=7Vqa=aFPm9~qPUJEl`YN0R1{vS1d7(6Kn;P?!exT-9VR-sy
z9N{wv1Sx|ScIh;Vn@z`>gt#&gC8f3@t!$gv7Ht!Un5`hPk>Y0yH-W1_{>7WFl-F3@
zd<ZdBHN*sFB>Z50oCT!2{;mI0LOeX|*NEQUdU4qdL4)lnS(|FUAy#}PR=h<-_5Az|
z@|%dH`M<~t_8%np|LN|=$i~9>KM^W^3ZKrKY_R_+e9m!(5DEmvE<Z+HU7D4&n)yiM
z|B*{nhMxk=e!T}FQHIB<tv8#vd(GCQg7EV|1z=npOBZJT{`04Fpfj;PT0OAO_yN!k
z?3e$F%tFs)#S`EJo>5vmRjOS*clXd2yR?qs$McQ*l6k$21t|mk0keBu-5Pf)|G{(b
zeR{g?%WE4c@IlbqZoiY}ky*~dA48da<KPLGhRw72c@H=cG)FD!n${jtLiJ4aNjYk!
z`PJT^dI}%kODk<c{5|Xj=i@cvLs{K;nBti*(}@g65+Y*i+nRtN))sf>)9a)^oBpJs
zCeD*ccOwU%SymuyKEfx!G>)Xkwe|5xW}p8`s|a@cn3B7e=ZQXfbFYPITy&SP7oht5
zWEXD^N%i9OJaxhznISqrsP27nd~>tY1$JWhrMWnJT81Fh{V2W7gg*gZB~*?xV$u3X
z2pL86W>V*a`vRrjQtCtb3FjCRic?C-WOe!Ata>b7P#89uV%I`5q4^|I@E?Kg{)QN#
z$zTXzRa_HC4O#VVW1e6SA%8jg@QR3YE)=i4c-jk2adVu1_*70<e50q>$7++~4f%Un
z=~=U8ax2H{P}SOXZLpyGib-DF^N)k}p<Q#GbEOK&k}>XU?e&%Lc4;H*n%2Rn#(kVv
zD2GV~%(S!1rLq-8Uc<kmvYGq~&LUL+nF8)r9*Ed>foha4Ru7|TB)_{rlQU~7K3Z&H
zU|HG+)xS6tFLfC33G$VI2o!6&URoG4*3!lcniTjdG_(OX82UC9w)qu^yNef>7iT3I
z84*!7mR^6}BJ>ng1{*Sovw6XXtJvd*J#wf_m_)UVXBYgUOnZH-`g_KeoifhxRYR3>
z;J%DR+rJ#+HkZox?~1PD<hc12e{bY~%@Eg8=tT9SSH<|<P^)<#OW|eLNX%a?hHBOL
zVytQp+RPxJO%@-LGaeXLTqJGDoNhsK6{fWM87G?T=2g9&$}(N@=v&COr<i%tz>x}5
ztefxeKr&Z!=Bkb^nNwh>_CaYi!;h;*G?y($)2DH2pQwo_E(#7I)K*WhM3nHz^#Y!;
z5jbrf$m0ImEeBb&Av17<gXmU;7A&8S*swB9-Ofp=R7c4RhqK@&fiX)A#^QA?EdKy6
ztBZO(CI)4WL}+=A-C>fqXJ?o0pUO>Z=Ymljh??h|(WQKQER4%9Myj`C@Uj}_WM?h$
zUF1K8<!85G-<~#&+V$NL^f_B0YbaZ@E@sy)JZbKhp9>OskYA_KgGBU3nF}vv)+~a-
zCdPxnr1j#^Xy$zM4|}TS<JDt+^=qlfg)BElYQvE^?Z)MMa&rjk2wh>-a4x~rls>37
zzu5fFUi$}Kkh(c@sA9{>-`6#boaz8`#Cmy@hy`b{M#=h!RtwpK-e!v4?cEFcMIDpA
z<h_r%IIg`p!OXgj3FSSrxTwuT{Gzv;t_!&4C4Sj3ov~TQ23}|u20A<7+1Q-9DssB+
z`mv^1^~_d&eQhqR4feTt%)e=s(l`AKWgVbE4i;IP{bt0{-?CNZz1XIF_Rf<&p?piU
zV}1GJO>@kp@+ua3kZD@Xlsn=MmN0(hT(8-Cq2S7EpLkfm9OB@uV6~uDuZ`B9S0Fql
zJLtcQWK`5FvL2>KO`RB~Ux7>+p-(nQjn=24>(yiNYZt5I@hb=QK7I)ukP+pp`QCdO
zx^0DNJ9rT#xaROf9Hl-$7yOhuKnGlEUFp3>khIz1^g=eF<M%@PaYyNs>_6Q_is<0=
zj#7Ap-@)9MLoR>cjL=IIfjpPfGy4fzbCw*yMC;FD#Kfp)^c1P3Wk0T$7aw;V07MCT
z`vIa_4jTZlcK7+YOP72U3u<&J!R|dF%GVlwis!jb9SYL#BbcME?RjPd*A2*<1I)md
zeN4gbf6NIt9p+DSogQNMrJ^&F7rb{-vCtoS#L|H+$LSefk>*w;A+9r*1@1SP`Oo+1
zYc=k8{g8Rdgp+bqCtuPPT|6Ocf~Jw<8{u~X{Jn`J3Ja_0ES(!CBD+CCx7a5lAg02F
z8Pj}97%Q)@nv=|Cv=0l}R(ee^`#ldt7!MMwos&^O`)zC)d;X6p=FB+iuo#6*CUhAA
z4i6XE)<y}WIxOjeQzP@LJU!ZYlLY#9X`iS{m@0!SU-c@UG|E*Oy|MCt{DI>-0(1Xc
z+G75{hnyMz)5QCKpP#>B8y8b2dNCVA7gG^aV|x=*K0at?7bjCgTWF8<|D>(|^y}W$
zZ!PdajO#t7+j8nr%bB}b547inGub$gX0z@e-hn|Wog@`Y9#0PiBtZjpcQ+vTXo@py
zqwV~<J^k9-cj|3}B4UCp^?F!BAh2?a=1z5y|0bDR>GiPU<2SM4tuWJ7vNG2M>>%!3
z-{C5Y{O-q{JLoi^eBRkx-{oxn9B=EfkKc8`<ng4E=ipt`!%7vG$Jc7NAhsx+I>4vz
z-8$_6txlNXe@N}x=-K@6CPqK(Qe7TCerV17=wqUcDIs(kf1UODE`4=+=tS9|UcgxJ
z<Jj`~{-Lt`MqMZhEPvW=M=U7<ZnEu?k9a6VQ-0U8hOnWjYv>y{Dh?B??Aj5{NVZ_D
zIw%qmsK8}t?wddxt7b8M;OUHI`c#A|*^z$Ii)UAAm@D*r><(6~OT3&)^vna^F07YY
zU*^R91E9#cOonp=%MqXz;BgezRGj0odcB379MOV9yHKi}YVS6?IG5Q6J}i(jBcd%l
zJf7X)?Rnr5PZ9pTUDHy@SX3pigs&MmgB30Z##|55%`a%A1eH)E!XBAtSP7dMj(!SH
zn^>((yHDY*FFNJe_)Wo#)eKvh|G-kUq_{o<nYyh_?%(N^%i2Dx8L51A<ZNxda!G2Z
zor?>|U=vQMKmD_(Y_d`{KsjH>r)kp3r;DO)rChaC3bRrb@Fvoc8SnQ4FaHZiwN4Gu
zJDU;Qo3^Yn;BMx&l$E(nXqG4dJAQ9Mi!K9V8vx}?1T_kjC~@)VRjs%Sh716h7;BO)
z5mGesHe{$G{X=Bw5|ql?`H*24Plw{OZ-y%hpp$jMm++gT)KT8ZN;=-{kT|;*@GI*n
zZPj%`Mcgjl{8|}Aa^5`mJK0A<bhS}w@|_+pY8V9nZNyuutM9UP)2F+8h*yFm#Y|Z~
zT1Ug^q}tiPf$x45*L6fd#}Olx!@4;E3JW~A$h4<@9u~q$n>A+AvZpXX?+Urh#6D<p
zX32AioG`pK@01oy1YMaDj1eVraFo`$tg7Z@3GP{+w)Q<*Y-=*amlH1y6K;MuFLaMc
zN9&~@E*<Pak$F14Q9kQMl!`V_H(eas!|Y&5dFMh$OG@W#dG>?rnomK_xs{(Ot_~ju
z^Co1nr6iE;q>i_glY2#}!TA)rNFQyA^KxXO1>h1cAf2@)T9&)JJD8vSZNRu&c2!TH
zdBh*3;iJ-7aLr8;=FkEQYLrotq(@r9coIq0msUZ%r}W4nuCNgt`|L47s~CCn<uSdv
zYEbuF_VyyJolr@O9&y(X9aUXS_@3lg;nXdrRrk;g9t`FC-B{~<=#|d(oL1Sgmp}d1
ze!Moj<g}%{Lca4jvryM8a;KcK_Ux^ig<txJ_Z8>Sl(NCkeSUf3QAb$>tFlFbt4(8%
zk3QzccOsfqFDB(j)9~h#W{jgHCd_APoQg^sWIE}wsP8U4%tA?&+`wYSDCpKw2dU3j
zOCoC=nv1??T$oJQ2;O_y5<Y<Sk(RxWqQ0L>%n;A8xD)wOPb`Ym-No8O$KU6;yfu-%
zg`9l8SqOC2-_Z1?c)AFL5NPCeHN)EGezhzLs9-THP~5&=5b*%b_rP!{#97)<Kl+<>
z5q&VX*BL4EY;kJXTiY@I!y)cpH^p5|78ovqzlw<p8c<Fuv}tWAO_9Bi$H8dc-(5bB
z(bLRw9xCgnqpCb*?xc{v#HE;7v<1h2Tko7)VO(p`cZ@nxo*VQX*KJmqsd{)h;mefI
zrq^bIGflR^TRFL0>8t=Nc=s}}X+39}GmIvu%NznXVdN*Snyk<l7c`;tS1>qY+zZSj
z7Da`A?MCP6*GjgEuc$?wDs8zGfe&{sGl^R}hLKJZDt~~5n7bd&qg})Rmhv{y+E#i1
z4pB2Os3R7sn?M9N^Yk8^E059u4m5P2XK!$-Jpuwv5yUSH$n!cr&p<cIaTH1=tql;_
zKcgb0Hx}rc;?&3g`EF@Ep~l3e^a`GnrSw2MM4#4cV?&pbrS*i?hkn8irj(x2{uwDf
zrDOXcY4>hZuq7E9ivJh-&>`|{yMisXGY4|bV?o-gacs|1q*4CkA_Km(w`iI@okBZl
zqP*$*m65Zbw&k!OO_pOMGLolnceIUAP}<~G2>Qw)hzFT;_))GMk8-q@eYr_ep@+)t
z8ThZyJN1~V{Bh>ws6RyXdh~cy+)6Zf`R`9scM=EJG~ewS8iU$aW5mB~(7$aNFW<r}
zuD{F}=U=}1{Am)9==A2#&8~swX!U#!DDLpXKnLmmTjpc=FLvC_tW5t~=EIt>!)d?c
z)1MQE3A8}d^Z|}Q7)yXXJPx&!gvU0n#<?Wu<RlqgDUGNlTjS;B#nbGAwwPkcd^F+h
z6qBP*%;Lkl&i#Fp%OWS|@k>s2MxQJXBH=5zNYt7Pf}kqw>~#Z>icOhZe33(!q-5QJ
z<NFM-w5w(HSO<22aPYoD@c2g7#XsdN@<pt<j>dr1jmX&?BInvX)J>q1=B}3ZU_4lE
z_4odI5Hv$9f=ubXAMtp8@p*Y<fHrAHk*YX*X#K@604L<lYKVXb0Bq1Nqz;rKS{IuF
zSiaw{OtIAb#;G0!sFIhl7Rpgb&M?WXgJ(Y|$`|9f9*vO@j;@-Mt2qAds37Wy|9gmd
zjqCMXxx`u|3b%MOH!*6~CRE$u2WuZA=L%d$<+}@{SO6x#rJRo^x+^xwv%<|M=LJEI
zQ0YvTc>L=Y_hiMaq)Nh>Fo(&6C@g03!i|~D82;;+q^2__F=C0iC6p$?MGlfkjl?GH
zC_!HI`%DE%F0iWUo?dFH`h#Ubnd<NEhs5tmD)%3kS1no=+L_Wwskjl^(}wr4y{PFv
zE7&#YoVa6}C^Rt5`+Xj~mB)i52CIDhPu#aV6MwIK{MS0U6$nGMsAHc7K^XKD=)oJc
zyc8*mb_Ie_RY8_cCK?orQA?Aq29eT(bs`Q6>Nr?mSfdg+BbO?3@cOO^o1vF3sX5=M
zD=`Se>XX&43lVI7IY)NF*?*a5{K)jk2(~SZbe5oyG`D-7y?Q+#XAd(bqze<hm_bUx
z3gAJxJ;mD6EGSB|ZyDFPp{%7KU@ZdS5ea2T^NtGC4z7C-E&@jskXSoaen(3Lv&5P&
zus9F#2ho_{T}qq=A+F7F@9T7EC1%iy1>;~lj47}at)TjOFDbQ8JvO9e{xW~)7xt8w
z%w_@T!re{Q<Dh}Nqpw(@%cRh$8CMk6#2jN%8v3@N)sa#f_KJxXdK^pyax9pxs4Z12
zAgQGNstifR_nbBOihTsuN|Fsy&^f?l!cl>SnITonMeDEcC-#`vp?Q34xll8#6tuIe
zXX~KyCzY7$Rm@TPv|IO24V@xWo!Yi`O3R0rW15X7@?DYe>J+u@s{u(GFD$GyWSX_+
zhWuA*2&!qrmx{9(TUbLF2vSoKoJtF5m|27qWf6OCD8ecNbaE+imm1LGqgj7r<`?ul
zaK*~XTxx9K@WpA&+S(Ldhf4NIQrP944TW;jo)36aX}0{5pBw|x+<hoWk#8E<nZ;r{
zq2Q-LBIPR)mDtm;?Alt`=i=N>qoIAM-Y?U=7S<Kl|2RLQ9U10aPtU4L4COLDT9%S8
z>-fOSVxCKnwFiWLsHBjm6Kexfw`}UfD=8P>^`?(SeBP!rN_mFgoLd9#6l9H5hRmqT
zLvD<MrVMC2wEh(^tu0SKo*APBxNn*+Jhb!mSl<C$3LKt~7|>W4=T}*fAcaChvE(qr
znTQnhu=CCxjW3~~(brt=w+5=Ro5-zVV-W4kkRzDd)vh)RGF*14gvyLuz{HDr7W$Qp
zJj*%f5@bCm#+Ve4Go-4U$z{hvbvBe0aHkIKmO0fZ$1w|SWf>*aqS4R=jh1Mf6=*Y>
z*YGZ>00j?Bn;SH7Nok-pyjA;a^$WE*)#f*WIip%e8OnV>AyGz3<~!@nwmqoJPHHs4
zh}Y0Tt+pGNF|UlO&Xz1ZzW>orpLAa$@XO3z8oN0mYgeLT#)se$v!~%-t!k9zC9`(}
zXM7Db6LhD9s{TJU8p)H&J4ah%ln(7p!>jREFgaxA`yA1mt!wK~_A7FFsvR5S8MBrX
zq>1~Q(vp_D;$updb#id$xFPWIqg(!mb&*J!e3F^LX56mvpdNKnKevo_dZ^;iYDLfG
z0xM<pk#75Wv?d<Z-r62}Bsb2YY^P}y`6$>CB6Fv?geW7S=<D!&#^a`-$4twnNN2zC
z7~JEfdvME^8JkQsS(Od#A#je35)e1(OAo*2-bz|$HR#X<Hxy>5eXT~>$!N2U!2+*V
zvldtVYNiA#r!ADi)deQJ$-Tsaje(Y1Jtg~tvcmivk`-Z9kCSCV3ak7x!7mTj>MU;v
zbrpK;jtQ12ED+hi#Ee6{fBNAYm9%oKOJfGEzaO1}>l2G>WW5)HH3Bz}si<*E)KPBi
zTFVJYg@SpLzwIK0*Zu`z4k@<Dj#5zx<o|hlU&{Gg^@|j*jNY@|C~?n>0t%lKx9v4_
zwtTb-x48+SV{}t2#w4Qh$X2qhrdmk|0wUNk3~ZSZChbS5?)T@oK%Nv2XhCQ303nCu
z{0{i@yJU2$<H&6j*o8>p*Md=H$?%P5Dma`N4ESy;tEJsNJex0S6-tEGdCV|A8yqrD
z;I5~QqT27(hhjEgaGsUS8uIy<Nkq#slzPKB@D}3cqO!eV<bjrM9wdxn^4hSR<?ud8
z6g#hrw1a)JM3i+9x;C27vvy+&NW!obLs+m=u>CR4#j`y*8okyNZXC~FyO<Bz%8i-8
z2T~n8&tO~{(omw&X-jVxb`BI(!H%tZnSjkw^<4!sp46F%CK19NA3ZjOBhR%MF6(km
z`7Hp*#_y~|;71I@BKLm=B9z2gT3H}GIsuGR_hL~Sja_rlhoZ)4cm!kZCntB%qpvu1
zzH3P^D=u_gfGw}7mDtxc#ditfS4ipiHh})=Ee>Z?UjLtS#U(YmfysBH%b1C2;>NNS
zyhv);B=1M;&R2m)A$I(_D~E?aP9yZqDhtQDMw{!(d(5Sge^&UZ$DbCldml9)RrJru
zVSg?2nm=vv>tm3{!SF`&^I#BgOr{;9njAj$)Tc0eI4xo*F|sSP0ZQR42l<JHTnQgA
z{8w%1qps!=lxrBj_w^*VOclO7F4=`1JbXF*^x>DQHyKs<msA7c`dM99O?6cxZPnGm
zQi#Olh`Bo}VupKa&T(3l)R*p>Yb($HUS3ylmFem(IT)F{sNz=Z>Z(G-cT%-iR${e%
zY}bq%;Vi(*qF%rcWCB!d*02E7Pb$BL)i2AzC&|4k44<(dRe&VMZmN-A1Rnd1toxn@
zyek5?sR1g)!xP>WhzD-+tXls}J<*HMPB(9QN};7NZu!fK?|JCqI<+d69JqXxn!fl^
zOm)xyoBH|@TK`wMoQ0F^e^Xx?|KB~~pX*d<qV+oP<}V=gU=7xt0`_oEDNiaa|EW7f
ze7(WrHw?uSX;W7R;~oByn1=1$<9Ru|mM-nocax{@{6nmCqFT2r`+VR23vc!VDnp-Z
z!xQKTp(C|+tMpk5dt?TgzlfEHKgS>9t0ndsGSD38`<F-8GpOZPe6Bg(A13-*?x`sx
z8vZbGzC*;Qn+uw$*n{+G^mY}~h5wWtu3ff*)aHtF&!oS&yDq=B$KW{sjw87F`)w5i
z3n^g8oHX4>^|khP%yjM)Ehf@UN&lE+Z?)2Adpy(&VX+EVWH+`1qOcsuWIHPwVo}A-
z>se1=6Ng1$Q0+%!A!NGbumZWoxtI=3VXJ*Oz%rC4d(df>FERA6q1i6yuqIZggbW8I
zvf!(*5L;%%g3|xc<t6#5>68F>ciU$!2$wzI<_S*~bNBVLGopzNtJ(bbc$w_@5NZI4
z^}IM;mFhb0z;#o1T>vVVDBtF~diC-lwG-G@NSbV~?<010e3Re0`9JFZIXyEUwGTn4
zk}u+$LQm$~Ka6>!L-F7lO(PhUU6^YN-uB^oB&M@bxAVlcS!!|0@$ylNzGY=$L6XDA
zmJV#_t0=1zSq!1z65$Gwi46i)B1Z?1%rEf$j1;oO{v6$O>#O!?`)i$R=Ds56lkXP%
zC?B^9NbF~RIedU#xdS%_ybnt=trLJwL#;i24BKQf!^DYyT;(i(keupq(h~-{QDdBs
zl6B*_wCfy~T*wosScM~svaU3flEb{RTtV~v8deWwpVPW5#FYw!^2^bjf_Z-zTwtUs
z##$b~j)tiUNkS!uB|z;NIN*^HtdNO#Llaq64xQIYB{AO8%|*3?%D9|HB`N2k0@+TJ
zD{GjV@0SjC(E?rBGu1P4c}E`Jw5@LrcXH5LD6t7=*;M&#ksKNBj!ib51zg1Oz+5`r
zJt1q%I^bG$u_pfMI=8bSWFSd7ab#8vG%^BR;V_NG6c_;WQ3P*iBGacfBJbX4=BXz%
zF?pXdqhy_Wmb@xS3XEDe7M-xuzL2>CSIUb%WakA`65Br47g01)Hh6*qD+X}{q^0{k
zonapEVC~Jd{psUw-_>hE;+`EZrnbD$)>bNdk=#n)t+lHZ(t2iD4HW@|29QkkA|Q0!
zET^fN+|2@c>W-tSxPSeN{A^2LexNdv#0}=QlfT~7b5k_hoH8?Q`!1L7dW0+ZV;&I`
z#~q~Zo-4F-TF@q9bf!=|_z4}J#mA8<74FvIU+G)T19tJo4qa5$YQ9x0@Z^`b{(gMp
zlv(EYZ@n%KJLPTuXz%isV&%PyE#=EOWG}&2V<=l&JSP$)%ih;Ew_U4)IpQXn_{>e{
z4cu`${=^>dIC{QWaU1eoPz1=);}ytW0#t#nxy>*T8A*|z-eS`aINC<raSGIq1V@;L
z#OxYwp^>NyItHUXd(lPl-Yb?6{T5?0+F06R>u(35QtjV{=uy~f`6&x!>vmZ22wc8|
zjsiXJ!YW4MH28(3W+oP+Cu)#c(4Oz8?#5hFQ?VYMDzs+BvQ|(8<;gxsfi$~lM4}yP
zF`(nguP&-sV=k;=|I((FM&Cdotf=1&(!ECfvMfiMO~7j*B~wEeX+il-m1!U>a`}<c
zX+dHmt}_3S8ljbES6y+HXDQk<C@s`<hfpA6X$_#Ubk!NbrXt2sPJcES9$BY@>}H0-
z?LjbsmC@$r*BKI%lGu%+K#rlOfiS*{$p}{Xp6(6X#>{ry!xI*Db1D-6Z2M3o!Y|T6
zC>?Vs<4O&1l?`wXK#%=rP;Rdh8tHlvl?@oEB3y_|`U2FK-?c|PD_w1~TIc-`G!2>#
z?s2!{XaWSr?KSqWJgNT%iA{sE2ol>1CjjwK$vgZ?0D1|+hTZ0<2w9Gn03m!_CJ9vN
zMU*5ct|5mak$E5qO#Ado5U9u|hZ?+eYW8y-mOi25&~mQ^?*?f34!LUiJrp1DKWx*C
z{?{zQ{(of&CiedksWPPXgWzn3>v^s%EA+L{T-5FiL+tem*fiT!CUe$+H9BEHHbDj*
zoB92ir>Cl`Ym2S->*`u5yyyktwO+-?EBysv>Hgf*$tA(6dCa20$jd7qz|Qdn5fdyX
zj~0LlwXiUAcENhmzk34DAK^@SXMEQ>VDH<Qy6XVGi`qA}O$GVz`=x#4J?>bvgNfZ~
z_65=7gl#+ii)&{8g^^=nbx_@O`|WsRA=jO!(!q7s&Pm4G`f&a&S`}7Z@*=g(FH}c?
zLcNU+a)+bD5e$pIN%r8bb>d);yGkU~)gqSJfmW+DeXkX|_jh_@_=UI6dbISM7(?Zc
zRkVAreA}12vWn42IzUeTJA0quSAKFX2Q$#}Zfoib`zH*!-si*7tNf1tuP~Euc;(lk
zwAtM)0JE;wOj5I$4F5c<gMpldJ(&TXt2zf!I_kCnVKwniX*N<y;9?E#7h+g^y|Pdf
z04dFdO^%e(wW^AQfwqw3L7ZDcx>BgX%#?JG%3-W&59utNN=HmJ(>fzU#Xr|=X5reo
zh1RPtj53CNczbsz!8VYx4oT~MJ}wcK22#@P4o!G()S}WJ6J?U4ki_aVZKy6qIbN+g
zR~)O`64Db90;}N0@#F8$tgT~_NX2k8XyQm53mqtx$psnA%Fd3zSqFO2+oBISpLY|e
z^16(7;T!{k=a5NorYq&z;TcP2g6#I+@juZRt*DX(S37o%flv`2Jgm}E9aoi3xvkhC
zp1|0bf|VH?g4!-Ip+<j*B~w-Urt)1RI|#=rI<wM*h(K_W6lF}=-{Z`Tp$ivLl{M3k
zEXd|EF3bj73g@QZ#6V->VdTrv21eWKXb-mIxk?H{=jc_P3BeXHf(pra$Y-h)B|fxD
ztnb~rPMm_K#?(l4HwNVsNX+nUFDWRK2uMx-5*-%aM8>3ZJHkIfm2*oRL1d8P&1Y+b
znB=>%8wA^4-&crgnKm0Wad=W!KcDX!jIBrk{jyD^L7CX|EzyfXA$3Ifkih+@xN-Je
zunavHppJY~v!{+g?ua10^kbVzEb)YZuZK=+VSHYIu6Xh`e9ZL4;|bsSEg3t=sE~>o
zJ4Sx@mO?T6%mWg&fo)rB$$H)~>>dS7^}(Y(dE{d}aoOees9_k`L3@GmWQun*q^F7G
z2y<sF^|3jaY2uW)reTygh+6Uthc;&_r$zwgia7<GTdUGzudeDaF)eyppbA9~9olv-
ztT$&nqfX|IlQEZ7B;?R9EJumKNM8@~$7EaAQ0m*a1;G}V`EMkHe;9TdGBS=m_Mo`N
zzi&3)8@^Z_=YqZuoP?&B^o-KQrD&fqfmP>C%T0``IbAt68A=g*IYU$m$C`V(i-$fS
z*@Nj^=fyCnLly@gW4ldk_$fSgoyo>A#9z3z#j<N!RmCdk%9NiYi!<YG?)rf@VPV45
zq3P8?Zutx<%~saS)Onq=ZzL|Inx^cf)T(~ADa9=l^Rl?pYt*Jc-cNh*me?m~2j>R1
z0c2tX*S0j2)q)nbT+(_z-7+LUXR#y?8u3!%N84s_?+fZTw)b7`glfE=K^H7whi-)`
z4S?nj9Q8#9C3$3mQ7ySdxy(N26;grWNj<~3X4KW}rk=?~FS=cm`7Xd+y-zp@p_@?j
zSG4M2wnm&~6zZ-=I$#6`7|C0cmMaJSn|S9;6e*Mw+(e#;j`80{<cVr+px|vEgj5gR
zWXS5dYmbMZp->V6H_@cw-_bxPUaEhVllKE}(|@HdeHq4n19sFzjT8<0pHO;DB8a<T
zM1c+1Qb}#wRgR4j4w>Rxzbb}LrDtR(E|Z??(<DNAX@}TOv|}XKTtmox>{%P<-BC7l
zZUWbGuBofMx4mufy%O}*O%1YkcMbj=UR)UvcPdc^I+8=L{*9#q4I(s~d)Dp)-;0F)
z;huu70@)NZ#8OR`BB=ml|87>yHzG2?URB7#+;saT4pLr>%ViRf*UxpPuj9u)-{e7v
z3e;i0OY>rs39cqe!Z?}I=(-DZjg1HWVXU`cnV8d5I!q?o!F(C6);NmSzknhXy)~K&
z$97W;?TLg&qdiSHnc)bdHZQ%*jexS~Uj3<dcf<H@7-%;BlK!b=-$42mq9P-~=!B6w
zVE0R1@S^cR#Q;^Hj;DrzPhJd)AQhiFWHfNC<Ms7j#3+~s@iglGvBRfIU~`Iy{Srpb
zKyRY_E7JN@Y`t6RU4XNM0aY2T4co|%vN+qTKPH>4(dZHQeYRU67Y}Q#5w9Zm9M{C7
zja+4gTf?u)T|+aut`2NqM6>xF;v79;Qg#vXafkIc@HJH_Db2>1)=~iezt(cpQp`$R
zi3BG<h=L-+;rD5L^17S^jFVviiMau9as3F+*$Ov8F)j0YtYRV;v={W?2w&k#$%VKY
zPs=Lv50!rYrMT3rT)!-50?&DCEA;wM#oVwTIOLopw?Bo0Rp2Sc1iNjJ>L%ojQa(@m
z5(2lCcZ|$~QlP+DSH2z37KA}yC~pmOj;!BuTH@PQwO0Z1SJREe7qn75L9;|L3CxHE
z?vnnp>q0&VT+G*ztAvKrRMZ=BeBI)z%jvU7|F04EpKyXK;gR+!D4_wpnZE4DfP`1n
zg9!C|lMe+d_K6!G>Nk#=#EXTZKkggcjKb!m*{<3gx#98Le62w5CPVd=jGw@c6D}~!
zSW~w5yDg*!2!noX+71{n47I~|D|~dUw+`=ezEsX^&EyGKg>S|U(*7v(@WWb2f^fh)
z)5x}f4>MT~+!*X{6)$x&WF0gvlvF%AZhs-LJx@hK{O&bqu|ZeW;(MuzS;+FAfqpEl
z-&T;ROQVU<!!fWbm@DLIx!2T683;W*yfs-KVtPcx9dh8fHl~~;tXum10LvrV1ev51
zISl)vl7;>X(|rh4vs0bp)7r_by-&qn7nhZqPNNy02van~N`-)zDgw20Y_yoKap#-U
zfwr+!&V6wA<;iU)hu1E>v?mr;ZvklDu5yHv$_jhQ9Eg2oAL@(XRk42CN`Kx)(hjED
z@Ma#NcShapYr}_07Hhilo)JV>#VJ8+&FcB+sm3O;RJc1l`4Qs`gx9~EW4R+j*7-gK
zy4(4A*~Fow*c}aGUbk1;p9OkGoJc}SXpd=YBzTG?W%eCKf6r%B&F+BhR`b)XzU-vK
zSA>?lFgl_#YfOkE=ZQ2k;JY3MDu>r=g|^k0ip4f@rwf`H@Iku%d4>(W;|4zW-Tf!L
zg5Khp`6JxIL$FG{{083+UStQk3EV*2jehdw!EZ9F;~kx-D_K9miCMJ?Pe(s>O;UG+
zol)M5D%1pD#Qca=>>JY}>Ma**|Aa$8Pco|@vQ5Ny?<j%2ZBa>REV9>&Y?!j9jeXsg
zGK}x&QA4L1n5;V~%f}>`1!74Ld$ZDpEneiL_;S0>@0>Ln*6sc6j<8BMis|0i_jy%5
z^ry>!kE0PFBy4mx?LZOc5|cc?L`RNco}N%hhQGL4>dS+17uBN&Wrhk1E3|=}QFMzd
zltML*{nY$8xr}Zq#PjHFl;c5}vq7>jViYyvG7|b#+r##>tj;F>D};u~$Jd@=psh}$
z&#;ct@m8!D;E<<UiI>q#HH4BXUfA%Ao$(ghy7Tp$k4owCzo}j}w*Oo8GO@A$Z?zGu
zu^1ZGOFn%?{#&6ej!YIS%Z?pbB=(%+Fb=}qLEx|NfJBaRnx&-;hQ55SvHh)>t0Wyi
z{lojug}cwQPWI?X<m~|EiFbRL*^nanI2MbkeV~w~x~}i~4**SZ0|4htkg~aCWX@iG
z=lE7MxD-|38wg$>M@Z*eY$p7euvYs0d>!=PY@x+5>D1}%WF-l2ikepH9??L{pWGdP
zkA%s7iCR~RzZF8Ye7FMCZ-$VecZl_c{t`9>z$^+9k`EyZ2*Muw-*M`D+J2MeHOpN)
z!{ms7t;|2Pf>h<80XdM3=~pjsqX;dk`AKLjY(fGwn;1yAZ!+w|J0qjw#jLkgQ=6pI
z70^ibjyo5E#veD09-?)M{dX7ddpj5J3%QM(tkAc1uSc;zQ`uM{PgaU7*pv4wwG;O8
z{4olep-K|#3H(tyUvIvcVm<Z6;FA0CKV2|QDv1P4y5#_Y;tG!*u~2r{gboy!7UEFX
zkiev0i0VR{(FjIr7g4B|)ojFS#hSA=d?k=ub(Gm2o3`5Z?jVnYkx|gem{eS9<aeYy
z@cLCzer7RMEkw+koabUl;u}sX|LJ3qjEFdjUvZ{Ws03~Lh0JDp9gFocbz?wL<A-nC
z9d1xz|3Ep+L*TiFvArfsM3Tv6&%?<QHm6>E5gOA~0e;BcWH&2H{y<*1Q#gy79~*{>
zBddF4%+XqLr7sM)V)Pf{1%=p_r)Ae-@BwP-q`hLqlRJuIP{PJJJjUyoOVg`7!3pMF
z{S9PzV3g|}I2$g=m8z~~muO>iAl#W##)^D}h*<sQAS14-yZA))0s{iORWvQ@TACBf
z6(1?Zs-&W`!i#IrL8;S*)QV*YR$1YQv<e99l3g?yQ4m?-EQ=I|Yr(eORZG4!Wh5*y
z@fI9rZ!S%!9+pdj|D693<!&Hu2Uil!qSCzP(Go+g$BPRGf*%P;Kb(0)GuGyJ;jF!=
zmCrz6u7>UOOZGn}G#AL8nk-5hBy6q?xFoN-q%q-9=dOh03rvTL2ddTs%~ZwlK4h4p
z)0#B_BU7XisFBl^OI3oESj(z!Vhu=l%-&}u22{$?95-B_B9{?uglMiD8uy!GsB57|
zET7YKddY~()HhH5Qj-aDZ(qK;xFnsJsg%>bZ<Q~`U+8Cpi<+m0GK1paq?Pe>kh_f1
z?_&5j<iYvgQcuwZf909Zlebr(#aGgmUR~4pw(sm2YHMOLVNt-i$cl2!gk2g`Cv#`$
zMM;6#1X7;^%xGkNL|31uh6b<MT#2ME98*QH@iU67#-^gEHxg(BpO#_+YRZ!yPesKC
zvC>S3CCr&E)n}}cg^Arr5lC~>g`u{V%rdFLdAJzou0wfaapc0#X-yuP%AwqDkE>dJ
zt?cfiGylAzhwex>S0RKy&i)Gp75LOQlZ$NHYPGy6aED8!PgisfXEcIsgFy-8TnkEd
znmt$#F4s@Ewo|Y$Y_BCH%che*je})V$NbEIGB)>|MnMh3l)reE38<gjS2*9b{4W;4
zbn*1m8j}WfUj<@A!iFij>hkixHZA*Poslfg0c&lCs|Y{>&zvWsgkr9C8@z_`UI%{b
zv_wo^8Pqx}*oCiXUIEu|7$UyAg4*jQ$Q)!T8ExMR1i`V8S4f*~K4nC@ROJ}-q3{x>
zQvD_%!iW;d*iv?pw^XaAg;<0BO(qEYg4B0UXMq$Bs@)^Q!~+IjEP#X;o5@gJKB=eQ
z2Z!s3K96ko*f_xYt%X6b6?>MS;BNpL^rJaMJl+L_YcF_N?)oRzk%#xQb-*ku-sHem
zEwGo<Wg5lV*7XX-TxX{~vDc=keusW+A^lb@V(;3>%vP;pd8~VH9EURAcj^|sv-fkQ
zmgdT5zoUvP)ob5%YKv{&vTZxvT}2dk&7$m>9=1f_FO!WHDe8l<6fFwOfv5+aLW&uu
zBcK^<c6v+EYUu|r&65=`iUr4g&8@6qyXqmetY*G5%In)jui2Ka!+2c1T1%AU<YFPZ
zu1++oiY+Fwg*|@d$iU4i=4L@It;**s6`P^omtoOOE%j<;R;83Ky1LI`s*9P>?V-)v
zt?gASQI$;snvJ3f?L#T=XvHUW3vI@{D$;&=#jXEp?Yv-%BO8~04d`0b1g@d6oxfA!
zS7{Yfz>z65v8njhE5$m!^Fgp=H=vMo44rAcjcfJrk4Nvoa!arVuDJNky&}rLh8!JD
z!FSGp(l*FZDZWjiNW0gD2O{a3izR78-Doz^#bIng8d|Jn{eU6_s1B$1TG8*Q$Gur)
zqkU?1Eo`C-t+%tt8jpJX&T5uxD(q^Gw=Uc0(6y?8my?@m)9%>AGzn^_xo@gIeRk44
zRv_G!6VD>dmTAGuDTG@Bo-s{&U$ug9+-y-KQ(>I~90~<<2K>7PV|@I31oyv}2}Urg
zs}QMPsDoX4xw=6Y_lNz-tVV=rJbX(NT{3Ik68u(h6}<x-DigE*+1I!y1m1kBEe`vS
zONM2xzT+Ox<LW0d*F;yn9BWIJ`Z!y^%x`KeSGLXnD-UON;tE$)>qugiIdi1N>X5ED
zx6*l{I_QolX5-aMop;CjbZh4e-UVH%(&mlTD5y|=-`{#|a_KfX@SeCzX+$26rjV{F
zf9ou|OM%Vs?y;$;@gya<Iiufaw+_}Yb_b`w=&(kC>5K@~tY2DmIV0CmaL6y&qz$^+
z#*hIx-?a8H%?j*pluJ%gYAhAK7sJhMUA?_o8F`OE`w^9kgtd@Zez>4Md{L)J1pjh9
zZo_{zK^Sf~R4!Mr>rVcKJHNsCy}c1@P{n}7f2(8{AaO4Phj*lJ*Bi|&KnFuCL3-0;
zoUaoy{f0=u12j9FYE90`=THw?Z+tNTdiNDc%4BG$Adk?hFW3duP)Z}v5nIIi6nRy2
zzehXetyRl7#cbEDmgq~4)bn(6RQ$bRyZ`PLFA)C=n2o9VOtx5h&fZ4)>O?k^w(@s)
zF4-}e+u>x^6LDFwj^MfDiUsqBtl@*b3+3q4e-lbST$%r~iNwUn`oFn?c589lAF?4_
ze^MvW0UP1{!RY}q4%}*H7w7=J4pI@kuAp}M?j5})y(6^5PtYB`KpnFWd)Rkyd%3mU
z9k7=#eJ-U(4-Ym}3q%}vUjeoVSkT`EC{_LgVxTqj_~!d~>~Eh1WSt657Z;7pey8v)
z?P&%TV+!m5!OwvIo9Uf(E^++((BI)`C!hd;#NrG?Jky(}k&7)!Ul-AlSv|gee!r3D
zu?=J{{p|fJ{Cs-)0_-Sv5}Kk2+lBe}@V^2l^%bGAM9DaFf4sQAyaB{vxRDN;$i5VP
zUhZrywm;HM&N;@CtgBT>m3K)jaGKm?-tsAweU5bIrsGUPNvMmmq08J}kBkh_5IHTy
zoBy8s%<6Q-4aOpdx$(&OgX}C}?|X-+^o0bXDcwfoTecVvy8+~}PqDpzb%c_vIfjTm
zy60()^|isPvi_BVU&sy6=%z$$WJ)$os3IXwz(lUZX%sqBF20EXPvYdUzSb4)V3B@r
z1;im!WK?WRtkcyz-E?ujI67><aF(9o=&|2cV>l#$Vuj5|`4}fLj$Vx_F_RRIjxrHa
z-}G(Fg&INY{OS3%6CiUIYm_M+LBIcNusX{C>&Rrj^S+v9$Ul;}Lvz=gPDhVh8%3++
z(WAwKcqGvsI^u2og`G)s@R00Zc+6e)yk`kNuG1xe4$aKFRPlg9tm?Kb`1ZPVP%vyp
zS!;}39cHJSTj-cX3FpoQFV+Y2S(P%P^d=C7xkiDI>cp26uip<zp2&a2w2s0JRoIdX
ziYr};IK8eR@r<%MV<n^sWw5WMG6;v@wpVK5b@qfS<3u%M6xCU2YJRb>OB(8rfF}^{
zvoNy@XR!ELj8lb+hX*`Y7!o(AYHJ7ck#Ha*^L*8Gb1^MdAPVqfuo3Q7@P65z*O`Jo
zfNp<F5yv`3n2YCTDmB2M$;GyZq`L6$9v`W`l0xVQ?tL-1V*jTlG#B{lT}v7nW(j3q
z3wOIrPs|7<$i1?y^$ZHJ1s6X#+s%v;O6_!uF>-fp<yl6|?3=R0WmzHk+`H%PCXP1!
z^M#BxxK`7sG#9fAah^z9IxRO}K@Z0j(l(Mj_h{nf#Hsf+BnhkVEdj)B2*}d;R4)5y
zR|%C#L=VPHs|@GhhCwGxLgQpOJgvrK7T|iN2xyO3Rv74I;Q1GD7}Ti-xDQVJy$Ugx
zXus*)Zb%bswWWFy@dw>{S$bx?y!<<zivNmojj-xN=iDfT>}AyM9Yz=lzIN81($8QF
z3$yBXo6^d1JJ${@+o7|0UX|=tOVT1|+3VI)rJ~0$5<X2zxTR<*%h{*c5V*i=FE{y3
zkg#>vzt)a0>2U>`gq-2hJ-i5=pv;iCpK$RUq627mu72!2UrUygEA;U@z3C{RwJ;BK
zD#Gc!TXM^yEYl^}Ja;>5y9abPHyJMWTdjwr>fO}K;R9N&+E6l9|JnxD4UBEUU)@}o
zNa8Al)!*}!vp@>DUAljFI3&TR!*5*%!>8y>x}Cb=yL7ZcpK!Uf!gM%fNQS_Fas>B$
zULw0d2omzYURsxhHOiCN`&9FCsAz89zow2~!|gv7i5y#PQud;5D5K_+;@Z9hwnqpU
z{HT?S%98WPW|e$c#YY}h^yHkC%m-Ed4I=$1?S`_Kw$$^f2+#-D@??XUxAt>yV>rSk
zZ3z?^&Bi^dWNUCZBq)P+I|S#Rl$My5-J@28ox4Diw6R~|;2hA>xS2kLBZ&EStCM)a
znnMHN$!|#Mr%AiNZ#`~zilp{X18_$@xxIUOf@{RDt1i(*X&UU|gik~2NV6Y%Bnb;d
zbrhNlFVZbkaqrhqB)M}5mFN3g6z>sp%j50t(!<;?m`{G+rDvS7=9ZiJr_C(#@SZZm
zjmGdT5cx;VEUG-+ri0%_%q%vp_=LUVdu;}7ffMijeu+;F9z^pUj=fF*D?RK_M|0hK
zN^_T>PNSUPo+?QS76J2F!OS%hgClUYA-~W4sz8-~*l3q6N7Lx=sw~Gat<(q9l#(V$
znrox@ZEuExgw2zyldqO$u|2c`^u32<$DTO&Sb10NLdU5^s+#om4b9Y>R=V*J(GhjW
zT<e8gnPW*?%Gzjak*PQLl0Pf{6cyPesGu5sv>T7gvoKLR?#=TMC#pu+1<F6fk)Ri`
z8M~j4x8*cXfJsh8ZBQRvkH5^@)i+^htX+Djp;_UiFOd&{STFhjm$pH(d*>FLpo?$r
zk21zYbvA>Hd?r^jlw!yWXGVliWf@yX$fk~~bWG54ZC@MV@byigMA%!}JOollvOZxZ
zs|L=MQL{f0Wu{x3ovg35)_8GO(xFTWZXKtpfdc1tGW<Wxy>(oc(Y7wCqM(4FG)PM~
zNOwqgBOM|Q(hUM4-6h>1CEeZK-QC>{cVeuy@7-&kv(7np|L*<%P=4R{#(d|TV~l4!
z&v@s&TXm<A!q@VrwvH%$QZ;V$1M_NK!_|ipt1ea3)JKyb+o4m^zLPbgsmC;3hN?0|
z1TS7CD_j&m!f5v6(-k!6MOTQOY27cPL>UsR(!)qA&P!(qu@UW%?@FixRulXf>B6Q;
zX_%s(d5pqPYHKMNuaJr8>X16ssu35OxdlXIg$&MEqp&7H$d>(#3yL`MOGK`U4dkbm
zS3)%;w%H2kwob|xrIhKo)wAx8u4l0@5XO@+$Z_`UG7murs|?v{N7alRQ8krBl1b1Z
zIk~Enpu{$DFvVORlMmLh9fp*k4CV2S8K&6MK@fUfl`<hVU4iy2y4$Qw|E&9^pO$zk
zzN%SCafwdX-WHZ3ukv=;yKrntQJ#dBt)R%~;Znn;dIEKgP>TNFm%@;`Klw97o5#t=
zm^;b#ag4om)nbb3((8T#_W#cIxzfK;D%JmHHjIv@^8A?7y*BWWF9zbdO39i0ZKhUC
z)|>kwUN`<P8}Gv%{RP%qM~-9W`7Gxp63zA{?;%t!pC|n*%)<PicZr~7diU=wF;V+L
z^kE$#@WRc;24>SL?7aBcm)#Z^Bz-rON@Nbo7)=;sx4o>}b1V_300L3yt+S+#PykvE
zVg)~*k9F=1@0}D~f0nMYY<lx~qPE6{8;9WbnJ{(UCgzs{=!d!&>L8hv^zHXikus!J
zn73zEOkR%95=g!zJc2uy;7h9?u+G@0893dBHtBC1@H|<aaFt%9l@_j|kXB|o81*UV
zy16>Kl^{aoRQ2Bn3GoMp%zA6(;K&d6B;tsud^51K44a>Z*mi|t0*hc@jK5FrRKAUK
z!5J2r?$$VtWe`wOyFs~4lN3(qJ$u8XKR8coU?Fit71G5|X4$aOoi*3%sPdzjAaQ<!
z=JR;e;uawmBD{Abvh(*@sDdD!>CSsek9+<CbC{*w;f>8p-Sc(Romb1!PX}h#^Q$j-
z$X|8o=%s&NX*8?1&aX~vQys<@#WV1sIfTExIvWd^yNTmv5t>3U=sY(T<sRbP>?0#2
z`Enfj{iD<}B(@Uvr%F;$-{$XcIA3C37sz+=ZyV-s`3vEQUYu2RX{A~;pp}c8QEuYK
z(h7?4pa*H(<n8QQ5pO)&G|OUJM{7*qNx<P{u;7n(=kUy9+|=Kko+L_tgk=g8B(&s0
zOa(>Fb(E?%I8UEeB=5Zq<mlKcp+=ut!8A60YM$yb01e}sXg5ue>6vHg&K<u8JMk8&
zEc{AF@w;WRhzs?NT8gg?x4Y|*-Q&BLjc?-9ZI~SEmtF<bs?IwSc3UB9r_RJdU#Ufh
z)yKKO;-)9!-`UaB2FD;;#d38`$075|Kq{dj;r)2}eZU~!b7-<7jQ)58+R=M`#7aoC
ztw=aVG~>h0lV{HbATWkWQlC6i>DIe%o-JspoFp8B5q;es=_DrFA!;d#=~=W)t{&z2
ziWF@@bis5m+)f?Iw)1Xr4?B0(71b<ocITjXx1K6HQ0Vz1CzUfZThT@LjjtK0%u}(P
zVthW(D)7D%g880lMuS?V1J6}bJ*4mbHuNucTvdJG$a$rlI+d-Md6iA44ef9%(Jn$u
z50cc-FF#VTcgaEraYZ&mz~b!)g@A7{+OH1R4i0E2w;UAauq{bhVHjD5^HPquBL>mB
z-h^GQW~ySGKi^}-E6s3#D;VQr(fgP~(L6VU`&HQ_+`YE-G=d243-{5wqi(anq#rI$
zCB~>jr5%7rrXR(-3Z0veQgDVKjWh)Eni#%#vn7ove9;_15SP6sG#gR+a5hL6{FL$s
zyJ_q0eIJyl^;2{Sd-Z3YxMDqYp{x)La+%ljQSc}sDQ7yVzC5IhL0Gd^_Pc|CWiVlr
z6Qz<$iwF=JywaS+eHJ4zoXf$0nF&R2hFTVLj^`bQZ-vR}D@7OnNt4ov`P3DgGq2)S
z!*7u<5}H3TyqvdLs(81aJ@@IWQQIHHpJwU5d>XzJS$K-aF&}dq712N|S5onTk2!b}
z+M8p6TXse<sp(A%R~~{sCDYseb_+&nqUSKB2Ysh^vixT<y<SA$f3CdXldq9Q{JHDh
zKG~j;UqMAIjnWRcUDW-Jc=G)K3m49w&K-@U*qj`MqT?uC9R|VFNo&Si4F<SXQdE1Q
z&FR1lRX8`TFQZ5C)S_OG(_Uzq9x$dNUTa@y=0#n{k@n~e4`RzI{B-g35j5KKzh?JA
z5$87ALc%i|wOOXonYB=tfAw6dM?v%B2tsqCR@u6hR4e>X6Q~Inl;=2xO_j<N%hFg(
zbvaMIx<ta#Y`;&_78%3XiZdc)gzc49#cPFM7Pk~Rz>m%LrjLpu4Js>(cW4|PE`3V%
zoXTb`U@YCOA{6J9=i_-ihJdyYfda=IP};Ai$ca{#XE|8QhwPKa@a@X>v?{Y`xL<Yl
z9(LoyGBA-9Ad^@2r95gF6lgv&#;39{-8OcYJ-=W5Zb}7z$lswXltkgp`z)58AU1yK
zx|ikiDehtukIcxb{SA&pXuX9m%CbOPg_g)37c~y%qs1Qryi+R+vu#bjOYKV&sPC|-
zr>|BJ15Nky6PI2?p;m-+pCUrO)rV+@UhilU{wV(Tqm+x08TWIo5q%s52Z*NU#KfGd
zgmd(aOZ6}TIG2MAVRx)hpG!5Uis#vJ&YYB`EG_!mi~Af>_`>QPDYd3gP#d9T^*{x7
zQWo=&f86Lx8+z+{8Dt`Fi2Y|VdG-mi=W~==pS3GznUpgO2^^kcTiPzj<SCJ^6VruR
zHYNL(s=nq|^}tVeul<+}+X%5yo7z(!iN9oJcYNXYnrXT7yzi%yqvNO~?1@&BfiGFK
z>|K9Hx`bAY)w}pJr7`!M$wH8~*6-Xsb`^?<{+EIAKdj62ckLxwrhjj{-8@k1v%-k*
z5E%7!w7&5x9EO;~PMA1;dGsRq0}hIig$SJog$-P+8g8^}B0j$iL3-rKEM&&)Fv51@
zmG5cj%7H;}^L0-9cuu}LJ)8pblbmnkK@bNoJ6(B>`1u}oQEEQ8vcEW#5YfFj?!W=>
z<lHIU)i0Y>)sHUNJv*TOa;@U`YOMZXz2RI-6?L;~s>AIf6|9g{lX|PBOn9kdmtOc*
zL#*cZwm*Mfy|ctKU%ex5(~JUojBXpJ0@B%CUHP3YY|Wska4tk4)r}-VH;=j-MVOtJ
zozOX6*IZFF9s;CauX{kS$v3L~SgT;`{E^uXRMJjM9v5{_B7;YZzO|y;&{4wMjx@G{
z^zij2$?%Hecl_^a>DoIM3C&%FuJxWCCgYAzgh<<qL_Wr-IS@e4rA$SFZskU&;blpr
zL`;XWkDl6Sd~Q~n8p*2qvDzI8L!esy{&cTf9N)qZUl`gZrjZ08@<iw)KUSV4?JI)~
z4Z<_pmXMWm+xx!vhSGtwypRJqd207XRP9HcxmpNFWBU<09Q)1X7I(@~%R;zjx89gW
zsH=A56R(w7?v)ryKh;MiNzj%;64F~|s55<0h<TZ((#rW$R9fMnF80>!`Jl)bQ>@dg
z?y6V4)rwb0q$0;ZUqD(YD47U^-aFsk=iA!TgVpStID?27i*Y(a4<vSnC@(Ejr@tkB
z@ct<r81T(7Gm(X|<|Hr9%+xfZbiCka^7)f=1JZfj(AJC87f+dILo2K$ao;i@mk`wF
zi_(hN6;S#pM`>ST{s>&D7LqCR3#9!x`|`Q&`D5O=li*5(c37z5M||yHpv+WQxXh{M
zo{Kh$#oK5wHX%Ge`Z;RaOv)1a!AzNy4KH8VhF3jo=W7ko<O_T8quE|cgSkOUG{@S&
zurrd9@l5acS)O6pgqJRp-2>0z+==bcQ$A<4OodRDHb{ugluaICFvhMf3<*siYy{J#
zn|Nl?wm8iQx!05~q%fq<_=^l(x#$_m&^Pnex)i|GSv^t9oXE6xg(#~~8#7wo7$IN_
zk`OOC-YD7FVP$3LWD(S>@362k4Ba<7D|!4>oq42kBEa)RhwP~G0)AJTqPFw2wIake
z3CEI8yJ@z7szhK>afL?h#j_HethNj-V|lvO4#M!eO8=m7v5cUZqb-4Q2l+vGqshYB
zhLLRcG1eB8B`NbneK;R8Rh2VDo2{I0gol!Oy~`Xk;wdR~$y_WW0^V36+Qb?cff1GK
zZZQUCTf#n5b`{1)=FW_lKR)t5JE{xJZbfcl!kONIR%!5`#n?u*ne|9D{5&N=<?+rY
z+D(FCb_IcCc^0-y$R0L4OH@n(;hZw|4gorff<FMuERuCyL3M*>6V1fD0t$B8bH*;8
z>AEArv%IxOIg^s2bV6s@EJp#|mf^hPg})3bR;=UMz#s(iOd!?|cSYxxs}Enq)tt(p
zC?^T+FFJOU7fs0e3)>xUcN|mm$AwVl&6mZr;#l8PvfZPFP|Vh^9H6(+Ep9uXZ(p8n
z)fely%|M}Z-5gz#sS3X2E!M3(JyM0fKb$@|-115=;HX@h3Q?if`9X^Gh#=Jq!Yt50
z?6IMJed=6C+G;;e;<vpnuk*02nBzw<dlke`y;~8ZHC_5!isxrxxlvB+YEe>!mQ!2S
zv<AwbqO6M~>==fHu_I|HN)o?YT&cg8t^U#mD`7fK$^Gq}Wk(!(4oTnPJu2qc`p%q&
zqxQtr1M>NsNK_-L-aCI$E>v2Ji~fWD1?o5cYCERV_-_!$(YoEX5!Vr?eV!TMKoJlZ
zPh^4`hbJLplSO`xTbDl*UQpdsR*ffg)8~!#zdu-`FKt=L>0eHxZp);PfITGXGj0BO
zj+dc#sZ=>w(%$x==(KNsE%NdrJh<}dMA(O&<JAU*5RUe!YrbL~)BH}0mt4KO>z?a|
zMY7+fVW%xa;NQqepv85n$1|6|U|x^%B9d*K{`e+X{eb6tJ8@tN_x2?bcl)?dpsTi$
z=})JLh%hU(n?kb51Y`8Kd~4%^M|mMqqp{0Vt0GiW_Ox?)lbGHC6q3Asoz61@`g?tB
z5+aNil1o3(Fl<r+b$ZlC%x57r%XFl6g7$cg^2amn<x#PwGX<t6gkF7epbAZR7w;+F
zy-8VMz~Www7R4M@NYl=bb-IJ*h`@<DrHoFD8ICD*i5`w_w6y}^<#cLA5Znygw_+s}
zK8i{HQg2N?hw7cz2u6)OHU?$TPZ8xrx0e`;L$dxTu@RP12Zo=0;{#$6_&(7uZz=KA
z3~oxtROlq3r?|Y9cI8aTi=%wcmKRyzuIAmO>-_|pvsCaE2Z3XH0(*P``yfp<onNT3
zP)x$sRdK3f`{s8~*^gTaQJvyvEnQ6qPGz?hp;{+VDe`AS>>*7KkfnyS1)=0=7E}E`
zbe-B*nJ@jUb2hVLj(>C+79Dm@-KX`8g$g&Xb$J~Ke2X$-comE)ve~K9r2@O_h$Tpt
zRTCE2{bnPI^%{Mh6els5KWHmaTbTIS!411OE8}a9_RP_Zkvyu=g=c5o9Z87h-Y(d6
zi-&=UwDwi1wZ<8N$8VZr1?b|#MHg1z`e}Z`3~DfvMuWNCDruhUn1ZG>@a0uS=L#Z<
zmf-q^W-gE)oFNwPuXEV>!EHwFEJq{THfSjI+p8S87_J4rDhU?Lgxt)g1Etpvoq}uj
z$*e~+Xf`?B;cT^BDBAi3AJtSobth<vQX-2Ezr%_$$>NJ!6)74o%=S&8l3boJ&swV$
zn_h75+8rG;Xe8I0!j3Ez$4JaZBzyJalS~!1rLy?UVUN=irG;i-h_B_#$Vnr6I_+wj
zLP~>3H8*CzN%1~WzQA5vp4jlgaIPwc-2*8|rrwHnDlRipvh@A+@tpd{fwMoH{uP#C
z`Om@xE$zR#3V!Xw+bztnb?b^3@b8fcW#Bbw8d{=}X2Dv)lxK3rju;df_xDmVkvO5d
zHEp|%g*wuS9YIL^%1m4svrcnZebti0eh=$F#jf4dJ_c{Lz<@Vf9@c?Ei8)p|?hhMl
zJG5SaDJ_en?(cWoKdC8yrS#+ir{Q?+t@n*jeq?)cnCLO&<X3JdbZmR!w~U;+q`gr(
z!s~3RJrmUZ_X+n?o5tY1md@0fE!X_@+3TN-&pj09s@A5bB_IdfiyG-)1<-Laa@`x%
z?bnSJ$ku@Zttplicj^H#e8xV0UELqcq7?7fzD)3H|Lm}yB406R?(fGDKV94JPhA=#
zq4_EBDKL}We<*O*O*v-5j0E))#|7`Fu<y>!LP8MkS}DwEn9p=4Z?N5J12gyrGgr|!
zkd?W5^xM%7s@OeB`Ejn$o4_W$BL-fX*OU0&%3c_g8)Swxt69OW2-=6Ad^foHca7eQ
z&zcIcmPd<k)KV<_-Dz<mwXjS<r+!1?!9*LK^l0ke-e_Ez<--p1hB-f8^YHz)g{)UI
z>xM>vQ6TVj*<wbZ?;YOA5b39BWAlC+rF9sokZOUbN1fj=8HqR~6blznjFG>LcaVJ@
z{`oqriB8b@+mHf_$+r$8)O7FTWWMV+M@pVfghBE{sK;nO*XW>m(fTYts0I)~D$ET~
zM755p3Lr(;?lG+tLEC$%D&S?Pq7>KJx1ZWm7S5=5==})R{8rIK28mp9?{S%{Z0>=j
zjHf=CQra5oK(jwMcXa9`5V?<br^%b6Fyo$&aEPH1(MrvNFlC;fZ>}2$y;xQ(GY|;h
zhD`QTXnT~iB}~2{LeKNLd30vNsEk7ZA};bSWq6v21dIRC1f&*2nip>fe1UF*khuC%
zUQ77nd|0Vo-o)TrR+C)olLHbnQvGRDPoxsBc6)M$$uAWMuQuo`(@#YB-Cjq0t4dKa
zO}+x_bz1AS#d05h3hvu+&SXb*qvK>x`<54bQ>U{wEZSp>eRHx($=H6MZr`n~p=Yyu
zU3wPQDx)a4D1%BtY)-GEz;1)ie)=QQQ7#;gK^Hk*GFLS-V<6fJ(s?%Yqpqm+Xn!dc
zCWT__m{|4aIe!6GV_3uy+=tEYCeA0`sSVo=4O`(U%Msh!48uh*C@~EOlD}eHNv);-
zZ0}b_vW{w9mkGI>-347arQkuIvV^4aVmcZ_#>OU*za{Sok1t^tg3yBVX%6zZ$VvxT
zxOG{XYuIxE1Pq6TRT=xr4Jgx>1E{ScF^knpQ5orr1V36Gl(o|(pZ3{CkdVZqlgx>s
zcf@=im898Yq9v9H{mhn?C%MeRd5p7fp4+HU%E7T8|4nsgdfcm4kqYX<PB$1!^)WWJ
zixP=vEB_EmKwETAqW^}?#wToZ1SWaaGqaU6gdIp?EY6iQi^`XAa*x$*WJtvvn3Zvu
z2bqe~xK7N9TfXq38i|_1Gwm^y;LqL?)cP{PEVf}?PH3=xXkfS2i14aA9iaB}(^$J=
z5l()rC1e>6!Hn$XVcanfbA&;PF2XYdcP!30)9Zd?@sd8v)8%nDY4zE^baMaU6~12!
z=_-Mhbai#CsAa9StsWNE6*ke*)1_84_z0HKWdv*M3h5f?>09B_gFkr<tSqH;E%?k#
zKAV~9np)v9Q}dY_n_0+y*3!|Xey?j|prb1z$oqE@URqju#@{P5{7;r7+Kqo9CwF@7
zMUC$ZLqgO1+#QW<4;tSvE7~_QGGk4_G%M?Sq^RGUpM;ewB3h=Eaxp5wZz87P5tyX-
zq~&Ii`MFULJ6vJk-RI9N$C-bIq!){}Pk5JUsaF$l5q<8QcwO5wGidnu(ZBOqIAk_o
zYKdVmqBu)TjEc%s$RYL6UQcwr!+R=2Ny^8~4dkJpTsH^JZ2L7jo%tj4uJ_`tC~U5`
zr>J<rhCU1TM|e=8REq50P|*`<zEGWy;g@(cwZq7;pMtO4e8jWuULD|KKP6(!TH<N&
zuzU6W;ix|cWupgw|2~o4jjO1r5z$qy)qM8)b%<TCMXQHhFbb|b%;MGQHWd}8f`TJ~
zaI9<f(!#=hGrY<H?y4`;Auc6%c4Whs$2dwKcI|L^E^XRJM^#b>7kIjmK?{_>$n_?m
zs_Rn;-?Q)V<cW+j(9*t!m-%V;3V!HGnjt#&P5l#WCy0N3^8yxwK0_|6z>DMi;@Vit
ziM@2GGKXfv7j@0A8yot69)-|7mdUKD${~uZbldGIHQ07?wuVLSPF4rI#_{}TUf%8}
zJEYyIQk7CeWb77?wvMhYR-2uPbkV4lRPkawdMn5Mx&8X9tuOst8dCb6?+OnWEcs<*
zWcm}DvlUMBWYV+p_l=<7v4?I$Ucf8lDurxdcMc@;9pYj;5v0V<xw@wCF6PL=DP}o1
zI4}nJ&zND8mYQ5Etad#PwNeU|h$1cS86732q6+j0fa-k&+I*-<>u?ZnJ6B#BPcCc^
zFHx_mND>pz;aHZIS@Ky+$P;=P=kByCy2@JPllC&t!wI<9BOxIn$#RTLc9Jm!w`2M9
z8=(oILH#+Q;U5nqEp1#FNJ)Lb<-vh19(Gb|8-uCPtp`uAdyEH?oP{Mjhl}|egEUy!
z&Vz!2^!{<paA6^#&o+A>8az-{%68O;WVc2Pz7X(p-RuvMM2Hs`7e`NkfmwQJfi68p
zfR~`GaMjO!KVM{G=i}aTLXLDW1cL2RgrwrcjTWnGu-NR84{b<#QB+ifLZ!O0Jlr)n
zI67JBkSQ@+ZZ;nXenNMYdWCE~Fge*i<-k|>Qiog-d$ep<#jtb=)*brgtaPa%V!DNQ
z>atRLQEPz~j^D=jhbCYw-&`K=?Ui;#9EYUohMss4+#2dKlp24~jiTbXzg;!5*_|R3
z8>w|Xr`2iOc`$`d?LKA}mKVpqyXv(Lp5MN4?pe0AwyIUzyo6tBbd?kl>8davRxYtC
zHX2WkWxD_P=|gK<n{S4wq~x#e(SJfK|MjV86gHvomX~XCu(8oLhttIm*JFqvY!2mg
z8A{cP&+UO&qLCC?vYCA$Zv`zZEeVAYxQZw6C@3iK@D|Mz%QWg;M5CyH&HT2l`&y_d
zN)?5lsvb`e!p-#Z^14-P9ZowFF3aa#ic3pttu{mJEZ@C*cgj)Wc)sU{O1%}>!^g{;
zA(?o0y<guQl$M_Ee1Chcy?K7Qp4cDH#>B$%f!%IvB<oXe%<1m5?d8$x>({TBUYifF
z8I9$V%4Xz#C?+B%hWcyw6s&i5Y6J#-sRVZS3dy|P$m9y_lkebhIBRV8=bqt=pKF2#
za=keeEHqzidIk?4)EPz$Q_WyHV|TdJy3~kBEFqUIC%p4k!4(1uE>|%>K{=1{U;!MY
zrmBiUsa#?@TLsT`Q;<{&=HPF)f;f`B`c^nLz`j5%nnyNM!nsJ#Ww&&kNFq+@Jv<y5
zjpo_T<Pn_qi_L|Fq(S|!P@)lJ*2`@#P43K)BScm=hcipz6H(RbU25$@YiihaxFEsY
zZfGHrXV-Uc3s4amQr5vPDc!wnMb!WgV!J(NGV(*3|0GFi`|{%Af{S}&e=JX>$wNCY
z^BU(3gCP<+y1_!@$cWrwMQp4JH+KV$>x~;Q7@9(1WOz=L$B$EujlW?WCQXfvoknT6
z*ca)cY_pk}?Yoa3xTZllJ3F_2l^o}zIQ$Cbxo{Fbt90k7*?8XB($?1Zot7^b|9ofC
z<L}i_;&s{sBdF9QLYJT&2D%L0a*P6q?>63BSQPOle`m4G*h^cDq*Pg2UdGvMY-ng;
znzQ3Of1Mc~j%0GT4~4VIQ`mzv7KT&fd|kG8Iq$HYM<^P(z{7q)(9)vc$`ighJy)yz
zu9OBbR*rJg!`%llGn`U|KeS9Vbj|H)v1r6qu2}TZk$nfTfKV9^*pch!PwM)cA{yl!
zg**gAL~zTvxHwMdD?0}Vo9(eY`{T7dm2$?#Mh__Xs)`DF!x14NMgoGC=H^tR@t?cX
z<*2UX-~bRToQy+5TeK3#Ci0<9%r5sAw6!6!VxplKFvLF}DB9Ral$DtxlUZd;B?klq
zsO2dKz+th<QE=Jt@;13aK&_NlZ$$bYT@Lm1^nm%jDV|Dj7^PnEnWrvCA*U6timTRg
zJ@E67JyQXJZ%(!E4~?7rGbIyUR;qQuohjl(qxRS&o>toz4^As;%+3>&c>BCeGjzcY
ze=YF>9v&VJPQB7XS4_-*(us))OcVaoB!hNjT^u4t5DSkrm1At?*mvVtGEB)8f4Q?m
z+~90`hlA^VzZ13)f7stH_Nfd5AN)@Z*ZNYkIr=IOC0JjqmZHPM4<-78&pA%03qHln
zyIkjq$1)z3UD&3t-tTfao*_ZMdNcl0&Ck!T85lY|-LJ{Cj%_?$miK)k2OC!dtAvDr
zh-j<sG)pc^1~@^KH;!jJEUtGo6VKt`Tqs_Has3mb6C<Era_(YdV{=+>jRbLC>32s`
zUhOygczZvGg}qq`leF2Nquc|}$$v-VSqJ5bzP!8~78*L|bSw_0DU0P=$Nu_cnMv%P
zeIOPaQ3Cpd<L-Dcmy0;N&wDeKN3B}m-?!MC^Y!<@=TrlOUF(hArcxn0%aTrQE7I*A
z%l+^ciF8|LM{}W(O0CNKo|cd>-E6+DH-<r_&gn8kJeEqeVqswc*EJ%U@!mHB4-fBG
ze0Ts6jq5EQOPMk&smC!l!@iy>quKJMmx(wfrKQL)93Hx<wD5*7aj3f!g{hx&CMq0G
zvukT>DO7)TJebFa3&#sQx5p`lr4A<>gL`{>oSd9<b8`v`3K#)ItgOJ}f|ojiRvQNf
z27s^8KhEztn{(VpzM%z^qs2qaSf<sZ-vg@LcoKhy=>8^(S|gHxl$v_KC;HtIF;fMc
zzX80<>8Ror{ws&mg9C3LpVu&AhZ^AKi8CI2UNKb_8@BU{Jh-5%%-AcOE|1359mVGo
z=nX=58fRu^rb><cxI%HFT!;0hJb0hp-Q8We4uVSq<C#tq4imgX1HFI}sPBfau3<HZ
z59S;hwi6wEMn;B%y?v?vU{ilTH7+jqSrX&Dcmjjb7!rJ$!SGsolsegn$L8i{U0ogM
ztnGp1W)F0m;kGsbV%dk5edW5Nu{M{!EOV}AybmoCGscy-JYM)AJ@o<++j)Cr(}#m(
zE#gkMu=H}C9Ss$gxFj9;&_1gm7(8HAi=cUirOL6c*|JRAML7QP4<^SI+re`l`Mp@L
zA4H)^rOrhj_UFI4PJrbzcaEGP7N*#2O-+g974p~n$2p4mC`LUE4ct0x?8MkNr{fx}
zl5{pZO399IsaxGZnDusUX&P&><@`I#l;W@{E_;A4{2_Jztgb&3%64yNb7O<VMdEPS
z_vLMVQmBu=e@St%N|n`SnaT9c&CR!$Z@vMS35NeMJPHgLzv&VKL`+Nyv-7P{qpOom
zNH(!J=0ex|yQ8xN4o6irH4s@o0tmanb9|5_kV*<L3y8ie)nFP{*<NhC>TYm2-O~N^
z$v2)@GT~+|*6dHy8a&n(8%s*)g+sxVC!C+HvA6Vqd<nhB=yGGHqM|}foLObP9aBMc
zIa3MVYnyf4FuyohH1^}cB=cr3POf-ODEK^08NZ`rJ;Y#{2N4@dw1AjI(L?f_i*_mN
zD^$K(HPQU-UiChBBG|JJ#h(HOG>naRdSjUc129I*OjK^SgSX$t&1K8yh&w!|Uikof
zzBe1DfkC3)@)+_LRNr@k=5$%_!(!KO@*r3I>Ev!f*_PPY*tpo>261E~B_&mEIty!a
zaDJ{)py>vJ!kNipuls)bVhzbUKPU#w*4(MlmynRy97=aylqNy)Ej1eFoVLKC``FPD
z@^+PWZ+W@c<<{YHbf(HW$LU4<V+crq)Rg8}@bOcOCyTTfAL14mgk*2*ku}%I8xnTk
ziHV7O@}1)#$!b+<^_q>ny}3f2_O^gG<woNYHmV{bzDnnv8dMbOb%vwarJz$4^VPKv
ze>hr_{x(KWe1}&F#}>VPeRsR%^KdwvxopeX3OR@f2v&(sNEYl?8v}E-j=*<`vBUp0
z-xbZyU+KexGr(m5Uh((T{PH)ZJ^=^*3jWW5u{}Y&zU0YMpjIx)<?%Sf#&+4AI`s(*
zBtm5Gcm<M~V3Y#P7mQ|eS=z0ep;juP^r2QwgM(bmcF3IzJl`IHE0ug9Jeak8v$I1a
z6soXqBbjwl=Xg%F_>W&FhfBAmgjuf&pi1H3aBcwt`|1IO{si25=*cI^_P|@h%@Q}L
z-CYQPoy3{#g7F^ONvV0(lTh2xM%2>a0nNV+auNhM?2pgEvk1X_=ZKDuHZ=t~)4d~Z
z<2|_sXMyYeAi+N_iqdqtJ<evekt)Y1pQSFFh0!0!!o|cCrF4MH_Xv1$<%Sd?Nl8g@
zaYCrS&xkAE;1N##@Gy_Eo?xe`0_#Y7CWQRYH0%o$=|lc!@L1*<yHF)C1|gVWU_5^l
zvB&?&vY`K&+x@Rz@>hTng8MZl4<q+K9k~CdgZ^{`@VtLI;tOcZ?|k$TNNoRf1n}^G
zI^q#Xdw+MtBO&C!9l@Rbw<8{bfb##!CDZ;7`u4wXxCjeXfp=!nZt<4nx;^QNAa5hO
z0saX%WJc4O3O+A55E>pvw!!0*?9Vq$rgB+wyp}l5Cn0`*q5z*Y!WouU0f?KYT&i4V
zJn8-n24uw#So)VMj}fnQaCFSqs8^pm13s(hqfa9o2$eY?`7$*%O_ucheIT&%@Z1L<
z5lOHK_=Ka`$zna=8K*T9U2d-c$l7hkaeaWP0L*|O|J&~ula9lnQay*?$~gB$CRZ*s
z)Gd$%C<!=iua|F#B7N1>)x(Iy>CTLg*Lo3wOt%ai;3Nv)r%#`hm6aKhm^wQ<zj9|2
zy8AnIEY_2b*VNQ#A8KlAryL|nlxjGi%6q|K!JgT#Mkr2|n^Bty3JOk7OrWtXe5kZY
zAJ`z`kdc)=ruYUzX(#ndB<iI{9<PK--yhg=t{lvV8__0ok*igmfV5?`8j(oM?tIqv
z<r%<QuH4*TYA9XM7R9wB`}{1_!{6DZLC2vo1X=6?C#b2JH1X=)Om{aq1%;Tgo~|y)
zRlBHv*rjPTe>=MV*it5wMNgy+toS#JliU)q4N@+7GI;<xJ$KYyZ>>v}DY3<bgkEje
zf!xOAZk?zl@nH}FZ0qam1EU*OnArqe;sfIoj47gB2|8nMs&xBcvAN1K0d4c}@URH+
z89-U!!_Ll5NZ8r}pxAHVDD=QF0n`44YC|FZ?;iNS$&mjmhyQT9{~y@%zjP7*ru$V2
zRmLS3RY~_%2!*bUAWNSua3gqd+7F(az8pnJtR|mDLu2<~QNZl>YIa_)4K@^zIlpMP
z0d@}kDRcfwF94N(v1Lm<f1SEFl<;4lkNvUW(eKN(c)+6kapiWwjZc1G1Pr9oqray^
z!u*K@>>ZkJnJbMwV!N?shPSscmDW2y0{l{chZA6V^i1&B91iuW$k2J2nS%h=pPyf@
zz4`4sxgD1U7Ew_n%--pR5@1u%V$<(dr0noL#2S4(%HXKXc)He+R{qs#@7GhqnZgJl
zko^yw?*C~(TST`jFjbR9MKu|N{hj<k$}mX6XQa=L$NBjAb(LS8?!O89&0;>1VR#4s
z|G*UgX+~520v!J>5rdOxy3FJai^U3X7-3;y;yY!!-I0Ki{3m3U1DWV;sh3y#gJXt8
zUs_$&-QoB5@c{@=B>kZ$LDuyIn?BZ}x4Rn=VW!k5YGO>EJ^+JZs=|Evt%j-ws7UX*
zZ}6zBtUO+K9sMEgwKW=Bdu3QZF`)#Iz^f-90v7Z7ca_gGo~kt;<%y#1`XB<pm#<!H
zsH>YEI|m#*7K5Q&<1EPg7@e<9yuH0McX^USiNtTn6>h{rO+`bML$xunG}NnY-U0^=
z0WNsFxbq>8z#u+O%z3<^`uf%xg~D!X)#2hmi;azK_Lqy_7;M%|1Z)z${s6bw1uqA)
z*&MsA?Kw95#%#5%m6g@ma>2*WB533cKi8@`O3^Ii?+FPBw=-b?KSf4HzHAk@u%LT@
z&A4xIEJ(uZb#(Uo&#CJvMk4N=4m}GWdV?GV9-GT$W^7EsR*~Wg7R3s5fs2dF_4z*P
zw5P-8R$pXFm9iUY_`~KL9Ivz8YY~~7W^5!c*vEk5mE$`1=JU+WrC>l}(k=j<6#3lZ
z2+ymD#ktBGyC;^ZU#ZfNK1Ta6=^&gyI4djbXf~P8%jM><{faG)&9<+tt?eC-qN1Xa
zkx`*J0OXlg>Lt3}*r8UF4IXfC%bxBm)LrQ8xbl5|F{Ac__;ylr>jqKVqSUnnt{6*w
z@oZF7RMOJY+H4#g&g*^gN5=IRUjp7Rv9hxIzC2uzJzDtsqF5wK{ki~(0h<39q%^=!
zLjtH`cVO@e9X(5m#Nlk`L#9MLNbgl#4nX>#9j{-o?s~w9UPxzceppt-dco~}P9}F@
zHBk^G+5+SEJFFaP3ZEq55)!TgeIYO~a4g=mx3iPQdP}acjrNxGoA%|*S0%GDQ!sv&
zl_H~+A`xxHdekyGqadoh1mFrJ?~=4Gf$GeyC#@z2xQK56qFUm4ccv-)^mph0liC}K
z^z*;e8h{{?8ki&B4cvr=!wYcUFOh;?a&Kmk(Lj4^Ddi~dllX3(^FI*g|9TKW+2#ol
z2#b#w|BZnkWhuu%fmyz_KQEYWxV1AgH9LDAYgR}3lR7>9FN#jEgwIGztEjqC`oPb5
z;Zjw<v-3^-`vLmjyBog1#^iE|DLp-hch#s=&N}Jr>Pq>2GLSgtJj7$`sDznt&XN?8
z(d<_`UjKD_qdOcIOMbHio4#&P5nXugK{%QJ;VjEFin3C*g94~xx$HLWr_#T55$^0y
z`wyHQtF>|@dAVGu#StOe`o=}qROy>PxK@;J1aN|m0291HMvjPz63j+f;<-#uWM}!I
zkQcePCq7-WX@9aI{U`THFErEm8XIf2+JzHARXfsYptL(h*`L56`f%ExV?4vc!_(xh
zgk*~5%)FPf70r}S=3j-7{mo~>Ro8kx$W@()c-X?SGW$<P-KVF6sqiO7TI+3a(;?i1
zhk%bHD>Yv2lK<yVpZZ6#coY=gz_RO`aoGbG^B;ba<mGh_lNA-N%3N<;A(OB1yW55M
z`*RFQCE%Uk^YhPET6&?_nZ9`O0pv>L@`j}ZyX3N;q$9{U>+i97fUNc4`Y)(2Z;tu<
z`sySkaF|(~Nf<x%Ds#iIm#fyhI4?FWEH-ng)j5qWW0yQoH7-J;qSGM31uVeb&9USI
z`9i7TTH7_}@`K>^ctLKOYsvKU0W^EZW5$KX%$6cuBY83nw(3nrMo}bRPU^agHo%gC
z81AGXmp>;KdgpL4i)z~nSSj}B4~B_s-7r{~dS_OZmuFjIcuE9dR6n$yo`y#`u5WCy
zr7RI6(Bv-YH^AFw1XxrwmN#azRRAL|yh{d#7Mf;XLBv2Nuk+_9{eh%f0$rpFs@C(D
zb)v6<m~rn9tbKmUq1c5-S;cI3nD`6OU>?JBzGSts69fcwF4TO5#VYi}UB8~3oH(2+
zE9Rn+2-AjIm1wtSS#Ot#h>Gd}NvOd?(xadND}wyoal)hh4ZU6ZN+dZpy+^<JkI^l&
zdbRs8mA4KjXjF_XYQNiZa&gei?b$TnwFM=7mCoy#m5|U&rC#%(R{i4CmEiz^o4|41
z4bS<#2oPR?NkdTkC!F*kjCx33e-ThC|0C4*kCal=AFv~6szN}*`Ryf2o&q9EF<s<s
zM7Ed6=wx#<HT%s&#`{a&9N>Na;)7b&5L>9#d1J)KP)w|8f%osh-1to_L~@U2bB2bW
z9`KL$U0+(a{gr8Y6Lxlt_14Fp)o=8UmOc;~za@vWm9;gp2yqxYsok})r~hn04C5|=
za3Q?w;B+L9N83Nz8rcy2PES>}Ml@<J_+GxIbG3`|^yH-4`I_+`SHyczbPgKBzrDP5
zc4h<l{`|!A2Zhx~{5hEP|DR>7{~F;U{O>Ug%5s%JgX#w|HlTIcY<BEc$g6>>ARI;%
zfd1~&a$A7Cz5SmA;~;4=d>bfuhGV%bb_WY)b(d|uy(WMeg=D)q>ZS%PX`=Jl<Wj5O
z_DI&d&kNjOoIy?mZ8u=tby~e&cYJmR;Ey&NNTZmg=f+%ht*414;tI;jKo$fB_aXoJ
z0@a(UD(H_!3wS?!Hga;&&@_<SqD`}qk!ACGLK~K1fMNZOsz7oESYIir;5}2ItLbQK
z``5Mr2=Pv@kN!;Vc)dR#5azEqT4Z%qU;!>MFff>unjs)2Mtr~vbO0~hXQ+zTd{{N#
zoGe}kI-HelcQ~o#(MqR#_3GN%8Bk|I5~<xhAxo)IR|6=+ZfxAWxhiW>6I$=qBqgaF
zKstJOJ|vgHVel+BOsj?m<Tb9%-)VIO{L%2tU`&KV2~Zdyuf6LFL7ErOX3JQ+vi_G1
zj6)^#LQwp9eYR_9X;}^Q?+NvykHR1UUu_S<Vv7t8#${}ovIGh}7ONGBRR_?Q|8O{j
zKUG_Yia^5v6SY5_8N{1BQSVxRbFrOY&#BvmRQMVLgAB#zbZhk3xnVR%5jecRNx5Qj
z(~loN5?!uHfhKvr0AOq=P>KW`9IghvXc>3^9EnZvzg>{<yK14=EgWgL)z5RrbXHsY
z9bm?QwdsG0<H|2B8xxUPERSc4o`6LGPWKgj%o{JbSzFcqgeewF|G7E9A#>&^zAHqh
z_*L~lF9{!Gqbhx-_4g_dQ|Zr{4>Q2y*{{?9OqyQJMvn*I1RnLz4gbqSAH?@x=-^j{
z#V_+s#DVXN)&Ur1x@?gL66V(ry<r<0_-d+hGh;t>|6Wty<sQ_Zr)N-6k<G95k7s25
zqox14KI7ji;`m=U$lx*We?K#JYJav6L9W=;+#H0%MRnr#rwE0j-2<2p7VC*LnMP@O
zVu~DGo<iPco%8EIvUwvf`$P)Ap3ydq@DOJOAt_QFP<Pt?P{9Bfef|10y(lN)>Hd6f
zZdm|2;@{$yo;i=j#l^ETH%RR8G?@IFAt=BaKV*2{YoUPXy|Y4}`vof1;T+rNPeZ^7
zAWx9baS$C6cZO$WKA8m|_m7fI;|LFp|Mombjx${}KQt{Y^du)tHtO|TP6iJ}=!eRc
zhr$<Zilud^;o)Ju-kAHgHy#jAh}v3OzPFQmJ}|HkWdQIN9Oz_`4@AsC=W}@Y`-|mh
zH#Z)D_rLo5j+vtt?BAoKC)0$zx*`OHhirtWwl*%#jPh%B7{{joiSxUVK>T=2Wgp<S
z5BIYkt6k5;<5|6gz}+9FN=z+KV*_L+e>7xvTBzG4vxNMKcMOj3pY-(D*vhFEc=G)G
zsQ<(F&~jvj*gYWTqpPduc_aOfY}tf%pHDV~SG#&g!!#3WZ@q(qN8dj@B#7*z2<#*o
zg8OrC!@B}?yInri-Jpu-b{{V51rz^?Z7_#PJ(tJpOvV!hAggU@dEeCHK^OvPMHAqy
zd@0o}EG*pJ-8+&eE`ehQJ=U3QHwmOsfMCI2ogE#ofE|f;X`*vD(DSR*sMa}E0;+FC
zY6oc38yAPmD^fLL@vICqG{Km@QNJ_E2a5cc;e-Dk00{65zp7?hep{~-=L=})qT=G?
z5;-Jj{&bND_GG2QgM)&?LOd_+Odx^)ViH`H86;ZF%)xwzOM81%U8M`l{p)yw>gscV
znWR=Gd-380$jPdow~9MCv4ad$82!l;iwm<Z24Z5bR=WRG)5R`NkI@Sl)d$(cYI8_X
zUF90+OMt!vdBaU8p=gD#&ftSwD<uEk_IytbvD(wq(`2f|r^2@bK!*&mXgY(TG*C5x
zKE3XGe?3o}=tKt!!T=`_(c;UipfHxg@Atmabh2nQgl(qUc7NKe9+?l9gk%H6Edm83
zDiBz#x5rkyqcYRe(Wmu5%kj^E)DPsb(1~~N#UcRZ>h<+&&DB;87Ml%41s~9O1p_fF
zL9xf)Y&G~~W@I2JC-=m&DXyFW@y`}0ETGT=3F=MI3?&_1kxaS>P^s|>Q$B-?^YwU%
z!7wO8(KQ2mccChN#yJ!O9SVXaQZfHT@PIEbh(4g<*3v(LGzFwQn?@{F+r5XV4#YDB
zSQ98HC^;pPCZiOE=~5%bY`Ls3Vo6R599ruf`5Y8X%o%_q1p+WY<P4pDx0qjl2S$?3
zeCg{1iXRVng_xEW@_GZsNBjGDj-nZOjFnFq<_EcQWT`1BC71J>@}-8OKr=(0p(G|2
zM5;#{w=3}W?-86sfw1og#(Spdx7rm^4FK{7<an;Ti{{D64|l%lZv_KE`4D;>DnbrF
zIe88kwGP|cn;Wzzo++Jz#y}opAMyY-OU=zXfU=&t3CeM_w+G`FGX0JxS5S^DV$U>+
zN=?j(*xT*{*uN<tp5-x?>)oXupdR>5x^1sdw}~TZp4nb+qy~L`T5dA^4a!gG_Ue?6
zk8jn4mfLXThgLx|Kss#JTS!<~wVxN8Kx`F0NBHuB^LnQkn8~(7jm2sgBviA-GDsHW
zjlp0kWXbSvgwFVS)55FQ*hy0$ThAYmz#X(CW0`XxGMY@KxfWKR&pDdiTpYI9r~@HU
z3EsCAt}+^szBNELt*XYCd7azH=>?BtI+W(4M7h+MvJBMQ{Gw-qLD&y4r4df2&EI~}
z6IC4$<Vi_MV7#sO=i&#Tb>%p#s<pL_j{CXZ>>z^iFEx}u6pSE~hlYbADskQCk?u@M
zu(s9f`97E_iL>ND<c+6VZmKpr3u;ruW9U045&-J~Xy~sBB%n0H+|qJ*_IAMo{$$;s
z@xH|XV8m8}D34YT{+7u}AX)O-;^HY#L9%(jHT2}g(?}uX;863H$?fgzrC(aW12Vg<
zBBqcx1j{W}OALSb_DW|MNNmZGNXI<zoh+`Su#wQGrl+^}B%<kb7aH7JPQ{q5HiFTJ
zs)hxDt`F+9C@WNyl)jVP0hSDb92(5LveMGg_(K5M+5tOa`l=dju<jBR8u9H=Fw)V{
zdH2zSQUy?l>9N<43rfu(*_<v{qX1zxI-3XTuj(yH!-3UVZw~R#-GTGT<bL4RYjF#H
z{FwF~q3YZr0FAb3p#m<U!DBpMqr9l73KYRO0p(ymo<{2l7&uT?rlzL0S{^lDV=oK4
z4%+9lt5R-?#SWMwK}^hSxor3S#K$_`KGHb~l$WSkAe3r<>+o1T+nok<*Z5lf;lc*Q
zT7RNOt%G%8V|6RE?o^4vY?XCbIlzZGv2V1cy3NG7KsAbx%AJ*@wzl?UkuIn$6Ew~6
z2?{DNE*2FQem$*SNj&Cp<W{J3+qv#45zk8a&0a<(bWh(NX!I**AF+`LvmFzQnSZL(
zI+QZr-}-7Nn=dpxUUE>#RVnYAoU}h!Xw=qvs2#ZEF(09v;+aFSK*Bz8Za7-)evv2D
zV#uE?<*if0YQ42FH&&t34hIBzcSj5L5^sR0{Rq_opVe~3Ka$C85r$J<UVgtCrLnSm
zxZDm*?%C6)-dLcXOh}(!3uyNE^wK{?(k+1aQw0U}5+h9|&pWjg6z73*gR$XZSxjO?
z<z#POWC&S`tF6%-V#XD|QuF1u)167#IUb}uwCDIMz+azjjrz4DP%eG9Z~pf3YW=2F
zG=q)FVh0qiILuU-d$jnvD}C^jTYZ)!CLkctnD2)SA9EqZjD{OIq@|^moSb~6Nc@U0
zo2{Mzl$l%}w&UPjZt~Iy$jOD1D9(DlExE*ji0X%Xs2lT@12tpKY^%QBN|0NU5|>~5
zfp=^Pl`0Ras0PgfWj{JNw2=QPuKZVD9538I=6dP<uU5KZruomSM$pmI{d=3$)v9z`
z>@dR3UX!0dMBxb?$gf&NOxsAx+l1Lgb&*B)eA?44ik`;_#vdj4A(%rr?>RGz@GRJ?
zp^>i$^Nyoq5l1qrKQr+s1{874<IHdU^^i$0p?B_SVPmjx+)6K-*UuNr^}t?r%e+t`
zSqxZ#b&gl`)bsRq7+~{?h1FO-@8x`}nn7!{_VnwFLSgr0zQKrbD}Jwac99`1gEZ?@
z)t(}1>U!6^8-inmwC9X20XR3~?d!v?i5Kp<k-WkTH;1>R>c4iY(*(QK9h}#Ht53M=
zOqZz2O|_(l{TkjpZ_aJG76qFH<u0deBnCk$)PxQNcC0&;`4+C=`!omd6>Q_jC0qB=
zS2RL?ZRJjL3^0jl7{s}Qs@ZAc_>YCjLg@)Mf*JYLHR?W~mlJLI>!sY-S&#6{k1~5%
z1-}0t+`eLY_4W>X9RZcn)5jGWw-^1%N~Te;r9O!tFE7)+d8x4<tuMK2XylsM5M5BE
zg#z<>1xmv#29ZxfSq!GTs8ixB)@neKM^GRepGvGPAIG}d`Eo>wZIruJvEKe+UGs!d
zssUqJGv`Ng`d;P80>XLJDeE);{r2_snuEI>6F#UE#uDPGZ*T%%*(g2AG`?(7delgv
z>{n^Lb4PgW)J{o<4WADc@p`NolDB-2=jrSssloWd+kByD?Z-shrhRf%#AWpA`YDU}
zy(mG_N_G6z;$jBxWOWrpDXtWphzpacZ&+vFd>E5dN6XVd<&5hSV97ok`SvEW`j9d8
zgQ@Oy0hSmm#lVY^y`7Fhk=Z~@N)eqFrFoCzly5JZq8rwo97d#?NLouVeJ}{&XnjPb
z-?o3ZDG4#5VD2lHc|m!Knwr(v5=NYo<%TBE7_V=b?rEOEhq#3nw^QX;>DxYoWobo{
zdo+ugkM&m4)=p4teBa?P;hcRwMiOV?mB}+!4NIXTt1yO(9#X35#*WX&H#{HSAE+)d
zy0(neeU|mKoCqlNmF~Aev~!*jM}A(C99FTALAy{1_o$f5iG25&!}#Xy_i;akuP(Ff
zp6{z&u{|%8J0EqUg+GnzU8Tb?)XBAE7*)6xHbTHC=on?0;rAeGw9O?9YgF$-KaVT?
zf*DHVl}q+MosGs-;H@?Pm;!HzpxUvjG2wdN*JT5BR>G0&r+q8><CyHL^yA}(5>Xar
zxh2b!N^jctr~2?J)s91TgFhIlACpyK^f>f73UU}>uD-?-Q%NlLcoDd?8?A=mLuuZ!
z!5R-KI%rws_Ii-Dw$t%-9=$Ep*WtG>Gq^g}V&5^);v>L{dG)CInBonRnLbXP=(H{=
ztAK^(jUGsgK1ch^>$V7;Zw*(~*H$TzC_T4y6*~0w#Z-2|_i2M9$se5Sls;(VI^!J4
z&$l1>5+c7~f;<zB&)k29&*1s`iXf!^mdcB3bE%)6Ss?O0J$as(;o((LBl^*ojS2n?
z&>49daYp<Fm?U2Y)`GOn@crT3l{D+U+4KZ1{Zb4~<S8ayXjJ#I*`6TFI^f=A%$SjV
z4^%#!=uh=ver~nb91{IFn()1it-aW#|FYPfTHD&wlqczhKVR&5nNK!NLqm^wQE26R
zvjrr!h}?SbS`v#!YTYPPqz4dgm=@C4q)mk7;;+NC!Vxv~Vfy7Gg$*1s`_-*rZGGaR
zfNP7N_dU|@{L~xHu|X6$KFA>;6Y+r(c_}J(kJhESu;fCaf*SJKcMErt*9kcmljC%m
z`l=_(6J#BSbL*O?$|sVFVw}qs1n1=_yx4^U#(5t#1+IQld<-3)OFZ_U7AY`<YKfM&
z?=)(xm`ZP<n5Du)DE?|1rJxf2)=}yp!S1Od<zo*btPbl(?2(63*yQr48QK~lA92L(
zf|>D>ewqtW2JZ-3L)3nBkbQ-7+*51M5|YJTpnA-U_oAdH2^NQ=mwLU4Fi6D3!M@}|
zTjY^K7&as8D!JOlBp0$=xJYBIW#{VfjH<|_#Vi-p_A{(m%r(dK^ns;RF4NkUAzQIz
z2itQzTbr?$pEUYC{lVLBI()^)x2Q**reyNE6PHo=*QZKM+%Dm+r7`6<Ias&+ccWyJ
zNBK4<lXuqhs=lthJ*&GFu7xbd7pQNXt6cLidid^I$8F19X)6?OL^Bp_ax%}aMU5Mj
zdkHBoMdNTi9T{0UZ>c%L<6w3Zpdfip*`jJRzo%<;j+;$mODu+LR97Z8D5iKIhZ|a<
ze~A~Xu5Nx4leN#>dsZ#C_nD-@Yb}I!&3G@>;i+zpsyx@qoM->{Hi$M@&a#3KQC0l`
zKBrgP95@`On_>}M*ScRRb%PyXsB--xwmzJ{{mcN}&z<wS;+ev`3M=GHfq^Fi0xyse
z_npQWydY&$HeG0nBxyUbFs9+Nh<s4q<IZa%$hM|sPwgl~|7;o&S*-XdFe108|HC5e
zQaa?+oh7Wk1R;qcs-{I_N(LRm>4eO6k<Lq4QbCUm4#saR^~q$GC6<>bKb-6wG{ha0
z1=_HkyCR*>8`Dlj%uPtMT4WuCTUi#qQL_s-f;M_8{sVV@Hdcju!o2CI<nbVS9EoR_
zGoy>`$6`xu%l36DYl~NW&l0Qrx3A6#e|+>>n?sIR3CrIoU7D(~J-QIZ#4u1AP@SF|
z?_b&`pCyVrRQQ^{0$0_`e4Xcg6ZmF63Bruuj*@%6cJ<N(qHeXk9RJgcu9O1X!ON4D
z{l|v*Sa8Z}Ufyw^&66XbwBk3m3PqKqwsc<a@ub#7L<n7{;<xA!{3L*;Z_u-Nlrv+c
z`6SAYh6#)L)mqz;=s<FceipXpPj*T{CJtIxv??1SV(xM1ucuxO#wQ4>7++;Co*A%(
zIq^>Q5paGldq)amP%R_FY@{^h^7Y#9W73zMK@PN-t~zs4;vVX#NAWANI<f{iG?Y-<
zvV}}kht8+Ha*S6e=QyGWH?8`gVmV*;UL>$=u?@-{iXShq(l0W*wDyr2I_w+W(9Eop
zhC?fxRD8V+8ezOsaj)Hh2;YZ7?#o@#8PZ3Ns*&qlUw>&_qILW|XYO6EJ0ap{Ydi%e
zJQ$SQ>%4T${CSf2?sH!Zou`y3gU4lA>n0!I&5)2dN~lYAs!nZ78hWEK!<-0$DAdPA
zW{z0P=ww^X*fQuvCyXdR4(JKox(kzE3XM>IFF~*sa84KVJrjd&C%pbCBexcZaRL+Y
zSSpz&{dTyXsiJ+6bn_`XM7q<P(DFI9B>nX4bBfchZl{~$<>SqJ9)iq`fAPK?GtGZq
z)*ZYQ{O>JrQVL!QCSZVl*vx2oPpimWdmT@J(u9fL9RgtpGx!Sx?~}rz=esZOI3&94
z<rOm>6_%vbMoHPqxLnQShPo5|)ay9n)ZuWbP*C%qxG{Bi<G>ui^wuZY4-76n4lgdK
zsiETHGBUx`Ag+@#GP*pv!Mu=YCx&q&{Br*oGm)J)wEn<)?Yy8%x0Wnc54R4s{_^|$
z#07;+40U^GU;c4Qk2B{TSo@1d^!e&tZ@1QVRe(ftS@(XvS3lG$-rD)8o0s3wI)qmQ
zlr$g>nr-uwqVmBSsY<I;vW2H|{v6h0bh${8NhA(>=bEKdKVIp;lNXs3ZDS3)ehaAa
z&t!uNN_gD(6sk=cfm+<7Pu^Fs(_1cjy(aMWN8I8}#3!F!kjzz=3`EdFdPKCNINr`W
zMy^ZB#Jz{cx9+}+`|^L$_MPEa_wm~eWo6ScqEeEQQOS&iNFkRM8Bxee$W|yqM%fg}
zij1T|MMz|1XH-@Z*?aSxKkoZJ{xAN|@w|HOH{C~F*YEoNzT-2`^Zb6>Sz6QN9^Dmc
z!4Y@1e7gPC&opUIz40TLBOY^|`F5hlWbYYv<pkTn7A5g>2@a%<yT<mLB%jIPN!0e^
zt`o6dVtDUo)ptcQsm?KiA=Z~#NzQev?72;kxUMu;_Cy(nvsp?fp4Pi^k9*J9PJd3_
zQ8znM!svcQ-3ZAm0a1H2V~^;pZ%Xhku*vQb%wis8JQSs7D|q$Pocsrm(v5{ioQ%7;
zD?Ck`eqNE58_7^IoeQ1PmOmJ6#lGH!*Y-}$6IuHp6S4Ln_x!*aPL4UZ&{?z7<8~IW
zPn}Btv3~kx(tSadYme-^?X~6q9R1p@DmSQh;`_D*(U&h`SKe;7CJFNh#R(57-Pkmy
zt1H+4<N-x;L6e|kdry7E7mkwjDK1iV6n9!*KuJN`4vGp^soC-2teKZGrdxRBbL~C!
zGBq=s)g4&nln*@m^L5kVsc(T_m-c-flk8QWJXV`3Q&AR5`Ls5$%jWM)fUF8vr|Iu)
zmI}G=I~o+(+WPut+;Xg~_T1r}JfdtW$Iz_$tGE99lZy9zXI~c29sfouY<G)pDBUo_
z)#x);s8V{6$>PS}iP(Lyb~Sp5TqV`stiRrtK8)8LEOx)OCuHEPQv8wMm%I$OT+njO
zuW)iqZJhTxKB@8Jou$ZgR-=QhNdh4c_>bJti)7>9a+h-Vvvr0viz!ZmZ)b#y_&P&a
z8{cZ@F7!|Ts7%o36fyZCsj-7k%vgKVj)k)4N1o-#UHu&=_Fj<JMtb{w*P~a%o63c{
zOnqAKT=>IBzfpp>`i%AMlo#%H!wo+uvh1^kb>?nbmC7b@wP)wht+Q3$@-*bg47JmD
z&-?l}1A>1Qzf%>saWIfMPua-b(Uf6hoZ6Pxr^Opzr*A?|{Hk(7jVr17eGgl(=C`*+
z^^Y2FDc!HxKq=m@F~V%!nY81mRml!&M@t>IUEANhec5g5xRLwETPgnFyDzwhNyokl
zH}+90_FVibx%+<3o+yW}rq(>6qwkhGZnBTms!o3&bC%Pqaro(E^?5y6W~9T$?)NfW
zyGQ}-Cu*r0B}XHJ-WyUcZq%0h{gZR{jq#4UgLayqttQ8xoyjqH6Jzq#bHUb1*X@^M
ziT2;G#{<(RzTOu)clp!bmG%h#+PF>g4|N){bcJ%l-d&7KU*5<Jn)3f*<Bp&BA0kfw
zpE<#T0)j`@jOYK86HKn9UN*lXf4{jUJNtg7Zvr#JfnzEbS}I1TCffwQY2Ewq<TK~D
zmN4HtrhF~gM)y;1@%S2XDZH%n|Cpp}Yqmo9=g%^wiJn1a>rW7L_SX#ez8x%B__`2v
zLP#tnm-+ve&;MWk{$w>;2T~4$krnQX-QC>@3A+<-f12x9S@z-NO!HGC<MH+4O&J&|
zTv;CRSh*Nh_-ti)rm$S=-bCEWBDwTQbr#RUij!>WC-l0c9qCzERqv29kUP?IkfpZx
z@`?ki?nobRv?pzp^A>rR^ew;psl{IFIK0W%^Bzl}q67YY^@C}Q^t*ORxy~{xD^p8$
z6gTD;7UFVK`hF_9VlUGTzG;4={SEtRXeOMehUwLG99%lmnr?M?NRiX+Uj0?&b*iLI
z`;?>gv)Y^Z8I7%4qoef{6%{+B6Ay}0&?QiErYWBP{`x1yoPN}@B!g{?Dp%ad_FcQ^
z{>Rs8=$1*zg{CrZy%QsxY_^O0+TFo;h1biHSRw^Y4438$^uPLk-DAwf71h?Jcl{Fc
ze`DhXwK?@!Ym)oYfC<OhJ&7d$tv}sd4-2ep<B-iS_F|kdu(BEkMFw*l*`8c)(_bI<
z;_@Hbox)}f$?W>oHtH6R!KA(1+~{AQQmT~U%x<jzYjCw$bGC4G{5;Zh_1}bV4pY&V
zc~(C9@+YXSd1Sh?Xitx?T1Uati?3sq6^?LOFHdE8MMw7ngd|fcXQ(P``mnj7!DX_K
zVf970hl@N`T!!CtSaBuJR-XN|pSDa?P*APzY3_|bx098BF6Rv{)|`9f{M6`aHV+TF
z+<o6NSbNDsGDqW$CD1_n(&Mm@!q+!6?M*qsETP!j>qT*fcR}QicjA#<_sdvALqom&
z?iVNcs%!)*8s+${Yi~YjU~0<aBQ)W^EKWgr;N3$Lqk8?ir-@k|OO-WQ!wg0%rCv;X
zcJEesp{m8q?BTJ3s@hg&=8s+qnvN=J907iQ-D96!$KvU@K6^9MGk#BRRyZA0AI$ZT
z!bxX3kC9T9|J{Y+9-ZEvo+re(5107Rb-cJtzV^mFMGqb@N=Q`Z&>Y;(61dyg^5vC-
zbnK#OeX7p$lV$_oK2Uz)S<u}eBYDhfP{6Wnx4H%M);s$I1s_!Q0JmK1HOeW@^GuVP
zf}(q@id9&Ibfx6o)Ze48a|iX+)f35XgL0G>uAzCjImxi(*dd+mUn5_5-=$??y>q%Q
z3M4+XQizSu5!y)&(iSi*nJzOUZ0sqhgw|A5oz3NEKBxg6Oe7fjz{;y)0z$n&)}-~>
zW0cJL(QDLzeTTT?D9ChS2aHfL23%H{_ZXFbeKu$}`}gnvv>qoY=@Uxprq&bS1t~JU
zC@)t4qm+1yBrX6E<mLBsa;|$qoVf!8;R!96gJpbNNpIi2MfEt>E}ob%a&Y^byGne_
z_K<S!(_^(cUBnei>(KiFaPpvXEtqp-okbg09||(c2H@?{I@pk~9;Z@mH|Xb;J!nYs
z$k+B(MNN=0o*I>SQT>*#01fs@R#woRr^Fme0ObelyMB+zSa*N_3ClKN;&CS92K^li
zr?;pkPaGpwdzoVk<$(`3<Q)$dxV=kx@ggGq+T4C6uoFoi(|1Wajeq*_gN>NnFXG!7
z7>asqg#-oZnTYvacjipe)L&A|bumx+OL;?AmPcM+izwc7NJ^>@70-kk&L_m6{zz8@
z><|j)*AI-u<i?Zv`Y(LHE}ShX*<(HMD3&eXVX4q(VS33e;N833pR3Eu`%3+{rMB?!
zDfUYDUY&aJ;spgY?1<uLuP`=EEiF?kEBC*R2N=qD$N<Y1aTwfAtjz<b;O^dDxdv9l
z+r*?h1v75lx{sd1^v#7Q5^vEsJMEucgO4#fACZ#M*VLQ=tlxJF@zgI~q3CG>lzHxn
zgd?a}z_k)n)HA`^14upaD{{uzODfJBeCWkBK5HoTB4=qPXnw$)GLH$-!+xk-m(WsI
zH#RqSo*OqIBPTY;Z)rt<O~Ox(Y8y%@)P4IF_;eQ&F@yc$3S$!!Lc+r8(f*~Ju%26;
zgU3q}&m8s}ECNG7@4X`f@v#FE3NPNg0oOpPdm~o6e!^gQpm~TQZ|>E7ZLLb6Z-Bib
z&TVKD(G=fsyEFY_1q?RF^H=2k0y7k54QXMXTi*cJA|@)DK<u@2@&gUY%7kEo4B6@f
zIVG+{-zfzPpGCk99k=}a{M8vBB>M06_EYGnb1Gs3X(`d2i(3x*DSPn8x!CwJ<BW?b
zPoGYVcAdIIt#{_k8UL+^>g((O8A+1KpOuueQJy?SI6?Q%oPGYj(6s4|zI&eYlsSkP
z;}a7shn-))evJ;>DZ)w`(e?iRZE9|Q_wL=@D#UMH?xP=xi}BaE{?0Z@m_LYPU~urc
z@4JPsaZ(<@6|Rj>x&plB(<3|AV^Ok)fw(d=HTCxP=Fyl8WIKTl>)zGHn>J8Y4SDfm
z49FLTgRheYz;_1kzQD9DL2703WyQt~8#it{tg8?-HaW>WcBsT_#iWd7UH`Vayc<_b
z^Vz21F3u_|tG?)x?wRK7L2KcWs68(qpX}R~hnc`+Ui+DUbm%?^F^}6JxR7e|5lrjo
z<JJXNYaXMqsHLr49VxJp!#I!pz#h4PmTW^WBbH#eNA(>iiOmuvuCU{X!8-N}oVtT&
znn8#e{P`{-Z{+uH^|?P)2Q{9<D($|p0_<>fbo40af`r0?kgMbh06J96CxwJw@}B>b
zfJ3F&TlA+Uj+)F9=r-8^ntmNMwfa=mr|3Ef(b2Jq9I@&ANL5yA7#|X%s-dw}!V$Qu
z4!0S0@p5-MHTR_{0GbC7YMiF*3_WlGguVbj8amz&<OjgE(Lp<E_vgEIIEb9y=-3F+
zxw*QYJ$sgSSMT4y^WgubrKjKg+t@#n{*zP!);l2Ha$$Q!L`2S;#OrO^d4BqR9CqjO
z(t>sSdu`PXeO+AwXaiQ%G=x@ee6U$HY;%-}?+s&o3r*yX0yqEoCQ*WJ&D||Muh!_b
zOUm7t<FUx~;);s2VgW4s(|MP%?$bs_-4G+F1*qWf;gDZ|&R@QK<273&bt&&DY3cP7
zP1!!^?{Nfe18AClJp>Rg!4>mLofq3)da|A~t-iC+#MfyQ{V8Z0(C6x;q?7r0gX~1L
zWsC7JdgDRc1ro+XGe>>&aIPC;!kX(l-}6w{_zMoUy-GhHC%HRwxQxS8J|D=gum;wv
zqc)e+=f3(2>fgS5S58Z7d-uK>yCn7pU~C^T!K*I<zj?W!8mI41Z?E8=7b$Le!M6(w
zJ)X+l@B2v3q0UUFmQgD`4-P%^?z;p+z|T*n^e#E6$SV})W_qBR?)pQQd?3&N_Vu-V
z{~i++RcGyeUDw>4mwo26ynMp~ltXY{agsn1wHxKp)up22^(K$94le2PeOr6TWxIfp
zcTf-=?=kVz-5ijwKxYV0*sU`AXY7Ec2WHX1#wL3-R9#h-VJMuMKmKgbEKnHmJNvk~
z?>VwN20!B9J0$k&XQ7A3bR0Vb8{B%XHN<1nRs5B%02mCt{^utL@<TEcZj);dNVg5;
zW@nF1PhWA`0w0Kz?ngu;q8y2nz@34x<;y|X*_hE$o5K;TyLQ#|ZZGgyDHxZdNo^U8
zw{;l>{|RwRmGhQ$gkv)SAPu3!Rj2JuO-*;COivT0ViTR(=b&9#TC7||#JOq8KwA1X
zj(6UTKP`*<(D`(k=v5{Te%N>?Uc;2`2vleUyx|B4{&Cfxi3=AlXfM#B^C}@OuJR&D
zURAY@zmQE>pg$|+YExF+L_^AX2Btih;2Z!uu*h6@3E#PKEB__kpJB@_x8A&&!31?l
z_n*nW_=c-GI*;MOIR#3tu8fOxE$^DCLqgt>L+FFW%Gwn+SY3q<G0~3g+h^zJb0&s9
zxX$H(c`+IvifBW9qPLRrz@eK5J8j?n{{5R%WPFMf?N9gXHZ~`>BY;=g{G-rAUrEV-
zT1tDior;<|M%YZPfUm@h@l)M(QLZNe%M?yksoA9TT>u-=k;xBpBpyu!7y{=)Ztv>%
zt9ZR0jZ_yGccY;?8ZVsvCQhN8RmP8Uhf=xSXD@T}YoL<d?kxZ56?)^(S3n)lI53%E
z(B9<fd&R^q!ZE_|@VQ!JLZlptNuY6Gc9Ji-OHOf}JMlQziivo6SBp{7ZY6OA_+`VD
z@NM`|Dj7#&E!Rm9V6;U9a>1Y-zf5ve@iaCud1dkJlrINYT;<+b&4>H0p|k5#o?Yez
zE_7R48=#`T`>{1DZrx+rj}4tOq$lA#nTmEe1`gJK4je<q(2>K<4{CN*{fV(`$<D~m
zrVlX;B0qqQ4?`~0O^F>N$SEQ5NTc-Dt#z+^T;b&Ih|!gJJ@A~H;<-`ut|m?jcp0vJ
zgOdijx@<lr3~qXapY$f@z7rG#-Y<LW^jDfKTXyf>-6>7U$j+Ye?Ad`I>Ex;=FUgfE
z-T#Q)9M<(_#>Y)c02pX~u3eJ$M)*oqBw##GjFkKu8ynG>-zDa7mVw&e+&P%G#Ywy!
zG+}xYej{`}9As`O!qET;Ya|=WAU2*V4X40sqhbG)i&R^j5Y3SYWr-ocT~DbK%XbV|
zq92G*9eYgjZVa^@b#ZY~x44(s(^=$&1kk1Q5FhUV%p|a^0D>OgK#kYWxvime^XAR)
z_ffO6j)x=s{q*FKd+F-x-lb*NaXQVDPyvg7E>6<Y$|^9fsRZ#?Rbxro9ED_DjqcIy
zEbqR(&CAo)(b1S=Qk5lfi7U7WV#J9_EO_ZI=~lRr+X!B%J+e9V2Zx3L@44%zcMUwt
z$*HM>x{cd8)cNStfc=6whkqwhP`r2UL~EhP3WC^|UDbRSUk7Nn-S*R?#>V5Kr`Ww`
zPgs0*>Rb3m`M@7JCPZY1m*Dz0oGo8^1YU$WbO%yfUUdSfHL0{^%NF&(;th25)8)U$
z)SWtr1q>CyZX5XihM%3aV#ZQdR(AXL?eRv<q1b$YOz1&|#EgnsTp<cj2(Ve#k^!31
zQerj4U1zghJLCY<!<s-7N%ckPehG;WIHTRMt#axtm0os76v}4t@7UVdofdqbe~rOS
z@9lU~hd2W0nI}g6l*&@9tT8}i9n?m^xdTK?xgZW2ya}e4@c8)8wW9<`;pyR@{(gsg
zc^+^5f{6Iu+(jW-QkS?wrT^o{978+S5*=0?^y=#Bo`sSmL=splc9x55&|7Lj?&0aV
zq30mqXnZhmaatEIR=N#yU}Z?3yh7L9|6li8`!<3dWnpM|+i!`49PWyR1;V<kNqqPg
zz4=KcrAmY|Yd<6ecilTU$d^_4I3l7MQWpUMTbQ*NYSU6PGM2hZ$W0FMjT$&DE-hhl
z4=#V)MeKnhQJI*SkdQjW<ULcHp|hnRe}^Z#iO5V2M=W%gQga{1mp0v{+f@Ll$LiTK
zkt1T?aE<$4^Ueelzz%2UgpV3ua4=?FmqD!vdzbyui^Vq}An{S$<fPr&cRs&&n!v98
zJ)Qy^-KQ^qJL4?iD>26`-+%ZJR)8;r0Yw*>o=%h$aOn3OKmG;@(Q)hc(51L!55ns3
zWQSFH=H<;HZzDY+zHmlaxyMGNuC8uvAp-z~uB-DId`YOFoc7<!`&LEQE}3ShuqhHK
zUNx=GBHq?qvxjWW*$80cshopRGT0q)A#~@?og*EE^7j$mctW27h(h&(+{5CjY8+KH
zHO);;G!zs>qUWe`1r`qq8Q5SdpFNt}+uNI)w;c^PghPYw39zVX7w{W@k+EnY1SWM^
zP~uf`d?3u>&{r${If6_t$M_3ie5#iZ5U8!@%tx{g$8M2_ee}Y{-MV!v7Tjm`cfv))
z#QgIfzD$JE<q$9;Gap2pyJ<tw_czw047~X6*<V3YIG`qr^2D%8B_g4A_vFdgWdDVz
zrBXcRjah?SDYb=y0!2h64HcDI5l)xi6`<$3b<L<(5B?YTI%Lj**G5`fTg}bQ$LG1E
zSbhEd8yXwmL)XfD>sCh<VW?B(1||OZPV9P4>Ky_`pMbV8{zeA$nf=PLTT@j!7WdRC
zB8N(|z$qzQ%>oZy+`M{B_|BX?`+H&zaoPiArbw(K91rQpGguKsi$uym!oKR%gbnSo
z5kZEswuTQ78bIP>8w<;qc*@$sqrkC~DpRpgs)(+3NnP4W4L5Bv&S?VlouaNgUO)Iz
zZMTKBb)n_=V>^Uz2M5!`Q{yX}sJFJXww~#y5*v+2?l`CYfo~hJ?wsNZPBX*oK7>q9
zume;4VbdRPUnt&bE-5Mb^l7h<5YwGa9Lx_fih~1J-s98dSD@kpWHIVXMrOy?U%1eK
zEZ~`fI-<r5WJ+Vd<e4#7uBJl5$B%1PR7b!R1y|Q5Uq$>1OqJR3+n`RcpRko>k@-6^
zQqyCL-%-UbiSGXM=c=ix!Zn69`?wdSk&-Tej1*r{S|Dm(UYt|hRfTAO!tp(vR6JJw
zRY`lh`^xgt+C*&!1rss2l3AgCG9~b&HDFKXl|%*y$ET#s09#5NkXOJ8>T=R!*RH-e
zj}q+Ap+koVIQNHrqp%@Ij~?Z2*O3wv`wpM9dBA!#A0iNEP@eK!-HOv6R91sJ{tx@&
z)6>(R7cwV*0h!|?4EIy~0^^An@tK*+FfD6`yb7r+ECn9l+1HmE8WH)#>j%QWhT1O%
z7dAMmu!t*YzjHXG%)9n!eX@Z2_mO-gHj*K%PBjnXw?ckIP(UgxPhmrGpFVdkFgf`O
z%3NzkVV{UdWX&IdCl#Mk>b`fp4oj@KRRA&@8#fa59bzSN*hKBhzka1%b8o-?{uRC6
zEo&xx`|Vp1OG`@=zb0U|$Tmo;16NEf&U4+*&ktiT3tXF^%j4tYee2{sV{|BDPGV<b
z8jPjv9x!6Ddu;PCZ*mw_(kJ9<pvGuvX%UucG(KQ#W~RBm{y}4VWKYbkM~{9ICXiDW
zdm5I-_QwWl4t2!0o40K5?C&Q6n$h^dDAL_eF&|OJLir}Gujq`D66A!uT#ql^B1}I@
zSM(HQIGO(}b=$C=y*;;}!@&+Rh$|d3EGD}XcoysyNZS#00<BKz)+4|kqYAS28ZL0f
zb?_9iu~1c6-<Ee2#mFC#%q%QGFM6iw4zvFH=#}=X2=Royz5TOG1!Swm*4?5B5U~q4
zV2YO}5TyYg=KG~#YOx2Uj)f{Zs@P-kpsV2T|Bp-Gzh@8duX}*t^i=+ctO)DuxBWhl
zlu^;qImR1}9aWOV6$-7|7)y$a*BrD2K+OX!FD~otzVL+dY|JrbhUP+K-iq?_e-MxI
zz@O>1pq5qWzO?pv>ZkHSfM~u-jQ)ri<xf@AQ!R$K9xIYTWiE5$+{F6vV@KZ)2(W5@
z|IfspQB+LWnmVQ7!xBwo?I^K5Ik}5St|xzWyt_$<!yU>Gq_5!7qhUP`!^Hf0JG!_K
zDHwgz2TJ3sSC2!B@9kSHE=fmM*Y}y3!+0y4SLtK5m<E1+{?(1!<N5UIQ>2x$(YFc;
z+>MR*E=UQI9Po0`=(q<{o8z6GeZ<mo=*JJ;wa*I*c8@uto{;Inv>hQZYE!NHjO1Ik
ze{6ep)=}lCgu)I$%wb&rd2Mo9HdKf&DyQt{J$@UNt4ysIp4D-=iHXMbiDwWRKbP>J
zrb`1)@PkWAX6EC_$hCc_J6Mab3{JD@1*I<3)d1tPSWblG;fq(VOm0Yi^g?V|QJ<uA
z^=m98UHBj1&R>FM2|W-XPt(&!p~WF|rAq6);|t$P2;#stgM<IeHV^yQ-fTmhq#NGh
z`BDHl*Mp7EF38Dk7Ht{j6#G(DwPxi#$#)>KQQjzXbOzqZ+|*RPK;s;$G>{Fuf1k$k
zCr->PaOtqv*@t1_K7a_-ZJn)c!qjOaGARDyR9PBQhLQFUr^s13jD_Ag4B6V)+#=tB
zO7J+$nZOLDD;o};iX7|3x^t)B4Q_c%F0&7O14`n4eJb}F$V@Zaf%vrCHH1~m)qF<(
zPP531cUW2{3nkJl9?gkR^L(Uk3^Lv>5=)B{HH0FHO_et#6DIp+Z`BTj9B|e+pawtZ
z0FHJNe;mo$so6>v&Zn*RPTx==C^B-pva<5&)~t;j>euPQ!-t$quUtu{mnsc2&#w<x
zlHhsXp6hRQ`SOLXNbFmL3J7z(Xjvc=lnwaz-%FM#iA<KO_Qe1EB_4)4^1(fO_R#s3
z!Dw!#*}3;?$bka~!s5iSX_1VyPF$CCoi#8pu<5JiI(^S0Om&m}-j`c8IhRyL?bxKO
zd1i;Wq~v2~mDD!mhcs*=*~J154h~BTGgpVD+~#vpY;3Z3MCIq~QRmleh?)Pbq^!Fh
zuiXeE2*|$sltNok>f-Op2!6Ay0PZ+$0C_RR1I<~Q@w1vcNGkHn7w>NP&=nug3ELU{
z(JwNSQ9Nv>`Z&PN6L+1E!i^(d(h#}Q{plV&<dxxu<EK8H8M-%-Ai}S6^Rbl2VS0)O
z4<BB;u9yhA8B!$s!gA}MPey(tABvH5%QxDz>(A-ymqHBz)+Oq8eklVeCm;(uHB!6t
zoLCDegp=;*?w}+Q$sa0t+1$M1`NJnOdAe5iyYC`$jh1<L32ebv*ES&jx_$e$?Ft2w
zUg&7~icmv1$}#xPr_$2vBC&~ydr9Y-K|>m!bA?6>uNe3XSF(aqQ+Y&(l|ZJ=e>b@j
z9TkO>7FhNRCE6GH5%7LR(P}Cx2FAvAUB!2Fgl~iXIQad|`GOG-z>E>CS!@p{c1ni!
zdq;txug@Iwub7yilp-M_q9s!Mlv@JG_qQt~I1vQ}-?%t^bMrLcW|Xdrbm*CxQP!LS
z-zNIlZLHS?mWr`QapF#D`ucG_>_oD@$eTOWWHq9{+Fn!i!za(j9Lq>n<~(wwwxUAu
z(#RYr1@J9Y*ITJqKaRi<N(@9hY&8CTd1WOSn|k(z@6$6WL4rL7*Wl1B{Cp3;egQKD
z+4}YC{aA0{YtKiC%y_u{ga2wT?gVWX9RCHA*Z!s(keKIm>=}bmM%s>eQnT+C5=(<k
zUt>>PwzB&AbuG{I>+J3pL=kZ3(e+oa{z3jyKCX;OfcgJ6`CtdkIp+d4G{LCEaU|Lw
zu-B<Z$Rs($9XZ{ZY;N8pw1)W7!SP-SKdhHVR!1yq1=*%eb;lwjB79bd<#Zk4HH0aD
z|M}zJN(RX`?-=$k@o&Tpi@m`B=l$d!fahw7fo}53q;l)(j8%jn6GQd>Y}LNt2M-c5
zGY`ice+`}tCT$xtGmm^MVS^NnE?+*uMr)~!{s_{|Z8#yQ$HLLGyriTc+aQp2QFR+F
zn+Vo_Z?}BlD5nZ7&ihiJS&`?awQuYJ%pGoo<r`C-vH?u3W~Oe)-v!Vj&<#MC_Cavd
z4#XhoIN@G3e+u0ewyru^TjNi@3^-uhQYfEmy)Xh<qTXdcit)c~^H9pr%%LC>cnTuj
zyub6#+PXZ5j8t$*#NsWQGl|^j4Vc_%AuCJf#tM-+wSu+Jn;y8<kX#LK8N52n^2pz-
zs;jA~sZSrjz1F!pkQW*5`Kv7Gpot@qG6+L(FZcE9!v_vXuWg!mnp#AcNDC8UPF}x$
z9f*AH0?#|ks^*5%`VlmEk)-Bz2pvIs_F5a&`0|nqtq&C9cF$+S{?0EfC{3rc37ZDH
z@?W_PA6UWO_Pp)2W#!!tDg-6*yf0m5tRS*cC#4rBb~BUgxaiZLov;WeL_~OgB-s;V
z>IR!@xaGyhwR!U18B9AfJ<`G7y&dr#Lxg480T|Dhxw&{T((p%Or;EtS$=T-pKoauL
zvcv0QNUW?Ef3C+uf}rUyyBr=75dkL+CD(d+qUKR*j7<!M)wM;mW@>@BAcE@ji}`XT
z?Z}w(o8!PS#tta`fE*BYO)sc4wK3ce43t})DYlrCU=1WC^<ej<iG6Nqc^fEYPe((8
z_z)H-jfs@FRh(xZ(%PM1Y1I?I*z=tX9Li@YP>)R)zf$Z?xtni}1g|B0^eDgX+jw);
zO?VY-?}5L6H<OVOJtIez0%Zf#fRKe`jYnF3e+~8GY=^A4$B(T{P2XOv;b;ory+vNy
zg_<RbFcHl)fAJmo!mT@)AQ!tUuA&UYHPjX)Uk&tu9dQjTCSF$3Iy%2o*RQ|0bEgyr
zW$ie2>>K_4g-J3lE;?Gr>1;FN(P#$^vj+4-C!tBBmE%MN$>mp;QHw<7H*fJdQJd&0
z8mB|0bw5;r=ZGi=J<pvV#*6Pz4-F26_SWYU=iCy-Dc&N9gG4~n2rKk&y;MWQ0x|nG
z{!7d&CfL8`78bElQTI}j8k|Lm?yWAwWH<jiX(&^?%FX5FvA|4V>6s>u5{efP;E;4Z
zslHxZYtNOhv3VA)f@_QafwJncM9k~5@tRmEkM_Dc)!1#Wqn+!A8VVf8HeiYkl>f^#
z|6O{d3E!HUK0vg0Gz#4XcJ@bE1#PQE4EH$?UkoZ+?m0x!J|_`xt@<L}bs5w^tOm4J
z3^~F>Lm@M}`_LifPfFZ-_kvIN`TKVQodFB~FfQ)=Laem3^x7}EFhvTR)njtK>EK=6
zz#U}BYDiMr1{vJd+T(z4M8q~rtA9USUS3|?*ysvy!GsO04QzYX@ON|zvyLRK4$6Uf
z+qRgP7|lzs@2?)k*5kL~z!4G&hDbS2nubXo;yZ8vY^jQD!;;X2we?Vf;WTY{YEQRz
zxU4s<Mb`FatzlHkMC&Kpyond_NK)bIrmxuU;<A9Sf@^g%epTia6Ps$6prPGAIn%M?
zJf8UFO~>+ak;t8@jtlnH(GL0fu=BoqiPI8}E&%pZSlF8Jayo{wDN&B?`xFqo61H^Q
zJGh?z%{6ZdihF;X=?sYqq#2w@j%|<P+A99^n(S35ukk+1^?P(1r5cwlC1t`)dY%qb
z0##>`=mTW8u;xP<QN+GN4UKLeeeoXS6I@cE5&_(@4}+Bgm{w0h=uUXRN8(O`YkO%}
z+F49NVZV$s<AZa)TwDv@`xXBA*ohM-VD&))TATcL<VqAT_9UuCQ8kCga1+CGR7v|M
zp<JE0@hJmAE_ZJ}PLGDXs(VEGKv!`WeKsbhhxo;{>AIpEkAP#}zFr){^w`+Dj0w^T
z*JLax%+!u?{ATFhw3-2KdBR9AltN#p5v3({i+TOFf=h+&LqkMAVI0BRbrCii8XD)T
zbA5H~xLRy)KybLp=BmAW{pc^tTwxsFwwe&<7=6689t{~HR}bBACgkBmrky(#OOK<%
zg7_24wPz7i_|;Tb4*>?TvT}3G<j-k03e`Q{r^?ItFMni{kz=a&50-`PKn1)Q*&o>k
zX$Oa6c(!L-NK_Pl`zg{1g7b%_J)-cWm82PH_R7R%S!w1n>gxyP7#duYDXxs8J`}S%
zaYe}|xa`<53V!j_?HpAVyi3DkqB-*w)$Ik7O$WEi89+_}lSJTtIQM~@YiBH9b`#l*
zx)a36<9`*PVgrS1wJ{cEX4hAgwYSdqHIzNK*y}hG$H?5C$y99J*oTHag@}0SI*u(g
zOJde1V%6N1%&m2FJa+AQBou7FS5oqlp579?9I-IGC1*#WE3PsIjl_V;9wa1Nwja6W
zSA&Mzau(zc5VfSKf-1^z?3gGG>*LyngK!x67aJR&p)&_6ta{H4jnVksPAa*-KAn(r
zf3BNXUs}2e6Se*VQ)JkoMmxKBR9Km=m{1E9Ma0JHM(i71oV$TSGj!VRhJQFA9n^OV
ziH(h|dn(`XCL}V_H)}J#ofNH;n<V8jy#@bI3V#0lIq&}2KB3+FX3gu!P$4V(HP+|R
zZ(|E5WmG`RZtd*o*oTUv`{H&MR^r=b1#S!TXY6jcV9ksVvFw{giy|$y^PvF~eLkPC
zaOlqP<)82^J3q9cWH{$4nM_UWVQ1#8TkAb-&_u)6)CE!WK=iX?b!Gy~LDeqx4T|ba
zfA+w<jm*>$hP?&m0qrXwO%Tkyc32<e=T~gUMVxOIsFbD2oLO4+1U<9&WY3Orb8~;m
zF%n8Qf_03Di}OG%1kRy}v2o-4Q2{+7HLXuZPLK2w%aW=U8rS(UJKMA^_nD372;SI!
z>>PW{x|h6n=OZK&slKfSPj2k`8l@?#LvQoGbEhs@XW-6y!mqGKQc9+6T%GgdH+s+F
zf0QJ@S{E*%f)D*`*%1Wx6LYR@1$(EudBs+LTkzISpv$K4<#G<*v!iwM{4taTqO`51
z5fAcmwe9S3Q3fU3-?2meM*iw4_=x^uV^h<B<O0#Hw0BbK**_b7%u>hQGpOgn<xQJ5
zA<i9a{~*@1%PlfGx(&;LEb`1R>t~&-6I5{*$x8jXeS~2MK<oaBXV5;5J(Bn%FfDbq
zzM+AIJF-z8Z;8BRZf<Vn0HN8AXyL(h%LRL4u{2z=FYH&1@z=R{FqmjM&bcn9=MUW_
zkb1in6cpZqZ#tSXfRZgzkm9(^3-@@8_UgZPEPC!|=%7<sUcMc`rkM^8elMvE(S|#t
zoJtarA!BcPWi1X@{0O+ORJo6ud4o<M+%xe+M&>vh&-0x0*{+^hTj4h!uckk$B4=Yn
zt|y?y5t=!oy@Ms{us6B(Uuo`)`zUN7k|pfJx8b9e=SN{-o-dL?J^8udrtHN_0>%5G
zJb%^7>TpO>bN5p!;`haQel2H2MQLY<{T+>0bW)il)7R07&XLQCkH6(!g|be>ic`T_
zj0QD$wxs9J*S>Yx!onhOmlQ2V88){#_tN*nGZs6~tmIIAmOZZH*lHVYbE2#jsC@F?
z+D7?IMox|Z^G{C?;!T4)lm+O8gvWIr<mFX+_4baB4^b0;%X^m#horJ+^XAPJA1HYS
z*4FGXV!3ZtO~=JkYdMS$OIDs8rP+QY9_6%RO37Wjb}?*=BQZph1b`^vSuxhtjm4I+
zc^2>^=9b#IZA2hC_(b`lp5BM87nw@D{CcG4-4H*OgTMMVrxeco%a?oFtq-E3{a%0M
zjJ-uoNqJz!)!8}!$&)T5B-%}6q1LNAug5bzeE=?_OPc$XNYQcRH#kM4$j<biUu=?6
z;^PVPiN0cbtq{5rDUW<e=^8IyQu5%sBVw<m8;m=?N2v~IEaf>$NgA5_kX=8*<ce;V
z+w@=*)w$88P{_|+yh!L~X=rUy*VPpVZ5_=UO%ctiw0-fyYtzpE+PY)7uaAO`m%f;?
z0R68EW6#plg_v9?xnAa+*2*P~I#0g3rOk5csiu=r=rs4>;2^Jl>0?XflPA_pO`xja
zkt4Ve<FI(Sj%gS*4b7pI273gLUFVxGsHvqb_@J(o=e{HwRQA+KZO+58{eA0d-TeYr
z^ex7-tJBE_pv74f={~J!OCXdq9by@ymK@tZ+bLaF>u7%Mnx+;@i49POe?w7`<<($P
z>Fbw)T?@57{~%EC++%S{x(fesCRe-X`(rzwUOp+x_}qSHiB}Vjh`{mVS@m`^BONJT
zpAe{=w%FjeW+ldZFcY~m2YXMR+T8Dqgdamt;rA8kmp_4;9`YOJuU;^=B5ddEnOyD3
zIG#Vq&CSK~iJdq>iW(JOR`*P}N~Lw_1&iI<T7Cso2T5E@a3;a(ZfKYoAGf`Jy&sp*
z*y|giEprr&8IwZtetDk4<G4cOe&vYrqu5x!q(W*s4Y^Ng+ZCZ$QT+j?hR3;7<I~#f
ziJ>8{ez3dI_*wF`>(>Pi$zr*$w%NkZAkx?CiMVEHD1HCHM!Nbgq;U(PH#u68?@z@`
zGeKQotjS}yv*-CM5l9;javG7&nzaUYLsu3xIhAO?46W6*FW=l}YJ$GQxN<WR(#CYn
z^I9La9nzUBp_5XVw7}N0JJK^U!f#2Z7G^gj8n68_j^fPzL<o(1%Wy$wMztEJ1>ngv
z@h-Bsl$8A}`aO4Ma5fAzO2tu+%C{B3EkKx79_><%__a3@+IBe9HZSf38w7-{f#8}+
ze_Q?dke3||EiEd#d81u5QY*wRr7H9G@rj9ytfg8Z(LueOBrkupaocBH=Td2KNx9|2
zf<Z&Fm3L_PJ0xNI5Ox!(2}C4<xlZ+o=y1|c;<6Q7q>wwpj^Ie%)rpUDt@3uh2Zm(=
z<v&8(0)2j7y7-P}<W%fzY^GNJ6P^fasV2wz-?0lA$)eN_m#Muh8IsPp|K*lX(rnA3
z4gm+oE>i56s!8B~&vKvr;BGx``_{4S>})tBjoIM9Y~C$>s844+f8NLG-@cMY;6XJ-
z4~t!(o$vKGmswf<MRAXZ`s8{pABw0mzN8P%9erq5h>4-Hyf8Nhx61N@v5geZQlRh(
zRbR*&unS$;Vy|mo#;2Y0sjW@t(|VTpJEvcGp1ep?O={bZHoEapjQZPDUa~n8PTFo2
z*KV>s?+Pi*<jYIa#I*#iVtwtx8X4vP9+2@J{hNv6PRh}=#r3&L9odGbmIRTGvI?7?
z96`zKO#ge+tV_B2qnUi`o_Sa8G2X>>`_RR&DO6Whcu=v^FK~qp%6AwC|GpF`8)O45
zVb}S5^6IFTq@FKP{NugqzM|Ln?xDBVg#nTV^3<F&gp7z_;awu<s;#A<px`<5ooWgt
z=(CnVOT{RN0~Mk?QoBFtcTrE}arcSpF)E1-v?&+wSL_q#*@k{b(XTPtl_S%LW7o>W
z7ef51dsnxSaS-{l^%)c=_DIo5I*#5d=>irt8BY0E;bI-GLl8ap+$M3$w=cCJ&+f`a
zI8S*Nct43ZYcLVI!l&wMO$w9XaVBjkK{d2*mIqp7Nxa9dd8|ebgJYaRdo9i?B*cm<
zpwo=FT=0-laYF{&c376bbjQ6ueF25vH<|3-n_UxoP&V-S4;>3dmHYDCw$pi@Ugpl(
zu-UlnM=#%QZnVqGw<O`|Q}o+VFMH<667#4KPPI!q&t;T`i=1-DdEb3p8^({P{Qe{T
z#f|UYQs86J)04f&MmbsIf6V}!fLD@3W|aH12>}O!vozb%E8o{F|0dCTpd4dy_eEfS
z{${Ro-j0mp4}z_{8~6XF6i@9;jV3RJxqjj|h|4b^9`5Z|owj!RXmTaryDvrO*T6xQ
zboC7kT%fG?8F3h5oVViVdTm17>Q(IJxA7%!Z~Daq78aHhqN47&&nUHJ%ODXE04*iW
z>TYV<m#OzIbpfD4<)8BJJkS(hiIkWVl!`n{kyUkTieiokqo_9EDJTLa&>cK*0Oh3}
zqIRZ+hFrW3bJ^L3s}_@++Ok(>=9NXOT3-Ro8EO9(Ifq}A+NTu^TNe!k;o*2?gyKZ)
z(|~B|BMEGQS`D;i?Z~|S^mgQ(U-cS^KjateQJTHhlol2e!eAc<{14o$NO64f#D3YJ
zU|pG<%FkYAL?CE|<Bfo`sIt=Efr8uCp3WqrW&9EGPDO1Yad9hsb$k96;MN7<Yu~Dh
zi;a+kH@Y2N4!bE`q(3&sju^-mEk#Ae@quR4CMyH{5@uFb4v7)*d20z82%og!pKveI
z?UT@Wz2|pm=ZUp2VYW6itJVz-GM0RP06h=$zeAn}wF|)+G;%ai1pQmjh-dT!#>K^n
zq#JFQIdux9ZUX~O-q#}K*H)*gu<gt{M1V;DYu$2LU7>_C$bUIk6t@IDMPmFz?=R5m
zjb2TUt`QmGT+X%T!-o!KWM%pp{?pqByqwqJe(=)j*3pP!X>@D6`N4V0DD$fPNW;}c
z)mAn(HXLn@f)zqm_-|%3Iabfsq7g_MrW=IYAWL}igzKf=AzlaNkg?UL2-c3V|B@8#
zzSjQ!1p88*0H>Q=+BVy%u`%?G4iu86P{evhp8j2eNO*cSghiu2inJ0^ccmvJrKIAf
zIZ!yg2h5T237opL9CAbnnYwvut!{z?US@FiZ~bb4%Aw$D(o>|wBTQ;A!@xe2>atwW
zd2R9ujuye-{$~e0pIMZKw_->HTCH{W*KW(ZNW(>`E@m67w?~0%r~R(#T!cfUn@gPi
zFMt$=(b04qhiZ=<zJlzY*M~fcBv80C9W%S)eeAHPV`%+7W|rqIf|1x6ENxz>J8eJm
zF{<Xx?=cH|J3iE;bxgJib8+3yw+6lh+9J8CbVxPx`2qe`)3nXngpzGy-XoOOy5``?
z8?*EYyiF@o9kj8v7Cmr(&sNk8L^I>%N;7Ze5>agF6la)ka<bD0XS#yLAB$4eF*RBR
ziz!sY4b9CdpBeJc@BCwKPBmo>Zv-?n!Y9J|LY0RIv-RO>5DR<<W)cC8|4J58NVF*l
zghGJs-az%W=f|fH<Zq@qw|wJISZZ(e6aYDp7!?*4MxJWw6V5$+um<wT2))x>HBor7
zZpmJaE<^Ak7#<*F8hJV*m4tkJe3Yg>JhLqgZzm@FEk|yy-K=eC8?(%3#^U{nmQUsG
zBCWx#QYJS-7%0e^^gcNt$br-~u`Qj(;A=5^d|KJ@axXWdE!sQg2nQW9{T=YR&vsl1
zX>U0O>n@XVjZ}(9o+>x%<)zWhR00k+#!JX){5KLTjTsMgbI|L%OvHWE0GA{l{r3I)
zJGm8JE-vZw=d<LlnW?-O@cfiQ%|I%hiuWnYv@ZNPRx^>1=UNaosOj!kgKdfC_vyiW
zAjlRWjYJfw2MM@11)1>aYm-@!tg5c6LNtu#!9Blf&enbzS<-ch$hbI=V6r~npP_%C
zzpqF$<KlOA$O=~*3IKx(<c#lS;LtOZlL-t=P9k8lKD*pg`_s1fUPdzk(p#;9F6pT)
zGQ4NE)w=@XH;k}z<Ax39`HUQ@Bwac|{iQV0@X+^}`nnf`4Oh<s_z+V9Z8J&Za{9Wm
zfL&4ri28}e20=q`etHOv#oB1AD5U|EaUTLxbm`}`{Am0ZgJf!t#p};rzdl_VE(Kr8
zVL!Fw?%faOcDLzdNlzqFPrA;I-Uhj(wDg{&1^N+ev;?QY<P<b-o;)#B)+{{&R3Ef}
zth%iY*V8H9!;NQ9_Y4m{<Gna#a<`8Zb1b~OhMp^JG~OcqCm;gHJ*JKT$$0xDVe17u
z(M`4ExenbboyG6akONXyy`5qMN2mk`N1^LPwIoq>8mMHPuh+=f+K#3?il(h4SXEKG
zKF+V_E&%roxG?dSmPoqjIK)#+3qRQQ_yLQJbbYYn7_WoLzVnx4%TLKDQna2_IR4pd
zYG9YRYYR0Vq%DIw1LEJ(!5D3AZJ@o0b)7WOC`GbPmF29x=drkxjC=?x$qy@I<V!QF
z+FlJ${e3iVE0=pK?`iVuzaRsES69x6p0A+OxUm4;F#!Ry=1g|DAVu7V6o62*$%wS&
z6p~eVAXw`6?NS&}LPA1glM5wHxy7N~^mYs$f45>VNmMUu<Z$^;WEm^qi#vA`xnd~Q
z6-;n#C^f=03X1r4@DL!fw}c{5&n$z=i}c?RV%E*HUz(Sv$eYRXSlUe=xiaP>1u|dG
zZBVYjUO_u#Y=v^jqJe9*A``3g+^;pae;Y@QjGkLT{``i+&ARgP*6@VmOU$V)O=GZg
zkHltBE3ay3NKf{1lV#xu<$!eDi<}(B!jVU~r4sopKtrfr5vPgsl@`lOkHU!PHanSR
zO^3%BzQJ8ItPU;jW<VsGI|&%;0qG>UgF&+~kI=`9t$CBGIif%Hx8GLetAz6w+{XwE
zTal=mmKJp4-(zejj7JBrxJFq=E0xpB2=+e&Dja0RQe%iwfnLHm(Ru6E$8Q-!l~?jv
z!<Q&hbEUn~j~(kEl6ZEp%P8Hj3K}aM(kLBu+^WCV@d15h&d<UAhvEM`WkTl`074oX
z;gMxxqWScUx(?INB)J^UE_ID<2DgFeS`PX8`TYgzWN6y&2?f~|6R9&#pG9gqJiVzn
z>na4o2K`4b{iN*dTHJG2RwlpbVPsSv7hIKd_~vDqy8;})ZF)hFBE$K(R)q2g$}vcw
z&6lWtmDO0*eoR@M(%mU6GBPNBvEVk4?3kT92H1Yn)6+-i_irR$jtSN@*~Z*Ub)53|
z*j-Y~3gH4uZ)>ys8X3`#*hUmKLPEfuyGB9DwKrG$<*P@ct;WX21QvfC2Xi%=9RSG#
z4tDXWN6&x@I-Hay5iH*p>X&GR1YTVDf%<<>cTsV1F*-WU&RLOE!1%CB%zy8zdntE6
z`piiAcs}DQPc-#y0NV*$*x*;QKGa*En>D48b47dquxPW(C@8SEu;3#~I;LdJz>k1_
z<1Zy}-;V|$XO*1Uxw%(BYJA%?iqZsJ*~FhLNtbUv4udSB=nd>}IU*_=jd16#Dl0)a
zRd4V@bUZ&l4|D!(>L8E|skmvX0?np)K8f<@ejy`)SBdf5B8Su0E~_}7-U`ioR4B9M
zj-(q6cN7M#{cIpE%?bxoaPQv6p3_i993(0dxET_=>a9W126#=txzsIgf$Tb-gxVsZ
zjrPkQm(R4dwh~>}QfU^A(p6e)d_w2Ykfoua>dX`+oc|;A<?yQREG*hAUaMqTs`W$n
zubG)`A}4QNO2R!XB2dxCF<`ESN-xl?2+vertY1BS;0?gkWd-uv&St+kS)1o{qhX)&
z<kaNz#YHP(w}z!QrfpZZ!0|GcLI~T`*oYnVywV?bKk>A`M}+0!{M*5=y!Yb%t>%uE
z^%zEYN@}Vp<ccBeo7!)6I<;jTVKm4xlzMx6(OpjaHDGEM#q{9f>xecgnIDzIvKO^!
zi1IlXpkcJMw73&Z?HLI;YQkXI;zrdb^!!lS&VR3P=@LiK)diFDb*0{eSB9^d>%jDI
zo<i##k44)|yXWw!wO{LzQBz|KDq8^DWnfk`H|Kg@xW4#oNbqEi*l7@Pu{6QsV1uEa
z1f;^#>@})!x{a-^&O3(;*-I7mwiI0a%T7<;+_;d;M+DIN9!n%ndfczDE0uR`WW31O
zn~)?dF1p@(zmr|v+|+KaH|By#M7Ofp567T_0+O?{sCa4x2Xm$$EUfGq>D6<jp$hbc
zsi`AS@{|Qv<m5gt&ZR3WlWS>h`I&8a(Vm-`L!E)Hp5K{3V{fKO`=w^Qwjt^bP^v&%
z0UaF`MGoA*QS|Ss90F2U_waE^Ne)DNiX~o%qUvjFQx`71`e)`pp<>`Z*Ru`1C>|O*
zSou{FbxthuTYwts83{xJ_ki?szqJug<BEcej0|>WX79CsxIzl}@Im6AoyQpyg}4L=
zarERt^WDzGblKS0`@iKvaPK63Nk?ZVT2!I&MC6HR;B?)|)ptL3L09+32bY`H*4D&W
zr0F;$U9~vr0rR9|*|^F9*Zzul%)-)g;B6dHj-^7CCD*!xBi%^l1p~=3cnP)-$WaV{
z=qKNE40gYVfT!%AnNVjo70}PK<tBdGm;CbO%lLKM#VH!zV}yAhBjAh;4VBf^r_u4Y
zpE1CFNO$jkM*gk>E-{|2qAGbB`v_6GC*FHtSQ@xE$v6-6H5<uS_rE$bagdBYQDE>|
z@^n<1rHxBz)$3LVZt@uwu@h`xQD(=)80d+^EbI`m;r#c!8g+GcX4?;F5MSL!hgQr^
zR@Te^tOPF`0^FxP{&!vhQpriTj7_HX?=g;ML|p)ZuRO;5s<&C+qi4+V-+%NgN0bj{
z81aKchy`-5H4g*s&a-LFHrl**oqMnfD^0V=$0ZULkPU2q?*z-0F&^^)4j)Aa)P5b|
zumA#f%oa!25VvmNR=#+{=Rc?V*0}Z+3_49<gm?AyB<1w$VDB0j90B<Y0T7YJ(3Zk>
z9OC11pYKmWOU-_;<vQxQfc?Xeb)`kN=>>7D*(+DXpo2~6+gPU?!Zq_jluLky<XJLT
zH@9a@2W6iwHt*!%Xpfin!UemI!HnCtr#^d@2WUkvm$>hmJ^NDxK*a5nAIDFyo23k3
z|KT3Ss0E<Yy@SD{{7XwqL4oFr8MtBM3Q~MQWkA|RkM8(KT7c!N*M;P>+JNQwrKGQc
z97heo8I+NcNn7wkCF__)i^@+WG{H0Q7KeM{`NLA|YTJu(<xSsy8fxl;E3SZIAs6BZ
z7DoAuVnae2y>a-Hlh(6{`61Rjc>XSL#}nUD-lHaf#sO%-EV<ZRxI6{sUWE_K)~$g%
zkDC83519<~eOQGi33=hkU67;Sc3_*jL+&2~0PE(S))jG*j2VK3bF(=qeA8rt_3tLU
z{|*)fklA459W>2Xw2`@i3NEa*(QrfCIf^G<(O^mzilB&u+{TY%{NR#P9Sn37dgyO7
zw|w>Q-olaKH<3VPcgYU87obMv^3~9oNXf`}^zh-S5nvL(fBP2hv<%EMNFzxmO6L2~
z5(Nl4YpWCZqc3?s<Lce~jh$%Y!uW&!=e`zRfBw8%?|_{t?hI1E8M}J*Dy`21N(;SC
z<m5g;q&A45#+~3L3xtqAf9<-oX!7$Y_#2XNxh+<5L+O;koiHv1&9vfMw{D>|i;YPU
z7l8ZRfN;Gj>9Nc<e?BdJTCrbxC)@Gk#{qzb{$!nevDez3hIHc8)DY^;U{BVQx_RQr
zZ`j%vEDVbQTz~yFt0#FK8T@<j5=Rx#Vqho%!;jO^sW}G@0l;I>e&y;VDRdGF4e)1$
z<QS?b$hSC^vn0@gWV3J`3h{_|8xGf{`J+B1K>HK;vxF0F^TI)8>9)<+9qFMPv5D$0
zp%j3kCJqA8AFp!)jr$wi0c=y_WXG!h1l-pZI)F|Z0dDjBp3SKjR^x>;268O~pP&6a
z{r2B+WRN5ALAHvBhd9gj2pS4Nn-b70P0b}Ylj1ryeelh1_yGXHgC%OaU)yxa5O#ze
zA??$Aj)1_!mbLMr!C4)$R$`FKAW%vH0799XnVFcFU~ome&wi$DZ<^}-RpudPRTZh$
zW|=yC?B;Ms8QTfZwa-7oPovW~J39-*T2#?!LJCN8S#1$-2Cp0FVoIdV5IrT#P}rC1
zYC!*uQAw#7*+*1T03+K+=RQAq21G?cK}NVC4Nlyf3pO8a4399VCq6+SZjRC6ASiw+
zZ3xNa9VkK+hzO<WI(7!sHnP`+yVTay)S$FU`4`IG+Q{=uevQEz?;g_xfKGb#<yWaI
z&COux6L=ct-MTCj^HbN%-b=1N;-lB#SSQXbIF~^%;dc~V|9$p1=TStt@RD3y>oGh*
z%rFw#!%fUY+(A5<n`5v!i%|YXP}dppfG><hFgJXQH?;H@;kM=b(~JIoe#GMRN00=N
zq=y%Es2!rKcfaFXeVSkQ7C8Z?&I=9;JMFTZyF&apd=COfH63DwT<6uFSD)@5n!Ucd
zhAN+d=~+MqeG;<?3kC|vgTJc_w6d(3e>Etz`g9b6?W8y<LRsUZSI*wOdu1<NK-ZM1
zKdqa%WG@?qiR;%xuFG`K6GA7v;ym(nTUl7_huXrv878Ltxg0^sS5vdOx>}jz*WQ1O
zU{7xmU{!>IKw=_Ue_92HxXUyQ<;2ufSf#$t+R0J5#Bmk}v59~8?hXBEIUI?4oPhI)
zGA90#c5(wt6k8v$c~23N1?=^$s~?$u;_(LL5!`8LyEr?_8NCi+WksnMxSdw0uOVcs
zCmkiCWPrin|6nIu6;?q}LPG1mKqxWY+*u_s-DvGRfX<S;E^vp?=Ki#3j>)mHkY8!I
z*qAtUwlfBHcES;rOOyRf^`s_o9xVT0tI)LTzaujlEaEzKcko9ZCd!;6F*E9jp$x*I
z-|u)gbI3BCRKM*i0e5ExTCI*#a3*gLp&9_J1gaH({``44^sD5b*%TsF>d|#2-?WKK
zk5u6W3?o8^%U}K*$YdwMRr?xy&yI9}PB7gFXXjH}5ZyoTW+6U{Vd>B{)Ys!zx}_07
z1qRZlYm|OmeSWY%If&zk6x?-Xnvb5!&*|`L<U=XVxa+^?k9C3sV`*U#YAa>wm@Qb9
z776U<3md|S?z@$`q17cgkOHtWM@Nx*5{r2GII&xai8mf8qGZ|!2qspKibMS7Fk7YO
zzrXr|e3!W6NqKVYI0em~>3ACw3NfihZWsPpGsPOO$uLNL+d&N?G2PZBSevB*_7l~{
zI}6GGn5xjn<Q{PKBLxez{b@fqK(7bf68LNl-~UF?TZC}sM&1uZ0Ey{>PQmUOpmacz
zxT@C?1Q3KNJi|r?|Li2RMNliY{F%?vpC-&vg*LxPZB!+DV)E<g%14>$*ZO;3hphGP
zpA;VW_d8uhaEPp|tm(z>ZJ>J$|MWR#+Nq>q)|%lR+%z??GJv?oZ&|tMpS8SB(3)T2
z-XF}Uf@ac~0XQ-d1GA?+H+i4ko#^txSss+djdUAUz1;I#=vD*t7Mxk}TUts=>gcPZ
z`r>`E<zzC;N(upw9tnw4tj3Dj;;4dxAb^|K6f_}A0%Cgpd;mX3)vWD)gTa;)qnv9I
zFV}zE%_AGqjR{gaWO`&2y0>jycDvSO{sY<Azw$;hYw`xy%ukVN%zgU;yn}De@s(ay
zewj;*wn}<#v(L3FLpx*sadA+(IeY&6HVQIqO?h>tu*n%O$W5Sh_&wXO;Y!UV21dpN
zwhrdi?NaRZV7wFD29L!td2jW55~-nU`;dr4=jY}QY3CSL@0`J{R_N0~+JZ9?mT&bn
zNuve`dzv#h;+neK!3Vu#;b)Ta@-lc=YzXWi5g?|iaj;3uFIX3h&E?qDdIcBVXMLBY
zH*h<cu=&IsgtdF2<)>@)ho5orBt=B9mdrT(^4<UW?_d^-KOWjfH_T-$c0>PE$p$XM
zDifklgtUXJBs<;Kh!xUNt|)fVmj5}p&i76SS&z4?o%1X;H-Cvjr00K6g(yL;a{#-T
zf8{IdMpKuE`(pBbT*}NGU^tH?n9!I?^TTDyqN4Md76jk^K|%6Ix};D4h9yw6VLAO8
z?UIq<;Xo0EShx*TQg>Vr<^%bWU@v!bY6}VpnZb1XscE4~=#v=6hghnisp&M^-g9^u
zctP~!UviMR0!PA0`+!?q{KuzHM5(1qny4favx`3nIV<F$1UiREnv1;#KmGjq0Wv%D
zX;eNXW_1a7B!hXJot=U6lEs7JByfj41NoPFXd-B2=y1_KfT)5`^0l5nt*t#g;31=m
zM<StM*IwYpO1)PZU~11>sl18|@tm~iQzA{>62AU_74uR{N=*EXi9j^g{_^ViFqdmU
z!w|e5;E7?SJ|7P3M&!9^CuwIss8<X=sDvPE+S<kV5bTB<=oJx{g)hz9+m}HFzyr4p
zg6#(FnytI{(ZbdgE{;<xH-o@Haylt3U4)9D83Qo#E3;#J@Bz%67-Zj5Qm!JpwlC9t
z=Wze78(^`ha{5NHJt390w9KNGf>byMP*K-gw{I84p148GEy7PgTLJzc8%oE`8FxNg
z0`H9c1t699NB4Q@Tt+*kQHTTc*Vp&nDd62{nlW?8;8K&5fgVn;D63sb%E<U&Sh88?
zVk!-T@b&Chuk3AX;`@QAn?>P3Q_`TX_6aPYr2C@Cp*HAlcv2NoP3^31B8sJ1`YY%Z
zlQ>SLW0t@+aC!6yhk006OUw7|mpdzqf6(PP4&RwF({B)b#Ss{%JD0t^WOb^*>jK;D
z?%R8J9Uh|(cJky&vsRYpz+GT_Y0hP1ouU_WbxXXA1XH&3N**f5r37K}{vdGxeJ^3e
zElHs2#~iD;!%^}D3E)!KdwTXm-_kYt_wBpuWQvP>caM@3E8=hNJZ7PlM2@HhO%CJ;
zsyfp*rqG^%y2EZO+uMQiv#tfBT|ge5)@AZ`&p@owgxeuxz-~cuFBJW~ssxUtmg{U!
zi95^&ZxQvj!ymnjj4Ui%`|G(eE>6Ig(``N?3c#hH<KTYb?THH?W?9xTEKW9@N2i+Q
z^%%fP8hu-YXjb4ZC*6UNI>{NIFZ<(Dq9~CA*FQ-Zr3%XRl7Htw5uq#MGHo$eJxoNt
z2wnct;;wgccDni^+U|WjA<T5lvW;R_eQ^Cp4G}b&-l7p>mrFtCcOIK1O7$4Guy8IY
zsVuZwZ*2?_JOQ-*<L6J5F;Ee{7w;CftAbuKapOZEiZw-gk)@y%Kk4zxw9W_{kV5CD
zgo031@5|9H(3r7~42Sbh`rDuu;7tc%Ai#Jl{suPI*5cyPleGmEcu}0hJ9HqB=ofqq
z4SytZR7mJ)SS%$U5c1+2H60!LEg|uHBB;n~IgmjZzD}KAoD*W<W{=rpmGI<|5ArSQ
zUG;%vL*-#ow*s|I^D}i`@q#DI`yi0qY?R@RsB`dV74WfUZE&>Ez?XabyBOe4GXaLj
zk~@zWe8ky<rp>^B(!&*G`@WGPU@i3$*RDJs+kVu%5x1oBdOr#`^%n#TL7p!8Cae{M
zXkBDX%nGiLFAj`?w9cK4=m1%nnH_L7H88k!A|fjXs4z%+3C*m`6s{b#Yzx&s!ucB|
z(_BT0W@p@Gk}gw7+gR)cwp<$O2f)%!*Z>27*_40#TL#2|<m7!X--`<fglzo-zwa!C
zG>3=B_rzTX+BsMEo++VtQPWb>jsgjxkm!&Q>1z}hw164hCE={D^LmtXk{SyJrU%Uj
zuO=gX6V$noA{h(^LEB2}xapHC;mPulk1-W09L8l3){Uuk$VC{A*Fil&*t%UDl*jgz
z5(!ZbW6{rclV!0hPDtDi^#T@O+@2E?!`iJOMpw_wl6fO}DIeMmq2U>ti5u@D&G1&b
z1kN4GN#JYL-Tnd~d*ptzte9iuy^>*jrs7MtiAzW@<YvjpA7rETnI3~7;nDr(2-z^`
zOjXEtzfX-!OoV<&JAucUz0<pYC{Lg!{n@jh@0}jzjr{%FhO8V*-LR3klRW+&=xst~
zwWj&DdAh*fbv;Bqf+0}D)Y=kx4j}1G=!7mA!<5y^CT)61I}c;t$DW){@OuO<e?CH9
zo0zoe_U2Z=V!wQ$WO&G*TjPoa82_`_ueneB`1oM!%UWBc37kgVC*{8hr&X;B=0Y!i
zyvL~j;cj<ETHcdB^`M<0-Q_J5Td%CzHOH-}pimBJg0w|@uGB!NkVIz|?~)G<U)mCT
zLhREh1QK>O!JZi;#YWZ|8q8IEi3RL<Y!5plRnQYUlo^Il>Hm2?;D7fh|Ert%?=LHg
zRyLCPkoEj8%9|nrLjQ-hw*bp>Ti1mRl#~!r8c~pLk#1=eq+6s*Qt41Y6qGJOX^`$l
zLZrL9y9ETK;XEHx=UV?h>-=Z!efFN$ob&SXtMR_?7*F5#eVh5VrLHL&FE5jffrZfv
z<J&OZ_TcvUKmQko%u01>%drR8RYxjI#)-_e<VKwOp9ynGEMHwh4gB`@(sL|PIb_s8
zksi`u_sRH&rlKoltB)}AE7UuVK1G<1yBNRNURnD!y*AvTTib%0YeGs&6v$L~jKqLN
z!j2L5`OU(#?RF!2mS4!KGGpI}dYP)*o=wWCG`jZVS1y-ItMt7b+^7blU9iT>!m|ZN
zZElx+aIx)m)2n^|LaKFsE@{?3hu8CVMCIw~bO~Fuj6?BdQxCKbPU00y3|f&YYyHs6
zvAg2kB%*{>=4HI2T%(+LLr%15vm&O%+;|6QTL&c>8@GHKWThyw9=0}cB^Gj!bZ^{x
zultb8j$XHrbJUK4EQjCV4jWA)S2885NNdL=Q|q%VdLHtNR}3?VF@0%5ZK@brw~_+a
zJ`*oUa=i{~km_K1pIv<E@wEkP6%FPV`S}$CWL3TT?qbC?Yc=0zc)5k{qVL$z%zb*d
zH(G@=RV~-Z<v)muBKd`6z8f-;Jas*t+<yM9SUCuF6#LPPA*$gE4s&^iTF2B8v^IQh
zeZoYFq*1wrEHv&=UAM~8`)!N4nK~$FMRYVdtj+pcc`U3G;S_EJ$aA4L%7f=-^KS^b
zUtpAM%fAtx`_Xr9_W`zXsKll+_P%%BtV~tDKF;ME5;#dNr_Oknm9fQy=wCmV$4~xv
zB;TSg6!P8YOK?8Z=iV$T{5LhBHm9Lqiirss`o&a6T-6wn<gW&$_E{v4)!8xPl?Ud_
zTYRI*cL}_ah&M_bGb;b(_Ha1smFASNP@yY|L7B-PZU{6|Hh&6F!s2wIaZ31FsMa^i
zP3gn-q?DV9tD<aM<4xGdjQ7%37L;CMJu!DbyzNmIB>q~xApdcBWqwvl;8w2(6V0xv
zPd~9~Yj{%Ki+FVbRm`eiXs*daZ^K^T1Z)>-({Yjpo7Zgx>T4|LBorj?+@i=~Vr;g<
zpDF4K${o?c!tuF&H|LS$&wD}iIV4^^+2lKB;-$qD_QBoJBfI2`w<uasT5QOPyzUL(
zvKn-4IMxiu4gNYIiiS(0-z@mlrCj>O*EYPo*)i>^S^l``!iRMVyiO|`Co(Pr-G#)7
zjEu3*I|cJa-5OKtpSurGP85HQD8`r%qB-u83a67X{W|!jS1o!@a4&N^D7q-OAb!)p
zfGzaFk=c~Pik<AJ={lN#?U~7-^-5*uVD0tM+3y}x7;KE{WWwP@gAXw%V;@w79yBU^
z=Dle$+xwv5bK#BHV$wHuKGA&_k}B43={nm!Iu)%U3ck8Elc^G3Djq`iIr)>edq>Rj
zk1mFE^-O;v%Zkc${&9iNt^{d5)Z0M~3nz;O`&WD?GtH$Q6Bg)?4Z24)7L>3L7RsBc
zeV!*+{?Rz-5a;ldu@bT_n#}0Lgy9QK-l9V~7v4<eCl(@2ETa&!T^4%XU{yf26{3dG
zTRQojitB2k@_~4U`PSTcpn{w-w|;%|l1W*9tK*i^OYVDWL7XLTS6>TV^ryc>zr*?N
zt{P2;Tiw<?E~gC-^#HMv?i>X}GLcG(Ufz-}sa_Uk%@n_*8HsM<en<q@eVHkfc>Ut#
zjlsM(nW^rQ4`p8|*BD<@OYXzHl%a!hx?D<25OYD9QiQm2aP)2~ja5<3)o&hXstS0|
zpN|MqONC7fb+*`ESvXx4cMTJNc~ly8Ek=<zK0WGr(yZXrd_raXZH)~B;!*o-eYMfB
ztMhw-#tjiF4gH@L7Kz{VaOWYf^gfTUbKbz&NQn%+{BlM$oF^(q#4T;KohszUrMnx?
zWvxq6B$^1GV5jLEO5W9ri>7qpVIp123YEcIOqVcDe#z*?ZA~9jK@>6Wt0$mk7&?)h
zZo@`$%cQR(U?*yuBH?y`lUkXNc=Z_8f*i-kJI_DSARoOq&<M+xlO#S^z2F~X@VX&N
zPpMPPjYf8F{Y}q?iB0r&M401~uk7yAR(IBmKPR3jN%6j^mhv|{1zN!6PsjPOGoyBM
zk4k+f+#M;t$Z>hAkA%4B<5{J0T_WjZIm(;C=NT_t^b;faDyB@rB80z=C%yLA$nT=^
zRpXwhBNH<jx!2;q#P2VCWpT5y_uMauJ8^K}aBM+h=sv{Z%bls{xt!7ao)6pTNXJFe
zNM{VW);yeG(V&ZTuZDRP*-ys&-tjwrRcn4BvP@3WVPrZ++n9$wNhgu6xxo$|<=WPo
zX{$x`1Q(avJ3hzK7r8s1I-cYNJD>@K5f2_Ks=2TUk}>`Iwl9m0B9MWs+_Mng-#2MG
zZZ+Y%-{X=y$f(*qz%3P2O*XY`BGGh`+cSqM5lBn4m*qQ+fo;>8K@x;(NUrwcxc3?7
zz{?Z#^)X?K0DG;pW>ifT*35gpL8~pjY<vQS#N>*6?h9FlFC(rUyRR|%y#1HnljDE)
zp6m~}{w?pBMbK(7`(X1&P8XI}D>;|Th1aiDeQu3wl$0_2>a*MHiEGk6Jc^E&@J%2s
zUL)35r(oYeHlqF-8@<L)_BiLQtkH9BDqatT@znrKW1$`<iKxOMt_#oDU*uShmxzO~
z(&EE_AbB$<;*-7mGh0HQDI|iH7aEa`UhTy$FPo5fyKa5W=MWZjx{<|Ok4Hk<%P1L;
z96fN&$^`RLK;uTvQ%u&pz{ZF{5$SSFW&E_S0+W~<$kVs$c7MM8vfa3WtV{Pe>VD5m
z%H+g_@<n&7>Im1_>JO{rcb!vnMe`CSuhTXAH9X1eluwYAG1DDkZuLeI?+KFV^+qx|
z-Xs;jP941ICwX^6iI=S2Rx*n)jQPH@sVFb4GJyz_9A8GuR%Wx*r*6W*yU|{grAX7f
zsgg*ocV;0T5TkXDrFlei!#_OHPJ)ck>sR_^!Pf&Xt51Kp<r03;F|++8=pt=~L*;V0
zndp)DO)6p&nypHyqWI31&YV~+w(XaN7rMf!y>8|Q<FYY)^~2#W{IvT4BZ~7{OoOk%
z<kZ<5*4JkAx$21${EOku=+qwFY=?b1UiDr2gs~wkKOd5jcOYR3aq9Vgdw%Fap0+cg
zvJl-CRn|zEwj+5-A6@>{55Xpi*Ud5vuVs98z8`eYMaBlokl^C&OzM0w6n!lecwNId
zV<#p6OF3VIT)ZY?beE%kP^?A^JH(&pUFZAb>jIZN%<KzdBvkR-X`2I8DEu~U3ce`D
zsm84#o20+Cd-LV7y-9k_-o+?WatpKcVL#9CUp9$?)>bE+hBr9lFXR}AU7Sy0)?08Y
zP>`UABX+xE5E4mQupKTRNaw9f@YabKld*e>Q#G1(Vc+iYf~v5}>AfDs+o8E{n?ng^
zN(ZsTBCPAC92NqfRH^vb^RVrGB5n(MCi#>{weC}{P69diCxUtDF$Qiz?q^R*<d}Fy
z@{uy?<fJn@#|mtJI`<TWs%1nTHDg$xEfdf6L<*|irD$6sEggTpXhwxKpJ~-mr21ov
zLe{CD*|T&;lJpAS`j3y_-%}Y7)xY@up`6D+a`J{?Avq<ID`i+zr>Nn*WQk|hWE>iE
zGJ^U_>v`t4<hb+Aj2NkH#-4jAxh>qV_oikY*7+{{E}uqVg;sI5+3#p%+K&6+(e#xf
zl<CC8R-~n{8Uv*}xU~Z7o&F=}vs%)9z3cqX=`ZFpe8Y<nqxIIOGt&E25OjA62|Mk{
zcgG?xnrWl#Kq6|Cs208g!5Xv;b@kUnQv9@q)svMU?hW*fRhC}xZERQE45c4mbaAWh
zHKt7Jh29>2W@4jpyr(6(wjBXOMgfp|;9BtWK0q3qCOzOB64d13y^1p?b75%m$Kth#
z2!o1YAEJs!y6i_84)dXc$?8o+Zt^{DBCRGhPr{vOA_yfEP2~7gF&DV6>^qoWE3>X!
zPcY6+DG$5BgPJz1C{s~bVqr#f(Z-Q*q&O&{|CPPH->VKQdA~;=(ANVU%!e&>+L%=-
zcJv&x8s$qq`*t?7*_PkQYIt|uJ>fvcu&Sf{@wMwdbkQWZM9<0yF&zzabq&2Rt!X$V
zG>fYg*NX5vUkVkFfVxfYD_7bPy=AqS0u#%R>|H4RrRFk%7yCM$0)6H21Rs3Cc!pub
zpRk=5%X>MEMyuI4cCxDV2b21-@Fy;rvs)!qc=z*<QjQW$*>f)bdXoPwn)`~zD+BTF
zk9`8KNQB}rJra)GuvO$6q6IRFYl07u&8ID#DCBOY&`Y%hmJY{`J*(f+zSu)@SzIii
zIoKCTh40?@^Zmjvz1nUu4*Zeh)FD3BcqWtd;nFh;+`VH4D9%EfCPzp&OKP%*-B$Hg
z#xfl!HMwe*t%xm{4q2`lu~Fmj)T?gS?J;IQQro(Bb#-UKvfMg8v6xihU5tJ`x2%mJ
z{Uh>E3)f982(EF3rgyw0-shzH=I~_pZJ5020ij`Y4M8<Ql=9_kA{xG;L0b`c#7DnA
zD!&bXG@RGdmZtNOVhQKz#{ol8hHMYElPXPlJhu$_U>YJC-xo1x(VoG&s*yC!_-J(T
zO7guIo8M55evpaxw6RCMW7vy{p~US=8%^W8K>DS;{g~_(@wWKCbT`cZa5w+MG6^g!
zoGcvwP7=5t?Wn0}aCBz*@Zqq2Bmp+gAWFmy{boM4LUD!?&0bAvkzRR8vOPYUlChSa
zv7R=(WX=4xH)Em-H}KfowQz7gP{!a-*|#1?<>X<op{@{5Yz!XEVvthv+iDFSjlGTJ
z8Xo>Tk^TSugAO6g<v@jmDX-cM*mRULk}-TgSDzlf$Cl{V^nT?<EM-%ctzCHtv@`V0
zoqe(Hr;_zIxu8amOdgo7wpmTw6i*RB2TH~==Aha<hfx0p(PO-GMeuKR_ms|Mh7nTI
zcnwffb^hftIr>#?cjcOg$Ao63({o@PAWhN)J?5k<5=vXfMI%4PkvVy)8S5;_gWv2+
zzr%Mv3YwVkLoj0^pNoWpIJDA(lf!F8d6D-X+GV$YpOGsveVu>uTTL8(D69qQO-ku1
zKc(bK!ss}7=*=<z`U%iqPnQR1{+HzW77t$Phr;bqOw&Rwcz?;*(}q2c7WU#)&P6Vu
zQ`~Iy45cHb7sdi0_sg+TX0ll2n3z*`)BEP*hgqAOk0`foM?VfF9nyD|ZC@Zyz;Ngq
zC&7}>cE~m6r8d~BL7wef@lKX<z57t}@%t0mcTY)(y;ZUZ9gg>{ZLy?IJpdsd$?w{c
zB-smyP2g7K0}=K4;l;!NLpAkm>LJfP(|%Kb#P#m)lZN^mch6ko`5NHW)RHsEn4!o7
zXsT8hg^|2jDlDw0R4GEjnb&n)%WQi4R=({`lIcy^zK9%h*i@SWqGW}%s9SSFM4t$H
zQ##Kt$;AZ=s;d(n7QU~|n1sjrja@9qN}u`TJRjMqEU?|KzEyU}(Ar?AzB<77aR?*a
z==15x+tHE7Z5yt*)x3L}o-p0QD^WI6)3`GQ5a<4ttHSz(J34l0@*0y4vm^zrRFstU
z?d{_4l3d(r0AN0tgJlG=?J)*W-8=e4GBfit9a{j!`sus-<ZcX4-Hg{q+5u}8*?zQy
zxY!}m&%BOe%RO`$(sCsx_uYUMW+AMzWHP&`h&{Kja%N-F8zkl(e{sQRu$7f@G<S}J
zBRlA`+B@tUp76oW10ig(UeAjcb5N~&@v*phs;ez0`L|oBDD;!C1&CaR%gAs))#wgP
zAs#ya7vu~@RQjESrAadFt<A9Tg)rH{><v`wp>Is{D=VFKhPhKh-#jUYJtrpGy(Cx~
zJTWnG5d@&x2OmgJg5K17d$`kNy+vOAOR8Qqfp5QoXami<tF#)c_$v&fJI>qAHR0Tf
z;6Ok0qo=n6xdBr6br0l=SmF9u7NFSCZJmol`T<z(UUP}vz#l4COp87^5tF11LzL_z
zhcIr!*~z7FW~QTSZOsV6Uc_+Dbvdvf$R5}6$Ej}wL09&|;uyNR*nnfPBgOP@o0}W(
z$j(3H<t&E)r6?Z}IrOI%?RpeHcaE}@Vok`?tL5cgF}wdrQ%l42{>#xqN4f_O!r`9Q
zAcVi;{9TVQOUn+_)h*h#yG`whNhZUuD>i5`$SWuSMdzP8;=s%VAfigAYXinETAIS<
z&r`yg1jG_e6cWV}J@~EP+`IsW08Wo$F!I?M7BmRQJgyv3J>ve=Ol6&+Vze=4()Vk-
zbGW7a+8sEp=MPOf0oii0qd2$e!NQ{(jwNNbCKqojA@-Tj_OGe=RxH1J+nUuAUh4&&
zq;4vz_zV|7K^`_tDiT(dX`%a`R^bKDE1QU?4?+&#bS|EMl2-zqS@i<qP-27RJjTs$
zh&9zg&;I}R)+98CKpPOC54?F>>F*DvUHaQh;r;m~L12$Kw@Hyhe<|A6!G{of_1K%Y
z{}@ELty8nko9iWhmDegQG>SkKCT&KjrC$-+6^(E}ouPLFEv(kPSJ2qRAYgq+S^&Hz
zmpfTW>7}qH%6B(5#088ne_!z3VL%{*EIY8s!3-3IJ>uxBz5<FN;0u8nWWa?WBO&c<
zOp;;oqhd+99n=?o>Rv70sP26HDEZ1O#H}`^V5;2d0nXkRM}Ulr<h0h!z6@MP&}m>A
zKLt8~@$vDkDcqlbnQ&p22s4=NhISK-+y8)}E_97z^T<{FGK@?BLV%_*uLF#cLFcjY
zuj>Xf)hF3NEg}34DzZzY*-xvh07UJGg@7Rgq5+1dphQ1AGlO<$0`SKi#WeX8aZZ<g
zD^nk#7}_RYh4|DUN^z7V)0-3vf|#;jG+~nDeVyk|aY-y~puyO)0+M~8&noj*1BLs(
zNdUBgm^Zs|R~#^Wt1)71pO^sI9<XgSmx~8#H_@{srKGIJ5K^npp<X?ARYsU|oBAz9
zC=p&Tp9&;@UF&b(ym<-qZi5#AwF(Ss1M5IiDC@+wVP;aw*aGtFQ$e&=T75FD?3|nq
zz|>sM@?&)sw9lo+3(8>50Vpra&IU9r!4AsWZ$fJg+Boyyk^k~0mq-=G%CyU0q=E7~
zwB!)&2<Y;?@Q2oHw&otNoIUx`*w6rdQ|K2J!i=HKR;s%-&H1ijqa-Jv21;t+q0j#s
zISLKn!02rigr@m2^uZ44vemPG!LzW*T_u&`A_dq62xY@KYVqq>PuTyYF5y9b0Qo&8
zWG6Ja&!Cv@*3l79_(;~TAxARgG4x5v$#|Qv9bC@0Xea&A9$Bht_h5rzIkt}@zsJ&t
z0^7ZXb{^=3<J=(C2Z19vM$`e&kSr^`e|^ULwd^dN7I%Va|ApE?mkU^|NbogDSK9nj
zQ}t;3-wPi<Swij&r!6XZ936q?L{2;<<b45g=3Wg0pNU$c*ky7zwx4Vx&&~NGxnJHa
zoBs1MD<GZ5YRLOSN5hWOy8D0??k+d1@*Cn1%3WZp9#eXUA7U{&KgIm>Xrlz1<nNVT
z;NW=LyP;D?$GI<(Q0s|tk@p?4$;B&uNn*WSFR--DCn|`R|NfP-5FqO9zRy1HIY`L9
zBH;vg77?y6#9h$TI+59%iig+(=O5qws{ExVMzt&dhH~**TTYew;}`$<05CV|Jm(vv
zz)J7G2CX^ie|iZh_q4QDQG?8rkfg&##<E5)*!;b)`ds~HTJc6wVSR$>npF!6ewfrJ
zzyJ4@UMDBNOH4e0d#5%Ii77+zB0QYifaL#uA4o%}I$v0u+^g$8#`yT`d1cf1I2G**
zzR=&-C-39qV=_{hDAqePV3&?PiIUg;T{o>@TTo@tDLx*ziQv!klDY)5FO$GAHP=iA
zvO&*kJl;V2`@M1o<;_38_V1cL2*<AcF}H!V=Z<woNd1-K#XpAkKVeP5P|v`At7rX}
z@#EHQ*o-LWQRe?S$o%hlQ_m$Y!Sp{);I}tJc#c5YKThD>n4Y_v->wL;?S8u~I58Ix
zHXadIs$Y=6`*SD$-@Y=Cmi+(i?*8^w|H%jDO9K&ViMN;6F?3u32PP$i>((4h32+25
z7)FNPAs7INtKrGlhptAko58qq@&f<yW-S^XErNxKTQH-)*t9<#n%+-cw(6k)2=rqg
z;M9TZzlcTqMH2{!m9sU|MzPS)K;%U``*PLk;VM`({=0V*yU5=B_0Fx^Tpv}+04Rk<
z4qheOnG=jAwDcFazraKv$P}r<7<HWwE|4R6{L@#vAFXpN3!Ushzuy|BsBK7kVarQ}
z(e7?cK4rwKtf&A|d95zlPb_-gy!pL%0ir`uR2K+(Ux9&vXrhlm(E#B$0zY2n;*GQc
zI3oRZESOvpagxg!>Eu7k1YM&tf0K!d1G=1wgi#>>N(~EB5~AXL>be`&w^T0vE+C-u
z|LldB=EE4Z4x!sx{ubK3z-?)ntq0WJ>^k2rj8`4U22F(ink%Os1TedJ`nQ))3_{I7
zPN5O7w)S%vdq02r0)IAw-w_DY2KeHC(lJ$Rdi;3FYV$Np`4)vPSi$p{q`79TWfw+f
zVD<y%AVTtC);LQ!BkXB~)wGA^9>kDFixrl_r)B^`F1O#v?luN*q}Ubk{dXJ_NZgjE
z84?@}QLo+p=Cc!FS^pjIlPjO8n_F6d#^`YREg8&|m`yFh(QHB_dlHzGWr%DJu_Z_;
zK&=e<^`C$U>D~i46=2e_A#w7BD0oIl2k|bX|H(J~Hy^_PwsZZTH=2gMmuP5)Ei8tG
zZ<L17hMlaJ)V+hfj)-~VST&{^{NCnoZ@0Sbjk^@b@R6OaCQ1I|^<X~Lz-5dxx<5f|
zo#!8`jExoi$7u6P{0V2bZl`_6{iP!>Z+G--SV-ueWtO@Da0^2;4vz{R?tM!vQl4MA
za!ToIru{NQw8}2~T2I|1|9)N31NjmANc9op`Q1M+0GvnRv|p}ROx)jj##;XTr@k&C
zSVn}|a*+&yk&%3M=BH%$5=VuF^E*G2i@^__N7nz^sJ>D}{jd0!|B3~qzKJ!EX)@Bj
zbonw|IikvOTZz(vnS46FPo|up(jr6aC|>06r5619PdcgJDan6uEb5ndL6^`GLQY5o
zJEPbk&V31djo-lqk)OX++5@>reLYhxRw-4kvwmJny0z0wRV{n0k?;{9Up_4@8UJxY
zMZivtMaF401?CY#Zw%sMm}G@q;Z_K8NJt3L19Lqi4&`?@$n($2qUF*EAx<ZRZ0X!p
zG1%~*9dE;wDwA411t;eaB%yIsEVQ&z$?|FaD>IuqKw(p0xSo|+)vPCY627u_-12K(
z;J43<;;?!AYw6jVkx%uhliiCKdEps10z;?3lFYfmuqIp(5DkZHgsd(uQu(RLkZR`X
z7J1eNdQhg0&Ww+{#u#)@37s1ZURBP#JS?N~_tMj`ceK-bYVR6#G7i{)#Y%tk=8c;Z
z#4j(p;)EQrJ~2Q04i^dIj~X(hnd-Zu<UjBhYSwJ#zXKPOFW5u4sIE0WPP6bW@tbC4
zfaw~Vn9x460yGY=ptrn~0el|(Va|)G&l8xqQxWM&^)irOLfJ|I{qBPY%2^f)sYtfP
z@2?RR%epE{f({H~TrS}r3^hWrj9#McH8I9Eu<hP++{_6O=qUM}z(6t0Y&hRw7i&I&
zY0!fa04KL^>j0|MrdkdQ#1p|MHJFhJopzd`>jCz|-@n#xTPHll=+!SZ(HL5o-7$u|
z$n(wlFSNcUMw@OU2;u$6r6*Q39|aDFf;2Sdx^P`%<0@&@$*)nsnYr)0g2`hIH8rKy
z8p!8*dgOm-LdJxW^B$N)1Lj);SUNn1!)Om|wL<9}1k*Kd&6gYFfV$b;@vr27DVB>D
zegp2k75>|Rji<yh%5(yGo7v~+=z0HRT^uLod-ue$ewchl-l*LCP6rw9HAMESgL56!
z!Q|6`%?Ffh#uLS{#&3aA+U-G~GJpCngUwAvS)Vlc$XU^L<C5nG0pjuDuSbH7b&~-q
zU?)GmZupb(rB{zdg@@mZFmWVW1W>)`525CFI3X?JC;7}+ru{45z}?NjzJNne=h=Dh
zY{V1B`D0?-gg>)S2_e6oo^5dbNq=_uI}68}sc}CkGllfp)_$)jcPJzl>Fk9S{#crV
z)ssqY6fD-f*N_vFx(x7AmpN`5f?%m)I_AxrL|&!T4?kl!wuT)&Ox%j(+Ac6U8*Kcm
zz6l{J|EX&PkjcEoirx9yI(~9IQfRjKE8pNLRK6nXa{~TyyzsSi{1tti=LL?x<5ry}
zi%ve3{q}o=?BHVm@`GUwp2)+ytGC<=P{)-M#eQXL!EdnUp(2*{S3v_2VZH=y3gFgC
zd_L;Np#EZG+<6J+=U{H1S_Ld6!bsLTZvw_Er=mwR%R<kjO#YpMuqO7Mo(VVeH)#CQ
z5!nfU)mUFC;;jjlVYD4o1VtaXPBfAE`&+*8I*$0O-f=-*DG7OAA$<EFgykx~dr$Y)
zI+tzL@AB*`#jr&fgKyB#U}9WzLs};EvlJG-^k<3b6j-K3>MAOUKV*#zb`REu(SI+^
z0Ftg^N836gORxRI(-f&Nz*7w#{vs&`b`&lKNu?mtW8kIlUvX;~&p9^ryt<ALvbkV_
zq$ug=%ubI*AxZl6OV5jw;$a1Lr102H-IS2t+skzuRijerE!$H-wr6R?G0cf&sGN5F
zPo_ix^X@`IT5IbfZzLFXu1?t4IO%o8JpuNbBv_}#WYu`hvkGg{wp8V;iy{{gVY;BB
zL=lWZPmDO3Jkw)t?XcLNs!5bok|YIrihkpA!o?akxbH@P^5LynfxH4DqcMs$8=I9v
zgVqP~N*|CZNlACnt1vg+Kl!!K0SzU~U&-+eF_1D4yW2Q*_F+42%mS9FGiAV*WNJtH
znP(53=c_Q2-$a&bHwPbI^zU?(83-{f4T`#qQTfL{f1VhvbmmB-ZSW)@AlO|+R#Ug2
z+n1EJwR4*XXpSyoMf;rfAY^ekQTg_P<-{Wk8>97+mq6SQcO{0qsw$_2GDed0#!Sl8
zl-&6$p+Lh_(N&>Y7Abrjwz%wXZEaFv+&%aiV-~(?BQW3j?1aVirK`qhv+zT-Esc(w
zG-NGIn54XiG$VmN3~8#}lMhi_){Cp%HRP3oktu=XL9aok*LWq2NkI5rn6NpcmQ#%S
zNmPJr7fh*#5z@m+t#1d9L#OhKNx?7;d*rwwP5Cs0_cl}>Z#t-#mY<?ynPw=v1}Wk#
z2vuIQUmIevT^Q?56aa061Mq)gB<Ekz0yV1fF&7()DP4!u2F27V=lK>q^=~c%o?BH_
zXXHwvPI4@O(6y|4*W91M5p3jSTJ-b{+O9^i%l2X{pci=e>$*&fpq0Qt?~l)>0{?IE
z^gEkl<W=?R%RC6i_POV@DHy?c8S*v$+7{KE?)`Aa%9+>=WUlG8u>!~K#{s7dSHYsp
z=yfL;U}7fvk!9K0B~~BX5Rkh3{6I!#Xma2uNebo#4`+)cZKlR3<PCZRH&+qfOu!|6
z242mWmw;!2how5Em@^^w?&yom;KR%AHs3w#8$o?KTb}*qoz^9z6(KmTSJ8qMx26>4
zU|N}xF+b+)zFe(AOXf;fJQ?C%wt#xT{iuO2r~Q(NCNS2n#JfMYJZk^0X!rWv*n7@L
zSb!4|PIuTGPBQ8B7k-GR-ay-*0c=2Us&0ao+8Xs!x6CxeLQ7zJ9zF}?TI~mV7W(m>
zK}EYNz`@KwNB16npG`tnH$kkou5^&X{BX(>0EO^*7@;rBk1|U~W?&G|sB+;Vb)g-D
znxtWSgF!1J<Nl@Pv$Zv`>K%DxnnzFs?!5Qae#6d&&|>VHBnZz@B9<@y@S#`Mx}pM~
zu^G`#Qc{29*qc!1uc@h3Oc&!~`sBD{`IRFxlYh|I1+al?OOGND4@7hz=jqezu@`rd
z{Zmo|&cCLEB>O08MJSARAIK#fHkV=gISh_vw%RtOy}f{n%9$_T$?D))XG*LaNC@#E
zR>LW9vJ;XxOFetFgRID`8nRHGEpnA>+1a}`KMXA@kjdOyaoy0#5jHz6xD+aCV%xq!
zu|cCLopNAh--o%Kc*E&qo)7ROI@%%lUe^FcbF#O<Y35a`y7UbQ2Hjwm|NNaWTwF9#
z5?7MHP$@=^6+Zdry_L>)qoYR&tMHZ5`+kV|X>)u9!@ID5kN~Vo_BN(W55=BcZ?k%_
zR%$cPeeVIJM*I6--jA)U@ITtxf8r8$p{}DQUqH7OKBMu+qdS}^vxRD74OML;S*N|7
z1!bMqWsb6NWS5<8;@=;J5~%E_Uun@=*U;aS3cK8`(<Zk=%aeJ>@n#buy@CjpT5*AX
zGQ@X$gV4>%p|2z_@=o2nJK<)&U%Mg})?}83yN>QGWNiVRu<eo-%+0@`2Ze7qUu_L1
z>C1ig2~byX!LcsrfSICa-egx9JS2RCTh<)??OXBB_p&8L=V>TjE#Lyi#$D;@F^Nt^
zK7{ofT5dEP26#d;2=bKSmvM=pT}UrnsQ$b?>XB5~h-Dl4c1!31xVRYDb%^HSR-Eqs
z@pe5X=IjUe#P9;!#iwq=Zy^^sZ#B|hmX&oUC;j-dhQtUE%mIJ@{VRZTr1e{VbW1FG
z1w8Vw(#Ro1Wv6Ql*d+Xlc4ZX!QkAQ1XX;S@=_P2)`D16uR8KGO1JD`_2PC#guzeOf
zzQ1`IPP-chhsTDn3Q*tw^dUrHUXtnN$<c;qG~5bAnuPla8@6r~4@`IyKGMVG+mHDA
zRTjizVfV_I3q!3a^#UP>Ma~(;bG_o8<Uv3OzcZiUcnwQDuW|VzB6dbbA+gCL9?`@!
z7{>m`b$Q2&G!Bmjun2oZ`ul5zGaNOnou^vo0kaLWAOzTr{wKu5-uM-Z%VduqlZg2r
zHxak~ZgkEF*-(Qltzkz0X?Qr1*Zy|wRTsH0ZFitbIO$T*A@fh~WA+9bB2YK!sZFfP
zFbfpx)ks`U?!KR;T{%6J_Y8at^CWEGzrw9RVwpQ&X8vN{Mvyn22LOMSA^y!4@qbVC
z?tH)eBM;l8Xk}+)v}g&X0c`n1zJYXA=NhCTQ2XK2ZwdsbuWx7wnJuDx@y!#;U4TP@
zO@t6vZDp$7+1ymk*O$z`Y_~Qvy|~Ena-|O#zMPzaEn=Jq0tdL`Q0bQ`18hV~V<Q!4
zlm7q%CqcFHi_Fz<_Xt|+KvM+K2yA0SRdX|rwS^t#-+`m71h_)(c$GrjDTEBowMh^V
z%jo}m9WB&@03Lz_XPIndXy~D+DBjKu=<UeF@Rr4qvD+@R0Wt_#z~lHy8YS?oB>}a7
zBDE<nPoun6Z1<=mZklcm&3Jt)<D(P=zkm%%a98~z6A45b^L+GBSA##0@Ysd&TscVv
zp}s3uuexn<{G3JrMkueA;fJD|GTdOd$;G~u7InphE3h_%W>C}pyE8E$Xs!cZZ!}`$
zx^R=|qinr7_dK`}P>W-!m|Of)tkS$i2_5u7ASiHjZ~!8V))T;e09HeBZx`SpPeD8K
zM^n+FhGk_xP_PNV<vl+SaK7er?B;1B!eL>V@L5QA6=@N4i!Yv90uT>u{Co$X*yiu#
z0jSK){oL=J7rD?J<XGlul|Y#46uA8W$GBYW0HiKZPiu{a42JB8>dGVM^OFi;bn|~^
zBc>Anm}swh(d@O}K?42xE9WB(Xwx(G=F+X^<T9hh76#DT-bC!QR3uCir=_Bf=mLR*
zy43Z`J>c4ULG!!YJ0mpj(B+U5>BRf@U%+nrtGj%@F9_)W?igN0Kd*;@^+4>d|4=_U
zub!RPOu#6@Z96X>I3TRdd3j|<=uZI}s&0R>2U3LT`VVis8H)_d9QuUi{dV@R=Jh+|
zu-fj#dy_75AOXjm_j^H#4VEAI+^h09$oihM85#al2l#)qglBRQVI2|8-#<l+f4`F6
z-s%Y?5q!>=+47WK<KvUHwK-1`F#+foe7(l=uW^)?Uph86HlH5wX=!q!<VnB7AkT-K
zzPNankMATQ!GHR9RWjL0PO5A4t92x++3CR3>~;IF6_7W*^V{Wk+~FH4exAc>P{knV
zQNCWBk<kgD^*=9u=XHQ6HY`0_2qJi?2yKAu%l&R$#4%(2T4sCrF8Ym=+&g+Oxe(R=
z&IE-d|0-LYvVQj*CVZ3wEcjEu7encot1IfRM?uX&m4X8?Rm1OVk@^XGgL$Ni86y59
zY^@w}Q;60JG@KpeFasf^07(OqNZKCH_D}YBGasx6MTz(vHZPnzROxrf1^M|<!dJ}E
zmiGJdI%AaxYQ*e>XWrLF*M|!tB*Nxu`ZEScMgw3qEKF7VQIY%#=Phww<8)-G2@|qA
zwRc;`VRM{vVRUlw*e{PAkPLYHZpA5QWT~?YzI%U_^5gQ_*a}EWonN0ikR?!wMRT_U
ziwWL+yw<Q!w{G-_GUC&yTyc;?NCPV)&v7jt2%wIPVx4Z*`YasDOrKv(oFpj*D#4TD
zwMFlo;Nj7?9iR3n=hHE;qyR~~w6LI1zQfVubmo#R%fo|wjf}rJQ`UL&HUq-TJ759F
zURpoX0)e+<i9QF1t*Z(<X<Gys%Zry!S69?k5|<a|Zlo(Co81SQQU^X-+Lp0c8rXDD
zUbsKu3H^$&qFv(IiAN8HWF{YZBVU6$r6W)@1FJZw0?f_n8}R<QB^AidI}T6Zc{WtW
zuYFqXU6jj4+m5?~U<iPk_blH~x)Kw42f<>^x`_9Ols~CB++xVt&~QvXAT!X117Q8q
zpK>)Mo-iEc2XV3<2%+wBIqgvYQ=$z#yMl>Zcy}t9u^N=J*Pec#6#f%%C^?db%0#k`
zPPi3LYjksQy}X7M5U05nrv@@+@sMZ4JdrB|`73ZNK_WqHz#k7LGDM^ss7kY8vTJyM
zcAl1rI`zF@`_3hbkJD3|+7eCguE&@{CP1&kYs~10P__HukuVQw_ukG9#oIe6KraOa
zE;_Co3}D^KI4nL(P6h)3EHST4vMEg+DS4qg5%c)>8J@88nmQjg33hJoau^B#yN9r5
zzB_UL;EZ0bbAPL?l}&Ukb#%JAvmtN{MaPGT?iwOry8hP$8Q41q>>E=xpfUv04yKfN
zVNyy}E(fM^Ns8H;!_bZ(_9X@?1c0*IP^&|&3JATM;*HZ3(}L*-E1tSPkmZ6ABp`xe
z3R7g%EVY8iC@t6bna?@=&zxI~Oxc&@<UG}@|8ZRIl`fPbRsp$40>=daqT46t>rh}r
z5Db8H32@fTDiHfof!602+C=Hcj~`g6K73dQ7~XfS5-?H00g(e#5F6|2ZoB|K4QH_r
z_$)xG4dM<#bsxwnDGdNvm#@iTJ`}id5prozu?`WrO;5iDs7MM!jA{Tw0>LT}fkRVD
zu8Rf2v9P?(k1Tg-Xng8o?%1UFWotRZWJXTaA>`0mKS|{Pfc^S1GW>7}KwHCTm3#vN
zs0xK4s4Z_>2ehyylo)}W5WT!x2AJ-^QG!6AWE(!PYa3`nBkd!rfv=}$Es$?)liodR
zuC3Jro+O)M38L#!Ece0DLqwdf2|MKgz=-pFao};GG^mlM`-XS~hAL8X<&p;Vvx<S<
z6;VBX9Nn^q&-IdzkM9<}DuPB9Zi=D4e&DFWbb!EVZb?aizkm3V4GT3LFD)(YAj_d>
z0`USchpR(kSw<ih!<*5+0#!((#j%=n-Fjc>@wGUs0yvl6jVpUAE(oJdsDud_bPBB|
z`gO}}=J%<kL-s-L{XU`JkDJl}sD&w|=9gL$**6i)PtXe-uLb;|w_v>f{B3)X{`k3z
zuvMC6KcR3|U^|ihV}2eF2S><4_DUJ4^~;z20xl=JgV)AzAM$6x<O$tU3+fV=^^6y6
zt<2}+weTogC}k#Uq1P!wQv-DNyv7ueA^~$%fyGnIK#0#4xpiPt4=$S@Bn2T2EjQ`C
zccO6N>PHaZhNrO02TcDTSM^l~_koEi6ro)Pyvp(sbA}XG95-Ra$0jYcx;BxwCM3nK
zZdXwiBBDexyS)yc4KHu+n=dw@l{)Vv$7t8o&MW|HfJeQ9Wst-=q}fQ!?@|ibZ=0fb
z$bOB(!)hfh;lw<4Z4C|g1}}<uW2AxP2<%$G|6jU%843at#&1}ukxxX*m=B-`YeQlS
zOk2KqbS>_*AbJL)LMZE7eAIMm0|lqk2!*bft74k6SGc^uc%-=_G_+xN-+H0{1`V$R
z1fq=J<l8PjhEsS_3pV1cn=2&RcDfLzC~bmQJ?cHh61t5+vpYi|D3C5f6S;}aG)yY_
z=v_LPA?YVAMSneZP>K;elZ%mP!YvcW3N*W^maW(D!N<=p%r|sZh=9M0?+dblH!$!W
zzJ)EmKK_-%nhy_#8Y$lToc;vbx-P^SCB2D*L;n5@YRGVkfB}UD29QpD#bHnKA1)<o
z=C$IiM~DD$dA`Pm3h=;77_%_i1G2~`K58(oV?)AhHbC$0?tW4Wxxgkzgm5IlTUT?c
zvq#V1|G@nAk_o>Hmw<pK#p|?4o1C1ytYdZRslz4!vq2dI5~D6YXz6Z%^Yd9v3k(XH
z{_!K*+3^52!<-dZt^BcSNKFB_{IanrL<C^u6*;l5JpyrO%NApLPsrKH$%Q(w$92k}
zFb*MbzOA-{)tV)OIWSoYudFts9?bxR%n*%_b$ExC0LNV1K;``tpjl&m4&)C2qks1N
zIfxiL+OV|&wijfLU`B;*Z{9{W8~G(wgUxQr*Hh3cyZiP4hPQT&T7lGsqQresNG;)T
zkgIjO9N1tD7b~=%Fy03rE2t^^KBxLrJ+T{dB3VWm5HCqTO5j}$UymD&un8sIIr;=i
z1de)qQ@^iNa`2+V#Kgp51iQs>J_O^G8$p`4*n^=SQdVa+US6bG-E<ggWKo{O!^sH>
zOd=&@goG>e^N6mu<2N9aMrp^NaNW})h@w|1QU~%kr;}SpJIjervziKuz^qjr|7?Y`
z7_!*xrA(bKH~%g4eHHBHW|Hx=5zyO$-a{sj;LR*z6xO+0+79tBS4HvbMbKg!P$R7;
z4j#`de5rl_>_1z20SGmo>uguc^25_sZoWcLcyCWID!X-m)A&w)%Z7#o2PY5G7IA@S
z!SK}6!(a+pT2tttA5?wCM$-iXhkVUlOaG=0N21anD}nxies58PQJ!YH`40~nFOx-o
zE=T8wBOW3xe8Tc`8k#4t=QX-C{(Ts-Ai=SK{ocyJzZz#}z5kDK_CMI;`y4F)7J8~u
zt=WH!6QgoaWgl4`hw=$}mvLOH7Kx4pTk5FJV^4|{;(KTA_$}e)@AF6JDP83l9y|&P
z_)7nY%+I_^JA1D+!u~lS0zTD#;(YmqNDq}nxpAKLh3}@uKLkEi$0jR=4}^sSd`jKl
z&H1n)g6dif^{dm16}()RgHEHBrT5}tB9F+J2U9}2J9_YkKW3+oR8Rgx6RpmOY@xU}
zOlM<frA<j5H)<MUx89$99cMbzDiyvNcA@|HX#08&(KTV`e(G|9+=Jb$1Gj?^w&;ZU
z*ImH_cWxTLra*b5_Z*vhVx?PHxFyiS8(VPn>_-(U>UK=kvOxLH=x#Zxi22)qrvij_
z^Eo+*>i*fU=6UGU;_k~k<J_=W5#s-j|59E5)`y?k9IZ)$qH{y{=5E;Dp|KJcq70^3
zV&1tUn#%lrw&cBT=xtdS#hZGfsgxV~@sz0d4yv@Su3pKKy2{4nsc<PQJ=Ky)Rylo9
zpJI@x(L6#-nzFEu4zs{V?}b|`<%3EEg*gTJkNrxGxfu_Am_y39Uei`}p+CCXveoaV
z?Gnb5Q?DX<@@XbxOrN1jYNcRyl|{|QrM;<#G=S+|t&|JPO7H9{!1eK-YF59UeP6@Z
zFz6Dmv5J~0FKoh|x*&!X5Hl{Sjg;5^vg&=c*`m&idOj+3_Ht4~QkFY9kBfI(vgOdx
zy(8@WlvOjZ2JJ9j4+WHoTXs?kI{H&Q`y6f~d7qTv)ss~(tG9|T6uxR<;v#*NN_-P{
z#X|}C>9Xa~7`K;nOCC#3vsq+i!MzFepY`cH`65FqvYb@rC`wf&FW)rDs!F4pNK`cA
zU0X=VCCSWmt0_e@B8aWqC&#UwX$g(Lf|I0&fg=`T7vJ-1ux+*C$)%$C80jLfI|){9
zOVf$3xiduNEE4;<-)lG079|qT9X8tW1|EOoc}7jb;=lQeg`V=|V#@QLA^*yn>)H({
zIemoMhEcT%%T@x~^ZrItN;aCc8A(6B8yKU7TlYj=&v9$YxpQdA<Jy7}b@9OBEpyK$
z%_-uHW}C~6AMVcLn{D&{lp=hV+avX3>1y(=$0N8!%?z`jACx#gSH!MSIuy73c>kr6
zXfpquCKl(SZN*^IB7vgQWd|f3uEpfT6oy@#Z)pOV?uL0+qXc&a%hIoU*Hg1G$|=YX
ze#=$8e2u;CLUsJr@DH-7uPJMz*>m;tz3y;d+>c}rTz3$DY(%VL!##-K>ga$<Mc94Y
z#c*AZhw)`CfoHJ#WW$#AsHEUorVqR14^pnOHB<MFmwioKja-aFs*c$5A_fd&t{w3n
zw?Y+^e@#9>(RjU|D(2WUPgwa_UMOEH9XDX^6VrxJNBfHrY_%?zl8*{06TEG;+ad8^
zPzu6u+D5InQZrJX<;{%w*KS?)r1zD*xa4}}5)a3){?8%R<+qO>lV^#n*L4kcCL<+3
zp*~T2L^b-Hg=_a3Wkfr@S^NjNkmZfb4CM=VNNz4TQzcD~y<(Nb3{87M-;B>_klUwc
zbK9z|Q@ZOS``uU0@vE#-qZ?fugh%molw23fgUQXN>6WBc-rO6l+m<NIY~~M+W;I%u
zytZzCuhaZ-$2`)3e66-MKfCjS$o=Bh*K3hq%{#2)+uzqCv!^uh+w!ZSF7_pieK|3_
zuh8psC+eao^P$lFX3MmzWoVB${h$7_B|`GEQ8<=ypqEco%Om7^yUrA$&^j7U%6k9X
ztee=r;JDwd)Ty>Io?>=o>kesKd#CM&)SzY-4zrYHc_ON=n+4x@+PFNQ`36Qlo}2Yz
zo(!IIrJhzzd+$2Z?>aK?y000#_k=o=rB|Unkxex!9e?p|fhp$K;92z{mUb%}^JhmR
zm0o;ui3?k8wf-L@n-7{y#2b?4O1DZyaKs)f*;s4JZ%RF?OHn=vZB&%|K0JLP!z@jQ
zm}~D_$e|BspVP7mXHMl81^hb~jv2<sqy<i0zdk)hKJ$CA`LCP-8_WOh4A}4gd(I$B
zRXTu}3cIQ^XGtIz-*<-6Pb>=Pa3*Kg+JDmc*f!IuKmEC<ASiTMWMi#?e~0<n6-C!w
zi?ZXQruo?)JB3G9{lK2;XOqk8s&Rus;!r90UY6Qt`R}GhynIYgf603QmKIAYjh)Z`
z)S0k-T8K{ehVMiuH!r7R(5X~s<bcu_=_A_b+%;X73vFcRxYr`3dh~pH%3T=S(*y(f
zTtKxbgkG3{EY|MsD!1xud8TTC4Q;#cX@LW)WKoQtp4w3;M?EAp*{WsgL<?^f`1+J?
z{4i&1SB}xy*g`s&uo~(0g`N(b{sG?$@!z6^^vHK#KMrG^5&U^4-e`#=Nc_{{439P{
z*@q?UcRkp?5{H6hDpApmQlfE$wV~c=yc->*69veAVog(Mvro<(X1<ztXT+<`I(oSn
zH5S+etu)E^yE^E*-3aie@m0*8z*c!qrI(K;gg(L~Z>(3D+<g_b)n0zYWqvzdN|3T5
zKhaAr@P%zm!pLUuoauUg%!)@K|A&;|x5v!Yq@uUcNe&o3k`YAGtrDTT`Ll=aEYKx{
z?@<z9VOMetM=?2H)gm8#ufpcs9W#e{sQd#ni@{99#LG5)GHVCbxBYRR*u2-}>enxA
zng{8LKFi&=P<v3-oX@6YHthH%neUf&G=;A^_bpTUayb`;VUvN;41qb4FXLphMn7$}
z%sH}?noL#Gb!*z3>zDlZT5k)A)T3~I*_34QPul943^X7o^%#Dqgql{7PJpBD(_`-M
zw!2=@@ZB_{N@qgs&DG(TDf-!JR6f%5#wz^#ngt`oSvr+`y#}9L*!hvaTIw^54d0AB
zCN22FzV+kd5v};5lUSbb_i9TA`xD32^-sKpWQ6P(HK<no_BIX1fu-$)l-xsN(vA&n
zGtxcc__JJ2GuMfBj!jRLJT-!<GcoK$T^3bUxL7xe2BX^3e>`?EOZwh+?J&KXCL#3V
zeO<rKMgK%X1M1}s4c?8#E<zKRJv&C6WZhN5PX%-yB31qem$;uRMXE5cBv5)ZH9S3O
za-G@PifgZ0Do__&%4%s^ukg~$p=jqnwHoM~Kuhnrn!I#ofZzP+q~!%mRRc{21D}_t
zCxCP~WKG241#gMDmB_TyzLq)(oyjbXj3rozOX%`yUR8IK(Nx^#dq7hVj#sofJEM(6
z^!i<G<Fp^zNrXm?5UQS1_Hvr<pp-F|SJY+={p6WuRsBe%K+%2dJK@LQ>&OlxKBRs+
zXmgbh3?)cTMj1B{qomQAxio?<jvP^1_9XUavL^d@%ohfh?1tW{U|AWyAj5Y*3S1)=
z-vwmnxy2AoHgE)|n6Sm5n7ibayg%DjtlTwE@2p_1QRB<t3*8egA7M55%B|HrllR<o
zb?1gk>K*yRea_AvKZ_sz{4H0XRa)>&JK1{<(GziW42tnh%B6p~$z5WdU;f@b+0LC^
zx?~tjWTHT;1OwCA^XErAv+yvEFu_=!krmNwzOW$IV$0m9#oNT5!_J;K!_CPBD~w6l
zWpS+bPa5QY=CUNY9MDDx7;oW5W+qn6Og=o6A-P+0d~x|~veoj2j?|<8%bg1h#6_>s
zo={84)xR@%x)ky;CsNn^;8p%ND>Ti49&~E2IYqr(#d>ey((SIB;nu#a2FrN&Wy|6;
z(p5u;6rICS^lfG$?}`Zt5}XuZhbX_YfA4{hJ5ZeILYSI`^<g30NF?KwC$+)+S>q>N
zX30rCUiT1|3+sj1y*O+J7AbSgB!WLaZSVIh`##sdNF7w;92jPKRwwR)CTHi(Jj85m
zY{hyNTdUikycjQ|ev7NK*6lf=nEj$gE?<1d&B1GCIL&mL8lo15gqimyuaA^FOX0U2
zb?r`Hl3s1QMj$n*EMHZh^t@rj&QF#<tH#&TLbmz7W$I^#jfKU?H1o(txzEKqR=9`h
z400~=_vjz{Tu81KkYuA8xJ^7PFD~;s>E5efv$Ii2Q9EoV<67HkGxxOeh|4pcnHAj8
zS`Dsi){CIm-1~AkQX#FMu{+`)@*`wYGsdBM?K*2L*ZmZ;q!gdkft0ng3v7JNy#LCj
zu>O-v`5%VMf><FZ=f4M+MOMXaK06b~c@h}obtf)V;hR{A&aK^tYYjHPuKw~!8Opm8
zuCVsj{^nXjs*%c?uhXk5SVDE)*Kkb4xvIpEIt|ViZlR(I2}?Lu$GSQ^_+$`LvBwrW
z*s<1TE{TdHg#BOpSCOT@t_}(@;s)++(02e)t?z;>hfw%<c1cqyltS<dnIe}aR0H;V
zWd+iHMjpakJ9eJ@AQVQf;ZCIsFS+y9AS+?2SYZ=CZ1B#Qy6tiSxkd<ufGFF;6PptS
zQ@Cl!3Xj<9n~)W@FuX4H1uD<aYQGGR6qgiR%ujFi)_97ep`m>f@}h%xC!l;&N>+LH
zY&wt}_Zo&1j*VZSx$r3^g|fzI=fJ=Tr2130zvB@jAwNNyO1Cj<`v|Q187MR~MSZE(
zoh>bXH{GqUr_nGyJmFWB)WSY~Y|?#+<9)lJFY8EJ@SE5_zgXmQz|CIdn!qU`5f!b4
z{Le4-r*t|9ci>!aKOz!tauUu)N$9GpyF&c1)WcJL4rYo$iTffZ<~TjQrn$NL-V)0l
z`ZX`pj#2Lq*HkjpLEx24z|H07oFq`jw(9{b3;6tq;cuXRF<NK><=e!pEUkhsgyiPJ
z#3%<rfjC}mwDm#?3g;`3dh2<^t~Z5Aw?Jbs)A3c{4Q-qwGIFe!FWia?UjWbI>jWZE
zcLZ#8Fmt%{#c~uVxw#$uyh1?alPnb$5wQm}=+HI<@;Gj2(K0iG$mKFHylsMbBrwUb
zusl&v$WTb}4+{(X`juyV2~kzqkTsF8|Dq2=VkZF2^2J;2>-*(vaqC3aauVnKlEW5n
z8$4X>xV+M@At_<4X0VR;zMqx%IDMx4#ny}g8=JW6360kp6-A-NxRQA>Tq1g1TOv9<
zgBb#Gay$}}c4q(Yt(~0#&xqd;ihe3D=QbdILn2B#E-{(F!~7xfO!zG5l*j!-9Db$4
zmPh>xwXbJXlo?ze2Id{jnwl0{tA)1u-4%vguI-W$oy7jLP-nyIPJnt35&nx~_2zk|
zde{R_JNRgrck)87GsxYLFXR@ZGr*WNX7^Jbz;75H9bH{of?rxCB_$2DE=&;5aE2GJ
z1S(%nM)`t@>o(BylgVtXE{n{~I4m|63syd*L&>%OW1Q!~S3Gi2T`pDmf`#FWmmBCl
zBiQyD8Zf=|!w(X7b}B?fYQG7ENzp^mJJdQ<8A|CGv()vanvAuaQXKrVOE=4-tEBf(
zF)<DC#m_j{*c8H_mV#i%&Ju_zjiZ(cN^p}qdhz!4M-L}SI=8G||B#4`TvH=uYc-r-
zRAjo2hb3e<ZS@(XKX0_gioEqVHt|VYsj{}-LP8F^mxnSDoH(b8FTQG4IMBeH4Fh+;
zEQv9TWZC@W9_xoheHuo_aFJC~0XJT|Ov~OcT2KWXO>At+%?&v=LR{l*PE|?F@^UAf
zpF-C$3GrnT%Y^fep@Mr<?5_1V?W#O3u91}-`E}_u4V&-dj{*NeIkf(8Gn}}3)JNXV
z?JAsdtK48>^P#S+OtJ2MuVc#lj`i+*40x@vLZ0_<Z2|(ty^yZ=_+UR5Rz^9czI*qU
zRqhBbdM)RZk~!)TMD;j*)Mt}{E6jC>4^F;^XV_d^)^lfV2FZ_}{K*eimgqq^^>YGe
z2ner1fo*C^3WV2{y&Q0_OGylr2&8wlVeHnVE4jJx!_B7SM_e~N22^_<nwQdybad#C
z|GGp`Vq)f3mu9lDkwYB!n+|@R>R3AUGjv*jp0yZI9hzQxlnb&C_tW6st*q|h^!I2a
zHq^(9P<eWW;lP!qYS;kL<J4sE>R=98ZK-_8oX*|z)%G{lK>2fZ`4Iy;ZY2uNnm~xI
zLBACLBVaAWYyVd0_Dn}-3Z!ip7jKc0c65T=AfM>~2Gc6CVZJKkj(d&Jli(H!D>r-0
zQiQddigU5aEiEm5lK8-@4*l884f6GUCLnnw<Fb))b8`SW<o3cFj`zuJIf>DPe801r
z-Gq4zKf<DcE|fcSWUrsvf2y}ZM?Vrl_5Ehg_V~fw8ta!A&%gGwZJ@7qNx?9qaHSYJ
z3JWELf3}&i`SP&sN8V#?b-wttq1svzY3cW0%+y8VQOm~sfAl51LBHJA_L=i@v6z?^
z*XM<pn7hy?_>xJWF(KyRiP$zMm@Qn$X%>_I%n$4o6?jh9`%Je{A`?<L{DpG~V@+#z
z9O6yJ+S)!0C@i759=MWF88fi*@o9Cx7mO^KYau~pbdZJ#^2<I~&fP+8sJy(qH&$%R
zG$$e9p|MF;SHG5y@z9e;kH8+u4WBVilQP}+18qa~<@c}Pam@Q8-KwCZioII{8xR{@
z!Te{>W7b`8OTg0l6M7CXtORA{xK-98t{LYFY9}0QTj}hvps8aG8#KP_#PA86?5Vh8
z+gu99Z2be9(|pTs7eJ2fN=axXvB2p#8MR};El0*n_;74JJW*7v<Rm#A);{VyjpUtR
z{$4WQQT9MbP*gxXw-H_Ra<VrCpLTjO3W`4xldX;62T+i%L3py%OyZ<g3Sng_A5YLa
z4N?xcMU&oUK?R))n(c7S$nnR_e=p>^Eprrhs|{HzMU>}vKJ6tZkcT!urm?4RfP;5z
zL3QJufj`n)g#QG`mQIMgJR{N6k7mQn#T9Aub|3~ueT*Na(>hh4Oa{~0Q<P3IaBbbK
z3Ccs<oO4fjY=BNKAo)!Q?Vic5!*sy;&EveTsiLBC%upufk0joJILhbV_h$r#Fb0Yw
z?!}r2FC-qWC&H2wv!bWs9(VO>m41JlN2_dwI{DfBtqGA26{svmF(&#Z%^O?tKP0R`
zn)HAS)gPST@2^8+R6#WnlRzRzN3p*9FeLC^1(o~<Dt#n!?FwhBQ0q1ZE=v6W@B0tt
z&ZMKUU4_?BmW7)*Cln^Iu3w+X9*5dA5Cf}rBV)&*+|Cm`4*?WWFfAyOILMd1#_myk
z#1;@P_2R@$sI5&XsN#=kO;!>xEUf`bka;^_X?=4zWWor0_XN}My6!`X%y(|AH~hP!
zdER<^gN`?hLODA)sH&;4WF@|cwum1E&do>Dse(nK=br>GBxEJGxh0KD%jp*AMu*0)
z&Vhl3cKwRDZwK_yL8AS}MT(m!=f2XWwPvq^Mr=}2Pn-SlIRWQYWMmNHwJy^H?G6S8
z22imszcZA72#SjT_{`65HQiqH*&dF?SHfaq9@BZtw646vxl4_^yR+jpEkQtAK+_SJ
zH%-g=&gu$0sd8TOx{raGv<o_ARyQW|qaywnXKx)<)%t~vD%~I{xTOT?Mx>-mx>G_#
zTG^BW3aA?-rKAxgq`Rb~yQLM5l$4+XB7#WVxzY1Gj&Z*`#y9R5{)Kz5z1F+lcg8cH
zIiG}}@Mj$uzaNkaaTpy{yewDdC~BPZZD&WVx@dZ4W~#s-%#a@XD6kidd%KMpZ=sLz
zX&6<yvl9CArKGENEPh+f#p>$nerJ<7qzta02%K`}8-)j#VBEei6k}j~X;*i5LY+L+
zg+WSoO-dLQBGx~j9xJy0-2n!=dT>}&^}O#`f(g<C{ciE?LCbFu+dPHPBfr|N;}v?r
z#u*(T)*61=2H`v>E2}c7)<cPAlxjhl{2v(OYB3C2Z(`is0}~UTP#2m{4;S4xfChJ@
zySv`^g)OSUOU<`;K^74FuL?H2%3x$MGxWXRgj(}}ctZc~-X33dJ51+&O3E-)G{$ug
zbO(^y@rYNQebdt*b2wA*{JLs&k+`U+J;-0aE`nxI?MH^Y?z~aPj7V*&wm3rp5Cwq>
za#8|>Pn%G<&iA#y@^*GsN=WGW*Ed$JPe4NV5vpQo9;#!Y6&A)U!xj16T4w8YRz!s6
zF3FmV_2U*<R=g<~k}cN*1;xY!2nNtXt+RtuQ_xB-{`Bcn7~T(@!pw-*wu5x;?mNdA
z_5LS*C4IpboZ8aa4+cPryyE?5TvGa_f-?J_5T6a`gMos~dr(w^QYL6s&*bZ|L0dUp
z;degsT(h-%2fSl;S7>4e4AvUrS}7hzY1FE2geK63x}gO)2Go)TppjMW3p6{@hc*MR
zVeEzE?LNw63xyAP2S<BzFbMd8zDz8kB4-v1!d@9y(E&;xxPK1aJ0pbQ1Yfj<<*cn2
zfxLbdl$bzE9yY?3i^QmB{cR{Y3X4Ju5SNdQ!9A)Ng_A8T1{;bJ-6Hb!jSZj<@^8S1
zgEkEqTu+alNM3oo+)vY|vCt%JPg5LvZv^_b$NH?QwVZMg8q6`nmz3|L3dyDw4LEN=
zcc=tN#9Y=96&L?HQxhqc0=I1^{^{9y0c)7jFfatS^7if9bSeg<HVy)MU0)$<SSinb
zQYBf9#)PAx(zT#USX)=;4RvPID*KT-ce|&qFsnajx2dB;q5c?@A_xh?fRs2=C9{Ls
z?9L%-oQ(i|F98vek|lL+c8s{7tbOk#H+Ofckx>{E4lE_8g+GtKyL1Ohc0E1de!>>S
z*Kq%<mfKJw1qq!Om+!5x27ZLe<)EGt1N!SLqVwsLVk)KTF!ecNKjCF^*p{SQ)sV^p
z`>yv=!FM)LQ+nT(LmV`g)Jb>CYQBGg=p*t4kOTdp-$8*+G?+!(`!Q;CaURO>5U+JK
zlgOTnn%Oke)@DS1U*-N+h4IqOy!^VSQA8Q<_tWUk6McGD?r&>*d65{QKz;_dFcnp0
zsHwLWLcvCMNMMAuUoar)W{{DktaYC5^}@AQ-KCDIYRA-yDi5l;70kyd6^!|GBGFn9
z=!5y;oO9fcoSm(LitXrDv6z{N(Ntw#PR`^`$Ll?F%R?g;l`GG3G$ex8@o@kBrV8P?
z!O#Msuw+$N?QvR~yC!jql5UEMiXtM#c-~CCm8xr*`LwmxojwW|;o;7Z7}DTE<Tumo
z?;p^(j?|7#j5kn~kwSowYV$3b?($nV#l@S*<)Ug1+ZU?xLRUFZr%9x0%z-FE6PlUZ
z=n<5d-EL-vkVX*_5G4A48h0RLDyLeo3W<oA`!r@ZFLL@beQ7Az_Ye>WTGJznqOfp1
z)L%+9QX;NgF@c~IBFHM18a(g*iSk=^?Z)tK&mdw2aOd85w_BILH6umD8p>&^sE7i5
z`}#F2lMwzkb4Ns!ct9fgXhhUVKoR3oJs!#Bd!I$m(?2;~!EiB{0kQ?n)J*fOvK*Xz
ze39)AY-eh%<z~Na?MX20;J;^of^wOu`h>gJBFl0XZJ8b(?w6OB3y@2RvU}3fiW#k~
zFuK%TwdtWW#wB6KN1GZeSSYmKyMhFK<ciBzCcphRTe`VJ^>cnRm>a*ZV~c<`XgN@2
z{CV*Z3|XjYvtQU#CG9}b9@`qA2MHBn;znb&$i7FG<>mE^jBG<6HJ&H3j59Vh^^VQ>
zlPG6=CgaML-c}S_+tuI8K#L2s;F)?)r-4iy62nx#6W`{vS#1lpBsa1Bp?B{Vz@qd2
z-fFkoW$XPxEr6MsnQV35Dz2~FE+S&cMvkCEZi^~x*=Lm`6B0<+lefEDLH`v>-^@L&
zvb6T^D?Fb*{emCD9rg&Dj;=K)heStbgrDH|N4wD4V?ghPiSL)-)9=|uL|&UW_#q-1
z;wK5@NXXnul%AR6;5-ixH{CzreLOtR7;{FH{T9g(6sTcJcQ6vE-IYmW1$M*owe#*k
zlG<x$7zsDY^Ij73+-bO}tqpR)BErH7L0Cw#5;R7UmyCF&R17*U3(rnt7d(W%ED7*k
z@_>Ib>boKudIsxvM3cvk050{rI(Ihf3xDv9ye9}yetA(`%www~e+qwKkrXhZr2s$%
zTl1yVvlgZQ{tDLQLzC|Vg)4$0H?VP;klNT{Ds$CkWwju6kxBR86+zYYys$l)++Tm{
z;CB2MM=@BL2L}w2N)cXvUE8wJr_;KeKaB3*YhMu)_1c<;88=Lc(dgg-?=gzArxm~#
zaA2z|E6jiY6-+yDwFslb|9k#4G54;aA(F~GrqkFD1J@u51&llJXxBi>sO|iJcYFAy
zW)23pJUai^-9W}VFgUpRp&{t+zZ#X?5c62m#^V3?9<jwhQ@78$9EKY_KJ(YtbKz&)
zSJV~w{`=tmw`DbwAo;?8V2eNDTsnbURR1!Jb)(A&`|rMwf3CiSp58=7;(zZj837_i
zJCMSV)%?4JR7>-qssua65G^60`0pD^w!aM?ACAoqp6UO7#1XtbxKrg0=6@GL8$?v#
zC*W-={`E(QH5ovJGjDZ7{@<TrorGO%!aWI=s73blf^E3)(uYT&a{<)&6t6|nqam?*
zMl2G}@<fwyAk|=;B4DR!4J<h50~%M_YFsxNUI;vm0O2L}6$AdUW2b|OhQQ#pLXJ|<
z)e6l35Yq5(%rl0bfxU$eShGdFfcFM`a~dkD$0&9TBoziEKJeMLHaTD)0<UDd=F#cj
zAtP#)omiL$?<?-P&Qc3V*SSY<TX69LJlx94N>E4uCBIf#%Cofuk0B+ZT1szjK7klj
z_0B00_vbcRT49_Jbr?wTsUgV}&y+yO-DZF-CJOV6#N1c#TyX!pWz}J{9+XH*YJ5D-
z=_8}VXE->&b~nIx*9*>J{P)?u!B$}CUvYk?P5HUy^d3%ATQ_$gNBaVgC36=0zYnVp
z0K2jbx}~@NT_P-qV-12(0LwxA1#)84m87D<%xB<8!!qj|lE4(RVz~a|-r2vd(kp9A
zT@9hw?*+oZjT5{9aqGs%k3tZUf$m&z-&Qlw8Uyet0Cw?&-YkQU^D(SyFow$D>)m<A
z(|y}2t4K5jG28`OTAhKxv}-dR9V}qk#!28DLGL(PGY?iA)It-WYoUM=EnEN2`5uU=
ziUz}e2J^ttKGE8G6S}Mi>Vty1p_54U3mG~2C^TDxex4g?b}ZsMs08oqJh#z8lAUh%
z`Si9xCqN6BmVn7U&=02{NWppIQoxoptFY+>DcZxMqhU~W*F`*p@qKwF<G(lNW0Wdc
z3H1D=BbMwo;cj3%y&jkP`t>Ww$O5A>-JlbmZ1APD<@g8)@2Ty5=1T`X8X=?TW#Aj0
zKhM4}1bwNoiFHv`aWOF%XjqsX#kS9wf}|2kZjkc<HI-#XhH4;w0{;jiYtVOim#s!)
zdreSK*WBEE9DCTM`AA8?M_&eD4Ml=N7`-Lz{EnZGPpj={XJ;o#5A(XwcMQzSEe3&a
zr2%e!A5d7ej)L`vsSed?Qoem}zVGZ@xaQ-L?+fw2l<T5`ZK@w_pN6WIfj!L~P`fx%
zq!OfnNvf@=LVKD(QUaP1P=z+DvOi~@A@)WmO9?1-nH!K=K)areCi{m`kY&kvNn~zU
z`94#oNF6~$Oblu@d2d;9PhSoCbe(J8h2dQwqywQ}PT2-9)BrgDun!8@rkiRS8axZn
zQqnhpL~7j=#e^VPg+8cEl5AnbD}`S$Z=|igo$J8z&o)^=F+X*(wOxUp9FQ*_E<Ju5
zOHV`7%^zI=j5AKobr_QcAuP2!B70#TVl+_7Xhx;|@4K8qg>FcG2?^L-OF&{gJ^z6K
zrWX>pAxAP5fpqvY!vC*B8amVpD+Eg7s)At<Zo)ZWg3P|)Ap#KUH;Ew_zYiGtkh88>
zJ$xvcGN`Gg_50wS9;7G~Of3Ms{@|<sU&lGLzZJULLFi>=brtout$}tPNZJ^H%qfWb
z*ZCsoYlVR^we|J7bWi^JxNINag$pUrpQ(Q=`tK8Bism6nT&Amf`}@^6C9!dTJ2QqB
z*}%%F@$v7#>icaf{@fiQHTQWSaO;y`|7S_5!cYU;ju7JhwiACo3A{`5T4Y95#BT$1
zYICWAQ9!r_xg6Nee=p_Ru`zd`o9aqJ;`u+*K}iCWblO31vi=<5UqAK)1)M?HmjuMb
zabIu#HkzjfOAfgOYMQ{0jeDU85fqqMGDu?$6i7zdt6#rvf?*KjkwpJpX!#`yMq8}G
zK{i0p{%4T^dV`sAplk%m>hFtF{CVAws?5&L8h9X2jqfS?$v%vDcyipBk+A@&4EojL
zLBxZ$-}7rDSun)thG+L*ZxO%?H~{A_T(G-+8|^X3z>}OSfv5q9-makiV{mZpFaIJX
z<^@7j`NI$<qN_G&1I~$sR@(z|F{H8<6AV~?Jxmk|l6eUIvd90n%kYfwW<b6$zSZux
zk2>{5-XH&@*<2~Ze>noT&erg7B9K4@g;De~J^>$f>R{07fD~+fpFg&=u5)0BiLrKd
zu?7U+aGc~8^S~29uSA={gy{b9(F2i`Bfwse+Pz^dzxGaAhu+t9%ru1GephxfYhF+J
zzmJN7;uSDwP5hW)U^aqMQG9!|?e`OM;NG&fhFZyO=Zk2Br2G*Y5M^gWe4Z?x0hr<T
zVFntU05}E>>PjNFz}`KhyOODof%eI0wDQyiXJ#_9vE@PuqvsghDv}JKQ1HR6j;r%P
zl?5^=D++Cy(SMJ*jC=nJXTDx;TYoGE8jRGYln0w$3k2orhvI*3VQ#o}-paMzN|A@+
zG9#c>vLLPd+W+KuV)xU3cVl(MjUW2++ZQ@IJ0GJ846vjgvP&Cro8eqQ-@!t`cLa12
z`~b3X;WnJ%2|cXPP*rMP?gxUjwg3u5ParB)xKCpONZ}w;x@B*l_~He6zsf*;@6>q*
zkuo$D)z38j=cKKwkz3F~fBEv|iMy2l9i1X-611#X;!|x;!rwdF$^iK`FlrCImZl(-
zZ2wQKtE}u<^fy`!vX`ug<<`l*VZj>p>+0x4XJvha1zB47V`>mZ1SFHp9m!GC0RiY$
zj{>49BrqF}Nq>%66p^U$Kq09n(~`2BFkX^|$~jm!KPsUxQ|RM<arp4hdC^_9;wAOJ
zgU`WYbB3ut;5!QoA^f5@0TI!XTOD9&e~z+hD(WS^0E<G95%DW_gsO_K932HYDM^7O
z?%3EEoJ)}PA~{_ZgAlR(r%^a>Sflr1GLh$*G`x_K#3ThGxOUKQ02=Xi2S5KDr|h9t
zQ8eax>FxEWnOkgfrsUSDD(NXHFFc}TYWosVzcZFYkm5x9pH_qA6)Q@np@4mg3T>-P
zsJ4iJ+KTtk7&_G3D_aW)aS+MHdh1RC?_YmiKP8~*(h^JQ1NGh-^iAIvssy)p#bx#o
zMiSr|LGTBKL=c-_+B(4f{YK{Z+#!%+;nVV}+v;YJzW3!|<-L#|#dBR3DqA~JTH!q9
zNP2*fghWIjY!z%XP!wHTSlEURlir2lKQ?9m2tk<vy)G-;i-5AWT}2TP5(W{`E`Ivt
z4RHYEA-Fg=%s1lFP|)C6SzR3$gx>V%7_;{g3Zm`{QV&JVKaJ;)0Z!OAAPLG%(A2dG
zarTFQ{$ZzfcXvNm(%07tM+wA`AuH`){r7j-JxO5xTB($j6zdz!Z@}?C7x$2zz@a)i
zZ4zAMm}I_94PmQn6Kp|f&U{dJ^w%<h0-!d=XdDbrgRyX-Ki2^D)219%phO^?L5>Zk
zH3SsKxJEH|f!;1?$f11{mJtM((3}TiXiVlXys*5olH%kOL<RKxR%ZqkAyyQ=a>XEd
zCUBlH3&w^3U%r3muN&nxF0%qQxJPz+x(r}@hmF!Wfta`?uOvVhaI&*AaK(nT)*O#S
zH#;W>z(GK<(2?GgLO^L~g=|4)Jlf>Gn;RNOteeY#W;r4P2PMFbV&RB@LnbQ;9g?NP
zhs8<Qfx0HzOe3||5ielM5H!~$?ZVU^=t`B3DV5dfS$0b**&X=O2DmKK4XQvCU^wT_
zor59ICe40hux!$Phx;HuNFB!NQknPLR>z#h4I;j9Z@%TMtuJ^-FlVt_DPG>*V4s3z
z7oc4pG)pP>?q#eaVJ9Jv4vH^d<{=_(oEU#iUQv4Y!vrEC=<(xTkN{me`S&QtisQm-
zegNI<wnq>+fE$fD?En7x{qx8+Xw$c|TLtB`)#YWT8-XD8Edp4fzyAp&9Rdx|avN^(
zO#d7@T`{ZqQDo%8#3VL1lu2_PulsG;1Wd(jg%k|01dd08=SCNU)XTWI?CfmCLL%~G
z7pNgBqNkP8Qi3KYvhD^O8(W`3Kw&zB!ISB-s5$==Um*H}9VhoB&&td+=B7>=EFmBP
z8Q^M{Ic4)B8I#I^drUSoOm#Ro=0h)=A;P#Nk2J^@h5pJJz>{|@p#d7q?)BqM%D}X=
zG|E7tYv3p`F)^P#dj^Nw;*Q9Oq?m_&NKlZ1TB!xAfPevA38f6>6F3;R4$<TJ;IaE9
z>#F25LQE79|1|8SM+}8mFnT@RTPwzx^(n~DM@gb^eiJa#$Hky30kdB)kNW|0f#Iro
za>_mZ{X1J*=LK!>nwh;pU^02&Ds(enLnrf=epiLL%PTnT1g@UA3|so;k0P_HKcZVk
zD)(TtA9P+bznOvD1~La2v#q0usH$?l-b2HLM@>E&8|8k;<GF!@g}aFXkr`8Mfj9y-
zr7HxECos!`3+~@+1f@-R@6+h*?`tDfU)!BQ<=c6>Djx){$D7MuA#OBVe7jaCd;{9Y
zON)yK{p%bjIBl)nWr35T-C+NeF98J`^2*h#n<oYUHHC(U5AbL~CU&dxT7Lo2v64!V
zOAQMQj$t#hh=Z678r2D}n*qqqf3}1(K+r4Zs{qVAgof{ZK4=wUA8Gn13K<4e2LVu4
z078T`&S`DHpb+xHzR_%ddTFQ=pO`n+4<MI?8e@E5U?7as+UIk~F4Ye#Nl$N?n22*o
z<xN0AhT;JuI+z&`mI?dEz=rOA_oIGdMbBjWMpN*~g)koO8-&XS-*uV5&6CBvwzqza
zk2GM4?SAeN>0o8Wn5Y8Xr}p&6P4XkA29SK%7jMBPyZ!nt5d#^SjLrq<`;kaVlPV^|
z4Vx_C5xN1JCA&@5k&h|Bf{mXKCMfPf8yw+RnNs=jiHTw_L9j4-dRbhVwt^>3_eLPK
z|Fq6awm5GmCxT~0J~(LpmNyh;YL^sfHn0~L#L|6gSwKnwVCk!zHNhU*23QDzVb0J?
zX%~?0Y*?-DGihKqND;Y{I6?;6py0eA=gruF;yNIzx7~jAmx@2!=Wi6u;PS|#6oYWF
z-<FiC&|Kf^C*(~uf9zA6tSXo3FLs0NPt(<0V|6;A><^C*>!49D{4Px&11=PMdacWG
zAS85h>i5sXExt77W1WAkFz+SNwa%_Co~C;QR5f||f!r2v=+N7Gz91Bx<upJOR<;P(
zuKTdN;>5TWZ~hr587xOf4k?}c{rg&)nub?MbfI(TCp72<hlej&dcJiOEzZHGc>$=P
zS}IE4eTBPNZ*}UMsBw>NgYSK(GrzrR;72P}RnUVyH&k_lw<bmz+-U|oIxS)(NE!)*
zOc=WNiytCl<C{`Y-;=xB+LEF~1rXySKfOno>(UUUe~zDr7}&|er^FZHr_a;S*xkCN
z_Y<W@c@Jt;P}j<yHb>XLY;3BWjN*uOF>w{yLA`x^yj<zo#%<XxhKs5BN53ND3^y5n
zuOmTg>$`XFE^8#mP8%B==VfODbBC3R5{e5J=L2C15<nUd^Jn^?9-$0cWeXs+t8$u3
z=v)U9gMb`4{&l7$*$?%P5fSb6o@UEvD|?Df#q#1p`35rYz#14gs*9|s?#qf|E~O-J
z2>1?I7mO|#^`9gaHBPg(w)Q*rcF5I~f023(Tw2Y8`zj!?Acu~bP63C}Wi%H_!efo`
z%?Fso3DvmK14`3s=ZONUYqlJef;EKDx5ysRDr@4V?TQU%5emuL^{1F=Zf@?@5470^
z?a5FEg~AI}5n{Kxh!$!9HXU3*{`mYJ6CFgOv6MiYfmTS@V2Fpd*Dr9b&>EZlYnhf3
z+D<FE5!cYM!}J%qr+1|fwU65`Pgk7<V~=K_6$qliYzYYF7ruZzZgO(6?q-La4s=`_
zUM7InaUb|!hyb=D^W#)ZOjtEFm!+Mbm^eGDmRO)MG#lr{Ge!o_QD{Pyu!C-r%Lu_V
zl%ii9Q=F#+yjg9(cyZAh(gdg}<>j&ekH4mZ16IxN2q<-dHPR1$0G1F$%`@%!Rnlgu
zCy7n-1@{$Phug8xPq)4f^)j=S28g7t2n!PdwGRkpzthfwWE8BW*&DC&`S_)!C0_Gd
z=hqN6qVQ<1_qfa5b#&qr7I1wB1Ashqj4wY?^J+XOE-HG{*a$TdvJ!O64_RtP!AFp7
z-Z3n-;4vm%8oWwAlw!5S%#@0~a^a_$mP<h7@C^rFVZ{tEF6@-hHVl&X)c}NfV4u<k
z(12k9Gl?eH0g#qI)ay#|7yQAsuZuUr_R1LfgXdbK>xHYwoaU7AKIa)M0n!5L@yW?a
z1J@LCh?<#B;m-K2H2i*l?)f5<uq(p%{3KRiaWJ_weTW%e5LGRYh)+xV`uVe@-4H@Y
z0@^(j+ZUk2ljwgx9wUwpWPM<^dCs@nHY|q^#kJ%#<C3FH%zbD0nw_H;?^|h$-Y@B)
zo`Pd-W@0i@&vX<684{X@82{|sw{I|O?uyl;$gGc%#24y;=k$#J+6J7=7&hkG0y4zs
z4-I>eMXdD!Lihf&fZQ;nu}qI{#<T0irn~W9HxB^Gsk9yJNZ*2h6AGEq$NQH8FQfA<
zDmZEdfNKn<pq@RuSXJ#n@<v&OlSwgvg@U}KR60;_8|Egq8DuXo2KGbsRL$#o%0yv}
zI-Yk%z;}vi<MFbLlyQGLR2xLKq0^7+%<XU8x^(#26z&^Hk`5eKuTHh23X8v}sk{zm
z^Yj~^V&K2sC)pZQ*k;hVYz1!v$}MYsFpJF7gPJh)E*(xyfC#m*Np(S}rLxXD2&SM4
z6_R@<s9g-Cl;M8m=F`U&M<$zm8Hk^mThkj2nuTsm{PJEO|A{of7hRj5-vTIrd^Q-L
z#7^5gH_asN-dD+2p?+WrRqNQnM`UDYvld^7e;|MZr;$w5n2E;%VO%=P+$(sqyvgqb
z20XMHD6Y<vN*_Yd+)|X1Bo=ZPcfVOt&;9*u<FCT6S9K{NPD6KTTmOJDU0?pGt2|Yp
z#`%43`U(_;LGoksCoCTs&838$*En-#h;A6*k-)U96KI?s@X>x7;j>GfaDP^Ms&q5)
z<x9(K>F2TwP!aF=tf#FlQy38Y5%?PD^NNQkdNWx%S5(quZjE<bv$mlHF+SZ51A7uU
zXFCt9bSp%~yz}x?DSStFh=}T;gWd$EIW`k8e?M?1i}xjv^XPSRUBrTW<0Ny9peMd4
z=4RU+f%+t86oC+fQI*&vh7E{|4P%@Zfubrs%jLIx7YgLu%OI6_64w9~W7|RT$>N5H
zDE~ms`$ZWU50GV_bDEDKaZ_9<FevQU`~guWxCX;-_n6=}`ecCXnZsjM3EJB6=iHu0
z(Y#r{?0fwx&;^?59D>8mzghlDRKsRcNwA?wNJAR9im7lta&&ae7SZK~#38iZA)`<t
zD9ShBMUjCMMaq}u`FYupOCRUBm<Xi(kN+vZrOd}yIxwIPdn;Cp6-M^h&+9h);FTlj
zSIrIl&Qsrxu1vJ9$3*KSAqFnlP%&8m-eK?Xp1g#U7xz<VE>IhiTy*Hrv!_4Qr~lkB
zf<eiv5bR9sKIb3+*=Gz4jK16A#pEK6<DiO80_T2N1K3w+f}|uzsoOax>&5+*!3#)j
zsdbcMz-QKu7i7Fle2OnB^p(3A777;8B&PT8-WebJUpa!H_{nf5SvI73uA&IrrwAn=
zP66vAsoL{`Pj<JhL`tHl=a+-AI4(4DuWif8&0SUPS*r6<>-zM2$MN7-5=I{=pY*)5
zwy~Rf92H)=x=6qj5m3eLL{vRb-ny(APzWF;Kwd!e;pias-}{>S?3sp@l~vsXB?Sd2
zLxHOzr6*cOBAh0se|_<aQpUk)2b`MEiw_5$t5CDGwZYS4Ma{22-=gtn^KtzJL~J89
zABnb}gp7)0DaJmgi(IkwHk+EV-E6*|e+wMedGHlbW6EJ~PR)EfJ#9)(=?BLNF!{mc
z>bGMXnmu#3l}hV<&Nr?@qya9gwIIIQ(@AcWYjl@cet?G!S+@QiwoARI2HpZ0xr#~?
zmK|q*M~AGsOq;>a)p<A^AZ3v+BYgrcn=<ea_i|{Y9Wn7GIl3L*FMSaU)75)>UnEX%
zUff>n1wGZ5z@7!K%dD)AQGyhWEICu}PAMrQbaW}Kk{Z*~V>K=(7fU#WaZ4>4r-5nG
z>Gnu=0m?_6n*jMs%Vpx>Uk2WhT4q~f=lamVG%=V1gIAzqzrM~}#EuKrC{@~@9ufN(
zW@^jH$*F5-)ZO?B^IL#S;eFNJCH_H#u(038tS{2lRRDb?(FI0;bQjdEv}<pN_1H1O
zR0X8yo#v+-7v<J>OvOU`*PkP;556mcH#7E}1{YlK6BT^JC;X~#et6CJ`0KClD{&D;
zj)6fzkYm3AQEe#5ExSNtbXIEWgH&+|#W}<wR7H0nI8n>MH=;^KfuFr7Y#+sZ0Q1pp
zZr^4l2!smz*wl#$qy5~DxG(W#7G?IdvE-|pXiBC*E)4Z>a405=);wW+^&rW`!2vrU
zFz_+_2kclatzj5x0rW9~=9Gj4n2*6tT}WIy6F%SgVQ+nXMikL~yMOY!pewE@`w8M7
z2wA|$QjHVf1VRmj$}bG(-iLapCJpWrGQs4i@-}=3Cea9Bq2?R58<EtlvdLDSeaQ{^
z0iARgE?y)iF$@G{j@2(;?%3E+Qc~&z8N_xN9vXt<>7f!-etu<iwA#@DucoH--FIcX
z;INPe4x%bdf*>)}>4}nEfGkw?PBXOZr_H_*G!qJVefZ(kbvCdHMdBl-sAwI++2GHD
zIx2JZe1~~X2YiC`a(M)_5>Ql_o0|i^&G4|6h~B%$sBgoqKuZ})5>n0}&Y*FIWGh=6
zFu;KYk7#7XKRQ{q5WQH%ISwS6I68JKD9jVdP2zn(NJ=@56Vdu#UmkiV><k1_4R!U}
z8ztPi8)tC={)Jp6+ekU|D?X|XwlU#d=aDv4REL|Gy}f8aoiv((X^;yEzQ~MDhxQ~z
zfLFX@R=UKP|5@DZ?5(XW2#meFyd?b&&zC$wWeY=|0<+NAAM+3qzzELPIk%5p8knfG
z%nMn^W{#6@CUZ3S0to>_J-148&5K{(vYC_#yX{&v-U!t@i}c(ns1RgToIF)cDu?J8
zwvCmGi~q_%lH*U{$Q-((RoJbS7}(>X&mNLWKVW7YQ(Mllii<z@(yDo{(&H?*a;_pN
zO8L+Pu`zZ36s-djuT*H{BEr9>)MLj)Df01|=g}t9lt&O~dNJQJvH)WV?EG{wCQ(u2
zAJ40wpa?ozVMqfM<JBAnU`&q~L*UP!KMM`-<QWa;C_T#vKrk-hFf9apzP!XHhu9dW
zeV3=>)H;KO#!WHd{PhkDT-S3KV=BcG6^)C+PEO_;7(fiCH9iVAzn-0&3sgjb)SCrI
z%BSxVpeW!MD<mCeunS+kI#+fRv-3sH3f(NKJ-|$1Y<)aww7xobzMbII+$wsl4t=<L
z*Fs526QOBVzqW4r=@ae9vr*rU*95lCes<?wTsC2vkoUgoDYX`$C?LTZ3Wh;37oaPP
ztS<8HYc}H(TGP|@JGm%yi^zMf$lT$V(vVjg+?q)xxA_b))!Vn)r}J+jp$^eAOoNHu
zA&~4fxh;nZ{*-?nl4?Gae1(+4`m$pe=VR~Jq+UKGm}uNQp|SA@;<)JYW-SQmAq%|k
z>RRV_RR6~JDpLDc3`Q4m(aH5K4Zpr40fuUue@~s95~x#PHPAdq!&2BsRRD&qACZJ6
zmQ#?E!v+Nw84_8>Qu(44G8W^(l&g^;Av%}+q2~jy_aIaZC-7Hi=9WW>h%eyuWj>xu
zJ<$N&Gv87x;j?Vo|M?R}V!=$hK6lASaBO6qoHhWu%NAghl1dw--{M*6Q)UYKn4FS=
zZj+eEgN4JyM6DN(u&bU+$LD8f5AL+~e(<2*@AX-v_D@lJYwK@Nwa6A&g$4~ex{Ux8
zn&AguN6+dYY%04!vjGonI*b5sE1B)0{mF-+ZkGr#hL5*5M<2E%+5iDaK!uP<Uo)+=
z%?d7<vkXYw!WmEa*)rA%01bKp*Qts>04jq|YZePUwne~?R#xr-YX)1q82vi6&COG7
zsj>y^e*Lg?nmp7{Xyglt;=A3HU3#(u76H8i(^^hYijOW2q+(zgUUs_x{1tL?=HrD;
z*qZwngEA5Nm)&8u%CG85du(bFJ1c8z=F^o2UT=?WWud@u|IKqM8XBq4xDwQkPee2@
zJZ#i4Mu29T3keE(czVY5zJN%Eq`UD6;t&ddk1sQz?bkgBA7HFB^DVvQl@;9|v~+a3
zB%|utwX2kfSa%Zex2-TUv-i;T{$~5j>t-?8hQ*g=XWdo#?_wZgF=55-LRpJ5tsM9|
z;Ir%<7v1XM_33|iKmB1SyK0t7#|^v$$3aWy6YT#_OiN7rf+%cgbe%!c9!m4VKMUQu
zNJvS+1$o1@E-^F5ethu+<u=p`S;<(Go>Z9a(2ZBcjWQT`)So`VeMqGvA;sb3!kue;
z(7k{x*r2|-Oa1t7m%=$FhK(k*gFnljGVilHr*Y}%;MvGXlFp=%*2!k{+`)nG(V-!y
zHDrr@eX^WtYW+|tj_Z2|7~0;CA2~E<pyvqa4yyO$T4lvn=OJRM_xO}qq{L%+OqlW|
zGv{7-lz$VemEo7^RwX4K7>`nrj=m)UTzJmLTiqm7z3<+?e}8;@Jb=Q0EIIZYlvUt?
zAO*^TbFKP$0v?S`<C~C>yP)8=d8Cq?XuH^pdy|vJG`vJ)ydG#{L^=(jWfYTqtkkU?
z21)>zu^EDhN#q1q<^jnpEbQnth64ipH{hE>*&WCS^XcU7TL2}*lJupPx3{-LqO%V5
zgtd`N*`#2xhs0_D2Lzz1O&1-^a)D$PBMg|~2$O*cP>?3VxOSj5pgJ&mF>%2&N<w+X
z812Q~vVrpe6O8nFLt7w?R}Mda)7F+6fw%$0ANVjnE?}@nhlf@i=x96+U6caeK{f{l
zfL+4F{HZtf$q44tU$`SdE(SbRU^e>MjC{5iv5T*A$ex?O|8)Du4;UJyWc_$43I&vi
z9jH6~zE}znpg4%B>hWGk`9g!qbbI?pAPlH*NB!f51YQo5PEe;;bD@KYT%ZF(S%vpc
zvwt%XFNTJOrb;c9t-(LOK*1<1SV~fEbbjMK2FMr8gvv)M!oxrAUmopqhRIPnI_D|~
z(8v7Mxk4MgpO0Z{x&>ndjw(D3Z2?>}aflyYrK<nE7TMd2R)9XiKSDLMQ75_W=y-6j
z>jWl4_#$kWa$|0=;^1v<$dTIRqA;(r8y3Cwbj9U~X$5VdgYDZ5tkAG9d3kwYvxE8!
z_`+!U7Foq5s8-EDM0)ddAv1vAAz(XjzK#b2HjJm6Tc+3&p!q%Y^XZFjWiWONXpJ<d
zk0hoKq|Rc;-kaw@X=c@}4oU}j`1tuH%B<+p>gZy6T-*f2Gk-42ZXF&T9upI@l#BN&
zCkJpMsAgi}-~j31lEWjEYKyFCpZ>j=AV>Tu5Ws=lTPsz(df@ugPY%~0G5}ha-9Vx%
zK)M1ShFWE*K1Hj5@WtL{LJ)IQN{ZV1ZtHS(w5^HaZIR7hNPqoWWGH%|EzUxE=G{9W
z+^tMm`d<M7NMKH@eL-$7Dnc;=2Bc$QW51uC4yBdCL35>myCjaTlki}C_u*z?=@=i6
zm7^V0o=^YgS_U&eIj?L#*nP^?Sau(YwECL+LfkV{uzb9cs&$MGln!%pl=#|OWr5#k
z>3^j7@fBT|3<NizxMo^&7iRf9raVi+$$;yys~|6r>DR}NRviKsS<>!{7YMl?sJ$wD
zdAK@s5#EcA5yyObC^E`|!&USJu$B9p@IXwh3ea1ycwPDz&}WUGheyNGk^y}kz<e%}
zk+t{(GEPfH#kx+}P+!k$_F8>B1Kye4ojcbIWxi|Iab|(}rEBE;b_<3X!QrNV|IiTN
z(xa5bM3qaSgSMb<D#FFZMHBP1Fv=sYRc|?R2UZ-4eC=A>3!I%|<WT>weF*I~{ZJB5
z5F_;mBLyA<2w%!Z3Kuy+n`3bBPPfrW0`x9&>eG(&F&1kV_C$UFwiB?dY<5hl7VB(!
z$>rtSfhd!cn@b}%N~STzFT39NG5=e4NLZNDc;2}~N`?|$1pO7j5n<S{ZK`z`nH<-Z
z>>n8~MQwV>w6wIg4g9LY3WkW{fuR6h4Xy_M1dVH%MtVT`xWT&zUf%wqTti-HRT#kI
zZ(Cc>qSpuVAxB%=_|#O5DJYu2usR9bUr=qe|L`VYN&sfHp98#<=V~tsRnRW$!>Yt(
zZ+(~5$YNn(0a%Q?TU$`7C;4_pG*1`tN?`NrR}$&l)Fr|hVCq22?lN4*w%iy95Vo)P
zKEHQ|DR>^sP=c?Dv4@j*Xpq^ZjY4tmDxjbaiKe@AMj)uDCNku4Avm{{;esxWJV(SE
zsJ|Bbv}l<;<x*s7JQRibZu3=#aM#whLMmEXk;jIa71z+e{|}6pg6U_k9t^+_Leoo1
zQW7>U?oI>iC1?mK=#n-4l|GfUrFX8K3Zc(If?fd-7co`NN-zUQ@M%t0ATk{4*L1iB
zk_l-4mzkHBr($V&0_DZHy=Rlha5iBupw3D_XwSR(^a9r{kcND|9IJ7P$sOJW)+Bl^
zuAnlSRt=pdK|#t%S)Nhl(Z>4EQLq`{GC?oV_Vonz`v)K@H*fl*hYG*GV+KdfjiDac
zx#(ZV#K2e?en~2gwqVCpupB_C4?UiOX@+(Gi`s<b7*8Lm6{dROYr0^dZ6e|mV1Liq
zeyZIxPV{T^$!B7Ea-rG8+8XObf`{kXC!&v=n|yUfzi<8R)YN?_Cu~umAq^d~ns_qi
zS}$_9#TlMC?&oYfJM!AvWMQ4(e*B<Dq@hl`ZwQ<gJ3?R(-6QlP<dGq6$oCMHBe($-
zAttV_s`~!zo2t6HLYER)_5KMMsU|NcH_rx62PiX}6pM_>D5GEBU16rHyL*H8&SgOn
zazat|3>>2K+P!J-P8Tqi7MBojgmo#5?CeMU1@ciDvlNR-LpM_|=&aG65Pb*rdsu>Y
zU;-kbHEJ|lg2^B6jXJB1I1@q+1Hbp=fStm|nS~mu?FMZ#3y{2?J{%N!&j##&Gj#L*
zd}~`Yar=%m?VnH8)z&H~DYczmkCH@*Y*AeWk7?KYP6l89I$!ANGpq6fj}*fE(+@$o
zMNL7j++!j45HDX=MM5r9Au>hg^nqE8fPZM&>_>e1&PaB}__v6d!tg(JZP4G+=+O#r
zW?nM({s{HMXyX>y-cH*}F7W~CQykdf6)+v405U&6f9{gfJ80w`&;K~oa73f35}J4u
zuyR-?V|<vszr3{6g1*%vUf=z#1rQwoG#coO(5aHOpGEeH{gn*&bVarT`62WzaL>>0
z0k_T=SgkSn_n{XM_FU)r22c`!x0=wuj1+~EmxYA|cuhFOP&b(SY-%V?d&<UP#szBg
z`eZSo5Ny9P)y}GWK)dyN@L+5`4ID=Cfd=@byWYF&akVE{xY5Sx(52<*SdfvyGywb_
z+59kvFR!X4VX`2BR8IBt$+Hd`QGV{%BQbzTBx5kp3MFnrLc&CR2wdWfpgcJ?ecgMi
z?4~?4JHW?B!DfGvAVs%1qd&aA?`39YcKYGD4B+(URHKI;!b3xu140OOC$<l~sl%fF
zO|P_G5^%S+_Pm29w+f`Zc=5~95*GUYwdGjZ*o2Bxt4eZ9OI@t3+kgWwIRwo>Xr34+
z0vD8XCgAj){)DwuparjBaVii_0ZXr1UtL{Qb^zEc6wS~9{1@b^fnr8-dRcAY`{5B2
z-!_$iozH#@WeK>fs?;c+;u#<`T_t%M7zpN%6yO5Rp;jQBz~D)EyiRKY%mbL_2w8Y4
z^t7R`%epcT9s)Q9IB;Eqk5({R)Wak3tZx-vumcVd0Si=JqgKwkxVXSA7GU7^zi8rS
zXJ;1|r_<v^)7ImbK~n4jki+od>%bzy&f!A=jk=(~4GDNJFa;kaL8yd`l1?ve8%vOd
z(wKpPSS|EZgv!AM0RqL&<r^Cxud1k6)Uv!Bhg*>jTqlD%F;Qrg0-{KZI#T;_?DOaC
zRrg3Ph^RaM?-I~ff06rHNFqeFRuk99zu$(H0a=B|oJ4E#!1~jka7Dr<h0r4ie*evN
z<bfb$e}cJudVQ-SS-{ay=|X&l2+-L0JnWi6^Em-u+TYwMXofEXZw)ABK+?QbbzaO^
z!XMULktFU+wbOk_+2GFml%Q-2CZ2?dD7NqW?LQ<Vu3=dyMuV&te8MogL*4QROh{Fw
zNBkjREsH@Z_5uR~DXbNTs_oxbh7ukGT~;}6?))%9@LgFx)Kij@>l2g2Y8=KOOF5!s
z*7Ybd&EfmQNjLTeP?3&vqf!A%lF%NhHGK2wQ(yc#Rh6?$tZ!I#)SSu#1Kst<?~8;G
zlB~JkyH`^3ss9gY%rFfFP+n&M@G4~qWX;U>rmIQ)J1z5<uY^r}JJ?Mj_lrkpTK=JI
zmc{(c1t$a)?8FFp6R~%dtH8G-4SB>^$#Q{K3|(*@@L$!kRPwfSaLCWiRmNvG`tP||
za2+y1edx(^)!tTE73#toV<;Mo7kGLMB%VCjnuM@3FqiN7Lm7tOG_=k|#$vp(_s*jE
zu?WN!M}T3L-J+r*eSG2p1f>@8P5K8~SkSzA^y~nLdvJ*9TCQg=rDxw(lRM@ieCTmY
zm`OGuoUtQ~%-qxS2gEh-4$uz*d&&g5C_m{z8!GzZMV52f+<2Od3eU-|%H@z(yo_4)
z$Z9%%0kJs8V$vTzB(&};DJp`IKK+eA#Uz9}>(Wcz3-6s7uv7JbjCpz|w?$PbdVstG
zKyirOJuX&UiF*iz9zFc8r15~YR#a59{H~MYp?RB;7bSo(Y($a^U8oIDpPmc)I9l(y
zxiRI^qI-K0fN15a>ecHlt?%M_8Y#c4|GV)ZW)>2fsoEYLOHt@~U_m0O*>h%A_08k6
z_vc%-^Db}_<AO=!;W5X*PD$9xjETDIfBX|1ANqV6%rg$2*GPG7WF*ZLg(L5kigR+(
zU+;O<+Y2dBWZ}>?0t1K3-j`UGdiOm_fBj?pedQ6=-v54>4A&RB7V|5=%&6C*qC%dS
zRaN_5y>mddP=G+Z3JtDxa{jDcIC)mJ(M9XX5(kG~`s9s%2H%4}`%HBR-K-$ekgY9v
z7)UnfQe$)BLJb%XnOaU<-2Bne(aV}IoNUHlw9^)qZuZ+oE+c`KIl%?WaU;Y3G8uct
zYHxMB(9vB8H0_F~sK^^v@;%BY5KksTPy$OR6J~S+2dk^2!^X~ztNY&<$1VnC%~Yix
zb0(4@HWnHCkB_f7i2w^O9Hq+fhlm*CUS?La0D}OSTzHt;hA)Xq-|DjdQ(lJ6k|YZB
zRCT$#s)7aeO3$tVN%-HoF_uwW2t<_hlcT0g__!uTIv~INyCQn-(R_iRE2yxCwY##g
zG)t1?KrJY9Q63N6095jUS-<>|apK!@v%S+W4PmRb?~&}yo4M$;Ty71pXK?aLoKYoE
zRE3hyJZiieTBD#(YoF2F9O&Rc^oiNja`L99zo9XBg_+SfJxl04G=qgUBu7WfqNB;!
z4nfGsUtxS&2g47YlUuUuj<aum0%W&<@~HNLY$}Y}V9gSE<y{e}oz|YLOX5s{miUR5
zSzQd=hFtz9)yONCFXuq!2+sPnrm5=|H3Ib&9O7hq+m(TZ!y04bTFu0*9a!<x{2Sfw
zok#F{?#yrQtu6)$4i**|IkWr%WA1p66QyD?031Qa-29@HiaA&;XzJOLt%rF&5dH;`
zj>gs1i4hwueEH_xtD~<EB(%zef&%F1VPj!=6Y#1SFc+W)P_g1`qZ<;OEG+0b`~Sa3
z@EGJn+6+2>!azraG$TQ*=AN8f8^Cms!a-IMV1i&Z0)y`vXPS_j%41uDmbbX(LwWCd
z-v0i6AR!H~bEKgsb6g4>!%HZ-LSY}ewy?TcbTl=Gz{@$*(nD}Pa1nT_V39SR%gD%}
z0gr`579R*?frE=ep}^WHd~pXxpW`pc!Qj{~7+(psjRWx0d9Dd+6knk$g<`g~tIKxu
z)pdx-;fq37uWD&%1Uf7`0%vu+2GEIJ;5~wc&ZHLxM!W`aNFWFiy&Qs`3jW0udQ6}}
z*3-iS_?LVOzH4T%q`+gystX(?%ekHj_{uJD`JOsK!=h2KDeD^)l!Afp2fwWL{B={4
zl(umR`;>jrJwR;0sB5K3a=0@jdxHOgE<QnMXG3>0QBt<lIF0~a3Z%*}Qc%cGIoys?
z^3vDW&-dMrQ%D|wnKqD0Va|1gSx8pC{|rp6*sU}x17Ku8HHu&}J|RIF^)kl^q92XL
z5R+VdZ=m+IRd)h#W6jx5@CkBW1WAy0$0$q)cd)lNE-(+Cq=wvbd$E&nCA+YYD@LY{
zb0wrCudpzo7ka7Mz|Oyr_HPQfdZY1oC&?3+31(9U0YSmV=Fby_35j-3-i|p!xv5|o
zYY{EPhz?522j3t>oX;1x;cbdM_~A?dO-uZ6M^_w1mmvKc5Z7P?E=Be|JOYAs_=b(m
z66_Ss7ox4w*iq-9RnZ#MA`b;ia&wK=trI!+0gubwfb<=*SjvPg4Y`oTdGKBi2GJ)0
z%07>k`My{gRj)w>>dEsD&lsd<r>2a?htQqb#iXNsBo8S{VL%b|+eDqjn5sVGsU8L#
z2PV(k-#Lefy#d5sb)FPXadA^Uy~=SC5S1*es)7}}B_s1FCB+O@)ylLkxsBq4wK(x`
zyPF~V$4?k8y*^c{H8(mj@uA{2p?aYtt^>=M*GlsZfFZ$;+%&2VzI+oh*yDa!u+y%h
zG>EvN@q)kj29)z4E#0;R-iZePaO@VHCba3u3W?p~hI~jAF$4(4J`H)*imoHBkK&+p
zc_K%_=-3zsyTb=WQwU9?SE8sZ^9gHXR632@6N$Itgv3+Wgw&D`euOi}OycJM+$xSi
zyrN%UTidckPtXU{zN`?m2xqS)tM3mg1J48O2leb1$UcSlVTuz`H_Hg*rI2;ByKJw`
zHJ^Z~Q+uzeS0$tc@5jTg9%KT}ZgliRXDvL~A)yc?M7k11md&$3-8~uY>cZK((Ki8y
z?ZNJPXVJ;8&1UG4Q4c!Xba(d6^-8Sf(uBxJt^5t~H3?A|H`D;6OQ=f<@Bt3{Wpx#|
zp+FH(IZ%vN-}?r5#pu!@5OC{{A#8(yQt;Q?FnOp`EgTC)v5$mrx06cqU)6;pZclQP
z9NxYqNCcX<ZFrF#82~Qr1mg7OvM%Lr)i%8`P^b{xfKen1tHjB9NhP0vwLE^u)?*kV
zQs5y0TXH_d3wl@_oSgiR50gs9*6E8$1!=mfesouz0N6filLke<X44;gdqFO9-LiC<
zhkirf>+eD_mG#YWXjLZghR-$>sKG>W1PMTH0#(+6(9LOksc}80^RY=T@)g44xL|(X
zD$%gS{!U=0^@x)s7!kjkF0gCRnMAM|a!&^z8p-4ey{`FuPv51Xd-=WZ+53{LEO+n}
zwjW@i0&E<h8^4kmNV!UybDbGd07p7vBaxJgbaZ`lqCiLRyAx<ptdT4X6dz5HlKL)A
zn-j-HmnVF4RP|+J<2Cb}yc1wyt{~@SY;ozTgqgITr(ALBT6SA7-VL=7f=*7AvcZ>W
zf_^hS;7~KMzb~}of|au$Os?<KKSO)l#s*KCH~ucA5DS8n)ih2DG?f6|#kONhXX<mY
z%!C@aIIzx=vIv=U>H+^DHaeaUCZFX;=6D8=fW}0AA!L?o9Bdi(SG4)v1&O5~>B%;b
zJC^#s4EQ+EY?&N`1^)kybNlyOeHkaX7sdXG`9Dp^6T15U8n`MbDk}JYHGR9-OfB+b
z?YXAC(k<5?{5P5yTGG$XT(e-T;Xa`L%+bT^mMBmCZfP@DC7QC3MW-SkmD^2jV&l8J
zU%;MZd0e~I&*xGzT=SAa)e^CzB<lV1@Q8UHC&(W=C~We^3$g)D*Pn;3KfA9EWd{xu
zOC`H!&wrdhp+8RED^>M6&orMxwrQD4?RtnQb-$J`$JN2&VhYRoeU(~wjr;oh@yf3I
z#^)VAu+Wr*oMG5Io>`rn^)k#jEdCWAici5i>44MF|EtLrWjisVDu^5Uam1k;$0JJW
ztJ20kvDBVhXV0h5A77|0u3wc=oLl)N_hFIf?f5&ds}{S1qPJ#mE2uqSc|OcF5zYTn
z+DhTA>tUmc+OMkHrMrFCO3zEHynZBs!H6?*qd{_WR#la2=M%-A!Ja)^o$FJ-@x}sf
z`H2z5_+Fn+_>Ze{gFkq7)pYU{Z{!^ca(&o7FYtt;A|QBXulEOGidk=MLg$M3QuXuu
z;?|v?XYR>4J0uBOPfBl^Xzo85G(0NxDvVd3L5`b^tWg@R{7C!pqDYUwT=Ds<l;PS2
zw{D7;y+-OI@2YEqs5hK-j8@lrF$QZ)X3zHDexngRZf)^7iHFSBP(6O$FT|&@f?Ogh
zE{b7^$S%-U)=)aQu7}dZJ$Bup=HNk{c&DL6SI>U2QbP}(L8-d##Te5!gxw8yx>;Y{
z$<m1Er$pV0c+R~m-d*8nT`)VGr&YM<`&OdOkoNUR`XFYL)qK+xldhHEmTzH^>g1(g
zRHym2LX9X#*JnGXnfq5<&$k*lx%WP)zOwEq)KBA|PapEIbT_~1<FEP?z8#!Vm(Idv
zWr~-6V|vf77P*dZG$aK~efcJ~z*6JQ{o2C?CH+)Gz%H2vb1S;0-<81FXy{6a(BMOx
z;1bc1;WLJUR99>QpAG#eFRAcUm-sLuRtj=&Mvkf8{p>+MQiVO-&pYzh_Y7H(9|mYn
zYqTgNJgHQiFx05YYsowFY0ZA&iE^ol-G|n!ftL)17FP9fhFPP-?$lLE70rjcdTE)w
z@z}X<TJ^Xqr`suqt8~X={Z~0IVA}?>C{d|=*7us9mP;9!5Z(Bk&XwaUZ2ON@Osw8a
zegEL@c)>?Lv1Z)N(PVEKjK`u(y3TbW>f2boN#qB4Z0RFx59BWW>i54d+1hf?Ds*cg
z3^_o7B{a(`oVVgu$m$Rw&!2IFLoKCEW(=dXIKC?{omS|YU6)t*^z)q67h=2-$@I@G
zgc@JQI2Vy02^m;xY4l=amB|<S|FoPEW7!)BK#nl+_kKCI8l|1;Q2y3cB<jJe%&5$A
zDe>&jh)7$hQC>QgiP2w5VWu5wm1kAYQcjA7y!LlD724~%e}^raYT4b}Lf78x;a;t>
z!mZx0$)_e3VOisDVI>>UwEdSkzeo;@Bv}7wx>K9q$T}Eu$E@f&8+(Pp!e>*Z?c;$u
zJj;~Ero8~>_=>O2DT%7>a(VO=7e=qCZ-?hoS`6OAv2*`tdb!VrFW{O@!M)ClISFKV
zT?C03uC#}Gy~T`eJbERsOY`$*WwD0y?r7^NuRD!N&Q=b8P@}v<`MtmGtbvb{L7KGw
z@~+Nan(R(!)wKLI_7w#zjDk?F4Vguv2-i;>_kJ>3KW|9Cws?N)Vw(~F_{P@2WNg&H
zbfAfvQ~EE(keKpJ0<TElSsjA1GObpc>yoN4YGXEu19u|A^qPE>ZuOT5<*|{PlJyOh
z(YNYG{rT$j&mSBd7K?HWUtf$K{+82_{i%L1YjiG>S!ax*-=XvMyzUxhm5Ui8)-4X^
z!+Ie?Ub%b)+fJLwbluVITT=Kv4|rCqKVA5eGv=Kiuh>^CCQ-*kbfxV}D4B~^vA*8D
zNhNnFi74HY7bXMoDXx<Zy{|0OXf`X!p78FSmp<?$TC(DzOcrdoqW?=IXMTT`=G&KF
zt26%bxg5zH`Ftmb+tj5_tcl;qbx8*n7QU@E^ENjhrK;X}dCki;B-(-aD-%{w>c+S{
zZSM49<~>iyKE}(mRPD+1PJZr064rfllIK?vz~)vmd?IHZ7g2Pk5pl7Yx|<id5Fgh5
z>$}aRW6Srsdo+Z$xXV-~t|;E@U^}z0FTI~D?(&>fkLcd?cG4bp$2n$tl#x@M!Nudz
zsX`u8CqO0cnaDu1MSaUE<XoTc*s$_v*OzY}b^3>rto4<a%i%v9+-BKAi9Gq5F4dNo
zE5Cnk{4DBjeSLt;6EcHs0kcFdm)-9K?t@qX%bjwIw`PhxtuX2z3SP?5-aQJ8a(k3&
zT2yK|l^1X^dS8HXMDS)p-Y?IGhyj}Th{;25=SQ<ogmImBV#m*nzOiIZV@M#{*WDg^
z$37gXJiMjJy#46@BCSAoxEGtr3DzbHxfc9w<VPb#q+h2*Wpy^`ThE^q#Tu+%%~A$}
zJI5bYe>(U$nM2+UfBNBmjyAQpvFud}=}kQ@{(-2#-Rx)B^A8cLIvXM{v7F1(o^X!q
zabOw*k3YpO)~2eMbK5N-W8MzoTq;~<P6*82PA=*DC*s)}$CdA^dDG&4$5T9xAvS(y
zd+ZsRVvz|;x#PDzJ$Y%X=y5kHG&aK`ZRIjGGLDdSLUn}EO1eZEDmLG_dc|{znM3^N
zR4&lnP~-OSi&ki~ZVSv3f6y)Lk;S%#$ccKjQKEO9V<ms6ROkBF;)s0iz+1YB6#DU_
z{A@JyvbB-=;et{*yLa<7{Cpz?rCtRhaxbeX8JYHWwZwJ$)$#n~qaPa2dbLtr_OPg`
z2sg)gH;T$vl*_PO_~t=*sB&ZnXM7b+p~7mb@wFWHkO|jRwW1i$`_^%5lGKGk`X%M!
zMQ+a~O8SO7^KEr`5OzMipe6c@$VMT2#C=hTj$LGMG)N@=wYEyI(4vxfZ}>l9DY+~@
z>_xF_7zH9`#rsGFSI?M)0^?D9Z?Xxm5_-Z?SA8nVxh|fEx1x0Qop)m`6%C1XLPJev
zxP@KVS*0yBBjpG6s;n_Prc8>S1l;l52-JT&e@;{O$q#ny4T&E1pzk~Zf<-P16mQl^
zQmuTu-z^2#$Ehn#6yF%t+hgTy#fTK@rOnU{!E$N)?rn3vU+z+)ws(l)>vBmMX1jgc
z<)}-K3<+qls|4}LJ*XeQSl<40<xN@7`4ud=CkV}A=`o#adjh*m?vYQz?M8bmMe)=w
zHOHAok<njkle00O&1K5m@eB`&_;e%K{>EyXf_du82Yo4@LRFk8`jjGKnUBbWnGfhH
z`}b|1nfZ+12v91~D(O4+zM!^aPQG4<xn{V7Ayg#I!JS<>AXf9PTWMR{{x(9JA<<<2
z0PiFAa4m_sj6^IOhTgZLTBB`#<BJ{o`TLieh1!}%@#BU2e6s79reCz^@;0@+_IOfv
zXLQ*h=UbaeLT=S9)fbaoHyzAK8wj)(MlMMzPJUS6$lPJnx;R$1HC{Uum|HL2+@Yti
zNFk|6c5#WLf^=*(|E2<$R%(#_N<K-tgrujcv0)nXFQji89j@>WCA9~RmgkvkEqb{f
zz6<&CiH2!(T~*&G<2|UgQdbVL!!k{_5u=yrtDT3wmNu6~;|u&KRNB<J?)Ne<T;y>1
zzRAHxlR~J%qMvwatWuTFwv0e?v(jF<Yo@V1aikyXFG_I<;t@CfcV|5=rP3%}deoFS
z)c?`EB62r0t72L-J!!BkJ=()!h*08D!I59+qk}~O<VwM<#ICswEMJVA-C8m=I*rw3
z_ATM5(toa_yxv4};G#N;-*RE^Y+maT|HyX{-yz*jptNZ{-tr&v?fXT2F-9NO@)(oO
zH63AlY!3Qv<$m1EF6P5~=rBTB(!e*+sytM>mpiwx^iq?rW}COk^M1u~A=$wmnC6qS
z6$LoL|A%<|s-WP1@%ZVqYAtI=J&!wlnv4Sc{Cr9dZtfn8VpqlZ{#D8m6}|d@6{}B}
z-qAu_B|9D}-RgeL@P@v*AvEfn?x;#JU7)^d7@2qkaotkwvj|0&lcr%W(%RdbRxZ7)
z-!b`TFV5_I8}RB1BZ&<l7q@3sUutGyj_f5IO=d98(6eeoU80aRPF+>F>AP~hjbjTt
zE1=R3S1TMDJtbrskeFD;{8U{uI(SOJf+gW4maOVIy>qcw)X(Lu$(U%1`{Zm*vOHVX
zAh!vrd2JZQHj6*T?wICukv+jqCM@xr+2?3(wP%*G%$qA)gnEf>!gzy}<NUA6f-oyx
zlStLX+G8qS&)VkmfA=Qs?!Q_%Z^GiYA#hdPaP-z=C5#}NXIQiNm~mywk0z>b_IqP;
zmVQR@U$$vt4>ypUV$<f8u1qK2q?Yj>biv%0vtT7-PvGED&+K^M^;j{z3RkuRhX9AN
zpJMBu<j=(RR>ggkv(wzFRp}MO3-}wtDQBdrc*{(-+T`q<&*IMXP~+oobUKGCV7OJV
zZ>4IaX~Y#FV;ADyAjP@xoyDHVskemRWX-||f8nO?pxyiBQQM5t2dt`}Q9lvdIg)zD
ztU2n63*;9rac*Tty);>%La|^7SvdJAdD*h-VfD#H24~!txs=ELW-3WFU9ulz;MUCB
z{dYoiS3VBf2`hh5(EiEOcA5R2hk__$6p@}G<0CIp=LoM}chngf6O24;`(@h1ep|LR
z8ZTB!!E{M3tY7Udl4mn8<4o8Uh&qO1G2<48p0jm^VPYX=ZZvbu9_{d9^!BfLTas6X
zefo~Pf1lOp42_wvYS+=x3uKBg+4@sG%HcwXD|083bO~)+SUPX$Cp^Vo;(s9W|E1gf
zPpPseN1k_?(V*k!a$JgYA7k&o$BgL7?__kt;m@aK{3qDzv!sk9bJMPM%gvYSLg|t{
zgrC1sm#$8DH5ZYFPVxN+tyxa+$#^$Y`ce(g%T(^GiTTjF`;hJ6qD$Y-y0sMVPL}P2
zbpP|-HIv=tM^n2kx+(``JFAHy9&bPU-V#hd>9fI0q?)v}uC2)5Q{}|cCe^Hn@o`N~
z&kmCPy2>K2`@Yz#DxvUh9ZiS^S2gk>-hk5mkGJ$J{~yM_DM*wc%Ch^lZQFWn+qP}n
zwr$(CZQHi(er;^e>_qHF%zVsWRz*cdRc1upJolb^4%Xs*b3Q>2ONVS*a~tZuNp(~O
zT&cbMNDY>%J*M3SPZnEzk*J3kF+tbPKJ}2@LXlYr+tx+QVD-iA-NCt)3NzF&2mq`D
z`+&>XGI-eF<u+q0d5-d9qC*avO6c(IaOD9zDBU;#Fvcp_O2U!A6u~PFstcx0KmAQW
zLYW(^Ys6RNVFuc#A;X6fSGSIjmTZN3z3%7N3ChpsQ+^F^?vDm?YHI4klYtJ7>`@yB
zl}yF`Fvb$P@iTFGTI%hrNaH~fwlrDeV&SCt*Z!HCgNKs~Q;o^#KQCh5CDX;+{G9}9
zPm90~`GUIuLPRopzC04JWFRk!4beR4(STx4heyOg=Q0bj_LH;z@~sLv?f&ot$pV&b
zfzM`v8k8v4A>G_K%i!~p`_qp5)1|ne%QD#kXDYA61g59n2{$Y(cBkm85MuUP+k`A>
zDhExf6xC8t3s83Eyyk?md-=dez+hBi^K&lM19=UG&|yXY<Cn<P<;@B=0v`@qM};^{
z{g_wnhss_uuT=CJs1o#|$b7I;VbDcy^!cOlycPx8X8cB;l+8GGO~ZV3A=|xBzwr;)
zv@V=KC3!Y|N7IA-OF`ZTJvtgrcgTCtOnMUqA<dU{Pf)P=O9Kh6-VEZAmhs|$-gDDD
z>?{D(@+Uv!&3zPeLv5gt-<$Xdf;4$_D-!-tR}|%eDT{AV62mYJu205)5I1#+VO#LY
z+eZ-YAF!bz-<}?idp(qYGM%p90Z7sC!v6J{3!N15At)yKoX?b0*IionV#vrrD>Ivz
z?{eeClr2eoznpKvJRK+{!7rrQTD1y6xyF6K*6s6JZH&8ptkN<1+tVZ1;!USV<P4bk
z4)i=C0rhq(4oY3aDVvItHDfuB7RSv-&4y(C<)(gH)w4{{U<>IHcWN-|IL+Q%e>%Vo
z2tNo}%8tJFwATb-Z#}i5J>LPX&Oim)t`LRwX1u*$SDvWw!X;)BmY8sjq+>2z@o-jF
zu>KCQF~fiR&G8*p6+=-w^YHIkBP<Y*#nUGg;Vq%maEO4NlG2%VAr-^sE>fu>Tf!L&
z1$_5nmjQM60FGWI_o8OHxlJfiLj&u>jxb*ntu+Vh*GPzpwaMa3g`;g8v-M8u?NZVU
zAg|ofePUMu^fm0sgU^>y(T2Lw2GSZ*yRn_$<k34%+J~g+?d>beiyM0j?$Xv)Fom}y
z!2`euZI>&>y0?snwdw01a?XX|BRb~RtgO-*b8hC|bgI%NhS%uCC8pPqChRq?XGX1u
zC1y1{+g>A~V03ju8^U9G%lfZbYiH(^NaMZTIP`?t@+X!*^cYD@R@2*LYpTfGw6vRo
zv=K8ewOGALzL;5E9HJBCNQKJo_->mgF+rgUaX!9&2X2-|{1XRH#@`{`<4jN&JyeEm
zq`oJ1z=LIfN9(^oI1O{}4TgiCPk7(%G%5c>AZKIx{{s1c1#3onhX3mhf&UxGt(=S<
zXhp2_os5Ny4Q-8#d3Yclog9qyts&huntu~(M=c2d;be4prElI)@v0FEmNq)`2h>Zf
zD-@tbR=mqDju}xm;PV5SB7u-1T!Gi&>>M||ic-b<A0dhRyF)^$tm|y<ZI8Pkhx&d~
zgWdb*rT9I}xNq`Nq)2X+!ayE}_qWIUIllzmr~3(<c=#CJWG{~I<>*TpcC6fIu0rfB
zbvfO4G%u}}uJqwS7XypJ+$UhqBl2yYuUH|mFLW%EOTpx765JD>?{C^JTS`*yv%2Co
zHuv`1dzE&;j7V<byjJ&i07ight{+0~?m@=)^XutyI~OuoN*xl-=YCh#lDHZH)(Zb^
zk@n$<wdai+PUKeRCtR=>mF3}Sb}@EJFSZVe0Z~My4G$rO41VWL$b~e29!mV%3lWTj
z5kCq`QCY!lJ29HhF+!g-dcMOSV-Sjo!J3*2NtR9#8lxB}?$vPdhsXQ$hr(aF)Hc&6
z^DdoFp}G`MWPYhwa$`Q}dBDv%8D^Hm9==5?g-h$t<_>Rxax*7;em8R)Di<^WC&W^}
zU}I%tJn=oA0hL7yT|G&&^vl9OyXc*@FA6qa?0y!~U+pDE{fufd&07!#B60XMLdheX
zKl0#g@`MskIuFbAsyPoTC!w9#sN>00JhcS#c9He=%Z#-TRF-`uu}~Tmu8&fRHLq3e
zDJpZCaSV1UaOtXDkoqbZUkOGIiMS>c!c@XU_Q!K$)vBv3y7Y_h&3{!fZ$z0^$C_%o
zLj;No{i1LoHUjZA6*MY&v6vH>N)kYH5|qX%<tw6QnF?3+r2;Cqs3D?B7J`e)8oTsK
zWvFH=Pd+fh-w^_o^s~<`&=cu1h{0R_lqQrEu*x?_)D@IMwC`yZ#J?G5q2V!cYpWLu
z$Z8bC#+EfQxjaML@iXdk&mY(}3Tx_8KA6<VCV~VyS1#Ln`ex*QrOsrluKSiCk6x&M
zzj~H@sytRY|7x=?c<8T8pc1xm&?}e<<#{WaI3z*Wl5;Ir*xYcD0ZO^t{hdKpnpBn`
zoz`EQE<A!hMl4Enj|IIje~y`;65pW&;ZB}g3Ib}yfVH%^rgZFscqCNPy_i!~tR&}?
z0j&mIQSI$1VXtau1p33Ff0aeNi%UjPP*}8NvUk+Z^G8tH;_wNwg23))u+u!GxsvhM
zFt8^3K|5ohp-P&7$geE>;W91qq##Dono+nLize6`pvbz^i}S>QShc<S`$m5%xKgT6
zYWBFB;pAKJwP8{9?RNA0@={H32b{XT`2Bdhj2v(rrDJgCnLRz=TG0?iNed^8;X*-8
zL;ToWE0v|Q@<wcom=%r4+(|iV3OSIr<VPx5Z>OXuPU#A&e2yG<y{bUMI~a2%-~9{&
zd4DPWh+Q*&6K0>fqXdaya2tR14YOL#Wn%W_2c0l35Zp8IraOGv^w!q|2)bTO2CBGY
zoHs`JslV1tdkBU`$8i~|GVEtHF8#z~9fM62Ps8DM!5!_AA4Z#>Y6XZ!i=WrnOePsl
z%G~yBn>{aJzh?(`zD!{ZZ4h}nV_2r}atC$`PGR!M6oS(~)RS_=i1RWuUk5q94sc*O
zDU@S=vt>cGy6{AUVb#W&w{8cjZCa<Vd3o6^U$2W#N33y<e$Efdz^Tb5F#7I1DK2U}
z-zKG4Xh~FibH<f#zp=uTy7dft%cPyz(wTEylB_DzvTF-PUs>NI9DemZ>)8Usl}V3C
zVV7G5SyDCLS>Ye2&F)SLBPH`>=a2WO3o(FoZF~(K*1eES&Vs9yqn)Wzl^!(8Dr7&T
ziXX2+tX`PJT!WatV<?hcThdY;<|ydVTA%bjWv#qCgN&&en=1P5*!_nZ?4Whn?aHUV
zndGz6^I$NwefUBtvbjSeoj&bo4(jH5E1>~dZZ;j}=w$^jtS2OR^C_A{JwY0MZUlUE
z^llk5t%AdH;CLvEuI|I)*lNj{{%Kj<jg-B(3Ux4+P-SUFyrN|^Trzjn!Vod;bx*uD
zi+FqJ?E4HuH%s*@!-DcWp=cPnuEv+h6VYYRBZ<u6VxdcWG)VSQ_)7D-%CSr=B%P=n
zV3L8vl=ouRf>7(%`w*GQ0$a^koGY(Xr)K=ogF9N0MPMZI;cIafpUr?p_12H;)pY$L
zP>unJI9mcyI#44^_f2LNXh*AYVLb=w=t8zsLk{&yVW>sucuKlbS*FWu(|R)HCUSK2
zMU^V?-%PaGU{S?NXlC3C;9HL99~K%(gf(c%f8AC#REVRpvc{V7nHGnWEFVNIE`wR`
zn+bmd7LiLmWaa1EdNv=IO0FifwR>e3uB%^d9AP)=1ZBI68Uq`&YQ@hjn&m(RX1JUP
z_dZW1LrmlFv3J!xuvLO{s2rzZpcC}eJAhAAFgRvq4h{fAn8wAw8EUw-^Bf6nw6OIf
zBmaHE;s!PMWIt@n&=WXM4L6gzL>$P~*(FaZ5~m`$6)g)22Uv|7lU0$|tiV(|u9NSM
zffG9l%x)P4x)f(1x-Cj+F}&9?(HsWsV$&xo4>TYZ0QKWouSV8vhPTV{9uLBzQB{q&
zNQeN5E+~JB)od$qg1F$93jVB3kxNr%5}tu-CVNs~>k8a>jC~NP0nn)M$r~1P{}p@2
zEQx`(qn(adK|i!N9S!!>$Gkzb<3reJPFzZZE0~+Rx91dYKCRo|r09)4CVXD-nQ*g3
zZa&e=+oU+`mcRK#H!?AN!wP11+>a?wXV}3y79vPG2i`On8~}o8IDF8rhQYWEDL4%Z
z&-U((nxBXnt?>)dBy$C?>f1SF;{#bn{~Z!6ZTj72KuKjfbCM<!E3wMo0DI9YfkXXE
z!il69aiiUd(IghFF!j<$y07ED#UHN5ZBcaf%J$kS>E!b!b)&VZvI4sAVkn<oz-uk=
zq1Z*g=V95IPC^;lY`;M#>R}#w0)SLS+c$^)eqZ`(eJ@Hd&>*mLm8GO`2x!|7i=}in
zisdpHq!4;kh)<MNj$AgitaE?jsSZk0dV3sARdhX`%{>`5qmmLUqGpZTPSkvp8?mbj
z@L}@d%B2<MRwyYg@)~^Mir5nJ*O6mNRdm|w-ycA{*C+L+S8l-xCHJ(ffc!JyQ81f|
zbhMDv2Csv(SQ`N3hZ959h1<4`D;>gWMlf&N?`iD%;bms%rB-p+1YtegoOCUEIC{Ca
zt)SR6vzb6g0OrgQzo_3Ee<48jn8mpz0E9DjGM~ROoGBTfzYd{a6KWzlZ_~$+Rk(Pa
zD!iqgut9&Xjy$N|wIjd8rRKPU^B0On3zf_irY<nOkvs$mUrBU&m>bR4riUg8U+w^A
z8s*)mbT6bYhMNa_8Tk*IrHitnCVlFr8_46LQ|77ssVFAlzSpcC)zV4dR?JAUH`N;#
z_aLqi)Nh7t*Ug8LoW1op&4m{g9y&JHqB#!k#Bk$S<g-bcQa8bBqyZjM7W#`=)3I1Z
zn?Yu+TEWDp1su*=i`EIt!dF8sarW?VlpXORb#g#t!EW>=!z`_8PSLA7pa4lF17tYb
z!9+rHdyfT{|KHd;bIB<<gt;tjCQoisSIeyDOy;QI>IFG2E3Zna^drY9Ss<Cl`>~fF
zeyi{d4Y6m(R+drQ3={WoT{DF)t3yfbEQ~W_OtVV=KytD%yjUpj`ouI!$}hc&Rul%;
zE*d>&MKkAlslm1-yidC<7$g~MEHng5y}5|XG=#CLnZg%Vi8__)k4V3d=>0#)2t~&&
z#jDIEMlK%B=y>o)3a`7@z*5`7@BBg!kvZMhm!-F%Goo4%r5a_<w;3U?8Yxf8+SSQa
z%B7qKYXrQQ@hV-j{S4%DZ7;y`QaAPk$K>DX5PaKf9GANaEFVJc;x2bUf-CL)zy&xs
z`~s!d1#qNX`1M^~3!hslC$U?xgB`?a>hD^}2WJZ^29S#D0@oQhm`>`{zNj5Z*>B<|
zGWFlmQ5!}bOiKQcUQptN-5@{D6MNP_88;tB2v?~xm)a}0YWx2saZ`e;6sWRg$Mg75
zF=q)K)$H-EtU{DX-|Y3Sj6%WPVz#x(kv_;X%Ais_{Nb(L4RNL5+jKW1wVnuBwY^`f
zY)YMMZM8Ad);5k~q`(fsS8SHqK9Js+r<jK{9ESl*A4Wi^#G76w6S`6s5HL4X(J|!q
z#0o{qJGk4B-&rFYj}~C%K#LnRTk>U;`2j0OrYioQPzW>K|4S$S7YbouXJh(br3{<3
z+GB{@5&nHpEhI(dy8+Yc$?=)EyZ~I9FYg#sC#!@o{b`Ev(d}m(AFHaUF^G}ZUUVPg
z=Mm`wZwQtB$$EXdHhyz^o9ql2XDAI$UL3tm?8CzLg^WRq&Z&Z*BB%HE%Jq4hXlHs=
zzXQsm3dW24ae1w#)lIT{4eHwsYV)ecA9%cdRln~Mn;m|+Xp|2ai5=o}>?-_R{X`y(
z?I(+tcDT>D+jo5XcUX#+qYeK`opIY5oBGouN7i&Ib$Q-;OS{iO-v3L>{+0NU^8GY?
z|2nQ!=0liDhx!@F-GwE_v`kB63Hv-&<@v&~{SJbox-Iz`7VTAK_IiJLVpXz}tuy3;
zRAJlsPSN8LKgpgoU`5byAL-|%IY>;w7?0|tR<=7~M!}Ll0WOHb3S68kfSxC57Ff@4
zN6wK)J;gZ1gQvgC>2dznK7_HR)rE$8Rp8+@9#H<oPhjGy*o7D7>Bgx6Ec%dV@Hsrx
z>LMj()^R;6KYXh?bRAKIGAU;B<l}A>DX^PndvWF)Jm~jFx=5tqwcf<Z^%s5giJAq7
zx-8Z9*=XXVk(|X=<sCCFMt?al&&~bSlAax)2eWIL%`mr_HyaO1SU#;wSPt~zwcf?)
z6?CzC2YZXIS5RX<dRctle%ACH{Jgee^Ru5xN0$5YnL9nT?mTTY&}oks`l<_?wNJ!f
z*+KP;&Xhw$E*0&BNKZ)G!bT9nVTj&UXq{X{5<t-E0Z#LgzM?U?jf*@@Dm0WCSqDB!
ztSiCBt?B1%-~rE~)q%j^>PlzyN-9{y@`vifwzrENwTMbuSTiu$s}`<h362#3-S;3x
zpr%rRz{<&Jy#B|X#%iN*9XPF1r!4p!q!#;xd{_t7SQO3pYR#3ph^L<bcl-HSEz=3y
zpGCJc*w<Bv>S?8zM(h^t^UsSzvN=p17&ks*%4banYznHqu{G)~OSDAa#{APNC!fu>
zCZ=rsRv3>~0eP$K0XCg#%q?8rG!!jZ8Sus8)|FLbbq-XE(Ng%QMK1C$*Rx#}znYQe
z&4k)0h#S;;8eY<ZYCKG?fQv_kuuI1R4<~C|@RfX8akF;qJxn}vGV8yR?wZi25Sv{%
zX59gD&={MRK<i_QcB^Fn6J<oB$txWVB*Oq=)!aHXg3Z*)KQX<XCXhB@Xm=@qsht$y
z5!gon^8MH$o8??>dy|YAGFl#s+K6|lkq5ey3O!=#dOat82N6r(UrhV|a>U<<R}1qp
zyoa+c>oF76uvmK}uF>Cj-`-1qKHpaE@Kf$lI5?IqHl@Z-Za4ZwcbdfPV}}bbsoz(n
zBbC(5-NG}NYy&k6!zBg>?c4Bwe}ev~pS9Dswc1>zi|6m<W3H6&2RO;KRqY<xaAYes
zs*zD;i%xQj?qb#{hY~gsf$f1ci_m>L+1T}(h7>I+5nD;%QX>T?%WSCb)`^vxe_!6x
zNjaQxNvR4xu@quP*tg~1VkA~Zz(F%xnO)K|Y$p(<GaSvrp8r}u)25pW%Hrhb=qEut
z7j3iJTUVo|MKzHgL{K?mtYT#i*l&gOoTaws`1haMd+au}<iOeIW!!sqE?sH8C7O3I
z><Gm~>h6x&hI03$dFZbm0ae)rQ%ds?HZPdZ0NF$E-s#2<JE{82NUc2{3V@`yg+KtL
zJvN^@L|+&et4`t?&)M-8kFOHbw+Ks~JMLa}vJf5Lq0p05EISt*dVD3V`$@s6i_yw$
zfF8tYI3$NXVCu*`iiB8*XPwK?%AXt~Po&@3Iu~3EFwh*Qx4a-)^3OS7XJ=g7pQjk*
zPh#Ux4^@1y5Gkk}PHUh@_{w)IjBAsh|5+MwW62uoi^uBkq8-|@4nfWM)KTz!%`TXS
zYH=kTZq+6^tq~Q;H_R>Q>VE`Cqk+ObD#{RU;cw4Qy)_a##$0Kr6jz=kvqB0Vc8p)s
z2cAm+J&=?0@aBZyy#g&ZE?`;_vx}7??Ift`c_5;hxRFpyOd%00lJrrlg^;5B<I=)~
ze1%&_?KjU}Q9T9QJZ%}(dx5d}nqbY?GeldSXuX}R*)K&1Y)4HIC)Kf0d38rdh1QKX
z_Q3YCFt8)jZ#BzmXovm`cu!STs!@AO>OFxL-j7}{6BP0;&~2PpEwCx{jUatf<75Z3
zrKg~8KCbZ=8m?qeneqs`zH!&pip&E8mn`GPvUS|?xQAdRag%l&qCuFl1>DrGdzrJE
z=O<3$ifOe7k>l7&FSPmQE<Hm3dBt>pN}@5|w6)1&`XT(xf825|!Nd&<KAz4{0uuE1
zd<dCOY%N8ywckR+ZYbIZ9ViqTdTNGJzu{8yd+n6jPoJl%0tn_Bpc6U<P&Z4jP0u9*
zw9h}7)yD;qz;^<1VI8fJ^ZgE*k8Od#hSM)C5Z22^m4r=bt_ngzu>KFa8NYQHY)hpZ
z1^~7K^rZ`?RU>1Ek+kGijvJ2(fS}nm)BXLKATg?^@C=K7(@G%Q3mtNiD|DZ;oSMeP
zdK0)1RPdaN+$4n@K8q4LpUVP_;J(=42EfcIxPE9pA9J*zzl2uiHglr2E*$FAp`2Zc
z|07PU*~{_HB#qQ!uEnh&7@V`f!Q7NYp35@x)x#)PEq|z^C?`k>|LDdM3@i3=c}aZN
ze)>-Av{Z|ZH@z=9_6l8fOdUeH-Q-m&jbRX@SM4vZm<fUfr2`E_B=G<#A)sLKO`#Vi
z5qpY|2}AhE_a}|32(hlvCq|!cLcu|nGG9=1SwD$UcxEmhg<vOclr&Hl7Nvk+(RUGN
zZekx|frbfLCs}5ip(!6qMmD`vpHi^U6B*;&GwWdD9_6Y$5>sZOFEJQMI--}0F_3X_
zDecFGS7%j;IXQ^8z@!j`X<WgLfn{Q9EP`VrG4nGhnZ;Xp?o>Y?>OV(Df_X@}sFj;L
zMEY&upXbY`!^_e1jjDDt`DgPtbU}1RROleXnz|)*H<S}eLENOpJD+NQ-`O%QRp!S%
zpP;oOjHEfdNk@M%wzZ-Rbg`;hfI6k#Mriz7Xwj`QLgKH{AF<gHyt({Qet(LsiXLr{
zBfW`t!HHRHERQb~$YF6-T&ytAMYDN#^i%o^IRl`1Cjl#b#8uD{$HgQ}S%iSmuBCyz
zrYCB4@IRX&7+tob`tKsjW>&fKKF==BcXg#nxkCQzW-`z$Vo)%|86%Ujgm&hLqWe6P
z7<=$M*n4N2O10M%2`9j1W2R;+A)0Nl*gjp6>HYjy8~%c^zx{A#f}!SSpwNMV7MP%5
zC~Zs-L#$t!`vrn*qgl_cYQmk%6l3M$k3<o1Tgvz=&hYexONybq5dPVqf&xwui<fs%
zh{8bd5WZP^FNjd{bJF%28+|V5WGf7^tlSc?^T)AJ|3tVkI|vkYKY>ry)#vl|?QAI@
zXLNY3jo;r{F$$|1R~cv6J>4aN0-Dd&pl#j24b#Bv+g_AsvU+UEL?^}5Kp_dc>34GR
zOPsLQFY!?od;=`#q&?gIHooEW^3r?$F0*fdgjP)3DqsShC~%1UZAVV>)3$vl!=s}o
zv;N7JNlBd)q%<`L5;}hRFn=+oS!=<O2`5jq5E7nEtU~xDzkan01WF!8>PJcPCM?eI
zRXkh{GeJNp<WZ0`oPl7l00rbo>v%cnCSN`@Sylt91atJQr(C#Xqj6ya?F}-yp2Km8
zrIvW=jK9rCc5$HlH>oI~psg@ih`Ut;B>L~(+C8_wwb|3bOOyHeiv@8B<Xr8aE4CFW
zN!@+V;p~1#Mz|P$=9N>=_zT(!f`PeC#E>uRjXgKu@aD_!@iAosfDzhT|Cn{H&uGWO
z&7BZc{VC7z+xUhg(dW_X@JxUTIhNwSdLMF^3GOJ$PxzC!0W{;0t6l+YZAg?W;pV%A
zYNR_bMx@E}ajK|lRYh#NHnI`Kj$S)bV(D_^y^VB5?)+Mh^D<HhU;KSdRCGLeo5QG;
zr#$U(2*-5xvN<~W11Nm=tBYc6WAtBzgWr*@fdwQ38#D8NJM}ZtvHtIB57U~QiCFCj
zGcT%iO@YJv^KQ!4^%iUk6&Hk>DlX3P*PTO<Ug1AISh1j(Y5Q@K^lg3z4t{_gyibRh
zIq_UyJ_)6?w5hhVyxmxA?g&*6?@!mQi35P)Ka}BU+1&2ML5vqq*B|GptQ5_Ivmm_q
z#q_WDudaQzo<7W!d(gqXNgTZ%a^~NhJ}TI%#yUN<)6-OHpJv$V4`Jc9&M8B&gXuG=
z&i5G+uj(_sJ|AzwDkwGHzTeLeyJ!j`mJ|icWT(w>da^&C-w%_u4^On%pKrv$;rma$
zT>m^XYV2ybzfPZGQ+x3Uim1cLfoLq)1^mDPEvyhQI&*ZZoe*g?mqst%qxk^-0Lhgc
zpPNJLFCrlR@qYb;)U)bVG9>{31yP136bo%2N9e<$C?kNRT8tJR#7VM3rl}yy2bOtl
zDq|G*;3H3uL*B2L{5>vJvhM4tVU@y$C^tR4-#IVI>4&`&mEoBGExhA7r~oeoT~O;+
z@e5&d(~t5+JQ#tfp6c>_;1lcnJO1I-(U6lYZf6oT2eZ%kzU7If;yzKyPbB-G3%~mm
zor;24%`m=tFV;1kdai(})6ikuF-`MPQMGbjS~b`s!`y`Evl2B2DvBiJRKITqXUC+H
zmX{k`h)hGvaaFrJi;)AJK~$$OG(p;moYeS3+JwW8N$x>NN3lbqtOc^Kz{@jOZXRZc
zOqs+YXbvt>mMfWUacS*?CvgojV=LDIMBa>$musAW50dvH^+CqAr|`)XDoqYRYSxKU
z^5maqhz=|Wx*S!tM&f^RNiyJ6zDpM%B3NOP)O4`HFbT?J9Ne`1{RwCW4*puYt_v?Q
zyIFdaDP4=HD#J)JV!u0F3F;51>_;$_XMxFKQpClNS=s-F(3s5N**vXQtNl!$C_-+I
zX|m5@!eU^g{O!(zRh0cA%T<V>F<rISS*pPU3lw*9HHGC~onkvRAn?^TZwWD#>~6gV
z+SeCDyrPDo;g$*5EDr%f?OovpV0JgqL57+CkM361k8~0!$C_5KO_J#bNrS+G9tccx
z4_lI^r=xrm&NW4N<k+IEkp~i`6NLsWj|7oy`>|?HlK0luB21FWb_#PVJiN&XeLS7s
z5*U`qc*}V=om&#D&iMo?vn4_CcJQo=kJF4!2Ubrg<@BnttPs%X(qhLKY-A$8k&~9u
zi~5|ykj*PbtIZKz73pd-$zUL5Fph*~D;jHZlh3SrE?@J&10a7mfQn}lkss=CyUQ08
z+Tw4K#Nah4Jeq%?2`A>3v%5?QF=8tV+B(xacOnlwk6srbQPd4f>N6`?9>x$xj)|)d
z{%V@>T79Ri5K!c`wKZ&!qiqy=Ewv@=9rYBk_P=SI0#h;%$94O4$Xt218xOzRyzjcP
zNiVm8>2mb|>~nPEae15E20i8x0wUo-@N|7p4E1pYhs}KTt%1KYtRtQNlKOyQ*g=EN
zawcj;M*UpV8^{xRCKyg2;9jL9Z>&!`0xsLmF53$>{kou6&_%s6s$<rcI0OD&bgRX7
zZtGwQ1($8}H?R@VkxbODrvD6~H*iqr6QRe&9jhFR`=(LV9IK<Fka}#h*ir@t)8JqH
znm$m^)ugs4xcpFXhwxvr(F;(@rT|X(?*YZ;4;szzK!JnvZ4vd!fJYi<sevb@mYf@b
zTPaR0%!;xm<(-*dXbdHEO0Etro18Nfs3k>|U~Wv&!9~<KGcNek|BM|McxIGiA}s$6
zaM`YQtP(2jdu|&yD5e2F8-lFo%d&6MygX7Cx5Cgj%yyAc;xx7&o~6=1G|L84hG_FE
z{%PD3Y^YiKD_}{+*DPw0_MYK5u(nfyUV<w__ilZmTMZ&e<ZZbOoZ?tk%^W-~UALVd
zgwnfY48pwUDwpTy{HNI?=L>{8(yu2AYkCP*s{1rz*`Y-rokGmEkyQk|E#Qc`M~vUD
zbf4~#&~Y2H|JPI@q|jZi6u{3-BT$n-bEy%y15=BI$8%-Lv{)1fnj_tT1re{Rk=-dN
zs`NW0o6euw2~9;R)8qI2f#Z?LT1bGqL>9O3x9<@%u)LdYJkL=#<ewfsfuQn4nk=TR
z&;i?EtCeR3-Eo5PX10HaL>A_bPbsG&!4l|XmyP#xn+D*g3IR{;c5DmlWLy=b1>*j7
zbv6Sn$)ni;^p1>W<V%&8Ey;|HCFy1-Q)I_;(@m)>R!ojS#*)uNJXLCT`M*0k5b)%D
zBWIv{c;l4W<WOW%kIQf}6>7J#Zc9Z&yr3j#2QMu)%PES0oR4l<Wtp*=SVjdDG5N$C
zs3a6^R`i~Qjmsj2+!KXkB@Yp&*o(s+GjDn8A8fBO>C{4x6{1!NHM5C?Onq7dr=IgV
zyY^NwwwuzJ`!v1;7GysojQV|#mPe2nr*VP7Z1%&{Q-&wtkZSMU!c_w-H8$dcQXIQY
zvE{!EY5gqU53RjB#xgPnHP+g|VN(8tbaMUmzWCNV5^z}#l%zxiKR&dvnV00DjoJhZ
zOM8EbTzS%K8WT5#6U;kh8IKF*nH6aYn*_6BHd`@LERzlAN)g5Gq7<%vWmHOZ4;ghh
zKuJJM){q?+0g9f174VVnj`p5Ldu4m|)3Q~LZN>G1Z-Sj*1sQmaw3xSJNep~c)Zp4R
z6g|$IFMj#nALbno|0Gk9CVda5x)M?Zx2ziLd7#W}AA5jbN+(=sS7Y$W{NX0U{z*g(
zEY>9N*%(2qg=WIAIQ$*O3@D4g>~?tlWD7WyKnlRqh?m1vt-^_p?kNg)7&Cr%bbV|i
zJ(Os*($AVMymjV$ZL`9yC1<@YS?aOE4SKX7+SDc}U7rv=TEt8*+Q|x7=zJ&^h2F5L
zvs`IKgGKhCdhW)uX&QvwwH=T?FjH$=u@j_n+JtuR^z*&_i@}Gz;@bq#UFcw`{&(!L
z!7a68o^^qcP&Zo_T|(Pc3KUQGKxTYyDT<lZ%%NcEu_F{g6Cyg&!j|Qjm2T%k4afO*
zRJHP40;Xkt&qJ*i%Z)u&AIRs7n73mnyCiaw@+Izmh9@ERxlK(hT5p&0{M;C;F6yo7
zir>=5I5Z;Ef4Z+v4x1=r>42C9Xh=z;Tb33ucp__p%G;m0fcCoQ-nrisy;uc5O63ts
zJoyQ?#xIyG<FzY5%VsMWr8n1hqQX*vK-6lT$fMlHYy_Z+bYUHK@L8f@S=c?Lgo8kV
zdXO#VfH4-XVw!`ah{2GDm^*HVYD25I-Pq}vlgYpqY@OnRKl4-#NAcv|qH5ZFsAt*T
zeCf7u#*%;4u@+}H*DQpTStM{p!OR2M0=61Ftm;$)PuahijI!E}Jozv`Ute!2)7n;@
zEX4?aP-M+j=dh`Da$>nbM+ScLUx6@zUZYNcI<7=QC%>jRZnc`018W*a(1MKGbqp11
zxDjY=`MWHs)L?S0t1>fX5)TSaQ1R?gnXN60qw+D~LruuFG;1w#%Dj-W?xKSN%&kZ&
zWk>8q>p737Tki1-2^zI8!M_^c;}vx7q=xQI+SO5hg0@A?zKN;Wz~=#r_#95R<);`w
zb<>?+eqL^%pE`X7Q8|ioE&9rNiAb_#Va9HDfLQm9T5f~H3{y2=+DvfwK&!%Jb+!2g
z^kU=|9y`imT=l;N<95o!6Q-DWGlIt4@qu>&pot~0^EBC09yGZg(3cH)c@N9WO?T-;
zc6PY68J)IPmk!txb$MAWUN20wopdwJBRb97o#asm_oHGm#uJ#Q=QLxN988kSk}hvC
zhay?Mt!X_<MQO`L-iDNp-=!#5$350xOA+yV5geue?G82fd}jGf4g#R(eN0#VejK@u
zZGlnnq|<}oG=Bw}CQ8g;)0n2ZIgQ;u2?B%%{OlUo;;W=!w$N8bKby^t<DCP7BqWr_
zYVBIdsj%)7=Q**10L5K!l{Ud~YCF7%R=*xZO{)94eveh#8mx=f-0`zcxFxDv#|xR(
z35H4U*n|RlSlG?GMLzK7!TpEvCbF7W(e_Zz1x=FPpoPcFKn<kuE)UsE8?2&Z#w-DN
zrw7LvaW>di>BELt?D|UZYR}$T3w-v8OP%{3_GeEhipVMtSGBJsXFL8=bt3Qj2jYv@
zp8Krmf5H?j^#2`G(6j#UDhb_Ml8J}^$!D6gf1#tpH0z>FEKg`6QhiHe^K^Vm8q%yW
zg1aYqdozc(^&=FHVFHU0g12INza#^O=Jn#d8;q4Lt(E1~BF2Sogp6YI`nm>q@>|s1
zTPj=b2cV<gzx(&GpNcjUTyiOd=0HQD9nz!uu2egV`1c<`2auaL3wX~@a1A4O=$x$w
z?MpT}c?U2zyNqm+XLvW8c7}%q(wPDlTZc^TZwK>v_$Y<lNz=~-?|Wx!t?ZBc1zAx)
zz5lOu@2%`NpoHE+MBvF(9Psz+>3(Zz@3WGny*j?VeRZ<M5J?zEcWtxb0x#r&Kt@M^
z9f7p&yiq)9rBu5rmsA)*{DklKtIL^&&v*UQFg{WJQz67o80kw(g1<eGEih4n912A8
z$8D;cZ9v`I@jrD(ID^L^wv*FoY$y<^vK)|amdQG~7z}0xsgBaFyC;;8)<%+cy@W~s
zTG~dhctl1h10>y#A(G@Ggr{mRbMT68*t?-p)saK&vFXKux4feX2Jw{m5&BIOHH!w0
zGYV(Ou5d!a#u_qk3toOW|3>a%DtG|?y|!~ts^-C#a*(I$L(Iu6)+$&;=!>+^*wEcw
zswkdOp=7mxvb8#_U6gl@#p6tbrV*jMi%uCS9vtGPb970hvZ9{dNVNSwvL2>HSfrLk
z;m{Xj@FTcB1}xpTx#aP;2SRi*P8mZ_J{SJB{uV2)xSq;1=?AKnELo-N_iVc*!6zSs
z!}r5|e3-E#sKexi70BQl^0gtD#%U@JF$PqkaP~uzokI-!O!vD@fA4bxL^(6WA`EQ~
z*C|fy_SzqKN)~Ed#Wp?7g2AYf755t0k+(I*a2m|Li6o94St8{^76<{oQ3;3(gD8v<
z!UZs9Wd_yXc?1_AM0BPSf2%+~uVCDt;T)cLPrK!pFI>>PpC5Ouxg}NbEyM(bR^y^_
zbn5t1Z~^QNx&BrJ`9+4<+b)Q#4kpl!LWc2w`7e}YT`5q2axKIT(s`_7LU?kYbO_qH
z{8$7|oT?l$kBkFEQpA}y;)WYNaV;O@i|%f;$E->03mX(bEmFsI6jydAZO8?X|1c_;
zkSAsq3JCbM6DS4gM5GQ`IrM!;C;D0_*3>xa%ARBIlnV-Zz-{Ib%FB;dv8<XLFKP<S
z&qm;-?iN4?dKv$$Bw6V+U^|H)U>38qkFsyCQ@V|qqO+$#Tki2eN~9Ri`ia>;sf8pq
z=hYo0;WlOAR1Tj@gd)?|;azhCQ6rPe&}g=u15^N`vQ1Gt8_}4{RTIyl*g0Gf;0t3e
znoNs_gD5jAcgbw2>%z!>BA&*vy&^uQ70sOjo84R&@OBEdZzw5yGR-Z!j;k3!#d;rL
zbAg1NAFo(oGP_q0Ko=Bph~Lnb7$U1Ofu9vznCR9^@v8@3mupj{kO=GNhwh|uP$qpU
z2m#Ys^p4srNN&xP{=;~nyCgvK7Nt2I3s&P~ZFN<<)3_l+J~r@5OQq#K+O+Zxmto|$
zom#4Yl;=QYET`U(9Q$fLgm+wyiG&U1;+(-?W4zO<1{_nezau^DhHEviJ`%BEPH3NC
zw^3YkBX3Qxif2f`YERa|JE}12<%&>e5If&fG3mC?*Fmv9c!nL8VL+=)>ub0=JP2z|
zH`QW058v|dZu4W1!=>ET<ZGYkSvgV$B>=e5eGP}A;~utzqSL2nx#Xtq_*G+n@G&Q?
zeUdYWq~wBLbxkS(*gyQX9w>s2f{-hu+oVHS1R|b?(p>4XsNB$AgOX7`R9h|7OQ|wa
z)t8-#V26L9y^_Z?0q!Z~+LAgDnF%rp;lWc_D<n3QFQ~d-5T<BzJojYiycX9~LtV>-
zpZ}*q8#A0+HPsY>`RQM_LbySBvN4(Ufrt8;-1wCIFYN0;*B=F0`RjeA4ssHhM2@ek
zGvOIAA_2&E|GCKa1JaO6j@V~kL~BR9-?v-(p_8PE&|O9zbGjR5!t{jL0~6TDex}I6
zE+%)BIQj)Y0QCuM%%?Rf#h5mm_|z7U$n(66lHnMPDk`<rhcldy6wejXXNDEb-8v-5
zjt*<D(oPP`ZjH7U1^X`u;XZCqU~v<U6p~WHoN}aK6zdLCKnUoN)kfi*CUlj40b2DQ
zC8?&BVegd+9ev-{_BLX2Q__N@pn6_IwD}qSGH_@NIEv-C-B_14F>57cIZ5wxy^Uon
zxN6)yUoj8Ym<Chv&I~icpc0!LxRj=UjN<nqdK^uCvB*@hXtBC#(=~JF2Ktd{-)6*`
z&3p}CiGaZe`j{7cOtf#Kd_zrNpz;|j5!WfIF3$B-6N?Gr3^iajosHiXULp&Bi)Is5
z+<06w4=6(g#}`dH4{hRnwJ<WB#(H~mgW*&3C@MBac-qQ@One-rleC46z7#hvjmve)
zJ){d^QFm;A45bIM!5G<5S9i9kqx|FZiG#obpT`7b90V~V6LnF1Al!gWYm{nTVBD`U
z6=y$&N^f5pW$|1!-K?R9TTD)~B$pL%xrK~bS;~!jP-9F6Ug)+BMy=kcV8s{uRa>Ty
zb;b3J(0KX$O-wbjesU8w6ruR1ix9A*z{XJf<5@vHs4u&wf9F7(=GG(~B-+AbXdii&
zafObb%*kqptg`SBEG=saa0jG!a`pMVTB32zC^9*VZxqB$>hA}z9NotD&tR5RVO+G>
zh!e4Dx8h@GF0WOMoG7*Dj?`tSgl`LYyQ+R{(q5uKPbbasf(jI-qq>Com4DV;{EAiP
zQ24c@!ly<I(f`?)hbc9|_O@x-hYPT+O;ep)Bxv;=n+TtaFxQ@VG)(%gZ|1=!%Eq9V
zps$BAxAE~FW*!z(WIDra*+d=P0o=5CtJy0N(D^~%LS|6B=vHK?pEsXqp<`oB_EH4Q
z(oEzoadL7>v2jWnl@OHJQONAAM4LlhS=jMEvI(!n4=0%PjRF!Lj|#~Dq{6Kh7c`Zk
z0qHc<{LUVEfID5abQ+q80j?OH^*F{9Eo2?^nATRAd!F<GQAB9?8C5Gqo?@GKTefjl
zRK1y@{u8<kDQ%j=Cem&AW}Q!JN7FB70WY0L%~T1J$%DY)*sFTgytv(<+m9?jafi-{
z(r2}<lcmaJ9Z3bt%gYo=%P5?OjW(-feC5IuUI8C@Spk8mj3D<&F<(xyHxjI_&D0X$
zMK#lTuhyR)+OSX0_Q5R(nwc@g;-k0%#XP#=6KhzYv}uE8Zx3B-o_xwg2_fQ(ef^@u
zQop2kVeiD8%Q!i|aIFSAiB-Q7J&w6DwzT1&ew;0tDZbUBKAD(&;{&Ey=BCTG_UC)3
zhMcorc}os?mh8PT4$<0I1K`y4o;Ig+jEGV|^5|eIcM?uBTPN52sa+8X@<I_wG1o=Z
z<wo%y{iFPu{K6@NpY`iYrK(l6{DGqTvu37qYS_t!Q4!`8YruyxzkVO~oe^grCR^5|
zsQxRvisA%F7_ufSp8Qbyu}sFRs9OsO?o6lR#fd$Q$va{QP05t)BHB1{istypf|*jJ
zqa;P1=ccgo%V;e}DctEqrr57(DmjD5)|&AxOI(E64tF+gD8BT1+FqBZ?-Lx^pnYn%
z%&u0XD7&)a?)sDhr}r8hWv@xHk*M1t0Fs<6b>rE#7)fd2Oh&ngKDi?3G_2j5sqd+X
zG5^)%Pcz2|9ia0R=hDZEwZ;99i=hIv_TxGbALTjzm7H(M=zLeW1i(X+;M_tSRB@JT
zX3^ISjNA(%Q-uYV>4u5>j4kzsoKa$f>u44!XT^HaetH5-X72Y3dk+Cz2DG5Ya!-Cb
zJX<m_q#(!r`3vvY<zoWV96M>#`7nJgNu!MXEEa01nA!1_KZE0OnA+ZA8oND!2B?n%
z%C>AB8?#)-LHnh8x}6<H?*&^32QA?rRP?i$;N4X9yUJ;gm_P$L`&6$EPSo3M?q02<
z$yA?=GFC6@LYgsQz?dj~N8~R|Q~`Ap>L#FvW58w;6=)#J`8PWhm>(@GIyUIrWZ@_i
zB$HAqa0Sa5zKWk(IrU0uZ;tUp2GDW#@Qc5uO2^Ks!9|f8&x^YEN@5NyE3o9+7QQ-Q
zJM3`Qwy?wE<5yJ=rkx_FEI~f+lLf4QDu%oWFXH$gbj2xZ*HXbtExt+?+q!xW?#puU
z%Mp7cctAs4r{l#!keM#zf9=}T(Hl=>nCOb`E73NAr0L1b2hG@+J5SWdAPOzSYS~cx
zhnu3CjZ1iAFnG&7-g);z*c$&w%);{DFbfkCJ^TNP{WPsD-bmd1A3YdmyE{SL0+J;u
zd-sGLo+k+X5LfN`G9&PQ?0v(qKOUtml;Xu39TwQw9`DY{NfO5;vXU0B$D>fcE*mG$
z_tU1ekTfCc$RavC-*yO30gJr>i-OT;M8Jy@T0Gw#Z_nwD<>!utaCSAw-q2sx=UBQs
z6#C^sdAb5!U5xl%Z>sNG=iOn)0^Rvmd45mvPrt4V<S$tg_U~qt<<doxvq6QkU&o!X
zy50NbP+9EPKTo>XwQzX5pU)-NaVII_P^rG3?<2hcBxwx^@8i(Uk1xJDK5mcFlFi7e
zD8w{1%mySh{o)~Z^BeTXy2AGIyKD-lenY@Z-fq!TjCmG=yuCke<~g7LCgW|3!};eG
zQo0Za<uLu!S5Z}vSkC&XEf8<ypTBJrUTmJR*F&4FBaHyHpRpOhEC%rILxzYm#$wSl
zUw!c2gn#fayb0}^q~cQ=`kX8FDaM~Za6PeeB41}-a6J{oSB9JNYapEwqR1eWWF;Vs
zWcujjOyuHj;nu16j(F|(eSu7DOpwpeDe8+^CT}k%Ij6fz^H+C6NB^m|lI7hua3d$i
zU`5<^eieVy?MrEHT!!DeqEZyOu0L`R1?wy9t;~q^^L4f`dF8>;CX*!c2if4#u}8q@
zS(!WJ6~+S#uFkIozV)vWX7rrr>tiZQI&VKE-R^q#Hn7q0(Vv6x=2e${XY65I*pk<j
z_0n7m>s0cRkBHQR-tRQP<28B|rqi2HrKU`YOAq0)dGUqI&>xnRjKNVT*iXYxN05dt
z?BE1tq0@*f9)*#|WAfwvKv=8Fdd!f-=i31>UpJHG5}oj;`+}S{%rke%5Jd4Mmn?>%
ziZF9_5brcf19Plu&HnsJ-TKGU^-|+{nbLoC(XK33sAG1nqk#SfimlBK2Ho~NX=BaV
z)m~D{u-=v-3Z%a92OU=Mb?x5*JMO47$~^`Yp`AE6TIK*;+X0=Itpy5-KHBDUF7Hai
zij-M<y|LhOunm+i`7#!ASRg+y_I4#HDZT&WRj~V9We$A*tPajC{|)vP)q9}rI>bOm
z;Ee*uiBx4U0rE5HAIn5iCKPhz;1BX^&=-jq5PhAy7zu7N^sIe5hgkpuMkd%Aj=6io
zhDm3@hsU!ZS_Q7u7<7-Pbt(*qe`=C2$>_*(__gl7@5k0v3%51ED0P~P+kGWEHqa!B
z&cmg)3Yr=4;1ojZ8j=}=EM^q~wCz>vh4r<oVJ%F5nYiBxTMasu|27xUn8iP#gC&&K
zroK63rT3vF011t~JS}bx&mi%!AX@Z%H?Bn-1pN)F_WW5q?fu(QXm%0PWq#Uv>;kOs
znk3B`DUw%YWMBZK&(wLmz8MM4KTNX`^pL-Jp?#$Qung%^9~px?fsr|FQ#|Tk<~l?t
zvIB;Ov+Z8ai;oOs_TD&!9c*M0ZJGX<7_@fQmp5r|2=2@d+BWjoUTf%S6>7(J>`xjr
za~3YpI$T`512Z2lLg}Y3PBE+Ipc^pf9puPDW%9~A#GrK*hzEt@+MUAi7jn^F6Sx8+
zCp*8)OhaM;xx3W%TxSXZR6oeWhd{~_1$96t$bAODu3j4Ng2UlpHCZFV9<TF4qra?>
zm|%K0%GV&Lbp}NeGd~I2$2c;eK$dZ)nxbwA(*2zijk1%~g(_Z?Xx`F7Tn+kTLL&pD
ztOfPqSnu3S=#S>A25v&xs?<AM%`byNWXdPDuanPEW(Z67B%07S>f5-w7A+VcVJ@!`
zi6G4Q85~FqPu3KeXAoCT(+}x*z@h39-tTfnr*PCyRC*Py>9~$waCGifkj*i-)z9B1
z$W2<!s_uD0*Yy2%jIAh(`GyBfOi`;(+z8^DKcZgF9;wtw@*RutFXD5Hhk=f_CIZtL
z(rdLPH6txC0WVPn<?LaHHL@X)nlY?^eo6aZg?yZR?nM`);i4e=DRr~<iwLC*l_Ey`
z1QQIGShLTB`1%IpflX%oqUK%zQX`b)-O_kzG((Gwgwce<*KF!W@5Irs{2^;>b%N3R
zwfm|x<6Q+YZ@}ee<Kb4Wp$D_7UjicafqGdcHgx{AF19ozN-`hAvu&JCh_!eMNwq$)
zq-6?uxpMsZ$#JaQLq>`~m`$IwVfj>SC6j`+kWntsxExH5T$#b2WUcxKfe?IZlj8O+
z$^G1n={hP%f|WDZzl%$aiGxytt2!CEPayG|T_y;zm%RB+63n*{Uj67;2D6ON7%KS*
z`lTvle$*6ZdrVJvI%(tGNw`L_s|Z}QFT8DK;Ck0O6P9UZu_kNG0K*v5e<m^Hv;@Ta
zW`}m(P<ZD`)RlpZbD^#kaI_~-yu;Hu%w-`o(XA&G79j+a3_P$>5m3phMd+z6b!y-o
z78Zb~rHmTwk?$IHE=_}^GbaoE4pQuSU6!mx+BbWk$xzdbZ6b#=__3O)k&8K$N*AL)
zS=P~maBT5979PJBq*s2LvGqea)S4sL|8R6UQ>We)rv&c1uuum!Jfsm_Nu<daFfS{a
zmROd5#>k<aHxH1rWBDB4Dbk9u$y+Jhati=3F_SjUW#Px+ZX7;91W;#sVaZ^)4T*2F
znOC;wx76*ndP;A2q@8V`4jp0%4!mbCL(B@`?bO&%^vvYc$06%MPitjhYF?u?Xx*CE
z4stj3#<PBD^uSrUltlNZ9Bi==GQ2kxAL_gSZHDofb;4+@HAknhfTmU#Ykd-@X1)#{
zemwFq9OTLF<l2_muwg#b8Q-iD>H7xO9LtB@q}&aqTxz&&?)tRsi-@cO>Flp5C3}rf
zG+?Y2qoLxurZY}=a2KS5x7yD5@MWk?#yD?jS7G3T>w9{jd1q5a98E|w9mfLG+;l^{
zqYk^e2BnMyQKy)MJwpKtoO6koA{MHz&E*vLau)JHI=Jnd`+VT^ddbrE77X|Rz3k?n
z)e3TcKGaU`a)GN6NR1po1QxqDXb?`(Ca_;3uEzsYqFr0lhDT?;B@<wtqmEOK|9!I&
zbe)VB?=ONgg%5m!&pHelrE_c<9*2<6agcMzt&W6Ij`ijpa;(ryX;J0@**D;3hVvR}
zrkx}nraWg&P;wt~xGU83E|N!uIXR5lizqLbMrbg27v`FM%oc}@#A=QIac3bMhsh$}
zmRdt<cmi`|G!aKi>zW|Y{_|w8BnGm{JH+6gliUt%>T$05s4l+vG{nKCys(_0cQt6q
zv)&l3s-1n7ft0o|yb&G&eY63g=KTzLG&Y41*zynVPS9N6Ox9OiQ8(avn=+x|tYac~
z$Ht-<{$WLBg0enYN3HzN6MP+R?B7TP$v4NXN&XEuFebVSQ9@LPBhsq_w&3IEs<z_(
z8O?BA$%cRJTW&tty(6}N#?JGi(SH`g2dKS6kd%QywDSGJ2bi-{dZpdwuqa+PUbZ45
zz#0Vr5r6TUhd_S|@(>runSvdb`&fsWOiZGBKC$BcCQAuGX85}^kuS%K!@#aJk#8Eg
z78&vl6<3+iehyD71V~sh>YH5j@H!33+SnwuxRz`A9;^-*%GKr6!je>4_)MPZCY0Aj
z&60kgcd8QI-Fl$`>w|;z`=z2imb-e1!}eI@WXfDIgXy9+HAcV_TUJ)JM%&5Q%E(Ch
z^rSn_M?A=>RSp~&RxI^GlR^?mZec4hjdDYSm9-|X2?!EC<DZJPW)6+TEsZ7S6?}e+
zBD<F8=p<$q7KOEm5s?W@PW@3pw>oG}BO%~iWEXN6DK2o1JG!&J!8{<iKv*v5|GY5F
z142O>KCeP`+S8x&59J5T`NMICJm+8P2g?P@!N;I~OnBC($oK!>265U$oy!MKRN(%d
z@AuJMaHl;yASdK{%R@Qsm9;Q1NYl+VG4)@!)8J|=tD>rUG$$@tSXBPdCR{i#jWGgt
zs89ynKREAnXLF(ap*+wWPx^EDzi;FX?PNEqR4>0B0Q-L!d#4~tqpoYP%r4uuZFkwW
zx@=orwyi8%UAAr8w(Y6+pNpA^?~Rx*FY=stBKC>(WaP!(XRp1M52o`;f38Rl2=@ck
z|Gf~+?7AVNoruol$phoPq5i)YPtcP%UzGU_$e+W&oloS!xF7ysa6%xZ%+{AARR141
zD9*gU;ME924TQr;v=`eYt=F@k(=+8p#9fcqjeu}$MH37HF78qaCB6<2&Uq;7eukKt
z@Z+LR)fEm_hV@-M(RnfOxo5DvIZZZ7N(N1cJ+_%hF@IB^LxLWosDjV;sdf6iT}yFS
znj-sH{v-cJ<WhV*UPNgZ%TOzC^p05Q{xJ4zRU%rYPR?BzC8n67NRs2!ly<*??v!h8
z95T+I@l0=;o%h@NI5wJ?GHc1yV?)hgy_295_u%WIgTH-}1C|b+h*|ch&AwJ$sY3xw
z;JK?wrKV@qx|YeQ;*X~$HUzcIR__U&5+)G)k+vscBV&Xt+^3`38SEHK=&4gSAG(w5
zXMOb8N${*<KW^;0Otw4+DOSpF*O2PZ<IdycfJ7*J8nLj7%=f8I(zcf-NYGOsXQVSZ
zP+u^Nltz;p%V6S7?R8+@f7ln^TD{Hn0`xG6NhYWCUN%hQWEseZN6?<?htN^zE$kgM
z&V^_x1n2&Si8cDV$#ZbidMx~5&+lFnS<crP<$ktkLg8;JG6w>eQC}aNpQiDMH9jQt
zoQ!@3JGk051CitKmHZ?6#|330E@HEoutA6%4AamPFGn(iG*7d-LM`XT(UVvd-$Hf#
zB-epy)U6%CS|srkecW`#FvL;QpwE?sF(r6TGVBZ{<6Y<K_8`*(>695c0)h!?8lB=g
zV}LJ?M|n87iKw&9+LeqGY9C>&%^fnj-o`#{mNL(#N_KM!!;Y$F6ixmK8d#+qO-7AY
zV)<ZUkOH(zdey#6N7u>^=%yZG)xI7XnDT+y)?m&*+Cod1nF<@c3hYmJ^+|zANiQ)>
z6f$(W(p;m5+MG)i_hJUqSsY|YThB?Zdtmpm)|!K1UqVkDm}%6e(JZMTgz$u>vj85h
zg$#(g^lHidK5_ntEq#Y#q4Y2H>Jp@YkLU%eEaRUB5WsY3yRh51#P(=<JnNN6q+B=4
z-+<D2a&_8nasGH5gAJ-sqe^#k!IQl6>6v1rEk$Eg6Jz40E$<AZ+E9o3KCl)Sx36{1
z+P?)Cr%+$9^!5GpzWF+6i3XoWi23A3^LL{re`*0*P^JM*iduDc-Ip^>Q{@3~QRk|+
zFWq!S+b9~4<?qsOY3`h6;JCX&=dO^okcZr7$m*M@=H@0oPWxapscI`q{I-6U4U}c%
zEXf^73}5!5d6$M)&%x(+@FMSkp|SuF7`H1Gp*;DK6>7=>`&JEQ?iFf=>&6Q`m-m+{
zzrbm8fl0+|_twKHfnBEp%JK#ke%xaQU7l5?u2>6I9u<~xzo*BxWRhm~l6Fe%h-(aG
zB#OI=jRP<n9GqgQ9F2@q_zRSx^n{3JNvE)Ums6&b5#|%z7ABTW`45XCDpe4n5=kd0
zz<Qyi#Gv3X=yS>K5v9=9k<q54MF+p?Tv*B(z@e=vau~U~J}ILs{FQVQ_x2ona&vL@
znLq|KcC@L=oS&9_eTwPf3oHcHU&NLI7q6vP5cri^c1SCNMP2*-dm4D+H!%f#;eig6
zhHeVm)T`N7W_0xP4h((2L7IBR5YDFgTs8;a9%`S!>|w^zqCl_fAAa6&wmUiDzeCsO
z&NH{=MgFd9_Wq;rzdh$x6{Of|CPl@X^QI1j(REq`S(R{Vta_=`J60w#Kh!Zz<cG#h
z8TA=R?~54H5P(o&2@k;!Yf-J+-ym^U0qDu>p@$KzqI!;#;a?SG|I{Emlf}fRNQOTf
z4*9A&D985k6kjVv07Aj%1Nx~oa781u!I!L&-mfV$_LK3})a)LwubJd|-=NFF-+uq?
zLWbkNjH7X~{qHVhE_HZfNt^#eOqDT*6t8_DXVfq#c3b|-Je~?2VJ#WX@DKU@xILoA
zH1RWI%VGxv+=oTFF7iiimi+kT@ENxI<MQz9@h}vjj6Q`6PEDul%N6)y=Y_%kr_cNc
zI4$1(_3iTRKH16erggu7GZmR2avRV0Y?%QWHwNefcIPotRd8S62J0WeUdxRh77nGk
zH`vpLaO}e}Ys9{hePL}peG_mD@ax<7GJ(GP8i?7}jQX4Tx99iUT~4a2$dJTPRr|;N
zI;;yNv<$ZcaDNADs(^3j$J5<&ie>t^hBc@M@X+fa0yd+?D^=|Ra9t8k!DCQSPy95D
z`c8Ub^?=ewnHU76agb*}<Z5BuB(`QnoXeW>y77E)eu|9eveE$*B66gPO@5f*P3$s5
z2ny}Mh9Hh(MUR?necKEDY_8jByLW8CaJ*qQZ*@48j?60sqsiRee2bnTst8c~W7p)B
zoJMa0;-&fLQZ>{XnnQ-xyZ~0QBy9CbD8lqHW5Nbwl?_kHU3{jNyjojO18Rar1rK%q
z@iGPYzT9IH^tc#PlY^#Cx6(ve3~Vkwz_9QH>jqvrO<XTQ$7-GkK^g0Vm!s;BSWA8D
zDXk?HZ(=5XbRDx->nYn08T{bX|9N1PtDyOiFe+(+Gv{<{y@K2+8XY#eqf<<~SY^kY
z`MtMpvp^SH`<}B-<&30~(xD+|-Le?9!3h#BIr(rxjb2l=8hP;*58A)_Z{>WFr4El~
ze+Nke2=NMTYUO=g*Y;RgU4Py2Teh*8MG0%d7erdWK`76DMqW4X-qceI^}|;}zb`}$
zXO5xjNP{*CvkjSYab@725Ws7Jv@%$bBOPL1QOCOfKWbbPGVE8I2y8_il$HP%7=|+}
zNo#A1&OtUPh8orG>8d}WVR5G@TXLo0V!g;d$?p0`>g5mVL6gHL7yFqWk6H({hGx|*
zon$UaAb5>Pbas7hA&OdiU23f|Aq0#6ib$=G%sx=R7R6$UTmi4Z%4XTmijqQm!!`$Q
zt3|l2Y~|j8<rS@H?v!<U4fTx=7IH2K@7IM=lFqTpB5L<|>F0_@jJ_Eoq#;<0KCcYl
z@8{*H1s*`Nvd8l&E+A4x<_{$GQOo(e-lp;4^6tkeSpA7mkksVXT+M1Ad7&}3xGbnH
z6Lj?ogQOzBOj&cGK&k1K&uZU`6}Z<)Y!zvz+tsSHSKR@6ov99@rydlp`+DL+Y6L~%
zYDa8TzDTb`kI^65VK2Xi&4QzP#0vn6PD0?MX;bH|=O>w<XZa26-k8Id>+kR3HP^Bt
zS9s(OLL|~<E-sIn#d9D0lW~XZ-9w5Z$2AyBy0Ot239>F0x^Z*=-jKQC)I_B`5d51)
zt|`u#C#9|Ebf6OyP2H1s@3%PjN;Pvwxbfhq3VT~LN9YS@@cR{RPt$~*hAWwPItECH
zm?Q!#f>OA9y!R>@ol@Jhr9twLOGB3pe}UcwD~|<C*90OXu2*pnE1S@f52`8*Dh|&B
zqJkVD=6T&1N5TE~0<MaR?t&i4XnDruWyk+2DJ?qeqF89s4>zE0qepP0V9~A58YRF^
zOFH&)9|iMgW+o+YKg-xEt(70h+2(IB-!!csz8}A)9^xj-sTT#$#$U`O8QWmPph$x<
z{W*Sa(A&R9MrzZrsMKCs%A{3b$&_eCYHQZDno?pZQbL3EK|^jy`ra&%)}o$1Oz7Va
z7MXx#Ro?`-n`sizh;xU?OE<=^B8d*#$1%ewpPP9xMA8{x?@3S5a7CMvWnW`Ea4MC9
z3jut%hXXJQ`ofv<thN9pp5_{zbf?|aF!O@le9FflOelM4NZEo|>3yAVl}<LLwBVFE
zI}WZ|lQNxN8ah_&oXRjhb6eyWp(>&Zr311&>jBlqOk019;3wyMhbz?}y`F9y@uq0=
zw66j%oHBOn%{m7oY;z=CguxeV$RQQT7is^{SzZ1s^Y9WPUFRwqHXmQxdp}5v5v@t$
zg;wF88Ec8XkzZLYG(mO&z29i6AJ|(8K#E3h_SxrDFA}i>xgu20Y;G@Tp)e+=61Fnc
za%84@vs`YhkWqYDxgF7p()(CdV$3#)MP;?)Wf)E{$2~0eN|!DI7RY<cV)&*Ze?Jq*
z&XPKJDhE(S-M2Dk_g}M|b|=Y*{219YeB3cyd>}z4#(i=8`cymGP4(0N5^DSl*GG@)
zRP>v=uM+79gLN6xShS7=nvPEd<Bc;sn+9~6^z4~h6Kvr;<;t@2XIQ|vG$!+Gr)v2C
zmBh$7`HxlG@ebc+(Fh%L4&(`644dS2%VGEhsww`~BLv5-vct2rtrEoFHrZ#zjxzax
zyI1KQVq`kV0om47s$GK8hJc+jEw_1H)U>3}G7vh!DsWS6Ox&YIZiqhZ;)g_{IA?;u
z8#0tEUDHAP^vU&;A882G8JIk^HWc;LwqA$&Il4HX<x()L)F|>em{lJ^T+<Z@_$kK2
zUEFcvEVHFHkf2KfhU4|+9Q77_kIQ$$R+Q`h9%~@mSvkpD;~wzMOlFA<CP<e&VoSZM
zK*lXm3AfI=zPL#D8fi{eHF#0rUAbApvR|8Zkz0_HnO$nKarC>hCcLuSf!T#3oG`1x
zsj~gX#R2kc;bWUT4%`#da9NCwb83K~#5{%a1scKW9hID0Zx!*U*O~4s^~83s1bPYd
zZxzh5Iim<=kT!MZ!Yd+|b0}+Ho`Rz}#^i|~<|co{n&FjgZpdnaW~{7$e+GLRf07m=
zQ=KRshGH^aJhWd0Q#F)1{C;vY>PE2iKa%Zw=FP+*+FVy@`yfN*g{v}}0<o}7aq3uN
z*R*<^urka~Kww&gq@~!z6ycLvHh%Z<{CSe?o76YI613OviTTGvlYs}++6|kzf&w2~
zr15_*QbwN+d^{7Ia0|v!b<;w!)Dts9N@<xwHb)6clrybrHGPY>?;}0fWSa#l+9EkG
ze1QqoA#i1vuDC`08ykX>1A-s8NhiutZqlE;Qin@H)px|1b!FlF!8t%PHW0@5jnEV!
z_+=My&S(|`GP$`6t20nXib?QJPM~D?1rJUNAd=bOMeDz$pu_)4NhvyM@qUW_#T0`w
zHYvFptWVNV;8>DF`zEytl0k!A13RO5IANq%Mgk$|KQo)1JojPWtkc8jA6HJ0qDBey
z%tgdw+Ndt?=xbaEsW*>7V^EbGk1HHhLw8@Umye36w*y7T*&i;6oJ~AUdCb7%%O(*U
zVj=48&Og4DwPREs&OJ)Xr!XH5LKBYwEJ43_qVdJvTY`e3iKHl(19GijTt=?SOPGYc
zuT`h_an}WfhyESl&I>Ptj`JM}j03!+0mDM8K-y3te>~*=!U~g+b*kObT8t=shFzd1
z|Izwe_E7!%Uu(uHSWEdBoh_H!HRy&m8u>YD$i|!G6ZTMl)`lR|9~n1z;oLOr>%a*#
zp21E><l;B1frDzA&lH_4ucI=ZPF$op7Bc455<-z=@eEe$Fbx9HpPz&SG5@?`Dklgr
zRVLa^=LjaV?#*9%4ufrLbVS31K?RS9bYks~Ge1Kq^B8u0mhf8F2Qw=qlE}~%r)j81
zdrUek(iK&hM-*2%4Yw;%g}Umcl1p+enCudkDw>KCzTUF7rB|Bj*z?&zW_ffD;EES&
zQW3ox)QCze#3|09=fPdkyven!IK=ktn|6XodAas#aK<IsBzwi@Tz6KXHDB8=|Avq{
z^tBm2G937U>Yv8cgk{-WY-sKpI-#wLRTQzQwKg@6Eh?=Q_*k;f`g)VLGFM_w1Sb4_
zLp#cn_ZLC2DE{|<O}zi7%>xTF+y9QeIjJq}MAD4Zb*?_4$PLw7a9f9^8n1Jy_KGeZ
zs=~#8u<KHl%{L&a`f)$+&Z-AGHKS9O0LpgoYsUOx*)HnkZoe-T!L1~<eY;crpT<$c
zKSz((7AWtY+nOEs`IAAI-YDep@#DjG(BrwC;~;_^9jXtFkE@&V9Q<T^PT?H}A^<*J
z7~>b8@6?M<p@YLu?oDO>A>r*9C7&d4pol8~A&#ASvd8D}Yq{iFzpHiM0~Vwvzm>mx
z<MZBj1I&FA)opy6e>)g57fe}>z7M#5gTDLYpLfT*2$4|Xj|LCU*XPWva}<fJ(Jwul
z+}F)#?^hoF4+uz_>y+<-yqtdfsg9>{ffNB^pDn3P=p=3qv2(dIWK0Z*aIfON0#X+T
zZ%<Yod`{%??R;+_vWUHXY(JoU5K4<^sM}{ll{mj1@o+*>Bywo9p5EAL!MwjhFnIV_
zee~nlW)V&;&kvNO>V{C$&0!H^{c^j$y1(c5gH%X{;PI+>_jkGl6L_F3CVvZms-B?x
z>ZjZ8V*G_J{BFYzexhj<-b?YXD^wJcc3#5AXN9Db-bbLj|NQ09GUP?Q3E5yKcsavv
zmFW#*SsZAYCCNrsVyCG6yMXPOMs?3r@ghuvioC=9CqFf53)ZJfSGkK;5w<K2rbyPP
z0OE>=kbl70q9}>mR8S-BNuOgM^H)Db8FXi8tDu?f_E_eb`-Rhd)>ss-G(h+VjL}9S
zEdtDALfD#@Bh8wJk~~^$e^!K`nyJyiY!<6O&qoK>u0^z9rlKEr;heBVQ-uf7o|Jee
znMfhQ0xOIrgm?<Gp>~c8Lv?VZY5>zBn-wRvUW0*e$h?&eCowIq8H|L`0BQ=;_cB>x
zgSOU&v_YC`D6<9ql8hx62HExV*3#GS{hk!QS?2wl$OG@BIj|nI%|&9MLw2qlF9il(
zY7XYSP&FZxnXE~$31B2;R-o-<gH2O*@%N81lR6sLEw89HaVQ4CD!G6Q7t!LN(-15A
z-EmUr8E=*6eBDSg{k`kzm4T_!(Gz=`=2T^6a0mbuyR{_@zVgwfq-iFMY@H{l7g~BT
zV`MK>$s9FEqN%BjP)6SzQMe8G{>21plXes@sKL-+6NjjfufCWRy-x;_nX4fzKx_&t
zUlb3;)yj&P30X}AxUGwl6ez}{R2uyBSA~(Xz}o?c?ckYd`jr7V)u|6oWe?M^bUd2M
zBTVI1yu6Hil>;%Oin2`bbXjwm5op0Ed(krIDTKpa&X;SJZOSpc9Ch)$9cW_Rgg*^q
zYlVSjB-X6)ExAulCT1}1BGd-yd%BPtDM2bWwNcMpG~1f@wKtD*bA&r27gnA9t=NoY
ze-;f}#Z4kEi7ArqL%_yOT*cB#nOJ2Lo2$+-Dx;vYPV*3<@~w#de6xsiJ*^ufGL?Ye
zCI5)kJl)4#l2(^&ZJ)mf&6j!=lWZ-s)pR34i&C&H`X)Q+4jzkd$?y0H9^!V>OOC{q
z{iwFjm|j<6ISL?eLSB5}-<Pwtx4#MxOkD**#F%ZYe3zPLL>@#R5Kccx$fxs;6H)3g
zWdXAse|ex>y!Cm+cIDJ6E8X)t)uc%9VN3shZJolMk<=?l)$QD!aRF!d{RV__?44&N
zad9f4eH!g>b17wrlVuino(=&82(8o!)1U0M)vk=UMIt1@u9)xt(u+QJd?gd)@wvMU
z1!)57#f1p^na%Gh+**q&0fOs@d_R+qXI72O2tQ+2K-(LCbvZE((z%sp7BC%U&CVwk
z$V7^0Y5lQ%;3WMd1%E`to5cDTP{}ZhR{bE#wb878?&CJsR%p}1R(~SczABtedU~O<
zL`a(gxC*E&ShcZW7Za;F6rBvIVC9LnWaHVi7{<rx4DKj+E_2EhQ+~n_8%Q=UuCBoF
zl3;P>5oi*0<{#uaj-elLQpk2&3_|-WCdRWAtFFkK$Xnv9_SVFND>)1!z&o2=rz&*f
zG?t~}Yzh<(BV1WNKUU9ytBc($uNo&AeG1KMMuZ+nDbLKC?q>d@-Ek)&&*r`Ef#4In
zj9PcUW~Qah;e}v-WM=G3Um*<oTwNI(T?w1bJZ&PG?kblrd&NQB2i0+xA*t8KEgBpU
zK$Gbf-<=-9pg|xv^Fmvdn=vNqKECg{b+@S~e2cwwvv7{{dd4~)d4Milyp_>L({XTL
zw1+6lp5NeUFXAiAF-@Lr?$*|sL(DRJVnvc(a?l`us^uy}FPJ`Cq#qts(Vm`vg<4V7
z!Ed@!sj)X>Tk4*GMrDL~qC`~DU>+?>pBb-gOn95hzmVjP<})3g`s?VxKmpG3?*Nv2
zcUJPD7>U}B@n46cieHQloGgONSZ)U)goFh8WrEFj#rt7d5y1W(+WLn|AFX9#aUKD^
zScQB#2hp4n<%ZgpjA^z<`TJIvN>Z+^b1-l+srQX&IZ`Wd{8R|1wTnk}@YE9sXOZH0
z2;>h2y@dI6w;6vRc!E~w=S>y(6O%ZnI71Z+7?~S|$lBv<mug9COxG1U+q6sd?^Hd!
zpHvMmTITbslT1gX>TkBngwoNL{B?>_MzoKddCz3=C((FQG!wB*5ZJNDP+h}Ef9|%z
zm+Yp4Ga6FniV8iCi4HeqQ=HV4VY8Htl1hY6S&QcnMx=-}u9f;yH-(k&1{A9X+7=67
zEx71eaxAjE(CA@x6JWNF>ISx{v^JRRRIG~X%=ZR)87*w)|0F{TWM?bo&xC+U8%f-o
zG=2<S1{Zc=Vd(N-k%`UMUA1bXX!%OuFU*&j80WPgzgf`=Jg+r9%86UvoEs^30ANC!
zQa%^cmr<0;BJhqnsQKvb$9*UBsF+p!rnFwC)QBma5<|6)th|^wCG4<{$xL!pahe${
ziRXesruvV-2BRiJBkwA^j4j2<q<J`eZD$^b*6PSmlbr2X-q}mFGaWOnnvx@Hm`~T4
zZ#USk(>tVp&6Q{!rDoeTrUhBC>^@i}GwU=so^mEOCVb?xE8*zo3Drin-4BUx)J*Gi
z+ikgV%aLsy%}&dcy)BVKKEKm*xORlYiFSBn`pW84rVysHntJDH>?zlr`t12zp6WW{
zzlrW5_(o0^Tun8%UY@*mSWMRLgo;|mvi!y%r@(9KI634f`Nd2xW_c?ycj)%-wDeET
zr7obC39ilw$(weiW*>)>MJg86L~ZPb-s7cu;@6fi4lHigl}(ub`BgY72uJf(cH0~o
z^W$Jx!Yqe{`Q;vZs^8l6rQYHylHQ<tTsVhvEO`WL84l)lDZ#ulxVPFsNxO}nypO3B
zM(kZ81h;R;a6<j@@z-=qWGL(S$SUoqL&F(`&)eIsBBs;#eE*UsCPg}vWh20SB_ZS2
zb>dy}f~8SUB8lvKj!NXK<S9KmSGuK}eWe^H+8duPHx>ONQgGVuNtKN{G@XPT=AJ0F
zPSh`9V)aPFE$WhYDhWt9QTN^`1=9N*_QT>zv$U94$H-pvhZS^cuMSZZHa3j}+I+Gi
zc*g{DtmQ`84=r2PE62>=oV`J6jS+Yv(hcKAy%gXvYmo$(tq=m5l&96|0@qs<Rym<Z
zi&4I*MhQ8l#l{Rp3NBH)nh&O>7^T`{3ljFYS}wJ0<kA&#P5w_=k6fE{OJoeBIEu04
zlZ+nGcB+;kqPr!s84%wQ1;4GvE#P-UyRkd#KW!I7QZ(ig?wFs0>J^`pai<Xm;?Ohe
zmYeA8xHV-x1qL212^uA`@@vQf*Ep}F`km&@cT*qgvSgt69=CyHBIA$@R-t?Xxi5C`
zUimn(FM&T1xtUHeNzppLQCj&}oV@0!OqobfHW+Q@J)dpom8Fi;_KI`vKt;(oviHdL
zZPVt&dS8X}B)6x_;61wH_Wr{5S`yV#fuY6tcU+o{UfL6p&YP~x-kU0@C$P+P1-dd&
z2CK=3G4$=JFpc%ds8Tk7Cts*Sne-g8yBVbaO0Y8fY@5Ve3lPVpci@n>vB!I8m3H^z
z=`3>ouBrL@T<kB_()KSU=EA5y(5pHwH-ZDTz_2nx$2GgSIaGj5vV47W5$7Runxmyp
zX0^EI8L?*t-Mqwe*v<4ph!_g~hVv3X-TQ_11plf!{+{E;I?rTVv(5dcZ5sc({V|6C
zpKQZ#O-2rlo1@I6edf9uYyK~6t81UWn+kIW=e*OC1rCbLW<Uh1bc+puvEcfVgHB<v
z2C3{uKx<<Be1NJp2IkwA9J91HN6$bjq;A5+N&>8B?LC$Us<1hDE!=yVA=k>9!n?B*
z%oCjLhP=tGV3uqJP)0zo7b)bWu@aZ`Y;`FyWhm3D-FeV-3*Q)YW=blqkB$EN_Tl)f
zAMc!B`QK_FPL}_{24!Jn{NFVnCUr&~f0_@S_cUh`_)vzU1I!{U0T)DQd5LuQ<fl8b
zDEl`uK`st2*V!7df;O9rqU)iiZt9@n1LNT7(0{#Ocz;a3Gmo1yul3M={f7(+Z`TbU
z;JfpnZtq#@z$naEHE{Ff@^b3`!v%emVfd+mJh9%aPw@z@lm9IU&u|doU&BFde<kZt
zyy*_O=ILADP!e{Me%_(_Dqaya?mf&cC}m2d0-!BPZf;YX32fangr{rRd%`Lv7?cnQ
zef@{TypsSF|NQ&hYP8^ExaineS=4)aPD17`bFP&z*vyZj>Q-E9Cvpqv6NpMz4yDiP
z{xSgdQ*{e!<jH+tv5XT?7ytZaP|gD_rYh5LU^lqY*O<d?rmMmgl?^xg1|)8VA_t9Y
zja3D+0Rb2)Zx`=zdp*h)qT6vv;4-LGev=pbxbMjEKjf_MbkEJIU7E%}=9^?MQY4tm
zo{&9r8X+5cApaT)bENcP?EeA9=kh?VY#=yxt5&HD#ece>>{g&CE-`{DCnmGKOX2|T
z!1h@j8GE4-3APYkny=gK80<K{v)mzpJSo;c$Cq)wL4QR5k^@l->!*Hof7-k}9wG3p
zu4yKqVKG=*U+TyC5K&>nKq#S`y(YDQS*m?~Jb9b|J~UdHcGK?UiVFsgvjK&th^4w#
zkB+Y`>But-S|(D$Z2Nk@6)pPIWGFp3s;dhw@P5c|V{hE$S}b`8qJJMU&}&GjOsOQr
z^+02i<{VHh4DYw<Ny5}4HW;JNB-ra)VhhrONk)niwRsdchZSY|xMB>N24jeENwQ+f
zdBg-8Fmu_&VM+^5%<O6q6O2epNnR(8oNxze^J4P(ehZ(l&d=df=7Ca)p>W3;F#l{U
zdvQ5%P~UgJ>6oi@!zqkLm@GsiGpJ4pBQ}xbq|HfsPQ@0fus@Ol7LU{T@8-K!DEb$F
z5gPys3SVsY$^KC$GjcZdeVR8WHSJhxU-$l&Brm+8qVrFvs<g?jtLIx<gd4AMEjE)Q
zUEg~|8I@X$u2^b3dvj2p@i=Np^5qT}kCGn6rQ~$ep-5{&$x5AoB4d$bKl;u>9@L@!
zrwZLGpibok7k&VbdKy0SIU;SEs%WkK4x^T>+;8QmIz8*-k$NaV^P^CuM9ktCAR`ya
z^F=z?GPMHj2VJ6Mxm%_ExN<k&@mr9wLWKlLJ1+;fA;+Pb=VkljqPgh#B+|JG2$m!(
zDRTzV5;2M*E`vf~iq<+}OBVXJD5nmGM%G{PW{Zo3m*L<-Ld)`4*20v(CiXcaG@bwY
zZblSb%a?oD6>G4N9B)wmUaIIvX6$`QZ-Hkc6l^kUUVI7y{GQ!o%NS7vos$JCK@&OU
zVJy(-v9gpDuD{Id10~H8Fi}QIcP^JS5OOXoKvNX6*ilH_`hnJ|X{o(VHTJ>!kfixK
zHZQZBvPGP_HuN)^u8Q$zTp4RD{Z(7lP)4VTzZqXs-fQ+)*}~`y>-W@{9eaiZ{deA}
z%AZh?or4<1W%#^Ya<O_i_g>0t+XMdAvr@fl%!NpTcSmKV6yPG`Puk~Ug+Wg32uMsb
zar=JG^`+5yW42dkE(I|bBY5x#B}z{`=p+~G@>*#@u886dT7|tL_q38q&Bp595@a9U
zJ^;v_y__sj2aMQZfqmB5x;JfE$7gt#!+Fy_w%1IdKPaCAq&#+zp>N<phh*Rb+IRSs
zUM;CVi}8W0siY$lJ&fE%nZYvGL>*%5Sa!B!=IEYtSLWva&gdso^+qpzSHYWr+s!9!
zjoS_0;5nYRy!~dcTUE)d`rpMN8M2=3jK{&P@ruYeGsiNp5ZSG2mz^6&B5M8e^pm7C
zdB)_L8cH5ubM%;T%~@jG7yx7-gDMn9ZHNOI|0U#LD!8CLn-|mLR96cIOiTol>Ph)8
zWJ;x-CP_jQDNN;G@NJGBwp`&dk3}gvz<?lii_!eni<Q`vFQBk1EOg&AaP20rd}Lin
z@0i8aU`G&hW8sJ+upTKx;;;wn31Ke7;v_~x5Tb>j!Luc@EC>l$Jq_YZjX2x%NVaGv
zW6unpRRzH>nP0Kj_hz`9bWhmiX1J8>sKf*gBdB8Aw*B@g{L^tsSy);QVF&WvH@h_t
z`v?mwR(IXP^ED=z{2xXy1p*?qLMsUjgNP5gPF6v$gz5G4@pzKCj!|>7<C0mUSj)5l
zEi%cd!-n#%k7;50sH3M&V_y|MSef#Z(Ytzs+9U_e=baZelUj<UB-IJSMWj<ILd39?
z&!sym%c=5kvXE*gn5P<caskBTMK%0Pg(LSC9Ozl(Q^wCzvf>_jtXXchZA)AkEn7H}
zFaV1O{4V<o+3S0;2-DIao-G`zQ%aZ?sdol)2XeE%kb@iDjrhCP!s^a=r(GZsJRQMg
z=IngO9qCNmBlg`ih|YOx^?(sNzb3x`oSyh8iHHk$h^b>5@&kmCiq7i2{_EfaMxVmF
z-MB@y&4T-zb-rnrkHYyvk{kXP$>T`vJK>K7ubpN;0M5&6L~g%Z-*f^-<C*8jj^~H#
zv=gWRXwvK1&qK^&vdUe^Z@SN)c(@N%vaF#F6yJ;hh3IdLr*<!Z_6rh{mui3`Q}}&8
zp}d-g+iyYK-_fpwc6(6*rM#UGq*Ok?70+nOyYzq#GyYaLedQ?0CNGNvIiAOD8YLSa
z2ZDS!JlaYgg}i$TIe1eqbE$Y9l0XgMNt8C@5jg@QzfX?bIx3u@a0pJfG5_WsFRX>r
zZ{A+AbASR#0=crK&clt9K9wD9Gpy)}t?fZxB`;}wL{9m|9PHXRCrm_I7gT>Dyw%Xu
zD*L**D9>c`QEaE03(ajyT%tcmFPsm4dGAaS6#K9j%_K*lGheRIqiA$L<k-;=674#D
zHdqGW5!9BE!luspBfQ>*uSvh7kq4L`;%Ti6RU#hSK62_Aq|p)!X7PIrYxdLXt~T{f
z__I${>Abf4G(Oy2vwfaFH1K9@YTNhyJ7P1qV8v2xmg7o_sI0-ltj$F|A}s`P{KYv)
z8C(XaaV#t>Ds3<FYlkn8*Q;XVliux0Myc{K8%iLeGP>s|#Vuz_CK|N6pK`~0bM~?h
zS<3*GpnGtBh*_$Pk8}2(OBhT*JNBr)(6%j?t%1)to6}&-@`knDx{zWr?n3}aGSaji
zg+=sybZLK~TepHBzKR$AVtu-HC6L!O<a%wDXnd})<TZ85JIpQqExmxU!Hi2sg?~u|
zee{bh@-G%@EixEb6&8hoW!zI*Ho5qYF-taF1xpvDY1?U%%Pt$f7&e)<tqr}7t8)r)
zXo!D~r2h3%3j{rtgnbnv|JEjK73}kUzyIHV@~d#oz7&f&h=hV((0lIajWXY5YKLmv
z1u*@}?axRqf~+M1IO%O-hWG3|YR{P%=ndA<60;RitAm&GZ)?Ay!<C9p@BUdIjoJ0F
z4LhIXVgB31cEv^$7$_6cP8er*DJ*EnaGKFt`lpn65LTCn&<QeA>j!7%ov<6VjlSUq
z<#tggy102<Sett$!37Ro4ncR)rlIPp&z9QPwCQQm?9>}R7lKvOoYb-6j9;YFt=Hy7
zTN0=WQP-2do#*VBGKfIH@2x|SkYX3Etg<BT9MXZ<(AniuIuUA(_yq^rRO-EcgPqj2
z&rB_CE5fwj-?1bioF*Qr!4Ij0+z74@*@JuiPq*1n(6|n?WUYVdwZ~j`atu?o9Ue|>
zJRL!I%2B9`8Ng6qQ&GiG%oB0^5sa7m4L;BY0+ScsW+PsP7RR3s*NHgBc-dyyZS!#q
zZLF1Mx<b8~%r${@WEfS_2g5j|TFf7-9sGRye&L7E!suH+ZlMe^FcreR#5DP-N-2$v
z35(v`tv?utON;`Hm>|9e)%y9Y+v%J{9)Rw9LR<<iAU58)O=R8nVcSnO%T3;KyNF#2
z7^fNqeH8=lLM`(5{}mr+RX~(MiOe;fF^=+8m)?ZwCtR$Y@suMc1M*VHxws0)nAkb;
zBLe4oPcXF18}VPu0nYzX4luL*)Cm9Q^Z#Se^V1TE-S~07p01g3@W#wg^<JL>;xmf%
z?=)cv1#0(#Vz=twau0OpfV#WBkI%<1Tu9Ke1ZA2jXd3WT_n}R&T=lj#zW+MfeLXr%
zoQp=;irYVW6%ld&xw9g*)tcudHbi5a*3Tf=k{CtLf%I-jlG@$B+1$6hx<>VgBK(Lw
z`I@qb*w4A~7;v&2CTsJx-ZQu{<mLL&f$116Amv9Vsj_<A!#Di3`kn*mZjWCLN|iUx
zAPg|67=B~%+AXrIG&CG_d$T%`n;AYGI6O>ztY2V<f*x9`ivyrMUh3;NkWO)^ft7Sn
zNn&yc<rf)TY9$a9#ix&zdr-pycNM-pyxrGPmh4Z4Q=a`^^`J&jY%1I{x$Hn*Q(@2M
zMg~0U5c+RWefYjNn^5(_piMFnny9|_Obw|M2?W9wWf8=4N3KzUB=04hb^mNkcSo}q
zZMR^hpbN@sD}x%i{Cxc$hbu55HIEBt&{ZOW4Yf@-({r&Yk_Jc;`a+sy>;x(caiP(a
zZDsB}G$Ys~@Hpfht8sCa#8XMn4}lOcwV|~bLV*-nks6iY@#Vy9(NwGs5$j9pP8(>+
zD3Mwc?i`8&0{+cH?r&(0DwYK^NmvyhbgkLgcA;0=CrESoc{&VFZO%DfV$oHWu|A9k
zn-%3MK)*Ls667idhx<lGaB+&iB^>nEq00BYD<sU+(QU!#YkqrbLJ7u;6fMIj0i(9&
zy9fJq({T$XB1h0z@mB}Pg;@~pe>B25br5p)^tcpGuumBJb2XB_Hb6VUB-o@W$5{_$
zDK$r@DMzk_aDgwfPf$rT8OjOO!8!esh;B$Vh=Deb52NqT5}p#95TX%I4xvd4<b`7r
zGfJ4wWGtDpgFQkeU1uyITn*v^AN)@=`%iQ&o)b)6vU4CPYlag_yMCudgb{&DASCUa
z+`Ek|@dbufs}g+kX%jNB@bWNPk<KMu=P$BRsi+xk$Es~9yniTBQ;meLFMIIkG^@+~
zSQcw}kY$C@ySFlEc_sELN<N-&;Y-zOi9=d~@?g~(P`s{dSI`+8K)hZ<*=9=kK9ZMH
zjeQYV=9Zr+MPRHC$;XNs-L^E9%nKHek+TlKxD3}7-<@7vugkKXQL#6pZLhE6vK=+5
zIw1CeF){|^R$X6n-Yu_AlP(>SJ7o?hl$BHw2{VgtoV(P~Rwo?La+k<>Y+Jwlx_qt6
zdA7R=9tsIjsk@IzBL{s2utj(9mJ1f!#}8ezW*8s6DN&Ue_DUH2do;-ensahOg7bce
zH}xXerJPi#9}B*|+3!H9Li22#uJDzdYyI?<m?8Ue9e~+v9J27Z=1h2qHYcM&MLbud
z!5KU^qrq8x*8;I~l6y7T19$F=p^O57vmqMh30s{b4U(oOWc%uOn^|--iUR8%Fxz&;
z5Z8UrK@0F@t5$Su{=GMJ^;ogvyiNL)*muTl;@!fEqBy`ZbUnjrxZE)>`tp!R32V$;
z)l)Y38)|1*tTmnF_U0ivrn95i;@p%+V|^|z9>)rt2*UDNeombkftGGT=)Ri?!%{2>
z^O_dnY^A46Z2=p~V{V4xEFafPoeT(y9q$w*a5TM#eqro%mOE-xi36h6{xYa5K`h(c
z&F}?AW#;Vfa)`&&Po<4l-{LnQnhB9ql*wP3Eu#Uqx|wpLuPv6wg#i0?Q?aJq)32#9
z<Jm5)Gh*Fxh2%MtUry{QxZ6|5@J2)yB7EE#kn{BH5SOGhUBzrLunLu099jifGz1y-
z5VyGCkxmJuJngK;ptoX8ehF%W*{!U`fluAjI5?|L>Yhn=)#LDNsJdentob-KHd8HM
zF{hlogA-@N=`j=x21b3v%TA8*>p|DKSY};V==t=;wo<MtlookmLulofgaW6Rj*VwT
zn8^pwdZ{|ks^y*-;(CA$sHXiglRH7jV@E13y#gKT7@9++7(j~jU5%X&woezvIq%qD
zX4RTU<txRNZL-wqciLhHZ&HMePNcku2W6In*=R+w`BR`OFJ?}L8&DMKn?`@;jCt45
zH^GcKK1FPPJWLe6zAjcYbv5M7$)zo8IUP@o7}_vRuHvKvR&@ES_H60o5udE|jquq&
zMQJi7DY-v>h+*W49T13gkaZCG-09C8$G6U6G`g(E^?ws#>9Xu*?&mu8TIy$vi#;)?
ziCvg4S3FIOfJRZ;sh(2Ilr}y6<z<mjH2DF#$%8&VmofsLCO)Z-j`byT<ri57f$3=%
z8aIxN_smQ+BXgCzq1mBId89Aa4{fwr>*^Fb7S7q*KMr^rFpoogG`o!FM?UW{{R-Cn
z;rB26{Rwm_(YTslH;L9?F`?Ur_A(3Kg?aj(Om(l~_o?oT|N1bYk}!?v%pZ`ARt>@s
z(l2}ZN@oIAifGv_#Ylx2$F1Y(dKbJg5CouPqooK(3O=P#nl;zoPMfsF$Iuk7Cbz1#
zBQ}vgiuh3Dr8TW(gt7FeoxKw5GcJ#(`-%frEq8I{<@ze2#_;)}j6yX<5qRWLpjy2v
z5bkkmtGIGjrO%%#DrR*v9dKhu45dhv+o!oJT{if6{{*`(l%4M@UO+W+h}V7pLC0YO
z^&Q6G?S{r@y4jp(j5WfCA<AA@$PU73!ZA`eoC$IqSsm-ggrUQjaP!e_4KaVG)49bR
z<S6}U<ip@0ggk;!-EgC<1;!?N#j91>j6CV5<Kvi<em}aiFz9xmkxL}4{I%?=g*cUu
z7Jn;f#VMNZuo7kV;9z#iZn>&C;#wKg9bH4&m0l78iqtu(Zb`Sz{%X~%OLOH&)VcB@
z3zOjOO~F=)b`UX5PSrfA-wvnkr11`O;IDc52YsD<Z)8crB^p3Gxb!jvi;|`4BtT1y
zxv$!75Gz0$Autv0Wg#_8@R2qfkc!x(3`m(MBgngx!-<_9BR2VRn7)(Gq{PRETO5hK
z1mrPqLhIMW?qd1iFG%FL{+>GqMDmU4mAJ0P44Qj;VmmEumNTQexqM%;UR)O84Tb9M
zQOR#tx38b#=L}?z@lL^eC9cKDVdlIcsQ@6vVPu}%xb07)51%(?Dy67-z24sizCYQz
zI^Evy85mOf^0~Tt7@?&dGm`DZ%1u<GahlBP3uM-Ibd?kf{NaRDsp)BsI{vBAQ5N6x
z>(`8k`3WTus-NDSHtS<7GM(IJ?_#+`OvciW*80x(pE=Kl9yGni$$WgK)46Yd)t0hi
z2?A$)NM(0Kl_kRm9sW)W4tdb-fqQrGFm5yAS#P_1pCIc7QdB&N$tRdcn-uYCrvBtA
zzj}4ww&a|+HW!A)_*4wgp{IY6K<9IvlG>v-2~x^)JhtbK%c}jjVt~DlaIYNT>U6TR
ziia-7AF|VZKpKs|*q&uf``4cAq^6QHbW6jMCvc&NHFE(Jc3WzO+32g0AXLL$J*xfy
z^$jZg1bqJAl1#4usX)NY$@IVE-RMX={U^yhSHGde11lhK_drwCtUD<0G%1iOvaBzW
zy^!`EF(vR5fJ5GiBMhzSwSHy9PY3I=^?qr)B@AhMojW<2pPXNxY**hKqO$8Tynkv1
zZp}N?pK+Yc7y_o0p6T$h@%5g;uJQ&P^lPDv4v%i@bxrQ``KKZieuKJqxc#dAsPG0H
z@;e`{dHSrP_l`gwPK7?f2ZyRijSovno^$K8LMTCO?91D2Z=Wm#xz!kC<OYb2lY9p%
z`v0p9K_fEx*7SzF9d?MkyIOo}T?@*ck+zW@^VnYM`W)6R7mil#jprz!7DkP;q78&8
zW^sS_D&PC?Mk^SZZ5ey_vn)uZ%TtSS(7E*6u38~>kjH2mKs8o&?Lwys%2#Z;zymiq
zm@6HVt`4ZTu2;t0#D7@4Jl??N&>oEYLtFkUJ(I&Yi<l2Rn9<D*VF~bNxs?lQk&=lS
zBpR4+Wm1G(=Xkfd>h*p8-cEK)gcU5*5p@t5#1V!t2$UEHUrI?T(oF{g!bG`~Ki2l5
zkj2~+T!eN@gmX9>LsJ!juqq7|>QH5JYj}ZoIjov87>bT&iTONR_cHS%s$W`1&F@^w
zn&D8j%^r@>d6HuB2l47^wspOmb}YPy8s!6@D?UXIKk;Q)ihMqTgyeky6~$npu3xKr
zdo!{1({>R5jo#@BS!Kfi;u2p7NfQfa@@J8zs9h(a!~#u*vn|P$7_c>7z+aSl5LIne
zE}<)F2ro}qT=iFlN|+em*YaJR%{8ICDJqCm3mD)>wdoh7sT4)g_<gvIQN`$PazPr`
zn9?GA!E+hKM{MGZH0{Co>at3&t=e!o&K$Tl!W)ExvIjJj$x3#QBrcllA}je@OER}W
zYds-J=13)uR3BR>9);@^M14tEFPy2Sp~RsoNBs}2*uT~B(0D-IE+PR{$J1>uwh6sT
z%P9t&=<YaBQ;i|*-OU!R1!h4a63R2XOy8hW-f(i5sgikw-c`c2Jt0thMe0;zW}mQ_
z&`+Z6f)s~H`xKhYZYfw?7Nn7(4GoRL;HU^a8L?&zhL4ih3~i%elsWaJusm+f;nF$w
zvx&4;uqYUUVB-##WUqIK$Li)s`}Xo^f4!D%Rf(Y?pljVbuHm?a7`F~Ib`3N)ZZgl*
z|JwBKdbNV)(Fym8OKxdeiYg`~bfU2<l+7n2F};)e5NU>2qaA$n^L?E|{n3YWc0>(m
zh$oluY{<WomWFgs*F!BjHR;}r8lVh}oRknGX>Ug~(Gn9@4>yk`8LONOh0f)I#}v^o
z-b6#Hsw{1+kTjCPHmD#T>ySm~LPIjS?03LdMqnCO1X-x7pc?iJlZ<%;s>eMiGUY-K
z`*QsG@@~aFZ{!}Y#s8B+GDggJ$)=WI#5X`X?D<zS<`J<T7xs6TP7HW{r%ueH_k?5R
zuu)0@{Fq_s|M#pVkqJjyL>l8^PkvLbB@r3NgPr6GF12OQivbM;28&vJgM>ePJ%s}i
zf3p$KfywdE|6Vy_!Tf_PE$jb*6!rrN`3Dl!&oNq+VQD`xOck|)tf}L(gTp8)%0D;~
zanFJObMg;@(6mDhB&N~-05<a1AmY&?(Vb#D{0aGaWRUWMo%ZK7HDl0_eqw%Jlb$%G
zJ<nPL{LGE4b-h0;>5o$HW2P7c{B=6I1dUg5kuD8B|M;``>yy@b$yG<9FMm@`h6{6(
zQYDyNQ56W!?J2I&Zz(0A>sHQD+U-520`CBCy~FRjX0CsrleaUY17lq5(aLJbDvI2I
zUy)C%F4kgxT4FieDJ1o`c6IHaB9oU~A$`O>m42lru4R`4-6PSQ7CL7xd{;gw47?QE
zl_(T{u@$&q+1u{=qwbKB**}#C3X0-9N1aplwyn8MSEijZz5Ya@d+1y_wZ6W&+qv*=
zVqD|S*iGN9Jb}}!c2U3##NK~GVCHn82E#STu5V(p)-4=nME4T<pEa|S<DI8l$im#}
zHm&|hDnl3|7q%<NDoh1QMjk2iYNgochqF43jGPoC&u$7-ZO$gQ#sNbY&)A8Z<A6t<
za$3>2NC_pPGzL|BT$$J=zT`<CUg{=V@|K+6YPGVs6Sg#N92r{Mv2<k#i08fwXX)jq
ze{zGAAj+%)U`dnut-&-%#)kgV2(eRVF^kP1`)y1oFy(7hk0`-sTheOQ9Ms2?g^vA*
zpF5f`TRY$3-MLX3Atf+Zt6+BkvQ-C2e14lGuI(R_<m5h|=5L7TL1-bcw3_iIb0=Ee
z1?3U!-52K5%{x6Gn@}cz0}-pqDT1HnWE6~{>jDa424FPHf9uF8i=K`MbUbKz@Mo2b
z>FvQLs-|jF*Bw?mPRV}!x@DP8`+cmkAWjLR7yg^($8=uRE6tgbchZcV@%Ck=kj*pA
zbRFnX*L{$YCgJw#0Y^O*ksXE=phPlUkAHQX#oH**nhtjJx<TNF#tI#)Ddy_c4;3+M
zV;n%Hs)2ygnHAE1JTWd>AzGJkaA7XCTAeM6z2Pmj?V$mEtR7|2T_L!UUJgmN%<AEO
ztK{x(st*P?(A#c3<&S2dwV7lpKwoVeF(;}n-!?l^DEhpvYibgi7rZ{lWL8<LIN)k-
zqsfTp2`Fel|FEdAx)?M9fJ|{R`*9*4;6QBGmoP9X;7A;vN-Mz=5_C;KmFQe(JL!D^
zxMrZ0(q)4_ZYefDq-}i}f{hSAX3;;_GG{0KGP(!nvuS#mB^r8|wOMVUH8@HG5TKvH
z4EKoUIj1gySB|hkX1+Ack<j<1xuiA*Q7UHuwPev_@AJMUVLVnPHh=w%3h<|7`LEWs
zGOOyib>?u|iJG7p`<QWOw^dG#5m=^I*l48-MkP(-9;8wUjKyfMP~}%m64;*XXi%sT
z2S8pYnrwWVRFbAZrw_}Z_6%tQpnYb12*yOi^gKF*=_9Prr7;fkHxY68KvPa9lN}Xx
zKES;aBAk7pE+^Wu%39{TJaKy_#@$Z=;|;4aT_)BSp<jzpt}Ved6Q`VK!D|zPqPZyc
zt!47lVt>5KrVpNX8g1ky<rva?$sCxzOTV4Nk8h}TwwAC$CAF`9L6v4zZDKGVHY6w+
z|BVasSW_T~x5YndHw|IXx~g&xPpu_SWxw@J%RyV?qpaDL>hRLEc(+H};s_c@xmp?e
zG)zz%Np<Nm>yQHhS^f&T*7=f5H_`9P6)9Ao78>@@K1I;X_mo$Mfsls5)gi-#*RSIV
z7m%5OxxK|hA5kaX$`$Uhi*&C~eCY?PLX_Y9xZ)mg;t2mJ4m-{sARspncE2bNT_yrF
z^zCBR+WEh9a(yuhSq%(Fi~5ZU+j0yvHPR4Lb@SG5wZ+a_4h%eRF#h>qJ9&sU*`j7V
zdT<)}=|_G!p7Y0yA#8e$@3r@=2yCGYKS#n5Xgwh?&)a&u11(Jo{FQB_O77bot!Q~n
z-(qR{NMXa#r(PUybGvHw`ID)avc!Ed&w(D=sb@s>JY!SbWc&2^BH{=xC!hEpGVqDM
ztbG$$5K-&QzeB+>GK|l%wEi+fWPaglY<eBl^rKh`-DGayO{2qa%5to9@z1sX^>KnL
zaAvA&IM<{37}<``%`O|u>s;=AfsQUP*-soGy*W=)Uce%NZm0bc_Im59PP|0LP2HVX
zJ?W29H4q5KJyW<yQa%M8y$@?_9K&Au>mex83=AvF@s`Z0p7(f0emmj(3jyz&!uY@B
z(p-%Hjlc1~^fIi{mbN2lMe2H2KQSi==~ut4t$>O%w)(3Bq4|PHxH``tplg7zdJvbB
zvzIw|Wjg+ZYuwtdX<Q4?ua*9|sR0%6_Gszg_wJ(d`8VPJ@9yU9TxbV0#t(cr69Qfh
z$~1|y<ICY?INr(P#;G^<=>JgIw|6<R_^Ed7gL-I)<Z>#}4L{%n)V=BFS|rqmhp>PI
zkaM#3Q4V80{^MmBKUC~-3*ZcWl^5vxdh|c0^X~G`ytS(|9WmJzxT%Qui{=$d(+zC}
zz781Q9)T6a&U29UsVJB_l@ni%<vZ3C&66f*Nts<UhKX0FPMT3$eyOr4`U~J6VV-LT
z_~`>$CbCC5I<gA6im`fG_4=I_<l}Y$S8$lcMH>m(4|njZk~J|TvU~OXup~BJ6+hy!
zUtQ?wi;p)_@4~Rb6yh)b<dH3Ii<f>*Pn`Og&=gc-QgLipt|K(xO`~j<jq}E{2KZQX
z#G14sOjK?=GX4$_-O<N)Hxz5PZ;1q5yr|KX)9K(mmW6NOQtX6)@iwJYyx?O8iorJ9
z>}<gK3r-{8kL+xM&SDaSBdxRH6SeF@;EqELtn^%lg%QlQWQGcYP|V`CqcbP8m&{vf
zhJf4Jp+`znCO@+Zeus&0EbUE8)MYj;T?CN?fijfzbddlRC;KwZ#el4!Kw^VGJv7;@
zcr%@0LA4^d4Go#`*u7+!;cKe$!)=33O4!MtsUq35PfsEri-1Iij}^3#{Kex-6@4a?
zpvw;WSb~Dm7Mh6SNd^mJAJa^m1qlm?oF=|w|47cejw~kX|BJP6j<R&u_U-DjZFU)5
zw$)|ZwrzFUwr$&1mu+_0Hs9*K_qpSaanBg{yz$=pqw4!sl9^;?;+L!>6I}&Kq>PC}
z0;jONE^7`iPGx}|okwx$kx4rML+(J*uCSC#7JAclQ3`4UFaI!AI?e4N4Kt1h>GSxL
zPRE;YUZ^QVX_xCjbht+&e{qM)b*ksY9G^YR_re}FqLpG<Y+b1)KAA`T$NOfyP7fq$
z#oG9Ui@H*RuLT;mlU_(cNMka&GGoJdYiD@a1nnxGj8e5#HvK`-(S;NQ*YDFY+j~IN
znae<ho%L}ykEi44;43~VKtN3JeOc|7t3%pg9sCcL$uuHtlZ@a3#E)$sdjA+sUPIGE
z_ONVhV7>ubS54p^#tRG85<*S*;%pn~ipe6t2FIZlFcERWRTGb7lk*f=W*}kh=alWs
zT@8aYCAvxSVJ0;sNl=q-M25&6hz_5muL>y@6)#Dz`HZ7yy5Fb*KNHm$x9~J&z6L*#
zbFMjF$fUW{u-R)HwgzZE20&PG*Q1$-(5*->i09pOxwvHd5qhH@U5qMuU3iSffZUkI
z->)7yyo}v-1qKSK+Ot26w_x%YS?TGtNBavwj+^t2R{f$9AXj5&0tWFgd*J#)5ivE6
z#X5<76`kZcz-*IXmVE&ALxu<KfKW^xF2B5iarwf@``FpJROXD_H*S76qge@|5Y~H%
zZl+G0;r8d!G*EQuXeIkD5}I%5$Pd{RVa>J^z{*O=K)OM2BCVF_FH|us>{xmJ1=$&W
zH}z?aHg$}zme>Mr92oWoSK?KY!S1W#=|NdiWII~s9Mj*JDTiBAkE9q=#od%PzC3Bs
z%SB5jUwQEj=X*j9Y5YniJR#f)N?c3Clrv)7xLfJ@#NrPe_$k}x=!{${>`9#XDhir;
zkghx}c8H{io^V@sXf3qnfNq#FM6B_P$~o7^fec$ab~rg6q_L=x%UIANvtU&<3pgao
zRIa!7N*@e*YQCtyy==TrS#FQaU&JugQY1-9qemSf{Fx`pX(8^^S@x}JaXa0c>}>p!
z9E$8gs_YTZlG`OW^CdXnNqcxA(p4zYLiifmQb%T&VN0W{AD2-pk|D!oT!vv3Qa8#`
z(^<JX19QJDs6Q}{p<aZX6-O|Ey=^}2GQo9kTe}76_(V>$ZPsaG1j^VRPW_aR2p%uD
z5C*8xmunwoHIIP7`|X?}zK1eYq{Ck}Hs)9H3S}BJDPlKLT(u12U1^{pApvfgC(=`M
z0=)wYxCa|4MMYU|-_O>(B^dNr;Cgv#Bvth3Q8{MVy>@eCgaZ8$sOWB)1=5I68$^f^
zO{+dEFjYjvxWm%b0N|lY{ARowObGTYPqX&m3de~xt&y3@yHkqjN);mG8Eg1glD5;X
z>Yuu+Fgs5&u{j(V2h^+3B=o8)U`V?1a2Jg<{6^DP$w<UO)#$~PX7CyiVvmQ0b-ZM2
zF)P+xGjpE#6JAK5W~mI0)zQ2%Bv9hU>1V_*uZnp&w$3qBEUhOE(gvYM%`)a@i>p1^
zF8kd(XZ;jcxGhfh5A5;H12>2g3U$M{4ha~*2vkw}BRNKGDPuszI+j%wKZM33B{p_b
zaU%O<M~AZyvPNJXdM5<Fj+-z>cYE<3%0Qw7#S#T;Bn9l3{d~B%$lxx(F%mK~`L*K0
zeN*Q)9;upzd_ri7+Jk+-5v|OjNSRJ{g#GIR_g=(gFrL+Q<M@r0orv^uZgdTFdC%C@
zCs8JvcWc0WGeMm(;tiaHd+watd5?Kqxn$-|IrArD)%1AH$1NA`9QTKpd1|j%=MN&}
zB!2JnErL`vGP!Yc@oz3@*OIm}VqH95@^7^-Jwxhv^*IkV>M+37$Wd`zyc675{hZA|
zPTKL#F$0w7N?oZ%qBm7Vh&yAPK3H``%)#(Bo1*5wp>2D!nJA;4U2XVud#u*uq3dRC
zOS-xw-xG+jStL!;vmOk?M(RU25{*O&XPO{AN{b#~C2kfuPD|*D%xVrgl5kfQy$Q^o
z?uULUqMmm87%9wJ^94D0@7aKBQcwlQq_<@~q<$$aC~sjmI4YTi3#zXVgGRy-JrLD#
zo5)~H*gLJVf;zv#-t}E_Su5fzr+=cp*NUAGJD4$t7|F8YC?fA&!W|o^h=SBn53POS
z=|{oH(}G#Z(PffMcrVj4{|;DQKcM9l#w3I_Iy=l<-b=-OTD3!`LO3L^8_E(QG<9*r
z=d>Wuw2zZNTgxA?T~{c$npH&Dti*(=DB<$b*%tm4+=PqN2I^+RWcMAzts4pqImJR<
zU;Y}{#<MF*`MYUb@v@x%1N|goTxQRgo^jE-d<g&h4r~m;41)9BN15z^+Q+gFSP8U^
zM`DuZah)J>3OZuW8^YEcMX#kX=@CKpX#;25!ZIb;MrRYG4lbu%y@WDmx7}~i<?pEH
z;wVsl0g3tCI(>Tz)`h<?b+J?qb<U;a8dw7(&Zh#Ax860d_KQ5F5MT7c_s7XHAkEM@
zx2jqV3P$*tBdUf^Si2hql8?)?UcmAnOg3+hMLg!d+z}I1HqS~mDo;QTbVBHktPN*g
zWWU1|CXSZ#I)TkrDO(h}H<hdFNS6a^(4MM_IWf;*zfR$>|0c-LS!(lo@R+XKELvdc
z11&|5J3Pdt9AP?$kkbi&{Kl~cRY`VUM>Gk8t2?&Ycq*`3zDfhjUZ@Sr2zlMLN#d_W
z(U+d<Bu4NAYH-*oQA=x<Z$R?c1A$Cum6$nBrur-e0q^cpG)v{BK1lOKatYnaPaO18
z0?WQ9y~Oklq0v+?Xs~cJut^e#yQxmFwRx4~TdOhF03BktQB=)*=;<(TERlWjt|%1d
z57u$uwL_x3yF{Hjl|Byt5^g+VKC-T^XfcZvA%eQTeYbSr@7rkZ2TSm*a$=Vl?mcw1
z*2RGy+zPq#PO3yO$eO1h3+gp*&POl5l|++kEe=|mXg)2&E2*LBA?rd&T9NvZE@$~t
z=^B3fUNFb-81CAnEn|=(kU&;QCkB65g@xKM2x4p622ESgAha7!Bm2PVpXO*-D^wev
zXnG?djW}ALhIJFSE@(K}FIn%G7E~*MawPkR#CdEl+`n^zeShxk=L?{zFfVttOq_d8
zrd%40f|$tuJ!_}novpt%XIpw)Sm^K74wHGCOUr-zq>wyDN9FD3QyzT@gJ@hb`Koif
zUN`D#z~Y#njm_z24UFZI_YUnSvh*Z@TNimt_mqA<I_CPItXn_&U=)x0kY1<YlzCj3
z`=)(#P<2<Dm%J2hL-X8-4Ab){6gaNcI8xr%xed1?syqP~ch}f*itchox$kygb#WPH
z;6t{dBpjf7bv<Dq&BBRTx#DL8hC<RR<*qSo1H$nMUcJQI^6zFAfCYIvf<K`XxwvTm
ziIUF9Oh^A8Z3XGX8L_8ozOJg?p1C9doRyKlP>N01quko*NFO63b0$f2385W1ECym)
zTwx=k3`QKm%$Ygr#OB^9-jaU1d!M^Mox|Tm&G_!1uIl#kaw_)Ev_d%hPmFZCmzVRq
zxm3WkqHPBh5F<T#@Dh&qeykn|P6XnOWMHfw6mh6~*Qx)IFf#mhbQywwm*jecgh%c}
zGM`QR&%xO39d2KP0H5dQQ+Sz-rk+&R-xg08KAm1(FOPFA-+=ufiKM$dJ)q$EAd51%
zD1w=O&RzQOzQ5chBbrOlki}@OV^nh5iv>iXFby%H;*r>2YX~IS;S4!G`EPjxemGix
zzV-+>d!{rzPZ4m0s1vROLQ*6GZjk71#1O6TG*=V4YViREoKtm6Z24LTcZM}{e33+s
zi9#}_g;+?{yF@EL(D=1}pZy%N>N$#Ii5?47C?CNze)#J;BtO!GEXMCRp8iQDC?=m_
zIW*%#UonT@nMdQd9Sl3tFg>&0u-qgtJ8y3qQL4us0y_em6OKMSkz=k!iO##C?NV$_
z4Zu3ym}#h2O2?EQt&VqSh|Vz?a%c^Y<K{{QYBQ9+%{TjM=5`kpsC53#SH$)k@C?0U
z8OAbZXIzJ43Tu^Yshz<Vy=Or1^?oDJDEZ76%Nt`3dlMSoSdS8((aDlH-zprN1tmBG
zD7oDxOwmo@b(fBZ^9xW))X>y%nccKzW~7}NAE#joJ(YPfj26)~<euYPivw`?Svf*W
z(e}vfq>z^gv3VL57U@VxEui`9(JIoBBm6T`?_n5B{jh3KPAI9kw>!5tS6VuTr5{y{
zr`Aeu9OpklUr;Bn?e^I<L;9;~f%Wb++|tg4{mvl{O?gK*T(C?I_6!Q7Vkw2L)FE#j
zX@3*&DIG`+KyNg-48E0&g**SiieW+CN#eN=`trqOo0RU^$RNoOY@Us(giS9-%qT8S
zXokRLJ>o`fqCXR|T+05dnWb&UimeIhG>-h<T{$!B-Fhdn?}+i1BU=oA;mU&H-U{k4
zNMO8u7j0UA$#mMdCcuS<df0;6)5jS_6StU(qTb;Ie(h)4-Hb<KzN(7M(Q$`Rdh2(U
zAEdfo`Gs6JD+&ylGY?(C5^euFy!?0#EGmfyoG)4u@f4u1<d&wF%)c9;_<Wk<HH&Mk
zKNXGaO}t!szD+hUJ9l{-Ya^!2C~Zp}P`Wzxn8OeUzO(qK>te0MG_7~A8Di+4XlZ5w
zg?(5@IZcVNUr+GX+Co^h2d~mOiav#A*A+e*3{1)Z1G^;{Qj=JOqK-h3tx(F>n?3#>
z+;&pvI1Pp|!&h88h&Wrl(>KRB7m&jfe27_Otu+K9Zb3u|S9y^x3^?w|jzF8z(`wHA
z8*~A1(Y-u#E-nfeM3AS|*?8pF1du6~-CTL`_|6e)A^2m+nFSW*Jo$sT;iF!Lg+me)
zuT$d<%5e;q5QHBTb0E^j>@i7EVoYdi7g#H%n(iwWcKZXo_SH<JYis(c&<S``_N?Ge
z9bhRAy^`7%JRz#mwd9cE_DaZMq#Yqhbl41coPaIa<xu8ItcgO*!%R*&XF<Fw(n0H9
zeei;yUrICMQ|N)UrS%Sw=dMRKR3TE}Wl&dP`~`zh4a@6Oqt&c68=(mmbuFci4X{K|
zG4J$tG_^cB3ez)o7L7h7h^N;XNzOAwhdlQiqgMu~VvS9Jz!UVPf7Sl<zhm|8^mdTp
zCth?#qoh|Pn$gTpjv17D!AfzuF3C_dL@O8Bjwd?*c8DcoX!et`xF}o)8EmGflNB{R
zFX@L7gvYqg&SE>K6t!rOtblY4Vs)ZPLpFmTP`r2)qoRSbMPQEN;1CqLt8P_Lpa1XG
z_HfBLpt8MsPJ~rqKx~G!G=%}&t09=SfP9=%n+F9=RmUrLO|1|&bBW$BMqU`)v1k&c
ziCaV@W14nhZjk4c3ngU0RxPU43Fily&`P2hOWHe`OXCKV%n^9ZLY=sW67XtAxmax?
z%^@zp14|o;Ya~>OjV=X4o+K_Kgkt-2rHw==Ws7I4^txC-ja}F2$H!QTQ0xV+bPwZl
zM!@M8&4QLXMcJpAxct!Y`n|5;6UYWh6g0i^xG17h)0*%T$VdR8q>CcFmY*S1LpQZA
z+3?lwhE~w;5H6t`9>zcpC0#kInhamrbw9q7worpPTgc%)2a1#^$j9hG<I}}!)1MY2
z;G2@xQ$A3gp;_P=%m9r7&B>zd=`ktF(_7t>&B}r<DBzQP+gA663T&<&^ql|J95)P}
z(Ir0)4>reX4R=w?*V}1-h==-{awLF>XugmfQUtSbc*f-rqx#38W|iDTiUo$?6LxqA
zloL|m;znkxs%ei7A{xU=<1$aqZD(K!KM^C@9kscsQXj!bD4UaudIk(#RI{+C`*VWS
z-p&HEmpylQJ`x8Sgz&FsVd=v-W#8<)=ZuL^)S}Hz8GHSK2wz2Zs!4*fi}sc*z->Hk
zEZ*YBU&ph(4~n~g_v72v++5scv)yT^1!sRay;z90vwq1RiToKmR0EcrJ&ndDBTEHR
z>B+Fl`E5NoS)jN-Qx!Zs?d%@HrIP+fz)$6Gw*3jEiP|Uyw)3%5srE?9IIa$Oaqus3
z6kI1K_t+cgdoNqaXx9q!V^{i)!D;xI>o_-tTvbLLW&$*GD6W(<oWCd5xuhRqy*V(a
zzTk~b$*$KyXfPsZJvELFO6E@wDrTyT!QUs1OCI@)YstHA%yx1RMbXWxpq`|5;eIpy
zaa0~=JXd4^CVa8B(}1M}-d~TDU~_#==f(PDEEt^hg44Bv!h30wv1yngBMCjF3}JnU
zfwm%MnKG%Ay!wOnc3u{H^YyF^ld8F?>MlG6xAC}?i6eY(L%=>I6<a``Enb6XyQY>U
z%$L?las@|Cw$dUZ#+(c4IladO@0c>|Sg~W0l6mHIo+)_x-g?(91rGZTbc*;$WNwMk
zVo%I$q@m1da{IeA?Ubnh>@jBg)DFMuTc{#xRQdshDA03pwL>+%iM#TV)>P=j*Aaz-
z->3pCPGolQ$a?-QP?52W**`Q9QXcuy%`~By{V&2yMcl)b(;A5wl$>PtG)!;LpPfnr
zqWoQuUg-E)AR;l%+2USs1gP+$KQsi^lqkCHm{^s<oZ=RedYZG|%E+#)0Kq7c_5FJY
z0~_H<ZAA|)D?%}};x3G*GUZR*Im<h<qZ;PqTJt@@NFAFOT00kjTa$r;AsYw31X3)E
z#Fcg;T#VrXE83B};ndHQldopt%x&ru)6Z!FuIFA{QfU9E`BdZI3Tj&LYuap!Uy68t
zDpb&ZK`)$YL`!_dx-UG9+FXN?EoUtuGjB7WGfRI>H-H9NI_wfw!1A?6*1ph-CR$?=
zXj9PO)8Cag4;&*RXucjvg9t)VVxopOh#U&KUA6<R?e<?Q@(~t#RULI_*^(NtdLe)=
zLw0tk|A}V>d$R}&Y6**&1b@T0o=PL$H@Ube=w+>b)oR^FjEw^|-c(XV;V%t4kdZcW
z?`fjSrs-d#r+YM+7lkd-sj}q0ou2Nr+iP=P+v@nvT*lkB)O3p8=3F^Yr?2fY6SZXD
z_^hrkigB{F8^wTc**;X}r(e_`;g2{w<BSX(f-XsEffkvH_?hgDOyO%+%A31o__Jb7
z-Tk5{$)!t&t}%~Z&b;kToM_s^iXj`150~2$>!EJ3Qb2^aKSG9MYwQr4$ouxt1L`Mx
z4f1+a*5yXn@Q8k~+XKr)=HlG6TiFkAypx3$*>w5WUs*YZ{gv#Z2Idqs>$xsKC%4y~
zzNWGxh6kE$6YcvuuaMz9#b<#MK`-kp(>Fyq?k&xxKig+Cm?DD4d2hBHOltABv#)a5
z*Cwu}SPsq1Q-g08vW+&!lTo0Dg)XCzlZD*SOS^&<_et8H<%(FIBgN3bvOj*m;cZ)A
zd0=ep=xEbkA47Fq)_yO(;+(9aBdJ-BIIT+dIHV>PHRg|q*-1UC<V@aJR?y6!AA&oh
zz^|D)Gi&8&HCHovWLkSBjWDTeT9Vow+0ZXg8vp6uBOz}Ha%)Qc(rfMqikODO&&Ay`
z#9I8wniuor2u*XJ8)B=@OUTM%cxXMcbi$u(S}EwJJ$Vvld6Qn`LZ8*BeTXr)Jn<@%
z4^5h4rgyWNC4ATnwWb1VPEg;&-5-$3r!mlMKF^#<F5B{!i-vxpjbr<BZNn}vA<OMl
z7^1Uqlp6uRm&xhf>x&#m6Mm{mx_=mM8p0^#qI6DZXdPaN0(efVHxu9tCh$BtFMbGk
zf}VIk6P|3_kmAhg3l3<A1JPqq<oH2!(AucoR%X18WL`T*Cy98sCpV<o77_k#91TZK
za^O)xU>n`+VgG76)P8|*{l=kAsj|IJRiO89G#_5h=3qL^(MB~O?9x2~Ue#u6<||uK
ztC;;y0{oxYwIW~gOpUd@JDN2d1L%&+B<1Ij9#ZV`;HLFDzqspXI1favsNBh7M`}mP
zwdm_wm48#q&mV55x%^bccN%KaX<qdWd?fvq4qLyeE}^JB!#KQ|ITBT(W6+lh|0GV8
znq;q;>)|!CQLMNZ<IWdtnkLDyk0|$><uh?S%12%hY$f!``wngUeA(u-`&l((7Ss>A
zx^BUTm#&%=0zBZjC((OkrSJFE^TYkIPw@P*`@fs|G5pUTl^B>98UCZGU!%sD{V5C5
z#w*nmFgibQq)OtLQe9PI*_rWByP`5<yyXE+oRqf@7QrR~MO++u12KXr(~H~m&*!`0
zf#PhPO<o=EHnEI9AuUsQe6oLP+rB_&U^S+}e3?Tkn!=x}uFim#n&RaC)tYdJ@nU_f
z(CHPgTM9l+`?J{wtoN~d#-Z?$?h$&|ev6NPhUjR5q$Bo%`C;eI^rUn`GI@1-C%buz
z$H(or9V8$wO!rIl$I|fOST^mu9|DQPT(?ihlW{^%0cw0q8c@k`^`~@q_gg-hy2RL5
zv2cq@MI!1S#v28Ic$VF%fkITfVM#&6#07&ykp5d=U*GwRQVT|Q_`02mODR-we+K0;
zxsl3;*WE_EaZieVRk&909?w9iwMp|B48Ze5BN5EY{Cv9~o^O1+J>=@#PQt8ueSUrv
zrf6lwAQi3PZoYJ3-|;rz|GfGoHz$De?&#nthAIGr+@ElCaGYR{p*2@LoixKpM}ucF
zWojIO`2@YXIh^h!sVBqdT<4i0ccW^|&(%|c5=X&UBEaR#N`cYbGJxIY)#1k4CkLyu
z&pIoIF*svm3fJ5I)0h}G_=!x`N}aOoSL^|1o-aJ`wqUUOoj7sz-Aot06Rv+`PRg-|
zyz37GiK!NEE8`>riyE_1wuyuQeJIT)&@ji6d<%gHgQ520@1H0jiE5gV7zLQQ#V17D
z;%AZS8TT(OjL^CzTD+f#CP}{naQ1dOG2M1lzyElm3Brror4MPC>l4RtiNj>QgM-nI
zZvVO@Nx?Yog{_$zKqy|WFMAtQKgo6OLR?BrV$F>aH%v<GYDT0S%a|9tq;wd!C`%eF
zrIu-n%~y0+-&ypnZOSYwMzwB^X)gOZRhis4dkn`VF`+j7cu{jO4r-rZZ5%mm_`OJr
zrTdblG6%ex{eoIF%(+gdAT1H037E^zHDzrvV^AZn#-otg6*M`51I<x5I)bWw(j7i~
z%fte%V6lSh*gwahK_B&grs4Zjk-eXM`}q3lKxZ^sFNu|-%%^5NYw2ARn?;?*R5JK=
zQ>`JrLz!h`AHtWqhN}DOBq$~fwo$NjHK8vu<y?k5Mmu`!EySfUM5cbGVyyk<oiuP7
zcu#ESmNJ>+r`W0KURsSzbV=;9hP*W?tB>S<4d?FBJ4y6KX}iAu@;?!mryy%^*aHkE
zBP2Yc@f-<<^Npie65rpF_n!~F=E!sMIgm!_^Vj)~<9<x2V9@zd(5p)SEL<8DoHEI>
zOb=>7KiS))yXfI0l;MkCY2kc$FcF?k^Tk6WAC!mWcjGpDw}dY&MRb(ubQojf!LSmf
zZ?p`h=u5T+bHg5J2)<O3Qx{7ob}_*%Y~B+8J^iXIvSgxatB9*^{JgRQW~v({{HTZ*
z?5y^TyAO<(F|aA`DlkD%tuBwi=aFX|18Y;TjqE5|T*oFq{mUZ&hsHQIr3V<t%-PPS
z#?MpQ`qCI<wHC^r2IlAb`r&;(4}%IK>!j))n_nboZq$eD7j`~+G(Yz!JG9!V-R-1Y
zG7hi+-1b=RS;-oHQ{zex`wEmU&-|R?i<yXd0&UiEwy&U!Y;avXQwn_a$#zx(qrLOz
z(!rs#`A!_bIwYM*ZOPgLnCA858_`r<OFdnka9n7RG9J~$z43G&a7*9Wt#60!aW<FS
zJc+>tk{c>ex09Ou-&;-<DiLuR;PO~RY`9F@w(We1Xi^CaajLFT1BfPn2sEshR_90x
z0IAhzbdEL9zxO<=c^+pI`-OcUNNIng#Z3b861Z=5MI*~J?)#oGL=PMTw~G|WkPSPh
z_R?kB0geJkQH4CB<Fh;GbF@!8zp{JMEx3g5hQ;%4^ODSa-Y&QtC+FT5XVydA=ZQ2P
zD>PvU4M~Da6l0e$AU)ZlyJ^NU5+|N{a%`Px>bC_aPADywqsW>j9uTg^->^Td8JD2c
znOnu7#kdW5K&(C<U`~oZ9(lOD`?#At3SlKim?A*~1X+ydUMh=Z+&Qo)yC-S@#He~W
znCXc~j2dpwM73%<&T;zOAkc(LooSGE(+5oInlivseQPL}&ZxcA?ouOWL?*#LXxi#X
z#&7P&Qw4}@s50C!da+@DHS7zB{}8-}oq}>b&H~CmgPc>0ps@c!&L7)|U-0G>Om}@`
zd_4NYo(`(7mZYOw*qFApk#c)5Ns2x(>|q8P6|a(<2gEL{KZ;@g7#o!y9I;Ms=;UxX
zTNfLs!fGgc{}qKDDMQ*zpiGxFb3Va%bk(6$MpdHS#?|`-S(*OZ9+BN4U1(;;Op>Ol
zr~QfGPGSN<gr2z30L8kenwTyP%i%7ZLYnpQ9uez^L>sVDnqvot6BRJB8|O5Vwaz9x
zZ6v4sVw~8A&=)~dTxWQ((PT%lar}B^?BBRL6lnRMWd?=$ik$Xg(K3%s)5~Osb2lwS
zCL}38-i7Ry!^}b_gt|1CbX;WetR;*aZTFv}UVr$g|8V&c%|n+)i^ToXr$;e7h5<j0
z!!G)3HhjO$tjS!VPE5}Dtnu1h?)df2R$bfWj^JHL^ij5qU0mP9nWU~p>VbTKWc)kG
zlElvh<GghRn&IUVCL732LkPbCk37nSeZo2*?cNmEn+?lAq>O<tWW0nKl7pwJzaEIL
z4A7my#N&ncpvokBZqvS4`Q<}8l1ifSS(KC-EqjD!?Ae8Dbb$<50{dI-@hGFxqm@I*
z*qL<}C6MBbq&jPk-$xI!P5@ETLAUAT!l%2vw)DiSD8{^5?PtWFB}7!`r*U5bc~dih
z>u<A;ydjY|groV1c^f<Q)2SqE{`fM+aZgw6Uq!bn5fbFzDJpSV^n<seJ$5mz@tGrl
zemV;5nFdK!r}N~<gLcfl#?Z%I<;1Bj4#NIceZxz3;X(j#8$5N`Q~7P+Y}-DxLv<j1
z-)CL*{(*SgNS!03Wz&%4IrmDB=**8!W8AeYzYCVE*uFkeX8HbgGQeIs^=hYtysZZ(
zoszI0Hi-^oXi7*U9sVB3&y-_@NTAVI{eeDW<i)Z6KsmfHqxX}lpCSs+9*uej?A8-D
z_4sFt(!1P=L1ifvH`NkG)c8&oTBGXAC+$duWv-qSNtm8(5tS7=^YL2;E;0Fo0tb*L
zq|HPm;GcN%fyV6zyQkml6g!N{Tn+Yi)yuZO73ndA#ZBxyN+i8ioelZ2dy3nagRa$R
zHJ<L<T9zq%5bS6`9xL@pyfKXdt|I6%I_`fTyc5&C{4|;KYX5g=j`80t<TLz-3;B(j
z64ASW?e>q=SG9-J>UCvL0$5;YSr@tvV|nmEaT#c-{BY334r1QkVJKp;OhRK+``b+<
zwNJHp-@S$GgfqIb2acm<mZplcd19oE>%_6xyu3NZya}GFQ9KpO<74e-$z`&;Ge2G^
zK}<h1p9KX7ZicUH#<ou0{ATT;M^xmqbYR4`KKyxY-ZUrA{WPty-DTf^yzL=nraOm}
z{dm(f%o!iY9*LV{mu|OsKktb1p!Drm|HS)Scs-o!&Q=d(3WEg6@P0oZ;U!2wR%eS%
zuNat%_Q~ktes4WzHh@G+P5A8Z_V(mucJU4+b@G#<Y)~#^C`GX};V?TVy^A)Er(_A&
zuSZH3t4%$Ur+AtRPTbg(z3nkt!}iSdfH~-IEr#vy;xslQGc@>-K&LD$3zNU>7Dk(}
z!eWhtBU^Sp1D<%f5bXOoQN?R{W=JaFu(y1Q*IC&)aoqpNMD!FxdXc(}zn6WypK5>0
zjenpWL^CFObbfXyW$GRYyswrM*e#JzMbH<NDma`P<{uWpJ*(&x?L7p}*jte=ZhbMu
zDw}jzloG;=hB~Rfu@0LSd5e+mBOvyPpWOIygGu*rE2l;k2}P_SB*0lV9Y5t(;NSdC
zd&{x5^|~1yvN<(7pn8>0kiJ?z#s#(|Vd^%)jp_}P-+b%zl(d43SlQ!5e~_1y+iCNf
zMBh2FPS`rg;igE_IO@vr@Pt^KoRFgK_(nSTO!6J^qzc%dadmpzbO+|do+k+7%9vjB
z=>sb=_Ttr<f!c^)AQNcR4y7o|f9Fl$C(l~xik+VP>9UPFc!!_~I`|w(01nPA>b{N{
z2Z`YRHH;wXIwV5_i2u{Co7PM%8Ib+C7ja>ev?Ivqj)okrYlD}T)TTX^=0!%SK8^Gn
z%a_oWHQ}^MUI_}1x_EnG16sMO8bM0Q(>Ts}SLz0uuQkZ9`F2JiexdVmNV|I>Gb7Kd
zPz@QYvS#Ja8k|K`->#jq-|7lbA<gxX=DtjO{5Fl?ZweTK2AgDt+xAA=JtOH^-!9~X
zeGlTx^f~UcbpOU(utxvu1EXaDNmq?jYU&iTGhd(uGaEHbR#MM3)T?xU33i9XxRQ+`
zfxKQeqi@lKRz@*w0IBGPp-lZWVJ?nFN#9RAJwe7~rP=6=-enS~==Yv$C5V*KPl{4f
z0<GuW=WLz4kdv^VhMN+N5T2NMbOy@Oi-f#Bm_l7UE0!n?zJxq4&lub$y*ScM5GvZU
zqA7i{#Al<ktj*9aL{?!grYuk2)xZpn7(q!;^$ZxyVREJk^<R|b;+f-pb3=bArg49g
zsbR{v+k;sRX7|)3Qojz<J2^B`y$XX##5)x+XUiOe&`@Q~w?=x0sw@}4<>^Z=mE2b6
za{O|PStmLn_-#1IhugR}u{XvUO7n6h9TzI6kJg5)5X4k1B~W9eSvkC&mRfi;+`YFS
zdh;DJL_6OiI647Dj6YA-dC}Y$+0IH?wJ6zkL5#LElgurqx<-xet9+_va?`i#iS7$(
zPKPbCj?kJ(^%VVsG<(I7-F>#*cs`!n_zTHOrX{ubJ+F;bf@pr{ySZ>{wd!WcnBh<?
zGApV5<3pWc3dOV$_9d`jooJM04J<okMFEG!q4b<l`H9YEA^y>V8r`!xqg^la46};j
zZk>ENLl=s#k@2+;KUS|Olj?clrR~8<?Rc!(`_RZtHbP;lne;bd-9E-LVaiE+e?%vw
zvE?(D7hZ(d(8Deg(h1xBTIHj@xwZ8bk|{4~=<nR4>>%a(UP&t0He(3t-&*U8`pajS
zAg*Op2S`mVwxyIQZKT7X#;BiaarhhPsFgLzSmKAEHYnppivHQWfPTIVbf^^#HH@9f
zyGT~^Y&6vS(H<$_4yeZ#to<y7d*X8}ahEj09+1^}a-ZEaOD-0Wto2CHgNWuQAW{%K
z9x@Rly|F7_DlQ-{PssS1bO>6uG&ZtD@~eH&2)7S3lGN1-Nlg35f|gxG!@T*DyI}gT
zsoV3XJ*{ZfY&MD5OC9Y3;CIVmV8x#Duk(2>Z`Egp>ali%KUbv*45=BaFlOrll^rB$
zHIL-KtHjC;g4)1p5<1vO3mD0`^o?F>iT5K7NYGe~IL4tIQdkCKn+^Til1EGyuEWd1
z@i)uz)%DQzT@!Pt_i^ao4)aPc8_=oL{TU7|XRtZ;A-N8FQY9t6$_C$D;A{__O4}57
zkY>zQ-h+-P3$GJ#3NL*Tb!d(4I~_xcYW3c9b>334F~A@Ltkh9zXQEg=h*nYp6N_Qq
zRc_hqvLIjRe#3mZ7CL+BAY@1+*N&DIMWX9(a*L~F<3{X^)1<c1WVHz5GZmZ;RoALS
zZn|kdg(AT;hr%xuy5*o8r-J?Pl3Lv*JDoyvAKzmov`LjMVwglo*q~%`J{FKj2%ngh
zv}A@Na*M{SGa7<k(|b^+w>-n{LU|uUS@C^3%O^r1+xX+Go@7pmf~A(rEM?8sNX~2z
z7%7ADE0QMM6q8bwT6+q#e!5(5`<)`lg?%wK*dbYSSF3>oS(tvSzln<TA(HlbuI7sz
z{mahn?~ai6h`6vanogXQ^RhMVN{!0n9SJZjYj%Q}0x>0tk=z9x?U31uj*P@58w05$
zDje0aCGc<ooi+?3JPLfSK@khQQhdr-E6^V#b)|mfu(h?6KYCdlA7aaD1Xh&M9^`_o
zfwa}E^8B$k`cMnwu3|LAV<W98gjiYw%Y#D;R^s&JrVC_iAryprtgNri(9IJ(to6z!
z8qLrFnhwm47KppRLEb&~y6n08RdLtgwi0FKXj$G6uw-mtS+VTfLQe83I;fp%S-$*A
z-XGQ}u%cP3BL5zS`}5G-bEd6UKbVg6{!a!j#aflPHOdEZLybZ0P(XI7*kQn%qsHJS
zhOGqPs{%+yZy%b>%#2Y9<`h4i>7vKoS^L@|$-{S7LPDqPP}78AG2^<KMftkQ&>M?Q
zPr9PA?9x1%Q01=dSnY5DU4B~a>|&Jz&dr0@=AhBmMZ7ME-kk_2&9Jsfw8q?vt_Bh3
z(p{Y8I}VkYA~&h!WxEv_a*JsRxw!d&*rYP%nB`T@<P+jwcoNTALKD!=`N+C=H-+<*
zQElHDtk`&_>(SVm&AbdMTM&3WpO`!iJRTIGt6FtEADA+zwAjtj;V;DoLrYX!=r5#u
za5AqGCb_I{Ouo)#>+YUNyA=`R?OC`7f5ak;lm`~1*KWV=wuH6PBaapp+k!mJ4lgng
z?+6nn{estX5R;JK(tYhWE{e8dJN4!Gy7p3EIKPnovW$}-b@jzk*IHbBQ`g~6w9`HY
z(K7?@3#>gM4Yxa~i<jwk;Mg)zO{|+=CnU~sx#Yqy=>f~|QDw^_!u(*9m*oU%GS|RH
zP%8)@>%4MJWd_J7QB7rWqSg^=Jg1E72GI18P9W7BY~u%vV5#RS_t$h49V12b*UIR~
zj}0_j--*<9e4u5X_8AigAQ4>$cy;Iej%-;uhfE@PJJvaJ*rxCe9nqLm)liY|#7BDt
zVlMq1m)1oT@rr==fraSBryfG#S(O@*mh_)y@rhcNX9h$nZmbtY#_bIBxY^$4b-2rG
zn~mMs$Mc<A1{_YO+>*r?6yp+{#vF|Uctvx%)tr=!neWS?$Cks=k|wOq#vHH@9PN^l
zCcKXN9Jr0=DaU7kgyLT9ss7Zz9}Vnr{x0EYX*~qU=WNK4xpEG8Vs$hIlvwzu#I7}?
zF~@5y^4#%}7v`S=vB!?<kp@Ne_#E_^=YR_T&^5UwOiG=!J{xyrY3KZhYR!GjF%yn)
z;BijYfIXJ;zT)krF^ALTV~X^KJN#U!2h<pRRH`&}ZSb*U4ecL#ob;JsYa4G457<TZ
z*Z?YE9<pX4f7cVW>hj?DU_e$hC-=(mcXP~Ymcrl0|Nav4meT{!Mxwdn6rACIIuL(K
zKerN{nIaAT%JFxJ6|Qzhz*o4b(!l4R*>IXBI2l2y#W+Ay$1Cb*!Y`xBSjBn=6e7B>
zck2@tRG$50dUii@Uh^-cACjlNrAe{_e}|AHcsdJjz&fh`=Ew-NDecEv0pcY|Rxo}1
z^j$D0Ae&9WD*VPBJ%gF`1!@iN5H`X|JiZ|@AzIa)g9Lv`^u<Tg*Gk5M3GS^G(^TUy
zIp_L^nY$WQl|?ne^UB(h&$(&W7H{?kxFURa&A*$@F#XTPX?h0s|M2v%szfvq83N$M
z!yP_df4QqDk%V<L_vj|ewN+y=aeHz^-ipMBH;-u!P_qtWy&Rn58lRV}(1ME4?eEu~
z;KK!1O>Ue7yzdUg6*y5_V7@?6#oI?rOsd~f<n-CPV07@<lfkTh%mMZ!8DFpcmP&6#
zf@1{vfUH&1pzeDEUWMD$Peb1yT@T?iMIu^b%X>3O8=QeL#^U^)C!-s+Y}JLwTj?s#
zzfE2C_|f97buW&=AZf?__E84~tPa?Ugj1j|Yx8t{cx83pN?>R>sreLmc#BlPJT8V!
z(CxY3kB{*dFf!WyJDcsk@Oyu7chFW%53}yDD~K{(OE<_k_(;bsq;n9#KIYPgcL=tN
zF)`x>L)rGY0zCG{IAjYZYfy<DV@onIsHHyM_Q*t0e->QZs)9p{&{{KKGhEYN>h+G+
zp)Kt|WEHsT`M!o`_2V@*M33k&aWgXzLLaK!7$kj}oqPZ<7QZxvedk24R1O$y*)+|a
z>YZu)0a@#l59>sM7PW{yrV*YD5zxM@aFgC<KxuP4&I$t*=J{>6SY0g7hRv!*PsEFU
z_YQ?yA?tO>mHWj9wF1JSr0AAH4Ba?YJVl9MCyD1aae=RO$^j|<EUKUklqp}raP+0%
z((UPOh5v)(QM^*XSqj?(>6j!0U<*4P9*N-1^}3#(WwYV7$8we^BN1=H`3TbYE9B*@
zijeU;xTIF>ouN$3Ylo7o2;<oa#cF}hp9eFxA79}JhZIFD6)0O4yW6RiI;4uJoC!My
z+iwOo5Lh{=zs`hyQp&_IMHE!_a|bnK2xLkcFZ9*7=^a{4TnG02=%4f*eZUrqhLs@8
z3y)-x|53G7@`Q(cecho=m?=2P2BZhz5{KxO-`YX=99KgM?hnTS<TYEI289)t+_Okd
zQMK=R^UrfP-h<2a4TSK|W5=t0@eN{WLwL1@kJgDR?h9dUr_4MLXlA2_38J3b)EbxL
zlVcrL%p6J2GlT+>d0Q(D%IiMs5*{ZpzgU<f;edu&j%Zd)SuJuEaFI{_*tNo3;~`zJ
z_Nr|?l^`9DZPB#&J;FJ1YIN8Uw%Q!+>S`xr1~JhSt-lgdWrSspxjteNvujb5siT((
zzALs{v$8`peZo)NaCm0?%F&{Y^OLbJq7d_{d?y}u352_#L1R{t)Nd)*#i8N&*l5fX
zrp_;5mk>Vqc##!*YB-LqfAI1uvBexlYhRhMHZD3wl0cct4<&(Q{042Nh%`D*AySAO
zGS8Sz)!`(UhH`gBf|SffGLQ9YWJaA@RCkx6Nztx6PXCC_kT^-*vKNWVUJn&V$W5p_
zrvQxAA8_n~{Unyjh(?W<2D5~2<B`Ld6xeO9(}g*6VGuc&=M&(hKA?-|HiTW3S}e$6
zfIossp~iI?V_BDpjDvAF@nAM1bJp)q9tbWIA_b3t50^#DAljtT?8B>5TwIs0cKscn
zeT>O`6t1#a&6qkt)=&b{={hN8YkM(EjeFD-?^3*)N)y}*hlZDMOS-O_3V9r9nYEMt
z>1qs`>*)BhO#Z`j2l}*Q(Ovs2>GJzFgQhwoRd!PP8cM}}{IHae`KW^*QU3FH1di;u
zxzw@`%egzAR{g4Kse5|8Eiy|r1AKc-CLgg>By$hwl<u3OmnMR6>`db4H$?t9&EL*0
zA+_Su4!sGdpD#QJn=<U<7XA}LQn%+vc?)STKD~&ap!aai`v2Cp{<lzsif*<>w5l@t
z=0*mNv<go8j(^_8tn^HbXjRM%9Zd;X=;&xgjm%6;9SN8Kj{;_n4zfn}f;LvRHr7Vg
zjs$G9f;N^m_6oLo21c|$jGWC3jO0ZG{<W8so}P}5?LWCz>^u`(Rl#H{7+!xN4BC+5
zBFqvLDBJ_}l)P~pZ!l8bM5}S9FA`Wp2;@Xyy*<o>uhEIXMu;-cbV1n=8l|wJXe+;1
zieBgbe$E>q?zPMQo`W6o+xw@zw@X$^4DZL6|LF76C;Qj0Ush{P^ua#KTRxwe-%5A`
z0|Q}vQ?}3L$WaR0?z%o+PWAeOVt2d8=89$Ha$jywtGgLM^!=L2czHjYx4irPe}Yq}
z5|DC*fCItH*YHWYBEWyeE70{#xgx+D@E`Z}1J&>IPuuGqj>f-SYqrtqc?k)D2o&JI
zbBzHcz3e`7s>@{aoOjb~Pv;7*KIhCL!CSAlZ4A_z4QBBefmO2F;#zN1OuWB7vI|@j
zfI_R|DOg%s3IeVTZ;M#KL{S|vNGc6{{}2qm&N*<qT5YsgI9HcAtt;5}d@f+z<vca6
z{JY#LjsibFKeq;%*`|{btGI<cEOq9^Mtd+g^SHm@uyh=0&^zwkZM<AM5AVkl5jX{)
z5BwGeQ1I}qM_lTK*kch8sq8k0c=L4dO-`rO)NO@vPN*)I`6AKwF+2QFXbmhyl^V@S
zr_Gv$*d&*i9&mVpe<%ZsL?stx$ZNqj;Xn1m^V<F^uR^i3(kTmnA-0q7c?QcIIJ|iZ
zIIUKj;dDbh%G27G+En|8cKoB;68WEIq#KM*Ofc6sBzyY2v|}&5t8WwuX=>v9X$w?R
z#oOqwA)zIQOs0pQ+K;rSE9}EVL-qlGwTY`}Or<JCjoGVt7{_egw+?s-LxUaEX0yx+
z(`@Px9G}azhDG;1A(hAD@dEX82`B5XF+tkf+dEsXHK=>c=*ihGC?KXD59uZY{wr<P
z_=;c~(A^M7_#OTpf0XYC$beD)pCYisf56xKmHszBKra3kfj@Eq3Ox`2eocS}@(%$n
z2mpb(^3aF~%Z0(tq{umshl%2eG%`w10QEnK^k$Ihf-V?LW@OKr9uL9oz;#qG>J}H<
zH+v(<0Ls?{vNZz**C;3`DJiT;i^eGLG{rZ-Dw8^OIR8a}Yi4Ff7?(?_d(CA9xEM|g
zQ2l^EUoTSp5Hc#N(H@87f<F~Lv)Osr-wjJjX~^Y{7*7}VWf$Tu;bd^^LudT2A}n+c
zA90lD;9?fDhL9(c+*)(9-`LqP|En%YY!;hJgIz<*rP=w_k^A^AcswrILyvH9erzss
zyuFh*#{|DYe=IR+=_r9*+0WNdaQ=F33}ocp>D7hhs=!c(L#)WOiU%DXN;pX<G=Lrr
z_`6=6KOYxoM^LAWJNP`^Zk@g}QnkBJ0Alh2*b9UQVs*9M6T9n3isSwJea)9$_#o23
z!QqcW_W0*q<2BftQ!oCc#5ZEF=Diw<#Qrm0+x*k6@GNHczI(|zC%s7*E|;5y0CoKW
z?>FROJ1MXQU|{Dqe|Ny@VpQ6-GA2|KAPF!UPv%)%Zu-abb^5$IW=?etAElG?b3w|J
z!!lWUe-~$Qm5oVGW}R{g)EbL|Jxlj+jS7j5F8(|50U$nsaRQ2`&$owyf`SGH2B~0w
zx`%L;qinagY{yt$1g15tbjCha66gV~!aDI9E({QvKSFjDz24=uHa)7Qo><`Nubev7
zpVnk&#PCn*YEKG<&Pl{d@r^HTEoDS_xB=TgDOfVE7hZ*dtV&#}g9gT`su=yffcgMj
zop*hK#V3;DEZ^bphY!+hwGaHg{YOxN;LW)X`}*&-cw+r-J|1!|H&z#=prN5L;r?kU
z3*CxGmEIhD!AyJM@mA95rAq(o@7evah`;64Ro0`$`~4-^0sRh#%RbVF&)Jy;_@6q7
zi+uI~=H@o}`z;RlHPz?O@o~wZpuc2}tAv}LTI<}#a7o_e3GF!C;qp$$%v{gs;Ees?
zUpZju2L}cen{8-84^W=P6M~t)pE=muW8nVNu%Th56_2P%6AwimJ-Pc`3V^XSar_q>
z@D(ogvc(u_N?flpKd1x*1pa$5pAm+QT%B^)=^m%2Kxp*;mXr_^qdWggTm5=$3-9h@
zb3K<myJh$v2F9gKW`&nnotJDmSw443z^=htuIIgXdOW6J|I?y}frRvQY(&Ibuh)l?
z5^~aiIwC6h^Iv-QmsI`HaiG77kgMu{{g+CE{MGav-B8HCoCrW$0a6P{q)-jwlgt6a
zLvbNsVsf0!VE1@91Nf=YcQG+BD&>k4RvT?eNf<mu4R|zwO^|{mS0yi`52YOPClS!o
z(_3q@;&i#_FkhfVM^BC=5&_(nFgrUwJ{|;zOZleSXuhEGHloI*Dv!tI^%7C&fQ_8~
zC{=-@rMY3f-qO1cpyfm#T3jdbrkt+E%-m#0-&lixvoa770mG%5fy3+FAsmV2=H{l)
z$<3`LBXjxj{;KbBJX?tG{gmsOIT()aa=Hi$UzP~&a><!<xlKvk`FN-UmW69(cpL&n
z@rU?57L!2rYw~QHPB!i@emA4eCRHY@Z!QdY-|_tX{WTg)bGbB%r8A|{Swdmm0?98U
zqN2PXH(aiB)ZJ>TD&reCo&#yaQ|)>eI`Q-Kb#qY$QSk9?_KwPKDrMa9>REy<2mJ?!
z0ApOHRQ~XEZBXlVIEkyJwE+r^7IIE1m1eosrPU=ZFCRxNb^|zzxAWrc>>MACfR6r%
z-GQrbMWgssC!ruOAA!{@$L>5ZGG0cJPbK{W0RR>@gJNQ0puEt^@y6zeUSM7&%NdZS
zlo%Ko#IGiN1%qJBrgMCYi;F8OD+>$v>x{+*q`|?1r>A9~p`!7)k?_VxM_-2V(U)B8
z?bV;qySlmnx89BiHxn^3GP1Hhq9$XlG|6&1``J%a;9zLk5oT4<f#|zqJ727;)gE#!
z7`UAr%ZNlRl|>B9dFS@Ud5!5vN{)vjlMRGT_4V!a_yX(8G+S>NeuJvg?o=<8C(%Y%
zDqj-9W^a0X#Rt#McI1ISW&agEbM1$nke8C>1zqDRjh&!!r&={@wC)QG?xxl7gU$2F
z_Wop{zP^62sk_^!T)EPa)58PcG&ivSj3`kTw~gsB@x+s~HsndvhRR&uN3xE$o+HZl
zgsK$Mi4?6$Cm*xtsi`QVPsz?K9va+s^R?hy@pb=94ks)gPgejJ?(Xjaf6F}gmY0_u
zk7tG0q<_(KaA5tR4r*s-hwnMNG`?CJrBx3^G*nbjK*tJ&CT<`J4(DaC)ZryXH5??R
zL36E98Hi8(-3}gV*lN9no|u@Jo?fmn%;kKw(QK|rBpR=CXZTuI%ngD34_+&Q1E`ef
zUrvNr<evwS|F8c6&L840(9rkC%mbj?U(Wx(&;!>moWC0Kw;TTpIPUQ0{0lh#r|97i
z{BZrdNB=7!5`a_u;{ZS<{-3OnuiyV0A(Zcef7~;mCGs_&Xvs8Mt1WiBxq`uXyx!xk
ze_;*R7<N;cULN<m5;KdXlXGq-@QSKE`mupNJ;7frq!E8C=zxC?3Fztci$8;cj70CV
zPu8Ll=5k|bSeX8=zl<jr1eQ(>z_l|w+r4V|-I$G*0%`AJF#N?WfTu@4Yjrrh+u%{T
z->w6n;ITFMZ>4H$YlkX5O5Lx<Zt=3CbpMsIDqC{^H$jqZ;UhsuuJZXFJrx+tX!w`O
z2EY*V<(8Hfcjrm>Y*#>Z!bHrr-s!4^anLSFj6dKZ1f1UkN?EbDN~@hzGP_ZKA?@kO
zbtwGrj{Gx6_-kbT9YFlQ8R9>ajsKO<|L$o1{SfEexnI3K-*UU()c`(CO-YXx+MVwG
zL9qYSSa6NUWxM^es;a89v$MUuJvBA80}yqlM!8ZQD)1|S4*-uA`pvWn{B)_B*L0d-
zCiQSKQ^Eaeu~H+BnA`K|58U}XOAUB@e0<<#US8gSv#V>%tiMba7vgKRcIO|I3c$Pr
z2u#cV6ciMFF0rw(lz0hC{;W2ee=CB%J_OboKfga1ECBD3u@DjxCg7oXL_|gcfDAxx
zKmd3S!_U}*j)_U}`}@noIj6^i<?pUcP8WAq*AQ-tr7G&I*Jwa30Cfc+IpXgvl1i79
zlq~2j(QLiB*&oyS95&F`4-n4G%-mRCpUmPeU&lv7Tg(*%M@L8J<?UAKxJLnW;}5_W
zAN9|pyxi<ePfb;mm9^a;O(YVD%HZ|URy;c0642JpDlILIj-CQk-DJ5UU#mX|FsPyL
ztk&xYJR8k6Tl-_l6B85u{{H{8ns0+EB_@Ud?Wf6PhTHBI4HKm`zZ*Cfi{;Y<>sCnQ
z_3kjhZ)k2}Vquxv*!<}MfKMPN2fY~?8TR|5nLJ)Ekv<-Jdb#=ey8s^)7Z>+%v7rGl
z<qZu?wcT3n&c43BH%>ddyJJbzuXQdP|5!(UF*jK{*Q@Q#PLJ-1>?n4Zi;Ub{y+|zP
zySqCUt5xF5hT7VdMvLSQ>T&YLh>?+z&!^qFBB^o!s_g8hV`YsF3HcS&EGWrIMHS9o
z!X6qL`q%8*?d{3V-24Qa_gSagyAv>o@$qx(tq#CCdrsbOfVz~*6kA=d^On7@wtWF?
z>S(jv>}Z`0eZ1USnwfD+t?>8r3kbPH^9_VK<wwKfla#eQ?ajMHQTpxVt@-_>qM{<f
z#N+;i88EiBGu1l0Ha0eHhouU|=8F~2bsF1Fa(@|men@vZM#kG~3D->g=%}bF+1$Qp
z-VgGo!MX;3$Z|TLA*=<%<5#IvYrDI*3FIDacD~M+D*S6WV7Z?jo?c!~PEXgCmleyE
zt>#PQ?BAi@BXK#E&k(74ozGW|3=IKn{Dbj)h25{B4==H#L`8p%jVY{TCnnz4H4p!j
z3ShGBa1^evuyDCb^~dfrHfN-^!1cP%UyzT0l-w9zyVc$ZFzLG)J_M)?=QGLQ1rg$f
z-DN8Qru5HJ!q>)|oHuYi0Al<zy%fXdV)Oa%^zrd=dwn=xgT-b^AT9-f^Q9`S{(l_C
z#6wQWC;$vytkjF}yj~CEr>CX8JU=&DuFTHNApWI`z9~#={>R71`uFJA*i=mb05FzJ
z`?;W^6MC-EY#sUz2?_a^?n_9*@cVP8WMq8zbOE5M+V?*qc@D5~&n$z^hK9rdv-=0c
z3Qp_7Q=){8CQ^ZefP|a_#+lW0ns3wR^TVqU`Coih;3iMu@p51JKfJwlRA1ZHB?y5)
zf<tf(?(PmDxVt;S-JJvv5`wz~4IW&A6WrZ`ySrQO+<V_w)%D&O^^LCX<{#jv=j^lg
zk~!C0r}6_m9bMo}0)sBi@$W{L;*5-p`VU}5{`jWgtT{q-=?-vK5jd=359)JiAZ@BL
z9isW4_ij5H?7w#6bim7(FF^_{4WdoG{kNv=t)Wy<oBjzxe`0XtwE-v#*1IFgC27>l
z(*F1@r}11#dadgJ=@nq*@Hnioad142eskyjK`T`9Ws*S|^Iyl*PUwBM@wu$*2qf44
z_^q?;ec&$sx+Z`84vZHQF|qLNwMm$#!jrd;ib@==r@}V{6rR5jqZR$rtQ$>!{^p@n
z&SH0akgtCjNEqtxS0X~r&&#{SyV69IhBc-w@QhqWUaEgd{I{EVa^LqOx;Q&yVP@v`
zILkc{R?HCpf{*HTcd@I?cm3?bceBwK>sc$;XzU6C<%pcsc#$){fWsZ#c+X$>MN=NE
z=r4eQ#gi4j3xYF3)X0;<sIalK%S=fr5+J<LttfhcXcZ)<%Je^rC`p@`nDBu3Cx>a-
z+=dl3B;>L;0H)9Hb%oF8egYQu{Om03@Tv6+tf2id<4}}GdqDf`lt6~(-Xt|4#6JlU
z?(?+?&m}h1(4%lyz@2SoCbv^CEX3cq*~)@D6F)yc9UUDL6VuzbD=X~*puVx(5q`cD
zKE6u3_J@UXilD=H_x}6*mu3S%+3|)#h6-Fd6=+>uP<@Q74L-W?Q5_Y<D-XbBmPJns
z`cSlgxCWK2X1zDY?)qQ;8N3k`XBcZ%i;b5l>{h>GPIks}KxjIiuPL6L1AgDgYHa4=
z_j@_H;mKryOWaod&LWHT1OoXqF3G>qJAclr6;MzF9yc{NKVOS>1g~WSZw_+SzoEbr
zMo;^Po73aw>!pDrB|0>IKE8kLPG$aC@CD4q-AxDM<kAU~<@!Yb+Hg|=2?+`BTp52l
z0|W5%^!!(FEhH?_t}DyTTpnNm;USYu@?TH16YA94udXMeprQHy<@SYSQVxZ8TR;8l
zpS^Zae)*j)|I3(VFSG9cl%an@ocmY&fByJ8MFpkjU#aT<)8Dxs8=IT-`ps8R*H2}H
zP-<m}C@KHKNa=~0nNOQ3jsX<)neAEJ!dIqi34ZWUS4kN-)xT^?63>30SHf(~keW;1
zW1Qe;`SnWlKL?bH2rw=Nns&2-tR~z9r2l#pTa!)o(!jpu^9SXB((Q4WJ$74)Qc17&
zfZ`Zj9<Yc1#^-F_y00e`mEHq%^eJv-KEjgTGx%sW=%0(F)sgb??VIy%F(L$xJ+9tr
z#n?-IJI~9@e;rvT@Y_FuPruxsdEO5JtM~T-548UW^zXj_;J*uszfizrsQ-xm{lA{v
z{RKt;9sGfT{FPAsFDSx)PVW8_9H9|+aBwg=sq&A98&yFT6BXTU?X58G`Q;n*j@kF&
z2E>n)q$KBqS!Pz&H{k0=SNwt7MtHqAy=?V+v?BJJh)Y5yD}BHH<iws>V64rEmyWJP
zE}h51!Xi5oWKdui{DDTyE!#m^r@>QGbkT=k;Tg!bsWLTeoZ_-Wq7KRYNzG$)V#4WS
zXY9<i)%VeKFp1bMlg|TS7YKj8ank?sdz4x7DQw<~_hw<!_PoPH)92UK{Q;Vqni#Zd
zO`exU-qT%OAI`Ui&aMJrk)55L^;{`I2H|;q<Wu(woeM;#JOIfXnk?0Sht5o&7A&<{
ze=n1zN~i;J6OibgmDJQ+0`O|{5adddF)^{R&!?~jvZ-nfj+>r-hPJkZsHoA1b|Fc%
z?xr31lWD(xy+^!hMS-o4jtnn?4LFyj!w-EO76L%2g8cl1n^-?Hv+b#3b=r|ys~I`E
ziwzJcZ$L_hiHUjENVP9K+8ZPmDJJ{t#mP(LG8q72%F1N%@3$u<C4oE+<1zFXD3`<{
zu!-&L7HUf!*1H++HNgcsHLx(7_9ypTq>vPD@RMR%k(`=-tP8ZIq!=YcmP_NR04xKI
zX^B!^e|rEd$fPX|4N<YN|J00N?ZaRfe(hF8lq{D|`=qoikQoo;OFvAbyfKf51$Fb|
zqT6!e0UrPXKq>*&Bau<ROo8f;D;r{%jf;zmrIgPA?jIu)i-H2!NXwskU`_Z>jkoa!
zv-<PL-#hUCt>55u|DA04|9h_~^d+;JHThr;r`1uBkocb}fpi<3xMdXVAKR>qP4(>|
z9^Y9U-_eCRgmJas*DNA({D&`aKe@!=**bR=dGVLYg82G*!i?GEAE?>pB+DX)LP<eS
z5m^e{IHiODhvn~zJ*(RbBn#M@JS}hAw<+HY<Rc{iU5YE<O$V^k6XagX>z}MNmNeE5
z|Bq+*ZZ^|iVl0|H(zj<dNi9;)>s5piEB|w>0kEJd5+vXr?)@fm+tPw@2X1gMK5f;Z
zB9JEWH|Y4CoG=cQ%EKUsbUXU(0W!62>3^QFA)Ms9IuacH=B2-Xy>NxgLc_svrQa}Q
zH<EcgZ}wHSs&?@X33CCRju85feP(%{DIUugH+=cJ^-Sdx&mhaMyI3|AO#jW!BURaS
znB9JtVW^R=bO_}c+@q7jWt#6_41mHS7aFEo&H4|lCCL3ui{IZRU0tv6mQ8hZ(g13J
zC18#dXJx+p1ggmY`YJ8vHmfwY!;*eVl42je&0nA2*m<YYs+Cf+QYmb53?LxsG?>!A
z3<p*Oiu3Pk^_N66LoG#XbO4>YzdrU`Ur4Jm_{`=QvPXoE-5_#MVDQJ?gOivF6CHS1
zI@Kv&qa)pBI6XZD9}AML!g&|}ejBwz9sxSK2*7ash4HG;0tC=fYgT#*E8Y($>vw?_
zPs6oO>H^4Xs+}!0O&>nB>Z235we=Uq&&s#|0s4XU6yX`|&XkC#uT$7WE+8|p*|8Ws
zpfgw0H4On$%C=dMz0!BN{!!?3IqJ?GI822ff8rqUUzXCQrXcm=1i9DKVjp0@TJlw?
zH9$c8vlIUmqkp~XZ~EhJ+G^z=`s)AmH~8{@r@#Iy0D_wQPi^ww044q_ZvCGkVE_R5
z>#YABKL6hYzE25WOuW2apkl}4v~2>_C*Y{mPJw3w<#|j}(%<mfFX)uLcJO>sa&kjs
zBV0N+4^Mv-j-ePY?@tib5DEF}Juml|nVG>Myep)g1?gnn(Vn5{rbf@SN7l^j<mBYq
z!y!m#eS#y3Y}QRo4`-K3)vc{Th3M(2P;WPuElR^XQ>s0aEehWdRv@3bRA;9T4DXvW
zhV|^}n<<+V>8f(6_FE-|vZEw^Q#B-jn$l!|`9x2TxR_UsZ;+&fJw}c`UU%&!46mXt
zqrLf_3TwCA@;J<UoLf+k3A{AGF#3-Ur%M;c^S}&kKz~cc)BHgy^NI}scPj0RoXNdY
zuAE$P58OhTyYJ8YHbb7+FBp%9<KteD_}YJcViW<dsk!;i!NKox{Z;^-S6fU_aB`|2
zxY^s!*4m_c6P!XT#pHr2+^;DJlrY@f-1HCd`uc$EC)(;<h)zzX_BTMX{Nnziu=bjs
znR%?<q6dqOOzso6i%ojiHd1KKb|!5*2_@-5!de9ge*W%Tj2UHA%jrskpu|w%IyzbH
zH2Rs9)iyot`gnI~q|N030@Ilb2+UnBzl?W@><rueU;f|V`}ws`?I1G8#i7?U+X+5D
zR8=Tbsv^T|PS?89W$-FYj0<1Z!6S3nDnE!QZzl<fNidO;=H%)@|8(38vrlsmrpUZJ
zJHzEXnj4@1gvKh!dN$Glv3lHc&rlu?qIO&v*6)nc-sxP7sjtHuu}4thyXGz~$CN|$
zC;qVN6w_5s6ch?|GopZ;jEsyVB_-Xg0<g4pt#$J3$>rtYcwTsuvNb@eF~D}E(tY{U
z<(|Q(XL!64*3Z(s`=v4bJz#FoN9<Z1)`O!(l`W<>1^s>=>q6Zl-6Q*m$Yj|PqKCR<
zHZ(Nc9DgMi*_$rC|3h%Fva?SO4qE>TMHLVbP%Tz_yjrjWRd%3~`VUIEv}$KtW25ty
z!g%AKQJ&pBJy_dztsdv51IRy9HMmnWIIA2i+!*-Ml7#sAsknUZE(ViXp<cX*WBdw-
z_U293yn;Iu56>NjU^wiv;45Tgc_}F)BujvYc}eZ*Up>nM{wZ$74vUngo%ikhC1`En
zQ=7s+ML67>D$cupT5o?w=HugYJzg3DR|PzVbxt|4`W)ttYiDIO03HH)TLb|wCbEAR
zELfE;{6DO_jlu2N=3oBquXysuBmddy2|T7F88)V-SC9|I*m!sgb92nF>;(XS0yXUi
zq&>ihO6N(%Ck3QTd{@i?{So)~_lZi#PY>rKe+aRc&osou#K163ku2pD6i`GIoItX4
zdO4%d3d?Q=wiSSZsA`exUi@1_N!%+eMm@T_>B8rSZNdI<5>iqC-MNDdEk76OcasMz
z8QBJ?T<7NI`qSZp*1+nK<O9rvnZ#wKBM3f~(~iLt|Fxr|<JIvp5iV{@OiV|pRA5I(
zM^X~Dq63XS$q6uWdhJ?#ZbvczEAAM&TpegSXmoXTG4t3M+?=fTkBnhsV<X_OykBZH
zAIsLNv5ds$t^_^^9+S>MfqWnggBIYH00Z4VI<mF2^aPRH0{|_LHa0fRO-%@hh=9Er
z$dXC`xF8!iZcwKJhDW>7<a3iUpIx-zS6=%SAz-Jc-oe5ctQ{XgA-aYgO;)(0F75sN
zc<s1b1HRN{e_9G7EEn{nY;=YR)|RM-7L?rG-JznQ8u5~nitt(cKAehs-~en>^p>_F
zD;J<`%Dzs*0K(cdqc9pHCV6A{*@n0s5Qp($c3Gn;a+w13Y}cT-p~44vVE_{W)-3$6
zyyZ^HxdeYGS2C85pFdfZne?5<#ZL4q(XKxP(%I$Z<NB>Sqo%g@zzA8n?oR;Su{WIl
zqG<trJNRnzoggp3V>5qC$uM45S*fH)riKq#Q^WV<p&AOH%R^)dGY=O9<t+z@o-@Vj
zj2>pTwzj~|G_B&^?oXFmP2@+OSr4VKgXD*rL*U`0GoDr*Q}-iNF&Y3aj6}r6#cxk<
zcP9z}nxtVJN|_A6@vm-Mo13cN6;{FG%g9Tc1NWOP8opt3iZDGjrRovQ^;Nx8t1DCr
z<rIIay}f;Lv01aqG)F8FEu>(U+i|0JMDV$3Cq<D?GD%eu6%$h}M-1^aa3gAaIZF#f
zBCs=h<w@hqOZnYTgj5{l<dD{B9FU(bCgep&Xr%2xkp!R+g(nr@UI1v`cTo*_`V}SQ
z5PT?ICdTysWMS;txb0#iy?#p?b^O~7HzORXgg{?kUvMqsmz&tEXMae>Qc#zZUJ(iU
zrfJ5bgMKI@ek!9BUbiv;|Lw@%ey;$__%w0Y8G<Z&eMWw{x6%%PJ_j{lDQd>;E-fMJ
z2yAAfwO^n6GmG?FpT=^;lQTfy$kt$#z@3DLD?rh%uCDs4<mBW)H<XzaxhP;<zkK=f
zUE%hiM6+sR#^zTy;4C-OTc4l2y23C(fElir`s(lR4+V<^4Gj&Dz=7^=Q((wOW+6xf
z?|b^yL&L&~bQ)L!lwgrf&CJ59`a3NFH4X)X@RjrK`na60ghsW<`EZ^+zu+{q1S|k4
zQ56dSE&;yo2h2#P!|47GTd6=UYH3--WzdzPE5K_rUk$*?@JLp(VG!v1>>S~dk;@fx
z#Kx|DMP8phUN46~#VuPg+buQ83*3SY=zeO%$;r8ug9QBWhkJ0{HxGe5!NS6ZM}pk>
zvxuY~rSJV=9R&qNKRc;8lhKr&wRMToH&U?(duL~z8q0Y1?+8Q!;DXSB2#rpAQDZgZ
z>E#vf&+oD~Ni-t3ISI-`z#Rc}cB))o|JD?NiW;!?jbIB>E9V0@U_DnU+K`5&%j0@5
z3;1l6QmsqiWDtsyP&a`sg7*lnwu9|J{d{{2d=jWVuP-kzFD|?RG_`Sk7uE^lnVN2+
zGirz`;7vCCHeTAeAJJSF^^-8Ov0a2Sw#+uVSQ#1;E7NGchX%VRncZqShFm(V#p@85
z{<|;wzPIGhf!~nud7RC+29rUw0kKi^TG0DOp#<7&Bx8uyn*1H;J~KcClz~e1<1eQH
zu=)jgc^Ld9508&sh8UQb-53AnVgL0&mNTFSoc1OQBk?!^Nyx{^=>}?fkbOtIgn*cg
z1&6?S0}m*5LVi!G)-NZB64KJe`mHU00AqGW#&Qt!C1W_OXW#03u66_g-rW=l0%ETd
z_}8GK1U|Fw2%N30t+kQ|`aruj;5C8UNsNml%LE~B7qDINetRasPnepT4&sAv8w%*J
z0=Pncem(%7RT{K_9UB>CpIQ6*ws6t16*$%a9;USe^qzp%{tkG7j2y-Cp&vkEx~;yt
z`ucvr^&K4@H8wV`S%PTY@e!2HWjYO&T(qu$&%XsWcZQ6MhX)F5y@!$~r!NG&2^kp-
zB{Y7^%{QW=jDTW5?SzGZV86)9$uS20x~m&m@Zl{3nk&frdBBBP?F@m3gZou$pA3j4
zU_>Cm+87y)X9))!99TO%f)+-<htr<m;NXIc*pw8li)C1VPB*$8mo>405_S~|;$w>Y
z$*Gg=QWN)wesl-5r+J0BfOmTRw3K9Iq3kPWnLhcyoz{N6b|3)6!tFs4v$5q=QC^eE
zrVmRK0~}b@U14&g(wQo=5zrQr2OE$11~i7~9K1$GX7q%Ifq^-zyIvE9fS}4FjMFb}
z@_lT42>QlS2^;RKsAXV~HR}Y@xa>LuE31VHV{yJ$5D<AY0nhHy2kqI*op`voG*{lB
z-$S(A)Y|%Bb@dZ0B!tj9zdPU?2nf<~SWT&}by~bNb<6;BP`=&{tO5@ko5|^~Z+lq3
zEuf)ME5SxXivigunK^);Oibbu60Ud%|8NUje~>*u4v312vwDMfxcq6#0BK3021fsJ
zaYb$Im#{D;y*o?ft?lh7Y~~Jti5;*pGaKGOKqyFob~eoLc0m6=dk%GUybF5Q{_9_s
z`)|Lizdl0-!TJXM$1dIfaFbpRZh(OQ`T5`6q_+!iz?PKr?5KRFnOta)INYuc<!#?I
zM1#$m_ScCcQ-<Wut;*I+8je2G*vEvxmtw|0e`?T8-wU4}Y4`jrXG|PPWaKR58A!v#
zY#b5f`NYWe#K<MU)>!q9iB$CZw)y@r(uwJ&VMW}5f!vqw>FA_Dpjm1a<=u)txYbT6
z8nj>M2*z?aAw$W)56FOS<dsw>chX1gz_lCZR~FV$SedVXYxYhYX9~8SxCb(F9`f{5
zgF$&?b|pp<AM&5moeoJ{KhI*Usu5<AAmUoI`$4`V-h{iJbbsd(c`Q2X6=^1BWva4;
zlm3Ni&-*AC&QM$>Ezm<dkE}GClEP4p<~bqj^7Y3j3Yp*^x$%Kf=4f`Pt#4(l?#J}|
z3v^!$2L0HkY|`-Ay(ZxQdF4F1yq8c&KWwW$@<r=p$qMp382yv4cj?pJjrZ-bG(E1(
zCrR6%em;KRu%zH92(l7b!=;{P{2B%Io-~c=BfDOs^KnwGMz6dSw^2xsj*;AtS?MHx
z8%BapfmUj-`E*=vr+c}PSPr!~{`OOLH-fhbG>#VamNz0D9?J=&RIolNclTP$J6Hcx
ziQyedNjV-W2;~klM|=3goLuKP=00`t*Nr%i{30<{*6&j@vT<s<?8nt2QO=A7S^4i+
zw)Y+baTV*3b4>O6VAk(9JWly!+_gnw?cSSWYzOp;j9+-=2DWJ9C`wVgLttAsUqdQt
zIf!4N#Ey?A;Lux@;N^bEeqKkt!~Z5BJ=~o5mS;oUyW33~v)UjFZ`|}(7Jt?J%xJTc
zDw+Mf^n~B%-5knR1&A~6e#(y`6<P9yuGIL8ToJNfm?vLWQ(y)zZSTl9WvmvAh7E0_
zga^kMgUrxy8Gai&qLEG1MqftsNeZ<t@08FTTJ}uCaqr_<uW{^u%@dO^k?q{Tey)y4
zFe6#-!!5s_<D=Z|?A_8W|6Q&9!v)J`oGp2%WGJe__Q!3#dL}(Wf%541M7ck;JafP%
zl6>@33(r`@-78rhT@k@qRSCGW@u{B&&KR0AG<3-J7-AP5Hj%d#WiMD=pb)W+95AMt
zw{Z{CoiOCG4cSx1?!$gsEAfu7nZAyy>4V#El+7>jh^jvpNKNZJgGeorueOfGxQMl`
zoZrlnoz@^Ui$&Zjm^G;xfU>UauK)8KsWsZ}s9Yosb|}VxmnOY<#@#JVu*rdLq{AZN
zhoUgXyyj>mqVNo_1KvrznL=Ei%p<%En;ULTS=vDd`V=EoS-kL$%EA;mY#18{L*9^9
zdGwurrcZ@@*$;<qOOJT^PZ4duR-Rvu9(&*|J<dH!ZQj8d4Q4?Z8R*TE|7sFJDarJz
zhJvH_k?3&CnU%oF=rL|+3B{b6-p4&QlJ=(|sb($2S)u3~R=6_Zlf-j!>)xDav?w6b
z7Cwsdxp-{IN^Cjai%$t!5pgr+x{~l?5kb!M4$7$yPQ~C}o$i=2uVYfn@e@mg4l`Wo
zkKmu6gdM_;_-^sJovJd<FZcSzkn${y7Nt41NM^u5>9-+->H;^!>hw<(Qr=L9rdfrM
z#dzj|PE!7^h!-BKar74-ARF^G8ttXn$q9&^8ksJBf3$c@|E{OZ5wEnvm4Gvcgri29
z!curEQ4&!R-m!JiVx{C@q-xt9-a9S7pRwd?EbVaexxOu2`isUF#kw|DYTu*Dd~bGp
z>fg&UGS{0ltq}=+sfuSWcb@g4<00R~e*>3s5YJ^`jgK4cJD#7=6!ZHaKL6X0e_6Vl
z2YKB9?c9*wE}!JJM`E`==E3jdGbzskbtSzY*<!dv14oWhGnygau)a{bap&D)xEcm6
z9L5d*Opj@ppvLf@-HPph$*Vg~&|=(0X%T!mz%O}fi@vZ2v5B7VvnN5IfQ65j|659A
zrx5+}L$8|+_Z)%~oTuk-)#&+-4{PfXt5&@FMuVxKo9tzt(5$a_qh4UXaPdLNPm)p|
z<eW4kE2+8i&6JZCYxc35-QP%GFHInsSYbFe6*v)JoG%QS!O=zJ^}Z^?*8I#Cs89Gn
zJ8F$$j_-wP1Nr>)tz|9PBV)kJLa|o>|L6ivEniM^o{SK_Bn&}<6Jjkkc^HCAb4N80
z12tHHga457Du=+tfr-#hECGwR=d*T3r~X;$<ZkR~ERDdzG7}Y56gttk-j6$a%f1?}
zG+fk%?+cp9dKhlI{z2y03=6af(JnJ%pRCc0ZfMyReskT4(SIdejOo8sb$I;7=q}8A
zCLt4T9J{0Vj<v``iz8QvMWMEvj&hbxy#Y`9Tzdl>Vg2U)Zd-HcHU~ec=yFk{)vz{F
ziGbW#Ll<+|KyAalR`6sBHT|9#8sc$N3;3^kXHl17%Ig$~pvGlkZhQ!D@%t=u1E$b+
zp+qhwqD<X0<YP+Emcl9xs&0<bz2w&~)Mj|NrdR51C1^uWc*(EEE=OG$?{rf-*ZplS
zq?BGjhU^e%4Ms}gd{X*RMT22%Wn1z2X|V*CzQ$;VBo-Q1b@n-6+g%j`?}d2aS{Hj2
zE&q^6VewaR%MI1%CUXg~Udg*KHkyefb%fWNk`_4G;SXk6m0L>*-!au0EvA0OH`W(@
zas4LAOW+p>m%$wA6^3d1<@4JBU-9A4IQ_YxdhDw5G1ZPM%S<U2hH75@-t|;Qtaz-K
zy2L-t--wrAL1L1~_3;E79B$?{@CJW`r3mKLu4$)S-6GXD-b*+5J^gu9k#eagI&6@e
zh=C63o7E0IVHUDe6bY_qaAv7W2(f|4g}+h<MbKacZ{VpHAG0~#MF{sZJ%t3ra$%|{
zMA~WvE%ZZE<trS+V)o|8`D90vtMQ{iSX}P`g>ErFBXO5Z?e=CpufsYQC4<`%Z`53)
zeuQ=(a@ALzlbdZz2SSOn^|Z}>eBr@yB#WKB-&A_;jdB*{seQ{#F;jz^U`Ua^cbEE`
z>Gc|m=rx8)H|K;Pdj#bm{^EZuS)}4nGA;Ch7{NJ%>q*9Le0t#5r!e;x%0#@(&_=MI
z3+a(&!y|z_5#LTJ;E3O!7+M#2igV0b*$UbJj92N$qpw|xo*1s?S=&S2k}a9v>hwg_
zl)MDHK6GIjG@YH1dppnkgDK~>R$}m-zFE!&l5$hW>pX1zmIcSaHyN4%NSc}yU8ART
zj2NcyL+X&SV{!OJbnK3BYmL#mAyG9Q4Ah;Xt=m7skH9_t(s{9cU15wG&1qYvMrkzG
z7PCSXyp@mUf66%e*)Q}icIs%XGJJ^oE%5p0r1y2BwPb{C{Z@jZnuOqQ_X!Upb&A+7
zn@Nf+n^p(she_Js@BPT)ywEr+cGOww$|Qt0-#h9jb3`4^YbzFZADdn^K?}OR=*LVD
z>^lBD$Yzp3kr3#agMf5yDGJjx;!jA3O%%a77QRW>yB#P>7hPj-qXMn0=D?gxlQx?r
zMyslgTc}^Ozb~8n?fxU1#x`C+pJF~;)x1)|+YIA$Ii}K(hDf@~4+7ppF5Ko4q}_^8
z-lskm6XHmBY6dp~E=?17*>DKU<q58M;mEAU$vycaf*Bpg1DdXWD7UqdRsBJl-l$Qc
zbc$Kq)+=uvRU1jfCU`3%Q$wN_;|8ZjS|L;TM)>|yKFiMiZ+9ACVFoha|5iS$-50aK
zfzmvpegX?4oKvAzsKVZ<S>zqg%9$uHnI`QW%6R+8rsg01K9nRKN*wJq2X6ToZ?K!6
zOOuzM`;+_4CgKOPBvpg=*QcdWo^88~_Kk(AW)MkYI|40k`f^~{<P58BjMMu@1fRN#
z!r`b_31SeBFSrD8U;94b>GJc3a4|Hgcf-JiF$l|Nbi8pcU<;ebv&iL$W>^d`d-iy`
zJKg<MKn?*$yxrcNo{V%lqxQe*7QMc|hV|1YVX7wKbCGiqd^%|1f6$^b=}*&qj(lo~
zlF;h@oI9N&5Ls}$NqhX16_RPr#Ix{-7wmi8$?eqd;^)?i>Z2E6`$=e9i<bQb0Y0l1
zibKz+U|5!yJ6SuK{tgMa(}ghVNz6yp;bmSy%cE7K)>l-LKWOnXQIa5Au>>XD9m_9@
zua$}vx>nDWPi8Js^Ed6L+y&M2i5NZN6oQr|Xy6Fe#}^@2!_4E1E*{^GQAx*L9A=;9
z#Dy9u2IjI^WmoLp3BIqw>A)D1==-=sM2c&X^j`UWPK?A-www;ifU;-Uwm7xWmtUv`
z0nEAKqei49sfn#GXi1=8e~!^V9Ldz3&_chVx>|y13w%dXmh>XA4w5KtpV}g88N2)g
z)J}z7u#M3=RUY1EqX$1N*DW~?pM<Uu3p;itw8#m*ac0bA@b*tanWuwQoCw0>ktmbW
z3k^S$aUxHwjF>VM2MO$n9Cvw+wI-Y>j5KakNDq`@e4KVgsPt8pq`}x_yO9s#SL?c5
zEP5n|O1eyhqpp!uwt9yy=<$(?1H@D`lt~F1Yoih2QrMR`<v$fS4(}cA!_>dBG52!8
zNJ^tN$A1ufr{vA*7|vSk?yG|K`+YN>;e1mXdlGuAH~B~O!8sY%&oy|GB@PL1kh*?d
z-cXwnd%tr}31-_cmx2p+l)!2t%M2F?eJj83#NgDg#o7$-Gu)?g`0Ai?SS8M}V+{F|
zT4+eD>UnY&I*a>q6^pT)b2cP<7n9w}H+z$+V#x(rs{K#K<LglyXMXOIgGau;BaKWV
zUk(^z)#Rg<sZ9{h3r3ks^yw_$r-$gppwf)<%o|K3%jkXMR{b)$#qM<*G?xJf$Co2I
z6FFXqwYfyrnHb1h_yaaYV2QFdp~_Y$AANA#b_jnj%|-ruZ4xtDdCdDoX4*lHvAmn0
znS90;5r@GJBRgsi2`Q`g;UdHr$&mFmvtK2&WK$(<m?>xIzgLry)DQSC$z(^flA52-
z8Y@9dZp|U}zw${XS{R<_*Q&{p_`OITq%Fy2%cmd8^0R@tW0++#qMMwPLh<WrE%~q5
zKG7YiDyNeJc&a6~ersHDp5>WuE;u!|G>;4#PPH-P6x3a9n;!P2q;eK|=-hsua%sPK
z6-r5dcPUrC<6mJ#DAP_Gzee}ZyVj2I&7GsXZftv<kw9mFV}&*P<rqPQL=Q`cZd`U$
zxhdiBLUlt?+thqrj!F?Oie`Pifo}Jan}<)iJT>i5uRUP9)9c4>zgq4ZzoMQ?Te>>6
z%rd%+L{Boy#OK=uUuh>Au9V&Jtq+@+uFc_Mui+nJ3@$7}&2r~#UW&>#WWBO`g{-z;
zEh)611{b%A-N>HoVzm(nlkDoQuo2ij#%r7D!yB19Y>(?tO{H2(Ybo!zxp+b;;qAPA
z?gc5KfNo<hE}QW7MnZ~HQAL0(2@RUpf@QNkj9NhC<S9w<RTY&zvHuGjI9x?bILC`8
zVN$92bSuNrOZ!bzw8%6?=x^z~6z1<$etp&aPCFu1;6hMjd2}+l8&@$Fdl*r&KmNF4
z(5&>Da5*)YbJk$w7s(@QTDP)$eBflprK0lHonfO1RA5<7_=cjY93E+3nnp85I@V_c
zU7@Q?MiJHF@{#M<N1tWLt7Va(>I%gg-8EWXNg*aSFE^E|eC?Y&w+R04IsO~14X#+O
zEjX6N%;wf{zfHfClM<(u!{bk~3V%T`m^V!0s-Y#H^a!b7zWm)%gi&2WZswq{EyC3<
z^jT4t#ckR>UTcu%yvv{NxLk8+NxRb#H!livLR;{3I`qgCU2T*-n7hoij>Kz%{=qVg
zDNmp~?3G9~Uw;DzYU40;%z0t1UJXfka8b_mhU(M?4`Jy_>3khdB1V%c0`-?Mfz2c)
z5p{^-R{`5Q#<CL%of=bt#Ym33-)&oPu#i0S$Y{Q2V<Y)~vk3HH#iwPG?h0)PFi1!s
zjuU8<SgC2JrH|m<f@Fv`n0zhfk7@H^@qyJm(D{5^{cRUE8a@;?ZgBrY#JU|1frL4~
z@rQB4V1i){n-F_hwzm2pukor>$?6ZeFLwJ}DoZyvtjObUe;HWRb<wSU^n~Lw;liFj
z%*)R!Qz<Ek(+ZodNbu*FETTAAR+f&2ruxSD&AlQCjfOF{{dz_T!h6+Y+K*J^d(eWK
z6fvuCN!X(_b=E^m%=OOR5E(8vspbnoDs<@GP$llcH*_@Ra#r)+^VYp-*`LHYrweta
z70Q`e`q;w{%+3v2{e()<jnxm%uHwtzhJ}6l%|CmNA*p0kJ}6>O@Su8SGwLhk!W%_v
zg%$Iu2DVfen}XJ9Lf!W+p~h~@!RD~7ZE<R`Ir)8#=9`mPu%mntr}_?@H|Qwc-pw_Z
zQ{{AryX#0G;u2*bgnT;llF!j^fVDa!pdP3a-H$edbM2h^{p*WS!vsojK8~)-pg!ZJ
znQ?NQW00hJ?ON`jvoGCL?ws|O(kR~Tj&Z`+7m@a=xNnK4Oa)fcJTq|tOG8iJs4~9v
zX|L|$C;zmk8PpTNpa}UvU88Uw+c<=n_*Tj|c6369+hM|CcPa^^9l4;wgeGSjb55a4
z;{4i-v}nm-dh$Xte^>Cltux?1Br9BO?EfQK`AZ@X6!OZ>HjHv_fuJ3LZ;npRZ@IZx
z{v~r~VgB!qCegC7TM$Eg0ykGB@*-;dy*M``qWD&pp<(4{^zATYGb@6Knud4{2^@O*
z&C?@)RjxI2HJxG@9+V9qS;)^S))79|pK+hn;K^mG$&{X~f;ggIk~{ah#rz0<%SD<d
z>D>cwR)z9@^b?T<YECD-8W}<6sw8^f`2y0qMOXi@h?FK=QoxxPs#5Q@ggFZ?n%)G*
z$U)>NL)6;oWS;eVwE5Uhbi(s4>!cocNwe2g<cZ?#S-3P5Fgm!2<WSUSYkW_{YtaE%
z-KOv)47!7!*!{E_Q2WfE7>(QTJ{yQ(x(|ldNo-mK1jH4DihY&Ttn{Zs4)N=?%;2z;
z@DOHxb=Gba7OME-fCKfZ;5&lg<yY&L?Lbuz4)@``HaTfrrT!Sy1U~JL()-mPqJ)0j
zehWgMV2x$UkHm$SeHADhT=0?A=1N6Ljte@NJP4wbMFJC5Z0GH!23~@{{g<=mSFf#t
zbA!%#43KWxKg7O)SdCGpGi*Yy%nh1ps?JS6={~_{>cz{<#9gHp46;~aL3*RoByG(q
zbFSw#Xc%`we1_i*mpQxIXV4VbWqw0P)nBJ!Bu~CVT`euL@eW2fo9cy!UcKRG$mEZE
zE1^s;nNa=LEM6aX^TEMH(lDm+k!2pgL}e{`gCj8)jfwWby$x9+hK`t7hY*?h^Ag_R
zC%zX4XUY-y*xs*q(dUqZ`LS+4t?vAa>--cbf})B18Y<+(S^{ZyNIsfJL*Ex_*UZQB
zYC^v2{+WJY?Qo-#PERE~Ux$r$SUo_;PrZp&eaeNOUfm|;ywoqA$TU{JC{?-Tf{BdG
zDfrr|={3Dn$>sS7-`RljL`?V3wYujAb@NSfOqgrSiu#O^ky`VYeK`Al(r~FMFJOj?
z?@YLp6Z~&pe)34oFTv`lPW1a}p_j=Pn~&bud<rMsfnSb-UTXByUtNOXhm<8_k_Ow4
zXvQRQF7>mQ?~1sC<|zk>KR3c>Q3ciDuH}HK!^4(~d;)@>7hk?BUiXAB_@z_;<8l~h
z`KZ=ZTD&VcmxgJl-$JvL$SJ#>@71|6$t2_O%!Dz56<$Qp>ukt7lsnWikc+r_;CfX;
zAEsY)C!)FiX-f0-dQJB91=h(f4ni_VK{w57M~B>8N(xqTjHIE}+Shl04tMk!J8^$b
zqc02FuFH{6)vTeC3j_B{?0kMcsjx^WZpG}{EV*^}AugCISi5d~P#7ueFHTt&+d#Fv
z?Rj8i^6r$rpU078sVhNzc`lXv<SvrBpI>Aads~z=g1MQ?z%Z~^nL#HwoWDJBdkN<{
zjG4&K%epf`zw_z*zLEs7@`VWvLH+Ulhc0Gw7pQ6;o<Nm0J+H4)^2IKhs$}eH&Ds+h
zv_9_W;&#_IjL_m-^ZVa&CWyaWS;D^OM4!%>P$czP%Mo$z6hywQ9+~2oYR%amL%M>3
z_44`}u!uzV{JI72xU9UNxTvl+UwPf;6Z;lQd9We-`RUv8BO^6G<SH>r`l9yOkB8<X
zL7%q$d2^9N_sWo8O<qE_6HdTtgeh6MEx>Bvnfy52y@a$9nek~e7lHL@%eHP&xtzK^
zKHnR#RCo?pE(p7<Zlr+;-b6gLm9;j@dl&hrybwg#agSaZZ*c`=?9Ep&u#rQFwoOCc
zpYKreKbyi%SX2Fm3G!Dl$P$0m-=fmex3mB?ckRkZ<#wMqRu&~Pb045XbS0UYp*EAn
z6@yNcR|EV}3nM4##4`5Pi+Y@VJ&3w4Yip+QA5oHCm(0{x`(jdVE;;O#{LVbtw(j&k
zm31FT%_X;9t3?hEx{P64ZDpCVU7A#AA%%%L@`h405RXP<-Z_MBLAES;a#RxRtoK=v
zjmF}l2xLlpf%Q~M6*|Tx6Tm)%c0pOAroj1Oac)2%vz#H66@8VF9M&D+duG8s^0j;h
zo*^Y_Gs^qEC%mYU2C36~pi0s`947FFj)36J>1}bs)z=MgGTuH79DJ=2l-yc5l*$<S
zVVet<uS9lc4PW(5)vL&jMb$1pw5xY5>{asUvsdk^GOFLA`?!7BvvM)ZNFiHk3CW+(
zh@KM}`x@KZC@$Eu!|u~uvZyzXv<Gqg-G~m>oOcDn25A+7z?S=Ct4?)c`6Ykha_8J*
z{nVjo50tRyc96&d(l*zZZ?8KX5D?oQbqp?I*IX`JX!OE}vh&*T_@APFxLry@hUym#
zNZ?gpDxFr8I<fe92|Z2)qPSD`<$XdeW}%FH-@Pa|tTB0E-#)KW|I$h+hK|2)J&I5u
z>ic_Bjn9L^@x9jBOfFPxdr1k>u+~`TAv3kx<EV<{=NBnuHQ)Bjtg62S)b3AQHYKO4
zWBD<q2_|l<hxa+MC+qnvhLTrdHKY>1Q9gmFTNatvAydxfXJe1vDDPh0AGM^rv5;tY
zMtP~+1l9W4>ZH8t<8-df^2fLt*(m9^W>5Jg^iSEl!U!@aQ12#%qm|H0lra%Qz9>%c
zS2u~*1u+B*s&Pq449ESXe<D1@Zxpd>QN=osLfYLPXlzVM%tRU-6dI>}niBbnk80b_
z^&6LbXxAA%$ot7<E=}cp<U@Mb%w~3p*+hV3;K;2vMf9{vYSV?x!*bhX>OkLGCxUHL
z%kBisAx+5pYwv@qahiFN33JkP&p6V^!^lObWt|q|M6-~~sW0El-E3>Oy!=wy$Nlcv
z)ybO9w%$B{q?#)Yu^s(LtH-yK%5@Pb80gFC%h1*;#GPRr8QpCm+$Vt!Ekcv|RVz>N
z=#ndJr~;n1atJBvC4YgSyH6d(n^%5meh2(hS7;BBE%kXv#U~VRb-18We4$Wmedxx0
z)yxci1lAr-HJGU-@st_D!7OzKsZK%es@IRIjj(jW)UY<21jjLKu9`h|v2?Y2x3lun
zJ$`SV?%txHTO9Jseqx>djQ66D+kK!tZaVZCiiqO#f1D)E@$bYVGZzcTe+#G?(vpq>
zK+T_<%_Gxw5{EDv7#bYv!uFQ#F8OE7QzUMU>fBxP=8DB}#XP8l5t@|ejd3Cm$LcKl
z>^omOxVbO;aA_X+havgzF15M}A+!1T=cg{YKni)lFWz4t4OC5cPq<{k*VHDI3HgcN
zx!zO4M^k<Dg*?c7@h)F*%q4ewbRzzm#zxrQ8OQTXDP#6xP$h*yK2G4G-WfTRwf@oW
zIMX#nb(H4z!R#rfb!o}X?>P6}l6IcRf$ww60P?7R<(Tt|YGup)soAy5b@==A6xtWb
z(q6|p#tfq-B4ii2*1o`0$V;Q4#aRMKFQXY`rMSk2qu0Km_sED)IGRYY@kgePwk<sM
z?OP8%6!Tq1Bq)m5AqA01Jl>p^zHt-qD@kC2LBAvWz^;XTh50KNZBFIwJ8Kl$8qp6W
z(UW(*DRjw%J9re!1Jdwv-{2iBB2_oKI>(V-<z!U+2*0>~k?TT(VRn0#di0cKmKi!d
zUg`(8dEGAxkgitKnTC&eeI6`jip0rIZH=tIc$g1fty(K>u+@p^P~&)rS6}hbd3ZUZ
zU0b_<WEn=QS2nQZfOPbFYZe^m1f!dvYk>YkbEoKTkwtHmAUiA+xwO9&Mg5Mr-P8G+
zE8F>|Qij1*u>rkFSL%-8MVC{gOzIU%2Dda~Lr9fY!*&01b>PiAYCKk#J{xZ1<||Ej
z{_T3w?XDzhCy+8EF#CAKEUqY7IL7tqN~retqWw_0sjWL|*f!>4$yAifAKa8W&XkI5
z5HndPP)>m6=bL(?yt3h&A8@B8x$6a`tVj1=toOy<8^sWkvW)y&!t&4*-^%aq@!z4y
zgJl#OAW4Jn9@@Nd3sLs*!i$wQQN~5|^GsV-zRNOe?3WSp#n~%idk;p$*H1OtY5s6;
zFx32QFxI{x(#oZlYUb-`AX&bqK$j4fTA>?V2v@Uk&G2fFS>-o1dwG}#Aups!tJ^wW
zLc`ry<NBRyQ5(U_eKpc=Zj3)dXq=O1BA|-v&4VG~9+Q}GulJVmFu%iCC|b9j48o@H
z!iq5mtPPY2B`0Zxw_}XTB0P3fb`hrD$$bkrB*?I+(q7<zR_iQHEUb@a&kG?~5_OG0
z<M6D8cJedSjY7PRc53|UzAh7G+-f`6{J04tmqL(NUu;+=Om;>S^<#v+&?ioht*x2O
zvrSU9mWvuPBv@TbqqwVv^OI*YGolWHq*ldDUhVlAzTueSZ901@%6A$sclWRM^VG6a
zKBKvsg{j)US$e1uiQls~%O}5EY&pfuOWB>)4XhScj0%$|Ptx_THoveltv!iFK<y?=
zk4+Y+_7#3YDfDU!v~-5Qzw%2T{Ho;%O?=c(gOliGDV*9>8W1;(uk$oISl=KLA($4U
zclNHRIjqs)b%y)t0;=~=Jfcf9+sz1>w@zacdQh*96D()c533I4cy%>tc!(u_QtiX|
zoK+Qw7KXCYAx((fWd<*t`z-|%OU|~Z!IuRIo5~JfW9(ck%rV!=wO*o@y-tFxpZ35!
zYeh^_TY^WOK(ixYdGR=$6wP9b$>y}_CpKBFQ~L_<lRL4k)?OwG*G7N%r#BHS-74L%
zw$t|3?@yh*=A4B+%*gModl)o4F>Z3+a<~bMNEZtYYAwXWeP47`u;|DakRT}~!?c8u
zS#mbG$FHlA4{r(b(fVagR26f=Fem=DbhbzF{(DaZ5d}eSVQq8WTz7&E$)|^}Hg^<q
zZ!MR4TiLMfA(IYoOSHdUhm*Hrz>nXEH+#t1YrA7<{}5g3@TWJ*bV|@R#HB@fIEci+
zOOoge&HV7qc=2b$0u24xJOuO3jFOX+??PLf!z@XBq_Gaa;YSqJ{&>BUR^vls1#^dJ
zKU~g|86&+D!D01?9)9RG<BkKmjVDk2=)#ZLQEpeHQ#sK6FlU<bwV$}p@)2JfWW#ux
z2EpRS;J$w0&_G-mUzoS_gIF`nFBP9&7`OO<01xJI?^W&;B9sdA(5zmCSptS0L05Q(
zPY?T;$33w>50=?rd?RQ&%gT^Jt(pq`-fcWg@2WUksJ&7ni+(keP`|93ZaX871XcMM
zJ@Ynmv31c_DH4uA;<W?0wirfrBHG?UgAyA13fYYKa7{WKOu0$`a=Z4Lx6#V|n*d(f
z&w8Mr1M28^AT2#NhR8Y6DM8nd#ula?Pl5w8Im77n{L3MjVx1zV9)PH=Hr{b$x#u+1
zOqh4?&W(Il#>gx|NZq)$?sBvv<0|{0c61MYSyPCit1WF`k#Irn`YjEiyD!8FvTo3z
ziHQE|jMA(^gw|fweOp@V+3?h67aa_?<sm;aPhg7z?8q>yKW+`e)PA6!#F?R&sMOK0
zZFz}=fZ7&fK11q$LTW=VlyYKTFD)e}y;1V6nOYftw|~02Y4z{UZ9mnoH``C;lD@6Y
z$4@V>p|1@81J=OJ$?-q1hJS9s%*?^g`QM7SHJZ|~3mgc4#oHue2mv!!K4Y($<gIB}
zyW}(wZ;zZ~q@R7;>!{{prVEiPO#~pkAzppW!f!!zys77HJ`w;RjT#psX2?uAlzqsC
z{@35;7Y2}l#MpYv!6WiY*{j>XqGDn~i3XOxEZ_a0wC8(qL)2zyhY@&+<vaWv*+M5x
zy+;f#g=izg_<;C1PF&I#A&QM-Y?H%a*Z@giPw>zrSYe5={Ac0LMx;Vau;7yGvRj9L
z)Z=A9>cGY^Z!iX)s;|>_i2jlg;>h!XZ|VBH`f3g1;1#JY-lSCb4=q_m+FK|X7izV5
zQ%$yFsS3Kin1zh<)T!@ZLiiiAD2DpgOVBp!y9P!(!c<BWFuGxJDWc$<d^lMy6t;}I
z`J)S@&#Md9&6?tD_`-glMhXw}$eR)}!j8))S2Da>`>mW2kSw;I#bd$Tm%sQ*d8D}a
zdxsn6%;qU4)25!T*LR6|ckxB>)Clp@G?NCQPId3lm$c}Q6jDuW{`bQV!nB4u+NJ^)
zv}cF|qA3!`5FFK*5iHUv^c0MSCKnOWDj3h?mgM%&_J(RRO%oQYX{l1A0l$0r&V%*6
zXiM^U2RE>UX|zjxcIneSvvuts%5Su6`}EMGGg;3*%1)!UlqQAN4k7xk@3%$|vAsxV
zI+bpO|3p!exnUgnA}XLlJg%Z^dx}~6E{#eom!?LPO}NNus<Uv2<|(nnT718#j50lX
zmv5itZ1Y^BPO%)}<%SE9G8Nr)lCqlLb~~igS%vX>wp@*i1A*8O-F>2&WWN9VYaaSW
zEGZNjN0th64SFwjf2{GwV(XCs>#^6eT<IQGh&gCqbRu_+8H4xeyQ;Fzhi3IJ4=Y1b
zd7RMZ&TfW1gG{lM;l+2{+cdI+>KnhS!oFyiQ?v`k*pd90JVa+MwGp@-w-dxkyc;~z
z(<dl*w`^uAB=-xAdDCrp2ihMW$slk$@P55~)v9-$f9lkisw2MJe6fK+LM$65hHCUD
zLz@f5x?Ph+AhZN?-`)B80+YpUsSUU*+-n2-evA)wuthoau=?;5V;5)}gH4@JVml|K
ztjA?R9)9)i9;#5oaH39#H!1nuEQ{l$Qsi;L4TOGL0?-K4O{Q;m>LTzE?3KMI@n4%t
zL0sA*MP_3Oyf1j&wOvdhru4#M3k6!Ryai<)U1x$w{JU)aR0__JGwaA;lbjrNBIV2m
zrAcXG=TdJ{;^k2KfESyz33RkJ7&e;yoy&-tw`%INs8qrO($SRG!&eZ^CMY=>X4UsC
zaoGvCIb7g7$6^Zym|mVG!tp`TjT#}{ZQu3eLmze%sBO)wO3hwVmz*ndQ!F&ZAK?!=
zeC<Qy9}N$C=uU3K@t-zY<oOab4*b@8>-xoz{E%CthF$A+DkGxHw%%7tXdvN#P8%`7
zdo3%_;m!FuGUjw?L0gMrz+!Npda<ji$0CFC>U+8n6#dGk_*Ue4-e#NQ!E3zu>a4!&
zp0PN_8xP5;PnEJTgs>b(#9DWDUwM}c`<w7C!#L({vDcnUWf|$VomA9LXj1Z{R7>0j
zs{P|7eh3o=vIiQ9ViYfhUfL$B{UCwsBi>`y@bD6nQ!Ozio&NbjtGz!Qt5@I^?TH+c
zOV=KUQL}bFMu`&KiqE9+4=J?%YN)K`KkrSb2zFx2v?L0^;5*|ORx2s6(Kt@Dcv&p_
zdE2kMh*WGvq(rT3z_J?kc+s(Rg`Jw652k#x3c`-Ik*?+M$EK+H!DCk)3b1a*Nsam?
z81*zCLVMa&<*>sEmLk9H`8KGXDD|U7wM(mM>>PUi4!yQ*m_G4WF!9}TKe0tyKzjd)
zj)8JXn=*U1k2Xc@XZ!0z8x8$)cm^UV#l|Vhvx=3t{lp?QBI(lz%S(;3Pn|ER1K|up
z<%v$2c&#F*BI5U@Nk^yM#T|7NB=TUjnpvc;qfY~!tj!YD@a^JT?One(Gt7i@wiVbe
zZC$>bx<_bqiJ*x$$UOcvu6{ske7S>3d^z1;$>;{ReM;ZG<NGR~ebYt(*(-=+x&b*$
zhGF3s{AUk~xgNM>!i;&0t=u03{WBV6`~910G?w>sHL@@|@v4Q9<+5m*UHhR2WMZ%1
zvAf2&6PLyaGxLh5kJlX5ie@VFOEWB!2R{05=4WUsk@AyspeEyCJ;m-#w0dj4+)Ode
zZ*}b(-5s&1uNQ7|DSjg2<Mf>TPhpJXe}&BdKb$Jd$;9<{<G}x>5BASgS#<U5dMuS+
zOH^S&Z^N=benj*ViEK_xnvKLgKxN0>H~T#M<Ck8=wg0RavOi+B7$gQSi9aOu81nOW
zkN2=&6}oJ8bkMbrySwS}ZyZxbvk|Wmpn_kh`&Tb7PUQ2`bt@RD#^ZLRBP1jQ<m6#K
zCaTLVK0yDIbXB&twwAj9Ms7S0tsgJ`yauA98^IS4ZTg#F2m{beZ~{?Q6DjD`2Gd1=
zR<KA;o`{G@lh>f_W3}7O*{1aF*2c!g7&1ExizAqx2h^zYEmBX}@$neXvJenA($Fto
z#?xskkr^2pGCS{#dYo^`0&O?w!X7s=G09cF^T~qMZ}!~W+naCodyew02FLUj${Y+c
zIa3Aw;``%y@Xre1bL`)1)YjG-C4jc)-86VzAo~GReSo^e_jWU}b+27s9G4IpysTIz
zl|!~#sm1#?Ffee#N<0#m+^SfmFeV$;Zn3e(c~=RH+5jV9!FoJTEu8@U>C@BGTFJ(J
z6F4r=N)IGnKz(z4bGnWmC3Qt{hJiW)1Y%%d6p+10-Gjxbvss{kRcQ8fd3FP{LxeQ{
z3=C{*6RIgFfrE#aZgJYp^j3ZU{#OFM_QnXkPQAjk@5+v-sA%8EUJ9VA06q1O<*iRe
z&1y(ubMZ19wRXz_R?{V*?_I#}iSO1I3^M|=Ix0LbEpae`4gd`4n|uPRv;ijQY>xC1
zfhihbRG!S)H=wOcD~aE(H17HGX9)0HSO^IyX=&*(DNAr`^XZ{z(hz5qt{r##joxU`
z!@hU_d8$bDj|!sEFzgsP>_rNvUD;SiZ!gtt4cl2AuoX7H?!`r2wMasKt=I_Al240g
z*(zA?b#3bbrpoqg$QJ>@Q>9eAu949im?s3JF@?<|;Gxer6NEgcrb|s8E=XHHk6V;^
zd)p<#Uo}falUTyey;mt~&IGE7?f!UaFG{80>%)a>Fs*3xh@yu&VKAMy&Q+fKfL6Wi
zyAPQ22PQz3FSdg@1i}7X;B2D78B%GgSI28*1|#B0fnh6K=3b*69VC8F4`2cm93mpd
zw`wqh3fBsB^Q#5*@dF**{>~0n<l#b{Adr;UwLXGT1!X|VATi+o5`NNlrd&T!g)!L0
z$?2#4Y6o~#-Z{_$y#HwzyVMVqcX`%IU^D{h@mGO9{+qi?E1-HbHeLsEIdegRxqt(m
zJ^h*wKvPlq4ga$YOzua1r!5&EjL6Fax}@xk$>7btSjtML4=5WAzkj;PbEL6+2&<jO
zK<?LYIh<ETAotlC$z(ha1DY3$#I(vO&Sh`XnpA9ABm<z=LPA2STm^FPn&&kzB_Xa^
z5|M$u20e9c0tiofsi+BZEfdzG3xgu!*`yq3Ue;9iW}aVOlFvgz#DMV<ijt*ZkILta
z0F6w|GKvZxl{pqW2?4nam~~jN^+!=4Ee#B~#_j21cf$MjOmYYf3#*{IzOhlNF4Omp
zx@KK+{ZTl5Ula(FXz0VS7^jbNH0*#NreDOtJb_Nr1UmO@>S!8Bdv}4UwoD=!gR>qN
zhEDGur6w%VZ*8fS$70Z-iUcwjIpa<+FaswROyns%hjU27FNjdZEs%*f)&p9)GrtNM
z(@<B4kCQPTV4cTzazaJmFl~^ff@eHL6sRXuhRCn8;ta-P_#&_OrljYA4y@2zF^pYN
zD9ldn3CLcdp(Vn>&pe4h3=+_usV*nRrAXhoySwik>;dtWg+7uqn9sF)vYDt4grNn8
z!Z^%_V$8lkR4Zy}(dsl)q)I!dM(>hmIw=5^hEMI4SH`H6D{-c(v?nD=b;$*Z43X6d
zr1?wt$v=-<pY`s(XMR-Ro2~pHoo)*TFpalli}eJMz@w&_tSk0?S1nGs(5W(wbmsNG
z`9U{Rs7x8Tert~m*PlEKRFX8vKx|Bb4E`$1Nkt;y*}+%M6!6LYBbh5C2;M0HRxmC;
zNtzBUl2KGAj8KJfU9nKFxv^Tm$z8FgDlt(VD4KHV;)%z7-Q@YVsl};~fxJ3#{xprt
z;ho1_pST8hL{`?@7laT(AwaAx9US+&I3q!(lquTt{X#&@txKb1ZKA8T_5wOTHYO$o
za^5SNuW<htcW)h4<=VXqqo^1N3P`gMr9nYT8c}JaB?Re4QaV*iK<RGj5J9>T0qKwu
z>5!IA$!|WW`+a}kd(L-$W1MsT`S#eu?bvIr=UMB1?t9K_UUSZC%1Mcrz^0cd0ZZb>
zhz-4w#tBtF`*UJCh@M)0uv!xL6$w+@8RP{cft+J={_C9|9>s7B_RF6Mt<+TTQb>8v
z8=57LijImBGvWn#)p;>a9v+D)5R?ezho|2^O34KsDRLip#7#YHKIolAdr>0xZGeDL
zS!R6}G&$w69Ye9?%nY6Kg#7Ti`3?^cQ>(xQ&^fL`Yu7q6Ja29Tc^givAFp8+LF4-P
z#%!n1Wq7L#cnh$2%fn^U^M-z6Wv=XRcTSD@tzWnbeM%rMLMN@_;_r*GGFFpYAO?c4
zDI=Yo(%}#8-%pXd{WvSMV~f{!R{`60p;rtA-Dn&|b>lAg&|cJdTix;LC?chY&eOo6
z{{H^pr6NA)u*oc@WmGMe90XarxFn3T(~L>mLPyQe3q(LSmm%(*n3&j$<LWzWuUcYu
zYY94VK})cR5tEct#wo|@s{UX^h4Xc2(@T|z=;uezw6}nhH#njz9w^}EN{0%oz^Ao|
z?H~z)ekSjA*-L%yEsupkYE?{U4q97NmnEih9;2a&c1$$}gt0nlpCP*z&+qc&S7G$F
z&%-UpukbL8W=*GA2)Bw8n)zt2g{=F>_}@Fbv_Bw<G0$Ptozjb?On0jxi+oN{h)u)o
z&}O9ErjK4ys7zg<4SvqZj|QuC>CANdHu~GQZ?DtzvIhEMfPu{6%Y(+W9@p8mpWh3b
z7ZI&z2KV=lpf$(_t=@7nW92psMGxfotJV{sX&5~K9}Nx9pIqSf1;yq8Fx|YClOCm0
zL*_NmZ<@{K5ALI)V6qm6O1z00TGltfy-LS%r^eQsfSzphYKW7JIIj>c_#)K`+nk!3
zV^E@cm{#Sug>%^`!b$EnT|_*J$5t}dYW&kFJG&id|I2j2o3&l+e-xV!o_k9>7>|$E
zAO9-}35i5m%Hc=BYt#PR(B=j`kD>2MZI_};N_IRh;OZ*DEjgf<4%#-MuZtfxRjHW2
zR#&26OPoWmpbRgqP9LZr&RZPr*+gOLcE*o1HHlWeEF%>6<<yH!RSvg}3KYfAQ|)vA
zEG=+Id23y5rX$Yj*Sa3jC+MJKGacLn-9K?}Sk>l&Kz?XK{QgnLn>Pp=<5~n1Viy+{
z#ADIXQUswrY`vDxl=0TBB0qjpY%D@%O=wxx#xHsfzqoL0LVV4YC3s_hb3S?Q@xI@!
zLZ)#AtTee)aqb-@A|BiKgMxVDb8{J^?bHqX&4$Ub`+Y|PiC7?Jr-YgG*4_O9v4JHh
zNWOn=LNLur486Xx`?R-l4X!b%t1aE^9;>j+-Tv@!6WW61d0yv7FsgJ{IoXk6S3lx@
zlp~3wQ2*}KsZ++@fq_?|-p-MRJ-cB>)pQDNv=y|5iovBpb7<zbgM$M(=SJjw20shy
zJn!M<CEf|r2x1?Td$?}gxWAuIWQr(F07#h56+H}_%z373KA8x8mZD`eLMpx`6rda!
zh#u%JyD~smq`Te|<LmYySad31nNBHeK`fT`!?%7FW9X>cn{4ody>``9>@e)$m1Ki^
zO9zN1GL3-4L0!8R=Bxx-i^dq!u4XLabSMi@3I;Wx)bxmm2vQ5*m2>AXFwBQbWrdY>
zU5*YM;`R$pp?#`(vF`Jx*mOvLMo@+idh(*Y0VMHVzkaVwD19&U0n(-y<pEvEwB{1f
z8!a`P7Kb!HvhYHuuWFuVE#RMD*N5O+ftZDq=wA^1AOHQ-sJH#|J2d@PpvWu+$GR>a
z$KtIJT*PFeI}K9~Bk1I3VAw)WGI~Fg$)XOH!`8yJsnz+w2`br;8=Krhx|OQ=Q@<5d
zPNOKBPGnuq+pn%zLK``B)II}+w=pgw$-pU)3`jZ88!3fIX=*~_rfUM|dF5WP`mK~w
z+cqV5bqt!don2fE+9HFzFhG7YrtmmTxy-krRmpg)3K@8C-U<XcDB$=loDZNo&NMby
zR36N#mY6==-<}Y3*-UH!^eeY9E+in(!_sjsS;{#4_?=3)Ca1NzR?J1{f;~|uX(~j8
zK2!OYy1u)IRf|r-PuM+>YZr%))h}>pa2nMXh_cR*)nHvJISxtO+o;PFHmyF<P`RJB
zHNIzKg7J-!>70Vw@@2Ch3i<V5-q$qGZ6v;l#B+BG*6fxJ%J2AllY{M;+-+N{uiJ?S
znAxT)Q&a93tdSUFl@2<_s&rJ?+$Y*(6g`f4O0!t|HCbs>wG%c!p>6an@;1-zYbJ*J
zBgY{aH0qiCAO(a-tg_@_MBG+qOZyP#kMZ<b^{#is)ydk;`C4>uEP}$^oG4p~utXnI
zK&niz+OSYPj?A}je~|l??-#?vaITW?)0S#P{)3Udc_H6V?rvZfzCWM+nf>i9gVOim
zV!5z$?ciF%C)`O-<o}IuuU?(#uSj<<(wpwq;EedFp1*2Pbsm#9GPc@fANpupr0U%-
z6SD<Q_m0kjdutc@_2;%XQ_mlEUqnfEKyP!}aTJ)>D?d)eBib_PPGjSo{2l({`k`id
zYHDhJ9wY^1m$osmQSX&MY6H^YCP+}Lw6}y!2?`LR`^5==ly(A-_&|wk05%Eqhe}5>
z*LHN6fx<@CPY_m_otfd#ZzBY_`|(|OHN+hboW{MhAPoB*k=;iYZ*$OH4jtTn!<R2D
zR?5oC(EJ-35QRz1mH{fG;t3GC3=R#|JSxA@(A_OprcOmo{iU~e6MDJ-+&!@dr3RuQ
zc<u{z;1}#29GX7^7LDOB&0+)nS~jEZhxn(_(LrLQ@4?b=*#;otKgvadZhZF8ln;0k
zdR$xAA@(ZJYXwlurs5wEfG8fu#XSYL11j>g*(qzF*8xaaspRlze+7DSV+P|v)D1e$
zxu?lMq10b`%){Hev5w{@2tL94-qfW(@KQP#WqDm_&fGU}ETO?2bY7u{-qON?kVW@(
za&mG}QIUdMcQu$d28L{;jhoXV5n|7leJ^nQw?VqTd?`<>Zf{Fsi{mW#v1Mqz-RA-6
zPjNW1Uy$Q0YK~G+u!ojx@PZOwY0i#=AXVx|h?*`D+*aToF15TV5zc582!{B}moE!?
zN#IDLn6*yN3ccJ&iY1t?q@7!Fkd1qu0yf<5&6`j?C8clBI#0U<2wZ-XMij&p#s!V=
zZbiq$w9@~PP3khiww>$Jf`<#O;LQmF#}Lbg37ea*!1=}kDNtBWD)G?W32*YyZ7}86
z=;U<ad%z<+T5o0Q8MyUZ_UihaCq`#17kJfJ2#U{QV3?EuRp10dKj2fLH@?qI0HQFE
zG37wmgFAn$*eiAWRb1$BM)avbMEp7ioFwE)z5*o_*hnuhz+^RXo+zLMUT-`RoI+ka
z*<~j$!iVrq!3UAo#IAzP6NQ!h<PLv%jeHh)8PSIn^oM%{-++r!HX+xSI|Elg!v$Ym
z6i(+qp9gF5&xx7*eIH68xch(Sq9|Qb(W`$hfPET{++=MV=?~C#fLV&Gu+ae8j{?Yp
zeFRPB$S*qij;vyV<a~~7@K8Z(*vifhs1!SYQZ5hyxM?N%aRZQu_HSKPWMMSDF06x2
z-AI*_l1r)Giuzk&IMT^IroZT!{8;>F!boDcdg)i>8aq3&VL?xXw)A;Hu%*G>5C2xT
zS?ryiS!bJaWBiAP;m2!k0`lgO1BzIfn8<-WS#F{?pv3YT18{cHRC{HiyT0Y-F=fWV
zfknAdqnYE9V5S2G7U=+N3s5t}Lh<f`0EGROlOL_We%<Y0_ai7kgF7rDfuE$i+<hR8
z6PloRfvQER+vEfxJgu<R`SRu$NboW!WN7N@T7%WWa}pL20R;uH94Bk`hU^Ep?v9QQ
z5X;C<9637p^_(?*uN<Ye=_DpXP_>=N0fGnszCnPr-v}EM)B63ivEsX8hoW%()Sqeg
zZ5_<Z|2%%O3ySZRzt+|YzK?J=HJV)@4!B~9t-n}c+|@+R`Ub-jRB}py4yc$0K>~ff
zShkQ4M=L5!{>^<q_g~)pf+&W;^VH9~?glUp@aMF~D)3!+Jh6OsqlIrmYMIF`U+TM-
zT7Gj#rB`_C>)UOI{iUJ7)5Alp!qya?Ac&lB+Ww5(9^`lTBEcN~HCs@-M$I7wnh}53
zo4N^f3E-VDCS?QIBewF98n@%@QL}!jzNM^h<U9_JAOVn$weAL0{GAEkD1+Ed6s4u5
z<BG7@pRzwsSctvCVbm+Kf8)lVqwHb;+(5K0oU)k$j9)e5wGb|XXvZWX@_~P=C7!QB
z!E?|go8wvLa}H(KHe<VL*F)~#=NH*q09{aG!ZI>4`m2c?fa-25WCYr_zAw9EWgELr
zgYC#q_oSarXAb@EN$PP3JZ?pi0`3e7k_D78h(k`kqMYoyyp79Nuxx_gNQfz*tF8Kd
z_EYVfJD>j^#rP;>77oHg^f`hz0KUs2g!++nd6nzo;Kpo2*uvdg(`r@nxhpUYgSGAa
z)StuZpuas1#{t5gOL(-JpmeRLr>ChYlk=pgtE;Px_1`Z~3uYU*3?{fT1T;ibI5Pj|
zj`_cA2cr4**SM@*5BzIfP-p%c7Z8K~Yh3;u6U4au-sgyMnH0SK*SP$#24Jy3`1!wc
zQRKk>z7ZnAcXD$?E$a7Y|FJG7cD3oRg4A!5@W-(J@t4RuPPRGXw|~!Ufy?}{$A7&!
zaxcXG{OdxZ$EX9)jv6X5nH?YR0<#A0Eo&)&T;K?_x0aTcfSxP^u{3s^;o;$wk1&=3
z9`ql*z%y0bNTB}c(IeQ>APL3oxH$*fY;;R-F5p5aAV>jJA#(9Qny#i+A+jnlcWZ1G
z!$Bmetjqy4<o-IGTnsR&&kPLGQNk}zpRISy=LL=t&h(<S|C={#92|h0ir2fEn~h)-
z{0morgtVN=y?4Op!I22|1ZNM&MIV%=AXp~5Yo;bC`G!rs(*6#3E8|(X1XyF7zo_P$
zJBVl@#~K8v8IcGg`Kx3sr)>CBHMRAb_6p$2z)ge9R4H7mdv0oKqk%vpW61Z66A17O
z#&7&+mM=8?$~5{FWNzj*+adN#nh03~K$Vxr{AHqtqK?76jWh7Z=3SwbHQ}LjeK^b@
z=@l%V2NGWpQ7J7T5>nuJrRJ~T;pu8>UT<`mX(N)Cm$!Lv_ihO=QEA0+*n)0~+#kJ)
zLUrf{I0jlKqEb*En~K=Rxa8za(L64Dg^Um=)qqN>_&z*_RH@LiE;#D>y3H^a+7{pf
zcyCK}ypvs$_zhKQw-SF4jt=lf6omc5BDw*!fU+a740{~&(L}1Ms;Nk1N-Y_A{gToL
zwPp~NX}bD%xdz=qZzqg}8Kfg226jdkf!}690T&mSfq{XChv#E_w>?lbmfA2B;36Jv
z9!LtZ+5S<hJnuLmNO9xFB<zLep{dD9aAP4Q0#4hE@C24z{%qP;)%L6MFaB=V|DVB!
z|6LP-fTI6TyF-hBu9Wk9TjMa&4s?0STn`<mCm^5!cXR@o5W~V!i2=A)wa7SCED3a&
z#B~teNRDSG=nAnnK@)rfIjVQ%<$D0`nhg~{+8PI18$dTcF0MNQ!A{=vl8+FEF+qq5
z`+#;?o^KaK+v%`DSIXz#-U4UvL7&|q%6V<lvvB{{uZh1ap!?IYtEtHlFf7Q}fF}6M
zmxA{8oF`BYjNrp_+lgHtp7M~8!~oX@6Zhy)h~y5!SG$hVAiCGc?>v5m0jNZw+<G45
zx1Ok~0v*HynwI&Vo<bHD%>P`=YbSAiEb5MqFwm)_q@>3E@3|l`(CH+M_h(ZdkzuzS
zrvj~=u;WPh$L;Q_p)fC9>M>-iIat4hTxwiH8-C3C;QVM64}3!pbth(!f9&oC?lJ5H
zY?Av+l7R&T3lyD%F|7w@db#?6NJ{`?6A%q4udNpaYa1Iy6zq3bwGi-Gi%r4L580`8
zh)Rk}N~D)-K{`qCeGjchh+y;zc8hd1HF})M-Cn-|z&@`f7i&QZ44{fg@>j@ksH&)h
z^`3-+%v=D!GL>@U6&&as^Y-l?<Zr9ws2q9kh2(T*#hX1qo_BW+#>mLXej36rtR-TB
z-ENOyN?6ZK!sAmONPDHPQdx{Ktre6ng>0?)e)oBN-t+VA?tn8uh%%n*dD=R&5O#E!
z%~8mi5ZAUA&Z*x{?)=_zEexgrJSO`r(hEZ(G&e<saEk`}_M2eglV3bt3p#1J${#1e
zfag~Ni+3&a1mI^Ni^N6;8qOKpYpbydv#l?Wk9+6xcX{8A=y&WBfjhnN)WHJa+4^K0
zhiNn;j3F_SJB8qQnHY$MNxyz>X~}}?<zFEq%h?u_lq5ZxI@7a$+hOQ{RzQFH>2RY|
zuU~)^`1OZr>zdBa&X6Af)iOqrhYvl{PgX&bIK^IFVFL0Z`yuDyIY(Umf~@kiU=<qn
zeeNAi7}f>$FHlRE5P%WT26~fwpzXIc0c6tf$Os_JSI8!h%>46gE?%R5LoT4Ma`bC9
zft~mf_?I6$1_^V2Oxox4ZB0=8efhE$Sk*9ZK)T4xf~s38`INxdwUE3%*+0nQ6ZN@+
zDr#`|3P^V6J1_R%QsK8>O<X??o`!>X5+xx3Lst`9{D<EPFg<$T>+0$teMe0#yfqGH
z^<PIk55p6(m>h;(nsCGv6&1mIJtjUpIH-3|*a<fOyU7W7X&W>_S**bk6ucn-6`ulW
zF^MqxEr$fWKQ8&;5=w!?4Q81*etS3p%5Vq^^w$hU)w<8)@}A$DgM3|mpg{8BFEy|Y
z6fTYyKr{mZ$S9L|52D2|d~8tJl$Ji(_BG=0BsyFiVJO;5#MZ>~mFl+>J7fYOvOV|p
z!B!#=dQnl|A+%aWLcC)y3^*6@wtR$FuddbYHruG$LTYo&Y^4fd*XP-FgbDNXYyjy{
z-CuA#K(pN4eY%L0nAjJ2K)ZK&=a!b@gcB~+9B8Y5xnq*h7j0By<jSk@smLz5KaPK;
z5ckgmPNUEL4o7jpo*5Q%5LW8%?cE?uhUc~doCIu_W|{<z0i}NX;jc#@Mu%?=N93rv
z1ByD>N4<IdI`G?1@TJzgyD({0d>_&*C(q+J3*#~-=KGu+GE!2o7JAQ~RqD4VoPnWo
zw+f<{jy65F)>)Sn+o??vfm63J_xq3-E4Nvse%^ax_(G?FRl;Xb{JG)kK#Mx#vN;ZG
z6H0RJqO7be82OIyTSEY>da*zVk@{^--`4oS>U!S-qKgY;#^~rMNTvbHM95|+1Jn@U
zGmUJ;TXg3jleBa4kb(8Xg6TV<N+GslyORx<;24#oNL}Z$`OyI|cyvw*YJqzIlHf7Z
zMtGJ5Xfq}picJVMp`k#YG{j*BIGiF9p6$7qs#eDKe2-%>i+&f}A^L=!;%)L)cZmC-
z4qyz%3j$er;fb6F%0Y&LjjeZV451|4rmPJiM-CA22UVapD)2&{tpw~2YuC-Wt0)i5
zm0EBuO~Z1%i5p3C?3$5AW%QoWSj^e`VX052bsD9ew;vqNqGDlmARIWrU;5hGH#!|b
zd9nT$-TfNg-I6FU9uU6*nE9~-@ok|J6(!}s&`>MGeN!&Cqe_6DkSY2R2FYPO-t$Cl
z?kE&W!(}UsPz&fJ5MgWrS{rbt{H9+U34m7`!K3d2ch+QiX>(mNooCykZszWlh|ze!
zj<eZoVttyg{n))<qm0d%UHfMWo&oTfM|#-&t_SvzFNXW%=jZQD`Z1}gsd-_Oqi}H6
zC!C;&!WevXm^86Cmr9<d;B-t3-j@LKV}Ri%PI<6b0L?{sLI%9aYw#QPII`%x0Jb1U
ztGj0aY3YRp`htg<N1NIn$37T(b)ak5a_0J5{$L_{h1`Vn_JC-*Z|tng`bz~_qC|~Q
zoB&(mtNZ#Z#QKZUgQtd+Glb;YcN}P02Vl3ljZ1+sr)Y@V%L5`2BxdEqL6mgn=TF9(
z$JoAFFr+5y+yfvgfnR=!L2=VT#_bmkN`c>KdTGf8yqbfo`S*M($Oxo&W~_x2*u@>E
zXS*NaYHC!C&N>$D(cE-6CE?pNfO_NWORrX%XV?nv@bVE@4$$_CU{JaZSQE^~e&FHH
z(JIx8VtUx?+$~Y8v*6oxf5Gm7xk@m?X&4}h0(LxPK#wxYmp~ox92RNhV7y)l6a#QW
zD(tQ~AcH2(6lrX6RA7VyJU%}>y6*QXtiM1_FzI=wB;D*nbiz%W+o9~)GOAWzYmRl1
z>3w3t@Tch2s1}4)3GSBEw`^A8b3IUAsE0u(&P}dR86zHX`_|e`DqXlyN>VJZLw+Ho
zv}^|QwDJ>|YGE3S%9AHqL&!#i@FsJq?B2`MfRuA~{2{R+lx_|aSaB{Vya8M0<>`3`
zD*=m0o#;nR$<p4|;<TQ6nj}#(YJC+HM>#k+KsfX3vV2>=*iO%lNx=yG8W2B*uL+8_
z<c1$X=9udUNEa&dkf15*$7WscjqAd7j#-z!7Z{|x-i*DWt)p`si~+=NhU>nw)kE$<
z1D+mgcqs2_!bX5IfUpJ*UOJu^RE!{lwRXnbm`+h&!g&VeFAyd?zWWmje_)sl7t_UF
z_dRv&cpf_`^%*!A#(s<jt>KTHT-8g2FxovN+}6jtl?%GD2FIND%I4sj0QjwMD?>rX
zE!k@A?PWhU))gb(<sZ9WHZE0&H=Bow_;!ssBB2{FjhQ)5i=U~yyYG3lpfz0O^le~3
z+0`#pq$(L=nDx^^XN?H-W}8E10XK*r!@*u2_=pZzt=A5Fw8l+CeFh@3Uj8YFmJ$z>
zV1CSf!Tl%ET-e7nwN$xQZd`_KDMks0J`!XU8`~ghxmZ|Hyr9cG8ChI<8Dqt4s0f?x
zMxA}v;GiQ=Rhc;>#ipwL?^OgWlvwrK0>!&pTP51UDyw_=ndn6wXxXPA^wJgdr+8cU
z1!C!~2KQH0RW5J<d%fYnFsbCHNF}&nV2OL`sb^QIxG$%rl2J>=tT~vEX5{HM<H*{s
z`{|}A_VEQ(FuoI2p^<y&wU(HgIvvaPTuzliFS@w0TiAIJstYdc!@i~tSb(UJ?p&X@
zRV=JH7$KnJL%4K_E08#Qs|jK{RA}>aC>`q0)fj^kpa%~Son^`uYPuURG7|Dyzaq69
z6Cv`3Id;tfJ%!I2ewZ6ZN7zsc=~80DXWSI%LGOH~j&IionP&l4G!MD`Ih5`Lo>%Za
zQ0z=2=;UT*ma&R!n+VC>tI2JQCO@LV?r@bUwlbA_D7>CH<MJDX)V^s4`V*{E!laFt
zNxx~EQdO0{VEr*A=Zeb9mUi2vHMk`qGIFKLp8PjlU`7vtyp^$W^q2HVrT6Np^=Yph
zmi%Xm2313ChD-HG=x*F}U_Nw5pna`$3*am^;&wnZJy?|x0YLqeo2wW$!X_fr2J=%W
z8>3&!-u)Ll%YyZnAO6dj0C@PnO5xxkQ@>7lq5l<Kb3zIIR`y4Dng1hf%_pF`ffGV_
z9h@J3NFIdJKFK&BFP^}+6U=<VEB(WF|9MT+7x`8vwEKUX4*wTV_P@)quU((|zsIBf
zMW+A||BalxkBo);=SiFp(<dz1$reJGvXd|LUpcrxYj~2#_@5JJ|2)&gbW~_WC)5OB
zU=kvye~THIuRLI-z)L-fWENC4fZP&zDBI`HV_;02g2G>nJy(P=y~51{j(*qy02ZVT
z_7<{pL$~1+04CM}+y;4T7w|2&)k+f)x6=nY(vS`!<F?LF$dn__<@p2Xfy0BkGxN2{
z21t`8A4EK#Zn4?lHdT=l6_wfFgvt((QO~~uvxT*_tw5ArCH#?iD6QpO7sa2?cWGdM
zQ2vN0vM)TaoCG;xn?2F_yO0P}DTQeQujO2FKx^9+@XWt)ZyOt}aXdRvt)j<u{rab_
zE(;j;H%JCwHW;!)0RCFnq3kE+oPttWMFseGo+3UN)`!&OedB@o3!C&VxcK<n7x{^R
zE1Z~^_^%1M*Kgk*s^R0uU4tJE-U;kMNkKsjtAT%c%h=c$46xfw2Pyuak)ZTlZrc<{
z0RrfmON!#HFhlsZYa`SLjHQ?Z%QW_h=7GK=OPoF_KN(9UOVb;4y8CYO9^3nt$Nb>;
zai$v@8-ezMM2(VBC6tfsT0+XhxC9;*F%c2y*3)694#%+QUHmU8xvwQ+<eoD4U{;x?
z<K?1!Kl|d}Du?k#aq&=|3Uo@~^y(@ya}?v;xxdtwx>Ft)X8`RV;NP;cUr5o^7jBz{
zAMG(ElRW1x6Q>jc$4x-5kO3v4Koc$i*Z<r2Bete$b^LIDXXoT8x-K2()3#~&LH(Ca
zANU2Jdgf<mBeu={aKM60yuWQc@<BNJ=&-?&XC{aRrvmvpU?6mngdFnX@5J3nyz`qo
z{*y&R`^RX2nfrey*>*odx@r^v7*uaPhKetsDj_TOFZNuMngYgUK>;X$OpqsjT4)%Q
zlVj?GLj^=^P^AID;CuHzIo%RN=yc@N_1*;a`_!+7%cE7<yp=F4f_FGy&(Agis&`K<
zCqH^d`iR7u)y_z3N^N9kQs}SU)jX#Oc>hK<%qgFq0P^(*5~3FETQ7efEA5ev)?167
zb~_mBEL2U?i-id8!9RDiQwPCiV2;4+sl&I2tqxg-{DK1cYK_Me5bmW-KRq(|WG8SA
z-Ohz)Ant^}2YU-0d{7euF)h28@{=b}_Vw6l2|~V$?3O)iiTWKkXUjECQj!R_z7+?*
z8`z=Nk&#0fDgN%=J0zxvxw`^jwhs$tmZsTM%sE=LY?a^Q{<jZ-6nH58V|o|>{q|nw
zWysEDkc9}D#<4v|89Omc`(0&FbP7Te6Tfc(QxjlD$vkB$d>5k<Qa;<^!Z!uAFP5Ue
zV?*w1+MFd(y`UJNJ_VANKzz^w4`f_8KpTB+O%?^@Ui+s{nozpTJp%XL9LbZtw#aTY
zT)mi<GxU#TByxw*){wOJ+lO%puz{&A0RTOui3<O_Dj9a7_<+^E0c!!X$^PY(5i6T!
zfZcM*0T5P?050T&qT>aFWZ(MvA(K4eM@Y`1Dy`>(;VjVZ4nT3j!QNsnd`ZL}&yTr7
z%CNppBppbyKmYG!^Zv0_u)=>fJM8iL--!-nlm8+0V3(h~Q6Cc<<~sNj@s;Zgag162
zYU$}g;`oOIzdxf1c`*;eivAoc{kz7h)%8ursTfKhKbSqH6tPV{yD@vblzzQ2B#9r=
z_Zt9<YYMYr?hMpU9*3+UmZG8p$iS0tf|3L^b#15vP-AS*byi)e4g9ju*>4+m=|{^)
z>*nB^9Xt>#k56)JYi~bZkavR2>c@|30z{oJUc3Nx{?xBpkTZnijqtm$E^QvDVJIeZ
zsOG&}W0H7FaJn%&?0GeBnFcV1*nG6dH{INhkvcOs@ga8@G7cj|k$k+=x^@+mM<l?Y
z-l`ZCUz9U2@uH%y55n6aZ?+26%r=i9ync53?E*^o7*6lwa6JYt6wK%V>&-xxQbTGC
z2UWKZE^%TuUSqI^Vek~1vX=VVSY4nONNvAl3gt)%_A*C;&;kJZ|H$Kl2^BI(?3#(_
z#K-GH%rjVM1avc)Fft=vs~@7C7q0=G(VrrQYhq&3AWrAV%*#6rTK3!EDQAJz>;<lw
zpI?0t;%K0rsA8b*U)iaoW&HJBpweL6+m98+Y{mB4TF4SmFf(^`bVwbk0Rji{dxUvG
z4rM0qoWiKBWzWS-hdjM99Qia!2uu<40mGc<gu~%ap$9GU*^jgPp`s?V0}2{`LfHtT
z;W2~A8^U+7ZIbQo`Pq<-VX)1XP5uG}UwUP*!(kbe_MpB+)ImqK(Y-|xoOe<U1bS6a
zl5t{AvGt9y!}n8Aq%68kus8bL6d|IR&IE@c#(vlKr}8up**Q!?(1eGzO@oP<UJ4EX
z%x_$$yVkR<2?JEuWs3J^V$5~lft3WB*{|CkJ`Z=N*-3Pydj=F_IAzr-WNb8YH)bew
z(y1s;*sRQmi5Y^-a^E(Tt2fKzy3yHu@3(`nO{w3Xh&2s@PB1Mn#{|~j6pq~>N@3}G
z95M>#67nFNf8}@2EiRgvnZ*ygXYo~7&*wt`_4`QGw*3^G=8hsc6zH>Ik+bN_QB^Y9
z+}M!s^n)%2TL}EQRoP+Q3JjagN*P@Q?}|;2jSQkeN^vq^U<P=mFyTx#J#X?!LFldk
zgYw7b=-<btrl#gFmdVXZ{XT8JXe?pTdhB9$V|HDy^=MRH1}WNSpr0QuD^SjabeO~X
z583!Z@Zpe^@`5ojP=5kJGz_TPHSH$x1hyzV94Ilx3>gYjY+^z}Xhaa%1mS*ZsLz2y
zebId|-$2Ae#|OK1{dpBI*puG&U1MV!7n@<|(eiA9+m}CwOabMby~VZ1G3^)dUft2K
zFJP6+@s7m3doS`CuT8^NfolC@$Z9}hdh6O6@V`t<OqOF1>Ov?UCJenVcV@urfzh*r
zLU}*~xdTNeN><p!#Kgi6A2L2EGz5s&nINE~<mu&g`_Ux$0C5RrWu%b;Z@zU*4t)lO
zcU-{2QY;D9;c?tV_7nhlP5q6;SGl<>bL6gF5&Kzmr~`tdD}P?1UQ2YP*wO)?@+WW0
z{M>O>1O0bUg!px^L?@{c#5AJL9Bw5Z_#uoRlxo2+o`JzZgh@Pk_X`Zx01T)N0ETND
zP9a!GsVEkh=E{gvjNKswuUHJ3Bbb!%=ZZ~)D5$<}0~!2vE<I1Me=w_P-g<+}qU&=%
zm)nMxutqyuW>3S%wCue#2Hp$;6x-dq<Y&oAATJ145FF!!Nxrymg_B)wBLyPeIQaN!
z#o*Z^LvvE12+|;5v97aPM>izS-tf*|hKZM!=?7pVsbfjWXf`9qb9%6k7j_5{i|)7p
zBU^SERq{oH6;y@&N$F4u<|2%k$Cj^Z?P4B-k)qzJ4x&fLUhu?VbqF+g2DY@`{Wumt
zma>WpGIu5%@;s)-#`LJ+8n>FkcVdmp(LPGnypaV{p%BrZ)cqlA=OTR-I(=Y72Jk8+
zA3xsV;8=!9SN2G?h_Sc1_nR)vt*>jEwO<Z_v3UFEh+bj2=W7pBy<msJkHjS=F1MN%
zUoyYx60_;ac{Yy3DE!Nz8+J#jYFj_uA;Y=9bxD)Ca==i+7~cgpsrY##ZKb-zfQT=L
zuVrbl{qm&<MqP66!f!Cz*s0DY;J*@;4+Z-#13Dn{0-VxkPTik)6N`&buKyhz6J!Yh
z<@Yo-%|YE=a|M)aNpD5=IR)<LUt@r>lud1Wn4=bcBt5GC47=K4gBl)qWw!tsRNWeV
z`#=U|sW9=!W!>D<D?k*M9Oj!JIs_nOs1;!}B@o!a6US<K*#S5P+5xdi=r`$zv0aU^
z%?1m7EP>i{j<I&+00!Z(|0`x4?2;4Hg>34yvw!{ZWG&`)3gk-^pmj=#?;V6jKu~{J
zZO$txfjKk=e-S#=<;i(aUi>nFD@)<?hGqwhG5J?=#15U|CxVJ=u+O*P`zn{1#lZ(5
zn(UH?JP%Z)OFhN>j|yFKPl!Xz!7V`{K9m|6Pr=n-Ono&>Ace~3r49%^z>bA%gY{sx
zCU1gzu5a&EHvIZsT)>Y(1zdsc(i0-dI1<g47vPWffchA|zK8DP=Qjj<2=W!{m#w**
z@)n&(764hmgrqIt|5>cbU+cO<odZ-TN=z!6^12?BI;tdZYksM(r_H<xGfUp>--8J<
zR6?=fuG639zdW-IrUgb9z2^X)PKlcwswUv$m?I(x0T1>Q(1AI0!+{ll_#CN3kcSdw
zD1)?~@1_EP2ai&g8zQbs2ngYcEe?E4hP((6?}VRrCTC_8Vc<7d?A8_Fv*_Qlq(F8C
z%r#P5q~}4E5@y>x%d!r&f|j<n9rg(chj^m-Q1=)%7e*K^$`7^u+fSm9`Qhx(24wcf
zv~+ZOU@i|*D@uXBt4_(Or>R#Dtz>2f1_alxr4}N6UfRmPB4fP0y}RL4be2cZj^n!3
zrvy;BKT7C<FD40x<a$=hd7mF(K(m&>RfPqugg%7L7bYBKBkDT`HWlI@%Aa6Ah(5Xb
z!w_455TSZReGmm!J=#N^5qkX@j1*KDB)-`WrIUa7O56z^4TMnjXld<G%uV?e3Wi$q
z5AdhK_%3Wlu^9zNX9&4L_E3Sk5Bf=9JQ1EwBnS91lflBwe8!-klaomwd;<a?UzwCb
zU}Fzx1Zi(`W*q@m7C7YKwd&rPWcd$jTvXJa5URAta>;KY0htYVXXKp+!m45fV=g;!
z10km*W1ub@a?uPlo=B_%O&`;@Aw#R)4q-Bw7@Iu_4<_|WWv+BxBcrGwGy0>lKuS&E
zGDjCoKYLt&Tus;f%~VG?+@a5{*O_F}yX4BicHQwZDu5pkl_lBOP>loKRvA4az7L~#
z84MDDDuWb%m~GVhTf#f9?nuVK(ufb!w?e09_t9xZBEC#{UE_2|87GzdPiFe)Tm`W+
znrQW4mjGqxzzKXa1pJw(P_cYTnoy(o`#?eT^@ue3lZBL#!WscB!T7Yia8E|x4C+4f
znPG^rdG#;h1KP8Rj-QWV6X8rOyl9ntkmm4aLLeQ<9F&+Vg~NeT1zLNlh$T1VhAi0v
z;2LsnYf+VIK+QoxS06hJ9zd-ri=Rv{bq&naSyeQ_RZ>#Aun%?+81kw|jwG_&8AEWO
zt09Vdqc!;YXRlMCIHg3m%;@FKt*x-hC;E<q%OP|kaHjiA^?Z}g38V-&2U7uh%wPl}
zMJ4V~{LYQA($~^0&_T(xb<ho`Doig*)1W!nb0@DGSxIOnRJdL8+TZ`SOyY}*iV8+M
z0O3%v0%>oS9#)9JjPKFmK)Ce5j0aLgW-<q|%FyVOnVHG5bLGO|z<^;cQe_NXrAe*^
zslY)<N`+2C%gtkIU$xdXby(ELkjeubBd#EUMa1V=*bVQlY1ITT4E?J4NZJEG^AUWH
zbTdw)Zc2v>%rA^?%1?vome>@>Y)!d8qYtb+G!mWEP0FKVk@Hu=K3XUS>p4DjVXPIB
z2UOS+!%*}7RC^0%UC9hSuJzf4Ic)EAtyC(Yr8Jd~mbM9?_7MTBAoSb{&O8ggVceTR
zfqTJ6j>=qgWNo33L51IG`~90dUCgMd=2zttNUO^T;irIC3bv37L3(ikr>E}84^e#-
zW%*h?C$ZiJotV8LC^-ohM{+CuW=P`*bAr*Q?u1SFZ65d^F(EgIF4&owU;FvV55lNy
zKyjbZstGR0MoB?BX9kK)<*>!?j{;nRP<o%5q6b?A#jN>aGZu;=0cZAI7D^;qX{{DG
zi&HKaAW8_)3-d2kn)}dS8fHQUIG$SGBvm%QBr1w1c}E>)I#L~q`vxlfTv*7ewCRR#
z48&O$neHo^uqkMIp!_WKT6`ajb}9+_?p24wy`qqikiF$Tq(f*qy}B<+DqT-1gE4@Z
zU7UL2&5-v^l_xV+%}lqDd7zR9cskJj*gZ6)8ud)V0jUE2vQ&rU727hxZdfkzKoVho
zd^``bDJ5b`>8+s;qI95%`EhvPbFx)+N@`?pR63=$1LvcAlve3Y6=E>mz6@EzZLW7~
z5v5SgmU4djhN{>loRSt2=@v{bMn*}L_!Aka8Hw_|f`XxZy-QQuohpz%Yzn#w5dpHN
zqI_v1I>q+FRJ;|2UB=#3)=?e%Lu+e!MTLZi1QzVLi(r%?Ek;|nUxzKwG^|W-R@4ts
zS{Dp*)Du=py!R3S3jGq)%jf0gMMCF}<k?ih*_e^hR`3?OX4E(7YM|Awmw{WMu+y*l
zQ8{fnpzq)|z36mxO?0C?XbZY}Q@Dxr0sSJND%x4DR1UmGu_<WirM;RRI@E&s@aPis
zNymW2O*t<hAaHKA3NXBGODG+SCoSat9JxNoMLHjDeUpo6dd2-2h`=y2NG%4h5nx*l
z{z`^2I5w9(>*j9g43LxF1s)Sxsb$l;d@oLGa9B=0%rqmc!k&OK1u)$qp^J-)*&IlL
zB9bYMe)3tJ#s(<~QokBp)mK0scn*SW7KcEz@MiZ8vI}_7UqEG)jSUsn5w}#S50UYl
zb1vuQ;-Nf5^u5_Yo?u)gpWBf$;OiT;Q1K*DZ@hzj`Er;~!WmvW`duVLLQQwy_flMT
z1k_JKOCvMX5dnZUEJ=WTq?||FbZ@8Ce@@%{3rFS%z*G%Y*sVZD3mTnatmPqi;S7y0
zr_j&>M-3s9*lec&;WI??$3OfZv=F$$ITzh71_uvl$B2!1i{~@k@ZIP-85-JkK8Pz#
zjEuPWd?+4x;nB)`8W@1|y`P)8IrtcuW-C)evjcbnNCL%`z2zT2>Z=$|p<S<XkSS{~
z`V2&B>MQX5F;XQN85s~FWMc!;f#Ki52OK`&J(uCM7`t7<K8<#ddnZxT%uF42V2Ric
zcEtKBSc5S7$2eR_-48J|v?4(TSTivRm|O@A7umyKwqZVRHx~dBVPWC3!GSER(3lIJ
zABezkENI0WG(iZradfnMjvM39WP97+<a>T9BeV&2-l%}q5TtScnLTu@3mom%ef@d|
zI1i-u6nZWIu{4;#Q3~&1(x?u_0)&vITCyBOnNqO5?+#I7P!aet+5VOpOnx-96bx+f
ztsPhr_<1lm^)8I>YlF$3H~Nq?zws;?kF6+TK>NZk;8KhFIU2nU34wZxxZWb}Z_O6!
zcxY(d?=aY({Ai1UvgWTX1U~dAg<I5bNbYukJ>54IoF4iy+aBYMOmdUvQwCb*CX_@9
zxE=scHs~+U3$Q1%*m`Aj1m>+dZ2&zv&kB_QNXMDu0`ytyTJvTc(%D0aC)mR*jl%<u
zd4_kXo$v$KuA`yp+G9I9IEb}EscLUHI4RLq2x$fC-Qgh^e@!~81495I4gyyI<NP4)
z3{XLC(C+6z5R(S~$CyEgieWNqjb`nu{IHsunsqIhZimLVK0cJ9T2D((qp%Bp7a6Dg
z66Fk?Bf!b6&vr7sMMlo9B0mQF&Dz!$B=jeooQ!Lt;2Zx>{vnD##=yUNY4m?2u))dp
zA39Mi94svVEuE+;)n+r+yExVRbbDvp@|q^k5uVolzV~vNdhJ#dU8~xR0A)nw5#BfI
z@7Yh}u$IDme`@zB@rZ3q)S#1BS5>-I9aw~M66#+gyyJYxg~m6&Nx*rS(Wi+P8P4qN
zLhkA?rDj+@I)A?-HKn!7K*G+(F-+6wOxN8PJ=ub*Q@Pep=eX)EC#qODWMr%_I9g#z
zr#ML@b3Tn0HpZRrHowQ}&IJ+P)FtT~2NsWQEigUD9sS)*1FuFz9XCBq(wENN3~~|{
zWu)ZrA^2dZPgxv<lGv2@)-XMI9L3pPlkwU##d~)`q$jm;wxtj6KwhZr=JxY01F|!N
zavqAW&JrbGN7*||KcC0UNEmreyj&kAXX%kZkn0+{mfYbJvW|+!WeQxTf6FY!BCmg*
zG~MBF?zsX+GDbI6tXOJJT7|6|rdvqxOv1tD5|sq)mB_Or5_6Ae+O%gJ){QZnG)UZO
z72fI$m&emb&5Rs8(Z#Cw7<==Ji{kks{8vN*A#8Y?`j?Wps|a*?FRXBk22Y+pmP%)k
z+}76WP*0&Mi;Ha;Ni5PeZMw#KZm>kK+HVr~x$;dyeUGP~9JQNqh_XZott>bN)erC~
zu*WA2V}sN^@?^XoM$yc^&6`bEmr+VAPR}UR`_^u3IvRsbSfB1RM8tlBi0jEi`$aW}
zQ|9&0mqZ&PmM$wDWPKD=Wg~p`Uay7dmVplLNLi3jDv_cjvsyPJSFmRIZQ9nMJ%c^x
zYC39qlPHcEiaQ^J%cm>$o^~cbVGFGK!7coXUP|!^*);!4*F`;=VF9^NtHFxs8kfat
z7uTu-cFS8-x6p5SjaAp|V$i-6W1ywJ;>CGz*tj~|>gJf$9Kkg;j3IFB&F?8XaIh|)
zML<duW|0$StZ<sQ?OoyaOV(^%x~HK($zBbFMpM#nW+YSd-)N8_X*%<jV<?7~p&_Nf
zA-w!b)4Aj}{_3D^PX#`qpf}y$IFAZ?1E$F6BE@$^M^$~TsKcAvui%@Mj$_UaE=kEa
zNc(F=KM0#<BQP42!Jmw!>1`MjSCb2Ee-Uh_=+3QB@swes$Vd9I$VW7*v<P|hv1_$m
z#`@ve+-E*Uc^%N&FOJb=u5M7?X=I|+F<2LJzVUOAH=d<b;eJHf?mn?lX11D6V{9GW
zc2aLyVNJ?<eEhnEYu19jgIVb(g|?U*mg2Q*F%dThBE_$A#JuxSN(pH0Ml)7<QHmb*
zp<de1s8q$`wI-i2HFWw_aeU=vrxp(T(lX?jN5#r*S?Z?nc!jDN%}0lVwK_kVJWDOu
z!aqtR!?!@QPiJ1?Ym6dh=6#_a1)84Qr+WzB>l&4NPjgdMqEE-XeWIFQ&qgomWXBgk
z8kzXwn^C>U%O?!lA4D#eC*=BsVY17L^fjKhDN8VT6x9C69-l=k+y-6jX2fmovrGxo
zsu%9N<4A~<Z9clINQrY<{`9`@nR|Cq+a1MlTHbo<hINz6HNH$;RjB0X{m8C6hiiO<
zi)QAU`Jz2cL_&F0<^2mes-nmxWiG87!q;4lrtXEmDA!<1?we1cI4&miOHV56omt)C
zvTSA4z2l{e&UKIOR<wq35GzgaPppSJ61Qh7vh)?RtF9;Nct}W_2923?9FLlvGFOU}
zZ@aq3x5NB(@Rq{a5(*|e1;LJ7_7|(+Lw?TQ^q=ks?O<dkr4}do#?+PszqCBWB$F)3
z#EmSry}wpI=kaQP`pn$e`I!{6aM2f9r+fW>nJ3YcKG&!3b6&WDLr3A`GrF8dvz)El
zH_&5yiq}LwwBOmFM|RlOyt#3m@`$+Qpc9?gx1*mW?~I<J0neW1dphHv9>;1j$$EzE
zKZh3rD=ru_TuaEn#V38h)iS;;n@=L>qxP)L1y@L7hF<HwY`cWcwPceEk9@s$IA2?P
zMQc3e5&hsebzggdRH5XPA%3h!2Hkgg7v1qkTKO?rwU?~?e+rWqd4!Vem;K7Z=xEQu
za?%voiJKrg%X%I+%S`=7Z(Etq!KJOZ6s(#j$+5k5CN3Jw;~efLk2O%|Q_1}J?_G){
zcuK2Z|1hIjyZtg_%JuIXdCYE#3U`mCt`K~i*xq0$ON*_H(3pRVR!`x|)Jm-<dPy@v
zGK+C0ChLj9)QwW|3MLKmJ|mXw@OG-OncXh`cn$t%->{PV1bpX1HEgX+GudJc@2h3!
zj7k()$e9l}S^9JeFx_XId!i#csq&2^9Cv-((S$SVC4Kpq@O2L1j}gY-=}J-@FmAOa
zl;haU&+<>lg`}~@nQXmWmKiblhLfRw?^`bAZlGr71!;y0QJ*5uJ&SDbo{j3eaO#a*
zX8NzzT<6_5)X|*M(~M7tG?<6#GZS^kUI~YC4d=94OKC8zXRY2clQ&7cUK6#Ye3zmV
z`=c_dUaI3jootd@h2?U(V|jM#9oMg-UiLK81E<_Vxoe6U?k<-Lmv{!u=hv)UUZ);$
zB)oE<F_8M&WYn9~S#4tA>UI@#58L!1w?k*f&=(QkK2pV%^6%%<iAwggtXXRUkkKa=
zSf6<t+dNZ~&RW#AJt|4~uw%5$*+?QvsGGQy9F#aJC%?(=^-eA%&h2T1&i9FI>GJyJ
z`D<r2aZicM=bing^^s$|JNUwK{F#XU(9eVDmzvie7xA-<tkO~Fj(%dXO{VZ&a2}=l
zB0|1cZ$ne=Ejw`UTf!AN>C@kG{Y6#pQZ?Y)RCq`J{QB0Z#Fd8rzEZ-HXiXb+?n3An
z{74bc(N8#Z#5n{^b}xLILg{SM?{*qtC2-oh@n@0twaQ%l=7q)<Hr83oC-3e>ET5j`
zU|z`irq2sYm(d;-k53N6n<N&(g!$=iU92tiGJ1mNsL#{g+sGBOTb!{k$)x<rIBS9%
zb(Qu}W4H_H7d>=y6ocpNF71Tbm#JI%{AWHjqMI#7Juo(u_qO|LrJ&llIU$YZkvz;%
z{DPwB^_pP`8;=%0W+QjoxHi!`X(X3H+TbOr@g;WxQ_dzCJ$=zQpVzDB2roP_%6_(#
zP|GhG;8%Sr^KEL2{PnXEc79uq1N~a>&lHe2eYTIB+Z6XQR(&nd^PMcCV@*Lxnf4cH
z&BmhX)XuhcBgL@HTOZGk=L$NE1@}ij5G(Zb9ufAfTJ!$=y1<1w<>hG27#E+6T2lI8
zJ!awQdx_w2NC0|sl$XShw7Mmar969*Ji6vWdmLkN*2X2_^J97o+7uVJ$OcRmD{%Lc
zub5tk(WFk+SN9FeJvE3H<J0jzUJv5BTTm#Ud&TuOlaT`1VTWhJ)a0{Il-eWfqyadD
z94pUBpZvTW(;zjZu<E$Ux*U-7WHelS;*Gg2x!VHgwXLIzrc~0M+oxR!11~2!yw{IM
zXWM<A0)nl33BD6-1{}?)$59o%vqbu*@s}U2<rOZ{mMh!wVm!P}c<TGX&3mIoEj%gh
z{`Fc{?1b0rK7Y_L)UjN}!z5!S#7_Rjk;ac(UHLhi`Cv#jg2;||Nb4ns#eMy^RR(s;
ziWm2{Z~A1Zr<WI!aed$UxqOUqq+?C}Kf<Cg?u7Zexwi2|Uf#bmP*|C{fcrxJ`Co{#
z6{$6wQI=h;T4Y$jsMoTf>AI2_tZ;5-#&Yc3^L#!ROLAU<<7#&ktL8AEEgwIoW+im~
z$Rg@H>N-5PWX`m+!c@asT56<CX=uBDG<$jVR71_ZwY1>7m+#?u5A7tBjMd1t;HleH
zo=J_JN?og{mHxHeQmc9ClLnf3?vocL$DbC_6CN1Oj<uTdo%JE8qggq+P%@^?;2ZHd
zTpXj^9(&pUXo!5w&E=it`B3HQ>hgf?Z{Kzzj;jmN<zjJ!zb_>SXbhv*zO<+_`;3Er
zu>GyZ+^zbcDm8jPY4>8#GPdjfVq;F(hN?%eoJS}<mEKu8nR=Ds(0q>ykyCP7bYbUT
zg!X-Mbs&sP{cJ%>n;qUFV(*0E=XYS~DRWPn$ie-b^+Z#uKy}1IhdIN+&F<^mc=t$`
zDda<AEljw56RtI67Ez0@7Z*Kb$ID2>xu!Omey<@cWB4@HO@+`vC#(s6p~v+~2QKgG
zDwfHJe{_jS5dV;!9<~3umYg={+9ML0GQ+9x+Ku;gUmVV<)bkl12pX>^t!8Dg3eDHt
zF;Hi7Nw&hjNs4<ao5wL=Gk&V-;lsJ64cE*&`aXSWOX!;MH*cSxIy}GfQ9~~ziIr$C
zXZgJ$-;~$+p0{Co%|G+SscW|iFeBx0X;wQk=;^Z}7QcLPVss^tJ(MgR%N~pgNjEvI
zk^0c_kdN`285<TFH{0BMh0?J4V$&+D^u^NKm%7euzPUXc#LrQv#OcG0-6Nx%C;852
z+_9a7FI>iO=xXa3?MyQ#%k|Kmk?r$es%~-9hdA08KOaV2wL6!;q|=Ndy7H2`-^wGa
zrDk~G{wFL_$CasNPPKX5!f=%b%>Kp0j7GN}KWver4jGLPNTKW2d3bGZN)G#NNvyOM
zONl^W_a;|Nn`rxatm;SlT`Kf(R#6W%7dm!FF<leH0%&OlaSsl$9rp?N_pc3=ukHzq
zts5+Rj2=DT_ihd2m`gI`efqE?pZ)XZm)GnZ!EvIZX71Jj!2^v^Z1}8d)l&FpjrAnH
z%{pFE!78=?WZuN|5i^a^fZ+zu;wkmP;(aQ^mBX0>5l511d8zjoX>JS(pL&$4bz#%e
znnUK>IU5Du3pVcR1RL)>%7;D6&)*_x<Ht(M)&Hu5-h_R#(6f@VrHM7%KC+&FY^v#E
zvT&^Q;*ZK;(dVplcM2z;v#Q=DWd6=oK)CsGxw+p$onBJ%vs~7=>P3Md327lrQ~E-j
zl{=Y<^30zs2vpO1Z|)qu&Uy9uS$ppqdy=W5<6Tw)OL2-i-aXYOG8@dscpl=t%PwV#
z3#$A6<Yf}Kt4B-Qv=~c@T<bBl!~KT8(3n-%3<o6teA61^$r|8PzDRT5PoXTTn>3)j
zj!k~k#q$cuipAl%7UANjI@rD!7`ZM@b4L*HPo_TK_LVlpd+R-3rLFPBN~C&9$e@P_
z<Hh1-jMJ#Bof2LY8I4QXrCBC4aZFwZ6cetp@4PaN6BJXVviK{q`y9UJj>Q?`)Ou9o
zY3~%K&ZI{5Wl^i!mRLaGX7N1#B~Bjymn<uX3JH7|m7n?UDiB!&3GH~E3aby^j;>dU
z_Gl!Msbm-n2`l4io_W|iz>9;+c0(@u>n)-)l&|y~ZK+?Iig{1*2swCMpBv_kpt`nq
z=kA46#uWbo=@z2^#Yoo}#o&&iP!_EfAz~GAnfA1D9GdSJttI*#2!2^c&!)J1rq`C-
z=6xg5HY(xt#yQ-zec$`ng@DOOjv|j9wI7Df0XoDVJPiwUub9P`%5JE(bKDPSD!d*%
zpZ_pAOGPA0`Au?$q=>027JBSe%`DUQ8vh-VY}d=FAKDyWBwaak(@kR4`DJcuMkC!9
zBL|5!7eSH-4mqZLXvMay1*m5{{5W+N=1`_wQeRI~+s)kf&5SMOU&t)Fe*g1XLsrd>
zNG4Lz_l|{lD>SrT^m8^@aSKt4c+X90yV%uF+h)8d_UPTcmqPVq@DdrWyw`Y<)|NM|
zc5nX^^WE}c;`G!|JI%%mS_F#B=d{OjKKqhlGh8lun<rPokE>`bPP0kj8RnJ9N3KZh
zOW+aam86Ev@MsZNEA<^^hj-XweZ*(+2WzhU@87Qvc!u>Drss8EVVSLK$r{A;3|pj4
z%ge|9iKp1%PFB~V#}QrEV$+q9XBCz6BQHNo`>ArKDwF56=GRFa(##X3C8s1xCvT;u
z2I74CKIr;bR{6DGEJgjbY%dXB^UYKnrb{mg*{ZR#Oh-Okp{_sUcst5u_b}_&-r_wS
z69pl<#z<tRDSs&w_N$6tJlW%Sd+@#0-Z_VcHMg5$_2#htVwZ06)9s5BH@!iV!5_Vm
z^+Z#CnyoLnMZESjtrm&c0|-})DXb_H#3r(rS0rEC_|3ku!xK;{(Phgy%-_`E^mxh<
zS?E~#oSANX%l&F09zKJykAZ?e@n(9GT-g%N@xk)DNY;M5l0N?)ZOQ<v2HG5#eaB6g
zQrFN2X6;c7xjkp;M8znVx`K8x_sG7L&v5ecUz-TMD)s9W7MG}!t?FdgVoRx_g<sGG
z-OOta=>+w%i*un5%Wmr0XsA4}@z!^~R<Yx+TSJjxDk3O7kY{)Ca_yzq`x2HTbZj|J
z>0MdLl{8lxowX#3ZLNqxa?)&4eMLAvSR|R|`fi3^?>QG?BPJ#6HG}qEei+wF56?9#
z`*L<#c~2gH<K=9k28X(=3+h_Rmm4nEWvQ;Pzn?;*7o=US8gBS5&oD~r;_Sh1q@F{Y
zdnfm?gY8V;8*~0gvrn`?<P!937EOOI>?k)PY+%!R_|^pf>88F!%o(b$?@2<(zOUFG
zU{i8^y!dq;E&K87Vi4oZFLnByQ&zsBb~%Mt_Ob0+%wxS5V*z?$_n#LIGiEA0>FY#s
zedrcv&52u)Fc;Ay?#t4DE|$`~(U+q#yuWbQJx@Qm&Sl+DUfn6&=;j4=r!tQ4&EVC2
z9echVsXM_-z7gL;wHlJg-!v>_F6Y0K9PL}?9SU^365ExPKd#}7$F1;_P^`}j&yW}0
za+yhBu%k|Dr+ALJQUgEa?)Rl|y<MYU1GXv#`LQ1tcUrA}qLp$MNUSjXxWwMcK5zId
zHZG>^N!#NzE2+`!W{INw<~1q_ugs^#^O0!FtC+XRBZpWT>_`VAcX+iDl)ietA-UIR
z${*-qG^E`2-S!Md!ish6vty}VOCNK?;O9w+x07RjRYjW3C|y#JJEUy?My%tZ*5&;4
zjIo=#cKQ!5uF-q<*Q}Sv#VHxTUOfExIi>lBWK?IuRiC}Fgmx#VYnTtG@w#uUQLEL}
z4Ze_I`P>lF6&#$QjiLOpNFJZP(RD+geQ|M%LGbQ81%v;A>zRQ0N^+;4XYBAiE<ZcF
zTNFC`g5iv81ONFa8A(|sXGClY8z!3O2kw20DWc7^op=^aY5t0fci}P{mwx#hz6+21
zd)Bne9(8zbb>w9~E$N9pbDjKC-iTIj^YZL!e6zSqf9B*F!R63eW<?T~7lG(cQTR-0
zF~1t5OZ-=3^Ij)3WO~stvalCalv?P7Q0KJrlUG$Tp+qG0aIayryZ-DA&{FGPJ4eFe
zIXghxV3tnBrEWHtP8T<NkxWgHj3?G-#_c%q{(*&w)6v+2`s7aj;=K7WA&V;P8B0zM
zO%9674j0=r8Sgas>Ygztul&3lf+1w+#i^ooN*#OZW^7pV#2}x0Tp$*<3%VScRoLvk
ztjiyJ_Rfx3dNF(+CXaU&5@KhsUw)PlC$&}-ksIgysAm2qYW$@|ykSw*fB*rBx5WcC
zdR^|$t8cq^leljae?3m{@q4&YU2|ym74tzU1u;=~+U#p#dTqZa1+gureWCC0CAh!W
z%~Gp6GZpe^k)e*xy*IPD_j%f=zYjuBUfuiBVg)ZG?x)jR%DT`lKGA3`Sk~?@E8w`p
zAAdPkxk)o#=eFdP$c};A9qL(Dv6)tq>|A_plp3{N+>!fkR%h3A=@M={%ujF*c4O+O
zvwt=&`QYLPY3$4W!`QQ37dnht>^q6SqfFws-g7JUi8>kIl_f<jar<&m-s(IR#PXRo
zSE@8UEWEH=DeSAaU`j)H(dFyj7?u9X7l^*dTNKXS6zb@En*S8_Q2oOL@l(ZLjA(@G
z+`oSky^_5dMU|S2w|y&#;&D#fwd3t~hX(?F=G@l*8Dy|BGqe0hkil}7h4a4^WHhL@
zTm2}#421XQ={l<OM~Y=?dL<*7gR0cehX#A7c_P{<oSGLg0!76sx5n0gTrXh^y8ap)
zP4G(a$c~h)o0H@I_GT-Y&8)}WafO|wPOLf4l$z`F@7Uweui(Gi-%j3NQW4jicTAR$
z@DYtjc9O1{nZEQY9sl0(sY+){;?l#o4<uFI@1Bk^l`PQ_mg7}9ERAW}`{z(T4S(|F
zZda<oy_*m0e;p3xO0!4_`fC0PcUw5-ujV_P&5REeq6<|RAs2|>^%TA`_c+~s^~H$B
z@xqbwp3@`t0IEzs<^%8LqrK6`($&I2=frM=^%Kj##-nUjvFktg^1%2?CYBsA&Wk$@
z^{!=3k*P_BD38aFra3l8YbS5G8H(X`$x)9c;*{%K%}Wc2ul-oOcuZ;3^%n0QIf_D_
zjh1=WYtl*3^4XFHyK)IfidQsVZ{RDEy}p!EuTWK7mO){UT|UfCs>~ZU^9d$*H>dfn
zKD`rcwV(Vt^Of1(o`G)tcpVjFEh_K1n;vm*O_%<Z$=LyYJpI6w)dq16p`I~14yLH(
zl&G+)?PB*VE>}4%G*(JeCH+6Ny$3j!|NB1PA`!}#Pzo6(TQbTX$$aQh$lhdcicm(`
zD?59ZnQSH5*-7>$d+*<M<Nf}A*6;KC9{=O`cR1?M^E~%+->-3v^SWN=sW#>yXe5Pk
zF*>*IM8qUkq1?}w#hY)8g%zxIy)1m`nN)Ls;LN=%$?$8sm8pbjAFcc1W7O@T1$mT-
z5(^dGSAE8T*)&&I`+<yKIHUD4L)tzPnU&u1_ZJLK{g4oDNp!D%C09IW!=qMbgZ3aC
z(zCwPx>}J?9(7R{Ur`=1jJcO$#d?wD%$kX&;>?610XbBjOC08<a~JtGx^Y9L=K{xa
zsyelqd*7c*sQ+s8=GF&s-It2ZPu~90&YKZ^&FN+PM5`O;^SV?MgJy^k?YF1^X7XDF
zI`L~V_pjAqNb4C_eYVhSAU#9Ia3b-ZaPX|`dnKv6)V;c1m4wMlc3j`s8*sFrP>$Yl
zWHj|zBJi}}8^oWScBL~3k(jT!T|ej=-Y+-}@;nivZw}UF+{QF@HWucoFN>I6SigRi
z?m^LT|0T^{hKtUxbi@fTJ>nOac);C|Q-3Ge<3+o3L`0cUpJ93FR)s@rOBWw-``kg1
z5b^HNPdB5Qs$twx{oY%+RxO38&r6D1>_*w6Z>4ukT9x}0ITrSC_19zV`C&e$7-m@I
zJLyK1fIC*NaQV)k*82lY_wjCCGq=6!#uT7$>cH>6%M~SFFm~XvYxt5Ypm=P4DEPVE
z6`k*9)PHUbt;N$2Z?0e2oV@qBTlC4<eRKfV2kx+ydvRwf#((hWEcTl%3ZJg1s9MIU
zJ~*?;P5g{|GgbcAJq?OskL~NLcc$4@D+Y00XV;eq`}dP`c3&yXVIFWuPz=2%Ik^;E
zm0C_VN6NmfNxGe~w38cWu>O|DjW^(@Kio-Cz}brht~xsw5nn4PNrP>+JKtmu{92PT
zau(mI9Aa*Xv6^}+GT@&UapF`7z17At24|eyqL*4_vW}}sDA)3>;+m^VKPp|`3CKuB
zyB0Y8xtmtk&ERkQ@OkHJH|x{%{BNDT+T&`U4g3(mpZ1D9O<q!ximtNiPn`_v>3d7s
z>5$uEuFL!RHVd|vSMVqQz0EA|^MtxH!r70;(T-oRI#_#~PcPO~$##lJZF3ej-W9N=
zE^eZ3WJ=oLGkW{n;p7J2>im`^Ypv^K*shms_z4BUhc&AOtm>}c9iHJR`2DQlA^9XK
zBavS7(q>m<R^(l4k)w5R%lFM4fi{oFV}$JyeuTfcDDQmNwkHYB^uP7!+C6Wfa6<i!
zKn1}!?TV&L(@$<u+J7}+=huyl?D_t&dCuD8J-yYc3mtu7L`T3$3QA5xgUPdBNo=g9
z*zfodoH^@gi}lU1kDe^~(z1NMc}hB~<W;4c^hM(g$}=`Wkx?|plY|D}aHPdl<#iu(
z71!Od4JvZNEg-`Pd;S8eF58mEgSBO%f%x^QOUaUVn%;Y7#=AdL=y}bpbR)WWHZ?Od
z@M{21r>aF5o5LLk9tP}aqGXKt>q_Gy_!s?Bv1qFWs%p2CT%)jGgPiFTe##eAm#93G
zt3Ic2#f@PooeIf*JX-gs`jzsT3u)p>l=vY`QbR4ikBxoYDBN9#lk9j%&FF&?zfSjO
zVd=y_$JTeEA`#OmzkY6FSbHIA<n-C;r4KVgRQ2YDA8>1X9I6t38Y<JA$v^#peK{jd
zrNK%}owL?cxa@|nt*uptckGB&fl7wDsO1-<Te??d%15elGEZU@wwK^?eiGp%a2tOW
z`rc@BG{cuQ)8ct0MlNAECPx9uJJy$v3`FBIw6<=ihQAH!-X4)JP<zCCLH-6VDKjN;
zvXpiU*?m#DT#Ebbf#uR?8EHdQ!he}dF1!1rM<qJ_CRQe**5K`C;L<wl!$a)a-7c?C
z`7}M7p(2NcviXl5O153BW=0y7d%Z-LQ-v)cMcv{<4rasIL~YlRc6t9#wQ-{aw02wV
z30Z`{cKnlt$(j5v*;PDqq(;fBs`9pHWm?Q#CnT8t?d55rz+?M0$&2Q@Q3$P?%%{-o
zS(=a6Sxvv%>-2ZK#HtTU)krT>_<b<(wW>`uZ)CqGBh62gx1s&oxP~V6{Ca4KlxDx&
z3TDZrH2dYk{w+p4UO&blvX5lOO|deBWH@Z(4X>Dte0K*{UVM3s_ZaUa$-PH;Sw>~z
zVJB|IcK(!e*0@=pt-7K6dV{dm)hC#xM}`iCKOg-diz+(u{8hn|(lQEhZ__Tz^k6b8
zncdB<Ec%Lqgb8n4hi+r~4k}}1biWsaN*045`(=lm)i*<%lIN%9*!@hBv<EEb%y>25
zjV^AQolYjWv2?;4%PaVW^yT%>%g>c1F5s;xF<B-HCTYK$3Rx|nv0p27_nE=-R+8LE
z@@OD=^rrH|t}ag2r}DU*(cC&0jqK4}RZ<D6HV39tJ0Gznx8xqGL|%zL=R|8*81Q4n
zRh{Lk#Bg-UBNnC%K{b}lC2nP}sJ=GR>My?(8aFPPrcV~<++5qG`ax2sk<*)5n4I)k
z+hr%h*~9jpqXLd;d}Qru`|}AhqKhv?M3jeoM!ZgiXUqQi*zSE{?Gyehy3aFFiRw08
zs9Dx~ioOpMw*%%*cn@HGUT=t6|DBByy^6=cizeM5iws!1e=XYAnQnRh!rnvH_^_In
z-Foa2+qUYAE^qD@tZxPQC$7}&D0i=jr%VN2jyIm*u581QF7Vg*J(UxxomAr^W6s@c
z5Vs^l>wEbiBe^vz5&wqM>Y&&PcL#mZxuq{@QQn?QqbZ40ZmAbgS@%a&u`9%#YM4zA
z7+M8f^U<+G8%0?P1vj=Ygeftimd13s!ZgSPmB0S!yAoT=k%6k8AV@O2!}K_F^X8LV
zkx?XE<^EKrzsUs}tS#4eKE27Q(E26*guXYEXfsXiz8shCqoo^{xhJYz8(*x(+D_ID
z8=l~x*`(p_*(j5>Hc+@4wrpdVFn#A$FS}KyICTYaPKRx(VD|e>VwB>Lu$lXl+7g!8
zdx`fZa}M(GcyuSeTuaXOmi#m77lL+C_-G?n<EP)wrlK;clrD6u!O!-*cBx4lYe3&H
zDbJ+{rhu=M)a9)^G2u^~qO5!IUMH@Q2nM|Tc!6Wy(vBI|rJ#dZcgMK6wWd5rgh-gd
zuHWv;>t5gbiPk4E_cpJ`*cC{t|FGvq$L=Wnad{OpQDouLwQz7^bYC_5KSjN){~_w-
zJeE88e?jYA(B`qqg|vCBp3-}yV{zJ*Y|5<k^cWZUpnJ3)T}hLh#4U$CB9*|-UzW2M
zs`69`P}eJe`4UmPQg1GfHm<FVdzXH#zMWb1YIvpi<kZR3gR1EiWsj3jRcjWvTsFrk
zCWSgIl26|W55g0!ycyHgmw86|=JT8r1M0Z$=5DE@Vy#z;a27&~UbrU;{|J`0r|B>5
z^yx@QPv#1|)qi@1{FZs8>;7s@2hsBuxfnqKjeYxR`WGaTBHX)^JHBqu2_g!#PNYV3
zJ8rJHeKBV_V7$&|*phW1df;Yap!)4KUAct*sNw!H6ZO6{zFXqA_TJ&#AFHM2Es=_5
zn+J>LiLJ(+H!sXQF;2eJS<U($<D%VzZSjjXZ!Z!!`utF_786267OOP#f16tH*%Z0&
z?0)Kp$ND_Qxr0WLpBZ;J<rt>P=wH2&H9mbSnQTws+%4VLLARboo*}IkbGvMQkxT^t
z;pg?|_q}ePF#bkzQcCRu(+$V_H?`b+g=RcazwHv8$UJeaFOrL#6g;WMaXy)0YFUcR
z=IeQuef5hH;U6~0R{YeeHlq4o`+b%e`6Cv4abV{MAKxr4N3zb)&~=VvJp&W54xF^;
zq1*DNgL3c<$=8QZkV_A%UVEO>zI&(4G#96{ma}M>d$9i@)#YzS9%qGaU#;z!3R@D0
zG$Uu(lGGf_*Y$eEDET!oU2DLeRh%YWL&si+PFc@{D9NT<-|_wlWf$C@Ho}#@hf6%?
z8!>-Jzd8|{l=y`0R1z6qq+AKD-3ac0OQVehZuY|BB~_s-lzZNC3?j?-^ap9a1xU`X
z{qi1u<c;nn^Nv21Z|QCrLR;_vDc9Yc>gE)Fd8;ao(C<ukL1^iE>r?8ax>P9Ez5dv6
zCqUJw`zA^Kn!SC!%UfRm&+WV4+FmkUp7U@Maz{#b3Rn;2Jff*8aQ0c%t!OZd_vQi@
zbec7>xF5aOj<>=yKCds9R)_X5n(2-Ea>?qHF`n?J>Py7$Q>-4n-$ifqDq|Hr3%nV=
zs??(U*b>Q<`T6#6AIY;~%umaA=TC<%HC(`oA<$tkj1ax+pEpL!5FhUQc08+8ig)c{
z5mihJrc_NG3;j5Yl!0D0Q$t=vDN*6DO-s0ypUKLN0B4Lw8<i}H8SNmHVtwEf#a~~7
zxBt{`-M&I`PuTBi%gIYbR(ap<yzkX<@1Y4OR>n&0db1VvZJ^=DQoY&p@7hh4c&{!;
zvTJX#jV68z5Exn0B`TK~d3YVa)t*w)n)NFWhNSg(9zV&8*UP!{*L*UF(l0#0>3SIP
zMppkcopD4Ug`Uc7efHkS1b!(kW1W(RrV*-dV=d>R>|UnU8e<2z$Xd_jrcg?Y==nzw
zeXl<wFA|P3M%mm>m}(cXd@$7$SBc*r6Bq1!M}S4+rB}_#z%i37U(Fo-^pxL~H<F<%
zepSgbQp@R;?`Uw;hvy9Aay;pE-^d*;w=|}bOIB7IHed4&o$<_K#9x)W!AO24HFa~~
zX53h}mm#+M=kAr-_z0a=jC<9kE5F#j+9{M6)Z8?T^dyi@`}{*eezk{B<)?pAy3jW1
z1%vrcEQ98IvW~-y6z?x2`>jPPq)!-;T#F(+txvZ)pIpJ)+&L1lVI_klzTdF!a-!+&
zs}XUHra$7=GJ8hh7%@!-Q^h6uG;8iRf;eXwd}|`rR2f3>vsF&?J#c@YCX-2MmAFpA
z7a(gQX!NFuh@E=e=XxnK<%1tX`A-7hoGPv`4H$OjyVZ#0^&&EGoULQWYjvBu?Qz7L
z8h5?z{96{mWdeTakMC|KV>DIWz0$+&>Mlfl?T3Q8+zlxnGWYq%g7*`uu5{>O*J)#8
zbcd@vW2=yVpJG$yEpy>@)QLhu)~nAGY!$yc8wH^*#-{oRJr=8b_>n*?_%6H4tUURo
zCpCwa_MY3oP5${edBSGt-Wd1Lw3ny~H$*uxnd&cmIu}ZRhSy;|x~b<;m>b6hX`kO0
zOKh-rJvEKuU%%U;TE}J09I;Lu<q_w^UTQ2<&OWWKCQs+#@>W2fNII5(YjEA)=)8RP
z4NHM|ApvN>ZX4Q_$g3&L77xCbt0?LtbcI3Jdele+d(<*x)AeVVoptyVDXo>J8*jFQ
zbX@NG{7607SEF`=SL57=VDkR%=%j+Ll@_$a=n00BiZ_n>UhO+4FQXKj?3C5wb>H0B
z;E!OHHFiQ%ZiM;%ag1y$WNEGyHH~=JxYIE~!^5WZp*r|up{|$golU-;EB>qazKk}S
zy)Bp3jH;i%<J`inzW*z5*>5TJl1GyBnQsIKHJg)-mrJ=WId(AOuG<hLN7CGMdlsk3
z*eu7R*KH@q-1V}%@oBp`<Hs2$-4EcYFGmi#U45uJ5g2`np3I-y40U?BeP(KDSK-XE
z;RGk{=(iyKqI<5a?|-AZ)H;}5$4bpK(pP!3W}`2-`hHvCE)jc(WiDJNlM`=iacbq1
zZS}0>DeeYG;-7UT7MBZg%^qDjz3^}Z(*c)zJ^`cr9mUUQUt+8L%Rf<n(?4xjX(JqS
z=B{YE^Y14VA#T)vKELR9KDTahUs|y7Iwhv52Wiyr1g~6sB%rYBjLm8vA@}9JZx%v%
zS@`OW#CKb_gt4d}gzpNv+5~=@8BpghUY;rQ-!i!o;<`CoRh5k|cqXRAxTrH#!TD!D
zSw;N~isc7|!XsQMX+)))$_}Xy99YaqWvf*xe;AFDq;?8tkcDR5blJPFc{aTJOJ+e=
z0$!cyd2Ic-&S{oY`5{vUgsYD+NPAofKP5bH(YV18XcAaeAY!Gr8QSKUe?F1g6mvJ|
zX$k$l7G|sU-Fd?3wJ3i_LWgFqqQwPkznbez)vsi9L{+1pp&!jHck^5n%MG&A&5c)b
zX%<W)a<z609{qHP_Grm^#qs0W{a(??Z)$9mkGt3?7dlN@=EUhs{F2N{Y|wnt=VToe
zQ9JqhbYHoZ_A0(#^U-DbJezT|;00r+>IyCTHyiQO@8PJ&6!c<~GRx+*iw`m`>4XJt
z&!rLNV3A`4x@tZo73J+IT~#tXc@861IoxE0qdawBG?2S}Q<mYjb3UDM;kkvOWQAO(
zlE-4N&}Jh2(o|^q5A9}2R1<o<>8zY;Ra;a;%JZrGyM#$b=Oi>uf)YR3MQxa3`%BXX
z(tRUWNMqjUeV{Ha+ojKVE%bSWF^&gXD@0o>>jZ~wL)QJEECq#Blj`Lhi(sXf21;RL
z19wCDhO3rupOg&vUPhCiS0-}p^U}BoTbA=Gs-bhDsP|sdX_~-!)7}S@QBEHz5`_dD
ztKm3s!Iibjg0C#s(SqXJV)Ehjj<#y5`T_1)`BSRJpD&76#+zEKxSA0s+53@;Y^kmm
zy?Odu^WAE_a+Er2-jiOb!JOo?G0KV7qh~C3e^&VhZWc{hVw(|tR5W+8FjhS{ZP5t2
zm*&LeXV3d^nV;<7*gresUd}(%;eT)XfkWG%LHpzI{}^O){~NT*%Fg+}>)*&GYcw0>
zTAa`QfiE%~a9QAa)pP2l#PY)L*;9sX{fctd2M0ffpYbFbhz`W_{~$<rs5E4EiECON
zf986ZZ!kJ;B>F;P%s_?AJ?is&?g?L>E1rq)@R*)+*&eMWqcDgubkL4|Pdv@vM)+11
zkKq1^xf5?>+k{_KwtfC#m-fsr^}~achUjg}5tdt(SRo0-SZrUV!dr#_{BJGRz!y^@
zi_10lcEh_)oH<OxFw*WvMt|0f?$7k>&hS>aUDfpV#xS>N-^Z7t(X2n%M@@tjq{p-i
zR#%f^=z9YTg!xnjq(#4^bzhjyt)n;RpLJOspLVHEAoj$Y+snRH#&Jo4OPndSL@t%M
z%PV1dT|Qp@Z3|IWL0-p{imwIUw+yaHsbYf@Ot?Nz$gC$RG^_=$xoxv7bMN)jkP62?
zV_3daWZbs;P&j<@f>p;4e+8oQ_{}$)UpuXR@pUskkhbAOY<QzXnlEQj>$0YvePsKl
z-v1F9sbpmSAD@>d*Or7no)gSH)!SnFr6C42mx2mimwblfzrug;TU(>HIAN4w-%G{E
z<$ahhfudil{52kvHC7+Yg-sISitXpu@W&lyd7g-=g=6tr2IP87ox8(;9#2Y3t?k_}
zqcq@Z35?D%V|*Z?_MG&YK9eTNtRzZgKtN$knKn%>b5(L<e)XwFlt<JTr{5^7my8#U
zDTy+M>8c;uzyCB=<UBrN_Q_A}&5-Kum208QOPa3(?tcoi%58|p3hufRuD)p$)@7Ek
zSsy`aRvMaAFZgZO=Rohoi56C7hwa<c1#CX+?M#;`ij^@R5Uyok_UM~w2xvEXZ>!r=
zAXq#j71E?-J`rma7sPll=Il7VjPY5nGq+2fW-`5k{pB?CfKi%xNj;bN3!`eQh2Vbd
zpR4B`9$rMXc3f1PeRv1&;{;oyiO4|8O15*oHW8+$+Xcc<{1jSIl=e6M+|W}aef2qA
zD|-CD3bTn)$5e@TB})oj?aI|sgMZK1XBS}SCg0N>A*fZw+A&jDT4uwYy8c6&C~BTS
z)3o7d1ATn&^;v^<sk;>KaW5-5IJYGeT{G|J<+8ojI7uEuY0qc$T|)R7)wq#YX8P~0
zmZ+9>Rau7s8*KXig!xY#X{?_)&|Yx^S3gh=S;t#;>P(%@LfM?<=U?Lhih$}L=^qr5
z?(*(E{}7}_e#$QRRC{W$mN^d{Be{A(MARK`HjMDPaLu|N&(ci~j(0>2s08OlJiYKg
zBBlbm(XP9kE==6soL#oK33oALG`A9psRvVs$Lvasb%bB&3Eg{waf!@F0M5!q$8>F1
z$h&O!?r()X4UWRO&g0kVRh32O$h<BbnH{HQsB-YO$=FDF#<qku%<eO}*(YVg1-sC8
zzL-HH#(~XSDUZFJ2n#OQ5>kEv?1Zt&Hk;tn9!+ojdELO4*Kt*)?@nZjwEq!l@6?|S
zdO=JQ-JUmhd3?t7mK5vT!N=pv_c7_bMxBd-5*~M$j-Mf56st%f9+xh}c>9ccv8Jx9
zDC)s^j{DP1-^SW47jC%FK6~^DC4OGvdEuikJcKV@b4L4b<W=!K%NUx;BKZ9Jk*x!R
z%aeT*Ts5lp-1Z+ljq_~Pffr9DRb%NoTjz$Rmvo&BiyQY7^Le_-Lg_j>G22JM6kyY{
z7VqeOQ-4gV_;sP>{kxo#_hT`3%E*$W*6%Z$Ita?ypmR*Dse4Tio*EL_omQADvC#hR
z?s~r0_JqBla7FOw^0V_xV}%#f4?Og=)u$X8W5Wb3JH?$WRf8Gi{(N8l%#`YRo8CJv
zq~q<z6Gd7RwcY)<6>cX_S=}x9PloySo&Pb+N)FF-S^k@fDmTY%771NLeFN(&+#GBy
zLWb5>a=Mnn=BCfg&2-JIukf%4o12(hDn8TH)@2dVwKde%RS*|qQIyfNGQPron~fE|
zOX2VSJt(*V+tmNNfA3GZlf9qa{sGGayphXS6AdUPc|%M&4MZF{B`BBWQ%z9vO0yBJ
zzM(Wv8h^bKqkMc(=wgNeBkh}nhbTsC!<U^2coyGtM6O?_3GN--`SnC4XJ@10wPe$S
zd8cLj{lZME!N{IH$FkU!GX41-mxGN;tgA0C2>qA8_R>!wFUyg?&T>cIn1_A%ZZrn+
zIUVA|5C4b%OUP?rv@J_&(%4@PPh<nG5b074p~-{2$^#Xpg@e7pgV5GB_aXAwVfaB;
z^0DJIoytGax_BP>(IM7*fv)O9UV@*<tQzLqI7hE6K2X*qiNGkMJbZ;bww@6__PqED
zJKy0?e${_}=YRVG`@2Wi`2T(@>)beW!3><7Ye4S-XXe!5EHfN(6cQGOL;4_BG6Bb}
zo&bLnXmmB58hCSj)8WI+V`F31_5!c*?Od8`;1oHW^@D?9y>JW^IXZbm84jVrnQhWs
z?o0mgxg3UAU>)tPw!8#_EpW$ZjM&0R6tK3*4R3Mie7`{1>@J~9;~J9K1hi7-eMSB-
zkcD4X;FuQEqlZAcb2z;AbeRU4BtS>A!dL5K)#~nkO7En3ab;HfdvSbwFa11hzF^d!
zE16h80f$V)980E%sAx4i{O=H&1)VGjC$OXw;qPe<I4!DB3%Z3mZ{cJn!l%?%45U?7
zAT8zu<OVuAIua5Ra6mBDehkhZMq-qeH#w*Sk7h@E2FS%Vc(CXV1INhGJ#|Yf*SEG7
zkD|#y%Ht={QoFfM{Ya<-Ifs=s;P^23{pq9is|=M~SD=T3+wDm&MpP$18yUm7ufLy4
zhXgoWl-gxPm>s5*!;{`~I4tWNy;&B{=<#j>Cxe$hJbo2BTC^ws86v!ScZF2l9(l~Q
z@A+*8M5~N%o#Alj13JQ^(XD56jKg@5W*y(&-VQ5I&5p0Gu1;E9fSLKcx%cr<Jtct?
zw=|S#ARHJB7qY2C7%xM-`=8$a{n-ZRAKnHO71t=!fTfKM$G)_}{-LzAm8GQu)jH6c
z4$Z;hexaY|<I+?EJlG-P=~}fuK$L~dIZG@Y-Mm7%PVKsR^6#r@lK`1WI-gW-#Wo}C
zf!R0uhV@r~w%Tg`xA$L<xRd{EZ3YFW)W_w4?SKk-oA0EC>y{|*1W*oD_rc~m)#}*9
zhzRV$Jv=rq?*gqYAjX2ThXrsT8YbGu$A=frVFO`vLm%_l!~`Np>F^3pod-uo)-w;y
zZI5ide=895#=+8(Z2%AUu6?(*vO@~!HD9KnF!mSUG<-&)pr8Pfd`Z_X5E6bl{0RC1
zoLTQmzW?p$4%VYk*lR${6vu1RfTwpAL>Uwl;7C54?o@dB5vUK!S6c33dQ@X-ykm!P
zD@?8kGzAJb*s-u6dn@vN0$_2~R%(9OV+N*`2(&qz0(aeA5=SQfeX-qiyQo6}EW)VJ
zP*2!(uw!hZP>d=$zd<d!N8KqK#EgK6dgSzHz2+!cCwxo~O<;9ck3orCpM}*7M`-=T
zt|24W)~BeTkmODLhK%*p(Jag|Wxo0jTqYD$RIwo;t6#(6)T=2RDh0YUl}cPYJ3E{l
zpOoq>pyFi%$!R#RI6gkk3Qs^)?g9hY-X;JXI>%!Mi@&)!2F^IbS=eR=i-!+efVU+%
zbrY>UEp57wxmtt6f>0U|$r|aAEUvSyuJ7(T13~$8&x1yRtP6;hJk_fD&J$RxBnRn9
z2d1++(2wBI>Jv{9GK$+ztAO4EA&sM_T8BfaKzkK~toM;QIIsHNIt>`vmu!(uk?*2U
zoG>)>bh_Dk?pN2o-c?pspju1-QZt2*BXelmT0X>1#3qTqq&SbsUMcQ*xX$u(@}+=w
z=DnDI(uNg?6qQDMf=L7VjNT0W6j;7yw67nHfF2}Y3)hMB$eju)w|!@Zr%`eg1nz;t
z@)jW}DIaTgB{(b!TzZYLLM}}@faX%W#T%eqJ6a;+*E7Mch<`R3hjZISqiBRi#d`PV
zro<KGmV$t~D1q-gkUenO50N4p*E1_I1|49g>AKNxH)l`sfKCxk>4&a4IyoUrD9izd
zKFvA+PS5%rEr~{yGANP0Ufx?>7#yt7A2xT}d6%cIudfdUxYsaP*$Hrj{Knf49v)xd
zVDRGMa9KIDFao<);R0GtBT+j&^(J!Z+E5U81H(OjaeZm&c!!PO%mm6Xl`U5|&OPTu
zQn5Y5l#XF&Xo&1>nOLA-p`{!KGOy7?6EJBRih{$|jOB4a<I%UlEFvNzL7SSAf|v|0
z2O!_hx1LnBwdFcmIJZ>L<J1S6-m9*_8VT2i^TG7l>ALA_>tJpB^I&4a@D^a9AK!6F
z10!tf(6D$oZ;i-~QnuE>;pyn;=nz`FsUL_HEUX8GnBe3uPJe!~f}gEtTzry~*8r@E
zN!)}NE<CNv2W6%oz<-uL_UaWDvcs;lf{9*VT$H2g1G1ZuBQqAG7YgUN!$vGJXEfYE
z!h+eUTQ3>SV?8A|O~sPwx*z!N-MeJ7Q(7d8wv991)UE+<=14W1U;Mp2!Fj-xeuNXj
zpxr7sQML~R48TPwR~ZcMb5@JwSCJ{x!`roZ%^Fk(V6NJA89h}$YGJ@OHSXv`gh}~6
z00jk*Y8=nNb~B@>*~G#E(9P_EDMM%wx0ey-ML2GMP0TOtNlD(eDJcA*7b!4}^+Ri1
zfTPf6VZa6W8^91%0tZ<^(z`Z;GGHD9Rw<x9vG$j0lVVHWarHtL$K0HjDlVo+C`rhX
zcd;VZBbKE341qu@XeYCjuPVR+bVfckuv&MmCj8X(5^iKl5Dd)E&kq2-F+3Zh-<jtq
zIZ_UKQiz4%D(44Y1>O%d0Y{5Mf2WaAPs%xq`i`ZDc>ZERmR7~DVsrEKwL$Pn{UdCV
z4`?iz-WtI{U(>}8LPtK__;#jJhLT}tnmjNEb@wpy$#MY^g)%5gB$Fs<!pIxL-WG0`
z--Bc9$NN7gzX1-~16df5I8M#ao9&AuR5f3{9HTMbx7mXREO4h(ZcV(aJ^1mythT<V
z#}3T6vf2vb^=>ADzJ=uKJ_xHFS!Q$(I$44)cp470gU3?`?jMBr%#W-sRhAlr{cfp!
zc3!ML3nY_Zo24WqK7mfIx_Ycjllyln!M#;QYimTT7V-a2t4>^e3Jzs=z-a-5&6qVR
z_yhz57#Qlg8iSB6Cwl!GyUJB27hJ5?Lmvv!<5QC3*zzcP{5!Y{Mnz!go&3+yBia)_
z2Kxl25PtL^&it?M{BK_X?&0V%|9{^~{#-t`kc3?ZPykgA&>Ask^Z*A~g)bQ9TlnA|
zk$^o(H}LRZ5#{?xn4jmOTwkmKJX@rp<YZ(K<_ikleJ~+Aw~8(=gC??wuA9}I8t~~b
zWpD%OA#$nGNe7sD(Y1BpjdNaZvYB{d*K9Mf*yKV3VOOVd?#9V#Oz<VE;oyQ6AC&75
z|BP2Ji`frC4UxNji;Kj7F^I2(xWG$)hBmY6-`)z2Cjd1R>4?Pg|LrnAVNMY*_b*=&
zn-1Sg(TUS~`2C&IkRYUyny3D!FZ2@(Mm_aFKS!drFYIeU@QdS;?Y1<8UG^#JJ1va8
zy}k4gqk%=G4=8T*fo%w^1^9B1N{j|dAuz`u?w8w_c5m+hnIR$)4V)f8n;QWFU_g(a
z`2Ic6bJz^mU<B8Tz$JHm9OvjV!4sPNDv(^Qt*@UpG>->`1RyQFpdJz(eV(E-6K!4z
zfe?^hAXh_Z`8E%XfpK`fQ{$pMScS=%ne76P-(Vtu@owO=jdUtnbW_25QqjQR1{$L>
zQGMq-Bf^9RTii3q0O&>F{w@t~fe!$)24p?>SXj^O>{z_|K$YUBxLbZ+9&l<YG(~zU
zfVX+;=T|Ub$K2h>h)>0rw&;9`O$Y1nqUHLbkiM(~B@_`JaHsIXblzP24P=-rC53OZ
zEWt4opw!YJZU^rejQY;FNxU}0l$n&21oktu<>Yt9&diJExnOC)*#Nf@a0=UN*?|39
zUR^Z-G41N7_9Hl8->C$gC6upb4yklzhG)@S&IQVrnM&aOnwXsAwix+3H8lku1hjnU
zZrw5idgwTp^+@1`$j=Ag4r@iXOXGdpjb$(1L}yStf&0DHaMi)V0r=+6<Kfv_SOlu`
z0SUC4ekc&Qa@>3ke~Tzr=ccKoqLbs}t9p7=Vidqz(KUUDzy%UzQgU)o!mfk4{(`w?
z#Ey9DjTpNDkYMt9+SQWR0C`-J4<A22J|*vJK@U(-6oFZ9=()?o13Z0Fm1tT)!NCfL
zZ@FpVSqe$uP^Io}tgk<mk>Pa12m9rPJJThh48GtY5D0_%2oRTmN`>-j+}DGm)U4$N
zw}Vnx1r3%NO6is$Z8K6VB`poYF0q!Gpn#EweK;xp`k_D`Jv`UIUu@VRW`D{=E|;pj
zJi^aZA1(TlH9Z<=_^r5zP|_ONuvFhbh=n|CBU%@dydVooLD$P%X+8U8{jqsIZ3e5W
zt4iG?!^45W!P+fFbitC*Z*LR>y9hHxP3{`<xqA6hAWaTrlp&>*90nIm<XF(6qkDV6
zh?b<9@-W6|e*)1QvFrolE10gJ!zHkv)tp5WEwZgJ4yh(ALmguO4LhJpQn(2uBMz~`
z7eV;^aJDz|x$>pJ;)m=w@5~izdca)%TFAAG^K58nU~HoTslJeR2y5TrBP{CU>V*QG
zmDEnbpX|EB%QCHP1j`D@lZ=dwDYf%UNcV$c(DVO!GmOJH9xmn=AO?cnja+*HHfyLS
z4==BPzVbzdfkBz$+Yrl>0^ZDm6sa~h0=$5bkn7RT!1x358!NC*;U5TdU{`}U1w{Qm
z8rSf+U~;v(_93yTYkw^!xe@OgF9sS(^Le0X>69WHx8MEP9N1bCA;1RdqB#(yNRkXr
zOiToAKWiYC&BJyTEq82XY|mp9-MC3dPtVjp@!D}~93ke(wpRE8$c0Ym2(7@IWL3yC
zn+8EdpsK42_3H}|a@&NXV$*?~(6@t0p5<^k@au*8`JFfP23?8Wg@f)7C_#=fLxL2Y
zUQ`metruBZZ^BqKZgnj~upCvnRk^>$VuM5XFxtQkoh+AIx&OJAn)aS@b=;CZMU#+%
zLeQ^bVPRouBSim$-}Vm~GUK&jdnj*zAiImIEGamplOV7;blRzz(KCD%QBFcFGR<_U
zD9h*4b^)*6v<65_fdCN*P+g#+5`c{d1f+ovIv0{4(4yVmC+CX5$ZN1_&|w2M)<<Yx
z_!HT>ND#LP=?bd&=}^^+&y9?*ZOl5;+Kh9{y8(vKR0h7@G+^H9AY&&x4=g>`4lTv|
zEKA^W*J`;hB84uoOkNZ=%#4n{Y#|1cFIg>j4j0Jdb(QG*A@iA`^sebNL_|bMzF9Oh
z-nqHCHv}0NrVagLfu<AUq_`;)u-n&<?T;m}C8y4n0Hy&1<}7-BfWAoK+Y|pPcj7)A
z486T_w}T36hcbx&?FL;`c9j81u(h;w)WTsdyd1rjCuL?f4=W#`>TL0?ecc5yG!rW0
zdT1~B)}yJfXPg;)?n4sFrXRT%5kY471r`Q)!yuGN?;t#ThTB90Q0zklv|WpVfdLb%
zG`O#r1PqAAe3>CaKz&4b@u5p9q_1H2)pUq09M`_~gM|FMlC!|;YWO(N?Qp$w&F0qY
z%(T7t-hNuK+Aa;T?IXaPz;($+^MKxokF3B}5=bL8^dPY*kODTiZOv*N7`dTECG4ZD
zoj=>oNgbMw%E8Q6!EEd-t%SfZ7|un??}-cE`)6Jyq%Aa&A*(5flpgTuc4w-{ZoLIy
zB0)_r6!Hil!<7^t0{5vy3=Xg`4iyv}+_eMxIHg-ow=em(w~j-C>wS2wPn9FPi9nS%
zah4b#|1F?n5fA95X|~|Le(5S-RL#?QZGE9)f7|$Q_<prAf@4%%0Cb~-_Sb9cclkVM
z#X;UP6yWhc+Pc-g7Phvw2D!Xga;5E!1&vS3yTC-5d_Y7@tecA#4PjO*T!z3j_;B62
zDqqY1Lc4-i*r%Ze_(-C%9uN})6minZ48i+g(||F|28b&mQ&7$kPy;v!c#R-lU4^|U
z$}&)5m1@laJbz|zWr@$8Tm5!}w1?}+H)~{al75bzi>pY)DFP_h3cO*N)F6}GP*w@E
z3WFypDG9)!<4AD~kV7{J4}**UTI&6WE>og4bbhc>@UQu;&teWjaJ>`AUA4KoJuWuO
z<|6{LR@5156NuVgpOZeEC{na8kaV>RNN<;&*CoUW4Guni_z8QruD67$`Fj44(@%dq
z1q{oFGXft)5VCAl&T#=`&lh-u^<w1CMd#BU+WDP5N}&B)=E0)=-q4UYnq>+2QTo~l
z*3pg^RlewWu(uJq^9Mrn*NJZm+AHGXsE-{4Fc_oM9hf*YYtG=zYj_;xrJdW~zf1hJ
z0bm;k$I>U3DaPBUhxg*+;CLVl!Jw>?+GRCx^^rd~y<Zl}B~BYT)~)Yv^PW6?I2hnd
zg6lg0K~>-*{A$2avE0Da9#vc`xO4$A$qEyDAba#_m8TN8QPJzN1~)Si2J9IpFz%F#
zej4)TTlI`Wpgh)2UI(%cu_Y)-fL2|B-w&cj7+Fg<P^TCl9!}sp1zHRTTdi)^fvAVE
zd}yC!x~|H7EMLGL$qw(sMK`pptPIjr`4W9QyA4A6heY|H<mU$lynx38qf!h>8S;E0
zFWs?B;wiD8@V0k#-G|!-sEL{K1#EHEGIb_hVruF*lAEf`k3Lj2J+pzB?c{09G)W2p
z=T9z&cMCz&flGTRJ%BD9DKd(K&z)h^fG{mB?qdaHpNGL%bXeH>uwv5w&3NFVlXY5x
z?QPRGSffwt+u;wR06>$>=kucwPmWLv?)=FIA+e0N4ghN1T$VfkP8YIUA3|C;w*#Q)
z0<cywnnsfLcbNmMT3e7`pLt3*Ha;F287X<VyxzhR_bb|#eWhvMk<whcP;nQ?G-1hL
zmj+inWKb2&h>t(Pm8OEj2Vm?+cUqn5H8GO_R1%>-1fgJGe>XSFxTwk}6%V@h(J*oU
z5`eTfMG0URz}m{eK7m9QLrUs#THNm}`wHgkP7~wS)>eBEj$T7+YpXjG`)w!=pvUjz
z``oz%bY)<gvpTgX^oiLK5eG0EQ*V%1yPnY~TfJOWl`@A6P*aHA8x>kFAjN__AKW~j
z@bDfOj_hAr*}9>i2LQ3n3?OZeW8`2=^BM?kz##y>BOxs+iKJ2yd%(JSXqy{~DuJvX
z$aewEbJG(A65%HW2J&WRIfai+lCvy<E*m(EVOd2fnk0w>fZ~SFPq|9IQ5D3Jh=GjU
z0q|!E3kiAQQO+zZ9DJiNKZAh*^sYI;J^uAIAvrlYxU65Rw!!rw#^=cW#VFT`<Ku^b
z^$^ypmPD<C2r$H|yX~zN4>%ex4pjpjs|x73rDa~Ts>$3MO2k)`rXN-qgr^XfF}7rb
z7aN_M>j$<(e1Xk-mtTNC2LU;#I@Dhp+zvu5>hDHEjvIloooQ#d((8%jZ;<`q&eZlz
zLtKHRQZ>wSECUbS0QUKp*M`a603pPLBF+Cjl{?D%j#4?O#r)9t4~f>dxO&D9)yx3E
zV*nNi(*`48e|HZuLv}tzxCySmvbS$>!^by!oS@}lgQ5BB-ZtU$@Kyw$`qvx%=Zo&2
zZ*Bcea`h$0t+Qv(dZ^7BdiGh=6Ei6@3+~-C2g7`rB8&g`6q!+Ft;VWkqy(U|65H(5
zC6hW@2<t0$SBuNKjQz7}Y<fByT?kjh9fu?L6>`HCPP-os=~@l@Ub6aM_nl6P;10+6
z@L|IG_tPh<+ox0>O$DUxUi#b8VD$7}9A>BGix1q6a$Miy4vm?^*Y!?9L<KqP-=9Jz
z@E^X7uS4S~-M)>}dHkuvOCDwB3M|O$$FupLZ=~%m1w7&YN~<8m;)oC`)UTQ&+39(C
z9RaEWD&O0Y57t$30TF{2fOrzLks;s|0M^4-uU>&L7&xv);McTY96CAhrbnKI&-g>-
z-`Vx!1dC=+e1JU%%zYsC3BuGDk&2TIiHHa}w^?*%=4=O^C0IY?HPD>gN!ys|$^;FB
z)y2g$xvZ~Fuim~T4a79Bn}r3?SXK%99jG@6+!~P_`)t`!j*V~zLen0=&05^~0imIs
zx(y*(-&Hao(*lXd?wvQam}@DislG~jx5*eIvVoTRGS9OU$JhK&`A=yzph_P-&^$o2
z2|JK6L=Nh4tt~Cx+T|dQcWe3sDFYwh<)crU4i5g6JLeQa?hK<m<w62l+)D-2_YpnZ
z>FHj;Qd3<HhCrAbq=sF(m?riNtDCB;6CiX_S62r>jcI}@9QcZlAJDXSFeHRXQzrTD
z7aBlHRt0dx-ph~S+K6l|e!+(osTLx<yu6G@#YcAeGEhR(l-mOwz?;#d!?rfx{}}fz
zF!KYz(9zN1s~#B_7Z(w6Ln=qLFk@BA;)3hJo?RJb?t69nA{n<@a$T5_F7dAB=H}K`
z^O3Kwfn*#qaL_Y<@Z`yFK=Htrumiz9r2UYT)(Yrpg3MF9CS;R*P$241cLPRuFzb?o
zqDWHAvS!wsCz{a%beME>DfZ#^bA>%g*$|7m7mwP&qT`w}zkmM)PozR>2*jhrZq)}6
z+W>|Q?)DQ9`6B~oN3y@S;0C0TY7Xb2(D?7Uae31|)R#%my@N%kZwARkHgIlBQ2??)
zb~GK?{YYpnb+)6g?=B~2*DgPBnv;Re*pBf8-gD`?+sw>O(YV|uKd-nHTF(EzGuPt;
z1&s6oK!AL#NUVw=^z$z;1!?A^Mf|w1GBPp}7&`;aBns9OWno>k;ZcaaZ|mq#bU4A!
z$7gA2+0@&u!}i)i-V7e((Ojz#$d`P5preztMNbFuMT04#(`sAx#&NheV=o5q1#2z9
zRW0CgL*mezNQ!+L=&hAjxDOCLtv)CivU78*jOJf)p}cOry=u6^y|J+oUg1^Sx5Nv!
zV0Mk0ojo5`PQw=-gCAFr%VeS-fpU@ciMN;UXzfqHZVl#o>+j#S$qY#?%CWVzj7K~j
zaE0matH)EX?uL}-r=^iZyM`T|{G{3a?{qwAPz1C;bC%FQfj3@h86*xMGlha#d(Dph
zU3KyDMU(RVBBL%Thx3E3JEq9j7o#`Ff@MR*TkGr7lbiySm5c%N`KEuBpT8VRzR8bi
z>UG#8!#A2@`6-U)sB;j>@aul$6;nasMwHY95hhmE8v}>`14_yq0yq$jD^?5xYim@A
zCZyb}x>?DwSeW35z~N*NgE9l(b4-t6V9SOHVKC2Mgh*f7L#~}ajntWhdg@XffW~L0
z`OSZR7>;}l06Qe1Kc%D_5EleMp%KQJGkkMe`$&Z{>H4r56h<#jYCtIoXj}2Hv7;be
zL_~z7d_t+P@!>oB@>zlUdiq34a&ly)1FYf9%*+BrZ`at31UQV-oLeEB*@61CufrY$
z4o+(`wAm3*grxm@xRZ5JKHZ)Y4<9P$8Zu17oI|OCBoGtUrDL2Y;P29B7wf3qINVm;
z_GUF^NFGG<$nWF_#Kzuoxu$cQ3?kP=^_GW8Ay9_`?BgDz6`-|$Plg_p+|SC{x49h@
zdu4AYn8*gd;s{eqCWMO@=?hbGK_mkBO*^)z=k|~XzBMZj)JV=@V=D;i^fosmT50?H
z`vU_502o3(6{3D^-vYwV<EwY=)6vqVyJh2nV&*hq7=SpAP$W9qsSoIat#Q3x&FTqg
zk3h~kGdYTddznX0JU<{Jg2&|NheYu}l3dtUV0s#wY+ROX%ka#-Cnp0Pvs(c8lROxK
zS_lAtnl&#0F9dm^*X+z|!oxV!;*S#tH^7J|a&ny+PTnyZf31;$1)0>eF^LEl&bf1e
zW;)85P=o_dq9<?%3zifh_^uBD&B`JZCt%P8X5<lP#a~j<56)RB$GN?ZJ`Gd2=hzP>
zJh`zLDpR1ZQaV!1>%93PIJkrG@zeD2A(PuUpqvBAq@$jcd<F#6NR_g-QWdK6*bslv
zA}XYY;(7lF7AS*}`eg*04KWG1J?U{=X1B=z6p9TGPj$7pf4>?i(z}Wehzuc$`xWQ;
z=g*(F0^1UM6QaW~(tjWq2WsU~7=0jG26eJ@jfwGb+N@)5l-;ATW`0l9*ETA@$rbXA
z0!`2o0w_FDBbPn$L3{g>rz7+_i20FygbK=1i8EMkh;N*i>*+^EA8$ng<#lHUN_vY0
zSxubkxvw72I0C`BUQ2XjU|@o4IS)AqOvv|%f)w$wl}oRLtkrcy6OdFcD}19-f~~WN
z&0PCjP{0xRK+4gf#MieGCV93iGj+EPUvO7$bnO=C1285*Hs4yd$jrv3uw5wel}nmm
z<q~`n!aL4MCt{E4rxSDdLhx<v6C*BQX+(R}={HHKd;b{#$N>PrsAK>DcL~8+fRrOk
zObQDI3V%Sx39AM%oR{`wVDC_Wfr$QH<ynA%)!zK9aB)HwV*l)!pWwD;Y+5^K1}iH!
ze~`X|)JAbz76NM39PaDaMI68EY%FkF^|`yd0|0ZQI2(`v(VTqG<GFopbz_Z{d3q}l
z5C(v$w!5z)mu;K1w6cP<ofm4vfWtAVhyw|-9%NLYKII62cp@p1NEuawcq`}QTrvfQ
z*IV1$VFvh6I}D%c2TJtgk$qn|Vj{uo1<?vXopiFX%yZD}KmkK!J7M!r@PO8(@QRdY
zHa0FJ{lb^h5l(1=+dxmv&_=bqvXYz}21wiUBR|?bl?z_o`aFQjkjxu)q)0m1Z!TVi
zc&g!D4}>794!a;el0EJSN)GfV^C4genooj0g$8~=qR$>jEi3QVR`f#wJO`VY{KO$<
z+d>FJxi-u48I5PI+taB*^CF*hk^zD^wgmqB^S$}vyk4xVtjVt0o_XJCZIOk4kphx4
zWvX3R!eC|%O9lXgk*)%v*AQ!ZfPGZvtLy5HzcRaR5Zs95#!b*tfq{a?2PHH0d_5V&
zg)rqx%gBVrN}B~Tac8OKIu^cwB#U6K+JK`wO<uUf`}y+|$F@|#Z4hK8$65agjzAtg
zv`5mo^X{EHYEbAMT~jZy6g`!aB&iLuB<)lJP+a``^U0ZizzBrcX_2t3vF&JTY9cN~
zP~8rEvf;%Z*!$Co1CT#Uc^p6s1wf&IBfk8N=)@c?MaR~AhEOABh%M3MxTN7QPHfoR
zHxg%$1xCh6YZUf1Fn70qvEE;aJD5sTU;C90eKo8rTKN$2NcM37P%0=$ee4E%1`*Ef
z^Rqxb-ujyn^pAP>g}u%r9YHzzn!qi7YF|G`|As&~aCobfSn5uH$bC|xz_JGV!j%A5
zweOJbxJ|e?L)>a%X&J=z1ghbG2Q5$-VX4?mT$cD;S_;JPi4ig&{C1g=GPyRFd6S(V
z8L*fB2z+2c1_c|$N~@Ko?B>E+AwJmb)sK@Dj3>$fXp;n8(5ttW;r4~t2z)8nDu9xb
zT=xJ_{skJI7PqdMwx=u7zi)eWyfo)?e-%c@#H7l+OWA+82?;kmko-;;*g_;Mr$&oj
zdig&vfxk=O0K^c!S<E`4j^4Nw>t8Lurq?m$mbC*c1e(@m8v{;Y)*%P5*V-ES08(NY
z$E&wv5OVi4d_%DHeFcUq^Q8;ejWfOZsXQ?u2u6{d4;TgPvGs6NQyUZm;x<KJuFpc0
z3x3nsvt4Ran1LkF<7od&XG33&p(GRg_~fL1b!aYPX5VLxjE@(Bj)Gcig#;SX5G4)w
z<p<?Z1_vzfcth;$L(9e?_;Akc9j9?va%>yKCxG={gX$GDsDXEDeOU>!4y{QI&CP{x
zdl9(GkDNWJFDf<-5D5%SOfY9Xmq;g%!xFt27#e_Y1Hwi^lmNK}pNL+xGlKnL;O!5_
z3+@A&@&HS~Zh;U3>7J@w>wXM~Y%GUiJ4_AWu5*^_AYer{1JRGXlZ3=~NIc5Q%TJ#=
zg>?Sa0+0nz9wOBZzVK){qp$4|zqy_zx!DDc0egFUgZ88Tpx3F{54Au+s11vWVZvHj
z<UPa-Al-y2>S4>}apwDX_Wln!C1mGE`Rg&T07)=ps(+VW%aPZcm=AG<fA|!5jDJdr
zw{ecjzW=wB7+#06Gg2TvOu_%tjld-#bxwh79O+*Tm=irJhadH^zWNWnt^f3Ya%f%R
zyXtQySlqUx<s+ekj@B6WWA}wUT~|o?*(x2M;Ji6(9@uW&g`8j4^|&#$Lt}z8cIj`_
z<v+YQ&9Z`&tRb66%HUIf{RV+8TO|x!BtR%I#6wVB-}2PoC3ib)OL&o0^Y35tKUYeD
z;`eUk7HuX1oI$`6sR1}a6U%Xb!_h6)0t^Gav-gxWkDlPT=N2x4RMNK!E3>pXE`dlB
zDBIP4`t%8+($*+WCK?(oKrF#e>_TA)a(txU*1q)C_PdxE7YI85cqGPK29O+zC~*M+
z&j70jBhQLFNFjUW|FSt&CsYWmwhai>f$b#r37t}Lb}oCx0T3XDdCyZfXqbdt{rY|w
zglipm_ORyxKKKZd#baYSFxb=Pkxfw0wX5|`-~37Z<q-s_BZ3Z;moGzAz`P4zV@^L5
z5?0uI<mbJx^)jIX0Ie=hNxn1Mm!@o&!Ulz03U%W(m)Y;{$PgDU?Sg~y*_2NhGa4w{
z0@K)_sa)c`WhvR5Edk=cM|UOUR(nSkT0W&%Yas{KMn?3Py=$Exx#I904#(9g*sjaL
zsF#(ThV5g9gFIbn@?RPoGodzC>pZ^P%smLDKjN}-lbv@#K{R58)+yr9KCqoxnneRx
z`b*0o6m+kF=jyt<v9<=Okg=nqV;j@T$_hwLf#wrDA!GT^+S>7rK{tKmL7dQQ8_+y}
zG$B%>K%orOGC|QtOk5oLDVG6T1VGc5lx6=<ek+K1#l*w_ga-2y08nf?jTq`*I(bJt
zIHA{ujg9U4^&~(OWv|Ql**Y9{9R%9WPEC;!5-OH4Re(Z9yRzD6Yt6&&k6Z;5^Ir6m
zG{pumL_U?6aPv+zn8%qtGD*4qlc4_aapbA%{?;TcCue9dVb`bxO&Yjy=_FLH`W|=-
z+c;E6C-Dr(1ldn0L-YmG894>R&;S;R0fD#~*j(^{yb+MDfWXnjLiOc19an%Ui~j6}
zY4xADD6Q@|O?$2g(5XH<WX|#mT4Da89I{)WaGhidLw9tY(S4|PcSCE^=liXgzE@b0
z-Ba?w8g2>fGJ~emuqWxCyP03VeUsYp^Yi045CK6{nUZWMe8Bdp?lb*8NY1V&(enj|
zX*B;lJR9XszC8tjIn;V=qfuOdEzGSzs5N0ezp%jKGS|cmr4-x8>;IIu3p5VFk|`Es
zLFb8#%(V(06pUu&*~}^%#vxFf+saYC4s~7k;=a5}E#4#+R7)&>qdqwRc{}U+Yn)_&
zwIiQ#J8(6OgZ6P&w-xZs5;zO&$kOT$+Y2?*M(XWb%pUsN#vA)(TD91;CXgmH)z^O<
z%?g6#QY}$?pA9<L>jBItnS)y9(dgvw1KO|5>lIeM76B$`J5$Du3JM$qMJgIDC>*}W
zHY9+#r2N*0+^6~X?OV4VZZG$`%ihz5PO9yl2Ou%UfFED|6qJ{#;&fr1&Gp3rrac3?
zM=wpmy1*0@5D+BwgOtrtZZ65xeg+eh?f9;7v9Roa)y_dV09>tBSn7jFnig^B!+=PX
z5>$8KK|$vi(6L$m)KKUF&~(jCg$4pp7-o>ngmw!(y_A0FXa#(p(e5#*CC9n#mqhdk
zLaqm(QiBZAZvlDvU&9KOd02oqcQ?S2tX+B=r0%*kJ~d@<(Qp$Pv`kVr5`A)L6ecjN
z&1E;2Mbq;0t2vHERzD>JF;(c#>15D_<_!=-3mB+}z~X518uyP9@+Kh2gM=I@eB%oc
zdZbWU9oj1NmK7*>r3gWTo&a>Jo0$B8_Ui$MNy;_Cm+-KOl(D9_|5q9Mp&cr=04a^2
z&dLEKre3|1eO%2)EV>HL{-4oN?TN7>gSHU>7dnYrx!!p~B_E(&Vt&U$`$E$`Tfi;C
zKFj)`uVnrhB$$Q<-ax#kVAYb`ETD<V+0xkFeiJ%Lzmj`zU!4{LfFw+HHRx#6qfih{
z0#Hz@U5Ce{W|;yd_mHTlO9TWo{ok{c;RQkhf+lCn3$S6XW<Dyb1f@16Jt5J<8h*V_
zP!Rz=DZ5^iT-@7N(O(d=+vTbZR*D=Nmtn^Oy4j$w43)Oapa{jT=%WDm@;<~n*-Ks6
z(D2zYIAViT?<wn{$>&L>Yk7nzD5R@`3o*<DN5GDFAI?}*NGlW6K@1sUJF^C}f?{I!
z+Qg~TH_UI7U4ktNDJS$Z*@5u}4*<iUSlQIrDE;VogZCo&w<lnG|5&=eja;{}I}X<U
zU6lkdGWcY-ht_&DptEP4kQM53YD@HVw!tWDZ0w}uWlva^e`VZz5M6gx%l-s;OPV7K
z(me=3k>Jv8DCdW7pMN+;eP|=O_R!AXAxBl(JDp?TjX_HN!$cVkRkngYC`I$F2$e!*
zb-E`J+L)kXeYj)F&sVCjlAQ-@s5gHV5h#9ACVP0Pusy)Yq1Qws>?saDe(^$<C3K`r
z>r;Zv9DDe5Soa1=)Lpw<M>o@<p^2W+{#ov1x3ck+<5F_h7<8Jp9?r@q!PKOr&MdT|
zYBgZXrC{EKgoKh@plm(ZWQv204F?6{e-}XmOxWj^FlY&#HeVu`B%0BF)V_i-&81ID
zO1hRU+{D5++N!!T322988G0u0-lZnSnvf)70~Z$;$Y3E)lr}eP47+)>q-ET?!uud;
zN{50vpF>`O_n$u|hmZ6PM#mYf-mc->)A{*XmP;d|hTi{>A}2$NTy<z``&L!<bOOLo
z5X@<UqGmgEs5l=w_pS=TgE@0I#bxLoF$pAY<%mFm*P)+#2L6h+UlOW|Ax={rsrY}5
z%bNM8);OdlUv)Sd6@o0#!!Fc9Bm21fsj~cm`$~tbV5l{HJ1rA<Y3;+9!gRY7#IYuw
zM>3ZEi5qCbqXVH^swuY+lqEGJzoW+_$>HgtJ2uQVA|ODnJ*L2XQ2nr><H#tbRKijY
znjggh#I21Al-tsOmjT2ML)f9AwrK^G?3dA7v&2(OnQ1C2ul)R&+o=4+w2{s&CoC+i
zqp+rVdk2b;Om8)znj%ePlN7oR5Of3*5`JFPpQaN4SSE07oihRj-jrtU!#`%pM>mmG
z*^bhOQ%Imn#clQYC`&%PgSnmj%}r>6$@|96R!~wQ1r;0G_dBo$ng@SFhn=G|U^qI5
z!_d5)Sz%)V*w=)K!BL((UDl+4PjVGv-UyA^^!WIA04WH}K?E8WRi|-l5p+d2`OYFO
zdZvA8^4FyKL!ToHDmN0&qXf0^;9cffY|8Eww8bZ9d)dkc55x6tP}mzU)NHyI4lt};
z+hw*S4A9GOfZPUlHPf8(>PY0qO{65TeM-up)`$2ni@CDAtnR!dIv!|kXNTlCtna`o
zLbn(rw}p)jsN6Cl=Cem{1=Jyt{&t_!2K&pKu)BkADSq7a&ee;9R>)@6G_a$vW}QYn
zc`Ko11r|_k$-4qcD0Cx-M-Ib;L5(#cIN5@nL)xR9M!H}C$c-HY6n7bo319IIo%}3r
z<ZC4i1k~F4pyLFB?qu5juV-iBJzxFKVq#`g(p;G7E6x?A@=sWT=n!yK`gx6LqjG48
zfLpx;Hq>dMyyv5~IGh72Y$SF)X*1pcbwmgo!clIBa<~A}ooIGc$!8M#1ZxxNfq^VJ
zglFf&cb{+_;ymn?$39t!uBw0)lty$P_onX-V#}(49fuZDj+*>J{i~n@1~!gZifPz8
z`H{#Q@EHLp<}|?2p%)=s%N3Rw^Q_I`)pq#Mwa%~OwF1`$SA$UdJC&g%U<tRrE`<B^
zyfd$e3<>#qxJ_@aBF*IOGtX)znsxwQ1G9g)N{iw&pk)HeL~tBVR7B*F&V6F&#1QKL
zWT6dtEaV8Sd~(oTihch4?^3&53^zJt=dEsR9oO_Bc_Cu2gh=Nyr%{NmTQM!N<OT`q
z4LGO-U7ew1H4;*}N$nV7w*`=$k0OCDAHG^8I(g~@HlZ{q(fjydpF2m+VGvMUYzN(&
z5v<x3n<KuU0k;n=4W(`e0*8~!*{8vpprus13+B3WaPYvNdjJazKQ1mDwb_H<8{+4~
zXZaPW5y+1KYS7&ZAp`UfxmG<tNmFG#|9h)^V*p}l=dESvIs%*>&5}693K`sCk>fBI
zTmpy!azSLr{Cht44=HCJ_I}fD5Yp(5G+rO>)Cr`p52@PSql9$?vSoM?nKOtx*&yJB
z-(YtgCx3q@j`fV2>`*_4JP|48BWCws^V7fY={u4{{ByPniRR&5oP;7v6U)Az;NNHg
z5=Z{ii(t-<ZxLBC$Zy4$_26u2`zHYB&O^za*sPA$H<lz6{ZUKQqg=WZ0oahEr-taG
ze~;?-FHPc5%X2Ew0xznIiHSMUR|>96L&+TMqzfrdt8Ht3-T<T`lfMdMzvWYgLosl3
zxnpoyIhQjJyGPARjekR?iv6gJesl2b844JC%Gm!wEMOZ!+xQXefdn#ty%!4nNKF3^
z^>}!OQK*_67XJ?y2y{9iy#WX?0U6p|=yU+=1df`1xDTlw)bWn0^H(yFQ-FUL{0|!~
zkUKzT@87-w&i^2Y3DN@b|G3Icr_%pTp`LHHKph%UqM{t(a2iBTxe#STg%=KJK)MLA
z`U(g!BTaBnwT9<YG>1;5n&G=?tAHd`VWLIz2V7zQLWR)L-d^d3virx)P5X<6i@Z=%
z1@Y@P=C*FgNT3ptx!J<Y!-E{N%4mkJkNjFSW3bQ=nma&)Uo!ra!1{W|+F3aMnFF0T
z+x$=rjD?kI@S_h(*meFC`u|w2k1c#+ZeWo1ikes|=lHxUoDjH_wehbl@R*}eX9+<0
zK<kbfKya~}RNonOh;bbe8%x1yq?Dk9jgRlG-`tI~H(|FXfE6PpB7*IKzPrX{*#6b|
zIoy;nj0+qPgeB_JUA6}SkSv9pg_QbxF<cgwT<9wxUI}gF9&|q4Y8#1q3E{^*nA^F|
zou)z7dPcn&=m@R-#ETqEV=(8Ad;ryM<ppq)NYFVwJuRBs1jlF}Jb;?ZQFR{LD*@Z4
zjD^q@@)V@6Fi-r^qo0sj0FVzwyE|>4Ha9l`7!L>tfId$CzyiosU|yh)f*}(+xC)^Z
zFvYUHx%sK21PMB#*4CkWNut!o+WP;n_ulbb_Wk>KrDa5s5m8CWmXVRsQW=#LnTf2-
zWQ0UX5h+_DN>o;4&xB+rGD1i~NH*EO<9u^n_jTR(^%=j%<NN*Z+v6U|`#j(0`I^t;
z`8<x}b?@$7E~5aPyq@<5Zog@~p^>Op@8v~5Orx}eg=wy;X$VJhBUPeegA6PoyT$Fz
z{zi`X{9-^aj3Yr4qGWr2{&ZfLbL^`C-Bm;GO#3S-|Jz=ko=eD`c44~(suFq*jhgw0
zXt?`!vj^va{-t6Af&n3W^)U!Sooe;!)k={v();#J;x*+K<>#zYyd)$cn8HDqwUJ?<
z-TyvD+#N#=W%^5z$Tx|OX;$C?j<+IqE>68(e{C7Ihs{DUN%SA(#PjFRWd>#aRI5(A
zxV#zl9swJox4WBY6U|9`Yc=*kVIiTqDhs?a@BEv--rjXJ_GglS;8}p+2|4Q(aB*xl
zGcc%)dvL2I{H_!z*u1cx=tm7J##_Aq0{2N@8e3agA_kwg?@|mtZa#++FfK>+NHQJ=
zrne(QbNA}>a#x_XtS}2TncH+2_H~6!>cZ5v`uGK(A&lasi^Sr=UAD!D`;jDqIsp%l
zH=tIni&``C{b@DG{yr>lL{moJp6YD2?ALj%e>)|63C|M?lZIS@1_4Ja3TU@&OPWg}
z#;j4kG$3n^`xQi|MwVtmp#aiBtqkXtsB0!q58<@lpz7<0OVewF!N&4cjd+e~UUp6n
z!~9)j-cLELBwpFj3S^S7E0HK5iP{urWr>4M1-R^x@xTeNJ}yo4DfmEs*i{muJ|(c9
z<U<_bHK_|#>_p5c@eINF(mQrM-TdisP|&VKCI%*^_4sZp@aaa6Sv<uEFfcNLZisaC
zC;~~@vDiJtW$*K^gCt(jrWUU@h?RlAX9BdOpAGRLR{}t)0j?vD7<|%>L_qiSxu;&b
zr4O+R7|`<>(_5>%$1@J*VI|ugaF=L69X0t=r=~_nRem*J5X33J-%(cV4@^P+5GS!8
zA2L8qU^pSfDHS>s`W)1i0qjs&W8~8u|FD1PWu)Og9IJhOq{>(ft)arPNik6DU{+4f
zd7P$tH;DtSmNqo_zp1DXCp>ey?(AIU{7Y+DtSmjceAAcE=7deJ$3Cb5u5kpZd@cyi
zjg3qhRFAo)4cv-trCyn4objGVu)TGlCc@-%RA-F1JSN;Y2xx+P+r;k{vm|=qdZO{}
z=Sho-ilX$H<EN2G02~gEhtuD;J2amA3&^I3q&<XrnBsLZ)$2<3$y<fPtyTr}WBSBX
zFp{3a<l__L=bK}g#VqI0Gg@8oNLf;)dW|~X#yH_C33x>05!JZHG(KlpKf4l%^b6m@
z#8T%;@83TQk+{l`?)PDw#0?2~tn1<ek;=p|x}*A|>?lxycDy-aLS+auzn2@b{qNw#
z<^kq|0ph}Lgl@@of#voZ>~@kB+?bgaKWm|{FpfA&=ub2!Ct+f6Db|^YOR)tZ>GD#y
zz^sXX;%+w94ghVbf&BBe?;vN06Huq^3XY6CMS5G1KT#D4yebd!*VoW6-=%w%crz;I
z4%WqZJINoG2V1NziC}&r?RVRF|4XrDB>i$d4Gkx0sxfNp+P6<FI}`ikqg48(m7xI9
z-FWJK@J_MNS2gd0n$t<vruV~b<Jp&DyLQC@#nn;VM1_qYK0;nYq?7WI5thrPU7Wq!
z;Y$ds@S@;|*U5lKYSoj0`Y@}ct+J&SW|N|hv3F)jHdZpu*LtzXR*`0$e!hk;f9zwN
zzH8MZFDXs%`nmJpvK#ZD6BOhPg+l)mV>{$tujd&yY0t|}$wnHLwDOop#e4ems2{^=
z+XR5jn3}})udrV`#X$b<q;F)TE+KAaBNd=qPvpDKwsr(o+u7S^lzB;#t|!_MyFxlr
zT9-Vq_4D<eM4!I+au_Es58LO)f(jEuyMw{8MTpd!csP*Blww^Ar~6(+D&)}h`3Iq%
z**R!`W(X<9U7JR$Sl$uQB?2xY8m4rb#TpsIy8GX%%9&wKJK5Pe4yl|uLmG&+SJbi{
zM)VD0Ip&`XVfW16E+G+$K;m<ID;zdrSSgql)4&~2$L`(DwrbTX;?Xj73tZ|X-!o`f
zMUobf4Ifrb@dl`-_2KwBZ9jYUy6D&N^|(Y6=`zD5)p{&{6%8jRr=iV$q&*Cm%38#$
zL#lfSs7m$M$HKzSCS6xiR~ubj-9Em-FWGd&GKFdYVh6=)A5YKiMCacoMR(>^sZN8<
zC~Z8<j%Y>NJlX<ri9zg(_g@DYYB6q61K`?+=s-=Dy@FkAC8@^`p-p^M6{5jtXo9}H
zCGl%7_U;m^EW=8MzICe<b#~p))Ji7UY#HXwDpa1+lqV@NLjYVIY)EZHTwTV!O9iA7
zry2xE+V~iqKr}JG(Q$)cziwYH=&rBPtmKxXDS|wI7J*=fRR?JUmKUqdf>32WEp+V-
zo6f#OiI`ACqwN@y#Q69Ii(ed{24~%Fuw~+s0AY53cf}g4sJ4kny3tJyLHtm`lQiDn
zU{>{T?@<2Kc_aowqpRnN8D8f`aIjO2QoT5C6>fKqML2`JA5v!-zY9uu<30i(CnWH)
zj)0mu<kUcE8FCU~$kGc*j!u60qHR5Sy5A2q@+(Ajw^yi&yGL;~y?4UmQDWj#ke<Zx
zB*d5|s5G6LTPSc?uILUYR1-_H!r*uK5g>v|E^|YtlFD;L?<)2I5(8)xvc{f==fGbV
z6eJQx40-opMBtEr^w|h%%wV)J_@xlIaw>6lW(LJ!$^HAah7g_M9znE^U-xhu#D+<h
z<Ayx*5+f_GqK@V6?vA6)5zZaFwbt?4wRNjkqdwe&>c?;7Z(NbQM>ls3&cvNFct8w5
zP;pUT`*!cr-w>n@u@!-zfJATt<7vzr^G$cho@pzVsZ6c<d;*omzrhaxzm@+0aF$5R
zh@Q4U$oyYGlO<$CA~d5l!5#mzu=@v0`paEP33@V7^!@kWEukio+?r_Izd@*FfD=Mu
zNu$&9SOFwKlK$5_fDVo2b1e#k9t4O2<(Y(P|G;IrRorW5RgvrQuCN7%^K<SIAGryy
zmcd+g2mW(XJKcXCs9BOoyO(hj(rAKSOVGg$Ts3wzxrhQCMhRgaJT@C5m|dsyQSY17
z;gt&EiNH!TI;eT7tNFen5@Y<_Fn1E{XjgsJpMF%WGso-~*|#_3cp8F(-7tFnK=RE@
z$oNU#MkJCq-h?C&Cx#(#H^lc#pCw-APiG(!8U7DBI^L6j?`RTIK^+n4=Rd9^>c$V(
z;KwDuB!I<LEdTz=KYvw(Qt}^Sc%nndtV)-!BZlICf0Ge=pW$#nKWpSn5dNW@(hue8
zH53H^wy;G(D?N=73Rv4!BN-=#EguR8WX-fNAd$}LsfiSLFR^1sCju@40?29*+&os^
z7l2w2@$YZ9ycS_&V>9)$dFKL=N#xla4=go?D5Scw@|Ll~`$$nOE7+odjPtGoxp*h$
z3rJ-_lwArs35ec<GV$P)=DKx%9~dAkL>zLf9X{05bRd3Mx6N4@bD7r1wD`^j^rV)u
zGMu39$(kj$V_Qdb{qhK1zIE5p6t8L8Ee}W?5;pRsz;&^A%jX+6ZXh{~{+ZL*If2{(
zu|>o@>_+k$fa2WQvzt+Y=o;MDm5o$djjIdh{zc-$*;NB4<mE?3N4eX!w_aD%_?PFA
z*qImwddl<J$51*P$x$(&=j0@GcpI1mw~L6_qPk8nGKi>CUms9(1fo(2aq*Wl*D!#Q
zGFZ^gm}p0Q1<s_Y7itPfxIxy*#LCKNH2n{y-AZ{iu5?iI_Dr`!s}4sHRx5#k1ok_%
zOU?}?i_dM~0fDrpf`^VdvRN*qXHV$o+p}lmSo}#Yq%?(@eM-z7u)qOo!V3q=m57AV
zBiP?gI0=I$2sax0ytelma^dPeTO~%(xki?w$PhpP+E-RE(;vS(fE=<1SN;Ge?4ewR
zJMV~)h}i1GuzRoIX79@%E~2csS_htUr4uV=a_K-sES5ne7lUkhI&Y%i{HYhPt6>O_
z7`HZsg@$UpILW$80yP20;jcH@92;AB-@%b-h=Z5+9;xs0hx0BJuIl>=>N?>d=9+Om
zAM3TRO8}u0J4t;0xR@%+{{RKB+SuRziZKER@ZD!_nuYyCd1S=MXw7P`s;XiZHQn=?
z$o;uEHrDe-$Hd6<^iA}A@*sW7^%360%d_Yn_+oX{?0a=i$zwdG0{$4~+zfs*tzD2r
z5;?UI`|BfL;#7IM3x}FhIf))L?{bRSOVncRs#BjE9hA&K3bRpN(F;psmz`RWtwe|K
zj>9mQ7%|prgfElE-7S>uPTC(*D+CJW*z-NO&O9fr1Tck4(k<X;okP3_=}cEmMuHfE
z%$(6Nj8X%7?;nIKQ~NE$$l>F}NRwoNG+C~$P}cUxM$%JMI)MljEvo9Za0W$96_q*!
z=5Pn7gSE@4Q#FVz*R5G|*T#!fUk&`$<jK>k(T#2ZI63#`CN>z-oW(-`HWS_e-F}QG
z_vCFNwYOn@-B9wMrPS`#(x>t0S8Q$53`-j=Q!0q+yI3-C4BTT%TPI4!Kp2VHxme@6
zK(~B3bWyZ|e@L6{tgZKs`Q6t4Sf!hqGiMNl6x5zYClu+Bizck#teQ5}pfqvS&Q6tW
z5=#-|;^rL~#wlJU11A&|{BJZT>m4>RF_Di<=9Te$z`Fma8kY*#+~dhbV2v#`<<1DO
zN)V)SvQ7@YnWZK@g>`;Ei)yr-+U<XqHb*-;Ie`K0r%E9Tx_)2MHfg88<<Vy@6uOp%
zjhA%{)G1;<HBU~kfsXvpN^wcFJi~SEn=O&MhQzNw{IJ^e(Bdn3hVm2BFSOE3^qM{B
zA=I+Z`w*uh%aRY4DHI1-^M9uf4oF|08UIuU3TipH9p258KWd-VUFSc9Iu(2T-0!Xj
zb&=FGG{(e))De(62HT)xxhRE$L2!duJ}aZo|3ewmSM|u%=v`)(9dD{hizdIreH2*u
z_t9}0f&)!_v5iK<H89vX_35^5HSb}Ab3?R++yYmPhTJ>aW5<qlcciZjClaMesh;nq
z&tbO>Ds6^v{Wk|O%o5UvRH@b;HZ_e4B~&{Q5)Y+H8PiXgQlY*Ja?<@wv&dc~Lnx-s
zuo)N|_kssX8q9`TR;VjI3Pqs2MDb1B-Gqp5G+Y(7Kza?k_(^M7=ACyEx5s^6f(is&
zORNyRLP=8?|E5MgA0BH(JwBc(k(-qzwJ0qhS<MY9QqaN=a572PHi_3D<K=5<F2Y2G
zM4L3}Y5knA?(pqRMX-rtA*fGvY!R+H$KYm!v9f0|3Y?LNU#vu$*5PrGn?j62(3|;5
zEuxc=C!7IGp$aq+&{_L$uSbzlojTR)xJeziBN#MG?;6^R<i5he7znZPMTl6itEpO{
z%uteA0qP#OMH=a*LF0^|CoFY`kt4C^jDHngx6T9G9dgg8oyK2<;+Ix<nS5JOUe4vA
z5d&;lAqy)hQE%C<rFBs85(vt;DA=t6Kxz$X$4hQUE8IP$x=R>o-CZq*aigHHCT!2<
zj{E@{${nXgk!vE&OZY&6Hxa#V52Lgp8!;mqjq*|Uc;g?6$se+(|2WU!?DTX{aYEFa
zx1UpJB9=A+6_kL9YK-~Zb}C{LR@j*0A?g^P{<U}rJ1nH$KNlAtp3a1xk#Lc7$X)9Q
zDo|vMmH+L9kq54%LBN~@5tFD@yY}7cuZ6d&n?Ej#f1_4I{QHqVwF1=h?emEUkljdA
zFL8w6zGKxse(VZ@deW@n@AtuYexoUI`RiNZOchxl+cMuBh96P~jh~*1l&O&Y{^N%O
zYT=|;7A+ZrD#)xMZWrr&aB#Mg-dZt#nLX>j6&g@$YL%+pX!tt+`d6=Bp_}-uG`fLD
z>x7TAU!T%hYPt6|NNvPi1hn*_QH((Xm9#||swB-+&|!>YaK&03z-XStLLXiFTx*is
ziPbO&&J}$$%{^P#Nt21kFWSbd*H7-3y*pTo?RD~+>~EC+Rl8eO;Q}t6oso=DMvj4h
z&z?P28YR#kkwwk%V1dXiET$m1_vZ;VnM`h_asIc%_TxWlV;*XQ5eDq!Gi+W-SaSI?
zxGd}@hBUTcq*uD*+Of#T5A|Jp1U*0U6hY81fQTYNiRdZdmHYyze^Avz34!|p7OCg9
zgC6>TY40H?l{?X}F1!zMZAB1|W><4?>h?pLuMQ#H(<g8T?8`o+hx6|5L(kODO64F2
z1kTzvT3R6t8dujvv37@@cngF<5Te0|2NB1|-Kaj1;(LrHpaZ_KDvXw@RhX^#&;S1j
zmbeSy{)nhec-d3*ZxK0(AZm^(%zp;!fBb4Ge1`L2C<a~V8sd)2&PM-5=*}0pL^3UF
zVUORAqSZ37blH#TZv@o8{oBig6bdtWuuRR2YYE*afSBfa%KOL`5OmPxdJ>U5So)E$
zN=1G{kn$gXM=CuLAAKtf2!Rg}(LM$%@DxG!{M(Osbu>AFOa6d#U@Q<c(!c(Q65Y~U
z6W9Y}Rg$C>Amy(=k{~nj+y8h}(&Lmwq$t?iPog?X<R?(~#9sFwvQUEGqubASn26B8
zc+17A30BiQWPzw~Gp<d?5E|>P=_qjR;GZ@7D`EAJ)_qS;WFor|vn*n7T-G)qe!!qc
z^h!?sI6OSU{w<OWUm#jQg&meG1X&KPoY5&ECPu8WfZ3R=e}^Ke4;GN7Tc)R$oa1Xh
zOrobEKlOSj0ePEc0_l-Bf{Tf!O)Ry64FK{MFi^RL>Q`{tV<bP|z1M<_<@}_WwDk2a
zS=Mm3oQnf_-RA4tm1wumhJ*`(C5p$~f!Zh#kxJU6dqF3}Bjtg7RZ42=#OIF65P^{K
z@2?3eD)9uV3I6{6V5_U<+$D;;KLjQ;gL(=6o`cN9kJ>KCQ)c>;#9hr1LeN!OU%m`F
z*B!yZzm~2H=yS^n*CwP--^{$s$<7A%?nJowUP;Lr6tF<JyX)<(mFF;G9C;mfdP6^d
z5_!3g=%RoGIPwtXJ>+fO@7y6Iuc(mIvC-4f2_x8nf?d%$FuvpH<;%(UXHa}_f+*!x
z(SOvxHaEn&Znv?vo}QoTAVdP|2*0KSZ<9z^glAz@A^^GE(WCZ1KU-swAO<QF*AV!x
z-%*rAh_L3YB%zKnMP7L(*OMTB#1K+mzxD#SG_1c^5^47*SCvS1vl;|pP5p~LRDCi`
z6?}<9`NIhxxDlVhUO?*Vukg??Dd*qH>Y?K&&tIQV4JeQHUf_OQpbWZ7%@@MH3ve)K
zA3h?81dr`wRUu*Ff%@2;ItMIog^d$LKrF1)0E_hiqY61e;=RN;7&n4u_vQQo5;hu6
zt)x<x<#Uv>)Q3t1Fa!eS)d>5kan?Tc+K7D$1=1<AUma19tqThU6KZmng@uJDe-I=v
z;mqrEzhA>q@rstLtSqkmAr_B93L%zpWLmd@**!CW7z!;LY7ki0SqKf~joqJ|&DASj
z+s70kbG>D*8eIRF$O%Gb*O9<Jx%+zT6#$RC#|ejAW1Auzg4MFF4<u%$b#xGAtI}Zr
z?j9>PCZ;`YlB-c>nKudofaQA739Rk|FpNPzc+a@}2SHbwUjR8wSMttZDXE0W^MwZC
z6IHjDuyw*}j@6`bRZZO>83ZEmIzgou<~e6`c8Q(X9q{7Ciz}_bRZExc0>*J%v1IUs
zKo=`dlJto4{ES^+HSq5-#`!_ir$#N{#MV%?0jl)qZBu-|-8G0gB_5CZqSWBGc#_gH
z8u?D8P$9W0Z_gh~N>5s0HWn68u=Vpj90)ocmRw3x0yKq5V3P#N-Pd<}zD;L2)Hysn
z@z~ZY5|Q^@ZnWzHk{FXku8ic``D^{6L7tS5l+--Jywr&2Ql(H(Rd<TGFFux%V6Bl}
zK(>zh?3-x_r3m-MlDq}5n!(^vv4lo82=$agQXB2sx%05(jlTgQjSJs=LJRJ2Ecq~G
zye85r`2nsF7)kb9IXUHwAJ@25C<r6y^bxN>cBwq%5>@g^v1W>wOR56Ao@QE0K7Rb@
zFO!poZ%4j$d2LC5KL!n}j*d={6_H&ik(E1sYd+tm_oTA2?@ZxQ)XmXKlMo+IPJ*oa
z6w_&(dE@~EZZJXif^Q^&PFdgnvzGI&vh2Im;r$KJ4-%8hHTK+osowA3)6AMqI<a(I
zpL%?9HtY|XXe;BgOjKbju@e4}j^9dXIqR3mqBG>Ws}?}Fe%UMcee3eQc1lqWF-til
zkFz1OZrj#TUtjO*yRq-6Q98<FkRe>?3Aja+6h{(BE>HgzPn1~uDQh&pIuaZ~785A1
zK_6avu>UN55#vcH{-rA!cOY_45|uwo-@QA|VozA8f<3G2z;8q32%`(mg8@d%r=0Y7
z!N}Oyb-eaif7hk|j&d5VN$4UAG`f+lIVE+kM^^GnY3Z>RweQ<VFWZ?-<p_Ls<??0a
zg+6#9aPHpS|C%rss#BRhgf2o^xT>s7xxfWkE>kUr7m>ky+Yr(jF64Wa_5XzyDoC)<
ze+Wp0(AOZCMcMf5f(~4wt<OLIw19Mxy5v2O`5k+FsQzC3O7UV#RRAuVVX_};Z<~`w
z-SLuWHlP=2<D{%HEE>aNSkfFk`T*%}Bs`A2AcoLVAuJ^H!9l->NOqB{zqE;t#hL9r
z4W-G^Qi>#y_R@+~f63snaUmDrux`~(m<l(l>%23X-vFKOeZncc`b!3=g5dd8SYS+B
z#15DfdqiadJI9Zp`<;YNFU<qFGsSDkWs-`-<^Z**^Wf*#hc%BEUJxYoPBVi^;(Y;X
zPbMZNz^FWOGC(sK#ge7x`3(S9Do8SKQtn^^<~_XB=*>n_@wff}dpc8u8nbTYpY<>S
zzBUP19xdJCWx++~g@n<sg=DyNbTR#E&%x7Q`dAnb@|Td^-o#kT9`!?kJJ`r)b&^+t
z`D%rb34Q5}caMF1SeNziiF(ZIBPB&XFW0j*Cp_miJzh^sP8$_<*Z7{sPD?4bm1k+~
z7>}$oFq&+dc;?kVv#<1kQCjY_l&gk#+wkP9bL7+~;ZJSjjIy#UvZxkQKB+_MHB5G8
zZ}ZgD=hqjDR-$}q@#~~}>*UcRN4^@X%Xue^g1NK4^fCe+zI_&n=A(Xl%vP#BJwNv)
z8u?ZCb%~(qU>oUwQc!W~GCYgpdww1HU7-N)NtZjSX=U81s;Wg(Rp%1aYZIi2O6_Ph
z_?}goH!!eVT-LhC*#CTpJ!u8uP+5D-m5`rqeG3Dno?iwZpBx;OM!93@G1&W(T<7EG
zn84r=qqKA1Klp%I)GK`jrqP_am27WsFC~p~Flb3}VGG<t=Yt|5+7qN7D#7qZvV3YI
z1H+w2J$j!Mui@C2w)!gS>L%S86|FzP!VX@(#W`6*@(8q5w16Vnix&%ok>;M{>}bER
zFrJ^CgP9%gYm&W=YLw`|9v^5r<-$Z5^0`kUj9Je}zHi~(;l@>X-{CQ<dVs?qnm3LZ
zMH5XO5TSM+A@ZG#Rx<kU)_3kPG&b%+vp!20Em(AQzLU%jrEb_4g=yPl#E^c2$rSS?
zHIrjw*vW3!h;r=OwW}|spyFE1+BIu597hgmYCcMaq(t|`i48>SR&qBqG+;yU-&T%d
zsC?+hvfao9ZnE6e_v_dBvuBTF6mz4iTgz(*+XaY_-P;HqyVIbRqhoevW~Opd+T!jZ
z%Gb!~Vl3ztgKbJY+b2BR%o|7ek&nVkmut%c(kqB>>DVrsj#;I@X1GA#UsB^1cq<ph
z`&7+n5V%f^p9`(>$4(v^9IU3Q8sj_<+G0_0anQ7s`Dej^a(6U-Vyaj4=g;chJM3(1
zKmvaxs8@uH9yIe4L>mnTYRbnKo9_!CS^y!;w1xwJr<GsguUErVGcZH0a#Dmx={8XP
z5v)gFjsOV+nooT}rocdUNKuFd<@Xk~+I8_**0ICCSQGC8e-nCodV~GMPQs}6+R@eL
zXcy=!TH3L;Z<^i0$~y6?=?N<Fs8VnhT<q|L1Qoew4IP~mM6-oB>Cx|>$c&5Q=Hil{
zrVrdb0jCWpgxAOxbI)9)mzQ0fvw>;H<zCZ+2+DlCy;BEgy%N@_-*y@iPN?(s^7e-7
z3)KCO5<bA&1k6q`u_pc3#86JccY?KgN_O71ImHk;XD0l2+rluhBqgc8{uT)@YjpoM
z7@saZNMuZ;PsW*iU8RO2u84GO{U3&eg!uXG=pF;L0Zt9f3=I2FRzak4{ONTlRk-qr
zM^I3|ZSFw?bl;3;ywU#Zx6=<~1PuNrP*uXJ_+*AMc-rRw`;S{>F#^`8>G2_0FCj4B
z@`Zp`fBo@)b)Uvn2}PPu3hKQh3D+U+WD2vJ=75s_`f;V~|JnOk@fg!{&R4)B=T#NZ
zqGX?rijUodLi#}NXSE-_Y7M$`U*hKOY2J0ABq8keg`|3(o9YD;Nwn{Q>b;o;MS=XY
z=XxMky61E*K8{g#puL8Yf%j!RY>v0e4k#hYO5hP@XOB8M#4KW>tdq~h#Wlm<$Od<(
z1ND`5aCaKmJq&5*eB89;_B<Gzf=e6R2_EarZPj5o{2}bcc|&^iqoG)Lk{Nyoq*BR#
z(rpw-a1(nOg4U@Lnp1MsC`hJge{z0~D*i17;sj;9?_j-R#wT5)EiTUs5_HwqSSbrM
zL$GOxA-^M1gVNzv`i)5Z6uOY&qDTLq--;1HiCf{y+vYFMr$97XNqWX2ZU>m3+Rs$-
z7Ne`3v3#vHTPZ92Prt-Ne1!{NYv0h=BSWOb=E1^%{T<<T1@B^Hg5Yd0hD<Xeid1!8
zJeS@)>F%PB27#TOHJ)O0{&-R#CnpCXr32jd7k`(^QmpCu{{6?LV0ll_&xYV~?i&lp
zijWFfN>+@*)15QdkzK7UI02T)4b_L>zBQT4UZ1f=y)ekN7=C7HpF2QBl;GjvL4d-`
zn}A}C1#=^kp%Z&6xw*N6`E~k#{`~Tu!VV53($Z}whhD$s{%+%~s-e-`lX(dtTrx#!
zn3G-T;L&G$1qGAa3l4fG)WM5$AmGHd1Z)$Z6XMLUhyWGu#Pi(5NHG$di0i0<kI07k
z4Vt7fLmIal7eD_#CaMXwUJ$lH8yOfFfbHxMC1fvPj}yI^Op#FZ^Z9wv*=ekK8#V~3
z%gv`&!Qm$F>5)JrG`fL_t&iK;O#sA!ME1SEpF2;SpPw4vc(#%o`FwJd`#3vfRlHzs
z>^b8EcAsPv4W>0R0D2bGOIllJPr1%~f}17~k==W4P&PMnkB*E;K&NmxiSU*KC7;JN
zAEu<BEMY>+Jx&1s<yjkgT%9u6>qHv7uyS;JzkS0S?{TJ6B}v&~m=pZ);lrRHJ6xT!
z=)aK2sYYmn0AThAyue0@t~PUyPf1C^_E4GWb?=@#nul0Bz>B1qN2bPb+Hfair~WQ4
z$<#1SF8wI5ea}sMkGwlUC%5s@hM2h9nQAV-JJ<cCWh-h_R#CA@W&OGeTbSgW9L;SB
zb)MtzxS=;hfP3B>uF#|3?~1l+@-j>tyZW~~kDd72)pg_!Z33pP8G|SapP98$m*YuQ
zG7VK$JaouyBFqZ1jyD$2xA+|0i4~CO6?DGpx89<nL(>!6iK#s+KwM&u-R=+{;^l&F
zWdlRQLZbVY+J1w}zOZT}IOAveP!7Y$NcQS63U!W5_n)d|CU0^sdH0Ug*?)W*=FUNj
z!AK93Ei$UrR`7y#7m>FS+1;LB2eTRylah#~c=<shb0e+3WbJTGZbR|mHsf1v%}_%X
zv+uWoe}0LjcyUZmw2_|UEq9e=|Mf|i40T?lj^kQg(Gisu6&L7zKBQpVt)PUngXu3e
zsCI8ZgP7(Z8t62Ayz@v2BT6+$Vo_631+w$th$$&4=W_pD0Q*Ry63R=`5Pkf*0()uq
z>>_N#f|72krU_L>XEB(b_BY-s6jnWY^|NU@e7BbFDH8kj>sO?R%WIaV(Od4ihdAQl
zWve6D)+PG5xRM)FmClqv#uH(#nDz>l`S2gWdqnZ6#Wo<bZ9e;Qz54Cu6ZtjmHMGI^
zSoddG*3^+tz!GHO`mThiAdTnlK0tEn1u#*p3<?gDxh{z&Gx1Y3El5c>u3;lx+hQ1m
zUA%lM+~2?MBM)MYDxtfwx(=Od9Qlxc_mR>)@NhkdYYLUfQEeh7RS7p>5PGoAD)2)F
zwA}PeOrOzqccic&*<!6A-4)Zj4v6WQ2eN{w)?J`ym7>`rA(3rpl~_Of+$$RUbuo8x
zg^e*ixdh>>gAfdBok*H568>#4CyuF(UL8P7Lz)mh>bqcf<I-*oTHDd{RPiPz!AO{2
zb95v>wPOxWSj-&+^Lx}-t!2*sax%sA7yT_h?~%i7%fHt2h*1_qM;MJ;+uCjpXV|o9
zC21YdN+7&y#RmNLCVOQV5ioSrVi>BV7S7WY=qlj2ib~i1&8(~&Fv40YdI>9`L$=*P
zA4r=^VzPpUMrgP+G?Aaa+c@aaTT}~mYJIZn2c;hZQxq#;cxnd{2ZF+yFtH@0P)D2K
zvr{g+HBbcXR*ZmJZ5aX0wBv%Uh^k22j;3A^+>Xj+S<B!IcZE(}&L&lfUFELQqieP9
zaO=b3MmVr9cA}7kl|s&Dn)k6A5{Kk2W<OMpk(xZNViMzNy4|QI!lGI&*^9wgLtefd
zC^vB(jDeJul`p(j*Vfj4G_h&NG*$dd;YeBdO&3~W*(YtUap8pSZo8~u@WMQB84+s{
zMq;4UeQIjTlvlp%<rSAN?0D%??65AzJuwVRoP&A!oXreDk|MQBP^#B(Pz!A86&vNV
zt`_243wK74JhrKw9^?hd0lSxao&$2C$?@@KZ300plHmVfure|*H0Fzgyd}<zBo?tz
z`1GV+CLePrr@KJx$lS*CmsV%KegIA!b-amwaA<W8R!$<V44dHgUM^m(3?-L6C?XH!
zJFUkE74fSnvMGKR6-Bz5c+d%reV4x;n0^XA>A6HAg2>KJ5OSGwAVmYjlyC~>+QylN
zd@#{9mNJwNd?596C!!9-US7|Qo8eRby|l2Hxa%t7D<0KgJ9g{f@F*tY0M!rN@MAb-
z<%inb_C=PX57_wfz9M;$?A1tg!ri&BOn(M`?UzOffWO;x5eC0&3gq?H5OIO3M0}dM
zcIp>qqh=PXydARZkrS=E#6>)nNaPT7pXp}jIt$h7^gaG~oTJ9gAmF-}=9iR~4m8vx
zE0~oTPjAREBW&8%;u{c26Z@^^6R=8|sn$+!8Xg}f2#46cy0eoykj-zpW_`U&#bu`J
z>(^oYqL^3}Z=>BOs!|X=0}qOA83m!ayu7@^gVz`U^)v9>oko;Vk6*a(73{o3ckFoI
z3g&R|_zSGtFW+Gm1G2rT%B+I$AYpB^bCLrLbvESx#(Mndp<#NB!Y>3m&dzfLx7=9?
zGva!m2cD05{d#35b0`Y<!~=GrXZF8V7?f$n=~!Uqpuz@d#Yj(2&cMgVS6xxz<>LbY
zbf%>W`mch;c>&U*t1j0)pYJlCbK=#9jE4E65gX#my1#wHIcpD)DFcOof=1qNAie2)
zd_qEfCcc3$#cMmc4p1`uom&Yf$-!2!6eCa~`E}nUu>Gs~nDyz+IP5aw;uNL;7j55i
ztHmK9f{iV1R2Z@O4PwKNF--V2G!%*AxB>#sw6$tjW$t-<=fYmO1^<u?xQDKYZ9s(t
zWFz<Y4e`jZShb`<8ka$Q6|-gm^tDQ-P7(N7)aX^v>V9M<gZ7<$fd>Ty(C0)3V2fc~
zY;3KzD<6Y*Ed!3BVSZ$2luE4Vn1zV#d_U}kW=G7R%oj8-{DHv%2U=^l*w)>VdO^m%
zHFdPXAH?yNnaHOiBLKPyyc|Jyr3OmE3g%x7p9sH@kgHwqH-t%T*-9+A0*I9pErJWV
zzYDwUZY3V&pCa)-fdNLfLKc7$J^Ft|-v1OL-$G-}DFKlstXU42|BQ+M8pJcAN&d?~
zoD}}A#d7@ne;59L{deH-$Kx)e0YvP=7_tOfTgC>`ua}H#^4BauoF)4jHxglSb$prX
zGhJ8@ynIy^7#PT$AFtZbN->`Y%1@3EPV=!F>Qz$}aM>NHbt`f@LZTzL=UPwBl#N-g
zrQ5)bVd3d2y93D-fG`I82~4RdLs~Y<?Nc5D^5QCN0fm~|Q4R#48QnrOsv__v3GQL7
zCZUINd-XN40hnhfofzzIK1sk&fBXorlkynQ1U*m=_-RFeQxe`JVUmCNN>;VPn(x9`
zUy>QrDgWtt=81_&dLB~0lI}`?rVP@!z@;R-bxtoMOzAleQowq$P0Haif)y;uO7im4
zs09+vwkTq&si~3RlB$IapvU-&NEb=$#()^e=whQs;_?ILDUMC6)(5!2BDAYDJz@|m
zE4(Py-g<iCng7Ba1c1kn7XxMW_gNVk^;Bkj!dfg6QbQXBP8tNXVg_+w5TdGs-C!MQ
z@HW3HX`V-ZkgFSo$QJcP<TGJ8h1QFYE<j9edU3xGO6iBEwNIZW{QR)Sm?;b(frs4w
z<MxX;39O`;(R#_gBhT?uO-+)a-y7J{;qrk8s_*Z!#s@j`^YZ@0*)KMf_q@E|J+q0l
z4yEYOrOsk{Z&>sA)~&-xN+XEG^3q<W#Sfy|js&+zzza;-M`gs)c}FG<GO=+u^-WAz
zn2g+UcL!gCY5DO-n=M|eBLy^Yooo(XWyIkz*9=dbs747DzUK8$53lPiL2R}-(^P<Y
zHRn7V797m*QAPfuj>|C01-j?YzglqK7pg>SwP%wdAR(Ly?6#`t=%9vWeQ88NQPsg(
zTYKyBQ+KCcL&dX5&%EhX8|s+MvJY60=}&Gz;DGK=s9iW;;L=uDc+h1SOBl9R0wI|L
zsM&xBa_2mDvJT`O&p8Z@BWFtrWy|eFjdQ<9q$|kFU%C>Hgr&o^Ym_VzQjWA3c$z<l
zpr1CD6>b%XbOj^=-;u&<9YJ;-NyPCaIf!vOxP^phms_I6yufkn`DH3y=*Lr=&8sue
zq6`^2X(pS{o}x{<0p;Vl9rGt$z5{2OwVE0H46H$*?q=Kwd{uo1g-rT(&FM~$Z6xgJ
zKiFSIgh-uCa3$8t9z+*HuGQTNHU$M9?a{yx<RzN>Mk1|b6g|7Ux^^z6q6&ty3lf^u
z*k2Q<F$12#yO-PTaWjATdHI>iR;$DgXK?FBEl?Gui2rI_JuwBMUFuvi%<-D%)Gkey
z_2ROlf=d4hqzGJ_qI0|11vTkTrhU_Tnc64^*ymhj!0Sc&ph4VcUqF#<!XAm^A~~3z
z<HVAgYG?rX%p-z=))1j=h&d)A(UL`bp=?f>2m}94Eo(#uH%M>L?Y0vkPW9P6s^LOq
zrh57`^2qBWJ6rneI7uO+L9RvWJT|8yi(N7^KX&qDrmOQ7m3Cv2xTy}OX3Wgab}XL%
zT~Ugb67{QrW<FrC8{tWH+pWv%jze*=d=tr3X%UgLE>t*&=2G_^MEslicV$7>PxMi}
zZudDE<Pze7oN+$c&C6QR9cA49Aj56=^XVZKB=@ie|Mi3=9&rS!!p}nt8<HuI;&$rt
z5p%z2dAL2}s!Yw}{pWrwT(+6|3YVWifyXzNSB!?xK^znixa}y8xWI#ojn3divj;3<
zC$q>JurXqWBs^WfN<<;1LP<^Cs4PQR00}A~f}aQrt2P4VNRI~e8jq!UA1~aB)ogpX
zo_wzfpzN7;Hif9b`uekY+s?AZu5F7SyB)<`8u!NUu`<en7t4n3F{I1f3l`z$)@ZeW
zgDSZArU=gJ(sHI6HVDma7^=ymJ1Sx5f}-SBq%84aB%7OIY|<%OM+2o@LM2jL6V=>s
zDGAv&2#veQl;oq{uo{KW(8J%E2ounKqkSnG>2Xr5wyRY&Vi2suhY~i=y}cD#c*ehZ
z3n=tHRcq&@!$^+n2aRmrs#EVa4ykIXF!nBCcQ&qSXh@Hae)q@%37?>vF@F(cIF|cw
zCRL{*2fPY$Mj;+iRO0gTq<A8?uEtnSpAg%z<Dhs8ma{r-8Ax$5e<lVwb)5AhPJLaz
z`cdjU!4Kbk!qi{oyZ_0749hmt2bpSUcXQc>`ug>mYhfIaHdss%c|r!sR|2<1wz|MO
z)yK+i^t*hgT7E`8>+)*U#@4)md{K8vSb9hDU~_73vR*L`_`wOk;Jpvw2npv<oKRI=
zBafxYToz)j<Kp!oZ+tS5L6piJVgb`dJG=DJ5~EZJQBhUL=~It!W|$t4c=8{_+1^}l
z+_kzDaY~kwf`ZfU;isb$;Oe3DwduUT<L5eAmsb+Yz>lPHmfM{!P*ekkl<1#+r7I8e
z*L@$KwmipFTQ-86zl7^fNw)bGyob#gVz$vpRJuv|8-eCmMdpi`IQ(1RqvR~+`0UR3
zxNYYrcCNJ+6X9yuM_NzO2ColEmA)nW_SH0RqS{@h7LhYp&oOK<33BH=z|wJB9*umh
zuv+7EPUynf21ss~KC(c@f@kX@_B%vE8r}1-KggH1Y@Bf(%Z`pdlC?V58oX=pUeM>-
z;~~DjE0(93R)zpzlK34M>%|1MKC13P{aoV>aW{(zX)WTZ<Z*z>=q-2{ok4!<7O{U|
z?LOQ7;xx$Od-^!!TB6g^&i+oqp-9UeAD5yd8BgnPfg>%dTZ-|nw;6tR6l@~~;NTGa
ze%x=(WfshzI5ll9!c2pdl*7n37>ArY*Oa5%f`kDeE!^+2ajNmJ5*XJP)r#BnP%SU6
zs?7-A?K)Su;P}GomRq)mTCnL&vjIb`$Ea+W6oVQtt7r;@Vy9~3)MjlpHG4=M@Vxcc
z<T8dc;&OAZ<)5Qjo`aDWGq&xwR}#OGot<6XR*oz+N5|nKM^<BbYG@`rQprtNX=&>o
zjX6RAJNj%B8{5K{c_Q<OFpK~AH2DW`I~%?{UazY`zF`Bfa;oGk;lk9p9!kpihMv?>
z>qFR<?)=$f|5k^a6hBQ|R85aFI%V@fMF&~WG<{mi_`}vtlh6INKFhl%i_t9hA9u2c
z|6YDui&v(fBn<kCI#Db!OP|(OQDI-~N!(CyY3)+?W?E!)pynoiqk5MQgjsqtRt59>
z@P4YFJFd1n9uv<)t;jCE+PE{68S6I!tmW^tGa?RgxvcRH{gKe&jGjD13EC5}@n<Z+
zgSq~K;QCAr%k*jIo~kXK&p#nv;YNXE6|(9Exhj9u)?_n5JJ2JTF<NbEl<5C`^&l_p
z-bx&bm-fN=jl)RP4GS!O@NMB)?%?!h81era9OT=2^wMHbP>_#R7Ak4o7rP<1eO0@(
z7J~aTN=6u^wy<RO?~H`}<1xy2L#$`N`+wL8^UWX{jSp2-XCEM1MRK0Ab)W(9OW?5-
zd=TnYAg79^NO&&cj75YmpA9}I>p>}jr!yUZb6hM!(7CrK9||5kHLX3H@SU&<EG>mJ
zzZXvsyf_46Q_MD^Q2+aFrdn6&11s7q%Rz)`*(Bh1<29+6Gs!4n-a|kHT@LDb*Smwx
zQLVk@uVUJ9^o`v>uzgMLNa}qJx#@$wm>d(O+*h)*vOo`gMra_560R0D_5+uijk2Cy
zo=^%YZ6+f3w@Rd(Q0(sMX+)i^ko))amT?rn)(>$WKRmZZDq(m!#Y<E(`NF|`tMH;f
z_ZSsZR6Dr}-0Gsio@ziGZZea{B0KO<^<=pCj;#$?+xvJU!z#HK78Vc;_Kz=j;k5v6
zetxi{vM+}o$cKCY=6k4Nr%!tzTk4YW)V)RwwvNO(&yTyumM|=eJB{B{20IECK?S&X
z^qJ`&6U8z{Y4<1aS5Tcp!7e)BBeGc=u#NHZnxl1ahU4$?Pp1esCj_TqPH2~TZ5<tp
z!19xkPA_*vRvV2GRy?A)pl!oW21s?o79j()75>0+N?uVB=l8|U{PAV~CHMhMp~h_a
zH-NPSXo!Lm2{t*S8jYI5|A4ptJM6<rAT?msEy#HMcNmKJO3cDOfq#LxC1|F`vIH@k
zk<KN|QkSY|JJi=eBX?-&pEfQTrmp@g7EKXb`ib}yA*cIq|HoHm+;wn*Zk)o<(!j>{
zilLqb@i!}dQwm{0p}pG$w-bMrk`g%Uc-c@u{gl3`;YC}4Gj{s6#Lw~;dd7wV=Pnu8
znrs&n5fnIXc*)qr7Jn5IICRO@M%nPnVJnNvR+fgAw%f%84qKU9T{&}E@1mi=5kvb+
z7Y&bH(Q_0~GPE?t<wztUVNU!JIC{z4*6@nJQFA?8!y|?ltqcrhWGI%O&7ozmaPP_$
zD^?UeS%@`WXY~Ys{eOS{Ukm)L1!Ty6JRd*(k0V;-Un6>X;0a0WmH1x`ynO9SOCB-W
zwppHWvJkzI6K@4ClwY~=bN8^NVEjRURfd2?U8Rv`gV$Or-h;Wm7O%_oU#lmS_C-zi
zCDF;w&s<r+A(Lm~0t`>yqJNPjd+G*__`>S@K1U;YR|wOd)OC$qn6Jnd{4LRS%$k3f
z3-7|<aHh-bjzT?}w^l2xWE0#zF4D;|I#b8$Zap`9l!94yRb*Z%XI(+h`K_{RuQ7-H
z&}aEtb=h&YrpZb`O5o<GS%Q9<YGqQR75i}1WtZD$(>A!SV>35Vz0G3aZub31^Oq^%
z>UVkM0k?Gv{H>ajLKb~^k5_YCeF2nD%eIGHNz$73cm2R8S=aCd=V`{67E4dEFB>iI
zHMZSZQL3ZD`E1Q6h2S;)$p=Cn-mlUvo~p_!dOA@m>-zY3<&e+s&7~HNfjM6Xiqs-*
zGCTO)TJgFweDhJAItGsdYpttey`p2{bfKH*Eqr9Z-@KVg=KA@Sq1p|_mf_#CJ4fOV
zA67{^*iZ4$J}$^<>!v+-=jQ#@wzlq&ri|C^cq;iRhyBa`;N5ffGuEOs!UI|F<9^9Z
zeG0ugr`2b)p+{n(RD<cp^M0B;j+ZRsW&59$?DOTkAZ2!>X>|7Ru_OD7qjk-L^D;k&
zDV%)S_NA^SbK4z{6CV#&C+WRvi{sDIS13;QTG;a|YCU^d=*HEe2R{m}bb9i8L`rU@
z(iT}|ZL!HiCPpP$qN;QK7YjZIUHFlDe{=5I-Q#}xO^+_U!rv#W`aMc4KW2VAtxcc1
zUccZMV{Y*+J{tPmF>{es!unJG8m=!t=U>#8=C&_i^PcY3=UO+>D8}aGH8sT}V%nT}
z-bV{A<-K&FOIdU{*3s!8Xf@QF`Od1spo=11&RBN*d{*4fXXOd=XYVtZUvHm@Rglg&
zv@dU8n00!Bct+@E2j=z9zR!l;@6-|crNzU*Cdf^eqrc!{yqV4Ypi*s!v%MnEPot1<
zr>|3rI+njqI|pPu_7s@z%N`K8@<U~n?-={~iFOJh^TXrLAxD#quJg<o`$cxQO33ya
zKBMD2I3F$MSbUMD)#H@hQ^CA6#j|Wp_WXJZ8%}$v#0;-ap~y~Pi%lQiS`ehHIGy)~
zc2)N<)q@{2yS|(4d?ESFOiPDFrlIPI>VQb(FYzh`jjOigX*XU)E1YU-t~evFMKL9(
zrQn+@w($(N!zFpLy^MjA=eNGURa@72>~3*RFs+B9i!e1=%{y8bvE2Gc+0`ce97+zq
z_5{(MZZozHkiC0nGbihw()@O_da-GNk&CjG?mJiU)vnCC`K@%dT;;uBvJ%Rto_b2b
z2_H7+1_wu#NW4|{`f~PIy2(NE#xMGOw_+6MqePWXzUVd7@9;iqUHZ<yYOp~u=A$9i
z5t*Kb6mx;6&dHk0#c!z&Jv?2R((V4uT~qV6<2I_zob`1Y&p(FJd3kFxWHM-1^QH6F
z9v7Kueo2#?dMhbfsX}Y_?O)9lO2^MUzqG|iDR4~B|Al2lk2pu<v;HnqZ}Ve4&JN|b
zDPySoU+$)k-J$j^am3c_&}JSP|FoP8xoL}trtDCLYo0HCwv@EzEL7e2^*KAOCTDw9
z+M~LhkzH!14u0|3z`Re|G?b_FTiKb~bBQe%QXfAJJ;hr{uUwt~0Ni*@ACIOC^~#GY
zXC-^wI<u_rt6bk`v$xXk-cIhdpSoXeyZ)Y;7OoCDDo2W@PS&s={<2#C$(6@Fr=rAy
zSFNom&0*kL(`D&f<ZkA-LnYvrj8YcE-B8cTD^IfgoYw6+XaBKu{$&K!2c59aykV|{
z!)uMH7w*3^v$CzPi>q#^I;K{!=XqfKOn02@XO`kx?L$xJHn)Ub3yi$Ee{%$N?g}0u
z$LE37j{^$nKk#np=QCLmZM;^8;;G|0<rQnH$mNdFpW+YSbwS?rl8(n^OGb5-*nK0v
z85;aqUO434+xM=k_f4Bg?iL}wTK82GkJ>pJwZA0zxSF4Mb@{fGchmj*QM5jvkGqV}
zMFr%o`c%N?bx}x$PHG+V?EAaDdz%l2T-f{R)b+-Vo9JCCnVs(ae&^|#@O9Ia;PaS}
zm^}(XqE@qsdOOo&s%ml>6$dLt%x^dKZlWx|<+pkB>jS-}^))wn4LsM=D;;78tcVd)
z6TY)gPA8sjYY4m1*&9_t0qo2d*rcgr#QoTNM1obq5Atsv`|Yc2?(4l{hdcEPldcM1
z7I~d#RlR4*9V2;bTRqN*+_qO{Zz}#stuTG<BehWFN|SXl;t%YVy~a$|J#{!0RLXH}
zUt#bLw#}m}v@Z(Z*&}`EJ)=#om(U8`LR#{CR*8PPa_jXoLAC2ROWpgOuY@~)%H%BX
zmfqdzdeuK%MMJ@m%{8d-!u_AD;X8Zh0`P=8d$$IugqQnClku<hVv)G=!8}$|#&74v
zz>UQRu9)w(`F`R2?cALsF_{-F>)mo5&F@wwcfQbb;sT4w#?-4CV%Jz!zONCz$y97+
zCn9tFC7Tz`E0%Y+PF$#2Ox^9PoW0d(6<6ob-kbBh#~vOyG_r#3_@4H?H;?dfDs7k8
zw=VQ;80|KtukTH%kIS#NzuK{qPT`kzOa~bq#l7eAD$Ii&=Eee&KX)~~-_sB!9#>y?
zWOGSxqxwkuQ@*(P5Ax3})p8P4?()&Jv>1J$z8z<iS$^ob)Naa{f{Gy~!!!31Y^43Z
z4f4FaKdZSXL&l7COFNyV!iGnzv56P_)7PCj#$fgIR<Tn2%k;v^un(f&PcoI~HRzmn
zjP2MvELow^8pFgNb>*nYgig9?Ra(;cmB=WvJ5JBnM7~P(VyH3aI!hko(ky$Y?JiGE
zcl#byPo_*h>G)qiq!>j^VlVaj?ALGfy<h(#pddC$JS-$gXF@fqwW(m&xRhjI=MC;1
z4^{5`q8f?QD%FyI;1|Ji)U~I!yY4{j)l@h3mmd@3F2o&`O5urTe!iwFD!21BcihPn
zS8pZ0-Icoaj^y=Ap;!7}JV|8QAjDV2(V`H$_1yM-$_v4(_F}ZG(GUFQe3L20<oVOD
zWR6u#+hQD?J0HG%tHrW^U5pBQDgFJ^6&++T{u<ObRJckU_n_;Qo95KM9$<THaABRa
zxQb3{lyxc9)f!VXi5Kn_E6ZG)?GH^~u;bJJWgL3C>0aC!FS<D2PAUcKRns-OYun#n
zeSYzAR}7hCl@q<@wmprTM9-SXe{9sf@l@m51-920BNET8STR<2Wq18Mnj-(9yPFo;
z)l`d2@18fv@N|3d?U(m3dylPLj~!>)t9?1|KF#FiU40;-%5ci2ZP0)Fsvm;!Z>v93
zyqfLl^;+S*Vy|chE%l-4BJ1QY%5igFj#V}b-8x;RX+3+$R(q`9Wc)Gpoq_tzJL_`-
zGN<n|7dfx=yPWpPJcQ-Ct&UTKY!{Qu+J&R3`Q=(~2E>#dzBN8E`u#M}lWVuwD*0ab
z@S#Y_j<#p@uWFvzY}x$WyOt$$<%Ul!)MHcIGbAXGU(TegQsnws@JQ=xTd95SflRa3
z`FCG-taRIXb^p#hnqT((!`iKboFBXI{iHQb{c*)pKX8j$+I{|WAvb&YFRWM>Rmt^p
z4f)76i>>SVd&g;3KUggO>hPI4%eXzq>Ski1UQy%MqxOc^hbcyRHCOCU^P1F;Q}oj?
z<$JBo)O`N^=EI?#D`ZxsZR|1R+NvFz=`(6>bDh2=n?a_o;PR%n#qO7UYa&)VeB*yj
z%igieerkwT(JVALV*T97?xZxf!>5j|RaIuXQ4yWDPu=fw{+CaI1NMuxJA<8Gt(rHu
z+Pa^=^85qgl6f-<WfrEZ8$Q1~e!{C;UrTo&N7G93i_a@D95!otHm*g^QRQcSoMyA;
z#P3Fmn?_8};#^ivS>B~Ay=-CIU}si2s5_`BvB!>XfPuM~=Cb>qves}Wsu006j&N<S
z&4cr~A5}&BeR{eXCR@UujDN3F<4Ni$<QKn|X!(nJ?&0I_9Yg?1rf9zIpJAx@e+KP>
zf+GK`pgmo)+VYDy)tg?fxmBuicW21lR)or{T(HV5vv~bNuJsH}>y#`n-(9w_h)c#t
zs|rKa{bf(rbNWXayci02{6=bP`!<;Z$8i7hmHuJv1t$*B2QRqA6fr$teId%DZ@zSV
z>RCE}oz0VVi4zeKM;47|Hk^u-GB8~^zIJ$=Ye&^!#}5e==`{O^+!zbF_2c4kj;}=D
zNKL8*M;zz7Xqt6)t^4YQzQ`$AHuG(6fnvKwIOr?)S9)*=82zG)-Nq5F+@mscJ9~Gv
zXlw&7-w)|`jbU8Yzb-FKv$vn`*tXDm_a?uxMa07Vh~T*=>$+0I!XK>Nt|VJ<vV(0Q
zFnK_u{6MPyMJhdEo(q@aE_<z{k}Q3eP*)^xYMc6f{G-5|iA~!N^18aXL>?*E<<ZgK
zdn5OJxOCFAuM*YyjLD?Tse$8;M~Xeq(n;?;*gk%Zj=irnfP0sI-nLD(-&4rOr^wb?
zIhQ~3+@$pSlE-gnS^X$l&qqSf&Aliec}i^I(0cBvZnEK7(-8x=Tz$8aHLvf{X>Xs=
za~oJGRJexyK7L>;CFkEbCSEu(UbedFuz}ma4xz#wVK)wv`QhIm6t8By;bHAIu;CH;
zE040(=Z^d&eTed%(8<fvTSC8X?!6r@@)@7<>|)TS@g;qT(bF*etCNsW;l#$!uTGvy
zmv1t8kh3X>)ON4gyMa!dotjR2Q!@Ua?GE{?Jr_B%j@&C=eJ=1d{uTda|0Z5&Be<8r
zr)?7+4iEU}P4)imrt0s>BR+B{Qx&ecu;qE?qZ?J0q6c<oGH<uhqDek;jgskoX#N_h
zdvsbBWmN+|PEf?XYdrk3s*%jVl=7X(2aB7YUEdizORen(`uzG5J<8@7JRQTo7KWY;
z-;L+qUpTR1`Pm(qII|+|DP5L6vODX5GaF~t0jUVtHT|4tRu&o*a8$mI+_Gi=)r~!}
z8B@7jPhDgS=x5`FlfL$e@o6aV4m$Gy9h3wJpVwa1EzZA9K<KV^4TDv~g530MX!sAq
zcjjXibe;lao`W39><Y&OpO{6+?RkD6D}=JmliWANnmbuxO^Oh`zR1R_3Ll5RuCl&C
zv3Wv{VwNK+dR6n=SIr+ciS06$da$c<3p-iM?{Tgcr%xATvIWM%uKj%Rb^c31NL6oV
zgo1BG)~nc0bgkq$*CPsaOnYWmW?VmKMYcyMd&4?Kp_uQYXOgv#eS7oy!*=aA6oT6-
zGbLx=dliQSg}!mQ^O)?Gwb&{*f5tVIhbBzIuhQD}GFg2+L9@<xYxIPn;?~S;_B#Gs
zB8flA-Ti!&u9?oBjMJok!NW1J*n8V4ruLPK-OtTb7Bb8-<~|QO_F5j#&V8Ozzg_Wl
zVBSRNgL`!a)?u$dZ0@a%wm$K~#Jrs+huZP9aefV1??v*pUWGSqr3WPy1t|P{Iv?@E
z=9-O|z`9IckIB!iLm#rXQc0`W>xG|6^nUvB=sp)6cP`EkcF*kM541bd)t!9O#+=^~
z=s0BZ=IXPTYlTIO^xjlFpOz?leUe?{_V&cE)*B6WLeEB1E?BDGxp#bHzICG+`&1UA
zd_MI^W?Anm|M_3`kpcrTIp>)Mown*eir>dCm2DdIrNybbvFi=b6`9>+j#uTVqqt<K
zJAB8=1FmVVdXVEeSFxgHVYY7V>5$o{@!Q3AYOR}TX=kKpk1>p_D!Mn#@ZCJy!SS+`
zP}s<;pVf~`(ynZ`(-+E(h{-qeJLS6niv;Ut^{4a6eJi}yja`zYZrdzCmOGzvp~Ig1
zyJE=Z)|n9fpW{xsOwvq@Wg${pG-+~XlSY&3K8y28He6v}_hs7k8dA1d^6q$B#QW~H
z*yP!on-tb}x2QK2*X_Gf&g|M|f34U4qWUO5RknB1icn|w))D%<?|*oZ4;*^J#PRb)
zS6im$ns2+J>@_PFav#T0r<V!_pOcj9Z1YhzQ8R1~oU7oQJg!*L_-ywks`K?49Y4tf
z_eXJ-wIAT7?Nj*>+ncv}W-$BJz|5|BmOc9u`DDJ$_%`1R3+1{Q?7PJ&OVKSsf5N6R
zZHVcM6WzT^k;(4GJoeM|tresGJ6N;?sdxY2@H+2i=`pv1T(Z{tklRW2`<y>GqFmfQ
zI9e%uxJWKZ9u@K=pln;f6T19896va|Ip8C4##vhjWfOU(gl@719&luRu`2%TcAhgg
zSz9tfHgP9^qVps>yXQcmN&5~q58;EU*HSMHHGYvRP0NyS^|I^HZ0Bk?tNMi_uZX`g
zx4Fo&PA_ZTX`X)n`$w9UU#KTfPFMt~etS3*n3AjTYj=03_C#_KAJdA@W^;AaSLn}P
zRv1Y>$<?TFC5)p>tz#jODv>;{OFk?;sf)Ja_t+%`@$OX;J7#%5y&Ck6Id%VLs?yav
zcHB*`6y;l=<=j4WTKBTCZLU0h@w4ymxmCkmwml+`olf+)ZX<IjT-*F44|DN3Uh3TZ
zAHKB)Ivf=CUIo3Mqh?ueCP$tLn%{7&KPr=DN=#aTar=wF5t%S+IsKXR7Y`p?jeclW
zr+&Yjnl;a1;9lnWsPko7WAke~o5dgW(t7r<I}w#cIcwbO#l=<G$h?xwCeqC$#Leh=
zrkui!v{N^KuPnbbT=?5pX1l}&YOW{z+x){nv`(Ll576wG7u-i<v)9T=uKlju&w`{O
znnkirp&<qTl-3g#{*MSjTvTu=K=_~X$>|TPudwW(x-sNJasDmEDvk^9Q-#eq`$RHN
zUauUl-<Mzi>5{<pd9Qgri*;eu&t*1pdHdF8y3w0V9KJN6m)@52i!JdC+p4Exri)R^
zl4)~td^Dkf$5v;&oLo1NSKE@DmMHQfJ-sgP$3YHJGw*}E$_t0h8{73#Tmv%#Tz(Gb
zST|>k9jLoEo~HA_C#+CG@#1TPj@cmNEX9=R6j#l@9qu;M$-49B%M)@06G9G{uV7E&
z)%30HAJDU#8o811%QaD^TkqU)CaV>-Ty*3j9!l){+C6QBI5}-N@7{m8w=LC|eLVDz
z$X(i@uykSm628sy`u2xB^ix85gz10fOUSBAi>O)ErG6^gaYZFSd@<?S<ppk^ix+;b
zkES->WG+atCCYH3H`hu2z{z~JF7iDNBEq%uQBN3*+&rk<zaEncJp7YW$JU#D<T%3%
zzVWvz&Y~6<-%4)nmCkwi!~NHk+mAJynQngMh^)~~o7w&Nl)m{zW!L?m4ll4Ky2RBS
z3%ycqy-}om?=8_^k1i%sz1z4hGk`K>Q<}Hc)x9=GkKTIa^{cfTM%xse)0J`C!SjJE
zxvaDMg&p%n|LR9}E$iw`?Ni*ix<w}9?-*=*>fz$L{pYNp4=q=7K%-)$<Qa8Na{rrr
z_kT+6y+<A1vv}g(jWotyHt`%KC3`9nNek+-fPPNVlcxJ)TjkqM#D$%GRC?Jlf&NaR
zo6W*eDzaOdfm?z%o%XBTl*m0ZsC38sVxnLmjW1>AQ7Y-1Q{niXg_-9|&#~{ft1f1V
zH&Prv>O0@JZ+-K@w{zY5@6daRw6gCwM^-7kiGM@-Azu;SQkESYJaR`~CGxNEEOF-*
z=$I^eFHY0>O`V6TL{?W|e|zCa`IQV8jXpD+@%GmKQS|2A`T8*?{`=`gAID?fTQDiq
zGP0huQ?iIr3wYoqKEL!p@80~*+HiO$S<Gd&Ey3hzCHHt-uj_BR^054}HQppgK*q9I
zd9A5%!>8^CmM>#8B#cjGymzExEY-YdQn=rx=2BlZe-G=%0Ii(!l{%8!BO;IVYNwc7
zmKl=F2pX%k&(ioB{UDRPE%t8dH8<Py{XdI-zuVwp8MW5Y#(CS`@#$JR){X=2g-w=B
zKA)NgCI%~HOYg*z%18fRND>zO*ZlmyQAkobc1YmN2|XLL?ZSe>`|w>Vf0UK>i3kcU
zA%*{;tn~A-qw|KQ#l?7ii&$<|nM5nacPcHkejm4UzgE1PZof%$b^43-e5z*e_>Y*A
z%ZbpM#;YIpOf)&NarHs=s~sVs<wMbZ0zDBR4~Q9A8toqX|5$tPcq;$@f4r2GkxfQK
zvPVWFE4w7C&T;HLGBcut6Iq!h$_kNG=NQLKLfHw&CNrCmz1Qz?dc9w-@B8!lynmnD
z_x8K})E~y-T<1F1^&I!PIeyR{I-HoinbN!Kmpa`nz4-hkB#i%m{-}Rj0o9a$K9CwG
zutfj*{GU($*KhrYA0iF0oMW+KkRHkJ=>Z{b86vO6p`cg^O^7`#Fc3MpM(%S$?s2Lj
zzPekymWc3ghUs753qYJ>fPbZgpQhWjiO(>hh%Jv=9{uN6zVK6#QIS0z`1{9e<X9ry
z5!{iov;RDV@Bk{ZzkZwWO=JM-5$e(4>3=<Gx}7<F7;zXe;gA3Monx4*g7D9U<j*k?
zelMSta8MIo|2#yQ@BqO_|LvgMW&Yc7U;WoH|A&M6!ao*49G(Hb0f<)w+rctO4xu>>
z$~|+zLk83cC{96~X_&)csi;}@s>7k>y;PK))-Nk>$MZciUM`}nPo%Ppx!B~>Bh?xg
z0=vYoKh1M*_I6>xb073#+l?q1PABGOQE9W?y7H=T%}J6LpO)xiEXsPL%{it`Ew{H}
zQOZ<S-CACSi2|kC=JPg=boH@kSJKYo-D0DjUtUhF+B*8NPa2rw)J$GLc1P!f&xx;F
zO;EUV&#r+WKnyIZpo6FcKbq?gg5wON@sxgP=T5p6k~S@^jQZhXT=d^uMg@z_Q1Bew
z*4mfvExhi62|+5vmZS;rM0YF<tI0M|1*L*gS+1`T==D=t_=yO?g}2>$3iyovqMwiZ
zhY~zba7491SW>vpEYbDyDb)x3bDah(#m|~ssA)0jNJ%cXPOQ=Ego`vcbm}KoInC*t
z>@S*VbOZ^hwb&Vb3vWM0(Lgz!sA=;QA;H{Y<Z26EFIpYJa|KFUx<pc-l3-<L7eOLL
zhd!qce4-YY==$!@a1?V!i|zS_y{pVAb9M>->8eyvwQVsx>;bvY(b+`3Cf&SObECHC
z<faoF7fmX-bmVRsM%k|GvAl6s99pRAV(+UNVkPUvD%n<DOqG8<MD>=OO=$mZam(eK
zR2MN+LCa&(h#Ai(-ejQ#&x^`SNADW5T_?hXB_Qf!*bli|GjKfL0+cjfBB;8*6YXJ7
z1h*Lk>ovxUb6<`lIrtucW)pP8c7Xrl^-N-;&o5qI27gZ(Nir>b?Kt_5(vi!*-n&n!
zgeU9wr4)oKr;;WibdJSq)*KeQF4+tBGon;kv4M~Sbe+?RzNLRsHTZG5L2#ol*WALb
z0$H6C1|Z513TzV}!*iC4d+T2h&C|)evH#ggv+T_2LymKD&wq$S_<rl|zuBLaQ?%Qw
zDvdeGsxo_zs0rW5C7;toiqT|V<;KRw2+*oeCbrnTK!BctMcNV+kc_;T&#SF65Q!Hi
z2hPU=P+5bBbr>W*1iz>I#5ENNF1*$VU%9M*Q&&}e5`us$d7$xEwN3mU`ABh2{Hz<h
zcU>Px(j?jX^6s<6sa@|VJKB3yi)~m#>=^~lUd;PPC>sa|iqdcZ*+#a4FCCdimfjJc
zx4e^aIOLVd^sgbcR&jb-Y35Sd`<Vu~FzK(t%MNPRgR~*3*NK)-(CT(6;3c`!&@CL?
z&u0xaDho|pmZ(g(X!FC(QFVn+oUTfBW6c-p@7Cayr`9CmGRf7)i&^yf>M2xtTYZV%
zwSO1HB>b3?Fs-WwlTcd=8v1@94?1Zc^?D6n79?L&JyZd|33_`a>kO*SI+kKykv;fz
zt}!`QDfZo}rCzwdS<wRVi#Uz<3i;~t1!<$#KA&4N7JIn#^Js~KBe>rK+X_b~L}g7%
z??lFyIXlEt?lagWi#{fPZD-<1SBZRm!>DsTjx&^vvctyD4_C&$I&@pHFubcY%3bf=
zuRL9+@aKp@6Wz)Vp4xu@?8p5kJ$Ymv!tI_Hm{6*|Sg_u+RvfqZ2735#xh!z#g*`$P
zQ9GJC4rBP1FCR1d#;t>FXCDkzE%H@4lK6a~(f7Flm5*>@OQJZ*!!5}9V+=~2RNh-&
z=V>>JQ12aN?HaLQOmZ<EEla(gc)R1^2HEhzrw0f1)&u(2FKp+15iC)RveH*DE(|X!
z3>CXy<fWXjymo43a<w<N!dF-lx$N-4^o)|MulZt8hH0oCQ|jP}O>4Kam-55YJqYzJ
zIO)|x$OBWdqhH-GUvnnzZ@f3duKI;P&M?^hY{}7}IZhU$4sk`QYZYrei&Il-xKDOL
z=jCV(w3h|t-etck=U59peOKZ^>#I~<yQs%MHg0-FHAY`BS9aZfSfkfl7$q9qh*D)Q
zeW_)+p};=X1y=-VKIy~;kupsk+ifka(|D_|UNNd^XI@gwnC9D=9mi%2eI!jgb~w#F
zQ`QE<W1!qN^nsk@EX~xFe37y0J=HI#8&WZHJMwI`quleLqKLXcf%j=VHf;k9b9DPd
zF8=345=DsQA|eqQPR&w}lkWB?ok6M2U<<q!4CfUg!xr@s_vnOuQI=0sbLpY*miOyI
z(k$flE554UikBYJ%p|%TJ$iJl;WvaljMb4TPI&o#N=GPr?3~>d!JElAZyP?v`>1g|
z*;IFYDRa?wgxzX+<(<`yCVQ$+sR4+C@9rAJB^*l9d{ii6_lE(_Bb8?#>fqB980`hA
zV#H^5i6y?H`S#4<Y@0Mo`sz(8UV&E$aUsx*at%EXQd9;F7b&hWGRFv$&x%r$U2c6n
z9P)DprFJs%_`}U^so9>WZ0#eMu*Zm}7W6Lk;~_+tBN)0qa3#)a&q5YQ53clh^?Tga
z*cIJ<Gi||5d+j@yGA!tSJaS>+*sHlK9~P?MbDE5-hHV!lUj~Qt(Q8hhVk7aG!K%3<
z7){n-`r+xq)Jql-qIR=x4l$z-cNm(9kkphI6!(?MH=<wozwi@FLNLS5t(}$cBE-yO
zpN1I3DYvKJv}v|sE&ZQHXBV7b_NAgkuSLd_f4G$${p4|>>g5jcthnhCF9C~h^5NAl
z&~$>_(>bZ#Rl%0=QmnsQ1qG(Am!m1LoXp6tC6w=`ZXQD(MJm34pkioB*tuo+$@4vx
zRlxm?ApfOAht`<;Cr>6K1Q|xtV?tT7^j0`<ODcFs@Tg8D1ULuB573XR8|KFBx3-eO
zu(q5dD)>t08$CvFmr3CKY6vz^{zBh9h+j}fauk}G;}u!TD#;6)AVd!WPX|EO3LTif
zAWMY`Ku~1%f(5kH>4!U5AR)w>r-(QYjFY{kP-J%y29f4DI=n%QWOP|)WG2?#%e8*L
zxy{Lq)IAX+?7qyUNX4|NYukx64d0QGsC_{(lS`TyX*qa*=lHYjuuSFq53;6lRt=VT
zSs|gDsIs9rciiy``mFB8JsG##?m+@b16j&ruAv6OZeNZQNt)AB=ZD7^YxaI0OP#dL
zfz=HJ{BTKy%C1Bqz+*l=KK@uPAR*xlM}45Y<lDYmy2o4i`##%`%HNa-wC$m)p&VTp
z(q*q==ZHeBM{+_e$}iM8GSO0uEwR!3-Q|O7&(%*tI_-0D&mS*^JK$57jPXf`s?kEu
z$Y;SV{;%HP(+~7hZKfF%=a^6zMw)H+7EE832yzQAzcT3q$sow<BOW2T9Et;NLCF3Y
ztO0>I7@ol$GgMbscdBzrD~y<KiWgIrD$B0W;qYF0v_RuPUS3dqarniiD}~V%2cN1R
zRY$%NM6Y722OCSExHsBAVxW*L-rcwUOU=zouFQp_R9WXJu9ZI&ZxxOluS#9Vl`k7S
z^T=kMY5NTuBpKUnC3AYXk%_eSh}%kT%C?KU;tg_jUK7oUlefkYLz=eCzB8oufivzn
z6B(WA$nk&^#wtCx=5Emj+R$^3%Zta{kqVLV@qAxf9=zQ{v5eCI)5)y%E=c|n?{5>e
zEWDA>rx*Lt#PMDJcfrvixX2pSk58P}NUq=dCN@+NHLZP039s3w)8hTpnDR5H<=3-H
z*<^<JFZ|n++=AKP-<S;a%d1|i#S-$#AY>V=O@xTGh6a**ZG_0!U#@rE+@v3@y!i-A
zGy`b*IM`d&4m@|9nQSKrB@?4VEsVM7Xvb%#`KA1WtKs3p#ubrIXOp|h&`NU5t5m~*
z(ujtsfd_bgup`u*79b}lpZGnU@_eA>&f&W$yr3*|c>+_RqTD*k=77;=Jc&<ho}6P@
zz1moJbE|j`@2lts2f7gdPI>|f&vS0QscJ4*A)qRHNgb~R+C^4S;-esd-QmMJtu%CC
z7l17eE+NswNN5zqq3EMRv3h0hli-eOXlP)GQsby6+U5`S_Yd-&DV6MbMw)<d8Om=*
za8#|4Y>Jvae7%Hc4-ngtsg<_0j0Z!imG&+;3^*GX-YxcEO#+;)ONjVK)+Muyo5yXR
zoPD|5&#56<Y<W-b!r{u(Z(sP8Z0N)FS`wRdACm?m)_e%=1hY!tJNol8o?yNBqP0Ap
zv35aW<S0yOhv0rTIi=+wJcP>U?VTO9)haDsSN?*H_0Mj+S2*IYY&*bz-u2T13Yj)Z
z^n<G_4>*4PXnHQ144#9CiWck5e5L8>jlJvfKc0tY(UCE2d3Q9_=gSvYoA|$#Z~%BV
zx7-moZr>RtuqJ-JjF*H2JLcP&YxZQWEKelM_VALht3lTB5-8Dek}osQF1%`B0|^Te
zcS6juV2}kd?ZMB-lvw$VJcxz@>xa@4VptEIKFi1X*86I(YSBDjq3dPl7|Qz4O?nax
z>lM$F54&JZN$cX0U1YePA&q42uesSjWF-Hg3f8H0Th9voO=cp#-C?QprAmTRvvOd<
z)arSwoI?=rK&%kkwJ=;ic%Y2QmH>Zw7&aqfI9!{~d*klRqbCL5_!}X)JaAtWs*vb-
zF@5=9g#{B`23)HNg)?}oQcE^<+*OSWo*Pg7c5mVyYY=YWyB&u9Jy?|W^jW%q$4fN7
z4-c=Y!vG<}IS4m>>x0|!Q(m#P8k5r^6H9KrE`~Sxa^hsgOl2FKdVGCv?y7?THYXfC
zgjoOe?%;w-l{e#<u*a{P`|MR1G@0uCkb8D(FRQk(E1h)a^j~q|@-D8q<(S}~1L$uV
zkLp_Zue{0y3qjw=>Jo=zLIcL|E=a<`3dNLt{#LXl2Rb%I@}biwpZ9$?Y9;f=Ms26v
zYxn$xX31NnEh2(*?Sd}Q^WRQau;bw}`t4r+d-E)cg|D7m*o&j+;0xQlGxe{YA6L8+
z=Jrw)#%8+PM;Dw-TQqz1-uRm`jP9*EnbTX~G!A@S=hPU426CmDU%G8Zz$U;NPFmeT
zDtMB@W8vfvsRcNPd_d+Tsc?>Z2%gbS%fEJuOnJUc<IBSr>^2w0ODwCesq1kB2vI&o
z_ZrRkfKFR_BZZdcvi4$13+<F;-aYFIzj;};qYCKP2xaFkHTk(8)ypYPLDn0O!@Ka}
zhdQ?5)eKRLyQF7O!TQ)(bpFTC$#B8Ew{LG;PdRwEg}NXk!(vpt0uH{ydme15CLbY+
zaDKb+J{Q#-?keBp8VCK3D@J_6(vrT*PV{+6=00^v3loLbY9H#cVYrV=$D3389FX_b
zdy2&0EpA}7bk97iE{|(q2>%Rrq(aQ4&t(MWoL<GHT~}hSGV9-;vs6@-_TUWs^cgL6
z#PaY4dns&@c;UDL>5=Ce3s22;2Q(33&efD$+*}ZdEe`7`R|<C+{3tqPrmd|#Tu)T;
z23dZmSX!<nHRg@gjd^0J4Pp(BX=fMp3d+<bt-V>4>zT^q<;Mi>z0kOUhi&8+wiO7#
ze6>`bGT{|o?c<=@X5J8cxP9xRtQH$BIfhR5Smx2F;!79?0X52WeJeE!Ov=V%|3%@k
z<+)=_QH-r)b9o14r?8wSk)qtM>dE--e`L}=>s_rK5lh%=&7!p-J65HczY+4ARX0oF
zPH;LKUV9Y7b$KUnpf%<oX5>Ngz$}mWLFwXFk<~zF0?RSJ3ayVm5FzqLVHEc8S=re#
zquynW3$>k^d*s0*-dFCE5ee*P8yrOBqllCFIJ0PA)=5mZy0<}@91>gP=;F6Yl~0OC
z)sem{@)BoNm`@2cW<X^@qR^OaJm~T*Zhq;x^BuB^W)iofA3>18sd$)k*UJ|xZ=kD(
zhlg9Y?+Ck>&#n-|1ZWd#HfUv|5U&tTe-T4x2lJIs$}6!KIWosEcTe&&wu)WNqAuq_
z#VaDWA<SeJ9Rs$jj|MidrAN_1XwebO5>IKe@P*O$^Y)*uifBvD*wIVAkcbOUEp-_R
zmm7sxuTw7(39ffFdxd41`Cr^Xu4pgO@nJ5;&~~t2IfAYBfRmcmKxuf-yG7*Vf(k7=
zAx_m2jR)y}Z3{mb7;24Mo0^D_quMUnk!nt9VO~}rmA!Sen5{&ehjCeV*9XBn=QN!J
zfQ>K#L#Q*jFqtvF%3mRbx^VX?%0P90&iFg6bX9s0vpKzdVK}_%>jQW<ogeJu){|i|
z2><<c{rSZwVPRBKgv?U&OT~R3O7T8|RNCU}BpY!*A6K;LN3w+b1>BGOIz5=1B-1js
z$S(Iqv<rtV$~z;YlJ28*<3<geD@1p8at1$tK(DsW@<xrM{EAY?s6a$7=hsYC2!Pg-
zl$5ozu{oVj>W#Jh<nzSfE~@ll@@n0AG32vrT&Pn!a2d!wSW&6Pu7E8Bg1tLTAOnVK
zkktyuJt+4**oxN4Txo<!x4H3^uB}<A-}9^MvR3uv^UIK=EOS#e1)J026XM!>@2Q05
z%Vqtbr1&ZrS(e04$|gS*DL{Qz$S3`t?k=aga_QJ@FDGI8<*-P^iPNlM`=XGZPOB(L
z$8??x+89I2KD$q9n9~nWdWFZXuM2!*;^pF)8==p8;J8|=JpAj2wOIH2TcHO|$&${$
zH0j;m*cqg?o(W63<@y6%#(7dC+xGKaB<yMrrn!-C^ea4%mMa{^R@)X0e{Plis8N1T
z`Oa);8rB`EOKDh3n|ya~<r+(SsXMdQRRbiU;q_T?buSYdD+-EFL7L6bU@AoW*~Hl5
zM~Rv{zusPbdrG1d9GO%POCaok3gTNF@c;(pdn_0Fn@N}s7OU^d22wF$5*FWT?91Y6
zVy~5fKP_$hM~V;aLlFa=(-^w)BmM>9P%aD>#p>#6cMp$+?E^pyb~YFBdp97!=0X?>
z3Wr77j?fOZLGv_}ntXln{?CSqn3@dGtQAe}JNVNK6&+GJx?1^ess{t?-)Wx3dbJp&
zNP4@wfI&}c|10vH4u_1)KqCLxHGG_!(7N;Es#8fjez@Hx3Z9rrA4v<p_>UfISkp7H
zlmm6A3r;GP$_1891w@!|ec_*fq_J~${_eNE&Dgn=D_ANj44tjGnb`<zy`l6%LR_3S
z+4nv)e?oeTxw$!H|0kf;9-p8<rC9t-LXiH3#@;%p>=67@99sis(prsA)W@Htrn1q}
zf^RS;N0$%FrHs?{Ig5f+BVLcm>$&>37IxyBMKAL`$V!r;kg1PbM>w}OFYp?g7#m;a
z=Qa7|Q_@iords}RZd3J^T4u4^oDd*8%^UGCztYrc0^2+ky+f9aDm(+0oj8Bx32oiv
zkV=()Z%}*&t%yA0-s#_Hb?KJ0BL2<ojDrUy-7$3**uxzE!`ycw$u6ejWvPijg=Uu!
zy1(gGEfz1a;|++#gM=Zt*~`OY7hFOk52jz0hL=EgE?lNJcyT(7{0XlIO~I86Z`LKJ
zzNN>M@?Z4iY@u1`_WRpP-xh41BUaQitAbF{z9kFO1p#t#oedata$brAw?c&#j|&~L
z#qvETZq>jBGMoAyt6mjmMk>BdQs~_IR=tq)ea9Q2by_J3leC)YYhATh%0aHdcV9@!
zT8=h6HBs61_;5$AvyI>W)APu)dwJ}K7MBI$$zNfi2y2W6sc)W_O0MoxkH6sin9``5
z{&)kw7xS+k=mtAZN2ZDk-w(MOoU`Nlg@3Segy$r-Hs}>Q0BeHCjJskO9lGGB$p^RR
zx-rv1HKecJ+cnu=&76(3TQ5m_nI6Vkg{q6D3^&Hm-X!d)ZNd>vFGbkt=+;&yZ&=h-
zC+GjzU975t8@Gh?!x7A7JE07Dju&P1m5niWw9d86C9Rr#^hjy%->e(A0UqO-#)LV_
zWtyVO?_BaYrLlbOiqE-cZe}PSuQ&uOBO<>|OFNsH)0@-tG*p8J8IpBRke|`PAcc0z
z=fUlkK^A<5?6{(FrSAi%JOwjh@dBgVnvfQ(`}<0qS+XxF1Ry>RNGaM@=^WAo>gU$W
z=diqIJg6Oe7|>xE54J|jUlr)X%2mV!D*@O9EYD)VG~C{d0AFlRlep#If9xm+So#OX
z=4ztOD_LuE83To{#F#(25DqH7qiKOoRqp<hIRNz*78Y>x!Yn#JR9>2$Z3mk$C`n>t
zW+qtEAi@d!spsuh!?ju$!sJ?0q-p8ti~kr;f5PtL{G&UNa^OAU$4Em{0GXi>J^|IW
zSF#1A(J105ly0uTUVg#*LQjK9g(?Y$@%Q6N#R9ii?Re{Bo>Efk1Z(!VkV^p(kUjh+
zj`a21c9nmNeOG5>=AEnlvx!_()=7H8zs5JFN*Cv^uTURt`q}<#WNOY!HA77*JNtFd
z^zY4AZ|q_vMyeuwRWekKb5dBj*n8APUEVpg-fsCYvbLKs@;P0?Y%S8~ydf$7a%fI*
z=iqE2z0~1hLnDdY>c+eh^@LaI88oWy)QNw{l;v3_0zvbCLDBs4D}V7XjSK$^MMEHH
z{x2w+|Hi-kPu}R?C>jm|vBdDdQZ)ZpywSfMcL;%4LLK>=H!8f0A|@x-yt@IJ4CP@%
zPh)})UI^%G4YmdXab%+iJ8c*}2_L=ymq48y;p3lAE>0JQ8+?|uX^K;lf~@uyOQ}y)
z1gx#J&_)Lj@cG-)Jk$A&ZhbTQopPriVs0{IJ`5G9&{jjXzW`9ik1lbnc<8;}{8rXA
z@?90e%UsgO?moK8@>68O%O0*`0-3c(j6CZ6sOss!Yrg=*Rs7R54u!v{t*g647y`ZZ
zQzW0%{@7md;OPZ2ED_-ytNOy9PO-#DptK0zKoc0cFJ_GkZI$o?;L5h(as5Zo9Hq*C
z{sjC%{}9v>X98PBcnUZSMig<;(E}xy&<%KYI9Fc@2XkVJ%qQfpZ@@E9k^L`6nQqqz
zr!Jj>a6ROc65jFyI9SR5au5P8NL@@f1Qmio7*+}3PyGr3MG*5#OG^vsmR1+fTxAJd
zeW}OaOxY2|h@YkW-Rok28NmIx%QftIA@P2S;+?Li1FGqE5a<Ejju0sWPWXqA_XgFy
z;o;$sn~3u9xuUB3g`cQLT=kjskE!zaz(Iv5Q2xruEMlnvZsKB!ti#SO|ML8)tX{<t
zq5J6Tfxhupl^z54C3V@$jaYDMj1}E?=br+&ygVVA)wztsXT=WOyD|h%q+5O)i^9EA
zNY{wx2k4A~l2ZG|4cZ_{fFr(zZraRc9^A{-u?!Gn)_fT&VG&{DcV)W7!G1jRj9lB@
zy^FsVYHRru6B8lYVGE)?KY7lMS=<3LKjW=j<uEa~v0CUNU|4?&Wurwf5;2OHo{U52
zn6BIxC!?UCf04=;Jec6A23s5VG2SQI0t3s*bH6{NNK^Z<uX~T#(sw-;_b32%FMa=l
znf2?C7dsF7U2G=17RwwhVVCLT<ODp-h<h!h64aob?a%kp)6Pcw(kVXPW+OXp@%g?M
z{S}k3p=yraH$xQ^q2nvp^wO~_09t~}7FJ(g0AxV3fTSlxR=K$4hihadL8L^~Gvc8S
z`gv^B)L(&J^TuYh+YkGQgNHZa(HeH>O!QLb@gLWFdhz$C-k=QXJXd{O=F~Lvx`EGo
zR~4t1Yx?P2Z%s=8H(Otg=ek~^4gC`p8Y(KCKym_VP*BfrpHSn0Za+B}#Y)b0^G<P~
z6+Hx_@7@zQO=#rERW@Zr!Op77L2G#DC?-G4Zryctgez!u+F$yNTbwXPNA0oT7yPHH
zYHNDVx%WltyRnyWE)7#+jRa)N5pu7;iCtx~ruQ2(tt@PMBPzI*@Iz1#Nr;sGPKS|w
zxPgUKYh;B0AVRRZP&nAKVq|2*m@RE=^teCD{K7A5LBG~6c&KU(<>Td_uSwFL!|0Hf
z%@4SS`(*OWF2!R@W4iF)>*FFum)K!W)~Sc$cD7Z0haT+g__#Qv8$q)kd<zn=1voe?
zT9Ph7h{)>tdd}_lv5!!yRU(Ex*&g5I%W{O?aBhE+ju6m&)xe*6)B4z__qN22QXwqS
zXT`F!6Bh?1S)+PgjI+*2b^UTGw5<cSb$IGZ;5jPSi_|vYi-yZfWP&QQ{XV?!&{uam
z*O+P79X2EB26d40@$9@LB)tX{kTyw&<CLi$&Xl-{u||uVR9=<!y=pxu&i$&Ly)V<C
zoG7?tT^F(E-=0s}q%Y~?Jc=V#<;#vN{`BR`)BBNnO|!3Ew022stft8Pp6H8nJr=9x
z=ekaa;J?lXkgoc-w)$ZAGFeW}>hIqLpo>9b&?QU#e$tcm@n6J+TZ@a1NH=fQvEDx%
znAN9L&bi*;ZyRp~z6ILGL$6dm#PdL;jItp;6DBA)*n+J)$7mBmoK*b|A*~EDCf62U
zLj5E_q186@;sdMxdzSGRR)P1rG>1&c&FT}cxubbSXmv4Y(`KQH>q1`b)3lr1KfbW#
z$*g@*q+$04K8`@)jg;JpJP}iJWp4iE%IlFaCvn@0>}&JKkxwEcYh8ZN&L+mkPlknQ
zT80I4YL@$7cuN6jU9~xV-rHZn9QJ!QB2#nsU@}iAT=^=u+Gi9xdmQ;5h6vc^0v*7@
z$hZm7lhud82S%FM72pHlf9DXaU%fjzJDttqpFdx1S7&d}TiM(sL5^mp@@2M<24qjf
z7wN0}dP(>`Ik{3OY1!hK{W*yyM*Pi45rYKg993O!nu@DX<r5E&H<iwpf5mP!^OFV>
zO?vITg7&%qcO||Mg(_4q*PEe!gIwHZi`6|!V*SO>V;iv|X)^OJT9O`))(81o=dhNK
zajNzFC+jbIzYxS>ZYY^r-4cN2^rM6WxnVwqcClQ<>YkwW@s}K_N}l`irtQ4*#Fe#}
zKwG%6_~iLnyE<Eo+jf*mF8dhZ?U0F5u52$pLP(zRr%}8B0TcP!?!C(`T4-RJDtslp
z+|7V*%1abSsz?Rvu&iA@sE~Z+cQ1oCTxpzor<j&o^9Yif0(0(`);=4Vg&sfk-Ac}r
zI}wyW9{nc1ln!h}eJ9USyei4O@V<ql)nXN$RcKf;f{wHf;-b{679a-eMn&EPVk9R;
z^sACCn~Swe^hY4;`SXdLnyv?Y;s5L~0x>KWmC#T+aPVEqsqbnPA)M5wf(@lAfz8<U
z<LRNK!Ka;)GJ8Eb>S$&tB{p(U{&R%5g+IjV6pAIW^K8QK?MQYDQ66#5tAp4;u~U-o
z-FG*f&Eu7nA5-ttTKVVgkE7fQ4Fdiq)s{Z^L55k>c5GbS8g~pQL=F?4soGclYGo6(
zeI+NG-L=n221CI@PRQ%Ej6F1olxGvKur#PT>kSt)0ZXk_Wd~^QcAM@>_M&Yig=p$F
zMl(Nr@<9C$i}u;UY)6hvGfPj}GZZ$*9U|<I*xw)Wg+B|=kKS{6_R&x(gWN}PN+32a
z=GURm=*^goIOO@$NKupP_0x#oBljL5xTC8{7b5^$AE}HQdOnoXz&TTn6E^xao!Fr0
zGQoklCiAfEtbR-K32ZfHV;@qoW};ylY;Tt_&vGFQjOO(1`QbhmCoAK%6Sg$~xTw<+
zosCpYqJgc8hIGmC{-tTfa*?-hO!_<Wgiua`NTfl5^{M?3VPRqSGFDVp8<3~<FB5W2
z`CTM;U3s7H6L}c3ltJpO{sIdttETrGrC?ImM*hZ`-SPlL(06_iq=4+?#lTU{poF|I
zJMEe1cT;;|l<$vM+8K!YznR?WTb%#gc3+Yj6Q~Mm7Ez4nTn`22{mQ3g$z;f^(dABq
zxeF3PU|ff+V0j2KQjK5$s2NgO@m^7v-PVSFsd_AJJ`tP^n8w2X;_OJFugS%SFTS6|
z5PB>6uGX~6T%_!RkV`67Ov+_9OCa?i$JGHyA?UYlgC-{k3AqW(3xMqK&V6rf6?Vib
zF;T3&<lZj)`bv(sSmXY7cm`a#pL6Ydzr8ApzdnPt)F=yMY*s+}lIs~#AeXfviH&`&
z*+kIX20^+h1pSung;9vz`9eIz99?GaT6(?s*7M5sT@j&l@3kXvxp8D$PT#s#Uesbm
zc@x4W%6NW7i<-H+v!Z0L$6ABB3nwE8x8z-{9u_vxrz!HTxtyk$X2DP;#?0*RqKDF2
zuDtu)qQBFZV5wa}eFg5n7Sa-4vHAVr*|#}YI^s;TQ&@w}ZYjw=tR|83pd%AYBrmtP
zfB*h))g`CI&=5KsUOt?WYPf=*lm|z}!B2l+-Spwyl@X(l7DcuAhymI-=mAq)f(WFa
z6&$X>i5E*Yhv4xst9T;Ep--3Q)K38_$d~g%4@0W(W+HyO)p>lD_LJN|{Bxp2yUDxD
z{r#u5!+p0|*ZcDto&Yf`RK-Vxd>!X*;L45_3~?Biu*JEl@O>H@fCUVb9b^5n3ZAu2
z$-#kN+%+^L<Vw<UXayXan6Um@r-XZZM!b*4+o-KKEg~MVQc7Z2Is3xP`3BC14%XEI
zo7UC!Vbtg6t5`8d49sDd%P4@M6P$&>I5;u9L?k~RR2RU~;{NZrbF1{K?NVjwkt)`5
z#XG9h`5N*G*F6v%&vRG{=EsFu+Dt@&yHdF-pHXxt(vM_6b9s;NU?wpWW!@7@dh+C#
zgDhB7f&K;%Cum<ZfO$DPn}?Hj%1jUFrqHRh?GbO<#F(WTQuHZmq!}`n+@kF}Flx%X
zX6naZ4b(O9)2RZ^@$~l%2sK)etsz_AB18xlo;pEpJ}s0jZU77+KS;nJx(p(S1T>I0
zlO*2jW~%8jL#-EqB5X9sx@k38#2$G)d-a-9`nE<ImE$zE^q6@(kvIr)k`%;2!XtN%
zpS|lA`mRF6ZrgS<KZLzX)}IU0<2-+!uf8Z*D_i+|0X&VUw7RBWsqKQhuiI>kBd6UT
zYAyUhl)c%^xDKv>R7@g}_eWQF_jP(}fTU<3u4DrGj*q!3T)I^JF2uCRUKhw@tZfOX
zR><`6bp_$Im_!F;{&Mp-StS1XJ;O>)v&lC6`RA1TJtsV$JON%rF5+gP7jen6AEfVm
zdk5m&MPrgLpk%G%K|w}$&ko|FvbEFtKW+f4*LO6pF`8XLIJPqRTz<}6!4U;7N&y(y
zF5+WftgqX6?;ZgFblP4ZzHem<kXAx!*9HO(1ezi;y1C6UtgNuM?VTe?&Hz+-SQ&W&
z2?QapS*7fhE$lz)Wb>3{@^q@a_5U!0zBani-8C&KskY<aGc4l&X8j`e=0~oxv()T#
zg`XQfO3x5qklpOBsY&kjIy5x&@{Y@R*{v;$V33XBL<&bd+l`i0dVkcs4umaLR!1eI
z{WbTpf<Wt%;oR$MvE)!L>vG6f#G%|LeGn;w<O^8nFDBz)^#E>g-RUxGc2qxVt360W
zw}tct<=pw%3Dep$pBWSPm3Vz5S!Jwo7w<cZvUJwwmmo1V3=O<LCGL2;ZofNdN~8rg
zx@qU$YY|ERn}QX7eXmxhIWXy?@!bPjiDGrFG7!;d)ysK6tvC|JuKR0LO~dp_0%FbW
zA+An3cEgObb$Vv{5u#7ildc=oa8osVfYKd1b_~cLZ-9d!4;50J38ClE`j>^yhT=WQ
zpOfD=J3(nXeT~`_M=BQAnLpi$vr?y?`O(zpdq6%FNU}|JD2x1V0cf%w|Lej~!V3G9
z1eGl2DHoDo0a7(SI!fh$Oj2NcCSn?^iG3hxA!<TqV@gcXG;f?5IuE(oeUtBQ1`m6m
z7ySCdsPJAZk;48Z)+oI5P$Q>a|0SuRkJUpW5OVr(H6?{rg*)0M{JnA63b4rvkaW1C
z5X~lM7>CX1oc#sOnsNaEA(5Kkkx}>KIOc|)btYOwucaK)1eK80dlqj!%WE@G-P}Em
z&7!k9A%0_cH1qqhhxw(dz^C$NDq^v*FJ9DtwgZI;>q*kUFQzX+5TZBZCsp_<Y}qLT
zWpLpzT^oDuvHq>jEG{iH9V&QzH*;APaH|E9c<5q*q8eSgV4>JFxrZgZ2E;!g)C?Dz
z4`u1P1scOu17|LrHL|3y=M7H5`$(qBTPU6-Tj=-Zvqy4dmGwEEg<(=7s|b_UUwJAo
zkRGw)GsVj=;V!+J3gOCLGu(ZWb`q$}Ka3rglWb``*IFQNuSG$(y6}|b%f=lSKXdx&
zA`#}xDyph}d&>=Vff}&)BIU;`pbuDg&7fC^S^#tIB@Q6~1-=z+>rRemR;Cqw9fB*)
z2cUXEvjd7nd-Yl$1HWUOV#seJD~}f>D=fXU#A4$p7aapKUkw@<*nP({@3Q+Y9reNk
zpmN1C6a;VP-a43WSk`y8INtAWDE<yCUF&9Vf+sUNZ(-C{_TJzr1qy8S5asOp4kRXO
zCQA|&Eo;9zK?pI^nT-di@At5u-^?CF9^<mzpe}(h5;>81KohQNGyem+3v-U5^RY1}
zf99m(E)@6Q#I(%b+3?MuJ>RW;D_<mEEV=gjo?aZtDj8hSfviSCA@%Y1g|N>XYMc0B
z*zS`XA7jPz$Us8d{Aoi;hm;o#P(*g>D`nn<t%&CAjwA7H!*Zr*cCS`th<tAK;JyyF
z44zAu)}bN-X(ywsq%;$g0r0iji;dEi=ZJidtljlF9$Tbh&Bmqg`Mu>%h>4$%5H#~6
zn7KCf!-mDZNVBm!-15uoyIShgdKZR;1l4-Jqv6b(e+#Eaw%beWb*U|<V7e`D=AC4n
zNJ;a0TUAeH$CTa?Hbi^j5#oOJa7oUC)u}c0c8tDA$#21vNKgyJe1w)WZ^j;$0)+4>
z)MQJG=Qhy+(M&-N88J?5yotRbJf4sa7BJM=zA82og)+bLN-Mq01m=U>)gaU}T)61=
z=UNPtr1u+zkN8ABxNIrUO08ag4Wnp8G9urK81ijZTlsgb@`I4`Cm^XkyG~@4B6>Ed
z#?>5mL13FlHaw0{Obzw&Gn?cT6!&IM%lWLOo0(>+QH)M`)y*8;YI=ag&kNi`Q-8bB
z?;riR>EH}cwhqvBo`d7{QJD%PaU{3-dmQLipAzaGac|YNVPRe^xO8pG`WR-3%O8l9
zF>I`R`|^YMVQkgOpFEgzELS{v_fuie^f-AhUt<^EOVH)iW|C`4{|Z)+7Q$TqYS!6(
zqwi?bw(B)-R}X(uVkF<Xsi1&BMCA&@52yqJ(u~qzveyh<mn6^0sWjtL;@reUn_su;
z)IDHlf(NR0KS1un)=$@khSOgwAJNEYPa|Smf1>W0$5@q~I|V9U!QrFCF~TNt9?WDs
z+Nrgzw67QG@|~N-b|J@%+TNu%NaSmrRpmWp319?M0Rt#|PI>wWa#RQd{S6E<$n)aj
zw`#GR&#vXm#qhRsh=Nu*3=7RiMB7W1W1WR}HwCAqp4upR#R28_e8~hOfTkA(CZQXM
zQ<{`hN>nNk9+{t1TB0pG;VPKWqNkvZB~KSWAAJjbQz}oKGPccSNXKxN7?tJu5C_$!
zRt<#i(Pu+xKUpy3te2@AH=&8)$yK~+%*X3>LC?1OxO@ciN%QvN*JmRuo&oY*aOe(1
z`e&!v;6hGYdTSUVzN)dpj4ZtI0eU`x!G*+5&?ME1qetB;1oT+3mRc20T4)*KM0o4g
zh_HI$l!vgR2S37Vg#g&sOifLJ5+&EP!Z1+pd2^`(Z`}nRo>epa+c~}qGQ%c?mei|Z
zObmEUNOrDED7LdX_OQ<Yf3r%2S;4tP{UNgH#Y!R~Qf>Tjf8?-z(HJPAxHvkNmp}vD
zey4L(p!Lg`M8R__#qDuagqLe)Lfq69@v+M*y(#|cW1XaFJjSINo+M>Jzzq`;eC_Y*
zI`OzjCc{A9Bx@}#7V9TC9ftT^QIRZhhlYW{md6;n(+SEP%`D)dG^l)DQ2qiOM3u3y
z<si^@;`{WL82!eJ1ekN?M6z^m_znDY4SegbO<(%5=kCfGQ}>h#?kRE1gn`imz4v-q
z(jHS6rOs9PZo5M;VY7HoO-&7mB_}Yf^O~%aYs*^Em!tP=P6WQG(f^fZ_j01pGe8Fg
z=f=*Hy55(u4U~ZymA<3Z#ELUzKe5&lg9)%-QBoo*Ay?*n2Eo|cFl54tu-R@H8hMbX
z>+UjR=mu#i;(Y!5pj*p-ZAMo@rP}zvl8VAo_H9>R>*UV}W@Gj4Zv}(?WGK3&el<G`
zn|=SfRO~r1cIN1Mf$V52FN^I{&i#}Ruq9&oq((2$o`RVbusmPi&BzfBlFS~~(_#V;
zGn_q3?t?<CqJcTRHE&U(LO@^zY{;Pf=B$FhFN9#~e5!eSixhM1nIJ;N(_io+&sfS`
z?r4v_Is%zS(0G&_@aL}YK{0u0-$jPGsLKI}`~WprPBKfm7#G^!oXXE=y8?Z98k1_%
zg`~E;eBHSw11&8Jo!Sb1fWSBv-GQ+OF6X-3!d7$o`SH<^9eZG|#SXP%wSJe>7$2Cp
zYG9~UD)}E{12r$9->%0E<d$L4`jZpq0CPlm_{YV}zpQ!}tH?(zvIRGxZ(E@Us?JH$
z>H_1u1UoxBgMj#D83tZbqn}(0zanc$0}zi8AFb#g3SPe4XMXYS8#|uD@K+mON!YnA
z;LsuW`SQbu*pKr&b)%0VM;pzW>~$jdb0-qZ7N#%oa`M<C`SL+EVbMM6gsKatXB0$z
z>0ts12pQZ7zNprt0RN9kAn&xLe_b9QWYha5h2#SWQ%9!_ZsM}LFJ$Bdh)CjP&HW5#
zGfMSXxxi$zq28>P>DB3gSzL5N(`Ay-Fl}yU+VSM6LhAn=bMtdZe$qnd4`!)502(Z)
z+nZ70aO$^)>;nSRz<CaN{o1uNmmF?`jVOHm9dKm;YWB>n8_y-4M?TnEvQSZ}eRTBL
zyj#f^s>T+e*-RQ=6|(+H<Z3<~>{82BHi@mk&Hw&zk{PQVHd!<!nNP~7W?R+eVjPUX
z=O5fzc$;!X_;BysX@=xeVF*&q_qPw08`wEGIL@m@E<qxC=i8Qa#m5l)3lzdFJw0o&
zM5<jEM_K{yCi`^1$SfK+;~eqQy9*QwYVx&#CZ@OIQYp^&J*1yGF87@40btrd4X2+|
z9S?&*P>3V`1mT=edz%P>rZDz&c66k?wacBuq;spO5E{}Uffkm9nK%Tj`5_P-64VuZ
zx9+YThu=(s7N=*=EGFbLUOi8vewdJJ($lcmtLCgCSKxrWNOT5fu@BM;4n4VT4ZkH#
zFGqy*wMsJtDPRLvXyk6Xt?dlIKZrg4Y`r<oT^Hm|#aM?8uI;cRHcR$!f!!<TjLMEq
z?(F1^N=oyYa;BIr03Igqs)tOGD<880UikRGku_y!!B9f{hk5~O=N|^?KhQfI3dew*
zAuw(Kks$h?9Md04iNMXJ{~@LRfB2Ijd*nWUadVIU8|U=*%kY7R{Nb$#tlWQ~vHr(V
z68@y{h9QBnOEm&1sj-T)s<g1>ODJbtDCn%TX#rA;@I(0EV9qB!^Vi2epCqgr)Crm<
zi`i(rPKFFrTbxni3gepb+qFShF;%?F;gyT%n}kB9zy6dkinzz*aaF!6Q9t7zv7GX5
zV3qGM$?<>DY=KqjcJ^lfPOG8+a>Ef!Yqdm0B7ae9a9aN+{|NNm0@=S<v%gGjf4=<t
zo&0LnCQkUZf0#VNQxG0!d9>&#?g+4ciukt5f9D1d0#jf6{X3EOuS+G2c!|+n=Ku8&
zz>IhF4>kB7pYhL&Ql06)n5YT;4f5bQ1LZhOhMz#q3#^7!Lo`U%ZqZ$kN1qCWnPVGh
zPg)upN$36|^RIz0`TvHzZcyB9E{^otON@O3%6`f^dGRgtb3Ds|?7N8Q4j1y?{mPoV
zEAM6(F|~y}ai%0?T5Sh<8o}F&pTMcL|8>7#$nl<xBujK{axNEbj5u!;GfLIk2l+R(
zNc_YRiccUD7BCw?vPI93Q&rhSAsQJCz;BdvPk@eAttoUf3$4@2cuVk{N+QgO0Fg!b
zG{Sss2$kxvC#oGKv`9dc9Mp%ezPPE3t^sudRmVxu(p#iu0B^#8Ia#YqlZGtIVea4g
zSk7;!r80s)e~P*&kD^lz`_+=k&$FB}ct*rVB0>1Vooa^fJDw#&m4CV4GTXzh8h{Be
zF=wic<ZIc<HEieX#_ypBe+=Z>s>&_#BR+U8^k0&dC20Do4t`PoeppkVh3=|q6(ai|
z9=51HDb^cQB$#)%KY#+GUq}A`aB>l}d+{1sVND<i8ZH<xEV6flA>CrqD=v}RpRR_d
zN4}uazh0aNfTXv{y&gtVmOeSA8--8zE}R^%^Z3v<OS8yu`cm%ShB(p~gy$!ku*3u1
zmSk>ExrX}so;<TdA$2FHt=-$yXDRzl!O~0g{qc#2RD|`2zZJOkNic&@uGQVWE<}<8
zeoBX8!>|K+->tTGbCDuKkUVcG=x+|Wms1raN7K~>Opvc5PZ&H(yN4#O;|DZi<c(YV
z)yNKq0L1Y(cK>1r3HT*}9c+4%se<RP!#|IRK)~>g?7Hz3sAPZ|Hla{N*x}&oz<|+r
z#ZTj1W^{|H0R}J@K+$T&2?}Ok=zf-bYYZLXx@st5KXB1dGhys|ghgUog6jC6*XKXQ
zx7eD~E6!{K<I8&~-YE!i!T@YXphDHeDp4IDqu|bnWRu$g{yCV{Qi2&?y$r|!n&Jhl
z%ubV%(mbuAQw0rM1Yat|^Rr&QdzC+)!N1c<(hLaWvc2Esi<V|;l%rs=qh`fit8~dd
zBXb-V>axpSGkUpmlB~9dM5SKNs)Gsk51=Xm!p|UPXKNpfH><A(%Gg{oZ0Paa@2C-l
zfyFE}1UG4C=hLc?%;o+l4BdE$p&!dT1KhPXEa&b93#?%_&=MGbT24$lwGFI?610m!
zgi630d9<Os`V&a+K|D@FLj&C(;?DgNASDAN4<=fWn4ERc*x@NHR@KQgZs^f)x4`|&
zv=;6i*8yx1sB8MHpbGLgki+3S2e8>28XU|o3j?~3tZoE4Xq!h!>G@%O-PcO<`}glY
z1U!QA1V%7XgncZ}ptRaks4kVdNOZQSxr&zC9dPXH-vmVS!@qH{^o_5xEc_wn;OlY>
zr+vOWN6&`L+0J!3gDT<X?k*q(!GI_m*eKvUZiwf)p7E9LxgK1Jt`EOf84JgRg)ObD
zghxj+w0wRSPrccx8Cla!X1lCNFd*Eqql7tiA><-J%dig5yR^DemVlas0Sf+SE7U*+
zIuBG>ZXh81n%!@WdY>PSM!&B0hyB9EAHe))^kGb4cmN8-k}*{QUD;*zdTY={;!4Vy
z3wduUu8d_aE?wYkyw@LgDeu*agNAJD7i2404B~Ur2gVR8O7=bsNrjd?=MhKa?VqU+
z=5wCIZe{-fu)V&TwY9b8rY85}9XC0{MmWcP_BAx=-?Hd~(D|wK57h$^H(V1?o;bUY
z)=x}KI$*rTy(h%jFO7@<A_WPb&^-9Y&ruyR$)io+YNR`u-iS@yPD}1f`A}pe^LqLU
zIP%I;Z<5M&z@qDr*B;s4u#gq=W*+a_{C<4z4zypG{xW#2)0Jz&GVrwVUONv5j`Vu^
zymeV<`%9jAlkhSV!nLmgl*ue*J}~1~UYTTl0nTn&smcqym3uM4A3t8!6o)AG4N6TQ
zxvy9CK$#>!UJABpJ96Cj2t;1)`IFmiu4Z<HM~BE9B2OlPtM1bb@9(J=fD9|PKR;N_
z4;Te0W1U>g#8al{t>&?9&Lg#B`vC|I6Bd|Ws@Z<Rtg%q#3*zE6M=)Q$?wxzUM%L#^
zRREB0AG!>h9wjLRERpT|m9)>VEbk@W^*F0xmgje8sL^+KF%T`1j8)>J?<2CBd~?;D
zo~<NL>Gl1^ip`lAH^}oJvlS0jM~HKyu2t7WHMbl^f|nU<TEYDA5jeg#%nl~L(ax{H
zYfNpZ4md`VUscvuC+)eqx!F*1ngD@@p8mPYpXYu&X)dgRi6+k$>22}yCQ>m3@evWp
z0wLlx8xSQ(|2I|GD_3w^35`qE9#V+5+2nadc!g^Z84@(I+-FYr$vlG{*|IDHoaW#%
z-?<*4T@eC&kGyZzj9qD1WYBXR3KN?lPMo`@C;#DaE?7#g|CS;2pp+<>NZ%pPCxt|_
zRlEXR&aJOxRuyIU+QnC$-!3<sec|V{iU)!q;9w0rf-^H`^TQq3(R-qJD@FeJ7ht%u
zz7Wd`uDVS@<BvME*bLJ%fv)t_wz-`P!?w?wVp~zyGOz_;4JA>h?|`@L(*oSnIJ?H*
zJyz5+Ov*0coXm4zfnBXT9_kyg&5bMtqy+$NEXgwL?d{E=?>0oAUHE*FgX3s&q|Kbl
zMYaOcZO$7Vxn+~?#V0wYu8jSl{d68y$zvv7iPrRDyTmUVik{5{Ahvoc0TOwwa1{g?
zWyYtmdiGNT=<yr1`XV;_`hQLxV2*MOC!>c043?Pg&B(?C*Vd}cyTa)0jnjXB`$Gr?
zi~DW85ZAvsxXk!gwm8EJpN-XZLG@s&Uo<z-I@okuZS&~DkLX2%O5rZK@8Estr?3cj
zQ?@zk|LzG0c?7Z}yRkbrVH%KD)$j-&@b-r7!@Ih@KMf-Eleq-(<FfvLpcG@ysrYMM
z5d)vi*3O|i69f!c#RKeSMR}2@A)xL#*L9$P=*w$h?%HN2J9dIR{jderBIVG95Vh!7
z+z(TmCH*Qy<!@1u&1<Vz%V%sZ9zNN#|4ilaN?x+QI-&-BvFa9sL7h{5v|YT{o$q7w
z$46R^=|!@c%lZXoDrbRS6N}ZbFbRY11>2i9n&p8+grW{8R#Mr@w0m<L#$H^=)L7`B
zGYOh3xXr-YHea;|M$l&N1lb2-hlK0HQd|l6TDbMBUKA?F(LnbFY{aZwfckvSgVTvU
z)9b)j{9#F1!xS_rYTsC+ivPviiRSzYG7eQjjP2LW$x8aa>0Z$RskEm|=}(bICC)VA
zsW*JRUi+fh!5DpX(k@XmMDA)_#L7XA#dq1b9A4#XH7@a5K$(^4>H^}(sBj#s_p9Ih
zAAU~f_8)#uMbO&Lj*?j->_MYDNNIpv{0;rUv$NaZc+v;}&+h4+2H^2lc^EXz!@nqc
zz4eAg#_}Ma;>HkNR?6jH>jP<CEl;r7Jzt9>!I87!{5}2}h<W*tF&fj0(%+q#3+0s?
zO?E*8DmH>7u!%m^D`1yMPDv4zmlLQbFS%UUTu$tVX3q~#Km}*;I#Hacz9)lEXtS4S
zUUhp)apvOHtGK%WH^Lf{flbw6pb~`SF!lBu^y>FvJKo~l`zh<+bUxq$Kfh6A8I7e7
z!wGd)A@|rEK^&>gAI{G%zbg9VHM}PBS<$|6e`W<Z;=;7(twJaA3}^cat~o#am?&&U
zrLrsLc<%)G6b|)3lgrMY2JH&4#}(w|JsFjs5L(U1sy^4Y>vu4h*ytO?j@-I&QoV{8
zTTSA8F8HOBL*1`8qy>tPPn;59ph^x73NmL&OzbJJ`3|>JY_k5auN6J0d%$o%p7{I$
zHr0EoXyXEblq*-PNT8VESH>#oevrO}D?tWpGQ3DVw4$e!0tgVGA(i_544Ps|_lhD%
zqqm=RBMMR7Ps=xMYH5``dq$O|qNb)7*_&j4F*sJ-V=wr!x~31f_o78`l@A~T1ID4-
z{J2RYU^rpJowk5!O;a$#S*l`genCoK=BcfQ<sf(!P=Z-)ITke`FcVn&IRuw3J_6ax
z?B;L&uh(0$1}dB0kI#(Tdwj~tnY4idC%g8hot=X6itkKVO9{}zJ&V=%YCV6a`lWgH
zU3cC7xQLAk0<WUo0V`LyS0!^R?v5(KnhuA79x`qkIohD28%Fcamm)&xi7-pISeiTQ
zK@bl@AIG8|<!4+>Y;3KA2e<42OBh89zU%%4D4f{g{fRA}G~W(rfda!BRI*-z4H4KL
zoAaPb=|~pQXhHbs-4sa&t5K2*9&nPk^W_14>S4U#-y6~F&p{^#lCJ{Pna?E&QFm;D
z5Wo<+Q7z0XTFcXYm9<!o-C8}jQJHR@Esh^}xqbRwzJ0cw%?32F<&!~WW$kObzv)Gf
zO%D*SwVzPcMa^yz5oGj#5p<zDlmCaGd$;@cztD4R%^qlK7t#G3gF?cV$VQB5b=Usl
z*up}KCL=ThjVlZ0J5Cj$K>|JDbE(eFCQPB6zVxz#**}b(9Amm*K6vj$dJJy#qo4FA
zg_gZK--$&9xmvUBp2ZJr!e2$Qk3$pe{O4ohZX5^H0jbO9f65Kcr`R1sQiIy*`?qh3
z)?+g(E3nQ625ciUMn2<k9NaJuelG$(d{Mn03hm~r3&X-rg72<cvZOVB%E*O>&8E@E
zd%|;~T6gz@t+vm;RC#$)E<;+PDxMz}&1E?(29Gq#r(mVEl{g?R7KH6!TUh>NVW3Sc
zl?Uy6meU&~*3ee=RahOJ-L@W?&sBQUDniQVR@L+xV*uiQ7R4KMm(2hrKj75YH8e5+
z<Ey>&aKlM{H~H18sEQtd;R0USU~oRYjSM2Df<1#;*7DVox2}DzN~d!N8{Yk$AmV$1
zMUG~VJ{r_Ly-zNZFy0+E)XNTWA7Vp`GRy9)I8k4h5g56r#OFcedQ4Z7s?WXlg^&^H
zD?~zfK%^aQ0HO)3D<@_6`NIu1BkzF7Q~8zrr-|*=>Au7M0vocledw+FdhqjB_IwVz
zQtX@1P?G(AO8Kov@%KvhnioQ9+_t(%AF31KhAZRrgx6Oy&RR^V^3mI*AaqVQsHMZ2
zF>+elZNGf4OPsHz&bv%FwPAWg)rxmY+|$!H$k`*;xJ<0Bp1DiUn%>|MyZ42cguNO?
zg((`=RDcG;-5G6q76%F|=yfyzH5`a+$f|q+SH_+Kjaai0TKB0X7a_(a^!HtW3!4dh
z2i%m`+GlG&p^iYvwQyaJ6nfw7V(6@;T{!*tCi{E;q(Hxb_#a%;5G&Ym$9L#iX!3HG
z8_%AE_Il*UE_bp*_nrHx({>Wf)v2Sf1~~hvCXz`!U*=u61KeV>Q<pw}*hsSR?Geup
zf3C*Kl?eSlNl819&w!DLzIh99%<jgH1ElbtA7p%RC&2Mv@gyny5gBioBylG;%tNN>
zSl`^SlGBJ$VCFPTTf3$R%v@lEnLv6+f6Lvyd=KjyBRQ+=b9{DM6A%11mY2V#S{NIr
zD0?S}+nx3z5n{O%ej@_W1um3#{X_AY6;5D8PVp$T0BSs2;SiKV%Rfe7e`XoDJ(pL-
zTJa4^rBthw0diDL8CZYlZQ)>tVI|2Y4NZ3+6+CvEndx+c^i9b?i}*VVOR@^$BL-E{
zPzcOyF<aKgn}Y;Jucf&D&vU8Zw-2UFua?!a2R_+%<K^~=X~;N;9zJeVFg866oZ}Dn
z$M0XoX*@z~u~va%8kVqqeIWWs6St`u@xEH=EVvmHIsU+3-fiso204r?8hfGiX9RA{
zYh)>Oun)#G*eVI)*P)h+1jz&mKKrBk805_?J$NlpkYp8jUDNaa(FdT1n)L<+ErXkj
z9Rt;Qj{>-bSgPKBXXJ8LXi&64M?SZy=`1c5Dae0;IK1+5di#g@Me|(-Wo6~FpOk?J
zn0eXzPFMNk{xi$9BffT=7&{lp=!r!@SW~{oHv@yRukWjx9ga~`M>{)loE)pz9L^kh
z4aRNYpC%(u^^IwU=gmC}%TC-m_E2N#t6h8pMI8uTo2XRwl5$O}PwTXfd@2UlolT9l
zoMWlW7rwPGYzNcRuo&t{vncmkU{yMc?Wl~2iMhIao-X9FgYe#yLjcNv7bM5%!jx3k
z+^m_py9>2+nir#E>%sgIJT%(Cj;3>T;VDQ3YZtDMMB;Pa)|y~6tS{&IV)K$^3YR|S
zuW9m4Y*Ewv&gD&9t@6Vm2n}VKi?v1naDjVMX>-<muB$mL1cOh>0xT6#)V&ZiL0UZm
zH<m>JdNTYNv8;^e%%khSVMH9WrjG@Q;*(9kb|W~$GkLR9_nvGrqO9(ey71<xHz~^*
z(MbrSur#uUR0}tBZv%;X<L(r*A{8v1Y8;09<(HpUEjvoy*q0SLEim7!N=0~;b?<=i
z_W_w(evKzs<A_H^3OAfUbw6+HR02Pk?+3&jS`y`_y(9*Oe=KJ<uioO4|1FsMjSL!;
z7_=++W|!zs^Cu%;j-8*>!cWPR-blBb8b~wt*7s7MO!-B)Hu_G^f#AH(v>41v^V^gx
zlA7tuj<Tza@-F@Kr?7gEgaqW<wY|C?ah3g*j;tGrp@QmaYNT~$S$c%-PD`}HCp2z%
z%hJjU9DrdaAA#!Diy<E1X78e<xaui7FfU>sLKiJ^%fj^uUp{I1K~m@9HDANgZd^}J
zqX1%D3lB>}&;~(p58ull7pCT4elYl|DT-lirqS`2WR5!AL?5c7WB(uK-UBGgtZNrF
zVE{!`f(oJ_2s9ufIV*@n$uvzyqGS=tNk9Zaf@CBoMY{=2lcORzOOPy*bI$p$cD`@s
z`~P$PbE<CDt-3QcRa6hWo%Xx;T2ETQ6~V+(tc{ectgH_qS<a{$OT|Zr7yAegD)S#+
zme<2Nv>r~!n1k%%@<tmj(>>butv_~{Y{LjN3|{Szzo|DcQr(e@yL$8H5%iS1M-h^A
z(>FgoiG5T4fGG!<ha9E6KGVbd!y~kNFtZ+yYuvjUJ&ZSS!64jr)alYRpnp!u<j3<F
zMO2N>`k%XPH*g#F=!u(sZz5jaK1aKjJo2pU#VoHyVB?%L=^yG%z@u#9&fxw`HnA>Q
zM3Daui}tOvNT&Hr(lH5qW|<DJOhtWFwNynTJA`8$(w`b8Q(fLoD!drL@w)W~2Zy=p
z7Zp?RNHz0~04YdJ$^9-S8^+EOT(8XF!g6GySa?}pB54wgb8!PJ?Z=N}3CYhwv1C42
zWCJTICYLP_8d02*dVRwogz$8H5qLKJ1IE)sX9J?FL#U(IR#))>FREKV44hz0?vd@D
zIr5t@wN*xN<i-rk+ERGsB#FSU*45ibSv=|8XPFD5E+Mf)&xzgWZfLIRwq>gm?B$uV
zEX>o)%`q=@zb~%6IxuZk;h~1w`SSI|$&)d`c>rs%YUE816tYRHgSz0scPR`0^c-o1
z2BjL)sk&@H-}ZAv7u)h6poS3LYw+#cHzWh?_ENS^_EIl)Op~4~{HC<Rqaa<3{?-{!
zLd}igS9lIjpG?4Qe%xwz@5}F1RnM3fV@z5^{N&i!cl`-wlR7r>WSWSYr#{+zA{Y|9
z<hKK;s<FCap;Li#s0ZO(9*3@zm&5fJV4IhF@?=7u<$2Tqy>KQJc72EpF!5lNJ~&$~
ze1x<sCMWuKSJh`QIr;mS?|t{ehsjw`dgHfV$4_R@I1^S-nP=jn4349hB`D9GJD2)6
z5PT#^>?q6+p8=WqhO<!5>T@|3uryfwDs$>BxPwN5J<r4A2&SPv$MX7$LG%$IMG)2m
z;Pk>m&!9HhL+_1q*tnmsPeFVRlaz6*^B0J};$psP&NWb<sb0?M5g56=rs^r=w##|%
z9tm^J+r9zo${VDl;NzR^SXJZlxZ{o*39I{xSa)J*eCD_Y{5KQ@p>)<mCHVmRwzn&!
zt^H4w+ntLhwZacc7&E{U^nMAbJ`nZ*IyV0^a{v%mH4F;+PLF*EBaluHO~n3wm65N}
zncoBuX@5pwhQ#A@H;=!A<Don5<_`t&o23JeBP{Q2hWU)pJKquEKb~E}4a&0Fj<YeF
z&K`NZ6Yby-Prj5sGOnooG90vv8*DTncxjY4#aMeD=_A(zM8wTOYw2Q1L6=aI`)`%n
z7;h(cy<Y7Gbtn+z19_$#&)RQ=m;6mWO73&T@G5<lMEQ8sw%JC0(}jECe;`@Fv-NLF
z*Z-C%`kTig0Lte7NbBIaGGL$Zzx7w7BK}XOui+p3+eP+Yc$~jjsQ-mX`hUe$y=eKj
zSx!%j@^2dL?;o!I{nOt+{PWO1KmGkbQKndbzos8~Ac|^lfmQr4lO}`f)B@28-i79~
zO{UMvi#HtNZEYlf=rup!^ZxUZdc0ZgkApDDjm91Sxsm+wt>LpMlmFWz_=HJ4vS+&g
z{Tl!4BmZ3WmFlhk>s9eCzWTQR`Fa0&B_WA_1!?}Dt~zQ!=5-6eW_Q?9U%!40(n(fS
zjvmwqFc_E*{)768>-vRiuN;GK7IzXvIEIS!c_WhssFMS=vV(`M%N&F1U85xwLk4QX
z=eo@7q@|@@U0omE0=Q3*pP!$deFZk>)85BNDb}0UHq-y)O@_yVq0PJy)4kAZ!F$=d
zBve+gHHp*geGQwqqPn_zPqwZIL}NnGWK~raaPK{^nFBqZmX-#d+Ds%7MoAA`Zs5(!
zR{qqg$$cTe=P&B)E;*yOj&RvhtSdP2d?<;|Dt5_GNcMBJS$4P2O8oqU<2Jh;6MMUQ
zFiWIp=>lccq2dw0S`l7*`9E@F?HJLD`ul$ePMSY5Slt2h5diSA7r;?V_+FJU_1jA(
zDF~iOPffY6`b~n}Ica2jL@-RM2rr&eF!3+ZCVdAo)QW|KLapK#&_9jktxbraCDS{g
z!YU~$Dl;m~fA(V(Dw7iiWbv&9+W`j<EC!s40zP4~L#yr1<~F10Q91;^1WlrfI&s`D
zhA6qS#IK*f&Z2;P8yE=0*P&)mFYI@xx}x6qvAz2w|BUA~ledISrKx4=r3-bJ0p*Xj
z(g8)P-%N#bUz=!S+kZInk`*HCx45Z-nr=&d!%9DU##j|xNL5%HNC^>1Q4jGNi{#XX
z%y#Gr!Jn~K@y;=h2Dl#yA=TD6Nt=5UjHF0KIVyTy<U@nJSFc`OezXGL+v_R>F;g*1
z4xtfeJ>*%Iu(2Q?o|_6h!R^Rd)53gtanu2lPp!*6A%Z5zFt#0xMU0ijB_-rsW9vYo
z0bcPN+V5QUiCh?O3x=~zBg0LZ&xxPx95hO$O7kM@kNCmZw>v2M`<8Szb1Wt$cz~G&
zS?sVxP1qeuFLWI1U;T+Jhj56~Jc?qX;w#Fi<0^SAg|ThlQ!+!4!Xs`2lB5FFPBhO_
zd-8XH+tZXxM|fcNCv;;FFSPoHLxaY1Rpe;@wgSi{Isc}*E~7*K5Of(H36`?LM<2vs
z+Gsv$i@!KDG>#>j+hGbXN+Hux?1|J-l8OEb`TX1@OV*(6i9e1%Xug_1X@Ft~fU|}$
z28g`W2G;CnrY0P#=07np+?F4*v#mi?|4J2%NKnHDTz}^0sVW#-fBGf(CH}a+1q|in
z#STILWX_YN^p5hw++yR)YxL}%xs3u#W!|D>*t$@98T-XSZ~!&{fs+)p0SaZHP2ivd
zO0ZZpJ-PiZ5#}-_Wx!AY_+HrA*$Z$frKJM)s?fL*G$NehSd~V_Sf*1_mmNyA`#)rT
zB8}OlPHWoT)zBJUG^>b3e6(Tggf0j5OPMX2KHLfCdW&9)vq(>=5h`@c>~CRB6O0#}
zot(raB#5F^!8sNlc0%zeT_op|1@;gB<}FNwSh1lvi3U>J>-5|eeap(uhfcDPOc}(y
zm<8;cXhR#=v>sesgIr2}P`Wh)kyXJ!ToV9?+0#|Cbq;{UNiiWSnI*IUg;I^q#%N6d
z`1nXlmjUJ3Z?gkCq~xFYNsE$O9tOD!=(Nk96QgRHXS-}hjmaQzA{rsflXccl6)Xl=
z-<iLxcaop#CgljT)W!Sz#I6Zyl+rXq0N&2(j`>W5tIkK1I3;#6AE)yKqH%kr4Mwp3
zYRVf`B%>oIUDk?q7;avA#Z}_^T;6_*;5kJleFo*LoN-Wp)l_psN{S7x5wKyb3!X!>
z&+6{0a1*d!03-?ovt&e!F$S%(1=hLDUictNvOS#d^1m!~Ww3Ub!?k>6|HYamO}(a-
zOs|137yI`|Xcnyh&Y|{hf$+bra^N%Nck0hRGpQ0gqv8Lm_Igl1d`uw9lBXCrI@}Vm
z!AWarQQ5ZvezkW-AI5g4>CTOfoy1h7q8?~Jk18Nno^VeHe`q~ykTPxCv?FER;EK)l
zXP=R6@C#xnw!&9-&W_=Y(rRt#szh1m!CzPJ>%R9P>&#2G40c*jjbYh`Uz%4%flf3e
zS>$A=rktGI=ed|tRByQpkA2%e4B7%T@rkH=!-tf?!}(1&u8z$IV;!Bku=GZp;m$12
zpMuPlg*x1qaxQ(jsL77d=TDz4QVVx!BtEvX=qE>&X}{9WW5!;h!*YvU!gLiqqIW<B
zK-LyM*J^!pd&S@qt`Se7xqSTeX~3hnvQnhL4kBGdut{LHN~Ad0n~hC#1Bw=oJ|)Dr
z*)b*$D&}`iIFAeBXa2>YkwKvG3a@vCt~_o(s_ICUTb<w<as&w(31IdsK0OBTS9+Go
zqW(zNXXRcYC_9agCO{cpL%0NrU<PvVX0yvPWg3`YT^1r>u?!p`L*FwA4m3HuOoG-M
zxC-}Wv+pBd?UECu{!nZbr9%CkV<q#sjjyaC*ANB}Jp>%X$AATVOy<nXKTn0{NMv5l
zv*8{ns^DqTp1WE_B!Ax2AuB0b<@fV#P4cs_HEOtGywZ7Lw|r$LHB{k7>GEdYO6W9l
zBssyZv`{_CuH#xjxDFD(&b>?h<H66uZcy7^vpEr^yZW##uGT9t@#aAS#H;14RlR<F
z%3?L9=1Ti3Rg_+Yk%*>a|Hl<z-FCzEB7_swYZ^lnx@CMiD(WHKSJdOxzx;Lpse_G6
zRK!t*Yf7T5>24TaQS!B4o3dD^8WZ|gR%S_WB#2wV>2;k5Bs9I2kie)tuZoBO6(%Ib
zu=%0%07_E|g@9Q7K?UGa(!xqH@+J`d0%?ykW1Fxmg>+~o=QfTy^@b|?s;3uTJaiwo
zNQp*?mp>z(ytTq8_5jlX1V{>bVQkA}SC<ND6NgFbqp|m(*ypzX<qGy}yrVxL00yR0
zf1^)<w5b#mf2DUfR-KDo7#x%k$B_-V67HpvXIm_D8@qt1cSImd?Ja#oX`GP{x%Ivt
z+qFciqU~2k#VLn-%<=&+PSag&6<9JfG(1a7F8!Q3M1TF&zEOJIjnM{0mRo~WCgNA2
zf&0ZUbGzt>du?$LB19ibe<hqqenYOXNBh0;i){40<ln;}3XJBWX8)HFubHZB8M3D&
z+jda_nNDy1q2h*-4yH|u7JuSG?nAwGFTF2|gY6FTy>xo-dy~J2v{KtC;xHxXqwyZf
zdA7b{+fNGm$<-zljlLUC3%dk+ut~u~fo=YA0uv8T6h#?NU;-zIWO&GXoA-fj_9dep
zWJ}tNXfW(O`3M^W>6Zh5j1M?1UnAr+j)qmP&{od{s84pW^S&f{g0SfpI_Ws+;fZIl
zA+xuieng_lI9)&^1-gf^(NSc!&JeJrljo}7TNn>ch$kWH>+1ubmOgbZC}>-N3f$Y>
zb#``kUu|H3#JjH*-iIs8Bi~Q>`|$9a6hATe{=+eO-RMtfWfC#Yehk5EBI_$haRCjD
zQ~~3yv7xFTZc4zzeE$5oxv(%fIXNhya~F<r$fqKrcW+`uD}=uo&FImv?1`&)Hgo%}
zUW=ml*ejS=&j!k9;M3WaIJ~}k;oTsFszQLFpPIA1hE*jPh~V%f3f9TD)OB5{frGlQ
zpF!0hLu<NMoG0M|Vgp5Pg?BeXE6vgknrLlmzJs^H>(*}Wen0Loo<(rO%s5q;Z!;Xu
zKuk^}M$aaP@cqtN5zKNF$)bA@a6m6-E^i(?dE!Kl$7kq8zBOb5NJT_QD0-WCl4hoU
z_Ge}j@`sEb+1+&zujV_FQI_XtSVuA;lg~TvIFv7wAI<sQTeX!^Plt0H(C-}}<@q#T
z+&K%0v=I?xmoIx+8X9ZBC85!kYK?sUycI!h<)8Eo(m3C~+Y_f3vK)7QJ>7hG-RR_M
z#J$-Mv}a`gNv|B$!)JwW>zBp$fM^Nt$)hDzlr}V^t8skt>NExT<qd))NTwivO}UM6
z0MF5-eh~~k22~(?cid+!I>rQ?H1rUATT38Nw&uLvbO)^k_YZ*fHA;j-*SRggC-C_0
zM<lNySt=yk28Ea9h7E9y$h^=jeR}DevnhlldUj-Bb59m8X4M1<z<I5guGPnSMr{`(
z`kR5i;lzHg5T^ayf(w9)DG6z4^c)0^TWYz04Gb^|12vF#Q=xo(osn`kBoceLJ`Er)
zDB7cZoVZIc5P6Awo1i*9MV6&mn01PQ3(6FTr|ZQU#B4d(*z5o*)(<y4G$A4)l`vf%
ziADzQ<b$v)sf(eJul5ATz7vUyu0l?yIteyOq$qgjx=bkz=64#$<17Z)V%ZCi+GqJ>
z$skg`Exu}tXQ8)whK+yfhJvRmn3@yr@3FHV1(&Y{1&q!fdZ_5ju>YVFZdi++qI3y9
zbxGmTJ1;-!d7<lbS5j*)6J7VrcqoqLNl8)6@yfvJJ)D0Wo}oxe#9i%Wv1DYtJruG3
ze5*BgzN&e#d7dwO=k~YwJ|SC?sph?3#an}q<{v?rtR7j`jC$0rub8HD@j3Lf%iJU1
zhvLiL)?IWYt)%aRC7s%0(Vu4uYIsXHRX}BpX~e){ED*Tk^^`1>4y&|N4{{qANx<}8
zayn7|$&+;No6^D1e~J1NG^@x=juFMwogbB8LA5rzT<ISYV4Qkm?qJEG)y8S8-Lp9K
zS&M{b=E^%Ck%GPw{hxHlbNpQ!kCs~q7bs1}RQ8}}`8jJD$Jy|UJLSy%(VKc?i_<Zr
z0nL44t1Uh0i(!kzn3N}dK6cdN3zeL|DjKr4S-q2e=fT=B&GuZ;_8!Xh6%Lb&hjVj5
zE5c@%0xT)jW<dt9(i)k_42+l3Yu-E|9Ud08C<?4zqvMbOZ^^XO_&`@k3=Q#WzNA}a
z^SA)|77Np}&ic{7U*SjCgLFAcwUy=N`lcq_8X;((YULy<ub_McoRQ6O&ki?#K2Luc
zN-Bt+4g$1y=qs&qzNJ->bBx1^rg15Jh7`K4o^M(2+<DxKSp_TSe7+ofs0%Eef!$k=
z$#2|XRY9@`@-GhhM~z8JkIS%x?66_!OA0R9gan%^8~Y=cEl1l)IJeY&bOJr!>zB1!
z>xSn}Z{6eN?LA~BIpwJuwzj^03OVAJurX7|l3|dbJ=gWR1yPdMb;7!-&%5GDwA+m~
zHzbp*_N7I9HW{MG{}FA}Tuk{OH8pjR_S(3VOOwxgx|xLqOLr^ys-r)cG(AMp)DUbV
zdZzhY{Ru+b7l&^z_GhUmyqCYJn}<8LgSc6T3)glvA(L9Vg_P0a&va8v^m4y=V}VVt
zDGp?7((LSYS5fJ4&L78>)@Yur4~GwC(axTLxR_60A3q{;y*}5L@UlwH_W4p3-c{gi
zU?y{mIG0db8b#3m>eVsdtmn(WAI>w-)6+vpq^#4|7w2&MpKQtk2&A*c+^%%zKI`*y
zryc%iU8^YI*!amkdniB9aosrk)WJ?CYLW+%{mhZ<T5Z+LEBD!z#=mT)%NGvhH?c>@
zaa&%uCrDR1hZ%h#WT0%;J*RecX|Q1GtC<*+@sldy2MuYi7P<$;FD>2`el1Nt99Nv0
z_6Pd~^)M)O5vBt7Jh6=cd|-*PH1ZphnXd*u#^iNXzr8}xjOfj}vN|)ibAyOZ%3u6Z
zt5EK(;aPv}?l1JcE>GvszT4jsjm0CqgjSnBiFuRVW(2$ex8Qy0gO3kZQ{YfsyMjs|
z#JotFpUAhQ?k^W`>Zw3JNVgLk`r$3tNp=Q&m%t`AZy?j9$sQvTY(oBc<FuCnnherO
zgBx>F4Hy_-N^E1-lFtpIP<L(C{rGBG)oW^4h=vc?*F>>QUgvw&Qyg?iT)2-(QTn`1
z2>?}<0a@qtU1G$AmuHmYV$K!EB;q3@q=McLMe@}n)_Hh8$;H1|r|k`oNJ_eO97l+>
z*AX_tyfW~th!5EeHq^oS+YeI@rk<24YqbNJ74uXOdQt>ae>LU>41U2>ozKt`G(Y@O
z@{C2L)VeS%jD|<u2-GOi<5&BbyuC5GCT-O7UZ$qRY<|n@>ztKd5aOwz?&F%^OJQ@c
zVRF3g#A@pc^U6qB%Y`Wf@_pd6#F}KwoYtq#nH)P;=v|4mX*fB%_Sf;Wa?c78hL#S#
zbbhu~^Ecj&36G{2#H?~5fQaP|-`*VAa)9b|HyLp2R3#@Kl7LV1UZtp{q@RMic)2Mn
z^+toh>+3{@hcXy}R1LwXrsDFKnZnKA6m5U893L;C5qlFo{2lgw{@47d^n<ToLB@9%
zMsB#h>yWudyw5t-OVfKs;ns4B+Xc15G>Z107Q^$QuJuol-^f>-93Af|Stoj9R+mS_
z=DmQRTN$gvGRr0is)t}i8`4sObgqu1Rx@^?6j~DlGcV93PO;nBMCJIK>M~e7=^R!g
z2${=KH-X}(zO}VrP0Kmvz=wze)E`E1?O)CrwB(LtpI0oMl>2qoTNl4hLFaC45PaV1
zry7U=jz>=4)wNGLruX2%vCS|bL1-nBSjJoGoa#nZXOnBaRwnH%N*_<8FuG6Krz%p7
z8(>@16ZM{s^Y3=lK)C)uw=IJooOJpJ_djxzAIF380vTFIGULOS%_TNI=HHcl8%Mcy
zrvYSlvesBE7H)6ng+6jtjb6TDE1+O{ive2TQ!c%nLrPSu$<vq<fA?TP-c!xZ*kxky
zErEeO`6h~j!1=w>xoq3g!!Db$-~!83y%$3uoXeq>fU@ti$q!xoi#Pbg*$%8A93A@3
zxe~@C;Pd9xiXtzd{D+}z49{F21!Ls2<~DooSw9?xHmXTvXZ*^1ZU(Egc42gQ`0EP~
zX!f9C;-x0aeSea!+?(^<ioLLFLlhe+$DQ`6PfV%C%hT!~<0!29rP8{RZk=t3<lPm}
zuQeGG5Gib82*y7D@+38-stiJR%5{#0j}CkG<=jEb4QC&>9I$b&aPb!T&pPG1V|BrJ
zkCj$mRGBT&082Ngvem}rs9=K}(n=6R!KjsUC7zp6bsnqZ=eVEvtoX8-Y~4+(#63pB
zT?U;uDn7#&N@Eb&rdCke`$NO^O)K-oCMtc~l%A4D7H8*uxNq7id%kIkcL$-H=q}60
zLzU>Gv%Z@oaVuQ~-UE{KaMbE}qqyR(GDD6jcet$oVTJ}d>r7vRbos&OVT${TFNb`J
zhw_wPO;FRGC#RaMFS;h=k5>77A2^!bt21Cfctd<cIv5}9xt21lz3t2<H^Ih{Nxkln
zpUdtoy_+-QeCOr{iyrcuzi9MnyOSCbU9Ra=?<xw6gH`>r-tr#;fd%~&zvJ?6?=u&3
zLanC1IZw6TARPnQ_Bzc8Lwq|F-P93hL}m!kN1?M@SGO4y86eL9ToK%Tw?GLF_VP}g
z){kRd@EgT9iL6ls0{xtq9abaPXIU11c?y3b(wsIci)wM47XZKAh01E}eY$e5N1i&p
zcO=N2U;q)*8*axJPpx_^)Pt>&)8Rso;vgOTaS)YMWjY8P&3~kb1H`9r7UXo?Ab_Hj
zW4hn9r<XRGZj17%cWZ8*)lTRkx=fA!y>~fsjKpqv+<l)MN9b0#9xS|jR@@wbIUosq
z4!1HxTRJ<HlNFxRiQsM!X=E0N`4U@4bSnt6h+X2d!urSWE=?QtyG2{K-Dk$=*Lmhr
zrT474H^-&qtk9k^mFvZ{?|=yGI_#R7Z?-R-KX1t0a~uX?97Bius?PJALcHtp7$ZWH
zR%GRoT)Sp!=h{#JDCAvm0<&Qi?2)EJ3Vhb&OwKiRCNZfP+L=okm5!0+FK3$vt+tuH
ziLxu>xgDIL8b{gjQob1#LLj8uPQ1qt+JNhufy2n?V>m!#_qe$u-+Lh-xQf<gcYP*~
z&lYAvXwL?IQTeQdLiNH<#q^PZBj-BNBb_X6Ctv;AQ!FW#rozk2-Tva!j}C>K)<1QI
zD{{=+sCO`+a0W!_J=j?chixhC3;qiAkuV9^5DZ~1*=>m+j7I*K!0k*Fb`*}O7&mM1
z62ZTt(V^!?fl<fDY-zg>usJuY?GT(gm2~&q?Y-o+C{=&!ThjUvj1JLmXD?n{8m_Di
z4wi?}K+q)M2L}N<0MJgu#0%u@pbfiIq(1i(7OO8k5iK+8)}I)BzBA6=@wD71^)!Ok
zT`5~LYpAd;#Kse}Pw(S>_*PhMOO2IVST^1rkALxtWhXc_UCgDimRPYlAw{F$L8*Y}
zd#}j2PTaRw?M?IT6eU%0m3MB4GZLzjRs!BQ2cme;tU?AU$)!v0LqbB{zo&UG=}1+X
z;cfWiqmdw?l$VF|-Hl(<8|1>dYb9ann?`BseSR+)s@e2|!R=QAB6SG~d1&pehbzwN
zL0=&NtDm$6Yz!hW011gCck!!|K#q9X*e;3E2sLRvv&t|jG}Cnm(8Y5?|1fU<dp7DH
zCJE0p;mups5HN#xf8nzp{r=dw{_7+Ef4J&j@s0fyqu?aOCx+sUV%2{b#sB`u>L2eJ
z{&(=djW@5YFw-HAgQ9S(NHs&@n^-$CbnJxvwR}@-EEzL$ouq{3=6~dCYW)xSnu<l-
zrX6YeJm0IT-ih+a%eU!!7#JA%EU-+gFFSK@6+$4K#zRdhXY!mT0HU8KZrbs9^y!gB
zJ6u?seL8dspP%{1jHji8H{&5a@iXQA!x{(UbYLRf-;^m{RD(A&{_VX3HVy{c{U>D#
zd6;5WzWA_BFy}pBIfni-h3-H6eV7yY!FJ%b!*i<;fO#4pfcdXifzQMQLJ(q9|I2gW
zGdKTw5dSIv{hIJ5!vFpkf9!b2Ao%mYyr%#9wzBgdHW=0JIjmE(KHFMQv__@x9PkW{
z9K30wcq_pd4oStx<SsLVpLPeIuk@!a%oas_$fT5AG=V|fgP+!&pN`uD%P8c67asxP
zj?t|YMj#N}-1e|9f@{CRVJSg6-Ak)Z*@MZyv8ibfL<^9NQUX0gx=hsLM~^VToq-Vw
zA9vNz01=-MwAO?B0lc|ck+XV~^8*~cT)9o4{SkFI7JA!|6L|f^J+W)QVvRxDKa$GL
zSl+rk#n;{OOsY+eK(`xbQ~!9q=Y|mK*}b#xR2k!u4)Y<<m=d~Pj~Xwn&yI23clfXn
z`_`9Fuw?-}h(Yh(VYEv(fon9K2UzN?sbkcff`D}8f@*?TxfNgKs2bhn<XX{|et5C4
z-L~5|K@4{?1ma`3@47huTpZADt>qF4uRLrnHQ^B@g+femV3K}<G5l-y%3QsdE0|##
zH2`D*D2$MpxR$!eufbE*VgL$2{II2?3$O}*hjE`h@uP@#l;OPx(SV=VjnS;b5^p9S
zI;Qi;AMui8z6B1`#M4SgzbVjuHKVYjyEr=jMJ>__3B{X0b|!CE(7G7A2_kqu(9iF|
z_DbZ~PlytOE@utof|P1eI2+ay?bI`?(|u1qt9~I1uvkxHaaM)lc6XNatl#-qdf3|9
zf(P3DaAyiOLO3Ps-@F4q5HUudXeqxR4tC9h3;X(HJL@YG=j?Z0Um_I{Ym_q}lfDxn
zzn6TzVvcxW>=IfKN(R8GJkYY_SFQm2pzcu@NV$dNvzYJ2(sc?ER{%h;@8IfWe<k<@
z-LUIIj&~<Cpo!`mg&=C)B!VGo1>o|=r44`OK*sR;dph!5cWOjbkNJjDdm7osNnmI%
zzm+ER=%UXCj|=qPM;qB5eOF)XfSv<{qfj7#G-^5lk1EsznDFeZ`ta7*l~80-`g<Q#
zYY&h)^5hBIbG|s1-VJ?(v|I2wjQ%I>;ukfKkyAqGZO+JbBE{zDB%MMnKvY&BK`fK9
z<n~_oTx3HnG1KxF`HN?|Lw8&a$&^n|kD3^5bGFNH8|Rp<2re87)jPl`bTGr&(mJf#
zlwvgYhHRUAgaV(K0l?^P4WNLbVPQ4;HY*H)j}5>X3Z*UtifPY(hn}z8Y3+`@I)R-e
zy^uhkM@A&;Z>B$!^*lX2{e;MY2Y{qQ5Zerv*b#9Jj(qn8#1&xQST1v8RZ|GUEz=`g
zSUlzXp=19+uwB4C>z=r85hqxLOYsx%ZYYLVHh1X??!6I%j1=S3(lyX+_xeVx86QU%
z(V(X{cA?BC^+>I(^ngO&=PN>I!zlHP*+sfOY7Wl*2t2lPCP>Zg`$T}C0a@sy!wx0_
zda-aS3n4<X7nZBUGt0sy-}j0W2i`-6|1`=CN@zeME#n~hRhf;AZ5{&&n`_~!gzB-6
z8VNi+wnxs8vcOHi@3x!wE$G@3WVzw<Uw;+>fqn>8>-K#)ROF?u!8(+P%r!1=iUqWk
zU)>bp>Ahq@t&XX{ttClkyV_M8yFfq>29V@LFuB@+uIy~Fr|NtjScgZm(J&vw-kzI#
z&DR4nI`D{(Eza@z@Glcj<n0?jDT!{*e6?zB9n+6h&D*mdyL*1_$z<ir-<O-qcl1@;
zCLMC?LHj2oGzaVtqlg{@Y-i*sAM?Z6m;4g)4^gWgpx{G|$;EO%`3mNyD*>OYo^RaW
z8i)$_1Nf@?jkhdOm}Me8`CD&S)?Z!n*P7UnJKJNykkW&{r9d1HUMA9`E4u*Mxy-{G
z1_JuUN)O0(FCM(_!LL~H3dhOkrzR>%3EX|^+>KA`q9Av&I=+lm=gRs;@ityHdZo>E
zh^aU^{{h;NJlneyK-%ZnGCcWk3RCrMQ0H{%%Ncn45zUla(A^a3sr%4X(^#^OmaCZ<
z?;%<8vz-@|WSc**&MnaPcAe~P_dh6CjGE(PW6MtSIDy1wQ(AJ~ylVZctoYE*=$;Oa
z(EE1SmkFYsD#WZ^(=vNUY>&1hM_(t$Y=mCVyMO@PnwIysSLMSn;pr6e(&K$vF<=EB
zySoe5ZD?_KCJLw)FjNE#P|zTAYF_i(gj`JT?58+QI_sfcwx)5F>6p+-(8&=|++{gT
zfq4}00VRriILITD$M0KA7Q2p0>f`|P=XXLw9kUnB_amPtOh_p8L2Z`P#;od|JVZU?
zc8bA>`V|4nWw?W$_B*amzrUqUz$X2BC}c&+)FLl?rIxEqq{s|=@X36OPNPh4_P5lG
zoof$Aj`Ziq_;|3>siM*QgJFNB*BKf3E_<nRzKvaZ7_sN@x)G0v_i`-_+I;QH$+>4S
zLjnUMN23mdmH_6V>akTn%}`B?asS{7aVvS*#&zWTLVNkl#Jq>2*X&D#53c6;=fH<A
z1^Du)_0IlU8yO)%j(?i2-$YR(BfPOqoYAny+^RrBLIMndc&ohztD)d%9USm_K3lMj
zU8#f-y0K1UNDNBxdwV#m9G39FG9Maicrr6n&dnyKd;?0mw<pR`8oTh-KS8QjxEAje
zQVF~E?&}Uks^-jmZbr82zg2y=we*{Qh<M3iC*BaMAnesx(PGO1ZkL<8q#KYR{UK44
zGlmK!esx6v27j&99>5^I;_Ru_ttZtFg)O?W3{%}Ah-h;f)rj}3)Mt@{`QZ3_r4HdR
zpTPNBqstDZ?)=Hc1_W4Pd#^k!x4E#TE{-!2+MN8LR1Zj6y_c6)b9NtS^9AGupZ|vN
zp-tEh0Ic#GhFIa)Ph5u3V?u|9idn_#U()SYyS5fk6Kvg?5r^>#x{uifB$NQ1lw|}-
z>MI-O-Cx8L5lbKEUQI749gTkbU<5}H>PZC8(PI)o>9l8$hqoAxy$SF>pgSl*RCzo)
z|50A#v7r02(#KRLwk{eP8rM=ZlOgp9ioz<#WD|b$)kDEgI;ASy_I_y&AsMMrkfXAb
z#{R{KD37T$<e8L>XDQYniRSY795@e-@W=*30NEz?`=YN#57ESH1I6HNhoLei>pxAq
z+i405(GWiyC!&PeUm(~2hhOmB(|ha#E=F%vbGDE*bK451!-$D-<t~0{5@+DJuz1<g
z_tO?6(amgoAc@f~dj73;s@u#CTIgnZe6BAHVsuQNx<7x0j`Bwv=R!#Slka?FmCz}1
zEqRc|Eb!L~8j+=)!h~Emdq$nlR%Pm{@`&ZdZ`V-Le9IuVpTA0uqr@FCURFenvo#bK
z2?<j$5f@*GAs>3x=^nE*F48$piP6_E2^3%4Z1Kr9@9p6E!8r3k^1Fp9J;Ss*3psw=
z(tabL%Fqm#+s%>3D}lfBF!5;r%pN@`;wJ(Un#vrKUp|wsm3d!E=MEtO&CTF!wt?yo
zHaaNPVP}r0i0T8?(@wgHi*N9uWGxT#V4<(X0{TgRP_Ta--^>Hw)kRBYmqONrIsw=X
z!b4XW8(l9w0-4%P!5p2w-XWBaW4QN7Nm0PfF@a#*Hw1jmuKJd<$2UqQf`q#s9kRaj
z(4CNE=`{r{8r40$V{Ty+;N$`Nw+w;~VRHaejT0P5*tR4QYC*>k3Fo^#!r;Z537VgB
z+27sU8z^4V1bXRcd+g$`z$^Wxxr!&TId(5^1ey;-5mWk0p4oi(m05>en`V?jj|>1~
zU@;d4MQbuqJ{_j%-X&x5`(wV>j%c}2KEgkU*)IDO&CUJH=21^TL*+bCrVc%;qJOGy
z8E<bJb_HcMzS-Q|42QScXQi5iGwQ(G@mTh!JvEs-Zj$Q{B`I%S`OD|gP#k60Ji#IP
z9V7jie_gl`)k+krO7b4kx7XT}#ZC#}f*?i;%AsvzAKUMJd;G}$s1&-VH(0~x2fA8?
zniNR=1PZZL2#!0;a*%k!4TfivNlhV;>-3#9<I#su&E^EJ0FiwO?g~hJ+u8dPD9n{Y
zx==B+t}RNE5aY~P@nkygd);6{R1^hJ3ao}7V`21#hQ{@Vep2F}>NTaw7V8gA#lbg$
zzIj_$P}y4R_Ts3C>bY^W>tU_+XN(f4+bi7m7NNs`7De(f*B?4~SO1L-*#xY=>5#rw
z-y0-`_{+WMEKgP5D#YXjKbXrvPi8#!eepYDy<D2UoL^a*z#28tnt6Gu#eI!yXSF!c
zRx#!K<V=jtO<lpcRn+rm3hAI0Ojq(R2J#uC8Fm)uGjr+@&hgG;W=iXgtu~_Ft3Cvo
zl%uyAUz#dgRh-=2syqU?XZHSlSF;_GW;y;PfQQjTn0CGena|<9FJ7I<-c1iOpv=;#
zmEO3oLrjeFA;RFLX=~<CTy$<u4ZT+7JQusXXu`0h<iIdRj&6qlP6~Ah;H+4zcX#EU
zEG&0?KM)~35zTHe-6KJz1DVvY_k$S!6afL3X-Bo|eJ3nFbsY3dCpvNBP1OCq<!p=^
zGtJ5#aKm4>G=^l`Ij|>!B^dTOBUNJ{ZyH-#$S$L&+v2KCejFWAqKE5s_RO{xK30BI
z$=23*b`M2)myT<Ir!(rP?3!AP5CLrCn9#&T>sD$=Y-P(=p81PgqqkMuV+$;_&fN;M
z$npPN`jpjORR#t$67<y{?=AN8;7u5i!cWJ$jfVN;hK6a6Xp=pJxws;Jv1dblVr=|T
zZ(Wur?YJw=TJ*!JH43g#JUMqIQlaQt40>|*1@gW8&dmz-WVbO3Eel=$(BPxi#@NQi
z8{xL!^Gge^Wtm*ZWtnd+s@(MpSwB}>5|;8cn+0X`lMY7L<w(Pnl@>Lrv`?R+oc({R
z*Kj7Zmv|Cn<XHwK>|xwLuhpt5P~)siG?MdMCRAlo`4R@yvb5f67XS~p9{MI45|^J^
z2Y{<8rQ3`k$z<;0sdyw+KXo=`WE8Wc2jDRWH59sfU;H3;_w;|`05Jt@Oyo58Z5)eJ
ztd+%#-|*J=jNWKx<AAgGD*~oVOsY|GHD<MJ{~_O0);SCPP7L2^CJq(|RQ2=aL(aZB
zxHtG7^H484INjU%IbR{9t6g@<R4%mA{;XowRt#lLwY&Vwmz={J7T-Aq3P&e+LofTx
z!T=`+k?r3nF``M5)y*B;J-LMCyH}KJG_-pwGDLh>zZJOodArKS-d83qV0@5o#-2ks
ztS-(1Isn}xL<K1f$-6ynRfAxx#ZpUq${QT?g~@H3A<e!qjjtym>#}X9X?yRRbU85{
z#R+Y`fcB8rz9o!d!BRMsuDrawr7aVtLnyCV03VcAbWC^Rx{$<I3*bVyWhrGX&y8y!
z->(zv=cg4k^+(a=?QQv$7BG{kjCa;iEKohHXslZiFrvV}9)egU%4^r0T7<usFhHYU
zE&Nr|x@M=pJyE8PV%C|bD^YNk>PfDOfo|6iA~s^%`4BO+d#se-c+AvF?~z+zuZce;
zdj`!ARY?^P6WByJE$F&T3O|3>FMh3l*vSjrRHy`45Sl4_>gs%#F#1IaZ!6QauZ_K4
zveVd8a*n^ts+!ND!JU%mly9kcMt-Mj3D()A;9v7T$s!)t6LlTd`2W1j;(c-fn9owV
zB1icT|3)W+p{3*i>x`VpWINe3Q6W=itq)f3JM34%+Rp)1Cf5G;jI&~Iv2;=mg6>jY
zsS^xL5N^ohiy7nQ=%>Gmz{&CyxE>;PaRBW#N1A?Vi1!q-nXzL1>Wd}F$2D&Zsb*?;
zwQ%HBIG;rkz<@A}eu@p1%JpgDkGvcKxzz~N^8N3K>p_18{pJu+i?|Q|2mrZy<31ty
zN`BjA*}hZ#=+0(=R4;et`bhHFUelfotzgE5G}7RH@yd3QPGVy)8|{<85I9l_IL}vv
z^@E8~b=TV>*}EUTvRb(!Q8j>r?B59S-mA=rd^`r~=MwnbuO3`zg^|Kk<!8B#=ouL$
zzZTzMmbsgFAKBZM3vjePA(Nfw)hKoGQtOMtLSqj?;fc-+MEyk$;-SZ<IsPLZPVaR&
zs8lxT_(j@L`ef4C6g!1^(VC@J#bRn9@>E9bnEr2JxMDYeae4_lJE(L>MrLMA-J$LC
zNS?G`b)wc8U+YQqX+Z$P6fL^Cy1Lzj$h~yH!ozIQu$}Nl&`dW?r+-LcWk^>qqN;~&
zYpl_*m*@Zt)A~1y-1fNH@}3s0GL-*LHr_LiI(1zJvh-HgR00%7FC9>c-Dr~D17?FS
zb^ykbz0)T7{396%b$|#itTBV={VZ1@nkTbnFB#~;7DTzrR_*Ln4khbhdW>kcWKtYY
zlc__f`GU+Oba8d60aJuu;U%m_-}}izdDuvjZ@xHQ)phqN+4uCd?Rf!Q%vwR9j<GHj
zx+b=Jc;zP%zw~t|XzQ}>IjtH#oV1s|Q`6(;)XktPc;z1YmJ`g(=qs~<TTnY;pvhvL
zz`3?EyMz_ZIMr5b_9Ucj?GyxV`NzT#iCx{^*h`&F@1K}!Up1eH1M|~{&O@WV+un3^
zZl%AutSuKEr&kAmB|SX3Mg;iN$g0s9LgF)E))$!p8IxD@f`81F6<V?@kIxnlKM4g2
z2Q^k(Te}(o!IdCDs4@0&^NM#8;p?ubjqSlM@J0lrydHt47Qp={9mhS0GK@%B2xM_-
zf5Yq%=}hjn^|&VPGgaHddOr?4P&v4S0jRQaY(YD1`_Vq|sZ<-D5>)sQe7Jc=%BxR@
zPv$Uz&-M*MnW+k78b1pC^bIeC$gJX^eR2J2JMehRkPHaJH<l?6*dWy94J=}-CPb}3
zVNOe?tTne9n~hUtuqu*?^J?|b2^N1%_o3_Tv4I@2)*dV}+y4FBCT6GzA)%&;Re9l^
zQ>82&!P2UYETw5_bn^tqm2uVT-8hT!V<2V#_8`6?#&t)O-QYy|>Zki=r`bje!8-&I
z<`0kSdwi1$U%;9o-tlCut>K`J3m~`X%*c3-zK0U$V;k&>L6AwG6UEU72~ULts>b1O
zG`XLr3nV+&&k{dN-#0Q=Cw}aEt<Q+eoI{?_H8!KCx`X0YSc;lI<QL4&HMjB`tG0pg
zah<v6v$6>QD@X<<7^_+gz%yR`CbzP-b~_ra+DeTk70HCI#k9bxJyFC`RJut}zM`?`
z(@o{p3w3JcWaYxsmn50UT%Sga2IjyFC<<1Mm&?Duow-P@M?a+SeW=*>3=<6t%e$bY
zdgsk~sR$SJS+xlhWqPM3p{A=Wyv;r)4vG$!?Xneux)b?5f&Da8T9*LP1%RZ;*#-H7
zijd++yP!%46SysXZ>_~A00Z?tz1ud_DfrHs%vHDQ1=TyuFFQ!}?2yqScN*!D?GPcI
z;Ggi4+^JUlxLjxd2$A)r`jL<P)HFB=Mv&RMP-y|x3<G~G!M_I*z*QJZqg~1!ujH>o
zBIvpuvghpn9nt&a<NKnp!ZbA#n=RsXzVY4;Ie^b9s;$}4dXpm~o~ugmmmu^47clF3
zCAUAq(b*Y6Dw`V<wuUoqaxVLts=Uc!dmpWLgV}~cL@N81ViH9;<%V&@=+>~(dCT7u
zvK})8g-yo5*5c{zOUKYjzwRKtzP5^S8{hMeAdE{g$wUS?-!vlL0TwVK*^jCaRU>!7
zVg3GsTvOsU>)U@*ced*89~mtKj|)}D=cp$KsMiFEU0VZwZ0!e_bnyQ&x2>F%TIW)r
zrxp(EwqcIxSN<O344Pg@WkS{+S!#mu7yQNBDO0&P>A5$JPB_UQ%<dERV4u5hvic8o
zcelCiP38UN8GE$Wy`doGe=~uVz!%`TJ`~sLgl%URfrZKOef6Kmgqr-yK3VJ^`!ury
z@e;O^&G=C<pP%YL+$)4k0)4H_QZ{ca-}X1%x6+fj&~5o!4Pf8)x30|#Fk-Ldr74KM
z&oWt7PP6M4Qj?J6IIS6{D>a1TTASPa>t`H`(-?wJ)gqv|ZOpi~DXH;qYH+0=-g@MD
zJJ~$r+O8D2sb?oYe<=2DSb1!c+bB{fnhqs^KfjTGy=kt^_q5?n<m;mh08||uE7|Ax
z>7%e)-Eu!UoWBaWZr@{NwFC}sWMl+D>aCR(a7lpodl38$;MZ}1{JLVT)?ux!5<Vu{
zE$rWfV0FrBgF)JN=cdOOR37cW{-9cF9ZWNJVrBi@SgK*xN?{HOwhoTE(XlbG)hPyl
z{!e1AtMNaHxzX!Fl;`J-Xyo#(X{4?%8JR2F(`=-;hwaz>=tvd<NQlu`UXfKZX(;}I
zU^DDg86IgyRnZzmx2_^D$+0|dh-*XK<!s0O-k3-J&D`MKWcXoe_@QTik{Q1uLI&_D
zlYblP@H`9+@I(L3Rm>qAxBnmhVjUlE3H%d&^xXfBEc)jX)7S8n)W0ly|5tkH|BDw1
zx7oj6q?#mVltv4U!s-KShfrC@vvvUo;uS;#>g_nqnQJgR?--K$56uTm-^r7I{{(j`
zp8JD;u27?&`IrMO!f21fvvtXuphfdP2!vX`0GDJ^*kNm7fl&l@Fo9rx6`*wgq4PvU
z9E3xIjk??@f=TV~{I^g%$Nx2n@Hbmj&v8%gTZLj)Xa3k6{~bMAeB<A91L2uHQLhGx
z@hse*zxpfla@_u(DS&@sHUF7Bh)>bP9{cAC@aFNfTm1t5>&B;Q<`CkU+Gi*K{S^Pp
zFW?D2d`#wlc~N+_6u%z+nt1qU1%(XYm&-pF{l8yw`A(rSY0ObabbNZHBhukn;vE{D
za;swFsil%D66breNaPd2e0I#X^nXXN9r$omkDZx*b5^$P^4oEseL!;!6!`}M2P<%_
zi;7l(vk8a8O}yaC@h4fAy$r2C&?K-+{EHQQ{iLKz<b2Gv>M)7$$G7Q(N_W;IKdV4%
zIX_(K0rogWMMXgRACJUEM#3S%8ZS2`jL8MDu=OOKBNFUBy(1_e75PP@{L$w3=QD-n
z%@@~Yr(CQ)87dtVE6;8)kB?PeA#hh+Ru$6pRdWM2YI=?FCgH8Z+m96Q`d`ks)CsK&
zO^EOM!@KEB@n<4k=6l)KSub4YbZo-MiC6*0=v=O|jPXkNjbcd^!&%<W-gtMPjL;B@
zqy@%E=x5P1%?Sxi@sYpGh;L-9(I3M{<HJ5KI;;vQkfEm_a4R3MT9}8k4|t}RNdyE1
zl-I<8;9*&u60p$Sb%WV_nTacp{VJ_2Z_g^q$Ms;mXemY{m|XQYXnsL+I(y0kW=f$|
zn90euIScC4p+(Pc5a^{efNJV-Dmx_nL+1kHW*lx1cEjWbi=mPTIw_D%L5T>!Mmwbj
zoI$!9EK#(xcbBcEADrE~$fdOth9?2TY1CtGLKU*pih$fXydzBPQu!a3-QERuYPLL@
zok{b(h&BhL;$1%$`QOYQ2gM;X3?Ht%5LjWXOGX}k;^B7_Oh>LU7H+}GB6ytg@@2?R
z@~K0F>oCmRFw}0_hjDmpYE=Cd7Bm>Zx%9|@(UYW`h1T|(z7sgizalmVK=<XC9N01J
zeGDCa<4cY0PQ(DHw4|xPzpX$u&j13{w8PI(rg0n3=cG-wx3{luv7JBd{S_siZ)Wp$
zzC{d~Y}j%6+W@+PF)1QVA_>~_+K_<7-#VZl#x=rI0ZyV>@PJ$p4bIVz57je)^_Ucy
zjCc6;^!X%p4k0Cm)~rVhPI~=|Es-0*qP~%tdV6N;ZCvTD?Wep30Tpzl<gO2@{9ONm
zi(}Zpf8)v8ukdq-<r`Mi7cw34#_|MIVwB)(hv3VNg#jK~T76)&L0nS`s$_lrQIOi;
zhejrg66{(gLcmK1s>X5dP2eZ)BgpX@535Ca&i*)QXZkr@%Zz(l^s4nRr&NC>!J&U=
zArql(LN4JbC)8q_Hvx6tu15f&d;k-grV(ENi_z2j9V~JPy=t)j4)6oQ&wqafg9a$f
z`XMqC0$z*Gp~L&|jq61DQpH{nTYnPHzsQzyr}{n6HSM=GpJ8(EDe?2l<~yHr=A^hD
znC`CQaaa5D&-cRd-lXND;#NI$)nkE_^CsKOsnrq#2Lhow@z{zlz^#G)(2M=nEj>{4
z6b$3VL?5A$@3oYeN)U8Xf^l#6BzJv31cw6u;0WU6E1!8Y@=VhNa)&pIQeM{c@D9AY
zUayRNcpf=zX5e-R`I_75%8k%6oQ++SN$-FNL=cC8^GJ^j%?)z3`MNkEmq@n$2N>GL
zan{6pqUH9VfkD|B?W1fdbG&m8nM|d-;+gYku>9oVoo73ui&OSWw2x6n57iew#FF0`
z5SC%Ne<4|uCG~1kfR-<czzzeECIG~<3dn<u08rrG_#>b-m>y4&mHMl8*)R|1rCMJZ
zRV3!NO>n??@n7HG4}1M2{fy{C;aQKT4A--Cz)N)pZ82DM>B>dIj<PaAiCaTo3509|
zRY_(J_J)tRkLWK@a)ZhX`{ly@Bcr!Rbi7EW*lX50Y|kv&L+%orHXRHG{k|eUulaMJ
z+vSPSs|%SV7Z+`H+d%x}xH|qe)g+l}8y`2qa|GzPYY%Qkr(h?SZKJk(I=Bo*;;~V5
z;$A-B5tu45Ka#91rIeWj^U|^tU08!K_<U?9@}B(=O;;*zRHwl<0-TK1&4p|+YSZuJ
zwa3u=9$~P3*)vOG{W^=|>8iiEle?0Qi_PS1rb{;N4|&cc7u-`i=6tsS+*$rZAVU1O
zq6uE%!%?x0Z1a)1gge>l;)(9(6Zs-7wmMW&FS#7i16+dg@m~2eqa=PF^bcCom=F|!
zrpm;~$jbhXEgl*7H=aVyBT3p4UhoxCT5Ek52t|_*c$C=7r_?m15+{V~)Hw)68D4&S
z$|rND1s_G3y@Bb={bhl$J4yE@rE=|q>hefM_UyH#!Q7*leML{n4eoz;hIief8zYN^
z9^STZeQ9_&zm@Qf>gw7WR@xa4l~B~3K=*FUe+;Av%U;!{CMrlk3KT${m95Pr-<hhD
zsqn|rR{&|!``sL{xw#5O?I2v-anjd?u?1ZXYLj;o?~`lt1AoTlq4-kI8v&)~G33+g
z>C`%K%0g(^E%f}5Y4LoaXZez{R+<TDOE3$J`36g_D_pk~IZibrps(Yo+yOI^Yd43K
zbmiRMWR31Kmo@t-F7(yAy47{6i$(%p5o49#@vcF5Gc4x?>}JPbm=Hk!Q;zS6DD%@m
zWop&vG(vu|{$#0*U?DTbBUE0zyn#gaE7ifv4weic(fqh}0hF_!oI#Cj!G7I-F?^?T
z^@A#7F=^~*mVlu{T}VpKjHY;4Ab2OomOy@aujO9RdT)#RM4QM_wflEG1yh6lim)H@
zopS1R_xQkSqh%5it($Y7bi$CV)0_>rF)ykgF*v9-KiQTjk{IlKCqbB<;*0*K;rvNg
zT^tJ~&)^I9`C9_=U5RqvqY8wYVX$L(B_886c9%|9(j$9p0l7N8)_V575}>CB-nJ*^
z;+NRRSNWX)%kNRAcMcLCaN>Cxt?GPgR2eDv>0GEPFq@u3`=I(=mTx14HMZUIaym$7
zJ>3kP2x7H9_YLW!Gpi8%_cia=$6)lg*W!S>?(3xq{q->8Er=u<7y|Ky`I9~)5J{3!
zEJvw|ii+lZK-3%=a$d*f<`!5F1Ngq4l$0bda*zkdGvs1EaPtSS5&VFcF*I?<P~z!I
z<wu8mlOjo=LE7&)og+PSqr2{6%i56$&o|?E*`RUz@QTL#`VVy2WV+609X~IW{lsn5
zC4N(=baz`52)>{EBj`h|7TEQfuwB~L+pUOtkRhWm3TxJ;MTb$Db>+z0M>d=~msV!3
zoYtP~a(@6=mu%;6^#RT}ArI_Uj>!J9*pFf8PAsV(A{PSKD*S?jO}H&!ao}g2xPOHv
zdOQ7sUPO=EU9NN$(wk}15i370kj{U&p*em<qGL-*F%wlCFqrihoo18tvTSA1(tMA*
z6>4V|owBQ?YeeKMzvB}WGg3Zn#-90EG~)BKQIYT!e&yU;S+mNB=7`YAB2n}QJgxI*
z-1aXYCf9~GqhF;8{CNyxng%(u*^g+yox-{9tXB1KSwQ`5HWlf^*gl`Fz90-iR&Fk9
z)*euz0Jj~ScO0W+h%DG-9tl||elcyc)ITWadqF(y<VH)^T%>v*j%$>`6AQu&dtmok
ztoKY3K{`%Bb9Xi-rJ17Gub-P?^+l$;!ZD0~qs-_D@$FzFw>t59==vN2^wRv8!XmWC
z!`R2$#wcyD6U1D68nlSh>X(y>TjgP9Cc4rLRTvDCry6>CO~{W;r|c|7$pVPd8g6DP
zz87>$wS%uz=k+7F1ILvsdlF^X=+vF)8(|zWq}H*mTACS%5U_}KY{|1@8{eH1)h7#l
z-1$Y7dA<M=(_$T*e^N{4+6GWP&1n?jbl>|(`@-4N`buf-wbL&<7YDwp#2=^lhrGL*
zqTCn#?V~#L4MCZggQdxUnCfiH=w^74sf%yIcvgbbr(qBh`<*++FjyB(CO_F}X?%d<
z+S*!ilF8kiMo3h{T8nrb3cBc?!HkL`9|k+Fzx$cUuUKDY_7b@P^;Ti-g-p~974~@Q
z!&E(}Ffsr{H`sFh`s#RDC&vzA*cJV9CdAs}L4F7-eR)I2_=~Dcj!!f4fsw*Rv08%7
zkI!>j5x*{g!YAeVofnG{;TheOJ$|LzKini(TUENHkq71Y+}IzNA2ETUJ8db~w}cXp
zE-0lF#@oJ9CDd(`dOU7?*bTdku{};ykBqqD<YnKy?!|uOoA?jf=_b<Yqz<dKwGDQ4
z7YBr!o^RytBsf@vdykz!w{JbR?#)S-y)zb{z}kp)=HdK>%_mj7hta>T62pg8k;Xdu
zQvOto4X>|lCk%jbd7pW*?ucj4@8-y%H7Y{f<;TP7o(~VhbdbojI7lLxnw;$C#S?11
z{1k@JqI`vEr7LGPyCLuZ<9<7b(B-2*bV`t}nro7tT66TR8~w^%KhUp*H5^vy!}f%*
zUrg|%=hLSf{pmmJ{3MldZX9kq!RqBfOjUXj5Ft2k1(&OX?EjJY1iQ~3ym=F0B$GeC
zbT{tcGAouy95$ldd(?#8$}9$-OC{>9HPl-BrAs7@IDfpiinEiV>h~?7TaRd^+KM_P
z&ZtMIWm(9_9N<%h|FZ0L@uM0(z<_SBSB9%f&y<SGrB;$e5fdBa&|WEZE{L<+F@Bff
zat5@|3t5(o)|CI|&>qa(V7_f|_~KhN0(a-!twQ)jkH=cP;Wq-m9xMPR@)(@&cx<-J
z4bI;}FyKzR4M5h7U9#}O#^~t7s%y+AKM4maJ|l`y4f!0|Tt{(#K%28o^$_pU`(knD
zNiNUS_THYs6z!pn2@>>9K%c~VWM8QUla=Ug>PboGmtUC!_YWy`cauLdXlxM&-K8bj
z;pQ?l2G)v|+7i5WE>;o;*&yfnFgyj|U8-P%gix%GS9JLhhZd6%$6CB9r+`OYVpZzt
zNDW@e*s_x@99}tEU-#bJHUxcR-Ih~x)5K7bzpt-^j!r_n8ZSUZ_WZl?i_M62v-4+&
za}Kv*3>tfupu69b0w5pK44b<=>}#*L?ME@DLx|?+t(%=6n@N2FnId-mJ;$6Oq2O^}
z+x{J3+$X^O>Dq-8K8Pl;YbGk3VgY#*)7`H+bCS-dln9HpLjcXc^TKYQA3W#g83X=1
z=qJ+Sxs@P){Ahj$?0KFnPe@Q)eDd=04#pAC6XP^SFQd{!!dxgm{5oME#(W+aG&;z^
z0)eWtHp?zeqHslq5)ZiR<Q<qwx6vkl+;YrnS+3xQ6P}F;m1*%UYxmHEX?FLJ?5sQA
zK0ehdC=ERgRm6hPkD=-5nl3?2u+b4dwfO0O-<dcnQ9|i+cvj5x^fTYPUaYp@S~J_}
zDTBoWX|};`Gq7>QFPDEsWbycit`p`)=ih%9V`S6X#gX1K)1iddBi`P9>l$0pFmOX~
z0(Qa0{>OsOI2ci~Z;IIn<jG!BX#*)<6*;%FqvQNgX%AojQ?0h%a1B^HfXpxO<%~S6
z!ASL<X%x|@*D1gCE)RAPe(uY5n{){+a%rn7ZT2>wCoUWB$dB<J3gV~Ge~i&CrJAFH
zu%Qd3qC0c^tKq8P+%P>11UyL3dI9ET7DA~$hpa^nUKqafPQ~JROoCJ>X@Fmxi{ZnH
zZs&Ov53;L(Pnpa(4r9U(lPOe8p>3B_CMNo0t&{y@tt<X)Lc7H-oKtLz_Kvy*dJvYa
zchSsHF>Ojk$=_^SnNo!5Lmy+pN$c#zk|=0l5};K*!mm9C;RMEsR>d1ID@eA~oYJ1>
zYM*6{z}}m&;w)XY04*NV^EWqK$LlYhR{?b{XfKf_zodW<n(KQ$BM6i2Lch;F2l)q1
z(ed?7%>50QpgL>dT7hX(`aBq-3S&xMyN!WV#8Y+9Pj!<n(tJu^2TSNCK(lbOp{#Qa
zKS?mK4WW{OpS;L0EQ`Nz%;D*}5T}NxNNilLe*cJK5ys0hj3)7Z<k$)xypZJpD#7UR
zsrbhq&)^xrWMF=12f_2yV=i&W(a&S>;)Lf&thLJx;avaKp2QV9c{8%gb=T4Aw*YQG
zP0F=j(3tw`nbaLd&$c42QBd7fn|b;Tar?}&WLr@B;*^u)j~QK$NymNAI<Bj5^~gSY
z^jb{q?1c+lA_w3!eQxP)FJ<*9JL9^Lb6)4`GnguZF!{aX=&%OKayM9~`uEmi3eQ*c
zIebXy?Ciu|2oG?Gp*V&nDGml$qw%@k6>dWVU%i{Rs3)Jl)S4J%qC$r{@G_eJKeWAN
zT$JtF_I(+kf~X*kf*>G_gbIR4S%8G}jD&PcNl1sFgmiZ(jWg0A(jXxq-6@SU2t&iO
zkFIsade(hEU*6X*^1~6%IcMfPk7M7q?Y})BZEw=f(1HkCo{vM<;QA=X(`|m;#tT;8
z3QUK3>AR~DOQ-PhKhvYzuTgT9u5srOG$?yCI{GYyTs<hFxcYSF3i|u)7H8^YGwQ$%
zr<s;$=wE2~N~5*<plx3*--O_viADMA6s=Fz=StlX*r)gN2$(`ARNvV3h<%OQTXUTF
za>nEeH$VSStJM!ZEVQ%EnZf3k5~4Gvhiae`S)^fMd2Yf8_rgmWo<TX{K^W~a`^(qY
zGoO*xT@}>$^x_zexF{B^?4M=9iRd?{+?jt7yMJ5%!1U-xUWXmuBS^8j>yr7p*ZT!r
z$z1~a6{JsZMSg>(v84yf?uLT}tf;UmY=9Sd$H3H}HCCwml>~T5(>AplR33T=qDNyI
z4@u|h8X2x+K~EyT<fW<C_7W0IP90XFqOi@qox=~|wI729z;nIokk*$FzHz=VKOxq}
z_PBZ+UxA=PsY7mx0Zixg8lP?y3o~cz?IAQ0#9$*@<%xpgcu7Iu3Ys6_D^klcn4QJa
z3OYo&cEIH60HVzX`joJ;m%95dHbo|f*&_b{FT4G6ReI(PZoCd72dJgsT_RO@-g)&`
z%8t2t(7ALcX%_CzF0i-lH;ub(9MQZ7Rwci=XnEG$LiX&);{8B`=|`{SM+ofk$bk#`
z;%$GW_-x1<f2a6D!$GJEOb~M@Kdd}D$S$!If)ds4Q@Iu}Yhgfz;2KB3u5ZJOvd1n`
z6?r{|1Xx_0mQoKsD><ucw9)3fK&GbpvxpzHXO|0b>Y7Yd(7}yMf-##0GBHIPH4>kL
zPp)mdfoW$aEO&!+OD5IbGSbW@CczggL2gSZ%i+@rI$3#<Xul^eap!+NC;vO^ki#J$
z19BXdz?ML)mj~8+t~p5IC^(498R_kP$T}PxIfJ_PnI``*LWcS^GM4uV&t4WMffbCZ
zgoQsnxe6Ovh)QhC%EamiD0?#gJM#Bzs?X2GN=lXaCin5hf(kC@8S!ugQok+XiN>pa
zARYlXR(mLoi{Cdv41?R^FZ#yidwF*i+~+*4T;CJW(le$RkIbnprEeOa(_%lfU>FlR
zGt%xE+j(D*@>%%;H>w)3>{Gfe?tlk_nmMC=qDYq`cI5K>Nw_Warn3az%o`+D@*ont
z&XBG;Mpb5NB|gTnPA3LCP=vlr#JLq6hghR0qmQwVMirJ9c8JiLVzHtV(xWgXsvF_=
z9=u&#sQTGe@|S9tU82P_;S15Q=Qv1a54sO>UBGpt;Wm6lluoL=wLBuujPQXN0u+Am
ze%V)r3IVQbCaBYx2}M-y-o1A<j}Le$@_jC!a`3sC$3TPgO(+X0IK-Q@yu0O<)0nMH
z<<n^4i(K3&NRBirwyqEiJfyn7^*eIc2$4HQ7Sc-{`CQ7=a36eTcR+-hSnn5;ae7k2
z%tJf&OovQ7@ZLdPzEd|KoB4B^jGuL&`?7O(U3#(fG#!bJB@GmUk%Ah9W*Kq3zAzb^
z{Mk4RCPFEVuBRc$tY>_<%$>E*3?cd~ozSJ&P^Xnq!C3?NaOnFn`tfF?d9y@Wo3XyM
zKUp{OI!+Xx3dyisahpa6F38{#sha0m7s7rkXY<8yqXf0P266Q0XEOy%>2mapnFLFx
z73yQK=e4!g$s8S+*Dc|S2^-1&i)D3c2#DvU@!6F*(`rS;#_~g;%q0D63*40MIBd)T
zyTFoNm_HPGA;svMRC-qlV{Hza$mc;|`a0d&R;1MfGqFXj#<G-=O0mn8=k!d%E?lUB
zP>e!L0=!U<Vhhw<54q>vbG=wvd%6Cler5Kti#c=r-x0d^`*@OYiD0*?H%YPbLb8;P
zGU%Ui&vxKhcy+;EhKPzHs7P)Kp)VbFZUHyJE4d<{Z<-95`^osBFo~~|`}=oAl4(=_
zRPcUB`Wl}MWVrcwC(1rD-dehE`wH@Wnepu<wFnXB$YyxU(9qHfne^uXU9i}wOCNSP
z5ERELgj7@{Z{L(>&rG`9<KQ-1_ODB@$gdZwREE;J|Ek>J+U#&bjrOts;Q<@CBEsW?
zp#J=+7v&J(M3(<ihrvbha1lJ5)b~FKHNQ&M{y+Ya&9~66AQFQ6lM8cputnqk{Q1RS
zU;USF{kMP2(k9Db!YlNAXG8ucWVGZN;W=EM442vaQ^kZp8#$<FkSzOmT^#Nqzo?<V
z<7QQe-+au!@?|)2%)t{*>2DqM??2%;C-iT>1Rq2>dh*{7`Ag^g+XMdk8TSCli8}_|
z>HV$0{f|d6M+MiBU(e*fJ`XsGr>`KmOVXAAGvD4++IEw-U}sH+kBjU0*|qKa^de_s
zkzo>gqN4SO2YbM*q$49Mv|F11J5fPF0hpkFSl`8wMB~gfd>DReQ0M1wP_)gTTk2x`
zg6?DQTijQ(o$wwXB%AngQ3CBY+-fW+$rys$Ri~wyK~@@yPhhLcaz7plQuz5*IU)(y
z2cBH`l5fz4%Y^_(C#m2P4NWPGF9CcrR%$DgD({vs5`?^v#6ltRc94g2`$Iy{!_NHB
zyDpcNgE(*eTcf&*OVWw+mchkX?Xxum19pI%Vt<)%>ZgF3%7^EhBm?JbQ^&Bkgxk3S
z`by$Ejcx{B5^S%H*C+`H5#OAZ&g{z$3Ote~-A0NB@3oC1HH4f6cAe<OS2>E|bIM^b
z8AI0uD@8)IVc{Sg+>-=>5@`G|A1DBr($a8=LUteF<iG*oEE*ddAp$;xr9SWklhoA|
z;v49UNmS=;%s^2G(O>Ep2UB0U(4&8+(CWToLc+o}APrJE?M?q-R*o%*GjX9SFtd*%
zT!xbV@N=PmL{BR1cV1xBV^8<jzlR}~Sjo9yvYNnKpICp8r3k-t*C<gdw|L1%)3+bw
zPuplANaJ7M-Q5kV8tjPe*h?1v9<bzqfW#=izBRy3?hDSXmGN<{%09SvJ(iK7U_?`9
z9U<F4A5*JsiqaSc$EO%yh<Sw#cfZ^ze|fb1Hv~g)28@@?`OHA>nv`U_)16lZ(QC44
z_r~=*#?-2s<8SK;QNe*n0h12PP$I56of8yv0h|>L23wG(zqBMkt|}$vR|Pim1u!CG
z8nR<^34SNl93BB-{T%k9t?AS6Bkc}jx}cGbzT)>nE{~SOJ8mRT<Yb^y>c-abD#UUV
zVosJlVc~by%|Z}j@D)g*1t@u_*8JX=PqRq}RtNK*cxX}U&X7C7wB&lQpuyw$%;o67
zmOdAJ*%Kst)Ziv%bhCRMz|y3lKhK%qTuc9WEt={B(@&9#Cu>8u&B^JMr$GixbjDI0
zWZ_EQZroS}da_?&An?SS#$1n_;f)HU3TT=t1kXK~BVjg70j+d9D)7!c@OnOQJR#2~
zL(K~2V%~z>s=W*6(kMCNuh9-)vs|8Ui|=G!k^QK`@Pk78AhX={Jm-GTvh<4#)^F#-
zy4Xgwi;bl-!!Gr4#Uo*WG7M55ILjd_3M%m=yt7!un3$Pcqq)-IT{tHveMAN1&xzm!
zR-gINa{dutV3kfy>E<du+ZRvnh>DJWTb5jT1+7vz4}mL>Y6N)JB&b<l@S-{xzvwJO
zYUHe|UYOm&@>uZ3yx+H8E`5VFU*zMEM-%D4v?~*t_7I6w3k!>iiVC<1?dy?Q%<`<i
zy$aF5_hM1sl~aSR5U825<u%nj{_*`r&pcDcb+=C2%tV+sM!tEOA|rfNB>QC7`Ba+V
zJ<8agWWShdXF-a50{07H*|5|w*^7rKVgOd=2Xwx7?AIAJHsXMHad3#M&8?1FQcnjv
zz6}aw36FOMukUvXcki=%va~k~Dtm-^%f-}urb+YoTGU43`{jNGxw%Wa-!#Q90jo3`
zri}LX_5t3FRgp^nx55^i=CpQJ$)gYRfUjcWgx1?nj;HpYw6A+W+<!AhsYRo<r@2fM
z7NsBjsBZ(UOcm?!0hO%ic@V-`xgVl!efgy0l5xV0FB=a}QFlHdh7Tc<_bWCRvEqPg
z_~>+1WsLbt@+A|!H+wk*4t84fIMpZNV#h@y8Wnol;aNg-qi!Bf<cz(y{6=Iq%Lnh5
zpazo65tjWmNNNqEZ@nK567*_>+#6NSa+`1YPA!VvD1)9J$kN?$AfLB##%(tO`2Zxo
zHM9l8(Ec{FfxOgsbaau)K+R|~XXM;~9p{jfau$RGHpF=K$Vh8I6x%^GdUYBtIWo%5
z=qRMc7TpNp#N0Ayy=3CI0ThGf9kvrF3mMdX=EQa3MbTjG;423`xIh~%6?D1ssD8F)
z#p=L#zTR0dJtaj`n4g2AwbhCU{~-uM-`Ze3I+dGn=E2>mSYGQ9oU9h?1fIu9f8*^t
zs~_sHiCH6|hHL8?udk&VU@+GmF$|$!pLb_yxzGSiZKbH~mL^%T9@)3nkNy4(34&xP
zmFq^rN_O)jK?Cv;mvqVAK5}Od{Q(-<)@GUWp3NdHKxcany8t=qSKd3Doj@!d%l^{n
z6>aaYI9<nLzk6kExlO77?i}4*&L{Xx&68m{Sr3VFO)^^@G8js1e`-QIjqgbMh$uNr
z0&78_(<+(>{bDFrM{IckiVUery<iTSd$jR82`95DCD3$J!`xUMStY0}Qpim$i3%Ic
z<)hh-)167{&2RXvi+MrS!$i&N6L1r*8^Gk*rt3w!<6|q1ro$^YUIy$cG>icC;}Hcl
zb(6DTqJTX~sh8`~LGeMl?T=2p>t$_{f}trJr4{DP`frMOT{%jxp;<s0dz34q<0VGT
zPMG9VS~PhF4GzN-rCu!@z`ukG4<UGJJz~e0Amr)3L2;e+>bdn?1HIxuK&=$IVLd0i
zT%4`MmIX+R0SPnte+A(9%|``<=9TlE^YK&?`SmP(ArPqTlG`|MN*x}Gl&WpiU(f=_
zJg8mVm2QxSlFuZ?JtN2_0o`ax=EDHVH38R;QeRbby)_&-Vy9<jw1xGdv$o7|3`k2q
z6G#3`zhNyCt1s)vry0P4;AfR$OM>5IoMWEE#wC-c=vqwY)eCeQg7NiwD7d_^8m)(S
z7($=pAYAc4zmqGw@(5B*JFkwNkoAMN9CV^aUfC|@WCywsd2ixuw{av8FTlS~@)AUo
zMaD#K(Lxh99=%9l3dukFZiadA<K@ijmb4B}>CrHGdI8t)p9YgZz365>oj^k?<?+3n
z2CZ@9B7146salnnrl;iF<XK2Tso@g74(3u<hcp(rew#<2tv8Z5zgczdAeza;j5`>_
zwbi&3R!kP~Ir0g3ICg8#3zHsPCcaaSPpl6(MG+^bvfv-bbT!E$SAK`s*y9SH81dEN
z;!86-tBIn*c^Y<y3{09H9C8nTO1J6VfIR6^`fCUE7G@6AP%-qq&Ilhu3rp#Xad-bV
zbk96AY)2>JD4bJMQ;=^OWb{of{QTPSM<RS!JD5!y#6()}$odS}Z50(dw@2BFG~%Nl
zGm<RgLs3xK>Pw;_m`o0qP*Cn_Vf%u9g$wt&v~(cDR_yaad{6%Q-(<FvI{%j0$h9U5
zPg5v)RJ4nOIeum18NO<n-A$|{R47y>W66B$`Xlmw=j2kasXt<6@j99nKC0e_Skm>C
z4L|wZ#>#1$ELz^-=Oqt#KW0C8{Yw>d|D}p0^GsZ8&65B~h_m%g28NoBkCJ0x`T=*;
zaWqg1e9}yA3e|ky#}?+18J+QhkmAFUi!Ql7Cf)aZkINpiUs9WG6hnP&ZOu|I|Gd+-
z4bBmh6^1X>SZ9C_!CPgHhNcH@nxKBh^V>p?5SSucVYuV(jDN5O6R7<zE|ISdu3MsL
z)7XplCB64s1s8hdNsD}=8;djj9Tp+cIb7mThKj!FmXBGVO?1`f&4Fc4zXptClB=q!
zfNoy@<qH(qQj45)zSx#@vuA(bNvO`9CjZYg8)XwdnDT$8*|tw^nHhevR{sulb-_;a
zTf0_2JQ!6Ka-OyR8E4zcFz}K}#=-ju_NYLpoiW<e*VjKsCsbx<>^Ougl7o5jd;Tgh
zLF4m`%2KJ1`xPhNg_bU#O8JK{cGW<68s;ZFDFG~c5bxW#ToJSlWknAIDZcPkA%^wH
zRy<*&9$fY|&XI{PYxI-oa9v=QRu~Z_#;5*1>b@FDcX~TcbDKMijngLU`B1k_y08#k
zqfZuAihrKg@4)le3H;U1&QLl6A^cpA7Hfd^Nl8jxRgSf+Uv@FgUKerGD7U8)oT8zL
zKMj}i-<p^WefMQylRY-Lrz){Nu5Yf+twsQ&js1r-RI~{5^7hr2%(f-MHE=68AIKO0
z59UL4EC9CwT;(>-O6f>;4ezCi2GL)Mw!bvDl2*jYimvDsTUq*K8X~J8+yKe{;1xKq
znsjd^<VnB@gp`$JDybFB=0<^z+fS*Y)uRF9ubIiu_54K<Dy8h`DA7dhQrh{~=su3Z
z(OCxZZ`f}e)57`-)#)!HkaH`qd@!)u!-fsWNJ)K>-oRqPA!BaF$usiaxqr_+DZEv{
z*#;JN8Iyfvbf@cKc=6Z^OZsvnPbkw)))suru-rUs5qEN)SgAat+6L*5{$iZCc8)A#
zR(cN+-h=}bucU+O1Gfxl`UHT{MN-mkaCL5tItCu*!a*lxF=_1gq#lm9M>v5R)Q?Hv
zb<X+~GLtKAej(Tu^-`_WHr?|6n-rNFsEBH()g}<gq6R!;iBNn181ES}Jn9{Xp+7G0
z1`So9)&HApNz!_!_K0_KDHi<7cok1xJfpf)Z&D}Fx9e+8zckX`YQd8*Zx1GA&@FBp
z&2Npl3UhIVmYeHlJO9_zTtbB0R;cop@_OW7luN7eUW~kT!)MzU)*Wx)vqjJK@Y0mA
z#ez%p*K7;2FY~ZKvr!)h>`4EQ$F)_6<C<h)3}4>kYUJ!oj_ZUo{FdG{Urtqf95Vcc
zvOJ@NOBg`ZmiuTFC$_mzp*5P+_h9!(QtJPUmP!BcFIpyh-&V0a5H5xGYF?ynBxFP0
zgth9YhPLFsgnifLDV@Qv=%gGWmCT1q9J;uo#I6;s7cX*7h$mdhd2X4IknrocbYtRG
zp;J!$EzGNH`kHlf>SP&?BEqB)f4JUO<;_u)ihcb{bSr27$qAmkCa5j!KU9wx+yV|+
zzi~Xgkb@;e1Vyd{c_3tQ?4f0Z_3vN)xu}wkk6GU>5i#6aA%>j~BBL=qkn@9-G^EbD
z9Gf^=^9VWx4ruRwFffU>f{Yd`^UNZ^E8>trDQn2A4l79Odx+#)=e#GIDLQHSz7k9#
z<mDeUg^PIKScWH@zjWsn8D%E4%+d6S#PwEy+q_>;z`k>ISC~=U!O<}`5v;}RX)^=2
zmN}UsV=6=L9X<KF6oW^ES)ZYufBCsU@h?c_XSKh{X9}fCFhPr0h8dt$1Tu(zQ+w-0
zA4kJ?90;u+eq;VgU{j)QP9JM#YFSULKNE$NnhtkR7U)f_b@(}=QM)NshFEE$yg$DK
z19&BRyz}P+A((V`Rw%!tBBRR-`HtxH=@)YZ-d#D&k0m4~(klO<$I1RnkGtDMfvT26
zvN2dL2jlIap6+L&*mEgG&4%JKeTdj{B|rhQIS*79pd~tMy;J#!kAg6y=ALJBdLYFu
z^t){}DxHp90E-TienlQ4W<7PtD7_obze)d_=vLZhQm}%9Wzu|l-h=SU(&VPj?TiN=
z$rJ*8iN(?0G}g;zMVewhy0oMm^s`~1x49okp?B2TFZUOq@)Olw6=GU1JaY`9%mt-I
zS|8iG&6+vZBIVH{pxKdI`Cs}D>At*#z)O7YZW4g}ry4ufu(O2IXBP2)C*(Bs93402
zj=|99!I<>pJ+WI2$~Z~4Ul1(GZ25-1e!)+Zst|E@r*un9-#9fXx#c${e6d_AXuauL
zxQG2t_2XX1#frXY2O)<ImZFa5(7VNBHaw?0ozqeAA8<_Ex_CNUcyi%%)DX2F6RP8f
zb-%z*pz3>NU7)6Z?y`G|&Z%m_fGVdcR&`C_JZH`RLL1y0VAdRmTpT**7kk(wx{<8M
zSO~w*$N}o^9zi7fhOBa_&BveNwW52=1No;)-EgeT@9Y0!WeVfsUj$7?GhF6Rtoq5@
z^j3G_^yU3GS|#N09J*8l=0j~y$0Db?T&kX@-9(y??HrPvw|dbHK{meP4n7BI8JPwi
zS<xb=9m_}-FMn+6(`+AKD-$iKX!|!X^EB@PlO<j9hP>wq*JeJO)7WFAr>hX;uMs~D
zRCVza>oZiF2vOCeK1l)MHa11^vz^V0oH!o6$R9Y^rQI2a?A^JU(@{vsq4Z~|cbYq{
z^=WifNcYElsmunO9^|Hf!a7Z}FYxVKi*UPs!(TPy=l&j)bF1UTDBCo`$46(VoRvc4
z<{sH%F2Qe!tkNF_B}Eqp!FTS%PEt@lNuNWYp;VXjO1DEd{LEFzEI0S9rhxaNtU}@|
ze}~nG#&7DSvR0yI`iNxFrqg&0-flf_<>4-xaJp804@U@GcnlG&fx|8wkh|0x^xa<@
zEy$upkOOrvMX`ng$hzgu_~gV;8IhVGGIrTyuKm4}Ys48pq+hXFm=?d<?oc9R#apmV
z%Fi`GqSG4^imEsw_R&rwKgOrg2Adj-<-wV<A8DAGS*WsEEXU7lY<%38#38vPnJqY!
zBtpnf0Z%#~pC4DS%}ebM)r~MWDtpl-eKM6<&>cpQ$>w>Ok>c33p*jQ(&I8R3l{2>M
z(3bLmuezQUF8E4s8{ujX{W)OER47q{P$(=6ndwKR?U*$xWLu){B_^i9c)@p=<8u2G
z5FSE54ICbq&8S;5LY#f_)#`o1bZ7&#%%sW5X_;2!H$lV-XGB}f11b-~x-hr9!m?yP
zR7R+JUj#zr9FOGX<$>vtd(m}oi8o=MN-34``rvitVfK3^ED~U2$Hvg6-v5tC*;|I6
zt-9e4AQ=bvK^&3tZf&LAnl_xvex5YHw&yyx$sgxF-f5OXpE?-v0*xm`$mCPPk98%M
zaHLNn3Q|Mds=RY;i9Fv@^1@jDE|+l_fs%tSM%fQJuQNJ%be_psZw?p5yCQs*#Cb!m
z&<eL-9^eF5XKZqfJSjg{ajebms#&9awN~GJ^x^j`x#W89TLUnfw1-c9iysE2R?x}<
zAMlMfjo;zCjs3;hS+N_%#m939(v|Sgnykwc*M5*O5^U7)i09g6IUljtwR;ju0}c^b
zh$&;8psz!@PZBKTV5Lb&L}UuAR-QO{U6#9#JrPcK>B+r^8(J`)IvFkjT>k`M-tW#T
z`8ZO0pbEhwE)JevWjy{enxr_-yA&Zg0cm`-5rLMALFeYHH4BeHQ7PEif%sq`Lw>iT
zzl*DgQ_J2jbDluja10#4sxRMzEDB5GrQVM;60Lua+y%h~;3qs}|MInyC=0N`-Dy|T
ze(sShsZ?jk4awS~vvPEllo_(C$zvdzwY$+Sq>a?>aOtZvp!=P#$ZCB!j5W687txw5
z&j_u<>H{7MQ<RNDFTouvT2SQR+bz;yFv>Eb!9@e(L%l)~NJ@$nL5^(q(q0H|(FkD=
z2lCTIN_BOZ&d={HN7>{ze3+u)T{f-7MY<XD<_X(N-nfV%B3Hfdv1C~)6H^T!Wlt6>
zrgw>migu>I3wsWj;pV2MiJOkV3*rX8mR|SSMP{Y+FrjMLp|O+)ERkDKL%Z*{Z<oRA
zhbp1Y2m2AMpJN6rL^R3jH>IYgpNHDA=u72@wbTWg{1Vh?O{i6uBuO9z+e7oTS#+cL
zAgRIWZ77f}mfRRo#-H!97lSA|8A~Q)52Gks6l)$3>~NDt7;&dMfSc&gkeyTlY$Jzb
z^g_5tn|sq10ZtjP8wDb&q!%{K|8i24pa|^4ZEPRx5KdTxB7LMdB7p(GD<I(1pQ8nv
zGPAC<Z?I8gU|=92A%Ti^Mv^n=u~sv-cYpLQG`YUJcD2E(w4~X)-TGR9A;|>TYOc?E
z0?0NPWHy@Hvczi_<v#(o8H{}}Qb@DGx3$lD_;<<g#z~XYnz~naS@mTgvXb;<FiY4q
zrm<u;F7Ib+^Ai5+REbRwX<qCGMz0kkEj3-2AA99=w>WHNhH|t{G`0O_C{MPt!e)uO
zY4mu9Ri;)}pX!t@xw<WZWuhLe?~VOqZ};0rLPgp?C7^$F%i&a_zjw?1j$8o~(fwax
zxZk-nob2|yT<(8WcyPj7?jN{^{KrqgmAn0wxBHcz!v(+oH>K+T-@|;CU?up*7<Ld4
zX49_op?$y$fac0ZTP^+tPX`+`7EiD=JfV8{OH=K?HNC&9b$^O^fFJt!0w4~kIy2aK
z8+w?TO>^RamI*hs9^_U+D&~|RVGH2zlos`&>-L}EOB&4ag;!M?kcnF&TwunjXcsDf
z(=A=;LD;LwKl=;5398!QAAfxZb9{#ihnB=?eF=K+41VKC+wr#vaVPAb-@!2^QEt$&
zhlk^39zHPh!g?4`;WCLh=x4#PzrLe{w+`y5;R&1|+ji^UP*cMb`yMzz=0CoJ!*JpP
zu=n->n`w6HuOq`X`u+1*mk4)2@GCfmlO6y2?}a0IstK!?|CMqnWg2u6e6*0^Ikb-9
zF)F=lmddOZ%CUJK*m-MH#Hk(k*zUnon_kRs(Yzr5GCUA=V7sINsSNO<PEJlLC@27(
z?4ev(&ZbmmPd#P<7|w^{&s2G%I$y#Ewc1%Cd)9eR{{I=zBl?o93dr|9Dxg%UPy_op
z^E=vs=iGqz;pgO2ayAz{c^YT!Ar1$muA*`XttB1~v%cZpUw;aK_KhKRsmq|!k&~bw
ziHO*DnS2dlI^OsaoHJmFOt@`~dhz1M%y)32d3-^fma7k>!-cj(9`!)K=-k{Kj3EgB
z#mv0WB#9^FFWIERP^rJZHfB_0&P+SAn&HeEP0FMi7SRz%f#;&f&qZO^gEQPcx75Gw
za+})Uc83k7S?o$4R{$o;!r(%(Q_Lbl65;g;UAe3&K`*`@2_AIC5|$4@=08xE-1VZ)
z+W&gh6)wj7-I!F(`8hX$y1~R%5u+4*NW&}@1rwjynZrtKs{qG``J<zqm*2Xj6bb^B
z&ntZC6ry60co!8C^MIv9j^kYv;ivD%>2B}7iFS-4O!_n(BgqoO&p@yoB8hM!CAxS)
zS3k<R+-fKOjH})3@G<?>(eZ@ntp$stRX3}ZmXcZL(OLVVt-TdM`U3;-Jmrn=-+0HI
zQnhh6p_@t~9Fc8s7l46#w$ghjl3`?IT;*=ic6IDc5aKu-Zb+*~3%W#k?;`6O==U=`
zZ9~qkb;S!ldFGm;h|sX-b3WMYYpp_fKh{Y58RFs9jv#wm5#O{o)LXCal38r}LK=$b
zE;VR^0ey02Mt_Yrs1X6%_~OxgK76#Up5B1s{4G2@JXX@-Z0(`01S6xjCfKm&-0Dj!
zr`~30)CJz;oJ^Losub{h=Z1WqfMlq*YImhNtGi><J<6ovq|)J+l+h<f;;NCW5h#)N
ztU%1g>H&yf2lJE|5;8JmhxwIY<fz>wQMz_GXe*ySeNtjmgQtyC9YN}V5TerltK#Bf
zpM~-8k;UjH=Qqz3HmS2ia|k|%HCLQMSrDMstV7SzTuG$K@(Jvs6*g(CF!<!C>bMw$
zIK1}bOEz0;D{CYt<JaP@0uzU*I=pkHO%^JsnioW2<D-cg){A(z<N~)c<BrZ$mc^9w
zGrBpWZOI&?-p9;^wmw_JWGSOcz{MzXnJUL;b?=r>(gpJzVkSn$Dd=pNz61@Aot>R&
zGKn}g#$qkRhT+NhdC1fqII8rAO)PK_5u!MJd%%_LR5<Pu!ND$4ZV<bkF4Fl`$H&o{
zArcwXwv%o+hAz(PqZRMeXnjX3bT8m(qzlSQv)FRevKy&Pi@ZE2#(^$?xW;RC$g)LP
z<@z`}b}Ji)Y(1?=OKvM;IJ4^~Nb$bXA((WiJwYho=y3Tmu^)1oNupx+xIygN)we9>
zw;kV9THWM8I`Kw+;j@y;AOy9R&?;;Mbkzo(a9}x}yL9J8wL2at0O#rv;1>QO(G4>b
zoUppF0~eXn&!2AzYd{bKio$DP1^4JM`U$Jy4f&k}c@>U_BMx8tuPk*u)XixE9QXG>
zremge7oK9GbNf=$C64cYH?N-^6Yb!7pj=l~d)MQ2jzplePru<8bg#6eVK<YGr*@It
zElXL}^u6Bge4_3Ps|VNkd5}v8AKxzrpLKZWk3P+3V3Jfy-79N)ub`bE_%W^V6!>)*
z44cU{#CuF<l87%1Mq3hvg;`mDfJtup$m4M7F=ibuknmZ#;rnyYI9t!a;Hk?~SQNsO
zh_Z#iY)~B{=Cz|mB)@(}ZbrMO*o(mv>kWvWI|lwaj|4vP3JuW}0}?h^NhEXwy=GML
ztw>I>QAk3+MzeYdKYWXKbV}@Oo+W3*%<8d2@=7Q&5|t3BYIX{A7Cxt3kft;Znq`a5
zRhT%=1)i%yz@UxnAo6uCf%{F&9bXF79;Wy5IttfesLfQDVC3;^xR}P~O2@hTiO=64
zKU}L#<juWpJXQS)dFd_k1N&ycH%3bg$0lP%p1o%A=822J*`))wWR-Fr=J{|OunBDz
zycATPl$9(rcwuYHd6WgdrG-PPYQ8_viVR`^UQ1FbHv7ir3x5VbbHt_@%?#3YLP_A`
zTm$63On(gje2j6g$mttqA6dKoO8Rd3h~Stb1yVkjLxV-$5R4TY2VT>`nlu(^+35|U
z^@WHQf-vrMK8rKY**F;Dk?nt!?fV|C938H7rp{VezCu3!(JULTrT5myC+`poambr1
zrh^5UqRGS+f?Wh9a3yHLvIP2MSlqa91)Z_A@4jP$I68xitdo$BqwcO<Kd!`tH>M;x
zEmVLva&4WM=7}uE9sjBlqtt-kjqw7Hc~$_}+tV^KOn4NZf~s;()xzgws0ip;j<hF(
z;Bif{!DcJ_c1qK06HT(Be&daL1CbhYNXTO~$$tLS&huWY*;?8~Q>@vC?<w=-G<91j
zeRpoqA;ZomEHAPqDZ};zkazj}s$h1$9@*@M^j>I=YGb>YPy3PBX2y8<&T(xTZ8Owe
zqk;=n_q;LdL$q8C(;aTT&)<x>01NhtNol0h{ghQ_e)j9tMVbQ_Y+j|+B}Uir$}(kH
z2%xH+lKb02e=51!6wN{&b*A`|paZ1pst_p~df+0^CQpHDe38&G!k0s7PiWkR^2hJ<
z(02NFB{u|oLBNM*-w<(T2{a*dDnh~>(G-bil5b2OUkkiFM#taOMDx;!wO^Obi>mr$
z4Z?GXca^;h9dY^td;f>HxJ&xlrE<xu^Z2+^_7GXi_84>`jSyr|)NCo9VcY!2h1__Z
zf@k;`?f1fn#*6kZd4$ge!5SB%{rpjB@q|GHSM~)hn%(OZO3LLS=mZ5h)?Dw0dC?ay
zzvHZJU?!X$-?RVJ+WuU|G0kYOudjH22h0wf2*hy|eF5j*Osshn9*95zA7u>tvMytJ
z07&d%(w4UxM8Nw~fi&ETZij<A=H4Xdnt>owEq+sS`um&7@P-fH^X@CFb=M|SNoS|p
zN*yqHF=kfqK>E?1UVam>0ART=mM%lb%gZ~Jch`FE1^}QTy$tY|HDK#|cz8Hf%6Ms$
zy*}2Q{#u3LIe*Hw&Cc+P{3~9gbo1sY@`jM!481O9&118f*8;^^vkRZ7e<a`8A2R7=
z)s1Q&bbsOlsW>55aM~=k{eceZjHI{q5BGSI5KdI?FFJEpH5x9P%({;Cy#r*vItf-y
z$(L-TBn4J-T+J0S?0pzgUC`Ga>2>C4I&Ir8USM4+rrvo~R#p}g65=P2VrVUZk~_*H
z*7##EFM)K3<-nkiAmKPB7VUJOk>8FpiF7lIeZnm<$C84XSW*3Y-RR?NzE3iSrqyX?
zpDiZqr_XQw{F%(|k$>XX2cXhL5ETJVCso%yIO_w<KzHP>9?WOuI7vQZ*cZ|1s@?D7
zP-d>jApC~Aw}Fv@84W9dk3o?da8a%u8T#BEIzOVSNBn;AY`znlD&Y;=_g-G7iZdB5
z1tK#lEpx)}TO1{D_hm}%V3$NKas?b>JQ$lRmP52oq=q(x`B5Qg=hTMFuVSe-#71_w
z&UHoJWV$Te>GV01@S{9Q6XLYw?$(ONp)U73m+kiTacfO7c`*Od=+yh=9%Atb$#b5V
zK`70=&;T(j<XEUeMrU9=s?{AtRD5^^N?zYz#js`SW+W~TQ{@e=G*2l*v)VFT;oOjH
zw#QVSe9fq~BfR@EOm@ud(^k5GzO+4}6CuiS72oLG^5p8JxOYk(uwb}$eP2lBm6adr
zIj}EVu4OCbCMC%nM+em)yoJXcR|NHyskseh#a)hJ#0&@Wn1-f6#9~%*)Hk<oc+H*k
zzAK%!jf>cP=$-e$d1|fP(3;4P9afZIkS{Qu<;iT8Asm9?mF#!1y5UbXoC_h-IY-lB
z*xc9G_wCz9$T^b@D1o{pohU?vsj(qnx97UaA1o$wwV7x2OhnRy>G|pY`}&0MZo#5?
z`T?uC<eS^Ah^C(A)o-1=*%zA-07z=-u3cVUR;)u~;WCj!&<T{954?ri7V>yTU%ZKH
z^ukqL6(@o(q#B-E)-(9=;5Dm)luPe6p+!J47@!*V)=S1e=ch|<Iy_LuqZb3)IwY{6
zQYGGpdcDSjaLi$Wy(pUgyr%%r3ltWIOGD@nK8=`}%8Z%CD8BW{2$64TqG=b)cfCfZ
zaQ$`*g4Dw!Ao%R&B#%Xr;1b`JE3f#LaeaGCbZ}wuM(W;D2C!}p+Z!OKeB951B5f`!
zlqK!C08Z4`3f1>lcPA8dwC*!)CAr}j$&8YAuhsB1#68uTy57D2xtuyW#yIETId@Ed
ze?N>pRDNm|szBY#A~>#eB4$=r&STam*q8d>(1?zi@!(}p1em8I5!mtJyvs!sZlg?-
z6=C0^5%B^jl8xxby@!%Ci_UzBRvF5`u;+ihGBH%-R+&B7lK}TzQ-vllQM)VQvEIBH
zEJ!5edgv6NO50?}U0+w18+a6U{mvbw9BhmF2?rPZ&&n6|H5?Z&2=NIxoPLW*5)4Th
zyrkj6I!WF<>lx{EnPq-bz53Gj_on*+Ww|LO5dRIh5bxYcrBwIDMZ8iwmg+(Q&GNz5
zG2V4U8qurdh%v=W8n!*V$O5V7@98REv>xxjYFllzW?Tg1b&1jvYLT|b06|M)V0Of<
z_?iUDst~%l1VzT5p8;yY%+PQSmVhv?0j(_$yhuPt0MzW;ooGO*-=H`Z3`r+eCCA<)
ziymza%(IG|K`s(+S8tv0rWdUR<1cs|tI>>iUcIBF?r8^a$4#9D7zyS0yloS_8@0Kp
z9xul~u_?KePqdj}8l@%K5QwpyjGn17GbwybvO)E+>+U_(drIz?Z6|KD1;;Ppm|HD9
zZs2c3m>)yY`y!jXNrT^h^kRmOQcy%hFC<xK=zu_eI`--PNpA+@WVaukZ4+_%mcdtZ
zC89IL_(+bRW9=-VKeGJ-20k=V<w)!94soTruqaCYs?p@oL;1U7W6hzaly!Cec>!O>
z`kHcj?8+7b#}rHhY+n^s1w<rEO40`-;ffk1LnA!;K{p2Yw7CTGkl+gp4BUJNO<f@g
z3DRBGfPOJmVZhvX6ji$si%e0dADmcWVT*b8mBp1|zHi1heD860>2>n)(1?gT2=tYY
z>gCk=#oKmfyH`n+>%xwmBD^v2qh8FKxOBktN8@>$!RdJsG)+aW?#6}unf8Wwv~$tz
zesac#2X;jFP3RoSed}82xps>mF&sl#D56*ouGQ~DC~K&|NP#H>sI1{4%x!IL4Tr`s
zw{_Pe<M!0f_xU(^bulU*KCE$z(G;4?4`64tmPxiOfBZ}kGDuvf6jia~8N2ghLP;bt
z!6*BKja5%PpH`yJ_c|U!YTtX;<28q#-=5$WC~EmJ`KTE0;M>AB+uFRIlPfD7*3UE(
z*hM8;@j3l?FggArbcN(?<428C`r)^XxMP5$`AAe0AidCx{%&7Vt`Hv|KTA2E`TBMA
zel6feYDeoqLB(rEJYHN%9t^D<;<@Xq>L=(H`YHanF4@#5)5Cp)RN~YNjQg9*TC}&`
z#FlsI=KA2tuD$3<xT9UH^1YTPqteP^b@Zq2M;$)4?hwm?EjYrutq!|&!_iiowdZXa
zgf1R*ru{i?!@Do-P&X5|xyi}MfG?N+?p>9(kO%0sGiXD|?*{n({J6hkV3S6p_QAA4
zu-}c@(0}lzGM?K&n`G`1!84IBB+O_{``Xv;?s5J|Ybql)*&kK*_?PyEJP$4LEB>$^
znDOUT>Q{$lq~*|YYAvlfCZfw)UMoOCLS+)aTpo_zL4eDGP07(^bSJMD!>37BSexfF
z)3;=&#Yb=?auR(KAO9{HdX6KZ;i67SF)!ZZg6!=aA}m3`;pcB$Rs9iUKc;f0itxTw
zJKN>eOda5s(dPytbGyz!|0VJ}CaE&)Fofz&_;rf$5nc+_#gk2lJsWM0X+-|38_IZ+
zG{hG!d>4R6#nI3296Ja9HdsjxKY!_m&K>l#63x9bA7Q$S#J4-iQ|N0~<9|MzeRmQL
zo%pc^4@iA8j)udZa-VTfU<8E|ce;P%Sowa~z7v~p#h22&x8BkA<#&57n=1K*HrtC0
z2q;vFDyJv3B|&v)l)hn-<!F1bOo^!p;CaWFGgKr-$tsr}*wXu|=HY2m!o`chKK)1i
z>6SY=vq1l2RFs7Dm-eA2P!2TdK0(KE4#5v+Wp#C0t66q=$nFI#l{{PD3D$p|e1Rr}
zJw~=cptFLTt)G5ji!Vy6<Z@|=C7HR#E6h50wluMxap91R5&G!^du{C50yEd4ck2n?
zcgjo{i4OFfBv2(2rydL)-+T4P?Au9Wr-XLu>S#R!&2X>-&PZlJeFJp_iB{V-2mcnf
zTZ0cZ+I%h_oF=1k%5_ttf+KcKtXH^E#BU$LqNssH=Y9S1DxBTuuhLG*g9Tam3+~s3
zf=*)Vs4K=>*%!<sr77Q<zifVdHIJS0thTTW+u7_91k4YW>JmOX?c9o3Hl6Rz{L;_0
zz>?)ODup)VmJi|Fo4jexClPk_4(cX#szUvQW|2;4avlNKJ=<vLb-HrWai)-+a(Vbm
z?`^4fc`nFry?8Q3)d7OnsSZoRW501b+I}CP5#Aa{n8d^nxx?<GOwC*d_nN{+;cH8a
zwS|RNg&WX+$S5c->b9j>o6KBgJ1CSy?%nO3A=+_&?ADhp^iyK<K@W^Kbkpr7=Pz7v
zUD|@%Wx>+4UiM-42jl{{ksjn291gnE?`1V^={TwBd|RiU6-_XCjY7cgdh+^xuFA~y
zONrPG3lsN_f|2QB!8cgFJSr`Hy&t0;EF49u)3_;UnIDYm5@TXiQbTzu&qD8w)-`2m
zDJjO5giVLFck24dDKdiXl=-(@UoKuCw|q9$GoTc7ZFQHO>L%Peq8(<*7iuSI5@#ld
z6hGOt?iXr6dn_++G1#|N&a^=f!46<@1|mQgzvI@=<!b6SV&UuHoU(1M_%e%WjCg!#
z!1z;Ik#M_ezO7fCzmg|^v2Isw4tG~tUAu)w1U2h;WZ#1Y7u1_+yO3`&CNcsz-2b?U
zhzLEIx8T!Zuygqc#RUi7KZy>TmT&>UeVTN<!(2PPNwhk2>j*=h(PpmQ%~S`%E!-6M
zXO>Lb$jeZH(jN-1_gmG#5*f~PQRM_S?4&D3KA}bAt{%q0*nN4Q3bA4?kP;i5WJ#A~
z%Mi;KEck|-ub@#pHLmv-JQCXi0~G8+pZP$Z9$ZL4CaRO-hTdObJacnz^gPwjiv?wV
zDMdcBHlH9Na+QfAOS*QTy0}?qMv24yslnIC1K6FgxcCLP#me2{X)MTm-80|JOqp4B
z?M*=G0HGH4>l$c83I{@VtFJ=42A<W?<eo?0K*C_8c+>$6)^zX$*M_9r=z$q=s=RoH
z;3R!o^Ch{q^Nh8^Vmd@qdAh`tPxjVw&YI~tZ*APjV)!)p7O6ClT9@eXu2cLts7dD(
z2TjBM*3*#^41?x;Xw3(6Flr%u`n1m$%z}k|>gbej@S1JIsG}y=K$FaD!M)cgRV(>}
zMrQ(@;#KcBStXCOT%!7k!?J7Fac}GKa3qHKgYrj|WR9Rh*o4<k&+1dyAhqt9Ip+KH
zP#h=AnyBo#f>z3e+huxsI$=`+fMr&4U*3iF68Y?TwBa^=ZD!n_?xvp953cO)Z7i_7
zq0uBy0?s4gEygLa81d0Ip?u}>=^tql+a-$Gv-<|Cij|%1#62sooP<#;{a=vx$lUTJ
zV7z!uDryy~V%`q`E>A|fUTIsA?C7G6L*M%JI3kv>BTU$Pq0WpypP=G!dNjdd9|+Vr
z1otr^ad86hHpNV^eSi)&m?d?BnyK4XUvnIHOJgP-)_5uV<<M*B<#WgG?BHAb_6p&j
z@QKcRT#6kYlDy{k=yZc|i)q5(3hdS{l90eQL*M^q<QcbJKwY)_*Ay5C>D)0cDYb^J
zs?CY}{wl81?#w@)!X9Bg!#m{WL%nf&A3B}xp^GmKq$%@OIwbJ2g{xhrdlq)kw@H*m
z&^^y7Z&9B8(!=?ri<;w+MKP(N7gw`r@*2EC5dYZ+szL{)6%^EiYZMqYXU?62iF_|y
zFZj;ff!m%a?U@qh&`O_Z_IQCb-3?j$4Q`}U9CnVrHa@mMs%fEk=)DWE<6O2z@1?=?
z!AsdnyBe^cK6#@vNDFo~vYEH{Y=7M9%!u2=iWAoC2?#$)CT;*1F)HgB6&V>BHTC?;
z$_n@!CMUTZHVk#g!C@gGE)H!c9cwGOXPWZaPM+a<N*dO;ri3>i*<NvWk!qe|5Yn#w
z_+b=I{N0=zdYWyrmERNHkX1bI*k1`g)WFA{-6Vu*2%gD_Cc=u#^Ax+cattyw3}Kw2
zNh7Pix+Du<RUz78E>l(f(EdI9y@Ln;(|=lpe558#J9$;Yd(IP%p1U4cP<e`uP2-K7
z+az?R>gVd_ECBdu4iO9IwoS$_0;^%AX5056_^GO|c0K+d$6Ka}`}uJEzYd1yT^amG
zm;d^katO8!atv0N;JC@({r89bJ{Tb-ypzY>?p?XyZGH^>3jRKQ@;K@k?i+vm0#g%K
zZP5el!FBB{rQDkDbCw_0@4goV+*^5B)Ujj7&Y8DKBh}&ewY6%t<+NSGs?`cvc@-NN
z@bey<eN~lEEnLfYMUxEoq~UbJ12Z&!pDEnwh35dz3r@nnUlVh$b$7vs2kx$*$PK+(
zcQ-#Zc5WIE&L-XwBprXB&5FxPc#Bl~zaW*k7k1l*g%kILw&4o$`|SSn7r{#cuU&P*
zdzWhA6X;i__w1{SU;npXggK5qd-yAPH+Z}N@1H;9ucQ6jJIMSL>ezoCTN5GfJ#{#6
zQ`sZUU&O-`!~fZLwy=^;ait&Q6R5B<kS6FX<vK~dId$`++-o?PGwVyRm9q=_So%WR
z>P6!7$>*CzM1#p*8&5gqU8mb6JRa`a_#^>4EL*KKA}+2By2fxDIk1|+eMwqUvefmc
zLO#>_dxak=^Q^7ed}~7Q;#jK0w(>!<sKQH)l6!(q_ZF;iZ{wP+Hpr%DS03JVJ=_Nh
zAg;dh#dr-YZs>$v1o-)3vTLAErlOK0DU4RBl;C+8*SIw~5jcikv^3|8<sx&g=zh#+
zGv<Fu?du$Uf9hz$W0bwoM^80^SdNqEFFQ6W={*_}P;<a|W>K|7Q1YRtvCO6Wqt!Q^
z9L>ytzX`od&DX;T4m^{^0vS>)qAJb~lwhY;@}kE`9X}%#`;rpqu27_p7J4{@J>*7H
z60R4pHiR2TNT&PMsa&Vl328=vg=pkc7sw`W7XmLNNkMQ6cCchrNZ^0K_Ek|<_WK)*
zy92`;I^^oxRLNUk&J(F$3P*16a`_70&5`^Fmzx?Tk&juT)3-R`maGj7B5R7)ua)sk
z>&w~-RCOoH3&lbt{6#JI%yO0d%AUc{MLkyql8E9@p1cH*=D(CO>ONXwsS9v`gw~nY
zkPE1$nulK!9TZivbS;&cczCd6%iZXwobJ0<OP&cqHUa!C$_fR4N^T5FD7k&gpSLbY
z!|lt-ZaW4xaEQpWoB7Xx5P59RW*{z;Q$fC)7$>ngZ^2a_#)qnckrhynWvT2D5z(kG
zlPRiZ>?-`20J3V^-Q@TuQ1)4n@(k^y)6>%{G#?x7J&yk`t&JanekWP7=y(d+f9Cbb
z@<4$KQfAWLmn28=ynL2feGfIpPVyFEX$?L<2ax!N;ByW^1JKx@TtSopEn5=Ecv#Vc
zHgs1?oquVsNhUe?$>J*UK}YB{8H7p!aNq!g`VqG=m7CneVo?@?Weo1fM~T)Eo_JTW
zvgT^pFzu*}yAN-{>3x|$MX9ej$H-~CXWE6l08BqJR5$YMaul-t#uJkMzFo<kuE3M|
zKIt^KRN&r#yM1hHJ;VwsRu1ZTDXs~NUvts3*i<NSqQ8%M(6s?s0CwqDba!6dV<WG;
zVe)z$=t__|HSf+K#*X6?QSU<<6SkANtC^o%We8*Cokor<v~wF--+$&AZtF}<&Db(g
zI4pR4*IvpZP#Un|AIIPf*kokM7PmU>WS3)Oq?j*OxTdiXtl`th0jCWX339FUccICF
zo#?iDw_@&j=ufRyR?@SO5`Xc^@|)$s=Jj4bmX|*=LtY7ea$VgimfL6{RuajMJbScX
zSKr)D+_=d{i4K!jH1cnv#Y~K<-_=TCws)O2zPsCUJ>*?me`&+!Q_heY1O%!Y)OBFQ
z4*4bJoja`~BbQ5?p;iLw#i`{E@>oUkZN;V?4~<IB5|v8S;~NFif))r1@@9qy#OBF)
zcSz8Rs^@71GQfHBb2^ZQH{kS98DKwKOhC(Vh5)aRkn0}S^y$;5Cd=GhTwUOfqn_Qq
z5aE4lt6nVG!uRqkq~aOjN*2XHY1kO`OG8`f`~2Us+x++m<FP<dwOjfuSRV<|bBJl{
zF>2R}cq2YNm0WF+a9c7`(gKqKl2Sk{{;c6NkLEtLHG$;$rIJPLD7;=>Dv3jWaW4H@
z^H6aCH7_qioj6nZqE6<g8!`yiZ4HSxf6gW*0@o=;{G!QS7JmNH^->=n9|#54m0iDb
zr3wz!0oFzN{k10V9|=_XTBB|bw`tSbDKcf<cRh(4B65pY-V!V+TvStRr2QO|MuoRB
zL?`6rT@E|O{dU6<^{tPp#oAjiP>uJ7t7fCzm}wD)Z{JFHHo%~58mm0+dJMgv@3)@q
zdf>Y=qo$@-B6zhn)D=R4Li${wz1^OIcL@bgq2<=Hnr1jbLk?GEb<A5iHijr9%ZDj6
zr@CFG)`@0qGl0m<4x#BmeJg9dxIQ%$8WNIY8x<6!IbH5hUI9Wwi=(r%^V_**cwcK7
z+ZfJG#bM84+^u1lJ-@kqdm)inyU=VTW+z|m*|Y9mdR~THAot!_iNB>skR#!-?A*qw
z#PxnXF{l{<;)_ODUGnR3tbbnR3G|&4==NKzMekL}i4jg?0YuqBhtLzXV7}6qP4`nB
zKv3gbdIb+S<~M@}_zE*lrx(JVGX<C(w=^8#ns?`Xc)gb^TyRMAZ-Cs{J62&fa(AX2
zOnN<r+~BiK<#YNY@u@xP5er(@b=n(rfi+T|93QwJ|4V-x-^0RGNOL8psm?6z`0?X*
zKJj%L8fl{?c!`SkK}!YjB;^$q-@q%rwRNpVI9sKVfGcWa!+OEob$|P&jibCgNsf;m
zYR5ECnQLwCp7HGmMgBfCWx2Y;k(@lEPRkF`VNhw%fS8$G(<7erSQ%=I<9l6R3v$1*
zt0N>va;x@gR+@93he%VasdODjWmZSF35i(&zN)p-<NQ&Jlb2pd^E8j(G1c1X`MJ=^
zTptD=YT>*0AB>r5L5GY<ioV3lFVLD7nnO^e7CLm=PGH9dZU`RX{T0rog`W4j`Yr7N
z;a$~iY!~Ej`)&9#MEGEgC<OI5Ce2<6Uc>56+p{oOz3FNX-V_t(<K)Nk;OxPv8+?2}
zp%KlM@1If3e0rFN&BtLqO8LT08I_&wADD@g7;QGTs?iXH1v_!YNLyRr&N-NZn@;=V
z3jO5V;p;PHW%{vA?LFVZBbp_QRjLrTH)jS^g@7LP(8u-M`ST4m4CRdID}p$OZ5?K>
z%P88HIqhX=#fYKDv+qS(7ACAjgf7IA9|HOJSa8q#Q2z7t5eaZO!0b^rn90pqud6mp
z_#i`k7nx{Z$HHOr0429F>Qk92b-b1BYAOUqvn_xhj2vZ(Y{tnRhYz?{71fpPPLnHm
z+?;<0EBB0aRF(<@%+U9hWxnRPt<M>G^S#WxmN_i8Kkab}m6@qeLr(5lZgtK+*8FH6
z94eZs?xxNUVE>gfk$miQP7`5d(4O}|rLg&xhI41ec=|gR$Ba-uS1WUcC+#ru(@aEE
zp2PgE;$C<S2>cTfwC6Uhhg*f#=7+q9XzHE(0q$1;ycwp%r<saiCaqzRot~OHU0K$D
zH(uMK#VUP+C7tokJD;-E!}2GY_w1q+Pogq|QD{X(?!D*g>goc{6m*C75EhTBCM6qq
zaR;Pa9+UnHjx~_qit|AlhWnrzn>mx`qQ{p_yY1iSrn51G#a}-~q<eb4H2T}Kse^If
za}MiDzV9XCgOCa5RVvSM>yB99|E1Jj8RH=L%_SzMPR2QE7FT7x<zdJSsf*X;I~p!K
zBHPoq<FE5(GlP1Udh{hnk9GoG^WpUgOrX(O41+F8>lD{;fzieLm^_yQhs7<psP*p-
zLAvrQPYLKmd}Qdy55<%QR+%0iSkJZw_1GS{Hi$K+1vuvJPWD8DUB^=FI|ifu`Lpem
zQ@zxSpKn~JHTko@95{A9OXrv#DyASO51_p19<N&h7boy|F85_8rl-$B@J<ctp%%I3
z0>uI@(I;^xTQ2AC>A}7EBZ?iCPM<ctbo%qT@<HxA|I?KBd3R&46*pYI26Ardw@@ya
zYvgDi8=4Xgq*cquvfV-b7?mUE6OT-28Y8(ehDj11HRriEBu5{_PmA_;aJh0goDSE}
z3m?U6-Jg`A9Be=4WR;f8a~i;4z5SP5_o-MIC)dF`coRZ8kU%>kyS@QX;v|cVn8fxC
zr?9>kd-y1Yya9#P6qiH?fi}`g)7K1GVepAC+|p7l+|Ez9gg=0f%5`l%fUZ$5rt4>k
zh^AK-wpZJ@pXfN&<K&0PNP}<X5~(5aZG=YPcdbI`$EyWyO<UEzG?0m0e{r`3nuVQE
zUs{gb9hM~CHQ!fCjYV76Jf5oR_+G$}o-Xms*Xa{p!(rwrFS)~$Ul82viM$uDV+0(i
z=`{3s#{!(xpkT`;DEcF;!qZYxDD8(rz7V2^caurd-(VE2UT`{6QH*D>GP?#!o3w&9
z5{XPoO3EpY4-T##>sy;jjAT=mFV7_N4MjY@D(d$KDtD0%-_FQoMBQKN+ihqjioO{A
z6W?>OsVLxL8{+OqgoQsWLz?XI^@g?CuBU#@Z98oR^6^JCw``+xKN^-ScpP_O(3B1R
zLuC>YS_DK3&_J;hr84e;%v(WyqCBWeudivU`LJNFPO-C3Qz~gYekq6GEe<%A(|!*Y
z_ruKB3j$rCW)e6QRlTM_z){^%1983K!KNk6xqvIE4q9<}=yt4>on=a-QD1-gE+epX
zdg*yG{?lcalRdV^fpk1Wlx3fN{JT=}C0?p`teo;48OoQr%TI#Nur;ZvE;Z*xY|+21
zP<t<%bKX2+DF@fq3(6i9De2VgEELtG<G_Z(y;*Aftqmzy<-ZGW%b!Hg!#NB_LK@Sr
zd>$~+>%0E2Fa)HfO@6$vYej44*;gHguszYCJ7Rqb(J_Tw3BBcs(ZuzrfyFaD;72k$
ze_~yiMl{#^E3M%jl%Dz@*L2Akr`J8U@6Y7OovTH}q-flveE8Z-cZhM<W~w4bcIPq7
z&V1j#4SvkH*ao|+;h`a#Jpt-8wMrMY@6$aXuQ8yns1~lDS)`g=CrzzFoXjQI6~$P&
z?ryNYhlMS7=>bDIRCxiQ8EpO_e}SV4fI?cGDdL8os`sTpBgGlsW;8&rJA<hq-DDA)
z7^4AXm+9+O(}%QV)h!Nbl{T48s&T+rs(CBCRsGZO=?g_4Hq=f3Y!K_*e4ZL~pC_3g
zay{HJSiHTvy{$-<K~GPAOqKnCSrA&_qfPwNAK=eH6s({5ypexr7W3>R<^efW(NzcJ
z4~(B$ollfA7SUbz@qZ4}ASfusnW9|Jo;h<2Rc*nvMQmgDXzGBTot-$(bWUC-Mo!G6
z;I=K_HT0EC@avS^|8^LklJCPaeTVi+hS;@pk;RLPLxhBcxq9(nQx6YQ>NyL5QNQvd
z5nu8%C1r-*y7Xq_VYx-tOQhsavUJ0QJFtsyFnJ~ew}jjadoix9&CRs|nol_>C?a7Q
z5tEkt&=nRgaK9Zs3}$qC?ok?cFK!5*{W|g9rd=*yb@nOt(Bo*u8GO(MAH3D>IG&^|
z5*Z=C_`}omWqZ7^j#uQ@Q1ifmfcnw#Xs%xJ219Px<@np115>_a(4ZNpDJ6d25Ffs+
z@d%#epJAI{bNL~;^7SMtH(kTh%1WnV3O6k~i{Gi7x{-K^IF{F3zrjfNIt$BbBKdZ9
zkmLJZ!C^wIR2Q=H?w9)2Mt@_ds;K)zZ}u(SvHY2uw7{vO=m6(ep#zM^@_gR@ANJnD
zE3PhSA5B6C1VRV|cTI42cM0z9H16&m+=4s7-CY_9!QG(=?(S~)@V+zieZM<%*PTD$
z$6_rG-F<rRv!$M@da7zq-lgK|B@CxX(v<8>D4cpKFKPnbySZ(tvkB)(rmcCvzywWx
zjsz0WlsW-B@cW;{s)lQGnJ25s&yu}=hB8miY@dPjiLD<1>sy(~<$R>c=~1!xP$QDV
zZnMG%9GjS!r~+W+Kzl8KA%F`TSaAIV0~F6EfOM^Fu)RG4>7ouIJa6aCIdG_m8@lMm
zZ|ntGmt9Oe)c4A%K@Y|dr!vONDp$Rtz$=G*7CpH7a(i+Jm>wP4D?ql7rP~{<Xcn`n
z;M)Ko5smN~bRcP%|68d(*i$xrrZA+&xBwJCHYR%zU=8GI;kGr1Jk*SvP*G4c+iy$r
z^ZRyxo}a5V1Cm%BHGd=8iYbpjG~-&|;CwH5rHn*WA|y@p;~zxbPwp7QQP%z-YhIw9
z#?>2{bPmwI;7bL*t<MjSkI)+0lPe~dc1LQ$2&Dq=O2|;)K9qsM;KTFjUl*eQ)ETu!
z02Mf98@<tJN)8r6>QKswRG>;=`PaM_hgCsaLI82<e9DzEHv$HecNBSc2U#0etN_*j
zXJ&$QHI#wO-+(`fyR>=jhJntT1Hx=2Cd_Z=0Z1GG;!i7CTS0?p1t3%>jx50N09GD}
z6Cd^dC?b$4-%1VmIVxi)ypm+lN_J?C0RZ&zA81Rz@=o@+xq7LqNJ~db+C3BWawct*
zFNKUJK%>*!=cKY^bob6*yjA<{QUk67w;;31?9`U7ZEkjZfB_vynG8jGkOD|^{Jr!}
zr3f20M(N~_-lAqxXv1ierL6b1*~f$Op6{DHvC<E4_pdI=dI=Z&%qBU~20k8q30At8
zrO|sICnY-VWLfkh@nuMX%)_fUpOGy%gGJ@0fSQpb9>0LxkGD&S{oA4`b-e&jSC{+`
zo-PT+-N{b5383|^+yD=`Zo?HN({oy4V(%Z<!BuI-AVXA&<M&Q);!2XH{i+eZMnHe1
z?i|%@Y!Hq7fucNatc~P3kvwbjo%v`1pW7_}%%>r~K2@YJ_Zc|Cknp5U(wiNMQDTqQ
z90vY??jA6Efap5ic;KK|gNkc_sRABaRXPIL5I`*pV4TOq!htZ^)4+Cv&y}GDwa_)M
z+CiyGzXSuu&`S44tc>(2q4T@_r1s5DJOT5J(_zO4H>}ZZI`CC1ptt`jUIDT+1&#+*
zvyeB=jnucR$pA*>G^s(8gaJx<^?v`t4ICB*k7hKVai}8u1k^1UR2|y|kjU=FL$4YR
z%S)LS1({PPyl*zW3#Sf9@a(g>c4FR31mF3KR0L{xf9j+N<YW~F^ZeomJr<`p(gIe7
z)4@Do;;CYeS_1sS92f{{*rX^$tME)FPmgV~Rgb0e2RBK#5yZr~3Vbz=2l5Q{kDY*q
zNBSv%BFY%`fG@tZv2kuyEzj^bf%us5Ub;ppqDi1SOq<mA7V3PUfIxx@7KZ@ClnmIQ
zrb<X(X%meneiXr2rF^ZW>5VD8lFzp*tslp|E+dJ|gA(TIK#!6of!YNU2h7>ql<@-t
z)_>Z)c{$4-0NUZL_R^VNy^Abu4bFU~8-YmK#>dG$nd?YZWHJMOfYL=OT{}J=^{9F|
z&I0}q5UI`J^>8yarPy4|Pl_c9cpd}N4cUR{nT2IyRb_HS7MgShWiT_p^HVsvFfQ7}
zu$WvEqMUo;a9Xe^Ic-$y=|onsuhNZf{C0b*yXfu){hkiX9>atr4WzX?d^8mB#Gx*W
zt``^$RLkAp+p!q{;p_%MrBYh*6;UY95yc~Q8=OQQ_dwX!y!F{(!?HGz(yCkn@_qs&
z6flhCfS1Hc0yzAU{763Ud@-@gIRoW-Cl`nq_Ix!@=JrMBQwS(*#VL__xl|t|jG|ip
zO1+H9LQxohX1P`R+ap7KW~L5sZthkXE`y;MARWxT6%#0Lfiqo_f#7q%um%!=haPl(
zhy^@hU}A<)-z1RJ(M507Zj<@|e3^tCKK1(p%J(}45QC6BcgFT_BDa$ic{w9Z?64)e
zD%X)~ZHpv=cfcVqM#uc!^72?m#_tQL70U+@rt`yMCrhtR4*{;O?uKXl)4}8QGCq^Z
z*hYV>Xav@Djwm8JIyy^UoD?yzg@a3NSq04D>t@cO`cFndK!^rF4GM>g;(bV!cIBy0
z!Rt{9u;i?SyL6^u9Y_6$55%F*!KVWnf0aG0E{w8XF`#i<gD;f<E>f<mc20M23Mes6
zVJwTAsZ`qbsn7=X8OW9MD%A&hFZODAuc!Qk0GWsySP}U^z*L2zN=4h98M9yCw1<v}
zTqypQN{&&UUgw|4GN7h`eLJ9dpT95s+#p!}3vh^<qu}9L*{)5=scnMp7=Q)2hnw06
zSNNush)Vg?M(OucFEA&y^+3sexjysm-@hvxekv9LXDV(0^PBdJ*K!F;*?2r))IrtG
zuqqD71R~eu4=`0Xftb!Yjflv|dT&|+GG)`*^2LzPTwk<McalNGpqul2Kuu;cY^?KI
z0)t<4c_M8dUUJ|(*3G=Lrx~0%>P+m&_-wnkO6BVMkQ$Ocpn3Ta;8dAv-Oe`$Nj}2U
zfbK~^umT9HG+L*HCR)bZ%3p6Ll_{xW-;6BbesWiEmFcSn94!`}H9$t&_3^SZowB&J
zG#+qtN2+z38mQ)QstBmwU&4fiQ1Knr%LDZ{-ras8bI?^0OK%PyrlSq4E)cc>NFp*7
z+F8KEKPU#6VdtS_KopI7L5?HjtR$}GG02IUR*E>P0N@yaEpB(ou(1GQ{PnHpLwqDY
z`Lk031<wn|@H7Cu*KaT)ntoMlY$OG~=-v#-;=Z>E%?7II#>K?EoM^=ai_7ED2pE5W
zoc|(|(F&ZTu<Uhz2x!?1$>Fohy3eY5wene<_HSG%T!$=x{YaV<1Vr5ucFo{%8#da<
zCWwpIZlCoGhKT3<Ah*3NncQXia9X=Z81U3*!UWXvfUlbes@ggI3V8pL^LcWz19)w1
zY)*g#V?c*@2g;P*qs$N*9B@LF@VWM%AO2E^*pTu4kKE9^=MR8G38<)O`Pzds4*iAn
zd*O!uCjI`M8v;0@|3E?gi_`&Vwa&lTB7pG&$YUQRfGGS=T+#oVlzQQg0NMv&kN&0-
z|G&Bdbd0eUS{Ky>{%`K<-y43Rsm_Nf{vzRg`8?qNNzzrq0j%5$Irq;CUDOwz4tV9?
zE;{~ocg;U<`p3Qh-5<4zXPA!=nsGRgm7%?Z9Z1je<;q&$433eGk%fSc;N_Bwi&oLq
z21KhWqi+T>aG+Ii)OUFKU))m97(}aLYUp4>z{*5ND+V$(HgNzh8E6Ge9qeU6c7oQH
zHr7@kD+dBLT0v_IYdZxSJp&M}5Xi~Y03>Rs=SnLHvN8s4XJ%$5`1^-e#MHt8WJfDv
zq2~Y+0vT8vf_Qk~{?RuV$YCe_>sPN{{rdS-WDH~d6S(}pKmT_G{*Mvhf!fKxko%7-
zn*D!P^gk9J10x&bf3@(sHTtbcSU-U8DequJ#@k+rKrMpYef}_<Q1_9G*$oihJ`0kC
z$QKpvl`}I2X=|3P(eT8Oo)7Avd)?mIJ$u~l3hss(EWv)I&GgoO`_QqfWz%Rf-2J|@
z6ghFUowmKrj)uDR=wu>|5}FofncE{}ADxt|^wUY=r&(I*u*N4SZKwyDvt|b!tyD4$
z7xA0T1TVQ&R25qnHC0oseE8Y7vdCL^3)`c#I7&CWT{mIIU)V~vHX5WEqpRbWU$H`+
zCEeg|uoC#b+jHyK|Jb<KH(B^#-YU4<6MynUd-J0wrSp!@1w#3o^h=nf&6CLAyrxx#
zp5`BFn$>j{%fHy3&eBfo-V762B2^@*Ol?5b#e%eANX?KyVhr(@;>DcbjP|E<Z<R#w
z^Y1_GtrC}gn>OYt!<7w<cM{=<TktEBbB<8aEC?%@38w9od*&}};nRvXzvKB~n*NTD
zL4Bg6?~z5F;dWDdlR~?gWu~l_oJF3?;b}Hh%{uJ)vkazlRM$Q6s7H4@+9(=rojRz^
zRjJyh@OsCMo8Wu4W$1cyc?gr{@fg;z3W$zEv_BNzrVJN6uZbYldN-sM>?$GFl_uwi
zpDH&a)F707T<1X`w|;9ot9`z=&Dk(k+nv++=K3AP+{6;KMlH`_6q8mMJa}8>Lf&hb
zK&IBSWlld3y<ax1-$UsX$13`?fgUpR@v}xWj*kku-Kzn{Sg5l{Wb&RNymm?r@Ss{H
zy>D++#YV;TXredAHFgdL$?@Q^cl&SrT4YIA*Qzx3zO?5j+uQR^^lx?0?6d^KcJQ|j
zF-Fiqe@cAA^#yL5hwluNzt7lcc-~)Ez&D*8?vpiC6a4%cN1ExQqE*Ix?n|Sbj$yOP
zV?2Af2F>SqOy0YOOQB6iWis7A``hNstAE1U0Owq<Idy#Hi`W;<!gsx7GI1LiGR%-~
z<=|2F0=rd7EZOOkhXg9RrU~_*Vr$T`)v&(^Z>SS%I<dWxl4|AYiR9`1_RWbZQ1tb?
zt=~=dmuYuyL_br20<X?fQB;OIBoYc`{K{}F_fpSN7pB&^k(5Plds{v?#9&yim-=M#
z2r>*LSWW~^)3CCu8W=w)PD;O~T@;(zxgx8D_xzl+wkOAzFr^jdqjzB@zJTaoV~AM5
z$L(t^zTY1&V*ay3$VRbP?8&wCH*<~KH)>494cu<rYD^?B^Vbw1)_`7pw*C(mtZvtI
zL2z}yytRq3*`=zwYK<uQkpz*E%;ldi;(b}9R89X-K$WSPVVG2hSn!)K=Q#GqSsu!5
zJdkN>s_kY;cf-|Y$?ez8Atbv$l(sKoHR0`)h}UIV*CBh!$44mnvaHY0xmb^kE75Qd
za47HO46Rw}ApG<<lIK*1f*tg$9yNb8Qabp93ZplUvFXfGDg9Nk;7z1(p*ma;j0rvl
z0Z(3i5-&?fi^_ZL_b@g?Oo-OB7-Zd1LCg^wDW=3bxb(Wy_eK%9(c?MW<Rsq><dY)T
z7Q8~`=`7uI5YM+l9>Wj(bz5rk{4BFJlI!I%OpixurF=6T>&Gke+^x;q7Fl}u__I_3
z4!;!72-2_3nkmp7`xX!yx=8XD!44!BfzPOkbXV+6#RU!rl#HjVGD%{qZ#Ta9>(gBP
zz+k+*f2}#A%m^}zsGt}z+TEg?V2(}7|C(6Wclh-K*0v*JrN`^RFWhcrD3y&FBkyCV
z8U`|>Hm~~HQ3eY84X+#?iL|%3WzHNSn+H9jAcEJAVjffK9W0qE>jV5NRMJ?zrE+n5
zpVNM42!cf+3$f5dXwz$a*?on=uo*t;acj0y*(jMgXle*;7DJhYJT1<M8T{5OlB#32
zL)|I1S#BZGwA*=ARJPnrXH+fvOZjm@9AKTH(j__<Ik??^5s$ShVj+LyXiUz9A=ZY_
z!70$VkJVX>gVZ@*%I&)0UbUJuYD2jyrs6<oT6-T4{>UrBkI+8*{Dr>)?=y7;8^G+5
z>#2l?Z#-_aqG+xJgY3`-?(yH_c6{8~i_P+v)q0O>-IC>RW{bfA+xhvYzdYAR+_*LS
z9fFeWp9U?rReCPv*vY!%b2o*=-3<z-{uHgays+OtpZH_#v3rdbK3UyV(Nx2E$p&Tk
z;=>UQejh95`B>X#O-w`;d4I6sn=#t><5M>F^RPk0eU>Wm`eZI&mdNc}D0tbLR_xDH
zL)Lw8%<_z@y$I|qDdsGB(Ldk4h0kNak;d3e))RGg{6!WkETGr>4n>3>l`D^A%D$1*
z!&#E14D`!s2GuE}!vFp|w|~Vsm5?lfM;Dh)T7`UD7m7%vpcpx&lc<pJV;7R#xR;WU
zVlD4MGQ*N4Iugji_06)pLTPBQyuz3d8Dm>Fa+3tD06pEB+GXWrhn#Hj;H8cG7c$^L
zDf|kQSxHcE)cEcFB)iSm%gAH_CqzBZpD8_^40$fVpDLI<lgvo&LZC?0*gJBY+}2KU
z9=e>-6Qq$W83*|yCmDPBvS1Z<aofRgTYKqmmZPUJ{Nml=`)}N#_zG1W$YjWIcIXwU
z1R4(nO_oHe#^@IDo<DL2Ze^$~+}pMFk5%ELdpUKb%ihILpsYzi+|-+R;B86tuWK0?
z5|mr<vVN@YeSrbe98$BmZB-+gr81wLM~ywa1;(W!C%D*c<qR=WYQpYT=Y@HliR_<R
zebv}<>Kj6kJ{A*BA`Jb6`i{Ko&O|R?5|H1Ug7Iy$^-F%vA5bo{yJOKfp=KnGO-Epm
zyay*AJ*#lsnuiyXa%BgumZuq~TuR#WJgbZ1hS2iL?c}+K!pAx8>tRBTMBMx}l(B~T
z89fjqqLv?nJ<69RdFdL1^Nli`PLBm=r#GoGzjza9g0>({grti22xw=WJ9F+>*bu{l
z)02f-^4;kSi!7=9SOkPq?IM=v6@O9WE3{^zl957~0Rbw7_dyeMTes`YXR>eebAQH-
zkhu=Ld(Bs)U$v%YB!na%@aon%<ZQy8d=<yO*y%#zQYbVbPhro!m^G`QQ~qXwN`^a{
zdq_piswohn0>iInu~-fhM2QCn_u2>4)0Ak5l5wzjc;1$#3hOWWoZ(ydT1_i*3>#T3
zwI|Y*Mv#n`qcrkHN_lq4_*uM~VJ!D*&MBsJg(=>eLn?*q?L2+pkbk!$S}a55iib7*
z+bBIfl?j<SsOQ&f(OZ!Jq?|Fa{=0I<z{d1nDQ9)+64rlMKDFJe>>ws>Mt7m>V$5kR
zytXsn(}-4B*qFdck$CnBLRrpMR-flF0fK>6hk9KlJnraEcPAHbFOM3~Lt^!meUT>j
zuix*yE6`bCDx5Zam6SjSeoAZOv7uqYHRg!6C*P%Wu-q-hJ0?|4d3}Mna=Jutb4%->
zxx+Jb{>$Fg>gUg$PY&BCM`G8XjBPoXhNq13B@FoD_;|hCA3b_=VEK#-M@B20pO-EL
z$UsRfY>&6+Bg=g6CDrMDKS>fvwm)UGEk9<6lT4(kc?-8^zGJEzAA(yq^?K|nXy4f~
zPN7Oqe}Y%=-g3OT7`=d@t`weQz=M?#_@hNU^tCh1@rMOrR;~fg;vtEXT|V^W6^GKw
zP*H}(7$bbHv8H(*7B}%D@;7A5ZXBd2w=HJ~ySG>C>>u-(Sr`&r*<9ODhj*rdL{f^Q
zHG$+w$G+a7Be{h*8Ae(~xn>u{>yC<*2(2h|#4F&}3X-;hzfp<98Fa5s@!s|Z=@Q;f
z#7U^o$ZVcz($H=q59UGMuj;9>(g}@(qu~b-60K7RvWbWx=oNK9gq~VI|LV%N;v>VU
z(e(NhAkT1Q=b`Gm(+d3q>({(RAI}HG!De4VRU;%`Siw+10^yz$uD9z9Q_`$cBlOj7
zUju_ciydR4(j`2=+r+wtJvQM!*u@)jJ~EUUw)nui;K(eyl%{>PUTnIJwD%q1tEPn-
zZoy&dryd41OW#**fo|2>7oQShQ0$XslGATigNjoJ3Ic6X(lPK79H9gmjq&mQv#KQ7
z(&=%CZLrM#ust;r_Bs9Cux!x5x&gsdpt#z^uI*HvhB}M#d=KN1|4Ct?2JD;kRqDKn
zVUjDQpl_4u`%Qn;9p@6WI71zT)uQa0(4rN;t-I3vki9q8u}KY5W!hEry0;@As&N(S
zB?_C((aYs>G&_=O$}99xaiZxQ+u&>ytC{|AN#;Sg-D{^OqI|@kJ6UkF+!(#x5r+~c
zkn!gGR8iz{;SoI+gzCFcy7KD02^U6&tw+cZk<=(}yO-o*U!u%+9sW3z3XV{r@W^s+
zq|BciIug^os|~a#{F}6)uVskkoG*oBO&P@J1<6SmZ7NCqL>NIA_YFd0Vr~O%-6Pgt
zs$v<bG2pK0xsj$qS-K28!_yJ1%^De<aJQ`mtv2VBLn^BXS>@-7{DhPUT2q&moQPPy
zuefbE%e`W(M`_Qx#MFrA5sPW!9RkaElhYqAdIUsHYsOl#yX@lFtKzhhe{6Hy?nFr4
zurYw`SfSO2g*|Gq-EpCK2&kb*ukSE-+VyRY+`pSIqA0+Qv-mXfO+(7~urY}nD670l
zZd#KBy(qDB-=l}hxYgekF4K+Zw4{h3&mf;3Nzt)zTi1sZ4Bh_zTtR4y^v+cBCVAQ?
zwaXYTnq!zQY$XiilGG*gdz)A>Ph+e^6L#%PEdu*3Hq_xH!Ki-1*W0v@=5M>gD=AfB
zJ~!vjE7L(&7SA_I)Qn%Ha7Rtq%fv6~$MqdgkPi3MqK`Qx%f(OEaO-}v9XQVul*(T|
zsNbd@Mz$p>9cCmgohXP@j~{aV#xhD5MX(nX``W;P$rYX@4gNc(w))p<dHJkRs2mWO
zr`)BAZF-wrJj(y6B$u<M_Pn$faUjpasO@77minq%IE!4Xw9$~x_quX>R43+&$X$Fx
zEWH_5acAkw&&K2TMfwXQ`xbd$6vltVCb{&EzcrhRVuKT2H)>?GvN|<figZ+4&2kgf
zr`X7_&sDQ%i}~<#BURN_X}`kFfllL++s{MnSknZEakW`9O#W3gZ{F=Hv4jn$8g)ZN
za<ZCzK_hnjO$Zi8d!VK}KXrPv?5nIyS-OsEb@f1BUJ=)55d=JvBlO-+wEtP1%Yc64
z7$WlI26NU-F>D!aTG*&P3Kt{X8_2;&kKy3-Du+AtR6=O<os@&$4zvV|ZPC?xyV8;e
zzw2qPZIVJtUfmpzXd@JvjM=%l?>jeWc6Vzc8&CPAmw#wN*G!7!&PL?nRhHyc26s^<
zs6mcrq~g6Lel3s%KgZJoshO`SU(sjU+ccm!@zEZa{1&u^5;FAFI?bTt3(-Nqg276K
zuJ2d5C(~7M&vx5}?}fbg3l(5&|5{<LYbwaOV#evqcr@s>>Fsjxj)_=DAA0qsF2R4J
zBW;MHDM7baSOs3BD8zQBD5M}+|0YgAB)EQlX)XDI&i+e7BUA-fpw|zabDlPX$($TA
z{pNMv_ouV+?fAzDMt)v0OB$g~HbwUI%-$2oQ*3A&ohM4tJH6lP<D1?qhTjz9s`pJ^
zbWNL`8*#p-a2v|r4~(Q>z1=ELC8)5F+Kjh8uJ<*FSyK@$+ONgkU3Ns7Onqo`P9!<q
zDyXK&C5tGsVjB|;4_uVfJ|rZ{$!<ucLbn*}xDAb8?8t!1FFRC`JZdx^C8zkb>Boz7
zIEvyfX{!G10yw*O2g&Bdt1G!#FRV!Rb9e048<IeHS}<0YovS~!--KjNx#bVP3!Gmg
zwp#d8ajYXlKbo|&<n;4x_){}kGO`PMK+jW}Xa$Nb@0(Rg`IXpHwgx>mC33wE5|E4H
zHVOGV(jdy;n+AmIBzAccrs&Gn;E|ZUG04vUDAX;}y4Fgo>bp`Dzs)!4QN4?^A;B$k
zk%(|DO<~z7_G5f1`P@Zyj)C7`@~-i0zEfbkZ^iNU#)$BlL2n_3H$Q~^7E4@JUI*8a
zTwa<w^a@ut(L!6|rv&kV*&EM?N2FKD0nwsM_;R(ZOd_ul?XJedJ?2plA6N_W%*n$1
zSh=Yk>(<|)xSH;3sFz3|l6vy*E=GIs9BiacPdL+w>YIOnq6Qb(;wOG?+gM3$vMc{!
z9WUr}!g;5y4@)QM1Nj>GDiA{5{S#cyBBq!TeJ_)qx%?@rG;eGbt?tP57HigFF|n9~
zS?bzbFGH$VQAJCk)XgKGetz?M(_-wgF6WNb#2JoZ(R_;v=DQ}YBR6~5ikCtGt4d)y
zG1e8A#gS5d`Dr~5OmN*t(GrhS$;f^0a`B|y|D-uFGyI$8M9;*=`d?{I!)g-N>;KT4
z0tQ!yPF_=eAd!}LuwLA%pHo;^l<?1X-(5bACNd4FMHzG&!^0SuveABELlj0_hT(pw
zb$6NfbP>TyJNq#5fp+;_G`jkeUxM(UKa?z*o{KAPo5vJH6%&y`%v=alIs-h^+Aeq^
z@cIm1*Lxe5=&>f;zcuIw&0$Gx2-G0m*%lAD^Mqc(6jPb34EuHf3-+tceFbf%6G!UU
zhzxB@yK|r;1M$eGR(&TZLh^a;hpXoMlXJ?()zgYaTE~UZui-TVH@+PdmY$A6bDYGC
zA>T-fYcP6SSxtvUp|l(WKiFb(9Ub?LHKMB-u{i4Fi(;veAtrz=wt-T^>jr5Mo&KAv
zw@7M8DJCu+{IGCfOfoV=wuc*o8?;BdZFP|8e0DbL`ef9xV9go3JB`U{T780RAmP}W
zG}hP;{n2hml<HkB#)Gb!pjgfDIYQA1JNEFFQCV9#-r6;|9d^-PR_TJi@&|EwBg#yb
z3v?MMN9|V-sw_K0MGTMLQ`??XvBNl!MG?J5Vd>}Bes2q(Qp+Sa6Jje1ORFv^`jkjh
zN10wL2Zn7^?khfE&Pwlpb?e8qZZZpD_(fn#nphDa^!c&1y@sPFfuuJ|F+}|xPsCY(
zS?!TW1y(f$>;pD8TB{jG+-gEnHbu~3cG>SQwl4dsxV_zUA-;p`vzS5!2xoSG=J2~#
z?UE|xl!L84#Inyy=v6<4lXznsl&br|xCt*gLcWxlEj1!jM|fKYB=oegj!n!t`~TF8
zdFN}_ytNR6&|e1XO8GVuhBbwJ(>W!JVS64868jjW_Ax3c;wCOGdN9FnC76*%k%Oki
zh^aT8reV}{utH9GHBI^G{1EnVB;ZD{jraO<(W&!$U+jHo!Ca4y=Spol?N`!Q18!NH
zb}epM>kl-S0>sc40#5kmzJVO?np0L;713rWDz^~kQ`xpE7wvMt6C5i=*Fw~5xW)WL
zp=liK(9KAWb4BtsT<I&}0$4Ayf9mkt7`u&|$p5a~={_v_!=?yX=7?w2Y$!OJ(?U}V
zNfE*Q5W%Uia$#VC=eUqzVV`a80Iez(Mg=Bnn6gZw0ribnE7suA2ypZi2DD9{+!out
zGC-MF?58F8Xn>->kiLT&TqELp+92{LZT}hD)?RB_;7{5e$r@WCVqW;lc-xKUr}r|b
zKavH{q81|uc5KHBxO*u-VG?sRkI=#g#|1}b@tz851n9Wz!>0M+6{sw6Fz`dy8bdI6
zP5F1ETGbHnQZ*}Awajdws2<<VU_E$^dTBdP7MVRX#&jO$GrPdO&!3r2qOB-7HrDxr
zEam*TpUPe#uhlDYF*UQdL6J_4>XjEVODE}FmHJq%5T<~B%C(RE#qAYOl5%!H(izGZ
z{(hr=?sgmzh3hrZC-JDErZ_Q;`#@E}<Z{bbsjt8gmfQs0EQ}Gc0G~I1QdDEg*~YcD
zZ|Q>=)!!clDO{z8sOwg|(uJ<`Y=Bi;zHesew1Mjhdhp=wKArn96Ypkr-x8P99J_xt
zfnur=t%F_ewp_vYU5Kgv39Rc|+IsN~8m(k(NxOGjy7Y(s#h&ww@>i?$CatJHsLW~T
z6n+!8OjCDRQ~FtkZ}iw_^e5x%YNK0nXYdQmEAnb6!IC99!vnkD#!gW-kuA!z^zj^@
zB%|VG+|zsQ;=NpmO?-yvl4Z2tO28pF>L`UX7!GigeTYn+{%We7A)`r6E!<%LV>7YB
z?M6%63wiB^c_mZmPQ80)b#!&Mzr7^^!=1gnPgS}vh`Gm&jsmv7@2by_UThGvzJ^VP
z5#KIU&4}U$q0OKpbKGylTei137?G%nY3{JLG?z9H-WwS5_DD7u<DD{-*aVzwG|8kW
zxc*z?`Is!B4K&Q=i)zTs;r<%#VG4K&c;B1Q0)1lQ!t-nJb}b<_X--wH``R@VtgqUl
zX48H{SUp6rxtbP@(XnTR^NpU+-=MiY{3mc=V)`F&`2S_`7?}UX;=Nq{H!PmKsK7r-
zJQjB5{|<Bx7hPo)jho+{gkqQXwH-SwH0A_{!iD+}+_#aZQe_7RQ_&Etg$8H)5d6$0
zYJ|dCH<=0{l>U=VA8Xr_9B%Ms>A@gAQlV&vI=#d`b8C{((|m79mqO)!MbHzJ{=_DQ
ztL?RYxXtMq;QbVf|KQTjqx~?;`*fPgJ7Cr7@^mobkAKATfVBR!UVssm!4=zcj!wIL
zIL3ux>#`<mJH%$Ed|wuUPOD=k-@YjCRZ@kq?gnBJF_z0^k}ha8=^EoAv5jU^$GPgH
zeO{kaSspZBNfPc<JJj(=dt8Sy^>8{S?5R6`y4t+<PV88(?Y_2S7!h~MDWkd`1FSDm
zf(l=Qo*y+3Ys*($99(q@_xC?>akUjVwTlglz$8-T5Qzy@OS1ytAT4zKzLOM46~>R8
ztH?i}lFfR7yQZWMOIhj^I7_4sr$)@4H`uPGuTvm01&w`j_aV8q4AYUs#&t#M5VgO)
zA{A2f=#33S9M@~02;6pUad=2yeEV6_U-6Dm`rRf^J@9qhBJR@`_ua8WtS<9njPRPn
zp<Jbv<H+S{Rt4@PW=Tobuy&dZ^R0$~j_#$QkRmN72z<VJ00*YxP_>j3i#WD5#g`mW
ztn$>No@KP06boiU@1!>i@oi3>6#fy~EJf>55Ak<j^?@#d@)vvMD##ThZe=&<biQYp
ztPiga1XT5eeM#}<PHCEBv@NBVhI1TWOEY5Mn}*?#<*XZ&;JkBPLJ7+m@A_t}+8e<V
zv&DKpClbMQp|E4x3a3gmgde7?G+KXchY?x;eBQj>NyuSL7#1B~-hw#$XFsdcaBx-k
ztkn_5x{u!7vwCw+u@6qwSBpkplLHq*zCh7miGJL53N6Z}qKu*^@5bl9ftfRA^cni+
zF~)`6f?xMw(&W#HrtmY0mPFy`xe#cQlV(mcmNwkZuQpL@#R8)T$pJG96lv=^S*qO4
z-?n|EfY%yMkgEBuMsV0Pu?Y7brZsPR$BEamA#tXFO@~0ogJO1A|CVVmQ?o&yv#wq0
zP>eX3$no{ZIX+k!C$Da7czKg33_Y?C=u5vMw3J!xO3mR6L%bV~g@coZOhaajAMeWW
zWJ*sPjI9ftEF~CDLZ%C?Rlan_C`gvuG`v0j^wpo6FjBsO_PPqeo+0`w1QS&Idl={W
zn-1KdYObd&?Rh=Os=%oX{H4Yv#x?rev&}}y^ri?ccC9lBVD=>d%k3NLJx;i+K4~iB
zyoYZ|vcn9)^!oJb+lqeK7ANy%o&`KDtSy<gk&~*N@Ne8@c_dMe=8m&GLCv++C)J%F
zEW(`A>7rMEd{>rgSVI*vo3%UaJM1G2F^GWP%Krqd{y4%1gC5qoCQa9|gnAHMThl_h
z&~c?MW1fc8_u!&>Xw7RJZhK=Rc;NwTfkSU+a@RER=CGn*n&H)N4wo~qd6@!;)VPct
z;R>(%^xBAL-fs&CMic+~;9{^C$_P!Vyqod?n9l@r(WFiy6sRerkwr_#DuQ9h8a{nd
zlqmeN5jxX#Dc@FdMPG#+Qcc{&C?SNd6sD>ESM(z1A`C#AmNSubpVthdSM@``Z$Hs}
z7fo~uNDeMHXNIL_gFcxLRjco#Gv_M1{!YFof*rCSk?Lttm22tU)v^pM@#R%d24Fc;
zXBuI$NqVPdrWb(kW)kZ$;@jo%ccF8uz`uuE|MV)SeKa{Uxzx>M?<{5MAhZsJK90?D
z!Za0462~-tb0S8|w}g|AqB1@^QX7d03h-!~8qS{Uf(ug<a^6TN4KZS4etQXdwFD)8
zXg3i_2L;|Yh7d%<IMAaVXZvv|{}emU%b3~OhHojaTF>pUx>2la=N<|Ed@NO01n&KF
z%DE^#Yofc1)B`1oftbYD)7Z0&7qQsw@kzNzjG-i7W0mR?u2;(Nr-dvI)u9xGBhEp&
zsAuO82&e$p?f!a~xl4LhTE@6g`k0Ud?CYw#to_CR?js|PZbd;3?@z=p_g8m(1<tgh
zk58(F9H|IrSx#n&1x{M&OZNh(fbjXo%c(2k^L*|5!T-<Fj3d&!f_K1tdDOL!Yh(l4
zsd{GV+u_ucmx<^%8C>fa*6wUhV&m#tu~=W=O@`^EiffoQIBr|?PMOG0U!kb9fHB`p
zHB9M>w{{qhffS`Tr2`Ubunuv2S8L015@Ta7T@HPPuG<Y3XauM3p7K)a-K7%N)$vm$
zZt>?*hMz#^(yu)ALhty3SaI*0y^`?&+4`*hx~}F4Xt0zQ_-|hZkhFISoc|KKz~#T>
z@V$FEkRodwb2XAJ1}q*k<Hr}->kVAaWJ2m>?;m?*NDpTzYk*ga(k~0H@P0lXb>Wi&
z75>6rmJ7@wZ3b`4qWqF_LJ1zJDF$v6;Y7!EVO<XK*OilbnN#BE*$(>kzPhGL?jp$A
z2k`j{OcpM7_G&c3d|xvpMN9p{a=dGNXnf+k6O(JL8+)4oJ#r;v5fgSVcpvEhvH#V6
z{Cv#SeN-kOZ2o!_Za#Ma$Z}sg4SeA52m9x<ny<2@uMwGgKdq4hAO3p?cnuEP?-IEa
zk1C(0me%W7wI4WYY8gPcfic{c!0=p?9CD6$$=PqE7^z%`$t-&9oiAoS66$+d*{;j&
zDj;M&I4$%`l(Dwuyo|I{W5lp_hHB*(LhsAcq>U}}YdST@J-gozdHO-WQnqq`)#-E6
zw=*#?A|l{#3oJIJ0$n_tXlpLw{!$`q0aIW|U1TDg`G)n_fWcW}J;n2@fI`hI?v2>u
z^QdMf2WRbOk4{y3Q$B;4-fLjKTWb^Xx$)DLV3$nJmVUhdK7HTKrFh?yA8wo4R(>0o
z#6OjM@Hrsb|1dlT?!xs&+{}KQ>mzL80aN4AKpLUMzff7)%AN@fFTO!$(9FMAtksG$
zigxmgK?^XZ3K>*uIv&VuL4I0<L|NlBK2ZZLqM2E@-f%{(6snDj2b1;>&V=-r;ikhL
z@{EV17{7J9e=D<mmV=(#UwSz`>7TQ`6(1ZVj1XiT@OvdpSLTOPVLGHvl=JNUOm5iV
zer3eghgHte<zG&Z-!+iV?X|ccue_f2jNMkktbM&jOb)7XVR8%ZyNA0}n;jeau6#qz
zOJsC9S=wOh=>-p`%DExqRX;@R8jg??F~?)O^WtCT1sK`ZytpNg!whFdKb7Bqex(@E
z*Ib{mmNOa$<r;-qAua$J3+F*Z5BH-|F9<^|$q>9F2Y-gtZqG8?zPB;k9x+NjKvgE>
zgU;B(>5X`Zbe^~scZ|Z=NAYAoaf&SaykGHK(zXk!l_OwCBS6k|jqzn%TffveJ@?O_
znwF$O@-l!aX*^xQED~PN<*Sd<XccIl<FW4UAw@-Cllsl<8h)FnA*fK3H^>IeKBNfu
zHD0oZ^brNQg`oRzvU%`w1bTaq;DuQ`5MrFoAnvZ}_7NGrHY~_-yZwXYKpG|<EzQV_
zCUAavRl6ot39f#Mb_pi&@80^;ifn5j*?lV0p|LpBvLq}}(tCzh;N!smju58ki|ehM
zHt@iQv)bcM^gz)|P6qy3=ZjllZc7U}*_?|Y_VcSzNC)Tj%+Mr;D|)+2kq1(a!5o;R
zs4k6%8>FpFB(xd@uFM&L_%W{|E)nFMXf!!Cx!e?`ebwsDCp2mOq$*p7?&<D1xcf5~
zI+(kU@g%by{9ds2lWs`yM*(XWcML!kY-zTCl`#%Rl3QUnxJ0AULezYo8D@2v{9)gu
z;oRvxpH;&spuab=_&cV4HFlY(r_}Q4(^qJLww<e1dmqSVHty;lZ)rDuuib%k82bU|
zEu|x*!Rgwyvw(-`j&MiHnXEji!{8lvT5o^Q?q0srabZBrVlTZv`Ws-DnoL9`GJy@j
z5_u8RW>(YqP;E9#-`?n|8bGwFyBO-hW4+k};{15fa<*Ocq{oIdw%y?5;B<aS@-o9s
zFvyg{r3lP=yX+=s<5~?*M|T(XcOofC8`*(fuqR>+o#bkl?N4wEpTQ5dL<If1^kMvm
zQ#7=P`yV}W?VJ$Ahw;|8=!bc6hCQ6OkD)$qL$M^}eTwwLt0TbwkzeV?dvyUD3MZk#
zHL}=p^pFUw47-;0$6p<tf^i0#7at<-ul5T{tP3%it#bDjKTBX&1UGq|7IkTT;f%y7
zv9^G;$YdhAT~=2r=uUXu?A*aO0DBEo;CUkZ*yKvOvm`-*HP&pYM0}VgqDGm<hXu93
zGO(KP@!M>2jeP^tQImr)Fa?o0X@JZI6+cEi5z7PX2>6N@v97gWw=Zn8Mc6-;-f|n8
zbG42=Xs&@P68-{wyA-Gr<EMB$i#Y3lz-6cuz)&TBNd{(Pv3))7U`*``rkOFIj+rHP
zqW0yH4$mGHd6x{yL?mE)Yc1QQ0N>pP8MZi}K15x_L34tWfXTaPR{+}F%=UbqZXs=n
zLTAPE%#njkJ)L`<k^?gPFXVc$5&r)DKk@4?F%K9Gf8p1&Hw2&CRpjM^FJk=FVY={-
zmzw~jb4Q%<dOlSE`uYNAU>K8j?PO}x_WKjj|Mk+eDwSzF07s|bJ}Vrads9*AS6*+=
zu-J@n*h?M<(OFo9uq;VU_v8}7Y+Ldky3dUTI9X0)hr&FWGKE>V#6en?rwpM9G_^VX
zIn#rusG&c6rjJJWWJNU-Yl(aQ0$Tugx%vJE(Er`Hv8SE>7oqZZ-^_W_*989OAx3pt
zbDe-U0q|W`s=M}j`uf2P09~C!U__}oA8dgQbYJoG9a04Ts?>71_4qtKP3ygmp4r@7
z1a`0%H8Hz78htwPK9B^~W{#P0=@N26Jw`En@l{B2{AJMAMmh3={rR_qLpFl?bBgPL
zyzRFIhM=;Fd|(E(>JqIM)72Gc+}1gnY(^^^i=)bQ?G}RCZ4tvuIh|`x*N#056lMUl
zda%rio;m=)?@9-}j~idvS|1P0wDz=iV;hCDL{8sVAMol*MQ;1TCOYe=D`W^jTK)XW
ze0DSbwz(b=C$$%(X)V_BUY!{_VNcGBrw>bjCZ&yIV8>$PXd0Mz9L?jLAjLWesUa?N
z`OK30f^21FEMsq2NQg_U4|jeK`O>y&)!O!!?vMy!nKtB0I16l1`#BAhngN0XfaTwQ
z%rNENVL{mH-l%nzuMoG(=E#g*u&EeQvp#9ZGMf{W$WljK7-kfCc#}p1Y_$n10H{Kf
zGC7?F!Ve3062DaObc++}w{n0*zf7ey<!etLaV5#g13SK4X}Sl?K1N1&x@d4t1SNR`
z_f@SH`6Jia`x931$k(w*8=j>uV?bq)-+Fzy+c~~=EIRvqM;Y~#tcrKSOLJu4=;2Tf
zn9>Hd!`1<A-EgBwRR-%#5qh=D@%Zp*jcA)07H={mib!wdSS7Y$UJr@;Zk3X$y+=SU
zF$x&uU8kPVF)-2=co^%-Z{-tj&n*;b!IwnTr?pyW$xggsvY}USev%+bemIocIjN}V
zPwU-X!A^Y<g!KN*R3pJY0UvhvSIAc}CE(0cuC<>6OBBs!A!a$r_d;K@sl$Tq%O5Z!
zzz;jnB=+<Q%?*N{A|$j_R%}F8Bz!75+6xp~0YdtsmAxNTH_iY%#oV>dpe7OOp<Ko~
zj0-;{B|<sMaLMJ=d<aGHaG0h9+U1mHq8`!PpQSDPNOvn`KqbOc+}1%kYYT+)_ncz6
z3p|;LS3Li@2$b=H?;t3Pyk1E8iQDPb39%$PG{BP+{Ua<p?lewog(n@or=ykUWH$R`
zMm+QrY(P!{;nep4RIHtiz3U_^k3`kVca@2=^;DWVq(=$!wp$!_oM#^U!}oK>wywB#
zoXc9-PECr~;znM(Gk}~A-rPAn`Ec@?%M+XS=_5qqs`j<F{3--kOE%gB3&h|9w)^ft
z<;zPEk3Hi*6o5YRR9>9kX2z)4bz2{sNVG)@>_+;8Gyu(Jm@`9%hBOu)2sv*5<dLqV
zD{Hnr?`a+QtGuO>o^43$`Rwfc^a86i)8!nFm{WQ@M&s3$7dSuWkhaE8kOr$579ryy
zERmQ#qCuiQo$&lqzus&(k$5b96+ApEc2b4tlOzC|2Y^59sr&SR9wJV~i)1&}V?LtU
zRqw3$D?`FL+MaF~3FmuH2DnJNnH4Z2fVM=iMcrkFXdMxQKb1&Rse_3>|MBq0ruB+A
z1av#Xb2Mh1mWJjc&+J5AEY6vN3_aVV7kRHzhi>^NYtq|iy{{@TAFT&XIkpw*(@1)O
z+edpRtwz9f@knc#WqP(r+~pD3QhtO1PX27+h{$BJICx8C@gms^JUjRtZVsx|8`SvS
zM80_5?_NyqE;s1#&h>m+0N-EG*Hr^xDWb1aX~o$A$jL#)7S+tq1qjmYv!`Na8_`k1
zZ@!q9u2fvbb0RuDlLJD%dPZBet|7IZ6{=<R<$u6Q^pmG^a+aGQJJle3!J%9p02(mv
z^AVSK$&)~Xh`KGiJ7*nZ1Mv0c?N^(1)QjHV{uVnP_ATd%GJt|>75SrFA>Kgq1x8g4
z9btIj#QKtOGp|I5&Gb4)?-vVzW{xW%?!3>Fz<e(`0A2gF?Dp+{d{=)%)L(YhKh2N7
zO)UVAt~CA<^5TG{Lh>&|<eylkUib+3?P6XGn}6ZenAi1cRLqq|ME6Ux{#WRunTuBB
z)|7)vQ87&6fbYz7#eE7g{^PUF%|XEpQXk;3v)zDb&SfZC+6FGUp7@VNB8=Pg1zH$t
zSq{@s-_`G|nrxZu3UH=yse^HJvM}DO9v*#o>-gId5IxuY_m=|};xEYb0{?$ob>JE@
z`HBC9O#jOWv`Le;{ZY&7efnab{>wHj0&h>Z6VDVe3%C*`0C*HM+iq6xdBzqy^xkNk
z7@-Pq?=3Ozd6SsUJ-p9<`u3=gNNLL8Rf~kh$%5sF<B94f-=M?Q&!r0T@%Gjp1>n#-
z=n@9fp?8<ZSgP?W-o6K8TdL~Qxou8_(>o4bgPN$aTy9XB_=w?o{wLw)FmevSv~6kK
zTgJ&xGhtV?us*Fz)#nzn@V}rE1=D@?Cy0M4*A*SA?Je}cxr`Jpsf<v7+7=U~0&W`G
z!>9jcGtV@bFDpDmmZLQJjB2zMNiU)zYBHXtwqsV$`J2A08|6m^7=VoSr=4loJq*r=
z>A}%@45c4I(+-@AeMo^RhghBO5OBtYQ#vu}mpq~qyT$<T6H$Y8G6?3DP3@!bN5Ej5
zt-m`)ROq0aV6%g-4u#!5tSUBTX$BmZfGc4lV-8`L+wr0O>=y7|SqEhS9;WUtc^PvD
zcPb<Ehr<?33t(+7t3O8d7S6b+=ahs6seK-X%WqRp$TsMz-vVKJ=RLLBemS8UCSk{h
zdVVvjg84-iL~zJE=|z_PB%@sjY$ET}%OG*y@6CCvZ`t*y$+aYUNf_AGYiqwg6tpCL
zX+7Hb+U!DHyV2fe<sTiT8&y&BYY~|<Q}@M=DDn3E+2uzwCm-9(7w&8L+b$Q2-k+p>
zVv)k5G0UZqRuYL90vk3XwNzhGz8yfhif;=Gi4vl3P-wZD!FNE?l5X`%Vp(!nl*C5V
zScJTS5$9B{cwN!*2uYrCRiKjZNuB~P5!(A<)=DJq0L!Yb{z+nyYAc+c#(lAHd>d(Z
z{5;V?D}X=fx)wjQo64<i?lq>Df1=cj$8UBMo#_hGsafojdi5IsB-9n&3WfUh^0f2w
z8^8BIMe@n7TjMiN$hIa1kO9SMKIRhQBX;6h;x`Fvy7T9Xs`R$EE&)QN(EB%R*PNV4
zFxi~Qa^RA|c(LqPsDKlHq;I8sKTf*u3sk3j!WoLM8hYOoB{YliyEv-yP6h(P741@o
zXBc^SUZ=>LOXPbw4O*FQD=1-cNfw%cT{(Yj2`lS&h|iYJt~omUuTM8fzHkNVYayaJ
zL2`h!%YK4o3^AByh0^tGrktbiSM#hw3P&1uzF83DOd}o*?rG!vjTANmILTnoz2LaH
zgwN#x7kkUqEIzqsPNI~@EN;{dI^&#8!?f^M84UAqmdGXTBZIKmY-*9|ORG3iC#V}z
zt1afIs~0g@>^)*=f8nI7y=#R!n}-WOVY}r|m&~#B_Sbk1lED;f&u@0ZjqPT&pQFUn
z#Bb79xTB|NO70B6?0~=NgG>#k-_Dz*wU3ykBCqi>x&pCRGz$@II#`tJZl<bD6JQ@s
zM%@Vjz3&pnx~9_;@Jl&P9aqrpM8#G2Ok$1nrXUC`SRXY<;ToKhoS&=)CNw`<1nz}1
zVX_Gou>K|a<yr<z_^NCAO(uQs9`7%XC(Qw;TpO4N1Jd5G>ci8_0j8g9nT7Fm1L_OB
zdBjid{nmgFjD=)fi|PloYtKVCC!udTOydX3y=SwmRR+6&<Bi7oa1lqf2XGb*s^5pO
z-r_>PjKj1QM<eY_hMx4hMI2!F!Jho};!)u;zAad{800K6BgDuBDAkXieTSAq0M;*=
z@B;Fk=6wH97*Dx-5&KqwYADIDP%8tnSQ^BJiGTrOIEfJ1{a_BPzWHfyl}ZC(vtUOA
z5{RBso&nqTpANJadHGKsBYhgsvj3ts{*t<Xf5x)JThuD}AT-J#p~#6{89N6M!wUEF
zygYzo?=LIhMLYxhc<|DiQao1c)OLtrnx@yxK(k_KX6k7k90WYcRQcg0?O=T7Rkdc5
zS)yJYy%jT>xmo+w;_{0C0X%ftZ4YPn?N9(&^NyjFDbn*2%!2p;Yt<09?%pYZ`VNhF
z0&tsUBM6>H|K?vadEt+!D=edm>d9`xp6_~jb-V$dr*$S$R3=0GFUrXo|0&pZyAu&`
z{dJjA{>27`S@U~vrYsY@m`<Ah9K1KD86FR;GkF~)|B2FL`S*Aa24)tP|0>=Ctll59
z&Vme{RXKsd@|Qe7LJf^pl?bq&RA3E~sduND3_1mQc_UN4jz$swcE+DA<k!SSg52bT
z_>6aRa%*?%cH0{i?GMDtcCPOCzPj+4WVLB7n$~@lghuD_h~wqCiNov(q!^^Z^WzJ)
zQ)^A{roNIO_IY}p)`t1s>jB17#O<T4rb|UQ0t$*}bVgd&=T+@VGHHA%;<^xrUA*v&
zr`?7M9c>r+*E?|2myi3K?X+Oz6bYkrOt0tLHGHwxlBzi2`mdaB_D5HH2YKPYOC~e!
zdyiim3r;W`iJM8y@}12t-<$xEDc|408wWjq&*FWw8A-G2@(%Py^3?XX^5xrBr=0fr
z7)ovx+-UHNH}t1Jn<v!989guPcf~|`r{w$IMh9oI#@A^?Q%#=8HUr3{o}4yz=z$Hd
zomD##R&ovbE;6Cgm?YuO`NCrFeK|{4-q{Gy*_@E91Q<=sy-D-fkO!kScEMS_3hI+x
ziS-d|_G_74+tBL3Rknd*gD{N;e|-%TQzVT3^_j(=vvm2Lm4h&4#ia0-#g`TWh&HlD
z%qFZKN$noiJdV;yqz%4G0J*Z0jU<sB7{!4c@|>4;_54Xb@6(Pzjc}G+Xg7+mbm$Z1
zbE-bQ@OWIVo}J`1g0SpFugtgK1y2aB86p#;9%tBRI2O8;G>Wx)CZE)V@L>idva{_y
z&`GLyLqoQcNFGjd!touYW8$t`Jtn@mT9B56Jm-<)SWJOqGZ@_nN;3)b%JNH`aDv(&
z-qP~iKB?v7kkY|7&yI;!`-Z6{^vgthz`NC*nr55z@$~O-6r+$YQ&+?fy&l;ZPRjhu
zfNy#P>KV<&ggZBa&z>*#i~M>a*E;1I;9ZbPnS4O%QkD#cm2l&vpKakcABK>q3}Kio
zzq9QujH(S&m7p8?snzLb)W4HduMl{S=7@L;(<H)_Bv)mM8q%NQ58CGnt?XN`apbGt
zI{Q6%^uYTgdQrQJgmaXVZKMyOB#Dhr+uy*$J!Vb0E!Jvavn~}>VOfw<EIr;V2C?BS
zoA66Z9J`n>O3O=M#p>Dg_A(k6X!(6WFUvwdJ5=4P<TjzLyfSEyPnBnpKSBD|L3`40
zbL2RDM_K@kcGyOE$Hn|J5|h7~@`lD@*YV7cY$0B?j5C!T+SAy%fK%@3ojV>iOKEKQ
zjH-mPNePpkq&CsK20!=Nz*)bN*5U?c0a8T#K}SJr`|Mthbi=7?ZODp!sQA?9!1&3*
zk4!b~X+#FtQrIr{R*l{?lc~!~MVfXU8G&hEq(C)~N%g~YA%h=f)SE+Ns(*f)*!|6F
zU9_}{t<XfJ=fIuiTEe;|t#A>>$M&RddMdZmZ1JG9^OHjB>-!rLcUbpdi*)rS)jY&%
za4uyy&RG*wBc~j_J*$zQ%jGKW9;?$d6XXhpscmRGMam@RQ)1ySTfwOlIP$BKWkbz!
zL-H;uhmes)2PZ2fRTkxQBx)VzPJv%}^{W^*C>pr2hfM2~aMNg2qgJ75@te#g+4CN2
zJERUJbP5f=QXpu#nV}Xj-qcPMf_T<)b!$w=@{Ojr7P_)eE0j)hTZY{<Xw>|e14DSo
zO)BL;RJ?W+<@rVe$R~cS8@wr*ckt`D0NFmFdr!$-!nS6T&^%|*_}aP(nacN!ZTs#U
z-GqW4iIN{@F)9DXpP==pFt6`(txf0Ln@*9m1;=q36yj8<pR+sK6FNxpOS)fWiQDWH
zId+-{m}neGB?CvxNe^fA-SgeBG?9|hkywi&#lF7#;`2;>F>!$giR0GDPaD@KXGV?t
zWJ<=#rrd_^5A)jAt0abPAplAvq+%?E@R!AHu3|venRWu>uK9a`kl>qxqwCS>xc9Kw
z1)JfYWg=qe^JVP8@~`%l9aBGP&LB1YoTgvrg5Fek8-vEBSTyd}o*op@k2y!9iL?I|
zWW62n&h>W^xZTpDZ0LP+xhTvN16$0ej)n;(>#u{2eE5E|4;8(iz_W-}BMFLY(Om*z
zlJA+#wO%v6Dd#LLsr%hn`5i^_y!wMZ4{?oIQ9@_#HR>%ThJ$@Uv4~;sl$=OgQjz~f
zM3Yx#QV2gQjY{k9kEb}Y)z<Eju1;>JY5LNk$sS6-l0Eerm9@V|9WJH5M(?X-I*%7t
zNpte!iMfW9kSg{B_^{+s1iucAvOb!EH58{K`teoFLWFYg^A3iTzKo^P64W~oG}OA=
z$EFB17QnMRCikGqtgsOy4*iHdu&{D(74d0|cB+>=Ko`R=Mb3j$`UjafH_Recmwz1n
z`VcJ1GGat@7GVuHb4%g3;XiNIC?RA&58aH$=k4CdqwtOYY@FERRLLe+jKpMy`hOUE
zr&v*<EnIWi#$L8<+qSWnZQHhO+qP}nw$0wRlg>^0oQLyTne|kuSzpcZ19dc+$79{O
zFdr7s6k+=i^2}mxIY&?I-RK07ymJm^&q=w~Ehwq8VsiVQa^7T5dX`HBhAlsH#Y9RE
zs{2LDNe7E&I2jgvp276WO423R1NZ9N;h*}4Dbdk5T7C2+WpH%a8$P=-!otE3m}IrS
z`FI*#eUduqVT2om3t=1|c!}q@S)O2+XZtK~xL7ks%6~(~H0`TqIQkk`z}k#(YKKXh
zqD7e!`-6RMscUCcBp9Yu)<OIB@P^dB&?q76b8vR<4OR%jdVI;~D6X<kjWOfwJt&j1
z&beQt`*TESvjf`sFIxHMj{rrGRnrwbRgZruKH-6csinO98mvlC(#9TJjv_nlX2{H5
z8oeCnpFnZQ02?c_*ojb)P5Ja1q#UO_Q1fqeoeJ2z#u-<ysc1Kxl}k<?lpy&5{M;Zk
z5f0G<D~%{uuPtjgoK;I`I?1}G4%5LknS&Odg+vj}IC$<i$6}}RKvR}%T@jYxO~EYr
z-j(G0ZBrM&yn)g`v?>54$BU^nSlKlfjVd)yj)iSD?<~U;4kLuDjuY7nC@3;$M6F*B
z=OJ@<(Y9$7%bX`+P5Jl}>KZq*M(LX62hEm(_ghmaQr4%E5z&Ok+zj?N84mAPlAMnC
zRW-8BAVIXvCo<zME*pjmv~$PVE>bdN{6hCzOD|=Z73CUgY=aNxL5>!586{?!f<@&?
zn@y7ygpZI;($t5)+@miakYxsR&CW6H^;&QG+iO{HAY&6fv46Z-hL1&G^Lw0<G{QP9
zd-_NgkL8y*x14Qy0LOnaa`>!wA{i_1u<4qQ2CPT*SW<ebS>qJnQ_kXn^d=s(1pflo
z(S<h$!lNiE5C+5A5v?jScIH+-2uQ(WmL)G}RmE-3s%ULAUq@>ziCp{Cq$oNu2&va5
z$s6<+<6J@GiLrBG1?I&#X7V^8y<J4psW{Pf)2*ms7aN#dKnS{gyEDm%kWa-h{M2OU
zbIw&WFxpwC?<zpfOb4axO~4p0(^O|xG(5H8YDm3#L{{N^vN?@uhG4-q$aWllGwiM^
zAM9V(4?kjWC%YW{0viE&d;K@2WMlfjGO_<_jc5Gdm=cR4fvA1{R&5C!EKv9$>g+6x
z)U*xp(tK%WM3br-J38O_$(!V+kXDt2B2dT$HE!&4=<Zw>IoHR%-K)#1J3kAi5h^N;
z_b1Cw$L<4-ooC5{5kQ(q8jr7ZSBC>FgBR^Xd<OB{@T|4i_W1?0e*r457htCc9n$AJ
znpe#O-u~~J+6ov%1aH7jS18+=&LZVMprmRJw>!r`AhxsZ_5J!7K!(iXxc2?1y|2B+
zqx*HofLs_$-!u3)JdN}vKt$JMkQk-jds^iy<<tFP6{ImW7n(K0pr0x2{R+4lj~!Bk
z9#X7nR>YC=N6tMiYBoI6B}qXlp0KA4=NFlZTYb~|*d5|E=L&#k9zZ{W!6e(C6Rim=
zUU=I^gz6O`mtJ^zki*aSd?@7TCj>qmLFR`8Vh5OyXVl;dJWCFuQ>17!QtWRNLQH=9
zrWXMUN<BM-K+;PCU<rJ)PDclnzQ7PFj0-1kK$K9;>sq6gNV=5>T78@0d%Z<nCwQHG
zvnORY8W{%`<gZCrG$92UP+Kpv5zw$!)Qex59o}W6a8|FyQXy8gvt>1kqtsCP46^(2
zaP%5aniDS>mW~(Z9d>f6!XVZhoSROH)MfJNT;XN1Hij&)k<q#si<bU}>5`;9QMUdS
zFpE%PWmu#?B90Iz@#GLy%G``B@^d`vd-VM#dsA*MLvOS<CWe=QAgwps-_LCaP{UDa
z3>bRDYuJt>Gr|kxh46d=FvNe}yLn$jPSUY+JZraVI!O_AcDYN5W`{3(ib<<8%%hm4
z3n*Ag4($8E|2c);Z*37dF}^y!)MWXmi5UoU;{>C3_p8}LQ;BS_L4otd#;5`SG=qx>
z2QT=Ci|bp8!=RsFr(X?()duJdpV8-|NqAN{3QQrECpYb~ig%$UfJ)~^=L5g4U3tF)
zb$%ox=hx4B_0MM=U~+$WzPC$Zxcq1|AoefBsVMZU08WtCqn#nGK=bDq9?C}rw!+h!
zrzpB%Q<Nvl5pPT;fRD!G?fyv1HzcqOpN4f=)J!Q1{qarX#^`9G^pr*HE+uqCr8Tb?
zH?i<?5y$&@5lt<$6LKl#3(ha4*xfH<m_m#LcT|qo#1hP4x3qjx-T65SbY1YmY{ki7
z*^j!BV8#j{t~g8uH|LGsgqxOR`k`i^l9XxUpgCzQyOyU$1qE<J8edTCsxK}6e{WJ?
zoI8+UCNc(GPJ0$ekSlBvW0VIB%Y{-E6WBt_r4T7NlFbR9H6*kXgBRxfa1kN${v}%B
zv>g~Ni%Z@1k{j$=xrlzK_%h0|Fa8m$N(hKc44Ja88v?Et!!`iSYLLmnr&c0hAJ$<4
zf{-ZI5#1<Mjcf_9(jhYRlf6{r#TI%ar5Ip0K{!pbDztSDn;`<;K_*QGlO7kIS>mbm
z<R&0Pw9T^|_;@b45*{yL!bBUs9gIn|hNk9MqM)fyht70DF@q93ntql)_G)q#I1js{
zXjStm+M*~rXe{OP(NM!67|%PpDpGV8xor><qq$W*BQTc?;LR#PIt>hg*s;)jqhf~9
zJO@FRR2=70l{8U7>hF@90AeExou_|XG}CKb7LcpK@aQ+##a0M#T_Up7S9%ghl;O&%
zw`e>^i+{{kelW@|2s&Uwe#?3J;As^sv!2>h3{xHNLuNvH?^vnIf1?T5cSm(92(q|)
z=yn^Zd6=r8<p2d#fZ0~b#Sm`3rsYRi?dm+sFS6OL&@3nj_1H>Av3le#Pw&H6;c6(3
zijq`ixX}y)@7`oq84)OMggfHR$Z6lkWN~5%zB+3lHuPS)fi*#$#cn37S-0VFy>Y}K
zV`3^RdEs2kF4a!er?%PD)NwQkyUpBw=i<CYBA-TTUT3r{$TABaIvD|QFOG*&RD{7o
zJ5;1QO?Zg1%!+#3EsHHo#k?5|x6H;YO-p1Sv@In&W-%>(C2luSqyq_EB3X%kZCGx-
za%!7#GMRsu-sCYmW908FCGxb;i$7+|dM!tkPP&J<-+adqzd(sam_Fe*RlnZMfKP)y
za-Sl<Hgxpg;>0cfg1l}_ct-98j{^0``&}*P_kFt0=kNUmWb((|n-SNG8?nH>O83Af
zqQOwPArhrp$)hQpC7*b<2gEEA6ydhhuEUmotw0~*-PztTz|?!a#21hRqZiz}<>~|h
zX0ow}=+#zSqCWSgxXnn0bgs5)tcj~UaYn#EbL>d)Psike`W(C~I}-jLpNq)z<jl;s
zNyRT37F>%*<8uEOXMw?kytU3PLvAEGV08kdygyj_$P6-(m}l{B-wUw!S;8);IXMZU
zAo7Jdt5S%BJxW9d!*kRO64iF$sXlmd5Ih4ug2V*ogAE;}7sfls^mT#o^f)+-WN>q!
zE<q+5bG(1hL3S*@yM`7v%#ei>4kr)p{h0((-nXQj6GMibq_<LsJCVM`Y^e-yE-o`^
z0Xt>SR71O}Hd9Khkuv~UQEjw}*VS}9?ENM2yaizW7$IBAA*0Qca{^yFU@dZ57>wO-
z-w&5Ci=09n_^b%e%F`j0C=jK*()56-HGddUpp+}YM#H2X%RbSMcU$s!zp)HKf~a=%
z41mMzL!B>oppj>mOOP%z>-5n8V%8s&+z1w_FPY((N)j|DJxQjWf&et$?t9TNq>tcg
zAZi@tB-cSL07}i&;-UN)H;E*3`7%Wk{)lpGmjJ9}mcyj_yzY<c5GB*ulrv(Usd~Vr
zbRxk9+NY0>0tb4ETYYN3k-yAdvM+%q&=(OWljQ!M{eKpQs#Y8{`yiDoD{^Sq3YBq5
z@I)@)7Dt!^=iZNn{@b8}xYf;v25msBVuQY|CLHEl;x+{5xisQ!io&(ByNc`&c?=D5
zVZ>{4IZ0Ke<Bns=XrjX57g|<@k(5Eu2}Tk-=j$1-&;FHC&DUR;Xiu<GsO7{jb75kf
zSmC0*j^Avtq*ucRhi?LzK3&NL$6wj5HRGOW=_I*p52^jJapG>ehR|N`G?_UDQL!J7
zzL?v@Z1XvNmwMDU6}s;6U}%r(OJ&uH(2~DBYGYHbz5ZfqAu`lbG<JVH$q;w76q~tu
zN@82Bp(B?%wc+Kq>I{F^oHEWzZ)puVaSdv@0b2Sw<Qe=|S`k^1*y^CPeUe@?3Q)J8
zK9TbB>jaBj@FX*T>I>BQ1oPk6jQu~)`xrPF|2L`qwC1GEHY>vRC)E?MKKl_2o}Yf7
z%K^Z8^E2&Pg>{4ZpXrj`T_MSXWb*TiV=O0oyo7(=cstxv`Gx6LQMTNh9G%;pUUi!;
zD8U`(T3xIG&}iEQ_Vc29;Ry{3onAKGn`TzHTrw|pOUwKFH@rJ4_*-hwD12;0o~}n}
z^Dk)L**4khHZNU_OcDrQFg4sqRIJ-h1x&oq_?cD*`xyLiTs$A&Z?94=<fec9<T8BE
z^P{VszBd+u*olN<(|EhL3!p7UR`k)ZcYthsKNdYZ--=Q=9aIQg#o5Rh?U7DByX|FG
z9ZMb1E$06AFl-D{r>Uy$!)?`G-)}3*ras?J^P8OH;>|{dVm9Q3=_+B*dRz*?WH}qj
zN2e>V=X_t?THRr95cxTXi1d{@|AOFUp8{=wNQPTKh+B_$;j4QRWVublkwc~A|5Pq<
zZ8^IVOO$Z=*wD<JmUpLArE=7AyW-m6mLyn<pU(gnG6}&yWz?xzd28eu$ptCfJkM(G
z@k!NA4*-b<i&PcUTzKN=1*jYu@&K<>*1`?(liV9KNJpCt3qIo{7~CD&M8c#el|4Us
zt>FvlU?hYO!H<$)DqfYyx<1}86m+G70A7CA)K^8ZV5OL(D!Wk5xOwpwMaMBBo2HBf
zt2s$hWP>%QvvLQT@AeAws5wOo9~&PmE!l2x<CQlmcSN&EP7R3^BTYF24bZ{*m~z2d
z#5&BZWQ%4Y5lgL_dqAcdL%4VN{W48QqJyOoKTO_syd$IW8P-MTzq7%A>y*q-?@rs6
zCs#wjA*6vEMx0s(I0sS-*_sdQ2!};WI&RS!WSSrqMvM(t8ZbB~Y2GZmWD6pBi)$G-
zbr%uweklHR9Gx4Sxxbq^53;nn5^ZnEE$Yas=<>R%p;7(*bARN?0N0w@z-Tu}p5p<I
z_c|02D%8hwQ_Vvy*%Z1Uit*H_5NtcIO<`*3g12HjY;&tHJP>^a7gmRl+-V%T9r~<Q
zzbRmgWoy(LN_<G_(uwnbmQCBB$Z;qgyhyI8E)A~_^~Rt)cX(RiJN2zkcw(?h#e~O6
zXx5CvNZTKM5cMSW+*r}Nek)K!mm`>z{pg+bqz-Jm`CJgAQY<T|VLocGWlT3cXt^&)
zlzbioZEF~Kbb8P@V#w;hDI1|R!~r)#Ux%Mkx8*<fyl7KCA3CH6$gJXzK;}YuvyFaT
ztc4{)Vn>~xb<eJ{7!l)Xc&Z3#EX`Rj#PLwM85HOPY1LH8gy;;4k5l2fcwpN}hnI4A
zCbUpxK%qCLhSPENag2ilJ(jAUkaNDKyea4uRflaoNoBp(Z^hoZZ%?eH?udCig(x_k
z6r%RceXvPMAso6FiRI$il!>{<fG2U{@TFjgQ5{ZBG^HmqZ6#CTbnTA<HsA0N!@K{5
z4q9)=7zB(P3Bn{qJ_~j(Wiw>w%<F`HQFnc`p`~G6J+~++p2iTPED%#IydE8#N{vf`
z3s;n=D!Vmi_Cq08VezP;z9P%eq>4b^6t8*s6_J^Ol;M2nQa}!bc-m*>H4nF~9s@c}
z1<7-hjR(qtQ&JczGvNCdYCMPZGFnE?RTvO%D<^_cwZo9zZ{8_lwG;&i63`BdD#D46
zAQkua%Y@8x6NU<|m8<khIu(?jM9MiR9DCVn&<bc^??BWE7>Iysns#RYt$Gy_A+T$9
z>JK>~<~eaQ_))zU{emX%W^2esN<d|b{$~HZazzTzXjjp6i&|fwDNvR8mdM`bEFeLi
zZ%&p)6vZwky1$fs5x1^43jj4M9y|M?s7vqrYZ>BzhCUutyb41Uy9b@wR+qq#36b5h
z&`M3Ao!>g_1jGBMr)lj(^y<{zXf$Xslwpv2tU0C;BNE*?-e$;z8u|v3q*e{>3%~GQ
z97JojWy~y#(DY(HG!14Zu9Ac%?<MC|xloQ9cAMQ+d^U{K(jGStD^1SeVy>SV4*e5>
zztt#7`11u|`vT(M)+N48^Y2X~oyT@*1w8vhGq4Gm>jfR;<E{#6UO)Wp8TV!%sm7fS
z#IpsFpT@qCcIjxnyfVmHy>^j}<XCvhKO%-YcHQz^Cu`ZFNn$yxT6211gSa`Nf@Jwl
z21q}siQ&3l8wtO&1<xwkK6#$nw7qcyGHnf_3vBEIjz0bv<k9Ggn;_G_!g`mqiwnha
zW!b;(wCNE=OL<+Lwf4Y5Ub>Qr<zMu45_70-`>qPN1B-+}J0nTs{Qx$7ucq;_+EamX
z_2xdlQTflg=tv^MmmOe~+!XvhiJPEO_$YdDu4WDpofMmxS8qOHcoqgkb@+Zgj|i&B
z*~WzG(i#?>K@w&DWj%+O@5MCEY*)LV)zJIHzDFnb{ljPfd#f3|8IcsD)q9{KaMHmh
z2Cv|rT2Y`tT_fAAq6L51WKub4bS`c6@7XhCg9cH55kK*RLBZv`{KZ{>$4$4{Mo2FR
zi2nA%#i??7f8V(&9W%1%Ur8lV2Q{?7`)fWFSu4y=|Jc`DRM<w;u;4_s=D(~!s;vip
zZQC;Drmii;d7nTVkqu>MPGbL{`II&OOvn9!ry_-N@Sh)bMh@2hzc24!xTcbmHJvOz
z13mq}5KIS0CwvZe#{bV-$4Jk@{y!bYENd#*Vznds%+~H5KId!0aU8{k3i)X(mI7*O
zBb4_2g(Z^Ds+Kd=a5z@E`1$l1Y&ICT98It)mGvWT>=c@@JIUOKq0lv8o_Rw`%<scT
z%1Iy>AZ;tl9RjJb9jYUDq$H)(?N_1}*d<O3V3cAee1Z{X&~QqM8|R=R2+67)WGqwP
zttW8!pamDgMGmBt7r8LnK}ejTd&6`*b_^lr6dyAxb6arLbK&#`1V33+gG8wegCV&H
zP3XSo-ODZD3@y@y?rAt1NDh`NrZ<&DB!nWJ>51v{YOiTOD}1DnQGyPxQfBsViy0;=
zmPADcl8&ea1RHwFyF>Suhl)jH3?P~!bfz3wy%O#vl_#J8&0N)=F&-@Hv8#gs4@l0V
z3H?b1E=Iy38APlB2nHNiLj+Y;7SZ>S<<Sjd(ij%Rqb<64H#1nC(x0x^%|JsgH|QU{
z<W;5gpQu2o!dQni(kqUSya3v-U&TitkKd70=$e`Gldf67=QdF_1Q{FC5Tmy{>YCUy
zKyt^K#6_uup)bu-b-1%KV2{(gMOoI4AtF~z&Nq})=?()VOg2dL*c1hiSO!r~R$k@k
zQwxJ2EKq!=x}x;I2(t=moCq5hFW_VaI7K+}2Z`6#t&hi6ehtDOE1&{4D2j^!NMT|Q
zgg_^LjRDzWQu)M>qgSp5vSw<mh(Z|K@|7VSrBn^1P_+g4=1}n@a$p>y2-aZa{#lfp
z{`}2po{l_sT4DK&jLqi#w&vzxn4WEXFLm+RLwLTad$)Wnad+qc9_)N6<Kl-<w+DcU
zj|NbGngL@E<Iy;ATI-@u^XYZx@#M4ANyme*;WvYhi$`QBZ`=EM!uGu>25mR0=LVME
zL;-N7P`#U~QTKDs3x+=el=YA7noN%i?IFLv0N|iDd~ltv?fqVSQW~80TX$ke$KDwr
z*%OSiYXCgIj96r*GfiCg`n7Ue%E$wjeso!KzV@>4`h81{ym|IvyU-N+j_h7<3XI|b
zfh<o|p2#_&i)~&84<n@(z{}yY-I*~J`?PkQ>+)`1?Xsoo7)15#!NF;r)8A-bywP^F
zt$pr%@oGuKYd-bnTv7A2<%<^Mo29MDMwIC*;2V155GBk`-ip+s#e4T|_u9Jx`>dj;
zzDAa6ntyC<l4}grmK0a+M3-9VQx9v4j$RIIkD*K*;VRR@J{trV=mN=%^R07V56YYA
z_rN{K9y;$K#Ks|FLBz*f0id%-?yI$bh710#{sn{kd%y4e^!qV6%BP1E<?ATfOxn&u
z+Xo2V`;{@u1kauG4Tekio-z7ZG5Q(2I#5LRA(Ek)0@H8T_^S^jl=|~}Mg0_TTWyoh
z%XQgcV?OM`>D3Zaax$Dc9@}wm1*P|OxQ+GwIe7cm)#b-<5j(bX7Y*-v{;KJHwd<Q1
zFSm)z5eK%lU+yIqa_`~FBV~tQ2B0BtMEznsiN@h~#lt1ih7~WC^mN*Y+gj}L#EId2
z1kYso@o@G!^u@r*gVlW2bp^#3SrEKA+D*!jSIo-H4$ZVd@JEOA`)g98l`qrF<}}%B
zTXysP&CFE=S?jh&IeJHH?TLW7n;ln%FPcwwV|X~(2J?esEB;CiJ*8bBvI4eEo`D!8
z&egHNsE%U6@F=7C`=Dj0@%ls|q`c@VNissaIH5rr+qiY8A1S`BSw{{4@xm(CD$1?I
ztT?GO=<yi0l%+(>I!3gd6uAlWkI3yIKzERg%YXwgXJIQ6Zt>GUttoB^oG`~;ZGNLm
zx>jMMOQYJ^l$29S_FQ<tmdR7?nXt9Bet|lVSy_2Z4idZ2!k<G)r58WIj7RXfr$1IJ
z$U(Iv`z#@Jb;8g^+xM?PQkVk?E&(L>NGZ>L!l??gVzl92xeb$CEm1ul8ZPGV*HC=j
z9hT*14L2K_^8)+MJ#Wlh8e3eYN)vyk^KiG<qyEMdYfUs{gZ}$9v-oA=#mJ>#y|VJz
zIzv8@Cs_>)J-Zq5iLQ*rkKJJT=OJ%l-xP_1AtJZI6`*PncP+|91rtuHKGidL4!Uz0
zp;L7Z68HErxM?AF?zUeNVpDUm!47=nrtf_1Oxxa7@#p6JY7xnr9V{-*PWeeDDkf{a
zKme<R`!^>E+$CnJwgVMRS}wf1oPR+V?ZBn*Xu#RxAv|7qu?$JfhDp5@mnYbrONxlr
z=t(4HWXCmejOknw(yqB6cu6%z2>nbF$PZhFZkYTx4ctK*T&xVBeb)XXpmoVX9#G=@
z^IBkCdRZz1#V(r}#8`?&3s<&N=!FM<=wZv~XS|gs4X&mgyYHqZ*EoQ{z03st@?QV2
zQ*tcWZGPlQ+|SIcw=v+SX1}ew^JI^c>Mg7N!tsrx;)~`UD;_(g8atAjk4v;E{O}~`
zU4F*6(yHo&a4FV!?#_l6UAi6G_<rTa%gKWy7i+8Fw5f#+U#-MS0O};v^JjPi6%=yi
zlf7+pyaq(#Dx2;~-risfSUcxmj6Pfm%(&=6u814fbLyWgY7##A0;7Au9fH58-AWFX
zuBFczUV$iSHtgkNpueZE>_P&KVeYpl!~{~id9I#iO_YNfie{-*_zuaaitST?2e^9&
z3IvfG0nJn^xyoVc-P=cL_2nsmnm`+;R}P2{!4xGZxDXAv<`Ub~9^b)j6l*+?!3;?q
z`#+1&FJyE;c4YR-L)HGOUpSp{Z~H0L6rqD)SwlM^;=9w3g19#_?xoW_i^$f-zu-;3
zSW^GV98C1=?Eg9BVP^SX^N+b4kJ=Dto~h730)~NLnQ@Jml_fOID`_7!*LOr)X2aR;
zNM2@N!4dEw?^uqaCi-E-@aLj^_yERla<i{`c)5G|ynGf*90wX5mb!m;D7xa_RFMKx
zxJr}iG@WnxI&WhY6NXPQ@Q8|6-q^oh+#Ze-RdJb6eiSbr#xYP&d-op$UMq4_S705H
zK}JIFPAIzP_Zemr-qfd>YC<C!K*cZK9}d4)!P6z=-73Z(pTGBa;b{<4s^=D*d_Uh2
zbkVDWUY6qj;Maa0;%(o;`(oguj48goLGS9O@^eygw?jO9!moeh(D@D<;ds5~GC!aC
ziMsq={6K+t`gO-a)`1NU{?S{Hhu}4;BZ62F5J!#)Y;!P833}d?Pe$?As*{-R3doSK
zOs8TA^gy)iGeHF4u1sg_jwAal2jP<UaFOClazJ1$%QUh>8A`SQuOpqqXtsR*ggR_B
z3RY+lC&BFWzsJ&rS?j~0k4n|w!S*!-8_a`^TJAcsM>Fjl<qpuekV>(s3s>guVfCfS
zU;rmeGcl>Eg4p=OWRnM{-3(_a=_)C{{*6ecOtkB=gaV|$s8JrATFADf)|mIAfkDZo
zU-+;pPofS309g@E9lFbWSNL%KFg|gn!tr#6oRb<A`Ke*gx;dqXKnglqE9k@mRKV8@
z+AhpsP9(0>*_u-jk-?G^@2`y?1Sa@!IltR`#FNpM{*e5l9_@woE+X#(u4IEx**x~Q
zK~$XOkQUYboiLMkr^xg~_rN;EI3c8>tyb)26eFDMm}5Suy{o{Ba7RZp?H-9wYJ~o7
zrJYs}W=Q$tY#&SyVam>z$K3bhM=I#?eR|;_f<GXsHuZ~*&b}wGD(j_rVep>_fT$61
zqSGgQV@NPTFQH=O;Z&EG!qPP^p0KRACVrl^!R*aWn=B6Ti6u>FY|m=#f@utfk$jeg
z@ox%s1Y1UoffCWD!pmr+6uSZBbAsSPLn;~cESg2~7??$afUqrIG)sjSV_fRISpt$+
zsZ!GBi=kvS#yW|J|IQq(_m?_-OWK*p;<;gn&oD%4d=i9cmIG%2(+Lnkb-nwFnL(o+
zP&_Bw!Zq#0$W<bwU*a|<-~DCy7WDI>F_@fCLD6XZ(Gl#_(9W{iEB2#iIbLHd{NM&h
z)701jhbCWQ8a~I2m*gcY=uJEg#VQhQu8S(hR8BqBa7JsxY!QIHtqnGY&)-bqC2o45
zttmv47N%%y?T)#ds9dpT28TpeShUckd{BI*#%m`-ioTC&*x{%lf}?-^jnn}~Fknef
zu8vatMAY*zV3L%JmY(`^QVPTy%FE|fA?=RMT208PL3oXQgcC&xbV8))7M8N+<8cWn
zmOHAafU5acepx$@q_-0Fc002!T1ov^l)&a>%5yY9!ts(7-QjpQ+iNOME_mICA}*a*
z@gFE2x{wrZ{oF)C>()~IDZ)%Q+0#n|Oo%_V=^g&Yi`z`kU>B(A9tpqq(rN)adzxPB
z&QpwkA(su8XJ$TJAm|$C#S29?KWiy2g!my-Bu+F9Al6|S%Tty&%^w^VN{kUU_^rT1
z-kl|CL$LfKhmn>t8kg)Co)?0aHdxesgnAm&56xxn)be$jv+H12Ktn1Rn1_reP}Oc(
ztF;j?c#EE3gR)`x{rQVvrwew3I<faY-{O6H|LpQavVET^Eek<7*O-O4Qv{!>Nkf&P
zXo#wFK`v;(gEb>$nBWL?!E{~s)!56DC$jfs2rKifcTUaH@yIC=cAi*dY_S^70Uk@*
z0f%S+daSXKpH&iW3hGewB)60!SecMf=D1e9SESU0%sw(BBWS5t^Q|r{Ux6ZS_ZC-j
ze4sc=aoy90QMsDHFq-|O<<zh;w3jdd_KU@2WNsyA8P(K)$l!?TE|l`MuLlT62(B1l
zRw@TG1e;!!!MyY=a`4K%!-+Njfr4Dh=9ma52DXE)WJVP$j9pY)@ba+bVjg|#y9qQC
zU1Bh~OP0hddVkt+F$+nk+~W%~2@aasTy)^={x^fXCw5<)U(!cV>$qtXa&A>Ze94a0
zpS&o3ip1tmpY~ylwXq>j%?;`K((w_4fkSV!yIP3MMK~Y_I~#+CmCFQrwR+lIUB*<1
zmk`UPDamK+ogdjOv6dcJtJq5Wz^}ETf1DfW^dg*1Bn5E)WN$G<I?<OckN)H^Y1!PL
zcJz5<u1Yf0$IC=_-?1>G^xVIPDXL-^#5HA3y3z6J_5BxkP5t2Jr=(;JW`*jbwSO9|
z6qYqG-6Zc=ff|t$8cEB{0AJoaE3=(&H0i~gV2+8^ZiqKnJ4V@?K1Nyk7G3qU#<>%>
zT$EgN5ir2gHK8s^)?&-p+48iOpc&Fb*j$f-d49UCaA732Nh2C0bg6ZIS0uATrodpf
zb{&&gkC{^d$`Ubkv~_6(Tbv|z_R>g;py*2VJ*}c_rFy1US!1n&lU-*2QBmx_^~4*p
zE>3`0nlbKCF7A0_J1zVj{;GXoYaMG#H^hXyYoJZl4>UM<vHs%{F<o{_B|;)W66NBU
z8%-v&(PP`pa+6fE%KWvFjem$X@^0>GyD6&%f({b2Cb)Tafs>LZHgYW(nNRFMmSHXe
zt{G-qX`$H>I32N&&Pq1359y(2s=U2Kc$MixqjJJM<mH<m{B{f$c4soCtr<3nW5`IH
zm2UqnEtg`*;Vr!PejpV2VO%f}r_%!sG}k>YBi2G2!cLWyS?`~Mbyk@w<b~kK)iX@Q
z4xqfRgPR}XB9g1S9o#-_+PZ{oL|Z{Xb{!Rg)n=&xzaG(^l>T0Pvs@&Pw1P3knQY9`
z=luhhij*^XX2+TH*jwM+hS|G75mOYLaagAt8?8sI1Ur-t-X(zSK)IkdH=X0ktUq^(
z8NuyJsUq8tX+{OkF+qt~l)2A~D3UF`ilEiyf>A5vac+`zNf(&a3Sbx<Y`pWzUGMJ^
zjMH^u7`@sLmP4j+9P;TyQnJd$>9AWXyoEVH^ZNY&I{<dD;oXZ(h9!Vwp)ZVxy;G8+
zO5eH8I!i`(j<7HTky;xWWVsNq9ra+2nNv1NVtPzI3c%Ve)Vwd~+(l)=dv`?HToSe_
zS3|1CsM|uPkz>ca?!4rwPF0tMMqr`M^XX%%ZRh!}3v=_V3r&v8y!-1n>6uTxy@19G
zwcp|0?iS4NJ9VBC`Ve$MDt^`bWc{&XSo-_7d}r=(!K)(Ud$|lttWvzi5ydV8m@6u{
zQjHDA3^&*=+ox0g{f=h_aO3nBAh)o6_rD<?$A3dSj{i-^Gp!}zNX-J%bE@`)=Qh&P
zrCO}`Y$9HE0Cps5Z_s%MXSaV#)M?#H^UBx_!tO+R=rXxMEtOmHe)riIYMWzoE2rD-
zk4f#Qsm{^$%L>uee_y-DZuX=XtOqu9_hbBapYCXW*RfyB4&FC;@C(QHJV}NCXA1E}
zIy6aEDV+8jvWjyXX|Ll|GZl@(tb=sx7312+SE_+c7swQibAFK47J7MH`f8i&hO}SF
zb`3V#)CzX;#-_?})NrDEXo2ZIKb8phbyytNz<T)l#<!z%@2PDqB_(QUqCBrS7a>SP
z!S*mw<RS}~MbU+AhW;aX^ErPT$SpfnLyY<;>P&TC)SZW6uihh(`RFi#Uz+6f%&*Wc
z4oYK@`^dT86CcSew$5k)9Q&Sm!CrGpLUV!`f0Dl>S=t=Q21+@BCc9ERQwct^gG9nx
zLy|+b9o>L3F6=Iwtb&Ep<TYf=I|b(z!#waC|Jqd%{%QP#Y#Tr@Fv@ncdHK1RE4OqV
zXX12|+shq{1qA~9L=JC8%=h(tdAGUfnq<-aTv%xzsA{b(lgDM@!~zfkrxG=tXvSJe
z!%+Q3{~aE)9hIkqHlY_MAV49t!>*rRjpCRi8xjMH-4?AtamF?jf5$fxvwh_r<vcTw
z+UQK~sl!y8cS!ffsLVUMc)Gf-Yp)ljpFL14Y`Ny+x?d{2dz*Ypj0d}L$x&S-J|0JG
z3Ix{gw!qJ%Xo>F_prC7~1_Pqy)1}8hI~SI;LD&n9JpqWkvWN=8{Du7hD0pu@FGHdx
zZ7z^I*AK26SmPo^Qmi-Oq!|DF`mdRGQ-{(yi;$dDzy+n3S74su*h*}E#`3R32nASn
z^_!2ir9as{ZXlL4h27&^@oh46;j`PC)_$CvRt;`Ac=xH}yqe}<)%6no?XjB`{M9=;
zZMNw*FE278`_72aaY5LeqE<LWF3jxdJ?I1tL#BA6Qu`Leqzu`x{j~)%nrMz_C!;x2
zHde)q;a5Wl3;kV;;?{~CyILWOt-Epw6S=8&yl9&zA6ca{gM)HI3bwrTyXF%m%dzza
z?U-eH|5CcVH~-FRe2p7T!%oE;Vz$KNe072L^w<WN?0lwO(!*%z7uB$;p#H@3>nOW%
z+UQrTy!M!tqHs~j#$QbkrYjTFfXp!2H)JtXh*|!K84(@RVDRvE=TkZ7nPOhKW{FO+
zKTKTK)QNv2ZR+BQ3b1U9owuPdl`QiopeR^F$pZQbm$f1aUr8HWm|jiw)eQ}kW!=>$
zG8Gzm+tj!`7ml=Pl+C_kP)cUsE5f8og7Q`3K}>?jR1`|+Ig%nLIlg>Dh)qs#z3|@;
z<-FcNZG9pWh4Q|+WWN<t_6P9>;q$LH@UQH4iS(ajBF7J*^m+F-)WM^T$dE9(2RR-2
z!ofM9L$gvZjppCGN}SZBCJ{93Tp_ERG6MM55c?4L1D!y8GPnrTp?SD0OvI=*^i5IG
zGLLMr%gtW?)vGI4rXSi(oOw!cpPh=Y*f|f)o`*!})Wu9w1jYj_T>w6E7z5E7lettB
zKu?rRjCTYt>I)7tx4{peGp~?r3&!?2eF|>}UD5BDZ-;9Ss-P27jme6nMGCr@GYhyJ
zBFjX$)G9981#MwT((6E$F(J4Z0^M4?tL&G6eIsRq&e2WF>1BFAJV`Zj&eei`HRv#+
zsEaD^m(LeuznR6HTsf()S%5#8>41T70VuN|SmNHZFlz!PeEaM8XKNPEvl{{?s7T;7
zw1@#W+l<A<{Kz6BO#=UY>GTE6ZY20S<VaU2*qFxME>AeU`!F?Wf5Z?lb!%BJ`qho9
zPjn?tlWv#7yR^3?1J`Eq!*T;~$cpRel>bTT>K9s9gfu){@i()&3nXZke6<zSlJS=E
z6Biz6>K*d4RpkX0+=QOY60c&bFv18UWdphnSS_v8381r2Kk$80BZy~gTEG`zK$g6g
z=*sA!*RL)Sl?}mVV%I(h2U~M6;}>uMZiSxUO2iE_{{;|&OJg4>C|hw$bZs0N4gYEh
z>J-t1a@qagCGOZdrHD%iu5@o~{ayXLpt{B4T9LK#dw7B?r_<~(2vMf|+KYBXi(k$G
z*<y~6IwcBOGc_vGFO_d-!E2%e>&^iB(?gp4H~uc`+&S<A>XNiH6zhLF)+K{*8ZIL0
z;Ck$|@NwEOB91Ji)s6ED3U;IyS+|aWxE0-1NejrISWXE5^*(lcd$MqOo+XJ6UtwK}
zQ(zij&kxCRXkqfKbE<Pc9#2X8^4g?WXM<KHy|sE$CGx~RXRll{#$ZLs^%+l{SS6TE
zxlAf|W`imVHFxSp8dJT_zb*~PU}^4wUt_ZBX7~Z^MaU^8JD2GX89C`FDH8b<tx{ZM
zt5nxlmmVgHSlW9FIU3@PLcjYDIq6M(*nG=On=%IiGf_kMcLF?2=F;pimpSP<m+8~5
zeLtJB8}`Fi<k+)#rE<;-E{Se)pYc`mtIi86pPGu8CD+Q<GJf4Cvn0GUuT^yw`w5Ec
z7v=jH6-oJW$+~{P+;fzero)a!cam~n^iqo4+DWJu0GLZwWcd99dX*=lUp$=ob7s`l
zl1GT;r}PisqOvrZ?&ywljJGMGyliqK8;(~lAh1^I682TW!RzKE!Ki5ie?_X5&&eMY
zo8mtS_Hp329Tam#m%TF-=r+AGSj?H(>Kks0CGx?bhp?TN(Cv^;+S<{A3`{2I8@`K}
zW|zyYMY$wRl^&_LUz5R2nB4|I7QC{XQfGe?x{D@=E@b&P^=D>FINZUDobf8P7<;3o
zJwJp6|Ebx-BUrT|ii|KjN`WSemDD?eMewtOS!b6mQHws=A?Yy(m2FhZGvq?3&QF;G
zgFp+C1_l-J+!>|JCRF5h$EgTRT1qL}jYg#5zS<Z-I(Y3C720fr{_1IZd@MQa)(|Ia
z@r@$<0gm!NA4Q}SJaM7#A8vWc>vLL=wPp0$`;>Pgc@T<hdtj_-B@cyNb`$fy@h&~U
zX0jeFcFmdiZpDn#aBDr|UDo7)D(*b5q9mIfizjn+YiF82Uc|_hGx5&!jkTzutjMZp
z<YsTm?x+~WEAEk#IbLV0Gb!fSc|zS1gUDp{uFvWxcWXI~6l4G6(-V%y1m}@=Dn;%J
zaS18!dxg9R>h%D{ajyFY7xq_K{9~CsAF3AI$CvY`AW5FnKMBG#9mttvcnK{ZMIX@r
zRN8HGdZ#vXOF~2U6~e(j+lDc(_up9j1MQ3y@C{8e0bvEYxthW75dM;MzaH&uggs<6
zxkJKw6`F!``p!d{&2TV9nzrvto4qvGP5TR&$+=_szmOmU!+)xV7@1lA*FKGD8zK(4
zo==Jk(1JhDbolhF_C9{?&o(xJ9Mm4Io^aXyGgtb$zI+wL$&xB9lP*Yo^~DL~2?b-Z
z%H#W^_@S%YhnqK(k*tAe{RwJ=o3H+N{0DpzPMlnBg!{~}wzq?|uZonRvqZe2lGXo&
zLpXQVudW$Dq6nJ8pIo+KqrH2MeU6x;#qQ2l1$y@iZVz!eXKuM-5?<J%oO}zz><eG(
zzb`M7vL=WSyGvetz1Kfm&w}iV18wwPp1b$Jg%W3ZV5h+>K5v)1gRk4GlKJu}gm-jb
zzJH0%vmqv!#&vxY2Yc`C&|p>=V1A{9b}tzn-Z;L1p|gmvgQSC4P;liCQv#6~hNN8x
z^_@``h9)Jfj;1IWDDT--5(qbPtQZ^l>6y;6nwgja-6<*sHX|So)~6s$P>Fi<0c0Bp
z>N(KUf`^fzmFVdyp@o`e>CvK?>6t4_7*ND0>(nIV3dfljrl*Q7t~CK?&!!XZB6v0J
zRn^Z0wxG@`C}-}@rAm)tl9(rHy>n*~WlYbj3~EyLqD-PtcFN@zq8J1>D$u(n7)}I-
zwQ2e=hRetUmU4^GtAnGQCXAl0Gukx+m@@M50~m6PFp5FUo%i)Sl$q3|>|~k#=x=SK
zu+x&%GpGhsaml=>S6u}@fDIW0BDIJjL96Qn0mH?n*mDTV`{Qp7+7B9lS6d!Ljx&UO
zdgEAAPW7uFq@j$g4eqkGo;q4h6+$`dOp|JjzqBEUXwV@<c_`pM7CFBsGZ6*_f0++c
z@;Qz^gi`NI9W!3)EvcL{)VJ`Da?j*!b{8-{`ZueoNqBHk<i9{PAE~CvSgjW<B_a*_
z1%dlrh7H97lPVMWI*s$*KW{Xv0u}-uBgnv`^9<nE*?~pjEYsM4kqu?^Z-jBMFxRgW
z8-E(erke+vKkw@Dh18th)*|n2_zPT6P_649b7RIMinU=?p;_e}+6R4B$$_1HVdZ@p
zybvaYA$*{QudGM*1&GqU8FDSIi^IJ)-M=ceSl+M`M?eQ-y|JBF4E)XMPg1tQ%W#xF
zWJ=koF+YxD+2a-_;Xx>i#Y}r{!X(c!W4BDRTF1}6UY0I%IR>C+qldCGkh>=Dr2`dV
zGFX}Ini=+l$OtIV6{lybpCmtcGf?A<FsdmH<EEp9<lW|<gBEcSn6I{pR19KfdMT{W
z?E+mlBK#BImC&$mw1HdoeM~aJa+}|T2(?$V>qjBp%#HIXFu{(P%Gohogqi}G&vG6u
z;Jn<#@kBsyRc6C*O)a2UoLs1w2-29Ngh@-vtJ?`_YTBF$YbT4J%-A+$KU}W#Cw*0L
zT#UOE=q!epdPO&AxoWQYXV9KbGUhsueV*xtJ^kApyhEK|Qy|C0f9q(8S9<2Ipbdh*
zR$J~}eUC!HiI-F2el07$xvsYy&KTg^*@>sB4(^UbTP>6^N}}NNjWfw!fE~j|f{d25
zKAy=1k<>?)0IEhxc-!sck)98&OjZA9^iNq$>Q|H~B8v7`R@O$^JQ<DSg*Bn><OW7~
zvyA+Yqp-;ig$~Wba#wb|2FCt3naXC=frx5_^`>x7MzbT~AQzZ-B64TakZ3P#18W+5
zWjOwC^Z@y=XI)jeAB-ZUQY+cfwf-dc)$Q8g@nklV-C`*}>y580#!tCZ7ww7o90-b2
zIX1RrblwuORa)mn#&=heq+rEPnoEv*T9)njM}(BjjcY|`5JjU@AzlN`fj?Z8<z;{r
zfOZ9<Xce#dd>R4b*fY`tiJInc?c-1dtDKU-ZXqkYRN%XlGPHPm@UEUPar53Pn+jG&
zXo-TaPxpb6o_AUW-#BZ@$A(w!)H$ZS-<AU1agO<AS~G?pGkhFtrp$*z_EZ)u5@+oI
z;$J%Ibu$5d{V{5rj#BA2He0ukzY&m&VT!_7T2fG_F#?z>u`|qK%ceepB%=;B^9XWw
z^U3%Zq;i%g)p8|Bk?FU8Dg7o-;!tgz9HqV0{w(LGK^X^bwv4!JY+|n)bne^5+f7%*
z$BG0pA=8cM!t&kcxL2m=t!^Pb8o(D9=Zv^hh%9+Fyacf%{GqH+aqJo@4H*|#I=G6-
zbwe(tq0k|GKN>GhX?CtW@K#ZIUcLFSsB(}@V*qJ2PlThe<kpDZR)iD3ckCvVXH)`k
z+xMD=iV03(TJ}%Sby?!XYHN;Aa2Hj-XRb9E+u>}zx9+QL;Tcv&@WI$Uke|gHWor<%
zJrsN3DJ`?U^k!6XCu|K>cp;VG4+X3NB7Hyg^RE@c?3RL<0U9W_5tzWFz2zXB1?05&
zs@HVU{dqk6-0k^J1S$DW4IQY7_d~kd{alOcbIhSj^^LJk7?KKLLQ=3Wohw9~Dk%Kd
z+x^aZ74;CntoEnmCU*;V08qX@J2xgtPH2)+N%uG?d}XKMWVWTTj+G{XxH8Kwvizuh
zYv_wHGE-uy<Zp<<%I)CIXWeIpvt{-JH}P1!ozt7vA`PUlw@kH<l-iv3nL@`IiO1S%
z{Eq6*#yDLRd!f?Dp?_(+bcR}HQ?4gwyA+xkDpYZ$*Xy^~)7*GOPI&^QRS{s1Xe?c`
z{K`$){`qU`E4QsqKc&-69;AL1)=kpMhuOm*$*H;&@WUNrN2O9M0SS1UJK<tZV+=&f
z7bi1ybi=mHfcJI{0%Gz;UMw!BtjFFZT8}HW2YUh>Dq>uu=5b7kkb*7b_F>taw$Aj)
zNgdAzJ!s`&V5)3ri^tPo<jy(|I0^v#`r}5I(=b(&n%!czP~y`N7)TYc-Eb(Ce6ON=
zZe01pW?-RFg<kPe1?o|Ns-q0;-(>-@5Hk#!$*nrFRJuM7FIEdh8WwkI`+Y`Qf^hYD
zRizWYj5f1KeErXLNP!g24w|~k2ipyt+FJUh6!-M2Vhl9XTCArOiFwG)rA@52(yngz
zU(<5!5;r)~haIf+9a!%*EnnR>hF6*k8=JJFF5F3WuGpEb%u^2SX)mWz`qzny+K5{*
z3&+p((om6OP>2sfQeAY^f>4bW?UA+@&7v2Rz*(Ba+|NY9pyz3kn^{|R6!d1u@lVm3
z2Udt}-S8b~BA2cK%Co-%s1n>R{6aU5^PLU@=zc6m@O!^R2*u802!G}7MfuWg5&mVZ
zDOUf*tf4$}JF*H`vN)nvu!G6+g?UuKIN4!gZw>Ntt3J!tsr*8u$VZ*NFfi#tlD65h
z@;$V0Y-X?!H`run?Abc*RLVpj@63S~_7?2=mm#~_ElnrsoqZLMIy$3F(xEkOV*3kE
zKbOe)Um5Seoc6IX{cnu7tSy&#MGDtDTWV*MS;379G?ZIWUY?Q8mR+%xkgY{ox-CQd
z2{%rn%Lf!_ZDiQFeIs$a=fVa?5Jc>Q)MosuMb>$Xr?Z2{tJ`#*bq68qeB0~74^QkJ
zP$U?c4UC{F^!n5M`K};oeD0AyDPGR<H>St4<8YA-Hpc?s8_4bB2xjO5twZOW<LOzk
zGvB7b&xyk0AuH$XF3nP+gL=8Crc~<L52~O0`{8FWa7rkSaP5s{=dHH&HFOv6%`fR0
zXx{br{R06H9ahkD83VIaVD<aH_hvN|k7GTbNx~Xr1peb?)E0;8onn$%JfXblPkjP~
zI8xJj6|CT>ShuuH9Fhd9I3pvmE^#_$0!D?pYV^zF@F1ZvGDsT7v3Rn{WE{7I7m|d}
z$wDeta_^|A?wQB5am$mm+6frLif06J26|wRI+FQ&20Avn0<rM|aioq}!e0|&XWa14
zt@mQQ6nX`_W1{g!N&<JdBn4!VSqdr*NV76)0g!VyYo5bGJQ3xdvI+%)<;~?T3UhIH
zzgMujvCIS=c5~CNa-V>Z^VIhr5;<bHg_HqK10W~xKwMQ4rS@bD23hoLu=e2`!5p6}
z&xl>vKd9zo%TRSp778J4#G8|V8VmFRhdh2JCcd!4)V8Ac3ro1gV;F0FT6>evg$e>7
zr>;-ls49gZ@2eQgXM_YDd&dd{35AS}?NpA!z%w_}`YYw;<8r%fIGGKib1>6RR(GhK
zS!oWc>};9u|NPNf`r;pC=Uy_`xSYDLb|EuN5nN6^HvYAv&OVK|LY%W@H%soUu4EV7
zi}it=vu~XOr>N2Bg2j)Nmu@-|lyQg~9=B$2eGU;+9>3#~W0h77^{I*#xSoFVg3x*(
zaM1eD$;3+C(g;zVxWepoK+~4n`_fdc3OdLK?`NGbcnaw4v1t*IUUH#K@e(xRHKqn*
z2PYO2<`r6ve($c-zLU-wS;SxHk%eHzq-mLL-23k{%|qz>LZJeXmQP}468wl|WGjc^
z7K#8wm-%|v<a{MW)<?V0qe^Kf(_zXU{(E_D;L^pdKBDcxt&RHi8=fd&r6c>F5}B6b
zcwW-UMs{p(3(6RnflBN(#-|d+C0!)R9WSHe$ITR|FazrhtcygLl<|KrSTrnB;ugOv
z5_uDXDiT1^COyJ^glRHOP1rf$rf8*7REVlsr7p#Pib+}0-AUmVu@90?6&p-)+*M8t
z?W`v&JpU#uPbM`h3rg`^K09o*IZ5;TX=1NlevgQ(iCXeqvdj?CZbZjKSe!iQs+t}K
zry9nYh*^sNFxQB@_~l~-X!e!2!faat+Kf!n3W0+aD_g=y4DZ_;USs3*A&FA8J%9xz
z7S}+ArUBNK$WvZv)z3jnc}$RJvDjkxF<cObn(QS8(yb64;~DBa@B@HQp1@-CtFl65
z$@?*dSm%PtKrXrI|KLnEUJMmk7%-6OE2U3MXe{nF%TXdn9vY&4L|GCfSM<CXDaNom
zYh-H37Uwq%eA?9-7wXk?pgACDWb4~Kahj4(E{xUL)Few<Jy+Uch&*d2SagXFT<~7z
zPiI6Q29QigY(fD+a6B?(nL6<cE-YT%petyb1`c<sfXFfS18%7fty0&e@h|WF0iMDL
z%g~zyXhxKznz=E5QH!Em*I$RG_#~yc-@47E-oOrLWN&MoUhekFo8C6CG-B7E`801a
z5uCa$Gd46_Gvkt4gzNYWX9dA0&M;pofhbOM9T>5}@Hq-MQl4or4Myos!-%B<tL~=v
zVXe@j#w0j7MDI@#IW597ed8s~_7RjTU|zLnLjEJO!A@e-$Cb%^X*^!Ej36Wuq~7v%
z<haoAaAXZq_QuU(rD-O~{fEOxLRCRh*7Z-_IW-i{Tf!?z>Nc2w5RfO+eby<9S&G1>
z1@NB{;-&Z?UD7D#<(X~2R*Yb#1Oko={iJ+EqTn>l2PR${ITh5eK5{Ddo^^R`T&#2s
zS%yLwQH7rEE&|5GdbLioNd#54(P%hgHoj=p()R;kcV|)i6ozgCE}~KUtb&_pT<D$d
zBLLUuqk*MvX4QYT9k3lONnRZFs<F;I#VpN1+N9EY%f04UE7nhZJJJTM*KJAF2ql)u
zI%!RRb|speb_%F6?&^zII0_9qM!HW4>n-9tPhL9Tv_trv#~Ggf14{MFRL$W~5zAg-
zFufs?1{;(r7em@{$XvwE<hiDTRRE8`&GZ$SlH0~ia#ORd9q&o3`*XOGP>Hk^6)QHW
z>jz^?E%U(@{)t97L+z!!O?~+EZdN-3lK=&7!0IG!yuv9}Js-b-M|&Vmij%)eJ)qSn
z#PUiYo>>N}rz#46#ppkgMh<-H3dS=>12azjQAO`T>INuw(5mWoY(S{8Ha02M^_yb@
zQAN*)>jug<RO$vod%F6}ZG#j(X7qxeAN=)#U%hbD_p4`tshGm6us|#SAI9D(NV9MY
z)67a#+O}=mwr$%sDs9`gZQHh8=|9b`?v9?AnLZcia!2f&^<C`!#J67Tq)V5DG<Xo{
z7^1KPL05gb3#Ix-!zU@OeTNO9*-pGnC^xabd8CpZ6gpxjXAzVuL`d4{+a_ct$}1;z
z*Jfp#k4VnWw5d=?dh#ea4yr*LyBgPptngKeRd)34RxN0IiW|1l`%MP6O5vJ2iw8AW
zMccB=y_D6&rmg6HXu|Ik6Ll<ksqmPtL=2C5L-C@}A?>F8e0j?~qkJmCTc2s2Db9~z
z_chB~8!q(zP2<Nk*97#+@jxfbpWq~lZ1{EwA=QnMA}&pPl}1vJg&CD@tZV++N{cb_
zT!aKSRtPWBA8k|yA6tw6%uI($C5Ykb3`L;=Rp@5J|B+gMpmq^(Pv1u0TL2NlPECb;
zj>Qv&8O?l%)GJ_!Tra!f@zwXZo)hs$BbvzbPuTIlIDl8->{Wa=uY7ks|EHz8{l?w<
zwM}p(d7l2vl@r7|*r-f|uzfxeOQHZSI5qY3ueVfe^RiltTv`h$eo6M0p>>(6P)wFb
z_|Uhp&1<eG=WR#+{MI<?;0d00p-Yh>#SQk5k0uS)^ABon&|K09ZeSa6>4Yq3`W5LF
zU&b$Y&Wr}-)iBdWI%|mI6$?r|c1$MEyX+fOE;<le9%C+ZPB{96bW;bWK9g4TsYC4>
z!pF49*v|dVkTvUv(GQe(O17qoGKOC*y=~|zXAL+E3#Q0>$Itd(nghkERt`@IGzyIK
zyt(2r>a*N4v*;#EwFy<|lMBb5ziKyWINW)@I-+L?XU39nVlq0D2WotJF_f$$U2*<-
zWkQ?Xxyt>0Uu^A+QnM(mjQ*1C`Dh=7zz%f|0IkAC^{Xj7^F@hGCs8rrw_Bt)p~W0I
zZN#@lnTPE7qQN?(9>@MPiypPhu$kDR3flUU@3f-e*mckNpHaN2`0~?K#c&D~*4YV8
zttqXe;{$metR&n8j1fwzj~{qP#^<a^`&R->kuHj=`}j^Wo7qI;wh-lQY`<5b#fO$w
z2T?1=v9zPS@CN$YjQEwPc`leh*}wN$5dy{c8ID{ht;A&?70kt4X-OjKeW25=Le#*-
z!5Q1&-3fp9kNoMs7Z{HC*&BqgAHE!aGz(uT<2-cCL&0lHdWcaeN-}!1;V`N^I4<Y;
z^rur^tkF&K33AkO<<Q#qrhEUqI3SKjMK8UNRqvbNuIbFY+;?9*=>v*Z8`$|z?922Y
z5-T`Z8UHJj#I=re(s3(n?_8aNoug{%6nMXQigjh@)>7>Q?en${A87;qhOp1X9y~<?
zlIW!%FIN{cE4b{71c3lB!QHPz82fJv55UpV)6)e&SI{_J)X)U*?oE9Dt1K^QNj4@1
zG02iMbB%tcZqJ#HRad<;z`5j{>2YU~oX=lr5Gf>oFWBSTM8M;nTbJF7j;qs6M}Esm
zpp%5Rqgg)rgHOS+lSR>L>7JCU;PVV%U&qgTUk^H){jC2b{=GHzcga8Z0Su9!>|gf#
z{mBGB5NUd2qoNQKwJhMQFLN*NJGV*p$rS7&$~uN)F^%N8Sf~b9`CGA!p0sm=iJ<X&
zEYF{1BPY)vhpduiM&dY|WNx)5kQk_m&YYrI{XE>V0i#(a;rAc}fiMv!&Um~DESw2j
zw@crL`F@{1Hx|jw=l7?iGb8)r0y!quxafSSxmQzXaU)Fi#93xVRoU+u97WUR@Ex7_
zZ4`^#&&>{+e}bm;GAmQmO^_jh487log;!#&SY7t+uqrIpS<bndRr5idc1~8R(?!}5
zdjWo3@3NHoty_U0ASfLygr(R0|7^|Z&H%rHpt|j#y4N}UAb*28+u7y<Nr)AQ(V=ZI
zcfZ^_^Gz)Sn>uz@4@2uM?K_F9W?(n5x=cu22LRVF<qq8U(U)nQ*`M0c2hb}xh{T{A
zYRqp2TPnkJlyqh9F>+nfsU*P=idalZBH;ri;=r5bOU{q}#oy`mm_;HiOQTEzNTRbX
zgBqRC(z%n$_OO1Jq}yj(us0~oaFLwyTjGurDC(T9XEqx+?dYN3;?1Mya5D#ho%@Q<
z(qbbEhBRP04YTZ&v&Ldzq|?Zb!Ry!0e#QaQi4E_u(-<ahf2ayP5fddi8_|<LkdBj_
zv+J7<r7{PYR6CB?CfcId3n-1ZVe_F*<C9PQH0wZl93G5^5}=6v3UP(qFp9j#i1U)M
zj(~*RWmq9|m^I@t`3761j?u10wnCYpLg57w#~B3K-f&EC<ntarVm|IvkKigBG81cx
zdkqLcMlPb{A>JHD9=3#!R$>QK>@cEnMhlfWhw(}Kz4$QieokPM4x`ht9--IW@^j?d
z`kwzqcSrs`o^9)p9=@W$2{VP!vR~HAKC8BIC$=jhnH9;YmI!%9X-l!3y|84$Z&J<n
zEt?a}eT&hK7=<=WZkY`$H?lce93(=E3kIIw{1C+oTypAeg{U&xuTXx-26k+L%sXOG
zzdJRIg3EM)>X?n@QAu%#8DqLc)z<Zh#9^`>^Qf_X9T-M*tgh6|>aN*9CSXvUbEt`h
zIr9_&*28dgT1j|r>x4Z`CR!-DM8;QHsx{1=R8>;2(FXFIJuSY;>M_jR(74QA5{e60
zkhA>(vR=<)L=I$*wG&8%mSVnMqOJx<T4M-MV0Iz|zcfyC<N^qOg`j57jS%pWqO$P@
z2_~CtqgmyE5dH!)TlO7$l%SQ^0+~*Jr(&-a8gjYlPda!-;2@`yiNEqOj@S_cvddUS
zHoIm^l|M4-R0UQEu+*;hB6|ipkmX}xUzr#B+sP=%OB69G=IW~bQKUP3K1*S&dFu0u
zD}8S$Bxn2wgN9yGr(<-|TG#}J7%F~#wpo?E)H%*Q)kyaI^w^rwm2y~9Oiw$y*G^Ma
zNp@|E08JwVw$$Buoe}fC`IZ8TXWku2dzdmwC%}-_J6`TCt07pk_$X^mVcI;o%#D4J
zxmHw%H@*kB78iOwy{Fh^8dKYWGq>juJ0l2Cu|W;oj7n=?k4;oldl%YK+YX<o@%jL{
zA#F8vo~fiydsXZdY^L=3^mak*76SMpi!~14uC4m_5YUwH*)BX*Ps-Y~dd(1iH;~XM
zGJD_{z}w3_NrJ~9;{%)QsAF+8`VuE$v0Bj_E7OJTUHB1hm{7^i5^W5O!K|Wd8QK8X
z5r!m-=Y_l?9-UAhZxBhUgybkTogzohZsNTaTVgju0hO-OxbGw+79ks$4Vb_rB?)Lq
z7N#Uk=Q-j?QZi~Yknt6hVoWSZR(`LNqZ~h{`g}m7EUA(j3xYFERxmdcyEBgn=C5nc
zmK`m6?ly%U=Ro$)3n{rL)vx^vFmUoYNu9I&k><aPS6NT#NizTKuzPf((}GY@ZG>~6
zSYl>jYpS%Vge<hr{)eSKTOH#!Xs$F{R-rB5JC~diH&h=6ICKn6X-l@(XevrX`_Lne
zlK1OvYga{%W40yASa+u4ST|I9HY-z0orn7@j|!bD@@-54<?JZC(vDd?B0+l1^@*f8
zXLL&7747o8R}_8f^Z~j>GZk61=0Ufc401B@Ff-K~VR@aZrh+>BtsMwS^u7D0_-2~v
z5cdzU9n#Q{N2RkL($GM<*M$7>BsQ)et)6K1vz+6xeTXB8YClJ|i54fyI-YRWR{e*3
z0#M+gqU`+TJ=Upn1|0K2z^TR60eddEo3RcwISi<g%QmaWFEI+jXM5R!jFS0V`R~`J
zW$re!V<Qxh7-G7Dtqm;=QmjP7?})G$RjtmPu*@d@O?J(8(K7hRJT4Hrb7hjeq_Ano
z32MkUOy>(_5>IgsV?0h6s<s*q={@q<nm>E{v~&9YM>G5v2l66UJ=cA-7ioHM_TM|7
zhwu$`46P=hbLnnC=cJ?Egf8NWJ;vPi0#IzdoIM$$C(|gNnXFSb6V^DVK51~5(0RV7
zjAb=e3|~W`KU3BrcQt>rf{}*vhRGoh(fE9e{$kw-w}%~cLAo$<z(C=68rf&=aOt$h
zG-obTq`d_dD#@Msu&+x@K7Z;qFR<Xnh0B;5q9i7_Z8jGCC6#d4Ar;<_cp4qP=UXBT
zeq8R9sY=91)H$SXZR+TT5Bu1R<h+aY*Povq4Vf$zCDA>9ICam#e(kHD*Ay<lPI}-_
z{9erCjh<uVOIw<InsiR$#Xqy3+$vQx;T1^2$9vT4;b@tPJpVfh*qQ{?AE{a92mMxG
z1xBq5hIEx%Xot#_2Ag5c0wiB>6eyTOFBw*Ldd_J*(N!xzm)15498p=+qB{51+=km-
z+d`x(XnCiLbTpUpvU4?a`|}DUFARu>mIK%p$`>H#=G^y`ONfT@XNKq%_$1gL^nadp
zF@TqZDU^!v;NPa)Rh5wPs!|f+Dv}bhDwHNIB*;yv*yT!hKrS3Pudz%PE5-zfENO&4
zb`cl+P;1l0Tx)+6{%t)BwI<<39!n7tM0Ro~ysE+>s4ZN;h~B9z%zxpPyNZzf$~a(%
z2(Jg$^+9bWrk;;ok8(p{0D1kMm+O?YMik;IMhYYOR-q=vRiY$iRH{jK8CPp=;(wwo
zQ`tD^9?QqANN;%03gAZEUEd!_D?qs>L@<jVgZqItUg7I{YPZ2R<ofZaIfnenI0AX-
zrW_yo(pTJ!Bb(o{22P+kZTDuE%^!*v#~9o87U7w|s}Y5~i07h7>{KpEUWk{X{1amp
zLPE@Tm;>Tz&zW}#cfkvK3t^k#e3&)Bofcqh#7*Anqw{(L{M-T7`Rs?gyHPiW`Y}HH
zfgkA(n#l09aOkP+Ar6$W+z-C%`Nu>2&41tzzDJpQxCmL|5A=f_ko><H760dqk^lc0
z$N%??3xYjk0(ie<H~_K&x$Bb=^P{2WT$Vq>A9mpohf0v1Zhsv&_&S|`#)$B0(^Z~9
z?z<_6k4U~;@B_|B`i<v)ivj}1pxBntqbw}%OENx32^NUwUSIhq77xD&3r&#KgTh@z
zaFsB(Ym`Ko^5uk#a-<M1!eb;X+{I{Sz0LD}Kwv4hL;oXiv;W`jAvVVU61cVhKfrwo
zvR`CnRaxq7-a6S<eRldl8R0vDcPRbs2gFAbnW&}WKDQ}%Em(vH2`UK$$`_D~<A-~E
zcX9CeeD_rqH~<+pd2w)^Cb$4C>7P;@h6eu6LnH=kuOFI+SF)U1i>ymf!sx>Jeti)u
zk01RYQgF^+lFP@TAoKID&}X+w1br)}IRRv`<YLCz$9T}VG|@$~u$k^3Z;QgG_`bhg
zhk<95<)7D9eoSo*`QvR$5=a#k<Nbr}K>Yy4K@VOEToDDc*K7SDdU$_)O_Wt7k8e?b
z{uvK1!ich#)SFn(@5d&bZ%KN80wd(lyFH2bkckoqj8K1gek6ZQ=^g6C&So@6GS|db
zG2!Vy44+nsT=6pdtokI94)XZbCYZ#Ma1x^uodL-FD{&dSrO*+%!|1mSye5|k@NM&r
zpGxtjHI00_exvFJxbsihc>veR+gxXE8LmoDjwR?!0+L{CllCI)nh>^kCVhrM?5zjh
zeB#I9g;*{l&n}+^HVL;o9T$^EGBMi5-wc*$u7zS*&4Y8qx~E8aJ;!)LogTD=lDM*k
zX?4c(jf=OGH8m0^i&8W=nlBAy3LBB~Tx%Pnhv7C_C)3mqs-7%_Gr{8UcPOw6rht6z
zmdG@olEBO~6C#2sffH%QwtL>x7q73L9nag$7n|Us&1UpXShQ@QqsZ0KQa^KxId0uN
zx7|GV-80lSX%JkNc4qW*1k&no=NK}Vc^>!q>`cbGidl_cv-B!qE$6MY)@z;h!ZaFL
zlVF+if`DztQ^0qmgmWC7fvg1<SfxU+oBu6>`giXi4KMI#7{^#Ta4fv7j5?0ETm3-b
z5Qy0V`yF}(o^tg6^_jB+{wvGE+5jN)f2*xZn>_XMuE=tER-&wEOY4-pG^J&$4cqxU
zbaYZbm9UWE9wY(<y2%epM0I|aealBopj|+a?PF2woD+U-g{Y?nyJ_-%8x>xRNK$Jb
zel8yp&R}6U@M(#N=r+F&C-T^pp96i@e??66n4ImcAu6&SNY_iYd6DtpX+GU$5w@D;
zfb}zOF{T&db}70UfnX;4?jT_+-_x@!%jI2>yu2m7b4;+=LHMW(SCQ_p+m(+yOpuLx
z0N3mQPAyFO7o<ms`Hu*4GUYE;A^bgX>Ds)%xiB$6_;i33qSrw;TMKZ_tC;F!n3#F=
zj#R2bo7Yy;;c{ChYkCF_>($8W^6GkIJ1)~W<?cmKA)!y~b}A?x<dNu<!lQfSk>I>W
z=MJ5t5P?}`;g~H2$Cy=KF=RzH?t}AnlZS$dVJK8Hc+I13l_OFT6~CuxYR4#m*L#0x
z7hjvB3pccIh0XqenCa5BGGNbYL?g+S(rZU0n$<LZF$VQ0^x8rraTDU0|M72|e%R^6
znGK3yZ5L3ZWbiK>q)a?!_gGm)I4W}nXU}Ci$E6M+S*X<PuT-}t<Bb*V!YQpSHLb`X
z&O_ORED||4NVMP_B6hih6hPRxfSE4vRKsV)VGi%xr=^ixRr=ahc8Agonh>FzhzXX_
zv<5+ka>o4F#j8X&;$(;$Nj)FZ93eE;A??5djxC0}MKl*5oAE@a>XhVEnP)z4KcYbV
zkSzw;XDi<k6XyJ2#sX@rq>tQMQ3kKMY5fH+IUt&GW>s>}YKnmq`Vnju7cGq~mnhbH
zY%|=t_%}81TUaO-Hpp4ZK$Zz2a%|IjsTMn!lk%)=MfUVT0ScR5oxXtcEVi?%S_FZv
zv@P0^1oHtcf9P5_Y`K2yUqRqN;g#pshE7&NIPz;F8&xR;Td~uRF-qF;|G5^z3TFC)
z_4R%=afa%>dOW>?qBUa@GLj+|HO!b_JM~W$hQ4h?YAD;9^ff$v0X1Q!IrCxNy+RlZ
zdOVbPFy^XCN!~n|nhDEa<_ETX7|da(m<WU7Xj?;aUMW;VJ(aNJR(Y~nP+dtku`Hg0
zUp5}EbUp+HFV+qe#aQ=uOoBF=h1OMa$=rB7Oh}t!+=6)Ca7%hvkOuN<kmq-KQ#}TC
zwmm+65oJYMHo8Z(IC9nzc^Veudv_XkYN4Td;b72FXe}jx&CU?BRL>Fo-3&#f!(;2^
zudKxoD9Y4ZKiN7E9FEh%;cs{$!o<7StPand9WOW%=4rfEUET@oWLwS~@3@9lS1Nhe
zI(f|yQT8`v;u-;$C{3I7bRo9S)c4)*V8z;d+kaK0kRU`l(NB2aBz2EqXCLp#v2tF7
zA=;l2#B#ed9Cqhi-T|Rr!M%U`R~3X%qWgq7uNZ*L_JB)q2L`ou>t2A3w>Vd;7U@1k
z<>UTo@T5eyb*Z5+)7i3%%DC*$4Q17{h$Xu$I-C%)4D1do7x`#l$w~8Z{+7Z{)LF1$
z-p+pXl`UA1_O~MIP+#Ay1}Emn>(*E5-V(9k%q2sk*tOf6#&w7+D!f?OVh+YyM+2NS
z{#Es5O>6oJ^P+`Huxe&tF$b6D=`7MbzzIxotPXh?6-ZZY1`+)G%k1tFStX6dn2NHK
zp0&X6lZuGbkf_68#d1G^8ILO&sOg7`htI8PpJyMbg~SW#8k|g5K#v<W#6E$nL2uUL
zcs3-}23j|QB+o1Tjh(FB*f37tQmOpD$y$L>DA;7LF<k#88Spr$wL}SJFjws(tt)3r
z)3eLPFaa%ST~G<@p*+!WMcaqQc@n{y<P`BrE@$(642bNQ@ixr5+#s&3g!1ynFPTK9
zYAN~3wgNcdwI7BTQPt-Pdpbd$oNfyKlir5)O6x5&^R390PaeOU+F}sXhK2kkShXOG
zs4VH_g~TYpx&@I3$IK5X;$w`_ArZy?yIB)metwxfsIkpuj(=K4;<ytwV#-H*sJ)Hh
z|L4GmdzqHsj38IXrri58eK!!%(%R~@{3g`M9CcDnauZw&ajD@nC&*>fe-nb{g;_wh
zcT7Md^i^^)s`^ntEb)ayGHfu|x~7}q7s~$QZC6f5FY5M_Yobu3{y>ZOj7071M-e{x
z$+b0<6H~i&EWHk2{Qx(J;I50o5kpjOLWRHubMKrb4(){=>5o3`>nuFs82s%&d08G(
z(5s;?U-V?3YwPx_%vldBf6tk@BQmKbYaU>JgITO~<;ABH(EYHYw5M2l_9)d8$EKRU
z&xlI>E#Hi(S%1oR+SMbJr5@!go-+&8@iH*5a_NG?|HN)f*W%f-?A~n2h!^CAZ>u$k
zt6FiM#U<{lNBhj{>csS?E%<SW=3N0M&HK}|)tx7}f@M@0eB|LYOnFsKC-fp5;?F?P
zF!BE5QbEoGw(fxwXVy7aQw0U`^OC#o3vW}06N9-m4@`_j>b1u)K#rQYx!IQ5MlPV$
z$*=MBKA?K<DY(2u!E)#@fTD+l=aF?b%u6bSw0r1&KJ;FWtEGiVpw7~=68^nXWHrG+
zpM&#qJQZl~V16t6UQX6a?sHar_cZ4^+rwJS?YMud@{P@13-<?PN2VY1KLIH7e*jQU
z_W#b?w4yx~`?no&=S7{O3nu>b<l`5{@6F!pUuJ3CJ=oJ!6DCPr<2DZ;l!+s4BvXw$
z88q9r(k2ZWu9cVZh3AiR5pw*!*?fAvI>Fg64Uw39zMpK2?ScpNMGlHp6AWU_21{cv
z7il0RIzwlm_>vU}cZ_#CgZ{2QsMA8wzxyz_d$A^bKA?SO9r8xp3iR!;Nhn?&+5u=c
z5bphID12C2R>&tp+`rK?IKSWTm!L-|6%Lz!E`Of-cKP+bzZT^sC}cU@m-c%5Ua^%J
zh-mUa(T6p7zrN!x+S2--jQ|Fl{j6kUViY-kFR+=^d-J8KfwC2=Qmkwz`M~L#*5bi^
zA#5R<l;Rb^n$+nfJY%<u{@@>yThkoUOPLJ35=HyC=Mt#vDp-5mQUz%gArj}gKf#nN
z>b)t50>Fl|kHicMp8IGKsTh^u5qP97ZvT4Qh+6)7moHs69mH}Nvq$?TsZ2l$_{Lu@
zO!}})&q+}hmgD%O`G6vq;>jmFgfLXq^}p)V#m*W=TTAu}gNy<xLp>_mKGG1TV9Q2h
z^2pCg_Ei}&80cvd5v1b<zo~69Gwj9QK}@?Nu^aSW7tDSz`nz7~rr#FZ4f~vRfgAM>
zt;<cjzwZQkT85*P4)BOa`->)9)#(rWhlhb}W8eK+>KR7e?z6i5TUS>|I4XBR1uroP
zz;Z#$ln{W4MKD@;`N8s*#B+sf@RDq}6z{{#n$|e%Z`cV3p#uRziBW(+%A%%jH(1q{
zhI4ij^)CU>vd46+#{#!uR*%cK5z#*~ys5JlGM;s*Io`h}<)`?(uIn?-5|xd`!~1Ro
zS3EDQj}lBf`oW`6LW1sykSu&O?;oZ-EI$A3Lm<VGB0{wayKjCAoy4vO9xs|!U|LRq
zV5S*ASTPmeiUq<h8Wrp_<aGg!8muXpj_Gobv&HN**Y)=Cy}2%BYQwij-ql)jf-ROS
zkY3zAfB(|mvuc19Pjhm;<d`ZDV3Qr2pAA9v`z2G#v<roH(y{Q@a77YDHMtr^vt&BC
z+;F&t?qD86!s(<8*|pgHM9>^Kp;}lxF3M*5=_Ny)MmM*J-aJ*h`ifU6wRM6yhhjFo
z*PdG4$8H%aOpbjkEsaiuX;G{d9*K;nXxOUdHn$Zf@Wi0g_9TuB-&&{bh9qX;0Dexe
zcnLX8OO6Xwnj{@rQtem|WhCZc#$^dz2?fa}HP@>OG8=<KmQeSBZ{NkhjeC{Zuwp$h
z0?&}@ZA!c@&2P#rd1VBiNyH-Q1-b72ZKd7h=3ymew8QjP=88?lO`?u#68e<Z<g3k9
z?eX}^2>r=5Va)+2CB*#cQjn-z`8(k7{ASaNXO61o#5&Cq+b>QvPJWglN7-uE7n63i
zL1Ldkm&&ARG@k<>wuV1=FXB?Ob~Prm&-5pTo`&r>W@InYD2WaO-Vu*qH`k$pW8td$
zYZ=<>bc0t=rW+Al6mCrpFim3n0ugrWR1&UDs1!!$Ar3so=b4SwQa)lAkpYvL@^n}I
z0^jPj*fa>GxZ2VY-K>LXLP5fTAx;~tyG;?+xQaa;H`}vP)4j-4=&8RlDOQ9l>7gLq
z>p`7Dd83|>Cfl(|=lvoA7)@E;o;eO<w<5pBu_L@E5p-{rp>J?acx@Ts&@|!0YM;94
zO8{BXo;&Dx@?duD-jsWiI{W36CoVO@=h`yJC_Of<c91vA+0<fUs;T3wTt(J1T{T3#
zlzFxqKi+dIfb5aQoktO1)O@K`oX#uaT`_ee7_A8Sd}>R}70DYZhaaE9wWD7yU?4q7
zKJdDNRP&N1UVS3LkChd6$QVshb%$=PL<NmXWq1<mToLg5)IGyzNU3(s+?dkj*HESx
zTDNC3U-fZ>v^~qNtBcr91<M2eWRFRf((IGui>ZFcTNcuLACAqY_2XU&E;Z;r6$UKE
zc#_&$5o;FGTkrl-+H91$ISZ}d{$;d-h_hEIHGRBse57p#I&JRCjFN_-;aDV=FU;#O
zZ8?oGb0wP;%5K$SKw#KuXS*+UbY(CK)Lo#&h7rxBwo`RBWA`g(qmGHRT6ix!?Sk9;
zc;)y)sUb9DmGNF>5~Iqnep~*Gh4=xM%ToV>In7^kn=NnhJ_UBr)Q~1d^a98!vBtq|
zUv}joY$N4`mjIl4cq74Ff9rj7glL{2H-}Y>BAq5}{($nT!ewyZkMHwWeb3=lJUSX0
z+U+Dr%@bq5#i@`Xj-azen;(2kGs^yw0Yd<1@_$Q_^AF477c@S&dd!=D^u`O5`^ofs
zHfO7YXx+y}fO(q)me)X(A(z7;nW1P6SBMw+0y)b1myigf6yoZnOVdA5p?9;Hp?qqc
z2b2>?o2=nE>1_C!n##Ung1v6Se3N^N$}QCBB*Z%rJD=cuPwD2fWihq`&}}1r5!pFE
zTTkF7Akp%R#64l-zRCOdF0f-Li65yhkS@62>O{{#?&_ddJ<a%;x>AJn-&&S^%I%tB
zwVopgIzJAzre>vVPXF^K4$<Fv=^%A$Nfo^vl>!?cxyi(ML1J9J&IHQWCgm~z5(UqV
zc(%-Zs2=EH6rg$r!QJ5$IMwo9Cczb<CxG;<`#t;jifh;8<8tKZpWc&4-(EXjbxRzM
z)d(1)TRx{o_~}3xuuZ+|B?OFkEBJp-z^{nTx*X06x9I%e3rJPhfUW(3s=-syf_W)<
zEeX0;KNA)|cQ^DJ3J!xO%ry~L7wh>xg-PFb06j+Hm~Si2R}fJiZ=41y*fZ_ko$AvN
zQ)Z_(MrZ6%>#YO_g9vT2YH4UBUz6k>IE&SP_V@i{IeOjB@DJKAU5RLL?@8N_s`M+$
zxa_2R<hm^1Y=J+orE}w-6`*AJ?wo7AE|;m2an#?B?b9M<NmwejrBROGlzSsMDI!HJ
zzhs5Ya142v2X&B=%noS#yIa}A58ihQqU=q5s{UcxGJQ5zt0<s4i}L;t@%kju$~SW$
z_MlfBwXLfMcgREg9|rF-Ppgq|)6P7?J{7#Ig$Zt+ZgQuLJwsqi6oG9?*$>5+b8FIN
zBfj%2oA1!nigC&2JWdod65CdbNp{0n@mx1fV`yg#A5uhmLP|IfM6^@bd*11;hE=3Z
z64hah)n~KAVf8(3K$pjFsEzI#a1!ggHnb>-)2#Su-F0#<v8vok-onfnb5FnE^PFXN
zsoA5A&sJ$yd|p(<p_HeMD2Y5!QLj!Wov-JkLi>CTL}E|-!1wOH&FrSlmtV;eZbEK*
zwtX{Q+OIo4>{QL)L2}nbxw-|^ubW&dzR=F=OswPEVN|4Jic}e<iR+c^=t&_y{7|kE
zMVpf`Z=_J%@KP7yMXi4JgmoM;$l;oKIE#fkvRfUKsiLxw-}Th~wMnj7P=`W{LR>Gi
zmfF*HUU;}M1<z0`Ou@>{M%p;jh7_pH>mn(Sc+%REs%X4;b?cWD;y1{DYQ9p?#sF#+
zf4DaPU6d*%)CZ@vcHi?XDs`B;%~H!w)UcgscDE&02m7lIG}FtneXD?V2tuQyzgjeC
zQ$TEq2x<rmYLpVVZ<*kcU$h+Q6v`A?dU+_>7A6a!OBp_=C7yt`WO}wA)-}`t;-oV|
zTiD_C3JN)oIv)d2Kg~w*z7f?uGg#ge{~EnKT0bMgrQ+EriWlX0svK))_C7jXD_xbq
zI8I`n3r0~1qx95v6KG39*U%>4Hjg;j<pp}si^mj_p}y~8X=0R}s0(qWW@jUIu!W1;
zV7q|8GBs2;dh%>EPUPC!6x!zihpNwJR!oxM=$d`xXSh=V{iy?<vj7fm^iA<bb+--U
zT>rP8Ss&52t{+0fHXtwliz$D#H2}vt{|C+0FwxUDJ8>I(lI^}-4Fp^HFXp}(7N_UQ
zFkmvGt~Ptv5P(?W#mB5L+EsCUtp6`CxzWq9(Y-EJ1f{SG?4$&*I={P;nm9QD_r%K&
zBzx)SB`)m;7~Mi(?LUz|%l|7J^Z&GwurmI4iPt^a9kC=Gh)*BX58#=Hs5}9JzjRl2
zK&s0Wy`x%4YO-;S*A{N6J;)|_(nl(j71ha@G43(Yi8Wl_=S0eG>DTJg`h1}{EsX*!
z`Fy|l7`p}Ua}>MF4HlyjN!K5roZp|YOLPwYM$-SI{MUaCAN%AwdSJ{PK?n4sbah%x
z?BB@x?A>#N96vgEe<^`_Id(k(?!jMZ;)*_$I2A7WQAg<xkMw=MUk^e>5VU<#(m^_=
zetVQ)$Q2P&)NM$0z~7+l4r7G-+Xj>=<0T@NB(c9w#GtyVO})f61t1lWA}9jj=3dY4
zvA$q@5aCAeB9J2Ji3Mc*3CpIc_<M<;Lu(A@DhN{_)X?5AInwOJP&rNSKWg^JXZsG&
z#)NThNzTyeIob(P)3qlF;tAa=3?W3t9JUW%E80K6Vi&B_X&ABv=lZ=kRD)y1N@Y;`
zkMF%oQHE#SvjY|v`nnq>IAbzB6XVOtdYMMFCnC;vNHB?1+B`$Y>pB>&#r{>gD55r(
zohay(7aj&57K%4NTru5|0QZFJFhImfsmR3TXDdb=SspUQNV#{b&`?=n45;`JElFM)
zPQLXNBdk(a_#clgRgdZ|EuC&{>Uh$)yEEG_5;vFJ1xXLK+aexy?&I)x#8eR%sgNn{
z%xjf|FBX<W&QkJ#nK6M@s~OGQQ~nBePRK=?*iUQ?-|1jREe$`|FcK*W{56vMTTBF+
z`%bIw{qLdrBeRk(waYQ>9PJ@b?86voZ7Me77K(RV?F`i!#rD|C>Dt)ZR30=4PzfXh
z>lffGs*4be7<f$(V^~_w|I7RR>*SVp<vfC-;1%N6OXyxVbWQDN_`;M@a2^$y!oXa4
z9taIW&L4F#2eE=@KRxcW;b5U&P2uVYnqT>ROS-W!Bv=nLxhuy+Z=p-O2HHx|+w)4t
zieEb2u6zF~r*5?nY~PqY>nNXbh1^E}TCFgv_j}T&X##6zY-mGS4*;1jWshZ~Fd)}e
zZlHoUO^?27mYX=eD_~h<RB;T^nBuX@O^`IySTg+vt#R99>)Elb)?NfPizuMN)`QWa
z>`SCnPn7Fe%v$TXoQxytlNQrZK6T;6PEN$wf<uF5yBl;JBe^k{Gs?7);WmzDg?^q-
z(x629yl>z3V#E7+!&BYTZ5Q2jNTDz^GTP9oq9(PsRAh)^c9M%3$(B?_BsT}fp4SSx
zYhA{Ptq-NI>?qXptUD@a%WfJO+!IbwmL%aC4y8h_dgwIKIl^jrh^_n6Y!O(VGp89A
zx;e|W9n*`L{<P&5`nxRQ$%fL;4$Dca>J+klhZfd+yQATI*&XpD9qDd_h@piJrfv8#
z<Tb85QE@oZ+)_#ODnPIg9@~X?r6?-irI?J$eWnBPIgoVzd@8D$fs7juaNbts6u6@Y
zjv%73q;y3eobL?GXmEy^=%qJ3K=H>FA%iSi@MAb2Z0`5+_$Om6tPm32s}R=u^Iak`
z?S@e1FMN>bJ1`oXp_agr?rpkKUw;`~RD?;7ir9{2Bx_N>EKKI{Vm4VlzN%s}C=uzF
zx<os<GJ&{qV_xzm>`IJnFJ2D`WXyy~IuTz1<fJZp(2iQ@#K+bI69MQXKK9BxgH_OW
z)_iC%fS!=eRw@&yKC}wHf<ruL)3G_VhwM{+(2y~d-rU5l8)`S+XLt~MA)nv?wxAnu
z7=s8_PrRAP1fw6lqo>A1k-T1#h4ut1dH);m26(a#lu(UNYvP&;JXMF3`fL7fiAh1m
zmoc$;ETkZ>mxyqpa1AAIrVUw-b@`2jxKnn_XIrObyqcba3e32ZyA-g>P4&?z%sF^^
zt~Ik29qJjG%)*+~4w-)(+BGP|PKnpE^o<Sy=W{|Elv}OAYp;x*VGOn(?WmFD*i?q+
z09JnCK#1K^0h5^<3~R|1#%iTC48P`mhgP?8A_PlfYdxI;UrMBVL_bW#eG}$qZ!yL8
z?DV_VS6$!2g}w&>9p~zq_~%L3|J3sQaqJOZ9rp@>4FUp*?e_wUcko3pA>7MQ;FSz;
zFsaJsha}Fij5YA$O_FF)!~Eh-O+Y7YfIzNpo;w|S4>>%Ql=qZb0!+w>W0vTIX0a42
zlg5ye>%3BAKVaO<El--wIg0HnmNvVo)gau|&Db^y4=njBe76(Ud0S9ON{0QlDWRW7
zvS2J>)3Z*cFSNrPqT;7+%4G1vHz_Ga-6gtl@m;0mqL7tF6`YnZ-0Xp*8FaU}+UoaM
zK7=S_nbj&_J|ND}MYj*htqJep&}EhuC44&r>)R7y#}I|<+-W<5i0v$Y3YBSY%1mBO
zOqH#rNoe%OkRvpadAl$Juddqbst~X$?5uGKmw6uh=r`^MBpZtD#(Zm<ZRtO)I6>?Q
zyI$S0!p1|mb?G!xN)hrW9xiG(FgdjR>yhHcq%%U_j)rkF8~L;yHJGvm#+fHmnd>Jx
zZWgsH1uCqiqeJ5H6U~U-ZC*vL8+R$rQHRUpFT$HVBc~n5feRls$tm5e4RMbG0ZNr|
zAUAP>yj~>&JjZZS2HDlEPxU@PQnmnokOxK_-0Wm2(ph1(d}lFPT&f>OR2|EI`uZpM
zh5`!LfzX8TEWe84z}Z^)k=xMp-(o3uFz-2ed3mECs*?sS+m(tZEUhloXLt{MXq*ED
zH{4W0bT&{!i)i=@;;d7eucI@xo&o`l8zSlt1U&JAphocHF-~v@q0!BFiEWI1!IZ4W
z-l1+oW6i;v&5huC=*ZvxQ~1r?<N)W|2$Sw3P52Ayy*r3&v9k*}LySDQ8=7LJv;A3=
z1m~>-f!+C;@mM?C(UZ%4QDLe_o$hu*9b#L8#&a4S^<^os7A9PJk9sXex>-J%8;+PM
z)7gLsO-X_B!e33W$TS65kZ4DTE(T15l$>B>#1Iy|dAsX@7m|#wI@xeCljLY+uN0-6
zidPb=Uo3kpG;-ji5EBEGlf@W{@l9tA;b!EQw(aEZFjsglEObiXRG3IX%K5-jE?_7;
z`(trhyY&LN`6A)_Wk4a_T}uL_|HKhT=H%Z;M6UUVgkJnzLN7L*8TzM#6xqrJ#xlsv
zKjJkOB0Z=-Bp>TH{V1^Ee{3rqJ-Ij*5X9#q_*4*!Oy2Z+JM;616*QCBRDG3ZK2Re&
z5IU1Pk&b64O!SS9+(bfdQ<V8MzA(W}@GY=`?<qOD`dSpqznL^Y!D)Va9y3uN)Tj>{
z0WG_GT%9@Csjm1!|4K|K5wIkyv*lX^9M{Qm&g<Y{V1zy9<KU+{;S1arK2U&K2v(0p
zmR7XSc9-?2u}GR&g&sYAvr^3dO94QhStkZJzbSFJDv$qBr^n}{Z01N@_W6eI$kLnn
zPl(9+AL0#I8UMR1(-m#$IN~<MfAZh+Z6ksq(|cJ10)fRVntko;4$S{Y{41Y8(<sTd
zQdo=$&S_C%nDk_}GG;`t@2CCxaDDOk{dgi<Il@#n5P9;o26p#r%s=obe<2FYKy2{g
z+xP7)#M$u9c`)u!N9t4K^ZHI3KbMpv`(M68|G@?o1YqyZdGsv)bi~!vt~mdm_-=xc
zFX21H%<&T=+tQ{m3IGH?fbZw|b03ie)zdvn6{@uIKvi5y2w0|w5%b#&L5w1X5=Xf#
z&`A%&mp1+eBCF-lz$Ed&bw*1Je2!9#wXBH%MEt$Cp-dN5Zfh#SU7;M0FIhai!=Wix
zuz$)dRyIqK`|xt}AzFm!4&bR^L3^U`s%b8AGf9x(oa0C3_H4{1qlz@~l%@_|ZwtMf
zc4V}UF?$}z%B)5t;Vdm1Kq^<utVoQtT3J}mrPY$P$hF0^6+??tQ!w$2Ot41l&-v4B
zMms;f8v#yRg|SbvqMbpi9@ipYWEPWhC^0|t1aOHhZF_9KtEv{-^!Hq{8&G&&9asFG
zhkH4~IxU3~cTzx}e`L4oFs~@2$W)Uss<R9weEztQ#-52*3*kpdF=hk`VYZ9a^;$f1
z3(3Xf*V8=7IO)|dk?pZfx+r=x4qzzAL54Pm;+M1P_>3iN!u`6yesE<twgyF(d-f^f
zUWk>7e=6e<1hD&Y3|lnIs^~KXGX$J&K_oL%0mu*_^!GaZdVBf5ypxVuy@rU(`uzQ&
z;FCeg7B~zXmTCHaoZp@<>UdW=k438-jfK8zs()K(qR3H?h*WqS0g-f|>aY42W@RSi
zRr=O)tSasxqWff{B%9}ZN%D>OVhoifQyqquF=1%j1I9T;7xxrqpS;h>LaZzZz1cCh
zr6$eTO`$TF-U%9=U%mG|<7uuyNZtVSK5yWBpNQnrx2ZqF6?@Rwy`R=^+TnkwgK@=#
z2%^95t>PH;K>@SK8!5q~x_C-2GbE=Z%3924UwoQ1=}(>iDs}(?8DQ*WY0*VhcC$;6
z{jkiB>e2T2Lo{Nw`R~Y;Y>7ahVv~%YB4RqFT3!uM_)G}nAso{A@ib#ZW*iETD<tX@
z4~nHzz$Gh(rv7RO+);QUzmq(Tgo^&%+W6nsnO@2CWuJJ`3~UafmO)cc%R}x9><9Ai
zr|+}Mrx=0Z#6osa(9idgwm~r|uuOKaO)J1Si-NWw63*d2AgXK&pu&x?gzvjYgO&w$
zMDi%YaT+p7KTK)?EUp;53Y{ktD73uT82RQE?RVhEba7SnM7?6SF9HMw<$56LL|~%}
zM|tiXW8NASzjE?o-X8xFpse&-X9Z9<>caioPHl!hsNsz~)IwDQMtUqb(HxW+hTn)q
z0@AGTt(;~%)ZLS%Pg0t;f0zQvw)(H+@G!NBwVY1I@s~ih`vw3WX)N;UNU@&*J7!qw
zS+d2ZntHdfyP3Nk2Ho`bP=+vo)kDKtVQ`C`(&o=Mdj5GNRy`tz2+Oq3i!e$$)zPdI
zCHjnGCEFQIx~D`*I#rojr;0QZI^uwJf@>IF{QhwDdh$b64^WgnDZ?8KgmSF_@7P~o
zp2mgXm7sWKf^6fHYB`j@=sO|=Fp=fe?=e}ze4iY_eQ57-Y`rJ&!mgdmJT23IuQ<nL
z_|4ZGTf0Ycxp(lX?ucEgids&u=v6(2ZY-wsu4-GGkMX%&<|eLc1P0fud*_;>dFgal
zqcSg46aq<fwGYlSazEZKdPV>7>aM6yjvdoO_i$p9dwCx=4wpUq)Nx!hQp}Ct5#9`2
zd+ewi)UUekqGco<%H4b=8K+K;MJw)ct~{>fsyq2eufgyhSY9TEcf7Pxie5P1RJK(R
ziA4%~Te!4d*nZqi8V==ZYWV40@{Vhg2G`0^Bg3NHIm;Hl9EX}aCS;bb*<K|XK9(Y=
ztDZa==Td50iktDtQ4-&t*E$t*G_^Riqo#UptmM2-F7;a_W-o$Xlb|GLQC?juJ<$EL
z8=YeV=3MxE-=2qeTkpn<pr&khW%Jw044ye2#q0fsaIO|s!JLS3Ttumy3K@}Im0~A7
z0*9!uM(;;ILxR{q!LCRcqB~g~XXi)Oq|juC3pqh|ONIqTk`%+JA1P^t4RN;5Mcd?T
z7uuGw{e=iC+0|k;R6NI3M*T=>wIkap*qL3#E-_z@w+ggqSr|~YOx%rEjJMb%&5Qno
zOM;r&m8U}Gtv7+g8<}j7t|?edMwH4__A+EE@kYldi^&^>E~c9Yy1Qyyk5*~I21c5$
zVw>pF(fOdC6JX8ZuUR#ArxjoO5#tCd04SCJl+v5KfW)$eimPz?>+8JIqip4i`sf`Y
zLpj1S7{AK-SJ9LI4~Bu<hMlAHgjA;9^KGrmf!II?p9>ppn4mPr7>@dMb$Y3$<8EIk
zO88v$C8df+Z%2V!6P=CSy1TB1ncJ=X`7a(xXIED!?nD^M<L79a1+{b~wpJ};0(O&C
zG}DHoT3FItrIKw0GG8}Srgq?M*Nn>Chin<BmXLrx?#ZKU(ri<y6wN{WV!HmCY-{Ra
zJl7*z;I5Zcb&I*``8<~?>S#lXt#4L(v*x;D4F5+ql>o8hb?ZM4Oe{<{oiM0MMerwM
z`k_$4WRn--?|ke@>B%xcOM_QQqVV23OF8YJs6ve3mqxigT{Yh0U)z<~TQ2UDt+g(1
zH|pP!iUWp-&F~!`YDh_lu_h%h{t9{U3|k#f|Kyy|?$VTvsB}$MF1szA^zoLvGJ(xr
ztpb~gkuOmZ^`RktMpe0d*Q<~R0gh`m5i9(KUk#IxlISc>4V{6ab(OIkx{)_%%Ro}w
zrWIJt@g$8Y71sIL4gQXDT!vSFL}xP;eY2yLhf~Nf7~;B00TQ*9lDPmcKYtx3e5SB#
zQg&qqb>qzSn^>%q)~<fo0#J<@4L4P{7txU~bOd)*up@>Uo7WMOS~m90tiQ`&vVj-l
zRt*awvmB3Td*?>^oD!>b2YOyP$Ve34YC@@7*YB*e80)Rp_e=L5di`LNglwPHYnKAl
z!->iP)IFRJi22a%CYx_@BDw(-gpbyIIIe*Lup`2C`8zU3N=!Ek&+pC(=mF7tyqE98
z>>j84t6y6BO{I`+{qS#Z|IV&pW~gBAThIa8Frxk><|X}eTT6qeJJBeZwfO2l6)WQ^
zY4Pg&2b@Ij&8_sT0lJ<g6(>GN=<$8aj<*4W*RI~~5=tPlvv%TH7Ln2xKD<y$Ie<b*
z?ahej8?5R;XXU#U><)1Z*a#^G6Jg0QDa+O`1OQ=+F;a~2Cq;UvY7UqrDC|eDge!sX
zcbtVmq_Frq@___E3Ch>_=97Ma^aJld;vwh%Emmb^`tQW5J^%2KgdC~&rEa{Zro>i8
z02IDrMM-_8nRKuFChKJBW-~2k=(W+`9|TE)U{96RxqOxiDQG_m9Pt-z7(Y1Y=cCJ;
z$NRnRkwFM?z18chF_0hO8@+@b3yTJjAqxz?e(0WFu}UaE`geY^7GUhbeN5iVcXhuV
z{Ej|g%{&U>y$^6Ud24uuK9<s(0DLcgpVjs%_K~(M{V?VrZ!F%q4d0HqtLOjo`X*UQ
z`uQqn=MdD*>J{N{tjLg-;eV|U`4+(g!YJAuyT|BbuD5U>ZP5XqaTk>r-|H;jcyUp-
z=MwXIbzl)5JOC*kNt~C$MjvrQIzwL80V3{5$KE>Ojwe`&skoFsFMpO2%=mWQzN?OE
z!c=ROWL?sZs5;l;vxQe;${%ENu+L=_(T%5+fE~^;HLM$ByGVE~yIjy<e!OMpykApq
z-tnA^{S^%e(iKS7(Sx2jM%A*_G}Nq<{2RQySMsE1CAEGoh=+XgvU`A3@11q5*<Q3v
z-)ql8r10=m7izzm7_FHYP36VBWVbxXk<>7ShAK{S{74}@-HLzGZ52ZEBVUvQZ68z-
zHYUGQH-|5;Z(f}*9(V=){d7XvEdVRT%^J`Oj0@-gW&V0}k(O)YM5ZKxM@;c~VDJCM
z=|cM>!?z^8N@A&Dilq%cks16)ks#R_3^q?8faHKPH;LW3UIMcRC|hxN2tm~=?Gw0U
zFA&5N)0HtgX)XBa`$2sTkLNdbsv%7CL&y-1_p_Xf&p-7Y;f`{0ni>dV2mqu3QC^Ra
zC`R&+t`>Hv=rO;ett@n1%{+@(fFiCUDR!&0`L2OS)aG%%IJehBi7E#<LL_wL4S)C*
zZPpdV;%M`r$nz%>FJxS2o}+1uRHH27^lMkKXy)*mp`ZHlEyvHK=RoReQ(fxDi(EbG
zF5K<T)n&dS!x6zp`=Otl0*qDJc68wt#f|o($O7cqDV3sIn5a$5*(-u9zm;eoEw`ak
z&qrjCLt94W2^s_MiK@2iAob?iQT;%p&ca_IiVw}NYnav?$2o(o-JSuwvlqkPTMd22
z^=*S69Di&%(=o3w0aCxW2<wgYW}(;`{FWoxoIs^<j@`WJOMB2~9=QioiB*&jFO`u}
zlmgnms;jUrf>TcU@ZhsZ`Nx$qNG>4!tRhAueBO1Ij@PKJyMCc45EfryXbc4xak4PG
z#Wkhfm8YF`Vl4HYCezpYFr&yGn-E!+k+2AU*^qSOAO)QI0B!p0`}Q88QEo*7su4Hj
zJFtTdLVzzjG*W;etLkri(AfXZAp>Dpq2$d0%u@(T-Z6iry)}txSevI$s@|m>_HW?+
z{?WP^0vWLHSO@95{^ISoHwWGP-8ba&>oXZhks|$#QrA<!DrHe>RABv(vmuIfGPOB*
z9hkqsi_Uy@Trr`T&sj}N1MW)cFq={a(7};&bPG)qa}qEGme}`27v>L`%Ij!)wtLmT
zdM}Y74*P-Y3!U}}V-WwH!0?-f?R+wjmr;92!nIIfb8P*PCghI~V$sOk6`5vgSy`@K
zHFdSgV6{>tvtjYw;$GgDxj5c7Z?9_SF_jhB%VveXL!Cx~PnA`1+V&(LZC{!)DG>b<
znlg+@P|LzBQC(KTP7;Q8in@Enq9zh}OxQ%?S{sJEeF-?!*SO!7cQy3a)2w0|f$}eU
z@okkzZ|+E?n5mPq@;m2T9&+gUCXuvjF1j-wsV%?d8ds4S$f#48{5oQ+gO&|CV!O<T
z^C{Lc=i@Snr$)=#vZN6ex+^oysYvC5nzM^vBO^xJ?DkSz`*f=Gzi}h&+O8!xDcne2
zUwI{wt!eA-;N+b$hXrR@q}Mxf8wq&Yb8qf<B$1I7I94ppUDuoQ>aj|B3A@9sggAQ?
z&L(p$hea535e8qgDd<o#&ANe3o2E9-k5qFo`761!E0D^aGv0QowiuQnO*i<08Zy95
z`lkdIkQCfYJpmaeEh^4?=D<I*J{Gej74_p`1LJJ8-7SQs7P4TRd=({t7yr-r(ESI5
z5n&nbV|)H)!XLSFm$0|ncGTy=*}apy*+h=>qYRq&TNF-9$ZWT?#kN+}0@sgrERGLS
zxi~H;ZtzvYC77_9opJ|*9sEY3ll<;xok`OGkMXE`*$#nxsbn=>v~V<IacnH}@Ysq+
z+BH}=xU&@WNahLWDpf&+?8oZ5)(92oA5EyW?Y2l>=DpfZy#%<!I2^gO`!%xKnp%_M
zR87*F=wZt#313%#mECFr+#7l7iJF>avI%N!++Be3n|jsdU*LycW495-WFl^Av^#A$
z!duw5lqZ|MOBo6*!DsgDr^dac);vdA6**Bn(`*7>)O&M-xj?gp9k2G4ncwxE2)-cp
zjA~nIZ|Bgxd#=E6Z=6?aUq+SGRN}P=4&P+WnSx9Yp8RoIkP{Iib4O2isNlFztrq4e
zm)+XeRsP*rty&#%g<CSi4@t1TW~T+2n9@P|dBI09gClS&Yv({(Tq$jL_YN8js#QIU
zuv>&h%r~Q+pU$U8o<=6t;-|LyV94`g<CU4)C4>U^dK(!$!jdOyND-fxBB?;dE`4Ko
zGGHF%n=t2s`VO4&^1^eX7q%$z#`{IZnF?MBD)W<We<0U<W#S3#8Qhx6#So;HN{1cv
zvH4{Ygy1NFwjpP>mTMg*W*8N7)lx;&%cRs>XPp%145v4JM#BufGkh2wxYuj9)=8ln
zZ%-B3mQQsRytn@v6&>2T5{CJeMH%k5byZ8J@CJeVi&bhP{UUCDUXZp7x*@1BC9nqN
zMzOvV#0t!G2Rc71gC2qqkS(3v$F?>L`EB>C)r$Tr7nXPf_B3Z`I?+CNR9x}Od{Sn>
zbLW3B_D?aQMcoo8+_rYxwr$(CZM%27ciXmY+qP}nHg2Eu=fAnh`5wNPm6iE4la)0`
z%^EeT3M*{3@C7g=Yt}Qa)vf=G+mj4Q{HT+6ya|bS1CtO7Ys=e7T_Iwy6>=mtcgts|
z(~JQXRsAJfb)3O(YhiAZQLdhl=cKYP7s`Mo&Wbj;8HhdX&Ckf@1;t-dSH;+pu5vhy
zssV&8C+Rv*k7P|GD9?bj&s)>X@mx{jDCj6lHyn;rKTOfM?Z70r{&539fGoOTig7gQ
zcs^G@(b~fU**y3W0AM$DSQtY8dX_YbUO#D5`Cd1@O#8Ubwb?r)b)r&=z16h@Y7=uy
zWpk*FnR&fl?wrRLyId_v5`~(m^^Td?03esou(S8*DDgs=>Z*jEAxq9fJgB<h2#{;Q
zvQ<JEhIDSm%Y&;HSWvd|*)?&B)vTTI-8FfL-oH0-s#&$_0M}!+3%k~4-Py;t8e0CM
z1`inZ{pZ<AvzKMn>Zy&b0XegwyVky`YAIooo4!f;a-f)zmRB#OK_c1*#@i$@H+Z!F
z-mLU!>G{d`50ZWGKfy#+mj4SASy}1-I}?FlNhx&;?ADB$vnxaF9Lld(Dhycab(>Ky
z*GL0Xr=+2sBib`*58k;({Q1(-0x+>3;7V0?e0aWi7<UNe#DX1}$NTr*=lAWA%-k4M
zN#Fk2*9wyTze8Vvd!Yh3RDN{OU$^^`oSZ`8ft576ctwhC>K|9Hm6;eQ>=%G<=-szD
z#K3PmFP&B##hzE)D=ZT8?pQ5d!-#viwv@w(U!<6zeGUGJFg?@L`E!f^fLwS>`U%r=
zS5tEX-@_M)^c!?wGI@r#gDVJ3fiS)|MrAOX^ZmMV_UoV}R#w%9UE5a3py_r@h&Y`S
zc_$c@Y)mt?OxO%?=SBYco$ib_Xt(71{d&dB{r2&au%4Xbhe1+sSxNuo+Fy`;*0;Q3
zZ8XJt_!G}7K=&N!bHwEtyqb{sG0SmAJ0oG|!BNR(Nz+)UfTX^R1q=CnEp5klR<1Me
zG#^&eLT=+6;3`;mph=e5fS&<-7f)7kA4s<*aQ!M=guHvjejff$p3|V{Ej3yV!@m;W
zWCCF6z%J@`_~Pz5ewIB|>gE(H3P=_IR3U<;AQu(Y;!@`^6pXwg<w;qufj}^jy}MwR
zy>EpbbGF_VHApmbO{p4F7eioK$L3XPJ^O{#t7a&B?_)d1=GBLJb|V<?vQ6LpaCCZY
z;?|n^b-n?N219vn4=;`7GP^cdZ@W`F@*JOosovW!usVY!owRk*HN2)cm|EEL5V@cn
z8EIJ?$;k|iFPo}yKX-pPZrL#iaIZ*bp6VRH;t$w}*q6Cx<+|+|PY~I*((GSDiZY;`
zyf!*Ao5y~1^n_+Y1M9Ihk={TUZS1ML8_+Or&s~j%!Cz~Y5?{vy_)U(~8@{@W7S3?)
zua8G108p*~oR4f!D=O|9KiIzez<#GvaYM2{njQogf!L@HMW&qz=(Sg&M<Rf>iJqVU
zhMA$7b=Qiy1?`YC9yRi!xYuuU;^GzYz6PfSC0htwT}jCv-*hw!8+%ou&;y`n+v26>
z$D<jwkz2=z$s^og?55B9{t3oe2v5Xc|I)lLtKjpn(`~f<ps;F(Q*;=o!5h{$V>y2#
zra|^eI!2D0sj^SVAC98j$sp~jTn~Uqt~xlrpX_#y)Xd*d4%MW8#`U@{&PEy!+DBDh
z?yoXz&6q04e~1i@OGBxMDg^hMhpI)E83OhC<eH2XSXjV0HJf#1b1Q&=H2K*eV(GZ(
zGz>B^bfNl0{G&ed<+a8$GMmWd!gb#<CY1u=HxYe?kRyulT8Ox1JV`IhY9Hny7)VyS
z9pKxy+6YVJK-gK;U}PD6ysL|&fK@*B{iV*ZazyzRnp`;=J76$qn{wz_#|h%=Np=)B
zM{o2rEnEiIaUDZ#g$$Nx$Ab+&R@I6Y<3Ei@Fao>juB0^~7GLw;acnfa&!WHAmW7oQ
z@Wm=uNTeI}y7e&x{W0SZJoaACrj8;a6~_Jw3~vX37Sw%BJw+KDX1x!udYV4A^{oYk
zKthk0?EglAfz?VB+s?{lqd~$l7|hRa;iZWw<Ldri)yhRN6Q4|WE5s2g$wt={B<0!A
ztp#z%H^nX{MZ~I<jN2rQz%y43Nb?exqz_=EpF=keEy+a5hK{A~jWmZalx74sbtVP%
zL8&V{jixIvQamC_cdGT;fhkTP%oDREdbVAH$YKyN<vj;gE_fv_TLfP$NDgq+>RI|V
zJB*_dqutc%(FmSjfNtHX#cl|5u}Q|3+^Yq*{nsQRpMF{myf|5APx>?8*){4>G4f7a
za*D89U2mLh07j`}3k^1FngJOX!h4sHVz+=Q3q8s^PYcazcT|tHY$IzWy-<jjLEw=-
z$17C{#!=eC{m#L7p*Xu;Bh8;vRpj#jwvWbSSLa2}gvvz>&2B6BvyUB;?=E(Ta1<hr
zN(oG#%HK8Lhg0l+Wk$iIC!M(bYBE)@L?}BWO)2m4)I0dWZvC%ROD>($3Z;W_J18u6
z!)@0QE@8}>FHalK>-4Ry_CiORCvgzr78s9KFBM@)zJ|K0*uAmn-5gtnOV8<jX#IKf
zBT#czGFrn_B^CPQ#(AD%Mn${X=@5AMGy&3+tX~;M{HET=vE9MAG+B9~P~J_r(lyh%
z3UEH|3UFNP#JX(7x~#-#R9H}tM)D{@MUzKJ`B)Ez>^|JZb*3&uE=<;n!Q5@tZ7f->
z$Jgkw5w};0zg`o`7S<PEq@ZjjsiB`E7wAy`N_35R@T9I8{v|%?$_ftyltQ?S8+ZS?
zFWw6KQ`RC%gW~@yWG>Zwl}EIKk;aHwNe2-*mRrbk6uC%2y|+mT7+=lwgrM3|Ix77t
zQbpI5{I;zimHFzId2_5ZkECJ#4M!(!kIHNojVto5y81vhKWDovs1szg$jo!>qbewr
zP7~Y6gLd9#ja51;kz@%U(gxyw=bmDk&(r>}v^1j2ssfYx$9bUsMX<F>BA9zYsh}fL
zs#<%g^e;JprTf`}_vPgZCMF8S%$))^RV{}T+qA(==B-2Lp91j0GMOh{$QKsyK$Ht5
zv!c4q2qy~q9tN6sqij$x)>5MRi7mHwh+^auPGtPbx%^qBukygj-cEgpqc>B$kj_EB
z!4kI+%E|DYp!!gfVj{7?U)D6u6n7iTnpcCb*s~Nf_~`el)8YYvz122c#3|IP-XW*d
zalG!thW0F2<|VS4!jR5FGwAdb+OkmwA72Dne}_7;2d3&LUHJwEbh5a;BqHoN|3%$x
zU+Qpf)^zJ5ZxJAy(#pHg;AIFhbGz(gVmeJkC;I{HMW<8Y<3lR!Cg!B^g7uW-uqoF=
zO-+!b+f(W^-4EMC_CqvIue9zBbKNeak)Z*x{D}Nxzt{nsgJySRqO0RLOeKcCOZ~7N
zMt~&*URbEH<m*7nBuNHH2B;l^*3(k5R*JUB!1{}3S(&+n&Aj45?kktNjz~?e`<!0?
z2(>i=)Tci+wjO~60g9Ok>YU~sWBwE2`lsbrU9DIY1VTVpBP)@L?v-(vv313>a4X!;
zlFyC1bG0NPDHzD9$(JaqCk0fTuGYyEnGU*5*rKp5L&lP~(P!MY^SjoPx5O}~AL)+h
ze14NSO02&_rb}PQYqo_a#f{L|GmJvIrKHFWN!e0*#@eW9o34Ui4#9|UW*wV2m8kP>
zBwvts8_!DS;tL&g!1gVSegYMcc}2o;>X}`C_DVjg#BqT-g?=Q&c{7J{Gkwe8kPKhR
z>t{-M2E@7aMBuvkAlFl7`a<SVbH8X#mj5YBWSlG?MDm4EAO2f$7_NQ$bK5HUeX=Il
zYGuVYj*1P<ghvP+MW_}F%1=@J{*<G=_Y+G6avQJ#<CH_?M=V?LQUql{!l?X~K>6E<
z;xCuCYo}AX<;=F#6F}yMsF?ift=fNDi>cNFZXSAkb8<A30vg5IG2C<yK5%e1PMkZa
zlO1g3hmNVLDnYuEpDr~hrMk05vS_)zOag5CzDWTxsGpx!5QMFp`u)HrHx)hBX*d${
zFA`KucV*{G*XA7%etf!&df=lC@C1Kfx!ewY0f}UHKmQYAW&01dR~!ugT{ifY)<itj
z|032?@L*DiTNM@-6{4x6ip~fot`xPfE17G;?ZX#f-i(;*WafDn5?Z>yg$4=u00F&#
z;$s7L;2y8vXRmjUTNr2hK@0jzdj~7Px6ot00*(c9;L!Lm{d)%&dnITTNqZ-w@DgF8
zF-JGid`}ll0SCCfKY*RQS^={^lQUsELg#2BOWkn+$N$T0f_+I{le{r(F02Ztd;qfb
zpFN$#;AI=53P9QQuiP?xzZ@N$?Z_4I59h`%<9y#eQFP*i_{-w-Lc|>3e$a*J<$SX$
z+AJOUEs;dcwJV{CFB34h(9w5Fo)em;t9woDC)jlNkPv{}-M*i^?7#_4#O%<0B0-*v
zKqg)`DsWjLpKc9C%TXiLgu`9lhMSuz;4YLLyUeyU3>Sd7b4?FBN`h8=QpgK55qS!5
zS?j-h8_?7`(|Kw_)08y%!!9?Xsmg`Jd4V?l$si57(MzQ}R?F7QiGnG{)g%guEvgN>
zU?;5cRBj({-USY#o=R`C@qRXq(jD!st{04hH2qqbxxbb3c>_j*5xvb>eX=@*>u(VM
z;rz_9@v?hJK!scNimk$|I|qno?lT<i;-!kEWd5KwuBA-TB_PMqJyWX{U^4(6FjbYT
zres1;QrE?Kpwe3AwvaJeKBa&^2;(0yF!>%8!q~6RkLcIYCH^Yk4``#F7KGwd+<dGT
zsJ=XqmYq86Z-1sLQvYDALPlEGqKGF)-UyhKYO3Ev=mg}oZA0PW1D1__GHI)0*&lKM
z<pK3b(p$tiI<D@u(Z49yKCPG6VCHcZ`u~+t?bKQG*4~BZ_`2q)?&cAa4J9@#Kc`At
z3lO`6^BD6W|N2cu$f$3Di7FLimsGx0#)sQV?L~I^=$1YI+p!vQ+t(&Fat6_};xOt|
zY%ohFi2cqbbQO6IlaxQnS^SVsd1tA6gId{4G;^h5R}y|zmWvyf#&RT$1taBkIlPt%
zFPe)Rnu};nr6j5<3bv-OnDq6`8JpI_Az~j8rRDKox1CS=mao`#39O>%%kE*|lz)3*
zjt#vg?&xZpgzpY*o1l2EieqZ*$mnuJdp|TQJ7Hd#?@JnYAZ{&!P2y$F33n@!J+Mw6
z0{nN_CrHm5`pq-l-&IRvKrGPBv;@4_6hMa=a>dHym-{fe%o%;+U$Fos1{!(`cWCe}
zGIz}`y|H@Vc7W0}>+Krf>+qGNf=@)hKpF8M83!aatgeC64m;_gaG@E;9vbI=*GEX8
z*GY?TVHnr+4Kv$;jN_Z;40w=nU>|;V0;CP?t~-0cOycK&0P!kEe7^ZmCO@w$w+JwM
zn5|gE6?NIQtYXXhQ;Q6>peB#zjhzb^xg1i23=ULc2^E{xzB=X+|M^)FMwWRGz&$tN
zExD#q;GvbkGQLIAB6NU!Ga}L!H~<pv*+A}b0(MA=XvE8owwKg~I*msP{i!d-2YkX_
zAm?a<Y2K0KrZ}K%sMRmo?~l)Q35J<0$tUgMN2AsaF&1z<r&bYEt+3M!6zNUvg#ijC
zbj%#AZa|Lo-!4$E;GewjcJ{&7l;7al<V3_>9JK)^&(}<6k4-ca!;(Iye%?5g2K6gs
z7!XuwPD;&{ZVtefEgX?JmsQaU7L6RM46>!)RWO;|SACL(*{X89_=_up&Vd3xo53o5
zkBJqn8?qM0fNwoiTSs|HwK2PxJwm#sZ7Tod?LvS7(r7!Fq}%lon{h4VdM%{-LcsNM
z7k*#lLV&g@w*Rns*&YDlQuN!>;NeIWBx=o3!c0h3#RE5MBeEH*6)oj5A4l7shlsss
z*3a_OR0)+_9GWLhDL%$YT+^WZ13HW*YI5<s-lbX}w$4<nIsr!&=7qBozCM5Z?|k{i
z>aDrSYnPUG=Q)@${;tVKT;5)$`(AOl#<MC`AC<=0L!soiRwYEJz^XiS7YzTS+^r}Z
zMykzN`C!4NFdlt$c_nfJ%(Z%BSn==NtN<p^G_W5tWHU4p;l@%SVV-1<TwL_bN1qX9
zs5oWZK=V))Mc+`9fpQ-o&bxB)LJ|{OBoax2rO&(yzA<;D{PPEtGp~!K73?qiIaCj}
z=pM*E;f)QZPdAiw5NFh|X73OQ2%pmS<(T@D!&)HAbnC2C^^ETD2~Fy^Rg1hWg*laA
zCTDq2gCih+8LcAWpbhu%)Ou?&4BMK6`N)O{6lWkXU({H3@iN#4KE`x}>D3m|1ZOD(
zmti{DHa$PfszCM>Yh7six|z<uOZ$l9aj51_xet9v@VyT1n%`?aY=P(!%4h~XicL5t
zRW7o9s4~*-z3;5NeFc#XSACjET%_tzRU)Cdip~UumMmOXn$IvpnyyxmwPIyyU%hW~
z9WYcXu-DQFYpP3tpXirqO?%V8TmxIY4`x%~Qx^vXcHKO)=FN=6Zcwu+HVS2{@e%|<
zZOx^{!gUC5C|t!&xvNEA_7;1_tqyL=vfYwtSfIiL;Ii)biXidkE%{98r*+&IR}(+=
z^Etk1;i`W~QB&WO$hbAzCa{=fn(NU$=3x1!h}Kq%f)PjV)VuXDEdu#8hSh9JLoI;O
z-*5?26=xS~%cT)Yhj5OLXB?>dUD0#?5AHL2L>?)RaSz$~{YK4?Om>u*6|(e?Ona~~
z2mPo7XQ<tWJfpEg)hdphDVsMZ<LS$)Y~V!HUU3t2*wz@aRyIqfDGlZfS9v5PRz$!d
zpNV<IxE<Xl&$DnHA!h!1Dv=6`C5sEbuxSd<fV$FVvZ8qG8Q4f|NpFKhwCs#v1@G-H
zIsdpawc}-vKS74vT#@Gy0SSi}s()ycWe}Ak@0M3f#(6{Db%c-eT@`{neG}5FxtSxQ
zMKT;Cv*hYS-y>@^KFy)NhjW(~Rn-W*SZ5OPn)`A=qN@YeaTiQ<T?Mn0``=Bh+CT$m
z?cqd<;R~d`f6W~;Pcn`XYED;KZdKhl8ZDp{f`GY9$0jmK25c!oGLm8}igbWYIsmqj
z`R%&+rGQXmRw&$kEnI%Qq3jj?9#(SW!l$OLUMTNn!VH?LJW+-r=q)d!^o|n<v1pTT
zd21{zJ00aC{^oQ8%p>8vekLB`^c@qMs{kgC`C*7ev>+;X*u=CzpElhF#vs*dOJRUm
z%OSl4sZb7Lwoq8N$u4An3a98iXKTc-X}!9_jIhsnXA*MPHx$fJ0dXV=fNoUy5Y%}3
zDb8v{o87V(>C?9;kX+|el5eVojFHMR*;C8%bAiE~&*|otyhE_fDn*|B=793q!xTRf
z9}!^-ZuP(?u>7p*BR_$~_{qjLZ0!3`rO3P94k-J>_{mRC_>BEIMW}U4{)o8+8Rbaw
zE~Mp#5c+U+2oghguHQ$5swcpz2|_j|TJa1*@;mB=@#)t9X|b^{Yz(pqA2|11S=XM9
zfj)$ZzT}ivy}^-1bab7xZ8hb;Q2EFpo-QE;KB5U*fG-Q4X0MA2+~V$U1>LXB<6xU^
zb+Bm!%J=VSVwzI&9deP{`%NcFzXmGTst!vwDt8?-Vs+Qn$~@*QtD~mGVqOF;tuA63
zn+y{cXg)52s-b|p`R=vJZ5?}f`xasfCJan~VNvqADEKG>Zs@$~M8w>*S(6>*uCG75
z8b<#teHs|lYECt|=Gp-?mbAw+<+HP0cEs^oR+{g-!p3I^z(VnbZ^tL0mLZ(O0g1Ye
zR09jR!Rk>3$rF7F#M|)cXC8lgzCatnH1MGM+!Ku~-A{V2h(s51=5H|_%#qf3*ml8h
zd=_$fQPSrgsqx?SAPX$=%Eh}4(_>C#-rb*M`!lr^;BVgcFT8_2a3uXAT9)?;s25R;
z^0c^JZ1Yk<%hApHXWxm-@ditt05GE2YW?F%PBamB^NEEI`0f7gQSZ9un*Tpxc6NsU
z%Q7>u{&z7AzhgFz53IkTtQ3zx29WW)j0t0rSA4&Wsg2Xc<Kor)&#dnIJ1ARNgt12a
z7A`UsB4_}fsvrX1H~Q=C)x_JwQAUOGKLgah(7lHjz*Df2p5C+k;XVL*8iUtQ&d<l;
zwpr)QGx<y@iI6-esh-zw%D#Ds+#i5#y(l2QZ^@CMYawj3Y^)J5DDjY7t`d8(_7>*w
zg*=Nw);QWI;C8tCualV@U3X0&GuaQNpM~$2ySMulsSf<ZXba^5Iz7ElSQ38;Dlqy?
z3D|+19@w59ubp@tvV%t$p`VcH?{6dX>d#numrVL2O(NA)AhO)0+$N*>o$yr3eo#m}
zlky~QrzUzS5%J=_X608B-l;|cU4A(FdUYKFR4$)cMhWOCM$%K1F$|GI=r4{j3<;oc
z;#}IZ_MT#@Y_W07)Pv%pNeQP=@!U#j?KVQ#y_LZa&d!KcPGi|ISJ5DgWToQ|*+>bE
zQT#?mX^DigJn7QR_(g7)I8y}&nr}97r4fe4X0@p<8W3z7;a;LWk9fQ|x*>QX{6M4W
z)nv0H#2NRM@u2z1KRz@`6As@97OCeHZi;)-9*%EvsYUJg1F+vsCeDowe(T%=;&Twe
z`3s=0j1_PF=BAMrVxB*2ny0~SRYt1<!+ggjuE}w&Ttz%K*{Y6H2puA^sB?8#;*f{6
zAdmSH0z$6ST})zSX@n><xjx9@9(;Jr1VPN!WjN{><RkoL$@_psz=?;$)?yC2X8^E6
z%xilGTL^SIvy~HTy(+HtuxcxCT?%AvCcziakS>({*;#^85EO*pH(N)5x<RUuk7Wla
z$8VFi1-Mr_F1r_<1I}@CbBc%1wgVET+9d)6;<}2bu}Go>t(jb2oSuIU9s793buBGp
zYE#npi7t!sS`F!7dF2{Xz&O-fQr^RU=5;Ya)Gmu~aW0D>PCYM+id#+TU}ZNhkL3H)
z63pabp60Z>2_1=(PK0A47?VaabA2Sr>?KctWvBJDYU6(ST@jz`C+1U-g~GESTx#OC
zt%PLORMjZf(Gu9<d89@aCd19wMGKd&K*h;+z&VD3CtnNc%)&!yu7`LAhaqIakrFBS
zOr%iKE7tpMY6wRzJBY|xD^jg`CO#-nh?e!WWa>C(4B?R-ujveuO6rEqPdvnGGu;Gx
z9?sMp$xoemT*L+5d0x@=u_`tux-hOQ7t>p>Ls=G%8`oqTt|bKpOv(|ra*u*}1)oz{
z70JHVD|z-<5D7`#qDGt{4-F~Fi&>X+0)y={$#<1zNClf$fxNUPTN`9~P2!4tsAZF)
zAwdPyYi$>}Q?&4%wfmex5GkjPlpmNNm8=Qhv_WD}tm6jf?FHrJ%HTjLX5aTT_@4Qx
z8}r(jUt5kmy%CSs+mqXZ8oFwP;;c?GYw<r6hdcO_;C<KxxMQ-!^6m04tJ3?j)8s!&
ze}+x!iETq@w&d}2MuU6BNF}(HB<YuKixcrnj_TmJlVMKFTban;$j1H^k&%}cHmLid
zQ)H=%RybZe7VK6m7KCI1R4O|lPuaq#84Gn_1<4^s6>nCMTNQ0q%+QVGn?ba>0oAA6
zp>QK#mEmw>BIr#q3un$J=IhtC338u5rg%2h%r#x}X5XjnCJnQSG^;pV4%>*}#^)yN
zOWemIPn}0OMUfcTz(%%iMhw|ztJNYj-Z9-vsK2gO<qZHqQ?pBR=DT?;Dnh^DcVPLW
z828y#EhWT)PHi?)WG~etu8j;*S-qM%yA{+1(3X7h?kCJ1)mgjNY+s9^!RP<(cjHtN
z4Jo%ImntjYS&nC2==#J16dH>u`b6Kr!i_sA-jt<kP|z+S_rpW4Pk5y8fFl+C%nr2%
z-tFW0%d5H)&nVV+`n^eic7+%7L;arjWv@Qt2Sr;M=5?kI)Zf3Js@V}HW$UwMm(Zxy
z2aYzJDb*4;s7qRXR}|azp(Eg5sQ(iOJH$xfrXoT00*T6^pQ8dYof}ZIrva6pKxEPk
zAPhB<sI?#rUhwL5QWXK%BnP)*{krsiH)R%?l`R>$luCOuHhN%$Yb{LHoMW|}BDyjD
zIi~9ROGto3HK_K#QQe|?JA663ye0h4ltABTmH=myBX5E8VXccZ<WdS6jGaG!1*cn#
z(_6G#M`6?R5(y$+g4&Q@iZGKfVAr{HvwXo*Z(JW<gFqxLDPM*<D3b=$_%DhE64B~~
zF(&r5mdZ}I1@M_UF5rM>3`tNJTzH~yqO4c0S_2zeXk=WK<Peun<Nc1?5*pE14}~tk
zy|n^vVxP3%0<=z(Nob2?zFJb;X;iM!+U~O~1epgPCj6^x8UtK=ms^{ouc_dS+u1tI
z?b?T5gZagWMrD#`X(|X}bg)|VFr1kn#sHpNVcm}>AalLa8ry(X7=deVwm?isC)M3Q
zJ0NR0J9#;GPuUV$6PO&kG|Nwrd;8Rn16OS`wiova7(6cuomS-OL$u{`6gcvvTqf@H
ztZCIRLrh^EcfK9S&+SMZO1_aaig3}M?8H=ahmh^W(p$xKuc|)bC%{~?3bTo?i?GE)
zh5LN*_hycSCD0^5iP<bf$J-qI$WAO0Y@-q<K!_BqH+Kik!gW0|F>B1!ynIqVZ7h`m
zb$jM;GHwT?;-tyCKfKXm&Dc?<Fu8ri3A1v<vbX0tO>|9~-s4gh7Ps?wMX{B6z-4JP
z!w_8^`#HA9g)7clCihEjONwiY40*pPVM-3KsTHZ@rdQ>@;Tjzw)NBdSax~;Xq0h<R
z+vcg!-?Sj=1>lFx8qnHN*PYXNk?+>BY7LD0X+nlqg*Ga6I##RiFtg27BaWaOq%?;}
zX8Lx*SIkWhV@y#jRA*#|YOLb$EF#50tD?<T3>$G=$Y&a=n>rjRWu`gW%NgIPMyN*z
zC{XN&Oh3Z^<%|-HNk@rk4AeJ9zz!GVt|HLaW&7rRpj7@a=I%9}7>PBgUY1v#(l%N8
z->#|(4XkidZ9D{VFG!f8Jkoe?I7nCJv-Aat99C!jz15jW{hQ-<0a5qFpy&`JxBO9<
zl0JVv<a~7k{ys-dKR%ldc~53nG+pJ*oz&fHp`Wt?LzTMIa=do=`0LVBuU>X%J)8X~
zpv!X>|7<RkVLS)Js}B-PiWXz{YUw*x_6@y{v-s=QVr%T=>}X<O^FMEPhL%uF4DA0`
z7RtcJ^xs)XZD~ryZ2ls#nVMU?P|EmxTz>lAYhyG2&Gc)U(K4IA7I0HVzCDT56Sy-%
z&8FHQ9PXEO#*&}ncO+-mHL`9s9VabYbYhHIO4)q|FV9WD7PI%zn!1|T*nagOTN^Fc
z51v(X<4@EnbLrx=WHY+goxWnbMc|`+eqUT>_|JUfF0rnsA4NBkJHq})f)*Z<G|1a0
zx|+09JXIHGl6>@czmqf{tqW<Ov0T#EbaoxSu1=)t1^2YhJ-!=5U_!a0M&Q#F_#56o
z4Ik|vO|j_WR!0X3pWV8-7FjB|I)!g<5!c_b!!h$|0XN?=Iz34a&M$5+88Grhmr1gL
z5(w}u5EpPNw*(0|67--&GU{@Mdy~|Z7j{aZRX{Uiq&yG!e~YV3c;}er=!_B9D+3T_
z{-PGy3EQbNkDZrFNIWxZ-_eX~9XRsMNhPBab+LCR{D?j#>b}!O;dal`4Jgeo#?31g
zwWLMEF3_=i$2B?z?1Yh`jx`FZ$HDw_gun_5*08_{Hz!;K$_lggjh%S0Z@}^k1Jc)q
z5Pz7Y3PrEV91<hy4{t)+7q!3z2>Ux?*!Nln{*Q=06)=3=dGar$2A%iZ8N5`JU9pwf
z)h9D|9`BsGUZZOod%8hIDDc1gbXG+0xcR)-QYUIqtDD^2dgO>>4z*f<nob5R1AlT}
z*qPd0NK{NwoH`j)m+~N*QC1>(0oHfZ87u;-V+8E2#d?bRm^-!Yv!zqk0(&eSIMtwy
zKQlO|mQ@IohFvydDK|nbI6+YD-SB>_0x8}IYP#IZoHT)HY8s(ROnfMr+IUgc{}LXc
z{^<lE?S%dBl6AO3v;P>U$RXa9eMX74K)}bXJ{I{ZrQ=@lR=S|U%8=2OQ;ojiJG}vx
zGz^uJU@T&Kv$^mCGu5Ti%|weqP9$S9YP7DC$%MT{wl65!W!EihEB+V`mDCZhcKV54
zpUBI%6P7jr3RZiqt|+xT3Eb2})t)|zrl*{ZB}<seQLDJ>MzP~&=XFairDf)%{Up1Q
zOG6C+OwtrOODe)o+l!EizPZu1Q90hwruNAx&<7`ANM6ZAP$IgSba&C6C$1&Th_)Mo
z=DWMe<R+T}vylzWOmmD%eSWq>RCXdt+*Io@ec9Q1_jE+2Rd;ude<2iIG>g8YN8!XP
zVgaE8PD7mZ=<B_Uv0-KMxMJHr5Z=9FoIP!M)TXZ45AAX|Y#b2qgC>Kd)oz|LcR{;x
zRm6V%nJ^e4Z6W{@gGgGI5M96Xj0+1xq$c0XO#*~YbqnOb_gHjpEP>@TKsH!|l3L+B
z;vgp@cC3wqj&bpnG+kvsG6^vBX{g(M1Z;k-udP8)yW6`bHzYRb)lH)V)dY1`CaBm%
z`vgilP}SYD?jf8M9p53FEaSvW-nH^dNrkP`@b|o|a$Qf+B{(HepBt6gYuq7q8N)Tr
zP~)uco=iLjxfT#9iw)Rejm@!LS7+N3o>lCJr)slW@u6HL+DyD*!nZ*e6gAG(!qx0d
zbJgMmJNUP)Jb5~1LXj)eIg;b6;gapL#&$TticF3TlSazOjL50%e}-{bmn<FG)-UH#
zp<=79&H{k1V(YFj0vf<EEm!E3jM6XEdrQQq)6(39kfvB0(Q~}7@g7;Hywo$4>nG|5
zI*9n*YN~bVgMuQ+_0AcWz_3)9W9#AI$?_Q1;=>iGbrQ$D9)JxU*hn3g&GORc`nZHc
zDTc0F^!0pfXLG&0Hxlq*YfovD`s*GLxDf=DMmS_}y=W))G#wt}qLa$?8UE>bo3S}|
zLPQ_r#3uQb63c@x<WCJPIl48b0?3(&AD)zj$Q}GSdB)1cB}&y2o1rekV|KD(I=6+c
zhHWwMMnknK*3<43m@E!hNM<-BeIX*Bb)-imkQ_L<OErSG<GUJ?qP!f#o;XzDE6u$u
zVUz_OcVq%S_s0tt>~?~n<|m1OrerC>+p4Sn9yh16Rv++ciu@puv#ZUcF0d)mV+hlt
z9m6DCXCe>VFZN|o^?g~H{kuw7jXHu>CV>7t&`*UdktzW$0`Tkv0gtb_;FW6pUNN3J
zGQUTlFh%d0O#p6feqFN9Gm^;_x~j2JrJz05nCK^vRl`(V{d1Zc0ztH{YlaK}f*q>P
zy_|b-B~D8?_pCTV)0_kVf5~cxn4Lq*Oj`hGDATt24jexa(j(t<T)im(Wviuh-x_PL
zWmu5)>z!O;)9<me1%S9>=L|uBezZHnC%X*OXJhP*_hB7SNp|$QS7MW4rp<x44vv`t
zuqWT<r-k!nk^QEz&d~7#P6A@v`kyKy$N#N}jLh^L|CIu+No&IPuo>~!RPhvi(jZR7
z3m{5#6L*8(HQ_=hR)l_xIO1+drhfEhXJ08%lcwxw83411yk6Z!y)wmo?Q3rA`pMph
z%j@sls>U8f<@t8EOD`Sq4}3fptjn+Z4R&k8^X1_;Kk}9HAxP13D&=S=`o{V1?ElP3
zAoZU05e9zKdGXwNihu35uCpoPcO&<BMA|j-9qY!^4e&(CIJe>Hx90eGeB9wTK=itq
z|3-a=^uQSrL<;Y9UUhfl^+PY>H!5Y30PeQ;_}+BmbSpyz)g#k>?*Hh46KBQx7czLo
z5c__Mxc1yl;Qgw&t3TLWz2UctZVx+dwtQ2&tp(hAbES+z!E%8i<CDWY^9CwFIC-*h
zBbCL@s1NlJodfa=9Pv3V=!oCpZ`<(s7~JQti1|2*jl)0Dl@B;*_{RMJK&Xv}2up)#
zS0uGrAA0<t^8t!`;Azc4Ej|6AxBjF|ebVjbHvzzN?d3=DUywda#ei;w4w(9&*~?3k
zSGo2J+~IgF!Lv?}4yUY|BTYnC!XvN0V$2|UN)jIZiS>M7;;9n0ZZL)ptpit7y$M9m
z#dku|cV_Elg>D8w-$@$qXqR+X8<nkfr3xdN4#zWv!<mYs>DGvbi0N|D)^f~3I;@&g
z%7fN`(<!f8q^+WAZ{^I<?aaik%&etewx@sZoU8;X4Gu?4B%<bOucW4ZBG{=AKE?7O
z%F18vJiI1-V8r|P$3hjkGJaG#)g@Y+^fR7zxeL7Nv2H~}@r^O6m^WD0eG(>zh_c8e
z{t9r$BbYu}8Apr^#l)%2F4v)Y)7WRO7<JZ6WqV{UCT)U#!6^-t>G4b^OO2#WNeUY$
znZH?laqb%AITZ~~siw7|fw}%hBCWiNM($rp=b%HIO152SNM)uearvXd?=;)w0h4G!
z8@}bh&^ggf=yoai^#qJTbKm;tV9Y(IQMxgN6}IMqw7PGSN`9aM1R4E7I)yQ8M7`n%
z>cnNYVn%bI%d*dxlhn0V%j5y*mp^GX?k;;#6vYOAZY|R#WF#MHlQpTUh=wPN#KLk}
zx=pOsFnxwk|DQn@7@ujUw8q9ZmW^;-=k%>!1qw100S8rtN)_3mdus-Td6FVZuZnsF
zue2_Ol}3`kX$c*QhXH#i=$9>|TVFWTZz30RHA5A;nABuJfLzDUdKVhp{VY`FnMdL~
zattVF(2Phy6iMs5PQgJBv3|fix?*w(ZKRxZ(oonN+Ddm?awIF*3OYzN^zVT<5=5Ql
zJ`{`<SjQ&azSGB13JE9M%qTPv2;(Jr$wwHoNKQxVvF5o2<)>~vG2zp6lgu=azhkHh
zkn`_A$~4ryRghk!7LEGVErWy)duZ_|keg!tH5n3>hcH>U_t$L{Vs+N2vB+f7-mRcz
zl}L{w8QS|E5(etjm^?+SI4zp(%q*<UZgB=P7gUrb(iOHOlxp|OjAO3m$@nK*g}XxI
z_Ecd>?=!|%+-psBubEMqsp|Flv+h*D8!R@w)_8FaVk$`Omb*%&>B@j6D?e!7o4H3N
zE7>Mfm5fb)R8(sWAQ@jopMBDYyKYXt4E41PFFpQEr2lXmIj1nf*(A{`Gr_r3*J*@V
zD{VWH!05|CXVyfPN>dN3@QOoOjesCpV)x{Z`&)mrT#QTAVBQ!bAqx3El7v$cE`KxZ
zTM;%uR97vE_PN*wWg;5L{7v6&51y}v_dZ}0mAxbnD>tQ83MI;aq?g^SmErp<VooBb
zT^13?B3p>9&vzu>33-GiNz%=p(?~sry8z6AH7h65U_Hknd;iZ~nq_``8+_5d3wB1a
zesI`JI+e)Dz!2lSCK~0tDMHgXv^Z+DdB_wn{pYT^7K{d@Lu@!r9~Lg>`(>mBulpM+
zAmj_Fy*|zm9^4vzXMyOCRb^uGl;L03%QtLB=u4r-zGxBPX2fQ=N?5%PU`KL@^n&y5
zCr|SyARNjinCRj;#}Uq79kx3JvXxcZjpv4X>v>5iqY7)kQf~Z6uo$kMXwZm#(@YUn
z%zFcu*9Dv&W4Uo;UEdM=P;e~T>Tk-HvH-q1EKPlpM8k=4FO!LKr%6%k{Q9^S)|x5%
zKkT?b#b}Uk6|)M)lw{HC`N2l0pn2u%gJ&xCNaLgA?P|#VO@OFcbR9ZlXQckGo0=Ps
zO?q}(XIwuay~ER_7m|zWutrG<>v_VG>moPE`SId;j|jN0za4F&AnxsSIs>xt7kOxx
z6PEPsKE(Yf`jKBYzyXD)_D~RNnUZmU>d?wf+)#4q`JkLdFze4j5HjpXknfD^Ta0EE
zp#HcIt~;v)NdwUrOlDSTWejq2T4|>a;XH8Jfu#bJzr{(g$kvc5U42Ku=VkRdn)i}q
zf<ZkbZl(+EnLN&AJ^JX9W2Tk%a9y~9U*sS#&b@A4t}Hc~#2Yh-1Bp|u>t=QUm1m|S
zKPtN#dDBs3w2(aySp$qJHa}Hc<bq0~T9n5lN=a&Zmr%Ez?&_Imoze1g57SsGjPq&m
z-AX<JeNx4FSSJekI8xJ&xA^yw8?W$T_vx!MPt*~z8wAjD$i+j((dcn%l!4L~TaVv+
zX0Yd83jU#D4Rb@`LeT^Zw`yInN*)cUUi_hE&93@q{w#+0j4eTZEKQIVu~BYF)PXrL
z7c3`f-YisG6eA07lq6}Rn=<6jeOt28rr-R5pq=$e@B|Wkg(TMogSI->EKuZ{h0Q82
z{!&+klt|>fTJSr;II3}Jl|3~<ecA)4a3LbY*xG?qmmyIUF8H!^5*__oebS7<>S2_8
z_m*<icc<zP?aT_y&WaPMP2Zo0<Qh6&+YuWdp56w=Jz2BH-U9qCXVHt}2g?l)-QIo&
z`BPWUlp$T!*et#jl>EkkpfzJ9Et<`MOH&ZkSlAF0H_mL}JOBt=)iF)|n}gbQBO|Nd
zyuf~-<~{tvt+PdJbcEnf%c?CnT5p8SDRS!E2nv98@GbfNZ6Cy%Cvqx<b?PY~NzNd`
zkKDn@knp)<V0SG0b?O-)SxQjyk6q-LqOuOpjVW$A{Xfwvmg|tvyI%5VK}xgqdp0Wu
z^=U7l(#BAT&wav}({XpWz6`ZlGahf=jz=A#c+-{Q>kl`etWJFJ;Zn5jX!6)lF{U$o
zsH>A-+~ln8(>nNdn(l5ZJxHsg7yXd7STii%hX5~8$g0^pa67<@I?<7l*FWHyI`<m?
z0k$#Hv;Lo!5(6W{e>b?$q%jtK$O_Z7qW1c0h)xn8?<Welq~+HJTDGAtn<h<rCUnf$
zgJ*fCSS`|A?j=7_6%?P1i)WE?{e9;A@^O9C`f)X2c#=qFU(EG=%Fo&O2}%<sodqm!
zoyX<ls+ZHmN_`{(Woc7|{L5lkuJdW!G6Eg`vKR?Mp$B}0osxWq4i?+=tQz<fqJQDa
zMuJZmC8Z27#%MQ^9PqDtKkP>6xOAd(Rd!#|E8cu>Mz(xA-uzP|ffl}=+9IyZ1vKn9
ziL(6Lb9~+STzw~_WasYj%`HsxksA4-16LWqO^nn;BM!BR!3hK)K|whLt9J^Y@9-$-
z+IXTxQZ^Li7i!Kj&8TA4RGU>)Lz&2@d%2+x!nIWgwMV{JRW;7;^8_qY7gHFPO6gyj
z<f{<hA9J5khcg~5IOt~3q%jrO`I+CIGOoL&PS7x{O_KC1PETqhufMOpBU@A%E|0eJ
z4dl1iwSryDlyCD=ab#j)C5_bW4ELt@QMfv7REI&*H@e28<%28H#XzmW_~4Ly4=)RX
z!+YK*3l6hPA9qeKZj+@RHKjY(#n?&&i^SPHgM1V?MbEEt)XDs3@5Y3|M+67{6d={Z
zp0fS6hWQw07IX&Htqk+%#HX^gy!JNFswa-#`Uh1R`ZjC)RcgIj)Z;@<=)cnwN!sc~
zgyDm16U)Mw<2px~^2i+jEOi`39{R2GDo%~o6V}J5G1aSTbVoE+$WU*lF+j1{6_^sw
z`2#;MN_hCwfEsp5ORm4Db8vwK?OwMhHIbqsDokSD#>B>hjuHJ;^-IWYa8o!_f&xK;
zkB$>se-!)G<T%|O<+16Q<_*4S68nt#bgPDQ$}0~cL#w%7hz}LF1#vfAhML<Mc1ltr
zi?R@|kTp;ocO9;}@`4K~Fs`5Q@`z6HDw1mNqqi37KP{rhB|+9I*hV12jJ8eyn6-7s
zH{=&?4*r`ZQz(f8yF1-DDh88~;D|7ora=F<%E=*iM+kv6$}v1za1;NqM2dPXEyWv+
zHO4VcxD4lL`>{fN+AxUZ-{L(S+BIWOy0j~n$urk|lE!|R`tCi-`W03rWWp@F5_|?D
z$%Gn|;R<8769rq^=(e#L1uLt;);V8%xt#)2**Q#ThBG-jTiU83yzZteFsS7*_7Wo+
zh-^VNtRcvhGrCre?V|s*q4+(u>bfi!@@?_XyVT(2A-EW=b;O$)*D_N(=PmvziV3v9
zksQ$EM*qLW6OUT0z613t1k^O1;j#QI2ox-lNUMZm39&U`Gm0Nnb)iV@q-d{XGQ~&F
zSRF-eOxXYMPP?oH@&am?8TxF3Xe0vK)9bGGXR`Jg0;PUdj^7Qit?^;?{vi`a57|`e
z{L^gBef{P`)^?0G;h;4zIa8iB`uR@7b+DwI<Ru}8p5i2$>~ozn8z*c8Vy`mOb^&T|
zW&3z5i)MdluRdya_d|AcSbK-Kj=0sl{)bF@(d!#JGVg+=jZVt(0_V^UC2K?sIsQ_$
z-$hny@%NYT;C=hs!`&ibH|zXPLhmi(sRN;Y(MRd70GG4@AX*v#R{>BE0=Vo|4`q0Y
zvh!kfZizI}Jtxv&ekJ6?oU!x+>@r*i{GZOjUsF(e{Qu#0xw+~7|4yOIY%KrX_4Sv~
zC1Qj7A40e8WLA@>bftg8Z({+jcBF5U)iX;ZfH(vcb#z8glJJSLO<OirsO2i2ToD-$
zuPnagC-d+uL@u6pr^ow!V=6d-v7hs1>ulpCq5n_VPtqeHyQiZfG?ByU%0kbyq9qwg
z=K-F$1?;!SC+_V+cNZoEyU<XJukQyL&1E1QfKO9xldd1Q?kc&q@ByArKkQHSBM(Bu
zIDSKu39I(_%Wnd5j@4uVK-H0>5<(X}E2ACC9lQsHyXO#m+Ox#v!Z&5ZBTv|Y`HmVY
z2U8LdxD305_Bv=t4y~m3JKgzTp8F?o<i%b{llt+CqTFq$iUB6cd8qRE?MBWW7367V
zcFLWJcFYx~W0^K#f3C!pda8PqWSB4zzibks>ohzNT`a#N3C;;bA&stE@MPNOo^l<X
zh|0GI=(*^>Vw4^0RV20s+ctCRN3H}`-SL4pE`xvI$3E(`G<Xz3d{Vr4msu_(;)U!K
z2Ch4luv8nrY`bSEUYgM!CY>d*Ivh^tw-*<}h8!EJtgBv@$a~la?iYER+pL++q1&?S
zPyfzYfOm;lVFWpwvtb5S>E^IyG>;BWXJ&?OgK!}lRf-Q%LY&zY`@DTI#kG=zDZbK2
z6Q_&czQltR00qV!QaSk}g`y{7d@5T-Mz=y;1G=eo?j5m_&Pj=w(>2~a-QH&nK}pp`
zqEdzxMfEOF5ywk>#)mlcm&J|6m`{8J_#-2PAeD1=4}fLLqtB4ChKDZa!P-c$V`IV+
zh~%LLTRmEc83o0=ry~Z2vz!y-DIF?FcHt;^w)9TW<+w#ni|6&>hoS1C>ivC7lUNqD
zjSYh#O2I1%1`mJfw?Z%LzL_v>>YeONzM3}BAY+Z^VT24ZJ0csyT@7^VZ>1q>`i-w@
zuF780>cubF5`XIP?g}DrjCPxL9vc!ymdS{l=q{i0VbGQ#{(1c@@A6gXm;?}tT}UAO
z{(e*8P~+wPd4Ebxi27!9Ppi^>?lqHE#Qb<+$YB7KNe5V-kc0OEgEc*N>@}NBXDWz3
ze+l6=it?3Y#kzOkrpSm8OtTP!znllQ72Artiq;7bj{&3Q!>OO}KcwiuCQosF!}8gE
zAsSwzW5p`jC<F**9X8;6Y{F-NLD<=e<VK>FIOC6wrQ+*x9txiy>4lx1>~(N^gxI9T
zX4*_H>OD1Rv8&B6wcSgl-5K9}seUDFUBlOzi4&lb`0GA0@-mp4eu|o#E=nrD#oSMy
z>I_S~an+5Y4#R973j=bF`vp=ECCNhc3YOE*XbaZcRv%dj9FJS9*uRvPEy}kYiL=Z2
z8vXA=E_S9u5@SQwNdp6Ywh@yR0*mL@M3I>pg6Cfu=4|DVR8@jfUfvs;{9g8nc|6^s
zLaDN|s7AT^m&r^{g>b5M?<(b7N_~y-mHA#xEKAY~ys=KOTz8U)iE4bval;VB^A4w=
znQRZ-stZ}4X-MG*d?t$jXO;^no=Ua4jf)(SsL`FlWaXc?P$(%&zqIlQ#EwOX9h>oj
znLFW2i>qKVPiBkC2eSErDvh?Mg|I|Zh~w#8xm5dGI&>>ZBIAcgD=V79$G|!7h98^=
zin;H*F<LE*-9HGUQ`>Vtx?L^*<mNQD((i+{aWctdFU*SyP<M^=qRqk&D>yHuqM7!q
zcNST`%}J>j#qI^OJ$#vQ$1z0)^l5Ugw!A~Y@A4%b46Z9qD^5-;;78tzJ$6jpFnbVM
zF7ZMP3*pA3_Hx}Qi)L*<gUbgHQ2KW?1i1DHxhGciZR0G9Ej_+vle6-Hj>h@WP?USs
z%Lk|@;!5=?%l{Hmnp*`1us;-3c5=uG7jzeOQuQGen1QD4BiPlPZ)Q&FUW^WZwWk>{
ztFV4q7MlVX1hM_Ya=pJ-ZkBb`eNrtmtJNq33Lxd*vg2|!=!5m7=?tI8z(005^elnJ
zKQLG9K$Mk%+>dwP<|JdAYGsVY;OwiUI8NS5#m*LT^B?PKv>+ZRARHm`Q$H=v%{yEt
zN636P|JZ&-*Nopt=CQg6S?DLzpWdQX@$Q$OLuq3%4io(6XrlYZ@|-V)$<H&i>}YGA
zBwa|&9vd`n8U+=xzg3&0BGqN9-%99b+DzjJKO*HL&7|ei6=7<KfH2?%0H3d8308Ex
zFPuaeId?QprDD!upknW1MgdE>_e2INu9Fb8m|jDf>!uT3m5|mt-FUANNC5C8nvXX!
z7$uN_5472J;SclgU9hC*qt(?7?up+<EjLkbU`yf>dW-XipE*BG?D`*+8ZAw_a^j*C
zV0w1@1M&uh@?u9uJ>rm$rpc|j7$a<pUT+Yb!Xt)$9<-VZV2H~e?)Fz=yjX|(_4okG
zrVb`pf5a@cke=l*hKq025rY{(3osG8t}d7O9)(tUUZaZDt+p8MY2sA`F-QEB<{RGn
zmKwNqOyV`AD2!6AUi(b+2Fg?B#v_QHU7;Gp39VdEUgOEf<u&abSa#aiE<~ivuROcE
z{#YpriJUG7VsUaZ%s*b3R>roV2}QbpnF*w1buU|ruR@=eW7Qt#CCW~ScnJ$**6h4F
zS#zHazY6%D+#1X7=X-GaA!gf6ws1tjEOAD@;)YGUtiwS6;+xJV!5SZ@ulahLbMo>*
zkK#KrpO#0rq>QaO+oV!QAUDTz70(PLm9sp~-^z|K$d%>Svrntw%io`<<*pE5f`pnl
zT>2)_g{`5k_0D|xz}n)y<7FNwoTi#MOD&@_QdQR!?@k)$iqcoq2Q6=YshQBEm{4vC
zNQc8_IKF9Y-sq}7r6Hg9B`&C{#nIp6n=@kh4Mr&+$8`?C6qJNvx~^*8C3*flWVVZG
zKO~{;`6Hb;7H8-Kwcz3GB6Y*Wy2l{jc(K1p!h+7(ywAL66_!WGq;Q_Q%@Mfog+=R|
zg+-cbchDdOHW9lE-ZV(Z+&jM8EgQ*nq!gl|&)XKY<N0Ky#OUUqP*HW)<%M8)mYQ3r
zqjm{=o0dv2sMlB*@f41Kom>OA`%%!7r8_W5rLLC&zZgLod@1E+H?0!VB2eWf*CDT>
z$!G0&w|)@(F>;lCcEZgO>oY_4S>2}018QId5uU{*)*@Q7Hl!j7=8A2^wLx+Y(5Ow~
z<F4yJ<kI{epAO(0?RT4PVc_v@mXs^&Br{l!THWSDyES9TFu#kqioFZWY4U6%JC+Ju
zvcNN&A*&pT3oP>mF5{)L8sAfOYzU<fBj3F`7_a!5#g1So!gpbB6T63*9@$n!p(`44
zGUIk`vCRUiYY97r!k45(yimSw?{~hPmo4o70Q~4V{y&;ZWe<B3Iypm2C1)EtS$qb1
zdOA@HM<-`|4tD1M7ZlR7vHn+}a7)w14r}DkceP#}-h$8SMoL=N7C!y%pbKU$D9i+C
zI+P&)#7ZEG5Y%Cr{kM1a8WD-2n21+G?y}!i+eX{gTAJCmD~1m0fwVIT`VC^J^suKw
zerm7FXj*JRcb~hyH&HOpwmX{4y-3I^@i?cjXgwEjU$J|Q9h51lqAfTjco8=$<P*;F
zlLQ*=*y}ER$cP#5G95RT9A%&zkYu>x+N-c57!;FPi6cL{9uSCXbG$R8oV%k7;^<9i
z5+ah<v!p*;s3Ks2y7mr<y!H;L^_CblAcj!66bL*eX;Ckk#tAe=dYq&|x1tIngWleI
zw19w$dr3}RwZN*BI45SEJb54s;b0M^(;x`NqDC&b0ScP6qB8IwJ(Hs+kZ46`E8<|~
zDs&S0aTsrm|A(=22ofdQwr!oVZQHhO+qP}nwr$%wW!tuG*E{e3=DUsC%SL7=BO`Lg
zUTcm4>C%<VZ?X*vBDCy``$;4@0frHg6bUwna4Lx+gm_`joP5cO&U9Ms8bDAktP|5@
zpjMNBogWXm(qCNt0fxe#5NimAUHKFNja2nS&&s3%S){$3R=8hW{uuag_k((GQe_&V
z3Z<@6a<~{*;6NI(Qijf0A8l=j3bblighH`oaHzioGVflXG9G9kX~{+$RHRd9Dx>iX
zk>8Wj1uLE~r4vRPzcx~_8{r%5c1^q`HcU$#PVGv$+!lNzc}E1Yw}utQzf)XZr6LKu
zO8E|PTt&d&&y^BS-m1|OJsC_z4Jg_%+y#}DLl92EB`1PPBp=!05Bkjb_&VDzGoEjk
zZI)Y6-sVZydb3s8w%|AO;N~vd?dI^6sqZqaXWfdLpB-O}*zQcPHaj;doQ72U(5sU-
z1TDWnhpt>%6DPs}NUmNFXID7^(9`jCVM+x@-}~D6;_k@eZ0Wlz>+jOc<<U*!yVax6
z<95g`-cFm^b_m=&Bz1Va5m0_`&nEaHFm0=RueC0gwBL?5-zUA*PI_HF6<QAQ%n@TY
z67Ex(+FOxew?ano#OEyDFGx^D$<&i6W5?En|DGXJHluMPhQV#@BEMVG`q!7I^Kxc6
z9(sB4<;MF`@XtM+Ugzy>m%=LMZl61pjZsY7-M^fzrsffqAB2<o^XDa$jz$a^feLMx
zuixnoK9xv=Xzc+jJS<_24**+tP^yLh(U!N1)2=J(WIL;EI_yQNZR=PSAejJQf&gG{
z&JWC=b4Bcz<ZoGkaKJ4Y>7>G-M<YT3hOIS15?H*xAZ)lGKd5p{_MshuLol>4NUscW
zE4vKl4k$3FVFPzOK*KZv3osylhwzZD8zh1!3qHKoG!%GPDgp>Rb4@s&C+KmaWaI&&
z5GpK#wY#VWm@nYnBI_%;&?~=xb0#V7X)!i6K|JqF#2jQBoZP;nmFz?P@AWxv`d#m*
zxrc-Qr+cppUX4xVs<}lI@dWDYIxNc7@EeJ8JbaC**o?|I0hiVyDtyq!>4P<p$Mogq
zdR^zeH6-}=;WymJYn5K-jpx0~;urUgZ^yfNooOHIZ*E)O$>HqvX@@V27611O9}@oM
z6>qOwJQL7@xC)iPI1tD_fIwH4Eh}CeX_P5rZflW;6HBHvZ)`0u9d5jh+%6O@rm(!|
zN(;sv%(_;XKFn+A-e7gJc~vkg^5trdEL9VqwAdwdo9<CzjD+hR@YHpv5wFk=CpVZ!
zUYKmN7_6q!Fl~W#lGQ&1QaNR=uo0HzhCE&+2J*qAbX;~|n0+&?5cZj4$53vk_CzB}
z!7M;oSaMSvuW>mXuQ;9snzfDn^V-2#H7jMD6JeH?a=l#bS&qM(vSzeU&8SZEO&Z}&
zsfeSS1;d94YqWvLkkJAmr=(Pf-6A0-TNLWgamjE>epPdOTYgY8d&iH(ye!(G?boiu
z>t=)_5m@p_JStA65C@V%9_Q+68OL}%J849D1kr^fa`_+FV-q5=ktgxkXdyJnzZwn?
zjE2h!0omqzg2=Fr1R72SNX{pJEiRL!5$y~(Y;LM?$BAn~#%}G*BD;iFU}{Z(1n<Bi
z<ZQ2$L_}~|p{`&VDC$_CnlBK_v-5{#&CL`z7&R}TvB17&)6+p{SWmyLz1oq(iPItO
zYxDm`dqPHq2bOA0n(x~v#v7d~?WJqXx*nRgOvkBga#t60<af~=dvMJwsqkoriYIL+
zw3eoKN7>ohf6wptv&v4xCFW_*ey;9+Qu5sMpHtu#J_*P5yTllj42N$pNWSrjdoy0?
z58)w_kZg;LH%BH`P(wJ3F2#nkrOtWOS&BV?D+troxoSgnR^q&(I?HqYaQ6P3FHV~~
zHV!Z#t{T0>&jlCh6VtH@jtby=Qo!L<eqp$u#?aFlr0KW(ZQ<6KmkKqg-x$Ei>2gba
zZ<LC?3Y$@ebtwIjEpsCsCo3oHSgJ**z#6Gb%TcA%eun2-pdSu9V(A#n4KxjuENnOU
zEJ4MK&9H)Yd*xau-9P1d<{)G|F;Ja#3}^VfEZIq~>cZUJ{lqa_R-elt<N}dg4oSeK
z77j?c*A?kBQ=O2mJ-X^~BSAKh8%>KFvQ{9dGu9Vly}bSfVzggCOpRDm3$NLpSKi`Q
z{lg;yu!e?V5+3ac&iEwzfr?Qz$)%?XPeJAneh^o~2QHra?90})tOq$%Y__RoI|un8
zLz!684sdOMfW(Y$Zr1+O?GOLj^W#0(xN{HIqZl<yEW$He*T->Qk`L^DPBSK?i2v`C
z$bK57Ps}h{Yd?G;;X?$Ptp6}Q<9mDPOPmXodQGV9b6x8jTiKQAcJG7rwU5R8qCG7q
zm@Cz*L!7yd3rnH+4=IF^=|8lw*;xKdQ4Q9(Emg<;P2C(EH*DGgC_13^yNCZ4;~mWy
zMf$h}Y{oBNu{i4Z!9uDr6~|a(*SO&|*5t#}Z4&qQ$J59C(_~jb653zNIypSQY~<ef
zHzf)0v#W6`l5B;tI6k>w_f$o$ukHh|VrEc2TtB!sn_0eS$aX=IukuGYILKc{NWM<*
zOSdoWvse3{gdH<>k2*in@Ybt2y7!m}mz1Hn!4!azQYZVIfm6!t-}n1AG6~}4ThTC9
z!F6{>Twp^f=Wa=^E6b`FR9p;(+8Jd)MHuJQIaXkwM8G?wXqK`*_?eF<HqdoRgB~q{
zyqF<7{$6$BdBPBU9nZjKjU(i@Ct`;cLe!Pi7Jo<oJdf{K*qY`!#FItFVE)}qfx)Id
zo}8rOc>UA1%IYaQx~I0PH2~&rWC0NffxvX%HWbZ3;D&2Yn%n)JIjDxK>#{@H9A=`r
zNR(Npl{n|XlI;w>T>3?9u0H!&Jg2$^`GVlkiYj@`VBxe$O?kH-=SX{<#Gyh7S*Qq2
z1rq!SqTa`X2UK18E}KL0dFE>HV5LU%ENY@lXb0*+9|FpN8OLEv8se<^+aWQTD!&yZ
zER>kf^BA&e>4&06^gMI2sV+v4S}U!6S1%z)@%TN(?tWpPl^M?i8P5KC3whHS^^GJJ
zm;RlYu0tEdHA$=*?Q>hZ;2_@kSVb>7WuH`mJi_T3UMux)@;ns4sd$%XxI}X3DMX1x
zfbp>NN|jjZup^zC$1yJqDnsp=={uc}ZEK{;9<r>)g!Nt`kN3`t%3rFgN)<J^`wdx$
z;-I>to*lNG2K`q_vTkUzOVIQ1uz)(U_xYrhEGD$fx*P&)7Zs3#g7eDSW_ui!V>{nO
zg|A}7$ENM_ES@w-Huu`27JjVSabiDCr?;xl6Nb_`t<Z!XXn0PMQp2?t4ylOG%ut^w
z*xJdXNS00X;Q2Pq$M#GuCd0i__6#Dr#X*8XaxaxfDzfdFywmS#C(d40EiL?i1AeYp
zlDNU<GSN_}u#H{UjP`NRvu{Kg%n%u(ONwdGCSfe#Q6hV<$kgkxMy)GmqHPhPXZ)Cq
z9jv$RCK0S`dRk^VwT5tz))!AAAWzS|B*s-Xjzr<A{^;)vVpK>w$m&Y5aZ&IO1KPD0
zZ1tV?KF6#V_R_M@yIO506fu`371zm9B?_~G#rEcpOo`bRwb|`DQ(0}%$=kU^ygK=f
z$Jxi2nk5D<uw`rSv6|%FjY2&}a@%vEmtsxa(g;4|nxisbkG6ee5X)xgu8CQ|aPv&r
z)K1!E9rVEju5oDerp|_9L>lE$XLH%$fz#nKnOdb<leM#tI$#&W4JfIXNGQLG2ref>
znu2Os3vk+;b&`?+v@v{&XqUhHZBdG8&6b~`C90Uo;1kox7nqFSxMjou?vu_oNMv{o
zYvD{wnCZPicKjyto`2gOzCu9Mbru<F>$hW^TAI6$?6t?qH2$1R$#eNI{5g*`E%Dl0
zNoWmdHjlO3s+p>?#q2p`A*btx1?$#j#+a4I#Vr`GF?Y;S<EE)N*$J%53IdReQb!yG
z91obbP3ohTS{K5W>cUNhe!0s+6!fY1<0JbmsvFCxrW7`jRoRAxCMggGxmp@GZA5In
zGgewC6`9cMoAdMWe*cswf&k#dS9)6J_x+x%>#s1N|L4Rw7ToXK_v7&VmT75Y15uGw
zNrvyP&AM$cX}7!xjij=}d1m8p1!eUKbvC}a0Rvu*!fAiUe1Wke6>|^<RBvYfc^e>u
z+a+|v8Rv$Y!;55l0c^F4((JA60Oni^2G_tM8tA06dR#x&@dZB*;v+)geu}2czeCJ7
zA>>-OI^2L?@cH&!HCD3YQteruSX`tzi$S7u3<AtqN9R5#1Q=RVJ%P2eob?i3(96E?
z-L4P1GBwlkHgO>8`FxF4n9N709y&&lV68+3r1<4pEzV~9`gV_%Wi0KUK1~ap;PmAT
zuGFkY-*Bt_hfRt&5jmPe>9z(l|5Ku!CSAQ_zk(0W-&5ehVsV6JZ(pgC8)$Ay_X47+
z05OoSpI=~Hqn_|qtlevB4%Pv3<m!073T}Qkr1vdnd!m8(G={vzF_~sKAlV`FxXlk@
zVsY$BKE}gPGNGnraOMw;#?(*N1SLYl6^<I#tjiGLVT15496p-)d4m%69QU~=Ni=x}
z4eO12qZFH6&km;{`zbbhP3F+99R`5utKXg|<{qh`+Z*jO<Yv#fWJ8$ROj9_yc7ICJ
z4ajpZ2xIlt46>`2#RmZnvSh7%oP#KdmZZq96SsOb{A9?u0pDO(pqI2;13D?X6eB$^
z8oMk<GF+uSg})5|dHA|z{bRXQX=R+ZUAy_t$n#+sKp=-cFFf!JXaRd@;a+HkO*XXH
z=ueliO*O;z08GsCdSxS``v?otG2fiRXu!Ux8NXkg!ngIdS{%72h|(Y4IF8H%qak?@
zHhEOyH1@};S+{8=3arB$SkM2=^VMxtxZ8fVjgF7}D3TP((DouBAC}Y>)0Mb-`uM0|
zth~vH00{lNlK(v4#k~2(u9cjY^I*ue!R8QZK^K5{O42YNi-_m3+!PLkGyKa_8(fA^
zBne2uHtJX_G%f~lME>#1-C1n*G*ReXSQ(jpcbx6cCNmfBp~xq&Wr)@Nvx1e}Q%o$W
z1?d%ZTss!xlPUYv+#4ymI*=R{)w8#YP3odU&AV1$poTb8K`lnWAK=&M1UpqYhc^p1
z#-OTNWnr21N<q~)ddt9IfP;&ID`c7c1-3g1B~T(46$^8z%z+m*+$&|DcX=Fhz|e0e
zl2pgOGUx%kf=Hs_ua*PtpPnz0nh5JSj`(U@%!TFyXIPS*kuI-j-N1>qsLD=)MXAN6
z_sEB}r6V6wDh&z{8-bI>MS9|lrU(ACYsS+%u%;a1Dg@w6E~mY9l|*x5<rm=hx+*ws
zc$IsXZqG7Qe{4(ec$eEqArBF9z3RzD^W)%$|J|Vw-32G&;E_yD`te%$rYhDODc9wV
zr%KUejaYmlcX6Y1x>7U}2`Y$=8Eis%Udl=&punKQ`8!B^xLOOTcuG(1<E<i}bzCE0
zr(vYKqQXMKfv(%(%99^G4U<y2=ZYO~w1}J;xBc8O&=2-4gR5H0wQ^(G16uE$FV`2F
zX7JtFuRX*0YN5=fuC+qpEVSuCAXC5+?DDfbHzj+WHNR!vmpN0Tre@ukompmPNzc-;
z`LCZEa3expg{q$BWSj+I9T+LI-;$-xM{nAZ$p$s^_0YZJ0MMt|Moo0D#XC%H9T#zu
z{VNw#I+2CH36qB<hS&_st5b6A+>&QD0I7=fJBh_aC7JPus!d{$k~c|xCgA*}L1HO3
zsUvuqxm!R^r>o^cNt4beRBDx@N%|BtOe#%qw30LO<rl>udpHY>3*0j=vAX_Z=(%Ly
zDm{$l@am+!>~L*_ggdJin#Obb1D>0cX7bjn@qxH1Iydfa<hu`_{xkN*14rg2NmHb|
zK8Ivn(#v3HP|aN?o?jFzZE|GIl#e<H6?3MoJ&;{Hu`$X96l_Fb%++M$z4&dzbZTBs
z-FVGlj|%T^%nxp7ol+Z8SKGid;AyyR%O2}YWK5DmDHo#KaDsH&!szMETMejhRmiT-
z0fjTTkF!3TjB_7G9R!)r@0a@r9X}-3!GDUx%=G_vBFoD0-$mkKZK=4U*8dgr%;BkL
z8s9@u{{w1fu<mKGIx%@SHuW0A-IMt7i9{9;Eo3Q<0+}z-DBwjs(?(kFe6Qc%oThD_
z{u^q^=a|SsQFYDU=JaD>|3by0#oScGPx+tm`Z@Rc9&2NA+q#p_{JE=tTi@5{^Zm15
zA9Cj&zDlPG@dXM9x1M(l6zBDErP~+$i9p6{)kiA4%}20%fOFS@GCHU)Rl*<UTc0Ce
z(6_qF>*uj=A)f4Ogd#-A^zHv-CA=!$v{J`7mw%TzaC^W*C<EVY5O_Hduxa&ge;mOW
z-63C^B5>VLmgYbQyQOhIvQGZ4DJP#Gt@%36zc}706OR20fkG!S2Y99*uq693rp(`i
z4Z|+Jz7vF#@j4)IOo$uI5YQ#TFri$<<Eo%#p{S}`|Lq5(WFo~M++yn0p!-50ernF{
zNcO;`eWeZZYtatqcgaRtB*BXW!bVXm<y~hUeKgTRrpSZ%JyS-)oHoAfCZCwuO`4q}
zVTduz^jVdHG#O9phVr$L7gQ?BryqaS9M5vePL$!jLb6X}-Y~5s-m~F7<BUw%ZtyFi
z?#MTYhP=Wt%d5_0;AuDP<@koj2epSxe+SqMUY0M(4lj@HH-lc%akkoJ`<73400t!R
zkE5Q%cryLA-tR?k-&fE8soNmNs98Gf@7<!FpX<>?+mQ*6n1qv2b8UC^9=Y!_21UB4
ziFl);Z*fJaWAB3-A5V6=h$?Ogp9pMun{^Bk2N+a}#7f-KV~{4@&wkUF?q!xRYat*O
zc2sLntEokKf5aCZV|nL;%~8`|{8JE-#BR)c0^#OArxg0isAHqvZ{k4<J4`{6@#}6@
zxX1#WeB^-S%3S%NpgS_~$J|!XAI!EEm&CUaKsiRDT?70nPf&`=3ZodA6*UR7^vrSP
zqS3gcHdWy*Q*Q@mMJgnAtD`qB2rFa8!AKXEEA$0LZ9DM+B63=V;|3qZBV>~>*9zep
zAI0&1%@iL_0yWtCPLC?sF=wJhFoB&5-_|ZB+As>x2d(ct5nqHKJ#I8{i}+#C(JM~T
zvqO4tP*Md0k0L2(#KVT+POn6<S6+_A=ev+<GF(XJ(agS-@I1*6Fu+1Oiwpwa)LA2`
zTR$T5ZSEYl@N0~8FQ~+h{2{oGKR&<AvrQmLv0*gV5KVhbb!FPe`7*_VRQrS5u)Z?P
z+0XgVN%(G}gH06Aa5^2pRXk4^$hzeZ#-|1(Ca9*P@dfo%Mw+a|W<iDL)TC*voWF&~
zShb{QpiJu>3%~Q=`Bl9F-+J?lB}y(=Vo22pNd>qJ94RVm8_f)fia$l8CR-aSFvH|i
zhTO?u%b8lJ0jN*&&vL95yRyBO&DDVE{ym&Ky<DTE<I`#w_~I}?M=sqv1nNf`ko3f-
z+Wn$jLF}bo;ibFU{%t+fD~YivF^7CC{XiifDyg<!Ce5KOX;)RbGpbRx%v>sZ)YiEo
z3F=Qf$<?GCp8S=}UC!;Ox#4gefBJ}uTaWwBp3F@uqjo2#DfM0wZ6qiS{VNT9*-{63
za>=2(-CnXM=U}Q<$ALx|W4G=HBEigV{ueni%s&^L6|o6l!lMb<TG1Up+B%#mKOI;i
zK=_lspW&D^vQ%z}dXQUmUqv)+Bwf-{HiJF9BSuV>(ODbqVa#ydg3VnSlgl}Petriu
z#wFLnC1(INEwDlDtn4ir(_^w?sfI2cMCyH)moZWZ-*_w2sW0c-G>E9it|*Ozw1aE&
z=B5!QUYB<U>yH9ni!DfSSNAkBH~r=Eh`FQHUgDoHiQTBJY3CSoE4hCB<b3ae_qPwh
z5jANqM0$GA8sEwsv^lgymGjcEG}-(=m>lj61vFyL0D|{bq$@r0jQry4=31OMd5CP#
z>H}&Z_|tcauvPp<cOg)X1|$s8t4<BeS=vdvVx)cUv$y1Ej0e3x_~nln4T$Pf5Zmzx
zU`-2yT`{bq$y9uSt`{;O6SLxx?xmbED4-(VdZ)bmapJZnsTl%Ue>Xh^&ya;64nQIR
z1+*#6kUk|YL%V615?{J_TUy7)TeldNaW9#fiAj*}yjM~dVJY99n}Z@N2(~34er9}?
zLgln}J?YdeFOtq>^QiR9fH)4<%zzfSLJ@o|fn?TUgF)a0ZKF-iuQA3VMrfJbNH;n7
zwI-y*?a=vW+U@%R2RPC($G^gZjt7NP&;nW6uVA)71hAsbCO)EqSOs$6Hu7;w9}nE;
zKut1O9}4qj=4in;fbRqIWe`?_SdV8t=RgOc!#;-3IYL*$)qt-t9jMG#Yi5CM7H&o7
zbPtX?@_QtOq~=Kd(Liko-(vD*hMWn_Z!g)vYzRw{7VAbBxM`K&xCa~L_8+f?Jwno8
zqxj3}lv}f{v}<ihlGaH6DWbAhSiupHUR<$IfdaKuTS*xN?&^~ovS*YA2p=#crs1fk
zo8?)lk0=zGiq5K9C8z?@ZOeDqRHKLWTbpUE3czAF$ivu7>8ar+CyO5q*B*2y1@kzD
z{StzfQ#$BXUKfzmQ3K!_IsCKF(sUWEuGpK8TAVze5bbR5TN4*cZaeirUx)1(((Z*V
z3ro)!eBLKEGFzK}{O2ZLd-A87R-eu{Rfu92P^3Ce+TG3J6Og7H*s2V@-I3}^Vd)Pv
z9?B!QDB&KSYE0avSh0>BQeV#ow(5ZlKXJSXm*RHj+I_{smA7~$Kv@#RO!#wqb1+C#
zhOKtSokMWU@g^c&@Wj)M<HK&!YS0rV7~lpmS$M8g=ht)Q(11647IbhBsY(0iI%!FY
zzz5q`<ktQ&^p%M0X++O7d70LiJ_yd!FbU#oB^))IkP<<2J2!BJ<(0(`NI<{-d05?b
ze+Aq51lbI2QU8sh@k_RKf4#lQGmx4AtraRItbLaD=m4GKq7IOs_gwoU7!>EE6Gq^w
z7#b1?svagnbC-EALL(w&Nt7<UN<{5>YJgc{eb^Oe^mQlWo3o7>e`DN-05hnZm7o~a
zt|3_5!{YkXu}hrx5j9K36b_kr_^fpvVZD1<M{C0%J)7PgV<T<Ri^Ue(=^W~dWFaTB
zwMFods?r!?4X<qiAVQL9(pG?!K511&J4o}S8?ALVY`Ce9?iCA56^FqbZC3rj*>tty
z$&nTl-Fr4<o^;_ZC-JM=eYWgVJ=h?#F{Z_s){XM2Yi+vAZQfe(>O2}v_~^67NHH?X
zMe3!l1Gd~P(^hU%Y1~B4Dh#b&WAR9g02CTsfb88FE+%_qhrkl4&SB7t3vb2r$u>mU
z(EaivVK^*2$@Tq<X!T|hF!s?V!0)4t5tXmp+GYUc7Hit{(*BCRUW2n<vqQ0GDqsz+
z4p4f0^C9oZU|W!8<a3)_9EY8BW^jXVEToWkbw>mwfriGSBFo}33I`1CaEw6J=yNgV
zGc(c1nVEC3Mj7gI{=tjWqugnb6gzS-Y(*;SM7_zy(^PMkG|g!6G}eG$mVZBkr4o5=
zCq+Du_xe07t$Errh$X6Pt-(5#FLYX|m%%>yi9#FaYeYCc9^+eO*e2X)a&{QonGtij
zz~Gbki9A`~AB;BJ63xni-5hO7Dnm*z_U`xDS~&P0W9@%XlC#qPcg^u*?aA27c7*@e
zlcOCZjH1h*=il`430OZabqmsvVQd=LF=l=LM$RiVUr%0NkOtBiYZPe|Ug0oSxi{eb
z=Dk1PTs;3iw?m6aGEfabSp2j@k_Rl4+k29)8vkFCkLNGL_t(@vl26W=;7m#J<eZz7
z+w+4+U<vl#SzsC+Xn?OozSK}x`<#pH?)4;E9_~?a;|Yg^azn5?48_8+b~ddH2SUli
zZGkuc?Edp{+G}2}yCvcrPUYw7YuPV%OAMmszoGi~$Kg0xJ}_xbI()pw(0QDnOdrp8
zPxx3NN7E0MkFVBPb99n)4a;^++W*=LKgdXagfmK-7X5@)@OZh*WIN7x7x~0_>lEtw
z$?a=WH1qj4&?FP>M~mem$=z$H77_}%5@GKi9z>k#<U~4Eg25=lh<x&YYrQMoV@>9c
z^$1{$e#k>=osuhONYZ3)9}-<8RVs!!c-R<V^e>+esJxioWAsn9Op;A{2^UbHp`25k
zqcdOs5Oac48N#%I5mZx4EN~HJ-@Q}C-w;NlK0N}J!&>zPQ4#_r_R&?Ar`M+jV}&jl
zE4Ll~f!kM@^UFEjtDq+xzq6~D>n%=RO%=*jAs8-o#CZHO7n-v9M@AA9))iC+|1e*o
z-PHU_4{8I)oqkY}_{|xk?l2!))dNhDAU*yd|A!1&X^+?F_tS|VOVt*P*{&seOUAb1
zcE(Z4hXdjd_n5QOICuxh_!KlHM;)t{>1=LWdT@jSLD%D~psLd15+7;=DKP}~D%E%Z
zUMucug8P`<yx1M}!4M}rwhs&KJ!HthYhp8b#lmX`Td^<Vo2O)7-L5Z@n#I5u<pt{|
zJ9qT9k1RQs&xMe6xMQG2at6|pp3@?Mmt5<Mbz0gf*l(t9RW$z_2%F=0G(R1hiqMBO
zcZl4b`2LW|o|Dr{Me^}QfPoEmCoGAy^FUldaEbVaVF9qOrDV6SKVo}@Vmlp4Xu+kt
zpcvc_x+MBcG`Rs!xtv6VcFH192m_j`nvirF&fN6z;h`n)##p=oF&DG~gEo68!30Cm
zWHZ2-+&KK_S$+a{Mxu<3S=ibP>^nGDsKQk}kP)yWG}j#C<m3}zy4-lI(!_=6NZg-L
zY!+W=qIH9C;nJs@f~6TlDVq>D*S&OcNr>rcn*Gjhj%5fnr5r;9mlBOcY+V(BuMKHq
z!qoM#AOjd@W+(>ZwH$D+Nt);5`Mm8E2tg;%iFkYCnPihP_{0lz<}?T_te{!=Ek^=|
z(<Y`3wg<aGk9MEeXHf$$`w!T2fwOKCawmjFuJgNlkgT?KwOxH@TU}0=jWVmjWTEKo
zp|BBUyS1S(hb>NT`uSn5P&ie_TFE&IX`XY9Qr8QJl4MNJS&oL&E!l(}y>*KouJ?wj
z;VMW@H!mSnJd56WB6P9ZL2#VyP9t;!;6vGEsE==n8$+jUYXu-8)#|HX7NXFEOL{>J
z)BF4qv!#Ci8L3&?{Y)?3QiVN4vWmrYhZb=KRMuAeBXHT~aGcQyFmMKF?J#(Bo?|2M
zf-wlD6U*>D3tc`ArN1<7kDTf<nJ!EX;}P;PDX*_iCc`$*QzdtbrH-bwXxdrW0=Z_4
zSERhGroMx$ywHa$`$5W6<3VTqE@=8zd5oX2!;0FLeuQ=}iH;4+;w4f;pE`a`n0T(I
zLcllRlF)Ys$ZfDV!q9myu_UjlS{G0Z*6TgJ|Nq6;@Rs&DOcM``QIDmHoGcOPm~^>V
z6-1@9RcPz3w6vM3>h_{jeI2yynFtQsUI7hGbi*Y&&oA--6Q3)uF?mvp*Db?mIZmOz
z%7IvDXW+zJJM+f<87CAX5aZu)nmmG+7u-XaAB4OHFey}Whn5#*T(Sf&2YP9b1rU9B
zN34$_7-%=-@~=2+#Q{o3m3H{n@Jw}=qUglb{ii`VI+(`Vm)<$$|HV@E0+Veneh(t#
zt?;K{&a3&ohO<iibSJ}E@dZ<{SQ9YF5}rwJ5sI>~*eHGB101Qd`arXo;N7d(#KwD-
zXRV!#scZy_51)BsfQJ}uw#S0?Ih#;Hes0|}VXIaX8Z@9xh_iS{kg-~i@5@AAd{n5c
zEe&&7m}e2r^SmXux9ceB<?c)ywGYBF)#r6Tg<H=-Y$UDI63UHBP%%4T@#3CQoNa0_
z0B?_c1~l7xZgujHiAG-;@_~Um{T>K9u-<28stmS&z6r=5v|dvMt*kH6P+QEk(6N~w
z3eKQUh?O<u1U=8oDZ-AE74IABaEohRV?T+xEGu%u+U_!XvPEI46*h6Oa-ZD(<V?lH
zBrzK{P%ghE6(lj)EIwn#KVd3&tV#?^nOA6xkX>bvWF4!fhU&8%6a_%0QuX9bkza?=
zkx3os7yKGQd36ABL9U}wsyr-&I<ZFWQfv_ud9qA}Bh@0k?aFwsXS4%uD@y?qD-Ttc
zGi%o7+N2>y@WKT!x$wSVb{jck_)uDX@pw7Ezl@q2ucMlkwQxsvV305LsQnrkVP_#7
zTP$!x)B#|Wk3wopx2k+Rcj}`)+^t$zPdd(2nJKxtQk*sk0@KV{t|ZE8Z#{PE^e(Jo
z>G=6+-RPv|Yj-Otovzemm}>1NX|u0=<q6IgAh!<hjDmMYptO6zs^aE5aL1gz`HWGk
z(53K9-&|kPQmqPNufqm}8`C4020_ABg=y;bH8Tg%EhyI=Q0Umek@Rqqx_o1@ap`c4
zD}U2wyrvwkJJ0nU#b(>6y=6>L`=XSth;uLXriyy-xZt%tkQAz1G;lB2Y&%)K;3F_M
zR@DC#1LueBMdbWN5F~1eHqM@iH_0Ao<MYkSB->rYKYHsdC=`(27Y+^n+kM*9M62+-
zJa|2r7yCz7s<kaN#Q5XwYjct38Y%l#J88t+81Q<HyCl16XpW834_Jpf*ZY?hDiN7S
zVLB}kSg-k-AV9oyhvfkur`zh6yFfjvRAqwEz*I>d?3;t~IF3zaT)zwR9N^MLA=JWA
zCv%JBW%`qC{_xRVr`lM5OGSsC=IFwsw4eg5B`1K%W?buwe~=g%yZ%AfBHpuoC1X;m
z0!FKrxqYfelbPT;Ib^fujyUHPCiEu})=M;raVDhXNQ-8^ADhj_)<Y$3EwTPnYbEr3
z{{zAi7RKX!8q3aJ#s99zK&!{gSD;b94Vf0XLae12E>W1J&Q;@I-MGwKtK{e#Y(riu
z0#bhDY&hwbO6bqK2D=>?lH>M=vev-LpC#z0l;*7GX(gS+z7xHJ4hVo{^!f$H=95aD
zL!esz>1;7S3J2J*6^g?+#{|1w%OfChhsd8QA5A=AONyE}<bHQDD@$0sK_LL4Yt0HC
zZsO2iQS;3ns#o&dqRYx2JoBpgsrcg78uz}M6^2_Y78#tf+@dju9YoTjQXQD8E2bsE
zxtM<AT`ag!z>un&RvE~>n&cw2vljd`y_M_}Xh|os*LS~#C0p75m^=RipMjO-zvD9u
zV|656wKDA1?ZHt8kHl1}X|*(kRVHifNSBPYCclBG9_+dWS;q@Mf{vg14`RTs#q`1f
zG`=?_OwY>2)s^M(wpet7qjp>N{+I*&#=k)z;&E=e1MD+1>+NRk_DWP>>%w(MCu<%a
z-+T3U`FO0r+XD-C?;WpW_3L5YDQ@fd96XEe4o*6K^o1<;UiKjfZ}t@EZu&V-6%p&g
zieVPn3<Mnl<P9Dkzt-vH5+@LgIG<Z}-s|D%$^Cj90AmuvBQXEI{Ura;V+O-kD$s_b
z7rbBUP2<V=9qNyMLHU0r&1h)d&7vb19%>_VelIr63bJV!*en0uujavcJzUNXa@U}8
zg2F-|6MX^*gAq#d1(;tFczOfI3MJoy3>h&0tGn@T<a??kcMJTT9ChFop(>n5s+49?
zk;F8hwzluZgQ27VGyCfxi0od_uS<?wQhir`aCwj*SgL085f9wp&KO!ABd^I2uIalp
zJ5L@!Yg7hF0cBVyfElZm)`)tZRX@VdL%<4YcNqHiF+T$ifmD-Xcg`sr<2LYFVKV8#
zMT}+w#EHJzMgDY$%EQ^IwzPA@cy<i)p(*9|9pIK-m@ryKUgCeou@XW-i4mL@8?*Az
z63esRD7j@8T*Hsfw!c55+D3C(HcNhb5U!{oj7RDeo72=xQd(kO96e}$LMqP)CpSfh
z1s}!@+y&&Vlg%e;)J^KmyT8?gs9FIr1mc$j6SWSxbZ!tnAgkVrtY!#LZp+(ASZ`ht
zoa(8h0mlza0Vti`zSqo>ucp`^(=lLWRTA>7hS{i^4rAL;idUTyZ$QYu5WLev8lSGr
zbr>@5YnALeke_2VE!kQ#e%R}>)85l}$Wx4$V<tPhSv>AGtAyAR{HF;3s$n!ff77}m
z_&mv(lPYa>j3PfSHDEZYqquP>5J}iBW|?1}mgM2#cwJ)ohoSD86C9~jpZC#=)s>(;
z#}eO^+fiJd6gRs5#&!dWSc0`?4G=^w9j=u9(2Uyti{K?W-Sftrt^ZnO6lPcE<E9cv
zH2_8bub#y|5ngv3!sJ1cCVq(9WZh7rGa7zJ$_f%USJ@H)0fT)OCv?j`FqJy>JVq~*
zg-ksl9jpUI7|FH%(TvoBC<%6|#j9kZfrU&$dTu>U27xRoZUVj7W6)K6!8X0{%)JaR
z^&~9Hw82)FhAT-bao5tPMVA->13EcxW^mR)0R}@`3B9qx6&mRvSNbI>ZTC{4Mx5c3
zROSbKX&T0Lf=p%IE;GohY*4J5XeVdfACMc|>|0?#9~gOV3INdneh%T3M-<j&iO*4d
z<2}1NFxNz{`V(8EqCoE)t!~==<}2vW9NN>enpkgKxljhDB~L~$lFOw1V)_v|cYV%E
zM1j2%lPnZ=7J=+8^#G}SHFA51`*OfEJ3XA`is!%e*%22yVTnx+f2kDBim<<duHr5p
zTnrvNy#<i18h$W0nH1N`$s(OcUgTMhlYbhn?*nmUl08{c^U>sda~kB3gyV%|E<7uG
z6|UFCHi$=O;*J?=Jmj~qcx*b>ejS0qKx@pMK$wo8YQmm)PIyPiPBrW{G6ZAXkVZ;-
z%w)Al^#paQ`}rcgDqd=?6&)dugzMbn>b;99ab$gbLO-Jy$JbRL9L6u_C;?6%1v87Y
z!3lQOm>Tp&me9f)08da9ssY@Y!%^t)WCm(uJ(i?t$z1*#LLYYM$mtdfI0gEc%oJXv
zrz4JcX`w<(X;JqerUvgbxE^w4>yfRtnxsvZDL}ATh83&6J4%QcWX#$IC<#DW+|p8E
zq;91POHyA^B6V$K1#xVqGpFB<nctH34h?W=<uh#a4FH4Q)Xu2dHZysfeS9fXGlWaU
z_)6Zg_dpd2n+ElEjZf7RN2!G+(!7meQC=K`LQ+6voxxp|Woyh`l~^bM_R1w_`GOwU
zocYS3vcKBp`!_qLrWX^DdI=Sxh-><;Hrm30D2eLyyUQ=oa`8xiov%}d?H0YU^bJDX
z!4`d9p;8`K_j@MT6#$k5vwoviLL2I<ncsta@rFd;N|^83PJUdwt$!i?eu03M<;efq
zg!$gykDcsJ?=Irj!1VvYNCzm=1Z@LVXH_RRDqRV{;a5ji)r{nUHawl;Mzy7^4ibqF
zo7}Ix77A)s5+$;yp@+hBm7l5}Zy7F;<)GMpMD5~N`}xkq<ahV*d=Wj=JI6?|BeI0*
zIe%A|xdSC)xKyIu84@oG=IN|wX&P|DX*z%4%Fvz8hAGX$XWklef<h1bmS3x-o@nU?
z=(OsbI=cxEi}=TS#`?ewQ`S$4Yk9J*N1)OpekJ8K<UMreULeh)j*xk?5~pi|9YxsX
z`S!&+clr2fO&VRETGMYFpr)_PO0O=o_Yv+hJ+E{@H65B07Yvs!Xp-L%N4Gr^#sy36
z%KUfaJIc9b?t`)A+lw}MeVWw@P-jL!h76*FbNg}eHoAPx<&vNI;*wkHa-JNqHK2r&
z1Fjgim~9cV8@6pN;fjojdsNy8<wOl^_S(ucT2ba26jdN4bhB?*Pl(QhW>;1d=(M#p
zK5S{p)_&GCbbw3iNhDZ*Sj6-jN7)mH0~ZO!;p*<0n)j)IKRbTB$*!?MQDvEHJ2nBb
zZ_KUKjsx0Irz)dA&AC>hnSwT4qq#m8JS13lDr-yuj3Ts~J{Q_a6uPo!-9ehQdb^Rx
z=c1^nXg2VdrSm3cYhhy4*;Bhw?{b?{t`y+g0DC4W!~2D5I$x{hTs9QOCdyYBro=v7
z5uMye!Dm2KKFiGb0g25+&&8@JcsDxMmjHs3__?08y}28y<RVj~=@)rgil*5uJ8Y~V
zmaj=B&$qUr*gT&wr}<J-lgRVhp3}eO6`I`u<xb<NDh6zJmxJi(=?zE!#4b>S&$q2?
ze+6?%uq>|A3|^q$ynzXk%l}=C<@}8X@8?qWg#D;yCYvDI)5Qb=sM3GF@zj{3=_8sT
zsJnhEn`_>>Iaf>7$61V5%jqJ`tm<`uK3XbEu+xXu+Y_s5&<3Hq-ySVw;Sfl%18GY#
z23sxF_~zM;nq}465&$dsp;G+<XZh}$8XCjS?CbHt32+N|<HR&n)opfaN1p}mqjEiM
z&ULES7Y;R#BYWc@>0gDvJRgnK%{h#XqlfG$ex*N@WU9%Lyx=;1(5nY~ZHg{Y2*O1Q
zHhKrikw2U>b<?cv@$-Eed;8rA)Ike%+^o&f4;jjt>>4|oenY4GIZ|Nzu%@z2*FhWa
z5>p|z_MWNv+LBCngLwE}bxp-BIulAdmx*daU^>6CDbpK_h3q@{LxN_i-rg(aBSrpJ
zg)j8dN9TKDp2fZCgP|MiV)U+8rD~Vg(Sm9f#Jp)DB|bGae^Hk-wrybJOnF(>^Zaeb
zqx*e}uI%-3OVp&k;x2=KxysCs<NDbYH>KYG`DX|5tVVQ3zGl;%<1IKM%|BH+mpa>^
zhV{;;IMIix_;&eZTAZis`aPNis5722v4E-bG)fBC`Z~ueoq<hzJE>KpY^GIr?fdZZ
z+>bo##QZ-6Ru=aEJO5{4`0rrnJ^z%NR4nlSl$uX)(f6MOF@z);Sj4XVSMXGOjr+9v
zgSSepRbLmfoH^(9XA%polPaUfp6i2Ud}O>nzm6a7?>B`{N9k+><L8H$#Q6b(dVlvz
zSdx$eHAqV1>=x<lvZbJAA#kTV6z|yIZuS?kbyM!<gn-?L_LnXG^d<X?oyd6voJQ|{
z@3XX$aavS+2}P2|#g8r(0zMzysh?G}FgoC#k?r-c<N1ES?S~YJXM<q%jl|urrp2Z8
z>52Y{Nu4S`Z?mJj?@<k;pd%~5UeuNQ{bKO(d3Pw@7&ry#AFW|24Uz+N;a}S5SGJzG
z!-yXMIK<Sa2MGG}_C(@$T1E}-1WAhccACd-^?FivTLrCJ-@nZcB?j)2eU5{xj&I}8
z%I*|?3tN{uN+RCJ{IKt2caN)OQn>tAib;weM%A75F=QlT`F%fOvguI?kzmn2Vs*^?
zYGC$Wf_gfiHJ3s?k4$KL7V({#3{mp|Wi1%2B@l*9v7%ll=+Ifq73aLuY;tEd(U{LJ
z%_qN+_Vc~<YJ%~adtQk)j#*+~UKK`YB9GYWwKPS=>SRzH8CuiMxO$QtVkANTwUh`Q
z9Ug~5GGvmmo-^=-SswN9jB(*8M-U2UpQ=mxfJ=gD$u-GTrpvN--Mq|ZMAma~FC$_i
z$(1@YVGm=KslBL7oVVX}Gml%$+im86JnH&hXAD=o)8sln6a&F^B{ugAcVk{@s}kq?
z%=YuKv!d2&AaZ)U7DAN?_8``9dk)wv!^xI4P*v?zCGI|mxg6n4P;m+0%DSM|JaBGm
zEeEBbs3Q#58Mv)dfXy8bF1sba4T{75PLxldfZv|Qky3dG7B|X32by5EX)ZRL+`zD(
z1)c&DXSmn;L^=ROaUu??KMZA?pWZUuwWu5uh#Elg?Rc5k=QuLJXiLMutI*5bTF#+r
z;-$zgj4w}tn<Bl{>AlS&msvmiP5MNxQkTI8C^7C**I_8A`IRaUB2lB}MWr|340eJ|
zeyiMj@NDz`^xyS_+yoj59qlB-E{+}LF24!cF;>D?_e0Ert{N|{T3V1j5*^iU&!3B3
z9l*X6bF4dRR76)zsUS^$0$J$f;S-HX>Ody-cZe|tGO5li({KXUe{A9?Kw%NDla9;s
zY<m(9*j>eTQCQ8QM^z0xDehL~B=|du%-P}0u)umPFg7Yr@o{A0XZ0t1L=^hGlgB%s
zf>=WAAUOQ@$vW>3oDE<9Ui!r+yNmZ4>A|wT(mW#tlMvcs(1_a5X(Bs>lO*FmS5l-}
z66h$3t_>MOIoSa&A*ZYw>JR#&oo$d}g7g;u`t&{+lG`!{=y_H!71lx?-X;xNmxNz~
z+B>PbgD<T5fq8^Iu%R9*MtD)E9;h>@%!we~a0#OvdIG@G@%N^7!iCFj{Mg#%RlSu<
z*>7ozd>nM$UU{o#!M#XjVaFPOA)?DuUL@E#h6Bre0j^3GrPSv?%XtFTx-)-vRHl}4
z(s`&i|CR0Nt+tcsoAkdj30=Z8f@GqDb|lJ%jNZ7ZM<IFl6meeV21dbAGbbK8&cfEH
ze+Rv5$Lg@1W*K=nNv~VlzkVR@@l2dVJTr2RoxBqQ&Fd@ZZk_8Ld7g)M$*i5;WDjqV
zOd&xxuX5)2qURxB2)iy$_37|utst|Y$Y_4KV&Nvc0*Pi}e>m6&hzBAwK(q!NrbK~$
zcv;$WGFy=jF#Ec=7`)Y2SfD{3b^S=y{sG;4ugDWyehQw<r3V)rvG_nrOO9O6=X5ou
zi)mna@W#nd{9?}8a+}M52!mFr<Us;p;|?UcMj0CwVq%I8OJ7S4OPhbJ+#r)v&xz}4
zXmjEtOT2(doZM0t*NC)~I~nw2$3!s7;$(DVGyL`!%d~S$`zaMTxCc9-fP1vx+c`)6
z{cq(1CbjX@tu3tZ_9%Pmi-Hs7T_(AQhb4ZF9FMJmosp1j^Ch2~X78XhMBp7q=nQ;?
zjcLnqGVh$6!96B$;SgHbX~*5;&#(Fw;a*v_sA4`6t5>D<lEFHHGAFkSTx2$x*`_$=
z<4`YzHG8!ImH`Qo&+GJaSa)cqQ%Q5YbQZ0B(o9cvYfM*j#C_tv3+xil5^20{YHWA_
zWtry1mCWcq3F7Fl-`o#UtAhx)V`$D<V+ATQxAFR-xZL=>Gx0awErWbKR0wu{G&mb%
z#gA1h^DUCedt~|`LjrScoR^VKYE;bQYls?TSJ&lb%+Yw`Yb!XxXga5IKTUF|M_<R3
z$`Jmxme~*GT*G}F-nu_R8v3^y5{J8OT7ul)eje`K9Pnp+Q^xi3Sq!73{-x*I5w(b?
zZ|Ti=s~_fEGfjEGYBJVa@C=^MwSv>W0OjG^^5xjA|5{=^L%GYb(b*rfUS2(zp4iB-
zUva9vR<!lMYzr@cl!o-$uKkBxng9&ti7{MkQURAS(M+u(TI(~{$iAE;*q$Pq6~Lt<
zjb77Y;j#Va@kN-7Cuy&+DN!5D<DJ*qz5(7QoKs}}TJ%-Jn(c3ZnbeH@Y2|6hUqoJS
zuq$6^<T>m7OCBpA#-&7MMIcLU5pc7`W{a>}L0@r+u)O9H+hZ*vA&@?$HP)+=mf}CX
z%7fD!+~#W5{K<Q(7WS<7z;3~F?kLsag4vTYA_ZlRf?5a5kB%$H4*FN?_U11KM^!3!
zS}?Z27-ri722y}y?Wyc?6{()E>N%7zUusF$sF7fLo?#--fr(xkOGQKv?z9WnBjeTO
zj-@LjZloZzqit^%RdI-o;iwZOoq^DHZ>HDmzOH0(Z8qhBL=JepoK4M_&N%wY%=f`I
zqB&;y_^UL-wsk4s2AQm2gc;npy;Em9dd=2o8|Y-ol(>ScD&y`vB?b(hsyCPiHnU}4
z88^?cyIr-s(lkgIjyo6^5>p$F$1R2EHUsi+A)m6F_E%yu6SkI>1AGiY{v0(L69vln
zaA9%d{Y1_V9Hmiu`LB6Bm@<z@2`1CYi;6;xM&tJ9{f?5dD}l6KDhthh>;pN7L&Szs
zZtelO+`nPk2le_ZiyF=W44#r*0xq-N;f(1jxB8$XJbybqt~^;VI7wt~cd>ud*93tR
z(z*vj=~ePjGfC0Py(y>jU%uwF+*Hrf#<s$;Xt$oIGI`z*kZEWU$VM2O0c{iHlwWnT
zrJ8vQ69jbpV}(e>XOdot!Xc7S`SV^!N$GE`9>i+)n20pv3$3ry0H~JA!@renC+4LR
z2GLs4Nj0XGBR&b;oP`L^Fdo$wEP$zJEy2qq6O{oD(=;V|;(+9_R0+D7y9Yhy$S!*X
z$#Y%7rzEOLEv%I8Nvjd0Sd|JoRi5c+aj!3TNQCXb01IRH;{Pchvi@Ik850Y~e^<_~
z!5+8$SLm;;Gbctnr^lupra4ELyTAasqX;^K?|-2XK!juDKfpdn>f;%HZg?AOW9NMm
zzYuGTY_YR@JHAEpeIFd0+#LK=13{b2Mpf0_`Ef;f^FQnhJWLbS`?n-Y?%?t1`TjD|
z!Tjd&7tX#G*_Zm~=B~`ww=8%NW#CYZyRQlVj*jGxXi@GxGSBO0o2L(I_K?G&uIB{F
zWrM)(KH|XyWi;+@3V<XSJEz~$<Ez~+_nmhMuAfl=K`Mx!i`U8@DgL*GuwCYXKSPkZ
zgJr%^ZVRQ=Q0(U26y42SgS0dca#?Z#j*?qJ>{c0SAxO8+-dMTGF8NGD#~tHycic63
zCG96j(mDaBfp)W6^;Y*bZftF=AYfmt{U1xYI0YLqSy|C_zOmj!=#h=9rjW&4!059H
zP7WzG>E&NfxeVjUr{W^R%PhNsrm*YJF<!C(v&yS9HHdJ9)z!vB;v+fKj0=u6{1IzI
zT_c;}!4iq?ISsUk$$d8EE7e(d%sDkB@^ppOblT~;Wt~pmseOdgBD{9bUof^uC_=W_
z!{v8)sgI4xvbXg4EShlyX#nW}dNc^g5i$Ra_cfh+85HvFa?DM0qzh9DKH!c>=9$Fu
zs8|U0C*297{8W{8QKD#)V3dO{B_51*>aJ|v(TeyHpcB=0M_I++!a}#fq$yY$yad~9
z+YQV0oq1<GTSz#lHpx72L_Fcp8^er4n~;uR4V?))$KYhHCgd2BjQa%_br$YNs9g!q
zYIV-yCp<AYbm~*a3IIu^n$Rj&1sAf>Um>D_#jC?RXjirvT3UI6?>06iX)Gz2w|WQ&
zv)u`xnsBrwM%h?_mxq113#ko@X_8DKah;a|HqdvMhJuOZ=T-_K1F}5yt`8fDMUkoH
zz=X1*YN959DEshB)h6mH7*BV_6#{Y>Tj)>GQRg_Zw1Adq>Ep_HreK~Vrp1L<Z?mLa
z{aU8&pM7f#$-1-|g5zgTAs-|Mx2RsdMvvQcZSS}Ipa)eeby4>e<4wp26PXiYWUI?a
zSSEEKkv7)K_pY0{Ue||LOSi9@*#yL1Im4hPK^|4vq>KQO9<=dhp&B&hYv@cnfrdc`
zg00?R#U2~L#VAfQ^E%_&-kx4>uY;Hb`YW50dcC~fpFDkED0se3{}e`U?{?ZB&w+-M
z^VE-%d~|-EZ`qA8>ohn{x_Zr{d8*hAD)~H^KJ!qIL;6-=(i-M4t}R6C@<41jpZ=Q=
zpAShA^~t#}1BtU?)-O>_8oBMLdDRQ95a$C;`Yrq)mPM4+UQixBVXVmT*eGT5N;u87
zr92odIM?G<M;CQhh?vA~Jl>5UBcoRiJ|i@TIYk}7Rz5z&)AG>O#`xMT#X`~YT-S-a
zDSgf__X#|7E-*`{ZBq|k-=$HNopw4rgC>%_)Tvm~2yxaL*YPnw_*jcCIz;AL8kA_n
z#2z++5`xsY|2C6w=+LGMoIj|z4{4dwz{p6LW;#=z(=OV}tI83CU$%5x3^+{u&&GL>
zKnF4Mc`O%t5gV;6_tKo(*uHt9Mrm(fJ~D4Bp?r)cVys9b631R!Yp_6@g5O#8yVz{+
zM6oqgv#%_zlPGt)8vS`rNvnd7iWv>~CkVK2kDa2R6VD~RRp_L-Pb0fqO+FKoYB3gr
z?IWaVB3>TP+tug*>(E49G?oPq1gwuNic#&{u?&<@Ca8m>-kde8-D!DasdNe1RXyej
zG?k--f+qr(VW+aLJjJw#CZX9xS~j{;rLnFe@d!8nc4na$8@lLpToOBGuSlJ+O)vw*
z-RLfbbF4tLS<GFFKdebAO_Mvh;CoW;sQHC<lUyWntC7LcZa(E`X$;LTYnIY~wiR+A
zvosWm`_)J!Bf{)PaicB@o3!lgP`}%6*_4w0>DTWs4{g!3!gQ0T!Ys%&`X9#LDM)l^
z+X8Ldw%xmJ+qP}nwr$%scJH=r+qU`o-1k0S)jc0Kl}b{ntV$N<Tyv%77~`rO$k3)}
z&arUrY%x7%XG|iVs*q$!dU}rby94X~c=`=yXK_tZr9H%kRY?Se_xK;dqx+`?lcyR5
z(J&|y&;l>&7@nW|dN2`(5S?_v=l59-ed%?xYrB#GRn_V6FJ}Kb&JdgA8e#sKfX%>S
zy1BIRj-su=JSKN2I|!*8q*KI3h@qGJlp)z+yfAtMW#Z9Mmq+3PIO6arVvHzifzy90
zvOQ|SPHm}28{1krf|fKDFRcKNT3@hZNOp>8&&JDNkfx&+b!P=7{12*?3Kz>BUHaVo
zQkld;8;OAWK=Uog^TAA3B0{UZwB}eUKHCy%kwGbg7x-hBDnw;%QfM}lbimawuMnI|
z$#F-xxMM@TRVGz|w2NyYiGQO*Ps^wV5aIr#vJ%{k{E9kUB>x=g_C@ZNI&=1bo;jJ^
zzNY$ky@iX+8JK!svodg!cHYi~O)Io*FGKJA#?xP3)5XleMwAK}2Vv?UqBWo`WSYrH
z*caSP<Pv28xFPmp3e-py5%f9USqgwFCNXHhdA5UN@j^e{gv}dM4sA+h4uRVphe}0{
zVzG8cdh#Vi#qr&}>YIvYlgb6yXK9y0{e&Iw3#Y(>UWO^ulHO;RLTb>CPFbzCY<J}%
zOt4&#<rywG6{dKU<ycHuxFu*N(^-&6mGp~pCd2${Bi(f+`D|k(f5A0KxNIZQYSOYZ
zS>xI^&yuudBjnC*G>?c*=0|X0ujg^E^_HW5#S>B?2Ii#I;mjL<nrkw>Zi)}88r!k4
zIU@E$xEpdSkZ!&k2HkB>of00PifBWw*H5h7BJLKC(SnfTfip>2bNK~cqCH5~?ggwx
z^MEvNItYa~qB!iK8$)c@A~?E?b%{+=ndI-oF`(np7;**I8ZvOwuw1~PM46;r)!w#R
zQ-(k!0xwFWTL>K!i-Aj)njPCBY$T93BDy2~iC=4Q(3Q;%tD#|FWTL}I|BW@y%WXvg
z6U-VE8#vU7h5Ei*tfb994|Sb_OaU^`u&)?lsk~&bmG`%x68DoV0Py)2Z!cOPubR>D
z9xWKqczQ7-&Y5Wd)<=!EQWj4Q#RkZD%<K-ft*B~8{+N&GSa*WoT7VIyHdj_W6dDlh
zw6EBj9gi+uv$^^`&s(Y%ou%&_Bpg}8OLlCo>TEM95!aLLp&ib>&gLQ+$(*TPuosPD
zVVe%@1_iaQKId_tgCGhW<I!_I)!02(X@m!=;ZQsbCn_@c&cd!Zv-p}6CFHA^PYQt}
zP$K^+pZ-Vqb{J%|Q-(1;dvxqcbw{rGX)v9DYrZLc3eURYuX3Qo4x-5hw`7PW=rauN
z0X<?k1Fo4oJs6yV473RsN;%B1_7rB@IXTT&u9n`CEOLXfV+k6K_C_W6`d@y}GNx;6
zCVU^7xWz$<yaRqGdrGZ#2d5i%Garm<sYNvvWIvX4B+{HJ`}>j2@RaT$xlZbufigQN
zwsUc{O$jjyN3|oP(!yr7b?OVTp82u&)=8#NOT$5lyKXZlD|VW&91)GbZ;<!XC>FXs
z93qj~q{`{z!6b}?7sa+K@-cjTu2aj{eeIHRL-kjiC3cU^wpHxshm1G|nPe&phfs#_
zIFh!N@9&dMNNeRPg9TM4#SOMP)AL#ArD-zpXTcNIwGw4s`GohUC_cr4DZ?fgl%C<g
z6C;690UBbgS&Pg81=-l<;20$N26tKPztaH}bH|I^uX&^TrDRGwXOvcpGt%ibe<@qh
z9zRr1^SJ?K%<m{fshmw#cZL9P?Cy+xsc0gc>qLw(iHw1v9qVCVBq7}}S}{WqgM7OK
z>HO7K-za7r9XTunug`~8qk(*74b+LC{KtfR6eeyG|6(oIM2s^bbdl!44X1ew-@@Q{
zTrOTL4O)b*!R$VFpYrCAIdc8XRiL$syaI1XHsF0zT=}Vh`|q|58ty#?a#fHz_$+A!
zDEjeRvtrOK9qYQ>_|(}yDg-m)QBY@9m{3w1XfDSn=lL4)_sHo7vGVjp?%r#|-1X`+
zhu*3$sFRKkqWSbM$`s;e#c0U*)Q6Y^=_NpHD5GgPm2~r|N=vV?RaYo%i(Vx+yPvm5
z^0U+-hU=Lcg`~AIRP<wU^56h+=5p&J8vLWM3b4ZN4w5(n9Gf!{cqdyV70yI%T^oIK
z54b<P*+X1wCVK`zMMbBwX{730kV3B8YDt&ma{}O#luP(Mmyh1F#9%^_4H<`U{TG%2
zU{2lyg9Pi>-8##t$E-Jl#pEt-79rvHri$8n1!Jt-S^N=`c20k3@qQ@+?OE7GT4Ak)
zu7!uGvLg17-l1f@x?F-8Z+3=q(3;g^JeayBG9B3*&nNC()dylsAY4X0&<kPNyf4m0
z7cFWQU&xAFr_urSDAZCmYmYf6XhpVVtOIQs-y|46Nt0+74wq0?*OJYZ5Kn<_R7SeH
zO)|F46g(oQY^9Z2nF378{VLQKt8KZ_rS4V}tT-(?(c8l^0GhuAF_{tXtuOI(UHiqA
z57Q*!2CoTgo3%yw6t+)RqqhD<6qNgyPIMyw^}-g@30aK(5CSzxW(@Vt*K##<3yGcc
zlKl7Ynay?%z<EOULrZ-)WrQY^@+ubYi_(T7Ifq{ytLq13lv#2ZBgWB>+*2*p<Mj6Z
zwwY}j;MLqaWiLxX(nu23HtsHpO=CFMtxyT385i{atxRYR6AP(W{Xq}@Gw@W}2<mcF
zA}-IZ{|~a^4BT@ajtL7(j%B{?FtBtJyESXkac#_6DV>SkM4X|dqTtnOR^pusQ{&Od
z)hI*=^B^a#19?J15{$9Wf%LmFT?K6^E^trVPapq_=YZ#<?<v~laaHP7@8c|NGT{CF
z!@|bE4?vF$(ENW}S+e~X<RSh4QQWKMmkaj)Lmtj~nbSs96Giow23$W+19fv`1Obke
zy#@fR2|XkIe4>ylPNl6g5#usvri!cj=<$fBmY4C7@%VUjc60l9x3YT-fox0D^~e#>
z{f8MF9#;mBp96YfV&&{aP7|AaMTK^8U9+R>bH=gkVGKwL8p5&f?GW)W4YXs@k!{9t
zYyQWZcj)DT-AV6<GDchq-YH{eHD1f2BfD9+8Rba6^Y(LM0$z?&;QbYb85A|!L46<b
zD48%9rvS)^5f9}L#r*AuM{?ZEG>n&4*=T#^i<_UXxXxh*)Hb7v&Mbw+nfRe*+atT5
zI|Yaa=%D6koW}U7bm+hN05UUmjw_}=j&l%>FHH=oh}`f`ivolZ${;04xvUpsLOd)~
z$WP7H&3f7gfAbvCxzGZs>Tf}0=}F9gzHim(0;(#m@7dfZ^XJBytv9EbmnY-R(Tbzg
z0AfsSFC~{DY_+7`%GfQ24@Q<*B25KE=ok`my?lH<-ftFk$|(5N=c*}gAM5Jm_<jn@
z^p1|;lT=mlzU*w@T?CJOE}!$x@a6bu@47@}@D|De+QWaISA%4Xbv?tQFnv?NwW>yq
zXbxvajG<_dR$R~tlkk!x1WVaPmZBc@D9Ei7v&03-URTb@M*GxwbPF=ojBS=Esmb2R
zH5a3(;S^~&s`#%tmLxBUXc?m2eUm|#5#>Mo?l=Y6XDSj&`VvSqFfWy`=t_B}<?2&O
z<cusuC6++mR=B~`4p!gN#_SXl8F)t=j@n*Zd{a{=Hl!$R@P>}aXDvV|{f5dO+8eG;
zG|vmSK27m#mfjK{;m)Qif2wimu%^{%OG05)bJqC_v@YGWS|Ff}9+a+|Ay}A-oqE`x
zrsVo_%Bt7;{~T6KZPU#$Y4$X=9uk86J^n1VAk9D~Utu5W)Ce-!_bDD?!~~|HQIoOH
zk!o)D$+kPVl6t5bH`Y+WRFF`XB#N4u3v9F|E2^1{;r98kdDm%FW3ucJH9%-vnN~>2
zHx5lw@0GYLCaAXM`|BhgS;(yO<NaY|nfv)Qp`&J$@=O29)TyagT@&IRk1DXy2~~m-
zIHtoe*Tp`b<?>LTP4X_P$tqTrl@=0Q@N!yO*HSfBJ(_4m57MAKO1w5>Cq_^qVajK$
z0k`#(Rv*-?oT2lFR!uwPjOP!FAp1wH_6my5-j-nWACq`aLiK!z4|$T(F16)q=o<~E
zykPtHsgDZFV*CBF!||(=$^^rtVC(M-+s16A<+X7vb@d@~b1!eJhzeDt6I{7O$A}fU
zesvFxkhoiZ4$4ciXmMSA)bO_czo!^Cmc_4v<17MuU)xLK9NT{nL&Z@ww0pysJFdv9
z68b?=Bok>r#w?4I5glPP8o~KpbtsNET_oyjKZF_*l#e~|(7?A*BZ0X|W4{V0xR@Ay
zncPg&yYlKq6{XU-LhD7LP?0_fg8?ZVHxZ8`_4{<>Rr|3dctZ$_i}C4tejk%0n=Dao
z%A>DIStuZyL?iT&`dRS(eMH-MTq@BKz#t~s+I>>1>|d}3-1wZm<78Cf#I^Sh#8_0A
zH3GMB$~r`WVDA$`>EJCbzX&QCqG4P?WrAWY@w1#ZwOS0ypHWVd1+9~mw$i8psV4o7
z%h&-`#f&qZ&i-q5oV~Gx=`aLQS&ulK(i20;wG2WggP3AQhqbQB4Y^eSxem0}gDda#
zU@elNdN8K*ZpLIS&4Rei+$&EiJ7(g(7wjQ~wOab-nu8HMp^37r;<GGg0=kZgZdGt_
zD)SwKo?<z%Zy(NjmE8d=7yY_f!?GYP-oz>VX<MU-62(*&RUAJZ?z4^Z?NlE716`Hf
z=rmn!Ot7N}JRiyU<^KF;%5G9?A7@9XLHbY$jo|IglcbhR>gO5qfo(VA?7wmNU>Xf2
z8VBmNe)$ag@&90I^Y}a$WwFfJAzFUb;{0NwwUbwGNRPC&q$c{2-68yzf}=IEHldqj
znrvc0^5%|daCm^k1sY<j$L`0`G%WM^_tl7$i})gU+La4rKifs!19vA~U2aGy(c3N@
zag7NV-##J}TM4C*miusEIuQA4pH$*Mh_;|Sw0}+sq^6&bWt%oa|8VKohiR-yUEk)s
zUPW#};_SOtT#+DFP)o(_WxiTm$~NqTTJZzo+w6&cwr{s;x@%K-RfQD1^Cm+<2#spq
zhjVboV&5H9Y8);4bW}wd{a`f4E}Z&M<FXP8-AZzw=$F|J+O2d`m0d@v<|yff+6S0C
zRo<ZJweAg9U~M+oYu#<6RLQi{_KDbZYR+Lax>4=s>|d}N+oibJ(;td{x?Wj?aMNPB
z#BXT|UlVXC6QTq(!)h$|mdCX^68rx3RMzOA`>2ku@{|GAK%3S#KC??{WtnotcGW&W
zz{%>?L>(5%r!smu%9<W;6F8TQKX{|AWn0*wrpt&!1#|m-Lx=-|8N0;Rf@4~ztx-GK
zC|Fair(IoVZ#)q<XDT>V^aR7eqf1>;wQ4OCfzc{KqmRweRSLSl%Aj+73OJRsNQ$TK
zH8o*QO4hxWlif_vU}do$HRhc%%)JM%i{OgM1@g@CpA-4}vw|ac3+{)e><S690Crq+
zb*<#Gh<W6{xi%WIABRpt_Ixk_RmmEq&C#a}F}bq5G&clonKN(c-lnEP&$}c_f{s<g
z`^8}sWiIXqB#KI26s})C;Bk5LbN@AZVENy93rwv4w?x~2S~5w^?XbVR1-tOkD~*5r
zl>aTMZ1LP!FKs7I+E~eDUFzQBHKu)jjShnu<pa7U0Rsy`ulbTp#F55t<>g%Z`u_F#
z_w^Xj+0>7yzU1>Y2lzwqMkn5FQQid5PvES_*QwVdQ2~R`#@n16?ycVK>Rmd)6=(!`
z>)m^_Wx4x<@GAeM;dAdGw{m@t=Nsp_-}J0CFaCvmj^!tiS3-gxO4=d{btTVB+vD?f
zPqq+Wd`qSss`Rz<QrGs=wFJxz7Sde%@in6LDnJb3TnS9foq1j7=i=A(RkN(wf8xKF
zG>jm#3p|!?xqJ;8n&<f%&Q*;`Yq&c4^n*MgP>LTs`}wO$UB~zLe0s+Z@1w58zT1|6
z3=%7dzJUM6(qd|B)*A8ZLugx(E&Aly*j!m*CuR)ydRQ(`mlkf*(%{b`4d{hM4w$S@
zm2=jk3#{ZbDB$)o))CDU-(yxw-^;|`1o`O5EDJzY3s#<uH-^SAW+cc&CNqG4=F}p(
z>X{m<+i`J&?0qBEI<D;UOXlk|wKx#D3v|uwlEmS7nm9f9qQIz7jm-ujMpzudyV84N
zCG>{kjbcnx*_$2!G=)g>Q#j<J2oU->*Y8w^gqf$9hqKp@<^KRGTse6{0YsW4$ob;3
z*7zeY;;I=Xu>t9`tS{RRg0-N&dd(=Nqokr-3DNegE%xm5dyiU3t1tvy8+b-mpeInV
z3``c^F6<gV$bVj;p~H5+cPQSC%*9-PuD=n{>R^Xrw+JbY&j26A8!nS?yN(|%#ZZHF
zZ>za$4b5hf_wDuZs*awPm1<u0fM9V}-0!IjM@Hs&yj1sqqZm9yG-6G>y-!>St8*r;
z#bX!~fbTNK=6BtP5Xh$){a3ihVISS|2PaOV=Afyu!k~BS_ag24z)MldvTv5wM7y8i
zWJRIWBF{<CC~yzJRtS7O%+60Nr#Nv129HKu3pb;O!N#Q(VLH`{CZ&7rAyO#17=gzz
z!IgI7x#6^fCzns(L6p$icsfUWX(QhIZQB^k?Fu)d`;DB(Jm8AIaq!7z2**}M@R!ft
zGp85DLN0rxWId!#P6vql9RZJWcsF?8VU=r;yUw`4l}9E;I+uFm&D?XZ=J?Wpg`S1-
z@H(c0gwWh1vMBqfe8@*@^sV)+|Hv7zlhekBtiEv49?E;tXeXo|03FpQ3>rqKkAz&V
zDqHSAS~CG*k0v6@<{ov2R&?k+=}PWyBUxTIyrfQK(eQBc8jg!u6S4~bE`8X58v&cI
zG3PmGaa<vdJS~PHv<0z|*KrgCB(xbLrNZb0$Dsmx1KE?~uC<TX5!d*!yjWdgrHYUl
zM%wllR;X<Eck`9S0&(as%~7^pDl#+m2~FXI_f1h)XH|HjIKg#HJmkWvPT}NHow8DW
zG|<87$G+??g!@(3<AkG4g<0c6a}{!em@&57&s2)-K&wV?JC0VhA5OL<(1}IiD3n)c
zZiG89V_=g_<Qk?UCK7}1O%pV=p_6@P&{NApDA7i{p6oB~*sZ3j^yKcI2{)q$je=Wl
z-j*mKtm;4ub+q4Vms;Nf;LVVH<oeLc<_b+|+DKA~hh&2{5?1+FLmrqo%8Ia_P1c&B
z&CJIdm(IFySy*KTE5pXnEG;<?J9TO+?o_JdhO&{LA@yzEwUxSBI|9+H@TgEz8gt`w
zI?Z&1(Ls0485c`_U>kGO_ZUT<&-GPgsEMlwk}55+6vXe1RIc2gqRj}gs#$7MW&{iK
zY}DE=#FT>fK}?g1nf1XK^u&@_l;2k|S*Ar{>o?VMD~CHY^i0mHu5*YV-Tq;{jXP$l
z1SYTZzsC<N+gIMHQ<ib)j-|`0&KDTJdWYh(cvUpFWIf16Ya0dTT@4(x)XCiz!J{nk
zAzQ;&t*$YRBwrVQ@^aFN$3;*qD8;d(@+N&uG!%=vj=p_ZWXC9^=w?0G;k|<8rmi4E
zp4{3s$L4I%-J?+{j;I$$X{WGN#7!=_U2Cn6v;|}tNS4Y;QWiAy?-E(77soV=uALFp
z#fi8cXkHDt*;a3awE=3CZ*g)%vc{I|`6RMrT0?Pbt2i_cO&{o5)7gGF-CP-M^#7so
zs9g6{AL@Q0FAo{9)D!o~uz}(_{$b>zgt*Fv$wG<tNt6TqC8bf4T)4NQ&{#=gR=Ya~
z(w8gj7!UL5i$CYzl6In&!Deo=eIe1bdV@>vY+=#N+NQ25mNHu{Y@S)*A@zyPuPgK#
zr!ZjY%DoO*84bV9fN)YDn$EQ_MX>Hydd($5n65~?`TE&8GRNNobHrdDl%TXOc@qe!
zDoNMs6IE68RbL*6v`&i1u>D6;8)Ya^t8pz|-CEQT&<K@$Q2f}Um<t@IdlYbSyO-Eb
zcTW9=3;#FPl@?_)a&m8))0T@0+nk9WZBxSh;eBMm9reoOMoYvCvs<Luhr!5J2-OUK
zek$CWa7rr1lM2xO_rpZumyo)n9TP7EKU)m`vXo`(3jtb)cU8sZQV!lg(|hY7L=@1u
zem16szXc`&4{=*zJHTi$lvok%{i8qk*oIh#!ih@YAg^|6eSMW`@ek#KXgOxVTl345
zhdijp<}rcJ3l1)J4MsMCI<|jNoNlsoU=fn+x5|F9GxwWx8-5TgOlFoJ#oPtVZ6K2{
zB`MG>`MDA4rzAZgHqa3yMOHELDBl#q*v84*%s}{BL>6&SN{0=e_#lz#V50^*w28^J
z(7L)YBU1|yXX(o1TeVx?My@-CXRT%SZ8pnUFr#P@ZOV2@)ie*()SNhYdr;@i#YD@M
zDrjP@G6e(E#*^Ot<7%2~wuyp)$Bg`owW8sxd|&24*TnN(Q?}T2RheACzSnuO`4lYa
z<5vN%mBRoJX7w$-J7{nyorfI*3i;)*eQC|&O{y&ZO=BEpAc3;6Il~x0!yj%IiZ=0N
zCRjr~*~JAyj*`5HWj9rP=O;VU+5lHyOI&!~d*~R8(xgyks_FDTvOijAte42`o3o{z
z>4M#is$o&-98iZn+G*ximf>MAQ|hWisg=&*HIz2dB9e2g8;80@y2+%vl47S0$e+<y
z5z;f2gabhpIfETX?HTZtPCN(u!(}rzuy2%S!73O2F*<cGjqf-V@VIkPO7~V10V$$Y
zc>0v#2Ji#?Lnrp{zlCG$|K(>6Bh&vijZ`XWlMJEz=07x2{Y4%sHa1I>)gvn1q-$9d
zWhBp}yo0~781pQ_V|qYB2~IN>pk<%ff}Y#J{oY1zSe)F+OPO9R6xe@K1mJG;zynZm
zexPDQMB!B-Op`ITd*Ji**weu3m~?wP6x}sGuWoDg@>1+q1>L&_FPcCD<Ro`V7j$&*
zI%n-(-f_;0vR`X`2qQTE5y~})zHLSxnADT{&+l<%(muQ&eJvoD5khuFoqz9ro#mCi
z<@Z7#AKi|;l*z+M(jOfXWE*W>?2YIRtd!|R1^>xlLigR<iK8ZpVVRV`HunAQ594dN
zM`l&shW!~X-I;KEy?wdm^35m3@InTq&Ia(4qsJ$Cm`++%2v*H?xQ&ZCz063MZ2wRM
zHPf9L5^Zq;T8hOkdJ?{p-zrw6&|-?~m$-V}xM<pAEQg@V7kvUyCX1;AEBDq6ThEs*
zY=(|3Rk6grVXcs0DU7g0scwa<3`jwnlx)b_oZrehO8o{hTOJJ}L-%*^F|wV@2Ppmv
z=VZlsV&fEs%`ZNKqvi#?XJi(mXzmc{09#6T-XviZ$7$PS3vA&lEK-aW4@723mycXK
zHDK9GisJ+T-bYG`EQ!``t^4g>fY(O~n6yOB3*IXnDqlyfK5Wa3r?yX!)K1Q*Mj;eI
zkW@JQE`N*+Jgy7egjoP~9}u3+3*n*N(O((JU5Bzn!o7zw9@*%bndZ+o(=K{B(~&dD
z)JT8gK>I$s$$kD|ZfQ3wR1aP-#xW401(UB^)>3UThl0_GqvP7;v=z4W%)nCb7*X|d
z-1_ggnwnGWFC+F@t+5hzsfT11Il|B)%aBxB%(XdVnf6w#e?|{ZniXx4!Po-eZ+*`e
z<EGk~Z&?$r;WZ@J)k~G{kc9vc1v(H&K?a;{GW<*iC%?_I;4DDF%oHv61G*Sd9RNg_
zD;K`)M2i3$Wori+D1Oo9Poxkem+!Nn<dZ*P*QOX)w}Os5u>t7JL31^bQhcYk*ct9h
z8323?D7-U(Q{c-4<+kUDV63Rl@C_dIP^>xv+&}(3D(=_9?o3Fq!77dnu>$IVz{FTk
zo1k4?fS;RT(HP*HIw4;iI$8ER(Q?3U8QBS2YhGp@B|V$ObLT;;TH>VwHNmCi`#bsg
z__krjD#G)a?Y8(cKH0tGVu$HZgP;qw?)6ePE!C7-2@P4Vw5Hl7rv=;xoEh`N+U!If
z^h5SK^G@!FRr5gSAf8+VH2Ew_n_tp_*9M3_9AgBko9JHD<Q#IouEvpl<?F+hRLfEs
zdkm;;&a;Z&Y#WR6ihmk4;!YQ^8$$BD1CRnMNV+XVZK7f$C|lB$l?r8);{eFFd8c!i
zA4VmMVw|Fc8HY8-54rE9EoX=ZAeZO?AVQVW0d)v-7A<;L#`ExZLN&(h@JD>94{a@u
z6GPHVW}U1a-|M5AmL2YLtLhDH+;uG6Lu^TJLA}*0b2d^#!SU`jZGV`#?FTf^hK7m!
z<OTJ5P--2F1{s{g71P7^U&NrLEArZ<byzQrL>V)Dj>>42OJ%<ZJAM?xj@s*seA*Zc
zuVlZh@jZYQz$fp2p^;O}Z4{e~d_a~F^RmJ&k>lD&Y`UX%%}aKH)UA9j{Vg8qu<g}`
zXmIx%DMRB9FAR#`A@%pcSwz*c!7)XQ`DW$bH%iXvg%_zj;nRns&jzf!Oo?k>3$4XA
z!}PXN#v1u!gnJ?56gbfcCZ{+{D@P}?)vR$90V$$2<{tJhF}DLm@8;y;OD{v$rxuhJ
zR!!MmH~WY=WV~q-ylSj-`cANIEngY#@<e<`lhUIv4;O=Fn&XZ`xOyq5Ogyo&@Y?g3
zfa^~)x%(y0Q?p@79tE-wSsf`;c-a_Us@r5{l;prtc~hIjA~y^<<g2H^l5to=R|35G
z`&cOrAjRv^a1O@1+`?-(_XFL9BG+l2+rVQhJ|c%K@)O^*-piMKyqoxxf8N)lzHb+b
z`QRmzGc`I+nf$bH3zU0{6CakL@bDq?JwWC@ba1^JK<odq@17noWI0N<Tuc5zHOvC0
zrD-m_w1I1HE<~Mdjs6)R;`B0QN4B5!O@vWsX$<Lqk|a}`#igPve@2^8@iA+4_(;c6
zx0n!~u&|rlFCyYZMIVs$)4PmC!LyFTYV$8kwv>_}WPQP6Q|ND;an_w*sP(lI%xWzb
zLfV>IAmbD9?k`%`bU?V1s7(i25%Koq-DF=JfrOU|&hUFf+e74;{_u#peLP?ekQ%-h
z@$(J64HrB?vO|EMJMa$vOCQItiadTb#tl^Scdga`X5)_mPSEo7giqE}efV^40p%i4
ztN~sz`f<DVA(i89yXz4~em65i%8gO(m64?ZsI=6VQqyaBxa^QfgXA?jw<Q~Uv}lcj
z)g&>9DyI(qQKe870!MI{_>3uqW=Pt3a!o0k?<R~S<j1_D4KTsxj;hk9fQ}}2q>f4d
z+mZ~puBVz)VO@8^j9+o8$6|(3UCA+6nJ5`5DzH;lkbG{?jU&ZfvQufnvK!6=A*P3L
z{C(A<t<R=Itnpb)7H&t8+s$wU$LtB6itbAEAZKjZsMe!>1M0wE9}L?pcd>uz7XeNf
z*{E(*$Nc`WdH&v*Tw-#`U9^9N8bqg<F!y>ex~9U$L~M&52#M2aqp<jypH*ndd|2cZ
zl|!7qmfB~BUCAxoYS0h`WO?c>G-8aZX29niUKF*q8H>6a;EVm!=D)dZ65v;pqD6H>
zjJ4-$H(!}!S?n+l+0?wotCZqSmLS_g$L@p`14gA=$mQ&p>COU}=UAB)JhvW8lOp*S
zElm{u9yrFc@i|R}-m8qGbJiyDGEMj^SeSH1a?ozjmZ(hM_9+9Gyrh5`ihnC!l|+8J
zg2}E7gkc3*PY=|~@UX3%tqn0Bd#;t)J<kHi96ETaf1=Ck&(Fay15_iGrg1}-ES6>_
ztGnoo1eIouWqh2)PEY$$&2<|6*Er~Zr%N)j{T~Gmx3D`Lso5A;)NaA?(9Abgs8kb@
zPpQm4swa;|b?qAI)X6_SHUO|rvT-6Jg66`R_=|3!fqma!BJzAaJAAu74``Rh7|TLJ
z7Dn>O1L)YlQE~9Gd>RGmiHFV(M$T5$ru%kXhNK*5Np;bF+}u}A>=fB?2)T9lKc}by
zdWw6Y=W(yyw@&@szR|YD*d1zpNW(d&{H45!eej0v7}*smD+V|-X}!L$4o=WzNshb<
zgfNa1n1AUst(kBIMgDALn1q6umN79aRuURTLOX>tn<HLFq~2e`-qc19dYotGfxuZD
z#`iz9$Q~U00LDfXnL3Osn-2xYhd}+45#cimjjVr9p+AX8<Wb)B>+dS<&<x#)2xMpc
zw42cO358gid#B}<>Jnd#(=_8}BKIEFY^|ckI}S;ams-r()#%3}n&hTZ((<j4t{9U;
zl@I_~Czj-g-vKMMCFu+oMroHG6yUm$4Zd&WZ_^$arC<$Rm2%scwmg8@jybV^hYsQR
zIrc(ja1S9$C2mSazONL|u?>!aphl1OMwK*FFOetRo2*D?&gQTfej+;Ro*g-hps7zp
z0x1wSiq#G({&rB&6Uh)}KT=@=7Q6MrKuoSnmo1qpDNA^_3J|t3O|o%!&{7p7yxY!@
zWdYPmj+Zu-Ks*>N1Rir%b^vv9FJVLXw2X*-bhZr5d^PG^yC6$zxX-8~HZ2#}dwco=
zEf>Xs4*T+WtgAYe$<J=wdH1c0i~2xDq1yh%{OS4Q{&ph_!})6PsNeJB4oR1cBF)JT
z*!}eWxY8r**Zt`fygtlEPT%cd)O@|4#c#5FUi?o#UmNj0zgt{eP;;;Oo;=g+<<RTG
z1D1$FbhP*6t65Qegk4>BwijHj(>AW?kObMa5sC^YEaY~N<w0_K4>hRC69arVR>PKz
zeX<T7D3D_0C_tJFTLax8X&e+S<WEt12K%>zbB(}><xAlO$T8GDLU$X{e1t9Mbgzn-
zRiyxR{!ZI4!B`A9+$jfW=M*M^LSqWJ<)RW`r@SkN)i_Xh0J!7O9Au)Bv|N#s>r#k+
znvAw=Phu6>uVrg{R6$k8r00zwi_x-}D~n|UF~wA0)7^$}4+R77!YcXp`)?zqu7-8c
zGQFCo(+HD6eR7mGt-&K#$a}k*W?UM%=2;b|L-<NBi8C{-=}tqLFS@G6H~eLebfdfM
zQsjOv(q!(BaEL98G6y*@C9WSPQC)=5Mo@A`eg;x(h2?#;Qc;`#AOImXOfk%~=C#VB
zF=+pcyrDP>M@Imz@v{kUH0iQTXVyOxQ@9%Did@g|xQxVAkeOFd3RdtNQ)ro1>}o{?
zjKQ|4DYKhe&P?Ij-UQDxKT!dg1&8{V7)<l4^eWk?!t#g2#O7X=Uf!lZ<-8-?`Krph
zQB7FJn=2J-&zmg9Q_-@x%VeD{EDyJjCVJu=xla3@2w4{!vpuCeWZky?)%d=EWx-o;
zUXghvSL14AuD;@mL*Rp`+619^Goo2oZ)_`ga5g(a(K0dQQo8pswnSX~X<SajEZ>w+
zNnx74XmCeT#WS&0xdfkvZIfL(Td#~0o?0^X%b%(C&tX?9MXV;#2xgun;@k(LiOik=
zTzpz>Da~{P*uuUrjSHP4fy;!8-HE2ul~oz-uQq|GQ!J-GCF#S$y0ZNcjN1hsJ4xs(
zjG56DKhSlq_s&U4P!fHM0QCuWG>e`#g%HnxqH}PojRm1X{7sx%u{s6j$?#L~c~%F!
zfHdj_#N8FWaB&8t9XY~O7dd|hHO=r`B;R#_;uRN^s<bBrJRs=QA8C-TT|5Bp2Unup
zeY{o{x5Cnbps4~Cr55VX$*QW>NplXtt%BN8(<q0>hEOPf{qa`^ABUgp=wd?x10a@5
zl~He+V8XuoPMoznNQXOPnVNQOo9Bzx3-eAAPZe#RMcjSOEvhkMBN<O0?eKYqqf3yg
z2?s!`LqGa0!2ONDS&D^Iu?b*~C<&VRoC{5TEPb4<jQ#^G;(vyCPVse1>44|0!_8K*
zZN(J0OGXl@&6RUpKKx*@dn|pg$m_y1qlo)~$~mU)F-2Szg;BlDab<iCSJz|N%D9%Q
zS6U89S2BvUV#{t8Sj4UN3G3d>r5lVnwrTU_Pg4q5ORdOQr<Ug0&a-VTMZKC!@cBAv
z$t&oZUcUC8cJ~<h!-B<>*Woap4*aaDq0RPGS+k5Ks2jt}CC2`kC%9ZWusCW$F2s--
zl{Jl#m(>nbWt-3K%rPYq+D|q>d=si@;#sRoS$2~E@UN`E()KGys~cTeFII^XH=^Mm
zRO^!mm0|ZjYSG-9!0_r&-M_~ozJmx<J!=+mX_tCU3%9%#O=!f$_s~&eP}KYH*0cuI
z@sphNY!x)IWVWimLo;j8HnyV>T~WupT7F>3Jd8%weE-aFC=1@1g38Hu_dVf$@+vGk
z3}jCM9#(c<&)~h^oL->-Tb}(KPU9Nk_`Xgar6}y#0Jt8N0PBsxF0u4|`{zeIfV2e6
zBR~jq;Dc7fFx~aRj9#c#do6kTb!cXVWN8uEOEuP+_qed_sx-P?Ym@q-IBS`k68oxf
zeZ+U0Bcv-+TU4*@Gar1Wo7}I`#I-2$&gjYFJI37Y2If`U=x5osj%>!SsII5&8;Nk!
zUPH+^`u@D=fwk>^4B10>rxxA#Chm{GpKb>3Z<KCf6F<rJ;flBi!4rmh+$}}(=Y!QV
z_X<xf*R8E9mam#gWJouCNu4zV6>$Dyo?Y2ET+^FjQz=KXgUK7Djc5V4Ui>(C%#)@v
za>hz$m{b}jyDKASy+=Hyc$r@L+HZ`n4altN)ePxd?<FQx(gl}lb){)G=Cu_(btlN=
zy26^vq#)fFG=GVpnG0F2o@i|GN<D+J^wZJbB{q}QD_LaNc+l6#m!~Khy4ko)Xnm66
z-66YQfLnrz4)PA(q<yg-<b<@OJj-z49;}4mh)M+Y@Nj!DZ2QuvkN^27x(|xvTU{8B
zWwqLu7LUgsHLl+A2nC|E=(8N#Cv{Gi5UkC|4R~vuwB1q~xku(e*<}QG@=#_?W}H-e
zQu~S%zXKpmqEJ}Cnc@dRm~&q4Dtcf5!z>QK&tDroIJjEZua$Toa#*{DN+|KJ0~v1f
zK2xF3pfKNMOveecVi^)Qs&<l@jy#ZZX?UhT!ltgL*KvmxsWJ`yjgxHq>w0x9;R=5z
zL?689^vaV*Trlcbf?FoF?hvUEv2d+qPPTK_R>ToH{i%JvH0AJ7)i|KNi4ZICHtkew
zgrd}f-=fDUDF^)7<;++{M}jMq%>rCz7voI(yOSuS4G*@ia!o0~i=WVAvkew00X^Z|
znU7UEqjk-F^S(xl#E=ddUG9v%5Byoz=jjJ0?cpZ;uxtm_6X#e^LSiHYT|hH^N3|io
z^y!edi(JgviavO>zLf~pK9K;81DKZPQEb<~bcW?#DJd$|#DX6o$Gth|&ObJ!PdDf8
z+~&$_Tk(fuGLD*ylz6MieI6iqZn*<iZ-8^Hh~c|Nzho!FneJi64LM_I!PSd|51Y@d
z7blOZxVGU%>w@$^EpgK2gFx8=5Ad1+;x}*X<H1#zR(U($gS#cd3%`sg!~HpuK(Z=J
zFgp8i{LuUh^k&`nifMG5av}TBX`0l!_wI5JX$pPP+CLFK(e5JY)?EJsJRurf`QL(M
zj{hyg#K6JI@P86~Y{8z0Jz%}et-FPfR)trD*JtP@$)cRqit0e+LEYlHD){&bQ*1m`
zZ%&D}IojpshLO1Yx*gpcAqd-%J&9XLa{~@S#!Fck7<mzx!zcMiCBa4mCrfD%gT?C$
z=jW56sMR^`9FRObdvw}V=H~6gFpeAy<y^6X4j$Q!DUyb_^Tef-_PN&6J=jCeeb#nu
zH5`6G_o($A1qmuSZvu(bt6v=Sc3<b`_vxGXCawsA%{R>Uef93}elrMf-i!$LbUQPn
z^a+eX8od%EKwZr;Q0EtRQ^v=>-tpN*^d@12Nc~1BAsEuP&VSl43V&wIuRXxlDQ?Su
zFPumiuT6jG1rl_}i;Jlr9N;Yp80_o$_GF~*_WgO~%6}p<JIXQJn79IRl>^YU%Vr6(
zE!It(XU_IeP=YS28Jl5k90OnDoXh|m`Y<Yrl!P@zG6UJ?3?TDPg57!=LI`sNu%eJI
z5gy+#%Vmhh9J@Nd_xZ}BKmc^!DGOLwzB#Bh41tN<Wv7Ok2n#mDz@#Fv2lJ(Tlos63
z$uwe~0x|Cp&B&RSGj>PwAm>f}hWX&_)4<4S%P5}{Gcc3?7_PD3>6wE%P4nidm)6_>
zlwC){(*%(I5bxYEAaNSom_09e&$U@QT;9#HISp|H6h^BGX-k=fq|OYbY>>EXf-4!o
zzcH7UGAC#;fxwonF;@XIG=jyBiR5;CYIjm|cgLc+{~PFOn0$Djz8E}zCGu5amtOt^
z%e>q<jBE7=88x!+?vsJR9%hE)ZQt?;F%X(^KL@>>$NOwS8#1eX&lRu{EK2$xD_XOo
z$49M}yiNI3?Xk>5B#tp{mJWZ8+F71<@R0LrPwS}Tn6pG0&h`g#=ms($FndxV@hEe!
zZGO@ohcOuOJ^!6MS(KKGfX`#-xH<}hBa$~Z;bt8tu;~K0VRt@c-iSnDMCyAbV!Aqq
zoBO`UJJeJtSAdj+#qlWrqHbEO{L}IjSW6l;vNkd1adcbG`qw(i4<F<WgSb%NO?tKu
z0zDODB`6S1H-41kv9l#6S02cq(rMA1Z_TkHeKQ$VS#qd|tu;GD60YoyX*v;u(~K?q
zuO*zG3t(3GjMeIQcd!6E%qIkm9#yEO(91pPJbp_e^d)vqg*x!b9lf8ZX29vP?9N<9
zbRn+<U9o*{mLXp>96H9zzcJw=R-9MJ>BZ9_dd=d0N6$`WzLStqp+<a*aT*1efjf(F
zJfq<?vQ%wLI7Y;${^+{0Uue4T;}yf(Lb5Ew6y!c^7_u&YMR|4-4ToN$b%Y{9H@kVW
zZ%mW8oPMEqvU>8K`rlEZFwLmjay}=@9G#33t8_tFPVktE7$$f$w!bRd?`<X`1kD<w
zrKAEOM4RId98NzMHTgG;n;2q<EPcd|Q$qb%2HZg1M}%8P$H_DXl;q9(+AYSU0hMk~
z;_B+Anmewsk^<_5{pGkBKt~;L1EZl!&>hROuo?USgekR!@iR?b*~)!au}3f;h?s}!
z)AX%m1{_d#3%_s+*SDr@lL>40)!byjGeM$yiM51v+y2YucmumDj<Y0#vtGSi{6F+=
zH9V*w1^R>`z%6EDAUH+?hLJ<dvp$hCN7m)$QI61KFRBt(>7XW8i|LwC8SHE@(2aSx
z`_04jXT1lX&aTJbxWs2*E#HS4QgS2tX1E7}*y0@H9l1%JhD$8R%f+?o*04I3ug)gA
z<BPUC?OxW`GFF?o(4J0P1-TCU1Mtu<O9OCTSuU5D8}{q>*Q=odH{5O((;QdByWF8(
zCD<p^Sn(b230|teC`%P6`<tsSw)aE<0uxd@i>5#-7Hz6wsAzGDT77~o+sgYg0i`Hb
zt-6@58(wik+9H2Dp717`p_W(I4T$A<o_DcA7Np5)Xx#BMSMI$ghHNGDO<5rawG1}Z
zfdSUK099=EOeccPdoG$rzp%<FTLsOp6)~Kl;(kdpKpOEZJ*Pr?SJxR%EyFtkSk3qJ
z2{GS>B})gW!KSafQ3v*Cw^U)M_b+=4cc0z7hk6;d^%(5qE`A++uDM}5lkHjhk)|Ab
z0M=8;m~JN18U#LmL8goS((%?|(fV4uz9jHQkPO^j^m#%z%ivV}35J%AcrqLYTuc*5
zV*h2!+6*>4P_swBa)N;^X&*O@(X5h2i;w37eaRV(h5ErO20k{7O`*JVQ?dT`y-?sa
zLI`87W|95~rL!l%{XCxi5y8yUx-KM^8Hl>~JX(;!4SkLoMwIpXOpuI7qYpdUVLLWb
z)1sQ>N3jpsz_ySf%kVxT<zkqIroG27M@fK|SHrLnr_hj4hZ7z6h?%+dE_Vj|gMF6V
z`ea{6@DX_r_vX;89`hf6C|r=2U>=ZxO#}Eoda|n>o6zS=g2*n+nx_nNOOAhjw~zUR
z*{WgW+~7m>eC}|O0>lSKA3;GPg$DqWYI1#7=ZYsj<B#QgCaMcIH%j-GIXV9l24s?x
z?{GSWV{iu`Y6-izqIaF$U<B?DqO-5xDaHrI|Ix#EWLVwrDvk~+0$-A&^1q`&H3R|i
zE3Od?2qN75P)T<!pDDdZ|MWX(BOX<Z#XK6_f~hZwTVStzhXnTM@yY%G|3dVT|Le5>
zzqt_^82*ns&o$~(ztcWU_X*`EP!Z}A9RdBWF`0;$Rf5)>Qou7ah3n2wJu|6d;+3)K
z1|D-OsU#7f@8{U**^Pu8Uk>fdLRyxKlJb}k^TJmhq`km_-Yh%jCkmi)J^j{B+0Kk+
zmj6NEfQ1cBZ1@dip0961ygU#iT)JK%@Moj{Kyq2fUi<P3b5He!Gu&0oiDw}l*Q6jd
zYg6(tqWnC^#fq0VuE+Pm_k@I6((9WmpN$N|c6t8S<H^U_g@njxGc6taR8~*!taRof
zIj!&88}9t$`^=eIf|Z^33bfwy{MIvMdJF@VRlwSMMzk|T%?lIY1%hiBzq@chG6J9(
zkKj=h%5*isd%G{H2KoeMKH|0X{x3a6p%f3&)FVGTui~P=l0sQcq34^v8vjGTYg%04
zw`Y@qPj)#C(}h}Y0IZXt<?hl3CsdZ|w#V#Ae?IJ=(*8ggXDN?DYaXFu!31r&lCK6y
zEWz*=DDQ0;<B+@<1GkWKBBhx+&A+C<%1dVxsyX&~UKfv=SYUjmIii&ORr&_P6!W#g
zhvzl8Z~Rr4UG8Qg<TVTyRdT1b)HAyvT6Vje6YKJgfF=-Kay!N|?xy8kz!?TrLVmVJ
zW*bthdZ8Jzx%D*yLg4~+A*)Dv*{K&blc+1q)QYu6xwkJ#s>vtksAAC-x*VbwVcZTm
z9KRA!<mF+GKxe)QO4>cxYc7M3qY_7dl3&B!$(lT?1njc^EvF17OaD<DQ|d9NQEREe
z%%0Dq-A(@M36+955~@n0=wkY}&8>AQi$69G@{pH(tWoT+s72w$YJtDe73GSfV=eNY
z%3*6eolT-{PlCkqv=q;vZ`@ULQ7Jq!H}ZyY0eOytL^u)(+Hod)8)=@xk#HwWtg=u1
zKEWc_vyL5Z(s|oACFfq{U;nDJBYcabko%C1JE>rPE2cx}{Di&Cqs2HBv1(Y+J}hq~
z!B7n|Kl|-A%SX19oDh^!Y@Q54zx_080b&faJ7jy=lL~Xa?hwVEr4e)x(isdh_>DYl
z2T=biayT^_oJeiYWo~U_3by5{b8*SG^j@G$lIn68$}MIi>jIl~POO@*6>N$_K2mT^
z$nmnm8-Ljn;I+!_MymyxX^GNyuIATQ-lL4H2|bq&=0gdQl8|<1&hnmi=q`j1f0{Bf
z7*<@jV+tK)?5iR2dd}7MHDHaHwUg%-=z;AeIjmZ#P&KYFs*I;U%g{2D>A^M3H(}FO
zS^Cv{>YB;1g`E07XA=uIgrepJHxUKSe}j~w&)CbS<6D}hri-VAe@R_Aqj|%9Yf!J(
z43a~GHHw5K*L#U535_Jb`#ygYpS?&Nq!Jcz6AeDc7}3Bbjwil_ufuh|Xei&;2wC+<
z<$Z{g!m>7Nr3I`k*k4kn<w<0QlC*$z^6GMEe6kk16pS*Q?{v({hAy<|$cqR{Mr4QE
zfFM^Cw>l$eb+kvE87N$hr%yL^Ae>E`kqouSTiEJ*YUF9ZwO&6qgm&D}uJ8abvm1xM
z=vo+UFfj1U9MiBLiu<%!VRLVVGS}KVSd^t+p!c8V{CK*7y|eQ1|Erh#|7Ak_kJ7#R
zuhhMXku#l=i=p#>l<y^M49raE)GUmh%?ViP>FLBxEX>TE|L0QB!r4jQ#8Jr3#@^1>
z#MYUBoleNk+RjnQ-oVI&PT0iN!pKBXOps36#MbP$G&2h`1N;B1n&?C=tes69=|rs!
zoK1vHjO>g}czL0moE=RJY@poNFLNN7)wF(wr-V$mrDK>Ql#ygI$!$Ajm}hPBJExM4
z`)2c*%cMhuOGqR_6sbbMjLK19757<&o;-C{E`I)aZ8iD+<h92zdgqyFoLn}|@P4$-
zaC|uBIz7wl$PgunlORP1|G%e2qU5~1y!dzoWaNkAsjSZZ)r}20_4?hOpD(*Fi5yNB
zsJUEjk9qwJKELm-loS<hZJZpj|CIKrsMYS6@A2~^BqR*zDXFONIGruv`}s_wGmXRJ
zseg*&`!46}3xP%_T7SA&>G4t>7#QH;=7xfTdhK7S(F!9D40UFe`2CPa5zW7@G!Q@#
znVFf_>n#_%eL+qq(<uoF0=wae#FCPd-GM>+r{xMo`v(U-UbZ7~_;E1YzMtuIIQBTK
z)~h*OZl}SCzco60INIL_svaF1yF(<Qrj}AuL(eLe&3Q+~`E;m^3i_Wa`&2|mMyB6D
zLqS#kdz#8(o1dRw#lpnY$O<|-I$E{*`TpYN<?ZO`XnBc=iu!oHBi3`bu&}VVr~gyd
z)a0bBtgNE4M9%cQHvsPE=SSjux!QouX8U`a;I^*kTQD;S99|S`I+q9grS1Er=f}Xx
z)s;2%E~;KzLnGqo@bGY*vPs4pLxk1P$mmYI-fpM6x3_nFtyYS@tGm0~>*f0C@eu|d
z{=|oSbYfy4z|qmMK07@<d=MU=7w5No#+-G#+#S9@J<`olq=bcquU6)yUrX4D#zEd5
zA0E7(FEy^%+4n1y%Hl~e4x5{qJ4b(;w*n8B#}h^RZ0oO^4q_^uUM$2ysm#H_?mw@W
z^%koUAL-=8M8THHzX#+!@SYwXlvGqux8-tqymh+&_G5rSs<k^5m6em(G6MqzeSLim
z3<`S7K$=xn-g(PQN=PXv(lau~UZVI7c$2a+Gb539We_qxM{c4)K|u#XVGkxE($g{S
z$B5cHJBf@>rgNCgrVnF)`LtA3hf5yvem={z+HAeg%+1ZYJs-nw!hc`OLQlUNWnp46
z6!p4PrS^b|JB@p@_n+Z9NbTh0q@40wpoSW#3zDe3vhv<RUPlL)LPA{J|HjG9EzAZ5
z4ehrTZFz7&K-OY^$4Quy*X!+GqsfFnmArpjVJ@!^XQ9i@Hm>jO+rtSv2+Yc80`VR~
zfcRmV-_LhAfE>zQTRQ&XL4HvYu_`VO&a3ZSu|yJr@6)+T>-+nYl|1|<fPnt8etvvo
zVj_|+oVuRe9F2NCz!^aseWHej27UT(6@~V_aRzT}7D{2c$@U2Zh>3|oG}YDZ&ytgq
zPtVW$33fW1z-tja)Ef*#evNKCdnJ`9Zxr?Uw`t(!$3{oneLh%yw+cm~+S!+vm&5bx
z>iq8ZMfj0vw^vuwO%iEceogPaSl+BKm)q%BS_iSQsVSpoSAK{c0^_cG{_*ZkfPm_)
zlU-d~`wulWH8Jsk*=+hwFpWu7_)KPH<4RVZvd#|p8+&+mkO{qNbYgY{GCI1Nf`S1W
z(y!L`fUZ88ouYz*0>I;JfzSnI=X4I2l%%BN*&^Bj!Q*ZpFl~KvGm5{E{x26MPejuW
z5Rnt1`#C>E|HV3}y)XztN}Ck)bV5M^VgDVwfEw4YlUPzsqt%g@(Cu(6sfDnou4eRm
zxyGdf-}7-V>Z|zO2(v%rM@IzQIjf1{>3mkvg5!P_dusYG^_QlmYI3>#$iNF4j7Eq+
z<PQY1b8<W^fT$wE!#^+de)dgBm1c5z0u;Ji8@m|k=>e$sEH6%We|^VW|JCIsLhR4e
z=kV#N`CtZ{9iG;ZQ26gatdD|OX_>g``H9te-IjG?_5STH=awXK5=EPa8$&=qKo^nU
z;R1X>@DdyAQ11#IGgl-={JC7Cm3O#XsZv!R$$s(EV{L7XP+L?qkd^oGa^r^%*MCAZ
zxW1k_^W*rzxY=yC<Lg7ycWt1=b0rO6wY+A0IGXVL3lj+8O@Exd-T6$h8<$SEE8PB7
zo|c4|7<g2O@IUU#cRh>UfhY<A9et!rmuINoduOx5>CG3yYlhqJ+q-l9KW5x2EhD2K
zQHWJX4dk^E`kqmBhzV}xX4F<+&+jfSK1$9p8lBtR+}vL_$g()MxG3E0DcbZ#1ecnY
zHu1ZSk2j`u`ePCKqiY9Osns<O(?GdUFfcUaX1{)X5VN!khSs#&Z2THXW?WDUpFHDh
ze^ujBaxw}EUDrvy!N78@&c*3z`>$<1ulnna?+c5H8oErTGdvlWTCLQE2km@};RCga
z|G!<1_W!W<o#9yb@!JR&**klrNTQOJG8+n^VT(#CWQ)qaD4PaF!-xiwkuBL0DP?D`
zGRn?KJm;nE|MMKj^Wu5)^yWS~-1uF;?`NE!^E^M_Y_)fvtqQ^|*Y`ZOZry4|GuZd#
zneNir2p-vu{rL{PtD=N^J4|o1Y(5;(93QXjwJc;}XE)*AZ#m!VK&Sg?@2Nivj^t8O
zQeCajpNj|3bMGE9eZNy)egW?U^@E**gNhBU<f>wURUK>TqMDi-jl8`4g|ze+x1T)|
zb9neb-eJahr?Rp@n)s5B<lbdQO*A@-<n(k&gX^<-!+j^s%o<9Ur@a}h_=apvPU!0=
zCMEg1uKt<y@wrXx>FM$OA*!;ib>Nn*-0TsZ$NQG<aIA_cZenAr?5SOyZ?l-PE&0RK
z&J@6M=hiJhmz7@?R*&=w<fFof8wiX;p?T-)IZwyG^W^I-phQ9L9~R0M!$%Po*4M4A
z`M7N_sd2Wnj<uxe`&=1orqv!|INV+T&=7rxoJE<DXVa!B>WW4o$M(!KW;B9?(2$U!
z;bFhT&hCL!+s?F1s`eujEr}Y0(<alGMn9*NNL!A7vvkurtd17!?!M4@Zt&z<_i1*!
zR58Q&lRsWIeE$$98+wnZ=jytLiy(9YRd<7ijg15yx%kqG?DvMwyUu(8-oJmy2cXJl
zXRKt(xAQAq84@_CaLT~Io9h!PGjf`J)_z;Cv9bNSbP0{a5N+Y=>_|D05UYzQ8yf%C
zsL)>O_1nVd4~Lty#1vk_Ynzs%2Mz?}dujbxB|g)QKlDd>$BxSvF5H|VRpc8tDi(#x
zHEuFGHKnJgZ@TxUqy)tT{gAvvjYys4XPj_HOjkr5RpZJ;d%ynXR^k&WMxX+^{&)Ri
zrZxCgF=~V$dU|^8W5-HezA<)f-*j^A_wj=VtMOR<j*d^!t{2_P<R^tU5%l#O-=nXN
zj*hlvn4jbC|JD#&nUZsL);eHrqJwQqqym@p>E@P}^Swu|DlY!)R*LzEkC-^;tDk(s
zt57bbi;W_?yk_i(wl-y)yj?rl2@Bm~1JVKhS^5^Q<o<KWHk~Hp!y~2qU%r>Zh)e?e
zEZt<;ehqK<BvyN;&~^0ojT_XxLE@?csjgCtb|W?6yjw&?JqDkrrMc{IVLnrLdka4J
ze56=qu$ySr)*gFe$+bVrvg2pUx37pWCqGvI($v)d_28jHHx@RCueT(<YWXs{S4t@B
zgG$S&&ko&1)wShgan>p6yvkzx45#uA|LI*@{WdgI<u1TNm>Rx*Ul>U5_~lb4Pa<I-
zD>zXwZWk3z&UierQ!h3)_PH~Q;EJ=mI};6!xTNGzXpOR=(+6t>zXO>b6XIoO&Dq%f
zua%dVYj%D7$Rsy=r!s2wckfzbI*V<6Ts(UB(yVT2gXioBLy}X(%pG+~03zh!N@PTY
z_2#%%X^%YxDvJgyeC+(aiQEMD$sTj_FFK!VYY8=Z89!&Ir>BE_UE8mp<9m}qUrAfR
z^2AC{Io{6bcdeigkPV;G6?$&s;K*KY60mzAENkDTphw2}v^s`Nnn1#g&CYW(!}Vol
zsz;6-*|H_2XhcVdas7WS9J5$k<{I-kbLNboV0?@jW0e|?%qG_P%9URW3pcmh^$$ly
zMUm08svP3OQU0Gu`2XAA#`F&!q~f3;{dkk~J+iVOHWn3ktpO1J`KGd#$ZETinwq*i
ziNun<UQ}ctw)Hr<+<l4d+t*fRbtP`rNI%^AP*URNNUF~tG&<l&l{-YAm1#n>(O?_$
zY(837l1<FOqC}IbJt}wHP&q4#$3Q_xXFND1A)&0b^$Lo|<<8gKJ1xslXL?J$w7%U^
z(Bu<kWF;K=5KIgX4o2g$czv<y*N?BS%!>_9ovN;g-+#Ro^}pxM^~vsHc?E^`jt&x<
zpP8A#!KIuyvi{w{#N?;d<){FDPO1WvDZmN4E;RC;In$47U}|j4&C6T=?%mvE?;4<H
z!5>sVq+;EIi2GnoxP*j6{1Qjl`eWP<A|}T-mQPJhWtbOlMC@(bMt&7I9LVl`hcBVI
zxfz{qScHq6J&a$8Q(&j0B>k@PVpRTz`>wJ+P}d;EeV|R^iD$<I1O!~=f{9gX142$w
z&6)~ECMIKjePn`S($YX#-;m7JH8t5fh~cwm&tAE51-M>v1I?wjj7Nci)#Gii{HhN}
z3JC9$mS$RG5fc-8mYhr(!4w>ZHsyz}`b_nar@HIve*E)$p;Dmn70V+X((fUC23b*p
zjEs!AnOTgKc@gRk9ha1&d5fyljvX(5b{7lobrly88JOtE!5G8A!QtlOqQNICo9p%a
z*YuY8f$C5LTsGd`Zh1JS0z*T{Z!aKar7IJvf9M;^Suq2h-7eXWlkls@xAQAHqjT9U
z%?#&Qe@M#tpM#ieRBH~$SfH(|TiadBYwzG7Eh3Wi>Xiy!doY_Ag2{cXdD}MJgRh##
zviZ>6d-rB*m^N;t*cd6OBDHxlk%(hIb9jZYePN>G#L1I;aF81|5c?}{Zr=TU<oov^
zcFRBMFJB@otZi-6o;_>JI(PVFo0#t7uDr{<tmUOYfBp<ov1{KM@Tr3oSEMue>^>ya
zE`Mdh2n~7Hu3bP=+qS)swQbGh{nXK6eEPKV?%lUT>8;DNUp#*<Dk^HpH4&<Mzv}kw
z+q2(33hwim!Z60RRhw+YSH?~~fk@n-K;!+=BzI$-|BcKj^CH*s)=(~KiqM_G2*})=
z9AfC}?Cj+q`MnsbO5VI#nhCb^&9vO|zmLf41;vI9J0vA>$;1YNlvx1)!Q()ct1}1m
z^_|9=Uu1eHJP=gzE^!_{ns47tmYtow+iqrl{^0A13iYn&psB^_bDvT%pN))+T>hGO
zev^5gT?f)8%d@YH+`8Ad5TPqV+5AB6GEV|+@6EeeN*?mgLz>JxBqTz}i;6~m3(=VB
z>G^K<Pj&Pk7#L6$dmi#qP}$Sb$>~;fG}ZCZlYigYfY6JKVp21G{P?ku;n}kysY+6w
zBMq_0hEUEOdr2Lz@}}B;OPYoX1EEVbVcecaFs1@FzckHf+*z?imH1RQeifZd^~dU;
z-_H+bNoziP@gn#g5Z>ubOEn{-P1=GhLh4;zUDtk1(B?0$EV-mVO->e)adbgCsqtxO
zXpk`mr=_JGK781P@ZtS?iYEjHo~mmsr%OGJwY2C~*{_!Q>};#A4{hB2H}V;8E587e
z_C-@n$1myZ>@0g&TIy|L!XsB6<fi~^+ZWDcCp?4sN^SeEr_&g)@VEl95<7NuXPvtl
z5fPD^Dj~<O;?3uW4wTT3bQv5NSo!nkYp$)Bh{)~u_(&Q+#2e8YCfixQeT;XwbV>0{
z1tq4a(x2V{(*7^&>*~t)Gx~0%qZ<^zvsH(J@0Q@XbLV1&#1;=^-ad3&_+L1(nz33^
zX1{Brp(|s{IaXI!hh{4$C%0wG7RlYIRmM5i+qQ3i>-~pk4S+MSEFgf=;8ah6(*`9e
z<7~=REp>JO6tZ|4rO_Wh7<3Ks9$8uaQ#tIAR-r*bWI2}KA3l88)!kigwn<}?{7(1j
zfojhs&zYg%<YX%k4}~CV&dS=$?V0^Ub$8m`QI384a&6lR-bq`O?sFQb0?T4<Zf-2U
zk<}^c04i&q6&keJn`<18>8?zmJeYZ%GB&2YPL)*N<#(3s3uZ8sMh3A!olO0l3m-?)
z#uN0P9MX9pr=zM$W=51Y&9yn@>A8Y5Lb$j4m7P%7olK@@-&@+<*2Z;x%hs*b1OkDQ
z(siPp#XD-=^sr|F+pgJ}nJEkKBZ>jL?#9I2n8%nbY;Ux)?aLRo!tA^}>yENq@-m&+
z-Bg@z7+}U*U!Lpzt%*)b3IaoqAGCu%d)|C4^}3GNIb{nH#FF{G^U}<DyUt>F2PDPf
zX(sFQ=U=3yZHN;Q5fNfhN2?(S5=_61j8L0VQ&E9r@Fyy8^6{~%(KNK7k*ie#PrXiM
zk-6B>(J|Tk_GTpULzUtq6`#`BSndWM5lU)mX9AKdZm*lfjIo(nSPSS4(rY`JaLCbT
zD!OZDqUeB7FsU|ApHfTEM%Tlmkpp<{Ff}#3@&aX%zQ^Xz^1RvAiYtD$NXe$ACY0KS
zdnr$!Qn+MgWlb6P39VM<lB<Hb0DVO<Ql=(aaHjG`KtRCIkd>YPAm(5MF~ILu5Cd-b
z%%A`=v(Z4d8>D~ld85SJ9QXx1S9MjD(e+~&E?gKg(9j_DYyRC<G`n9sdsfR+Hq~iF
zc1xP&zhiGt<2p@Glarqh07jN*q+oU%jLL*SI`=2)V^^<TyMosaLIhRn;jVKvcekBr
z|Cz3-CbeRbXBT_x7K<abo_`p|6*7_3?3ESIKKi*chv+Sf+{RmPZ9Z(dtn9g1W>P}t
z>FDxyWrmKA&tj`Sn}5aEyvvp0a;Q`s9ML;W^U4;4z2vsiPwMSNYg}DjRsDrx>Byda
zef_8p@TF^^g&@aw@7|3b1*XdkDVXUwP(^DuGd-<;@?^uicgXkl`Jdg_kpqAWiHV6c
z<W7tN`)D$<wu%F5sH>|B@m~Qwwc)qHtJC4T-JP8!q@@q<vHRTA<mBplz^Y63IX#BJ
zYd<<=RoS-Pzki=B0TdzHy_L=(<Ao6N4b%YZcUo@JsAZR(zsbn&>gL9BCtqm;4qJo~
zSi*B76BEG#s6lkgmi`gTyGGR%y_DyuRH}2oS`mrCp`ipSDh57<sBl_pgNhgDFI<q(
zxY?YXlH!+xapc?=G<Rd^63iq0DxQLJC<^F#c|{`(RtFALfx9d-n7BUD!0qPdR&Z>w
z6sTa<dZzis3E+u^pWSI0mn!SDQ&0;KC7GE!Eo<8}{E*Qsgd<gIr{7$gm)N=UQ)T5K
zGPuO>33^%q^$AW^R_BGE=3HF&Q2HuwC32FuPfyOUO`B>(fZY33N=(5@oS{d>%i@TL
zWXLl<bt-TzGr(?-jISj6`Qe8$7mhhOO~=XFqG_ofQd29B-fsBodlT`^wqeT7itF}W
z1-JOsPo8wG3SvP7Y4k)fOhxFJR0VvQLl)u;tE#F#v3pmkN$etJj8UZJ-KW~I6w9vP
zKapQyvSx^G82$1r3mF<>t6D+ZCehXU`qCFw4WoSp6&Uq@U=pj(^xJ9wJYmzQeRmPn
z_HToOITxEmn3-i+SDzfOXa#5D;w4B{j;CvA&>pfv;NTq-W}?_o^d?sG4jGYg&(F^*
zDq!$u50jIZkKc8UI-;>nBhjhxu<rViR+Y!6pv`xHRzWyy0eL`p**@`Y@gOxdwJMYR
zW0{KXuC6|E&_8rlM#jeEx@eyX!YgsFUOAYVnR$Aa;4H>OeuXnF5_JRkQR_@fJr&gv
zAAURff=?Yl-wN!d6hA{|Y#W1~URnO5yZBx{$HpMqi9-htoNV(vVwP*e!C1~y^V+_<
zv!-TbcKe?HkYBDgBPS;(dg#;#2Ff|Gv^h*y07LsQslWFn-vDwsLQo|mBcpL|?erL$
zAAmmcXX5-h6O&H#@=Cx9b#-5vfOqfSp&&PX{mL6a#rfmM4;9ZvvJEsb!p2c_);hq_
z--l)~Ki&W%U%9*QIsPOjlG_4Lzeq`;qN}@6ICAP(c<F1tp#Q+{X9XHdx6#jmg<|(}
zNptjz&wgTxii)L|CVcC?|16`|#>U0jw`V4%r_&c#mY%n=BD=fYFr$i!R<{j=!Y}_h
z)Gl;q#~@V+=3r(jsz9K~2tGyDU2LqZsy%>d0DqQWaJXpH1dCuA8yg#Z%uGLY{HcWn
z($vuYe*PtyPx%L6zI3B3iUX?=`)B^=`6LzZ$Hi?tjk%&+{*C9-ZMxnfx9#+ufIj5A
zg=o8sv#o*)3Ovv|Y1;E0EX*uHHD`vj$o^0{clK<VBp@X}Kfh>DSx^v-Ts_nGI%OIX
zd(v}xeyqizr^Ig(-J|Qt;Ya`-0P=t*mc68NweD<hpJF9QG37-RYMQmbV+|81px?V8
zLi6Prlb6ivpcMct)`{7^zN?E*)h~rd#oH5Uokh<}^)>q8LrD>linchaPGDd3w6O5-
z@(%znWbb-<R*}(Qbx<_JT=!p_b98oAKj;?)1Z&t9cB$sZIQo*6$1txB)dNe>!(p4s
z6FTxPo4UEpZ#(g{0ph^b!SOqdlzZB|=a2p?76X=-+sjq`+;^VEShL@Gh{m06!-WKT
z>Dy&O3hBf3_b{hj=_+udQFIyzNJ|^P$t3t|vR81Oxg~#l<ISe~%)L!bEXRIS1k&Z@
z<<Xqctj==0i#fOLm4#tj3B}8drDXA~TMYp8K^mVwf1Zgc65I#4CE>2Bs$}#Gvty2<
zkN01{f^5~gClzRFC%RdS?Y~!a%Ox{CJxp43|9*1~jVGuAsDn4pZ|KHoo3)G?6Mxmx
z(t3M+fpI}OgY$~}VBA&APG<ezo1Wy`b@-PhCyPp2UAR!WkkynN9u~GdKQ8akvrni2
zXrp!V$JhK~_sK)c38E#ZEiDuL{VUMJJicD`cX}5b9<Jl|Vd^5PzWTBDdxx#gb-un7
z?!!a%UpC4fV|@DbX<Te9W(wbJP>SBJE^iOx)hdLPVs}GG%K7TmC15b9DhBCB<adIU
zcrq9oDxPAJ_}P&|phJZo`<gFJ2iN-7nxGoxU*+RFPM2(&qR?J)rWkyYni^7h@bKZy
zDhhkgc6M|C4joiir%>!4)O2#%l`bOxLunnqvEVh$(>$F(b>-&fE>Ru_(}R3L)q0Dc
z4>qf!vJxeI`_`?T?Ch@n6;xfF1!u?$5VC!$2y@avCLwV<I+|bEQ^Dl|EiLU-urJT;
zvz51yvB<#BWo2adq9P-iR`~hi^Ya&=+n_&IRaYZ3eB~CE<Y>e;$3d`&2qk8R^@}$)
zHX0ZhXfkNxMca*C2>EaFR|zG8Q#<>ID>*s2T`%@<1m79?CvkB+7HffHagVQ|bx)nO
zi=j+rI#I(|RXPyPE6;d8=k@EVgSQ_(q!;o}#5`xWY-rb6>ZQaN9~w&6dyuNHLV}`7
zbK@MPe|f3g<xU~PGHfcyD4+N{k1J>{C-R|V6#_AhA1DjYpj}#CuJHfSRXB@&&?4JC
zXZZnK?wM?W|CW}P{&$(7#l=d;6|PKRX!&^pqcn;`PR}sdJ21M|y{lvj0bra-km{hw
z($dlt5d)QF?fVZOyq4!)WM(>I2H{IVBQ7jdWIa%0W^P`7SST{zT<6lI(aM`lPOg0Z
zURI(c!~mo?x?O>WAO<E)OpK6E)k;H43;Cx1^mZw!8#%Lcb3<RZ%gGrWIAFE9vc%?8
zYO^<QvOSA9zu9tg<e3FbTD;ucXFBTZ>G&uQ-C|4<KA@{RIn`f@dJRf@=+Pcw8JQD@
z4-=Gcy+m|E4J<}DN}dy%ykNlo4_Jw|aGLc%sLQkO-Z<to`z6t)JKrHbEUY0`=0X9J
z57@E3Ch4~;{$+^Ffzo2nr5T`wGt~WyeT}KM*Y?x%DL8&kKGGuVLMyxvNI_IIfMMzE
z<;x6T#3Ur%KV=x;cp&iy1wBFRcv)g?IB)*NrX<V^!WdENa=$*pJNSD>bK%a8j%4SH
zuKd`GAxlA)mNr;e<3`E#1u1E1T@#aKaD1NAS+tDP`Mut}N>?OJ%}YI(0Gj}>c{gne
zibAOu;@4ySwhU?>NeY(7pWAB3DPOdSNHiO@BS-3bd*e^_J+IS!ypNVx*ArpT=6Z&P
zi77un|E5n>DR=_cr$Iy&Esv;54U@7puHSsBq>A_;1n8-{=5rUcAE288UF&1Vjv?yP
z*ZA2=+VAf=2N2lL+|u0acp>n{4K!F0zN*(bIhCa@BMqZ%uX>>Zb<iFRC@4^fAR^iG
zUcWwRU~ofssGhn+cvIYi2ccGH<j9pbRU&0E&xuP&>{U|w5-Y+eCqb_2(ZjZQSIVMP
z`|pGFrIN*gjQRLbkxgE<jBw%?XdNJba2p8;LS0IB3Ku1XnAgx0FWB2FU;LCX7N=&&
zxx<9AmvDPv*QPYb%YC?Gz}s8d>~Va&;Djkg$xp$1xm!P1S5rB%b8!)NIdm0}Z$vkw
z^Q4?_?;6*}fKMiVJSZ~KA;V>rmGIy-NwckA#76>}j$N~eUyCU%DA@Ed3d1ij)<BP4
z-vHPyF^bD}c4M`C>tp(gw%>`<r%x9=BYOdjG2`q$$3C*M+S0CF(J?U)Bg+EGZBQ&#
zH?0W$s(AV4$`2D&w*1dPlyh@)MedW0$A8Vw>u740T@G!|<S2Dso*PAIqmYgKpo{$K
z2o=E5Km82-m0hy3fuog`IbmV+c3a={h%(CU0MK@XFwuLP;^#eq{nxY~$nDs<liB@y
z_1o>cY;DJZfCinM@Cpl}*FzAVXDA;4(>OvI>xP|3x|1gZwq0LY-1S%ydO1o1fK|Iu
zocIb_`}_AtjnwthFMNC~83-<|oHpb18=h9_En9|>e0>Q~QF=BuW8fZ*2V>{vT+@@@
z|7d=3CoOHe+2f~Aw;C0mtBnAmCLD;eLM#lj_&v<}X!C&OiDnoJ+7-Mlb{sxjo2YSf
zhH7MZxZFCMcDw!-WM|Q@Pt)tUl9^ZbqW+%rB>V|zYX!Ciat|?lfmA3luZFrd_4-9p
zlJn^2Eq3Z#u1@yszi{Cj3M~-9Bt}CeCEk%^n6{l9ClwSG%?ljKm02hQ1gm}mjR!;%
zN5@;zxK@Pj?e2<?j^-#4E$&wKUI|($`%WHI&;h4J?bmYf1a@P4T}J2yq?bmnt%`($
z8I62_lrzb*GX@PnSK|HWT5SY>p#nHGRlfE9X!O>ypl{}Gbo8`u=PQf1ARYbmkcMhk
zfk}&si-~9L%ygo}Wn|d$iHV8*5V{M#GRX5}=7R4!m1c0Jzbcq5)7e8D;nDdu-yw_6
z{ulTdu(D)V;J8rsm=cQVmjaHolqs*JqT>{jWnf?^P%pY2Ox&S{0h(u@ha>*O$fq#n
z_Tyl&+xXLrjGujdeS?Evz=eRnGH44^YkH<DD8=}AM_an_3B!%7tgMF)9jaVwg)c#C
zoLL|mg1^?0@YB`R)mN`xwa7Y`f)%2-J~`6yC_xz0<P;P(X%0}FX~jf?a$k7mvE6MU
z=BfA5+bb$65>9JthsDQRV2ne%L2mSvT-~stJw=OagwWqvQc^N?GWaV)<WsWm5&TMh
zeU7@Isrs$Ixjh2oH|3t`hYn-JKu=%wMABCQTp0@cr|{PHfG;PFjM^uBw({=G^E-d}
z@(#V5$~kYu9~%}<3yN&s3}n^c>3A4YsiTvmq~yfJ#Mn`_)29U%EUS7eK4^JItysF8
zI(s&$v{d=$E3`~d-)G93oA184)7;v6gI{BR*#`$kMo_a1n|78pG|(&Gq7qQL5>|a7
zN2c#IaMvTCBXkYsMw0?RTaO4dTf~#_EdWGfGL*L}jnT#F!CP^7)2Mph%bs(hGei}0
zkBcAEHR3x?voACQnshirQbO(G)zmw7EMl|L&(cl}nWpRd1D3NRLXUfvo=WO!X;pgd
z@>yMZ_>hlt2K5f*9Qc~slah`66{x7ff`TqIfSRe#DO#FC_>DX58<@^6p+fa_gLyRi
z=Dt`Coo+zoI3q~9h(L|XY49a1+CveHcM32n?t_mkAoJ@T5;P${0|w<!4Eq$jjiqJT
zF=qQ3l#P>Z^lTs!%9Xh)dUI`BHkNGxsb@+-a0&+PYG7{u=#Al}-qQUAl}5loXV07|
z=X#rzq^_&$0O~X;DM_-bt!)ACB#3tV0-g@&wDKoT7iisUaf<xo^YjHl)f>DF?!6%5
z(riuAQCG{!D130lcW>Q-Pez{4;!h#zR~`Xj^t<L)l7cL1?HDp2cBIa{%k_u20#JH3
zmI*VjO|zTmZbX5w%<1n#RQ{eDB|a+$OGjZ;#+=ZKvqC=|8X2kh0`0F)pqy<G!Cxl3
zZ+lEk3`vhGpxo0>8YT!TK}u%<5*;sRdH@sd3=1NDtLy8tJl%&c8lE~8UF-#;2g~K|
zA~&*&2M$n99`GYor592=EK0kYo9DiMIijl@85%lh>uz%LBs;u9kv}e@#FlTZRqL~B
zNj>%CP`LkWh0femzXqrMxpPm_(&%1KA8P#gvF+0*SW&1nYnA0e!jnl&qUVF;c@Hj_
zyW`zCA))WVrnK&qQu6@u2UdsgS#U}k>6y#)78Mi}#Kg3M3#kp`-f$;q82%Oq&xrMt
zDsU{mWNZ7ivoo9go_7f8F`ibdmC?1w!Q{Ys1=aEVd-n8!t^R8$pGA)S>aMPH;H->7
zU+usCD}h?cd&L8#th^lI7d1XN`lUsdpN#xdLPEmSlp_j(ks{{csayG*Zb?#*ALjIb
zcJzR0@s*cPo(#e;^!AMtKdA+xE^`0e{YzK_6~*sxfjqRUn1*tU9cG2jRAz&Ze#<H-
z7#%-;+nrM8JD^oTBnpU=;2WsRKj_-YgAg;d2OnSr6c=yLFjrz?Vk#){_4S3n0;P~Z
zO`Vh;BWb)FqU0eX7M5EWuy>dheER&ERqD90@e&LpDnI2(48LgCjq&kMm;^|_zD`Zg
z4gl6SE+k~*Z|7&PUVZF2jw#LuBn`kPbn3vsz_JeLR(%tepIlvCnc040V3_EAEBK4q
zetMzTCo(cpS5uRjm36P{cV_y{S~t1U5&o7UrXSwDQ}CGTE4eyrkYNVO`a|xozpBg!
zBT{HS84${-VdUS73EbIXGl^-~Syx~G48-RJC`go}j@$MsDn{SC2mK3XkaKK8uvT4z
zdbIC!1+TwXgG3Uw%atqCTb#(HP4l$P%(^Nn&H@0cegi<QSZ-WP7t{iY4NRHLSziv8
zARXlT-1jEDcsfqWzNg=aEO?t1KXe^1+RDnwQBQpS@+HlddWBMayMzSz8Zg8?rCy;r
zVCUgB92nn1Is5$S(<_TpXYw|V=X*>KfQ2o0M&@%y%!obSM(S<sJUm6kjA%a0kTj!-
zh!!FXkslze4b2_+V8^y?u#laCR}bF?DiJQ4t}kD{)Yp&W>q6a#lPn?l+zdUStKs3k
zc63zN*5)uyT7JjegTC$Pg~2NujSw9+1guC%NN3K4$n5MHbV={`zWSux8x@@YF_;Rf
zFZ?0NJLt(}q^(Ua^ls$Tx7?fPuw_ZWBlvV~I)mDd%ydQ!baa?5YP`q|6+9Om?P>IW
zzTdNJ*O?9U{_fx^+S=L(Mrwzk$H>d)e;j>(m;^p~kIYBM#qG6jxDT>lwp|S>Te)Dl
zn%NFV8{j4MxIQlJ4u2}TcHa8BJ7R;hJ5^MAYiqv&bQG)@jbBtzQ8}=o#J;B__&y^e
zBk1sg&G=OQ@w0HUCEvM!|9M7+G7?d4Qp4?o?MH4IYs<rO3xW>9drVeHBW?0>;B|{*
z$1)ysZ@^y;HL96OK!U7DsU8{Jkmye&LMUoRm4{eUv2bR7ur?A(Cz9i4lZ&fsdIt9r
zc``2wN(jex-$G>0_wV09qf6r53cWZ6i%;90(SR*hFOLAjXIffWf#S`1>VZ+9O6|D$
z?mc@*G9x72Ma7_V`+ZjT$;rjI>y1O<Ao-Z6cRYj+gELoh8aTks9SbULEKZruPJENx
zoZxL%%-HVkTx@KEkES`*_kaMvqqzvDy9*~d)u1o{<n!XV%9&8W5Iy0sL`N6Qm(-Je
z?=VxxMxT2#tF5gKzZ2acwKFoZoXOuF<5;_dq$C8<9}qA?^D)+gkwbafq-y5{)rW+!
zYcTRq1Zq;+H$iIxos>4;kO;LFvrNWfXXxI&dqGUS#yR#03AmHqXT6#kJ)R8Z*U~}3
z>408CXQ#>0qhn~b{Ciz3KUnsWFSOFIv<G;BN(iE6Z*A;<sOWMCTvE)J&!16YF4);M
z9AXX+4!#ugcgUa4<*KgSDHJpon0s$C<82;V0v#P4g8pi;0J3_!xOk>ogGa8%^x6IU
z_fJiu$|F4pjC__)iY+Z1neQ<_9K^s_UXFexDkhfm#*&sx%1j*o!NAwd>)}u%gv`s7
zVjKu7E-bucW0U=DPwQZ0xJCH0T#SKz?~Tz#HbLuWFf?q#+d*aNb7I`e*ze2J&c(yC
zu>sB^Nm<!>i~uA_=$yTM2Be*{@9XNovxrQC46drJ&3`R(kW@>ce<~}U_y%f7<&F-d
z<;X#yhDS+BB4qb&-I|^rtZi>^N4#okY1LL&!_P?)x}1<}dMXoN^Fc+rgV{VlQvm7u
zD{OmJRR?-r!vY6FB#MIT_Pu)qU)T`?dDX*7(U7;t*woY~J1qf2-TV`=KW2<5Tp^*M
zZQlB{3S;;)I(pT^!;w#N#|{SkyDx99%d1;X;}i4|)U>Z(zn*+2Dyr|tkMK9Dx}|C#
zO$z=-FdxZNtD_F102v`CCx^s#SvPEQe4p-heIbwzWsg}II=!OH2qwTmR;m%4@v(;r
zEId3WdU_~Kds7bW3<i<2(-rKh!qpio%+YWb!xu9%JA135Bj>^yGC)n1q?D9g6i31R
zUK>bEl%4uP9>XWJY+D;1Dal71hK<wXIVUr7!t}u-3}tHfWswV@h86ejt;l|9UYr*m
z7Y7`wDrZ^d3+N-nNJi2*kvBs_u24JzTt`1s)hF~r$h~?3PdxVJ857xicXn!nbrEo2
z4qCF`HLCb_%UEH?JR?R-P2~QlGbMhR9(Yy(74Lnkf(nklg9+d(QQ<lsd_>0j`+}-R
z(?UyBfmHx}-=jz75T(>~lU1Mjk^o`LD-5+kJTQ}dyuWL+|6%BK&oA@rmZ>V|tWfJ4
zMWHLnmD-o^`0?rU=ZDcOEr;QW?(gr%%Zs~pOITc7!M0WMQ)WqtCnigTu*duuM_x@$
z4TuX?(~jREhz8I;9;#7IRaF6A+mmO+NI)<C6<}yea`H;oJ~L*nR}(F1CKhWhtEqEy
z$}(I3{Hc1oHa}W<{ppJAp4Y=e?P=63!inV}Az3CSG;6tT<0Rtu)F~=T%FG^Sk47gK
zmr8$w+K7nDok1JvELPUKY4QHT4u|ukB_9Y}dDF<g`ke2(wrtE>uixIB!BYv|(1Qd8
z1xM<qM0=wgV(8h~L+rbrM}+>aeY*yBIlkCkMGOzo^)~e;(;6@ThZuJ$<IhKrf3B<B
zwtqkWeJxE*BhcsfpI>~QGy8n?kNc0#K-)gdP+>%XiV3#>ztYn28w~cm@*yEX4LNz#
z;vL#PMf<P0+V?6e?YX;gsd(j7&sz%?HjUv=i))T?j@fZ*(jV%UwlVK>{jQg0@UlH;
z@7VLPxLqN@ut&ADM&F2!^7e-M+XHILPd%JnqgnyFh7ZoYHQm^#yTEC)Eo)_w`U*^P
zU=`uR0!>8)j`q^Ui<(D{%>VexwR%f=4HL&T>7&aLMIO^4Q~A~(;$SvymI))Gi@^2{
z%MQ~1UPwPOKU4PF+{Lc9h?47nmfehvV?#shL4q<TB`3f3`h5ktfw^>ci!%TNLSkmd
zSwfd2Y%M-H4~@d!zA*+kVG}`kpU0%_^M%({<<<$Lkz;xM6^787bJ0ab%K)(*B}|#l
zB-B@dDu@aT*qsly05g}c6+FQp@N!?jwu6mOQ<G-AQxDz_8icVYzX()u)Ws<R@Hdm2
z{$)li6F|2jh1BMqbVG<`7!|a1blxDHz(ga&es&egoT~`~265q|-?%X(GV;Qm9!8!$
z-!VTG<bp33foLrvLcv56cICLCLuOXk`md;KZ}$Kj>B&cF%g)Y@`4xD}=On5qe66-a
zb!mvf%J1b}^APWfMov8gU#u;dv3)8FFdfr8y0rBz)WGsu3~1J20_Gg-2U6C>2VQNA
zEwy*{UDpllr01i&Xx;cQc<&epQ@nqbjlSbePYB1tk;0&VP<)_{@WOvPlxyG3(_A^r
zVY{Ic<=X~~4xszW_^Y>zcYe<-eTV!BfMgmFauF=0zrR1)A4(>sDOio*(f{@9mwo0w
zX3%^<CxxJLkjDcmF~%eW>%I18VJxmtE5qXLDx?KS6Irxj7+WtmIFO9+pvPTYW>G#W
zX)zhwmHCo`bWDFjNy7XhjH|Bw{at7Vrwm}~QPrTM3q_~6(Lm;uIi2?W`9+{oSc-P<
z+6B3ZcJ&;$Fin15-f7%=RjbA+xCk{j5c2TNKxoN;ux>v(Pk|)EREhD)t@N@GsZLat
z*FZr_2cBgk9MQ%xf^8I>^t}$2N8??4);T6V1+v$|l!qMfqTn8ypP4Cw=llHm6ucx=
z)v|m1>lOtTf=`~aiwk~=w6L(+EYt1#2#U=Rcc=PCgFcuB{AST^Xly*RSRKmMKXdHy
zzPkK2)J{BB>7pm87mE&9YHX90bptU6AP5zs^R15xGtoIOH<yYWm<4=A>=(VkugnDA
ze*W6xkhF^S_Dyt<>u%k;)gn8)+#mNM<C23NII8h+l%}Gy)gkEE=u%xJS6y(Hut%i-
z;rRPX<<vPiIf1w6XlbD@9>gD2RZA$ZAl=nyXzo9L+!KWcAjjHYVrpWN=g=$AModi|
z15P$_xtF4KuQ67JKnFzBXNh--{=&<}UQ%f=!VOi7{9KS1I6j>10z&^)-M=%G3&JPf
zCN91S-O&8YF>Cl_g@oR<m+i0$B8DG^EDR14qw`rky-&bTu+-4kV1Pg4_WRclOTGcV
z{nwfSV&Kjb60!v4(;|DW0$vTG7G>LUdwctNE@;eurxgZiI71*w+jGeB3Je0Sten`b
z!M8F>k%Q8i?=*O*)X|q|<R=~vbq*gG=v@jf*`G%PgdXw|gA7R_9<0#*j%vfZ$5XLn
zT#iv48UvPe%`gp^<lCFWW`^TI_@M?Oci_Q!Qs(CB`tZ>sq<69FXvGcs^A)?*k#0+M
zlqCNgD}vO;5pEk#KCA$VG3HSYkBoqMUO=S~*X2Cx;GhC&e=P1yhL@LDnr)mu<Xuo7
z{#1a)1n!sV=>dU(*XO>Uv9SqRGdiScenOojaUxU(za;>6kMw^2^ofiByuIgY6~iWV
zK07&7rZ26nHH9%zQDr_bgaE-Noy6`$O{Cz*`tp*JeR7|cAt=6l`Lg5Z$9R<m?3P%q
zyxMQP4gi3og@5$CF<py?h;xw7?iie$SRvccjh9Yrfgwd#R~M#iK*NVh?u@!>zFY%k
zyrn*CJzu{*aHDPh^5vYlXz6-oTQ_pZN#IzM9`>&;R>1xvEg`|f&(GlSu1J0A{l|~5
zP{EL4o7Xdt^rDW8<GF<Rr`TbpfSC`k7@?5!LX0GJfU|}C@8smPeR~?_5jb&btE$rB
zBn-lG%u-0@zXYPY`{vDl8-07y<N{t-5N=tAw|@-Q^irzICo}y0DSRhDA(?62M`5EE
zRDQSvuFjwxnM;oHnEC<9Bmq|++`kVDN=HMZdGsje>p8%3i+kjx0Tp2EQW3m`m{Owe
z-YsATlPM<zIWs&j@5-rD98Q_!u3&orddaMIx-WwL9|wJxF!kusqiOAmdNp@<_g|jb
zXRC2SbUo^gb#+Ts4ZeQ^PBT>ld?*Sk>^giTaEXlU=-T_-@ayYKY%DBld@D^G85lU&
z*i3V*bq*b(elMY-vWCv1+K-(g*zDB~-F1P?em{uK6%7-IB*gmdC-SdXUhH2)Cs|#-
z`~!fBgQMDsIT{kyRxa`yM`Qq|&r}#3AW3B#J}NnS>=-ql)0O3g$w)T8izTM{m%l!J
z{+#CQI`4t%Kkf?i0tj_~y%N3vhNiZ*Z&WGJbIrFU#I&}w{P>g4DP?AK{3fVwnn+{y
z<-3@Vu(TKYfGQ(O*1o)wcARoAY7F2fhzc8Qh}qhbxObepV<6ya6;>*Qd-u!W(}UMr
z=Y~gp8XqR+bXZD=i<g}ckJcFEI>vU)AZIOB=hi&XvWbZaK11bhX#zlX`<v_TQaod4
z2!xwOtonL-1<3cV-qZ`FVl%|~cUV0_@wSWgBKQ3L>$0UKRxGYVbgDVow&yh50EDSF
z48CwmVgGT4@F7ndysl92isd<j)2APoUHI-Y((ko2lTy)p8#XO=_A2#j*RH`(2*=vO
z#>TR;gkSq!&Yhkcr;NGWkxlhZ!_U^n2AWiku{w;b$l&QCjELcA2CgrbB7b|%A^U+V
zvlr3p069=gPGb&-XYLG}uMA>5EG$g=XiOUuh0zp|)XHaozRRix2F?!j<!@I~Dh7%v
zDk=ixd4pbxMvEyHFRi<~`^53%C>iC-_ifP#y1LBl;5*R7ivn)OSso(It{7B1{YOun
zI5DWvgQ)|4yY>Ul7PV^dyIsd<gyw^I0AD#jp~<)62~naIOKPuP)}sdcR)8!6t9N#L
zZ+jd33N<e848v$O{?^_dYFoE%wNp5+47gI;(lXD>c9X5Hp&{T;x(bielngx5bVhFM
z0E`!{tVGpGG7leL7!{tw<7!!kX*eH0X^HCl$cP+r5%>>A2kJ(NfBlMMbbK5R@!Tr~
z7&S0mR1WzmEXMA|9u<<YV+Ik~_j{_hzJ9xkiVr@j#tr%x8Qh0m0qww%+JIyC^DBo5
zb3A^N|3B^Av;gl6(=ut8-T=EOU~p9!wbRtI=SfKeAe68%N|M&UXsDv`Lj9tvRTKJH
zT|EryGSjc5vU(5oznXJGAJRN<EZ)D4j*cP12W<F#8gg<!%(QJXtoZCVH|h$(ywAeM
z2IO@U2M5{{bF#LDr{})LlQ4UMpaRUkbMM|axQ6=s{me(2gGkT^;KV}s<TJo^pr(5W
z3Y-kW&+HIRlz84(;awW49MKm*cG%__wsI=Q@t!sxFf??>6MGh8<_`@G&5^#bE>J6I
z+Bl0aI{OnqPdwi*XM!aQaKclbGvAvwgM0<9!bEp3CdS&@x~Q-aUa^XI@~MKDcS$s$
z&M|Z*^y!0MT6`<Pz_(m1EKAs=jluarMxgzbr2{WMnoW(miAh{m)~<b<d4H01LDC%_
zG5qWQ0`wn)xIkh;qHSQ9a{I8bji3J!b+J+qvE-XN6o4IrSt!64K=|9WeEbN-A~b&)
z9%o~W8-pvt|0>zZfcKo#R)C=J4`IX(BU*;vy?YYX&uOZbR%Hl26!3d9b8`U<R(|Lk
zxAB2?+;#jQwq=Xc8$)zzC|VERzxvyD%AbnJ$mBpo&&s!X?#E3?ghK6<6T?P$TTDSg
zF_p%_&Q8t16OhEo!NITK*oT)5OU(2hR!+_v3Vk$Tit+;8P$|b_bBO<)mKcZ_duvsA
zMk0vtAO5jgyM~k<hkS)JN4^g8004}D7k~&Jk<bJGC3xSmuS@)INlKr`(Yp8V|9e{t
z9A-86KCE?qHqh5!`sLITM2tU)kS>?a;2hw@i@9{+0zm;Kuz~~Z8wo364gl|+pOkbK
zv;y05_<yhLwfp30W8)4KE6H=@eCe>S6*xT!3VMeR6in<jM+s$PXSaI0fO&&VM4qh*
zd=CDd1BlpJHZ2n8&JBW>>Ei>zCi^m;w5&jihl2<~2*{3t&?W&9fX%+Nv~VypXXfWS
zWbS({&e#L>yY#^hE-tQD&?3+n?PNto59sK;ylVCz%zCYd2JrQzsXPui;Zs>zACik9
z)^FrPIUIF8!BMPiY=Hp*?R<P#!%_gZ60^g!eXF#zG;Cv}e*Bl|k=0_v@jx;u0TH*e
z&7%yI{a&!K$Zb3#$!-g*okSkkwX+k@{>6T4o__`(pXZRD8y_m3P=W{@GTwZMZ{;|N
zM6?(mABVNt3RP;$7RX{i(b$J34<t*%wlwHqE<rzpI@m01&5=^{FLZz3*?A2jdTu44
z0cwCc-^yO>>DhitQ1fi5#|)~U3T4>{ZuHIC1Be;JbR)kLpCDY`uR0{FC-?8eSTV-j
zy_*md5gxvU(K`@hIZ(lqq@+|^WyWjZtKNvBO<P-8RmGspqh6vi6+&sL>R+z^YwH4v
zBA;9RbA6TvW!4oO+QY+aV#g-B3UdJbcz?tyET#algH26;{=6@!EWE)5!#4n3jHBPb
zYPQ+Xz#zsuyi5qM&r1M`r6t^1plcXnVy$b+(Ta#fvu@|%_Z+SJ+2GMNJbwItUcq1!
zat|ZYZ<pCm$%;Xea0JePwlmX&5P}MZsj$ufmbaSAoj^udKP9nh>-KINYQBb#ALHd6
z9>a9-AbaB>CMJ&b84K(>K#^F=UY_WXQB*9{&1)p-P^7or_wl%u)ezWaecNejxN#8u
zx#c56mqYj(pa#H;ROYt<rcAN}i{)4*gXsti8Ke*{4h~og(@-s8qrl9Ga@L(|dk^v}
zSXiE~0fnUJ!AZ1N$ih&EOfOY!ZTk%uOjUUi;1dIlbNHG28#iv;vuBTQ$yi+e6Vp69
z@{I=%95BhXIe+n@$3RsuC&l`YAklVXZLi|*+&S@M3eCjj<)2_;uCyk~tUUz{nUsRU
z{T;<cMgCqTx5|&oy|NGe4}EjYLpy?9FSjx`Hy88$F}{^#tZ;*c>RZiR8Ee_@Q3=5%
zO9u(zOK#`a_pgj?P2HjulR0W}bpxUWZYea^qxen`6=mfZf-Du4l)iR%^FNN$NySvV
zA5P<#+0}Ka7F2f8V-+7PIdvlsu@?nS9|i^m1qG}yb8p(T=l+XJ&U^>9yh=~^L?zC<
z2D$?8G_M5Z3kq42+d8$2dZ39efZb8{b&FF)1EUjj+i4GvMbfPE<EuNuEaOZAptSx)
z&y=~LIyOjt_ENO}WwBhM$v7NXI~7pVVQOXYoI6xs4haC^PH~xJQpfU@oipk>HWcbJ
z>VVpwzjy)8(?1gZw&V4sJL%~Ypc%fWi2qHb+i~OsY_~!lxqSI@p-B>mL+}(xS-XdO
zPR?WI13-Y$9rr%)gEAw|?-3@|G}5}$NqzlWj~;bn>W90b#>T@)0BDl4vD|5}=0mK^
z&Fv!gC;h<2=*I32l;Zt26bMlGf)JpHl!0`iuiu2V6UtJLX>~olo!AencFs&nqDU!c
zkSn$8$i}P9JTHl5+K|)U|0>7qCzJ^+tRmJ!&wq+?V#e(A1>M0!lHlU(Oi4*;l%<Ni
z>qX@-t<j*6D;NG;{SR?{P=yO3KR5S$J-qyUgqsY!)(#H6XBCZ3pOz98Wfv6uC=r5h
zePxmxa<vR$kCA#h{N2%i0CcMamct;2wTvX><^4k6FqI@ELMgy_Yng~~A5FlFD;R+l
z#w~LrW8Hp&xE3(UwDf!YSinV|kMQKl6D(H1s%E;IFg@G<+tbt2V*_pjVQ!DQ!wxQm
z1Usf~j^TZQq2ZCUGjepygrXcD-wRhNe5xe1$x=tnTZ({nAryBx&qY$UNSnM4a$~Ym
zzm(v34YTk0^RkF7Jl~+c-`>Oj=3kjF&=Hnu?r@VRPkVdf`93k{PfblYRM+=+w<W9r
z$U|ZyF9r$8BUJ&YO0o{t;=>da7k?WbZle0{6=t_W3xqWa+{d1A_4cS`W4;os`6ee1
zp&4G?06U_%m>6K0jv>{5u_aG{>Hwoh)k|lKKM;FgrKBAE7D&RCECkaa&?>qQ6{f#B
zy!E3)!|DHhsMSo-=+vn#A|mK#l?Jcy@2ku+d#<iPK~p_*M%bM9Zy9W#m<M3Qb0>8=
zv}JTh32RW(ZxX`rbc%7~+H-^)FE1}(9o6XK^0IS3p6kBc<+!I$M_^n+H{jtV=xYsL
z$lm2bk~v;kL=wp4m^M2%ehq$1+D1Zje=vH!29S#dy292(GEA)us#CEUj`02?>3~--
zdRdCgUw?%l$<39$m<`1h<8S6A8(t=YFbG^Z2ce9NhK2?tboR!6nCt+)y4bbgyIh})
zYj9nI5tf!Nca`-z{TWxWfI>^jnKd;ttc9Gw=*^jtnCJv)wx*^W6jDXy3^~uuO_;mU
z^G7@X@o+H-K=K%aTNw#L_zepw@>pczd=SW)!7Mr~hsuDbP0Mn^y#ZW@IGVL|buS2`
z2~&{6zY<PWBu7eNw7@Ozy1U;1Yjq{7YlAWX^Nhh_F)6ukyPB@t9)^q0q0(QD12aZ)
zId`%&vfRA0W(w~{iW|)mWy#UE@wggiH>4f7V-enco9Kwt3`e!KCAWAzz{|zvLFw7r
zqbI&N`}jzdQJ(;IffMlmm`vwA)qXo&TbxRSoB>P+ME_@HW#9h&Tsureu`_@9GXGS*
z@`X)V(9QsVT=ISb6qlGYp^Jskxcx}OzUm0)>PyfcljDT6)Km~vP{H~faXW(T@4E|=
zy%ul1_rr&T$PfsB$8=jz^p>EVKE)*E>Jc`5*_Je%50;E}h>6{#1H!*`_wJBQ1z_zX
z2vg5WxO9#lrAPo_P0zr9azWkuEH(AL?1(U_`y&i3=?u{tM^u6z*!154FxA6cL#{kB
zqHv5f)790{(q&rW!RDvNYn_-^8G}I}t$}KhOh$<-WaZ*==u^J^w~3-sWFR8+`kHVv
zM`H!{ozhV)Kv`vEoPN8u?8JA#`cODebx}Zq$w<#W<b*ROVb`DQK3-S>SXid(iu(`(
zXmpFe=2Qr#DiOq;sa<o-un@shkZ#zAnVFJ`YIb&Zl!6pJBpsbo3U);BFH>?12N_Tb
zr6nc1ECrjq`3PvzKYQPPB+0DsaiZ?yItt6(3@S+qq@-Tg<saO>a|bnl>-O#Alam(a
zU!0*m^!xUq1#Q|Lw}*fb<7Z{9AR#XhqK#XRK`-V|2oludBK!Lz7;giw0g=&4!61ok
zk3zsvba4B3^eYOC16;;NMw!E?Lzu3zhXIRhZEe525c}J9SWkdV#$2Z-E-c*ia6~ZR
zYqj)VxI5*qNl@N|o$M-Z2iiONZ;11UuY!&SgCLr`2|kFq`TpIzv#&Zb7kMsaLqb3x
zM@P4SQo*1zt%dHd{q51e=>)462LVq>P7_{UQ*)W&Pq0aXE&x5)PIF1;5m8S~&9Aul
zHzsB9gZ&JbCiK<;8lR{W1<zLP`(;Ys>|nnDZH@6FPVKt>zooa^V{M3nt8X7U;js|;
z!s`f|^x2Z1R;Vc$7d;`NHq?nZ#f^-NS6o~M4N&m$z6(z^Tb_8Tt!8<mzYiiH=1&+u
zQ7B0oC>Fgxe?Gpq{@=$=)q?v?K*3Qc-v*NucF<p9|6X5JHQ_P$5GoYQZpg71QQ`Ll
zUPlXdaBx7XnVWZ`^z3c=_d{%Y_~>b9qN1aR$7h3h3Xn-4Qc~D-{?Q^BJTMDl$J1{I
zoQPPzdSZ7x7?%E^(9kbm_5ZyJSqDZqK}ce=DU<*>)n8_C3m?|h6vr-dqwK#=9=D4S
z9UWch9f1_=lGJ*B69hhQ`U|FiO2Gp@f?8CEPZ~l5`-)wufq@6{1~3-jkvp=9MB>ex
z1X}?%US8}M)aYpZTcA>7eT62ubw<~Qb+B%6X$gF~BioUkAaLmU_uhSQ$$x!I4fIY;
zx<kJu4hp5U|8gXs;_bV4DTRRYdcafR1EY5&WN-N!ddJ}p)DRYW{_-XEOeKYc5;8JM
zA#U`roLKMPNDiACzM(hAXncT<@*SrTxg?vX(5=#&Aht$1Z7tW2*<pxu8w?om<C-E~
zIWR00+O|m*Rj)&Q%R7S!sNx)SqhBFxQ~$+$e|yq;Q2fomb#>`qhy6Pl{2SBO3lm8B
zbvlHU<*!bFdV}nJ?`QO#JC$3XEZHEY_g$SyzTyFu-`jgNKRL{bHzCp!_cb(5eyjDF
z9nryV$KRjLO7J<`1BR*n#R)MKcaWkugRCiZI?=6Ly)W6Iw!k>D8nXh`BnJaRZYAz!
zFvO)3(nlrs?7?jHyDJq_O2X`udNwgbWcgznY*Rw3{AK{28q&)@+Z49lwqXC-%j*xm
zhM5Z1fH2;Nl=T+<1Rp-a+X|+}2l5~5>Vi#mw6tu=R2C!oNT1-vi*5P1lWH8Uh`()B
zlyOy3=l;$!sKdFH_uv_q;GqM2#WZaSg$QE~I4TeT??uXXKw!E0E=pMJ*h3?fDmz9&
z!nDi<jG*w40roVXT6(zGZS0<fH)i=TgqMv#1ToWFv!kH_29D@z$|md<e~2Nie7TJ^
zqYGEOan3J8LSm!H&dQ4NCX_ST<R^)VO*t2IA=ZK05cUU7(B$Fby4hQCzK?_-^X$&l
zFA57aBXJ`6m1=fZreTpEdCbkjBjRO$dImQtgrO6_oUL#Ggt;IAF{e-isux!Q(8eCv
z2x<3S3+X1ODKs$=CbyKB*x2CD5UIFvVM1^)E!V%11)N0_Q)*h8Anrh`>VLUXxR8V$
z0X515W)_%fvOxP%!eoaH0EEdKk7)`=m?b$1D(gvBK~x7!d(<p^e0+G(h7Tkmr=t^t
zSu65cD+Lh75pxr6@oe(rhmNH0l=bmK*Yze$MQ}`$oI_c#<d5Y8`=Bv^DivXVf!>l(
zM7rwfBI6Wh-t6Vcg`|S>fVV8+!{qzmF=HBRN>C%h0vlk_{np3FGE_PrS^rMez<_fA
z(<MQ;j4kL2?$cOYTm%J8qKPAkQ0q|59x7J*S!U_;c*v<{khR)8UPww1)na(i_la~F
zR?vB{q@dEDOx;SsWHlS>`Y*NYw(^x+%QaC=5MJHu+E;Ii=8}^mqYXLmB39m3R#t3W
z9EaaWOYRgGr;(zkbJX7HB+0CzgTX(N=oo}ieISCu7Hy=){W}(Q%lxy>?@U422qR*E
z>I!6-@y{#ad${Kk>*BjS5gwxrdxTV`B0WcE-*63=GWkCaGsbIz2XyM95%t;oDS>cM
zliQazmwEa4`1Hu9?YsydiB)}q;#j-Skfqzpg(N-#z-X*PV3~0YjY7~fik}1I<A@Th
zYLH6H5Wy$FvF(WfCku-o%iy4piw$$vQLGkclI3TE`>^jt;LSQYVh85A%bu$NMf{_$
zU$7(zGjgI!7;RJ#R}Fhl7@gu|jG(wzTd#&AUEJDpS53ap606Peb!t25-e#nxZrn$8
zgPEd!n~)Hhk;^6jCeIhxi77<-Pu<42?X|k12s^7QZYk+7XwJ|~eEoVZO)+S9K;3yi
z%(0q~>FD_a&X0&x%2qOxh!|2?o2o_Wr)`}ORQ%rGPrg<`0ZkU!u&96wu5QDQFuNr}
zT|9%O*dBXqxXGeSQONSvlG>1;7k^sDlIPXA@A{`1D{i9F&7{!#W1Z}=v?E6E_#vTG
z1rJ^8WzdtUwj+jtyM|83v_64dfXuzG%+E*w>t?g5ddH8m+DRTC85Rl|hS2EK>%zuO
z=r7>*TvAZL{T>;&SdR*DK3G!w6#KXK1`}0|VRx+1ZJbjo84DS{hoEefCwVg_?S_n^
z8XfTtL4hVRG}s3ZfnQF0yV0*%V7BAMFSkleAz{F*Bvz1j8_jYgV{NIbhxx2~8Qu5(
z;XddKYP)JAtscl4*EKfYI6=qAXziCy5YZmTP#JZ~AaVNFgLVWx@&k%v;VgyYh9fA;
z4G~V4ofZ5mm-{0sceh`-crl2oy$&;@2zLJzS6<=5%K0@PX&W$Ms`P}7jk_5K2*A7y
z+}6GO6&c|HbbO%`w}*Ch{e~P8v}F)#9?LxFDAh;9S-|B5mfAX<TnDUCHHDfyc%Hm4
ziO%L`S0GT}^>HZ07-ZdaY}p_)a*_vNo6rF?B1*8WwY9+m;?CQB*5F$EDPcr<m&(s~
z7LqWo3$T<{A^Q6pS@o<Er-(xS@B{xsS=&Ze;yx+xJS2pX^1?36^HFzU&l+c^^>A|7
z2KW)~2)mh-OhGgf7<Xr$RZv>gFEOTi0Bj6*K}fp4{RAz!d>_7wyJUjWv3B;r=?5gt
zu6sw>Z$(5fZoC3|4y1xr32~w7m4-gVOG}QQx49jYNq3gT@I2mcWJ9+1)yIRovRh^U
z&Y2K(Mh8xtQ`{@Ni2A{_zYCX6#ddH0%m2X)xa(p?04<_EH#5^$I<33DPboD!o7H=x
z6xQ*gGPeBJm*iLzV+xQVa0rz;sWyGw^%!LPS&N_lslR-px3B{iK`b!UP*+E>1#%e~
z9gThTNN0Fmg(A5w(lJh=)xeexf5I%NqTY&f<OuA&65F=L-oFoHwFa7~wzeI}5!aOU
z4SdXZO*T_~<v0<zKc?pAajV(Lo3_#*Vmqb~puiH``XYs~vmVtdw>TXo1nCG(f79K^
zkI_0CWesZ{actVO=*HIt;wUju*p`NY-Fa3W1gmdk`hxg6(!h6xAH_rrYyP0YhYue>
zKrFfOHAC;b_7>~pj>Ov4=~_X!bwMM#zQKd_h7xhf1(sbXL_f=G>FH$#zRgy;IwQJR
z{Suu%(*5)fW!!N%71I+y%XaiCJhs<QOV>GN7Fm~+1J}mf=83ob`v06P0VzM+AOZ&9
z>a^|oHrCd-)MwGnn{?&qLqQl^&ER?&-|NS^S*87A;Ng3SlRJWTE6#3i-B|F%gnAFi
z66=`J8f>54SCk9R*T<}#o7|U>RM4YbG9e`iPt)X^jp8~i>~s=ANxlr%C*Sz<Uk~k$
zI&dWH4fXHd9ub>~9U)EayfIZ25+V|dZ+6#t4o*mB`BMD8-njo^tdiSg*Y)_TH@79L
zkE{{F1Q%bMw>fsK@#~6%IBXe%YsBh1@!rcXu=YG<Ieh_iDLlC1HwQif$2k2yA51K?
zn8P#UeXKtGeGn-1T}fB;Z^}R5|FFZ>?*5g+gSm`2B_vL|vEn6C<Y>xgY$E<i8W!`2
zf@Y8Rw_8pw12!K}hR|VfzeC65<_-V8sc928_e*@u&=1k{h{p~K7u7d#e9%mKs6uu%
zTaQ|LRrvb^-gwL{Lrt|<{T&=2uEl+yxcCLe7Zi)=<@h7ycdf5|g4u!M%@k9@@sy*u
zGO0hI9pcFW;T18-c3Fi2r(7HHGU*#{-})3f59={j3+~khkC3u@fn=J=5+?h<{^)=|
z+YHO!u-?j5R1%<Td{_M2wbkFC!o&$=M=3#ZY^0@qaq<OgNYec~4<7Ie2>cu#4ongs
z!G*c)@Y;gJ6Axki03HGT4bh5%_3;m<u((%PJwZTn)${j{eAnc@@&R3xH_)Q}<*Xkd
z=Hj|NT<Nh08$nP|5PsJa?W5CK_7K*vbGy%|p%5?%q$y&Hfr--v=ttJ-T@)HuC2<Q6
zs|C_8w!gXgB;+$l3+!5>o8fT>bU~u{5keK4$#8uSY&_ZoIswT>E7Z+2gCD>uQ<g_=
znjr7K-0ug9zNI;b#ot7ukO%;?m2a`sv;C3ke&#JtbsoIGC$0M}1FBYbR@Tz$?@1@g
zjaeRaPNzsCqG_Ha{|j^jfT+g%yEY83ED@5#BiQ-luu8J`csP%2l13hME50DsqtOzl
z+8rG9-zB7^xM4|~T}wX`?!S#367Pncfz5F-324z5jV|Dou$Tj<zLr)!Md|bEz=2|{
z)-5eAe)8{HAuaR;1bj-xfaJY2d=HNx^NRZ+uCc-5Hi%ls1FLL%O>qkpnqA~rGWOnY
z1O|dShiHVw^F@u;E4XSlga=YyZRpJNtw6s-gM+Q6nU>{frm>rI;V1#e?BAQh)EKX#
z`{;gNo?LD^CfFA4=c(7PT|2Jf=;rnZmtrYfJwn^D?JC%{qb1!~#(ou-Ie_cbNW(37
z%3up}4MsGf*2{^A_?w6@9>zcE_w3lC_w=|PD`249Uqzo>ITpbMQ`((lTjN!HjtSU(
zP7WU12m2WKtt7VFIL$!yasj$Aw?h3P(y*=K-9R6&=&}`i7f`36a(A7A$PI-e-ryiZ
zj6wB{B}^YlauJo&!kw@YK51X_S^FJ7){G4-gh6c5^)cQT9w^9lN5hw%id7=f+HbGT
zepCUo_-8Ou1zrPZ!xr4*i{a~2DiZ8q+S%Tl+Rvn#iw9KM$=r-xePN`Ey~Ct<OMXKx
z59~7|V}wT;xXC)l!eDs+{sYehC*&u-H+*{*?@d}vgrl!8O+;pQuN3Jr9SpR;1Kn-f
zn(dqvj!L1>Ot^a@+%KLz>%Ye2O^$18bS@v2t%^FDGC|Y?NV@Rqw46=zbM)kfL}A+F
z3-5lnE}<h83M5P2H8{B4>JTo#sC#WO<)^1F(>tHwq1QshJw7c>O^GratzSc6g3xAl
zL&<II!0~e@HCdB4)Q$_|fA}yj^oYIAYWFV@`bSA&*uNiM^TR&)P5*KV>jdTSB)b|f
zS8bsAcg795uY7&U^7KFbFWTM&n#%rtAJw2;O1mVLvPtGLWKM>ZdB|K+8Oj``P-d}{
zBqGTinTJSG(m<5TOd&*>Lz2v7IM<`z_xnBP{MS0`x6c1>t#@GW{p@G>eD2}8ulu?S
z>B<HeFN0_&!jwbe6_OAaw`^CwH;a73VYM2Gu~IM68__NZi$I{zn!CYbl-3&gjpC+E
z`=-)TUzQ*BN$;Rq(){HS>QZoEl{Gb~uL+NHw(i~_vF7x~tdz0r(VT9=_}gFP((s8s
znl)}AM&L1tyT2tmSuU`OXzZnRylkBSCVG$jBeP``jumPg43pMf>*sFz25*dei*C;d
zK5uL`{RbL|Uvp6t9O8-locrw#KrcW%yTXJSAo#fF+_Af)h&{(j=RY07Zgx0(EIJZ9
z=jF?nx*6!+oIzKeHk)E~t=yu<x0KAOUG`sIz>h|;ut>WIO@%!N<clqwxJxlgdQGNV
zNR9r0M>q?MxZPIwqe9bKSj#|<w)IuyC6ESgl)Zm{jU`Un^@5$<p8aae+(8Lo;T>hu
z^39|qi!iNNvErf=N-MBn?{-cfNK8uNjs8+_j@8eGL-^>?qsBh9P*ruek2<VQ0A0bt
ze93PmfdMy$k4dXEWgW}o+K)^(E%PwVEB{;=7r=KVF`;O^DW&g1@u$=?HjKRlk>kg8
zD9Ic>dFNXR5HZL7gGAY4aSQ8VuzO(KU$-rnC!Lv_{skZ2SK|Qc>)8#;{uwYyL;8Ie
zcvyHRm)k6(B#Q->otlh4;`H_dP>LPC<SS~CUBXYAQ*?QMnp}aBNAIfhj3T$%o9p2v
zK7aZY5v;GP`@!4Iq3rtR&3yW%$^-gVxHXb1X*~A8PA7rM(~{?_2Zx%Tl;ntJ4oguU
z9&3;ctArdM2N}%+xxwB%%-1u&J>r}Eb&G!K86e|T)({}plAFIWx_AJY@;k4W`T`NZ
zfBy#N8^3>rq6eK{7Z!3APh7lqTgkexdLxKWf^h}7W36#XpKHkBb37Da8XO`ZPPU^2
z)V;xjP5+49Q)T-r((m`bPWDadvN(JB-l5Z$mKUMF_ax}Hpvm1+&#*9E-QBaiQty0|
zmvWe$optK}+PCOlkCok}y<<6z>~$1=!eZbe-mYXOltIEE17<h5?IEm=o}e4ZKX7eO
z7|M;v;EksjQ1OG*z(QZlxk%N9uOC{|4<*dl5I~!T7g8FYoO=c^(UDyBx{@p7X{Mg>
zw1m85Y(jp33)b5bee2dKYwP#My{R6z*c_dM3zKXhpJ}J8NwNkX(&@>e6?Xn@xO0FM
zVDu5S-Z`S^zpiY-NBcmUKHmV146Yt3`o8<Q9;KR1rf#9@;xSEIq42Ke-8=t%KFO&H
z-=c>2RbeX6U%F%!ub&t-K!bvN0<dt=k>5XyMP?X{-oAVHu>obF2wuHa1&qr_9)i;2
z?|2Q=*ilvAp`_ht#%A~4Cn<U5>eWX|y8yeD((hKW9qRUnL+#1A^YVal+H)|E?iqxN
z%8fkaO`r?k>CV=_8bmGO*G|`o<;7<_xV@0P)wtLz?zbSHZYS<;3$;>4*D0aJ0NmiR
z`uc}p5;uHLtdhG950Tl478Vp$h8{EOHp6yDc4G1Bm>yQ<gA(M+vMKw86$)(=$!TdH
zc{xQq`s;QMhn~{%ikxb%&jbz26Dyuk_Wu<E{V)E|u{IYKX6&J)9e9<|`zo2$I72rD
ziEg7RQjQ$Gv>NUg1(_OA!fiL5h|ibZ!Ojj;V4s{nx(Bxu60pAUjni6$2f202_Yo}(
z4dSc$*~iEoVHgCxk>vC!Po7zSZtN?k@kl26O+v%MlFst^#b$n$>l#JbYO+!}QyQ?{
zJl(#VNK}?Ht;E;RS%T4V>ftn2TU(W|MTHQn+kVtZAdP(R9veSjcm>q0C)IT!)3pm-
zN|PYZ&mxFeSGA=RQC`bUaorQ!wrsJov{ceVbc)RADQg8n3KLX#fZrW^e$gI8;%J|y
zMRbPdiG&#1Y%;esitW;qlPku}Z;C;@S`}>}d<;wzgW}9}IH5UikP4D-VU6VNL+4jQ
zEby}h`-KQC|8?6QfgFfi_%Rn^kbwNQN8|VEy1KfWnzKl_K>j+l<K6sX2R5f_FeoZ$
z-f2dfrV@=^H^jZhYhlK!H6$kHGj_Fg3p#}hPkaGk<)RiUxjf-&sK%@5l|RW3BTUR3
z3J=g_$2GPNqizf_t{*);j9a&+3s^4QHMS-7*{GMCsmLrDqh-A244rGR&u$f>ep>4m
zqJ%==`#{ddcJ48-Jz%RRcNg3swbV)$RWCZRwFGVZ^8AwEIlfAi9;`pW%(m?KAljQ7
z&r2L!HWp&74==g{rSWq?(bg(0H7!lg4t|W#qQy0Z_MONzUG|+&oBATR;6mY;!lrjY
z-vGeuG%>`0bu^!-uv=DE3!VO~JlrB8VOM$Ar*xw2HGS9t76n`kUIytBANFd5TwT`7
zX=tVmQE^97^8UQ2Mh*LK?N?<QMo%4fM%7eS65S2v57BD!u>nwPF2%{gp(8u!dkK~^
zw113_A1B(|s^3*r-3Z^)EkFG3Yt)WDIZv?S*mkWQ7Rr|2!?ZmM+{epFYiMXXHnCOn
zGSXddM`U6K&hr-^o71OX2hsm*D||hc0CveTid|iT58_b7Yt0$?qvUZmT2BDNZ5GE=
z-sKlO)nufhIbinCs+Ph~hkS;&26aqx;k-wWc*anm?sXk}En^4~RK&Uu!*I&Vw!aUM
zkw9I&3SLFVvjUAeP<4)={egzY?;v=vU?wZD@6<Yp;z#wR^XGXXed*@|849kSh|0Wg
zlT=;7#=&h%o<Q9A_S2`M0ukN`B%45FKw!PWQgH6lGkppy@hb$%?H3Mh0t}glZVhO5
zg76M_M8Nok3Im!&zMkh<MnmJli~{;j67a-dT?)b*d*O6Z1Q+=kKjT2gXWAiQdMtJt
z#vMg%xjO0_0ySo5=kZU}?^~ll$Izk*ldz@twZ_EcB-i#XN#ePQq7*s-9oWXjcc}0T
z!4qH|S0F;tiN-M681GI|=;%tw(OM6!tFlG>78e~o3pl%A=7G1FiRK|Qmh2$+=e%LB
z-7^NT)KuRgqe_zLv=6j<ibM${UEMd=cdf?l`akz``BoXi_@)OY-kqg>WD|2hBl(0j
zthO7zlP(Zg_M7w8un!>hK^ol=MNI6?#LvUGIMo%{>v1h|9-&Bp>e%5sN_AGiE*g5^
zj<btBEiBAFQOH2Sp3)waAv)wOMWJ#cBQuY%-UI;*xWCpPK-{{1>N%dd7wAAp?)?Qv
zLF_eJ=?`Ocpx!m461$U=VT~prhv$P0VAvrUHNkor*MBu@^3u&&Bp$Jk*ZT#~b^5U4
zKvh`eA?l$wCpnNycC?_p{C*VnV?Nzm*sDVT6u$aspus0zNqaBOOZ@nT_;u|U_PC(B
zOzR}4Cv1`{Hp^<x*OroKXh!x?I1pnU(gXF&g=GP-7E<A;$~5o^;IJC_$L_pDN(*bX
znT}Q@#&A!L(PN@BH!1^>J+wljpjfzHE&#4k>G<)OmqOpWx<+4<<#I(76cy{)K_cut
z+xsovJGG@*Q$OoSQ2WZ)$hU2-0<Xd?cX4wQc7*juv18t<vrYgCV`CR>5QGh_LvE3_
z<AMp0_w*f+p4V>NNTx0PzR#iK0GP>0!J8+*=RgmLj&mRgyg5f7Q1bXv?QCw>us`+F
z+W5@RkG5cB;L1v|Z8S93OQ}emQZhJ}i!&zDKyLZUKzYWYyTTT90|82yY{JuQwq*{9
z)m9x4GY_L`gu8(|%N-bcGjf~DB5U&MYxTzv9G4U16e7U3PX}@^BxSHD2Z1ibioL?Z
z%%awH2o_PvfhJ)}R#r@Obj5EUShN%qonkM~R)WWAA-v53`}Ao?LKPzQ0(G?YA!xAr
zW;R&}Oa_`xa5-tN*ogY-yY}u4RiiQ`vv4hbxXJEap9B3UpXZ9gk534sd9>C-NdsOl
zDtPVuGepgKekK{ZoJIt05?RWvE*$e4-cvhj8BIc+Ok_+91gn)rpca0s9HU_pbOt{Z
zMhvMeyWIlXnx!a6p=@lIFF;KB^~vr0rL8{SN?`iyc|N(VYg3PtCZqUi%pThR{ksYL
ze4G`+H_@9}S>I9Wdny-VSmd#9|7V1JyxcrI>6dRot|85|+~HEdFE-VXGghN}EC?me
z?MFui))M_9`uz}N=#%!GZGa8uc>i2t0Kmo&lj<7{-t%kY4~y{RAw*BF!Y=Bka6m1_
zdetZ+JybxqZr#emNH-Yk>I(_B<oA9rX2chT*)|C=jr7pgQgO@g)g$5DlA9$>7S;vv
zl18h|qT=?wIoK%&Z?>Nw1b%F#wH`lC*W8BS!={8#+7u#p%eStS5Rs9I+o8N$aJrZf
z?@G2nTTSJSh4GJCqD0Pe){D;r$i}Gb&#$-jr;)~s1{lcE&|E$^&bB7l(lkdDm8siT
z;d8a-Dggss>OItyq>7k5v88CHZ#y%16W+@{_qr0yEROZpe*$pMLm9W^xuy%{zE0Ar
zpkZ`kNo==|7){fNDxLS?(EMUIzMCh0>Z_8#kbusd;&Tp3?(Pyf%#ezxz496>xVM5(
zi!yH{sMX(2+uLi7AyK>$<#W)EImP8j`$Ao;D7Cn_*tYZc%uIbU(3b?+6*LDr-9`Wo
zR-I&rr%tN}WBxi_52|OPpHVxoj(j#xp6d8K_dNvN%2jUfWkD0cH;PCk_PB=Z{mJS*
zS8{v|B?q4PNuS><RI6XKM0)3k4$(EM-n#$trRTEeZ(M{xs3|bTP>o&`K_j+ls;Zg^
z2*yN3h34@{TFAU#_<3_iA}h1$7>Hl`4xNDKex}#r89$}WVSb2BCV=EIGH>4?r*k)V
z!>hKpP~>okwnwVR+FMdEb5xA??Y$~KgUoAXr=kIQuOF0te**wZY=LE8Ulg?5J&H)Q
zmog9|NRu<aB$b{Rn*-uwEeU5tr+rV`4mH8TvsIvL?8+f+7q?SMg3Zz?;6_a%X<H_&
zr9&qSIgbKaxL_CD`K@iJda^qnS1I{^0#QkI6TJA%3APCEzXAUZd)K)T*w31v9K(mc
z1*SwI<c->rUzfZ9;}7BI;0?><kjN}jZ(=a5B6JiT5;iVCo_sk^K$j|q6y<#h#znrp
zN$}(|1(CXF9h+q|9cx`emcN>wc8TVQs4V<ML4Los$U`}#&|Emm$7k6tU#&5=jTA~v
z(6kF>4#;`b^tI=!U4t$fvWMm_?%AzCW`HD6VhYc6ms>83LYr&kc1=~sDxU5V_z1^u
zElAWI_l8ZdDaKVjAEI0ReXFys={2{t2ur?~0KGi9_{g>~6Wm4ZI^`ME%wXiZYqa8W
z4fmwxS|i37%A<kzn1at)wSLF@8l^L)6vxy2N+Lj3XMsR&9%sw8hofxUe)v5Xf?t@=
zp&y~rgxvMs+NmHAC|Wj+3ObIhJNi9-NdCI0O#~0!X~*{q`fT~?Lq8tfAn`iZCn{Dg
z(jzziIXI{mRe6X1bS>TcN;*H^LXBDR#o#S`#(-vXNC7CFRqXstIK*<Gy}X<E{5Dvx
z%eT1*9TPvc7mS2~6XQol;CEUkoo?a>?CLk8%NA(nFz4xO+UV}?KDs<l1vtvnPIF*y
zksfFzgNyFlY(IE6DIUY7L_`!Fp<G)fGHrexz1J7xGHdIJr3zZZZPhzF^HfrFnNvf8
zeqZ~_?e2h0===Q5{1!^Lyzb_K0LujHU{~HVxF=7u5aMsO_`vbo!&*{G3ah6{n*8Cc
zvXYYUcErLKUd-nxAF@n4Kl&18Ze$IZ1-5jgQN}VVV!u7p^v%gPA$j=3y54uW>R>Gb
zEM)a(8zan6opeEwt>GJ*_d=Kv!o<hLRcp+`Wr3?%vS1bDyPr;x_fgX0P%W(N{Wc#s
zzEEk<qmZW9ES9@*(>~zFsT{t53-2S>o~mOXTLE%Q>1PRsy}emq2FnkI@o@3GL~JF0
zU>fkpDvpb?$Aic`g9K(*wKm!}uN&}ABPK{`u_o!81Ux}#8!f`emkG{~zMFpT3(|^3
zG?1f202hw!M59RmU7it%BaO=j{Od9a5h)G*J@7@KOfbTRU;alw2|8`hoX=pLz||3`
zs>C7~PrxOiZ$p9Ch0|=AgU=#{qCgqt_puQme9$$BCd$wN>kefgeC6RKLNy31Zrr#L
z8>Xv|&|Fgu;{pzH290?(d||GT_&IdwH5eZuE76X0X{foR2aR9i$j&W?QARv)fW~+X
z`zFEsSNo-}=VhS8qX?1|LJ#4GXst2^_D_!?69MeXmUrQvl@R;T&|XFeg`E0cCmPD2
zv%tW9ok%OnCpu7NL*?^*Fg6S#g2+{n_5(jIvI>DP6bb8S6#ak}3{}xHi$=<NO8xUG
z*kO<YTI!K`K|8nZ^5e%7K=UK3Ko7ZA)<rZ>AA7_hBoq=Cb?PmE)r`*}$X_weaIttC
zoX~-HV~E<*Iy~XF^rbKD*|jT4_i=QN#R`}5j}MkNrU=OV&ODqb>Nts3|9KwC4Ci+_
z#q*QLw-wW6hVkj=Tto!(Duiz_exSQ{oy!af1n~T0F_XUEPvkmu_GEmd<(?Rnvxd2&
z6!jW#8)O{&jliI2^Fe*?aQgd1v;MP(K?dA~5o=E%A|BB>O)H^1Big8G&5#|`G_Ywy
zvQ)WX`UiNYgJ%@d4=2i`&HrpUmtY&E9W8abuXZid>S5KfCoJi2R#rx-MOGAgpvoI&
z?Ik6ZE#<WF#J3kc@s?H5yCgHtJn+(ImG-MB)SWN_l#~25U8<Zrp$oY^<zsTxAV<Lk
zADC@sv9rD%eO>-RP!Zn6vbIq;Ae|MG?TS*qnd<cqi%?iF{PE1Umk71JejQpVAz@b0
z0PZc_VcWLE39P;huS|6Ku162ykup9r-`?2h=+=8@v`U=0JQr7RmZ;{h=I1f2$={i2
zsx-;PHQe{YyvH03tk)0z+<`R8AXX+aHWnieoG*$>OV0tZN?V3mK~hZ`eONh2j;&3B
zKBMbpsEXexe=BH5&*-55g5c!$n_}(m`719i1e%u%AW!*yeeq6pRaGopE8}KRYtXmw
zh9SecXQT0lwbzVZxIDBvkemLxb}mEb&}XN?rss<XbQI(klF#n`(e4RRpDdp3sm~6x
zc)o_m*R!-vlv3?H*^8VdK0NGs`Zwl<hG>~?|8@2sGGFpYh0108cyXb`z<YLp+ttr-
zA$g;}qLC2>@%R^o=MDC)i>(Y@KOUr^NWD11rF!VN_gJ-rne%By&9u=ic2eP^Hhk5u
z<3+7Wc-X3Ej<m-Xg>Q<B_^|NebJi4h{oRiXA0H%qEj1C6+MDA{n*E`_d`?y7llC#4
z0G+~&v%z@aQEJB%Vk@trYH(UX`MR7~A}foYRM5u8?Qi27I;}F-=(DCDNzVxzBnPP0
zwjEB0ZC_5%&r%8>r!QJA=o;(?eM(kX{BVN{yLSmw9$#bTeh8z)g;>t1Ql~1$=?R%~
z2cBIc-|gB}pX;(jS@(N+Md{lq#F@`zK32#s?iOq|)zJ8OBp^vCAoEhM)?nV3hPUUD
z!>+npk{J`W{rCrRL`BbX<u<`n@R|a$%MxqUr}e(SxinlK!$3d6?QDE7u+BIC>7bO&
zq5OS?x`GZKV#dPGB54%~Hopj}EDHTk<#*_<7PD<gxEoZS+-eMbo;gA4G;q{t84eC}
zpw5Hhs0U60;PUSKQc5kWrZ#LT=ilKp$T<>LdQD2>q5kplkGqY9)0_HO$?N6b$G6V@
z_Lf#j?a*X)zxpddIQib7l*OvI`1*j9K@JN)w?!9+F4nn>Ag;OYlJ#RVVXJ;xNT-`5
z?ehL`bKeaTJ<AJy39BB10akL&vv)DcESnc^lH9X<1)I<Ccgsin%h@!^yrt$_Hdb$w
zx;^ubnt$}CR-(+}J1TVwk9_f?9K}x5N6w~;^#&=UO<zKpb8gt<^((`K*h9<<-6apL
zruAYD&%dMEdv$0rdxiXBYou#*)}6n%(DLf70Uoshj`Z=|Xhf#m94q-&sO#Fv&u?6`
znOY*`cn3Ci%cm&2`j~vv9mdIdz7-4`8OFtQTwL&t+p^eQ5~(w%s+S`~5v9APjiu%%
zGw)zuSoCTm*?#gSY17SLX-^{OZ<3Bz^=YX|D+OYto<d^QF0s0AoDZK|8O#e!Ise$A
z{)M^8mw=bMbseRr0cnD`hmhjq2=WX#2cS7NE=YyI{|DEsp%=zJh$0g$Y!lh_2Hb#p
zAp-`~NWobyrAGJ34~FU=&nj)DrCcF(mYnKEo#4g^&6l&czi|fWAB_2^K}Hf=kMIWF
z=X~KG4iMqV972!}thATLFQGQ4!PSzBA|D(A<C+i22o!{7fIzaJ(+v&?C`TZ_wt0v2
zYH}&+`_hIDl>5v_)XlATJ{orsjFMuGI#%a<*q~kCf;+5Iw28{tCet?psiK^-L|GoQ
zN{{!z6E*a^ygUm|U3Hab7lk-Wd6x~8IW@|C+Y?P~vaLZed$L+$45Lb^CM<Yi>GfP~
zX=}6ZJQv8DXOvz>*m)*y7aML($~yl1k#rpG;q*S^rWKSu0Tj>&!&P3VpL{O<=L}6t
zDEg6G`Pwx#Dvn(UD+=;4NsyB^tdKTDimdt6BBB>F0+1D#)lm^lJ>z==s(J@;GJE*Q
zgB6NN!DF|IyoobbwM`bko2TN<LA-c{6zGfiW`GC6yC<HinhjW6{9H-g4^thNkbU=0
zMO}%=e`ADv-k-R#k+tq2Zkw}bU(aMa&uR|({W*T3OUQUK9~Z;6fp!1ffGZ?*9}>4_
zG)GgYic`u9g{X$@1c&19g$zs*Bzyh}Lxb33ff3_Do3H)3ZNxMF&{^U(@jbBDjp5Mg
zvkcpLEM{kuw=x7}da9`6^XWBIklkA*6w#)2`R~2<AS3zsAX-=PGD%Qkp9FuBkT_l{
z(HcoO@%2@+MAe((BcWV{;U7UcSb8{DNI>l<MLvlaoE00TiXtkm7~D*ck+i*i`n1+v
z6TX`D1-XxjGr<9#<lOE5y{^&7=&9-n?`Du_o_@v(2F2atD)OQJTQ>^|E@8S7Q32^e
z!&!uIxy6uf=DxBPDk8}k_BRMH{@kKwD%YLV3%mV%@-|8gM1SA^RH}rU;N%jrgTPnU
zg=ND1ATRSkqad(mOtslN1Dp@@O`sIITLgfL9s(frJlke-mQusdG;TJq-jf+NVmx{?
z*lq)9>2Z}%*5xpF?Z3~U#ctuYQrxa}HDzVW2P%<{ioEb%^h(h!cXI%Psiz3l4CXjF
zZ!gs8tut<Q$q%xoM|4Ce6@)j$FFhj_3gdT`H@XjPuLlMBk$h5JbJ5j;$uE^CykzQU
z+=9f1TgV;T`a~4|c3Y#XEYn<dX(g#^NvHGE&lFwk?U8Eh(f%j~85+vU=)MvFe-OkI
zvug&4W&gf?QQze=ML+;RQ?aA7Gp9_+((dz>R^dOXp1Di-Q70#VfVlnP%%qbRW-@nq
zV|U3ooMf~fb*@Lf%l6C}y@>a2=*9Vj8WHq<8W?yt5v#6*p+|mj%tGztH%sFe$*0*M
zl;lTRCY?Sg&kzDlVt^ON$RA!IhATVu3-@+uCG5FMiQRkp>ir9A7t(fIO|BT90Oj#)
zuT56winT|x=6;)Z$kUKvZxO1;NuKWlwzp!{I<(i#10s4#sj8@WcD{R+h{)2ahDSk6
z-E=G6qe5Ppz3I-<t;BLQPY01m_%)~=yW_PC4Gj@gM+GLn0if9|8JYs_ig{I6cL}dv
z8os+!Yt5nbpx4xqguWA7oHO5T(d(akcENKWzJSS_Jh+{i_kLW)P5uwbNkY+xS|S7!
z1v})Uib|dzT@b(fUXx*}qCo`TYicu^5Ek>?(4@pWE3L>U{G|*IaL1bbrPzCMa7!of
zjFcv->dk3}fT6WmJ1JGvh0QDOpOG5<GH9dsxh3g7_n-Ic^3MzZY*}bN>9@4^E8Xi}
zHH@}@&i-25t#~548=!6ybHmu6-rY}G5t^%WjIx;je2#Ct=ib@LM(G{bMfXMWsegBW
zg1en7>Z;A|K2>h?g+<VaAsae}glWpA+^#f_aP#{*yyaMOhF0$OpR(Hq$03jhFA-Ib
z%pM#SM501T$Wb&kG?IPLwTLn+a+<)CAr~-512u_FEG+i-N>OK6dZu&cWagNdD;`@)
zBotR@Wj<UoI-oSDismm4ZQ6hU1LK{+PT>$yI}A$DCVI5<6h3=~-uTeR+(FkB*o9FV
z?tATlTmj_N@^vET*q1I}JC1|~yPbDb3`|Dwc+vU2V`<q5_Q?}^_$V3N^T9qrduT9H
z1c(WU5@U!06hNBT^p@H{g9gQ|d+90CE{K|MkP79lZ|2Migq`j1cGEttr943^ctP*w
z=G{k@*3C4h451XwEM{Bx3VARAejy)&n`4IJw;9^mqFmoo5sG6wegLk6s)YTyb6R)i
zN-0&<)r9>2V4NILmLp82VOLIIbaQm(WbH+-;HLs~c;Y6_)CM|;B;Y3h#sCe#FyPx`
z`c3GuRzMzT;G8f;`x?MF@pE5xpnVxmLvW}B1(bn)-W&60e%;deIs&0H;^6CKnlX<g
zAV;75Qy(|W8Y}<=SMR7*#jAN6z|O(A0pVVhi7^eh6@3Tos&A=i`olD8SURGYAY!Tm
zIs0vaqUG(km0fMS{~YqEY02{MZ2s3p@KuKFSBce^JM9uOv&9T$F*K`0)q@~{^zPdH
zH@5Hc%+@WBQsB)iJ(g^3@@eMJ$x-j+lo<E-z0}VROtLzbNj@yrX0YEZfbkdMf|vx!
zC<GM%u_5YqN=Q_dk_ZNam8nc*T--dYE%TCZ3Hg<f<sTEQ%T=K35LTt1bIy)t$<(U;
zVlxN_>&tei%G%PBD~OAe^P$N2U}NBva{y4Wdg&?>?^bkOMv<F{y{zP#Yya-zFpH;a
zyWOj1KG|hU-fm=<wDo;a{lUth@rmoA<@t}Tsv>{&2Cl}UQh!FGzfrAV8O*MJbLy6l
z>kla%VinVsQGxV$&ncRI-MoT0kq&?Nx;hP;&3&e<Ki+C8eX$fPJwrq?Jd9-B>cy0P
zU<%d9iI?umoFQ4tRq|*g!`tubCyi#t2fl8>H_6{~(LSp6DdSO=<K`)LP2^ipr)FqN
z0ZZ;nWnIyA62LZ5Omiv$#x*Es`l_SkP-G_vXSgiWP`43a_6#R8f<t{Hl4l<QHNcKu
zsrW-aD(8NMFHa-Y;A#0opR_?wVPxyMsCA$nVDH}Z&dwIb0?+JPgv)*l6+c8Nou~*$
z#O*z;gz{08nNh6%L=+|YV%iU;s^76W=_Q64!9-XQC`GwFsG_>}MY5B4l$^ILo}-r;
z%%|c6x2!C<+>?+u+=_pY4#4ZKA~QWB0YbD8ZUIp|O2>91O1&<>f2ubcJD6Iso&507
zbW2+QEw8mA&||4CO8D(4%|rfK^E0(imI_}2vnq2)#^;D$%jQpt_*Du0iA$nmkxT|&
zp(-er;s;YJM(aO*{NDVq8(KIWB?rel&_|0FM}u>Bcakr}vzq%5Zs|t}?o7ss<HvV)
za|Yy$|JgUw&ikMu0|&}0ZqtxnkD5-@7LUEC>S!xNs>o?DijzSVu`FgMt|2pBC!zBI
zj_-Jq5`E?f3hyV#fJSQOx7gO)CqX9sb)STF9SyN?LP)?d&h^=BlPqEu9N=@>NNn~v
zNa~|iYaB&K+FB(Ue&-ix=UfzapV+Cj?e_R08Aa}M&<X?ZnB&o_SFcc=5T@kk=Jrnr
zVw0Vg*~-FF-S<(K*qM)V0;N_yPZy_OvFQ+Dy(gfNTHEWHeoAPe*AlsTc#zio{QPD*
zKPvS#5LEStq+<kJ0?Y{Bs^b}tv|*x5|F>`4p;=zN@|j5w2=5a;M2vxbSbGb+&9)C8
zZe8Gsw<UN>`m7>Q@L)G6L>&>p2Enb7xPJQ-nE_!fwBU)Sgj-uQKik^q*z4I<a><2r
zxBG75tO5wn+={9!N(U(J(w&I|Ukp?|ArTF+G?bn)7w@C-ooKHRAx0kb@qbLIr4kf;
zFp@C!0bj;5<b3fWQQ_)_eAZO;4rn0|CP=|W$V|o8QAE(VmmF>Uv8jALu`Y{A@Ik~N
z0t||<BwDoyBy>K}$H#~9f#(9(`(wgUH<<T-mj_D$5gdBDuzy4Zh3;+5#o8ZYXB5(%
z&?`7-GzlYXrvRq<NndGVLI<9f-Z~YzVidQ~(t&QJ+r=niL{1f6{K4D{tZV=aX{ZZ;
zyxGy_r{_;kW?m?vaC{qP=)zjUvzqpyG)M@$Oa&Y_zNPvjZpYI3yE^~~WW1Go?&Dzl
za|zY!z8mQbdAz|ciZ2cT9m5}Vo+K=<h7l2AVIJm3xw+F2-sBClJqM4|;tDVh!#@r<
z?@pFf$<Sw>SI2&hpWgzQwzt^obA#d81m)^nm9Qtl;)M&9tVc#owY4|=9y=W^8nJ(x
zXWkYmi~51GOfmSijN|IfQ~<QXjH3zn+}RAsQFkwsoNcG>jfhL$f095t(|8Uj9Z$df
zA(1RVjeM0yHjO5X;S6bI($ia^Fy7KGs8UC*{_H&1lyx(KWZ(V*lTFTUZ9DLus)Q8%
zp<f6o3C%!zSlU<0G@pC*YE`w-TDLlGp7G~Di}b#gfHb*P^7h8&*rV^kl$*HC<x5Fk
zD75`Ci~o8=^YzOcZO;yC+T);W_S1E=N7b2<JU=x$H92_l?xkNJ3k#1rz3yLFF^eO<
zt-qsIbuTf`_CdPP-h@rkw$$a6`#lGh-_^IP)=}HtC#+kZq$H4D&b`QxmGmnF-O(#9
zAm}p5-jgMLcw5Z&wQeD{cIHkj^5i2qI)1j#^=|uTlAI@c8<g%#+pU!DKjjKxHF>g>
zqMNq+Iy!oWaZ{)6$eoA$>D=nt-k!qtyyz308KY;kRcGJ#)vk39Ge4sPxBW#stxWSh
zN0J&YNAXMhSAib+)(!T-;(QK88~auE&)m-iC28W@rN~`MHxfDmS)e6o-mZML>)Alq
zuZal@^Rx<3&%wo0UV-u;!4eze{a}{E#MJIg%vl#4xn<wi+IRF9n*`-X@EP907gxSc
z(6<>E+48@kUL(l~QP!<qKd!#6*ZC~v<a|tljgQNQU#4q<NSMsAi9$IGg|Lo}4R^H=
z0VuL>v<cjna`_-%_$|5&Wdtr?)9n3fJQ(R{$2gA}qz9v*{qYq_**L@I2M-|@Z4YrE
z6&b^4gk{wlOHkly+<SZM){~;ca3b(?x%tLh2redy+>xDt0Rj}Dn^tlI74a~}nW*nS
zu{V(QxEwtNj)f;8aWFMvqN5=fcf0)NwQC4yNo&qRo5$R$uTK}I<O-?SPMU{LLi9>R
zl}9#6U29;lQW~iZX1aCC;otCn#x=rIBY6_}+E=Fc?XIn^<}4?MY+T}O0@+RjTGhlz
z0&zJRnU@GyAXS=A9yDg9Cww>Dr?$F!R;#WTxo^j?&-cU;(*duuiJ)^-(I-M<xGHJ*
z+?a`s+7sQW>VyYtm*T4Up%SPUsyaER^ohUV2py>F!zF8<j%Q8219r>d;r&ervt}y=
zi#@{6y(a<zX+tLp(NUW7FgCwTU20Ye_y>VNNGqGLPm7XugJkaQ-D8wb#2k$$w><v!
z@aT`!svZ2uw(u8NZv?={7Ch1PRS{w9HZ@F7w!Ke6AKiTmlf(Mi-R``}k2SQG5m&=j
z2@!DUYK+Gpl|}FOj&0j!%8Dd(6e#+N?MXX0BTygK*yx+dDV=UZ!K~lA!J#?#V%A7(
zliaRq@#jV#ORoX>=a{lGx48I4SSis8*~mr27X-QG5Zdj-)KXB9-t*>CP$^Z>V9%4_
zN73!(6AVvP2RBODHt@J;uXqqBRW`D}N^Ojt{>Hkcn5PE<-37Su2-W)B$FHC`o0B6T
za$lPVA@$p$EsT^seUYt_*Hce)PF>%d#ITe;5Sip6yNnD}_j<8C`oSv!bj+<0Nig`a
z`gO+^moODiis%!eO>={+y@-X_q^}XVqDyXrd#<gE*oS35%Rb#cG=snY11$~8Y^7`-
z7Gb;g>9h9wKOb3o$#h^OA^3z=56jNcpe`={wGoJ9mY&?7S2Ly(^i%D1Yf435U61Ku
znuR0t{eyZC=c)TsUU<t);8$vyOBbB$-|zb8D-Ek4lK_I=iz<i+oG%Xi2h}0u7a0nl
zEJbY>ekcYqEv4Vn79r*4u>CKp2!k80hJbqnnG5$N!L{yNj2~TeBJRGl;XxUf;kJE+
zp_GaI7s<Jz9EwXPJH1&Z60FtJr&o}fSa}Y95_!y|IU5$S?!}=D-x4b4)U))$BtP3?
z`6s2$I)sO|C9<v9oBDi&4vw%t#4^Rv0}kQl5z~jG2}OP{U&-BC&c5^viR;w5+{UeY
zGa}*>Gys?WIOjno>fwo5v+G6eA8ITomSD%x9%$U&91ATshi7rC+KEX?Q$&JgYUU~g
zkHPiEXKH<Z!ZxwT8<zIpFGAM3Y_L*|*e0_%fwwfL<6eKTUfT7`{&O$AoWVr+3bqFd
zo&gG_^ILY_jUmt_@>Ig(mvM%CM>O5kXqVnz)<y-Sd7+RP01F2bTj~=myz~NVf63eG
zn5T~gkM1j6FA*L@p3Ls}gV~&)>@e}3SG>nMFK}WDvdog&U#59Y)iTd#$D-v5sq2<5
z%oOiJ%3e+dT!N_Q4`|mZPd{!7#b)SN1B5}7%`4Cv(`!|an8B=$n4TX=(AXI<SeZQg
z>?e0zh+6p4$MzI1vPxg*xdu(D`TW|cwbIj35HnRf`{R^$mWW^gr|zIcN6CX>dA&0t
zwPzQ4A+r`{5P8Tym}Tjv@2UT}WwS@hy5TMJquR70tL;-;*PN-GAfB*EJGG%ek6in+
zB-%A}?VooSQkDuUI~|JuU76Ru_2WFWKT7k|*8}?eb?ZrJFQT45=T=66iq>=I?{@}C
zkvM{OlC;b0xD9_mb8>8Nak=ldx~5N2=yqPZqrkO51*7aeMeJ!6j?u|UOJM?V3ke|-
zH0G^jUBT?ukPNwZf0kP|1*60uF+gaO61+P%A`Ai}g^}aGwi@qb2grcP7jDWesJsTc
zXDZ_Yt{EN7na6+atdwSgv&sE6B`s}C8vF$NinM1$N{tXQpIg`Sr2VE%#<%)aYB7h0
zSY}B6LVmTQFlFt6R_psDoA?YJ(9>5E_BEb_qTm_R2g$>?BRE=h-w<{r-Di;%+Om*$
ztu8Nz#MGv4`41*Y6TqZ7AHPo8(C9&+u6I_eQHIr0R;;~31`I&iC=Ku&nYL}SF*B<N
zxf)nu;avMY!qP7-=T3P|7cU}VM+p}^1NP8s>kJ<l%M7miHD|gp^pT~#IaRVz)N+pF
z@4H<_vcAe1RZZnGy5u_E{F2D`LzVt#%G<CL9?OX=$JF|<Ehaw%+}kS=K)-GRwqgm_
zE)jf%;vU#M2%wwih;~81>!3tHhplY}kL&ORVb7U6Q9+tHd8uo!ORj^Ap<HT_YcBoI
z!&~1eQBsej{<WCWx_Wvb5`O@fS#mv)v){4KgXlS&oxOv$_^0PJEFnf~0rsevurQSC
z16+-%pp$M}lY@#QH1c4!i5*%y#WoSu;5gd4Ya*xN-QOpDVH-5P#==toW#0Lw4FIBc
zyaSnnOB7<+EQ2hjv1sNG!=Uzj=__uoe*4xEt4dTOGzrqwpnoqd?4`Rpvi5+|hQKli
z?N&iKzfCNRG&R)Xh*{Rd6Sbh*Xvdto3e;+-tLxy3265n4Q0R{G82yQo=>P2)k3)S+
zFkgakHd@slD;b&qqQC`LaeXW=w?QqqN~n<(rJ?k&6R}t&`nUv|;kb}YI%kQAiY6yq
z7%nDeR8+qk1F|h)ax=KyZH-%9%R6tlu0d%Pmed4N_`*g;wXCs@_n`q3G5H(DO-;k2
zkxdLryP#liMt&9{=Qf@$MMghcl5Vw>FXcYVXLCqg<I`yxw$l9?awp<Q=r``~`oV20
zq|{(B;GUu8|MZSM)YY<UsF0w*J7YwI_HGz4Rs4qM-YV5X=O!e(v&bgP(w8z$UaxZa
z@R4r>aul`3y>1vCmxIz!Jnk`a_LEVW5jBV-*~H~rE+b1trlq80jIsTYo2Qc}i=Sz~
z;q{!E+^UNT<{H<!%N57sNCW-pV{9xR#aUW2$JiQ4&J+ydd4bS>en<K+KH1p(?4b@W
zKsSrI@uEC=aIi8X$k4TQwc7~fNN|=v9-eA6#i6gAk%Tm@8M)O^o6OcPpOUh1-hoAo
zlKkauyK*x%?)jyA97FP~WFzf5YW_~C`}sQZ<PV3L+)RuN&%H_&s%vX|!J{$XRkqgB
z6k_6PDWgc{MD0$W&V%p>SG<}AduZLcV}Iu8JvW7|p_^{-1HS=;ko^mlIFmftyMz#d
zr%dam0fByb?sLU(z}xiC@^#3c^t6?o?Ynr8vhFK7gad-82_*mZ+m96J=*g2;U1>_#
zWYlg{N+Im`J!Z153E#&}BMFCp$T9}3^lxMuH+6AJ0t|@-P*-)mCir%5CXx+ye3UQq
zajK-^*>u6iPfg-=om>Gq4fSP(%0KV=Qk<MF89hAZcbd}&Z%9y%(M$i;0esbNho4Nd
zx9K8Rb<AqFx(A7Jao_f5{9=nchu>O12hUphF_S^Xx~F8WDNojw)g*XKL}8j~yyH3f
z@EX^U$=JzwFTVs*x6zyWb6dvQ!Z(LUCXh0&WW;9V(@Gnb_qNxczr-+(r+Mdx5zhGY
z*O)=76#2sgk%t>c6CSK+lsWBq8ix)Fx{<T1zod@Y;7eJ%4h7ka_^IBVJ@Q+2>n10V
z;@D-<WmfqeVH`Kr=r**i^reUvJWQaxlGT1sWm3|w!S&rARC*Z{Q}8^2wEyCO3NkWK
z3;NUFQuS-{b(WD=N!1fimQX%9=pM_@iMdT_{dc3eExr#9Le=EfRd9f`_KN+UD}v$(
z{m_?}pTB&$8wb+(tUfD3<WGunJ<oU)mo<<^nOP>nen>rryE(C?jQH2Q1Mi!+ZviHQ
z-?(KxRE_-}E#PQ*4@QSUOorKy@{sxjRj(u;2<7SBhQJzUUj`i{PnRdJBizQHz|l_~
zwy*5edB`3C3wm~;ALB^5A2@<?lfO#pJ_u_-DTW&|Hic<D(&Gz<e~H=Wn2K5l!lcMK
zXlSflwMxJ+_p)yZI`TRT*ZgbUD;=QJMb|`#W<SkJ=|2blDlDOzib`{?*@~&+D1z#!
z!N4lfUx|cf4Th}6-RY7sA^+#Z6yG5yE2oYZB)>>qMQKnVt0m9vKlpr4xVu5hn8aUq
zP+J_30)zwQF@&roT?6v7oegqh!>dy`?p()B_<OkN8N1f}@{$rW(88Q9=UjN|H7_MS
zT@EYXePsiDFN$^Ilw>cBgq@!p{<!liq_wnB8e9wqGxSTT@fzC}>1HP_*0nD$RQ=Re
zq#W<{_vRs#JOv$^;?)l^&eCNhY-dJj>ooexkDYXzME3Z{GyYeg$FvECJK3$6=W#^s
z+?SPzGrVq-=-Vm-l~}%U%B<c0kCPi&ECHd>xDPEWP)3G1AE|T+VPdqg39w)=X0Qps
zl*%QP7!XXa%rw|eh0mWOMd;WPNXk&wwU^}*PyJi%+{2z;bLHDEroi0>nT$0%+1L*7
zg`?qlGh&0krzciN1&1XPK-5?Y;hgvEAs@-mRb&q%YN{OJXO>bA;MoPg&51C}u%EI6
zF_HZh$Fiu$Ps_bEHm77T{5{$$YP7@2+p|o@EAB1$yY2jYH4omC<>p?=3PY@@8cO#d
zHKDsx2DAI{DY+)<`Wy=T>Eod(@7z^wejIBe|2bCfR%mD?J$O(IS7j**si@}Wr^u&+
zgM<I1CnYy06P|~;<_3A@Z<mO=Q^42#o%YY#F<giC?yD*si9#V)<E32_v$Hkd`CiR9
z3;xy6kR8rxj)$zfc;MyA2b;CE4$l9nCXATwy}~EwIs($X=!>)MQcVFMr);4InQ@5}
z@aI7yS%nrb@9^o#T%g`S;4~IMk1i+V0En2`fUM@kampSg&1=^X8xWt_bAGB^6pBn3
zVgQ!f6-m(kxE#^j05K<kZA2(GD1TC>y<=Zzc-VJ`FwZNyyZ(MvA_+2GOsuRXD3G8L
zyU1<gdO0E+E98mi<26dOUw9b!A?1RoZ}?fVu6o*J&ADR}(tjW6uepRi7oq+8kFz(5
z=brE6WRWR@Q&)RN6qoLuw&u%h3XO2OZP^g<*9zQ`C8vTDk8whym4a2)N=k3hbJJ0>
z9GtLy1F8lZ9KxC0oCv?E0gubBp5=Ohg1a{7-}Svmli~qI7pPz=%(%F_i$rOT`jEt$
zhzj~cr@SNCwGGeCFQxiOCHqPuBexzu{uzjfAZ*4&2db8UVTQy0-7D$^B!t2|Oo6!U
zil-p<HtN!mam{PVcTSPHI5@7+^D78!{;FOhe=2~8Z*jrehomDR=p)NbXyP{!2~5%)
zSn4HIMA0B=eazD}G;nkgG<dv4{mtKP8@-9F+xr#jMYU8oO)!BWP3=okpK{i*^?bps
z&VRd=cPRxxiFD3eNLm^rTBa~p1APvsq#=d724vMX?xpGoVMA?wNznf7>*!z$SBr~{
zWu)`N;6nv=G-hDNjz7Uspk`eO*fFPAr1O(A!qn=)qkpUi+iS$KTv~U*x6}0YUc}~*
zn!Qc^*01hQIw^f}bFkV+n%!rGokaE)JpQ|&5Yg`#>3HXsD}RD2EVIJ3@m4GIW63`o
zP2MG>ABgMY_5b|3fzqHsHad1JK!F|Vx8PEuab^48t&0DMRZzi=!9~GGV6FyJU>U+g
zj6P{7hYrGfDnpch^7T7BY(jqtCD=FzP+2^n+yr3?H|-SNfS&^4Y-HWz2}~tFb(!7S
z|2gv9N*EZ-n$_g{*T!_pSN}3Se0Y81D=Yxi)&dkhB*;U(;-8N+tRxm7{Wjq@=!Gpb
zBgQeJ#_(>D6c|p*czwH~XYO66XHtH6U&Fbj(j}2A#ba77>p<=M_vz@XB?SBD6U#ml
z5c?+eGOpe|uBw@7GSE8y=SxNq?F6NK4UdZp`bG+zajsv^xAo6M#xC6X?@>og??J2n
z61*osq2aCrjoo2KBg4ZV!(_oHqubcA<&gy%|IJ1-+%!5zpfAt0h~<pnlNc0_5yWfn
zF#dB`7)=3GYb_QO-*-FSjlnRfjAMCypoqe>p;AlqR`=*tuUWTFf5H^4PLSAnABlqF
z>a}Zk%My{N2Z)_Hy%l4(&S|F*a%Im*TlN^TIr8kj&hlFQl>L9GKG)dgfgC{6hoK3j
z0WAkf45HwEaEG2|jjbCb`8C!z#-d_1YRR>|qIVP_0y1wv-yzX~0i=QEZ7!<<l_;H@
zg%4@PdJayR0<`21&S+CBQD^7h=p6aaS~b0gm_U<m9-Nw^PBF~@{8Calx07|-B<z}V
zA;76EU1kmm<A1{!Xr%6uUE+6z#}yTx)PZ;0{qGH*w!uK&7-<)<zuAoz(I^iM(c<63
z`z@WorUHpL?P6=P>LwI323PVBHyllNL+t29(oj*+E;x972L@1o>NiPGwDgoFR5V}h
zldx}19w#3SkGosQ-)Lojawqx7*=HD5z<o|mo_w*tHOVt#m;kD+<(2QzrS6K*ammRS
zVn{djIh-HXC(h(c*&$_T=#*rQDA@0a$ET1+mk|)18Z2q0`^_BD!hKSntmIl3N}sKx
zSVs-5cSgY^<t_CsLX3h5XE4|v2Jvn~{*d84IXQ(m`TJzH`}i<#zr6CLG?d#Vk1+Jd
z_pWlw-{%+LbD6YT3QX*`Z|jfLQRhjM=RUj$NFdGg&FU@oCBLNtIu=GD;&DMcDvXS_
z(voJSSC~V^XqQqM1efAQ+}X!oQP=nBm#magaX<W9tN36Y^*n!9e*u90itB1ODm0xP
z^-J|@^jnEK!RT=R)BOn3gKa7=X5$lLK7YdCBuBnRRsHYj!#ebg<7|zFnFY8%ZVSU6
zyDAO=e3nWN(pKL%k`UMkmxwRt|8+8QgON@hb#Kt=41Ro5tLy|#13CR{u-iI5E0&#P
zUAaU5k1+T>_a(h!e%0u{U8`<@=Nbbly^hKV05ni2`;oV!>vFIGwqoxm6;P`y?(kRk
z6+k{ZGya(~z`@KPS&PJ=^g>oGI3!Yys=dNVX_aragI)5r?*(KRaJr_sdQ-->UGm;W
zb*h@QNsg;hQbIcdn0_Qbb8#>~b;@p2r%Ns}GATKoEpYjEoqSu#t=fl<;1@O%MNMf#
z32|G*fH>e_JW{-s-ST`^D8o#q%Tnw<JvU)>{>3&yd|hj-PXdX0GJR)OAgP=5PQt!X
z<D~R7sQsCNB&e=dUFnwJKd9bVMulQShQU!uqP~+UKuLD~5KSusAL@z&{V0Cs5|U^-
z`vb@0+bi}fVOe+{{t}FNa(FZ`!S?w0&l<T#myqb2rqU{pg9q>8727?H$C0DRq_yVY
zc;W~76@@4qdg+MrWOTETlTcIXGx%{<8Ke#1aK1}m%}L_=lV(a%hCaMp%h~Bt_pUqa
zc}81l>IVDA(6B@!BJ3DZ;m94{!imx^m=Y}@?>oDLtnl>~r%eY4NN5xsvg^UDEAzjp
z#!|jf=3FmG)!^E+=GVTdz`mXy1!ZN_yCb(0$U=rCZs(poc93bSI<yJs2+F&uME87o
zFQ!$D>kgHSZl<POrOOB?cQ045F}3?Q|L}4pv>gv#t1UEf0$ygWYhv<sbg}e(7O2>u
z83Ic}K?5=wZgbOnMHT^3s5<m$Zr>h*kccV%8yDIJLv=#?!tFBn9m5W-xU_b^`akwQ
z*XclF;jC<x<0kMK<e(wr=LMtQJ^yh8MPs5P?bMzlKf`$G$V`f~|015hjK%)>JlvfF
zF_c-44bkzx-WSxtL3wCu<SmDL|2JyslqZl?xe}^)4J9#mcuvA^R!Pjhan(QWZDq<A
zs+*eUNs)a&#qPCU75;fa_8+s##xp@-yl<LgUwQF+3;n+z`1fV$)&Bu=mK37?I+y>P
z5!hq@CUXf2E+Mn@51%O5Ce&(Oh(p)Ve>t{{o1_d+o$Z~RyN-TD%lm&Wbk783u9<Gt
zKZk)7)&Jk~AVT}Uv8ep{5jJOD1QVHQuhki%dO)H2&&3%&lYvwhlb*!bt6z;Ss&~Jd
z{d*}RrWwToe1;{rnUwzX{{O;A|M&gTd#N_@=g|M_+xlOZ^Iyr_|L8vP(2<K=$5K&x
z`_G9RdOSsY0<z-w=b_{w3yFQuT@d2~KK<`ErK27Prwo38_jD%-;WMZV|2gG^s}Q{0
zjA1jN(^xjNw|gQ=`Ol}*cWR&hH8NtV3{he(&FH1_%%8lUd3oa(;Np2nz#O4HobW8F
z?9@Xh5*<g7EX8QeB?zLZUx9)U+F#xdTlV>ggAyZjgfIN+MWxIzjTd{?tyWKQy)1U_
zR#b8L!Hfij6?3#J?%QlM3T8TblH2~8xbk*`IEClR`)#Z_^Y_W^_B$x=89dTgWpk7<
zRd&&?rlIzO^hmte)>a;yaHSfq%ovdids9?bHK|>+6py;|*rQ;d$RS5>AJ((hwkmoy
z5i2L(w5r(e+aLX6x7Be_<Mn<nuWWh$>Q&tV7e|eK?$K*+=gmKt*_*$Z7F)gie)aN&
zW7LQK+xSzS#L{nG^>&qSubp*h@xF6DuWe1ohST|ID_b<}_r=T@W$d=9xHosHcPQ%-
zw~v0!jDKUbenM>AmvE6EHxGE$2wKf<9wJNp$}_Eu@)DcBSaKotQ6g<d+m(tI$>LbQ
z=$mPUt*-~I6yMzrwp&*HPCq?1W8uZ?a)py)_n6ajX0%^!Zf+~?)o?aWnVCCYBXF>H
z$ZmAle{HzP<2RT58&5oy)!TjKlSg`Npq0dg@mRT_RZ=OT!GXvEQ8_GU&s!>rJo*xj
zAPb7EN2KFIbfS;BAP~#=)d4XW(Ca|fs<{Cnj{6uoWhRWat?HfVML+<J1QW!Ky<%ce
zCZVlB=lkWp5j1&YuwP~G1v2?25b4ANc>ljY3Aw06w19sMCZGsJXE852h~FPc<~*`O
zNaw%scSC?jNP_UiqX`Qtur~Tpb?3A&1sbqFk~G({M??i0%H~^zI#30VK4Jhy>?f$|
zW-@^CMW|;(g`m_A<;RQtb!JGtm_G-DQy~km^x`;Za&mJcE5aKg+zp!YeI3Q~K&E!?
zK&kxo;l!L&kZRCUhP1$+dVK|my_nnrX#gmNONp7`JfVFK2n!<+9FOmf+xqm`GpJ_T
zqo{$-FTm~}(}lO8@9%LFMEU}1LLn$<eYqUjJ+fZTxUDy~?k#O=n;Vn6r0bjDcn|Do
zE^cmul?L1aLJY3ocE*Wp-M$^F0&Z|p`dm(T6MTaDfwZE&1A?F3dax-fBhrTbT90aL
zhbYCVLMlR6Bx+O;%{QblIYjdA?!QqC(CNeg3Ur7ZJY*8_)qC;eJ#BNeVbUMt3adBK
z6bL~No1=^hix&dksJBFixNZt|^W(nTN@vds?3c#V1bQU&8l<0w&myleE*a=tV^!<H
zG*#UUQGWiQt4Mx6=b9bjir=%1nK{!YMP754q~!EFR&R3|6&Yz~L5&%0Q)_$sY@LF+
zP*TQ}w#>~$kwKKrg_01SCAn<lntfIkBk=$=Qs`Z^oo&ruLoruIjo>rLTTi0YIc_0$
zIAd9k@q#bVye1kF7ncq3IjAmi#dEx$Tx6z=<A4ml_SYKeZ^seGA}bIITbLWu8r?Ou
z=isKT7_n?{3(bIeSjZ!6es71?q2Z9f!;=lAHhr0^>$V*%#z34rjXsna71%er@AHJ9
z93*~x0*VCir{KWA)5-%qR}Ncr%Ac)FVLzpe#wub)g-_W9)QeHQ>icbCe!%wqkU9mh
z<5mWSL>pR$ZQC-kvoRAd7vu%5cVE8}g^RYLVdP2CdThkV@S{0xKs%mNlGFVUXScW_
z(JQ-aV7naWKZtg&`^WP7kLsOxaq)*y1yKwz++$fA3L<^!>(^$94=Pa<i3;<$+R;G>
zQNN-NS)-OSz!<@j;`N~8octX%EKggIW{XZMWw=8S;N>?n6s+3R>7rMZm7yr3&lg^m
zy}sdMGdtQy1vb}%945vsx}KiC9do<s9c7-nj#(nwxcCGy3%0+Z$o@O1W8W!MQp*k)
zT;;CZg|<<M0h{7XP;7!!NE8$?q!AyXkdCwAQ4P5eX*M{6QNW-erRmUKfaO4%I9x20
z{xPiFe5m@IzCyHz8MFbe)tdm@h0gaobFA{)rB>CvSHLyP&>y^Y6MA!Eo0y?OnWUD8
z`4W>e;GrQod1BcVhT;@9#zz{zDu#d_$N*}1A(6#Qz-5~}PvSTV+yoAtWF{2{U4Vuj
zWo~^LmHK!{fji2XgE|GCv{<Gn?#Q8UW0?gdqG<Sn(!5Or_AIT(j=b^Es3<%wb<1h&
z3`WBIjo2@YEq5xd)xdObEi6p2G>!g-Yo4`)z-sS>ePg9oQja0fcKS<@2Dg5X!sOSA
zXqw+z2l)JP`JXZk0=H#D5fhUUtqEZuV2L2~_y!b7Ay)+#@mg@OxQIy7i1z79GYEJ(
zF=T1;0DZM4<dikMfGcI>=r)Q;NLaZ^SNnMFcfYHppKTOgUhL56i#d^yNkYlHaYxWI
ze{rae+a3fioN;{D8h}<pGfvA@X}h|+31$+;0z)&g+D^Icd*Zn<xfC@kLNYmx?zPP8
zhLFBPRso>8O}zK_JloYCp~27py&<D}c=$`cjh?1EUt#kFm_#^ir~sY0u(PiGv%&f{
ze>WcER7Y`5ce==FqGiD1VA}XlGT`-}EqVT37W8&yxN~gVQ4_d%iF=)&B8)`RWtFh6
z*R?k8o#1|9q|ja&@4b*=qbR@e7n(4O(5R1_D?*NXq|NaZ&<c!9$^VzG)nx{Six(%s
z10kAr@tK=$u^v#kMHMjqMLWV5@CDZP620HMA9>7N0Id}oR!o74GmzG?GB@{l{|KZ8
z;JUM~=rO3-4a=e}7PNdly^o0tK0TMXxw-q?m*p2kPX?nomxk5>5Znh7AdU;u-DF|f
z0YS_n0szB0&3E@Em>os3wel#8CQIh0d++j##ohIeS(H&*v4V=?Vb@c%Kbxw6<xi8_
zp|u(Fa#;3VxN@)%-e;^x<IRLATrKJjh4GIp9Ifd^*&-@cRjby1#abR0Uv#!YiKHtB
z$f`T;VtzIV!RJkUXZxzR8(n{LH!Lg<XMw%pH9hee+wm|Yc2<D{E^$ltw}1ss+!V=X
zPYG^rR!k=r?9QD#MQ_5dDzHPH1WIyo2_)!SxqsoFYu@oFrr-1GdY0KEz0dl=Fpb)s
zkKr2Q1t@iZL~qUoVt`%^BI!R$u;^Paw+5S)uOm%Ie|yBuJ9m!2p4%iG^rP!J#YMCf
zOU~l!!H1E*=;+`O39)?SPx{53n<*5SAi&qZ#TU!^FzqFIr?}VL0kUvX<|jqSL#_@T
zhwOjli|uoe!+QcTBV%L6HIFD5hMS}vdI3UH8@F#)-)-KdCU4z~-U-eXe2mv~W*}%1
zMi*s9KPM%EZ<R+E>-YZt_g8F^iLuxSSMEkRp=qJQyy(SS2W*8=u1(dOb4XaQS|S*w
z8z1)!#QcgDj4k~xPxq@i0|m~+9ZGeyb${gO3Kka}9Y9Ky?83oa5QIQ70FufXh&*&N
zZnMkw9jb*5-t{RtHT8nk#q;M0N#1+zLOXZLD=Yi|{#gdijk+!*>Si*MP=~xYh^9f_
zZ*s-O1Kr)PJ;JAf`aol9`WqYKCEpfU;h7X1;M^E%#&q%IkFT6jl9#)u8#XBTsBG|6
z4BV*D#grc>AI8?7$ggvaDX8AT;!x!w{?3H0wi~mVH%f9S1SUDre)w9);G?a5$aADa
za{QOi-EZ5zJ{a)w_^E%^WvuMC?eEUc5*ao|I(`9xg{k#9+1V5K@40w-HukpIH4St?
zhsLo<_5x-<d^QPHODR2RYa+C@;i_+XO)JUOf{|u~M7jA+;ACEl{+XNy586ePAAH&#
zST?&KvOSq4wY9Z1rkwZFsHryV2Rlkie4-l%oMQ9tSPYDM$S!VSxv$R`t@5_}7h)-s
zlMraQYS#QY`$%BgI+~D+&)f4D4=Oh#nG4S}o}5l_QB2bLW|w_L&l55@I6Q<5^(Nff
zzhD4R17lUYeDT*}2tA}z(I(9tTiF|%2U+A<uF2Bt@fYg~sd*Q)<-xJ;30OhXcJb1s
zcW>V|efk8LTu?wDe?R7tN=pyDIGoWEb|Gock*r3{144i|Gcz+aHMOwb%fmxMOKbPV
zvQs)A_g3>jVXlqk#Ia-bR%l20eiz!QGAs=mj!>(|_*MAYlGe61{gon7H=eu6{pf&7
z6E5!PRIU7x^~-5yv0<W83jwt6-Ft7_N7REL(y^yfCiR3smM33A?Ki#1HucRP2J+_@
z<9;kGd=XlA^IZ}f-vRjGPKRk|<H-IV9eq?>%o?EJlnWh1m3!KCIgW)&!$U(FnwkT7
z%Y80C`u;*Z>B<0K^|rW@$xx-Fz5QteXCcIni)}q(hy%H{Vd4Ide=d8E(_49`oS?Hf
zYwqZD!NSa*_`h?fZ0STqgv8l|*ogm?mKM~oIBVr(%_h806l^!*M^MSe&dK6}ppu=L
zlZArC>2u~5va)o458={Uc6^^BXTiprncc=Js{+eBZ}%Tnyr6iLH>^xUH-C5ba*<ry
zmy@+qq;)xWOV(Yzb7DQ^>)!qq?Los2mW7zenoRdTkhm!*@iRWWNU|@;0SZ^&6+4!_
zGb#C%B({ZnWuDpzV`D+NFOCBI#y4GTYQjq^JM-Q>FD%k4J$+@hOThR2X|zSWJ`pFr
z^okg3&Eypq1=engSNJ|GqMWDw-hW)dPxyMb#a`Vr7KuB;Ovs<&O^Tj(x~j#}ZkKyo
zFah<gvkT|*?j|()uGuZXYI3e?MfOTPW8>TYn&Z7wCIZD@Jc|2_EY61rZCdR3_%mKG
z%Vgn><_<}Xs%){Vk%l82uUlW~d*nwpzlvRD*(4P1lOY<{@!Lx9YTIqUu2a>TUWSrf
zV;O!Q*E~*>+Fr|eBQ%BM<j#@<H+;s{6>YtA_+0fyQ_JeTtJHc^nG56&UHTDuWVp5=
z?!pbT`|%<YTV6&vv(nc8w(1tV8`W{2ng2kg%qHWbUqZR>L{1+2m|Ml>X~J&tBsqv&
z<*Fsgb=G3dcYno8+1=6=T^bi3S!hH@-2XZ)+$t^Zo;pR=$-Ap0_r&Ci+wFaCvywSX
ztbP|(mGLv%DnC`EKa`YFJk?abBj5hSrf+F#Q?X)G$2J{zx-nYOX11O8dX;3?=QC;t
zRfE;n2>xHBy=72c+q$idO9<{Tad-FN7A&}v;O;I-aCg@vIKf?mLvRT05}W|RB@pcE
ztbNYj_0>I9_f*~WW36P>oMR5_Z*Tn_qdiT}*VLL0lbeN=w32q{JLJ-2x;zc@3_shC
zW0BE%t$llsv$To&fo4q55<|+_JJX&g56#>6+{pKp{M>nmRqKL`_D}VQ+`$RquY#7J
z&66|!axPai8o!zk;Io@;94z6mPMhaMBpC>SD0eh5)U=KHq_n9}s$O?QHon1H=|RB?
zRW%a8mYDl+F`*jD+LMh(qS>HU<J!b{Z*vXoJh`>U^%WlslJICwq)Sr=J1Ry+4e<}`
zRQKI06wk5<7np|JH3wdm{&`o#HO3}Ae_x=I_KST{5Ek9+SU^pt)-FP6kzdx4&GgnZ
z7FJ-g@%yl=^2c}6eP21l?<(__KZVG@u}$gtm0KI5xHtbYBl_?>i^k{8u1lN4^_I4?
zMw4E~^rz{AAx+e26;g90!b6rd`*oL+(O$aZ3C+M69YPxxioj7gA}YvtOJg~n${7R=
zW2MpT>mY;IyHX)C!wnjSign3p-=O$98@ctr#D8t15!o6{nh?0Ob~e*Xqzy}D>uX5W
zo5UB?%V{!>5%41*mSdz<Y_@VtucsJ)NXR;~f=7YonV2*PDP!8tfg-lQn{jtP=r}li
zTpU(_8coo`hB6UG`GIlV^wcpzePD8?;T>81$wI82Tvf$RT)fGUE5cQhft7u1AL8H_
zIChNY9t4FP8hkKPL++Rn_QqCxGuC8)gj)+@iaHk}bV$G{$Rl(2dOujwh;bijzC`jR
zwwk{b*U+rQ`SSNSpK|?G7yzGl1<FdX*aFRYl&}@^QYOqQig2k8jUrz?YgSD5?E_pX
zY0wY{)GzVkYt)sQsmYXhMP;!!O2_YaQ_<|%@y$e%RN{Yd;#Ys-nJd;i9XmSA4Z}cC
z3?NS6M1|iMPO_>(nsuhnS?8=8m5FZ_nx=qW<IpEAsG$315yX8!lcqh$3w@ElR9%6R
zG>W!>a9EslCNDcr^m;iIJ3^8Jy?Md4^<mF>Q>>CB91piu*`2jpa5P^1vobXrexZ>3
zm3e0FOU_%p32HrzQN1dXYD5u?>tElFp3pu}ziCOTfZ8q@_+=|HQp|^NSQ;jJAU@~|
ztrS4a%4H|059j<xA^R=t!+P@&Jqjd<Cv&J|X_kfX-1ev9tkwwsWsrHQq#SC&0sg)B
zT?wtWh74XC5X;nv#o-hsCK1;kKNl{ihj^}3YgiLs3}oN#V3YddCwQKw`4#3?rHc%+
z8{O=jq>(vv7<$nD!9g1pgfB{j|DrnoRjwcYy`zES%5o><t9y=LN&$DITp#pI+#Q3w
z^=`w_<oe=4C7jd9zS$x=!os&kCE{g9TkC$AC<i>kV!KJ0V!n8HMs_KuQ^xa%uGZ6b
zR4XH#Ir^z{cII`{kF1Pvb)0%3`N!GOFM?|{OPkE1B5UHd+PGm1;coOQ<wv|KUJtDY
zn8`Sxj@`_m6~2Zgrz}p+j&Gqoqnnr(l)uU)=!%oevpv#_&p?`jDHBmCmFt;6@O6vA
zY|icaAp3&I%2a#$)`GI1zbGcB@zKW)h6y83A3NNBf_HB`kN3FQ6!#O7e#uL&sfp^(
zi&)+=R5k+D@~1e9CR6!hvW`qVLu}MvJjD3IgUO|`>7evDXtdReP8T^6IAy3Zw-?DX
z{y1|E1`U6*J~tD4MNl*M+rC0OnJGpz<MaNnrWxnD{i$J9vEgO!Ua!Pf8ri-bQk(DR
zg+Daizc<Nfr9$t2#i^d}7A$V#aSboczeN4Cu3Q1jkKMw=tKQ*dmD;|uIkPs!@-c26
z`uy!EL$KKwotjTE@iVDO1&-um@;1Y4u<BZK>gp!7n>Zpp5{PUz!*O2)6?CF*J@AZ3
zJ-$JW@xA{ufSRFOK5^lgGIVt9^hPDo+tZw+=nWm`DyN!a{#Fu3_!YqV3JF7qoZNh+
zhWR{Cf*ZJuF|WJmq!!G2*~LW3=Nl+uTSci~1dlA`Ud3xf)hbticySr-{P9ywbg8a<
z6xTqGyT<D0*RpOqP;PDNS?C${La!nPU-IsluvY~hH9Ih%C-46z4pQnPnd9rrlvv_`
z!syjmRz>w+a=w%KV;d`I`Q&#@=U_+SWTpf=h{u7Q@ymG$#zCIs&@M9{H7-4;-CA2}
zZlJ7^!PhC6bTUh2%N|>ykzxsYAyMXNS3|lf=-6SII-{;Na(zCm5(g!k0<WtH1+PSk
z(|jf4k_q*jAp`voWP~ug8j_po-;%xuy(_g|UHmG^w-8>9FHv3~#kUaU<?WXjcY-1H
z0j)6<-&ck9SIp4GofDSgb2mKI;KbUom~cf&1Bu|{&D(U#2K#CpI2*M+ewR6e_{bPt
z8i#{lisW&0Nys0JM9O0aJa+SNVB0<%8fdhI;Ly*qg|YBPVooF|6XR{TMOC=H6jmzn
z6_N@Zrr(u%t%WjI2_XtqQ^RLUJ3@ZDj5`zuZ%kwD@ABOs=h2R;OZxy0&1`X^WT};6
z+Dn#RtNe?tV^IwweYb613Zuw7vlZ385DN9=<!1M4HFzlFl?u!E-&cEd#O4Qy;;YCH
zy-L)=aqHJI2rLCFp>RmQolVMEg&t%XWnn+ge2-6Qo=Il<yx>saPyM~o8kb=YEpcK-
zhR+^75v{u_nnwBRJGL7Q&WCcE+;l2$TsA{~3GVK>7rHXvEnscy>J5_W3@fOz8!W!M
zv{qwzV3-QsoDpK5=4oq)`=+Crx+W%at5%y!cFunxptiyc;t96@yI`%fS;Ji)CA)V#
zQ}nJ})x~<YK<A`4TFIZ6?+Yvj1)^;;LkN$kfT-B22HiONrj~xR?D93GC6_epNWypM
zsK9p>2Q~QAvPV&6K0ENvZNv;1npSID<~VO-Nw!;Nqi^E$Pl66Vc^g_s$-Db0Hm>iw
zn3oHUc|N%7y6lQJck!-Fx~%psPeGWsNcHY=$mV^_SMIzt?|rHC>lNbnUuCX|FOmKd
zZiuoua%+31t$szc>pT+nNU~Gfbagb6`(7<F_A2|E<=0#cHTJ_xvd)K3Z%iKiKGOZY
zF?+dc`~BzVO0IPl)lHKPxL-J|pO$Jh&t53!sO^#<#goSgG=B9Fz$@4y-O-vb3xa=*
z5#e}L{d;b);TEH{Y^i$qQSFO}yTTep@lrc!$ujQ5@3jrwv7@U|ERoyuq6J97bl1|`
zldCR~_<NIK9w--6m6><xzdgSaMqj){UJW^<bv-+tS)OwJ;SpBzw((fd$|Yg(QurR5
zkgwxBN87Uiu4QZHErk6Z-B51!>=$f6oy$`TaR~wqmX(+EC#uHBc@<w-!bfzTb7ar|
zNzZfq4|<-DljDD-=i3d0;;33Mf9z?~x1ppC{DCDvBCxhZbK31Q>z58a`DCyjb${{p
zLtUgz!#Dk{9PLcJR-OBUA;#<TX)-$BzgIiwx91~aQ*;Aum~Xy__=zIl1oU<W_A1%F
zd4pt;!vFEn`|kkL-c*l!fP!l!jsaT7`y1{(M|m#-3K)aU`!Q>1KL+Tu(d={BO55Ax
z@t>O~5pR6v+VLW3MJawxu`$l5LRe9#8nkGCm=3!OJYC;5;lE4$u3LR(d!OUi?$>d@
z7s5JWH54uQa50|dWysW>oJN-mnSPi3_;-H)Pg<vXDb<H}P#;$GOffzL!WR8Cvo~UV
zOx~D=`W)JaoRVQd6x!tX_;*X{`2MEhf+@f-9y6C-K=%<La=1qizL1_fmmc+%jn^Zc
zs+1=}q?$GD>~|`}8eLYyx*ODR!vW<qQ6401zwelNl-nb0j<=3m7_!*?{ao;Y40Rb!
z5L3l7LITsslqM?hZ+-SRNYc^W<f!C@_GI>ze`@B8@P~QIZ=K6I`YerN7+#Q&Ou4cT
zB=Z^nxF9jzawjv@3G?`3QoiCF{rrJwh8LRbXagMm$%xUC-P(|DEg}mktH+z8=0}sr
z@h&{|522#{IEaeI5B)foFy}C{WQW^+BR2)hLy}EGn68G!!IXqqGAk)T6Qeprcnzy;
z2vVnnYGQC9Z1RtAbb&~oy7bwTieDCa?j^4$`)8b&+%xu@?&w|n)^iaq(7fQKT*~MR
zscy6aj3dW_ZXS{wEJRd#?(giwqXSR(_kR`q5QK%Kd^U^~h=5wG`-6TC=gN+`Lfx3k
z-IGM~>SR)=M^d}ZWX5_-8hL9;oM@47@Lsqiud1XMZO=B)OslL&>bQDVJ#t{|i^_*k
zJv=3<<OnpoJfLt_l-fdTsnqTmgvMd$r#`-0wj!I(m8(Qdlo;Zfwnz<ZvuBg&M<pbD
zm2v2yo;z}KhIO}Q1Jn5Z)o0g()WVfNKl)|&a)|9jF6<+9xqH9a@x}c{9)goC&p%)<
zqW|rD=;S1;BkwJ@RqDGr8TP8)7+UIK;a*8J;@6FH3e|<@$;wO)Js#V;Ounnu{mzpz
z(wt+n<(x9QVuP#f^mEM5O!CplykT}|bSA;)P%Xt6un=0rwoiUE&8CAdOdqpg@>skx
z=?RzQ>A3?0Y+Io`EG>ekusOLTIF7Sz|DF}#q>ol(d(Wc={+2cHR;w5hf4}R!IyvGw
zt7M7o60e7>Tuh=MON~9H78$kF*;ft&JDo{C62iU6VT%I@?x9{kPkyuWBt}wZ(6*GR
zVjqyw$1o+#YrHM_(;0VHtjl(s5<RkNC=xB|Z4n4d)v8Ls_lkX@pZ4%X*Ht}3uaC1q
zwzcWFSv@od@m<pVYe)s|R)9gG%xA7oZzZ<$Ea+{w<3A}7@9{Yj%3A-nl|Epe?wCS1
zD8QsWy58c`me8x^o5y~Xkk)47n^J;)ojRhGSs7DhI(-tUb3)Fs&L;I}f+FsU>)wM#
zm_jl?U5Y7(O~g09qDOH2>pGaA?J~VYlUq^0^FW6N(l3u(MIaPh^PZ&m7Bjak#>VfK
zFdr2nqm;o7i$z5YS?V8F+dy6QafvFLnmyT@k&pPWji6X+%E(+uja-sXz4oI&|BUBQ
z$w2$Z+s~E;mKnP0r1XDxhT8A(?Ze$GZr&EcHQbcM#86ap$yv(^NpieKaBb!&h%sLo
z;RvGMZRCTcN|wW*T*M_Tr{#4)Hk*rDEyUn7O+hoSD%eTx)I6D#r^a4i5|h%*52+7V
z{W#a~E0t}@;Dy*O9^<^e-mJ+$B-vhd3<rxxu^H@MLHF*|$7U>#Q?}L3)DK>SY_`&k
zBGV>hM2F#Q<H~W~cISHe=)Yfu*wjc?^}hVI_$_dkBFcPMCt%mH%=*<$ZtTjyyZ0We
zlY(WTQieg+AInWblc>4laGgOE*_^`WbdJ25z+VlI_|k96VFZmwRZXPiLS+J@mphHS
z3DPD+kzMdpAB`p8o#fh1^tA&G_Bs%%SsS-Ae-1A+s`&-k*fU$@`3x?$R;r9OyCUys
zy*^d{t04(1G5k}5_Q@aS1eHjhPz6&m2w{8WUGbebteOzwAER_ALUYT=<9jdqYWlCn
z!(Y-$8~Rv-!`c+cXYksM*$~rA+&`(epU^|?V)18q9t7C4F{ni<z>|(cn%vx$(oT=c
z5(2G>X*0ql-}iot%41S%(uT8+zQ)eeH90T3wKx2lG&&fOA2K89qhiS;SCGBDFuSY5
zkvn-Ic(-<brj8z$zG>#ERg;VDJn<W&W}t#qW|2xOtmtFtq;Vh<Zx{j@j{tKloi60v
zkH~IgLf?q?4e82sVxkY#B@5-KY_)vyb8*f`oSak{9-F#ap(XZ64)^I98{XN*mcg5m
zxf_$REHQ&p%(f;xRQFP%>%rB&ff*WObP(1>IitGBtGOQ>LKLOEO`f~gx2&*}FxNci
zx(I<Ks~quTvrtv(P60oh3f$iwpd@b&)OMAft<&|3{3fNUvJk)FH`<<Jr=zoGCTw}h
z+>9L7E>lV4s%IxY+w~*!@O|;7)>cx?yS|R}Py8Yn51P!kCAgaQK`CY{<3+Bkv+UN#
zK`QzJtrsfqw|pbyoqLmyQL|ev_1{n0(BGZbc+GjNThPnT!o0yfZHS>nojIeoXL_wG
zb&=_fc|aNV$X`iRj#s)gsI@_mT{au4H?yRVAFsy&`{*G1c0EA2^U|oWl#HN^PJpe?
zl59CCO^qWXn4g<SkX`{&*ww3rUTz|I5FQurFdXI*Dro`UpS`eg<f+aR63zPv=ZoUP
z07>eZAdEibeI0m^cxuBdMbDC4Rfs(*&_KKSnYfG5<ke7SII_a)?^}<Xwo=iSPcJaH
z2xG$u)b+4xLX>m0E7HlZIKG%EeGjIw>$@+L{{MZ#+3lvc#gow_*=Mf6-Z1d(kMKmo
zWoWDhr;azhc#?|BA0yn(AchogU-e?{o5+V{`l-=dz^;8X?5Fi!dMEv%6J2)1h89Of
z=)tE|cFWK?P`F9{y=2EWzhr;Psk7Tm{P=B+^{=oXUEw4m^s`Y@a#Qts-g?m{tan`a
zMJw_;c*zf)mj;__NkUN`tA`S8fAnB-o8M)sP5a8fao_09UC}mTQm&4upE}Tg!5nRo
zYbWFT_hj4?oi(767<+<BHOOqgj#ZI*bEL^E&o3>)$QC8geb`j)b)Ql8r_*UlM}6N&
zHkSMQ!In}lH_B%APi69Z^72%w!#-1e`Ju-%tsYlqnQZ!493Dkk6uL9zmk(ysZL9mO
z_Z2dj5bP<x$qifzZgs6VHkS$NOAp!V{ncwRnLPvkIkniNpX*L1Wwbjqh!W2G{yqXt
z5;C4IJ;D?uRHGiQ3htNSD7@BRTeElzBdD^dOv2x~+|c;cj>Jmfrk}=Wvf!3TmA9zi
zU3Z4kS6Dg;zX+(SCQElCTAuJA4OFI6BMTC9MyXt6G4?-L&9XI5hHp_}s#9cmf<j4x
z-CW;oA^iE24w*pH&ZqdxBw51tC1>P)0dd>!kNAChwB*!Skx?f-+XkOwJ0N$Cw571U
z8fj}?j5KbXe{&$2q7;LXy)<s$7%#*MlYJ$cAIRX9*9-qAxeJu-|IxnY;p6(>$zA;+
zmu+6G_Cux{SX9_SMP^Kr$awkE*i-vg-vmCKM}N!^iDrBJNT-EGu4|@RE~27I*^RhP
zp-rxU=7;8TbA9>u>~b2{sk$ej*5~HNROMdm+>qwBD7TxD=4J85*4^o$HbcY4VG3I^
zmOtCbPZGbW3w)JWefse<pGOnbGrydhseN_B+i%+^$F~Yzq|TeI*S0bRm!ZB9%3Wz%
ziEO<rm;n-^$yFm4j%1>uL$7H1vTe?9AI~nIzWS<#MngsPg2=fZSU8%leet06eKfxZ
z&zr4%L)=_)g^eD*$2)uEDWZ*f)cDRv+OVh7-s~0*FTQ({EZ?ft!41FO&Xyn+BRVrU
zgm+MWC^~3*UVB`>k;K<Sr2Is`|41NYiEfQ4EjXR&m8o%!?0Z8bITKD2l|^f0hxMhq
zZee5O0-uV<FR5xnv_t3D>m6l0ShH>kY(4lI2|Ni#79F#pBD{j$_CRU2TL-VdddH0A
z<L3Qw<I~H7Kh5!4kcDE>HP_;>>br%!jobbvORb>Ps3Q~LwTtAzwKpEIE6M2Rs^Sv{
zw8{U9WO!jKoyH8GK7Wx&O?e+TZ{oGdXquR@P@90?q1g9Qd0$tS!S(7vL4ppcm{2)Z
zjpz)f(cVc$gX@T?Ax<rZP{=gc(#|01(i=N1P_)DSiG@gbO0U}KkIZ<Ty-~!VOqJwn
zel_#U;GhI*FVxbWCB;r9MujTxiNUH_QNITPH&5SN-dN&#m8HwU0~Ujru{7q9`)@A^
zGO=u2!&Lpg`Lyu+Za0d3WLv)oOzE3&Ui{i#$!QX;BJ(HwTyHG0JsagPc>dCg^%E{q
zUBiqfl;>D>2G;QM;ys(^xL#d%oeM3ro(HLDla?LYOmsf=#}8Y=Ww~vBkYe~SF0O0s
ze(~;*tC#_bm7ne6S_TWh9z~?+ZZau0&x6bH>|R+gvc|ELQgzRlJGQUzWzikJPnbDj
zH=fct4!%_t?@#JdK2e;dCuTYyr8mK)U$70pA1EGt(aWI_Mg*@Mt18J&H*bPu;`Ayd
ztT>xGsi<F<+tZ=AB9%T~<EJ0d)hcH#!1}j;dxt}kCQ$)bYkz?GGnGa>soJN}WAbo>
zq0Du@_vsLaSTW|5V}I;?K_T9NM}a(sy!L_ETw@!38Gij{VC?%bwgvRUmONk9N~A8-
zI`xYc#ot918~!$e>l_n{rHgsPI59I?0k2~@hKj=TnI~({Sv#z8?mp;;oJ0CpYvjUw
zYfYE}335H#KS<Ot9g=K<zdGy`es<<(yr$gxJ{j$Hc(z;oF5jBJ_>$)D!QZyr%VDM;
z_XbZgkc<(kx&hcglSuu`aX7d>9M5pCWgJgiC-Pbt9HPs9#lQUJhGFoMe#wiFgHMCy
zJUx1?X}SX*m@}UiWg?ZGNFoC%_?^Gxv_VyP{J`c$s;%cAAE6+DZ%o-Xo_li9LH-qz
zSm$0jI=nTi$I!KzF{+0n)7m~-Tj7HCW<qJ*4x??UBx_x6^tbFnv5*{Bf_<#>fz&(I
z_Z!{6w^J@me%NTkH@d3Pq(ifD|J8l*ft-2jQ;oN~ht5pi<4tto(UU5n!tw|qp+Z=5
zLQQ?eA4_KH@lhQ?<V-|1Pr|UoSvSRUf-v$v`9BrPNFg@EF0PP26ub3Bk_H@2eMfGI
zAG%iX(q^t&H@gB6#&HL!T?^LI8p97eR*&r6FZH~Lnr1b5x!BC>1WB`IDo$=*((3Pe
zCw*Qk(m&%~{*+CeX<=)?leJUS_PMij+-a-m;X`mG1Yv(DAiiSxTIc6rKX<y2%LS}|
z-sT^NZ&Y;JR;#}Y60&{%UKR<t(v=haA#h|3QM9le>o>#`CP>6j?SC7EnwiI&b+5RK
zX>_izL0xq8p!@MNs!>MN+ppg!gvBm}=Z6I-WNS}^=A_>-kh?zCu-e4ZHKHLpl5MV{
z*0`Fz;QxLWConH<4)-^6d!%V|)re8I<I^D?dWlV|7Xh@o?oUs=_aC%~?c_*2?#2Tv
zdWG>Dqf@tTf3cC)Hm)Zm33k>tdc4w&nd|<o-?T-wX4J!h^s6qz$kv)W2==SpY>*n>
zo*hAc_%hLb#+*<HT`?}Nnqgj?QX|^9O%O}(VtWW;jd=rmQB0hxrtFoG&Avg0Lpbw@
z_J%_}$-IG{nVOs`^XnIXSchEU1!Bgl_M!N3c=tbF@9;|!k0xQKzd0i7YE$hQR#n-N
zQc;MJLMx<o=SjXzn;tEI^Su3$$CDkbs&7a7<q9FoO5m&ju|^IOv*C5BIP$LIHBpMu
z-1eOVRziqDoQI3M*BfF3ahF-k*2X&LLob0<S}zr0BB#6^?y-+qywkxn^Y}`3lf^<!
zrhC>OiC5hSIA@?X9vSPm+U!m%5?lSESxYed=JT6X2=q$t<=+I%v)qmER=7g-6`QJc
zV<_4T$L18%n$g8GX&P6-@+%G?jB)m3#Tn}oZ7c^Rt3fx7u?m{|LRw{Yo0TfB87{q$
zS_SPQ;#xb6mk7*mDp**eIFE=GHb#iHuh!`lOAjS_R9+0?Z=ZCx*VbuyXG?YLSysnw
zY-!l3{8o!TBZwq<al*IliGSh~@<$|8gLnG)V&}X)EvTNX$TG+yD=?w1{m){vdz*3o
z(1vMS6EAteEK0_{0>hDFR-`S%A~h9A+9AyzM^H|ONf;=`w$AN;`0jlv$Wl<;vA|Ok
zMc5EcmCRO7R*@F&QMyoZPK=gTc}gid%xh0K8HT(_nHdawmx21NmE!tH4nYU@8c}o6
z7XBqqKhopkL6yB0Dbre&cUaThOjFirA*aYe%3Be<YV?(GK6<h7$e*}Q$KUmQT&aA=
z<t(XP3L}{u3&S$WL$)T*j<Sn?mvW=a%J54%l#s?P9BiB}tJ&X~1&Mqn$2GnAlE&${
z6^0sM19!q7O8rA(+pc^PtN2V7^#Rj{RpWTAv}Byl=Fo+NFpDJD0F}b#KojQeQYR63
zCs;Drfd|b=`(Rvf8}~;tY%V|MzU5p@epP@J&Wa8p8@HQ#qGwFQpc(AU=Pzs8is{4H
zLS0Q+A*lAI{Cd{|p2_<Hy8A8Vll9J(>e=&QX0;K8?_Vn5^^CWMee!!hMNuUupJBB8
zZNYy?&PCLz9!BG0IrJASyX?|F{~zBDRe~~~OWl;f5Dg~mC#mndKiE)8Cn)<rx9Gy<
z(_$)?@8@P|k_hjfzX&r4JiDpC=Ptj6&KG{85hX%(fIU}(F(2c{m}`+dF=hx?=#q_Q
zAHund|6$cyXy`#h)p3$<MY%BdqirFzF2fhvbgs`D$RuXPlmsYULX(Rwii9CtN_pxJ
zYL6lBP}=&<bv}?9EqjeHncH|@jaqLgvB*SUAu>!w>_6?`a1~7=jYDs{jb>LH(>P`J
zz?T(9kuIIdj&ch~=x`aX^{o^H5Hk^IQ?4H;QT_IVg3;P^`=6vG7w7*WEjjqX|NQ5l
z|I3>Z4Z0(h#C$JT@0jjkwJyo8pkBf@IzsAOaXhu%G`V%*rO%(D)UZktxJ?JF9q_hl
zEa@W!snHp9Pef0y9(#|^zrFW#twv|eL-u(3Ywh?eq)0h`KX>3GyNtKxb(P_@A8AA5
zmr9hkISsCE@PA47W4{J*B~UBojeJ6be)-@M%Bp{IEi<3eIWSGp$>uY>;A9q~BfH@0
zy_NXd`r9ix82-_V_)iqJ(LxVbfA8L0B-YD|_>DE)&fm?i&abwKIVRSwU;Ftj?%=OS
zyG$OpYXxB3UXMRcA8(a6gs~9?D!-58@3NE=iuohLkY=cP6R!mg2Rj9mkOSL6#e!g!
za4Lb{jr=8h`7atWuTu3^$iOdknejGJlD5VH#&}$Onn8I7Ttav=^}CyEZ&BgfElJqE
zK<$V*{tWx8U}q=pnN!(G{#W}vl#Da%wlg}ynMcxwY%#seN|M}>CKiPhl4{z`FeLdd
z^^r{o!Ve}BvUNXUE>%sTg(E(@=Xy^%?SvSf2vJBEQdDq@IKjZdv$kbm=SAQ4y&)L$
zMWKr(io)iE!hdJ9{SpbClbpeT=^GvaZ;j_W@o-9pQ?ue1l{M=(`51&Rhx4-<lrea$
zPg2g9^e=Y?_E0as-{>YSB8<SPh@{W9iIQT`3Yts!c&w$j!<KpTaelV~Z$MZxh$-q(
zFZNWZQF$^NyF$AzslO~T=;}DuAP>3yc&B!tM{@5j>#Riz^K*Zv^=C)?_DE;TaLvlV
z>?RuLVs2JNvS{>OF+2Q8<2OX&;^uGti!JAS_oLDB(61X_dA`JY*ohhGMD#!Lpif;1
zvQ}n^h<szzwlYMqWykO<-);YGBv0&~Y?0(xC#rsAhWlw-JdMHm^r%?8-`kkh!Glw?
zk&COF@2}iVC74>+94U~GA}q|l_%E2f8c9dvi}yaLTlttfCaiWm6M5MZV`=l&V(kl}
zH)&1%Yow3qYnhy}EdH*ucJ_7aOC$Kn0p4*usknWy1=ZA}-QFzWBQK2=sm|!7aH+U!
zv90I5XHRVu(sV79rm9#A=#?9_HFtMC9C4%OET)j_Sl+z&>?VjmS4%SKU7t_MX-&b+
zN2Ry<Qzqqf|AwqUpN8>^vK;01FcV_q=LZebKW;f$-}$Ql7{w+;dZi>2rCDy!B~$J4
zr{0sEa8PF2&(lne3|>r$jbaY-dI<D!c{1VYdxb!3b$fm>AOtS>udt_-#|6!0-u06)
z*(64-?G+ENIt&$0P*%&o%i3*VqYaK(`vD~>gt5A5K7wAkph8@BU3L+9kp$m>lMo)I
z5e51CYyUW@vBAorV?T<HCpwgxb{W#(h!OjFnOfSAfHWXgGDfM7m3q@7Y5Q%7_jE9e
z**8jcH$>WD|A8fAy<fSdvG_(-KO{)JbQ2{6m&y~E?uC7vL>w`lcZwTp$c@(60xWAk
z{-w}_!d<{HDm-g*^XhAym<xJai%cU-%X2CvYqTp6Ju*0n>x%LvNvHEQmxx2!aitAz
z7tzSEgkWxA%g+gEn*!}`%S3S0lvj0Yc04h!{N2?uKQO_vT)ua)p@G$NUZB4&V!95Y
zHKjV5BBz0$v+0~AH&<YQ%L=jX<|Te#Zb?y?8dAxXu<g(P2d=HKw4qh*^0V|unDepQ
zLe`*a@%ON4g^&0s4()LADreR|jJg}-gd4`*|5gb=VHoPJD4icp8M9bEYt7M8zh(Z4
zKGv}JBgVOuIU#B2<tT&u`Tg_<>+a&$6ham3*%=M7uDwMNX87r4UIw+N3e(I6th0yu
zc1#UU5A!M`UkkJCbZEFvc5OlIR$uJ9<a5&(XdQtA!3shapI3ZVVfFqD;YOs6@@J~8
zi7<<aaTD6L1-VlM$|fAe-#Z@5@V>%b)GUqr{rCHuw{PZurr@sr{jj<TqPJVy{kLIX
zJL~p?6=rWa`FqBFGa|nQx4=0$lW~L5SB<?VLZe3Kld^4LWI28?qFv-TWG#cr?<)kJ
zN{^ikeviGcaNu2D^9*S^^_Hxa6C+;q9B>{h`}}xvQEGJ}w88rr^G&!`OqWFK-9j;o
z`!QmJIagDBZEt1Ur#}(#iueJV@0;cdiBCLnjMp8ALI!f@$ue;w*iAyr^xvTiZF3)N
zxds@RPRhNlN1@_fTsU$nEzoZ>tp6hfY5(2dpP`js4M~*u%tV}(4B&UXZKUh-{q1x^
zQLol7o{eF-_)ExH>}!op%{}5&wX`Ac8)jp*x{*}-sVCGQt(UP{y6F!?ce{5%^e~e@
z@;`0GK(nf@W!rIEuqwE^l;YrZ(33YNrL(#;nXn6_+$X+^)9Y2EJ&GI_V*437W%C!t
zzEyJn-NgYx@#yB<QGA;>?P*!Tmc~wR9bMp4LMy#lil+mNogi_GC;@|*5oy~x%}%8W
zv(WsAKeD~s;Ztr=KxEXP*1B&SPa!+$e?;~6p_=}uS(lURKeMiex05-Wj<Si3xv3kQ
zy1R+n^WWv{jV;aDw5`p+n~J>b>};>itu3wG$hpBEC9K_CRm@!^9qpYQ9n2lv$obhM
z9qk-l)SZk?&Do^PJ*-X5)nq09`#nc49&Rq~|IKXv`!k{r&cNM(cufFmvIz=fWV`9f
z<if+^`}u_!T2<4S{xy@|?bBJ2;qsXAc}SGSIc^lim+#&*v;6ThPI_=W=6|(=c742d
zvbDv<H_FmGdiTR}ZE=+_0VzZxR|RAr{=fX^gQ~dra~ND7r(EK;>t}|@JnPJHo^5Z~
z^EMy1k{k7G;OPKJU+^Nn@P``}y$nFCc}9T%zXl+pu<WS;Yy-e$NZ2M+HFWv=qv|AV
z;0WA2htT|4^-PQ%-xCDvbzOR#^On$8j$&W5(FL1|i^D~)o)J>O)d1d>z5`z7R9$i^
zs@}oDUthC*1)Vp8D=RCrva;$Gpy7Z?CVY<4&^7=Xnz))W7h}Bl#WuMMhQMH0IPBvy
zMv5B^ydLic-uGAh3*Hk4DW<Y$l!ysIgW}udTFU19gT)XufO6WnIonbc<m6NXFT*bB
zVS^2XL!`h5Yka6#<QZ{7EG8<+tPXV!?u8CcnQZhiZuQAqdaAa6#)<UEBKOIE?*{DW
zjI1oKeGn~yiW-xgoXn{}D(F@WJVYb>KEPqX`_$IfGv21Dr?+HDiV`g#B$V~fS+p=S
zLxO{;QFL#$e*Q(qH@6a#QEp3=p?^Wh)1DHV8*Qd8HcusTmJV|3Dy@t#wZ;2{vil(x
z@$CSWgF)K^HL0P_D94bKsT2m%-++z;3<Vk7FMYbibBr@+0Ot}Vgzdfzyqpfh5w><`
zM=Blv0aqn-yT*vyKRooG9RS=stl;L`68+%k-yc>jfk{xHY^Mv({Qah}k*3}e>HP{r
zz-pz@)T_{Qg(7Set4#$ir9yrD&wxJ(ymIMouV06Qx9{tX3r5{l!-{*pcp`IK{Nv09
z*Fi<D3_n3h5~Hari@S!3XbtP}G=2e3YN_*Uo~zY))`|iHKa$L^jSW=;gVBi!fTLh9
z34P|51Ctlc5v92cJ0DwB5bQ>ZR6YmQzb;x%tp@t0NK`u3HBSR&w0(Gp@*q>fOsm~(
zGxkq@D+h<QGHo`H!v<dFm)ry{{`ide_?+C_v;AqhesX}R00eg^a^S-7_42yWcE<R&
z<pRgk9V2Ec^Itc`7wr>w+IF_!gCnV5nJ8a{X(?Vtk0f;>p2a+~-nt<izk^=014B}S
z<m={^77ze$W@eV`%3uw-SB!kg62hMFpG}c!otO)@(-fO6)}^qVDM0O~NTJ}~aubTq
zOYh{ltYzJ7(KM{YN&Wk$A2rx`(?Yu!XTDgO%G&vfhni_5))mtsPK!j555K|9`R_Ae
zgsWZUkFgq_hwzgQ{3J2VU|0R{HRjbexm5@mGB!H5#Xkce98z9-(Eg)9`!306x96!P
z=?!vnb!4bn!;X;@FRE+YI5iR1f8FL<HdjdU+oxZMT(Yeklzt=Uir{I~s28&!ies;Y
zD`4XM=YwyqTtf0hf2#EJ!ii0s2}_@vLTOo|`Or-sksuMp5dPPo!8K^d39wSYSR6Fq
zpO1R^W-krDm<v(7{TPcig8SbaV|-2a;#Z0?9N0T|<)NEbB1@f_t`m%@M8^{UKGvL-
zs~Z1zclFc=_lwH$l<SU%B?5>c$;BZ&XNi2`f?)r3GTQg|_B=tx1dD#7Sl=+SB*u3D
zn*`|-b-*h$I;u!pS6w~P+q(sfCeMb)CMQM3XyB_%zP!GcK`U+9`T7y>0ZT=q>nF_6
z<+#@)d6fW?;B{!`o>Cq-;j=M8{uLnWXzY`Kx}+o(3roc)C&-g=1&(KjpB=)$I|-Oe
zMd!~o4?sUa4FTcuEg>M(4=n?DjYNQv&_%_brPCkd2sPK=UsRYX30|J7794|Xq%=70
z?=a6Ntyl|UmYk445%WyE0a$yGF&IxPR|}9hz|!*jnbXYJFD?UIA>wGIUw4w~f5}tN
zj%KDGN_RY+d=%Rz9~VFhy!q%#9?>5SMzG7y-kv2${qhC5F8zSUA0K3OYd*{raqQ8$
zZ6A(rzG@+XcAgz{=FKg2<}LBh&2sv*v%Wq7u$k9e0Ac_V+P+4*vV&TvumAqVGerR)
z41es^OifRN{PG)vUPikvPlns1w^7j*Q7mh?=hZ&7NP_qfab|U%L3$xX@w!(-*v2?$
zThAqy2_HWVjOSXiG-h4kI!Xl_24)w)QHafNZf*u9-pBz)X69hkgmIqoM1*@ql{?A7
ze!L!Gwjl|MlTy0Y99MXl0G%S6S4$6?ke7=2`b$`v9WBt$1)dBI*uR1IF6}KkV19iE
zo;iSWV#&s=LmzrtG54Xtn3qlo49e;LgWSq5zZ(5~bK-1~L$UJn6JV)4<M@HyRqQX|
zxhyR$0i)BPc>Vn-<s=r_xe)Q@JD5VE*a(Q(PO{jQGv&-cI+!rJ!Eq1~6T>2I_=%Aw
z34mNiV20z^2QWi!n;9#>j0D^);4U*4e@=ZZQp(aQ(?Z31HVr3r0(a+4aejXO*RP1r
zc&6=~1Ko~?_oXGcH5C4%_@%L?l!&^j^@B5CP8Et~yPtndoOU>BD?A1RobhI8@IAWZ
z?M&!D`gZZ@U%vbpOPj+V6sKvp2cxVMad9#L>2;yzC`Uq&!Dl~NB@a5y<$w0K6xT(C
zkmV<9EN3l_gP%BF0^_grLGj8L$O#=G?6hfGn*UvTx~tT*6092W7rO(HO1n7Dz_vHZ
ztt{DFb{R^1{<*21XG}w;xHvZy+s!>FHL>Dau*-eP%=C*EEcves|2etO(`Bc8U$WJf
zYi$b;G`&Vx0;&LgQK888ql!Gwe?Fo$Dd1Ms`W+;6yu*ZcQQ7AwNR{TR!NN<BcN~4b
zWx%z)X{e3Py<`c-r=q`dRwQ)+p-Jz7WAT{=6&uT1C+6gI3Y?~m0EK2~XejN3!M%&=
zEF7t3^PdAs^v8JkrO;SMw`MkFvy=W|D<>;U9~iR$ZsQ3ACLZqZ0}7E`Vq&}>u!jS$
zE4TI3_t{xZOvvTUYfleqtDS#W(R)4Vy<`Xh$Nb^pyLeeNMhO5{1xZM7NngPD0?y68
z7m5JoJl6K%_8h=TaJ$6`O24hq8i-EFTmI*8T;y|JGKcK_R6W=S`(gxKVIZ((tbom+
znH|{6YhZv^uifK_C@ItkAkm|!<yu^;w+|<0(9w~1IRCqx-!rP#Tiu~dErcQ|=FiS?
zc=vS!ix5cS@?e<(K~SLDnfyb91>qSM?7E%4!pX{S-mqTSfD{}k&#b>y0Nzdrfq<w<
zJ^HNL3zRH9WLQz;7fh<P+Ewz&z+MI7-1dmh+n@n71}elP14o(yTy#yC%yP;}9qr9K
zM5!f95iMn50eTE1e_Lc^<WX5$+e09GB2wgl$sSk+fy@ISNj~EqIC>BoYPdW51F%hi
zA^(`e5^!7$-CxcHsJP_v<(`Lyf1G7_JK7xcwfbY05jVM<8Z)5$ZX2ce@4n?-o%?b_
z<6j<pu^S&`NCPZUetv$mA3-`t1U8*~YCr-tBjW*3anCVTOZ6o)A!CshQIYa-_>)}t
zzKjL8cB9R#r9Tey)=`d+j{zoc9+bdG!j-l5tuL{`pgg9QAV$;&6rO5vhr|v^q5-I+
z8@RLL$&q?SM}LFeSEK<c)QG09hitApFD$D^K}0kNf!1_}z&%_>ui#0Y&w;UZFxm|i
z9&j1k09Ti_H881158Z>H2ym*OMO143mSeHS+)o!fIG1V{9AaV=QeluT>%R`Ly6LKE
z0*p8&Vy}CDHvt9>AP;cz@_u1A4iqC-p-0?0xB7@;3>`2R2-zk<1PJW~000J|7r;MO
z19^#rR1A>6z;6FM?}ehM=k55K{SlHffsTnC(nkvfVMAoXKHI<(DT4GO0wf9G;p20i
z8Ch8sCnZe-2{_ytYrv0`d7b|10#6$~KIndv+@KU_Q;-Z8?|Z(DJR)u}ML<d#3yvU?
z;j0e!IR(e05``lYCE_iExAUp>3wsmUcy2uumv~vo{pCLPDY*MRO(2R23PI1X<L4Uj
zWPSbNTaxLJz&UjcsGs+#$G`mIoIWV$HjpFU_I7V>ngG6<tE(#q8mPWvpVE_(g21oU
zz8K2M(#NLjrFT=@Ugz^f6m#<D`sW;=BZ>nMgY8s7L{n2!TpW6@IB@U*Mk67Y#c2y>
zd{Bzr?J=1GDQY{3axSPyOj9J(ESxn|Ul<<G7@y^*$Ik99z)`9I`@O8}SHQ&rz_Iva
zK!32YVSJuL2024H>v8_lOoPhU9qBo+?_{%Nid6ysJiA27lTp+)HHCEnM{fiQ9&@wL
z730$?p@5Ecsvl+a-yfk~SxWifM|Y-3ZY5MrQH%{pD|VkR94=yt>R|V9%{>X*;5Ayx
zgD;$mN`nB>5}wrKWweH?Ettk%B0k5I1M4<0<N~_7p<YIDaWRO{K?62&;2dA61itMa
zpjar7t}ZXlxBzXD7cegXF^3rhHFz)M3+T|KDo^IkKAJhwQ9f`9(z>YpfZS`mKFmPj
zikex{ix1Kq5eE?Lw}jLqg&^MtS|6;>QdauXFr!}rV$ec^t+tWTdN?MPf4br$a2LnK
z!~hzliLtRNNK!=b2Gb#!+u8w<Nk?Y^OtFBN#EJ?T6c5o85sv`R@p=uyUr{{&=pzJ<
z<(dLLSna@$Ww5(+XC1%_8t==%7KsT6X3Df3fi45p^^?d)AFv>VRs<P(QNQuwQsuUC
zfGZtP4AR|zH9a;KtxgFw84-u`7UP-zgoy>>dH6UujG;?QN<eaL6r|j2gsn&_X?5tV
zGT2{ipqvw}rx1kIF2n&RCNMh(5!$7s$5Zm(&?Y(9NaG|N*#5IO<l0a<;GCIRTGk1K
zgoKFo0x;%KNRa;$5F#~X0C0|kQ$#PXpnzhKHS}D+!KxHJVrE7I++biSB98wFU<80^
z8cv)Xk!kuo;G-!i<}_JSAC91fri&Bcc7Q3wm_q048<H5chZDaibyT##uBl4dl9Q6L
zLGdq67=keOOn_;c$cYhE6XvBA9HfxSY-}Pd1EeAF{u*DG85@qRlVZI{D+N0ign;Uw
z4*1tozEr`fD&>f>VD$f6co~O|?-BKXR3tnw;9kzBr;Cf5-SU2t*aNG_aC2WSLQr^`
zJx|`zw!7>M1Cq{%o6~N9@&(Xafa56b79Ur3lSK@CUgMM{2La!^`Ehn?Vj|2fMziLp
zAb>!#JV6cwsZ9Y6qnUK?U(<W(f{-x^Xa@%)BMcG}7RJ?|2rBePH+1AY(J#B7UOWc3
zsTsez85_orL=ePfIip}8BB?vr5(kcW;9q99HZ?W%;9LX;2|$-62#Deqp2%O}<b|I{
z6*d|cTy9mBsQnW6@i~>j2Y|qQj?-pM_2oi*y1y6)I5)0+pbA-7Si}elVxprL2BbWX
zmim}`0o;X{vDdc=+{a+GImI7-n0FqszJB-Jq-q=o)ISuoKuHo&UkDaR{tT4c2Ykdw
z=&%gJq$(l^9MW~OBz@08Jdjn3<_->MxrB=R0Uv>ITd6h%Q3WJjC$SI#$~$Rl0RROr
zX>R6M-H*ckFCi}C@tfd%`376(Jtvr0ZzKW7e7W9wYkT{<0isRMDIYJdmcG6aH8q@o
zjsS+ivFz@7EA(t+5Wt}X#7~Tj*aEUJ7)s`7hz0cW+S>E|{W9kKNU)1gmYed_g`v;o
z0@OG_YXwnzb}P*s>s7!i4tCzd%`N$%l?Vc$0I@@ej_%-g$v`3#@;rVSZQ7CwVC-L&
z$*{4p0X6{K%N(+%Nz^Sy-GQLMb;sq<KBp?z|BQkpr@cU)RKeer{h5eOeI6gwwn45)
zzT9AoD543tPJs2#l>sKncx*DVDozMo`w!;DZ`b*4e{95_m7J2xrAZCZzes9zVF9&1
zPmk@}GDQ=Q0DHDWjO?PqiiMGfiH>yrh1?}x6oQko)`*0T4tP5zu3c>866N=zO1TiY
zBlJQUk>ra5dQfeP&n3{b41ho^XW|&Ydm@V#M>~%Ux=+Xxcs-Rz@h3vEioT&1Z=2kY
z9}xXJSk!_c8pdc?$ySkw0uLN=D)Q6*HhtpZcOZuTJ3=UcDwnE%1rfFppb>Q7jB%Ok
z?_ff6shm4I0IJ{pMh@V3c(U<;=>{mIT`1b^oKhDIn6BG38lnF1cILR}^A76IH$Z%#
z1q>o-C8aUILIn8BB0!+46&<XQf&d<2L1$gUvTq}hrAE>?1Oy9(3iyj;DdK31B#7E~
zX}?JrTkMIRkN)>u#WHAmWn6-qok2H2&JMbT)pd31a%h2#a0#YwcP=c1pC@f`Zyz6C
zyM=gf4Pt>}6UfSA-ZoAIC*On&+Id(Ipl7rK_Va}{=zHCP-i6|$KQyey{)0FNX0lcM
zCkLG885S3(E<3v3^G|%x@JM%OPca&rluUw}UWcCZX&9NW(VX<y*jP7VhEl2+;N+Fo
zXGa%^;7rJ*O8hj_+j$;=u4tT<ba#(=gW`qjIo>a4r&4doAEa5H&$^YZZCOc)fY<3d
z&_TUViWV*|wW+CdfC8(L9+V?X_6u(gXh7Vie=ZUIb=xoSv4Lnpb#O0%D*MlTY*-c+
zK{Ne3+)mG%#c*{ktGc!}6Ubz+AcFE!jBRLe2a8#=Cl;VM!1dDufD^!|9ARERIs)6U
zF<Jr8D1dwQ?h#krKX~{hi|dpB+=okkYcd|iYduw9*y#(tV)6!A3au0Y5qA1nD~T~y
z%%sMhp!=qP1$jkPm7}pSwCh_Z`{TQt=#|w~iiI|}eLysQCO~PI>vWT$VQCBzhu{hn
z<AW`0cvB#nYb!boW$FHTDoR;B2dkVy0N<(e#QLg~)o3|RS_1<X9eQnE97!JT?kR4?
zKvn!wH;w?^{=*oFe-sAj<N-=ZxVc1Zd%LiU3Os-MqwY+9Y%@f>#DAxZ7!gPhAdWyQ
zzz7>VX$}H<2KP}vY2-kok%^Hp@)ksV>;vr{nd0o`_8X9@K#sDsYUktML(r@Q3Li_*
z?^zm^z|7YFbJ~I4yaDNqHKRZjT3cC>Bc^-~f{~bcBCX*{*FYW7W<u-mI?DstN-;5{
zL%@e2&#?v3kVQqi02QS3Ern_~!O<hq*3@*m{k`Xq2`K$ls`;T{{)OTY7Jj=l3usD+
z$Gjo={{H^M0(jmR>fNu?@>X8qzG@&O`t(n`5Wu5){K->?KFyP}p#i+qT!}iOK=BNu
zz-sbxU>n9UQ44pX7#DF5ptbU4(W^6_XN0dQm~b^9V);2A??McvY-Pn@o&3Er1}i7%
z#?^2>$18v&8YwBbbr&!p0CAt@(Ex-580a&{hT`NP5$<zRk|e{sJ~*Cjks8t9`z1E5
zH&-3HMvP!q|4>OWPK+S~n~2jCS#`ykuC|m`3iwa#i*<X0OG`cg(gY+v7_mHvekjyo
z-@;Emg+es&&;U#<$=mP|h#SBzc^b8+e?eObZp&JOTrmEBzc>Km-eNSJLdysc;X$1+
znPP86aF8c`RBJfQixx?B5dPcXhyyZ*Dx-W=!7?K))NOx%f9RL(pt!HgsSUNcjj|bD
zN%Bq^E}}8zFhuAE6b5j|SKk7#S*o*W1^_aFD<Y%e^F<FZh(MJhqzaB;ji!kI0%|<a
zrx`t<#qEADlVFhzl-y*eBaZLvTt28n>-Z^&iC6}VVr!t&8zdmFs7Q?@@!gREaU?+-
z%ifA1`bVzj=sB2yz{+B&*_nzOkbOX`M}M0T878J`tA{;Mg}Rj(`1pt-Fye+_^6QT!
z8CiIzriS%C(GfH=f!uy>lLFk|+qbKLsHM=Vk>ZL`&l3*$!=oSfT!Csw1_l81Mz+9C
z0!>~W44%7le~H<;4-K2^K*54Q+6;YCE{jnBN|OQ7jKVBvfn8M8&?D*M>IxMZ*;w{v
zA~iM^)&h1-R8*89NOv@+CL`J-Mnsn9p2*<;o2f<!1Tv_G@{HR?8N&>C?oQ;a(khJB
zw>Ec=M#jb)AYla`NLaAk1xv8ur|Fw6aTOEsA&G%H4LzQ6cmbUT2nc)b$jDd(DwG0{
zajdu)B+wY~5#{u_f27>(6PA>MdNHY;;?*M_fAN#%{k8>eNg!CCl%)oNejAW=tw4tt
zq_ODpDn)`S+t&?x?cs=U;w9xf&pqfILzqDZ(t;N{C@$4D0>v;AXpuv-sW$dZS7J2V
zIkisDXkMz$0)Rm*;X;gJNLYNkIjaBSxs1)vf@oTr6MnHbJK4^mkb8^vVdh1dJNKeU
zvc)1ah(W!TDiA}UR?aDW3k(XH%?Qd_Fj(qRBghnIYWUXz>cw$cYDR`P$SVZxu3m1<
zCGzKdD+cu<kU{-oxb6L)*h2$AgyG@gS#7YzZ^i9Oq)}M#X1iG7cd7nhjvGOaRUf1N
zK-D%I5CkM6xI_qEJ5`;JM0j}pNun{15HmcnrA6%9e=3_Q#Ld&QO24U+*)T3T+5qet
zsEh$wQwRY-qEvuPb^(D~5GMo>XvXpQ1O&xpWrurv_}JLfp#6%1&&nPGYCGtzFvO{)
zaGgk_apqY)_Yvmz3IugJGFQlgTXqT+(tH4oOJo)lKrre-015>nO~YuK!DaA2%_Lob
zvTo5R1=Gya^|E6czaUUxAbbKWS!wQOH#av6oH4|7%d}XE<Z_q5A|9+OgD%0?0nF#F
z<eK_=Gd^epG<g?70eYoY!^B<##!I`GF7e<X;V-(tNe1IJ08uGh<N+Xz2D_i4gQbS_
zCGa?Kz&w3p&**2!6TZ~e-oDe*4UPj~AVB;z<-N`8%cJG_m6cCJ(c$66x^=V{ve1YC
z^oH<~0E)_gdngz5*GQ>TKglOis^E`kkMBuuWU{LQVdsqN@NrxjE%0618e!Z};d49D
zkvOSNzObt5<lrE)FG1{d7S>h(0+u11p80nmuhee25f#f{4~#V0q{Tx9kmWT6@PdFj
z^XKa7GmmX*NeRe0GS2A4oX`c6XD9?rrEz<RAq8K*N?r^f+88syBjEMLZK2HW72rN*
z7bJlObE>X5Cntq5hz2C525TE$7Zo+jpK2DKBMZu;|ID+XtAl$I>%tF6H9x@;@p=oS
zAU>zDBP0D~W@b(wEiR@9%G{%hP6ke#fRK<Qs@S|g2n@5mP2%73Ac0guV&>sT8X3ue
zmx|n$c*iQ`HW#8X{ip3W4vEdu&wT~0+To3LP*ngTd*HDc+^M5?cNoV<3pGJQ!Ce?V
zl0zpFyP&C1=&bU2UV6_b0VO;Bz~c6xs~g1jDW<bS7lpL>m?e)x;LLF%^mMQt$gKZ0
z2ZNsO2zdsrDK~13&g<7mP9W|qLw6Y`g>+?qbF-%a)YQ=xY&aNRKm;aaA_Z|)&~HJ-
z`v3u23^cB!3r|=O5D?UG;JepBc>%K`n?0Yj`|f%T9}S8)jKmSMq0$-|oHh}^(>fRf
z=n9wwqAn_m@8sOLY$=bPpQ6y0f%&3Q9%0CZAwX;*Gao#1WWQbK0qKFpQ{Vw8iJ;rW
z>@2*Cu<w2P!!ej-1Ka>gJzc<+)>WO8k@1B|7YLmJ@wOmBX?kQtVd+FuA5BEDAWIK3
z0=}L|uj8*%^@oqMTy{1+;uDSJF0b~CXu)yEhcRB95hqfEI!eQ`9o43yj23E{VwX>`
z<$E_MZoX{<SWzo!m#E6MC9uyAP2#4XesVuQ{bVXWJw0`~cG(D;i4Q7)QDxHa;S5#N
z9~t=ASSByu4b+#f?O(i}YGxvESKM0w)UrXFR}IkQL_|b@-6i>4o`L8%A+5QiDfy8a
z2QI9yDqmi4Wm5zt*7H{cRdK*mh$fRh;4>xJ6*vNy6hM_I5rMKx#8dKM^A=0iavXWy
z74u0&<?}tL=Wive0)i%3y2yel2-*+y_CeBoytBUQwvoAuQ-i3$xiO*Ri+?=@lmn%#
zqmJUZ?;lCLGqSek`#l^J?3!sKD=7XV;ut_egWZe?@@p&bS*1A-wz>s#XHNO~y%b*^
zmIcn)KP{3U;vjDZ@WvoQ9?VQkQDaii-}u_gp#Gjx)Bj(G_1X6vSMj~#`Qmd+<p&6<
zFR}swxxaof>r{93_67-{APWpKp{<4If!)D6x32#HuFI)y%9qdcJG{<ZnVS{Q6U(Qd
z96+=Ogm`YpH3LaW@LU=cdGItLKIqj(jy(iC%n2XPodn_&6oopxUGB)6SDD$*OG8*h
z!UXDD5V8pp!skjgQT%nl*ags4YC1Y9#IpDxw38BAt*17GkbmuLT?6E!5~?c&eP-Oh
z7M8l^78W27GjnkhM06a0?}7ywG9t-Ob(z#2?F?=`5tr)cu3;7klOA;ecr&)!X`n0Z
zZ>x+jR5;LJpu<+^VmouatXKxcNnPSKy}bCnf&szKrTSdJ3SZkC1L$x9#O?$znmxf1
z3NW=myA~j?hG5bjXY-_^GP$iKFoGk&;!LE<%>B}ycZ~0OZqEpc_Uh=#PlUALDuIW%
zphFL?PIsiOQc3zHm<hn6TkSU+;OUn==)0DhGlD0F+LQ|KM6lHEpI3)ZU=343ia6fR
z;w{LXCvRCXD6YAglaY}DB8rmGkdUkZJ-j{<JgfyA_wev=6-~6j{uV57A9$IA!+0Jx
zWe>R;#Qz^@Ul|o;+x<&AbSN!d(jp*I(n?E9>wusVDj>oj&CntZf&vZ*64EFk-6^FA
z2+|FbN=ThOKL2lLt+UQq=l$@m^*oQxJ@+-&wXeN@vA1kyH=C0RMgDmNfLj}ji|twW
zY-}X1Ugha`w|0XDwQwoCdYT8FDk091FLn0!Uj}mSZO8c!cihbD>u;3$+2HEtze-F@
zOiwRTCGWfjIrhK2xE2S>3K<H6j}K48%f=sZn)zdtD7p<HZH7*E_lDp`%QOCv^zra1
zq+X9FYt8zyMef!gw@@meoQ+IPbKSn%P#cyJdR%FvcWgpY0V>X(8%FbG2QpVG={D>x
zmUaXz{9i^2=_*{l{N>3F4StX}TZ>rrWg8)_5(c?FcJ`5RmtYTIuhH@dRmH}XbX)3h
z*d+-KB6n&>p<`<M{ClW=AB{`$PPMo5LSswe3b&OZXcYU(k3Vq+S*B?Fkl#Dy)@~DQ
z{7$8-OGr!BpOXQjzf<RjFV<t1oPBg0@I%Pkp@nWgOd*G^f8vKAcX;OPj6}r4b|WMp
z$R$SH$o}x|C#pmy0=`SWG~&c?nfG=x=_VhX6ig<x_&bFbnj5?L@cXUfi*swZdi*d!
zPz3T;QF|%p3Sdi3AW)@IToC}%e)h@29+GZO+PUeDz_@E`f__ujOod81`R;*di(y9I
zb?OwAcmJ|9<A=f~oO$xM2sBKx;Rpg`-`@1QKXYdPg_;7*ecFuw{*aTfs6T}V$0^CX
z`~i4Rffgm_V;D#ItTToK;H#*}NTG6VP0en|lib|g%s)P~e)6P$(ZbYJ0@fW~`2w$Y
zXJ@B>Dx@Ni&%@!a1po%P9h)g%19ysg^&DU;-@K*u5uwZc_dM^Cq~wWwxkO^=Tuo*g
zA_BOU++1CUhK8O)=hGBBEN6|{wpsQ7QVWC4$Zgv@I#guX*>Oe@gh?V+-Ik9)&;iHv
zlEzSb`+F$z8fER4rrU_R8x2sG(*K6+F8>Q6N^NF6u_ND-)d;XR(`6+B6J7-5MgU1M
zUbxWW4bg5Lj+tt<>{Xo<vEVK;Qqm8GITVU|>oruu@}S}Iluo`(^JSN}X5w^7%Ie<&
zcM3Xk@agZ%1{n((^T56`Vxy?{Jz;8=*_(qV%1~J2hQqL-%^#m$bY=171MUcI?Itcm
z42Mfiyy;#FvlFx?rc)c>tzE4$NpX|)npL}Nu(j!Z0pdPH-7n^NUp4R8bPMI<VbO6)
z1c1@aKbNuXS|J}tw*s*q%`Dx#IP$?{6P~l)0hKM>qHm>Klz$YzfJ_W3OO>d-ji00g
zy>kJsyE~`eX1P!Fy@hLI-n>mpf|NO><KA`_ZsN!11J#iC*`&y^t;L`*Z_{IYPSmk`
z%!QeZk`e}A@E*AUPz_{B2tm~_q&}x+Y3xLR4FM>)c<<J&sFV~&)u@GInaQFT?|tZW
z+_=1CsY|P8eS)^x+u@wBnrjv5ed~R_rJEs$y)sr?nvRc;;fleg=7%tMuM^M|<ylCZ
zVecTtRFsw3S#8^?4%VQ*a2XVyJANz*(g*CZfMPq5a`wn=|IY$_HD8}Hw0eAjz&H=P
z`aBePr%)k4=Uk^pk2r9fSZsk#Jb`;cI{*e)fMS3p6CSvMy9s-XeHWS<L%Qn&bC947
z>e=bO0<U_IwJnT806x?U;<j_%s1CJfUz2yn!)&&(p`<O1q#sxR3ze1M#TKczmoDau
z0d@qN7WR(Q<tS+O&an$v1ti7~NT5mdmlW}i#zqV@<A$PjogViERNMjH&R5=lfpPC%
zCniI8IDixh@F{#0o-J?txH>@v!cr`{_ES=_<>%7it$Vmu_W!-s?Qm3c6CQwKKkt7J
z(wcqcMeyGS88de$j0atTX8FH}Kq$3*n8!_Kc@UJ~@<HOi9s0nLnkw^lMeCfd+XB>P
z05^VPCfE}G!}0P_SZMg&LF)hBXzBZGg5onb_8=p>|Nc}h^T(*{*QUfH3S7JOarXcH
zsk855-#>}%O<^JP#02!9^lwVa@C}c!#*a_z{r-ENpJRA8|Cpcb4gY^p_KXWx-<7hi
zG?<DvV(K`K8omHr1SF1!!ZDV!DgxMi152w<|0Kvy6u#0EUob5`{O?(M`~^J4ZlNhp
zEcTS(X~vQ0&9W7J%J5rJK{;})9RDp<w~cURW?ub1k$LRDsJ@emPPxHTv&m#iqI(<3
zLGy2w1oA#L5{;j+x^P7pr2Y$fwF^jU{=b3+^Yar|b-UcEfAIhJwKyF;WABt%Ip4a1
zU8kGOK~<a2X|Pb>TiHVqi&J5&js#x3$2{pM_9C7-p&52PA<yyUy+O~F`@N!dpOOZ3
zO!D7{i8yxgoo6&BQ4xE1<KL13%FjzTAdA-E(wZT-7R;s?0v~t#Ju*ScYi}L;8{h9w
zXE}e~;C{;b`Z}}=j$*p%03g`J^{rTRL6Mq%UD~`KH&%gXy1?8^A5mbFcO^RP=3+A(
z)>!Gp!E{(Pq#d@YJK&|eTU)@zgB<X&gG0V4F67=2w*iF!xUF$;a2sTcolRozFK`aR
z&xbDNtG2+t2(*BR{E~h7rrw30o#!fp=B$X<Of?ma?xbN8ux=w5*gAHvr#+g@Z@!bd
z6tA%Z&L!ma&lcI2Rd4}e`K|MQ0r(|ICpo2-r>7eqrSxa<alWa4rPyTo_{Ut_*VyT)
z=1W;RLfM!Hjol*vWWiWy&+#lO7IoQQv6li@w-3)*8W|Zqe8{IO5TBTsosrSRdd~;W
zBd|c>i7B8=fnvfx8vY_+d7K(?WlyQvolh-ZFe}Y%ml6cK%0+MjIUyyLTv>>~=O@eY
z-k5?JisBD`f+R!CQFm(OXKC(;Rpuw*!L6i{y=QPO;<`*d3E&R$b@4Qo^E3sX9xl1I
zabB44Z3q8wFtGx;DO>|{kNyDW_7NtPr5o{ba^gvKW|E6u0|viT3}CCrKu@s4uvy|E
z!P^Bf?w=s32N}h)GZO&NAl-&u(1xth!_`k80AVkEXF&QrCi_|GWTDN$+ryDBrMX9y
zr;g9%_0Nnf_em_3GR{7a(mI?__l~bNHJl5bN-go^`>wXoyYMX({infkiT2fn)LR{k
zLt|~VAH<!v-~xc0V*%nOmPQ7-6%K-wgd{l{iGnyvC5UgE3KIh3egCW`LbU(rd;~Hk
zh<cQZe0+Slz(RRfBo$fJ2yn{j(uN^#emV9^xrOjsYxvXgq;Jwyq6*JMZ@Ho4iW&K5
zils+?rl<U+*HTlhL|GH&kq{z5f`hVn0EfP0h?|B4ZqY7D`q4>v0*K8ltDFIN(3TMv
zCa)qs>!}hUnK06&=+*-I&<Ye7Pd>Z?swP*TOMZV7Ha;-D*jfo2`xYL;<bYoFH!vvx
zx@WVovFWe)v;en7OaDsf70d^C{ra`ol{o-NU>HD_6&5~%Yf=wxRS+w2oqd2KLk9dM
z_gzS{Ng3Qx7Ih%S{0@98cXal<cV8R){3dS~!%Twq_I8^TAYcmw*gVfpuNx6BO2VZK
zx&u40sX8x&%UyT(irn93VbBE|movk-N$TX}B-_dX=s&&4BM(3riy_VNuAs|hHHxbo
zMrYvM->EH}1f2_Z8J13gt96071Fd2(>^bK)a~9t!U^T2~!;nT^M+rfvcgS$IDxOoZ
zSOW6fPfj!pX$WjGW)_xT>+7vzcWY7l!H<>WY*8o_q&si$*9C=y=94two(&6Q58Ih!
z|9i0g4#ax<n}Jb^j(K4K@jr5O>{&WDf9Q<Onds^1OSR<{Vxs@gGG|*8ef?NaxC95^
z(A13hC<a*+5G|iay;LBl2R0FA0>L~AlERr9kIm^y1<xJTALmLSXZZg8JG5b6zsbhF
zv3m<tc@ZD(K$ywSUcf>WziQN}2qe;8y?SMO3W({GiK>~!#rFe`USI|>D;V4i#_Ys^
zqW!(STk7i5lKhppJi5j)@1QY%#`E46s5k);k&yQ=>znRaSN}{p9Yo?d^?bj!xO@d*
zD-|q2;KB>Q^x`RJ479Zoo=$?=#dUXXZVpn+mDSb!Wt}1n7seRUBf&yNZEdnqVN=3!
z78o;a>B#?pJFg03)BBLUP|?>{KGG!avNT{D&GZ1Z=7HZs2`2<^dDhjES5lJBJY3dE
zOuB%^!?}@O4S%^CUITqt=X+w_E5s&;!a0OthkDnE0>uF<e+n=-fO<~fa$`rQ`W`}w
z4js<Tt6l9NKZAVZi>>e9-+3^68n$qnoTn@NciIPsNPX<Em~j}&QQ|Zv)`h1CL(|?t
zkONW^j-z|8;Mu@WfE0j&wVU*_{!kcph55&$ra(fNKH&q^QZ@U15VKeS<pU)lbm?}+
z%3A>6+6O_n@)1mn)OYc!ReAU6$rLLS)oW@d!?H%`Iu3CNfB0y(Utvt!!($)zWBFw&
zXx7Z1`&=}WiGe}!4`vfYG(2WFPEQ*BaUsZVjF!&9Smv=VK*D7kx29jfiI#;+#h~gU
znLLsh=7OSm;i?gg9F3!5mP0wi;Q%os-XbpMK@GmVvdQa|lz=qBKJ(-#|FaB3<UqWT
z^XBAr(E~UH9x&wt2w-hqZT+g^>q+=|BqZ<GdLQEOl0S<Pat2NSV4v01)hITF^fwI<
zVSzap6=nDY-{U=IHQ-34yipGaAJ6UMyd_>>6pW)(HUUxL%Nn)=*dNu!*9A7TBOpXH
zD|SbogJB;w3NNFgivF7+(D)P62%L3u-@LWZuaROlZx(^OeS)Q$`H<Q|p8O`oy1B8L
zNfNpiax=82sN?b)4TSf488n>ijRW(TBDN*GtaihQQsCSo?$>+zD|;*hYH?qaq4#__
zEADlos7BY4^A$~iN>E4y=l&knVkm-vB;2=gGVc@NLg#gkcB~#=X->QzpG!LE9tji5
z+TW&6&3}bi;=Qc|ZW^gO#asvjA|oW?@%4i`(dxjkWa0!&IQ#dQRJn5TM-1HzOTnJ>
z$##~Oskv?)>F3Cq)cX`sd<?9>w3xKZ2YAF4Uy_xQY7}i-*Ln)YI#3N)!27mW2lu78
zOYkl4<AYMpd_)Vo8}5Zp#T<V#&HA%Z(6`=~B)oUEWCuzJGK8S@S8matz{&VCmcI*_
zm(g`~6*I*%6?=t7o{4@%H&6_T4u@VAAd&!j%3sN!IKG;TTkrrbEDqdXLhVZJqBH{2
zV#mJ*++sqMXlpf`B>eANx0v47@(lA}C#Nm~`uScGT+4`o4-5#7*Rmi5j)&11gpRSP
zMNcUW@nvmq3U7Uf3|c$|A`le`f97*&xi1=7-bdsTKW`aTcpm)+Xsl9Zf>*ed#(d;W
zMabukSs9P@TEt$2yHFL7<|D#v<&}+cF@-G{&GnlD$Xz`P`!(e_=AZ*oCc0CZwzTfn
z-h$DEtI6%Da-L<WJIMRUm>AJ(*FrnPE{a2)7jw38&ZQd~{-;Q^mQaPipf&yYa1dtt
zjs{bNG5Md0N(6ioZJ{R$d<_E_%&3Ra0!s8HT#{}&dU{lqINEagCdGXl45A=`zK=(C
z=bS#bDJfFlrQ<AMB@nk<<r;(Jh)wnRISimO9{vvF6@}&H<(v*g4GAn5%*rYttc6+J
zSe6hWSunstVST@iLQgcedBD7n&QsJW@DL_Zi#uup_iD3qc6#(gM@J_+*}YqY=-k*a
z<$YabX-f{wR#5;xV2?|58xDDlzsUXff$Rzj3PhQ-`gIV=D(%09obXBG`B#323s_Dr
z)ygS|BX$PN0368+aNapgODZc(RR18s!j;0{u{KtoBg7mOJv}&a;|);xf|(%$mJ@9^
zA8Olk`Ew(z^YQO|h2bFwNr8_`Caa$!d`UfsJs>s=kxbQiQ2tG5YRXUvMUr%!67edS
z6j{zx7vst|nj;yk<pXt~Qa3Ij^B`w?Kc&a|kO)mn(BPB;+f}1zyz_?W_ojr%6Cx|h
z0Gvwy;^NCAUvbB<Iw2{d;3jj-H7KgA`s&&uh_vac+O3-H=r_8T*gYhPm|$QK!Q+Tr
zjsd!JxEq!x=F|M`O%PO-Iw{E?+S(FGvj3WNA^ZdbiZ?2@Knl=nV387-atM&kD6xRo
z8ZHhAvu&VbB6Z|A9UYxE6A=;1C3i0mxEU>T(6~!5drMoP%w+(e9c1Q*uvNGxU4;E0
z3PN&x!9pN_SYXXV^tlSE^E6BYWYjorG}h*0V`Bq0>!;>$o5pcjq%a5Z6L=J1RY@DR
zDp8;*e_NGAiKfM<(NUBYBE%ygJVNSVPCX1mAN+IUPpSYdw&%J?#ss&O(=_E*@jWN5
z;3y~8w^w<1dBxZOeg1MX&PdKCe+>2_8+?l6UyuqprGQq(^~_zGgz2r<*Fq3$*@U(~
z_wrJ?T*3wd$a)yoaK#bV>t2WMVT4)QJ%u8(@^ifwbQC}VB-iNSV_*nSB&P#b)&bxp
zZ(3ZOL0=)`6^b(Q&I#)v`*esQ=u2MbNr;qD49FfKPw??<)*nTE8fNo7n1OuPX&KkW
zMw;>`|GP%>(H5p@UG%8tdaN9Vbw5iVb6z~ypCKSk%iqU+^_zyv+WmDyT$~p$lO`-N
zfRV=QE^9KySAkMIGC|sp(|RBx+neD`?kexmT);}?aYLls35gFxzPuCUV-N>1LVJ|H
zsFn4*gB|J4M393M*hJ)}H#8_*t4{{$GRUTQUetN+o_n5h@r`ATJ2|J%9#;^DoI-n=
zJ|Kf*&sIR})8a358wulVsF<AbNf=<>(J>M6UItV0@KZJM!=g8Ey4f-uM~y9Sf~HPX
zwM183`M7U_d8xajLrqCdz#n_8-{CSydg<nPQ%rt-_Ci(&r*PCuW(PYGYt)%Fgs^K|
zdq}Et7{;{7!)RN$J<jR6*MEsTXR|iB&i4c8ZlI4OAHaw$0HCIK(&}>FG0A|sN_H4B
zF7CDDtp{Qr!I-O<Ybh^(!$3IpFGAW)wXxTy2=4o@w;%1$61Yqo_70H$!R-hdjW-js
z$tQRX99FMZ%J1XD-LY}O0%c`oBwkC>t707Q^?~akL*C<OdC0R7sT_&k;sUYM{pAn|
z?cS_lds-R^6F_+wr<=3znBv1<Si8}BnU*qBG_!TqF<w$rx{6G1!QcwO9Gea&%$Hwx
z8N8YV#};^RmX?+R+<E`|mbC${{h}FrI;l_)%V(gb$n*=7KQ(@guVe9p&jm(FoRFQu
zCgd^Sxu>^EMRo23>QlGhY*)q@j)u-o`?Z@VUouY4Wo{mZL<9(G^>t3Hf*>%R4k;95
z77{p!L-xiqJ;_lR{7(SPeu{y5N6}oVzfxtXjI>1bW7)pn*bGP*l)sF<aoM5mbwCo?
zIpCoMBG2`9>gfr%hrb8#d_9f#Ch&y_L)-i8VDW;-P~{~dxdW8$!vLEBG;&XL_rVUN
zR^4)M&p{SK*IO457t<yGjiBisC=royB{2N&Aage9MTQNPSKw3J;;}z;YHA(eYbX=E
z4}}4g57)ktKToPYvw83UVYmiK9Ml_M8XIv0@-VAPd?decE(j-QqEAHY+b8VpSbciC
zyUj~&+7kE;KYskk-o$y4lZlVANlM!hvf$gNO$4;m5=`2Lkk#Rq%E=NHn}3Qh8??FK
zqgM(ij^ZX#U9>z3n8ucgbKZ=+a?|C#D$gSZpiV|I;S8gt87}IKkB-8yCvp1{-Q2?n
zzqWRtfd=q)Aj*UEU-hf?xd0!9sQ3&a0iZCYpq>dd4k@mq^+sc}Gr<M!Wu|1WMyY-g
z3nexsPlm|{y~(&Uk+YM~RT~!_u;G7etj6PVO}gZRK4xAY80MwOF)FMYg;jrUk)IOo
zQGUl`A^4tIHOkCP9*LWe_+xwNm#7;$w&V)eH(Vjcb;k4SEtFJ)it0q)hhW?sq0SG+
z4n$7Kd;jOnC5`px$1_wATORIL7zNI~Yl5J{(Ik|Lcf~qO#&w=ZabRFT&t-Gj=j>#E
zz)du@8Pi=<-+a)YiYC5EM0-)pmT2v9q*=2n(QBW{T2Ew>q5*^N4S^iU>cVW;8J6TV
zh~jnk28kFlt>tlCl0cxe-^N2kYfXj!K>v3hGY|;V4<Wrl(x0~u?C?1xynne-spR^e
zjx&RGk!3X=FCt~=&Op*D_Q{~{Xh~bP<bdbRt?8b+bT=!sn45_Dt)2uirOQ0i%(zIc
zXooF&s&zXLj*iM-bj3`6k4YI)jBkF3pGG0_ZuvBLkToCC@H$Prqphu!obaTxkkxDU
zeDVVPq!}TQ07ZDcy|qCCqs2hBK&G_rxI@+F|EYvjhw38UCNiW;$fpFRci=e&#9C%(
z8Nj%m&H<JQSv9>28zNLxl+75Q#a>2tjTzv;BR}D=(Vh?xD$?a6%H7;0P*C>PSmHFK
zISLoVs&JEP2QevYlv*h4(30Va(kRkDt@OYF6LTbTbb?7NU=&aT&>mBskV=#Mx_e*3
ze?#!8gum$FkYAfQzp8emS^i`X`v1XD`;1QKD@=Q$8AmE*XliJ<I6EKAx_mf2JsqmZ
zQp`v&f`R7KkJo3I5-}wuqFv%;JkRzwtlS%PXf)e^4I0h2#Eq7akf8B!&&dLIfUBos
z9P)f9sst(q*)%^9w(@t30Gjq|WrcX7eIp6;kd~ev@}8SXPd@`DpMg1P(a#llrN=kX
z5gQPo1ldZpG_2Oyde?NscVOW`CNfr6(?fkjAC9i^0C6W@0w)bn1kb+%If>&5SL^+!
zAXx#m4trDjOy)z7Ks0xDswZ8Lgi;hlMei{eV)Xg$<bf6k4nu6Iv$;9b2pUH_hO5vS
zfCZtNBd}L|>{THH7?`K>E)43sT;n1_l5MG}tBat}OgGhkPZWqlKRqcFGjZq5TKKf%
z-H9Tq_f?|t3398XzNFr1*B>X-wxTBtFYELDnPESm>17z3`{lkl`7_^G5KPAamHE?;
z6qv<_cVZN=vIayWcSXObcTkT*LjH{xbQ8ii+Xz&@4h{w!ijEjVLI5zi&*B=G0I<xw
z-`~Ao7M(^BT%me=W>U+Wo0J>vMWHy)7yIA(!6`5C<^y<R4pMY{T)?+1Kxopr0(_UP
zE%%58Of7qeL^d9f9*B9o)x81*rV``9+b><cy<ocwNlK#IeFnuGB_-uv4$O`UWiwu^
z)1ryE#7tA8K=hcDViZzIEv+651_R!aMqMoH4YQ8{HM$NdE!2gOXxxh;IG^~K(+cWr
zcoQJg{I!%lArJ2Q>cDW_bBCa)J^e3-!1W|NOG;*{RG8SFI+pcHOM`*U_w?|=x>SOL
zv)=kCyVSdk3~Rs%tgNhXW8nt{NFyCKWFR4_|6O<o|8Smt?ruN>X(uk|9jp!uf-?{D
z*wvL44GX)U5UgPCS=A-C7P~iB9T(TvnPE5@Gqd;Jx>=EU;F<EzswHHmo<on)0nwc<
zZ?KtDol~ajDDs3^vgb%(dYX7@V2KL|@FqwDpJ>+1_`)GQ5ZV8h2xX)PNSIPdR33$r
zUkqUhH87@VE{MBs8OK5GMaQ}()_W2b6chvy+tcXPzV2>^(Gm_X|AgfCx^Exb*ff8^
zy=(EwA@L&#Jsn;1=3hVqN;dU6m52taItT+zTpwh$wYBM;0%VvuG-tz62QD$nD5Rt#
z5~;U{m9@|(EKMd?fzD>WH%)ebrX_UUEXDFAi);m87o#qG#tZDw|3Cys!LI)%eIhF!
zp>!{)D*XOvi7K*@pE4dBLcdTR$*~bEqU7~vfSv-F0U2x3;j95Lf8Ptt&_1Ec6j0Q=
zlHw4IspK*$c`mk&gCkU#Nlwc|+3q=`wq=Ig{;50-S_`ZVD!!0ZLk1z_&NL0eF<Q6d
zYb0$5id#U;o#cUa-z&UpeEgzoaye6zAv-xPCb_tDsriLHs2yz7KESXkAdi`ga7OIE
z<qiTq05P(P(BXSQcqNBkXH)Sk5*Yie%Xez|=AJaTt(;Hr!)myrIXO9D+8S<vEjUf8
z52ri9?9d*i!hrw~4`IKDvwhhCvP6*B#q?ex;$&y{H#;<|cIA2VHa;GI<UzEHT1E`I
z%$cg*5diF?waVUK(XQ!dK@N%mV10Kt%{9#FgA6c=1y&&>I9SR2$4^BKuwmT=M6=+U
z2FdYfKp)rDfa0u@uU8IJrtqZ&e!s{vAR2I55c~rzO0@<~dcdBldZ%vb)2Mn+#99O9
z0)Y?2-dpdMmmeNj0m^^hkEHx|RTY?m+?h$FiC!!C`NC1b;8c-SS%A4amP;+``b?Me
zE<nakeNfN93BlNGy*NHOX<`KYrumd+;<lQa+D@G^4T-d=JP_5DoA7*z(?n1>?9%9@
z1g^nu$N!TY4FUl;HuoFGVWHqUuGs2VICFx1C@&kT3|8J35fPv(!rDN-+rgZzp781E
zzW|llDu9^>JVUF!^c+U1687i!jZeTZbq1r^VV5P+8w-?bD_Rqp;g5Jj%ts=TkiH&X
zd<Yjk?9q|?Q$19P^8QTJ)VNa%^|~AaJ`ooeKY&!U<&Fw}Uia^U%NwF+H<Y-EzTWV+
zcXf3=<N-u=^?PBD&sF(Lmk_h9)6>Y({e>9^fAs_EJth)k;vZ#pTCj`_reycq;GPU+
z6suq$`SS>mgpx_(A(VA@g`};UgP<zA_aZ|t@m*$SCUBlzB}Urd7g;a5%15(0o{*L%
zTrp(uzL96UkNy9vsl@-)6{>5br-vPkTB;q#eQ5b0BJ<(v#i!(+lm2qWtX2YvZk3Q6
zL3y0<sdL1<9_4>k5_#Cd@45xM?{uSPvsG+&XXn5YqOSsJu(JF}%o2I$?yAxf9W8BM
zR?Yv`I3gzhmTW9orT=kqR2Q-Ao$5b@QY?DZAO7Vu{U6S>JNpS_U6|okP$AULPQT9y
zr1i}bI+MJ9sgfY^c>E+z?(gf*)HggY4F+SWcPwUqTYa<4-iLH%k3K6Sd&`HlkOD*)
zK!yrW(OO1eY!5xw$)6n<cPUJ>i@KCjN)MzAh|}MBne`1z9>Y_Sgjc@>OYK+KG8_s<
zGP%8Fw;!$qiBYkXY12knU&c$JU~99FKfscQeOf^nxK7iU&#c*Qya=hGQoCy?@C4om
zq#Dcn$el_>gGWH({`*p*3?Cy-{I+=JUlqjO_I7xQ{i3{x>k5M^8y6g~+}72ZX$V;c
zhK3C@l-la*d?F$@!8!SA<+q0<!S>VvmMuJ?(AUl@16>0!mKwpFDq0!X@1eIKo&5H#
zf=z1}v~XZ#VCYT1{)WBvfe%Q#zzU59emErL+zUB|n}F$M!RT8Xn;|eoLgd9NKU+iP
zj`PW;b=-@W1^y9$DIhD?FI;Q3(bU26&pXT$4SovFG29wJ^uZy4gBhVq7$3L(>lX-*
z+N5B5@RvqTa9)561K@hlVwOdyl0PB-+T+UJ>GW6!<(p|d48I-N>Gi}NXH{?u)3LO(
zCj|_A#ue?zg5u+f!?#?q7jS{NNb~V>adE*&ym@w$OU=2SlLlz{{b&h{=sWeR!fevH
z&zaZf)%liV^|8!QcXnYRaGK3oR+g3!0~(UhVCjLov%9-HuDuC6+x%&Qi9q5K5`k#G
zR$V)JCLEKxy!Y~nK31BN*xRJE-+=7zRC2q<v%s!I1)*|1?mY}_Rdv7H)&%$pw(6qJ
z&YGTuY+=2LS~U&w#zQ%n?kt0><1}&hx;xreRCO$7R()8vNyCETI0#wWt;O}cbA5RN
z2CtWD53|3Um6B?D>3sUepv9V5I+tRbJ;)Uo7gzfSIEUd3m0>Z546qWx08gQ%zpK?E
zGE;BlsP`6>xH#TQex#At0)zR^6;`9`6n#6~k8EPm%a8NBn!p?j*kYUp&=Q^?@DOv|
z0vRuWO_0b4pZ#^{Z0hBVEvgnZtO)qdQ#$`M;$y#n9qh>AxoHEo9dl6H{?xfDy*xkv
zElcBqu5Ld_7Opytii(Qr89jHk0r7ZCbF*r;W6s{*-h;kWTncjw3l+~rUB~f?UtraH
za#lRq;1|to>sB^#$(H-m9HI1V)5x#Ne{8cD!teRbbP6?!%$hu21a{&oR_h5!1x~Ow
zg^}0k@gdm{WNUKM;B4CqE#>syUdFRTSa8>vLHf5b)EB6nzp2>lj@0V^J1$~~%5$-y
zcQ_D(KaG;mB#ED0wlvQF9Gv;!f$8Qu|EYq<X^wBW1O!bJgtQ<U0ttNL#7d^CRMq96
zAB@`hIsdqLxQ~Z~ba#5kpQK1OfYs``gi(b!Nh}&XO;4U|0)*)SEd-!Ag^PFV&!0kX
z3RW+bN8TWs^}tF18Zfp2>iw3<hQ}`(MrN+tIjLk3VI#!$9u3jyl@|aeJ_Co#jx=gd
z_|qP+{6ERzq#p<F1?hb_7L>SsydB3htxXf#{Ro!7tqXp1sMizwn;iAO8L-W>f4+#N
z#fuvBGT+};MYZjv2VU;~&AsI1H?yKYkNNCEB>;}X8L0aASBHCAT2#zSfkwcx7BxN>
zy!AzXTC-?<*;a<N##+V~x}b;0wj#xYkHkd!Y?o}F5CubgiEwd-`U7-R(lDp`qYW@^
z-@krcJ1y?*?e+HdhMCnn0zzwHfy9(Ac4*kEjdWeQp?8Ot*4)2<4{AOj=xkEBH6p%g
zfTm{hmg_i5z>M9&UdvQ#svciD_c5rtIqm>+V;yC+s7qTZvidiG^(0JB_kUD$7$In0
zqw|Yc)3=T#-MDPwb#%f$6)b*F6O8hOtU4KKB8F(9p>^ZN`m{i~K~8bY{ZfBJ{H{fN
zv_c|g%S!Skv(ZzENvH@wVy>GmhxO}BHJkySk(Fq~j6e@9Pu{|wU~D*Wy0MsmqlNKT
zn!PPC+?}K@0NS^!2;f$+t^yDNfkGbY<_MqAOfhNjtZuBsHv9K8_k1^CD?9Em7s$lV
zFam3L5y&&@nU5pF<C~=z;<et4{?8uN>w;t};_qr~Dz}Dx$0jJIE|4zY6niVO{m6Xx
zqb=8l4VE$no2;)8=8YK_cH_pQ;k-kS6bUgg2k=vaYt9x|Q1AofzhKn`a4zJH-_hP9
zA0Hp+o&$h2mIizzeehKrE{-;NeC?r3`{*vSbnYW&3hZ9#Vh&14Pp{U)x;k@zFbA!s
zyncP>vm5ac@OJn`8c~!YA|n2^Sa0Gb9ntPDUtqL#dS&IcEU0jPfco&yJkr3rYnr|1
z)<aq&YJO<>QgGyPPY}CyangM3s-KGgedic76&a+`WwN7KG++9To&m(F1=lUmoPp-u
zW(e#*a18)itfxZpn26nvI5&%4`Dt(_t4G-N>}#A@J%{EZJDKfpF$upp%YR?>V_Dg3
zh%40j0Dx8rF;Bn;)OzlYMnef4T%yl}fJ%q1L*K=v)WLa6WhJdiljxjT)27&`OMSK3
zaN@8ZsJ;|Z;;*NMm8wZ1AiuyiIoO*6SlaLG#KTOg%DN>4@~Cn)kh|F1+nX)YbKEKb
zib^Yrf=xaZ{?~sG#F7u}?0iZ`#WwpJ4oC&6A3rhNd1FD`PxxTF8rO5OVW1jXCfKCp
zE8Prs)tgdI-b?Xlk<z9J{^w5k<?8G_ojxN5<_Q=M-r3v)Kv#dNbo>c)3lIeTt^GDe
zv*<<pd=fK#vKj5Yr5dLnon;4i@Peblw?RxjFdCxXnR9qgSR**Iv?!>kx}^Hqjr~eB
zihOYdz<MTN5=tghd7a6JkB`4N_zK$o!*$cz8l(XR4W-uo?Kw5D`amdCe3%&9xxa~r
zua&WMiHYR`IOCburxg>bYxEOMdyeEjXFuxt<r;Y2>SY$q#}RT@Jr*e}ZGmlR0H!1e
zchIy#?R~Sy=A>1bEitJWC1o@aFVRd_NzQ}EH1ZM9QL`X5*lGj54i22zk3Yt7=m2{O
zZ6(t1{EyymZJG2Agc5IdK=ML?2)L?bUptM}t=g!S$TBLYfsch`hj1Pv<1EV0Zx3n<
zNP6Fj*=br@GRGf+8w;YDxIG}6vCLN0uEAZf`F}ylIP-$)5}2xUDG?cJ2$>5b??8m=
zJwnNIR2^(PfNc_>IX&%6-z=E7iWN!zsnuFP+jQvNcr|~&t+xFIvk`(qz&yW0X<UkA
zA|}(*`p{z*6xz*|V-i@m{2+7cG;d{mTYdF&0v@I%&4R59YLj1SK95)_D?uUxW{C_B
z03SBj*1+*DK|yyubJK~a!S=O6(i2*(tChSquJ+I%0i!G}HU2_qNxQeZG-<(i-qa*8
zCie!l*^YTA=iX;E3ebi@{sN)!>%hP=aNFQ$Xs8;#<XSSzJi(~l_lNWF{-FsM|GkTx
z^3mAE6No}n1zSbtYEo?YYP<%qe74Z=@2Yn(<FCNP{#4KCZ%j^TDEK_Z0k?G+&IeYa
z^Fy1d@hsAa;^(OHL0-&a#==`4t&~w}Hj;<f?-%&_-q)u)%WDF2(_Nz7-_q3+>mRk@
zJ=qDt+xWqSID}!x-~yHLJ}7~a%t4{pTEtDdzla^t5P%+t<zl6})XX<&e!6{)2meMH
z4s5I3n2phk%D>2XKeU?nHM}$bao+FM(cWx;o`WM*9ub^E0a64bEgfB7in!W4BalX6
zZH=x>&}8As4MS);#r?&xN8uo1by;sRl@UJg;EN3Hty0-KllPw#xktN`P-oe8e&bRO
z?f2#}=1)Zh1ROxrfYm76F8l$SN+_%V4txLM!#Ic*po0Y@<<iNVRNJfM!LCeO(XWhN
zjjR5B;HF*gOw}F$P`&yhoD(?Ug4BpF_aJXf>{i;UoI-0D9$i)qjKur;@J6i;)Bp;I
zrnJJsLI6PC=jDw7MXAWKn=XqjZ;8m>D)>b)7qBCs|5znqY`4v!?46jOn^Y+`mXm9z
zr$K#E<F>+?((rIFGf~h)fEr<Se#NCoHQP*!n*a5t&R(5bXH}j(90er5Q4k~_hrLQ$
zc0@P<ZYO{L;2&KsFaP-I6O_X-XxbT!j?i#dEFWiL*ACbty(W3eA_lG;_B4TpuE?VD
zR=5;HjOKogi3YQ_5GKICTux)`Zm@qw#rJ0!6DKs%oqiNWrlp(`5()h;G3(HP!nEvx
zmh<PGy4?D0&n}h4v9{%8<MyW3)>dft1X7C%#}Sc@1`;wd+rB&coD}T~gN2U-(P;4A
zLM5Yi$M4U5T4B4YmvH@Fp@`$QfAZuq6%qH~?sOCWJhDxiDDB!RN9na71B0G3Y5(<K
zs`E7O?%gIw076&HVFjotjwa765?6nx=(BPmj|rdoidse4N&6S&qpA2ak$2eLOhrUA
z`5St{m3limIwm`S6{fEV+L_^}9qvRfUz1=r@&?tuZZ_DCJ3IUOutrQsbRK^%7~_co
z`P<CKMp4*p13rY(pC=>>kF}isFi#l@7ijHi_nh1wg?wSN`$$e~9uw!f1vnv?eqy#c
zfpPc+co(2N(8_RaZq9J=Ns6oosH&bpXOF5LDXiaOFq6<0a()?UX)rdRhcqeN^jZTP
z6YUT6jzIKyvSBO$d7|YFH8r%>?Y>O|352ZldFvhd70P`2*GkbSMK^bDZ-I(?8JzC5
zV6R9*1AzzS2i(AoN<eE;Mor&?-0+fiGW%RMP9RX@Iea9U>JZ$!wylBiRS6;E7ZUOY
zXb7nBmTn-8MnEbdB&2QhoQ@+aMTN{b1f6%-464sGlqR7No3``MujgqNNd~Wc@JGjy
zkdWXh7_=_uKctBA;F<Tr`Ta)wdA^1I1;LqM4;uX$(oK}o1}WT>qzLfk>Gk=9LNO!n
zwWpo5hSLYC`cc_Ps`Egos}*Q}xLuf<mX`7x{y_3q)ISB1Q)4c40QuLSfX4e&?lg7_
zx=+~&u@yK!;>m;`$r$8ufL(BLawbQE=IGw>N2noOr!JTdG85!TL4!98F52zLbcgO7
z)m9^EKyB2H9OKPGznk4Zp}szzN7~rfKrboKKs3Dp%N9*9wR0%D?I!`a7T~Oxu>mmW
z4I<Dh7oqPClWJgZ2e_HZm?Xl&lL0!e@i}tGnFS<uu`f;Vmx-be7zAS5sLx~{o9JyT
zKFEkM;O}|xu36KR{RV~WvxRP2_aQ98x`pkGA@6(4aBKs=3fN26^uHZP)M>K4RokQ#
z9wV*nV3{?iF;?v^e@;h&`pPEVX+a@oBRN(SrgpdP`6qzz*44H!kjOccX1#wOIQ~96
zyL7^KL$0~_9KGb#tIz{O?wHa$VNVdr1gg9LjDi2WE5&8e0#5hMm}3*-`|tT!;4dyL
z6?^ZDt=}5VDZpdBT|7h(U;3|FY>TCqvjc=t8(Kb`09M|s4XngZfY%K~C~n$*6>eA=
z4&M^`yCs;vd0lx*xoVN0u~V8sIcnG|Sv}fQl^p@@W7-jjQnJrhTYZnGKytqH*3XsH
zzXh*t^tyxp8>l^K6XoMb4(|2hpR0q?r$yC-@cWIHGR^dt+efBGebTj>j}Ch;WckoC
z)7r>&4g(~8j*_)E3qi&p>bW!TN4I?|9UAVva&@5st38UNW$h!k)vV4du3W-NTX7hO
zDqOn9zp@>hR7=d)v3gxM`$&odF&u~noN;BO_-##%my3&w0{?3<y9n*_ny@f9t>w{G
zKp)mrRfQ$mj});41!zBqXxV#953u<TVCBZdKqLDRS~|E>yeCqfF8p3<bhtDqShIk>
z@tDu2SkI++lnYV)+3yUCLIRWrDyWRSJkOOO9^SK8cYUE@QDh{&hOTbt8UW+v*ZA`z
zz-g8TT`y|S{;J_}Jb$Ajd#1zUr&72jXy#k&W%$rRSIAo#kEBO|+dOLcMR&$H+@0_W
zl9H0(?B4nFhmhh{TqdNMJ*cb6d4UShsH1V6#=os#6vr0O>K+;?pw@fMK_UZvD{^yk
z=xJyOQXnc&;mgCnYQM0h1EzC3zgzEPsG780byr&M=YMJPJ2&GW_wzo@$89FpErnQW
z4GP%oo|5mOEY+TC?8Fi8g+=IeC4_<%@S8~Z2apC%dq4~;RK{AKz8ID@30wEqs20}{
zIOm+|WbBb!BM>$Pf~BX7o--^9Z?t>rr`bDJvq@C{(kWxr*fF8Z3B~?~z?zFoNH7DR
zSyMqQ-VX_%xIOzEIFo-sZVpZl==_j&G7E1Db>k$MQgQLmC6xQTG;of=s6vm*Hu<yK
zE5#JhgyeDUIh7R7in*H<5yZY}pgDq`6YlPk@u0JQcCcl=$1!xdz$h!L<gS~G3#e&v
zD{j*w7z4%#a=va+DwmMR2MKko(i&zmjbMvwdDo;R<{(z{S(T`#JfjHJ;UIM8(66eR
zsn|0P9jK#MAcq7*Cc!<~JliFv@;hFvJub%sNJt<B1LaJTCHENgJcJelv1qZ?B^+0b
zc#{uW!$%6+3d+4scx_M;>jFm1W-VXrV@m}$=q|;Rl>ad~i#UV5ouHJv>t0vobFk}C
zdhQgkMxGl$uYUD+HGv*rN>nx}uH5%P(E9%UedxhzT8~Bpz9!<TNB0Y*v{%J-+8So=
zLiJHRd3{|S3l@+JVr8q3JhwkBQEq_R3GnuHro1euNhToT+A)KEgjZPjmv}VK)N0BP
z;=4Hzfsnle)BhfL^Ja)5J5^5>Uv*vj$?@S+XP&NI-o_n{2IdD!6k<P&Wlpt<Nb)Zt
zw9buE<-fy^6;h<=zRRZI`|+$8Qtz)2^57>UKp=X2(*Q-s<^j^u)v01tB0oC9Q6eK;
zmz=E<mew+<i19YL*}j4CJk&c~g5rtvffql4VGq;cFt~M~BnIP<AxGg(R9EN`S}y=}
zlIEtSou;SrUGY@QxCT9!LztPj{kIMvf#9MD6dW$Tdv{&k@!%y8lneDN(aba#tp4WO
zUR$}=_AMTb;VRCn*n{qAg+HRY;+JgQHzv9^#0~3N6?_Yz4GP@%v}c+>Uf!6Ww)HTm
zAD`nwfJY`SV6Bsy=;UTJs}FE=aF;@=Hrd-tW_i=K#%Z3ND~?Y1vQK|ys>#x~Bs2#4
zvwW{mCDeVROwpYQg{D}|A76c_ZC;&Xdtg9+1M_K}$p~meN{^WN3wW0!mgri7A7T-X
z`LR909hWU}{@1fg&^5R7@+;TXUa;nda0ovNuaYx2W-0G}OHWQh@)xwOAjhr-+5X|F
z$o%w=ocWZ>PNhGy0Q$|&Lhq61C{$2rXwSAZGwmGmX6(ZOADbe2v*T<!gQpaMBDL?p
zy4pQOK`PDnF6;Vb8PLsS%X);Lwb$~5$^9mg&W&Q8F9%mnF7uCX>7z`l-9v1hN&~~{
z`PEm!ZTBJl-tEKP%=Gkh@Qz0uSamklaeV7T8q{O!83hDv$)+9f$G-*uuJ~>PEJ1r)
z8wa9#=Rp0UIZ5%2E*ULVic|62n}%Ns;7&1bV=j-AW?*CU1E!-p<vbVSjO3j!u_AVX
zni`<~fP~>D^lF0Fg^n}8EWQJoDezm06lK1Ht(wd6$FwW~%(iwc7DLaA&x-5Ta6Zf~
z|433Vkf7ER9ldd74|MbE`WG(DJ&3M5x^dy&I+Is;9kS(oYoI?NtyjjF1&wMfFQ<Sy
zH9>IQf}dTzU;39iiG>Zj>eT(2nUtThHTx4OPy4+3*4MRDyJN@<`{kqn`UH+=WvM0%
zQAx<Jc66SSj+2#!psBk%A9mfW#y0?>A$-zq7QMNSHJ<NDh>HV&`3Q-`xCejx_6>Nd
z+qrrffR;nPcI>UJ+zcfL7%TwG#`X+k+(Ej2e%^TFMm!hW^`lsn9$BSm_9v~szf2RG
z7~ZFQIzy}n`9v%F?;xG(<C5pIP0cN6h>^kuSwvIZuUdL~K0uXlu*bZKv(3!R406|m
zxVSQC>_(Ma(FowdIuzKTvW0$)P>fk{+|4tBv-x-cqXfMLpp({TXr42H5@o+*+@Zs;
z@(#7XEvY1z8}N==Q}C7BHtTFFy1D~U_<W<hGEni%DqVj_>h|ZTycHcBhJe3L!b+fG
zp9<DAnLh0Qb;w#)yAr0b?;er_H#axq+2$xEbP~Yt*GU0xSQ%&v_N@Iyo%)TZ?^q9;
z(J&1tNo3{iF%$JRM)KCp{DCar%f{!B$<G&?pgj%z{_GNuATx*M%6Wm=-S-Yh^xw#n
z@g)8{OgF0}y^91`G5kDFl8{-Zi!Y#;aH6UYSHW)t#at2_XWLU(*YhSgDG_<@P%2OP
z?5D}xdpZe8;PKzJVu<UTW965oszj$exBhrNd%MfFcmD6k@TgYw*tF9(gg`A#p}=xT
z-UC)S&z&qYI)6fH%GKL)XRq>acZ4BGsVVWW-Fj2Upm>1D_qnF#$!G~~?YTu)H@Ds|
z3z%}?0e4lLCxDe~qlIu2S+o76ukp=#7uR|vYI_!otwu%N0g5|2+YU0+_jjq`@+NL8
z))6)>!7<^{z=`hQPu$Bsn~i6B6O(fU<(2>6cw4e?CQ-FiGEWQZCqGIr@$b-Fex<ks
z1`vnTgz920)?S{Cwmc>j6(yYr>`!Z;c;UaL+Cg)`ry7BgBdvn69BdP%_d?y}J3~!l
zDFdJCgC00GEsdo(B=YhH;gvt}l=T{CCFbVCXywOcwB3kZyoOh7WzF%i0^Nvkgk5g0
z?qse*73_Zz*7@syfFnIPIAA~goBPMsIUImG<?WeLF%eaB4=^3M5OzL!{CMNnuP+oJ
z#5f#((vUH|x@ryzi#vsI5`aGVa+&qFXGXrO9>d%R^&x5xYRdc3j>Yn|(VpDO8gj{A
zS=rgaAt950uOV4l(VuBAuNLMH$;0OFM872qL&$(>KseVK%sXHLM@p<NECfeJ52wgH
zbAUwF(7-_P6V{KtIc4&|>6x{i-NMXFKv_+VY+pxf>pUr?jQsIdpvB$_@iwN<MfSGc
zuQ*>mZFuNKCuledZ2~=~LSLao)mj@&-bm-H-L?or*ma_z@c-RO45V4#!cXW`Hx5aP
zm|eHT;|T^kcUL&1p!>9iMhf%LKjKYtGH^OhQ}t<gL3A>OQNmB{2xc<NnnXGKg_W%;
zamZ<Oywry0O++72p*DRoxmkNzeSLi~wNN$!4xV;&4cUWz>&-OflS7eNrIw{8xH4$F
zh-X7R4oO!~q8oKThImfyq;6OwqeLq?%mTaJ8!&Z9mTe3Jn1P`BACA5Wo3t%;h*rSL
zfpi-~<|zu{N9xkUCk9VYa#H(AHj%gJ^*C9N2?#FbuTv03Q6{#c&1N7W!(zw2^)FCV
zeC4tyg&3_z9IOc(lypMst>}SW60ff+lj=jYt!TF3EpFpa=;qf5&DkdH?fEXyU&hD8
zP|Yfx@QKT*^veEw?6Id_p3xn^8{ubx!tLbkT@uRv)kpH^nI@fpZuuqJ^Ih6<2c==&
zAM!obr;qBXDf7iAQx|rXK{)>4!GlWA-TPpjdVN{o6Do0^R8Re1^n>abz?=L11b`KP
zx#bQP45($!^`bQsKY!<F+7C8*q$PJi9Ck~8<k^MAsoPYPVgNL9b}Su8+%+QEo@Qpf
z^e=`xK|2Yk+95hEFE4j`4uHHF1jqiKTeeyGUDQf4Dd7kO@?wMGKi7j+uF<rj8{wgS
z3k2|Gf>Gph1T44Sa<L{zQ-%Zuv1ipn@aqONct(y0zx5D0OsXt273B~tUD8~!<z+p#
zt@SN_{9Kh`GfiM=Y4-0-6I|L5G}g>w!w?F#vmHSniXp0yQkLz<F-(PfZb3pUb$K#j
zAO<DMH`8))Vw3iKH|!UB?)Q0b*nAcGbTbkJ=~qJ$x@j_rEAFDLXoD9cr$NqFUZYA%
zwy>HIGV-yv=O~NrHbCmK{*KO6P*F~5@a~n(BxM75t=+KYgZbiU(w16U5;FsiZH%?x
zZe<Q3i-_nAIliZ?1m+b^V;}9mf)Z?5VXg_J4#{ucfI;aF?@NpYt9N{nq{l74bk0W~
z&QmU0)zT>a?9yUz*5}}}yTtnMqoS8+SZ6d@u-mTJa|_xR!TklU<;sc*$tzd#46;eI
zu{{tWojBKpzKEA_BEpxIl!DT`G+Sqw>y=73oiJp3Gx+a3_7cb~$apHt!wMrtjn+vy
zSS7q!fB*TjumFJvN`-y689)dNH8AwSdZ&HAb2$|;nD8mv+bR&Hw=^4o+k{58c#trL
z=jg)uy4N1b!TPNl#Z^oD!|{Zgl8J$GHoXJ`R`xFd@PlL~+JlsgY#k~^XoLhM>pl>N
z+J0>S8ACEM2K_uBowoOe&LH43$$^3hdJFa6TtS8Ez@->=iEsGivVe>iF@K<C>(@(^
zr>94oPs`fDg&&4^CNfFj*S4q#$G9IcQfM&(@xas!TEl`D#{#<D*4rJcDP`#S<6=!m
zoGK50fLQ&P%o#Ksbzu#vc`_vrCu>{cTPP))q{BBN8#&)&Q%bqfqXHYR82OM<_<X^1
z<!si)!h^q6G2GENvzO2zg}n_{h6VMyat{yWI;p6rL_u^0b1#zLfN3Rc9d;MnGsNY~
z#odX5;vDelSXInzr64@0k9Wr*OZJ3gQBYtcfCaRE{VHT*jr9H?uvyTv@H8Cphd<pQ
zn{_GYp(n@&di=J-5abpq4}Ka6RcpzanxbNs+#s`C1H=xlW~2usMppISp0{ge9guL`
z!1r}sp#Zu%=H%p5DLXqmgI@SRkO7*BF<f<^X>f;TvVrW&p@^jk1MT|!oKS7p*<lE+
zWLM+C`LM0OKJqV2s60hwWSA<dfz5$KbRKSvV2^~SZu<_j+sJ@vr6}OH{og<$g!BPW
zuKx><`lYtpUx77&HcKvqU|faQ^{sLmE5nM8gF%x?CGtZ-jg^dqMBq+sRUT0laiq-3
zy)Bt5T|an!?d{HU*Wu_BOqj$)ZkLktv2DNqn=Mn79o)*j^bWN=F*<7cGG1o?8RWj}
z6V>InylHHH!kqv|ryVAl!0JLaaj#9GTcPLzC7l6lWOVdYt>@G7`(G5bVKI8X@L;l_
zR^L)_UIvn`jTR3dU*Tw7{8SKb$g9vjm<HMkz-PS&NfLMD!DmqTKNzO(L_k?Tcf<dV
zk5dN=nX&Iv1HZ3Ad~69uazSf*kCnfTdwX@<=|2)#aTrCML1G?;c-s0GXW_C51Q>cs
zfy8dGP>?f<WkD|x47Zop5uBq33aI9p)zyfPQL=umsSv5ZKC}+P03l-?%Kim9HfG{f
z6ea~=NgXY?T)aWpSAY`ToXh-mi)h38uNU!4l%B)(?bCM!8v?4_OO&*9s;su`ti9wF
z=lDDRLa3#~E-%Qp>LJ%i?gSOL5s~t7t}K9C&1Ild3)lg9I!Kfn`@Eh%-@zRMH}*?L
z@pB5;n-us(a#GR@Q%aQm2>c61iF?E=&%Np2p!6tbACVdTV4+H{?Lr7hAX!z(ECIiY
z<!p3<E-#Z^3^Mdjyc&<1=!e{%UE_NSa7~7T;?G$~``DS_Co|i73Sgp>N4ZT~gwO7m
zYuKx4-eU>Esd1hG4=L#jSotkmXHYyqhH-BJS~o+D-+}|3H$h@r<RXotVzVnBai8?u
zdL<Z9UHSVjljJtWw{?75-M;27DjCDc`i++I@n{KFaABbu78Mn8(WDdmo$i(lYzAUJ
z)HGdWf5&e`o@)W<$2o}%7qLx7POs?6_V)W=lpmuSP~!kYPRGRFR|dIgxn$1V9@!kV
z)qS(`+nCAL@gvo6+q1V)wA<`2U#yaZwj|c|4^`#PM85#F{B7QG4fQW*SvEnhw88z?
zmQXj)^b2d-Ly9AQ+s{Z{?%`<nI&I#nyXHVf1r=p?Jc5*JrxPt4MSO|#&(Dxn^mDrY
z-Mpb8o}Xvg=`x?$ccy_!1DMY!Q#Bm6Wd^hh){#r56_GPpaegkn&AyF1-I@M5l7E?k
z=<ifZ6iV-GHZ9yBfaLOxwRh#DoRsGX2qw;R2p)zba2sCb7t4i1CP^NI=s4u3o9qgm
zxjWnBO}***8zCTfqU{>bUp1F09f+c;qW$3D8ia~L4N##(%`RJPoTu#8tQ9Oax-GKO
z!I?$w`!DLSwSHKwzCd}Nk_nGe=K9b=*5%ozKu*>`PsdkoBYMPn@YhQcrKR+ZTN3~K
zsFFkd8C%=akF4;mBRmu~TDpn-5L!2aTd@%{_K9)gKK9e-P^`7|uN1^>A?@RweAh}@
z<n%^jTn8RgA=aR2Z1Fj3O%GB+xJH8!mktE!Cb}x=xx*0)>~|i<<*zDP|HS|K0T+7_
zKDBxM14W5lqozfrijdy&<MC*JlpgQwb)DP8a@;~SIrjv(StXq8e2ew`AmfTbl^}Pg
zH;zqJLQs0?vo~^b*FF<}wn8Rh--A)UGTHF6rpFIuIa3CdGlXU+Jisnp2UzEOS^%@Z
z1MC8{d4G!dC1GKmw9<<|KrafFYj8ZMf}w5F00*V-0zXRrOTY0Pycy4(UfE|YEzO%F
zG%~-P{;fbKVoK2My>Kh~Tt8Ob1of1>y!@RzcfkG+%?^iixBmWdgci_%OoM|o2OQ-9
zf!$XqgkrR656%a+DRFQxVg5=G_NYo85pvw6*X3bNc!S#BROn?(UoO6geO;8D<YfHm
zO*J}g?2D8DR19q113S^z71Z!XMNm^0tb>)K`|Dy@1g#*}eQ-(cV9mH~{Y}L2qYJZ2
zj1h>+?~7HULuwhl<r;90*H##fzBv^zzE;V}Dnuq{3inMqK4-(6#kOlC!xbB_=UVba
z&<QgEVN$dOPNMTXnt)R+rNP_E^EsOu$X{dH#>lmK3FgYC(?Fsz?^5ND8kqi+z2mE7
z^5zpfL5`9n5a$l`Bd`gD>*Zrb#rNE=@b1t~c&i2aaf2yOr+~YY4*Gc~51ULcBD(}Q
z3(P+W08z;=zW#ix0=-m|b$k<;3`<MPxX4&zd?x5OORQmatM+K)+tEn)Ti*+AkDZ(f
z+BH&+WM-S9mA|McJ9A<;&SU-e+9`0EmY2beSyxwgGZ;{|-nR;G6B9M<AcbJN{=Aif
zmgEK$---!iwz_aTI5{cbz_u07pQHN4M$&WUL(9*WP>!<K`~!K}^;@Ky^3M)Zca+d0
zNzXT$(WqIa*`|r+rlzLh;lxBx+JA+%fG<e^IAE4?W)}>F<hJ_Wn-TPdP%=g!%U-P~
za_`S&ay*fDx)#HW{uK;$p-4K$h_W_U@!Rjc_}l%Q@C!e(f37+(+}8&vK>z}>njsLw
zD^amOf=mtgAD8nsXD~_uWEE<Mpxp&E89)(yaNSgc&%B@+9Ny4>8~vfI!EeLUd6B&B
zn<I(zM<~^f+YEERLH712iuA3hFdXjbPlOROM}cr|wzz_hp@Af@t-?Y=NF{dFq%}}i
z00sAA*zB`g9&}Pn6fmwo-qDiO;1v+a?`$ZATd;PEo%2}pC<sGfCP5F$0~O_GvBA58
z$GKEzv8gxZ9&Y^@>e7<BctcKI?!YAsq{-3I{il`wNAfd$)$>hU)S5xgV)o?!LQcWj
zzGFm4XmE10ixC{3V&Uuk%TB@+|EiTCj#qL-b%~GPvOoez_heG|hSfC=nPBX<_B&Xq
zRW=ddzy^~)E$IHq{BGHtoK7AVZdM`GtZ+k4fu}m^b*kV~F6^Q*x@ycEV}oJ|`}TCR
zY&Y>K$zdlPBC0kEHcWx59+78!dE3A95c$X0iJX%aK1Eb6*C=nMexp7~Uxk{IPH(#c
z!7sSk?^!{+jcIc1c``!tVe?<iV_<~<Jp4P1T@dO6(jru!ZUS7kXRUdVHk4QN)@7Tb
zLa~o-+(N$Hjd-M_X$FOxDW>fysiApI)`MN9tN%W*%&U9ln%izJ_Vx4}SiMv7C7<5X
zjbmH_RKjgLO5vnLGImM}Qni_HNiPGs!%th$t!P5%#5QFZwWHsubAMI|pK9|T0JCLz
zL(Y`wlilWpg~k{E{+K^vP~xg(@Y<C!M^0Evs{W%kguw3zBX}{%N79UWtG?#f$|<mm
z$+OV3+iwg-8Y7fV4@GCKyt2z&H58bOU`dRAxc*7ZZ@Yrs_;`)zvueyJO>RO)FN8qu
z$+VIDh@+wXP(E0M=nB<(<FSoPu#UaLpnntxZ5vFA(s?bx4}12j%&l9u0B_$Mue>}}
zeJ8E-+a5Fk0r!9$C>a6vi(Y6lVHCGL|8;(p7q$fN9mU%2`wb;)rDfAQ(?IGY{Y6z6
zxVDi|R0*qbHA7P?v&0<RNbBjQNye@1ZAMVhnB4Wrwz7tF4E(kLJAfneLusiCOjD4#
zr$T~2gu=KKFv5DQ3=xj_3Q-mvt4U8-aP&(;l_a+vc){2m5%h&|j*>76VeKYb2SK}$
z%sLWER1Vf&>X*?(r<b2i`@&tEjsOiNNE!i2*1M&pC4TiP2*W<#OoPh<K&)4ZSVqok
z&xr;=enZHO_SQJLxL`omZh>o;w&#|PfdRS6%&K6DB*ivHDS|#dD}O`a_}N{vX_PZq
zB|x_Hz>eV_WM<ej^ljxcar!hkx1L-7i@mRm$|_pdm2L?M>6VsG=`IQBmhNupMrov^
zQ|a#RZjkQoP6_X1pL6fN<LrILIKS_>{(*tI_`Yw=HRt<2?-QV?(DS~h6aa?U-*a*{
z+k$bRv(zgLfRpBt9T*jX&s9Pnj8?2fbYYp6K?&+G7T|cRm#~YB8w}7$&~o>yHiHit
z92Nr_WhCjmU*8>+46b}k@_fGoUNuZ89MBxXPHk4u$;KYPNyaA%L``g~l@ww-KmA>U
ziXifyCQ?9!&pk)_dAgkg`0Si^(~9^SEG;c<EwuqYxoJUyRRt*B02-1^d~-Pa!hYRA
z1k)t3Z<tb&lAl8WmB_GS4=@7(54opkoq~Vxiq7BPKbnC9Qy)-E#0}FD|Fp@)(t)!F
zDm$^n4WvGV>c$!&gL)f?BL<Xeqmzt^7{)$~t`^zrjxq}z4fXT{#e>#FKNf6@W7Rw0
zm;%LSLkYX}!b?B(`t@svnFMf{0~wN({qFAhK$uOg_vU@+fnW7x)8w}!NzSJaC_Vr(
zRGZKEd48Lxr73cIyWkT~Ku%YTI&wwM{phuyd@-L<XGBNzJXqpcf8(1dK1+?s@`bt<
zoZp~sxa<N*`-_YN50^9iwN?LXK;P<ca|E{?ln*<<EoMuB2<?XoKe$UhT4de5y?AP+
zgqJ+JwKk@cxl#bpOek>!#W6sZ{ZcoJK)VbangnY%M)YbeHmbQ{sEMtnexk_A6C|6k
zV}eoUH1H0uXjrbd7vbF@7nufg9D)4?T5_rG)vCI}af*>P5Box`iXV?pm0#rl?$fb8
zNxt^k&FD^<J|)pQKfC?yD<#1v->(71sMl&`ZeF1LY)eGAAr8pg5j+6-3!S-hL=^V{
z2j}j03t*8pn84rxX7Ni=VH@e^=P#SdeE>Mjz|&mw>ii|N=ku51qtjDc6O#m$kH_|k
zZ{MAN)-4A|J6NP*wHu@*)OxMW!xGm31bJ_Y03%dhQ*(WG_W8!x$_lXk!Y{x&M<w8D
zKX@6Kg7)YjoRj{=J;7TMoca5cdEmGscJ2+F1YL#EHANiQ8v(Erb#`@C3|I%qXxjh{
z0oYRHKYp$d^5S;|jBEgz7W}zyRsdWD@Gw?hfd22$I1_X%<wEJa*{?SyK|Kx}VMxn{
z51zsK{drjKuQ+JB7r<SRD$NH#JeX4((dhyoe$a`|O7&6$$Qm#(z>h$C2{=z|a%3yt
zOJX|r+HIHwnegai1b&V80?>m_rI6xT2W%MhTD8TgH82}OlJno-)O;m%Yd64rXB5Cf
zqfbcQK#>&FI{@UJLnDJLtzKcnkzy&l@|g1YzPP}9;tUY|@NnKyP^@)Y>1LbGo&aZo
zrM3z{V|HYLbpY)y>X2$Ppk;jE<{oH(AYB1YBSv)TO-)TjMMb>>;29s}H41|0M0O`h
zrU6P7P&jqBArT5K#Nq%eAz;#?Fp5I$=-LPj{Ef8msJ3}YGr;r`<ZSx=H($1hfUTkc
z;8OzaHZaQhp^^q#SpfFPY_?<pcpa>9z=QZkRVM8RaF&flu&h#lTSfsA0hj;+!X+RN
zG6>M5pJHNSQlz<o!H0vrJvRvUQVsJ$@b~@M1T6V`@5{@}!N!%r?I=EJ95!xd#=}OE
z%v{8?_Xi+H70Xg?-@oCbwJGu-KtV#10Isa-&kR8A$l3>3k?rGbk~c6PP|<b+PdlWM
z4HzMgCx8#hl-=<IAKz2=*Ee@9ii(OGm+R~6U=Qc^`cpj}rkzDTUI_AEG*sL`xYHH3
zW=9h_DA<b`=G|hOq5PR`qg8^owv6uRjuQ#MY%UTBP?Z?W4?7Uru`<J|m_cwRT6OH4
zbtNmoq%}cw4yT|R7*!A)^sR8>6m0`ip}p?IM`);5;T6C!5s>&^iZ-ArEjwSt1v))2
zy?Al;RxMOE2R-u97FeACn0Vn*fKOPh;FB27NQk&fz5>LmE^rJ2Lof@qS1IaRTDM^G
z0>%L^_TnDf#LGIlk}-gcTyd*cKlka*u6i+=RHXHOCk3qdZ(oyMGH#3pVu68$@n9T!
zhJf=yTtmY>m<o<RgIY4yCc>318EBsI&_?=aTQGA#8=of~@9zdy`34mrK3Q5>4eOBr
zDn+~`ubVOez(PY>hpyrs_g28!3WNydFD|_{%gJ!~pqdZfSQiF{R$zr=!#f6C`n>}@
zAo!1(Ku=l=atQ0pLB|EI8;nZ)`g%<(H2Z^HXek&M&|GQrSnCt|{oTXl;yXrsi7M&@
z6zE3a>;l%hF%^1tsGi4gaP9?s;NdX7U><3JOAvru@Nsd0{nYi|govj%kSc%^UV+Jt
zA!3`_89qrM8K(5rYY4&;LK%{MCU$l)HEcRn*=C|UKp^NY<CCtV>H#JnYmz@W$90W?
zOI$s8)Iq5m(2&&$4y-|P_q$18^0MIoKbA<(`j(5E`z5_hEXj^uyMjAACuhA^<EO7L
zVp>vctp5eE&m*(lnxHD2S{8908mhk#9~x24Bk1$4VqD&`dOy0^pzjI2h1?b5qYq#O
zSq^MLJe$8|>JI_>X09wa6OOqCjt)HDIUw(-m7=qPODEV*5{<-pO$*GYi~}9bWRI;N
z(~$T=R>#+bB&nSb*FDg+Ny$yVVkE;i1iPFKlUvubm^6nrWhPizVu{{?x@rME6dB_7
z4lEJoC+Jo4w)CT&9azpFO!c%1HtEFoI&=46#0di1;$ceny1}B8AthQOY4oLJ*~84w
zyOF~MM<so-e;KSAm^wt9XA)t5^gmE7)yA~|{s%W)V}PTWi7;-&j4J#WeSdv6D_nsX
zM_QCktJccz&1dt<IT437^6q2m%rthlXqw>MSB#r=Is4PNGL`MgY|R}Z46}a)BE%Li
zC^ur(fnRf2!8Xva3SGehYtWq+1J^-tm<H9Fg&e{|-H~2JyECZMX|1Dh0dF9^aq+r(
zkUOx%#VLNA*V9ABtZ55Q#9g5zWP0UK9D%y$IwYj{6vNfYu&BaKBt+D)gk*GaX8TN`
z(?F$rB2Hqhx+%QC8~TWZ5~fGBTq;XnxzS|_ZwMOi#E)M2xR@9)U|?b-zL7cs491^#
zf_ZyZS2N;cx+<+Fy5I6R?EzNgy&zmAY`K8;o9dtj##d4!FVY4yRI)6j`abblAcYpu
zhVYXBU8M#EGBOcw5=?tJi&4xt_IS3FmY<+=R{N5(NJS4GT`(-_OQhLRCJ>_j_$?0Q
z4&55}ftoE1RHv%!HSTs`xIC^erN$*{)I*4wN~>+jA+yzCE&K$$(dLsTfqHLr6tI@S
zOePzK>}Z%T8u?-Z6?%(_Gh*ZyHlyz2l)S*Xh}{N&G!z7Jsua+GSVjzkA;=AQErBc(
zSb^!ASLHJ=O*}O8S%-MNXm0>ejw-ycaL6^doc5wIH!4v0?@D=?>VT`*ZZMoZd>A(B
z!daS@GZI^(TRD2fk9fH89ic#;ay$<Y4{gc`s2Q-+ZfbRl)Pve>sPF=8YvaGdUxSAz
z7%kF=Vr4_9_LY~Q)|&Vnsnd&xE1q+=;T?RNBC!>`Y0y^UtOgjP-qc6Pz0#vv$wbkK
z5*5@28<&LDe&!K4=vawFjm?n0RDnUmJg?R;Qf6Z&(>dCE;`*U<xSk1;anKu6*nE<N
zuMK*G*4L;(nU43uIpEEGG!q0A);+<SB5yC2cc486PouYb%NJ*k=>!_IAD}#ofy;lK
zfP%;5?T@t#fOzE4`-O=d8ad_G+yq>4DkeJCUnY>(<YNKA>LGR{`;U7Eu#xFW2GJkg
zU~FQJ98V&+?E-fiaTV>;jrbBGxi6We3C1!T5vAHXIyHP+KY2sZgbC>0Yu76fSnZ*`
zj3j!JQB7eE22fBhW<*4p!FdX9lQe01N(z3ZoHrK;R4L}gPYni+e)5m#=}%`}s9(;k
zc%qU)(=eYE5)yJ($Lh2<{!%03y@qQ8`E!%aqX1UuY10W2+zcab1em?N;nW;BFtc#(
z)?Y*`$hg^A4C9<90?tSrJbkEu<ofD*6a^K?qP2Mypt9HTDG#whm9`C#1dV&Jv(ED-
zPpk>8_#PpUck<G72V9EgoQve1*_Hp80ana^Ho_R|OJ7DDfcwXLEtsA(lH24DZ-%s5
z3?YEh5Z@ceTMjmk<8Bc|`ar?Rc0+`{Ux4D`v}xk0`l{bxAeIu(8BOSX_hQW{Pz4X`
zt(r93nnIy64SxfC3@@?jU^*ONs;+I^!DHO<(S?nYudzN{?RR-$c4DRXn2rS)06R&D
zuSu92ppO4~83rB&Z(!E4@cYF#QWtyo9dJ_={W*pj6Ws5K?-~rKWf%59!63n96oClF
z%UdayRe_L^%pgwyhQ9z-v1Kl0-(?9|ryX^I-4io$9o$_zbWaZQ3)dSvZDd<p+po3C
zz6Zj74V=CPm!>@Pmuq1D!r+4RneJIuqHb?84@rJ^9lM&C)wn0X;6*Ag@b^9nbECuk
z!Y#!eD#G6OL=2!wpe7g|dj!#K;Ecq+{NVG`@a^4US({sa8_?a`B;EZuDnQeOY8kkx
zM@2yq`+*`$YP0o_LIN2`l_<<hVP8(O(@2Vn{HB2lpT&R-0#wSoi!CT%cM;HCf7uTq
z-|Fhsy*js_eK8~8mw_vxh@@r%__|7&Zd!v)9D0&*{LgK4v-e0yNaFSIwBt$ijKQwI
zlZum2P&eSd5%=z+p+VWd&VhwDv9TG&9>>0A^kgiKMdf4${sM9I8nF@U<oSX>MqpyO
zGf9_TXs-AgCBd~MM_W;wV6R>-)?ihIa7S}o&JRvfBL6srwT88h|06h4=`NS+(R88H
zq_Di0U>~R=RKRc`&;C0&OC{N~e9htUgZ;tBT6JG-U_B5A*VWYp;_W?dkd43Y)a5wd
zzoOTs^LZXxHn(v@i@gl7Z-mLlt5<l3kV@MOafxip+8$y?t6JQj!Nv?~P^7|sGBi`u
zUC>Mxpdg)5*s*}VRt#1682!HL5{xWNFQ<ILwnKUYE3-k7F&(%y(cKL#)CmL|c3Dnc
zJ!r@cRN+XTBYj#R4ht0T-4eq<pa|Mn)cNtMqx-O;B0Awi(9ME~YmoekAuLTn^hvQA
z3X&mkZ=Be~@<@s<<qu$F34n+9#r3Lcj9<i0iGu}iT7o6AycQHP3vVu+VA$|Fr$KE*
zmbfeD2963V(J-+gw4*S_R$n-AKQTh^M2lEKX^FQ(r5Jy|%54X%X*5_t-tN!5Dg1V8
zzF_i0Cn7+4pY#`Q>|wkObSD0qori<&L)^@sUS3&6ph*E)0vIgXH^`89kk<P0pVuSN
zNU|)^Jle-D_35tA54u`zj~8+F!<|kOg;X+`_Fg*;b_OBcyDfkSHFZFa7|9ggV_@k6
z<p}GaU@xJ+Sqyt5W?@bZ+{h?{ug*!RP#4x4vnY5?1Chs#_E-@;FQ8cPas$T`z!@Qf
zS<TuNqL@QNLbU|xLs<*j!g5#sB;<}Y6mQ_$fjQSs^Y%@;Mb9@Zl9$NM{gsJ_ofO+~
zC%a#Eg`QU+d*wI)fimzjFxg<|<ouZcA2@yl9G3hw{iW8uf5V0V!G{sEWN!~E0=nl#
z0lM8pC|tBD&<;pGZeV;8<6?drHr?dl_T@|GI2gNk!eOAVebw_8x`HGMJUQp=Sl4$s
znuCb%8YtEspTUZ@&AuKF_UuI768${3@;%6O@g3wU^N$El;ZCqnWE#;R0)?isAbMDh
z9v>g?KC>ViAZSPiybG!m`qgEWd!py&5gJmq!|Vv09c-zua@2WSH!-FiK&c#k2CR1{
zTjYN^E`i`MfOpslJ+iy!vW{c(j|VQHd4|A$NAa=)DIw?!kn->MqcIy_MTDQA32^&U
zY^ab#qjs7Df7BRM8faKp7HbFrX}$nR3Z!$iJV#qJ5Iq!vbarfDy)GRoi=7IOpvd|#
zQhou?1yK*_RN>|h$HY!-5@--cY4mjt@PYxt9{${eEA^1*4&|ixbLEx!MXgVOB@N5F
zp9#UY7bO2(hxf1DUlx-{-lyaFDhWeB7^!m*L%_*dM^LAAeR(-#A%zH;<qMi!945}z
zhF~J_?xBwPDr@Y47@&3Ko!{rrl2TFy$qelh0P=~kIhC^pFc=BvkPU$Q%7lXuVQa^U
zJIWz?2YP4z+@AWZcfus$>oEvf0M-(4OC#reaEiiZ70<rzLJ0Jf`10is89F@!1L%)>
zOrudZ1XgjiEtXonnm`5Qd3<7(sZ0wf1~e0|`=R|YUs}1kV$crzUo6yG4~=ns_#h!F
zn%RD97lQ||qng^<(UI3FpbZ4e@w*wE5u}i6J{San2f$H-``tTnD<P71sRM-$VB~3|
zdP~S`Ao?@qEeHpZ_PDPBK-xu+BSN5m5SUz_W!^+${SZcjPV&9e_q(}RFpvo>*{;_n
z71kDwUC{Dc*RtbYwMl}y&vcuWfy_EPBe6n6CtqE$0sRV?{aoMNl!9ukc&@)c)YvFr
z0Zd&|A(+jJS;Iw^Y9(Q5)k^7g4V0liSA&5#2@!V60XWRTwlu#Drb&SPxRwDtRW^+L
z<{Ux=`%zH;+~0RB-MPW%n}R#>g7EF+R02v_NC@29mvCNSP{}<CK%&3kh%>MR=JN%h
z!l=TTt8`0~N4fj+C;RpDB`avwLCf`YR5doY4H_!YNN;9<5zGd;7k?84no(x}-WZ&I
zEt4Q9`aotq5PH1i4P^jx@<K3U2R-V=?d>TbOEx%c&m{;Rgp1{dVbO7bY6W0hofL{M
zTC9bp-%Q%Ap17`?P2TyewRyh#9k29{fa=yNMweNpQ$QA&y6{APe#@JG?upiGW^DYD
zkpe)D05Z@#Y5}Sx*i6Cf9s1GvVoM4-VIz&&=i!($i7NjL^vqz^251+ooSb78an8Ua
zmr^<o?Q}e15#%U?%>h83OO4K!V4F2C415yiIOk7-7S7rhBIu-@*9SHMlY$iXp2L@5
zWS(!gH%{8CB#HkbA!{YIVl9F@0dKauW<y5}c%JKlcWMP4Tn!rfUy<Z*X;H}vw^C$S
z-feiS+ke*|1LX_ZHyp<ID}^8U`B$c=<7`Rt{4%`%6aesXgSET{Kmo>-cGA*_t%raB
znA=JY`Ujymwm9%bm=d`_3!C2tavq&~T`H*MliBuD;X_S8o)qiT4cMWxTJJi#0r$k-
zWUbu~j1@i0WSJXXPfym?1PLXWl!-7y+?gOe2XDmzX%8}p^#~>#a5=M)zoVyUfo8vQ
zVTBOApci(kd<4`*V2i#69GxGn^~!*nity&0!YX>`OHNlemY~1q8w=QJ(+?|u9v*U#
zhghbBf2szzV_o-*-72EgRaRAX-Gk6H;E}6k$k`4qBvt*y=|=)b^tc{EeD@24N8m9s
zgC&MwI;R^^ju{v8%3=<06-O5eV0+VZntboy+W<q_dZ-hnR?mBYl2B1nex(B$G82aI
zkmqX>=br@yUk$!6)3)nuLBiKHfY+-9fQSl&VKQJyfyH<3{`Q40Sg;A-9RQZP_zzo>
zbaL3&1nR(S`S{YG<i$!m;t@cJ2^sRTvfWSj94F9_U=-+vHmj+jvHg6lb#q6aa-V?)
zg#}c$4O-MHh0|al`;Y4Z$0tGVqF;Us>QyEm0ezu?-5knA2VfZ9dz+g94^wo3&|to}
zA@msl6<h?&2Hjw23`{^`4nfEzz`S^;Y=MbI|6O;wK4=30${iLS-Yo&Z#k*2a0-ejF
z*KF*ptgKAiLxW3zC-vpIq-{6)Yy+MF@KE<#1XFE}fB=jOR&ZT@*AoKhE*wzmcDx0k
z+NeO#d;&nsA6kSV{oI&5h?GrvTLCBR{l#BEkA_SU`%Yu$^WnpXa-t8ZAdwTmruQ|#
zV58CF_I(*IfPX-vec9j$Elj%G0f?QT8mmz*__SmMURV4_*pP>paN);tfJ=Zq-4usR
z<q<Hjddd!fB+U-c$iXWP2w{Q3qY?^o-@;qE1brEh!9|6J_IpDdgW(Dw5y!SAW{oEK
z;$dP2TK1eO<+GaDlj|;nvClPLmM@vo(4c+LJ%OJr+B#YP75E+Vvrf*=03<EG`+e8l
zExPr!&SrViO^?VGN^FDtfE`i@H3?*51H{OoI5>Z9WrYZxrRJ^QnfWUupma9@g8}Xt
zpeNr26t;pMc<84!P#A;h$AWnQEWl{)3)!I|*QcVRqcfQ&V`)CeUKvIIH^An99DCGT
zcLLm7<R*y5Hb#z44#xV{FW=c3SRk^puyBwtk-U7%%gd<jZfDFWZ(yP1WX&i?!py|P
zC~ofH=tRQJ#mp#cY-8$VM#93v$qwAg!5{E<R!+tajN(@MPR63fhPFn={QQXjd3m>m
z-<sC8o4x4ICwdjcjCu)8A09NJ;j~X2_93ht%gY2c`<RS=)M5=~3oEwBgx%EL{=hO1
zi_xR_1H;{(5WIB+IQGM=tlBZma@0cNV(Yv<RQmn?<MoG-Z#3u3l9)Q_`k=plzR01@
zR_ffNFrd&!{%UkZbRMyVKp>7#azT<Kq(NkioS&3Lmc?G-b}R^!AnYdnHRBW7LJ72{
zLS%+;@lA(F;8$UZyn9^?o_k(@=YfSyC^dB?*fsy2Sg81(Se#xA)Si$EKP<|#FFo=M
zC--qA{@F}^>cu@9*@i0gCQ^6f38wNfeVXOe-!sI87A}W1w8k75QrJaXL*g~Z5?|AE
z41NBcIzldZ-%dW~Kst7GhW^p}n@V0W-*kC0jF6sw7Uil;=K{2I1<7HknA=a`KzQ50
z>)?`b!r;eZ+;<2xrB2n%CWDA6l}vA^%)2~=koxdZ2YolN-lD28)2vo?a*qtb6P<h?
z3BlsBy;Oq4a_5BH3DXD&y^kHm+>vp|Esv0)S#|8}B2AL+titglQ^+MDhg~n(xMwP;
z>;CDFx3Ycwv3^(WmMy=1(GTWNv$TKCU0mP0e#v)Ay3jp`BzbG&qo);7(BDf^{o%X$
zh0oeQlj5tLc_fCO;-WW?VxGY4p&*iD^r4X^2>bKD)fZKcJJnEVCWKat;;xj0LoXS@
zgd7m>!Mr0#A<CjQ;2`GJ?nb-tBrt0xEY_6mc3|dH?M*gNk;*G~P3qG4`lP~J2LHB&
z(n%sIlH^_FP3u$b-PK|<{mQV33i-Fw!avo;sP3sew9nd})GHg1`ku;1iJLnLbd2QM
zJ+S2DA2K*txLl~ntz{)VdO9w0*Af1*{LsO+E&WBz`Q1X50MlpnZ~Nb}_sJ%$cS?$J
ztoQB?+&>b3eUB+Vxb!dk)}~UmZ=@P|u<5RoHk(saG;s8oGFx_gHak^Mb>eb;eLz-(
z8XeO;kzXDDZPSn9sGufyE9x63<ac6B0p*OsjWo9!>Z+O8+@K?gZ?JX)-E(gs>ET4}
zRV~C5-hOV_#Xgbz;$~VoN!si_iy^E>N8Zw+RP7@@q0S-1<_^c&K9u!pnS7ov&8X}W
zXHA#PL3Dg_$<UT@U-RmNX#5QUb&vetJ1TKVn(8EmUsE|+xYeC2kS5%L#5)dnM84`R
zDO0$Ve3QvR2RAF+W!%pGYCo*GuyV=$iD%xJk(=X-`+ifFOyNw;cGLP;^yf1_ULqNJ
zBsG6z^3P&ZgCNnKTfJoB*U5@X->}57PKz|4a8C(iTSnstz0cF<G;%a`?AFngJp8%~
zl$efTDw`v#8wGg;ljV-z4BfZk8@l<sAb*NcEHoB**Y#$((wN)5cD5`m%9)*?I;Cr4
zv+1_y*M*E@^-cz{#4hCwkL937)!L*%UYB=6e!_LVcJaaM)@wEK#?wN^d0W0SE(*uM
zb9QkDeM45Yo7<j3lEDq!wHoV=#lXltl2xbZHqDKRO~}4?`&<zg=2%)(1Azy_tp0gU
z@>_RQ@S6QQliew=@`Vw^UkNIsNnU<h4O7|2UPq;3GmPJ=gjUp_e^V|VG<_l|RdRiD
zf8;r-O)tMy-$*-pa8aGEosh^L6+G58u+2ENI5~&?MLZUPd6ve)IWAw%MXg$cLxm6b
zBTi4TOl6OO1A9h7-*zIp4`VRh{u;A38N`tzL$^nWCLGOZjCbM2=Qho_syCvW+voER
z34JU1L*t?*D9QD@=XL8YQAJ`#@=O*>{!?TDkZpBqqKE#mA_C~tH+JPA(V^%ggdba9
zab-t*5{Vuv{J8aPk$g6pVRHHF<$!E??RYwZi*C)bdJVgXs<k=(>Y;tjz|<FIs4;`w
zbjz#M9cJ%X4hHOCBQ8@LmelKjMvQ79?X1jIA(gL0>QJ`hX1Jg898Fx8?jP@~c=cv}
z4^Tkeo3rcc3f(K3$bZ5j{Dj>Wx!bq7S<!4(`qK-o^4zb{K4QHW4n%l9H+W^@uv@;a
zC_=5TUxyk+J%^Cz<X_8YFxFN(;hM$JMRADY=zCLWX7i_!K%SbNzDGbe@)Yyf;c@ia
zO2X!@%5tj8loY!<-(mFcdrB?eM6!|(N%4MDqEybI0iDDB0~>Xp(z~oSku=iZhy8s#
zZ3~;|QE}HidvoeV8(gs1Qt0cLwLVaGIQ|#LyM-zuD^Ve_YbnSYLpZ9)x`_;r?TQ@L
z5mGo0)dfjz&2MY%LGxAb1B&mti=^hw3Cy<`T`hNcy@f+&$6H)mOEG5WZ{9iE%fEy5
zt1>SG5P33c&$o!wv{d031H7@z4%-m-=G#m!oV#vhS5|+f1pjqibW$s5xiqPd7=dRq
zV)ggRY*sg{k9_WTr}Si>Am-%bM7~%mkKYacIQE;G{@!jMChQ_Dq8cH_o-Q=JLNB1l
zNhzR5{}<G|ET7lusJ6K~<XDNh;(?@3gQPH>$4`H6O-;X_Lx*wk7wLycY232fcp?=0
zK*Q#axCThswi$9FeMD{9b-OuRao<n*Xov`v8b5YZn#S&YEOZfNj>^_yKJnY-*su1e
z`l`<ol}D`U=Ci4ottIN6v*tWORE3%|`|*B~&T*CvhqX%;o_WR?z24af&g1znVtH@i
z^1sy;zmv1mRP@fSU0`)(vpkn$JYL=zj;pwJ-|n7*`vW~O1u9RVvD<I6cG);=uk4-K
z(tfB&|6H~s=f2-=q!LvHyk*Zld?P9_QG{xjV$USp0(|&$Z0{o0AeplErHf`~@gVRc
z!kKy73Qu2O>OXhQw~hTK+Qemmn!rxI|IiHNud6T}%PkUZ%X@?Y%Q|FbIGvBI_>JDN
zK8@pYXm?*%j*MJnk#ae6WEUZDV|+~Q8D&dYwOy%tC1*dtd(GpNxe`m9ScvCvjwA&q
zP2`sqmuKFr8GG%QB%X#&ZQ0r_R5Owc)JL8XqVr_)`B6G_mGi<-I6g}S$1jkQ@NF7D
z>!iu92&jxa77&dL`KW!Lnz*Xhl^fd3&`XmW9hI(yr1jgpm^JopYPq{l!E$>hRxP#1
z{WrxFD;v{4i>d!CnAljD*#1WaQ;D`z6hR$U>sVQS4zwZl`>2#+<m?z;SY$mTI;uqk
z%$|hT=_IPJc29Rb=qr<pOKB^ian0(IN7QWf28g)5lD4jnfkS?ZDQfULMt>Dud0U?p
z%+~0;wpev1mkVhGQ$WD8|Fa(yx3KOUg(GE-tV{IwwOgHQojA;q0y?9!>!H;8cSX1T
zdmVwbYudjbmX<GW_qh>f+WO_ISp&U)v@%y5#kz4HQLrSOf6jnp|90Z-{;(V|Oz(Ah
zV>04DB`*4D>Y*=1;GNZ@i*%E1E)(<IzK@b7_TXBZ6yj<BCVHnj&e9bEe#X;om#vT2
zUpWNdr^LF!uYt9+!Ta{lW+m{GAzV-C0&rRMXL8b_3Ln3B6-LY{yh;qZokB53J>pO<
zYP+<+#FRWloTCPI)x9u6HOgjdRPT#mR!^@DvJU*o&a0z4GVW+_69gbz1<xhCe`gG-
z3VNm5c&FH`7!_#lruHSERadyV@}pTxsUUnJ3Rab>rN?E=j@tUBoSisxQatncK1@?E
z9$J5&gqJ?}I<|s=H!=R(X=R{kJXKuYMXT;=OxHKHVtR3lpa@^b&Yu2vg_TaG2zt*?
z&&*w3ucqsMKRl*t!O&41{?ytfHQtb*ysg!rYPmdV=Jw$9VC@hok>Xdev$|JBLN&7e
z`7Pl`!6K5HZ7`*un;`>=WVS*LDTE=6SX1a{s!lBI{>>Kj4lit|5d5)@UW4DkkCAIe
z45vvVZ%Mf(MCRT79QT_p7+)Fne@M$bW9t@@dS7;V8=S!=5HNc`5FT?It7rE05^?IY
zPH(@w0@8cW?0QdHf*@{%PT}?NXbKM-rkQe1o8o6X^b)^(!DUV5o5=i3L*dp2*7sE=
zI1bMG3dZNg5ES_#?_3!7{LxqLuz1EZVgp;oVhbp-cXAEMPuD1v&h@t5yx06OFn*~N
zrmgbs$(Vt=C{ex?EzS9iwySw~w34OBD)7gjZf|Un3Hy$u{Q-nHciL%6PsXYitaKfz
z#7kDhRBygTGwHa;wvg?wHPAa)KQeADG~9ovDXXqDYf>=zILs$gYw1<l)>gzDP1Oa>
z!th_p@|#;TdX+&pyfJ)#ZV<`*H+{jn0lvPZ(ATfBPl_rRSxqr3=5uU;p8J|rP7jpY
zq=srGYNF)#rQLOD6Px?E+@HfOrNt&O{PW`9)#j*m={YsZM*p;`+vTnIcyV$~Jj*p`
zN88W$J^Vh*?Jz}Mk(aEObL4N3RJIh3OcjhLF^v_>c--7aUccgi#*D%k=Gp5U9@rn~
z`xPRoupmmT8Qn%;NQzRGjlDI7oPES0i5aZfNrK!*wxNdJ{aV;GkfT)Wj`d*fSKSQ%
z{huTOYtrseJoygZ>wL_by6-R(3rYT?^~^#u3VcF%M()uj5at%#oj?B8H!S=NoE^~f
z`6DLxx6^g?V*!Q)>F5VyDmrDsfzZlBpu?@IIx&1b5uhFc*NuDq%Mgp@OAh|C?l1$2
zhyN5I77v>33`-Cqs{I!XmId}v#VUk+M~!O+x!Arh;n3uc4AWI2ED@sUO`o3X&BB3=
zYWX{*!(Vj^t|f~?+h=qE7jhFx&W?E~zjM4fJ5=|Ai<^S+s%%7r%?~s=H6z!K6KiZD
z-VYWk&FGxp;?67&I!!kUpqhpX!#4FI9JaEP3r)}_MTj~VcIMBMxYv;QM9#~iRL+I-
z$uh9_k5RvkG_TSgctwq9%Q}9DBGbje!o-enV$93J&ep_cVsZZia~{1aUV!K_6j+!X
z@48+2R?_KvbJdo;`ssw^K=y^~wMEIS(d&0Q{yP&X4<xCCa*ctq)4s~@ovaK{lo{#W
zv(hrC)$@7bLQY*bF)SpqB#Cjugx$jnjhL9uS@|qt$R_*bwzm|XYLIPuLJ7~?&HOco
zrf=v+atyNRK8{zs8+V7TPPNoV)Q~--(j#QWMSsjWbZAHvjHw7%WGCF3VW2K)a18NB
zXP=IRjWRZLJEHn%$%`*_Tv=wx*QF7Wu^(qqvOA@p&&<2u<WeY?&rB_fX_;ek(FOax
znNjKtiA5t_gY|y5f=rP=Cqidmk+3TCZLw-EPXfyT+45jVbnr&9X%DtnDsicRjAT*P
zPPB^@&(cTIKyIRR6U}LFc&GtHlr2B+@HcpP0_dHgS`}Q_or0x1zMnc$^Bol^?Kx*Y
zYd1e&cq`Ke;um<WN=2PfQPkrZ(ec?7eYJf4ysX)fHfN#mC0E}S>G#(XaR{|$f6cxb
z!nk)o2Wy!V(xUTZ-&SC=5%xAsB*hbH(w)|0aKc9J2u!JmZ##akDWPu}c`c6LKDO{$
zoSCdAnT~Z&`587|V^XT`%0(fYrE9@Pd%pLp6ST^>=?2BN;&TZ5=ufGVB7xY<u}q^p
zg08<@rC(*ni-gYH@bE;f$yvT{Rv^}iv%_P@6!q`1CCZcd96I5Ix~0-v&{Jv<j0q#0
zPQDjUeHZe(SdF1d*N5HfYTJ{d81^r}eBN*4Qob=Eocsq1juTGP_{MZ^pV4q<4_M97
zH-mv+3?=$ko0pR2h?;N<T~^_pCa&bhk@FEXrr7)KEkBbX620fIEw@_yzCIXlo#r8(
z*mUL^qApqau)wMgpH>n$%dXB(m`p-LXP~b2nfn(>o3*pCL94BPn3n?g(9oY+oF}y@
zR}1>5dd=8qBT*WEPZjZ4VpR9(V5kTS!34ve!MjWJEl*5^w~4p+Xe=uX{?TPTd6J*W
zG%?z|zsTS*mm~gJoOWO?ow((krggNfM02#ue~l))-F+J&sj{-lmDk)x;HsE`Kk2Dz
zHnUV6qfK_DkCg3Qeq({Lq9_A%oy{Xa(@mUj!v9-qG!#zxO+@=j!l!(3xPrr_`d`T;
z{o3fmvEFG2@ri~<8aOp5Y0*4*w0^2)Vy1LXl=R>1-!El<QWKMv*uzpr7+^a<qPiRJ
zK-ktIBelJ<rMES1EJe#{Aaj|jE*qOf6xkMKC(iQuiBEmy`{&#D8u4-F>}1OJhgzEB
z{2a&v)<Y5JJH6?P!kx_5GSYgZIJnt|NZ5Yws<$c`)lUyh4D~qE`3mw<t@1K#4-mJ9
zpveoSG(L)4US^RY6(e5QAGT|Z_3cv)<s&Ie#sB_R_PSgra^g18k}}aW>EO;y%I_qW
z<XRXh(cseCL+rtYCPBooOJ7I9WKLL*I+%kvAd&nGIVGz7mijF><XCx}A(GBXG3KUR
zHDl_b70EtHGe?Hn{P=02X#!!7RP4Q)#O{apz9xO@5ev%*ldb$1b7@V~amM8jutw~z
z>%BO~l=5dP%#$SzGV>w5EM$^I)sy8nnG14X{^m0BeJ86FvC+P=-}EQyBkSNKXc4f9
z3V%b201=40JZN2pKtBDmu;*fPN^6KRW?~3kajtteE?V&yO<f-8PlCjDIwyFW&v6Nz
z9{4}};O`xjyOz$GMl*NAyr-yLOiQETda$$c8cx<KG|j5~55i;V2M6MV1Wrjh4B#0;
z6Lg7O=L=%KRVTP@zTJ~Cx*FKXukK%sYS2Qd+_dQ*=H)+WOwn%%W<0d<^b(z*aV_G;
zbK)cyt1c?oejX<8A2+~t;ZzqKXI?j^!n@#<?hy@$dWuNJyRJ+Az`kQ!qOxRhnsJ57
z01qP!b?-U6Z-G`oPNvyMTCK29CN{|ft?n9~EUcBS;-R@{ABvaj7>BT_TpA>pGv4;A
z`*IofK%=ix@$kE~5I><7p3v@Bj*Op?<O$-cN#zD`2|E+HH}jhGzpz$uhYCpH7=D2e
zv6iZLM5ZB<osv~EKX_Zcy2#0oRkIZ@6rghN9{FqW6rSDK@@nkfm#U0wau881=PEeM
zoT}B@na!j0M;mjZ2{;KCVC!o6l2iu~Sa~1231ddfg>4r_MmW$~kn(qmr^O!(2!v$o
zi>#?%S70n5q}Y)3Sv5k$@XyIHU6sRW7FdwKwXj0)46i}HjUcK@#<Li<xH+?upf#cj
zXw`%H5qX5)a+FFyMubrSI}vtYr<U$P%+ngDkCkfGGbi8IN1m@fkk4+~LMI|eWmw|J
zljxU*li!OUDra%Fx0tsgUb$J|xhL0%b@%IJ8-u^Zy!_EYR=u3sS(Qu%ik3!#ujZXy
zvCTZkOb)-kd>$cHNmlTmKv;ufc{9P(N=0H;OfJ?yy2!}oCDDw9K3EGW>wfo?K2l$1
ztJiY<LxIXS7}@AFh$@18SO~;lzEp8!aa8X+{ZP%KySCuk^7S^5vbWYL?H#NzQzyrh
zh9OwJbCH9(Z~p#vzQ7O(QVfr{5u&dzo~>T(%!2E~4B3PszpEyNj0*wIz=j)pXn?BG
z9v4g=F@G^~?jfdQ{G6b3nn5XgyhxPkRk;#Luh^9Ek4Uw%P}Q=UrN*G-FlHjvA6#)*
zb5sNYflkqmKTC@%jG)HfIWeNhkfzyYHN=IzIwkm1S<$Oh{l2$RsOaKZ_^=jB=tD5!
z5Rbb4%%l&;$=TWOk+)O)EG_u%swg_^G-Jgyr>eF~<Zw-}l`SbD=q_C&VkkGwhqT>7
zjj(>di%L<d*zoc<7msqwodYPHkM#C`xfP>3Hkg}5qCV7Qgtg#pxx>k%LfN8_eP~HX
zD9@vN&}*vL+!#G$<9q$FI#+BJ)%k7wYg0K{NPK$3Aa9iOQmyUFv&yU@SM)d6;TAET
zzgsllxA2SMTzp$^{xs#RZJE6yxqFS0(B^88CL_`XOOZCYk{p+rhWFJy|4dVOHKM24
zz9%8uZD`E}5w3eJ+R912!c#wK?t34*W%Oa_BTeU>W)NSV#%9U*L7WN7YUFRN<YNg_
zFKI}0lEL)YSJECmSgs6yE15Fn{eM_7ND3y$%v~tU37IKB)frJ|O!mQB6Z{h9C1=~v
zJJWO)H+u}>$ku`xH?+oBWnXJcau5c3PMA)qtRG3G)9k{Bm|=EiRgu|;^yIr}{Q-j&
zhHqMh;Ff8g&C5PG)@yCPm0`wew};EQry}mHQNCsg7YR1!J~e*JRwk4|IyC6?tAeyB
z@w3&RW3h<bL_JC9(NZi?PWW>go(jis$5V)oo*u5-{W{r-ZCzK>kKZ<MPwsj9gLO{o
zXpn@|;YdmXBMspwJo`qf-R?A_E*o*a%DD2`q{v*@@S2>c^(G%8+LhOpl8pZ_$KpvE
zY%hrE%d(;TLpFoy0IL~Dx}BX#Y4gT1y45<-O-TgnJV?$^8pCQrBRmvYwInm$*@r`B
zq68f`-bZNI=dKGyQ&sd0>2iUhIbCv$qPfiOUVe2j9)A1`T0qVuo~T#^^8S7_RjE_%
zZ#CI#3p7D{teLe#<lw8vMMM}G)o<c(BKaatm7R!%Q_A=ZOPW=)4Z3*AchXYadXHHj
za4Pp)(BeNfcwQVvpZ?|=RF)JkFua}X{t01^lJTAw`okVFdx;_IFp6fMcoO@XdRz%<
z^<>9BlZ%lK)awlOSwX%JBXBd4;S+QEtYUh$)=)-`?C#Spc`Mh}?}Flvb{q>myG**b
zEJ6(&-VdJCz<o{C?z$)mv3GRzgglhs2{N!@NQ26%gMAy>%s#HLRszwxVT0mVp40r3
z3)4jnIYMk1dr@KhTGP!^zIO?)|8>xk<v_dulIoLt`ClmYK$p^gi(GL23v$82&H6t>
zF7h;GzT?-Sw_d6Z+k2_Q3vTXPsgpKj)3%n#)Xl*q7E%yb%ca(Dv}N?>_`z}!IAEzn
z)CUMbEl2duq6)s(!*1}LBNRAUN$a53g)VB=<@eUb^nL*@=Jt*@;xS%NUz|JN_k9Pz
zg=2T0B!X{LUmMi(!V`JEFQ0O^#4&M;;vFm^v5xmOZidp+ha=*tTVA=J*YeJyTGUsh
zK7Nud6#TC}@Sr$G0-oYz+8#0lp3A?hQ2g=wYtQ*@1A8;)jC%6d?DFXchy{_8@xPs@
zqR$U<q%ve1ZI1)yB04KiPbx#(_ZdxzGY@$q*OK4)D(5+-i%H=LuMJW$Ae@vOmpmfc
zV1FAcpku389qaDa3LelF_QL(nNxzK}Px+Ni9l&rS`4mn!EwzrVVyWXEjk#!Pt73AH
z^NVOlefMUY+4LYLA-R_=i`P7S=}lktFMSP@DwYK#e7~)aCUHB`hWL55Q4uYbY^^e%
z(=Bb)7THhJ*i9{?L?*0=rY6)ikXo-tc-C#DZi)kx2n8ax;L=ekV0JSuc?OdtP&OBG
z(XF=`E=h{8ocPHHR88%i;}m0u?muofT#VhhIH%L*aid}={x#h7xFc#zGx0dL?xm0)
zb~!say2221TKK`9#5_ng9B74#>f@~@*{nyYh+c;nPT})(GBJ0SS#_;;LD#-h{EBrn
zw0Y1lo{xr1BitrgmTtWu=t}u!{HK*>Kq%LJ``+PCE4=3Mj<HJXQG;wdjFoI5>y~KN
zfu>^(v<w3#@&TM=O8VE7L7ycssCY?^XvyQ4gN!_^t2sMq)wNQ>#dj?_LaN2@C_b=1
z{~2A3h4+ehRj@*a?u#|q*(iEgC~iw>6k61a_(f^k_YV|~_~dDwsiPuwlFRoC6QzB$
zX5WDwGm&mDdzkts6=wn%gIldH!d7S)7`uvQTGzRpKbQs)=(=Fu{KAucHLNTA-DmjW
za=M}+Q+=_cG&`@*Y~;%)4{^DcZ>cRcp`LzP7=cJ5QRo~vm?f+Pg>_yl;z-qTp|JsK
zoo;GzH-`~k$eko{C=_Pe3lAZf`OAXBggWBFa?jm3ULJoXwYSVq+HxBryzfpU_(+4o
zz0;0!3#=Wlg+8yjrXAA#{37-(X?$wf;4DuLDWKyB68mFyuo<d=KK+-~1-nApEe14E
z)mT!dK`KOcj63TTMLWorelEt+FsV0x11B!LWLO<^sH%%7R1rgx_zEOe+QsnEmaZi8
zYH^4NU48E&zW9-_mdk#@#aTd%vgdzGX$4R5b>^0V?bnw@Vb(!gbu)fjO|$J+4ICG(
zLAvWaD|5cD1G|6P8shW5-v7;Ye!GaS)uSosijku^gxh{V1tYP%xw8fFiLO?#$LQOg
z*F&~UxZdi|bGP1_^Z=(u4~J-e)5tkOW9Oc~ab<A<+S~aK`YUQS{cUVC-xgbkGmLgJ
zw@H8D4cPQaif4zk=SNvSucP`ysb4=~G{I`?VFe074N>f-v<euW)wj|#LQ$PgNZ*`4
zNW@#f&=iayax9~mQiemt30Bl24?uo3;YnN<l=GbNe-*tKo7qIBNx>l@w;|6^lR#g?
z6z4V-%fvcj?xOkW%;VIO5@dK8=}fZxDZu|#ARZr15~V|^uWBl_NTbgGyG-IwY@hux
zLsV}qmU)FIWz~Db3$cB%kVU^)7f8yW8Gb0^(4Ep~NjO=aVaZSS3`>!rOj|B**BlQX
z$jOdt1lu@9BSMQ$;J@aIQb873S$fcsLAL9BN!zc35ykFOa1{$(JQsPT@0GMY7fwuO
zB5AK#k#ZgoGv}e2v>c2p`p)UGD*41wOWbhq*9O<BmyHZQW4M*~C;SD&Rtj9>_{bhb
z<7yV<sD*S#>kQi?Zmrvd_wTch+~rG4*U+!|ro98*-atklQ+!TSIP52o{NqJ57!0MR
zKf>T6q8MvruXaA9x1pBwO*h6p2*P-knyzF-Uu=i{Z)OLgp=GR_3i0kIbjDGemf^Ws
z4A<L2tgQs~<KV+R&vls4Q9C^)wxkDZsv`a9gY}RU7%IMlbybxi_w_V<DVx5H(Dr&@
z6hT+_>mmuA;?;)oqR{IvpR){-;YF+UksV6QgV(<(-1u`klZ*8gQ+7}IJ4od!tS@Qx
z=ZyOpSaim~9T?7^Fna6}(JL``*jSO?>D!7h#P98CbR-S-^S~l%5ndzJ1Zw}iSU%u-
zm(+gK;V99`6LqwG@t8NJ;ws^V@4syE*P*1E??T0O-yZd1-@rzNGWJQ5bp2rAag};m
zY!EPO*myQ3JIbX|+pIr`l{X~uGZYU}5ql3-CP}jQnP;bc;ZXUFM&oIs!adYJ>l|O#
zf6%$)N#N=H7CFJ!ilJ?&pj~^U2+j5F1J2v103`V1>QJ9UCVZO1GvxAzL=@l1lvjLq
zrTy1^DoMK#E(1dIG?VUjfzwKI1tGf-z4?B3({&QHuxZ+5HW2XkKl<u1R&Y$;EOJkr
z^$KP?NSRturfAi`T_K_;$JeqJ-9C6yXj@^b(u+wr{<t?N+CEGw#?aK`gyu{ce?L|0
z2<1xSud-+Z_r@sbj!)sK;uwVqK7Mq>f;!3gWFdJcvZP>zGd$HGP}(P-;Tm3CqP08z
z&1{rHSTc&h7hZ>HHbU!rVZwX2zkQSF5?RDcG3OsK_C_bv1*nY^oWECd5QoP|>53ah
zS`W-^pR5Vu+YDDb5*%p^P-#cqg`f81l6p_`>Y23;Rs}3FS?#I#S}#7j7gL9;fPGS7
zXL5Ury)a^Sy|MQR*XYRUKC5qF^V@kNT*$Jj?YZ{UGsZ%&dZ@1>>Vf&UdpNr`@)@=L
zc6mKzG4ryTmfd-ewa?3GA$oOGNtx*%ROZ<nuknP!DBReok!^9thd1lewMve6;=@Rn
zJ{EeX1+nPV3L8IN)S&z6kkEhfb?#U%3;V?Q_qkI0{=%@$0P?Ll{;>Sd_%-7NtD-Ct
z==tn(26C;YdILf7U_JU%Ww)Wu5}Qu!YCUb)$}m>Q>Vex>?9{uiSW^i~!S3eGF-msw
zkBw?up0P{0pP*?CZ%_2}`2QYHF-ntk_+SKSBb#W_25;9iXBWXILCtMjy5A->xDnz|
zW<y^|CNWE$o_(0836>vvzdWfJJ$O}eZD|qF37x}E4jVrjt<=ZkZEH@D3!z2#@iv|l
zL&+<cUU8GGB~Ovo`it0?Z0WN7_0wsr0&g9cjKpP`DHG_&#G^i$x}UO?OpvfWSPRMH
z83E^h)#xp<((1Ox{<Y!}8fmP)*)Z#Qoox%_AJr}NA(h@lijgb4c^W-d!(<xJRQ?n`
zX~0p89Tr~T>G4se2K)C-n?Q?dBctTC-_ztrE|jFI#*2_{yC9oR9Vtyq*MnuHVT{G!
zzIkakz|te<$872I#{e=UMkK?E{y%d;UJY78(g>C#gow#8xpcjTCqX-vULw(7`lvn4
zEGB-Z?C+(|DJqPS%rj$IPF#t5H>)JU63c9Igsbv?U^q9Cdrf%x5IGC+dGq4)Hg~yX
z$xr<>CFuz1<UKfCx@ozZ(V95tZI<zPj4xOBCS&<V?)5^LL8@QKXbEmr*0GR8SctVO
zH*Le8cfMyf644b{rWO6t7E%5bL3Nk9{U*#3Mw#JPIU%B<DK_?(gIWIqj9|2!E=7$z
zz6z$66+@$jod`!I$HH~;Aju#ZX{n949vquYoH)j+Ne<YtawJ0IZBrL%8L>H~oT88~
zk!h%PO1hk;^~&1*x;w4#3JpdKcXYOu#Hbfqa>W$Q?uwP~y@P%@^a$Im*eUr^yw!5M
zYI7Q_@bZww{Sg&9PqFzG7baH+<SPA=AyRGox;u?Ck#gDWH7^Sj@!^Jc=+b@OF0%8`
zwNxUuSEu(1OJ<_!Hug9;>R{#bh?ujhlJ>Oi{onhUhrh21Av=rzZ4J-*&l>(8R4#SE
z<uW#OVpMWAaC)J3Nm=Wg8Z)Yy8v!C0Gb=L}qolF9shJZA2MZgcu(^|?g0X{$t+k!4
zjj@dr2^XV?t(C2VlAXSxF{7xli@BlkCkf&I!vo`H=Hz_ghy7O$82G0zR4@g72Wv+_
zDf`DY97$MTzHzp;ar}?p`)_`kf`hS<xuKJ-1K^zjvg<#YWDrL!AwHPDZ~32}<X!Y3
z(2;0qIjq?B#RRSLm0`S~{G}N`<%#$Pl6}-9K@-M9$c?T1&@y%uYyts|sVKNgJ9z4R
zGB5kCDkW&_)bNpnNcUmG+4W{L?Zeox%}&D@_t?D;=RPs1AUqmm;H!W6!UhHTw+y#8
z7l3)fzGx)S@c;G8+B-#?MudO)@jw2ppNWt!wQE(obO6S`yEYo6Xv03yzq#hie@=#h
z$X*(Oq(hPZ_b&`>`=`gheL=oxX4uf74xC8EY}EhqcK+i%#-*J6*SGPXm&(*YBN<$#
zFQ3f(H=o7Je+F+B?*G48`2T-r|L>*s|MW6m8i0R^0qz4@Xpk!D`8`h5UeDy+*m>(Q
zT_R%!tK|$ZT4`Pcgk$!T<=ftt+jQ4W;CfJXbLw*#!8xT`c5a=)=QxwU-c?jo*u3Zi
zf@i-E5Vr`p6dP**EsK+YgO%y*Xquzx*!y8dF4vRPrl-e-%UWw6aBj>zKXKYD+0_AC
zz)KK8&?6@HjoyP3sBbP8fwVcD?*TY}Kh#cX{c~r!!w1jr6Q{3i$bW7tW-ib`pP6a7
z01OveBwn&|d)xkYgx0No6*m-kx|-T1F7>>9+6KC3V-u5L$POwZkE5ZoHtx&qM8JFI
zw!d&X+A6qR@nXb;_h!2B>MN*9<ynkOt~E@$AxEu6h4BZ{g;``Jc3Blx_T~PC3@k)o
zyJ|&KkX&r`t#F`_mhM2}J3k_f|81Rhoy&3Kw?W!FOx-`1Y3WEX{#k<MYEeY7*+aBP
zz$*dW1{ub`Yt>sq)k*5Nu}TJ~ZEF*smrk&WdZYP%0$;hFk84vQz6<@1=%MV;!Q&%R
zz1LZLqI|~v!BAQ2^FfiVg=N>_Olj9KkQu0XKkgq~AGST-&I&x=ZT_h_41h)U1oxwb
zcNK}S=AV1{8Y9@DZ}lL_QH1|vn;oZze;fdGj)zmg!T$>=|E&)vO_h@uQg8TEnXUg6
zu$msorH9X?RUh$8Ip`Aqv7knKyvvctR7u3tiB(bj<fhR;HBE~N4-73N;xHBcBeHXi
zPChVY>iu<H@lJVTq>DNKw~}OF(e+5X+U$H($HPr=N9^mHTzE!NX#261VZI+f;#JsH
z*!6`z!APycG`t(eBA`iUv!4VKzunRMxHptUx(?QXMft@RfNDvNw$y0cL`|2SEnTRp
zqST~6^{%(~G@jz40;Z5Emm<hrOJVZfDCb<!zL>}8a-bjLVEA<=IdPz38H|kgxW9DY
z)6+*!Wg?d2!p38IlbZN|`u6c6GfyU=Fu4@>r@kh-oY`WbGc_6kr;GL%@!&TlRXNN<
zU-R;M?r+LxSLYf$>WpVK<ff~0awH;cmLY%QAVOPObWJqAp#s7Q=Y!!HhQ%9GV=+VL
z)l@#$o27f(widhd$-L9HHqY%oV&IeU<!M*X8`#PK<!OsXDce5}yQMsMwCz?IM*1}D
zf9y1_OpqYf^#L&J+x{FBAtk=|ay&`>i7N)|qst0u=;#F8&J?7~e-x&m%egg+gm5s3
za5vmHxgO6~7@E2p*HF+L&io94?1&^JDk>~pPK<Y6YbT9N<kHowx0^^(G-R3NBq%Fo
zperjzkRWTu+zDK2_Rr~)wjlQm#366)M<BB^6xxAi?nKgmt&1)fC<GmYW274`Gemm>
z?yjy1%9A~g?!x2!jEnWj{`j-L6znrSD}&}tIBIVTF(R}?#9MRCT74ea{EbNc`Hayt
z;0W40S?MtJ6;i)8CwMzQ`Mu;f>Y~(Fj*fb%Bs7IK>b2Y5M`nZ$CQ~j}ixp}4gc0*{
z!w%WO#zl>!G|7O<#z+;Kojdrg^nj!}MQ-}*40lBm)RUoc-nQjTuJuy&z*h(a&f|Q2
z%|e@SBJM(whY*v|Bn_1}s!m-@)<gElT{c)&=!np$yF2=vDx6M3Jkidd`-cMFg=egF
z%ZVP!iEjR#9jLVVQ1Knef%w5*pyz!((g5}-FOPMgic4&Lc)r-CQ}YJa_X|Mha{=ra
zUhF?y&qgGb$p3eZl#KII!GtkJ$h|yT?K)rEZQGu9K9p5XPL6R}7nT$>Ja)wc_h4X4
z`8@oiOm=a9p<(E9%!!s?v+%9S{dpk@K1YdOM#Xz;zHVoMm6*gWct-v|XXIIec8_qI
ztL_hJ3M6;WbN2ql32_NI$SKZi;{mtJ;Nz~dntvC1F{;yKBAWsdH6s~aP7q-Ht+i?*
zU%6ugiEo?fY3sLx10V|m8{TYi>o;Mi^NPvJ%5U(DE!R8c-YpL{&z?vQr|^s_Gl8fk
zy9E5HV+gS%gR!~)i>I#)tE%CGmF|*`BM1^ocej*;AbCW(yBn15?r!OBICM&PbLj3a
z5$@*w?!Es+e!z3sd#zbBvu1r^gZEjRZ?FR1yNf<+XdS~qb*@OVgTqypN5ORGGn}C+
zt`UNfEqL0ZJI2<3T7I|+1g$ev-Rt8acP8ena@UT>A7x=&4&j|FvKo+6thE}BTymbN
z0fp<%V`TTTMVki@QiEqXdS+5itrnH0?d68&9A))pG6;>plPaS#^2>FgeO+;+o;r>m
z6&0;nr{v^2<^4BTGmiUk4z{_8p!)Qp+ygX1hSbZ!_^BFHmt7p~AJzImneaSKF=6CK
z+|J6;OTGoJk(oi+ICi#*3wAoCY_)0AO`)r8e^F7x0E5{;L!R@sKW<o`0&v;&5J-j4
z$>hBPUeNS^j|{>4giPA4Pd!*b-s4~9FsD1H(sknh1I11JK<n{x`60!J_Fq4G)`ti7
z?vwrjM<ti4F82Nxll5=urn2Ml@u_jm`oU9nS*_f|=?sV8E*7^hh27YT!;EbJMH~?W
z=oiXDDU@Fh4*`{oF)3c9&WyO_EVBWOgn8lXPs%z1(aCDJ>9`?a9-6JkSrIaL904NJ
zB1rkUtWMlT6X(lhfbR0u*;6b!{;SZl;pTlvki!+PlKatUBL&ob<!{gx>MH2p>4M7u
zYAj{&<v*)4gzxQ6%dx+QwcoMWF3p(fDJ(Sn3sa<6s2CO8c$_$@o3-fiJs%v%ay&k#
zd*{-7z7amURAR00V0id<OK5-Y_fc|8-klLhWEXhJR+E~AJF>Bfc!Fmea*6+x0`)^S
zG~YC|z@nv?l!9<ztHsAkEVmvvg4Js|+K@9Q!oiopLL@wfowOm3xS;Hi&JU8m#9|f9
zi)8=KNUQ}T$gj=2PLm5r`!p*}eLrOGPpCCEa*<1Cf$A8Dv!FzD#N};I8a|R{sUbNp
znl<kzJBKb2zzz#*1Kwa!SJN&gjv&AjO6MsaM$6b6*P-}QQbI~MnO5NU{s;+kx2PrO
zfQIT;0qpirl2FRde%Blh!m{F3n%F)%OkW6Tc`fBY$mSu(pIx0_4!2K@0A_gvyfGH@
zl`lY~;)jHnzBOACU>W=MtzSPA1G?#200$`O^*?<0-}t;{!2y}sr=Nd59uxxWY)m>W
zmiz1Pn}e`kE{A`ip`p_J;xX&H#1Ac@qyoYj6IE~eOtJu20Yp43SLtn`jSJr0RyBTB
z`pCr8C3J8=3hY2P*w2<6AB(3`Flsbm150!*pk|%V*u0`7Y^Jt!*vHU3^@Sdkp8BVK
zrMy+rf7X+eZxi`Jg<*edYlPEz%x=pT$!ITf|AbL6IQZ7NJ1qs_(i$0BI)bUW`d>_*
zAO@@a@o{?~;2fHt9}<T9<wfMl(#rIw1<H|&5^ItIL8qY^sdttyx{$xC?{4RST|*;z
zr1FKTOwU#W015zGW{kAPb73t30d_&`+R=FOgv1V!=bKZ&qjk6h>~#)`_{mGn%87pK
zPOoyin!7n51zx*0cEg-4l4!%>hLa*|k>Dm{jZT%6UaRksdb}-qAH%sEVDkgu;wnwH
z3TA^NAbt+5bRWa7Ng9Sb7ftO@G7k1Gyd;Q?_r2_VKhUQZq=irMvk(tUI*)1yN?*tg
zAT~txm>2vmVR554C%TBV)cN+Q)Esyo4l){7dDXrV?7{<a(;U;^M+p=&J$Atv7tWEt
zhpu94%x4e3?v3$#@{Akv@rqBWaex%a)Zl;o6Cj9~>pN8FJyhsFR7mIZ+D)t7W72*+
z9f}fuz8q!-BA$XgWR?B?e_5D*EenNOIH_3wHS%9DumGmL2R7uZf>pp~2a#6Gui2h9
zXR8h6-|vo~dq8bSOh{}rBL=EKXaN;&CYMnv9e}$=fUp0#-|Zlnb@>LAsuilLsG;RP
zW!aHmA5FXLhJ#xv?8)HmrCf7p$bc4{1N4Tuv>qaI0;=$_S;>1ay9wius{KDP`iVjq
z?fS4)41eNp&8xIe33uWD6cA9Wl=CWmIR<A%F-myHh<%k&O3`epFr_F`AQHZKP#8DP
zx0#S%4s10IUaJt$h8`3b>JW2mYnV0P1>=Z*dq0<k;N<6A4xDraGPJJtS+m1VOS7w7
zk39^4)<3w<T4f=@&LjG`9amiL%c$*TM0dDc5JF;KXR*fwbhC$=gd+R2M5>QVlLkgm
zh$AV-y3rb0WJ2QJ4&j_6yrQ2~Fcvqso2!DEg$YDhv}N9K5No2U<}~Wx7L7izXt~8;
zCo;RUZ6{B<i}JijWINXDFkRzwQrZ#~2X~_G1ocjUMgQUAw@2msVK8)L+Y2u7Em@pI
zgZLOPKN>P;)GUB4s*bz6bNj-CH5(jLJmlRdwj?`*yu^LhSDLd(Xx5{Y-f-$!kL<gc
zjHY|FmZ)*-$-olA>MDt=^`uismz1PbVXyZd6XmyyXSyAKhK9ryj(a)*aPO@B<@s>d
zBIE$;SC@KN;Qwbjo`V2z_e>5e3H=X=H}zoO1AzN=Gy?J&@OO5m3*;8-^tN2qyn-BC
z3L3b+yTDo~GitW$PZx|b3;liivq6O*^YJ62$HfET<K2m-U298S-P`acVrr&Op9V)p
zJ~1(EP$O`dB+`I3`Z6GgxP`p76le~1jC_IyJp<%kzyPK8VfplyN5|>kU~5f7&?wt3
zx=Q*MaaYCJ*_2lB3Zk-|D@ra|#hGpX{X!^yDUdtlYc>^%3Z6;?f%t%UWwml3;(sf0
zor}^p9|kAL)dpJQ2=x{)8RjC&6Bo1iu4^CIJh84#5O8b?p4H0Tj|_}RWl`6{RZF%b
zrZE@}ZE!rej1qqQ;GmP(=-{lt^yww{gTZ#QmEY{*rbJk_<$|oBSD-nNL8sk%UwfSX
zg9tvj(z>e3xxh)2(jXHDnM8f!1-;4LE3=`@SB5lrgfh8C6q#Ex-@~+=pPS}0TycMu
z!}e-)43l(j*{9ST(ro*!d-s(NjN>&|(b$dNn72|GxQ-qI75q)h>$X5P7^Hf^e!!V*
zIb>&Nr?Ak5$3DMFC;^oG%M6U5ud0mO0IKqg!<Ubr)q>p0rJn08rPkM!93&$2+m9q+
zs>2xK*V(CZJE2KNDqomtaX4>o7qhRQ-JS%<u6eoK9CEU0b@cq~G3EP@Q(D=1MfStY
zEE(^Bg0Lcw{8}^tKn%pA90ODG-|J&1Abcn03?1kKwI-uA84D|mSpZOLZ+i@`nacYs
zTiMRW!Tl0UF$8SSf*UN|eu}=YJxJuyHq4n1XTPGX0;3?84_8P<k=^tffaddmU&NyO
zvKxM(H`+;P^SCBsHwP*?iD)dZq&I>LexECn4~{|q^7~p6Vzw~yM1<%(qd8>)YKds0
z@V8QvY4~%n&;6oGOu@oZdg&h*bkQ+k7E%Yht8yqBazz&puN}`Be;sc~sf?aqZsZLN
znVa((y0_A?8WSKuo9wugQ1$(A4dqSv$`jIc(P9}slJL0t<$V9q>Xe+6j5#<S?UfO4
zB?*s0$9l!3%+Z!#7Blj0*Wc-AO2^A+KOc)W=MS^bQL@#*VGC=qXWwN)B1qK<u5WMZ
za!4IW3SfVPPF;Rk`{-?oeoWF-T}mX2&3cMoBr0<|j>}<-;b^Hc3}$QYecKII;Ym(3
z&~4@~JDZ^3A1)~&4(ys^-rPOx^vE%8!c|Sxu0<YzJ5~A+MgLTf^K<^RtyHaY<1-j`
z3G4GxT?TLMF7WyL{6q3Sji|7Z%hiE6Lzn8D@CQnd>4Fal+*_r(a52?Tz4r#l)Hox4
z*9_Ed8aUC8H^tu9TXfLyz^=**gB+RuFQNGI-@7jipM@#^-&nwioy=wQWu8!bM97te
znfYY{X;~P0fx=Hx%A*ZJg<q~i$0w)H;4}LE(*7@rL90LHo6_W@7a0882%oIf8^4k>
zBAn!b<(p}LMUK1L&#ou8E8z@&W(9tUkR(HeXq{8x^XG=Uv8;|ajZ(?WT+w=JhV(Jk
zFyhAK&76&`Uba$ai65+VpRuV=mEEiN00{ri(q#>pNZ!XRI2uZzIf#ksT>wWGfR^R|
ziJN(7q-|OEW0zC*k*_bt6E5Qb3)>=#=x0@@^%DvbKVNfMIaUQW`htEqYWR<Y`e~d{
zIX-b<)<AW5Gt4nvbBHm%Qf1ZHlmtOz9y0fQDl$%{(+T3$o8BGYW7cqG;U735-snqr
zFaDwAMDg^;>}200EXmRRFvZSbpgE%&sB&|VfJsg4HY-|lK5Wg%ob?d|=ygE6hf6OP
ztv5z9PEyi)cd{=5{7T5C>Ktd?Qk(7J->IsR8Z48m6kya2FD&r5zaup|<3pDc9<S}s
zOM688sQLzg;>PB2OUBKk1D^x^9P}{9`2EWlKc^oOMza%Lm37vO_5j{3z~?M4vi=^M
zYv;=WMDzxcTXt;M8ZnrI-p+9LlvJPyuN^F&YPG1S&i}l*;U;AEx|^OY1umGR8G-2&
z8JF8E<{8^duUY>eYd8T9aNlWf0=aNsiScu8O53_7*9mVQ@^BDv9|~(r`2OqoDbxXT
z?=X<FYR%jEu1>2-ce4E-7{o8Ym#y<lG@e)KlaLxoR1>04>Oqe{KDT(E8BVAQBpqp3
z=OWj$wWWcEEa?}qe1clVci=H1_sSRdl5-XCXl@0@qM%+g8h-H^ZumIvd~Vu?oC?l`
zG@l-D-d8v2^NPyh0GV}4c*a^X%dV|mRaGsB)nCFiE-N3X8%{`t*&@)Oy`Ew)kLG@J
zlO)R1bcbdCs+j`Ul+ZfcHlF9^!-4(jjFRRkwWN5pX3@2R+`kiq-8CGEXFK73vJa`5
z!uuVs0+P@jWx*w`=Z9&0PtWJ?#vdM6>uPgO8dPMtzAO_(IY^=)$6jXe+X!-J*pjDb
z@K{HD*I<3sUs`EVrbdhmWzd9pBvoKFK$ryudt`Slx_{&@yd%d3Ehop?_=O+yRPzl-
zwqTyLZ#}Mx$+Q2FHc{HAcwV;Cv?sWzsQzJ>WdcD)filpiX_2U7=`=EV^`7H6?0Fu(
z6N|@#*6|76IMQMN=wJd3$LhWP3&gK&`Rv!H=N`N*)6+nzt+m?%sHnE0cl4#MM(69a
z<Q|^0C9?%Px-0^;(mVJA%k|BLPGO0^@e@W&+9Uhw_$}1@{{2iM)V_aAt6lSY`FC?9
z9H+Uwc0Vm0MET#)U@Za+4NzPB1G3i<BFne)#$tYNB=u(3qEe^HR`hM(EWNtTp9MqS
z$-26F06)1qddNg4>}4Lzv`7rkmQH_Gswlw_Z{jqlRV+%uIY!=98sOaH!HqnJ#acB8
z_o;0)l}x4%pFW*ymOUb1bU9w>L;XT87mujw2>{7IYZK(84kcQ#TU-KJH1Ah&v^z1i
zJM3K69uY_6zi}vC@Eg~9o7?U^J2JudyPT{@Ok)JhCc%a=qp6tUmBtMFn4f$;A)*^>
z87$EdS8N>OE_jR<C?nmcj_h{*Cl|{=R#hQ_m!R33<;TrKU#?}cbn#iA!J*tVkb;_a
zL?dSjKFNNVd?v#ZUpyWkS1j*Hj^3Q#$<zsPPCw%7gbQCZKJy>sn1ZxFmO3E5>%jqj
z>(y3?NepNWBhAM=n7_c}TF#?AdfCSMbp_~IYB@_Ozb4`BySw<c8g)luxAoz+^|!0j
z-w2R~@91v8#T#aAVwml_x5AUqXnCNVcZml<Tn=uxw?GK}^B9R*HoVoq7!K6>9KRq9
z=NU*cn_YJhl~nlCW|)@iqX1;6q`-unPsKI=PpLim*l3o3*`>!>`@rTZK3vUxm!sE;
zhk<6JDX-oWe3?ch=eQLT*<_dHxbwnyxG-rG3nbP07(vM0@p2Y|+jGI_m(;WdTtT<X
zYwoplh$D*U(nR5T)+PPts}KJpknYm}3+z8`IE5g9QS$kP1{+Ifz1no)YP564ZT+L&
zUSj$hm`%+C7aLTM6e*fK?pBGXO<2~--lbE_`!}a9A+xPMJy8*1k!?tbdNY{H6$Kuf
zCW0f}=Jg8R<tYr?**ToW)?D5kvHcL`>4r<#k$CuB0fn?qxLxhZiCD{BB|oS82Rsr%
zU!oe*DMY&t{XgubvNU==gtAge?ctNf;A3b|8Y)`+t-%Z@W3_(>O$Zr?RYgAGlB0Dx
zuci2em?sx%4Liu;UCmAqicmpOoL229-6qM+2CegKvDlpC6&1UftFm+)Q8?bR_Tken
z*x_-~Ef{@j^NtfgU~c<}kq#%571EQKXDa`_{=J}cnCr4)|A-wKwn0KXY6m~J>G|fy
z%9GDOM8~5Qd_$8W48-nsJ$~}9&rPaTedmoq+=t^R_?bmW=d$9`jL5YEu)1`Y1`7=~
zI%mr{;(<a>x_@turbMfQ@-3R^@Dm0!5i|uZbv?4qe{M_L?LL>~=27#&UPbw(>d^#m
zG?a<ii>`!#&{rAYjyD%o`@a~kt7G9UvHTk*n=TyPpKHEa;mu+jcOXqd`3&{~P`NfS
z#|K(ktSGu*%irFN-<o8J?r}8A2g;8rxM0*%KZEPj_fwcVT5K;izp7WpiZ6HU1Z9rJ
z{YM%szm@|l@6THQ$&At=odC7tC=US=f&c}^4xo+4GlhR0&K$7j_4~tiDN#u!sgs>e
zJZ~scXec_y6aF5NlWyFVI8eOZ0C;R8kN{p~CITuKQ3ADjxb~{=N${KMMY7gT0RdQ5
zNT(0Ftf$k_N{U;is034$oZq$;GLH%K0wv2hJA-&1qF&AccJ|ogK*op&=MJ@Wb^n5s
z;Ta3EVW9XVGfY1`m0dw6y>FTK0MY#~m@r8Rydw$X@FR1+;Hgw=!RP%vV6+2DqkINK
z7!guxd788N7Y4RPVy`}2pG#{Ti4<xqZpxz8{6C?Xs_jH^y}Y&gcDAqp!LDj!IKB&-
zu#(!;I~O@<Q#!InYQ))i|CcCyI&b};FSgnsRj(ztdYQfY{U7XZ>+00i5Gj~<$gH<^
z&&Rv!;w3!8pSV+P10Arg4_4fuEsp>0^P)s2jgeq2>}VfdgQUpa*SmnQq834A4$9W#
z%$P}tRHYJ?WGZI9-ZB>hhH2nzrArnjmT-;b0qWUlJx+TG%?KsPP`)SI@x$!JZzOG%
zPR@*3_8xDyQEz-)?f94%I0}ZvW&W~C40;B+TyBkzJnF`K5-2B&pGr!Wq0?M=$nPpk
ze6Zm(SYcqZvUI9UH#XtGNMJkrG$59fkdNvr`j_>0-G1W*rIBTq!$Vyi$>$6WXz{ep
zPp{kKKG~$7E$#{N=kdb1*YE$Q1+W9);cT~5c25$3kfWiUuMft<<xcC$lOlzy%@J2@
znepjxDwoq^%gvFS%u=(({1ASrJ?7p;?$p6xA=33pQXHJsogH}*n@T5a3tBAaaOy7j
zgt`iq5I_8FP+<PK@Aha@1&YJf9<%FK{rnYbq`w=?fuld+?U^tSP~AB2EqeJjnRZ5V
zjcc<FlxO9~i~F$Zn2$RD8m#^LwSrWWtJ#pJY4@Do;0$fB;?L$Fj$3)p1o~uol!#xO
z!Pmpm^><0V9nG{Aqd^mKW-V~Rp4Ry&*bXd;%V}e4Ph}{O`bz*-z=sPCqU=LUN&qjt
zg;rn!LpzU-E4MDV^}!1At=XERIhZ1pK^n5Ef}YZ`^V#WY?y$u}QD?4&?mhkhJ~j)Z
zZ2nPGB{K0=_>^WtufSn<x+3oG<IU;c1*#}R@*x-qX94?BLH$-(#>p8(7!MiK!$8B#
zjoLp+eB}-HGtgF9JbTyyGFlE}WvR24@j%s&2ut!tscbGK9>cb(o>a0#bW!2LHvnt4
z^KgIU2Q-+!julN%?mmbJX%yQ~HN<4S#}hkx2{|iGXC0&yAOFa9<K=lbULDI;Me6+<
z<xS%63vNR>-EVL!t^}Uj6~u!j1Q8kkVFaD_r(xz7t2rH7lJx(}D~`RYbQunerZfOr
zw3VWIJp%DNi;;N2rZX}FB8Ll`BWg~tQ&Te})aD*V%3o1Z;=Hm@og;ES`(Av}kH5us
zxQvECdg9`7$`_f-Ae&b9qU@L^__Ki2_vomEL~_3Ly%vt1_15!qI4m)ls#UL{PmvyO
zVn7*0*`k%p5@PQj*8X%4yiJvDk4pfRN7cIC-kr{(e;?c!h+(lEK8nH{bhQe)TD`}5
z-U9KCppbG~J#TM)JeoIRNwbn=XNWHbQumvUjv4?>sqLSH%xaaMGC<1T`yg3Rwcb(L
zkmfq3I3*6twPv+gDJ7<)??*8r{o?7*&h{n5c2bL53|XAylfLDmxK{sGSAb|Dg~Y}z
zUE)zzK$1%*F3FMa$z5ZP`=Kx5&Kae-Bn_Ku{N7>eCOBWJ`0JB1Nna=xD1c8}<Q#oi
zsUC3`-B3FS-eG=B@ZisCght9RtcjDNQHE*8S$Jzkb%u@x{1N}now=kW;!rp6`2m4<
zBSGj*JGA6q-HaxpfnYy~U>w&6FpSQf28cdRyHL|v`2uqls<%|d@W!a$2@V!jWSLto
zf05y)Q(xxB?iMX!fpfRQ#83WvsQ4Q%L_%)lDs@h{5lncQ_;{)S2_O}0rO7M~#a^(i
zN~3Ih!Qrd$6+o>vS}pEkta+z9H(@H?gJ6`L=qw5kY5bLvF=sx7bs_IhTl2q)j2%yh
zM>fnq02PX)xduA<>y4@Xe^cfY^Q#Kp@g~Zd8h{QELLWdaS65N>dbr_MD8q#3mdB;G
z((Ql&>U00~0fJPh&f#2})^fuv0}N@Y9GY)>v#`ZUs*_Y`R89B1odo7RW1#uGMb1uw
zDxg-S3OXoQ`J<s>bVC7RNZa!?K+gvAYASp*%SLo-XOZ2+0GV$H3IsOxokT3w1e0tj
zgR;&s-|qVBRdx`g!V9|!kL4bX&KiDDSZ+8wzkTr>o&%`L&bE`42w1wdNB@g0fOZSs
z=wS_?bK>+k4577S12e!UfqyN!xgj)FPwK<amVw)x0S?{4c$}4SAdRFAc&@I)F%Sa!
zPDOb|XE3DuZ_UDio&h76vgrEh@K0EHI%%1;L&l(exliFN;he&k;Xe0_Rve@Xlgv~2
z6+}%EZ*)DEvA4=L?>SNvN=mqCD+>dN;;p+lzL}Ijf{-0x{io9+cX&z)YO^hZedpxq
zASYTT7pJ~m%NG-l(+V8+TiM#g{t7)MFa|eF6jXsx{%IyFe`#2z0>7A*s+SM)MV}Ce
zVL6CJSu+Q~b|G95W)G8O#!w6?7s`$P`6EIXyIbccS>DHK)%Ypd<+<#XM}^bR76opq
z`MvaKe@Z>eZpAi%`ko)Hio}x1X$=L$6s_zjBpKr={DKSXcChq9h<01iL1&xZSH%(M
z;z8&q3mcPg;6=J_i?VS5;SPk8)Vf~0J?=-LOs%F(3<#2jFG|8pLDqOdcTwWZJMArY
z*z5RrF_1+^`gHp5O3dW7sl4z37OIycCf*Ddpyhm^Jb$%+ocAPDb_kJ{^P45#0{1G!
zeRKyfqwIajHu^R<4S=TwH8wU$M?+hi7$<+IWXVIU92lkK6cmI8#CeZaTAzS>V9hvC
z#gf$DU+>PAD}e`2Hj&{bHpsVIKPe0~x*LO8G@g*@`$~aPk;}OD>X^l}E?bTWi<eiR
zi-&x28oS5o(*v695Php$-o3`)T=}xgNg&@h#f-2vptd{Kec<lM&M)icM@}{aG5r}E
zEQhzbM!eL~2G@Ih-OX%E!4Xw-1R;M2;}A;rq62tPK=EtM=t*Zen|9+FELJXPAqNc&
zo&Zu7T1KXLh-S1%K|}g}UGcP-&pN$dAdAQ0W0}peTd~DF0Jbd;*x;n?Mr!h6a}yHO
zg>L^)jk_Ff^O<ejBh@-2GHX_w*wFaj(1e+}6#8Ij$4AS3`SOJ`jZlKPJ1&DxBVZX*
z&M(GL+&mX3Z@M@N80h0R%l+YfoPHhLp>-k;L~(NFM3IG^@Y>khHiILq$z+QZ$R)=N
zE7Whu*;X3|z6=?Ur~VtOYV1kxWc!_h(yUjm(SDgADxS!!lfpr>!Y`&O&vPW=40c19
z&IK_jsJVXl*`ux=)TVb5AQg(av+Zs%zo8iZ^9b?E-$WfIv8}a~eU<8QNe;;b<)na-
zJ{2B+AxDh=d#`9HWJ!0BN!q^Ek&M0K3S-9$%|FY|2?}rFQn!edqBW~J3V-D=46Y?9
zIdyVB{K$`5ye{X|=rmnXQY=dEN{5I=y-(4srF0|7Jj#)$tBmC|mZH?rrE&qt>VTNG
zf~8&o+$B2;5&lsZ@P9iUxz8J=(_g(txiWxduu~_zvmQ990s!94-r{o226-Z3R$AI5
zLVO&#KJUlfU3~QYWt+?VsK|N$HMb)`59hgUtCsEn01NPq4vW1ppk}S|sOHDK3MUUu
z#YB%p9c-q5YQGA$na`4_dz7I#G(RBV4txb}MAWB^knK`nK(sUg2^3I2kyQ}FXl=e`
zu-*K&hR#=BU7ctPx?dkoZk=y5b+JwPSuML<yB7~wi0j<c1hCMKWOAD5zQ9}&GF)4|
zVju&_2;O9`KaHS66>#+Kp~CWsZCSqMvv0swf8-W?%Ev}KraFuH%rdw|YwpIb<59An
z@SyA<-VIl$lNhSL00}9m&`6l(7tJ^>hsTkBV5HXNt4lQgx?e?BrE`EihAsnR&N6bV
z7a*_DQF;X3#egkllOww2NMg<c^{C8J^}?Jael+pOl9rUXxV<^jyT~&u_APGdc6qrk
zE|6x{?PNJjS-43emGieojbD{9z|+J&ccld(|J-nja#8Zpc#?6$MC%-eDD_S$<ax(M
z6}dix;Xo451j4RQYfz==nb4uFCP#aoMkTe{Wh9blw(`E|C4fj$p^9Qg5_xj=Lp(iI
zRZ^-K^6EJTJJ^O%RHs|;^I09v0Dq(H9nms5;;q4t5fhsmYa?+2lS^9OWlo7!m{T0*
zR!x-}GlGUpEQx8GLzH30T)S27xrc#trR)2r1#{y2?EyQ}InqrqxQK|_I-2ddHRgYB
zGfrDb`DKHB&ejTbARbFh+Se&)$Yf!JP=g0Cb{6~<!=yooBIU>Lb*H_H4tte99%4Ao
zJMO>2Vo1GGup$nWBJmL3$=}t08nZqovJN;pUgI$Vqm9Kth#6lz9szZN+3--<xDa5j
z2<i|NIO@#t?asv?R^KpS9gu}WoB_EVpv8{nA!nO?vhbn=M!&1}P3f>Lp}9sEtaLj5
z=FS(BQ@pi%=pGIllKqlr$~VcAXU=C+K<II~3sodGD!iTlqiAmcnvXcZK8y=#ei|$(
zGnFG|4{R0S!9!Na;(|-y@Jn&ENcG8ZxVprc?w-)^vFEZ{B147Ga=m879$G+ike7-i
z0!%G0S6#oVN?05yYB`{uRZuUW%MTFpt3ti`RMou9sL{&C9M3=bS(g~or83c{($Oyl
z*`H4cEiQUQ;%Lt}vTR=3q1cdH(tnB{n{#Bppjsmcz+Y73fNMM(=3{XgD-6N!b9W5t
z+}#;Y6rW*=i<GDTPP$<`=JpV<s)t23qTT8`m?@ajNCn!8M7^)3E!@G$&g3&KvFl*6
zbvIC3luI?9D!SX3uV-xrl`Y}&G|(06VjCMAH<p_te)A&mi89{BkR_Fmh*~a`&$u-l
zpxqrw#c&t8$`v=L-5FLV<4HdGy(u%`WB9Xwm9xLv$8q4u7)ohDV&jVN#f_hojjR3Q
z@d%Jj0+*oSU8y5Vtx|VMfT@ASK}T&$d`ps|E}C1*#3uw~=(jdjf>b(yxa%QZ0uB%;
zs>l2Rsw7D+9k|*~S-@;~U`Q8AuxtAu4z(dD)TIs_jUwiJJfO{7X|O^Hg9d{e48W7N
z{#krRhYgI&8t2Vc$(T#n0Pz@{&dmt*Mh?Cw#2zXUJ%ktyWYea_;+FF}J&pj+9-t1>
z)`)koKhFS0yH^jgbR<ykwwor@fPddhQJw#dAO+k-r`reYfmZY-?+1rg+rWN0?;LfO
z8?qR+yD#khad|Q0x38};04fO?!FOEX0k5W}r#T><0#JFMy&l`nj~$Mtr?s3|JF50)
zbHuZL-7X$3AyX~eWgQkPi<FU0G-l-&bt<T8O`^+lKVBaLpe*TU172BcI?)u=*8i8$
zA9|ZM-m<bM9+iqjRfYhkd)sHJIN|TLEjYqSc=dH<x!LE^GG@(DKLN*Bc?C*dSx`*H
zKC>4<IhXLgqsE*eM9;{r*E`*>uKAqe0@w&rQvSM$a*TX<R^_Z;*&Wq%IT`D<lUv3y
zs^{wd#pkl><F~l3iO98utvNhK_(?#@K$om9)L+!Ox^b6aE2|*)AewcgtfP&!zrD5x
zprUg6U57LE6hzp{4kaaw61Y7nMo^@p65218j=bD&D~ifO&vT248nzZD_ctY~MkZ_j
z%C?|f&thn0UJlLxbcxks3r694>FkHC1KNy28-vf$3zWfIfqH(p%><R+5dqowUtJxN
zapA#6_=V}qc6L2I8!d=6qH6EZGiEjTTwvAj`tiIslJPSH%)UXm?&OGsyM*10dqK=|
z7{pLJCs41$0*>6$2VNsvp<zXDX-j}stVy;yXO}v4FU_?WD79E#dl8|-9$}`dvXq+k
z=?#tj;d#AqX*7%lH0HQLuehKh`^2!Z-n|cThKc7FZ}(eJ`j4xmVVjSlQ8T~8x%43E
zYGn)gd8>LUceP-3k8RL&EuEb+2KQeZ>d^miTxca0AbHo+s5Rw5+D>J9o*6BJd7bHY
z@1xqjCsx|xb~%%RM#v0g9{+nSb5bwf)7$?Op@(063Q1aud?2xPozPBZ;GD?=8X@3<
zGFw_NbzEQ1QKeTWAL2z1nagDe9cnn?k5&Q<g+0w~q(B{NfJ5)Gx4!}^(8GQYXxEGH
ztX!|hSS}0fS>{jp+a&8JV(-*%`iaCJ@%R$6Q3>%|6sq-lWoify6$Ke7%{JZlAz5Ao
zjq;)FlxnB4+-3P6O0;XO3_VXSkr2>}(&B?9l(?+ul6Jh-o$ri9m)H`~6mtH`)C_GK
z(IbqCSa~jx+7u8Wv87P=S&H4ubCm{~iAQtUzEcxI)q)%*NKPYqm{k6)b14esVm=%h
z;Qwf2w%;#J@z%A+`rMK~*~@gYyAJ^B7q@tZ^+#b^$tvyPi^}r{fxk`NjZ{aczX~iC
zn%HEuo}qSQqmH2}7JAQkWV0}*j_5#Zz*XGS>Neo)v_40_LUm<za;z>#gWwt_#*e>h
zws6Ze{ozpvU{_?p-?hP7q-GrCU)hAMkzxlEEh)NTRepp^bK~yZ_1d}O1n3~^y@9iJ
zlw$XZKP=d(u>NowtQi^><$FC+c8mZ{FmyUFkp(8BuyMgXM0<m;m2PP^9ID(R#F72E
zlYBk)D-L%~4r_CJ(X@K&cH^!ZpLJzP%NB?<&wZD+K3aFU=VCJLtkPWojBowX9P{_4
zWQ4jf>|!j5zGEC4th1-5Fu*uj99T?>Z3eq|tV9=Lvbf$8MNin2X;d9G;rPZ2Nu)F=
z?UVcF%z;fiFJPD+?s&dk?zpam&EWbUCx!mXNhQmQ;=Tt)&KDHe`HC0M(-qzx7OZ!C
ztq$|nuhw83@J_jlRFtT`fLFg<(J%s+u{|JWzV{0Rh|Gs8LUfaZMZl-L?TUhM2oSqP
zEqX)*@K!ZL;|7vZZf>2!fPKp6Q9RMltvfm}LtRlx3Bj&)WRxd{m*o>$K4pvx^*5nA
zF!_~qmqPf*LAyInE(g(XpTP)YmBbBjBsKX(mSHcy2oBK-<a0;5cM(vo2HT{%9F5fT
zu70X~Q;QTb37-<@Ixdh;lR_u2Hk&4};cguLVkMqh2cb6eOTr^S&w|I+`6c6{`bIS>
zDAGfCM<VzKV$MQ%5{zH`sVRv50G_k++xb_SkO7<zyT3iKe5|4)<!f7LT7{1~V>S&|
zF?o62?w}4fblCpcZ;S^J-0O{)A1uuP!1CI&R>$-d&Oe@<7RHuk%eyQ0;`=MJoFTFj
z@Dw7z94<)L5xWHSNvg^deUTh+G9IlXljjD&ps7d)3;<v=$t)osf7v7%Te!!{{k*fK
zq_@PVy_{L=mOz3d+qF>=Ph&(EMd7SN9WZpgzOycw0V^o10>jh<7TJ$gFd0N|sm<fo
z6PUH9<|+&pCj$rPm>0e!2}4#PNl}fEJp>G6zjrP_4BTsWIGUe#PX9QXY4E|^*+<<U
zJ9qYdq}v9S*CvOMloB8ks!vkbkezJ+40{8ho9<kyYHzd^Bmy3JslE&I?LE_ng0!E?
zZ$v`}8hh%mMnu)>acRH4dTL)qu{Y1c5B%a@|5c-Vi&q9?eqB26KY&JcKFr+lr{<VP
zNsF5ryh9y;yYpLC<o@vbnmwH^=ZctCx+1mB<*Xg>u&S2M&3ta!G#J=~byB@qZA;k_
zCh9O6=F*PGG9*vEJ9rvR6S@;39=LMC{ga<Bf?x&>t{p&T);FxrJs@%YF`yx+O~+vP
z{i4boJ1<+o2IQYca_)*zJjvIqPG5p)S5C3RTYzzCKq+f-F_nGecEft60)|o<mJ}B;
zn<glz#NtO9<%iYK`03T;>?0PKV}S(n<%boWof{5Ba0=WpFpvfKK_f?jBZg`)e9p3Q
zwxZvN*J?d$<vmM_zN5;lEemscm?%A_!phRpq4#&XGN39WC}`{CjPhqM?{?XfU7Fc^
z^w+l)ns6PST*!E(k`=xR-2fKBYzFq!IM%a^Q!U8q+U+WUm2@dsr;NdqCFb3ycFNou
z52c7Tv{g%!$#mAbf-G3&s1DsTAC&=q`WDJ+u+Sn0ci`l7-`fMzY*L?pboOmPcJ}RG
zT5CNHQaLC!0e_?{ob_if{a<GUBLj%m#l$#`yIh(yO+bRrqdD--n$GWLf^<A#g1EN6
z*iTNz9!wisuC{VYh`Or19c@UbTVXLZvla^vcL`kOV(vc~b)u39;cTRh{|?u9lY9br
zr&jz&^q6Gg`8^>y^uf!k%&vVGhoM6y_uj2VX1g0bZU7gTsaP6#a`PVL{zeo9gie8{
zHRmV+NA45{eZh9iNke_wC8{YX4FtQqJ&gW$jcer1eW$}nlJ%da!t=UmruLsD6VC^@
z_VA<0@u0VOI#BN-_6eY##C&vt)20+Yl|19J0Ctol5Of(6D4H1l7G|XS>w?$zbDE$B
zCav@xE7tkRatE@MKR`F7{H+1ZE_8Q2y>!fd5mj}1VXFHf>i<~Z)219KDP9g#b&g6;
zVt@lI4e_Gimn??TV9*B~|8T~;1FtOWa+Z=o>)IU7okf<id|QHJr{ZajBmFjxWD56g
zyWDOM7E7m<ze0?MPYbQ#l8IGL%ty#pB#>q&(CHH3t1A^oLR-Kq;ic?Sb~CF`m@ajO
zoo~L&SGS9hw8;!SZqZmH`JitjH$E$pZCgM~>dJuQny~*--e`N}b?EVTskBmX_9Enj
zzx|C-CW1;@qLyub5ujLu$i2|<WnBo9@EybiK2l5u44;H!pcfH<Cv9j69CE7@=+y)t
zTn7f@X#To$-4J-fCx6X!$?^GZKj78D3BCTg9SNsx7Eaef1A8iEO$Cd%NjC$x(_<$h
zh7^hgb69L6t0C(vW?n%HZVhinzvQ}NzfuV-MR8#A-v8EaH(dabfD`|)T#GG}am}d&
z$aRd)=X(=7_1|C?9!ee}kO)YDCmE0kX5^4|4U$b)-(?ks!O^(uK8oiuot*T4O6-E?
z=-K>C9Pz%&-t}TK7OUUueD}!48wJ*ttX~ez^abFn2$-KoI1&7*k%QZ1Vv{0?*A>&{
zEaFFUN#_75?(x|pRJcR@1#_zIRfRwOPm{L;9WVg=DZI}a^GkGM;#{8QT@=l)57MaY
z_(>%t7KLBa-l4*K-0df%Gz(J(`vR<&kP~XW3nqf^MzHR8fF>Q#S4XPF0)6B|C0I!n
zvOaFYqE#3g9NbFIOTZ0F!Tg|F!}dohUt56#k-Ch8(5-~~ovgk|-vy<lH?0v>E?Ly>
zjgTC~>fKakTu_Dqc>$i7pt%d74%OH;E}p9mcfWC*#}8GqO6_L*o8Ajt@HG#Q%~qrl
z{y-a5lhub58RD{KBtL&*Rq5Y#2EWjw)H^9a@D63*KBk}^%A)rc-{bKxeM@c1S0N6a
z-Vb<9&4NC|gH_)r|ImOa=ietng!IX1{ETup=j_x<!Ztx<n?W%{G_p*8R+47R9rg<O
zgkT7u+5J6B{?-eC?0O%AwEXqicIqVhr6l#7w6g6P;IS{!(B9+btfGO;Xa@i>o*<m!
zR7=U}EZNS2n?6-+Or8g{D>=6R8@K!_Mnh*W-^R{Of}7%Fr@98c*{?H6%o0x@Vu+ro
z5brXy#mM^~&_sumDmzPDbIU{-8gV1~^=+^`jvh?1`X=Dclsv8EI*pmUF<Kfa1{qG@
zD2TGl*?dDO!_qzPa64+QKL3gQPA`V03K6g4NP`uY0r(u#bU8QLkS@pKaJPj7JpgLf
zOu5jFne{dV`Ry*v8y_1o3>bUCD`Pk~uCLcYWru(fyxsO<3V_Y_Nly|v)Bm%4kG(!A
z8FnC30{_kqtk+vG2a3+OxdCBgnSw_%8HyOlgllOjA?CN5J*)cfCv?+#FZ2N$20)%V
zJ4cp=e=H%`@P|U~bo-A}IW5ktpu3BOnSm45EN)^6_-|`ZOh>9n7$E0vD$yz|;#^O9
z<KYT^6Y)i?_6&a9c64Nl#*>p8@kHp$6fl}TO?I|dbD|XnqM0o)^}igW1sOq<DL(!z
zADIt%u#?7e<@Cb3%5p?O0bE1*x|#m<5Xb2XW~$y$lmqNaOa<#>LohxE!W;wq;2lFd
zH#<EZasz6^A7+M2q4vK^Z_(h-QS{q|7|QX&SFYiSa6NK0mrtrVl@gWO_PUbf4YVg1
z)Up;uj|fyY67eynjYQcHPTB_XQP1SE#dx+u*W9(<+==y5`nnL($3}>5F=Pr94-%Vq
zGz$Mr#+?|@RJsA|pxcu{{NXR-u89R<tuwLs<ws!|9&IxYcbRf`pKay$0`@XJw!3cI
za~-Y8b*Ny&J~x0$;he6yhLj<Yfh1DPb4C~L+~;uoZuDvMp|<qjkK)D6`Si-BQ;~Q}
z-cxfCxwP;y#xTf{C7fU?Q0yR(-10Lq%%6B<w(CB7vW?a8oZjkt;9QxE0X@i$>ANcK
z*q;Sgh8#NY5dwRbq+S;9TiPjgW(MmtRXM*boIBE9Q8n*KZBxXs>;8bZiVZ2uT8DBN
z(q?)~?few)EP!1$z;kVMI=<QI*IAR1te33H$tjM6l_G2d^Eh7p0cT<ZoXx-Rr+UdC
z`U604kyScj-CAlyqtnL`7l2Cn?M|N>=l{Gst=-M1j3onI?AiX6lHZA*@&+`XXm}se
zj<tJW=Lp!OdH}d8ohH{}fXatMs!_^gV*0p+kb+CW_EgRt2t3v?JyL&0uw|8e6wLyg
z^?C}tz)0rZ>0@*p3u=A%Vgn5Z3Uw7lMcadsU&BfHG3Apbz=6Sw@F_~)D6Dv5@I*f`
z*UH3ca9XNzbNDIgFYEDEFvyT53z#Y+=0T$0p?obZ7%fY98!^PviC?%O6!w>d64%}<
zYI0R5ouUg_GVWrpLeEv%Pen!KTM-3Jl!Pyb*HSp}dv0{4PX6?aAK12XrE_^R?z+o6
z%$tNKAC1TC6onj%BIMBM{CP{l*!U+YLsC>$@h_a-u)yWOWNrRW)rPLo1@g!R0l0=Z
z@I|;S88Iic+K!EWl|Hd$QSgtC>zYM-faOoETRU}C`a!M*&HD2q*9CTa?-oFW{FHjU
zKHT~2!?KKN$a=gQw^;6w1Uz~?eeXAs^rtMQ5-??E!CLgcJULgCOnK%xG;cvkixJgB
znwWE9Q7zq&N|VsWnvz-v$79oxtPaJj#&mrb)#<MEbWs4z2y1H{IW;JA{9Er52evz%
zh7;4@yt|~+=Qi>QyhYu&Ti?1mz2*^B;l!;yy$EyWcLbTTL$tLG++5_)6@#)q(+_#W
z1|fs=SvyNQ`IiA7ENhWAd#uMW`wOyk#~lnp;7K$YEDG~#z+3J3ylTGOHJRtCNt*xG
zjjXYNZL`K8+#A4D=JSF88%HD8+UNw_LGnGM93o*J%(C%Gtpm8g<y+97PmH9nIgOUk
z2G7ON+`NHgcq@I_e1s3!@#gt-0R900h{CHh6DobLJ7lj+iiGX}t9DnN=OhOcxMz0T
zM20}T7;3Q|GG~ON4bQmr<&>rmg{=xwFr!8DiR*XJd{|F5r(0H`msR-FwUEqTzBgw^
zz6myvzg)nHw?f)1_p1nQ^ko@V%w&(gpcVvZ0QQ@P{cQ7=f}YsVc~!Q!b37`ey+7_|
zF0M{L;Fz$#NAEgWw?Yo@T%&O#FzO5J{zLF#pDmzDNKHVdfpJSoWXpzF>}-to9kQZM
z)*^Ilv&9L|%&tqqI0t%~^5_}~F_`Kn<aY+%17Imqd#x;>EZuc)Pe&vg%FC=%ZU6j)
zz#HVSNaCNRNDW$p5^YHQk?XGvG2{NN1^L&8fe$|?sdfcO%9P7p?GU@6Ko=>@b|V7+
zcN}Vytg2l%A8bN9(cb1W(cY}asrFQXb~9G#oEAMCc|zgIYV-3gFN5MF+u3~ysRfG0
z3Iu<3bpr@ssUGsax3iI7mhz40Dq>}jtEE`z4J-lE*vQ`d=M|fxRTdtsx9xDMUAanX
z_;PAD`7D>w%m&p@SCYaszznHjn(ZS&dJC-gAYGGI0q=7loE>cG3&wCTY*|@xZ0-;$
z+6c^(z5+vkx0QxvB#bdZpivgE#Ml>>RH2o3xb_kz99l;YOzlsvl+QnA$hu?8D+zuM
zEJqU%;936j9DcpL3hb!@!`VRr=PmRuT8sHdl_E?WqpaF>H<bji)0W2b)0a;Ci5(3K
zQE<`GJ2-SC53SV7&Fy0}LC>S^3KCFBfgKwZGz(E)Y7lHmwSKvgILAkVZ;1VwWi42w
zg))2*2dOe{IPm-58>~5td5`#19=<%<hBwTP$dnF3weBkOf44%BjLLp*505(v_XXyf
zW!Wq)6&UT;tLt2OZF_m}i+7c&R1JR-q$f*!80_o0cdPP-MYGdI2q)<?_>psnFVL>3
zz@S=Q;BJLWVB|es*gYy({5V#OcQ4)ghi!WzSHy~c9VLt8n*hA|u6!R})3hyVxS9A{
zHZ%n!_mIUuG6+*p4>o|kAlrpk;g<jx7w-UKi~Y4m>nS0oCh3O56&l-}2`68u_01_v
zO52rC#@XmvMZ`J8`K(VDsj*|ZQx4Tetm!X~`TU3id3cxS`Z;CUeRfC^L{=F?+Si!d
z+j=27Z#?~qLV7t)dS#})Nh?~lO(=$R&MO-e{E6uVgyH(p*sm7Zxy%reW+TF2tPP_3
z7x3u4uB9wGcn*|AoQERDtY&L@M8l+zkgZW578aAkoGGna7<IM0K}<I|SC|5eN^p%F
zRg5p+MfzFguMsSCa8B4%U8t+b2T4a~LUn+hB8%xcsjfEIGE`t!OCaFwkE@5?!91M2
zN|IKJLBM3N*U{#UBRBfc7|>7ipItR8;#Hx$y~$g|g>}J%UHkd>r0jrQDF#i7N88_g
zDzxP8WVR3xsfLrvMX15tU<#%<YYNviK}nA+Q~WB7z;0<^mu7g*%3#gw>ziTLKcDL<
za5=w&wQzTNz_Zas9(R3G<4q@OzSf9xlHBt2sqqT*B*Ez+)$KeqY%-p25nQfa8yY^9
zDq6`-%MTR$`L4Evq(qypVI0gcSfwTGm-q!4%Z(`93fl@740R*71l0FZtbJoY#O=ly
z4ypDP*JarIO%Is-rwbT^u=c7y@4N0=Ia&2e&8`eMG898+23w$S+|ba|p^WmJEK`3I
z;aaFiTU&jT@F)=6;TTOeSrlQ>9nNFg3U+OWo8Fv7Mt^xT_=K?yFN`m-Yk?@!&F~Fm
zZb$43KlmWTwgqA|OPkzvmAK0XWk0hQb!AVdHn0P>8DQ91O?I&{A?^lY+tX2S6H($h
zpDfktcI5(k;Q2?o-+^X)zp0Iq+kRbZ6hjEFB#^AGxT^a=Myd|!M6J0_jSl6|9z8{l
zB?6H?nEtGk&u8PCx2Ky8cIBh=fson0*6{h0tOeVD%eutUftO%g1B*azElRsez|{!&
zw)eKT19(N}SWX4qpIn{({v{_8I5%P-OahGg2qSX~sM&<{(HV>8xHNH)In%N?IT#k?
z5z|Z1$rZ#>)P6@Ztbd7y>$6Q&eXzE2!EClq8*t5T2&VDK;73bftK3H%OL*ZqVqW|1
zdwbsSn=qEBLI}B_T4GE!tDphgGS1YJDrjQ4$?OWy6cjWe3i0s^A#sFEKc037tF+PP
z=T>=trTu4f0_sy1l}P;Gm#4QAuLjr;um?x;SnuvHyWo)8pB^oGLWe@z4F}m&YJq*{
z@3<Qm#Pc;7sY@&~K@oFqJ!wjgmqI_x0|T7Ow{~X>v`BZC8q5JBJaF)CU&vZN_oKi$
zHH*1nT7+fMk<quA`YP!JI{JMR#P<UuO?28Z$+K`hc=G4_Y(CfD<q#4$#N!Rghm9Cu
z{oC<;3+coVokB4?sTpLsEzeLMmxyjoR4ZCRl)#?+)Wmyn;Qa)AP*i?;X)qFF=))|X
znHX2&TTb0wj(qQaSJvVh)_1A?D&ay4?*OT$5c-@!F=_YUQSz`_<7`SIgbbOk@IFY)
zIW`<_^rqOhg^)<vX*yp!<J|{}zzyFrX>f32pyze&dliI{+Ci8l?b=)_UEqJPn;$CM
zd~J+)yLrfICXUn*o7oOC<(=Y-y5wpAxVZ_H0zW*^D3czZJ+aVk@8Z8Yc16~n*Hx`G
zoXf`a==s@=M?5c@_e|un4kA->eb}!VZ9+x{G1{vX)mma?Q~2?iPMdxoLr+!_TfFjk
zj<p?in@5I3x&H+&uX<gi)}T65G<%o?)7-Y#8LqYK290@6Ui<eNqw|VVukn`xW=o;R
zi5}3Bu>UTywWsS(J*97N$55hktfUMQG0YBEi63dU+9iSnrfcE?#ex;zfV!#I5n0|x
z4@jptVS!l`(Z^q8RkcVz(lG!|fG%+?fr)wt!f=JCbA4b-FVm}KIugXc;Zf)v(E%!7
zhPdd^VYhaiwB7&x@zQa%(JKR>o?iZ*jilP=ozDifCSxtG2N2*I@d^3EUIfkmws3S_
zi&_VX5=j(x6DiL9L1EU)G;omgFOE%z-?95)yje{FYt)XZp3>w%oX}G1g~Kw80i%qP
zmYVmsx6w?vpN4#uxef3N5XQ5)J`?)suzsnz5cxrnDQI%L5*{A09~{_M?K#5U?e}R=
zAWk1UJ*1~Qj(p<Hv@;oP0ZBn6C5PfiZTKJ-cDLJ0#*}<x*pFXCmy??^K4GWIs5aWB
zh#?ism@SwhH(EY~a_jsucvZZWA(c5|5+&Pi2i9MWzha=Iml)3!n$H)u*Vgo46fed4
zY9D>XwhfX{KJ8UhZS-6A?n5x%J+_vaT!$wH!!wK3TUAXPT=dg?F89lc|JwN(c1}gj
zcGD->c)>98&25kdCW2*A&QAzj=C~xTOT5d|67tKUi7a5COw+OTN06qaEj?wz$&4v;
zXI-(Uz+C)y?M$MpKM+3GmN5_knr|^02CReGZ;$N%0z);UPQ#{#4ETp;0~zzib>zAQ
zRpaN#P=xU~Hpw;Z_QqBUm{{kTNM9f4!?d+V=CGUzmN&~a8%SL{Jt2Z|)-WH_9$J<a
zIBr0io@Kotn5~{Q*OMDyOGhDt4Ij6`u9*J7o|O61hz>=`ilk+$aq^@Fx;&XGR2oLw
zGo-7^%L_=c0JfMi+oEH}YSwc0TXGMmR^I{4%^LY#yWZk#z1}peIHxN6JFoo>Ain~L
zN&r%2j{D`)E;!Jl@Vq!4wg6XK?FXDc-JK_Zy=RJ@%kXqqKj5Nj4(=)0WN%v^NtJud
z^I;ZC_($`_0Qo67jyXVD16w^d;s7<!aS}TH@D}z34&HjDsoI0EiQDP;d_M&EP@)Om
zk3BROW&uCYLah=eOqbMOng4qs@rAt*-N({YAKtVAjOUg>l;$KRaq-C!RoF(Wmt24s
z6#hHI&Y;)S1x3X-n+{p3=8G$y&<*hi2cLVRTt-HL-)A6-)}vg;rNZbN?H1%(P3j5+
z)Z!(gI`~OIgX|l-&|+bzlb$X{tdTq!5XmbagT9|h=_&Ad^4I`D0TL2Y^6WA(>^dgl
z<kseNjqJlp9XF;U45;QtTLA{eOP`t`;;r5W-}hzrh-kK_YO|LXPNG8}y<e6w?B^W+
zbUB|<-!Z%;rF0kXs==Izn3k7Saw30!aT1WUKzet2wC<7C>&Ens3A3w_=<;OiA(x@!
z{P+qWcc+`Lyz)425<v)R!7C6g{kJTDgq9kEhrs}wMI><c?so~zk8f%l?HE`3{#(@)
zI@;Qq>P>*y`{mOH*YPk)Q-SdueFYg-1$4bUI=j;SuA9pbWKDVp4;5l7JMi&cnrXwU
z<gn(KsZ2$GHvI{+0wN<l*X;hu+N$HJTR5`B==4x=|77IWf1LstIYy<|-Z(uU(kcVX
zPsTUjnT78_sR`n69bV|mK2L6bZXLDtwB#p99@3<ViQjr1fu%3CdC7?h%pq&EDPY+P
zaE!s-HH582BbEgP^6G8m_ElneIrY`xx386!q}s|SA3<x?(-L$3{l!YYX;08oD5{8P
zy4Z_ue4}?IGsnqtMYf41aMMZEYPvMWZ8EOa=6v$1)sjDfax4Dl@4R|l4H@qUbVLL^
zQbC<N06%H_z!nMUa^$<C8A0Z^!~c+GlE&>AXw|gs1Tv3J1_DTYW+UoBiU#P@YAP7y
z)Lw(R8tMI|UR{xas%mO>UyEi-#kFy=H3&ak=vNqj%VV1f@`?(%kTQXCoP4-ckpa8?
z?26e^oz*75+4Tn|iC0%Ju;5UyZ}@*$`o{3Ox~|>Wwr!)aZQE#Un~iNdjg!W<&8D&4
zn2qtR=RN2DzV@|t*31}V-Wth3_aEBEd3>#|<Kib^2ZH_OGWwq1;Q^hK?dt5H=Nq88
zo(D%RDgCvCV-k5|8j2kgpWfW}75w-l<n*9T>mc&85wJK2-2OcP8rMF)*?ZvV)7gUS
zO!l{?Te3kQoMyC_tMx`&4Dphh_!<We!pQG43{aciBC4uQb?D7Q%*v5&CUhgFi}=7d
z5Wan9tQD6$Bt+a$_S$sCDim$HQQrcyHTT11U?ZrAckQCu`+;D$<9km*wB7}Yb}F!V
z6EG|`;24VrjZ1u`-71=>UB1<oQ5UD_-6qZrC=22-q#$G#5h^Z+W$Uijx}H#)WbWfm
zeW%<kXeI*qq>^;pIOgf5T649}23>l=_w=wvjQx7e&!=&l0EAfN8`L0N<uQGFogJnf
zYvE4NFd-?$@Q>GxN@`db7HUXs|LP}3aMc{P`$11`l;7@=l@`bMH|8r36bxH!8E@$g
zR@{U$ENhOa@^2ogPBluie%E-fQ?V=AnnS1U(1YP5)xZ*Bz$g3^s_Q+K%yBk~MU}~h
z9FvSwm7<RC*o;jrZCr6Y_cx$Y?+K`v0lnUGJ*Ro3va;6JBh_^XgkJ?;#_frY1b*<=
zC$KR{M|Qt&k(6_%b@~K92HviSIZR7Ps-w#e9KotnV7OWdV+K0$V~BnuXm$&O_sNe%
znM@10_Ri0Lk!gC~4UWk5eZHO>J)Oiz{y$sWti*r7sIR@1MQHy{cz@iEBLOXgxJQrG
zz~E|>+U}IS_S+LN!OLu9sa(c7Cj}W8xq1FxW(JND@C}UR!A&Ad#F~7QZ|}dXcsD4a
z!ZkWO0B2H{$K`1BaxF+qnIJ=;YFj@+m&p03rJ$4aBL*3eKa>vW_T$rnQrL&n*>${D
zuBHK8w`*89(X3hE%Xsq*7zC#Ovfgf5P-Ox$)vXjUBIFO?Rtd+J4I3S2q6^>vsBm5-
zoP>OExytH)tkufe92V}<2E_WI6%)B7f0-);x{oJP6J58v9NAnq*#SHk5Eu3yHAckD
zSp8t3!lcvqPy|hcGv7wc9GOdUkH((0_z##?0q6m+mC<F_3Trq?kEW|5S|%5G4u?Xo
z@NPtvmHs09c5?(=2oAp|G`$KqmCpNl9rBUuwpqRG4I>+<JIj8T?BlcegVQ`X%7|)+
zA^zgz5MVffdfkk^0k#^xg+<@{0mxy3)bC&aS}9Ly5aQJ2VE4mx4BIq?fFaZQ++hR_
z!+uw=<_aQf2yuj0WllvO;5Xvhw6^-^Q|Oto!4(ade1VCqZPwtw2^r~La0Km2kBb)j
zY+7%d$2NOnSpKW6pAX%ri>ShD6B^zzTpp?URF=Q9i*K^3^5>Y=FWnE-l`Fs{XO<VN
zS)?-rkcwq~Z2y$~F&g+R&Y()j_!yA0k|eLA%C}jG&JC|wH4E+u+U{g~m@}E{H?dMK
zkr=X<IOQ0W@=OWJ*hBthp9Z-EOlAFle;uHc^9h-tl-Ou?zAr<}sjjlk(H+{Sgb{4Z
zX0h1muzgy68Vz_~pl(mx6F*2gVSXZBeF?)}T(P7OfCL>4Eeug#JeK&j4ni%xC0^jI
zoo*ThJC)7^_@#<W9J^2H=l{y?IEVhSGA;df^U_2DG5WMYl)&G2cR-Bc{n_?0Q4Wp@
z$WV9w6KVAky6}>aIzf}%oD>A+^Y}wiyZk#_0;1PCKm4D_yY%`T1Yi<s5Ey(#!KBE%
zToOiETG7R{Rm9zi7jK;lR7~0P*Wo}<P*iJe;%r_3pp^*&=J9-M>`56X>ZY_5v!lZp
z>B3*3fb-{yoB%eH_Bd=Z4Q*ppNK%XS_tw474`@4lS&xvtER(DC9%>Aj?UhGErzO>a
zV1lesU|oAe&*$;JL_r66h@=_M=_xLtzh`9NJ#zq5=Nj&pMME8wrAM^i=Da9c%W)Ds
zA?$uSof-H@Ux1fvyO4EjqaOp2C;(5qEy(9_;a9Zi3P|%4i_9Fr-zH}#@;F?V-@W4o
zlQ9(vSp6k|tUk-R?3vXt7!40xeZZEB4hQjcImKpQ5p^_7GVe2&FCVU<C$S`<<4*j6
z&HQ@NQdGwT=uNDX5jnpDDnUT1)iJlc?LCO-;u~)E10ukscKno(qJdGG4&v{R>mC!F
zM5<^sCb=btOqQ$hTIrODm#C0?2FG`g*d3BIJW*D=en!pmkkTw66@F(w<3GiAIFCP(
z1va}Z>q6FeWZs@RWuq<lcJ1bB793{&I(dc%x)29kKiB7j1O$`B74P-skndimcg_?W
z?2}#rVQIVDZ$Q6p{k73k%@62m9Wk8==FO+J`PKxmBwMtn#RPBZ!9{t^m`bX;Av>8j
z!&Sg#4Y~Yu_L;1vZDxBWwhlTG!x0I{lIVYQ2n;cv`cF--B*}<2d#Ev&3ibksj-hHN
z|CcEZO0IydqS{B#2GIQKnSYW`&@qdu&Q7LP14IyjiQajLQS&e`M|kY_D)l&KiIEO_
zvcP*no<SNPBgH)T6MIX`8!um7UHkpVj&^tPPu{f=Zq*m~UEBhaKf-CjLm>#76PZEc
z-4*O*CH$imr*130g?sFZ!&SnYNMW`AnD0J303Z(JDE>x7>SRfdQv2(2qna!A4lXRn
zO_CIMBWiT{mqDNF%2m~wu8gKVAC|d9H^R+bSh>s-DGqVke4M;Fhc!XrKFT)jB%7dP
zM_5u%f)R~fc6t-wJOz@X`W%fYI>%GxH)Q6<`%9b=_$57!Hllj?eB=S6kLR@S_JWy4
z4xi;!O_uB39emG*txd<|qz%LJ(bo37MOaHivFecITnJm)U72#QHcY2c#VryfZA=?o
z4Wv$8;zC~4vnHSjfZhNt&}nqb^7J!$^BDgE+zp=B6Ng?C^>}t|jYwt*ID)2s##OF@
za!Fttc`Y9P0%GH<tp<kRWrev`VqXX{FkstzI~O>0WLla=3bjHu@J{SQJ;!9$h3VT%
zXvd;HNcG-0tDN8)CyQ+d6iD(H3QuVd_4yNvbK2aJ?-gOcJNOTYnmblZV$@xi;-`~T
z1Ev$e>l988mbSNV(=?^j{rz>rIp&!{R*)b9GBgv*Z>rl&)pUSR!JPXIStT&#I|fK*
zsfZn04aKHn2Ls8pvC`jiB_Ny&sG<gd2FbiXvBQ3&vj?E#nStOoibPxoj^XcI9IV$K
zI*@Hb@Cc70UrHiCjEDC|m5un%=^O!b9*}D$2R*Z^py#2?X0RU{iz`8^0sV`T@srU>
zLtqgZjr&Y`Z4O^f`z7GeO_aE~mzg)U7(}$$4|^qydTkv$WmRYE@r1j+87V2YnZg}9
zRFSUTRTgjY%y>t8D@I11vUa8YY>a&}IWX1m#@|4ML^^}l4zwSpwzwJwu=0|PVg`$U
zDhb`1(V)~+RgPra7uWpwS%j2Dc;kB|&WLX_H4;N|R#uGLSQW)E8sTmD+O$Y10f=Fz
zriAoKPv2$(3Jcf%E3ub=2K&t;sW>3CUx+Z+DDdqPFy7;H+6i;`uI3EIr>7P*plHX3
z?6?rxkY|e6=sI^<Bqn?taz10YDKu5Gu1RD|8ON)X$p$S=GnmRXWc~+eW5uA58wjNH
zA>6YCf{TQE9KQlKf}h3w^0zbGcL5aH7f|V&q;p^T2m(<`99Dg?Gn^8cSx#Fe%brsu
z=0nYD_##b6GSW$JUXwpob8<mEhVHZkKab;s5Zw-wHT-Xj=YXpq1!nqyIbwHeZVAuj
zHmC_?Nz{$HBUgGm{KW6EY8wC}_L~gMM3?MlHwDZAfM!IPQ=JrAPb1F2m!zg}OTa$s
z6CvH54ntk7OeyDclRZx``#s6gxu~$dec!9)y8TMesk#@jnz5~KbTYroaSQ4>qY~=7
z&#k)sF;d~cu0rPTj<-(4hvq*2(X{-ABH-L0Z+(GC=I3A?Ms!V~U)A?Qv3K>32pnuk
zq-w~svaDv@rbqN|>5Ot5k~~Yg?lk%=@S+%Ki?DAwvQioQR`vqmbR3{=F)+H|_u8E@
z_WL{h;Iroa(ufH#@6z2Zeh2kve^Kb;H@sCOqbqj8gL&@ZLo>0N8BmmGu%y%(9d7^f
z6~AGfSvPd!Jo({Y%$%JZcqp!;P(n(VnNQ)>;Ne(I(l{V1)mFhKn1!topWo0}ecBCn
zn*8%QEd~96x>!dHIOxUBRxX>hh?tsO)?^%WE_SD_Azmwn70(|^7CB%6y>{N(^~1eu
z)}eG-rPMWaCg^)}>H!h6yehwkdL*v*anJVjlnFLsEEt3K=wH0mk^AvL7?=0;I2ic#
zwiEbTV4N-VANlb=d%a`ttJ{?)J6Pc7%SjC`dl&zgsD{Qr;L=sV?LAq^w>NxNWizz4
z7J;BiignPC<*)L7nx6Iswhy7S&CswM&&ie-ZX@<D-TErzPwkd3;GXXQIO{Aqtc{iU
z?D&vPK@6BqfdxgsgK&k7wo762={J?lE55bG{F1YpM;G#N;vKG1r?}*wuwn7)^M)9R
z$)y|y=2_8rW*B;{gys{`yK0c7BlXJ|YG1;3mdyh$!qbjtK%6L23_Qq}ha1fIkllOw
zWbY6O^!ifhdWU2U166eqO-Pwzh|`$7uF+TmMrji7QmVt<>zRDFFQLcYA-VYdvrnZp
ziJ#(ZP!Q5!Nht#mxCopgU^1B$xbPvuFr=v<N4b<3#*~bcJP`)=xmmN$7ObDrP;ZtY
zh^qw+Fnz8&IhDUozmad;!2Wf8lwR7>SXx>V*!s8Kb-i6xBvjlSLjw3sa;ID@+@CAD
zp=FMb2rlC7KF1r%(y9#&nysSr%gJ-T{#5`}UG>vJf}jIW{6T;8ftxbY8sr%<kB8ss
z?^0m;sPK{S!(ywpktokN$_xH89$bT*Rdq-rJkFFH8)wy^IDl|$Swo)F^_gu;t&mGV
zGAXKzT1#Wf9AQY9zo#2{4z4!&4H_xx1QLm~UG|S#`gz#E9D!N7!o2YPxWWv}fKQk&
zG~HX(T*==OgizQ|b&aiVzunTT&R1Yimp}?<ysQ#Lnq6_yH|q5?4|V<L<&{1>F+ibY
zx6N;7Fm^6s#T01^ixGND9Ay7f3Azmk9-Ojw@9q^|N@t0ii}g-7$^^`bkUW$VCijU{
z@?OsC&8M)<C(796{T#mDZh)A(yD&%dnJdgJ-KH3pAB2?}M0~eYZSn%8x8%^f>pI_;
z-TmKOytigT=sgG*Xh-8si3(AdV@RffPjsZ@VV>vC+hK~5>frtL{W@h&s~c2rIeTLE
z=FTF1j^Q%0FO=ZNMHBMZbNP8)@A0%@4>q$7m|vNe@uJFNlFm$nrXde&`yYA?`LQD$
zn3R;9R{EPbooIf&c+bnFmE5hi>+rZZj8Ku)VV955QH35~v)8AtOsy>K+>ZJO&075r
zAeGYeREA{=2EHC5u9g1owH=dK^sEvH2UJKW)wE6&p*k|mFuuGY<c{3ULeTl6Ai<$N
z15t;s!)=P-BEp5b2+u4dLCA1LyP$gki@_a~OTc}U*kVTn!FxJ>j7%i>yGttEc!OTJ
zqD&2m6EbP{lSG)=tKKgX(>ILia8-?Brct2NOs>?-5p+ZHxebb+G2s`Mj~PEmOW5QZ
zc|VLtmPlURYcQ#pWAAMl@B`$tEOg_IE^Ia-G*8%lpIGJf49)8FJA5V&X%Wx4qH(J!
z1q{_K<h1)Khd+gb)mlt2Q<R38-6ZhnT&i2+FB;A1v6ysg&KR<&aG{EnOEwch!Q{RB
zDeTh^sQ8OFTT9T0t=q)&ki?<9nqv$>`7M(k?Tyt<(X9?~tlhf53IB!k^97=m%C+xY
zKY_p8oB)@@$?O(}3iLwD!Rv<S6`q&Omx&m0te9(LLl}Ix`P&Fn9}MM7D6vaB>9vU)
z^6D0CDNn8L;U@zMuJpt$yhJceWDV{rV8ILqhc1`Tce(IluQdBE>po<oy~7P~OTzM*
z8!m(I$&?EvRt)qsgU_+*$m`WAbFg5~wb&MFV!t8k5xvHATgSwqNbVO+Q<h)}K>|zP
z-^|^wdft$J!^jO>9?UR4?fCB0thw?wd0QL_Bk}y*>1PejB_Rv>g8(mc?Y-1|$x`mG
zxX`)%zE$w`o{>V`_iPe)wpguXNs<?QJ&jO8_2hZ1SqiRTN#X!?*ReSdq)30|`+nXn
zNL}|n88)0NgvSsIZHqUG)VJvn$g#`gY>S6lbksr$7cwtGMg^_5pX6VAeugCozSB@R
zM<HQX@t6589c3Wnu{dp=Bf4<f$=SY<BMbR#ubo`1H8JNFAXqumv~{qTlVC4<ypy%h
zuZ4OPHeGJSPx)=-6XRRUW7Q1$ES%2@Av1U<X(;O3?tPddn}>{vI*q`w)%Rk|bO=|r
zYKSghsIkcVzTY!5=(UPIbAuTiY5bx>@K$v8HzJpKNRAjZM~2J?-Q_WGA`IRpcxuVy
z4<G^uH@B^np<=p20t-pl4Am>I0dl{dwj1W+{m<HzoM#xE_FyE1_={6D9M*z}V1D0~
z2fY0O64lY$moX73>|u9T`x9Zh_MqjBFnRve51|$`)6*=D3Vn%|)qei|1mC~XDhyP#
z_J|A@@PnC8;ICvwX9Kg|JKD_`aSB&!%-Vx;DS{^{fWa_=(3OeQLU$uO)gt$lUIpF1
ze~QUBLda!OW_MxF(TarF&}p`sSo41~bUr0X-aX?yZgbf2{>$<JGZ=`m2=!_pIJ|d`
zCeTFz&A4(p%lPzub0E?yn!ZK1pbD{mM@8UVl0XvwgERNX^j=#*1v#8Oq0VuYezz5@
zXWDy;_4Ty<>gx06>Pb8SACjnG1kIR+|5Pg>x}lfJN$nOaQ<JK2mrL=mv^eM$xNJ_{
zS$9xX0R=v(Pa=H%^3*BmuxXSZS-`|be4809Th6S&ea(1zr@R+Y(#sqL9>Xw=;z|^k
z;a&(()%s{J_#ONcul{$`7QaKOro270BcIaz#(i$H*sZSXeSK(}1G4veS2VQrVI}B^
zcTaMQ$v4s=9C1%%ruR7OTrebla^uVY%;WR?dY|R1>$tOt=Vr~{^ytPO?Q<GPn8pA1
zEZxv<G6Z5gqRd%i!Xwmi6@EV(%*nA5fzf25Q(sYJtI_x*?WDJ;U;z?;0B-qu{x=iT
zrNe}E@@MiCR@V6)QHT7(Ak05Jl?d&cY8oMYm@bNuQJscO57YCDHBaL~c=Sf^8OY5Z
z*`U<^y(PpV0vT^uH3fZ@_Cme+pIC4A(QtP*>oSo(ojSwpc3t@=QiW*ZdMd_5clC(m
zbmNfm2oXaiE`&^uo6mrOpPjR(L1$XkP(@?QOuRpxQIF9D{8qzpyU`xVwMwB<to89c
zMBj6uA|pJUq?o<%30R3jw;v)VXm$Snoe_Q3>l8v%@C8GEmbMDo-*@nK(5I3%6hDFt
z_Nh))gUva~>C^EhJ;k^5$9{Zcf(c@{^I$M2iqp^iB!}~H7lu^ZGbrhChqqJ4;Zwv{
z|IhDp(xSYb6|_NXC?G#q83+PHu_})S&Z&5qBkW;SksjNtJo)QXsa&%-jL%yyr_z7X
zurmj#GnZS<CEyF0Ppb2j_oRMZXcQhDP3^fDf=7W_EpdFGR#UP(J9F$fr3$u(UBUM4
z>=>KKA^(X0_C#MEWY%z&rt0Gd`PelKDa;M?tgVofGS1kg_b&)`4u6?zHV<A#BEa;;
zU(oUD9hgDV(uT5p^72OZMTafNktla@((h-xuKiz6Eom>_cXSm>;(p^wTkhrnQ^(<_
z#R=D?Kwj|I6~Z@%)#v^XF{F!)uZ%vw2O#?Q%{<dtO|3~mCLa%@zN=dl8#0{`TV+PI
z*<<|yH$mXe>iYrJt|F>9gmOn|DdB@w&l>@^TWAIch+~%_iJugP>=w)tctm`?3Zj)c
zYVHY;*H|v@AY~rIOBFl)duHKSs+i{&4Mtn62<P!%E3k?*mIlmfKX1gbX%mf<3P$Yy
zf5Jok7HB%WI!JGi#KtksTb<#;8~tV0ylewryA`X41i|BpNfkqo@P#vU2PfIB1}Z=i
z(3mmP%PSf=3gV#}zK?0iCA2`Wj*{S0o3u#7VRu~!^95ci=Qq1dF3-<}gPCF%MU@N&
zL(!y@;K5`wwZk;6vbX@#hir~>adRYhuol$oF4VLQK3BPPCcS@WHej*h?J9S0@rh$_
z6a9@l!KtHi-ABbkhhVQR<V5PL?BNqhHr2_Xd}tEMepm!)I4BsuhMB0mX3?@kN?3GO
zaYcL~<)c%30G~Sn31$v#`K}l+OFbUrJL>lMARbo2=A=wjM<fsh25dgB;4ecPDm(W$
zyMV&r!5pD}9K#Cu7IQh#2c5$kL-Vns>wQatzdvRk|FiV;b_;1SQ+4Ns4cdPlvv`D&
zrb7+hJNp5RtW+TQ9Uy0p#IKUs-3CLkfmCn$jaq(+q{#T7ZX=lb)QKT1bAD$-QYJU#
z86MN#dIREA;xyC5Vsy4-+7*HPbJm;{rq10#d4rE;t+FXZo{vz__jn^SS(x-%hFfRT
zAc>FwPaen;oX~>bJFg#Kez}c#U!K^+){dtLCG0l^Gr~45UzYb7t%{brdbz<$f*)?r
zAw!Xg$GBsR72zX|0>j>#H<NIy3Nwt3-3_PzRtzm&!*0_98f<TVwV9ut6%ar9St=%h
zR(CyJIhOlruFyj~+ucWxdhbquzwio>(|aRqL%!M~U)<cp<BXULcNRZP7s*Ki#gzcJ
z;dS7sks2W_TKIKJ=Unt6p-h&OF1NrgzD<Jyu|+rr8J!<CukEhMu<iwW<s~RS8@Udb
z`B)(YEpuQ%srqeFkkMT<#*0ZTxO)|HxHslPCrF64XS11b(FMUQGlY4jxqj9ExhMjd
zPm+Z<<96*4mkX_z%u+lPnaNE=7?#^(UFu4yBv8BN6cdmiI;DFv4-OL@mj5FQ&ssa%
zL=nJyp=gnqAEmG&54#JkB-LD}M&lf7bo5MgPc#nVH=u!XFedZ`Cb3OWhCWw4`ge!_
z<xZ|Hl;<7H68B^HFYrd<Rt&IL3+iDdo%^QTO-PliJnW!q{)UDgb?erUEI@QXv54>8
z{EmZ3XHa2RF2t>%D64Ab(Li)tQ~pfRUQB$v4(2AA*IQDMSm+lNK7<EdZm6<XfUf$t
zb)A*lXnG{upFGFgRdoPWJ492(qfe+pI^5$VTM3BgBT2r09}713JZFonw&Gow{Nbd7
zrrsp*_4l@{U`x<z@=u2D1|RJTVZ}eZp$ogK#&{p$KHXeaeQuxbO3M#i;0~gY(-u^k
zP_Z7xzzBv5@h8Y^Bf=Nm7egLT^x=p<#^JrMLd`>85<V)*<{2gQ0rkG>@vLeoLEzO&
z&KAEfC2RBQX5oMhE2s}6op5i|G(mfJD{ZDatd{F*=j-i$0+`iCkllT%+@Ss|bR?!g
z1LZSS{cUiH@Ln+2FZB8Obh*(Q7wE8wZRB^mAVu7FKl?UPyMq~R0F6Yj+2vSp)E-9m
zY$2R=;8}gjBI)*c!2>+A2spHlxEy)S-|%q6I7d58b%Wz?$25tTKD-EC5M`NN_&&y@
z{fwZtdBGV7?T=NqoDP0sHN`Nph%K2nyg03)=y^Nvm#ooA+3DZcl%33nh$I;(Td<-H
zS`YNntB1@uG-jqNJGBGB(Q>UaH4wj}jcKb6dMJX5sj{VFT&x+^fG%!l`Cht<H%-tg
zo7B1;r{)&ZQZ&P(xx+{E1vErTu<*b?8?x*cq|70rm+>(N+BQrC8$7|mmPqX*r6aDu
z7)1dS&BZQC4lLxRzbGDkHf&+&Pyh-J5op=BD~LB+&NJ<_&MMF*K-Yaluj#PS0MCE;
zAZ>ZS-cV*?5h3Vv(QcBC#>CtB0)T2W**sm&=%Mq{8rAjwW3|44-q+`d=jcpo1f~ey
zhKLqmmiKThU=Igi#Ay)A%HNvK*r{HzA>kjqFcX;Gt;}B>hFk`F86V4s&0Qn&haf|j
zWgc@WGlnqtANwJ(e5h}@_hM;cv?6lu^P;~+ArtVj*{?sqq8l=xFANQpR=j4G7vOT-
zt>AO@4n%_E+M-+;U8sB;PJt|d?ElT0-fBe#=Y^|^k{LHNM&y41Tn~+MSy&5|sLrM!
zO>4xNe0IZNx!}an$+oFWwLOddw4^w9Hh;Go?-f%^3V=oiLs|s1igMZ6>G$~9XU--{
zjW9pf3lcBhusV95Mx?9Bcb}&Gc71Ley4t;&LSNg8g(H<zT49dV>g1Pzh<2t@%KqK-
z8B6+v2um<|g!<B!2UCWC`8&$^IpqY4cIUqdPe?v(^mA^7^ArS_=muv0tC7Zz{yux<
z{}8%A9yjdJF4nr{Zc+FsMc2t(t&;L$>-X%1dj3gS<^e%K9?=7)h#52VdTrMPFRSA^
z>?h+>7*is##q8xM4oRU6(r=UXc54*hG0k-r!$J_)h?k0nq$K7r@~Z1AK)F}sItF&-
znW$Yt_~43X;w>_R6gyE)Q9sl%jLQcgl^fLHhJ&;CYwbCWIPa(1j|tcKr%ou;jW=i=
z7{?I@_AoFd(YJ`W-OT3ur+k(R8Z74w8`vA?WiH0NTV%OYE%go!T}+!qhL5NDr+MzZ
zaP)VWv4UNfV-KsxZFe_oZQAeZ_p&SFxbQjs$VON_$K|bAB$E(T%tEBtr|U;(=JYhE
zygS%%{)nM`p#Qq0*XrQ~2Bv{zAWYMQ4;jguL`^I3d`9RW9Hwid_f#v;;GjFxxM1Ch
zLHRPy$B*m`2b6{)kOv`*bmwI==wsO4g3NlwoN1J4k`&R~gz7QNMXAONSP}L5o75&j
z%iK(F37`~b6hoy$c$8Cw|3vaMHH;Ai2eXIgh+pA4Ie!5)w)^ECy-|(~ztK2H>=078
zw;i?F#u?FD=$r}=WO?jTei-Gy%DNo&t2m9X8Ow81!tp8wBS~kI+J>agnW?2HWoX5b
ziiMk<odJ$pR2*TilxMcCcT=s+4@?C&^zGuqx1}sCfxf>POf1>C#J!9^Aj5Xln=Ix!
zKG)ClKQ9+yVx>R>d3CH~U3T5Uld%v6BX-iQ6#Ndgy~~c!{)|HnMIQ7%eG-4(CIsd#
z(+9aZ{jcrz<oq|;ZEyX4_|a?2ego`wsftdXI_?G7>p8)<H)!DVGi?Em9ozKGN*lel
zZRfA~M*Kbx&#|~o$8$!J?Qpkrs92p&BZX6MPhD0c(Hh+drt>x3t4j*-qX^$h{TE#@
z+0Nvve>_PyBT}$mZ?>F_o;z}vUuW@1pigmAr5r1xrELDTS>lk_TSBNE6{)hSrbRuB
zO|=vEGG$b<P{6kB6%JloIGjROA6K2{@K#aGp4PCVR&2#^MdOBdZ1naUYFXLxoa^E9
zyo#?8+bcW^en8uC=dhs^goO0|aWMCT*l<MwUjBNRNS@T~Z=6nRHLhmk#LP-EzvDhw
z)&X6ISfvX?3%5QO*0+8cF?`+I`L^)3QnF6^&S-|9kKW_@m&IWziNS@Y1wF+Dg+gP4
zI;W$;DFBgLET2o9`@NlIdp)f&vz?KXx;QSvgt=O0(`oOUN^u?bi;#*MeuE-d5iCdi
z!YG>wvq<CF)J)(^*nius;Yjz2<ixIvB;^UEJ2jJ>%r|}`iVH2wgGXw#aSN|5HQ=t-
zgjXtoi`k$v7nZiaL8)OMD5c1FUTZ^rJ``-XIfNNWWH#?;Apdl;cnxE_f01{2O*BG>
zBiK7wyYj=J-xZ~|A~BiWXG3NM(J{nc0Fi1&#_=dF5oN*<W(K(e%E}xm{u)Wg75d(5
zjhI6bh(?`mJa0za4_nJg{+Iv4TjLq5z-V*~W}B%8d3N*B49ESVbBj}cLr5ZFfQyXb
z3NZCiaG(h(jpqk_*rizM95u&ml%uOf6<28K`>dt(d84s?r|ORO!x}{jB~+b`bBy?N
z=e8Jdvh^hRc5%JaLMg_({a?#{R{XD|1yVce@&3JDM3~NIC0En>YTAIp-gzg~!;T%%
zaEDNFXN$-W4B&^}S-l|bNEJ2Y$ktmIR#ZcqkvwB)($k4Q5wU9><R;aj5=y*WPMX<B
zBUvIysUARM5Ad7qXxnl*9~hESsIi+)mLO!dM9kA@%R$sLyr&Gwy|+=VX{oBrt;2G%
zc=lx5Y^GDqlSs24%czrMVmC*HeN3+yAEJ;f?j77a^5P(`&4Q$0<6JO!(EP27yytZU
zS6q`Kp1FD`%0rCr<iBchgaKX$0$ZUgPZEB*teWD7=ySQ?E_xiIhwt4AC+|o$`~%cF
zQ*o7vuC#6poDythcKqufKBr;eOba{~6CCl!40<1~-^MN?K7NskT?%&pz326|tH(|+
zm%eWgOj-o~q^~Tp2}it-gXsFYvAI}iy)wN>p4HSf%F9JC+dJ-|-)V2hA(C8w0)!7r
z4qQ~qiH5=6v`Hsz<1b1ZD7bfY6i&+#m2mqMHhl^RkjnM=)MLgXV{swZ<LUF1Q|}o}
zEglER1Rq0iT3xfY;$Jj>6G!gCoIGu%9)WI>QdwQVZdqG_uC?pgp4*2A!xgk9<H6u+
zk3z-&H)9W7rB)!{@1;AnH00q;*MbPD@5o`RlM$%Fd)_Dap5j-G%4IS%!<7CsRZK3k
z3X_Vmt-o=JF|v{wXKh8Cz;BpdseT*_Z`QTg8VOZ<7P!vw%f%G2>@I}&xjidOHZ0qQ
zzdZ6?2i%>O@07bs_%QxgyPP`zyD}6wcv29~J^4ZPz61M4dY|{t>-zrNJR<b^_h9Ga
zGtk3shxH;DTB}z3=rwG)HQs2`{Ye5_e|3F*)X#t*!fvzKdk~(v)^Rl`k!?gnI|<UX
zT(FT$0wS5#13W*#^r{kM63lZFZ%Z6JW>{!!hR#NDK^As8)xqakI*f^lO`3kjt3gy2
zhWd9jzlSr>%ABXfE0kWOrPTkR>0~>shs$*AC`Qr7(JY$yC#6Yu%2J3hM9y@J!tNNZ
zgO?sf3sr$M(o-mAAf@q^sM|z}@)-^=yXhe9>yk{BHrs5~t8b~R_q6ofpO)wM?sgNm
zwRnna2c<oNKo|d10Z~;T@wJ^Arn;wHMkTpIt-j-C6-4d8v-}Pvzpqau3*>O~QW%Ei
z20|<Sw9sCsRY^(-)BI^Zi#xto;po*oPs*&O2Tr6v(+Cs|KDTF_FkbR^<>1dFg}dwB
zm1ba>SMbN1(RyIeQTw!Dzx5sfo_Jdgi!mUMAg)Pi9#FcX&!=ff%9^MJd*izjJ;l=A
zEMkT_lOgjJ)!0By<MY46^o@Ps3<<E9n0N`jfr|7hUg+=n1{+fyb%?hjUYD>i3^#Ba
z34l(ZS@{;X7#PbVEFNdxFmi3Vy_ANeGZ_$75OhpH_&(%%aqdoj{&<03snOB;SrGPi
zF-_|=fzjWq^7i0ic3&c_sH<OtcKynRDts6u=RK>UDNylEMkqI}4TxglJ?teF>N`N<
z6c3ZS6Sws>oBKwjthE|nu4n8lMb;xOPts=UJQ?~<V#=naohZNHHWQjQBTv?d9S$oC
z<``6l{tNLOpNBR`W3kcOeKb}F9*ZH>n)YP~I|wAvae={{KZzh4MT!BVYa1wTeg1bc
zbnK4;MiHeM8MFOgBmW`4Un4c;C+|hG{q)Q|fBJSL0RGwkJaGH*-PBiVY>!+jbC0SH
zFaF%uaId>*j_3Nq6>#Jy%cQw2j=IDl2yOh6v4a(M_Kq5Qmj@M#=eF=qex|d2&u$fK
zmc-<|$@ho8>U-)1#Wt|hMNL#3#f%7n0;9y&MvhV;JyT@(+i8M3oFoa)5F6@lq+e3`
z9n^BI)QxY3f?SimUiS{Z<C%m-k7Qqe53Eiz<&t2?qY5ld0i7M|%4mCWIq7`odjEb_
z(DG<<siV89BD)H=-Dc6kg?kfb?CGOe%WQ+_SU_GA?dtbw)xG%D%(?W-)f6>cD#uW*
zgmkQ1ctbJbw}efz0y)NZ)7HJLVI-b+3odYC)hBx?b_VZHK%QphvRipyg^#n3rj4&p
z#JMS(K9g%-NZ)4X>;`Z|Ck1amBZz(4t3b(LsR@FQLMhgUQq_gJE<Kz|5`4Ek{LwK?
zDILcFZ6N?`!9g@`Ht8!$`bB0F=$a9+NmrvAbaU%|59XKv^Moi?tG^l`MzvNZ)xY*2
zFSg~uJ~8DE7+}UAf0MroWi<S(tjeB9Jpdgu)kkJ1rn-Y)=W_)bhg^aD9B|VP;{IjF
z+S=R-{iDaxZrkO9W96Z1S9Ds8@=r(2lm;1f3?GLVrTv&QJq?Z=+U5=1h{b#)lJyN`
zJ`0rMiga1OKz+~T1ZvRv3C4n}S@886U<V1D(a`Ga^ie2`M`FiQf-J`)MJ9_rWt^8~
zm(K|BFZwtXbwTB~G!k|p6R0E|m0ag*^iIe1eK9XtJkfZSNW4xPz3_eb1?#6QTe(Xb
z<tV`(@FnC}-ui^>UYc&`$qT+S{@>B#V*6!Xeu@BLu{#9-!OwMp{!ZibKy9^seZckB
z2z<=41m+kZBXbXjA>x-yv`2AiT+bceG8f4JS;Hn;7aGI;<PGG{$qt!(a?<>3k)|pd
zHP(9<Y6pXW;?hgtuEcyPWc(Iya5q?yN@&orS)v_Li<_=T!z&0zXeXMojZl1)JC$L)
zn^dG19G$8(sczp`NkvTQ#<jc|oOefs&j>4UquO58pwKa_5o<q7HAA}&kZla86ba&~
zpsWf6>?_m912xsSWl*kLq#=msy3ze9X;);^4%+pHwx@}Ix~r7CVniBke2g{d;-TOo
zWhvQ4V_tbA7!ANs%~2UU2(in<h$DvHZfQS?EtyyZd0sqXsCJP&d(7oB3IJv=vphe3
zs4V(bUl_~V9(yBchKr-+Tg{Hw^IujdYjlzkJ8OG-bh4`qK1XYN>|0d|y9jxo4i~em
z9$0UUCTX~hcx3*`nWP^?b0Rb=5MqLq%{iQZk?rdS^{7|mY1A~mCk|3(Y*Sl^MAite
zf=Nl8m#wp0y77<6pNunwkEI0X6sUE>>L7If=J)x%ObTY@d+vB+KNbF!^p$E=Dmzhf
z?BvPfQBv6rf*JO;#UQo(#*HGG);p40Ov);l8ZiBRsydaE*j0TR#2_H}qexh&BY*xC
z%Y85`y){0^vAc2wR{-dJ=IgbYdk{6;M0&`g{(IYp0-YxH$;HB_Q&JLHc^II?5!|w>
zX1p}8lSG!Xp+9PNj}`PJg;-x#Z>g3;Pq?OUkFj{JW;O#%LisngO*A@^FF)G8T@h+8
zl;r%c1yBCRw3hRmTZgzo*<^t(vYgE2JuPV&ef)C_v=wwVeaz)aza-#wkzYbI9wGEx
z7?f!Hcu1#BDZeuXlQb)WN!EaXAjmtk6jzQ{yEaoct*xnttYlePA$1Fdf%=t5$8xyY
z2dPsAL$8%6#Nn_$E%((8=ZdE;hmjyrACnkb2HAL<Tg>KRK$O!>B#O}^PfKylY*2PU
z!SGO>J*g4WX{_k-acHc_m}I}z_{$5uy0X$7ptG8R|K-A!y>rH`q{L0#b}I8AE%Bj2
zcdrC~*gPJ{VJj%Fh0aPJSW=}`gOz_avfxxA|8x~9*W>IT6NW|B_=>P0OhFE%l;zWK
zdZS^Ny^IxMqzsNcrW9P@C=*YXTH|Y!525B$Zzd5^(?a&9-AKSnT(n@#KLw!Z2mnrh
zKPDxBW51{sX}*~qwCLak=&ORj%o9QUsxf#?HjnAZW*#(71N%qiTO2AA7TMC)x+9N^
z<!}p}8<`^<dLCpVA#IjKEfx(rycjL4kW(aA6HgfBPlhqX)zRI|$S=fjU-CMvcQt~}
z1o5L#y(j@CAr2~F5SwELOxA^%hff=O<YP#LwbHey|H$EQlCiqFz4ofdHDiaM-3g>G
zijIi-35Q$G#%aA0H<JzGY{oI6y>n{BNV%w^*%FH<cI`XBz5}2guO0TcfLN<~LdV<W
zQn4%iMzhmKVRP>kLd3i26Aa8#$h6_(2_(n`(uhq+gy1JvNz8^#F_2C2WRC@Lrg}^P
z@879SK2QHaGk2uvD9Eu@>cQWZ2-`#j36(0X``)&F3;gb^qj~DgKT?)u{^L<^fQshn
zrTAL_fac2CZ-DK~F5<Rh@dn-!=Ue2!c)HW+j$h}GYD48ryH&3*UeeXvC`@l&>*Ml(
zNwJ;UGeOEQ#+Vs@5VT7NgGSZt?L-ZYU=Bf*6XS{0I~2x17o58VCZ}C3o(T1$0tY>5
z;W{`dR(mpW)g@G84%3cr{rKc3-JNd;|3<OpsKgj*Qxwsl=)Mawhh*yjh$VEn%2oL<
za=zVtIX3ry=tiO3PJ;!q3<?@rbD@Q`MC5efIVq!5P~bfTum}Y+)m0YBz{(P@;m@PN
zj_+$std%3O2hm6!PBx>QsBVxxksD?vPOSiJfvE0L8v$XB4R5Q!tnw6fJWOk#rhBpN
z4^8ZUm~=84AHrP5mQLdXk9|i+=3*8qcF;z}3=)%gE_3F#hpymj{%Q^m1;p|GY=sva
zchQf(7Wnv+2_)8B?*y`Jh7v+^AXdKnk?G|KZ?z2b9I+?+8Ps@>h6}jmOgXYFq-!iH
zStrJh2bHN8p;)+K3ji|Pk&u;0(TCHPk_T5u;MC6kSu%1x(kOPl$wlq#-2wJI{Jt)B
zKR3jV7z~^3#xt{perbijZNlSz|J$<j%cS;)HYzRkM6H-N4vEeKj~4$aBOO=pS`@Of
z$g>Z(7tvE_te2A)K=Te1A*=V4^GMvoLC!IA2ar0A_EX&)GO;o56SUiXsbQhHaqUM!
z6()nweZq+UK9ose?x^1Za6VE5zfg>VD)KsfJ3kUx<5tc?UR0@{=Px1_5EAU4Czqc8
zDM7y6(h^($$CeVW{?{XsFT%?+-e#uzfvvVL|5l6wZz7PFf9brPRcSG9_*+b~(`(hJ
zYHa;`r<O_xa2^~Ul1fPaq!T<NuuA)9Z8psPU0kLWy9#((OnS`LxDhIB@e<u_<Q&l0
zI{6PsHkPT!Z$sN1zDjlMKDkR>bIbM{N3RoDgg;JhsTwXjVR6IL&nX!Sv;J@mFj2}o
zS~HGx?hoIV1(WKq_ji5Q&Nuplptcm-E)@SxlbxjH{%x*W=XX<*CCIzpQnAg&AkNo2
z{Lca$!rk<gF0i3=J?ZueRdcrAXe6n{N{V3sYZO*RdsWOhL6Q-BiDuAon%IlFw<@C$
z4SRo-Yrnu>R5<XF68fI8k<KyEMc#0y(ow@Ugk~1dWfvjlv?YYtQKp~&&04jrHekLq
z1GbX%#j0KYQ7JSjO5Yu9etj3Gy3pywEePV!{!oWt1wCoBf3=aPMXAiPthTqoIR$Xt
zP9I~IM$h{^>w|=Asjjh``?Klbfsk{+w;q(3<ELGwpCJdX5t0DbX2mvq2`=)~YH{<9
z{-QT-N!UqHs$EYSk#K7;!nvKU>wYDNcIVgFtYCcSn$zLDkZbNP$DM;2YBQ0#va)on
zeH?ZmF8p`C1smR<nVACwMG6;mDDeQOL--SnClFU(raCmjY@;R4<J!VIh2w@xFId4!
zK+Uq=QpeTXkVT>e^Q*dZVLK}aSYdM|@D>^|SWox&>F>X7`=zOnI+~R3*0@~SZs?m%
z@_5N1w?pWtaS^}cfNCqlfOHJR3kT3l4PVPzipwM|1FtN94Gu{J{9c<Sn^O4#72ZlM
zVCPM_F%wA8J3afJ`;gC5&(qg!!q<nzGKO{C{~$R3|3PrQTMG|9y+^n#Zr`85k+%cq
z1g^jKLZZ+Or~fSNI9<AvtblIMk;hv?H?G#sNeh*crMOOLBaS@!WsIl89O8+q-(?#j
zX}V6*YX*l)9Lg}AsBgq4)V{;`)}>ZfTQynC2)Y3hG_!@cRdo#rf|S@xnAQ_J!x>ws
zlVWy7v`ot=m0b1|>wn<oAo@`8wmQ-(obq~H3JP2rh^ow-<~)<tOx~-h61dK#E`&J0
zRThvtkPHN!e(`8%aTH5($!|3)ptE>|9M+It40W6b*^~dmOTD9Ti{<g;kD(}3m;#hK
z8g-`FCcpG^9<aZcA6N+Y!kP6fCo;{753~sHceu)Ojj&~W>*1*;AZ?fCaO7ufi3<)_
zNula!m-Hd~k=_kA!II%19xO$99Q}bY!)VW};Gcy^v7liV9GKG*QmYe%_d-Gx0J-0t
zloylNudDTQ)&}5iYaDZV{~eW;VP%F@UA-6>7TMnJQ})dMbq03-Ms6p=mtoSglHjS<
zqw^ewVoLEci6AJFYSVQV*W97ahMB;$)Yl{>fQv{hD7E(^7hOWMv`Ay!V)VmzT3d&6
zcwTk4oEr|^>+2bIj=vuBqi)<&_i(?|Z2L<IetLOTO!#Jk%{1QfMYETX`_3yQCT4NU
z&nK5&N|skn&G0BQllP3M5XtoW&7_+wab%zCaXQWb^epesv4x)s#V6xn+e;Apk#AWn
z$5S2x@H9+9YpAd&;M~lGcsMUwrasIgwNa)9uVJ$}Z7e`@2NpWFw0MT+oQF?@;suKp
zUXskY@1n_Xcz+=izu(`_^&PcH`E>pdadY?&H3@9H(&AlTG==JYJ)@^ADB!ZZf+qHJ
zI=tBffb1nL{Yu>BN5U?u?wkF5fwzgw&``AAP4hz9TM$y6HgOn)TuZGm?1JU};3HJh
zTDgw3(r=t&Vv_~Xp&eEDcP9l9ZFE6nU5^)e-<;V%2Y(I>L2$<RAE;nC&~k@gIBY>2
zal3UCQb2SVPd<#oMM<Z*!Lh|S#YuPv1yjbXmeec#&L$ewFwIQ8d0fz<m>q=eG-YoD
z3E_lT$gpM7Z}yt+2lfH$X}bGQJ0uBb(&VgQY|W2bQ5Snf@_Ppbc_h8#u%?r{6Vq7*
zw=!@t?HG9*yYMJU|IxMaNS^eJJGb74tnkV`OmV@z0Oe)-0;ObhTO+cH14}R$r$;!z
zkVm3whkHWbZt~n2;IK9NZs7BHm#LZR<as3fp}w{Nq?ZK-d+p+aPL0wv3SC5|xUgu>
z)*vMabw^M_d42tWgs<y6UiZE37%O?#HpP%qcltd%p;jip)ni#;g=8gDj7o1a2|494
zg#pqY+7C{<k~-C1D^4xiE8maRKX`~DU1zF4LA}prhy(&6=0xw;e3rjABnhwr&^tc(
z8@Vm;I`@pN#^{|u|HXB>rBo7p(SurwNtu*4FfFI4pBeqs6zQZA5-qmSAE1+Sr3`5p
zn%N7u11UJ(ycYi9)-G5MSiqrDXWY-{_kN-q<v#o+PY{Mmcv$VI20EM6qf9(!+*ZxF
zM-c_pMiiCfiGI~#2BzqPVNvLId91BhE+qk6qi-aJ0g;lgrNhj+`=aGE1Nvab;2RM1
zfvF#!I%C$fAHLa2UjDa)^Mrfc&xBFwbA{Q>|9MO(kKg;QS-Y$x1mIcof*0-n(<*)I
zB0Y6dyvBY0(CZ64|J4TIcQG-ueE@kQ6!;0Yaoc*mDdAV&dw~57ya8|MM}gt)yj_{3
znDU=RcX*^{wjc3*NAl8GB;Bm?ZOVULxKEaYaYJz6NtcG?Ot2_xwQ<ev0q*N!=_mA%
z8h8W_ffZkEF)l=fBoS@oSkce(kZ6JjX~{#EL~Hj!!aJfAoT-EObgND}R3De4s+T$A
zxF2i^;+pw?%dvc%46n2})B-u)iwrFvyj&g}B~JFhq@+@Y*fuBKHiP6S=k+WVCxX?c
zMSO}5>KO>&!#5JQL|t(avWD)JNB;GfR|wQ97V$$&S1Tet5uD)lyMvsI1UpYQtGlPj
z^x2FVYt8T_Qab?x41eFp&cnf&K~a5qTX_d0h@mb}a*pBLBeld&lzsOkp*Ud$FY?6s
zm}c)jC%9&(O(iz;B5o~1_GkBIuUplCHm5?tpd>6{v=^_#iDmaZ#@J8${&EC!*mr>v
zd%@X_{>M2K=XMIie2*8_mpj4Wxeuh2)Md<UD3&i&H%)7`zQOtMJGr?vlo^KWb_;GU
zC@Y$#CicnE$FCzC+S7mzZ;kVfrl+`-t$JTaNLqg>&N<=O8{eF_z0j~wF7Pn}a7*a)
z95%vGEv;BieH4T<mN!CT*bR<*y!zib{-#02O8GWy&L0K(fkQzt%u(usbH|uek0ElY
z7-R=!68eya<g6!QNQDO?P_g4DXe4y+HKPh_lIFhH2CH>D33EnG9)#bW`XfbR(Q@S-
z)<qPdUkRYY_IX{SjV@mA3%&+VEy|ywsR#{t!A}wVeE+8d-qtU-e65}9Z_Y3RCRU+u
zcH)Lhmb1AnW)sQrvkbqR;7h#-sr^|;VoZ3MklZK61u{v}(h<zJ$uuRC%VC1iU~L!R
zPM?%B$p+zmTr`QNQl`^(TCdkW_0JcvTh3Y}QV(_TfLN_>;f)l-zQ5L^ntqqc6{cUw
zm@2!ZaN=vSa~V8?!{EM!@jznLh>IFXtj#s?4%I()a$CYIH{lC6zupKmUM%~yr#MU`
z{gxh#U~1V8qElz{?9n?aeDQ4)^LoO%Q;X)TOyR6uv>fs7xYGnuAMKE`GHO{)@g|<H
zWi?PX4Zw%SLVn4WwNEiPw1#y5m3D+|!vACSsJ5!(XnWn>{1pV;xncCA5jgb6X|s)l
z<3W;2yIMAf(X;|*TH(eXh7o6ggTf6VO_6wqhhv!AD_oepYNy@je3n`X#PUa%g&^ix
zop9Q%dTsNZY`dS#Nwh?QQF4!8S48rfiw9MS<iL@NPm#LG51-Q%@enumiy8h{Vlm)h
z_oD2^Z6`Gu?_%#6qArGmoVwzqN+5FlZ;}Rh4v_%Ep}GXt_tx{M66A2Fl~s@PRzfW!
z?<F4pxo`GnQ0DhlV4CBim+epzOZLd?#_t|v9pXa4Fk<pxk@JN)m}Yf<3L{tIn$iBL
z&^YI0u;tSsrCoD>)R1U}e{=LJ-GxtWDkjrwX|W8vz3+N^yy$U%LmKFv_2@E}N^D4>
z;d0qZDF8jzZHr1m`jKk*5R(yT3acof5RpR#B1?^|>+xJ~BE);1`xqZTj*yp}sEq+n
z4!r>`m}F;o`MRY?{PnMncH+g;|1u+92++k%{&$}UG{;+1>^)(30=ByY`phvPdUn%o
zNC}HscTk!$GB8~x>f6u*yn)>!Uiz=W-CB@5-CF8pu^R3{kni}QZ4P>ok*Z=-R4^pa
zYT7(5gJw;U>Ut<t_SfUQ!(t9y3r-3{u^KmnE_B<g;p-m!ikPoUxAeM2`iymR@kEZ>
zo&JEj)Yn^sTC@caJTw673G=!h$I~`=8)L8-^SK=Vld;`>ZuvgVWpnu67MK2oIFl+p
z1i~F!YqdLU^h!s)q&WNhPp*DSCA4{394Q{&#S?b5Ic(JCq}Rh8{tRwB<!-47rv>L(
zjv;-@ik<ycp(94b?Uh3djY9<ymMs#3K*uMwKzE1Qev(WrD)R$5k`*ORUm?lR!L-AW
z&E1s~IoZ0{{|Dj)X6OFCiYY=d2;ol`Rfo7D1HMTWumv`&IC3kCwc1$TN<sya1X-L7
z({lYne$*WB>oN^d=MBZ=lJrge)R!mBcmXEu`g))s^Z`NlQJy_HZZ<$O!sq1la62Zb
zLpi-b5s<>d$(3JYh1rc5=^@iYMWZ<Vof!zfYgsN@5aqVMD;<5o>0;SI|5eKuh?IOu
ziY%@vw;i<;54|K=mLgrtKrFBw`{0|%w`%;usb}ze0>1}<`_AU@8P%pMTl_&7itpUR
zNyg2c$1-H8|GvBY=M#SLsfvvwyJdiZdVNd<o>o`QRgQWFJh(P^)Tq^N1dwOHuG=Cy
zA6RC#P1DiYxZDspN|W)6y^5ut(w{ZeP>qcKG_s1ecpknb8h1kVek1!K;JE!m2Az89
zg4*k?0|h%#Y0O-+8(!G>t_msnUP&$a{`y6fwWr$+D%p~Rqa)>nTHqBtFIec~RB0-m
z!T)}S>+7|xuji$!&1w%K*{l_r5W6~yU;y<)bit+59=YO=bU3}Gqb?H8pp`e=Z(1x8
zSifH~ik#9=Lsg2!WneCJ%Aa0g#o#7Ha5JdVZ-8gz?Yie0Kz$Q9G;m6b837p)e!zd(
zzuH;p^tq6m8_Sn`Co_8W7Kp&6(oMj5HdZI~@jA<Q_1`ubZfnjuB1c|li@I%K{9bS?
z)>u;5YIa*>1M;V6nE~}4U{9d3no!0Fn^?MT_h)%%qf#{cWiiGGJc0EZLuYH?7wCF|
za?bh7R0h)=fI9%R6ts*h8b7#CAItlI381;6F2F%FWyi@SGJP}L0~pJ^%Id(&e^3@G
zs_x`<xaTX5+)vX@fO*5~_b_db&qMoVdE4D{xl+OHR97IY^*)9(kA5bf=YK)q@kG_4
z*y1;KFx^XN<t+rEW#H7&oAOY2;|f91u}>i=!n!N+B_@1^!NZ|R1h$zTX@vyZ%SwP5
zTC<x*#LK&)m;i*Np9Xm2w?q-SnuGPgMi;@oFj%23f2TMcN9X8u>O|H<mVrG5F&NhR
zWGsIuxFo?c;k=tNU}{fSg>|Yuj3nloqaE<Ob852#jeadVF!Y>h0e&KI7{chOuN#us
z(cEBM(9;J!^qK93*e^xXBIwMNe%muYL;e+|OrX1~P)wRmtJX~L_K{axj8bb7x|OZ|
z#vw8qUm&AFp017ge_WmOU!VWm#<PuOFWa`4ZF{x6uxz(%w`!Towz+KEdguGv=lk3J
z4^)q<m(KHe9>*E@8M>?a6bJZMu-YC}4C7V{yvu-o;4Kmf-a6kh$oJF@es=PQ{Pc^A
zqI6Jrz3xpcRf}ffW?@H*&=z_0-UT5P{Xm$6&)gbl)&SO${n2<xv+J5!cX0G$`|1Y$
zToVhoXd{1HRlBua>co+~aPWLnMfopKQ7_;BRkEOrXkS$!&eRCfj&2h_9^{C52oko@
z*_$ZIJ~xhaEJ0fd=l8CU*F#iSgJeTDK1ks+JQ;-xl7}~?Y9UJ)m`!I};MX=Oy)Qgm
zJ?&{As&sTO3_KG=nJrE2v|TiVMWtvb`A@=p<6+FkzAkNCC(WtsUzPA&DZpog{5D1J
zyhNBlJ-k3M$H%l%Vg0-h$qXb8)P>#?NDAxA7*pVISk_xj`Ix*S0_N|O)r+M+ZNMob
z+%}btF!NKq2LH}x12AKZ{hj}~xiE^`|GDOHaN7Bf&!7M#<%6qz>X(s}WJduS-UAjv
zRV^ncVEL1g*wrW`TxTt{yL;?_TnRjwy`M`<7OO%Ks#x%WlbKmD4MZNiK&S25c)4^x
zE5|&Q&dxnO0<6Pa!Z*EmS!*3i<a+cz=`{dy(Q7)7JMG%2d9QnitZIF|`8Rc`>u5)(
z{@^znF}rjOVR~sf$@><|>flZM$cSJU&sD-Zux|FpG*?jKbC*elS6gh|0a1%2*Ndrq
zre@TYtGddt!)<)d>|jcdCbyO%)4|?m*MkA<@Q>uxZ_9fNZtB-nLnsDv7MP8OxCRp4
znFFlut{x&JavzVV@L2j-McO*6n;QwJKyvyy)O;#Ap%O5OI;%@x>-GENHmj+PGJ4Q#
zGDNA$9F+6ha81{jFD6(SvEi%rp0t`v@mlyBHUkokCfu!#^C66mv8`GPv2a;YoQTC8
z0_VWGA|T)fH<2U#y58V+zRbV(PNqhs6n|VO;W7l@+1#J+7?6G&A&ye<qlAxOZvjd?
zz<T&;?^6U$u+os1uI?(ZwQFr34XsI%vi$=_jTu~I80Dje<t{}u{k~7OU|@4K&(ZU-
z_OvV8{FV#70Km7OoY{F2L8?@@z!hr$KB53H_?rvy1AUhgV_(Kn!62Ebp3)6!O$YZh
zB}7S(OkwHJ%<<hy9q;~P45jdo(9+e4L13XhdI74DQM<KG-?J3xS9nMjdr5=)!9taf
zf$vFNr^u85wR)dbhPArObLeisuo^MO6c;8kKGmzmo>vT!p)^WUDA}E}FfbR{%l^zo
zuj$*+OPsZY$ZK(cQj#_P)wqa3AAPXw@e#i^9Df<%;)mInhEMg+U)+I)xX*u#e&@#W
zUF&-`9_XaJ`u#QaCxU?klYWiIO+PJT<+(h@q><(zXkX*kVT8U;g{e=lYAwQ}f7Y-5
zV}yM=7z7w30&R<$ZiO#bgS{(GK69>i6&>VAkW?_KsSx^>l)qx_d>?asOt)v(W*0z%
z2lvRQkiR;C;hUF!GXp<`Zqh8ZeZif)EJ?e7_a@Ls0A^#NeJu<LIsd`$cHVguh{)9C
zIt{Oyhh1X0S>t>2ix!Ag#q?^f@%!ij@Vusm@1&NtR0DPjL)aUUBGm1&Sf?|ro+?IR
zwfL+m1rUFK(P~p&!ly>`x4DWBKml86C{xeva@}uh8r@#CuE1q<t>xA*i^b`165+8w
zrE;|j-*v!W?io{1JD&XMN;%rcjiHTNL;*I3TKhQE!U(`_{+)mB7XqFF08{?Yaq(@%
zwQ+SIe<lBz&<Oj~az@-d;BKl+KF=Di3zavmW{rbkmN$%cvQ;@iMwe0ly1**WaUG7x
zu0<@*j=!^vpx$!wplvSQ>=l*<Z2PHT|IgYlw=S5bvg}2J&k>F}GF#kCKgsIsS~Ps&
zoatvNt^`&2&@M?@#Kh9WF%qmFEGIHDOUt{OjfBVC+8DVblX6ZZYx`O05f9YYot41%
z<O#K|3yds|9Tr9tsvlD_9Y%mt=*M3gPb;zXHDpS4fhb<sobBMo9kgpdaHGUmF3kJ=
zUixbJf7hkwa-Lt1*3ZOTqJ-ioh4~aDUnj#^U3y6i&k~bCwnQiB##_vtTlrkF*HClC
zT}B)$kEk9NbGy@pZ@}^4duS9lW*bWY1s6Vw<offKEt0S3H@~qOF*}MxYych}MkmqT
zC&nOXYy8_XQHWz`&~m`hRSKhy+ch*!tdCoTw_CAiQ(-f9dMmSWqWeLT>Efl(GiVo(
zr0Mo`U)(kAY%&4UrrLkSTIp&s$^$`KK<vW<wN&hn4^%%702;$mC}~&r)~WS=M9k4b
zqs<np`XterM<X^{YNvNX#WS}zC9pbg%31;uh8)2>#hmd;_3J*W8=)wxb141oZ}-df
z51`G_*ch;Haq>6S_qN{wVE@@Y2sW31-solGu}-KJ#q~JLdF_;t7f{~_y&su@m)pI!
zhqxS0uGU|I)hmqwR;FxEh`PZ!>EmCal-%%1WNt+J%4qXE4Y;^DJfqMHfEfnKj!iGq
zn;8Pz&_Vm-wfWi~FU+hxf1Zwvh%Qh!V8cUgsc7eD(!eW)!Qf%#6#F=eSJKNG6mDDH
zX?Z-m&npSJk%I;Wiz@2Kq^6ZSW;;z+vekZwV>Oxf{eB<t+G{IA0U<seKu*x!Uf)Wk
zga<hd4FycPex9{Ei#L{iuUqwQFO@ll&3bh4QQ~UCvJs|p+T6QsH7Y&fUzNCeu5T>2
zV}~*H>?eDFm8XLna9EWz*t?yPM5&b8`FcxUnw%T2Pj_=+Tp|+L+Hqjp_7gFnwGwJ4
z=2uqg<qNwr^8%RMJbkZI-QRXj#;UV~h2iGljL<L1!psKvmH&>@mYp{%+eP+9Vsdmu
zrYg7&Tq{`yS@E6nx*Q1JTotmRw&?R#Riqowve7uvT*b!Nl0O91R*aLzX?=%yakavY
z`050`dJPmYKRQeVrSE3|W8GAYkNrZ$GWaZT8c?cov)T69ECr0lXNK-$!AT}xg(7kE
zpBF+|+6U%26->G@Ho7a$Je&!#ob{Lh#E;_LCj>RvK@q7x+aBmnX2s^*m8xi&$tZi(
zF3t1_=XVtBKcK2RwP?~*@*?1w<#_-Fb{iEX$~&!|7uv-_8bl<&`p-L|=g?^}UtX?5
zHBn<X#r*z$-UOuKU2b+mP@l792e`s0rv7`J%<U`w+64)x2LJ$?>DE9hsV8Va7G`!S
zvhkEQ1E$qMXCE<riEuPjpV28<1I+!qDRZLmiKpc4D|txgj52qnoncI0c18+C0hBYf
zGx2>zdi^D#_fY**^ta3Ldy0s<(!>(Js;OP0I`dM?C0-4y<UB}N6l6a~5McG7RpWN4
z5uAKL)3){mE67^=<iv>Vy<6bhtgHc^iCg(*ZYv#}(f1Z5Xy@%FeIoYY0+e|6>!-lJ
z20B@AQ00=;1;#6idl~ZTTM#yCi{>odKdOU&MOMXfW~XsG^dyxso1oHa-D*U8T|8|F
zvbk@5=f0^Tdzh@#+fb@%BLg~h-$&(s1sXzxaW^$6cstFudm=(CK?xy>N7?FI)w|Z<
z8ak69j<J&qjbvC=s3?`KqYxCd>QK(>XH~Tnop*X&oPU=>-PXbg!$6S9aHJ<5Zbv|j
znq$b%Bx36Y&#mhDAkS@}+a-#?C403`1%EwzP2expJs1+&j-RAuaFKNLeL6OUbFF)%
z^ZvRADVJ~@<j~Qf^^+KgXb_SAlO;VjyFz?G^*PK;K^PpxL3Ii3ZqYw*(b#rGe7)}b
zDo_xM^#@dl^2kD}NVLf2+3}%^p_=nT8fAC8PqZg%wtwcKMj9IE;H4nFpj`XAw$fx(
z4P0#}^R|$X2MUrwlRA4>!_a6ZQKN6&n$!|aOR_V&CwAKfLNC7v1DoKm<()X{&>B19
z&yXScd&w<$LTz{fINf-BZt4tBVe6E{RB>NJ{)m(MwZISjG7CD?Pc+%yy9>w=0d3R@
zGOvAIJMnNC>wJR^pYC_fH?aiX62MqfGPtFbYV9Y$ftO40Lg=$O<{XjZCrBAKl`g0?
z>2K_xY{KgY+UoOrJ2m>C9?QS`^`FiZ^B-zkIG5?>`#jkVIA1;hsZ)N{pw@dJ6qz8T
z8Xupl;e8Av9GxDrDKjASCY?Sdd3<wh!(F0A)@ArUDj?Z@tpJZ?cXa;%kTLz7`(jd&
zal6rUGRZAtwLIQioPZ_Ob^&TR#RlHOLf=ou>=c=eI!*V>v`~qJQta5NYos)g<EZ+K
z(&+>oJ8lZ7Gpi@bEth5O2G<DExi!Ql{zz1&zq&0!_GwmlXb@ZFMbp@*ds8SQvkZK(
z7S9V@3ck+?Z^GfMR6gtIV8;>3bT_|MihVz4zG!U=fUxlV(;wn?HI0svY~|uP2jPjj
znL#O+BFy34|2Ng}la(xu`SRA!Fm|XFKE+3e&yF<hBP%yx-qZEp5=}LYdFh)}U3VLE
zH=OtLMB(*^oPFHvD1oFK!1Bi{^`$%U!VX=9EdR<OxO7{zzULe@3K-y2i#kSkr$PME
z&4YZE=g#X(rW4699qCL?K6{**p{F}qO1UgxO$GGK(tuliuh-X(`l=G?PIIi-B1o9z
zFZSLXQYK$~wgAUYe$KsBJBzQ_p30}XzzbKbN1+y!@~*G`P-M};7Hz#XpRIz8c9n3A
zpon%JDn7yVAywGTT>CkZooBa%-p82b5YNpn<j)zHKLk{+y4$Zbv<&hH#C=ia$>6v8
zQ^R9LLL+lSXQ40zbHckVlQVBeOnbDwXXV5j5LlM!TYH4lTK3L`O2WH9J;$!5)#WJ{
zEJ7i&l0c4s=sV=e+K(G9D3lhU6kt!TRe0w+=7t$Q4X?i?WqEeFRO@xI=6TA!<L(4N
zzV4Rj`#D12q=4G2wY<6ri~5!7iLuG+(pT|pNW37hkrYh*c$_xGbWxfF+|l=$DStM;
zh&^tyoA%$T%-sntn(wOruW9y8uMwr=vJ+@ye001`wpINMkE_`UKgI6mLX30X1M&Kb
zazQfeV1C+kARQu{Vs;JSU3268gQ4@yVz1=)$R`_~$JvJL{72MtfL^4OOT|l>;GJ(g
zB;nu4DlHfOkUk`$bun%FV+B-R(m*59CdkF;Zq49#aHHF0377-hxn^C=S6xy0wj+T|
zwVSd7^r;w2G}IP&6~b|{+6q?47DOheU#$F0jv$7h@mTm+?jOE0)OFVhO@Th2l0O0^
zw90_IS#DQbHJIx2XW!Z}a%bl?26A0$Qvx||1_f^H31tp(W6^<-U_=ek<F#Olj~o1d
zEnc$E-XSGaRa86<wl%kS(fIK6JkAw12hq#Gvy?fv*XfN_DdyjEZ|9<ocRQW+YPp@y
zbvW*ygKodf{t^=FKn&fug}cR0+bxYyriwIm**d~EwEn3bL7fdVlou>?3E`Rwuz>}I
z{T#M=aL#p37jwiG)_bxxUuK~{I>l138p4B7Mmd4OuQT>Cwz{#*IwK5jm7N-KqvIHv
zCRl1$+?KimOY76mF;Tr0v&#+7C{%F&9Nj;RAZ55{Tg|=Zb7=67USqZV4sTiy<X3W6
zY|<2X$A9yrcG(T1r|EYJ@%WCk*3FWQW0dANpGU8a&$1TO4%@PD7kDyGo0La@$P%vy
zJ(@QV`qhAymx5~gi=C=59$})4Iuj&X|4bRQB&6=wJ+nOL_Q%vm^-5+f<Flqi99!?A
zq!qS#UXR|Z29_k!3pd9^!tw6{#_?C_s@4WODfC5t)RwqJLe6#oxtj&u2d^J$0rd3<
z9|R)hERmvK=Rbu>1Q?-zc0iAc$WwBI!pl5@`eH}SL*X$sv#W5nuMZrJfZqa?>i#iH
z77zq{J-d~5Mlg<D=A@%9Sm6{bN0-LmYr8%Kp|kv~7PF7L=kVezfmU46h)mSxDIN%j
zCDEq%dYoXfHAx#}%-I*%udY*)TFI7<kfs$@lw2MqpfLq2oC4CW+enb9$9h4B@;B4V
z6n5DDM>_!LAO|E&@+poW$;f=un(N$80L29mh_&GG>2?tT7eFckMvZ0hyBiYjtbix0
zhdX5e5TzydE<9GA6^w>s5M~YSM1t%VSk-Mnq%=cpA1KtuwzcR;N^zcl<gDq8j?m^e
znpwk@%y3<$|2vG-Fk&Ydq<%AsCtt#|jrOL;OJ!5zpYL!YmVmBBPsf_5mt`0W3WcoH
z>Kguaw@2zIiYk@l97nl`1lUv?YZDEyP2;jMh=IcTdvo`oDkJB}-^&9Rlsn8NtcpvZ
zzRGN}TLqjkdA;3o-C00%v$Kb|cN#E@;d0%5D*^Fk-FsI7;2|OJ2dHeu=J%wxsHTr2
z%p+0AJ`N|cyFR)_A*N5tP}<<hO|%E7ip-_es>AqjcBLCjejQRjjD^_cq-}-QXD=Tp
zPW%WMFkMOpxwU+V@K%#mF5M&e=I2RM3c#L>Wxh6+o`Cgwbg!Hf6tAbN{hlx7r3L-^
zg|g)H7_rgY@*GJo<7-zZP$RPJR^HAn;Ui!?E%R89Rn3)+K`PT|0`PPYU9*hgC|HUb
z2N`AuJ6f`t({?ys7?Wc;ZT{&B-mz(~!J*!8T-wX>2A$pmaWA3c8JwASRBg!YI^ayx
zWYqJwx5`wXhlXgB8SOE7a2A`6+;N(No2jyxMY2wcp1^}Y!`OFc$C<;_;KKXVZGNua
z;pYT4Isk|6(!E6lG-5@;F#bOCz-Tz%(LY1@Iev&iROUGIY0o?nkg5p)kDr!d&Y-6%
zQ!^6T_;ootY79{H?v6{ktU(NC9|5FLhxH7`GPYzEZ3s&R<U0{;)kk~TF12N9=kP!k
zw=n4HC$&<}t6}xbY?PeUmX{D@>ByU4SI_aj&)fZCE)WaFf7zL;bq|gB>4M+U|EWQz
zf9$?3-dFyOaX~eB?x8~rXNEQv$&S9kaBolhy%o4r?v@IFuKS)=PB2#5=gZ722JU{c
zoXP=ky^muB?($Fz+45Jwsq`+uZc)Ya2V@Cc>0QLUWdOVSuS(m?)~71uZq_z==ICqa
zFpZ};EV5>)lQl>ftCNQ+XiFUn$a(6efbHXqsN;-Kib}b!eRH8sdQE9~H!EulHgOb=
zUF`rv)`xRwY;-qtHt8x$0ZW{qdVy;eUz1p!*w+YY$4=STW!efan(jdC)Sn(R>%fZt
zgv%=vz#A<<ciB)D-w`z`KTglJivZlU!1ixwwAvp(1TNjfY?liipJ%b>-&uV2x|6_5
z{rqwFR$5R90fkDnGU`-w<EM<rPML0%_eaMg(B767;L+cpO^pa6E;LPzvrLn(CS8<e
zqAoF(Xxul40{KJnXJF=wGB!=n`bTT}`0rDZI=HtM!A9((OeLG}!M~A4brL2b-KEql
zr>pV2%gQ_0cMHi{|Hf;4#YnVO%oW=0Aw#v>sEbj@MSB=8UjeGbT;H$b8Gour$|9Z1
zgKvOVOm)XeMvF__{Yf)r;C1kGfOt*R=PxKiUm9sh6-QDl^copE${__ILy=oRM>V{2
z`P7e7dpMdhs!@eIBV9PN8E*f=h^XgoAN1^PzuKes6&?&5p?U3zSt)MIM?&F7RRw)#
z*=h?|9jlz@EIdvL%cWi$g`cVvU;!p^x4u5llVzcg_xJ*71)`8JOh%n-3E27<arrk@
zT2}&_Ocbq<*dB=j{I#>2`eAD%xEu9T0{@yJ=P7!S2COyoH>U1qOB1YDK!Q<#Cg@0;
z%imd~qaF-uT2SY?Ij+S}^f_&U)}YJl)WPrNv%%tjFAI04e}qAs$Kouj*8eKA&5aJF
zXu|uZE}z-^gG-t;$T}m`GwaR*QNAy+*(d!TeLsxv_HW<Bv`FVLu&h<b8}m*2enlj1
zE;Z1mYDJXe;b&)<Je<S>7cn5NNxnC8Wm|u<Jq=I#qNx4GPWlQE6`5S$59@1hSDp2{
zw9gugdhBB=3@Q?7_|UakB6`%3TSJ9!3m9Qp6Dr{pVHLOr<;z=HFznSh(JUF?9Ca^&
z|57TW9+{C$Ud8X$^Unh#Exs{^V3EIF`zp7gW?a`fkXLK<S!`;Z6urB`%PH{K*o^Yn
z#1PfGxW=DoLweFn7@fgPkPZP9@)%f+Bwlu(nb*>><+;zxnYwmH$C!=kAO39hf3=*K
zsauoaKDUrd`8a#%(QC9%d>l$fhb)mHHElxCd>T1|3{2MaYuKBkR6Byin-ukUesU4F
zC#SI3L)jgLmnEzraYm?~B%(n~5%4b>D@Eof&0V~1_n!%ABI&6aQH&kT4N_;@(4c-~
zP6dSaWS}SQI78{gOxfa`_j(#%2q3bkA1COwBvKT|WYi_46odvS=uwRU5UaUW5{N_C
z@?YfO)fJci(w=5Ymhe<k<vW9vxvI5)#VSJBptsy@Q?*+GVqXx9;g5l$R24QBFrK}Z
za+cYh6`&`G6?oBTGvX9<DI<ni92<3u%I0d+UD=XbN3N9+r71-58MVv=1cY?FMj7Ls
zD`avBKOC9;3%>FV52{?<q9*P^_=c{c##Cd7FI)uNM?*sV4-1(NHedNn_*qu46)wkV
zVNr;5+AjX^ID&&629w0Moev)vc~x$gZVG8+kRW=ne#YOj*uP<eBqr=FfW_b8Qc2P5
zPFqAieG~S+@*J(xw4h8Q1|lx*G63VPP{)NdUj`r?dxLsl<vZ)_4Imi^yJ*lVwJ~H}
zEx>#dG$U<zb=Y$HxU7V#&)4P|d*xGF1nrI>;{u3yzfcJ|>37vS&E@wE?V(cF*Gr{w
zsK775qLM7v+BTWAK$}POaqXhk1360)2UrDz5y+&}X10B&cAi<qc$tk8^Aa0vYhRc!
zm%U~Km1fGAmyzp#NT`ErbGu>m$L*LZ7eFAm;*eh`7PZK&%!tw#734bmbgsvTLGWWM
zq)yCCwwIky6}}beU7{c3V;w`3SBW^7*1IKe=_eT^O5wX#rOq{aj?hmoZKSvD)QS4A
z335AZd<i}ouHyb?1*F}iK3w`OW((tRVj?JaDwwyUHYm6egQF?B%Jz>blhf5_Ayn(+
z^$ky4t&-MdETyZdS1;{9OAZM_=P<Ha*#40P=`Xh*-~b_GHmxP*qq@me;;c6!0i506
zYIb^F7vh}_-*9k_f`cA_n+H&I_V4g4eE|l_Di`l%Cpg0EQBqCSEUWm>Wp_TnA@r@^
z5bx2c{ni1Iei6_p5&GL6>Ct@pkqN)5tmo+62ZIH~A*tOrDo5jp9AT__$`>f_2>>WR
zp+EkjSr7YqgS3#)*Ry)F$1})uH%327`lgIpqNB)c`z#&>)+;(+@laKE0{y&ljHt9#
zg$SbuxWB={emjdw0M!NPp49jv$Z=Tkk5gVpZ8Me6cm^?d9rCN&91Su?-wfH;`bl`a
zOe8dU<zfRgyWO189>HR_0`gIjh}iaGCa^a=KPjR%_ocN^_(K?%i06Y2q)9$FEp%S0
zD+uNk0fTMvKW(_{7*Fx#L(>XOV2{N62qzL4Yqrs1o0?_xttH0<JXf~GB-~8P<F0Tx
zl3`J=?cp9MNMDaC>cP8Q1tPP-T$`NtPi*FjqsEtHo=sC$BlOYubA~wRP$<LrtE`qL
zcLwApkG&3%7Y%V*u);A-=oqQw>B$%wbtZTiO0aOjb{rx`!Zh}q6Fuueu76}ySrgp?
zL)t&Kw#gn*)7d<~kS<{~lW%Z<&&DbSrJhny8amGnhe24had|UCQ<uw8x*T4)DtG5f
z>MIc7fyO0=TAN!<;0p>@g5vPmplvw|juzBQ*A~DNx@ObJR?Qf68}`nGe3(f*XlJeA
zH)M71fDDMCHCB)nunfMP4tKuJ(kD_YVfP~hJVmd{yAg-z*-(K&6ptX7blL<^^6M*X
zh)+pD8wGiD)Q#iO-{jlnIJVpOxD~{g2WwTc)jEgiQ<3^eO2K1nh?1x6snswvrI1K&
zxnljG^pSY{i}lxzn%1hVgTh^LlO3>Em1|a68ma0@y8*RwcW!R0Rj>>R5)gF|Y0x_f
zuP@U}l`qy>9jo131}oVj4ucQ1%;AinlRv+FDL)r+M=a)qEdO4+K%u8!b4!29(bSMV
z-=-8eo+f~raz4_Hm)8-8sP>jToGo&*&g}2V!D^x>L{VbNLZnOkxOxkbLFDW+D&|h4
z%8>dvw%<5;7Ku#{m?+DwTkYV(MEMG^Lt^DdHeOJPbJ*^SWHC$jurRdT<%kpi*e-D8
zS|@k#7PQcyU47=OkMIfaXVymYlG9t|hmGSqF0Zco)DZzuRM_T*#gI#S(OH&$X5~by
zd-3DF?Xu%z%Fp+`lBsbv@&8r$N?mx<-+cxfjz2(~A3iVH?k4y>MbT+Ueuo~s0}wgI
zG0Kf`_jEqnGD_Mr(X2|sE&!(6{(OQDS?MzbOL_bb3G2}Q9%>ZhaRfLyKie@Ubfztu
zM3%W+P_sXxRYxC+wNRUq&8Mq9oC`Ywef(GgZo>neZBZ(yI1`kCFxbTT8qX~FPZjd4
z$&?G5?PZ6Y;`2xoES!C8@|oE<#mF{JgBln^s11g+2arTI<H9^%g&%f{kKJDZGdxDw
zH3O`zXsg<=+~-UZc>%OMCa0f4gA4*UR)Ye-bGB=H<#gbNvQAm9(Gy_ec%H5TY1gO&
zZ!I8JuT(sL9`lTh@u=to5m4@^pkqq}m~S8ueumxUo}pkrH^R-Giq`p~kHvgy!2Cmj
zc9IE)qYp(QJ(w`|_XA*s2hx3jP!f}a=yZw6kHUqA$%o0cE7&kW?uQmI8Ht7+vASCt
z<(6?<+-$b)DSyO}?MP+e)x+&7jE6i~WJSE+Cd*Z)y@5qMX}MW}=t6}k+pTm|KMX@+
zaYFun!CMKn8B5V<UL6+n-~F+!j-wA73dKQ`U<xXRBo7)E+Rp|g=QK9Y(b{6n@C!ba
zAv$15|BDN#ag5im|C-KJ<E2aU+35Fvleq7=z2o;SiKKlwolTVn=;(U5@-|pLlZnGC
z{>@^$Jdyn?s<2H+*%^6KjCUKCTk)`;oi{fPz~kTfD5;=x)$Q!7^-?zoYj$Vwl-!$d
zqi<#_crnSQ2PtKfOkOv|j$o{$4>bUUj~Pm~IHXsus*<HXk=H|o8kC7!>{HOPUm@gk
z`Lr(xWG0Ng4O;-?=>=`=0UrV&-lO`vOH4quWt?99(c>~!1_gVrUl3F&t3s4#T>&P5
z8*F-@7c|b(9e-}7KpniBkSrZB3hwFn>cc@Zm&tkpj<OhQx5131W>Y)o5Rnl}lN~Vo
z&x-U+<K1Mn_2}@H(;{m~{oOQNeI_AE*w;5dP&X^436M$;s|2gLE2x7oEAo*J+wKqa
zfOij|V53yym>h?Ae}He3Ks$-2h4|)&J0@Bj($&XDe6R2_py@o4iKBNh4S_GX_ef{r
zO55G}4miQTV8AS<mYs%Lh`_9ok#THZgUc78^6n?AsIuPr9v4HW{Du&dif&4krXP(x
zq#R08A5Fjk7|>81gG&4?jcc;y$*kFQty&~TtQG#-IYVDbieRq9IEsN(d7o5}4Pl(K
z_{xNHz%mzuv=#A<GS{DTs%x_-+x*v#z$JN_-x-ExSbTV#L_E-Yopd-Z0SL>4f|=Ts
zFG3uvk-yL^LsQEFLQ8^GsPFGH4)Q57H@iNzfK)wCNw2+aY4^+J#rO#l0oOFY;faaL
z;&V6`)7zA9Kyt})yM-RGc3Eyzky!d6u&byZSjawsPU2(WH?w$>l#O@#f)TdGUsp!T
z@o9r=zv&B}C^oNIZ;FDjOi?LPWDBM+UEg-B^#gG3G^|KVAV^Qe7)#BKZ&xX*R)gOf
zcQ2M|Uvi8#T!2RgAc3NG-Z$<^>$+mIG!FPObq|T2!JGVIx@A-6Olu*zlEHN6c<fW(
z`%~}dLhTKkuN`3bs_nw<2yco>^JB5zXBa`fgR=k|2TPLRp2vJJbc1^@35}A$lhMvW
zRpY1_11>9soBy^<UYGc{_}R6nZLas)AF5jVvJ0F@-`X?*c-w_eH`%Tz82woi9i)UI
z^UVNY{gd8C&8)ZD5D9pj0}%uON5}ITHylK5$hQ|Z#{VQ}rt3&@R1DqSg452|B7{pl
zs!Ds~%cK{}e{!vYW8k<wo?dVFTGt&jwdsz5^wW{4X35K#PJKdT`sJ46=xyC_t+?>x
z{Xf|M<Kh_gC{pMhR2_TxaS*Y|pwo&JuYns=L)^o@kG)4p1<#oZ5z5>p@Gr0dAXmV|
z5>te{9!Pj*5cX}KP@{%#0#(lzT@$fd<gf}uZEoGsL0{=Xvc<}opDc;^JwU@_%pD1N
zeOX|9&}*?!zmdm^(fC@rBl_?R-BxcSB_ZnuDJAJyotfgUR2#0Ru(ca_r@w4e5c^Ni
zYRFci+eQ4Gx;^IcCD2wKrr56gsvWTPwHd|b<1@3HOJs3^a^mg}FcNZQBU65H1h5bt
zBt(<1l5=<`xo564$U#<xT=Z3+PIxZ3cx#1OAmi13`Ck%DYie$rT63VC^^Xx9k(tan
zmGc<wHICa*UjmuBt1I=HT;H;}9Rikewc?*YdO`5!LF>fLPF@d0Xu%%^5j5%=HJF?&
zVC*~+KBx_iL2V8ziXf)jb>aC613B`~sL`U_4A!P05W{DFrO%jfcQA_J+-V<EdC6C_
z605=bxGL0jU{X#bCrI1FdggDjb~%4KAsin8S1zY5U@U8k74xQw=KivP`v8W9_JTr)
z(LqVDT985Gm|{nZl`dn#m=TQVuS&5`&;q|*IO3(9{AC6SVT$|<;UQzp8LWtD-D7wa
z;bqYkJKl1fdzDUy+py2$R1RmDk*UV|w}2|b&BsUswgBrqF^|L?5l6l&vQ?=RuN%*Z
z)Qz7n3df=F*@%@AIKAU1fkl)8{tuIc)D5>`gAG~v|3wewdYd*ll0)f==Jk{Y3U0@7
z_9%!f;Cf;*BI19lD`n0N&xOLnPNt;tT%;FZJ&9oVy3gUaGW<0bLF8?+DB?Zcc3SvT
zSdqQMPsDcTPF3mhKXLTs{a0amRsb-ysr?vq4aoi{A8$|0?ML|n=2O>lPLK%aZhpx?
zI361S@_9WHk(%H_naj^@v^{pOyRYA`x$O39p{=x$4HB)_@eVPd2D-qdu#nVYQoa^i
z>wbR1)QLuKUNdNWbfz;C?}K4QQc>pUyNyB`XOdHT3Lj0Vg=Oa_9{9rBif-vwdi%98
z7b1wAHcb=Jwn+n$x<$>>5<`oKDa{QV>0w7w4tCS6O_ug~``{+0Y-L?PJpl4g8h3^X
zfkaEm5sc;s+Ds4R0cp{vmbSAB6|~&y$>~X}YC#)9R|R|JVd+l8mT~y2k*M{H=OnBA
zor0QYsEdr|X^EwlHY(l{QZP+|f=&~kUBNC;k$u`WrCFYmCjc0%0{|G0?SkFptNAhK
z;w>z4djp_e-X1b0(W_k{q$PJx?oHs&HNuC-I(|Bw&7IZo4`6_5QNq=*GPL+n2Bt`z
z56q-6g^F%~YCSc-^{lyT3s=!YJF+YZIx9}1r%sa9tL^n0;;kJ<hy^zSrX0TkNM`%x
zhF6GS%Cw&}*FStlO5|*#BQ@f>tl*S_-~8BFL`defT}p=CJy{FAJ82VXYX`=gYUL;a
zHcpKU2d9%F!CZ?&Qo!AGKzurak)$2|=6H|In+9W`gGXYC@N#;a0>{X4dw58i%A?xu
z`64`2-apm&C^ADo&q1f>mdq802dfEQ9=!dKGJuup7ZSy-q{}lOtVZ*I3X^h*gvlMI
z(oImmP__tBt!(#;ae@Ge1uOetkIJOqtmpDE+F?nz?f|pMSUpC8-<`*8#DW{zaQym4
zx*7@z@1Qa%tJs&`YS7Fz8Oo1C_3*khlyb8zqW250OQ3t%MTetytoffcVD&%R<RQA`
z@CCTV-ySzFH!ll-vImG$%T*(P6BF(4#C@Sof$+!T+sF$@wFKT^o}T!AalN0dx}P;+
zbZ_yE?ZLvlSeyp<O4~mpSH&#lM7WL*9b@;PJVrgf-E52`7RS{Z8Gp{PUyg)<No}6@
zcK+&gqmn@2)bAScvq|5vU?IdAdaSZ{#Q+f+UwLzXI&7^)f3kdr38`hW6EbY_Y`zU>
zgKAbq8|TH_4wkMAe>cX`nJ2p_N!qFeLqzzHgdk|K)XqvGp3Fo>QnJHv@wnI5sdC`6
zKQD%k+p_?tbK+AW_2>ILNL0O*(2Km4jbW|!7YPR-EG$XpvUqQMr70ds>B)pLw*bGb
zq09U{5L8%ThtkX|KeXOq3z+!zInM%2;)z1tRheY^M>+M?&Su7<pP>+$b+(+|;q0*w
ze`5MbecBTd)OiWsWHlj{UO3EdJQZEZ!zfexj>#sI#PC%Yn`Uh<G)aX}{sDy}UqAz4
zBqb=3X!*5FOAC0_R<}K8Kn0eI2m_lwi_?XRMfCHJsZm293<>`U^c#o*LRCj`*jdNL
z{zSzBOI{0KKMLgxk3H->n1~jeU3Zac_d9H8(2K`Jdl2plhkDJ*95su8@?#cnD2EUp
zc%aSz4O$@F*&;yS>Bks&I`$)FAPRX)KIZcw2fgQpI>u(zbz~aU#%)`fFN22|?le={
z5|^DF;S*^1AVqWNTq+2ViUjP}Km=7C#G5$WBf1Adv{kXP#?s<@!@RwOp}iwbrAf$&
zG}P}qnVL<eyy6SHj+nd}ekQNM09k>{>K(jBZA{9yCgM3U4A~n%9hdZlae)r8eT3iT
zOaAFZ*ZaN0#{&cdkQ*q%2l=u6pLu{Mz~<J$21YUGbq7dy_83zTe!Ksz0<@e=(Cw}(
z4gNUK1HdYQ_v$w|evi)6=htI{mXmp)Df;bP2Z2}9W0{wPSz1LVQi<FlQqx$J!%|j`
zb`-1DK$vuMLYz%LPG(!}8+sRqXlDf5`k?7dw51poBq^6js@;WdO@(Rd%GMTq$o2kZ
zlMZ8kE8Qm|`d?oK^rA6RaZGM-C0aghIP@P0_yJN{OiZRkpCh>L!gV;93W<zxUCcmE
zbD+iI*PhUTQFx!Pp`~b?Dd*(7_)*lTPR{GWqD=T0pZVc~NWa!yD8`iTGKi_|B3;qi
z!Do!&y--@W8mML@ON-6vr9F%`?2je>osmpbLmW%=lpS_611Kzj*2d+yQYDWQpQRPH
zi^5?9hA=%;33?6C;;!o4t6<%yM=+X1!VyDgH}AW)6=A$OIZp+!b)#Gp+Q9j%UHslx
z_kWUIXxqWYIpT6o+*H5GuhaeB(_l7nIs08!sd!CXA5X@(>j<Y!`<g7alCG|0e3yo9
z&1<Ry02-u8e0j(!Gd6kQizyycCg>};sl1f_(m?P!Jq$xgi$*}@x3YM+qC(|9klNwB
z2>rAh=Knr-gCIHe(|9D-Zr-)vJ22ci>#aY{Wxe5GDznSUY}QM|Rl^eNQYxsPM{<oj
zmLS33h#OboZV;L(EU}|cpqx`Nyo{jLB-?9Y8TbRzgJa?En7PIl>eww#<bW|fLe6K+
zJpIrz`{>$zFxghvn}oL=W7W1Zda=6?9<|c2Qv=83GiboY(gw&`M!3Qov_bQ?!Hg!@
z*zKO-sS;U=SZLI!0Z-rip8H@)asW@Bo^ESV$$t?GH~;DyPpa7upil=L3V;0b!+)F_
zQE&P_3Qlf)y&UJQYI0e$FUAu;3p~nOs4$#svbyYPLM{gTI#3dSwY^TSVu;rk!ZdWR
zu^cJIW(D=2#9pAXcGQ?P3Hz(yeCc!HdEf?Bvo7Dv%KLbl6);U4-|CSH_IIL!)30?Z
za+@9=qFhO&pB@%jM&31Y&|>)1uwQBJ76nIoh=*y)pp7bzIdz8rEbSqoR}U1|5h)p3
zNu%V%sG{f$X`sE>4uwHp<QySv3+WJ#GrRFBVskmOhVErDl{p0{gRO`9Y*`zt&6Sr-
zSh*k!Lq~uQGlKErwO}CMjc2B~e05ss_RwA7qZ3KvNLaOR;RLWI-!+fU83>u#zke}!
zL$4;)GRPm?v5w}UC^s?`)D-zD_w%tSpsJxF%%B5x3&ixnX?aJ4U+siuUq993EV>x`
z%|$F|R9785A{pV+bMg_h6_gt?yhu^>U@0gt0*iRP`5GC@fQ(&*g`urd@C5)A8MK>!
z7ga(+c>SU$##Ozc(yR2XEk9Wze-S0dxQ5_C`F)(YnH50n?1}9px3g}3KBy}YT7Wsn
zP#XqOZzlH2%f>=~0*e3>|C6h!;r@J?9-uh@`bH!qc*0Kc8UeaXSze?iER(`r*ZHrr
zvA~U?i(cZ5dxDUagmLll0l;}N>NKd&(d(ydgS(pF0)kf*Q#0T#7IvC7uM(M#hSP+w
z{F#~$wT&Wrz16Gt#o3zAgKrGZQ-!yO2e#Ay2b@du0X)0pO`N0xH^9IAV|X6)R37xn
z&U{UQ*@WBx3Ed8OQG}lE=H?vMQ~%Y`m*@Wuy~l?I-xB7t`YEAIHXG-z+o?rlH3l7?
zI>G<J>%gn|X`leq+R*5J6ZXY>zUE*$KEE<S22nCsL_{H~#`vz=nSv*`Z$~eiHv~6X
zAGOBC`^&sDZp@Ti5*K>$*f-&UNIy^wDND58@-_I0xmtT+2CLl?V1L&TpHnDsb0OW4
zTro%r3CH69?ULMA_W3-@qF_JpWUvmrhMO@pe!SJ@G_({yr8txBIG8pg-<pd5z$Nn2
z<IwQ9o#9w1ALoq5g!*^u=;P2*0<N%?SFxEI9n0vVUhf*&<e_4M1$HIjx#^l5o+81}
zUN-t1ljT;M($B;xY8*@OoSzeK6Mhf!usGFAujxE_il?jY+v=AtD*8i(i^@k!T2dS?
zHH#-X>}X*Uno#cHCK?6hV7jY%UA43neY?Xf)M7BA<PQR)snv(-0~hNt!^73<wSf7W
zGARBa%UIuGdER<G)m7uPOTaf`5l}R8o0+KxB(XZ_e1L11k9?Kmj+6dPPz=_5<BO2G
zZK4?dKA-kZnoEOLK1C)-ARj4JKs)9%jGv3<)H<29^J0**X|cSXrtQR0Rbz45=qoJr
z0hoavNzN>V66u4&SHbHj><A&~7?#cc368Dbk%7a@Q1C~Cdy4uFCS4e6t<C@{whzP=
zRfSU(%^faOJ{XssmKDEQuiv#~B5dfIM;>)TarQ>=3><AG%ql@(parTkFxCQqx<l8z
zgbxD{r<}E0Rgo}=59?0R^39+O3SlWShC|xRyh5s0C~gH<xNK2L1pl2Y4z1f#a}5&D
zO>4Z2NOXHc`-S=&9qu3OW;udC+LxJKpF&QmQq0{r1YjBjz?SujzD9U|{lqKh@irS1
z^ep?Cu@|VUfnLp`kxR@PwcZ(*4HMJ@WQFvm>&BD~f2<x2*5>PO?)x2(mFiof6@9!l
z2gF#8|1;H7+s~oe$Nd|u^kU@d1jJ?lngD!lbTvu=c_aH>C)vL-BG2~(12~c@oavdV
z`#l0AIR|!~tMcSq7A%kpae?y%;-i10x`X$=`FIP#GmYU6$l&msr<=*+jJW{`hWA(D
zMj=kN!NZl|BT=kz&f1dA%$Zu)3Sqy}k!xB{$$lx5Yu$QFs1r`r5Un?uOo(N0HG>lQ
zsq;K$%ffdUWd&Kf<OszRc(mAvbIS2p4hnl73UZ~1eK~;4&BRj&EH84zXlx)|&l5;>
z2;wbgI3Tvdg)nI(T70B={;J(roaw3Y@p0A>k%}eRDEwZcE(*Q_5x&|S-ORij$afN6
z5Z+8t5cmSYL`oWX)^-*}TE~pc!GVmtZlIyl!mM#V-rVl9G5$0maLCIi)Yav3Ss57_
z86G|EW=Z&$uLM;>N=ibCDH-o^W9rR&hoR|??DQ@{w|Q;G!2A6#SROH|>XsfxJyAJr
z3d7d-a<ch|VSh>ex#_Li{4(=PRy%EPcDkQF-V9}hMiiGjwxpGPWFBz8Q*FjoI&qT0
zGp7mmD&-S9?+ect4Sdc$#iz)!8eyNIuh5wEn?X;9hT3gzKIlN;wJ^AeskH+Qn;OWH
zR6K@E#|^=Ig=<sQ=%H787S@!I)Shw37EM4rIQ-~rnk0NJwSA7-h+$?+qNDiu!sC6v
zFI14%{&l<2O@VzSgwIh#M#@~PqqWP<am;P8(nxr}k{0*)UU>l*j0Z43=yiG*g;FE+
z9iO~zc=tG}e3pN}oN<DGBz{COaiQ1%C&QtQ)sMM1g|4#<8Rpd;+i{QizKzKF=e$)2
zuho^pXX%rEVj8=^MTC=<q1)3M@9k1rTCjf8ypAAt`0{GPoe;=!FUEqoC=gu0<+=SR
z^rHBCUAe8u=32M!+dZBXlt2XH3x(Q`;eyjpaL+RbrEK5TE4@^vn2C_my2AO{{<86}
zy-#Ap_*sju4IymCdx2~p2~A|Q;F_d&;!>Ep?f<6zbsk=L<@lc9!9*z|>0y2jaz}~A
z3%DE%f@nMVTW*||>-)gFjo{}zHgN#YUETfo9XbOz@xK>k4#~PhcRoL#0y+N-Mhw75
z^JYry0DSM=-woJ#l}Xy{Gvj?ML`YX3<7JznA{>$1>zYj+_l&8Sdo+xmVO7+RYt;!Q
zYZx@sL;6eTyw9x@m;NTVb>9<fN=k^NjXK+PLuuZ6&_*iS(vCq0mNCV$M|9qenv~8$
z4^XX_@%{FR*oEowR*$ENb&%J}sg55LA^wD!X%R!~I111nDZ7n-{s_M>?S4BS+Fglj
z_rxcpEp;2)^;au27Z0DkH(<W&rJ;L_Q|l8)Azr-)ato_ttkKvS_Q3b?uU3BLo#)4<
z?Dju7?Zp!bCr9#-?nO~<^+|I%kHR$Ar808w#|OO0$(b^{Itg^0Tfa0@4<0*wIvo(l
zXbpfHzGXR%zsu5Q*USSUt;(xi`4bzxMQz2WOSp3X`1{r(@3w}y3$^4@F!Nht`ltl9
z-8wG_`mgl!G>35Pxqps;Q6V#=z|cL-fEJ$gyel)Vzwjb%;!{DuUxjY($)CH=8U?-z
zu1O9+orRXx2zlIjWHwKj@jX9q*6&LjLg+uI!2#PcV8PZn3eiNS!AFn^UyCLRkw~+d
zV<B7Te*0XCBN`PUOK$;}or|g(5@MomNoMWqV2bB&<nK^sZCZ0f`WVyR<d@qCM9wtE
zU8^Wn4Z2fvG6#GOCLD5Fj3K|3M~x<lNhT5|<(>w32c@i|FR0;QL$E7ZDRX&zgHpx4
z9m7~xuk};M4sj}Vy}zj$wDv9`gpEgFvbqie`*@ObYSu6pVR)8IZ?M8`Q0wSgZ!SBy
zn;_w*(#eg{k|#2Z{1xDrU&d|{iuIpBKmA_SS_gFNHEt0&szolVMCSG`*6+VIzDw%^
z{iv6-2R(<E=90svT<=f_rhJm%Fw~}Xw=u@*_M4Rvrfyh1-yoqz8-DPDHcx6kRX$FJ
z3iwKD@5Lxe#GSt1<foucm3qb9l$-9ahT;>o%%8A5FlZZXM-nU?x&&i%-k$=&x21Me
z5JyF|QN{#`a}-C^p|tK*6y$$JbB6X9N2Jp!Ou6cJXd_nC{Gm_=8OR`+0p2B>;0&_f
ziW}1!Q8+4Uxr|%n-&xF5>{)7ZZbo~eAp9|>C+Vn7=>0&GG_LSOFCm%+X-7R-@z~(C
zq%n86135VoCvynjN0_`=Q^{IwcoXk_;&O_ny;d=oCi!WJQ7Vm`C>Dlq#pt?+{0XY)
zMii-t0}0S_=KkOPZE%YAWlHi%=T_%K()K8bg`8cl3n+mtUW|q*874m@$hdg<>M}`I
zMVpUf@eMBdSz+UNP*-!`Q<SVU06&KFC>ewQbf67KR2umxWee@c2UZ=~FHVVL?H5!+
zeC`MMRkz;X<FPrN#x0$8;t=;({y$uMw1>&Q+*~jk*4@p75id9Tz$$-ROvw64DA;H=
zG<m1Q<HyvvE3Gn+d{heJa<YiDJB*?PZOreaJH_n(2<@A~@xPzAor5s(r$VmJ9*O>(
zDLhF)Nz$}$Y&#Ueq+I3>CS;uU@SXx$Z`mdNm(JNM9|kq7|0QMvFCVY#7)+!TL2#+*
zV{2(>2&L=V5>B%mja%b$JJ&DyO|0Q>qVP;_V<vO3nUCkfGVd0V9l1wmRoBBJlU!59
ziWsiXXo6c1uc3xl4=Cb`<@715u~XKwAURPB@1eosqpmJz3*bE4TDxs+MHDj{rd4~B
z(nqCP2p()a<(@hC2>)a6GM)><w$4}+zJ}e(^?UYv8Z(liyPb_<rh***4!ObLDv(<f
z57tGlvL=JY>?B|W`WwHQ3f70FDndmKLVb>;vb7uKpj)HMUh~&JfKY($Yp`3q?<<l^
z*=Jj3hebIXJ!D1ih_wp?D@<WREXcUtDS|mULdcc!We*_d55Z2grcT^p@>6cnXCo`1
zTD;|ivfi@|zR6#pNHolpI4G}?+qiobAwkV&N=e?b&$B%CT$Hhq%2E1LY@8$?>?VG|
z0@6{gCP+H`l2RB!cbeQ9+~?WJWN_tRqhDu$F%onXiHc}Z(`TU6s3i32xC){kDgoBq
zVIrvJI;Ll4R}&t`zp^O>)4na|=1_LQ8Q%)o4=d)@^dmsLX^q2sSSP48Af04(X2h$`
zw=cA5!`rqmRN<8;!UEX`auet6r{liDYLUD8+Q>}@h%H}+qeH|1?y}-a6Y5hmu0EKs
zkde^Lvd_k{xJ&xk-bghCBp38mp~NF$9i51oFNZ}slKK?t5?+rUYXp=cuew$zu2iz;
z?E&!R=)(+%7o~KRhN-w<gNeZ3pvqMq3clns+eQQ`byE+0O@g+xL-Z<dY~q3sLCN>N
zje~2yLV^neOHiiG5%!$j)2H&9w-~`eW8;VLz@0t|9;0mT4e7hijJqO{Vq}DMPCr`?
zf;o5K?Y>MNli*|)?*64p1twUQOq=62cKwOjknQM>ol1aR)TJ6eFW*LlSErjE?<X<&
ziW~9)i2;lwGB_=!R5z083NZTJ^R*V3Sph;9W}tr|U-$3*8EPC8N-FTX>Uq2C{j`e{
zSZ1#p&Mq**DfH7EEq)9@`Q%A#<oC7+T)B2fTV&loTfHeIczFga$L>eN&;G2H;a6t;
zL_YN&{LL^d4Q+y@VXR}fXklleNb#)h)%nAz)O-o%ha+=19qwmTiXqqr&rWTP(FvRY
z{nwnf%Y3x-8m#Fh+$72>oxmd%mrubruC|=yxR?ih4)|2flpy?;tWau!o-^tbXKqoe
zQ)Sy#;7b5sToN&Lcy7yf%W$Y#S9`2~Y)seGu`Q>i`(z(Nu0Powo)dSx`cSE^LlTa@
z&!SqjAKGFHmd0x|4^#DtvYBhCb`~V6ZPuzkTEzQ@7C&heJpPbEdKLC=0{GF`m(}JJ
zM~F7T%TMSsmMPWfqomMbLYyoXPRCjA7d81D!b%HLG$56y^lPO<w&k`<{P>%y9z^mH
zGD<>d%wut5f9Ycb@CznxOU^6HC>#{?32kr;*IpVZkkZOw<b6@zTSJqTeEKujcEVB>
zofp?E&loCp2pVU#kT0Po%(pU9IC>IW@1%-a{j2u4wYKubinhm`cqJj$N$~`_(BqiG
zx<&R@(a|Q7u_P(Gy}F?<X~YM$T9Qmp22O)qn=ot?q_Oy*7<}+xA~~0M*&pI*zs^0J
zNt0q<W*X0xVhKKgQ3@tC9XTiyQ<17YA*|(f=DITsT;Zo74J_K%?_b8;)ZQ0{m0(jL
zDgi(&i-ZP#$r?XNl*Q$sO&k{47t|9L%z?nAp-H+e3r+{!2|2D(;012^4QU>a0CwlK
zOo@Hpi$an?#-;wjyQk9R5<H^2lpQUg0F0I5W`L$Fct6s(D%DHqeK$A5`j+H?yp8u%
z?HGYW;X>`${*~k4=3jmEa^AG*eLHdZaZ^2xX$P@RQ*j$f3Fi)lt{iwvg1X!Q6RZIy
zFG*90T<%A{m!#G38IggRXQ#jEOR<8DGZq2t8jf}ZmrdND&n0KMx1d(KP>NiS)VnH2
z5-YX75a)ta-rurCoM`V_bq1a8PXuC$sX9ey?(s5&#`DAQhpQ+c(E{h6K{ZoLeq{G`
zfsKNW`8p`k;tV^X@Dlryzvxi`za1?b3$fSUOsb)`Y%cT(_(TGnwlVcE0u@_9{lt<O
z0v6c40-U`|(iWVnokz6U&m87fLqR-(oUigyc~^BmWz1Hu(`2_`$xQ+*%{#iJj;uvb
zM8#FBrbU+Sf^PcZr+|DmFh7M~CZASfb@8Impz78sqemQy55Jcff*`<&MLo!ZyTgm0
zzANL%piTd%06i(EA9q{=^JCZ3>r)wxz+dI52vgYNm|L!T3H+S;4p5uwIvd6nvXE?z
z1kqw?)!ELO7S+nD^Sx9KgD_!!DUcpjg=EBk$4ejIp-o*TJ+|lzcSm;4zwW$yIH;qz
zIlBPMl+F@($*8e(7QH$dMTS%7G@stGt#18#Dov_w@Fsp-%)w0f^{2eeybJ~Y2{Kv#
zbB&#MDHs*nGpzu2w$~1v#=uz(;T<uwNPsnzVo<QePOUbiYo5r_{pc8z-}Vy@f7`8Z
zT-FiOm^}%Ps|Uc^6<(6W(1IcOQPw;b`D2sXwIO|uF-e-`7w>-Ib0o^*M}D;;qHdGf
z1@p_$Ag)7ga{l>cw-PD|O(@si+n6N4qg9KKHppL&<dwD{2KkX5>=mmG-TU9g44myc
zkV>gVcp}wM))E61$T@*I$on4ees34<3ih|TgM1yHDIanFPQMSb)IYq|YuHQe$3am{
zpylW9r|H+qb)?r*w<_>`ZZy%|%sFr-1I!xNg>dHURU~X5_JllHActkEAcX(p=q$sk
z?3OUR>28!xrMnw8-QC?K9n#$$(hY)yAl==ukw#iNq`T|9-}%EYE-u)tHEU*`d!Bi@
z!qZ8S)f51&pO>!<kmY?XL6vEi4adp*3(!bbp2B1YqLD$SOdk~g#i~thawQ#CNF_{A
z_GG|GCCeMwnZhESVyXI>^0Q?UD1aH_e2AxbNLb6z#<>+Zj5!}0Vn1Q;lt~aa_{AIC
zQ9=0I(45rNI7>TpAb5(!79qR%^;I27O(?lwa6F90>ZnE{ziC`f-X!zSo26=}6~~#z
zM~>~p98yY0fic}Z&idiHelo)F5TQ9O#*g7vgVyQgNxyxT@lRpNX@|d#^)h5_VwgGT
z^uadK-}1*bnZiAKn|Gv!OB4#1A!|uBmykQq3smd+4jn6TXoN6c!G&sdHJ<Dje0&os
z6qIgbpAi4r`s<Z)o?Xzd1P$Y;$ai!dNS`9ma+KUX#s?r=(V5N*aG3NDl{^YW(aS0E
z#0wc8UcBB|cb`JEImyb`8a&^HJU2L;-d<Y6g^6@^ox7&RdJsDc)OBv{G5_I<M}eeD
z5hUEd`bKjE^x87)Rn?n5_5<k~(24nL+ZMn{zUG3Nqk(|^m=NRIhxn<`Brd1nI|`4r
z3IV9P9vfTn9y%J0;92IIRK>U>=&}7JrACDLNeBkA0!}+zC-E72W&qv=G`{O9BND95
zg$F$bT>0aL(Am|(xhWeVR}Bu$n9mFaU`{j1H}73}hBCkPka7d~->M<nJjRR;>zz~?
zV}Nn{S9UVt;NP;sIHsQbh?-oi%$GUy$tY1rv#?9aF9w5zCnC-^{Tw*wUDlsl$w}~_
z9I#xgw3N|Ky^{xs{i4#DI_zSpj;{kB7i+rT_#NXHtDKC2_di5+Kr5yQoIyB`-`})1
zSDkwNzpSOeJl_VM{*k2}A&gMf%#0MS#P2Atu^)w}#W$ha)<gbW43jo)OkjmI`-`p%
zXBIr*s>1x-otgQisI%b;rp5r`>48_>^^|Qu{oL=yLhE~YSdVyTo9Us+MuZiuX}ecT
ztVwT#SmB?3N>DmbnU~!4Gtn^VK#Y$;D+`}a^5;<)CIYN^N?iM($@x}uF|bz=*WQ&|
zxd<dx^CcdA@VpQ?V(q+~7TuE|M9@+MRK*zWl@ldj)i6%o2fH-W6S>JPpk*Mc%%#le
zU)Y)u<J0x+uPd>!RgI>S`EB;sRlt`|Gn-Q|{+}N75p&!lY@sI3;N?-VZe{G_D2_ZR
z)Y%Gk6W9h|pFp2vxWP~fx*D5nQ*FPJjG+^SL?A3A?7%s7@xg_kRi_-V%(oiRnh1o$
zLljx!Y=q*e&FDQbmieFuHTZZnw@+ba2I~@oZ8lh(x)kSh+Q#ca6f7rVZmlOaA9|Qp
zfr9BrFP&^uai>SkUrgx)Qkd~Mgz-5HaGvi~$yl+}dWLYY80);LoGh)GhwgL(d5EIg
z%6igR6>|B0Ed2LuoD*vc`VX(}GL-~Bw#GcWraihA1{^B2YpC+LMXL6%oi^I^mH17{
z6yOj^Len*QttZz~zai`zo1h*UVg3`5#U0Rgjh-Pk=&9a`e?do{@d{WuUU0$>*qQKr
zyF#}r6k}G1D+pghmcyVUN^$u);<8g@%fyeX;~8^$?JNk}7>d#n1qHvYV+bgoGv*{!
zc#mXz+E9N`gp1EYX09)t087QGz;{bCx551n50^&4rDLx<T`grxS+Do!?<NB4Gj}_u
zBMG6e#9Yo^t&ZdtLZVux1Xp!g-QiHCSZrg{uEbgHsjLU&G%=M_QIP}K3(Q;hx&G!l
z!AS@9DJ)4AxW)sJ(j6slyQ9})sWC4m#5QR#19FVC@Tj6`(nrK{0~Y7T!q!2P;`MLR
zUp|umGrLYjNlj)3Dvm?;79wb4KTj>{ExTqqJgpjC-}PxXfQTL*{DYtTZ}ihB7af`y
zD7igYGZb|m<`Aj`T@U>;+BexdWei94l@`BJNH(^h*B?nY9@b~5V-FC9>8{yYobd&f
z+VW*&=FLh;z_M^pE3+s`-g$^DL7lRBw+;4JoN033fSJbA`C$<fs>}O+<p{{8{Pt`b
z6K>soX#MGfOQ9$D@s6~RVS1-Vr|!$|e{40KFF&J8EQR6o+OLKV^8Fo;=4}B<wi-KH
z_G5(~vRM4`=)R1vEEVF(MP^H|)Q}2(!A2Mgsl+dbQ@|3$Rj5|xu`AoPkCy6<wWF^C
zSy$|3mTWf1F3NM_;fh%d8JzFY{uws_cKi}{{4_l$0?0&VYZSNR3A#}jd@%<S<E<%!
z1zD|tEDUN%JcZdd_cWe|lg|wSO-!<#S8x1#-NaNpK?~$O7B9?&k1h#$%j*C`=peUG
z9QZ(K;@bR%o`EHYK%Hd*KT=nS^w$;#aiEU_I)dswK%Ki3LLbq&FClJv=SrO0cjq^?
z$md8qcnWWduSt8+ZU2bOccvK0t$-tl#m2W}lP*mSpS80>E?12!ogyWT#Kxw7IJVN&
z^24}!3V4Y(n(Pvtqf?3ZQA0tdi9ZK$Lw1cY-+)tg0c3&sX2%oz?%CR#8l6C-d)DY=
z5Xt6SPA3VUdA}bKxqW%!0YS4s`L|#uOVuzi3M1n~;4|F`K7V<f3c@z0yExkwB}cj8
zE_Q`0=_yGhAg8f=T7TKJQKaiN&;B8JN}o;+j=#|k3XC{c!Q@B(-4`WIvHewrIH9zp
z>|74Esg$h%93=6TWQl9$f_rfl+^FUrSljgn+DbY$E&g)2+Rx>9AyH}k)_4C9>FKw_
zi29nkx3dbbvR<uP&EOvquHup3we}bJUKACYMZ?%OUEtrHA6=DE_pcii31-Wfqg&$0
zbjgscnb<y)24UN4+(CnmhjnwR<hA>jx_qa8;2Cy)`IGJQ!(DutIi4oAX$!?2N)#8h
zypA$N^hkT^BQ@Sd=;!oTl>-`RBs)ZT7`0y4pS!VgOyJkxsCn>4#2u-~%%2{o4Fl>9
z@9Rj>`Hpg5E(Qw8cKbRD92W4O)S->iHz9I`eAwTH7>J+34V}lKof#och70V7zM{XB
zyD%5^66nbq7(a9CWrti8^%-D{Uf^wiy;5;DnEW(An0o@L(Qh%`cndWN;AKs<?TFM{
zHS)Qnq9!Sufw$hN730$HpyrUOBCvPLhHmJR|NQy`!aNffs^MUHg;M6{P(MS9E#ER+
ziS^LX<+|mp|C~8v9+AuU&qA9A0}e1ll4VZ$uX&_$jMawsr?r~dt-~fWq>7(DRD%m{
zdwBOa6q9uqI)!f|D2;9+{vKh5S<QdPU*GrsY}~S5906lFa1-;G8BfrJx02sDNzEwv
z<JP0*9dDb!AJ%BsmCu@Pcey#u2jSb&?<&X1NK*>L0do_dqsS8#N<8sfz(#R;(1<_b
zql(iQ4>sOjMY<pNbI2xEn!G6|7%mc9|2vFiE}A*!{U3BXolSe}3$}Rv{$AgQ-r&Xb
zgsS0H=~Ro#={&Cz`jFb$QC@l2k$HR!PvL*Goteb;sEZXNy3Ew{H);VlY(Iu-T&!D*
z)%R-1u)_mkk6k|&Mg7k}o2aRL%xnV|_JnOJrg|<MQJ;&S(2?fUwpPcAsM#GMvN~5(
zK^<Q$^h{ZuqGd22*Ufj9L&fFPQro8}HbfuUZlb-mpq8OfTz8Sd7dK2JWCy%8*pL~-
z2%94>+id;eeHblw?9l=)EhEhV*bLP_fxCTEF^cQ16ek;LEYu-b|1i6*=z{omA74V!
zCUCyrWZvkbg)R~kyL1Ls6&nOS=RZBEcp~;8koSiA&D7QTyy-lC$Kq5Ix$@n+8moQ*
zydgL8`Gt%@?Y7_K?p~a?LL?WB{3ih5lhM<Z%LsEg_*$2!EeIVhDOuadx$X55CvQsR
z{)LGLO`Dwt<<ySf=xYzW&LP|KqL{I13ic*lBK%=}5Q4l-@tQ)T?^veiYmm@7r&6~K
z{5SOUE6gg_yOaCgdO}WKbg?15FUaT+6ECh+{W-OD&0D1GkJN^hpAKAEv_g5|yvEsn
z!>y*3Zo=5nbsmQ?5G-aqYTCW{&3vPhN5~5phG<sqwAcFZtFLo`Lv8%RVcq6h(|dJ(
zT@EDgeuY^7f#S16^!X$oPWjYjZ9HzCC>+315hZU#@1b_%6}cyet9esRh4P~J{nvQn
z&ZDW$*;WlkPev}B3z;Y9%GQ=HAg%&`s+Z|_c{kEj8tNBviJ*(CF+BMKzo+Glz(;vE
zxV28`35LS|f}Os%ZgyJW@rVCBx$e#!+WeMkf5>Ox5XR3&g7}b1a(aQ~*=WZ!SzQZ_
z26ucVL**DY*Q6sD9Yi~g7B!H9bzY^WYQGr_!jN_F_x+~;5FsB}I{`}JO}#5%VEB5H
z)TgiQKsABvovpY&v})f~=+9@^76;|tKPWq}E8vfN-N=SviX+$P#5R8$JbO&oNxnF&
zc3K|a<#pJK4)~m|wH<_FG-EEMY3|4F1Iac(;$vU9((rezt|XEj^XiUsD8ut)H-9oA
zSW<lmK?{Io`#H>0$i9oXXa;v>)IGddt#`!4Hqy~eKsJ5~!NY;smxPfy2Y>8dgRy_f
zG!D9QrJA6aO`5Gcvz<awh9y16uVo`jLIy0(I2>qVf_^29!eN_>W-b)nSs;bZQCs=Y
z*s(WsOU|K>r?LL@;xZ;${Xu7=gPPX$k7uX{su#*SRi=XB?0^^{OO-T)UKLu=rOn9e
zZ3uYk0pS5JX=_b+(}ZDr%Auv@3?@^=>hUT?^nwqB%n5Zgq2GLWKIi$buus4<245q9
zwn*W8x7I**40pm)8r{3;HcEB{#HzaDh57p&4YrGDub6h^g9sQYUn1a}hVZ=4ZXUt8
zLIF1nNBJ?%YMWc)MQPfz);Z!qP-q70C{9#2$kG@(5UEF6&OrIv9+RhomEAQ@!X<)W
z13y<~O#t@vkA>l7ilO&*ur=q2wxNeNbBe#mcEii&Q@w4QDS<ox?ZlGNmFrbAtzhnD
zvbS7IdIpX?csu+ubaWUMYngISUoC5T<u9E1TSz!#zA_!t?^!7x8!=~6kvX(vXG#w1
z8pNCD5&p8IRbn0cT>=lhxXH5@K<SG_CuGE_6))++@|D@}A;VPLw%L<Y*En%96yte*
zr>y}rkvG-#g}{B;|2#tr{NGp_HBL{MlX!5K)WxH4DY8_0g)wf{M>>}yTDULHzW<F~
z(@u`QT5-8^%l+4Q9Dc=*&6<>tk&cmtNm?S*@F(YZWY{Z_M8suslL99{i&iJ+VryoY
z2`6;5v(KZRVPhd6_kfd|=|E>B`f$!l!P_q8A}g0|Cr{sz@NFA~1v%mIxVqiSq~{IL
zb!seD3LoHjmmz)PzuFYIGsf~yPD+WbMy?H6ynK%ufo|&#i{r&7lOXFrU|*LV@QLo+
z^5`mJYRaW6hMuf$Kxe{@(11=u!efEX>kyvNZYz}3dX-q+ydmqbu<+e?lz?IBqvq44
zuYdn68nK<!&lI0+04hYF6cd8J^=~ih^))yRS^2wazq%cZjDX8*>Z1fgrP3sZQL;Q1
z{e({jY5^?9N{Qy@F~Q3Rv870zDnYm9LMgp=NwcA#1I-WvivX=`_J1@ZDZAGs0!r?}
zPZ!0n&o_VT_r$wN-4KJdiU?673l#FMFdb%|N~cxV-G!1iQjT}N?E1Yye}RZMGD%^_
z?VfGJYb=zVsMUjKrZIxd$zDR~IqO~hv<{y&t?0A{@X%XGfSWDyC+UkLX8II?;O<M&
zL6u=kuv1nkW02DxGcJdOpoh?Pc+xJg>QQK|bSL4X)Tb;*t47RQ6GGRk5LG!_4slBC
z7{Vlf8DoS#Ny^{!j&rxc9}+XS)Zk}rY1w_k{eGB@xp!`!h=xv)xZXXX5$&MM*+jJ>
z6`!Ixncq2J4~w$MzH%e^9yVbP(wc7_OK$CLAq4Yymz&f{IepdqlM4ur5(Z_k2>z{K
zpQ-SS-0yTll{~)I-2w7az{$*r=gG=xByw>AmpR;n{njzM=5){k^BHy>bCn8}G8%)e
zA9Aa^?S;n-FV!4!fBoiRl=_raWNAO(+6YwidFjV3ZU@ZEA(ruTDplT?q##M`1$o+5
zqtNE8+Klb@zi==T<9*{W_fJ56Kr<}NcJi&j*T@hh;@qY!zwGKwz_GXc7gY)59tP3y
z<^DWyL*e?(nt3E=M0n^w!BBX}aTLRY9|40n6Ie_1{#funBy2VGjU)`*Vfa(B0WHXu
z+wbtP9A&VMO|;tJknI#6h1m7r1$;-Pt`PM8S%sjmoOUHzjsiTqg9BLvO=9oi@f&mQ
z{7bfBTTEkVZdp7OS{7w%{fk*`-2q}g_KHj-{@disV!WKjG6~o;-;_UZcpn(%JC&Mb
z9-#6GWQ<kobwz}+4IuhG3V5DOTGj(63a$g1Dm&P3_1ZY#@`nMn!1iQ^3UT?MmQzpr
z&Lr?yyZ7}=Ch=j=-PE!R%==gj)<~@hnZYWnt)#J38?^bHm?!o;LG1cbjSF6eKVb-U
z0!4#nqHu&6zmj_bfzbY&YH`?0qh{s88o%Ge0EY13KDb!uYI?vLgYf3XZ3<5=UA0~=
zT^v`!qQz;w{*Pq72lScaQY&4?7`pHdW8)sg6!1tKDa-ex&lm%Ko>b<R`c47i^TQ7C
z3hW#M7vpTHEhijniJ1pchc#G}9tE^6pMjkvg`bXYx@_y<r8o19Fhv!57Y_exrk+BP
z?bhLKD1V5yH@}~y+$iYl;R!Qp&t<7RbEnO~Fvf#u>Yk&?F1*esFSM60-y`d(DhCzL
zjbK|2sd{tn?P<G%YOu(K;J-32?!4;yHMFu5sP;=H*!@02DiY_;QA<l|vVfBzkU6C_
z?P|XsGXCp*u*os?#YqJI$=Q)>X?521V&K;*-QR33t8m`3&vxXv{2eIirxB`rKW{Gj
zO~e{n6_w>%l%h)qshHz>D;-ceZgQ|fq<3v{y7BsJaBmK&$-N|nvAk(-$Sb=fe?_iJ
z{S`O@E>`;`8e`{&v8++y)T>cHH2)p1Ria)J;-g@vD#lMX#NUd7M2JZyAy!9_kl>PK
z&A)V*O!9f3Kt^qaeKtI($3VSzM#d!vQof92)>Xnx$79xUUnJ?#Yepq^J(;?l=Wa9A
zlOw@dNJh4yhhfbL8H32wOoww5xDak{x_=bj6R}%YC86Oj>*TY1P~>5tfbX9%{faB4
z0H1bBrKs*xu2QZa$t~RnJA{X3+*5QcE&7>$Ses-G7Pgnh;eI_Lkui_@-HLV@<b!}|
zrh(O+UwZm768L7VKJkpu(deY~a_;>f`LS{R<;`M*Ki~v^Da`+wl-C(CI71a*>%lVk
zcc4^bORvVRbO543-(NWKHdIL;)1g@dlM%Mwwxdv|4@n0xCQWdg4qOtt1lgZJZSkhg
zcTGDeV$RQi&_KZBLEHd}R<_)OXtpWb0UxKV*BXXsGX$f!Y&ewQaWfM^UeogMXZcl>
zz<K*#>6YDM^+ED1@adO!3HdLa+&@r#9cx5^EB^J$JD#v}FuXb@Ce(xU?XPTQ3upEp
zxB)XI#{j>glc%lxE+)YBf2E<ua`i3`R>Il=7&hg#JX%#RQ-T3=S>Gt-b3U$A)_4P*
za3mEKz{Gy{Zc@nChB?Ks%ZJI;r5Ila|8%8^CS&Yw{EmW719dE<z!JEeiGnv50_VRi
zB8#ChXlaYwYo@_%y5MfwJ?1twUXnJb%05ivQ017@Z4|I5I^D3PiX{eUJiFDetkF9F
z(|>u8Z2S!-awV}~x}~J+BDT^}R~^kNDX1NraO-;yYGBb9H%V_^*h&T?XvnYzlEx!9
zNavqfGF)}bkGY-GWtW|YPf;|`7(au+C<^=boQlC3+2v60q#!6so=iboS3t>jcyKm_
zr4opkmRd6CviiH1VbpGpV39q|{2keRziT>w<3v-K<2P~Cjp{NK9P~qxS~}biGC2%_
z-kXI9cUJw=0hCaG%#8h_d}zA$5;3oO31gPv<1K<zyirT{A|jk$L5^t~fO_1xC3H9@
zi3SyrlA@6=>QvuD*G_P<X||#KN%AL>_;VXH>%esHe-X?E{FArXFlHiy9!U1A#1ubZ
zu4lSSc{|7vx@pYYsVHH=TjY>JBH&DQ@17gDtH4{H!O5h-dz2reac8cEX1@=~-=B%}
zN^PP<1tS>4IQNHJW_8v+#fQcnFY(a-l8PzKW5Lh=Jd1u`&gL6l0PdsB^Is@rVD5@@
zft@uPX#8ZdflC1cU%!k*3sSp&d#VY19b^8ML$lrq{eJ%b+EY6A*AMk=f#CvxsLw}w
z-X;JD6qu<nV$TUzX~u&^wTgIZ48YekxCmTyaryV4U3<%3B>zUF!hVXWG7W|76{2Iq
zMUO~ocv)FG#hfs}Hd)Z~-`l(HwG$6mK3_$iLGWk<IW9xoCjHNr52B<~?GMQr3xHn{
zXb^Zz@Q=HMWMdynuQqQyj<1rnzC#6I4%DwEkghiSOCMu#Y0bfDl{^h?J*zF-#gUGp
z%`T1ED)e!Riv0)`&%~yL`=&_EYN;OfHX5{6=XM_vPe$x$caYcw>;-t1S{K;X{!~6I
z+A-dkYn|^yoiQfNdh`amSWVj5d@4G9$S~$ViGwPe$m2-Fz#|h{U_1W13^*}rv~$en
z#CC3k6_nJUv?Gfo*Hewbq153%94-$QV9dZHa;-8;ieYQ+g%uA{h=y4;MrSQme4`gt
z@4XCRjiLXrh_U-G-qlZVc|AF7i(bS^Zd6&S9C_R&>lPXwKSPa(2d7AQf(0>-YF7st
zu>jh@j516bUu#_f^_|(MarX@Clacr12a9Z&xhRD^K<Gz7vGjH>+TSN~{B2p!()I97
zB~hzj>uFOwdGYlo1-HNL3egpf@>{PSIb6rhr@{&1ae}^Yjd04CilDWo6}&1|<y0>C
zA8aLYG9EENM{UlvbJeMfDPs&6ZL|BGS7cA8t;aq6PhE&klA*`lRl5EU+%5oTIqAJe
zn~L*QPl*@O^=A!V<%$IM<$11cg_ME{d*iS->*jXu7JqHjIy@(^GDdDD*SuNGdVx?R
z2VlZb$g88DQuAW#lUUMqeCwbjd?Xcuj}iKjrB5_tf5RA1LL1kv*Z7}!9ROO!zm2l<
z`aOnJaf)3gXvg39aiqs^)lh%15y{2M51^HLSB5OQbb_bd>{F1R_0iY7M@T}acpFsM
zuOHGVsO9gjU=fQrseh!RkX88A-K+w%fP9Masgt2QP%WtZzgj3-Z&Np22T+3B8j;6T
z|F+6ycLP6ciC?O#NNK8{#&6LITd*=1<*1BMjWw=QQIkqSybz~zH0dFh1o#Kve2E0t
znrQ+t{^^H=<!<5|^_cDekgvD*=?JNJYVQ=#@S(`%s9y<7KNjM8t8k0Ref>R$6UH)k
zwfCQSLEw#!bfIC#b0#O%jb+YfW(fBlH1iv_y|zf4Vf1WJM3(*XcYoX*8mS#jFVXzZ
zm3%YHUCvu(Xx%|gVO+rnD5&KxMq1pypJ0Z4{wsuU`mC0N*@1K^;`L<~*v4M8OJt$)
zEGqNF+@|o6U65pol}XR&cR`EE?|%rT3jR50m2|8!J*A3YwH2~k>OV<-PVLx*oP}Ve
zGbEr)(uMei;%6RN&$`qfXO+<izD$3z<vm>r4w}g;Rx!JEffGGMqns9YhG3;(=?~PM
zxsb9gfSFaC^{cfPf)J)}i;<mI4l{<tJTXO+w6esAOjGvC07$R#ZdT36MmYylG7-yJ
ztGP_hx!Ws1+cN;XVZfa7x7IYJlx<a$HpKC+`BXrw#%P>=2N`o1VG4bzx*E@~Wr+@v
zXTj{yog5J{l({8g5-o)VoA6?edP}Ok2fy&FX&eSkyOq~_0QLviIr=Wd22CO}dHz>J
zN3GR_PZ9wGel<e<7q|hzwFTk8zWyB={ztwB_9L$f<A=<4%_b&Y-)?X~Qfq}7EBD1V
z*Qex5)fb(HONs+3>g5gwV|!aZ-df{#pf+dFsxmVy4|>Ey3vw%hazK9Y9>3&vT6bOO
zodcFxvXSl;ouC-w_%#$-1GbjXAu%fiGftvA8M1>(>JZO9))~5rQ6y^)kp&l5h`xGt
z&we{)(9jVZi~*`dh;JqJzhBp*>cS*Yu!#2*FM`RILU)sTwDi=S@yhNV=FM=s2Giea
ziZ7$TcD1h%2yh|(c+eCg)B^JBw|#*8yBsP|4wtldLkIeS{{PoU{Ebe@#_II$&U+4@
zjJ*PSYtl0dQMz=PR;lJ0dMK?oIPR|CvN^}V6L|8bZIN-Xizqs*f=W9GTm?sD^xV8m
zybmlR>A%Pn3%Jc6uhJwA#H&45Z{Z7Etbi<_#}P(GD45#$$FJ81fY?22eNB)i^9e&k
zgzwxzE&vDA?0IE(nJ_&c4dI}fb&v+OJtSA0cp+tLg5s6*3dMui==xzMqe{3-PY-{k
z4me2HbZ`GOpe8!#w5Eyjl=HOI8E<4asi4?SC}t(pM{n*HevAhsi+mEiV%R<oW%2Tt
zx92ly1}sEq#2CLFfN=-fya0(@eAhV4znta7x0Kz=59o3{D=A02hygz}r1l(|=3xDp
zzX}kirsFZ`<&IlS5&1YCZ*hCmwnYROxcVC?X5WFJYYu~<&^c6g*@oklrBvmtc)6p3
zw5RYCT~ED#{>D9$jBb6=505XU8<Z_1xKIjetuq{lScOi1*x_JYVpa_HnYl?IJ@x|p
zKBU%#>MJ{BGRYyvz|o#NS#}Ha)AI_ZdEMRH36aK;{QFva(mLks$Q`ot5Iqwx9?@uI
zssligg$zWjb)c=A+Syg_+1mca_$2j0eJyd0A(A}D$XPstQ#IE8$*hcV3l_n5Xj9BM
zoTFKHU0?S1Fy%z*Ke=B^pu1|k%oFiz!>k~R7oex0u7i6@gh(F$CxIN1HOS(%rI;yg
z@=+37wVbEVc~=_{rP_8fS{Qum`WzN_K6T09S13VS8$N`X@>v#QMK{Sj4sSLi4)(I^
zbs|IX=ReH-2$>rl6+s|o1}LBp-^!08hb8Rd7csW}_12U=CIGubmnl#+d`6a95z9tW
zlbn}^GU7UYOIXE2AmcUC>{pf5!`T_!arA+}=t02e@$G$EfM<>Q4{cC4D6zshL-u>z
z%fQPI6nz)=x1PvWbg~}5`-c-%*inxcCNnydq%{~>wJeYq4N<F#4lxv+ClBf)9HCQ#
zZdQ+3X(HuoH`)4bt6o-j7Gb+ueLtVFlN@ynRam|k1pyAo4{f7>uLqMt=?igTv<P?u
z)}iX6H9J&;_hH@9@UmtwW^y}u<oy&sxubVCF|)ViXyH@!2gZvS)6HZlP%Sg>fGKXS
ztFvV2JAIpU_Ahz=YE1?x8AUWLKB3v?M^IT<J=Y~T02=~UsN|>&eQ&v}bGe;|o-328
zE9qgHt<SrGCXZfq0Vdi~m-WW}1R3_bo1BHK13J<soClm2I(IjFwSw%4)6O|uU_{ZW
zPquYMi_F(?mmM42RP_+wcntb)b)k|+@D%b&svd`$&N)&bt#!2lJ2`v;#IYGt>RfIp
zwi`!kQ&^dXc?iM9=_MDU4(TF>!F`8~u7368p7aPk{tsy~c*Lo4-*bV0?{0v$4@9h)
zRCst1eXn6$wlCoP;}1Qt_@sCdvrd$a2(+&Q<=>qu)?!(vB5LW88nA#-HIzyVV;-$f
zc*HE=H^EEYs}PkGqw(R#ivb67GBSa!tmq(}YDwL`%!4qHj=6tuE^Zek-wpyNDn}gs
zjHevlMpqf!mF_gkSYnnAH+fm%m-)7ULTfI8ndQH;u9ziK$u+UL5@de0;R#}d#4gf<
zVJujheLUO>DKxs|N>1Ki`#eo8aVOySWoY0RC-l^o_$INf>bB3$w&4!)7kiX9+T8Z*
zx?Q1a;p0;qxx{Xd=AxQg5^m6wBusTpU_`fkqJFDOWfCJT0sICaWeFBKosA}&UsbV2
zg92POzk&SBxJDBj3BR)p+=gzvZ(%5!xjCl&Wqh;wf3h&Mz7jYX?f<xE{9SG;=3d)z
z`3b9HK4K3!31)FywN=K!*$XFE?ZSt!6vpKQ^w6SMN!VoGE5C)Nbj2h_<?@}ky5}%-
z0-(vQsuu+1;7^y)i1mPu2a|4srNBvLo7HJV7Kb&YpNY(2_3c6&gF)8)u=OH5@jaD#
zFkN&@*Lv{#L`8zs?2KqMWQNPc`Q|JnnI#=LLYFa1YK%U+*>r_!ocbHh@AIz;<8+$u
zCQheQ(4X62>P?5fX)<PkMIv^qA4B>!gh!PVdT>v;i-gRP7N}RpkpuASI|k_Ypa<X8
z8Nqa}l%5%O<j}+{HT^9JV2PdVh>h_Kfew_;Rm)@P)$`WVJEEE%sC?O8oBsASjN)jV
z{0Yl?pkJ*$TC3A+|G3h040PMC&DGlQ)GQv&oGj~vMAEO<Ky9B2NvH*;Ssh$oFQ!5T
z^^2(4-c@WcOl~uM4OxTeQFRJ=75}CxQNGbi<P9x;PDdG1!|&@*QW;zDl$a5n!JhLh
z;$nt<G5v)c2h)ykJAopFrTRK*e-5ZR1-^HWlCc@UKF!ZSS5@RD_il$I?{7H3!P!n^
z-^#r~FJvGGRc1)|gmY_-vGr^roQ13A)&ax&+*!T&yA;KQxzps1I2Uj~ljxcdxalzb
zU_-{Bj9b1z2Y>Ly{VDzFmwhI8*!EnP(JoXl@#iSw#plOxJduZ9d~ng?nX4I8JtpfW
zDn_D(^kK)f=+|v~W!vk56Uc_h7Y!hSEmrGeP?C`M6ATqC@9cxl2@GUP9*Jv3+fjz!
z($j%}a5$8a%UdC}-!QBZ^!pTyTUsjUwpWBAg~T+hql!(qAj$06k5b~o@{Q_ufW+sY
zBc6gY%ddCyw%Ud+d(k*~(PRd$`mInhrAP<k+$0)&ra59z`#aJ%J8IB-2Gf4Fqoe;$
znC;h5pOV=X@~}f#pT1uZtGXLLMCl~l8^7Onyg%Qq1`hT&@mXCX=Td4H+PW4{@@&uN
z4Zptv##YyhPBIQT>)#@5A4PG~e8ZIMiKDC_MultQCnTb{=^w_HKv~s&@ehemD2xRs
zznxaQ-LBK=RpL3Vm`1u@W=(+^kI7uV&lj;{EpGtP@!v4Qe+e?014^UY2eL_kF0n=@
z&K$QmV9^;R%T8k>BqQSb!;z(%UOA}z5p@ncb3zWi*cwSU7yydxA&2e%fXGeRXRmQ~
zStR=J0c>vG*1gwSx|hz_VY4E+i?~olcgPs>dF4ZlE|RgR2b@mdEjWze*Aw>2QG%^N
z64N>KY~L+(63vi)9KGTmC|=<kMsNcP*mmI(dUV|r9A2*TvsAW0d^8+<Tx)WwZWV_^
zl0h9d4qq}9*7@;(sJ7k|nmWtj3-a|olh;YIw6y}lnLF#SgB*F{B!M7s_PTq_s1nbG
z=2x0c-o%z=ae2Dq=aF1|AD=A8(_4u788ct~VA|p)SyaqwsQm|vYbu;-1mULPTvych
z@TMP<g0Z1G)C}lb{g9oNQ%afczls0D{5j-Yy#o(P;~!S_bzyc_nZMWSEWX+2>3lNF
z#l32)C6tp61atz~Y}3^Q!$Bh#nH=5e_{a5V7Gw|HeiMOwvNQ&|z)?SA_h0RDM4$;&
zxIErt^g>SuN(?($IG@@>!zYqBDMHxRiKGq&(AVGV-uk_4+8kMcj=<$Uy$R*-M$>Uk
zX!Ah;j=WL()q`BL_#>Zn{GOPBN*-~zPnq&q{dt$~?hicUt2a4j2cpcnukG9R=*=)J
z;AYt~*1mRg-@X3)uIDr0o4#rr!!SpzTS#HT5|%uv`4L}sZFOLyDe~l&QYsU<|NJq0
zwYz1$8?O4}8?PG*tOmvSCLrb^BIl)Ep&g-Ktj49r5_|mPofdFu!0Pw~e;x9)dYdy8
zAvsHI65j=|$m;8U$4hnXlW_Rgrk`uhVn7CXkDhX+uTJl8WdDJ?6&>V`lU3euZxR2e
zC?jI;asKHs;yQM34Y!?7umvX~fD=Jo%@TOsr%||;5w@x-h_Jl|H7J)$kP<sfzkLbo
z7Rxq28^BG%7|`+uoH~(qvtRj5oFMZpG+o@$I^M6W+}DcNy0;-Vvvw)gQ~3jdBIQ*X
z==&7KL*FDpwlWvu<5`^a{)2-dq06^St88*wMLK_AJ~fv3sc5B8TUGMa9MR#IX55@g
z|MsZz%$=Zd8G3oz<vAnaW1sjk7E*Ql<ilyq8UtVMp`T#w+l|5uAEXqfN>v5c62XcK
z6*cRKupN^J>m&@Czb(_5;8H#+RQPW{)|>j84h`w2rO<Q<-(A*yX3t<rm-|<|Qfn&C
zoI>KduM2&;48&0F&Yt=`6NVh|?0J}l2S8pr$~bi!VR5qNiVo+^rrP8wmntVeO&?-a
zxBA<#6T+zL$N3FVIP}IYo5}XdFjIpP@#>gSYbL*ZL!rw&l|d2&Gqd^p6cSN5tT4ri
z8hEIMuj<X&kT9C?_`yiqGy*&kJ@@Nt0O25UPdg|h`@w!@`$xlio?`R{bFTKQb<UkL
z=vU3Y1m@3$>pua=rr<rnaPCYNn8!v&)XC5EyDedB+SS1dChn!QhCf_=yQQR7QV6Dc
zjvwx-;&Q3YRz}ArUrI5!Ign`8g!P-gpj`{7dp>N@AnAZv$zGdZ&coZ)z0xNnbc0xu
zQ)}aKV2D_ZS@jZ$En{r@5yKX=_`^j%3%;#}4Huy!IPw(+f+NkPJyj<v8H{szoS@!)
zaIN5`-N)}>HV0s)CBD#0#6;p@ox`M;+#P`V?oQHZe?wWW{$xn!EU|I#t7(6U0q)WH
zZ@(4~WHbUCcf=`-n>XXn>!EfigKwJ@Ul5I9<^*&e>TC~6ZYT1oK<Jx0z_zdT;8*xv
z_T^tuA-I#;6*usij0u<&#++as_Md)lEbmW?rQes7dj}sX^2`9yv*PdNdfs^)(7wq!
zDaH^HyNg3*fWxqxXj9VT4M2pt2-$Utn950YXBx2vBsLO{u^%u``+&<#O!DmL9XQ$b
z_NeL8IqBnFqwN)WZ-`Z9XIXCGVki}HC}d$rKKgf+Nv4JTz_%ZSAqnJ3L{M1nWA$m5
zvl5>cJg_rogUXnsBy%U3FEPU|Q@X}qhGdtAStkZ@o4y`b4sQV)$il@DbL?N4=23DI
z())jPnBdBr{1C{bm#VaoK#;(dZ%J0j=@FpKEk&e29tj^o;`DZx30U}{9pN~pB~H3z
z<JafnKhOwlKH~AdItE@4!c+F7jtz`{pyLWCMmq(>%Ex@^exh*#_D~Rv(F~EL#z#PV
zRUi<QuI_6LKt}Zp!p6@mq?H-avU-6HbA2lDIVyitZAiYBgaJ$%jgd{u&x0<Tb74ck
z9*kPF8s44BSWLJ4!pyA`5t<*9x-i^&x->qX0MkSm!mRr?&zCo1is5FqfU*G4ZU5Dy
zSv-)jLCxXhEKfXCghHV`Ubj?=l5Hui1p#zg6>Oo_`ruNvJL6QwO{dsrzYFnTY0c+$
zDR05#h#09&Mf3oLHUriwz|~kVV@xUkhX3@o_c`|{PwMnq(nOYgJ_yV~O{P>Wx(No1
zl}<=jJ*ih7#XQ@gZ@SqPB1Z|sxUZ<mt6yavp1_mP?7ics75k&f#lZje>Hf6C1?Pb=
ztjBCD-SE$-bH0KKG;@9kmV~8<Or=@^{uO+~N7}x$huF_*jVXqYV=2Zx5B(?@ElrYp
zq+6X~igLLLzONzjJkXo6YZ#cW3`wH#?-5s#tO_|-$le<ZeG3C~Wj!e@Vp^TEi-m(N
ztJQd0Q|Sc`-8xS<4O2aDPsi)ObaQ^?Xm8?n|HmcacJsZ@SNx~S_q^^l-0je4{n8UZ
z*VPd$>ZSGthJc-4lA6Se?8Se?#q;xghf=3a*QznN{x^^a=!QgNDqxpEY4B~E_++>1
z(jN3O9kl14;{dd-X7;nc0WX8V8Kbjo>;P%Soq`aMR*;g661<($%#hf%m&dN)x@Ks@
zoF5Z3l)}6j$|e$1Mji68X{$xB_$iMhX~LRqG%xpw)bCFIPED!P!hCXPVx~C3<_4++
z2ZR%~qMBNy;WksrJ>-fUYy-0XebUcmY*4HV<0xisIl4sjNi0(3fb(irlhR)tWuU1f
zsT>~|!#g`ee?cxfbNO_Cm@{tC!ifz>#^+-XFlmE%M(h-g^Jy%lvdeI5=1%kdX|)bA
z@)5U;FYU=pJzD_;TJ&E)Z$Iz21JK#{dwVpNM-4?pZ}P)scLLsCEFRclTZjzf;gUL`
zD|yia^MzX{$jVJXRe3dsl!xbfHtnX`6T7WPR}8>7LCVfB!~L+P<I+Q4$s)DGdVN}m
z>&>x^oyZ$t0)3cgw^>o)ZN39v=JP$@OoMO;yLREQgZ3V}wcRS3;*<mSD2-`PuIq{q
z{@@Sw$H`_q0@#mXR+YF#Hs!HUQr?bLr41ddz=0HS?C`(O7Xk#2ICor%stf~g1e0^j
z;a^zS5dcJ-?<xl3QI}_8Tth`_mk!mEt*SDl26P;bW#}6C*-qVaFBX_e242rYfEE1|
zDJ7&AOW{dE`ctJo1RNYvnzd<N{o9Qho8OQCev?MovbQ!eNt*f``-_Dd=R)tbcsdmN
z@yD0^v*ktXG94q%zMfwFo_dd#+_DLDX<BU2#ze8#-QyxwE)n75SQeu~?#!!&R1yKt
z1|QcO-%R~8H*F%7CvfA!rWgbdHTCN4sGyzLk+c)Y5QP6rgWvtDcM`2|A>SWg$=-TD
zl_LLmQiC@ON*X5#`Q}h;Y+Mnih_nu<ZVT0Ljvt=f{7tw5U5PIu6Wi~FwUlFRb+ze~
zw((75*edxA(zH?%qI!@&^Y{Pj4GzsP<UgsG4_N<em?VDb?m@Vp18jgk63z-B--cOU
z#*$DWW27gSian>yz&YLJOn2y`{qcB5vl4)dqlAMfcOx;KRXPJxmQE*L3__5pzmBDv
z>z_v&&6V!KAm5uF!EMRC3f1V;;xZ^d%i3~ct02DS%4QdQ=OPOIn5~=`_+>E9v#E(}
z3a_vEjx{4-@`Ml2pN<%+fk1xK=Pc%Em%x0|9~4Nx3KJ*+rjAs>66vyxHA!#O{#F_}
z4If@b*e};NZq<>`3MFHkoRr=_hMeQmiOMDQ!WWgqfoDn_9M0NWi?m`JAJH{k*rDU<
zR|};36pI+zo$qgedY?P{N6Z{MMuND<gwe1Wh<W*6p9oxvAIle++!fXs@vY|Pa5(i4
zqr1wXj6TBcz>9hSZAMQd7%_HvR8%>eXk!GGB@aPAn`AtWvsd!_J-c3;=%8PpX3R(L
zzq9)#0HZ|XAsV}F*ZaWivjdQ)ClR4=?QU-!pibX|Ns@4%Lc}L-0r!Hysi=<1^^bB(
z3jvtR<gkcl6zVWjl>SkqiUJ@Ag?pzb3)FN+483XF)uh0AwN9O!kx8v=J#;y|b2D4l
z8f+w?U7YCOu+LvcZt1j?#g1*Um7}<GtMRjHAmf_sr@*an7|jYz9dqH7E$l)ZH!bm9
zr~MI|m#RWu!{4f?*+Nu`N1HhdAlN{&=yv?>aHEVxJ_<a|?_44tB=zC+p4USmTgtQV
z;npeOKS}t_Mr&mP=h6269o@Ok^O%M=fWO=So$o);w1{*UO$Jtp`faZf=wjJl`)zNu
zO&s;4-E_xl6XSdpZ|IvSTZmEdYV)~HizAFG=LlRKTaMnVf7++6QPk5mQG94kWw@8V
zgb|Cb!@>^Fb0N3Q-bhlve>x+wTg76yblFZ6i2g$?9<e`koCQc^@@rY4K=of#Lu^m}
zJDb^?tg6%~9J@S;q{lY4d{Mu5W73#6*eyp%+AN?aK3JIp<Z?=DEw={vXB&J?+jNja
zc9Lv$?L@B?Gj#r{3(N+BtAGfj^d2c#R;P{^k(==-fM@XQGILuJHE7%?n61Je7^J)`
zkM~?<YB37|u_eKr&tK#?c3-@pq*8KHt?r?5b88qBX&-HCZm2udNw!VzlGGX|wjewV
zu*s*@%x73Wl-K(+`4W6i2$7+IEk8-f4M}&^Hn8m0iRs-A`HVP$BK;4Ya$R=vQxVD~
zlk{F<+E}BQxv6^TfCburayXE3uykKwX9mi2fUWE7A7e;5JQ(oys+<h^+zCeoKrfi-
za|-$uU^+55Oa$<u!IZzG^q40qZs;6el}bR9AEP3gZvKn7k_Mvfy6*+0<_`yvm+1lE
z%GF7e{&_aLFjen~CGY(gSxxP~=*n4S3qf1xOb|r0`iX%C>(q@tqe%DqTdT}YJ~NiH
zC(myNRR71U;#Fq^kj6$W>lHRM+;VI(O!-A&HA-#K!v=GMUoJSzJGPu<@m&{|XqQ#t
zgn6o@XaOi%MCh$Tw5*`+cECXn13gz#QT{{dF24!f=OuYRSIn0hLs+~h+j%l$&T&c6
z`jos!PwUBUps4zH`!Bf*9&AENOA5kUaJp@9bUM%MG1(f9&v|?4d9T#z;jRQg_(<pf
zhzN`=W`7p9my<c=DSYP@zR)RZDAvO1T4L0vKVLiLde7;)9Ujw`=PTx<*1Q+LOUA-r
z&iDDzEQ)x1zXs$>AJ?^qgh}Q-Y<lLXcvbQt&-M-z3Du!nF%gi`Nw%$_S(M6F&LFpJ
z+n1^Cj{qumgJt+pXS1W0-EzHNu`LNdTPxt}g9{)4M<;zvonIK*I*fAWznl!O_PZRw
zXnB3<kFDvs)62)nFJNLQLuykuf4#MaDQ!ca=M_pPD#?tq7=gNkATk;!Wo>%(@?m6Q
zT^tlLCx63^&I{L(5uF?IlqJ|g0C{V_{TydL2V_YNrPBO}_MPU{{}A2!ALByqRD}__
z9X)%S7er10C&?qJ!gK-XRk@i6eR^t8e#$HEp!p&cw(rn>_Jpjltn8~NT)O;J7sa(W
zP!=Gxv&T-`L3nF8s@qSiMB23^s?Cr=ogeEjQh&pY@b7?uC4*LHHsxI{cBTuvLD=Gl
z<$y?Dkrz(ld+UQdKnVbLQw$}h>re}Z2XUEnVJ3o-aN3}T?Du>Yv<+$TndKY@+m`0L
zzF;zek6aTlzC9?}`U>PT@8y*9>=;QjAD}pq2IT8E2Q-==0LTJ-JR7E!z;fC~ttn1Z
z9f`7e!{rWcUuz6v`<l!5>ojd6-?z4dIX^8M%z0e<q;NNbY2HnYBE|trSXZr?Ljvg^
zlkhWeU04ylfurljZC~d7i}^dQ=u|H^>BvhtlSY5$1fKkZK&yfZ4Q#pUf7wf=)%Sn2
zQ;1;pCB6B+aXP|p^BSsW?A*7yD!_{Sz1E7W_bWet@_I07%#QbH(ukELT!bwiVB&u*
zni>Wvfea^wn$+#sc>oNHpE2TLC9jGoJRC<+e%=EhDRv!35CP87)qXESc!6(53qup3
zoHe~0fQ5$QS<k_rm#u_b5o5nCueW;3aeyjtdnJ#8m}+39vu_$7JWK(|*x7PmkiVjQ
zQt`*Xo>QX$j<%JB^fI0bk^4l!9Hm$UdI}@I#VWU+!Ct$4X-zD!%8P$nTndVSGJRUY
z_(D0a>lDx6Fez4Nk=gqezNIcgf8QDaLUjOl#ikR*P-p{tv}k|cz^eZFUc8C_D}T8M
z)1N7$i+uW}sMi&4$)AE1&fYmm1S3*hl>NpS7QbI|`jeD<D0|}#1qd;g=;7_Lc_*;Z
z;TaexOcde%=+@P{T=XK3<0{=+N^-!97ojOEDMply-NTqp1dt^CCEWyWl=DFg^|pjK
z9_==%CS%wj$RY-$nk>7SLEohmP2<!yLbA<7-ow}S>B+Rw|5<5xl)PnkYY%Q<t<e_8
zBmyBn!7U;Qt_wk{N{7pTCGHoQ!+|vUFUD@_y4=~i{^2@<+K)+Jg4Oy`CMJ1ZzjBuv
z1aU&8O@%{m{u9C-YscD8LI(GtOE*)kYJ9Y~Cc*^g#+RK?X*jMmdQTr(=L&CfQugBA
zee;gfJ4Rbj{W``x)Vh1-8gJWHy4NZ6bp;>{TrNy1F`VM)LEX2`tLv`_yUZG)ZxtFD
zY{r~xpewA=DA%rbN6EEKh|lK7ifAjGzI=A>@)P;(REVL^D$(Yk0eX}x?9Y;*hDa-p
zKOp#y3=D9Q&Ppv*z~}&T_h16lxy^nMx~J!Qr#BFz0ZwH!cdq5&9vg$3&iLxq7BYBg
z1QOQp(l0%P&RVAqH$6u85s~o{be08qQ=L!d3c2l9CE%dduY+rtNd7`&cK3&*(bx)f
zhdv$A{pH7RWZ__Q1NSd(&-QQn068o#qc$nw^(Nr7#H8PxtBDLJBNX#q*8w%^4fqBG
zXoh2v*HFBLz^cM!x@(^QxicpzyC-wL{h*C>i6I9=Ap`AweK`3PQ0umsPcss4+-1d-
zBd&1<K%I}<ksU{llP@%0|Al4ci&CCk_L-6`&uWEaU^FZ2JS>jLxZ7_JI5vRx7p#NH
zDzE)1j0rH>SZUH8O^YeUqqbr84aOEPg}I^KBM;eYB1WE5(SHmGl@5vkaB78oPVYU)
zpWj@)WC@V!vSRhw6F7Fa)gOKuteG=Wk`3t^O?}z%;8{j6YOS3Pci0Ib-7@1yT9MYv
zXEtMPv|rW}3MxlRI?Z|>(9%9=XF1MQ^gp!5R+{WEazIQAvi8JH-Nc*ZR3gycm@4LL
z7!aG?NJOR3Ld7xIHCVE!&5_;l!@gY*R)&`Ri%OM2{YmPDt;%u*l|=2C3!Ir{?5ey~
z16uS#uN{em5T;Qxao@m?lLV6(i=33FQdzTuL^BGFZU$pP*BkuM_UXJPVo!jF0Lm1=
zx~Pt3;_}68Y`xCS(?E<xgqL57INn^|Mj=NILmuhh_0}w3K1<aBZl{?Z8M5&8j~g3@
zO$annud9wyL=2)eEWJ$^XMn;ul`E8xi4HHOi=v_LpFcIq0peaC#+icoh`H8-Amg{Q
z9RW)!PxwtckLv5W-^MMZU%tQ|i@}*qeWahWH>2d_g`LZ>_+yGCPuUr4tNmew^!?*k
zV#A>9f@XW)@oiJT=T`Qg!Knjsl3t@bD=4=)!b`3jLig+bi-=LA_g8(D`wy9SC{sON
z+ZHsfghV3x{dHlA;FDs;Tw8$+3S$yc@gg$r&6sfUK?k1z{JVWWK-v)zIDY{qYgZ~$
zY!ncB7F~9Bh+td~cMA8@|B`bUN#?lOZe=G!H3jn>L~G|%_>C2-hP5?xI0HHK^Zj9m
z*(366hGZd?G|2=!G~X=gqMw>7%CG9F81KJeML9PZGruv@Mxdc35a|6ZdGP+b0&d_D
z$NMFDWRllm6L>hY?BdY&WNF7)f4&?lzt7c{J2(O{y29tbvs>X~{nmH=1f&$_UC;Tz
zFjAyymZbm#p7O6>4jk)Qvb5xli4ai@zO@C9$68aMK@(A=Oo3bcR84yrP8Ss)BCW|=
zz4HVI6Y>Lchuad@;QjBzR8PI`BhcWaR2CtFnEUE7=KYxtmMeAW5Poo?pdnUgW;>p}
zgc%Qq)zgDotsY6qxC&L_V@u;?BaW<BL&^`BstG*Z<<qe4A5z#}$t-NmH)(-k7Lu%@
z?h1yO<--jy>NZ>YqW^O(Zsr?)WLuCk4b<U<!0z-PBBGl18wJe8BiUa$lvl}KrxV$A
zzKNOh0WDa*dA;hdAS^54&y&?@tO>QtF7^L;FZ~YqcsJ}oAFmJ0kpja@KG#gh?6@&M
zSdT7)Km9%5XgLQ)A_%z9H^N!$axMKhw5)y~RZ&-iP_b-+Llqoj5HiW6rzU$_EG!?g
z3dIpDdnu52eV{PjO_?6W%#4sWG9hrqgCpFEBe%(CI!-~s1#tf25T1NL+@D1}6wYzt
zHr0&J{5$-5g_fD1=}ZC7H$|Schm%2TQR>3d(bFf>?*d?NmEDgO2SGs#syYb+LYJRD
zA9<;oZx`;zdk+ehZNu$iMCNmG{M6{?sNvl!3yMcjFz+%krl~16vC7$GjESLo9&_0N
zw4ZWWM0#E1VK2&3CAU1w368i`g!;PA2Fw=_m*38fc$?OyLa#<={fp!Qq}Ur#UX1ZH
zcn@98=YD1<IKQM<#4sn0hT`u+;2-TSSY(>1a!6JG*%{%Ewir`=bN&|AWV>`m`R{u}
zO>^RxsFNteN6H6r)n?Vu$-=t_xtcJ`JWPJ83)|ixEuDxm!^aRcWPhB*L`?hp7iS<a
zOp^Fa`Zny-UcRRn*KUQ@zKn>x9g=l^aYWzfga!to0M{en>r3!mpz8?^pmqU>`!Wh)
z=R<s*=(sBh&<I;O24c%`3bnSPS<)(2@F25VRb5dzIqP@PFDVDlP-@DuTXcw}Md^vG
zhz?>w<t~m@S(^G#4TM^wcyTF)e?1fN`vwTFhp?wFpf$LV$Av6Z#nW*iyq=3nzig;!
zIgS^nlX^;Cy~g)H=5RB7UI}BB!Hi9>Nq`B?+qIPVe);|e(aaHfU{w^HX3gU)+Tm>9
zc1Ny1pft8HI}9d^fWkijrD2Q<aYJ;Ei(%!oQ94e<D7w%vOhzB5GFa`OKpm#gwp#e#
z?`gnNy({YUaR6yV`o*E+D2Cy`tMBubB#4G*#|~Te;c*zZ$_gt^{FehTXYRQ?#d6BK
z497lhf)jEz-5(ws=zZ(}IQ~kJY5k9+YmAQT@4A`TNz*uu?KEs`8;#xAwrw=F)v&Q`
z+qUfn@BKgTO4h9TkXgBN*Zm#rz0W?3VBk)49HkJZK?SrNFL)8afT@gIx#VSVrp!ne
zLS6wy*#fke91e38-rT%r5!n!O=yaYs4`q-M=UoF3wwFDc_e4teVC(@^K&`gGCi9FH
z^c785YxuC^qRvT<u_s&bZzLl*u@SDldVh{ByRV-Pi}$_L9RZR-&cxrzjOEzCjFj3T
zRuoi-@J3K*JA$zlTx%398BJ@7J$|tS+SQ-RDhc&a_9SB#SBmEWgeFVM2w)=@+jz0f
z?c#!EGVXR1VbBLZ@cZMyyyrsMQMpYQ^(n3Rm!d#YTG$)pv<IY+L+9_L_-ZYzdkPTv
zVyC9x)h{X$sR8z|0OBV6i(bTP{f{>&z;A<MC1{kPV%BM$sG(90k}$v&l@l4`$YwQ8
zyfRktD?Bxg?|CR3El6S)VQfmy%LbU<HY8xXF3)qp)%`A59GOLp@r^N#B|}||D}G_1
zBlwz~Z5tRwIZ&%m4=#<EZp<pLYQCM|03-|BsCU@Cehu1aJ`{HLg<BHXKhr~=Q0B>A
z!e!b3oJ+*Usn7i5()H8(_&WY%ZYxg>8;+;G{#zz6b#5;o1%Z#OK9l6n+pIsH$!Q5t
zTaDIR{_wmcuDA$)u5E=Om4j;_#TCBCev4tbU`=L;V;RU@eNzx*j0)F@*$X|cAEYzp
zcUaFgY8(HIpj}n>tGYOu&n8x^6l<i*WV5CK&p;Y85sP$y%zutq6MYDm1Ht1mVJ0lk
zC@t>;71aiy>G;p(Bz*F6+1~CVK!0x;VeC~irr@(mSsFIBP>U8KSmPJF59szE*j9``
zw;?V}G4H&jYIF4R`e`sLK{T!nCodb<JJ)i7I={9xRUR~hc>zYK3!?~cJUh-fPkA!Q
zrViEn?V=Xoa$%G{KyqAQA^9Af=5!EAjgJb<d;#RNcuaGsPvE&x3J}^2_F#ta<|tkY
zwpUwSlWUAy>=C?)%3O$1kszbj+4`5Qz}<p!Fb>*>^AUyc8a0HFb%_ka&}q#VWi$lz
zO?yw%oC{U%p0@NjovydBvOWD5N*IrILbtzDW0TQOYI!u`ok1)Ru9l0^HCb=j11ZfE
zi3k6r!BE)LinR!wd~FO%!1TMHjDyFi`X$E}h3rbjh@C>VH7*AU-Xa8vNAM6NaWm4c
zbp-+~WFCod#6=PKY-r@Ni9<lYMx|8{APVh#?v;K{?HP+ZMFF%@p3~d@@2`xywTTV$
z=<2<kI6;M?`R0u0wLLGozHt8xN{1r<UOG%?L@hUCfICB$|2X`U4`vbgMM9^l<@OZl
zhCae6SMrjANeWklU70eJQNPv-Y7=^apkS5D`+6(L!z2E#^Qy{26$Vrr&QWPdnswU(
zXQ&F>MS~$Bi_aT+fmEwmh(DIf{H0SUEo|lXIy`DW2bz;Li6125yX*tTIQaO8Ec9Tz
z*$y0i?VFEL;A&C$_Qo3vhJyNt6%G5l`pW#FvL~r*Q-^;j%lcjimYTb9R6J+km5l!v
znoGtjpLHp2Q>r`R$1pyep&LsN_9Kg88*Lz{tNFaSx!Q~a70RX|YQ|T4!3~m8`$O(k
z#`kt(!H=8Wq}e7*RncN+Y6gB|fSYp?#ZfC=y!sn^Qp$nD!B6p$%0N_VY^3XUL5mjC
zZW!{;P}uVl86=wySinOQ8Lt%A=DNZwU=7Jm#zS*j{A5nDo9?r|NYTB$ODb1Ic<433
z9p^km?hl!K1}V~TyVE1j+Rzjv5=JE9Z<%5dS~(<skaXG%vxpDqeDy^}Ew`=TSs`*D
z1|H<^gJY!q3x$vg%1=aL-}ps5QKl#D^PpWSbdj>*m3rKf9-F@X@M?|@HWN4}f}O%=
zW=$Dk$m|N6@MrXs&_zrQwjE4)3Fg+Tro+SZv+Q$-Ey(Xub;*+=j}V{B_97v}G9t=f
zLMiys1;#IzA49a2ngo;`m2(d7x}MIzt?=sD9d@{bwxlUUEDIvi(IojG^&rx~l?Y4#
z6{SU>8XS+|S4tl4(@jMlNwg`KL6lVW5IZhdQ<alNw6!{D$n?0%ABOqKNQjHyzoSoS
zremLCWxLoJ-7}x0-(G_R4<lKh&(&Xnc3J#jbi*|ofVfypR64F-8Qyf~T=J!2+x55C
z`jZesaFPXA$f^Xjz1fbDm>uA7AKqZ&@CfI9L-ZL2Gui+pI1l&D3V_rJV=c%M9}ziH
z)Izv}hz^%ctU$aQ`|QSyBg=y?B=|8fu+JbSzhNCFUp{sz`&wFlz|rIvK4oAH+E0)D
zOE09FqQPVTcy+L5x!t|%=Me>YwDyh|yq_MIt#^|E{b-{2l$;zWox&i_DzZV^S1f|b
z_o?IP=fqyCqfli$QMw&EL-;u#l<v#uB6>$u_Me7K&b!QK7>Rc&>vb{l^LgqsUnO6o
z&LsAEAxtz9Q#`DqQK!01Kl6NCK8Jk|Fd68Ptx`Gm7nOv(7d0wPRb7COE`Sv}v=}%N
z3}g^PrkUo2{`RADg;)~<keIjll#a8nzwInW`nv3#Yt#j&Fs{)hqmS`5Voj<f#Z2Nh
z3ljMnuISbm_doG<aYVNm6EkQYlUto;fzSVle#oF8nLwn6zDJ`F&InJ5NxuQpKlo9>
z9Np!ITQZd;*VQQOS7p6|>@_27XcHP1H_m}fOANO;I4MNsN9QsQFJ>lLuPfyM*+RZR
z^{)Mr4;l}HW%u8bi}v|OgeGZ&{~?e%!C)Bs0xhKRA`E~_zIV~oD16u3fI39-A*=Zb
z&`J63xMT7B@Q_I_qU2WiJCA_P|3GpZb_xpxq0HzXk@JGIvRxryBxJ>6+yf;h8LWlm
z2uTDd^fyf0q&ji0&gA}%2f(Zve_@rA1rm~Uq_&MM)iG}e8Va;mL~9GUv$#b=oZ$At
zpi}m?=q$t1HBVUV!mH`olaEGqNQXCrun?lp@++|9rtLuvNfe3j-y9@;2x~in$sYgY
zZH|>k7_2eRDA3B2{Y7h;7FKGL_(V88XK-@Cf`wjR(Rm(lUIOS~Ai$pR+w~f`<=J{o
zkwIF$T=Jl2s*id;2ypSWN7MhVIXz$-LXV4>)jcpBPO&fh+S%o>Fvi&B@I0)@uOg!T
z^2qwPd5swMVFA4b-4@YDhNd7_hWrbu2{r|V(2@qqvI5vhRiz^=)Z^v&%3R4*nDiA4
zZ;hts8`I42!wPs-9?e!9etgl3)vG?sD7l-DEGS#xVYDR1Y_08ffLeLWW_z3w`+w6(
zLAe}NUqx@1h-f%Q7JP;q$6~E70h)HmMf~9I1+c7z-gf}1S-{+ATifaV|Cx6ADDQxd
zujK|fmjhC5--b-x#bFI0@wlUCdMyHRX$>|-nv32ZVG-pwP&b_wkzm6+J93R^Dx9L#
z9&UyRri^O=@+pF!QMFN9zOJ#6PA!bHCPro2RHT5tZn8;-NecY|E#!dXVUaX{NoH1H
z5=+|NH+PN9-Rq@bUXGHp5sbo6R#u$Af1%I_E8*{!6=Ys+7jlKYha*RkxSyC@)O!Hd
z@FUR#vWNoBvGG&T6n_R8?T;#O!pTSYV37lhMW2(|gS|n%c)rkYKCz1X!MUw4?n^E|
zA)=6|uo5-F99UH6Xf+=I(sDA>Jy4E+@71!3{149~i*_S0KuW7d{pB{Sx~%$TIgbr0
z`W~$PT?+GiX#8jVAP7MjLwxodRL%%w7oIZ*;y#V=-P$Q{8fQqj5|rrom&S9CpJs3p
z7sZeN=BHS7rDN@utlm7Z>N}9;433QrQXKA#)O<EtkpQ0(Dti)v$CC}uN8Zo%yKr=G
z)?D+mu$%BS?UTRc%rplTyEN2Wz&hr&3%~{<Iz0OeQ)^G0vb%-4bPG|xIp`=eK*nZl
zG1kX^;{jNTx&t@Jx7zq_q(%T9&R3cXP_3R%gv-9$ab5~F=~fNX932S{dPpMpv^;#J
z*ne)t*tqPsUmkGj%Duisb&Jp@%0Y@S0KyD9t`B&ssl@8`y%bki#2;_Ez*E84b$Z>d
ze!$}4T+@`pgG<hnz&|XsA!3ZziS}ZDY895eH>wu_XbN(&5Hc?qyY88{)TjalBdfHE
z+@p$rptcNM@B$75f5xhgQ!eUkbA%Y<hJn$Wqe)Ov>s?5;W5phsLrLhu5|-iWi>36}
z5c#_L7%0`dPxIRJu|dJQ!+-x16#VPl-8#DQd1~95NPoEu<Mw?x?+*v8szhhEmDG9w
zE{f>*Y;;)ib|qum7ka~F?PkPE)zGXbTV>!iPEa;F9(k+=o~*%o2*!@P<2#$ZEiYx-
z-V2}_1x5(qa^P3zu}@(|)N)goe?ZgWEtm@Wxx*~{p<jbylVh~w4dc{@M43+wM9oAG
z^MN_8jRjxICj7yd+i0~^Czvz=Aw@jKE|qMcnM<z9iXcj=NB7+pD#G@J=|VtuwKR4?
zhp82TNuq}-lDy7LU35<DPwH?z{@kM#7EW(GD%2hs6Ymg;uGeL760I6Btqe};h|q8}
zq`8ei`U@l&yG#aEOp>0I%R{N3Nh(Q;q8W6PdIsgxYDo+3{EzQge$jAcU4MpB*|rg?
z7=;L05srUe`vnmKPCNHWGWN!>>4edYmOombO>|y%0S884aua8o<7sCRy2T1G5~k)x
z*I&idXPto!t+#s|F5)0sA>O}PVK4>}MEzb^(b32L1z-3>_dg60@Fb<Sl{?@T(&o_d
z@a<58Zw&C*^ZHw^|04!vGLh(fA>vy-SORz<5{!@*_|HT;PP{^9Ja((M?Zu2xtY(ft
zp5j)(ewS7NyDU^}%v;$5G7b_lo5W<gSibuZMJk8j5!K}&5&+Lg0jD}(u)~YRm~--i
zmP4$yjcbSiHD>;I;r?JcuFO!deP^H<jt2~GC*1~AAJpvg{ybj&-LI^6iTx@H&a+Xh
zA>Qu{z^HbsZQ@}&;+W6d%n%^U_)n<r#WI?8Xqjm2AU0eBj>4>m#s`^<@FBx_jI{bI
z^+zQSiGz`!is`y8`B!f8qBCzu*4%?vN-i!vpOXV5px6GIx)-w+oN2^oLO;;Bvko73
zNMFgZElt0(5Y8ghsc@@EvW;dtOtrsuzS{8wL_kYR+gNEj%1N)>Y^a`V^?9z2R^zEJ
z0rvTpz#2I<*Y%dqZvSR&c8~g?gML74^dD9KmasRL6tQ~|u>#_m&zu}{0Yezc`33O8
z7q<NGyOG5)Pfz9ntg+<(@xt`F@q}}aM>)QmjB6!;XF1(ZvEmQR=czDmdPt4)(&=OS
zSn0u-x?**9^G<#BZnj8E57xEa&PZHjnvU=yjo^f>EZ>(~<5%pO^veK6I*sCsWn_t;
zpmKLeXqMHfFRh>c5B>GLeYGcRxg3WE_^6PtvH75MBe+r>86j1_{^ju*C<a{rH(i<(
zI<1?H-OeZA0h4A7s9iPYCoG0ZV9Df(AHpWOfm(_lmOgPwxWW0!_efL>f7oi?2K@sa
zeZ1Z-a!!M;wuLr_dcp0t!MCMZnE=jfe6}gPttcMmn$db4Fi7x9E(Q1RjVyWJP4L+1
z(1W0RiwT7;&sv@Ff;<i=i2mXg2#=a5!11dyiUg}KAsj$O*Zv}z7QQTMn%eTTy#&m$
z|5NPSaru!JoWsoY2!{%0+l>c?%lJ4ovod@3_D7+)5vh|OiUnX4#=1R?mJ0=9N$WrT
zg;4azE^!}b*vfr>4$ljOCcXE-)XPyH%=LIL)c?BS8_pMeWv44rhb78e8z1*Wp1f`&
zZ?VyJGwlV)(!;74Eo_la7po75_EO0G{Wk~DF-Iw5Nr|xTd2MHrUC0jE%q4--ot@$C
zknLJ|od{cl_cU>JTav4!@3QuBmqF&ritsjC|89a2un@!(IZEWRTmW<>i^G)fm&~<C
z4mL_uQW!}i-EebKESg&2?&d>olhyEmv&gcjFgQO_QOhb-;lY;x83N3(-Ji)kr`bUn
zO;-O|ivWy$9Ji7z94g|qYRRMhM3BcQo4|2S!IBznGSM*MlQ0DeHxjb0&q*-pH>`7A
znVw}}hG{a(lS+Zn3veCCC4CWI;#3&vKBc8gV2L!HaPo#_hYc;3?YV}oM-a|~2<V82
z)xWJzJOZbCjXwL?ZJ0Yx8>7(0cQ6VJ)8Bo5q%OWpIKF^%{&VqzXw@(;(To|*@wpx8
zdjIF^gcSwY3N;aMN?kF(aLc}#4xk;|3Ee{^w5~iQ`4S6(m0*S)o*A!^q991-?6zjZ
z7|D5g0?Y<(BS*~mH!M4gUub!^sF?bJuuxG4LPpJP`q<<Mem`}$<BL_W<LNK_>ciD)
z&%I_)TmR(+j3+cmezi-eMmdv%lF^X&8$SJh!5Xvak;fiWF83XyScOUImsl+4yoX5X
z5ywpiuII`b{Y;PE*1q*Wz`4HI97Hkq&|A757zY#QdW*VZ#;rS2zhW4V^!rez%>Few
zcc}Olye6<#sG4L)|7w8yivTaCQFD3AZmDU4i)4fdGj)~kMZi}UaPG+d009BmmX)|R
z1lJ)KzgIzQ|Cgnp*Jjz_gXH1*U0h|}FJBon+ns=|7r{qaSD{}NK5&2X&jX{9PfcNQ
zq6g(NzUWIRY}>jrgV5*@UAjSQsr7M5DDY&bv$#N>h;$~~GBHQ+L8aGjb_1#*A?f$x
z4|OAQE?tbF{*YJXH;ehDBlIi5c`n$w(%P{eYxb`#z>IFh!0gF84B|P5{v3?fzI?{I
zs0IYP;{`)^wPKoy-oGz5q}UR_$CqG@xBX$q&IfNiK~(6Un%i1JqYxo38v4`!<YwHS
zReC&&UE#)AzdE!F4XUa9)$yI$e4pFQ-rOXxG=RUOOcNmb89i#B<I@6+P9}~Sw2-KK
zdqFtk_zw-s7c&jc>)eXgJ}oQ3aC|jt7285F_h{OBly;X<a>Ck77C@_t*eP#aBL*ur
zoLTk4G75ya#w{m{cicHn)jU42cEiro*Y<wF09291^5qU0bp{&#vo*#v4>4b}sylC2
zrzW|(3hEHNKsVc8hMxA?9YiWiq9)L#7L(}-kvT1=f%<$FNPTxb*OF-<+UZpo1J0p4
zV?Ft)p{|$PF~&KDvWyEr199c_A7Qa|B9zVC=T+}~2Hdo_-PW@g)FK0L2pmrifB7Z*
zKGgUG?M~K_^6QDx3K2(t$geTsHYzpCI*5?wN-iLJ3i`zng2mKg`;a-uDZVAfHPhW7
zn)6DtRX>W`UYf0+*hyo&F$kD^S}AuWCfu#_HkzL|7s1T=6L<_|(_r=eDpmHKIBqy9
zSV0Yv1;3MKU!{awL*R}^R0$#Lf09*Mt*S`0i&S(}%k#nPt+NxO(8Crmi%!BpS#$f0
z)E`*F+Rw~m>a0Ty77xLDd`)L75F&Bq-O7d|g4jueh$xW2ge}gxzHcNvn<nLZHYm%4
z?cZ)A1GZoVirRt-Gl{J2=#1hC(~{t1CfUCtfTI`fi5$gbwP`SpfyYi%+jId)Dj?j6
zR6*I!D%2yy%LAHEe~nHvKLI4NJGBAxe$dZiTy5U4H^9E;5T<|r>yPAXo@?uU=7XWp
zhf&=A?>)A3y!t<}VVJ`WeO$Ynp(`XU+RnK3Z&4Z=|JzjaQ%yTO_mV3tT`w46ZjVFq
zGP~N~KxAz#kVzl>Ff978b-N(%F|lsc<>3!l-mFx$^&#MU43ww0NK2o+_C%#&GHPYF
zSj+2ShS3@}fb*LWARjm!E!TkX_II^V`McFQ89xyVeHpUKSd?6o6{k7iI`oZdDZnoP
zym+!HajXhu7CKz+FW|3Kbg&&{O2MGh*^6@K4zKgfJoBb9qHBCfU;rK<#mNc2i;Lh*
zEQm--b!!@LYn~?`wxQ~S<;}b0?Yd~p-*peP{k(KB1SlrQKm|nEc8itm`hin|qz#x4
zIy+-^(+2ZH>bdBAf4l@q^YBy&^=CY!1C~?3o`kmreAm&KRIeT)^bWfl;!r&_)0h7g
zWGP)<&ZBgj8DC#+^!|&szuJa|V(|G*uU8z8#(z~)b`>kY!%8X%Ie2qsp^S|8FByWE
zl975?VGRRGgrVGAb>!U+24-;0XKNIrC6T^^XG90)9}u6QOYjVr`rJix__|!WYc=^4
zMv)y~P<hvCe<_eY2Oov%XhEjMo0DXfdSD@=)_)--mVr)fHw4qHFitn)IrMKEZ<~4@
z4NiRgeb9{i<l?O^i`_nB_+@tfj8wlaz(B~b@Muq|41ASEaG#{S(qtt_{mcjp0-u&w
zj8`#CF4c=9iR)$=CmC?`nE}=;tk>foC`n>t?n!qLsgz*2@9y;z2FJ4z;kYea6^k#_
zvH?3~iV;bG@pDO=|CSp7i}9>iZZRapwUJ3-nHn1i3I9DWb2=Ul#^2Q1GwtZTV;Wiq
zf#)1go=Xl%50eOH)8DYsMvdDx0t2Ql7aoEI@jtr1sYOZO;S*;b_zO#w%B&Y+@%&K-
zMXQcdMi=6+b+LUko6RLtcQ2H0v-87tEIyyi7`r>HCva%{zP~{sybKEAtc-oF5tTrr
z80@%l&UlROA*=@^PrSSa^Z(4~JGY({C~rIwFrj?uI@<vs1z?6ic=8+BG;iUz{-ozO
z*Ne@6fSEwCyeG_j4etB%n|-*tNDPA#UOvMdrE{HgBvZv1wC?YjnC=efYRn(p0aOz(
zF-rBvchS<tn`vh;gpdcffQTCyOW?R^5S|n<_zr#Sy)c!GONrer;S9>RM_sj6Z-Gns
z>q7dC0A9o$=C<Gfg!~O0kuFvkJK85opo=gOX->1tpuq39@FB~_Lz6*oHxbnqY0=Jx
zN3-=)wl@^Ag&Z=z!;3+f)(%!+Mb)4gXy48E_CyVETliH6nSSIb^<UxYEw!0^#aagJ
z5Uozi)&u%fI#cL@a$*o#$L&Hx%IvC2h(M3epw+jv2XkO~OUSO(zz?UKkeO8w8~%4~
zg&^nXE7CLU-YDU<o1lJ&UQI$9ODx@Y$X{~_`U4X=A<SQl7*NJ6y!dzUtUR4>Ocx^=
z3&VRftEN0-UM?##q3NSouUpN$hDNHs7YwyO9HPGn|NE?a!q@#trq$&eA%O-#BG?Zb
z5D`E^1Kph&P6Ho7Bq;cWtjrI-l&VxBCNy)B81PAioaSb6orjxTPvhqL1-<!)hAKj?
z<fg;<%Im6cz~Q2*>Q9yOhK`Qs)z*i#7cm{|eF}J24p?TR9Y=GBVJuxYL?5_9R9Xr_
zBIv$atb9h$i4{_U)D}$opzIIdY1(I6j~Yw}8fQsCY(>bv`N<m;RB&>!iA}jqhSsV;
z6k)dq1if1M9E9nQJvDibC>*~&N`c1lJZ5Py&df}m0%1G2e&W!lFg*}OD3k5$Yxg%e
z4Iu{Or%Q~~3Fb6Y&QP`|kd3GfmpyS;H`zdH1<FQ?ChxSGxh-f3o^!?Jv`$T1=~Um+
zmc<wJl_fUaj|t{@CLfd(EP$cT2v2;hOz|z(nFxoh9M_La@D<TA^evbv{>Qd)p)9&D
zExP*{kVb3;TKizO@s*U6t^>s%p%LM~`MA~M%8L2eZlEGeu~$-@5!cAb>NCU@iT#o>
z?|-M7_e=UpC<h<N96cF!Tc)B&)4<Mh2QB;AlhxgGEZN4+N3K@tM$$-8PW_YXUBNWP
zs43?&Yn>;kQaE$JkPULF1kR!oS=1xpK<GB1Z+=3cjT8?Y=hpm&M{YY}J>G13Sbj<6
zi#1-CKNu23LS_~$99xEQ?T90(x0|w_ofqa0w0f?@P&6zDI3}`;fy+KJKfn78Wf_zT
zZ{*QLE{JV+<1~50Zd|FRg|#)uj-i4fq9v@?vlBdtj%13<%LO56&^XMXsZu{&pB>8Z
zSLA@d7+~?;udKIj?eUI6{P*p7d7SL*a<dzeo?_qM#u-WTm3tQSH<>5K)T;S>Ki+<7
zU8RY!un;*xo2c02u{<<YKEbUX`FuVRx=BNFn%sxv8NAH<eOGlTrf#v=VTMoe2i=lY
z%2lh#hOg~Q(eY4atSiMK;M8!V*?ZS3bM?i6PfLLs1YwRF?%^?l!T)?WfUbm7GIxtQ
zb0|u8E(v}G?d`L@;}&tToW`-HZ^O?z-FPSX#j2Q+`5?Y3M)%!CwWfQ%LNjQGeFtun
zN0OtXrIxa^Wk+dQYinbBer<E%{mkBOXt3xL0>R-{M@mM9@8?G#9!DKBOI1mUX@1wd
zrTNGRwWL}V1sU1kkWN%|cytSDLTVhgBk`fv{!7?(-HvJRR3@MBKLl>)*ZuGWgOXmk
zpln2Bcw}U<kAMK-0Acx%>>tj|=T6J!^Vh&fSW2dXa^PZN`SIKtm`}a#gO5{}Q<s;Q
zo12%PBX3d|1}Vw3r6eTUGt<+PoenG<9HcBPG{wXYqwzQ#Yi&0=voca#uRh*kP!J*s
zco}>&HC1&rHWpVmmzUSqXBRiAFHi8m1mV6>H%Ifa?34acu?#vNS4c4bDQ1$+?dtGX
zFuI-Lw;;Ty-v$)o-CXn<x{b2H5(%5up@bPZ7$qlbMW7&vl@g+of-gv_;AVoP3dP!*
z=mJw?c7=Rotj+bo;<lMQX4d~1tCHad)39TpxXB^XB6<9||4(QdE?XX0o8UzyGmq&z
zt9yIzbfS|Bt<z?*^ZAqPP0%X!A@Q&ybY*$mmObvX);z{)Hgg-fG&)V3x<BRv?*^04
zI<dK*z&a{;)|42MZ^HwUHNMX9OvDc<i!+fOQxaTUykKn18`!^Kux!Ms5^$yh&)hMX
zj>Q@{KEsKA1L=ao(ToxKj?pw#d*MCBe-5T~%-=`@s>zj0i4O_Yr{g+8#4QIZYvK|U
zt=Px<UGik*x4;s*q15)!DDF)ca%D!OR0bDNzdRV=!3VM7yfo%l?N2KSY+~(iAABzO
z!N>S*#25Ug0ox&CC@cNL>9Vh`X^}3l$PtWY6@)s~KEoYT1rAF_asC6CYjsT8i@!$#
z!-Do(bY#@f#NS^34zG8G?rM8~&6;6+TVDwI)MYcFS*Jnq_pg03B)G1?t7L5ug>tLC
zEsNJMB}9!P+v2gMlZAF<cJ8&GqVry^9|TqDW(I)@<ADSoGWK_aP4Krj#8Grcz4z;b
zv6QY08($Os&pA_lv4f8sZy4^?wG?j7g{4U7LWjg<Ww?>()MHuI50=bwOtOAxZs+b?
zir_O;>u?dTj0!pW;D{`Z===IO&L(PJfj0+HO`&^q&(UMrlmw4||6T$D+`RKrrJ`Tv
z6yUlRK`M;8PhqI!{`tj0hH@j@P_uz$M({;GX4o3?WW3%NS-XeH?D3_4Yio;}oA-MH
zd)wRJta8ORC8edi(i0s{424wggTt=FrsJmfhQN{WdFp+y_<Z#~b=F)j?^inA6H<;|
zkuEJ4TUsl08@vu?$>kz1&aJRk)u*ydjcp9RugzN}kBk$h`*WHcB2*l%GSq7rUe)xP
zpPVL5o~<uh-7X*38E0Hg>Uch#b9!AXbl>a|_YjxJWpm-XaNTMfS;?#)^wXA=nRC?D
zIyzn@Cnu*QCae8ih``N=r<i&u?8(aiQUSSwNx}$eO-`A;e>1T~q|yso;vmB#ZAxwS
zB~C@qcIQENH!K%Vz8hv`4)bYNC;sF)n?$@8u`4I3z*4dRTOl~3HrY|VqBp6NsVdT*
z!wb3bm|f#n*V^i?kV(ESWFjk{F+%FU1-s@X+A1oqs2_bsCEq(icilrH2Dc~*h~Hw}
zNG)m6WZ{M7ax>9~hR+@s=1CEzCT^xFCF)H4Ss82<X~nPwmxHI4g9R+B0wUYGIkg&8
zwnLm-6o?o^!G-L?x}NlnY%9YMqPIi>RMy5S{oT&ZA++oF#NP^qGVZp1?#35%EPjj+
zhV4U%X%v4%{P8X6axw1n3&M8SY9f}qR?J<SYi`l<L0^dz0o2+!oL`Rxix#24f*Eu_
z=sMg=$lh2`NDTP+osb#|Y)&?-DbX10R1^~8C>5`Qpw=4q)dp>DGziK$!J<(HM6wL^
zfr-6OdF(TSPNRwGy8Xr9Txx{4hjYi#6^oxfYg-@Ui53Wjx}UrY3k^n;62(p;cHDbN
z!GcRuZiw*~Vo@BOknrbA^3G$CxEZm82+fX^gjyta-urNrvd+VuV5R3y(3(d1GF5Lk
zB{||^;-98x&@2U?T9e&Tqwh`a^u5q*TwJZrbxpx#=C=z*?Ym=1z?yD{76X&lq>-c$
zgE%@(Sy|972@~fE=18G<0M8h_GEBB(<s_IxH@Fk|#T}!Xx!D@RI<Y+JLwuk3xJlZ#
zbzfAqVz@VW7GKRWP+8O=VP6)?C^{-EnBS$vgvg|IoE{9(Bc((mh)UWNd(o`k`#rGo
z0^hkIe|EOEw>N*i-BUwjrOe!1jX_UNMuNwc{dRt&CbzpAkhNaiY<0|;hBl=vK|eiI
zF_fz@Xm$2)_PFfyx+go1%bI+4Ed=@Vak^%AOps894$!7BI6B&|)YxRD91T`xbhy4g
zKLbvlLsPz6m%K1f7brX~Z$=VkZp$ZOET)F}jM`C=k+IRzlCA04=^peinsP2Kn_1i|
z8X7MxEv|RBx0m~vd|uD{KwhMjr=g{>_)}g^Lu+Yi@rSk?S0sNic@a1?F^`XvG01OA
z%gbb6q&-^O{{4G}euXIv4BTb-D?=md`|j=i4*Ya<Y@mfVmzTGc!Esojveh@%vvaVq
zg^4sQu>DL*TKM@BcuhLMz%V{AE-ES_Czp?mj3mIs#K6Q{Z_CNa@TBEFI()#xF*jfO
zt^AuoFD5)DCOm=<%bp!rCl{BSy9ewES@qSA&H1dlSmW@txp&g*8Z>?wEgJiFagIe_
zlWl46;_Hi0F-~uFsKvdyQVlR-6BZm(4+L5rP0uf{(BHTTzM%u!A`*5A2I`a3hw+RS
z_m^PN*wIniQ&HKAEJG2{!VGy?RoR}w?T|b*W)#T+s4WlE3?a-#3+e^R-m#$9spm8<
zek^R!-)(qMSWTau&7Zi(t$xm-r9hRp!~0_TON7g>MAfb7X)_?{%uHrSM&sB&>_oz&
zfkn<1ULkM@wqVDHf5ig&R<mo(Cp%ds{UQe8XVRb-GTL8htWmo{#NZS>%H@KP6y+yD
z4TSpW`erp1Y+sVNpyLAH1lF^;76}ycR87sU&ZVTJ<Rm4N1fMA2WiIgI*x8$ISs~|%
z2W9w~yzwIgART_oKkvFR2xcpl3Od+gj7!NTBc!<6Y#FeR?GN{m>V2UF?ab{>(G<Gl
zp}|TB_RXD!NI(KLz49McOYIGk-l^I_-*|qhvf^v}wCho1I2}m+duES(R5M@AK2;hv
z3wWdI{Z+8IJu>4&Ug`3sxke5n{Pb^FSf=b_M&(bKDlvw2Eh?50wLkjjILpcl;!LWl
zh7I)9*mFs<BH&2MACN}<ddIEr1+9$WG4x*%`M<`eC(GSF{MJ$A!oQ1&!*&x$9=t$X
z>TVNufu$n$mlO)GE%?ho!Q=($<!ZKb&m63YKzzSn7g2#&-}JLerR(EhX$EINE0vj9
zS$Z3BymTF$^7KmmVgZGJaWOI?f?)3<Y<4Dr{P&w_*gCl1N~t|F;yNl^g$LVe{ftF`
z@N?N2AtXHYt(#w}659gGiEDF!kh_ZB2hnLe_Y2U#Xly8;YWFu)F3F-Avdr07K0v=u
z*K9H&dR=jK2vwdW-?OtdC9A3M(b2Vfb90cAve2{f@_iWYw+z=Pmjng|#1in;4*51s
zC0l2o@oMX+GP-KEE+hAAX{Pyh{5CPSFyXb`G}PigI-p%&s(9F}H~+U->G}SSa>0*k
zcdss?KizU2)|{GJIF`aPH&<<HvN<tE>vf&`e3L*KOTZB(l7){C@&3+SCR4ufUu8b<
zwDozv7t5&E?(y&OVknkSTwHvs(>pR}dn$)NVY;lz_Oria^MjK*HYzGNHCR4}4|o$5
z9qs-04Cn37G;-ADy7xl=_j*asH*eEU&FJ%Z{vQhAHOA23pqQB0@J%~$$m>zgUPs<5
zA-~-3$d8*tq!l>oz3*|s$3n-U)1$!OZFc+`!D8m(dfn*@O{v!H@Ogh7h{OW^{gQbD
zd`o+EQ&U|*LD}CMb9?)nJ<6xsE0e8ucY4kPB`>f2l{!jcA>fv^uPxbTJYk}^J}s1I
zW_`nC%;fOZOa+fzDZV>W*GNyF_}-*`)@l-&(v(xZwIduVACw)&gOz4SP0Y*7!tr#O
zKq+r|hF24T?6@J@hZB?N&RKXbx^0@xi@7^Q;z~nj*{h-!qEtzdu!vn5%iIBRNk~7v
ztItO!R&h8l?FGfcDMQ|!4~>PDVztt4T$zCvfKWzFzdF9b=;qe7?JreC%&REmgHYom
zw5jtk3Zdi|x^x=y`U*thjSM?#e+-%at#I7^7-nWQDGV@_gebFab2|N?ap^iA>f9zw
zqSR)Jh~(?c6?!!|715!S#pctRWEAik-hU;6RSF{qe_Ak;tID<i{89IEnw!-GaVM(A
z>QQA+Q!d)?3W=yeXAAV0BiL(IPAV{303AdWqRaL|YWf87k`X*^R`x?#;PC-*I=hf%
zl}eyUY<+Jh$Xs#e7p-F?kXDA#N9!#Vrc7ZPh@%sEW4sN&C|EXB9D1_)Yflm5LKLoM
zJE=n4O6bRYv7k}})6t1kQ6`9vf9nAH`L#x?Xg*&M=sh)pN&{!iOe+{AK6jxml}tYs
zF((kWL6qlO6Q5fW?_FqBQr=lucrFhDrmoI5{xJM7`4K4NfGVlB4ve33iLeMNLwMM(
z*tElVntbx*Mks=VHWqg@f5;+A`48KU>12gg9i~Y_QouTwhusk9HRvxL(G;hwznXce
z@B1USV;_?;GYf<s2Lubv)6&G6Z_7f7e2!BHH6cy&>g-c|Tml5eEr=7aGFFTG8`a6V
zH(Q>BIQ=8SX9h0>OjT;6775D77FiEco`_W*m2I+dz(OpEc96P_a?U1j+YWsaUP)QP
z<KL+W(g1mpyZei|ISyP38p)JsHX^V6`Snc4ho{Ts@?ZPF(6P}Z<i^K4x+9sTzi=b^
zSU))FbUQ{@`Zm1JOs?KNAEBWjGZA9LBBEsEq*~+Md9Nn3{ykkLB`3}=%r`dHy8xSA
zyuCd-KH6-%{=Jms=yrMpvufo&clgFu{*l<~o5CqG`29*;iCb3f^L!0@3gH?8WX~wS
zULfX2ZirR0fs^*}YN=iDu{Y5?_+cW{jr+izy6)iQgnrVex83#k?_c0>7nFDT@{9IB
zZ^%)dKIa#Tf;orDdlTI|I-v1+6I>B4_4nfShWR1H3rwP1pt}>_&Qsb*B{eD2gB#KE
z{4&}JMIrS0yl6U_%06*w_5bj1TwFZ;^C!Qzca5`2Nx5oZXfT@(1Jmw)*+!55N;YG$
zxx&Rky}jF_D$aD)u<@k#1tAlYyOpcW&E4j9?Z6YbeLztB^~*3pxytxx-A7$rMe*m{
z>ZX5_h<5mj-w5I>JqS~WCG9z#P&A-zptnBrj;cOsp#cFd8mv^L5@*T(hCI<C4jh#b
z{!5}oL-+f4c#EIrrPf+t@bC9!H%381Rm?Wosl^E_x7vj(bO{fZOJ$(Xe1%k68F3N4
zmlqf~?{km7YnH_hMqe#r$zhm6X8oWA+2VF+;t9>XQABGnqK5q-wQm8x{6|t%Lmw1=
zjUy`RC^xwmq)fa)(Xff869oNe(x_hMAX>l+aRMWnFn8UypbE}=-q+anngjm|tuag^
z{D)=I!U4lBdv3a}nNXkDw3P6f4o#u=OK~6JLGA906;L)i+}-BovyI$`;KyoaWm~H3
zgMnRVxIs@-muU|~5K`btffCb^@x(B4%>z3Tgh~{U`Lop^UUJO4G$GVie%OLIX*iv{
z{cg_u>4j(KmZqmgRP~eC=G{KERiW46<3sCW30oy=Kv0YGcNrRrO#%js3gXCw>>0+p
z7Y#Ui2yh1&KPiRfhcUKeri+6s;9Af^hq0o0B8}+9xxU{QDt4>&8?O83px9NXsXc8p
zj9Dd1Iw-*x6*2y%K^lk}7pFz6%sKjeevIY+A^WyZ=_8h%2KvBBRcv`VI=XUxoXFq+
z3OwJ>r!MnU+Bi28!!u4W5@?WgoPk+t*(>v}c495kf0IeGIQt5HA+Q!ThxwH#D2nV$
zo!@Hrrc-#@w;dXUuvzLwLo{+(aRtudcwVe4FMvx94i@FGS(lM${`SR(ef#-l!{=jD
zSJzT!Bbl4NLC5Tuu??|^j$r1XG!}Q2UK{esc#psTQ7@@se!-lwvZL*myNZg6t||pu
zc)yNLAVi>M#O&Gqe>NbVrPhY@PF7Y?^A13=lyqe%WB6@B??O=^eu1>pPTtW#N=j-(
zfS~03mN3D_JF`~km~np57m89i^gNN#1dOS@Jm2gAbLl>SDJ1{L^B%B?)&1dgf#dU8
zOj;4TLr~h>Su-m+zfkF9yIVld)>g|>OU=~OR9xIqSy|lE6L@vkaJQC~1%5~7O1=ug
zp^ZvOmzVzmit<tEF)3-O2ts}z74`f3k%ZqR@?+9tV9f39?ZvIQKdUOs%BUEtt1B(7
z%*@O!#Q3}*8EtJX&(4ogUBkaH=#X5h4j<q)XjU)r`#zonamMrI_F#(t{T?6;-(JrE
zBkh+ZphPgYu`X_m-Lkud-Vq{u<J@U2Ns5h$nSC9?7}knHi0}RrU7);gJy6-{^>O>}
zs_S!)@N?R{0SI37vJ;}g!9YI&)b05sq!6+ATYE=W9p(w;b$38bvv0TUetwji#=X2+
zY45z&Zl2Ip{rR_GNIo<xCNf%HR(fipLrYs}b$VWi^h~KtFh+A9`Uu3eyYxbvD3DEV
zT&pE<4*6n(NF<wUu*$S&{z5<Fynt)1s`>~tXO~z!kZD#A!bIoFRaR%m@Z8Q^98U~$
z$gl<xnU?r3juINhkUYqWM+M*a7$nxfhBu#=+cBh~1@v&->2o(^OJS(W*iK$4EN6{s
z9_8j@zznmIgkSScjBUi1l19+oGV9~Pp4j#w^Q>bO*;1o3H(F%RVD@>_!;;mgK=1;l
z*sV_AiP<Ss!$?FPDU|&nlba!br{4Z8*hv>4g=%SDW)nbv&Ygq!qgX4s<#%sjFjJOb
z(Chz-3lF*Vn`JO5EYgM8t5ikk0{@V0T465CSB!TU9X`uv*nQwn7Preh2kWj)Fzi?N
z7C|2orgJ#jsF}u^PJ%UreVXH}!^fdrWQ_(h?1u^Wl0=$qapTK2g&Z?uq9b1gnyK-2
zzM=8A=f7ZHIt(DCPMjcx@_d5aU+%x3{bs>qSQjTjX*hc)B$gyLEirO8Nn?^59q;?p
z=~1PtuQ0}<qOp`I%xv56y=>H#$l{UPLUcsJ-DF7+?Hl`RAa|FOD{u?s^Ezb=+agUJ
zO-+iElQ};gQ@T?wqwI`K@_CeJ45nys(S1!0MBcp>{_U%WC?*yN9u}qohdY_ct+RRH
zv8%OMaR2}@SBJanTUNM%Afxs$#!xnN1$bsFc^g<e6jmDPVah!53^JUx%ZSGO9UnJ|
z)P15lhB3n-nPVs!IB$q(S^5EasmF^I&zs!=gc&TM0O7~is}<n!I-|kz1W5(AjGT)i
zRHSgzKLCxQ<-bRWe~+PAJUE`j6r9x5{2Wd7g@r9;Wer851**RPCYq}Cx;`$RB)3s#
zd2XU91!oE=ALmNs^*X^S*j__BdVc--Rd^WU<e*||WV8?AA4!4XjK=qVFot!0Pk@8-
z@^ZBFUl57B>LMh-J-s-=JHEnl4PVpLI<v3@8fxJD5@PYvG`sTuGg+KC<4~%i(2Iyg
z;rD*MFH<HK6dXb(T<J2c{$$v~U-N3`uon6|S~WR2Nvm3Ar+HexZ2z}13rIcx6&VP3
z;{}uNyz(03><!FK>P)WLl@GlGjV^ML_*p5{>TRnFeH-}T&rL5E>kWW*K^O)Z5(YX1
z_}=c1R@yS-N&WrxmP@rZ>%(o1luwUYB?ALMV?H4(JA5u=&N`LD`(JF}@A~}M(DRKq
z3kVN2MBEbB4uHJ<o|{FI^L>^2)}L6Gb9g$Cw#)tB;}rj=*UfIQSS6;b^G{V39y+S#
z?0EO@3=CvUW?PpVcuThGt;;jB<55Q@X&K{<X<hmUwFf{0^!R**{DS}+XvK+%Z(biV
zfV#55`9TRs!Xa|DBaT}Fa+0i%`R28ow8n$d?KCip_I|9V+nM)%aBj|iCcTehPBqYZ
zJsiT}YCi*DhHQ=5n0^HRl`PvQ`;&_l&Qn;gtd23FJiHR)wq*7dA=#!U0_IY=z(5`f
zCc*ziUsDGYhI>Cc;}H^kR#qx9tjtfSoTcW{<7Qa`^jr#6Sb>)~($Gq6=F%WWnF$}+
zPI8UWCtq05(E*eaW8rGUb#KgJ>xa!)4p0&W1I9RlC!){XL#Bd4;2j`YX-P>=&89C|
z7*)Si1Q3sKEtC-0R)bUpixqG}Lb{N7wwBjhJ(3~^tmYaQT{N47kiUS1Lb4;62cMQw
zXBix?I9ZEHb(e!Z-=Wj6ogqhXf(5M?zlc&HmrIF>6(<D#lCXn~o#mb^<^vHxei-#!
zb!rwAfVonPZN{=Y<R;1;bclZevBGlcv&~_jR4UVH=(8>LELPQ51dusURu9X>%Vvv)
zU{r20m3hiH{3hzOuZ2sVp}<U~!?{L6GLo7aYg5xv4m}fQpR<Q{R{24dbmjydEjEVe
z>iaR#RM-aY>uws9hJ(rs)Z9&R8n0UO?w=p;*&A)b!b0&$e|;68cNIfI{9jxQ>7lj0
zJv?KEYAdK%1$W<7un@h<R!-*T4+<7)ul#n$Zs`$A=<|bb(oRs_iD8iZ>3mTjE^o8d
znef|zo{NU-*e!R9VGHSOk(=|!t)uG1cNV?VW~04VB%mc01`h;atwLbLmV>dnqQ1VN
zrAUk9<#O!`$hon;pZ`#t`r;^}8jfQL`G0@!D<u3qpCI}E!txes%U|X%Z;$(RbAD!K
z#`S8m<KG!$Ar0-mLy;2o7inBPTmpWN`<mQsm&wZU`+M5(aIE}%Qn?&BICw6H>)flD
zEG-UC&(>;JXBXYm=Ke*GJUcAj*vlaK|DJ5-rkZT-DU;5I$_F9O3{jvVkVB54%}j2s
z&}=YiSbJ@_Bsg0F_<;Mz)%FR`GZ#X_EuFT<%S}(a&mLiz>8-qi&{)WC`~W`X#eCnN
zHn(#`O4>IS)Yla@>=-|zh-Jvo&-97*6-`j6I<@(^Ig%(tO~}x2Hwi%P(Y{ubk&)@P
zy1^M?MMeXVT9?xenhgQI^Um8#De&npJEo+)q!bL03|L<X2nbvbMUKO_bL0H>KtnO#
zxSg)2{X70W0fF=4>Z)nXb?FhGPpip#<^1BHWyCRXsFlrfJyS=OUaw`{yeuKGprdkw
z#PSSE)v<D%{zGqO9{BRQKnyC#>bQ$Y!dnDY^xNol0g*Nf5iU9+sy#C<H7z^d^=NW0
zUmS=tmGqTpf7O9_3)rjif8NjK_<n@5xn$dNBH7BH-tEsx3lR?v4oX-eQv;R9$C=*e
z>qpL~3iVF_-H<2Xot&DGm2CB3;Mu+QhtTR^=VsLFi8XJ^=0-W)$9r>mg(g4=i;0kv
zY;k$I?AYD?z4{6rg-@<VTpZQxm3F{G4v>gHJwYeUV6sjAv*1?9D}h3sNKBj-u(uKN
z&;BD-HS4RJ?(eS?LDFzkUoj+}z_8mx3kn#5vbc1VlwvGzVdb-VlI|bx^u1E8^G*@K
zP0jp8$J58rj``3E*=a!v@JxCno)H#8)+Uc5%DtlF(A%9NN8<!wY2<iE-|8qdz<Jj~
z@^(!13_`#>ZG=p#u-FC$`vKri2*VYY`WnC=6WyZPsk#k%x)n5u&xlihzN4CpPai|3
zVVHIJ%XcC7Lj^%X9UX2KWwCjLT_Sksbq4H3lg&Z5B9YZHdnl?177?IyFi=+tpt5Ho
zw00<uZ^5;9CG<Fj9cC3T-rnafp2pomJ198zoojev$Bhb3s`Mp#4rwOPiO3vE31CX!
zWKep%LUXl}>E9JE?4)D!73}CM!@w;>iDnmUn~E^muzG<Z>B^cf;OaO#U$_wFOyF_M
zjC(!91n}st(;7)zdg=>o*OJS@C#8=a#kS@P6f<C^G^Ec4I1|ju---~%v0AxV4T~ey
zUe`p|D6e1C3@^^kpASyZtwX|T)(}FsS<Eu@37zs${X2?G$>E>cN9)tt!x23G56u%Y
zJ?+xEdSL8AmN;PYXL`0Oc#9i%<5xY@#YKZFMq&oj$(ns&LA;}38#pDwPV~Qku?`jr
zO)q0e&+hd1P*gUUDFArhs-i-XY+l#5<L^22Ynf_Q>9)qfNdm?Cwspha6RR7YJ^<|W
zl|i@lcj_}xW*-()q^U%#&#<C0mYL_a1%h>V=O9A({g<?Hiq*ZIF7anABp7#L(z6j2
zcS{LtSI?HZ+5wpBrOuSOtiG(Q42VwGJ%O3$=awVXGm46)y#m)41pJ>9*?3D;E;>LE
z@z`vyR^2~LO#^_ft<6Q7w1m0;n60I%wcQ5ye~Z>W0>0dK=f^hye+G2Sq`gyOS+OzE
z;nK~CafwcB|4ige1|KrdwcKA^2>~4aKW<BPX?_7<O|&%ExVg!9FLhgl2L6;z+mtHx
z`8=5~pcB1ta*IS)gd;8(!KDD)Pz#f7Jq2*jyw1&?r}MPSNZz5RWkE#2U<rsXtp;dn
zvY&nxb<(r^u+5Q{XCfFs+2p0iA_#a}^sb3G82P!+RDlafOq}oUrv<3Z_#=|3%$OhU
zyYqkh#GkG<j+(}nOvf5!3GkL$Dm4Ip4=r>cvZ5>G^qcv)^f|tERu=D)?e)gryWu=x
z14{Wp5j_cjrbx)rdEORuH`kDM925BT)2;S}!r8cv5D{+s51h;)+4?fO$;!??Xv|Nt
zad`FdI+nXcn=}d#6+kHi&>C<zSi*q;uGU*Shr5Qnt^{}VV&q`vh8=bfG^VzU)$g89
zSv>a44D5U?dme&<2F}j(QuN3+s4&mL4<RI{_zo=nKaqkT6`AznH@kaT`sYv~S7j)-
zuutKk+8&W4Qt0i$Pll?BaDPG@{@SCHkp(i}GfW>(Wb_}0+-q#X?7c%&i~df(JuVqR
z8a3RI5W(?EuuztjZxoYE4xHV1FlgcOf|K$oLWFuMMa9!-p~p8=x&nMQP<A-oExh3#
z<njg@MR{n9<eF((Kt6~;5cF%!M#&afBr}c{D}jo~4Hzdh5+<WvK-XL;bO2*dTa+#Y
z)D)Vb0xW(wJFm?NYAF!QtZTf%=ZkMXD(I2Yk0Hk0JSB5L>QS#-MVisxQC8m$aD>Rn
z`hE;K+%j<Ol5d-UIu13zFvRZ?M>f?BwbEBG0B)BYa1AqQwb@KTT{=Ji{tqRYi$PLO
z?l5Pe6GYt(>6zQ9${#EP3PA;z=Xa-PXsM{^T{n8<;^6rG5vZXtatIUBa@<oC?+7ba
zStp{A&_TR~LgTLiUfdpw8aLGu6-zqhpN4UuM(VFN9l&&u2b9ar9&9Vd3x7U7Kn~Ri
z8m~x!4#6=K5s>+GB_2tI4g)CsA0lgsyNp-=brKc-brN-QV_QY2@tmJ~8McC>QXPWD
z-<~#^;C^W&+<;}H;4gnQt0!vJg7_^25z10{lok*w5{Zws_obWw&`dB?SNPl?OU@>h
z8<faht#vFpzK4)MM$*f!d@l!p<w|>vi?&~s)1`ZiheoC;Jl?CDi^YoNzHd820G1a2
zeZ<e#@=vHc02SbzL`b$zBL%wI>U(&7?m%e5|JC)kZ?WC{csZQNz}oVyRxHoJ%{a3>
zBf!8oFfu0BZg*lJA-P_*|J3RHuR3mVN&_l}s;%BX=IZI`8XD@8Ilmh0F583eFqg;m
zYKB)m;va<8f)T8-R+s5DjZEu?eRfFNf+NqZtj?`5SLkb<)s32#7j6D#brf^}*|`rm
zV$ZfGhC{~EnW{hNby}=~$Nv1Rnq&a*k;O`lwjS?}BM%hDDOmtA0g&8>t%IYhvo7DO
z&c0COzKWS3IRd<txVUETjM(xzX}QOX_4PIsLQD*P`pVKjWms$tjjr1Y<>77o##hol
zUnDguXW&<SQAYR0nh;90Fo+@z-El`ENz!I<dr<Z$h&tMt$U%m@_fi-DOX@6uS5go5
zA;nc-*CaK4fsR7MCkoH&4Uv`gyQ=E4^)+O-qzI7ZY_>X60_@4j=mMaTsrzyyNA%dq
z>*@MzwbedjnYEUVF_x?~&YopyU~WLes?l1MmR6AU&eg=m?d=tV&*wlvcW$Vy?O}0R
zd2Z$R&&)h&%t*8HvIzi;-2*@|g<*R$gCiXA#Zw7JZosVe`}IyAFo`u!<Ls%W1u!WZ
z95(&BVQ@SyUI_OIcYdmv8jMcPE-RhRYXG+Zlw*`+x}9lBIx6?-^V4A)-1~|7u(+CV
zpTv#Nh!O`>3oP=jkn0!|0^wuskj%NQMGfyBhqCC<QOqJ{RNls-3y+m+yM}usasaof
zV`S}pwuoOlT{<4)92!Q6?^?`okHNIcSqUyH9V&7uT!<5?N=WG7Ee;sgfuU30Kgu3J
z7X94&Ka$Qetg5c-!Zb*CgCNq~-Q6iI-Q6JFC5=dT$D!-c-QC^YDIv(WpZEL2FRp{&
zUVF|t#y#fZHH%e*sNpigIM!RDF;YZteAWkNHBOAuzA5my-1l`5t7Ajw?9M8Y!(xI2
z2yNt?I6+iD0Ix#SPaNx<S@NIjKO8jvmQf*Dg0OnmAJ`bJ8CBE2jM2z^ur9HS=8yHb
z1~)tPIyxp5EwnT>=WkxbRyOnS@SHl5+i(d(RE1fTy;)lAYPU_Aq-+<g6;6>4hQ-Pu
zSG%p*``z)eNKCD5n%Acd$;_Z3wOQs=iu0||o$thEG=|He(IAG|Am)U<RII)xuE5Oc
z6_?!D1s)jNk<UjSE%p2Pne5?ITR90bpF-EKIi>5HyM!5^5R{803+oh`KNV992o)$$
z_(+{VtR3{DA%4HtvMmGd1X$meG)^NJ-u4Ihv#ah?#@D+;dAYg8kiSp!7h_cmdx$ua
z_k2)l#0ma0BJ{_rV3%E@Vl_!?9Z8QR5WrvM_2iKVt$)`714pf=!h@-~>HcoR>*6Lc
z?F5ONja@!R=+eq@`P7Zsu(J<K@k@*I?#c?jRN#JJH@<3p5gQ8^*PoYrytz_tHnzLH
z!*R?0en1R6EItIpuw{0(X-SDtIN0kykFK724Q_RXs{XV~+m8Y2Rt{|Ieb}K`(@j|o
zfJ6?A^<4L$-478)%XhuTo;8cE$G#u;c-oWMw!Fh*8-0BX50%-+0wM6DWmQ!dxN8iC
zZ4-;Dg8<D3JWjwIE#&<?5;cQmNij7zIAqwF5rZj>ha>1<W#?e$AZZ|Xmn@Z>Pm~4A
zC{+m;mCf7z-Z-X@yxwRN$&m4RB0N(Dhi>V;*DFkvz$i7W&Pf;IkOUXsFnZ!`;{9_M
zlVTth3uTG(8MY@kN7FAYO+dkz8t?YmADTQk_ytTMzz>Cm4<i<GvWnlnnk4G{&)2N;
zOyiC~;|+FvhpGh2lcXBGzv+7o6#vA#-e~)-s}?+Au<iFmg4#r;O3Ja#WBsR~bL4BH
zP_xtf-|Y{7E&+xsOs(Hq8(??--M!)-Q%HbZsnTuEI#+CD;rwoA+2Rjs2PVUvo}Dx#
z)@8>H8d^pW$T9D>ZMGy)0OoX<myM31kp<S}Kw%aI-6HcsB2vLR7G8sz%HTiQyM9gw
z1gT<wf$8XAjpg!5BSPV6WFGP5SDqoERnp8E-AIk(d2)miakw1ejAMQiRL+-h>BRpG
zfP9Tg4(#aP!F-Il;ZlDfHcu}k$Xf}Gtd>vFKVve5GMVJlp*w_4{9*~o&Y2?|*zANH
zh9(uDk@x|2;$(uHLVNcRQDdc;I=eqE1c(mD`>nA>cAAK&v9#Gz1g#0gZSy~HBw#+w
z#7Y%K*^Eg$G3FoxTQf(JU-V%$>rL)oB(8&HzKFEEL-%Q^?i~E0HA3{hF$f7=C#Hw+
z>#;lSg~;ZChL*9ViSYH|7(|xLaFX=vr;^e{2BFqgE&t4@usuUZ&RKx7Hq9du+uE>f
z6k8GpLqS~J?x+TPBlcXMK6CcqGxuoc&7Pr%`=)jUeDo5t-vc&b2L3r|J$(MB(uao;
zGM6{9N@8dwG)F#d))}3*!jbTsLWSv{;S=>P=+%xZUG{ovYHZdwuv~a>*s#5b0E;$o
z9N`cfJ~ye7vG4acriUF+4Y;Qd-nf-f9?;S<Cp;P(YVZ`P5{wk1L-Y{?%d+O6w770y
z=xBCkN{Fs|mw(j-_qfZ&b)m;jf&fZGt)@6E_tW-%w@-HFdI9W`*K%AK;3ro8u32gF
zlU|8R9+j2#yuat?=gSce`1kMK$3^;|ao-~H!f7H#(tT8w)D+t)OiFTBE81rv-@lhF
zCUvdVC4Y=XMJ>F&6*L}pL*V~w8oMkmP4wDp3L)9KS<`K;t!?f1Pinme?+$G>+kZIR
zZ0sEZu+Ff9@CWo?G-7@i1|FKPbZoC1t0QG)NUU`L(D}W~g_4(&l92*9C7HauEiSJ2
zS2%*+0J!<`ENkujX=j(IvaGtQ0#s36Q;oCRjq2<Pyqks0aQNDJ&NUVnv@{ogf78%f
z=FnJ<CKaY-WbpRw-Z6Ig*O|2r=B{^PV{<f+S5$7xNJ{Xc6#A#8Hn+Ix-yQ-L0yLuW
zBq%(sA3MN5E?bDd+3SZ-tE2Y$*-?kdqS`Pe0M>zyr?9RJ0fN{BVC!t{8vu?DsGM7w
zTz_9)NqBt^%ub?!2Nal*EC5%I*-J&!V0K|n&cEL(SA>Zk4MYgzH1||K!B-ZVlAPOA
zuSSU=8NZe!PLco>NhuKrfyj!W0x!_^OeZjU<3z$>^~f+q;QO1ZHn@&*v$J5xB1=|?
zk#T=+P#zE@prN4vvH@WC)?eMPfd4_a+NMlPYYE|8OJ2)WOzL>|E1$PF<?v90==kZ;
zF>eixg|&?W4H^UAkN4>s!+ktl{%O@uFKVWjB-vk}XY1B|-P|4?&zA`~ZxVzVnWoyD
z_cvBurO`hBx!Y)6S&CciF#}LSV7}VSXzFVBfQ5zo9vLUAJPiO8wMra=+^7op>iuBr
z`A9T#Ih<nXPq4J0IP2_zMQzvv9TUMphN`LxS{j<r6(zY!*oSL6PIsg2AV9D|`1Gb}
z+gMeLbQCs%2_jVqyC+DYdebnY-_2hT30fee+2o1k#V@M0re=<~P@nkDW)fpc2VthS
z`&F$B$E`HvVwDW#Q48f8omK*YA3GcSq(wSIZHDav;-|hZG!)=@9I_RkXj^QQ_5c-4
zef@g7sCl3hAIY7jls^otMR^ebe>nrM?YWeT!Rly4jHICC;(Q78wGKB9B7@S6h$Qrj
z|G1+3gPNYsZDsj_24CdTKf}&coLb5#hbDD8#4dBc@Ku>Sp2qtw@#o4(45T~^kvGB7
ztA#OD2<<?`nA|wpXTFIHane}$r}Eh|m<%E_wgfXy9-5W0je-&f#6tWRqTRmYULdPQ
z`w)|_rCnYyf)}dU(5&S4{;HU!WH+~<bo#&%$I;lBgb%4JtOoG`_Xny;b9`WvW%sa^
zm-vha9lH8uKQ2f*{wLrL>1AYm(^%XYlL_3#R(BW+)F}B(Xi;yCAD)}&VfmJyg<)X<
zF%1BIv88*CUjbG{;Nxih;Fs+rV{%mvhmG^E(Iu7#g;IC|TUj_xHNl>7c&8H#w_gQA
zN9|8n!0Vsa{cd53gdY&HMknO#S=ZeT=ghnKJ*2XPTm>oaoo5RHTl;gsGX|`@>DX0b
zB8H2+V%rEywGYr|?++`lm~yePiPituFaLNrO*W|cd~#g_DhH<qa8sVz44ksJH(%d4
zdA79Zgu45eh&arUe@i<ATyNq_5`E3;yN#VX&!-#;GDT-?GqZ=m=g@;+3+LSR5HoEz
z&ayh@iXa(=Q}acCXJb!g;omx8zyC}P9ceD9x2kN+YpKI)nCA#Pd)CbI+1r1Ys`>Vv
zo{jJ5Ojk9#tXzI0c3554;M+qh&xeDk$JAxx!vuY5&CtM5t{~BmJBgL@<kCOa>)o!y
zN;R>WWa!@Mow1j1yqX6S?AJe68L$A1hU*@$%1O7)Kl?8r=Kx*-KsR8xlK(l1t>Op&
z7*v)gBFymWNDWb=kkEYobX-rfi&!M0>g%LlUvlGM3}-F)j1TSBrj&0osvGZ9T@LE@
zf2TUHj6R=%As&#YKjt&L)Zf^ox3OZNrR@eJ1L|YVvBiUvd)1&}s5g%nhppC6U&{>5
z#t#Feo;-sV^#S^++1p;o<z!~{=H^%BD8MHI?q>#JraRvzU^ToQvdc`x8QSQ|AZUVB
zVO>a`UB93S=ralV-^T|aKRjt^ecJ>Yhu60sky=kKF4_~XBLE}B%jmIjvisv;W*3l9
zurM(JtR9$W4&x;y)FYB2g0fIq-t-X8A<dlG_w*^R0TF8$uoYk7;p!Y#=lN@$wtNAb
zN-~~w-XqJb-+DE3B>XEBf!7BFF!PiAwu9QpKO6N;Zd4bDfDT~`*(JbR=T+_(CFt?L
z*AVgXvxsI$7wbKnI=r^|;znS%1o;_}2s^zLK2}p&eSuQYPe+Frph{+t_n8@_s`9qr
zir98bM2fOZb34R9qP@V6_`00-7;fBYj4$k>zo<Rln+)*-#DNWgfd7qRR0)F6HyCUq
zu#vIilfaqIHn%-<h)Tj^X6~S)auX>1x8N~vDTci<pE`oR7ZHTRCy~MUaDNUfC7Jks
z5)o1Ao~jzj#zt3n^h(8lO_V^&?=k_v)>LR{y1Z-$yOwCVu*jBBkz;@1ib&5lvWbbW
z@$k7%AY#4GkScebVu<_%59OM+Sc*y-d0QMBf+fg4h)G6b14sgg`$br)sj(8vc3+F~
zf;Ju-p9<Pam}qHr$}NuD)NLV*NCmTa+{E$=Zd~JDuRs6tyT25cby2_L=b@OM_#eUR
z1Gm*3pxmI6MBpGI_buC?5;|yA;R#mZL2SGC;Q=L*FvD?s>SuJDlhcRj6X*GiX1qqi
zk<OIP@tKF!N6-oqYm2LkZbn8f&<32&Q|Bpl`$QZ^P7o@ZbO`(Xi_uD-omFLvXFV<X
zcdwkFbUL2OEc|vOuI%Xv_?}mn{CN9&P;sR90D>3Q34Lxs7(c%O!=r^arMGjib8^<P
zX1iY^_|Io5E3Vaxm7GzK{n+Cwq))ooJViUGsv2i;fm5dK?+9jw{_@PmEM-1UQUs?v
z%BG;NAW+Zw_}kBkGDbBO!0pR2)Qe4DqF4vJ+B6qtDOtTi8EtT^*ELf8wpKRQ=h*BK
za`fdO$207w8N^%TZ;J29rRM*oRt`2s$H$drI$VU<_&6R8@<XIOE?2;#riFm%DeVLA
z|8ZufS%E_8-|h<?_gR?8`O0N4g}=pBUVi+?p3ki(wu%DMx`7CZMbZoIQx6@oOwe)E
zBX)xdmDXHY4B-4C7H1UnAXP4+1U*Qy%+wVXZCC_760^IOC}yuR(xq$0up(j|o<-}#
zu_275&><!Xsql3cwYSt)9QB5XB|HgX9CQk!F1k#kkkY4uYITez?<+C@dqmHwVQJAG
zuvmdMO*WO;($XAo3nSYH$c7FM>D9=oPw}mopx81Atw;VooX&f0`*HKIXsdtE28@HS
zvN8btb=cT3Zkd6r!6{-1n9@&8q4#UH<M(**0{BXx2mtygeT7Y+B_`V*8#PQtFqtVX
z{kJ5kQd5)2wCtUvt$(6gsfz%3rj(7`ALa+GChO_=Z_@|M%*6;^*7^~kyMU1wcr#$A
zXg;gQ`W#Km#dZu=S-?_BB3va)izJ}wLrT=1i+(}Y@1xv#GuFqkIz+@7$pi(n)_E<<
zSM~=)F3RYmNMM~na`Q78<Am<tCsO_Quz-7^ckvc}{oXF>^B(-Y^hY#(g9SNQN1Gm>
zHWUL~hN|)$>l~waYLAYOV3mO^#)Bjp3~}WN*N3X(Dl%nb(AccTUVXfO%hA#pa*zf@
z%sjI8LB*gD9Q8nv!B%oyGqxR!;#IjD9R`-DUR5@{HG!SiA_%OJ`2E_v0{QKb#FX@U
z<7eJ@OZ?^Qt4n1iC9r2r;GAMf8eDfqDu!bD?A$N24RVCOg1`uHv(f_XmDH8@l8-c3
z5MdYL(aGG+Jt`fiTP*fdNQB?HM>S9{mcKDoJz>KV&d>~~s&IEP3si7IyeJ<sN}poR
zu$Ynge17MZE?v+iamh)p6V?TZ@<WpL3d0bLSmpc?Eq^<#{A`i7Si8qUvGX4!bgS`}
z8rI^t{qLq9?c?>5do8I{rF>DBq!>2J_(mxF_Yb7bkH{3=xp^%$wVZTk!8e7_+?ref
zwrg}#KcZ^L)~XU{{1DkI4)uAudUEsp06wO$u&~JCZ5d{eJCF76L;A#b35Uc{s?VRG
zZl-fZlL5z1kcia5)_$_h`CiV%#7d=h`Zowb+q1IFs2WtQ>iWO%r>0C9<YWR&!DqjJ
zVrpt~4*%Y%mB(u{BEpztmF@i~go-}77nCS{Dy`Q3PH$spV{dz1&;;C2cTUvq<x^IT
za~lu&NX1&xDDJo3s>F141v-38wlv?|+UYVdQ~0Q<hJCZ11YI7)o2QUZXE-+z^DM@$
zV$)UF0O79f&L<Xndg^~Bed+TU!||}se_wp;?{qIO>Q|<MBJ9Uan#(uEh=Au9knIEk
z*DihD5)+fz()<uxkPerVL+(&6C^35ejsQ58yPI^27csU6Ed-;`YU%prYX_%4UdC`3
z6|ly-QNrGk#GD^#Y%wYeMu<Prsk+PzhFl^R&!0kG>B@Rtn9HTeDLy`+MiOTJl$QfC
zGbJU(0AiTUN8oYNL!N6dbiP#Ou6r>eF=@oP0<e9UJ`swfIT3$8^I+lr&dJby1}=&)
zEp$_PJzJ!=>!4s|E82}e(12%p;C8%Od;fE(7h<rzxcUOPD-#bf?8e4_J+3`mosIn(
zfw_2nok6$70tC=Oqgd3yc?90_jPx}pp`Uzi+#p|(5zRg{ts4^@ED4{W+`nj*-Q8b>
zhle6MYviEv+XJ}0=P^^7u&k8~J1IzQGQ9?4Su_~AK#w-tUxhA0X}?DGl4B#sH604Q
zq9NRu!V<$?KnykCbS4l}!nv4v?<}HRDv8rP;XTuOP%3EV?b)L~7a(TE)yDx<8)%dO
zE2Yyo&131rSovDyIX&`4Q<Jl4U&Iiboq@P<x{)F8=a)vV*uFq6((2|`LYdc|#2V6~
zYKvkH>ODL%<_C1w?_vJ1`@dDfbg?k8O&WR#SBq}SIQU4qsfi^o>$eQk#Rdz1K~bUv
z_M94fKhN>hLZw0!Yry_etHMnxmC>nI*^{qqTC)Z(<yU%?m4i9?&v7eL(>fBMb~=ag
zn4CQwrBh+>wO6h10(;|79+dus-`vYJ{ceJ<r}Fl*qGC)q@d-Aqsjq;eG$TKSwi3+S
zk;axRSX64J1d^%P@&9V(RBWsiE&OfrV`DZYfc&e4xdR3!F=84Ol@TG<jN9vd-r~8n
za-M$v>*I33E^^*81t-OuV@?o3%3nw4rwC+m2PP)q(HJZd>UoAbrTG<Or%<H_A1g(j
zLJth5R<{$HN4P{wAS50S1`0+{TZLTd;=l8j?b?@1p#E|onh21jL3~0F?EVNOQJ=Yf
z_pe`DZvl{-?zg+~ogO291Ak!ada0PQ-Tl^H?FjC5t!%IF@2KK*@Z-h{K)!4Oj#-PP
zqs&MYy;k*B`$ug-#qWjJz&3fKoKba|qvcx(MbI15#_y%LlHXhwEv-fQ<;9&L(~wLO
z(^VkH6Si>O8jvGLv>l!cs6N_8?>r26!|XFt?HKmbH8y!h5ytlH1nF+g&2G%~u0iW)
z>n)R+y#eJLBFb^E-N9Ptu=6jP<HpF;L@Nf4J2?%DS<Qz*UUH2~bA!$Rj=cjgd(KXg
zqC332hQUM(4GwL$IyaomVyT+4gYiTc-VZ!8q%|vHF&YX<IfB_S^BE%*s*A!}$5SAp
zqaO6;^5C2wFNmO)dMg*^f0mk4`WamkP~LGO3(pgXc-78snr#O*cB<>EetSQ;#mY+_
z9?0f%KAQ;&S#NRBJ3cxxvYmG42J~EjzVmKv2job8JkrF(-<ogFvhuF?3C1T++rYhH
zvOIj{>4gxu9el=_WYI)gG}y5yD6mgghMg{yJY-gY6$warwQc8kI(jv(uJ@Z@dO-LP
zx-!w!RC1QJPA$?ZT+F{f>x(69B#y3fm8Y*N2jOvYe7R$f#kKnDQtTZu{b$~|R5Et5
zaqNKnyfmn)hYniTQLbhlO;Ns1!a72SzX+#5Uo+#=k<-90sY;3im2yJmxG?(o^DV4)
zV|^v8KsuT<S$R}nL#%;|(3-TR(%Jq$aY{wnP(Q(_UY^%@0t`${tZTLHEdfB}2Ouzj
z(D)2pWs|PRvh<A93|$_X_CHT7$Wk4rZo4Ua-H&G#I;}bUyCxVsyy_te2wl0$R<(+J
zZSD5fweV~{dqGbFsUdfcPJxcSDnW=}F!{*r;8txtay_)kQ8e+I(yw_1O=0@s>PZ60
zINAhj5OutnR0>=Y$o3;`z^O3-A}|rt9=fng;$dp<C$sy2850}_&z7QG|EX`K!9;eX
z%G&5_g>0KkJ$wc-&lW9-jOx@Ms)8DF8Q!5-?jSkNXNfz3ErfrH?Ho==wvR4pND0>G
zqA8H4eEV~&ye%ZSt1ot&Qm!GfnGkMn8PEd(aa6eGir4Vi^6V2f&mjU7F_UE*0Y4)|
z$~C5%YLR*T_2IhAde~>+y8{}Vpml9@7<zd~o(7-gGr1iQ`G2rc(^ux_8(Uc+RE`xP
zDz?Q4^#)ksLQuU;ci+V9pqvA7<@Tp*0E(5Ck}~M>UENyB$xZd-C1>3^VFjcbAefmX
z_t<L3=2IP39Fo0hV|i?^*9=RH*5So^d(|0{vgp3pR1gzGG&*X1xlv}JD(>n7pX-`p
zxXo2T-|Q(P5zAP#C6f%S2#DbedKp&vPLIji?B+(Ox^^z9Nb)^-cA?RDN4#*kq8g{E
zu?fg&1Bl4JW|dfKYrB$JXdeh@pts0z3w{$x9__QbLFn>O-1und%=MGb7<^MV1cu{d
zO^3O~UuE~Ry`uTjs3f;rjjQIx_g_%U*b#2sogVbNjkk0BER`Y*l{%6#Qw>$WiOVNQ
z`OJ!_8D9A@g4k?FI6$aEFtk5I9|dTVZJXe-A<5=o*@{D+m5s9tZxw70B9Clb+SzH;
zt!c2yCz|3_X%_L*LM$PW0DBA2E`c!`sAP(Aa&ZE9Ki;%^9E1a&n$CaoFq9a(-XO9s
z_qi<AAL?~*!NbCQ9Wb0|cP*{&P@_uuuQ9R$QT(6;d@8IO3zptqOW)^VA|rYU(fw`z
zdyoVJt^eCsUfNTDApGyFgWU)wJ)+?Ln^=KHZQaioDh_^NcyP1JwY057li{>i7!hg2
zi;wY(-~w*A%uFl;5V^Jz0Inyf@-c@<z^zpu?(c_Ue8u)UM(0J?O&>_9%hM<aJMdNa
zV~dPD-LQlFBZn<V?84-3xZ|3F6+S(fT||93M~E__J~!2=oIj5!u?!cnAGI>uSkF_~
z?eYKtxE$~|Z%v>bIgmpsuB@n_wAQy;M8D*nAFWYQZS29-6<;GKlgWe_BXoC|U)vMs
zEM6o+mdajl6;f0w;|$+5zPK;SKIFT&NXmmH7E<c@u?`GWgVV0S&5tg{V=NpK*E;5x
zkUQ(t&6deO$*;$Qd8%cdg4969k1hN3ecE2Ao0u6vH^sREKhD}+wQOuE&`Z6y13B3D
z+W`J9VMk>Hx{F2~0^K3u%ub*oL_z2@x%3;{u6hRsI!ha5Eoxm?k(r~7oSv#GcyPS^
zM3t2;0`2Xf3(u*fb{1k^i3UU(;3+bNVKe4}=zBrD5M>t5Ze|ALBPCE`wF8DyMS_Vf
zy7TbpG{W;OIT-z{o_S9<`f~FLb_3S`16n@b`TreF>h;e!+4_{)eSxa@{FO7kULmhl
zs3Zv8nST_UY#efCv8iqM_@hVVH{d%PD?Fc^(BM%K^VCU8%}9uJKeb~01t`tOhX)`@
zVo_2G4-eN??P4)#shgIeJDu4*VO^ejTtX2MKQL_QeRG3>P;eBfpd>%t$kvi3JbWhW
z{Ap=e1}KD};*@zXVt>QS%2do*v{e3%{pB8<LD<F#e%{YkVBko8`>wD`Q{UO9`E>?<
zYiV_HX24QHL<>)B8m?e|P*7~ag?#^WcQwQrkA^zJIZ=gd$a$*P(nk9e8896miRoL(
zoNR6SbaFZ9Hrc@@gL~|Le{w2js{h`-^q)ehPTeoE;pJfP!@Q}e-BZ^FQ_(cv&+pdQ
zRo(~i=b`8eHRXCf1So&vAS=uVyP3a|bB4TodB3tt-$4l&mE<idI(Vm9LjGVFNi)i8
z+>iRSxONIFz<wFU2)YE4JAgI;n6{|EvWA_1A;6UP9k|WYY?4&g0dOkqqOg{R?p<B!
z$Hrvsvyf4#6%&At!)j8(dOD%}cKhSsK9E<#tZS`meDIb!O0S)7_j?EQ!`?AQ?(H#~
zN4jG`WrvD~my-UDz=PXrwc9sW?{Xw@n_9;4E6bFBWp2%%ot}szLZ?-=!_7w5+eu|)
zzO<N!k~!exK!;$Z&E@q(#hV(2pIR?}23fNyS2NZF;b~+4Gv~=0q%zy`R;nFReW9Yc
z-IwV(Gqoo;Df%JHUz)k+D$)`pm7KZ>E!JYUJ2t^%ICPxq1X87JY-}ywj^$-lQnj!a
z!U9pNX(_5Qj4?tGl>^IzyU<?;^N?wBy@if>&lgq6$CgpDx8}K&A>_a4?}?N(I|yiG
z9_t>^g4JS;SAgpy#oS^hea``h&5VSJz!cI*qt>|XIu0?ECIs7A|GnXN|0BxHHG*N;
zAtEIsNn1w7*9cPsSP4#XYNyiHOk9$X6_43-6*XN*G|{gO33vo_xn!fkvsKns7HQCh
zZr*o#QmV;xtPL)}S>*VNjXC(^3qSLDWvPY>TUnuyXcrvxCPu81Vr|}R@9qf$6fSEj
z6+Tx!(4cwUKpf=s1eoEv?I=a)ECmV&oSm%+)BL^4-46IC5T^amUE@Cyvm?y^UnDHR
zql-Z6@9uE!P32FRB?Pq;Z#gjvh`R+V_g7<(L!|8F-83HajXRKij&ywi^yDNMN%gFi
z>#g^Pd1cJHqMfaGSIaLRcK14GtuQ{nYmKaJY}mKG7w>R{^n9nXP^UW*zZ<BAJbb<j
z_=kR?tGch-!U=@CDKm}0O`v3WALPCUaCt}QNi?RzW=&Uqd49Ph!>RrGiH+^G4e%A%
z4&mnR>EZ4v>)rOX6BsUJJ!^_eyK4;TPEXITE|VXEKY(E9dx&7?dlG3^_m;o4r=f7e
zrP-$r3E?6gOnL9Yq#UMa+RGE{7SG7_we}21uCC7f@+SE?3W;spJvwRCLn)p4AoWdr
zw$I!1R|XWm*ZZ77>EY99!Tc~RKByt6I`E^}dsopFYa4Gqqod0}R9V3LpbW^TR%U-k
zNBAtfE)!U0;WPb6gnF_08A#eZ>aO)iNTMh*Hl|{Wj2F3NykzOXr*Lq!FzM}?agBck
zQWshzy1Y&e_jkzouOTxN)ttx#;gBoqC%+Mzl13qh;yfgC#CN-~Jqy?&BH-k@BpsRo
zwt<BP`X7`o`E-6!X-!>S@rkN4a8xuF!T>y#-_N7U-MvB2svi+Mw;<NA<5zB`o~ACa
zBem9%_{Rvw5Ikc{Bq`=~f=XOJx8rsk;0X}<Gm#K8Fw%~8{kVSvy8^T@xNq=SSG!uj
z&F*LF6CmtPN#_6vo0m2~aLh;y@$clm;MCp#?Fyu3GZwYA63zy7Dw#n1T5CAkcX2Xc
zYvyHY2S+4J#KW=~dD5(F-^NZ{E>6~U>{7#$(G1bb>a`W8bwmgo<+%^a1dOh?i=*zk
zbx`FZA!9N3Mq9S6wawAx;iEfQ*r!1vGceX(OA(58x#m9lr_hh%HpRy9{S^y18GI>&
znNHY=<{eQJR}CmAX-ID}*PQ9xcwHVxe_!8@*&T#p-Tnb66|4=riTotW$X2A(JbV~(
zBu*Qs+j}~cV#`-0r(6X{$^o^&9Vbz_E`|K)_@6C0MkGW;%iLu4_Sc~nY=V`0h^~Fg
z7RMF_Bmzc+AT^L^nKk)OV1H5+8oZJP>NWR`q-zzBs!$WDRMpI9R|xe&r(@!kDGnEe
zASEAFo#4eQfty8~aWW3%kD3!JMCQl0{MhmXG&LZ0F>3sPBJ(KTLnlhVpyIuTM43V#
z@OEl>fzwa)2$Z|g|CPIJB9I2%q037{fUkIfq_GSa2wgHbUq%ClA3i;3(PqgWu^85?
z==$G_;6G!4%^v>L3FMgBx;%aFrKnQ2#>T+-c*ou2%e}uzz3m;?e7aI_Obi6QpHEVX
z4lsj}vxbqz;y+!V`~t*b>(z?7i~;5M&6Sn8rQTtWrOnL<fdc@qEu$8v1eeydmv$4`
zbiYHZ)fJVhZ)|Q+I-H}oG<pD!@-la$5h_<*D9K!mwDz`<K+igj4T%J&C%&iLPeMJt
zAFG>mm1n3I!rD4IVVnnRuS2ilze|pQ4GD2e&DC61Y$uqWKimHXhNHQA<;^`kMESe|
zhnr;Y>W>l!8i>ZYRpguIh$D6%empfUp?VDTMOgd0>U5Gr*MOZl&f<@kr(|`t>1_Uh
zH><#HrR~KjCcvGUd2j;!2O#*tQ=@&_FiB0rLIbdi%);HG3lcwvOFEmMokZx?TYMa4
zoXUD$Cz!dfqG@NW<}el%E8){4bUbtAn!?>TjsC>Q8BsayU(6>@!ka5Be~}45K_Ckv
zo!ZEsyNF8u?rhfpA<NK$M6$AAa}EqsX){Cd*AJ_Z!KQetcUJrB9W7Z1sIncuoIb8P
z@8;Nw)n1?onOHAjqe-Q^oPPxoGZq1rQ|5aBkTnN<nyQ3xV1dM(`n+jD0U7}21D<*?
z_&zN@IR!lEQd$h4XFv=w3dm{y&QaBcsz0o&*we>iBLi5i=!+`7Vsji3@&$;zWVK=p
z&!_O*Pt0`Q6N~GOGq}%Zaw4KdVb7P=E4Dix0g=r$gMLX+E)XWE6l;%3I51-z4qRcF
zH(B>~SECM#SpHh|2~_=V5N<fB3O^+)rvoDU>j5(M#kM|&5=R4*W=c05$i?Ansgjvw
z83sxJc(fwf6UR24uGkk1w-gjkB8H68No5Ah^%vK8c*rK$;+;&hOu}&deA-tvKIUS|
z|CE57dmqu?c*fusE<<_rA&Kt9Drp>EH#uaFWfh9e;#IEW+C1+M#TOIZR@-CaFpz%J
z6DUG-9#)meY}PM|KRwd8!X~>@n(x&>x%Y)R+Kh4CGkQq2dhy&G@tdo>o;uQB{(_}o
zWP>tss-&!MO+yqxDo{?g=t)_=7w|=a=Q#|gL<ztetK+kJKIRXmaEg51)6c*1QzE!x
zYZKxURV&K313tL73Z@Z0d@TN36~w%84m_?vs{H3AvDK{fgn)<|SPhNMQQ0M>Fhxwf
zE`)1z^0o4_|3LTU<$4PrsJg!3%jumtfGWxp<IZ~o9W^Kd2!Uaf?;vU}$?mS@{oVDg
zmjvYoxbTH__xon2Hh8cN7k7a8VZX^81sHrtvNtz7p-p6&U6?(e9-xacY+@!s*+2<N
z+JfA`sqBGu8&0tnG1rd^8!B1SG}XSXX4+4jW0_4&v^t}L>^<=3e+PE2*gb9qOOMb6
zW*HCBY^aE*HG;w5Vi1U=^{M@<V3I>#u6U+ouJ+ve>f+|oEUa%#2ZL1^cf!Lu{x*Nb
z>dZ!;GD&;@9zy)F;#s(@(m0k<!_lwb-?rM4JzrerPHWl!{<+#dsB=y)#A0t-nOm>}
z0!qN_BqlCaqp#FvcoTSvmP#+WP$^<g6pU7SF>x~$_>Dy_S%@q<vUQ1wOh_}~p4d<x
zlnxg`{51tq31uToUgSqnBVP!VkvmKf(P;I^w>XC5L{+S+Si^4rdy|d9nVp31>p@&o
zheU=(oSxa0>%D$|nX}^CauQW_)#esf9wxcBfmCKUMEOAknpIHtO@i`JY;p`faKeCC
zX|irH?&9L{R+sC)jhB9SR3PFy<xTpN4)x1oD)hLLlhF+@*01Wc<oy2au3)YrEvyci
zOTUzxm741as1ntO#8k$RFxCMm?7Wxjp{pZ0@PSD1MTi!m*h1r!Gz)}j8~%jMP*-v#
zD*Kb%OcGFLRu+(9`3)SB*-O74(ObiegIQyhrVnLJ?IV#?d4_<~e0rK5IRuO4X8h%w
zema#%8DjfXNdW|Aj`ez<AVf%K-4k}Xv6n!yJtSUst+bVj)M-zB$hz2}C1Fd-5j*^p
zeXl-_98O9`##!wM7Z)6l&!P=Tq~Ry=4V1#DE1Vi!{ma4aH54Mz*X>?7iaYGkZw)7B
z==l&m(Priu3?Ypft>}JaX+2W<kVwG#u!3+ZX`cu{ZIP27cjAp_p-E0YZKcT$qHoAh
z9n@2V9OjYB8+;VnOC(cA_>-%Stx|id^YsC33@VN!*U{YlbdPe4)>r2AfecA-7%TXj
zt$r@{103#W&I$-uOC_-M$ge0(WGj?(Hr@1K*j~~3EgTjGF<}mcKBZd^KdI*rum3Q_
zkG@Git~Wby^1PQ3>D>Xu^2~E7f|Ewz#5!CtVJLFqIaHE$`)3UVO<$8sT+b(5LGJ<;
zn&~iUuh#bYJ+w*7=^RJSWk9YcS@)2T$m;a`ATKYwNjANwskXn&MzPD#w*Il0GOh*Q
zC`E|{W}Ixn&XA-l=d9+U(iZuR`7Ls5o71w8P`y0MtsNVGdxDd!Az({h8ve|Ih#VJ%
zPE!okZLsX%^9AaOE~7n3?4+;VS=rh?HDxBZJ+NF(wB7F@U*wuq;gAtwL688&V42Lu
z#@_n!1SqV7f=Q$hV0fdy4kyCbXIEF3$H!$vA9d)Y%5mxPaB$gNQ)B`4IA|#lIw+%)
z_Lq~4Qdf5F05)9m&v<vgW23#U?iR2$>^AG&7wd5pJVJmB(Q;ec<>NLM&|UKLw^vm`
zrO~vKG(Tf2jB$bm2M_gwr7HJVzp?A(<GZ6Ll+4>ciW4Fp@EB?KQ$oFXIr)X@hM;`3
zCK!si^+rYn4?D;56C?zuDf`k$u*j*`zzE>Og{gh}t_Mg}*Sd0qMn_1<TlGK2$IQpm
zpYSjsfi(?VZL5DJoMZrEv@Ze;;Kpd8y`PqOp1wjA?<*q_h{sv414<LH_<%cBm$9N|
zdU`T8K31^SG52pqQ`P9A-4U~syH2O=->Dt4yy)ZYPAy%7(TZpa8f&0~bz6W<`y&={
z8cZtat@d4=-fV7!FEB0kl0EcHrgs^!2+L9^ny$&CZ5sLv3noEa7wgwTboj*AVX0)D
zdCu80vdf>p%XE5UeJr`C<wXM9wgRC>!s4+Wdj_>j0Is&!e985gM9?>MTm#<K<hHY8
z<{vB@7dsB=9IMhlTqd1DNGI=u^Zn%mu(t>mSM2W`$*4*p5b=Qw5ut_T-+7xkRG_`g
z;2x&CIG`1TzLl>W6$tgj#kT_d2z6+)rD00KY*O}6SsXg@bDR<NL8*bsf$73Zk-<!~
zl-N5Hnr|6NNq0)l+zSaP0D_5(i~z*W`^15=?cvamK1DO63y)Jzw^gNy6RR1qkJrj~
z@%fjiI1x!^>^a~htxmK|O^q2daqbi^oh1A~u;9rofZiun!G?=HiZ$z4R^hXlfwVz|
zYNjSWErZb!FLFV5%uPt}0&-d9cylbYGmvm72b)<BSrig<Y-@YhgE2QBxBu%fPc#4S
zD+@oNjg7#7$fzrnY6!F%7Ed};=zpcTCHf}o2nr;4+TZy1_5<iI5CZ{M=|^GS{72o3
zYwr`QyTa@7$3K#a2K90~9QAm8|8t}3Z^FXVL;$1eDXaZbmit??%Kf{k0~<f}UtFj3
zFDVMhS;al-hz$z&mF*<i7lNjnQNxIVP9swDZUiV?4z}!Z(5Y@j%2kttgRueQFyC68
zzi{}|L!Y4szJaZGZzW@oO#3-suCc4(zCY}@0e7&9o2u*CN?WG35|Lwn#X<;cj;ls>
zZb3JA0B9du_;yivLu{m;_Nhz*)Mc24O<G(NJP}2zNtC898fxFvKk{y7n-(4bVjk#E
zTW9;{XN_57FV;59nv+BJe^2TKLx*5t)ocL%%BL@|&-Pb6TrRL8Z0CGERpD~zqhHQ-
ztooJ}VZJM0X_PE}@XYCn-0+h22Wm*IzvihUa*o8x0Br4wlV5^kR_aGqRg`&{njlC=
zlI#G`jpsV&7v^*y%J6xH_L%|m!3NlI3IrbiaQ)OaK5Je=65QB)x?c6!^<4m6WUXZo
z7K`2P$0r9;B#DOL&PJ%51GcU3`?G0_gqo&qMtTkqQC#yJo=TBV1>l|8JXv7Iz@0MO
zX<oyr+ylh#ar%ljAL`ylFS-Zue)```?3v~+hc1{*9CQfNcPKkcEV%@CPiH)8-1mkU
z#{B8-F=t!vb2Y|t^C~-|%ZkIEa&agp6A5&m@WfC*$o=It0sF$%2OfMmG+CFB{-fU`
zKelDr;;5QLTO{U(zO%55$dzBY{_Wz4XIZ6+cRvgLN~Q<z5ji3I%pn!RT)MlL+)M+E
z1FJS}3DyA{+2!fw_1PKRi(w!dbmL)kvk|HhI-f9&SeV7qE}g1vBM|?yW@YGUg@6vG
zq+cMcYH|Gzb=YNcg$dk>)A}-mk^}s*fNZ;isR`MEs>a)<BQ~T=x22HE{Nt0HJenQV
z^p|^jIi(@@%|A-cjvqUD$VHF6UXD~&pHZSJvC4Aep=Yg~;5bO_3cp4~n8{g+phI)W
zm(>=^2lt5s`0Q*(V6FlK1|5owYZ>HH5N&*&M)ljvBXi?)Z-6d)I(RF<?0*$(grooG
z{}Xn4Z%SYALj6w&=BrE8)DLpdI3XWDp(&TP{soQ^(k!cw=e16q8!h;wfrC@8$>cC+
z(}!7NSRd^BXf4ZIZh8aZ5G4JJ!Wd_yb7S%p7RU36izBGV6hcy}T1nQ?C$$xD;o77c
zCw(fPN+y93sL?o2+<&o7eWO1_k{dDo=ta_tMe3V*UJAjugSm$FF{|rNBH4JDYK2n}
zsdbQ*eM0esHWfRESkP;;2%g|`9acUJN-WSCe$VdtCgku7&|=s42Vy7fX%sq<zT&e(
z4N8Y#Cn=IBd!-KW>_KSZy8NuPl}!2OPP7CDa^$!7kNg<kW1giomE&iynr_&)0Kx}s
z<KnNq76BBAw~qENs)aY5JB=pJ@IRu;Z(TxRYR4DO+{ygn3=n)VnlhHWMDAe2$Q1EW
z;iydV8pV8f8n^Fo6tU${q~?Md&|k@;iP#wcD4?tY?8om2o@e7HTF+kh^s0W+K5Gqy
zqZRXJ>nUpo@wPgCfE$h-jyEbQ%fHd30E-EzgoPuSNm3HZr$pk0Jg>Wf4IF!eKfAWJ
zfE_%5j~+R1`Q~xuV$jt$7gHUrmzKfgtbGCQ-S6pw24LHHHTLMA@-}A@sZ5l1>-9}t
zN)o*)0S63#2-VW;Ia7j4;!pCkNg;pXJmT3;u_{$Pcr>>j%UsiRu`|VH%u3uss|PKn
zbq3y?jQz#!4(g6Qw0~eZS+;ako?2fn&D5gHaayQ8;oH?Ii3oU=eCh{+{RWUmbl$bl
z;U`B*6qTT%@};ATP-ujki*ZFaNlfjp&zPI*7=U0B;&jBSR>0C$Z?&ZnolM-4JXaAv
z1$b#dq&2aTZp^^G8CBjxfLJEsUJ6;#EK#A*i&C4s>MB{JMuR^b%ile2B%q*YF-L)o
zoFU541vsdbw6u{Xho9{}%tO=*QG{SNSelfF1#_C;nbMwvA*vM|;k%XxF}vy<kV9Cp
z{J&>t2ehR$Lw0X{6ERLG#7<EN(wBjiyW-7*Y|(YY8N5u=!D^r-f^b>Cv?`YB#s-4K
zZ#Ht%EG55wX#xu&RjN%IGW&}hXmo=*VA!<^GUjIQtP%_(2wpzo{{vh;u%7Db3kurm
z5%UKyWMibDpCxIztZ3+EluJ}-7c=VbqARtMfgNXiXBWY5D^|kX%YAc;=8s!~e+#Ch
zj}JD0wX?7t+mj$|y)^7r@})+nrBSM;g4b}_ZyjCv;HFL>mPsEJvyr`*)Ag^HPnAOF
z^SBHpz?Ms$wmQGG_EZi*lcc2n9V>IxTv(YKI^V}BXum@r+|BKG<P=#9UD%Db7{&v3
z)d}+-qY7K{S8OcIZEctkhioFKzPAQ2L7IGy`Fj)Ae$qS1IiwK*KR>^|y*3M5caG4y
zwZ27Z?O<IKhW<$(<*S9QJ$Hs8$ugdcw39ihFMM@K7{FNyY;Y&Ig#fl+pY)*hX}aaK
zB<V`P3=mo|(c_;}fWu!q(gwM@{&%t_!hj2#UCt$95AX;_=PA*0tM9YmsuRa>5FTR@
zjMDc{GyUsXYhkJX)HC=gZC>>Yu#$yf<Xw;~2T`~5=}(rXoT9jksqTAOSmdI|2p1Xm
zl=i-<v;k{r+3uQK0&45PO=Ty}Qc@NRFDIrE@pncGiFE!(*Ih`*g$`y0@CMGscfgK2
zNjv-J{fFItX@>c?17MZU58#5K07F9^Yik#^xu_j<fM@eSvQJ}jmRD~mj+5^Y;3na7
zIsvHvg~o^a`e^yvPq+<=Pgh&=|7>>ba}eCVjK=~F0roqh-$iK8c#}7-72ot>?G?yW
z{NRpbpp+x9+zoRnC}HK-mRGcOpj;sgEIY*qQIzy#XL8F3pes!>2q5S8<yC<S=Ya}7
z31z#K5>)y*j|XGOI}Gt=^wVq{S?RA}9Mc$N$hQn~Jd53r+9y&2HDay5H@Wpl@>7_R
zUnOt}U=U;vvP5OfqKIKJMGqKURw%EpCX{kcbJBCN?!T=1XB({6Ggigm?07mXIw&f-
zd%DLnygWW-FYd0iR62%LpQeL+u3d9~rXpIM=~E16?Wfj8?_bGDVW65;=%}ABRokvt
zJa+K9;@pm|V2C^7b=Ijc=~ZSBO4!P4g-tRxou1svRzo75a*ypZHL_7SdLM6AM2`WR
ze^@S~Q+fROaFw^fuHDsJk@HE5Rr4jp*K-jC5tBz5r0JgvOWJpwSaY-{e`3c{V&DGp
z%93Fqh#F*FKK=Lpc_w7u1sg}HQ12u{65Vg{S`6EXK$1`J)o&=J?bJ9b2Ds@hAb)Pv
z_X~=~n*==4N<$Ba|J$gejJe@m=JuDj4-0r*g)77N!j6C^gKv_D_nU3frk^s*Br)1X
zIo2+!=S!!m3|ox%p0yB~tiMT`uP=e%QSbqSrvS9xoXh`#VLF3|j@gI-9f%MPS6K$t
z+O!S}2AR}Be*1lZ^a^bpf8g1+fntGs!%%)KbKelD3by{Nlc~Q$#G6s4)3A8zbUS83
zM97fQTseGBVYYLJvxhrL23GP(xi&0gPm9YH*eo_Jb;xetssL9u<8MqcH%h&p8C%Uq
zO$?hN@|Eq^miW7-=5iX<L0xC8pQpZw>ZhTQeUb}>Iw-H}323)67Z1^4XAL?DAwmce
zC0!zvEzEFw^1lR~on`ZTGV8V30l(R6*G#*jx$T<T*;?IvS*kRXel5Clks<a*PQB5b
z(Z?^-4XD<V1q5Ry=tS)=v1+p{<SO97A&5^pR9pBYQ7g%a*fe4twyXL%FvX*4iC}Gm
zE^buaBn%#?iYbH?clA?!w(lO*^@B&PA2+*+g~J?*w^KU-L1oX)jC#}AytY-`MM>q;
zAGbhKe)>cPffgd_BgXF$N2S#`6S&e*X#Dsdx8Y?@8t>_1qsgnc&$&>FzW%nwcJuY+
zO3-353vBUtCfNR@|8|lzyV9fL<MJnTqaePt2m@&pJBtU$bk?oYp#<*(^Cjf&lbJ}x
zs#Y|Gk&pN-u0zZ~(rSlG5NjR>^D%!Id}wHxm}xb{j|!#U$#mUYO#w)EPz=?0eX;y-
z9AU~N-a}s8tx9>L939&c*ZwQ2+A0HilbSyM!n{KXyd`#$HLm+K&5pd?5DT*u2jun7
z0-^cHzFI|e2sSd;NRu`gvBWi9@$cD?_x-GB2<hv(xy<zg*cTiUa;0ZKdquT4mErB-
zlVJA^g#}3~6{tQVg#WqlLu<WsGer;HJ$bG(U2k>(QLn-z!?MLeOF$3#tF1^_a*T>&
zK!Ta&)VxL*bw#>x{yY)K_)0vfxc<GmAZX~jfe+G>BF~vHGLmG2opJE?y|CSX^BC;a
zpMoUJ3Xa^zKu5~&qb7|T*o!(C*=7FB_e_p*B2zSXhet%`m3oI^QU4<R{t_Tt7h||7
zVV%As+zQP|^;FZf8u0x7*)a2N%|4n#jkO6g7y7OD2B4n)T>AuUW&MDD{Zlx-%`EWr
zR+()0y}{ygwvhTtuuQ$dUC^Lg;8n|#C7l7a2~m$Jb(pqQvbGvJN%-XUm6!>mAN-=3
zp0VljGUo1&>m0H=cAR|_;-!c%P;D`~K%l^t_xEVfD>mWa;xH!AX@}7HJH)yzt4kr~
zVMU;CJw#LXS44NX`OEdE*N3zBEB?$=DewjMrgu^@#+%+J_}Q%D80n>uiQ$Y4a_@SL
zYzTKHI2eLx1ce{Gg`+l}(GKy-cd%sc`gd6B8%7*Hi|SuT0<{A(ob>ZxK=E5YVpF4+
zp2o&fwtZAQUmw8sdY%jhjb44Nx6A_D$x)$C7h9&lNz&{DpUw34w_I<jPpl1=LY6U<
z^=*z3^JF}vpmRrOw6Z43L3W$M$psWQ@UEqoEzwiE-d9@hdFZ=m)j%Wk(>NKMxyFvV
z;Sx|^L`4fj$%$#!o-KMu6+HC>iKp8^MzrQb;EZ8>n`dsOP}<+4#F<$BjB~wIeqWYq
zXuuKhFV23QDT`mrP%Blo6xa_J^ATfg^$M(>i>f`@7V`OXA<GWp6YBAMf8;jwy_Mv_
zu1}jRRnA_!yDBNK=@iOy6dshg1wI3XNWksrfxuR?M$l{L=8=f|kQnCKIj}G;x{ywY
z!U0Bry2-T6!1L4Culdxj6b6)1XhZB7D6YHxYW(;y;q3W_0jyK~uJnVf0-j>`hmlbW
zCr-2Y<FTK|m37(1GcI3$uQWV$j%!tqrUrp{e(=vjQ7PsynuH<|vS(z;cc?0o+j>oZ
zldWBV6HOTnEMm;mg3T;@9MqXSRbYbchB)k0@*2v>;IHZAD=s<WvP3+>b;0FGI`H5F
zr_L5JW0-(@v}Z>(u)4WHvSvI?1$(GPVYn)s?n_@J7pGRF+VB$NNd%XG3)<J(_&9m^
zW(UOZzf3E-cf7>M(EHO%Pt7uucZK;e5!XWJa3{){{-S}rWJl`%r~l?mlY|P_KP$FB
zOpM#2El?im(DM~4Qb+NHS)MO5IrH{-+R)B$u|z&1?yk=wy$Xtw|Bnd;wfwv%b&f&{
zHs;^O$%c<-nEy^t`(eP_d~SMfxD&_FbEA{ypJtGrVKxWY=eT-c;%cUO^CYn_mBFCx
zy7S3#qk@*V>hG`7=>@Uip5{%v;3c9-kJ6pn*Zis92T$;Q(d5o7`XmK}j$9Gumpi>^
z{LQxAN0&_x!PcvGi*z-hJ2x-%pOca~VWhuDP*c}wTkTPMcQG@O)c*>oT#v>3A$Ezd
zpd<UJ7CXoyLaFn8S;Z9;>Sg>?uuSRY8<em~yOPb6*>By%&&U-qXu_>$N7Vr8wUa!o
zY2s~*p{m^ZeHLpW9x`ZEPdLeyJm^r*HvxR#QRCY88!!)S0Lvb_MCpkKSTI-`vs0+$
zV7M~)Fy9?0ivcVQt(_x*Fmg_fx;5%f>p{2c*7YLp&e_fusd+!ykq;$(MuuEqWLq#Q
zg}Xq*C_$+QyHs!U<IUP6q&3}9Ba?XOf<l?%DsHNf3Y%Ar+yS;Hq6c@q;-3i#t1@5@
zWTb`Ip;Rl`(}1N)UT+7doj%)otwyuJpVixq_Mec(H;v{R8}@@@QCACS+kPOoJ!<GV
z$%A%ZsKFGiT_;11D2-ivI-b=SXsz#gQT!5<GWP@V>vWQWr3uYjwj|DM8YeoZZX{(=
zmN-5r$~;0I|7}kG8%_#v*FgTYlxbl1WZ<b<@HZ;HS#7ZZGt8$u_<p8TufJ29Xu<+*
z*7HY_NO+1_EW3kU@7S(`?VeX`9tUNENO*p=0UyA6QokgPk>mgNOE{uO@fi;chE1Lr
zJul1b8Y=&L92S1-rIzJ(+~D!O*@-qEHwbW-W7fHXW>P^jpP=e&^Ig6B&hMGa>pErD
ze6ng&kE0z)trjNzc^;<-A9jk8`YUYSo`PoF<v^i%z3_0Bc<~@o*pRgMo7_Xat;05u
z`wS)XK9M5+Hjh;fXPBh<uXdAiZ!Zkcp-U}gzm88hO5fTbr)kB&%!O2r=F3#gJfB#u
ztb6rXal}Q_*8b^Ub93YW!wkXc%?{@C(^5BRx3C9#s)0{6RyJcecd?4kurG{FxW6>C
zWxmdjH03XE!#^KbT<7#!>DD_r3#6D7JUqliVdtfv;_50DWnTE&<?SWJ1J}C6CrlmO
z#cPZ^&o&I8S7-k7lPO|9ppB$bB3wp}IVKW5I(CCVq)h?(@SE^QB8ge~A^5HOBhbEU
z2c^2st2Ax;C}s`fKN1xf(Y)oF=dA)h)R#5g{;1^AJcA5|>bUQ^6Za)^>Aw5J3Z}cW
z!$pLLHMW(FG5<cV2H__EU0yrN^?yB`rS2x*{=txEQn3@}?&F8Yq_5lHeF?Nna~R>r
zZBJ<e&U0%`=kd7$mw)6Qc8SJez;xZk_ySej8L;sO_wbYfGmI}JN{Lc(u3wT-Q}s~t
z#^bM;bX&{5pH~2_*tnLqC5w(-2cNPFa`UWN(+V+aDotxAX^rw7{96#qOwQlnq=P3J
zqaSTBHJh}$_Hl3xi;!Vs+7JAln8~nGg?b0|drEUQ-X`7QaJ4O`2GRAT{c}fd<u5`f
zted2{vu2#VabRMlSxT=>(lXHHEx93dFpErA;PmIag0lZUu?m&5nd|3d@%+xh22+R3
zNS2nec;07;e}prh1grZ<Gc_Epsa*%p-U)ak^g-%yZz>|iOx`&3TeHw6KlZ~t*EE`Q
zUSBm{7zc5I4FunH^{1?y-Yq7BLL#>PZ70wFD=)(GX79c&r0P`NZLgq3>3Gr_5a&t0
zS@LljYq#zMC3Sao89^M?+^QXF^yJrhM9a7WDx3?-*aS^b4xh||m8k!M16^XVgMaDV
zA>?&k9f#wejThJBae0H}!8{QM@@!f-cGZN2ukG<Pt}vXh+=dX5P1V4r@spWd#!PNz
zz50j6D6GFvr}7^pz5aS-MB6nVG3E11H99^QBeQ@1kEC-9%e4Kvc(!fZwrxz9T$5|E
zC)=DR+qON~wynu_z1Q=9zxJWy)_Goguk~AdyI(=R2Pl6%&DV58z4Ic4`n@cEr@;Bd
z_y}8F%NF!{T;2YB{CL{7VAoEW|8;r3dZF9Gn96TIP`e9=HYC137#hoL7SGzUO{8Md
z`MsuIUGb3!`Mmds7(4ol9$h$62{dHBNjZ*pgpqYdnnnB4!5@}6f*?ZA<3MJmfZ+7n
zo8VY4nAIy{Hau_o1kA{P7*>{-4!GuFCaFq!iq1~Sin|(LxDR2&Wbr!NZJoEJ8si~L
z-tp2;jaxy0bY3dhn-j3v!*##4yG2Bc%qNAkI_l4bz`)R}MvLq{m*WLz-1mWDG>J7r
z66<%m?3;6{JC0$^yW^U@P?n`9kC`5({%8~03>=}wC~uwRbK(UTW|%#0cbnJtApso~
zq0|(do|3*f20m>@-S!F7x>1YaWL~a3)z6UXu1M-{&+-HA!xH;9Hq|ncO4CX~kMnKC
zo<@ye2|ukbU9L8o)H)uqI(_N#m0W5E$6>21oKzSUP?JNtEyt8&wu+<%+xH-3#K7l`
z9ZPG&ze{%W*LnPQ{1e<aEBtvn&B$|SIIjLS9lLJq^mY4o;leYOBRKAD3zjECfz{)E
zIM&{DHWZB~Nrq`lr;B(lZ1h;kyWVou>3zrt47yGupb60KjTHYit9W7-HA(B~4b++s
zfdPyjo!08p2BS6R6|EaYO*7mS!4<l{k*G0dgj`=M^Xq=Nw^u$EJIy1SXG}7+LLigz
zBFl8=9>BkVk+v)!qY1i|7&VKoBohr^qC9KH&ld%hHRtlp-KRRG#aY#`r;@la(s{$(
zt4K8{&Vb4+acI_Tmp1crX2N(Fw8S<#r&DAa*mlj@G_PGC^&SSYbPS&bb)JZXW30MG
z7`%^;8RASz|3$YdzD#@|p%G#p&9O`l5m>$w1sXyYe6ZTO2qI)M)hCY+;1-z`yxH3N
z*yMBPb$h4iV&Pnke3!O%KfD0iX6UBSd5dZL^>&lqcS#fJxL#>iV^r7E8OQTdRv;KG
z;IK23#s7pmKp0<_gzL@3?VBwl&S%M!K4O~DI*iLK#0!O*w?TVvY79-zni-O<U~&CR
zHDy0{fe=NxcyE+Le><U4TN_3iXzzAssSk%*9t2SNO?c8}iZ+R^ipIaYU9ENfo79*u
zRw5KKO*OuC3*)j4LnIRT`5LXy@0Ka-rRmT%mo@<m<%h9Fj*Z-h;hjSGLNkll7ygfL
z{SUko@15D?cLzeOt=^~gqcsZWAKuEhOQ?Ce$4um*h#8zFN5FTFM8xBL@rszSnI_ZA
zRevTex#L8KyP$=uf&goV65!ov3}60~Veeh|vWMBh;dg0N%AFH&+MyDSojhh<H^-Nz
zw|g{cEvR6{L(aEpQ}D^Z6`&=M#w?O{Kbg^NvvD13j%x1NJL?fS@PzW0Ge_Z$DE<;(
zTnckD5J^B5;$hi=Ab1hJvuIxAeF|Y*4^213YB(n=^DID9-oh!;H^x#RX&n0>BI4O{
zRm*CY3pN=&RVMWuu0AeS_+UChlHBpPjY7}-h!7X99>483Xe5G#ebqsI2Pux?e3kdd
z1X_vNsGe~IJ1|yD(}PqJ7F1VL1WNx~{kQpG<M<njeMl-_#Cn-~U%Iz)u)Oeo%4~jj
zAvF0T_|8gVg#an(ozdqqbuOfP^zwUOl^?pvazWdsO)jr848E@uX`nuDS-fgubx5vo
z@NpvxS~6y<3pXBQ4==}xuo69;vvX{Fe%AlfMm$P_0z^8NGziz4`!R*QKW7#zmX~YR
zug6+%a6mLQR9B1Z7o&XMrqcwUu6N1*;%KJ}R+wkf5G^4hVx^qGhLZheK5=Sjs@#~A
zOiiU#$**VQAgp9_O=U8zfGd|Vq1M=2v8-PsU`jcGgy;|^c_HCr^z=dMwQb_X(Z%Cw
zk2d#ZH0zK*z;<wh#XIacHJ-gi>}5>HiY3xrGp%!Wi%_V7t@C}_WF%Z~e!CuFos(3Y
z!988<hl0keZIkLo^`=xJVFa&#g&FEHX$`;|5Mg)C5x@>aK2Z!?vbtEtRd5T-g{JaP
z=?(A&noRSI&BMnXV{?&pEEulTD$aU))?Z{gAX))cqlXeL^kWZF4S0q@`9iz!oxb}^
zfc{jecXLz-i*8868O+yc#c#Y)3siC!n;kVrQ-ZM>IlQ1GmwUg-LYyFf$@8TouJek*
z`1e~;Wh4&Xj2#PG$Wf0ili0&b?gLer(dKD0O*&j0c^}b>e>r3P$GRyC$&8KYZbC*C
zKbs5<S_!VZ48LKn*Y2(PfVUIeW%{~1Fod$p1Q?t?{=Hl1u!)oz@!v)_Ics#?OmV$N
zPIgzZLi+n-Mnx=vkluG){HP%!zfWGIh<T1vVC+utXT{4Pc03<a-Hm;>>;Sr3#YMs0
zaf{uWfY(s8C^M3g^}=@zbwEhBdXE3WUOcJQVSWJ(jbzXP<;L^jasx5-#I!3umMq&T
z4Hf*&f<`I^7)dF_l62n(6-t0AM^8eiMTcIj$-a=*eLL>Jk-#i4XYzOCZO_{r!RHB5
z>^Z8(K4%N>Jk6X3gq=;KDT6z{O_FKeyHyBIyTWJr+i7JnJ|)zZIJ!=N0$ek$aZSE1
z#-Rcib(}Exl-Rfm;3`Q1m1X#k{=@z~vbkHg4X|KcjCsz2WSJt&?F!TOh`p%U5@uH8
zw8p|Y?T;(=U^n=<d+)f;-(Ns~e!^)tz+)}x#$Au_OeVshKgiqnjQy>BDN`n3v_fWD
z>Sz#i<j~I;EDt7PsK6Zp%kAnDd@cdYc=X$|&#ATQWxJCC{wGbLGeiv%%Du-dY3k@7
z^@AbyYVEFO+lr+qYz8uT5}0%wq!CdWei%#2iFr&sV%9;1lbk)Qmma_W9ja&c*EeCI
z9B%i^ynZ0Hy>Ykphn8c~<4efLVmh<SYk4bU(fuC+Rou6D?a5-wFxNjs=MS%72BXMW
zv5HA48785QHkgANyX4)SxfoLq4;~)|Il}&0laRAN-P;d(MZ!p&c)I*CbVB3^l`l{@
z$hrbc@g|1JwFn16I`Rkm+{Jg8sMtkd7TWjS8Wt}d+Ftl9nGltl!kUJ5gTmZSS5BWt
z=ig^tj#Qnupgp}OtL(lV{f<v#L@JqQY^P+7hYTihHn@e)H+81z#5Z+oO39&_3G{Vd
zICZ*nMm+h^)&AojipdjrFA<rv!!Dk=0fbyi&ld_Wc2{N%NLQ$NvPV~*?=>wCX-*C<
zgrVkkxBJOV!lHeRkI&fI$J>38?Cg3|H6EYGNFOBp)8(dnMf=bh{Ga#HlFD97fwCMI
zIOKeGEkf}Ld0zEQ*{>T^<p8<#Bph?=GwO>G4o0Vm&;myc->^=(d~tZJ43o)!P1;o|
z_?wX9nS|oz2oBlf;0!mWp-O~*kSkWe98D;fXb6U6E3RF6GsPw^I^LfAYTYK+R@p7$
z@L21=#~tIecLPebT^3&)F0a#WC*&0{W&C4}Gj9FvhB_zyzyVBU*iYJO{C%`1w|4aj
z!0bk;b9l~U8%m}DgUxknLOaM1k7oF<JMic|*&&@TQPKtahPNv}hr=<~sL6n|u5YK&
z=A~+mx(f>1FxM1TBCDjenn-U(Y>A2`&fjY`jYOs*;njXpl&+b=&CRCGH|sE;BaI))
z;f$Xv{TP1CVwwe)tj*XzjwH{m5rMtdstz5CruCMt#TKuwSqTi!p>5>V!oa(m+s|!_
zJUB4BK(0Lly>OX7r<Ivu)RDetXgo&~%;2^c^xZQJC0h?eb9{{#orQpSIYm~TcR&m#
zgIS$gtEOr=9Gb+F=qLigl_{=I=lzNo-F!!fM%CHmCypE-3F6BTcK;0Qr<{8uPk|l9
zZIc|!PQYR)Was_$H*<FM2yRqC>XPZlLjEidxV9XH+QbgAk4v4RSq1Phv-2>ssnaDw
z!Ug?d0~ZS%2hW-09_wFVVKZ8jH}DssaRR=6V`o#R7t_!WZSeDG1T}PdiU?-#PmetJ
zR!ECa_zc^x-rfIHe47U0ui@k68bd=Cdm+TSvsGX@?y9r<`HJgDG1nLF-l`kx@e$BC
zk-zu^#7p0<XHqh3U}exTb5jS7wNc8#<=F#L#u3VfL1|A+{E+V~qc-7S`N((*V-EYz
zSOO;nq5<0L>n)W^dBQRS3l^)sLCBWxvm>PNGl}q%t5^a`;I&g$D|1dY5Zd6V4pmN0
zWmK&YSDSB57E}5vYJV@%$L+o)a;z-Z;nil*C#9Gysn{>*a*x|IjI$r(PiSVSp=|ud
zCM6+JVgfBn(<L+zWd>1&wd7$4?><?&9z;<z7Mo{`behb1Ybn2Sde=<#tfK7^hl9yq
zCVx2`K^p8Ojkjy+fauEmbc%WD04$((kC&{ND<waZSNRVEoAq|B<Vqinj9u|iMx9p6
zZLhbA3j4&$1`oQ)ftwLoNkt<B5KhjyY?(rn2#VlJc3u+UansWRI@N*@2|hK=O&x;7
z0}w^h*VxqQh}a#wJ}n|XkJM}(HKjuKNCtid=?OaM#fJFa%*o^*W(j$0_13!wSg2wm
zu}R}cc(5J1U;zgaV5kDWfgP)BYSh!f4kg!{K1uZ99yUOC44ftlxN+0a2!iV5a9+EQ
zIpfacbIk8GnlNkF_Wqb`+4hzG%-y@@>88-$d%nG!CjNSC>2c-FN)c*rHJdMPYcNN~
z<tf$Iz{Dkn2l_DrA=8zKQ}(4QeZTvYx{Q{cM!DoM?7avMViEWwMKG~e7aO>evqW|#
zKe*TG#Y-GsZ;7{ZuBN7kL(f<gD1K7z1BaoG;tKj~hHv7t^7&@hYzOfC?&sbnV`nMM
zhLkNZNqknj@U+4AL)G~?VRyrr&EIe#e_MPm9M~~^r{VBbU>Zl;8&%&gh#MDsioZ#)
zzf2(UoXVpAQBDpQ3C1Le6+46;(Rd^+D+N+Lc$9k+6366rw$P?LZ4lTo0xiW60rEhe
z-VrZ_qQd4Z3x8W=T-w%|Wz`Ez;@n&<(+|7M>ewi+{(*6j0+e@2pnmyyM>#G5JZ=r-
z{v2fQVzNm*xe)eL!Uc<#xNP|GChCrJ<0&;rc-M28+Ri+}<GdgSWw(Z4%ms67Xexi1
zW4ynOX`hVr4!0-Iof7ofafJ4%4kaWgSRBNo=$-n17J&)f`F$q7uQbDa8U9ZYNqH$e
z=i-k*cClXm0Vv>S%eI{!r{%_n7Lx3vr?;t$f8R9hIr#PR58yq#<D;w0!&WZJ9T}Cl
z*r>T^Uc|97&~tafE#N~ayRh&V66#k2nsjHyc$^MA0DEnkXj;EVwW{=Q%N|F-K)QqD
zEaOcKu3+0trfn~B$k=bI_-gG+B%vfe1T&8P_Q^;uNjyl^{s@20MVNo1JX$;WVG`Qt
zwEmpElKC2Vh(B-XClY>TM7FPHWGc3%6RI4F`Y?+)11HAX-8cs+N6w$9VamwKYn;WU
zi^g;^ngp2GT<}`t%Zx<Kb`OMAZ#yCxpWp17`5@yUkSQVc0_na7=5GU&n>J3EhEt`l
zBhi}MypaC^vsIdYd<<FHE8x%62qS`D&J<sai-%Qd)!okYA`>qxolW4s<G$>cDQNn_
zNIl;H3p{1+=n~kmdiN8s6iJgbQYEXtGa!M6>2*7f7f%9S>io;O2S+KGhLS(VQ(x+|
z3)go|!z>lBSJuQ+h5Z8MgAzWSums=^n>B6oJp7jjh%5#VOO4#cNce1xKzkG&q6d`a
zB^EW@#R{vfb|0remqjF;AXiM}htxD6TgCQ%{J4G<napeM2)F{Pal-G<8OpxoaD9tF
z-*&OW>NmAP3JyJX;qbQz8B*^xeGwvM>b5cI&6A{^8N`i>M8C6>PR*mnJK%aG#W4GI
z4gVR0P#je$OI-BS#*Wc)puSiI29M^?c*Kx$Hcwf3zSG^&WUm8;uEYd%)g4>93@4ZG
zw`%-Ygluupcbn>HJ;%9}cRY&UxJVuuC-Y)peW@Vky^EdWPAj`Sek2&c77sO`)&mwS
zpCoXau^f|tijqoAo<^lY!soBAR}4F^w_K=1&CHPxpjz--fuis)Yf1V$ERhUs?xWD#
z$sJ)!UIB7UC=$URxu5*vy=|~BFnFfWrU^8g#V~v-<BlN3j+e_2%@s2;Xx1aoTGx<b
zNGaiwjdcmfT4a))qy)8|>;_%Wil(_(b*4I(JpBzN-?Z5vKMDxtDTd>4+1>6_-;k|O
zAuawsGj}uc=wy=%#7=@VN)vyDqBy*>B`b{6^xPsfrfz&)LJ3>j2wZ<cbA>st?BoH@
z6=A3QfqtC=9YUKXfs5jD$xjgU+7MHl0&U-&Bh&SA9i$=|BO`6`kk+p5+;wtU;0~1V
z*9HcY8cEjJf3(M>)aWHL*8{)ci1OkPA|f_L^7urz*yVkf2<y%>)82HAbW*pAh1!+}
zipaaPtx->ixB=uf6UD61MkUUy#X*vqfDAB^?7)3AGCnPlh@~~mK2&d=dpTbPy{iYv
z4WiLYj0FNFGm(K9^SDbg3qY~6H(WGd!=EQw>Rr~mBMu`3VI`DHEtsP8h^C^y>99hI
z!gP7n8?hv?AXsYi>UmtLLkBL0%-2_Z=10c5zkt%scbX&%{E5f?va_FZ<`|p|d`_3v
zS`_0zOv*5p2}#D>`9+IhC;is(nGx`2VFKr)BxX%UU6_#ns5|xecfL();s1rCEu|yV
z31i#$PNX7W%y_Ln9a)JM2}QTARz+`#$YD+TX5=2maFfnI9Gksze_0H9cN9m+*$z~4
z2W}ouQ~E>EIBnjebAVmhqpyGQ^!9aoT5@p3RID(ZVn|Ns>U0)H?E8EZM*MZ`{so+g
zHg*el;7r&*5`>4fnD6gycDN978XpAyx@)oB{G1((a=#M1mhpWl>9-%n!eeg}AP%kK
z5o3e2%2a5C`zh9D#U#-g4J36ZP45mgV23fl_9Mp&@R<$Utee*oMDJOBd+%|a?qQ5_
zNaPWgqcj^FPM9QO2NDq=K=LLvoXupVA+qDfezd4!bXg=CWgw!pcZgl1_w@|c>jD@B
zd2$m~#`VzTzh&U0*ZE-Yrknps8`tW@sU1|`eZS(|B&5z9MjQ2Dqf{2l<HW_iO>_>?
z6TONv0K;r5fleUL2hW3f6ab6QI6KK};S<N?Dj2pfJpiH0tuhvUXr$sw#q<laI^)h#
z8!0Fr?p-(m-Ybvq9>r$Di3Iu>XM3F(4cc_xHe9|k{Ha3rPT|R;2o0D5Vv!6)$&Rnt
z`*U-Z&UTkqjw+J0Not?}B^bqVk+xEv%(<WN$2hQ0XLQ;;>Uz8LKUcZu;q?7*UYS+$
zdKWqFw=*M`nk^C*8B$rfF)H&vhX?~h5Rv`Qcrv>6Gc|LJ678`QSx1;Ct1|XfyqV5k
zBZ*t+jo;g}<OWAtW+0=@<>}%o$l9ai{>wW~L71e5$abJ|jCKYb`#$k;1xIlt>FMDv
zfj3k|4fryrHb2!)ZNlVTynMJi{=OWe`JdpAx$6>|M-fKaXV^-i|GA>dsWk8Q%mOIW
z*}bl-3$+H)QRh>Oe@EW+_m=6|@Rb`z&2!8wf?(v%dScO3&k_vf%s5pOpm-wxT46HH
zOFaov_dZ2brn$Jx6us`v75DcXfO@bL7*xAlOhvz^z3>N;d!hYKq%FXI_^f4BzFy{_
z%f+`<kx*XH?soQo@wc$Y(^fyZbZpPZ(mz>rDxu{n=(=J+Z}%)S{0|L0Uu|e9K0>0p
zQ7UAEXqm_)T3~wNG&$&U5Y+W13WSg+1eNtpO++5FSA|a`>$EVz>!nIV#Tid)t8jyT
zws<J{+e_A_ad-Ev8`WN)H7<(*w#fUK{U4|smBw!ubqtAFU4SN-Eg0suUt@C~LQt3Q
zWG<T#H6Y!YBj&Y!`CSA=X1etC3E@J8-p<*7(eb?8DTprtFvl@Yito!_5s;>wNaOt}
zvh53D7FuGtUhWFW5>&B%W-s-JUXR`GFSi;<G=c;%v(T$D2=(H!>T)-%vgEc#@lsME
z&1%^%1O+$7HZiluhEnLGL((}c1vect?x*hjiIc><PA)V4DUVlQ{!6Nfn48e0c);Q#
z2w0VI=r}TAJZhL|k@NBsaGA9*m!D=KIAa4z%5pk$X#)E&BrRfer8brRwFvQdR}Z4K
z!S`ff5H@CmNW0PsK25EG3{1?*1r#ho+U%Q9%7~!<Tr{a&GAp^VAuK%+!p|^$DZ}xw
z^6MlR5(r*y!$01HNeuAUdPL4FoNZpAe6YvUY=}(6IdWLh<PT`v_^$U=nB)cU@|EdB
z3l8TY#S)gOl*`IX_1aC-_VhEqYM#jfZ;Q<FI;iwU&lm=FA8*9;HbDyv`_9C?vbsB_
zgJl}2I^HJR2vMfSC6qBSS;3;`9c@go`TfbT`~MFPpyXY&j3+Yz%M2#(0~RSg4iYqi
zS07<ev_ebk>v&BxW8UO01Q^Q?7!zU22aiD3=!pMpqHP$k=R+S;0Gq!DQK`|x+{daU
zzN8)l>{+-|E8lH2I9ok}RzvB%<Uiw29@_<l6t<{1v9qKe+JlG;{}Lla5y`s92d<xC
zY$EkuI7GTZ=iW@J^sd{n>2iYSnOCS?@{vx5LuPE2$%@-@XiH#paA}{LXAp|*DbZwI
zxU@me1`$cDbv$(ejkRyr8;l*ZKnSv2nSA^)f-mmunK<&EW2=N8iD`2^=0d7Q*^A+U
zwa84>B{%77_<H3D3)zKJ@9FKy0qj0cT<VX!c3s*8&;k*|-&aUfq?Ss5U|{mYZiv%z
zlQXKrz&0Y0ZUyQQ)Q^?z`*{2Q(tE>6-E$^TcpyT6M04=g{?#aC<|VFy7Jz)Mg;hM`
zPRGoG{9d!Y(e?8x@Fgk~nQ*Djmfu*&)q2_YeZ|qf=iw1>4S7Nld0CD6sMkjF?8`=X
zaWXH`I#lqnD2?c)miuQ@r)ZdnmQ3pyQ(Huko4(PY$>-Hu;IyRe=uAzdk5c|2ULFoa
zdMMpx-JMBR%@Zf@r-O?(P(uAj_fLeDGqq?Wt71r`cqPkDA=$j)4J^-7ydUTf1edG4
zfFdXt6x`t7aPps;^3wCUy8YH!6HO*V-<=QSYW*&(D5;+3y(qw}{ZBo2!8tyw=^^!N
z?h$hxWHt0;OYnH!>8&G|Uhe#OJorw<26^WXSM7|AfnZo`d)7Q}F}1Xa?-QWHmZY(c
zJCRlWEJPsa>Aku3klv_qpG{(Xtj=W^p0AQL6_<?*`uAXR<NGNB25bs_X;UzHokFrC
zYK^a#;<{$lwqD(DsQNMKzxKJES&b6LH<^)UA&SQNWpp6@egUN2eGBvD`f}iFe=)!7
z*uuj#j7J*`&Z4m;3xP>|U!abewS1rH72o7@^?k~6kBTvZw)W;o(-Vz-?e;#j-Ztn`
z&f-%H&4)Z7G;y451oIS>!{C{5EGCcNI))F2)HgY7-EX_YUTdlbP`1QP*xV8urs;L-
z`&~Lu4F_AsdYlRVT--%ygE|{f&|3M3?y~QJYU2D^lbe3e4tv3{oQv)lccY`8ClK~F
zIElKgze9A~5uicLaMg?i$`s(kh~T)kO`Ammzui@2jHc{N+@vaa3nj>XUHbWc^&<<%
zZx3PZ_Bi~f@8x{udEX4C+#*=8l$6cb1+7BWovW-+ddmJy%p8xf71629g_26b{$SJX
z{hRzQnj?IIu{rRY8vHh~2org!wWhy}z!w-swNmR_An!0v9Hv0$;qkJ_!P+!xJlr%O
z5jRZ_`FG(SIxr7r;7d?w_QJJ&gA4Ga_z)2X{du?7pocg`gP;`RBu$$|%dBrzA_Fx9
z$FolG@g@jTpjDmQHyX?*+MO<-tW+`NgxCsJ7tVtmU`Je@UIl({)AthX#WOQ)1GD6~
z!p3cy*mLm0a<j8HO`A{ien_hubgF}SDyTBN6wkEatXjKU@d?SK{Qgh{^93+QH<|M5
z7u2NolLVVds{>OMsn)MrqA}02=9h6!M@%dDG|nCQjLlQX0h^VJ2i>ORn2uCiw-O{c
zA2veM@iUk=;Y!X-0ax?>*O;1~$5+5qK)dTG&#$+y*#chhMw9c?q%LQ0P>fKq#J*$O
zakggs_>E~~?8^uyAk|{*jRg*dPhM&L#uiDABY6pW7HziQChZxw8a>)QNLiHQG?k(O
zS5+V+OHc%JS+YJBFvs#PL+?_46)QSXU^28+r!vi75^tXYBMi!4|LDC2un_DCLn9j&
zi~1LJIujd~te)%HsSEyOX5G(!g`><5SIT!UyD4yoVLi>KS?*H+%GjVI4D1qFx9_3!
zYd&2;3X=c&d?Wt+eCi1c>;=Xv*{pY(T~@9#Kb%cN^z%2{a(nFjEvr%Xd8=x9y87Dr
zw|M3Toq>x)I?Qp-%lezyp#8r#E>DqPJSfdqM39Y+Hd8qdZ!p~rlAUBBF!@(jKFa;P
zj5tU&&jW{R!RVY(y_iG`Z_iqswL<)=8m>kS#4M>;@xrCXekIkb6KT?vISRY8K!Q1g
zLtA|(W!e~tJcD+!Fj!pBd`W*oD$zAqn&(sXCg8bge>~!yckB+_$zqe4vi;)w)9&!u
z$CXsYRk<yKaRBDQ{YOqp+8FM}V1saZ<2SOT4vBI=p8)kdu3%B$*}oWSrEEE~`VeUw
zGb=VWMhH}*0PnbK?kICe^09bq#j3DIbkNktgdnN;GPtCES}F%gXcV{LgkD@wY7gr9
zF?;e?DQrK2)b3RDV5@-?Yd<z?;-#ja`b{pQZ1*iffk^?=ovX`IQc;76bslr(?r3O}
zG7L9>KQHA=q#OLzOlZm4=66JW+tTw@kjMK?QK@3ZpEb>vhMWIH!mX_TFd4cFl}js4
zLT3``$@2qnc+p$?KYlZlw0C)Ei3dqCb}h{qVM;_W#Gx)Ac(G__mAfXHPo`>UsPei`
z>6vpl(UN(sTh{AbFx7$!b{DgdBdxrG+0{(jVK0E?S>M>Qb+4b9oJ%>u!@mcbY7Or%
zhKr;#Aa%sP+bT9dj|{osu-L^FQZH0mDeslsIf>^=HB4H&^Fg5M;j8R00>4@VViE8S
zq8xdU{mf67`<dGPU#?Wa;wmX;e|igcznXZ*O3fzy96Yt6ss^IDIY77wm*bya{cB^q
zEn4h0x1K__I9sk}H+h|c+61~Cr<PTOqE>apGLTh04;=4ugHd0ez<T*s5U^ir3kOYI
zPiSeC?@=fD|L~Spw>9{ELCdEw;By+AmbejyLvGdCcCZ#mbv6B~+NV}ZuX5vgrB;iy
zU~r0#hX<q#<#wt~){GFcl~uh0fmn%(StpKc0pGoYSJ$&~GI<j?LQ)D6iO7=#T-=aY
z*TIC7*zlC<!tWeN50UfQA;u<yv3jO9s*2{4CH1`CY(tU@yeICS@lM2u!;8qBe!0@i
zcKFLZ=rEGrpDMOhyBlqHfB>gcZ#nwRy6@@6Y|gRsR1SD8;BP%VfTLWM!IdNIwe-XN
z{fN8gWjBo2@-BnarXO?MfkglXMgU-?fEjo|J`A7Lh9_qE@BY)tn(IWVY7`dJTARu8
zpnQt$BNaKpAtwl!@OMkVTQZx>>bP$>OHd_qsjDQ+Z&*4Lls!%tLT!qe`)BU0L~H^^
z#G4>z*yh>fl~P@?t;2|XRyH}+DDLPXL+z_xts4-XNY6osmn;!LURbXZ^zYbJIzg&5
zLuJW6cIS~#PmCiQPzftxqIN2=w6oh>sB$K;AzD(3{~X`<p_tXn_;zB)C5p^AzNDVo
zD50JtC>>k6`*)JV-5s%3XtO2BS~!0iJr5}&YTh|;K(+0uY{&;Cfiw6OZO^H<4?q_V
zc`GsVBuo|0X(QzjJT};9_1t_sMXCH^6XSxm`XZo4sveQKuZth$+)e%4u1Pj6PD@Qm
z`!wrO-~Q^hjr|%;%&z&lDO;SIikk+BEr-MIq|UNonzpaIsu%Dm|EWM;QnX%E3)G(O
zzcOC;k%fI-KHrHyuc4{b8hdT-#1X4V&=DjYRPcSo2J7*z5|+Ok!=+DfMWZ1HlevX~
z1z|JIGZGUQIq34{YBxawmg^3=$i``8|K)m%pUBrIplzpCA(f0@Jc_xr)8_uWc?lw|
z*c`P%!W}M!oW$HPXzKPr8R0w1&wVgO{drsDI2StFz`})ZRRU>hU(qfQUR;X2_7SzR
zZu(7diB`*nX!PPiG@V0E<L^0mAWWy(gnYhKm9jYnSiFt1utYwv!`#)}b2%xo!bZd5
zjarQyN}{9J^m4Z8X9^M_vwr8ELuC4R^ID@I=RHPd4QW-VnPQ|6;1bY!_#bUTCyh$P
z4R}l4vYukmlQHD#KG)-QcQtqKPPD;abv9T5QpxMXrT8%B+ruTh#bf5v<7LO!4k<79
zb%+yrBS;(rCriF#@YvCjFLf@8P6BCi4~2IM@-IjdCb%~Cmy0@y-$LxCX)xfwdAq@B
zxRS?5O{i35Ob&6stQ+X1cJ@&?)oG>_pdu*0WlExPw$(yI!ftyyYf#BA+7@?$FwHvO
zX-)z$7U0(A@j5z3?KiyKYNK~NGs@G~Seya!mW*1pUtLd^8{HQ-9ztKx@kRS<JNG#+
zSJ?uj05trvs0?g{{tr0ZcQfp~o&~^Q^Xb&t9%WeTg7W;^7rV{j7~Bri-M!6@N4*ZW
ze(Sp9TQd*mYx2aq7H?=Yp2)3eax)B;3#|E4H{1x&9+JA;+*lXdC7mkrO<=3Y8$f}i
zbDB`?tLD2tJ5fM!F-tvK@2REck$C0fCLub`0$c1+afm^i<sbZ-*k8y{2r;dG9yFNF
z(}y{2)3}3!<1*>B^7#B~OkZz*XHy-?oho)wUn#a2sC0ViBOhdDD3E1RJ-sTnu^o(O
z1|VO&7v|KVBkOCM?7q>&l5#pa!<<_@+3RHYf!OcQzy5^SVkTKdrN4Mb2wA7()s$_c
zK=?rWN=$W_y2E9yH39XufZI6)ZklYFOwq503$!vSV$i<mKRg&H%H3>gB*FKyDSujl
zrphpA6fcHbLx?=87pzzVK9~$J@I*|j)w;St#8L4Oq`$6*flqZ_`Jkm*^4Wd`;0rsV
zm3L3)OW=>DDuA}I*@d_I=n$>*H}M=pmv2bql3fd!3zld)=H5D#x=`fG3Q^7U+C)yF
z`A^5d6{&_8^eSTnuU8{Qwb(`cNv;m~vCv;1@#^RB5P?684B2%7E$L`~G{9T_8pW(F
zR~vV{s(=UwR%vW_m@$20A-=u>a-eCg@hkurkHM}AVhuTtU>^5M!wWifc)>&lU?b8q
z{KGic&)@4U0PtZD4)hKLj4U7H9#ATrBZw+Ci%tY@p5-FdS<gzC^Exxz&%&0}=^;eU
zP;fD0<|S94{=C^ur?F!zWeG)fECbwfm_P1kg0}AOCOhZ$gET5r11XXPABB#!qD^h7
zMq=tPd_@%WCpnNvM_8i4xF7{+1dLaYa|dklTd3cc1iPA(n;%!iTFovm!H)YUARckm
z`oNdC)?Br2w^2C+VKdPisHB^;{q5&&j!*$C2re6I7)KIk2^YgWlJ*{-OsE6%2wyOT
ztPIDZUm!I*F!zHY$vQBr;P?0;FX}aHHq{8_&s@k{voUTqv!S_HoXu0m+Gt_sgDR|9
zF#mRjtV!9=o%F?@fNM71!O;?IztQS?{dd>(=kjDKm&r<eUe+4+F6Ii7U$^elD7t}m
zlha`tIyHm&9d(ZP?PiOl7k=QDI;mrN=LO(jQKLnyFD`nyU8W0~NNpQ1vtaoG_l-1y
z->=>GkPf#?UauoP+o}Ky6A<%`gZIsM7|Oab6^i81Iuux%cU^N#5CI-_H(>Xvg(W=`
zIJcCEY{eiH=!y;3yn-^NCxsHW6Mf<&>z#Q$ub_xoczXFC1^HvGPhK`cRJ5``#(j{@
z0Wa&}+GryxlAya6m|>KBk#0NZX$=*0bjK5!iVFu~L8Gsa<t=kX8fc02l}^8;=fE(J
z_GF{I%^x#QXyE(}XgO|kqApXVei+RCma)Ki`3I1eLdGlClu=mDQlTjB&bX2?oc3-t
z@x7Rk&#;pq9-=o)b_0jo2faUxVr_iGRLxz<5$#H#D=XM5J)JR4ISPM}8H3OBJtL!A
zv^cYwJIx7ZTh~|1!lU95`&7bE$yqo9#U8DQcNERO#Em+iS4Dg1-}*h;=0`7qG|%_e
zB5!wG&8?cMHDjlgx_$s46Li?wNmVVdMGB`x^%9w#Qbjln(AS5v3T0^uhl6|x8W<{`
zuCaDxQS#1EN-hrTtKZmQ1IxrPYk?965W&aS%=~p5Gk8ZY`R`NF6UZv$S0HgEawxy<
zkK!pR&Ft3+$imJMYcy{X^N9<4ZQNTiW*{1o%6qCnmpE}I(+4Y3n6v>hs?B#(%F{ve
zC8eo_7YW<bP0=fZ#yR&B(Y#}mGSY4oJa5^E1tXwEekN8wDqmuyZ$u&z%pCQ5C#Ahh
zt2#T?!5Ck}#_sz5<8#mtLlO#4iUz)Z;j})>dh2$Ra5f6haLQ&uA9+<xCO0Dve$^z1
zQS86G%!M^4H?s&Bh5CxMYaqM%Zr|$&m#Mw9EP%O}696wqt(>`0AkV6@TILh>eMk{#
zF#l5*BE#RXEs}quhZ2znm(!fi4)77hxnEZDc8kP2dwe5c%vPYyj=@NhHQ_=&u;3x6
z>Y-vm?;t6-fs}bvDs+00`tSt}FQOiYR|B+W8=|r3(`ZfBpHQWpp6y3z?8m@m3p6{_
z$AO`?6wsD*@rTHvF;wF~d*yzy(d9B)AAwGFfBFEXZ`}>_djC1pL17H(!hX}$o8N2n
zV4pwyek%FAhsvRrOkPm~z!R`>HHx`AeAw9CZhNci=Q6M%4Z4RT9=A<AqxrwBC4xU6
zfmxc<*@DV*ohCPhRD3C`xFe&G$(#<zsBn~0VScjFf3@_!7jp}y8@q`xBvjULy8vv)
z>YSN165DmE13E4UO`d^aWM!HG0y(wXYULPzKW-Yr@2T2%>u0Q^Tq^5GF(|Z6%yYqS
zoH3`sn~BTe`SsP?ug+Bx9_t2oNu<BaeGQOs3Wnj}CH_jHTLw0u->T7skwT>^a1#e(
zPw0cjQ3}~s>qSt~2o6LosoG7J5f)q&FyQT=#ZV>P9Bo$V>`=SRNz_CdA=9^EFbJ4J
zw@I00hX~4me^>%MQ)1$XfEyPRgeZr;H+a8;2NZ%|^k7JMY4jj{=J^@zZH>i2#L-?n
zVdb}E)Z61fpUZd14anODwD>C~ckh`>bgj1k&Wrh0O(k$%&Ng>)vV{~&8n{X-3Weg~
zG8gG-gKukXRqQ#$g_<$0pWoe5!pGIhW`CBgu9xa1It{e;R*ogB+jG2ui`-v?-0Sc3
zh8B1?gn&rQ*Ppw1dv-tN#+qpE8j&22#BRbxLOX7ppmln`$t7CDM}~0=hp$Pjc{oL7
zjl#zx7Y(B<n%FNdaZjXimH-CLqFj1nnV+_oZVYLTLKviATm<I}Gn6319Z4uu)FNp(
znbtmz`0k?&QYo(avelWGmAj1jQJd2VWbU#8>3Gl*X?^~9^oAKkN=GmU(p+1fIWCw-
z2jvjDY5j`+#;wQ1PNWg8R9N={Lw8Kle{|wbf_JY7Oa9B+{QkQB;o$nBgCGzg3yJ#9
z1*qK@F;-n=G{!*UQyR(d*W|y&5~UnhI6H64nfGnWMjN^2S#l{24fXD>e~fy_Q23?R
zVHEZL>Q*w5jdzeo4hZb|dTLafdTcOz&x}@i6Y@=Nyc>ib?oN=CGN04%YGDO8E>y^_
zc6TB5Q<r~lrVzIx^PgsRkpZ1OFkudYlo+m8y6W-y)~)8C$N34ANr%f~={zq8iXN)J
z1fcq+LSwL}ve5Oag+ES}4gJ(8B){%*zEU#Ly1zcX$o!r*-6K3YtD7JD#lrxF5V)Eh
zim`N#kqD)Lx3aa-wgO|%qJ227_7hez|F3v5J=UJ){GS%_Td^`n7;g{W#cx{oFgNNt
zjQl&#5sA~T7hHA|T207Gj3Wwh#4wV5;V_W_UjDn~f11Yc2UBKw{`{eCHpA!&j!ch?
zX)VY7EGD1vjo2+;3E?Pa(@w9EuS395Erv7#n?ADJ?sDHdpe0CFC~>~B@wNc?B?>v&
zGZWk~2s+%Yh(G1-2qveaM<c@!Xx$33-LiBUi)}CQ<$-!O6T{RwMP-m5;hXIm4X+oR
zc`CtI8-<_;_x=f5nU9mZ#^;TJ`I#~nb7`6HVJVtQg@5ZrTa0i!<6s7YC*H1iQlN-<
z9As`x<O;u;vL6Z+@bD=8`Jf7yMPhi=A{+JIHy-7{O{}yXyv4NDZ4z?3Z_Vpf2<HNX
zrO#fAPn!Vo08d+%vR=c_SKOBiYt+vgx7@*lrKUTenp2}8D4Nbv#PsQc!}K|H1Cp-S
zG65na09F+W#V&|c_|~2@@3ZIuU>S}LO|&vmEW3dx2^1LmE?(Y9Bex8)7$qqIzs)aP
z;1jSJ_3m$H@;Kcu6S1Tm9Kjbdp$GO`gVCHra0wEOym%kIvq1Bs)6Lp(yp)CDcSa#&
z{$>ptFBeupkpvNSbF=`5(y&LH+z@s%zs;ty61R(4{$jbdWzoR=^+(;g6z+%!-DLoT
z(~fK6JJT4qM7zJc>=`-vnDydAI2aa?4hGmmpMA}y%*g0jij-95cs)8eswi+w7zlE`
zTj-Z7cEP~m3hpbyr@5Z#Xd$pjaai%V<zd(bcsp4od|Bqyga?yP0CnGLa-Pmyb%s{G
z=?4rzWI_Sc!`mc46zVIo96=4n@`a+*<-zOmIlj!G|IX@RL~=hSX3&_gpTIbYR<fK+
zWwI`NVV4N`o8GjZ(nWEWhM<|Z3&UPAH#WTw7L8t<0g?zY@Iv7!2fH{lf3IjT-gRv&
z*f2#K>Wd9anM>%^!{`fJItJ%gg-%d@1a2fOcE;qAd`fj^i~IH%CcR-6Pn&%Ww&f70
zdUfS|aNWLUjpx}w-M*`<YmWE(&3`Dw`6@f#>&K=$CC&rT3cY;3UBSQf^+O{+UT$?+
z_67oE1Z%=I5XnX{ZidHZ`XMH2UIxM~lk`mP&}EzjJ8A{uT9K?6c<`CY*WqM9UU~_M
zhlCzX$u2?{`#tuiT49vzMVyrp=R8J`_ErAZY!FB!&+^D|dhr7;Z?X&)T%7Mv3&MPl
z#hu8^NOajR-j5h7AXx@~#JH0YIKh(|70Yo%W5i`;8mt0Qmp;A%n)@e3?nj?Oa!OIO
zQ(l>bJBG<|r`5gZmdC{dV6Fh_ViEqLsf6D5`!dn|0JL!^lE$Q5CQ~hul*0N`z1YOg
zn56kMfsdclg<qKhzuJMPFEF-cWS7g7sWcmr;A*UA#fpZvT8~2r&_}(Z*sjI(H^KlU
z@8UHE)@1nUY;*WDkbvMdCqF5B>Y&y=!qVL71xtrhpK80Nc){4|r_F`E-`o0d9PwMw
zk5$#Va!We-?qUI3=spKeiqgY*qk6uiJskpv)(yL1col*%Q&$+f(PQLSFqQlDOOyks
zppsudCW1f=)72>I?2ayhN~NX7dcNZrez|(v*lt&z`VYKj41x99x_3EOc+OAAzYl*-
zjTyd)VBt&x4~$t;TNYi_DtqXobT;jC>#s^KYBwevJciqg!HxhlXpByli6yK0dXo>s
zl`Ty_(38UtyvOD#Xi<eD0Z)NHnH#2&Z||gjEC&v4yfwtXm`F;0s!)-T@M)#;Ti5|!
z%!;Z>s#}wM>3T^VT=kSu{c4oi7HCW?4bTPl==6DOKQ`c{rLB#juC{`eK`gbb=!LU0
zpP)q{G1p+seATx8=FRCA?dD3^;nckV4*b>nQ!jwKS``!lRAetWG?Q-o{mLx5W=9YK
z`IwyRP*=X=A`U7pmQlRQJwBZOF1w+2$w6_1DJp`ecU$I|sq@Q_0+*pC33gmiZs-^^
ze@Fz02UMjxO7T&)-l319d~w>Go+{}2R^;Q?w8?>(<uL$jMgxGWq8WXO{6d@LZEeq9
zpk&L_=SNtHQYMhx+r(Eh@GB$Mt^N-hMC}g+#E@s;Y4Xqdr*j*DesmaXk1WgkXd?Bj
z<$vokp&xsC9Jby8FOrpaSFqiUP}qDU@{k^ggnt3DHd0FbB=*_OoIGT=i(R44apGRI
zPbe&MT8$Zge4WHM6hVDH7o9pA$SLSN>`Lj-^obB@Yt67B7#eY+*%5k-)}<5p!%5O1
zoLDhImNxXrZNidgDinHtbfW{8`xkK_!<@$JGxUJJcnfx#8T*X2r$SRIsfgLJ?_`V<
zQFTW?b#xm|Tov&{bhO8x{GWHBJ6X~Vq!<v&sA=>0jvB#e$V=rKMS|1z&a?@pS%jtq
zH3L1+;NJ;8IR4E9GcJ@sB&75kX|)K=ZEf@C`%4tTULys|u_#p7nx+v8q%$ydK{jP?
zViY_-@=)^JQ@^WDI;A?1N&#a7&FwcECSuy{X1rArJ&VkP5zi~o{kMf^D)wK$E!wrD
zZgf^nnzQ{El`lmlCtkukozF9_Z_MjOg9)WHR4JV6<ULkGdPz&~LX21M1^3W{F@&b1
zJ#q44u~~%3wOphy9Xcw^f&7-cC|kZ^cM8;;DK&K1QEpC%M1Q|K?{_H$lNDo&n*wqb
z1m$p7;E5}&!7O6ussSvW4rweC%{Jo;c(=$%RC#Qhpiql0n&c>@Cqba4JDg_NxV3X?
zmD?yb3x|J9l{GcD7H2>=RTP^Z@gZbt3{}ObDW~qafgcDQ&pKq7noGL?Q!jlh<?KHA
zoc(CrZ5sH0L<V;iG00-PF$jqt_X4T(nR^=$JSK)YWu@ZHVOX8u3z99FefznxZY_Cr
za`;2L=LL>(zKm|Qv@r?5SsmMhNw|DyuWR{!#+e)fNhV+i(5B|F%gi{_PtzS?wP_!b
zGP3xa-3Lg$ouT?<Gy4=>*iMF9u8|PO_Mp7d<&uAdxh;_6jeerO-|GhVRj;Er+Mce1
zkdc_Q8%+W9>O$u`YSdxTCwMZ0p7-MfZ-+RE1%{v)S(Yk7hQlkFW;WQ-)vRBiQc?Y(
z62!e~n;s3CxYe9PCzXJB=Hc!)!5`WZQQMHz_9ZG-WPBq9tMn+5@%{P5Y_?uu1+ZBJ
zIuo@clHU{xI5RR()^PUmVuzRwyFEYOC826Tiq^)DZID9?78hHxm{`-sHULEs;H2=m
zJK3FIqw~O_-7V*1JAHvWA1FhJ*id-pwzpfZULVdPB=$0DKx02Q-M16e1_64;?|B#6
z35Y2g0)hVJ2#@!vi*^2%t5p7<ZarH*C#<aeuk`y=m3rN`A5X%i)!i2f?-O>pYqJcA
z%>)Wl_uo1tFlzxFUy2c~WF^Y3)9peEfG-g}9sM_y3KlUAP5H!mP*nyy$Tam-bL^t2
zzab=Q(p7LX;##Yud~+kkuaUV^QXUv_x^`g5o;ZY0-=6n0=?BuVNVq&OmX-;NVda+M
zV|DWP4s-jXU@WwR)dWg-4=}r0`7_VqqRe;v=68}c#AL<{uq|vp1*+kt^q(waw&ytS
zQ2x$g=iuyk-N6!OrO&wn4D*n)rW{o$wE}VyCp=lAjtL%(FvfT@nW&Jv=+6fc{BzSt
z_nAdCSY7HeY|_(2kppJ-B&J97Ow8F3SS}>z6Q)<fmE=86fjg(bpoG8|c)k3vi9llX
zziRQev%wx<I!$wPm=WsfQ0MW)rC7VmTYw~&(bCUV(F9z3O+T0Cp(WY_)w^a9PHYWT
zP`W6(_4931hq|5i)}tc=<*vYX=C7b9sCR7&EOtdERY4sfxm-80a=<1nB=(_8&jM(w
zX}5k_)}~ge2*{!c+Ql7+CN~J9Jc{yK_DJ7G5$&|P@kok5q+YCWg|2k8Gv`hjvFXa6
zGkJ=KPuNgdQ3vMrGFVfBbEhVW!kMBlc(g9oo*@)JoN+N__NXYgCOx*{C~8J6TDga-
z9BeTD>Z~<cxk5pZ`fCjBtnUO%)S<?rOL%+pg?wR~r3N0+uG!{VzaGwH5Y&w{*Quyk
zg3cB|X*fa;38A{D)#-6vU#a&I2lZ|thTCZS%>T0qT)N-@Ih=^t@9<C)jGIGCl99=$
zeqHPOHR{G8arG!P1d`WmWjw!ZzURgLYMDsjd6!LHDeYjgxy**sWoIXeeoMglD3iu>
zbT<f<!p~uJujm1a9NpSz;UJ~XBM_+&0w4#VrUJe3$rBfaLdK7OgLh`k?7ReKf`+>+
zGuX>>SOqA-47iQ}3d2xz-a}n9-0vhpq>wS&Rd1m$XNS%*n`Gi(gwO)?p%_4*)q6Oi
z^X#?WSZ<8f+v@r9F+%rvF~j&`n8DwJ7n!^UCi&QGHJSn8Ss*=l2E?RnSz3VCnQPwz
zY)?C1ZL4@as&b%FI1c!F-j4$mBueO5`|cO4@i<~(H$H;p8iUPl*Yydj`M3LX$+*>K
zjWb`L*KXFs2gEb(Y@l+IiHQ}{S_$c;6{hkNDTdU*kT+`PI!ZpGeC(9@-05HQ%%#qq
zIZEd$%nt~F`3u*&fMwxErRcvTQE4((lv;}yMvJ%$^K5p#Lh*~u6$5|kpIyxwpZKub
z1t4*z7hqr#$QNod>CU!%zcP`I3Mm_@)rwD?8w`NUA3+&9VNda>aD;2jR~<g#)(<Uc
z6C*YBDD?i0MISquct4G;UZNH@O)-}SkdVp*Ut?Jv$OrrLSL~kOE{HQ#AZFl3U6Mbm
zlH)RE3Ne3xfl%3p*36%--lC$s(*B2*7@V(Py}!hrP&nsjj){OeIE})z^^p&4QNaNn
zBI`V;<B$49Xlr14z)XuFKxhPT72ofV9NlWKo`Fi-6kQ$L`#uRbipRfmG{N+2{RhI9
zd%YcuqNkj+0U=$V*T$7hq>>XbaQ<*>P3?@>Lb)TX{2omiWIGmJNXks!Ku;*pgjgH8
z*|Ua~bOm_zAM$BbW0rf6R=iv#lT-2mZf(CX_IRE;wk(Gh+j=o1n!<BE8!+x=Kxt9L
zIOO55NTgP7t6)0xjbutrltIg5Gb+<FM<bO`JPEGh9rmwzOtvZMt^@@C9O(faij~@G
zIlxbUucj4;8nNU%t>K{}e&%67%20ENycD59;M{r_{ew;DAn)6I;o9oInT7QBWjNgO
zac?}LIRdGUq#xfx-#jw<f{3VDBJWb*!BgKKl~l*`MoXM|*Yum7{lMA@vRyL9X2R4l
zch7X^M_#}QR>u6LdPvI*xd|>0yt=RV!B{H>*>nnoX*`iflL{0fQ7_`lmZETUq{$fm
z_?mJjUCFr5bAQ1&U}kc8SOVfiUD8~&>mo4PVyQ;g?`2GbDzYA!P6gEdz%!Gz1pU*D
zEa?Z@KTDr*NpU>4><E@KWU$0p2W$E^bqxPD8pGVQI$x+7exkGnF&CgVvmlL%_}Um-
z=g4s3NSHW;gSD$8W+V3rr;c&j?F*5=GoSAIRY5gBvG;hlffAr{(A(dR?fD4ULy~Op
zOfuExj{rUf3_{tc^Euv=j2ez1TCZ}(UNfo&c%}J<Wgu|hY`^ISXx#qWUvYl#;s$h?
zAhW9N&);*t_MLMm^gA7XJm}ghr?ce!EK#~o6^RkDyau*+10aXv>8*wzNqFvT`OEr&
zC=O^!YzclI%6DheXuFj0U_&`y`gEC4vRR;eVbHIi*!?Dw`$ye|juUgl()xbq#}4@n
z7iYmq)F=+lEeK~?EUJn%Q$|Y0=r=91Bs1>*h1F>=nFf<luP4ljaS@t&)0;WBs)|@*
z{xVBsYB}_`(;_lc{%L|i8w(fg0fCaPfB=+hF$y(po3UHf)%#~yBXsa7VdLI8_XviU
zAQR~r;Ozo3C4v=&sgVnOvf63o9y3~1+9bBIh{oS_l**i`153TQy*LFaK~R2_NI(JC
z0LZE3Ds5?I|K!=LHZW`QEJ-W!@Rgij!m{qAKO#=JfC)?eYp<q!L4MvSsOF>dzbL;x
zppi|A@*cFvaFq*L{$%|sKoOULU6rOw{NR6;{KgrqWA~oVKV$4fZ+{E~Xohzh37KY(
zkwrs1SFOIGZpf!jq~1}YCrAPK>|ZY*P(PN!RJu{v2!P@VgnPpyR<fyUt8_b5*=?@k
zzud0rM^L6BK@gYBrU*WHgm#Fr3)|5w;Fi8CzikwC(WzKt*TL7#kKU4!2&n!6^;1X1
zv$iRH#iNss4@#ycF_}=w@!?-nVIR9sRPyOF*XGm+-+o=BJ(b{}QPU=I4$Quz@Ktyr
zvzi0!1Z}42>*Z<XYUhyn9>&wAmMXUr#lxZ^Vz}{iW)RH|ZJ*=k8#N(6n0E;n8Ni&4
zG!$nGnVN<<J0ty0PL3aKuFTFA%s~eSH4=Ia6^QCba_-Xhz|UYQ(!Dq0V$ycw5VeR3
zCrx2EL+Rnd1S`mPbsT7boE%Wx%^~M^yw&gXCX^~{-wWiWC<L?s-!F&b#~v^+7w~-E
zz(`BxIlYD$kRXwGZ&311Jqn#VsjQz1!$qfH%5<IR@)s9OcXpVHQZ?rXF5*qI!zi4=
z2oKKY-AVtHKy((mjxg}5D-js7sXDfG8^aGz^%m7G`~?elt3J<ykf0psAPKs;O`->a
z`n(=HwR?RS_;u8M^!-lzaX5Ej>?K+PN^DuA1XEc5$I&^4Rr+>Od~!|3WKFhhO}1T=
zZCjIVW71^vWZO-)J=yg>|MyGRRaYNQopV3;-g~X}+gW?N#pB*?S6b~RD}Waw_v&vB
zM>-(fc=JE5pVi2(v|HW+v|RF<$2k8NbnahoPxA_bAFzD3Pu(@B%J0D7{Hg!S-yc}r
zKpKY3sMRVdAG29#z2sOp+;x9g-cLl>rlwa|>g7g~qXtg_F{s@onM}S#H83;8@^98^
zsd~esTIcZy^cau#Et{B?A3Rm1j6@2Lgnm2~DqtAHi0g<b;IVjF``JZShG#=n388o)
z^_nJ><O|wREymQVf`YjP)9YJ-CVnLR47O`~2z!;n_;s?Ar8SSb>=~(KlsFX2P&Z1X
zkQDfH=_o=}HwD!yf~Li@UI`R$GfFstKBZioK1XSE0Ib9wTQ+ZovMEa;))yD()9BrS
z4-KDF*=j`<X9)-I1{%v{50X&boVp8R*fDv=h><@=#$JD;GXTxGUFR#LWO0h<aXMJY
zv-0@&1Axcg!{Nh*G{l`O`uddQuhfK{>F=lg@tL;uUn#gT52OKh04rr3cMm?$^$<d?
zv+LxjgT}2iY9YSKD9mm<zl<y!g3qyUGdzgujs!-A{Df*^q81ULRmlX*gj^iR<#c)o
z-!f5OHe$}JAz+aHtB!m|6tgx)<16-DwQV|<xrPSj;wDDkvYjP;Teyyi3Fis=rnISd
z1C&!6k*_lkEph~2!y2YduHQmDf0xWlFW{gpG{9l!T`3|b0XQhksEr*LctW_Fnm6Gm
z%4aC0dTw2-i8US#O~Tyuf;T8KkOaJ!Kr&D8$xu}$*6aP%a`Q&pEz#`8S&SJ-)WO?l
z1bxRHZt%r(wI-%lEOag}CpO`q&m{y49VWg=!aa(p9tz02=aR~=>4d?pG+*jm;Vcag
zS)!}i2njy@>^Mwb71^tBOB8Vnr<3fiod?I9s?g!~??n#iVIo_gkP~X=Vzz0mu}lSs
z@@i<96A$Bd#ZRUv@xUT~iwQo!j~Qm=x|4hIZWZ??$(S956bHO3;CRTrQ@04#NL524
zN&~y&Y*zZ>za#pF^NJd{e^@8Y6YOU2{4QzL$L)m646|>NI$1RaXk}I9f3gK!GqnRU
z3CR$+OIF3f_roAUS`(mSt^Ufl@cshUfp!B^KwvCEA)h`2z*?eFHyga;{tHRHU4nqw
zC*spPQ0@5vk*MyE`<nlZv<o&Nf}Y@MtzIifaYRS;D-~%qK&faw)M)u%P8kN0&g)7*
zML(w=2n?fBDl9JIWxR-cg+JCE1tCGB1xIdjvairSTVJk2r3T6(Wr@Nz>xJ$}i--2u
zbTJ!vs9v6hNx+5{#t``{r^-ahA~r((5=SBn`CDY_&(z^Le|UO|%P*|aPP(0nD)QRd
z`MSCGn#-4zr$muv$j4!hYW%|NBqJwTgrWGiKr*=wLCSsGbelqhR3vaWO|t433nw17
zK$33j$&oF$)o+g?SRRP^^e`asoFYvX&Yf;)vK9`i<$hUXFHGMSQizmfi@K`HE<*9t
zBu*1Xoe!i{q$v3CbVTubbKw|X9(}6rCr3E{+y#PdS#%uQc|M2-T>${M!usek?WNn}
z8AyZ3a7+)a74&%)6wyd<;eMVsbk`j!j^8tO{^u1MjTLBh_kW5npor4Dfw5Sx&+M$S
zSWKrwS`97Z%+!ldfs59%qE5DQWL0&PUfDbh{(V3JlZ@#6r9}M(3W&PFd>KkyWMC@m
zK4Q>4efOzJqMT-d(ekmZw_5JEiiKDw5OD^-3b7TjR?Wuc{5{hMP2QR-S-BcIui|V~
zj>^v8T$}>GO({y<C9}_fdRK`rGOVvxU@Ql)7?<|0sfXzCCOEQ0G1e1X;JOG0e3&q^
zRFv`%!ZFMED5`=@i3w_{xjObUqjyUG2;o-PZ{gj{<pXWQ!e7pKarRyt=NE|Y`Sf>S
zah!P;UtWFL+yf>S4eP)1++c-w$^3INr1+!T783N}cg_##%fJebg9EP)QkirC0FLCt
zKY^h=uddxF3^q=p61Wu4X>^>4rgkka<d1I3Dp4T0sk_x>`wE2n?vLkEmdSh#jynxM
zf`B~8>N&%UuE@CHUVM&Zu|**&QFP*;P<rE*8T8@?p;Ymf<AiTfHl_KnRZbYNaZ(aS
z_Y*;`Oc(U2%$#xZ^K2P3YbgPdC|F!-l~3cTON79l+{>vA7{wHxtm*xmWiya>O*oJO
z6RFtJcRsr0KWx}j2)QCbp~WbhE1WoP6`St#MizM5?g3`_a=Ba)qg(%#cuSz^$`e*#
zQ0{y`wkXzd!uGa-P9^gnI7#rwr|%~}N`r6I@+v@gVeA#!UKmZfs@!67d<(fOetM9<
zSVbmcu>tszyzZCfWo?CKcuY|PX(tHDFcOk`MN_AfH;&4NPW#YZT!!~L72}x%sKQ{Z
zT=B+7U$g0upNqn+c*O<sWr5m#Cufqg_}4J{iEKk97wUYg8)B<>C<1<nwROt3>@W{M
zmJJZ)lPOS#0?K~381*K335Ed?UdRArs_Zwoiv=DL8}*YM#_w>Y+0i|mm`8OrsC}$i
z^NWCB4n!3I)N=&~uidc8;qu!dmr<>!?Anou=IR;bkJiQcDaSMuo+Mdw#hp5mI8q!$
zKF8Vpc;R3Qg=|%Oi~O~ns3cj+#1mb8c4n}`M;SuM<)>TdLa7iT9&4t(>jDz8rgOFz
zU!7(b05p5Ri^s`g9X^CMKnh`(gAMtl67b3RY~l24i1JvXWPrh{?djk@1&ZPFK=AW<
znozIZ&1z`}D2?vg<YqV3X-9*lIionj0&eT{g0R*a!}Hcfk!4JzXhK6JNvAYiDhgx+
zvt`-i&u+tXe=kL$tNVh@UNo)hJ>A8f+x@z8;$swd0nFGl<HYr|Hia7~MLDoOujoww
ziX9Z^P~l-|D>{CNsjv=FwP$f;yClHluLV$9%ZJE<NJ=qIB;kR46H`W&SM7|~I;5A`
zKm&*$jR-epF&!Vq>mZuW>n^qL_ha*>MKb_>0c#yd7>9fM%qCI1O#8Og%<ZAOHKF{>
z_%ot@_B{2i<w{`2*q&B|aDs#P2^Zyp?>7#k`gq}Efr&B=wEQY6mH54m)IxZE!w}c(
zBS{ed#6ZdLKVYO1g(yIp^QrjJb~>HlgHr1XhDNlzr<EQXbUfSOF=ssw(h+$3M<t(E
zR%OrSeA$=DZVi~(*6Q@UZTujvR`Yu+Hg#PA)jVi$(6ue&HuA98tbiq3cZ_(^nI#l&
zGUBfGSHBe9bgJB#nd^>729%i5n2*v96^fnXjb8)zfwacN(~S(iOJDKqxRdyC89!#^
zz!_pF(31g!eWQ}jZ{r#Of-n|`Te=Hv67x@Pj^xg%DL~A+3vkK<45FK@j@F+ak6XHJ
zPIDm|?d})i*<za5Gc8-d+1grbcsz{B0)~MN-Tz%w4fb77NX`E5@X`N>X8p<dK1M0K
z>$Iw1Ha`u4a8^GwV#&(Kf7PMD<J$lnx_Y&yC*WD~ek8J}&pAL%N}b}FxJx48@um9a
zSJ7UgQ#zHi>Pl6Stx1zZ%e<_p4T;EXc=(%9X~QFR@GBoIPdrx72f2#q^#j^ACqF?#
za6<Y%@<cjE@BpQfnSJ0)1G?M~ux>oFiuVgOd(UEG=GSj28RGcALcJF12U#;G$Z2pj
zp<{^?WKj>9CiA^iL|PN=nSt-(iY70$TMd7HcD>)(U`S_8&zq6mitV%DLaFk#g)ns~
zcezs}9ZhFV9XffY_D3UFH?K;>;V*WT$MjlI#9$u?+k9BiF_^QHju}1|Dj?GECimTt
z){&L)P>zX>?DrH33vM6KE_HMT7?B9^q(u!RE^iKzjtC<HNyPc?w-gF5x>ZT~x$ghp
zQ;zVT02T+3`WG$jSv{Pn=Jp<5zyXYNnna%BP?M93#$;nKaWfB3rkEITWu$HQzG)JU
zUSaTc35DbH;PlQI9x?@9lP-rsHxSbduU*rbNKF(yp`oVVA{~6e(om2d4z_^y&=aW(
zt=(l|bG&Lz!zU#sitz;7|2iC(ce>E^UV%<IZqH2rTl!0e(!Av$oLz)41JmBNjKCS;
z2GlI%DUDT?`PRP?!kR5JcerKdkhu{kWa7mG+7{lxlNfzuiNE&t)?M*;bT29NJ#vTe
zp(KRK9KVOJe*)I=NXlTr9|kdwZ?H}dphX4`CCmtwyOA)OrFwQYO;!G28gWn*OKueh
z44wGbf0OntON<6I?>Njisq}04o62!Vmx7h%cKns>_}orDD{}{ISK0v7>pvH$=WQws
zG9e3<u2ZWL`f)%+CXyvE;x4mc(6E@`JS_$<d8Hx5i~uLizxffvHJaU4e8iG4m8!|B
zx*zEhwr3OD)X6<8J*UJVbmMlSgsszXVx!`sy$RJqEb9SB8J}rJj8RtI#SdhMjm2Z9
z4g#I^QTSO)rwzx}oZ6Fgobp5vpz0;XIcH5ERjI3ByG$?V((Gzn_J*qT0LHa`tHt@B
z*$rN-1l(4!mGFXobviw_x8pkC1;1UkbbG%|`~Szrt*Gf%XB!0811HLo0hZWK*W1Gr
zd)bSXLKxdQKy>^M77Aoge1KGvLJpsby#?_WIEIW(y=JXr@jO7t3``vPVJ+d}&8XKN
zbUxjY8^CHSIl~)Fm%&G53nn!y%|S>=nK9;2$)4GrnSo`B82~rTM1jVRw1yMwfexSK
z@?8W2LF<=0L>dAx&+O5-LPqt~4j+&Pw3Gr#nj|&+n9E<LA{J>=2jfo;qNpQyqi2Za
z!FkmfY1I~(ecU`4JGaH;Dku&^QDJlIr+3@?KhdQm|H&6cVhXgWVf4-F%0DNO3$RF9
zYm9OOG*TcVuzLlIq(z&^MX~ri0Tif{2Y8DD>;d_-!pF`ramV&Jp4r&D^?=uC9~2EQ
zI`(-7V3ZD<CiZW({B-EtG})50i~Q~XOFEUjza9y8$A>9m?5$C4Lj5fqJRE4QR&iLu
znY7BckRNKw?2IuAs0@ru*g1ZrI=_Af5Jk<k2~;&+$Dr8q)xm7x?{UdQz0hX%)tyzQ
zY+gPc0ZhY1|2>yIZ_dLcVnHE^Oxu7X2Y`8Wd(LCt(4~v{0*~(!g`6(uCBnJbUqB=l
zc#^)TyCQD*9j^bHC6$oF77VmYehUN;ZnD^%ev_F^F5W<V7ht2VG#mo?qENH9@-}Dx
zy&-o_`^_;7dxLH_V_=X5FwJ@p`gCt->rhR*B!E1o=V1iyZr+}6zL?6<O;1|E<#&?8
zpMl6;oNMRuvwPPfgiQcF^CieI5^r*X_(moN`tv>PKE9?Plqf+fB6}a{85cVb=}UPI
zicpG=@a$gFD^jJKaLk~h<G*^EKvK;RalgP`3pem!UZJk<W>;!|qN@N0akvb&6UKgv
z*gZ9!dKlI$n-LS=NyUjZX{K1M=Ea_PoT>8Ns2<6B*Q?={_tkU`Ym5Jt5yv@zt)I@}
zKk7%0Th00EGrlhr4CRDjyvm)X_>&!x!U!DvhvIT*p{!ksOkF6ih*X0OAx!bO^bm9s
z3qE;+^;f{aGkW^<`<O%&xrf9KY$OtOrskh9FE|cmfz}SFKjmmBJqpDQ%N~H;?y>3=
zx2>e(mqIuc@q?~Dz5T6uvQ^wr!7+0=+JZ&>@>E`TtM6;<FL=yYY>w<Q^vyE8R{+^w
zqg{7%SEe%L)lcO2`3g)SfcL^2FqwaQyVCgAjhz|$`B>xsUa^0l%kOKp7?}bb^z)e#
zdfSlyut`8vgEqB>%jMs{rZZ0LO<}5_jcjhZEt2)$S~Pqqp@!!ROPkA&n|>6Q8U5k%
z3TcwGnp3@!sIv03;|Ct#;ky<>j28y+XY=%v`}=ZiPFynN8a%kb=mCd(oFfO~A0N<V
zCP=;kI9>n`>is?>cc@O2V~Wxz_B~}WU>+jNtKm7)yR0rbfjJ_65+WT6r*q7~`nMG1
zAKKjUFg1}$F+}+7NrYcZReBPi7_3BENoHL=N0JgBNK`q+a5r(OBu53I&2)%(c>gA>
zPR31W^eo=EP$B_+wVh!6i8I6v?)j|%fW|f1>fO%{%IwV5@@?oL2~*mYneY@IdjSAn
zMr496>m$N{Cd7Q)Zo^F9eLKirK=HI}03Y*n4ej#dQP!r}J!%w@_vYzZ&~<g~PYFp#
zeMYiKQ>L1sY<Y6Q?@10P$bq1xBhJi9B&=FwDS?1JG)z-@{5T>ywT6{yT@bU$cE0tJ
zJ+pN_qo-OBLcHy7?11tr`&MAwRv4v$eS5fkodkc#2oP@(Gr((XZ;D0%#eTK+C+QTa
zYwLQ@aouXp6F$J5X9!606{4ko4hlo|4M*F2kwzo&Mh@vn+}gXI%ob<>P#`CXrr^KJ
zoZ<A-wjJB~09el-U{>mR8ofcC=wB+I1N89c3tjw{aFC=(X{{!+i9peWGhYG{t2z6u
zd!+bK!#KhHmzOBC>?n$fYKn@Y3i$&?3*!d|7*q<u3PiX~z5P$Pz%Spy>a(+bPXu1O
z7^Peer`EmAFBdEJqJxOf9=$6PrE^YCPmyjXA2%vq_QLEnG}JU6xqbgU`sO`)0@RC(
zB>sQq0IAEAycnj754OnbBOePBy|SvRpVnnGKMqzPm~s0a3V0qFs4Ksmj-zKM;w|L^
z@u=Y0o1rs%K3<QJk{BIaM2f`JZm&*9VAd@osI9Hy9mAI@9(*Ew(KP0aRXG@>?*g;x
zHbzLEBd`Is_C*r~T5$wL>aRHB_SA0t*Nj@3`!$54?8VFE4SmLckuW<k4x@LIr>g3_
zI-Y+2dW1o4R|RQgwcqtiZa$nOEZg5HqC{`LU=oy-H3*Wfb4wM<r$e07v|-Po<LMjp
z7&VYmvm_jH;FvbMwPEY{kL35iBSDLzI{nBI*C0|Y>?LdKOk*&3wt0ts`gXcjJ-X1K
z^LTAMT%}uA{<3T8A1}wipDBd3)FRd)M$92@-*=p&4rt|T%JUo8co>6Gw}J_pd};Qx
zcb!DvW4}BnxlrWF{y+t3rY#q+Y9X2juVcwSCS;u|d(!7ZQu@SOs8Aj?uNa-JAOi1A
zUJb?4E-tg`<0$_i+YtJwPeIE5u0v?j<}AxQR9FHR1}i!t%iKH>f&pmL0bu#6(P~Vm
z3-}H=_v73>8T-C=3OMebrqEU}>cT#~e?AQh+U4VO9UAsmKMBT*e$2<<0h?G}%d4u|
zMr++4FOd%!v7gs(8)0G>r%L2=Z?5a(9xqokRCJE+VR}s`8;rU=JgH_#oroP<GBa0m
zH`*W1|8C6KkUT<3nUN)mX%C3=WlXvMv!qnkqA;K&j5^g0L}o;bVLdm<gb1twTYvpa
zDiJ3*Tk{zR0dtS=-M8mj9_*5!kXsUzxs$wp)&1xPX!GlK$Qu&VYh$J?=umMee8T^-
zAUO4-rA806Te&s8%9f$!XyXg#zgew!c*<x=oy3>I0Nhzv;&QA#;=XS6OIq~-PiSzx
zH%9u-1y!Cf)vg9`EdDfWjoaTa%?*Uar~~keSg&`-ZEy1jWBvm9I=>sbR<U3^A5M#Q
z`g84VtnB=yX(smnED#kf2gJq0dH0qM2?X9^l>-mIwHkM#s7ZN0?&}u$3_OJrR?V&i
zz9U(NxC#x;t?S1<2`}(&*R<E0L78i@K8s7bz5x&qPq(deE6R1;mW`_eb<>wZ+Uv;p
zMcV2OLL#}UkmFw_$*eGCOicr6JzJ?JU0YH15=;(nQG4Uk(rNU1CCxde*t4g2S!dOM
zlcwZx_CbR0F3%G<4G+Ga0gZT*&Ep!YzDlC3is7X8IBb8$##>TSl4Eh$wG;W>((qwJ
zqs>6Q7P$hpFRa1PHTn&yvU3(p>})tk(>hF#@6TA}jPn;qQyTWGjiyIfATT?_I&{_3
zw%bKY`Rp!-I;yM~*N&P6y$XAq63Q}rn=WPV3+WBlb;-A(_KN$(G%M_@mRUm{Eyowx
z1MWGZxbm{jX216#cI&?hiH5oowYO0lKoR=pp=E9Y_L%o?jyCT81mw79#)$#$p;r&X
z-6P)7ZOM^YYm|-UD3@aY%bfl685Pqyq3^jg&*$Y~LJQ>P+8&F9HPWn>)2-@W71$*)
z^tb;pHmsG5@nW9kb|#F(=7-~)roUY4A_HnJ{fl-qVj%S>-q&EJ6s+7*oo9x5QL_&A
z!pb6-S(vC}adw>mg&m}m#8Q~NG$;|Kd~3v}OW~!$uD<m_rW*Pd<c*P0vRk&6Q52Ah
znrZyqPlVnYl=+7D+LdSnsi?__Kg>a1RCn#T7GLNB7$y6ufOiEMpR40I_(x%A99g2V
z0{_R;xz^SBM#Hq3J9x1QUl|O}xXs|)Np?Cvp0aowxF}chWPSVAMB}=sqBFg`{_IBl
zrIL)gk|RL`xrH>LSNJmE!$Mf#$Ti96Y-2rk#7>4tI<cWu#CNCfwyoo)=(qzU%hGh;
zwk0AyVDCis%GFN}rzXpkXQ6|>Wcoi01eymarFV^FQRwMUW#v^fSg?zK^*(RLA%NN-
z9rz4{|D4`Y1b58MBe<CW*>^W|`m6z~WKBpsZx7|@;QZhnYj^nf6$HW92#SaOlXaEV
z4%+Umm*lcoQVxmfXsIt2J8ka5m7LAGTU=$H4`kAckpva}e`9a?sWdnqP3h%xN28_0
zJ^1$2m{_kZWtPb|na4`a4%U>eE<-sVJ)X=u1HWWN0D18XQ-DU(7M3DUMhK-O09?pR
zhDL~h#(?-bT$7~g8NP&KtO|EYXtVMOXO6~#3grulT9BY!jsHa^07132vwV+yz+Mbx
zB3T>(lP`^jkVTHbEUiCL836j8F(v$W;ij1%`V7QKJmBI$5A6REQCiZ?p7Y%MJuM)_
znITWs)ae<REjmfty(F{X$WMtIMTnef=xU0Q0J7By4uV7s#o4tn<?S!^gDY{GoIyvT
zIYEpFX4Wv0x~=(*C(7%*&sWY1Cxh}XO?71+pJnF(tv((8iVgC%PAX%Zd`?eYe6A-A
znjceE5vDC`=Avj~#NQ4Pxw2UtyP;lxu%|YO2akRq1y>O5nD5lzx|<D?$)LC2SbVv;
zxu{UDHnaQrTuSSe?3VR3R#88U?65*C<nQ}2GrrejZ_whFvvC4`M;8RPq-X81*E9;S
zqd^{Rskxsb_r>q8$ZFXF@KoyounwCROc5%N07(E0NocyJQLiIYtt3kqAM)qat|&-U
z*(f+87d)PCW8R7}^Hke+KDX20a=Gr-i`jhx*hEv(#caB{?&$%bM48vl-AdVZiZ#Zh
zrI)I;U^3kNXNaPoz`uv>59w6Qmrl<v@%Y^={6@wEgEh!I*DCDYcd>k`Hu1|Du#eyR
zt6>7X5UmS&=pP~srEJMH(W9sFCdgA^ASF3=OjcHGdi5~t{!qCwsbu=_9AT^n!Z<Ys
zQEp00M-u-md^8A%*;(_io@O`7?u-7HPX1L(C>73Ni27gHTF6%_7HlZpd)c`dd>y$$
zaiFI2U_M$_$PvhxvCwNZ0AI1?9;TgSaz+L>2IKi|F)bdaO)iJ)Dmnqi48Jre29MWm
z@|1mXF|D-?zTW*Of~1i<RC&xPnGU?I9>cSEh*~+EvzglV{?V;wD5tOz4)ex9m=Py+
z2IL!Gu<?Lps?DVm{;VdQ@E8H>3SSFJGHCnPEoO2`au>fd<|Ew(?EE1-5u;x7U171N
zro0MVG@<+&Tu26t8sVmBzBi36Lnvm_4(U^SIp~+_<jzWa=f~-(%v!UpLF;xf9Aozl
z*^iAl_&y0lB4+!k^-j<uHgMBw>)5}!B-zkjZFKuTGQ|Jh`}FH_J?s8>F{md~%jSB%
zo-0NN>O9B8z{HEu(vl7V&y_hKp|q~%tCYsa<vZSUo;`l_RZ+gf=sD8)h+)Qj&<*k{
z)wZ7!IsI@G3PPz$K9@CpxEl4*=r`>ReL<?<f&J8p2-`8Xj9k<Ew*xrpZhOr|&Ys(W
z(^!YoS^nowaCSs?#?3RoRbzG^24XZfv%ffOyP|)@g0=85#vpi{><)N@ZK{D}8x0X#
zUq}<A&8cJUg#D?YB<z~8Z4*N5er!@_Bw^U>t$n}HCCwe%mMp&c<tlBLG?*%D=D?ib
z#1-pA(=cCTj=*<A5?E(J@~aRTmQc+lGWxeGQNBNKF8JJb5i$q1q$f#A%mCV`NvJ1F
zhRw&J9I^G=d$<1~+*^s>r5Fph(emxELT+CcQ_>(|#8CIkDYAU73!=LW{YY>^GIj2H
zEZ2l@pdm4~lJgp*77Mcd<BCATys0vVYm#lTp}eX)?%V#}!Te@+v-_1_d<Tbiz)$T5
z-R{kRPil588@rk<a!<eoYSC=F(vAqeb~V{sd`?E_SigH0fnHjJ5&Rp<dT8AE&9mOj
z2xXn!@+X2=`_NG_DL;X{mKM9vlu<u5XpyW$N+C+FAGs2wEqK()?e#P|sa~3EhveG*
z2pntUVOobnBC!z^n5UHXvm}EoKD&u1EJi&`D~_xxH;`zm$#nAtFR_|{j|Uh2{Y#Kt
z+!l2kJ;xDm!zK>L-O6G`lp$0M^!&=0a~i`EOZb-?e2f0Q#fqEaqy@w`+pJ*fgW{Tz
z<;?GRb6b}V&k(|S$kzAIgADw0SjR!#lZZnJGOW-+$hEb&CCJ>3tK=g)FhPmKrTeC2
z8{vV;7@D}SPN~-6{v^qwCM)KlOjxQ0W4TkU0rix2!aj$%3G8N-vo=S2p^=>~VJZ*L
z!FKlBHl~YvzOO*l*tv@Vd2{zU>wkqPHzX<%ORs;o6ID2zII$Bjoaf>GaL(AxmtW>R
za4`F`Szt`G<3e;Xj$U?GWF@1Kic@Vrxkr-p-Gr(E12dV0H6SG2t|jB|RxG&+cJGMn
zAJ36R!hR-=Ytl&axn4z++Y)-VjQDM*4?~X`6$s#AhD;<NZI~MLv91#T#^{%Og1~hf
zREd=f6BzTHW%(i^rATAU9ugo;a8Y}bJHVlH0rXuPjh46T+|I)>zIFaYRBa^wFAwMJ
z*C}unK;P~C5<S23v}5nz|0<LC0<Fa&mK0kyn-nnsT(7n|<L1+ugOWM-=YM^dXA0)G
ze5@D8)WVPm>}=RzbA5Q0GVnT--M`fwNpGmt+XMz;k4GjJrx(cmLm9D&FwxH7{n6i)
zW8TLjZq_egg84O+{U6pb-uO^f>py!hC!2ZZAel64;7{^Sb_01;sl3Q@7d1dFlxh?6
z#4j=UUE=@RKX(4bq0B9tx4W7yp5-$g_>WHO4rjJ-wbP}VG7cQa<w%wTIM3HBR}&nW
zl`g!-z)hFhTyrO;;1qtBg%CHCG-i+pX=Fg^n+%=kztpp_@6^L!*8r-0@Nm5^_F$1q
zoVZpM_tbMLpd-NxEAow`XBjgCyAtw3gW%n-Gl_(AFJW1UjGQN8c!p(N-1*oxHm?VM
zTy$K=15Kd+%)jwI&9&xnZohQ!art_cCHI#?vN3eE>*Xa30K;te2J5ug%_pv!oDEfB
z{CJk_5^tO)a%weEy~O(sZxpyI17<X=ftv69o&S#r5%~jzWob)?&u}`{(1ee5$>7&$
zd18#%<{K{IPOn8Pub)eJ2o)MNE44q-W4===6wVrgS96uhwlFmprD&#Z{Fw1F^C;{!
zahJ}lt`ArsP_Ss?Wj~0`H`9zlU$bal`}MeH!6fJ9Y6SEgfKN90QKaQKxvv?Xl-fk3
zp$c790T*YlZ!l92^q9-FNvKU?UfRXnq!6bM$m)sf`f*FhBmTrl|9&!El+ByNJ^cns
zXWzs1Q+8|ji%pqO2;|hUrNq15(ESHXaW)497hVRl*42U!U4|B)&R_<x$bLr_d=)fe
zJMI`Rq)xKQ5!4N>#QVPcYpNki;9%ZK{W57u+wU2DgD(8#cOV7onfXMxCX?=Rscb`B
zJkb|S@*Gm)W4`_EWQ&<Kewmzn5rmmE$l*D6rqhyIaV<?Xj2lgo^TX7CeOGwhKF!*S
z({2^K?8@)G&VB~ybv5+8EH_qN<Iib^q~>0Y50kf<z+s&^!f+xe8=#a%Ii0>WzOx;;
zMkTUJ4r<3;RZ)B|$<DI`W3r7ZUq_nH5KtxAIcaF+V==HG2WdKSBuz$LR~J#*ju38r
zFd_qdEHaC`MxzM24!sgk*$VilSoYmegd<SeUQltai9|9<i72nAtf0aE$jpq9h|AGz
zaobpb+$3W65u3hL{44)~K*092(XzGKVd;F4LVKe6^0!Kd)85Uj&rz6bQ|@DLn5Y{_
z{dG@kcpjcd9*l2!T)2uv1(=@!a%?NJ@zGQU$a6jwAbc5(G5>MW$iJkO&-pm+FD{c!
z{a&Y0od-=rJ8k9!1s<-L`-4y4(UjLt*d9GBcJ1Qce!Z4CK^|VL?^<WNm1kdM)TOlW
z2W~{dWW!hgeT-chbFdvO#>Gk<5E5<PJbwy@s6|*ehpeO|_tzF7UDAx9eRC_{<cs)^
zNeD}v>&_HkNqA*tx|e5>V|RH>ud#hj;9wt3=SU2(h|?wXTv~IDQQMpGkUEHYw*dC@
zTL7xK=x{je0ZvOnb26EZLsB;oeKH@0NbrJIe*E@!n<EQ@@_yP=%v$<Ic5_Y~G*ASg
zEMPppgsqwf!RJY74@DSM)8#R1!~%hN#?V-=b_wU#4!XtK&U+N`S7Ch<*SiD`UsI#s
z{bb9h^VKH*jpx?K0TyKqSymKZh>bEcf9vE<A5_?fA@<cqhA&<)dC4?O@x~psE}dKp
z${OVU6k1YDhobLuTR}g7`LJzXUjHMG@kV7(>jSR}6^D0gRM-pI=iBwrF+UnP|E+YI
zrMjt=W!<n5c3>#Jr7}TN{8@*#4Yl}Z9k<k1Jx|OQrp@qrJU*hZq0J2|=Wz5~p(CWG
zmsNq~9a0I(P97y{X)#9GQns|gQ^ks-QghY_&%-j8JPiVFSq(MK3uRH2+w&IlZjS4w
zadt%{!)7l0#Pf$I_k7jo@S{-gFZbJyuK_zXvxeGjM)&L7x5q4gUcZD;k3y~dOA@>-
z>Ur4qa7~D@Tt%kP_G}KG0q46Vpq+jor6IiefaUi(x2~^aWw>0H6E^Mtm5~TzEbxqT
z+ctvtfDj(NVAToq%C3}Kv*57)y1|^}(TV$ZbiCp2`F`=+pvRoSx|{1x?VK3yvPg%`
zmM_F71}xzctlb+1xiZ9@HN%!6W4KGWlV$9T#^2$;rjwBGt+($7DOGh@uL5#5r~4-G
z_nWT3=l3#v-8br1I$@jhE@avTigbh~Zk6)3IZ?=FD|OvtZ%S6J!13>`_>NbB>%1`3
z200ABPkEw@@t@|(WI598Q%s06z1YIR{w@@!{Ha53wi+x)okrXOUCH$Fi|;975+4i6
zQzva^PAr<x8RwHsG*=2dhT_0CP7r)2tYLIGahgnfefV!<>Rd<!M6O3->lSvDfZ4*I
zAMlu$TP-zp1=oN{Y$)pwulv;|$erk_<Z`R~b3BI^xQqe*P;WP&(oaJ7_rAkmPGt$2
zEl>ra^g1JXS@kT8s-*fgx=dgCYMfR(_2cwpgh{RfYi(pZT&OnHnLL;|e<1Y|T(t78
zhqofHkg=XjV*?-_Viupk8Nl=loFE`E+~=}Kcz9EZ-!jGM!!UgPFvw*27BlFvU1CPs
zK-g3_>k$+`%`7_dT}3THI(QqCo0f@$Ii5krT<v;otgg%jvzs@=CeZKoIQK)yL+6>J
zVQ}n6Tj{sc(D)X0!_^bGx*U5;C6TQRyRlu{cz;IuZOSB^rll<+*y$|Z8VSo5;C=Ej
zo7rCFLE-@OK=hz<%%FQyEt#v#$I|-Q{)DmLKP!kL{ncgk;@x3XB5_T$eEoW&Cn!Uf
zwouijScvTF{Y|%qJNk5;%#i)m0fi9tA@fF?vxJkgTbLC!T^yq!+nPxW^1gqEpYqDc
z{j?)_5pJqxJW6v#@b4J2vq~J+*Zb2>w+kL(9BA;1_%*!R)ofp{2i%8=nK9?ZyXx!H
z`=N>@JUqmbVj}vBrHV(w5VBHU6debpb%|`LhE@DsxrW6ZWwZW~#qYoD>Y`MR|G@Hz
zAuFwVy_vA<ae6hEm^D-eZ)KH{=9?2YPnEfp^tyUod3`;+{yOOQk<p+A0K139&Mr_h
z%z#}Q=Tv(Jan#UC2Z@DU1<!-b#-(yOZWmsK^8Uab3-3dAgHTt1uHd(0G`S6?kd~6{
z*Y=6R3h{{YZ+}n_TyD|~G!NAg8g9J4b6TV~V}Ko3Lw?3!I{YrB7!6V_savsema+q(
zFLJ*I%co_DQ!(Wz-Ds&v39sO^P!Y&C<+g4Z)s5Q`i9z}Z=V)1L*XZzZbDt%4fYU{<
zeC@C%G>8waTRxZXYnD8mWn=3n0w!d>&35nmOB9c1i?ItAdVlv~mJGJqi*?)H$+O6c
z6QtcIQ{+sY+kC8f(&h*1X=;9v4<ujMMD;yR9k88pui(UMMC+={(?C17mDwuSIS;nV
zu0LlC0S#X;G@o9VskOSyN)}Ey@*@7dr}NQ&HRA9Cke-I=F~4!hU9Z>{&%Tons^>g<
zhJ{X4VJ`9vMXsw(U%eSYF(RH}gh`K7^=WAHiQ2g2^F2I2l+!uvLe{+Vll%yN+Ijc7
z&t=<LR7hpu)l=HJ_vn7T1o^+g?wULa(#g9){G9cF0ql-CEqw)wz?})e_q@yx=@2$3
zRqjvqIjjc4app@IKV6;!-(hMNInml7lLl8SytTG+JFTp1eg^UnO@D7GG$Wh0%(jy&
z;xcq<QYzw48YiW%-fwTpo6LC6J%314$rMMrPTjeInVfQg*{!SS<aIGO;>45X2f5~b
z9O~2~L@kI~V8xV5(nU`|)DYePYYmOeop6JeM^#flqMe?Dhy!Iy2E_{wX!%{i;vZnZ
zUBYt>bGXG{vT`|c)Lhja`B^WsM=_60=UGwk>K~~}Qq@&s;|0KvyCyI1phP|lCFx57
zl1HP)UM*drOI|W{>-HL;c{s_Ol#<t{Om1{-1TizrLuQhTE+bTVJ4ohHJ4W`u3aT_l
z5!k<7ePZzkZl#Xvc^{0a#PJitNu*0S9bM{GC#NaTK`XC(0A!Mb5{1R02&ifTRon`p
zNjNI6CBQYD>+QF`jFqDdvcWEK%i{K>Q1q$M?mx$>;t0!B@UH5z+`sM5^V=OAD<=ud
zp2$C8igMo+&yU9ho03zHk=7HFC`|}Z_MX~j5DL*9SAedqi^aeXHa+fuOhIwB9!=a+
zvf9D(lFx~4^xJi8!bb5`(xuL4?VnIh&*L?CKllhv0lIs1{XbpN?n;vs6^tl=0GqgG
z-2iH~UAyiZQF7kOQ*a;hxgoYw;8zZ$|A59sv|;JuzrFYc?{b|4k*h0_ip{;l@^(fT
z;nm^CZwNU7P9>j?MZab^hu-^LR2EbIfYmQfu~;Yo?xBG9AOvePL^@bUtMFW0yV9+6
zQXi6uSqtB8MjN^OU4ZKQnF&Yn?2O5JmvQO&{TagXm*FDx{bApa_^G&PU5ws)kRqYM
zmf<qU{S$_W&g(M6Jc5z==6Q#R@wK)56x{Lqu`gO3>nYFAqPV#p5*@09{$5qu*S%D1
zi^fy6dD}e5=glfxpAy|B_#HPz(xtKJpM4IYlH9!&&EE{*^YlJWZU~Dv{#NS4zdDJX
z6jiPsOcmlpFll`u2g8_Y1AjO|6)nX{5IO6cp{a@D+F4EMxyyGr!I|SmLKm=rJ@`>;
z#Rsu8`<qh256huDp?+D2YHoSKXMdzl4v%9po7wrP_9$ch9*`fVa@nod{@!F}W{UCo
z-5%v#0XV+xUWm8Hi?!};9tTxSzcYY&*|cW!jZ&WB^|%mj&%kKE_Z8?M?IzOOoDZu%
zZTf+ZLf`Y`z@|&lf_gXHOr2I>qNPz}ELMkOB3Ndo!IZ_*qv+)y{GCuZstL<;YO|Ph
zQ*OqI0uiP*Scgzc_yzDtDU^uz@?3u+3x_YkA*AbFhI|Wm(Wgp5{bso5xggfj@a$&B
z_2PtW&lp-COh)L{w_Jw-_=Za7+m`fm?jEAG8dveTW$>Ow;P;NvFnCTvm&zi#EUjN7
z&ju1blTS?V)ZGMjBrEv~<oY>B{uzcHV{xKU{&^35DMT7>b>?v8ZWpoE;YnI+{Yfe(
zjiyjPan$S3*mPn2vH>Tt=l@>w`Jo5Ox~75(>;9hhdbt7G8}52C`5c}rV|y0^&;i48
zR6lFo(Wwr4^d~Ri4O3Z~T*7pqTqC4Kp9RSFy-}~fZEh}da!mUjZCo?P3OYBB?IeQt
z)-ECco|lLv{B?ZwFTn(h3S|uF*TEmCLG!Lvy^2bXp|U^lGLafqJNKj@_b&(tD&@cD
zqK~Fq2N?77yy?gdROTb)`4MNXEWAJmV}~f1&F@j2w)yfE<O&HD+D%Vds4KOpJoRvd
z$$lZ&|JxcmYQtX&d*<^3XA?vzmwqh#a=QGjg|`T>x<&&@{8$3sHkH|!b&ZruAsdJg
zy*EgE+10kt0GNdNZq2tEoZixeMxBvC6&Ck11}}TI^!BgnoI0)r5IuWPbh45UkazhU
z$DqHri^Z>N1c}&3Ml5UjB>X5&ZWf3hzQT^$xw+k41)it-)IH1H53KEa^Qi&nlV9_i
zhOchJ?gr9{vkOR>hIO@}n?0#YxV)RKo^MhZ<GdqaDFpAI_O7OKeGNZ_&l_f6zxO4`
za9XXpJV3Y)A?z4E)>pK-@Uebbos~zzd*NPAS(dr>C09Pt1w&~&(Um{S&B|cP83lMO
zW6nnz4<;1wS~*72tnJ7eTsfEdjm8lRN4#CfD^;(2_q)mmRPE#`Q?vJ}nrph|<=G&D
zQc4(>KYU^I2Br9H8DGh@b(qlD@#SXLXY~#8OG@09F8P7M!129Q612cG1{}vv?m8f9
zL6<J*e|5<Z?2=)?_^OG3zZAH~##LhedO{fwGkpY)1Ty~Y@+G3zVsTr5j}{bknb|H=
zDFL>N9c~AT0&_0iPIU$KKR=19EZ1Scz_<ho4544RX%kL^0wohW8@z(Zss}|!v~(>3
za)e%FM0gxxgs6;~c<WdANRg&5CfF$EamdvEa8?I492bp1QGNS1_<ToBB8d`<ry;^9
zg1SwDs{o6(WTM@&^tgM>3=EX@+I@bpmE$m(4}YCMk>3<35gg0IT1<6!`|`ofRjk!P
zfB<ZPw4q7r^eCgikD+B=>vdQ#Y6jpC=A*mhY<W0s>IKDeF&Fr^#7y00qq48|6!Hd1
zp5Xx2^4On%Df}H+ZVqVeK*4a~Fj_?Ua_ZJafo%o3xRLR@K1jFV;P1s+qsBd`>H*SE
zl;N>I4EevS-h@R;Fv|P!@)_c%S-%OBc6_*tr233N8MSK}DoX{oQdlhh>)rf;fe%>w
z7;sOhI0YZ76NjIbxcSzlQJh!VTU6U0V<clfbt;-x>ATw0hqFwCnz06#n-WHu9m*ka
z*+YhK<k`mRj+|XhP_P3*ZaH;ey|wtSNsh362jN5mCvWnVJ>W}=g4?S}oWjG?#&O*6
znY%Z{7v>x20e_K%^x?C#PQR7CBt8MaQLwI`w(mGf2w*K;+6d&iQKFa68KNeC*7L}k
z;l*EkYxo6BSTEYl!_;WmdPXl$ME@lu2uKpO-aeeT4gHn!ELQ1)uYc#M2Wf(2&ZQGv
zS?b8<0Pz8kF56w&VjG3hhc3K-?!yEBMUdOY(C!6b4m<%goHB2_<65a2c`}h>FGFw-
zvz+hW*4rvRnui|O4TI3E9jJo|N6tdiYY^MqV+(W!5#ySUse1EkjOux6)6N`51Cp;{
zpmw<V6e*>HbO)*1EHOAOHCUO>DmhW+l3xnV(}C@E@TB5%24pXEOAj6p(#?WVh>9|V
zHF@yt^{Y8v>rRJG(l&s-z@_-o>-q16y_|Pcc+7wGmu26893Cep8#WUsUFYnq<-A>h
zlO&+5V`-;wGdQFamRBl`c{<3jB^cdI7v*x0irJYj)TLQ1^<A`ySOa!Av*Lk5hieXZ
zQ3YrbYWwORaHd#2VxyYXj{7sSb%pwxK(drHSsDIjB12a{;15=SUrN+DN|)WEdHdLm
zxB_g%E+*oaj>$_N=Vr9pW9Lnx*S_>Fe(za%kE?n@#yGsk(K*xMI6`OcoGqW*$sE9t
z4#*NmQv`n?J`d6!x{3Gzhf`V}uuBm);;(Ldqh)_6&ip2D;{ij;v$ck$MrwOMpsF5C
z`bH5Iqfb+ZbI7zFJC?51<<0_$2ZM}HDCnlkHsoobe1Z=#Y&z5&m8#@vybks{otZ+s
zoKk!=5MjU&h`AB-C3+1HRj{r9<mE2uAX;=Vx-U%95%HU$d3!)mFV2P-j||9p{n^o&
zWQC#f#|zW9y&?i8@W`n<tvC2R<k0Ets)f-s=pTR&UPdPCDGWSqm=)zwlce_z?Q3Mq
zHIKXB$B*gD4MF_}xK0}*4K+Ovv+aB=@#wFR{Y7;-?_HvVbQ+3PHZU>QvDt1sIv9n~
zlq%~rR3Fg0-PhkbKemAwO)8!K?*WoQU?~wto@9yTR_^oSSBX8~e@1_y^KR#y*JD3^
zU`E&r*s6jHs5Y5!qA~r67giq@-M4%{Te{!#O(39u{v_HWHOg<kLz*~)U(l}-^x5aK
zgoM#~@VVQO06xzulnOsvLGtR^)n%PS---rVVOivjIQtDDhjZyfEyG#+GwT-{i7c43
zQ33$-sY%)zt@k^yECjvC*oJay)7uBEWtb7vDZgUuJ{oAaY}1fM=T__-r2$X<*N9nv
zO7E%ZAe_y4DV|9Exsn;KVb%r^E#JgGhXHDup5sg9{6VTo&DtIW7QE75&-d^6P1gm$
zjtxd)?*P|{oy00PN_YR7`Igy}P8quz2C@buNCCf*T@S=};efF;J*+b>8aw;Az6Cw&
z9$g5Na#9KTyOY@lcP-?Vi~ASR0!39PZ+i9N|Db#wd_NB&7lDc4!h(KqeUOorXgh>)
zp-C%rZ4iP|B1NK#E`$WXtm5MNk7OG#JK?W{*2llQV-^-7d)C(3<HZ{mN2|(!y1*{^
zKE;ISJ?G?fk^@&ZJWj9X14keG-IpWt!p^?u3^|fOFHcZ)Yk7i6AaQLUH<M9ArD28K
z)As`fPK$>uBb3;&RE@)m<R4yUG~kZa=knjVokk*+9~n84@Tosf0!lFOZ{%LI&^n%q
zN6Mo;RR8L(#n4VWQLA}t9KNJyw1vSIXeav`wf~Hh&xk>Vfl8Me`QF3GJEu4kZpTqr
zT&1StzHx?(iKxo=)d|jKAdy-Ug5-fT@yY1Lf^ZjYi<l}Pgcu=^yj2CKM$qH<AF6-8
z1)khAw2nRjkpw<{qGMy?$z5SVazmR)!HoD-?g_c&D%sCi^+mK<EYLMfe@slMvzi*-
zUwa=wUIE~2_Y7ykU7x^*NHmx5KT0&PAQsTI)2ENRUq4>MkcoCP9o{(qLJqmfSXa#N
zaCz;IWTsRM0Xcvg4STP0Fa<vD>l6(v-&%{;hv{j9KdBq~#j1F|SqT<=1{S7-jF*N@
zyf2Rcx>ceEn61u<TJWY!m3Kx+{hZ};^H``wKp|C3h(;=af*0uZ0OzCkO6YZPIgyY7
zoG%h_UV)b8NC|zqlrIEuxvqpPIT`(Q<(GeNQz#%<GkjQbe*DRhZ)N8m$F|hJ=X%6~
zKJn?DF4TsFj`|ZUk;7<V1VcPG8<=zyiLvVn9A-}2<tL9Vzm%h?*;e0|OOuwCt^(@7
zY@Kx^=D=O)Z0v=_3V~u9r1YJ(I#m?P!f=jJR~_xO)@#%QKB3gj)gf76yEAs?w1~{w
zg+T!4_2cJ4ZVw<>bbW3M)8*d5^i|m!_&>fsK3=sX*#hB&V9<PG)MKGZ138NG%}emN
zZ{~oaf2OoLc^SXH89)FOvFurl*)^N>#T~nB5<k>ydGBBw2CN_$mliJHj9OZowa$O>
zkjG7aUm-}8xf#(HB-;nWM({Uo;l_^S%C@W6FlAe7<K|!eP8k==!3Nf!`C}UeeKwFA
z>uJt}KT$%o2VpF3@?oq&t<eG#M;<R|oIR`3EUW)8S(ib93B%PJf>3bArAdO^2TW1S
z-8+Y&W$|)=nYbw~y$hJg7D<@lBhSJqV#T8}r9m@0_<mIwDV!JLXJf;uU|ig`z<Sp2
ze4PRTpBm>)U*x%wk9?VjZLEilm8(twEV*3@<5%nOy@94llpW=3iy)mQKT09~Zwl>L
zv2K2<QG8+_n|RAdD?Y>*G)20LnQu!yg~@}9cQymi)ALZ^L4*cEw>jj6wcowp8i%K{
z4+y`4hXWgj&mcBEGEXZj{&@Q+$#JG@WzGIai1`qbj6n@Pi9`Nr>xU>WVP<&H25<e>
zhr_Y2qO-jPL=>EhTWh%amPAS0x-e-{0a>mVNzcw*VP*)!m}Y-s#cY*EGg0X1VAQK)
zxNb&&l}4u{)ESFffKZ8wyxwB(4~#l6Ck%i=kbMC1{Ax*(bY@2C=);V896q-;U~D{(
zH*bY{W81;umtGXtRPTU4!W=6{qVv1qs{fMy4(=aZ&4H{zgMYxGv-l>?@{J~r@dnc;
zP|^0Hsnh#`{@tPsEh=_?n}|Wb!~N0D*!gzJz~A@t<9&_l)%WbL?Tm9LqW_k-|9`BF
zDtda?rX~ZQk7r(1b3_8R^oJgAx3iD88FTZg^fs@>Owq77Ha6^5BAE4@zU!$2omhNs
zbrlsEJF!X063t<qla_zOziZ+5?oFLmGu<@DwQ+6~@o3vRX*DW&9JcfGlXF03cc9s^
z`;}wf-+Egn+GNafaUvML89p*m^38&{1Jw~X7A$O!3}<zQn?BIBh=XewY`77{mW9WK
z*Qy2gv_lsrmr(j)eDGgJ4+FLW|Adobaw=hvrZE=Htw_@KzW*+9_GEz9<4KDn4_M37
zLKA=t95r$$@Yl!YERen5n{Z3&wP*kdX@jdRQ6^Rj5k%;2SL}cNB}o^!XKE#IWZ&_w
zap^>vBo#I}^&Wk-0g_m^ncD#*p7of@`IE!L&PjF3?|WKxJ_qA_2CE&G6)K_cE>sw<
zW0#>6^HIh_Q-0nLgEAT;elqb2nCD1(iB$Gdlu;wUa1l#@v8P0#6%{*9DKNuCR?K0U
zf%k^kzJd}FIKPt82kALU+Fk!uK{u~B`7>h2==kS&5Pi*%ZC=uQP99Hwv}&}N%rYV{
zxV!vDYko;Y?Q9iN)oZ+i7jo@bhRU{v(Sl{oMm`PRbfNPP8dw?%s1r=;fm=5|4cIS|
zsriMkZPh)rtx;aQ#SODD_Egsl?V!Lc{uct6K{fWAfp6I%V2mw-=wg$~n#k*(*-5;P
zcRs4npfhMPze;mCas$zhI(!j0p2}ur2A?V}Bk`PY#zX+N0pzPVvAT~DSU1Ta@Pw|#
zuxS%iGu%_kI&px|RbuD;7&B5iS}m;FA*|hh{GPCK*9XiDD%a<H3rVWEB6hVL=f7|y
zMQ4~3!qNl5v0xnGxH2*_M6>39eYUkJyZer(-jpSk-d$}6`Bf@xWw4q7P<p^{kU_6X
zl7Z^{!1TV>qFeVnTMD5$wB6^5;h=J)n6%D>!(il_b~+pENxm7GO8apaIxMS#s52k~
z%-6W*1;Zd!OjiCSR49RUzz8;uaEM(+S+CA0?DD#J>1eyfWYn%q{;FWk9b&#ap3p#V
zYnxdPY*e)sqAN^)fe&M1BQT@;H!c^R@XXp4q^-`8F%dqzD^=+>+*S_1HrCR~`Ko7%
z`*!X3zUkq6Gc^5IBezO+U`?q~@biteb*sfLq1(PNa}J2G_g>xR@H-ziG+7H1b|@IE
z!d(ICpv>M7qwc@nsp;-kMPJl;ua+%IvQ*l}0o2e`Ce|eH1-8U87Y>*}1i30|vvMQ_
zv;Z<a5gkKGsA~wz?zA>YU)Wl#5Rz+{K5VUEl}i_*p`jY+HWXjX^v}p{ji*YAV(#B~
zkd##ffi=9)7!ktJ(H+fqMi!u(^=I=izT2DD>%01dd?C~nPQpY<s}Q(<4Zx1P%x+7?
zj*URYZ9kGHOIiZG4m3!^(Q3^)qjPv;tLlYW_YhkI=lnHGNX4(=E6)tR9Ba9A=3U04
zs_M^$N64WDgpYH8!J^fnE3-?zt0$B)5hp(=j#8QI%iTTEha@}`I`WzO-qN{KOH;GJ
z`#_Sf*WKHh!DsnV(#Js>uWt906(b3^X!~ERhs%vP75{XC$s^=0*0YQH#JRSsY4rX?
zD8JZz&DBd6I2tmjO&P%1HkzcL$Kx4^P6GC2HDsN5JKI+tOYvufd*q~97H&V&rT~vA
z;)Rg|Z`3=v*_Nfdv~E88hTMEb%W6*0QE;gfm#N))NEgZ+g!>G=u1?iJz-47wQz$Py
z*~-ZZTi|;ut63YTmJLT3f~LDr+g_vR%mreri4d)!5hZhVpsuzcJC7G?V?T<AnXp%q
zE-oHJi-h@Q<LO&5u=3M(qrS>F)J&0ntJ4};gdWL=gyow#O^rd#l2mm`_XMGVHhuTx
zI<5d4_GDvjbFYP?G+`dH2f7qeHlc;{2xiuUMl!;QmH!Oho+CauSQfH~lYE34@%*P{
zq7LTgdHkU9-ga+)-2KJl=J{iQ$2vJcOKce~#W4V^Z$=i#&fk3gJm{xaA{B9BDh@Ts
zV!}KUAzi)|CD7Wtb@tX;KW?trPZQ1v3oTZ7ubd@O*wSv9{g0xv461Sqqc9E9Ac%B_
zbcb}Kq=IyN=<e=RM7q06y1Tney1P3i?!Gh3aOV6xU+lft`>eItwJPj?K`DK<`DMzV
zE>rX%0JEkwotMiZy7UPN*)e;f6K$~maL`atEXwU;%rLIzC6lBZl{=fE*659VK1X9q
zs1Xx^TJLkQ4jY=i^?F2{&4;<|Z@UWEqK{qoI0`Hl#rRQf(P9TxmO|s1>VqP3BV|`w
z31i#uW4W9~!Owl-ip_ow#0z5=Vt!}OouA(T&2j|5pIf>UDUUI6L=3mrmAkq)8w+dp
z-07r$cW2X^kSN>4<3~{O;{(D&eK_&M5!;sAZG{E`rj>8#PPiS<BkS`Ov>FX(K@r2a
ziYhsRpDI_oWQ%tP!XAp8)E9xk)NtCbOr?5k_%`;c`cO5wd2vT~$WlO_D%BbSfaf>2
zvd;&0c8F42AM$Ns1=LAQlTgQTH5=H7Q}w7<DzL;y%;j<}Sz}lvvh9=(>EpH3Zm%C<
zMrlwapeA{cbw$WE1o2Ere)b*Qqc|s;@#wm(JO=0Uu6zBAB$ngJ3}FoZ4LheP&33!Q
zVFJ)<YnHXKUlbt&l_8hVsgkrO4M^H(UT=T&X)U#7^MBXEhd)Q`0znHWbIpq4UX;_q
zNrnu!-IK1e7Lzp>kYsPt!Qp#CV7Oww1Jx0n&!apns%l3Mzs`X8IcoUfO$Q)Bv{b(?
zw!4t3ILvFvYG`~57p?G5E$!fmpM6_V_^a+p3^E;zcKiznE==47R32muL9Syfy6xRN
zw}@glXrpt7DT!b1Ck`%=oXZ@Gm|S>yo4Pj5Q>OElv!;|a7@c9>f;WQ}W^TvIFPf5u
z`?rPdH8Wjy`6@(ug*L1zs^2Wir1y5na!#klMvl&mIgj9-K&(p&Y#UP#Eu!wn2#hYc
zSV$VlG0zLqwa2bI=j>p?nd8MIbuO%xtwAdw$$jd)yTN(QPVqk5)zjO@^wJ{ce~_K=
zB{Ds0Qh)Q~hW}hV{_p4wZ#GB(N!R_-^#@DXHds|uF$?sBV@ER)g+%;gW{sy{_Ggg1
zk$3@p$CF$t@<@yXgQAsbVQQYEk?@|dwIYr|72G>*bz9;+1(++8wiOO|mR0jNGEh@$
zlSTS#GpL*XLO%R!Da7ov#lmAFPumk)vrrp6ocI{Te08>+*#_cEaeZJ*?s3p-mr<&<
zHjm_J@%V~F=XUVv{Te^EEZsuy{);)cD&tnVq>zcnSJ?%>6!fpo%60m_j`7kV`Nwkg
zp*%gr@K;2P9h7u7TSvj!7l&a9|5I4bXV$g+`@N=%=90Mw+UdiQ5{iX&ulfJ<D=oej
zhb!7WQy9lyvc}8gkf9(#N{&aY-bfk-#d9kGdkQknYt-^;F<7<$3d(EK8JM8wc6iZm
zhVfL{_DrLgT;tw5G>wak-hB1%ER*1AqdSBrii)?6P5Kc?exPiD{L|%w!`|O=ox5E^
z<@EizcGF@ZFFlgisV+B^Tj1f)a*6Y!j=7;jvxVm$IFAy}OR$2LoVHgFMiXD=@>{;1
z1qUm}^gyK`baN>8b-=oTD>qlJr3jQ_x}=P`ZspK|u2|!m2m<+BQ&S#?UX_m}*OH$|
zcS9&*7^cc;R$`@OEwt!kYWb-Hu&$qGURZgr)T2que+7{6zFrAB)Ljm~M6_}_Y4R?9
z3NLeR)GiBx(|AuZq+Bo#Cyr&Lnv`B#E024G-e@8D{G|+P;7Tq&LP0S<EOTqYv-30-
zv$=pm#qXR9LVYg|`hDbFcg9~YGVVwPI&9X?e1tvk&%v4n{wdMcZM;1H^CKJ90s^4b
zLe)vD%i5C9dd2LiM5YDVTv4OF(`MigfW$-Q)6bnhM}EfL<{*y#<q!c;ltT5r`J6`Z
z9btf9wrVM)Dj`dsmts&iY_xY}XSk}62A{ntRsz3~f~{weK?%p<Cso6j8nGIDBkOSO
z1nwES%}=DEC=FcEB~sPFn!J(OW>|GR1ai{_@;--^z`HCUSvJO{<o>NkB}97R7&|^~
zOF@jF^sxzALy(2<(=IszPmRQwp*L@dILxV^5{^v!6P-A-+l(#Tqo9Kf-cgxLZF9}S
z<v>;M+P5h*LJ~cFK#J2Sx_bfV1`~J4@$JKt61OF5;^Xx`*X0A2q!AfTm=H(Ar4p^I
z07>D24ae1zpU3!)eE?lU|Ir*I!eyVCtpkt3b9clEKBP?qsm;|I30dYN`#4Q?_z^h@
zvw;)Tqvq3;!7N)jOX$UwYKr*NbZ*NsrT9c^r2X7^M$+(s5BzxtIN5t$UR);hw@S=J
zOGUh|1?Lfqx!Rqm%zd@QS#l2=L*GEIOJ|udZ_0!@)x5qQ8}%`Q@h2paWqn>vx6>9`
zdk#D3WOVro%lLG8uh($ePB*pG(P@W>Wm(w2WN}eOXMk@N(ta}dn2@7vvP_=KrZsU-
zRM{-~B*>20u+Y38Y=8#BpFZ4%4Nx~oF}@MPortF9CDT2{oI*nKM^}S3Fy;5co<by~
zsot5tcxy5DCymYP(Lp?eng!`L<9{o}^4+dy<<?d3<D*#CIY{CFg}iuY|Lm}M>ugqA
z9sq&p1CrzArM8N38TJ51Hn5joFumT|lo0Oe-SY>(`fr(r2f(3!@ffIPzXJ_ct{kuz
z2D%rU`d@i=5Q~Y%8Zrk}xN|m<?fX^rS{R3QJ7fzuhS1HKvP@ehriP5ZDG_3a7qG7C
z9wgd@KrZ1Cdaxw_m=2BYZ-lXZNHgLhW!9W3S33iXUEI|K(|Z$iS=`<8X}-+Md_;t9
z;N|@d<l;K3f9T8UW2SKMU6T@R6aHxi#`K<vawZL|a3euKj>;NgqlVNN-Z)QeTWy%i
zm7WtS_k&U~f=7j#&VA%76#|jtB-9?gbJ9o*#+Od;a0x3Ll&>{xJu|p8-gq_v-U!5O
z$oUWJ*SJ1+0ynW3z2zXb@pvwRkVM{!LOzEO&UOqIU2*_UtvYe^Qb;IJM)j1U1yM6+
z44shbR-A2TkC%JbH^uFw)B*(F5b0ozSEnCZ$Tu{Q$%vlHRx1(p3<k>fD!)NKwhK9w
z<qAMVC3-n_kf{Fo$B&zAtppdwlryns0G^#ZvX6X7aH@;zn!mM53wBR&ch7r}eDk}@
z=PJB_53D^rm_JP^ZZB#ciN81Ne7!m!L<UI9Bd7z&V7_y`{|e)%Xo%7_k6SSIe+xXQ
zObq=^z71XX5ISke#InZb@%Pwj$vN^<144st7Xh;S4CVdOXGzEE**E5pm&ol~c3a$n
z<1A>D>%7J~f(*WT4r`SzLDsW+yqWf;prV6~cO~L)!^Q5vRF6!S1r>$;*$&r>?7g$c
z(+KRBTPh|f(KBQhViK8vCAQF&Jq&i?$!+YO^)}UQ#x*EI?n0S%+e+h`VlI|&LPhN`
z;6u}UIk2~GU86*a@Z)4&G;+qya<VKKX!~ga0v(dR06U?Y>EQNdPrKnqBy^XP#c;B_
z@}iqhPMHmH#F5f!xeO)8-Bxd;4TJucn+lp+^H_hU8K&*|N$!4SN>OWkvv`H~)12l)
z;Qjevn(_kDn<h_`i!{R@oGt}rM1_`vkSh(UseFOF1+ng%FK%L@)lYFFKc3iv63{=?
z`Wv5E^AePlGzaf=f)OZ~;AV6FKNcu&dm1{#WndUjjWc=(X@5pDG4gJl9;@^ObAip6
z*YQ_Ryl@F-m2agQwU4t2dIl0>2m%+|+n>sPfO6XUukBB1Gs%}o*_fqHDf-ewgfPC2
zbAB;x8skM<&aB`+3MTgcoc63lJMZ-OeB1cHDJuh{A<<$o<V`WJa0<)BH!4S-vX!aD
zp}@h0l|@2w{h6}O{NPKnZgQ0#N(|ZjY0ih<J$`4Qh1UzPO`U$BN($t*4|?ZHCYF%0
zUl}z{7dI%i^{LJtZYS^)=_CC-!qA|c>JwUo#MXDc_wxJniApGiwk~9hi4RdK>gmHQ
zneSepCNME-d`=c`_JW19+)f1=B`BikYigpzEg>Ho5Q9-b>+tKoUBCB@m$yqJY5vqg
zK*;YFR{NQAgK^313*o1YwU^s&jJvZn#oQ@GbRr?wbJmzq)vJqSc~vem4bi;HLBskS
zg6|32oj)8SSk_|&lHu#sM^l<5`JI~VXNS17NqCn@*}J=#@{Q4QX=YTb7d0C7Oip_b
zzK%Y{q$kH@))dzX=bMmxx>eC@8L4oNio+D2c9J&Rrr3tTAAXOJQ)d0B+>a!eCENYU
zhvIhwCQaBVdE*8@DvTL9&(z=iN<x@onB6`m?p8-V<<&(0^9@U-UrCWfpmFYqFxk#$
zGqwHdmLKSVXQSTZCukFNo2?#q8G>3tmHXBaXRqIVQkHSpj@7LQvAkRc*ZvNgC@{n6
z@iC1FZ+nZ2t)4)GrYN1M55ytmU)S_~3;o<DB6U$K&A!<ZJ#|rE<OWQ!CRs&5iQe4@
zXzOD!!$J=T!pnEzaXALD{r)qpUqgSi%A!-&{nZV;iEC-E)A1iKri^CsShk2UZXFV>
zBGbjKxwzgHVY|wMSDueyxGperh4(2}!tIB+6gk!v7e5yDla?3qijU{lbbrAl%n>a<
zaOC$?ThLhcKP+4#b{Y}N6ZT;U4hC=8lDV?2@e9D`yW_Kwzr>#o&L_#WO-$t<FtO$g
zkf#kQnlwj-_v?RW$FRzE$hNmzR=yH7x`eNVLW6?F`DrvtGkmc{EBw#0o8P9L1G3&h
zvhuUIGXvB<fW$pe{L`+fmM{u=d5Q2@Cl;>AS$lzB8bE9JI6G}EUt9#e#Z4x2r+n3G
z#Fy71Pi`CS9tk(^1r2AKt4#oXd~uyr{dq{-(<O6D=eUn*T4RwgmlV%R+VE2*K1*9O
zL^DN-EoKLi9}a%rPakq$VE{as9>@E3?r1%`6+4tv@H#LeAEyyoei0wsF|zQ<x#XXi
z3M^!!XXJZ`*-6mL%HrDlOb{`D!2C%S<Z0HEHT&V3STVPX&(;DyT_86ntR3oPbBWl@
z;8&bd%Eo@w7PuOMH}`K921m~`HbOANXlK1c93JBTAoB&WuMjIzu?cBE2zlQw1OqUK
zMm|x7gx?rbguGYb<+naBK{zZm%;~$|J)47$S+(JT_1-sSwToyk#oDJ}kj<ofFZykV
zueQ1OWX<btB|`Wm^rZc1r@Y8fHD2(4qKx&0TF&L+e53nsUHM>s2-uL7!<&oK#Ny@B
zio)k9z4;+g=l_mD7>}t>rQ`#KUYq?*QYTq-meiNBgDgHOja0I!>;6zesWmPui!tBN
z<$Os#?%1{Sh%3?#)ca<p4g?nO^0vYbFZM;a7AawnYRdIYcmh5twm=OH`@u)P%I9Px
zlGyybp1X3Df7d*{gB~d2G>G`1*gi^L)W{Hy5YuajClNM~S6OMc6^|$NySP<zyv6~;
zNr+A$zO5GcX%R7q$>)t#FgKh$o_TLBc3xm#37@R&hoe#P&y1_(Z|^;!UDpj!#nc`~
zn(5;aQcV;z?Eh8#V|uabbg2Y^XaA3}U2>-*mvuk(h)2NiLd~GVC8~88I)!!$T~qbs
zw-e@mF-t5}C@1|41zs^ZgY*Yb?W{2%V2%yFJ^yww7%hcXK6R!*TCwGy514dD#CtmC
z)hV?1@GJ<JeRif*9^hd{_}Ulan?Mt@&%^!zvn823-4T=@Uqa5YHMC8UZB{DaNyL`Q
z;1UT@sEm>`YU~ogFa2mPGKumb_S^~d1y6pb1q0D&IpfQq-!&w;!#|xXHzlV`maq!m
zU|LRlreBNH!fzoCc<Nq@U{bLdCmD`yc2gCqI}~G>`uyULv>t(6L3k7gviH%=>*nwp
z92<@I##rK+v~C}B=Mu54QxWv4&2I*`X#M@56av;-+}@ty0<rjRn1$H;ex=#xd03As
z;pl$yJ1LXy<z<7N1PgEm*8n=#Ks;nDRp|TZVQv3!5)=B6=>?+Ouf74943eZ@B7Mz$
z9OVXE7oRtsrL+jUgJ*-mCTRR`De3*TvUNnlLj#+)mKB+pbhY;1Tn91oFJ(Uv`oF&!
zqJ-}Ho(5&i&M`NzeRWSbEGWxke3e=1^y10&Am~?a7tb^m{PE3AiPi~v^+GaBjj;pN
z+sr++ML;gM%te`|q9bjNUBuG6xPy7YTUvTe&35dTnMQ$<F8RwP-jz~qY^67ju?Cc9
z4Oe#pNyZLmCS$7HT!eNq{<lxiQBfutgnw1ZR<bX)!5R}VFZ$0Z<M~_T^4^$gd|a$T
z3j59KiibJr($n1<3&V$2+ZTbAjm`kXo;N=C=NmpxdwPkb7(zV<rl%jGF~5&-<1@91
zyblomzo9W<=Nrr0Z&xQ5PcTfCkA1rjSMNsL&b*ANx0Q}hDJq%<=4K@V9H#XEQ8IW~
zv8aPy0Gbx57igPXtNQnD9|d*W@#xn46K{ZMXt05tt3=#IeKa5JuXT-x+1|;jpVFEw
zoMnsp_)kA}9Px!@B>1H-c$^f*E??w5advv<?{|@{-h*2J7AZKk1GDVysAb)lUIR`o
z__l0jAHu1CMGcqjyThYk%!2LiZyd3VzUG1}hn(Icn1bXO0&o+@QYakD$Pzlb5?L7Z
z6U=YECeI6-xEB*|!bj%TX!0zV-3HrB9~@}Alo|jqCx*7w+O%$-m>o$2QLJjI-eWHP
z3#Ale>i8VOs}H>5m=1?^&r4mExChKKDWnTuXNMzXKD&*DRU@E#huPjSWt4d#EFE`G
zYtodJn9n}jr2Z6)S=F63D3T<mC#a|>IU2MJFSTz!rrIaeOrN8E)N`=sBRc64mT+$%
z91x)(m(qlj0`e)7*3g=(mpup$uRIgHKk$=%$5K?>l-Ykr^H1~CT5cinvx!|&St6nq
zUt|puGG5jM0dL0DvZbQh#!;G@L{Y9Ba|W()wv3M<ojB|^4!12d--1+3HCs(Lb;x5l
zrXj9^Q5E2``R%vg>j)I_{itFknHK%QR^qA|iL3~VShYy{2Y%FvK6>z1x@?Uek}X==
zrvnN0E2J+eIId?w4PP15y`HbVzNMFR*$ho9qN>;5R)e+5!{B}(9T0v}Pj{fLe$~U|
z7B54b%auN<tEvi|C+;8iv|M4`K;9{Tq0rM<8b63@Gq+m#JDO4;pS0Mn9)JA9`<WK_
z-GR0EWU;pS^bk~oXquY6g>65`n{uXz11|7n>&fpqcr9G$3j9Ynb|D_mqDsNCRG(f=
zwIy7oXhhu&80B5K=LYGe-3=*!(hW7*e1x+T`CC?q0Ov6<LNz2#)cHjWHdZiTqj2Se
z;N)4QC@g|*n<p^&9CDpBYNO;g<m@$q5yV92_%APmU&j~ZwH#KraN33Cc1*hGY(!@L
zZ+o#nK);0p+d{}#^-o2juH@d(H=g{a-<VR%sWUhE?AL<g%VV968;J-xMt3z6lV63O
zE)?%h*9vqh)_y;i*f*Z7wAR~<r~5c&26(r*am2~7W$fgm5f-#R<WEkF)WmeZKAiG1
zi_AcvBchrQvWWGoXVI69q3Pg)zFIm($gDs9bST!&h-XAK(QoJ(R@&dk!Yao9rCQKb
zW?wx>@6n==u=K%$>O4Mb`~6qqZYnJj(kHp^inFKDpi_Qull6Tm#HxW;ndz-$lNpWQ
ziKfIpSf=uM@%??xx35^V|L}R<)0GS@Sz2>1vCcKEeBjdodr7$NFZC?z7qIrwW+{kX
z-h~}59?#KBdyb7(&V!bf(|Wk}?VrU>g#K<#O8!P8EfF&2N}w{VD46?j)XXUW87zJH
zv6X}-RdRgmPl+DAMUL^6OBq%MOvXg-19XcOXw2CqOwixnCE!d3?qrnxPZe5}EBNP@
z%_+V|26kqr^N9^+zY;%@GKRPj^B3+$E=v#-#8V7)2etP8l_wnEALC1JRffmsQmotx
zI&{x`LSzx&e<#s`+tb(gFgsfH7OJ)r%1#=%L|#l*&Uo2WNQc*fXrso@1~gRy%{ej>
z#LA;)Le2@>Oh1*8cKQcvQ<)WGIvcE(eiwnPv$^J#I}%R)br9S>o~C@$N}7}^kir6~
zauakja}Oydy(alV?GU7T45ymTzcv9;>S8}gnzb}qM5fz>8d^;n`9dgro@Y)h!!Lns
z4`0<KZNhby2~#&uG!HAFVfRy8(y}MU%<l|&c^6rmA&E6+LoFo+y{Cf>i(q{;=q&oh
z9HEc_Pji+Cs`iVc2Th7Jwj%{6%SuMw_lq82sRin^g1Zxi8zYHc<|9RJX9A-xyIGe6
z^@qn48##?;<CDC@Fi?WBmvT<J`f;^&DGB!V;cqF}SUqoG_^=WryPVFm`J66qjHBlW
zx?v*t*gbVUpxhlVXt~@ZdrMurW((55WUin`a(gMm@TU*yUvQK)Q@>A?gQao>R&y2=
zJsigzMd%$@#JqQdwo}CK`IEK>yozvbLUPV)4cO1G11}&yTNdfyGO>&zjn1Yb;?apY
zdDCSquzSCI*?*L>ZQT6~Chj2fL<*57#tdD{b&_I|HW5dv!4OLIOuL1hI(>BS?o7DC
zY|GeMg*k=+$F+=I<&tvB&0TO_sv=FG9)j?B^n<4|McnTmY^cvHHGf_$VM}wc6mj4`
ze>KZA_?~(Tp}zuoH$5khC-2+o6H;F0v!i4(NCZibmd+Fqc@Vr*M_(?$XpB$-6Exm3
zDW&5fxp^tI)4O?&{FPtVa}y}$SIg0W`-2-3iIU%Yu8?>2o%Mg{UD-YdQwP=USLU1J
z8+cGnPvvo}uwYp&&XXeEf--X*NIq9w+dFs7EpJSyqbOUKQe)s`8c+zzh<!cFCiR=#
zTQ7pJD8a6t*E!oiUE7TR+RSU?DY4VEQvJj^<TzTesy~G1nJy^;8Lh=ngl;o-S+!Xf
zwKERl<v%-WtnoZ7wH78{_^fc}_=!;i&nv#p3cXb!M<i;@vbGt5F%*Kh(gw9Y(+X{^
z{i}_$vN<;X_^V06NQ4r=t_nZ;{9x>`1o@*cCJ5;C3?EjSYQ?#=hkrccJuE@KDbP}7
zfQKUiPNBU22oFyoCp6`^Khj|vL_<q3<xjx^oaFF^OihW*8FP@#`3@D&@ab+e*-a^9
z2hA|76^Gyr*~|$LK(#Nd1Pa7qS}^X6Z%mE{FG2ECV<OuNFdbPf3Hvw<rS#m9@U9<7
zl6oq+mHk}Jm|RlcL6@8OS`#F;N^`HHDY9Vt^rfc%H?iQ~=0E)dd%^!;4LYM63VMpD
z7>`C*+Cp;d+}nrKQkRephphDMlslJS9KSP0Wme10;#mOYkUaR<{;u<S9+kDm>9hG{
zfhj-q{LL6D2RPqVi73X6`6Ihb@`&mMC&@)K)EFu_2~v+jV(mvSAoK}%jRyXeynJth
z{vz|*u@V8xKP`j8|AwFue2x$_rYVFo@PgWpqQql)T3>&G4bIPhy}$FZ+53Lfk+Ig~
zMM&ERylwXxuMY|OWJ-3M+f&q}zdgNqyF-OG8~4`I`7F90!-2QFcK-g*E{W*h0}TfS
z8`ERF*$ulkz5;)^qhekfLrr?SQaJS09AR=S@r>r>gIHEj4Dq+4j4`LGd<%-b?}M@I
zK4ClMvDOYWC7t-_AD&e{2HFp}&#z-Ev7sN_NwBBM<ZD#b{H`ZcGu=VpW$h&>`mz4<
z5?K&C=!Ze{BOpPJpmGBvI0W4-|4;5UNBW2sw;mLxsB}{be%QcBm^UI)Hgp$q%NKPa
zsD#MK3EgkA@JjNh(;>G_Yz46Bzax+QV~O+hFn{*ET(S6gofF1j{)=l^_Ia7kczK*_
zpZa}0O30(<b+BWlRrm6I=2Nf)e_Og#k3U1bFoDTgo3(H((P&Z5f^;A+`CFDRum)3H
zi4iTk{E*tmFeCoxKI^kR{xBK5n1PBQ36wd*f>kxYS%E@}Lr`V_6?Pj8y+xJ~)5M2C
z0(0HPPgKO$TkepiSwp&Q4s*^#X5A)s=j(mID-?<%*qR6$?L|7xA<i4?BLy2*_a64-
zRVbyeqmU!}Mq<Fbr!v>~?&9!ge6%}$=lgs$cza+1!wX-ad%|2If0jySQi@U5Vm#WU
z*4OTN<NlqW9L674hMn+)!s7NI<=q|Lmh?dC{C>1?L)L<}ZrgL3##ySMmQfY&vdqll
z$xQ~Y1{(;XI2&;fNqW*dvEiqB2{Co~EClygKU-5qPHSt&$XfI?+wmRKymb|?lAw$T
z5l#Sk6W89r<+7N(WmqqIv1=QY4wQ|kCmaMJ4ME<F`MqJStjtD>&YMfDQzWdfh5kZi
zx;DUOvsY6z{djX2plk)`r*mKvjf?vRQUrd)Q1Mf7iDy>8Uk~>>8eLPldVH?muSrP>
z2l$16_Y;IhD|5D&Go*lnZ|6me580$_pP0#hW&y3tKC>!lrn&1@P;Wxd3wcB_iLsU1
zh&H4FL4>p3a<$23&0CL5!ezgBEo|l9l_H1&g%LY3T|@u-=SB(kmDwstPnk+AGD>jH
zf~TnOL0*DiViSKgutp=WAPs3_sq&GFY}0UPgeVgTJ1<XC5?xj(Q2_yfVLXjxO;gBc
z^$CQm-JDIoC~WW?{Xaf>=p#(wwrugd_$om1`fqLNVcz@Y!DLhO2^Q<kN`oCG_V3i<
zLiJ|%o1?e1w?~DDW<KJGsGb&woPR$3wX7mTl)B-78S@F^kV^kSkVX78Zc9Mio$Ps~
zkk{#PS}}glo=N*FtUFE}R<jmzJG%R-r~vX~eGMIcMcma^yd$@?r?6*%c@|s^_G~7q
zPMJS(uL$c`yTh7ACp!-ySmaL6y2IpYu+?b2dGjM6xVcbHFbXCU`VQN;+UYd4G0Qgg
z#|73a{$Wk0*`8Wr|63)hp16A7CPxx2u$mW;s<9HSJUrlC?T%0@VO5P;GAkltPA>KL
zu{k`v6U(ALAM@raco}bA_IViA6MDYi0>s65z0IqMP+YUi*<cG*aMl75swC9GAquay
zwZFFnuDe1ta#mSK(K}i-_!RFL<kHwqQ$B|}`_88aA-I0{G<7|ZXhGGGFq{c7=CFi3
zT0&9zQe*iDnYU=n4O}g#f=afdpBz8?9~4R7xWlTI^RRo~Z2vIm)>p|ZD)3lL!^RU7
zz3UWlp+@yP2zb-?UZ4Jal^^oywkeBV2u=yUX9)*cy3`phg9vAMgxk6<+`%kGc^aiF
z=a+HjrrKDC)nt-ZS(9z0+8s&6-HD#v;U9aFoTSge9)7v<|CYpds04LHi}};6?GzeJ
z-A0YEs~OCB61lAxr>Z^fzKlosE*h7UzNXLXTvo;u0Q~t~zx#m7$BXUU_UXK^H_Mhz
zo8>W79&iVG>`Yd!P1y;m${tp`zcf?{g~7m}2@~2o3gcJ#Y~FbN_t0XC&*h|+I4k0T
zK0-85owvDdlH0L!<^RNpyGOaDd5LEAszl53re)hz*RCIav*fq(mHM2auX%L*pJ`F?
zXo$8zWk!G{V*c)%nth0VF)e{dD#CJ=GrB6D*1IoK8=GG}-{n~^LhlW|<Wj5cv#-8y
z5y{o+Td{~vnng{XO}+k6!0uiF0v3@$rPV7NTnOk~U}=-s|5AOyE1ken!@gu#WnbC3
z?p@hjaz$y}H-!gt-8}IOMbD4C{yCzT8$FZNT9*t*1B4<Z&z5lgxTv>YP*wek9Qjip
zE7xBu3G@*Y5TJs<5#3t*SuN!PG-823gvdMKGO)f*1S*f~{RFv2bldao5}^z%(V0r+
z+4K2^cn|fpvTT-;Ly<8swk=c}qq#o4aZmSC6ygGz5ujXYsr^;tdyXFhYs}L4F7lAp
zmN)req&{!5zX0Fbf-=L`jy8th7OWAWBdMdfd5T})p+h(#u4xe%d1Is{w2R3Rrjp+C
zQN1dFT_VnRiLDP`w|^;$YM_-b8Xo<<1a~&Rjl795sD2>PTXz?KOGDdFZ#^U=hccuR
zFwNk*u!sEg^Ujq$yql=V@^`Q^M?&V0{<sJjMDZl@ZKZHblG(s&JN(e{4?Xo4D*Gx#
zOIb3g-mjEweZnE@1VL4UIng&~pawd_ZtVpl{JH(az8*(2{rGi$q&|N<>DgM`>zPvc
z4glWZtWy<W$Gf90@X+t`Z+Z>%oGALdxB&KVU~ba`=Uo`R;etM=1N0$)i=24K=AX$w
z-RcWM$hvOs(?S#WKd;<~DhleTer@u&6LJr#VRANWIIweX;Ds(LP*t_HxEFKJTu&aJ
z_$15>^%T?Wd=OyO9HeRC${o4>*MC|ZuAk<s1}%!X<HE@swy6QGeT)^$TK-h)g{`}L
zza5mrkydX!x>kO2zoH<ajw?zQ%}w1fGl@%M_n2&VAyt9)(bMHtkG0N#kfdQd3_+Qs
zg$;i(K5f;}o)7z9K9+1=U?OGiGgnGQA5hS=01lF2T!$>o&f`hf$M$L74#@f$o-L_0
zCHOjb^(*e#!rS&z)x_SUBiuaa8HhU33F!RM)-{uP>ykO0oX*#Y;F3v_GG^0$S(G0=
z-#S1Eq~ckhoy~qf<d-XXr51Wk6+BIy!sJ}i9n^Ty6N=6L`Bg0$)hwH;2l#s+i>=Vu
z9Mmi7urE58Ud{6_1FhgubG)CJtY1x)V%_TwMt=qdkHq;O$KA0~0_2<TzTQgP%4e>R
z)z=gmWi;f-<h(!sRguRi$U?$EBD1STpiu)qvg5`nM(p}!hwhI8l?;^OJ*PNj;9jQ?
zs*8g4_m$ZrAfsHscds)Q)M8aHe^NSSkQUg601AW9JT1w&779AwTv@hg<I_yvRDw(a
zbd^HA<1ZtP(>O0%l(BEvGaU(fBy<cPh=B-*p74h|utu-VGq0+OJ_(_SYZ)m~UAHd+
zl{{GXHkIEc%hTc^JBaib=|2j}{pXPm(X+PaF}w^P@13DIpzoiqz-F(J`aB%~o<kTc
zVP`MTAao;u_kP8rl*uX_4K2fF4dj6R3qb5Zh<G{xOd^0>HCs$C#?qPr?uXTm7ef~9
zbZGK66KmyOYO1m2ojRFyrFsZDL6yZF$wxTfwh>PozVxBOPwE?CG}9W7DLEP00`b1~
zxPCmcLd2H2AMLly4T6BY;*e(zF1&?mse!Esce4Zd9FRd`$`1RO=_SVW$*LogQbB?}
z(4q(XO`n!ReweQes)`oh8<|I=E3NuiP>{8=AQvBV+NTyM$L~MHAU*lZwB|(&p>S@2
zjy|8GW<kdeS$s@nXk!gtUvNY+n+==bUpMKiL@_JLLfpSK;q%wJ;Z~QsbZ{drGN`=_
zBI=l$p#@R^^2_ULv|a*n6Z}3v_itS!k+teJVb#=*F0Lk*p^0Iv?A?upux!shi$(~O
zHMVj$4-*2nhShR~Muh?H*{j?;G(|;i(lSKsKcD>3G962Q6*5wJck{~%w$+e0nU<{d
z#_l+{l&y+{E-CO3{8D0&t+%!EE99TnI|=J|{0?<EqVYI!a*<N1P;iW}?jFK(8Y>uh
zhvL#Vq<$^JrKJ>CN<{HuKlLmsFj6+W(a-8#)(m5QTeWnAcnQC7(tWq@{I4PHpych6
zb;3x~v(>e8=&Ck56SpP0?K$9&XG{Ldr3vgzWaUVhFCExtwA(&}588MFpyCk$_H@n2
zN6UHI)ok~0NSyfzx^mg?-)K;tBiGgPjHQtCQ_si&#}VPu&P3d>z9y<v*#GuAa_%hQ
z*FPM%a(c4Sts`6P$(^%zj{*LPD8z@^CU55p71Fy;p>IxBef#ks@qPvFfrst_kxvS`
zt*h#7PXE9(_U(@xpmKu5nzBKC73-F%1EPgdhH+%gio2AtLu{z-EA66YO{Y_s>VGQD
zaKW~3gB{3ZxAID7f~QY@<)q66&?V2`b#zxFcXGIEE+Uxe{$zW9n@(>g<&HC_um{DR
zF&eOXbGia#hn4=!(Z(EGN{505tTz?af!!liq0ivlig!^`OZR1So)$ubLqrpkWSK?D
zKO~#*rmrqTjKXSAg(iUo{5}OWDY=60_*4cR>+((*5{~&C?~zWl4vEr->$&lhf$#bi
zj04LRJ*_(4Dq@f0)9ZWOxmk6UfS3aTms@T2sC$q#bq4ZBuEy(6Ghe+Qb`k)s_<H@e
zLhgAZ!pCv=5-8;FnACyiMWzX-1%x=|NVl?@+<K&NTYvPpO+CKYxngp=T>iK$HVVIY
zc><&M(Q7U(!w1^zsgfnSA3My9gR3y^8^;g3szpm7mQaXjVVM04*Qeou8sLe}yRl%?
zrhu~NCLftnL?HOJ`~R})wdfdpjc$ncki9H~KgHY@-UFkY^yjiQ(?g8UJ!DE98v+wE
zqM_}VjbJ*ys|Au?K{m#30-;nXgs4xf2!TZGd_G7a7Xu?v1J2}WuoRr^?}o&&&&?;9
z=Dn}Ni^kEsZa4E3o>ZvbnV6r6N2yzAVcmJtr?t9W?fFcl{L>S90uLZq_`7ZCo><&_
zc&t}dzN1iTz1v~_yspBrDhZ5GyOEIXJvcF}`hF7^8Ep)`mN47`Ye+pEgOMqB7)Ppe
z0zyH(>?Xr=BrPnQ=3#7aIKtb@2~N`_xJ%1zG51tC`Ju|{xV_=}6I;`V*N@W&>vnFn
z4oc^6j*#J-%s1!L6M;;TA$^{-pZK|xlSj#Rr~_ibp7`Q1psN4O*4&et(*kqgTgyih
z^ixYt4JWa+=AI~Yq8Fids(%&1F|t$oxz&ql9~o9Js+zaG7EZ9=?vi036D|!2gBgd{
zVI{4At}yzD0dw$2uc~e4A5LUR4@u%>?W5Oh??=fRPB08q!^)a!kzhLPxROdHLcj!r
z3h!Bao;Z4(G&bsqYbB<xS2!9VB*R_NHge&+C~#&!5cM4a)B)8NpMUS%ACbFugc0?X
zhK>a<+9HxL3yn-?uhmh_yu4Je-L6Zv95X5x9A>N*T4pV4x0qC<5V?}=w}QSlOE!Lv
z^cutK87vzbDt_^L6aeE>^n~10<q2$AD&$ft?M>27I#1-uIy|O0)J~VIc$m$2(dbHZ
z1mnliT@*tXF)Lo>%p;6oL}9@SkaW8hEbmW#kAt&+*$_1#5~yUg;ZDisWdsi?mSa|g
zL7og`j{CqK|GpAZJh3RXOMdbZO+R;dxBB577mFk-#uoyRGHM0#&5qvcy>&EJzzT90
z@e;YL@91}~G6WI&#-oW`AoCQATZf`==otV&i6^YvbPSfmMXNfSmwzc>?OO3UUq{$<
z?i0DruwHKPd41slknoMGf?BF(({{Jr_r6<t2Fi5Lh$-vt5Ht{w(btETqL5lxvG@S7
z95KHz;7kWXznKt64%oGF)jA=irBCZq(A_>c1X0)`^FT?I;Rc(cw4-Wi;GvaEwFL=S
zvA_Agtt3McoI94Jyew|JJIyD-a%N%99Zd&2&4+Ltay%}>OOW&YRc>wc#&Xay%Cv<H
zXAPY~2euEoPuNRgP|#)UUy16U{F3Ftk@=2;Q@MC$ci53zGlc^s_8>Y!S9HFu!Vro{
zu)jH7DD^Nh!gW~DZd1-+OqUCxWSf93=tAC*Ic{%%*wx#fezr-xeQg<jW^oU7IbC)7
z?0H|-k%P(0s!IxN$yJU6Cl?E;*qtUY@b&JG=$?zl9IaN_R%{;<W;d|aQh%no$@|-d
zjTXI&>j+aR5iYP1b>R6Ut-N>|-S~s-QH)HJW<>Gtj4mNJ#1vekEOF@rzt37zQTB6Z
z<Q78EEl_DIK#m;a{No*>^dQKVWTlYMe}bkj$r&+sW-IDeJtj%vZ={24@(qvquXZE-
zIu1mLxLF+AhxXt_AtB^pk&`P$6ha7Nu--cvbOeCEaNVf%7;wT8wsJn9cnGHX-g{8(
zaqIbm>--oU526j*$y3ns01_Pj{?U+TF6Gp&lE#pS!bhrn=hP}P6%Hyn$y7U{0}X`O
z;c&XH?G9qdX{zOY@jH|EgnSB-*u|mi1_hFX`$?(V$CJ~5SF~(q0-raBwq}X?`G>oJ
zlD$VjXT=MTzjoPzWz1|NpIv<z2=Ol0yaJ|s-tgd^Q}kou_1NW+-A`ATbj}>bsE~>U
zrW$*mpsIcj!+8gfyW{B=xM}jZF#|HTj_4wL(fpK4%=5fM!<Jkh_N2ibt%H`g@vqdx
zO%TZCgDq$J*mhI#TlD4=^bU^3%~^{%Jm&k3gG(}#V8yt=UcVj{Er=Lh;|T74(=#Oh
z`C`iJdLD2j;2~mNbq}8Kn%ZwlG||G?go?O9>?YCk^)ueKXId^N^5Qe3RW{qqo16Iy
z;Mc|Y(H-pueta=h^k<TxDGs_FeO~(E)ggy(TJ@GGAj%6QRw1pzHBXGAyBXj9sR!BY
zD>L)g1>p375sG?W0G=pt*awFHdh4||V4w*n<N+Uu--ML^owZu1F&jFHv8xrf>k1JN
zv|XIxQ3G{G`{z;l#^}EAEc}yMe|~>1(!zK7y-Y9%9641FViZZv4Up#U-ji-l>~NPy
z{=Vxh46F}cQ9KL-eOrUG)_z6I3MQBCZMadUp4$%N{|C1Nl;t4k!wKp-{9^TVB^+k#
zhksOov1aPe<Y}W(HktMa+9P7#-yMRoLjshK{M8XEGK;PT_mCy^eAOBK0s_qMui&71
zmu%kl<CW>yKqbG4T`Itfh}xq|T1zKEEL5;d&8|H;YL+>D(c4~oe&szmW9Hf05I1w7
z6LY4oA*Z{@yCURP5j(sJ@tr3!6MYj|0{2xx4`~VDck4pr8y3D<_OE1r*xMJ0?(jUM
zJJNJh<qgZS2Mt!|E&s;OPF_E9UY&Qm<S{2H{QVQs*K(@a4HPru@nm@^JG5qH!zWH_
zp1hQyll*D6X=P485Q}oFOaCOPYHoqx`2=-kBe;jMU~(q^QNUwG??_l9ja?R6w4hGo
z+CXU7>m9BImu>eUYaBeIl;7cldv`Z=fAdQe_$EzSN)y0uh_G(j4t^A56TOyid><=t
z<}buesc;K|&J_i3@5jbnphpwS({}y=Wn7-H9tem*&kmX?B8^b_y%4AUew_8nk!P5D
zqMuXZ%W}scV73Gl{S{9mdBQd2fa2G;NGzVvqJgoEN3S8RXSUVZ1Dnsv#);&#y}rJ_
zWM6QgB}Kl(62arZii$+0a=^ffqqDOf>=HeKzg7H+(aDN3*b&Sgx3!|=q9Uyoos|R~
zHLtUWi<9)SvX9@0trgQ8y!TuN$X~l<nkM}xE$jKSUIc8daH`Tk$HBtJ^?91FoR8<e
ztw*Kun{Q(%7g1Oe0lnrdLm~FJPrY~X<*wG}N-cJ0tBu1oAEwcXc&Sn-WX=jJQf2at
zx^t(}JJjPXsQ6V<C5nm3<Z)nXnlZ0KFMnR8?sJlaxvoGaQY1JiMIt|24nQ0(AR(Xn
zuD=LE>bfdj*%NGXa|AyV=p(MR)qXn+s91YYkrH@(s04Y~(b{W{s&9SxbGYMzIc<n=
zAqr*!J>Rh;ojZkehi$Q<_M+0OW=yvku`HeU0x!1K+>Wd?a%54s+*P3sP-4FyFmlrm
zgi~J`k~0<UN`4H?v18L)uO%<2h?L2*Fh9Hn**=~JRTo>jhA06pXKZ#)!-T@m*Px7f
zP$IjisHib-2OL>=30rDwaTBOL09gTq@FR;OXUnDAT~LFu+Kj>0J}mp0^k~+k>;6J^
zO{d9Ye6{4;Y{NN;3Kj#Z{%3h@<Dz@aR!gfN2w5_g3(dnIa<<lL34LKq?mI=?C`8gs
zJ@G(_RE)s+V@{>79($6(u^V+2=STq=p&NQ2#kaZQX$>}_@N%%1*UM-_4{1^)$A({I
zogF8$kE}eHsJZwhwt3Qs8zcj5uhk>4jTI>_B||~yp_BGos;huq!?SW<C+=>7dfM!m
z5$QVxr@3<<87)WGXDVY)8d~wx1tp{p{XO>Ahyv+yD^2i3h_4CHcS{_N@a4XIwBBvB
zQMgx-0x^_#DdOC{cn*|cXrsb6GO9i&W&J>HT=H45{{zJ7b~&%8T)uP~fB)szh~ne7
zaEXT}dcMMI-8x+x_T)dtI0!ZTg3R^nw~qlEiBN<}GN17xp@V;V!zSSNUBKK?Vp<>y
zn40{$;#l!t+qx0`WLrLdf(0^xmbSK@u3WuD829|htyv(h?xA9O6DZ2vzdDt0@clZ7
zg5lMpF@_>Mf_&A^#qGxQXErWDS!4+@BTo)~aTJHktm!1qp-OA3dZ~H3!0GOU>K!4}
zq4}3T+(C;28&JY$%5G=_&l^;ON2oM}RAJ&}yfi8<4aftpKd_~EUC(7ebj@n|1^P6z
zM<_}7Ux-aua3OnAP_X*ZHO;mnim!MllU{~)4k7%Jbf-h<Fv7e*PYD>sOJq7^jjHw!
zDi?seqB%ZJ>lhG(pM0n{gzM|^N2RJq89Gk7I`_ImfTL`|qGNhoMW@`z-5xy)E{j_a
z`wL9b7~LxRdA5kmmhBi~lBqFV*l1@5=>!Yb`K@tD+JPK;68mCq8cPG#2eRq@Hv*`l
z=ndVwj=%<rwevBcckAX`sz56!Mt@Nh40kyJw!hSyu|Gd!3Sy7ZDn518e7}o=XysEu
zdQ!3N%BC=<#?&iviBTG`xhfA2sv|5v<yRZhGntQE=bOOg$YSU85iiOcB_+N;?I3^5
zFCe=-&l-<Qh%Iw{g$=A@T3n_s8QLNI$>b^D$BsuRo6EV%2{S^IAP#4nzfF*B=cY{l
z{x^w$ZDC8UBX~3#KrxF93V!DO*aVZhl%nXQFZpv!(($Q|2k?J6ZT5VaDr6fbL%mHz
zmG-sF<B1uxzdxUh5PG`5DVheomMTyZEDWT4miO5pX+YS1h89WB=lKR?ZrE9!$2YsM
z;XR%#xoB!>5niGh$C!<m=~SDH5nZE-JPfp3@1#ki`@E##+m!3Q9vpD&EnEAwr9z~(
zc}poSsKi-b2a{`Co9^b@ZAeQcRQu>KJn1Rlb;2Zy{6JW2XI;1@a{9)KGV{ocTJP@V
zcTlc)wotni$DnrcwQ}Pp=xnK)lO)*q>lYw$vG>cPFB8dG+V>-E!gZ9K2}FEHJs{Br
zj<cbPcI*g4h{tuH8t^GisDrbEj`31qw)Y#AwEnuf{j+%-QyCb!?{kePt|ncF^f^aU
zFVLlAB@KVPw^a{CeG!ft!>swjiGG9fdd9H)Ovup~jv&4_r9j&9`Z8`}^ZLavI~6WG
z+&I+T?`9u`cbO!u(~wgGmqQ?H%vp4od|@I}LQj_@=PO)8-Rk+%guxDV9;w$oTWNUw
z148vzS0Y9e)(<+JN<AkR(3Z1^f9;#Y32*{^|5qYB!MBIDU_K5=L&CeQ_xi4B#wE&J
zT{+atxxF+qT@HPI`$wq80fWYsABoWG)u6=q>(`}cF7!vXeEdllPDnDQnHmIzW&=%9
z?f+2Unt~8m7~1F1JoXc~sB~(kNb3&x`Iz+L7MI)tEv;F*LR$ns(+Y7%q^@wJrmW(m
zB}$o6m<(aZ;>$nrni>c(lH8e}Rb@~*dW`1?z9FpgTKfs6z(TDyB*zX~G7PP9YlNx4
zp{Fc}dp#a$M3Dt$Kde5`zLkQy9fKZT2r)*0w+4!#pz|@6mv|`Z<RS&rQnsZ^%&hal
z=1#x+okq1WQ<BOzR;{$f47|TID58jP4HylNao#SPLcAm9R_J*BWl|s3I;wPio#ybv
zV9H|f;WLnzS9UJm2*ods?pzE$nucm>h$b_5?@vr7osD?5w0cg)6n4B{?T+$#&wc{A
zggrq2s$s%P{LvE*(OhXVO=|2m3jg=0O8%Ey%&+FXfR1m=VlIK>*?bgHm#TEcFJ@iX
zm~{O-oR>G19V@KOFAl1ghh~cxI85--@}I&}I7!kSm2fF4{SOx!lI#Cmi51hQD~&LU
zbRL4Fae0eSYGooq*t?!OJApxhRa4!IfywVg<(Nq*6sxft;n?j)*czXFG!ywvtld5x
zTpA7AQmUYgJI^aPDr(?MA6H(8NB0?dOoc1d->K{z$fwkeWboU}Tv5K4y6%KU7)t|r
zdGVzPX)kx}|DG{~Icw|cHisbZ!$530I7bnN^<F+i<(x6;+;;6xawfOR7F3bLBFg)C
zJs!?l{3aDPVAZt!={M7s?KZrR3!0^TSW)5Thgy8U%JJ5nnkq;B{RENdMo+|1`JlAI
z;g1-ca}nG}sHsF{Te%?MR-{d0&{JUm!t%)eyn|4?c7m46r=YOhC?pYYKPA`+d3-j}
zN0aIfo&Gnf-)(+#qAU9J*ds4+T1T+@jJ1HuBf{`^nIZ-D1%oh9s4R2LyIDk(&xGXS
zwnBV-zo@b2QGzpxj&s~R`hHzTPpQ$i`NJ-{;@}W6T-%XFQ2Ejf5Jp>=_WaJx7~lV=
zXZ{kNv>ZpPcyqc+i52lGcef960%HgwRR@w#;R3!y%&{<$guxP2HNT4;Es-%FtslB4
zyaT46wue5)yia*#kbpt++U8Vimi#5Oz%JfnRNNLpp^#~-O3_NXKNCB*pKV|S6Qmb@
z_%`!tlbCrX{)OtDnrZOSwkEkC{M*+q+YpiIH!)!7a1pb*q*5gLf;H6I9Xdsh7d*P*
z*wnJ#<(s`Q5QK&F088$}x_$Y0Ob3<c*!+v2VR^|qi2I%g6VbGqn@#B*!`^H8$_@<+
zMmud_SmQfl)}Q=>Bt0MAqxQ#AG>tGO_Jo<Ke3a1GC&{apYevU4F(%*(@fU_vE(-R*
z)YI*Htp_5d<b>XHkHznejHMc$?<EGIAhxNNYf{1=Cyq+Se(iU_HL~fZrX2XF@V0_a
z+h2J53%BnD&b#Iq<(2S8h-aw&YD+4CjGE~Fd;8Y|0-dUlMufzt6hqZo#0uK`QQd7H
z_))B9PSJf+$ylmdN4Y!kww3(o<V~D-;|>Y_&K$TLFAF*i&7|ur{q(@V=iSf25H2g7
zo%}~tg`tdFHa8;(@>9WLP_9#+&rdt0r1Q&WC5ygI6OAfcfIxAFz<NUKoBijoOioNA
zqbst^P@lwF3_gxcU)HggZ^^M7L>WJ6z9yo+ISY7OGy<v`$I=DMKSpl&=l=^R4HKz$
zq@?8dBQv7mIiD~66~Yfoy<98yTsa<;2&cIl$4nD5OgjxFwkt7~m*c;&AB2)HqRPLC
zZaS0rsdN8F-`hrn8;YwDI?u!*F`{n;JRnH%|5UlH{?=>tau`&%jC4W0G+wh6T8s7u
zduoy}Vb}A_qBp<J{|CSBx5TT$UT&Mu?6dEC;~0+S$~=s^dURJ7y%*u;FS`;eL8Bc=
zLV2w%J_3B{5B*dP6B-mHf91FWP=P80uWhKJlyJaO8ZGIvpIL5$$X9?_g1;iI+K2uJ
zP6Z`0Yc3wWv$i4fVNRc2gIE2>&z~`1^@!0+lQUVybciwy#@YWflt7<ggXutD3$>0-
zCmnoU?fbQ7hn79L^oJu23$%BPq?F;OV3pJcs8DLTk8jD-u(k$Okg&3%q>l5me0u$b
z8$f4ea<couXw;?kHUYstiw|Se;~j=@<lGGp(^H?^>EFM9%Rx(uNz+z6n4=_H-9Qn+
ztJkb!@CjO_iW5!$ek-~o;r<Y7=4qj9{)G@QS<`4zX~OR%g3wyppn5cQtixu<EF14D
zm0~1=JW+5bHu14(!sOX@*V3KK%n`&<m;Zzg_V7!-=yZ9h@D(a%w4T5GBV}cyTSzO+
z53dToc0xKQfZW$RAyh9MRx~d&2eLpxd6MBEOMcN_L9EUE@<6vxdy(-0Kbs@*0q!QZ
zz!{zUL@>)cy$)s_HFG5p*<)kdw(z*u*Xf4lfW==MiBhD~)_NfMRF5&EZk%>YuX<hy
zRRHP~H@PVTp=3ISeh|GN<qAed`hy$8SbBSdx0lv$9(EY=ht12l5^+SdbPkWPn`HMQ
z`n-4Y?V-0B0#HOFm+Q6?gwXRgA={JcC)T%i%E(4|#&sWmQc!{D=NEmt_dn?yTOlTX
zsBSE=h-x28W=>Jy<W>12Uo8-VGar%I%j2hTa*5}#@o>z4MOpt^cl4-jF`Fws;<jAA
zzLec+j<2nxjv(Z4p3Wcbo%)5?9fjB5R$;zF@7^p0!ATx7vS*LFg0NqiN}&&Sj)gY<
zq031c8AA$Xls6hC38RiI@Z~C%tn?dGfp9vY=h$T$B<x`B{^~0DYKBi}zj^-#+*4;v
zJ{iK=GfKCNF~h^mjA{d|JsC)9k`!!K8n(5?z<RNZ<@wMj&G`CTEEFv%Rka4Dn#8ps
z^bJcIE{k@H=jk$hKI|SVgBUi~Z`&c`bca5}&454Zn1yqMqtd@gw7}T>XgVK_fYq0o
z_HSv+)%=&71V(Mg?FobO{(sNyFG0or0a(9Z|6#?^{ks5F11)Y>Lvaki*ptHH@OMIV
zoc*(ukgy>*=7Sod;=)1|BMBGY6;P`;-RuzoN|d5ANcOVHIH+Ud{6%?p3lnW3!zVU1
z-ZyiiP&dpFe*{*)gyE9QZN1WB5`Oa$Et{54j_lKuYCo&UA^Ckq=yr<VhB%KGAm(0^
z)gp+E4ekO-eJ&r8%(}gYyF{uCwUR`NOsHZ;2Oo30Ce^pXQ$`3DKZkxn_t)n|hHvE1
zFdU2_{{cbFCF;NW{)6+%>fuW71+Xx}we7|2v#+lwNz201Y2N4SqIB5e?+9cu#<FI2
zGLugfbaNA=lW0t=4#VFS(gj1EL~{O4^h;&#G<2BamkgNfA{P!R<mXWcyKTD^IC0g=
z2lsc&=chzr)oi##T7F?BHrs&`HF4<_Ms%TirP-UstN?_Bgjnm9s<aVi2Gcycy;a?8
z4N9jZ&a!XiJW1pV`*t2|Rf5|naIjgSOGnjwDnBdD{-_7T-cUVC|7B!&ANC_`>dC)q
z(RQmi&Z11WwXa`e|38k-GOVhmjly(DOE(D8jnduS-O>#L(hW+7G}7JO9nu{l-Q6Lb
z-@M;X{v5atd-lw%XRUk5Ya|eoTtA3zIwrhL#J7Z_YTI;6qVE!PkZ7bKHSY$3O|5*@
zsE&IzT~5Ci?gN+qh%D85nu}2|G4Gu#BfVegI;Mc%l=(SAdpy*JoZt0F+Fpcq167Aw
zF3WM1@6G*2wI10Q-@x|CLqww9MVPmbA+iM*59<}^0*ADdF6+NIhIKzhe~+(OXv^7(
z0Xu_~f&s7#8)!nG*>-<n$(VImj@Poi-1r}v#eT$6^4aAR@qZ2aP?l=d?owcVbEIg1
z#}9Wt+Lq{+$ziti^LLg_j*x6VdyUvTOs=1&$>shCWxNN4v$@6L!ijV76V+J8H86O?
zrGI(1vXi7=5;I_7J)9`>sZdU6qKuagPiEC5#=_fac-S@G$qy1MsHLZK`HHB|eU5I-
z#Yr#3t5&0rWsS&3t23_5Q8wQOg_McRdd=5rUDN|nKEfC9?8XMx6~oaVqU2`WsWin5
zYCd^*ofIpB<tM{L$wrHV_YZl_(gLGp4BcsnFOMN%)tHzcph9*bC9ZFd1$`_AmM77O
z|8_Xz#`WtPsDuX`w7H$l%pZ3?9+l}96X4hWUQQ_kjR1b<YyS<vzl-MNNk*5upLYNp
z?9}Js5m}l4pxMJ!3ZR%H29<=zyQB37>kf>cBWN!M9=6Cq%s!@%nmrhOdL*KJ30#v0
zHF7t$-gHD!vZoTJ_rgpaaY5Uj-%w5o{LLI$aRZe4b}@RVgHzy>ggw<&@IMS23uj4v
zn!vlod&#NiD`%gd8k#f?rB<z9(qa?e{~DZk$!;+zAn3A?Sf)NBF0ozWEt*#z!9rk>
zvoX{g;vH_w#QY>**zz~DCIfr3(xBo<h~rDPFAGr=HJSO^M6-coRiF3o4zF&Fi{+Q|
zx3rau6W_<fB1g)B6Uqy6>S7x=S{#}^NzcTolx7u95uYH&fQa#29}8VIZHM}crV6?Q
zD{CdYPGMAzG=kKmnS!5Q$7-S#c?d-^j5cIRs@qIIQc7jX{*AnAxcO0V&XmDn_UlA@
z#&Q;cSF?><E9Z#O`nYi1vk7~qD9C1D&&hL%>vup2>Mp9*se&JwTxBSQ5Tza=ysomc
zO_XMb@5_&gf($q`g8YwugyL^Yws<?pSMC92Xw@S#To(K=ztz|19~&+8SV*=oK4J6y
z&LHO|>Pq#DU){ZZ5D8@N?$IXm4|=y1WZ+SA?+oF^yP`gC9Zl>Yd%*#{iEmoe#dDb$
z@RGGy{S~vKOUuh^e}Z4}_83&N5ak8?9XNWPG8w_Dif3Z-bA+Uw8&mLyfqH}doNs-X
zsEg+^(bQSY-EN2&FJe>pW9cz<Yo@Tjki%Na3d(U)0}*(3TP+uE;rs%E@#0=TW2o!E
zbnkIN4j6houH=4~&9tJ{Ii4>%?}bGw7+y0CNgm!aZL-V{Dvy=3Ahh=$MzkJ|?u~Lm
zZLHU&D#N>$*c~)I!7q7V_%U^YgTwa$AsMm8FqF22!Nos(U&UY>HnxJXj%1G|)0Tq0
zY#!h7FB2uD4fHx&Y22YpWe$1bPxuD5jDnC$v3Km$1+&IANX_n-PInXf0*@CaDvaC|
zdnxFtE)py66-O-QMGv$p?Depeg;;+nrgMFE{e#yrA+~iZ2k$*p0u4F~A1K#k7HpA2
z_PT60Emw@#&e!VhRzJ!P#(0&0AbUORT9L&hZ50^fJI%G!S^Unq00dDt-05m?uX>&O
z6I|B=h~%>?D?|t|%!cjFR@(!IvxWTAcdR<nllH4XMfvQtS3b$%>-yR&9yy|5(WgY2
zB!|6`Yj%#jiWfaVKwuwR4*8f+o-SpE=*#;fQen})nf2#{RO+Tidsi;Fo#jc!Ohflf
zsMQsSPP<bfon@NxnA9%d;qU<7Gm(4*w=IM-9jz11tmBIv)>%@M)38f3sw+cgLMv7>
zgKra%c^>OIr7S*BsvcVMkrUDl`iPCRDwx;yam7$;@$jz*hPhAAq(X&GwK(Ipy3nO@
zZxi3^E@#rl4)3KpXj18JT!jwAcUi?gICSq+wix(;oW?Et1Bd;}vx=(&<faDfEUI$h
zFgwd-Sl;EpsA(g&e&7ZY=rOD!723dytiXb-ESc+i7`QGiC0P8nCBaiperjHObrcoA
zU17d`d1Hf^Tblb<efz>rvy}XK2$C!|Er=rC`&glhF--?)gxCY~U9X<$$H<LoAQXNy
zck83X|9DN_NeH;3doEt>0;Hao5UWQw%fzrB$3G9`=8UEMDDt1re%y;K$mI9>a?uT#
zg8ULXTu|$D_Uyecb`tJn`NP}(MfC63POt7wQScTMfl|^0g&hbUcY^^4ywqTpNo!_~
zPJN<RBK0*DIj>CcDsQ@)@VY(w_aM3>dvJxY&#V{M!DD42pYghA0S9(7P1v|86JmR2
zl-|J$7*7-z3kKeFyMns>q6D2t^P5)f#zq)h(wZ9em$;cj7Pr=V{iR$^9ds+RZlhWV
zX<jz7f>ka?f(n<8>A=Y6Xo$93Qd<5t2HeNjmyk7Y6Hg#{UPwH*^oaf~f9<nm+xAIZ
zW;|Kz*#BA5%X)!laS?a=d9mgz3{`INA60_uq$E;Rg3<c;Zmx;EzxLcl>Fm^F-?xow
zIl%Cj=BN?o_q5cf*9lz*H(Gk#yW!Sg%DP`IGzFr1HYgZr&?7n}X|`?kp7=_V^cK%E
zGBY2#ajLug?FR*rO60}weon|{^#|}46eY=7I2IV2Blx5YH~h?H4_K#E>(<z8Q?S@A
z>BXc?B)ICPPfC^)9jZ))m|jm8cv*ymwnFv*YKC)A2*a-~qxt9zj<SlJoXy++chvOl
zxj8$JX6)?&FzQ<Cw_`NFdr^J;)yj;vHT`<`g;?@W6`nv?3wjh}w;QPk^5X%(cJp*#
zNOEab=$W~-9xi`OtkG2PLgn~N7LKA_JX<&{Nwvo#bogp?-=Il#=eIQsy2)g5<0piC
z*sOS;1C5F00)Xy+m9P0Ecxl={3`OT<ZbU;c$@xh}`Y^>_<>)h2EUE_6ykycyXv;H-
z1fu|-`vJh*7-5U>wWGX5m)ec;PIvnR(hL@?DQd$AGyELk#9_!g(z&T77p3ap3veHC
zgzdz`Vq;FhIvA%lG@zpZed&G=QGsKAr&Mn8-@bwGB0aUvm&&BAq2VUW&`<uH+)|r%
z7w_Nf6u(vK@EIWLrbuaAgar0f|AwU8ck%n+TwRhv=sdDRo6k<M;lw3pGDwcKxcA-r
zl&P^%dr&KZuf$sAYH@?*TfA&s+3;Ad6Sn0?oyH$JkHt>yn++cAd>S4r@KXIPV%p7w
zTTK>a(X4LS3KsR)OWiZ5&EiFzR={}?TZ6g#-<jS~nD%q^06E8*ix)SmC~6zzbQo;a
z-a_p(vh9O68iI1AK4aKOp!_io%S-3)Q+23CaE%%IQ@`OdJdc?7<TBKspP_3TEkp)t
zk~5eBp!v>}D%o)&1m}Ax-8u?8TQx!~SW|3s9Ix-g-U(5n{s|VVUPw=7@~g3XQt-_1
zUMx_8MTue4JqbjkLz;OQcqXr%#nw$5H>yQ^HR2*t_p}37rcWqRfg5JWL#Xw<E5&if
z*~KsclGQViD`!aGdzH6TWA$Tf^F}EgliVs=?SXM)+mL9zpuD_2zaG3&!<f6;&Hnw}
zEM6IyrpOvSgW`^YfJjRD@>R_xd_Yd-b1(&K)GZ-DPT*QgyM$;4N$9SN4kfRq6GsNd
z!%tV_B2$hY(`Gd(GjRf*G$}!2->1ECUBK3MZ@oem<e=Awuph==a*u^CBga}7(D=zl
z7jRVlRS&HeJ26Gvyq4P>K8AP7UDF)FnI;@1(klNaU%6ebp6?^!?0V{@YsZAYRYA(T
zg)x`wFA`(Z<^~df-XVtph#-zkQcgei`{+J^H3iq8zu%E?I=ODY14)6%V!`KclSS3M
zfyYgqBBx%93-?c{l};bR)O9q#KXcnhO<9+hZ#o0|#BVgg;J_ZF{Yk&;9{4J`lp!p*
z^SKUQCbhps7FFTlGJ0I$oz@KO(-Y+>@jt8!@~ZPFlcd7A#SggH-(6ln`*SBdxOl+i
zEsnU5h2TAptw;#vs7BNIk+om17U++RfLw?ioXmGqHcVN{_z`nHd=d6sTA#t%2k-)9
zdxunmmw=))az>TGeo7Kfllw${y5mgwM2IZDjl$WPz@Io`dgv8s==I$7$!YKyNy{!;
zO@~r3=~=CS)L=!#7Zrj{^5dni!1BfZ)Uyp|i@b!%8+GQ`{y)LNe$fXQu~e$Itc{cQ
z#9kXP&_PQKfes{<ndTxQ@$>^-;#?QXA@LU1|LqB&LWX6w%B4Z6JVKUZpamzrZPcQs
zQR|fQgersBfnNQem3q`CvA%u^^RiidZjT$&FFFiz@pDv+(&C@cg=OpG)nX{}OkwNP
z0kb8BO`8W{((EQyg7J$t%f=}akCD_5S=cPSKROmH+Qdvv^XLB{Gtx8>xK@;|rFirj
zVLx<_c$8kFXXJ#^8QRBa&|~$OVQbnrbMYexpJF`Qv0&OQLb~tvtNg;=rDD+~4*r2|
zA-z!*{yswsocMxD4#?<U&rTQH{Kkq|nLAie*LmWOoK6;oM-BDzvS>%jU8fjJYnSz5
zsPTQVeR{GbdcE1&`YVTq<R7JApCh1c8*yvL8{8KeG<f&Zm@}Z77WD0}I_3`LNwP^b
zn>uo=^W;g~3qC1lOn`b}Jp&pOt6HvmE+~mGSEg&E<jQjQx1gL!C>yv+Y{7$63xy0G
zQHK*XGgK(d7<qcfoPIJ(6?Asx3?w(9F*}rLj(CF~*P(fmv@rx|JsXbNoRE&yxDxJM
zb+M{7Vm!$XI}pOsiXX|79j@VjbFCx{H=5<N@daA{;*V|`EfueMkx+4vUV?w$h4fFS
zz$b9ik|Fi|KiC~-i;48N+@sdG2l6tQpI%LyIv6ib7<28_y|$H*A%*y=8m>I8)2BFa
zXn0x%DmvO864p{v{E4R35Ajx~(RjJksS_LMG$x&KsI^ed#Y^(dj-TvnxNgq{o3~!D
zXA9*BFI@ZxMAsV1y-$j#!Ae=LvuXNm{{}Kw-hd$tEFv7Mt^b+%0GS_{!0#YyySpB*
z=sOgeYz8Mh>h_Td*@B}B+?Y;3U8F)*$!E5MPX&JW(z4kHQH;e9!e}8>4wFl^WTEvh
z-Knb6%7xTEc;S6loc7k&oH;2r%WWX}x5L(PBLQ(kqum=iU;Ppvc-C0HDN}`hKlU$t
zg7%sHivaPodz>rwM->Kg{0dTa3tO!Y>!=B%+~t}`ym>7>IJo7a8EsD-Hk?xGMlY#D
zP2A0wMKmYMq@i$>`a5(bx>O>CJ!+oWT&eZo1qXD~SFo6Lb#>Fy()_5ro(4(%)WTn!
zQYxhxmXNU&1ZZ?mky=&)r=i;Dmj=vlY0||kv-$J1B2oe*A~H-!{~pUL;4+8jTBo&4
zCq<+t5Q%(JXQW+XEDjD-pU>`hY~W5W{Y!zBp<#YoIg1atGQW7eDpe4PoaT&5t&k?2
zR6d1q=skrK(IG0eN$HX_NVt@_?zG@|LmF^&8WabeUc190gyqHki4_=!8L^#!8fu+U
z7j0YX^FutWDW*!*uxGilDfmn?T+GKM)Yj4Izj|oS6oPN|tvLQ@R#eRo%Ef}{qDa?r
z(YBxAKsaPMf}fy&Ymz)fo80V2W@u94bL4^!rgzUi9NM9tPNVB-k$b}U-j;WR<+zxu
zM|gG9B~26VOk5Ndz1qe<?BLEY9f|5ibvi5d7giC;+v8IgVJo}>#isE+PZ+|5-6S;P
ze|8QbZI*icd$Gx#a2z4a@Ymn?3=#LuV%PX*oUV2b!CU+1$XXB_vQ~o=_F7t;d}4X5
zRG~DUSR=GmWM4dYvEkg0x4vuJopM)5<Quu9GqlegNQ-WVb@g%%kB6=jaPb1kMe>2B
zbY<b{<_WQ)EHYnWma{J<eg??gT6{><?{ryYSLE;TnlzY6#Q>s3hwmN(H<mIYlT#>c
zT>E%EJ+a6`v9~BC!8*Wj&(~goX#Ts`Y!!<FSG^(mJhPALz!^zMV2|Y$0@C|}c(2a~
zm*D<gI@b}e5dQPc&SbZocM*W%-bd^c+Q%-CS!rVDUgn<&wau!3z527H_;=jfYj0?i
z0#Bf^h~%8pZsZThChB#%bEitJ*tIPG`Q!4|6O@#cv<#$g1e)vRaYjbqJkOXt<nh`Y
zYgH8R_%=fi`t$M`A2RjZqM%f+7t;imU2ac2sF(>1T3vN3H|RdFE=C?Rz!Ebb#9W0O
zAGy#}b&dyX(AhB(BPfbR<E0<sLpMX-yGgMK#*a&Pz=gT|7z{!W7_q*j^zXxe86($L
zftgx>HDp;)r9C9W3RBm}WMZa0oy{w%N8qL+EMm0yB}BqK|IU44sI>Nj_g4Nq{fpEO
zNGm^+#Tm0tWP&)dQ--Tr3>b<vyMKLk<@UND<DV>Qo51?XyQEx1nF?fhz-Rd&RQR#_
zJMRifM=*K{X<rf}nHbe>njP0WVIFp<^~YLn4+99smRXfp*N!z)7VSv$w*I|(?iDcB
zJ-lt@4;(!H_pUoXO17LFt~quR4c1+0xGWnj(E#&VnDitj2!R`MFG>;tCQvzMWB>VI
z?rSztu^Z$A^4MhlOoQM763k105*c34zd-sv(msBEs41wnc7c#lcKb0TASnTcDhS)c
z#j6jOJ2@im4^vDv-~36rWw`%*>2<k|Q_FD**FIM{;Tk>1lOFDFYnuuY`<ns~a{S0n
z(1K4Vpt9k$2{B45=OPLJ^$x2h=6`3_vO{H@7BBBRWel7f+r>#qu$9~OH&lZYvajI|
zucdxq!Q6|Bs15lB7r%k_81yA(e$b8Mr#8gV$&5}aQH+Yuq|GtPzda_r*OhA3M2p+A
zWxuUH=@a;N@7*Kd!)%~LYVHuv@7q@QpwwG$-V0<#R~ZR}GYZrxfOg8OhG?w(bvJJA
z_5>xobaUwFUf)%0D$)Fb+n9X@QU+OOYt6o4Y46(c!?G=tK=be}j(?Z_RkFe(Zot6~
zf<HsGD0*k1M%TCFXicilz_N&MhiV*Oo&nP4&K;teH1RdwDN^=HN;fXmR&*O>5h9_U
zeKt<6C{b=*?ETiT4Dx_b4y6t9kUnlImdxoMA(=`05SC>*WKjdA9Xc^x{m!9kBr=Oi
zix2l1V59beDc9ZEA~jO5LP1?^fT#I8rxYCLU_t56VZS1cgQXbGTw37;*f$nxFvn;a
zLuiZJQMjO}NQ%uEq=N?ne60RLNJ{f1lG|r1VJY4H)Y?+pwfe;+W5q@3(9{<D!Lt1J
z4rXn~b<8<QS6Y_GO9IBa(w*-2zSBSW-*XK&Pj{qxowg6ai<Nb92V=+QcD`Z;B*xKX
zCVtQR>e<8RtI?`oU;}stD*G?{RnfPXsXx~M(8%ld>+LGJD@DF=lEYT(lV4Bq=4AX#
z5O|I<WO;AE#R?#(UoL{s(1^IyRCUkMey}e>8GC5c@j5BP@F?h1SNpy7oawjvw)#S^
zlE8m}{Dv;ZF^?W5yu`<^A^n$74p=<y@ADEt9Wdq;jx!sp_lABEgB@#?NGgt^zuAO4
z^5OT7v&0S*R=Ca&LNv8se;2wXx63L@@)(iFk5eqfM($<}4^Q!bpp;ek%NnBE6Xwgh
z?%q(^3QqY<86%b6%H4SialeOGsh~(QZ7kd02q9D7x@?^M9_G%??%<op5>eInC9chd
zIZw<2Sn(pQceOsje;!dJ+*=-du6vd$Uo%x-Qky%Lm#EXffG1+APGYq{V^b9kBz+Mb
zu!BgQbqyJ`c1~qP%^My$K~N8hFNSJ)x1x;g9349NXI!~p#p_YnVX^)j7^c*tKq#}-
zVG9doDB*y5<A<N~-ME%_Si#-A8}JmsVg>ft)x*Xm{Z1h%Bx{|~xqx$u>@e-9-|>ou
zRMm^ako~Cw1o2Q8G<E6)2L#IYEZA5iO%Nnv5Y}fxL)$lt#qRf!hwMPhEov{6QW0|L
zRbg~3xJy0zVRjK=#gum9Sgbi$I-h2XnoMToAQBqiKf)lf9p9e^fvzlfRJC}N4;|-<
zm3#dTvVhN9^ygzkSE@T8&YvHtX#8z*J1@2Q>>A=Vx>o`I@uG12qbrzuQ$c}qaujb<
zWdrXW8hNr)KA1&A(G%iAQDk7SW<j3mYzxL?t8{{X7vUc^L<oXfeoDj8G-WW$YXM62
z3lXKW_{0h^67$fG6%<v)3JzQUDo2@W6)aZjJ?Q1ArF~k#;twJ9eqPzgagSLmlAPqO
z+h_S4)o(mv{BxCX3&1=}9I?o1iOS+g0?qV}wU!i`vUe`v;HngcW)3n5%rNNqJ}M)q
zVeVrp(?<=&t4b$uDxelX85u8o#N}IjjTkn?@P<@_Y5oc4VO_GsSca)b&S~#dj=HFv
zk3sDa(hI>-7%w6XAKl$|hyEuWwwVrPGPGP2CLo<Rr6~a}5FTc+QVu@8ABgVBMUVV|
zl$XRUPHJAuc8mQky_A~HqQK^_Z4JBdQ2m0&aH>ICIZ^b0cfiIaCU%x@>zCVd?cX7&
zry6Ui9KAq**txc<a?ome-TIF7)3F)$WqxVt5gag?yPj)-H~{ETP4?^dR#u4jLCH*d
zdJmhDpqR*9db-V(kl%PGeX-8XcHQH4rC*4q*A#WMQeNlG>oWHvf14|>%YA6QV3#+i
zkDt;<I~<pB8b0N(wZN~Kmr06&Ad#Xv_JuFCFv3fG$Dfjw#=hbJD-=RRawp_C)f;^v
zuzy|g<D>!)z0A{gnb6N0D4TTCFH>%jdQOx$gV#jEYH8G40FXDD%$R(nXwh#<Ctr7I
zf#iW=9QFrIo)~-2uT$#OulQl!(n(Xl1JDXyZv@*}UN0OvACeYV@Z>cN$r7;tg_+Kr
z#SSJoadmJjGvqA}1mL!=66-aOC7G=*DbmxqPYY@fu+b#iy0vtNqXz#*H;|xmnF<OQ
z=T5he7!Q-jps^cY^<Qc}-1V0Ye!oSxIG_G*=iW0wO^$I@Q&vwr;>)%R;+V6=nFrx`
z8k`GfNn0hS&X@*G-HwyjYM>!HS^PS#fB2%mx9DfWQzgyCi59uJw%~ch#DNOFZt4~y
zMmK25u}@CbR>H5_i#bHNoa#CJBxhIdJ{oN%Yby^NUXoXIEie-PrVitc#dA4fQ>cmo
zANF<191rC+-V4L>*r+9q?J83$Z+qjXAc-mPSERzamoJWuK7(r)jTkA}kVUa@MG|Fi
zxy*Kin8e)QxfsB`&E$V_9PRXhFKXv7TU+tiaJ~$Ya(UU<%+suDm@~Y<GISo*9YY^`
zPl_ylB5r)hx=zyaDJqb#XFKN%cdGs2^P^9A$D})}d%}cK?*-`qX=%NY;w6Od*gFj5
z@9MNHsnYA|ujW90sx25zaech_w~bwAK0!q!D+-6I(_!5(ZJK1?K4cKZJoVt)OzpVw
z;ow^h^utSl6mE&Qsa3!BsNJuCr3wlqobWz2gz(`k+${P0Kigwm<7sQ9TQ#K&_H0N(
z3@%xfZ`T!^Yde2@<M9_rb|fJ^ninW9a=+<)wm{(iDE~Y_Zuqee5?_iY`vB(ybjvTD
zR_1+vN^+arSPWCfC?hWGI$m5xy-&6WN=uzHTsUq)_jPRffpzB93yCImGf#3FANL6F
z1%A1+i3dK|MaUJ#rJf`*m%5&=)B)FmCUwV5Rdue`{M8j`Vq(&%HhNp@Ky7=O^n3i*
z_4e``0ODPaZ%?y)A4?8#65t(x;#dTh<=5>W{k8(t4sT!`O7d|=Qi-3>l$-$RvPjPN
z^yN<DWVNlUwK0>bhw69`>d$z%@F~A-gF)MxGQqDy{OEyv?u%Ho4=r8`fxEnMA~{m>
zcKjFN4JJ4#hX<nRvS=_kBV<@k<kv#gP0J<p-JbiM%ZJDpNwgigLYkR4QB-xy6tT1)
zYADSuhUokh!W`&z?+Fzx#I*GyQyj9Vx80ePgBjklIj|6obFLT>p(t&}v<ez`UOM<a
z@vdJ#^-sPY$=Qdjt`(>*HnW=e3cS1D+cS@cU8oBrcMtb_dJ;TaX$i{e6MQQfGMK1o
z3cC?^pa>mf6_nFRdw+j%v1zh8!n<vRp$uyk4@!X!ze4*dvf#ag&U;^5aEcl;Y+aUT
ztLf|5VBoMt$Wr>bTito6p>I&B;xjg39A){;UBJ(Vg&BZSq`j83+5Y{`QIT=<hms1p
z9|;wG?NU1GH4OO^R?a$fPVoj(D_`K;gseu0_0G~oH^KHYSG9y;TO9HfT7!!Zn}*j+
zZ5xY{70Y6}JxxCYOtz=Cu#j#Lz9NoMEf<#p(+modxcEelWs5HBo|Rj32l^(AKP!iC
zy&%HLk15~y={7Z(H~i>M5eaQpbBh{RvXOO-uLc{)86a}#LQB6~XOX2z(>l<g6xk7#
zRa?{%n>N0A^$=8ggUj+Lx;+`?o0s3vplphFWtop)$5y@Y6bUaQi7$tv_YkwK#W12P
ztO2}|2)(1m&JWMzTZ802rr+<9x&$q|yV>`dhB6BXZLGiPHrl3f`9cjOfWbf(-{BlE
z6_>KPPcO4+YdFu4zmIYbT$9mNWYd%G9+N0F%gX-^iBeWgtzecdgJXfo9MvXs6OV(S
zx=%3dd-t;nL`S-KtlpcNS3CyO%+rWr(i$Ugkbf0J9B7ePvD=?mrKVGu@v`xhu)TI-
zL#U@rr-%H?Jy&=5S<D!*I~-RBK?eK~aKHMa*_=F%8>Y`G{EM89DBW+iVDb7W@jWeJ
zAvQb-@3{mNQN>`hlf76UQz6%aSnCP@?g|b`v)xpbEPHR@w%ZRuIiD_|Cx-JY*E|Qy
ztM<#w<G=lhh{Y4mb>9d^$98de{Kd`j@$pNst(DX|5G6{zpSfS-e^=Y%_3d21^-s!!
z^Pb&O`z|m8Dn%8no(I`@eqKu`S}JV&!kst~=N2>Zm~<<bUpq#q@HmA^CFjNDg(mH%
z$3(xbq@*4p5V1=WZV~Ugv_=(0*_Kv6B~qCaU_j{e&KbR(#q%WF0#2{l9d@RLf|a0Z
zf|fi<#=?x<OiC*)lP1|gE}xy!D==M#GKVB{_3#S5ciR0uuS|*vG^HMCbf04gA!pZr
zXI{UrvpQCgT|Du;sZv0~`GgZiS8*@s(B-qQFuK~N!aqLFD8s6o_PYS9e1knp+D)7|
z4f*>AMF}E_GQhoJzwPb`5-=R1xX)0i>bEXpv__34j^u9&sXN}9<=V!+y2+z-6bEN?
z;n7YFB&Ljkan}|8cSlyAmJ==0c^~OlPNPwqp;F;kI28De;AGHHsM2uaIPxf!O4w7z
zHHv=2MkjLA@j*l<8gUudwW(c>5y<|u@(A3?t<Oewv{*w@<gew)Q2snlDDE?kO3jiF
zPoZ43QG{`5Y8MI7Z8fg(F}B<4pRrNAmI)3&gS5jH#TtAMDW&pP^OEZ?9c%fI5qVF?
z5o3de8L)8RRDJ&)qA#7>5b-qTJ+oXyN0gVZvN{Ozt>grSEGW6c4Q~nfU5-3NXda9v
zj1WAa9t6}De>2j9UTERNY#!dK=EWLKoE@=BJxA(3K+6)8GMPBOi49kzpZ@ThhN(fo
zzwuc(IUmZqW)h2eVDn;#hdzS|cpwsAE#D{GsPm=^j~2`wMrxK-`qi4RZUh8@*;mQn
zQ%BSDODhlYVmK{gY5ATh8D70w0@eE{_DjqeRrTy~&Ayw8VU>43^1|{xg{$!PWRtbn
zuJcRj^LayC;0k$|e3F<4r%}1k!-=>($Mzfx>ON5K$14|zokWiFp`zS0RxE>H0B}J_
zlhNguQ%MZ``Px?`G=Y=onE3;F0kxPr!(SfD2}W#nJ}Dxi0#?D|kap+_MY5<%;KlP>
z)Su_r+ag_v%rN7k{)x2Vy&*lF_b3<nmtx_x^Mh((dGivc%1pRq?`5e}4ha%aUe~Vf
z483;ez@R<{0B2q3K<01=>^W#>f<~Wp2?z<kemVp37Dh%!!Pook4Bm0j#Lzsu^8B>^
zwq{)eVAhU3D-XXz1SIKcnQ47;&rRYDGMfRVmcQ)|9B%wr(nsk2E>hW0L8W}B#jlHX
zA%lvKiPC;Qto?y<cF{~Wc`}Q4UA#Ybcu^~t5z%mL*c2z$UJKnBW!hcj@7e(IR!L+z
zvw!e$j+B-NTt3A;6cR4OfsTf6#k*>V;DB@nh&Hq#a*XT=e<x$4m(!>sHqcugm07_;
zM}jN~oimh=h&dQduzlqo=G1qlT&=^JTtsm+>~gzrvYMn5E#B~a^nd3dDd{s5Q#hFS
z(b;)?gJZXt9sBZk(|_PbLY4(fc+SxOZuA(bSvi`X^#s;0c7oHKgZ}vdpLE2ah~4wg
zhot8hqM7VCZ99%-IqOg2I@a<f8nB{pELA)1b{$_XbNi_xcm_X{RtS*<B-zn-Iq@nN
z)STd(kU&H9Y{E2pleY5-1iLtT!jz4h)wLx!VqWK?uWNY0OZlDNHcDgwe_egi_q-^t
z5NFAv!JG^KNz!Wsh}6O;i40+z3Mb@JH(A^FPID?bG_we$dCGiRz?Vc())AQ1*gV9b
zV~HST(J`H?W*u;y0~=Xhs2hrm^zYFpAhs;ynJ8=KDh~7GKC?j$)}$hn(xqKS&(7t2
z5~KcFOZ_b2>>AlG{G4J*|3S}pKd~Z3PNcsIZ-$Wxt83Y!Mv?335R$5k-IdScXOLV_
zED|wIWTBZOaKqrW?2%S28sEYCFOsa3|FU`d*>dC};d^QK%zK3(6d+AKanMbheZ?7=
z`F&{P#Icc^7nPXjao~6s6y5!xO)&8N`V5Yo3363pU%u(p3T`8yUr#_5DkJ?-m532(
zafqBtFmACzu~Ob~%jV6~_k!Ldl~}N?(BQ9tiF{Tg2-(W{)znBvHyt}X;z$80#)o2+
z2XY;qcE_$C-+fhPpt%%xnGs<~LjM?G+#MW}hiRDUh<$p0_*%<WYCk9SVNG0plImUE
zB46LZPF%4xX*9Zg{#6v`FaZ~TK@rQp{%xN}I%PzekSWF}&Mt2lbEkx6q?>*9?aFJt
z9>-4^*sS<?@P!k|VFQqM!c>85S58Tpv#0z%k0+1cvH*>@fjw8zq<5pL<Lan1Mxc^U
z{m``qcvjQ?%cd-M{1;J)(r|YEhxYbLjuVy1=ixLyqCQ*H@#;cer5nSfAA;s_R&nTH
z2d)@9C9@i7@!E9GDN+_Q+N`qEDXhc?qS-i-Eq~LCn<X#MNFWUQLE2@?<R?3mJa0^j
z62~k)DU~X|x+Vc8>BP=E7Sw5X6+|uND$Zbv8jvj=Q9yxwkV`6ol~My+v;QUgTij{B
zzsmxV?u`p-%QA&r>a$L=f*F{06oB1a8ZIK!p;_2(;P>=>)4Gpl0_$_v;j_)BmN!n#
ziLp5Be>+tI6pnJJ69WnU1B{j?u<1NQ^{&kUO+RADcGLROV|fj@{X`3Y^^RAegGHos
zqTpMOjjYh%VyUkCtA;%_L<Cjctrg@j7{NQmY=+)6wA{aXi|&)d68A6{GaP^0fIUf-
z3<^g%W`V9(Bsd~MVmOiJetqxV{WjRnPVlhM)3l|&Ly*l314+L$^+%4^SqnIWFk^0(
z>#dMxD_w@}m>R@;iH&;hpL3LGN{Lk}cZ}op!;WkuVw{@!o@J(S(tbxA1@Nvq&T<*z
z=`bes_-e=|7D-x3S5W`bCm~%Ka7?(xom$Gj`pdZN<#O0+^-trFRsqS*A>KaAyVm|X
z^e1Wg{ZL3VA>#X|lZOZ$e{L}}jBr@xWui(Unk$4xV`W8jW;Dz?VXLoxQWV0X855y1
zqu8p1L*Lm&I@We14WXe>EqG{%JwK*j$nbwH4B9};<aVx9u-Mo!bo{hHWKUaoc+~Jq
zy-dKv_kgAS)VJI%t3LuqQ=fjQ+uIcGM2#l`ax(pom}i-!DJm7wT|oMRH#8!7T!5u6
z;vajL2np^d(!OH$MIw`(ykKh**Nk$R&p)`ITlO6X1n@<IbD^%9L6n`^iRt1UMm#MZ
zB#QU75iq%p?-EL{q$pQ<hF0OS{ELgtIMM<F*Ur#1Xoikxfm=FIAr3X5Yt~dkP7QRh
z1YjoI?SED_lA{Y?TISG>5rrE}qz+Ft>z@yA-Yi1XdD0D1kfgpBIsL|hPWK1!^a3et
z4-Jstzg&PoMb0E@CaCEH3_)ufn=YUTlD@89{|9~l`}aRhu)_=OD428A+TQ;4sH|#2
zO~Zb*{b5x+LK4L<6Oh%x_mON!)CC5?KsIW2yktAQ88YSZ<ptC}4l)LP@5-E?Xwa7k
zufPkHB!!hQs!)h@VwanQc{bo}b^p8fd0^O)fpXRgId0O_xi%}id?D#z0=ds1u|7rB
zlaV|lhrd7LVR-QU747a&<Xc=Sf)R5G@-}y?foc#rhnxi)2FHB?Ew<nNc9w4^ve`iJ
zANTt=dHZJN|L#5sw~*<?g7JcSRh-&(-3Z@3y2wMnxJ^g<v3*}O-VgrpkZ#Ho6b1ts
zDx$Md^dS$%)<AB-s5aITts-eCR=FH*MIFWPUlqn=K|j@<b+tGeF%O&RR2{GZQ6-bn
zTOzv4IM<n}ml(rb&_q#X;jpwwd>4Az^|zL=JF)NdpWWT{$2dXdTz8>Nc7UZ5NTf7a
z@mTs}1(dY&uCBnyT?8-&4kG2YT!LtY%Y&8qk@x;9WyP=50vf`S;t8cHU6gfU=7D`Q
zZhZqXGju2wub?wTCsyuL@t{qND^gxuh|sf}2PpE8W!Y_ng`0y&Oap>-ph^+*y2&ci
zHViE-&rDp@IC<SjsQzc6$@KWI`*KDgJxLNFY&qTTldpN*_<qnxppG6|X%_F{ha6Nr
za$GXH1oWL}ys!9nuY-m>7+QxCNaNI7{=d%1U&XX0DOS1}bsN6{1^LKEHT;TV<4});
z^lR;q!O;F7{hG&9hQO(Gnu$d8TBBXtGR1{+W~x2`a0=l}DFt;T+f2p9p=@s5F3Hx5
zXyeZXo>kix1k2Ut<wp&x5&p$Wb!O`i&KH%6NKyTCm50z)S+6>{M?alM`R5oJL?%A{
zVn8pESW#>K^=W-aX$B7QHwtI1@z-Z-F+zoIiP-rc#udVaG9DcBa0^r;+)!IJ%n)Rh
zD%us6P>9A*XMA*$pr+C4mF^HKKiI%K@RN=xB~DS^_hNyhzsm{>PmcN!+L+Oq5lPMu
zaAJrC!4)JDF2}TDJ-h$+e64^kpg3_^10vo^%_b5TJR)@}!Vb&VKu~%D2pyoK<-goA
z=p1pdYI&OU^IdhqeQo=8e0#V47CRnpsP}68vh>ZN^YyX&b#h9EC)t(Y!sQJRG`*jv
z2jOo?OItNgk%BKil0o?;wqYsuR$-5z)2Obna>86C&T1EU-vgQf;Tb4!aLLBQsQgAL
zIi+gEjTN@*utF)QVin>Zh(b1d?ns&0?%QF$lwbuA6FeLgA{9<nyspPAr^9M+d1nO?
z!h!`HZiFTKAuJ5#PHI-j*8J$Z$OfWgpI4qf&!Ii3r>qQ9Ds5Z2ynhBf=1ylxT63Ga
zmIOIe8r<N#P~U9NTJCh0+|Pnkg}~O=s!?Y)${}Ey;4l#oAk11U&B;{yMO~dYwhKZ?
zrYwNuoAsC-x|F7gDm<LNI}e@XSGC-+{(ysR-9A5qX0#piyw=*8Hktnn2#m`vQGa=j
z-zmj(*oeW@_XmGm9OVgGG#^~c&E7!sh^4i!Yx~0}WSaja&ij$d(X?o^Z7;qd9d#!m
zsaHRreIIda%YX~{?D7%9T37DYR+&|?ZZ7Ata5GWXFUCBAnC_OGSMTvd$t&M5N>5;N
z+U2N=hG)bzRT+HuKAZF*_@+SgkIP-FRHj5cuS5O^K#Fo7?jpWqd5=)Qz!>_Qq8e6}
zQ2XSIC>y-zTS5I^zVL`-R3$Q`FR)A4zZNmkS6DMUli0TE@083_UCt3Qw-<N2-vse7
z2TGE^`_L!-_-JTvLs2*j5Xdl5JRH4z49LlRq7KQmaJ+?I4Vl<>biS=Vej@}0U*GQr
zJKFA2I4#aVl65#qHZQyWb6@CU`N9`U_IJu+`fzl^E+(iZm!{xd%1%3*#XW|l(I;FI
z<KOoe%806lOoG}cl=W5)fj&@1Yj(AMC`FbbWBAC$GiW;33}Myc=IAp(VMKgX8S1m1
z!c(Rsc&`R)D|esYajnCU+I!%Wi-NSM2OX4V#sge}cu<(K%a_ObVsRl)gYOZStaB5^
zjKiSeQ3W0rw!z2bEr&6DETAZgqouhA%DM<!>dzaZg4tZ#BjyTn`n{A`E>CUu+0D&*
zOnBXCuEb=GN=l2m2QeO#I&K4S`8ithY)w;e;B&fOC`~LY>=N)<n95lNNVn5{0e-Je
zwy!g#F`61gf}3C4-fr)@-lkm23Si66bq~6L1EaBi<z<%i&9-g{&?%oC`{6MMyX4xv
zwA6u56A<k(SYIkW+x#WrziJt_RraKhJ4Aj_?QJMOoEsTg^cE$<$`rI9A>CC7Y0JBW
z$#Kp57VaUQtVfq2uDwpzcbGq*#TIWyavjN;Q+@nrzHo-!7=JJ~an1Et4Ife1PbCiY
z7>imJJG_L6^!LsWM@LH_vnBn;MJFkY$~u+t%EGa7|6X<dZ9ZyHQt5s)ZDrl}=@cjw
z*GGR}vfl0oDfGJixOTw(PG(a+q4%`M?~vS^QD`7<11bB%GVE9ATEgzY2twSg1-RKl
zAqdCSN0dkrr-MzMy3#32p1;B_Gf)<G&1ow57p2DyU<c}SX&Eu5ry+C&9w5MF&~H`~
z3fpn<0-VauZ=<f}v)+$m4;8wN=c{h9HN%Q(Y9xH_!HBBIQAywrAJwb``Mg6}BUAM!
zLobcG884B^;ztBWi>NW{CH4;WECy{#_y*xevQZ+0Q*yd9DAccG<!6Dv3>2vLN`fe%
zDx@?^zDq00jN+8(^DZT2Neh?3md)U@=$w$C1v_@cqD=g=WJy1Yq}!XatX(1!#}hA|
z6tQ5T_)!`Q*Z_%#AL32aJ50tN8^k_?N=n?sp1MHE!fZ;eJo(#dEw8kVVXLatS;l*)
z9IGmM;y-@fKTxuysb^k){j5={O*`)TogYTyz8t+bl@@+uNoJ9&Dx1qi6$Fv#yaQX!
zn>|9NT_Mq9`)%Qm)CO6E!Nn|ulhGCZjB3#QwwT%EWQIi#0vyFfF|U5CBGc13My6)Q
zgnQ}L$&6ca+>9|rNX$a3S=#cd^5l=TVAk5C6IG&Dv(gV{|H3zo&EM^zm%8Q^N2kzN
zDsa6+uc;x<B_d?0{qe!rTmcK)!d{VJGP}=#Loi;hl+Ki`#r?z0Rm3)knyl4(L|FUH
ze0jPh2AjVk|0QMEabseaMj2`JXPP;+kcEa{UQl1zJ`T&^L9SbPk}A3h^_kE?E=Q(N
za7gql)XNk!HDr|^ueBVvMXVP{e&J3XF8y)Io|;~+Uihn=DLj0`H5!we@rrBMG=pnM
zgL#NP$$*xa%IB>E8~i>H5usW};RLvc<GIhgjpur`?!)!-g~%jav0qfZZc`zZNW5Q)
zEa((NL~{K6-d+GnI_=l5Up=T@^D8SWe_YCT&sNu8&fMTSe9zBT&zW@v?62_#;sAE!
zIOS)ac?Y}A_>p2X$wPaiMu_%e{ukYynWNA?8O_Y$KqP3xCE)yZ3Qzy(==Lcy!6&c!
zDp>D(wJDu!>7t2Wh|9-(*z$*Txm^!GmUa+di(edR*W9GRf*JgsxQ5=V#kBYC9L|Wx
zw9^R%1ys~O_gq}$hyvjRw0ZJoBF$5pU<Hb2*+tw^s~jp(-p2U(zL4rfyd`dSwGY4i
zH+kB{HC$@*y7Sxh3|EgEF)c`gNc)W)-IrU_yn&S4XO1}=f+-^W$1GgEii>jXbHn3&
zQKx|EMja3WQVuV2=Jqz=zdgwD$!ML0t6DVLt$SP_d=051!*zRId8ktKc{!-TG-c&z
zI~Eh)omJ$!*)e%>fvf8={VY8EHDvBg?~!Rowu|Qw8$#H_$G!^9VfzOfzK(%PtH&%(
zUpq>oQ(%szeg}yc%8dkjw%_s_z$?}O*Z%cC3Xx5`O6th)aX6|-zn1p=>pV1pYxfrt
z0OGi8;SUr<ACRF_%&J>Rr;1p$Y&?ug0>#1q6-0sbHPklU6o$Xfzu_f(hl$uibW$Gg
z>+S#wL%j|KCRL58Kw|ZePQQOOu$Nh-V$F~4cH!-W9}tQVn86x+Qg^6b&=*>RyN1b8
zqBPemTaPZ9vd4oaZqm7e(r%pnXo;A&`D$O4^%Z|(#E!|dA5j2dCl0lnMnXCJC%zeM
zM@?w$W=e0Y9A#Oo71pyb7}G_>mMkg9p!+cX6_Ct-^k2Jsgi8N}|42vDv$oe>Qze)F
z2YBt~EBP~fl`2ZtV+}xkCrQ=&U(!qXLz8+2eq>k(5~h<*1a_#&o_4SUOmk^aJ)wf=
zfA0yY>ue-THpnIFP!j2at2mW&^)D<THg!+!FSM09gFHJ9F*0$l5Ou0~8wxe<76-<e
zBk?1e{&ck7fN1^M?x<mG6a@v@vNS)AWXwpohi&tS{ky|V87H>iztls+rCDPYuBBLi
z?_g2n*Kz+?M*N^~%4<|uck~4W{emH+8`x(dgBF0`RVbeZ&o|dqSg65~FcR4tl4@)J
z+79cPm&zYenmqecv=LgSJBee4(Z~d%sV-7~9C8iUGPKX>{jC02nV@8@Sg^lG>$Z}u
zkx1wc79+&M_?{dk(z~w_E#4THo4q@Z1fjBueM5H2mUP=Xqm%GAheOO{JOK=hB?l=(
zrvGHl5os9e2VKvXocB(+;Y8%9dI3O)nA^Lr43m0O#+IbVQ2m?TawD<7xZjr*la8mU
z|KIdL;(t7q+va$$Vs5@lviIlt>rE99L(+709R2GHcF%yXk(AFpHkA74_~`azsll=*
zaC$PP8yh98^gmw~39Hj>`Qq`sk7w?dX~&w-#sS_rnO9Wmt&{(&&^-uv3pnjq^5ks=
zL3ln(kO&a12NNYKgL0~Tfjl`5YP&uY6GS0H8vM!T6@H(jlJ{8?G}iX*UsZ1#Cc|N`
zPo~}32oz<qUV*s?`bp5l_Dv=>Sw@Fbd5jaCS%`(`<;cyiTl%$sDeSHP6|C^pr6o^*
zyfoaA0F7mSK{E3=ZGz|%rp|uPvsrI9FL$65$I{O3<w@|hq<(pWB81zuY8tt}JMj;}
zc<hFcc+agWW1hSh<bRRc%Mo?h8nceWg_>}HYI!;77EYO5Idf@SF^HMxbh}*vcjQ!u
zN#x*OntsmUc%!VdJ5w-71?m|QQ`Pup?UKGilOyo>b#{32KiqH2*?Ro`F-aAf1roV3
zdq-=!oyLF176B}}{}cxX_AR&6b9k!&C@yB<gG2&X4{`mnoU-mI0tPMCd%WzxTq>G?
z^lJ32!X9`>n4`RYmfW{~8@=`QdA#*II%KXi2qKGF-F0luu-}Zba!5!{U@~Z4nAbqz
z1v$=_0LSyQSXh^n1;pg#eMir7*XlJi(ue1;K@w6PcZAHJViW*ycGnY*c6YL>s3C$x
za2HKPANW^-s(`S7jcmW=s}7r3Tfc0VtZ8$(&|STdD_yJu;%oNcanlk{bDv@sH+=ux
z=U-Kx@i;<|1~KECe=N@-d(vh_>7a%TQ0>KRVb_FMPS8ZC>XOWk!?oFs1DH10_Zi@_
z;yQIC;^%@KIg#^%Rs3GD){J9s!EUJcaA^&<V+)U`G^9T?tDyL}-q2LD2RtWss4^Xh
z-133@lZHe*?n5AGc2M$AJ9<2Or7<Og0L`cNfC}@uHH3&M=GHAyZvVEPQL!k*v3>)|
zQkyZ>MMmgKtSHV%L@f)5OSlMD&!u{0m2EBsBbssrE{^gax?8}5aFl@a+ln(SATHrh
z2`E<2E&oW*XK@|;BvDCg%>HRm9S}!RCkVxi`DSqUH8^J9I(g-?%?3E9U?Z^&9A<(h
zi;$~thY}z1{wNseP#sDKi_t|b)~RSK|81n?=Z0c+P)W5_{?X*(e1i7lQq0&(<&#~&
zl#Nkb@u8}iM5A87aJZs$eDN@P33_zkqMWghl*N~B`QFMgnjSe5EoeI@memB*OLd`K
z$VC6rKiW%~3K6O0AUur+Cv2ry2)p*h)O~h0Aw{4zWZL~5y9p$OK*i1XNFf5}!KgWr
zq#u4}k*ovcF9V6UGuj@<*L`8(OlG?dovay^Ut`BYVcr`zaMLSfKH6?uIGay;d>PYf
z$EF-`2C8G;4?#hP+w_cOH@z080`jabzc&GnF5nPmWqw4Hmj|;6(0B8AbNqCBqRry4
z225dYhiVcD;4yTir6*)NSe!3Xc=|6*dLpQ%uP;kIX)?jSxIB^5BbqozmPr&rt^S1}
zg$?Cn(&!Vf)K02_1dI>BKcgTNJ=JH7hu4T5ZS{}N40&TE4L)+_U1Eh#ijLcrG~)S>
zEw8cipuiK!B^<;1Y4s85su4MTeD@M5x8*4@whpcaU6vxaWk;G{oN(I?-Zb9)eBNGG
z+t}Q?3dC!Kn%0I)aJMNydJQdv@kHmF{@l&6WDJP~2Yv)D<A3Et;P3~g(_s4T0kaVQ
z5Y${c5*DtSQn9po3ljuR?1JaBdHdO~(0Cu08&9<WCz@QEW3R@9Iwp%ty4?7%Fl#z*
zq!h&fdRQP#IhSBKB)d{a?q;DTp1;-QbpDQ&fY}l7+Q2wCL%=yRb9x(zLHqgaMB6#_
z*5?WJ_V~&DttgX2o5&z*{qRa;WeqoFelOD<9$D6~d;126KWVF<Hx6H8ZqDjJN=E0Q
zfL^^!)2G&Bl!E}Xj5Vp4b(WBP^7Hrn6V#@pjMgM5ryYWabN0Qn+ND)&YK3p~=U~c7
zpFtvwRF!huJ1P;Ok0gQ1ykP7m^Uq|sfkgzeT;CyZY2y`Qq$(+<*eg81yX-2JJtF6Y
zV2su~lu;;4aOd_Bg(^ZK3srennLbb4^Y39KwL<#QVudbvUa+YZXHUhbB69prVvHSd
zw_&^MkX|=N63hM@Z?QXl!h6%xK;jdl5GH8h83V7G?ZzddK+<;=)BIM8yEx$}s@v(f
zhw2B&dmJ>)foP+%_64&@N{Iyp)9JCQz4UE9yGJ*`$`S4B3-iF9!-I5aRJ*{pku!T#
zO$J@z+B8>~f(#{mKG_&a_#~n`1&(^aEvDD%4IAnhZ}+h_|B25RDbjWw_Cu(X@J_H`
zQc6|2eadqBA}!TtwNnl7PPZKUY$@~49O&%9t|{qI-X(BcK9o4rQVMKI7`4Xs18MA3
z(OLdk2;78GqtwoEsP+Ex8of?~(z%6_tZXdxyMzHYr7y#5#A;}m!)l^=-qi}<7}zNK
zCI4-JJF|<1_m@8hz>UaZzmU&^y9zm!&He#yYvxpENAFcR1tG4uQ9~g>Mb>rAJvYb+
zd4_cdncX})Wwc7CWp6u1<NYzu`yd{DaMw#%3c$-_`Y4MQ{DCK^WNWA7o6Y>Vr^R0Y
zwrq7flOdecA03rMCgga^Nc93F%VU>eu`AHm4zRJl0Fn0sj|KoH%V4)GYIy)kRgKp3
zOn*d{-2y=GtN?DrRGF@ujw1M9LTCA!kCat0f~OPEqybRHp=L3azBhzMl~?oY71p>q
zPQ=F)uU^)PnoQ7u<8a{UH#DKHG*18C#CP<kR-mYbmdB9`*cSoudgjx)DGpqu5Y)Yl
z0f(IT5Cg;<g&aGnw0i=W%1#jd*3FcI{sQSUXdGs$I$QCUgsokk`Bo%vznFFdI0;`N
zI?^3uS9`X{x6acpnOYjxs$EYX!2~m-YvJd!?WbiB+JHb*c*ztp%L@E&`VOOK|7CBP
zz0K{A$KqCxYOCI9_Xr39?*k3z)O0(GAY6T|`!>wWL_ZgQP-1bY^~;6rE39m`=;!|i
z7h#=|79Kv*1D%y9f0bXP+F{+^QCR03<Lhe-uHe8un|t%;FmijKed;ZY8jXn0wuYOS
z39b4sQUD0A1N<5=d)&Q)OrE!JYh4YEyW37I8;B$TW+bo3xQOJ=?%~tWEBUXA(lJ|I
zYWYRO#<N;HvTbH(j@AvE>%PZ!3GQ{!Rq*%eD^Rhiuz?0WlURTM+B0qOSvW;`_n1zL
z`Gh1EcQ4&gwf&BIPfEtG?of{<K2X@<$YT&GQ4QVV?MDXFpZP)_qxt;}AB~nyo7fiv
z?jPtTGprip6xL8FVi%!V!$tV<fk3p-w7fJZ=9&`fp<NHuwBCosa~XEi-*>!bPa9|P
zJ5PWm%WT*p^BCkdtf?gPOQ3*~P+sFB2OD3X>DT*b;U8A(r*nPY6LO_Gn{p2I2vA(F
zmUSz-&*ETY?;4_ucv%_r=gePog1OG`EeH6W4SlWnPFGATMQ+iHH)w28S#F;Kb-V+a
znzA25;3oeY5w*7<JvP91NP>EMlP86yc}b2y0@wnl^+P$P(ftoe?1m5V=4{+V>KY@`
zduW3e(kED$&+li|zA@}9(k*9AbHVHs{CKC4rbM=7Omzb*0?zol1}dJlxomjRa3VkN
z)^#t>;@l#3rG8tB*N~@aj&Yh7CXrChE}HT@S-Lu~tqzcohJ940FJ@K!O*lsq8Zm$w
zDctK~j=Fui;|OF~PuFgtj-GPjQ6Xwy3e6yI$COkIQu8H~$k+|Wkg;Uq?=^$QOw4KT
z%RlAY4ag~|4MD}CUL(8~DJ|N`u_aG!RE7^H=KM(aK012taqO6IJXr|)Z`Fc+s+|jr
zj#k|Nk#vnwnSF0N*JRtaZDVS(YqBQWZmP+)&8f+@ZQHi}p5Ood(5f#zYp?U1z3+Wr
z_XTWV({{t;aCuDvW0&&kpGc5F0F4HyufUJnCHq#ZOIz8Ax{AsiP<0Vr+T}~u+RUeR
zA>oF)o=FXJiHqHV@JisS!m{@lhsYPC=NcIAEc+NRYrozfuT(2+MWl4wVaxt|<m|-5
z&%bqN4bEFCtEYyuqIyE331|`xGyOR}l87u@%qCg)Fi^wP4tQVjBLA2twc?2o65l@)
zpJd&!T93X#ga{Yq4J}|<y&W>Vw%~Wz=^X$r!i5OF?{(h`S2b5eXSflTMank55;;Kp
zE(i4o_W=jp>}?F+ODiw-4=!6@nl@RGF%_QVBIszFrdi=%MQ1KnLRq(~i5xkV0%{(M
zCn4AkYkPOEK?1Qt=g&8=B|3m26u^M38?(y!`CM)jTbvGAUac$xcp`}~*i+pt9(l87
zPVH}>c*_s#lWEtrpF8T$>l&D)U*7XhwF#W6Q^ly|{*u59dhCkl>4C2LeC{Rzm})g(
zlDIx+(YUicUYS-0kfqJrdB6{vpixlqo@=1^5jbc$?N-0{m3Nhve_HPy0)W~~o@1rr
z*s6W$%E@6;Fd0_<1}j$qzAtMm*|uBIl*Ps%H6uh|Tq>5SidB!J6DxP>1G;IZaWcIm
z_>tmuD8BsjJlrS95u85<5DzSG8Rm>xt$Wx~kN4|FMo`g}YPnsnhsmn6IniNToQeTp
z#kL*dd-2CEM_~tDJeHH4Hh4oHtFcrfjF<6d-Ru@I*^k%ls#5UKJqJo0%od=NXFQH^
zrB?<bvC>2M&7hU(p|rr#1D{gO1lW2;+<qccq&`Lh$|}kT1jYLcX$o{};N<d7lwnoI
zm0!JJY@+Wb<cIkrJ~D>fgb3~yRGNeVzz@JsP7eEv)fW_ag<1gB0)Q5oAPr%uj``4$
z^bvug$*afPwbLA;e+!l2D)WCVV0BehA7`uVnyo1XOXI&xyu^<V43o~Fh|QxlR4ECS
zK43GSgcglcrPOiQRrQeh1!y+wSF&bp%NdLiQS&5fV5UK5(j<&;NE*G8Dj`{a!*mlW
z80yB6KnR1(29OPJR%o&oK_c}0TQs(2I1JVJaf&$r^(zO!#0G`TC&pLpPwW^ddXDu$
zaTiKW(e-37O<Rr<)0)yCfx$15(DU@k?IB+H@F&nK6Qp$qQ-I@q91CFWjW~B~+5`5@
zYIGVC?pA|<V#yOz@*f{ClzBbh#+lpFt#ecxxIuHx_(1Y|HwJ3cn%ialN1IRE3Fh)n
z*W2k6Vq3wj>H#ng-txWs3$)WLzjl7|CLy@De*L~-0;<M7nhz15@4NWVj%bgYREh6)
z8tmEm6V~pq=&Lz6(MWi9>-+qaAwT)kr40jRbVudG&nQ)X7u19dX6(me?#7F*H^&>%
z$59eHwRWY3V$q<3`or1gF#CD~bron)PHf96Q|Tfv*wCT;-T_ykC!C3{$dL};i&mgT
zT}@Aystfn)RNDGD^D6SeE-}Df?tn$F4n_161SCMlQZXWJYFw)U!QP+9A=C<K{6&Dz
z1=^4TF<x+k)so-iN^(RBUp|Ysu^PtuSd*MTo~{iNgM?E2f{}<zj(IJ&D#}24vvcQI
z5WNiCQ;ejZLlle2f|UP~Bo(4=hq}|rcML5y%wqX8-qRNp^U3U>*UM;5-&&wd?m82f
zj2TNXVy}CAeFgz{?#)ggtO9Y*(i__P=O6!W^XWAg+=A$r4={N%i9ONCtwhQ36T;mJ
zgZDmhnPA-(pZ!;P>E2^Ucz^XA8N<=wL}KbruI@dsU3n`pyG!F-4+kxuKu*_|!$Gw`
zfe;5F5T@vCzpYvva9fNg#+&g<BMFEKzw$ut^U(58!fDJPIuXzP^1zYLtkkX9?OXK>
z0#fSUMShd869L_L)laXQy?6^1RDdA+bET}D{rfdEyjjK)(Y=KM7TI@hX1ZA^TEOO_
zU<FaVhPCr2?<q>|1EBXq8sOs{rYW*a($*!+p;aNsm3}7~lqo!*8Iup6kxMF>n#f{n
z0-iT_<^60k*3%VVZlXrylZ{rm$Dhp+nFD563pBHh{Bl>eRcUl;{9s&za_@PQW?BfW
z(IDAd+8cW<zF3Q7SZ?3wU)_W)<J^N!Rr*h58rOE~*DC=Zp(ZTdvnvj8Y)t5?B^D!z
zk~^-0A^lM`nqVNUam!?p17gI^<B)=}p>%wZ=r`(BCj&!fK{MG_&jtg~7%+m}$4BE4
zO0To_q>CU-jiN5jrP~u7nlQt%*)Sk$7|PWr9CL8V2vT!dH9#1rUJmaC;7LKi7?TPF
zh<p#-#*{-TK`Hm=b$eC^+^~4b7SCkJy&spAj8ilJBNqU`MKsaVu9s{I!E#gep{ENy
z?#E+>U3$9h$MkPtP}<hga`1S;w*^Gcm6<4ORPzDx2D22hlpXJmKz6A#Ha^^c`wN%Z
zVD`7n-4Vlb^G+RbU-QfgV5Ss)m7^8Ay)X_-tdtE_sfYh!k1wN->QCO}efOWXwsv4z
zIb9iMND8OiW=MLlA>j?-JoXCkLv}|qCB9MQ&15YR2=p7trw9}h?i$IHrU&?_Z2{G!
z2@@y=F#CjmV)MW5v=Wsky8K5z_JW9xCei>4{C|LeQ5NeSu5rD~PoIV(F()+g=^o(W
zahHDi_elx5aSP}fa0Y@Lp3E7rKAg-XgehkWI(-5F@o0>O3*!a(0CvJ(@GYYn&4A(z
z=VaLO<zEZ_@s`j<xO?|T$Nv~M<V`3zU1mYWh}0M$>J~`wgz_FN<TRi*=@AU=qwqwc
zye!HVoPloARCY_J*#M;lB4Oq9@%d;nT}4m%bu!Wo!mk^ML%_FfkMr?4J!7A0KZO&X
zx4bIr@Q)qL08dKrZ-{=P1T2=tr!gQh9jWsL)GJ@X68J#!B!sL)=ge`mWMNi~+(=<O
zhfi)D!^v6F15PIXhMh#gLmbzJEG6BCq9+ljZoMPHT>9P2V(|tACzQ&~&jP~5gw-`?
zXjK+CgXQeU@}>7F0yaz{OTVFEduuw}$mEkh$~OR76|3?+5KtTiOn=`<bv<;M&8XD2
zzLIJtJ*lv&1Fg<DO%FD*n7Uq2TRN4)4DCB1p|@Bin_7*qBa?STGg8DleJGY{qf-9Y
z{g4ho5S)?Nf?!wrIo{_tz-^DDFNPqR^#=K*Sf!C7Y2=WIA&eIvL1IstlH3}l3^k@?
zw%dTbjl7s8dz87Xz22kG`H<*;9;O+!UY?)As;c~w=YA<t-R>J3ZI7jhzemkynk#?7
zKGANW)F-Mz8cX01gPcTaPVAp`(-dXL*Jw%?NzmkBEmy7Hrp;+JQZ^3OG&JzSA8>m;
z@@vJTUI#dLi;_d7UC3FpB}=m$!m`0a`<mL+Ijip@BbYhz6fI$->Cx|~14JJar+4(Q
za;t@oRyQI6M|63w=mFFjO`}ySm1^C;@r)V2Mh@`MP7r02c<2PgLv%4qB4nqa`jjFj
zd_Q{!0EoA%?Ri6kN*-Vp_=Twab^29eC>GNGFqm;BsHNlZ-aKL1=5aTs(`5IbOC?~S
zzS#lvh#OYFezgEGB3qxYPZsQ+2eSu1`CRFLwwvre-vBaPlk@8;(*uLhE`Eoe_u&8V
z-u>$%&kjIbZ$gBi$BdjI!%ulH71J1CTFsIL$_lok0&w!^4gnjB$4?(<Y3TS21R3g!
zFpPbAOkz;bbqIpQ&%&&4SD^(rqbemu0(;l9jUJ%aOo)D>Ct<tah++QpuhL@mqm}7=
ziK*OUF}KtxeN1p|y$3EfoX^i_2ClHhiN#ZEmatX*d^<^1GCrL6Y0$v4QQFEBKJWtl
zKr~$-yELKG7^j3>?)dFH`98O`*Y&m1ATCwk%g7Elq9m^@WX8rolhVtLQ)zG?12bYv
zI%b<?gXY6OjDH9fh4yz2Z!Wxh3sxKe`mMY`w((o$3U5-RVHv&UU<BD0)OGfU>^6bL
zPCZLzo_+n*eVP*h193Y)JF&v;ucTKBw7SwU(FpkF_Z}fZilT6p7y6vL4*+AYPSz8Q
zRyGDJE*EUNKy}eEXhcF+k9h`0{(cY{9PKeJuT`381WLTm2T<$<iYdCPnYn<Jyp0oo
zKipLYgno0zWBBXrH!$kS@rT3}0$YY;=9V2suEN9wF|~l1E8%S?xeiJK9W9T?u%Xw9
z94>v&@VR_QVLfJDJJ?{-PLSMtxYe|9t>DlYWQ!vG?VoLr@_xpzz)8z;MsBuL@s~@t
zRo7p*W?5Ie6KcGcXG?%$4G-1`O93~Oeinp{kEb6=V=?Ce#ub<X<RG4JeOan;%*1RX
zaY|QPKxfXCSiGPV_l0nk7of~f`uw3<0Gz|VhvpDfyIZ~Lq?-R{wz5R2An5xJ5nW3<
z_LnSH;%Y(FRM(0YQz6{r#Ypu#uD?*FPYLxJ6+`&<Mx(}5BPC4_dQg;~nkAfkiml|h
zY?&BNwa*GCr3T!mNWb^ux7QXbmVh%DFKQsks9apFJ@#8zJsPNxpjHHcxhSE^7N*r2
z!iaZ6-AntG#RhoIE(&G)+o#tfl~5#D1$kphAi7<?2UA2Oo2>2Z-V}Prs{GQU7F{GX
zQz&MueNyDb+_(m#=W;qL6pNjbSR6iL%`Z?ktYuL)Q9l6uEJe-l02@UFDIgobjK!V!
z7dpq@azLycn};560-S`AMwL1Nzq9*Ct4p1^D08L^921}{U-<i$8b@|Cp4sx5&~P-B
zRi|y!^|f&Y5b8Xyw?Y30s$OlhZLP1rhYG#tbv8Z%0tz4sOBc}E0!khttD&pu#pk__
zw@4hGSHN{0APYCTj5Y2JDZCRscE0@x&$oJ0SP7;G%Z*2OBE}(nU@0KZ&ljXcO{~l{
z08?d+PlAV&)i$Ggbbk7sow>)DONxymAq**U-zL9gYXdkDxlF(cUcGhbE`8d!ww|iG
z{Ft)dpjtP7%JwlqfoA2(NbCCh)`uhUpiOfe|Ld?+n!cvMP$Q;lqjlbOwZ(~xJX9ml
z2{D@q57wbTfhx;}$pf&nfI2RFYjeF~>Gi#b^DiY`6<QgMEz%@iaLg4L6~I*Zw14MD
z!_d1Q{KDPuJXW~n!tnLjr}hxEsORCdTlqYnGrsJ4^34%~30wWlg~G|0p-Cv2Uzp$Y
zyYeX&niN=5@<9Go&!fv22FplMvi(;2@Li&Pj*3WrIX&~eUaUlFm&oY;*a~j><w>M%
zMo7#SC7EGox(`Wky+Or%OxQ}63@kkk9&E@e7b0{_7BjWza#yyg@J~0D0EsS)D^wr9
z!ucaXoBCg{Y>%dSd5fpdx4SmiQ~q29{j7R;Qqd34AMp!c{9n5hOIzH=uV=}d()Zkf
zD639KKGO{<M<tSi!CV1Usw#T2>h8<4WEIDtA3eylh7%c0sO?&)5b@>YS{N^KH^_%G
zTYE`{4SA&+ZrHn%gff^O)%Jc`H$0hy-oI#?DHi7C7WY->b1|dK#RngL-=O8TZpv0`
zS`HGF|3FznFubs97%Q+K85TNPL*WJ4p${$n+zy%0X$4m5JRRG?O9Y`hOL;22XTA^j
zPGruX+`mc9fxCr@?!Q(!0W*LThz$#-*xkOsZu}BqIj`R3^t907`Eb5y92X#j>S7BA
z^MpHe>(>zZp%l?~(xasMmXYZ@tbs*4y=^b!o}1=qQkunUy82^j%YU+LN#ARulFKZZ
zRLNVh;(KmyQk>K5c#fFm;ts-_uB>ZNymYHo*uA83g7ODXN{x7{B7G9CxX_@93UmPo
zml-#DSd`C8k`$qLYFPTUat~5<P7I#wg$PE{&0p_M0cv!vT1bNIJbIOR5)Jxzk&<(z
zq=TMW9FWb@OxC1MMpk@ppCZQ^Rm}dX#7PT&EkS!^jr#n{Tc&~=ld%9X(yMGGg-%7*
zVhFr8B?d)QxsrzJ(_jeNf1I&7ld}zHWi-=8lS_I#P*UiZ<CC%}4EoLXaWeriyhErm
z1vqqNenY9N_GFDx$7e^7KekDb<2Jqh^QoLY+W!;sUcTMw;qYC~gc7y9gk(@h-?yf(
z0xsrAc>K>VEXJFuEPDwJXSh!ncE4;HlfDB~?w9wwDZu*E{&fE2KXJ!VLP;mUvlS+8
z1`0UR;e?mXu9X5m=lkFpRDL%!YUt9!{Nvf&(_~(sq#%pYI^Eu8CWqI{ZH(K+>V-X6
z5|kb@X8Z{IXQVyZ3k}v7#zC8X4go^jX6A!g&AKl8h<Vwz^ir4SFk01`2=?clCh37a
zcbWMN88L0SM*$?gUj6=R=Lhl;zU9av9QhC9ymdJWD{qhe;~c-w7p{Q9gX^(ayhdu&
z=uSc7!*DgLRa!inPlM$}6RDg_PSU#8nZwep7CgSH7b!ZTC+2>nrbDYo?HWHVLWk$N
zAn5EicNDi8WQ!*qqC9$qP=vOp4X34zrcGx!qjyd|W1bC6IY$d(zoQz*oUQbPakHkX
zh!FH!mCKvL!C@!w2y|;#EK~cHw(7Ec`8GsuDM6Neub(O3_6Es5uMMaxU)fWnI!{9I
zLJ)95EOKhJt7*S02^#iRY@D^y>(-;k9Ka+KF%q8Lss>2-2u_?Nt9I#8=|;Gl5)D=6
zma3WjXcO2h_s&jcMnc&Ol3>FBW^r+^`rN}4NUovu0uR<=cq3}^%-JSe?MwhCg+}Dt
z99su<nmnc9l4ivgYH!uSqVDeu{k}#@BwjF4W6+bYW7ldrOc3-CLpENMri>KLG(iaM
z#tbVTgCExKqgH#s!Lv>}DB{R_{b33C81=mH#(v1G94!gz)tv5OR18ZNz<0h1!?0zp
zrDu|6rbr4f#*E8lyLn+Spu40Gg*8(I%b*!#f;G`=x1Yy_5}xQS6;%~KhF5K5`@Cw!
zxvP3N&rkg+j=u5EQ)Nano79Qp+|l|6LKDZU%IU<ilWhhjlTNT;oXudoR72~#>&9u9
z9@}o%cgWa~1%=o=x#MhTgP-@s;RE&9g-MiN=I%Zz><#C`7{+e@cW=-#I5`>oz0fpT
zRD2eS1j)$otFYXt3~*mn!f@z}3&amLsn@tM;oV-c?Yt=thQq>pOKl+b*QQtFuA1OY
zxVRB7Utd@LE2NJk$|&;aLP17QwnB~NyvnBa0Oxtjmc|e1hKBsp8rQR%M0V;(R&~5x
z`jmS*)%0$~JHQV5o75g+iuuQD6r7+ow?l6~N=!`iLvaQm0QkN$4lKXwjB7g%n#j|p
zw?CMOVO>d|2sT*g?fnjo%2&|icK&3p#1HtrKLRM`vtK`#kpN(ou*qv8o#U$G?($hf
z&*M1&y6R#r=G#Eqyz8mk14oc;*ItjG(DnGSd9hqJuiIri22u6z>i}V4r4MREY7K}I
zftS2O|F6i7G)1F-nw)i-MXWsUy~k>UYjwuRetOiXYaQlWy}=PVM0e{hAFq2p$8&F|
zY88vy9N0mMo9-V`z^b$TE<M(%0H>YVivbuVNiGQ)O<NxO&v&Bae=k?g&qu^h8u~sD
zZkV1mgL=*OJIgi3g^h4njO$NK+ie;5t0vXwRC^Lw^qMACj@y3YiNs_4i&kU}sDzz%
zU0HsgcTNGN6PLS0H1s>{ey;v=P^ZT2?&QOQ!v_VpWwNr<i1n6?A+o>c$xv-CZmC^M
zLiaPtI1IG*-AA`PD!*T=GW?-}YzWi1M&;jI&9aTrD29DMw^$C6=5u8~M-0jDQJlwu
zr|ImDrx=6Y5IjzsncDtg0L!Zir|2@7Cg4*0;ks3<MKr$&ak5#A0a{u^$VA4B*DL9b
zY*d`GWQ!!8ViLjbsNB+#ZmKw`ql~0t3npmqE?Jt784l6FEM&TI5LW^A1%#IH`xE^|
zemT03Z_+pKYsA;H`NL-&r%BzguePH$Ut#mDb90{(ZZa_Re07eR4m4CY(E43b^%0WL
z@BAi3&->S2u}w3Q%D?wt9MP7Jp7V|AV2CABbNR`UNyw8IrMzu=F>|gYXN08M1@KOT
z4Lg|Q{4qN8kfCFEuu^?-siq3X&>a>yOoOogKFlWX3RD%oX6PLbu;miU-oQx~<z=ed
zWmP2t-R5=KZUrE-i3Lgg-6UF10y<u;%4xbI-KXLMI=D$04gP0kFlvd+=G9sX#$Jtz
z{Sc!2KT(XFdW7%gs{B<NZqNrD@1)=`VdE2IYhTaVrt~(>$}Q++9MlbiMMe3wXt}Ut
z8OEg==lwf8V)L?^J`-hv$tR@_)@e#5U5iclnJl4USy&{U(QhIeeC~GwtFrjEn6EvH
z!O`5rz$B8BrwNu+iR#8oT03sFl%|F`&VE@-?W~)1S{VDy-wsA{z5bjVqA}|f5eqN6
zjW3ct8u#T(8r2RV^xxWyb&X>RB<OD@&)S=~e~>nHeD=Z(!cravQKjZMa`|H3wlA^j
z3z;DD{bz7nNTU#i>vA-0=u_Q_yQe^Zf4=Ag9EXxo*gOXfBeIwbRa<)1!;XN-=gpo_
zw&2HE%iWPemY~n$Igpt$Nhsv|z8xSYsJ2mO^#lMU?Sek9PE-65vy(Kt&Xw&|#f`wp
zfA;#&^^vS|*><!^6(aQ6n|eg0kj`&)SoyS-(PnWI{{xT9X1&=1kVOALOp%Izpl{`7
zW$9N*^Uhn}w;w6bS?y3{CzK_(`GMHB(P4JcN+Fv(k@hk9+hGqVExXY=b{+Xv62F|=
zv5gnLxqGd3l2OZ;PprD+Nk@LJo_brG@JH*pR||OYCyeC?`DKg$!P58)C4O?-@?Gmv
z$Z=qb>1eK~R?ir^xLnu^?`hhQN2E{E+04GXoTr(iZX(Pf&6s+aC~JS1k&3YGI#_Ht
zTU7bFy(yYwNR#wEos<!box7{>kW&pmb!+Y3I=i9O8z?CS&C$Put_LVj^ZvIC_+i1l
zEDY!pGR{4U5DTUV5Tc#_#jhOt!gp^$dJ9KMA9@2+Bjr-+6fzI)py2*ovwH}^#*9t<
zA9qlJBq^6{_wJ@L>f~BU=s6k>Bjx&!kgEFKSkr;p!$i~XJ+b<5&<_}-^S@I{f{o)W
zz%a|83tquV7D{L=h2wo}Wn5c3sQK0U-7aon@|b+D!9MX{h|n%6Q|~IV&M?FIu^_JB
zg+pGm+E@laOk&=^fH;C<<+WeO+a%iDxbt*KWPT@f#zRq_w>0N-m{b~QPbVUBWooAG
z?K$}r%@4P$dD86YQE-lKK^y^AaD7(7qG}#3W}_RZ{W;JsQUYaLC&|yQT(xYX7Q4@|
zQKp^Uj4|-P?X>}kQ-Tku7YE6u+KTS=p73~!o=3#qQGZq(@n=*wI0lDs$eG<%4IF{E
z;7k9#n~q;*xlH>7FPBg@7MyKNN&@kPy~hB9q9&F500I-Ucu**rcj_uA9~iSyb2~Mg
zK9`Hcz!t&)(ja2Hb?0`R2?M^w*%KsvJj%Fg*(b5y=BCdNyRMJTSdpNa8(ZNm-?@?>
zQkin|hA7Vgb6O5x%=w?R$vVghY>*X^qZ8V>)a*MP6DgHe{)TFh0<T#5xWT%47LjBU
zkjXUh_IsM;6bM7sxug+iP++H3!$ng>kyI+k*)|Q<#hkUX?(Co=eY?3SEoQ^hoHvJY
zaMLoYw1GJ@8}6g6f0|Y_JDMFfrN<>Ffw(Uj`T*OD%{n(=XSR5Ui4`v#g3!mW#%6*N
zx~MRRy(%75?g)g^>!-&|l<H31+8`kXaNFFuBRe*0C{>>=bzP|~v7fJs8JCZqgE*>E
zqvE+UBpUhn48qrDksXC!$k|Ck$Rn5}6HWaCJ9KIx-P*{Sz?yCw^|<r-+y+w9Qy718
z-bHDBd3*+UYW#HB>kg1iasT<lCS_HA>G8tt3*h1c8~3oC!|Na%1A))Yvd2JHb}UBd
z)BSK<i`w{ctlE4k`{SW$>%Zy^#PviC>;eoUoi^7CVCYdSpUr!CVlD|!<a@N82YjA6
zR1gaxWiW8&>8}(F0qu+aMul6`<j+>u^0SS*@6IcAUNN>=XQ`jA%HT9z%!e<p1bi-x
zApdxj-(vKr1guMT*Ub0%?F8-ruxE8+g-*N5(1m`8#d_80omo7Ls*x{t6yBaDRp@!V
z93&rNIwpHFa_v9z4aWFDhMs>t&ncK~YKM-E@%*`#Cy7NJIRsx@TXx>?2mNWuXZ**A
zAL2TzZ}9#X_4B!!iRua;cK9Hwi0~N7h7y!@x1gV*`bT^r?|QfWWJ%3|Q(eJ=wr3i7
zh4xk;qVOhV)4^_dW(fvHxBncO46AkGDcoxy$#0YUj=+mq(Ku=LOC*bM+{E_sIfl3!
z#AjpF@rmSxaQkoknWzvVf{F)P121F~<w!8TWm>5z<T2=V;I{oD?Fem;w(IQ}O=OM*
zK+r`?Rk7=2rdyykIc`Kj-EEaX6w~L?fR^24vF@>ZWnsN?<sN1<@hxud4riN6P5y2l
z+KKPg<5d|sFu&)<4o`2f{7}y$(q!}2Q-k_;*YzM>w?Xp)uFZ_D<Ren2kL`${X<I%>
z)@r4&_w2AIu}DTI$bc`BZLalTJRNv+0Jlp-w-}@^F;DvT?#ar}^e*x)FjPlv$p_LG
zwRVbU*1_Yo<=ZyLV3k*T8p^afx?q*UV0F+{|E|_w{%_k+CV5yZo}?}-lIa-Rhu6UC
zt~$ha0jz@VrJ(6==T3Ip<w~OuwBEmFVAsZ|8j@N&HUr5Prr`};ue)pi$|=p+vzb4g
z(6#J|(G6=qsx=$T`q9XyUuvaNstTk8f`c+=l$a8jVpn<x?{c&vCLcC>Llw-7?;v@r
zFdoJ+!-uKNIH#qJ`cKkEFr`CmuJ^I9-6IP?H*O*8f%~_upfI65gVYY9NnV)=RD3&;
z=XMWJqQ<H(LqbZNJZ6~#aFsBu#<0y;uL+r;;2>>9SjQC2JKELxCi^XW0uw+PO=}0J
zvBFJ@lR~@93ZW^Iq1n07;?wI|0=15rPN8agekX<EM}TdQVA7k{@=RLwv0D<?T8&k!
z)y&t{8BTpt@}*a3(!xoyJiiB+XM6IZN9CisxtHOKa4oXK;SQtH5L5QAa*toMwzYLo
z5e7Dc532^sG1|ST)BkL3m$fO;6Y%sjT_79LX|&aEw-^ED<b$2;uFnNPqBP-Zmx53@
z2Z7j%p2j`C^O+4<h>+mrL5<yHW&`^B0-nv=(^zf%h)EZ}<NosvkwWJ4%TWRqkO<fW
z!~$LbyZQ}?h6hN-qT0*0O<cgG^7`a8CWYY*NSM;9?z&n1t)Y*7x2{5#<9(kOO7!{!
z?uJ0XdAzu0*K$*=W7~cbat4KnZ+`(e;QV2P-eNhfA`!c4JJRnX@;PrYA8LPoha)5M
z+npp4t3DI**FMc%v(0!Jq>mw20R`Jv-$S4W?AZ0+^YzYbrTQ361jJvy=B+{!k8UMU
z+4P)WhyvGvm1@xA_6a}j3J%s74<QnA3!mUSz7e=^G(4;|s$A?JyxjRQ*9}a6{gU@&
zhz2+LlsXFPzIfm(kxv`MmwuIb<l@P)y+$qFK7l094JONjNu*JCL8{YHK8H#-8uPaw
z11~M&TE=xx)xp13(K*Z)qhX>e&P2#))Fj+S&SMl3#`+t6GFP{fB@bYow4J$nDgvm{
zAl0nMCp}*bKh)oCWs2Ri#p9QiVoQkaYv1Rm$(jZ9M3!>6b+Dauu8)>2oUE(2T1-xE
zwb-`e*TA~6Y=6xaGYmJ<dhF~+{ROdF^uFB*g*$V*#Ekh<1)IteHaffz&w2#css9uZ
zq*RO$pwSR%DeJ8y9j{xcB>8E-Uo+o}B3s3ph+;F}XT1V^gI610<rQTFLbm*M;B`t$
z;K&|?(b{0r{JPgkgC9FhGwsOrI$+q=g)&#&^t8{X!N*&GFFq~X<NmM`LC>0JCd!My
zcgsX~UYMWTKhY8$%y$bw8k=}H3%+sH_0$Q#CrV_Jh5~_smCxLZmNAtUl)ev$-g*`?
z&s)jlbRwk+jHl#7;1h&4IzD~BZ>mWH*MWaA3<~wFH^{?`#tgTfCHB5$L(6?s#`ANb
z)YvA*EP#?_JeDg?@`}U-%3~x)K&jy1SYkGN_JlThE-o|}TZ?@1fA5%ofhcqWz8`^#
zSd}zZ`fBUnPMxDxptNvZ+8H7CM7<cL1yv{wvxiw;3q!cmhb`JCbMr8x%BDhIsBAc=
za9jvXra`>p4fUQUB|Z;!uttkA)QCGSj%$6q-4A@*^r+#&kq}qO80_btpO<Tvhssy^
zP(t95&*TYio4y7(!+YUTuu5<|$ic4>nt}IIt(jvO3=T}$STBQj1E>u@o53+_TS!fX
z4Tq}Ea<~a4zLoMn1JpJwM&14wanYUSPs+E~2RtFYMx)7do;^LBul_xS_p?hc+s>Wh
z><upi2wQXjuf?4%xguy<cb&>|6L_8OV~X_feVzYxHSXc{x!wXN_-_9&x}d{lx6x2%
z_|+-dwsC3CZK8z$1y-J49-dEIPD&SE*__Gfr7~9)5^}M2wp6Ltv0VTO?0z2L5CCv1
zZ@{YGr`q<%QBU2|iO+&Dk`Tbd<BN*oeC{+M(HMC4IH~`Gv7Ydv%#%HiZA?Vj@u>aO
z%6QfBCdf%y%<7!=$EfUm1uIl0)_vx{57ANozKAT?O8swDMVSvElBNC_5#OE86MKgV
z(CxH%jN!y)XK0(<zqdye2p@G~i!bX1I=pzb9Q`1BTl?=~YE%nEvF(Cw2|h;KZP_5d
zosc4uN5<a=xK{B!A33~i+^HClSdv?gnd3sG9ycwVyjy+ek~X{|2{~}5tSFx$-~0BD
zjEbe6#xtOzfdtzWsbq|t(@<$}e2Z<Q-{>6M$tJTtzl*npVH^<xKW#?MWi4|}bo(<b
zx%6m_21br1-*RPbj@~X?u&kD6{vCZh0Syl?LBp}LZKp6Z5Z=WEV0(g(!^VkmMnjVu
z?({T&=3NIr81(tq6TXz<H)7!plqEg34P_2rRPk&O6*xwA0(S`AWwxW(-`Q8r83&Lm
zH9J9m{2V-wX<!M~s@rUJI%clBhS)<EaO%v86N&i))#(sDw>BGXoGD>1ozQm?dXM(U
z>k7mdO^&5|%<HSe14-a|yOv@MTzgrepJTfaemwzoT~8u*C>sAb33RmF^8(&M_2iM3
zkY$qJk#Xd#ZUvv|>xg2^oJ5La6~A16&-?x&Yv%{zOAv{5%w9(M8g%0TR~}|*oY}vD
z<#R>bsJ#U#{<2-1&$q|5Eg!qfEV0l$G{Y~eSGhF)frA@mbN#U~_KHVkQpk6`WR|JH
zYw5a^cwW^(SO&V-v=h+xG^P^_@4qQhG%FoVHRes10LV=6Ek@-;UC*k#?xxm)8ZGMe
z+^%*i=Sze(;UJY(+Qf7|QYZeT9%U(QB0?%IFxzCalP!2uRNO>?4M(b6!zxC&XHA#^
zv8C)>bk{6F-qS`Ic>C(Myd%QTIL{Os3$motI{UQ_rX>yA3sd5%{blyiH7ECJGxy7d
zhAV%@yq#57&@h`B&B!t4Odb~h$YLC!ErT92sIc5qjX7?{5`w(!Ll2)|2n@q+1uf!e
z%lLvdy^XZLcPEu#={pFg5d2A$O7(pLeLh9)E8FL7$!@vc<fgXis9nJHHW3K3<h0yR
z`09N61me1o&V_QHFa&6h*g}T<RE}qg&bmLBHhG+v>lB3ej0exo1YhPa4VdW<;UtPT
zkYk-XIiRTKC>0B#;xvs0M%6m1k9n*P=~+YglXKzFeErU6ldC`V!CR(sgt`hkKOau!
zVl);N;C`)^qL8W6>&uJSj*S@)`7*yTO+qJU_UWT3f@z4)wXEx{V8K9$X>$Xi1Gypp
zgtp_BG3s7eYLcVT1y7<kXl^Ptu(9mciDai4mCMuX0^B8yCrLVke>+r1R#cQ_@&3h_
zS8}^48fKY1h!@Z73(Pi*=9|B?_cD>Vfp%8S7Vr5I>^fA#cGf>eAtn%}7g}_`?9UK}
zjYYe}*sX-hX<6%S4`iQ8`;ecy+?e<d9araf-4j^D6qEIEXl*bBb&fnMirnn*qk6NW
z`dq?PXSg=UPLz}bTCA9pwH7a)Cnv6^Zf4RHy4Ye{^a)Z7{nnxNpgb2n)MTa^wX>wD
z!4K{CqSekHA#6u(fn*`vdHQ1pR4FZ685RT*P*@OZ9!WLps?!9~F|-#<6(rQ@-d92K
zvTL()Jdcw*Xc}2$j})3U=k-5!Jek>mF)e_m0v(UQYuSw1+Hy^=ajc#^x82Nqvf(l+
zcjg)bse74s&!w~RDp-y_oCQk3Kx`QaGW@s4{=O{2Z@WPL`8p}1a3%CoJ}q59E+&g_
zC1=}!a2-db*4>EHl%ShL)8!vIi}_(CoyJABsr5n0x>Wmlw1H~Yy`Kg4jFrRk?6h2s
z;%H??Ek%m-H^w1R;0|!#tVsBpX6wS+Sb~*XJT6)l_~VV*Bjn&&FcH+$WcJGB5P3K?
zs<rUK`wd6~MFEqSI!GPR*t67?&@F6*UO~8B?}Bm*{uS@-Sgqy88le@`C@R?rhGi?M
z&D}BJNN2#m!0zD&c~EwOPW@7jE&mCH>6Q?a-C`zyEm-Hga$Di`P*okuulGrGE@SFp
zL!i!b>l>H-LgcNfp|bXsE@;wY=}ebBYU1Q9C_$OQ#;GjZn~?w=p~eTtwGh=lz8`UL
z2yOzWB01^dO_wdV7mKqFllIb@Wqp_YSF(<nko2~<#OXzjQ=xMW`TOGja%0EmoB~&+
zUY+a2Y6IX{mCgpR;O)lKIbB}&WfDr{()llwZjAMD`5udEs@wTp0a2B7HuuiLNCKro
z4(G>&mfghHb{_x=*{rl)>>bbn_^X=uZ}}>rEF1GZv5V$htU2evN)l@E;CU5lwSLkS
zf=lvwrrf=kn<I=c3V<YU^^NdLK2#R2g~^Eg1sI@s0PiJmZ;W5Nb0fr#5^%f3KQbBL
zu`~)%BjY{b#K%CeFkFC+{$brE_0{z<uKaFXG)8upmr^7<#Co`hTBjvyy76TPdc^G&
zp}><o71*8#he_WnmKvz>5SIBxzta(ov^J*8b|e}_iWw@I(MQLIX;uda)zluKdxV2(
zZyM~W@gLkV%*shx;KyR1+#2DRqyz?h4x<z``I<X?sj>sfm^Z$+J+f~)9c}njcbbhY
zUVIWKmLo#+g$w;NIyh|$-FN`1loNWsP9VB-X&s8F(|v{D*{dO;41$GlR_<xxH+-9U
zhQpX|3!&}kFM6>bLE7R<+8tY@`Vc*+&hD^HLv$9M4$U2)2;XyqMe~8xV16QzT}r~v
zOL}1BVCKgOtq?h0eWVX##uZq*Qr)ewdp)jc07GNvX=EWdKTkD%XxJDUUlmJU>dU`W
zEwNT&wdWPWCxeLNm?b*-mE?*;h~r!+ol4ExZz$;mu<>X6e*~XT;B-D0cR2W*oj5!9
zVE{%toRouL``CcH@T9#rT;>sPFK|D8TL`O0JYLhTHv<qV>eWpC3#Hm-ygEht1lx+P
zy#hvbk{dk7m^w_H-g!$GA`~BHiOwI?2U)u<)Q|`eo)=YSw6~4mGJpxfB2dFO*{?kp
zs^i|wyh-I{DfKOEd4^BATY??qL0ePGIU}~c&3^mY8-SVfdczGiqGXufL7pG?5pV})
zZ%o5d1{m-st#p_(eFxo+6m7EKZIX7N|E^O5Prp8>s#A{Vf)hn`h09wnm0}AFCQEG7
zeipN8?Ptk@Q5PAq@z7$2%Vo14G-yyIOe*w8rcO4ZX(fmuGk`su?jn*4+rY(}dneiZ
zktdE`|I8g!_PI8#Qj{~T1<6Lh1YHN5Yd@+w57Re0xNKGf@d}_)+&c|+P@CAPdYIrj
zVFnWhuX45arAgCpINf1Jjr!>N!LYGbI~7l?quJ;7@2uI3oW?}<LgOb7P<~R(Mk2S(
zTqZCHJ6t()em%`l==V|lUawtxy9P({^=LjFc9%-{*nuX@<n(}TaF$9CK#9>uClja5
z8u*f;2?)~5w8gpcvLb<jM1mkO!}bE>=^6PCS_1;9%fg0bWz=7w1qJEr@oha)8^wE*
zRgM|C;S>r?o4LD{7R)_P<1wI2D4tZN5?v#<K`CJhRXI+x#aaPUCJMG^7NJTE(To<;
zgsEYwuu~l4_mVNX3g7;PNLuN#&s-@mc0|)6CLGiB1C<D`$h7C*WkjcGZ(s>jKmJyf
zNnphL_#&v7Ol3UnKoc=8t8e)ao;utT784p9t7$Ec8yQPxm6~?hdQMvv23Rc-!QXX6
z2<p3FD|qM3c$PAIC4yE37lh;H$^}Gm4gqMP+Q4<++U29VL|PTwerQ0}%#qX>^O{U;
z7TDB216vv!gmPz}QqH8eB;pfv`5Qel3iuJ1$CKWl>A5iS^&SK47hkGKKbL`_ibx3j
z{qL!@Qwm3L%xyrGX29=hzb_{l^XCb-&g63gg63qI*(vM{SN0&^f>{j0>n_|YmF%0z
zVF&rvg%@mbGC*c-jb?cF3(NKAMB@<Q#f7fsm7gw}x}>Bs6kKE@Eu3Z}0;dwaNfJWY
zuMpYlaYklu2T()D!gV7#%ofS4qYynx6EY|Nl(b74eftx@)Q}rg3j>PXB2RwRv?3+I
za|%r4Fda~FL!|)mQ}Kgh_T<4@Pf!vUM%i~$$Nf7q$L(R$Nh53a_#zQQS{7T^)FlJS
znF^<yN)ihM95L4*%xFEK?bQJAsyg_UVU1RVnZT%+K?@Qv<Wg=Y$)nMUDXi~=)d1F!
zqIF}=hH#J!E(u|=sP(kjq#irDfBU{{{jp$rU6E0fegsDmofja!%A3g<{OW!)qFUS%
ze1w4{VZ>(E?rYQaDZ=nua$r~HkV%+7iN))}f!M9nLPSwx{s$B``JM|YNd_IZAERS%
z{u~2>pK~)VOzfX@{D^s6p?-2Y5R$(s&|H^wllw?pqwcIxG=#M;fyEN)-tDLgDWium
z_#OL2l}s4>f`E`W3LM(CqGHU{3Ei=nvFq!Sk9N;f%Et!)MHON^!%sbG7`t8``EIwE
zY1;T%C&ir3k`CBA*lUDxTRLcmgceN@n>cM(7b}J^1lrt&hyyZFD3WgKCb`u?6c_f{
zc`1fz)(MGesdT7Uc5aRM#WqA!x+#mE)0*rHim(V~xc5HgKxf+z!iT+$n<ztd_ABjb
ze(^#j`@laNX2yv!4I-8fu$~gS)Uo^J+9r!Ym*Ux<kN85yWA#!iZxBtV4va${1;c>V
zD32JJpdcPSjxH!BHaP^@ZEAO!O;8=nko0XGJiIgj7)=8(l8#z3UrEnoO!?ib_IS~h
zF>tfHLpaik$bWfm?@=-8Z#wv|nL<K=HcZ(JMZipNUXIvk91tUd5#@042z>e%_puGj
z=5X&cw1sTYWVioa@~bUgElx%!R?eDXNRbtVgqhfN<MrQ#2wl!XEvG4M)8zAS+92(Z
z+DEkB=tQIF1yb!<R%~gY@0ItHH--`8&m+Xq7Qn_{?u5<(uF_Su#-+uwn5FS3dSA<M
z%wMILF*2uA6yjU~7*5<hVXcR|FGHNrTuccBUfo!OqF_?e5phi<qMd5JI@|p}aj9$X
zWl9CiO)Yw7n&^NYlnC`!g8E#Qex}NxJmUscx0u59V#_*Pi6vt`R?f6P;78taB<Rt;
zjLPfNi50vFMN-Ddh)HY90kH(J_2SAhVx$VMwCvF>HaT`1i;klwF^C6U1`~7Kw-x9S
zv)|?`qH%k^`!SGTmYCxVgD3W~jQZ?k$C3d<lKMo1ItC7&zz7y@s$=>sJiAjEGJ%IM
z7y=b1zKM0)m?@SaF;kia`znuIGb1G%$hwVRLExpe<(Q83Np)m|ZYxZot8-;FMz}F4
zZ#gjNs}lILf!mbvja-X2r;8!kz>hAft-R#-^<pJ-XwX@wBC~0{6dV*(x+a*!Y7<qF
zhBmvR0)*|z(D;F@$*w19(gBv1#V?-ju;Lk^KacvNr)4B%f_<Fq>`nLvTjhm`eS?TF
z<>nn41CDGQC&x+!XO@;Fa@pZIul->;Oqk>Bb53Q%PXIocG+_d`));Z#0Drq&uJiQM
z$@kpq`@K9I<uENI<Osa+J<9I+mBylCPAi{Sv!}yb7=Ftvz4!C<MW;TBZ2NWB`1a*;
z1w8H>=fv)qOlL0)V_Mv@p4-Hcn?Yg$Ut8YH=pOoB5t7TSh*KALvr&N%qR<#NluA^d
z&9H9S;Yv&cj9V>~l7ayzJ&<?jzzmQoUy0H|>{yI6rm3ArTG;)lGu<baF3d-FHCeyr
zsS=i$mVFiWRs1`l=TZm1gd$G%@B9JyMevGFN-1?m?QDlDEEVd1>-c^K7;qzlo%a0I
zX?ptqv8e0ShQfp0&!CEM(ZwH#O~h&Z%8W*dpz&-N_}iJDW8+C0XMTr%VWl*jcAJ}j
zoqpzQY5=#=tX5OkZ>F{5V?S7=US15>e#nb|N!D84`C+8h{mBTwzRtybgzf`n2xR#C
zoSX*YG0B_Pqy=r(rg7tUy%Sg>C1m5)tc1_~nta1R02&p=lcusgXs|||gY<p$n_&9i
z5tjYv9wOg*<TDhMOk4X?ZCsP6a`I|gR5S>7_Z^s#SfhOyc3SgB#X&fo1dajRS+cj^
z$e#dDPTKa4FBD#<)<lc<r;9k3An4wmEqxqgvJU-*tpzcTU1UsrA_z@6y`3eehIHX?
zWd)hwU?U!Ec!cA*Ccm7{K;dBgD!%;y5c)I-K{iC)LdiT?V+avWjce)6|Ay-qs|smi
z_GB&)Ll?2=mTG}(wY?sBs+@r5y8$^5Q?iqXD}p^rZM{i2!9g^d(G7j!wwk{vy*@n~
zGU84FWoWEp5BY*VlAb{Ujl2x=HrY%KDGy}4QVX{+--@UfrX&3>ORZ3HM3EymZr9%8
zTB}P{h7h0;n?%_ik*k@vKBO*&I)s&AwcC1r-2GHl0JLzNmSJbV)}E`<6s>Is`u<iz
z)6Zv(9y+YN$T3dIIWe;$=Hn>YLnm3<>cWG(xyOGCbMWVVJ)T*1rX-4>x?u<r6*wo&
zmQ#_W9vmD@C{l)dbOg%eYZe0dn|_<+fn_0Qww!*yss-D=*EbR_j)SKyKYhlOvcb^L
zkM7~29O)8e)wiOy$bkgJ6!iJtH0#g0sm^DB+0ym#h_`yo6oYU2Ss@u*%q{64SS-=$
zz=oAc|Mt=xd?g>JCPYN61?DIVQSrQHg_UuE<QHeQGM30G>sz~|mvvv=cD2$z0EQ<W
zPb=C+v77AH(m@)Qk}P-iSBi_GTY*S;qZ?CErdP@ay|ez^UTLcM`zX7B8rtM8wH2<e
zqz$}{Ex=&hJQ&k>IG%j~2yB-&*1Zud80neG)zE}ig?ocdkzhFPT^2L5R&NhojdnsE
zwHp0c)3FH09Z!ey8Jxx&C3qNyslWN%2T#Ve%fA~o@ae>LNyRIfx__E8#LLMuR%&}Q
z(sUvn`qdjPq)TRq-Ii~=B8KoBvI#iqk(t*gR>0=DdE~;ouqSV?;Y20C`eF?SAS-Mh
ziU+rYP!D16bbUS2+OSygs<vik0_3&ntoIp|zMN)<Z3N)9$>`Elj*RM9dGoYx20fM-
zvkgV{MyB-uc^%__Q>6tr?`6mK;(<ZThN2qSK4E5=c3~-eXVon+_2oB&A7Pzk2F#n0
z=jEJ+KYA3^s&_XBi`_PT{Fb-#{tP^|eu4v(ykc!Fs1<knLhWTm#ivt)YE}-2pOkYL
z4wK`8%`9o4#m?{!cNAyolJ4{r=wlN=hb{4-bU~VePJ~K|I#jW1@h5$+8|Y{O^t}V%
zK}3zqN$0`1p@WD@2bp(Ekb$E)Wm#a3dMOAnIT726?;AeHe;EDSMw?d~#j3f}Lh`nF
zVlfJ~7V(?g4V#}W8tayU<#%a!wx|?0kFT(w3RnFRLE(Q5kC#eX(|;-OR@L>`3Vf9Y
z6OttQmIz>6n?{$+`D$e5h6vVC2wJVqE8U<=tpndtiOEv*;3jy>iFz|gWN}<a=+PmW
zB5&afrY+1YJ!|I2#fPOaNS6W2IiMxS=X9sm5(k+IhWLoGkui6imz8xLJ0MQTv>q{m
z+sA6`{wdd~*rO(L8rF|Wr%~1Uxq8-U11&#l!5^k?!1e}59%2G#svaqzHiXiFk!BrT
zPu(B^+C6l$a-^@zyB)^y1&nS#sB#toTonMf-FLsj?RTbxrc$qGvUr?AoOEGQLsvO@
z)-%%$5lcdY78t-AU#bW*Hb%aUi#>!t4?#A4bMs8-LjZLrEIq`XCI$Eu6}}P()SdpP
zNU!hdOt~*XdH#AaR9PdUt}IVX*>Zg~Ou1#gKW-pR7OvF1<10OLHKWfRy(m_o`3+rD
zR?>c*H4<nzU%0AG(Q(;sS}sC6(*Nx}pV8dwnJ;w`Qe@OAfNN9PZ3of(i&z+}+ze@0
z;%-_0M^vDUwNxVCzpsH=A|^5i>{|pDQ(Liy9i@D(XRqOsV(-v0bc4GZ(U%JaG%E?_
z0h_bzZP!&t7ckG=eGXH#9^3RO9t0~kbF@^hg56y`7p9U&l&bAJc(97a64sh-uZ_7Y
zFqtoS2HVAhDfHJHpx@CA`14FHL2EYWQ0AZ~ACa&A13dbXDA*fXOsNhdKEbuS>jEx>
zd6QD@gy$vp#_GPO*G=5i+k4->U|a#+?>Ydw&GoYVx<H~ZR6q0&W`M+C7`+saZY-tE
zxaeu>mgd)y*j32HB|KE0pkHhY|NJqWDzs4zEd%<($pjf`SCf09^mb%69jMxYZZ@pz
zQ@JnfFpiqO!A2{B0{^3U)1(W4qI7ub-i$6g{HPq|nh!%WFsT8nS0r#2G8xfy!3-BY
z3eOCxRYI~9ktml1vmUWVNNVN+=6GmKE8s5YhRJ}jSjh{@LzpU`2e!&r2to#cyrwcG
z4y*9_nux%MO@2rOvLK>mvrw6Gl1k&q7W(tHMWKYRB&_`54;^tX1~+!85aCHcLr9{+
z`saKy1S}`$3LBj*;NHYs8R$2T36LEQ_X130{*~l-wgPux0A1K=lOvhj#oN=NU~S2~
z@ev-S0f}tyNo$u-dNsiX+mC~Krv2vEX*yh33Q{$0E<UqI-5Y#sr}B|-5}t1P%-B+>
zL?*ScFdyyXLEmYGevcZ&@NR)Miy13d5``2$`tJ!R&q81@+Xf{4ezg!7fXx(0vaxW+
zcHlJQv90Lxrx}J!m^=Oj;lm9*x4Zv$V={Vd2J}o{!m2O--K}m&o&cI=ZQ41&j+9Z%
zk|Pfhx;+k$5u(KVk{SVIHt#7iqwn}X1>MOA917*h@4gCT>GiCXzhR6=_PQx)+i~N@
zh^4AKG3wJNr)TIO67t^zI}rc#c6g?`^Yd`+s#EB(?}o_l^#q_@mdNLP9xUlkW%As;
zU3KLMeGLT*eZ5!vF#-(ZBO<?dhs*buEx>GUJgu(l^KlJ`JaRmp|0NTT`@{jj)}eNt
zAD71(`b2`>`_tz7^+3iLZPPy7fHNnw7XAw<t>U?wH^7t?CypOyoRz%%>#K;y`v5Dd
zkFrjO+82ZQ?b4z``g75Zp8g(P&idS5rWT1-5WE1icc)c|zcg3^Rd9d^n&%^uIP_Kg
zx@e0Q1MErn^Ja9#ZbGZm8-rLbW(mpq8l|moqoZofhHI<U!%E;X4C0heVxx!=qk;Di
z-7VsBmaR}MzXW(stn@Cp9_qGJjRqKkL#t_=vwp?~<mHy4xO#3aZ??SnHSGl`I_ZbY
z^suxL=;i}5_Gjw%XIGg+<l?oJWtGPO6#QTk?#yh&-wJO(YU4&)8g1-^SCHd17DN^E
zat)WV0Zno#7md0?G==G>A6hMy9C|eO#%y441i0=M7f*QN1|f5NJVWAnoJsM@u3dB4
zn%~g9<sC_0B9VY+rZ(PH?KkcbrCnh)A;?1c_%h$S#n-R)>s>X0>dK0;i`AC91Tsld
z=;)iminf2psBPo6r7RhAu?u;T`0bfB`waVE;{T@IohiMPYc-V8cv=n6hjmmMDCV3A
zh`$SwYJgLki~T55U1YzTD}}(Jsci|=Py+hitdYY=Ir(uV6Jt3hfus>E=1$V9@L|Q7
z$^+u9G0S4;*|rFZ;=4s_7+ZytOl1>p@Xi4&YuBUFI0eC{b0_ZQ)UBM>HOE}Ck(U8w
zo)o&l-I0Na6lK^6#(=*9&}sX-+3(6NY4FS;s49nwIBh_v;CbF+>?reQk$2ccNz~A;
zOMExRmMN?!#RcnJ*N=yiUC79*aVa7O7zt8z0X}t9Sr99(cg&((YaCm1*a24i^0XPu
z;}{vX@3C*zgMiuK!Kr_7k`5ws7kwGK0NLP_oE2Pc9AJC#POKU+Ii$|*{d^6WHZanF
zwGPL&`K~=?Kp_H8(`MDKt?;a~d9>^G)aq8BlIAXJ8Q<yiHA?51%35p3(d}fX+4dV^
zt2Q;RY+)|3aGCe^wmA@l@ba2r7xQ8N4zxuIU$<jvb>_fmYj4+f_)B5+6Oawu9Zlu%
zI33=f=*t7ViqO5GuCLGKxB`8j#}}*#257heWm<3ynB3t6rsEr0YU-cNnA+Exl*hqn
z@QTn1f@50WJDa4{0dkK0_j0x^M{!p~38Qi*7!ib>KTO#F$I>}RRsOwie{z$j$+m4{
zvQIWA*JPWMZQHKNwq28LYqFmA=lgs9Zq+*LwD-PsU9Wq0Kn%AmPl~^=@`!t$fVCci
zdR!V^C?#$2zZ#!Y*H>ivlHbe6;M*M(+dhPTvkh=xZN1wFT8imE*Y%2MkSK5Pq4T+#
zg8~F{|2R__jcQKbsN)hEFP=K{;yie?#*T8)<OVx)Qj2CQ6GIG5p%gQ`inhi5Oe4ft
z0izlgQT#IT3v36!Z%y^C=ouVP15UXwj*1zz{xnyE7t{B+As<+bQPqS~?bPLw=tRUj
z&*EX!eqf$w7)?`*SCJ2sv&?181v6~_aChGB|KTrVDkWr7lRqq6J1YE548D;dPDVR{
z6)y}*#87@ADUIOJH|Y1qYq5NmfL%``)#@9d7y58}mZmx6P8xL>3AvVvLybStDO=aE
zDUNAU;cOUps}ESBf&Ml~Ptx=H{&@IsI_r2erE8k57^PFpNnoS87rQhb)@ZOj$!I&J
zIh{zY3~0%B+n>1l8$fB)OJL`4V6e}B$=tJWCc#HRRz%;=oncsFG^21bXwmFD>}j&<
z5ORq^3aowi&#TJGcvU2pAWEE(z$0Q29mM&uy1r;k2ff9d%IaZcXhSxX%~E-D8c^=g
z#x`%xA)aymm+i1yqA?VilKBjufg#y_M?g-SNCGv1sI=$Uj)#4z4E+S<_)--8j8@zB
zvT{DYD3T=&Ri@?zrIOnVwI2Azap&lu^66&iuZDg$JO{)X;GvUF?9-h3C-B&#+hoD#
z?m~^HjG+&q@e7gSC?so=e}N~@jTK7&rsKOU02}^h_W_fF1g0xH;+<Pm1SqJ0P|duP
zWDkt$=jdN{Ed@Lu=!SU*?D3DyGx2FU4Cj-xH#{P}yb4zYO*8uO7b404_buAOC>0?P
zM&d~Iy&vD}0P+3@paBH@F)ZO-Ejf<p|2zM{eE^ih|A__vUAS)f0ydRiksNfHNdOds
zBLn%TRvj!RpG4e^VI<<xNrbF>1zx0yRl(U9hK!d#p6D>|=}%O-U4$i7Z4exV6kMVY
zAf(I1NJj7eSOjtF(^jH<eKSC*Z<wc7?#<cPl}`5V*nj4KwKo1Vpt@a1Pym9l0Qatx
z42lvuapvWP8JF$oiWK&@syO=6YOwk)>^+Ap0x3Mi>R01__D+v2r*}jZ6ajX8XA4oM
za5y^3iTIy9n`)3-{CtY4UlxJSx9!PB{1KfO2Km0`rdo=J<5KHsQffzmia#TK*ud05
znd8-VbC9Cv<M`Ja7#RNUOZ}XPL>YKGi9ItL^L$wO@V<I)-tt%j%I|c%O9Lvpk*}1r
z0`lc(jts^>!Uw<a+EV}h>Du<JLVf&5iIQ*i+qLD%C1F1BqDsEGFt{K<a6H-ny|i?B
zMmX@>b(h8V{Z+>+fiy*sk-;yR`SAWCRawT{RC1-0Q=zzOj-)9T?<Psgv>Qo!YOHXC
zw)>_{^?y2#VNinDxo2K+;kyhNCk+BcIqqS8&1q(l+(DK)pC`XFY&$Ffdz7d(+=50t
zog5`)1=q21R@wLu6GM8T?cn^-Vbr*^IL&%VNB>bfQ#aLA1b1!&<)gZY0_w<I^_X-U
zi{-|U&pccR!g}tr^qo9A*$@Sjx7t>nxMv3?&j^OR3AgGme=LJfC$Ni>G0sU%EQg}j
znwGX)?k4*Q-4UvoWR_X%2&I`Y{)Uss%eoON6pw_WODpiL=(#acr>YYwG~scHo-pf~
zI<%_r@V-`Y3o}turd40bHW@&df0JFjaO$Xm&~oI^?|Dkr`5NSY_VxxOgX><~*C`VJ
zJ=4vsW*|aieNvf-y*i#r;lEG1Q^vQ67pTJC<z4u531GLruv`)&zFPuo_|2OXQkyh0
zKazAADfT~ZuXlzCe4cz$d|zw2Kf)=Mo(TEKnZ7gH6C&aK*ZTzWIiSp&>zfVIyT9Dd
zn~w3hzpo>V&f#%>dEA;-eP-6Z{1CK<L17UD3@*SAeDCi$bG&_8*4N`l!3T|qu>PKj
zwL`IXiqVwjVJfh(<5-5>vC!ODod*=Yha-ps&MLJ^od&~xQ-M~rOIz2rB)>!CAb>$j
z1W7%HheCV#T0Vw-3HYX7*5h~WJX09;*$v++2)$aqj7iW6h$?K~&~PGk?Bu}^YIWPg
zdk8foTE9Idq9AB3RFu4E58IwCRt4q(o1+~(JL*-z1eHDkl3&gF<{pq0$47akeS!)1
zRvA#03g@!SL7XEP4UzKIbiZ7i2C)*~ua<xzV3Z0N2xm^g{jzK6n6Ud&bGj7Rvx1Ju
zy-xTP7Zs33jX1aJzcoJlcq*U}<5O(7q{rtX2>C<((Iu}=i)yacldufV-WTgG?M4Sg
z*o!&vlw*A~4p<kT)Mf6iz#OYtq+`a(RnaFH0X-_3KTUi+$*llrUf3fd<NlYW12h|a
z0;HFCAK^g#0WGj4V+YWG0V6)W`S!<1i+OVFk^GAHf|1dJ*wt0K)G5TUEj1#-%8B2*
z(_f;-CS6KAmuz^Zk{dpB@<MA7S)USEV>?u7nhWRhlPX62xrW%$!ob=@ro1YKhjCJ^
z{>s>CdrGAJj5{<cn{c|hIr`OE0n_H_m94T)_ctsEJT6OqW3Hk;Ltv@X^;4#%f-M*)
zm_a!WFX0zb+#IEn#UYjE{g&@lr)`HoLR$n$8#8Rl<?gsFg~D$252=5&v43F#TBRRH
ze+HWdU_?vTr^^C^6Bw$`+Lc&imZ*&lpeUe&#zizk?IYK8sUzw27|F!xbbj_4y<Ph9
z4H2pJCy2r9C6Cp8gj(9}x`$F}VlWry@IbmLR0BNr!7Ru326%(B_}s^^N+Y%|(pb%$
z%P!dL-$V;;mBxil<6)&*P=IC^vMx^w=Yv9F!~QDhdKM@Wx<t<pFuyx|-Y@B=m?xzm
z*?QkiHV=Kamj>&TYTZNBVJwU*LhMM=B}{TQ<GleD1)xgda(?;ue0M~;bN$)2(c<*_
z=BsuGR1P<1!0mb3Q_zQ?_ynhOEQ`d5_DF&&XNrFQ?Ij4+Fw<K_liElBd!l*i3J;LI
z0m?3oxxs=(7TVL+&T>MUWMQoq%+qTy)oeF4mx4*+Xohq@oX8X_=}XIm`Tii)(&-n(
z3=fwxJ$2<X2hX?O+Hul?hWkVMH4bY_sGUL%>4)b)F1Ylz!45R9?=u?P$d53@i`ad8
zRhnf^K^Uul+;JT(O@FEqsA%2<li1Uvq05>HvATTu$|^F`S+DSOHZV_<iiYd3Unz<q
zWnfyv<`1{=#+m{+9|9xwuQ!=-Ku+He1q{PyyKVK)qnlQ;K7^cC?#&V3T2b72zwAk~
zj3Yt99y6GaL<91^9cU8Hm;nxr+^KlT{qzN)I2m>n1Yt681#aaSve%~pYZf~=@zP(y
z0&hsyr&U`m2P-AR|4!}2bmcD4b=x}uVZ1kx@ct~)l$oPL!M`?6`<hfoQ?P};#R3;>
zlxCc{YaqF&Q7?vk#Yc3kLI(5WpDl?P50f&sWPQO)s;<)Suj`0KHH>fOv$!?WWM`;5
zNK`#Z`p3K}ob<hj#uiD$1BIq!q$**vP?uk_Mf_ocZdFWHd$D9@^2RmjsgHF8(fM}T
zYHBv}q+Nz?hoUlIHhhM&JEO3MwzWO3`C{sMZTc8EtSDy2KVLUr)9x$c)96a1xv{xp
z)Q!mGhns72YnnD7V;Per*Q499p$5K&NVZPu;aZ~6U{db>2?=%D&5|l+F{Ox<E$+<4
zWXFF-g!<&pi#;t{a{b58ZM1*hY^n|2j2T+{;1%XSJ)frY8(k<*xfp$Z<!y&lq8q4a
z|Epyu(pZzXVcY*U0}_F1{ZHxy3grliI@V<=`3*5O0#cX^W;hY#8Y=7pgz$qyAXV3Q
z+INT}eiYN+B3FF^N?*rgyU_;5g_C<ICyv=UvmC;Y%l0ck5$JunZen&%{_gehsqO>p
z=u|FIAmn}M2lDzxJTAy@`m;g1@J#(Czg(KhzX(Z4+55sj$Fs)dmEO$XXV`Ro3>^*e
z4W9(sex;~}c$@E|!o%-qRMV60EdmxkIIQOR{TiPF<abNe+A(Z^mR=h?lGpHU0EjRl
z1YWR$EsC)wzN&h{tb!$qtd@B_k#dxNGkuDSOL`ySc;lkw#xT;=P2K$`!ZoG+CtuA$
zEE_Bvo#m)TRSH#_PtLjo{0D1WdWL!jn9yd=&CVUVP}cK^gtMT+tb>SE*#;SVg?qqO
z0L(z<nX~cWhWp|){`1Q27zyV$X0<<@BAG>O^q>9VCC}oRMkbyI=d*Jo!XT2&(EcZU
zFLJ_I;15lM*;XlHdo8L#w6gT>GkIGP1ykNdbf=E#+7r-!5p3@0dUm^x&yINfdv+xH
zH~;sc?o}Wz^wj(<yStVMvE8g+bn`0IBI;fWzia`~u!@8}52lrcT^C_Q?}w9r+~B)%
zz}=vUzUM#a<fYOwMV}8TmIBC_-8;?|n70YF7Ph2O=3x^49rL&1Guu)?(8gzhT82Fu
zsyMoAV@;hXjm9QjSpVc6^<IXtZ$4<)41hkF&hbv9j&7G@!8{kmbLA%cI%Lfpmy^M$
z7EX)$q)~Vo^%}T5^*Qvvl_G3T{kX|;-4dxg&=4eNlCeXNxM*!r@kQ&jHFX2{PX%?L
z_Hl#m+=Y02rL^<C7(~L4W4q@4qBg}8A?9Spa9>wA@)2M8V(fM6LQ~@!V(m!S?;F9E
z%ZCi+kh7DL-BRzz!u8ML@dm~)u#~fpOmSn0WAp&XB|pVb;BGh=elh!SG}2W{O(agE
zPG_hWs!lZgpL&8Cv;#jdr=y@LS-ml#n3v|VRf(TEOb8pGv|3VYca#78oR8&A8{7UY
z(6I9d6NsNgN5XqMuc+yMJ{bsW=>D8q=^bxoxauid_W!S31$=zV$isje;Lkx%uPjH-
zBxXnW*748Y8U|D{&2K^OyMBO%1@IfuDg|pJ{@KH`G+ZDr6j5($1t7Mk$j~B+R3<8#
zSJLa<QAqU*I+-$E!!bCLF4^~o=bo2sM~h(zPDmV~H{ucxH+vLapPL=Obv<<n#K?>Y
z6%>E5v26VTTUQi9!W@i*SloRsKh0?je*3~#@PUG1g8Jc{8>FGiJ?9d_n;R?=*Fu!u
zN9*cACRGh?Ti|aVf^N_evCHIHnLwS}0&z$$Uw~xdZ!>dw``@bxXhtJ2=nwwc{^9;>
zyX^F<8w}`j?YaKkj2hS#uFbC197`1et<d+@+NH<@P`z^$WRPp-u0+c6IOS@<OFRue
z5ZQFk71?q?*+^iB0D)nmu@@rFFT7}PsPomD)vPOQTOW9sH(WO@_SO5%2E6L{570T&
zzk(Vo5VuZm7XK(OSQD*$+=QzFl~x_lqtNl!&51m>PMpUU=PN<#sH)%>N#@0bF?ED8
zMWT@=9bS+dQq9ANA?s#w$z(*y?HD7oNL$gU$G@Q0qM?u><<%G)aYp7IwaBhnmQ~2p
zU_i@w_B7$((5Dlw{liVeikmw1<GD@@9&;@n&(u5hB9xWy>%Zl{VF!|d1W7X=4>(})
z_9T$fb-G`ii3}U;C96SfKbdn)RlNKG9V<{aRqJ&QmBKUw4E?vf>k6|yVxD`0GP0pC
zSOnv=zjHSWa8kh*rK^KaZHj1CMJyW30eP_8$%(8h9dVV+I0}bwgCaTJ9{>d?`{{YV
zHDU+P#x!Xa*ro^j12^Xn$k=ZD1XA|=6grhRN|rWP+8r3B2{&&jsw=l!kIqFnk95Pv
zzGm2VwQ5v6Q{H>MZagBD3=_hVE2kwxz;wU7WPBI=^iteR^1Y>In7;S@1p@+qR-lyk
z0#a%iWbfoj^ge+8)bzvt#?P{hr5>N~BQhZWR&r$IBla7;7sJugridjue<CX$)ELjX
z8i^U(7#R)6W|3C;={aRbaUy|NtC>FOt~ihlWVjjg8jr$d(9G!UKDpceoIftf-kfjI
zhZiH;G1p$itb_{^3gunO33D3>(9$pyfVuwbk80#EG;;iHFk>3cOQ33W=Qq~50c`Q#
zigjksUGbi^KHOjl>;fjQ4+GF1v1;c77rs*zdiqym_iKDaPdyL&bHFBZ9JenN8-knT
zdl?Ko`pnbO%*s}Oqqo?E!n8CHg~XE%a<r9rP%xrXcPcz?uBD&K*m%V7+Nbb*ges*$
z*D%|QN()5j)<)dxJMl~+xQp(a9mpX7aKT3e@D&WTUQPESj2>^x6oKMP45KqgG2x3$
zS<T0=Lv8pUMo0k?gV!Ge;<kwc?e7irF!=$7Lp>IJOvqT-{*WM;y{}lvg^)xR6${F$
zisps%38SRV?ra2CUp|6(RKrxpdjpfXa<urm=+MTwsmN^Fh+9e&g<?hU7UFPFsCCQM
z`=o}+Uw?ArC~&M~>GC>Xu2j7O{zLorB9!m}2T)g&W$}BY{LoOy<idHHHBvXOc&WIv
z_zI*7XQ-_ScaKQi@-VI1$)8{6rHBUXHH8S0r$OgKQlrUE6fs<#Pn#BeWaY0%W$b(H
z8cZI8QL_uzND|+h*WK>^S<FY-e~dce5N{qP2Zji|?gCx!=+CD;V1j6hO}^=``~sx8
zL$s7oRJB=*vRN#XSc}N_*RYpiDjYGDfF&gTc|OIz5+!{hiX{KhtrIX%`TRScMdAZ@
z9N7w9{|~$YWU3vko=D2CpZkDOW&1NhFl*fEtOqHR|MfNOyWoH29zc*9Grq&KG->ab
zb5AWxAnbB#=?qZO0Du)UWl+(_vF2}8^Yc@ehYnI7UlQ-(Ap4gof2Y9psD44;`%ul(
zsX6~1hQHicas#*$x}S0|O%|`5&N__pO7o#NJ*=hxQVyTz6tA0$QY<?0ft#f2$F*6G
zdo?iH=<>W0>DMJ7#{yp;x9|5zI^Es5Ki{Is-TR;@!HE7%!YmKR@1sHY1aWTezTQh!
z+C?<D7Lp7n*+S2higOU;aG?6y%#Gm4M4IcJdrxu{5OBmp47*}Z&AYA_O+CEx57PPa
z(qi7U<6(8bYR!cxSWZ5s@(Hkw>B_vTz)(alTG`yWbpZMEjF7lo$V|<e^0!vGdqchO
zep6Tz%7Fbr-_TO`#xXayW_96-Un-n?McPVqh&<zhde49o(fI=M%@Pnjn!fxw(Sv$g
zjO3f!H%ufgnM-5KlpuymcX+WE&(nVXL7f}iWW6GOKXE9xG2UWUN5`fM@MBnHG=ISI
z!DBcKqTZ&fL5*J9ktJqmLjx|ft;daKPg_#w@mtg??zAZv1F%A7DElrA`4HaNdFJmO
z_6pK7S$7wjH(yJ$=P#v-d6ZM?AXQS~CH4U4%Lgg4APE=Ke+V4+bJl_!@!zSLdPPZd
z6RnUg%|B-21t&`LR4l9EcFb<85>G;p$4g@^sg)YL?#Jv*hbVrURnjL^QR#izJjrbQ
zTC8D1jV43Vl&#Yh+Hq;>e|dk;`>NyYTtlF~d@kQXvw<2nfhi62&*lXovZ8(N*~4;f
zq>aM%7pRp1QO*^P3#F=AUDCfi`V~ks+1$EBK|)R(dL85I!KE#)HU9!Gv;U4Lyx>bI
zFl%(}XPClEvDMx5ALqT^TJ;C;6adJkW;puY?&<m)lRu>riY;D!wdP{2)m6Tj40qB|
z*?gf}9^;LQBW`Vb+TSym7BVzm{DnS(8XETsDw)HZ7g4<IV{Cke8k2JFR$Ot(w}ETQ
zQ%lPng7?U2eK7FHix5a}qyoq?j8UK!jPolR|DN~Ld5M3`^_->rmd0mWw_?KaVMJ)t
z=6u59+cZc>dW2@zZ)*l;Rr(hzr4ocH=KP^V)m2D`D?L_t?DovTgG&$Rc<<VN!*SgJ
zF_qFUSAfxyEDW}<LB_jLPn6N$_zrW7U9L>j(sL50D&EtrD7a{Z4Y3=NzUD;HlpjtZ
zraV8p25WpyEK?EWM#0C8kLHFfArgR>0A16C4RUQMG$`et%*gyUI)Z9uZPrw|RcrB2
zxuE8FwF<aMV+8yZ@~P&oEjP2uI=MQ(%4C}~^4%LI{%B$xPWlbE<2S^4Na9I0WsE*^
z^dJOhY08T=_#1oD#nCk5Rnp|NdUM4lvwM;~U-Dj^ua;Gp8(6@#{|K|VuBvw}_7=}7
z-sHk~)vC(6pzus0M(rp&xi;ZTGx}RO)|e-^K2LSLzh5O)`(#6r(mW7Su~Z3&T|0N$
z{0XGQA%60_wuO?zqC<!jl+%1I?l(24Mt>94%8{Iw9dgHhVx9~#<~eeZDG-mom2~I=
z?fSY<hnAIi?|JZu@Rs4Ghhh^bRD<F2##=P;rkqDg$E9c^*>S8N+yKZvi#*t2v5wzO
zN_s0f3fXTk@(udl?}$uAxClan4h_?UUb{Dx7<3xuP5ng<@OOd5STMU~vUG9bjY)Bo
zu3vr0r?WdpECzis|JOdyVtPXM`jS%8(gBg;x=nJ~1I$Ze2tRLt?p*Aah}xUV$YtQC
zPpwtS6rj%vb;m2jIoYe<i4A|1skd}hYk9q<FCdctsC>Nv4;Rc0_WPD~RSJ|M!jyvf
zI%+C;4p!#xvW`Pt?+@q|e=9W~sY$F8yiw4Lp1{=~Pn{oY+Xdd83T{@?==ByUyF8;f
zLUu8w;rw9DB;TTeH!tVUN(;@+G;EZ!XU7CAl!_%T))Io>)^Q~VL!f1xw<^~uIy=@v
z!X_jBCPm6od?JYKvVtY4=b(&4#dH4f*WR$)dWVT8C?cqUdEWqY$H-il5`?RzeE>2G
z8LtS|cEpf;V7ZN>H;K+zC8gEk8k^7n>kB}#Zz}NTXs0a|D~D?&g3nIlN0YweNWT1D
zMs2XrTGeIP_Y~&sp|1+KFMx3-@Y0lxy{6M9U$>_5DN`t|#JiUfk5;H0s7xQi70%1X
z+GXe<Cr2EdcEq{m>P3`hxm9Xd9;R%Qz?^cC1@^~~)``APL3_gl)feoQh1V4x>dfYM
zu_gS#wvq%SYRCEu#D)Tlry!;pk>uY2dTgdccc($q^rrG!0^II->y@Sc+zneC;uNGl
zHIJL#Tco}4;vYq<7i4zEyvb!o2Bo@%WLTibZ_&F(b(#aXk$NRS@i5v;s0Xd@W;k&N
zOyGgvLWY@%$B}gP=Fn$av#Z_SUy(W#?^ml^2(TC$X|2tqMe~>M9@tY6$x@C5LmGWR
zo9_jRYTa-Hhui=1z^wxIST(`1wv^F><*>&a<JRXdj;swmh_ho!LPN8$ESsg%0-u>W
zsPW{@X6hn=b-J|&Btka5X$-FI|2uKMw@5lqchmYl&-d+DG2Ra=m~|SvAD&mZPHXr;
zgFdoEs2p%PQMJfB_oSfAFz7smIKmCX)frSKL!v43L;~VPRD(dzd73k<imGxJY^NrY
zc?x#G5qAIktom%VtYH{_F@$MhU2SUy=xMHeAGMJx8zW?(R&~2|{$SlkaBMpeqC#Eh
zlZ_Rrl3rg3faclw@y9sSA3+R6n<VP70G^uSkmP?83bDSYYCK4YAeM>&b|@-)2iHJp
z^M`jmRcdM{KueS4zKAbgM>SKdV3HZYA~FNal})e2QGlwV4fJ>oU;c4k8>NEZ4zb$a
z;P%&v#;S@$2$p{<chOELEs%9^sL4gjwZsAMPg~fzRkHDA&ka+dh0L{~avMsML^SVe
zT6oOG`1Tf8i1wq7O=_dD#C~kQ4C3X%ik!oDzm8t}>}GaGmg?A8($tl-4w3|;c&b>t
zdGWFRB^{W+?;sd&f=^qPWzM|EOduskomih?^C07G=X1f71&8W{o69|F?S1Dkdyr)q
zHs<)fptb>%Az6gUsI?1kE>4b(!Hu^#S&78qqgO8IoHouZ9amJ3pu}TI<TmcCPJe(r
zaS%aL`+dv<yASb+Z^s;rvFq9hsJb8?-!pV#^lpYpEIT!f_6$v#nF{fw1F*gF+&0k`
z=B!LG`=d}iw=jGnajA)uEGQA0|FQ1X<=LGB{-Mj|S%ZHDdjg5M-!sr1+~I4u;Jk8`
zWF`WH0S!o68c_n{?o~TMqm8gWW)!715}sDK&)%l@>u3_D%w)H|w9(0);74f5_;K$w
zPlP81Cy9SrNtB+}lS7@42x#4nB&pasPW&*G(1%MtYk4}G-}Bwi>QmnM(PTvK*U7>7
zIIo>Mi~$VL5smpS`0fOBl5*rTxg7pgD7A89D0m=F*)<QJp%u%rg8w}A3S+Do8^Zwe
z0sOKtKVn3Ge(gPX%6ijbJJMM%ip=>&9QA7r*DeA;J>-onrwqFQy4Q*M>0K8zf(h4+
zgsjKQ2G#Ktn+Aw)Xl0P$@~=d<c#hf1u3}ADc;M#BWek1XC<s<@rPG9HA+$YX6mLKB
z4}|$^$KHNv2E!a>giPc|j7k(6hsfmj5S;qupG^h~@z;}l43Pv|{q<P|<gP%0U0RA|
ziFbFNTAjty{cm?X^QeAhBLLE3p<su~DP+DNSWI`=!@{0efP*iqGC*^H_LW<~pBERZ
zvlx?CT|eIad0C!ncKKV-Rg(9tM*h!CyLv`Hdi(!#VKi)Jdfy>}mf&D<E?ZqE(-lft
zsk+z#$h{&^LGEM-a>Jc0T>u3Ed)^GRA<E$IU#UT(v4dS)uG5t$sVcgN5iu#=0qh;3
zN-Tq5yZk*;#5B=!p)rCfli6OnRy?T<BvP;X^TQ{}*Fpx#3{4PN^&3>fsMIu-L^;jb
zDGSPU?0eIp5W=Ix8uEp!34bakkERddmt`iLpm#NEx9!}s@t7T+>thu24(-^Gro5Q7
zxTw<9!tUVLjSS4^bh4}Gp`H6${#^wdVb>+@nEdqLo%wNbV(qW#j2tuP{dB<rb<NQX
zWt!j}>5%tD)*;|OBwVCkpix>%lL48`d;&f~w9`huu@0&Yby5a-jf31(MfJJ;AZ1SV
z@WqJ1<9yinl;ajaY(P2M3IyM@LBr@5U|ao843dO9C*T;g_2f${#@nmn*#mGVA(HQs
z@@DW{Y!q?Q`UZ;}Rt%hicMHcrH^lk8mr>+0`eN-s5eUwo$4CJ2a0h^~Z2>Q@e_%-b
zi8WswDxVxnS!$%&E+FJ|JqlMND-UPD>L<HCfELlBDp#XWPdF<ZGeM-DStR1`Ky?r&
z>wY*b%&6|(X<VX!a<*+4-)mga4ad=FG5<RMDrce5Xu07h@UB*=A=c=~;H<<<aO=N|
z)Qe&);Avb4BSIIgb(@^sse^BHGbYec1R~Xfe+@JY4~onLw}zB{MLA7gAxDWH6=-O$
z2QPyY|0w~${eV_Mygo?HD;8nBX|(5z;gw6&)pR_$<CONN{xi^?R|x_B-JG^nF!{pz
zOr^c1mn&x4z|Nu5Lpu+yg|CS;xAEAq`7hvLHaArisnJchv6hnnxh=t%@XnDLH(*t4
z%zF*M`JxpW!L4u7<iCuv=Mp)hiG!gJLc-qNTm#z=K5odVC$e~bukk^-|K;4*7l9az
zdRpHGa1}7ZFC={TM&noVOBOgVzxWmd2?<04a(CQ-|H$6tmBQu$N)@{F`EHwA3oKG3
zrNuzRZ&s<KT2ng7`WHY*7bEb-bOIt5(X&_Rij{5NmC`8)1;jS+D^&-FN}6cdiv10(
zMu1ack;+Yzt{JOBBrH<5-tFZ5aT8O9OpIqw08}aCsu=FoVvcru^mE%XO|I%SYJMLk
z{NFyTAd?<4>eHpoo>Cl)=gMu2$)ZA8UrMmi*DCIWan{cI7D^1@XgR_|Fz+I1jG#v>
zSH?~qO#Z;V4MxQ$&JTIt)9?d23+*Q_XCjrmhCs{GfjVHq_xV8m4b3ARVi|Lyt7^Vj
zz6MBH4~jEH;!@)>*PmK}#qk@ELr-y$l-dcRNPsv%3Gap%yj3BSi}|2YYD>?WqFLs~
zEyY>yzk0>4%ynCVpX6N|KZsHZyI5I;%;tL*^)DCBkc2coSCY`dVFP1d`PCBj2MaAM
zRNf@De^9FREkSyPrX}s+``6+b1vDMycb*}?UNe9KrQNw&yBo0MH7XMDxS0b%1fk-|
z&>>gSnDMg4^CDg~m>9s)Ci7`%2%?fnnTXGdtaAH(GM&6z;peMF$@^=`PTU}=ID&X_
z@Cp~(9u@uBdc$TKs6K>15v3mnFxO3VulNka;CvYyopw9<Ko@7trfcys@p@dr8{<u;
ziyDzR92@@$DP?KyOK+z7J(T1&z)9xXvQi4%LT2wA!HDkY9z8++tq2_s3o<LO=iOMa
z+S<1Re9y4D0V4_n`Gez^okx0*9kzIG{3>YhVt(H%4CL%(3C!v;VdnOn@~pw^?R}LD
zpzF*W+`q*%-^j{ZiqCEgI<&`NZb2m%NXJlh4Wh-L(3SH=I<vIj;Ig*r<Z0yp@Ol<0
zZU4`jWgv9V+zznNoWS|ALlU?#hMh6TO<0cy@k%M%`eo2%3MZLjR(gis<u-Vh?g3@Z
zr_b7)$vB|pVg^`sG2}}0Fi3_mMRgDjVvk~&6}S6G;bP}&6PQ&H9juzr5iKjPApPzJ
ztD`A^2es7FJ=$V0cAGEa0gY4$O+rSe@dKf2w#b1;B!#xXi6VjRUaM=y5?+X3WG5BL
z8+&{lQwCU2(b;k|0f<X>(Tl$hH)_JYL}J(2uK#IW@cI%%5^Pi^+a>uV0a3~jj~>T<
zoRSB!`M8)!i+B~4)RrK6{4=Ty9Dx+ZPV>QD#=$_{zwYiZK>yp<3a00~&*{Ja3g>kP
zqHM{p6wn1fD_WLTPm)EEY9y=3)D9|U5B+}HQ{li3zlr*=-$_Q{+{Z7dKH0vj-kK8o
zJ`GK3O9--i9On|pt0Vp<`DmiiA5W>#`s22;*_8PDE+?;@uy>D*@!Yn*7m9Z|`r-EO
zQnE&f03-5^f@4So#iKWcBw)NC)T_Rn(0m%?S4UP6!1bN6XXg}l1VL;spfoHs^hk1e
zgGyNKjiH$n1#+va<z7M&f{V>hX{1A^-gw3Pv+Z#-jue0)j$|uKbM@k=DD10C*l0JC
z0fNf)NdF1?GU4yx?0+JKN(0LlcI;H(#ReY<1smt-Gt!yTGkXn||JfvQWC7?){+H{5
z(tM3if%n6S?7cgf(Y@lLb9nPSyGltipEFQX=Gcr>t!epFDz~y!wK&bJMW*O%fjsGY
zz=plY0@dWREDdJ@F!{E_!F|e~d{Mptm6U6YVT2RZxbA-^#liih0y9-4&~)z(kWv~D
z;#cgJX#>FdOVrUh&SV>vMXh+;Fsj1j;u1tMTjk5d<a#?MFvo#*1z-qkmQeE0pKSm@
zFTidfR!^m<I#E<j^0}Tjy^RtbI`)J=#)+nfCMw9UH8j?FDgi3(YHLyM6k;HF1uV{w
zY~Cx5Vo5CgbMs%VUhQ9dTC^Ngc-EwCP0z)f3Rzo~?!ZtT<6Wh&aSQ9Yb2Uh94K)?M
z<T;kJw)V9;9x?3t-vLS{Ay)A{sZ;O3YumpH$|OnhX1yaJDbjEr$zXK6NuxWUP)WMD
zsr@KCYv3kf-e^jy;`^_c=kozkHln?IKb>J52(PzDZ@x_svDhC#HA29xMV0*HFyv>~
zFan{OaOrQT$(?_pgf|yRoM|hxQkEbHgA-M+iOAjZ-RaJ7{q;rD6VUxw*G>$dnz}Nt
z$HKh1WBK5;s&&iM59BFdwb}|j+sH{swZ#v>=fsdg_F<S;f)w;)N3xQJ(_vWJEDly#
zF;I)j@+8>BBM57cX7krbK60${HnA`d+QAVjfNqH1R~=PMly8t}R`0c6@acx0-IkjI
zSQl+R{S*zvGkS;>+<*7%q~fq@8}~5mi9wohYg{OiCO8mdEPI<#{ZaH-rMw%UG5Pgm
z;?eTZ_a8^<6;$Rmj!T*equ43HA1OA=!JY@O|G>Kl00**J+-YbI<Oq1u>H*7Um-mcs
z_9yb4O`UO^9O6OB#0_-WJZwkaVFHs|e>q6Wi9;VMhK(QsQuJ@usMZpC2{d?wGfm)a
z0%_d<fZHbev7f9Rxsgi0udx8;&MxLP6{eD$CTT4`A<pn6I|WK{ZY97(NR(liLMw1=
z4*HRZn>q+gy*`~@v*Ub<xPR`1?yFlV@o_vI^lvRV^0O0=!g-Ie+K!R@5=X6Aj438N
zE128(p<}R7-pG*KzD}MUf|&8P+3*=zyCL>5I*n_MbKGxn(<zR>5D~IwaY-~;-(JRy
z9j-gKyAy7VO+PxG6{4}qPXBgluz2qdOk=)}g#VyP#YI;rj<hG?LLd;s!>VwhOc0_n
zkZqf-fL8GerBrZXQeAWS&I8^`pQLvQ)C<5%b%4Kp)dT1^x&hHPpr@mp*Ia~o)7b4j
z@)HvQR%5FwCZ1_dqK6pVI>9Fneh@9hA^o0tZZF6BnNq?tc4IK}a{Yek>6qi@HPGQs
zIA%$Np#ZG0_k3K@^Zj`^B)2v)h-)I}N1cub^=({jv=pg?@(b}xRnXC5eil*?XFEAI
zrJzhbDuP%Vn#q2kkW}kwS#9d!SDmeluwfpxW9_nmwf!-EJ;M3nkGTlO{}k~!+~MbR
zBr<x72YCfEfH)~EwN7!jSrHlTinB?(Y2??wi-0c^3Ig*l4Owzk?^zXnl)bxh=y+vZ
z7?NFA8PC|=8ipC6<Q+~E--P*{c%I~Sq()TmJc<y{@unAGmcG@Xjog{|i2i*j%GDi^
z0IpF@XYtkmx!1rRDQLU2PJ}F(>dUq1bqHnoFu-;aLY9>m54agCSDZ-79nz_W?Oi;0
zXzR_KTA&ZF<khEQ&ZA?8(=~MnHZE`M1Cb@eiW%J<`PQ0Fw#9!Tf~3S{PF|=vO5^;=
zB)(~xk%Ebir%=qjvhi{^m)Ee5HfYqa(&yK&>AD~PUY9A$f|_@u+aP51bj<3f0`!y(
zFd#(8+a)Ol*~{4pmJQy&4V-exLSQ>Zj-}Dx7~?n~luK1fkPj=R(ItLu^w@zPM17uR
z66R8h2UMs3V9EelU#!Yo@{`Wikq%A9(F8!T=$n1Y68Q~$%kpV_<Twg?B!20#X6f2U
z>D+6Iz1(lL!9T=mBj-=r4u&$Cf0c~$?8Avs{H9b^0W!DG^B31kz2ED;W{(Upo&W2Q
z?4I`kHUs%=KIi>ob)c(WD)v;pU2Z9?zQj$Y72Ps>2rIGog;B(CT$w9zJ0M@Ym`!(I
zlW?@WJKXqIwPMLas{*ml)U{f(0f|WYXbPYg;;<KB_@c-^o4Q6_5J(-u#43vbytRUx
zAdGN_k<n6Omy-6?dIg?h6BJN8xhA|=ny~?X>LAPdm(ylW?stEbT{4krsH8vl>$a!N
z&ne-8J-*Mgc8fJojj#-J0%lXZ?=>f{VRrtmg`#E|1vh{wsLpX|6PQNu@xNG<Y7pdi
zM~++%_`)6PF4pWo6(BJ*tHVn~rRLylg4Od3^kGGj51vuSCP4RhY=QX--94`i*nZ?v
zA7PhmQPFl8aG_#8A4mNxTTZ^<b?JPYqgB^ksQ?<m0K@y^t}lY4{d<m_6d1XWZ--IR
zEUBoAh(n$V1ugPB{)CRTO)mrWp6DBf<e_!H|BW$pNZ+kQBR@Fq3Wm&I#eYyX$&573
z7*Ujsb_*@2Ki`Re5{8D6?+)3+#GM<3#stT%!Y%k)ljD)al!TJUi$T~JlYHlcgm3*J
zwsS+T-Mq#APd;j}AvlBxjQn#tgEPnYV;qAHYkSgs2-woyy?}Pik|uvw%cBxs@JOFR
zt2OjSnP7o;6+3z`fXG>Nn2f3aG+o8Hq0<dv*?sr9JBpyyaWMQ%rmze~s<2_=_%|P@
zu~|EM1PaK|`3C5$aPE;Ucd%LY6pHFg6h0m7qy3=i2aT}xtU?=d_m*$c;XUGA_|hnT
zeFsQJQ|HZqk{ZY9?DOXp?|LOw{ljM>Hnxxonxl^6-z43omUf0izc!lGB~=w22pS_H
zF=nE8m{dE*OV4&&ByZkzUWmgMD>YFuC?5sq$;>isRoxA5sIGS(eZgsg*UV}G=kk|z
zg7fAn?kjHpmlQ<DtNyR>%e~+qv`@R^jA@<c^<V&?ghl=3w-@pg!Llj=z~LTFrr+}Y
zcSC7fk;Rxeiee-_*2-9AETVc2LmTM^esp3z%+WV%tj0i;nzG|WEzx^d8DhLDs^ySF
z?6rsJo5Nx;C-Uh5Nw6QqslMqoKbfNM;!c*`^XG``(|xB;46u4(;~`1K_e~<hz$Y4z
zdpgF_7m4e|XUCAwJ9JT*po5J_RGt$&SM=8zlN{7mm?PYGO+_|ii^$dTRqsOECLj|a
z3_Caoq%g)Ed@T$RnsUo~>g#~0*P+4~_Xnxib@b^nleb<JV}$1hT_h&9JN-3z{BX?v
z$-wW*{S9MK$An|$F#ChX@I=TwvJHXrpNqbefqi7eH3wbXgd`UaA}QLxV0#JTw76?`
zHLi3!<tibZKQd$|`hrZZQ|p1qss0w2jNSik5g|nODT^2DQ@^)UIBTphtwj}AuW=q2
zm4HaU5U}<Cq3tLv?je8F#va7cjV<kMGpG^~lC>^y0S#EUmwVZiuId8b$yMqS2K_GZ
zNag26rYUqb$H+3=t$S(p;*9U_6Rlml4B;?m$7W(Djm85Cxya3+^FQoLO_JJ3{Kwm)
zou2%y*VHYvD3|SzA6?m-tn$401vV*}W=S|3VwIbqR!yw^IgbdDv8cWUj^N#0Kf5g9
z292#7zk)@ANz8F|Js&ZBg{$u|f>M_tq8RH<>m`Vobd#nbB|#c?RXa;i6(A6jaGul}
z7KAOQAyH&hKil%{Zoht*0^r*oN4jrL>}Ooz9!f1L>{aZ2&v#yjX-B`oh3lOLiRNxE
z<m*=pGF)N6c)hSZ{nm55ZokUjL2W^F!FVB`OCeity!2D%e6%kTRQ<(a=&us`C+3s4
zI+{(@QWabC3zzk+wFXUE$tIF)MVa4`E;~rpqJ5+9smI_tQdxo5e8?XFyP`t3{$NSs
zziHWf_3*gry0dw7>xy{l5F4FR2tyX%g^i|h(xl60YZlg7H|vcW<2OMJs`9hUQ~B28
zbE)ML`@13%<{Q5TID3@xZK$(OuGHpjc=L@;rZ9d~3Hkupk=+-taTBsMIIezr-kqp}
z3W~?e#ijYPvQNAwabab^I86Ny4b*`GEAGHxNvmO%2U`;2!js`BYa-#>%DvWx3J7&C
z^@EMbe&@0PDIv%x;FBv5AMX5uZo419z1_{Lzvo)l^?k245P2{O+_&{<+UoDaQDQ*(
z+97v;b7hu6lOiLtL~MW7)UMDZQ{DxuD;&X}mic3IvNGFmVRz^#nw&R{a#Q)n2uCA2
zYF4^aF#PP_7(Q>0y8qq(oAR~$;Htw2*}+b{fUJPx9LsE0Mu{vyjESuG!H^}kO`~53
zio~<0URpG@GM>k^rSm;JInt326E*VfdfN3}_eMS)-@h@;D2>1>Z&zbSTuq8ts#-WT
z#%ozMSAtE*sy>CT`02d1JFJ`0M*OACxar~cfOcm5aJO+FyQ~5)-N~gNgI{BwX)WMa
z4_!R-Y`HAwHU?}9Bfd_6-}S@as`h($yF~M>U+>trpLhrbQPS<dH0HyTa0FsR;;zMq
zYlbH0pU>|YUCu;WgV%8Z8>8##X1|C?yEA8({r?@ImYW`A!MCTAl24l$YnntF9mvb<
z{Pi*6$yk4<8+B_YRM#9jel$_)EQhYdNoqfG$cH$4#^<9j6%Q^fmo!DMtIkJ#ubsh~
zu7}@=s)>CX5`R=HUNQMT_Ftkz)^&lkRyDmxmER{9K}39bk%bMV2hJpECjoPOSkXnS
zHhFYD)k0M9bs+wH5@opESl*U=_zwK2d8c)%@lFsb%zlHWtvgI$f8NJE1gH)HH<^rl
zxG<n!mrn8lF^5|2mIX(lMyJvAGP9M+CX=y^5$qc0wegD?G5u&Zs)#{Fom>2c6o>s%
zIoI|_UBOsPWXN9#16GvCkzgZC?><GurtO$PlG29PB_bTTmaU5(kR322!t>r&Y$;e4
zekHs<98<l%KbIo9!Fs8`LWuY(8#X%XoJURjt0i!2#D-XowG47k+PiYLdIQa+LyIu5
zeIqWv`{M=Iq05|o<}eV_Ma;5L$!c4`I9gHJl7#KwkCg=iVLEXuOXcDF3B8oRhzqt-
zoNH*chOUxPXwvABcb(N5)eFFFzd-$7s=BSjsio!<0v*-6jiFz<Fk*R8Y!m9f2lBke
zIGX4~($%)uY#Myn&&lArM#`vJIF(z-h-rYC0Rx*QVM`wz!xH3+R1UCkxu#IDjH=T1
zwY$rD(01A+h;isQ9mz-K)2OPr?)plwVqE|%zI}lHTiJIHjSa1y@nVB$f3@B^Pru*w
z6afF2#&QoR%iie;{vEv#T=Q>x8o0Xo{n*Z?8VihT!9wb7W*-XIWW3R!q31AZbQQlC
zcH=Wv`)GwBJ4ss4aF!uU_IQeu0Nxd3!F|vzb6sS-wTZ28x9k+I6CPa9-C(^Em8s5M
znx^gwpYh3_zMK&O8D3b#{6>4vTAiB`s5{En`yq(xY}Nuoq!)g=`}&_UP#RJUmu;^r
zJ!EFO>ZgF!h|=}uZHss4KhG7V=y;q|j_bKQV^bx7Yw8O{N0{Ih%M(}k<(F&>DOazK
zL;Yap)+hTs{TcQLR~*`9tyeO$@@$F*>QQpJ$dQa&;Xqm=^4P+KbvLeeXGV^5BT^at
z#lvGd>v{OVR(@YFK|f!oRj16}%1AMaGe#$hO-ZCR?@dWtNzOK}(%E64&mV_Z@&uHU
zzS8I1bYDDmK`DCT$VhKQ=aZpdEx;WrHLGY|&97C)&VhjNb@^l@EVUahyE*Taj+M1Y
zK~D>@lY*;RQG_w-dmCAa-}*u%%{OS3#X3Gm5DKiZy34xVIJr*~Pfiqmu>trsu%?{b
z;bSKY%d}mktStTV1N%mF(%N1*4hngB0ra%WfSZ=J%!blW7@08Zj0KseWqY{)wy3Tv
z=Zw9b%xcb)Lyf3iu|UCOJpWr6@$cY|#!W2j4Oztc6hzg2U**M@?4C7V*GsR@K;)HQ
zLN$zD<Hwc#67S8!VFV+sh;sT}UW1V*uigLV#`mAF&a$~(*Z@~ydtErL?_*o=I$+s#
zY*pFB-x}+Z`~RUz>%UO7H=N^T*|bXc{QX|lQZdo9$CjZx{)XlRp}Pp9D`Ubq=?68Z
zypK(AhYot&?n$rRL|l4?mxc=6WCQ}E&&QyQnXU`KnpHl@RDIx6vwvU6HzS?pvglQm
zm#Qx9mgFU4RbxvRx-laLQ^lk{7Evo>YNaBNF}%`&@oy?2#KDamIDX=2MClp`en*Lz
zI^_GgeT327;<~lJ=j*y&&i=kmZh+!yA98A?c~Z>pOv0{afnX}U%!S1T1WIrIP6Y-X
z$Uq{e9gzichQG5ze0Xv+!TA|V5h^^G(e}&R27;AD2e&Jk@zQGoM7+VkBA@4~4c7Qf
zbc#}B7}5npJd0#nvqC(9^KoO{gK@hks^DEf%fsRO(NCFWM-uJ!x%0@XuiupUf?&c}
zHZ)ZkU~#KCX@8`R>Qn1j_N{;&{6%i#8fOU7s!A?n6a^!){TtilK#*+urT-kBcCqB_
z-Sh5F2%g9J>~N*#k!(s4iIAAG_KE5EWUs&JGTT4PD6W^8XuK)wMrtFPInNu>mk!5I
zw5Y*RVY&*r{+n1R{DEV97-mgw-w8#D>1Ep&8Ae8vU-6_WWZ0Mq32-j;Hh=7Q{C<OO
z25y3~1R`-S;{J}5ju!+OwOYo_5w{6k3ZthE_C;~7FE7n|u8x-UaNYYc*^sl$$XYzv
zKqZ;GK7BXQaSc$KLrT7}_gfKPd^*>4C{!tnitK8<57{8t8qyCW1ay_HJR?-LgVLl&
z1jcqUrs-{U+c%WO@~6D6hlD*y)PGhB<TjQ!h;5J+FLhsq&$aJ>H>u~bC;O)70iU<)
z`C;vG^EEfml1}_S2#&94WC+Awqx~gTTIN8So%U!$MYBe0!*{C#h7VoP^RyQOVVFhp
z+3&qrVZ~|=SVOY!)7}s`nm91IWWsJZF!h`#iQtM;(76Md-OEBfQtg=s_Haqkg&#(_
zo}#pO0o7lpRCcOiAS(c$eBgBwMwz44U<2V9AsDc{kfhXl8pev8Azj?yx7<3!iGU#Z
z)FT2B*lP?dtIrc%r!5>9`c9QO6}FX+sX-0u)pHP}!{>2VCo&@gT%}{r`RYd+-<+q^
z468(?<L^pGxg!mTv7+5ALnY-I6KNc}G914*?L7h?neU)5f%*F5UDK@lVU^_;H!)H1
zd;y=UZZ*ug2TM-kd8m+`lZ9c!x>2H)nU)h{0^<koax^JG@2!8XXlY4T+S9AYGYC^T
zywyy>uf6i&CZEm%4+J-I%v7QKBKCWf^5^hSg|q6Hot9aZW%T^MrcuSuy@@p7HZ;`i
zaXK?5SA<TEA1>LJ;zYRJ4c>Nsc3f>km=>~XE`Oo5(>y?VOwkp-;$r5)j*|E$pYgDC
z9=^@5%1H#{uUa4u3ja8!U~%<R@&gp1>5k*;ODXjINjL#`7O6%}t7Iiwty=Ru(~Pik
z+FWuT^?4_|PDxq?&;2oimub^YS-h@3z>Mb_wsM95yXx=rXKilR<%_5TOh@N<a%b9<
zv8Iu)^&23VU+eOIz4YN}R=nKkMBeJU0{ibldhmns0e}L4WW#9wuvz-RvQC9q&}@*U
zRxe`lTG#e|Q%#C8J2X6V{&)?cbgA^<uPt8Rw{i(WU7-zB9go4-lSHTY^D_m$%ctHQ
zqOlhON?iR08<Ef7!Y+9lB>4AO27mh33hg2PID}_18tv)ES~!Ef?Rzn!M$t-B@1+|H
z;K;>lj_s4e?ytM)@Z-siJ(GA63}l@R*`ABaUZO`)?0b!RLp$@3I?@tpyfOupb=#zR
zzn4+w?0$4WD@Xu))aYr6|JK>jZ8X!-Vs?O=T1RDuQqSOVGBv4=?vR03MBP*zP;%{%
zFe5eMQbj-=BiJ^6p{g^;=Jht9!iuIim#58kyDd{z5P*WXYMAD-I+<H4tyyE(oiMG4
zBIbz@q}~gV32Lrz9!RmcGo;U!*{srVfpkk!`Hk%4fC5Edm9EG8B`gU_k~&Tnb6j;}
z*RzC<cM8f|#dr!{3+``l>!3qp88#VOZ}S(~`1Hz37Vgs|OMcX-$yO7Fm4k>Z4J*3`
zI-jx`FpbvmHYHXt-+|xk+0YZ!!9qT382fAPQuB(=oJ%+ag%nyfMUF`Vm#eyT>J`pO
zW4;)bs+9}koZKbuKjA}_C;O5QEVJ^Na+5(BnWqR3suIQ)9FaE8jn8`o18cAjzW40y
zB5n+mOo;OjRJ}Y_BF5p3T7mBwMLsh|t>v04U~Ma_4|mh;_p}Qk>+Or3|CbeiRPbNF
zKWYoU9q$1t*EeU9VA#@D{XgfqNtKs*kgGN)I|gK=#L>(P1-xj3@W1d7-YytZGG9X+
zVM0XC=hnZ~5JipG5HyVp9j&vNt?pB+>$;=2f-u4KE>R2R)EJV-9J-FF<JZwQFSoT~
z1`}M~I}Jm2rbNxLFyZ~0r5Z(P3if3Hdj%fZ!KHcKie_k+w!L5326L&lZhMmBlt4*M
z$s^#V40ufTf~(Gq<Lh}fAhTiqQ$XbCs;J){2Gs3pL#a@$uRwxTq_b%Sabdvq`!jYU
z#}!o3AZWyv?Le35zU}`!{CT<Dn@404)#hb6%L<@Kc%V1UGF^BI@AI3QJ60v1t{cBg
zV>qCYV)yFYws$L889e<e*<!X)khd`uRHVW>__7w{Xc#7~OYWX+xK1*;dsh^QgfRr)
zaTdehB%}tnH*i>eXaryS12@1ev~nE#BhIsy<bleyQ>SW0>`l^J$tDV7RZG`-Ix}Pn
z2gPDV<0a7MUWsCIbu`HyGW2x~EVOZccyEy>|Gd~qBT<Yd*RFX|!Irc7B(wUbL4%Td
z8JweV51}bw8nD7`Gx7TM=}}(ARu*QG`LT_)Icz4r-vFzG3leiOi+2}U^m#cc;eR6s
zn3Y?GA+?{7y){xvF3KOwYX$41G}cWjWnv`Ed{C&6r<RV*EOoE3p2*!|9}fT3*l#Hw
z+bMu*>+OyaWEHNI1i6fHmb&&JQySHf;vx+2FTAWpTe6apyvj60<DR1e!A=WSE|sQS
z4J3Klv8IjDg@M&p+O~qbGyOJMp6&6B<nx;Am9$AyZNr};{<XN$j?2gxk|`9*npVzX
zb9oG%1Z=rlFCD+=z3Jk>qFo(;byE71GN-jHn6?R)N*IF=#5G;M{oDGVw@U`0@`@BC
z%cF&7R9PUbx={OqGd=8$+5-?5G@Bb^*PlPW?_Vwxe`_b!Y;my}1irVCDiXp4MI3}v
z@`zmf?VS;upxa!CyWU@m_}kGWL=7a=D_zfA5up&k1yyC*t&#Grza#0PyY=J=x!tEb
zzsjwd<&=3cYW<_*wxJYkJYHc{5QKoZnwRIkkK)`c6fo4)19us8-v}@=_?wF_<yJMz
zq>YsHXVF6h?@YwlS}sd_gXWZ#Ejt;Z{TCZM>VhWw==ZS7TndzIy?;(X`(##vAWb>E
zg-JDcqSK^lwR{mP%2S;$RGwL!Q*PVk5DaSGDoVVaADnF#T<#RZhLc>#BT+zyLGyMl
zv6DjA_Z=}?-j43lse*<eOL_(tFESxu6Z`VFBz2j{m>r1_Qod_t?^T$t^0y?eB3Zro
zs&ft`=!dMfM?$lyNJ@}T(8i~om-a-N_!K*^v88S8|55dpQBl5M8>p0&bf-u+2n^j)
z(%mH~HFP%;0s_+AIdn)2jdXW+cPd@y@&9|*IqQ68F$^>Jv+rwPv3CYrfJHu}D>h<j
zrV&`Rxi7NR0Dgu!FNTopJQi8M)lo6}G&(E5xUk9KQdgMX5W)h8b3K03MM52Sl2`VN
z{1K)8*R%62hn~IzZRw=P>DK^E+3>GHbURt7F&|F46}g+RPE)OWlrirYDs+%GSwjh+
zVZ!4h47c2EjRPh^cb~yXRSWC$qYVCluWy3g4eYAP(V<Lb#U1Wa-swDi1e$J1aA{;q
z4p-wRY2>mf&SZ)uaKuKg{+v49bet+7YC{;+WrFJ&1^e)tQ_nIx4W{^y%SC;^J-!kP
zr#YIsjU+iSpxB^(j+A-tG*E|944XQ$KSaM|e;y^T3H@^;@Nd|^z<G0e+8XG4h!N5I
z?+#z+>harMl!ksvTrc35ul-ya1u+qhSm1H6XN@xG3wzFq^J4F}ZzfpKld6^9Z0s~8
z472Iq{oU%78xWJePqoYvc>Gbvej)svFio+A_P*965KU4S%h>WQTLMBUcW+3oYeL}&
z#*wPo>Xrhx>$fNQMx;Ig*@@oi%hjSUy1pp}jp>vDYRs{%d!;~t4D9CpOHS1#Lwy7d
zDga(Pk}2m-BU5z9NMuX0Z++>*W75pcUf?oFx{bFOohvDg+7*xAHn|6)iXulBRFzF9
zPS{_=i#J+kz4AdEM;<7j(ipXDnp&ai6)igv#SeKP^{Ae(0!QX_yRpI8-rBiKlQ62u
zRLgAVr}Jnv6q-3C@(+Iond+N||6aQ-Cn}Y@{D>{GI62d|X&LVU&P&%fEo(;M3Q<I2
zc{R9WhhT8T+o&dz)`ABU=tIvoYOz9=)2N14-CMvg`>!QxtXbzc`*gOxCFEcE51_o#
z304EIq9pR%xM*#;i*9Ih^q6=mA<6JFS+kA_gQY8uAshj3(nD7>qE^)yvZSh-^fW}5
zyf<_yAuFFCIW8VDM}Kz%v`4q=_74D+@pE*y7k`RqaXQK2hOMm<?|}+27UzoQ;Qbl4
zWgNc#mYov{D=)4H8zjd3)+0Tm#pgbv!|*Bj>k13W`>f+x4EnTE+&_gw?A>>g;83P~
z@WSUX`JXnG?O@<`Yg~6jOyxUwjj!g{0|n^j=T5<473#dXZu^sT8>q^c2l$)qX;M1h
zuAuUf{N|Lf9s65JHQI+$$$RldmLhm+{^O<CJR(t`G8Yp;miJjm-MAo$>L^ffHSV79
zkTm;En}_Fl1xt{RY#j@0Nb(>+yMf>?u$AOywi;BSJNvt&jYyK6HA{STHY#aRV}L;D
zoR_G@qM_;i^Nj8e&3$WWk!M7_jYVX^D#voA;-`q+qZPi!xbmf@rd67EMf@Y->+n~f
zRWd8UL!Ro#x!i4C*TBDE9SUv`+|@Cj<7tT}v$ns(F0H@Rd8-vhDj{Jm@yR>5(bMrx
z9D*HAV~rLiNDeDNy%s7az2&%~Q=wfc4(1*)M^NAa97xHp5VqFS<;HluTGa#TffbJ{
z)|c6V#@HEVEs~X|AfSZ@fx6l6A&B-Jf6iF<*>6AGzRn!)R2~28XkKM8o*s^?faHAd
zLD49%P6qxbB!TDq@yoRT<sUXfkw>Y9Y)&6*Q<!>RR$RwuFlFrJ4Etb2V`Mc-Jw9yr
zkOc8fr}D+@{8}gf;k!~1r6MetW!L#JL;kfID5|XUt=YE6p#zsr<qQY3M%eNs3rbbO
zk}CZ(d~2B~K?{SdW5o>7C9HAxcW`$6cfA$~O(BQu0UK91LKSAWTMyMZ)Jn}RoQlfT
zO2aG-n+He6VCwxkCqsq=fZC|i6L~}8K&4&1$URA{!MbyWT_vqWj~%bCwiPQNMnA*o
zD{Tvl92Sq@l;>)po+ni@Y$iK&nbw#q;WGb(RWK?_uUo1<=!|w%sU6*eZN}7PnPuQJ
zcW7y{;X6lU$?Pi$->RWdH7CiXt~o{9!M^O<Z+bU2$A#lBMZz+ZP1GlZT(VRrG?aTt
zNjc^AcJyYUIb|=3(Z4=Dq~EnJUxqyFaIX9mm_rxVV{3;0YeSF?&!%3F3<s*_UgoN&
zc!fS{vOA?yhYfkhx1j9W&(6`<O<LNtw<$lxD~zA(r$4<l8(+&&_}L)t$3Y@-M;Sf{
z!ASF1L1S8Fs!8nz1oAaA!<i{lYvfJFCwWgS9qiyUAU$U7+Sdp5AUQ5m7>)MO=g}+R
zbZw*I-I<D-Bkx=mdAiuv1I_?2p}?sF0n)*c4>FneBoPsit**+qhGp%vP$DXI#jE37
z%B)T~jWsxifA8e%|J8ruHsD{i7$aY3H|?J{@(0THxu$iLIxXEx2F-ah##x1a+ZoV~
z$)Bwbn~TvN*=EQL%ZfEPut7^QNs)S0<HAO%B&0V)D;EK}3n&v4^V5$9`KYj6GQeRX
z#l)w@bsz`{{-VB-#wNQhajb7H2uMn|xJDkb^rWWFzK%FW<w8s6==RRfq}BWu$a7Aw
zZn0Up)|BkA+EPvjy4k<ZmlTLC&3iJlWQ3kBx*t!4PM2&f>QGi<MF&`H6{)cZ$VekX
z@;`%T`Q7oH*a^ecT-N#c&aci93}vP$w4!NLQDVEm^VI<Y1aFNzT^B#oWi+kv{}!$2
zn1l%|%g40+kTl@n44>uqz*ll94BReFr#*sS_5esq5&)$hrNu%T$WP~16Vz}RRtZpY
zV9F%KDiBbFK}bQRLi*GUXXSQ1vOo6p2>>67$wn~TI)MfRF6>uyAMsXx2N|;tkxeKF
zDDi88NNy@RdXZm47x6RP7wnqmCdM-F!%)XWL`cI`$G@{-l_epx4YYuNuOgN3&_+jb
z&0{&OeoZX9CblLTtzx2ehXtsUlPoM3(XU5+tM045k9Lg&avGP6O$l(cCNQTz`BF<~
z?>3FPZI`=@_p2!9iS-)6y<xbe7Ec#&#m|H~ZR^UQ5h2zh`LeQ6+OAn?zCgwvir3wb
zdQ>5*G=xR#C6lA}P(KPw$@fkIj|21nSlZ6pfc$@0<KNT`U~#c`iMOo0ggvXbyj`2?
zS%rVcCGA}`K5C$HLJwRGueE5R28uF7P6<P*>4TVVl+3<c_<?4Xfr$MYT-ki3wB-D~
zsv<*Zez=yQfg7*~S+h!qu37{a$zA1Ylt$zeLznF5p`jHss~$WaMFh4rj1@FV6Ei~0
zKC#ZxHcnOP!2qq+Hn-JK;J2l|7;D1@%LwM;Mue*a+p|)A-=syu5-e})1m(ChX1d}y
zS!#tTthvsMpaTR{y_V5&s5&$5;rFgui$$#CebMK&CgZr-Vkt^TJ+D?5ZaNv%bJlNU
zAVwzKMDg&)W>s%fwiuC{J3zuqw@>4AWFm~k(~7-MNWj_e6|6tdF$TT<<4M|2(C#t{
zh0VU`Gv_2}onqB;R%>TB+iZ5$TxsQl#3^ZH<DCbo9x9<ZiQ+}UzNICMj@DMcEWNeK
z!+TgWK{m$5KhtI26}TiQ98JOcy3G92chx&U&Db$cBFPVJq0p}A5U73r_`@=nO};D0
zKrl7<(@cUHzw-k_vBKHv@hBb>M348kmAe-ux_g;mPV`dT@&sb%lzJ9Iba%IbK$po6
z=^ex9n;HZLruS|a9{{QaFrWRZ`y6dr#3h(VRFtXY5Xa}KYT6)u^JeYk<?-Bg%Og*<
z!vB6&Yeow{gE@f-#DQldnySgHwM<171dpO14ezAsi(&xFxnHDbA-|selnmTce!wJ^
znM%LmLJX*X<n0smU$b*ZzTpuNypBsdD1$@RLEHWVSyT?q%iMNoQ9UfWo|Gw}MEiZf
zdusz=&-rs3utf~MUz?p{KZy@`k964%d700-qD_%;_WdTdE01aKv@8lpgGCY2KiDH6
zOM#s-nEJ7;euTAcoq~?JGdxx<5yyEQ7O(L1hfgRJ_hNdm4b?7d7<eq{Z6^dWM{9Oq
zPu5CeH;se6LjK(;hArfv^`W7=Q!-NYn#H}z%Cs6Zgx`Y+xKYJUPFbT0IX6-Uw5nG9
z^p2FNv5CljAlpE+A8JVm$+Ee@kyC9JiTzuHz<}CS$H&a_%bM?vzBRs@Zp$|(5Xtcj
zw<`$mtk2%!e(Pml`dPONHF@VqXXs*OS(==<l@ov10;%~?Gou`Se0*72?je_1Na5<w
zylt1S9HL>x&NsH%?dTP)7b{Nn9~SUzG2mBiVcLi&)f<=4yA5iLW4rU~e)N;;UG(56
z1(97c`)F~>m&ozR<KDS)qVh<EhDVECn?Tdpl8ijB>vi+RCKX7EN-#**-tnxga>+LY
za@rbVM)55F6@v?ME(IrZ1lVa>yIx6u)U)e4s*aM7ZQeigX1)o0`wk`8x(#y3qQOHN
z1;^xJW9IUBbDCkFX2Ac;eE2qf%`0s69`U`OL?tJOBI`g_$Ho+R4z?fxb*$ltam&;a
zqk+8XwTr=9q4_rZHfT}PJ7HzqKG}ERdD-_s@Bb=F1tR9t3%+E>@T`|qTrZ5nk@Mom
zvBgY3-ZUT`!v_T>sxY8#$qfoSrU_u(a#Pv2wf?O?wp{lZ7dcnRH8jE@D>XEHAU@+t
zl9g+K@l5}QtTebesVzR3tX$zltq2}bK@T<UoY*UQhoXae$Pg0e4S#)PV^7Mlo%83#
zg`B+`*!CM>oT&OUF)!g1sT#FtLoeu#+np8NYgo~)4eAV+`*BF<%s`M(t{@u#lD@@?
zEn8PiVraX9-h;8B;Z_XNS)PTQNV6^BAM)S$jj$MzDe`opN}52-nyJp;aBtK93_k$K
z69(LT$7R~nf_e#JGNzwY`Z}5s;kX8oXM3F?wGDuydRp-xY36MC;<#H1x`;hettPfn
zDA?dmYoL}Y7lWyRn&Q|qji*L~Yvc|rNH8f@T=@u&^<Yj_C>U`thXZ|OJHxk}!Kr`#
z=2<R`Di3AI;i|mX0DVJvB{hHh#OArV3sf(n>02c^KAs00e$7NDf<e@#UM}%)i3{i7
z%ypld%v0?2<4-vB^_jIG4m|v1#&MNh!uX7?iBXv-<Z3@BmCHZ_DJ%^}pE6tNUMNdu
zE2v6|>1VJ$X2=i~m8DG*=PtHxvNEV2N8PTy8e?pno;TPM5v~5vMqc_pNs%}hxBN8*
z9CUT)zV7>;j9_v8iR3gIC&{T^%t8YRr6cNwJ~`HJG|km)yaV{fn&;}{qfiWJi&MJD
z;D2ajO(<aVcKN(jk5P+^D%eoZ$1^HyTj}&y0#z>CwVi^YSD~o%rF{BN0V{4|CCR)?
z;LP!R^aTQ`Qgg`sXLHRnO6@3RUoNcrqqbh1t!mP|<Xa?j7IL<{L1<!UwsWBRu8$Q^
zHQZRrzsQk^7nfBO>Ci~Gw^i!Q)jEB^4F|S!9{gUdAxe+ON}$?^ZnOQT>GJx%iWs?k
z?5hN3j!ui=+0v#Te;q~GuA$CrBG}<o7HTuj(?$w4?Nu#|yzQ}WCmXyvQh@FHo`<`2
z17oco5}Qc&`tgH%R{S3a`i*O4jFaH}@cnEfrzvDQ0FzGM6I33r%lPr5?s%!~V=LsA
z+tN`?&b))!$1(h+sRccdo0iW|V_YCv`TY~QeM4vzfr{Q_MJh9i^!A;a%m?gng>=Ef
zs$%!Sxa2{{WR7S)HAmxQXA%Y9IU0Y|P|9fS1ki$Y47d8Inl?EJga@<S)dz6_kEvdY
znpo)g2Pd~Jsqy(z)P@63vg#NZMW_CZwj&=xGE-()Ru%+@8$pZKpaPj&EU_&UV*Gry
z^m3geg&b8BXmnTfv5!d#=qU5k*OqqSS5N4s-1QoMg;NE*T$9|aTdo5|OZ2Q@a;6cy
z+6>IE1S~dH5E8Q%CsnlJ@Av+#Pg%3-uJNafnquWv^o=X5*kHWwMrm$4i`!QcX`g{-
z60!rO+y6lkQJ()fp|p(r+pPX90~4Z3>HQ9jV)Q-M5a2$++;zI*t_0{!$wcS6ME9sd
zeLm3bL$lAr%|d_R#}%EJ-7*|kwY5LBie$R)d<nC}`vn7opEn2SDl4n1);s2c3@Rn$
zmI0_<viaJ+V}X)C^kIh=NjsDUaia&3%CFFx-HFS2Nq6!=Hl$eJmJwVBQy8_KaZAqB
zG#g-tLhzN|N#juW255WIW3I8Yw*qtGb47WL=;18A0W!ZoJcnW!p#`y@e^W+(kYI7n
zCK~4pvaHPKxygV;@J10ASy5O%a_QZB@RzQkzT(ImK<4?M8>020S*zXGR!uV^L^VWr
zVAq^EHZ;p2M!0#S$u#=(_%O3}zCg$Sd3nP);F+0LnsnL(R99&ImJh*9iP}zc#(`3D
z<$xL)G2H|vY&2?4hF}0kk9#D^BE_InCL>3fmrHPekFR+X8k_PN6RI~Y#6?b*D`PNA
zp`&$DW&)Mn4NgDhwg`61&<6U<QRFes8Oae>RenR4vC$H<!}95Impc4|lvAwa0(W)<
z%et1XWe-`X?l+pF(!CUyoTImiBStlBXiZlUOaM=DRpl=Py9_5b%XJk1QPpz&77rJ5
z&C_vx0|><HcDdTqMjnoXe%|NKfaJG_YQDf%$$$S8$p(<;rX%gFH*WQ7w$??fld5gG
zs(c7Lo*A82a}^>kf4<OVY<~jhdOxNAhdUOYfhNZHk-*4@kE2;<j6kbWw`F{$?cW=C
zLY33^gDK`*JL>LzC^-@+qYuHpa01rwR0)N(7BPiUIZ`zJ&J1*R+PoTO?bv8V&GK~n
zIFE6F+v$45)_^HjfN<Y<SzspQD!pe%fqN~>7B&>;?0G~O;tP*@l)M3{Xt&AIlcKtD
zDprTZ9cP<BD{CoxNc!_c_O`o(j=M(@5*8(8_j!O`;Yf8TIs{@G{<oG#>!k7a7ZYpC
zZ-uAz=i{&m=9)jOwdI#tI?<Q5tl4gpNEF0M(EfeC!Z((mlE<3gnZ6sc0<*IGj6wIU
zby|BqT5w`iaAzYSMIX(6y_l3H4G*|2_zd{S%qeg`h~$VBeJ?W3!22;LlfM{<z38Rr
zhA)tJXrab)D)U>%*R5cd7*eLd9nfc8OSOa8HY{vt9Ng3PbwU=`U58AeQov%u1nkV2
zd2BgNilrR6O~VLMq;Al(A>1PzCKDuP`pQp<P1SE>9iRRnca(#BX-%dGkI$<A%dh**
zhMsJE8$27bZ2h7R{_po6m+fxDVCN2z`8^#^^1_mtG>$KDOOEvRvqkZSXDzcl!F_(y
z*?O%hOO~Hzp&bM~5ifJgE$uL%$H{*qS3z83l|Xeu3eBf}7iziq>yYK=WKHrjgM8a$
zun|sCK){n#G#SrV+RUHhLId89neL0ycc)asJ1?ige_8JSbDJc4ZUBv~LHpBaa#HIF
z2kN}dYlQ^ksjvc1k_<t_u6WHB62%~f?`;6$I7;Zf!B8~c(E)IQBE$%dffr!M!E$z&
zjIr9X89Asi0hEmh@S$*Phbve=E)|3Q`mBoC^cqOHH&jLYVH7>mxoFJs)W4BMyGAyW
z>B>_b2{0s6r5OMq6VPG=H7RbXl2n*%^Nb#z?lI0g*8kQiicDWh|JD5x*!aPG*9pqF
z7Po4kQxjEUSzEio^LqyDW^vkA2lbJP`?h&N<KI@WZmQ+{6xnG-Y)Dy*OJQx-NKZb#
zXB)5lHtf^ws`o3P-OtTWw*YR0#d!mQaGH&>dD(3j4K)<Cff=yaB=dnYf-U@?UygY2
z-w5>s#nm_wHV|h-ja4KTk8YSn(LGq#I2S&I=o%y@J^`Gzetnve2pcVT6^N^oykOA;
z+6=7d^L}xRd9$l&ou@f_Ev5F8ElRbizs?`xibYBTdo^Q#Q~&-rV5>m9BG5W*5Q{l~
z&+ulrX3_(G^|&<CQs!}>ZF{4ll6%q2UNtpAWArPRt5`;=dSC6DHE5<NB@}2WXydFh
zsL;+e@*!fCAREnTey56c&)YVGn^P0}ufw<NJW=>~c11T9DP%cmArp+^CpauC1BYpW
zS30`lV9%M9YTe?iry8t?@^<|S`{jnw7pvxu?N4WCzWd{U|C9N_*PRV^%fc^rIQ5X_
zt{*bYSV}Gu=DkKEn_s7JUXX+HQaG70BX)O!%5h7|ZQY=xRn&<Z->~3q>|}iab#7en
z&&twERJV+@=2eOCeWF(KXuA7rPICDpQ1!9pl-c39jhW!7WZNnC<)UNcbr13`7qZ8m
z)c}UTJlQZw4J8P?yu;Y^&=`5~=)fMdq`WKVW{+EsqVL7{fF_MImnun{g<Xgj;HWvU
z8)>V12bkch*u1A@DU}N@)%_#P#-;(8L|&K&7l}?4n?I2vNBCd?8I0-2L9>hd5I(ht
zZp|a`c+xZrFU)4X%X;?|M>JSDV3dTq<xiT>>U>)p*tG!=2+jk6h&R`N&p6GGMtN2R
z-y#Pab??k`(r6t=O-_!gBPC4Lv`BnVb{1@0v3|F=rr56^)FNkJPv-os;zg<9bE|7T
zgK@BU&tA2qiGfF-=by9l<3pjV>TZa>J`-7EU=1R@GeURH`!CO-QICqN-cIL+LBEY{
zOCeMTGO2itvVB8^)|n(Cma%x|q%=-{kbV&>sKmJy&ap31-Hv+)RGhKK+DjIPp*c%j
zglZ$qZdXm;Nq|>Uyn<=Fh|X$5QhG0ojx2_KKesN;5aY|iVqz*U2Mo*FF&{midA~oI
z!yvSMTDI@_6sEu;Y)W*3!gE$y2K3CDi$Wl4{JsgHpZ63iKS9hNo2_sb(b4)McdgOP
zi(|;veeNdvTLxK#ch(f5uSfsStfu9)@mQg^?tT7EwjLcci-ch*IH86p{;|q8t`>zZ
zmN!%EAUR#~$Agu3AQK?RX?>4g{*meOgv^XvLxLP)`ZL<5_*0bLiv9LD`42N^Sw&V}
z{8Y=@HIAptye?&r8`pW$-#5U~A(k6L!Y+oe#3p|qvm<qm+crrZ5>KXPIyl|BJz`zu
zPaWEGD&fsW7~k@34j#QM;9_TY+4%ch1$}5ngX?~i-rKj>5{NnaCrs^fbppO=h0%~J
zNrqvY;f)S-gTWcT!EQHre}jW+@Qm-MPDfJT)gCYE&*G9`8yv~oQ2Una8@S^0b)a^v
z|MES%?-xofm-pLfP6%9RRV`~=T0<#SL8F61S1D<j?Q$AfMRsTXx<&tsjJs_1a7iUa
z1|<z#D<+cMv&YtcHx-v`$gM|qSH1SNc3iGfX2?8Ji*!Re-%8*4dUjm!nG|A5XKL;E
z9YQPxAVCK9wV8Jutf$`_zvN9?$EQvy5XgHR&)KD{w9ErJ$GP%jM1e$pS?w@6&62r$
z<Wmi<SR9qcfV{4k1Y@1~Lkp|YQq;ef5=;RiZof9$&)b$}@CqHe?)R59+&U`(qQ2wx
z&tm11Xid%aXydx>%A;v7LUhO7)_xXnUfcGL$`)~GT0I#N3m_yH#HI)mD4nRG+u7q=
zaiS|$C$HsCe>5v{&)OSygm)S|6`sEQ=R<qa@-ps^CIh(A(Y?ks&Gcq(pGeqV!?>R+
zR*vmRMmPE&vSb)SR4rPawE;c~zA%BT%Jxf_gYET^E04Z^x6o@V57q9D4+Fp9SexHA
zra7rm-nX{|H1#tpr00t&CF{bNET4V+Y=<~p87@S1hX!S1du8yi8^iLG97S@z!CAYK
zzFIQ1eFpW<2dKX#0xd?+UeE-kg<$Ikp=({JCteJ6$0%esjMqrNkC!+9a5f6E;rV(!
zvDaiyj|t4*KR~FnYM*taH~lm<X23-%m&Bxzyhgzy_;;fF^<kdMHknOAH#&glm7!C7
zO!$$^r4L?x-7uiiR#Vyc>27d1{-)}E3g><U+|6ngU?S*zzRh-De+Cq!4wq9Y;!?%l
zcohZLWhgsgK84t-Em_ErZT`e}<JPfpBbr8C2*k_-ZCxB%g-^r*Vfd+#m+V2%vsd4z
z-Kjr|h8>HdNo}8Ns9_<;bHHz{+OOD<V-tBD02swgd!q9uZbV?{<HhTv57r}hOTo<u
z_W-5Uyu+5<DPU^e;f@La^XBuStvfYezA5hBUFb}|G~o_X(xfMR<8zAZPS??Ptaawt
z)I3-GgP!sJ9LcC|zGwfbJ7CbxayW4WAogrv>?*+kjYwMieG#1fHXsEZ?2{0pWB@-+
ze!8#Pdfxt4CtIPipKF8rt{7bt6LEC`E3|2O-O|WMO!#+YA|TS*JRknQUa~)(z+RDu
z^H-muLt36adB4xzDE+xNDhP<J3dDdT9$;OK{?47~Zg%NTG1KHvnHv^Y%4<eTY(2`a
zdheZ&>p;E8_=eNA-bl*$GNT%^Ykrr)acY;K^ek(b3#?B!7>Gg)(mVtSdbpo<t=@fz
zPju$UR9hum8<WXs&<F%JfJ8uClCI%?c=*bN!EOTH=c@JxFIpzc#GsU05>1#7qze4S
zBHB;ZGYxRL+5|e!9tKazcoIH`G`COwK#yVUbY5^ceWTOnCUlaGJw=r2T{>mJ!>j(y
zxTfCon(>4&HO?e_Ce4+kvd!W~Snc)<sQz0&KfU*L5o=%QQhdiAfSVR-MIZ&{LIW1)
z3^&=R@uU%xUEtCc#R>G(YK#jc*`7EV49P$veipErh2<3~C!Z>$?~Aa%mNHez7I4h4
z^vl)F&gt3lNWUwGXnN1s)(eq+G^Wxt`Gb`HVT83S`p1v<UP+yDTE09I7T;E19zNEe
zDv*Z6?iFPacqSpUjD%&1q<OIjwt5D)U_FPbl73ojHv?6<t%pIaKJ+a(mFN70N2?qY
z{*TDRUA)jmRL~{&t<vmJ;JzRg<B5F?#S7|Ze3#k2@#yGCctnRp<U9}cz11^{IC?Ft
zGML>@$6Uqrda296myT#C=#Tt=230o<pAOgKmh(2F%9(|xlHvm8Nd9HGHL2Js57_=5
z%w<o-Se5faTpv$ogsP1p6vjPhQ7G&tuwaDLb72$lRD_C-fO5NQ_5;(06>!9AQ~H#C
z3NT<MQG8Hgj8C2jC3wDa`C=zNvh{_&MuOONM78yqOLNwJ#JBFifeW$|*#s;%sdvb&
z=Ux1e84L9>UF~}O$|bh7=|W2o-WMZ$RRVRkub1=b><G~^Y_xSN4-mxuGk5XM&nw{}
z>pUmtll*xIrI~dAEmP0-p-2YjbaY$xjFP33PAvk3?yq#hYh?;jrGhxC&kek|12iiG
z>Wvq`q$+*><621k1K@Fv5qe0m%);7bPcL#lYb<JP3U4<Y97K!!N$T(892vEE?&K!H
zsdqZ5aMCo4RR=2NcI7c&C>N^(#jzzR@sopOIvy+is{v(4o6>W+qERekG03@c3b7g#
zDQPReJ~Z2?KxC{1fYI-SYS7Da4p2}aPUUnFIGim#R7VR0yo^i{+2cZNk(3lqOV6?g
zKEFaW<M@YPin^xe^ksdGM^Qz*X$m3hm*r_gsUda93wR>;f7e@{_00Ci?14XR+naV8
zfzCVz`%s@7N2_j-U~keOZH=bhJ|;}{V}n3y*5&5po2+ubTKntE<LVJ9kkit<edzNA
z81D<A>SCb)z3BR-eBSwlbSjp+uTltutViBm;_TS_eG6(tJ&&VpiQ9?azll|DynShd
zngD5rAJ927gxKI!-+3Vnp}*CHUa`ZTzn48sWQTz1-5Ky_w0a(PuYEuy=*v-oFwaH!
z3ds;%Bgal5Gb6LR4l-#(PxE(GbVMp(PtEbnQrZf6QXOk^{+H4JNh2vksP&k)FS0)-
zb2<M8#t7CL9~HEJvLK~*h?T@Ui@GGY7bpgn8b!<>xyZXG*FjE@g7YF&gCxMwTITy%
z&NJoe{K@fua%EbNzxps{ew$|c)h#|bdMvb8=o9;BKX5#pWaQBmqIQ!u^FOC99h<=e
zu@C7t(cXOH3?}UWPfGncGK;?V>lLxlyoBz2En*|N<_*%Plz-fCD(;ry=Vk?%k`U@H
zSOVy4Ra1Zqq*s+l`Im7e^`JDFp9BDvw}Bv3DWtiCmWZqD`fX;tUgeQjBuRRRj<#|5
zoU#bmooMf(dLaRi(rxUu|8(?2%cFg=e(m{88v(ON#uDMJ>a<E}$kI%Jc4S!}dNlY-
z>dsQzcfcTYnk4fPLe(mCNhfO8$G$Y%%ukB6c`2mU@q9hMu5Lr`{w-=#ZJFwIVX^eY
z1gdM7leMfQ4E-q^r2-;*E&5)nEW`8l*V0M=_Zl)ZP)KwQa9jvIgHEdU|H)<NfkJzM
z*?>>1J4U#2ju_mO13ca8jK``t{(4m8Y{isQ+LToLa+qHeowDwK37SK<qXuOijp@H1
z%j}PO2f$LV46|x}E~?v<AuZ1>)#?po2mJ1VK$7fP$^QvQvK~#U33kJ-i6kOt$Bm;C
zwNVsFYReG$&%}x`Jnt!HEMcjoTZp3b+K1-EHhtzK#%jx3amvJ+<%I|hn=IYG@)Wpr
zO^a5P%`Q@IG30~K)YI68AG72c?&ID}#n%Dv)PWlPT*TF8i-T~J`}m0U&R*lN*hzL0
zeP5}3m}%T%`N#H_TcNVc({0yBU(-Rksu4j;3nL|a;M*Fh8`r6E0{MJbt48TV2*Mi$
zIL0*on0Z}xEby150$H+Ii@Hg6<!u!6yj`|4(C!OX33R~^qU-)Ch=oFOULNtUUL`jG
zJ4Ngt9u-!cbXMfTAFF$vG?Wqb8g%_pN$F*lq%F~IKo|PbIvH!Gv#V|`lB;$6>k$CI
z$nLu&ESKC)nG$3lYYyv%Wm<{#W9tYFgQdqIkUu#kvjyIIYE7I@;yRSEbJ^o8Gz#uf
zKGkb<cHRfLolgQWVt@r=Me2~mvhRBOEA=)b<Ft9$CERIlQKRHd{(ite=ehraj5Iqm
ztDC>*I7BhPZeK55ac1HIE5bWhT>?|;dJS0n7FS<@4QFRJ)%NTEG1#wv{wJsdT$ELV
z6g(j_vC{Gv!WIk{rT2d~wQIGR^J(RCOESgWRxVN3z$j-Q-R{kkZ|w<3218cvnvZ$V
zVT5=`PAS_nW#nwT9e@Z`eq<XR{!sG6JjT0sZ;g9YsxKN?jC2_@>$T@-e~NGuo}t+1
z2XHR$jS}PHxVXkb15`-QToPq17rAv|#|x%(1orFdYWZkVhn`-_>&`n~yX5)l#=I?y
zw_|@v-2N5+e1~B967-dMaP+!bk3NjCq!5T;o!{@b8MFecttnH(Y5F1rfDuY_w4GzC
z-%`G^;C`mR$6gPk4_~$0?B*P26+Ycj`*;DweInn!ecL<9cqd_ju)_a6-FcW%jS7S-
zm2R8|E@?hvJ%}Jibsa+(Rv6u&ZTl7kYWy%!b3wcsgHoh*tPpog(kDi*!`HT-@$8)x
z+*-t)d02~8#*_J1k6CN=BA6#F1I2TNS2v(fMX+woSRi%VF)$xfTM}>b$$^Kc8-9PI
z()I1pogGHE$FCcOQI{V(YD~?1LJWDoP1@#^g<!moxxT0?SEG$D17pxGI|Nv<BF_V>
zegX%e88`rpII)(iUy=FCwiWW4t&ibhZsbueC5}!w$C{rTmcLA}(*c3730r5z{lM~E
z*}reTF;JAGU1kgYUjJ7FF0dcu{^OfAItJkM{O>kbZ`S~8JM~{QX(9K@n+U`>24rc1
zPgs!;dXC^0$=P-%ELadd;D@J-M|IvG5!&g5vv0e0ODUlz!)st?#oYRgJO%Exmd!=l
zD2^2MqSz_L7Ql=cl7%4#g{d@jd@uRYvM)-%r1Ey8M+67*YebIyY|FCtg4+O>EIgN6
zH@d(iZfD{o!ne7f+&d4D2G-k(D^Ax3)?7(rojipHetjQJ{0;9cY|fY^fon38v%Nbs
zdmPQTMW68S*ds1&uKA%m13d>m$z@Jq95%VCq?baQ`V|IQvN<<FuX-bV+t1w=cs3F~
z`~G7bd9g#zjgy0Bvpt!2%<EhCe4>|AYje;ME3Ar;ulo%5gW3^}c>Az&z0?qEnI<d|
zx^3&Mfo)G_k$7DL!)zV==2Hc}#b;#UG{zG+49#mt)NnSFqm`Y^OgWUN)?|IF(l4Mo
zI1vH-<1HG-S^gGyg3j)wi*RINoq!95cBTM-vks;y+$3m?riS~QBm7S<Js71#MFy?O
z`Za6HB&C{?mYfaJoZ21g5MXcFzW2uCc3BnXT3<gCnVu{EZWqn7ejdnHnjzt;x7H$U
zmfyNjk(kcB6h<4F21Xvf+mQnmf>En+w7$~pdVF>hhXHkO4D)^bk8{3byY8&&P{hoF
z(>3kbfKtJl4bK7tglv(7FeA)%eOSlYa1s|CnBZU`^(*j;m2D48zC~xg^JMOBTF23{
z-{aS&U;yq;Rg*!h)U7(o3#I87f5u{8qgWI!X1-qgf@FWWDE*cbzO`8eZVp{5$(pNE
zn0B-jj$4gUMz+2Cyf2iu{$Q#88AFg=1YbJWs{iDx6PVS=g?UX%n*6(4B~`dk@3-~>
za{*6Y!skDd1V#W@`;pL%@Z^?u($uu=E<3bTlcpCny4qKA*$j9A$C5+8VO7QYI4xhQ
z?`?*?7Xm!T0I~xje6C|G&T6U5K4_5$$5bt1hg_gfJfsdt(a{ldRS3N$tRyyxWgzac
zLnl^0derSGafVup1{r!eAzoK#rT_NtkjQs+29UXP=>D3uLj87a9_?@Q-XW((*wO}G
z!~}-iHD%YjkfR*|wH5}@O|5>#m#1>;W|9kJe6H!9O|VjA5yw<)7+L$*$>{}jyrQ?;
z1;&!#!){!d-_lAmTLRS($CP^}SB(IXioY+>4|ZKya*0d-v?9cIk#^Gl<&{zGLKq*6
zpAXS1VSojvSa1f}V_b;;i~sj2kJoD1`j=~-^%G5_`=8#8&GS5G|J~)UWaM`@E$7^J
zek2hdgZ3txZl5fq#>6xB2H)|C4~qrg?)YHUhSY7Jwp%H9&zzFy0rk6Qa?B^UU`LLD
zq0#H~0qVTR>W_m}9sWunS0LESGH*}HfBctoA299$Q2Jf|MZ2?3RXrl^dXk_!FRGr^
zNXmvm28yDv1v{y8)jOPcB1oij@IPuuN$mjFHclK^2j@C56X~2a2u+a3Ll6G_k;dqU
z1kSvHfcIh9FunQJl=%!n(aw$!Fi-}y77Bj&wXJx_z`(<f3!kp4>1HP1IRuW*0B(90
zsiLp5%FA_%THQy{;+m|u5v%IkRDO>q+9J@vu1>PXbw1OML%@;2S#%CKc%GfwWwuB)
zxjhy;a*EPfVA5+mSP8o})kIr8RxY%!?CLwnho}3VgS*=0x~!?!j=<m={kFT`k-*8Z
zZ+h`c9#*kFja7(-9?anS=s-YHQryt+!MbT_DW1qKJOKm^cRfpZlU7*;AFoNyX<8+|
zy@7nQ!p^OavP7%`9!pFh9zoc*)_rr1g24xi-}T?HM<d)+zm3s*+QI446Z~$nEzfz+
z@%!icS>SmLFcKZz@-Q++OU5Otar9x7U4TE&6!i4EYn@+dt;q&HJ!?5@c)ZP3tvHs_
zyFgb)#+s>gW9rVpfZld4AFyry-&YSJzq49yzz}{aQLR`GF>@xrABWA=9?DyCdvfbX
zAp)LO^;J=hDPsc6%nXk6wa0(y)TP%q4I}CLG{ka!L=m>Cfr+nv+X}#EXqo9pYqfa9
zs%h5!!oFI%eui!T4J&e5z|>J8@@~Aj`=%r{L!*f#n)1vMW|JG54SeCTDfFC0U`IC7
z!C!IGj)!Mc^OM4MX(J*x3){d8s?h;UnpeJ%;3|hXO6O`<Hpvs^$oFD5gkceV9@~+q
z3YKd{V=5{n<pxqZ2)GzI<9Inl+8i{Cz2i`vFxW3QZ_EKy2?zu+=^pPr7qI@}Te-_#
z+N!whJqkVboqXFt*6Ucb2RCsAJMsIHd8n}(j~-Qi`sP&Ggmny|l-`Qi(AXX-TdT*I
zPDk>e5Ur$k?&CO}8i^^zcw)pgj+4(yVonM7wjc*_$)%gY&Y6h2Dcjs;ebUW+#sxG@
zWgffu<?|)5Y>ob@QO7?61XVhZx0fi>T@1(Tsx#(Hc^ytG&L}%svKr_cKezaEqXvIx
zG@7P0eZ?E{I{ga&$5VIjjk@hfwhl%O>uaex5C<dcd_RQd;<m*4?K~WQ#j`|_kRa9t
z>WbU39q@5=h#cFt_>-2p{vt{lK>5iswQ-H%4P;W?<JW!uKV$trw=5XLcD?<%Zd`cY
zC0?R~=rb)U@2POn7t;C@>@D~S_z#Im4r})ORQZ*?tOvC;pD%wg;Tj%iT<r~k8pqn~
z&PWqxN-?9pf6gF{rp00SQl3TdA=&Ak#J*{;QYAHExQ?s?b8XCPm`?iF!VQ-J{Xqq8
zljF%*^LLoFP@~a^<2|U{PyWFKmVHu4-%tsiXS+muL^-;EN&WAJQ&}N)!Y1IHRkgUe
zaTgZ|kgZ83fy&VXFMrJ}Ql{T}b(Z&MYp2LqI?QtsnuWMXS;X1}P^4NGf0s@WhaATv
zNPJ#cRAy({)lVu3NM1N?gF20_zdSTD#4EGa#4GXwYyXT2h0~x6`3dOyc}3HwH7mAC
z4E@G1sNYq~kN1{&JC?To?)`pe&2ohe>LMQ>?E}mp3!Q(j?)2X&jtSf<0UtAf18C@x
zf@RUVv{Z8|t%LmJ3M{@c@D@k>`fM&+gbqsMM-yQ1Z7SzU6E(AYK6sM1m_W=`uQ09F
zwLi)dKW@r~%4K&tRos-US*t;C9FnnRXZeGpRp!AYdAupJwZI&V`aZy1Z~{E!Z0%3j
z9Z{u4v&0HKR<I=>TP)N`6kw!<C~0TtWY8)My>Kvf<8~s)IxZj~ktZ*h7gY>k(TT6b
zu4QmUP0_eOMfV1bO)V>d0MWn8dX4rU(ZBJ^gW>bYA}hd_$Yj{U1RYTi8$Z+LfP?O`
zr^K2iyd{<EL}DnIT2!W%isl_Ulpk}9NYE(A@VUzkGXTaH*fSg9MNs#UL7O6*-;L;!
zbwn)qQQ#)>Z@!=B0MU0Y7g$ins$9`*(4mHyD#t05@(l$I1py}LKa!Q+YN_S(w}|J#
zKh`aFgN7BizoU(zl8MR=ZO&g$O5+!B@JZ7-;4K&lCXVrex0lMQ0QtY|GBmx3d$G%f
z7O_)+FkB$Zr5gW2hM#I%go%NvxUC@X9l~$e9M)Z2EN>%r%9gL)?SB*JffTy5ywyv8
zD(8H;jux{P7C#5C+0xEFhgKzst8Uf;h!Q0(ip`M8N#9LN9CkgiQwgBPB$T1mHni$$
zeXs!YAoP7PAQXe2FO?#uwvXiy;4z|Ke;j`Ve9k$%s@Vx7r55wPLwO;`GCyhQP$G7q
z)Rr^)tQJgK%AW-Rv$Jkg&cS*w`*O6^sRk1}`W(C6U4D#UgUK#2;SNf@E}X~qX<R21
z=UP+GS<Fvy45Vjw1{cRqK~iyZ+melJyFZp0(0FW(^K?@51Z2^9vkkGB-w<Q9h4uH~
zJ}~PDeKrZpTLdk;*9&pTx-n6(<hz|!+yKV0t%fgvApCkdyx>p%e?*-y#?#%j(s|pf
zSdDsa9w5$mRpHvidQh@5-zU|gi2s-j=Q@;N+{WvrnKJD;ToH=@#>VD*n)8gY&|>ez
zmGu`zhFN|q>c<dc?Pww+s7h%E+rB%0Gv_OJ@;Pg!_SkY?)R%B&cs#lI4ht#bvLytv
zWZ3Y=FS7jc@rgkJv*j&Ee{)6e619w^Y2I?8>(@8GPjaTAptgF9q??L+W4ZQpn~ouJ
z_p`97^~^f=@~bCDX5t?jfh<00HK=~p5Zd4#uKst{tjK0a%h$I_@u;?%u$0{65mVa`
zU?JunR85f|3IKjUb4wm=e&_CyCmg-1pBFk+IyWQF=(T47?oLBZ(^utcl0&|hikI`N
zB%)2^lvg0N4^q%N4N3zq-PHOyZK^?BLJMp{btn#`hSAk+(vX46UkN2K?GyqthD?{p
zpfj!TT{6Cr6bBQ26FmBhC89D14t<%ExTnF~ydz!!IW!-*rxa4T4gq3?qW?3rNKs<u
zk~bY*@#vdvnrYkm@N8Ou56Rz}0Cb;sT2<is$}^(Y=X(Ov%TpBD<oQ7~n`UIB3{pg*
zSIX#qdfp}jEl}xO@{^6R9co*!^;^6LO;A)i@q}`{Q$4HjZyT=b_kK@~1bzwth9lsg
z=Gpzfi=2u4?mpQ_7~puV4eF`<@u{O#jb`sSg1>jtoUvGA(v(@^5lv>9uNGv+Pvamf
z;%2Jn>;&d+02)@8Q0lG^mgx+@&X()fCrt$p0|s|Mw|5StJa9nhAp+-O$uF|7t{|Wq
zUUFxBb;XLu9{iEP7E0x!WY4aMtQpkIQd*F3Y0h0{=6ufz?NJtP{R;g`n&?^gD9wOO
zZrb3{2#lM04(-zMZ*c(_|C-2I=l9b!9mH99N`>Yjdm{b*3}tg35_f^@;P%WB8Qeiq
z_pq6q4PNHtVN`5>*xZz+c>R9yxsUKqv3*->x(VMPVekF^rsPKhWPB#K=Mf9fdR&sz
zj%8#1$_yIzr?d(zZe9z^;51b0l&N>D(4bNQuD=}1`EDw_(knoJ4c&MJvNQjTMfZ=+
zO?s4wj(_ilhO2<eM#CQMly%D2CJM;&wa_iFDDF}NUOjEDO5t2Al4zy)(3}Ln!3rT_
zO#v-UdI>5I*{N+w^Qz6o??^dT-g7~NebZ~Nc#)csV-2aZU<Fq8k!Y}W0mlF^B!X{Y
z4iZeiWMsny^vhvF7syxl`$ik5iDj;5L`k!VE%}Tv<YCMw1k$+`JgsZi_O*{BfkL<_
zg^Tl}V+@{WAJr=6a%8C|mTly1m6S`a&b-gh#r+@t|0lrtf2JUR;m1}8(d3=P>t;q|
zkIG#ZkWhrn{jTM9_7yjz2gtYvNmYuPa%J5}a7Xh<6?*oo9h%1QTMQL^z6F>$A9!XA
z^spD7$!q}Voi{xz-c%Y>V@9Y_PXW0{!C2KRnyH$E5Y|gK=FcXKGnIlONpp&e=S?RK
z#=K1XI93T0V0ZrMnjsG)f=q#??Fqo6D*GjGgalMB3*~Rgor~(5#;x+>#nwVG)xm?6
z0pNpR{Z$0bl#^dMr2fr=>T*8oG*RkED6r;0$Ng+;Zd0uft1&ghG8OgrWI;MRuQ>~5
z1bv7JZ5h&<nKg=Q>flh9^54pq7e_!J{y4h%gt)JEny{Ikh5gQOm03zUdSQy4%o^<4
zokT!3>KbTJR{u7#$AwLynj}q;cI5Wf{%$o^x(cu(-UoU^y{?x4cV*g2i`J`G!TCaA
zIERS_am*KEup%!oZ-g5JCoOGr{V@-8RVgZ^v$&mWl}yp0U8zZ;P)jW~OGbI`0oOcP
zWXv}!mA+aGZW#DOrc8m8^#yaE!$E~<4#IP&XqDvqaMu}e%k95X_(7*dDHXWgQ7%fq
z#9u{m+kYhZsboECK0!x|reTVMW{m$bEow!(-NCy;83G{Kd4POA8xe{7R-jKiH3L28
z^Td3EQ}&OMb2A4+F`%c0jQ%VC5pwN_f*$`<ZGEpnFV&c4Ppa>yKLxU%eeI1#l`8qk
z8)OM>Eyq34p2KR<<{B)5q9loWkT`a@azThuf)U?-<5}n0KcI@yZ3KTS$A#blMi4C{
z*U=hjGAhxpCwZDb)D!H|M;U-sJox~=m$t)V*vcRe3!0&=(05tFOF+NnpxfUx7hND!
z$>CF*a81U~ibt(g{XxT@X6o$Ti{7W!(El+|+05A30f@s97&?47W06)oXLWk)yVB^A
zDOZQ6KvoOP18wIs^e5a%^iZ41@<R(NEVfNMv;@tEi_ihIPFpNd@h6`(E1B9#uMrUM
zZ7k0@&8-2F26&UtPg?|=k10&wKbqlT313pB;+o>j^OFBEiJ?-b7(+K&T>QYf)@Gt*
zJN-3k=7>=W{oZ5|{K3s<_6YRa^!VPwrT|nM-gayu!P86)2*%bxI{=OSeA$d72F!$D
z|13CZn_B`j{3{uu#%d*UZe=XH2{O!qOJ0lfZ@Z^Jp?$69)5#=CH7EmC%(~24v$R{(
zaCH<pSI&_-ZvfHwW?M<b^RCLZu{2xf(oX;KtnbxKT1bIsaOUlT;E>}6&1?KVL(&-h
zdx4Q&aYY7w*If-J%}VU}sW0Dn3o~j1A{!y6z&4Ei)rWAA@vs%2j=TK-*qIBViF(BL
zNANI^eq*(2Y}+$N6ju>|Wtq9t9kbkpy*7Uz?L{l%?1^byzr-r5vl|=sz6HQlZ^K}L
zqk%edk~2V9?M{&WPC8yT_d8Y1>gK78U{@$Bq)kU3OjAohJmg^+O<tn2GOVAanAc~G
zEemQyxX5)FQJWx4>T@~h40^jAEzXgNAj?;Ir52g>k4=dddT>v;g%{1VLBs%p$3tSA
z=&c7&sK9wrDeEP6b)ZU#G@WsX=$e(VYxKOH8mMV?tc(rsGL{%4aFz{xSE{r^HZF_#
z<0telFx;rYlc@>6iP{dpp01Z61VA(d^%h09&5Pv`W)uzLIIBJyP(nuU(K`ZwD_Lg*
z&r;)TCPJj!2#zpl9YAG9xioSC_O2O?E*SW7ab$WMEop>bCP81V@!M^Qk-AFLXp#On
z(d8jA(tfa6nrr?;msC~}s{kI@FFE8{ySV*^&btvpK+AE#fN(Qq8v`13f>Zc-N~!}J
zrM-TfVW{Kr;hk4aIrlY?LUnz9Z*z+Scro@bQY^fCzAoNOyal;97$>^v(JVH^m`Y{|
zppy+Gz*k$i6VFb#_=Ik0M}PUXl=cO66_oJEo+M`6AhqJ%S6GArHQ*o8`fp$h5L_E4
zzk7MCNMHAo?$jH5F4*()i9vfy6^9C})$MV<r#x#VJBdiv)0Z4MySTN?S_I0)VqtP^
zBRo~0F?>leg7r*AhCV$m(@YI)VYgvk(bi`SEMtbUot;*#b|&rlOKO@dS*CZ5P}hW?
z2vlbw2?<mmKOFghacYWfn|E9RcL1wiJ8V55l-)5i?847LLN=~mJpa($o?YLP(A^DE
z)dVMcoXfYK3u6gU4Vo!PaAmm-Ccb%u|G-4xVB~Vy1%D6^>)u_N=7D;60MIf7h8~B1
z)s*^2tdqym@KE|UrdO=#?H<3NSr_Cq_8s(;C#f{#+B=~z*?0}7MnLt=>9%S$;C89K
z&rI4(cGCdQbq-x!QEHI9ft%ILaN5^_g4Bwx+|6Zh8=C9aUyF=C=F|BXX`6H13YxwV
zYfUK3_>h2ML_@Fya#d~KSc@aTvS5VaV)9=1d^BX%df)2*GV|FN$fP%|RyJ)%6@A<)
ztkFm7^`7)Dv*za!$kdhO)GJzC17=@jy>_4YiX*O(V=VD-twsmOtXLb&u4D+j39aIB
zZLx(-%C1tah!6JBmla%)F9T63YE&CqRj5PkR-2E@wtNRppWFUe{zskzDw;Ha&9`HK
z<?vam^szAf4f)vjHDieg2bpE>qv#?FONTuKv=2;7H<fqB1kSuRdy}V$blkKWPdI*}
zxBM2TPU>73EahTtF4v3hn>O1>tF>|`Lh5z8k5i@h>ebcH={2As49^c;MowrWjiiHE
zTc6>&DFfLv(T_u@qPVwDH%?8-P4n@ntzNn6$kN-zSw2Gm>Wd#;UTjmLDlrjE^t>!A
zoVf!Cya?kBJ309ckBF|N1Zo<VK}dpSo*2923yZeIUm7qX?*w93O{1q&+^P22gjkF(
zIXv@6TqKdJfuMX0ky-$tXEOjZW}%??L>n|*_mCLdTi9q)H!Rp*xg3k%#Jcl1Ld$n6
z(L(ux&cg}Ydq##2L+xs}!hp-7&WSwX@>|njX3@YxO$gBDRs%*HHiB28cTNbSDM-k1
zC>0IP7xC#56UwehdlA2YkcUkaZEuVa`TIR;V>M?O`)$Yi!-j31wbRZGc2*t=C6p4H
z61h6i(EvnBfg!2vUv<CVDF@8L{%eK3eQvKBuj7{4T}SGS5{2Azv!uB6x$2P31g924
z5Twew1k~K*v>YL+1V55%BecMvMUpGya+aY7K|$4+zO<iLc4dv?YXA}@01C$u(0WR8
zeDCy%e}3=$pK=Za4bu_*mKsi<F@yop(vV@)U`jYPdz-o_ibl00KMvJ|rHbV7F^Xrj
zdB=XI8UI{YJqa}uh%_H`$7BfHY96sRzmp<e2G!L8n*c+7bcYsajZIN&BpA=ye)Hj`
z%3Poa=hdn@EUJ%`!Clh+0U|*;ihvVldbDz&WzIhEZgzp#zNOXOn%_HI^PSXL`fcaV
zS{r(n;|5^7EJt^zPK^s_PcI7y;4%5eq}mT3sLW%JYA8N5RPv~$g?qM=jorJ0N^_|a
z@(AAK<5QwBqJR~2A4R=@Lx1=7ymh;h=uzRHHS#WiC`PC`;9(}z-mxY?&T<gfu}0Xl
zG1s+PTMLt6N!#!vq%ZMQ)=kf@St28oJiIrCcc;2I+b^k2Lo1z`_`~eUz(Ty$A>fW#
zs{ZokK{39rnwd7K45IM8xz4$FqHI=*Jux$cIF9EYEjwbjbk{)z42ytU<I5qG-a^?}
z&~cbo<_1ya?EBPppqxWvjj6y|VWnBy297&~D}Wtnp2Uv#ZhKwx9%uIQx&h6<oU-}1
z`OFo{8F@j>(*)d3E^gkRWD2}KXCt>N?9z5xz>C?|eojxXT$Lh0$m`<JDgpJ$%%3=s
zzVQCR0JxyKsuuL3<O<<y2c0r^Ql~CnztIN!^|M(y@L%9*?~`kf{%?d5Fn-ho;|W;(
z^4{l8rAD7-d(Dj(z8*T34DSwYbZR~8g8ZV+f!Q~CSLrBtcTzXLjq$u|UuiPCH`e9~
zK=b|QL%}r8cQyXY_?&fa7c#uqbg6v|_3Wi6dl?KpI1E*Sr-XNfoE-J=XQZr5E)YS;
z-v2|@JGf=qb^pWJwr#s6^Qy_3Y+KW0+itQw+1BKmYMN}@ll?pI=Xvh$d%W$x;5s_j
z-fMld<^w$(C^GuBxKbPF)`HOgl!{KlcYaOAoOy~?l_?!*SQy)4t~j7eE;zt7)OjAI
zL<EL9O0UaJw)A6pUtHFZ-LXWxsA=7pqwGy_JMhiaR1>4MjOFYdl5w32jBi#FmRj-y
zoLyE-xfosL+4j2oJLrt~<cZ%qD0vK5d#^8fH@NKorU2q{%y=QcV@3dSor!CxaStdg
zwczu62v(TG%^dkr!o%GnOJrdz#>>A@a<M=_B>eUFOqGt6GfbO9OB=ZwOQL|*Fobw<
zmzO>}Jux0&<6{t5fheRfiaF%TBshI7Rcmz$u~W!2BC<R~yb?LVExFMj99s1V-?yn&
zru%p*j1CVb_b@>FZOWm(!E}g!6r6vI+2xR`6AV?E9}JLlA3$b++sEg3ppU-n*qs9U
z|C7^W>JRs!k~CwDuk0GMlR)qlu5wxaap}5s#J8&7HEfpDXV<cr)Y9^U<rRkn!57$v
zrTZl|-9jebKd_aGSs!OTXS5*B8n@OMM7FzPH+^oHj|27TfRLR3Z&(Lh=($1xbq}zl
z*m4wS>(B|51?Y#2B4|@)*xJWOvg^=4ls>l$T(+I=m()6^;UR<o{*IUN$_e1%jG;0O
zyV*7}n_Rhbdv@C`7}XMP&w+?35}w>G=leeD3T}k9ss~b6CXOSg;9YRR6`55g%a)bf
zw!*jvJjl`}^gMs$7>t&Ih~=_;Qi`&a@h=z5BVyopU=DBGY*<~p_X01xNI?|<08rJd
z?{)II%nlDPOe{^D42uzKLXU-u7pH9m;Zw<0dN*mzpnglJQZW^ScF`%-y%C=V!y-{5
z*ck~CF-u5cZEyX~>(}W4w%ym8sgbe1t7UuII_N^JSu1BUBJ)89mU4VSB0dNIL51%P
z7U_08++)zlw$9MmtMV*^KT$6v1+_G1`uk1enfvDds(ZB<QH(><>g1<ZB9B1A{UN}=
z;6c(PUX0qeX|74Cf0M1ADs*?jW@;33iH^t(9x=KAbl>ZMJ-@Rku27X>e4ZdBGnfM|
z$cimZdfhIzj%m%pRx+T~{bJ4Yt-0KKp@KHI$!q_ISG8N!{OtQ5ITj#{!Vp+e>he@2
zSKG6L!26N&e!Kb6Z;4}3S#8sDH-F{&<45wR?X|z)mHeWno0id>UBO}such-gu{uT8
ztx$IGq|Da8;Nj?$E4%KdIvoUObK(Ho-%Xat#J@9|Tm3(oYL?kw6^FXJ-JU??*FmcZ
zd^W#Juo_$n5W8a2YbaY%TlMUKg}V!Pwf|hK+e^EP)udY|b-@?cJih<|eW+>)qC<-l
z)ddGm)b6k{ZS9o{f{}=uxa_#<=ehhy!Z0=M5vR0h+U*+AKFoiBS4l`-#`l}Nhr7F5
z61U&vkilYI;ke(sx&Rji@K7gZ!ICc>DwEkkf2g%xw=KGO^*hX<rm?Qf6G)oSr?G<B
z2~+R0;dZ2a5&UKdi&&1?He^N;NsGmra<Bw(R|mwetVNCx1IlWFZ>pQIi<{~4a#x{~
z2y?u=C1e*{%%aN0#5lJI6y<x83FKRVyw1SzKFnj9EgCD|x0SXlmbprMRm8mZ^K98U
zCX%#F7K{0Soh|?2T%R*P=p`<kpo$n0dWq0L!0(hMGgOp@RerMO@9)=e1FvK3zeEM9
zl>*<3c#)IMXhtkz6*)e+F#kB{0{9X|?|*i-0a(Vg;Ca(!&&x$oyfFW<$8EUN;<>v~
z3SgqTmq6vWdKP}+w@5ebigo{^%>C6&&HrQg+bK_ix2ZWJUp(Gcn{TfusZ_ea%Z;;c
z#4!!6OI$42VG!k)o(-EsSnxVnG&0+wmP$frHA5NUy;7nb!~f#!|1d3eur2lyOP$I;
zkN_4B3xZZ;w%rA{UZWCLRc{LSd?mgfky;gF>c!j<9%NSX2)MVV_;-J)DGGjONm<(}
zcj5v+{tYoP2e^1*A;SQkMu1SwX3pHb8(2JG*RHyck*;!IAs5ZOmQgB@HKDd%y+&G*
zz-UtbIc_jHDLYT5o5nG*ff5ssu2!oV(16#pUq2J3zG^{)s?rI+5B{9DT?mYm10<AV
z0}-NhG>7`#Y~_(1s!n>)Lh3y-QmL2~^0?-39@QolCk-ok>x6f~G)Wm%o#Tk7^n8c5
zK-BQdMfb-RJ}mG-O{Mn$Op{;=>+kg+#FQCVDh8UzC>s}t7@(&51!;|VMm%>`$)Aa)
zae_9z*0Xw`OZtg@8(-I18V8hr0P}<B{csYm$HE)dy%vYUl;w#qDRmfyIW6dZ8N|EI
zmg)-}`MFySTv*I`FEwf3Vj`Q!if_)5F4Z$w5(4R8s@<t#39C{5x+z2cL$qAR#yDW?
zo&%wWz}kB6a?cHX%Y`5Fo5Y^#07f8zDi{HNuW9JV{cxZQ@V2aZ-zpw^uk(DK9kse^
zcx+=W6zPe42NPc|gwBU>4TE&uevTgL24T6TL(^*9fo54&i4qGrC&wyGDLcnex)K};
zEuePtV4N}4PYr#ZN|u;=uWNvifFj7Z<9|o<x=z8m$sL^u-0?Hsj&-<Jzg0#2_`^NH
z$P<9}RHZ|B>;|NC?J`Uo#2aBbTU@4dRPyvjE(cFHqz3Ia%ydQt(Q!L;{QPmP{h9N#
zoo2X8gEktA4)CRI*u&rke3=P1)DHm)<DY)#su0SQeEr1esih&wJv_(MCV7m+K^g(t
zrKdJ*5a56d>7egw#T(t-mqqTSWz_TzUQ=b6URNRu8J&l^%pKA_D7XTIH)N_0c(P~H
zGQ?!0dWWC@szkk>qjzG_?<$;Z@0S$$TqC*&VZ;%Wz&ioN>J0mq$^%HM(EZV7HxO@o
z)oxh9O{NSJY$7&05;4e3L*rPE#8h;#d>#j-)3#4%*6<g8goO9-d;1rI<2DbklF}V8
zyXI}t%Hb8U2r1&zfY_*kb6e--cC*AThbF-MYc_A?b-IrHmZ4(8qh|EW0RyC*<vb}$
zmS<%I;|?8U9EWG&J-On?Fs;WnWLH$|{6LEibXWzZRs<huj}5((K^PauFK3-DLs7Um
ztUZ4lCLVG+c*S=ubMNNU4B_9i6n4&S6rk&#_mUA&fpB{nvqI6#7guCP=yf(-J8Rxb
zN3B{+rhW#kH4g@D;y&}AYV&b^-;7}6^nZK{?+9z%^i<kP)_0iz@(>4Od(i*Oxjd3A
z#gaT;m*s7EJ_wC0>GLdt=dO!}+G&9M(aUbyJV<?&X7zX-1Ej?NG?6I{yCr=0JvIIV
zKr)*`&fLS#q_lpl9XZSe^<WzMI*+WbtjPkvNQDmBLm2ISx)N)N%-MBHm&9~9s|`Ds
z)+p0E7TcZ_j4o+sK<7DW09s<Ij7SqzSWJ&IDgUAEl0M393q6dUgcM>jE!63;l2V~*
z;LNR=OzcW!+6cL!zZbGQAH1#At~Uf_appHjX_x&8Evg}RkHC5MPKXw_Z3sFsT{o!(
zXhu7(f)Ig$ABdy(z95~_?=?2<W(TL*I$bLP$nqqxzI~>uu=MZT-Fg+C=ZLWkMqjWb
zQ>i?wx?1cfh=&|13T)Ir%!o@21b7!1zE?6i;sd(~;;B?KxKE##n@7=kXuI_6+rTsQ
zY#7tJJMX4`UWWbnMB}|Hv^Y<b&BWCd*I@RyK&yRY&K8*P>Do4%tSxh#K_N!xCaGhI
zi3(ZmUnYxDzVmk9wct4>&s+jXjL^s0*MJT&KH%Ln@OXb>)@^}%S9=?y&HH%*G=L`6
z)AXT*SLyi+zqh(ovoLA7jgql{-P7H<H1OA16<IIk6y)zFO<e>iOj6M!C7=g^a`j4d
zH2pW;&L@3GnLVywb^k5ztIhquYLVBvr|Q-UF4I@;yfKEIVW~4|fBu2XLwTD*(WaQF
zZgvhJL+J_7e0bj0k9;PU#y1IY%d}mvZ`(ik1pJcHl+>>-#(a2yjjEH#ao}WzFKSP^
zdWps+B1dK>`y`N!ZWW%+!ERF!z+Fv1jBXvov)q-R<DkS<%6o=$nkaT(Xy>)?%@Nuh
z5<`F6^@Pk;56qb1(!a@l?{XRcOpA&VKktGH{S#16eE{6v<LQt;cz9!02>SI5QaNDS
z)Tn-uT)dWSc*%CsLyRQ|7Wno-!5i>kG=dx8SJYLWrp0v8g0Dq}S8#NTK|iL;B!uSY
za2-vX=i%}nX<RZ8xa!N1L*Co2Ki*ca(Y_LJlra&B<HjxX*J*I#oIg-|K87LLVAYBM
zYk13HxCS2&G_^+0W|fXfw89n<%2?axG7fAZnfS=2`}6UU0`>D88ZIg#;;_cMtumVS
zG?upVG&U*#wx*^Pk)#}MoVyR{ml?+=ft<>(hB>aZ8Q^;w%IH{9p`RXme%O|ls_N^n
zk4ZznEu0RaDrt(`so%{SC%Fqv-}r&ze7yR_O^Eb4^ooBl_2qFoY*D36oW<>fs$5Df
ziCiLY>lVa+36g?$5ip;jYByNsiUd!Ri=2@nzn6T47OLj`6ZrF9$-VCS{(bsELa+PG
zU+(L&ogph3r-<LH2T$_(iL1Sx^|6)XW~*b~vHMehuwhXN=O>&SqUIoeO){khj&_PH
z77{-h7wBc#@z2ejWm*KkT0~z4Ly<HZ5+`vhVi(FuCxU;~^;T>$F*uFd&#36ddI$XH
zY+?QF15ze7*Bg<^DcFg=c~4dbIrou8#U|#pcZ&nURa#Zsg_bQhdGkWWdG*cXzfeKj
zX8eS3VI=-`?@71@l6rb)?PrfX<~83JSnxTmRZ5Y>oYPieZ?^IIb%HC61QUPOJAdWc
zonvd3s{~oxIP&5?e*5N`T!AG1cxE|x+2Zb+!)g9wz{2oqZuZe*73Z+yj?nl~bNV)p
zOCjDOfd+%@DsC9VW1F|bR*DSg4mJp9T2eI4qp=amP&5T2rOUZms@=-mTV`YzS%x4C
zRk{RZjhL9IY2aypve#%F_;52V<S2x!WvX7Y{WR=S5ePB?Wv`~%w35hCVkOF$Yg9i=
zvJZDx$k(WvH(zS}2usJWQB`FGvBAv97QAi!1YD2;<~zj?8@h3*6w6p3g3!(zspJM-
zWhuiV&Th3CtQFOYszN_aD>MZ2(2WD+-vJe0h7p5&SDbA9$kSOf%la{Ozu1PsUkSK1
zVC1@Au5e#&c&iW*B3R0&JFQN(c71HPl;{l?ZJu3T`sVqf2=D29`O1*RJJd%l_2x7r
zUEBQ*<MS#%mB2q0p-F^|0%#c?b)1z>%u7^>*9>UNdZR)4j@rx%Cn6%NE&k*8zox&X
zNnGSjB(I@7fc@w6eZJ~z-SpEv-LT2IM|yw68B{K^ZYr<RL8Hvi)v%?E@f#<ESGe-%
z?Dkv2>};;~{i&tw`v54?374gJk$3Hp1z&>6d&v;`>r!O2MH-#j-_RT9oSWFoCLmqn
zON#Npts@`Xr0~t$In$zLBiWE&pfjU3Y=^FbeSbD<Y#`6IzCi4}kXDgh@m!~06>kA3
zRmXJ)p=Ql1)P_{#QAk7TgcZ{TwEZf<RI5A6{Pw9PJQSr$FotdU8ERPh(}tItVj3+S
z=qe<_smjyjMv`H5ts-`IKTsOpUN`$A(_Mt=D)(DIG>(7=;kHt0WCl5s(eAWeE(vG+
z;Lf7nTl@?qsOO_U98OU=)Th2&gYH@jVo~K#Jjg*PqFL1mrEyJ8mN>;eLMsl=Jy^MM
z-qk&o4y@&G9a~A>6=cL7Cu3_PbVk%B*8CM;Ffy5Qc_a^I8eMgx+r`S{qvHATMMf#{
zmA>Ct$;5kW)3@1DW_!Sd^}O`!_>dA!_~}>UMf*+77t(4$pNKEL$k#k;^7X|F9j%42
z`0%QN@JhA7n9+M|mrFdU!MK+Jk@{!t$RB^|%COMf2QBMtqobOoL=ScbyBWa~>SieE
zj%7c_-FP1-d`+Wk@LL<~e~+RyplC-T5`1jd)2JSN_5rV-shVS%CPN|tVd-yJF`>^V
z^@YQR2lY9H9O;`ChsC>lzVdnLJ>PfoZn{P{&Jrt?Ld~GPa7VVDcDe!I0NOQ5*!A0y
zdF$fl5jZRthEl0Gdd1A|`n#ZeWvVbj+`))cNryMcu%GNZYEX2u)(N@kgm63w=YogB
zpHQ*@qeUA;sb$vpOibH?2`!=&@;*6cPC==CeTioowtW&v=Gg(2li@+-6swZBUmQ>}
z;TAd-Q*{CF5}_3JpyioJvt%WW+FrN&X0!}FffoWxOWf;9;$E))U+oZ-j>`yGA=Zb~
zC2>j2x$+;omzr-+@tSajB}5)l(7%3Jt%3@qibdf#WYu7q8|c7te~+*>6WJ;2Uf_x^
zH0~EbV_ghKE;#Li8enX_`}CP|mSpfrSxxOb7`QcV?0s;I1P0QVQfyL@PeeoXDp+*d
z9z$^rsv1p|6=V2~DEHUB?;CnuYu`rf!YAJ^1TK$erykBWVwHKZri5*GhQH(d5Cfjr
zu1I+nC9Z4n&Yz*N)%N}$r>VzGGR=J>6)qKmdRpXZ$Z`$nc*C)1k6mPGF&H?i8QFRo
z{pfvY<-_M&97cMK-%*H#<?qH=*+<DAShv48B>#I)?|J0+_WFOkLoVy}m)ETBqBf`{
zMjF0W3Tu1MZ@zErlSpz8>L&;wL&VUtV^KQ{?WSx<g`haY2l$3a3F5>J0(L_{E97;!
zCZH0~^r>qy9XoOuAG(7LLg{Sn7Gv4~Ci`j_X}gWyQPiqQnk!}ri1dKq*|orTp|a81
zAdMQPV3%OvX2&u|i#*k|yFBuGK9>5>eRiYBu#O8R4zb5fnNC+KisOBL9mw(U5j&L!
zPb?3iO(3kVIL%WL0lp$rXMW;L1ANk%2Cz{qc=}wb%<OHo1SBmmB#Qfwm(2!kPD_5f
zd*>pM$m1of@e<C5e$BN<pzuKHhnY6{jK%KKH$FMNEW$Wh9K_)5dm}8|TTqRomtBWv
zvGVDLZ0r#QbQl`^$W*q|3UTW?NgPC6YuNjp-FoRIwE0j_p++V5O#+iK!q%&e`#gm5
zAT7m?D<u%P0p$EcOC$m@5(CDKz1PFKS2(HmS?OOYg2KLdKloZaBlEwlS(nA<UCk<q
z9CN#$uYHTt5WRjn%yZmqIj`p3z&pj`9v3#5cHpq<ao@Cjgt<C{i7u0x*NInw7jnQ?
z20Ms0FG*q@wcB|%sffW9P^O=4XoujgsY!K0B@rCk=a|YBwWCQLJ6{`;U0L+L`{4Tt
z<d4n-`tbjIdsiF<?v99kb);yT-$y>GXMeV?!R6;)=bXyAs9PrWx9fRlVW$W(_!;b~
zp>WIawcYpZ*1Bo!=}__O(wUOrJ5@6&o~kJvIJ+4EI0_s^zxMt;!h$nF4>0!4@EO4=
zCHWT2tdpy@q6wR*ph)iq8p+5OSC@@TOaw@;pPSZFgqcVV-*vqlRBJDqu(!zLEk>M*
z6CArUpdnRjw;IpW+QJhT$+`J8v!Zb@jNydTS+fND&6KCi#c_rOpLh{3L}+-RVuKN{
z{QkgNNGZrpWQpHALV`<b<=n_zWvol)0DcN$t^wb1Od}D`OG;$2f!|ZA89$<b&S5ka
zY{R>z0DGZ^?Ukk<k0BX)8#h+wG;Za(P-$w@V8;=MW&-29&e!ad1m$kBU}5did?_)O
zigVp;T=_ibjk&LdGi^iK%E-eq%?|Si{i<smXUFEDfk=7K=D)8ye=-CwDg{!H2(7lL
zII6(r<mf+ad{xqx6VqhM3qJeUkz?Gfm}b)1t6OI6JlbhlSO6kq{hhb^fDR4z^<7Gd
zRnYV&6mSuoOw@`Bba!`aew<&TWck`Ep*m#rt$M}i*QkFDrMR2fBI9qNqqpgy<K)_|
z))Tw9U>Bhf`|xdeg2?R@TUso0K~tQ<mIL{CyL>)Ay&UPjmYX1Q`S?R}aNqv#1z!s5
z_qaV`zv>=Rl=MwLdLRJUuCWDNK*Q2Y4HM^bmM$Q;eyf^^b7Vqk<w)&n`*TDbCcVXm
z&$7jUtPs)>4RgB$4QbcMLY<gvY5#g;_j7{G87hyfP9RfVBDodAcbQ{}<si7IUy10a
z5E~bS_iqha-{LK;vRH88KxFyd4<vp1u47fv3S~#wA<Did-f0hUtCx73f1E#hNq#f#
zQnKt4*y$l>kHT$h`SQ9X@HoaE=dkU&dSZ+<rO2=!ZQiPFMNBC4|H@DF=CGeDM9uj7
z7*zUcZ{-Ohw0IH~1(zvj#m>gI_v?keDJafm=`79AWh}G#GzuoiFOu}_c|yKM`E=Q)
z{t1)a_q+o+ltL!+OA3QZrM#6Y>ov^E0Qgs1<vhAXdn>PY4jB=eq2#R9oX{kbP+4MO
z2=JRZS!9d=HBvOvK~fwuXH6RovQ09zMYLyNWgIfGQXd)HmU^ScR}|iC@Df;E6<Cf}
z`rE&UFQ{6tJ%T<*zX6s3rE?SAdhbwP_+>Y>GC#+R!ugGRk<aeFvi6mbiZguI`~G4;
zlhy#L1G=GY0-<9Nnu57kImOWHx<s194~kf4d;Is!6<4@QV{B=F4K$*q(qoIW=aOPg
ze!1C`D1R1Ug}r2cemYd$O?Wx{OVv6c)c2p`B<sC3ns~`De-Q=-NytK>9)LJgP_yWV
zMy;r;Jw@@t+k*fH0W0ylZ3Na!fLX8YsMi1ew6eBatQ<Wp2+x7;17GpsDa1o?h((VU
zCn#+}t&~!Qo`QOdi5$z?Bv~;ne8pvL|2}h}Ty6Jij`k<)ThP+6%g0j`<3g2we7tVr
z0D6PT+VNk(x7$K`OxoVpO-(#fLIzI(Z>c_mQt(SFrVTwVlIc^oc3-l$enKPGsEcQx
zsy#S=Hc29c{7AOQ0SXJZ4#5H^eFz^qCT_2`tC?}7n%C1*phi-B2di7xOuG9yo_$5;
zkmP}746&dedFda;z!dQJ6~hV{TV{6{-=$rTyw9dat0A3Qky@&iPrsVrZ}{%e^m9%b
zGREc2Q{i#Tena6j{?vhavxzEnLIF6NkaxCUD_^iCL-GWUSz%u~KeW{W%nh+<plgW$
z12sgDRk6=jG%KEAAtWT^YKy~q^Vda9L*LMo)=t}XN5Fx5)l1rCd$A?}m>~YaabQAB
zN8-o@YP<ge^GWaV@F%{}wMO?mRs3Q!$h%c;4ZlDQ12Kl9n-0_Ky{3x8py{h5+;Y^2
z!}+xO`HFET&1b$A1AITPF_yJES_LTcb<4a<>G!93D3o@8Zq1v}|Lfp-U;)ox&+Fex
zA2A+e6kf4f;*vWUAd6F-4z3`ya>BM`6(>%jso0hS&gbrnrhwi3%f%Zm#Vkds=*|(1
zR&AG`s5MIUQ)q%2zJvD2s#RqWo?@ws27>Dh7L{_AN<k-CHi^>7z%lSoFBcY7lZSqU
zzkU{j={X*O2R4x6-zRMT**-aPmC&kG0)e%|7$<Rqwf;&_%h1EjYbrQc4K5=e6^>Bh
zTy>C8sMc-s%>IP+{TAw5LEYToSTs&Js!j#E=(IT8#OBM@tM%b0yZhOZSi`QP3`vfn
zk5osl!afo?36Tfzy!L&kqi*+2PGFk){ynzMx!SiCJ_d^2VAsOPreqxDqXE23wVIv5
zvBFuZ>BFSr_1BK2WBZeZFRm+;WFiMS+s`w*LT=Civ&M*H$<GLJ!kIcvq&ULgjb=B1
zVfGTRrDSke|NZeLkBpeg=Xzf=hIJHZ##uAC5v3Z-6f*^Zv1_r)pxgJwZMX`PR?!C_
zW1jR-7MHs%S@oLe()N=~pciatQ3fm-%}d0El&+zJZZ5K<h3x+LtM931Byz<mf@3v*
zABs3l=yWR3tHAIZz7)sb@6)suI_iUPpI~MFI{GH(oxXZVvh|<Chb{Dbb+wZ99w-A{
zemww_v4RK-Eo2A5&FrGeMmCWxTnD<lc@{|&=<m3T>|$VnA!6a(Ea*-MYd20Eq7GwP
zN9YLfua7D!$8ZtcLM{oT&4uHOf!kA5lS<ou>g-AaH=eJXQZ8W;#)unS+XxteB`F|(
z=RAFaOVom>&1e-y-=qaQxtlX5LnpJ2WL_&D%&fa~^#r?8EGG56y%sp!XzwZca_R~4
zynK#)MtAkYij*z&TzPZ_`y0EwV!Uc!q()*xMaru<8lbI!_lk;<nuMHVvQ>&P=DPON
z8B7>@EFrO`djQy6fvSYZIc3~v$hRAm$|$n%5wsYuJZs+AE?$DL)OGj#C{wbKEMp<q
z1121?C!3n;V2_ZoMUud+C)x)?N(JoD24Wsy*ljGVxKhWcRWl}7dL(Vtdcj&VSikQz
zM8zxbi>oXI^ERt;ezOoObsc8GGB1L%Zu6o^TNI(OD=dvaDny<r=^(MTBB~3;2FC{z
zP8#1M@jtu%XYji}v3>LaF4>bUvEjwNPC!pkw>-iRTIY7Z2!0T8dSPx=%;19Wzq~2U
zy#Xdu0BA2ka0+SNhWXL#5N+oKpaNCuH_+NF6?FqR?{FdVXkmoDzg(a*WQmkDPT;op
zlRg);fffz_cxNr8>YPvE%6TAw4IGNRFass-mgW+?2KnF9ON~k-^ei#GaL)F=d~w$C
ze5t9B#WBwxjs05B0xHBA>kv;&PmzIeLb>880mFWKE<+%IsTnv|=Qa||^#PVxSjy-1
z81;+SSE$<?k_{d#M-A^h_f=F@9J+J}ZhN&jAupAcha(PB^3J^K#Yi@v1&6(TjAeeS
z?R^rVgt3z`oaXZ^aEZZ;4>ojj?U<f*;U@Jzi=+|n`Hkhw55{@5?1-b-wB)0U)^AC$
z$)i^Y|Er39tOVF<C|+5_iPei{bM%3$N3iVNAqap$dft;pl=XluXXgAGvv=!Sj^~2;
zyE<sHP8cjgH;e=O0b_aUS^H6@La(1aQe_V(FKEIXjhVGSQGTizl{1r$T6LPqkEcxK
z%RoOxggwfNm<<y6<&>1r99w3zy948Rilpm_BzCp~5@wkBjT8gT;Gu;Jv`{hAD0Q|Q
zus_p&6`~aBx}>b}DcG<v(*j6{=^e1h<1x2UdSH86?&}k+*3fO=De64j_;eh@ITBAI
zQNlvo6X||V65}}VmM?#4W6u{9>-dE=Z5}~PrF1cn5*pu?K&hDHj061bu#PM(5+hQb
zCXEpfqk#NS*Hb$j#b4raPqT9`;YIT+w_(J;16j&_Ux7q;_eP&~S&rXKXrs>R$~Ebx
zoi>sb0m4|&f{8GB>KM1I%)woB)lvhCgU_YVLu^BQ+kR<Q^%mm}eyquHNT!Ap;rE?j
za0BJE8VSQ^`e*baJ=+8{7HYiZIWNHq-Wt-t7p%d$kh|uEN_0v_K>u?VnLHei-|8}N
zvxDOYKen{qVPlD9Gn}fhztD0|o98or48gSNm34N^!h1#XrL1v+(2{4_rs>Iz4?|*X
zds2vgjkfqLo>mRi|Mg_o>N9u>XO)b4p#~-z2v2`})5muC*D)v1h=<hg=6cx#rpTb<
zQ9{}zefFafhc&SAJQ%2&(&guG&UTXfS`gZxDpsIV{j(*oJ)!r9&gq!h4FMnfv?gsX
zgye)RnWEJogu)SKTB?o`K3Y7ocYrcmM+nu41_})vBDECr_mrfX@f>AJmK(BLgzV%=
z$a}Y)nwb)v>VhRl6&`CP=veh8`f!wb9SKrEJ~Q@oL(KCcf%>|7blM1v!9aQx$bIIq
zX5IU0lCldKb|UJjatwDbNN-%rAtZMao8<W+xOL5^(YjXP)}`^OP-XhuwI;uIp7#B;
zKA9B&Z-Tq7PZ5De`&syq(FN7%KPS6`<mmVU_yZCo2IF0d<jC<HD2xSgNbNY7!V2MP
z37El0y1@(J_tsWgJSSIh(AeY3f=!?@e_~EmM8iPUtNe6|R|<#$2$nOk1gSOir6-~(
zWNCWzJDQzU#k7(0yKu?@;ugptI%+0tQ^%Dx_M2CHAas!Q#d%9Vc7cT(E{XYT%qjD(
z`PgrZw2_Z@SjLqWzPs_qh!Mg*wWDiY8KBuJZzLvg?GRMJtd=s3G~*A>YGA>Hg$gtp
z^jZ!bA%w%H>Kh=ji;-ZjH8dmjYZmDI2c^Z|727SrpPl;0vxn{qo>4Cg$=uxFnP8ba
zzW1H=W2?z|_h9{j)oi>F%zFOskW0l{Uq{>ZftR)fqDw`Ux=l=jr4PBd2ID-&=w}we
z*WZ-DKBTsZo~pTy>vC>g{9eRoM|a3@-sjqtMr<X7Iu2qoS0@_PDdZB!I!kkxP{(*8
zV}IPzq|Yg$?|ZLLB&{!3<Qg(-qXdMG1HL$W7{NBeh^ph^H+$$Z7DA@{?$=e(z!%6^
zJ4gLTt*o4ZuhD4lYCxM0U8kTp-xOtyCdzv^I7;XI<6CDC`el%W!%y<pr2o}qfH#=O
z)(ENpTF3dUQN)l2wABEiV|(!-_UT_3DxF#;C{4khd^V``nfwZ}C>03nkfJ}~q&0%>
zCv(!Ip1-U*$U^XAmO+hr4*q-u?lwye1vbW~&65e>icw`?!j3I^YrDwGrUwbov5A-$
z*U-TQR;LFjl&H!h>m>dZ84iU?0!?9mik){^;y>0ZfQjLzig)9cR5)&;Cm=#m>&p2D
zQfEbh;8oaj$u^9>gP}Y`Z0GK^YmubyRzN^lM$t^$G3}--M3?jnYuF{uco1H_M@LhQ
z`D`rjuRx)UnjBvR%QV8<3Wn6{ygQuzq(T-?hcT^@@<jvh^&Kodg(3+}1<c%QdYSJN
z2xhL8a|9Q%`Wi5wYAPz`tl)nHxctoXfk10Cbb61G9LHcKo&q$gjC7+)`k{F0jKNQT
z@<X*KXTIT66;6=Ju<+jdO+@LO@>~ia%MN}Gb?oKtU!_&dUg~te{MhiZ{ey2wu_ZJv
zyu<3>Wir+{O^pd3Z8)EIDH7lfs#y3UyW4tey>aEhV>k!<x?Hu}E>3UABvtq)=`c<G
z#{u>u@T_)vGcp+Txl4_*(TXHb>`H@$pokUcO=A1KUeiWjJcrKl?Y)N*O-%k<o3Dz3
zPDj6u+pA8#die$><bg~}KeXfH>nc6PzSw2Qw0XS+rS56@U`Hrc%=ZFGO^@k`aX=)E
zEJW<UF*y4ilyMsl@AoU@5K4B?4Q=sQDZaE#2pXDrg}lKB2#))$WSZ0rQl(WhpQAO9
z%iESXH&RMS7~JWn=j*lEfb}+rdJ-6`TQ|L4@OPC?PyVVlLI#->M@nGkiO>zH0hlGb
z4KgR{Axc}f->Kw<-7h4zF7_M)3HU1$I*ceJf8_y{F|St6M~k$cw~ML_PT(qgWH-(8
z*`c3}(T0HUn@B<+@Opw2vx$l0herOMion2A0|vT<`P&p}Vbm`EIj<rbhazMg%|~ii
zXQhQ&GV02WAvt;|`QhU<Fzx-tkZ}vY*bVbD)Uu!hvOS0uZT{}bL}fEiTSU(GhG)$|
zVOPB|v)fP=_9Yp3@(B4YZiXnOdrW=ANS1ngsnVXGpA)74l`>vh@glIc_x_7!yfYtA
zfv?h}67Kjvr}R6U!*qj*{A7pF94Q^q<j*qn`sq@O%AIqIB4tMRWKP;#+J9P1g&W@p
z0_jyfDmCgA1$Ir~qW3(v{Hhsi{k6RoZAXjkL*x&&yOuo&|KeJXtB}6FfKJBuyF4-k
zMxK%tM6h<1^S6QgR-6SG6>W>+oIZ~$Rk-@cvBfI;C!--*<_eECM-AyRUAVBj*oyM|
zYq8c5N2~>ufP+1HN|R6b$B;cxEC9H@jThjCHlh<>s$==;=KZEhfsD$e;d5ILhGvJh
z87zfQ_WIk1LxgVY0fx&_>%_`FlN0qA=2b^S(BPgZN|oPFcbRPu6*8}7c=<p~nb=#k
zDI37H_G?f>O*IS(54ySDivdD{0?<IiTM>q!C1d!B#ZWi7!V^P66jb$|9k~K~cM&af
z-8u#=MEqAgNqyQNJ_;Nec9*_Dc59#$gDG}c8Mly8Qn1|r0ZsOS#WvzQ@QzbV?ID0#
zVR%h{`w(J(=e$Qsf<p@W1#mn@G>(MlKU7Z(+*-v~mFIpKfMe-r-BnP;asHK(*l4!v
zd0c<k<m^E~d=iYZ-t-(HCau;|shq#+(U9&HAwUULmoml+Y;`+Z3ni&8E(hg-7YBzV
z6{3`mSSpnN1(3)@<lL!)3VJ-xcB(Y&6MAfeB>T|fe4SJ9?&V40Rw=+N9+%T8PkhHn
z=QR4p$d|#||37@<>g7L6%iF%K-b->+e=tvx0Q(%ADcR{0$M=+nS^c8Pc~?-Gh!T11
z>;&A$KLI7F)!7(rtnELsiy3ppqq^|tU{><k0doUPs8H<r+9?Gi$zlleVDJEphgS<?
z%f%a4u-WN0B9XP{mct9@G>9yu!u2$tgb=OZDOSFHttYI!g&U?*Szgw_|NO4D_u8T;
z52p>8nbs%WN+Cv4;k0k&Jv8C7vOdy~=nOdW8pt%(*use-pSUeR*xnf<P%SMVMK)4h
zFvDR=^Qyeue4x~SgD>6EZ{TjxyGs~Kvs|1F3q_im0ZV<JK!6SbS@rOzu1+$QMqdUy
zFZ5Bqd5}D!Ka`pbs`mI5y)q1yeH8rXtST^)yGh-Z-7zkQg5q*J(UNb%r^97r8QsR(
zJiind^Q_3Sq;z)KT{YRQ^h@2nYsmclZ_8&Ex3?w2YkCg|LRy`T0Zs1BMYsEYRJ<^-
zy}9UkkT0$kTqRG3IVcnA29Lw1hG>+!cFPD)sMeq_PraHTmx7wJ{j{e2{vhiFZr)64
zp7iS$Oyqvb$yPW5m*2qn1U$EgcEw%x8ivEID&HQcW((qf^#}G4@Z}%k2pBH<eZ*QG
zE4w2j09GOV%Lp5Pln#y>ITQhu$LgY!jN4e!M~rh!ht&a-4GTYqA{>chfc|ORDdfeq
z2L~2RK>MdId-bq$S5MIpo?H%D*xxSk&_0|Rdv2DCh99qKpDnZ7Ys?|B1vz7&2GDEc
zn<++(%I>b+yidy4@#N52H+(OHfk`c{A>QL77<-U;Rt~s@UE-xo;BC$bc=6(fnJpfr
zOQ8BcaFw_sGbm@;z6DA6w2CeWGdQu4MSx;1_L$DQL^oyZj{mq@OttkHH|Fs;tQCev
zV|@1Vbt#+h?f0CuG<-v)gw`UqpK~TqU&`d}Uts!85I{JR0UuFlWZ$<{?OJ2RjxUfr
zi5a;D&0PGMNvs=dAh9Bu#!&F9Mki>yCU78=Bv|E;;f{Esx?$wht+H9S2-2ge2BarF
zW1OuK!1b*;XbR=8)YB2F2%hf<(I=O^_%wY0pLcA#=kJ|%e`GV_*sM1lS>d+5$m2!r
zzl>#O2H{*?E4d?$w32UxvCRKyrdJ9i!NmR1Ofri~@<)?ltGfO!)aYmj(GnOM9<K&c
zolFX}IAr-Wt*a;Y8iM?Jz;-i1tnn>@h|50LK+7|@?t)>C(yml^^YrKcrcO^(A_2G8
zS>n#}2<SH(M;J%%H?$lIO0rlUf_R%-^+z%mM1^g8JZ!^Gzk$xn9#pL&%z@>DHfQns
z$6ydz+HWeB<P5zc6}_^F-8<bK$mn5=0xNwyL83$^6LFI~cKUkH=&yC8>23z1WFT=;
zeGr9)O8KH8kkDDtID_Fuo-5C4*aPgtCIx36K^7l>)&{y@)(G*0`C}m*C4wBjK_=zN
zMP`J`hGy4bns+r(wUuRLD{|1jPOYvYuI6H9VB+<?OpG96K7n)l+su*^1=t^JcQ-&3
z_6k5u?n`HlENw>fBbC>rZ_0j$>+aI=c%BbA?v0-YqNu+p;a4lI&96#Rx@h+^0}zo-
z6zd;as3jOAl%O9v6U+@;DM_k<D$<gFKucu?FcOJbhihENzc`1SQZfaVazbHL5?fMR
zep%s&7RX`;3W4OY^C0ROGtHsPgDp&<wJX#ShyI|-KOWwB054_lMQ1!g>UyI$ujD%H
z{Uz_)!mI0uqR*;%bI`z?I!&2IC+D(YO;H-UM@w$p#`>&jW-WT8tXvw48lw?phF+rn
zpLo2Nt^2=Eg&RHS`??ItFPX~$lXI~mwGiQJBDEsz+!x^gIe{NaLQY3l#_h#NjB*vl
zvqum7U%<{bx$Z3hvK(qRvZ<#-3vJ+Rl;In*N{lMyYJt_S`TY#x^Db!;PrI+)wCsz7
zX!4%Lk@Qy>o|I%^#Ub;*45~l#A>v@zhElMn`w>M_GnJB~;T%=1>u=!6m{^`uEj&h%
zh<Fs3(3ynTMRud75j+x|eF}F7Epk?Qf%9$f5+ayGG#ORx>8_k0R4R?M@K7+dh!;s8
zJIL~8)!zW~%g>)#*FN7uR<uX5LI;z`Vm}5Gyf!9f1K6YOj^#yhW!KZm*2hq?pKR?W
zpIItTb$Rm9Ta+tMzD6D*au6mfEA;;pFmIOWCYNKN!5~|brO%|01IO6Z*=lbagfq2^
zXm&G{TfG{~)HqVv?7-Nh8-iYKsT&(HjmiH!@v6+qcq-aWmKIMeF!gTZ+Sx(2C+v2H
zSI`asyOlLve*n_&e!Oy!`EB0~!jTc!AgH9yQ0WEyWSq+(=!#MK$me{*oBYf^)eA4}
ziUv8bEsLm<f~0pl5)nml-O!+S{j-fm&CBfMe^|TEuEK_y|AGS#2Z0+K8vtcY(cth|
z_QV;*8|rtkv3?P9YG)TaedJd7u+Q+gS?LP<&sAl7v#fYm`5Yg&w9cBVr9)Jmm<NpL
zlg=VfNQgcc1Dv9CFqt{2pDC#m>io3SRXglfb={CD4UIP&<bx$%0Vs3d>NVaDhMc(I
ztR`d{yt--FArN}QZL%VD71JyySH1(cs{!wej*VlZQ&at;2o-sFxU3!Sv3n3m^q>nT
z1QpJtNYMTrOsJlM4n$T&q9er$#mohH^DM=%oR!+{G_<(sVf);41AONhn2Nf8Uy~#m
zAmOv&JdC_DSaIfMTlG}01AZJc=>d^I$6|0beF+l?WNQu%{7m62sbZl#rtl}scp6_A
zzobM0>-kV|WMvDq7`~4q$&}_ngBm=Wb<}J%ZkC)W0V0kN8)C}_7!I^l{r(hKz42~d
z?~at^2}}amqQ`p-^|Wy7PB7|1t^UJbIo8RZ6Pu;X3&1Vt<?Vfkd7TJ|XX0R+<f6vO
z{3Z^}ZR&;Y<`1LIhmVy?f`E{Mn;+W5Q1X_e6D1~};iaj6nhu)RkOAGWKw#cze7Sn3
z^UqIfQnYdh2`^@zxwRAjHaXt+eXY`Qa~-YRvS1f3-sn*a1288MhH$h=_->bsCz9+~
z1^4g+sX&M7c~fo9(#lFg`itiQTKtTSRhT%f!U%@J)$>(nEjOq#xbU+I7to-Ngj=<~
ziF?$gw&LHe&n@^pQKvC`*nm5Plgp`MD2+MrpH!wwppw`;LYS=mWe-$A=2^MPa%9f(
zn`#o#^UCPS(jWM=EwOQfXu_XGn?5iTW!FzyUgYEsa4l9bD>V9!BPqoShKzABT&`_^
z$p6Zy?at&|o1S<0=_aZmm)|cffU^vjV_mPSPwhj+Va0C#p<x5vZl*%2sWvBuq+P+a
zASEyNiNvA6ZZJBmHgtkd9n!q}3sRW5OB)S&@WE=TabdHbf!|FmFA$tg+SBLM0LBij
z|9xUWsJqGKl*Bcf*(&lVI2tGR(|kFItT-h#>KHL;f`PIpXL{eE?Y_gY=LyCSmqggd
zV_OZ;2$fB}>z_IDgz;lxVMSJfPx`;yjL16BM((-b=em!M8_jZ@Q09wK<j3|(7DGb!
zQ7aaKig-v&m%y*ucpfuNRR8|b#$6YBV4=)}Y;!GHut#V&Uh&tE6Bcn>qbdV?I@iPm
zhRJ!oN!*?PLGzCvQ;=_OZ@}zfF=vD?20_=h)&p0ql>(qq(Bx+~u#s_wA#v{<cswJk
z&)TXS*a*_BosUTXx~7w7U$aO`1RnC`V8#zkhy>!^sc=E~$M)~;+5MR!!+!*6^u8-1
zJ%_Lt{y-=4ai`>eeB`>jYxfC931&z}aE2{dh>I|q_O846P3&evr$<x<O`HDWw_j(Y
zEdVItmis6cbOe{WN1M!vv;B&wQ4<<I!Xc3>Uh?^i!`}EEOo^w&wgz8IF!%luihe`u
zJCqNQVma0OdOTg%;SDs)0ApcEFhW^}TsXyZ`LiUrp&SZ#<rj|GJjd<)097pRBt``F
zLTF{Mg8@uz&u>ag0N;@>r7(PZs06C}?dC$-Mjc$$4ZT%<)ufE0-KU;zQS*oH9yTT3
zO2ZRH@N8$XdK{r?FYqrsxV^>%#p4}6JYCfS+BMQtlcdVVuLI9G{iFecHf=!$BzSmB
zEyP>)kE~gY>Yc4RN|&|j9maNgAK;=qiK1Orz6E@L)ET~<kT@l5?qBP(y}iPl4!5+}
zi($fpdoGRN%bM@6QcB;l@E=nWx#Zny_VNCg7VQ(ddU@!3*z`B7aN`Wxs9|AE&5n#R
z_-!PL7Fm4iL}ih=L4+iwu)AsnPXz@v!x{mEh@TvxQ%>WF6xc^&uA9K+=x>0={uI_r
z@)4j)nZ_vU#9-rpny$0S9502r$Bz~Px{G@h+<~@t95a*np;L+tFYXh#A9iIM*n@TG
zc`Z19?bFjX_F<Nx9O5s54e<4k&sVWet|6Y=<sq;k&gSAM(o}9w0WG{<2HkUHpR;($
z@+nE8%k&2ww#_+mxX<Gha~?NdNaIKaE_SqWnGJ5&<~}C_fWX|CC9D2D4>(l}%Uzo|
zOC?UmFF3xC@5)TK<eZs^vYr`{J7=JAYnte9;Kev<D_H<ja+s^1Bwvbu>ljlMB3~mD
zJ0c#Awbw8s7NAqHk07<jE6%nvQuBLh6sviB6p|DVPlg5q5NOSWlfH}ieOP<EsQFZ~
z^~6$)Ju*|`7pJ>Ba(BqZP%<&6sWX&=o*SyrYCw<p$50Z9@ftOM{Lc!4^lEI$H7bxT
zjmZ$(>OT5#KcxNN#b-bL|A87O)C7{FOCT8b=50(p?6m|1p^@Hfw0K3?g_Y>R7BBtB
z?j0;JA*JFfAPr^qt@@G2)V=+P*ZepXVFfzdBe{HL>(pUnk+j+UNYL(f`5c|b;?62n
zBHjuTM&A@BSy&{f&r^E$%Y%;QRR_sRQVJ53#3++)9i9zJSYhJxG>DGCbU?G{n=wiM
zdFwjRfwOk*VNi0Pw4otk=FBJRFy%_-A#+6_k|kZ<Z&QLASVhAgvW9;Z)TK<B#p0O6
z2b-o0Zoz(r@;pFq5%$C7_grpp-!1^z>b|TA9K?})0y+jmPVbh9^Pz2U`rtn%17>G7
z)v&Cy)CU|MSp_sxAjMzirWWN@j*u`TUja2P^tX11lYz)@+%T*ecyoptwkZdeAUNba
zl!zl2XdlUmPM(T{MZD9|ase`u>4}(N)b`??{ZpHgh1FGQKX9YPc^mAKI97wUh#aoC
zNy?$Xz`*L&(<`3x%_-De0y>*i=O6%bs}3Z9va#l^2{Y5yxYghPt*<wV9k$w-Hiu9I
zW{}sIz?yO!`nDfs{PPu187H9do7a5W6E-)1-h}%XjQQXt;q$r;*Ll$B9Y%oJsl@qP
z1AvS`yN-*dFq3nCDgv14f9tWYsrwM3rO%lDAT#}xivT6TYv8|YTiYFp6Xeu1)8Wb|
z6qV=)8-z}u_@~TZ7f}pfocOJ6CSu||4a1ssbnJn*Rq40$j~;UTGc=SA8}?NP@URvb
z%ozz|_!{igs-K!W{p$uIdw9Bw4aW<*4!+x=xaeb~1r@zWQu1-uqRk@JlWp|4IeyH_
zTg2aCURrsNsLiS;cEdd{RR!m*vb;p245npZg>~RL42i2l5upnBJMMr|GY~wb2%i)-
zt}=ePkdK=;#xqG5uMb%iq%<<^7SYfqkDTP9`Ye_{%EyA%Wd+?XAFHyfmXJM(d+nkp
zRn4J@_M3?gOU_1&h1uX$o+)>ZArKjrCL$8LJ;t({(dG;4PoSGlig>b>G)%D)6z0ug
z>?638ZSJg%4{cS-=0lGjW)bv<5noT9;=8G7|EgK@Sguj2efS39;jK%&!BUC6176J~
zBYi0)&{nT5odCNJ#y|;MFvY23>e;RbvIKwB*ZV#p+d~&lQ4f$(W7siMeC1UkQ9Aa|
zdpnws$Nd1Rwq@1>^?y@6*bty?fP6vqQPEJpon0%;cMj8yE#=oIVaH0{8B<=OYESFv
z7!Fk&S(W(!?1<l%&P+l43stLNu+@KS7Dvu=SWvkcNp7V~@10z@Q)%j4j|Cpqe1W~N
zoKc&Fuycq8y#aaV2?%8JvW}M3_VZm0Ib7wR7?Q{81^U!110f0WL?^*;CD><o!^+K?
zXELX0(bgPQ8^$mmSmH%;#_EF)L8x+Ekh4hyRo~5CAPWKaPoBsqU-8o1!c_XHt0y1_
zyEznrN9<zF{bYN-a>w__s=Odgr<zg<GIo%l02a3H#{zP2tkf=+t&ewGHux+!LnB_%
z@F9d3PxE)i?+o+($Y}2PY0-I}3qYlE-H21-S#aSx|ENG_L9qFWNCw9`0R~I3Ogrp;
zzG6@Oc)s3vnEKHpc=${>C2*(*)l2GmL_Xdc>QLTWo986;dC{NnLSe#$O|ZC}l5p`y
zzXEY8@$;y5_)%z~UgP(+i-%}6+Pt?tOeF@cA3S^Kdi5b9el=vLS(V2=ZXB?Dd%_tT
zhqhn;$1gddX7f4kyc9_@&N~RsFT*K{K89eLb8=nln$R)86?kx%W7Q6)CNgQ!Ao;)D
z@sj%P)#t+n{UPzM`(#Epl#Ik+$DEpj(F8SfFl>FO;oIf)nqFUYMw!O!joV}%H;c~U
z0?qOWQJAU;)`;XJh^k0&LRsFX-4bInXR2CQP<j-EXX5?}Ec3o0_1?+!*K0i4<sKv<
zIBz&`^@L)FSSfJ#G>tuU=$jXMC$nQfiJoSrDH&)F660i3i3$5g4?j2%Wip4gCtc|5
z>o!J-D&lpojk(y2l*@~5aXkTSll|4>)iPi~DrB+;*kmcgioiM<Xey@QEXc?;6O2Wh
zm+UCn+lw_iPV)sg9Hh$c?3$C56uG~gUWgfg`2q2bjVSeCL)d+LHu@I+MBUe<&ihoO
zusmHk-{1hjrMYvYiJCi87m$GwI}T{^KEtd57R$`}%ENVUM_pS+0xpc_mzPUGYyFib
zj~Gqd>6;Y$7(qEoPGkN$M3f_>maZzYNb<4|`bdY4J4-TTie3R6VN8XD-t!{~oaeys
z0w8e$d_5Q@dj1(x*jUXlo&9;4Db|mty9tC3FlqlP@!DV&GhqR^{QhEnB%=6}K1saD
zlB7r-e-9TjVkiYh98|fBa*W3-&9(iTWIQhb4PTp$?_!d8KF%mDRL!0$OE2k}SVc`B
zH~65R9sh;-iIWShART0gcDT-YnPHhLL8*wB$uj|u3Ps5mMUq<+y)>Ykz`TZw2A+f@
zt#DU?@GN<&MPY7EXhLPPPi^3JO$|KfKvy48yWvXQyt*4-#)+INTJ)_ST<neU3Xwco
z9(LO*TYn#DAj_d(0ni#FR?YM7_izGO<I#+m=)`bKBdsF;P19^z_Z!I{V?@^8zbl_C
zm4GWjjYd)DedA@}rV@T`X1y^sf1L&kFHIJ^SW9i97{76g`3s~Y*_wWHN=OW$WX`~x
zO{;Y#yD&Rznh~2JB|LA;@(!TvMeL3I<>s139?)PsVQ{kC(d}>4_xun(a_)B;%Ug9y
zISD*wI@1f`BtDOTn32H>`|rN&(0EySmHPEAWQmTsx(04k%E}#}{?ni{heN27FUx^^
z+K!6KMp7MVUIXV6M_0v<*rRU^(^{UXmaTqCDUp|`JsXU~By!o^*y8xV8?kTfp^tad
zhcON7HjCuV$K4ycs7kZA2sa4kt#wEy2K|deO*CH5K<kDVtVLF>L8C!0qsbt}2B?ut
z$%Bu9dmD`zgODfcejw8c>Ubtf8Eoh(mQ;dIKQ{nSSv4Sydzp~C9=p86fR(`<%*MP)
znmW4~ay7|zHl7}r;o5KfA*cRB;s`c;II|uGWhfeAMEG5lS6!AEK@OClWh!ksldgT%
zvr&?Gh(3m;zy*mJ!uG2=0b^|cvi8wxpBCA(!an+@M&J^k^UrqOvgL=VTiQrCS%?()
zfragPE#YuC-$S|{%WlM|>o*@3xS7fW`@VdI>L;Fv&xQUvpRSw)9Jp$5Mb90C6oVty
zWgocHQ3C8VY=RXgW+D>v)x*Z>fq@w%$#MO@;*Q`9kVT!XwA*e7J!F3OIqi)FRB0Wp
z4ier|*ip?F)fsjqv2~$AdO@|i)pb`qT;dkRmN!v0k39%3_+bs#$i@9B9cd{1P1Z=?
z+6Y*7?Vtz4ku@d$mI=6>bh{$NIgR=v^kDT+{}0LnzX-$}i6{NIoz0<BKe~p=3`iqW
zjaB)2suGCml$Bi2d!RMh?FW9oaQwH)-gzqTV^86$L)7QbZmyb}Al+W2w}?g7pF|cK
z3f#<pnZk-dZ79_cTZJi91cu>aWOMPZRfsxrHVLd?ki|#|@=H+JmKMjq^+|d)j!4K3
zc)%$Opbg1VrVjM=H-OT#EtmSlHvuOVv1SM6oP2v>I2dMXA+<~P(VJ^SqHT3cNfnU@
z>~<(S<%CPnRwK_DYZQp$WExDaco5NY29&Wt`)<8EEb?%61JB$RUI!&XHrg$_A4PJG
zr%<~<M{dHn?Let|D?cT;mNShT8{k_aOHMVyi%QoYM%Q+IZN|X)^*N^9{a5VZ&Ldw^
zB*}_Zm-7E3>n(%w+M*>uT!RO9cXxMp3GVJ6F2NzVg%BWkaDuzLySux4u%L7D-n;kC
z)Z|D0q>5B+&R)A$ukK!I3bl}MD0VUX?zf;JUAn~8w_{|kjkn{XY-D1-FCFth6}o79
zy$^v!$wiA6d+!qQoMhR6N!`{hY4C4+yocjW9QOpRPegz)o`#xKvG`^d(~9T8?Z3|j
zy8Zk*topNQfVvWO`*XdT83Bz($vTn+miFy$=A425ujwGrXY25~3-|+UXh;GmJ(+AQ
z5z9v>dWM0@07gB2cv3%_H3by9uh$c21yE`JHkgaRBu3Lm8hb^-wLJbc7zedyNQQ?l
zbtaJ4AWJej8z~Ac`cLnamx5f^VU&r$iM0VegmHPP@ZFS%i$Mw-;P9ScFR58%mWkvT
zi4h6i)UX)%xyanbXB%(y!~dcgd*Vm4?;nIPA%A*;q!<8-N^miD-XXQNLl~MMn;k!L
zm5l<iH3@U>XL*FPs9fBQp0@obKNEgnadCqb!}X@vuwiI#0fZ4R%=KPZy$kfb`zeME
zHnWxx8K^0M4F_%&na()hy6>YK{LUCmeJkFINocF0EMUN~?PIl+o6%&Tmy(8|!?Kk^
zfq=HMT7;gR%x3Y>mky6^D!4Yl{zW?#l14(G<#G#l$rfCiq{|%yCcNGpecfo7ydSXP
zL{8z~J0C~D8*HSTz}aT7$cG=_n=*Qmfe+HZo|bE;k3f@zo_<Cszr)LkWJ;351P5_g
zQ;=MecFQDsL5JpM(0|R`BUuK|LWwWK+`9p|vGS%1jj8Rg&l+Bcz9|2Y{%ep_L_ANu
zD1ZgQ#tZAU2KMh1hZ68<B<#d)hl`5YCnViQd(?XvjM6+u0PkL}z3Zt6q7GP&M$iU1
z-^b+KJI1nW9nUqcMU&c1AOQODrzCIYxGHX^Vv)2deBPX+%2kE}zKKQvi^6w~^a^Ob
zLD_4KJe{YZ8etqU+&e~mE(7=dO+a>4>n#tfW(=o!)6$C;nRYp49fKD(3k+~Z&%+OS
zVq$UEgOS?T3<CrbsBXq$3%jnlH}v`bl*tnN)gfpB5x-c%q>hz8g@yFmEnk7CWp3?=
z>!}cwUQKN{U6SsP4_Kp7sw}*A<w0DWYX&qUUb8lJ#9~8xQBV=4Ta>1bwh?fI#EW{p
zmxXMA1O9E8+U{0ov)OcbBuv!m_{rD3{L2!BYNxgun&b;G;KmJTGV(vB91?#!6n?qh
zh9ws8y4u38Z1>oXL8Qv^A6Whi56)`+#x~ZQ_TS##bd?D=sF2A4@1Qxrx|k8>ElDC+
zFl6h48?-XAR%_X=OX|>$P&v#@t^S&`op;3G7PA=gM@fXksZ<ECO&rFIBNB{iJ^Gje
zv|;^5sGCIpy;W7P0FZe&Tem`BhhjY|k1yM~>XN}CSB(IJG=(FfX`Gb7AclqCxbxL|
zG!T39x?ADpo@h2-=vQkS1|ug?wbE3F?M1T{xobYm%I3Pu_>^5n*<k)T63oPnX2*dK
znob2j3(a-i5E#v1R0-h3aa=zc@3ah95pmJIQ)vz|i+Y4y#K6>S1}W`oD}f6kjtogY
z={=#*W=VwJGBw6O)!f$Jk_%;jm9Z#u8d5z##?0u|6a8u>wANu*86U6jGTomi9iv}m
zOezGYS2KvQS4!b_C62U{JiD(`vnQN5DKHAm=Y-GO07^j&3<d#u89he9!ew3;B_UJq
z`{=^;H+{MkmMDhi{1a!|YVH@c^0}`<LkJU?SFQHVWf~;JIKBuCZ2sKo^d!E93`q%S
z`r&kT@RuQX!z3X;O>v$Eih8luS|82`-)74a4FfRil-jOda1)#!0aS3V*O1}2AQ3+9
z|I~DyVyn0URBb@wFgu?20j$kn(?q@;c`-FF*00WG{@bCE3cdTU^&>~RM0%|fiHMDn
zaD;?CfK-W&ZJO8hlr6_WE7{hA@Y|*P#7MqgB}Z_Afwb2t<o09dm%3Mk_=Wo8{#Cj2
zzpx+2YwKzogwT(Ui3uBQa@y^5+baAq4h>DKoK`E%VT2eiw~2<$g3isYr9eWi^oFpm
z(|Oi%7MPY%<FGOJhIxE3#-R=m3EFI6RH854;(HR?)E2M_<>(;&cybNBD{iwLb8GKw
z_8IqUqcrSkVkVO^51CG-i<QLgwwUf4mrR&Z(-m${tm1c=z_}-h6>PZU-S(%$+&HLU
zw52}^^zp<K#YJ=&A>B@2bORatMFaT7Qky=CPa`p=zddhiMLN43BjVhYLf(6==fI@F
zg~q0H0a2cXs`qFB{Jhr#VfV50;k)RJ+pNt&>9e*l4DHvjMVQD%*|PqK)C-9m46I4l
zl#aDK*|!(fr02AqZZb^8neK9p%@T-LtwgKcvMA8itNlC<Y^#@VAH^wK_zZSZ`s1h2
zT2EH{^M^OCknPIOw<n-n7b>Q<y#Ickx4pY)_5`eA*Lj0&?QX|_dD(NPFOvFj5|{|m
z3q1y!Ky92{UT!wKa(C3(Y}(tRH|2iw!e4<JgdNz2k#OO$KZyfLF_VvhUVj!X)Xt9#
z;(wAldp>hj@65DC0LVsL(3i{be+6Y-C9wW{9`>gV%CBN@uTI!~sF+G5&?O{cxl-#%
z`i|Kjl~qsc{J?9nG3I$#;f4kZy2UgZSx((sUoi&a+5sOUP!Q2Z9w}*}?!Qr(0L&;d
zoGHQFY2jUCSquwX-ST+#oyhu#i%<Yd<}f-Mh#n)8DSlCS2t(;ex**>sn9mSTQGuGt
z$a|j?_;T<2Wby0B(Nk9O`wb~`#yV~hb7VN;0??P99@m{Hq`nWrLQR_*3!|GkZ+ZJd
z<W-T`>Qe=RCXa9WOyPs2m@K_3X7~RNDrP4$52}`_Grbq^L2U&Z3c5_ECMX2I=2}G0
zoynao*EGtl(pA&=Dx{K1>o25M14a1*wM*4L(brS`^sg?<@VWmOKHjkx`mW+>!2$OK
zTKv3}vG=I2KD4;7f&j6yOWpPQ*C6wL|8TzAkO}m>;d#<VUjFdbXxuh402J?i)7#wK
zL}9N(yDis_?oFwms{W|l<B;xwCVqcCn2#L1=A#=8;!#ySjhHHj8rY`$39qlyt;PM#
z^U{(MyXw^&{nBN~4?S}-<>3Dq{Vk$LTmcpqma2gFAPF{dxUlw0IGI2rwkxg*tjGZq
zEef7uv>!;~e)@@|c3%9VGz>n^?SST+-F;>`hGN@k&bE`@+}G?hIksG{&bow#vg}~E
z_n4&EE~iisi&|$KxKiTR4t07CWW~7;u;OaIk1`NiEZ=~VEbKu#-_my7WYQoeKHj(p
zvTi?LSOqm*z|dyhK7RSSZ*FfYZ%PwZLoJqoQ{R>TBbX^~pzk0-J0sH-EuOmkyF+vW
z%?9Dmb@-(>oI_@7P!kyWAE#!iLq3BnM#2j3hjDy%m+yf2<be3y$>&81?hC77xjS}L
zL0E9(5FV5m09tg{W9MiflCywE5KBEFC)2lDn-1mP3C1-X;_mw=H#E=^rvuR8{v8s?
zj;>w)w3=5L=vcU3WueDE-~RmyTlO;V|K7fDrNndoVl(TQ#a>i8JHNO{{RpW%kuB_i
zU_I;)NGKXXn-dL?R~`;cWG$q~wHj2|%7x4stgdgn%HJ{N4O+WdVLkXlm}9dD;WhBQ
zbyR5&`2&AVOqF`^h0uNi9}=rv0N%A_e}2xtzd73u><;?>`hYM&CqSAnXHlQSLTk;j
zE_1{~Bb6_~Ml%CH<}{R(BD!4x-mW8E<g;Hg|F?;i-&&rYlA*vVDx5k^5g`Ze*LcfT
z4?WYn`00V@J#)6OHJVpz3m)PI*%nf3Xr3Sk1*(DzFu&HVMmZ1f>Agy-&K*T^X+BNM
zaQ7~dFk2#8en8C87FZ_OE2OHHkIYZHai6^f*Gl2eQc!(IL?x1%G^Gd(GvDYs;XJjP
zXxh44+~80%i-#4T)}R_(Fj-rlTk#l{OzDEyqGW!dPMshdGo#!3`b+J!frw*)8<IZh
zQ#6`PV452b?p(5-H35=smBqKJ6|!osx2h?2z#Q86HHtZ91ZtYkr~ehU!}!HuqXAff
z30RO0PAk>jH+)~+e;B^)ATn!5fy}({e~#;VR!bcFIo!%F)!F<89Ft1;+!OQrtTUto
z`!mMl!knC$Dlp!=1=~$^4D>Ni(65NWgcA{l7GYmrR^a8U^9e4`3!d)hhygkKT!6H1
z_<y(&$!lw=dUv&9(IOrHwGdSC-azsN7%OKfpbfW6m39HC4&kuF%Og2e6j0)TM{jvY
zax#qESt})_yt4E8#10gLd;G)i&c2NlSEdU&&{9`JhO17wGs)lV33GSYZ5H=T${Kaf
zjGcTqLEEf5AYK?pIiIh!4R7(mX}j^w8Hd~}hL4I;=*?@KuyH)}*FlWV6;_BEvd5LB
zL{9l&)N2q)7VQ=UFXA`>)1z+>!0rPj?A?2pU{-pTzG=<y$Hm0k&@nYK>voW!r&|lL
zkK5aqYBS)G&#rt__mQR!^1V$-)d!!S&29Y?Y)`qCIga@;Np0ps@A?qlCu7ufNV$fb
zD^7REW_?x&`+^!>hU%u8UrtKo-9%+-SiM_Y;_sFD#dC>p&1GuK2OW<$^m)R4oBq$|
z81a+D>@(oBBhR0ofl)N1f50e^V4<gbe-0VCX+*%IQmzTSUv}i=)@_xY*<kGv1Wbve
zr$!ZT(_1nWbmjVgSWQM`aO13ghhCJs&mDd6(*Vm(k?$K0_17$ziG@5v=cK4;r#e9#
zKZq87c_eq41}SZQedc}`c?|oHml_kC5@xwtw#sI=wE$$v;y-X9LIQDmjtT+!<G7p+
zk*)KY%QLxZBm!d$j1l0DQoT>Gipj3}RlY7WXibnD&FuHQ9;d|pX>b=&O%+up8ClPw
zVt}!i!47BNGJkMa9mWFDdMbjJ@Gv}2o5-c(O_PR!a#k_S@Q(H5w4z;8JVfgVn@lQ3
z3CpQI5NC#XoI{I)C{CtZrEeFbCxHqDyUoA*5K-v&m}}wP)g!sc@-EUA!%>F_4PI>v
zl7r*=dkO$lUw;$7Lz(3R=2Zn8YNbPPSWih5udRrY51xP$gg*@lryq$=5>uWI-;MQ0
zqeA33dou(kcks_*=NUHNJwOEQH<CY0T=5uj71?919?sxsE%(>>QyL-_;L*wBv=ueV
z96sL2T?UH$(wolZn8A830M`k89=I4(k8Os4)XvMuh!|qY3G$NxqB9+hWNWCH4&el~
zmPTp)H7X8yvP%!uUCl~+k{?yDB`H;Bu)b8?{SgyW7aKUow~2#Jhc+O&Zz<7dnDk-z
zPS_~(UjVU{a@Up98~z(Q6^;m|7cA<?8Nx>bO#g_YfW0p@#zns?pYT^N%Gsx>bw!7C
zqsK{k|D2#W3l;%uv{kQHh%CEvBsh&^o;qUkpdWCiCGfo~HfwMZFZkvwyS6HElxjRb
z8X=7Cq>RgX&K@xl<awy_0Vv1HD5mtG!SMpvcoWdxekQ1S*IK{gWzLHyPxAc~;9B@$
zr-`^S<gDQ7{Ob_(Q2S7AV_|OQ@pB+qY+<kx(F@PQ{7LOK5;~l=ci5>jfP2Q=e)J89
zGl2U_B04&53H?R%*hN%RL(ZX`)-tTr+{GIL)fTts=Oo4ojQG^URH3O&33P#OFZuQy
zg~w|jkN|A^Kf~!}l+FS?jKFQ%KwqYGWvkz?okD$&?0$d?3xIRfL_TeejSta~um>9+
z`~n2tCy~@#WgwlpEgJy&>}S9CR0TS!UU~R^=3OD8kP%P$0`@pu-E*cv1&eRz)6Sp<
zhS_@$bwqiVH)`KOl7bmM@=F2P>9`cr79G3>9$FIjO@(!W|Lduea}WT~U!bqg{{!eR
zAAn%3L1MdgK~&3V-J|$e?pse$aeFE`6lqGqg*z`uGImbgPE-O?nW2{;V*z-Yse{RY
znt7g!^3d7x0rYLLpX?rFgdZf)I@ngw-33S@;0SgQ@O-f7AF?8XcU)ulL|HB&o3Ew#
z&YY(#%HJr{?$@MX6i<r*tVbhtXY*&PpigE(z<fa_&|x@uyBO__BQ0V;6{3xsgk?tX
zs}R!{xBMfe6lqkENl*P)x)4j_mPFco_?yF6BO@9HQd3?IXzYcGO{crfO29r1%yWqR
zkI*D!ky@DOHTg;HI5gbn*JHULfmWuu$9TP7H=9KPjQZ6A4^c-ia$yI{B@}(nY!tH`
zT^|_wbJ{KT@5p{^F2@Pf_9dlA`98wE=|B32(dM(rv@|z9yKi`N=4zGsSKi!tPh9JH
zvE&V)Z*u9lF+h+LK7Mg_WDE$7fyX}br{Jn^=##B&omHN{u>sM0KKo<~5Cobpfj~<x
z`fDK}HbnWm?F<<c;>Mt{Txc#bxw*MIvxwAZ+v{LMozgny^Yw?+U#Cib(15D()w9oY
z=>J4HfCS*T*U!pI&_kAA^}s2Rx%DN8%t?mi3t3=+LTShSQ>Zab)Mh;meyEd?*ES_E
z`n!%{N2A4Kd%WkrH26GCNtnsRVQPRHylqT_uAGmD6b>W#Iozs_qfg5{-$J*eQB|3l
zV_Ie^(0U|-6i@&X#7j`VDrGP!7h&~Sm<x*Bj&j9q#txu+j<=r{@^Y8kE$>}J<Km(|
zGPDyQ3<j?Z$_9+OQIwRr03VO&hg#|e&!&pj`<6jspA6AO>VEaUYG}t<Q}!wq9E$c`
z3{g8)Gd2$~*I`6vC86I8BEEut11Fqzq>1c?G^NtaPYL=tP`&Ggepz>}u=EA=>%exX
zHDv3xbwP!}*iE)1J3&qMRCsl{0?pF~-QtXqJ+|f^lUgCeY#`3x0^1<QWr!LWg-H|p
zd%Zkf13{G)dF_4+(0?5Wf}R@y%o}_o-&kBM(zNw@XXM%li!Zz^F{qe&20`L=PE;UT
z={rq)r}Zb5@scW^E6yg4C6CSGP8i07<NKT{6{yvy!X?ZDtpVk>xSor-3Vb}uo@b`d
z=Z!8v0@3D3`p)y-fAD-+;0hQwQqJn^c(zR!;6zoCgPtORt|wuv>R0D&?G{{~!~9e7
zLhD7F)dis|csd7G`TERu8k_SBKvVTcXLGXIam1PCD5bzkv>`oK+|5erl4w){qbmVj
zV~iZ0ue>q!YDG#p$hO-%-8A%DLD7(893_SREL4=s3~t7Wx<uftjibAZ)Xb=SeftOZ
z69D=Lyr9_u{<5>Iwr%hpGz{AYtr7dTDS#-yYg_D_WY)u@*>B*S#&EIikBu|ez1qO(
zQt;hXIt6(kxI=_6r3MwGX1VQGe0T4-f${~Y?1XN&LK3vs5CW%9Z|zqtS*i7fvAH%U
zFc}KI@6M9S>TxM6I}dW+u~FzkX4t}f+JB;TlAwRNfVn%NwCD~FRi!n_haY5b?rGd4
zs=ncVxZ3UoJn1bj>kk=(gu<`4vym9oLfUUaP}{%_&(f(Yu3RJl(I<7=zpLRom7mr=
z;I>NkZ^(tlXkgU#sM`EazL%>Ha~-0G@$wl$k%Z6<xn{7=R>E0CHZ_-G_6&X;zg5W5
zJv}DIsk%W>@+18J-?C`0zX=3;--7*b)dSYJ2W>CzqQzLw?!o*y<Kr3L3z?u`%OO!3
z7m1Y%P0rfO#{q5@77?4tQoC>iL7KwP9G}b_rnGr>-zHbNX^~kj(v1e)tWV}w`j?Qo
zkFS9^4d2)t)=R-f`w~e6-i%${!P8gyvwb`wjZgt0>zyn3?IU`*wLq3J<w;(13^{)B
zE_PqB1Sv~+na{ib)+QG2a~~XBPo8^$Is5V(iH;zaRel=zv&MG);)Z}0|2n`I*L@-m
z$PE{XL7V$frt+PU>WRz_kM@aV2V&X<?@4Uri^$#=En7i}dDfhGqY_&rTF8rwv;r9$
zx<vUy1nWe^Y{S3dA6f3IK-_igt_zoufXg2LK=!%;zh`}YgAl+6%4hjq+-Vs(*g3rJ
zbUAf8Rt~!U&ftv~zBvMd$46KoKzDa{Iqlto8n){ecp!leaWTy~Tk&NCHX3oT0`Ni_
zLNAm$D`F3KzI3e^ol`b#GY{XyIdQ0gEP0Fi#k1m`xyGV6KmlAd(YnRI{qI%cfhT%A
zUaAH}uf=6bIZkCc73-*&j}T-;t8UHh&USqV*Ah`t^f)0v$sI|`j=|@AN=^qkczUpM
z+%?|rtv1C0{5t&S@WbNo=Gp379!@&DQ*m6=RU^!S<R8{PG6DC-Oze?rD+4Dol>8f3
zEVus0%|OEe1BxisA>56I{$Go8@hmL0tB8hC{s;u%@K?#vRAg!plbqIMKTuq6z&~X%
zqt>@EReP7~>5(PNa9tB4nG4E90`^or#}0QO`vJJmK9|kZ=t)i)W)zSYto3nA1S+BP
z7>>A;NvhDyGu06pChkmW(9q2q$<@e{eFknizf`|v&&wI8@~I7k>WdEOtdXl8FQkKX
z+8%N92G#)qeCuTg_mdt$wA6b}(rli&3r&oB%xBB#+}?0)3A@kf?d^A7_`Am3ND#Y|
z#hSLE--5YE1}@e1^{X|`FP59C=S@33=c?D_lz6JWVPuKw;RK8rx_!ah2mSK^`PsXa
z5y2+ce;n=PaS*@(3;H~4&+V8}6vl&L1d0Y6!@8N0zBU1d+i2?RVv?NRomQ?3PfR!|
zL+|~&VJpR!labD62r`pIXZRCKMstf057RrK$C;Rk*$1AMkEL^%VhP0-6)q~QL@bjZ
zBL>iec?jH~(`}%F(bG{vP1OVx1Cj<RA@Oo@O)HIl_%8GTl`^{D-^!u4TSddD7A)eL
z(Q`Za87f><m#KY{n!~{9;u^;ETB6R`pJ$1ZsQmOUx+;?Unk<H?J*Y*n^ZMX@g4YL#
zwGh6{aMdbjgZISVL&r=9=d#&_F+**@+^=!Yu40R`zhkL1R~YQKCWSdiMMdT|{5twX
z4wnCw+e=$y!;oT<P%Tye|9mW)n#w6YF_0Xa9oH?Vn*ktiy3{d?T7Y}av=}wkstPCe
zowBa*2Ou8yB=eAQj#pE_S8&H|0;v8daNbns7u-Kk2vNEQn>8kDfG5VtHM2hvuRyI_
zr=m-`wTFDKk7%nvf7661k>2;ceVcx&^_rU%&|HIdIIk7|@d!GhyvPwf0*L>YLR^}&
z!|f??EG!fn$UJnwnG`P?!%NnV-4dsz?{McPxCpz#U>2fw0T>Zg=(HM=-zw=GV{E8k
z*3Ir;{{bs)9zC}sGyU+O;80mmf6QjmRj6jQ>)&M|Qvq}R_ur6V0e)r(_2I|pc_R98
zn37?P>D*{?J3rtz2cFyHtS4ll;MHBLzLV?0FEt;P$)doGM9eb|jt+AZsI$CI0CAJu
zQ$0dA%T8&$$+PG^ZZL_|?Yj;K`N1IlT=QS2;RYMdMaSWVPprsyeNEJ59?W+QJ3$Sq
zat$giIxS#D_0zE~44WSxgz=M(b-w;DHS)<A9NG-TMM;mZBq@qd!bNPn)&vcFli2%t
z!Y{_RMGcjWmxcI;04vu-aqRB@bX;_NsFcoPS!T1(m9e_I8f_JC<!*0&MH@vqe-r~r
zr{RW^?ko@_0NEPx2cjDg9{(X`Y9FhsAyfkR*{GVH4HR7nmn=Ps;a;B}V!oHyjrPtv
zG$8Zf6!i5n`@h!b@ZoQ^BuMu_u?t=lPA{H4JeL&?t{!3~j9*6ZPai~Vh4*LaqVy#g
zxFYIrrHc*v0q5?(3N}{n#eJKWUQND4$njLhzJW-|WqK)7cyhxBAa+IqhMxydrZD|z
z6uPloIR5G}*jd+^^ZEx9V?q#4A2WA-J<!Y=bc(<qK0)MAvACvilTgl@(%8O+gkM?#
zCLv(z2behZ?yhZR<+yash!?I8ZR%{3=~_7NAHh0sDMF%PEYm*o;8~gHbpAP*aV-jT
zdH}**Du6~kE&vRw<m-7FJhT^W%1A+t+?pDFGFinh1tm=lvZy#bH#Dxhl;z+UY%aSD
ztM5!ZPm!ow`%xj~qh+^EcwFBmCenC+mp8sW^u`Obz5Vca-U{rsd|x$slbeSR7sD8c
z5d@v^&b9Xa3~LO>c&_zUK^1Px{`_=d@85))IW@iOQ2bNVR%iE$XQN&(iuzS-YGhPW
zMyRxYhsr~&U4wq!RJK~i4OZrkZXVLeSN8Gb!5L^qFLwoRivBl|<nP~)e*~|9Wl=ef
z8*u0<p(0cH23YQzH0VHx)STW~Hxm9T(+~eQpUVL2e2>9%R{;J<vRk1cD1MI?N2JkN
zVUEynVFvULHmt#TiQgo!8i}6B((S>q{C!@!!|vf7*di8!O}`{GBdx^f9qGh+5XA*E
z>6}kgxHG2reJvRC=D^18nf9V#0K_sv)~d-nzhK{Ar(lUIzC7VMWX;<nc7a{qt70Pn
zvfLh@sSgZU90lK_<>2Ys#TXoW1c2}oU2f06J87$0o`JZRNC}wI)@v9IjleG`=+`Gw
z=|L7JX&@AK(6&5_xwdu%Gmn)*P$QVB7Py2K12JkoTl0MBqM^rhKVV|!dd2^ImlOa*
zq;XHCVtfqtc{t;}z03Lgv^%c2g&$l2Idq=xEef%>_xFzPZm^$|eLJiRz7vufawHJG
zq>SF8gAstp)wh1;m#L_K_=#N2mme@#bZQ|<n6Esk6LtiILb2a%fHZ(sbMq*AclH0p
zQzr=ku`tiaahXsgece)j!(B)LLqA|0$PMBU{g9)%(WFDpn~N!_zO$+w>kJ6PuXC*c
z`u|Q_Ad)VlApes{D0BIKdm_}L@#L5aMctHk(B~8_xqM??uUTnGiPEb;oZK4MU6crO
zx3WN~TQQF0sQ`45m^2#l!Hv)jID+G|8=;3Y*P?j*_r8~+7`LTe^e}@0uNqQE-@LgH
z6+skUcStGMgc`&;K4o<bG6tgVt__zyd=r9*p9Kkl;jO~D%~630p93Wb!P}eU5b#%2
zse450)ylbMVDY&Nia*Wb8S4|L_%t{+szJmSev}_U8ThkTp3ariy4vW_y1jA=x5MG}
zCx!m)d&HpPI}Ne$1|sCjRqV$omud-+Z)`(26l&`PN`rar4go4rrIiqO&#F_TZ-=(p
z9E#8*T;Xz_d#n&x3`$a&#+4SB^_V;S)sWD~hRY?JllILD04#N`dk%U|{MRa;djAYC
zHs0)*yq#~z#R<)9nF9IIi=_;r=#(YYFoS)zuGVu6g3I~N5{pWfE)&t8IPBH{xIqE^
zjYx<l6jwt1K(JJ<v-QIMJ!fB7C>QWPSh?^@9k9&IAtuOV=z0>y$($J)F*PneIo5~G
z^uD%;#H@e}{mVAXfl<pv^C5K%Ave%>{qkH6`TTNxUHv$gz*!Yww&MD;kakKqDsXH3
zq&pUfXL~o|5OXheopvOj=J`o&ZEnD!h3L5{q+q$>b?I21pp62lgH{aRIffZ!F%>D-
zpmY>kh|MX)n1|qyFNL|~C~uL7Zq1>^R0<?xa6idEjmW<PIxYZ;(o~T2=VJqnFyNHA
zXu50gxoA6meMnA`x6tpF5%PcG*=T5)Ssx#wmget+mXY63nE@X;vD$TL8mVa%t?%co
zwaPnj1R^%>0H;sKO()zV1kL&i?jY2<;3|wiGh@>)P`0>ZiQXg0zxW79on41C@^@!R
z9qG!x>sFcD#h6m|jTTGc7GWvQi6jbO$x=9WHuI5QyxFJEha%j}J=OpZ{YCt^_zI{Q
z9ozRcOVa#NDdeS1pPWr4K1h<^=`TB03Y)IgNB=PN<kdhv$0f5bZ5{nU2#M?wqG4rC
zp}C2LsN>-y3bJ;yQJp$TR!5`VzJFp=BAqjC^8_4FiVo+dJa`CtipY!FJ0Ii=#kX;C
z!KpQUpU?MeE3*hcNJl!K6d1vw-CeENapttnrhh}HjpyFFe>71FH|%!nJ47+B8J{x>
zRB2Y$etSdEUiW*e`wirzh3|4Vd|+jtqb<dhTBzPo%F&jNs?aR=Aemh>F13vvUk<B`
z+wHaQDArUs>5`Y*e{>da^8?tD$f9gki<$g(pZlqIJT5l5Er-~`Rt*|8zKU(-{FKsq
zqWT1>qd>9>-P+(rAcji>weqo-_}#mgV2QU(my;QRhL-Lv7*JhR7E|BHcapftm(Kx&
z3;y*!NcnH)y}`fB|6jr3IVO5MSOv*j%&&TR&F*4|nP3DS-ai5we1t5J<7L|obAX`3
zSrhP1Z+^^1MPO>gM)T}^%zmHX0X^-6Myr<a=EN8UQz?+jFMXH*N^@22(rM-C#+Oy|
zQIgq?4%Z*ds5Ln1oQm#Y^1WZIUcN(p^p95L+DzV&eRYO>#@#j;DyvxhzC#<&+;p-#
z4<undwK_p@<#Dw{58O$do^sb^2|2z1Jps0nFYd2Y1l2BaWYfovz@@1}@J0qH(_sxP
z;%cVn7f~4o?8v@5+E;sie^RK@5xvzFP?l@${|ZjU-~EBuO4Jor^xN1zu;B7kNS4?o
z!0=<y;{s&Kkiy*2f0FwRsB<OU(5PboY*gNh1#Qeuw=`;7-`JRP-FSN~O%y7-TawE%
zSGES>g>RNmPfI6<rfl(BgQO&2K!Ylr4Y^D?sLp#|E;Q)Nb$z}B9eOl<lJ2;Rs)#7j
zSy>kI5Ohp8GIz9j?wBD{=~<(3e|u2}GSIrc@$c~e$Aa)k3Od~$p;#33AXn@Zcb#IF
zVW^8AI2ei8FIOE0A0HY4LBjZYjav%}K>V(<(`R?LmR0^1l2FW$T{^IqXl_UejL16K
znHtgjm0k7^oUr|H;Qf)Xi$&TP2z`_?T}r`nW~wn<nHX`*cuD9B>(DJJ^VtkrtFFy$
z*udaw?U#BAlOHT`c+DU-N0jktlAjPM6;hEToT)kR#$@^3d$&{h<d5RZ6xNO><njZO
zae~mFf>e}v1p-sLH7K6xp3Scd!Um=^atse6rIp!&&bdJF5~u4ym|{5`#4+HC-9x0X
z46~K1j|%6wbH$why0pfjO0X^4u2Y_l-TtZ##PS_jNQE`%7k;$_P7eek-bFb*%npcK
zw4ZE#3yasPI61yBQDALxKYMh>?OmM<tkumkx5n@%L9b=>IM<{LDZ<f}$fU@nP8I=j
zh2nw0<039|7s>h+hfr9@BNZ5@QD8Kh9Y@H1vwCTh0bGUI)xF7{e;)%Ubb+hcYBNB~
zShzp95jX`koLoAZK763Z-H3nZ%GD$9SlPv79M<)G42@XXez1b6&Ci;%ykkzvvqZ&6
zhveFT>wrQmouJ_R^(%P-&%U!r<Rwc9MPU_lf%yoZKe+;0o#EFHlXA2ji_BFRL%<gb
zjHfaHB++PZPmYBYkt5aH1`IF>lIqfIHO&}JNq5jaXt5cmO3;9zzp`6#FK-)=$jAZK
z%P5f>mh`6(mS9>tSU0-EWcJ{c<GzxWoKSfz;G6Mqu+wOr(f5IB$w=xvqD8m{RUnCJ
z8dgknx+xU3rA*^iF$d9Qxen<QmzD~BFX_`?GXe+)^zL+P&6l*YG&9p;gYgFI20Rp}
z3l*6xmJD(2lc~P<KpvWL`_&^b0d&z3txyREW5E**E~_G8;_UDGh)I8^1Btr@`_wAF
zI)y$B>zBlwEEm<{cFgv|^l4%bY&uJImhJVsZ*^JmCn;>V7s(f7bD$l7Wlj+OFK_@5
z;)0#|?xK8v-FJG~LX1${qNKLYD`!;-a$Kc7#R6+Xk-S2}84hT=bpaUv&Gz3v#Q;{$
zdr_+Pb{E(QdRj?u7atVXND4Sy(NV?ygDyD4FhqDzm5J(>_pdSf2R{4dp{YEUbK<~`
z*)G_o=g@Ki{sU&VIItJd(jVUAEl6MKQ2LX_Td0<#;tPpWc7`R`90ai;8&wudkW@yv
zjiu%Ce`y3vYZC}!cVIEq;zzoBzr?6P%pN$&8+u)r%_;!`6JYLneBeQi9A5U@rCZ4G
z&NZBsQK%xN_KztvJ^IAnjTf_HyO7YgaZD$DSmP2=5h5UMw(8KnV`V*rF8oz!Q;~ZD
zkn)0y-hG!%jCuR!xZ2`=H%L4Qgp~nWY&P5R@@&K_5BT2ICH2!r*xoyf0=cyrnX?bh
zE{1u($P2sV>*eAzYHQ)=o)no3>u_!*TaPVIz>)N65=T=;Lj3xd2d0Gp1~K8XW7g;F
zfe+}AFYm^04*xMeZV>_G5#Y|+1VRtNJzvFWjKO_(zCzxEk@r;Lq3K2?pdvE{?lTmH
zHuAS(G5Qu-h^K#4mc9PS$rS;Lp$pG*K9SJoR}^tx0}@6vE)7m8u|AT2|Jf+Iz}J!M
z5hdyp-;Yj9k9(l|+rjl)tqMiqW;)!Ddps(xMYUhE)m}H}o*U#@J-05eVGH;bt=-&i
z$U7+C+^h?+uJ*k18Q@IlC9KaZjRDvA^|I7GU_|j@n)x7U1|Gx^U@YRyF@1bfX1DA%
zpA-Q69(JYU0rn?Vw1NTA*%B&MM3aSP5+-uQULiucdta`AKb<+4hc4MImY!u#t{w(r
zaX|J=%C+%wDfH?-y|NQI;oj)YtDij@t6Z<OVfHxxZ5rSztbL7*x3;=Nj=qr+JO`Zl
z!KM)O&)+LACrLphIM~~FnP*hRc5kx}5AYag7{Z=C)U%3G@fimX%TmD0%IMm7a_|ez
z&dxSdA6AJ%B-&SjLyJII1QfM+*x8komfB&VuFMOD@4B|OoM_&$*OCkSUHV;IteYjK
z%Q!YrU>6`Zur^Ga+U(Q4;z0=W6V9XLPUG$#zQdXw%hCBwq*}e+b9W-~b(0s!#r~>0
z#M+x3jov?La9D+&8t>E-X+GSvLQsZ+1A_<zBLRc+?Smr#W}fp}M!kF+X_h0>&aCua
zmh-h%86Ho90i7vS@O6rwtA&HR@S#9%|87D`Ly9s@u_jmMu+Z(0*<_hmkY?(;OdpT^
z4UJkz`f-V5VJeNQ7njRsN#SYXx$Lb5xer?JXss{)rq~{r5$c!VXNQjS1B5+;5X^6f
z3QCkLZNAqVK9})GAF|#HDnB?xAX>6DPVnxQEp=_hS3bcDUXbwCWtl@gF;!#a(Y7FG
z#9;IeI}54!E;4JTjNEfO#CvmpxleKw!&0VsES0>+%uXf9!jT673&ir`29_)pqp2is
z?w9)VePhy`S1;n&?}o>qy7wRVs6=16Jh%4>$DfP%RRz6-pVTrvZfFx;+p|9J3<Ljp
ziIwL4U1^KY!$k}6rj)~sO8m9#MM4iu^#K3Oms5u_LN4OPmc&*!bZ}hr7#KA?Y3TXA
zP7=^yBg5b*rkpRC!GRS`-KrV;lhYj5(l$QG@fdN;n+2EVLEsT}8x0Ix1Pm1n3M@$c
zV-R4QuhpnXc#meb(FekSX_Bs2Q=$)74YS~j@(L|pql@zh4)lZez(53_CWKyN^5S@I
z3h_Z`#Ku&Y2EiF-xE|;qJlh16&g0gCQ@7-86+@%dX%c)6_Qn+MXOt(bV11OcCUEpu
z!cJDj<zS$Y6KN%O1>>`g7`0^_w||`X#xp#h@gDl`+8Lz?_ZH_0J$ChK`94H#^QevH
zz02p8*{cTF3;!Tq|Dd2+EBZMGYUAm+DU8R}@q{wM93A|XVTp@zv5KGt*2_t0Y3K&C
zwBx3rvJVC(-oRzf2=tC`V)RX_e506qb=-rxCXMf?YW~7hlS;jRoMBL&g1+l|MKL|Q
z2(e1}ZBJI3QD_IySx3Db6A%*{)mUHiT4l*HV^^+ljyGN}WL97M({g{q1$plMez}&C
zNp7Q}XE37qM8~O7BIqttvOpLQ^`1gH)9><01$V=`LFP-g$K6WjJGm6da07ZokXcyM
zi{^h%9b9A!Rm^yFp*%F#-fX*%W@hPWe7>{4dEZ1MC}#EI<24r}_0m+j0tSx4+tt<3
z(Y)6C!R}xtOPzm+HU<-eDZ`Sytkt@{s=^<rQAI=9+Cn`KklEEw*f~9VTV$OAJ!D1~
zOC4WAW;P>ZuVyma2C8;WXU>vhWJ-IX8rEjZbr=Pe;ZZcUR93Z;tXSJCEUNGWPsKzv
z@UHZg2I(lOx-wb&1DPH9UN=T+>e<$~Q|WQMmSdU1@9acuI`oO0{tvBoak{<|_QqE`
zAcKtQ4{0RuXHun#C&kgleYy?Fz_PcWeFrW{JfQmLqvvr7Vb6g9Xz)*M{jQVmAyi|q
z$OJsE-_=tXbYEtonWf%TfVbsNo!WnA7YK#o#ZX?Vx8wE(-%gyN)h2U4hQkxWb$pEN
zTgq_wjJ!y+S(mD5w>yyooA7gGiATLtQQ7!Rx7MIy?xOQu)X0Ldu7{u!K@+L%W->sw
zhi9@7d_T*Im_`5@C^i36ahxK^s{Eey3gjgE_?ObtOyEpXLQVVYasfx{sBBC^Gv`@j
znC47c$|1JbWZd>oxpiOB`55EC@4O;cGadq>JB>Cz3S-$`Ds%Sc%x-L{I2BqdoP?6Q
zi8gFY;X$?&RyU;ySLtqoojjUl$H_L{qgrL3ZU3-$|F&+S)Ei<+GZoTbce0$eC5lef
zY2fDH7MznAtNNt)Ahx#<JSh`T(%pGHhCF+3{_%OprPpI9mDuNnHX|xv=bG5(DkmH_
z7qiXYeQ+cRkA34<s@8}KI?anl%kI1y4&jhm*Y)6QiOWtbSJpdhxE8a|?KDtD2IKkP
zfbEWJwN}bY@dwe&ZIAtvq@n$Gz^|~1_x7;@m?Rc{zN)*^OkXQd9$HHJoLD%1gXWS^
zU{ta2EO!Tb!2R-m9q-@f)h$MuX&E=(&XBwI_$&S_6g$rq!J6eT5$U^+kYCyr!~Ya>
z;Doz!;<|uMH0tG?O+j)L-44>!x9B}09|ng_bdPwNSyp7JW??Wk3qjKj#A!_W9><LV
z56^a}m?=Y3g(<UKjA==njDHT36i9%IlIVB-fD@0URD>dH&#YxCL)%Qr<;&@QpQk{f
zp7q6rC$2T>br55OyYp>44~y9OoVcQwU7Cw3;C?n+u+>r~;|Ho+scp;o9-!V2W``!P
zNatp3th^+;wrpibYImEx7>ztv^7|dhUHyYMHO!Vw|K4H*xNDRCPr}jIOfg6w-mYMU
z>+DnHb5&ya+9KV`H{fg?XZ`!2YWTTS5m>z2^w)e_RVrKr5v@0(_DAFS@setYTlb1D
z{h#?!gdbJdSkqKLDURF63|=<9f>I28hj*aiVufGc9wgipuMp*bpS#pgL_{eHODc|q
z+ud1yP5yma>|>XoI+()n?NY%LEpZt_xThrt$#dO1MLR3@t?$+Aaqr;pGx{?$%f#s4
zmvny+0R{S>my}=z?rf<#Z^IsK=k<zw6qw}0%$myLjidIQkzm)aw-F_2jbUCOuaae{
zWcfa&lAA@FLWVqScy=mM3Lqx4G;)RAPBo~wtyogAr+lk~xhj!BzpLrtz1QhYRz);x
z=uM%B$)1SN4S-%OWTQzD;E9vg7_VASJ>q7Mjca9V{4u)u{)RG0>^CNkA{^Bzx!Jk%
zTpiTS?6(~~#xXo<+DwO!e_8(FVHT?}s+v2luuZ?Z65S7ijhEWxTS|#PN(H471NC5=
zH)&%PX>nxJ`&+p3>)bH#6V!$MZf^D`S=NT*Ri&xFdJa-YYivAz$Iy7fZ*Q2kHQ8qU
zY?pel3M>O%Pd9(tk*|yh0La=EJNe4a?KAd?e16dEz1GC@3=nD6r*yUcQ2faCnV;ml
ziPaIS_Y6=v%DIOq9XM1M9p4zG_A<)Q+i<&GTiS`GOV`U~-1N!Kh0bV)z#~{$n`I9&
zLqANox*@Av@IG~SSA+lahVsUDO?YlP)VC@NHSau{#TjrkZEc6Au=bJMsHuDexI{Ei
zJ|f}BQxBKl=l+q+xBCDCW@c+a!iELr0;!z2s&^>GHs~?BF+G7C8e1--vafX}-D6wj
z5^RJYRoNM(Me-vt80|*(xb!9zLY!+FHYx7w#e1*6eB)&7by=t|cqm}b0&;fob`&?d
z;GSb!A6+g?DyHnd+g!7=Np}RKZ>^T%CvQo{ygy0hgJCPv$V$35(K)xnzwa&8vDT6L
z_j^`qb2J%W2ZwgP&CeTl@bqVneEO5SX<4ccYJfWyTb`KNSuJo$TF?-;R}V(mX}7(j
z(qs$xlN2!WF41u_LJxFjetS9Lmz}`6hs*YdYH6b3;*)^4SCA}<&}Np}dj)4&w-yW(
ztSQv08>g-V>ux#cBhoz7T;n{v#q`=D@zFfYyi7($GR}H6(XsoJ8JR0i9>0+DkKgC-
zZx5X!&6Mxojq&~|-~eQ#fjjfEI`m|Ey}x|B#Cv+b9bJML5KHD~DwRu;DRd%ZO(nO<
z*HGjx3T|vM(eW>PzNXHK)33;9W20M$G+!CTwD}hE{8jqtP^l~G=%~+6E`@ac?8*&t
z?unjo#XP$7VnEPPAm?Y{zD8yn$pzkliR~JwY~S0PvV1+@z4u#*j~C+;+iiz0D+6_@
z{BTh$oh?p-f4=UVb1Yi;uCTGpn0^&#D*pW7Xwr@&DWTG$>GhspSzvn?_?TTM_<XEi
z!_*2;CSqi@s?~bcthARhOiU9}GB3OeG(qlzJ!<2<0v|``7dOA!X1)f!<hDO-_?%y8
z_oCP>ikr7IOG{qA49w}l8!?nmWJ;kVe^oG0=H%}UIv!rC;M8@q;2|>4m?|Ba*QMXv
zhl>Qw{$A}tDPAW87P4)^2LI0RfcWmC6yWRR^$!v;79w%l>IyPw_3#{|l0hja@ivxv
z8mX7^YpbO~crDOy)!g4XjtLT+8q{v}SVu!NGkWUrF4e5e8?@oioaQ8SVEjXTLYYiY
z_3gNWV*vN?15fg<Y>7$UU<_FX4PM5f8<lwM_%pVyI=jx`c=IbqULi48YMQEPfbE|W
zpa%Vg+L)7NxrbktWpFp$f<2vW8y|JlN{DcOFLKOo$vzd#zW9Yui~1cPMYkOR+jcn<
z&r6r^_xbbLc&xzPHM#RQsJ<qa8M5DwR|c?iv3Skx{AFgj(yWJ9cN5d`db)mf&G+>c
zvOH01-<_0J64qt+0|$!I*hBZp)(j(&=tSQdVParSjG3*)tlY7&auI=|j1j|}={uc&
zRQCtf__<{1^+kxM-{$HEXb1&zdGo*Kd~|(xdB;6GFE`p@&ue+Ub3=^y(FE{0as6|q
zTUTPn4?xGXn>hSk?oC;E5iSw;P>CvJ!@kwk!YxGaDEe74`#N>DN5-9szC3GvbhtLm
zfM&Si6FEvmtZ@lD-h=wjV^qG{k;U<bnc0=a?$$9GDej{*kGH~hlzy+S6rTN!x^<Qh
z&80i=`wOdYzMFn1Qv@!5qCR0~1a#wO^C}kl3g+SCzB4NPc#m=XBKiD`Dyd9kSRs9F
zI9q^zd90rBqdBf6Pr2T_h<a%0k7?w9oF#nf@59cQyQaLaV@r$7FS>k*NNh~qGNsXJ
zphXoMS1p!UpJ*omS<}c&g0Y6l%MSkjMpgWC<lJ%KX*;IPe!5L_7hj{^LH6J?Gs6&E
zURkplGj8P{h~;IUJ(7<p^0fg&;{U^79bRYPa9dnPPPAc<n7@Tm8L&iB{o?f8w(b3I
zJrySoT5o-bQMO00t5J;h$IwKz$!F7Ds8T8MV4B>YG$E=9E6<k{u-ZSVzy7k*Wba&K
z!pFA8+{A2)bY3fJG}D$nVMU7IV2-%<JEnLc;o3`#HFKF`^M3a*Ed8<)D(x&?X>#v`
z+;n+VX}IpE^R8ke1D?JYfU)pbP^~V6T0I1nrm?^L+Lv5e(+~1iHd(n~keGJn1EAk)
z1*LKd{r5fo(!@_oRRk1KS!(X1i<VXwQOpj2J7=y)tL){aIa}PIK2BJ)FcgQ%EXG%h
z!Xb~AVEo}LldhG5{$b4G+u<EI^2#Otcq0(@?ZT-=^ZH^XY{bz7)>vLhOg%;g)r7$`
zPoy8c3g}G-d8+rfeA(D|{}4zoRA~LlP`KYeQqk<(gFTF{i{IX_m$F;DSpWMK-f{=i
zJmkFNcEkUBj=g9>f<E^O3XYaEBMkYYDtr}%R&}Xo7hi~Pl8(FVmy~dRAEG8re?Q#k
zj-jY{*^USH;P`$MIF~~N_$G}wAy==@Y!0K;xp`pjprFRYU5cpLiC%k(x`X|8Us5(!
z(xV{yy5tF*xEPX`Vwa0lt4-HQKhZ{Wiah+}^Kw1mo7<=)^RKG}(H;18vo^62=i3N>
zwg^%aOl^{FJvv&1o&uek(w_$;oWL3%dfi`Zpu2sau-ZM$&2CaUANPSu1Jujxm&d`l
zl%i&xse_w3&7$bps*vBqaO=aVAF$+-<P<;BF(;RT2>PDRfvG^sQ8>H%wA}*Ir_*sw
ze!h}T!b^BZZryV{X2Eor9F>wsU!SPuWILqP*eciz`VHIoq|#~0UGv#t#q^=XZ`~}A
zR{2_|<3Mq^bb67de@P8;2{(iE{HId}dbT_%<H;<+#Plrjgn?-d6{++>SQ8H_J7;{V
z9Bu0k5;zIH+4GKG0oTFz!^6%9q!0d2`O0?%yP*cDnlezB?c_8x+d7+3Y5U2PuF5?1
zDZASWyE=zo=xxU9ih{3fI9C@}QE&6N+3aN|Gd@f92uM~j60iq#y!@GbKRhCje_6jQ
z>ic8KIAh5R1YSw_T#ODBpxSVbww---?$J0n4a3xtQd)DNgsCr+T?|hgj_H0iUGsbw
z&lWn)2Reb&qRbAD^Uq9oJDL!QsReTFvzfo>sdkr&HBZMq!g2GNn-w~=D>|AjMn#{?
z#e98qMKg28_lkw!1Y^?HdSEaH21E9B#mi0&%jT9o`e}F)Mo}++7qTi}rmAr;hgj^g
zUOZ95J6jSjSXlaV(F%N#Kj`s+=8Hc2V+O`wrr|GG;_*fM8_X?i=p`#MgVP|wq^j%1
zquw8EYgmE$!GhngW7COjLxcO===ElC(Z~~0Q<R)KlEa>xdMGrG`r}$Z9S?QhacUXb
zu`mxQYZ_frX{`t*H8PahWiqGaZivq`xzb!BnV@U_<^<a7MGCvo?`ONeG{6kLqr@+y
z_MN@kgg&l|UqU6t`Y`l{>2PlOuD;Ih*L1$z345KEVy2hP5gY4kN+5G}1~lw@B<xB9
zZVumL$&>*Yn7_^HU;#I<BQ34`&vHK-_(tSUW9|0o5=DBiSs~ok`E&Gv=)4ZSMn`^Y
z;W%LLw!GR}^GmjngQq=Ns#><UC8i+Gf)xHFn$w*c<*nbBM+!4=#=0m8uZ;WS(I3(Q
zWxy`*!HLYj$0q*w*u;sk2r&OWwzB^@w!(4$K0UD%`Ye?X@?4Arn7oHk6qvqPV+e`3
zQj3b?=FYZ`=6&fCTa;}PcExE*($BSFik?9`Oxwu0Z`w?@N*BUD0g?MHXoiIsuzMku
zEJL_IelX%5;^8U|60Q^Zy!--WZ?T)!d?%w<V<lD?QE8(oWc^FcRx)&S>)NALgnzIJ
z{K*2b-O`opa9*|;*@B_w8RLHPr5gWmT$24pJH8iLkXXOO!a!Y28or=J&CK%qcr+c<
zaSjk+k-%j*55`rjFVUB`9b?3g<tjJdun+b4YWYYf$0nV+O4Hn;@ylVg<Mrn&1A0en
zP3s_aFszxBOXv)?&_gqT6E2!|m0aX4O+4*0Xj_=H@VgC-`mAy^j@<@ElkfzC9nyFo
z)-b~cLqxswMK*SG8sXaXf3??|4T?T994tryz0$%VbP@FbeuiLB$evIXk&PDQ%hYKU
z5-Z0*MMG>}jtUAzyDM{CcbgqQ%yOSt4>(*TdgvFsR>vtcTdA<|sinys;7HFg*ICbv
zEC8b~E28{?3UHBcyjpM*shPLyC7<l#gb$X;u?CU_-!!JI_TX&dLivNJrPQ^rnOt&+
z#}3^T<tbF9ud7g5ozqj3yUEF8wws!(K7LT-hWd5aH(9)?7N>72K8|Oik&zmyM*T7A
zV(OqW77Mw{Kg+P=1Tp(nU`B)g;t`hkGf$cN_rUywh>E60*B{~6!-}BxyL-`R(Y0z<
zD<At@T~IPzy@!{hfa?k@vDZ=Gt}>3Uuz*1rVJ1|ZV4bePjJJGEwRhfp3Fl|1d9gqD
z9C`;zTgW@=q3ZxD0(JhL_<L+?e~)cV8U_~TpJM|HdcRn1^(+}k$KI=3U%rh1ZY8sk
zjMWG3MJr8jjhP!tYy)(BoELyGK;ILnHOrA9UmxW+6v8R^;Mg4lmj;4$Urenx6#jD`
zw*PfN16T{eRxx@e$%G+$*ESij9WM`xF>f}9XRWdP`FnX_FLE&Z3P5%0Q;LMli;U7F
znw9AdKMQr+k_ng#Y~3axh_JC+66^b_?BgS@^87-RKAK}X@F0`Dw_uV^C^L884-M=w
z`(M+S{koqee(#H&(J(ur>Bv}nA;_ee$iyIEAl8p-G&@wvP&Txy8>Hrfo+QVqCa6>?
ziBVRlu~?dI<F<DwSg>zssh6TxbB*&73AUYmT?$UHWsBu%S%>G#RnqdCosA&0Up6S2
zLq2GYZ+^TaR=H??{MDWM&l?O>J<_e{((00fTBhE@UcWR)Sv3g>D$38y>)wa!rP&5{
z>J+%2{yhC!7p21B6(c**_RE}h$7VpG%Vkf*=qMrJ7B$8K!M-dQNiy2uSPHLC6fe>=
z;c`xlEq;*Cup1YM4&}?PPDLIi3OI>TSLO$HeFG6$U1LSe76=p0X0?6RQHBxx1JLpp
zsl2xZ=LfRxVCa&{g~zRI_ttc~2R9S;n;bbQip^wbmMO~#S+&0Xa%I0f?vJf1HUoQ_
zZ{D{LW{RB~zw~iC>s2hESR((9Bjl&&%)c=xRkrYaSjA{nv*TFJ?BEO94tJ<qdG$CS
zkiGQFtXjU<!ba#W0R{6qxbf}gf9%w*aNK;udI_PB0;6ffA#OtahqVBM2<L-n%EYN4
zJ{Vz`y$!T`KI6u2v-nYA<813ZJeHr!hP$$-Sse5^5|eY?%VYfOkGoKBII7$t4aGvd
zwHotuP-!Z|>;FU5SB6y;ZEG8(AV`Chk^<7*f^?@eNOyNj2qGo9X{1ZKTN>PifHWJB
z-qPLhP0qdd==YN!`*~Pv%{k^6Z%w}Md*;K^xS0Ge1<S1REgP%jn5vc5eL_w{`m@rm
zlip1Gc$ggjmP~wkv;Red%r~sY)K#|NbJ>-)EE!|+7)d!xvpuuh@>+NuBvbqo@zD+)
z?M=-yO4(&pQ`pb_p2enDV6ZsGl3Z}x%CzvpH={I@lC~v)PE_QY_I8O(b}v*KBkvYZ
zSe%Nbu3b1s(mUGM%vw)k+?<@YvtstN+Pt8NC~Tv`p;2fu#)*9&OI<5bUB2mCJyA57
z+T-N9RTrco>tA@Mit8uzb^iDCchdixBOW|U;z!w@sorSZFhF7B=(OsLP!_KZHnf5c
zm_Fri5XiHla^&;SU<r)%c)9_w4uRiE#&@85LN%+Zmr$L5ZYdx?F3-?X<eE#t*9<8a
zE=UDmol<L^^!S5QLcvj*JpG6OPaGN1?6dZRd>sT>XF_3y$zjWsqq8L^e-$gQ0Ndx`
z58dC*0fRV%7gv<gWN{gmZL5~5^agbSZgXrxkM<%%>C0EIvQHXx@Y~BTzc~=RB;~oO
z+b5&~$~6#fL-M%Ed{<*Z_W=q+8mFA-MxaFAyN!_AxpY>?SkGDzWG|Sx2*8_C1@7tL
zriE=WnQv1mwtS2N8n4p9d;|b0S;#Fb3YuWxa|=}`i^H|uTyxC=ew^@pLFb*_^W5Cm
zyjo!MV&=cQp1-L7<gD^_o<!+A%Ksgl4-nR0Jpv5}8FA*)y7+iY;Is@qWpc?Nu0lx)
zYdD|Pygop)Co%1>57dQu*ZZlhC(9tOl@+Melgow^hv2{P&?9gVx+VB>Kt8W%YEN8p
zb6uRV5K1PAlXlf8+SJphnck<3pi9-UYm6kroXy`sbnP@7MYrKdXgo<2PWSs^PUf*W
zs@1M5m9NHZz(VEH8H6d?{!6LTc(3E5=H2V_jGmJx=!aR>p3JEZIBdJvY=kt0x7|33
zna;ghg`79ELKnYO8BhdsBDHhW?dvwNW+kjvNpmI0uFJLUcV}y_zTR-O9u2Gc=zQ0M
zPG~nng>AFcUZTz&SEH9M?)QxNcGQr1eW{#-9jG+^K%Y!~RWXP@25Bqwh!)VvnZRoA
z{0=o2U0$;)ATI5W%3G?lneqB56yDf|1*+IBYUjn@D)vEqsa3CeHodYxW!gm>=hb^G
zHxcwfLRBlS5*<2q%9TTjcYDf5+WZ_Oim#*?1qApa(+kuhhxPIIK8;8(=T)!IrPvwi
zF;OQRIE_As9Ty2O*EqV!Odxu$Q=+a_$uhO>U!#R^bz6q&602vZHvTN6gH$~Gutk-j
z#B?|8Zu~e%__8Cx5T|<u9-+`C#WZ27(DGzaol}rd-L17nu1x|6T#lHpx%u-AKYK+K
zCEGX{Tz*g7$!F&i6t$`RdbuyPEU9s8F}f7kgNA>*XMe}6@oKb2sfuVC9W`xz0ZdV6
zyU`!)p`X4uZq)D#8d^O^uyOD6a)I_ENQ^{85XdB0C1jF*xXeo^*-^k;>_zi^T8FU{
zTp6T<ESW_8QDsejX>MDXKim52JcY{=*<Y+`P!D0j_Y9blaq!>1<J8N?#K|J+uFzwK
zFb}X$R3Q+D!0F*i*>*E_lNJ8XC-#{>R*2O<qYH@#->>G8yE9^E)%ITu27k4yE1P2@
zO|9JLUt#pHoGr_*zA3z>^0)7jqNs}F-C@Zg&fh6c!E%vO#F^IWYTGJn$F%(fb-6RT
z-VBM$P{`=3SCX{Vp?dg6;GA^$B6$@a!Q_^`<wi)|d#gO=k#V<7X*~@`_cw3?_kKr7
zlSb3x;-EH$`b7>Lw%_cO;Y16+<YGBrW6t0%&Sf>n2@Vg+Et?=K`IR5obYkWWW+LRB
z`c1%@cUZ3KGLZ^cWvN+4pVai&AwRcv)~KD;qj^I5>#;e(p>`kv11E*r@bFnPq`s)i
z(@T`F8DIE3Ri}=mX2i(P1|i<t&nIdh_y5R*2QR)$zVC}lul0&!d1zzNME90AN&O&L
zG&V0mt#^ok;xW8CNREH&YS)_u6lfD0EZ7I_sU0=u97>@cq%viV>JS_}QOwKAF4t<k
z^6xBiUcA`aTcvw>du&|qu4(9g(y3#H&#<8dK|<gj7uo1|=w-F@83^a!ETQCY_bZzk
zMU|M-7t(zon<BlZqDJdniV^8r;moe%lU0J{TuB>iW!cDZWVkG@FF+zT{4cISp+oH>
z@dg;^$4dT`hiq<*vli}s&v%A11idf(7sxv|`S)7bxx!{Yh*vMdQ3QnCaz@Lz@1bjU
znuOFyjfv+&q2ykx0QGdgFw&B9v2?Hysde($Ox4GBVY(ITEbd$M!8vzZG~oX_@B0=B
z?eA>*e)#2wu6L?b-4*5#{M{y-c02WP_@ohsa+d8tn`zg<sP&x#tcO5v+uHGEo}Fi;
zqZhxhQ1ZkO(EvB}bd~bESa|q>G?OIW#E@fWK~7NwTuR(Mti(f9+B^a@kgwv>WsC~!
zrDsn$%%kGf(fC#wTCbQf+T1#(x|iKzC~PgMaK9^eLgWy>7fiSWNn>qfHfpFw`l$%t
z_{rduE!Iv)|Ec#3_tE!KHM%?HzVkM*Q3m8;F+;i2rTsh%9<>Ln-}`#A1aiXo*JMi7
zisa>lB;8B$8%w@3X>8%7X5h%7Ed0QeDM<IR0>8^owZmAzi+C@iTQrestGvLoll3;Q
z3c#0e9o^>9KWbhjeTxs;1WWcPly<c`ki!3YO&AmvH>LG|B2QmCb*`X_f60nG{M1#H
z|NdnbwFqUNRgcO?u@6Bw2$aQUPPwpaj+K6MNP&9tH*s^JE7zXHB{jo(zw;n+Fc9AF
z)<A6wL)p!P_jye_9t{ZT8NIbZ)&n*CdzV+sh-lH8gx-MpGx<WuBc#aG6SA_F+zY?)
zqpqOgwR^zCQ+-ZYH{#yo+_`(Hw|uxz!{jy1v3m~a@1wzDRfS76@R(`r!`i|m8&&+f
zOVi!?t=^dWK3Apni%jZP!QFMUda*OZg8B~6+`I54qk~G~UDkh)Uq3s0j0WCJ4ZraY
zQ-PmO4MN@)x5+t|n~E(5F(Ymhb3DC$%fULy^@K206ngyCmvh3UXMZ=F!_Q~aRj$_U
zX_6r}Nr}X`!kL^RowUDPoGrTg0VI<L=8kx%d0<eJA<plbB|Yl&r`~SeTe;4iVdi~9
zB`E<WzeRKF`UhE7rrIx0q-p&QM@<!}Vl9~szIU_Pu(6XNF-LW=$HAduY<S7`{);!5
zH@jz32x~um;2wV4DGPyh%9swq^Mb{NF;`<~mO9+&K3Q_n{c|1$)Cs88;!F;l#e-67
zI*C&n;vuoaMl<cMZ@UP7_rfHZ8o|0jNc%EBTXS*Yi_F?~GpX~*o-zbZ8>)yk>=idR
z7(wFH`$|LFok?w$5%#r2iqR%IGNG{dJ^QdX`*gv`WHGn)gjeg#?q_p0cUl^5@l&oG
z$k*<ZVZ!~Lvm<6)rQNe2vTy&L9v=iWe*W%%d(@Oms7WZfR)sq^WO`Ec-rgjHYwB^F
z>-ACQjUxDd>;-A`Ij_tRFP$9ge?CneP*=!T<1-o3!{f5x)n_t@XSOc5S)!R9DbrI#
z;)KOp)&lZ^awdmt3{_XTdT=EU)+iy7s>~*vlktb0=xYALFbLeR_2ns6-IHkJ-Ot)z
zPZqgI=#Es?k?a(TKN2o*zxoU)j#&GrHQ$m)y-e=P+Wp|lk|ac69@|}m=p@P>@%8%M
zek8yUTD0Tf@2GIA)a|LXUJpJtO)k)%v^a5D&ELRj)-?2v$g0ynK}vpR_H(CAHbGmG
zZ!&Uguw>%))8*E)ML++$pAXmk{LeeA{^?Sy|0Ckbyz^2|HR~4{W^ew^ygaDS3SmYh
z{SomVUuNWb+}6MGXt)`sE20~?>t$2KyF+Dhg%M-pN}n@;7Q*V6W|_q)k6Ef?mbfI#
zT31HvrHReI=9Q0JX~liR_abMcDyu1#L7;B?Al1Gp;gs3l+*NksxF)5x)B0Iwi=`K)
zJEx80PF|HE8raoGXQng=TV2Vh=PF-Xmewe09?Q?I6we4YZHJ93i045IGM)ZZTTeOC
zt%WnUw%Fg@2AwunfwCZzpj{#uqW+^%@=sKBXo8!@7JJdu;H*>W%V*e<MB9uCADXo7
z7moW<%WqHY4|~{Fo*86gZ#|cVQdZ9Z{t#R+kx}^8=jz5EFTknYAHC!8KYf59!s*`P
zU1vBm;!!r-N1N#-DzRL?+PIrzii<(dbWF@#ZZ<{o@#wF8slXY**F7o8y}D8KR&7el
zxPGD5j-zCOhCXvch%vd4Zk!)C;C=*bMX@KPFXcV-#kdLiO|Oxe*|pFlhlA2b0^_<Q
zZ2I`OYR2!orker9uJzzZ9fhc}?6a2aYl5yCy?k7O45e7QiJ`TwI?5G%`<-8UYIDMQ
zMH1`dA*SBP$#9+NN%U|k7Z&&Qw8A&<E#6)qLYaU-ib>0*n~{65v*HIt-_P`}V|8v0
zYcb!oC+X-5RH9jRX9Alcp8iClWHRZGiG<mhgyxqP6*THUi7V;X197002Rp{VQe)cA
zwaKs?31JWM^BR5qAO6_F_5+$<_Wlp<njWs*K6ryb@V6kbOkqs}-fDd#gVvFkI)Rlj
z^d&YM<kbUNYO;dy=Nj|;oAgD6`P9^BrIzdEeivKLBYYbmN4wiE@E7-HCcn|SU?L|h
zNxai}IzQ(#{idDok$gRqR?WH2-QJz!2o4E{O{l_K58<>%&9c}L?uHl-1T<gay6pjG
z--DKuE=m!M*P6q4sFIC4i42+_?Zaw!Nofm$xxY9&eXPhoHZd4?YE*Gv_8m9`Os(2B
z1WV`>ZTvmpHBBwqgM`F<JK3YvCF)b9>kD7=-xYRP`g)UOb9CWGwL{fHz{$835d%dU
z(=g5%T4l0z>7(6z!AtSy;i<xiV$m$~?FK;yif0M2^7c^V1caXRa{5@}ql)+!F3ihS
z@?RQQug;J^Qk_5i3(Nox8Z;Tby|nCF?AC)4KKwawqS69nHX2$j7<K_pfm%AvpWv}z
zB7^q*EikcI`}se85^+}U83JPyOtht?Eh5SQ)u+Uh)0n<uP@uE4{u*9L8P&2=C(OPQ
z2&HS?HJNoi@(n8)6=KCyW1rD#(M6F6^5x|yI-@kYg|CZo?sxdYX*=s;hi6;bw%DvI
zk>mMBukfBEav_}MN!oA=pD;a6NRjMY7?`Ea)b<-DG6$~+h^`9WmHE`ci2ToBXBe~5
z{rr2|ucUujvLhsjm3Ga(V?P=&1w8!aHHo@46w8&WO9HjnwJB?0#{?o;q+(rX=~|<b
z_dIv^*rz?_+@na>6e+;o|AT?}*Lm_6NlQ=unFAj`+S{3(uyLiJ3=Zaf{}yzK9{sEt
zN;Y#VY@YJmgb8%L>}QAe?(+$R@_3)|0)%+oUy`l%7Amv(Fi~vq=2?Xhk!GaN_CO29
zp7SETbJwXn)MBHsUwvYFU>&ZcAr;n28#F6|i-lJyc}uHs;@%P%JGt9*MOKz6rXrI@
zl4q`Ji8ZZm#Ie@k>!87{0p`J~#H^q5E383L;19OtN8V<rzG3`dMnnVbJ*W<X_6!;=
z96A68B5Qwc+^D_S<T@vEjx|$f>s_RhI$x8FA;g}h>cHD*sj8)|opxTtnPQBCV`%x?
z<hk*Ck=k7k@}ckawMh8d%o-PK%Y*o!!<^YI!t{5BCI;=(kqGHLSQGLQvi4B<Tf;XW
zKg2~S`DLO1vO*q;N6;2qn*my%pGu0%FO_N0h=Qcdqx0G9$nu_*dzj&|({8JfJOcCA
z>rXrG*Sn)%G3Ejza$Xc0#10LW(~OHD^(4bcj2g2CTE(8zRbIgwlc)=oTDWVVa-V2%
ziDK&J?%X872?^cs^|W}KcKSw$)Qe^CwAqJXfFZ*~CT`Tmyv3A*)Ep+8NXr6UL8Z2y
zzF2<Ed8Z~nlH%6>+tDZFo8jG#y*ruzF)Po2F82;7NNe9ill9wN=iJ0(_@$b@jqd(<
z6+BK-%P$nhTkKQuoY_7)Qp3vC)77yGKE0YxZs0m`cYAu*FO?hM6&<O_rIX&DY$6IA
z$5D7PcuRA>*ZeQJ@m|%FD6pXYT}2qO2z55F-fpQN-5xjR8X7l;N9=(m>5_QAAhP-r
z;cnTc^7#nv97RCNMw@YV!3@UtCtfth&`?GBxt#Dqa5UAM9l5>l`C+b6rX9zuwkf1O
zScCxS*fsr}k`VErB>9tzFZ6(ra6p;^7O#xxzZ_N|?7nLbDv9`U_ybsyJj4lB#lC6G
zX<lju8A$mk8m8oa_?k^ZVw@A!y=q`4Bq+&E61!+POLij>2HH_u^FNA90|l8K((46&
zm#gMe-?pL@@A0Ir^MRQrTb<ektD4zht9QzamQhYT$!~~V)|_h>Z0byTZOwfG&*3Uu
zMCk0Gri>?V(<)xy>;1qLX-`voBLGk4Rn2IS9|h6)GsLJ<??+fczAhBtM80nDz{BV7
zqWUDURO^378^PMtc&qJ<wfuYSv#z)u+si=~IS7r1$Lu&%!}mAH*>oEi-SX}#t%(~|
z>#Z_%?i}qLjfXHSZ?-Z{9i%QWWhjyPg8QU}{FXO{dCe`}&KEfQx&#C8`Jp*}g{cox
z=9z?fbhHBUZ+KjIaGQ6|=0PwU5IWjf@T5;VWI7B?ZXUaJbCGguyo!akq2~r~6v%GU
zWbfXTa_aF-bk@LXT-FhP51RD3XvYn*xVMr8)IX&R+tWJ`{AS`eSL$C39Q>^}ZpY;2
z!t8Z+u*ub0=&5lhdwfz>xhiDM3Sw17{rt5$A*M(aY&Qx*r7}JsN}bSl+eA!7=Ta+#
zR}QoUFxH|c`=DgK*1%MF=P#T&Es%^B@15cEWuNfxakrmEaWBOHi}@q!XRiw_Y$_F*
z($u;c%%^pv%3?QRv>6I7G6Ym-_}>QqBBOI<BeTpM*RE7c*4sW_xFQo^tJ?$<)$^I8
zZsI(4B#m7Lj|i#ZG${(p<E2gy(Mpo&UtG?#EJ68e7ULZ1D&JBMJ}_6lx@^ia+|@gH
zfgLKFzk=H1($X(s9l<cQ1&@tfpgc}2XFr0=N{C}F$`qvx$@Uiz^x}N}s`(u{l>39d
zWZ!gpad?Ou<~0~23(j|l|Ay1tfxhoaE#{v+FQ9HMEi1fmE8xeH2!bZ0yue-{&BXl`
z?B1tD;kDmxRzlftDQR2NoL2uJUH{<d>zsD=P=l>|F9pMhY{G2a&YbY8!Vb)!2TFEG
z&l(W^d(b_6^8z;jT*hiUdo6%*{kf;Gk;C}2&8lYhZi9-|?+jXEWkpktzVfeaKtKle
z+qik$A30UZF1LnR`@lXm&bQDMMS(^Eo<6rDc*H9N+noHLs$TCq){Z~f%?x%VP16Z#
zy}ufB{K!J~>Rz5a^{Obs?e_(N&k4iESsr$u4<ueqg3VJXL#cCtf4Z#jBcYJnw6@Qg
z?~~%xMCeUCP9cv06-NE`LBTN3NezGlfEL5|Cgk^k{K(E+J@c`>%5HCBsZL&seZALe
z8*5hS<W`LM=+3Za-aB&?2SQt#@#|kM3JT4OlMHIir86Q!1E9@3cpD*6GE~ZfN!*79
zUX6~u?{`#0Cvxa-Hndk~zYnFmoFDy<mi}ITiTv^VF07fBTUa<eMBT5QL1^X;V^l$P
z18R8AQ&#?LbL^|Sxmyxq!j|J~+1SsxzSL-y9r%9H`7J)d8N3HyQ|lV?-8m9aRp+p~
z6~g>y-_UJc6tV1m<7Mw$gqlBXF1NE)eM7B2=4~@Hd$IviMY1~(-4}W^*a{TM6kmB5
zm_0M#TW~oleFN?ZYEz!)Oc?DnS@3mq@{4-loAk{PS*I_OUd3n|$TlRQ-c3bWQD<%=
zD6OQGp<La}t*cewjg5kFZY7GJc<6LEeQ)@3?+nzJKHBpvGtqZp%4L(YFaQ_VN`&t=
zisvK#2NoTU|3ttb=off&AXTKh6LBI|4V~bdB*xTHd8J-F(aX(rJqxVtVHx9Au82;X
zwO&C&WLlbKlZ$>E=~~8UXr|8|{9p{i9{k%&J$&;){2p~$vr7LGEcMtm&55EvB-<<~
z7)L`y@)A?w{AzKJ|LU0CZ@+a<_Ov@&kSeB$xL}_@9d&{|nl}FT`Q|voJ8e4!6?`Vb
zoQ$4SL?0Vi1zO&d8<aygw!D~Qzlb^?3SXr(AE0zNo!b-ht<G-9%QlmTVx_U?{V`0N
z*(giHe6<<b68)k+{&?^RNE0Yywxv}S3pl%Q(|I@h(iPY=Vx)A)N%zX{XT0!eAXe4C
z=wblxjM%<y;1)|px4qF+R?~_*D_!I<fpX5gJ)CS7^Z;r_Mk0q!rK2{^ol_l=`zUl{
zW$(70q9z}>z9o(lzD(*tGuq-tH^u}DQuCd_3~tPYepZ7_X`68$1}d)Mk1Np!ug<Ol
zG4UV&U2ld_-xEvcNDLESo4xq7s;Pp2T5CPVI5tZkBlod(`-u~=S@#fL=t|C3NXU(X
z9#(Db@vMu4c~vdExj~b!aJ1IW;=IIw_YTOvtIE@?sK(ZjM`pfyNm)R6QQPE-p0Ag=
zgj($mF1@HfN;YZWS4q?TxuH8QKj61*7nP6>{i?s~YmsZ}8ZNds@eS*GJFtQU=zJCP
z*;4EVP#W2-K}BQH>l+HgArHKlN&yBa{jw+y3`Uv!8}r?Thu3J#yNm4xx7_B?#L~Fb
z49c`>W@95K)jaYx`7}j2yY=3<0fe$qcoes`?R<VXe*lrLFiAT=8!EDOn)zv7(G(D5
z>z-M*9?O&-_&Fi33L_?=#uK+a{#(-@ptKku==<(0&g}1}rwlyVbxM{H$eNBa*NozM
z^1N;mQfI`_Tis-;y^sUV@HnJX@a1$&NMN{Dpo+=hg^})nqAvtUJt6;Ha}B}fD33Uf
zsu1=qLeAyP@nv%<@GdB;QTvO~p{OeXF~9Ak9i8OaHOh2vc4V1#s_h=nOo=_RCvqaP
zv`Dc`QZ2W64Mmfgl=ON-wB49(*=RUI4DT3PTMIC4nX$w8cmpG0XL2qgp5$)3jr<Rk
zNp`lG6J5`5-0Zd!dl3$+GL@af{NvV3E3-yFBtU8V`dDux!7Y4hGBsVMl#WhHq1lQ<
zXJ?ov$45dDr5e;A+JXA8(8_}Wt5zTh*&Fzd;=CaI=uP9IaLCfbzg^WmtGw8Lcj4bc
zxGLH~Hp|9AnW3aqI8y^04)D;_9-JS&qwzQtnu1lV7Z=XL^(LdsTjh6@XM~A6r?QBS
zZ^-%z$$gIxPP&L2GZ2HMEaW9wb#PJStpk%gbn<&FlS3AG#PCE{b~ZJTL^cYE8#|^V
zDF?>p)d!cv7q8!erskSCGrCW(JUZg>@)O9nW_>?=ibvC7TNK0d%$=&%H)bAfgb%Oi
zL(^CWUtC_F%YX41YrB)r7!zu^{ImF4zrg`0XQ!@jQ=IbZTy=t5PC<c(D>+xg0J|6L
z{P}6@y57A>go{Y0qHZR+D1VoM4j@hMPikdaiRdOI$iHq)1wh>&by-_SPI~V>Tom%S
zvw65C<e}N6@OK%p>|sp<4Zpih;4@yBQy`T<MDov-nqD6IeT=N1$JjQss>NS}JwQRh
zoYFn3T!}Jhv~vA&!PA%_G$68JNU2}O(oD7@R8VV)U6{+p@fK|kE1JJC@Pj`+{2OV;
z5p#R^qI+p5)!9|WRwy?pCFFt^t4>``gpN)JkJ;Q?eMVW*zGe`yM|w-2ChVokUw^6f
z=yr0BU9>7g4<NTBfKROG`EUl1?F4~%qcdY*;{c%3B==pW-7L7OW#ln3hJ(Kf=$!D;
z2_Pj;^JA0O%es515_OkDCOA+F>_^?aZ6d<*C~sdYkHU73k2e9t+3+IGZMYV$I%<>2
zpN!6xvZL4R?(7)Ufr<JM40qTSOZa~m5QYJQTicag?PLY179MM-ZH3|frzSg|5G5-#
z#NKfr93-;ukR|yvY{Hn5HlncQ4bXZ`9EN0-picXa(qt@tF^i#^d8u^95M;KD0YH+g
zS}}LL6<r)Pf^|RUa%kX>QKP>4%nk*|*o51Wz8}{fr;_yD@z!IBk=G4a?45YxkchD>
zl}99zzMOHK*)^yVc$}}}<_m5pEQ4Hnmx@up0Yh7(c_uM<DT4zA--u^fxL6y_NA%Z2
z&9)|Y3`LCq32f68o2;Z{6#rx5RMWzm$&rdz>Q|)>mDI3xxjffSl)S@urzM{}&A5bn
z+QfT%4xN~oYqREN^AR#6)`sf}tOq}C3@OoX{}OgfEaLCy>UNfIyU;m1A$4(-B8N)|
zeGIuok9Zl`*JZ7UI;fYi=7!9xdKdTF%wd;U2ZKv6gX?9DJ|<qPnH9xYohHjIGmeeE
zqn=755}kEXteNTLG-P7GRiDp5z0Y<Wj%1d9t}+oJT8TY~@vOYlUA0)@3#%M=PO|&F
z<C%kxnq4Wr&zzC0$#5UeXK;F2ALKXnSU!-*l$17?e#k2rWow6!D);shQ6iYfb6Pg;
zo8?E}ff)K*spV#fLBLfg%ynwf`h*6yVg(+9gL!QhR+&-AmZ4fUNulHX(X>EvY^4I#
zaOGUe(56MHIGFWXrj_oQfW3R~>8|Ob5=nmhF!EU?m`HS647vXfBs_!9&ZpY-^LkoQ
zXn#Lv3s_bbH+6w3dAT*96S`Bf@e~&-3Y;8+riRw{y1M{UPO%cx+m!WtTtR3J7_c1f
z9%GKiIX2p};f<o|vX$$MKOJjB+grA7KU*apZ$`6$p2|TZR|5>mTwn_3F634^IlpE*
z*T+{(7>QaoGLZ3^H9dy4%**}U7>QP&Yfqv>sWjNN4z?c4*LP=c<`l_I>%5c}aUFM?
zp$ZHR<Z;gDu6}>n^Xw=hafQ{%4j(%){QMB$e?ZUHebeO!d@gHfYZ?YuXa09r21IR<
zGd4LBs<V0k>n}}CoYl>aLbg7UsD@?HF1Z;UqfG6(Xk{?3&rM4=v-j4_G9{L1=A|{G
z*{MmK;n9u*Wy00z4*#H^FJ?j?;t$_KA8ZGVzwa%n9A$g1-hO<2gECA^^Jlan-MPxR
z90j>8e})OUO>aWBnJZMQ+OF%Y?0f+-a#ck%!Ih@JyZ%H`uyuK$;N8!l?sx23b8A#o
zR~Hk$^@7f=kC@0#L9)##Nq8EvvMCOtD|L}j>$8%+adLxB4QuoDU_B?^)#8aQCV{&^
z0LRH1*$Rxi_24X{BG$sl<PqKLZMBqL`k?HK<m0ClXVm=kR*A<oZ=bG;en#x~>6J2f
z8QYvYIy{s{z!W^3E)e!qxrZAJ1eg;BZdb^evApPv)vQ$WU{4K<b<I#Sf}DDMN7;le
z1vsz-H!DWdmM>fi%9c3Haz|6737FNbb`*}Igcviw-bBdDolfZ5w%y|TtBoMWU*(9B
zo&LiMbfE>Dx$RhS%%Ly7fK=xMCa$cZ3I5RUHq6bkq#^a|lOB$O5Ud_!XU5ti+w`e&
z;u6u^<*7yPGqAlf&Q)VHC}hD>PX6c-^>$)HCs9(^BR8(%*W+=guH}(#^J#4XL)g!X
zY_qP*GXAk^cI%1&_k*@okM5sUG+}pV&G`HWK0AT~*p5}qT-PKg%ddYSWau^Dg6>Kf
zkfURLin?vrL@6dRvp8|7$t+c?ot0tIM66uBI*qfNE`4(_i&;75Z<AX?Tw|hp){@xb
znbS2y=ckfa(NC3jqw)QxN&hLh%m{VuC-`p}u?JiT;YLZ)``D&Q(fRYbF<~l5=4_Nw
zGJ?4@0?DFbTMQcI_vB2nx^-yC&G=#0)S!t^vQ)7c3wn%?Z6d2xpE&hy-OBH6yIp<5
zppGe`oF~}cXsL?P;RZ}Egn!G}EY}`0DKKTq6}ec}Uu$pE2F%(i3dTSnUL2PX{7yxH
zm*Wx!af@Ysm&6Uak$}i{wW!n_RaAr{xDGtxcFRDo)uW>mDDvzY*WRTRNz7_um9leT
z(bVIwKcvle=UcfMb8TOnPU2xtYqqRiNJ*L>Y^5gLj*u5PyB6+rfnk;#EJPn$1~~cW
z?su*KJ&qkTDL?|t-S8hNT`P_R)}0wOnw(h6&)a#*6>5p7P^lsVw*gZCv)*F}qV_<$
zxVsX~dz#_BOG$YcQZ}dG#sONmp~o>Cd?!m-=9}VS$u`>AAvyr0zM$`}n<033Xgv_a
zmU5d8ZSw^JR190fn7da4PBsnt3#l{<gQfkl*MItA>>fz)lc=lB4ZN(aQ$r#?)~hW2
zDjYy}aws?XS*^IBl_)=mfAHplxt`qp*XaF!V{qUbf1z9~b}Je9{i6!v$<M^yS(rsF
z!+0F=(7V#n&+}-Cg`6po>uNnbt=_AL+O1k=?r>}MEHzKS8wJ>tKe_7!z#*;>BF{h6
z28kQWHfY}(H*O72RE6aGLq;5U#9$Sqv1D>I>>(zz`>kg}4!xEIi>@26VA{;PwQ|ri
z?aywNnUnm`<@56`Ph~hJ?M44Rm`|Ec64gE6OZJNA<r9n}qt2S|4exmyKZ=bOq<Wly
z^?hGpifA)|!mwq<-p1+SjJVDFn0-J3<r97L5c1_tpFD=+yU>q)bKYz=;wODZOY8Y`
z`Qt%wHXJD~EmNvEmo&wLSMN0tZ7K>J99tG-JO?N8_RI!NEZxKCHTo~=B_~onhoVJ+
z$Ak2ovz~dA7Vf>9SY>+lwI#Jkz9r*p^~gCJxH1V|)o%&?ved9Rc#ZMfjt?0^Vda0a
zF2BEN#GFt>8vd^#4@lb{J_4;XKCXu|U8n&D>FhWmT8MFF<&qPhE=dbCBEm+l7e|II
z&$BVmkQX#w3qTS$NRlkeR}P#tVSjX1XLGZk;KC}LypwR$$0-+uzSM@m$0$~{U@3Zm
zH7o0bZ@N>ru`P4QW&sHE*>r(|mL12XD=U2es<>!fV9!E^jX!+`lJ=uUi_Xa@oG@f3
z*jJfyMA5(KQx~947q4fz0cr!T_igA)6dYDk^3f`r>>{jE?3po}s)?&8(-UcA^8z2=
z@*-h+@Wt~z7lu}WwB#<mX^c;q>?Kd~ezb$RhUNSsjXfGA-h0?b%|Yq^J1}hK+Vt3}
zAfqlc^it<Ywy&%b2ig{RQk+(gD(O0^Jwy+ZsjSLfuXfza^x4JiaScPe3rU$312+uG
zo|&II?{(2EId&&L9#?9sEDj@JbH|YuV*{PXYg&k$X;O16M2d;YS}H$>T?p=rGvF7y
zf%AC?2NlRf=U<|ezj<L}$E&Z|A*>JWPEMwkG35{$4PIGm6y53$vpi`Ld$Dp5+_^Sq
z(mByCkp~MWc-^u)yG6vLUBdlmya+bP=tR_8LYW#68F`d5StzL4D09s&Q9AmBU!{N_
zzfLI)74SzbLI(cy8=KBRterznLPGqvI<DhA0v2Zv5CpDrGdvP3z#=hOab*-60$;K*
zRAXK6N4mQF9m&zE)n;@uDoN(q$=WBhztlC@(Exc>&fg*j8|lVSzZkpn91~-cH)osE
z7mAT{j>hHHe5GC$LdqM-|4`O#MMo1Ol0TBa|LLDueZP}Gwf@&D^7GUtO3$M3Yhb-l
z$`bT3E3OExa8Ksb=-hr!&#Q>Q7oPaNQU#pjf9Sq`c^wNhzRij}>(gM)%Y*1z%`llY
z79!~os?eA;Ozghf5|mzou__;Xn;8<oKgU5ui^fjo!bT}Yn$?Qx9~Hr4u&pP&7^Y=g
zfMeibJ<;<7%EL^C^y;kTobdO{f8-YWFDigbsuD;`jBAuqO%r}ou|a^rF#Pku8pFGE
z$3K?bDn9l%`@NaBs~efT7r(xdmGXhG05(t9yr16kwB<CVxjetMCYu1u;N#(@zTdkQ
zyz`F2$(g>sUq)TVpA6m6`d!~@ZfiT8b(Z^a0*w+NPNXF)9-mESOE%P{GB_aatd*co
zC+8HkiOt&dQKGJ*#@oIuISM61jya|9D|2az#&A|F*YrG1HV2}P5rLiKm>97V=;V9k
zB4$CA67G~zg)B8Xb*pqcmm)4Lor3Jim;lGB&veX?c<H0LkfF3vwK6+cov;4&v*G&C
z2R{TCg_J$W{@Dgw_om!x&9brMfrC`Camv^_Gfj9zSuW?8h>X+k=&}}UL{#Rc_+()L
zw&!Q;$%qrfO}}$R`M$6&<a?_Q)!=U+i3gSNQ2Jr>@22!N&ZW|i$VpMD&xr@ltqQkq
zxE-e=womHesTFu%1itX)OIGQ5WFIYbc}@h}9yT*8BD4eX76ykeru#QT6vX#p#I*K~
z*p7lA`8&%a97TnJs~?v{1+uL|44Y)M<E_Z*@F85Ai8)pw?-lkZ!6Jp|`&fd$={zUe
z^E^fpwLO7DX3fgor)uVp%rNaZD1^=YOgM`)vK0ZzZU$)5M4pW*-dj8M8<&cUK7ztZ
zyBC)Ooz4&cf<^NZ;7*do?;MtsuYc9(HNDp2wvnd(6JKoE<Oy@-MzBOD5;^I;=ydQo
zE^FR<8T2WVpVv+<Ap@s|vx|+fVDaoxH)eWw#snj5o04~);BtLXlC;NJ^+3G}%eenj
z4Lqdh={fXUBrWc1BM7IB_7Y<{JnQFRQ3IsG-&-IncjC2tdt+Q<kkpG88u4@$8%$GG
zW^J153!h0iY^Syp8p3r&Y36;77$sjvFzQ$A&%Uwb?(#^M#Vxn1p?W-gJgbs7NlFaC
zbd0M7^P1E6dc~zJ8>0oVR{{C&#{6w->Qn*p55aQaZsGZ70YFi<0rlGh(a*AEe~BFv
z61UQ)Ay_U|<<qF3HWaUH(Pdo@T(^Eb+otDEllI6C^Ak`4bCD@tQENX`M;0~`pM`IR
z1?)3RJVdOdbsJobc{-}{Sp!l@zw0Ts4EGML!$y=%WmFl{2Y%3EeOa6C-;ka0d;Vb*
z`gB}AkF&`pqm7|X9_?d;a?33A><Wut-*3gjiazU2(}jS-*YD|bfTYydo%Ow_XNGLg
z!mNfY!YFU57kb}|jLAyS*Lqb)vW?w@=&Rr@c{mlR*j^EB54lNJt&=c+<J1OWfeX5~
z>8W?-;IZ$7jXgyFdF*kN?b+IrNsBPLr_PZ0cD9xCu_^YAU*l0WaunJV(UWI_XAAzb
zMFqFJcYfUA^R7-@zxuuScC;oU+71(Acp696K5-j~g)}#cE}HPNenHWS`UrJ&o{#me
zqPIw@3$H#Pbw7P%51hw$)0KUUdS1pq{Z~=^7UGLT>9SD;evX944!&jC@Or;+YvRff
zmmo6>kl7{KwEka@eL_rR*o$sg6&d~4BSzB5Ko2m_yFp*XVc4&mBiQQ%k}QBHyfF^m
zVU?ah4zb)+f2ICo0_o+SLV|ko_XUEXGm-DifxFZB@@&U}M?h(}>$hr%`Co=208sG&
z0y0gC)r<-^69Z0%;Emrg-9Q2DG!b3)2O`mnbDuNND@G4t=DQJlF(3_wMP!#*AdMmU
z+y|xX8E}L+?Y=!>s4?SbkIWAe_#zNBt%O_qO9%M8N9IosKLQENNnwv<n(HM|Lt#ep
zi-z2hZM3j6F&HA6?*ddo;^}v<+U_oapI5QGh0g^4%mu1E77d<MI%9jep;0ZtqG}Ar
z{rTD}99kcni_hkrQa0{s<SAv7Tgy#$xu1q});^V?C5$O&R!D^ND&|eT?M!YyuYwWH
zRlDz{-C6RZlK~XiaQ|f&<v+AT%NsykJqGa-u^((+5U~_UB1L>5DGMO7WJ}O$I_?Fh
z3~0BS+$@s#Rapvis~hl=4f-R;Dj<TlK=C445j=mGs(b`I7nbs`52&M7;~1AwsY`SO
z#kYN**x_gl#ILVDQD9PgpEkJqJj~>Iy@dOG%b59=y;_(c?A=X_Bt)=oTbh|Whqn(F
zg`TbG>BUGCL5rLHNdl9t?#+9F*m^V`_Jtm3sF0}4gdEfm<E?=iM|ulOHi#8Mgx7p5
zIlDO|XxD9d|2tMZ{9Zth<oQjr>6>Pmn`nw(Z4=DN<IuBq(&GGbql1*V6l45L#2{Tl
zN%}5Nqe%pSHvt_r6i!}uarX=uX70?<6;b}W8;($cm(`tHhcv{j8|O)O+f9E67A9P-
zPO0n`<R#dFtS=L#v7c3D`5Ka5T~f=diW$eraf-g@F{VbeG4fehp=XC_VrRF!?CYUn
zCSGoVbLS%QYP}V@eOsw4If{})V<c?ptR_zQo4!Hs|0H&OAWvHf4-P%0R~^$5c`p28
zDC&tR5dU0ZNGVr`d|_Y7NvR_kR36!skO&_ZkSu3;ly%tO9qJ)^F`jw*zcilk6?_C-
zw2@zB<R?20`}$Se+$I~GU(SJUF;h0*U}1K8TwFp}$6Sr@#+$3Nk%G+Hi<{Tr!Ou3(
z6p{aVaGZdhxy9@YBqO?+TkThd!*Rfx&oWabc=1H1WwMIrq7|s^+s}0f$g5Wivre)<
zG`8?G8c++V%E*VuArs|$bh{nKCVUqW2I@<X_k1~MNmdS}`8v9GuPMbt^sRz5gmU<t
znL!NmRJrEGYkPwt`;SnxB<r!oSL)-^DUXGM@~fGR0?jEY;rdQ?(-@GhxQhsp_J$Z{
ze2?QwS|Q&tX)qC?wH>O<b2HRRl)xm}S(>K3UsLUpt`2-u`IyMM7ZE9m?YbST5B7h?
zT^}0g>5?+rt&fh2J~)kxW8#zjBMeyQ|4+*YA`$X^ddCQ4U*?KSn{NKJrbv!`8Q{^l
zvW6JD#PCi?_7T*C--ardjz3odk}PeYT~}ZgVq(nb^|JCQRQcCyA-s3rY3eDSj-Lh=
zEfsFKv5_`T)(Hlzx7hKxoH6&N!W~Xc*XP|AgTj#re72`oGCZ7keZ#LrgEQXD<*)t3
zI}v-ImO)u@^W<ZZkM}s1AoCO|lfxUY^_Of{^Xeu)YvyYfzehy$6P8S&`@+#S?iDzK
z7x9r{2#}X(NH@XzW-QB$nt}F)w<Paa$?96hlT}>2;cFV}SD14GKERuDwEy;_<fH#`
zu>oDk_XcB^zTT^~FZYpG#0FPL^}>nyIyJx8A+7yW%GON*t)&3%8ZO_ZD^Yl*dp9pK
ze(Lm5i_6W;TfDQHrd@-}-GDFZhg|<3EWd__CcUV|p6e5#Xs3CCU^u!3o3drUgy3iG
zP;Mo@pejJQQ-IPo8nR^i`Fq&P`l`VU!VNBcg5OoBA#Kj*KWk{!g~fzX(YNQtng5#r
z^%heQTf)6@CC{mrpH&(f;auQ|Ov`PLkL%|Q0h63r$<ZI{J0Tyxnf39wM{601nKS}u
ze^q-V`6xF2CA1>~{pOo7U`T4T=;+Le7C6k3B)gA4P-FMu=vfOrp8e{(LF!HlvR{bH
z+8y0y_6iHVo9G~xjZ_`{4Os5Msd5=_TASalHGj(N*EOZG>7^WWS%P@r5OS&f&iM08
zA-gT`Vz4wOQd;<f!KS-GuOi?v&rY5mEUky9*!>*LoW~~4rPrC-@oRqxAB>MmF`!^o
z{qSM|Yq9o*fOSgW{RRt-;J#zI+_FQ-X&wYQ5`AHtc`~Jf_{9#r-{|A%dz@;wMxx_c
z%_s5igC#sxSZ-WX0k9MAz3U&ckMI}Qcz|+!)_y;{fUnMDWE>z>wd3J`-sOrZa9)v=
zh%o0q(iC+$E3E7O3rm*i?$o`TKLlpV!yaW7BA?%fGoYrKLnkoA0%M8dE-vKxBg>Co
zEpJ`C!Ikmh!y#p_;Zfs?)ZpMyWIo&B6`LQnbgk02S5|e4_(mYhL@M90v)!=CC6>1u
z@hrC2N=>;ZiJi>H@sKEho4QnMtoRD_#oj#wlm;<hVDMd20D1b-k~rId_Vh}_XK@IR
zunDR=`3J~S<T0J_hBr~CnKZ0+h6#zvCx|2Ug;(*7<LPT{=jytm=jX-20XBcLy|nu8
zgIAzzFSHhPjI-0qy=Bu&9^{Vs_8i+x*4zbFG2mekvb6>R&Xd(g%ADtQiy@R@H$R6e
zlalk-f-6l0RHe;LW2dnKFoTPvx8rBj9C%ZH@c2bG6%zsN-5~D#4-1Jl%@{f?Oum=q
z3PqRqodBE~^3?Vda)G0yO0j}*-tpV3aUum1okQyoF}TOF!=NnwZn7<RpS+ml#Uo|G
z>A=kMI7MM+&@bIjDOa5Y>?y-v^qpqJ1Y1*9V+mmgD5Hl~twh^6@5&}Eu@zcICGsXc
zHquGA7kPC@=nz1`DvCzK_{Vhj;RC`WoqF392`$qkxi1S;v2f9%&WU*p1&k<`kBc(E
z8mRPWzrWYSflUy5CijWD$VT`vwh<8B?(FDG*f~DM8K{PzCmQnkwi0|L#s91xK=NFW
z-6FN^JIJuvY$}D1>Qz}_9w%#*6Y-s_s@{t~n|7Fw(^&Yer{^o`<P7WVW(>pd{AzsY
z)iauFO`9Ltk@OZ;GVn=|x@3&LHC<4Cv#@qnW&r%x9Tx;(F9RWK9EiG6JkqzRZ!*r?
zU!N=Fbgw1T1-02~$SqRx7hKKxwo_sY=Drwr9v(^;$^m#rOn5Xh#y>~Whwt0ZK-S=Q
zH0HKox*=m|2Doz<twzfg>#gfAsUYnMm+=}i><>W#v^=qd^1KW2UpSqoDq!VlXL*@7
z$L3v>sO$J`@%Acfw4rCu)Sv^XVx5h1JvdLS8WN!V<KioZ;|MF7xxk3c*ev`tzq^U<
z##vY}6;p5jfkHWtu0Yl|9bu=g6Mr$ia&<<YOfpBn`*=i6zomL7!$(ZhLyLnnlD%#v
zm}vu|XJSuC?gY_pbPFGPzYq~jpX{c$rX-Hj(DPIW>ph8D_CD+^1qsRll-#qmNW_or
z>-ZD2gWT|H8h>nnV3M<4!Ro7{Q1wvU`?pXO{TSo%--8)wM8dv<;4cu$NGo(&30kw-
zDSbs}Z0Xc2i42W+lj?zZvtZR@+6>*>O5<n)&=mQe3F8V#8$2z$=3B)bMWU7XX}kIf
z!Z2$W)t@9jTtLer@N)=n&9yyQ?7{M+W6mNe%T;XcG{71}&>S7V5qe9Ra*RvJtPlFL
zSr-sGW%#;($4fK5UXi6vs4V>I!j|VW&fNg0`j0}n`KhsCnD-}1|8%x^5ZbN75IKwB
zy2|nkXP}{)uu2D~GRotX8sv~M5Ce6(Gj}F(i~pfK8snZ5Bt5NQ|695rJVHhcIC=XP
z#0b14WW<VaE_yFBLHZfd5mJ&u$5azV8xQufPSdeOml3h*NOlLFP>>=2#f%$PPIxaK
zC-KFHV^tiGEX9(c1VEQ$lidy8aF;sup|2(1;Eb!BO|80y7oR7h1)3sW7i2=vLx>uZ
zc^ktIFRr$~DJ+nR2{F%n<YIAFFptnGL~S6m%CvhDHw|yC@C89Fe%B2+U_&h446j9+
zg|GL<aQz`)hsFZ?3I**rj&^#{uVWBNIM?ofCvsu?K^_C?{u?6ws6hVr;m8>uU7hVM
z)?Dy<>M>;wAIR#n(pJJq>L2iAD9J&?2F>%O6)2KvOuKYr6%{e6Z_dq`V&Z@BnWeod
zl4E(XxBf)emC1~eQlNKUe4FFzki>c}yA~Zr*ky8Mz_OaeIAdr7HZfRn2G{chiLZs0
zdx`BOV56QR45t7*O$-|Q74s_nxUEJ}p8Itl&W7vw5fQ(^EAh5&!^AI8pGpBV>Ohc$
zB5-ufo{u<zo<a<rj!Wf&${DZ!BK{w#X24_eoe{+C+YX8c@mE1TI4yrcf(I`gHYqBM
z`-Xk8#0sU7vI<qS7YA@9x8@EG54$sjJ{a<rXp~><u4#NDexBcFYD04iKiid{-(tHn
zb)nQ`i6EdL3_fPF);8QD-7K}GEkJOYbsaZ*t~)9lSZCU7(ze=G6PG7zFR)u^Oy<4q
z$f&Cwkug%s*oVe2yfIvW?AkR}mALX*^zztewhne(m~GWEX6N;MsyFP6uCbhB=Vkh+
zDtz!`+t=Cq$eYWjl<0yO&94(n(j9RZ&769=0NHovbyEBGY~!&4m(G3V#pl)h|7y^@
z*iz=*HRMeUv9gl>B27x<&pJC^E#qR^cPxamBe>I2Q$L*APAAZSN@cNOf9qbw(5A`F
zfZc0*gQpvQ7&*SJZ3glQHB9+2+NJTTNer@6tcq3ZHESZaUS=q%6r1t$4hnazqx4Q;
zJWZ|JT7wcxss>sb*83>);hE;j7D8ARY@RSM8*FbqIEFlN-C4&=WS{#DOgWKrbwj!`
zIDTUX-atrzyx7hLc%$plzSrK@s3~O5w)q!s&FuIz#>1~c=t?=bm}mZ4JU~_Q#e;bM
zM=u(l{##+L-@YFq^SiR^a-*QL__A<b+eTd(rQv`_G@4KBbsr5Aw0GlhacI6e9&MZT
zy<yFRicPML5-A5=vNuR#vL@->Tu1M4hk6HA_rlHV%O%DEEN+c>VXl=p6lj=hf3o^{
z?98tNrRijTb{;8bGoL9Jf#mWT39-GI2(SDKfd*1C>7wV{5B2k%I@|=-;eK~l!uOoR
zr;nAwcNgX8>iNtJwPraJ`I7g6PBTT5l(^9UXb$a%+82pDC?bI9@SNs<18V;#H8P2!
z!W8;WI|c2m;mHb1pd%H7Z8H%JWebH`(|VRn@)4|DKT7R75w)J)GBOp7bk!;F?Pd0+
zs^E7l%xzVH2oD9S@e7|Y!@bqA={KZP`4<ItGqKBaszQ$=YgLGL;A<ngw>wZEg6p~J
z9X+aQAqeY&zcX>Hjm9&#704<~5s9U2K>lt(zgzlVk<%<C5wGYh%$ht2g4TVsQ@`lB
z9h(CxO8eznIgkr>mH1l0ibFUS%O;<T&?^Y2ZAFsuyOHI$i-h(WKVB0O5Hk0W`HR#%
zK=430ud*7s%9`5L?+f<(8B$D?A~Y^A1I0^5c`2A&3Ded8Vw?s*w;Bjjt0Nd#2Sp!3
zjLnv1`_hZ@Gtw<i<H?zYo6oth8D$mK`zdP*2H4MPIy&NuS>;E8O*2t8?zEAi06L{H
za(8du1Ktarj-CEbu|Ja3ganBCBvh0U$9843aoPh#!WLgfFEK2P_!K4i9^^%-T~>AI
zBnI;|+U=vkgC2U$-cd#~UHuBn1lc<JP6s`8sg!ArO2yAA9lgHPDR+9hN04>6<;?im
zPcbcI(nJiA-*3On5$p_vzu^3XfSH#S&KtvN!QY<C3+G!#lWrSDx|DT#K6j*u`{Z}F
zxObo4heLjK4x(iHJ54?>(xu2PJ(qc2dn;3-b@9!$7fgPtD#==6jjphFoat_#79NQu
z#H?Jj*$^5$9UjWU(5%tvo#bHuraS|L1NH`|JGItRp`>r$rY}%<*b8KR@p1LCeoFUf
ze?(5PtF$_KiYb3hhJGPOv07tM;Tk)1A?1Q5;)wzImyZk@rMw&8s_)fm!e`ZRIsEbU
z;c1Ub-KS%gX1M^H@YUjETL9ikkjUZYUtM8^@$x~uH-@jKgWA7eS4a#2h~qO2x5Je%
zW-C4}uSP*Ik7yQ^ynwIg2f4l{P`5raqaIn?%rVm8PSb7cA25de<f`|^S<>TX-V4RY
z%u&p<;=^gq-!qycV1y#{4ScfWG3=hM$&<+kJ?oY|G*+A6RcbDedd5@?-hZL2!8=C@
z?a^91rneb?Zpp^%Xn~oV<;7Pziidxk_EoUv<!xlSKPABG;ffhGZm<x*4v1!%qO<uM
zjhk7t-MqX_Bb6xw!VQ^Bq%_2TKU_2}NU3wzu#S8&gc9{h)#2dUMH`KVEEKMdSpR|&
zoK|hI??z+eU=(PUL7##tvl&XgA_X0`yw!hM&gRuG%`OKz8=yxR><mUeaIPnmHw`43
zEz@|P-8WzB@vMtj@fA;K=2c7J8W=_-@~%;yj^N;-1@m^)7_7_ooP2>2SUNiGrd4zp
zRsJ$C^bsYRNox5cDXaw+OGhYKY}U+R+V1qZu)Ys!ioJiXuq(2w^VfM~JiBT1e;)xP
zo87hCHoahISmd*sgT@K`mYaNnHW0!}!M5x5@tfsVeyi7zQhRH25Z{CyMp!7$MwY}I
zkBH;}`I25|C1FP1EKca*Sn@)ZNronOHK#M&lU)(lN8gO??6A#yPxv;;nA~GYWEa-o
zA;zXhuR&ShNc*)LxtxW6Htok(86@a$_`diWbJA_G77B6{{U3G)4dM{hl6pc3QqPMK
z|5mTfJCyIJ!f(L7@&E8Xkm28Tw2t}+WKnOq^aUA?naCs*fsd6CJE>d#eqfE4-TGh+
zJ(NUV@47Tx2z2V;HW{H^IWv|R^tDzkSnRW*b(CmUoy~kwpXd(=BioyVJdSJ7thN0-
zaA>_oecoh9lyA$c>KBO9O$n|?2IH);LGzO60mrW1agW1jTN#0}KAG<BhO46Cg#DCV
zlN1?hg7s+6#_9tXmc<H^*m(t-U?LX^FYR$et5x)y(^UI*Rz*L`CDvS^i(-6=C-)be
z1u0%EZ|{XsE|rdaK$J?_`?h;Uq3r#fKHoUyOB(&aJ+?<Hz%I?u=H2rjX|tpIs#6No
zk#?nJcPaIwhiD8gCS_=Db5s_Rc1pYlp))<V#xEBJx&4S_l9oY>CPTgss-vaBrBB6h
zt*)lMHsq;qI(zGf9jJj=`6~otk0G~Ts|y*tN9skZtlb5&5Qr@Rra^s{PpNsHKI)?F
zExhf?2{XYSl}vD{hM7*g?pNs+*@mvqh0Z9x^3(k~AG+3uf%)%I`+?-q6($LgYPe!S
zLypy-7}Ff{6pfOKJ1u7Afy<w4fbolg<rc^|uS)C0B8RH-sim*c!_{BJ;gT6+*u2C#
z{3cj}Zod7aMsNHSLF;BP>X@4GIC5-VQWeMvY>|vwVzf8A4ILQ&e9AX7eX`ET4UsGA
z#$)|uBQ|B%{!<Vf;6DZk3eTwVCa@<ja+oDP=a|(yasZ=MAV+G1+SU*$f7V6aeDrWj
zDq_6%f^Y8rn}3IHpPwi#)y9%pnluqo3rl0fbEiq)X{NH!&AAvMvPz`1w2NfGt0;k&
zu$eZb@Bm#{v`neWJatO`FzqHp{iY{5$b6Eqj**C9)kD<L36V*7<hh#!`z-))KQfA0
zBNm@1o9=h;u(x8x!QL!>rX0aywrkOiNp1OBsQUkyd&{V*w=Ql}5fr3B*mO%sD&3-l
zv~+h#cbA}a3bJXCMmjg$C2YF8yBjw7{yEQc4(C4i{ct~CKkTu`z_IpPbItXezZq+;
zxgmF=;nxdG{aVZe2*K6B6_4)*BckPFQ<dfyFK)$<?$%KN-o4Q)hVbICz6a1AO5wi(
z-HoSZ9wrR0wxvu;gV~p;+TSnjYkvAUydemohsI-zJGS}%VD{KYj$m?!y9T*an;;TE
z5>vIjk6k{bDXrBDeCrhmFdT{uWkqHikOMoux(6AY1(+9w0d2BQb{2@jBc3l&bR)ND
z7`8i}Bsr5VRe%}VMQuX<yk|SYkWzZzF&{P8`b?*PQT&f+g&eJtL<~V+Rh<F0dm8Lq
z>=NpTXo33k8sidU5%J%`?!|-$dfPSV|NdYf$}&?t6qaiO`OrESb?R?%Tb_FNaAi4k
zz@u}c;kF|n2%F(ab}lVz5{d-Dr^5nH@+s?B;-kySnqN0u`5FmUNc~?gO-4=I;JZKL
zahvT$%Q6a{RInCZ4Jk>rbNx{KwztsQ|C^9MW)En%A~SD%!__Z!T$8|V*i#tvqABG1
z=P&h0pJ-FfZw?AMI@vZR@4i8zi$@9K9Ib(C9c&9JlueeT%Wr|Qo4`e1^Zy45y@tEL
znaHQBton_UoRIk0LIYx;@Gfguy!CdNu+_;*)oq0`LO_{Tq5+L8Kk&!UDT9dBYsy}i
zn(RcOj^i!Ev%s5;Pd_Yw9+Al37MW&e4knnoXUY-^u;NNmMliDgy<lcpqTYKrUSZ3c
zq$3yN2i%z>oKj>k%keA=f3$O;)+&&(5+Mj&;^qGNWI@)J*C~;`u)yl<%=ZcNC`$gk
zMZ@JV{mJnFDM_9V0HPNmc)c(F370-}z`HyEZf5}1B(aAsBu$}UUVQWfuj-iy@ADzm
z5T{SR+i1fqBbypOAf`3nWuv;)$9pCIl#4;WInD8@kfR`5Nm6g~k|d^BPFd+h#P@v3
zX2)p|^EELYBE*J+v&_X;M6isr%9C{oiq<ckorv!SuKjf51Va|Shm!oPXz&OckB-&E
zk)_QuR+W*N6=&+50U{yrg4kc509~0lJ+L1W_YZFa7zZ^s-JkG`u!t7~;dIIZ2^ruP
zp1wGXkWC0Cl5D;j^vJ3aHm;OEfuy{op|`9D*v<FT6Ba@2dxNM$o9P>5>tJ>VsNVCA
zl=yC1fl%$IQm^B&(<&F)S`hQD`f=o*G~K%}JMFl*&h*_bUIRtz4ay~$7r3}Ci6RqL
zR5Ia$2=_Od8+zY@3=^oDFOaJRZuQd_v`;gFuYjidtk5P}RSr@B-}y-em0YO*ts`i-
z50=Uj!W;POTwwufLj^syxf!vui%M<?e8F_!350TSnKdtuR%Bc(rlJjv2P4z;viI!5
zsTiO5b7qP3-+8+`FH?0Zk-FT-{uq4v2N5G+FDwl8fCrbV=!_(s<=pLUC!y}7Y6t=&
z2Rt4k{BXi`Bt}#<135g;pp>KXnx~1f(nY8!7D3n`H)WbVTWol|+=W9dM_Jjm<2kh4
z%q%;<wh8lP2UO*tq4&n{O!(%_!gT{xR797WO_4)XxO5S>caEv_?9L0hDx1XoGNyEa
zF!%c{QbG4S?v1-4-n&Ci-t~5|F+ao7_tY)bEB**6gdpm`r!nD}!h{7q@;^eD(ti;`
zD%v!?NUeud-~5T6j22Ta7bPPW;!Mr}F18d33iPuQ!hN4DKT4{WfX8_M<SFo>n}C2u
z%rm<a0So8ZJ};Z9QS3t@6wM?wsIeHWFfh6H_FVSXxhis=k%=q>tu2u#YT2r4eRib(
zqbV;|Uay;kbm)b_YqhvG@1l~b<&n1y26c9DH%op{9;zygzngBiXjAvrV<J#|<(~*H
zRNqV5@z~8Ey}#LL`jr@#oeiRP8F9=<^0ngev%hYzIeU>1Y%T-&!Cc$A)NjwOqRpX@
z!I7W36F?)A6t~`_P-Lk9k66R$mmE#;^$8pz8StSYt?TOQdhiTgSy;%AG^U{(waw}*
zVTnrIMXPm>xbXb&tgId~TSAo6!ehDRn_<<IfnSL<kf}ZOu?tC)nTV2Dg{M^0!X_r!
zFon3Oz(O_UCyX!NHW`MQR<L5jZ(lmbVo?N}({q_0`@S?x=U;r4KTlYvt@<N<z(F&O
zJS>SJig`BPVoxP#b<eqbgtReo;ZR<Spk5#ULvd3{yRojuo=#==EIRpxArU^pgTVii
z`a|%I*JCwM)sDfBepqi$*rBha{e$LjYb@wgM`_aNJAQ}RLA!e11GBx!-$^oXC0zRx
zRy|5CPZ>o}+s=|ZSF{DM^_vPpaxbF8%A{=LvOl7;mRFXOiW?R44XG0i&=YH!rSdaK
zG0+fXhO~90Ddut&e^KL*<O{7G198lr@w;v?`(yc)k=hD>bHEf9z}INo_0+}?A4t#2
z1)9U(_vkh^Jwbf+62lHu_BrUkqDCZR$!<PAK7Mv>rtM6at;J=+MF`4TW7!usEsEx{
zInA7s5WC*Izu8Q^>-N6vmd#go-e7(|ni<lW71O;uvi-rUsB8noh^ga_ok+~M5ANKk
zOpI4Lm6ly{OVi_?8VRy%3#&wddJWVu*PRClVi9WuPxG~Ccs(nxR*5NNatDuh#vn*&
zm|wCA^L#o_qvG}Ayu&n!Bk?BQTM{E7JdyRYn)&!O(%qfUWFPBqQFGP71w;@rCN}O=
z`!qtjPy7+{XcsN$@{HAms%yd3dPH%gP8*sVAMbCylf3(DbZ%N>k5iQ=P}+7;ZFMo)
z$4=%uzDkH8b5cLSCe6|i4QcZr=eE`wXX-)<&QY~S5j*)7lCU(R1#`3=<=Vep7W9VQ
z*!O$RBre!(XcdCfm7A_`Y5H(cpwdlmf|Sa)_NVr@WXTisT)p~)FD!~HlJ3|RuhJ^4
zT^q9$Gw{24?6Wd}r_@<ZF3>~D`%<2eU+nU;7UBK9m^dV#rJel+an{^#&+YxQ4ypM=
zS725vT%QSqs3fn)s~QtCO*AB_sM+cHh^Qr_DTl%7+qFt{Z!}m7Z_agx)q1Go^Zn;o
z#5KfNGjj#hHU;#3t6VG-+%#Cfr6bJ@8=0UN*B_fNT&-6{PrXP`GmqM%?UCr(JYP32
z6n6hTxHrI=#76FDLAQO=S`;j@0y|IdteB+~3-52&p#as?0N9eaGcbkuA02<<!^mW_
zx+&MCF&9(Iu~8mqlKG+baHfQ4fH}}Bt8+L_uLicZCReqfjQZQ?ZqSae+^*4f3D;<g
ztIxU3uHjR{9Dh>n?s{=fI`#hf$k_mR1=MuG6faEHd74qQf&gZ1rEneM6hB<QY=CJ^
zC9E6GoS;JS4|bLUY-wPYZLEbu?NP){?&2>wyUxTCoZjwY)!W@%;@*#}ol{f2|HDuI
z$zSNN!}#z)i>a!nHg&Q1vC=*hO2IZai^1B}Fg96M;0c*W{ZMI$RcCt5tFb5Raf5B7
zA@!n7m5EnfZ|Vp++v&xSTUkY`wtQAz@kTIVWiejQk#9Bc!qB$21cp`_0&}1yGlK?N
zq-8vFNLbx+bxuWPozoi_hY`D8r^Gx)-M)$Mt}IEORh_xu{?BB|QAzB(tb@2O%Bq0W
zW5HmF#$829GGq)pC<PJa-<XG(gycgh77_mTdV)^f#GKdbuESA)N1`mGveKg=7dEKl
zz4@So3q}Vx?CGDuQW|W3aHkXSjc~~$BA-|oe8RuzOyx^$ELU=r+-*v3Xq_(mR57j`
zw%ai31gOnS3F?ex)xuG@<h;{bj0*cMJ$wE<iA96$&<AI_NNwx-p&FZ3?=nA^&w>HV
zDc4?gWo35Mckl7+^8vW99;1PJlK(wX1)7rjdXM{gPa~htjg8C6Im6>`*2&SUR*I(2
zx%nq7ooCNlnl2aF1*?@gEp5+_?4oF9zIHS$bs8%X@`jYJB1?#cENhFrJn{g^u}bkl
zyqcyoXVaxk(`bhh0-Gh>vxF?<X5G)mbec9>C+?4U?^&dQN1}X(hoJiIZ47ffyu^iW
zD(At>o!e}7_WTPj=f~^Q?68?nG6IhG*}bg@n05oL3d;nK3QKj^n${+bOUUV}vin@m
z3&Hxh){e2wmE%_)tuq@v-a5_%3Y3`t$dRYaibrmi&{`j2H5m@dT*<w1^FI$Z$0arO
zkCx@bg}03^9DRdT=<21I9OD_MTEr!OiC$cH=TGP@|HF$f9)7>L=n%vi#aN+~!2nHO
zs~$t%jZPJmpyq;)pBpOAfr61K&#Aa!tOHhQO&_e*c9AgbQu)pMzyE{8kuQ^Y9JlAM
z_Xj$;r^<$hvR=(j$!6J!v8L}|g&0Bc^}ZRI=6ZRzINj>pZ9DyL>`vPe*%DQd5W}8o
zN_3E#ue%ZwFm+O+wl|fUK!(bD<AW6%AT_{)bb3H^gh&zndYnc*#GP<GDF?*7XU;e$
zcg}KOJCB^O-?M*x9D(Ln>TvFmPSn|8k?q-65<&EuoBQ4^j1D+e)I@XtwQ<0)l2KT!
z*+Ac}p=(T#!=??crqw{q*tbz(<nHN#{AL1(P6mEc^_q&cTN|O^`B22(<W!MW>Mfh#
zQ-M%7)R8z5TjoBt6k--enuB;ryQVDE^xO(-;Bs=pWwIBW6q-uvmqHtO!CfadR^puV
zdsE{|omsn`Pa8OGqDO9!dfI&xO1+`%0$rXW)7MSNY*R~`1XPI2r${{7E;pA!BR6>?
zp1i=xvfjuUX!z%BB{Gy$S62YN<%BB?>VyOkL8;0hw+5{hBQ)$pk;+$R3Fh9Xz{vpK
zpy+FASm`Go%fU_+Z41k8Poq@@tMGQ$I~~^-X&CWZY%DwpQIPaA1UrfiD06r(B%TM?
z?v&o<dTIn;we1vnuONW5y<qqF8P+_=?>YC5uJRJ_g&W?r6^J+PR^JuPxqp&r?7kQ6
zz1u>nF}cULd+`_D0X~{Tk@)}<(G9nrJ7HM7>+xl@hw4d4?TNgL+MqymlUud1i>J1m
zq@1_oOtRqJR?~C2Dknw<F?QdQH(icVD}}r|E>9~xMT$M%REI0uYY(ZsBASq$5TuDd
zUpW&tPZzRmv3&+T@0{7%yIyEoj|y4vn5ttUHa%gn6}RjQtvivn8wIGQ?sVhcrIG8j
zaCvL1=f!~*K8o>;bqI99<LQU}KJVw6Pya*D9_HH}O|t=V0*MNY#Sr%$e?#2~v+OGN
zY)A_KooFQ@*;c{d#FlQ=_=PXfZM)NSvhO<B9ldpo`^0~5I-Ges5yX>+d9i*HG)@ry
zD_$9|W2J#<TxzF(VYhy5Tqoivd%BDdDo!0x6-~-ji(1TK30{rVh-SXt%dly#ft+97
zJ-bn&V8H}t6~yum@o`e{@MwHTtDl%_8D}IxI}$$8!$amD83xevS?zKRI_K^-1TJ7e
zL&dl5FPqc^Y=$~YG%<IKSKbf(5*Ol}v+`EitGg&mixM`^W;HXba;VT-FeI&fiz~uQ
z2tr<+Xo%Iak(9gDEUTFAdLs(mNHwp1Ri6<*K+4DEW6k=JjO7*C>UzlEC<*W?5xC-d
zyLx?lI?T_nk5sj=m<0o&!x67TqG!~dD$0&!y`4t<ig^!EH*^lanHNjW2+s{gxnwmA
z!<M8k>iVTTK_FbmjhQ@5$F6wX6zSph0Ij2a6%=i`Kj+ek9kxOw$E}#HQpc=AWZnBk
z8l3Z&3UUgnfpr}wzXjdjr8cS31uad?yMt0j%Nk;}-CTq3NUDEMfyh=<ZcY9WKd|io
zP;E(F7CgS2KL%qpzX>}V@{A>BcxetyCC61jow|%74aYtCcRQ)KAm+g6)4A8Bnrb?7
z#SW*m`!Dovv22G#US&E%hiF6QqZ@1Bt-)*TY6Sed8uhqyXCIYh=iOs(Id6i}s;xg<
z#T0AV?7PO^x2<)|L9q*S_HL<?U2J<!gXdaD8o?|0U=5df+I?mwCaXe0>V4xA1i&M>
z$G1WF`+PisbM-Ok(q@=dRu2|LHlX?aZu@)$ww-6?8Uqmj@TreVRoBRMqCI`owdl?Q
z8d^41s^M{l4=#%0-1!Y^$e2d3+3T&r;FZqZ&w0jIuXj=b7ptA{xYLLi@=9rpQIJ-v
zXI@@$&cT8+(k&|hXG5CsW^sCF{(QaaZtTUq*@mVVfzX&#@o~MWJM^emb5#>E_@2kF
zNCPHTvXOGFtgL*z@X-u|OB&G05%^5rN1cqvQq|U;zi$3E<w6(K)uWJLq#<E6a-LyT
zY_n7~T6Q>i?$Kr!g_`gGyUDxUfQghI*GUt&DY<e!-Bz`hW&O2$MR7M_oW#mI#*!f>
z`d0-LLxcpiwMM_0KLGPvEZyED-L~#n7vsrQS-G6|%j1B1A&xZWzfpKkN8W%kR-z^N
zDps*>@O=fqkB5i8^~MMSn{Bc!*;_rM4e;t2Z-giRrFKu@vo#j|J_Zb9SY^5}5bC@}
zZptV!R+TH3`$d0-x`J94TdA}{NqYC^n#uiu*}6g6EMBnS4=%3gT-nsn7MI>Pq*WM|
zp6LP^L$VvAR%R6#S!*<)o5J-Y(Kv1y_Sv`E#GVN1z?;ki#0_c64TonXWeaG>IneP;
z=t`>VdCk@x^5V9!1~4Z`Dyd4u-&WD$g-=#iR>SqV(u>@w&k}D;bh56*zpGUus_iFg
zT;3)U+aJ`NmYoc6UqE%Pf=9XB;~0$l&-ysN(B=Io_|24(s5PU4vIhHMth!v1m3tJ}
z#DE8;XQDKZAug{?YSbe+tJjy}oH}&B1B?KbSpw)Mgrmt^_T+pD?7v3je$=sd2x{lI
zN(Wdw^|HO$-@e`xA58cZudDcNhmol-Zy4!`fuI@CtxK6M$8#o+JWCsU<pA)C&+=8&
zWK6Njru_!>bJ+?LxGd<*#0XuVnf2KWpUt<KgH7B_UNF<s%;cS|L-(r+Vsm9{FPz!q
zWtnONKU{?}-3LIKxi0GIi!MWRt>p&+n!@COE+$J!Dm)iWqeHxOEv5k9iAA<b5)vl!
zcv2F{ZI;`+*NeR!+nA)9!JT$S3S}d*(ooZ=5nfUjnwmZ$?@^Hzx%z~~J^*Hq5``1(
z5+UPzj9jk|de=H@q*k)mxU4=ym{lB|fd={~Xx?8USsRot*rjHUqAqS0uIJgUT}L`n
z>u(gs_R3SI!&CeC_t{e<7vx9h-1oHrWe+W8*_d?#9EX|xANFm3PpnxA8RhvmF_OZm
z4+UAdMWAiMXs}U6WyS8L-cpDfbD)xEo8q3wf`#;}3vbUEVx9ZWD)nF<&0fA~-&F_N
zk@J=L)=wLK+TRGj=jlm8`^!<^^k`!Qojex;I_bj=pjQ?s&E08FK|gV9_DcobcQ7_d
z7<Wwcb;^n%W>_Wo3Gn{fyR3~jtjWJSNDCm1-x7Sv<OKl&4<<XVwv~U+@d;c|^N>MT
z`|*m0$Ln`?Yl2Z`(+TnG3s`4In+iVxBcCdGXaY*Qt{w0?x7SaSL$)nm7z({%yQE$*
zbf$Ao`gQ31z0CIb!koLv$lYkK{7kN@xT4ZRTD@Oe_HLw-qoK92Ew@f*oW8Ia5h&{~
zic1+(4$Db(1q+hH_h$n-HZ{7g`8XMDYydkN531ZxO4UXSktUC|hGW|db}t$IYu6KY
z?k<zP9okR@aoPKj+ora5kW0iQ8u_J$sTZu@wbZqouXJl0mCkZD>^i~x=kqU=Vu|{T
zT{N^OIO7>GoSZABCvG=IgG#&g&%%vg?bBMR_5&-YKEE{}gHHgcDW`op&LfH`zT~@}
zJ<+$b!=e?{qWfg-vI(=T670v-Z);!a)?X*)jRB{lRoxNj>DT;VAc&0y*J)=ao|%p*
z&B>H!<IEd4?%Qfrl_jc2uze6S|LNUdaeH%puD2k%({O*xe0Q9Vtm6&1B|+xcA)O-9
zrsOboy<3$~Hh)3a<5fj3znWY|ruTmuI!IZ7i#h+Dl!dDga0d4J4&HUj@MY&hHVDDd
znR8Z<NG`%HSiIk$X~SAIj(uXeXSB%5YvlfD8=k=Ou&lBb|H>Wkw_KI#k;;|ll|a+~
zwVchFaBGf4CjkEG(h}Bi;g^OzgWBkeZ>B{+jli8-E!3K5zlAB06-Q)9Kw?&5W7~q~
z1wWPK%<(;%)E+99o$Ox5QtKl>PIw8Z`Rz6O&q7pOIZZtwiFS4mX-iy4uL@WiPK9UM
zvTE)qb^gMK?X_y7futPtD!3De*Imq3XKORvNTkJ%kM2Xi*bvT~Mlet4fQn8{L^M%R
zVV1(}>y7Q`?9{tuu*O4laBkHr1UGB~4a@PGO49sFMYxWY^8FuAV@-UVzXJ>J7YmSy
zc+d6QCZH6GA?^s#-KbE<o6i+=46m%v!IrS4CfLjaZt}-!gA1jg-7=$Lprj3xJTc0~
zU~BX2Nt?PuuM-v8u`m8BS8``kj=w`3s!MkEYfAOk4ShVT1Us*8!0Ay8dj<DMcCB!L
z;e&gX_b3Kq#l_Lz3Kz528g%pYd=}=$7+|IwnR%9rV1<y%S4x~o({Ua^j3zqPdwV_6
z`iB?7pmdIY)nTh7yQSb<frfNZjTyvM@1mk?M&}TEP?ZjdDq5vak8`$~r?xiavZ9bc
z=qE$|d7IS|gY(ck9lU-?orVQD?=LKt7abs-q@m)su?;c6`K}rsS=D}=7M4#=id!vD
zt$8&%nAAzciUSh^%{jVgz><2pO6eoj^H}Fns}eG9VzDi}xw{Vngz;RGjx7?nx^*p^
zLe`RU(ooxRtoksb3+nDA4_nLd5eHOPNx6WDdcEQ@7KX_X4U`nd_BmKbhsYfsM%;5i
zD5fqiHJ?2mr0)pY++KXN-P-#BrFraHg_d!Uonn(o5=Ziiz-IyF85`b5P@Ry3A1}9F
zo=mST_!$`IWN)lYwlbv;BO5px4409G=6m;J0FPa+sR!rbhp%+w19~b)tHHMKre}#&
zdA+{?`0nqP4zSzES&9HeskS@)_)kP6&dMhyCdL!}PDa?)RB!a6f>A`8CLXl^l@V3j
zDQ@@cT=IPjy4R)u33K&w@W4-Y`uu@)xeMQ1hHhXhZcZ=#H9VB4TqL8w8dghN!bTm%
z&)OO)C38+`;ydX<`H5hIm`tfubMqHyeB5&j5=B-s0FYSMiS?uY<BH8l0IywxhWU%4
zGmY9rIw%+{v_G?#Pw3=&HmP&{GA`QxdfR91wZUjzd-?3Z{l33OMi1(dl@|{~nCw0u
zMb&<V4$6;ZxUn)H(<>Yv$}ePVJHA<NF~^2xRmI;kl4-Fqy=z!w7Y@tQfDw<fwFx+E
z9zw3H?8{ln0Y4TyA}B=oA7S9>ERJ<_0A;2lR8lxXT9C!+#<w27^YLZWlHx5>8F{@T
zpWm3S4OrK%#WvM8ETh^$4i1fu{X2S>6|yo!V!C^yLd{<n`*x7YRk>_1QkJ`Du=RtY
zG?z=0^qAr~wPSa+P;eZtr`XmzJN*{Y)%1i+&GpuIwK0eVZS`-85&1E>3;-Kx5(@VD
zJ5~WOFwqA(e!RId^`w|35~?CmC$6#^qcvEiQR@x?0g=nLw(1Gxu$#E4gt9o->e<>M
z{WA|9&V}Ndl3fI<=w|nm4`W9k=kUl&HKRKD0(sjU&l*WLSjWUSdPqCl8pfUW=F98r
z`PdiOnc#g;fuASj0&ckdjTn=y4{=pux-I9aU30-*3Gt5AFE-UK8bz~lseJLI`vEFm
zYv#P_mcNabGR=QWyKTDtc(n%Bve&yXY-1h4LF!%!D^7;ieme7t%*;TBdh0|^O9p^f
zF4jT9vrL}q?w6By&%U+jB72+GRH{9=B4;#z&i+vi#89%5ii!#VVy530We~tbMyIs9
zcXy)jM-%j5EpAaK7R<kI)|frlfH+vgsbJ%Ef1Uc4Sf71jP#W9Fu|M}%2ioLvhQFV~
zn$wxoSzg+-y0uqJMes(milgEvRjKCegroUrLk|#u9tt%{^|bnKBC7z;@eo{Q9p$EZ
zJfhhz`OYuBvI=8FzGd$vIwGfyNUaaq=-WFH0Zscg)RY^~CY<Z<8xHRO1PG=(oPyJN
ztURfK7}L7HQ1M>D!B?%xf&zKY=u~u3PyUo}@Twf)UE?n^V<*naGJB&P$2O;iu?1?g
zkvJ3H>^i5W(_v`h)$8R|L^goPS+xsOK2l6ZYMax|(pGo_uTg0gkmOtKDZN?)rz=yu
zF=l|)O>s}unCC2W2~{<Hu5;1Yrbin|?u3G!LX7<oI8_(6)YYrCCj~T2>wNUM(~U2C
zYH85&k&8g4HGT`IRW>{4AY4!^WD*oLGnEJEfyP<4oz5fkV26{Gl&o_)=oa<w1Id}V
ztARJFTU<O<E^hbx^J~`vI%cel24a}I8eHYLLk$>hRW<%F&qL<kb53Mz24l8X$2KO#
zPP4gckaTo(4+kXrK-C7bjxGdBB06T+7c2x)s2`BgjX)FmQS54<N$*9ZtPFZ$%G_Be
z$(J`}0eMz_a?INto-K)c^o%!XEBgz#dkb?wPr4}%qqWxVNq<vLHBj=dd`cotYRV_p
z!ZM?6({IxidqzW-+ZH#y>)L)*%I9TTTbr^#`%blcUz-o`GLl~-uPZJ$L;vOxWF-jj
z;SHDiRaKwy?#7hW_pe$Hs#iO#r^trecfsjiN2yI7K(em!mJw(i*-vcLFzg0B?!mR5
z*q=@|q`UA9K8pUz_vQ{k=ql`&1;#{(qSic%Mcfu6!%kG8^UeG1qQw3mlwwuQv0)mM
zRu7zbFUY>>QP2o*k~uQs(Y@uEx@z2kwZm{0UC1a<M)TA{<{71(2kS^Gcp^KN7@rkV
zRi3d1wRQIF+TG2m)bI`IgwZX-77&G3JX$J<=%a@irz&TDemnGFmWy`Kk<<r0I0>@2
zd>nZgac+RPe|rmSl0Axh3d_ag-X~aUoq|9L85YP^D9dq`ryVmlZU*woPf&wMsyQ7&
z9K?t9$_yZldZx}^N03nwp5FeW@bu-58Tr^<y|M}QNQTc5wUif@ba{Ci+875l8;xsg
zYvpG+i+qSus=&c&{e|)mVIyN1ffEZSDfDVf`tF;6ulQmtSRFSq-#>XjF9uZ_+A~b%
z7g7bi%0Ba7%PLNY?yK6tl>?#0SZEp7q1p#+h{f$}o}hAk7;u&Xc?m-9o29&yt6cyC
z4@3JWn1M)EBBX0-YU-uutcNivuQ+og4W++k(;)qREV;f^aUu$9?%>K^(*#nV-q*Yf
z`KtAd1Rz|V1Rxe=S!H(Qq}Np;&sspcuSBCrzf-iG-^Bc>iTG#bz__{pS!>!yLK2i0
zh`ql6qlzNH#)SVjYQA4Y`HyGg=}}Z@^sO13ZcL*tNs2>40NyF+s2vLI9vgmIT)Ta*
zby<5d?=eJrvtFdUL}y0dKsJuz9{LM#+`|d8d9CS*Jx4?Li$KN~gp%u78+=fvd1Rpb
ztW45UK<^=95-PBV<!Qa#k6gxBZt-x5(zH4LKZlgETkr1hZW6m(BiQZsOz<w&8z#D^
zc)lwx!ren9N65ZDK;vRgQbogS?d)KSZ%W=hn;kE2ba!ZlYdx^6gFK;0@&1n%+(-6*
zbmR$Kal`rcvVL*Xpy82|fP<8IRQH${k>hpW!clM<kbDr;vgMNREp`aK5!Fc{%@g&U
zKCFQ#pA!n<T^uvtci>R_Wm<cG3`-CG2t6uq(ht0<V3`;varY2h;Dv`o0diD!OPKu^
z1gs(qi?7SuOdhiE)=ZyK6be<-i$ap9tmC<)ke6(nu96+2d+Bk#2*KucjcgX@Hr4J<
zKMQ~KbVIkpf(AvNDOG7T<pCs(edNt7jwb6a2#q-9=EOi4o<j-HMFki)_diQ(4n+j|
z!nBleDZnSmlcPqI$FzOn=pFwVbHeKt0tV`ECwQ4%uR)E6B65^CH@P#vlbYs+Py|no
z`DW(MC>mN0-~iKLo@gAhA5LfpvJScO5Zd|!S`0Ws0)+Wz0WOyNVR@f4*+d_WYa?7&
z1gUj((}bo5y~erbksLt+8jWMDOs^H)HfPgWvY^W?;eu`V%t5uim_Py}uXUr*XkzS}
z6_-ZXuCo$rLof2)O{=x3%@#{dU!^p3($)R?PR{ZhFU{)wbM$zPM`9Bf`H;~CTGT|R
zi*1#3=kMmuKutiB)LMjK()Qu6ur@-gn{FU=p&eYroA=f%xH5M}7Gz9M+@Sx|=?-o<
z!LaxscS6ETCl`RU0(VKdx&~MmiOI$uk|!^|@f#pCmvdZdBln153JfLp&DciP-w0)T
zWzH9l*Gwl3cBdKj*IWG-a7v3V|6CZ%pmxqU@xsF1S6#2U$Ay%?!gXj9XkXr=yGyY*
z#l{4v03zkL$8&7K{E!PH04L1=1q-XSRzdOGesyLzqIvRJ8v(Xe-aEzCvD<_sZbGYW
za$b5)(fi#8$i`7XBabjwYExGpJG?WQ6uz2<Mz+~K2oRQ?2S+3)KJEmUrH={-RA&xN
zi|OrIrEbal63w(nVj%C#1LQKV{RzE1hWxBASJzUp+Bg{68^(gx%&0G|CsnRITaeSK
z-K}!IEjZuL14Men3l8OLGYTSx-9YolWvqYEI0b;l?YbX2*P7O8_w}n9*%tgbMU9?a
zw=67>R%<Yu^omTnNH8y|TkW%kjR?XnDiZ^O(x&r%%pKKb<+g1$&;XIkpW(Q;jPphF
z_PoTO|6{@&1f+{z`aDosPijSc?0*1Amk{Z}JtywV4U~fl9N(&H%q+T9n^!4~ztc(m
zYsh=I6GqNC5A#{`pPjuGKxgV>1IVjNr|(OUc_F<8AfxPdlrm%LKH?NX>~2fw$y%)h
zuk9WdnP~kO@_z!CLw`67WuCDzB~!=y6)Zlttk70LTs%Z(tC@7jr=~o&r_#iNzN~$Q
z4zD-L*qc2sKS~y8&PijPFZ#J+%NZ!0sIGRY9Xu7&9e;?s>;>62{7Z-EN*uOFTZ`&M
zZZzdS`3_gRymWqN!u#fqmY8L#&)kxCr%b`cE0f@Sz{@Fm^p{hY_0as;hrURinu*3I
zpt<O(HA0h=1NXzU7G0K@fv}Jdi@iYRGF44SRP6Bt3c!79Sp>=o`b1M;b{&~kLtNYa
zA$Aa5)$koZfKVTgsd(SEP9$G!$?A|3pn$w44lWs|3BfA`4OOyEH%xA}59VSxw=M&J
z0Fd`N49zA7@e^Rvu@D0uSAv)XZa#T|4(@9#N*-m@pCAk6O`t>52;N4pz3a1w2tk0?
zU$sugIkj#!?)L_q`yJJT-!R_vLkfUwWE*-NWF-+&PpmFdyekGY9CC%N#mY)X_apXJ
z19%cD_@0UMao4Zl0bf$nqB`0CSGTclpIu^5<XDAgsg{%1Qq>n_<x7}!)vWDCcrLHt
zdHBMX|IE>}=ocfJH~9;d?gPISa<UJWY(x-~lpgF#Pha<=U4aBW;0F-QlWrT5qj3x&
z<A@(3zQ~%ZZ_lapTWUSf2Ks0;YRw4D=vE)PJ9@cK0aXvz7Oq8w{D9*!vnphth^PqU
z;8d;;>HY!SCvYps0Ihz?jOaMhEt=T}1+Qp?3vTZmIe4@kTTIA8y)b^)I^LoRnz{hd
zQ9kds^rK&VVzq49#*@tTHeWp6xlxo<KDhJaC#zPEQ_YrI987ELo6t3Bn_$_m92?tf
zjP72tWV|+Yd|*@Zhx7us-a6j%u+-~WP#!OJNat<tp)awZB5GZ&JBFid5*N3T>{qCe
zF6%?aK_S)t(Hj8*&w%m$Bw3gLY{GEqqEi~;|HH<!eMWwCDgDUz`^Lg$XBD&8iU{fb
zwX9-lt=mwwUY)+Ht5=e{cb&75kr6M@PXg0t9Z-+H!Y_~IHeZ|L;BsUb52hPvNVC_?
zQAU_%kQnN0RUTqNN^?bld){}&oK>&b-%MfEXW&cYh5<Qh@0UOY$<gXTS7sU^oQ67J
z%p?W1t<WAdApxEQND)r#8xA?Ht?oh83Gm##a#hZ8<lMY$fh#%Sy`bk~er)uKUks-A
zfhb>gP`tU2ZZIIuoJ~C&x<}pB&|aL{NICeDPhTN2nt+?Wf1_EBFk`Ed!b5VAa4tQ7
zBisN^*Y@^|`pg(wt3OojAdVAYIy=9^%m8F=U^$(r6f6)xPbVE59<9-PT_T_ly4#eK
zvVBm_Tg#lAIY+fYLZ@p9F>4?i&gvUgvuxx!a@7ReL2o!)4w(y6k)IQ}Ivo-tB}I9@
zf-(5Q#}bHaHCU3ca|ZuI+FAZ`P#W$&5VY_V>HLV63brc`2q(SL8}<0$0s%NcapTp_
z0`MpiF=ym<B2iufY}bB37L^9%+k6V(g6u*0QqQ9nK}wo+StfxTjo!*_(a8^+R~L*k
zq|C<WzN#ZYxetVW1bz8t5r<tH14xG*G2VQsW{0Py1>y-pQvM(QL6`TsY(6w8H@BY%
zt#W?YIDtyTjW6}8C_eOizmg2;oKYW@20L-W4pIR106g$%>0tsNO9o(JVWa|8;g1VL
zfD1q;9ZE1Fko~AX=a8w@z8)2)CX=g-Czzjr|Gnd&;)HH06Z2=d*J2P&&XZ8d_qyN*
zG~F__E&SI$$}e69VHc4~kXtS3buj4z)p5j}Y>ZVllNIlR-Hqr*q-P9hN*G~tE7|=O
zz=p-+8hiZXcRq#RMvzJ7VQ<>m4|b`}%B|$}f$7Z-^|lJ7SE6kAr3Om~1|CzMt~)a_
zF^w%eW5jq*M)UqpGG!XvKMMK8rw;=U<@q&<$0y;YsSoeS`OGSGXY4%N?8-c2!b!r*
zLgMx;@Roo;ydF(S-zAStKO;as8$?K`s(k(tMF~5+*Zt@S09+InB#xlBT2)DvrobCC
z)BFDX%^f8;2z@pdoSz+cVBeT6P?h%KeEI+ndY-|xKjfbwzP>MQ2E4Dc%I{`+<w#5N
zHw6wAEC3KQ3rcF5sX5^}Uw1{M(l!d1VZQ%W`h?us;0?h)MTm%GHV}B3Brbf>27lK-
zCwlf#ZG8I^)Z~BzV!pp{-Q3gDGgYh7|E*aW7x>v8*;w~~#o4rc07uBL_Pd$iTl=z*
z$?5jv+!<!c589JJb%d<oZSQWn?HZy~A|Uz-e$1j5{}P;%Pm68|&=<n5)K<+1qdh^-
zND#HPx{v^{Ph}foKq=SDD<HQl1b{Y@{<~NI!q?~r|2fFtI&Ve0i6sf}#WggDv?qW4
zr}huP^VQ-MOw7so2N(1jW`FUhEahZ2hl|K#L4%5I=dINYNmg`uR^V(V>-^5>CT}xb
zylPQ+V_5)m&QtJJ{^d5KB51p98I3q0e={3EcN}55q1Bj@Llq1xV`mhpoO@`KH0_@<
z`wq@T4Dh&jx$JTOrC>5Di(@J}sOd$ofQl?ktjxa3N|yZ~Os9jI3m|Bm<F0lp_Qb~y
zSfd@Gz+*kvd)*#*aBQse-h6atFYy?nJ!Umx4*g48YwfMgYjnQ1Mx0%1fIAN$dDY^_
zHSqtQZ=et?oDkf{T8s!8iC~!FbqeiX)wtwsEtLKN?87U56sE^o&L=+pa4%9^%Jc~A
z>ELuG9WRE5<=e^!Q69=)faS0~-FiH-Uz^tv`MeKx+fb`Ua@_$dcs03_XkVCw<Foe$
zP-?AIzKgJbG?L{St}DDDH!=K_r0#IAB&jVRqW6y29`GyUfJub%voL}Fr3?klLD)1z
z%nT^yB1Xkchk!2z0U5z#aPb1WXwns5{>YfgY_j3OVDR#MJz_~#>qm6KcX%lMqc!-z
zXb+5?-t`jWU*oD_0F`ky?gI_-C58CsEg-bM);0SuL<==Q<c!x}YTo{{S3Z`nBdtiN
z)Aw^{9+EK-tVLtYxDUX8qw}EE#QK-|8NA#TJZ#t0m;uBrMNO&@X@w}I2xv`u5{myn
ztN+?;jC^S4UY`17Lrhl-`0JeSzsLAXC$Ry-o<c_b@t=aT|7vQ<Lkh@B2<dSde|_^o
zdcbpx=(NY}{?9j6Ai$_y_3N>`>OUn9&G3LYXac)GKOTVpd@BGbOR>i1e*yYm(5~wX
z$QrxdkoF&u{C5K5z&I0(wEsKJ|J%|0>j!p#advQj{_iy79^wh`Z}DIL@46fTAv+>V
z{O>fSJ_BO&jYIK!6zX5!nrne^Y7t)l?==4}hU!1-`v0ByZ@e)bd5&HhSsB_p*cs_r
zK76y*H+#v-%Faf{LiX^LkB?c($jZdw3)wqvc5br2{xOT1S~wWlF^gL0IT#5W8CV+{
z2?)IW?+`B2W14f}GqsqF3ps<x3~l$Mx}~{$1<y{II_7a2xDXRgDd^l}n6Ivg<v*Hn
zf%mjRWNdH;)zfn$R=eVtY(9z0EKqZMx=l6-cyJN$B48%JbaT7qp{pieYKn^7oO|^X
zYd&}Dtn2iGhORQe06idP7>^4sN$k>bAVO5+PZ}1{Htek`f{JKM+J=tj_m!_3N3gdG
z%v@_zhu;IOygnTC5J#IaAq4~^jT6K|eLQ+LR$!#X-ky<@RCG}TIj{hohI)aUyJkY5
zhqHLs7BnKjqmAIT)M&JV{2#n*!8>kG&&yst4Lz(BOs4y9<${gU3~SbZY@?b>Pv>jG
z4QMHOcwKXSyQ_@Di4V7a8R1s@ntiI<QNLWm8+(&*^x3{+=(G#J^w7jOlT-29ELrQ-
z^FPh<5?X2SFFLq3`CovT>I>(UeZO;u=a!zKQT2)n3p*f388XiE>IQKalTHW_erMM+
zymo6JZVWcYWlPg4{M}aTO!xUBy4H*)NdiaB7wqyo)tI<P6Q_T+H=cjJPVGYT1Mb{x
zqum5~ovq#pajRF1l98Xr_s)?bciuFmL0*;sbqb3jnOhaJsjSa;1vqct5;3afFZgVr
ze6lh8+VL5mP1&F5&3#$x{1B`=$Pks@V*h%$uC4|oIk0}>NsjmW<9=ys`DfD3i*h!&
zmjq)&D3&ze$~jM+um)F%EET+8EF~`mW@QjNv26Z?J6Ec%$?)M!lV8kYP}e>kU%?Ez
zt7WHGTg&+LKH=a^7^RHxEB2~Iw5X*hT#|ODP(v0r8tclZD1k@_DmL1?sM%fgKMWT?
zNfgHtF1oJC{>cbF^n$hw25v>tay6pBwHv-7=k7-n;dny>cYN1^yebR7r~I5*^7$R#
zSD#<C+gNG^MjTd!^RL+#P!-wBzQ(2z`thMdrtmm$w%!iEYgX}7=TcS7)$|J<mB)HB
z)VQqdNL2iZDQHnFEQ>$kN0&4;c+B!^kWnD?GDs{L#ZaM+bf7YfqK75e+#eF#e>GFH
zKO|~1N6BGzq2dwyl#ehjdWtJ}g9Mct++rQMMPdm*z^{LTjV_#xezBqHj%h&9;TmAP
z+9$)EKmMdRiJU{CJOKC5E#N35kit8z54PJ8mZRPBp05{YyE(Ke{u0du{(xBb$F?xl
zwTwJt;U}T6WpLV*WrjfnM{Cq+pxHD1Gf8~?Jvy;h9R*V6kc%W~<*%fCxf@*S``pg<
z18*yy@^@kIHf2wH2Rz?K(-fm@hT(Ovf3#1m_?{NFSSz+$J)h={eb8n<_<@=;*u;Mw
z((Mqz;k!X%`;~n;uyO+nt<QGo=Q~ZPhd?7JjbrddhTD&{HSKy&+~z^iQ~iiZ+x@ti
z3Yz)|bb@E>AN=IXixn!QmGi_1Tihd5U5b?7q-EvZn1mR0omfT<nvF8-zOsz2+#V7=
zoTCgcQAYV8!-kf5VOFotX)ZtVi}y4){6n3beLXM+IA4r#D1%>rqfY&14qFmJgWbTZ
z#)8wH>q^q~(sJqW`99q^;yT?tqV)D*E(!ikWjn?5@d#Uci)ktqj%P`o+2%x7B@D%<
zA^8=tmYep+Yui3i#_2f95mD^n;4W8b@YDh+if*7=T_{UC<PdD7WAp6yNRO={WYLN0
z7T-pMYP9l6V>31SnOfEievx(Vk@+)fJiXgom}rD?#XJ*&=WTx>WQ@ePu%&c3eSX5I
zKZ)t$pcuQzzf@jT3aFeen{_1K#icX&6!Mce2w_Sqx+2`~2ny*m|07_wKazpJRVAsY
z(`1g-p>9tTv@KafE$6w<h}99v6SU0%fAh}NPJ)ZqOtfpThhSIKbQ@DFRGwd-Fl#yP
zN1m9DTYaI}gtTtT>CF<Oh7eP@<_e`z=ZAAHv;C=z_+4D9n9xSAiCq-)zT{mgou21S
zbyF1nUj|f}4kL(6T%*_YKZUGGd})y3D37^)<3CyHS#^;So-9@?xniQqmHd4o!+4fI
zLGD>66c}YSu5w+*ZZ`H^F@`p{{w|0ysw<L5XTg(-@GE6Zc`Cbh9_>xZQk3u3iftW(
zV`Uw+gr~wjmADr91W_^52`>J&&yCT?OdH;KX_~|_xTz7?F>4qYg(6ojWSw8#>sGh?
z*V$PewP^vnGgf8ePxG?zjd=<S6w@CZVpRKUhUTlqcqop?6xDK)S15&Fpd2;&fzM_=
z{U;fh<NEdmKGN)asXl+BZ^Y*s(wStjp+$6CxFk<p5ZGDtT@u?ham$~DPM4V3qE;l7
z=#-HIyUpH_wB9m2J?#LiyOL)ttP3qRs?hQ`=E=BaTOE&Ik8>CmkFnf<{$9`(HMS#u
zPoh2ULQ3?-5U;f`N=EBfqQd9j+V<QgD8mQqNJ<&;cdK?ZBW`G3J~z)1k<X|X^B?zR
zB+UFJMnsOluV3{9uS)Vm#2pP9<O{M#ja&<u#1aW&5eaOK$_@W~!u6$l!Oy2HxZ!xr
zGOlX-s3cTkS4u5l!W$G+tj1;IRS;B**g=apZ-Pd{Qun+9e-FXn_}KRT`A^Rl-rD&r
zGYxr%jbLUgRjMAZ7e-+5KU`+Vor6iWj=bE@2uf0u$iW!3YEzISuNwKch--bYH7_xX
zp0z4^El*N=+1qAw{dC!m@0+VP>gx1M?dSsczOz~og4{(va0UmueFZ0Lct+^Q;*ffa
z3nmRBk;s&ZWnuMiK4Lg`^nqGE!7)aG!#`^LboBYhoFE*FM8dU_I#tc1u@lk7>Mr!5
z5LgColWgR{h~m4plV3tsR||?!X+4LrgiDrB;PsLs1gdtH+PoV*Z=taZKfiX}C6@Pe
z``*zGHj$nfIk3DyC3~@O#PjmC{h>faXXgx^k&uMz1-8H>R-+WU^0&PBq=qc7f#2ej
zUn;55V-IZM|4hCl6R$+WFU=Z1ZX9V>&(8`R&Q=cdh><&-@<1Ax$T>C=*mUfBDSK<~
zcFDu<A*!5wA@ClISJ{o*74T(`AYVciCp5ixoGP{M?f1`W;!oR`q~2IsOOC?T4)-T6
z!EyaCHp{Byn0Ph3p8&6llSVzJlj7z7VqE!?8cK}~Qm7AT9(+(lYv^q3!hL`IEcN=i
ztS3biYk$V@a}Nr18eBGt%}^zQDmd-FX&Orw_daO>Qw2?;hI-1Ru)NYEQ?w_GQ&!#4
zsI1u};&|lB9fi*3U(%(J<WD}8Nlc6v(7ny4W!pIsPT)VesK&U2^<~n(PAa!1OTJD;
zo@!|k=He(C@DRx5c-pHwXj#Isu9i=8Y;`Jm5uYfgBJk`d_lqCuOGF9-HUbO$ZQQeA
zj=J$L+U$Y*A}&0p+84g%T`e<bOUd%ME<U<~16~ZqZ%?_{QdasWsp^sHZ?@Tb_r4r#
z9*F-r_-d|HYAcikFZGt{dnFf+pR?FE)kXoFpo37nV3fuk&$p6X(Fj|cyt6yEY&jd~
zpVTZRHV>yJb2?!0g0Dno*u9<dVAth4awptBtK4XqS`rA9lkEGk&-^(g#T#0Ghve%?
zEvNKxn=R<x1$yk`F9eKyB?D_~5*VE`Hh<cl+u$>?R;=&eQZN2GK-gDCj=kyh-KvnC
z<B!3&JQFnnf6Q^I49G0~*tB-;p~r=<k8s41BEK}2+)?lbMqW`}oipeR8AYJ_$u4Yq
zrS_ASsh8FkYWrThc31X~1L@v@u|xY5L|qb<ZrW;GqZ0ob%Gc?8moxl1Qq<>@+!Rvp
zD7I|E16^Ve69tLkzqP)mE@SK@Sl+Jcz-OH=b^~qZ)Y5)r<RS>llIx9;m5v-Yd=)@s
z7I>Sk&fYJ&ayNyS@s8SxZ@P<{1>yq>`M$ya-QN<l7LO{OOAe<vaIRIj@{Xw2m@*Ov
zS7KWKGuXrOefQZtmp3S)AAK6b?u%@NwZGhoprG3GuQ9G$8s@2O_|m6LTr?qsCF_X{
zUY)y%H+47GBX~Erf(1+j$NyJy&-MQz_pBV8T>qcQeI-%16=mJge(@$LOTaZ8o~Xr?
z*L$mzMUBXhEPtfXScT@H<V~p*YO{G$)9UZYB=w^^rt?4HcEpk*>v*}l+<IIdr=2dx
zakh4^om^8aJ`K3{{g}ZXg6*sJ8a$u_T}a`I#nmo898RJ~V1ALxU!S-k_L}b$<Ei#C
zY{>K}01&ag2glE}U)=1blCGS<(}X$7A`Y!$9>3<71I9fp8&ohiTIL&+eZtYoe}8*Y
zj7}uJ%rJTRwsFY%AQj!4?=##}B4lWl4tl8?eap|-5a7iLJ+R<?jfeN8gwOczOdpQ6
zlO{offV4H|)e=y%o>k%7X8{N=AUhq?&59mBxqE#?p+>9E567Mqvwb%GO8MT#&D4UE
z={!`EjSS(IAoXV{flDDem%w@P=gMxzQZe5v`4nMV(Q)f$o5Gbp?|AxF?InFp&2-|g
zY534CwNA^iCh0v;hvQ0K;&?Gn;VgujyoC4lAbk7H+4WU1eEb&e?_*{pn=i12#4%q{
zRdJITkMN9<ULGO$6P2Wkisgdbv2fG1adj*u3R8cv1V+1=UpGWOguC5D`dt)~o4r}8
zEp&WQB4*Md@&V}j)+&Q=<5|h6BOll~hje+SmDTMMvJoj^A+dC*^f#G$qThoTmfU{Z
zc07q;&l-v#7?IlyW~X{HU&a?bpfD!-nP-|E`ME^TNcz2>^H@>FfVA@G!n7+m{>gL3
zf?@dX0_CwcDqK2NuD=*xYSQ}D8@Z~pl72B^H<;UL7qap2rVL+w4G~1wjFPq+n@5{!
zqE@S3c0EzVrqk>}>Ugc7DYr~E-OrU)ZPZzfL*GxvZ5~tIq%x-YQBqYwQCp#>!iFK6
z-td*61H_D2ih7g+PmIVP98?I$-P=tNko6&XBH4nyb0C61DTn9FD;G~$a%#WC8Jo`d
zlV$0b{U4DRZ`LReO_`Pq98`xs8{A4BA-6lHf=7+dDabv8-?8{`YfvVA1`U*yojgr<
z?%MN668N1bB9s=1m2bqMe&$tkef09DC42PX1h)Ej5wgZUB-F^S?oH`zwJu(^x>4h-
zE^=QOS#bK5R-EG0QJ%iZD?@|P#Sw+$u(x3IoM>NiON+Qxn54vJii*1?e-B<p8FUi=
z^sVl*F~ZSa?wj!BVw}iVQZqOeGVe*XWX|m*KD|aBErNA2>e0GPu+xud&EwhUJRRP_
zftBk%-6StZ^lFG?q0GVXF|*}Lh>6BzH`ZKw_54#}V_w|hAmQ?f(l%O|2}J?7YeOv*
z&#y0-nQ)t_#w1*0+I{c>d7bIX3zsGfg@Mc4L@P#_HP8_B1?&)oPsho=1GL8TZ!t$O
zExE7NuO2^V9zI(p%`}=1$hO{Gh8d!-W2q40aeapK9El1w7^e+}mX?s(SDX7vurPe7
z%ytWiNt?q8)KK>45T|W;m*?`wZS1A`NM2bj{gB9vOdVQz3A3RFe*3)G$P;J2W5mVE
zTIWT4>Kc(*{HH0HPX15b(`C>D6MDFzfl9jDPv0<cq~oc#zwb<W!tw6#`H?xi3xx-2
z@`Y%z0hSs<v3s6-nAV1w396R;UOk0f*jM$%mzMU-uZxP?a3R~R(>o<XN?3CdUgq1w
ztGJ3()pnt2-{-1^@qW!2W(kpz=$M8kUlg&4ep!)c*?UPmjb@KMfJ2W$oNm>_%j@<I
zipt0D=c2l2|5-&=bocO#GE;#*w6yBxa&qk#-P|=lo1e?9Nc23f=c#KFgUnT9t$O__
zEJ11Z6l|=upF>CGKD+gjZ(~GnGoWrE2=tcFqiI;ke9+nIw868D3(Dy6%*TZXm5TEG
z(*NnO_yWsmI1NRe$FsES=Z!w^`IZPf=iNQq)RGDsmx^lDJ=^5aMo{M%3u`^XK9PsT
zJnP`B2c_CL20IP8njK*+BDmt2y{-VABS{JgPuup7h>;7J(Xq9+IP(?CAGk`tna}HV
zNVe_Ue*WG=llmgSc;&4r_rAh2*a{ph_Zd700cM0Whq?xxd$#_QvfWe7tcRDbxsZze
zU^8-FMg7a!EM<j5e^Z+~NJonHO__>{dEQTd?wa(Jh@9Ojxil>rf0u0}QPk4R6xN^9
z6#V;>7O8M=jQw&xFN+t@2J_M<yWMM!>4~qzs|YCT-<H1bS^$p{KFNJD)pHLX?fuTi
z(>|g}o)o5wb0acL+IX|kb<yN;{er7>t;^d$rSS_!G@=(Qrb$%mSev`yh5>c(qc{mT
z*RW@sXgtP;$;iSS2I2d{qI|i0y`Eai&9cXyX)<@ID6uX-@lJ9<`E$@`bNl#HlLb{9
z^Kz=`H*bzKzWW-isNm^{b`%GV@4Iik%($!M^Uxt_q;L+4VRAa+dxnO4XFTgyLZ^v>
z8`hR3{-rI{{@3-N_(d`l3-*gw6QW2+l6;dB@1DD-t)%hyPr!d=Bwoux(#1i<q4UVu
zBTOP2{UABxV@%!7+5I&xw-63LDSlSK``YiBcXP=??FGu_+v#u)C)d%^cKJX=`WR#a
zLB>Dr`&rJm-?L_EGU4jJ*coNVmtvB!tW9Knq{F_Ll)E_d03#YVrzNc;qT+>N4~v+`
z#$Xd>{hriv9Rz7%;{<i!i-Pxd<(17Oegrd%<gJx72hLKA<`;ycUuS0ZNiu1)%iT`7
zF!}WyB5+3`6>h#Wh|G$Xq$YG=fCRt$_@XXd={YhIIyp}CGa@dNC||yK^PY?m1jfou
z<Bq*JH}(dF|HIl_2gUVm+rqd72u|Y`+}+(>g9mqacXxLU8c1+=cY+ffg1fu(_V1i~
zzOP>0TlL-_kFJ95X48AI_L_5!G3H)#eyh!xZ*rJUQA*R1e%c&X*6u9`{#s>*%gQ@L
zX}w6{HVnc>w=%>WBqi|GT!W%@;+#?uxPxvjBMO)5^yE=*TSYEcS)PTBsVSVRRnN=5
zgKJ&xrOjddW0?`;ZpkG8O*g*HA78oTsvYSYsm3V_*8cwI1jV=3;JTDT$Q1SaUu|Wr
zSu|<c&+Tap@_p(_6pah&ed^k>SE{u;1NQm0gu`2T&bZbIQz~wF1WWOpClT5Cwl2r+
z7A)c~o@&Ck98QOZ&Gy@oDXlzF-gl~vh}HL8h;PjeX3_0BDyhD;Zl~F2JeZ$fN@Wil
z)mat|csp$6#_KE*@oVcWXq?2zIIkkD@~E9z1e9~>YV;*#eUKl^?=m6oE0_Arca73q
zYgbp2HDsx7d(mg;&z%E}v$6B>SGI1LvQOAl=2S)@+GkZrl}nnHmwl|}X^rupNlo(_
z8S1aj*JDg`#>(Lr6Y~CKw5six!U0X{DLhSxoEWhAh89BZ*4(yLSGA;az<0(@>bDhh
zr(#kv4eMtyV-RZrvl0B^3!(Zrp?ol{Vb;Nh)v<V(+o?+*UOkPB(f4cx6Gz?!=~$tA
z%<{QA4wYGNFxa)9jP7Zrg7*$VY(b*xC|^d+YT<ZQHnhXfUwuEalp*+dcV0YJ*i?$x
zgMgE@JO}YkxHghF-s|kZwtVroA$gLuQIo<(7C3wAsfeC4IL@#5{Rmg(;WW6h6~#p@
zOY*!6bUqnVe*Wy%;INwl>RxryA$mIPS2|SD=Gik)hCt;6a_m|)S858hW*CB`8GoWj
zWRs|9GNUWs6HDBYkh!QNMHXv9!J_K8218Yg-Oi|dMk9LFR!LjsWQ7sUBKHXciDkz*
zdeQ=?C1M}`xA-RGpBYydTrO&HzD!CdUyMtcSq@$7zbg(>j6zg2!4tV|-%xwo?~qP-
z${|zVEf07VNf`Z|UT3qG{gon8k{vRXv-R5GXiay_DF}}!wLUKN%OcWf$U$V)8yihe
zB1-bJYUkw1VNrt%+pC6MLg<!EnjTBZs$cHeu7!@cUN4vaC=^NLWrO27xh3Rbj*8-X
zEaRf!%z<VyZc-zgS&ejKW`*e^xh^G_`erPBXtijlXCKYx%_gSelm9Kl7zr|t_a94%
z>}H8yqP@EEWox~I<pYkY#Zs?IMq#vTRO0L$f6K|Q+NHky61gz7GFjg?kFH6@=59j$
zM9hqRml@}BP_jE*n1~kWD6M659NNvW>7j=2?UOxVVfK_Yda;XEocn|Pz`R>>w%iDz
zZd^iVTN!zJgNGUOZVX(;qP7c~WBK;PZLu`l&-H`h;^~d|P0%gB?OW+S6CV{tZUJX`
zXt;ap2dS$l_S7fBcF15>JFQ{7brprNSrZ}VKR-*e2ne!JqA_=*scKWZC_9mw9Ti#f
z(4|)=6KJv4%co))SqujuGCg;6aqz`gVv0Pj-zP0wVKaT~c~Oz&`5P-ae3x5&1S~1*
zW&P@J7|Z$QHfC_5M#;N(r+6TfVGtv#zeJh4(zZrtL}wzkUUWi;f43jS`W!*c&}oBC
z@sLQbo{+F?PpwmdO2Vp3ro%Tup3r16j=5{*CfnqrKSLepC({I#F5HBg{4(a2^F~-l
zV7^4QiaHx+Bc5=@b&u=sdxnH@yJ)hH{+_kX{dm$~*7l>DqW*8Melm{Rqcw!3?0G_W
z52*vgL;Z#AoKZgN@qX+l^2~&^?VA2@6e&s{xDwx6lI=->kl%i=L?UKOTcU8q^xvX5
ztoqe`mOQBgg#^iUg2st;94{V)%(}yEhH>sk+(H_577aq~YX;EFD}_axkz4OpM5$n9
z1w7wYicatDyD?N2A*v2zIvj0qYP%$DaLkv`btI&8-i5Sl#C?4gn3F|guiTUSBkuTj
znI>x_k;#hbx#9XNSexUnAm@{r{RRwaR<WGz8CJK`CF#z$(t(pc?xD^icjIF?n_a8o
zT^~v)Ie9Tc-(8D(meYCdNIp9Giu!titOL>T)S5qYEt?E$#j7<nsJv<WDno?YNosR`
zBG+TV)`t1|5X>-TG;Jjj02HAOrJh>iUXkl2`*T|QcRZsy=)k(zvx#>q*~+@jwRwmH
z_u7PGQM;BWuWku`|Ep&0(zl6i)J=G-_*_`5_H{ZpRjl^*dJki~td4Gffmb4*O5Df)
ziK}w3{&%j*%EkJB#Z?EtjmMp~Agw=W%zxI6-T1s&7+7`u6TD7#x}EqJG3zfalXrjA
zu{4|HYtbgQT=DgDw;48R0u>DLlk3$BPQRDCv-i9EaoMA6;|(}Ud;RWpD4*Uf{obuY
zc1}vDdYQGZ_m;OC%7fXy17xC!R+US~%l7d6{kL#5LMpL<c#n^*hJfyrfOvfWF>D{g
zBlnMYGC#pfoO4$}bV=3?62UNgwGlOZ0W7LUbqWF4LntY-*O%_D_xqt_nH*?5|5(98
z{i*8ak1l2CN;8?`RiNTm^kO5LoTq1f9~0id+0@6HpZ5#zUK?9Jwt)b^Q%8J)g5~cf
z0G|44#g-3qhg4?YV}twKOb8w_%Th_Rd;em`U3>UKaxVj>X-?4p-H%M;{pzG5J(#v{
zmQptEH=C+ARxx!M=PsL;LT;CL%p#1ox3>Dftnvq>&w@#+HjB_#l9Qk7RYyG+7m!x@
zKfWr<u$)JOV|#tYETK~!Q?+{c`Y>#L%}m)Er#*((xFA!DnfNPa`QsPE`VA;gOK8TU
z6s#L6s(;EP%p$3mZANU<e~5B!S+<{77*+5IE-C|8!WN%~9eImsfML{ZxckX>-tbY+
zoNKb*7N*V60-FLkQn;_zX4?Kzr34aPSH+Ywp*jX410xKoA<wVM@psgAZPvgvI$|ZL
zpK9x|wJM@fE<!2B(WP^k!>Q3&yET$0_L+&#(k1+F@{ZzSf6O83C^kw(rfZN)4YM42
z5VNb;!LAk~-D<$nsV9YRKA)WsyCV3wJY3wD;M#pK6f~?k`GaqGnfa{Kfvo+*{xFaQ
zw6c(UBuQa7prwks2|bGb#|mM>b{=aJ`a;O@QQip=f%L<0>r4^J_?VTSM>0;nXK8G~
zt`u2M)K8g^pYw}ZA|~Z}Mm5gxS%(Tp<BSP1q{^}l__UC}Y1QXU|NdR8Ox}3GwY_JU
zCEpy^bGRXaT1uXP7PTMYPDBZM^YlYL$m6yvCwz4hOYCE&Ml@VPDJ%v^q!3P(Bo(}n
zTN3i8ytqLcv?SZT*{rLo8!V)#Rw41t{6%B7bUDYbnl33>U#92e6Tn5aWlKg#wjril
zjW)$Q6XRM5Mft;m)yvT3c?<}gPr=>#uE5SRp7GbrqnmnjV@zN>RDV(QwEOm;b5|)w
z4fU_!EI{oQd*mlt))8e;gNPI}E0d9HLL51BeG?rjGu%!izPE(5A199nmq{QAWcma-
z*j$hH1l>RnBju}QNt08%wvs~WyGZ0WXZ(qYc528wndKUdp)1D8^YvRM>omNi&g(Rc
z#K0*dpQT|mqEQc|u<=FANUmOFAxxaV530`S$vMj{K&_(m8Ekw(m?;acHi+Z*33cGR
z#FXpYrY)V&7_lIDL?|W9Ti>MJU54!vp>tf(U`W9j+7`==b?P*jkT{@~qp6xIscuNT
zM!;WxBeE$rE>OiTDjGBuoY7{%6WO9P`|4at|A|)l2MLYU6rX!%5$V1c2ouwv-TX!^
z`Q9ea;WS}FHST4ZuYjTM$)ZtqW*Ri&^66a%m3mk~u2XC@6rwaFCyQ2D)1<SMZ%~O;
zbQR*8<9Li?D58`5H{~&d2HPox6l1-Y4Znz1LUT%bE$MsFLTOmR0etUAk}Z$T@C>i5
z@E^7*O<jjKUn_vSRRi;`EHhB$3YJS1y<v!CzU`M{(}L&q9yhC`jecToY)cR~`CRE;
zrS_XEs%lWqg}&4RX-&5hk^6uY&idB?m8Q9{#MhYfJ37;`DLA{udW=0{k!O}?HZe)d
zK;uIK0_Mf2!zB_~k}Q{qxD4a?Ln$soF-f^~GM=OZ_n0`7fzO3xr3}AxZD(p_(x`ci
zj?)e`AJ*^*P+gl3H7P~o;-Cz33(0WSk!KP&?}>TP@PH2nEr0idl^kd=5*%t6mZ*>n
zpN;_CN5o`o2^)wF?CHg@tSyKA!6W>B@V<?Ny5p*fci&>x8@aLBuS@&$C(+p<S(Zxx
z(f$&of5;9v81tRX+vX0$v)CGAQBVbA<`2tsIHMH`CG$+@I44`UO;zR~3)A~(RH))b
zPEeypOm-@gM5<7HGxCoFymb;f-fu&Mpn05wxj&LJ26|-=5(ksHSJg+&Yql7(!;HFG
zDeF=}j7=$mv#Jw?UwHX^Tj6a9qr<94QWaGcwIg-SX3PC0?e7RiF}`f~0sZO6*Szj5
zHjMr#un>CL*P>NnLaQt<tdp6XdX|n|?neN^{9p^Sm2#X8v)!6L={zmL=h$yZmh?if
zHljZ<vdYxl?Dn&*N^%aqClo|fRk)}dR7Kf0X#BW?PAu3__}K)}b>5jQ(?Y>3uQtnk
zv5QoQfIdmz9;K0EcBg=8pq_Wamqbg=$eR)o+pUdWMNi1S4DM`mNd=CDw1$blSW9c9
zSY6^)c=FWu6YaGhsK}i?9NBXGDbHy`ikX6dnO%XI;p`5csNLVSU@Pnn;>(h@@^Qfh
zrTeq~<ts)L>YV(rq9rA2!3^TldK?eyNQ<zRlpF+cv3Aop2BM6aRH{s)?Ty0#!^J_G
zZ6c7it7fycUFK}(qC$xESF76TS0^jsl%oV~z$jHd^<5D+_RO1F&5$hZsiV^@!^F(-
zgvu-so;qbF1|2xN)21kK!nD60WA3+Mo?ZXa5DVL<$k^)4l3^`wHgKD$!8*+<3=7Jn
z=vrCgv_h$J{gjkOc*<<0WV%uWv$aHGr4;v~9IV7?8y@3K?fWhfJ)oJPg)_XczSdA$
zU^KG64aeR;C@nCt>Pj&p9HTVV*lYa)%{)g>u<Oi8qkOlD72rt*<-8sfV~j{fdK)98
z$4c*Q&xKNjcOtll(bC#l^R9|js%@8k(yiHSrTbx6#lPSmVjO-)wAa%8PBssFA_#3b
zY%zbd*dmy<=pS;^`)n_c+|SreDbs4JNq1k+CRFYyBh_*Cil{|C#`a=(Y!ukPy7<jf
zGIFf3>w~RVOw{0}OSP10o5P7%uDNe(w_G!1;aJFvGS6XMM8>NqL*?%a9qBAG)>aq0
z=nod~g7cb)kvBZGxVMlJHtC%ZesVh3{So+2%)?FS*mEQ%e(Lv2BN}~@S(7(3Ms4hs
zPQy7U*ri`_dR8PDetb2|{_*v9<8Xrvms^H9_3)czsuGdhFVs|qvM_W~sEX3ih7{H4
zjZG5u%%Gw*r1nsXMT&hJ=RRB7wYc6)rDm*BQ6HpvjWMu|ele;qb_{Uwy?D;+6>vW)
z0!EaE8m+hd5I0lzr7FQf+8WatFb8N%L9#;_n8h6HB$8vPbTkWoabHNDZe>28b*Mvb
zglRe{Bz1-*He?EN9-39)?i_31d6HC1N))=*0vc)mdQK^%8vJ&ht5=N^HL;N2RB0`l
zL~dOZDf<ZO%<~yjDbY24+Yzc5>m;<q<oa<my-UX4l72-iFOYr4MB2EW;e(Q(a9~kl
zmbVQRXx#F4jo@?fomYyd7jvJWE+PI@SaF4qfiaw=Z?ArZfMYK@l${Y_GM@CK5`3Zb
zmD{pcktEh2U4#k171{OZ$Hj(*rts?7J5!<JHXJlY4X#zIEr%4J-#P@SNhE@r4|Xz0
z<Xsd!Y<@x*hONyy?_D2{tc`CpZdLCU($ir9ch|1-t5U8`A1qE4pi4mwxRC748VdE%
zm@7QJBZTjjd-YSBn04|hjt?H`RLe>>m#~0PAs*ABMScwJ-Ca)gV0_bBULnzK)yk1?
z;fqzEL)P)0pzLyVCNM7I9Q-Y%`PE47Yc%s#nU`+WgBV2t$x}kKM*-Goye9r#lf^qe
z^s3aukSXSt*M3dCwRL#Pt+MT-8PsoUkkjy&;UY^U(L4Y1N<*evpPw_kNJhK-zBEJc
zC&C%H?&Shi-MR@5W=P2^w?jDlULAA%?(j`pF&3(UtS-ZJl4O$+?=M`U1K%7a2Do}~
zIH0XIY=$#u<h)k#%rmUTB}31GSqtT%C#MPO@ngBjo-cL~QQZe|P<O!02QZSdir~$0
zS@e&*cMzTVot=#ECL=g7x%%DZoY=Nw4!h%7#nv#*#bO|HNUYZ6{=~78XVw*PPBJYL
z<m58=?%Ck;8@8A6b7-S_dJ$(W+1kcBiu{Us=Bb=*dd^aheHR|G{VRy#sPRE;g#QT3
zyi%w2V_k5&DdO;qFyiRShLLCDUNb$4xTC``tq|2%^T{`Y-m;w(@slGMlL<_Nwfxne
z0vIwOGqN;Eb0o>3m8XI0l`J_L6qUHYy2+y?AlVu16zMav5L(V%!O#&~)cny-!qLIm
zmmg>1nJ5{^P9+qo^n(wJCLQJy<&(@vtZz+Gh%<X!rCWGBe~crpr^1@Y!&}0iT)E~c
zF1k^LAM}@n2EG;ApJpTcREcUx)Bua_`0I0JkbY&5x*nK4h2km)%`%2}_bW}epF6|O
zD_w?p&i<DH>bJNzwDsv<e&Q+i_p-mMi}cZiAKuVVA8yBasFU=udPu9KIiL2<dz8Be
zrnRac=wF@Ax?PI|^Y!gNAp2M4_y0!{#`)h!7z+mn^Zy+Q)0v1nqe9+TRNaMUh<%c5
zqbYx%l<LgRTWR$A!wM)x(MJ;<?;r>$IWz?)mE$BQg(wtbyqA6se1SLb?YX~sdV=|W
zZPM&o`)r!bzVFQ70-KNU_;|5-bYQG#rBg(6)qlCMVX8GcFb-6<O|Le#n*H7>dJR8c
zdi36K{sQ|*4oJ`F9z?2ccDcyw4n0Uz@LBXz#JGD6^N3U|p6cY%gYl*zcLl+|1j6s|
zf80OapevF(?#s6z>ECw8PyP51PzF!M_)VMP_i=?mNChv)!wheV)9mzoqd!Na=>NlH
z++fHEZv-B6FqdB-I(IEQTMIeQ%qejqF)YJ;KxxdEyd|x|Eb~)|DOP5>jE3M`c>)(&
zy$tRyZN-+%gXGw;FQ-V2TLBj>LJd<zmCF)=GURVwH%^=?33nWsV9gx?qrjXQLS$Mz
zmHyoII7^jMnCColYFM~+s%im$eb!j_<&|69=kP4a;F4>96jpAL8P-ZF;e_7FE2kgu
zMP+i<tF_NvKN;Y8GEk1_1$J!+KbOhTOg5xHTo>>`GN>cUZOaK-NzC%>=itlfADz)9
z6Y`*=&Bt6`(CbF#*RHA%{XvFh(Ri_tbv-Q7eX;gQk#Y`?Z?eY_AiPy3{n!y$LF<^6
zg#QIeWb37WT=a02=-a?rvxyg<rF=EBesto~aI}8y7#qKxR9W1)BB}F}SAeH}BDl-t
zam0p`zHtI!qC&|8YSeHYJ=A=|?Nv7rM$Dcc7e9%HY{*vyXK-3xqr?kEXQYEaj-I}k
zwl=herzfrSp=8O4hzLiAXE<Z=TMLYi+Ycb&V`Rflv}h2`CRgzAEry={k7Eon_TZsd
z7VpBxer#wwtZ&N5M@xJYRQ0WL-*Jy*_D1^ip|6h#(HSD$GmI`q(S?HkK~W_<yHCZ_
zp~vXc(f*T_9z9cIGZjBW4X)2VW4o$@x;VX23|tW%A;z3-(59JCga#Ob{7{1vA&K7r
zb9NeTXoZMh03uqCNE?pjwgI@{Y!3~IN=Gxdzor8<zo!-xZBwVI94(HdfdDmd2`fos
zUh61i*f76EC+WCRd_oG8W+=Cycd$F^@-Tn7K0%ef(o^E5W$o}!Do$RSWxhe9ICkJ-
zZ?I*+nH0BORmSk<aS;7(g3PSu-5Zl6IwoSy!=r4cea}Y`Z=#%OG<qe^QLXf$%Cn_v
z``g}m5@ltRKii@VamaFFZgTFy+qN!I;$DnZzlXLcU|C9mSd>yjlTAJt+KXz&>-%?d
zqaQ^W!=XXil+M4(WUQ6d^<p}9^b2FlgHBlWDEyM|Nsazuoxn|Tw2oD=HGy!6Q5CR0
zp;}fQI3j44B+wOh=%a8werLE?P*gKTjjkL>`{D>T|K#TzRyP^a%U)Z-o)}cI-%=|r
z@asdww-x2XO;?TZk7X$xd{>2n)^pFe^#QL@=FzF-vAS1T5L%{8(<NdH_9LbQP|I4j
zLOZ9Gt{D%VopFvnwu5Kj0qDSEQ-vm0W&zsA>H5o@4I88t>-gzncRg}8*4SeP+s=}5
z@Ha1&;|F;|=R_Y=cEprucK(B72)~`TrCZA9@(#`BP6I{dM8?MTKgU%zM}EasuV0^R
zUu+$ET?;1j^l5eHx&6$86C>k!I$9<kUhf=65;UAmp?CBa19LV1D9*3Nf-G1Dpg);c
zFlvs<E|o4$z|x_A=Vh=vJ5KOq4^W{iG5EGc{#*BQ(ObP^SEXwNn`DhVpkp?d7{s{8
z>(TG0Y$ZNBQl9Hh^c7RPPck$2i*ds=AMd_L9JwPyP9dL5XzLEnmjXC1vS6_mwY2*<
zoRzbVTx%aATiXM5n-y~6VTt4O%2liR!eU8UJ)=+?Dr>yvfWJz<)#r$^KQb}ahiXi)
zh%Y{wLyU*CaAozsq;{HBDVMV(yxwsNyrC1l$K;BJJajMJN#cYleq_F*a^%5=5QHrT
zG7X)AOfFFNlMM*5dooW&M2c2W_~MgW;#qM>x#!;)v~q;2sn<JWo#6=S6I(>W5V-uI
zO@_!|I8qZ6AUa@@<x3mkHb6@1N1n^DJg#wCc0`$H{a?;mFNM^d-G_S0DF=@h?JjJh
z22Q0$(HukO$Q<dhEQ4%Kkr}skBgkfit!Aj<*ks|qi=55C<508jV2bSUm^p9QJW#{^
z?ggg`RnuMelO@KTO&9TEB`TbW#l5c0kvMG?7yR_tMeS9J^^U%mdurI)Ez|Y2WHo!I
zvP5iDHr;)URJIQ$*P6|KisO0j1-BzU{slyEmLJq6F(o34n{$tYty=iUPR(wXhg_J6
zd@PdYEG>pd>zF^4Q8zi!`TIN%5z{a0lHHyZR99RVQVDR~Pe@fQAuAF|5TKa=riNyj
zQjFlvV*#ju>m4q|Ats??6O4otJ}n<2p1ggFs(FO@B<@|hd#T4!X0NLzJY%x^-6(3L
z22QJ@Q_{<y+I}LUA~X~x)g-QdH{v{{x{O;dre75cpz43ljhWdx{eDT)-&@Nba}qe8
z)UkMZE^UQPO{sonmf#zytf-$u+#+eFNoyy(&$**#IZo^T!5>XeC|-qIWtQ0&hQ$Ae
zblyw>7lC-?=-Fd7Snrb|EYdJ#3TFq$qNVgj#y}R%&X`K(9pgSUiwCn$g=`5uYa!j$
z-dXy_b-WZ-nbD<OnV`dN(UQ0<^Ma%pCPGEEgY1L`Yc)?-6Ho`sEXEqa(_WT;|98v*
zw+*$s7k9#)(fKL!2nB|*tNCA6)fKc;N(us}u1DL}VNCkP@z**Ao39>dFv<!Gtd@ZV
zC3|@$v)>pC680%>*M5#2jk8}SW1YBd(%+7{L1a{WFG=LX==_wQx|5l5EUo|A>|;~)
z?pz_{CCR`TAvPOW&+91-JL#oYjN)0{RqpdGSh3GC44dE~k<3y+>3CNlLJ_U#n=NnG
z*!c=3=nLe1uhK1~_bhl;ue;nyE3lPj8X@f1=m2@T3<cq*!XF)!K4@oIi0&{F-mlqn
zTe7^zU_yQB#r4;Wuqb`QU_w$C(crrZdT0T+9Rsn4z?&@R&|o6ATlfJ=zCPdqF=B$q
z#3@Hm#4qB1`Zo)f#*|K$z~fjyTYcgk!&1HeYDz=#SHTz-$2H;E=Bg$CIY78u%kP;*
z%2%k}qx9x^K3s{gSfKcU>JHxLyjbA7i14JJ^08={e!??K6s0PWh^h6E>7!-gM-Xsi
z7x5hl#!lp*@6?)<@3RZ+0MG$F^X--r^D7nS$<b%H96$;on&!Hy!D)LeVIqvZZw#W)
z+S)&fWP>A+w;1cCirI>SXLx^RD%~^27;5OOMm9d*ArEIxN#7BClulhto84D5R{5G5
z_p)cG%DaJSh+f;7Y=}P7uMr{-IinaNA9CbV^CLLS$)O%P+{fWS2nEw`r@+*n;}=sH
z>z8@EH!i!AlCw4^w__-)66gW~2QY1g@8UnZ@$0oYU=segb%AQncfNqD4@<vx4yNx&
z@}5Fp3Ri66d!d;6fO{bB3d9kF?5VMDTn?izt23cn9x{0kmmWS0AIyD#^Qkn{{!fsV
z^S|c}*t!2}{|fQ{VFOF<*9p5-Cd956>RSjy@yOyY-{DYx$k(p(>~fr{Rq#WLKEH=6
zr{xeWGi(xCyd@Hi#wS@$gxBtuoW!s9_a?)m-~_V{y9aw2+(oN%hRO>;U0K9Y-n`6c
zQ_^%IXq|M*Onc5JHsi%Bvon=%x}n3l?kNwZ&Jd{GyAn2M`v`0zHf`0CI@LO~Zw>rV
z9{(9(h2}Zwr}mj0cg^+#-`}mN)PzJjZ@=IhKT*5$BE`sQpb?Ax4+f$H$;%lW7l_jV
zpLh0)b?Z;w<N9KIQ&CKRufo-($;py>4vPAStZ|F^YRpE)a?ipK8PyenXGP#7FOPpV
zs(ezBEzG{_imCYey<=g9a9^1k?5QSXtXwnxDsk=mMY?h)SZE9AB%@M2*X?B66*0yV
zl{?i{0zcs;Ti3_XQuAS)Fd<;?*EAzBZC?}NKwEXxup;IR)&R4dPoDVhuFyz)M0(48
z>Iyi3JBisleFOyM>Tgc=q9(yZe521#*Mr+EqQS=4u*maTv4Bnn8iS}ejm20vrUG%r
zg2f=6Gwo=0zf|rX3<^Kbh3a~!0j{g|S0=7~siP{ZmuD2GK{t&BnYTq&Y!}#)yLmMS
z0;7=K7IEE8s3KUz@;`dpqD4sO&$Rgdj8F6fcde=rU#yhV?JhhPFB-d@b)ZR>r%4Oc
z(S6fAKLUjy6kdtCuC!?mPF^X)#O7Yj?d)xJf5Qije65$@S%e){#<N+~OKzGGYHiQs
zMM3n^sqdR`vFI3lUz3*UFYOAyy4)HbE%1C!Z+H0Po$En3{9yq(qV+#6Ql|f(SE+eA
zm@;ZA8d;edyD+M{8oB)QTH4mo+>}wn(gfIl#lg+ZC}nDCZs9`A$;8AcV(H?nWa=bp
zZ|h)hXKLp{%*7~bZ)5MI>R@PW$|z>)W@&7yA}R8p-CG<iO#gL<=l@~xpXS1PV2I-l
zAF<zV6icNts0gB_P{5HA!$~Qpl)_QLsT9LWp@R#TgrSIh;uj)@LL&*;kd-0_7x4<6
zUv>KX!IJK}-mV)r-?{T9$j8_0oZRW_Znp8%$y#5pO&=*nnke-J?Z19dUo<%FjXCZN
z4Gs*LWH~QXXzdKgPo14OIy+}G8TKvLn>%0ToGe!7b32z(MQI5H=jZ<|kxhqyfRHo)
zgp4e~!N|xcjG}k++QLCZMD%K_q?B-!qnO7h%wcbDKesO?CPuva_VU1v8~c=wiiTEa
z*dJy&S6TpPV{6Mgc7J=MYU1GPTC#tG`y92jw1h8slR~YkbaLGj5)Atp*oMr^z`!tW
z&Qu>ODJiMn<@2=B=}k>dO;OkB^K`vGk<Dqh=5e(h`m~skkzrwO&gp$`!`9F3ykAz|
z^ySMJxhyuzsXYF92Rf;EB0&bVGNnKWXtUoLhHU*n8>w*ne17jCWY4_69iBJuZ!bvB
zT~|YRp`WnRkheG!=9ZVa%^5)iSXJg{D~;T)eZ21Hn|0W9n#77?>_bRL3za&VET)&v
zWD+qWhg1A~d|90Kdc@ysR~lNq?tU6_=<EM|`0KAf7Uu2xcxlL5Jv}&BqFf{~``7Js
z3Ds|i3D4^T*ljMojLTu;@qBC9Z&OfEkneL1PwP@<_2(xQg);->fJ&{A_^b5Q_8_fV
z8CW+i4o+k@zw>^Kp_#79+w+|~uc*qq)!}pjmHh2Bj;Ro)cv$mazt!L9Ec$_jtiz0r
zHp_LXv>Iv1{r&yQS56#O0Vr5sf2vtde;2+MQ@eP&Iq-kn3J2Z^bfwpB0RD*yn;m6k
z<x5KOC=A2p=~7@R7)Lx<@J1NPq|yx5rP66RJ>MRUj*ez(2l7#|yuz>Hu2}w3uS{t1
zeR1>l_D1&)l@;+DK=ssuyee|2F*G!s2hocQe!SMtJA#lnFna>5V){$vvvK$F1$;c>
z3Hggfp;|{G!%CeD4uNIX>a>Spm^gndn?hv>0jkpPjth$kgHa2EeR)2qo<Ut*TB71{
zxt(dW`qSqbhKO^$sP8|K#s16XXm-T&_HagyacX_N(_*G@FbY#N<PMm7;0$dBy+IDv
zQ8S1EL}*v&MG~<h9FJHoCk`?|o<S9g4$<D=ZYUNHn?V;&Fs2_EDeD~V(|P+v!VNHy
z$toz1#Cvm!g6zK$l?F7rKi+xSng2wkDkoFOvy5F{4HJ<;YU_4EihjJm3AfIY7&*^@
z^4Wbq-X1J;-7hv$DVuDTRn8GuAYfQmxn9k~kqHvcfU1jwvES-Uq+@Z}9a?ObSq9-}
zX0?vK4+c_gc%9JW_YIp_g%%Z&IR(L}s;ZWV1xNz_F3U5zhu?Sc#j0;NEieLJw{Xpe
ztM=U`Ll&rLP~f1LJX4^bTbZ$B9R4UW6O{>S57<FKy?h3faH4?y2H#rj55@eI<>gEV
z`G}&cGhq6QF?2yH0*{-Jo*V&CXnSXUfBs-rl>=+BR1_xfAhT$x**-rdLXVIdFu=3}
zvFcbc-KG?$!)X$A6|TsX7LO~VFMSU(J0(E#W=UwU19M#6+w*lDJ#;xpNYYZTyW?J=
zr|RVSg;!-Es$N=+Dv)0^RbHZ%iHS)UWUxppI3lB7hY;=IT5CNi0lgPsodN;^6;0+)
zh-OV!cPERiA3cuvvfxiU?lUlZ^|5g136*|t&$GL54Dm>;xCGP5pHPBqq6+-xdIDs>
z4nTO8+O4;d>_^4LmEM>6wYIb{VYxax_rn?lTN(tMw749hA<tm1Nu^^P7oN0w-cY=}
zzCJlG7#4uba4$=vp`-8llf!qi&LRQbV)Qzr)WaMya{0&%gno$)gyftoREoZ(uM}Cz
zXR*PXw>xZ!NZ`nU^>=zdELA2bs&Bh%jci_AOsN<!YSZFZ7fPPb6iG@XP{3n{@!2RT
zqbX=s>9+QTe(Eo&$4L2vyyd3X<wHHST=k9jl{+VXJ~H@oVo*|&IUcVY9oom^PJ9eQ
zFeZ{{dcFBn-E2r8Py|N=0WmQOw8!xF2PmBoVxOCnQ&v!}4s?DM$C1zDTfU?fYkag?
zsxfrx5mJ0R$oDTa=k~ZX@bT%$?AhGgYjiyUTAQhJb8!))aiZF<tgMi(PUiBGEB3j(
z+}l!L`d{yk#u4ydiuyj^Cer9hNlB@xVF-Y}?FPdHRbPfv*g^Bf92Icrsz~%0_~DuF
zU=k7%Ufax-%8PC@cE6IA#NczIQTxB$rSq*e*{)`^PaYx|8yOkw0|SY(0UBJJxerZE
zef~(NQZ6(ajG8@z_!#+c7)zmkME4}k;)?>=VKqY&#fTVaYAyi{<&<$ynE5>}iHm;q
zSt{#QXjbcYS7rDJ4~{IO8mpXJK3;B7RRD3R2t~La@J`G&lTHRpb*8f5kVrTzA$uv+
zhC*f58Oij?24@QA5gDZlxly-36~HkZFv|3N@Yp@T8*o+_SR_?sr_C2jql-6Nupv~b
zm80?KqmL)cfkC#T_0L2y8xM_y6M#D1k7mUak&xt3lR#FzLC_<t)(dn+#)Gpp{eh6M
z4olVgg5php^BfN+n09j1Y<II2=g5IQ;xd;=?~RfuVC;JAV{+W;3!T;Pr6q+vkH*ci
z+B-RkrUJ<mQ&US{VKQM3x*yL|yG0Nd$Y*n0*Ga4jD;eN|glcR`LcZMojK*eEzr{T~
zpLwA@PV(_n?F!q<w(nxD@PPZ(3wbGM@fdaju9?i%b}d7AOsZ&LV1O0S`<nx~Z<xeJ
znJZ`Rhawq)yP{0FdL^z3p`ah#6b!DUQ78f?OPdG>PC>%V>fQUx#rOw`^5;JAr$?Ie
z;%!AR@9RWGE^>?#K1f|J5s@IHgrEK&?@>3e7ycir82o6wuE%1r%6lSo`herIJc2mU
z?J1!M0Iv#|@Lnia9cKTA_B0*{zC{@gvC?eMxEJE?MQ-qT{Ht2?o5**hZD`C*N~?v6
za<DLTN(BrWHMrLfz@T-S?Y3TPpf7`S;CSTLz&VI!erYhGdjjSKbhW$s6NO2OaTJY3
zFVfafhL_kr<Ck1Xr7EXKHIP&3GZuXGa?zdsFc~2qQwO`v{<J#?7@ngAmn!_JByewG
zEK3_A06j2>beZc&)(lEK4<-r3gib?LP{(28reQJZwdWHQrGBDCDKa0F$miSq{Q3(k
zTUlf&?ml!D>PElK<!GxvJe$*A0zn%GnH2)YWv!xHNbz`oA0|IQ-w++xA;*}`OUkkI
z4wUMtrP)0?imXy6lo=1)0!C3GU$B@=LwKU&2vhuVGbQMxV`OZM`o1$IPuDAM69mGa
zoF0f}0SA1Yqb>+pr^%*#WdGnG{FdhH&zk<jLks-AA2={9mrIW{MasA#uPvB#BSF|n
zudMIppHfR!L$6cc5d_emO2VlFBDkOg@p;`4AEAE4zB)`gahX8Y+pac=lScIeVbK0>
z7eY0%x`}a;kT5njHU{lagx4k?`lxxw@1&ix#>B-Pi~0WVk+V4hzHNL9OuFEv@3=67
zs%SVsz6l+Kf?WT{7g!T6DpDzskQT7F5E>Nr5Yu(0;|cOtr@<mNY^UKwej~U{-}iTZ
z(d#sU2v+gAoxbpzjLR72lCA=_5$pB2waxt^jt>qW<4880!L$|<_$?|@tax%dIs_iE
z3LV5^DSd{R*VFZ$yauR-sBYA}LB7J!z=)9SGex7c7FLmrIIpSkJmBq2!)-!ij6(3&
ziExW@JTQA4?oYKO2(Ek3->}<6ZZ~He*t6Rp2s1oTG_O*`B&<OwydbwVo8_L<<$C$G
z5Fi58r1oe*{u6}pz?FcKT{OZ9aEC$5U~yZY!Mm?gXTpoW^cIlBZ-XPDTZ5Les4NM_
z;|1SG!g5)8KK}gDz)>ENEOFSg5l&}IAJ{GhT!7g#+*1HJWCHp_6mC)0-6ZeD$T;>f
z5YV;|%3dMD0H?c!GAWi=sZp)xWMB{+81{=%6HkoDd>gr@N0;!l1C6OhSgqd?*z3#q
zZT5F2t0R~HyN@_W&*xb@HVcF8MJ+t2eA?M%k+@pp;X2W#%`t}wv}lXs;q8orly@26
zDcCn|WW<VV-EKdifv8pJMu%J`%VA}nfbW=!imK{zJ?mksW?2Ln3UwHP*UkRKRq6mF
zg2Chcs(q|L)94Vm+r<X`6B2&i*02o(<a6os-3eJ%S~a;4_U9UAUq6mVGcQcwK6(Ih
zU#zvxmDTr|)^B;aQEH8R7De=j>yE45E#g^*4dbv~*`zJkF)zTf8IAZ0lEby@?+Z&w
zr_<^~#^bchp<ZS31&X+Z8S8%edIW&$7__}e|Bp9LG@!6&8H|Ae?bZGH`mYfVz_)TU
z0TsKCe{-O`hOv#AtJxUnqjF|`MzkU6y+2#UJhJ(&=LHr;f6!x3T~|x(*P@HwJ!#iv
zR0?{&Thvb!u$=wUY7R3?TSp?7ponX?(jeWnFQ%fTgp#d2;b>r>_!qQ}98^pgi^r*;
zFli$l!^SkYz+L2u$5G4q*9Tp{pac&ZrinUm>!)86OE`UHWNhr9IUXL~oQ4Gjla#!u
zSKi{>9P3;GWO;ou?))!^;7r>bdaR`C34^HUXi=WBFnKV3pT`9L1H1X*?O&_S_6HbD
ze4f`(-L(1<djLA%@V8i~SdPJA1-MLs=3fZ^Ap7AMoT4&BD&%7SyFVH$4M$Q<dV@LP
z#Ti5#2cvc>ewF~pw#l48$Xr_bZUn972M(DX$-0&uT1l1^5J-O`fmCfE_YW#`M9<iP
za56yxfo4Hf&jvTpvt!O~y%zdWHVO&~r;awjH1Vue>NMlVS`=GKHoXrxh|O<o5FP<!
zjFX=f;UoCA&GlrTi^=D*7d8_{tk?(pWDy0e?t4-!m=wtuohGOQeG*9NUl|xZJv1>!
zI*_){IRw9|4r`VujJcr^QBhGNFgCwu&{rfv1j^;-GtVB*6xGB0<JoG^v!69gJ|QBS
z#HAC2g@x6j3igBY&&bT59pg11U^tGc^tnsrbvAh{rim@5sC>Y<rE(!wN1!#V8(3;4
z{Fbj3SQi;<2MquSRG3z#P*Pf2_idad$w$~G4ifSt4{=8qYhUVXI(;AK8dr*Zt#nU7
zjg-Lhvy2hsOKGzLJ2)a;S|V1Ulo4$Zz|kn(al;Iubwv9D_{eIvF!~E1?>#<M$s@yr
zHK)Wf$as6yfvfZn#-1X`Grs@EWvX$#^YinYKU8CAun2PlC=7r$sK2!}G{D7fh4>(K
zAAvK&gFp<paZEKh{u}{QLDsK1N4dWzU}0=m-UDqE@F~p+OiN$#>qmJLR2Ke;%#&f)
zh~fOpb92{EsxHG}(Rj-ih;%HSG@Gdb;HD<V#>Ot3Ae5l&Y`j9L1uZG-pUZA_5)bvM
zA?R3inkc`GC*dHYk&r0$q-juBNeTd0S_NbDS%2KR3j)Ly2hhTVnFt|pL~w9$30O!R
zK|F~C8|Y9dM8uUqSczbDc+e8M^m@DdRG+VbUSCZ<J31(X%yzx4ku+C+LpoZtu&{75
zaE1mQ9UUn<a{dSpGh6j=p)#8f)i!}&l!ia4hcpF<jx=uh2JwNi-V<A^KwK^)8Ryhb
z3d=4YmAnZI<To>FIn9K(L-HeJPR#T+O1M^z@dHeJWdE`2^;-6b)}_6yhiX{vfm|I7
zQYYHFMFLJZ#AmWiI2dNu3!XJFFwsha7p>;qeZEZjbPKxQRVfU>@?koxSKkvX|M5YP
zQO66F(HMhO&^Er`eVqL`Mv5{LO#RvchE;0IjDSdt(*HGu80GeeA|U?==+4!`jb<+g
zpyQle^ray4xq3=D8^bW4PuNL6i$P8)pSlR%ke@NiSyjLwLpBgL`cYOed%%PSJrRdV
zr>#(wQs}f)R@|ks?`5btT1dcyySoLwHC4JSSyw8yRM9MgrBVn%>9Vzni6VdqK`xWR
z#BXN?0h)HK=hpQ^93+&p*yw2GGcp<v05`tgO5Yxg?nfG{6z%Tzyx9lX)R=H5z>jJa
zy3kVTb5-~ze-p+fGnNIYnoj5n^bQb?rNu{m$5^^?%#4EsK|sEKQCVrUHfIxDUHS>c
z^ZD(0#ujTfJr~;3Ev-Ryv(qcResSL@ic&C``xv20?Z#bCP4qCx#3l>lU$=Yh@n#5!
zbcak=mjbxNSsc>@^2Bo*ebVb#48I40mnZ^yUpXdAmAW6dKXJ!zVQu}eUvDEPhrmpz
z6b(mK35m-NCjJyzst=!t%Jlt3;)&+9AUHU9OE;ikDg-i+WL8z}lsXdV**8MaGVd$3
z5f=M)f3i4rOWY5DR|i8XYOCO5j1^oj*oM*=J(}TX3`@`&3P^LV=i{((?}AVTEI1_2
zcqm4z+mGJ~AVrxa^`^BK9XH59`XNu_Y4ngHHRiQ@^_HhH8-6dg=H}*rlmaZ`#R{!t
z`j^C;Ge9d?CTwCgpM>K0p6}%7sMTyoz|$i4gJ#KjJB+rB(`Jc@ME<l_A^@CMjV_wU
zX%|l03qZg?z=AsAuv>%s03#~-p8oQ1kxug!0Xve4aCurm>pO`%<I3%iCQMPR7|rPm
z$ji&-NAwyjg}tLA>Z!152Uw*FfYb~f5b(MMupus*6zZ3$lps{$=s}zJFxFKbAX2LT
z28W^Mdj-S_)QN(Kz0b{ck8ukuT2Y&;I1NSQG!$^tuUf8v+5-vzB42v|xEehh(Nhwr
zKQ<fYGvRNT)tL}f^&;&$`p8{I9BC;jgSl;fuUjs^SFXBTUU%g6*jGykXr>x|*<~w!
z^{d@ZVBbb&2=`S>F`#yV?&~YQ)-_GpfPYhzO6Q~E_?KK(FG;o~&Bw==lar%apF{@L
zAI$KrP0Cw!GpuG$&MCx$+`vFb<L-1B-JP&Gnx7pWIb>L!Y4&ofKhYMTzAb!Q5*`{H
zq+j$wqiNJZIGanoY0dV3!zc%kLj2Q2@Su1BxM>Pm3|FL7aejUd;4HOM#^JmW_BR_|
zYNAG?L4>)RNo+>FlnW0VI7C2xli;JZg*J2fAxS+Vo@;0fam~!BrhMN+jINEN4YbLW
zFE1~z$dm0O@zx495<)zWjtv+Mea<;5;<5St*C7!MFN8UzPsnz~4EZ|&HKYh-(J%Mu
z<IHhWi26}ZaT6oiMkIG#F+6j|u0%b1a~Dy|X>9Kg4D=v~qU`=6C1BdAOhu3W<ruUe
zi+UW6X33Av_@Z(Nx7KWIZ0c@G4OH}^=wDf%Nx#^xyz+WyM(FO1r6#ksk}wf(h-pxH
zf^k6Gw<v&8qg*4LM)86mKn*bVkzeJ?28V}>&TD;#ezWD<VpwtskER9Tr5c&_=5SQW
zd&7zN?ftIAQFtKslN_#G%A3i-=Ea2$OGrS$$iIz}%b_7Bhj#zt0DoJdRlk|A*BuRa
z2#|*KsV1<cO`WlHM#iZym9Xf93k7(dzPFV0E|R`j5Hkc$F|4hW75YzIA)$as_k|)B
zsN%>1x-3Dixf)Sn1S*K-Vjg@D$ihVv?@a)g%xO#+i*0*<Kdi^+;e6f3_H&mw;ouiy
zQTG8BxCX785LzLQemgiyv6MiYfACq`B%pI<=t+?0s|#gRY`W>H2*I%M!87bn7l_tC
zzqLfZjZKRv=JJHv03FKNS1y-%6pQR3MCP1bTn7G<eqrtqcltnt5@`ZZMm9({v^GKu
zkij6DnA;w@lnW@N8vCZ%a$x@s4Gc)>F&Cv{B+OGcSf~Rf1S8Wz?3KM|1_HM8^75`9
zNMySDYQnzTeb3yfo)#*`HM-p$30sNQ)4QsJ^eI`Q)Kb2}Kt}3T0T`*v-%JMglur?b
zOZ|cDoiVzBeLl`BExQ#cJH!h$&CQ(rm~i=FTzv}*>b1<d@XV4wmrAs(9xE_|iI(Ex
zW>;)lFPlGJZ$9Th;zcXbc_u+gLP-Q+UQWT?vaIAIz4hhsdpjr!6-TAgP*H%H<O}$g
z-6hh|){MEBF@096*5gNKbg7|>1;h{%!dJAfOe%h{MYwVJSy*kXP>W&JcC|9K;jJ-7
znt4ZnFajvh2~Tulq~*egrT;W|pNtSYhi?%8L~B4qr)|lHlg4GYGU2+GBi;L{*WnR(
zxxBpGs@v+c`E{;3iTNjy|7)%3cv=|acEgyrsZkEMJ)a&RzWk8O6uC?LN7VPMI9qL&
zojbPzv3~iN#9u73-s$ZQD5BXMwsmodJpU2>|MwKY|Jzq41hW=wg;cY?#EDW)0C<~~
z|L2K3AeR5_@|Eb8k&&rmF&-sN{dAg@lVc-~1s`M(i4__i&S}4~cBsKLJSKwZ^vMKO
z%cK;LW}5+Nxk$g$<En3LY~+~^;EN@x;G%#?0em+yDjfWF56C<l9y>@y`UL?7DXHwF
zw4%gL>n5n+kbig}VE+9$)*B_ogb)>@O9kEr9{v6p1o<ebUo4=0S&2a*0_RF0Fwpc!
zW^5I8=FD38=b&I{u>QUAbU+z?1Xu>6P7@s|X&+N?eEbwbW>8R29T=iA^-2`4>L;5&
zHU6e&Y<~#Y<ndb>6xMoJUz_t@qj>uQhx^_@GeHqS0fBU+(RJ{Mg5P$wP$=|w%VXd`
zaG?3mhyW$Y<8`m9tULpJuJvLS0VZb2hyx(BGWDFgJUu;`f`jxd8t9RW)oT_pma7tL
zdJns|o3a#!TZpPIDf6Qu@xdXvdFn>80=I-?BGf-mCHz*!L1jW=DkA`}Ve&bUX+c6F
z@`Ye>x!73$U7y(+wpAj5msRYGL92&|9uEcY&xZmOEP&11!aF}euzkTMlsTJ5bldwZ
zfTiM*%0UVNQx`!uhRh^%hFtQ!_Tmn^>|Y~SLq<$u!<GN@^XpU&w=o<dAY}uoj}80o
zw|hXl!3T#?wUsa8f6;w?g86&3)P1RBn!f+T`n4WP*)}!;n2b90XcU>pmzSc^n+`|4
zc;HJ~qg2Bb0_4}78o}|tMIAQ=92U^45@2|Ly8Nxph52-DeQk|GKC1ypjQk_74GrzT
zKHX3#Z13!-{!}aPWdRa2w9N7VB`I@_tS$^FEB?xvtV1-<=ttG`jdg*3&rh>Z27{0=
zPCMN~JV-opp=5D_)Ut;un+MQKQ)#~9vRNoK+pPoIxPZrH4-?Sry%G>ARY~N5B+EDp
z6JM;CR-d}6M>YjlugqokKVxePG~5}AbwKXz?JcbbvQQVnp9Po_NvNpc{LH%Lgy2Zz
zEyOO(MABTSQ$9}dhZh7f)~j&*1HGSvu&;F@YEXRO)r(tNxO&@e+q6<tMfLvC_8%V1
ztM2}h`c{s?5rK$mmN$JXfY?8IlqSH?ZA@N|289TPsE&vTSiC4cDhQNeIoTin8AyOF
ziG}H|hxZEV;k)OVFDY4tn@zF{FtF$2p~_1qh^yPr0Gb7m2s6R}>FVkl85z}mMx0c1
zl_kpNa#EA{=R>Vv!T^Y|P^P>CNLbecDJd8=;OzlROPWVX5-}iVo28mp2pEdLTvzqZ
zK+hHFsIhjo6C${BkuLUs*2qX3y$I69V)b8u<i<dVmH^2jwXy&T|I;PrxdR+-r`;gv
zPj<y8K;q6R+Cy2#nKK;Ams+=&h17u9K#7vc0birVVGBUI&9dp8o&2++qN1NLsCtI*
z0aUaGD9eBr(*oqpfJvTO5kOqY**Ln=4q1AG8eZUFTbnxjgfj4I@sANFy6Q9l*`G9e
z9o#d<V)AJRTwo+#&$s1~H!kOEIC%x~Z5zH1>k$zV*p>+J@Vy^`x^f~q@3Ztzp?lF<
zCh{Wj2J-s@!^7dBp<2n}j*gC7TU$83*{U7k?+lyWMmz9)Q-WT%QbjX0Ghu)PNF?a@
z3gjPgD$^&u4Gf&TNIP7G;*o5Tori<2G4S9(Bcvxy7I(*gfZTzwd7|ZH^M;o-kSL;H
zF?nW)l&o#jige;-m&Q*;jb)mK3&xcMIq7<SFu?20+udMoEWI#toPI*#B8kTZoROAE
zC03<=cNdW9D3!#>1SB2+iTVL5!6{g<9maM>6(;W^{J*ax6k_YcbvcIK%P?fY0wF)g
z{Q7uh2d&tb11~aQx}z&*3xT+dO``_z6wmCrFL!rRncU740)RD6<qLl7jwVk|P6AQJ
zvEFP#uG@VT-(cxgN@(ZIF{$WZmy6h7<gEE?DMH&!Sm{wR#$c0mSk7KN%-(`T;y+iQ
zQQHMb#KE%w)9g;?3!n`*H@8O=OiXql#n9#Za)0rcXxE^n#0M*YgyE}>Re0$%hy3QJ
z#DAH4U;qgtO|ZK6aVFO#B4cNB;j6*~mHRr@+!Zba4B)^EeG<5PzzT8;dl@=L|Lx{b
zn0EID`4lny+bJ9Pe@ip|plbUzcLvqGo94hQ_xrs+<KM2mzGoAwQ0H0Go`^EyKeO~i
z{Ndqv#}8gc?5|-qO2Ffh;hpR5MH)hh%B1ulgS98f^qu?9-ydW?iw6JS@Q5(efAXV!
z@o|r|_dFD=9vl#CDc`bt&GPOGRdUw<CEp{2l_<{r32~aS*+rH}vgPVOC!<kziMoG8
zcieD&e?!;BY5(XEO+3q52yclb?8vR&#G^!G0(OZh=YjXnLDgPz$iwT%%OU(%s0uzF
zh+cVNF071adLsR}N3A-%|F{%&g`F@zc$AM13HV^xpa3rHNFhEfOxcCydc{CkIK+Sc
zb}V(kk})jO`W#+#@|xZ41C|h%*NL_2@=+ZH9$27tR2l(Wo5t%EVke@&Li|@~3^pC5
zR9JhJ0=dOiTYNBb&oB4;`ZJHxO4O8UORm#u6~h=vVK#gvaH%}?DwZ-8>k$GT29ew1
zm6By#HwgOnf4k$Y+yr|`h_@?^iUZp;>ZFi>%a&T;D1W*6@5l!SqHFK6Y)tPPWL!@w
z`ICh7;H*Qqw;J6vI+G}_4uiPiXw+MY+fbd3OGEbW_-$<=c2&MT6b@k_>=i!fM)Zl$
z_WOA%oeDgY(E99#{%7PL|Ft@6f+Pm8_z?=-`~#-16A1A(czU?`Yo%n%y~(5JAraf(
z0I&C#yHELFQ4)no6N%|CT5kUQ3LBUExVB(e`rNJ|G?x9J6;KWfO7gTVX)qn)BrPZS
zC#V94V=^&WLaj$PWfswP$TwTjTE8eH7m-ElIU59+71R5^K0_$}XGVN`dN7g;-`-iK
zI09&zRqK!K>^hCI@d>m{275M>n{I1wk>zb@{xj*W5W&0W&+QJg?KJm~zuqu?<~x(b
ze?q{NHP=@`SGF@M{b#_;Oia1(SUM%u;WpkGyiaPK^Fsezl>f!sUxr2beqZ1yjihvU
zmoOsTk^-W1$EYZs0@A5SNUESH(ny1VbP6b_G>SAxcO%W&qu<Z(ob%@Y?wo5xuS>_7
zC+>U4T5GRI3@Jg!QO0Pf@_$zhg-6BSu98zsmGyO8@^~h_tsnD7Ll1RkL^OR~D;qyV
z-|F=$|M&hvT%25M%ld;R?e6Q%!+NVqZ^P2kbq23wpe{2#LV<sC1&8+d-NL^LV26=r
zO9|u(xx1k5Xz%DCMOfp~>^*ujSW+)Eq`~ZzSNq1y*X)wsm**6y6XC-OU!rizXG{q-
z6OFt3?-;`YC~s+L0e}|}<MIdHDS&9YdwGeO8UR9K05Az4O|i`KlOFGHw;*$MvDb;y
zg8umOK{_9gd33pJ(FgxuQHLpVl(}*AHIiz_9uKux3bDe|(^<H=xqrq1q=8aE<mZ1~
zOs2YGiiMo^0jjwPatZEbCJtn$9Aas;*wAmIR-*HzP`QQt?f%XLvp}^=UGKYiP#+FA
z<<VPX+c&v$#9GSk8-KV=i_Ib=mf5CONQ~^CmKYdT4tvP<T&d;6Y9g`fHC2O+kglz-
zuMZLv=oxZv;Bo*d@;JZ}2&MGvVV7v!@o6})xg(;T@m<z_T$TuC#;Mu*5{f10;zHwD
zYMmezf_{AX%sKu}Ys@&KwFKGTW=+~m^QSS9tnnioJT_F>>Of2g(7GdFOqsatclt9`
znC1LCi5VFg$13ga(a}J4M9K18fAA8GFr|=vYFZTR`i=oy2K$YL1&hTn%Z2`%*r;!^
z4{2rBw{8s-Y;4}~$^5lh$wX<tIitV@JCfS2nKCKc)0gd;*Rc3^VXu^RIgG#GOh5>2
z8I($OplxAe9M(=(2!W?MRkT6%{j=0+s@<odu@RkS-`W&1N3_V3Cr>&$IvUH^o`i>g
z{_<t<^J`|crqg|2-REcd-vcN|bC(z2x_=jEl}o+?so3X^h3|yGfUW+q*wPv%=Fomk
ziISD4_~LBrwbr{YEsjT?$}>MWe0R(()N#9ar^p~yy4U7K45|!^3Ymc(0SFga&y^Sb
zE>4b)kRv<fq$MQap*mfAmoT-|=h8_GV%!%sf34s$m@@rd^MAOlelzMz7n>+%+p<mF
z^>BOQly-)IPwr=!8CzmRe(u+Xo-MxgI>O5t{3*8QzCb1c%PM^M63xbBqZzErdtYMj
zzd+%zzz{x2dk?aSJD;#H!euaVawDwQOCLQRUA}$y=ft<O`ImP{lwosP7q_5!O659W
z7jDkqTvuB?+b(c!PoFmI&U{l^_VA>b=Y#KPIG;rv{hhzv%HIW?{`Dn)9L?2FS1tPc
z`)@pQb6XxRW&TzHEh9kozP`Rd>^4AKIADv3uJjHa^SNr@#vWw~^C>m`n*Wox0n&eO
z^Vg7>n`O@N8HtTtf@7swies;*TPvQX{5fteci4%h9^sp3;+E_voJVydMP#V+liq9I
z`SVBZ801||M+@}4W8p;fuzGiZF=m$a9ooagqj7I6aQKMJd{G}?|AVzhRN_m?-xxpN
zg4P(MXmqq_{_W>1S^5*9-c{R-A>tTKL$mmx2IG6xCA5KFvB_D|sMg8=Ln$^om^H58
zY2L)T(9qQTFX3^$Tj)aW(;9iGKz*5Tn9XCeLGVr?-go=dFKGdu*VPDe9@(Dbbs4tR
z(qBm$%hy38{F|=e(|m(Nc%KZH)bYRgIC#W=*sCJ`WNuA1`xYuL1|ox1>sRYeZueN{
z)Ma(ts>^%-WH&>yFVG^aBhxO<OzIF1RsW)1D_A(WeH`nrf(}FNl!zcF368wVKxLWo
zOUO&=L*+H++S7%#6qPZ&@1#t5pi$iLi4~pHio?QLA7)qim!~jreqdleSIhgw7jUNy
z67*uENi6~uN^+s3wy{~93c8r+>50AaSZpOL7Q{)gVS~<(xZ_~wk7**~jKklY{T366
z%un|ISy7L*==$xxOMPC&)94$J;{U$1MG*>b(=7sSUp)3Uk?uamEp4a}fh|os2z+Fn
z*ZT1p2px91xM8F292|&!zXn<}fIYvz5if;}0aatqJ)}b0s++D_XTUAD1sQ%0S#oq!
ziG6{^H^r)bO>e47_5YBY--N#M?<=ez<hm6W?7E&M@hpdch+XfOCZ>PSCnDm~@@_n<
z5(q5{I;U4x&7cj-%j5BH@>(|qs5LP$afghi(#{2$pHxNr={x^xfd@AhJ}ULD*s$G_
z|5hvpU1)Kw*%!j8cel&S|AtPeB3`z%c$`N5Vp*q+f=t0aarhpublL2nGjq*h?bioT
zHW+c7A1$T<DjH7==i=;pmrXnv-TgA@$O&U$E$1tRC1N#Cr;O8{w<hzpipih$hbzXW
zM*SK8oy$iI%opu;)DslHPJFg6PU-z~UDp4mw<YLI0PFMHnhs$L*ch(`1wkS)G+FgF
zy-ehb%RuP?=<^`g0VEx+mw4@>o;|zS6~4xtQx=n|R-dEVaN27W*~J?qf8$D8l}!1h
zWp{De3z<HvKYAnD;fuQzg*uQEq9}PKB_*Y#MpFa;|A>l=>{=Qd8q)Xl1%1Jfmz|n!
zHE+&oqY8*qh%Rfb^J>A`<uF1bSyC#Bi;HKBfWtjG-R;e3gL}C)Q9s<C51g&I_ttdI
zvT3dR{afgml~dwRLfl?j<W~s3XWF?fNfs}4VQ=4dcAQ-1?{7*@miB1UugC!-urfT1
zizOUFC+gLDH#lTjqo&i)ddBK3piVZ&CjsiU&KHoq4&d3;=b&&?Z8o^64%7_0RW1jI
z?6b4#@NXBQ?sE*xiyy}#WBppIvK)OPqPxQj52C5BAI|pdHm26T>lsHHoiEa#%Clz1
zFIqm$<<J*QNrcwHjgf%NG`l$}^8<VJnAH*A-xQu(R#pb-zu5T0C)e++a&ne$nwAn|
z-F%?+_BYwH<%!eYe#(EU_P#doeXU)x&*FPFp@rMD9oe>1Qt1v37Y~KcRW->qW<HKJ
zJb4sLGsC<rE4;cv!0SJ0@~<}CszVPGTb~iel}Eory(_b2cY<-=fodq&a-k&&Mb9?+
zE@0L4Hb>lPR6iwyM%v2}ZvA3!nic558iI;+&+hMuT)*%tcog?Wx+nho>g306rjISn
z_y;My6GzRBj|Y2|hmmeU!H8Y_i-o|mYN7)-$)-10|Lq{fx9K1EN$F5$-En<m`X1xo
z>hfxBZLQ?jxtW>P0SgO@{?DAu9b(Q>paRI{_9qIt{^r1o?=vCG!%YdjCPjAJY8kSX
z?B-AB>+B3IlS?!hJ-_#qk@c07;{`_>WNElcME@dK=@rBcIs3wDZK%QJR?=^0{;X4O
zNhfQ3*V6Kt_E-IV&CR}|CYw1YA-DWdd8X4b!ZD_WSoyB;!r&vLNy@0PmGkV`tAY}&
zj~M@ll0ms~Ge6Whd17=qqucLGE&ojBj~A$aQ))`?*Ni!LGFem4iDWv>RSOMEEBhV0
zwN2YJM|}46@dB>r>NTtU^@(Q`Bm52y>od-(3wJ9x{};sKAGg@72V|57@X;?az}Te!
z9irU7s8#ke^U#h1wV0xN0+m{WgRiBjUUaB2XX;k2MH13WcXz$7{qX<~k`?MjfgkB3
zd4bc*WG8f|aBP(~p~5j5_4oP)s@0k8FCRUD<KlUA4ol~j6)VS>vz$R+pqr(5sE&30
zgWW9i?GOLOeN{#a*HDR33#fW08pr!tS_iJ6fQhmV;ZS;0S0(w?tE+rzv-F~<1_1SX
z<>~=-lGpO>(91ISHpX~5J)f&!5gy%7D2{8HwC<ncG^1Dj*DjDC8Va(M-tH_aWtfi|
zp59**ITVOUaw7Ev%;l$rsx#|dPQNojqIY*Bf13YyTHoC~SR+So_pYc(PszUPnw^`g
z>te$QKmEwF^(4^qeuh(uI<?QB0|yC)%rnYA#{|veWVzlu<T8iL$wyP01RJcGr+C``
zC8J%&iGG%h&(1`CL{kOz7(%8EDnrmn7(3AB;E$%cY^e<9az~BXtJZA$onn_7U*%Gv
ze_hnb%+BO5@V&n0k3`qY4};c~=pSQJAH0z6Wh(zK3wd*@QNJl*d}P^7L&o|YB)0=s
zGJ7JR8vE8Med~0Mp?C7k>9<v5K;5uuckO2vZyv`$j-o~(|Jkl;+3%D$d;7k<u0I?q
zs58O{7vSgO`rOfRAB4OhT2Q@qcz9S8&&$IT5j?Yxg~OGM{NOz5Tn#t_)#RKWc64=w
z6tsl=f?4nJena1lM%@(2Zq#b2=acN5X3x-5-@{Bq_UT&SJ>zKtw++uG|4}k7?JpF!
z&!2<`&v@d7FAu&42CE4m^quOlZ=U~eq?HuW(9@ExHC9VJnbjwDwp!@!>O|>Tibbdo
zkKS}F@cK-7w)UCcZ+FAf=wGD(;;9j=f4J%{+-JGi9(H|4Uxnr>`yCZI8|GQIVSoef
zXERw4+*<Ouc~QqNa)3oJ)f*GcY7~+ITKnC_UaRK7#^s|+P+NBQ=WNw|G+683v?r#+
zmH#2@rJf4g?4?fZp_JuNNljMJ#E&;8l7C1DyaE8=!6mfmeRl>lW3rHyoc&Oq@cpkh
zgz7~3n3>ZCax^a|*#XQHX%`g{p-`+@$p*lB)=!hyDkg`8W*O?6Oq%ysv+9^MYe^HW
zm#IvgY2Dsx+>^(^(Zt0$n6a$<bx;<VH}h@R-<S-w&%;x$+1vt&u9o#BLc${5oUd}P
z8yXr8MS%2j5TTolqM)Y6y#NTm<MZbS>zCcA9UE&4h@~C?4s-2{jQY9b2$hK~ZN67$
zba1OwF#5oyQO@eeWL3v%LW}0JN&@RXav0$a=;8mNm^dkk-l>chjs?zFIsBd#)`dNI
zdi%UrTiXUA`TZ9aEiH4*Ld5PrHW$NbGN$Y{dY^uZxzW$|E*nyd3G#-zx|_SZhQ>gQ
zI2@g3@?#lbvQZ%ikuj!ocX;+VS=&SEvqc=ml?k?$zc}`@+$_%^U&9N+Cwrws#9G*X
zTI;grqPP~*eMsivUkR)J&**)Vto$`y@u;5f_)R_wX8ktW925kq6u9U*pv?CN?dKJ;
ze|RMc4NXKZQ2c?5sgJgT21R58PhUZbWqp2-_haLL+J3!XUg>IUSeyncYGt$_omyHR
z|8!bqf9ADl#`IfbqtiN#N0lNib=D$mx7&{lVQm8Nqqw#2!5S@@WNdxVMU!e4y#>?`
zJ9~RF{)by2OuJ^GtJ0zn3(sWB>hGy52nx^K5$qP*^oOz)c4Np_p6Zb6%PW#jH=_<r
zLTQq+iL?fA8b|_uEB7M$TW`PedEyeZHzmtq#`v39rU$ngx>qt%M0}PWY~r=$qO^>|
zH1bFCs+AtmdDl=8u`{-X?3**e$3IIG7$m<HZQs>l{|hvt(scpoiro_oX3V0SQbIwb
zOtsD!IDA>d_%yGeabHk7j?_FoS!1{mVT~K-{a1{AWwdB;R9fbW9q*7C(N00_YCd*P
z3`{+?_h<7-+EPyQnCEcmJHt2uSVX<3W|jn>Mt5Icp;o$8|DD^N%H?ky#UdOTs=z{3
zPg*se2d-lc*n;9G`2UVe)6FA4)T@=~EXz2C;o+YN161(Srp;%jc_sKbYLlVAS6k%Q
zm*OCn_W2%q`+sRg;9LJhEE!yqUvD1TSj_8r|1acVgs*wco32|1WsIQTI!EKuWlQ`Q
zvMn9dlKE^GEf=G+i9{=9+9x8Z{%?V~W6{TA;mM{-R?S5=)@3fEJ6B3yjLEi&8h>AC
z68ztJtt^#1j*emUL|6H+o-+Rk%d*n@zY9((t3@(ALG3I3Fo)<Mte_qq*6`USkVD4j
z=&NAZzY5Ns(Sq&0UvTIZpOxS3M;*~g&i}hCl03KiCZAD2>Cm)2HznKi@&i06&O_p>
zlm8(;0n`6ZcA~!)C)!bnTA7@lY2o|1%fAa`f{x%hQPVj=7>@7Vpm9&CR)31?|H`+k
zsGH&6|0`<g@5r@Rt6jAU`O~T{xb?(#nMwz>Nn7FzH5NXeV0sWR{UW*VX4d^E`EZNl
z@7)L={?yM%Oi>`UG#W@NO%!^mf9G$~fGtw(DX+?~oI6EFTrxW>O}l{(s~WXMv|<H?
zO{CYxo0Z@>)lx$WsX9N~-Nr!^83<hf9y!<Jr%&m-D`5HuZ`t&tJtZuxFuFY!Rfg$L
zngr~BuxTPC{&%>&{fzzs`VadyY3t8J;-7rLS$S5To0pgOXVT9FP;_j3imm;9(UkXq
zBV2p%{nuNrK5W#7`OY!1Rmm1jea{%Y+B4fMRE^pNP#I{bQNPS&T+WT7u9I-H{E(Sr
zW_|sB`)-3B()H0J@m+n?jrEy$8a3)PsTebH%$dNPx=^AL@$Y^cZZ5^jla83JYy7fL
znh1Onu#kW4s{xV*fo#ds$%2WS=lK<XcUv0&ur2izS@su%pX9^$X^|)x2Gu)D*VX8+
zV_iP8Byzg^usTawv$}I2<W5&2Ct&Ts+^+$Va&oW^(^ipso-0Fm?=qa~wzv}#5`+dR
z=;)3I>Z}8(0G%<}+s9Sm{JwXSuY&#M9i_H_d;*!-Gj{T#++ZNpU|#6jvEfI7>n(86
zOTJ!F_J7#yoR_iW)eUV^0dkwlL=bT!RJy>tw!E;gFj9tHE!MmY*&5>kDxJ8m5vnvh
z#vk%Wp$Mz<+jpRkavbi~12!YkCt5s46>7B3e%j4<66<^4j-S2<NXA(?3JNF!LT%7{
zgEGc}4Nx24FqIqfJn_lo<Wfx?>jU3qj^FCWx_oZ`7+d-wYC*YsaLv(jiMb7378Pjd
zasULT0WBQF!Y~0ct>o|TZ)?k=DdV}KA4$dulWM4iAfV{je@{QkYgQ>^_Uf(CyLIb#
zdU}j*d~9)!LhU7nR69<$QRb4Rr*xBYpJXFQ;l5lK1?o%j2|ijlOvue(t_pL3h|_D$
z*newhr~JbQyMb(VU;+6Ji(vB1=GYN+hwU+OmiwRz+UB#mpUgzRvNj_oL!m2R{|;zi
z)R#!sk?@KAdLI;gcRgU<kZ6eZzR8CVAFc|2C@4`U{}vB@^HM*%m6g@tH%3%|0j~A<
z_`Acp|C6Mr5_W^!xbm&Kr}d-3tk6onl~RK6FUTxtmc_W4F;$rH{u;G_-wC0oDj%vy
z5BnN&{o<vepTG^<)L0_MIJaT;D=l4%v<OffzlYfnN<RHxwGVAn83URF&!BzRh2c8T
z&%N1T18$&yJ{h%4tKuowMRA!4{sji8lK2oyCssNRo7`oMR`@Q1{;OSI9Q2)lx(I7*
zY^<;MgQZvPIC58{4Ww^=drPgm{y;eb4SktW3KbvHZ$06*JE>n%3UgaLN~S~QVHBcl
zNlUY5=ulZvvZFG{lndywG(+CxIe*-H4!|^0ND(5=<1hfywziwdsfj~G_xF=5jnaRd
z&vU*QO{kLn7T6=(S#a=>9hnY9K<70i(wBtY%WdE(Ynb`bVx;}v5g$6v_%9ekff18n
zMky(&D!akYB8Z^^1AYard~?*2oPXVLcfa)Ab&s;Ws+MRJb%XjbK?fY2up{mAlX9<P
zE&B5VvY{$EUCx$Z2OjP878pnwnSpBt(~&N}WhN&lRideixd%Zq`D^0^<nQh6ZNn-H
zJ{e3WyTV&49gS}n623Flt+|WL-}&XBqq-ndh6;2LHlI9F$KsAidpnaOj!dbQK6Ah(
zjE{-w1~JEo4KD6e7}H5%x|7a}=kBr6e{~A0${o+GRZu<zMV!G8tF%j@3<J;5t%tlS
zRSwV(T;;_gywHIWx13Vow8a8Xy-{0+pqxia<z(3;z#Os0z?`axqAK7ge*cG-kVxH?
z*{JL*;YH<|70Wv3w|y6A5$1IsUH$zZOe!A0;70pX=!x0;fBf)l2}T!aE3v3w87>V1
z-d*qgt(QVp=K#Z^mOqtELTYIse)&(P9!_Hv#ryY$#BO|v*i7pa3U~C3kUo)-Z`1&;
zD@D{H{A;`SG^~_qT;5&)HYa@+Q{jVC0&_v=%#fOL1>c|5OSZ&e8oPSsGa0!gimZU2
zaL488dl~0)vSNfXZ6_Uzx8oGdl*plS`Q`@va{cevmPy^=_JoS|KZB@zPm0LHrAfE8
zTr~KF_5gCy`05eKGgK!B|A8}lLpj><%OTsCD1o!}VXh|XE9})^7L89tU3#P8+`@bR
zz>WLeg5B`nI~6PVM*y$}dCMQw8;@dL|3_w^{=>-7OV<w*q-TU3H^k-mN|TJq|9yi~
zG}Ic7=O5;XyThL*Z9f1_3hFz4V1$V|O+JJ+c#rrje88uoJ>_qFuwXLYQ_P~jVNH77
z#pYJ(NfQ}DTbd17Mu9})d$TqrH?{lz{9&H<hXa+0)RJwe08%Mf8@Q^(>DQg1@OMFg
zd_~Di+D)ha6fT6tzC5=ehxp0=Vy2Isj&b5a*5Gm1mr}u{Ne^|@8d!%F^rntN@TdA1
zZTW)phP$`(J~irQ&MocyVtX;4c6r~W&T+2OLT_Fv8H;cz7Gmionsh+Bak(WA`f>s`
z9JOzw)^V($mB3CE<ydy}vilL^kH)0f;1D>Pu>0tk?Rca4$$Jht3zoe{soAK=h6*t%
zybaJ=H6cw`H{5hB_*wC2@_=eHW<VgEe6y_vwnil7UUUCFQyW>TTCC_FKs(k$k^6z=
z@0r3CV|^n<E&zD&F5f28qPuJIUkKgZ<H%XZJF#lY?yN#*%6Gw48xHPcEVO`_u$#>a
zS2pH<ZS0*(?rr>~jNCo!%PwIzY&2<)8&XqLN_3)v|KASb5>0tlyz=^(9lfIANKRLR
z8aFDDboS?e$%LMw0`pjw_2T=m?0?5Cg;;XVSG`~mH~$*)Wn`W9-)ZK5%~ZZ6J*m{(
z;}cB4oNyC$$x(aWb34;(M<rF0w^FNO$GJ*M?O$5st6b7f*Y3CcLI{$g_wOSABauh?
zUH)JlP&JeaSqI?S{d@09<&sHgXVo3v8iQ;5x|XGnIgsgI%1}AApfTbNG#3?oQ5WRW
zjJx%3wd)nDgyvcJ81W<>-0<8D8&9?UKWh719Vze6@nQm{6sFFvS>*q2zP@vR@!re$
z4kuAyF8bfa|GV`?mjTyRN_LDJ;j0%(w%qRj-!)vR=7}5(T+N>%-#H^6f1b?tKW<Zr
zYO`XkKQgPd+~hB;EGeb=q4OvoD1lhoubV?_2mezQ!L{bH9J;09=aIQ3)@|tWFcl9~
z@kmNR*n%?fm%&BfysNkz*Sse$`#W)me(rzu(2=RTCh}Xs@c6aD!GD{=#*J#^_|p9~
zkMG(xmlvtnOiNTQGs)11+UrAoA@ykEtqzlsk&%cOJw0gRM!kJ}{-HH<>E{On^3HZW
zR75^_grNI+1^)_j;TV{hM9ur#+a|1fFjIp1Qj-@67?u$hZzZZ|zkTP2dMJTvqA9=+
zvYZ2qxfv8D^%JVAs{;ft<$s7sX@Q<x+GFW&_-Ki*DBt~Yt?d{$R3mr>@-QgOE@tCE
zNwWO~CcvPe<rij<^0-8ZkhXv+1k`P?PQcsFbXHDi{8H6@XX8$hZ3b1Ute+S=($&Mm
z@n;e33E^!6&_E7`!Fy3JllbB|ZG@Bar`Frrn25c#k%thY*LGV%Fkw{FNC^pGcl~P{
zb_gD7R97uxNa!V3sAXKf*ACy#!=eH0%)#>eG>xahPl4uoE%Afe%KrnX$X<DL6{r72
z>5+;Ko2x&sQH+&bz54C7cuVk!VWx5<@KPVz<j)Uh+b+wCV1<KHOTe<>uvh-VAKKTy
zm&`hs%(^t~a7l#vowaPs7MwHWG()MizKpQc)YM45k05`85@deR0Qkes&dy{zd?KO%
zJBxC;=ue+M6?HL)WNvM_64XuGF?DURQhMc15@{h)04|<<tRmS#1}qs=@e2D7o<afQ
zfSPZJD=b`1whFAZ?n;DdzGPm#1-K7fx0Y5{9q!$`c8g0*?Mu6ein4MUe>Gi3u)=l2
z2RfOa>O@rQ+qVj8u2HTxu9OV12X@uklYVjY!6XEAGlPh2yvHwq!lBknNlYijp}W&b
zgqP_xO>42#|C=SbWq#}uW&wKebvkjB>a@|@4%BNv7z!t_A`oS|IT}IV_NH4g<+K!7
zVpQAA5`O>wJ)KtEl?gJd$+~Q^=y&Z8ncw!6EoSF>yW{28-q-0N_IZsK4mHMbsa0CQ
zXHH6fLje|+U{s!frit&R>WSfKQ~AB~nOy>!bSO|QKuI3}a{5{sd&Cf&G^s}43oMiZ
z%p`);aB8qQFjGa$Qa(YPij$LbqQ(ug#p<KIfqy)e2=MIZuow@~^h@;K_do-Ym6_Re
zc3?K>_7BEZI@sqitW@{2AL}vd6hhKKleD)y_!`$1B#Un%Zfa?HdU=Hp#v4(NTU#>;
zTVMLONkqG><hRX|{(d7iHunAtI~X_?2mpc5z)%~=%^=|$0+w=H?r07M2LG+z+Tzrv
zI_!9WN~5Ut=V=iD2ZF>lV7es)WcR027w5@u-#%;a>+8e7#xHZCRk#G7!zAWt3_BZy
zvvZr9fAbL4eOd41#=y)IJi#SOEBWhYN_xfYNN=YQbh?>bv+rB-+_BWV=rrjdz~knA
z4^4wQf9}={ND)v9^n$nYL6fka1_$-B`d`UGTZxibGNk%)89wT}s$?P^Qyhof|1XHM
zN2&M!oXn!nF~2e<no<=G5UHBY1eFSx!_&7v<UK~Fo)D;4F8rsR;!O?YqPjUY-NAk9
zU#$9bY&<v3tzR~L74ux&{2OEjNK-(}+z68@jTW7h?ukTQm$@ZN?&x>iQD+tsVncn4
z6n_q%-1zUEkMHVMih8Zee9l$-+aXHUfwU1<K4V8P6_=+`sbZz11t?DaRaD{<9%}tk
z$IbP1eTe1vUp`UoPn`)xLebf{MRP&MJnP2YSO0x3D|<j<)!8hI(L2;~W0+XA>wk^>
zFcTQ-L?F2#16l09?~e#2Z+DmW-0lcL`@gs92$x#8h5;`B5F6TuAiuNv4)XuxxazLN
zTV#;gT0f{$J=#W{TW99mM5~5he<}m7TAyT%$J|H=4zWjZc5Ok9{r2@~eLj)B=6e4U
zZ{q?LfC!Vn3Ljp2Qf-%Ov7mHMV89bqYJnCk@pN{7^VKhA<Q-Xozjux2%gYu!KyjjT
z8*2)BHuRH`-wHix(Ulw>QC)Iz{#O-MRp(!2ttkQ^LKvyEdoF51K|@n>?+dX*`BHL_
z;xq#!sK=&54Uh4VWfc|I&M?fQ2SM`(qcE^<LA?vlg*JY!0<#j)k&z#W&7hWCj{|_J
zq~_k8wNz5XcvC<<U<yC==VfGNrM=c{wWw<A>Y6UjPkvhWgVIv|c!}Mi+!pXv&m>=p
z>a$0iYMQv==$JTQkMS$zYo1yh6L>dNaYaMcTAP`<qeb=6dEyuj!0;ZikUf8#sgA$j
z3iYB7su)1M@Myw`>62Pae%=4y6)B(-clJNYh(p>v|0^B^@R)#5>8a`Xe{T>&$oTGc
z{uV@OTK_*^ta_fy135rX6A|iz!~sfl@iI0L{1z%llKr&S0kmH8CEtB5svm$^fd2;k
z707pBjzajj3ep8g35_z$yiX6^KHTkiW_b-i4>UbL^;K<^Z8hMnfK|+F6m_wAs6y0-
zv+HYWnWGJmw&`K;#8C3~t|={b3BC%lb@3JySb>^eOZmS2=TEh8^y%s8##D1!fA#h2
z*9(X^NzSeH=e2JRMRbC2tky;coDcr0;Q!Cs=KoRG{-Szkdz%)@0<Z~55->?jNwL3f
z6ie_fz61j8jb+f)(nxvO0t<*5B5ZVso0XVb9;sN2AZ2Hg@lI#r2XlvVIIl=@?st+8
zRo|U!7*1c)jgNIn;10S1AT7u<IPvF!CrBUz%0c!z7byIbO8RY+1TCUgo@y+uuV3kv
zhe@SZ&A@)12HeSnXLLQ<I0c&5&5v6vWo~tqKY)^i1o7ZIjrl1kXcM15&)I{CoBicM
z8!&KzZY-G$7Gm@93cocFwpGVggLs8tBXM~=NdFULQ<HeJnum+A4Y@F#s%<%_9l@h<
zBq-H+F=CUJpOM=|Nl(8zJ&k}VtbD!qUkC4Wz;O3X<8AO2*={&b#XbyH;`q&P-DwpK
zBZ6z|>nNj`*Vk+w%zd5_&)$Q6_Zn_Z;F<3-c9yF%h+%DrwfRO)$;?~dN<KtU)2`Dk
zpx4!`W*n`w)7zYEd`d`*VZ+0TgN==v3Eo13n5(66{0H09Rfsw8aE3e0-+g0T@};<-
zcP&79trI@p-v7)8a=!im3Gim~Mc~5}P)vL!1854qFHUM<W8?g%nTxv1j5(-&;joEt
zV*#?VH`s5u?m1rLW)*bi2|g`-l{xL8RWL@^N8C|?lI#`bA#)m3Lc;{`jQ8ChY53#2
zRkUg>LuAzjit+}nzwYk@b~V^I4U!fE^q4l?DL?Y{xXmSR7MfIWIjBKqfaUQR6W6^G
z<?f=oW9iyIQRg}K;U2EwLj7iw(tOD3y=Cw*QQM5G+6CcE(8+qWzy+#b9?Hi&ZA*J`
z#q4KqZ`^(}!WuN@Fducgloh9-#^++5j{PF}&6_dkq`$r7r>3K`=XFLeNr(NUtK|oE
zl;#DLW>WAJhm?bFLd_^n{_0iI$g5Yc=Goc$@{Ai6n|95{o>KBr`ep(~n#8N8`Yx7<
zkDZ_75rTi3!T7<eFw>{XBAnKpQ$`APb#*8)?pwWLQ&AsZsDl+qi^`xP`%KpM>C-ft
z4&Dsf8&L<Gpq_z#gVv5#{~GDac3_(pvQ+ZKT&+-|ClJm<ZKKVm+kn;!1NkH10?>k4
zMr0gJ?z&S(!GtCd@+;)Noa;2&S5j_tE!KxkdhvGXcfs}HN{7c5q5F;wC~UPUBA3kj
z`hG0}zK}ls%0+EmQS{!j0=xb#gOSi%%EW}A!(SPzF4f`BPTvRToclrz;a6ZD;7(*-
zOlR2JBWKcMZEQsPnCP5A4fbQQH4=fxhGCb{IIu_sNM(hNt@a-W*1UrHX|^px>1kh1
zxw4DPCRBkHM8;AcOGpQYZv!y@7!7h~Qf6%(lsKly`~1`gV%9X-kMSBPC^($(W6<!=
zM?!-$U`K_E76TRtL;d%kE_tl2t!<&p_Tw`9v+UbK95cMF6$Uq%3pX63K6@vLlz`oc
zS7L|~T}_YYdc_`NmU&%$!dVV~@9rhSX<ZdD#O0;X055lU#l2Pr7B==8i6r9p)ld<=
zd_lpIVNvh3cI}lD2&mU~0r!Hv9(S4h_PZjpYC(EFJVHVNolLAxZ2qbiFHh&%Bgsh<
z^J?9?UtPTic;o;JRJ@!9qULG5i1ak<atyNUtCkdf0G$1@d2s`zi7h{P(jlB^Oe{gk
zry~N=e^PeU8x1K5luMB$owJx4B=bB(BLEIwk~FwPxMY8an$68_)o)s)?v`Bxsg+im
z93%7RXSBO8r&ghiMT!7>F>|7h$K?u|&5t&j<k~K?puK1c<$mh-^h((cI)0pQ`%C?p
z%p_+SV=sln(E4TbtM0=z55^yJMLMj4w7}zhz}lrWRJ4>LDXy667udOebIfDrlmO?V
zq)W^gg$ANYnP@n!b_r}M9N7s)XvrFe99&$VqF{uMCiJo{VJr@3X$dH0mBi%$s|9BF
zLoa~c2o7)!V_`uf)=qJMGWbTur(8n=f6}`%p7oji91~6aXN+H<aSJ>cvC%yT9IE_l
zqylD!lqS|W>nnqYvvB8M__I5t=im8WV#>NQQu}ZbZY|hZ5PKsD5Pi3u#%mJpD|}NN
z@$vDosC8GF+z2jp-z-){Dm0YcX~TI!86hxd*b);2UELu9X55EVc^R*sK(F@$hp@)m
zd<(`gsDvyX@7)W#e8c0>qh!B0Yyzs_KU*2M)8v{j&g);q3Le6QD6DUMjB3*qEgJbi
znKt@d5aY2?L`6j<KKcj(R#1UT-qZK?<h>S3I%M{vN<dy--k7!6R*~Tk7+9G`nJI98
zv!{!vU+?JeCyj(9q~^*uZ&GPXTE_wOVcX^T^%9^?68{*oBFt-CGoXJ6H{4qvGjhg$
zbpRR9Rv5{_>jsg{){qYiCpZKqGg6dbMg}#JOGPqMF(>Q&4woA(Y5FHf6}kUh*v8rO
za4ozxDPm+`_`Ve$Fh4zwmPIq6{y}mmQ-wfZhl)u3&1aaIU~pFMjv?Uk2bZj!CRMu4
zau(=D42n%(Ha>O+WHDl4uIy}W<QfHMr+JhR2|dJo)9w~DGUJc-FF)b8&${!I2$%H`
zRF&xn<@*`5ArdfX=0bB&!z;PP4t!=+Mg9%ZF|UJv<Q!rUV{QE2p8ay<K<q3{_YLgc
zzC4Iw6=A5wg_;HDkQ*j1jrcpL(M%uk`b*zq%>|%TqtIEO2Q6-VIioZo_eek!+WF<7
zd~&(*1|RZQd~vwV=f@8IP!!~eutfGz-e26@EHy=<;jplXI$_rtwNop|wa@Y>+bNI{
zX%Lm_iell^(&G-HEUAJb%Adzwl?ysQ1tSc1ESQ#hNB8-rs{xR>@-Uq&{VXmnIZn@M
zvVnHHNX%(FQpP&Zs##7-pq&NgRD<V0r$lm8Aalb;r$Ee!lHWcByVyCg5FmV{q|<0w
zmB_6pnaZ8PT$sbCK^lk?WqZ(io71)GaU;MOL}lffa5+XL;?airPglrvU1Y74o$R%~
zo)~S=B#x}fOL6JfLXJ#=*(ao!erS`iDaG9EET#}^5Du%Ah|{QuhwRSBatMIyKMJW&
zo_UyaWYKwFw)?mX-c}<L*~P_9U~Oj1#`b6rqaezf=!>pKl0f7}BqL-AgIHB6j+$-N
z7&^N_LsCqf?1VJ1vU;I)6r$4%?f^;9jz=r%|6miuMkuSQKCs8uMh_q<qY7_f5MfRt
zu8ijr{DDVk3diVQQyGQY6EJd*)&pvTPZl={jc4|<Bfp)J9UsFM&df^YX4^<&ihTv*
zH3)@UGf*_B*05%fj#AG)hF<fXjF0Y0><_WMViH%n>F#ut^USj+0*Yvo%lBIqM2C2)
z7$(}<Fqz)$UtjWS#Syp-DhiW7QODsJDndmMkYIbpmO}Za<^nd?GgF-mMQpLx#Ae8T
zP$iB`<K4=^iKb{l`_i#sNh*YLL)&p-_TNG$i0vEuBwnuA;JeRvm*{he{OOiblbUJ-
zF_cAZRJ`hnq%ZgkqCWbzPeXfa4eFHJHt}dt3N9Uraj#y{w1-0t+JYK57jgJ`la5#S
zJxTosPq_$osLMHn6b<JS?M)RG6+xVsT~-2;kRm@KNDVNlxT+5BS(ad@2SF0-2opw=
z?0JIGF|q_L8;su^JbO1D1tVJOhGktIsuZ?zhrzjzaX5<w4KE`)Cu*S08JAdn){`c0
zm8D9w@VOauz;+5XAXae{q1+(8B^2~jccutxnYN#H->~Z*#0R$&TQ>TZX}&o)WE#ib
z{RPED;yckps8!JLws4RQsJSPo0Qp?3rXPc*|M&5f_iT)Rkm0aDD}#a_*q5>T3)2yp
zT@k<zvt<MuLfg^Gd=Ux4FO%ulF-Zi3g^bG-WW$4T`}WzX&>sHAeB*xqEY^;H&VwjA
zKGegXKS)a2>aj6SILr6zn5|EtkYL#^ENxO0br>n<Y#0ZZq6lMkeCC#4d4Y$9O%%+5
zm)|iQq3mgzv)Ua^FzQb5mlflBn=rVi%b>P~Jf0-RADrw5HJgq-n$aPVUF0a#pYgO-
ztQJl3VN&~ZUF0>T(195hn^H~}yEo|Q%jf5!(`0Y+AKH7n7!J-k1VTCAVNWL$>AS!l
z8WyHwHvJo)N+9C$0KUI6u4=IY2fs;c-kXjv31zmX(AP%}q$N=%;7iHskGx%u<{dx4
zqqT*EM&QK`%visaROtICQnWHZ&nemB9~8(?0w#2y63rBz+bpw)o3jei+ddxU`N)W+
z{oP`y$V87WRDmYc7}JtCSmbVdg#VPv7Y5eT>(`CjbH3*E+XfTaQiK~33xx#gZ9k^y
z-9<c?Rdylu<>ZG8d9=W4uLn+SAjJVk63S&BIrG{XrQgB$*@Y`t#g$qa%(H(>8rEDI
zx=FRo8{QEw$;sKlR=~W`P+eWE@$R+{=S+&uAfW?Ap<Pr0n`XY9Nd?B!8!rycpJ-@l
zVT)m5VKpSui;GwpgyIJ4gzuifbe*a!nytoYO8Ebm7wP|F!}kA|ckbvsf=&d8W8jcW
z&hdLTS5|ys`7+4ZxVj>GXsM}fq)=|A30IN{pFMll-Ua<gK3jsjd9t8|;@F(eJI3Fb
z#ImxoR{Jux?0Gr1m#D1kd7^3-1Zen;06xBz7$1+68iU%fyDLc)h9iE@f#IfpntExr
zsKZZi)DwQA#qwnzdNfKN?P&lhbTu4P+z*VL$7@2YT3zO+C8r>fK{U(F%}wbE_lr$y
z^4+(09&d!=+%RAFJ~W2<jqqWpOl6IeSl|I#Rvk~yp8?yyv5BGhJ}YSUJleT`Is;v-
z@Ol>b^?g@%5E3fQ5QZUMd)U<n2M492>-eF;%unp1JYrPC9N??nk~R@N<4^*@!k-OW
zehcUS#RXK@JmUCtdD^)*Uqip4o66moD0hB%PyZWq?1c<8e1=8fe<=A9^RH>%CpcxB
zlxs#I=%!8nl089S;Z-ohvHOInN&!dX=FRWa$d=2mgU+EHqiH;ZLqTzmrw#-#rfCI8
z1m~BKJ8&ivZLMLr*ScR9wxD$bhpnBV2*~Jr``hIn$Y5%aWzR|Ulm}-32MeaB;YMAL
zkP*?*{3hjdP^2q}zTGQm*@8Y%h3>}JEHoS)iftz<)eHB_G$`%Jb1w5&@82(?JYm}*
zx`MZ?a8Sr~lE={B|0p)27QZqTf-Xh{v*6Ef%VpT#{z`<BoB159y&m^`)k#Z1!SRGc
zl`@HIxAfY36g7VgfTknPPZ0!+N67>niwt&QtAl%@#2C1WI_#O%(2dL5#Sda3IJmiO
zeAC>Xrpfr+Jtn?PFGeU2w#7;XbJ-w`gi3HsnF$;5io{;n-cDCHQL-gIK0f$+c=r{8
zmMt*z;-9XCIDTWQIjEV9u{k38yD(zfWoAiOxKA63fcCWBRO#WR>M}-RwG`Q_=!AzV
zbC3DI?R49$K%RkedfF*(;ki2r7|mezpRnA$KkLUb2LX_8VH(Dv(Gu0@Pk%=XMNtdI
zI_5=%;SdS09TGFi972;^t_K~l4WM42*gV{YnFfom*I>hBm(522<=nQYXlEZ^f=cuA
zju6Z5)IZ=huXqq6vkl8RRm$^jvsu1w4piTn&0<(MdmT^*5C}e-7ceNGXy9SEk^5)|
z6~6<;#l=;mrmeyH7LCd46ZlqwNADmHVjSEIpiErbxL%_PD?;xrk<I{2t-2X>2Gb9K
zgP^>8GZZy#mK}U*V)8rS-KxXMJ<o(l4^x^@+RSa)0gK17lxGMpK$Ky0JFl>lmxx}Y
z1qK+#vee?BE&5#sa2BNLuopHkl6|Qw0jwym=C9v=Beo89O~=rx$H#KP18HuhDjHi<
zIhy?ZI-R?<x|47OG<$2uOd%Pk#tnHXfU+bJ?d|Q)m}KdUI(pC^S=XIBpp)|8*~vPh
zcfT~si^1j}ebQHKChpp#gA68D60)ka)$!Wi3WP>q=j|0SF;t7+-cj);Za0|06&Lv=
zl=;_bXKC}fhG^Ogy%sdnJj+I@od!S>!7k^G<iiJm&AEY9jJJ<iHi#MsWQFMtv(=x2
z({XEFI<PZ4j=+m3MWgw#kX@`-PjTIRg;qLH|Hx(T6y_;%9Ct(sTX2$yx*sam8J3uL
zxwAETt&c`IsIq@b11!%<0~(;%m>8l!I-Ad%7&oR<`u3pM>3sV7hT~8kDZ_B4O7!$i
z*NCd5e5K$SSH|#fg-G6d$p`^}B(RG@ZIDZBxH$1R2sA>puNcT_{1K-wHb?PX_3uzn
zsy&2>klpF;63V|p6AKO~N*rEGeIb03ynljQU5PN`>!BIG(SEV?tWRs!%)`m0VH<}m
z2TvM@aI@(|@D$Je)4dj#7eZE+w`yTz{*nUCjHVMa>*qoYOoyTGcIjDa-#%RsfRk2@
z>m6{Pc&&y}Qt=%?)dznvVBZhKjgF0JwxB72@__LWD_y!l8O9D0E-?rPxTxq9nb!)B
z$NJ|Q1dh!pOk=tw*bbq*k}2_#C{vzs#<zr@#6^N6sK%-0z=0_Bt?Vn8>sMPXX3uA3
zTsvw^IrtjL%djG$9L-L-{goGqbQOH>yQeH_KctWu9UTo0E&4aepKNgu4)vQ<yl8J%
z9&+7_+<7Y8^H#F$90)$-S6huwH<3#NIRv`kTEvqRyi1!0KAUonb2^`@&KSfWc8gEF
zs)+&k_}=sP;j<;KYBisrj_+~_zIfe5hDc{^i=6R=ox^>0<s7S<;c94H+-1!V6%}n(
zp5Q<&XCgHQ13bS*;8y-9G9lmSIL$W$(Jvoayk|!cr)Ta08fG31FTisddrO?YSFwnT
zlK6r{j;QQvb|4JrZC$#^u@pVgBWr4Ea`d1m!FjqAN?G`ffq~)G58bF`Uc7Kz-9>uP
z2ix{~9LMeVpmC^3HSk1hXWM*m6VFz?g02L1v$pu*fRn1sM9p9VmTB@8u)}SmOZ?gN
z5xPjy*Q|)?r+|@#Np!^Dxsfe^hRMfqEL~~Uk(moYhcSq=o8U%E@PS~j80<F=jj#uf
z@eIv*hm=&Ji?HL{I#@B#RqBx&Iv6<q(N?@VnUC*MV|(MZ(*d@D*yf-xYeD1AzmX>4
z#x?+3o&UD->VvS?WC#{p;kxtIpFgQk_M?eYJjx0?C}j^4k<}6`!DBS6l>oDHgdgvJ
zquLG0K$)>%ZUTrWf=_rUhjMn#QG^A7KzwTn?im)<g7_FwJ`%w*8X9?U@|oTxef5Q~
zH9jF(+qLeulH`$FE?gz+0N-?80~lE480EVZn<RL<{m3LMiEb0@G#+@IwLz|*V+t4b
z>0oDtWG9YKNV4%#XZSiBc!L057w&Shzxw_LZe>M=q$DD}gV*o^Wq?SB_u}w+lVOM*
zKtZ+mgut>1zg0?LBfz6^o@$Cws;Q^Px!b)3{RZvCXVC7^SnPoJ1%0-kj}Py_$g5&3
zLI{SRSrsKaHp}^6v=Fl59Co|1k`uTRn#sO-C>eAv^W$OSQ40d5`X64rY_NZg6FxU*
zz31VI9d#{B=z%63MGI1bsps=l-@Xkk4Ta+KBZ!-G-{QeHrLXta8!@6((XkIG#G#Xx
zocta!g2hWa+XZ--Je%jyM3>n@wX-X_FxozkmEBPSYe=;iX5E?yfRowJAT!(z!vjaY
z)&Mbo{I<`ZKl32UjtFYx3Qn`vFx!B^i!s~*{4E=1igk$zFU?SnzyMsxo>kz{6UJL$
z;`|h1p(vditPiy2dAB-_y=-@+tCyD`QkU!CF;F2q!Yvm(18z<?#K+MeK#p)n29Jqs
z*E{-B2vV0^{?ocO4AWpPLyKjDY&BWFw}hE#$}7TI1q2A6g3ZI#Q>7|@9dyF*XWb2t
z3kw~RLeN9LFNipelw)&%wF0>hlku0&e}L!T1K|3jCi&`Z&OGlQU--6_Zh*0UZKOj*
zb+y_%5VrCRKva*moOIm_i>bWP(lY9;diSb|qoaTs!JWfZ9@4CqmRoNs>L(H{&5+2K
z{@-6+jq*sB7$DsOwGaweQJv_Aj@X@w=2v_u7U8oC=nacIy})u<5>eLG4$G$Q0SkFF
zI*?M^)nH~Vw!%U5$PNdyGh_z%`}UcYAB8dJB_efZ!0bb4*U)Qw9m`+tsTTzfco3CC
z_BSEhTsu`dpip2!3!BMQQ$IgHUrGhAHVS#2Dg0WNJ0;Dl>V(^~8t=N7%I*L-FxS!t
zN<#(|x~^)0WoG^9w?wU3Qv1Sv!jO<FT0uG7<LX$1AQXshC88(xW6_z6>PE+rU}Hcb
z2nm3;xxL1~{Dhuf(;u!~=V7P97N|o^2-C-3X26q*as(9|ad80t>X3;(y33RcUVEx-
zSZx*yza+A;2&epaWA1*eKoL}5U9~V|#OT)mnK^Vq3twKRyOGU+g@a<Jme47C^(qWn
zP0(<>0^8=>!7o-!9n9k~eAvuRxN*b#`m_(QJcu#Epx^0#C|qD9@T#5jLnC|kXVgUE
zED6((Jx1i}cghkpbOJwn19ZYmvePnm>hBOLqIPOaqvNFcb}UWRVhez-R(LchFak0X
z7Tth3D}9@FJ^m}4Z~&mF`1Cu=?pCmhVi8~1Y|2bKG)ziZR)}4%tQx(N^PrgG;~ztn
zn9O22gDCh_U|~>9dRH@)B=Qw0zoDZdi85LFpTOZW`9JTxPtG^BN;uq~|Cu}ugs~9u
z#djx@e|&l+dyn1~3t$)BeHjuG7(2RvDwUJ@c%lp3Ut*c1<(Z!)xRDkPlCL<~R(LGy
zwk?yx%K2f&*(xR8Nl8D&=?<nL;q1!5zubQK@^M<4RSGeF=uLdW7i0TBc@m)5NDxKF
zb3|%5<6J$DX;HZ1(Uyv9EMGuIv&5P&5VY5?Hgbs&)B!S*Y>mQl;QK0TRp2A*a}$p+
zHa2$eeN0Tuwh|B5?RPi(AN0J<CwwdY$T{UX6L`=ng#Zqz)bmsrn^4~*O_U_;F+8K8
zT=)hhiP9AKo9Z^+HZK6w@e-m_YZ+<^KW}gCtgvwmgAE{jeAy9YPB%fCa`{+iz7~Ok
zSO$e(<3>kB{B@{~Hd`)-2If=dhj}KMUcGsTnxcX3?idCYKu?GnBv^#<)2Bv(+j;3Q
z>ymgSBAzkz`+=dM;hxftF6rGjm*r36&nn%SajN$60U_Brb&YX;N$X?kgF*9gG;iYi
z=>eeRdSUZmSO@|Vj`6!^F}^KjE>m(!%|zF8ad5sS0iMNUT*lHbJbwTS`T~Y}h7xl^
z<FS!_OJ3&hNeX^qVq)g6(MzhY#QgQF`x(t6ezlVn6qoP1?X_<`<8V<e2`;eXGu+E6
zdb8ma9HRVrrB)Q@V9)(hR4jw!Cq)>`UmkuP9Ul@hdfn0K5b)_al^0m7Fb4Gw{2PI%
zGKm0>%a5I%@<OJSOOdz0z>$l()yuKZ&$FocJ{X5}9%<(WvuZ&j6#bI4sta*nv0&L)
zjdp$uas=Pz^XsTi=uVC*mB9-~+hE};Ql!MA4TiJ4(<?qNI7CEVbd$5-8o-Qd!A%sw
zLO1cXPFPT|1h5vzOfJc4ARTpCGe>{AzE|^oP-Xwd{|1ms#ujfA*0q-G2Hy@TlYH{B
zC<N^cG$2c&P#E6Xnr`)NQMjdUDX7KtEHEV?+5>EcBvX=;lN$&)Bl{u_W3)VfN*8%S
zldvb&X~F(nNfM8cZepXzy_J<zo<YVt-ieQ`m!VSv@oQ}Cn;$)$eBf_bt9Q*9;_F#l
zK1*%EZI}X!ep%$NDkpl`;?u06=UZ1oLY@gbo>xp_K46i(EK3=Je>BM?>$?;que%Ca
zC|ve(qopsQkCE|wkJujBuXipxmLK;pma9C9t4n~;m483gk)pvf3)4dv10KpF#<2@4
zD^k!{LwwV@Wm$#{nK_spIuPHx2BxMtu5A?+F5@-Z$+z}MB}slpey|(llKcb=))JhN
zb?%O$^CG~z*?t?NA2o@&AH0I6jsPwXgF@=&jHIO7ATf$!7ymxL(N07!M#7@gc=P5h
zH^{s^qtF=bGsOiae?12^SeK={5Gc0TUkh!%v{r`fDx{bDY-VE*95VEBZ?euUO*PBQ
zV(O58<XU=9^ysZz;3=x101o8V0(3Pg(^IZr!AXj(5jQgx0Lpkgx{182Gcg17eVQyb
zcWZ>OuxLrgYNlZ~s{STP*JbdrxmK4hPxJ@Do$2c8nr>d__s3b$<m+2~ab9jEU0D^C
z7I3;NyQE!Ho~D&W-<@sh0vZFmKozPZ&M2gRdRofLUL`zeC4E5Feu&3ilx_mw;W%@$
zXaP$34M@qbI`>G!ymTmF9-1GUTy&TBElKC$DyNm&M}KgEB7ADtWe^CNaqq^Q7uXRh
zl6GDjl-)N)_rL{n^t*gJm-Xwwl(_<i*L`m_)+eyeH7?ITPa`BC$Z{bSTPnuMc?1rV
zsI$s9C)~V(yn>K6507Wz4tilubY=^Q??p&E67b|}%A%LPQr6bj>j|zyq+trW?ATFV
z-uN8z?%3DmZi82H0lvV%?Qy-o9$&Br6@j8}?UkD`^FXZYK5*{Od~%=WMou}X)M@2g
z1S{O#-9ey`Z~<%$>zlm%ab7_wsT5Ok7a0kf(3>3gulVaP!81~NJ`?48zWp*qD0%?>
zuo*mzU~v8O+Vph#+Cz9Mi`W?E3fydWq_(y;@S23;zd~3HBCn9}7!(Wv={^7?vKWq0
zjt1p@PTQaO1TZ#ZTQ^G&MAK*D_qO0BBHjH|#COkjKGPio#}}t@qJfp?%Mb_#z<mvP
z&{TKcHv>0QuHM6$e5P70_$=!E@+bwW-$}mR6OwxKi{&}A<}cmY*w~h63B_N8u^8}j
zaXo<AM!(e)REa=f0Fq=??^QghtK1op*SNJ>+DuG?X!-m(7iRa!xo-AJn_`5zTNss?
zzlLN0V+whq6M!|TKj&K3;t9RU4agNBEzcdCK#10O3Q7!LZ|~`k3M5$9|9|vLo7ff(
zAHcR-yX@CV1il*^V@10gCvv>bLK%jko{f<drc!o^igzDM9q^wK?T!Qg|0@D90=;bR
zYgC4fV1{CrsFKOU#li7v;$E}&b|$o;c7e)x%Bm*DteMYn)o`Pcq`~MdXh<{!8VwpD
zd>A_Xfrf*IiTc}NqG&Y28|-LGg2RW2T*4-^)Z9L5>cef>dLfC0&2xu0EB!c}eR;qQ
z<J`^cbUz$%l7iSR*y_e!Z{A*f!unqC!t>#*R2|1<Y4&z^wD6n!0_qj5W(>=$9$vG!
zX;ew$WSeRlxjf2fIB;fvzgEeYjnJEfGnF9mK%P9`@OV8}Vn0#AQ?GZ{-@&HoJh16E
zQ^_mFKRro~hsiDkkEB1nEyy;<4Y$6&`^}TJ`_I%e&+R!Jq{2S_{|;Pi62YPK@QS|F
z)=hK2+i$&0XjTL(Wj_s<Ek-+jv$e2lNvas_>Nlcen0~|jJ3oKy1WSId60y-xb?weE
zrPCBf>u2Wbj`>HG0mOSQ(Q!+M0{c__)S0V;*>7eZ7&6ZbA{uVSn=&lW+U`sB44%kP
z@~rPuHaD`om`q6FxmJqLd$)>`4GmKuR0S@n(j<C#uy@y4&fX@b+h@v*DfOAzFn_nK
zGt8W-hx-gZ^wfdVymx-9gp5`MO_D#7C5=1QOjgW(sxwfCrG;@(MRkW*iE7|OQdknz
zQcy4Q%vVNA<=Se2wrimncI27jb2LKNiLy&A!6Fz4n&&|ca|=SnHSDo+{Mrtca|f0A
z>1(%2BFU#NvJv}<Wh*5$y+#cr9I`Pl6W#5JM1ooM*qCfx6*LL1T;84OT2`$mV>1{_
z#;1IdQj1M^J#y5aCofKPy6z6FmRiAH49p|zO&+^Na<^m3ncb|0A3Y+!T70Q(W<SMA
z3!S=|$!y(hpcL%LL%w4TW+l%r?cbB-)z%I4kB`&vQJRZa*3z6PQ<-l(&9Ys-LZAZ~
zKV(>-iC6e;Mo`Wp*B7h)_w?I|{+*0@GGLqtOKHaK_xy@}Moo{e<0>9mFeJ1MX?|LG
z`nv<kD{PVji&nVg;~!;>5qIg@McP%7h*=Rvi;o|@UN)ZYRxjCfQXs~oU#v*vW@5<N
z=g8&{t5mtIZWP}C$ZIYxfPr8cHW5QaBrX`%)EpZQlwDK&(b&MmI#jlZn@W{5Nj7ns
zlZ^M^+V3JhWP06XclSC})eK+2*osM6U4L<5>=UC;LgX90d(Dlpw0UH`JO1I%FvE;q
zwfk1nG0(|b<^;~eV*rke;%_Hwa+{M3Ar#U7?yFAEBwWHd^G)R&LzQcOx_$Jv>GFOI
zi`0{SWQxX#wRCw*NEVzt*s!1&gWrF^^Jivnziew!7)ei!u9{>Hk{f)kc~?@aLTPj}
zaOFGqj){OWEjrE=P5!B%QxAu$=v&N()2}<JLIUSmBSyLjhZS>#dM~~FV~;zV!u&gV
zigZ>&zx3~V`AXhGLL=l9z(Fj{uK(*SD1IwMR{qpuhl;sh`m17YaUm`Zh^&>LSY%1>
zqb7X7zc$V{xb>%+G>wkkW#-hxm!-Je_B@Ma1u_KictF(M6dCINycQ6;IQZfGZdp6m
z)Va6h*uv0&i&k^uXh6oP(S4zyg_n<4DB;e<t5V?+&K=8Gs$VO$r<UQfu-Xqe{q^uu
z?3c`-hR4MjTJM4LPm}VDvbcce2k<!qJcM$NZIS1dM<?%1n%Bd4@}D_Jhx=G6QsMX0
zhuk^!S$F(?`Oog>M{C&{rxp`b%olQ<;YKe6Y$I+jIX|g(m$xEkQq!F2O36AO00DAK
zsF`G@bB(hx7E*!Egb<<aSvG7m_PXU%@<fg3Lm>APqRHR07im+(NCai^wm+m)AZvc5
z0o+1FCxRfmfSmC2=g$Lt81_==gOn3KE5Ujqa#mh-?6A30^+#r=bRsQ;Zq_et>-F(n
z4-XNrJRl{OAO3&Q_LgB)cH8$bjlz~r1!?IHX{AdVq$E^82|>CQq(Qp7QyS?IB?Kg-
zq*0_B2?0_3&#fn(^X>P3c)y(MxvnGkzGJPq<{Wd3F(vP7itXB+rtQ@)#Rxg(l+QIK
z*<0Yf$3<8FW_uR(=;Lpn_;<KY6$N&-bt7UC%u!)s{ipcdMVFp`UK4A4?B%|>xAs1L
zg&P~!TDy>g50|n*vF@5bGKLpwXygW6H87q#l{Wgl6o7?&YM_T1rAL&cc8J?yMh^st
zHlU&bj<&5%PM!oX`L)pBFiZhhm-sp%$#hahP<{5hT%Km!i^i~yU29n3!o1P`YyR$M
z<y_ml1CcH1DJ_8k0U8x5dt-Vdp`<S8kJ#-PE%t-PxU6ZGe$BHE)7RLIt7p!n#ZSjL
z?pVy+G;rZ(&9WQ&&?XrgL}szCH+07Q$yM<o^I_#&oJ=n9c-~7Rb1ptRLP8EWbIiVV
z?6XN7PWp`W1QxORiqKdexrp{xws2dT`%|O%Z<t$HxUBS(y0$%;6@s}Lf(WIT`yXA8
zqa@Rw_+Qv=Xa8uw!;#7($@mQpp)0dZ`;=MKbc(6NQPy54qKyXzMgk7K>H|evg*<$k
zey_Lddz&RI0+kJ(I-h5S+M*OJqCU&2q@duQkBENOeuHFUaN*}EEe%0sqAM300}A?R
z?*m^taf8}#Ia3SAx728}#(w8l9fnfGq)-#Vi`RzcIEWg0DquL8$fKcjIGBtUi7b-o
zXN7-nze<k9RJEBWc&;$%{D|3-ih>J6NykCE{09%4lR8Cu@qHr!^ZRPuhg0Kc6wepD
zYEfT4T5pKAiQxab{{2T7rgMQyY|%vtmE_2%!nsF%;@K0S#xJrv(BRo@Y#8D$eu+G&
zov?!#rIz+}ww<J--&2PWV`UO;XSLs-CikUVa{;d*prieUQ~~;4uWM@P8GBvK);;bS
zu@5%J4xT%|QC2K3E4oL-rynwkfmOOX)gNMNZ@zMQ7*h$e!CJv6^2ejxp#mC<4w)6j
zetqwHuZg0e*4C-Rh{&M{(+3u=-v(dD@v+H`t%iKB7u-4P7+9KXQ``~NII9}@>N=yA
z|K7fuj`{Cf^!9tpub=vK@r}pJWs1JDJx}_c_IbMj6s)*SScahS%Yf;R;^(~Nev97G
zU6ClbH{Wt)B*|#$vFA5L(lT^Z9VOOxzNuiyO4rRjleuZGA=K%m7yGQeaEG_JK5F=M
zwcx%|$tz_A&wRch2L(4K*$28!-xl!MK9oDJkA6YP`u0p3ExF00q{yF;V`TQ#PDiap
zPCPyPuDL@Lj*dJE5J#p_WfXHQSB$GKE`EQKs}~=rJRy^*wR*^e;;$T!-eHzETVge$
zs{N#>btk5mjFBkC9F-!oOHpp;_1Nz@Q_D7Ej;zRUp~N&<eg?Mh(O*9l(3v&4)IS%c
z6`uJ7^;_Dm&s7I!$&r#-bqn(-ju&M0mr{Fy1Ru1FtzSb{wQ|D?=RlG8+i2rqNsAIf
z&1RMznU7hS^LE_szA(?0WrE~2Cm5!QSmQ{9^skC&q6CSfA?z-QjCd$)`qAGNni5V`
zGxUgcUh(t0SkK;CdwxEh%$r`Axw{>D!Th<Zgk++0#W^kG;+YtG<>X4B?1X@Be;Q5L
zhDbBHLrIbM)#+!->fFSPpL|9()~^YT9!17lPZeF?ao`ocq8;_4$gXUDAz|3IIgbwx
z&j@f(BkW7nb?ctNYZX7b{VeKUT&U0hHqSDqS+-}K;aq&ir#b!|E7f?F=^ll6KN%u^
zRK!+WT^sS<&QRtek@a}egl_z26UBx(*@FC`q;)U8Ein~)?_y)4n9qquJ`52YCoH0M
z()L~f>b(QG5}zqT2QQ!feB4a>s~V-`wViKG0zXasV+332w(3qzNdG`Z=HJ}_mV1Lw
zH@2KgaLcB=HwqfM^lnuAKtXt*QhYxRvrU#%;%TD$)>_!Kn;%*jJUZ`Yh)=OAr_3=+
zbMQeZvxy*H^oiM=$J;6^b=^l>leL&N^&<5g4|vWfI4mNz_jSz6&xI6DIjGJrc-Rss
zB_(^eI4P6}1rJ*>>cy>#TlRaF#53P33%bmWZWCllUC;Y`xF7y|tNrv)&~Ju2^KD|4
zy*TodqITR?&Hgnk>q<nBRzH3yytm06DxxKeiXg*B^hs|bBw7?ioGF|YbAzZARm*1t
zUOi;u(ZCRrW5#`NMmbs1p<`L)ut>uiq!==A%eyUz*C;V**PEVfMD>)T*8lwACC~KX
zo53mUbfT}d&2$91N4DiFEwiuLdHd2o(2ZuDbQ$86#F3GG9w?5Z<u?c;pryQ-5*;g?
zljwgjT^BEpofU4mLx~J6wOClzZe{K4cRwc?3iZtnG}52;EVFz-r{K6w!4g#+WUY1`
zUA2zLW<to}>lLp)bF=IoNAjDQgPM1E-l}=&N$2-_X-E03yT1IroBMI&@#ZI{2)u*o
zM#}C9Ut!CTpOvH6pB;8K6kO=5@h`rh-K)!;dn;I5E5J%h(bM`SO)Txxs!qH_y#&$9
z!`)ZgN;z2<zv@rhifGS^7a!FO-9mnFlpu~4?Far^v)bX$B@~|dyOG`0LRK{|SvZtB
zMSG&Kmz|6FEF6W-qDyVtT!;}AES<GJbj3JcY4c$(*~JZH(lzv}O+7Wm>^|wz@fxv9
zuC9CM);)T6aef=}`T}!Pr#Ug@EiT)8r-{L5G1t-Rj%u;!HCa=KL`&(=&Q@Q1Ic%N`
zG$E95GJdbaM%yE{%`E@^=UTIe=HW1UOgWB~XYW)z_$DI@0U@-`HpLD0sNeO*x!;F0
zf6r0TVDRV*@p!+QO{{wVZO)2JOSS8hWYyKpX&bg|X=i#`m+Goxu3fc&J{1(dxFvpv
zF_~{2{Jf(Wvl_|adBlu^l{YUAZ;`s%=;`-9@|biOx6{v97$S<wWx^sQ*^x^>Q?*8c
zojP!dLfpS#fm!Q=<=gVPOZbbH?z`loDE^q}of8^6H_*o$zlRRhrOQW*VNPn>6ni(f
zJHJ?F^U|^@j>11sf52fTnUNwQTKWUh*r`~pPb;IpDE9VsbnpxdtOy;*(?)*ywVm<)
ziBx1APuVmc>~PduY~d$i-kOLPMDZ^`L!TIraC!L2sp4{z-TcyHs{xUm#`)(<XbC3=
z?7Ja-F9MfEVX#B2-7xlVch<ox^rN*bqD(Gg{i&-xh^AP1V8a7R5;-vyWDQp-D0<Og
zln`{&#2miLiix${BzuCav9NAF4=rDKFgu~84fUpl8bs@yj;i%S*NTUn*l~Am45VJd
zuLxG^8op`Tjhkz()WoIUWS2r|J$2Z=Iwlo=Xa&v{hJdor$pit%_dYa8;V8eq2%s>S
zinjp&5QzSOKBGd(OobaQ=@WtTsio1Pp_@W2VikpggNULc;Bk~_zkS+wZ+AfwdL=ZT
z<Z8L)!6{9Hy}i|nAZ5io0YQhaU4V~mVFJWm3qr)u(3zAxNb}iX3Ip^hed&hLIyaJj
zAG5VKq#60>=^RQkqWG_%%E&ry_TWxGV`G9uY2TFIF}0#uko-?A6TZ2}+kYnT<BLB&
z%wGW#Sv?FaMhYc60QYd&TQH8)2DO5ln;VMEb&Md``fCav9zS?<q+e`Ka=<1=P&7`8
zUGv@?C<@u+gnAseDui#=T;0#3wFldyMIpEKYb<#kF{53O%G!(kSwbz2q_t5+f&kz!
zfFJ8`m^E>+l4M#|RtD`J`Ng9(JyRbh1^8AlezfJ~ToIh;fWVypr~{qfEa)vUGJr`)
z#G&HRmAh3*<VB-$AM?GnLjQxx<&PAHJi%;vkb<fx)0VyjHRkAf{4s_h8;=IN{kVaj
z3ki15s|I;29{pIkp=ZB$B;T*PBu(eM8Ph>E+Ig0Sow;IpOBSz)Fku?V^MhHpvVH-y
z-ajmP{*vn%=rCZEs8%F6&ojVyXND6>;it5yJ#Q-IoZlX{W^3!{#G4t^h==Fq<t04-
z(Y^n0=m68%s5)V4Xz=7K1JPg~RlgSqAQIzmQ@xLQ#6$E}3t(>Nhr^dSXgmW;nnmHJ
zlI2))EzciT&+?kc2mY+T*DWz~MO>`$0Ak&qoywKH&7R;JD+vOnCT-1gUSGa`)vC2r
zZJvTf0PP3_DR8z=j*hyb1G@2*WQ!g@YftpNtFu@>7q226LTTq4M0fcEN>H!4D*k*$
z=y>~~J5$LU0#pi)iClM+hmC?{m=LL6FLYNYh!xIGY3o`YA+^5XjlK<N?KJqI7leZa
zGPKx!oSlsvgO3EeZmDd!5OVjU3$6&chp)I5?8+UkP%c`PHF@5GT>A(HEC6Vnz^Z)%
z$iz1=gO*T<wuo;8dyB~Og-c_Y_4<=xGFcm5U!~+<_C91=8(Rm8F;uFX^DQUf5DG3I
zj=wK_$Y3sTD~P!wDR^_?qx3LCOdc@CzRpyYx4UlYrFUUi>~SsuKSx49Q1$%x1yT!)
zhs=(*g|U82&ZM01$3-~9vHdAmpDTfKrfEO@@tIY8a`O7>s_VhBQk&@;(Dwkz9qA+4
zBL>&jS<uvKc-ZQFZ~XqP5ww@E|HNnAy&iG%$lkCy!OPfU>iuG9E!MDE(VM3J41;Ic
zBK!rh;P#z?X^|Fe*7_iMk$%dU@JvS~mPSDr!mg|nS+&>E7-&de2Vj}G|EdEh{cS(q
z{f?8hxF4Bc#9F>(34gOC(ZA4hz(3i#LMJ{wFnXyY`aNee2|elf6Ao7{Md_>XnR@7@
zMu#TukCCzjtA8i?$-q(SiOJI%@$=U&<|3xjU;MZTK*OQcWS9F9wv!U+D|N+t&t`x7
z2RxY7G8AP}Z;M|3nnpl<?NKBS9^Q>%f=5y~I5=Uo94RxoB1yz`*jn2vmm{<S<p7!(
z%R!|WkR$gmxS`hSjd%P@QR(p0dn4H9mnmRJT_~x5o3_FFeMNi4cl@+(H<}<;H<NO;
zDOUvuNaLe(e0BW!RNx3;krc(Zp{XN}FkMMpu6QkMG!-AYAdjz}gdHa+G@m*r-jDJA
z+3Xe{0FYC1p4^LkgFVig9fdI4u}toi-mP;{NL%R`l))>_*%??_s^Iyu{;s<l6Yxe~
zIc3$>;(6<`JoYH$UuVe4wyNS6L`oe=zAb8ML4ybvEj1GUOFjo2L6Deam10S+`IWTy
zHotXb;dKsl=dP0so@An>8Q<;9bW?lY#|VF7hJ#A+W&NN~u3nPfW4_<(Oc>7!!;A+#
zoGLSn^dx_GUwkdq>}QA)+!qq`^xYb&M`O1#6tOvXd1OKw4TrPftxVJO@vUQd4<RmZ
z`PBtvMU8|*@3{$W$|vh|B(lFwXd!2%i$FnGm)&I?Y(9Swy3^d2?CNOb&F8tx_UF1N
zJRL5NDj|!v^>P}c_Pt_uyW&G?YhM1Rp%Q9b(?;C4DI!?Nmw-%9zBJeBL24dm_hw!2
zL^hw$SBmMY>|m@By;yqNrpdyJTwxp8e=pi{?-!Hf2f3Zb^FqpGIMOFAl-=mH-0TJ#
z0?)s{Ac!kaYC_&E@((6KPIRl1I{C4?Z`4|hJW9~qw*Ki(p}xMGr;e)E!jkT%4ATEx
zI8D0iGR?-28MU)v@Ap1<;^pR->(BDX?=(<rKQprHXIkW(hS7EZx?|P+_aj8JkKMk?
zNg<cEW6f@~0Q+8;oHIX}q<c#w=HDBDf4uD5t4UrrCc4ABhM_jF1>d5VQX)J-(n%?!
z8en2Np|pBJba)BKcid<fw`rcObMHOkN!0!`2Q3pd0G{!BcVqOG8{SKIOQfegcWu7$
zuS@%G3DwyJDCxT3a;ofosh1==w1peFQ@U?Hm_3Fz`|y5o%kAfuEtUVd_-?xM>Mu2%
zm_83)_;iJ1eYZ<SGj&+ze|0fRlKtnwobGse%e~4lv@Blgy%m*kV2HeE4xr0R7IayW
za<`oN>hE*^_vsurLgG?4?>0|Vt(h_{m%}<XZvjv+Yu(kH?RA2Ys}aeajj+G#2L>36
z6=emFRsH%2sD#1cKD=Vl(VpN3yz|~O&5^Bo$$f1E|M?tit<rH?!)f|HdIMs}m&PT>
zXoHohoI!MWEJ7Fg;W82|4?^_M&$mIegbwju+sQ(s!xHq!kiVvXBqU0fEn%NK?#ji;
zTM285&d8Zgv9w-YCC$S1Y2>iuPzD;jbIN$2LB+*s!?1KSvutnxf)S<Y@YOnU8+5XW
zXXL*Qyn=n$#3dv6B42_c*SEa>Va^~S@sK?Re6ku9umA%v5EK5}Y{Y2OQ{S3d#!?+0
z0_xj)Cwx+y>#3eRUH7>Ch8=F<z(viafETmk_Spyf?=O(QLc8#wT@6IOG=ul(`8X}w
zC-%<IQ^fGF;d5x*T}c}ynz-nH3`*u=NRb{tE{-ueJ3U?gkhGUEnY+cLlGwFDLPP{Z
z)rlbzA|s-i5*aA}yA2Y@ZDZztVU%K~iWHdV{W~7k(3*_@^yYL)3$%NG$}!-MihKrR
z;6oFSjRJ>wMHp00W%|`$p-B|+8Ak%5TNR7HGLb-m3$vkQXbguWESnVF+pXhjZZCDD
z=7n&ZwjLwD!pp6x>W9ep58dt06vQ1U&{N^7(0o*<O;aMO@kUO^9cT@0FrQ9@k%i?k
zm=XYj`3sVzh4nJHpi>j7)|s(ka#?fI8;bOEH7t1c-fXo@aRay)(0rRrZUCnaxurS!
z$wVZ`;YI8Wy!xCAKL&lO>7E{W*CzM~)Zow_9IOBmK81*@m}b3X;TTH%!|W~;xoZxR
z97>$+c*yI9ByyWB+@W*C``7il4^!!98`0|T^+b^mkRtWYFu1M}DBhG5@S~5OJmI`F
zW0Avc+Rg!*Eg+=(E^>0>DBsDzc@=d366}Jv<7hnDo`G^HTI4sMU@k%>P%anW*W$9}
z`n4+*Zoa<*>9s8B^P|@RMsh0=q9E)aj-1h!e9au6AXs}bzM_v@t;p?EAvRg}HPTh*
z!8=T;Ix>(Zr9=p`_}vGCy8iniFiimrxjTDRLB}GKxfnNT4LSeg$B)a@KrS6t$6@g)
zH+gD`$s`nvoA5H6)xKYe&`y&B(KcjikgwTBq;<1^2^S+HW1G6}2QL&XfW29?o5|R!
z83M+wMu1=fY9negvO;VG43(M%me0y#fOelcI`TPYL7EPtkS{-MSc{}rzTG4LV1<?S
zX~<`=Au!EJl|W}h^JAy)^}i7}dMg(Pw%n63J{HXgGEicXaNV0yE`km#DTjU)Fb_$P
zQ)%F7s)WUdfx$m>%EpFJ$}hpz)cW<S<dKZ<dC8;%a$h3f$-RcF<2%Xq{lcgBbeiP)
zuLnWe4A>a(lH)2)wx=PZre3mu1CK@dTde3^Tj%JL2+cZX<f2CYeWRV=<HajG0zp|7
zm)IH<FC+3{kr)^l`25%?0z#rN%rE%1hei=k#5kBfX(SxEPe9~eXM_Pe6Z!W~=dC|F
z%v((gakchyDyyk|IzI;gLLl^8bZMhfn1kI!GmVO|*x8!I;nUE+9}u}^{pn)`$EP?f
z-&-t`NZVd{@~LU}S-#G!lnO4XtJC$$fgZ1aAEt1|JxJ)CztpEScB$jG<9C&e*B_4v
z_Gs+YD$wCdfA4~AM8j?8UGq+qoQi6mI38m82H9tl!&z&q`=^w2F*o`DHVlzRT{aPC
zOL@kL@J@|ciB)`(ADy2%<~y|dxiB%y4Rq9hEiw34)gDnh#vaJ4;UZAKX^qLud@CFk
z6TbP}$%XvCj}dCZ^tf29dZrI@Hr7ewE2<LLb0Zmk^Yp*;!TeW%MxhvI*fO3;W%{D(
zw1|ROjotcw76#a#)yxJ9O^ofrf19XM5)am{zNwzP+L8UNyQ_z*F|(O_tScPN{=yFb
zKTo<jGOjY?;qYoV3<az9E`24XGkwa=ht~_U(&E?u`-^$cX&#d+d;6%N@FZ74##)}P
zT_k)Rzlv%N`NqsOfRHQ`gs=j(9BdD0mC!8dLP|8j&e5Mbglz?XxjBFi$?*w&!RGnv
z*BU<dVQ%qq9>}Fj=P&B;HNfA$6A1a~85uyGX$Q4E{2!=f7A^1M>TsEqpR7RzO1bIo
z+ZD-b?-cO->eWa0S1uug$_rp!oIpH7F5IeK61GYs7Q)BQIT*fDfj>_C^r1Pq(=N`y
zc$rK;@=v%7&s+v7C@4gK15FcT;rPhofL^;8dRj8)M=B`(XTsJ>{7Yf3_15j2bKH}d
zebOJ0EoxeJaq*v8hC~75XL_J+009f=l4XHq9kE6Qq%){2CNz)}u&PBdv9Y1NdvAzj
zfWnE$(U%MCfo{XZfe*R7k78frGTAYjPt!*(jc<H}YMsb^&{x~gkfk+YHb)p610Jc$
z%uLwlhjYv8{3In9Edcxkp$A=;mLw2Jf>&v@H}Mfqwgg`IOPdMf9qCY-mzgYn*Rc{<
zgf+(wc!s~Pjf)ju{d<+(Mm%a!aI~Oa>hPtzjEp)qj^D@7WTKYDeWg}B3796TCP$?q
z7h^W8Idw$tK64Ea;hB+2NDO~@>(9GE*ntv$2O7XL2?=#nD2UQeDSBEd)HAG+<r!zu
zUE-R5-YCKv+zNqqQVNMOur7dJNJ>S8`1@8vhy?+=$CpwgL`ASo3^6~rKmQ`+ZBd}C
zWMw?`Gkocma^22ha(#Y&{>>hYKf&BS&k%1W(yr6Q4h0NOXH{b|2zU=mX_Y;SsgP-S
zAtWiVraZ5?h!tXD={-mBZ|(|BtzLVLOOG?<c&)X&^)B+R{zk-XA=QKNPs%oX5G9UT
zg2tR3#$McW5F1(vjK}`-QUC7s@xohsus{ifP$+bAJKRgYNpW%l=+|=n#>(KI*YY=R
z)P-KH>e0K4M&YRC#*okERrh*-@!5q(S^T5_xfXVFLV^o5Dx!(|BuUF^{-AI8-f8c$
zxbZ)4i|KMIROdP6>9ml&YPXnhXZ-|?C)t6g2-?3QD2hLhuI?murU|yNFCF&L`Of%C
z+%025#y_tOSv#Qc6l?7m*L8fkAnJ(5rI$>0=XE(-_OvJbvn~FHuu@W^y*J$okP=l9
zxvRd3n@KVqk96|n%Kh^y|GXrrLEt#JTCCjThJq&!M`iT2>!<HK@w0s%SSl<E|BXsb
zu29Q%R`FY-2Eoz?8d^0}=8#IbfTFt->4phvegLfi;NW9q>?-qH^eEFA)>Oe292NB;
z`o}G?jdz>v^!FzQdnD)!0E`9sqlCx7GEm$>4Y>|1I~EOs0DQ^QFEHBU<m80rrrkZD
z{11{-0Q5t8^0-J?NTLgy=^qd9R{V%fN+u!cm`g2?Oweky3r3D`&CSgf=hyzEAtclU
z;wg}2oiKv5{j^GU3=P-6_Plre$CL5U*cXnF)Q#P}?~vZifrF3FZBUb+ltgY^sb8H7
zaKM2AAXrQRJYr^3OKHv1?-cR<i2p47ze~)_vh<CqSmrZsEv+cI7qp;;^#X?S_As!V
zkZSIW*Fc&F%GS+;V~A7aV4gZ)1Nv`40Re%#MaZzQ8uaW$**uN&=Tkb~8%BHSMr0{!
z$kf4H?arM$PoGvX8IZulBO;>Mg`(0G2^92FJ^iEan-2GI5kt>%l{s=_#lW5$1|2gp
z7um(Mf}97`3cnSEkGI6c#I7}=$Mk^6fWb6e+?C5vHqdodcjut*XMGPImYfJw+MX0O
zbM0X3y#rYW@|@B)xAv(NTlSpy(M3VGH~@0C^ZG}rjTo>*xh7=WTLHOlI4Qd;fGP|(
zHHV`yn62eACfXO?88x-F(>qc_l#Jeikw_5<2@@c;KyE|0VtHepvBBArh13-?)pX%+
zpeX};n@1Vw;2KAwxDZN(mex=ODl};9Uwsp*jBMT?jUO+>&kbcD;-j~aG23>s_R(<f
zL!dn%Q*&QA5!&~U49zH>vYq((^_NZjwn3V{<-ytkb<N1k%n#sow}OV}haVlih>J}j
zPD|6)eWE|NXBOr*3-TVwUSPE$j}PR<DtUJ9A^<}HPe1?RxGf|XK~OllS#WIW@#UsJ
ziEAp@(JUbV^J(_)L(;_}t|^dHq6UFu5xAq+W=q3(cD0YJ!2h&MGm`>$rJQ?@?vpX;
zCk~9~`WDNU&L}@o?<<oNAAU8K1C*UpS(S+sD=Q%1{QQ6sCnE{3&FpOl4`8yRSetWA
z{9C-maDz}3Vfnos9v+^VnUSKSp`p=aW|LhWl*QG$bLS=Pr9^X0XXn?C4fssMl{a09
zanuLW<w|!B`sK_T32zwejkXWeE1dS4AKCl~ny~;IgF3=y6WZ6ht?u7It3(+ZMR(cx
zOJ}YvVO?NJWXpY?@yO01zJl6Ht(%V9F6ruDVvsK%Yi#vrQKcXVtAPSa@fn4z!xvZ+
z7U5GOAFwk+q35yHiyO0(BJ%T&tKIS?h)O6F3oLg$F`$T}8pBVj>!{k~@}&!(IhN(f
zpZ)F;TSTke`v=5{WaH!G6BefU`w_D1PF}OKGkbygFvofp_Je0eoSXx1xM(Mlx4%j9
zH;*8r{!TpIE9OgA`(vs8`^1XMtLIpMJ`nO(zpQf%<cbl_lv7B~;G6ZPuX<VqJHnA*
z<So05#DS18(>JfyJ^X>l6<zP{ZQ;uu(QgfJ=cPC*so)0IJDz*X{o~(#U?5AF+__E=
zd&$hb7DEZjUQSMu=^kniI{H7igba*S#tYTO-S=VfB*{Q_;^2^LIYkNotYJy*&EJ*#
zCqzVYsT&*xw$u}ii@l)i$SyLca~TdB6>AKC`*%$uLwU!@ieNbOkIS)>=FoWZU&)$`
z$++Efna)VpOAfe)tXTi##m*ida|iM3KA^vmWJaMFb!BZ`J<KySbhp3h#w;ye@o%s}
zxVhCT=3E8B?Dq<BU%Ci;hpX4fIXQJ^I&X!<{@GIhy0KxyBk{Yq2;J{!6i1{9xmGJ>
z1%cM*r_8+nb7vM%%xSx`Cw)#-0hYmUR(gZ6?dU`@>zTtYoj>35&&!3fb5qpXM&4d7
z$?G~u2p9*M)n?qZ-pAmFB>m$z@yKF4Hxp5)z7wM&@_~rei9SQF_59XvKB<4)zo>qf
zxqGimCT5V{<y8C=8!NwLI$^?pd}_0i=Z1Ub5LR|D>7vkO%v!tT9|74(ChmWJ=~tE-
zhBdjevqo9jfPcpq8tw_#kE7CKGTyR$f1*LQi5U!faModAWJKm&5DOUI2y^o2%yq69
zm}TG87&B499mM|@)A#2oVLa3#WJSQWkuiZz4AgPRqL}U$tyS2x)o|HM<-YKaeU7`O
zIeCG1hrFF}{stVZ4yfmX5zsL}R;ttN^<n${Tm0;Ycs!>)iWA?ho<KGmwfOh84^fb2
zlET8m=|EPGhyp$k@Ca-47{noGL_$m?JmQgwrP~vuFCuVh-qZbAh^wdwkepc0*6;Mj
zFn}9@AyCR8E>VX51k*y8mBFF8KG)>bCkp0|;3RXyDEeJzXIyM7B&9Hk$oxYrrvB?w
zPn?M~ZSKOVTRtZ8<FD5&P1Kb5YphemwLA7%3`<`JKdO~j@-Ra}i`n#Mji8l?cnwyY
z{B6j?d*}0;Hvo+M_>st_yU82~-756Ix6Q$9%E{yGsLB+O;g+{PylZEL?oWGZ6%T=*
zn;+)&dQ3=^dWG1%_Gkg#BeDf>mg4)m>+nbSL|<PCq!#jafq%o(HeepEIkM9kR3EHy
zD{qWu@mE}0wUC3@D5mUb-;hYw+z1vTf4`@wS#RSrC^2FB*B{81h55L>0?(&V?6K8p
z31f5pF>h^dJfT_RH9Fmo1~b1XvKapcm{n{9JPI|4yn*VAji5lPyrCxl*EJ^216*%^
zz9^|b^>(C9so=(1d)6c7w0Z|lop2J?QQ*Zv{@;KRBt{r}%*DxhBtFE9`*+0{pml;F
z6HJS4wVmdJ4k5IamSRx<Ks&Q_+x4#g>w6M`vL@P(+4VXSKO6@l&qO4b>-%?|V(?_0
zf|g$cYEasuR#OHfE>6-sZidK6aKOi7u8e{uEUF*B10a7j!_{GID{e`L{$C@%06qfc
zoGI4~=*B!w_wRr5U3?dQS|e(>?xv+k_&1~@qhTZ-nT2L+aZJnB#&4@6r#FOP-t`8Y
zWA!iMzAVC0lK&b-$#03&hgK#eCwnd|A8X{qXSm+&>Ci@dM7Zm>b_Q9Ekrv%o;axB)
zfzwD!YZQK)vyZ;iB!akQ8WpJ>5}}VK1Z<02L}WUagI?iLqd-p?WK*0I6#@hC;D2lP
zx$7e1dBVBi7M$&{YB_&TgGK_x^VdQ@zqT3!<RZbN98}+69Y8d6EJ`3tc11Xoijdld
z($USWsDaQ%%F)GT2V7N|43^i|r}7l=Rcb<%KLZDR>|CdgT2|gS2i)jj$Zw(!B?#+R
zfXoed3e0ivEa2?Rh0R95ukTBD@lC4_GDDX7j<NUat5za;&<Mk);JJNK3kwNuOOWVR
z8(ru_y*)i@zO;ty3I#SPq?HpBDgA*X$N^q6hU3vl&Ve3VrRFiXCqNZkoqs(Da3^6A
z5d*EOFqmKiI$7|`pd#nK*9bVFAXvOZ7Lj=WLnlk$Cxx<O2PeTJ38J5qFcHBRipMwS
zyDh+w1AW%JCa{w+&`SJwY=>_rbtSjH^{TJmQX>!(f#&}iKqe4zcS0A`P8^v{{H<z`
znUnY>1fyp)bYsiGJ^*eF9S!ZJ!Y&vdL9^HL+GEJ#VY$_EDuOJsOs6E4$>6A)+%#IR
zfiWD6=XGA)<%IWHIte9R!N~@g-9cwkcYV;?#|H*NW~}l<o$m#*!NhoZ=)2K?QxKDZ
z?2{fa*n;M$PDTVJe@`q2xD;)}!y~`P9T-JKc3}9N;xXvlAzQoXd~pO81wPOhO*9_?
zxi^SJgjK+c()3vh&;1un1_H2gKo~02rVdzR$&f`pY>y4nMEmx$pxD0H(U6})q)c#w
zqaY@bOEf1OLDN-&3xVMg(%^8R2-Fgv{!i?95q2p8z#X9UeeBAN3LaN0<2F#*sN$)s
zsku5kgCWu(=o=Eg`FMMSOY-vaa+gDTDdOAVp%}Ow{HjC@#FIL!(vhfTmA=gduvIgc
z<tEnUD_0~6G-PE1e#U@3uEiU$8evyYhX<H`^Zb?p$RXnH-DC!d!}X}^J@fO%9-nep
z!obDrXdetYG<)z7h&k8^aLOcXv228cN~fsT#Kgo>vz>*I^;|?)n30JoCC2FE2P^R1
z4-!%JFeF}ijgF01%vi(<+f|Wk7dDjVBJAhhM$#i@%uxjrx=GI5uTMd?0o}@iwVnhy
ze@AL3-+26lZ!lI>aW$?*yEDcJLKSidIqYcAXqF(JgOP?r6I6Ex+d>B@uoDV`i|kn$
z-gqAOfBNJKb(y8Zoku`A1Ron-@I<h&v2h%+T&9HS>!YxJheSsXKj^L%Wj&jLBM=OV
z7W(^@htlJ<pd`xRvrvZ9ilyB%vaukT6>RRnCIPb20t{dH-x#o{P*oldgL8nFef@be
zY(v4oBzC=_QxAF(T2CR*Rd9H_0I8AV7q~J@)uBzFSQE9ldlirl;=RF~55@*IAgYl)
z|Lxjgq{!}b|GZ28auVpzz=;&UUlXFx&ZqLU?UEooYUPxDaLy^#$ax2B0e#xH=n{tk
zmx*D~ACD-|^gUAZIUKwRcq}+%8v@KBWIL68@XiI6Fr2P;(uHGc$j-0@b#79UOynOe
zJF3NP(Uv|acGgg1`2Df=k-A9H^t!A_gmR+iP{G0eJ_{h(FnzIvinX(Z%3=<uIlD~J
zPbYqTY)%3VDaA85TJYLq?Ywoi90$;KbysdK;~~CC0MHuvZgSaMD0vT2(KfX}y|-L(
z4e4zFg%eVI@%&pi`ErMPrg+~w)(@I^Kfy1fPY>ou=u47%f_Ea=ozP!{65EF!gYtK8
zE==S)hDpS8;i4z5aVr}eO6`$#{vl*A;=o=$xI3;=s7(2ia-Rx+(glQ+{#gs{Ohyv(
zFOXTdh&VVI2P7R0l{jjmp%-8py!}A{!KCQ*izmCPcj1SYK;Z4HS`U#ULJ%EPcWML)
zTVx}@b*7Gea{znx<n7xB72#7r&74(POkk+gDrN7Q_VoP(lKFrybmW7Ca=bh|(Z=t)
zx(McD_#Z{!I^vf3wlp_uIcnT#zOly(u3nbZ9#-FpAO=AtN=8H^J!#HBDw;r3GY8-y
zRAW{BA?epu7@SbCJa>W?Wo^bo9*RUHM@XH5%_ZRWgN`#urxDN&CFo0~EA0%zF~5N-
z18(Liq&inLS8N_*3YvvZ8JP10Tip5)f<q$rK`y*_J6aR=H3(@tO-QWM;tB>LDh+Cb
z^rgR_ZOd9Zx-B0$lJEC&R?({)g16lEXDGQF9*@wkieB<cet&}}JKtR56^sqRGyg;=
zcPC5I*iJmQ+M4SgI6%;dIOfk`>QtFYwc$Ml^O%E`4}G^-vOFL#gERPpNw7tS9h(;m
zh}zE~u{~LN`I>!jf{488Sz059PZ47H`7vhQ1w}MCp+?Epazt-(hsp3d&C|=6(Z=T7
zz1}3;Q~{>s1{XiiEYM)4?GeQA^%nzldFDT1i27_XT}lXQS$X;V{QQO|IN*ZpFxHDM
z1RJ)$!sIg9GHMOP^;)|)X!dAp8+#C2zk5}0uK7!PgWk<9E(xh91wvObUT(U)OI-Sx
zK3%LrNUzF3rbJp3T9fWI5)xSvRb5qa2{UlslS{oD%^uyNpOO?37WSaRcSP)v$lSKV
zDE<Zy7*C8&PaACX^pPU{U~7_SUzK=1sJN|0dqCu<o3@pUHpG*l><05l;Q;&a&VpO2
z;<lt{@a-2f)JQ{|EqG^Hx1MJPB2PlTt0mg~n+dX4XwDWVxUJnzHpYUVM)r?G1Yb@Z
z0a`;}MM<EVNW@h)2=**m`5`szR_h;gz!nlhOu2E(-C-Q-QMPjZR(D>p%@A9<_JmH5
zh*kvU86Elh&AG#5(hW_}5iWrq0!E(c=WU&E=~bjt1wZzq!Zg-k`jFaChC=kdba|bu
z-;l=Vh|s3r2W|@_hB3E*bMpfX`bu;aB~^E0vV%mn_UXprUF=}~D!54K7}XR%pl4rR
z)KgYSaUJI_^#o13yhwzJBPat6)LNW3ViOV)sDkG(RU)B_gB(`cpT4@YQ*LP1=zZpx
z!e_DYST0985DbIWRcF92yjRt%G@dmT$Iq|H!BeOv{-cha4U&7!*REy9Xq$aCF+J_4
zoOL(2@Z7o5C+q(GEjG<IjC{^Zla^6uA5I733^0}#G$xq6lguM^p$t_8OCHfjzRyEw
zp^klR#Z6D)c$dG7fT%9XRNUl_<#7~X9S7u|niNI|=17N-x!t^F+9)Y_*}ezlU{OHd
za0Tq`l(%EbX%QNRQrn_Wzua!7wKbD4yQ8x^&Qq?2PqEyUi<|obTmx;KGY&DPl(Bn%
zw5&ggey*K%&CSJy$&|T)y4auy1M7go&d#pPfO*)Lo-G@T#8|^g=`w<(cvbBJaQ!$^
z$Khc}t_Iihbu85Kum}ZN1e&WS;W+JgfDUjHU|swI;2z&pQ))3UyNQ7%0k@QmqgVf;
z%doKv49E{=qs3aru@DEuM2i~!>k#0s&gbNgTTa^W#u%}Kv&Q{!RyVosfK1n)r{DUo
zU-!^Df%rVt)CVggkBCANQZeK(J|-qj^Cp`>LtSLA)M+s{2%pzKolD1rB04Rjn7YAf
zO)bJsF(iW4#`)o$cq$m4v)r*Eo#`*?abRBXN{^05BZ1B6?e&F^IDCGHoN$mKf~7+!
zrj~$3V8L}Wjj(>Xa>F`-X4bjovE|DyFqrA#sd9n+>OcxdG@%%;_im~a^m=hf+3vM@
z>Q1X4LtfYG^K!W#EHP9kF<cHf9zW0%;*a2h5~GmOsjca2DUr=7fl$>gs5^SuZqyPB
zens|7$(Q3GOmRnPL~mVy$(=xRN-_B1oB7?pe_vUzQrgT+#{D3dw1CucO}y@j0d5i^
z2bWEfBONEc=A9Nk56eWp60wj+@>M)Xdw&pKcTW2HqO4vv5EhZ}5N0J;QHC_M*twTj
zxsOx>J@`xTFD2UdR}}yxHQVl|2;EN*A*R3m^>)g@$_BV~LD%M$PRToQp}MH$P@`2Q
zwhgdVvh-(_b;iJdY53&H6Qero?yue;ye}h8)~bgMy{t5jKw#Xmf1L}SeL9DHmHN;N
zNql&+_gQ_r6f68C3i^Gvt*?ax@>iTd7~Avez+A`L=ce^|h=rW;M>u|0DmZOZb4^Hx
zZy@;hb`}?rQxJue5>mb32oVb(19PUtvhg{FTr+;&DO;r#Pj=E$NE<oMSctX0KO}A#
zP?Mb>K-V^*m$xX$j16_1hwX1bt^sF_fm)ul0%>AtrYSxJ3n%heQkED#v@(<;<=Ps5
zaLi3gV<}0>;^pNfh>qU$4FuuUR>Qz{0NwH<yr&R^Y_Br^Qo8cAK0a!WTBL0*qeqFd
z1PUEU0LP8KOv7tI#_cU={``cP=nTjGm8Nl|=?0`n2vcyUt^unDvXqbq5cApg+8KsF
z*4gNJ4nZ~X!A4@7`=BGs=R2jY#4>R)p%lq`M{+-wiR5^kaTLZdCl?osZzXv2C++Uv
z9vrO2lRCKt#%Sr}miTD^sO|W{`E6BPDcu8q0HNbZA<t16!zq<FkY_UL&4pYRko5qZ
zCHN1g$y`X%qUoTam0Ko9UKSO@)e}WCY`C(F4u-ig4Nqp}bvZ3=l^Vf8{edXv(sy7I
zUgt4G=`M<eF{U~91eifp52wme{TmO?6slf+0&>FZ8V7w6#pYKRw~##u@DayJGAfHt
zT|Pm&Asn2sx#2RFh8FMC!ojS}Ev=of6ruI`**WJ2EBGFH={MB53PeLz1gwd)q^mX)
z9IsB~FH02PF)%U`6A>xt3hV7v?6=avfc+#(?&yRZ{AG@r;4Hxo&e^t)A@qOC(UUr%
z<9Bj<VhcnCO6kZ~k4JKyrPSs8fD`f1E?xGg%rI+<!vl1oDP_C#tF_xV!261hTmkZK
z8>n7(n;lBmSzUS}sXxrUm}6BQ#6Ca)SGz4J#7pdAnQh1TXQ!sX?D8c?LRv5s0wgKc
z)qu`N56622=&+Zjg%E>XTk)#or~pmcCSkwG(M4YqtDe8Sf`ax^2w(of+#9IaVp6(Z
z_3t=48pq$KBK7?CHE<M%job4sLB>@>sYw#)MaR9nT$`{6l*6YG9l%B*u+pXj%0ryE
zQ`!dLB1g;dmsg~mBJ2o1L3Hi&UP-*yMDJJ|`J3BgAZF&0&XtkQS{LFPuwY<)RAJPh
zX7Dp~xA<5g#$vC?aUN&`I3oa^JNk&TX!FI66(1|_W%=Gs83VzvZtPe*y|lKah(#vv
z1I$2ng+YB9%oKT=l^=z-k5H7u^-bxU9CA|y=7m>86IvYDXqCg`=c&OwfXO{iKEUxa
zPvlauha&!PV&?~tkkduN9xT6lx8eUURrLQ=Y;8^^3f~w`8wPs%JI=p>eN+JR1VEa_
zTS5xq6IfGlqk}zO6$Z71w$n8W3kyAY(3J>zHUs7WB(8D_3VN^ZJ^?06Z?Oi~CcJX+
z4S_;%cxDpnZWT2%ScBG8-EZIe;cO#qzP*E!lanE7h`e@l$o=){srwH)(5Deo9zcn$
zTdc7Nt{lBt8j%~&e}TbcZd~^o_Y+SHAoBzlJ1F5w$t3~OYW@v7cO@EGVLt%Tws{^Y
z9Ua{fJJ^m4Ra^@S3Ig7{C%7)!zdVEDYfOFb^PA7Imj$3?h4m5lUfv|oQ3Wit=qw!j
zgTW=o;PeOBjZ2vVX~aoQ@(?((y~8l@N0Hkbp5TaQJi6~7103h&D!&5VrzfwwyL$v5
zkdZ0p@_eZ8o<U91n-vX8D3HRhO9lmkH<${vp)?-wB+!N_#aCFGK}mLXejgfZO?U_>
zgaPt`nFY)TZ1L>P;Ayz>v_IhJH)JT8fnpjW^vqzU;h*lk4eiU1=#@qdTwZ=KswUb*
zPCk_g+(||OFFjoO;P&f>%d%E4pYL0ZP3$1>DPw1Ug-CSPMn|`Tzh5T1e(TxGZJd=>
zFqOM4@)Iu3rC((R+<Yb_zVP=#ggii1if+%0;r&E~5(Eg@n0l+jQ;Zjm2nq}-FxCAE
ziRceFe}ImRoEHSn;{E;XA|l|eI1MOK&bc>WA+@F|ByvB-p%|=y_p3e$-9Tq!aNrG7
zQgtU&wX21EIrd`QNdamDl69~?I8jWyX}>#?hfVknXzy}SN!NKkPQph;Ho8C!2)-`S
z-k3fo5qstkO`RJ6+q(Ab2p$%E1?m79g1iE(zrrFT)rcv2!ys|FLL=D<Y#lfQrU5Sl
z%mc9}uj&t0f*&8cZGCzI{0=}dVDT|{nYOy;fE7Y+R`yeKJ(gg({d<LXD8D{pTLsLo
z$Xl^nl@iRNY+Xf>mNMZ^P>1|D1i9xy1C-FW>2wOTzJTnr<Py?`n@biL!H{S$>y3e&
zqWPoEy{Va*QMcnlV@)a?a71D_KLmUzw{mPuYdR@w4anolW{Qf6Q1#qmLHi{-(PW=q
zE=!+v7wU7!^Bk!jof4i!z@Z-#m;YjKaWUpP516uktaO6|4_ttjb#}J5-7_b{Wh#8B
z&?K&jM1##jY+kmT^bHO3wq)R2aNI4U5TRezRlL%1`@-*vsg@E<zYMLb6Wq!|`eua_
zT)U=nkq?*}iAkvr>#NBJ1-C_l{i^)l;gYc2rW?SG0aMA2-Ogwke2S^cIF%Tg;$k8;
z4Ks-Jw|*UsL3`WRt`R7CGG^KRNymBZFq#!g^aqnVf^l2?7i!3%kGu&N5q)E0wGc9n
zkoEg|QAxU%|5loP`v`gv?Cw=VP*L84odNz;wKfvPQ?-xqy+}!dcjcoLDuR9;!Dmk3
zg~>)x9i2e2z|bNsQU=E1T0!Pxc5o!wOl2|!)bkOLUJ0A7G`&Uk{wW^1g$quOb(`7S
z7s$NIB0#o|alQ%)on>`MP*e+o&{%6~lmSv<$J|s+0zYXcOrrGKTd>B}FZ`s}m6nT;
z)5Z{pu6><^IHSCpJ0{`M;HP=fL)Qz;EzJRdLn-P@u@Dp;R19^hfLq96CK}!j>bSi>
zY=T#p*2`LrOjuafoC$w|0rbmv8-vk1M*&Adu!QC5z_;tu`38J;?NbP-F?w`@DCmL_
zU<henTDx4&uiK>)9ZFSvOsAj%l=7#e4?SgwC^3TIIkYa-eY@PWf`_`QAK&q1Wks&w
zrN$zF6pp-Lj>99DCgxTVLoSVf+u{r+E=(-pyz_;v8A}sw#5>-T6RUPz9iO5dW{%)c
z!f+IKcVXO|Re691dl~wSvB4{n8+g2N%m#{Z<lje^xg=pm3yD|Mvs6Q;-f9iCQs^nf
z({UR&ccr#<oplY2y9ynUT1!>&ae);|$L$7s+WT}mY1*uX(GbA^Z4U}dlfUT~e3g|X
z&Rea}SL(2k?y@5u(qTmUed#&GRX?=Q!-zK#0CZ@*2L5S2<pLC3+1a#Y(of|iWkEf-
z(550QRngJiE%%)b{A0`{DjZ(}fSM>>ywhQVW(8&*Wpb965CiGRDG(l{TIf9atasb*
z85=uMP!#7#)rr*M9^KHW4r(Hz0QgIe&hr&@-$)z0=J9NvZOwQ&$$)mn@@s8Rfm_gt
zreYGV`;Cw1qrc`LxL5mw5uL>V#0|~Y;7C?|Yi%G!=J`-EuT14H>CdXf6v-lGp%_vk
z4DSrtY;?$$wM4+avp-m#Oi^n^<;v5rLohGx^V>l6UtmG{H0lSZ)!YKhQa`j3>Wns!
zC|)kW&8^25T<hbfcuqIq&D2(P151zzv8GrX<WEGA?B{7(lvM%9A?z^9>j0AwS!pau
zSGsg8l+y83v6etdD&-nDj*f^=xr`M;K;@DOMu<K(kfDko5IT%w>920%7b1!Z2}nf`
zP=ff$^iu>!PZZ0+ojk&)Z6BbcLXoBz!MDp#KjEK&ZOf+I;mJwSK7{b}U(-Nb5TMC}
z!6jB`=uVP1G7K+Ec^X!kp{_J{hm(JxgNn2$8oc_7PDnq=s2!))tQTM-0MFKZmZThU
z-6|TumZF<D(VQ@u(z6YDmT3!5Ov;;FI&^d_Vv383Qcravg|1p2f*FSm24VR5Ek?7P
z-Slt_NlUwCj87%3!5!Jw=9Aj7fBNmAwt9I$_Dgj1l|v*Kt6E5c^;LKJpNaED22&-*
zl^96MLbVmT;1yXCDH_V@7eVRzsa!GMgn`#~gLwdbx6mDGCN3{8KYn)Nk1v`1)Z-mi
z9qZ_k+A;x1vR#c5Y<<vIFhg_o0S<s#f^||>`<prXku6$I=G{sY)uz|*vyP_C4rbYg
zCX~pNB1rcq;3r4@59`4?uhAS%C0583*5ns1pIfy>w#PB6YkdKw&Jq3j*%|Ar1E>y#
zNFJ=mgAJw(jfjYdm2)1xvji9~3jB(t<H5kfDx7A(k!nIi$l}_%+riazj3p;R>+(ua
zV>>%w3J+>Mnap0x@>qOFhv#KBo#K5}(cDWL(_33xDxaT>7sr?gZxNe)kU*sA@bOgj
zpowG&Jl8WB=(hBrw5#e1x(@z)E-zW$`_lI2Lj#uN)tA&?Z~7BCUs2uD{(b-!)yT1P
zI)NKL@z-HZVSu;UvtK``P_Q~J5`YW>3xINn$J__)D8|HGsF4*N-HGGq`>2z$tr(_}
z(X7B4V?H)wVA5iVOn~4>sKX417~q22_QMC3>@*&t?52rGOsuS_4jOv3#4oExOaa^^
zl`BRTN*4UMHa$yUzaFd&n*v5I;KgvI<%WJE0`amEQ6yIoN<&E#YlcoUwMW>`iySu;
zzXAgZtQs%TL%|2*xODMEHHKh5p|G5q6LaD=Ce|n1qMPBNSMH`>r6|5$B&<`!j^5VV
z%2EZ%sAr4;TzC0fZ}mqbzV;fiS{WiBby|J5zpXKZE~BMI<YU-iXWaI*H1*imZ}52H
zTKpYi3UG|KS!Y+Yt<KM+0n4XmA__?9Kl_{*AE+FZUMYmMbL+m)?7(+8rwrVg%RYVb
zlCCe?0KolY9_~SwvYFY+{70{KR_(a!m%XOwaT%Nh=~%9jQd6rAVzAkGVn*}<$NciJ
zpPwJN(X#4SWiNbb5A|JyW@Qlp+vEPnN1PI7W<KXfW8k(|_9>IkB8-nmL?or(5<H)a
zGagKpY7KFjg)>Qri@*FSYsoB8b3YqUV%jlj>-GDzdK!TG2jkQ9(}798xf2DfoA2DC
z?Ax%DZJK&<!_DEaRzC={zoD+KUhnb3(SRw<yZBxp=+Yhv@k?Ds*lMEd{_<SP2k^%J
zUhRN9<x%Fx&!4k!9AOT3ykh5Vxyy)L79ju>q#<!GoqXc}Qx@RBthMnlDA<1J|Kf#V
zy=}^Bjj|9EI&n@8)ff~=qqb{&zu@VPSjD<;o5(7eD(%@#<07`F;}6;T#Hx3`#CO~l
zm!b2=#vhV5%a&GFv=?SbebOWS`-9xV@NgyUO3#iCcmEkvPqA928#ih$BktI~BWm;Z
zIAd!Qt@oZKpkN^#wJni~-thHH?IT5Wkx3wBigMZ3M_|_tyaq8+%K98XB8p<pEK}xE
zF0rkyuv&<r9+zxw<8%8SK8L9-Ocj<xepe4<$93d`@M$Wr5f~`oL2lWS@$v8uU{sI&
z78@3BUpMNP(iJC_upU8L`{zQbUkdTL=kw(>;csx(ZUok<sWqd|r#x$h3^ahiR#vy^
znct>cE{9OJ!@GdX%c+nYcdnYTXKX@*q`W>32kTGf96}NL`d{G-nVMyU(AE@!wDVC{
z&8wOzHd3Ja=@-G&%umrpe8OF~4@^qnwSl|>zP5QSd!J@BH~qe)M7pwBcHvuizv_Na
z2LQ~@&~#M@?Nq2Ed*lk9R&~OOkr^0+y@Hu3vow~lGXc6~g&7l0%AQ-~Q`uM&PvHyp
zv5p*`z)&!lHew!A;}v`?7yXJZ>n^lM4iZ%oIW%_zyW?YHwZIPE@v*Iu@g=^xOB&aT
z80xL~C(d;m^+y<hr5KR90EB})4%I7HR?5Z2#fvTTJ=0YqA9PfKD_e6yOG^vijm0A<
z|DV{}vKRtP7B46`TNFroh|J;Q(!Z{kYvoJNH`tp7&B^lV?DigN2>fu7*rKAQo~^Ta
z{~)$E3>ssFYgh@n@6BcxHv_fH2gT7t(>Foi*ZXD!zM+pHB_|Np#L+vk>VF#n2C^Y?
zJQoAJ$7%u*^<2DqL4h`jF!kOXY^a86PQ2AGDXXKRqQ)1@f(C+;uT2=bNfO--uC9eO
za~#xt+P@&9?!C<Nftpk|Ec@n0QmG@(_$+p2sU;5M{#`6jFE8S28ixU-)@))>P?9`B
z0;i9mU6o^r=0WjY{AK6-+#JG(+{RR+(IhxOr?=o^l-BIq{Q}L-fG3r*rvfmz4G@^-
zoJI3qs89fCNVc{64oL?P6OVsTfs412DiK=HpDf<Yi4;2K1uuIF-S3N|Y2^R|`=xNS
ztqx3PmyayWotb-@AeH*nBQsUnQ;e08+$N|uZE-7fM%XisKVmBOZLs;U)srX25Smx+
zVG4<PLhGd0k076f?)A4P#CPb}(x6U5x?iR`k@xlWp0`{j_sMwB?vUnbP?n}?uZ$%q
zP;4CMfR8Q}&B~~x1(g*f&rjUB?zo1gs(HpyO-Z?XV$T6~kugI%?bXPPA!R@&hCpWI
zH=1vqQMH}LLEQ{{r9X)~t7Z?za`F$h<6U~;XacQ{B=tuAK!7FXZbRoW9%B0leswnk
z<S^R1ah*e7K&V6!i~eFf58A6h*XxtLS)^qOBXVHEBFI0GhJBEaTEsPjn!My?Juq7#
zPqQdpB6&f}KLJ%gOhD`Dxj+5lguDIp-Q*;>J=TkZ8sI%*U%A>P)ce5tWA<A}PHL-O
zAuyy888t~{q@`ithw9Dms8Z1)Y#<$_rR_MK&i=omBmdt_>nth(&Lt%w2{=KKisq0a
z0D)2{@;|j;@ZB*qtVe(L9;h)lh=+e&TH@@ZQi=s(7DpewCduGPF8Ea=nGAzby;RCK
zaM@+aED6DNi;tSe_J1AgaM=}>8#obu9DvFC^|B!UCDy+j&Hlr{D>gwIkbxquO57|&
zl`!g-x}&F^rC=GM*;CK-#>48TQ|jX6jPp_P(nNn~K7ZZNK<JfQOmbzSXp9TVC&O%T
z+_N3IXd$y!)7A3q_(IOi>lFVR$9@GY0I2VsT_z8haO)o9E?6f4jv+PD9}J{lC_w^U
z;i<Ki`6816WWi@>^}lrsfwef8u4QRV8jGp2exHD8kXtWSK6CLrlLSm6z=3MT%^~O#
zc$nC@`(}A636s&5e17_GGdQGrDopXywl_VV2xt7lH&d6hU7^U-SzP;&M<z-RtckaR
z{i44py>w>N__5{(Ru>;vFPV<-zG=|3#J;)P`h}9CqzS~|W`uBmNN!@G#?#y&w)biZ
zzppuDKfi*v4>T=iv;tX^5ys)Txa-=ogWYCUZ(VXiiMij<G%qZ%<HTOYg?CBz7X0ae
zvl1N<rZ!^hlfQ!D@iAxrZbkUD-wh4x<Vmg({<`2JQP5=a&Z|ZHv%ayy!bJ&7N!dSt
z+n(E+q~%NngspP*zizI7Jo#gabX+*?gMw9PF7<GW;giFiou`ji226DaIW+$J3Klt6
z$eDjR8W*=FC9NHR&y&3uqop?~imw!_$++z|LiU}-=@02or9u?3fTD@KQ&3i$Rdmp9
zHk15FCs@u2ZjSu9*U;^IdPk2PzR+%m&8p|+z0GEuh1tQKLsF-<f8J(_S62FJ<@xX9
zbn@w*h$Dg_Op7*ARGxh1tj8sJHk)3%F6k#xNsiV3{2NLgvsMA{*jqi`?7jN^=YiOR
zQzN%zbc&Zu4f0~4<JLvr1C!>4%1_9gEQCO9^Pji*ws3cHE#O@xxnjNBN=D|R+09!h
zo!3a?iz&Sg&A5MSKf17Nj6Vzn3iQ7X7<qs^JF4ME_r@+)8r5hQ*X=DkM;bdGiVW7u
zHuK#HRjz`7WOvO=Dro<G6Vc}#PsOCKK2`2L@r+n@BM&U9zEQ#ls>0rZ^i{CSqchf1
z-=LnWnSFqU58PXSFd;;0LEU^8rQ!vao!Qt(YL^O)$xLn}SJi;_<Su?*{`L3T%rfoZ
zI^m)Bas>qe`)$N1sVo>D+O>@^>JRG6z~JBy;LrjLkEC}|G`H~naAT8ye$5|wq^+k+
zR<{J^LoZaH3piDvK+EYjXx3o9=n+sPSe<$S6e)sB04xhY>MJse<T+DikTF%~9O2pZ
zkS;2{G;ao50be;X`w@V?kR&WPFY&Gl6hm4C<^H2Z0C6F@a`ruhu@g1`CINR*2FR`<
z=2sR`jZRJ7*>6s3yC-P7xI@0Is2EaLcMNog+vP#f^8nY{)If;GIUptl(m}?p=j{mz
zq*8&YetsLW+2a?-qaVDD8Y3%OjV)aV;U&}S#~02t*u%WktDkl>fU{j(%+lQMacBv>
z-j`lxuxZ}D@S`*MSk{`>{FM6*=Sf2FK8KLPtDtmjESU2Q^?UW!{{Ft(f(J7TOGpI(
zcI-E4AYGFL4#hD#`d9?d4qxI|*WWuI=C_33oaIm?s~X(|DXnH^O~|umVBP?2&KcTA
z&jH;vxG07u6i^_LV%q~+p6%y1KD}0dlI41<=W}RppfcT>pI<5h-b}&OtO`h0fG0y5
zIi`k|-2q0>Rl+*Uki`sOQY1eR{uTFT{quY5GpnRP2JWuj<pwsS`DQ#hNA8QyK->b@
zcPQ;<RG>!Z{pH>m#{j6Ucc*LL_fj`+9F5tG_!OTqZjq3YCE6l=GKw!-DXjt#3aG8E
zzkDh%nP#b!?z<>=qR)-2vBE@gIQfA@@0nGYpP&4W)An;aOk(^B08K-E{r7`|TmsK(
zI1>+#k1YrxH$WcK%XqM$3dImQy$cJHe|+d12`@Jlk;jDoo>SfQ5-$ia4!eVQ_M{2x
zT)SN&sdYkZ<C)^B9R;vZNaD?;=);kgB7^jqH~Y|X;A;VApX~!unG1Odi#a3`zyV8Q
z09H=2q@ZIkEQz?A`ah6PFZ7?^ZOB%n)p8hQ%IQ!V#ER(nX(c{O6T^7#@Y(o_KEt>t
z`N^;k7c0|EPI|TwKx+52wHpJWkkBjEcso6P1~>z}=Q5bsI<fbHmV`ya*Bdk33E*k~
z5-FPc09OT#seR=~C>k&_;H|9pVUn_h5)ZYLklkCm<+Ql$|Nn@4>!_@@=xvlx1nKT>
zP(ng!kOrlZE~UFuKtx(VQbHP}yQBm`q`S)mX;8X^J73Q^-|rjaeq-GG#~tH37)KrF
zb-#PBz1CcFJ@a{<DYLRd{72>2faC8~<D0+Po!;Rx<1xQ5mCC)(q?^2#Ci`}0FAdss
zfuHu;cVJPR0#9Sb86^U!WM&E|(gY+DOj6eer{!<8D;t;)&G-ND)(cnZ^N(^3P`K}T
z{Q-<Ql<zI_5)iFOf`A0MIbUDugRMxWQ<L3<#{@2+M-uFKx7ofS0{=FCz_7qK^X$Z*
zpK!?HpUL`sVEB%A>I*zWFr3Z9G%@)Cv``>~+Paq(L9yy}V&XLj0bpLV81ZVJN-?8i
zbe(ONx`5;=YafwJ&m(A56g;A$u!6pW;HF>`jB`~|c!I<<K=%Sjw5ei8HtQEMn`1(k
z_`+Wi6IVk%-d+$Rz-#J^gTw{}R)lfY8lWL$0=7ghD8yILuL;29DrZ+r0E&T7V+rY6
z5xdz13xde3hbk>vxkOWy?daD5{%Gfg8zlp-G-N=WgcBwZ3Eep6mr3Mal6|n1233+P
zIsluF3k|pZZ7;;qMR?eVyKHbduTIlO_kLY{kOqNB%>tNEB0`eShg`sJwt>qHoLnFZ
z_Rnc1p2i~FZuau65y5~VD2U@Z#ZmDMybEW&@NZgH>J|2jOz6)G3kwqxCIFE?+-ol_
z-GnOoTyq4twJCw5ADvy^8{NL(cKYqp_M>4`uF{RGAE`plszmZ>Z&Q~)CN1#mJkqz-
zxBb&!79d!6_8MEKIMwp)ikw<#I0Fh7kzq<o2qCd^Hof#wz%!AnFRMbYH}~DXqGbAU
zXN_zY&#Z`&K7BtTwCeWRYS9gCd+@l0BOL8Mn(kc&4tfN)+V@&$zUb)?YCnFM%EBpO
zA4R>NJ$Cs+;_$s;GMfuzDWltS<LM?IQqV8UJsce$@lQS>hdzAWeU{uEt!>m~(iOGM
zW{V^}hm5AZetoZ#hY%!cp)Bbq0{9Gt`zNML;gEl6+%OFvNiz9diggR?^U&+PhG-<T
zAIPYsGSAr@qDe%~IA(thk_7a$3uop&^I(K6Udf3323n(XKg&qHpevX9S&ADP&sr`)
zwQTG`mTH|sT}PE!FO4{d^y^uqO*x2OiOqK~u`pOpll=QTI>XzZM$jl=o3hn*bbO51
z+jDzNeRqX-aJTCkVP$Rjzjve~7-Xhgr#tigbN$l0Xi|@9Eo3tH8(0teeD==eU-aCL
zH51<++Krx?P(v-FgvVCf=%?VxCtZ9B*ci?GcPgw1|AQSk+2=%PggNwMb)TRXcyS27
z#DUMgl1B<XzY+S!MPuTjWo1P}kO6hhOyM#zRx=tJ0e70Jf>oH^%wq3tY*L>gdEEaV
z00ryx`NgS8zqRuS(xp;z%uFdN*#Qn#KWY5!=v#)eKi2XN$#24Y1HJuZ(sfhxs1b|c
z4&rIIeN`LZ*z14az}pO74bD<4HkAVFI`__Ie8}wzhG#n@f1dXUeHhmE!Mn^dXWW{?
zH02@S=}o@PK$=Bt=fMAZ+;#Ka*oV$_ts!Eqf4_j7j7(#1MF%tNDQg(IP^;7($hmk-
z)uJ|E2eLaW&;HV&3*<4D{%`b!^H92Syo)0&{GT}7b=1u7Qy>S&6($~@?$tEXg{s3>
z2ODbxBJz<<6sAW1eE}}nHN{N-`=((y$8#S{1_)>myzLNc=ut;@jNpZ{hW%QOsdRbt
z;$+@)hu6@U{hx<1?ZA7@dHAK6N=1zC0iCz^=#O%QXf7QURr5VviwS`r8k;i1`jn4I
z`^B4-G~7ICm>kRa&%VVGc^f@uRT1246}rNNShYq(rWI1Cm=$`{-btgO=^(|_3@jbe
zL?ZtCBC?K-sJ+H3wuM_*OI@rhOHVO@(mg7|Q-q3RklXz&NHpCsPiNxc#T&xM>sya*
z{__%rQIVt#D<-|dd}fa#4i}$KG#)i0^jlAmf;kvoWT`}L>|kqc$=|?~Cr~1x{C6f6
zdYR7BvI<m{VcWGv;5H>#Aym|8sAvVI1A9T<HjBtP-+e7Qa?)vc{j_i8`205o)=9~F
z$pS^$Yrf5WXt}bYK!N+?5#=dT;_0jHymsCx&oS@tqtWPgfbG9aDnliRUciKg#F{4F
ze=#-PEw{)R;r(K2Z)$qy_a5zPH)~s>Ph>^FMb6`;V~pwOa3m}?4+sxc=EbL$wY<}A
zt4B+ly)^KSa-cDiW)NFtGo9d`R@Jw9I{N06VZGpgR&K`5GqZ;;bmzM5L&YvmQ_gj1
z5$RzfHFTF+S@9|SxIsBM^>KJg^xA9Ijx{`(p*~un&pvz{*rj?j3`MmfHRr(FSa4rz
z)6as~#dR{>71*qbxmEr4rD8pw3jVtm4^XglYV(H|hEU73jf7&AaYx@F4nP)A<A`BW
z*lu9wBXXi9BA`9eCH{B5t)q~kR4Uu2w>&a9sm1l)<3@Z%a!?Qih+IAfk<pSE%0|KO
zW;AUt{(X#3(1F3>l6_^aPv;1O{5tL>DH2vUYeJv<xx_>E)6-+|&qy7n21Y~ip*J|B
zvBgkNCUDU|S2@anVyJim#33$wMRY-8BBxz_VNdI*I@E4OMnzFV1q-3lepHU=W}ZX~
zAPGPS<-izF$r5i{V<G6szU~n5C^Lr+J<C_?-IZI}e{)37G<@(?=8~;v!@T3-PD<FZ
zF_LukMHJLjPJvxLt8gDI<CIN+NdUB7%$qkgG&C}OH`oIML7NQ^;&O8Eq?0iD5B+Ad
z%fLv)qW{z5|L&B9=fs52PM7AUU3+x<n^T2Pn5N$DASn6*g`bYE$p=8TKRG$6+Jr(j
zXlcR1z~9tV_%#dmeEFbS-RDj}lcpYYnt$I)9U9iMeyu9o$n$6sy5(Q77dMcxV#^>U
zJ>2~g2#Q?L)IhO&soswC{4p4hxkKAa&)CLBFYtPeL<<FWg-Lb15*-qn0sB{zxac2a
zrTd<+DG;RTAO@8N-ryaf(<y(HbW<bux86rxH+V#w-^2N!hsJMkVq#+P<F#vAz<?4I
zOcqCF2HH)nUpNDdLNUXo0_30}H;K~>)`3htT(bsWMNUqxho%7*be5KHxx{Vt<)-&E
zo1(>%^4Nq=m;K~_E?$oP`0=CmWkC%Ievb<xtuT8=&mnF5WPNMWqvbes1Sxd<f;14?
zwnBbfBGF(s;|QcMcd2D(=z5f_aBmFw8oBahJj^+ID;RrD-*Tnt1>I*YGxMwDJpY}N
zHrL2JKPd>@<%lAwFefgYKp~?sf_fX!+|>*^fbn>Kb~Y{xsMJ&%qT`Qd_YFkUN_o)Q
zy7gw(JmHSw??IB!0!n%W7!Z6KH{p;4r=oKp1OpV8nwnbtc6D{NQolZO9JB+wdSJYQ
z6nevY)WYY3&2N_h70f5kqs+!op5>hEE5t+npap8q@S*`yT!O^=_U-6#BSS-|JQLH<
z1OVL&YB5yMO1mT5&eW6xZ^a|-an#UFLH;j;l`(jlcVL#TJpz5|0i<ZUHRgSgqk<eE
zE;Dm@9AWr_K)+1hFOH4<1|AIJ_!24m;>I-$`@)dZ!jJLJx-gZ&oaL;3#@KG+?GOO2
zplT=$XoTXrJ&;u3=dQ!N)`1B`ZRJ)rKVuSVe<FM9?|tRX&L+5~oy`L(G9~yJ&qqN7
z#?8&$%o2!bRcI%VXebQRVe`Wc2wC6g7qV{^=T*+d*g7tIeo9PiK_TO3Knmu<OlNY6
zCRrfOlpA}@x9g%v6pcqtk@r}z0A4E^@D&cBcn@s7-1k7EOaO8pBFx0*2LIL&^Rmd~
zzms<j8zEsv;t{oFj<-187uv9vW;My&;D#kp3`}%0IP=W~lR-<UA<#2SHi9uDAkC2{
zOH#pThwBroZTBB$2n+D>@r~qy|AOd~K5u+xR@T=C64eLp0oMoQ+`2qT7kh8b1rC_s
zDag+!V+W0ZhNfoJJA}fBE#i~Ht7U9bh}=Z>-eS7%(wK96WM!b#@8Rqm9v*HgF$|#v
zaNdoUYXrb;lD#;C!DUr>^R7e<(${M=&t=!w*PmCLw}yibSZOCV{oc2>F3T?V#+Y=M
zH>)QExM8b4-?bSn@TOR5{k|9>PTG9}ax9-)m71C@7$1q~?9+3fqNL5#P>ulr2*yEV
zG&B`&%fQ(X(Af<nCrFdQCw_$C66Gz^!Hx&>!E2|KWn*MZ{i5&vK1N&xXegQlh>KQJ
z8a=HL{heLIXqR4oK~Fubmd)$yzCBl42*U(Ct;R`rQ0~pn>NZ2iaJnVu^NKyhKO4KL
zoSdegxf?JZC6TJ(dxQe-0f|hhcMXSC!J&}@HFP%w-R8787Cm|lRnHCrFG3?zq_)M5
zF;~N>466kp$H={IWwn}ng#-J6bCjSF$+qE8Mx40FiA;iFe*wb812(T$V~`KWvUc}#
z{hZ(a+IK1w<@@)d?x<Hf7#J8E22e+E^YUtkR>@*l=!h%TdL%592Xack){H^FUf{?I
zw+s7bel!<{^hS%c9+r@679=w3VhsFp1uXEtg4h!h<hKMC6haOcBljC@5X%*wke!TL
zwng~9Ic4j_m5%Y~nJ2b2xZL(KZX*Tzrf7+I4Tq{apu&tq3f16io~5$MU+K{EJ;eGt
zlCOlg!@s-hcn6-hQOWjODi(`{u=>ANIMVEt8N1rm)fOT9gquy=mjpHUmgHZ2LS*VX
z_d9+K=vBAY*3>Ik)3RLuJEV0^?$<CNR~o!QB`S2RoR*#~`2-P?=Sax*uI*3EZSj^i
zu?m>s|9#32kL0{KWH-n=O8JNaSrz}ikAeyi`EluN>^KRA*;3y7QmX$u5S|<I59z!-
zC$k^2V;>TpgpI2RGZ~J=xl`j)fHjHI5+h2NNd&?Ei$cMDtNOZejK8mAc5$ud-u{m%
z`<76Sip4UX1=UBnynxWg+M{4qsP_-wM)lldr&!p={`+a`k82f{icVWPo)$LCeJwI`
zdt=J{VDh)D4&`C){)rfNRds&?j$iX`{(tW$n>`QnVWRH|Cnlw^(<&~j!oLsD;jCJ)
zJb>zz${D~+IxM7-7m9GLK_aWObU&E#HX=#)VP92itu_f~KMeWjE#+yThSnBqL=F8s
z`#c&f;w#<q`5$q9osB!o=chQb+5_x-2Un057BqQ?A|80J$o-kwf_C*hnw!AxMo-UA
z>v^)jFDy7YddD}$#|$++$J^6xeF)YE;(?VY$Yc~R#UJQ3@UOpm#cC}^_0J>t=Pl3{
zZZVCh^s8_`^FW0I$r)PcOOzF!V%BLl<*vhT;5-MPlp>YWq3f5>%}6vF_iOt1k~m}x
z;{{d?DK<G3l=4f+Z7&dRa7bkC=I(KiSS?~j=$KYNG$X?t%s>wQB2XP_F?L_xL|j-|
zRpp<DJFsv+=q|rkSr%{mRb4~+-=TPX15G{NikM5EEKss{#e)Q~+#sLM8-oODR&(LN
zcQgA{a^-pWC>b)@OKhz5kSLOV5x4d?Qw0To=Qmmg1vRCu>4=fAx2woy(g6)(f&{Av
z(ThH`wa}|m8_B?ooGLv>98ZzZj3}2@2GdI^J_{Mz<NiBTk(gwym27U|8IAihH6Hf<
zNTJ{Bb_QO9*prZwvh?j6F+Wdm8UvUnHX;JypruqYsJu7sTpejql=GV<!0KiG@jo}0
zb1=BYec9YJ@jPP2>mFPto(>(T7$xBC1r#J@1g8e5VxR%o2d+Ledu3%}LhKpLg+*A!
zI?wJ*^8C*J6~4Nd^Y;bPs7J6gRR)Sr^+awqNeokd=t9QoM$qGiQh2*Qe2}^x3Q{@P
z4bP7dXQQ0uP>KLfDqHKXAAC&k5C0hh6^~o$wlxt}tRt5`U38u|OxZGrjkL6YXSkkT
zhZYLj0^p3+Y=YPhG!O{CDYolIeeq~pTA>uibF7AeMN8{rFzB!3amyu!qljnHIZc7Q
zeJ#NSj(lpjt{y`|2}SZX{p+FN4F}{4NI4^x2=VYnfuzN7iN2%<4}V-%2X!scC`dcm
zaz#<uL2GP&hnyX@G^U6CG-s|lp{p~CYLdy?Pf(9H*Dnkc-ZOHnixOOm-uZ!RI{Mnx
z+ow}1X#2bPpOjabsSf#AD3IBKPn}9qno(&-$NmL^dQME-J?`)CkC2Uhq>YG)0ebU}
z5@$$Ahzm(UVIg6FGn$6JzP^-HP#gnvfS>wWcy>3KvOe|+<mMOeCKQ`vi<b8kYJ<Iz
z(5Trtq<x2bMBnK+1mB_e&O*4N_T%t?Z_7%mRi;Z#5%@?!_69x9s0;4e&mt(R{peG)
zfmi@Shf6$X#~?Gp=-W-}{i(rOv1%ELar%6RkwX)LtrYh>MY;XSq9-S999OaN4Olq3
zo1Gz;W+5dym7ig}pfbkp50Vg|e(|2in|OZ<0eT$bxF@L2!g_UF^y^)%tK8Apx$g52
zL@*X@R&~i0-A>utT7HLmU+`Gs?<E%>PDjyj8@F0JIbA_-Y#{7Sj*ap1@qz82X;n@*
z;k^EEFN^C5zgF)zMC@1|%iHjcndFls|6mZH(z^85&83AMobUt$wEyrjGj~r+xWT9%
zUnn4<kr}bGnY^+J#UQ@5*=XRqf8z%;BBJ;ljNWYJnsU24^Om*kq@AEZ#1~;SqqYSY
z`I^$hLd^n|_o4>9e{9w=Vh5%FZVklrQt|EQ@~J;5X?njnWbq>|gIMt`9&Wj!$N9aX
z6q$3Kw5sul1^-I^xcHr6ah64C0>mdwl_-t4zfX!-H)vHm3G5mz4w1V<LiBqKf1~6Y
z%{(RNoS(BjW=|K!xL6y)9|jSlUYacdgfit~VP4-N5y-B2|L>hFsmn}$xhU64ruRGf
zcs|<+hFX6=jb*tM_PxN;xIdJSOPv97K6QQlc|e3gCiQ)HH&8j+7z($E?3$2?dyP=D
zAY<uQui39S-xb8B|C|-{PozN#{#_=LCmFE+tfL_W>Vy~RkJ<8S6=@J8ARzNNe0l!|
zEIG?2ZU6)L@yBogD`A#d0HGfVZCies@b<K)W@hNU_+Ln$|Ji7fu;l6;M+P1bMO!|x
z-|C|dm!}~Pwj9ZHz{Y|!2-p&X0|N=SS$m;=!F>BRc1QMHxT}v(KmPT6`5i(H+Fz62
z|Bf?Hm502QVw)G+xrqOtMm0Ky*hoS$pm74&%5x}Ra=7ux2??!S)}2Gvce-Xq&w|vk
z`!uW=<L|bBzyvl0C7uKf9=uq|85mx0a(o#p>g#Tr0t95>ZBEgG6z==Zj`7Dzpiv{7
zmn(^F$MY3!Y;2lHfba5?LU}%dO@!6!L&Sv=S8^?V1JMPc)3~2ZH-qg<FjIR;gfAf1
z*VpIeH)B4&Hn)s~#xJKIwWgapIli#iR`T6;$5^p9?>kbcK>+0cOECUXCkTD*AMg}9
z;7ZKO%A!LDiXl09ZgqZKc!dSVA1`NntJ>Juy{5&ktKz}I2avHQiIvF9MTW2X*v9T0
zRah>tBcW0-V#_6V_gx8<J@|@)yG;0?H&6b0C@?vXc9smNabU?if}<pUdklMoM)Imf
zb+QGysx+>n*I?c}(OV{dt_O;U3E2&~ZYje8Py2h3#y@G{qP|3?;jyuewY5<(F>g14
zDT*jAoZUl0yDgI#&mD{LzUe(Ld!={lf0ML7(W7UzU;60Rpz;^7Tjnmkd42;uJv}fU
zA3o$j$;bUSG-5A8mwLkkjfA0s<m2N*{6`)Mt03=LV0UQpA4)s>uha6x>0c0W1Cb+=
z*E#lALFm>t9C|zdct2Py+E{R(jgN0XG6~}GWFx2y@Zc9Aq1%WAiYc3o1yS!V(QwP*
zLZ8VH$MAFPzl#B|lj&y0@q&tgpE;p>!;5#Deb)#Kq#0u!XDY~w`&it38r)WrL!-Li
z*OC6ae)?-*DPQ=fbkH=rkB>;dhENk%Kfbrph|7+O*@kjb^pb%6)4vl6X%na3cg=O>
zI=;^}(@}LAhjtd&$jc5c^-)kU|ND&TNtWNe)1}V7FMzo8`SijDKfS*ey8bf{&)@48
zDWB3bKvMsFc)Hxyu1>edl?&Hn?)hV8K=`E6ePm=On3dQ|r|zbNbt~mm{X0IWWefT<
zZuyrYPBao9A`lmC*jsxX%5TbA=_qyN<jH=D`!_Oy0Z_A399`eON`yVLxmO^hH1}zc
z$?9Iozo^u@!)c%rQg!@Q(1=T$8*zd~+=%WmwZ}o|<kQ+FS{*mWMrM5(_`H3(JUxbn
z!GKG`%CyUs&_zd>=06Dx%)h+&D7H4qfHr6Kxf}A=2xCQ2cxLHh<M#)9vJ-0)y95nR
zrGGc@{*%Lj6^;jKxi)E{)Yqehx*B!Z{a;Qps>jdBg?rqNqAbQ67GwX3*>Lc`Y?a{C
ze8n1+1J=&kI*whkh=FDc5j|to+xS^{%aN)ktUb}@N!EYnpt{HY7W}NL9E%T}zDrTk
z>RmVG@hnyzPAH_pr|}XbZ2$njd|vndKe->6I_{j$e2RKpk8Mfm)D|L<%0M3rD<NIJ
zm5p13b-H0tO3TnGpz5vxUtuv~*c-vkCm#qlpqB#zs^8(14v;5xeWj#6008hX0_@?_
z#QX$d*wri<BNbi?pQYW1<ZN<Y-X_?Cq~ihpb`zIai*{--8Huij{QbYHt^j=Nftq?I
zUtAYUA+V98V{En#v{s;7ix7kJ>n!+)fQKQ%n~bd%9}t(xy#A6ye81E2FY*)}rd1kV
zg4z_#@=_tw1}7Q3`gK6-c$R~HdVUUW3vY}>goPc!Z`Nq!jnIDd-Oq@<q~BX+p_1Ki
zU-0usTDB$DB_ym+b`O|GUbvVmw+D!ATT)eH$G3Pl5i1Zs4ah(SPHo-qXQ3Jv78X|P
z&Ts?jzqvs<An>pWQBqaC!ISMHU@pTer16EiUwr(_CF3s!cvk0jd^l8Gc^Ou@X(w|&
zrQpb4wi8B>F}^u6a9EFKBO7!{W__#fp=t3w<nV)%wb4dz7d0_pS74*A|0oSQu{Ov?
zL_`4IGlw<~cDudgFNxFv)oQ5L%Eh-Zw+KAep;?e`E_J$Wq7)L04gYb^fdXtx*z+Va
z-R_q)#XVUNj1i$T9G;XaM7XZL%R@my2`2$GvivRVF|vPv{ZtEaQ8t#xR0-wJ?>od}
zS$Fd-jUh--P~MAG{`(9);poHx_liZYeRw38sW2V=*W$I&4wT|Dm{8+yRM}c=+6TyO
z?hv#wz`J9D=R?NV<7|_pX)&2AG8@l{_ou+=$ZpYioyZzU%Q87%l->C(<09@_#wi+1
z5@&?9iy<7}5q!77Tqqq@9RO@9BrI&);u)<0mI;B&z4(7Gy2Q1rP+em2=*-6HGK*kS
z6ot%PYa_QdU-ka=M;vPc&ZM;Q*~&eL_v?sI!lCKeKzEM~)>H6m0PfW^U_S6)B1E{g
zHXd-5i6hrw9gxE&o^gvB4ri{`L>!q4c64+3J=^!L$uxN^4)}$%6L@S9BW8gjWNbg{
zU&<$P_rjr)QlRx+x+)KSY(E|kQTX25fyX*uo(kqmE}jgVeKSnkF+$+=h#P^ULB0`h
zd|OykU-s2*6Bm6f7A{t`(D)$FhYy;m{zaRX0keW!Gjx~#eH<w?Uh}zPfXm$xxqOK$
zWwwgi2xP4JJGP2Atz6MJ31%nix0nuA3SA$z=4=1}@bA|sH4B`czSJEXPjvr0EKkyd
zj0`CV7W;*_##39q%3<-BB6L*((Vt5H$8M7C`kxFYv26#(M<va!n+)AGsq&%7p(yG{
z5hP(Z_Uxq?+w^_+B{E2yQeGNDH261)t0~b2Cwvjlx1Z$USc^MCM&p@PGVA+k#kASP
z`9`fT#P3Nv9ueo#<t+rp|IZ`gjpg-j*7SPr!ukA$_?D6a;2xYT@@MOHYioHA#70Z=
zxZAhh=w3VgH|Ijk9<kgC3*|qa9GvtLXPUj=29#*lamY0GSL*Q@ET(^iXyRgvzO(6h
zF;T2RUiLqaR3$alQ1+s1Irl-J#M*QqQYe8o3oBHOOl~f7l87IguJ%Vu&}~J=?OaJ}
zAm-EG8>GwqK0c=f*rBWN5o->v^ToAiIZTPhbS*{%CUdM>p6d6$L9J(0>;eKLNto}o
zf8Q#=*pY-QbF43JCTH67al;AH{Hy3V3(v=&3Z-1((npI_s%*RV;#83Y3HxgQGk)3h
z8H!2GXY808ayKw%+?bH0Ba~h(KZ*EhL@wgAI#Avsl*!}#I+ZZeg45}LyF~)qPS~BN
z?mF97hF&l17L~ezUr<nha7_W)re3SJ3ml9-1<~J;)~BW?h@W!#F}X^;wh$nuJL|3f
zU`Sr<e=JYMiPE2le|aTfKGf<xGT<e?r6GcUxHabnFe0ekz(53SV<M-)PB1~5Am3y-
zvhq3xwy5KMls0{bZxR^9P-O8z`Cs#{W6hm>?1Q%jFOx`82kY4*`%wji$_)6M-|p>4
zc0fSW28u2m7aw&<eo=c`L6yeVWJ1u6OK6Nsf{roSLq9k0VWQPze;%C|w&UXe44B9=
zkQo^6&pN9=9asJkYSr*7@D1q%W*zYus1r?m*O?D{fF^n4`gO5uh_x3N=++s}yg74u
zkDcw2&vr~)gG|(cJ}q-a1RiN7iB$+j7BwLYHG$mLir>lpP^%-CUR{9t<Ja34@_{!G
zEQ|+Vscrl>z2&T+YcR8shm^g#cY``J=}v*<<&5*#t_GJtv7iLjZ_uS=-gC)4dji(p
z;C~wVt{rwkAcB&s6@jA?JZP%zXu3$eqthnscZ*%)ugeBJz9yG5IV>}7O+6K>`4*N7
zGJwyL3Avg#1i2rqje%|)&e31LegzIuj{^SoZ1s=4Sn=kCmNbh9hG*XZZ%4<#;MI5n
zZa?`|RXmKc{E#&Qa8AaaiwL&D!usu5Ca(tNsj)x%?-wy1Q4rpmJqjL5<y(b)PA%~1
zPSYxMvW$F}<#-7|?J@BN9CmYMm2nJ9?(8N`c<$>;c?G^tr&XOn^>jt`Vy~ocX&RFf
zZTl=T80h@sYU7^7@r$a1vab}(P(`|YfLCFT-t?x3ni?!spz%;{kj&{+dte}=Y#q!S
z5RFl?GpMSL0HpC{+i1wsJ_-qKL-XY9jG~yA;-0f9^dE?0@P%4O-wteI;HX#BWA#ju
z;Njk<sP8xAUvPY(-gqZ8AkhCQ*aMi~yJLA__XVD-0fPkm-AG-omfkW!`aJdC-;a@4
zXd{-82gV>fI~$4*Wb6z;^w`7@K(z+`3M`1Rr5?0@LT?G2F63ooW-p!qi3v2n)xX(F
zW<cZ5?fbvZiGSa{H2F2OabQM)g5s`uFWy<8q@?87-S*R;>EID_v*n?=J2*ky<a-4y
zrcA#bA$VedO28rrzM7*M2quCWGEg3(<hLGO`V3yX`GQu==w#qIvU8Gpez0B(N`A%_
zb=ckmul{&3%F6hQSy@@Z??a`KO~f%U&>jE&-L%jGSaX5imnTo2;8F_%z1?+01c2xB
zLB5fDODr6&P+n6(LojMU_-L9R4m@;F>RRWALv?2X+8r^<PQv~M`|y|TC1CP)_w?A|
zkpZR8rLq4$xy+n-z9<8KZi#t*INNTK8~j$po6FLtfIn4f(hhuYFTdj++Q3A5^yhT{
zPe0ijI*{r85lt#)Tv=IxFtn7JnU<E8mF3jS#uhT}k!GTG;TXH-_-#QZm+kZKuBv9Y
zb2Ln-L^RUUd~l&!D$D!vhRwYHI-^kVk@EbXwHO>p^~Et{Hub{PB23DW;o<i`!su`C
zq(1O+f3~pAz{n`gy$Y|f5urW9!NCD}WXsdjnwUsUP-fba+5&!Kdj^UpXhmh?<n#rc
z%|TI|Oh!ofV|jV9%1qjK69{%QDnythx7I)!n<4B0ZUwOh<54k>ft$85<o=M|J`b(A
zrvZNmY<0xZEFupZa$%H9914KURdWEeuR5nMnUS=_op6>cIpzlaI0I0%rGSKI-{K!c
zhYm8ps@mGT)*?WGpdd5ZxVW^${sTBn2T;2$5oxlpw8SC5lMgTD;I#c>Px;W&#Kh$6
z5MEWI{T#qW2^5uva6SYLR?aK-NDv`P$v0F1X=yVaDC7ajG}P2fInp;Cjf4KH{PPtY
zAFJZ0!5CxPK2fTBeG1~hULPwMec5INNT`48?yi<;&ZcFL`<zQ@PdNr2$;MEkgY>!1
zacN$-4J<5dEwRC}{9)EzCmOyE07O`snX5m-#00yP;Xx!P-obIwDeIB}V3bupDxZ(y
zeqTgF^ZX4c%QY+taZkYztK=9=<5KN|sGhwK0yRi-N<xD4QRmio&=*dfsFze%S3~*+
zk}`gy5r7oHxm?&a6qfGk>1lyperCAu_m3ERzhG||yDI^Pyzq9i3V>zh38a5_d~$L{
zy)tTku<-)h&@C{&^oV|o7Lp4cD>!6%A-CLQyFCALRibVPFl7ZhNwX#PG_;+6yuTCy
zP+pbO7jfG)gpnN!c(7#jvX-mK%-_o3z_il_kX89`2VA4vAJ#`)ywZJM@x^bq4dgfr
z4INb=6MobISvB;2>J@y#N|@K&rQ(Eemu_}BhJGo00IHnbwsq^<o|F4{uqj|WdC&Rl
zGkc+=%m=DJk~Se=gGB`uu|z+9&HG&3;<3bD!b~9WEhx`(R<`azZMSw@mN`0&rdd4j
zQj}c4Ef^uPHJVT$(yBKyG6Gpfk*z@?p==c!q9X_O^(3yJXv@SK_@9Vyv}r?Ev~t#T
z?8)$_(jfiUxJpF&B#dYX145C;*%OJY3uy;7S$XDP<&B>ajp209K=0l6olI?R@l&zw
z#Hu+XeIhr+Pb=W0sTR`8sy|RrpsN#cU&Etx&b~O?ACXx11=lAnnYkf&`^?Q>eZ&O>
z>H(kziG+TI;iN*bn*3N2w@Id<ihKti71dK^1`Co>Ml71%@xJET+FDMnq7DwCh%_x%
z?}y!c<I3E6sHi@jVtI7l+3!U;UA1B-aYUA+z2-$3Y$ctLLqW{9<f^Kqbdy*p_YpBo
zz~!mcoU3Dt?UMQKWD7SN+o*yHB<L@<Kd4Ys2{7z6vu1ifkKV>(f}>wpf#%iorpA^Q
z5pqUl{?rE`RjV%dBfafv<N5<^Jd{Ij{2cd1MmRz)#w$2JV2=LA790i^k`agT;3LpM
zz%I`L{LNzafp+jF16^K<I?)DRF2!|Tj{1^YabJUSvsqrE<Y3mbDicROulJFe^!<#3
zg_U1^IA@hr+{R>^SX^9ibeE4WVKDlTkxbOuYAb3@W&8Rh7=cntX=f~gy_Ohz{6!Ox
zcwc&E2AtXK$tA0l$A7}jqvr$<g|st9bv9|1HY8MHrn8>_u(^V9XgvX#Q7BMv`)pPB
zeE<F(-LY?EYAh4s;@6Nd?T3Pj%wZ=S75*5C4Q1K|SAo#yZDY5@$g~_8b&hn(3%qMM
zT<tXkZrvK3$8A6P9tAw;S9%}7{y>-E5N{xf(`$QP;I=toea=CU6ZA7HliWUeRm3S3
z@~R_L+yR`eg=fdXP!r8#nsuzvX5wciYx}Gk95%I<i&Q@bzyf2heiQRd>bA|~PL<Cx
z=tv1XN{=tTT>be1s41C8X*3R}jqPU}JTh3@vPF9CM3XuNY(?4yl5B+9<8N<T7@-JU
z#~S>E*sur)a*wFBBnJEYb0v!;lx5`SN+{@3KL=j!(?%m5{rc9$E;6ZFs-g;|<+Bh{
zV^jYdn{ungg_%Ikrd7w&Yqp&MLl)@94O8-C!F0l{hiUv6H*bzd@!>5j>wWyWt4o8R
zpg<EATXAuS?k^T$->j|hda*LeU`I|?RIhEdHp!-ISMa&k`P=<m#%4&smSM}7I)Du!
z40x{C2tj2ku_hhD&hTuray`wLWO|1ewiHOv#=0SCDvr~b=zB0*mg*pT{bk81lwD?H
zVMsDTjd|t>93N=$rm3OP%SERC3KA~Zo5?Zt(kbaH($mu|?WUm&P_{sEvk9_A-|Fie
zHf4lL;)~!hR{8!cvR-p?S93R~f=Zyqo!7kg6{yaO)@3Nt^o?CHdGO+<OLcaS*uX;3
zmO3OOV+okGMg3~~B`lY4rX?yv3zDk;4cz2}=-Pyov!m@6+bIrD)cTSalKA-e=63GE
zXHl^40j)NXq$Z*)r5VXyJS38KD&yr#tfS<xup5*<C$Je7`&C@t^!W<-aZPRQx@G-c
zM7an4X4~W%_oG_Kh2{A#KC9tg;mO(&@y+e+AyWefvEnG1xwtoP9-!TqjhRg^(6lRv
z@v<Q#E8@=ftllN*eo4+(G?a_EVZrmxBB>O({rb8&E|@&Jc915Okz?cHvPM0ckf;ym
zLM%>81BIrhrs<E4oZ#pMh8fcogW+}M<j>x`d6(!xo|~Jin_FAUXRsid{Mf<K98+K1
zJasGPy_-h?xzMRyW!qbM13J}g!zU^FU>+sl5AeJBS9f06S|Ehi$Rza1sFxC&Q(Qy4
zPq)d{K+z3Ek4d(ol>)X3lXe<XQhCyidB0pt4~O8l&pzl+ZfWrKe<GW+kFrly$(3BY
z`dB^bp+S3xfvAq=du>vY)l9A|rbh#(ghZBMTKR3W-5SL!;7*l9QR#kpQ|>_5thrza
zEjZz7g7pyXV9zfuh<hf5Iq?OTgX2CJ&F$~+=WQbPc#ofuy#hl_{ZH<D4ahDhLkzfE
ze?32BM5DMC3Kb+5VCQ?iyq}>*1G^14ZkP`+5^O|@bb2()WAa_ff;mIp%VsxM*P`AG
zF~8&jXc&|L#7Un5ZkcWExrK#=BXya_f~>5pNoEj!=4F<ydHO@^6ma^jmhULi*<4@m
z8}Az>LxCB;*~gk?{z!F8zM)e1rQlSZjpo_lmcG&LJdwC|ILbwsSplLq=>qP0zSV_B
z(A<b0HpFD0<ZHAC_)FpOV`9DFKggqy1w_ax_B&8eZrvm>DsrpW%9(7^cR)d{b!pg$
z{q++teyepjHWd6!s_%v{uf;mlXJ&udN=eL9GSm3rNxkEeFBGb_&LG9GY!iYU>F_t+
zW1&u+qb&nSou>->dO2sQ5-erC)j;S0np4A8?>MDIl4slEvH{8kEBUerr@SOK!J9;c
zgwh_7XH_31fRn9Lc=Z0{J0^!|TwGj5Ma5*kt<?a^#yK9rsC)r;7T0Vzm8`tFw~voH
z5tuht#UT1o*b><vV?~Bk6i=!PSK}GTI7ovcMw5<O*gbnsm9)&bN}iR0fdSTecp!8f
z)$!byr}dY)red@QIg_Tkd6(}2GZ=Mtb#(>rusVUY!>$7v&KT)8Z2xVkkjEK7H5Ra+
zb?%F2p8QQOr>$dC0W2<IB^jy7ukr1?R1C*&9nlZ(aGAUpHS4@qJ`}!GaUGKck5T>G
zVzUS5N{HO_tHszdWqq)to2oto&pI3$(G(>Lff{&KeS?FEx8?mb-6w1F&Wr!l8)KRq
zsITY7%l*;Z2paE(lbPNbr1^unh=vrm)!#KWsS`okscx0ZPE{hFsGuF;sAkn~&+u6A
z=8Kv<VD`JaY|Zjo&JZ!y{sse;Se_(V88P=el1@%ecO-2!roK%>UXjopHx{=Pw?5~2
zjCJNc`H=3%<A!Xm4qezlJeYgYFap->tbRmO^{Bf~5;NTJ{t<n5)-c{)@AfGOzq5or
zBHx8GfF5|l`%kW_Ld^Es+6hcYwOCc|2FZ{sXjg|25T^<y@q4V_V!wAUW%gOk%Z;cr
zHDYZ^TL<<*Qj@3Zn3@+QMgHVZm^bLE4!|H=bBc}O=9IJo5q^Q1aGTx!L;ZmB?Yo=d
zvmH3vzhF_0u~P9{;~AAQqM0*(Bu*5E1K;NL#z{ix75mj~L<}GmtOcADysnq0CdHdi
z!5D(?>522%uNPt)gl6$f>UPS^2^4gK8{INo@^2*XA|y=nUmi!To<g|M_|2%1d!!?u
zcM}L$wu)|^o()_!2&*?vn@KmF-9qtuoHkJDs`sS%l=J?XN8FPiTDTO~jYeY=6`YEt
zu^V8EEoXd;tI0=MIDCEX1kMKKEgM<`DbbpNQW;?asShUUAEWiY>T2l1oW?H>MuojP
zY6dnE_$ldfFLQz_B>LH}$u4ajq6DLhy#a1G?rhA}veH*z{r}+AH9vZVdgtK9#7zrO
z{=2(cm^mQ+<n-9~CLiD3dvrW>h(C&oazC(ev~stm<GshrL-+R=x15c=yM-&aoV}U5
zg^a}$CvyvN@tgl%!h60=zb|It9$woy=K<<aRquDSYuN8Hj*#5avz>y4u;bsjrd(W^
zMH1LV$&Mwn3*GREE7CD9uVm}{)<YL2f#XCQrfL-(O!)bkW>?QK^4jx!)mQH0X&;c$
zOU+JBgf9N*DOs$0=b;u%rs2r9i}|O|kYZ=bNoJ!On^2+rnUD-idq0&Lqj|P%V)FeE
z->P-SUnShKS(>_sQ+QcD9!HrGzdpj%nrhVk>-J$*tOV|z+lrB<KSoZvui65sE=7J_
zFQL0GEOK$akH?*jmeWR2hugWvb8#m4RN|uBjJuO*iSfZjHW2KvwgzYNf6yuRjO#zS
zNgd48{lnA&PuMuk{bjzCw6H)vYd$f>IB)#tmyxTHQvt$$x-y^WkT}hyeG|>a@})CP
zOv+vg^3Y#ts{2y8Q4|=C5*0@-c2RKivzy+Z*b|jb-=Wj~l(ZVFPTx(gp&-zMqRpXK
z_Uq34J+o-zlwiJ-clPpD9j<eG12y)mvTC8RpR5`6TOuk>k4~tb<0GF&GTo}ovgf9I
zXNwa%pgoSsT3R2hpNy%(Ry1=dH*CI;`CKL2ldnBpj{Dp=Y*&}F<B@C1uINvEwldTh
zbRul$2VpVtjW43ED`O0<Hs8{t^Nq$<E=g(Y|6mZTYeS?|DiYQFQjWQZODj+04Ouz*
zlr0OPl%Te7mC1+Uo2!fSQ$CDqMQ_?iy0u3;uP<Ap|0%^VR&yzKx3m%#;bQHenx5Av
zKd7zpe8hV7;k~3_+IZw@Z#jdh)V1{!Yo*)%vUDL&BUrw?w(z^%0WBlmN}Te|VUNVK
zjcV};SdNP{N8oht>L&hr*T>TTEPTo_Nk0*zGu5JgwntWk`K5;4)ORVo^oxhKV@jAm
zINI#d6{x#P_K<o0JbRH}Z(<|-c)at6D}PhZ6d?vhS>2nML&{F48Y+#9x<|X#LNDuy
zIAnt@!inUH$$2oU@9lfgcXKde*|2PmSQ^v#WyiUs6%&Rx^A=7rj|vEjd16@eHAXV~
zPtpwghTl6&;nkFnyJ~tmvhqLy-HALs#*Kz|uki=L4*yTvfe?IgBcT}Ug6W;R+CwoM
zcKB*!2Je!xNUUpa>IiByZL=YJu{~{iL)1}@#-v8L<1ZE^Q05yf=oZLPPghIquUhxk
zps@4B<EMt#j4dl!CCJ-UjN05UV*6%3Tg}k<44iK8QlDAYTH{CO|C)X?#Xllr*mOL`
z{B5#+Gp00mYVA0NM(<`Rw)~=`4L!>rHROOf?rWL`BTW2nEQco3W0Gd+Oa?+(c<PfH
zr-tt0pD3-6MMlNo_r=#A4;bBu&D<kOKypryuN~u!d&y3lkbXX_7WL<PMTRm?y)D-U
zmEVM>&e!GXlo2y&?tAqkRL<=keImCXbP}S-7t=j2)p}3wyeHV<;9(-~7xdn>o!7Z{
z@z;AQ(Y27*Gjb<SS3(Wt#ey_*bwm`8xAmFh2zZCNXQBrS30U=UqKc)=*~&lWu2cSL
z{r>W#qTTrIr#JLEzAkBwAqTt+Bv$Xam@)`>33Rk)jt9*jAL6><7`nZ9>ai@fAM=f9
z%*nI-Dc6IR)Fw90)^^vui9I&DsCAY$UOtLi#&ShJvVvbErF~S^K5xAgqGgmQs#uHJ
z@xAB-uzq3WzfJu0K%;BA2DP}UcqekQDycyc8;yuyl28#l!8vTQYD^>{Ey+zW|81u7
z7DZz`g?QLdekkgWs%^>()_98d$=mrq8=8fRzj3xbw0)&<$K$sug8-KQ4UH~55k+ii
z8)lNi7g#?r@<Zu&C2h&RE7(RxW?>B7nes_|Z*KBL<Mle`IF;0b$UzAD%{bJVW$x{v
zk3F=XSu&Cc4Dy+Y^QSKyYe^=9_F`Ad3RYbDF>^y_up4@2-X<Yg2x*i)`1r?H$;$q~
zn3;3aXUWj@1@~Y#OG$JsmcV{grqzSuUp!`%nIW8b*3$K}b#^%K@fFy+Ykl5CQU7wV
zR1+k*);5mSKhCyED<HipZTT<*Nqd}n`&Cwa{nppytQ!S4bVll=1Af*G<D`wpc)l|Y
z8)%p;uTiO!mnbb7sXk^~cWeFpT)fAcSw~S&&Rb51JMNfmmHE5wu#@=1V4}S0wH_}9
zj{}dS&ns?+yS&MNh(;Q3`VXVoiw}etCkT9RiE6}b^Gv%_;8XR=wzM~5OI+0{O*|JH
zJ=k%9MTwG{9j`U;&3FDEhME&*+H4He6K;A*G=fL+!qQIfiTJvCvj%t%zZVuO9kk}8
z_!@@}X1eZYzK&LDSo6`08R$;u?s0A^Y@mHc+OViF{Pt)@em-UCwEX;VTx0ZYBVplP
zHDyVjyN##)A8s2duu>V;);#P#8Gju&Ax<vcwy(lXx%8u$-H9*%lR^5^{lozYFA1Xn
zwSbaL!^`5-Z`xIz<eA?nxeIG|k7K*8^3-y5|4;ktJ-+|j{(6t+?mycr-T%Y>TCX$Y
zxWI$;{+#^)MO`)`pNS6rRsK-b(xYX;E&ehwV(IhCF!dU9gNE8miyPfF5B!-VE){q;
z)X5JEMBcQdCrGb&(EWOpK}*zogR$MzJWkOEX$WWXRPoO&#S^_xQfL%E-e7U>)2Yro
zP+g0A#)j%i8B6DcV?KLnqk1RCg2prbK<vkITYv#x@fWk>)2Jxz{$A5K8TZ74aH?<t
z=fyvwftwFpip6(_mllbYv!7f$+${Fn;SatsOps9gv;=8{x+jISyZh|-&&`DTjU&}#
z|5ub&a`Gw@2Py4hxYK0y_!9!cJN4utICYqhf-D{;PYpO7zA`m<mz&>dU!s#9VIY)=
z%#<0LS`@)bhH-}j(>Ra%yW)LLB|`-=R+mYtJLz3qoJvf*WoFU#VMd)|N)JYlT@+Lk
zZ*}<7Vn)~apZNG_)h3;f;nG#6E*b`#kMfrU^G8*F8Yzf2XX<{exyLd`@YzWj&oQ6s
zv$b+!Ke|~rl&K!N-TG12&n7c&z8Glk{`MQOoYlCF<w0}<k=*NsdtH@$6l#RuJJyKH
zdWxj*ta@LUu@)L(xvdZ8>&A-CCorZn-12!I_L=jY!EwLH!ypd*&$q{~iQU{jKYo$C
zL5jP3JU$S4(4NX-H~;>O>tl9afe^!}u+s;dUoKq!?EUrv$_a1w;x_Yt#d-gxtxP5H
z{~#rd!6cxaY4Oq0BC@yax<gFLR%x4{Y(V(R<$c*#G4HB23GTWQ@SID%=0179cYtiH
zq|(@nY22Dnt`>0Tajk3nxRbt2f{H)aDt@wUoCr$#)2OeR<yt+<l(90Mj^_r_!@=>W
zdP>u4_Y_VPU$@;JH)L{1Qb{h8yD$8TMStF(o`4N2<(|xUwXKKEr3HntRyI+W#ny?p
z9Pl;y?c0bYi!0(4wftuAGA6%$+Ald+^B^h5qzE=yYw5_1p2x+ezrLErB=~jaYZY=C
z5sO`6$cf^ann9zQN01o$e9!N<UVcVr`UiHuW*96{Hc$U>1jVbN8A=S=ya_HnjICaD
zzLMh3OQyzDw!BDAcYpt^<@aZu%doP~^E#Iq7uUV)>&yR7i|($#|13H!A7=}0UDe07
z7Ej!{H9a1?BmS=FU}j~(tz%;j>8_xV5VwMbjg_@K-CZ6Y?)x_GZfX{;(oPP}PL39i
z?sWIKrJd}ZTs57|o>*|pSa{kzu>iKrf5|expdjC!|2N6<R>qB|FOO+^>=I^Mj@Uq(
zR>D<RZ=gL{X6tY!pF~nnG&WtNt<Qn~7!#ipJt7W|iT-}LjJYC1jn`h;d0^IghTBg3
zKr@q9)5!VCipTN5zUcAGFZ<_hfd)B9|J$z=L1=RW+zl*rbhdulhl+|&1~>)m8^mAD
z=6wX<<P8l55uW<{ZReR{ewM>;QZq9v0&AhBG9In;;K8kz%D=!dAQ({1A4YK9iSMAp
zu4JEsdLRD9B8Y3&8o@*{Y70eEQxj}>6ePgcYYu8GK)7V?F#>u|-wuLzn{z$`9tu!L
z?E*knN15#c2#q0Lw*f6K2Eyj$5uA353jixnOmJP?gYs&&UF^%3@ZEGqHgIHzX4Ahm
zkddC*@Y+tYB?}z^m@ad(3(WG@NqfL5%f*n(D^=WIxP=wUQOQh-o0Wh|<VgT9uX0!Z
zt#iLav?kfwFf>u&Vq+`v4hcE|6Rh<Us9AZRBO!HYg-1n|>Q+C7zAL^*TZ%UCDl47(
z2tz?XH^xmP<jVAVp|GGp8*TyQ^iWSNpoYqT(lZ{l9o2?k079=={^%qDx7?2a;rN3C
zyEig2Tod|Naav#lQ|j}3e-Zx^;H0}ZHo?CJ;Dflo*&k#zzzm<8pO=%BweUfl+l7Au
zu&01gWKTm&ySu(l4o(kwJZdbiaLh&&7=G_Vjih<yY)e5!rJ=F$OO+Ywwt`+0TjYjB
z_oq+el~D9b3y6!^Dn@iVyDLBdPW5y>*|-3OPG8t5DD*;m7q+KhB#_A+uJwIwP2n4O
zV>SrC5qz|(;I#$ES$DTJEpULSC=7!NYX{hl!3Zi}eR=01z5mFZ9Se+j4-k!$#@*Dx
zqCVRVDiyDGvc@MS*3kg13xFyP{7f)n0lEpwt332704l*o+x%1-YNzmr{Jb%T{6bZF
z@Ja&2SU=q#xM>H3AhPQcK}SMjk%DhyUkW(+MPIP5wFU*QOvg5VfQ{iu?$Q5o{c6To
zQfLP3^aSqaQ6q}KGyDj66r45cM46P$W`UayNErXIKR8Z6k2g2b!2Sulsgi=GCjdxg
ziu&ZH9@<GY<Xxpe-GSZV=gv-MYO3OmO2ei*LJ?<`Oa#q6&?(>$P^t<_aff;ShM91D
zMy1zW<)^I8#{IhW0clvPP~E$wU6S}SIW2?BJ>(lqz=LN{?{!`srWPZef-;~mvTwis
zx5RYlescehfje6bk&)R~pGz?NV}mvszRJRV09iv()`liJ_r=d2Mf&v{fYCsC5Y~%c
z`IiEd9Bc$;)f$Y1WO|`?8bh${^~L|SKa}*D{HL~S<x|lu(f~=ysW}2z8R~pOR>M^N
z+-$2ry4Q9R1Y?u=HmDQ<x&b_o1Bn&)#JCkUtTNdaV8WS`x+`hD%@p?F;4INhE?>P^
zH)kWPtE!p>bK#`Q;>3q~WZSdg&+sUq@YXKy=ArCd&;hP?#%lMw&Bc1KFflp73l6|^
z-KI>}r}M8}?;fHt>45XHrLg<A>ry}*Vrzw|z2WS^+X|BbeBfMMU2~U;-?f3E8v$B>
zpui1GwV%qb2V=I@*2u`n6rgY1Kr%?oELgE%0CGhn+ZFB_j>vX@{5Wdq)OMKc=pc!3
zWg}HYThpSI2zcOC4AKBtFH;0)N0ZH7yE{P40F~v|)?l42LT2Z{_U273VA)`=!Dy9N
z39IfcA&bK(G1?jDc1%Sqj3|Mcpj3ol($1U?2iCYwzp+P1Q0MAUdLaT@lDwJ!G}^zw
zFr`L9S$P0FOG_I;=kWI8eml?TQR2WP$W!1=rS)9xwCuW;VbclaIFyx@P0}Y001Z0}
zMF%=J;ed`sFyxJ>&z-x+=1{#(YpLD2sehU#b;d&_=9?(UBA^il9wnjo*DkJ0^*VCb
z^Nb2!LA^A^Xl<I(`op{u;LQcZ!}IL#FVF0TI-t&1eTH1mWBC55-prS^YipMYcR#lW
zW`&Hp)a|r3-<Q)NTT|55ia&2)g^1jZ!hDIC8GCdVo3~!9o<g9|oUfSj9(p0g=?vpY
z^bYzHA0-KF+F1ssorchuoj_^V!{G%u<YfN>=+VmSQY*9o<geD*Q=zU9w##iV7BNvR
zjG<3K{sXn8A}MHS6?&haHHIggC%>@pHM?f0tC@!a>>$?1Fbg*3qA*am2TFp$Ex445
zRa=d1^?i1<B2QiCI+yypM+S1Bk?uwwVikun@fnzzAeS%AA;^u?Egn#^rL&o=Aq@<?
zx)l;+nnda1;^IvX@iWDYc@beNi-zC?5S;)E-3CJ%)_9$#)e`EY!zfT~@+)EPjL*OH
zD4Ar$DmyfT%0H&D1&E1=KEsZ-HJmjSrsiYFDay_sQQy5PO$J@gYzCkVxkl0@Q0Uhm
z#`CeK2jm<nk;?$TPi$q4zH@ov5!6-}4>g!Sosz8pAJ%a%?HiJefdAY`bAxS&PKj1T
zeu#`ZXDckKw9by!$KX(}KN9X?cmY^4aeov0nj6qk+^Nf8y6WAOhDd1hnjy0j6A4Mu
zm>fKp2}30WOf3|Z`s!bN1Jr$HKBIo7w02C$?WcA9p`9<h8dGNQwBwxVjf9#ZPs_dp
zgxN=w%@mlJm}d#im!ij=9UN02YAEti)g6*p+1mI`>O+j!d!UjqnkDSHbpl-dS3y;b
znm&g9Pi!dXvw@p&HuB0oXwT%V9Z~JcP&tIyj5Dn+#rKLR1f`ZLf3Do%?n5)vHU>o4
zUEzC7ObWriZ0`QC^|Asy1Tc4OUT=jiz(nIgW8*fR8lH23btZ~|YIio1LFxi^inEX)
z_I}u4_IxqGU@wRI{BW}{T}1O<QpUc1m!!y@iQX6xc&)0EL01Rvn_0kPb9@+n_9;GE
z&8!{LgtK%q-Gmo~g{;Nbg}T5m$knY9ybaeI-=F@ZT%tfblU;J*HJ69+o>!V3o!<iQ
z^K6HW>GD!ym3cXdxB?y!pC?`B7j~f`Aw7E>7TK8Cs=0po9H&hCG1zGOK>UfRwWlNF
zw|*61X6itq??$~G`sx+ld^I4>`i$&c#;v_*IzP3^ba@f`WApKghkx>S)T9KHENJg@
z+_H-iR_~Haw|&Sw<V#@tS@+R?AgF7pLW0C04T2RQyi6t(L!iyZ>e;*ZQTO3<;0+}|
z7&j9L`398hf=VbdR9>3+0JaX<L4m+A=~CORa~>HP88vSpTK&3%g7l1cAJjyaLSgjE
zeeCT;;7Z`zYwq}ho{=#R>N<%g@7}$${uyDfg#0Iu>DKMr#YLMxe&qSxNh73N+k>`P
zqd9DOI$;X&A4AW}!33$Ya};{K&2NE)S973}l2T1oRrYy@?;6xeYEv2N1PNv1l>UHD
zZhS+(gK7}$W!(^s_Utga5Io_c8)BNeb~vR^p#+YIzh*0M140|h)N-Cbx`83+_2SXU
zI2a)5*(H92&0)<B3?896TaTk$SR(!t(uq%Ub43e|YpU+AYluP-+=e+^LU<*=B{L;$
zJYX8tZr<tHGxVrDSws)MueS?=Hk*0xvxD`-V^G}%i`UL29gcK?H?XX+RAlE{B|5iz
zKvh|>02*I)L&LNiJ;K3gD?VT)R^sxV-!YdzwfNI})U7~iVP|KD*DQEj63TeB>8C-5
z^`_El?9xn}*>0JzFd(n$)dk+kkl+(Odmypf!o3iw6y{e16a#swy!`ebJfSHlC9k8K
zlcwEG4oFkPw5W0CLo^`x1hqn6Un;$u!S{u<^1(FHoyDEvj8-^Z?Kw^nGoozxgz$ln
zukW~+VVyPp(g=iJyXYvhJfpQnXu_j35>xSd_|@d|2Cci=ugq!!Ub+)A*~31XTZt%H
zIlxLU^bjwbP~K@VgXP8MtJ@d^x_#yr?Ae2ti8HH;WZ{sw13xG&Nfjt;l5?Dtj?l-(
zRje~xO=Y%sen^1c%#^%>24-P_&=u2fXxV2oER&&ZmI4NAkECCcK!%%ko7(*`Vo&|~
zAokW9*Hg&sN?AR$Mh+BuX5O?{!#<oi0{T7MrP&uRUVzTXdg!2NZeP>uZ4a!GX_9P#
z23btVPL~3R1`~BX2)jBvpJ$fsJcwOZFt>oApAzLg8-qke#|<SKc1}bVDic1%fki3!
z5=X-*;%96Paw~8tM&I;M!c?f1I<*0|p(VK?CdURe{I<X@CZnXHDyx6!>Uzf^&p^Ih
zEv7081oTs+_9S#eFgr@y+uI#y#QSA{)v3S(QC6aV@v!n`HU{+uBr0G7R%p`a>bsnh
zLME&|q{!}pfs0${Q1mfl3C-O%NQobn@6MgIVKq7fektr87s!l=Dl&gO_lBGSSdr@+
zuqDhXe7fayOX^&s;-QWVO+h9}gcPo<*K44vc{-lS>&2s~eJWq*m$uqmUY3!NFYKXV
zYoTmj!e<{IO?JmPNC84YZ8{$x_%^7-4^ce4DTSX6mcIIF@<BkA>=}UE_L-tA><LBh
zGAtU<BgwqHj)91>9%Ig~g>Vgb<blNPwg((>;sQcYk&x6he(&v300{z^NPjLapzXN1
zl~qskAJr5d<hwlzDWX2EkTV~SU~A*CfSrKN#9P=$Lteaq^wr~Ys}ESC7VFoSyycB=
ztg{~b4*U&oDoB1ZfzptU-J_`qE3W1`5z|nZs`#l)DdWv1spMc6n0YCDkflTDhIRHC
zbaPO&8YM7u49{JIrU9`5uv}oEq48zOPU{4G`u~$(ro#}kT&Q;7p^=adfFSlBYVY*4
zB5w-{zU`#)BUEN=-^Ek-y|I7eA&-wUz<GiiMs{e^IGzcd8W=NEeD5%$qwEALCL|`l
zNH-m)yDh7K(+nwm8L1QLRWifv5FWbtbgtA~264hpBr_yDG$gZUJpUJMZylA@_Js?p
zDAFj6lv2_lpoDaHcOwl-qckEV-5}i}p-4yzN{6&k64E8zedl}5@0{bke}A9DF;2j{
z-@VsfYv%Jj^UAZ0c4<c2(9epypPtr?Jny>f(sFo@<%G+!i}jT<G6DnfYw)`$emwc&
zXsz2u+n;0kVy_>R#9(TAVdoPME?>FaMB*4kR#x$UE0sPFqgWhgJ8Q?|f3@k#UHGt)
z&_qt|1{R2-8F$OO;454Yjb^Piu?PtvnOX6fnwc(BeKxef?Zc(PFm^cqJdAUtP|0Fb
zR-3`5XP>8M6aChWp?YT&B>$^IM0YXnPUmigY?h$ry`RS@ow@2=RPv_Ff~1o^gYQJ_
zsl#iFc<e`!GUnpVLwHuX<(UWmC#TgP+K;)`U2%{J1~=#wLvH-@)NfGLUfX(!^`3X@
zI?3_{;0b6(r-nT#EMgGU|E)au<RaLp^ljNLgn(R^fm>`Bc|_Ev48><HTv)rshV6y|
zw2QK^B1kXUgouO$K6+DYOwcTarA2HD@O)xjme@)2ret*Jb0CzFn`_yeqAYq!>=}VA
z+j&yPe!caTt_Xb|Yfr7M%*Y2Cj8bB>O|9TCy72a&f3Gr9$F;6`x<m{0>l_pQ822)w
z&(vP42^Wx}+a-CP<$GSXlUys})$-8%ri>IJpu94CI3l+=p2XH-)?x7dV!3<j{X%g~
zq{6m`e5D;SV$r_o5vjNa$og0GN}g?m2o-3jQ0>q)-l^1+*WCzjKD(KJgqj%8=-INS
zjDvi#5UggfeeA|#kz4Pr`Qh<BQ^L+Zf0>22eT<aNlmnADFGO7wBW^4lh_LX*GmLeN
zH0)Gb>3KX+8T7$!y~8Tetut;<7V&wI%dY`61*~9iq4-&$h#JLIEPfbW<239&i68KE
zzfhlU4sWDRMUNGucHy}VG6IMGdKJZ=6P-M0iK$~YeV7td2);>^%k5PbH&B&DkVd%G
zUk@^hW=UlcJWN)8`U5@nNyiw*L6@gvpdbmY1FHg0QtZ!sEkh0W&B}?OSE58${J%m)
zL#3PyTRb7L;6D25#+S>@WTBbf8An+gxj?-X8SQkRSI#sl*=2*yTv_iYyY=Z>Pk(C4
zwR`%H-AZTpq^cptCD<Y;6NN$4sX0NARKu^p$!{@An2D2qW_Cid**Rih#<i4aR6jx<
z|906u)uL}%$!It+>0U=Td$C=cBG)#7Fybzj{qJ#WC$F)nc|3Tm^kZ^jB+)-dvBm$}
z1i2uabT-bD?q&ujOV^kAvpCiy6)4gPT89@{?MJWGkR%;z$EaA)iZ64m_A~2l)0BQd
z$xS{M=%|+6`PBHVk0ifXKv9!T*>g_GJVK)CvQ%qod^FAX!`H_eEbL6%cp8(pS5oPT
zec=U&q7npVQd`a93s3HC4bZ#Eq<UI#nWD7&wX(FUkN3}sb!KSO9r8)XB-1mdPCNXX
zV@IQjx$hA2sOxs}(4^d=zH7o;%a0kXuI;V!gd++x&L(vd?ks8R3r=gUVLmu%MfHnn
zuUk@$oB4BWQabSPbvGKB$lhJXk9T_6hX@j@G?AzxtN;gI;IeaT@!B=0A?tLjqrG^7
z5nD5)jKxPlLZ;NiER=`sd`7El>8{jpZ!=>aw?-svUTxVV&n3tRSE`Es$({i7bb>+t
z@vFm>{nrVPeG)2XKZxB?_Rm>c_0edM*hzX<!S}7%o?D6BK;|-x0sXnh+vG?7>eSv&
z+FA~}daklWtJG>!@7aUABRr)#-#<ef+;1{O+{)XDksS&5=}cPXtC{d5in!vBq@=WL
zn9R_Q-}$mWxPHIr=}fb#C%fiywCAo@jc{YH!G?Bn>2wlJy5nOJzw`Pm1(%vc&-M2a
zM9Mke^cH@Nxu0^!NQjf)w>@drZAi8DsFrLS;PUIu4|5C&t9*X8ZM6QVUCp3$CEdD>
z`DEJpbXo-L+AU;)&Ju}2-pPSR;U(AUqb9Ec)<D_frsMtdiBA@bS_8L|DIW^v7V1jP
zhET=~>2IvjT~gwogdO*C2g#~=zQtH}^#4kU+-aY0QomhsLE%|^=Ol|9EB3x|rdNae
z0BfG>sX0o_Y#rlHQa{yctLU$?l3myI4@XVX50fN>dGmx|vt|(m+@lmQt4%5FP!wlt
z&)&Iw$JyF$;rggy>d}{a<WTWk1|9k}JCAg_cg*wjF+q7!#mH6}rBhqW&gvrp_<`wy
zUpaKLiwlb%YW_aAAg7U1kNdQDDd6fB?v{2ZIUCdF8eLmwfvfTYbNq{?LNEV~)56D;
zvewIf(PA|>w{iGiYmWpBurzTc28dVri~3ToU?Mz-aeouD%2UsmCIM+?1~ti={U1@W
zQhr1YMortMgmNh9v}h#!c2AA$V_KE6-IX#`<rgD1Yi>y-yguAn$#zopu|pz8BW%u@
zUHrz{`Qzb*c*%swPt^4$3!Qu`lMYM~GTya8Pp`{h>ar*nc`0L<t`qvW6D~0-Vm~=;
z)};OM{Y=t=dzsXSMXu{SKp0*WjR6rdU7h|u_t)u2#!Ak(y>xjCuH9s%SxLP%GrFlt
z&G5{yPP%ldUh)%5!W<9Q;xjc>2|Se%ZiaX+wS$_;70svGKRg$T<LyX?KWKMN&#JC@
zi4A|Scvl_g;jJ{T2{z(=eV<^>WoZ{NT{lZYz7Yj=S(db1db=UcqX}GWK7tpgq1~Tb
zU>hG3BDF}dIu(e5BN7cJ47}+uYdih@Nye9Bv5y_JJ9+G~sOj>_JFm&Rj!UL0<}kIs
z^x05jD<o*O?+>+oKmRQqQ<$ATxZZRcFX~%}mM|Ljq`FP^w&P)#<FK}A*)cBicB_4i
za<6!D*E5yg*XqWa7pF;4<)!pB5rpd>$=o-ZS<zO<DzR^dXw!06jPX*GUI;|E^ktUG
zMNG!r!?+cQ!-~;{+gTzeqaJ^J<8!Z`v~wUhmjU{8dG}*ZafT_WUYs-9o9LXp%`cVO
z4zqR^!|2c4BNgNM1uY2NhI}aVjr^XJJ@x87+!+>T6<b@5rF4JR8z}v>cir!X>f=W%
zRRNfqrZ%RoVzTmak(uM<HAvDC`$>DxnGTe~AK1CQ88J+}2H}kmQ3#HE+Sq7OmV$_R
zr(>YA@PJ~H;zNNiH)*6>$vbaItB-8^)hXAL*=&Y9B{kEVAv~jK=^kQNO0al`D&g2U
zUvw=cow1wpSwn#D=!KFg6=%*6`Z>cHuG+bIq6hIW{IpeuX<UX5!mno-xm-^fV$$#L
z9AHgF#%fjJvIohJ^(Ef;`!0xNnr-OQX3+CvRQrT>cE{Y(X~F^?*bDeXPdwIODEh|5
zpA%YRJZq3zxJ!<&AG%05_GFAAbvE>O*UI4XHTq`vsGgS#uTYT4wqrTOl?6p<Y#$xD
zw}>}8%h0@dLOHqp=4v4N$_%#MkL568+hh8$eio-mb<xhYT9g9MfxJ9w-vi;<SGrBN
z?$12)XlT9nLlEb}m~j39*>gp)=3%WnqT5|GBIDai8|o@8Ugh^N-qK(&QJM6zojs9W
z?h{I6lN;M@JTUkW<({zkIOft*Rp0>wKWZzl^-K9ky|^Kb`;KVCeEOWtvU42xPhSn~
z3yfXp3Hu9YJ$++zu7c)5+o#;?=zq_G%}7>5<gIo8Eom3y7SWZ~<u`*GH}sdzmvN&X
zrl_<50m^u^g7cB+c2$X#LbrR+;!(+u+inlPI2p%CkMzU@Gast-Z0C`G5E+SAQQLh_
z9L(k)7|8qh@y(kzi$p;bhQ(6Ir)PLIHrC1y*d83Wk!PQWsfiL>ylXKX9FiRBEayyM
zS$dwO{&MMY%+OjVZRAOJ{-{b)EBQ5>y9)H<JVY-}i`odO@x56i?BZ2PKEF_zw(q@$
zhE)1P(?vYsdq(WO2YuKI4N@%ipcU_to+ww;06O<f*_RWGJGqhV&8{o*c~eSRi%qHp
zY=eV!r8we5!|qMJXR#b-DvEAD1WPP~H}Fqi@XPR$(y1I}I(hE%vpQ*=@Lj~%vx|LW
z`}B%kKid6Xj6(a7v}S<sl3KHR?c;f$MbB8=K;Vu63px)5r2*30?fALf<jD%U4nwEy
zAu)$}9HVA$S67#(zi;)v-jgEJ!8()dE}<p<VR569t@lm8ht;E`ydTMBb5vB!U`&-$
za#c$fB#TdkcvdQ&d`|B&Al*fGQ!eDl+9CU38rzKX^9CVuS+qgdZCpb`_A~q%i7qlh
znRTXOVN6Y`1#^1d<7a^>*JHZIn+?A)E|A)OL`jG$I6AAh?&_NUY{+Z$b!d=8dGMDL
zpUn~7vsUHjZd;{qNTT271|FRMP&E>*9BZ_eapJzs^fu)cPrs9e4cQ<~wRQ3FEN+ob
zXJT;y3i2e0ADe$rP-r$cz#Zj{`*R@`0&3H@9mPae!;PW*zPil@8(YZuSuC2ZwJoZX
z(mA%Wy9f$57H!3(+ekax>H1<-@+lj{<8aDWSbSLCt|Whsa}rcxS9gt#YOyFZ_wFdg
z%Kz0Q#HSF)6yF>hd4rf;H`24z(K1Vb{^yk+Lb85Lh4F2SbTUgO1_gSxV~(tDsB8zb
zAD<t*%FV;WznP)Bu)lYGsT$q)7~ge#8+$WR7O5O5hFkkM*jQrhhp+A81mVzO^yjrl
zqBrjs4c{KsI%2H2qev0>+y5RXX9egp!PpZhM#;;s_h*ShcL_R0JxW9<i73H6eQYC>
zajPTT$`4e=aD-zv$!Un)sC^RN8ifa*QV?1-w{U9yutsUsdcnJ!{DnZ6F*j(h$4q^*
zixgcF8TNa|%|e<s${`Gbq_Ni{0)bjQYJ!cE1lrHSY`zZT%Xg?@8h@yY+-0dq`a&kE
z&P>0@2$A4fUfv6nJXY3swW=Cp%q8u{@3L4-BYoUMC&|>*34R-{QJu{fHGPnr{mk7-
z6k<FP`bs@Oe`4zQW4uypUUZTVAJf{80*uGX%#e^VIvf)LV_v53NQL2Zwk-*%xn`pg
z6hvVjr=EhTxp`Y_tAoV1g1Zqv)LH5~ldjVfPeeHtwux13WgkJaqh)Vz?-i->5>mth
z`;4@>bK-BL1;gU%(7N4vZVC-Gc1_K(Q-=}JW{Qc8(n8O}yR=K;z(K^^$w?v|=~KR#
zyolIYdc`rMUh`z%n}w;AF1_P>QQ@ldVTULAJ!;=I*Em%y<He13yp1vB`}PzW8Vjv^
z!_!!yedi5v-2E&}T=~WJ4UY+xx$O*<u2ro!oaw5+XdG0GnQHiOFyTsQ`-$)cr&5Dv
z0kQbW0Zpz0Qm<q2ond;;pm0fpxT@6WKU{}Yq+`%@iI(q!z8|3CZ%?iRon@fL$$~dr
z53M-_pAj&woQ8g&AQ6_DnSzUp8*4l<dc*YN;(Hn{j_@<JITiMkv`*9RBLj5SY>IH0
z_a-mT+;Mx5gI$1!IpfjicyaqC;ceni&lWOtee#be-t~wc%Po6S#$dRyXQ3wjzQZ-y
z^TRkNCkJzX=<9aUr#Qj$rS(?^16b2?3FX!N%-QYH$?0Fq0`F8#{K$6`@agNosO0WV
zk*t0>n$2zO{$lxdd3|;LL#}Pd#dV95N}u%9(jl($_U&&Y2Jf1lR>_4hi`y5{TM>Uy
z<QSglC3Wz%@n#FnA~1VjGaOWtwNzthYfiq?NB5nCDEiG;TamQSNgFj%3|{FLaiuYv
z_F3ub!ce%lg_uYHf}R6*E4G5x-`_!<+&j;|a%K4aPhg3e*vjKr|4x55oT6pbVx!gy
zag|^}AqsOB&23DT@VDfN7Nd)sOesub40%YsI}(GrU5nnNcGV59ob~<U<Ok15ojk$=
ztz(Q*vEuW+GQWSiYtwER5~W44KY1a3GPvM!Tkc^Hk~l$-AYUut_UxV81tGzj*Y|IA
zK&7CS{FPUfwH{I)a=u9J^yk(K3%C8Pn+&RFv-kJ_<ZYP|Rgk&7JuSN3PO{?)b63iU
znvPGrZF9@FHjy6mXQpMZSU!t8PfRA8brM_-WPV<`rE%{OF_Bwr3p7W4j%IH5++x@)
zCq09jsH8zZ5Obi3=Hcr4<~<u-9PX#mqxvSgv>B_A?=x@C&^Y-!PVeJcDOkRf>P>E3
z+hCjsVA6i<a_%+3k)zEL8^b!R8KvUlM;a9_1iH2X`HKfH1M!h(?=Mi3#Qj>jAmyx~
z$iFjwrt07<jZfYBLQ{?EHeJ*F=y-Yy{oCq7jn{ANwdoQJ9>gU~h07gZBr83OYyjXw
zh<nE3GxFeq`T?3J&eI*wux^Be+IFISb+>$X`Pr;Jb_UPW^{CbUhumYkpSx}$sT;~T
zj1l-#dbn+OF%W!-;(92+0XZQP8%D)3@d*etKy7boYATroEWs&)-OBrF?792**)JyQ
z{LMIubE`;htP0k}zcHpa=nVJtqcQP2^9aq>`}zon*Yhe~;d~p^m}s@1TmATXJ6*;E
zZEE2gmCi$lh2Z)yA<k+xC7#cbj&Wv*b8hJtE7CD7=E-*)N=FHz@-p;YBp>2rwFf^P
zamXC{NpiQWJg@1ow4H<aJr^CBv5hgd;GK!S-rcl#Ck~a~(Mu-5cKuAeVxl6v9bdy&
z6+?=oD@@-%dgU$6JF-5YXr)w8W7=2y(cItD*^Wn-78g)P6;Sy!<N-=EVwi#Dug%TA
zCO<zvm^%OI5ilpufM#8Bjf|8OLXx`}XbaYU#eztmy(D55QqfiAiMMA800@_*M8z|P
zMe;zI2-VIqL`_oLTzk>X*~H*`w}AJITBa+0Y?h9(^}&9<e$Lkb-!u%8>hZqlzH=kb
zvK!A)okc^uPy~Z%R2^_WHbgjnJyU)a*FjJB^YejR_S2%{Q1!xT{;z~nTcbEh35!1j
z^(k|qZ1yxn?I2Rjw%hAL;+%u!VMN#Obbnu!muJDt(_Leo%QVbio0K@;1haLV;4zgs
z;Nr3nUb^F^Y!)%_OYoq##@F6?6!Vk+wzhH4$GHI&H*w8{(JkL4*8#fxA1DeM>P<@?
z*Nj8Li#;Uza6Jud<k3ZmUcVFz&f8LJ>XGM1;@=xNbb4ygMLnn08K+<}!<@CV@s#nn
z;EN5(T>_6>_Sn|^j(k?W_@!*M$Qrt5G>P#Sdj$O1G?Ttns7WDSS1C!7WG|<<zb$!4
zI|n$*+}gpENkgM|6d{r^rOqen#)-OW5z4GBJExxWyWn^E_j2x{#~9?D7S#oBZ62uw
zpwcWBlaS=CZzr{pl>GLaukq{BcGBE=h<!!Z<AKj4GMY3I%Cu|2lz_To{eAbk56YF@
zCyt(uZCX!-Ii|<9l}J~qIUiH^`Rfg7*ji`3Ufok;$GPOJ`oTbdl@E_j@U{z=f=2(c
zbJ*o?KE=()=hMR*W|A`Qri~t<RyQ_kzP{yP&XqQS{PAH-Y){$kJv>3b*ihciEm<Y?
z66I?)hq-Txc;Ie_Zm2)If-j57uGx1+Qb=D$(MEEbn)X!q1Fy$!{gZ3bR3tPnI1Gng
z5~hQ}a60-}8Eow_e}X$Xj_uC!wASS0a1#GsGgFKlil4F6;=FYy{Rfd0suDSzJOlKc
zJ}d82(E%m&gi?**8A5cL2v|9L>ik(1-31XgQ!JL-;)%H0#|0901L!LCJ#PaDJ4IvE
zTL+Uy(A?)Az7T&T_L(k2x#2eT51n}!T6M36U?9FLi^A;vuHbuzEJQO%djEAO1I@ce
zMT4vTC(-quTNG2X_bpt9tD>BrjO4_Wz2o!G;+C4nOCK0aQd;T2jlO8ar6R7+dc-9D
z#EBnY!h_R4%T&@)2Ki>{g0{}*6NRiQ6Ny}g?>9=f6a7u+-c>T^N_np*wJ?yNvq!kh
z@z=y;?Bt00(#GQE?G(BcB$IWy`7@!m{a!1)`tpW0Q{Be^Uk3)U)#6ft=o3R4l8sZF
zYvOiSjM==$$>Qg>;-#AJA3~N&(MjKGFiZcDQe|_<^WOVfDTbX*q-bU;-R8=7iTZRO
zq--g2lTT+~Y@X?DqM6mCkoPDE;xbmznlem1F3_p%xOX}bokfpBqN{97_@dfKLXETO
z3k2bmb?NrnM01p9zgUz^6*|<Q5?a8fIM#9;8Z?8;kK&gk=9r&fIQ1+3MK>eToyy8p
z75`j=^JlMPx4w4}crkvLYG^lHz}SmfY`At`H0pkX$%3uTQ1ODojN$MMc3mdFp}bo1
zlcdDLwTrb=^ja;I+vre?e&cT4$GeRvC>_(%ihkN6KDV)He|(+fZib@=?Y@&*fyj_w
zy;szOMXvy|PHX{!F}$zfL2Ht(?+Dj-l`uWy_Tvk9(tRzCzj*Y^)svA`6v6M(cSDyq
zw#qgSr7XVQ#Wg`(`)*<`dPqp!e4(sgp9hjLTNNU;u=Wy*Z+%g@ba><S@f*oFO6v6v
zhCFSHk3U{uu-)wt3-GNb%}RDEI54R3x>uKlyLZ7-OY>2iU59+}vq>J#tsF{~1%`fJ
zk58xGv3z)@c~D#TzdBD+>FL&RN+b$jYTa?G@+1Tk`(2qw*POXQY`e|Us3F%QQF4mC
zo`uBnuOnsgOIokFHVCG&gXa}ha~oJ0bioa;R8~vKs8!QSLva;453wK6)Snm8Iwq|$
z5d9RoxF!-YkIzizMM>^PX%HNTF*S{k=!?`?lm;mgp{GHK@KKBf%FdBwkH8H6i_YD4
zg$9P=TB;3KI&w6;XGmfyp)~1DmV`6RkAM7P-!IO;<mYU=@g#&Rh5`$3x@@jI#i=B&
zY>d(e_x2keM7yAae<^;Fh%P=5suCF8lUl)XnQPmco_D)_@Z+7rifb5=sgqR_G+Uxm
zZNr_a`HJ{DwbDNCaK8$?c@){V1w9&XRHYk;mBG6>ZK8)dYHZnCA~oW>7_aB8ad_P7
zn+3634Eg*EJ~UO!HXX@cGxLPd{}8!@am4BEm~fvbB{XHA)aTbD;W3su^i`PXR;MiA
z-+hcXBx=J{##$aDQLTJ3vWm#)8s0QM4XS@PUZ9dKH#*1KhtAcM87iZs)8*cscvSEW
zQMdf7X~8jLTtBf{+9@ph{7GiTcM+aVnLCC}o8*7DfUYzdv2f{S89D7QnLIrHEhQ~)
zVly3gH1+fEZ-$d`yj9?6{suQNMd1;{gK*{m?O8jre?Ke0R8Mp0p|@!E@jELWOWJ!Z
zBx51!AgFA^hL0n{x*76RSV)wte<?|?x&H*k@c;TKpG?##|04mTk7twod}+)g*Z=&e
znDp1-l3)d{Z&z87yPLLV3TE*Fq2UXz8v(EZu%Ob-xB2;sUh&Xlvsz!CpXYSjnFnM5
zLh}v8eg4-yLqSs7ZE%#I37rWf9=4}s^NV=2rRRXw45&+FY;rkmAD9pR-IuSA!p-;n
z`)80NrKP2PdVU7PZIF8d^vOcw<wYIDNC2q$>!JR7Lvub|$K)f&MCO?--hwrOGex1T
z8L1~nN0w-Qv;cLnoT{&(TdkYFO(nskU3UHNJ{!D=urzPS#`Y3cV%n=`n+;4D(*_T$
zjgNr3h>)4EFmwzCfmSVDT||TAujbs#8AvWarpzMS)kx3!jD%?tiHM5Q^HEca^)*0C
zp3=DE+jp+d`v}1j5Wmz$f==yxYNjcW^ej5R#JOFZxrvA%+Ksrx#KC{H8<`2xTl3?_
zJJm;9TkqfbW!Y#TOKE&Gz|2f>x@t$o7RSb<Tlw_M%d000(f|U1Nbw~66`(@bK%}4V
z;V(CGjj%h&FA;wOapcNH;aP&Vk(Upvmjm^hKkhu=F40KJV^bov8OUyh8Qf(*Z6Iw1
zASf0JO2_t*WelR<gs_4n85xPE{Nwx&gjfW-)d>Q_55v@mK<yrCC-KqISy<2zHw38F
zsT}1%rZ5srkeVQ+_@^Jr#zjDE`)YZUrL*bdPziUy42-8Pu|nXW_A?})l@AERp70X=
zI*Fej`v42}eFCW4!T)WAa#kU7x(5@tOzue<bmWSXc`_yUFD$Wc2f>MlMF)!#URSy!
z6kPcnnTYPeQ85$7ebE*ipP_$`k`tA_Qe5dQ1SfjFyT?=crpETcQ8H)pe*az|Dj{gW
z7Y7@SPboX)GZRKtx+omudl=B-h43P(ueZC&p6OktI>xg|d*s{Lhy{UF5z?{tKl4;2
zrz>4m5Dv)NzN$?m0)3g+^M9Y<K%Cb-RBo9q)z?bsluaThsBb4K?`#tNdj<n1ls?c6
zEVv5alu;X{8DTR`I1|SC_xSR}Kw1<w)f9+}-v8!H@Q!eouG6A@`@he$n|sml>bc<S
z?x=vQx6J-mqe&6>dZkI}$%=~(Ii16d)kVNPCUM&lOV5{FK|c}o(!XEieN{bp)SsTa
z)0e-zQ<*UA#<Bwc__X#C6?WX4x%A7pt=h7s`8upy$YPFvKd@-<g6V4}45Z%0M&TK~
zH%u7kr1$=K+EBw+sgqoSaX!Y}J7O9Uw+Z`${<!Xpm^hEkhYI|}^y!NJE=7XCV9h_4
zDn@RN3f|zU!!apc>2`V<VT%h2(*G>0e-<^z5!`6I-!6~RlbS>(3|am^zBlF@_Qp;}
zmy5tLUZ5M-{N#xHpYJZ|S%V|~SG5sO+Q6r@LF)~FJdJeQN&&p@z7@~KqnuPTWA0m5
z{;*`GuVPG%E#szZ(rz-;38kVw2(7^TcR%ZjNh>=LYFgh<jA1o%z0768_;8yxH6bLM
z=+2$w=;*6|we}I4gumibEQeRd=vi3tZ+p3^LtD8A`FViVvID#@jGw)P*?&byV5KgT
zkwGFNA~bYpd>cey0PkZ1Rt@k|2fxRV{<+%A-$b+2hoamBaT&fi^4MKXx85jy^CId6
zH2uZdF)A)C!ji!hIsg<k+PcEI8U|=J1&pHrI2RNz{@TunYemrrG6YF{s7awP2Kc~0
z!$5&VbOKD0V3ak^v>eVB4}}?eHUT%UUxx~)LRy&T(M(Y8^?#3JN6f3RVkR5rL!XZM
za2+C0pH(V1m!C91!<)*L#@!zQfeOU`>)XBr@IRnFZ(83Cq5S7newE`cA=N(KPU7QY
z%-up8FWs^2l^PuU?Qaa@V*pJ#j4Am1w-(?nIbUQvFbPbb>r3G(0YqDm!B!iI9njTj
z7R3G@If%n%2D%U+xa~k8m;CxQ%Ea|pWUGP)zi^Y7B8w+7l19kRbLPp`HhaFk(6QDU
z>QEtCo`CV#&VZwyTnE*iwI4<K7bgIYWYw*d>)Rfn_WhrgM0S#h!duw=WKN??R@ZYo
z=Yuf<#Se7(<b{KQfdLqAiH(J2ZU}goXLfej|GgW_JiYQAVb3WEjVX5)bWyI;Z%-6q
z^<C{j#sTWgdjR1T04=n_%kS`}VN8ydZ9QNT0oqhxgCB@N3{O%Hl479X<w$z-zpEPI
zK1#>v{3#S1I@8sQogZpQ6hZd*ARZa!)S1n|xXk<#n2VlFatT5{u$eM=F_S>(9;yP8
zVMN=@161dYh<4522cM_>LpK#8*F<*;+vv^CX2!tOZqC?0Hnr2E$=LFX&5{LX=?ZV^
z&HZrn|E!7;K`L4_2_bT7&rifrD`PwRm2maH)A9xhd(O<*(uLZtX)82nBC7O^;Lp#R
zqJ*$(=xN(T^vSXY65E!sJ#9e$<HP9kqwFfuZbaHV<KhH?{Mdv4oo-}~nK_O!Hg5hN
zKejz5HLNxaG0cDM!dJaFvkaPqXS;W?pW#fBfmO#8{qv_sL%QTUY?~6~Dy>{tV8K}#
zFq>XVVIdY5G8Wtjf#?9^cdGDHol4U69nPx{|21aFPSErx*B-mDw3XD5K@-=u`S4wV
zh5p8$zi}@~RcXVpGqBsNq}^Zc2CBp-{?y~#ynhTh(%sc{)khC<=Vgdr#mwIcHV<<B
z&!T-T2BR?%yzMCV!T&3x4nbh=!GDIy&+WlX8DH687po=}<JkCcMyfABjQ)x*z}1D~
zSAqEswOwE^u279AByv!G`NvN~kRr`)60M{b`ENG{?X;=t{jrq0vx`nO@mby5df~;U
zxcujyxPPn$>n!_xbd357r<yU>ugo>g%@k9Vf4oB5?`jHDM@-C1%=Zmo)$%F+_}TM4
z?t78N&K?Q79bI3S;$+nWh5mD%=ItYi^X`87k&fir#}SSNT_nmle|*;VEMYA=z6Yzb
zX^Xm(&fk|Z2FA&5&$aXO^MASkdBhx$>VsPEHpo3{Z@7XE63jRNm668}9^C#nK&$Z$
zFD7+(uv9V7B(bp`^^I6uhZllMQ1;=n>^JI)qfefxtgHmRkz76RjfonN%@~?=0CNM%
z2v9El0dgfUHWKi4nE#$S(rqP$qlpzQEGouMTuTl|WGZP{?S(J(Zo8ZT_t@D72M5`^
zLvBa_>=l|vApA2na<KjrtS$sR|MSm0O~aWL->iDNz7R4h607SJ(qPm9od9AypvZud
z8Zen$s`7_m#=(csQ1sX!7$Ry~LyU-J5I+2hb3u__IXj<v2K&V8n1C6wn>hr5U^OZs
zxC8Em)8aR)pOx!J+jA~2YG7C`DC&3sv>ZmeLfp10cmab55C!qe8vtI%Ktt>52!{Iq
zn>TMTjJg1dp8xhO0`CT<tQwUTb5Mn9{xfb8HPVtO`~FjtORXvMmnP;{hg8_xns5Vw
z0ivs=#piR8Dxk_)UQw~Pwbfr}e6-b!ZIuWpZh%QP0-WD<>z8l(9PpNEp;%eO34)cv
z|Kt*H?oS)Um>Gpe3K)-i8%`~f5W<sy@dJ$?jq^_k?hY1d{{+ma2O;$AcXvd2jhp;|
z#C2O9=ZZXzWy}@$TVuSg4^b593$M<(&8`4y0y;^<JI`WU0Qd{xbaGTwrqV%3ZeKBy
zsXxxiuZ0D5kj}5X`T)e0VZ}uLiOTPCrJ^L5WnOA~A<-<>cJDrZe|S5bnwcLWP0-K#
z0pT*p+VlqjBc*f;R@r5J+y&AIAW>1kL{fC>f+OH$Ec?-goSz<U4(4@3&tbU)xDcpV
zSb+%hig<HEwytqoj@QMWhp8%{SdkbpH34_zg_s|FqKIk~UOfoZ*~3UQm))SMs>8I$
zFGejK13^})9~qVnni3<IjQ<)BrRLC(onbvIMwW*jX3J}Iio33l<k*Y;v({IVA+qrL
zr65<yk29@iYcGiP$6+>XsJ~mqP3oDtj-v*(5T}RKrpcNBTKu0UxFS#B(b4qrJI5Lx
zJBw&KX4yD-@5G?WKWXVz%cg|PJJ0W};SE04*uWo~X?4f_(>l%OsJ$ANEZf+3WfvO`
z7GC-<j6*IDRK<z|+)tia9>z~kGvhal41d}Jg}R&_Zj_Q`CS(~z{%Nn}BI*B!@k&+h
zEyMr4-S;HZf3aBpkFVq@->IC~*c$3#!s(1u=#0l^x&5aRS^StDp22HA5S(><Pn+l(
z&L1|<={}`;`qQjH=KLSQ_e7QAJ}oXA1{yuLqW>ofzv67=`YS&9dYk^j>u$g_a>&U7
z&f@P3JhPinL0j&Ymy(6T{XcAXs_{Ab(xfX*n%*qEZc9n}e>aL0#Y0G2CE)428t)z1
z>^z+J>W`;OI+7D7X8bVA!lEupJlZt;!i46(i(X}`B}?4@b*36mzKj+9heiKFxxm!F
zx{~rivNIfSJz{4uDFIO~`YSQHE6Wnt9oWB6-F3@b{lGQVOaG4#zD#x&Ha2dKubd@k
zv@luTxLc*f9Vqk1z4n{V;@4=*w3CRaCBOQX6%&4tJo3knt8uBSWYngc<!c#Inlqj1
zZ@<m6wY3GIy*02iFi@e07&Jd?VcswFnSdr~0wJsRhHTv);JzWXhm+mY(ZN=YKoVR)
zgW|OmZ8+rUk&)Pt*_MOcqB0)DZG^dbw8EQ1j<VO!8SQ!|U&n;8G3$fyIWQhzeE2cQ
zNm{KgfAK|v{*}`pB$(K|ij!m@nK=kCDgf5|X+ReqctRw4L}0#w1QV)dIw<#s&3LId
zts0wF9vCjloe040L)7#<`aoXPyyuQTDMH#(6BLp;rBxBxD`<;nBqStE9_-~qB3n~i
zySQ`o?;-4nHtii9QeigMf^q{qH9|r6tKUt?k5Az=>->btIVt`#0HKC>nS#qU6{HFO
zwcs+I2%8^Q1Ll(nxIc0)1*&XDBIXRD)8WR|inU5XWeT{6RK%I#F!W}h)CW8ha8#|{
zUR{5aii=Qvg>p%(FTJQh4UnoPZpOBO(_v@FIsED@tYf`sRbYI2*9x^6stCYHY!!%Q
z0CEaKP&D~xU3GQ+!MF{C|4$6zzAiwrc$G(hEESDD{Gnwux&bg9RiX*L#NAUa)~#ZJ
z*8!BfB=E0u0&iHPWPA9S^(!ouXHc7A(RYFO@viZaW&bb0z$gC%!R2T+gfs90iRdmM
z4rMMKfiOLg)3Coi2EwK9mpBl(OK)rxgsVCoZs<>3SG{Rfz>}W`qJjbUF&;d40O9Wb
z!NJ=;fbi${0YP~07+Uoa5fM*QfP(_nod^Ljaq(znZf@?(a8v+n(-jo!)szJyL$qx+
zGmi{j&StXKB}XD0s0gjsKztrXvWXlV1IQk_;fRz8{z1rU5cU}H;Wzh5+Y3m@%BUC^
zLYCYn7!e5Y|L0GfT77-JL<H$rmEANr&nGh^CbZPlriI9`*O~NUa-oitlKu-=nxNyB
zHzmNvwv}Sna6BKYn82c6=+ncl^bR&SygDbKgF!jx2S#LQXbOh~7JnvaLBWX1qGrTy
zfkX(Klp8T;X!Xnd{5(RG@=-dwnT)F|_lEWDxs45lNIMX%gJJNiLNL(0M{@@%Voa$^
z#~_G*4yORLQGbIJEoT)4zdL6@iEb5^O*g><SU%^UK<<&T8wQ#qXS*OT50BS&D<R+o
zMyu>Hc-d%Z%s|?6uMNZ!>6w`1$h=|sfNujKR3C=LaGq%)kO6biK*<{BvQHZNoCmm?
zeY4htOoxnhYj@XV29e@S)_w{N?FNBi2r39}-_BAx$bB|dfBfnOI)pHwY`hB@>>M%y
zA#lIp6kLMdt-#SQq@FDxT}Brf!i1bAe^d&uz33BqpfAR&&FrlVf!hQk8U{KAhdBnB
zFH<zR8PJ<La2!lRaoi<<M{N6}@CT$p-+lMOpx>Y6P1A{?eFW0Xz{gbJ>4y6Njb=%0
z>>Kc+`cLbCD-9@?`tyUId-o@PWY5D+ReQ<nzDKtr3=xSGw$>3i@gk_w7ybe^GPW$g
z!_{YU*1~qA+Wck_#hmhaY~!O{T_F*}pn9jPt6N?hgwURzZWKI~s|ww?f<tLSfb4i>
z(!MMJY+7mGd>_-j>|1CSzPSWoH?(pyU9)yxzh&cd-$P85Dt;3i9W7kY56*pg7cogE
zwmkYvxoJ4{M<_~y{M=p=4vWk%D7%H}>g*|57tmn9Jelv~w4f^5@(cRczt;eSBiq)|
zV(>mdO~~VXe?+-cNsrHUD-%-1rz=YknG|%ZLb(sI1@XF!gZJFaLT+^7f(~@YqS_hk
zGceFLmD^AefQZIi<M0{CKSM>pA_Hcf8hWgM0{)8F{T%#2Fm~Y-g6U7V^VFEwK(##m
zPPYsSLnDGt6Yua&E_bR5)|H%`ygI0j3InVNZ1M6FMtnD|N)cbl2vTF;xD{Z2NoJ+~
z2$v~a4Tk}mWya0O6DlUR0wI(I?MHKn)uBLY><c=zal{2?MsNit!DNszhpnS^m!9%-
zWitL`CLN4Z<3=SpJp)SvV!s()Z^g<W6hvFd#(?9fVJ8T3W->Y}qF`#Pp8agQc7A>i
zV;yZYh9w_;g2?wEiLX#Di5)DEq%*q$IWTyiOV)J4LKdJjmtmmg#z{_!f9>j8{d_p3
zy8Mmu0%{2F@^2tPTU{;T7!2$WLy%)e2puyTq)|-u^_2h`jIZzK__%d~S;P5(F2l;N
z1cEv?S9f=WZ~;s^fvr{ygldT3>9)Y&0IgKXPSocP5X8KWVDE#6P;{xz%cCJ}(0Uya
zVLO=H39n*w1h#YGk~%_8cAk@;{~QbfBp)i9>@KTnID#krr6!UGarj-q$yRgk-qG#W
zxIh@H-fmng>wFB5rq{}C*jA#zlf&;S<Z!o!e4i<GFH-%9>K;OT)ZJsyO6>_rJlpT_
zPoV{5Ilzm~y@h6mGz82Z?O*FZtC*RXmi<0|_JVW$ox}&|@Nt+kh=;NiDIt%og{(y5
zosuwFgXKW>X2WTOKp1d=z>YF0;fXV(71Z7HrSywI2%RIe<W_IFmw%V9h4KO;h!e)A
z6}aZv4yw}?&5o7lyh!+QogpqQiGE#PR+b~9j?3fVspI{Fv5}F)4Rv#KTFUrsQSlTg
z<tVqjjSJergGsIg0#Nup5aP0(kaD|)E+-=c0=;@lO1Cyrwo#fP_2``at{(^~nhWSf
z2bA@yrqPh`yH`N=@Y_N4YQh{hfdX-C&3+}jw{a#@pv!}h@Nj1R+6s5-&~aP>CM`Da
zGgktHfF2k}+WYtC5(sF(4FQV6$mNDLY|<7D8_H!ePhd>G+1J+5DYUWuk<MoRjcETG
zrrLEA=D@{fFqA+*?PgG?E6J+!l`^Zij*QHnmUB7gDot<!8kOwQuY-eOz(^LPa9&vG
z=s75tTQj03a?+%qX{&uDXfXKtibi5U0`LRaPR2qlZEY6_|M~gT)a;>L-*|qIA~L2Y
zwLqp3ygD<ZT;piwl7>SkVZ3#CXdyoeT*J7{cXiDJe!8{J1tL6nwXhOTO&7qlSPvH;
z@2#lPV#79RJ!g`Ed9jfHJ#^bKeIIeYr{l8R?>i-chOos5qCdgKu&&)ecLzCXUS9DF
z3Sx0MK-z{2z<*T<{yAYwwW0w&6L?qCZ{<mL3|bg-%pmcJJ5ri>Ynl&A8uQppe4Lzd
zv(I7wbeuO_UU=-UeFTL8pf@UQ41w%?MFIg81Y9{cVSt%%fXqqeMwpD5iu-=8^NL8N
z5wOaxbt2O<W=Tq{41z|%J91_Pa;*{r{_!{I4dLOK;Qu(hPMqDJL$H>!Bt-#)+#kMs
zGJ~Pukqup(4j|ux3cwVc5ce{<V<I@N=mGr)YGE{EU-X$}VCve|xJ?OYX6S#uZfbA;
zQH1B3SIg_$rkkLr>JB0e-x&~24-7&%*9Q1<6W`ozHK81?ZIG4B6--jUiTn`qU|@|G
z%eP@`e<&#_S)>+Rwfw!bluzP8zoWV$p&ySK_VK$Y_UrJW5r1kWW`AqYqOdZ-h<wPf
zoR_1?&cdR=sCVnSyB2uk+7h+^6QR7)+XMs$N82XvzW~A8T`O6Y?*WK~b{w+ZxcdJ@
zAOAo7=gZjzs7Da8=t;@TKa9sc*xu%W!|3T*e>Kt#HZ;U3E7KBq;<d_9cxJ6%b98VJ
z_P>7RN@fPc8J>AxM@vm0JZpq=_VMFKk@HEQFMo!B5(L4I1)6dsBS(1%4a^HWfhsS5
zKh@+S-2?DAz%V@o<%;ELCU9lmhag7Ikh?bC=p#r5bH=Z3UV|s)mrLW_|5W@=t28@4
zKK=oEGgK_UDa2E^;{_N)XasYlLg4@#7~c{vg?){X1fXbIgU<;yvVkfz@5T*mkU4$)
zg5?I#Y%A{)yJw*x_3f54UexC0nfW1OLT%S{G~pG%;1m@VdB;I8+-aihV}Nj0u~~M;
zXt@Qp!A-c>NdSUytTO{D9NAkCZ?0~{q^FaUlBPm=WbP|Z1n}QwrKBKl0f~i}=ZNu(
zU`d21U}goy`A-W|Q)p!_Z?bR0)jK+*q+@m%FyitBfj!K?z;L)E2q%zt9Ije+W&^b$
zOG`S&oi#$_+A83=!|XhF@6e|q3Xyn>b(kGEpqsum<*_;A0yh<<X?z4?&h(y8$XSho
zAz0(p!24e{4~4J`=3nXBv;&PY7qlAU7S16Bffzck2Mp;x&~OJffD`Zd8ThpQ4YG+W
zJuD*{y9<eigB$Snp)7ONYkBH(-Nqxhr?+xZE#@%94`g$5d!Q)B0Vv?f)XA@3v*3Ea
zvthP#wb$u$Nl8iGSWx6yB}5*MYyuAM`cL3#%BMV9aNRN0gNO(I^~X7G(x^p-m&B5^
zWKMH0P^ulB^o=t!K9-eb3mwz})0d3R=nDd@6|o4s&KEDAslsUtKqnONK6~>aR7}>U
z8?vaON7xhsEhKOx?jANp#l?a81qWKA!(1S*{pH1Z+&Xks_wqBm5fwTJq_iap=<%6S
z-hP2dAJV22N9(~r=#!{Qfia`c`KEx7?-o4U64j7p4avT6iU<y-2M@LI`_<k45eVG}
zB0zZi$2jB?j$jTvpi)k6gJ5h!`jmvP^m@@85-vvlT2B2t3%wBjWq`zwmwCP<PcgYM
zh*u!Sjj!BVsHEl*yf~sNFM#NES@04OR*mgb^71At0H_MW1BlCQ2eQRsOS;1ofbfSx
z{53dkkD*)#H1-Hmo){M2J-H#`Wv6%s_2oS&tL55d;eK4>5%Ei#b54FGwCT!cBvZiT
zz2ffgyAwwyL)U!-?~=}Gq~{1cy-c1y03AOJD`jib^!jJPi-3TV$$yT!D~G<nx0glM
z*mY+f^_gPps-C%)&-p&#y6iFiX{}i|!3nS36d$~ytw@0rsY&-35ba*DL4%s08sP!`
z&pf?-IK8W6Fb<wh43tR32lTUhB0!;MiuT^UEXd_6#-IqhpDJj!U4f||put^(7BGaX
z9MsQGZPAYt2nIm~#VTfT&d8U-bIX?(F%}N5%xZ|%1@se8h!ql=?4hNNhA5-`@7aVw
z*W;~{FIW3y!Mp;A&0?h-5m^0)T;-*uSzo5TPvtd<;FW<*=SN2mg?wOfCl#UhbXq7o
z1Q{eggIcC%2D^j4akbM@g!wN34bb-w4<~rRNc7ASFa(1q!O(75CJcmleP9H^G+%dx
zb~3342W0)hl{}zyVSYEw6Cr=0QT&d8xeemDPyW8eFGZkJR)N)Bn|A%?O%5(2d1{?Q
z|EFxS#??NT0s)8114wpYB(Lx{37P#wXB`FGHg;{H5^}47)wnltaTVYzL8dtjqUs*s
zeh{r1PkTOq69|Nf9<qXO3cd(|Bpet?_DD-dXBa}~@^w)`YNFaS0#J6B@l6i!qQ!u1
zF=+kkV`ZgzjtJ+uaCynKIk1u6SutP-8Gj;8FL6BCvl@PzxrBqzvmOsb+fVkLnwru!
zAA{sOxhX0=U13DrTl`RXEg#xL^16el{pQY$%NzlgWIdgo^s>e<GCp$)>YTsA;XyJb
ziJTwvm%Vy#KJc@<y9g4tVh#hSsMaz-NOLd1vUl{0zi^R}>@gF`_e>g$;zbB<nrj4&
zhdmW(7KAXhidv%XiB|Y$Dd$M&)NTH(imcSDd`f)EYctGr0;!t8`yRx03M1v&`Sd&M
zan2HsR%8!!_^go)K(%}Synmb<B$`x-z{nU3q3hp$R%tWBa>55R_4hf)Cnx6g7>XGJ
zd0Sjlu79ojOFq<QSVFo~5Ox@zhME(+cRSc6<?YvsW_zpHo{m=(JZTL9FVekmhOE}`
z0dDKLn;As157%~`s4$)$<b(g$wc#+o>+Ww1T336}`t$i5g@3{*2lbC^b><w`0bEBA
z=Fm1@ur@YQzX~!ypQA8;Lx^Gj@=*#5eHtDBZ2(Z5fRI^*TS5bTT1OEZ?3Hr8H~Xe2
zn3z3|d$UkWR=<Ua;6QIg1-2M7Gc$d8_x4k50jZiLH8S_TWnmwkDqDSki1wy7fbA5u
z;TYx;jbsDJiGKrx*_;%IXMX{D)%P*&hTet`w?RxifQOvEG&Xwk+)2GOILn!Ydz4hZ
zwh8iCEmJ3RxDPsM*pxmRF=k#@w4NW#1W{PqiV`S8Zvsd*Kvmp_Y)^ef34*jCpGqh`
z6xhBHZi6%ZiG5-MMby>R6;_EJ9c-qu-R{}3lOxf2)0bM6il<}j6l)ex6Dy0R*O__x
zJzv8NJ>`~Y`4H7wQRV`P4$7-9ZLZcV6~ey7eZe2tm<$_Io22h=Nx<Ifr@}y($~Eq#
znHdKd>w+LJBt^Vq`t~T6x&Sgu&<{|?_PML;3YHk`sG+?vX!&rygD{y)-o`b67D7qq
zYjHIeiIU|R8eBK$C5Z4?YgwPK47Rqm26#7(jUlGcn16n_*Vp`oeFMBB3_+G$6{^jK
ztf4$TXe=6Jw_b@Yy?xj^1Qis?o4I`ababx+M#sn5X@>BSVOUfxs&mOByQyMZ@@x|j
z?s9}O#cI#<&8eQnEMCWX6n(<UnR}vvAJ+!)13~>QkwIO?#*rDGOh}ffOwUusrmYnM
z@iKuBHRop|AIKIa<361OBnY@70ghkSp!?Eq>{55vI^mUz1-+W+9U&6Alt;{pO+xv+
zkd^e<+`l>6;8mv)>i2;yH#Zl9LLhZr1&UfQYDSfRfojd|T@aLytnn|C;*}%%%yV-+
zWZk@02LA#MOa`aGHc^8;R0*Dzc`RWMF=ik+<vp}to{p8Ru7T<ZCP~XS)B<d52}np}
zPX=JdM?DtYd)XYsEj|s19y3g)s#vsq0=iF0WhPx~h@nK|VKOHGFRut)WC~T2dwc8s
z*D%Bh6zXw21gA?cwx;Rn<wb7~<u@2rRNJUPDeUtj?TWTMggMW)K+>$N1W^D>Lk&6i
zFDfe1ID++|vw6#2my$waHWBy3nmJE+;RRYQ`oQ=$m|n01FxJ2b@_K1=OA>yIcY&4n
zPLm;@-N^X;`!`hNqzBny<QX)X&|V4MFRpX6R8hgvpM-Hjmfi4TJYnM?T1WboRPbWc
zJHW6mFUH;(m>XWlkWNg}7sZC>y|UMzuSmfk3bVb+u7TnfSZMqO7ufibA8epnym|{K
z24aBv{b5~qjeJCi`UNZuwAbJEvJ4Fk5i;wD)mX!!K|g(g@Dz`8K4tylwuMMxwf2rO
z9%#bBV`zdNMo$RYBUFosNqIzg&k?j=@(z$^_i2i&EWjJVnxUWOgT#j8kax6Thv%{g
zgW9`(2<#9pu1F(Xz81=1^gTULyC{C>@d$C(&e9;KQ&yJDW-J161fme$N@#cn=BhUB
zsvMRw9>h(2pkfFHd%y(~tP^Yus1mr&B0l^7*&ih#7~~AY!?vMdM_T3V<n;5Cji!LN
z$gS%W7O{{#NXOVy$la+AbbWor%1YSY!2#wF^uiPbHAoYo)w(?G`j3u=Vt<dcf`SAe
z3~w47j8chznSIsvm9$?4o|V=1Aoxi}M#d9>`r8>BXEml@MRc+LYHxgFW8UB2f3H<a
znI?F80<wm-V`Vq~?XOD7KI8#c0fHyr4Pgk&85BuZ#=IeZhB9eGrm(v^FC`Bp&tLtn
z4ortmD_Opyu@QMzo=;U(;59y-tbs@xC{PwyoPy7Wp>3Iw$WZO-<Pq-t4m&f(mmUn^
zuD}0vD7@1`)t&2WE>zuD_0SZaInm91N=psEBuqQJu8!CG1TPQ*MhHoD&t0IW3>&4t
zoIjabsrky^e|Kcq^WUto0Q#(hV0?VLDm`UN2$g|DM@#GI__*}lDCkha7yjF({GvpP
zjzdE6OB%mR$7#33BG)w8TIZzk@_ZBOtoIe)2za@}&fOnz{a>RP0UVwPjTLZ-Q}?}a
zdGX7xU5Yj>!I2bEzGH1TdIGZxAWA&pLV#ocg31X+fSyR~hr;o@)?<#B0=L9F>;LHd
zEdx)6bFEg%A+Z<@BrnEB0!A^M{^*EB(O}Q!b!rCe43{y)Rn7tp_G)m<*xUa=5B*|M
zc))!FCN;g8VF0^~dNsxUB)30WkAZge6<Z1YMvaBKIiLfwt3&V!?Z2**AF+tcR#4SZ
z<>xNk(eG#^o59;jQqYitb{C>KUoH`c$CV))%I}ZWZs0D0scV~DWWTOWZf^WXqfLff
z*$S8qvmgt>>%k+hc(hoC;DL5{WTX6tIf=tiaciw%Tggvp`gu_39roL>yZNkmttlWM
zNAo2O*&i*<p#ah^>q!-}l?bg3;}Kcuj~{1{)91J*anEru%1@y4`w#NYNauA-6QhpK
z5>*V2+SIuU4oJRia&I-dqV&$F1ZKtZ|JUyOO1&i;Re2oJ%!2apnW2nGL7Lua&e`v^
zrr=+Cn9qg&)|WwaA?e)w^JP5FTJ2cSF|q4O=f2d^$T{F+2@quD-2>&BD`JiawV5|m
zs7Yu+Mh?R+AiKdBqvVsphnurc^?**A;iUYJr3p#H5O$y!NoSs!#O2eak2x=@O*uWQ
z|G1L=W-_ejuZ4~#MeHmZ5@i+OfCE|4i_8r@B{4>6DtQg&`Z?&<AIQy$oABuV_bgd-
zH5B*H*BQN!t*2XoSM>AXte&K4jEz1(@?V`CV>Zb%+*~s3)&Ulg&t9qKRsNv)qN*Rz
z?EA!-7qS0@sBc2s&5!48UL-CJqoA9k`TMI_m<PLfmb+3I(g||Qj(@xo(c77E{y<-l
zD3KAF^qSBl4Dg{KdO9VJ0N?y&K5iio%>sCOf^As5jy9{;z_vx)xTu8hi!l3nw?CYO
z|MgGZ>PQ2C8ZJh=^+SAJ0JWGlt3~P>qA~n8LZ}1xmE$qsTw7RE{dFo+H8T7kusAw3
z`?(fe7ZgJ@%MsS-N>da~>bvR#IZ&n`b&e>hJ;j%{bdRF&k^Z=Yi-beN;r1HQPyE{!
zHFx9rY3fBpdDk&}I|f!&xQXNFP|V!d$A5|>_tY8b?}b1Bmi>gK7V(yv=B<(hxmz_y
z^FH62KO6!5h9_bSf1g&wYkE0`D*7~VR_r79W?=uMFRRVpTTVZ6SFjqr-A~?#o#u}f
z@!MsRH$D_IxKTFuprNAenj)O<Lxml5J5&piJTa_Dba?E)-h$EdCQxf3b7fL^F=W6{
z8-qD=%Gvl$)*dTS#f^^=amBv`j>>TZ^XUg0t~i7g^{;lorgP&OxnycQiC8|6>%!aC
zih@4%Ux;W(&M8&!PborUORQx@K#Sp*)7g04Y&ad24#OQpBmQm-5(*MsSp_GrzQcSl
zceA|f?vKSl5cvQm83;K<T1rRRpZ3r(e}*^*0+^c^L~MW&guzBGP?tY}>Mo4EdJKRc
z&=Eu6Q~`(q!VT044!mk0665UtY3@n<<V2>^8L!kUxBR?$Bw*0CLqG~X*xbsh;CUc4
zRrXhgmUb4p0bBD7V(trc1mFs8icQ(L;}#yv?HxW;e#LtB>w`^!Yik{s8_cyiHCL?)
zun?V79w%Bagsv}i4upiRpo3uQG4K!nXR>HSiLOgK?AGux@8(B%(xhkx3VGYhkWbpv
zsC0cv39sPva6jBQU&`<q3_=7?kQR~&z<gIof-(Pgbw;&Kcj@sC=W>}3TS039nC&!K
z75F1#=@~wv+$6MlMku-POmTN6`q17x{RBpAEUmoOpfQZWglg`726%yoFwBNua2xP}
zBZcZ~>x9mmQ<IaE{Pg#6aB-=qsA6Bz<KpAvQ&2>n8V!P6;4j)YoF_wrKW92qHgd59
zGn8N0y)~Z({G6FWZJg~|3Jz&(n1|({Rl!wx5Lm8J)o5vJE3CRF`p&RXK>PaXDyWz3
zWL~U+TZ6(`=F)vor{jfTQHq${5gc*&NLmksFp)Y`byhR(uJ7Tt>*L%ZcQt#7<aNGS
z_VK47hP-eeoZ@3dmk{bSbzY~Hyz>xzLe4KGug-jP5W+c?Yn;-50d9!$*+8J{v`rSg
zSbH@J=0PJRs>`86PU?-5p`7bd*9Hf#&A{DM+V3vPzu$sEkEft1n1u~hL?~;%4h!@D
z8@EG*T~GHy@2p$$tV#Nv*{O_^=x479DGTG25-EbKvH<a>n{r!9gUvDG`V~rHTGc5j
zx**I5s>&J=eLyL4Sr7%`17O?Tj`~KQnl95_?+?V1ggZGT%Z52fXW)mx@C9w#U(d}8
zh}USb%1rXh&(EL!?2&lQiSM7!-n*Ujmit63b&kcE|DuZzb+6<-nc#8Z$J1#0OKmh;
zyg`-ND}TpEUtV6Pm99H)4c4p}v8A)<`Rws3<Ata_3PDCb)qm1tM$bQHVHz6Z+-c;k
z$m+8FkU`ZrTj@$qxt5n5{g-IjU#^30%>Ql2L)1B!Ho@VsZcG=PM^m``PNmPm6#Z*D
zXd;i~T1)S`;Y+<=46Hq^E@9~sCD~|ScxL&3{T7n2ik=58^FH1b|JTF;ts7;5@AYCo
zE;P7qUV;F*7M>(Y`}*fYgX~f<TpP}VQIik}T-k3HPZWQbw&H}h#aM<&<Va(PtBKKP
zzedZ>ejSoYnf;EA^4g#W|MM5m8((PaMIe>ln{c?Cc8ctqnsEH}?J-9qAWDKgpfo#>
zkd1JL*CF5|;`~a#Q{msIxHMXql}dbg@xXh|Wsz8&O;<dJSzH%{+fcw|vT<|1a%ND|
zbJSk8Oq`1leXQ&48)VN{{fo<0L^q@ofZn%2-pz<^LiX0>jtK?Nl9pmY@(r%VxcWCL
zN;@`sa;VSf#}RHT3JJwe05{_vYr$T4UH^{Jfsu$bu3ObIzClwBTeL;MwngTD-`dZ`
z`D*?S?S;Mgi<2kdxWqJ#`}hgcHAa1K##Jg1&g(9CLH_9LhxQI>TqMq^CDpnwPG<WA
zcC1m0mmk>+{da3hYSfM~POqz13;U-Z1X0Ja$8T~o)i7}f;GOM>qaY#f?G>poMA9dH
z=VA_<UT)HP3PFw&ws)m7<Bk)rp9V2W<h<~C^xxM8j_`-spOjWlrFeP9l+>}dOGGX7
z)J#bp=cP9dNHQtpj9j8u<XIwOF~ntWpySdsx(@8Iq&zoRHfm2~@41JH$y76t??3H1
z7J#B`!tDTa3!WjYd&gD(nIF;%c}~VVY)wJE^#>uLU2IJrSUc`AH?J9Ns-VJi!-~JW
zi=cDtMVTr%qTNpPK>a_gy=7Qd?G`O8qBPQtba#kIH&W7}(j|yVC@Dy%bV`?m2nr~m
zbeA-ONH?f}baUpy_uYGcKhF2-ynpt!FCW%(Kli$0&N0UrvnFEG`m2MIY#HkDC4-sX
z<ZedoZnKsdsnGv<?7>XpJSoqUjUm=Nrr7HXpRdN}agALn?IZrMA4m5;zs+WoPI-K#
zP4P2^9(O5oW>}I(iFsAR@rEU#LANUYe_v(OAR1ARI4PMagzEF7(*@aA&r$=Qa#e!r
z>wlh>N01%0J7F`|gE9PZk>}fZmsUuXlzR#Xr444*e_wW_z!TAjRUMO3@s|9#6?h%a
z)Z<MGGS_wq<^O$)^|t=rT=jw~Ml;Nj<hJV*eA}1K5@_8MnX;z){$0>~cg>X+av5bF
z9=c|FTZboKAu_)F6P+tc&OY<suQID{#E)b<U1(Fhu|?8dCXv}918;jULLode_V07(
z+6m_CmHg?0W5OjZws<ffm8UteS$FO@ObeUX)PAU3f%!*Xl}tDtx*`WoL3Q=_!NTs3
zah<mB-swKz6+R7nzeE$Na?@Lyybo@5{CNxqvoHt}Sgu~NQ02j{p?2V^S89|Vd%yd-
zT3#^C=_OwV;);lT>!=@nRE)6u;Y7kKDaLD3k(TDU18>H^+}v+TSojzh&F_0o<Hc~e
zB*MpH9;BXNB-`>;*h<dUvMJ$O$A|AaLf8j)+G$I$Smj<lKv9G+q4?4LhYufux94*e
zBcpMkVfp&_oQHGfeT(*m%lOUTOk$j$+S<ig;>dg#{M<(1xvOAaA@zlXRbc$_OTL#D
zL-7OYV+hNBp|gM;T+E@^L391#Z8*|L-Wi?~?=49K{BT@x%xfwhV!Yb@Nhzy`U7Vg>
zeU*vAM}sTP$B*yj?07s6P*VWdD&*7L|M+AI@uhd7CMQRTV$;)RNEkvM$LIGZNFU|l
zx67{YTD0GY;AAeEJZ{H4b=R1gMqNyQ$$tp~jQCI6aJ4`+>NDsIIB-A}QmC0J>*n?i
zRUL|86HpL97-b<fe9iZnen{`W`m%Aa$C$1wxvPtjYa9Q{rS~u8_fh;%$Sn;E`)x&?
z*E~)Rp8<nN58<{U<e4-<=!OmrMd!Dep&{`SB<S;5mlb&3Nn=Hz7>+|O0A*(CHhNl|
z*ZY2cT?CC|4`64CJa;av4LoX3`Z3Q~8HjnG%}N68Q9)Ei?&bVTkR_)tDu6=m#UVIA
zLt*aqm2&?XK!(9RSF6q(;fdJv_AD6y!z~M{#Hi_f(9slP8)hl&&&9z(K}u)=YGBiS
znQ*fKKUiJq(~ic`p_*Gw4dw4Bzr6@M&rKC)_t@x2KQ#fIzIaMk!@Dng@M&l~ukix;
zO-TjP>q=-!2(=IH+wT!hj1xX7sbaVm@Qik-ppoc%aM2x)zEH`mMg3U%>45iz>$~Ne
zEevhNxMjSZ4;XVc+Uen+ZB@5Uv4tgM49S(4pe@Ag($dYP$j?Acf!^8Y5JP|}1MBY^
zUa7f^wKd%Vzn`9oQ%_Tku@r?MW+>0)eYwkxX#vbC;1b*b3u7R{h&ivRp<p6ugB-QS
zC~dvseOKB0I*;r5J&(9aQ=;_>cI@0AKzwajc;@Ht|IVmN`$gB*F+`xPz+eN^E<Zp2
z*ozd-Ky)_+EZ(ag^{pv{9BVzhe%@{Sbx+EI%8j&95cY0)uK#7&r&qSUt<-nnlEaRB
zZnI7*F->0A#EYka#J(5)UQ1LYe<%iOj~<qPbGhg$YkKpbwrOt2rX=Y+bpLzR@U~uJ
z4i!5d?IZf`uE?3Tv2=^{_2$I~oQNaag}*0vC<S$lbGXSObG0LzJ9DeRz47U}V|6Lg
z_(Opy%)jG7E-hX(j4UVayW<WqFXp=+@TO^g^`UaVJL;U?VrQ~)gwSb`10JqaG!>5k
zCjs5#j^tNQ7vJ8QAKhO`l{=2^YD{Xo{E3|j$*7P6=7P_-CWbPB#--3?kI{X>%~lO&
z$2HoCvDfL(P2N}cHV$H)8X)rJe@3~3&!U1CBfo@yz|(}|oY|{YyEsYhCS4Np1@5nf
z3HnE<=|7dxk%aBw-*Mo!G1a+59iBX!eCYGGuO-cZ)^o^n_<I|!C=ZoSi^OwA{rNjH
zC;mw0CVd_=Y&Nn=%88!bKK`2U!d2fhJ~QPGja%Pn2(3p-x}(<UN*i1ojDsPYc;`UN
z-*-TEQ7!gybI8!n*?LhQ%gDB`_8TM42cEH^@5fa~9V1*7f`d~$Pfq^lOGbv}RimF2
zRqn7H&F5%ZMyzf7nq1jjZt!xpl295x!UYM}f9CSCs%(H2Q9H{QOx(StNWXaW17#b3
zug8Mvjj5iHz}l+zpU<R`7lt}P#gv6_gbi<_S_zB3cv4gC5f_ufwdJ#OIq=^JKi0!g
ziC^N`b!tzG*KAdrWI7z7`lX+8`{8Nw|BNqzj!SRsd1y`E0w&PPL;2S6%<NxV+&<56
z|Cy&H)~J~0M`iS^`vY(8q-nPJ95w2__i7*5pEdg5kJ&T|uDHCv(V(X}qEl$r^Gbwy
z^b)D#|E^#uyh;99XBycTUB509O-i*UPg?(>Xx01Qhm6AkTQ_NT^|#B8YhcC015=ON
zPeW`|{4xH!G!2yJnL^z+v!im;)U@(#I=<F18NDKXRgM2YGf64qfDyeHrk}^w*H9_w
zGshTZ$tH~~7vvVZkbT7=8f6LJKE!bM^Vd$|%Yt6<k=_pH%2DJs^R3cpkTJF0oYfW1
z{5JWEnc~anR^0*Rv0K>roBF$tOrwIXxnlbFs$=VMl}cp#Uajj%p1Adr^&E*nK0^K^
zG7ozoscjaEH}l8LFV`^UTQ}Uk%!qy=`k!6RH$WHB8<`^8JZ#{P(zcJa_}v#X?ztna
z=$6-1>x=URem-JLoXwzXE4umnUwDr;i~4%5yE~#=6IznFdv;jQSmh`3WUwJBmGiw2
z#?U@@ZQ!GAKTqVNP4mGX>=#!L#_kf44YM+^zbMF?#@7|!^Y%e9rGG+79_z>B8*F!+
z@W{B=%S=5+e}y)-u^v?5{Tb$RnvixK9`D+5oL@2`R$CKTCof9*&q|cG$KG_3`AowI
z@mczdQC;Dg6HAXRR1b=>>Xw%#dfYbz0=0M=XS*7uD!#)!%;~Brou79Q4>Pc^-1{@z
z^5SUs1Xzy<KF{>#%^Ods0KlaUy)913$jLFxFZAMU)-&~AkAwndm=nR>J2p~PnhZ<L
zzODKm>)8-Pniu4~KOuF*hz`7BVefSg-q@<{P!ak9(*UR`!^6XI4;%XXwUpDLC;%54
zNRXg1HTn73gR~`j9O*zSb<thA7hm-}{Tykwa66Mal0{!hGKV%4i~Z8gCy!*iH=PFH
z5x_F$SWQq4nS@+5wHXNNSR@=k;FNX+1}!H#I=rUH?AJ(~V{Ia<fj)7Bjmx!QL*-S;
za9hf(VorhP@j3dW8FsFSgp2b8pl`&28}iNYIkrDQZ-vZQr}qHpcq`NxzWk27GUIMB
zxMB@&4R7?nEuz?4n+fx$sWzq~l|F~R`CQC{A|mQulXV&gAl-Hf^&UXk2wJH_(D52N
zwJ?wU_yN>P{gVK2od>vpGu`xqRYx?Gu$=m##SP=Z9Zf+J0)i-<?Kr<2tEKv(gSC}+
zSMJ$r$S%9FS<d*EGw$gy5@~}LlaP=w(g5uikeUS=GGq+z;<&_piv4hrsuwt}4<o9B
zUcJiIKnTsfl1FMHTfkL^Dqz3uH6Z^2p)KhU*NOm2fxSxxzkp5+P<mlV-`>2`5{iCN
zuNB`4pj;=_n$D(>V5PnoKkLCdLcQ(uc?LHdC5+F^55){}j3jZ_Qc*YQ@Vi;^!N7Bc
zRv$==M{vWe{w0)tsVbO4gsEc;jMHXWhK63(j<yDzP3lzQ(_>|!phMz_;NvsHB2ZYN
zdQu&gn5fhSt*BxCQ==!LZ-tvGdTnhj6zD?`;DA^)ku4RQsLq~j)%F!@wWHGHQvbTh
z?h~K$r=XAr^J(UA8&$rC5)D+rSRA3RmG%Ivr(}pRfOiL+mMFaWGiHbunpyQB3AMFo
zd9aRNGnsr)*Y4Se9Jui67@?L3tfY(#x)9Od^5Jvv^7h1EVfr<^gu79A3XZ)}iJ4cv
zn`M#0bIr3?5j%1SKmr)h8_h1HUSz#9Zi$@v^|x(kg_p9PvTG9){G&?;oYkH&XrM#$
zPgj|2?!<R0N%f~B)U+hsit7J-H4-5BF5XM29MZ8ZPd6CRAb%Pryk&z)f#lBTtIDro
z6jLMS)3}xKFW28U8An{pGOP%Y#PBaVNG?)PT`u+V)zhBIsmyOCk9QRs7`5dgS%y;;
zIrJ361r%o9I?9rSa%r$V>4AMTDrMzg#(N$D{Jq#^(*FbzF9CpY`<^T0)Wv&Rr|Cbq
z^nH;O`7g{Db~XLBeN_UIbrbtP|FlW$kFLz%bB<plbMcb3=rc$8hmKu87JVef_-Omj
z0FhsZ;j1NlG1^H)+h~!w{1XIPZt<JCty?n;L<Go0B9|hJlKOh}{Jv7k3eUCi^cpH%
zRMp4dE;k7MD)^$v09oz%5rU6qNk^wOKGP=YusF<2u`)bM2RY;P8{s*~YtP>!d|f~z
zE5j|({g+>z$R_(<H{b4z{b?_s*5)Fics?h{)d>lMP2>@4ko#Ikbl>E4L&9Y}_mb)s
z-xdi)Qg7rppC8xee1&vm@+*&hY>C?ST`-%}H$Lt<j77sWOSRVgq_&}0@#p$2^h<vo
zN#MkmLuuq+F^>lA*!b^W&q{_cF-9lR<^<thbBt}f@^o59mHnm!SNRZeIpp0SZ`L;x
z%{4{Q)BjBnUbOt#r6``WnyhA>%-IiSu0>`eR~08tA`JYi$ozwg8~cck#RB)y;y3Sx
zGCS0$+*3EAUSTMz<)FS|v7L&nlyjYJUj<#B=h^e>Wl|m*H#X<(y8ZX-68=eRjiZB&
z=;|I`H*0^GpWdRXJwT~lG%0?qKqGOwRo9tlf9u_TZTN9c8KflLo3$R<^hy7Yu{_s}
z9tXnP9r8a)_&=V9a#wMwJtfH>dF&clVs;vB?Da~5`amxG<d_QGI+d<qd8Hmtq`}KU
zyS87t_QQVZMvU+ePt{Ksm+<>p^>H*}5`&QGh0(%oAWNQaDoM-p`lu`}Ddy5=+QkR!
zUrT<L*fo6o7*%q3GoA2FI$69sX7ZDEIo#{s4mUMSp9$T?Q6l!$X}<9)wvqAPKYPqp
z6<3X1+qFYXQ5?Vin~9}qC6-f?=?;4Z4#)bU=-40HbmWe_e$Z}kvNltSmLqeHH=U0m
zG=Wb(Pj#uW*=zm%y7+t>UNY_UR*Kwhciz1;UvXu#RN)`A^*lF2qV8i{LjW&iJK3s+
zmhn{v+DS@LPK@xc|G--4uB>>tz>06+g8AiSArRkED2Tmg4u}lUexV)RcM%j6?U#HN
zv-`)ZU5OwnE+@y9+j}uHwuLk0IGvYD=2_D~O&h#N7}7*}xwRfR)L*}Tjlx5_1!O#J
zZS6?BfG%K{gDfE2L)!3en7?^KQ&q*i)8~sHj)PG6G?Cs?V8S@>E2ERELNXr5XRDra
zgWqXca~^eUz=8`0N%zlB5s<3^NC5i{+42R~fh_lPylShI%nJB9(9GjD8cq^i_u=IH
z7Z1$(z>MHABwg+3O_$G;(cRr$G$tI#M}ZynecJ~q=t7T|<d^h<P%;OWBON5QymiNf
zNoIW++(Mh5wgFTo&q8YS|M&m;yslBcdiU-(@QY}~U0#m&*+3Os{r(?d^X_CLVk)5K
zfYN1UWyOxcvb(?k58P)p_H$?s;6WP^Aq5_FKnZXf0D!&VCD<~&O-+o9d=FeGN5?JT
zi~e>(sHq1qo$b90-gXqjUjQo*Vipy>`$c8Ia3>qGNKJESHj}vIDik$W2>^hWrThA%
z#JtT65ixPO*O_PG9Z@PjW)_yU9~Hgb-CW}0Mq;TPTqIdBF)?f*b`~lsAwaDS4pz3c
z&HrFcci9mXFlbONfsPwsD6x9YNV){l=eyAFE>PUN@wRfE%O4Bym_VuoKo*4yCI*FL
z1@zHBLE#EA4G_m^45Luo)<wt%KxTRXov(nf@o^A@tbHJc21+Nu`))uNx1${bNgiD6
z4>#vKfH;MVfuT77lyaoW2dE0xfDfEis#r`a0D+O@a4GPSPJo|cHw%|xp+QNpzOsNQ
zGeSj$+#Z4ucOY&D;L#R42{UbMGsBIXxeZsQh?p3o0d~w_p2kRljtz8H0W+bDp2yHM
zM8yptAz-da0tE+X2m_cvRaQ9!{o3Grb2&RHAP?SvW_!b=qE5?v7R^CGtr<u#s`H%4
z-;8vsfToc1H7cR<f5IvUY|L$e3bp2B7X{QHHFO_3r`4fg;@9CdH8t#M9Vv15ULDfe
zl|%+A$f~K~+@Us(w#Al{m6e65iQ`PIub0-Au@*OnpfufMGuWZ0NNt%8$o{Vd1mr~D
z3<5t{Udks7`c*eJ=1=8)e#404)DU-MRTzLcJo=WGmjfu3CgK<n4&4#ecMSfxVg-P9
zguwv2iW0(IjyAxmbmw9lXaxfCB=@D2(_yv2ie8~(Wb`dA11@adF`b^CwPp}k7BDbF
zP>MlN2ir(7E|5tb8s%w)gOeisZ}7ykojc?}-j^ri9XibyyAt^eq8gQ}Gdbj^NJwc0
zuiq)IFq~0~%<<<p?s;>9EJhsNa1gz{IRu7YvA({(lk`{MS0zOQ-LwOAgM;2s5rJTG
z$i5$BG1^W3T-+WG*iF!t3j8rNBtqSU`M{l5KA6xBH}~hA+*c40t)+-K4%st_2D&j=
z-Pih<3-uk%;#er`Av_|EmEz*!q1J&0kZ;36$^CYN!})W@8(%OdDwHGIW@aK(jC;bU
z7zXkksUhe|gCab#Ez%BsBBj+;M_53PUg>dAn&mpQng9Cc3C0eK6Ca!h5Ze|=sh(U}
zd+{@lPg7Y-HMjNz9^iHKTR1moDEZdB3y`U-!>-<VngK<?)sA^TtAO?a1~R0*RLlxU
zk&1x`3&oyl$E7RGZmzC+@9yeJ*ucI4HD``5kk6iszL&Yh3!GyXG7`)j6y1q@f89yg
z9f#my(D?x*k;YzH2DUafH($LwWQ4k*)?>=&Kfm3g5CdrdL;i1&e+w7xPnBASI&fiJ
zTo?w;8iAG@)za!8(7Wa5MhoQq(d;Xfdj#wn?a!HT)ZNEM{dmt#_|mov`6<}R{1$uZ
zw2(D!fJD-6&OboO1hOb59C%yM{YpzC4FXSKD@Elh@Zi|R<@iE{(hE9>YA=ot!4wxP
z6~Qim9p$uv@NC8_wQl8c8d-ipEoKE;Vh(Ugf$`P(13J8kh>CKwIn}78(zV%xrvm5&
zbiJ%Vxd_zrIDwlNhY!qvZ8Egm+S*E~@vUH%$$BzpkVYBxFs!1&8PU)t4cJ}-G*Wln
z_6*o^L@a8N0(;zgui4ohKk30Td@7P{(D-zJV+Q04Z=V!g)}hRTEdZ_wH<g{9J$p2e
zgLxJDAL>)E>yg7=hN~eG`9WjUQl2SR5m!LCB(JCa%td<-4kiwCY*H>w!j;lkO2gDV
zGiiG1guAfvzO5}3=)ZpbnxhS<ZlLeU^97HDp?8J`5`nr;;r{9L7G+ug`BU{<-ixaS
zzdn?g=Zx#TX-Y~?)_Ln6B!GeF0Cujp)OnsfUmGtlTHIb8R>eGpHjv-m8fa9Ns7%}f
zbwM1}b7>7fdLV}1rE1xR*Sma!iODY<u?u6Pl_3)XrKQl-zF~ahaxfoLN{V<Hb;_PS
zCGF)m(?;tBBDgwj+#lul9hdv)*|#L0z@Q?oiYc~%_$(&~=<FZj;)Y>wy*lQyT6Qah
z_oh2hPBc<<CtH0lYb<YEN<~%m?VYkWo7CmnM@o3hE|BcW1>k6@L)WG<fj3s-eLMF-
zGs^a5*{2*%+tqfQKDm4a(B%=v4aKtS<acS_Wny=<qi)#+T6977{U4v+iKcrO%K_PL
zA8N>h{&+k}q4;)oK%P>rQ3itBvUI9B;0IaoHXZ{97mkQ2kF8hHV1%TSa2>E<6X-e$
zcgzraYr#84g`AiU%;67=nznsA#Bs(@tcLpcEp!zF*_LZ?L>ESzdi+rUvMMybjg8s+
z{o|%}4af>tHKAQ@fzrheInb&CHLwa$UJeToYtGx5&;I&l1o<AYeZPA6ls;C3?3094
zuO9AyT|^Y5A<u0CH-qQF0@>y<aKZEC<beNow)(atU+7H?uTiDh!|!lMODDV?qBgE^
zBazAN75CVWJfz&5fz?tNvBaqvsq<E0WlZpCI~ts;w-_?ThcAbOynoL}#SJ{noSebj
z{F)jG5`1PmsqI6s<N)5Q?_G{m?KjwZqOQQ-Q$fFkMFjW$Z&}^=Ct#4lQ*c91xiC?*
zYhTkq*5BH~B5N{Pvslmh_n5*;XPC8SB1lQwSy)#;k_`93ffGr;+<o8-T}~SrP6kb0
zgkH0RwA9MiWew;59>x4W|CbPcqyeGByb=U|YY{Y(9!j*Z@ZLq9<-qpT2fqyoiJA>Z
zOCXy8G7m?Y!K4GVO$-qkSpuB(R#sNFXG7rW5+ROlU|=A+1!3qmGc@OQJ$(*%$l*0+
zX6C({07z^b4pp$Qu)uu}1axyaTY-VOEq?Y9NSUIdqJ>XD$Ouf2w;re`CFz2?%e-9O
zvm4vnIZ_Y_;p`A!xDb*8)b?05t=Vty46CFLfmN}b4a7OGk`)85jRtf$ZQ4^2A=zOK
zh;jT=NPsjCSBy!OL+<P7#KhYSykMFQy&B!Lh~ow!DlwV8gHS`A7U3{jLGC)&Qxv58
zsRnpDWyH`>Q5gdJ1AZ3g6Cio7l!L}h1GW*!UbnxEj5xPP(t~czbIgkyL=+%90hxsU
z<k;BQfH0gk30MiW-t*pELA8h)N#iJBXlMuoJukp$%io9%HWhq54aucKjtdva-|=aZ
zeS<J+d<D8212U3FhXG?Bq5{>Xx0}sXb2h!gp<-6e2q`GQKq*N_RK(ZebwR<3IyXdq
z(&^!pLMmc^z(awT8wc*eOedgQE^zN0ZSTT;u!;#X_t;H{?xcR#Db}T`)A4IgM(6ea
zfI~5#6(I*}Z*B`;hK}5#-VPa&*bCiBg4^y;CkrgZx&YbTi<h786Z%3BXUEFXs0mHq
zQqMQ-3aA{qx8A~3$fYBm2~|>VBIe|R_tYS`nwy=CMPjbp!RVNnbW`&7OUqsaSFgUS
z{yqGbER~jsXzRF9+Eo33h?G>neSsl_!Jo}qsLekA9kiFb-M9g_DYS1hhuKB!uxlTJ
z7m#Q+%1}c8<rSa?K}P}zrXc$2fG&~o;0%*C4X#?iV_$8Aw&$OJVLGK2dw?)n->qw_
zdNck|XhCOHY}xr0NOL^bVUE?)#FK35q@35rp?xb3^r;1?ZM!T_BmE7n*P<GZXy9oX
zBmtf@<{2e=0~D4ehUC{MaTf>|u3!D}{d@mZLCPlxsF|nQO#rY{Vn=RA!7X@YtHPQF
zhCuIzg+IuDg1#~cYi00Y>cw8FL+RgE<ZQWsZmtUV<<=bH>eg<n!ka3116niC8pER0
z(j^6LJ+w3JgzdJhF?`y^$iM)L{ARi~j%ER<H&t!c0vjPnAqC0bf&!)7%r=-lXurIl
zj>yi(6`=K<bgzHI*w|-~sCP<iJ5RdifmXsv98$91A|XR@nidel!<{2na~opWmEVAX
z5l~VVraWEl%Y?*%Bl|=sT{_i&v!I7M0NiXi>t;iFYcqVoJ^-ZC5m!{=xeCHRa8`c#
z!dVx3ssqux2~6CI^h#}kSeF#-a>4*vp~0U6Y+nZn^|NeQa`PL{e~~kv!m@J`2W3(x
z+o5}`E3rx~(w#HRM;jJq-n~UAjtxBlUU8<z+5Q}!D9hLO6I?LM={~`ocM)YIUp;Fd
z88Kj&<iIY)7{Fo1DL-=<5Lpmgky5bFip5SX`u(aIM$YzQB^yFic%fBr**P79?9(=V
z?$<B1;O!~-N0%?WWB_>{4qngYXZWIx(jzNFd9XW2Iy#h<frCvc_M9tKaU^H;z7zlQ
z^mN7*0^sU*r%R>xP=MLTZi(j@cd)pBCY0}n`1^<Rqw~j!GwS587QC4PnH|yUi=iki
z6?#l9(Y9lMdKQ*|p<GZ@imnv$HA~O~XZ_UKi6!?843zxoaE7$BXW<PG$5u?bsoH4U
zo>nl$)-8btgdH>oH)d0KK`!t1jXM|=fc<%%o}QgQKo*oT`8qK%e^62gNm$>Bp7@3o
z!4blQ6PuqCsSzFDr*SJ)RPW@jqoborCOO^UU^stj!=*#9g4=LC-x|^f(TIb2JNhGY
za_IY=JNo0t125E_F4(v^lQK5)cd1&*pOso+jG+?nGT{tmf~eg1s2tYsgFOpUQqt{R
zh`ql_vO?xs_35iAdF&P8J;n<f?kZV;c!E8E>sSRz2H*f%liFv+FUGJjQFm70W`MRF
zzRzbvu~46V13{qn6ugo+w+-=iUbaY*`80avi{Fcl#R#RWIWC`!C~Yj4LsTgts1g?y
z$ly9a3Joql@p<fiw;6Qk2Wb6~IUQ}jX6D6UZ)k8T0HTUi*CQ31TNC)k+#H<b)q9if
zqI;4iO5_fvW{?{t3}~d>$gEBku*Ry<qfzV&IDODdj3y=~CTauk$`wWwE|@O;;^ZX?
zYP(6kqqDue>qO)9%!~{>#4|7Mn`dA@KU6xp(@#c%#sS;VPu*VU6dLd@9ncoQ)BpkB
zQ3_iGmh#7TrL*PRZB}V$agwQ^u?WM0mdpyAIK6^_^mKI6RbNjOd1C!BhT{AYuL6U^
zp|^{**YjvL7W@zitY{v)AhWI<#J2f8dz_jYduryRUz~YUK_I;??oZz<f-|oUflYTA
zwmUdZ{GOC~Y_7R@w1V@lI+E`gvYG9uyu2k<Y^25RN;Wh?e#;Iixc-}4(JD3I<N{rE
zMFEIkQ|r3Qir_fuYq0<EufA24o7aHreR&@4>WoI{6MTy8o1UI7@W^aYXXDSGC;^)x
z&aX!ark=_p(SU*D2^=hEM>w^nn<|k9;8MkdVi40c{0ZmY)D{_<_Epg(h!VZj&FPlC
z_R4oqHVFa%@i(}WMZb45>;rH$2l7^jG_&BegZo`453AabPxH>ThohQHRPeGu9)*<h
zQ_XXsZUXsx3|BEoDMGNWfZ{<&C{8jYBqXZ>Tyk9RrmzW;7QK77QVwpDk{Yp4vC&-;
z(S*RqjS~kH-JL^jsgpO+S4&r}E$~Y}iZvuG(I7Wji1(|g5Ki4fy1UjPimUni!epBS
zXbYu>6V6j#PqbWIDmT9jiS2ZHc|f=<Y9m?4cj@xwIIcU|OPRi8gaP0;Gbk+nn<gaI
zH9Mqy12jnb2Bi!iGorb5-h|EKY0QCWV4IPcq!&jkM0g+?uH&Txe8hI<1rgdu&R(zh
z*QHF;0IWD2UR~%A`<?zSPFV)iS77B&jD}>t<dO)hgf{I>gXSt@Hd=E!#;gb>WKsZN
z$Ul1ry^(vilC>8P<F|F-`qb0$gw%BbCEf<hmqA<<FB2ca!ojq~s0co`rkhoPlCktr
zHpSkE-I(YjFre)G0lE3Wk7MDqjJouYfNRtDiRg7fTdX?uric+Rv3fc`nLF8uqFk%{
z<-Kak%3A4^8#_CCqy*YF{W{J?MvFf&<Uk5ju*XEphHtJ(`}P<z_1IW2vw_@Ks@F1Z
zAch?uU>Dd`hYFCg=&Vf-`X69p`o>mW_~^i0auVAB5fI#Fxm)@707F7rfFGIrjW=fC
zoI=xgMH`W3I64-pUKu<#S5i^QiUpDP>hWjrVvbx{`4y<3p1+j}3#nq4316Go?FVRu
zf7^(}&0xS5-gzfb<ENs3baYv!&-J}-k~tWN*dYfTAF;WnyWHpN)BqHLUmGIiCfkAJ
zBD+>m#G8K?4Ba@0rVZ}NUWKN%JF!<u%l#4TrqclKu27!Q;W%m&b#kV6{>iU}pQukM
zdy!o2ek~NL4PL5Sz3A4rE?7TfylSrb%~snMn(CJJ;ooTn0B7^K@YvsbT)K*J3gQV+
zarbvP8JBclLj0nKwf3Mabx7hmIor#P;}=~6SjF;439a>;j~|5&q#-zR)aY*4A<tor
zkp}3k!~=DKoWj5#i^K`-Q4mBeU@_zfhz?RhgWPPzppTb95bXcnxLwKR40Z00(CGl*
z)yVmGaIES|kNM&-DRf@c)K!CoWDd#f+MU)i!>EMAd-_iuU`hJp8E-rX>CxWlW$neZ
zfbD+pbBYv89Xo*YjV+b^8UK*8t*MA2G!SQ%l?W=G9)RRjEmd-79Ry%oJ3GU*@j2Lb
zhu`6n8oas6(zRW-?LMq`)yLQOqttM1VXjl^<t)^T7njuz8)4;fsLJWA%XtOt_Vl>Q
zrh*gDN8l<I2`2V&lL5*J5JkbRmwX_!r1n2O^|)>w4*UE{EKhwL4ikR{P0Fd*80SOI
zi*hHPYqWXCnNo5#!=aUxHxAFV*VR5LrX%>k_aO&ZRaJG~;(k5--v`C%>Al@ttrs30
z4nc$E??c+g7^?ni1zu;TDhGs2csZg~g3TQr_u<U#R5b34j@Vuh4Rr)xas<VpFs5(U
z$e%7>&x|b7E4m%!grP@r;Q|hE+ERg7kIfFvf<V=VaO#ju(nAVz^8IeXf~K!q4c#L*
z?d&+LwbFiYK(ZTRq-ujd?N3L<Na7mDiha6QcLwMNN4f|a04bv1qTqHt!R<A23?Ap&
ztG=?*aV-qA%9*XO582qVva_FwnnIMm`q)Wdr91EF!RY%LbXpJ1`7?r5D`j#(Q@<pf
zqG=Jgg;B(~-(K`52mcg^%R~8Sl>{k;AB#VomTm?e*~^R2_bCT9)leH4Ho3yO2h1!3
zdmav@2fU)OQeZX((7@n#I!s~`pq*vSwmbUZ!n)WPFoM?sw_4}s%}!jTeYLx*E4>Z;
z%Z12fs7C^M`gq9z`M>{f{&Q2feR~vQ*T{3lUV9*7jtrRt@1af>+nbymx;>oN^Yd0#
zJ~G+;wl~7%Br_;4#+uE%hkH4J2(7s)6Y0vV#%Bwz9P~K(0T-Bo2@vE6R4buNY%&;4
z4-^=c$Qb1Zzl51AHMKaQ$h?ZjYP_I{i=u+x7<o&8^<6c?(D$}Dz79`dE^u$-TU^6<
zRIUrvHaiSWAAGbG?VQo(5CbNlY)E~-m0@6WR1|yHUF3^&z=opgW%CQj7ut3=nw`9i
z6HaK!Xd{?p)P{z^ibXfZ*;T3ZjY;JFZrtSn5)m?*Y+|%8()z7xyWZ>Kmn2KHE=Bv*
z9hdebxfArf{C8A+%YHKsIk)XA9$tJ<A~6@<^k=$AFu~oNhUq_#VI+j6uLbst)&6(L
z(z;Nf0mFho8-?1sK(h!3_|GBplfjSS%ce}59{b3=o<GfN?R089y+0!Ix{G2kGB@YL
zzWvL;fvfCiNlgN_vg?Q9Gci1G&-Vu?lIp%UkZL3H`&;giHxA}dy3Z$L>!a{`IC@)Z
z^P;=jS{zfL*drpnVf>6sp&$4>o~hF?@9#66IY`kYGU=B^zOeMVAeRCKMdW8USg_-e
zF*ZlD7`04L?;P=-Ft1*H6?Cl`sQ~H9y(Hv%+$bHAF|N7)r*I==Dw}KhT7#bSfB%>T
zXBTfF%KI?hXp>VoUZ%rhp*xNuMNa;ozc*WEc7b~d7zW_&FwoX^^Hwz!{ROz;(u{SI
z!3Vr99S&gx>IHw67B%ujwrB5S0xzG|8uK~gdG5vSWPcQT@-X>s7FadHa5Rrq-h)aM
zGDu}etH84@gXMK{avi_`$!S1CEY5+mcMpy$=wF_8!`=dnkO~}m)f=DENv};A(#%&~
zeEqgvu)y+-chTOxe`Y#^Jk|_G_{`575;@T&O(~O)c6z3!rq=n0avmx(?AIIX<qK-w
zjaMZfkXbEx@~EZB-hsd*g@K_5ip&|5KC`SikV_sz*aCQU1;DZF!BF->Vz^!kDbHhw
z!a`p0jb6To%e!$|f}HO(>^O3fto7Ov0XNC{dSE0#U~FFCgTjtCeX*klhe3%^m25C1
zvrs>I<<<j9>QZ-7Os{h>xa|QFp8c^BfRN3y46YEMZ1OmLrtZvm{Nm?M`Ngxv9|RpF
z&W_BtDg|W<I$C8wJ}$s#1*aBd_K|q`0rgK+Bl-1jG-iB-U`Xm@TSxS%gM(dvmj7!>
zprQicvbdEBMj2NHtb%rk!1qBJZL|#k5<NeE6bVnB?81ycTD%d*-rZWSVUmM;H+j)b
z%oJsSZi*V9W@lHZm@R`G08QKtvQ*}M_h9~^y2Nu8s+iy@88Q^cwgPDs;)NaB#vM%1
z;}BT;=kRCejFq}wrKxB0z#&SL5@kI%Cams|uiOYpWC6-2bG~?znujk|J0SzMw6<1g
zYiMYIGM?xDhB-feK(hU2b#v#Js?m`u<Jb;kSN!Zaq4Wm)C8%0z;(E`-O+z_v%4O~q
zEW7h$7|h+&g5gqQ_9y4JkcaA&`}Q&c1w}!M2LSBAssV6<6^^@8Q0Z@;-IF5-pjg|?
zX|5tOUC)u9>0rR~nzZ_7<<w8B@QI5bdy@iO)mh&NLjrta1+0n6MrCz%W5JdWss!l=
zU(x7s_$do!9~Z~6FQspc(YiZ%Fh%phvBqpg_jd)si9nCp#4EX(>Vubh(0K3+HN3+Y
zhqDCZKc|5$TO(bNHt?~`AMZUC5qMB5f16_#s1z3p>7_9#1Nf*hJI8){lgx62rKiW|
zsd>!t{P)pt%)Teu+-UsPnCicrP8ZKVx2%Uz#)3@I%Qdu&OY}G;O43<CdU(74I{i&|
z%oBn1uKnZkm;X6u+<3szf5F5(#E@Wn%p7bBx?8d&vj5%U(5=|eyxmND;>G4bsKT&y
zNj1aI&U#Dne}?yxZ}}Y`wFjMs^DhfitCMH40lKV%x8YR$_qR1>q&M>ISBg*kN&lfx
zwLBZ2Tv}f3n6H`Me}2q-rg}-}$0yFZM^r{Cr^Gcts>P_peWi2K!ajn+G392i>NmX6
z!nciT&Icb43(bhy*$7}(K2ycg`DNPrgNi`h%dF6_Un<AU)s)7{)~+S*Pb>LSn@>w7
z!hqa-N0Ca#9R73>`;iPcTuRQU$gOhKg+=In?3WzXD1|8seyvrv+cZ|A^G#tKJtYdE
zHPkLGxEzUHk2mJ@W~Rg*j$Hk8SjhM5lIf&)7xkgorsRh{gq7;-3!^6f#7Es@H1WYV
zwbp1L-IHrwN$&+7S?XhAd-cZMCK3=jaG_c~Hhf$vUNIXos03We9f9HEB8rNhWX$*f
zeM%Oehl4iT@$l8HB{`c~{N0gnfj1(QVh%+%JFf<|B>2bL+%aTe1?psMXSEIg{3Y6K
z8KR5k7nUExj|phR4|uN|T>Pjj`RB+&wP|(r?Q6z2r-I2s?NbK5%I+-co8BsN+1nu7
zPyZYc+-aQi?VAU`i3SyLY$kwha*EN^o0NPfVz^Ey$;LRcO#)~%p$}3_Lx>-&#fhCr
z-aH5@Cp&FyE3l~O?w|pAmqGRn!d+ZLgNxqfRipGock*?Iyty1N$Mx)iTuS#FC&xcb
zBzk)NgY~%P&n6OeWDmPK?_%-3YOi*dqj)Al*aP?9-;s|n<2;_p5}+wgA6z=!h`T8@
z_I;C@Q{UcG(x|A~@RNT@FJ?*NR;=J$FGeS|Rteh;gcABi22+;YBow*y)*k2^4-Z!P
z4^e&kN_~Oek?p#@C-dE+c#%(hEw|VrFLNu4)IN>t`tZ%9hb+PnXw83DO@{J(PG=I8
zyrj~=lu4_IJPm3l&2_GiH<}M7E@^$Ad64y1^JSOmmKvp_%JYDLqn*#!Kyx#J3)?V@
z@7OtzqIoS+#XyOyKY(UwUc9q$gf^XPKbebChSJT)AT2paE#_S3jl3G4*_!TWm-i+{
z_WDmV=8MEm^JH=N$3(bzs>*6oW<X$5P}0^PEmGjV=t#6qGo#~ziICt>kePjUJX@x(
zqbB5+X05YWmi}and=-sIJO)kd&7fE2yRScu?brow***Ho8IirIb`hb^NF(zS1+otN
z(<kG&mMQ6$13v=iJ)im~$V?_CdYbh7hC6)vXS&%_DgFnq{W6Jpkl93Glc^y=TlI?L
z#uUe0lBw>BvS8N%M%g+!>KFb@oCw`(y=>*<Z>kFuU5WA%A5BBd_4nn;lA>8xuz$WM
zTr_StO>dg#{7E>+n&KfUh3l1?-6?hU?kV@>%wu=e*H6)D>!FzwoJFu<s+k}t3fw!Z
zzHer%jKy0Mbr9R|`auBy=-Quol`}k3{eAn|1s0v>oszWS=-h~i^Zc*)B6Jia#viSU
zWY$LTq^zuuoay1B#>g+!_EwPjWrb`VB$Sdmaa?zIk@@!}_xZ^@aZYwqkGE1E@fZxy
z_%_!nrqB4XEDl{xQQY-zrqZktey)b`{4YEENa5|c5*XkF?n^%6H++51-fWJhmNl#`
z=MdzmCuyjuJs;FrQYOps46hcLU<l@%y`eFe3&eJ*00C}#)Zft19a$W*<Y}R0=9`<?
zU$1O*M+WAf%(kc`Iil{44}Q5%9{T}KZQj-7yMHp-vm~2c!xZh(MaKE36QLt53NoY6
zs6NtvrmM^c<rZZ(x}(=YtIEQJTgAjcN~pt?k@2pH5TYH{9$MzmTSX=#g+u2X5MsY1
zkfr2cEp9S;4!t1u?M_)`VKZcHw%cCNxa-dSv*A*O5grS%Ysyr3@7=!pzwXP_6Dwh$
z{&V1_{;rr``-PfkA;6K!qqI&FRq{G8r7zSvvf`d(>+RMsF~a6;!JZn+X4uo#th~tP
z`{#EClzuKXHv2;neMOmo%1ZL{jqLmcnbW;1oc9@cW3B$g-Y0P*TuNYUSxxHMIJi%b
zy2nC?13yt^WB3LD!2q4ANqk92NO*n->LCcGB!ChE@ne24T(ba6&<I$ygQ5eRc5^Bk
zfi|i8Ns#Sc#e0eDQWojrEuPY|(PFVXT^y$`l2^P^bi&!^a2Qi|taZA8+)Tpy*)Qa}
zcbEyQTuZWjmSQi<7$~MDF2tHWhN>lES#$m2=mykc>}C;9TK#u+=H?(#9jdgqq@to?
zPQ7FX2`O0bnO3{lLkpBkGJP86iP5fYUP%W0<B`is_?(Y8_Fjq|htyMin(Cx99DN>2
zp_1a>X2s1hJm?*v#QD08J}G9T<aF1kuOZwZL}N*>HSDKbqUkDo<j>Wn6QaM9??H(a
z!bt<b`9dvS()jltI7+CZ+uPd%4FkNcYNIDhtyRGYP69f2;S_8>uG0CEep~#Baq@x9
z&Y05&)5JQ1q=nm?B1Q^Rv&SYr`9)<DvvgkB`03YX60Oc6fy}AP*ZbVJ^xkIclyAhA
zr3Usk<#V4WHeE4%`K;w1X3nZRLY^1!{4*GJ^JOS`1E0hJvA+WdyT1WUd;mgvn<fb6
z{9sq4jORbYEA>Y03RkPQ{MpSG3M`uloYnn%mDrvB8;`y21=-UivD7Tpb86(>o%H&h
zL>kM*vq!jUcuz$7z77aq&qDFGg1V2^<S2Q0sY>@tGsN*#N<@z;r#$XxO%#Gy4Njqk
zCE@_6H~<%#6F@#G9<3~BGsBRp`Pr=ph)zI|itBZTa&I}5L5<V;0LlWYGOHAJXBJp$
z%^OT}C*#;X;eL|GLiWIOPn|RAdl^@4Wc;?Mm)6sVozgdPPqI|Lh)|s^bq<3cm_+x7
zWxsh|jXchYKM8q4=6~2r<|?SYA`gRO6l>P|`mb9K4wt`NUOF+xRPI1uf3qu^rR7;&
zr5CU>!4NdRyu4glS!rd~zHB3O$zR_m>_a0#0GP@Ao~knysRfQ0#7{11fMDN8s31K6
zt^t%5-|M^vwvNE6Hu&h0_8{JX+8mhA8qN<Z(**-a+;_(<t-qG0U39r6cv!4xgS)5`
zFL_3D`_SXQrCS0!(an{+r%yG#6_hD?lg6#rsXHlYfwX%|l+#7N_V{;1Or}L1>(uV|
z`<an8rohD>$-BV3pVG(&XfD(ftb~kMH4uEZGX?Vwc)Q2O5;MO5n=L?_hbjySJ6FTs
zmS8CoYVl;=(LkxT?)u49u^+;x1&YDA-s*$NIi>KpbljTU&&hPV4(+R<fhp6b_JVys
z{>;kaWiCz@(O(vL`)smE`p{V>)+Py(&$>@*l0B#9Gs=8_4-CmXKIqN{KzQ2t%g^hu
z`#n0MnOE<JFoFLHK*{XDp)2vD=P=H?Qe4hXmSQ2#xudx^(xAR}%CWWx1n0}8N>QfO
zqpXf(dC@H6)vAr5lgcA!@gYvOd9-hLauR5t`#5_r)oOmz3VZaW#Z8g`&kfchC}S?y
zPNSYM9B##$@(&ubrQizuJ%#v$=TX5RhYPYuq}EnHF_B8>nwAb`KTshATyGC>$=2p(
zHda%BVW2z=aNAqGx8Hy?5_cv{R?BxBU6r)MM%FMo%i*5@bktbMTp-YT?02ckIejj#
zm(yHxx;EBJi$@xIGi6D+Nb&9rGh4#OcA*VBZ-_Nfdm2kh>>7S^yc?GBuY+g}@yu2l
zgw!myC?r~#cp&|DGhgAR*s1#QgFW+qHYh#h($M3iC8gK@!L`>=)NXBSliF&-zC*se
z34lS<WcMLSW{+_PigjvMnK^Gfo=gYBGk*};-rMP|Wvkx4(jP#w^oL^eE17kK>W%_w
ze?$2CQaey`_e47Dth)>*Q!N)RxgCAK|D9(5yG!FqbI5Ojvvv;BaqpKS?(P@=<`CF1
zvZMiwGHR{?6N!7cAc;+{PIBu|n;38X87BJLW0UEc!nK;NfYf#q&E>h6lFzrjy<4lE
z6zzxN9n}XLTrciC$=Zg99F#MI5ux`UwOa8!h`&;^Xi8`@)ot?khjO(x@-b~mu767&
z6_lPVaI6u2P$~5#@9oC~Dvy-2R%75A#^4#cztVEv>QK&XY53K(y17XGOmQSjp}8tL
zP<T#0LaAR!t(u&EGK9#~tva{<;OO}`tD(u8rVj-;{=L19XIr9^<USq_7LtjG$2`$i
z6XU)1adxZPMBik1uf0rCZN2mc?(DePMi^nT^`vK^O-HJ$(FO#q=ea|R9(ipIUBqBd
zEgN0yCKHo_I04t%hUL%d?+(1?P0v2<u#{n2Rm`=FeAp4ZTPRB>YOI}>#`cIqs=WiN
z9GbNu*WmlxSaL-y54M726D6+TFCC&PHv*+wJ1IsZAdRj4-AB4}80|Kfzq!uhU@T*b
z>m~BpaSVIRuVdk5waiPp(h(Rk<OqWHejgLHjcI|VreE?_@C|hHf`RF|JJ^rD*51DG
z(@*Z-k2IZLaCHIv49{3Pg-TE@*0UyRXS<mS{Xi$PM3?@bFFW?j{9%63*tyM<_C`Jh
z)s;QCIIxT+Bx#vLT>XXeBRHcG)Q_JCZG=y&>e<_dIf5X^NX*{(I})AYDG=zRvqhU$
zF9rVlu>ic3`OpH~5&mjOGDTbxnLMB0)g;S5ZV9aEBLvySIqn3o4de9(AKqLvvd5t7
z2_;%;Dtmr4*FX>3RdRl3=?4>b<6jU=nVjWwCz&R;#8csm_E$FNwk_@3gMST6ijDnw
zP>h+Tb@1-{<#V{P!GaM7CD;P)oD>SOXrDGl{H5vd;~_*KVp#pv^?{K!xOp$lr%FnU
zmPb;FaQOR`Dt~%)#ui)PL7Hsq8DzJ(V{Y2ZzxJb$M_|SL-bTJbV;Ymh=QOuSW~r+$
zh{vc0ggEkGJ$kq~X`YjVY2e6fV5xz2lI&W)P(~1Zm^C;l*&HFly#Dz4jUbZg*D-+4
zA)!2pIVb?Tk~&nE2ZgTqVA)?W3E#NZK_B#P?;eq{KV(>o==&U`u1C`BI(KMR?{spC
z_x<EH{L9KesI1Ku?@KtkDLVQGl-^_v@&TeD2sEg`isQ(O)oHmH#MUgr_LNW{XG8v$
zRjGLDzK9L+%{<kStQJ(1hYKX;%b2x=@h-5$fr^e2yzzJHADUy_BH&EyN)Y!PwL?aP
zS8`(|%4QRlQdPn~9^Lg>*sPq*>9*z2o%1&(Zmt<X#$x6IsN=RCdFk6SCgULdyh=<G
zXB*rbPGa3Zq?#XLA)oa4-3Zz=u85h-%_4kIrV8#?0=WTN{GGAB75sI4znyQ$y4?`M
zY6|NPWULhrJ^vt+dHdmAnX#3&*x%MG!(HCa*ntPx$a-tQZmPNTp@_nu6$kGoU5~d6
zmamUu<fkc%o%FElpfTe!3$nuiF9c4Ux!(3BY{BvK^@hfUgVm8z-jmGUs{^J4_da9K
z+^#_x#+N3?k|i%z>=L`FJO0(*;{Ht-qldX>kV?X$5pm~h>)-DV=Je<>Ikn-tF`C|D
zHq~2hzUZ|!Avcju@ttMF!8*m9P1L#L{@`2LJdA-Nks#PE2+3(;`KYl;3z!SaB0_tm
z@#%YH?1|9)LkTQQ+lme%zv6AYai&0jBrPP5Bxv&F`Te#woWhJ=lk!{MH_%)3$%(-f
z0PggG0FaOxJCk3pkUjrsnhyD=35y{o2-xhakWZSBW3Z^8)4s2GSXQ1n<GE%-OPVe7
zA%ijt0G<O(xO{XSOU7+)NH!WFzYNP<^J4=`<TJ69TwH()btW!BB8QqE0HD0wWMf9r
z(6%>G#I#HbYBnKS9Oz2|dpy_&i=I(|I08h!O-zZXh4ctU&yC|GB=b&QM6idJQ;5FD
z`vaD5<B7bhFz2%>R<~zT-6ebX46;A?UWZTjjrz;r=N2brM{$;z^$^WclYil3+fZ}x
zU{ZG?DR=HGE{mW7!OIp)wknQz1TuNoWX(q%bkcr~&eEla{D3nBu%-j-%Ln&=KH;@m
z&v;i2DnBX<bt&GXrj@oV{t}w=KO!{8P$7M5Wg<h3*J{g%z)+ive99oOdGi_b>aWr~
z(fkrh_sX;@$5z5KWVFwjf+jSGFB6=D<-v$$4W4s!2_;MA?7d6K-<rI%QvZxkzqV~T
zRf&B<eK6+!?8MuIq`Rs(@PPes<0r!~b@e$dWZ5p*KYvb><zH7<7fnM(CJgF1h*&`!
zF$?wn5@qt!-CH_qs)`nZln6FnO{ujqP{n5t@X>PVhBwjFES;%9o>Z*RoE!BFnH$Cx
z87_KMWK5<wAWncPUb>qsa1o=yqj?_)U1(@%NXG-HWp%c<vpt1hY%$CwkWXrO*=(2a
zeAsdx4g<MLfT=wU_|o@1jn^`s8_#KcB<k#YAY_Z@0~YZ=$Gn4q6AYL#E+(ed<R{Ps
zw6?yf_c!->TfNa}lV>johbkS8zYG>{sw-D<70Xm3+goP5N00Vti!o|HFd+EjA}<cu
z?5<4JodKuM^VcXp^B=&Lp?a#>n*_w>2&vQC(~TQ}C4E1GSVXTj_yDjNz(DTt-ixg_
zy(sVFbkmQ(ebXoJ+Y0_Zt#xp43py}Bo0xH60|Ja58|4RpEz*)q!Uh^?f>Y2VNz>A&
zGmXH#{?yk9Zb>V5n1I0(7|4nu$HK*h1CkPGQ_})0XU8ketSndYl#)EJ*Yr%{C{U>+
zJ2Zvo8cwPQ--7A~nx>DB<|Wh}q`}<vhvZj*_x=Xz5q5`Akkv0XO6tW!TBrf(wZ)ee
zm<GYKWh4^4zu%!WlbiWvDoqO{giyNQWqXn)G~HJpZMeoP-goq|Knu4&l2j535vIU?
zTX`vZf;2c_Vq#l~wY9K_?!_Y=dK`BFnN$V!FvUrcz2B^cPZ9(c0#c{(B&^)Ngz^;9
zQw}uF%(jFMS@g+hYHEU0BSIj~0m7{_gqu$A%l$#F0B#6?XF%P_S8IlVUEQ9D<M-mw
zM1_F~!Nr{?^>Cu$-k*`a`U0q9P>j2!&j)^tuUH;%x=BgZfuwc_a%o6Q4x|{MPbAcH
z<oH=6<qo?do|SjeJSBJVk;1PZxU4H@=ETf~A?xAckw(kR#KgDz%lrlgDzd;rB*++=
zOM0#GZVZigeXV74kq<$x#r<PwC<e%~8FcAUIvc;YPSY`o`Sznw-C?1oAGUNRWfnb%
zwvipM%x#4M*Ga4X_r;BP?z+{&r$Tp$k#&g=*rT=#`Abr@Sfi%qq&1)2H3PsXRYa(p
zqi-PB2;%YpHNZ&R-W)iYim+vg-th;PCW+}6Rk5UC6PZpiDeKBz%K;Pwfw1M~-NGvW
zi2htkwEXfGBEsCFY6faN(z$CeT*Olisb#rn%iLI5X538;tTwpEKKRg;ElFg?yfE3!
zAz>nnC@IJ;C;$u+{(d$xxHmHVa%fkL-)Xo<%f$Ch=@L8AO^av<Hq_}Y7y|!4p9KRP
z66VKWT56w?+@+rXxR4pkzSlz!o+mTX^;>Jna9$w~KA$igzdYqY-%)K1a7ZwjI8<@{
z_#uk6^1Ww$)&R1i0G~vrU7D0q#zfm81`q5%O72Bjj>&;;JSpMn;)lcr&Hw$F51XrE
zaPpJ)GwL49@9yL2aaDqTyI2gI1>}X~<-lj@9Dg5U#LJ@%1xGCP!2RgEdtA%EqeJf}
zFqrOWVZZ!$WSZSBDIBjSt<zIU3+*P#^1nnJt)>9dPN<b;s{|MvvXrGmaai!gx}U2)
zc$8}C+9A^h^vUnPU2sG*{(apqsfRZn`|Kz4x}V(+Utor6oQ*L)wa;w<EDBu2h6t&I
z3~Pq&<8&ON!1FbGqX%i8FR!=51(Lv^kVfCz_wN*F;+C0ZIWga5QeIAWGzf)EB(UeW
zL^#ogcpl}ykK#)PlbA>OAcfrhBtja`<O*xr)csYJUx+fsrT;7r-6mGQTorwkUsTSS
z#_FzYk2-5!1f|dLBp#F*+PkO|{m=74+q?A&$c#zdL^c+>9g0+phke}7$m@o5-s4)y
zbi7CZ`%J_nNfTGlS0<82A;j=W=^q=vc#(?cECUHu*Ttp?gUf9<EnFxauKs;6%-*mj
z3x}~N4Oeju%er0WNxlE3okdH8B=hmFCh)8v641o>XOfY!us}q#!WZdJ<ErEyDJ35Q
z2fs&uN##Bp7BLT`-3lQT5qI@QUfQ{PfSuC23Z?_rp!mJ-$Eut2fmk9H^KU9*E9>Rn
zW0~Pa^BtVe6{1drAp}?s4BP%$|M9Ri8TO)S=e0nop@}wd3ptzr3Wf&cxFP5EgMZ7u
zvC8d{>i!{A#br@EvL{9{w0G}p5gGZz3$NlMII=?gN&czrV99i3?Vx_zFYTL(DzS)?
zqfvSN^wC2rul*e~#<n2!n-KB*=d;mtDxi~|IDNRQkl98s(Sf&Z_prrlx|w&2kZZl{
zC%*-n+<!mbYJw}m5=D}DROXuC2WF;1*v4LLmMl_Uf4<*0AJ&_ZzOWTeFz5nBpuY=?
z{Fx9j^NLKg?OEi>{8vdS?h+&8Ry6~97!L$savB?syqy~MfDYot;CPZeKxEL|BuD|g
z$v}otN%F(2_XVi(NMrT)t2>|AE$B_AWJgFoBY-UZY{J~8sX3CE07Ch{TT0Oo*OdiA
ziipdhbw{Njq^-B~u;i4iknMXr@OUj%Zqp%l${p*=T05t<qzb9Rf5tf6cUyUJz6$@k
z_rtko*IRBaJI3~&+{hCp?VpT4+s~;k9@;hh<?`Nz)no_*qBPX=gR3wzD!N^HGKut0
z8D8=#i`39}Tdzq=f^FcY37$IZry1XOFi4}6VbHQA;zBGn%5~4{qB|kFCY!HF@X$ZN
zTPwnDe9C%yE4g|}fepdk)C1=Uf??ca51_cUh~F&E&@xTTZ!C>@kARR+z-hS;)Tz$P
zeKSy$h9mYbGa{7PhwaI`XFO`<m2KaQh`{y4H@u5MYx1~&A+x&hc00DguG6%X!y;=c
zN$nQ?H`lVd<I&R#&lzr46SGf-lQn7Q+Jx(pyJ14Ogm~f(B?2Y{RtTKT%wYI~+y)m4
zGqnh!Sf)m990r{@CZD+ILCpe44O1t4Q_ZX2MKcARvubmzi>xCUkDs!79#hQxz;49i
zbq!V>{L;Ou-dUh&zd$k*$m!Eytz~O#n<RjAr#@%81JwJ(8{0P_VFW701d(XBpu2?G
z3$(Am7K7&Az8-u=K0ZE@r~8)ka#4}(%6o)>BODaGlGq!KPu`qMhcpY)G=d%vUi8`;
z?y36dx0Oeis~e~F-#CCIYP|Bp;VzK95mJFSb`!86p`AcTq@eBUw|7(&6v6Ox;Jv*E
z2y@<d23kySL7x#u+HLArba?vvtU+m7S&l(28{Xnp@O}*)?j)|V=o+pKG>Lj1x;BWl
zZp3GR&Po%gg~JuVCq08d-V!;37G<1KM9$88^ENVd)}RUPsE-8MrMj6J-Dbm6=s&#+
z{Prl&2t%<S{`{o}pKbo70&3I4HVV#8Zeh^H{(!bYJHNpz|07WK!uT%yW8r{8d_&hq
zHyrJ-Ezr-ex{o>lSPx+of{Uxzr9c=3?z_q505#x^LZ=Euf=p)W!S8kI0YA*b+8R2z
zJ^;b79JA8G{(uzFBTUMBK=eV|ohoLK-4Dj^+If*@I)sR)1Mui7$Z{a!19b8iNQ{A%
z_-tk1@_|`pxoy7>B_}*^6WWmh6PyfW$C{YhBDLEQC8l+??dTN1wRlt1^mO;v9UO}P
ze<`u}D`>C*EH!u<0|ufOJ$W0G25REO#GvDTL-5Lgu1(JX`O6#VuCX!#bMsHw{KWYR
z&^&8Uxx2kRBbr2k;6R5-b0@Qh$Jq@6q_xJ`8$7*@on<|gc!*y+&=kRWdJ4J`K=<=Y
zag(K>aeRGO{^m{26;O2m#Q?gGDTBPEH$^0yqZy>41y&D+3wID-_-m9_W>M0({m@ti
zCPY;M8ipGSrH-Y5#1dM@CwJPy7P0#R<WBjxD3vl$8^V0r-4rx$ymS}<(oqlSOJ_yg
zKH6vnk>ilKwmD*bZ#N5HS^hKoa}Ldp0Ca7FZYVr)(39-J@D?KEiu(X;7wAS-2Leim
zk9;Bk4nrr!JIpb_Q1kcmV_R{O5CShPAV_GMLl-Yyi8E*n8LRZ$4$36Orgg>N7Br+z
zVzHvs(}W)A_ZMF@X4>E1*UU#pLsLl-GX!!uZf0(U<{V;oR>{g{j1w#aeCOEKPUsBs
znlV=xQSv<xE><om$^5hi34=5mDd<3*<&3?M4$jBNL)$QVXl`}nL?UM(JQVx=`*$q{
z*Q-K0LzVp>f`c*d2WWS0<HA<X3uPP#0|7-Fh>j=>6E#<;#hyQeM%7RcFd(X_sZm}X
zI?NXpSwZNZ`O~s;a>`W1KN)L)Zd00<$9gs9nPxwtVutMkjZTxW{6m6yGFrnZvU7C2
zJwg666r!l0AoOT#8_o*tMNQb&!?Ai1pP=O_n2_wueg<Bit_{)u;q5KMqHMb;P<sG@
zp+jM$QzWDvKoF!G6ancL1tbI{ltHCCl<t%+1wkp1?hfhhMp4cl-|w9B=Umr0zs~!^
zOXnS!d7kIK_r3SpYp+E@0yCGU@SpAICvac|I~BH;d3B%|HZ%u{Uu}erg%b&T-WoXE
zf4ZfO2ynBsv}AXD-2=?g$0G0MQ-N3i<6D}HES21n+_simN|zrYFsvtF2Qi1+Ftnm8
zo;#g_H_?Z<xV(oo7ZHTgUKHY3`x4?S&z%z;a|rtumSAgZxxE*ADM*V|YXLVi1%^=>
zV6eiFF$|gdL`Q4GkYUsui*g(r0ZR#RbD-*BH<h%($Tt|4n8;)<_NhM7)63uGfuS<f
zc2S$GWr@N?C4Um;fs$(}?;7q>vJY(W+fC`y@}>lK!Z1$G5<WHGJaCS)E>>GPJ2`2Q
zU<pjYRD5dQzOSpJL1X9m*RMRvVzAXqyT~zwHM;o|4g>9)b1deYHy1ZI3k^8c&(TR7
zNK%3K(Y*y7SXM=Kzc>U1A5hDuJkSz};TvAP!E56C+&Zh4%Lt%&@cMx1(gv^~5``ec
zzNcaceS;g~9yu9Mb?I3vdV?lCE-6dS5>`CNT5-ikcQUqZYN`&HoZn-dVTNlA93FmE
zf<_(Kmq|;nu)_3Ua6o`bP@;r)9kh{PjwZVae3@*v_!XRIps>w?pSkv_x|+iux|FHz
z#=b8m_8(vejGQ^aL$OV-`nx)f5kx=u0B!Gikb{4?AL^b7rPn_;SdWA!(OU>N9BR+M
zLQ_*+p8atLXyNCW4SsN*-vfek8U{=mJ&q<{tagKq_;+ySdZc>{vn!cEECyAGyvf}7
zQ<zWVU_`>k0b!uLUyM)-f-l^S(aVV*phCdh+uWQ9kMVoVOR{+^#XAHv9x5D>%eO(5
z?>+dVM7h<>0FPd~9|pNoCZ>U)Ue)?Kx(QZWFEa<9GCZgp@atYZ2Xky%cn%Oc#(u#h
zKdbq(d%%*?{^R)kx%*zHP<S$Sq*skRU$=sz^&Z5<inP`B9Uw8A0)Ng4oW~`4Li3mM
z9k1Ut-h~An{sCNR+3hY%vH=Bu^&HF(wsqD)ILM8p@dNLedmqNt;9+t;EM)T+6chyJ
zyg(&Bcw(lzGXlx<Hk<%Wj1pJsVaG(ZvI3Dd2lnZ%5>bl-SzlXI9#%mgnlx>#*xSHw
zMZ~9E<eLi55j%w)0{Fp4{oYTXKdV1XL!;so67*B0E_-`>i=uc<n!p4qFG1;T(PZ3t
zcrLKp2f-mgB<q%Z?RB|_X0Z6uFtD-08ae$3jH9^?4QijB$3uYm*y^-Xl4`1y`sOcU
z7<STnrVX>r`n7A6<mB2A1bTYZL_zvt3q~W-Gq@jDW5>J%E78C#<9)oM7NRbA2=mCl
z>FHAJ;K42cMxv#GrPG4m12}mUWE`#Z1K&^tRuep84c2{);d@*o!2FZ*j~957)m3!y
z4Kn(gGaC$O^9u-MUfdOBcQnxk${>WbT$;hw;EXjmkTVSY6(Zw`z`=|2r#;MT*sLwb
z?f<7r<p@S{g%m8{#+mH{j-*t$=V)P=Rj%zzpz?y$!48}@K&sR$HmW}x|9mgH*@x}i
zw@!}tK_;Y<>e7MrMkHRuW|w=-cG-gj>*?;-)5Uznsl1&0Ev;3EcpP=;Gr?;AVstC8
z(dg5=4~~oOn@o{&9=iF1X~L&}|AtHooLt#B@mSnD8MGmnAU`nZ({E{!F6R5tVNrH+
zqGub_dhAle6|Hde`;UjXW!dQbjgy>_z-AMz#S(nh$W|jBIhGrQ@1#=gYYpb?6WbT2
z3JT7m+ZRY$x<|5R@5b4hpC(cS%8)rPJU>}mTbmq>*Ycd(5&kZlu3mj0s-clT7(l`h
z<DY$mkXbzo2e}${j&DD168FDYbM`ysqakFB>tEsb<W|t~55Av1YP;!jCv?A&K{JDq
zinWIdov1STOo(A{#VL)%wbr0@UQVb)i!)jwh_$Gcv}(Lyc&+*L#A8#m0y(~Qw1QLj
zl^}@3^Rxbg{6UwR#7z!2H0qC+gpWXcXWy33(3*3tL8!n>G)heS{~JI1FZtJVck6w8
z@ywr@IXXQzH-3u!o1KXb-tF7MB6K`-*uSFD+)C!pES;?Ac=>s6)BXF8`@XfUllgP*
z`?khT=F;Y-c4p?{;&}i26|Q5oj}+|~h4AZsKkT^S7Za63NkmngXx2Mb{i9&nx_5(}
zt@P|nZB?>U?zRg}QE#IeDuQiyZZ4hX{ofBpzpC%lZPtCEiu-~`QJZAot@PvKH*ezO
z^s-1<t*+_)UxI&5a8hUuW$SRhQPrU(5Ow2!F!kSV^X$I)jZpnzcjcsSq0sPHba1KI
zTW;==iox2Ad}+N0%lA;c>K5k>Lo7+C;}2)te$=Foo*n%9O_(SD+}mq1IP3i~+3OE=
zUpJ;+w3&9yZM>FfpFN_D{~)1JyEh&CbE^D9(%;EHy^EcyrNbE8!{QOHM%zX*a%+NT
zS1sle0z~mTE;l(@z8~3c@!{unk9l28{mxgg#3#lwES5WF>5<+dNxWpYn^(g!qv7>Z
zyfjMY-H?{*e9|+ElhmmL6Ye2jgQmzsHRmU5Y@WE<<7&g|;gM5im(AbtuddcvAEvtM
z?7YGyWix$j@9~0oe?OA8SZw#HOCXEZ`AI9|fFEB1e-5#Kew9@ExSi^WhNl+MEL(K?
zS?mEp2E%G}ozF=s!=A>}?ZY}_Zm+8Z>JNv_uOEq;8-xVYJ-K429vu5hZ^z`qIBn<H
zCSieN?|q}_qmiQfjswl8A?vG8q>8=}S{56~lpf-pc(FvZnq6u@J(gP1*?1T6So`V+
z?UHsw!(Yem%>{f@2P)iZX?J6SP`i~LB9T=L3rr|mk_0(x5+=K`goHP;tu+?vQ5^+`
z_a2Qu5?V^KdgZJgTqDRhpk3L??-*r1PZ-$YlV)mx6TAAU>!hkOwahDIS=laKyIMW!
zJu6|P@1r;^(&mAIR(ndG-I}PpjAc#b#bpiYtnuNQ;M_&e%t)7v9DBSnKS}co*^bjr
zDJx4q`$L;(YnAAEqS<6e_unNy|GZr1k^hDEb=lQ~vda7g95*CwMQQk@C`W;cXS%mN
zxjKa((8yZ<%9ZO#&_6lCtB@EF$uLMae>zmLn88CJ+otVKY9^17_SBWTT~_{bdDBgH
zYb*C9cki(XTl|#X`6#{MZ=1Vawk2N#znr@lS<?Y3YLa;hE3}<6ZsBu|@6>y*Q<sg7
zR!7FTh<J|=>o8X@RI60)Q(EkJJVvI6Git9bIw%PEQaGbuUe1VVwQ3I@n{;w6?ko>g
zB&5$Bs}9DEl+JC?8qu<Np|h3fw!S02irT7PWte&zs~^ev8@&}M{s?h<|C07FXWP83
zX<0}6FRhu|$Tu>oKZ0i8aO#i0tmnSmpJv`&;WT>We1#@IW?bppTC}qxMWKD|$Z$&P
zgm0*Z$>61>_la{p$p;<Z2|C+*^Xl0U{F_X@S}*wy8MJ+Gw^@wN$TMl}MyPPFki`3@
zA6Eo7PP`oHFU^h(VD8Y9C2SWjtt!>46=(fMb3ILV7m@Icb@0bu2Me>oA37ReI`Tg7
zb`Ol}*7v?7FQ~%|zealXpR(@BADQ>Aq_Y<83U3ysN(_44H{rqaq+t$On!jU@%DTsy
ze6dh!U12j2&swka*2BVc9_L$<FRlc`xU*81<=b!80YmIO1;eeKDG6n^ca{vygAdvs
zl|>w%NyiN3{AhM~;rcgoxlf<5;!&z^xtq<C_gR|p$#-g$Z{3RF?w+N1_9Oe7z(mL$
zLW=Isv*e`h!s953uXgh)w|%2CU0<SP^BF#<GmMRUB0N}1gcW?3HVlP6Mb2Imm2&*$
zLATrcIK7sUpR1f(stiwx@Ue6i)3`{F!nQn>_jl!p`PzNQ^DS=(CUs+@h?rV@wdTc(
z-uXV~c9ZLfzfc_OdP|?_)34FtY<JFC&ac)i{A=95I;S?P41URatyy|khfTgt66WHq
zoYQ-Dv4-N*y_EdY75{j*UgAw=<V0%0IkN!kuWzbu1pgswjsKeT@r+>mMo`ej@c+k<
z&hvkVbS*b~b8a0~6B~0=CvHt=6DRDy<)0c`nsYz2HiKX<$iu@eXKrn2<wVC1f0eR!
za#S;ae$Vcyz1=hOXHIm&-1qEk?VfAe8=IPQOPjj@eO^OW>i;|{_yh%Ri~N5%D!?DW
zl}vm5Z9kop&t0huj27{V9Jf!}1r2g`AGQ>_6o%$dmd7%eq%+<UYGf#E5#>2aMH)z5
zV|qnI7k?-bpS~Ab##|OZI@!ATnfq2@=lXhiqS%pJcfwxMyqOdZy)O~=kBpaK5vmIH
z0dS8%utgkHHVC@dL`8oAFA6f!FW@>3Y9(Y_P&8D*2ndWQI1kPQ)845BMFC)hE!Jjx
zGcYnbRDrI}n^+&Bs(I+T!9e!`*l{0Is;YKF2@-TofFm)fH}&2-KwsAAFIql?L=nhr
zifgC8){1ogLUL?fD(mV300e*#yq-^BE(??v_#Ydj-zT8KI;Qs39bwfmy-LAi6g!^^
zdfj9sBx&!qE<9WWP$gts#>B*gg&cY--J6d=RpK(`A&j&SoK813G}wCroYd&!NA%DE
zvh!E3&YgZYbYIDY`dT>&eFDWc_?#%iTOgr}M|E4l>;wp5eW%7+`@&?{H)>C?28YD?
z;P4PD;*CM$BpN?~i3TChs}I`Zg^EbgI-r%!!VcU>Rwz((X4lqg9G6svy=Xrj3rD~k
z?+joF&H_I{Wq6Cra0EuQYhE}2uAn~xc5PN(+i)A3vHD)o=EU@cX>f;+sVi*;Iq2yn
z>#(z=Qy|slj!uWR({#}NP}Ec1otWY3$!-jox?)b(o{W|<Q^bH8>xP(>0#ptqnjX!+
zzCK=&l+srLU7_vU()i|{6<`Dg>GCOH4h+5tz%K(6643FYxpBj;`nj3eJlw!+<@mQc
zwvdHBND>6$5A%AT%WU6v&H(oSWQL401#MVnK~T8|10*+u?Y}u)?w=#usLmCije5Ba
zev^=qHyMCE%j;<=df@~ZeEBpxhc#M_X=!erah8_8)NqSFaqjFnNNB-|nwEi#ohVVh
zy@Y!W)ikIF{{rAa7yt_RJtIr?^WO2huU6`zqNk-ig)U&SkX=$Vg(3}Ch%wle0-g{W
z5+eBWa*~912<O|ixB2g)Gs{R5z#$K+3E?ZY($yDcA901?6{V-8ZHqE7T^hdef^3yT
z$91ORazDoH*M4qh)qpF}Ph#<qCxOE-fdn}|h0BYxzn(j{>T=;h&rsQqf=s*dugq!w
zogMhR_|2<Jeg%LM>Xm8b*LRzNcnyfHoq;5jJ+Ix*I<A<XU)&sF0SOO)q=MqG?2f1L
z@7ZtIIpqG0JNtYj5t2FJ*5>s#*M;%~?qJCn=(xRW=pP@~b0>QfI#+ema1VCC!R~H^
zMB(n<-o4SZ5xTZP(6P9uvnx;{a3XrFK3BF&K24a5&wCZTvs`n(KZ-zLp{^6-fg(f0
zsTY^Ow={sLJi9n{wcMmcU?N%?ePM2u=!(!|18_uCMER;=hb_TBv~5^6C#3BaEvhM#
zNG~jQDu)#+aT%|=Px^pOJuJVH(y#}{;p|BwUC9!LbzM+Mg)HPG_>mG%=f@i6gKWkJ
z18kU><(iOl6?0zIqF1N~j1!OI=6exG3-#jm6L97&zL2K)8;uicgv;-{B{Mb&nut3O
zU<`Eu*&+#hocys#hxiRp3{!#1e*T7ZBBI6m+D`3$pW^PI{U4dj>mZ^aSSyZI!SZ4p
zNjLc|eEH;i32n>c`%-*3P^mU=%|<^o0Hn++4+R}!${j`N6vR<#TmT9ARj`kyYhaLs
zl_ncRRRYG%WG`4%67D+0LZu54s`cJ=Ed@^PJl}3Xn*kh2W`#+Zi!v<(^dNFb(*89$
z9_1|>GBPKCF=<c=FUjaYC$ZFeNmWD`rCwh-C3Fwk06Vi#C+0+^M|=SVrb|vb!mis!
z@o2wuCm?~?k__a7(H(9aQ6kUj%Xr%Q=dDkGG;q_6=yGWeK8Tj!Jwz^KTa7oJxwsO=
zl3i7&nzd0(!s>TX%z6)a5Y;>d+yMfv4ukjZo5t^y1nI_QekS%z5DagsrJTDZ?!)p1
z;c$NDg74XJAu)vIB-;F$XW?Yvo2d6u5Fn!!yl>4Bh`@|j4I;wARD)NfpX!jdr*bum
zKvSK%y&E(-D3B(GFgo3k_;#}3aSeq-qz<*gH&4Idxm(TI_N6-1iXUBx=*<9dvM;^i
zxlgP@N=dqNtul0wXh%Mu7VS!8cpiPns)x?w04L4ia&Zml{LDRGSJz656MV|?I2piI
zKcZ5IX0B_&LQ~&*17b+ki|+nWn~_BPdwMt&yGz~NdEsn-#;WW><F+9@cDk#nS28g(
zOPcB6Jkl9$29=W%70_qQU{1qtLQdG0<>la}OUv-&`28&=_OORK{eerB!`yxL44?Nf
z{0)H&-AiY&D*4f$<i4(7AO3<j;89O^jY=uH5ybI+7ej)BqYaIlvUNgcwQ3Q>6RBf#
zs~53Cgymz<=`XnyP2jf2_tCfQO5sI3)e@|zo6b|1Cf|L0mx!%H-B81|81&tCW6a35
z@bY!t#o0E`YlY1ZONcD+-lCWl{ejdFrXAK_c>J^#BYat=z2K4k0ol;)gK1FAAk$_k
zxmL(@%0`Gf`}%7&Ux&b1MDcg&z6a3*a;}YCg+)N_i;1ctwW^3wPCe#KOmnZ&J^>(+
z7MOTxH7)Ir!e%|CfvY5O;2&U|===rfU1wC>t88j&D#uSjQF_LH{s*z*jn^y1SUXJe
zcJ*tfSQq<(Wa1DSsdumDH=IFmCh3$ucH0OrN3vv@SC|XZHmpu@|3b&92yku77G?lV
z);yrQ(2ZvI+bk8nDyR0>Gj;a$0uD9V4FL6oW4QGym3--ir{UB^d;irhiL<&0J2zwk
zgX6ofMmJyI0L$Jm6QNur@$D$EYCbUzERJu{8tbA*K(i!T4+kH=DY|@K^2g<`loNyC
z(5dB7400_xn&?+>5G5?PJX!c$3Fe6|0-S1E=wh1JUnVew)j_^y{A&3UKy!yS$5`c4
z&RO;*|FjvoT{<sJfNH8%(=HY{RfIFY?qu=0PWR8BhPQ?oWCG;wNS(jH`^b7~eX>SK
ze0NCCj^Hmhut+Y)Jj~UQ>7-6&`U!+*z+2Q!0$EL9+B1?tZKfVhISz5=q(J$d^nNs(
zARs@M&Fq2SmjH~+T$SW~c7a&bU|N-IE}*59VY@{@YyHA7b-RorZ648V{ta6n93JLi
zJS4|kwpkw|`k>rCOxC1(J!oU?Q0TM#^?g!m*&Xu4s#Z)t7y%*4xH3u5xWN}Bcp5x+
zS+iZawBM1T@CgW-lL2oMtW`*>DYeoB3bb|JZpFN$Lq#*Gk_gyyeg0Cf>DnV18y*(a
z0gPwnhDKmOw~8HOwHk(k0s~121p4fh$r}8Q)uuwAGgLzcjS&OTj4SB^ERah|krRP!
z`WPkBh&-R?Mc|vPQ9lKr)%`a*AXhel$(mznx<{QG-eZj-Ln^74$BCJbsCm{{J#V8^
zSf|YP0Oc`jdIssJ4Y!+gUF!Ua#>$HM@{a@GAY>>_aOwRZfym!a8BHbdO9EXUF6TX>
z$9J$=w+Dt*%t)-zRe(Tdi~T3N)c4Tq@W1n5fI34^9)39GUHg+24u4+RS?U*XJZx;!
zt;XnhL2lIi{4{I+K<Ix_Ll}rh1%5x~QPLkXbd(3q;P|=4J)w59efTv^bE7OfO8UId
zg-7n+jRU9#g+a-Mpj(b#N8aM$N<DObW>m?<0|tq7ZPp~i&n>}>iAzG<s25+_J#j*L
zJR7y6Z|1V!o|$W=`Nm2h%6;e1?@#S0^-q%%`YJBI3$Fgahz8Zn=e_v~y`~hm%4X{!
z{bi4n>?j58tdBzwb?>14ioV&Rv1#2(o6eEZ7&jw5HO<cN2-GPq^4G~p>r01M03;iW
zkTG(5!=hj{L=gK<U{XL`a_G00Uw{)x4Dh4VA!4?2_Hl@Pf`;U&znmnow4X0m1mexw
zLib%}qZ=O+@(?nX>TlZnW+exNno_9P;}dvdzjSA_vz@>xu(`4EXS)8PG8%gp-FyVX
zZOW*osWS8S2U$wg#L0rsCPuQZrcGf$@34Lsc3&&yY=(4y4s5%zfWr8!oj97_A2*%j
zd9Qk1wivGk#ZaS44z_3U(3Q&HF}jAmB45Z}<Fee}u1i0pSaAy&Fw`=LNa4?csUV75
zI>mQyt_&9?f0-s<^ZQHw9I5hwPX<4ZhYH^!6wl=c?5>!bk6<sE|0;~07#NubOjNlh
z6%=sq4QyvJiExOyuJkeYTCs#uHtn>F;n;tJ^EJTuIL*Y)MgMC929ED<HlGx-8mCI0
zaF&!YBYnUg*dIelDZNkG?4?}wmbp)cwB8k(4+Vy9(WnGK=+R#4E0V8t!>DA^sQgn3
z5Hd{}%QT$9&1LhDT>cKk_LH(gA`O{$xY7+QSJc3oF!z(m1zwVDs2p|)6+svGn-l=n
z)~U|Otet=P8+Z@F1o*ZiAs(L8^xJJqP`zVl#rJxq@q&1!9xkX+nXMiYrqF9ye#JzW
z<Yks2*DkpvdN=$k#s~wNYl7m@?3yf}5$a$3Hrb$er3)5$u;C9kCj&`vEuw}#NO+5X
zd1ZQCz=q80FC6PKbQnXikWPhMhAW|_hCI9QD=Oh-5awjOLKb=lswDOmrA$QRndftC
zw_d9Onk{d+7)=`F+0krCRKS0vh)EE560=XU0fY$>ZHEN9m^s%DcOnpP!%>86&5vu^
z2Xf|XGly^hTqWnoz$wvoCz2^gGx?^=%<&AM0T5>d#i~0l_~F;m9&3?gFn;zv6PKdu
z)%APfKn0WFIjhra-!qE`F>tt$R5gJy1pJ6UCmlnlmKfU=?sL)xS&0uTc2jb^QV228
zpeXrj`6Eck`u&voWZ8_#8SzV&hq}ul+yAq!ePYR4tbx0H-5(IE1`%2Msg_tjJx@sD
zK;S9dy>j|3J{%)>F!HnlV>|Yq(vzxrcW3Jwg5M!$D=D2t+@vsCV((+x<>aoMk5Ol;
zAxfEu2`!J<ZB23KR;^P9D{7hkFb1)!f`o5U*}`6`kukFIlAV~5{ZUtE2EVGD*>6OX
zI;|HFj373t&3i@K>GbZ2h3GS>U<AD<uek23Qk?q__9KadjfB0}V+r#pOhjhyL?!ph
zltrjs8S^=B=sI(`xgNw@27j5-Dx#GBzSw4|yR80DvK}ktmd?lcvkv5yRAf=j)9+_I
zo2xGVZpTPRqm)#rwcZQa8I+xveat>b?#Zn#R(w$SU`ui*<}~JeCuTL$K7kw=fCmkE
zTC&kd`+P;i+H(T;!;>Y&fBT~igUj#8h<^a30_a-Z|0}6=bpf+H=*q(dX9WIbh1=3T
zIJi5U_g;aRE>|;N1R_%N?J^8%k&!<fuv3mdFOymVaZQ?rD2B}oTKI)>Us+zs7PGhn
z%w7{PC^NW6?g=&c$RxD&2k?C;oUMOQuFYvesDTl<z$kf*TP6t=E1Eo3wC&`zEJ@;?
z0R=O{sTSc2N*^u8&OPCwve$4`<b8hQI@yizJOZ5!&{oN!B>D}TVt9g#PT%@V5;>h_
z*%HI8<_fZT<sODamf5S)Zhitc7g8FD3w25^<j<_#m&>tIZho3F^6|KCeu)inC;Eo}
zaliXDtZ&G-<^bYZo4W#6h98K3LKYRQ6H>ZC+y9fEeZHUZ<ypn3F?4gJ_=?D<*|(1>
zY{(1yBA6Ao1J%%XW9u$AnfuMz3t>sTSufJ9+P>K}?Z0gDzo}b}Gu}dL<$g+78{|u*
zh+=BOP2qc`w+X|ELyh%*(zjssu|p`_r{litzy?EsIQAhlA`D|B(AC)*o0I&8wJQ&|
z`Nltv{P!;UF$Rq#TqV*qk!0wG`eAZ`Eg;&~lvdR)=XV{<xelS^gDz)CtdDS1P7d3I
zHB_$^+?lTyX~aETL9ZR?3lGqIi{?!Kp)~)Xn*YE4*p#-ly~%xn-Zu=t9ny!+z`KAG
zp|Ia%SrQ8w{lEQH`W;YYdol2Ss?kgZ4UGsCnoR;9ezFM{<?H_UZ!7>3O)V`oJs2#2
zRCOQzhc(Lg5{#Bcg@w`mRZ>#&fC@}Uhf3U!vMaP}gh>(m4&B#(=_EgOuV>?5xtMmF
zm`Ncm1A$BL`{^AbRpiSP!C6fYyfnyEa;=0t|DJxnhmm#<Ba-qXEx(XR-&S_-K0Go8
zxuD=+XX!9GB#9J`L-~WGtg!pkl0ItmL%tNo+Ss?tRd3TZeA7aZk1?fZWH7m^<<o^@
zuXdIW8BmxCkXA@!I2ls_tN%V7z-NiBJiNSEC7`Uw_h^|I+2wrQjPFPP&cuE#;RcM~
z`5DH8Kj?~RVxK0sVm~_dah6AL7nrmQ?uu2mHB0jO_$7<Id>eeP4YGcL>r8#vTgJ}>
zj&6d2Lv?I@sdUohlCvODLU08GC!_2gu7~H~W0%hL0?T@oQbu%xPKHIO*Cv1xhy3yL
zf@$+LQ+X_+$0)sO8_s)w2^EF87~<zg^kHINl0i%kZekXWqC#R~9hz-J_maPfQkr9*
zXFq)CEQ(FTGFubL;Um2wpB7giioHBD9i{71lix7A!Vmv4sgM7=5n=m;14eP8<B8}T
zp1Mj$DeOazDD)2w7SIc-$ww?mrm_6*MX!>G?B)$IHde0xUPvRg4Y1NsPE1Tivl+J}
zs9`^0`4ZVAhjtE&v&;AWlrKqi7Bm!+Y0*z);PJ@)+D&O27^R-TSYV{fI)7l_p23J@
z_^!Tk)xAd}jBU|3qR$gn(4q1GTp0BbhOKZgyDJN!NkAf#zXBb8{pc|*PeoyR-yRHG
z`BS_b%aYMl_-{el1VPZ34Nr(^BQtIJn$43#Yj-u~T^Dt06}F{}-&xrIGpM1SBpQ!$
zDn~5DPU5cA0i1?YV!#nFFnUNHy>Fz8W|LJo3O=h)K2nCS>KCoiyD4&MMe?aw4i1Tc
zM+A0x`DJRT^S6l@a_*-n@XP=AzEO={ja6HBH3?J{!rvwP-)A+IYn4a9XAPtx7DnPW
z<FSkF4Gt&?=V+_qyZ4-+qSQY8_o_e0RzXm(;VZc*w)`BaAopRS<I2BvC?~T6gSTp0
zHaZ-A829azQ95<BJa%tL(`9xIRcd7*4i+W*>3DD6Tn8j(XsMgQm+0s|_6j@bP*LB6
zZka#t&+{h3-e9uavXy%LWyzK6EmwKv48>XD)lZxjnwI?->QPL<JrYTV2li6ccCdZ;
za#kfgj(f_Vu!+eb<EA*qES@v^62+P_I}5@Kx{^to@sbSmOo=99&XtO2Hqkvr%+ZSE
zq<a{=`Y%A_Zgq#*tlBQURC9e~Pzdy%&w2Wph>iS5@~R{iiKGEUQUkAcHO;S33FMX~
zu1l5~goK1pw5lc;7WQGRUt@nuH%d!{zvBa{eBOLTvyUGSX4appd{#i(yg8Lmd;8%R
zEO<g>zz#9TMb2!uo`Bd+FNL`K^{ERSVmTtaPCwUm@8LMqA$Eqa5~C7#oqQ0Q2c^!`
zA1r>r9b1<?IZ}ZKnXoPyE;|0EtTfI>CP=S*(>$85R6O{CSdI|(DMyO2Wp=q3=jI5X
zi#Dj7E3UvJ!7dgDB(hD@a;*uoTAM&YEicL78^~$&<gF4m#7um{z(Rt@q--#xaR1ZI
zN|p4x@T=si4^u`5MsFRZ6c`NY-`}O{{l$;>Z`B(jbrljf8Z>s)@haU%Q~64#f|FsH
zggLEBa^{vj*5+tvPfbmAaUB)mi&z|sWR*z8u3<5EwWy_P4)Qm#uW6H=>w+>qAA;h#
zKHv9X>=LU1lS{Yb?bJdA8Eb_=C!sBp!VEdL)p`66*pD2M1>XoKr}DRM-}?+W;7YRB
zP7sMAaMiHY4#~$Ll&&tt+e_u<$R%!|_O?ZhuU*Donypk)nx+JbpEhgBYp<74&Plr3
z4*U8*x?Exb;@@R>Vb{Ki3R`$9S1<?mi%NUG&ZW9B^Bk%Bp|2rWj7>I<v~APrC-!UP
z)Y}HA47>NZudd_IO1eegy_R98iARh5jT+4w0V<U4JcAG-Ix!>xBB>hx+l<a1GnSpA
z5rY|BUBxK=>rA@;Z9YCSdG;K`QTG6qSXVW3*o8f|VOIa`H#4?b4YjhYxPeiX!xUvr
z-5$&XA8G3_@mnl#xei2ZdRT}F%sN5HT@Vs)IHw^a0|(M^NQd-Eu7DAmVl6^DtIH(|
z@iei|2c_lsaQayg+~A-fCkqxK__|eqGT#As^=NNzZ+jc`Be}G4)PNKQO8i*V^k!?k
z(AKb_H!TXa*>SHLFxGbNuOIs`I+Ere*7=cRyqTB9a!81g3M+LxU;F<>>`%-f5#I&_
z>;UC~y{G-@vBSynv*QJjfd(A?HHG8+4-j?7&!!!OHcN@704gebdB7hr_!2hDIA-_L
zxb?rNHbElo0~8)r=L%r?5z+|vETBAgf8M+anqYzg<D&H#LAMVmr5eP79QLCb|2r&(
z-R*}bnTkHYz0t*_uu$A~i(h3gg#mW`vKjUO>FmcIDW8Ofiwt2D=cjTkRAzUCg>^ur
zn>1#gw?)m}{7WWTV&Opwtw9!oc3x5@j7?X$A3}|ooZrhbD2zet;KQvgpwGci9!%H!
zX3W20zLLXpy!bQMg4Q7mv2n;JwtNZMhLH~6ylxodi9OrozV90W_?38K))Rn`Ec$aF
z`FHDm;w?vCL_r3e3eF3XA{y0p@r%A`2waDU&@mB*M+r_K<hS%)hrB^8%mcDHU_PZI
zkm5jVn;T@^o>t^s5BkD?lu~b0?Cfz#EC+#9_;LGlX58UB$POQ~Yv)D1yZoiL7AC)+
z$(CHalhG`kg$h!pr)U3sSD}B||K%Z>?3nvDhyx2N58J#!4dBUC+m#*R!e!Px2Wbgl
zsQrN_4LqtEmW%BhRH^7^6T(n-!k7?9b|0;e_B&mLW>h=lMS9CT>;qhP<T;O`+07Oc
z?eD>iXqlO1yv<|~Th>M9;bNoCL<HPBEa8(IdJ+AQ)R)`Y_om6fb7%<!bp`1-Sh?CC
ziO9*{@Z4TB?1-_BK&a!@$3OqDsk|bf+gF&SUVvu%^6_b|9CEb<lO@S;!TY^>zM#F7
zUt!MUnnNI7MsgiR%3lS}9N)$h4z$%DOaq2(v$!;^M^Q<^%fTA0<=SzW)m#2j?ceG7
z5>`VKxchkgwx*%Ghtqo<V69xJ-R<qSv6x}C)5_c;5qr@F8!ND`Y(!azZD4c!_;5Qr
zIoS|YK7b<H>CQt(mjXq3PH=pTh$oUd{yqX#F=}PK|8Kp$hg8CIj$yKj$~*NvoEH%&
z;OnugrM!Bly-p>8^QipEFk#$n&#UHsa%R_viC;s8o@-j&1;VUQeY=ApaM_6rrZHGK
zy*Z|3oFlc|l?d_CiXt6xI}1U6^w~9nO%z#%h7$~m{Ya>VgQ$4ZzyR=0$Lr;ep-)&t
z2N=GS7dTl0)sX4Y<z=gW*B*tWji!uwsrDe{blG&<{;)|De>|MC83^VKx?ad6aOnwW
zp;C?uij~F1dWtZ#X+}Q#1uZm?{m*)mCHk`|_8okoiy<Q{0>9B#?EjsmqEbuUxH34s
z$ctC-w~vkt0kiXC2Kq8%44|1_z*Kg9{pKqW+Q8leASP^c6`Du!&zQ84D1UyzU<v~Z
zz5X0pR5>365;<hn7|(3ENGcBfFRZrmhPhILu=>rd-t-oLqqj|9&p|f?x+OGB!X`~P
zWzE)kDCyT4wx3I^_wV`=m<JZ1%BODkFr&y`zoi;}TicIuKTM&qR>>*P{>|cRc4?ra
z@y>EiiFNr;`s-h^rZ6@dmUn()o^uCUlYzmHzBU}*hchYIkEiazux`R5y%>9qM4OTc
z4rM;)%G+XAt?kEA{J}nv?b_}<8*+|PqfwjjL{dnFFXas*{MD>VPkS@IV^%dS?{s1w
zeEE1B<=M~HgDKoiLCZyP+)itbe{+t0Kjj*dP|U*i+6MW@%5SQJ7zY0sMFkl}-I7gn
z#bgb$A5PMcCWu~OkGx(d9&^!Gm{dy^=!;?iI0LmY4Va|Rb4Cf+3_vqNAVIaB7=`c8
zD;T_}XUmhLxMDnsF;TesFgn1Z!iS-^`q{t_IXn_BWjMxR-#`jv4{-!!p&zxOB5zx&
zfW^hmSa&x!8=G7e>Jqt^kYMn$>ViC{(pkxqC<@V?lxy7MOpI>3q<X3`uuvd60jz!D
z!UaH5#lXP}qKiVK*A_J2VQ3J19KOAUSB44@ElCQ#h&x;A<cseQQqazoUg(^@_QO^=
z{yP6>r#$JN)O-*U{9eAY=LUb3;-t(ASh_ms%k~u@io_Lm6HI@6{<f{v+;bkXehtet
zY^WF6;ERxuFSq`8n9rhupvMm*01qBeiMcz$z^l`mq?%)GiM7+h!2ssk8-%i^Wz7Vp
z(*4mr>ei((&D;Od0|z8lRRx%mu!7!uQWEApDq94R))Y&m3NGBi%Bk?`JM-90F<yZx
zZ)_)wATErEI&IQ0Q-B<30t52auo;#>nI6~(u#T(kXQa%`n95y2s`-ZZ>CWIo!_Fx7
ze%QhX>-F2&ty7K!b$j?ojVFVV=wpbQ<wuIcm0DjwtcTXZp*-lL(!bnl;%!e)PZjKx
z^$tv?LNu0uyoYRMLHcCH!)9~5#hka35<M_l$U*Lao|&0BlnL)53-kEzHdtWk-*8q#
z6$btm+d1m#AbR5rv$hZatUf)M_G$l~)qCaooaVta)Ewn;c(VB#>aHTRokH|{W>OQk
zTEpbw!uT#!x^LUvdtP~R1<p;qP%a8hwt7g%Y*$HEd^OZpGe4*-J&wpy|8~~(U%Enr
zusH_pacF&v18af^Ws`$1`;zF(*Y@|OPfxA)l+z*7LAKi`xMp!pQATEUeP8@6r9VqO
z1v08Q496c|T+#qC_B2=F_s~;^g){ltk@c~w{oGU2kc!>>+0G<7Jj*Df!%o>PKaE~~
z=gK>sjcYyDWKiDwRZhQ{gnl+KHvq1I&k$u}vw|}UOMXVXJ%_*yQ?(RSRLdUZV>9)~
zP5;j3?UCBEC@ltu0M`OiWSny1M$BIC9g}}IFpUc!^Ea2k8O`W$5^@djE-8nt_Unl5
zc()j0${yy-t~#6j$C=Y-1edb)ObjXtNNw1(F^voExi^q63O@3{1jMkxORmSR+c0-;
z&A-LaxEG~CVVp8s1_^URJ9Qg|#9HZJl~uj~lyO3Q{Bpji6rfgMLo_AoQuI*9yiQGw
zk;|bAB$NZ~)cqI|4HltzA1p|a#x-2{7^It_BjPM4=QmsM<HrxcY+PTy?0(P7SnF!P
zDl&JsK}bmac08(I8`WgP@HB2wSAH0qUS|E5Lt1sgvMn=fot>F^{jC@}xPE`(u+fA*
z4mUUV&Gn&zc+b7vPfZwN)mhIa>Mu*zq>`CdnlPW-aa3rdD#xCpt*$ZYD*U?untC`-
zVDe$_`*pi0pnVwC`!Fj+!3pdFEfOROy&N%D(v4wocKMz7eTvUib%g<?l?-PLe3`hc
ze>ElbdFV=%s%LafQP`|=X?nPMm_!7MvAN5MYH@CENV<iw5%je%M_*~^foSEf_R2?H
zA(7w@;2<+YcjRyFrxajwV2=7XwJ)}k4n$J1i=qH*;XJ|g0*;243(*fj_GTspx;{s%
zgP_YZDSDK$?32cZWK9s1U`*{tTeU<{K;4_+qMmL3JRL&XOPIBqUR&d|c;xdJLdqZz
zP$0=WIX*Tr2GWXvYH(&Ww35jB8v|YA5=4GXlbU_{^K*)?YbPZ)c12RwRV@TPSvB|X
zVO7mTOol14Rq#w$iQ4Do<%J21w*)|7$-0Mx!YbyMvCG8Mb*{b6WZbU(6lU&zOvqu%
zDW|<Y7WSEq0TMta+|maDn{eDglWNU<&nR#28!CwBFs8L?WSf6V9tBmNdyfc8Yl6un
zO#JV#nF<>=i8pKT#RZpv^3sN4<3>oeGC7fHNkv&LMimup@~gB<lGX}WS)aT?X<?!^
zuaifYU&0pW&syO`28UZLq2zE9at3<c_3^4AFErKl>-nWr`tAbi9Zga5M##2Wn&kI1
zdlSd6`ixBV<&dIXnCBTzKV3R-0S=jsflUc<2#E}Z72yHX0KjzLgC&Qhtpo7}>@hgO
z!w!l8ysKCH{SlhV)yxPMj@@`(rU`q_UI&zKY;u$Mq?^!6*+pFLGt?HTS;BAJGgyah
zg#j%MV#av^D%dNI3bTB!De^B9ukrCQhdC(-ml_?7zZ<S{d%#>rS8U27B!`9F!`68s
zQ#qX{o?kF%*R%Tc6?zBm0zb;lMhOUS=s^4v%IjimTpzbzo^B0)c&KcMih6ii``_Yp
zcnM^&P#Td^IAf=$r(qU;EISCoLN{=(U|I++qJ%%c3u+}|S8N{>Lb?qa#I<wY*FI)e
zosCRodT%9b6;X+jcKgIP@AV=2?0w<L?lEx-DWNQj*x8Fc2me;t^oX}NRN2bcZzu|m
z(s=JVkYJ1TI0BN+F6_OO8kk;M8isxrcBl&wSkR}<lA&m*1~R~l)dA}!KVoRiOx3(7
zaz>`<YkWxSziFc6xk0dV&3(Ifuc@#m$vtbNprh@0VVH!J>qa6yuiN)GO&;u`U9~;M
zzwR@L<>+CnAYK;Y7b!Uyz{kw)s*wS7GqO*7=~<uU#fuk}Evu6SpXvhlrD+X7<y@ej
zJwW?u;MAMys?3zok9zy}N_tNy^fh5v(`OBJQSS&WK3Gz{;GkD|9Ol0D=y^QtclzMs
zrJP8Ma(U!x8vtZYklx>6A!PB*c*3HZv_3PVp8tSLGcgk{uLpW*5P!9A+70FnzWC|M
zdeYU>(lVf|V*u$l%)P*g_ASdTwu6s(J~<@aTS7h`L8%|qF51mjACNzHhVN<e#A#bo
zF3t7X$ZcWC87X(QDpXYRE>ylBVPWDpNPyiNs0DTd5yPS_s&(iiU)W@RQJqcPd>$N=
zFd9i&lq%vnRL<E2$P>iJzI0in-qH>Jt5TI3oWaNyFZ990Go+NBA+5))q<zmX5;N9c
z5(_||psZ8qSw?%3uP>83kC>a<-{5#GA~gGbn$;Wn!v||vM?oGQEcq<Uj=mZn509iM
zJw1J<9_o!|QP>siI>kn%8++TjCXHOCa_^r=PHcSLXyx1;2|MkgFr|9W{$qRXjdZ_8
zFU#Q@)&+edE#o84Gdn4F9#1zlDRI^)07glr_OVCcRWi17u^&94!Bz<kDbk{Da7}@t
zz5kH8W+}a@p=LZdbZKhT#&Ph^&|Fi%*XzvhUxXX0>O>r9s595k$d78KH57#~^<Y-U
zFaf5HSnSLX^XeIBhr!tnwn(!uY6D%ijv*AE57?Srx^!uJZubovDLBK@E4+oJIFzG)
z?~@%cQsCg^ne<~sy9x=icT1&rWhGY4lyuc^`(99-mHyhp85&{BL0(YC$lv-RN&O9J
zH%lL8b4C(3d{Zw--YiVc@2}+K0RRgf(VVd+G__S#M|ZZqTx8DuNruN}X!sMBztBl6
z@NrB};QT3;gcf)NRRPc}AmsmEFYAUPXb(<S(H4x+N~UF!BmM^r{_h6uXvN!IEmyT`
z>)6E?KmW;0WA>_vTNW4=-uGHKkdD@sJkP1>C52q=#{?*(>@%rHI{ow|>Qp4}C0K<1
zN1})$3qw6@cE&8dQZsoV{Qx^C8F(Q;y33!?m6B=*IeQLO{d^QJnIC_q_>$TTbdsPC
z3RAC7%f=nT5b9op!nW%?b5%X|>q!+CHYD|*ze_v|3g3FcMlY^jrlDRRYDwBm6qG2G
zl=JxuWHS+bX0v?auG_Cr1D_G(k_^dABxyoJMTViDxP^sB)@$HA0Sed2e57iEhgQg5
z3cYyhAC);wL&8=3CmjUVvZKDsAE!0a&pY1axmg#noT8yVTXymL+>RlQCG)sa>F3-m
zwf_ss?DyekjiSXCePeuJ9tr1`-lG|?C@cUQ!f{t3^ALL8&wELih}SB{l8W!FC!1Nl
zNMh=&d#~fxDPi4XW9Rqvl&2LplfkD`NbvfLUu~<$U%B=Riunhlm>5rd*Spvsu1G3}
zE&76EO*|wqz)1vx*VkyDOo*-U)Errn0V#H?pQU?EHQPeDYP}qCjfZ5O<o5A%li&Ex
z9Q%kkwl7gYyxuJGK~@}Jba+1y_|^2tR_a9Ll8Lgx=PC{+mZ1R%ZYX*?L}$?D28r}&
zaMA%s+lda0k6(6cV+gC(@r%+7N($<!iQ}8gX)_<H?G?r~c5IJL!%0r}Ft~V=I-f5L
znKOAu<ha2BEK2C{s5?o_bVb^)e+073B(xhEjV`t51P%6HING9aJ=VRsv%3oxV_}e@
zi=a+bSK^HJqA12<0(#nB`(@l@pVVZ{49ZZuJs71x{V4hVYt%qIB)wfrFn9&IE_Uh?
zxXj>crVHb35Z~dMkJ(C59qrKtJh+op-ZOz|WoKkZYhJUCeJrmwXD-0kK%ev}cEA0k
zsqP)g)Od7{OHFFal;uxa@9M2X&5>kmO-3);^Z_O9VP+V<iGyT0yA#2n88#ED=f5gp
z9;1kitP_pcp6xqJX3^Q<HgeEEo6cThM2$Oy|AT@rB2DCNZ9k$EWaK%wcViJr0g`Xx
zpRtazrc0zp=sr{3J{^7`;~jWrW!Ruf@9!t{P+#%%MBgNI+LxsLY=<%7_s-O({G+51
z{{QMLphF1vLd1pD1T((SEI&CpfkC-C09IhvypaDYu;jW8!PeSMXAuX9$8DPahmcy<
z#f*^e=I02j$e^xfDgw4ku}|w9e6rKkDUiEzW*XUIxM|?Ec=;_my}!ebP4>Ss;#nhP
zX=pJqPY(}Bb_O9@fsd#Y#34AZAS@kjZ@WSsk|=SM)7YY%;Up<<%3{_j#&*P4dFQy>
zdvW8>qUs**Q!>8lofjIzwYVl;jWk#B@BLNs(->aZE_x@w^*6sOZf%YmEvN8+_kXn`
zL`9-=hb#RpPGc`{tboV&sMIVxtJ4{YWN3JEsb4(gZlebA_~(~Lg}Jfb;<0Q#71!2g
zL*o@Mv6?yUNcOe-Cgscg=ufl@3y-)ma5hv={!J{ko<dod&?(t&@=T<~|1OObp91Wn
z+KOs`z6<f!={nNniHj3~vsEtJ7h87C<+|+*hOf<m-?ldjcD_DV{&FD9PR}kPD1I{V
zC&QmLY3Pj60FChFB}N*YjsLypZYK9uRmNz<$n?uf)_gx3H0K_P{Q2D-`S7tSjXl#$
zV!@8~@b|B!qNs3-cv@UWRi^doc(T~Hl`j8^55EI=mX_vkU;#rNGA@R;k`aPZRA$xI
zwnf49psIf4e%#Z#dX1`XxPi$ixle~xf-?LU8ZRjR9bKOfYr1q*FE}J528&rU5(+FA
zHGPkttzFVZ>z=lk{Wv#+3MpTDg~b9Y!9w+CW8)k^6TH7SVTtT5v<rbLiqJOQ1!`ZO
zYgyXzFLT*c6BBK2&(&JCYD$sBKlMlI+XNAmb=+3*Kb|_Y3@l>$TW}Cm<hLk3@<-K3
z$HDnD<_<-?Stgysa4_qQP4_F#0jlPNeEg#{62Hu5tK<E>?=vTcsQ4TZb$B^5oxdLC
z>!(KCVpq(JjzS;v+&WQm>I+eu`^p$DPaN`s5sQq13qbPSH-Kb;J5MMO=6J39G62fn
z-}?C#Du<+`BnZ!vof5S#)iRdLBUHbiWTsge@ec+l3-8m@UTB;>3n|E{6!bc5Zu*gZ
zHYy<1Y<d)NKS_?QW;J<`Zdr6K>CBx)aQ$_K(lg2%O!N)pZu8qk)1}8H*(8nHemD2!
zLD_kS*5y07-ZKj|CczPR*@lilf8!Fu&vcN@Z{m54K_Q%{nH>@u8tQ?X0v9d<m=I^a
zJz>2j$u5)ag|><l)l1&%l~CjfSwrfWgz-OoF5{6NyyJJkQRse4Rm;H0-?*pO_v_!p
z)O*U+cD87EV0d58O!<@P<}UU;q9iqzcQakzV3l99yLPel;e)dl&(X!}+oQum)X1#w
z)yI?+e^{TmV}f_SCJgKNVJnQ#dLx*r{|+o{uvj=<?q#Gz`9DBuWncmS&aAGXYy`70
z*O!Fjp&{3P%P7dxx5yKU4^%Y1EL1A5t<*VOelewNQ#P0Uw@{;OHBfQxDpz3lub*Oz
z`bMrV^iJpdyYHDqDBaLEIJ;xLnc4KBid8Pzhlse9P&lG^-DWoRM+QW|u4X1$KRI15
zE{BGdrp@0Z*X#mhc>+Bz-x7#2Bu5e>C7vY&v^(79CYAE*NBf0Iho|)~prG-o*Nr+o
zi_=Iek#`(^JjeLCRw!9|T)1$+y?P7w&yv1^?bq9>%XVk}zjwI=#cgOD(8j9oOVV@W
zy!gDYN5tyO+g42KE-|c~?cU;Z5!2Y8hn7<7dob!*fzZFwg@Mjx%@8KbK;;E$z-Bn$
z^P$pLGU=3X)nKkP)!9yuven;#;=!s=Cwusex>!ne9(?a9YS`MT5New1E()beW_pHH
z7Cob5hjlTjW14gPvP~y+H~+6;)A5fHO@D@}bJ)&91^_I87pfSd%A=rFh*EM^2*h8I
zER#%gjmTM45H+5VmQ(8cmX7;c@k~PM=9imw2SZEgDol_-P|-ZoV!et0UWKQ7kHV_$
zLpQa~C<^Fi1N_c`|A58p#o{rx0n%9&S-tW5tgs+JJ*=~ex~5m0uQ-8@i55F{kn{jB
z@4KSADF@I#+(~I;=UVR%xLR{tM63Qdu8GdGUtH!EwQO<}34Is(`knc1U&0zvn}X18
z*4w5K@{L<a|7H>3$|u`SeClNs=zD{0ow?K@234(%-#t)57=8*yEEwj?9dUKyW+r>^
zrK(0O{)ry3GX27j_k8ZDU<^sbUtzahCZ#GM+beQr1iGmAUS`&A5$VTNs#qK6Kj<K=
zvzI_}Fh~w?0<a0<FMYxf{VIPR&-}=UTRST;E}HfoagE{h&0tka0(jjV+(1@_F&Y6~
zT0(35$h`ehPu&u}<Az!>wyIP>BcLBUCzKS0Ua>Waal20*V@U2T8uyx>e#IZa(fy6_
z;l3IejcJ6!z<Ev5EkGsFRjoF5qbJ~?o|3u=1AYF9b6Jj3B0n)cWxety9dEYK7t7zv
zQIkl~tV_xSsEE1tuyBv(MQb4p&F&)Gp5TNKk_@7Wy{;vnj=sLcOZs3=d56Y862~C{
zv9ZiNYlR$JHIFa%S(|Q9C;4@hk|xM-Wo6Er;Ck3#Xw=Evg9y2{IPVhU7gXAEeue6W
zPhBjiBJvO$Ys|B1$gGm1Ea!q88V^aND8F1eyQ9>6|Gb6XtW4p6_Au$A0irZLAjwNW
z8*k0?xB3ND04{((=e}*E>+Ubyg|XhMwU*&b)AO6!OmLdPcG1M-%eMrJMIsF}c3&}m
zx2Cy$nY!{F9|x0k_D7wPGypx^cIQ_B=ZE9QZsHStfIgW395~qd8yKgGg5(LZeJF^#
z7EsbV(J=-LA=@brYK1y&WI8mrB6u%sE&+7(9u&gjT-A8Ct$gGrJl}QW;r4Rk9A&8_
z>K5vYSWkbcs;b(rhTnrjl}Ho%{=jJLrN7l?J-_>C^@DXPxYDrzc!vBtbZHGQDz9Ca
z`#I6@`i<{$BFsj00s;UI=RkHo@gru$nv*WHk;l^b=fZ+qb}B^x`g%^)`d~!M2}4Xu
zVjOR>{RZ`2_EQyNL7CveqOU?AP~4Kp7nMJ-xHb!jvoSH<`{4xo8vtC}z)-~~IV`@s
zw!6h2@(>lP%xS3zbuG{>ypy&SR&Vr;8DNvL-@GxTuTmyEFeK1Yf}JaOlo;a{YhkG5
ziwNWn4F$;6;l<(BPYK(RVmPW}+0+}>@zmATp>xx6wgA&{Lt;4WIZC}N{E&6su@bC?
zIwh4zMWxa7ChPkjZyKo*MSuF*s<8F?*c?4#WnzAKJUJpN`Ufo|C->F}n2-<9%NrX{
z|I9&8%@KYZYevcqq(W{P?Tn-%hTGOXXhBV1Jg8S-PdVd~KQS>eQ0XC6-x&F^dfndy
zb?>b&@Lb?@gfy_gp!S#0PY+X<X%V!}C;J)Tnx}R2^29q`6N7$z!DqxEex2GLRlCEo
zF;#l|Qan@;4atJe3(d6*QVl2G>b|~aR>(B`N3bri4W+H0&~N(JP%3l;S%Vulk^$%*
z6S`5$h_)Jgsd*3j)5lVA6+_mNX)14dS1qTQfk|_K*{6KW`_{LZURI4LB3z7Z2{xo3
zf%3hE@($T894tm$5RK?h64m=l)lMv#RtP`Yf(c01h}^t$i}&+1pQ+pr;{&)PAE*Xl
zpbP+%<xxzGhUJr4$)~>)<Rlp?bC=iNQz)B0`L*&nPZin59V>VH3A-H4i#95f%88_J
z8a_m{7v{j%uMkQBKtZ5i-VgzR@B#Ejv($ar-W9?03mKW#Cwt>BVbvHyup9&?tg8cr
zpRh2ZC=0m7Fb<awC;iI({W&btqJL9T`HGA#E{Tk9L`;I6D9!gAM<j=h8jz6$+-1d;
z4-+l7liK(Be3uDNzk!)-C+KK^T-SL*@XP(EV*q+OSfBNugftJO?nBS6c^5=rU@e{J
z25F(d1u)$P!N`};GAAZk9l{I&>7o~-vgXJj_T8t~c6H6_h(-fA_H^p*5yyXTnPyd#
zWRK{*CGsqRqaQF$|F8lmiX#x9o?>S${bjBP15BuNpo?2?zu99>D=kXr4+=*O6konO
zaBE0E<PD<zqhJ5iq)SqKEHLDlqn#h)`Uj+K<WX2RttVjIn2OFRJ<9UTOoM%vUo1|q
zer+i%4WTLWrfJ?IFxYxX#aE)=82j@#vF!Z$JD(B1ft?JANYI?*L2lUw8#x?u(gAAo
zp0xeq@5sd7bKKatEg0<Oy_h_%B;iJa)A|FnIGS5pz+db3b0pAz7TRNxsU}x;=x(Oe
z9c2|{rj6E9R$q{%qNb)MB~5SrSOMI-{4{F<u}`ycCBs&cojX~Euc<DiS6Rp6w&F}9
z48D5(LA&h+p~R}OZUi+hiTu_YFvtjQi)w-vmyly)FZ;7J)RR3`$V&3iMBmQ`;#@w<
z_H>acc&0u~Xgbh?30I92TcRrM7HTRh!>4z@&^-O_hL>NCTAT+K(g$eb#out@pBR_@
zDJ??|U22b1(vSG2<$sxaKg7AXMV3~6*jFgxC(My>OAft^+e-^r0xY;m7ytuU<*_q9
z&`UgPg|uS=+&3K=)RJNC!%r&owesA5DmG#rx<fpZpoouyOJ4{SGk3c@c5>&bC8p6S
z<DjktQUwX+HXswcyZv3(H~EHH5tWoa{%l=zE_5t$tM*f@S8WLtW$r+)^@cM1QC^e|
zJIZFQ9`c6idZ-5;;Sta3f@<qx_%+B>TqKZLwMSXe8oV12V1gOptpIsV#YI&dOb34r
zMtL}}Ln<DKOF&S2$PLz?pkp%v!Y%F)o&jVi(yhpY;n||v!?;oQx&8`-zz3xvjojt2
zm9fF;WpUPj(bMWhwH>V>`;)bL=!f6j!w0}9l;Gw&oZibYV0Ey!r)l<EIvhKiXsuNH
z5@?oF@fM0Q?;u;{uyM@FT5y!gJ>_s7&-i%K(Z2AIx+W+o#3EK*%Bda*E}e_PAHWp=
zXb2PEF==~WS@iDh+)9?Edvv3}La*hGr^>ve)Fb($f+ZI7PO%Y>_+HEtyV7<x`f4Mv
zd4Px}=xH<n@GrD4YTh9j85t`)r<iW&VFy6Pl!l*O{)2BC1(*ucyIy#ti>TAn6o=)e
zC?Y7;9z$c7+RknZOdI4Huc$LfhCsm8%u<ryLXjyyWpkz7QB$(j_kEY!CAZ13FA}64
zN2;K7_45Ad_1*~#Y4^2vDjv?h_gHayufjPF08Re8cd{^23;-jYXrR~WPD_{SY-#}_
z6KIB=vdK1bIaD6ut3mT!tdH?ceJ>s8hucd>?(Rn2Oy6(;rs%RCG&NWHv%O&g63dJS
zPK7tnfMRLhz6G*K*U$te_fRC9r1%quqNO{;s>T3c%v)ZR9}i(-&gUff8uTH%DQQPp
zm8f#|yY})iT~53~z3!{FZY)9%M<*PV+p194#dNHwu_P!zgsEm2cW)2Idc~1Jd11Qe
zX5ur!pkD2jI@+-eTREk}xF_4VvmZ%)m$AbcpEbp)%W}(r`~Tp)&s?cf^%2kgB`v0e
z{tSI1e+QFacT|Oa!-$3*!5(_c3>%^0q=eMOdLXA4P=>y6M1dK8e-nlfkBGmc?}Niz
zUeQ8^nR1o|n<v5da6+ZA*B%9Ko8T7==cG`rE0!R}5ABQdQrF6y#)RHQ-FE2Vxlh1J
z(q*ghtc>yXy%pSE5%5qohqDp#TTQOI|Kei*EuVr%qO{?ndX5P^LA=4_xXRDF^DNgA
zGh)U`8%!Sc5v3VoJ465iul>Z>1aQ}gE$@o6=31;M&CzId)wecMWhy;e#Yq1VVkebK
zXa!UT&ddm)`U0_NWdg$yKeoP?Qur(3eP`dwJUms}$|vu5^I&PV+6L*S?tOnMH9?~I
zM@BK4!+GvIf!Ir;h4{AFc0%N;$X%jGqJU>NHZ~qi(B$PC2U&2E<Pw@X@nI)FfZ;;8
zH%&Y{#rEs};Owh|vRvP85owWb=`I23M!E#)kdRVAx<Le_Bt>bYySqUe1OY(_>F!Xv
zK|B|~b7syT-#2r<IWv3q-m_=>^71_QeZ{)gT35I~c%0cx*Yq*CdYn+!6c!HZVCN->
zGaa(03M;nLW^R~9WYOM{Uuf~tl^E&RD}y4iOleI9Z5aKvsHMDS@*C{D1#zYZ>dU^7
z&&KZ`DFqyt{y<*~4A=rQCBTYd^Nb1+VxV9;V6@5JSc0HjgEn(}FyQoLUTaJ8Y}0ue
ze$ti;Nw!v$+IRFd%Zl5jD!0$9;Y^PlcNq_IuXdpZ!{{1daw7)fq@LS0gJPTU-l{3-
zxuf^t&GN(;KzfIfAWRVQju_FTxbslKpG^R?tbaY!T%0OIq$+nMiB%`cqIpk+9Hpjq
z<GV=K2+;?h{sE*944j;A$|G=She25Y=~Syk?-79iDmTiIOT|P*VLDNWk1ej!u{-$e
zEW;<yx<0VZTa`R2cl1y2qu0h!?H>mg7JLtuzp{s*4d<hR@TUQUT^m%<`q)_wsw>Pk
zY?NagE`P!vPfAu6>F5jdOmoQ<(wd&FM#E?W8x76DXI=`ZzHhc}DGq&ki?Kkl(%(9I
zk7xm?H83EWVS-AA9kyg-#o5a4$z}9pVmC#c$?esa%N65`I70L|N!oEzA*cVyZHue-
zX}SkVL_KclBf#GqZt-$gc2p@P0ISGP0Pw--dUi&QK{Q-#grCezxto))HJ!S!(VfSh
zN?NL@uOi#2WX&_a1y2P+3xqX&Gn{27;2pX{(ROWOZe|wwtUYh|i5}@%dxm5HyQc&E
z+Te_bw@oVSVhMD#QV$&2M`m)RVVvKz<{9QM3af2pvAtC_21h<eCC6%`gdUE+y#7fB
zo0{@e9>~7?Xg_7+eb*J!jh3N1Qwy{&U0rvDE4Z=uUZs1#+<6I~whU^fSPxL}Six3L
z%|Uc**tA0@EIf_-Fn*FQkc5cnJ-+)=!<N=Pjg}Lks8?GALHzJ@bm(DKGCbV{pcw$r
ze7ZC7)Mp@*<5`tOUkQsAs5txS1L*9bZ-L0^yw=YL0G6192LruP0M#6X1NBx}S=m2z
z?I4oO%|wNHD+K`oy|J|OKD%3~9uqx1J!qPMBUkGs{^sP4ze3sCwc}_ZnfwoLRT)MB
zE&BcYMW6BewKrdrtiXY!<{N#+P?mWfx}^MX;opsJ>s!ytw_uV6Qj@NO6;wp8TNMd=
zIhvxp-hl};8ZhOiHl~~RLj@95vEBygOyL-YUlDI0p@p%ha<I>g>Zap^F>CRpntc}u
zSc?aEaA_yd*nE#+`siVl_vACw3E>mB_0$v-S|HB-%CYAzlNF{Qg#Y<70ER;;U?%4Q
zzTz+J?d?bHSlxcemutUU?1URw|07f@%E2~%$l@;_OR-|A7>IjI;PTZ%Qxz8e8G-~z
z$V1#=3ay^2L`b9=C`h1uVId5o<NT-Tyl-$MUO#$>X#5W#H~Z#;_j)%V`-h>(!to09
zfZlifd19;?zNn|;>%Q1}GOi!PjkOLw2IXp*ozO0pl*0h<sg?}1a4i;nc-Nsj>q|>C
zuIjKfR1=Pv{uA8gx7eMCK&JZCR9j#QQf25L%7Banm-oO7OgHpkQ7y%F!GW(+|LCXo
z_QwsbFl*v$-tUASFm+K<1raHjaYD%r7^13W3Y$fWq-Tn^WE5VjEOFOUfHA$jy$>FA
z?0|GEO~_e}J(&C;+3b;QNPs2id*LtTSmr2rc18cbE-X;p++L>3(xn6u<sKjr$@+2G
zPV-)Kjzg(K0o%|)w2eh6YUxHd{YLEwLGq|dECAEo#YI+1SEb+7)bt$Lb#Hzyk4KiT
zYhYjp3Wz=y??xyKIpbG)4~ONcDl2HaC6B>;6iRUjY2f8x&AHNYHs56ck;UPL#%fdY
z?QG^9sp%N&7wx7Ik%{<9@WC`RG@!}U{w(YTK?BHL@2^zT#pfW-f)A{=;U{>y2`rlH
zFn$5i-R)xmy@|da?y?&Dpu{F#&0z-5T_men-q_4mn~5WrKO>$(H9tZL=%<aM6mDNt
zQ6y}ihj*jhcmu)Ub|EWZmayNc5J)&0XI8U>z!zqD-uf0OK0wXs<_`OW8N^5_^A3Tk
zarOKc{e+UL#AvOE{OD;d|4yl*C#b@pnSyY!Ql9wf_;!Uke_jua14dF&UGb<B8BL-T
z^5OrtofwIcVIxHE%^j+{8zuY&fSm*mBeM=E1!1XG?dnJh&tHasd7TQWTWodX=AA((
z3y2fSborGNVZJj~IBH$q(la&{m?KhV+(H&j2G4%$Yrt044@K~c#Tv2sRDge5UqR6t
zQswGoGCVvCOAjn(>Pm-teek44@W^uRVUuCLKmN5LgA|%|M@>Rs_(UQCm;W{%_X}l`
z4??<dC<opfzSOLoF9d3vk(^M93yYM!j^p_a8K}#($&}o?NQ{i51mB)Jn3ybrO{~5l
z&hkrgVV44^V8L7EaDSg!tKcWhO<i{-ht={nl#6^wvXAM9`}BoJub|@5rT2ku4$KAt
zY>8q(p|O+p@Hhuk4mChZhQ;#Wc6y!uc2w-Jj6rqftB=h6Hr)z|TXBgZDR(p#Qe!?X
zj3NoNaG$0Kb~pv#2p}Ney2cvRST3!stbpPd_(34b&yfnk04jmgFPg&RjQREPN`m6N
zzjum(Vs9I?*3(5o@u-o$ZVGO>uCXTr(B(ll&;cEZ4Bjw|QhmYp60`$mw_DKRx1dm+
zunVJ`-mFQ*XN7}5Rz!Oz9mwy^F!9;vR>lT?9?-<4{2^;hSDFVmBP(uDj`@nD;a{fY
zh$-hk5MU9>q6KRk(DW3&g>qKd-k&;fQpUyQ7+7;*Alh!~2D4I`Cw~iHVG9epx=lO!
zq~jY@yW}M>z6g%|R$WNz2(m9|9?@yGuwoGab=Za&4~8K$-QawEb#(=Mr-7kP;7HXk
zUHDh?-)QX@H<Iim$bWUW`4915vJO4$(TIHB4Y@;42WR;-1K~OJ_t5&o91F}E4-kEY
zCC;Nond4??8+CwJh~Xe65WHRVBNPU;qRRPYP$1wHH@&_sXByxR15+1h*Z_>BynFW!
zN}z$ezo1DygA1-(ZhY?d&6)t7EBH=w#-pO5-cp5)?&9C46eW@H4Gg|*vkvD!Qi#OE
zpQrLUmsC3NhC2p4;*ggbERFH=v}AKopR?ZT2ixFSrZE(xn?{_o2H$@G?7BoYT@GON
zK|TQQ&2#Wtg;^{=s}1<<5)b^l+~F*_hVl7dFi(Q3C+@I_pbV#81dSO$?~QB6V`dcf
z!5f&_huzYZEQjI0W@|~_H&i!KGT%A=IYrlUEHvyaD`j6;4RPt#@f8G!n)=G+>SDlx
zBOGR<?S@fb$Qncjc=PuMKo&RM2nuzGj%wsDex$WPdtG2d1qPh0{sS@Axud2LZ;Edt
zvyiZGlh5Bvz)h5EfuEp~e7hcP2d0OLKXAYR7aE=R1x}jo@qOrjQo$;OcLo;noPw@G
zZH>dQ!2wK1-kko1tEKccY3OO$8VYpk&QgszDjE<a;fX&4CsR;qG+4sYlvIA(r1NSh
zXKyKm<zV@_y0(^kWKkN3W90X)41Q$D)x76MDqIdAGmgIbBnV?8C=h`>cJCjZLbJy~
z)>PG8W?${BJxR9mi1$vBqJGX$+9UJ#WbOj5$2P%%Cf@EHs795x<X~P6)|SN^UI92+
z0zTkx1Gg}|y$GTJE7Lz&+$OMr>3&}j7EX=p%Zi~b4<Xa-40`rLg-wW;aIr3qHWnZk
z7FmnKeFs-67*SVW_w!o*e2m8&81f*x%~~qt4bHeqs<3u5?3)qTUahmc^>|0atW1S}
ztoR~D0L@z`w&~f6t?z1ZqqU-b8pKp7q<SB&D)MP(0PbA_R=c>C=Jx=BA$|!X+7(cf
z3@-4|$cVwCGOB+0!2`}7Ay5&&gwFy(1Qx+~l{6M%!kz$`Kn;*-xu&cfuX>9d5TKq2
z4Oc*kVj3Zs2$owVfaQjT#V`<Y7;XZuC`Ta*?;jp6ZdA%)Lj}y^AgfzKK?K-VXIIz7
z7skMXd<ga6af(w8aU;};@sCig8V`6+3Ho<~h4t$9&(+|`0QCE4r%>gc$=KZ6TP%|Q
zL-83aUejxh9^HT7a?*Uz^8onWD~siSiCMV1!{O%b!I9eFV~`5LDg-){n%XCizd%L-
ze+8>Af8p>Vg@U#KN`a?PZ-K*91hx8b*$c5oou=fnIibmCQ9rd~4&m-VpudLkVud*q
zP4^mvq8uPfEc4L{cH3c1tHUmJ|8C%fL0sm8d<f<T@}Hncefg|z-~qzumCK)@jyJNj
z`ma>_p<R8bk%OiQ*~8_w0>UYT8L|!7C&Oi&ueVRfcLTLSB`}u@aK5}J!lg$+MP;)H
zt`cKpMDe?uMKTwOKW31QU%l1p`YWf|c81^A_aD@JA0))2LVV7l^1>nXfu0_e-EN0v
zxXtbF_Ni?WC47AnaeV@Bs5=092|bS1h*#aezHt7Fd4KeROYTt=2P@v?qgsX3L9NL^
z9J6=`Q9qLS?y(*#XM(*G2*@?`<RA%-4s*$gY0+1V!tdD%eoF-|2RclFI6I=?ngNrf
z5Y&M<w&49+A8dG3GErsBt44{rJb$swV_ESIA#dJq&A8oQQ)YAp2er8corV-A)(5te
zUM+`nzQ?92BjIR)1J~+(1>8MGZtjuauK;CzPhAP`e0<bfgh#>UqiMw96FaO@h1oXW
zHfWR4j)b!OaH37OAk;y?hbwyA_U|vY@D<#^?n6LpVBr#!Ll8}OfmB{2Zl|tvW&|+}
z(r&oRLfFVWy1wun++k@pJm9&rz$Q|N5Flvn`POy6paQ;3*7-xwG8i((>lW+S0vgCF
zz?^$U^m!JC-nhY`!)B_?G=3<8D+2Vb2Y~+~V-O?)DajWx5PHfs-+xdtG37^;pY895
zgofe{1Y?rYs&LwAR&tJb6SHc4%iL<cZ3#VPJ6nY^)w=r9CH_It>iGl4oZV1)JFR~Q
zzcH%=VE~CV<RfoUn$r4trGcJYmjy`BND*N|IXUd9bts`^ARoiQqUd4PrXQ<E4<DNX
z^7pdXQ#2Ypp#8G~{00k>y3@f~`50Wn!t9t!0A@-04K97pa;|O<i>At1>g!Jh#aq*Z
z%QdDEuUYW~6Tuh025v8S(k<P7k`p=+r?H{<0BgSnND>NbfQD)heG1>l=qIDzZ2hmQ
z(e>!fG=Clix&Gn;G?tc?l~pJgAJ_kOmZ0gVP@sIv6|C|1dl%J24>}LE@i=z<L%>JX
z*zEzh0S7l`mjE?K0JbAV-oxAaoOh)o;}7V-axL`T0^yYJA7#<v;^S+=L;{?8kWf%g
zH;1L?<2IqPf`S5Wk0xaLwRB-PikVVDug6Qa_Vy&8ySo8-`0gMy*q`746#W~?lt!(*
zUhD{fGxat7)~ei}Z{t`hD*viyUbblu6tb-2lJ5eR#r_=q6~Z#QZ!YS*{=^Rv8KO^)
z9wixBlL&GUDlX5h3AYC>$nb&L8Y15VC#K90&R;_Ww4TcLvNAHzPP_!dHOE#M;Lj{0
z-r3CYBi?YD>%e$t<baxsz<g)F%?Cy0wpD}#3xBK)3>B2wphuJ1?Ep0pMBBpMVF(HK
ztrzbl@Ie$xgN6X#5R!6Mh!1QE05#Z!Fgzd*OuQ5R2q=A~LXe=sF?m4&Ur_Y&@>0nc
zwsJ^e=7Quq3l-w4?5E9*jbLg=&_aZmKFqyxdkOts>V0z%?cPF2&=>B2EXxB>T98aB
z`l=>!NfwuNf~5Sr|K?Tzw)i<X@#Hc<BDDoBF+@rc*KGn^TvL$Frl>)a0Jv?|LN@K4
zaFEW!bCZ#|9R+|9xxC7@O%N*R$#%E4SX?#8xJ=riAnE@Siz7we3xGceRUpNvuzd2n
z(*|x#lNFE|p~Y5ly+=DI;NJ}?ET4>C7(N{!>3O*IFfG3|rs|Ir#=y?<$Hl>+9y~XM
zA_N}QAqPVs4yq%xjnMnJ0tr4E6$-F`bJZFSDGuNQDb~Jtx(t{nlr4`x=>gb=h>Q_J
zjcb|&bCF_bANV2;V7LjK<!$hzN%}?*{Nb_-ZZ*$~{y<{rH1P*|BS-@{U%(+PP=^yJ
z?s*s}zzSM@#Bb1Nyn+4-Vi^cYpt%*W9)I#IAFdAoZk6UMtQ&!N=6HVwMn0#kk4;TG
z?Jr7BQoSm5`U~0SIF{z*o~*Vh5dQa?kA|4i-)6OfzW`0Q&Ml@39{r0!vjHG6PUzwC
z2PgY)_;ZuQ<)K2E;5ElS;!UTZ72!~C4&^#bd{UucTm6yl3*WW59E<{t9Ds5PUjk|v
z9!AD>VD1zG_)daxN2tznSn63Rkmz-KTBmU7S=R%bpr*0<&wN%dlJJ<P3t6=Y*rkV+
z>2LsCh9Hm~T`R*G13K3}t+==kU>Hym;qELx1-&82$6vR)gNO?-)oEd-*insF#dlCg
zft?M7cW`j<Q64Sy4Pnhj$Iy+!#ha_K!Z!xWHma&ijpLJzz$WjjI$DL;5$~>dms@0E
zbg*JsAKs`-ntcjbFu2BZ7-8H|^p*05))6@BxA6=TP>LG~)V{>dQiRNE?Jla!%^_i7
z{-|qPtwtWhApQfgp*DFAk25m3UN~Yg&zLgk)`}$!q4UAW1MW~Hm4tC?WTJW}!{=E@
z5Jpg$K+Aln1D9o44d*VcLJ1g1g7Hv}P$mJX3=9;6Xc{*GoW!Xt1J{iOEHJmI!GZ1K
zq@-5DEu2#7)ig!COj@vqxjRKIY@TJD32PkbcMbUdfIwkoLLOemLzLez#Dw1rINUH(
zlIbU~+_~k`ri#?JwfRDm0jhn=trZv?($2X9vGMjNcQM1?Yn09&YP_OjY;jz9Kgw}C
zd%(f6Zn(9QG7SLZhgz?DP!7-3S|{YX-`4gv&tqsFLa8X^b;|C`9te)A;O%cS-`q%|
zLPbW#O2$)fGo=Ah6oPCf!PhSp21PRv{lI-l;=84{D)X07Z|T{+)+;+WI)shggSRed
zu79R~MYi0oRtEUxkYU05EEV4B@hjMIM<zWCan&RM(nRIn3Mjkb_FTd)l!v8URwMNA
zT{xb=%!d;;8ThPgG6v$5hD1ah9~?-UWa-U=<0(u1BnTWp$ZR?K5<&#Wh<7mnXn8IR
zpreA69z3nJ-sHKJn*bNmzBi5q%y!J&LZP+l+gKWme>ejjmeFPhRBps@29!8N?u&&K
z|Ak`Wmi%?y2V!~5Kv8C8m@G1p$>JV{tO2T$1ZyHZ?km4}^Jbd(9_W&9OO`*kwSbP!
z$yO^cU~jj?>uGya!=xWKD~ZC%a$6GF%?m8mE&2jpkEE?&P?KD57neU-=!_$<&5Red
zWI(4pwJuD(T^Gawb)9@3{M-om^4SH~fchn*L|Z+CVwU#d2E1){v3p;@2Web#HL=IL
z@X47ii^43txwf`{09M*e7ZPHtCiDLRd&<687W^Tsjv=_SEK~W=hoHp(T&YbP1g@|2
z`7Q$;JFBtRC?0-Il5kYIF@UpjH{KV>Z?~cGEpZrw-~vQ4b9bmBZeX`VrDP+j<0Y3a
zEFi4>IWzMGeAI|J3@bGJI$on<k$>Wx1qcpAF%_Z^uHfks-fxUPq<*eNqxPUfDutpI
z%~J%SbpiPQ#8+Rx6d+RwyP$0__)@+B1HtB6uQR)w*N|gn92~eX>Y<uG*Z@8VuSq*X
z^JALZ%CK@Dz6vuXA0`qAQbIU!7DC}^kMU}JItNQRPzk!h%$&JheTYq}5%w%e6*7+y
z{x}4S)#WQVM+Y4c`lKx&cz56(5Gi^6#!8-x^52gF7sRWtQb0V1W-^q}A+(|yKrnDh
zi`$cGrH56Lzc9PMLW>yKHt+k-JCnZeb}d986|OlYcu>|xL0B?JRvty|veeV3%mFPi
z`em_tb`Kmy*r@}+Do>?%UjSN<`rQl2VUDc;taU{#E{Eumyh1ty!t7wO0CrW6Hlc9*
zkNS{5CRbKA7;v;({LF~H77`G{rY%!|T|Mq7o<NDhZdhld4_GCI3j!#}qx$ZbsUyJ_
zFw;O7Uw=lCKc~iUj@^dz-`lH_?!-8A_!dX<ktoAot7(p+O(yecj@NSm;1?k4inzfh
z>l-4+fDO)PclV$`mZQvA`xbR~n%k=efLc~4UEiX&a$2gAZQv0T>sFXZ_60!f>DmFy
zV}+gY?Lj~QzsmM}In;ZQpSMrV3<8L4l<Qx&XB9Fkpp0IS*<tGa5Mv$?2NP2T1+?!~
zP^`d6BJQC9|A!;7LP@Bd3)E;><^+~IZrx!YsmjO%Zpkafvxk}!l|wx!0g7J(F|=UX
zQ&=i0;RfS;^Vn+*RNX<bb69bu{Mj5CGDW;-$mWA`u)A_KM3D*oH2OP=4#Ve9?Zm1w
zu&$@8UMOEgL()^A3^bK3&U<@xE2u^L_ccc50@DyiW4tFOFuC0~+Ghgu4IZ^iZ>V0O
z{(G}iTE7}FoRJ9`5>S3*nou6iT+`>kb|!&6LVzd$$_tpjYxQ3}he_rFC|ucYAiGi#
z5>@O<{zg7MJ2<d{cLw0olB7)b<RWy!U@g!+hPJ%Ahk}f(NJ$RXYQ`InB*hQnT^W#(
zkd9#^NUH^$+-TG#nS;lg9gw%9Q$|ox^6CEBSXveoqFU#-r(;5Ka9^-<gliv037Hx3
z!UFvpTWTwk!#mRA=_}w48!d+|<s3$GKRQknEmPV!?*Rd2il}GR+r7dO@V^|klFx#x
zugCoS5R|L%CAcj@5)u+}FC?KAU@BqpCDz8(#Exgtl*@F2ilJ;7+A8LP++I`Q19`1{
z5E%1vfFbF5!*vubEoWnBe*!0XMhUP{euqW_tPI!KsMj$tF=M;$mr^fAi%tSe0D>ub
zd3oiXa5OEC{qutr570Er#N|uCiYkvkfOwDK1=gdX=ce(D9LX<rK$*sphR_d59eX}b
zuKyjBs&erB*3i;>KRoF5`iLWSjL?4po{<Xp_>ilgZ%=0LFG3V-zG(xzW(i7nn1BO8
zAVAP020omwHO8v~R4uY$1XCb5#dL-7p(eIe`Q#z|UX@aomdx=`R->@B6M*j%q}nPH
z8d}eNC=elZ`P;&<bBSPBT6_L^4a5P#_YnGefcMke3oE0{p|~<TRn!a=aP%07h6uM5
z(d70A<ChdLv9O6|h$#b89q9oaM`T4XyQYrBc||X0ylA%U0v*@|gmp9YGL#{sCCL<)
zctPhwPI%Pgw-OJTp_+n%0$7)({-{XmFe4ZpfK(+_(l?8#h{x)%I8Jte-Vo}ThcLfl
zC%_&ipo9Wi0y<;%_f%n`JQaPtRKDzPekS+5{4my@V)}yNbAtgUc8OXhqM2|bD$T8+
zo_4)ciX88T^I6R%;9Tsf;-`@(Ys!Ek5tDO?hyPhwgx;K$O5je58!&}X8ExC3UDtRm
zIElZvydwBc?gLpUWsZgb{r=Re^f)%%A?+hyTq(3i3=kl1SdE(nJIROlh6%o{RibOC
z@&~>|bA~b?J%#%xLLQeld1z2pcH}prG7sV%b7W7fc(~x^$ys+lenY2zPE|>QSPy{7
za=^<oXc(krEC6IlkkivMgC5z($LBClJ(Fh)6$ts!nmt!+hhG@;YnaSeCWEF-8Gojb
zFyEqH9ysdNdXe}YvO1z1;3<=Ziu1HkvEh%;<BS3nuzXjFDFuO#99gw5R3d;rBNJk>
zcy>a=0LOSQ5S;-`vT#nQ(`uu=1{vid$!~{Lpv8k&MZ}3g!0d3oZ@M)HR-d$&>&Spn
zESG_M6DS6e?-2ow!yFFX%U&Chj3>hU1tdd5$6FugLP%$Fn^WCzh>5|-k5Bobx}XUZ
zFsJQc`ZPQj1(hhJ9w>5Qc=5GlQ*{N9d05|ei&gnbJ+`v3gUxHW{I~eabW0HUEuKOW
zujCE82AF*E1~_yhKD!W&z`W%visv3Kd=?2TK2q|ya4=_bfcg!RtM0b+4Wt|0BCWn4
zO3Zur2GYc?V2a3Invr>HzxOG7x8wFoahISHHp~V95)7_}HQc|V>L@{7+X8zK0bTdF
zf{z{&YXB4t{wNWP#wvUbKlas?I$$6X5)jk?qyd8`Kmx7-O2ov%iuZMcEa7o4y;Qe+
z9w1ksVqSvlV0gXVLhFab%^wo~o~V9C&Vbd$@K*n5;%NG;RELc}`-&IP!S#Bq;P#0#
zt1p8|WYe3IqRs%l5rA4?LbQZ{<CP}p*f~O|prq6(d}~4b*3`!4H$)vU&u6g71my&6
zYRGH=fMI_Z1F#{7=)>^c;~IdO!;fGZTms*BlAU7!!ouwl&x?%15{DWSfUbx35>jZx
z)u`ZD;njRQ_lL$$$`tr-pw*5_%OHfcjL`B_v_tgDfAU!a#-g|F?96SD8ixk_6rd1*
zi2~yQrFy7R%;YP6>(I(BFL-K%VfMQsQP@zNRsTOe&Jdfp#vp5caoN&<(QPB@t#ps+
zDP_Zh!>@M$f2D?EDQs$zA%p9!8zlI{AQ%&V2X}!_G}181eaA*!$LR?U83CaF_k#SE
z2empFG60LJbA_wi!yUX;^ctZVI%jA9s2Q+Hh#mvt_-Fa56!SiC84R$Y*OdT0D|1Wk
zchw5aal@83;OHCjlx&UOhE}D>7~)E|k_IiV%+sIo`q+%g*2J~D9WbReu|})PasT4F
z!+3VPZVgO+*ksE=?mu}0(lOgp3^dzNdZy^^N()$Zr_mnrLy59G{U@!e+G}~)eZ3k>
z^}t-S=MCI|NXC}Uz)3mK4?=~#W1zksDAE7V_kx2#F>ugUgQOv?#o#1?feoY^og3V4
z8MKVRdpJ*6UiK9lw>0Tc%6u<KY+i)+IU3cG16Zyh8oBQ2qMrVDkIv2nypX{~74<DY
zH@C(!ZWtoKMB0;XACPedY>XAyJf^q@3)4yPWX?k;V9lG>69$AS7@a&_l8oBXNNV+i
zyErc@2*?b3!t^B)v>dnjpB@{eKk_#m@i0Q=6v`DT5alNTzfU<fY^(?T-+O!I#d~N(
zMGX%Z0Ra$Wr`Am-Wav2Y^Yc4&1r;=%rvp@ZF9jkuN%nRZ+NjD2w~`#)Ft9-`g<JeN
z^)&VN`h&j>9m5SF0kc{Lq?qUv0&it{Y$*XaC}4Bga(!!eS#-7t2aLe(cGKIc<iVnb
zIOg-?4Fbw$hb2T~5R+Xdp}!194cpWZ&)5T{Yd`qayiHwPH0cWf=!xh`!^ZGQDz&OD
z;QogniCHvcmLwlNEsmosf-c-}NAi*1IV8=sGQYVT<47G*G^Ok3Uheo91|;ts-!gI2
zE9$eTxAN;%=04GXGDU{LqSB`tDpgfsCK>&OI%<rUCyHXK-~r7GNjoV5ekTF^<#{Ii
zmX=hJm$_bZTmK9!he(~p^Q_X+gsAM*)up8ayp=B2jxNOeJL>A{!oo1z)vc|qFHxNz
z_-41X%&nX7lKom498|d~5$bv6^fTqd7Ac!(=-ix9i;nfN3k(^}MS8E@UOTDn#Z4FF
z-!nMy5T&E*qR5bMEcte{YC?@uv4?`Nn7*+H*b#Q`uV24lp@Ft`T-M<1mh75xW1hW-
z*Dr%Z@|2X6F*A8NIXioMz#aek_(qeHlaY-?$BERH8$ExXpPLJ73loJdpYm|0_DP-$
z!qV5IBr;7BP7V$wC8gOT3oR5rp+}EQ?@jde_4W5Fb-FLEtuZh$^@{Dl`a%v75rZph
z)_nda#m+y7?*~bRbp{0?du4@(IkOY-eDw12sy*Q0;hApk>XNq2Kb;$tY*D&5)^>1k
z_Z??+0Oi-OU*9NQ<f3kKE0M0nwtk;Wzlz?-I2<s1Z7L`EXKTyK&aO@u!^3HReLb(H
zW~`*Cs%qr>@4dYV3s@xz0oBwrb4t~#+ri7rAeR$njLpr4xG)9VUe8#5%94HZ^2Lin
ztM#?DEq!P4S5p18RaM;Xs`Pln9Isx!eBD(;RBIxDs(m?T`YVz!<5}y?b|oqK>BQBb
zg2A(A-+var`E!2m;p*yYZCz}1J~R}WcDX5xh0wl`vwij#`?O+lbaWI@!m+Wjk&$Q$
zjzK|T;SuYYAD$i_f`Wov5=}*ztQ6nh5(<}oZLzJ5R<2bYUNmmY=xc0j{Pj!wguB!z
zX7^6*)s+`t)zp+e-+18JR4aUaL&KD`G@ZBUsi_Q}_+#e;NxSU=c`-g-?mj+g`{Fat
z)~~Od<ICS3|86KiE5?MVoz%Joi)-N{b#;4RVIycv9vxLDD-8?`ync*P%lV}vH8oX5
zMJ2teAbs!m>f~gi7q4b$A>vNs85RQd*Mhs-Gp((y;v|^k<KwJoWDg#+9iC5bC$pxd
zR|Ub#6HouKcUJtb2-$i;8)1&9v9hv*2?;@C2Ig_Ce*nUe&Q!ZC&WNro6|%?hKlXMl
z?ZaJ!6};{3?c?J_>8Ss+U2p%t|5v;3s=!y*`R)I1p^y;Y|6M5be|E(2aB)5QpY4cy
z&=}1oz>!}56B|>-e^Hw+msOGHjrP#E$mfA5_dgcTD^#zqP<}8@1z-Mb&Tw^!p|(S$
zPD;NKJs7P|*x49(>K@9}Re}eT6W<ghvBhH+!MU+*i$IfSVWS&oqviG~0eL%0#7-Ox
zM(=$LE_F6^S=c||L~=O#@G)<u6SzoTzZWs8Kb}c@{`Uxh={ui2dR7?EB3_z8g1loB
z?+2OK{&MLRq*9Lk>lxVQAR0x#X`r=0mCpIkcPn&6iKp-2`09-KA40ClNGCcyozqj_
zzh}q218o?An4fuee1|vcSw6|}-0KOrJ`wP_#1^~$jH3MNsc&%R!T|U6aocE{m7n~(
zhbwm~=<hTzv|l5A*H382qK<Wm$ezJ6V|YoHST~@C{7Uu2Bw$)!@^x>mPLaBZhJaeM
z3EMiFhlU7!uEt{dQSrL5Bd9i`!>C!(@wSq+5EVPruGBwmtl{P~^>^dmiT{>8#xQsC
zwETjDMaO3d|9W3^NbYjbUj2lXgKt~sye&Z;`@EcdRF!-rqLf;P?5>|Is#ip2eG?&v
zQou`Q7f1D7a&nnZj~a;_B=LU>HIS^GpkzNGydY?npE~?PwMC&9&pntSa;Wl+O=eC*
zGGKX*m&4M?|N8)b+VOt$aqf3XzIK(QLOKHSp4luronpge{ZEU{`JZczzoaESE-qNA
z$2lnNlzcTn@_}j?J$SPlH#KBUHz-J{26>=Ilxd!htmWiGjU=_S5r0Y2O+z2jO=I;>
zKHDtOhqIXr_05IIzpK{^ctQ%)J%<>qAI$wFk?j2T84b9j0}{#m`?E0{Xl~TrI98I~
zGE5{1Xy!&{gJRsm9K`9v;-X(^*LXG6Qi{KaFpjl1lBo}cdt7_l^L|E_C5*w9{pC>G
zrE`JREPvf8&QTrKl*O*Tf9EQT<Hd=APT4z2Tm;a|ccC?5U-)EZ>xh$OwuGTmQV334
z%x3ARCZ`6M)DeDf{!vD~63@Wxm#kR)stY$jJ$~wq+O~<@nT_C`Bn6!Fb_W0TYYg$8
zP7*#s42-cr@3uL<=+rUH@F%37`LZpOF{b&K6MZ#m4x@Xg`*V>f%?7$m=MPWGGd;wj
zn%uurW>=c4T66Yith?vgJY2Q789wERv$0u;M|(c{g_19?ougs?h0<>&B_f@~*w#!z
z*~uPP@eF=X5)5OP5Y){bgEM<>L|weZ;g<w{qni0nNb{0L_+lC|S^F~4<`%>&n^8Ui
zAAg%&Z?=>X>qapDDh=EGgb@F+?MBqq-RFW|EeF|o_ShP6ZzP&`dLO5l-9#{$^!~re
z2nw|Q*`4?YYfn5YOsU7vgyvr>zjO+rag6WVW=otcy9nt!BJn|wnOO^b_R=YsJe@n|
zq2j}5Uy-(7G1i8$<<)xmXcP-_O)Ev_a3o8XbYZ4uOro_i7Rv^I5AzJac+^+w@!IK2
zx+yl2=|fEoO7*Bg>TxLk*cfgyeiOyf&aotO)8iH@6W%bLpk)Sszr(X8SdJ$rd1q3%
zcX>;RRrABs%=fCK=}1OQ$R=M>nve!G3zhxI5WDxwJ03I2@=KqRZK)Z(jvoKYoBvdn
zivf4S>h0Yh4&PX%^<pE?_K+_%gz+4oZYdz2#7mudkya-r#We<4<RNm)^jXy6mao-w
z@pxYC=a`f#7oCSK9yvA07s}`3>?e(?Q-2}SIrTUk2+(h*my%$6`PI)|$l6Y0;S=xI
z=V$$nKA1m}MZZicDyClGMC*R`c1wDzKt39yJUFP6R_QU<$2&sq_6~P&B#Kb8e{S?Q
z3*m>If6S2=I%y9bV{l9|<UdvB+%rnTDeP#f!aet(dK>DoKs*)@bNuiMO-4algpV?E
z+v=mmbW*2?xQ!0ULl0*?7CuwvH{YC18Af*hTy{7G%FVX@Rg`y7V<29Lq_i!ap7!~g
z;(W3h6($r*cYXQnG3#{U+3S@C6}j_o6g-P4cS~uJNoG|GZ&oqSg&9&<voe}LUcVPb
z@8_mPO%mZl#bd0g=jQn)k5#AFNn4_<&U5HYKz-K!vXX*3V5h8vTDjb)Ou41*P6XmE
zLhBxi0s+s`6Kn@Y`=xJcp>__62K=v-Qx(2``_12hB%3rHuJg@Gsr}KbHytvbT?T}d
z^e5A%jTuqHhhkBvXuLRCztbM?-CsZQ>_^+$Mi(=AA5Y$b^MQhcNGah@MNnNQ&e1b2
z&C#jWy98gY96wk4l?YL+pD{A*R0_A-`xcZe>R;PsQmobs_v{fR$Y1YrrEkE$+m~7f
zCy2ErJe|q&EVEER6AAg~$>ALP#@29zZm?~4KJVS3@>ltqAB_aEG<RMc&RKr!T)L0X
z-X|M|%lLY#QN_ebBo7(mbBvyKM-xvUsS4K<wkc_E{}YN&jWq;O$&~p*pXs;)vZ%zq
zkz;)mwU4D><l&y+UlrC&oITw!s6YKXsj8y*$+01xb~?$suxKK3>z$!S<VM76jF5Z1
zuOyw1)dJ9j(&j0@k<lmP{@q6Ppf9oLY@l3z(yUXE5bctC>{);sdcQZg@S7zvg4}*m
zfTID{9RaY6dZ}JDFZpMrUWHN#5kJE#gu3|kN1xxJ(^J-$ic(Dzy45y67Dn1W>5)^A
zK5}a_uvDvc@;%ftlxM;evyCTkAi?<jW!SZ%HhWerq6L4!M>^V&Cs)qn@)I?IQ^Q`A
zAwR36()#ae?be`6%s6`*JM9a%Q@uEMLD>QF0@T<@a$}LE6sh?U{%u<Z+k&efG+wb>
z457|b4q@L!#@q_(Wl-?y%04DFJ{te#OztwaS;gXx;bIc}BWfdYs8(*{9_NedkG&+e
z&IUi9U~eU%@pDhhpQCN^x%-gDvCQx_C!BguFqo+1o`ijMYR?gSg3g!s>s92kE%&Cv
zk-QDI77}A|QrX+!&C9Lbb1i+65_3A^!}tSdhBdsL+L&@L^CH$8Lr&&GS?1)k=p-$Q
zLOy&m`th0JvOD%Z1d9^Y)(B|Y(hE=b$PgZW=3f66wf_=_?U81s^R{vm%G>w59-+~`
z5u{xuq~Vr5jmq<w^SkQB?LHi0i{iq0OF!9Q2W$y1GGh24sw>93pAm=Os=7>tM>n=h
zG2BNksy%j+XdSgCie8F~+J2-{9;hh$akbKxSEw7;6BDR7MB{Yx-u%adCPWXR&n-F|
z6B7rle2kTII={$Diz7o=Y+N<9DK1;+)?e~|u^Ss}%%)D2cjj|ReHB4D`JAkFO!{YS
z-%!n)E6oR6{kUn(DRH)MCX^Eb3VyUG3!3FUT^z8wkZz1?{K!m{`$9*+Ipa{GaF5~V
zm!pX3HeCWEHD{Uot#rnmRO;nZGrHvF_t+9Et0KQI^AVDkhM#MUuU%0-H1U2c&O&bx
zBwSd)Jime*V<w#NCdZOBmW2}IsZ+SquL1nEScD0CyV=v2q|YJzyn*9Z#r4r2-6Uok
z7W&FYSO;Q?;zWLlXfMRiJtmvpF*MNb<CMkZGu|T{DcPnL<24l6%l7;|?m+&6E?fCT
zJiLF_uShK*yKZ-7B<j5oxs%Z6x3%oijlUE-bc9aq$JrHgjPp-t?qVnDDzMrc9GE|Q
z=G2rcxZ1Q8cK>VW8_JD{FdC6_Ws$!>x9L=T9#s_-Jz%H2cbUt;+;-|#Jo}F~`+@(=
zL`sr*L#QoZrA_zYh+Sg$PGin?{FnECDy$wDG8Mc+_haiBbRcQ{$)9-6*^4R!(LGRK
z>;3&-2Fmqls7AKq+J6?uTao!K+mFM`7K_f3IJ6Bu_||_hC>Keqn)*CweWzJ_Pj@OO
z=!0VPE@fPhv{K~2g~LDb`9LO{w?3Pl$%Im3Bh{?(w4c8&1-Vx@8$@K5yl6P1H9^V6
zet$)de9EVIB8at=Nw6Z}R9e?qt(Mt$kJO}dsY$a~do6i(_$Q(<)!y&TwXjm%KmL#E
zwv=Z$oYAycZ71C0F`gB66&WE#R<lu^)x9gOo(Wd}8d?2IkM8fY_qnXby%k$W=R)hN
zbg_{_%0aa^$#@@p5wzojWdGYX;J>bl27mwmz5bK>|FJ51&R{TdmLIcqPxBmE2ls^2
z<Wc)?3#P|*wr51EWBiT0!RsRvoF~@Fb-i_>Uzp$QzcEv$m(^I6?_Zibi*!$?OnXix
zZkW5x?o3v)Y24=ej#KakITHKBE;|AbUeg5CTmsWjP{$4_=hwzq5B=*u`(;l*;*m!p
zd?ZexEJeIdKx&q03THD;xSmO+?7C858mfKzIZKtNTc|avHDcL=l)Nx@%nv*A?t|U@
z^73>l4`aMn^MUD0a{o>aFZa*x8f7@e>svJ~_=*M+D<Q_Hx-25&kGo6!``s#bE!Bs0
zvVm~Y`(gM^j)dQ;dTx&gV^?vGM3Asr1KEc+W6Uj`aj(ai+R`ZR2M6Ky=ksv~fmvGm
zg2>_O#M8frqSJ#y{m&h;9#ZCQ?+3pM%VxS8#*BK`o<;V<_!#X#7@K;+@sm{f6js8e
z&-M6?4u7qJL{_6BGY_!(HUH#L#8}QgaMN&$Deas~82<8yvB5n40qYOkFMcl&d2<^k
z?E1O(n4D@SRjl8ZkgZL1Zgu@SO7_yw&&w&Tle4x;Xe3lNuHxldZf%(PU~00?67C|q
z@VH2{y!0%K&X^nQPP_Q*+m?}6NJn-z9=>B)VQS7aJi5*`lTzjF=9IigYq^y2hmt+B
zt7Us5S4Gu2vDS@t!!?U;$U$i8L!hdb{8PL0Scjw~zTsk3HuPahBY}%gtNw}{EDT)&
zdlj;6x?w`RiM&rtHy;jpZs89kd07uc#K%ensBg=WYIe3fy6^Z$?#@nPFHLOtQyLLF
z`8fVJfoyd`4gPo5g@ix1XP~so*6Tbn_tUUM!&s!h8^|7rr(1oQn&VpKub+X(@}vec
z%<5<tO`f>}#{%zExp=IdG3(R}9j`?ClO6g114+bWiLBfn2{%&=7Z>IH=~sb9i-zvk
z8PVI3UM+H;Y+_3;ze{qGXm4<}^7*}NP+}hXf+3fd(A?`GeW`>H>(CS@%=43;=k>7&
zd1CFoa9!!UU%6hG5sBK+4}2!Pd~8WefECCnWo2%e#irN2UcIDf{iEapoBE16_*5$`
zxAsBrepYs9Y_GJABt_s4rSH)h0bhRVyJJ3kCbZt*9iG8ZiJ`WeCTab|an3ywxgzBZ
zZNlIVJA(xGQsjX$$$Cm#eyLM#-tdp+xA!^?O^M!l2}<_#1e$R<ti&T>x?^DGHdHZK
zhfw?|GH~+U2|2{}dvm9AQ6blF$m7eQfB@a8RRPK(^8MPq0J0L|@p1uA4atwO{S>Q5
z(fZ|iu}sFTUoi%dOH7~5gldo3VR<JleebZS=Ll>}iv2eK-Y_tvxxR}sAz<Mp&U@_c
zPU>Qv)Wa}GJ^2~uzZUHxn7c>!t%C~Ccy0CW@jiOIrAv$<CP6$UbS)-5Vq*Kyi27oU
zZSST%A>Fa!xa^7v>#zFM{%-kh{ZE|dG!c9%)Da!(TLvzkOo_d-Q3a<2HOG~#)5@mV
zh9r+;gJ&#;GmJHx>Yv@<jFiYMle|JP!b_WZ+?Yv98@oQm;}IFTG}lYHiycUH@%Wcg
z^|kezsLh?_PaUVxlbcxz{EK5K#JOIkU4;o!bLi^B`3-5y7h3*k!!qFnNGKSy!GVYp
z<=^oVo44p~`#bblku^1WDXN2XtRgHoE?4n_2L56E8H@GWA5_l5iCl|StHgJ!0#5=l
zt^4{yju9UmqGk>+Y{|EjVw#;;?_aUbRJr#G>d9TSPCb}#Cv_^xzFb$EcgJ5-%>LYS
z_}Q{to=QNQ)0Wlr+wW-eQ%WYKlC|6vnOBe7X;rSQ%v3o>XeI;2Yv;6F7C1-?GIRBn
z8#uH|at$h^d(cpV`F__nrb;(NYnr~nkd)I!n@ovi=<(TM%KxUsL-*D|BVPF1>%H_?
zhWG#)4Jx$5FN!v?n3?y_>|^3yC{WieEC_4!=1%Dt)-2ZIENFPsI;XJ82QU=%m=`wA
z?mAykiKPYOr<`X>HK&hObLkpl<U3(JDhaF8NGq*GWSVde(|C@z{b2_`e?$yJ#3AyR
zrLHmU*Ehd@HXv#wl{PNZH>^m`Q+x@Mwwh67sW>g0d^lB>g46BbN@;TGq(w?5^l~iL
zKwOoQ(%o3;l>KR?4(p!hLKJhOc;GQv(Rl1aNiN}^0CSpv)(@vE!-xMKUkoZe(`<dY
zxSIK`C84{=0$I<Q4#kTK)@oa>oe~nMu^SF==Pv6{^!|XAy_Y<T9r=v+5{Xu;Nod!?
z8tpAU4xVe1gyv<2zixL6$05e+W*}Eew&Cn1v{cxo_$oXi8}f!=`8`J}$?;ozyUh}m
z(HTC96ux6ad=G|MrUm-n5{I+|l0yL-Rd3cR-m2NE;xr8nQl~pG8i`CNGiIw_kYqfl
zNsHWMY<8w{m(*>PP<dHPfPL}Kk;?ZDtq0!=U($4Oq5!4vifRdwIPY=RMb0L6Oae;=
zC5(%whV7euYJOV?@43)-M;h#%$Yk3%t=uy|;U(!m(x((knjQQ{nX5PH(Le2VDvzvt
z#@Xe66tsToXVIG6&EMDVd2DpS?@Hnk5srb?a~7H&?@@~~I6YC9#}iPN6{+Ve5NbR0
z;9J(CdGsW|_$OztI;0q!(UTW7wzJ>k?`uTpWN*Y0NrgCEhQ387w>syePIwRzI)l^w
z;)d;)WE#^YuAiArvXo2xv<t<zsGU#r^;>NfEMr3;QIRpaD9DK#@ic5XbHW+=;T`S~
zWgRv2WZtl+bn()*nBrB-GBWFRc@$+2J&s}0)K7tG^P<GZ4hGcC1zgY4kt}w4>r||s
z9vzq(i$B&6oNlO&LttHyBYzn9p*>+pC!4NSc0RF~A!}b~-WI`~bT8Q7tf~K=n1(`g
z6=B(l^4eE95vjLWX(NM~?&?YBE&UdAhPB#M?S4~Q?U5rQttXcbCwKH%4?1)k_JZ5c
zHwdp)<fytA@aL2F1wOkX`0lNw9|*jV_}HfD<703xwtf(~nat&5u&=V$Ao35{<w}L<
zzbS!uxdi?T63Wdl^glyF2Mv^CRvPZNo|XTkBEeF8S6;-SJC$~-_2*$sqwOD-zTFoa
z6<&P{*a}eu$cp0w$*a^e57ntDsUE*G8Q85?b$h;Gx3C~Cfgr-vtgh<5;QNO_?9*`}
z{aj6+vl9BfqN|(is|zC~+pVVIAPvNyYn%1?_Rbeb{iA4vn;HpEu?YVnhjGmL?w}f-
zUC%rh|GE*X`evf#Ral;{^z%`Cv=^~C10xDDT#l4v`Ic@kKrsDpOXX!0fp_ux)ymZj
zmAxUebqPY~oP<yFS5!2mcN3{7lFM$sH^o;cf9&I=GTDPZwrw8mkSo3ln&61Oz963Z
z8$)>`6F_smyI{5XQTW4epS>GZ{1V(#DaOvnC}=;>22{frr5Qy&ap44ZSEP5#4V4g$
zOXL^jpuLY#sW=KVVJhW&i6r8m%>TIN1DcFz7Z$qKdy|)F!U{ON79RA>q^WybH{y6|
z`@_%Qx0}5{{2bBdf^`4X#A=o7PR{G!1(s}NJd%?cs#8v!p&#}csz{D;&z&TuaI{M5
z5aS+>f4D=VfB4K_p~|$TUitJ{4yr0iwxF1&ktZ7((_LS-<y>s(p3~~=^6Oc_V06Fw
z2q}X-<0C-~)$?I-$pl=RLlNBCfMz|G#PG;Ol_MtJdPCW1^qEYm%vK9g)Z_@|=g+5}
zSKq;5d4~MT6CvVK<|6|lca=eN!v2ejYP`S%*J0`?uOXxOov22+H>Ez;7Y4n%Z?A;8
zum~!xG~_;udXZlrS;|OawWD+e-fwc;nV`Qr<)9{6AKvn~IOpuG+%wcyQ}+mFaTu?E
zUGT_Q+}uQ`(#$t7g&8g}S}v($t8wg(2KOAFwK%f<k_aD37Al>sMk%3FCv;KzR^m6D
zs&h98zdZ0v?LHD|iQr?2j%flrB1{q_(kM-X)iZUemQ37z$4MWX)+)m>oTgwidvn!;
z0n;%0YPAj(<|O_5uR{@8c&mn)-FM{JDDEN%lq7j9aAZxR<ujq^aB)Up5tK2mZw;0y
z+qm5Qj@3|6$832MngYIt&qzx1xqI5V?&x|CJdm7_8OP_?V)l<IGf4QJnHLzBBcMB8
zw!tBkSJ%NXU?{xdDEx6;9o1^p&VcYZ!c*TV$e*(3;vMB@KJ`^ge~u=W4OGMkEG>gT
zmfm|LXyOcBHZEG@q~vOm#FhuXxL7@7s>|&uT^)%DeB@k1|Cj_PdmKADJdX9rKOyJV
zkfT~Da*WjR>F62f-Os`EQ#b6Ddh!)-+>Y*Vxop_X60-3Fx`>o-t*gE{_VTq94^oeP
z_<#K*m?&KS>|mBwj>3~N`y&&9w$3mdu|`_fNM{U>P-##@XB-_br9!Kor-(BIgULp<
zIm10ApCLJ$2#b(PY|=2j!n5xo3DZ6PdgGsb1ATu5=O<@BIyb6>n$OU)I!0ugXC@A7
zJ-e*-lK7&G5WLD0_QOX?zCS>X!y)XRkX@Yd)Mj07v_T;{Lz#|a40;Yt-TNY*wqcQL
zpV}ygc`Ma8?ba`d1eTdKYXq*th%2rvI3E^_pD#M)lA}<Qrn`!c?B(tb9kcGPkg%M3
zXpvNIFL?~>)1mUbPd!t5733WG(ayR3t_a_+iO(02SGWpn5uf)-dHico8O`Yy3!j=k
zmVEI$oaRSCE5c#Kvg)_$`g}EU9DK1~R4-*UDEB5etNhPev%H@7?QBKW)+|IWvYT=g
zCGg{1O8Hg0A<RD}c)XWNRf?urrXJ6`Su*-UcO5~vvVEi5S#GFQRo)?_GzAQD?0uL_
zBrX>Y;~Ct3iYQzvTz}Yon{0DXrD&G&K3{uQVTr3EPR?wI-p(qU!~ED~&&tN1UQMIH
zN|xPSq5H>2>J7IE$=7%A7`Z<VR@4YQ&tmqya?|Q$d@yInAEF<Bs`+J53Z40;TZ~Km
zQhefMauh+10_l^F@g#y=(a+|ia(e{2uUKA1|Gp%2%C){TV=ZXQ`xhSNN;o|}_%H5h
zf=?Cc>$KXmLr3q@DH@`(OM=I@H6@<RIQ4ZxQa_IW<DKak*1i;OdoF70vZiT0l^we)
z-Pgaz*X<FJBKf;s&(t9sK1~i#6;D$$_B$EZ-^Tp#7R7nmeQ$5asjaRNpA&h}lsB2e
zF~sw0>x4jh)6?A5SJ7%I!Ul#)PD&RSrf=n6jtTF&4o*&@^k=`<QTnDoadOC#hx<t5
zJnyg7rA#ho7@?sA_ke5*$<$aqV^^JYc-5;v7mEI!L~k;GH{_(H-DhS(Fm#q%6yG|0
z-iW?0%xilg{+EQg2ODqo2-({r3w1GjgLKEXMB0Dsudvuvj>=ti^m6tDF3L5f>IToW
zEOm{-zj_6tEW2u4rdrBIj_;p;q*ZI`zj!Fz_Hs~^jM(WZc?}Zof?&c!y4h|c93JD5
zZT^mX9?{8*B2ivBr3^JRzuF?V7lvmr>@Ka?DB`k5dO0`W_*;CE+G*O&%f=nfY)KJ0
zTK>ga=8S+@XEH<l`nY)F@*(P`{FhNf`C2A>R$>;zv+C!`Rh=I;%_(@o(24Kutu#1{
zvU4=-w@yD8sL^y4Ng>`oDE{!#I;p*ZSE}!b$$h3d(V^I?(Bd+<?cI^7dn{d$le4|p
z2F>^%{i6xFAN7T433OYc<FUlkrd~<~PS4oiJRpi8t(mPBm``@9;?Q{8_(XZQWs0`c
zhe`nD>g@V00f_-k0n-f~3;z<1dnW^hPHAG_qiiGNqv~dddPjy-(tBO61Kpy@c%La|
zTVzv6D2Wf5YiS}?5;aC-Oj8soI-S_ug-uC+O&4jr6RQ^-ESLv`KAKfHvLOj3+&djH
z*4+B@yvj*8*23!dVUvciQElDN5AGjSQR2t=rfycaO;&_0|8BChPacLDAU0#<<#&im
zATh6CF+S`)$o_2?x%+lN=;~TWZi-In!#GX4fs%<%9e@AQV>uPyEtb5P70>5(cV~)4
zV;*>rV7-_{tuMagHWMEqrDW3X5Ej0HyMY<S{gU&E1JSyM=0lb>hm%c&)vAW~JaJvm
zSrWVw*Tw`tB;J3RJYuHwBgj>n+b?_j?hJkLFSE2vbVoOXkS~#l1%F17PQ<lHZO^GX
z55~kg3~?3`cz;c3(AnA+ghhT~d12e>DIrv1EsNlFSyEQz_`p;zNZ?M+@9n)k=1=UB
z!s6Df%VSQjXwF_v8K5>T<;r(3KTM??>Xo`^MqM51m28}4dgFcZjwxS|Q*k%R!JJBT
zC?oFi1D(L77GnO(O~?1MzDr|cj<W0Kw1#BPDQ3H;4wG-m^AkUEs_>?LjOpBxE=rl;
zyw@9b*+q{ZEEt*l_S0(Psi(82f9~8-yxTYavpX-$ki4V*y+bignc9@VIeO#~^erRk
z_*`njFVVwq!BRBCr5DLVk|22UTKY^$ak)^ZP&c4`DKbWbIkvX^w;&5f(0K5KdkR6z
zi`7D3%^4Llx>wK2TnQr2QV2-R1s&}9U-o+|GuNtb@V_1<`dU|~DTdcqr@!6CacJIe
zH%S?;Z|j;9CpYHEP(~vE^<PhyX2q1US#~~-zK7~*(1(JCxQ*?2sjBtTsgkCHn*on(
zKU)IBWV!~@hh4r7zs@?(wG8Gpl&bgt3HC1pM*Kst{Nl9v-xaivc>YU4%frRT^FJEx
z{~{lVT&+70e?BZx!K+`gd{+;%Vy@b}y<X6Z=dIFmiv{)pLtFZ?Cc%k~qB%9Hio3E!
z<vgbbo&AFS;px%9-<bhG*zI)$HHhpMZW>s8s4vQ+-;VwqWMzoZ0Rp#On|C}O=5zN}
zHKo#vs_VG}rWT)({sXkN3yt!0cGjC*!snmFmr~2+Z+49(f;MnfQ>-Okg~YyWPdJyk
z;#N9*;X$93SAq3Wli#qs&G%|e_GNU=KI7Z_RDa9M7mRO?M({0TiQiqGex%}Dy7QW$
zCLlSglUx4gdg9N=gtim4ooCVcXc)fdGwBQVbE(gH!donel+8or*ma^WA_dHwe=ez;
zJ>tF_s;~5&(ZUTMKlD4lzk|KZ@27jJgHm#w1gD=UxUfUH>EB1ZASBxjCOlv}N(zd2
zKJpszuGrdBdS{p5K6bVM8NZx2Nn-{L!jh!V&yiy@B6qJZk7zd@OMa56SNIvNp!D(?
zHM*lo)@kVKp-7dV+am&YG{jBa*uC=f2&5aivkm<<g?&ELIMxZYYT*Hq4p#1tRa|*p
zWCR^Mo#)c3mg`#Tr6!us(|fe6y7)8SSYo_-r1i~zj91@4yv<I1<q`d-R9dk+Q6*eC
zOtDB3SzWH=RkEhP1LY5G)8tj}z3`Su+U;FknUWW&*A-6-l=t>~JLwa6=!U3Bv2Nx{
zMZvHx*!`PT_+76y+GBLyo4Z<QHTavqRS6M!JD;+M6i7YpR=?+7g!73w{+f{0B4yE<
zb8y^2t4?>=klaC@XR!QFGj+57>T-ujbQ`4(`NI#?RSyD1k=51F{6)TIN#3h+iPEb<
zp_wz*9;4y6a`!V7tR%qt6I3P@vrb~`#;W&W$`bjtm-cr`X@h59x_{gK!GGk1xfPZ;
zr|Dz_|8YKSYHsr?v>}!5fP?gxrPnh#>ePynC*yHPJFIMvRa{mIMv?J|7%wXcEZgtd
z&k^!k6J+F%=MDw$c#lZq7~(yx*@$A9Dx)VJ__i!Cxg6!G{l6GHryx;+b<MVI+qP}n
zwr$(C&E2+*-Pmp0wrx${iFrFGVs1TFy=Bxx{+Zue-=T0vUsFd;vE*C3`I2PA_&xCv
z^^X)}hgPV%o{zc2;E4^doQ_ptU@zVYf^jO<lSd#!^~t_d+~1ji#~fg%0`r8wUv&0A
zhv9#34X(DX{yLt<cgA#oeLTM?+}Z;iBj5oz%NeYZr~^!BYiyX;14GR7nE}cNuSRCN
z4mfJ)fKCuSEe^)QMu)ZrDd}Wb4`YS^J@{x7s!W_aThg%*H~EFJfoCZx+q3-24BwM(
z!pdHl(!V4tCXWojJ!V!{hws7I;vO5IP8rDCNJNZu$=Ov;Bq6wCLkqKnTj#Nl;;xAb
zk5PO^*XzBKHxX%+gw@}p9cJ!XG&k@}^)L}h<7DAa6aT1M3h-_a{#J{lc*p+}>oeWe
zpCd3yJyL?_FA(Gi!=YW<ca{}u$Diis>ZG$rD(x=OHf@sTsN&`B3aQ|^Qj^gYa&4Bz
zn3&E}y^s~66||@zkC<eK)a(>GFYIySS9NfiWy=cLvM^hk3o<_GEF>}3UykKdI>jLE
zZbrV&rEsCyQ8bepV;z#SLwd&X*QRCboD)y}HoEOHNzryY5mn(80GG_Q`~G!qLPk^!
zoDq&MxJts`1%a8`AziBw1_!BkNrm3SO>&)(=e>5r9g!LS9zxjRdeE8j*qtMYx8+bM
zgmH8}Tr@j}<RJ@lFQ&03%hB<6BoShPmJQ%g8lSVrQP;E{cF`PluCNMVA6LEbH2t#M
z!jnWc-sIPP9Vx+OGg&ta_5A3>_0*UW{T{S06AVJlwJsx$@hOO{t^&%|)vo)qoP{Y}
z6bx>&Y+R08r^tZ~`tH08ZbkAS!+G8(z{c{CJX>!UgzFOnHuQg=7&c=0FwI^hdBWE`
z1LbA<wF*b|O`(h}v`ZQAe1}n<YqVtydER3vT{gS=P`sp~_K!OS__t0stmywR0`awX
zMC2F%RK{D=5uuBX@1;(Of*2Gvt+1mjzO7pF&JlWH{gaz~O`aTiL`&qy?KH-X{2BS)
zH#D;wi*R<0b1-i{(L~^e4ePgp_E!v7)Sj9y4P*cEQ00oEh2UgCg<MbDBi`(vkD9bJ
zQ$fniPAL9pHNpA2o)$xF(b13+Ya|4M;N%#=i>P2Wl?(?4G7|2mCK-zidY5jc!hH~Z
z^R~T_kZ1nMEj*gzL4EuLb^c&bTsS!k_FihI<xM&_v*};lSlGX$!qb}k&U8OFt|<?d
zzy)`o{PG7-bQt8orknzG#C2?%%1P~v+oZQA9nEF`+6lQUO^gavR+(<O1F*I3fh>V&
zI-w+)03$cL9Z@|BrI*A{+^5JG*D#ZSBKe8+mg@zBoJU|f>Gv$=9F1ASZg8cHZ9+Vq
zbFaYub4f!l6*;Co%Wuf>;u+Y6AAq-3`?*LEVAW5JQUiHdw5-XDEUREBo(!j=Qx7hm
z4W}{-`wd~mtKO`lK`+HsJNjxc4QxXz(vpO_N!LnMtdF3^g@F4#pswyN!gZiq2z#$V
z#IzBt680JKO0y{nj*7ERv;PCW!?wjSbr&$BpDSA*6gmQ<?(KlZochyzG}I}#PDnS+
zCoGipM|V+l&@oTgRD3J)DuVP2R{Rfy>|Z6BC_G|3Hbwdc!O-wjHw>eED^A`0k#sIB
zEl!e|c}eqZ>CuJ>r^6eE6=-d*$5hT!aC(j5f9VLX3!|!XiOV|G@+L(16MFH?Rw=8?
z{lqsH;U>&lTKCoQ#Z`&^wcEDNa7W8da7V{~7CDGS#@U-7=iQ_R#n=i%Sig#4n>0iR
zxdZWz30fn?*R`uX9~&Es5|*E4vx<#j(Pe2oiZyj~4v<t1zJcZ1<C0g1ES4}%lf)!R
z+c7G$^q7JT)X~b2j=NME8P$guPkJ{?wy8@UtN7^8Xxa&ZC-~^5zUDp3ko>beQjX8*
z4x}?=3TZ!1(4ym{7G7^bR$$tHJAKZr@m%TBIB(PPM$n^Njv8^K*eBl<2V;erawSZ6
z*;qONS-+tw`7@IJS9e>ijzC1Rap1>Fa!a%5k~#T65!Yr(PQ>K}16XXRTJ3W+p5e4=
z@PjpIKa@|vV^CLF1jy$(HmxrSf^|+=jTqMK>hqMNyuG4pym)nSZxZ<&+x7>-{5UDs
z57bv?ya-14SSoeTgfS*0D14P$aM?SV7vg}KfU;FY&B22Y8i~kYZxn@3G_$knaGO2s
z-1<+)EhClf2VE{)b#IAJT)ZRf$+w*)eSh%w=f`#))nwSYY>smi>5>W(ZPamBJ8iTz
zxJk6tg2<<_cyEBlshNl`(>!Dfe^wEhy0_c#$3*s>Z?w`}^Y{Z?g+71FHfF8qp%7x<
zY(M{)xKR47^*#7Wf;(!%Hsj1!g{_FE9Qf@U2y#))8vbpzJkH41dMZ-X+if4OsKG<l
z;y6gvd(Rv^WaPb2EymMZavbEF76XXfO*?F_4+1hY(4jc|#^|sdj%28Hw^9|Uu*0hR
z*2*FP%&GkPZoM2x9LIUVG-e{xb{k<|9H-36Q?xxl&-0@02v*4**t2B{78*23m!Tx*
zk$#<J%XFGV_y3~HG<tYKn!a@Afi)sGBQu4?Z<H;KK`+Q&0A2GuU*7zjjTj^MJ8vI?
zaWif^=TO9+gyP1vG=4^EN8O3uQ}gN2OAx>((P<W{B0XCtV|H%YXr%2J#ISb|x_GTT
z9m?7LvnoKfZ?S4pP->Y_B*3Lxx^5=0KqIEWq4bHg!oKF3Q4Q+`2tj%U-T)mzIA<YW
znlo+TRMVjwj3aS^T77HhfPeFr-<+}7ANR9Py6IgUfrrA%=P&=`M*JC#v3g+y=WG%D
zV51AQ1@5Tl7|Dinyx4eQXULhLkh;XacI9Bwct*u4jzJFBV|ifX_`tA)y$OR_{eb^Z
zw&DQJ%`t!le>JK!)eb+a<DkUwTTTzR)G}9{Zuq=Pc%wxDx^@|ImFvNgrPlFEL>AcM
zmaNp)=XhA@AQY@hvySXs@yOd*$d*k-A{k!zhNsv$gZg5UjdI#Q0#=g}%)L0<wrL*S
zZ^!9<y?putT|RE?S_la;rN_TPJeSx1TU_6(Y=wU^N)mJ4_Z$AI#6I}HjSftV|2aA^
zGqeA%Mu#44$#~p0#NL^@(;^UK4795DwPe8KSQa3kXyQYokCcBMBH<i>MsF)+?QCoD
zJ-o3_9_M*>jT)a;hP1StU&cub;aw9WUHshyvT7W#YO?tLnF0C|-{>U#EMAQO0|YYo
z{IdLf5*08%bf3hOh;Qa^+jpw`+r%7rfO-D69`=y0^dR3AE~@mO-qSDFcYNLPF8ewk
z36$q7!n{N5rxO%o{DZ`i3gNCEIJf7!c>dq7U%8XoK)piFzCU-@`tB35mcT+-8Q<?`
zbln6XfpW3{avrh^b^iH&9Usc1iW1U$1GqmQg+*r>X?Zb>J}rUA@%$5K_Mi9Ocrn|u
zKe596p?gW_w+sAb{LH><QvHC^f>=ea<e;|$SVRAM`fjPD17B(|U5#{3kM6cd14_)+
zEVA>?d4!ma4*Nj*wnGD{WL_jK5Y>`H{<xqi5L7i%EYdJELmq<#!)e9pN!;rL48nD5
zLHGVh-CB*L$ON)qW@Es6jr<J|2TW_2<>RSJ_#g_QLGT+a?}W2!2Z`RrH;wh3R!<&%
zs4AXH4@TI58G`#K0+jLnJMLB!G;?G5KzOtO`G|QB<1$}+d_gZBecZnYZrTr<gS(3W
z;`Jh3TYzY)t7gAcxr(X4@kV4R)~r4-!@>@X+;)2Eb9+#dqExC6OLG>|EG!EJ(>qwQ
z>H6f$q?-g*C^CAPqIP*&b0^$U*3$Uljs}FVsYjgsARDpY*nl&Ri3Ic}DN3Qhj07Hh
zKE>P}f7u)q>qg1wMceJYw4AB=*>auhYt~kWU}ajurQxD>j8{IrrW~PF@B1b)m_mx5
z38mFb7jZanldNpKGzBM1o;<0~`EQrkSB5pMT6y8?uN{|Hig$ODTu=}O`5c=JjPZ?v
zu5e^Y$F=7*QqwD{dz8Tg`7#rD3nUY)M;@o#)+XRKxF!9=GbfcMDr`N-P2wB_-7`H$
zaTV$sgI1EtQv6eqiYT_^Oj44_NYM-bXS~CN1)ez@YApe=6NlSMr(CAJH=Fs-vhT4b
zCthXZiL{=u8uI2uLJNuESM{V}uC(4nN_G3OajPT9Ow)w?Tb<u{_p;|>%3I(X>*7|*
zR~af3lj9b{JLLuIUjg*P(pY96);#$<{ovJbd%sr&R6#_DY20*f7)v9ap*T+{qiDp>
z2_ohh6<TZBs{J6UzVwF+2@eJJPqfK=CMJ5#ZI#=mp(UN9?6HW}Yfb*j{l<j|26IB{
z@pq)$ozz}}e2e*mr(W*n<>uI8v}GZF?At$4({K%B^_Lb~P|MP?WagnxtgoswV@j_k
z)^}rGy!>197d+>bWy5#$z0wi-NnfJh6&7|FsGsf{)RHyY(k(33WX9a@DA%Wj4OyWh
zqbPD-F+9-bwiuqUm~OT%YfmMs63h7NFXFhW>6FH}{uVt#HClhQrOHlS{FL5C%iC?>
zTi?5=YwEQKmJ7msX5@DbO$47&1}d;HzMOI;dmE~qUYItV`YRgL(r?$2=d!-7dn-BB
zTjLErg^l2Cbwtd9L^FdfTKzga$>FJ>-j{jiyG6J&-IR74#X<;cmZP-!vjj+Ul2)9X
zqNSTustjZc6-)^(<mTgi*#?U%GvB4#M%-2`yKSY|sffegwpotX@rF{nDDGLuiRa2P
zH^`V!SS|NBY-yx&;im=&QNx{vr2UwfRvt?Q`BM5JjWRd*wU(-trPLkeXVxq^Haf?f
znVTF!B!z49R8}&!`PtjsO3%61SGufR2i1EbmJXC?qBgVB<XGBI;4=AOD0LKjm@xsJ
zhXQR`rk7DE?(j@Ma&$n~E>Gw6((G!c$le1^rH~OgEy~6Gul@zas)KO18ROfzvl@W(
z>j2)yH&iInGKQnqI9r!!sh3~Z@v0E(ZH+c)_izbeXVT){SjI2l0vN|9`|frNuY~G1
zkti9IH$zv16(lqelx;9lbbGCXgjjHtIn;|_+pkA=vb>ufGmyW<I=+^8>YAos*n`Vf
zaScYM-`=|-#YPD%+cx2Wq01|LsWt>t+eRkQxqW>%gpIOBGRpRhXJ>fb9}Vu;8BV~t
zQs8EoPRQM9$5}g%*a^0nbtr{h5`GT;a$cR}{D&Y*6OMC1sL~4|Y5L>qkh&9X_kXHL
zf1TPKG!+%@kw%wJY4az18(lo%b~qK+8P@P7xLcQmy^#;Ji_%ey;(cuM*pY4X+tbB~
zp;FN6^@;hkxxKoM&g>XcCt1fh^<(4KV45Sar&XqOYiaFju(b|KIo<X_vh5YifY+mn
zPju!51>`eH%3)fRltXf8U&TJNyISRu-QZTHt^H#V_H$=`Y>idRJlPeaDaO5TS6w?h
z$C6g<B%s7+={^tIc}80fxHTcE9Cv!2R`Vt7=e8H}69`AYjxD5Yc7^hFLgGtbAYfeL
z9h(O{t!I}w_vvO5+lRbxm+mUtZCr?|>AP!TT<`vm8(w0;JdsHRh*8oT_Mh)({vEmk
zusz?(H;fZ4A#DZjqYp}dpRS5bHKxVR>?Bs^$8}m;Ycg-6MjuGt%)^8opY6aANby>8
zN9|?`NeegI@j`(};<;tQ$#|fxs-q0V6g>$AUgG>VOu$3r<d%Le!0&{6V(i{l=tvuC
zgg=wsoN);FF;G)vLZ%YY3F~Gd@wC)Wlp!3fi@o6GbWDr<CgDh430p+`4dPp7Q{~;+
z2N=H;UYu!LV}vSD4}Xil*b2g9P=OwTAOM`Yifc?FabT1hqn0ACe0Rc?`IW?`ci|-E
z0ZA)QX%o@?9+gkXogsv|11W)FP+Ii?Rv1B!(ya#ph+;h(DqkhG-c!R+AZxPUK8C@_
z+HFI{d#W2t<N+74rWe?;v5B}WM&Rxz_waLyh~i~$YqG$-pTWo;2ci))oW<A$F*zj?
zr1eK=ucvoi3Rgg$+dv=YEsR(%*R9(pLVmEpmgFOjJP}8%9)!TY7VzuvWg^J3l~#S3
zMKUi?oL;@T{F$+0(eWp-duA0lM$>SjWengt=494(D^XFXd$gw0h?cO_;XCD!T3a$P
z>U$AL)!N@pS=>X}ieMXM@Rs?I(JFPl%Rw>Fb`@1Ben!<~R!gt(NM*0a<wxJpt!h%C
z2029Qwxfdsv+IKUY<K@qPe~$EZNulX%48}WtlDLtar1cH#&sust!3jUU?XWbKoQ{<
zt>LJ5M;qgyq?eeG*u-UcTjzlMawV2Y_uwR6%D;eWr5#XeXUHXu%34Rzf`Sld1!sb9
z-H=e`*8|}XP2i+{&v@AJCEGjeGWE%m4$1c?<g0#f%s;&E0Bkd(ZcSp20AD&MI3D4d
z@jZE{CCgU&xjpk62sFD&wkLJ2&Uf1^vH7+`OG2R%Mr2)8Z4?}#*KZP9Ng_zuaNT}%
zFM4>%dTDSCme5%(F0{-yZp$b46?!&%lUGqTh?lq4Pi5C_w%x@q!L3GqO-H+Mr)-V|
zBk@|BZYtW&aV2R$#bs*X&Ig?vlb2h_7pQ6}xYgeX)v@q#y{H@C;tP$n3SFf#=AFIk
z65RdHY8FpkGv>QT>@{Wd^)-1(zmgc=L@G8=gPnzI!x1zN4m6xAy0)#;GrdV86L_tf
zy=oc^Buc+#9ftzZ@aQg%7qu45$L3aLTiTp(KWJ_@gmbNk3EIPDna_+V$g7%u#d<KD
zZtION`|xDhJ7Pk^+o;`I;enwSG6|>MkMRWu@Qx{DbG}saC+lXUWyKAvX=5>0+aSfI
z-`?bVMUb<3HMfpR{s-l=^e-AWO{xc5XRCki3joxQ=iC<Suk^6A>gu#rxk=U!vCyb>
z`dizke1Ly8qVdCap9v@4$9mVBcY>AxjIKj7-ag7n0@wAn&(&diYdhc3MrQKq+ykX{
zTu)E+2#>VhslO7E3nvlHH(j{Zy7Ts7Xb*k+lS#wx(ywe3-ZYx+$u7}AQJj=E=o=RZ
z1>CrJ@{|Rw;}(BV95F<$fOdL%z>VFyC@2UNY|z%vm@Ke=@-6=QdQ9{OY|-u{*H5!6
z%cI=j{^N^KZ-8<u@Q6=}9j-daaCRH9>@}<cHdBt5ezA;L8SwBfr-P>8Sh4I_t=i!3
zFYA}Mw^n7G0%T~VzFfMDWG|ElK-L;qwg*ZMZ|=K9y>$JQcN+<uQ<C5-7q9G8t7RVu
zI#YFR)3Bcxg9}E6x<QlApwc47sX*p-7d)#N@mDO5@8RcGn@r{EL0)c0by%)^q&tL~
zO6l3>Yy{Zl$Y@xfCOlObJ|Gp!p2(}G2ci!Mw+oQ82~x{fJf6j!h2fV5jOEhEp@*x?
zbt&MmUo*_Mo6i$AyqaTRjGk^cZh0dx(_(1V&T&Vqahp%PXQC_|)IyQp6CU!UM|p(C
zrJHtcr(?|(q<7ob3E$Rp%nMOGC-Lg`zdB{{s(vmuh}hP$%}eDvWWjasYVjOgUUE*i
zK7QJ$^7)?`!e5Foo#DTZRsR_=V*a1p(f@t^KPl@a9qD)yPQ>4cQTaiPg11fhV0@vf
zlKT_O{I-Uej@t{!rA`JgG}7%2B>#dNA2qXips=f|>6+m$Y)Vgbd7Ms<?)SDeIB^mV
zZ(l$crimh^$M@rd_y}C~H)ugP7M&Pmfq*Xmub!XBzK)ew&#bf~oB+LWK5fsz6JkgS
z+Q6K&4t^}sBP>XdiBG!yV`s_C^ZThkyo-?jLkfk_C1L&`{jL+W9K9@UD$wJB0=MC(
z$4Wk*|4SdTbPneg|9ji_M%fI!zxOGo$XF(t!`s3N_)joUf8Thtzz9luy8gG*+uPm?
z?-~<E86V6S{`=!Zy7bH82Z`~B<gkwQKe6GU<lQAv>yf@66w_*nAPFYXq+}vbV#f*z
zI~HZHWrvTwsVK?uamGTi6h(2WD3(B$s6UD--`5rVBjyo<54XH+B{w<E@$`;pB!v^V
z<n=o|gwGE##V(VW8#56i33{;lCFw~qFcmjL_Jyu{YDrXMKCD_`#RZeJ@|#&aA6@3O
z2oZm-D5d#0wG00ToU1WpiKh_#i-wSN_|?^+U%AON=*qvJ!xH-Vydhr)lRp_#_OjBC
zaST!mqikUH0L3fAl7IW9mYEpw7Zf=QX`B{P0DC`WxtOA=(o>0=Vg`go*+frMFd;^b
z)VOzO2I|4a45d7Cme{K4!MBUg>Us;qtL@qH@B8{)$hposE=qC~cTgQT1Qr=+9L0Wf
zsAPYNo3C}!C*hHl358CPPG(AGO!Hl`onC*~D{^PqsJ_;NT9l@d5JedDHVH<b3O}68
z!m=Xmk3l%SGw!TpDQZfho*Sm}vnp=ObVk^6m}N+B%l=DzcDJ6q<Yj9|H;lKO1HWj|
z=(vb{;$OLmVIX9)>U7^|9S$J{w<bLtrn2eT=))*VoQF?KOM8wHlqPHqwcS~)%*1n;
zG~`i^85B|qvOO8)$po`(N+}eXeaSQm$OWv;YP1>*^fHY<i7-bR5epPjTI=8fIi^v|
zj4IgZ`7V7i+=EDg(qK4jR+v5n1c#DVV1fpkQL#M1AYouLyKD9<4*UDS2{9WD_mh56
zQAW(*)M+00qEXGSnQ%a#)7?w_0XtMfng~FjY?N1G;DEzS)-w(W2C#zZQ#>EE_r)!L
zKFP+~Iz>c1bT61W)cm?o&97CW0y<~*Xz0`C*KcN>K-Rr{vH&==;0;yK{cZy9L*^{&
zMzF@IbCt0v1r3!ZOObNWg)Ml39w3)nR8|gJX7}A-Hkys6uP87#nrmKc{$NJB9NZsB
z6rY6X)klbNIqtX-AV&Rt)t~_n&>CjLWd;w(YK_q&wllwNnw!L6wv1{M9yLuYM0Db`
zF$e9)F*66<vT1D!GJwO@5@avK)Dq(?Tf+=610d!`wlCs2pwpS3N!DBFNQ~hoGv&KK
z@as4u<ZMh5hQTa&X!0wE2y|GuB%`__@*35fzQW^loKq51emWe)L!n@=cqN>`%q!aw
zKG!q%tJ}8>Ji=vE8DYww5|fUO(JR8Uuju__tV^n5`q`vy0`Bs@#qZqEN$hOq)w9W{
zeKAS}?4?f1jKSum4GGt_!r=r~Rn(q^DPtvlwx{q>KPH2X3|on<vV7<*p6lNJf^yut
zkT1PuZrWP&@T>%L9uTC&2vtI+qOVlMX<s*~lvy+R46TT@O%81rC6Eogg1$Mbh)w9!
zL#vE2++@fholh>IBvR1)NaFt*ssAm*?<XGX%l4RGc(%d>^Y`=V9$bG`o6ts5D6igR
zzMOR6Q%EB35OoZW1Oew@d`sTwU~<$)V3$c#sbVnCjzVIZhH~!?7r{~0ygoi1)_Suh
zkGiT+)mQ99jT%(JiZp+|yzTKq+E0pu=V!XgYH?yWWI_w`-^Qw?;-(#BokiS$S;BjQ
z?3T3A0TWT<Vc82!kzbD;hh{;F**dB?u%zsv6n_7GnGJz@p|VhL>zz1*U2tFks{CMS
zHJ|QaWS{{Im$G_gGdA}F53oJcolk!_`#pW!X6|C;u^5%G00frZ_Sc2)B@-<`WR~KC
z9snS}h<tYVNY$kGY`vh~pkgEjt$CmUF%M(-sEST>!;r_59xF&Oe@_r~Vzo^j3bjNJ
z#VVwpyMw|6`4BlVaZ&l2iYAe6>ZYhzWHc0-0Vda+g4fH9C@3^23XuSRJ3ngh^kb`?
zi;<{Vdm&glcJ-Qpo4a8by>A{i)L3p<`dCkQB~b1miSowX>&4TeOoQ4MPz&Y~ZR9%n
zv&BV+_~Q7ARxIAxTL+ittr~PumPuYpMZVW?)6>WTuqdF{s<@F#A`Lg8umchFrkQ!@
znr0~90vZ)5vs*wNQQf!?1y!!PM+_Y;|Ip>+Kw;HrZ~!_dh7jHRjyC}m1r3u|$P7hZ
zF`!@Sg93@$aq%bENg>xbU+<b{9wm4@?ojH345#7RQx;@n|GOwq3r-k<$k@k8PVwwq
zF85T*epT(0{{swWhmol)e^Cpw38UqjD;KfRxD`{#csmzk@f8Ibgp-&Hc$Pr<fIa$F
z;A{*x4|C@OBopoB+Z-+c1E_ewYYz(BbaX3-sNR0|Ket_KB#u72^n;S0vQ?YEA!;P_
zIGz}uqa@d)4UUKinwER&QB%cF{QDei(NXXY<N)zOg2XCwnD}Wsu0h`#q*C^&&MZ{`
zoB77`<1EfzFC#|~(geus&o(eiBKD4;l4Bt?HbUNssU|0&f?{q45nNCv)7UulDA2yy
zAxh#@r13lk&G7a><eDqZ_!dUEzAf!08^KhcrycCJ)>)jut(tegS18iG>H9MI4%R{o
zn3~-BCe2d;CDHsnyo^!ATm-@@(FlZ7+;6?Z%)-d)bE=!m+mSL$GGRcQA3iqZ_t`@o
zAGzRDy3Q|lB3KP9FNKn<?j1btmScCD?CwHra?Me6Y<$d7D^yp!vg!C^Q0d@8`q9kp
zkYmu|A~+>TX0+?sn4-xSS1flMFY!VbEe$acs#RvysFv6yEfa}2Y33APf(F9#z`{Mr
zT0h@+`1m{gTqj|wK?}o_!uzFXpTz0yJc@vOB(4kAqS^yYf*93j_l}$|3B2tIFi)$&
zJ4&k%jdFmV=J%98K%A%DOC3$~;z#nNDyYIANui^K&IPR+cq+(mT@wCNYXc~Y7*q4h
zLw!_3tcXnEIbxToW;z1D@mbiw&zyObvo)`UN|yRY{kH{<W%>*?%bW~_^&5)&66jhT
zSN6dhxdLgwqV=$=QW$ing7D`&$f8?Npe1<^DmIerEGcgVYJFQfr~H4m#Gf_Cf3Tbr
znU9gQmoNle-xhPKs++i=bU~ewVgZk7{vcrY7z|SS>zx71%<65q-CN=dxyQ1mkCg*T
zf(G8IN#J#srR_m#`T)MDdTFZgl8Mq#88Al>#$$x=?{-utQ8Az97cZ&R8w?7MQ<mm+
zjth^INnFuNYD~C*-pgQjrdSK;CauTsGB-+<m${%<aRsFLYCuSh!W0jV!xSIjuBR@$
zu1CrrTKizx(GN79BaT6~=jF2o3r6Y8KK68z$1_u`<k_XBvoXf!nB_KObqpLMh!+0Y
zI4Zmc<i&lCph+CstuJIJ5q)S+ttn=kSrY8C$sSU({%*Vy`3>UCI&=grpGwkNaGNHI
zsU0UqVqW)mR#9FTH<|-UufY*?m~9=N&&eE}E3Oy{TQG0uOWkno;Ca<bQax=bOfiKB
zL-sCi)(&YcGfotu&su1UZK*du)(=&x`6wbxaHxeO{K-6ZTBFUePwLFKC?BWL^AVt3
zP^1!p1(y_L?;LH1x|kO#UlOr3s_Ea4T!0!6B%59vs4m0BUz>W4ikm$~U{W)Mgc^Ac
z1{M4kBV>702f3+pCPXiytoR4Bn83&x$uNYa1E%Pthi3o1Oas>2B7})DTBw)N?S6JP
zc0rI;koWqyU9*Um=Qi0E)0yNgP5ek5{{TyEpRqMQE{xE;x9SZ8Ajllt$R5Bsp&B*@
zg~g}-XB2Vx;rl*K4W{_~Vc`%NCPKfYlaz$XVQ0V9UwZBRAfaQRB+p%?eV=yHu6N&c
zpRd796Afu$Zt8tZ$&g6(U#DWRm2;TJ9_$aUPM1byyN#^Y8-XvoQFS%wWk(UyQ8J?m
zO;7bQmGa%B7k~1ZRY5_9bpI}&sW+gnJ8OMWmlEuy)(^Z2a2KX+6_{J@v4~t^15>G+
zs5c$J2k^Gq<Ki(L0JRoRKvBb+A81<&91az|*sNnvc%7F!$_bsBx%ph{wznal?!$ID
z`o+&mD3>t(S=0+#-UI7(-brq{)IRHtEzQZ#mqX5x4{RmA`|+XI^Xy4|PA#YHA*c8K
z0S~$S0wR~w5T#z$Ic8?ZR)SAPGE6A%SZK>Xvuh$#_~qE{oxqk5xm*2cn@dN1+J4lV
zB{sg(?p8~Zx7q_N$y#gVziZB}pLk9>*{f527Yto%wys6@gFkZnEWne~w%I~|#JR-r
zA-n99CEnGMMo^QkP257@UKd?zDvgWy;aooxGEiZ{mD#bUBWy97;b~zxw{^w4-rb{e
zet=JJD3$)(aK`+fcV0{!|J%5n_Eh{qC+y#oy4wa`5o*PqK7vycX3X6!H*SygaCeH<
ztt-Q_Q*Yj8e?LiI&9tjk$0Q)!zGRJ3%5(9kVv11r9q#R(tGJskdGiXci5xVv&KLXK
zK5(2L=(t!hcQpvJ;?r&ZPW?O%@Zc51{|CZe-zwujD{&Mf;L$s-YLni|L3I=O4Pe$O
zzhA?7d*Qu?xL+ynmmgr?;^ax>S8(WoQKlwky$f;$_C%iy*4^-5d0-{=$t(Ky_TKny
zv;;G)GmNReJQ<6_r=ZAioeD7F*u8x0blP{#gbM#wy%pKN{@6y9uucY&rv3dK@A|tx
zn4h5GwerQC;Fn_O@ly1`>{kD8l@Sj^S|#A`zYO=j?uSU*86oMJ_WPA44a!VFvrW_5
zAm%#l*8a;Jfu@6D3tk3p7WP>rPwV4EPZn-)6O8YA_MDioB0wq@_|UL2(RUGN1q+>V
z1XO}lA>qYLpeeNtiNe5jrc^(<n#acs?NF*2&wc#7kv3GnlV7d{RR~y_det7(643i9
z^KPH4sy%IJJTky0HJlc?FxQcKGZ}D@X7n<zIC358J*?P!ei;)N++A}g?s|03<N`9C
zwkB<R+WoQDNeGYm?`^B+x$U15@DUSQGQEBx)#kQ;*_R{W8*eUbvz{!@43&7;yDp$`
z!3|$~C;Y?erF#J4%$v;)&BM9>mca`wRX|cvEn4w#f*d3jT!v)0*-;+((LqITruzh`
zDL`RBKPkad0w^PG8yUWIXCtUAl19roq%ShHyn{u?lF2H|-ZN-}gJR~#^j8JzfTu!F
zI!E{52yGDa`?ARyX7(`m7#h_JGid#4+UA&JiNj|Dbt8>}$I?b0QeRHLb9SJmc*s#a
zZtw>)WDAvAAP@0?boZ$&aOl_S#H75)^pa*C$T$E#J*=<E25W}j;0V?otU;R^LpH*!
z?^tp44YR)Y+J4J^WGB)O$$0YN$WKd~&zr%)PE(LBp$iMA$<2UM?`6u<elhg4MpXPk
z6dYuesKcqB0R}Ky{cW#BnpTK~aT+}@P!xxl8IOQC9q`}R$B7Mbo+OrDg)-DKAj-0+
zOGPB)Wph;j1h`6bY`u#iGlZ_5Gy+Rq3_(&y)swMG(_@TuMiU9j_F_O;^j!4T3tPtb
zD`urh69=m~bnCWlID~uq$W2k#nS$RGglm+2p#28RGV^Z;#XE|L%DaPZdO$Ky&_<iE
z!jWhV^H5lQEa9AT0Ou=#V4<k=>%g^)lHyBO9R?TekSs!-#n02QRC_f0FW#4!4#i+w
z_E%o4K^`{aW)EU#R(B(_5>K%?$m$jfKAB=;n`FiilTdEkR)a@KLzAeVf|A1^?ad#V
z8a1T6LVMGP;5t}6;%VF@e4~eIFW9v&6W-CHZD+fJhxBm9$f9i(^y*g$pK8#$r09o?
z%pV(bB&Lp`*2Ymg{cOj}T~Zy&3>Hk}t-WxMtdJNVNk!?ReKs8Oq|nt>{ZZ)B9B7?q
z&f|}vg8%gTKZ!fXWN&XbPXEms;u>QJA6ED-xR1YY&jV$iuKdB57T8GCng!s(uCQl}
zwdSnOhWV#@>Wva0mJ@2M22SFCk#JrEUEI~DOZ+iA31T5SK|mbBdPqcalymCimjC_s
zwUOWZEz00?r7se-fDbEIV1s{#rl4MKltQyjZC34ETBJZ4d!Ayr4v2J!E<LznS2n8*
zCCQpN{~oM0N6ZbX(1U|eFOtDsp4ABRufbmS>#H0)TBRhKBZ3BO3o-qXZtK^&P1{*g
z)jL!Cy1di1)n@m?Sy&0HLFCB8^bGb>s(^~&XJM#Mn4R6npkoA;s(6QHekEhgyPcPU
zY0f}GB_Q3$N!2y4jp=MxF`S#7pnjZ?9c3Z=0RuPn>m{6t&UpTloCFxMON7*kxcB|B
z692f(pk~VF9hev~aH+TVXAImLo(>t-zY-~{9NBTYP;TtvO=-xQ9R0l#6K{3zq&A5h
z&Soe6^1+H_Hs88Ny<kr64GJ`J0XzKyGb}U60Z-`9G^k;wl*>Cn0<<09P$Uj4MA?aw
z)x^=diqeI&Pf4?T)Evoh%;)ydy!~MdLM+l!1>JI}d1DDK>bG2BInK3GVdtU)80kMM
zDKHe3p*8b2dUPn`nL#Qii>8hcO$8ZVkhAHtuIp$-*$RVkJntoWOcQ&ZIf9=#*fFKW
zRdqw-p)??=$m>~jXfHksc3mo8Itkn=#I8;pZ42FT^|(;!ZIg&*<!qQ9(8EUO;w9A5
z;6KyVOSg}2^QS%ro_n#Co7xvq95%E+`yQDrPSS5LMwJ-p&`9VO&4=UBNZaHVkSR42
zw_y)Dp>V()WD>9qIFX6x{beD}bF*Qpg?$P~>v-l+_MWh7<L+ZZN$C`CH4PLY1+y6|
z4#8=WtT`<0VNG9AdTQTd>e`rI^_om!Y|Errc7+a{q4xS|TN?gjqj0}QL=kPZTiTKI
zDVztYzegPE$bGIAXtYmZTCflW6>C&#n<}+{vy%f7{X%l^WF~>cPn$-iSPzbdU}jKX
zVHVq;5w%<#BhhADhphuL&iV;#F|{wMeU$KQwmZ(p2`Q@(T5bkO2zPY2o#h~S*LGU~
zT>}8Mw$R?UVDwLVF1eXH)uL=T>4oXtGhgSp<+6Uj41IvL9@vz}R1*QxFbCS_w9ZWO
z`C-ZJhC-#3%P}V2Uo7_MEm&T;!(iy6c{DB(J*xCr9S$uFS$XTmX&&f3#jwNN=R+J}
zE4)ueuuX0)dO2nuv5)gy^#^9Uk}~IvBifwd#kVp+7#HVongJ>}>F7=+OS705WtYVy
zWp{Aq&jCt?__afhblA|B$3E*a@CY-XvbGpqkxwKfSn9Z?l51F3dNAJn5oL)0>?ntz
z<owIE@FC7I-aR!0@OXo>#Bs(#t9u=19LGyEYEog$TD;|W@XISBX|K{_7-Vonvr??n
z0*a?*E4<Jx4^zQfedD_}ppNXlqMO_gwD(q`Pb=V*>dLGpVr{KbfyS;Dx!Y{e-2fp%
zZNo=Bfdm&}3BDZq)1A_~s*g!cMP6FXS5E($aT<$=V}xKQ`NHc&+u7i)0qGzoDV_-0
zN$BiMIr&MefQoA8FDinRp=KAr(i$#uM>xMo`u+i1?S+FZx=B@pOU}$N!WQC_^l5=N
zVHScO23kQL3yua4u;?xil?t9|9!<>|Rr8&N-kz9tDCaWE(k5SbQ+?4Ec6csATkLbA
zT?Wshq}0BICqBApOr(`A+^b~1UwZtEJxNuw#h)>$4UA~`-mJRW_loLGKH$8>N7@bm
zlO?+0Z1x>qWuCjZ9!0-*ICHvAe+mWpFd8@r-x+*01YWh%?Ai*q`;V$1$uSV>{{2w9
zV!!iSIdA+(7i0P6|8PMyCm>BOP{tW$RxZUz@G~G&fB%`Xpb4-sqPoLKp*0^!Yr!SC
zTrRB2Vff*8VIQmqsEex@G2UIu16<@Lg-#udX|Hn=*Yyc1=rj}?U6?@=eFOvg3gal5
zK~a3AUh2DZqE_6F83r1sD#eT!)Krv|rop(Qx-|s*WQ&K%h<aSp0p2{4+GIay3h0V~
z9NQv(&3!}j6lAJ>3MS&fAWO~B$UEK<O!V!Zrk|)(9UC9Ic)p|A{Y7g%XR`z#rizuH
z=n2CyZ8DWz6X#&>e~T}YGfFN)4+a$wb5!yK4!ciLMx46`BH{A#Ye2l137`d>XKp`e
zMLVlBtE;4wLcW<cxWBoUu_lK<4Fdgea%n9urjJY#aXEKiv%e_GbIQqc9DFM+yC1gl
zzuxE(<Zc%J>z4aJ<&(eC8K(c$eaMs+cQP?M-1LdMiv{=usYFnGC8ermGq?89q`Pu@
zJKH^V2kdm@*8^-4LAFWTqLQ_XyI_^R4Z<e`5Wc|J*^8u{exBZT9$&8&Gv^*+PS&<x
zc7UG5JE$Z)tsI<y2JFH{+gX`8WopSmC>(LgvI|G&v&--xIBcGlKmjluehy&gf<XAV
z`IFR0WZl=EuuuMBvHcda^uNzgu2Occ6tyHcY1-I1jE_P$fu6yMy^9Ta`J`|X(fT{4
zyy31Hd-)$vEO3h&wE4HQ6MOl)&^dx;bz}|1=G^aB*Y~Tl5k)KLeDEN7t?1Q$@FWIS
z5?%;gebd(jFf$3qhe)`fDTEw}U<MFqFu&5*EHGOLrpLDqlGU>X8e+c#c*Ooykvt}_
zDlMSgLe9D(06~lcF_aWE69CXqBE#YJvTGjQx`;)=5R$vHI;Z#0?Joga*V~2silesm
z=$vofFXlhpge%C?inXY5y`2Oaa#A4p8kaOclBAXcS46doGKIHN`e@Q&hz=><b4dae
zizGY{opfm;))d))Bc%i4kn?SI{7fxxonp+VR^sXonuxQfc*>_asC2qUT0+y2@-}C9
znLVo_sBP2#X~lZSxa*LpziPlxaboAKx8EY}6VdxIc0{B%AFfS^kq2}-;x%%_tPeOt
z=Wc=T_gk^$oHEZh?K|m7%da(<jHDc<IbXt8+PKL_yxE>(Vt24%E{~{JkvRdWX)eTR
z&v~$}DRW=ne82PgH^QO93K5*Dmfsd65f{hpEo2dI!U0(*7C^&80i6a+J0oatxT~$t
znm?Vy@09J0@%K)8S(6H^DKZUhv7t@Z(87WbFH;tDcZ)w3RG=%-?lM!1T3<!uS)?$)
zN8rslmVf*infjmgd26P^L*h;7Ym|Kya0BVpZ1%y|g-^O@U_xx4WllYWD5vxz(?VN~
zK&=yDAHkr`UTR0we~$M7R&)^3e%Ovks!E7UNgMc688OUp$=M4ua$t-coM_y1<UV>M
z|7q*xPpbF3kb%DoWPTZZoAe99b9Y101M*o96W@A|w4?YLW9uQ;-AQ%3#alzu)55fO
z`Ha;N)jsBts0IGe-=QC{fST)<s&xUcCP+Xbf|2L1d5g>tI<KckM4G=NNyY{z9&@ya
zIx9u*|HdgvrzCl8=X0v%OE@(zTq}ifMLYP&N;IS2c^&9;F3E({#k1xVW8?@yij9v2
zEm>(j6$&WJG;s*ThuI6~4ilwaj!%G@(pS?R#R9aSy-XJ8yRWi_Rxbu}Xds%Glt1kw
zFa6{!DLmDvDOr19J=n`gWJB**KmSR_npv0veKuz+!a1etn5QgK9g7QJd`u;1T4tqb
zZLdlur9KzuV5jm3C!r$BnH$*j5R9=u#xve$n*JPVtPFYyjtIpo|5b~^EGJNyvOcn-
zs+&{Hc)pQ5vXq6K-TUh3hD*sO+(h=bM{_JO!xM!#XVlU29_Odf{Q7%%{~_KRST0Q1
z;s?XNL`Oz|I%_N}84pu$PhnA5-bvY&>Red$E!FM@Zw*Xm9pVBn<=Y(lHWHI=uasxR
zmN!Gox7y6fYN^<cRJCQiJ^&T+=#h@yq597?Pg&dR;HqoR<1rZ><M*eR#;QXtkz@$>
zGqZ~HCiIT(qoHEjFy4&0o1tYca<=%4vWd!U9lot`5gPFf)H;aH7R>O}-%F;Qx`nN)
zzSK=!FQ9ur!7{wFlg!R+z%$RCI!Rm*=gaV}NRr>$f_^bQ$*U@@8ddOm?2eL~Ghf8;
zo{{T)1wL1LpgE)@^sD4;)A2UjXXT&M^;2>}Zpq3%oV2C2R=)7@w|%k7y`YZzJHOnS
z!pWo-hWS|$Ys^DE&!<juHqCQFJ2tQay)d(yh6Xv}&_DxAh;VT3`hKmCJ~1B^7RV6y
zXRdKK*QGoSX}t{f#z*|2R*F@-R`L9~E2?YJ+iCA{$;u6FCf|%~wx)LTBP~`*I@yWl
zisQqVGE-xzZ96k4bq_K9lC~bS2TcXHZZvBHvvq9kFOAKl@-vlfdJEa4Px7GCwQio7
zbtb#M4`!GC@x@WL9p7gu>x$Zyo4l9PTGj=-UC&7$9#c!=mIiD5B}+9$LY0@v#>EYm
zw9P>uCJ3Fj<cDr9co!RY%YZ{|zF&8h+Ac+ZjW<3j&=$uF&(@TilH-mkCW-M>O=rT~
zRfph##uY=NZN(mIhqZ)%@~)0UA+u>h!O6h-U4Z4YZ;jLHFjHT?YV{VmFe`Ou<B*3{
zxTTa#X1IBHruo~`cq`&_uP*1U7KrhPPKrb#F}sgFsHo+iWDfKu6XeZKCI!j-C}SZq
z5YAXCR`eDO5B`#dYhf+wl28~{&%yMq)GR?HpRB;z!+EfG9WhL(rp0SW$*)(eLn<kq
zJxnjfOA@KVKyR*SS4`G~=~BAhdjHwn0`b_6i$fnXA8ZgL5DyI%D>s8#X)hEAU;2xd
z;0uI$wE$hB+f2lEVRdngexO8dbD1G8Rfi&#iIUlBjPYut`8IDSNqslo;ur-hy6x%h
z*8%;t%H{MXasl!fHgoI8wYr&aE|$5z+IrqwTJfuIEhC*l1mBI;pQdppvPJf?Ls%&o
zt;_V3O%s2Z1~CiMVT66v=7wm*!z@oZMKj~d1AqE@M_=^4UMtQ_fo)4l&D?B^+LnB$
zn+o0Z*QpbY%jx7)>l<H+u8L>dYKV}Ng2Sjftg53ybgNML!&tagbiP!0BTjm3QBT^`
zf*jM&N;{~{JtPjE0HH$R5qas1&F1jSdU41LpTp7sV?mHJv9XhH(|>k;7Cqs69K_YZ
zib*Vv{O6x}DgcH({0l!ZX^4EY_bi8`S6x9vgEyr4z|q7j<>D4?ciV#0q>_9Csalk0
zp~=uSZVsrRauiZ+V3Ovga&sx1_w&YoZyIwwR?jG!n}OzN3%f2U2^9J?0r>288c>J*
z{W)GBs4q$FH4S&XR+65G5Hzt#a<wb@oKu{?wr}Kut$xZK(hO|fS)iT;#2Cx$bt0uc
z%NGp{{mEspEVM4^oH+6#k-lKHCLx;t+YX9_;Xi`<Ozcby|EnF;|7CpJueSB~@G(SD
zo<P80AC0X7dQ2DEhVqgw^_aphN&n)Hr<p`kZBT9W^x<tTnM$J!u~ViU-TrFMZshCm
z<^8(~@p5j#a<KUPumkWU-l3E1w3s;n3=zZP@dxSkOH|To)jOM0B3wA%cNWS#zk_xy
zK|nfp-=Y@;b`VaL7Roy3p3=TwX7u+T5T`j-d}R@)ScH3G<QWbXljx+0Tf~r83LOQi
z1~(q+@cDVinn?@fHGKR&Jl=1RcT!l@3Kj?X{dtDKu>=d$gM|b)aKB$)ZZ{TophI9Z
zq40l-{_flax~O>^5sm8mr4If*-IHg*2*&?r|8<h3B@X(B`OgA*2lONm8<2p7(Dpeg
z5WOb!O%uR?5&p6PBIIUjkQZ%*K`8wW^+l_m1P#f{Sjv&Ok=aZV^|25jJdwF%fea*O
zVZ{Q-3#^bKk@<dJrIOVy5(flZFP4|DZ=XZs6b?xt>VIY2A0&*CK(fU}HOS)YEord~
z#a<m#{wN-kZpF&$Kj_XLb;#U~oGe^MB+3wV71)uMhT4lmb`P>6hrl{W{lNpBs}KQE
z19ugPY;SeTV5_Qj5H!ae{8vJQdH?(=$~oyn7F=TU$u@?vYn|BgN^;LFC6zLH_u!h=
zZJ7tz)Qs7cQufL_h+#^S!l`+!>zQXd##5!J<Qv*wx~X>2O@gz-)H_9IcZWsFYH2Mj
z*-W9}DK{o->$-)-UE`A<zAjOiYyB%fS(m8;%_V{();soeQB_P{C}+4?Bat!FidB(y
z5R)25Y!k~9fR;C7+cS#dF?~V1g7&dB4cR<0fe*q>4d&}@vUlIY2qWNWAC8qof%1Je
zhOto;k#r7;^gsQ}?}xvdo1Y8^>bWDA42R8|Q6Rx9TlmwurNecu3SSTmuqD)6r7ii(
z#9vHhh^XVfr2NEa*Ia*o_9Un!#=6YkFYWQ_(acIP_d$Y&%_z9v#-f?t(;q)MPRcyS
zz#1jgeT-_-k`Uyl;wTVy5P1@PG2w-XdFL)n*?9JnKb@2Z>RH@yjC&tSuvC+yx_8$h
zR-GqMo{}}C$Oi0%mFdIAAg2Klg53cjf}e;+sFV}dCBX1NITZF2=agx(5gtCmqBLm&
zF{|$_G${dNMus8r4RSD!UJQ>>cP|SBMup`N)|=*|Kz70U47d8u(}Q0o6Y8D9O_GGh
zWqOl@6EB`mhs~k{54$0|C+$}%u1$7#z2Yigrv8nmVb`S*ABXzI0EQyf&NH5bia(Qi
zSjv`Kp9Uy=lvp#WF`T1fs@41{jrEE)EhjG|trr1=D^Sp^{_PL6jmVlGVn;#-i(&5Q
zE>uX?&aqfejI2zVm4y|PEL4FLn7s_R8N~@Vs|DCn(B5d1I%sLED9m@`9Z0(=`TE?&
zZT{^IqVJC=XmFm9vMQ8qI-$-9Vj8G06Ss~k3WAtsZ}#UF`S{bV+pMkyS+)nXZvJFG
zF^;N@7{a{AEFs@eEOJWP!yZ&yC4uiC<Hs$D|EbZ%Desk%B0DN}$%?~<M^VF2ib~|J
zAK0Ww9vWrjpQu<@$f&%E;oABPS)?c2IZ$qYeVCwFM|h3-<n^dov^|n;vpDlWiv3i}
zi(WLn+%sF=^mT*bD)x~q))W0aRGwez%XV(Fc0i>F^J7X?SdH%T6cyey!1;o?pE%{2
z?tzl*8|h&nQ!nTm1)ii>)_e|hCjbGYN^L=oBR_q;Du+kUcbOXI;<CKVqkzSwbU&*H
zNq<F-qAZ~j`tptZn{B$6V;nk*hNe>fN;7QGRLq27&Ti0E{7swnCSR?OT<VCLHk$je
zd!wRqt{~jMS;~F6Q&SZqj-vCZ3JOGx%UBp(qpH9_v9$Id`audNw}X9BFDYBae2h~*
z<*+Kv0`*TVwXG-zvIn;+%NiCPT{-M-b=aXK?Y^B72bO>$7F(CQ6Ya)}de%?L$$@;H
zIRq%b`7)m#H+Nj)#NF74x|RLm!1iHzmIZy_yQ7%it@U7uDx$Cec1_)1oONSTV|$O;
zf2_$yEyME_bHsR}Q^=Xh#Zkf$JHj0@Z0=BST%B(H{pO?Q*!i!BVmiu02f<O<MEv`$
ziP+9ERL7;DT6FUzi=i_Mb&kT2Rd2020uEF`%+0fjXuSf;@lVU8Awq73A_Fa<k6atk
zdWP=C2vf#)oXFBrB80z?_`W06wPSb+gnZ;wFaj6q5QdcMuDq2*%n(GMQN2<A<LGw!
zrgknCVjI>$z&--_sNq=pl_m<Y2e{AjCQrKd>OOAj2HX<~&v6Y8Zu~NIYZ|7`q<osK
z*TbqJhJA-(BTvQp=`c6t6_#>cMK1kdPYL&1`;XdjISYAJtW5q*^rCH{oF&?FtUn_m
zo%JXutbem3JjDuBr{!kw>NnMr+rf4zXO<3<A#HU&HsVqBX(W|6u^gD@SZahaG}b>b
z<KSVFk9~O{F(Gs`T#7Z|n!~p!Yg*ZgzL~oXq(qbC=-lSG1OnDzBytt^>wl2r;CshK
z-Z|NF(G4roklKZpD1j?|M5t4|=X6mJBS^6#D_)#wkhq-Mh5$S9>IaeG2sq#tY%^Bp
z{o-+yrS+K&C~_qvoOov3{Zk8gWZ>0W$)xd0$nKVIc~gMF-L`PAyWU6hIu~4e)03IF
z`m`iFf6To3zm>*ZQ)!|zmK5%^QSwFUw(|K78=1o)kp0?y$4AUkpX5^lm8WivI!pHN
zv9{zglrM&f%S(7wCo&8a^X#&}u)MW|rJY8Rr9>Icbi9}9(BxQ{S%(sxe_Z=jTv=N2
zZ<jq)7l~-HiqtIvqyid!9@ph4gkYTUZ%@7S;;|F3*5x8Q9isPcpI_fAl!NN%i!k-i
zY$e=E@*FSCA<Zv7&bq%C&yT>i!n`iJ!HR8W=sj_fmz3Ml@s{gfIP$3iaEgF)W_(O8
z>>c=?3hOaPyn8RiE|P2hjZ+$?cqQa@z}FO>bkazWQW)uu3svjs6GO9J>S-Ea<R*Z&
z`i0XhGEd|F8Y2Id9T{42c*c}6#<6sgT2jM}eGN|n{2#{7DLA%pYrC;++ctM>+qP{x
z*|BZgwrxAvv2Fi3RsY?0F3wHgb=T@u-Sy5j=X}P%THw1L>0dO0-HML2nDg3h<U2Mr
z;9-><g@xJowGHN+4a#^Nwfsj}Wu#t(hbM#T#P%*On$Q3a2Pu0<tW-Ac=2+lF6_S=E
zw5ysiQ3a7?Tt6@NmXNjkZ0mk94C*b>tJWSzPg&FxM?Y@td{5dAH<KzRxUo9uJ(MP8
zxmvFGhRzI&PhS>IzzOf%jTBPPs`}QM)>UnKtq<u}%MrawrA*w$>ebp1NNAzx)bFMd
z`gC<OI%<oSnKK?NKxHtjH0b$sau58pIUs?<3qh<RxZKqgbPZH~p6|p!P+3kv+xdwD
zgL0(2kj**gj`3T$@W*KREogBAI6n<;yj;rEQ{Kp#d`JN2{~nv5H(64+yUTFRhTEfh
zZaXTs-59nbsRJA#xr3Yh9wS6bWg4C4{>+P&$Q^60DBCtf984B)H6=hIAgZu<!Rf%p
zQ?#JtBr*Fsx~zp)`=EFXJo8Y_q<8C1SMle!CC@u_BnY=o<|1$ls`P5(GU$>1A3N(>
zj*~A(REaihn5~aJ6x^)UF?Q(@FFRyzs7FrTzL~yYIJ>IFT+<T<=Pb5=eZKLBT8YB`
zYqiVvf2&;<*8gpPx?5Aq?tlZS=R@tZ&5I#(m_49ReKQ<*s@lv6d&*$GbKC0_=GY<a
z>rG4K*JqbPI~-2#PZSx8bYav@$?1)Kc-zI<WzC!2gxLjw)`g|p1Lu#6^b<5BY6cEZ
zkQKUj<nnqZAbE7l_0Obe35!0q7kKB&wvzB4#9#M*+Y8u#iC|sIJ(bA1V&Ax~TIX`k
z3)(+@C^k9y&*fsHOp^OY_C<^v3tsj(|1t$`e0|>xh|@{?xLICGQ>X1Rf1UG^fQqns
zy6j5^ED*R$08|kWp08hggR9dgk_B-oM050?KU?!Kf<!VWI6FM^pJ!&ApV-`fef?aa
zd$TN$$)Qq+HaI=-Kf><~+aUmL&<mVaojU#$T}tAApk#DB&>^6My0lX1gI?2kny*@9
zk!rx)0Naz$;+qmQQCl>(Xq_ky6eA~%V6LO1rIX`)rpnlUnQWT4pgo7U+VpuAwAb=&
z$Qf(|q-t|XBo53R7{gbOn|G+M_Xj!~$?Drvi$JO)5d81>BCRzqL1!yZmb3%|VmGu9
zCag$q1F!Ng&-LORhCKCsoo<Jn2Hj;Hi)X_Bc8D*KP(1JBSc}gM{$cja#jin$>)@$P
zj86Y{=BtzBHMb>FlTgtNo24@pILVqPQqGnLH}!MoNHeef2|^qi4PB|nZ#ys9l1$JQ
zm6aL`^-~MJB-89djQ~v(_-k)Km@y10hZUo?xtwJHURLwHn4zFgB}sseWltq(5i+QT
ziatq(VyzE#j6e*{s0`OZb`UNFl!d_A3&{JGl#{Li^N#$9ch^hD$Rms?3N#=ertP)E
z)<bd3w?Iv?lkG+6#ebXi%b>TIcZgooZ5I!fgF3Z?U#{M0_Ai_yZ|3;l^UfEZ+32Dl
zAFuXXwtLAuFMFP|x@m_8<s%)JwTaz9`_r7?F4QkiroL$Q4V*abvAh}`YPI%wgTE~1
zXfZe)kONWLK@O8hOqCnzF@BZ7qql~)bTYC7Q7B{8|Abnl_^k{E1nVk<VL~pq7Iwr9
zElfz)mjrc+HSgs(^SNdC)rv0KegyON^y9|>?IdbxOaF2zu-9DG(uOGzhz~o?8XxDN
zPOW~E)gPt^>bO=@M{&pgVZN48i~x(i9oc+1v|!t2N4!&cr`dKRvQ%sMf?Q_R{L5_%
ztN1c7f(1z5m<C+M36>}-SrYG*VEj*328pqCLPZ4LsOc7|go45v)ZJ=2NP>$viK%t$
z*~Nwp4rHV8#rz>tqDnhGS@4~klQ*bDu=D)Y@*f8>3t8sn8|S^wgWYUzxlQj3-R{gP
zkS^F8?4Azyn<^4@$SMcofSySjtT`h8OK5si7nT`Q8J8ZX=Q`)KAU_yEGwO!}Rih<t
z#MBkzs&e9)%{i)c?v>2=OyX62h$~Wp961aXQ3l$P^dzt;%YZ@F{QYg~&litex|tSX
z+lEYg&$9o8zOeGSESPqom%qA{D6@r#Zd1R-)9%e~sKjLK>_JB}Qr=b)Q(uv?jwrEe
z+tB#wAq*V8!5iyZ@G1zeQOIE5A=zXt?N0Ky`($l3qU`3&S|od-dOOo&?o8*0x^MJ5
zl1}udVlZuFL*spF0bi4dDDDSCVUgG~<*G*lN@YET@+a`{hgIZft5c6ZMMZcR{7bQ_
z`dNG%@|dH@(yfC6<xm{`Hv9KtgPzNlS^LkJlhTi5m+i3?AU){7ragIA((rI6T<xGV
z;L(_SIvF}meYlk&(tG#4{-J9R+HD{~ieLt)(!D4Ybu_tR^Y<VSyf7V9VUn{`K-m51
zycj%A<f~DYKbb$g=shDLIKfpscnF0TmveEJG%nd_%qU~rp5UY+MiGtebQ;ZB0z2Lt
zL7r)%32O6}6ApBU44c)42P=I@Jpc>dO_Dx!lK!O2A5)28I9zk)DrmSHI>}_8FKfAj
zhh#t)m#rkBit^Id*EiOA(<0?N7HsUfYYCKE$YmsnF0*bFr2sB1yMFK4F_!N_NL90k
zFU{aaZWSmek6=HJP3pn>gsO$~u2X;S!nK^c?R*!Bji{!wmbqClA9<ZN?%@r8blnhW
z^;`)iu;DqBT}@%2GnHl1$zB!8)@8(|oWF9sEoo>vV@4^?gwYx4g_gPa2Vq%AQDP$9
zEGw8{M8N12x(yilxVwc8>#kY(4&eh9pL6*&*y<PPW%*6NJQf9p(h;EG1ia~+7AQf1
zFBme1X~mqKr5Xv<bOAedQsExR0|%z>URa92S)qhMnT8)UTmV+!cDSjwKWDXoe~T6}
z<$8TT-ec*1fC7-e^Dr?cjBdp#2d#XIQu9_0x^9XI3-I$R;Qvyvf~8UZi%MwV%;9eF
z#$*%q0Ne;|dE;g+41+v1+ISlm^T5*FU1&)zMp)j^)_w1`p|Qf=O>nQ2$|o-va5_Ug
zN!55vbGONvl+SwFHF$7Y^R8(VjrKE!=XsyRR2Rs@otqQX%9tYykm%tBMB=%A^ER3^
zwNVD5t)CxL&lAM?uDHQcI7tcJUI=c23RzITUVMhWl;fE8H$EG}U>HE)DOoQk0)nvw
zbN&1}N<MkAFw1WQEe@%-T{C1MN|Gq<M=F{M0sYezMK|KyE4AyM>?xu+x1%CQoUgh}
zChP#28)h1wzIruwdGZ9XhoT-ZyVM-j1~B4Fl}cKx679|b5=A$9di6(JGNyUOhf9q+
z<w?dBs2K=kx>efOXKoqwT}JDh-isO5(NgieBPbXN4)-_~9toWMI9w=xQl+kLrc6)E
z0!NLzCXXwQMQ(CL0q>q1XdLqOGknM;<3vDsT`x=`6}z{FM1ybPg`uVTD%qz}4`a{p
zubbS3a%?-f4}ST@R-F{Sk38V`xlYIe8+HO1+niqHC<>YIxp)6r@fme|<sHTh3YOE|
z_$3N-l97ZK96dF3iH;fS*}F2EC(g-4SZqYA_XJ^Ummjkgisnv!CZZhX`(5GGr$E*8
z&CZRrySbW>L$IfL1qY5sKo&Y?C1{jObPLte(8<waWvK+-YNb}L%Xox#%#N7JWpKGb
zkFmTji-^wn+u9MxG^LSKpNeMWsT<kkYvl#L+34MAr!md)lCx9kae(pdw8>cm>O|4t
zilSPwA_Y{lYEtpMTHIjWxRw2Hi~B7;@q@xFl5GOgNIt&_rNor#=x?tnw!~(q_K6`X
z1gkP7JK-y@NcpZ=UEn=CPWN~lz^qB_4#%8Et?nOZ*YjSOFN&CfRvC<*{admL<!DM(
z&)wYPmYkM|Z_C)GG|4}%*lPq2eu?Wz6#uQvV`ch3WgasF=l=#S?ADU9BWZ{IE%P4d
zgZi)(P@ed^!2bAJ$8FMiL&Qx}j==yK0~z}Rd^`x1lOz{;I=iS4hBa*KAc-jA^C?y}
zba{7<P0qUM@m}-x7WSy`K~!~kdv^ou#(g4__*tA90tO4Td3*n6P81|fUf%~&Q;_26
zC;hm;W#r%?V#NGUpmXr5z`ld~Qn7q1-FkEvcCXI}w_+WC2_$JG3k*U$qui$t6yvFT
z@tVXT{~?e}<mG-p-<@!$OTT}@Io}Onn>gFVD$1&ZqYFv8*EaHG*vxlj_(m)8_F{Ko
z?9p;rnrQ)E8#W;pd1rb5e3Y^bPR;N?XWo>Qu?#PKqEHvrUnS{1P`wv~xcN`hQC5S9
zt7>rPFkLQu^**4mo<Pth%Qv_TeXUldW#{gL$g>}@OE~%v$_r9intk>3AgCMk%b^9b
zSm9L^kq9ZjXi@hXGu#g|B-1-YUce>{HOfN@f&McN#WW_4T7?lV44GdOZ~IYu9;bog
z)CSSmm3@o8>pquouBs??U1^oG@;{0luoywof;0}J{Lyyd<)ZI9)orEjjKT{UnkjhF
z5uME+<G4l-R9Ur<0IKepcfneN&vyOU4lO}Lht1dmkx8eEhViA>ZwNQgHswI&gJpnD
z^~H3TXC6d6R6xNPHxsof4*rdphGJ%LCQ!`c@m{2AYGUB_5GwJe3-HP%4i!n#RrmEa
z(-t~Ig{&?k=o_MGc0yzRZHxK-T0Fm|z5->`&!AK@D1it8vo`zs9+iXFhnW@q1i9w*
zEdqdQtO<M)#Lg-M$i+))e&MUO+K>Lny{)aP&w8Ei93A4i5?fW28Lc0;CzT_ZC0-{(
z09`+(zuK{Alf9u3z+Y>-1*=SgVpieGzJe~uHriJ;o80j;$d!pWN)kq?F>Rdaea9o`
z(0E}XThe!-aV(pyIIg#%wnv%548$miOJ0-0yq#c(f>K%f334&RJe8&df;vS<VfJU1
zD2Ft}IonZcTolszoNSKMmE~cTHt#?hxZmoDYBd3QGTwA2r8p}f(BU_c$rIGK8@wow
zr@<Gt*jyBUmWz^VYYg?nXLbzt=c|Zlu~S79RM@~{TT@ZN!@vPn5EP*$(riO*LQHVh
zPh_FgS+7L+erHSN%(f!v@WT^H{{S^iz0gr4{6UkNjCvelv!WSOaW&6IQa<S>^b&`q
zIJG}84di0}W^ZTFQt?)sXJRJeNk<Ka{)DWPPOo4RFNrNy#kS>Qm;||=e3`qxZlDXS
z?2}&uJNr~Ahkab9!qu>bYoYXn<(YIDb$z%{TZsfl1)5l!NA*c2a6$hIg7Qmy<uBJ#
zu46W{cVqobI`t4l8wdS&oltgSeoyrT)w4juXpH949wQE>Y4e;m4Tn`DUfQZT6;p!|
zD%-15Q_TXKhe~PU1H<l&)peo^<~!AuSu?POjr0Y`xZNPHY3zxQG^_sD?p_?O@7I=(
z=SSPe?2>yisLuEI7oB(Y8v#-RqEe3u%+KrS(Zx1>_mp$!NmZW`NC^JtYOG!Ab*7%u
z#lWN#2NSmC0(9yr4yCv5v1PyiFCi0`gDc+#ha)XHG!!*VH^UQ`WTXXz3|vR;qhdxI
z2V;ON&IDBwrNt|AhbzmdEVWFR%7R0ZTgZ*qFSDZAG4Q&2NF3bbMSBKN$csXdM+xnS
zpJ$eBrXKA~`WHc`zNke8<dO4@u|Jbl?njL{El#G8f%z<WG`rpt@NNoGTqSkA58^R7
zd>|bjk^{|bk1Pp$SW5gh@9vFz?(a>mS*6J27_Z%m;$p7guXqq4ryuJ5_L#@Y69e=B
z<n)WRH44LId%^`*h1*S|LWj#e|722*C6;>-C8Z2mEJU`HSSOXx%mxnCT+%<ypFrWL
zses*y#3{Rfu|7q$I&OMWF?A3UDXOMU7U`BuN_&TwFg({M`NHk|6ZMsFEquz`Z=~7_
zk$HJL7v;Lps-7;jGDgammQXcV(4xuGnFENSCTY<&jJ&H-3_v|RPBdU$ww14^!K;@Z
zB51^CAumG<_99MS-~2X0C!}#T!q8g7^g6YyhCYbFX~uTkL7eS`1*CMy@gkM~<~X)a
zk+62CMb=Q?VB*V+OzltryO6lz(H-3UmlBJ5q3lY2&lvyBm^y0N{xOp(GX=6d3q~kp
zMw#{QL-Q%c9-aZYdOTBw&-T$X%cioUeZF=nw2@&&x2#@%U3Zx6scd5Th;}RG_=6D^
zC0l1P=n*y2_$T?nL1|OBa^B9k#dr~{M-ugGTBVkV_D@>RwM%}22ECV&L}#3{5|$Y6
zi3iU!`Y|$=m~t7+eJqC6G`mXH(S1u|)uA<v|4M%L4(tf#3M~2ku;v5U6UQd@$vdh6
zuW&zd>_E7$$ydb%qHk}^=I+Zw<6r<OmlhxHLc2jt1=fKNNzNMM8&huN0#$zIcnOeh
zp0e{adoy?gSwgUK@wEa6Y!ieY6{mwQaQ8o%%VOH|NkSu%BAUN;cLBrIUSmw4?%efx
zA!{~BzmU~Oq)xGZb_o#Fz<TsoPyRd;w32Y)jBT{xb`fSi=Ryy@MYJSogG!{2XM07I
zcaiU@Y6jU`)_SKLz9f~vwxsJTZ%!1f39Axmy|9s%Dt3G#<AZ?qL{A>!{W~R447N@*
z;FITU$-k#frv$t)WCqJa$eP#ctG^IeVWdkj2+MqV8Jherq-G@5LE-%mMiS(^a80`>
z_lP1qQgHdp*EC48pf*HL;L!3B`>a;<QONXYqDThWe^y{O0*qYwoHlO9z0<W7GBlT)
z{a#no6qZ-TE(k8qsvI5J;oJt`K<x2P6_o6dFgNuCk3jID)nW!;s$pK_Z3{OJ4g}`w
zV*C~56Ur3>x|?CNjD}y1CZZH`ox1x^aUz@z0=NyWz~*HET%a*1Dx6_1NJpNxt%WZm
z2B!WhR%DpPxZUhXkiiAv3>-NbTt;xg1$#}yZP)Y9u35C6VGkjDT>RL&(n2ylHDa&Q
z7*|4y?uerS9s8f}xH!B$a`Da|9Z;Ye8yGx)CB;kSBJh?J<9@|tLHX2qSD0*J2j$A-
zp>SZDq>U~wg6k~#0k|j#8Nv0~&N+l5_7nG7u=Pwmw48S1B#|L0CG@ipXOg`n_xsIo
zqd8{Wga+<Q)U&S6+uzc1iI<0*v6%RNsjJwV$JfnfypoQ$d#_1noUG&*&$MZej>Bsa
zQ!fW;FC~=k+Rb6Rq0&0tITj~+D>U28N>3T)M4c?wwIAm9veG#!t%9SHNAHt}%=CUq
z<fL)<F3q)FF<jR^&3zgHmiM-R;<QoDKdGtEeIa7(2SM*wv&bp<RdChqy7$dj$`^Dq
z>9r_P5H9(D?>!`;Bf6&?00V^ez4P>**az%fyu#BXw`A`XThlGZr8Z-C;%5osO4HR*
ze!@WDP{|QwkyhkL4yq1&6t-tK6i%3qS^GzxUX*WIHBvfxcw7(v47>tF2_py3I|Mf;
z)Dl+V`hXZz7cnEYNZf-MY*zt8aBnvR)@9OALtNA~`ZHW%z?U4Kb&p}?__zS2`X?K#
z#nj|-9Ch^^6v3@P&c4+q_Q*q{0W=pLeO#A1V6RRiT(0Pdvzdp8a&MRU?EMaHQ9JB?
zDbrhVyGI4euR0vIr+Bl^9Ki`*u^GCHfNh$A>Sq#=Uugyl9-aeMgiOnJE+q8+kmH-~
zoIU#LTv50by16AAiCJJ*Ikuj1R5|)}q=JNf$RexbUUNSo9qqBX(9h?9ktXfA_;eyl
zwii3YB{^zY_<)bhPtCLe6hnqk6F0e#E4foYVuiF8#>%_5(p4+Do4o@<*^$yNqo?Q0
zx_Fcv%e0)laK(K5>7A{&sa<r6IH03{I;og^`NMMq8$a48OMdr<iHc;`=0Naw`3|#l
z-s>TQbwIPz<$#%S_WCn8I*kqTbKH6fO2;r{O7!ZK*t*GPdrCC>41BFD*;!QJw+#$B
z2LAm*2plje#B>nkgSER?=#93yRxsH)FFtf%aT?<)P`N2MC-AyU4E=RV9JcPAaIxa2
zW%j;u{h0G!qu8zW@^+f(wTYvF(Tj_#XPxEDp1Kn`?)K#IlXIYe!O2VNINci+qnmzZ
z=GME*d;JZLOdR-MeTTm+Lk0pnLrbVXf9RDx>`mz93@w$MZRlkQ7#SFTNgR$&&IFts
zEcE|hiO$6EKe`Z?HEry$*^$1l_4;GG>9uEWzvMy06P@V!;nCMP^V~uW#-7;}uFSWW
zp=91KZZIVqimFVsE{5s@O44;;$2~lb_h>M!)~O4tj`Ao-5DOZ5x+Er)LiM{NLWS$7
zNYVPsrFHA4sXi#^qbAx<O+^mB@RZdN!7)eNM-@gm6^)%R<4`!Iqud(m66^7LMfG+#
z#R>m{bX%K%C{u_r!mQL2rm*S<-7z(`{z`%MhHLf#eDyPj$k)blvc)y}%ZNb5WliDc
zBF7^RQA*WG6tlwXLemVYqT+7lf7{X8R59RbQo|4X`gat9!w&KPdZ3x~f*l$56@ned
zL<W9{%&G_lWQJpd>(@~sjhvA9{sj`JXfRQa(o~YdmkZN^9^r)3u2Y;aqH<Si6EG4W
zli-_#LzH7=WIL7@VX3xVMg}9cV<eMaB)FPI-CIxmg^-AHyzN8J#obd=`vhj`Vz(c}
zROS6!I-`Ds><B(inM+zULxKuZVg|w!2YXK?6d^KQ*=KERs49>EJYN1!J03zYi8K!>
z5K$@aFG6%wF@OXEkpaXy$^#MuwYfl-auJL?fNp%XdZ4;!j0kO|IjK9`YwS(}O+bEt
zMb02bmof5dj6~iUP06Gz$swU>Kr}EHk_ID`F6l9mi1-6y85Y_cH;P!GkUDTkoHryO
z6yaMShIEGDywE>ICMM><m<Vbl({T+P!B=^2etOf{BodXf)G1Zz(8P7J#+QTG<<40(
zmnTOi3oezfrKS@XFJt-h<F<FRrxMQ)@TE-$7x`QlvCAkh$FW4R1R{nF07I0h7&^Qg
zm9WM|+2aqBKka3@y0gA3(U}r1>R%`SqCemEgjRKAesooCMz_P(%XzJMRePJCJh*$x
z=ru97m#@kypKen9yzUI)f37Ue8n}perG(<ma6SC<x2vo+?oJ&wWirN(+!bRtEInDd
zD&<P>O?BKfa+_Ru&NgPrb-3SgrP!}d`c*fxSb+!iMW@AU9suV@n7OmJuIH=C-)HZJ
zUn_X=W1Sa!Y7fh~XpZq`rtwWF{s99E^o&@7k-PU5{6W588C$4aKU0<)EsPzWc`((^
zW}6lf0Xu7~C(1TrqAV3J3*?xxWP%{4;7m;M0X+mPylZ6Sa`F`})WbIbCrI;<|9ap3
z32^0SE+T83oIJor9TUcYvz(|RPW!QynUF=Ma40UXP@6VcOgW^c23CG@UCN&!i#Un>
zd0qPQUi%%!PU~H=C&F1>YK=QHnIwOaxPu5KgIhz5wL61>)9Whtet#%S0toGR)8GPb
z4%6Fh9$(g7$5gb|zErwBmBL{azC2YPG}H79Lt}8B$;&CNkDWl{EiIa^<SZ|GuoKjN
zllp&`zbPx?il?Q-GWHCMhxi)TsK_gj?YE<Pi4~9fdv`%*M-S~r5ohpCO(tnzU!Jcy
zOL`bWuPmVJL(i?H#8jq2E#j;4vEO8zH@_QbqJx~};WE^vw05Pi2bjJBf$PzlSy@y(
z*WVekrx}^z^T6V{fBUNfz(+FX|5+ObZZR@i*zcDNYsNsHq=)9{11@P|J!>`Per~w>
zjI;LAG{H_<ntkSR=kU0iGmQsU%0qWdVZkE3B-4+;4L78Q?X?n9d(xYJ^>+4Hm|00z
zuM61d&c<6oF>W+6Z!-@VbdNk{jR)xer1ftBEQQUp**6B(y0HHp2QNJ4!+HYbyRQbN
zoqPycqbk_pSK2E?0c>wibt;v=EyW78iQ(KD+>7AsyT<uuewh;E!9xhtbt8aaAqDou
zf`-^QtgS!s&V%udJ#|u0J%)ozS#%00$YGP$$0l+>kl%UzJCp)T7D_S4>&vU3#c@Dz
z8XDD)u{ms~hET)@u|R-WPyz_Z!W8)dKq%UR7whulP|eoVy_i(0H}{ohEi%!<k<MEc
zD^00^C#&nXVtRapBTEUTgC}!kdn^eGxAkV_!IYVi>u~jUW!z?H_(#9pzh(us8kJp;
z!ggjWGjb9KeV6IiyVZbb3!nohxG1zn_klNu&T|ptlDq9>h2+>f;)HYP`5r^NAEV2?
zXcc|f<>6+UD0D;GI9~5&z{dHc3*k!H7Uqz>x^s`U>LJt5CsLD%-Lc<+3)>xjVY=AC
z!&&P$xcD|6=E;Q{OSVW`a_!keyu0XlFsR`5_8E)EO8vsAptR%JC8Y6j@sQmgX=H%g
zSCCa06C6YP&#qkzXj0Q@b5Bu|PdfoSZIiRKwwdHpS+?<dB}pmR-<vnlbVs2!HW9k8
zb8<6&mw>h1)01tx(&*eP`{w2;Vri{fJbEB-fPyaM5rOUj>?$@3kFISB`7G^W@t~nN
zUy5#XeJ-=jT2YR>ZcSw~OLqmo7cFv6iF@pce!QGuZIwG46PH{o`esAxFNrMzWabCL
zc<4$NUKey8xEH)k1c9pW(q;txxz)p=*B8IzgRs*Vg3|+zk4r(jQ^0d4+_R8*_@G-+
zu~scoi!O$DHO2<-xC0)NkX4UVWe;L!ZS(H>&P(Jcb`YU}L#rQq4YhVFBHINH9;f3b
zl=te`Q!U<hcMaZ0{N(KA<Seai<cd!2Zo$<CU0ahS`qurTnXeynla#(UM;^T%8)Ou2
zOq7qO%)r2mNo-wXHv1EV-_-f@gZusWu^n#z#(syhS8@%dwpux(=JjxnNirSX%}W0H
zG(y=ssPYfIl0>8Rf9qVb{a@t@BPScj|Fn)dtu<-8&4#f3LH!79I__}`2!dXRyUFh=
z(+%BVnQ5IN(s*{`CLZs3OQmK<0xhzN!kWdIHJZ;bKfNl-Uh5WX=hoXzz$)v(2A9!F
z9~}>_`TYy9qGg}KFqvnu+s%->X#%5OEz~YzEzp+r^=f}O!i!GDBpWlKdx?qq3=yR>
zzqut^Rr<c7f4TEWNZfI`X%M#E^$&Xl5lSyUYG6u);H)8jh>CrgJZXh@<H|HXv;})P
zY3=m8u`Hy~3dzXe>)jp(4;IZ21CtMLpML4hq<8CeLy9o2_WmUO>5&b0R`sw~I4qs*
zc$32OjT^`Ce$dN!uhtS}|6%bP19A%NUXWmz01MU%bOEM%jT?R@!U$SGS&39HO9wUi
zvMCp>>}y^-blv&8X<?NvW=nnz7Dbe)g0Q!KC641S<3;X3jK*&0z>~loE|tIm6RwrO
z0kZilIlJJM>Q!0Tc*b!+ILq`lBO;kDvfrm|y5@@;N(60;#&<aX`dA-@b)IuUv)i08
z+$!OFQM#f7J{F_hwAonwZ{`vc+$3oINXANp28IAp3a|x49=9BNAp^8h`?NvR0k?<W
z5m1jw(J`-50q$tLk2T@Kg_*(bP3b~>9=1R!<Hh46NJf3^s`=fD$;jdF2l{!GNfng2
z)JR1UT{zo1Y{`hFqWE)~xzzzoZeu=)l*^Cz&xSHFzMhv>w^Fd$hPZSzvUd({5rc=`
z><WG5Dymfyr0re@LF3<PR8!B*jS)%)?`AgHT3FF|_x|Q-&I@aP+&B~hnq`1)qMI20
z>>;;d1@xN9dOGRAZx&Pqt2KZiwSEI!t)U6QwGptjaZ$`aIk0t4P#}MC-@KathGf*c
zH#DEj0gf|7kSd^($9G{1?)$czi@ByP#9|vCsk`Qqc7|n*<mD!+S8S0wJ#l)Jkl3}_
zgK<V!KiU%wM>Hev2{`=Y?IPlqGjO#dNsbOn>95n+#>9vtZ5fnc5Z`Gqj}10x>~Lc6
z7wn|&Q<GFrTM(|OKZfvijr`K4m)@baEx(0a3In2)`;3&&mCVFESDM}t@JXrGDwB&|
z=RH@d^~rqK-t5YKr)lC<$_196T{6o*qJ2sS@?n1>Bo3nk7gNM<5Dj52gRfV!#edSE
zHOx0XOyZ7w7w@Cwl>OE|Z`>ZhJe;Gd-R5M~gHh#*5<i&9=+*hXM1Rs2IYWOCfpnqr
zSyz`6eSq@uPF|N9w0ef}@U_irKe!hGa?|V)pE)rosa}Ye@}~4X-sMZ&EwlBGkdLLb
zs{`Jrjn?+r=IM|2SwUE)7fyhuDIc_g9mibCIX5T#B>_;gm_y#Nys^H!Eit$ZDF0{U
zk!2z@OR^lml-WIAanZZkW8O{LECIMIL7Q31_rIEVHrIm}5#O#ZX8Li4q0UI4!koM6
zhr|YJRo6H$WpQ6M5PH;x28zh_$85!1VW0l65!Kw`pZ<{~Xnu(IN*2sQz7^}-W3vq|
z&~N38n_|9bHQD3;xxR6^>VBJhCxG+xPa5r3Qa+`;8!|qJpK;uAAAgc&45vRYfD(3a
z_Hw4$|G9hrk{g}jUJ!qj2XuVtUe;~*U>p;Gx_Gl#C+M`qF@?f-JMbRRV;t^AU&UMF
zibkpV0<GB%z_4scx8w>!2<WjcE%uteUPNK2x6GkgGG)?&z>1x3_QfkjXrz6KR(8-j
zD?B8GJ@#&9_LUzJ>gkee1>{U60eaw{mnwyM5i{?pp9lp?IO%=t6kqV%m|b@3QgT{~
zZEMOW;?#P679egmYs!||DRn{1QyWv$Z{UThH}R+lbSN}z^QAoDk#sHACz%)j_KPXK
zI6Q5u;c8TFylGZ%u!j0I8`E!Q7o1PnLO5=rdr@~LcF#R`jMVjeY}w^UPEE9X#t#V;
z0^r%9!{<fYv|%V3Rb$t!0&<kEB5p8_8O3V=J|O&t>Nf3!EbSh;D|Px)b_4KVj71Y`
zaVUKCUUV8A8YrD<U4YMQDjlpTWWB!#!TI-Y;ZfQBDD8@Brmw$jk@t@}=B-^m)EYeh
zt}*mEM(WQ*^zn|g7Nn8MOS<m|sud1^`<=?8Exr^&g+0L~pU8wcjI*%g!kIg=6HVMS
zHf>>dHW$w)2)3Q~w`gYJy5?R-<)dVXxXt|KDFSJHPeSv;?_2p*f8Q=i#!zZnWV?1J
zb6`nu<3ASd6fK5Dc!pf|e=1*X!B?3{vz8_V&fcvt#Iyw_qR=H=D%wxr4Y%PD$Jr!j
zSCXh=JpI*Nw^f?G=JEM@M@h=3j*Lpz?IlpD?xj(wU7*fgM?XhZ2j#q!`NtG<_#7$Z
zh9FHh(&5rlG;4mCa$+JEr#Vt$*LDLvNHjW`$%lr-8bu!eT$M9J^Nf+b4!J#`c~V+a
zH+6ntoEcROnuJTQ^CNF5|NU*`bwlAKxsA!@S^Ba?ij~DnY@i21O!B8W9o#?RrEk>k
z*Z?A8X-*{X5QhZ~z`%(j1~fgjM|DXT-%Iq+oyixllI{b9`Q;G|`ULi)o>)=vZJ7Ff
zXMfMMvmfju2K_l)FTqPU%!TqJC_a{*e-<AgOv-B0XEP$}jFUcwG~$(v;Yeq{-%pw$
z9d|?WuKZqex7hOs{*qur`M;h1?Em-lXJTUi-@+K(n$rn~6mUJOO0S?1q6FVWTucQ_
zm6gQDEsZXjX^T_O+AZ+hsXkw*>iSXH77J%vK@X!Ng6{}|@BDCnu#T_aj^EyI_xTR?
z1K9SKub+$nzXD;%|DVQSg~{a$+wGAor_nL%_%*JbIX-SKwsLk`-Yx+F>DYU^Ks?F;
z>sWrO;ydfSw(@n4oJjoogMCcGO}&M>i^;hTRe!@6w1~~&U-H~}IsG}lKc7b;4HAkV
z*nT6m&%3R&sXtv!U|41#lI)*%52U>Ou!3G9u?p&EVZA@BpYCsB)Gf|rXSfFY;9u__
zwfXH2)J8ZY1~k78;Lv&^R0vnApt-yGzl;egp$vkn>Mh&rf(>_qi4z*XybVI@=yZwi
z`#l2{KH<GVL5Ow$T~R<>Djn7x4ufAALa*?_$C8sG!CDB2BiiRbspxmILP$kqzfK5%
z*3MfWcHYFpjYo=RG4!x~lx~)Y1pHDpYL&;4xF1sZ*xG3OrnY?KGM$ylIB;yf=b)un
z;p^EtU%M(HL>?j^m97?`Z2JAeiTSFR`JjaLOM%0GgOBl{7OthwIAXpxdC4`0*5Z;@
zwV-Vwblq{Idm5`=!KT0fCcy!5+%kYfU@?<lqGozoQXuN=8-Wlgj2qtuOk8hx*cnj+
z77N??=n}{X+NHV?f@oOgTI6%MdTvu;hndz=kK5z)@Q$poWU}Y77cBB|%2F3Vvs4&Q
z9J*2wOdAHZDGu)xyDhR#JIpoULo+x#fFbo_HP{4{s02pECChR);yqYt8WE6^>PuqM
z@*?m4cBHzQc`$ODKX=F|RjB$ZwYpvUtkqU0ZK#NeJqBHFH-}lh*#rE{4wBfSrz?tW
zNC^*dUXtiMXp0dMFjtyrppUT2FP0ghyfKtb78x!Bd6OENDGcpBa@!v7-3f1)q)AW;
zhjzV_dvczy(5x%{@INv#ekrzxw>NvajIdh>mljNM`oF7*TSzqEq(IdMg=-SQ^LNYk
zy?<q=S$q<C*oPP7!4@ZEEH0bR-~{K-E#ms9&SD0#=xt^@IgCbL89cPCwY=Y>cO8w+
zL2fQ(6RVj^DJibm+<--)7*G;oLgQ0lB&p50l#(b=(NS<(&0NW;I-7U-K@Ay($nym!
z`7#=a0_e(4$0HY7s|w0b9yeTx8aD%Bjf)T^H^>QvvVU`N#tBbmbbnxeTDx|%&q?Z#
zv$s{ZM8ScfLasSa_47~#Q$Zo61xh<<E``i`QMFEoPS5c*q_5HDTYW|AQQV)aa<2rp
z1%wGzbcV-OG+|_GC)TDg*Vg&QZEHCIIN;#qds9;rAB{UcQUoo&&E7o5+;qN7Emlut
znQ;SBJD|}!d<{U8{L~*IAAWUfoPSW!AkLNPIynQKRsweQ@BwX_ZyPWJ((I`&FbNB<
z`%!us@ydTL(i&{Sv4)Ul4wMp82kTB-qvo@bz4z>r0xVcy>4!<owl~FfGOqTCdOO9P
zB>U^;1gHh}FM!BO_R4#i)e;?W=9V$6-kQ%hOB-l+UH+(^?ZV{C)21qU4T}U)x~V3x
zq^KHkJp~}=q6f*tfB_jv&nGdJ>{~j#tlp^kXmAn(B?Hnj6*nC!H!vSXEx9ygaS9EJ
zz|oPdDnIC^ljK1C)L}GOy3?Fc`R+H5`Hc2=DO2^Z&Mp6V7pG=D;s}Xbq+k2<3Z*ka
zRZ#VG1=uXNUI#}Qc4W8t{|PC8-IWcG%zX`r{}3y7iqn<Fgj+di>c`Q&I_>I9yDRdX
zRHBi%E#5Y`F0v;2#=?H-NNjoVW>I_I$eI9HdZ6LUNnO{uK5p{&uih_a=x{<7R;mM5
zjy%_d68qZO0&Y(aiZ1ep3lIf0vtF=XsGe=>gwW2uk0~j&s$xZMlerLkr8J$!4p;La
zpbun<{LbYk_ytr;m+DFe@MEJ<*v6<CM;_M)VRBq`%lFfq<*`hv?b^$lU&UY`Ck<FG
zN_Mf49pngZej!r`?J%7wPmc1CXrH>U&OM5z$Nx)AMHbsyuCG%)wNxvW)l{-+H<C8Y
z>Q}vIJ$R_fX-zpfBGY@GP5O7Yk)m1=TYjy>l#xDzBhW?@Fa|h<5rCYjhw-^>T~aVH
z<+(eeg>U}7$IGTfJrQN=g#NK6U5!!=^c)ZLXp5Mg0SEzq&1r4Al?`O^PTb!d<yhZV
zoseNNi(dIczqLA*4jq74v7<AyNbYt+D)Ca-#>mAai|6-v3bIGBJC8&emsCyvmU9>M
za(60T{-{k3)!Ez~9azA5CFBkVMK!0nzm1Br<5aK{*5@n3j_`MVW<_<x7Q}mN@J<V2
z?vu7lf*h-}d53?HE5hkq!ki;QhfQ1HKwGIVo8ofC6&O@V6{j~~fI@zqDO<f0bO5e(
zt@;&=!2Aq57eu_Mp`<$zEZ#IV4=#st_>bFv{;ssfO>r}S5`4F43+N9N%P3@8e#pPI
z0Egeh)fKxI)WqJ?4u8)xVAf9Z%0@F=Sr&`JUe&4B;(QB9>fia!?REwvZt+u6Wc+@;
zywab-rqE^_Qd3a}YbBP>sq(6X<nGy~J=ORWCMsMelZ-QcD#)X@)vYKcDN-8~n=|u~
zWt=*k{?npx7<m;b3&p}EoH=cRhN+MP32`&+`%XKjK)>j6>Z+ZgZ$>YTh=s}Va&QQ0
z`wXfG5kOE?5F325sy->gQs8!Q9BoSy+us*WzGG%3M4DMAtos!ZpSGc9ilzZ#YHbF^
zwSx<-zoX6qJPm!n-&Cp5y<+9`WT}<EgF7oGeaPs;k?0qDJ?6QXhvTlh9Z+<E{h-+O
z6VWw@Q7rOY+={nt_XL(P0Rbb&l>;;ZeL{`rtB2a5HwwFXAfRBAW6VAGsqAU`k==aa
z`^9#dqKWZRR7grdJ4~wG`tiMB*jv-e_&4mgDFq&=FPz5F(5-T83u}js7h8~Vi!;a*
z!7wi;rkwKoJAe&L+?TkP!&JZZ1?v|mK9tS212)LEznx3p@*;gSVZ8%+xLV3@{cGL(
z1HRJA)%D+MC650>+hgYZA6NOYCu26*ZhQ3d@gW8Y!|3(v{jYu>*{jcX0&_OVQjh1f
z-#>5_NkkOw)@KCfC$L~orac(CK6dV|7Z=~Y&JI)z&C*o#g<rm_fj#@r^!_~-s%8V~
zYYM)67k<1?wa+<cp9#(s7fsIXrd*#N1%XH*{5uUyg$E1BA(<n6?`&Ie3HRGfh5z>H
zL+)XnMJzn;MJT-$?Eo1*TL1~CNY)Gc=dTL(a6kHSLXM|ldA16Msi3h?3(-l0mGqOz
zKrmk{U7&|3CXQ1ut$ZgwPuib%R<A$AUxeU2blyAa-dnVMpGx<gaD@8%g!h7nugjl<
zY}@h5>Q}jMiCQgrNRDVg1Zv<SuD#^#;qDNlJevH+JMWJY_G?5J1uA2XRK`;I(CHhz
z_p4RtL{5b}`(TAG^j_<#|NO<{!SX<o4HX;=OH)V4YD(!}lZ}C7e$y;GqqD2<d1@v~
zzmKIq-Q9d&+-xk$FlXN<vE)BMqzo1!;2E6dow<H)yE(pA^(nR(>PMQ?D(!^!1vuQ*
z{+KA?V*?~!hVgw5e0;rb)AV{!qX-EWb+n?6>k{lpL%8GB0XrqaL`>y_us1=I25ivy
zpj|c|nG4$P$2+dfq``)wQA0`2>kOp9Ml(^u4ENwf5qc$6g!@{N*%?Xnm%%d$MrG)G
zm~fF3-<UbzL=$lZk4?mxGBG$gkGCvPyld3?dM?vnKJH=xvp#X6-1xj^tAAY!>C&k+
zs!WJfpMc)mVGLpM8PmWjJ4gCF&(%22M<%dL+dSg1smaqD)3(i39e%l~1)Epzs!R+u
z#e%mMt8PsXTAPBq2%aEfp-`Gp%=u#WQW<kzCTLOpVHrv+8;{pZ1JC(l@XyuzXgE1G
zpWaF+k%ZQi>HypMZE~9Umd3w|s3!-c;<)ioq=?`QYw9Yqvm40Bl|Ff{*oO=Ju8&Av
zUo!nO>lZzMulUuyZNBZBogOs=tp<C+DjMXX8PK6F<RWhLc{qwB=cp|=zKH0ge1m~-
zlcvK2*=BciAB?}M5-d(7j>_On)PK+P7`hd>PFtJfpgIZ1sjEf1pDdapI{8QLS4t->
zz#+n$C9qAzbI5nzqvrKEzhNCKXy~l|)s?9#6R;*K?r4EVP(sb*xYIf+?0+HGB}uz?
zSSR%vNUD+WbB42=3m_q+SkO{ImPrCx{?68wI*0)C+~MB^kN|oWCP5)D>jMUpYQ+7e
zHtwc^NcvcdYu(A|Ii$}TK;|GrgAWOj_{A)GmtW%h9DYAgKcM{R>jex_-B>m4V<_O@
z_l<qQxDSj#3<(W5UbzhRX>+LG%?1<ONfcM2^|b7d=12F}e5u$i5~rcMM3Tc43ggK`
zM!zFS=XEQ?J&Yc%qvXhYZW2?bJz6ZB7OFv?>b;E`RRKAbi7=Eamy%&Y6-*73grvFR
zqT{J$mL`;`ICrblxggl$@YFTV2h4<po<*+sG`gp?KCTL3=T2FQ5V~T&J*($h&M)YG
zFTOH|BQnK=S~{-M<f_cjVa4`Po>3t_CKoTCgq#h<ppplu)QTGo%>fGqSw(Iz22&da
zf`!*4r=lV4j11`%JT9E_`6~T*0+g{VC*rN<yXg)~7Ey~clO2uKsQOS_9PZorsM`eP
zv~7x~w9sSE^#c6n2B(kisuZ&5Qp@bsa~_d*y5hm2=8rsukYXk5qR-WU5#^`!1|d`!
z&vN3GXGCkv5=EQ8?X9XTJWGZU@fi~qovgu2bDd2o-~_Bk<J(g;0sdGvfCUg;U#XGW
z#LI%pJ=TXEfxcbUqPciIkm`cu%~)G7S`zc)QCzDF<V;;#o)LhG=XjH7J=a?)(RJ3X
zX$<M@{?-o!#oIH*0AS`~30=aO4D#?@mY5kJ3a;r2sj$58dmjh;&3PR54PD#z6X{X5
zX9uN6wccK0F>P2J>DS4zN7_&J()$K9;6I_M8#j9l)#PR?$m&Jh<^7ng#ukhp@+I6N
zwRpXBroi2ylGv^$8%pzdW~hJ(W?RKCG~$$2j=qYF76wHU_U)4jq$&;Rxu>+m>dY~`
zoRnF04G@iT$i23zgheIyZcDT)e?U5hH9$#?d-!TU`Tn$ARqFLvXP7}$rn?L}e<_CS
z=rlswOIvadjG>Pvz&3eLman3L@#*5`9>Q@hqCjC?hCr{DFGWQq8E8gxWQ^sBDCF<(
zvyvK6h;fB)&Z9+y!#T1u57;jzP0T7-oyfKrp*S&tYu<JE<s@5|^fwtYvpUzMuYRmc
zM}L(~FR2Gk|4pyBpkrH<_vtgmf`M<qlo^i3<bt?vn_fy=u`UV4J0C04!F6%QWy_B<
z9lm1h%1jdH3)-SU@g0}MhZH3|7!-vZJyImpM3!o_bik5?F_BI84>kVLOfo-FUy4X`
zea3~RPv(?WNI->)j(u=Y#g_ifoX<^i3S_4be|n%tk1&xdcb;!iW^=6B<()98`tIOx
zgkGw@f)t>xn|M0HheBc?@@3uLm9HR7e(F2BpH!E5ba+(3DWAPZ<VhsSGtzF=mv<02
zDNiC78Y#I5K#LpLS)kpB)<-khXMB$Em{!2e%+IiA4G=rI_!d<+Bwspm+)zF2j-?05
zd^?x@kYyn~9k{WRTtNC>VDNUUl<uc(ajvtTTJKFe_%BhN?W&6PDX)j^@b2@~w4!>|
zMG*EG9dqf^F4}UK5_&}kdcOgrC=Hv;dJ=zxjcXh1zAj4d3tVCkMmZ;5LdQJGZ4TL|
zZLu-}2+`Vwt92?^L(k*UlG^r~w}CDw&L=Gftt|=kH4GbOMyDA<gd$(Rsy4~ViBMl3
zI+Qh7=*XllK$#Ge4aAz_`fw8|1}mvMP9mphnIzf1jyO!;Yx+%h5^6Yq#oIB7z)h?y
zk~-9z)Ub_Fea{B@%jM`MRhuY#XTnqKEn21-M-BE@Q;e@MM4*uB+^+>o(t2tQz<2bG
zhx&z1yDoThF2e!u$>mq9KTIsw>oIfnX*D&^sw7TG<G`fEw>a7b8QD<j8O6-9yf<x>
zXUZ)BiB)uN<HC=yW2nK6u+LpF&f60`&4n+^12>1MvgZnCOVtcBB#s}U4^7;6Xc;d+
z3&fCwKFutN%MzK)*gh^Z#Y-i-)C?GLC)1}+#4K6g$n&nar$mlIJ)<YbEN4)M7=e2I
zG}NOlJI<^3^A;If-=H*og9rV^Z2zyN+J8x1GP3+{>-bk%Qi+>vF#kzi!qZHG_l-`?
z%&f_}nmJl^wo1#{BuY7%(|vt_QpOcfXwh$`<9W_aAW0wq2>c@sPTpRXb+rpy{Maj1
z1I|XqOJ3amNR?FxD(aP7kc!M!l2kAJr6g~km{B$tUkIa%TfpeV_<D65OeQMic$V>}
zdU`nKosLs<IQy;EI%mJN`gufLh09`H<fIcl6F)LGr^R_mbkID3FkqbI`%a8r9JRj4
zfsYeWc2u5!-+kS_<fBetA&hKZzu1MX0f^2QUIT<I=YM>gy&T;-<;PJ<?H$m+f4GJh
zH4a#zOZa}vxcbQ-ke^!z-h4~w^d?&vzo@-X%hU<4L8n7tYx*;Ysk>@|;HO>k1IZvD
z4Y(pvb2e(K^U0}5Al8v)PR&p=aa+*tk6n&!v0xQM6CpUTb+)Y$Vxgr+B#+mPA;en)
zjj_aO#^?0ElqBTDZ_fB+Dz1(SFG}9v#6}v)cGr(G-o@ab$>IQDGM?xAF}Zb2u1;yN
zI4B>eQae%OrSo=&D0x>Fdb9mR{2}FKc@X>zn5@2|T%5@&aZ?z8{}L@*u{Qo3#(JjP
z3}nqilCAX06hfG>B#H^7$x8#KjbU0}9ojuaFZ<GXPt++7*}kMJ12o=$ow4K(SU_|@
zK#D{&X2I#0sX|-{URW>&w|L!QuU?$8g+ya9%Ceu?Ivi_BP(SMor{P*V=-eczCo0f{
z1QK+x>FyjuWaucp?C{qC%=%D=ysRmx+}oK?f5q`g)Pka);C<lWEATchP2Ljo$kdi8
zi8GCg5o03fRibhtX=U|9PTO4V`3cUgB>o_G^+?IaG|o|iH}|kPUYRO2Dy3qkF9w%N
zYNzILsf84za!_ir3f_TgGATW~-B0-7fe!#)Kf}g7Po9hPTtc>aw39~lw&e7DpLF~K
zC<qf3Lqh-?K3P$?Pe9*Lqmq!u?l|C0l3D;KhKB|z>GiKgp@^#5W4<;VGxSvphC3pn
z+rC>MN#o(TO|&{!%Y<^#e9AHPdp!z^3K2J{b(BeyJm=G7(4N3rU=UP#4N$r-x!a}R
z^MV!D<h{vHiUFRDldhwiRSrF#7Ae$)s%cDB12JlPtR=sL>^ts4_%tri6v9O7UwAsm
zx$5`th>+FJ*MoZra9e#*aiIb5Zc8gs`A{TtXMMX!s+G-UJZEe}t_$}`pR%cYUvct*
zRL71}4j{nqXn<!HxvVP<e&8bpSzI0{s)rQHj;ubR5d%Fn@MGrzB%d6JwSrep_;pDO
zP%FIZ&V|!V`_c8wqmO1Sd;FDW2)WNQxJM890V`A=gvuR;ErcIjzZ`s*|2{EBc`00*
z{KfYh?eJNXz~Kv{od*iul!)Cn&Wqg-+Od<LVr08v^*8f)vQ4C6v<)w~1)zG$a9Agc
zF2qCsm=y6$l51P8&2U@kO%Q&pBSjd1%^hTGwmk`=hEnxMFb^gMOas#`4UikjP$C_Y
z+mC``PH6>!CxTv2%wNX@fyF-yQGeE6WpK+-TeTQc6QNu+`M!;Tic6VWL(pq+ejiI;
z&+T>%@|K?y5I)c?FS@R%WMTlhtD`1j`W9Yh<8WGf^<OUO8oG}4;OgU0dpf)wlD-wi
zgN6M-7<qPPxGz%CpKV}1tiNt9a-ZUB!WiY(@C4)Wkj1u=2iE_&fqw@O@qh_nMQ8wg
z^Z+<h<#&KEUZi^R6e0ydKQR6oSE2udwwElQIK_y~nrF${e$3kbbAQw4cIf)gHI)~>
zYH`)J9qloM;1^4=fX)?wMfMJrOeb4~t=7$Vk9+^Z^qsw-dT!OKE%jb74f;){*N{9U
z`66LV{Du!K4;lRT--IdyCGcq96-yD8OCOb%82_<gOU$mSd+6@ezw{ol38eS;YV9FL
z6YFFUsqPEsnoRpuZ{W^}#tQy(s~IpR)>CF5o>cu0X~S!so`NG~&nFqWf*#+bKPrtm
zPLZCrxk$ZASs5q1FW27-V^wEBt?_0!NO!3XQKFgxnQZ_$fumX!y93_plD`Vi<ma!{
zgeTW-*UCnI4=le5@j~t7@XkKBCGBZ=3;pXfQ~r4==6T#rxQsVspalv`OVP;(-=Z4a
zd>kA6;AK`q^|A%plQ4rP(-P3Cok!cO2~#|r>~7W;<=Xv6yk++8MK-aO>G;Kdb|Y6A
zx5w+8G@U|}dqh@}jG6B57OaG65>-c2`B(7eICDdtXe;q=?l_;;f6(ts{Mg3=N8p-X
z`IS^K8mqX<5KAf)4{HRA%z7yw^_6N&M|n#Y#m}yh?RmT=W2kB1<+1WPcxrZOl3qE@
zN3Tp(%~jPMv-JbF24C%Fz3x($g^X~v_w-u7#HVuM70yoq4qb(@R>CEY9(XP#ltV{>
z(YAjEVt}DL@}he;85hQW8(G*Dh@MffFGQ3kr#rcK=knNmaQ~07bBgUO>;rUdyPev$
zZQHhOb86eRZQHi3sonm|+1btFV!w;;@+5CgPI8hrKRyp+54KdLrq0=PC--XR*vImd
zdke}_@KifmeIfc!$L_PdjFuTENmaK#)$mRI&R@-cddw50L(W@zg=MP8%hg8v3X7P<
zhJdB&q@l_;^;K+>P$ret!4CIYr9$ASd*`B|pf@v3-GYM=S|F%CM0A)>8$504r&a{H
zJ3F_!ZeVwoE06(u?_+d)v$f0^RFf}~%>2ddD1utooq^h^Im}HdU^Tf6f*7@c3?@@H
zuda7l!(PV6Iy1sWoVU>W_Xxa;;@(O)Q<g}#9B`T%Uqy7ue|?UhdaC{ax%XSK>~?Sk
z9>@bGveP`=d~9ZzJY;tS&hwZN4vn-sb8mgj#`^_<im=hxrTNl2+$xyQuYLCN!am%p
z<D8@6DJwvD4jJDKhNX}~6k$j2Xro<e!(1||8k9$+r`6qGkizL1^ZyjAaQz1$8vB14
ztc>bRJ3qA{?446RgBU^YaV0;<UXSEn)Y-Y`)@?1@&-Gh!_<e)!k)kl0Nu?=ZM&ZJK
zK!Fh2m%QJf4}{+t%o#B3>y_fry5Q7c_5EN05jcLq;`Xz8Hvr8@V)gaG_xCu_!|7Y}
zpq(opnLRWQck}m3J4yoMGjR55g9GxxdPKXTY4p>-HgoldpCVBATOF{K$~i`Se2M)E
zk36wURHZK&5tf2rk8}OMKMq6Bsg!<V)!;gD-zE0giflwR0Bx`DasQ~1f0*wJoz!8d
z(4DJ?GnCIfNNS}tJgX8|;DdyF(LVrk2bB~$gM7*A%6ov%N1Y02I&eRk?V)^xo{kx@
zjG$lne`aocvNXIYoNJbIQYKv{pR)JlEPq*03bE93$RpVWYTaUzV?iLm-MfmJ{qt;-
zXXZ66KJO`TW0|s{tI){y1@S_vaK-~zS}~C?WWK|?&S(~dkZefsGY&f+ys-4nWb#E1
zSh9+1%nQX|z^!KCwxGsxu}Vo2C{BQX^InG_sp2IP{SBatuE`L`)yZG-;FZGUUotk{
z3wc4a-%tL_->?6Z#Z+2jv6gYz?!a&ulr%v42$jZSWeE5b7$#s4P~PY<<~jx-EFU8J
ze|*jd`97(*tL7E2I+`prSWAH8%1ObCdzZ~vvQ1f-gq3*YnkAbqb;R7&A=nes%nR;{
z#wpXQ6QWpu?UnycXA2tQr|=kJsEq@ULd6V4MzFXO$GbAV(&C|(Ec=QCvFUrSbPkqg
zI;Sc)X!k2illA5(=64`i1A5v~CjIg9(Pc#iKZ|f$bi;yecG`D{*BOb8lb|t(_K86z
z9*?NubcNkEGs4Q)q~ya(!ERYuBBemi4W{Sk78?Wf&&|_!d!h-m(4gU=ACSijd!E(m
zB<avJo8Bj5zrpAKDx(mrP?zrrkR9aq?-#)yUC9dS>T;Zcm5a!)i=y2zU&tdbF&3tu
zw4-mXV_xGts>7UvS<zp=Fn%5)9?NvY=kTV1F>VU(<)1Tum2nN3rlqV=((=IZf=^Zw
zw8Ib5CKmhZuxpoRZeZa`%W2IRnSrb8{)RY}VfcU-_6vD88+H=9H?Ds{KPIGvY)J2&
zgY)vES61C3=H13LdlMGF;U*bR$)2qDEKJ0l2%WG)^<~4&5(|h4Lfh4zq^-~fovL3J
z)LH=x;x31e-ZHVe!4(s2&=*9}Wh%AcrKO;?B+K_I>w)hYtuz!f8a7=Q-^D(n&;IuV
znX+HVRn0QSn|>Er|Mm<7w=%?#p#ETB3)|&Ianpx^&8)oU1L*)Wd+wg=vLFP_LIQXd
zCPg#!?hRT^$E5j<jOz)~qp_{kc)YBeu>@0EVAaJ@tcA2kn+@=&zHP5^GE;?=u67Jn
z(D+h~!48%9V}`o`-XTgbW-_4Or|W$IcmzLzBZb;u6iNQS!ViZ7!t+I110XjFU6HY(
zZ9{|XfHM$(d+8EFyZ>I$OfPnA7pXac8_vXNfQwFg%{m&h7n}DFgl2Uq4eYEGgrAkY
zEAF3qaL_SDWa|7YZ<(7`r+mSD;bMIj<}svKEP=^C4#%`O?rx8oDR%3<ssPWRU;;-C
zKj8vU4|(5wlsea!o%(lcIr(a<n=G%olyI<7)i(Yt;<{hm#PJo(58pet@P&+F*U3c=
ztA~16+D<Dnlg<v2O>#{jlbBi3hlLSj$R4q1T}0+<L|MP3+(y8_vFPH{QODZv5qBUS
zHc47*!PR@lcY{;IJ|s!ay*ty8wfcy^k&<N&%HATswWAxCj!S5Qd&iHL7n+T-C%Zz?
zmE%Xx_+2E)YA)HuD_lUt^(El{NZ#y`*Q;2^SMmtF48Fbg<Eis2F9p)ju62L>UHVST
z;13o40^D+l&ArI)FeN016%k;r+KXc+Dqzdc{LYeVLG4-nwCp52(NY3;moONDi2?r2
zj7YGX7wI_A0rnA=^RRTICrv9<v&hk1$b*wA_C<GDBg))|1*h(e-1u%zMKFiWfk$K`
zY@nVtTG=Y3XdE=*FfC;*>nE#cKUN}gySqae=)ukb<+h`a(0Xp>VUkbi-bGSeFadCD
z)w`qWN*cH0@Q<NO%TD!$Mj9m9MVg+uoeyt|Diyq;`qZq>2pqwgTHmyuaELeL_6VR{
z*5UTVFbT;b(yf+rlcAPSde^CNKF|5Arbg8u{#XJ4lBR()T0cMiuLJw4M|20d-VFCs
zi!y?}>$P0K{KC(mep_FCw`OHklZp)%rB*KBz-zUH@uK36lhKytMb^wYg|JnZiC>{5
zS~jyaYz$)J-z|H=CC114%1TJQUdx`}TEOnv`8q@7IXC&+J$K0eo=k&q@PiJy20Dl+
zNSsY1bPpD{UR-*}M`9gd2KFX15C&4outd6~^vte@mm$_Z&wyM>^Gaxx!bX^_NNrAD
zGGsb3Zzhp?dC)B?PV}#}n`j73DO@SNa9J6%f4m22#>_cG)t(SG!M-}(^S2dJiVIqi
zpj8Z|VkJf%*#nt9A{r*>=bwf+kZyCHAYN&;qvaZ%sIPDP!(KD{YDad9UOYW`zeLv_
z(n6U~8MEP}NlEJhiB_p0U@F%>!U|2P<M3kCM_GKO%%66)!$23x_$Q~lLqpe6@%-F(
z_L{RoO#BaigcKWGp}%P>db2(9T_zj$U+6<d0$a^-4XzEr_sf^#m)JeE)#`dww><%}
zh(jEiFL#~#y>~aYb%SdcZn6CvNsl-g<V%T~;*$tE<9Y5LbN$IpfI>rDPk3)!y)*r9
z>4&<$v>b<h$Q7|+&UITko}Lm>L}cxpt21g51u)*yxF=4APnS66m!Ao!J>tE{1G)#w
zj%l=daRY@p#TU#yE8hLu?7We4_h^;hTHwuGZdsY{tYC$C-+@6Mt-$)sTx@>b{k{`Q
zcWK4lDfT0A5b9O0+T>lOMwjlbBZwIam3G<#b>|9B#@RhZ=&2?^&zdXG)kz(5VmjOx
z)Qi*-w70*Guz(G>y}C2CIFVJ#ijt;(P-D5?iz1>YF^S~M6aIOkoJn#D9BzzHhw=*U
zdob0!?42Zy)aQOH)vdp^N>3^v<7;2<wh3tJsvFl3j*j2@<`UCo%gW%&ogH%~TP?n2
z$ws5diL;VFIn_)e)p=yV33(<>VzC0Vv_X{YJ22BN;U!x|`cV8CPon3!ZOX^Zp9M?G
zd56fPEs-{IP;_xKtP({^S0NLYEuQmD_mBgUtz4posDrna^Im<PwYclNp>j;<)}<<0
zvL;=TgXr-S(Fxm?u`}mQc@lDF`c>W|`x2BJ(rJ?2+d`-zO}=8maVpF~jv*`X&!+N`
zm>+?jE|`#}CoJ=IV)9m+Kvw*!ypTlKzS1>JBH%FhBkdC$b=?HWCKcwx&IMS$wq~t%
zov2oysvI5RTpqjE4D0c9AGSPYJz+QJ6ycq|9?r@M;1ESD7UTOEVk3NDEAY>P9k%2;
zBIYf=Mrm1E?^mU5ZHLVRH;>-F$GL`-7k+EZDeQ8o%HQQ1F#H8A5R=LIj{)j`s-`jh
zcSB;M+LLi7>?pnK8h<w2DB07Y=*8u*)i!hC0p{KIZ3=mfOo7{h@K`MQcjIa)%9X10
ztNfIiREdKJ$0j2rsm=hmgyHh;^!4sHACX|xywv64D~8}9&`1R`r{%|k6qL1!!<VP~
z;o_97Bkw^mWt`yY!)}R51w?VL6-D55Z*Rwv`)NvD?sV^X8;!ot@^Ap}rj*^TN#jIk
zsK2~|Y(*<6eRP&+n2A6|`9ia||Hs8kHjh?ViDFgV7vP`ka@SXJErD?lK_H+n8gYX1
zYX!8ifg3L1#{zJAyKP>moi;S?0QhP#UD7&ijV2EONFn?k8+K6K1QYmTQ2XBLN*s0=
ze*1HfV#Hv|Y|992G~0mN0>zKkTnx=FBnfk#z40e=b?C#bS~_aLr8&}FKgfMa`waR<
zkYg62O$!W*p(Q#QkGN3~od&g05E6tA3(S<43@KVj_6Mq~fIP^yW}G`RjdIYSlHYX7
zwK#jB<k0;BNa0{inzPm(UiV)y49fzz2FKAU6o^L5N$qIM3@c^$V|i(W8W~xqDnz-K
zzAimhQjfO;RWVVvxQj`ZR+Afxf~b?I6=4{33eCMDu!BZTf9jNj*zF-gp3v-HSxgt5
ze2%S=iFt^Yr;C(%PUJq!*NhArhR$)sSETR^Q(T_NfCYUMmzZ%fNe&wQtHi`}vQ;@-
zbGV$M6ffPKY6O6*uMCJ}Q)@=BdDfi*sHUdXZARYcljAGZrEiH|buW59v+wVpNnoD+
zmRk-!TE1R4b~g>@r9^BhZX(9gX+EC0OjCpW%?N*;dcU^<&CO-1Cqxa}cB*AXVKRdV
zpT+5c3|lpwxZv{%cv?4dKkrts18_5kukHw0W>)r9wDvjzS*i1M2FnbIbqef5;)y!`
z420wK&NeZWfS;@dF??pwx`Az?t1?Sk_?Yz$w4!5|gWL3m#Jp>+h0K{u;g`Lr-;Jr#
z!}@oV{p1xtX=)=a?kb-%d6E9}gy)XmPyMkudh5hs|3GhTQga}A!s;AU23VNwos#ed
z<*;i?e%QI<+t6I6r+KBv{{d5$hPA!g6F*~;=D0QNPnhfWm82+00@jMEk3v1q+Q7m#
zQm@bMEu15{Zcl~HcGZtWrjqx~v8-jIJBwcMq(2gQ5iYKZp_I&m*YneCtI#l=Db<dh
zfWG<uWyNMpvY`y>B_OT!sK$7HoCzfr8r1=2!>waY8Ew_-z4wG%8H7Y3x#@M|e@U)l
z-V7&&T3C%sFz(cQz=)tj`Phzz>!?aeV9&rQX>?ZFMV^#nXIyNd)RyS~Yf=bGW$ZbM
z#5_ba(@!hEN#~)CDpAoIeW@g3s9lXX9nlw05mi$%;OeVNiZFSo19L3)eKjTIqHEfH
zawR=!x>--EhK-`T<S+Z7+k1!n!KoDGtV6Lxj0co<*Wn9qvVGful9LN~wc_RG(il4J
zuy_qU{xVffisG0)s`wmxBb_r&{36y{C|G2nW}BHJS;T80?rK8*PUU04M8f7SmcRSU
z?Ux}8^k1hL>j%M#7IrdPbf-mx5a~Zr^=6WbCjNHIA699_zLh~YQdiCl_DgOBnI)5a
z`01>(Z+dBxI=r@aM|$-$GF%hZY}p(;`M90`N)6$f^7#Y;9W=e+w+jx&0vB2AG)m5G
z4;ie{PDJ?kLk&>JOfI9vQ47>sIv<?8XUVyjTW<XVE$^_-E0T9_*N{-YwWWF|RL*d2
z+l5`=MXo*1NXbUK9_hXA#Nw(`EuPk08Eg_tY7`70>2#zG9dPO87(jCPBOk`gQ08q2
zg^6Aublpx#Hg$E+GY4(CAaZ3T%b(bLOywM;&WL>^&NX^_qA<LS{Ngxmdqpk4K4+vp
znEB5Xb|oPa^0_hfllJY`N;&)=jCt31IpaiA#IEwT>Oy|8^$nN_<7Iv7E%~tIxztk=
zouj=<dRaS=q}%z!^X8E-3d(hS2ayk0E{so13Xa@kOi8pkDt_CzQOpXMOmV^+R{aqV
zz#fgGjcdEVdt0U*7F~GOL8SZTU7|9q*N5BDn|p0^z7@pzUwb^(#evww19_(_&(Gsl
znVS!>yiY<|T$5W3*6Z_&I~;M4$}H96JZ$=Z>0FMNEr*TmVrX2Sc7Fqw27R&|7})Y4
z^!IOMY*&to;s}~-hbTPyXs9VBT2X;(8wE6OHcmC6XmjyXhsK7S*xJhkOrtOA@lVyp
zA=(u_;0|lANr9*7-CSgLDR#+4g@##z9~5XV5UJ^}h?c8k7M!0Sl0#D(dk+q3w4~jq
zoVeX;DV-D=oE~Oar_)%gIB)mI4|Z215rE9Jd@BhV6G3Y8)A^t48RRup+B^-?Ha>e-
z6-5RwJ$YG)3mQ4@O$U)F&@9|u?E?(5G6tg#V$Q8lDOj&s;+w_(^nW^79qkx10E@O7
zZB57D-}$e$3;vEMtNrQULM|9Vm?Cg3v6krBnIc}E-TCxh<&X^u{L+85!{#V{TC+oL
zC+Q`mALz7VxKjL8jErh7_v7}rAlIjSji(r>-`p|a-&4(OXm!7Tzk!abAz{(*ioDm|
zrl`ARc#H&v0KI#NC|T=5Rl&<PH&gC-8O8`0bdEDipjYNkK}PekzP~6Cr_@XW_oa4j
zhA#l^^ccI(Og-;H`=~S@{!hJKkFKj-Zn_`CFRx7wuK0AnX)}HQ*4T`Hx*n$dUrdNU
zK+;1olk`@3!NI|y3q9<^FeFbCUbz<-?s25F(SR^CPPejCLbkk94eZSyXBP7s>Zg?<
zuJ9X!GhI61Pt;VucCBr>?^@;t=@w<-k8k*-vU#)r5R0%e{{Q|Z78d6JuCq6*D-%!J
z4&OIlUvYeEWX4y4*>G#vW#J9HZ+tiVKouDav?l@wHaU;@QkA4RUsomW{O95VClZay
zA5E$;+@s&E=kw9|`{Qv}Psk#3d1&hPsy=r>!10TUlb^M#S(vd{^zB9G?a5-UbZ{Nr
zV4zbO&;;ln_V9Zz2^mHnI+oGp*FpkFLHkZV)87HQPWN|DRRBIEsOMYrnMLv}CpKTv
zk7j7&>4S+t(vbX|@`M&ny<d+n;h1@1)OWqO%1niR7PO4Ic4@~`ZY<RVwWQ4z5$>+K
zYyi&V?(W1~i#<uu*uFp(@Vd<R!?$>^s7tsfb*^1`C0TI&Br8R$F#&#hmIt^F`%K>-
zfRy&<p}=%cjJf!q@C>aTuRDs&Z=1}_<{4OyS=+Il9z$iTDo=lgvckm$Ab3Q_=~m4h
z7%I4BgssqhObyyn)4k{mbRC{aF`P2a&HH}GS<F;wbc>-@h&6{@IFi!LX7drrl3D<N
zPbXreUE{eH8z!rzp>Rd5rKPw)NSO7RIICr3*k${*V0>8C<l8iR?L#5jO-k%UruIfm
zTC9mabXB|F7%ea3#YB`wOs0!@j-P01Qn8(rWoOMVmwzsd{IxSUlEMsf^3)hEmZfFQ
zW7(+1#}a*D1Pk@nWc-lGp~^YCP$`xVq6}4*^%-QiW2B73Gw+QM)D}=8xoI<S3W!P(
znx(8|iDHd{a!mH(GsmHf#3T2k5#|`4E%f?BQ6{MI6?Zjh`-KwMX)`<u5a7x0XZHC+
z(j0p05leUq&zM0=FBEmh-`<R(ZoXdU&aRf6d2(8AME>1+NN`5S3y@ZD4=KU-&~Ipr
zA<OOeq+kwCD#NG*?Crd(<bE(gMthIU_#gYSWpS)*V<C?7CSIAzv*e27;f^x-kb6dH
zYnW7z+(PTBA4>J6eR%a$HGrlPtr)e3LhF?k8+UKmRP~wAgVH=@yTg`HD&CSJomz`s
zOb<ORUWeF{7<E}Kt!Sq#I}5t(CTHf{u#m+jDw_DYU)r#r_vNFl!0B))S^C?om`F)Z
zCK6lv{F*@Vkkv2GZStTHB_iWSREEMurlt+#YP!30+UQ#u@45ljN=OMPwfkhEAx?w%
zhx*5)T9!f?RIj%*-bS-%D_b->j0S61iRf!iy_9*0&}|)mL_A}lkE!ud^L57bv!1vu
zX{5dN=Aal|5zf8nHl;bE3fWLCTepgp3i0f@aC3$zOIHnG=5M0hDC0H}-&CKJEQqar
zkdud)2EYmryL~ljB4HYTkS`cHEDn(q?U9GE_B7Y$kfx&+fM1fRg(Iv)TCrfAR3LRn
z;fkd$r=Fv*_ku4@v4THAsO@z`()-w5?c;E~i~7y~IH7HPX`~nCpGD2=z8?QUp3-w*
z{;nB1wR1#wifw1E7uD%uYY(%LtBss+op1dkZn|#Tj8Fj7uJ5%pbua)ym3*5NrdT$~
z$B|L_rw?E4ignT2LnNg_se6v;hVICSIBs#ua8F>x=XSWOc#lv9_BMmeHnNg+NDB@{
zIw10;&U7*T<OiZ1s#k`0Q?8uR)^XLnxq8ivAPU0ZlkIg)h3BuS%;DQl<s`-sYW&1i
zUTgT|eaZR#pGDloy;b@vUF_jN*eVy<mxxsQC-=PASGnCvSIgKpp=lv{Dhv(Csp9F@
zlt)LjEveiv80(tJD$~iYwd26zWQ0u{&rf0g-T=S%b0~WvL6`x+ReSn9q~dQQ-hU)v
z_JeMLpS^!Kn}u{MJ-tMzlZk6Kjj`|vtLp)ity(Xr1}_7Oqvqe^_<iojlX2=d4E$nE
zF)Atk+{$Uih6IwnyZ9$Hs&ev_5bNm0c*T+7B|~d3a<LGwm=;=qe^f>e<()X1k$@OU
z#`)_E8*xc!B8?5MG!myl%$rlVnC5s7uiM|8QD2C;l86-E3e-0SGz4J1Wv@>PE^}$$
z9+Kr&j>w<M{6;@z7>Kd_N}NcdMmLKpabWWO#~HTYBc3`nl%GUVI4anFWeYe|P8-&U
zSP+^w^Ndi(_xF6u>r%2BHLVuZ+g4%qEUkFL*qknrdrSz&@Quma<sC6*3d}g3*{~`1
zgTTve@mp-yhy>cwHisfSACiCtWfT`oIEZ3;A&IBYV_)EmC!NhSz#*rOUFISQ8v*yJ
z%T9w_5(+AA`UnV%dJdIT<S!^O0lG+)Mn@6xdEgP_9C}^7$>>VeI<sMf_h_oNlyjiu
z*;>3B@o7nHjTsHXx0I0iGE;w%R^h90JTC;~EwPjkOEiTI&EaC1G_H5L(c~oCet2m7
zez`lL4(VkPwpKJs5Hc>(DL<TC)Q61U@>G{;=z(5D7Q$}7A8G>6q>fJ>rSC2Ab#tbq
zYyS=Q(8<|oy+X6qZnoYw)En=!eTyL18;)@a-|^ct`tj<-udnOL_hIP|e+U8;Wt$7%
z+zI(-ksU-FK6KNRS!gEPhL1122?pzI$4!`OjQnfP6E|Z(rp7|<(iOs{=S9AofJC;(
zOJ{^sgE{<pL$n7*C<CoScn(;Ep$SB9*w`_AD0(F#LA9%3e?U3JuB+if>9G!PBG$Lz
zZe7+k<JQ*kD&`r<eD~9X!~nu+yl9NkM*8RTSh8%j$=}zh%P@)E?tL+wP~$n*I+;_G
z$noYJIT${|2WgKpojc=rH-9KcTbu4Hg-ZrJlzJ5~xY4~uZmn{cH#EOz+~$N0ff}QC
zaDVuPkds+}FeQJ!VDEmy0=|GD!#_6rj~5*He(inleWSrt(u6D`iVmT$ER%Kh3@1Z^
z){D9o6r{%IohZTlS~pX|p--DE;`5Tbpiheu;S(Zqg7e&UB*D6M2N<>;cgO6x0bXyb
z`>}lyJ!m#V%wFHv-x!X05bhM#rLV_4z5Q;}e8q~%pTjGg*>L6rqDPoU@EkF>Dn}RO
zd#N99OKT%yfXkfkNfSjRO3;U~v&__gneFwqp=YL=1X3Nim_o#Wh2T#1xaE7)+liT5
zF@3NqByqX=TU-7XVp4@v|DR{sGMPFXUK7s!Q%1&`VJ4-!uWE|Ijd$`+x7{xo;ZE$F
z<vw8w9G-$)^J`^j&^EdfD+x6ATR1r8FXsGT#rFwrZUPG<tuEDjtWk9P-qOcR4k*uk
zs8m#<X6dk{7uphjq|h34|5BW`HtVipre%Q_{MB)B+Fu8mGN3?Q&Xd_pF}k*OBXWwG
zrGT+~lIlPEG3HU`ZIPo39o-@OrM(nWnPrUvziWBhYFA!4wwt9;3Vp%8xkk*yGNgez
zcoj_vFNBSuf0ieNRK&`ANO56$hBU4X5<=^+!1k}5F1=axxFaonQ4w3h^x}U)QQ}1+
zkZZeV2|!H2*rpB{$&ppwwQc=vX|RrV0-tG`N;ktE>KQxd*h3*!w>4E4tBVb<6<nKw
zYt?xmfDOh`!&nykzWC=N-oTs<rOCS(O;78~I!E>uX^yK?OC+6cKWxF1Q<JXzS&2H+
zXIti|8>=-PDio3tLR;cfM!0iQRqMs;#X;Ds^G6QiufyfODKgEAL)XdoivtzYtmEb1
zf*(9IWNW%})dA>Mij@KAaMY@Ulp}AvoSB9C?pcS_7n9a2RrX?4%&NZ5N7?q9l`aQ+
z?1Hc{#CAB^>>))D@nwx-IUW$~xXI#03!$>oSGb85+_^xw!VX~On;1Q^6R=FRwQfR@
z3j08~&9-vEFsv=;1TEH3WYdV4Pa+cDJ`DqDP6%w_8|VZgT=_t_(hdY{W|93@Ve{M<
z{0YUXtAxVsfA3Tj;X%kWQz=L&WX2DSd6^m1s)Y{OVj*01<^*L9If9sIVR_!RpQvft
z1IHG212vs!;{#_NS&$E$DWZxY6fxxm!cLh``)hNM<Krdl_O1W3s6B%}pNA7QxrD%m
zR*qpL?NnCC>ANo*quLDVa2PNx$))`G8Dzd|tG-P-IXjf{x7Uu5(>%>mQvaiQwyCtM
zP|s9C!v4n^?2zoN$I+UesQ4|z!=86@f+O96avyI^6PIoA=JU{jLQ%@RhG~bUD-nXm
z#<!Mx#?!ETINmsNHd&MT#kUG=qqz4pIr_=&-K8sI<G8Jq5lzqIIoymCMQ%3bM41CS
zbWJ4h?T{?2-rlC3-QFA^fO)7}ob9!c88XK2)5c&GQP-WshhBe#tF3+)-jA}1qWdGR
z%HytpOxzxhQug(4z^+Y7DMyNC458-Kfit%qn+xBnv?02-Of(8_uS=Au+)mudnog)6
zw{I=@YHp<6gj0k3=c+6@XiwRXeo05aSx6+ziRVKwo+|a>+jCExG*ieTh&j;|l2-<+
zp<<Bl=HM6Eq$T<9YeCCMmO0cC(%Nk>DrhSf?qJmunZKDd+(Fs#1D?j84M>0b!vcI<
zjXxO+dzScR#m9Bl-_UH5FoK8>dfD>#k^s^0<ImsUSJMH&{4~YG|0$7TV*P(OvHy(>
zV`1T7|F5_oy||OkR~-+x^$Uo+V4upT%Bf^J_15;8J=-fe<e9pU2$?ee0`LXEAacsO
zH=uM{V8419fHuA!?_m3V13JGw-?v3BqD}UoC>R8OdZ0Q659<%^i`g(Kz?!9X{l9v?
zUZ*-+9Cr_va%jNQgYWS9XJ4{W@Wms5(mVWJv<CvD2gshveFxme|2b0W_7ohKns3pG
zvc5umWjs_Vi#c^M#I0*(b0Ei{|7AAy{eJ<cp(Iiu{9;e<567qd3@MZ}q|0jtKY(`>
z2TQVy$46s4mMy)XkM+~}^nMK}aApgB|KR@pazXCt9#`ald!kR%0*7Y|jGEv9?*=#@
z&GN$xe>=ZG$u&W)u@wT*p$K_FXBd*e36QRVhAIX+5VnPCMrNebGC2{|95LX)vsdO|
z6?yDynR2&+x}=JM+Khk$i?^T^Sr%~WgbqFu1>k=i4#J$)9nvU5VL7M93=Tomm>E}m
z193vXT)*V5d<80Le1{zA_nN8MR6Q<L-aFIp!ToUtbI>eURv3rzfH-2F6tQt?A+adz
zX{7dIw?SsJSS(H{7Him!%aUpmOk0yqb}Q%-XhENJ{ZJ;Qneay<&RUKC*TEr{`WK2t
zS~65=^(?j}TlJ-$U+rXB@F8kG6HX4`jm+x6nxDAWx-~^YOp)V&XbW=A>M~}i;~<YA
zj0Z{&bY6&;KdH%>aX&oXHmjaC9G;eb<NIPs=gLFI16;qjW*~s<p}?9TC)D&!0z1ll
zhj$!t^2+B6OXDu{7f-X9G^q5MW<q|m3dbZ!C*Yvo%XNu*!wC`%MLWt4HqPaMth>La
zl&=PwWKFgRH7}t)Qbrbs@i2`qhqAy}JXo>DAz{ePsN+zJLI1XKjBn|&;AJ~;{X&))
zm1Q#bM2=$RGMSlw{FQZ&7wi%<^TsP2ox;MOU>(G~0Nq}JOJc^DgI!(CPMM-C;xSVB
zm}R1^RL~EsP8CnLVu=|LIX_mzaj9^MFlCh>@gr~qpBCD@T?8<Lh*szpkK%F>&#*G3
zM6o=JP}qTGPibP^I-_yg*|1+>ftx}_czSbvt6ln*cYJ_GOtd#QJ`R4&L-jqThK=Sb
zP3&sG6@rVw+aRoTG6mcF_)RFHP!xUSN+u(EhqzYZ`)ohv4V%yI==Dy+JwD0P@6YgH
zzBf!cf1MyL23|E~vbb_9S(w9Q9*13Z8I`WJe}?{RmIh22%l7f*{vr2^L+eWtypfes
z`%BPy21IcDGCWWZHI`@x%MW-bb7>Wr;zd{jMZm4TFB1fw@G_q?`!?O6A5rwrKnVUU
zvDf%GrHp6F_<89hQnx`;3TS@GFeg-;H7jV>maYa;<*eg~m3eMKg6HID4u~n(j})hS
ze~T*5JV=2Y5gWPYkWdq*>+23uUCnuO3JCQH;~9&tv4Co6W{7O8pxogIVy4$D+dY+?
z77$y>iU9kQQdS1jr51ltXqj~L#I{L9LO7tZ!)&Z0`MpxqP&$s4HdKynTA`Hm*ncL}
zhd?-*5A<xR+IqgOvM;=#V>fP)IJIzLzUBDnA0M6PdPAGM0EslgeUs@PSlI|>5^gfa
zYx=|DR-~(3Aa0X+;@^6?K0&h?C3^4^nsw7@3oD1obP^YxIPQkYv{R{NRO%K2R{*pC
zMMeW$RA+&LxrrMbs;#f_(C9s4vy#*V%u0!PT8LsH=GR0G?-6)0S)?O7ED`qy6TB4c
zzB!u`_+SnG_E8Jw${-`1T=pvI9;?1-^cyH5ZAq!3!i)cA{}Q&lr@%}&6zQ|So2n{_
zq;~6i0>=gkf75-K;75;c`H7Y3m{jeKV7P~@2)BUPIJ<!YAAM*CUv!$Hl;xeeQV+28
z>jao()$yu-fjWz)zv%O)e*`@}QmkrgUBu^if}f!|sS^=m2yXrwFqLR7%``3{MlhJ!
zKbfy{u5@)Jg(iK_GkDteyl2{J+{H$MHk`UmMzV>ePYt_<f=$IP5PCHY$V<g$)<6z4
zNfly7?F(ETa-Lc1-<!Zq9Ef$OCf*!s`Gc-0dpWpalD*jtJKejrAGmI^guVg$Pwe@o
zZD{Vb(<axKVBsyJW-{~o0@vu-GwLWoKHFw5w-DuTe_l^iUblXl&S`V-W9u1UFHX;R
zB|WnyMe>M7rnwASQt7d97TGkxkFU_04m3L50u}-`JxcoD>dDX7vS~6=GK9B`hbV*^
ziajKPuZr<mj+1uKt9Rq@u?JqlZMsQ1AX@nPGKEMQTaNnBg*DkshpGnoMuU6>46FzC
zN6ptBqO*B1ohO}q4SZ?QM$VZlcuGy6-W^k;Cv6n(l8H8%E^)?Q!UuGfc8Gw8Md`AR
ztO3{5S>dQ_xYO;JSz&sN<~h`ue_tnDePh8tZA<ZIGmFV2PlgarGh18e;<dFB|Bsf9
zO2{A7wKAx@QTyG)=QJRGBRlJu%d}Rip|0`g%kKM`$kJTtS-cX}RoE3!y&EA?UWDX%
zm~}o`S+Xl8tZqLs0*^n?2&Q5Is+F(BGzcCe&uy;$#00T8lvyE+ohCffV(7Nc_o2})
z=hki4?jv{@LPYZq;x-T$nAkPqmZnEXzd$b-6Q=LTD0ecX#(S<RD{tf_Cp_;%35iBp
z?uNhBqMB^u=6AD$&1QkCxUfAzD=l|zm}Zgb=)y@1Ula+fJxn7qja46p)=rTKk1&+?
z6s!ZR%+wj32;n(u*r-jb@evZqYHK9mEKIZBj-Xq#d?8_D8G=M}>CsAR8HS(8HH!*9
zPbcHBW(Gwa(ZYS<>{sPAm7?$ma<UhFJxZlA>B-w5+BLGkE_aN6#aOATLaI(sV}uCE
zgnC=js7e;X^yZff^C9r4RzNZmguhlQ&ACzTBU&phM&n=BV4Qk~NnZtGU}wr7<d=x?
zwV-Re^6VqOa9%GI;SV_Xxr9L*zW4yW7vX??y4V%af-ALhnS+tpkSStDsZ`lQ#uRXI
zRb;79o~k;zEOp@C**|Gn!>Kn6?3($qr>f6M1B+|VoQKRV7JqlCStK(qAyP!lDvCKQ
zG$%<oH+X8PTqoOh;6@QGv=@X+07Hk=GZ{Y<bSE<SY@xlEv-9Sz-$XlhW{@pE#b0QK
zgi@2W^9+fxngq*GitmZ!Yt|KvJ5cf8ed6k=y4+B*GU@436(3z8sA#9Vl!S<2EY0p7
zb8L$B6N_$|$i~n&>9W@E=GeG`eIOK*Af2fI=I$}5A4oo8sJI7h#3RUlD^uJ0ztCzM
zNn3gs-ozH5mA{q$Do|zT9qn!uj=jmRN(e9V_Vjuyf$OVMMD&Y?c|l_D=#ygl*JLI0
zoQWvE4oO&CNqeK)z;GkYn4x5_msNzGp;hL4k?vjRrumM*5+>7Udis_@kogFz%xZa1
zOf=AJN)Z<k<{}-0l@XT`oyYVNXR*8o3>$-|$y#~MAet@3HBa*$WR^BIJ+C9mlU9gt
zmg%$UtWJphi;e`m(`UVqI(FA5M4|w7yBBu61&@Z(6C@774iJq9EXV~FOqVVno|5#L
ztKZ_{%{*%__k;}d=6rPaDog9Oe)<5mms9(%eO3pZz4TB2dx<e>zo?RlBesCn9(WD1
zl`9=NsN3Lt>5AP;&uxjo)>yIEXk(#~J<x>aZ3fBvPXtGdnU6)Eq^Bgh>uFsES``AD
z_|Fyvx_m}Uk-I0qS@H<lNGlE=%#o9$wuEn78MHi+S#~taS=gJ3IXjggdG^K^p`SE;
z8SU#~lKu+Bd4GSpi9zkM+|L|)vLWHR{gdTVk!UMz?Mn1?yS-MH;kA3>dH3bEf#wcI
z$}ftC8P78c?sv<R^ugM%&gnyo!%g`5h9rw_E1d)E6z;JQg`-S&D<i`w3v<w$ZYSL$
zJ;vj>5)mDu8kxVus;48EFJs7xrW_Q@(lW@apL0tDA{QItBY;9d%3gA579`xDS<S})
zH?j`2JUND_oi7tL1H1??<%`Zf?t6jP)kcMn%Y%yBOceX)akUF=ht)<@J?s0pxP9dV
z?*zr;ZG1krydg9RpIgG77W@b~nXVslvk+OlKt^|Qt<rWGwKzQ{ZR&c1Y`4n=6cR@=
z9x1AC4b6;6>RrW4(?e=mN;<VAg(=-ud-MZ#TvE$VB(lQpAO$Yk`NiI*`59G@TusXE
zI!`$`dCO`JG@!FH*ZLw$v!iTy&k+);-@o#x6-3E1N@gG9PD-@OU=BxyecQ>{zn7^0
z)TCnX(<5@wQe~R`ws?;?TTNr%H4yh2uR#l`eLSx8sxrhXK~u%3vpQ7mOqCsT?QP$2
zomCrK{;FMyeRs2SW0M5?C9l@pR}8KiiN#eZ%U&0fAhw&98p~IXFCLjumW`}kCzn^U
zQ)r8{%7>9!v!#ERq>UJmds(UWnmrx|$jmtQ&IC4k1*5R*5kP0_z%AqRYq*guwML{|
zd<~=yJA}!-ZAF%S{l`G@Vf`d-JJ#~7>qCHnM~Cu$e9-mNyMD5o6ftG1^cDX77*LLB
zPWGR&uYVn@{|j^eKiL;2<A2E#(B(-!X+xg7(}-#AcE$^lI96-jtIF)uOHZBThoqNE
zw4-VSp=k{0`z3fhIqoPqHF!OfbO2efzY8?MAYcGEjVG`-5b%AfJ23vAU99Q#e0mTE
z$1fUAebx`-P($&^>2(A8{eDM!cz=5U@LcLp{OCSH|I3vH8EzXAzQEB-GkDNnsvv^Q
zzEd9m+vn|Ge}p#@dEd1@q#}8R37ei`U;EJqlLoTDr9uKZzshxje}1T-<rK~*@-H}p
z;{XQ(fv?YDPN7p?)E(WQpXqQ%&=kX|$>?VQ(p$R0U$6IzVGk19i<n~;*>C~C)p|a|
zaH;@mY_1@Vi%H9^B`qfkKK+<y*-)*hNvulfCnF71#Jj?EMfYz!bk)iXWP8z6UxA_|
zSxFXR7<5j#eZ~Tr+{psy-tMI$o9P&J5^f3oU-VMctj;>cszCmr3mj7@J>?Fu&J}lu
zk%VfErStQkDF*LW+4Ui5TcziN3x7AbI_&vT<E+NfqJ+zCe5}9W<iO|FDxjBu>7~#t
zRgZuh9VhY572|ce&{%nt=@l2y=yhov4%~@3+A4NKOboFpWlS`5ughJ|bi+Mi&b>$E
z`=6a|U;e0kgGj9Op<;TTk$N75()Ud7JUr{3oxU{QQVtJv%|H{qUde~M!a#bDu(*{=
z*t?D=qJu6G$fV7>rrD(yDYMY|Cf<OqRfuqh7V#Bv4#L7mP<ZCyH!i|GG6yl-tG5tM
zeA2yNRDk}jiyWwRlZQ6!!cq)He9(cZ0(^Uwnf-Px0G*dMIlvmfXl(TA(*QdA3IsJS
zREeoI%y^}LpUJuOMtAPhFv(za?pueJ;1rk9#YnqH-7seSjYjfFY=FWQKIQ`uZn;fN
z<19w+a^`3ov$``@DD#YW7u}>|G@%g1P3diOpD9j@Z|<CL()F2yt3a`Lfx)s<Hu2n_
zeFXN(Jryxw!xm2(I><HoF;O`yop=iST>a$dGo><}Fbflk@UluAVy1dFok)ry?5siI
z^{o^dL_x$YOGSV|gyTa_piH1W^kb@xB=n9`)RK!G&4h77CS$y=n*j@hny#s_N+NCo
z0cU$7KlC&}u$TOVXc>Lzv#>x(wn(p^NGFFXKsHT(gmYOe)Z)tB_V-JH$xw9`)m@ON
z#09mWSeek>x-BXdU!N&=y%{RMjv;YldC`@pHKsnw8mN-OK7PxvzDppw$z(gD#lBo*
zpGx>!YigOX4b8kl!k<(agac`6DR^lHm6ReHNx1j2Rz9%=XAw<1>exUk^>z_}h_a2>
z1R2L=KcCc#Zx?)>Y1$~#d29OqG-EQvI(HZGl_>fWA@c~Cq=keUg`3b1Tsf6r87W$V
zOLmm&K$nyXs~c;&nlzPqJK0tPl8H*8$sKBi*U?;3E8!!&_3S`qH~b}^(1!nz*aDBn
zMQ8`Q<UnMX46>J*L=NvVmrrKJw*?{P!=pC@26{~K#7$_VW$rT~79hRnQ7{4w{VSmi
z<+X#eLxwrTT)*`<aIxfe*|LuuL5mh87S!1y<J-X3ngzY;!cZp3>ZtQ8)+zbovs30r
zw#U3hdRPU={Xj>>CU2_Mp>|QUyKpqD9Mv&cITi%>FxVR!HbjXO{O$YHT*Sni##bd4
z6X9G<RqSY3B0rC{Km_4+%96p^=L-?EyUA7CkE(F>r+qL>+^HRhS@WQm>}L6xRGq`G
z*%FLFMn2-qJqhM;c8N&HAx#~hGK#JQ6mMq|4j+ePxYO4(ok8g_zhsZ)qmZq7Pt#|J
z6EdaIuU{dVEq$l+{%Y73Z2{t1xez1YWng5H&;o^vQB;XLOw;AATSU>SF56@IIkkwj
z)zDsj0Z|5!Bca~^I1_a*dD_nE_eNWgHR_U|t@{Hb<zbPQ+EcaHlVN5-d_0%!`b<|w
zG(J5hxL+xJ-J&W^!!3K%IMg9_#?qDNa$!RNIFoEtxv-=tE1tx*&``3{hScEZ3M^Wl
zjqtNe=#hqsVuHo;wy_DB9K%XgsE&{T#>6FVFdb*(QWf~*oYOz--}dd1^78-Sf;m|j
zGvXs8PSjin2oDQ$Dc2>RO!QsPEtc#pX%0NPUK?5<h?Xk5tmaw;<lpKVY_Yko-Ho(R
zIDDP07BEAB6B;Vhg)e#VCYhalO9xmKGkDqr5zBMMhH=a4R2?+(S4a|SRkcAu7^qu2
zZ}dfRf-*!nrzXLaJva2~_Nap>6Y$?_8zx%&&u7A86%JD-&zEzI)>`}ABVxsAFLsu_
zEE1F4>&1A9-!<u=N5C9Ou@3Pa>bDaRo&}2t@AFwLksSYqq{fcpvPsYp8D|uHI8%5^
zl+R6$_}S!-vHaVTW^v4GcSYMZUpFvEPB%VS2WcB$DqB>Hs$;jLyHQLWN_S)Ee0l0K
zXcF(b&}N>Nm^bXibg2V^8)W%5j=(YssVc-SA%D4X^6qG}ycR4(gt(w|hMQ0F(idA?
z_OUH-dV~lcv|IL_&szd@$;<YPvEXEYz$V69nubuT(Yv4DzrmFZK7-<c3;X{i>Hh+C
z@Q;oSE`L|B{gAl`Gx%|Pm*E6`1j16e2CBg;>-LAwFjzn;N|tRv7cNc-qy4suh$RVT
zbZbP2zJBo@(c>FNnPP((k6GlE^^mvuh(#f*d#?RsGBjuX(KdT$^E3%-=VqK6f|hiX
zw@s2Iq`FJ8<BMTj3?!o2lHdd8xso!We?uQ<<q(_215pOTm-Ss)cN-VvK4{nF^xX{d
z#@sC0-gAkI_47wlyQ24(zK8NS3AtA@`)lQr_>G2<{?boFhqKK0qBm395yK6Omm3eC
zUd+jj+DCd<aZH}I`-AFCPjxvxSKHNUC~;-EcLg!Nf3)O{QJ@iBp17DqHq`K9b2JOw
zyl5%PtVgB^Zm@r<S!(_{&(h=7C!Z~!9N&rC&Wu$q*Y_(<{4wy2kt&S7G7zS>w1#V1
zhR#jxLd5cf*as`t3d^CZJf%z!e02EeY-LO?UZS+0_zWfVQkhA(nQbHF?bG3CM`s$R
zrn#;2_Quf<waLHfRnDd3_OUskCLN+Aq}ed-u$<9~(D08O7Nf-NV)ju=<5zi&i18E%
z)cZ-(&t_;V+TE#kJN8_?C$Psd){l9lcXU1$Y6ojJ+k|Ny%3)_$WP5I=l)Q|8q*?X-
zAek8%iW$-=IZgbk_m`Tn3cyk}ymg{BT@3P_HZ$0YeF3Z9Y&rjq;LF=@#;gc)RO|Ow
zFVR;i?qylG(b)5wrh}+VuL}Lm;T|?@$1Slqx)&L>3Bz0Q6fzNgWLuLmW&!qYSYYGO
zcPvP44Vf^llv*1^#IWO$HU5^rC9C>p#ZmMmsl}SR&zHMPB*?0DNGamNXM`V_uBW}5
zm2Yb;6I0gF+|x@gds9hos~If}t-U-fQ?wQ9_B%~5vN#rfDqr!Ii{ryNE7M)?l6gB5
z(c=XMIF@`|oa2Jeyx)L*1)DhnUWkIV7~_IH%Hr-aQF84w$-Y7EnW&WKnboHo7qA2+
zQ9{^6Dh<M`Oy7%KO*9b8wyc24L(o=xXpj0r;U@fDXqn}>-_O>u4H*kXh>8E~FF#hP
z4TXe612H3lLwmZ8mQ7^$oQLMhctXxgo3}vA!7|*OYNzmL`*r!3etKbyEbN{()2a??
zd5X$9Ue0!es1P~BaTjV%#j@nR)dh)V1skW1zi&ya?QgOss-ap#y3SxmpY~PiXC=R)
zsL=Pui4ECH1&wk7<bJl<%ZTiPN(AY!tH%^eyF-}|qr3?ZSB%$k<V5~hmUZ5GUr^O{
z<P`QF9f?(B*XXmq=*;9>A>dq)Thbrzcgd-z$`y@1M-x4#lO^b~MZvHHq|Bx95CpW-
zFBC9csE~;KD$=V)jv4Mo7yf5;#+jcy&1tO<u>pnBg9$->aK>QK));ktNQJjKL4v`O
zTrTwc3X+jRgQiGhVpp$YkgB*BM=Xh8f%_sS9#ey*88ecYO|ODyljhak*{vPvss5U;
zwWeOYh;jTy!m5@uS`y3lO^n!XN!w|R@WJK(wN0&+k}zY~z?VRVJ;9uklc>IwgLDb)
zDz#H7l_Oy7FD>r^!yrva+4ULCM5AT6OB9KYKxd0n-o0+U8Hp>mBB)W&5=O4vph5v%
zme9e|C(hnCy<{tYrspzI(X5PU7yoE<mGy`6??XK)!P;C_bp!Im`%fZ4f@HpWY|a7P
zz4F_N#}!_rA?9ug$-@OpSuDZdhQ0hHC#?z}z{f@NS-kh^Knp&^$EVshsJMwYHNy}!
zx$G9&$HVJIGkPmS0PqTVp*M@Yo~{+jnjw2mo{V{2wTHUzTuW%{W&VKRc<Pox&4U&P
z&shWNxTY^qMo`_ISotv;^IoiWSIM-|X=eznC0EOpa7jfPvl%6t^aT-g+q6=79&QAE
zN9>34w)V1&S`SJ}bTAv=Q2s*#dTET!eI1uCt;;i;NimHaFk#&0e4H0nIs4O~jgVWg
zRZ6O`V(=n^sd|)@l+bF@`8bj|Yc|NNFkAA~;V837ghrk0{5eOW+}uZV%0h12)JEQQ
zi7j8D2>bX>533cys!8E&oQaNIwsSo09FbhGs-*=t3j@o^Y+Vdk7I>=<o)$sLWja@S
z2G=cyytf!#V`sU!s>F24U?7)~_>YHc7klM!in-jP{UIB*HaCQ}ylIW}VvVTuu<2%g
z{%FMs$(!!GO;xKxp9lgxg(efd0*HBr@VCJpA;nNMLssz5+pt4B4sv`yw51^0n<^8c
z&7GN;;*Ed0!d@W%^AfJ~wq|~z$5)MM{qV1bnc?=IkaqJ7k-HtDGC09;uO-mAKKTbP
zjOT*c?UXeb;oYF+MnchWMm4sH5KU^f>SqrS(CsD~Mu*!yp}z=!;6C003suop54~|1
zJ`iy1I$FV>ZfpHesjqxyJzMolT|H$WSM&lt$$CMxeG?Is8Xao(-20`fA*+0_*7T2;
z1)8sR)vbb$U2>O+#*cfKH^<Mt%lX2`eadSi6khV?|GA#{2pYZXJhptd1;D=EO;$>Z
zhFPmReHG=aF1t=oE4W;+FpMyZk`8|b=lunC=<6K+kI6FE|CcN?bFlw+`zN!yvvDVs
zDE|vmoJ9tZuue_MTxDC@ac^@d=9;~oG5c(A*<9ZrfPg#fmyIdq>tLQH7a)uP5JL_4
z0$dJ;6&&8~90ER5tQ%&m8jJxy<3NDGV}k?F<v)WUj3h>Y9}~a}&H2!Q+qjGqJ()g6
zKhM|7odGE)?Y|{y&%eg#2LUMeD9h!5kt^i>?l*?R_Xw7sHUHQM?~l-qScMXWQcisc
z8EP`O3W@Qql~0HL0RQ7gnH(?FzA(g({q{SDzh1?7qMT>p`vE<n+tMJ(@^Hd95C!4=
zKWtwQuO2Yk((}O5@xO}Q$oLK@mD#Y4GvVTko{e=fhm7Cn4FV8O<`@KKXTd}nor6ZI
z_0N|;ctSy?q^BL~*J;2(jVMtL0KstU-rxVWfiL0#=U1yXlO-P2tKB!3@Cez+e3$J}
z*s#$bH~w$AX3|yoFFvz9N!2e7Zw8Lzv>bSsn1-2xpan18yFajpneEX^DoR?J3z~^%
z%lVBzm-1TuC~(-Oi>RbP!(xh7wvpmil%!}cTz+AHUr17*R*DjET`5Cdi)VDTqa7Iz
zV-)V=oR=BKu9$WS1W1M=?A9dsu)EzWn#r>(>^a<jk3f(124+ANit8qr4^ON_{M{!h
z{;jvQqW+AF+7$Z-{xuRonD`C#gibrIMH4eE+Wa8)^8{QX-h#q_CYkWQ^R@A~Yp_J@
z-k=2ePn8B8*QS7#GZv9n3|4$TL8X@GF%vQU7!X-m!mlN)3_x@s>Cs;)lc+>^kVTu{
z-lD)*(R|=sK*vfztGNd1Ekc%MgXc`iN;THje>Y&h>Rhd7$1=Rbj-?mbw?8vm4EBHZ
z07Dqf(P<-;cL+I-ra(k_pXf*i2qgDjnocUkZS~a^Me0ZjC_F2P7&32(weE?<mI9n5
z2^hK=CgGR2=o6Tw!Z<xuAA~K7o5Wr;*po{miBlU>Dk)nrWP_OkjklhgLSe2q&`b?G
z9XN?B7185%AJf|Y#;}Qg*8yS1J-YSPKJd#Qh)LH0?~0Q`$YLwNoC{uuYO6?UOuv&p
zk@D1cHwp1hG4>;w4wmD2ju(1K{;80iYYug?Ym7^u`3QZ;A<V_GJi|>EuAyYbyiuVS
zZdk4vA{&<WS(J`_1*_+sdxE(Tf;6r~=3#^dz80hddkECIT>2l3y;G2AQM0AH%eJx0
zwr$(CZQHhO+qP}nwq3Qm&gqW0H*WtC_pI0XxMIc}nIrQXrene2ok*Q=JJ@qj@(q%2
z#kmT?0qe`<AcJu<IeF*-OtL>l$O=u<ee{%yU5#G_-FhP9GG797IeTdC8M<b0GmlXG
zT(h*ucU--4L*734#r9mBvn)-fDxW!m`q+b)dSnjflxe62`99Ys;_)V{w-M#SNm?w(
zjVXom2}yRR(JR5~RTsuh%p3Xh5<Z-^O+{dTqgq6+^qPR{0_PEjfC{T6bHO3Y>OO%%
znpi*i0hAjk0|H_sm_a~3IfMfNH-?d}1mh~gMP>B4CW1uCXR^-bEVGCVc34u|h7sqo
z#%H>W*c%NjqX85o9Rx^<Z4@Y2?#XExQ<);UlP6pS*--`y(>~_gcj{x|YYT+Z&9jk&
zWG8B%r@Yon@5|8@C#<=>lB`KVmybvTbXk;TB<9O4A9O3ss<mIyHW&(tg}3j~5`P?X
znn;>EI~o(Ri?4mM?qKJ4m%{p785^C=Aqw1=V<#>Q10R?-VE)n?1-eX-KJRQ0ll@hk
z;n_?SdMzStHA8Ua+??+YCv3E#9K~*>#7<Y4T98oHl9nIlzz`7{7kKXSD6p0H)Nc1^
zLF`3vk#C7Ae#HQ8S$9naN%=~-giI?q_^bY;Qe6!?kRU4S7}YdIzM5%hu|!IeP<?;a
z<RIwTFDn|$yjO0=qyWLsg%Dc;o`~1s?NA{(n>c3wDi5SgT!B};*GVZs89{#vzfo|6
z%+0eJgN~sR9El;1JY_iU>PW$yYJS3_pXTVpX|Cjm!qgbCA?9?7a-kXCF=H7C?(C^L
zXift--E}1`jB@rYrS!Vbg`n)l?N$bQ#V5$T2LPoUY>CJr2}~ZT+-e#KrVwC}|5qT}
zQaD4JXAyE7Br)s4Y|w`sg9;q@dd4G;bnt|NUs<_=luhpWkgo~NHvcy?6oOW#?oU~l
z0k)8-0%2KdBj)w6xdx%%U$<z298#8uwpptpru-IBG>by&jZ+qd4wfC1!<6kPP4@9Q
zcv-C0Y*rpreaImPXT<Xyff5F4qMBXSqAI<u%N<aiLLJ+^BD7POF3>CeJ>dF4f21KG
zpjydQ(4D$edg(kf$`?|(A|`xPizQ3O%x3QLi9Z)^GqC4@+tq|iKR!#sFqqY28xE_M
zI}?{=08^a$jixjwBp>{izM-|0SO%*x6@PpL?%O>OMaagXMMRFMpjmv86}X@h@l0Tx
z4837{YqM(jsaYTaPilMb_jBLwOsTMAT)teXJ_uI_Lw(`?frnbd`u#m5w5!u=2b%(8
z63?=gsbdUE$+17y4RvcU)?#?#|Ew#+(NWym1`^iCr9>oRo#K!W>KGrcvPFz_fy6DQ
zoKSD^oY@F<{0l$pQZ$={ozAIQE{$sF==*w1)~%;;VbfOL-EHsa`IpXmn)#3>TfkER
z>(J(ue%Xcx#-7RHnwlprX!s4l%2qsRtBcQWoGX{xpJt_u0L7wn3=^$VMZ&K_ih%uG
z(4N`#l1~f$1oN5}F_j-waeY{vv$Vg#Gq8K8Q%ALqsPng@*5Sz2YF=VA;I{GT=&;;h
z4I=zEH`LQ-BZ1&5F~K%v&i^RcP_?`q^a*=eJ-4cao&h&8x`qQ8<rF+470Jz+Hr4it
zVt&?tKc5Qg(+f}BeDoS+bxdKON@!}`@$d7S&P1)o-pVk4fqF`7*|3Rh3Vzw)hqyY+
z>jz(k!Ncz|h_30@md@N1gnRs@Aa&oZ<Z5bbG9UCX?meS@dL^3`<CTn?p<VZ-j2oLz
zRfO877;nb01D-99Fwm+RmI|Q9OfGUHqd7+bPi$8p6W^9#ffsXofZZR7vojJcCj^F$
zu(yOi?^T{=sG~h;f(o2Ln5kjvZdocQ>$Z}Qt)2(Uuu^|lMqjC@xT+&h#c<ELp=ZOJ
zx-Y>9oE6y|qQ$cH1u7ouF3uo$E~0A-Nj@WRz3O2mrDf&hwf~Z&I1)NUulw6RufGW*
zutpy64Y@>cd2q(=w6%AeJlv$R!RaF>>ZzPaw>DkEh~93#c;$nq=4L53zn=2RWE`Ff
zF9IX%wC11MIU}GOE0Q}J);MIFur$_1fW}}k%tnAZMcUfng9gQx{L3x+cJaps*T~Pq
z&Jw5YhQ8Ngq5g$1TTYbBH*F);{vhmQ<0I47rUw+CD|0Vr92!2Tn;SF4*}!5Lfb)D?
zaKmT%qzBZ8;i|Hgra~4wPo4K7?4o$mTlu`pfzQdI?Z@9l)xvaiTdE$-=z0sox7{(I
zc8Wz|o}6}>J#q&0L6Hc-0;lwlWvq6K#-PdZ>q91Ay#A$+#CK7QKpQyn+NEtOcH_q?
z>EbmEr4{5_XDA+YsgUGC4f<Q~Xym$bCYEClnXo%_c<6{VS;g%-i(ui{8dF4^wsP=G
zPlCKJx*(|93cVgH?c;8#lNlaH7BOnKuXv_em|5#CJ_sPje=ML(@wmlUo{}75Urt?f
zKUo@8YDgyhO9!GVU(`6|);t|IZOdBe&$<(wxbv)thVyDy0VGqBOV;rZKzR7n^FOI#
z#{b(wVq^Y4q8RTt57f%Z*nw8WO5e#?$k@=<$e4!*($UGmSl=4bZDU$Z+IEu;y8HB(
zaTr8Mtboghpf|qpr`vk6dd*zZQyuQo*w;6yiY**rDFVrmldOUIV*NH9F*opJIB%V1
zrE@qGY?Oo#8FJ++2j=|G1#Q00lK~)to2bdFMa~U;^=iw?NZ5=ylm~Wm>hgm861c>M
zSu?bh$B9_xwZ~_kwPNg{Mfxa|Ogh#LMS&CJ)hNTk0uN-fd$f^D!Y$#mPQyowr<Pu^
z(8i*?Z?EQ;4z##C5xmsN=OqDL9iiQP2K;jct1a*J<AGyrov29IyO?k5DPCZxnP7<5
zN61MiERRK<<op+Te(ksQ+!urHGpJ!7Lfs!uZ-8@dJtkBIDU_E#_%r|x9oPl;<6R)Q
zyQ0KFM1pGxF7Zs)bDr^uO^Z&A2gsUu<ZjZO3&@LA{3|)Q9Xc27XMRD;3c40D`AUUe
zJr&Z5;!H|6%+U(0oVtcLBa)EjwfkbZ9ZDPq-U&S2If~)2XXL>B3)LVf(ieGqfMw*p
zXi6L#M-LX&-^9_6Un1=!kC^k32m!5(=flusqz>mTqL~z{CIXhpmJHPh?N$L{HYDv?
zKvLf6c)U`zA~%V-_CcJ&VN;Cut*B$`NH?SH2bT`v%CSnLg3;`(I9SgaQ<g%tTFV7f
zMX5LU$XmSgmJGG@-RK68swn5bEnM&Lfe)tfRA?BvK+U8)oB5aDptjJi84&kd*kR}6
zG*<2K8<B?BL+MlS;19}r%0*(EZNj(jshX!$0Ud<lo5G27UW)T^wD_h?*qIp)A$_QR
z7#IR~u^QGept%%~JX>+K{^#V%`6>$W3a&=xaN)$8boJ6_URzodyW`21yEU0syvxwB
zi$U4$d!7x-znCJ29gcf#?4|3(KN{<PggwQ*sK10c*RkpE<~We=qBfIr9q9QtqmSf)
z_*43|f=>bPmw!xai|M_!WTe3J#tw$UiuaaT*g_mh%}&9cG_al0mFHO^O^^^nq3KN#
zEmuLN%@#-8@9!dRpWrx$QeDkFBSfT?HyN@K(uFFC?9pb#09>W%O%ELN!Hw(-1`x5M
zlc$>h`(#)rdW|b4qdBhB?yWV*GLs_J)^#b-Ji7cC4^LY$nKs0XkpM$wZf8Ia?z1<r
zKwjiO-Opcw*Z<ezWnldc@%``b|L1;IYe>doxBu@=#Wy;QZrsQjylv}P|E;TM>UBZA
zK3Kq6f>^<aM?vu-q4Fg-PD?#4ZDmB=g0Wbo@#{;H)>k;h+EQ6K-tM+Ga>6XQe%st0
zE65$OH{Ef(0@ut`4RnZFomO7lq9q*(&j4P^V#YhPJ=dU0Pag)B0dQxppXbHgxH>cr
z4J?(QY^VXyNeE6DUr6wr{aK{`fsy6b7d2nB7sMYg)0CiP4VvUSa&YM^<T<%M;q_C}
zwy~)%Eeum@%*roL6Mnf#MhG=E2>@E@ts=S^Q8?AyqMaX*6z$&kH=H3R7ohMgNDp!o
zyMZ^SWg-ryL@rC`HVagZ(-G1$0p9~f69DFHU(o}7>(AuXFObo9f$XEPE+kyhhYlsP
za)yf}(KnzT{AiuIk_r!a2>?Pu1)Cq|8ja*jO>Fj>G2%|PBcx1iu>``8dyHpJX3Tnv
zajO1ReAo|D-1%!*_Eh?i%_OpfRAzEO353j@DFKNT<t_Jc)`5x^`xTcVQ#0ZOKL6sX
z7P?8MnY$C{E#;FP7Z-tarL^LzLqv)59OZc+tIC}N6(!gfgNF^xn?ntD{o;+Pnf?;V
z-1C#<g7qCMog%#gEK@9mM=O*dsbxppf1jYGZ_i0&d>i%=Bjo+mrG8dCA@&oZ2L`N^
z3Q8ieiUr`9I_1EM7E$hB(Isce#8Ih$zYfaKAzA`u%@M;+L>eDWb?}h)Cz-i~8l!mK
z@Zi`nZJpezFqDWHCyfH*BZvrogwa+^{DAsp7}yLUS64<Ce~z15$me4n5K~oTXmKOr
z&{Bkn@@$Py#nPiP2n;yVHD5l4@3qH>1%e+Y8*P<Snc|PnHVNw4Tl~27CMS78x=^RS
z=V|bw{b-og?e+$9l$?H6(Jt3HJdSU>4TPFSEsfeifC2(bes)rae%FyuTIv~`GLLg|
zy0QjQ<!Cx@ue-v|s$|~erS}Yb$NFBD&1u%}xQa|o2LOp+_~nxo5lJ94b<&l7CZCeb
zA&gYAlU^!3-VPXmoly#>WIVr22=a)Y#)~l6fss3)rK~2Tshq4n2WO+xt@}Yu9NxFg
zgDNG$u=7Done0BXcwo?mVFf`t70EPO%|Ei&w6sYo4Z?-3TfF@lB)!7)xj`k&M926*
zAY|F|X1iJ2U2ZJbz4A5Ck!{&HLTGVLXKARYYxq3;6EZExx>Yo}C3LV=*w68Z(U1v<
z548tHNkb+ex#w!@M!V<;rE>Sr=Iu2ig?1e?!0u`b<5AF1pM?=bcLCiHRGi>vCW^?w
z!|{D&>z)_TZ2(iMtu*^B@}SI15E1rat(dKfxdIs94r=ejLaJT{(qVS{^hc5YPQpRQ
zXo32n_65??#o_T9Gi<O+{E?z**J;?-YRTbZl6RHx>efW30f4rm>!Fr1mJC9DS<K5C
z)#~-8NO?hFlh!P&?DFT0n{!bW#&zr&)02}30p<bL7qq`d>IKGW*xp(1{*>CAgmSL|
zaTctT+iAvsI?|JEL|V1%lLc5{lL&Ul(c)z4IR8@5(q<fE`G!)zhe=)v2rtCPmZa@u
zbswFak|x7;k3eX39V}lQ;LwAiDh?<Ms)31F*@j`AH2Qr6NLCsz?w=&A=K(PK+$OBU
z3pw=OCRD$)$+Y9__5I{517lCAG`2wMfn@r@)2ZLtxHn8~&>Fyq&Q##U=vn2_;i#d)
z5<*S@;6~J7<!G44+rSJm!Fo%nFLXh9C$?~+M#pm<qE7KOof!D+bxJU>(b5kcxxn~w
zgd7Sms>{`n2#R}M7RN}qt!tZWgf3LbrC5&D$KYIRC|Emyt+Y&<<BNwVeP~N!9gC0X
z!6br6O>n6!x#J*}n%dA{!8|2;sM7UbcR^d4oI^aK&4d>(iEYRk2Cw{97;>l*%ZVa?
zUFp+lhbMUs5Rb|-)>%|ya3W=6ryTxHU#o1A?t9+o?9mX0#uN~-J~juMmC2daB(nx@
z+3t7TPm&}AK5A@D94cP;ieL1=Yzs4K(C@WRcO#})gi0d{wIKr<edXMd+=l!xY-yS~
zNekd=2k?128sXnKakCPE&=JT;dQs-{CEuE~?j=*GtfE;ufuI9kB;FcaAREmTm55jO
zWMLa#kJ5nWZq>X;#35a_{zG!Wq$`7Bto5WTmEQ-WE1iPLXL=lzxZf_#Mo=0{F5Kkl
z<@s_1a=2|3KuTgJ@3`jGbf-%4_tn%Pg;J*b8|wD~$NPPFnab<)Q;A4-z3X1KMU^I0
z+}6FKdFx215PXfb12me3J%)#<2AG0ol^XS6ZQyz5?peq2$Cb|B_k)^OdO)O-ITN~s
zg}$$E#E+W*gJJk3>CO8PPP}5K;B0AV%rzZyX1bLw*W`HfYxLHfZ0ck4<<8ji%}>R7
z=QFOGL|{7wE>)n(#aNlR3R1*=vPBfZ=S**mg51wn%VXln8m`%|xCxXgia-pc7xTP?
zkz3zmkuXmf`3yn(^c3xi?i^A!BT+!lpZlf(ouji%;8F$ZbvIYnen&O!IY*l^n-t_2
zLGPM+b+mXFqUmhNX}Pp-pDDwII@mh!zQ`e;d(3rJ=+Bz>S{j>jZE4L~XlFD)KjK9I
zXB2&Chz&4G;T;DHp_bW794ur7^cciaHNKpQ`WQ60dDgIt9;YD}p^$yG=HZrqcWFcv
z2TL-bElJND)?mMSegjkt_C<T-HqNQfo)RK);<r$HJ_JW}BIvbXX$WxswN`@*1cx!K
z$#w@^pj#N91G#L)cY9Az%-n)aUOUj9p+ymM>if}V(^A?b%_dgK*OMYC^B+fVeg`i`
zK9^~JA8Oh-9m`V2FdfMjk#c}jZtY<nW;4skAB{+aP!Rk=JrVfXPY~=5&DnOfhuaD$
z^`0Gsw7YI=%}M`0EY$eTxc+2Ze=O7kpD8;-t={7t?ay~eB_qGoSY6Tz*z%R;T)K<N
zg<7(QINke)CUuVo*&UkLs(c?D7u-}&gfR6nO^^c|B}=>?YphUwH4g)`@P*OjkxmEr
zl*Eo5x7<nYm4se1&7toW+!ts^?$3{4@8^TD&ckUNmBl_YSVCr7!MlnPId^jl)@WGg
zI$Lenu`&a5XY}YCtUp5CTe$CK3eqY%TPY@M+--#S<>?dzgkTyLHsR<N%*ph?i^of_
z>fX%;NBX1R7ZOhyvo7N?+l`gMVww>@#?R7NvWQOWyY)CatuyiN8+J@-9#1+b;u5Ez
zw6*bPT4he9C$eH@jUFvo)`JEg3Z1d?Gu_^GYcpQ7l&LL6!OcgBq*X|+f&Gp^Y|nIQ
z-V}*}HO38z{h5L{9HK8X*5Tci=g_F9xa(am1Y+jXu-1q!J{BVZy^{>skSIN^JSqY-
zE;zkEQx(y+-1Mww^PA__aa6_Zb<OUTO&0_}f%<CcjH{LjdKVMZxvKjIqgq%l@Ql-g
zilzx&xn!+1Rj^Rncjtfyb&NMGow;tI-}0pt;MG}a$+lxsecUsLYj?z(al_?_!|2NT
zE-*m6k=t-9p{2;grG`%{;T*I@ZQsPZVy0{{yXg!!n)e{}?9$eW^`p%iFIxYwSiMqE
zQ{w;u@I(l4B$gjIBRFL0mx@9jA01rHqgCZ6$Y_cDz!HIlT34z8mXXsm-!WFc;M9OG
z<^PlXW&XdRVT_FIO#dVKyZoD$gw=k}qdU(_r`0nK3K}TZV2sB~>jmYYhF<Te2DUzp
z-<wOIE{wWvRnF{!8V;at!J=OE8j0QRrG50UfBtoTURqf%2|r(+mny4^3PMZtRZ9fz
z(F$A~#A@;WWc@xFnY1x@3XMxyOn(P)$2rjI>CHkh0PP$Q>){QnvhP*6U%2;5%)Qvv
zwUONVk%XW241zEDR+P>!o~Sp#)jUAj^<Nq}m#9E?mseXmYe6p?e4dI&4IqsGrnK@l
zAN<;2DH0i#XP&L`3)Z%!kk@zZG^KH%c2Fdp9yaaR4=2+H2m>|kkM1!>+5sT_5GeMu
zaSC@ahArWmFyHPs0I(~TxVN`5Ku<Q!nW=#{0iyWC!Yt!frOZb>_)O41qmMZ>9tn{c
zr&L*)8DSETRuq~=bJ(PMhmCT!brU$BBroS>J$LE%WCfVok{UrUln>PV@)%mt2+@R2
zZuOJ-5GqO?el!t^m=^(YeK^qV;}^zX4fFiszpj(;1U{CaFfR?OO|l0*0orVl4fSEU
z1YZpT8xg4jW#aaK$r<@47j!Nw7$u$~B6?GrtzTtcDEGrN2+dg@NbR`st_DIp^>aOi
zfcV2v6lLWofEju@Mzb-6%R;fF7gWp|RUpKU#LT?#b~cdCc~kisqq+*su?aXQl+=T+
zBYm4UpixszduutfDzoA?pQnq*pk58V`d?jIBCgmZM;B$ZERl+G7Cux|*Q9oQWOtww
zJ3hxyLM<CqT+&o!T9&z6b#obweb8R|?2Auwg1%y!20{jXN;9XwVf7BdF2`O*tQt>i
zIxTO9V0e=)G<4n1-nCjA{Ep!pPSqSP2bqg-Q$d@%snb^5u_{`9@)XL%HO*6-63m>f
z!Dg>(Vnmi285{8y7qzRzT^G>Lx%*DjZS5*iWt4aIAR?RlR`_L&{0b4+?M={;R!MW@
z1AK#n4nc@fp==?k76@I-=UfT-v=!awKJR?E?XgzfN=ieBdGfm`*(jRY3*xyr(Y*jR
z*ak7V(QZbHmrtNQY63m<--Cq6oZ(=p-^l|MMr}QF%9jX@0C$$c?3M;`o(kjHs^fXh
zs?SOt9NHk~%+5_Qx*Z|EREVx5dU71B?jrmCN=<iap(KJC9-4imFsxFw+33J#6CE6h
z5KDrlpUpenpb*H^4w&`YXQACZ+!phWoh_0mWod}${MCgKJsKlmjU>rSpc{Zf!l!2X
zt>1Qik~@f^7g?$7msG;bzQpoNhVZBCB?KC88MGIH>8k`roP#9#uRQOfRjMRx?}&49
z;}}(y<R!nLuvTZ4*YAynU(NgAS0a0am>1EI{BM@6&Trb9ORZG|Kv0x=HPRbN3UePe
zD)0hgIJ<Vt7?PG)R`T@1e8R&^*CpX6Gt<vaH7giL=mIa%PwjJrU0w9eEy!k?s>NkW
zudV3o_Qaa+Q+jsm95-4)KDD>DV&E*&R*!n4pn1#&D%)T(+U4Vc7S3cdkS)UX$bMtM
z142QLo~3qNw1XL|g$}p<#NQRo1#KoNWtDItGR$aM7rxtYquekf&|h)lZyA>-t>+Ef
zr%lID>tnM{I3!)u9vkjC9%o;nf-~-j*}MNWp`ZxaSx!H^Z!X$GrM-s|Z*sIrHj1=<
zd5#YXV7!6l(7QqP1aJ|v&6ei=ygyIo^870RrutOPpg=2w#RrL+Y<)PRY6}C$c~<p$
z82~FgKqFSU13_&FJ@VHFq^<(WO(m@ay1F62`eu$6=!%8EUpFWW3dff&f?0##kHcG$
z-?&xf5mRfXW{^gSeoD&Q+js)MG*|YJ!u^p_)L-6BEv21Y6g{eV+zDM$3*Tr%cKy{=
zBg%HU6zIe3l7b^pbgXD$l$L+2*D#KsR`{v;lJWplL=lPtbe&rz7^CSr%)#{#fH4bh
z?Y%<bfU4;hFr@~n>t-Neh(XHA1a44aTX%DN_d9CqE`UC1>~a*3EwXiUticC7v-OTu
z?qpOEN*Aqi{EvljyLHxs4G3Fn0RC9L+UuVocjRE{<H9it0(6Vu>W!c`U?5SgdD$gH
zaGI0N|7kJQ<lT}BesbHjqxF|Z#g1q>l%akbXw3;#lk^{JiBqMRq_JKvl&>UH!e%|0
z*ANvPmValS%%s=s$4C{Q-+t~jJ*KKI0r!&?j2ih{$5#AS4@6$qMYbfb0s(K0zNuzp
z4)5TfmiyG%V98TKw|(*4%*;-3FKv5&>j0uat27dLz18h8lCAcFj80>*IFqq%-WZmi
z&5}v!@)28{8&dz24zdUYCat2rZ(K=2vy>K%G~Vd)9$gYO<(@|X+dN{c_s(Zc)d<R9
zDb!NpQf~blnqiCN=#!dkxhm9KS_)KLR?2qaPH4>7+FbeCU@~OPWw&ND?c8eqLPJ<N
z;$lA+Cc3sqKOI91EVA0({vE}Ao$uZ#59h95B2^z!m|P;kf7^7bh_WY$R}IC~YOd+6
z&`oVmEAL|Ak*|AJSH!c9@Q*&T)C)r&W9WHd1`+|WHV<-0u`w=7IQKI1W^d36mWM@)
z{8`|h3T<ZAVa$+)j`@}UZw?Ua#G(|`RK^Fi%~Aj+dE57^bv;&ex5~ey6~{BWQ%8;6
zf|`f!oU2vn5kK9gR>=@0A|WZ;Xyv$LY)aHVgiK|AYyaVZW`aMonkumE-OP76E6e^h
zDrJS*)pMw=h=2AQ5eD{r%ap{nwCaH$`9HucfvtlRHEIM8c>jL$RwG0q#KnEZVfMd~
z<>7Iv5zbjj#LCu=bkp&cvScp;IRgpmr^GTD;8tNxxh}(iUZ9{&^>_H2+_uyX+6V-m
zCTmZa+Jl<B9*rQ^KxK+ZvMz+?e3<C4#|awYo32U2ly2EO^*82CV$!z??WT2QMr3@J
zn?25hF~}!&v@d+cC4#!dsm0{KeC`eHO-)!viYKOpdh>CG)N(&)L=&&{3<fwohj*S|
z4&DJ@{a7PSJB0~UWNnwC4R>0iqaqAE-ui0xaCu|ruXqYGq{mFY5K8WEMt?=I&Zt0K
zrS6nQ3Pye7O@Rw6R3dGxNGf@P(KM$_D?g!T-21g*=EE3k>d$*O-5y}~D~@`UtT2af
zia+Cb<$t`6M8hlQ;_nteGS=^uUshUKZ*_Y5l@>X;AX=$2sE{jsqp)bzmYy~$)ugh!
zLYHPgy;yokc^dK`W}ri7c58a<!}w`+J#p=q+uvuR94HNVLq=WJLuhl?TjsbdR+!a&
zheWm<K9zU!@1+Z}-QozD)j^bIq->1v8oGB5t8sUfJ^aYnhcvz%fOk=gn$I+|CB?E{
zb+D_k976a8`VwR@?I4$Tu#)S#n6D2J$rMKqB53hZt0$(2bpd3OTUqBUykOj##$C2l
z;|X2!%HPJ7O{P5%#;;rdkp#KMX+3R_eYfsRoZH=n;3UrTP2Uz-V;{-BZ$#0fzO!~n
zDhZtm=@fib85pjHQ=Qp=uj%TdK5@^)su3Jh9M=?}`3W83DhJV9Y}R_!6dpYoqe$B3
z+i{9_&14qsn5w-nW_mQhMlvQ#s1K)p(mhdO9k9g!rWF7u1h~+iv>w5NLlEFj+8Tsp
z6z8cdF%xZYNYHq;3evrmCWjwVZl=()axOqb-aNetxfhg+JdYMId~WCr6HYBm)8B`7
z>eq>}ff7}5neo&&eOt9zk-@Foa~zGHrR5}emWN!vq1hS%RW^>9Ed0pBi`P^%-!jZQ
zwrkz4H44BRhpxwJkDoesVqPqthW#q~RzoahM!u`N{zI-g@@_9lcOFo`qX3)R-qiGL
z0K(f$NNrOOLTZ9wD;qNY6QZrY?-t|=TD3YBokh936O$`GqhkFQ)_)9dBbVG!R~!dI
z=L6hoit?$$Fo~aj_R2=q8?oYV{u3^vGzVtIHTnI~NMzab?76vN*%@#%b=I9_$-Syo
zm{VyQHTbtgGWhIu*2Ik$+qBVXL{<hI=Vjz@can=L=qZ$*-?lieUFs{2wdfxGJN1Wa
zAiEt}0c>X#37iGzz1`|~HkubUeW=npXd{|iQ@s2dW69NV2ULGa=-=%&NG4MeOP!)(
z|1#i5T9g`rGwBHrB2tn}_?h2aEOZF+K*--6cxafTdHSfhqgV(IWu^^?Mc6nTXpT5?
zsJm03IIWZFqi5t3H(tPVy<j+tz^8k$A_Mr9HJG5wJ5kpsQ+~^LZ!JSz%i5EH8mb*w
zEjUrCqYjBzX)oLQ^iA)ma#lu{E)3Mka`-CJto-bdgAB&b@&&z4l?TnuSL@HLber>@
zVM*Kdb>RXiVNs334~O^Z9E5eM=sJ9>^(!wZ5W&%eF|PU=;Wq?*KN-Xl{Jp$EoUCV8
z3?qH{AIt1D?H+RR@xo7~HlDwxA;c8>TbhgpC<SW#Z9s8bNdJ6Pfsd<V?SXx^kmQ${
z{-7%y6vmv20O6MoE3;?lu5|0-w4>ske4_!nyvW{)L*K>6j@s#^GgS=Qh3#2iWX#mQ
zO-&VN2gJlBpy70RP-aebZLw*h_wM?7e86S!-mw3tdW_{i*%7}*+W*9kxYq29xoSh0
zKBasDN|kK*o#I3TdU}$=TAO+?ii}>_k@f<zJ=k@<|BmY@R;knnLo!<uloxK0QLe{9
z&izSydrD1holVW9$s-x0f+Q@R-6iFRWBW$Q!eo{a0WaZ~%H=K5?Qx`G*0yybp8+^y
zc-ma}tJ~{ew<MT#hFH5Bg;>rj)&cy@^ZC!Q_l{+x^Z}X7?=y$YvBOK2kj*V)NtUhS
z>asSMxA%_8dnW6J_uKmG_VDp|m-f&gj<ZBwy7$}d9u65CR7GaM4pc+R_s8+$`?!{H
zodLbn75#(v>2j90B>&hpJU0f3w|yn-FK@G6Os2TyXdiFHm(ma?(GD47m6dm$fI7L!
zcF{aT9){__kehIoP@+#S43L>%60lSO^j!1YA71UMwsvIWG(bh6Ksm+JCTu_`V^#zA
zO+ciPXxm37UK6|-f-i{23yqCnO<DeFGWYxipRYT{hBe3i>h#@T_VkGk-@Ot)-=U_l
zg|HNpgzrD0Y>HvNW`(PFER$+LlWPtK$-Uc#GC#guHqBI3=2oT&X&Y@p1R^>nwip68
z?Zmy&YH@iuFiA@SVO}fg3HttHD`|ql>tg}$A{2oxiU1uKRAW@^hW4SH#G#tX3+2K%
zB{U|*5gc<QA`C8uGYta4eHC)aDJMGl8x0$bK&}Ki=*Z`Nj>|~aFaAi-Xv&YI>m)Mj
z)TfZ%JTs$(y@F(x^8?NEcnt*6jrpJ9V@k!7cq<|FS(c3~C!9&lgRh59hHk9tKDOSD
z2EN^*R(lib2P&-$!wliPiomHIs2^qWNqBobre*?Yn(F4%_ewre_3LmuI0)@81yo@B
zcKz}$8kuq96zn_^1;~;oE1UTBOjJ^SdI&~TW4i~B0S&t9(Gq6r8rCR|I@oL@4SxKZ
z%<gkYM}g0PpfBAz?nx942ZU;mfcJuM2Eu##@V{YhD<FY(HtT5U9<)%j!+dGL(-e!b
z@vf0kH5i^ZajvEq^^PL^ttYhX>}e~`GcpQ|J-^r(%p-ZKia#7CO*|4H1Tp$ukRvkE
z=%y7^jayoo=PE;=>!LGLsB@49F7g~QzWz5C7&l-N(^b?rqI|@kGc47@SGdA<5t_?E
zN3@)3**^+$q-Qi4p9V{1Omx8pQJCYtC3|dO(dhzzXG8X>zIE`#`QsSz;lGbndA<X2
zex$Pd*Ux$LFJl>?bAOn>w?&I7gT(WJ^cBo1Z4piqjbix1jVK|2(X0lxiCAW$0Cg#+
zts8>M2IFnQDr(c*@W&K`#W5f$Ky;cZ-E&zsrv;X_^6DM`Rc*5<A@NgevOz8TI*IQH
zGt_67u8(BpFE{B_%N@VNPZ}@?4kIevxEg5LZNv0o))gJ;;Uczsid!u)5LEVG&o{_t
zvl>kxzsHH7aRQ$xo1q;b<~}g6i`1!aNl<dnegYcUbW=0_x<=tm*s{g{tWf_gEC`X*
z6MxF{e>pZeUlb-TJuWnDRN)0XteNW;i_NEUD7*GcR=UY1xSM9cHW?i$#ek|NY7?yN
zPAzOVq1boI1#H==ZYlH5S4rrHr!t+#EsyyQh%7ZXVId6P@&&>s5$nqM34<A)K{uKU
zT^=x^;)hq<=Bi9I!4Sg4AE!xUoi-bBe1CUqbQ<`tI7TqiwdsM@P2=ytcGT1@(HmTf
zQL{ZudR5COp5BdqXF2m6)E~MTReGi#;zS$`%2}}~c<oBZuS^9oGDdjB2PRbR9YS}b
z^k%Xq9EB=50IN(r?b8$>lFOt6Q0ckTFT}et2c81ZgPdB&-Y9Hm8WbDb3lB26Qv)r-
z98@IwpGrgSdGQL7>oP>q(!}Z~Igk-5H%B3;l~kqQ?*T+Xzq1xliUP}cNF~%?uqe20
zXW;>p>gTn^z!CK}l$aa=T3CP?q_+-6?WDR_5;00dbr;r4iX`*q1kXjpgT@qYr5DAg
z+y$*`EhDHSj-D9Q#}Taq4lPFjr9!_&LBL%B_c04HF-Cm40^UIqm<%`?la4Ky$}i14
zN{Op*S)c1XfVCaa$6aZw7C$D8y2;kb2tYWf>?eZ??wG6A!5;0f)5OB90C++2b0e~E
zOB)<c5vju`Je2}p;xSKax1fv~*1%Wcir)ANI%v}M;^v|3K@c83_4i@fDrhlLw}^}}
zH<?H16Cn34q{AUhDai!iDJU(KJ9(CqC%U^!mpqadhwTy@=S}N&VozBe{n*e{+vR$>
zK#87?CDo$Fu?o`kq^<Ig9g*BfS}8OUV4T!2T?X>Y@b@r9-%>i_w}@Noa?oU!1Zi{y
zwje{b5Gd7)g-1o*wija+0ey*F`q~Ovv%1<Ap8SQeymE{xc0wLnjTld@-_oag1!@B9
zs-J>*CEBYtY9Qz1=l+~a$y3a^&WzhH+R_;Az_9=a^V*5zIj_lSuZLAb+IPnc&sYba
zvH`96QwYQ9g*4fQ-od)u;1W1)%F0F9G!?2&p<7d|_*saz{&2l_y5n41hZK&h#TA~L
zvG{V7BEkBdvcn*JB!FAQYJ~Fa?@ZB-0KW_FGH8okQrQ99^uV$<*eWAQGhNA!q|SG%
z;*fk}PI%RMuimdMEc&gs6;FU_8rEhgSJ~1&1X`|^kBVVLX!x-u!G4l#wZ)*vcv2zr
zN9pjkFO9fDc|g<98rYImQME7NOp6zcst!8J%#wtCR#|B4$!F6KXQ~!dSiXCphGnw|
z#=~Q}b1g~GIBQV^!VU{zb9x<R$jE6dH!zJ!3S#|SHAQ#x4QLG;$#x=A;suO;Fn@iH
zRb^<AeS>{0`tKH_k>?m0l%nICDYh-4UnfR4U|v|yyW@*+e)@m~9_@Ae)lo`zWrguS
zTG5ec+<0Rh;}FAK7__bwE9DQ|M((boPZ5<{lX1vO_otBD!rviQ3yJ{(?OS?@7u}+_
zCsRT%8<Xyg-`(+)4c*@yDLL2L2bNc_?uE}PSp=nhWWl!xIoxo$|NfWJGtd2fgn|hQ
zzCC}Wb=ApTI)Ry?P8m9@JK-eIVwQGpq$!u)8?jM%(Z!^v7d61HawTd|y~jSz7P4<=
zm!lj;c$es(UKrIr@74O%!Aw8cKdgxs2Yk+72|v2ga#7akJB~*WU=#9(?$UEbRA)7e
zoWY^#$7@;W(XbkNaM>FuJ{pgFQf=sNPZ~RH1zD@!|2?6UiMcki(Nux@IBzK~PQz?p
ziI}+>)&*2Yz<$-QNlX%{+Z<J`_{-AQ)wIwAtrH|0oUn%v7+KZLHLg|Av7<RD6>%w8
z_AOT~IKZ`4V5~-AWJd4Zfa-&2h0T0BLbFKk;Qo7KSc5~q+Wu4hht{y{9(zx`GdkIt
zqIqdYiW`$FBvKGT$$B|@TT)#={$B!ej?)Vt_+VAl?q0O15-Q_#+GMuLl+f7x%$mDN
z+ec->Y%i{K9Uq#`2FZ5$h;@1K7;lH52~ff`Hsi2k7b=;}pcg9(W!fXU1h&@O3s@hx
zV@>+9g(Zox82ce}j5cMOvvOVn4S3T|N<ti?29$eakDMxYt*{1G{MPjSrdS6$6_>F4
zH->qP?&k*3=Z1}&O`(lp*&XsywSmm}i_`u@{Ek;vuxk(T3u;~GP7A&5D}^oNtZ(2a
z_6mmHjF0e5EZLNdoeg1)^qbR@zhmyJ!kWuMa#=x~s245k&OfO4s8!kjq=;D={=<L{
z!~e{o_<t1f2N=!tE07t}+#;j5|8!T5Rek)m^8uC{(!ZWaqj)L_D{<AufJ6wUjFd70
z_0kHrh~I>_)$rV&-Y%c6uj87haeCXJ#L`$^89*JoH%d0Hrjw%}J(cjq(bz@XUy^w6
zB?`@SyMi72+wS1DRuB1peo)&0G0koks+x{?8}OHJWUplVQ}ej(F|)YbepCXxR(%nY
zmT<b9Opu$l&wsJ+3?ECan`Icy<gF>ajBQeFHN@VEZ|@{eBzp%zs_Ew_ysL@xOb#b#
zfw&jv6t7DWnVM2|_sbjVx0)#p<zvbm4v=IgmByQU)<livHB(Y7scB>reKfsUlfanK
z6^#xY(>={d5FdX>0PnkX&K#-!+A~Fr9k=6nmwHQ8M`1oZq7r%T$!pHTE-YG}yVmv=
zRO7=Ud}5raFx(jTbB+7VtysBET1OXa)dyQto@+1Ikq+y#3l`-G>yrmN?`Nt>bQxor
z^ouaYwB9D_-r~~!OLtWwCkbHt{vMU|%1_pV{Gg_Sn#lS2`rTkj?UgB3V0{JL%F?uy
zIz0&pytRzONo!el#)P&@m9!XsC3GG-#pme^b6j*qgc^RgL`{4OM8%%G+Z5wnOKdhX
zZa0XiT!_93Umsws8Qy@$b&M~!mTTS9JO7D?82-dU-Ga#TKPwuR{}W)5{%kW3Cq~+~
z6Yl5NiTA3nTFg2H%6(nYl;M6Op&J=SfXR_?@p$)-&wGQOSSjg!^Y-fr&g~rvS*1Vs
zCo|%w?D=l7SJpRjLJQ%FKU==eL?DB>|BFcd4q%uC0kkXR^6rs<rG=`at9jA@;GbTJ
z0~(A7ojEdit5oI%#U8P2M?x#F?AAX~A`d>n(#FS#Pa8iKztjYAyQEXVTWGEz5y?R&
zHc~)Y_t0bWs#vMX5qYe1)vnUkaNoNJ-7kibG$wF7xb&F7_M)a}sG-&k7lsKlY{=O4
zvhQsTjd`CZ`0Ly%Z|k%toFGfk8r3FTL%X9Wct`oq@Mg@VpuVxFC`4z_j{~|GaZxs$
zhlR1E2{J5piune@L;XHlX&fxa=?T^#OjI#Y$R1V-R&_8nzq$K=(Ph^wH;@)tfdPZR
zBnP)(`x{C%9}=}2P%{j-Yaqo7^eyMODM(%QWP#YQ9TU2T_o)IuR39XU;>z^r0?+L@
z^2Gz@ui~S#>W;+XnPO_`{kd<t;X;UDbN@`&g)J5>mrGj^)KiS3Td@5Dkj6Xx{dXs?
z{x1UZfr*4R*6yW!NGQ{yvpv;D^iIj#PR^bJxl26XI56@hdzPu64}k28A|jb_a;urw
zh+<i$(s-e=1dIK;{fO>Zqh$U%QCsa4?B>fA(4GkQ)x=PTS}gh9=o!chlEtUf0vlw<
ze*e3^EL0|K86l}=hzkD@uNA39mxTsmM!+8g1IFu%89=6d&R_jkHvZhfQqZhfNojUI
zRUOMO>3K4ygf`QkrfP6lQ)-R$wDd(l79j%J`JU}_6F5e=i&>O|DDSat1)xu8(8=5x
zkQU*!g&maUFg0nzA%XV!90bRe*0$_GXKlR^l`$QAf|tnGRw^MtQ$ZhW0>Ajd0@$WG
zPx)s;YC0AJwb*!PMJrt))T}Q&Wjmu1<wq6KaDXsk`m*4Tp##4WX*jc93Q_l9zk7?d
z!lP9}MTwdLi)TbNNZ&sba~oL4mR&q%Qs8R^LpM#BR)L3xw)ue^(wcc`yIX4>XEy5@
zUGO*q#JU%8NtUSW6F>u0CwQzn7rIkFgcH76?T(rEnw(GCJR`9zA;|^EH8-K9esEta
zL5Bt=h>qsE_`Gyz>w@s1rDV$XTXXz5j=x2<`q_zkFtyYiWiUZG5+&)ipgy2}QtAW#
zR7HoLW-!&!ccA^hANlE`ul7F5@u(;jVhr2Hm<l(+)R=A1qK<ai)aGD}*P3A)HK|i2
z4~puk@5E#Pfs=R7cWw4Lhn3X}7&YisY?5c2Ko?eE%DIfGo!EPGO7yRP$BVi5=l}J5
z)-qo5g1QZ+(<RI_FBfBA+lihJ8AVIkW<_o7CZnK~Pp!ntBQWU%Ww6Y?&_S0HT-uPc
zPAd^dGp@>K<wuD;*l0{5sVSxUxFPPL5<j?7*hd`cWgKxtoQWT>Ot~612DDGHk0h=N
z12JOKrRIn?voGdE`P(qn2X)!C<t<brLYK0t8o`_>v&gt<`vQ({W9srN95wBu-=f5b
zoA!vb%tl#t0%!o_oJt*_@3k9CsC9r)n|ggqsb<dCYT~>h_E)8D_JFqb=1zVj;v+l=
zE|CW6_nGy4ogn!%V*TnD?y%PbMonN}jiDCp_yy0CRL5M=PJ+@+T2^e(xDmS9%Fy2^
zlD7O5E4)#dR!J=RXR}Y|<V4c?Aa2>RT#{Sr?Ying;#KjIzK6}UG?Q^n)+s}C2FYBB
z4*{s$UgoZsqm|~os_}&C3FB-g92>Pc2mF46HhZ0sX7W-KHIA*A+UK*f0G6PUlyEjG
z&aH{iQMd)y4s71BW6S95tp<nMF^i%?=O|Z_swU}Yw<eZi)KGO_Z&F);Y5rX>FuVnO
zgj5sp%azUwlvtI|4HdtO775v2>*~Qk>h_xk2PdpUDts(F%`{Y}sw^&~2(MZO`?HKu
zkjeCrD+f7IR}Z~=r$5}7zu2%s_#!Z=5Or&%!nMHVZ`F6Ph<o|ldo_ioKVL0*CzFXC
z2#bieXo+Np7XSqHW%S3uLzfs7#+S#2P7)ijne1qjR2K4^f0zWPn#wX&mAf$vrn`Bd
zH>iY)Ef}TYVB)zfB#}nR>XO-F$oDvWM-Z%Eo>p&TsjJl-iAGw4sbj##>^>B#@oJSk
zMGcKM@iNW8Ie@t>U*^byjr&MQt`A;IpS2|V1>h~YQ?Xp3VwmGRs!eP<#DOFnNv^~|
z^kBcfoD}8kZ)2Qh6~d}2wT6*tj5*FpQSz(x2f{`p=8UW+4v{cErSq%^c<Q~r-!hN*
z+%^&UtMw0Zi7%Fadjv8;6Ovn2MMYGk!(-%=ni0;3X)Q6MJ|8SNJ}p-CU*s6oq*KUf
z8hW>g%Ps-*lFUEIe11_>Za6u43?@pPN{-g0qy$iGr$CM3bpk1+-%oiO9tB&%cXlNC
z(5|0VPqi&Wb|-GxCMqBeF<L9>(xNj!TW8_p9ws%MaOo)?-09+s+|;Ync3kYdaWK@Q
z0B-qSQ%X_Ks5dNphKf|^J@3ugFUgo8a{8s*Nv|Ldv5w8=6bC$9f>hd5vja*L2S4R6
z9xy=FS<S6yUlM03>jIlIlr9h`+`SRvp;esyUvF0xE0aM)C&m2JpWEh_`eTP7?4%#H
zkK+bfBe>@pP@`E-a8B7?R^tuHMl&IyDb=j%E;Pc_nyXGVN!>5vrWf;n7i+FsoTA*t
zvdT0q1$$G8i^>f;@1^?``Ka$LtZO+m?tSw0v72Y_e%e%tV8XmBye$RQ9$ykj5|GiT
zmy-OvJ(=!?v`MB1BEN3DFv2G{Tp+xceJiVPbn0Vg+)@|s$V}}-thZvDUW~*VMnZwk
zxgPeh-cxo-Mha*K&K?J;xYbbIv{B+&Ipq0B2|a>N<O}5<EC$itN05-5oe#ek;Co#v
zNtS9SzrPQUed|GmOaFu7Wn%t+EmAAF+Zof!8dxYgS<}kk(bLiWj-)#{I^nUiveW(_
zA{he%6aD|FKfJ7AZHv{8@N=WvAG_+a=O8u{$At}elyKs~2M2zIe8x-%I4qCE++ZYj
zNwTWDYucu<{_r83aQsB$*Ep!1z<9*vWg<o)G&7P8Z%8lX=0Jy>068ybfU@RRVLRkT
zLy>o6VQh_YN4gWXC`&+sfXi=3;y%whZ#3^QK$5DFE(NjS_z<GC5R(^1R*pCx(p3qg
zFAu(9?hw+%K!i>#Gsl||KNRMQp_^#78THeEiH5XsQlP<D<yvInIKZZrOi)zj;Yh#J
zC=nb5x7`U%E__C2h6VLu7_}m4t;t2CTde0*!67${HkCRPLn{nJjiHFV)D?ixMzjHO
zfoSX<3Wq+5u)`l9@vYGz9(vTd9#Awb@eN5y6emF@0ZFmXE5$<eS9hME@bj#RGUPYP
zsRBs3;PwK7sOA~5NMY9<DbOQ}yGYG5<Z$n3L(;!UQ14bs>O$;gUJtV(C<hAc>~2oQ
z!j}iDNntRMk*xeEaevR*svj#RUt*rl^%gv#Bbu=1cN`gAW*HSPm=VDVm_7}98Ne}E
zj2e8s59<R~E$+`}7ch7lQ)~<o=rS&b!h(C0K~Uv2f-Xi|a1hT>|337;NTF+(yws{V
z!#5=fYbdIc3&j{2`I^Q7$j7Q%zXIfCcJyTNEDPFxJ^1ESBL!6Uhd%2GXMz1oC1iUG
zVV4M>m0<yAz&f!8*9{1bqrGOsJYLttQcLoa4VAjC3deb>(zMu}mEo(Y`}27?UM5?g
zPIT5;(?<8rEUg~f=BL*2t44F5ELxr@G~2%x9g8Qm)3ei&txG5|qo&uJe8UdR#G2W6
zVGi;UaaQB0izn^;ucXtmkF1X;#qT!ZB8l%;vn!4N78$=3uW9w6Eo(=od40Y~RXW$I
zkOA7JXi66sH7)yjRp_Mf`cchx3QVL@{Cher15EQU?3}H#i>NICl~vLYDoV_Crb-`&
z$A{0XtO8i*rZ43;W6&56I}dbwT6nXY3SF$qJnt=dEb|XhqKHzpmmeT+Dz22E{3w%;
zR6*pZj^|Jk7Gh~DY4cLiKUgH3jF&<C`2AIMEmpn3)PN;Gw^KF7v;ii>-^)<_2p{y7
zVfOvNFEN8nXERnK-`B_6yf?di-YeW0e>Q4PstPWi&koTQ87xN|UCNV|%eirAPg5fs
z?;8p|yq_oE292H$zbu6_{p=O)OU-)uB$o;*Y6Q<4pa_(^1=P0=^PvY?5Vl0|Hzr=9
zNXtSE>ewIuOyr>onEj!GG>DLdwR8qrI9Rotf94W;Y=-2{A>+(s;0DRLao7vVb$vF$
z^kBR`o*r85QommXd6f|hy|ayIA&PeR_4B32tUxFULJR^+^L;^{pl8G&>=EuYv~)UM
zeF?de8Xk-qk{|Vv%Mp0i@fhd`z}v%`GY~fQtsoC@ZH0i>v{Vr7R}5p~G1N-A01JvX
zO#}>S0PVyPAQ1Ry6<FeOzJQLr>EL&T#XApJ7U5fsv9?^AJei-8^;>BkO4U1J)vDD&
z3q`IrOqaf@v46mEYYk3E%pP?LsdxKELj_zgKH9GK4w(rrZA_-ev{C1+O+$pdI1lx>
zLOeDiXAE|a8P&Hz@yFP=9Z(v?%4K6&-YVUDZn{n)@Ctfy|5f?!dRuHp9=}RI+AsKY
zc-hXLoKE+m9>PVApWdGaw~RCy=3Mud=9Ly^Qnd%Cv*URt@5sQ{O8dZZZce)Nq(zlX
zny}<blQ!O)&G)TEp1)x`a%)I!?6O!X#i8tw0PK`47en_)O0VhiY9%_f1m9jZ$|jS$
zUDo|uD<A>Hp#Qgvn1W?~Ba-8#I@5WWU@*#>nsig~s;*$%UF7ud=PZe&frB4Ts{m!2
z=Tpxa=nAB~a*ldENDhyjnA2(w5OFnHIxUBBE3;mwb}kSyQ<G1)r3?cbfa07yuYnWP
z2JeCoA%z?^PrwJaOs*#Ya$cusa~^zw^GY7+MJ=gt`K}Lvoh*3U{B<&w_BR~KOFi5b
zGM8#oWO8x*)Ae)_G**^6=eyE{K7)F8mZqv_8;_Ga!Xxd9#j<)%g${I|+9H9ZvHVNk
zMM13t=*VpYj9sS&-?Z6(VX&(Ez{Re`Y){52JLjw2Pn;NdtPMAaXNuIN=WQ*HgH+N?
zhwC2*;_6}^_ixE-nxUo_YVh0^Ss(2X#efG};g0!<Z1>~e{rwE7>Yq`C^XWYZGXUde
zBcA9{f)xfjD#$6=0ga#iBG&iZ^yb>Ty+^0j=l_CXeYfi(*?!4)XqRVCTN1Ilkl#mH
z0=WQHj`mPEpewa=5iRGyVy=+e4btIS8gf77lkSzHb%stUY9@Ya?C@gaGcPkk`tPs#
z;%1oDXjhCu6Mt=NR%<)X(wW>7$CxJFj&|t<%i7!9BEG_mXvdy9T>rBH6$Q{rk%C%|
zihI;Va?w;!wFKQ-C&3*sdOX!b3TN`+r3eYuHw%Oac0!XC_Q|d?lzR@b>P`9jZLGnV
zitS-bS@<>yjAEu{lOdssYz1zO_(h}Owv1I!LTOh;n-9Ky88<dujPR4iwf=&^OlHyW
zs(7VrZGk347it-r7j4jQ7;RW!GGa7QgGcn4ce$!YLcNJ(yQw?&-jzo+_7L6}L-*<-
zFd14)&I0$uB0v(*qg`;291{0!1gu)EP5jw7cgG-gM=MXA07EpsZnhde2fo|a4{VhO
z=O<F7_H`FW9mk^E3f1O(f%j|f{F6gh*I-+AVZ=^@5fX+t9XP*W#t$X|WE}PZwd?<4
z?46oLi<T|jY1_7K+qP}nUemU1b4}Z}ZQFLObMH$<o&B&Q#?$x#*;}^EY#$sY>D8qZ
zz>4y?+``Cp=1%u+UcloWKy=C47YpWFQZ^aPt(0d~=K`vw?sUl4Je`FHsz5D#cPGeJ
zN}Y1of%SQk%>VaI{mdjB1N26oDa*(COrOj9LA*ztIO-`^5h^P!<z2@sxP(VY{t-!m
zb6hPK3|LhQmXKh$X%T`Nh=_ya7*Qq}2Mh$m34E5dCN2ywNx#X3>Xv{nFaYby^#Mii
zn7a9sh(31wAEXj1$A7?MurU6Ya!6a-9=8?IZ?$es95TRShrS0QC6*TWJPEddBL0Yw
z(TY7Fg-tSG<hN(nL6UC#UwAUPTfmCHk4qcQOvqq{eJC}b@4un>$HC41!FcyV(*KY$
z%EDfg`ha79K}YjhlaK>7NlU{<&$CXmrK9q=zbyVIq;>EXFYn|;3o;J#0FD3PWWzGy
zggOF0U>E+0cQ$It%#X_c5Q>UNp4$0qPKLa#kd;v$MC<dUGpvr@11Yc3;9ULVthd{%
z?d`ZweZ;Ci7<I986xu64#b9Dm((|mYO#hdyyVql}PRqkFo<aC`zg3@KW_xVYWTwwh
z?jPMEh7&Z7ry5pfzGyugtdyA{A=DB^rdS&^Z_B&BBeMZP=c#Hu?J;uR$4c)Uq~6_>
zEte95*fF9ZFbEJkXqx%WXv@;c8qS>QJ92_dyk1|<>^fcTMtWkamg2TbH)y&!SDo~A
z`RR%+Cd*UjvUoGoxnI*;JtUciL%jB5GB~2Hi~}3Gd~f>cmJ*JcFa}Oz25O&%w(2fU
z#_j$n4}a(PeET&p;EFvH>W}$C_+HuQyETLixB2mt;}2b%fgY9wnMR+PMKr~UNngZx
z@#I+Nxsdc|FjEiSkHnzHnl+{MhsTmmc4j+`u_%&OQ?yww{$)SS#sG<x8cBhDlT=|z
zUH+`f6KkcG@2s-m^XvJ+x63`*`ijQ6Mdaz@K>xvRrY-2=0Hxw4JG|#`FtU<qCkVLx
z8b-}CZBU}7eRE)FE+^fc_I`^7e~gOPDPeVr`k3&0BmVAR;WBEm$SN$ez;qy;1u=2d
z@s7rjwB21^vJUusFJ0f6OPxU^8EkId!KjRd%$3OcrMKm!<kHjicaf7tjI-#`#Z=Ck
zoPT)vzdp>WMAIF*^b_5j+XOTcj)A-lDx+=!=|;IN2af55i!O)2jkO|VQsW<VqGjoJ
zu%=jP4Kdn81k-bjdMTQm`0p~-dFsTqdW;+ETM|-_j-t6l*}?J)_0i_yn5^P=Mf0su
zCrXlL`8suw@{X*QA_|0J4y{*iJ|{_hgy+b|wuRm&#R$GM@Vw+eG~uL2q9}Fnn$9_D
zv(!nx6dlOwetG7up6r{WyAa|n?U#>oz4VT)W}eDuTpQN30XhV?ks(C>TqGB+@`5`=
zS+7e?b!4U^>18I@(2Dw-KFGs#oRhe*Cn43ATmd<CTw&1!Q6Nz-s|y*kJI))DK0Vpb
z>FOjGOoz)|<oBPprF}b(5R&%x(cZZw^Rwu`v3v%NdE`;=DE2xTIVP)hSR!f1XEUZ)
z$HRIN%6c(hqiDbF5%7)U%qA|L85z2oPw{9mVpPtFK#j_2^GA2*S#q=63ZT6`?JErC
z=yRwI(j~&eNZT!!m7d&K46e?!iO!K}*7=tHkBtO<qdOUakDLSWqsGuaCcm8=`dX!|
za+zY0><`j=^2Qm-E&BJ&#)r+vg<OZJlZO@!hWXbzS|v5hI+>3L*;0RUFEnx1@kL`G
zJHy$Dbl}7n;Y<hsqd5`pv49y)ea6DkO(CYJRr`a`ipUIv$Y_9Jw`hxCH*~|HwP;}~
z3up#^F;<#6+Sd5!6UY_hIp7ss@K!_<A?YkOf!3bqf0e7NXd4NA5UZ{IMG(H_hB`dm
zV(#G>>9fO~sokv2ofWTKvZ0S`qTgxqs~k`23CV$UuN|_B>HYPU!@_~kOCB)Pj<aJ^
zlyx82TE@c>FOC(`S2X;k@!IR8Ka-<H(|Ql4nn`MP;QnH2P2X`#qraO%7%Kmr@zf@f
zV7+wg1PG&e)J>RE>$A~YHo2)xtm<2<o;NBw1XF2tOD#zTW{wjk){TNhkm6DA0Wk|s
z>^sgDSI<*1^@k%1e6-7R){3+xk|5HrDU7?4MqN(05g!v4qeVW5hwR9h$kMAxMf3<{
zsq@P9Pkg>gG1lPx+F;a)UDAr}nwc)Jwq<D?SnVG)1~p&Sf=??XM^ffwF9jPqe_G<p
z6sTF~S>J`~h@OlYNCV%aThCr?CuyiiW^#4RSZPY6vY<|A$YjGYYJr~4Obe?jz%U`X
z@n$?|2?gSHwUfwF*!NKf4SKJ*rh?>0QJ0#g&|g~kF{Pk=pj(uBTA)+)p|<trnLH!B
zNF`yeO7z4LSdBQT3!U%)hB4Gm+^q}M35<<N3KfBCr?5if@}7aCF38vuk7WxRIvQ#m
zG~LpLDtp4Tx*RTRskBp;UtQbRqdv0=vI6}@8@+-uP3r8Dq{_e1BU}N}`dpEue^;=k
z1&FgbYtH>dtgWg{GON5^K3js#^z|*rRbYFZz_n&~1Gq*!8{+(wpU2O?%Km0|zzoSO
zZxn$V^yN?jRA{?EPjy(zDKW$V$_6d8U)g{bc5&Q`Tg!9$o8(=}))(4VCH{w)f>Pyg
z2WrpGge!VVc`2?{dgv83M{9CE*k2rRpu_Y_+85d_R~#pr^7y@25nSmD6y{2DCZnkd
z!#`eCK3JE*{^3EJiP7CzK!vw4ZOGdk6R9&!)13AMmscQ^4=Etsi1p?8XKXQ@wM`34
z8+Vsdb+=)KMyoC)yV34?njzKO?~{#;*q4~<+i;P$BO|V0j_R<{n<NR$(Nzh_K<TUt
zUq4qDS{j{)>ROaXjg4PY))D2NsrrDfj5cO<A8Jc3nUSX-p#DfcMuI?(cYjlj?H;Gi
zmziq-IrOJx%qW$KR5kfi45b5xTL(%e_nlI!E~1b5biRBvUFW3sn=7U(b$YD2$vT>i
zWbbee%w^q7@zBFeR|;C>jd%5Cb)IO^AXTfDt8cqfrfBQ@Sh_#{vgB;?8HYYvD@__u
zf90&qy{~`CmUDL(w}{NAK;y#Mt4V#ou_UTvX!1;7Pi%WoGwj|pA*t!UAFNY-f=w^P
z#I}q%4v~6>@5z$n9Uo*EwN-g7j)?DcZsDVV$Ijci9wHKCGs7w&f=aRNHTYXAN2u4L
z_+}NG4c;Re!po-pj(4vCAVA%tNQw~r1{0w~2xOWXCHJ5`CbQGli!=wRkaGwJUP_^N
z${A(Amy%L?&QvGv6~`~Zj&Dwh%1c1#QLF>w3|9mvL`Su=^?rGl!6eD_0i_Vq_4Z;}
z-2&PgsFHFzU2K)M`HWS*5kfX#q}A4lx8%bzxgo^p7>h`@{v~V5^kEePUBc`y$~*`w
zAd*VgXpJW#Ojm}Y-tgprK@qwx^!1gcQc9PWbn=ceV7)L=r`FkteTC3m>ze~XsFC!3
zI^;_gIk0QH(!{CX3QskeJt9EFO~ViEn7c91URBiuPf<y1i;c*JJ{?R?E6h4_h}}*`
zEdS#(c{Pd)a;c!{MnZ4)K4J>GPZk2PsKA$_Xz=CmzInJzIn7dOI~2IGq!^>`Bt1je
zK7E~5t?3k}-5olu8$jkCmoV4fuj45~^noJdr7Z3ofSQdb?i>;c0WD>jE&)3YR8hg%
zhGbsham>YbbOz^93p+xt<SatI?&SIMbiQbwH$sTJFKj#>x_sCcKUC_pomoT~%Ra#M
ztR|rV>VLv5=11*rfvhSW@%>Q8&9i3*I1%3WvDlo#NyC(3L!oz-ywlIT=sBntULX+K
zD0Qur!fq~SDXv^BK29qi3ByBRSEUv&&s`7q7i1)ji2;2`vL7UMel0I_QVyZCO265#
zAZ93xXu)w6swL^oK?24lpp_YK3t!S_MjS+?PN8o+wRgxYM!=`Z_t3ejVk!fr(zL-_
ztGw<C6^OA<(@Wj!^e@34qgRd^(GwQ694DO!62Kt{T)@IZ@*~pcHUI_e_)I_n2sI!Y
z7IgGIsHsL&5RoHQu?8=_WoZ5-dx<<3{r${Nlhai;I_R8{Q0Ff9Dd;@anO(2)!*SX8
zDG5LXk2D{-Kf|}=9_ggI^LhWIRZIblE-(`FC`jvY(z1q#@d0nH!<`AH+y;eA$?G*1
zqI>m?jgMj(At&quK;(V7s%Sfq=T<~K_%Pf{w?OWK*1ayV1sFk(71S1NU0-m)C(UYg
zZxCAus*FO@LhSlVBsQ0ZG=20dS9HhEF)pVztIf|{j=<{RRwvStz{ll_27MxLCADm+
zNbT{y2sjy)BwIb>b)JD3=AP8Rb^5(Ruj(ax!j)8<`3Y|iAG*>AR>_BT%8Bn>aG9u=
zEMbd~lsJw4KTa=qDekLy;NO`#f*<GI#vQLJjX7G^5AL|k#n8~vjhc(*uKBhC`~l8w
zSa+Rw-9K4`x5}i%omOOqNRJJ^F~;kfeiX~n?bow@iW@4Pq?;~tBbJ~PT9aZA3r*{7
z<_eG)O%u89HM>-fKHc71eAQ6(uzjHz<>#tm!9rOq?~iYHS^#>uWK|h*bzM3?FLCbN
z=uUh@JDHPFIaRE26H@Fbx1&5jM}_=1RTO7hPGt{$dd)gNDYX`ESd{muIdjg~y8%XN
zMxq+XDY1omuPjFyse9LaKKo`uJ*Rq2x$PmfCN|N_Jh`!toiD*_*hMkVY+Nl$+Q!G|
z#C{)^=pA$3MXdL4Ilw}6E!#^2#=t?!(kWgYDe{wy@bWxbWcan2y;8b6KOgXPl>{;W
zsX^oX-ynD<HkSV`US(Qa`X52#f8tf<98czOkU;cf7gZnE98;Vh%<N+2W8JHC)_K_6
zOt1Cv{fS5u5-FaIl{hVyrpW*i00a^6#oh3^*L|HI-@KfE3+tT+u#fY-zvn1F1Q+y4
zE9Nt2C`qf;wlB+WKI=+Aq-TkFDN0tJI6kh<ZDYY}c%BvDpUMyW3#7hT=ZelAk^jto
zA6P8@Lcu2l!s(MwB;wtiCojn)eN+g{9^Qwo+S{8T^`<U!p^(fIP~S7hzH@$r%8vp|
zo7mV97A^p^9{JIrZ!A%f^GbO66uc_mnELR;oyQDn9Jc-5P2&V`I^h?6-p2Qq)ysG$
zCyX8v|0Y`YNoRJ;_!2HiKENNAqYkP>CPqR^DrK2Qt7xRX(xY_{1cF8>e_C4yLn@p-
zy-e!zo2;9TX#sWBOy}o%7a_v8VE6V^^mKc7zpqYZj&KCPb>j1H-@%us%;BAxqo{q{
zJj!o}W!+jeD388G7ALP@DRZ<Wl5n}LPy_&IPS{wvNUG9HS``}(=FQ@FRvv#3!8VgZ
zs>)&>H7pWA;a|*nk%6)DJXlFskLS{*t_rndl4})38F?;r<W_{KTRs`4URJ29`|y+V
zpM;i4IF?$FL{#IU<nZQEmMo7Zdz2nVo|)UGHr*$)7~smPp36aIuqZUiu*Q=H5JV64
zGjP?wTph<L&6%@kW}*&)3XW7&)J%Lj@n~QY!*AYipT9t(1a#NEQ_>W3#}pz0K1Uxs
zuA^Aa2!XN}S)j8*GzHI>cd%Fd5M5Hy!(>`(widSxcHA(XQ(j=+mM*|^Cro&&WI~oU
zuioNDHRi8b#?)5aq=(3QXUTNqlQ(kp<NFte!8U4+YjW^J{9{;SSye7eVR(B7ELE-a
zzq2v*LqY#cuvM1ZZA$cBzL&di0Z}#>EKpC<S!%jErK`^iJ$C$obh5ObQmZ1lI8~f7
zv~`|FtE8~t9_M%BJ6@<Y8u5HEPk*`Z<~v$6b$b-k$XsP%5{LwuH8n#y#abx}nim4s
zSKcyY=KK)eg)*H=AD59##$K>%tp6TaTi;<Yhp9%JiN;YXkG8DESKW=$2oBerU2J8^
z%r041WvVVAnTY}@z2SwHmyXCUY|ZRSW0u8~Anll830R{{L7tW|gwf$^czp5<C{7jP
zo+Ro06?FSkg5O&*bH)7|q00|zWAfU1Gn+kbhw#$070$p>K!I(6kct~FP+=<XMq=oK
z;3C73GbA^jj0#X7o0?6Zo@vLdUB(OwE-6|ONX5ofvTl5Ke?B)JC%fzVA>&ucp~6?{
zmbB9z$KN0wN1bSVDX@0qBjjaXkn2@?nAjdpL{=}$KipcNb);965_Qb<HoG3W<3vCU
zK>u)sOnId<DWY1)Wlp;)NbJ>dk9(Zjf1hEY=O8FvHpcki96=)XCrwF{osuix-ln!|
zaf*G7r*g|IuEzPB0F#uaKL(PGiFP@8&jTP&KCZ$_F|zQauy{p>S2NG2x+YyC#Zr~Y
zKO30p-u{=US!x*O6ivJUQA>#P0xt^`VXucnS!E*cp=;zqCW$|1gSFj-v^l|jx_#js
zS0Js(X?bKGGf}IVUb#hpI*^DOOZwdD*1ffInv~<gP21^xa9dEbvtP!Uj?Q-+CNp%d
z2^i$mEK=cG3k83dWvO%zhYIiKuUo4H3Aaz9acmMaa`)mekRF%A+{}^ZJoM38fn`>t
z=4|!zcRa)Faj^(e|FTnU-G_7HdE2Ng+DG~@JvF>9LBU_N$f^zu2QAGSm9%7LMI%%r
zGeXdsc!p`GPo@(mMwCGffaVIbv`Dn7S}2JfS?Iv3P%_WiyaxBnVbB80W6sR(C871B
zqP$x;t(~7weEyc8P{jdhZKqG;6En0h4Fs%woQHzx;f04J4Y+kUSUJdv;z`BcV%1M;
zpH9bTI%c@g>SOb0Rp(GGD!u~Lpit5{YiYd>1vt-L7OXn=aIwdMlA!hL-OT_{3b1;3
zxpp5{cYsr{cQvcSOp^R(oyv#R7PP<eid5SH=a@fVkK1ajJqNEP(=(F$LdV3wNhy0%
z8Q{T6VcldWx&q|@C5-AKuQZa!EaTUHl6I?VxQW2+UJ)Y8%~MfOAZL+vHq1&*-jx$#
z__r@dVjGx%e$l(|oNllsNzJ$dmK%R8rK>40^5RGc9?R~B7H3w+G=<U?kAj3l2TJ`{
z{n9G|W={wf6}WJ;q%<P`jH)pAMb_ztHXN8au?34#W``-kYa^h6ItbUQRov)6KvO7q
zaEAS*c)<9JF08QuAik1qCg%*w1KeFmDl3#CqTWwUS<R+;>f9HD2TFZDG)Vc`Zigl}
zDfj@>H>0(&UvOf#v8Bv<;dBFOS=F{l_6pOaGlk1BHo37?uePw43pJE-I6e$SOd3${
zmkOMjWxxoJ3CuHSn0T79r2gF`?p(QZ%3)RcrIs8Te--+d#~q`v#D@0)CvLn}eVl;?
zJ=It<H26)IPI%;I#pLeNL&hz4(i9-_Wu|VT@3}S16!WYLabOC~8n2qtmC@c&G{11G
zhe*NIZ;%=YwJe4js9EP!A90RKA@=RpK~z9l<LN_79fZm%iy<NPh3LGU;V`E=@d>q=
zJ*=;PQzIgCpsHuo<D<)EO0rS>={UqJRY7r^F$yxToRR&c8u76KxzC@*&Ff7{usrq3
zlhXtfnO;(*n||uWAI>QjP7ZY$gD))z@r~d_OlxPE-WhnJJE)y)LQjzw(v+XQwbDav
zw6*!W9*?!~iyP+=s}-QZI@u}P37Vi?N^;MOb^6dBTIV(vAP3h{QSUKUK=dB=2Et#E
zQ7_Lch#SaXOQP3RoZwpPnj2_w6C~~6Ir1w?Db@PmnLt>lL4EX6a6waFH+VHq#wgFk
zhm7GQ=FSV1;<4YfYo`Q-xAofh3rLS2VLFHDJ`6Do-@#5sBo_(IZXsO(X%{jmudbK2
z)1To6DU35^poero1vC9;w{I_U0grt2@tXxPTja+F1w&~w4I03%1XBYus*P018@5fJ
z5y#YVgi`Q1x~iW2Ht+g?r}d6^kBgo3B?SdIb`?FsQL9aMIrDhH4!PXfJAf-(yhNjX
z4SZ3t7_t<a0^w75;#yw+rycMDnXvXdPKIrc^9LWK;l`v7jSc+1;nSp#jmf-bZbNp|
zM-pJ%-W}Z59$D()6O2Rrqx`KUpxvWq&}a&bPW%!iIup)gS1eh@9>+<RP!_9@U4gUb
zT94wcime9(X5;PLe|G_(879UO#fs}vrPCoh$>>oDEXV^(zL8E8B(bi^P8)b+5wdFT
z;Tw#jJOuu4aR4e@>O+}Rjcrj|@)UO@<}gMd?Ah!B5-rH*_l>^?_q|<6>nJP^d<flh
zYsdlUffmOQA52_of#&BaW7cgn4B-sQ+G_3}@T@fWNBWMg8<rLG1qy8f_sZEi<Jq)i
z>oPYV3!|8(It4P#X_f{5@wF7$ph8gLp5ry2xDnR(5e|c@f<e4hld68`s7f4dRI<mc
zg{EJi<;XCC{~%J?82<MZfr)|rzgsb${+9u`#Rl_Fq~0-diXckz1{efP*|7=8*><#&
zl(9iVv4nwvYCWx{HJHYU@{hhW@lMT1e@G<!djVo=d(=9-yt(*u^!O4hx}eZ_)a3Q4
z_UKuDL8tM#v^@ciAu7x4m&4SxiVjCJ@Wrf&{5<?{-`01!2_U+F^d5n4^gMt)c?tRb
z^>@o&uRFY2ZF`1SS%HStF*P5FYo78c4Ij9|zuI~~pPp`t8$amrZPORGJ&QC3<=j^m
z8*WnfX{DR1lOU%XkDW1AX8*c8YDjA~L^rB&E^3B;;kLN|M4gB8buYva9AXRNGOSTm
zp3P+dt!bF@!uVtMH)^dMZSTb0xM{sxL~hUWHNeeg>B5b=V;<|!xPL$UsKu{F;=qCk
zQd@jE>`=G(Z-toYb?>DNJ1y?cns*~Bu7@v#f7ClmJ+w)~ji10Uga!#vH2{*#v_}*K
zpSnakFkExXoWXzH__6uTaq%toP2bH8v0)R2xO;g~?{s@|efuRETGZ4AeVy8uJ%tb|
zpsb}2?bUz$w0}5#C>Fld{qgm=*2o%&&ewEyLr*x6-lXD0oY%tKS^0kdX-3NH`r|#H
zg+mR2gz4(oOz%EWj&N**RX*<#sx!lJ<JAmAc?T?+JubF<(8j{TAm35x>NdVkl2PU7
zH<mHtHqXewLgZaAsVW<6MlyAKo0at9tcDwByfT)d#8|8*fO>arO{3tm&?-;`Y3?&u
z;4H%>Z^LA~&^wvfvTQcss966<vQgZ#?b~pA>5JYj7Q!&613z$qrLrW335~?M?ar1G
zdhbl^5t6-o-0uD+{J83q0{<W$(K3J!-`Ug}a4bgAG+J%jgF5nTCzH(=Kq|^8F?5Wq
z;0<j^q-@WX1`h@Kd0IJ}@ClVQtKvPTbEM7h$X1t48qslJW8dNBa$O^co4hJF0m6*9
znjRUZ@|U4(JdHL(hC1t~>4b*QGf%6veDne?vEn`?d&H;k9lK7)B0Ex8_A-^JIrDWD
zdqV8k76-$si^T!Fhy_Ncz31IVO5=|#6tFM9CRg=g@?ixvF0`mb)^G!3d><`zz&XiK
zRBc6$m8(sy3O|qR?UI+U(W%;^f6@cGl7sfdaP3V)T(eN$?3_2AeKX6-MYtNY<bCo|
zix%xb$x*PD&fY3<yv_^pFv$CUQmx-U{qFri^65`?n&0Kn*+uI`+0rB}0C*2F3=E}(
zA)JOfO;%PF*b9&$CDk<PfmL`)W=Yo|?gtGGxLy;Idy7732xWj%CRHbFYyp|Z!XqcR
zd?u;z56#K!&7(D(YHms9YR`Mu>a~194l~d(WSgktJNE@)s2+nJjcTyYY0&hDmQ$dr
z;9~7v8=BE^jDf~z&_O6RVVxc>{UZt>IX~VA9qpm<OhnxzSk+94>x_;lEJoB8_a2&+
zff3xyk^p`i_98i$UDg;pRqGSj4J2G(RhsE(Byv+q<jf-u2>d&Sz=X~L0Z{8iOOqD+
zu{;En9N|X*fyN$KTgf5;d}{zUfN3E19!@_jNtxtc%@P2nI{2$?Ibe`t{SSLS7{~(;
zD*@VWJQhMCS~+0Fs+(1xfWk`XFCeP{$6p&BD-=`<>+%5I4P}>!jG$!t{*~zYEKAse
zSH@~F$d1G%gs`3c$Uad%oN*|Kp_FR>b8{OWf}t)&yf`yZkc-{{D2R;5YJZhGwQB$P
z#sE|V2=Ko^3kv^!!NhOY108wLxVD;rhIc0CUlUO~;Qa;art=1;COE1y0UfISt_B#%
zFMJVi02)eaEjW|}1hUJ5y~+5f4b`Z?l=sCrV=D%zN4SJm^7k_4F=Zk+Lgagb2z2(t
z^|s!4eSwQKVTZc#&XWE4;(6#e2K(}|Gw7U7^$CC~T1O>a6{_w+hp8(1PHO{n!BIvS
zZ3}qH3+_k{BQq|2MXz;67568j2y@{Pp9A-M83hstph~1}_4!AU=KwlFLpp|3a|V$E
zH7=Qg7DpHX_nSUvS^4FWM^w1|Fw9j8&ps>S*gAECR^3AIxWSsF<w-!e9AO2bgvbmb
zM~hwaqPt~9qa@)RCCFkr9`>57Ydivd=zi8=C<Z+lhhE~52a8w3C#G*<21^LbA}e~5
z#zcr~>c29BL0I=hcT3&-SeOlA{C*K3xK0lPct5hS41lofim;FeYMF7iy;M$1zLX#2
z4GbcfFMWd(0zCt3c%)Hm4e`8^?`_Gn*jV@~JA{?ULh{IMq+H#JZ=eLFWnlZqzz&8L
zN%ec5-uff4Ca{-&?}2biIJbhxW{Jj%JyP|=9dZUvzBcH?l2jvrl60*aKK(=uh;PP;
zQqQHkY$wzs#NuRlZ)lM_UNt%_=?Ux<Hofg|SZW5dh3t?!Nie4fs9snD!O<)4UFK=P
z0KT9E<oFn;>XsXI9Wjr0oPuN<qP_qhylZp9AJT=3j#j9e0B#+@;b^>xqEc&Tcg6Qd
zK|$okAX6c~_D<S9utNRym(f?24#-p|`C;EtzhfplPb7zc$TJ1f&mahw?_eapi+@9$
z^n#1ZUHRR{!p{?`yK^nYzq8VuE*!}5fT=}Lt@eiQ(EqasrRg#zZCtBjUBvLYoTgAm
zx9ke`%Uhy2R`>6tDJy(PkjC|LjDp$=aGS?L!SF{wpt{3-G8tHIZ3%U&^L;W^5?-4@
z;cl5q)-^%0{n#)thbf)R0&JVdTx#7Y5$f$}#haldLh(u|T#th7#V@gK4;5PLFP{Xg
zmGKVffF57(zlRAJsj=ce;bhqu|AWq8WBKoNMzglGJxMEK&!zg5qbow%FjbGyH9+S2
zGT6&FHZIUw4@W&L@Hzn_+St#Rk4d9<=GCdhLjf!Eb~X5=?sOvGm)l%e9D7+=R%;v$
zcwHc=J)C?m9t1qgPvVHIIee}JXX)VL=^*^87IxTWuym8z$nibz-Xh*@I*{5NFdf&9
zXF)?QEZxAX^T&4G<Xnrl6&@2<>?>&FNQ#=*Lbk(3U|JkMudjCz^k=X<mam^j$NQ~p
zC!l$b5pCEXpB_p0U`a;?@IfeT?yvjF{ne~na2Q^*8ljKSUi?!9ud={<$YXf^^FjXm
zJ8ha147lG(X?}Q<r+xjf-he)|gM1gUDFini;c9rXVFc(6XsAg=%ZSyYwebw*i4fQ7
zV$~+v4I85X1EUR2Yii^mOU%hWTUC$;BO|O6(ZpzoH2O$+Fa*b(iEigdni}`x_lhlU
z+$7M8(Zig@6F<Q7!87mQO?0DE!~*44-8^n+9LTDM4nf%2l+{oy8x^{RD3~KH6mDN2
z3)0NfH<3%9brTznD5)6@(=ixDNW?LrMzqX^xk6EydI+MH#TB(^XRIHY+R;>_Oo3e_
zgn;Ny`Q4GXII{KSa4}8L0WAY#w*K77+o$)Qt=D7sz+%8i>uewfPa2lzzJSYphxWix
z)YvV5>i$vPuWW9ehj_Km&w*`bH`+TCZM8^i_l)nl#Q^3o2!3|C5iW$o<Tr)$y=i@W
zZiHKN!8NK&ta$wT?oCypzkd*G3wmmik*dPKgAulOGAwqwF3;CB4MYH-!z}!t!P4;f
zh#zGZcyMq3#YGd}q>Pri-isHx+&zMvc7vILpp}rO*D#`d`tCB$tq&T+jmOHEl|2h!
zgXRgRBwY8l&M8!TlrxOTSEE!i1wIV(buP!8J~0vrXN#lVIBhqOSp{PbGw@g^w3&qS
zDLDnD9cBtL7<t^O2&#htZmP^XmnV`GDF|*m|Aq79MFair1cz)4rl0_(%Td$&ggdys
z6oF{(cbGv$>X7+5dEH?j;L$G*RDgYmRg1~eL~$k#sJqOL05fMmh?tNbT<=Q7(2GiD
zm|x%vGdN-l{kLX8e1RV0!Z~@^1)o8H*(r)AEjm9OQl5Tj5IHg(m8cTEK*aF9vqe*M
zmBJoDCFLJX&*-J2o&ix<ioduHk!hpIbEY8JCEWM)Go9*65MQ7aoQPs14gP-T$~x8%
zq8`7*z+V`qQFbIknuZ1Z?bBJN6mMYo%%cJYFP>#AGm&v5-84QQ`SF(2U^g?_Gt@A;
zFPY~!OB&|xF3oF@dJ$l!im?pni+9UWULM3eQU9vKkfnH4EZuT*aO#1cmI6$-IjY(_
zCTKw&0mqcu&Euv%^8FWNQ~b1Gg>o)lN?dbxS=k}uwJO)PNfi%L69l$^KKI*nxl6%_
zs-XvL{3PYtlQ|1}MG9v5?(%$#1pm1fXdS;BCOa-3)eMsE3Na$&&?K(BbgaF0L$d^<
zs}RjcxH3VpHv_Yxma7J>#J5c>H5yEL&1S-hFdkC`aw`)ebjNOp@yW_d@Ep20*co;;
zv2c1%T&sn(RbYu&bSkj!297n}e);Y=YqL}2@sj-##aLmUWwUtN7wVuXCA3(RzDRLy
zXvMKtd-}jA5C}s+A@U<}_|N-Cp06h$X<Bo^HNhuFMc7!Tu_}goW7fAmdct*iAsZ=}
z-rqVgl<^3RoYsb+DACKn350$bSRcv5$7O^L=D~4=Y@imnBg>*O0#n5bHI<%fbcOO}
z62=oABT1c_e6(<+$aw|AwQ4p37_dL6Q*L`gkk~T+;!pM5QE*iSM8_$FA_nN{m6@~!
zNY1l#mBc~i53Go=9jQ=b%de^8E8DQ24iRf=1We=-(sKGiJrD?lV4qE0cggzP49qH8
z9b~V&lj{j6mv&GS`21@Yh0?HcbcCr()uOeuE<4yQHjQ8#a7h|Dog1ke4FRJ4N(^s0
zVNcm6ER-epkciV3P&SBHDOhb#&MKHrv5sC{>p9B;?OuB8ZXu$xSoid~oX7BcewSUh
zxoG8X1@G_c-&wr@1jo{Y=eBSJ5TJM=7xDyHEMB{K#Wf`HN(;z)L<D&Kj%_QtYS92-
z|NOjN&uzUw-4C#-$cz_Qb&08q)2kbb0Jzv_N58xJc;_z)iv*XB??@GPMz)A5TY9Ft
zrI+a*Srw`}j-8#WTo9?%&!SlesnK8cCTVngY@wIfG4<;5#;E36hr7J$U<v}W2e!MR
zgFzA$a#UgxXKaTE#!Ix9>*>fTkR6Db*TBY1x1DF!rEysG3ecX2CJ<8(xiQMGoehoF
z;@Eavt_!sh@)xL7ra)?2)CH3t?EkQK<#{&abEVM#i_<kTA3;?Pl}Mi|DZ!VTQ<8la
zV@?qJLvX^=5_u3l_vM0Ae&dsQ&ye<!RzSpR$zT6xo!g6+5oM^VzOcV#>~>huA&p<B
zu2(|Bx1eP8t<M!>AAbNj`u4P2>T=70y{ggqGVPqWn+L&@)^};m?d<T0$#>~VPZ%F|
z1lD({N?#b?B8s9-3g7I;zvRtNXC>m;2Wit_YXvec7C{`c1n2pP#EI&D|Er<9lyC9v
zUhfrmi{`3z@&k4Smu>Clie+P&)iN<Tf7#l-L|;W5_PuGOBi$GHLYhCJAKFDkF=H}&
z#QMB^q@#wIc=fXAbv|?@Q5@XGpjOvbesz(w4o2!1Oc_teQ-?+rjCWPRNWYZ#D$>=*
z_Ig>vT_hPE@|Br|LW)M#?H<UUD9Nh^Y#(}IB>xKRFfienuuzY;I7Z4qw?awDVS6u!
zOuiP)*5tUzM#AVzj5r37wW@@I2s2NO?fV?Iu>-in9=|Bp+Cn+XtR~NZPk&==U{Xms
zJGOpZCYnSZxs-D7a+<FDcFvG0>)nXATgU(onHbi4cDJL{s7oxba%qE$H8AsF{{GfW
zJ)YL2J2npiVoj--G35E9$OZAfp@b(F@r2^^y{j@J-5b2YdLk)Zv9bFko?KG|XS9z0
zNYb*QZ!>M`PB+Gg+&po|zLVd4B546Ltw&|C<)~kSPW{I5op#!QERFT_0rLgci0_X)
zc7u_}&Iin0DO#Oq{h~uas*Fdo<kWK#{Q_Dk2RrwA#pqf*1f}E`(-pz!=J>X}uu7g1
zlVw*ZS#oIk4(7sRkxh~|8eDe#b`2aYx@Voz2|GtGcJdRgY|W{=OVN6h57by~ZoCIA
zB_@2)2Aa~~f?omdr{xdz@kl>)oH(p8b3;56_1N~&<bCW-guIJq0c)ybhU59s7CF{#
z9j!Cb(p1C{Q_{kIdInWJDdvpK<%}$Cp~JyE^`sjyG_c^dq8&Hh5<Xy@SOicDJ?Nlk
zb(&kXnM=A3F8pluX8p1iiqpd_0h@fhY?rSa@UqTgKIy7uYj)T#<UJ6N(_ZTG2=bbq
zRbE9xbxiWBox{~`%WVw7xqwcq&ob8&GqLu!hP6F-#^UHbMQZeD-RFhXsK$z@skD;?
zB~5IgdzG4a|K&)f50UvVevNc3@qco{m^l8|u{|>b^M5(6#h$byX}zzlOD}Ty{{^b1
zzZf;91Ko_7eT%%L8zM0@NoYXz_j83&(zdJ{`za+s*!<K}{IeM^W%p;-V&|>h3%>nM
zH_`P!E39kX_p1r`NAOX9?rUy62V4Qp@b!N3W)!?8=5RGvznGP%m*)@MmmU5sB2b+_
zxnKWbH~HA*{2lb$W1|oK<$@jemsKtjNHjcV!%158Q@*0X222{>zq0rpGjn8(n&Xq)
z`m#AWn)SQYG^HUPNpbb_axCUY020#w4;bArbL;m#db2pKul3hi6VyP{^0d0f1T+&=
zg8Q^k=-o0N6}g7iS)7L02y^~Ypqpri?O#slW3A`IerJS(iwt2?d`#^TLoJySVM~;w
zAQ=Bzlz*ew*0~9&${u>U#&wkmXdMLl0phSW*$kBKY*5?6=VyQ%HGD4eP)NS5!q^67
zpqpoy0Mq;DW1poL+@SqHvu@Ieb1J~*zuR$c1up4)FQznsg5Y3wek!rI2$&SG4`kG@
z44VQo;eXE3gO)|sQxD$`H|;33qbuOEjo#5<b9r3>u3)#u{^a=z`Z{UI+<`Vnm@<JD
zDkM6>Zr#J=7e#oFF0I<gK!*f=C@$iV6h7G`e;$-mBQkQUt5L$gplw3Dl?ach;$L4c
zuS&aUxC!gkgAms+)4uQ|TQq&EC4sByQzjWE=5=_Jchb*Nn$aWte_J#ujcWkB15Kc^
z7-}WUpP2IG;fwx6=-RnYL?GE9D3X{guyZGGGEqT=d)k0H`FJrCmd>`{OA7GRV{@1-
zIkUb&)|1jlNeXo>pNsc_>R>j|5ula(uFRMi%{sS4-U%-@EEDx*U9~S*<>DR{X@Kxu
zW5|yTitz{ClMDhLHEP$x;^sHa;Tscq1HR9}?SEdep0x(dZB%wfQf~q_BNuCu=RJz5
z8EJ!)-=u_mLaQLu?yExD5|T0ymX&f6<IoMu2oyjw2^Z<dlgrmqQEogXNXbcggcc6R
zb@IuNWi%WN)!PN|RUn{0DY-a6=}*fs3=Jdb&Al~~vyJ*kJCaE6Y1|69>$hWI*1*OC
zERNm;=yH&f{9Zcy_L?6?W*DCy<Fma_zZ40i0c6(D(~~?^Dqgk5LBHS{U!ODKvY|#=
zS#t!3OROo+p@Jn72n(RUJV8ks7&QXk4rjey-G8pV^{CI6{wz+=6Qgy1uHMusG2!`R
zuK`2oDDBC>|J`vE8)Dr8a04=Aq>{aNK@e9OIbg#2G+1XSwB2``G*S@j>fGny9B@f>
zzKVQ5KGHb?+C4V!@ub_$d$e!hH*3sDuZ%WESzY(RO-wt9mqE9DP(cvipbIiZ%rWk|
zCX}pfHbhU3+1b}}4+l!aWJ!5yH>j6ahTI(jRopDdMdmzkvZ)3y9u@hlSTZbv66zsJ
z0BQ&!74P!+FofxqmI-C#&@#Xpyfr8ke_#gvgzXVhNJy##({{?Kz&|<0PMJxc{?hui
zNX(nEQ)$4=<bGoy)jiiNplaU*a<__&zH?MNScFN<Ys&_Q$8whf_&a;p8ezl^5`6Fm
zes)+F6(3(?pv!I|BL!$>nX$TBdmVpY(Y`dM!IfZA97e&FV-kNA5&3nhWPUM`nQm%@
z4=#C8PkMKg;_G07%w!|-*c2O-@K|m6!p#<Ty2-zLOsi8`e#W?zL4e-dH!rrf0=(V&
zJHy?jFB+K5ZKY5A-a=4cr44~fe)*g#?@KdOE4S%2DNhsJSqnf@zGvKhdCBhny0>+G
z>;8IXw}}4dQyjn3eIiNIAdD~TAXaFwlrr4H6}n7^k4>8|nbuI%IAK3HUS?*v;MX}y
z*DanGo4RA<p{$k!QKpj7O43a>Vp=U>IP-5Fivd33r;tE7;`dGpKxX+Id$`zg+cGUh
zLZH@0(~a{`|AN=riN!LEf($xZsrA)%_x;aHS`P<~Ic<HyC%&)GOg+a1ZdklXs%Il<
zUSgfEPQZ5XJ$0AsbzIGslJW6(QuMrYh*TBH&kkuq6HhR*<k5j(!YejOEwRz6B-@fU
z-!5*FkwI~?FRgO4H%VTQ3GT;n8VcG4h(QZq9q=ndRi6(W(~;DejDphM46IOtGft0+
zwxbTWJcyYfWisKS);Qngm;^!-!8bfI{ybka0ux)ssLy%qQ82-M(kc0ZyP!0@Ce~-7
z7thOeSRCC_;z^&|Y|r6Ky*=dB2oSb)GxE*HJbpVwv04rAehUI0Rn-yIln~<mN3~rP
zCzQ_ZU;avouNwPm#P!~P&fJwsHF3j619gy>IzQCkm6TOB+&zC+p?W&rNuX5>?elep
zN!sDHybCcwPN}Pe=$ejKWxQt3WP@Xh=V&0FV!{KQ&3zY;F4Zwj)#l4q`;HUxL0KC{
zL3~*9H;CGug|<gn6UO2Q{Ykamtr#*?7me^A%LmS;Tk=XDgCTKt4?uX6+%dx{#M4$`
zl@NDuBw-aet3>#8QgoL{*)Bt;iL<ChI=4)@&-6)cW<Tz08)9+PGVGM6SvDt8+>AG|
zOz;`DL`4*=1fOeiz3g=X1d-H5n@Z(bSIUN8Wq}_c?Sxz!D~&Q|Zt44&o7izB1<c*_
zRfj(Lo(n?I;HRR@76M}j9b7%HMPOSKjwdj7UaYb2EWGIh_OA1J11^WfqLRz6EIgw-
zy0zb+EWB_%&%;_5^}>Lc6L#J?2j1&iSNW0vw<+1UMtjeu<U!L$``B!gf{@IQsf8ps
z>{pafNSn!5&<}ttqBqyma@eKA({l8!CkqEVU1AJ7uO^TlJ#U7rupQW3Ww6E}LwCt@
z(x7!78avQt#&p*q$X}t9Fz)%&80}`|l@DMk9!Mf#qXIgD%tt~g*V}8LLT!A=ES2Y`
z&=;R|%M_oiT2K0_0FR^Y>MKyQ(Ghzwxxos#aC;5{w(JtoY`PX3SU+F1Ulf=YwVOId
zcs4R8M(NdTaBMhlC#hOK1T^y?Iz1*a9?YSH<N{R?duI4lYEwG3b>YlsAwT*@Jx};|
zXHDdX#|E&nU2<UTBF_x0H1#oh?b5MEO6Re>SI=FZT`mSwK5dOKm%itMfT-%jf29!*
z0?azPeIveow<?4}v&?(Z@$az?7<&4S1rCm=3T?x?NFsn!W4h9&T9@NPGD`@kkh=S3
zEh&t#q~=ptupw4vEeVcDg^h5O@7~%BXQ^Q>x5OA$AFoV0I3?NoG2iVI?eM?M%ggUU
z7Iu3fkNKhM^BXv}?a5kKsnhP$uc6ol&2&EWB?jFb%6U&Li0D>uphtPkkH;<yy$rKi
zJFsGMFYAlzSbPQ;-<%&?#P}A0bQ&n|b!jo;Uwjq=)M)W-B*5|K8hz}2411?tO#^L<
z)3><p9aN1O17XmAMdODf4miw7A;~3Q7>tQ|$q{<m%3R-TFJRKXub|UyNEh%ffm&U?
z2=f*y_}PqYl{$MB5VX_*O9jUHX*{9?Vzx@aGs)YP|H+=qA~rx^d^7dap!N3({)sh7
zzBECpcYZ3RG!#~~dU1HUdo@M_e$KQ_WsS)CFcX|SUvMh2W;IL8{W6k4>}L`ZJrcwF
z$x$4q1JL?4rZ7F-P4w>ZUNof(!-OiPe4RP|7?m?tfi%_Fsy{^{m#AecM{7{m!09?*
zl=VFV#n3|jaZ4JmjMIDEsYU%F*-4$UP9slGmsp4hdm{0OeFCkR5)VP!#<W(N>zlcR
zdUZ(bWpl{?xSyrbsHV-}MLhwQqrT^3vKop^(mrA3t=a5id_lUudnGKlg0K$Dl#p8g
z;ke=c)9YDBhd)mvYO(v%)IwYD#EQOk_}M{igQ;u=do9Piv7D=YAwj8~+m54jH)Uqa
zYqBc&FkzNBSc-%byj%<U<V;S(@@2(;u^WIHzVh*p6PI_70&&1=)Sn*YMw&oIwpQB{
zc`n`l1*zHugKBQ}^}tvwq!T{}ScjF=e$&+<7{9#tm_-!0{VPF5*m8dI%GH@L(kB+&
z)Kc(8rJMWA!`dzIEPp%4Jps{cbLd@z3Duv*N&K(Y3?9C8MdvRY>%%wp+bi84a9yX<
z*8kKh{Obz-UvM5XBOB{~l|(T7f0@}K0y=2BO{joHn&aX2^{x<mZ^<UGAnO2(Fd$=f
zv!m+l<9@RgS;E#!+a~pu8r2__nMY;%a$ir|+mD;uqv{oMshqNE**koB5cC{BNTb3<
z@~PpNl4Y&*WqEj<=wPeNbi?h*Z8SgbZ_D!Vi8xwwnCQH`8lvspC~XkG`UAd}=-!u4
zYT3k~=lIN~vD8gXL#Fu^%(~xCkDIUVJdLv%?EAZSi@V)j+uxV9=wj6UO+4Snle!+Y
zB#h@n5yhqx6^nI$xP5y*jpduwQi#hV|6$&Xr~ohKav5{S@2D!Igk3<FEy^V#eBe>b
zSHH3sxZzxd_-g0H)}QE__R~mcn9-)<c@Pi3o~-w(Js$?$-I3**jFGO=B!l>x!BxxE
zE~h|$)H@s&%U$!$V2GIu{y7|%i$VK_3xgOBq|S8$HgCx7_PHus&^PJFrTB3w?C0PV
zi4cuz$X_hfwFHwO=RK}}b<QqI&lMOLEO)KBw5+&G2X{ugS!6WTo$F+<=rjLH5{pQx
zB$8#kbfKsjid<CKNfYV=T_1{(hEpDrE8ShjNO_zHRup+eWk?MNXIprU%asImHuzrl
z%VEcql%(6*cu+^GM!V12a;H2a+XOP>{-AhU<_bi{)#c8%i&By-s=fM<eOi!cJ5?j`
zmSkISiG?As#%c$t<Ra!-Wty?lHU|r{C*7O2{5ZQEaSgXO5DhT@_++!!W&9DZQqtKU
zPcBSPQ3X|(MAp5`znZlD5w<8D?`{3HsJOWcG{i~M8eZv0ycpA(uBuCll=qpK6Glev
zxOJ#fCL4Ca-V==_$ZL4gtW(VM<E+$>kSHZ?8^nEdtmG<7EY5V8A{8zY^h=1gHmIMQ
zyb`Bx3W=_DmkXi324$9l(>X_MJnw39=b~On#q^)dHG+EzxD{aJ472;Oj}PMb+sS1K
zo(>12oozFcHgR*<tuUXf$*d8;cQX|39}sgzNlO!Bh}PlQNG*Y+b1Xy2;l-K|j)o@a
z@U-(>6LeUVwvriDnrs{!7k}fb13NFp+=T3Tq6&v<7(jL&E1kTr&;B5EMm5=Cjf-0~
z=BA6Y@2t~5I?pu_RSL=fXxC}*S?dZzTNIVbsq7_A7h{Cu=R%$~!l<K6_t@qfLAak*
zxw}53>^aP1<e26OL`c|67hzt&0OpMz=KS`DER8gVMF1p`|9eXQ&+z2!JV*L{ukj-N
zAj8*F++rMI<2J$79ww0CzdM8(C__IW4F|p?TuF;_O!Xg*h@5Ok5v8)#aiajQ?M2<m
z9m-&H4TvjEZB223L_^<wD8D`shS14V?;QH9ypV_BN!y9ex!(sA8E=?`D3LsBqBNiN
zFKvj<B{zmKMeUe%*hh!vffaHg&uZlAKx~;Kh6vv(qWI$-y+*2!V{}`w&Lrav?ck=2
zgLEbdx{w?au=a(;{yOjqqy4I-5T*mkvCyM;($TmD=d*b-bjOR<2{$Rpk<kWSip7dk
znZ<`vE><2$ipj&7x2veK1|HryV^Vo5%LQGZ#qlQ^GZU$zdmPe#10zbFj0LLd0SyWm
z)6w8`vvyVBNEsiP=jMTK42P{zbr409JFtQRNjAVm+?78y?eM={`pRr=MrddYl-ln|
zJ-K6iWTi}5m4E9tC~4*CQM1&h>nlL&?Z)8wP3R1ZLp`Z;?7Rt{HVlx!`j&1GWh95k
z<UycKbq&@G<WI@3@vx~m;vUzgUFy4~c^jEu)HOvKRHMhVpH!={rPDVN=~sB8a8Zv?
z9bQXNa*U~&lhyK>c*{)sTGEdsPmLjoE3ZaXX`xL4ap#7KryYBioyyUsT^VJdVT>>p
zp?L89)V<a;dz!$a^Hb|~ck3d7*2?2o*>(AL>uNv@_X)b~JHIS_QOGm)3a^Q7@PnJR
zy&fenc&0CH{TLvr1$q|YUS3rbNF>kJnsk2}Coo__6E4Y~Y<*DxC19*_4)Ch6M>EqT
z6lq*oaet>m^qND@>f@tWy5G3MHW<iVH;*@6ccM)u+^qX+;*1auC21_Y{ZODsT<rO+
zV->i7w^lUIJIG;_D*^}L*0An^FH3$+RlSBuMb0rkPyFPU-~d5e7)(+i(1jL_ae)Vs
z8Ic7Z7O#vKae?A73JYXj3%9afs8y!qY%%G<3uVN8q?)5NJW_yN)rVL6hTET3P&XuO
zs}gO`b$xMxM<uxt3*-lB2t5OXu9_NSTT&g=pk{$*>vxqEmGp;GTE2ynhHQEkXhIc=
z6N3e|8F2w*L*OF^CP2oO6!Dgtru&D&<GYL;mqJ<ql7k&E1B!pK{uP}Q0!^9D!D7f?
z4RljA^5|?~EIKc_o{``X?-Xj=<HwB;5lKpn(^_*kCdHhO(C5*|rfxNTZ(YSTL{f4e
z(rz!AytOQO7{9M21SLs<uu(zGkfMf{VgKE#=Q=7VW_)aNF%&-F=rMXHYM{u5=LvUE
zbyaG=Wd;^;gFD`5tbP1(*KD)UD#$vrdnT25o7|c}wjvlvNVv8;r2<j*Zwf{4u`S`a
z=Q9>%`P>mNohh7E9^M8!S~2<CMAV^rt|%ZiUj;G^Q6Hu151x;EiWDc)Qu<}>%qoGj
z6=ArNNZZrX0JGY0ac@`OTr{l@QR|(Q0w{X_gONC^QrDoMB8MtXY(nArj*^%CE<797
zL@a>dSiol<EKrAGJ5#lNYd|?*390<p&|$2tX>#zsL=zY<!sUBT?tE8-*lLJ9W5;~Z
z7k^&fl_SBI_#2INC-RlifQ=l>u=kT6K&DcBZiC_yk&%#D3-FLeI<zNr@99$#x?~_-
z>TRzP8lN{M;ci9HxY{-=Dh5!^cRWhWp60?-fo|~fTeG`Xew%zRe`WRz*PaAEfgGha
z;woB`R=Lc&mJO7t=-}(*`by{$!^vX!3!#u9t;?WGG0py=gN9suh>C=3mhf!a+HBZ<
z1cDkganQ#xBVJq-n9>On{v$-8;cYEO>a{(ji%$E+*^W6b55Yf64PoU+9L!2LE+`@r
zBnh}smza{Rql&@s#Y-^CTObu(QRJaLK*YYvLdt_Z;i7`^wVwvG%>>vp1XmUDhSN<0
z_7zi-W}d6o#7`q2TTch@W2{h{1}ASIkrHjpRR(Jt@z&l0DNrVdhRVAIa2$$eAx4uN
zow^F5z9Q?u<Sa&Oe|e4GsC?`$lR=loMQ&xY44${xxUixN<gh9%6;J6R797pFs-cE<
z3{JVTDHYcyS6<d&7D4_Rr|4d>-a1JD2?Lu!8)HV-{~yNQDY&yQ{PK>ijykq&+qP}<
z8$0O^I<{@wwr$(CHJN#<-kO^FUHs3@Id$%zU2Cts*JnKz!a`86?&sl_RPYm)Gw8<o
zGq7?FaxO#5Q!+vdk6)8xtD42--xWwr1^aN51{?^dsqS#T{g~}G+JI48a^r~6J`~aU
zGnC*2h)Y?F5=jgI3@gueJi`p&tz9QezvSo+)$?z8apw`yXu%BjlYQ8f4mwqZA*?Z!
z)uTY+mfU@hR9VX2lJ3|a__g&I8Bjcqb%d97h`T^+bP0M2n~iuEbeYyBacupr7V|$Q
z!iFgPFj1`LYkcc3zo?TEti3YXrS^N!rXE3|+ZC~nC+^iKTUmTxW`KYqRg3@@Ks$pD
zZlAdU9kCRsjDgiuUCUQ3XHb>)Emqk2wLjIzKoe}ZgiPw!YNTic0awsx*ic2d2DpOl
zZNL@Q-a&eBb(a2l&nU$92WL=Y*ic+8-SDEz7||7ekex3z{IiwQL$#OY4y0eWXu7T?
zm0HejH3ha_2Gbt{uwY<%nRRVC{GEfi&ZzJA2t&V{3}7wC9Y0Nmmsah-T6J1hwz$&O
z^`#YmGq(pUcThGj2jjyw2OYOSS|7V={+8R3F`GQ74AVDcw%AhMVw|}I=Qa9gXbyaL
z3oneL?Pb<@t*zh4&wSZ(L8__Yiq6@Ibj?_m-%@XS++vv79@_Pud1cB-44$mE(5~Dz
zu;bpq3WU3$tOs)^#d!Qa1h6Cibdqb7<+1ljsYJ$Ii%n`GNmiY|ZFEG+;VNcSWSOZW
zdi~~c(Bx1=Wsbv^e?c3WQ+nUdyrJFzO``s$E0zo+x2NptcZ@4P5TaSQY#rB@!%Nhd
zQ`%(aR=25|J}JVbCUO3lf0X0PPm-fe#*Pvy>QCfNhlR|&>mBv=cDbia0Z^3Zd+(GO
zLYL4693swJZk5M7qO5~vEG%-itw(b|61%4uyJXoo^_ms8IOCOKa@f|mZ*siNd=vS6
z(I-SA&HvX{<tK0U|87;7*x4EWM_ZMj5Sl&7PsjJW`hsIbPZ&v@CkYDZRaR}lrZO2H
z*vh7|t2tdN4?3)gpRYusimDRto8qM`!l7NcsMMbH!LZ$?&#Ch>z|Fn${gc_?Se0Pk
z_W7-q`6J+{TkJk3e+oMvC*|tV`gzl<vU2~p2hj??C;Ip~K>zd>SY`+FC=~zxWoP!_
z)!>!!<BQ@u*&{UVP5x4O59rqCsNyv?b;|-NE8(?m7+<fub>8DWSby^1asPa?`+5zO
z0|3L~>i6zRj!%FA-s_5vieLHuT6}fCnGJ~$RIetub-SHRnzU7F&{^|w1-bb-g74cO
zPZ)f~Z+W(Ch`IUM`3_6(5ay`B#R!4!x?xnoU2gO*^>XA70afHJ0Nwm*Li4bXO+8<=
zAjy`g)?j6z!AT@BwANA+<Vco^>E6cZhyY>$k%otkU;Z`p1BW@X1JQ`4Ggf9w8FPrB
z)H>Q(NC`DD+$<S^-qf*Zl}S6|39H`rlh`s*=r^k9wE?9A$5YMDClZH-_Q)EctJ~Aj
z7cV5@o?i^>zbZE`OmPHYleeANBWz50Enc*$46-0YD0Sql8wLgIre=a+WeOz_wytbZ
zz4}>s`a!E>YE2XC!Br`)kUC@gPIfN)UoUH|l`dn#?U<VWN(GO1_(>_@k02f@Vhjl1
zkwnxgjLNgdd*s86_;p+vmqg?Jbs|OUOg(?Ah>uZUef6)qI;TlG^wD_1%H)mE01Bj0
z5NqvO2jichGws<|_!I4ZW>##ah2Xu^LfZ5%#dQR<aRl5n>kguK9$si5g`6#FB-`}@
zSL{K(><TOz%Aa@Qd<jn1zMFb8)=@_c+A}uU6JDmgiz-btf#D>*jtn+WFP7SdiZ~fT
zgkC0dB6rb0QHQ3Ab<bFSn=yoy>P)+*$Md7pUXE64tyImmEi7*2pdtUJM;IKMQP#0h
zt4~v}a0|2<?{JNhntcKkIF&90ky9)3--)3Lup6(U$i<0ra;$`}S)a`#M0CoIEVT;C
zw&^$=2JtAyK&ExHAJmE}ZuI2NG^X)Iy{<_@d5BncdCvP4C%Fu|b-(bFY<@!^rh;MD
zS6&(<h<8{d?w|>ksMoFoOp1wqYmLtov6tUXJzqCHg|Cge)b4SH<3$bukZcQ>FjBP1
zc8S>py{z=Z0fy^)A4Am=eu{J!f0-(k{aUZ-A5WB)`swK>{ld|J;}jj$kZN{tkM&v!
zEK;)KCje&xHh4V<rpB$IhgCAq<<sgvE0>6Z+kSRJpbBxe-jFTmZw8|+R`c?7EsAL|
zm5kqP@rF(CcpMV3iP9`?uT3CcZP}Frh1+{ganut7#F!|8a$HfI%5lqxoW^kh#(Nwn
zEZB-TbJ76gupqW?zYD~)A=&R#oV27s#|e-fj<UpnDT+*Rg-9821-8Cc1bnZVa2)cZ
zf+`ZP54_R~Gf{gSYsVp5m+t9kL^<go*}S3X<Z^m|F1}EER<D&?vf}|0g5l6@|02%o
zg!lIgrg%4nV7X#0ScfNDxNov6sVSis_2tsW)J)ZOqn8Qo`M>#1d268CxLk_f3(ryI
zBzeJ%2uGI%2i|?uI=K_`!qt$`Qo0&QYt%Ll?!+w7{_qxnL|q0Ds}8n2hB;(a!-ot6
zw=?R2aa&`QRaLPI82-D<gg9m`XPrS(`GqcZ9RH!tq6Ae+zFr$9WaT3F+4L-G>$ux_
zPHtMi20-6Va>No^{}O&UaeO=uNR?Uivg$^Z_6Gt;W6sF@nReI$+H*@bVIzaDUagvt
z!GCOmiM`U6qwK!frEE^Eps*Cd?iN)fM97%;TRPNR6*YCT<74T5<jZ1>w}O<G&!JA}
z@<J0#TMBKuuZUE<dB-kPHXld`)jnEQdX6sj8A0dX=x4o&7)Lk1Sofc|_;-KGvFHPm
z$1I>bJiUuDrod9cladW$IbJ@^;Ciwho#EyP?{VR-9L2Di--#xk8yQ_ByC%2uvRwWh
zsCinF5u^*@M;XrSld+mNxD7P|iAZ%)Ym)Cxy}~D^&9CR94)g?Kl^d`PRAaAh1@wMD
zy_h^1_=--I*dyxiQ=j_HYY;uM9DMHIkL#%o`ebd;B26JD3IGAPWzqlikje>MZ2{o_
zv5s1fY{Ej|SeJpce!F<}LX~aqSq|`u9eTKmHb%C9a@Rr^k-auVl4BM3gvS~6iFs{Y
zxTU9L?W~>T6umlPe}U6!1<nB!ET-$=vi_=Xu}X0PUl3CyY31jllyHhGvoi+)v$9i<
zuoQRY62%>Swm$p*#%D(ydd;&dk=M@787E|0s<CxJ(r{y|NZSl6BI1TrQc83|Q_%0f
z(y0*Z!u6MCthqlx+H-L92QV-Av|=l|V@==L!db&_JgxeYhjbthu*UMU&J<0FIsN(R
z>cd!uI53^S?rDKPw#8^p{Rn2TtDl3$Xx?0YWT-m4AhLS*^WWDpNAdN*E~wspI+DmX
zKxN#ibk%p_2IXA~5n?~z1PZd_yb{mVj%o-MYDVthG0>>8=CIk6V_A}T5^YA64St$J
zjU+w1oM>?cc|d<f(e5oCl^L9IdJ$xYmUY~GfmVWFbl~ZKadB<vZy?z^pUDUsUB<A3
zg=N>N{u8uMvFQAY<C$}XuE3~uINQ+P0Y-Jy7WN?H>6K&jo1+<qc9^v!WgM-}wdw0>
z?~bcNqpvmV(rl1l@FLr~XGtvT{bFSVudnj5<;Zgi+@<wMdZ&6>YWr%wJarw()VhqC
zZkVu*?R-AUeNvNiIf-_-QB8f@)zVzwF>Y@;{akOL)_$}BZd2-;3^K^`>0mDvvl1Z#
za*+MvPmN)%qwXvi<!r!^h(_4z9y>=U7Ms(+Mrj;(?Q=b~O>J|A=eo5Z1GXG5COIR0
zrmY`$vB>(p>tW`dj?Z#@F}nuZX_eiXgJ4xfRI|+)#&9ai$~wDCM7+KCE?&g@|CtS$
zzVfYtQsaN)>iRzL#H{#Ip#6vXC<$hw84<4^<Dl8Y2(Q1PPU{ysQk+GTQiObc#sUrd
zkjT1Sa=$p6;Q^gJaugti>^rWhBZV|m4=`r1v)dK$b(=poJRJ==LH-Q~_C7l-OYIA4
zh2Uw8WrS8#k;^J!Yn9~+@7T72&xt;~7;IJheGAC^`cvKpMwx8Jch_GwAWYY+AQo8m
z-PFfGj;e8FMw_Yv+^56jrNz_*!8P}yye;F+*Un9>Z;Vc^h5SXLNt;}uV|k~|4&%7i
zUZxUMr9Mlsz#`RZI3KQo2#=k96l|!+;LuZ%w2umkjVfl$8O)))JQ2}o(7NADUv?Br
zUu1fpl|2lIwy=<Y%{UC4Yml;~P)23N?ai3T$!;nS^s`enrHsmsCf1BQY^2~Wgp-D^
zdsNC>JFfBNl1`ZkmIBqTqID%oH6flvshWhy1Co(eW&aji)&rbZ{AM@dtn>xX!R<ze
zm2X_m&n@CV*^-}I-7<GaXDbS@)XN(ODQ8Hm019QJd0`L3L`rC$!HGfN^c@IC3FZMp
zKX$PorZ>5m6KayEA&`*-`CE#{xfqupnwq4;PN|Ok7r8q-3{$3bPUa+g@6SHW`7oYa
z>Le)wF>|I;ZxtENDZ>S}*ZjoWuxR{!g0CklttnZKHe179wWh-Al6mEMsU`wA=;YBp
z%hU>DXUW_naQ{YcJh{VIEq=4<#$;L0-5BSRA22t+mePOeV>ZVB(8sL*Gb{gpLmD<(
z{y#{AF6BJfiyWCU>->|=;t}d_h|1*^CH6ZqeqE#X>pL)|5)omeR&|uqjB~VzFOVP*
z*lVfjb7B1EGa`L9-c~2oCg(0Z#|59yl|P=u8*H4N);uSWA)0e`mn%P?RA^Gw@3w{n
zsoBGm_G0<_2ep7=^c$JqJe@9B0IwitV4v=RZ|U}h^#}P|PQsF#lvq;T+CtD$qN353
ztJ7n7ex8wvF4W$k>f68H&%5*e-S{sC5zIo;%b)w_!rnP(P=~R2)`o>Y7k|v3Uw5l<
znwk<J>k=E5QijwBM*Z9B*`UzyADf5@3v*{J>te~UPnY$)lyb4ZKi&$B-S~X@sj6r!
zp>QxwjYD!9m--;LG-MZYu<2)eJUlO48#|D9$l@&4%op$EEPjZ78ScVksU^#W?CNjc
z+vk>hk<vmL7;Emi7Or8rwpYJm8NU5a#nY~YziFb@CN%fsXbo`T(NrLX%wSpEtYA@P
zopI8!$;hI}k;<+f<+WzqY{S^p$>^Zcc%ocYx0j2X)rM}Q^0{(&Ta#VCXfCxI*4N5U
zg7BNX^m7spc#|R1NJD7<iGfv;{kwJ}o=X7-V;5wpM;DP2SnC=1vyK7*GSuLbWr>s_
zz+j#RA!vcUlm-#$z~M-DH#W-yMwTBU9YDtI>l0$q?FA2@bIQLHdLLyOT(AI0Nm){o
zClgZG<exXX0B532$<9QB(P;JySb(5A`5!fu+9oRXKjyH8DU}@mqyD3I1a_-*wFeyv
z)xgk%r-lV_v|&qqf6mj8FkcNf32luEs>@i1*opC?5M3r0s955Q_dBU!C2;o2QZv#u
z^zt5SvpqMqQ<%(}#Vx60KEAW$$qI5V+BXvvf>dYcNOn=RqKz(nK0HnhV;^HD(M5l7
z6#QULHcR8;GTbhLzJiQ*OF8_>?+;7mI^?f<Ui?1qr&-3xc-d5H-6+m>6&1U^)vc|(
zDovl}YC2O@Yg0j*tHZ@QXAPk>J#Mux!&8I1$$&wMA0$F5@!4xE1+Btv!DCdc1!{wE
z8=)IYUO>x=qX0KQlCY+}oDO4U**$49GzCqfMBzb~Vh4H2A;62MceqXt5g9;;<%N;0
zs6Qq(kJ`&<02tL8^Y+FCB_ExfW+X16*?Ui{M{K~IT4_g^f0tsSsp{7{r+P1(l@;sM
zr4Q*Ns9L06rl@e~jAK2uv$KYKfL=WQ*upMd$g{J#j%*MDPN;Mf!~pFfcg*UV7ML3n
zR=O05;s2s(*FV`sT;NDz{g#yNMldpYuD-H&wD$vDnJ2;uh`edIp#vI^g0Wq*d9EQ2
zR4j-`SSM7o60<wV)Cw;2hY8-N%hDfM$c5T~wSXydz5#iS?}HPXP<&-Lio0$<`<@d&
zexfQQaIdf))Kp01x6X;lUWdcDtC!xkyBVD|h&+Z@u3a~4J^~3Wj^W%R)h1Mf2tNDk
z@#AGe$=Dft9|u8Cm9-KJ;qH~LX>@iUC46_NY!sHZyS$X(dYn5vVU!DgDjkZj@I?un
zSU`y)WZdl?Tn*<6WWfl@mdZE@3I}9Zr>hc=t&3%%xN!>KRw!lTGZHB^mZT6_f=iPZ
zT`^WFzFw#_7frauhZ(-uH>jPW0qE}&ZuZtPo^P-9Xk8mJG)atEtOg%stSFf?K>VJQ
z*?oR%u^zPz6g6b9>waal&ty!w%&PfYdwJ`j3k@P*ShRnL5PQ~hV3m1V%&qBlf1XZi
z*F3dv&lkU6U623aRiZs7|6}Okzcrr_()1f$(6dSU288zfuGTsCPofp#VLD3{$)KUL
znG~JeIn5!WC^PFiWC@Z?aK+n)(m;&+7;jF^C30Y)Or9~c^^1X7Wzo#sOD*mM{_kTc
z%7Tx9jJ;-2YHF=6oa8%{MDOuGCI1A*%^sdNz>mN&)6<r8@rLt{I%?ajBHZEZ9#ywH
ziRLNG@OhIlm6c^w*=WWe64Wv_G3coVL~Ab_aPX*RWTy{0v{Ca+WqR$lzV4M+24tUD
z-geChta6A@HB@{&n3dPI$l_yctlnU%3FAkW;Ym~TP2!Tub~&bq<1y}-QU|cMV`Yru
z>|!K!sy|KQ@-i#czP|ls@<xBH1;w9W1iQL1g^n=;KI#+fd(=*@qEGGqFrZH<b&0lf
z{~9{T6bN)2*cZVN*CFd`0e3?&JWegjFv_w>^G<7>V@ZWFt}G{R*Nv*?<qzkRpkN%Z
z7ong$Jlh|OpIX5@qmvqC|K5z{858f2Ii|gceJo`%gMvKC7qVa+uojQsf@&jk@~~j<
zie?PE=U4UOQ%DR_!3UUgieU}Xs<`lZB?&i|xRFMtR3sMsI(PAB*}IF57sGK(W!bBI
zju#sTxW$O!IHa-|RYq3{AT6^xH5oUTSi_8bMoo$_xSa6@NyDa14ezJ6>eC^t)Kp)%
zmy7|HwqnP_>G9l8*&IC5t+#dT7$ns%8lY2~@UUpaldg|(r#ya*>{T>Yp_hK6;ja!e
zu|v@eF{!^JIGgxFnilHyk{bnL`;|mCRd1pqtE|Z!&Ih$;isYi`RX%jxmc&vuGn~30
zbW3HC&}2zg6QMEXEWSQaDa9eESmO~)u?bf!$Y}rajC=F;M2jr&Y@%#Ju7pk%?qtwY
zJ$%2tLu=N`dKJnA__`o8+>{3OsG=R8-8~<NwLG4`E`Ex&vtp)MLAZ^`HVIWkz~O?f
zWDix-)#pfj<u7Qw=_5}r<2&i+vQi%B?%_h}?S*lO1Yo&W_X!$dY&B|nGdDepz%}g>
z%oo)g=l~{n$1_*T{krWf_QxeJEOqRil_3p{5<5+<FWN&XQO%>Esxf)bc+79g?r0B%
zbkg?`tyxae;OvXctt2*`W(F$&IsS~ivhpTx5ND$1fzmz;n?$NiLzGk}w1U&07(z&6
ztdN(ZGu8W?|1f>UP{q@;-qJwiu?(wKi&~`x(dgCVw;>=aY&vgW-O06Uypy2iSf_C|
zeSP?q9x`ID%e%U2zm!m^FH2IIH|*H^G%#sF3i+1GeG+rv_luivIPk|>rP3!%yftmS
zb@y5$0r0_{wg&j+x}x~xtC8o;sLoc&Z>E1)#iKnyj9dx(w<*O%U{aaPxqED@1FF5Y
z|DNeUa*#uXgO3Z^)b;FG4a!v>${ZFIYRJc9pZoLl9<{azZc;a@e5wjLB-GqIJ?5CY
zW;1N0PLxuLccA5YU_LB$S|7bG-_w~3dsY@~DXN9cug**Lm`a<PXLmp=zaQ2dGR8l(
z$+5h2HLGgTFU+_?9iG+P&TS6^{NM^EN$MAca*tGbFY#%DUKGPHDS9as+;UPd@0bQB
zex}vPEY410;%8b>otUH#JGP^YAXrq2(&GQg{S0durojBfN_~U<?3EnTUlpPDdSB|X
znV)ZF5Av7^oLi33iC@W*W_G=sB=;Ugloi9FMWYwPQu7B+CQEMz`r1qn@qV_W#V&q5
z5U3Czp&Rb7FOemVWgL;v<oC3+C(r{g>|ckbd<}LXjRD;B{0TO|Q6G267Th5{K-Hgj
z$cYS7H>G^eb&PAikO6r5cMD7om(`yqbFkLrd@d3EZVLw{+0}&CV;ReT)aH&{l0h>(
zh+jDjw|Z#n4`t^+RbKxL;BILBZIX#wmc<VCwyPfEx1@g27kIM5_%A`s&h{UI_$R&j
zf1aNP4jX4v0KJ%vp|h!osj<C@DIXt<lQY27&=$shtx=0Ro}?Lm>RH{{9UDjXMt0e=
z%*@kXw-8`-3Zg0*b`g)FZ$SF>4j~x>wz|n&j|j;dZHEj5jPwH3r1RA{a#)R+T#e}G
zIcs)Hq;X^Q`t}a;BCtm;(Q$gQ2OKzG<New7aYs@JaM!(IQnHFn9^MA-zC5iD_9xzn
z_?^M$1<ImF-2voRed>psY`%V{lF1B?oBgkDfI9Mlz$MQz)Q;ox=`jRSa!>k~?ptct
zk)M7S-^a0`UIa5iPw!pyHR5T&EP=B^I6{kW{l~5C>*ww@^WxCBM+c-QFyY~yI2Oak
zYwQ9h&?04Ky{iKG$V7iDg|nx~*+3$eN~|cUhDxk#w_+jsZySYQW`*y*8t(a4oepZ#
zDx_FY&?aRNrI;)+9=+J~#nW<R^%_bA3=$<i2cqz9z84)aN&YA7)G?=N6L&CWIdNI!
zc{%ZjwilaZl++z-6v|Z|4`PRGP?0<qDr${%R#Rt*51%@yq?2gicn%DD7?;=~5<^~+
zVPs{=eTr49aMh}O8j;pe%*CL*ondvu+&gAtcOKij-bjk89?p7}bBnDNDxJ>DlWoj{
zg=m(fk>k&{E|1IY`QWu^V4%}g{TkF28yID2Y}iM2<}J(NXpJ|s8`SD>LTN3Ed}Dd^
zlbd?|`hR76&ByWSAI->TVfKr==HJ&yh=YC-iXM;)L6@RhJ@y9M?Xft}u-Fqgqi9z6
zIbwHGx%8YSuE2iRB#;ase&li8A7sgXd;Rzh4r{L8lfvEXY=GD2J7p;}cBp(8W<wpq
z0qelSKZXn<MC+0oxZ!_0k_LR@Rx#p#E_=&nT*^FI6-i^GOwu(fkU3;S2qXM`hgXw2
zr7m=neV$hxwqjeF`*fFrt|(x7bQWMiEGS1V+om=)81FcW*46yGtjgS>soy-YBO>zh
zhydxq8v6buLUdjo&%1_le>6^K+>gva;a>b)LzagRTqYV82<Md4rH~?rF%Kq`h{kkC
zUd1<>O1AlEYr~WJq5|USyQgtwUqikAjgH;~h0ojHcUxd9ii-LD_GX{NMXXm(?(&1|
z&?nHzqZZ9Wv0x1nYin22J8PX7ch^kTo%K+KzPqN6OeWHcG)qo$1?F2KOkHi;L7s|K
zN(8HY4<v^<1=C=)rV34>70zJkHz*U_D5L+V3SBWRwP=ilQSh89^<;m>ND%HFGD--p
z#}_k9h_hRsf|5#Pwq0r!OPM$Sh?7KJ?lsnufMCsS%W!Ex{gVW4?$75(xL@tkY8y=U
z7v*2s3;%pFoJ?kVKKa{kwlYb_X72nXM#giqf-{KoowTktX}9r?qP&l@2lKds$#SG>
z&URBeh8E|dt6<iyWq7nSUV@TLwH1TX?PN;KRi2+b+Z~2AqraRAb(2?gN3VgWI<`mo
z)O0~AOug)4HhwcbBDqWYdi<Ya4{F&arUQD;eESbkNF8>N*2y+VTU?LySd9t~4D-KT
zB~8|@><sQSw|xOs>{%CyH9#R*I2*iKICW;hoF%EA+0JREzNY~$GNh5}SeWY@4pd28
zTh-|vRcne*X@QcX%92{b#X0ShZuye>+Bo7`saY$l`7!vK;bA)$9y)8!p53!z7HH<w
zc%rl6YoU41mAZEM6FNI;(LxkP!%<RUua@FRBJFtZd5V2L3M!)EK$mJi=TkP$J3Ag1
za6Fynh+O~RSI;J*t-Uds;1)6#;%<^rTBt@ji5-M1c0H<-+Bd`J*>>h-!H=c;x@N=U
zOT`*b^Pl#Dxf_4~K5W#>P~^~H$HqS-H1gEAwg~ytFJl{MR*<)Dw4I=7MDtXcI>;yi
z8TmXKdDtrKK^|srx7p!mtDy=7^c`TGPGL3ay~UpKKv)ME-NWu%a6v3;>)|+0^b|6B
z{foh`;2}dVDub^1fnP??96YZhI>Ei5#0Swaan6pRO?AO4(KFgt8Adl{A)v0>($t7{
zM#=@r+O@>ZldcW+)cPYB*Zbk?526w^-or$%o@OC1Z)M%as)_wLQrw340g|JS;gOyb
zts_&0d6C6PRaIVbmWQPY#j72DZJ}b>T>x81ek9A&oZ8Sj@FMrPV2^pe_U8gN1bA%-
z+c(b#u}(rYeMX3?PBJTiJ&=n~71V5DD)%Ee@kH%14-bu2tlW2G?}u48Y79BRWvst?
zK0srb=vm2(jqq^jW(Uw1Un47AogGrQO>YZflNe#iguDZq<Hy!_R4H#_{^Ge0jxP?D
ziND&QjkgXUp2EJ2u(r`rpR%-<zx`WIqR-{txXRAnQa{tuHSCaZ?(KGo)1Xi%jy%Fh
zF)v*iTjb_NWih{8q_ViU;z{Mp`Qe1)XBNH7y;!FwEj|~<z8!IyrQ&8$>BRp1P!S&=
zTr4>pyokA@>7zRx%Er2*sBgUPV2t*@*(kR>_1P$w*fEHX4oz<pSwSh))0M4S7*ASH
zVJsS9hJ8wj;?Z3|=5du>izg&BECpZBKRV1bSo%phqHyLomW^NSblbYjAfx!ig&pbi
zFmPng0)BCYYFwkRSWl(q*ETkiRJGT0ay)GPsvSho5E0(DC!wW>0U9*EG|LmCLKIN&
z6d*A4Qic<BqV4MRiWL3p<$)UOGgzvSAtmXxAiEk{P8>`Uo9*UB)8qS;ez{eXJmat7
zRkQe28&+d^4|I$j>JV<qa+p7ZdCDF2Ugf#803*wQW0H3VWEOB5X9>h3Vowh%u0LC!
z!LMOTR&qWAyM`OJ$q6qwz7yIjv&l)uL4+g<`32ba_dYJTb31Q3@gEDhN(dE8mWk$b
z{m~X}wsd!Bj(H^*qFb}QI{P^SKH0Smq=GxnKZ$9$ShQ}_GAH__J{iS|Y8-m_)(H;*
zqS$}4W88Rf4i?FUEEN=XtLK!9N*QRaTRN2Evo+#!HI+nP)5-Y|ez+=;PyaQbi%(F@
z`7r75c#vAx@T9G?dc7)e=!S(lcsa%bK{XPs)vX=(AP0(bg!q8L0Q7x;tL?b<Tm4!4
zeBqBVZ~GqGm6(sHlRMgKAw(C#IT@y|AP&JZE~U$b%{4McpRqN&=N8_1gyI6-+Qy}^
zUG3aACMIrIdc2hs=zX4(b~Sz_>aO8@3$Ae=**_{}v!kMa74>)N%F_=f7_(d0`LVCM
zkkaWScg49~zZ9z#{5{^UPw_e`t<ROpedikn;ScvJvnrPlZAU+1+0Pt~yPnCNnittd
z7U8Ul-O}xt&vD^&FpokOMrRy>G(C>r|0JFTc%EmfM}G=6@^sdNPai5ZaL)|yT6&ho
z-$C)C*#qw|bre~84H;hJcUh*Y1}G6L-3V(l)__85_o3dJXK8OW)wVM&7n60)rwbL@
z?OT%1bi?zys8*^*Js}(|h?E@e>UkMQ`z2lX9hO7p=6~GjaE>3pRa97`&}|w0O_^#1
zsm@)OCN|9XJTsaBOzC(8Tj_f)oiuP?IG>CFSxQ!4zs{dtNAfBzz_I?t?ed=&(|qa+
zTaC|t1HWb!05bHl-dy4ppkYQ?$L6(PhjnF;)70{EJ!p?@7r1{LnWzVdD|TpUcv&Be
ztfvp02k3-#THJhT{Cz2N1H0C$^WHN1aK3J%a%DTTgGuXR?EOSsT08wu@Y#RVr!cW{
zu>Fs8ZYTD5+(FYNpMEaBMjFXW1;AF-?b^A_TbhYISza<D{MkVN?#J`x6H!o+KHK87
z&=M8|g4&Y}lmE_qd1!7vA6}PtQ;0W}gQYHx-X;$~<oH6vfsD+mK$`GR;p>v`{+eiI
zao4?7$^hLlz1`Za{(1us6ozu_mG=HLN|d-ExJ7pNgZ(ydd1=bZ9+Msb`Z^9w-A5y+
z+(B|U3taa5yxdP_&ssU2NO!>YJ@LE7xA*Is9hEr8lyiPter~=I2r$?=2t-Z2vwJ(|
zpTD;EE4F}TgykTc%kBmdV7`kcktbKk+Q6v9_JD~JQj4FB0w$8X1ItJR=yzk#-SS7q
zEs#Zny;>$zx6v7*@-0&^M<Ors@kl-c-z8<Eoe-D??j-Z{bsx4fROJY4A>+k61Qk<Y
zn!?im(|kq0=Th_4G(uQupt(NXwUOB~>h0yJ|8p8gkf-vGjz=%cjyAy*#zrrkf!L!u
zZUlM8b4dy<OLwj%Yq;Nu@W><3C5)+U7*i}ZXxHtdbMj=CA<GO4PI<|+sCDXh@5C)@
zQ9pi1cuM-Ofsgskds)>)tjfXPw=V@bmTe-tNedxjIXoz3{W!mQN&r0+%Ymy>ar?{?
z<I^fB3STJ4s}VFL6s^Wd*iyRQu)kxoZZ!>*rZv)I;ZG$EQ&H2DLSYKZ0p^FJad33>
zRNB5jwb7fTs+REeOG;rM=-YI4$tR$*!`QhgbYJ3c{YI!I{{9(AA|1PpRV_wS$}9b1
zu11x!z8$22p|7e2g9$~<T2_Y@j-o}1ka4u91Cq>jsOm}FkimKt6;YH3_j4THu{32I
z7bjn;u1}Hgokia>TFLu!({h%RywVwws?__3nNvRc04S;QT`sR&r$s%c{G5tr-bxeH
z`JyDnooQuep;o7R`Ih*C^h}B4R?**pQbv|9zwKkT9>znit-odBi}G`G!@uRNrYoKh
z3zTwxE>RnA!r+s4#5crQQa;bgRmBVZg)aI^XtfUUs8^;wvIDS@uuZt@Pc5aBO}<O`
znyU$IqJ#=%l@B(6gU8{T1CSXLNT;62pw^i31B`e|Vf!%1WQMl|euD{78d9)u&GtK2
z8sjm}Xu@R_!D^P|&Ghz#f8pj_vfh^^?;IYU<|3%sNwrL@w$+7Dx}*R<O10w0xFb)j
zEs!uTD-i-Oa$>eG1*(uoMomv<pj$|M-6=8sRF1`Kih(lp%g9LgXHdUl{kvDJ8+;&;
zYU(LGdt;mKo-$`heGjEbHTvQEl?bx26`z8=Vd~S&kma}y=5po&m|~JRw#s$<qsDJ&
z83$`uy_pLm(5S#ot)Q}DpSni$D8K0&UH`3A%@+RVI*eM5px=h>n&1?-V}OPKJ-NR2
z(we<2_s3G3n{st}d{=7>(Fz=-(t>a3Zy3(8sOP{Lm!AL20w{)4lK5}n>Rz};D0ERh
zjP}6&c;~cTQA)VJ{D+<jbzJ^BDp6<Y<<`-a0^4^FkeG<}s4$IXRfQ9)nS7l1UyTfh
z!h70OeJca8l8HL9`_UR15D;666U5=ne?06cS)9E|PSX~qSc_#N6g&-#66iU-I*`9S
zuv`ATA`(>sdTMekh4A2yn?<WB8<q2;_2b2s=#7urGAT-Bnd^|lP?N|vmPrtm_HwP3
z#S7&Wt%$3tH!@a`ZG?TVIoks9AUd+IGJbiI_~Z9cm{?RR2ZH{N(Xs;LN_+ooi1yhv
z5nzM{T-(Hn>!n8SO&SR@wV9IMcsCZ!Trje*uevaRbTY?B8Kx0=0DG;vDNg9db(>oT
zUzm?umh#lTwAvJ4Kn`Z9o)Cit$+hyM)rsjFT_&xEJr^p+5BbPi8%v#Afucn6t5be-
z*}#)5a&Mi5j)Lr}6zKknYo3Wh^f*uCLcrq<_R>h_bKMaZ9ZM(>27VdJ9Y5$%Hn6f%
z<99vNjMa+Z&Unky;ypxUtyfZP1DV4L?o@Z%VyEiwxHf0KiPYL>a_jIG`T;sW6SO_z
z6j1o1k(Pl4Mezb_#zU@xZ<F!z#_}1E1Qu4!e$Ja&pQ=5{v+K?x9~;}gM}*P5hme~0
zUolqsJJQlr4z*pTgL?Lof^OZ$QBnA=(FwRi#7X$g`~rrS(Uw)Sia7(&w5e*7*^*4o
zUfUP9CV;r(%F7<ixcEZG8m!sDA?+}G4fz;E^wwAnyH6oVBWpfG0zk!RY_hWHmZ6>$
zjmaK|pq-~+q~;}Q!*?H7H4MgR-X4$DYX%<-z$sVN{za#-Aq)TY0H3ENBV@<OULQJ$
zl(&?cWMw#1wP}(mjkwdRn}RQ$q{3uKo-e}-p_IMekCTJlBEo^@)E0z=zs@E;X+<I(
z5C_NwQ6!>5zbg*0Uc5W8x=965nCA~zZUIPVs5!qbox7vxiaD}Ti?Z-Cu{&(ysX}bq
zl!n7LaSAAtj$zTD*p};`*NoVkDN9qPf43vSVKZJTOF>=5yG5B^Iajpj^wc`J;D2?*
z+cWb1lc9B(iQ99%Fp8#d5lUVUAl9eiWv;ur%;X0aX5dYcfbHOo{jnDUrAC5V*EZIR
zyf2Y*9CES7V&Tw>_wEdr-@Wz806R3MXWq>sLJ@(7Q8oEqllO*#fQ35PKH9n7hZa+{
zfB3xJNia-ZE1h8TBBid~aqRCI*ZY#v0i2oF6pJnQw9pSp=E@ZKnW}WK?V7ZwkuS~~
zN`FtbIy3i4hO4ts<{KQ?Nz{$T2}f6zKS0ieIllQ4GBBhOW==By{+2}OFc3@4J+Kqz
zLrqI&r1tKpu%zhh#gXc|!9eMipuO^#nh!{g$zApAGhDgSd9M=Osn+<Y^5Xz55zw1Q
zBqtdtl-QOvflJH^Kfu|l&RK|~+yxoa_uHTEzTKsV!l4TI7M-~cg3JC&`NHIzaQ9!y
z4Mbuz!_~mE-#J7%`oN$yuR!|?G0k9xH&hc(?;xX98;SmV!8*+CzCLcoPh3_x#mMrz
z?XT=0wT^YF>eKqYTp{H!Msu`9rC(R9d|`rcbS$^JeCI*1O6@JxmYDQ-TV~Qtej+UR
zj1!d$>YyC4P=h!2<;cI1#;X@}86(9(j~_hwIB%p!?y-G)(r}W%%A&Efr6i(nFDVTJ
zA)Jdy9|NM<J7^UyzBzO_<LbC6ERl^VW!584;O;Bc+WFZj)f-J$wRCO$)Dm$nEV#^=
z7BXv|%fgbby_Y30Lvwn!L|6xj<bz6Nq$YH(CyzcHpu9CkrZtAb@$29Yb`Z#S<xYAk
z!k<rPyq+|}L<lR)O{0$x)dJ?TjJeEQOM$YL9EaBli#D?}2V--mV8FZTk5G0>Xxs$V
zt^ZYM<hAYWj2>Qqny@rSTsr0i<Dj%MGtC-fLKh#ilScQcdESX2*n3OgzrIgIsqqR!
zgY?P|EEFtzEuzUUn`4%3#DQMp>8nsxt}w%FC9DUW57JbG&rklS@S@lM@>4nfuSVnl
zzqOcy`G2G{C$+}o_L~qkpVcim>n(x_dX^^>%Ga*0R7{g4uaz_l0P&g!Q<YL|ub)7u
zrI3#`q}FjXCaX-)e}JH}K&9BTu{OQ$9ellB&d^G4sWfhiJl-3Dzl86}EqqUG_JI?G
zEw=dB`Mqb>IJ&01W0(GUdwk)&*vUsj1!(u6=sS2mfVJ{sb@_MeAGld2d#*p}-zSh5
ze8iAicNG*uMiR}NwaORcZ+H33Xnty`A4}GXWmawcb(g_@o#N7>FxFm5o4QlUa`zS5
zSh}Ma$k$HrTaQ(7W=h%UD)sH`;QioJ?0qZI={Y){M!bV0q<|}e7HeJkxRt&??{{mQ
z$)yLWp>J?nit+|d!xf3F&*3ph#qX*S)zF_S7B3!LvpMh{MIn|KdbYjOKeJc%OSR@t
zQ_V5nm!O``KbpNbKm8R!cKVVga`JR)qmPmp?Yffui3QMy*-?8FQUKOH9q02CfwY9F
z{g%4-8HC?L9G?7ZSl!px!p4P02={&@X~h31xxK{9rg3{T!*~^Dd=mC^IGgh7-rD5A
znTBR~41yEn%>-&~;c+`ly)CSVn*5~Mt}@rxa<id)ULcJyds;F@&GK`6KV28|`lUrf
z=y@WYT6%g!Lu-p*aXRx9<7a~~{{`JLOP}qf<4UIqies-rueoEfL($<jsfB_nkj(E`
z!1Y$a<XC*QA1yEYc(6~foU3S{L%V$PIx%9sl3S@At&?3(9J=l@<$wId-YRI3(uDuY
zCQP>6WwW$h(-`VxSomv4;@cKEKa5X?B={4Vkpjk-=+?4ZWX~cDrP@&_{XT7%$DT~?
zFYZuNMzAf<919&qSCn(M+nASMfr#9Yw=BtAN3F2Ryi42g3L}N=ZBL2dWkREP8iYsn
zzKTnM$ltLy?a0VrCptU>8<Aq9{zULbNupSogK8O*c#&|JN^YgRza6HPt=0CIEsG=s
zmX6gIZ+?4QpFXb3BtvVl_wJ3>02T&gigvTunQ7L7T@b(QUznA779t0T{AWkqAIHrb
z7{(S~GP3+`{U=9$lh1Xu`oAWqnV{KVSSz8IA{7}2XqtW-G8Tz;Y7RuEfoCDEy3|IH
z<W(M42_dfX#RQ?)e{u8~;hJI?)?p(pQCPiEMjGWaFI{Qs)+QlZ3U2NXR}$I5>7v7C
zl)44lb&A>T9EhlSfPRxqiw0Aid?fF-4rqeqZx1+GvQLvT$Y_dnOheWSDM&I0KHfed
zU^Ld5`FBZqwM;ZW!%5=UVE8LyPCq6gR0hp@9zwDIQhN`ArpAhID(_CNQOmy~u3$EK
zfbr=JsYx4&f$jTER6!<n8jL_*k+?w}*l6`hPf(F428pz!MYJiaVI8__*Vo2(9#*py
z`HZpzDjN@d9&j}t_$UyXTC;>qZyopuWNhB&sk4CE=b2#%W5}yuL{`x#(k`ZP&F)HA
z0Vzj<vh-N8Dy*?ATOU%hl_UmZSgD0CtU;R387_{j){g<w!JSou6(S?p$17r9{|}pW
zA8D4%7qicbruvtIQ()t;F)pT0<?Sbip)Fy>ySdAxme!`a+X~h0P*34VDJVjIIU9@M
z>Iv=?uoq~Cqp%;%I{=1m{$Jp>(2zrnz&o+hpIbYi2VF5|!uVzB^|S+v7-ZIem&$o}
zA`a!ZfQbyt9<Epa5gDp2fzo-XPzw^9)EYB!1+ajjtKUcDn%b#?;!+uw7+_rc76r0k
zTm1bJ>2CZ#Q9TAD6PY8%h-EovDvUjDrpF(jo7=scEzI!CiYQY2_3e|G<N1moaDQ3(
z{^;UG9afWUj>;I!X%S&8%Ga&W+|{ENnmYDU*;|*(GbtB<aRVsAklo#H^If8XG^)w@
zGxhfAr^~<9$)Z%5eULiAS_-e^Ihnz%9D_RLtKoJ^&4~Qa{}#boP8cAQO{B{FX{%~$
zf`B$EGtpL;cfj7$qN1P{{2ca1l^06h;8C{ZIzPXOLL7{NnlVPfmwq1C=r9;#Sd(<{
z=h%Zc1&d?*UdiUsLfu?@1=$16N_5;jRo;WBI`M>Jdu|<8o*Ww>KU;JilpJsQj1eQA
z@|^SV1brkO-0D0um$AZ<&qdop$<ftbA{!oR_Kn$RKS`OzK44`;rhYUxcqGFL=G6Dl
zG{{+v!27ROkjciSD7o+;En9j)7w}1sTECIzAlsnYXZ!AUs&MS5<kwR_>O=pTHo{wj
zn9=|UZj{`V9wVM66r*(17BfhMAM^ujvH?xLQ&XJP>mv@bPTMYpH>5cC;QJV1R=5cF
z%}<N(=ua?H{>WYTj_Kl8?;3kLssFcw&2g-e-&EW+(dfUTwz<<!K-Eux#jX~IL&*})
zj?Y68U85&qq(Wq|Jr*)WV(@Ta91NjYjh5b{A%?qoPJ~)7Pml~4LhCXFqcM^UW)n~z
zC9KXu)=pt|qaztjJB5*2Ors1c^mWPD>7N6!+$<^YtE!QZn*$nXZ2)Z%;V!d;izT&;
zAf;8CPazs=tKB)o4ProY9f7KiOWH$1=fs67c8V`OU%uS;+G9n~9!9M5nH_jo1YUXC
z3pMj_)l6s5ZwIc1TLG;wjK~0<-Ftzh=ROeH-CZ(pFnv@>FIL@7Sla+z8ewgX+;2{7
z$d6PkGmx&*EMXCpfkb}BCSI;c3U<5iXSQhg70n);!6KV1IWd-ToOlbD&Vy?RJ8j`N
zgmE4aFwvjkfAu#~FpV=#d&!KCNkWbwg1-lggdFh%eK=cQOQguuRoQw$yv#IvzQp1N
zt7@xzI+9d(^>`?9{7aARsIS*~2pBfqndQ{E@&v(Se%VPN$4qhk7*io2dNF5_g&jI!
z20j<L1xeF%OmGb0bM$=P&MgwW(#v{-t@pra*c*@>r6F!Tg%Iw!{Og41W8(gZWnXNi
zzObfwcY-A}HV_nb9K)K%6~x!xaWZcYGX_4SW#1cKBTB_cY%lBhGYfd5=5(d=!Mhtg
zUp4k74LrHRm_yj>^60Y-INJYL7fV8({!=!h*@An__*m!@JRLQK(?wlke#+Za!##Z>
zT&U)M38(+6&thl$p9!Z<ZE2@{Hsqh=(hY|b!ykyYTZvW`(oxqkz<8DYY6e@njw?d?
z_}3?kfInDIvc#yesTMLc3zk0n8@Ph2pKIOw>FM3=Lx20w6pXlm>hrBQ<cGkWUyS#>
zf;n*D?2M21xvx*G5@r{jcW#2#z3FTFzC0JNwZjI;ly~pxDpA3c)-CwQCzkKAYc2bu
z{355)%onL*l6O!NdI16p<!on$@B7QEB-vBp4VgDQ{$+J@bn!QDWBlwAw7~l3`?|8b
zKUSb?**x+~?fKW6{PXZUJ?$zoHDp90OK?daj3UQ2=6(=@yl07X%0ha=$b8auu{kj)
z!OK|kMKNK_JL31(TfOzw$9}OR#6+w5LR78^_{hwfW&q7HX_OvyR8<Iv5)2rYAF;TU
z$8M#yHgP#ZEtg`<izz>%MW8t)tBPR&BnTtut`{P(;50(r*RMy<f4i%ko5f#zdv<J%
zEPqDiRsuEIJW8P;%WQF9h!^jB{NU85t<`Eli53bx0=0MNA@+!@ME~V08Sb37Wh1Jo
zUmEs>SgE}ugk=qgumXb~d7rMn<~sLbY(8Nf5%)8oevPttx`c>-?{1E3zUQob)OfHS
zxMvk9edV|`?U~PxnNVS%H{(7CJ$VMOou0n??jt|+v?txgIU4fb%r#X$21e_NH%jXD
zT<C8p!Q8~hqAx|pQN$SZC7(^gt~N}WzIn?c=Lim=6g2)C5u+xyRf-mnEZkl;A&sNy
zuV=7M6DGEftL0Ih!G%jx*`y?m=Tszt`XESUz~^5p8v%b}D?2t51c>>kDrg_LIsAHK
zt+)lfbLx|tkAie}*QHkIrEUh(KSl$oP!O;^x8T`Ed}%-E#bd{e6tZ+Vw`o8*x!ACY
za=$)+#qRtwTDC2^kYjMpl{(YWsX7GRMYe{~D0c|OsDbFUy#|bGQdXEhrzA-RBx{KG
zad%hCW5@vCi>(%oLmc7gx=nbc5XHE=(rM;C>dl4h%^C?xGb;ok?aPPvx)9Sz2gZlm
zu?5$)O)OSV5dk>Y*!<xDf@M0$HJ`ch;>{mWlpZ+{YMg`Slv&p7-KACUpG#WxQt8l@
zwD_ys`I{}Ioy1qIGSkse4Q&cZKL)F!?QiW{N;8LZSLnIIu?~uJ1PAX08<%K5N9Bsp
zNDW1<3STdVfVr^8>A{}gs?l|#Ad|P?PhtR~79wyzmv!eB1jiqP8A67%pijw=Hfmc&
zs0|hP-5ps~ntv}pJIGOLSh#E7{qL8S`qk-U43#OXU2JkVw1TS@Ej<HWbqpR@D|p3T
zCyrr3An}Ku1lPaJCd14z752I0mOo1<QKL{e!y2N6+W$^vq$sW=2(ZVnp-tvJak__T
z#0|Y(9JKr&X%E3O$5iWzT=Jz+9!6lThe-Ia7dw!{L>#1HVdgL5Rg|StwGZcdfRATy
zJfi;!lhO!DJ5pB2v$hsJ?ij2*7v+Q!UfUXZm={Lm{X+o{r8VU(IAl*$gpgKJ&TW2w
zt(S^PBjUHu|NScM=q@}+QvBHiKqV9}0Z=65OoI=tk16`hwv*lgJ=HN<QbJ<5QQs21
zm!i)_0K^xP=KGWO&fBc3wBtMQyGbx6mBiHhLcSbtV$p(~xx)rr!nYf`c*G2u66P`4
zVd^>3FQBbp#??H|aPQy}cPpm1uW=EUi?I+6){SlaWYO3TCQ>`_L5X?QC__fOi#;W_
z9%|bVBlf#Tlw68+%n2J0Ys&FRzyz#7Ck?-1oik_?{~mH%v+;Q?dpdY^mslUE#cC)S
z(<aUU!X`XXwb^}p025LKv1rIqZOIM0t4bc7+1(v0*kjG<;T`mhxehr(Ti2Q_2}ytx
z6?uCt{yuY(hS`t;nM#gGQIFS{rWQ_^S#P5{&{1>|*_Mc2*G&HjyA1tt&tT8iR{gc)
zz~&A5J0s<`2L9F2_nb+Y>rzgKgB;o>)P7&nnjg~fpH%&UQnHl?n@r`pl#$a|l8#9_
z`l}1M62J;Iuc}IXWWRK5P8H|8lvZ?&pk}tFT28a8&U1oGy^F9Bw!tMa<+CHU6+LQ;
zII`R=AJvrih#!Qh0V@uw|B&rmFCkuazA(jO?HtiNb)<vyJ&)pS6nxd(<!ihJvPpG^
z2RA#{WwVKqi6W%5OG~LTf3I}r4kBrTQ+&rHE(6F{;GWYHSo~f~dVlm05DMZi^F#y&
zQyAh6I)vkw_laBzAp|grnU6>I)z2wJzoQ{S@OM9(H(LyG*YC?OOgRZ#Bv%@LffO&y
zAFb{x<*xRYbs_P`{a2vXtsOEuAh;Lo-KDOAF-xn>y10kHrqqh8DhN~FqXMjO^ytu_
z&7d{z6KKe=ihX8h0{Aq(XC{LmHH2|%&alfJu9)THaMO_A3pOS{!hQ!>j(~e{8dxNt
zW<NC!28kUO>I&$9;TZd0O=-^7g*OOMFH5<mfCV~IJ!njSL^Wt054n6sk`D~Qm>1hb
zO~3z)tp(1pCn&slACNCHR+KQO=J~M4(w~fH?U?-xdje!Br3XniRZVQrYvf<qN4~At
zvHd^4jXbMgIR9bvGXhPWK}{0vxe@?U?rB9x%<tWT#M>n|(0@VHk!;5PtAXduzZr3n
zi)j{{?ljCng|-%pws`OOwRx(U??uVh6F#N|H&?rY9*KR>Dz0gYF)f0S7+2|rhw5j=
zZT570XQ6>b-qkM<s?IhjRp?A~x5;0~XxO7-Z-DSp>a)t#jX58hrJm0<o%)AKV=;&0
zkqDyc;Box*eC;8DjL?O<y@9cr0|7Eqsnj!L{EcG-#%^;$!19ck;#@eS?w?(eyDmOg
z$ByA{QEq&KgJN6Cxk^>c?#<U&j^kgl8O{yb?0ozJU8_eCx12d#?b8qF3F&-8={EgG
z5y-bcl%kzWk0(fC1z5y?<?DC2C!vNq(Gth{^}hxs$X+TlFAoHPUfR^J>Ih7tBy+BA
z_J=U&x+;`i6!n*nkfk;uWVkVdT9dC<{4_4lthh9Qak%TkpwE4@43i&Yh1&kwnH88n
zt*rwEd$Z$&tpeweA#cs&L9uc+wUk|f=#TH|39KRsi|JWBurnef2;KrJ+iV#6{s>CC
z{MHf2PyjOAG}TU`viUWnE1j%&X+ZoGVOK*An;(?0zCJk1R?q_4K{JTEV@jE%dvZ>T
zfWQK&TvEm?Mwu@$BpFR+J``%E?8M9a3<)dE%%0Zcjb1vU;b}}ZK5r_l&u;h2U{Kb!
zVSnU0^!&cy5--0}s$WNua-!NN=2ccBl3Xw;0ZkFQP{U_MUwJL^3!vxbEkVH}Ly|7X
zCwqWcA-^B%<-@0#@kkNQ6aKs_;!PU9OFHGuqklg3t1KW&CSY=1a(`T8tqAZB8-dxL
zK<q>M6;X6a%1gC=(ZSZ7sbea>p>|m<NsG&x(@v?O93;%iqd&BZ;rhJili@cze{%m@
z;xAme0KIpZr<18oZ<<4nBqQvQKEYGwb%-UX<R;dx3hmba#oRjwR~mGUgGnZ~&54tV
zZQFJ-v2ADO#+uli*tTukw(ZQ`nfLu_w`#wt{r$N&RjGTQKHYu#xS#Gmos#m!CEEM@
zi3nzt!!-Xdf}?d@%j=>kn{JcEt5}ca-TF)f^B=fNZQl{Ouv|u6&*#>o-j5wZu}7Jt
zvlR)4jf#1`fUzNz#jdy^^E~&2eRC*e`nEjjKvC$^mzwU*lU4vyQ13FJVy7q_vB-yG
zIg@8hfZ#V?$zs<#qp9M0y0IK1QK%rF@dwhNEe`2_35+=yng3rqG+CHf{tu7z<Bde?
zHr?<W%t)LNr<v3=69YgC*G@q?$y3wIMShDyMjj&*#ce6Xeyg-UF%=I+>bR-FRL8I;
z)gkob8t%&`ie^9($mqe(xGv7%Z};J&@*+Oblftd{aFB+9JH5WWxIR&mwmospPmV<m
z(2wr)e(f(p%Hf?*oVr(5Aj(kUC1UMB_Z1k}GS&Y%_^v8I36D&OQ<txiagq=iiiUG_
za&k8YOjWj${B`-6&2PMmzmxah2F)6(_S+DKEyCR|U!((;peiC2dkFmAmUhni49w~i
zQ^>G?+(dPLJdPRSzR-@bNX8Y_V>-qZ#gUmfSF=Nh{;K7bi$Rt0le}djqb2jijYB)-
zWG{aVsnaM+kwYH+#w1amPc3pwb|XcWo90Z&kyau`lVV^AoCrbD@4gomkB5Gh`?bA1
zEp{I#IZ~cAnnF<$ucmbmCYdy@X+&Y<5M-fW2u!|vKsVx!P4c2X4jni%mC}F@u5PYF
z{?p~oM)NvW12k!3adRBYP$rmFN!coqfi8-%7}Lix$wles{uGhcYBQbXULdFjxh+nw
zX=2e<k-R;PKt$Ofo8`3|+Tx5d2EJ;;yJ7y4><)N2<D|w!{{uKXy$QXny%=6vo$v}o
zqnzKA0#SKn60FiMqO0j_Ecw;cEav#N7SO!enW%ASe;){8%KBai@isdnZ5ZP|;eaN4
zt_lEj;0}V=7jcVeduqJvUJv&Qn@CLb{KFM-Op#e~X|a&)MvhGJA2jn9-fe4&{h%?<
z*^IU=EXILIaJ2)g1pAq(o2V;>`II_fL!s1m71)8~m)>(Qr#?Ks5<eB)68Y`so`@6R
z?1cmrt@b`>eA}eDye?2ihh6t_=D{MYLwFwb8lnhi$mg(+RxyWF$QbbzR3j)oLS|mE
zt;a=eNaxj*er2@KBRZCcCne7&sSwT;$|&tVp-G(VZC_nfTV9+QC4&ovzAl@3;W(wK
zbJEm`Hg&p~hyG5h(MRdy7hi7=A8R>M-_;W7mK1RIRr{$$@DNHO?dgzYvKe&dM?k$e
zw)Q1jLZ~W;eZTkBz8AMWC*OOJJE(>kFCUZGFt_c^77l>9*p0Wg80b3GcsUVD0W00f
zPMU@yd{0m_+eXA>PQ`L~SJ?IqryIJ*%*UB?rZ{jgi2gCJq)~i&7fZ28g5{!KutMX+
zUSh6xjH1~4(HdmuXN&NA({~6<h~M5KZvf3u+78PAyevV0VAK31E$@$CBs1*|l-M<X
zND7R~#ZJyu%gt07>~P~ec*$rVD7%+Y*5%NKHMKH4Pyh#POhxx|-7|;|^^>dpTvnZ4
z;XQY;RpM*ho2^Di=i_ze<cFU`=!5UZB;gu1EHum+gSjY5s~YQyIInhwZ8$v#qk%W!
zqVi6sUb+-AO?YsW#w=J^y|PR<FsF#dVl>|QS_NIk<pcc^#R>eC8VlG>$+mKRaB?+)
zR$WmIIp5o_YuUzf8VapSTqd^;ZXS2z#`eRcXV3~ZZ3s%<;MTL{4m7Ea;hVIr{g)Rg
zbMNiEt;6X$LwO#TJvU1d-CU6<qvxUr{18k{suHN^&9&p(DkG^%2El!8DtgH5r*Qc;
zJ0z^C^^-q>qlC-rdl+);wn&l39O^3HF8IKX1NERe>|-Y(HoGBh*&WPA2Dx8SgpLnx
z4R*YK@JIQ}Eb$n%&pzA+`|&LCI64-bW#;JbcxfC1izSNrmdYxw`F#|YWDNPcpB~%4
z#FiJFVo$YK9&MhK3gpkZs&h7JPB!o-^F2bp6X;gLg4JtdyhaZ#2rDOXfNtzLNQDO)
z2EAr=L4aCN<}5w0d1x2({)|9yADi^dcz^XX*f5?A-9GK{<=T~FGd+1N9g${G$^2C(
z7y1FmhtGYa=e#iE+Os>1+)K25?DcVMFjI8OzjTGE=!?&TF}DF0>gw5|rHITL_9i+p
z_k<B?vVbbw|9n!%`yNmDDH^}m^YmwS-aav$_x<g<$-{LA%t(QoxIf{<L)AO<Kqx|x
zi8#-e3R(CxdoyxZdXg2E=$>DjIW2%?)8)ZvY}CjY5!w^Ew1BkB9dokF3c_cfhmK;?
zdWVO8c#6*ds{kDDAx<U!5~~&>yH<ls6DLElRIEMdv%HmZLiMrR-tiOss+dIW++5hh
z{rHwbIRNd9)Q@{H&or~;_Ym!aOA{}{NtiN;v3|phy_jQxJ!$>qHeaW$+izgTBckCC
zNOQvrR6NdAsM<8u_Ik7ElGSIBK>;_e3M(!<W(qAc?4?OE6?nQE^Y&`BYun&-6Xaz2
zjRECpN!{x)aVzvfs+aupJ@x#^y74`ay~0f>X-=CJ!smRokvr3%I@^3m#Vz?5EJyZ(
zxWuMfF}AVACOh`q6JnuyXURj#8+-8)3i(&)UqP^LUN+YsUKA^LtF!TnrrnkS)Tx1I
z^#mG^x*Dc5<Ic%v&y_O%lPu`?uI6%;>Wy%6c<7LCWsV)ix3^<aBPL$oh>cf#620+h
zDd$c7OgQ`_I9OxHF36%)GzuN@*MBQb-D8yxiK;XKR_dn*QPQI-NQqJ<9=8c5ev7Lx
zK1;Ru)XketMHa+PM~lLa6Wo@iuenR#aRpz`nbeyS$aIFwPz{bKPHhwfNj@|u8R~b|
zmF5-IyG_z!6g;WFe(;pF%_PCH-zE_&ahk5O{l*_*udK6>e$s0*hCOn`TyU1wDdpZ%
z%E?4rI>Mv7>7H#G3y0I?jVSfpvNI?DB)&;D?us}<mtGvU4aU|VLc^zDQY$8ws?<AF
z^77R~M{7z}6BX4eJh#J?E!I5O9`J@{{5IxR#9_VKj?R`)#EHKu6+@Taw&r0nkGoyt
z(e@f)6n17LbupG%cap4ymlf9pH<gxw@+<B30@pmjywX<Fua~fW+bldxWlqAHYl9JZ
z%}S(D{*#0})i1UZ*%{KgS=!hxMHW5Qi)YQ=HB2aAo^ojfsLwXM){)reQ$b8^cN(=Y
zy(W6;V5i^}tpk*FiQxBv<b?&TX|We$95!F1AEe*=Gw?e~t8?2l76-0{$v?S3I(Fx?
zm<(bZUEkDN_1FQjbrDP7`x0X_J^kz;MB{wpM71>(X09Vu2$#x_WNpP3d1v=56s#rs
z;Bi%qu@ZU0;3|wIE9!nfpgseLBC^`cj4|=njj_(}+AZZokW5FdnER16<?5#5oNvpG
z0J(ufjYG7QXPA?WGipWnOi;^5^uM?t+f_Y}9+q$Py}Phj7L-~c=-C5|6l06RbsD?%
zw82Iw($LfD%VamwWI4X!;Wak66dJmny>dhgE4TT!w)sO7ZIdP!6B$PtdA&HQ#U9lg
zJl2#2r3?jj9i#DdP%fVW^U(9?9XyNgf)c^;*KZEi<qMLqxZ-PM!#9x&QYqGf(`%dQ
zWTw@dXyZ$3%HvH~QBvK~lTAIy4En46)-!F86bDLWu`i#yRZXxtM&%U7UT?IQT%}ht
zyauYi0C0!0K?9#)uI?C1HG^0!f(z|;FM!%>rJk{M(v9|5=*ur=w;rOQt#1`Gkeo(U
zS?Sd1m>-e0dDzO~<#`xRUgYT@^8?8R28DMvYVpuzpWdYb5l_ARZ&Iy&X&g7<FRF<r
zYIw4DqeOThGy(&4rsL_UG6G$%-vjw2s9r>~+(-P0bup@(RYA&@MhDx4AAi!`+2!=<
z;g%sW?JX_dzp#Tr6TIi#xf^U5%q=Ua+8W*wn8gh{FR@U)gwZG!c!e9&@gB*Rxdpnt
zX!I10Jrl58eYTDHy-$5nvm$+}SlJFX6WJMs&#cvpu7~N(A>gC|bG5?&Q^MvFewUT@
z?UD(OqqtZ&?d2HOB#i`KztF%yM_E$+K)Z|EhHsa5!}-24W%ndfX(M$h1>P>_YTwn#
z&=vvkn9x}BbZm_99uh0dKVVwJ+zwK^>H2;DP2=s#ul+kHdPolfayEUbfFz6H`-j4n
z0G0LDhjDh&m(mM~;i(|ZN|6{b;5`oSvJg{mNUJUGU7xK}7NK%IC?5Vi0jjk7^fn>O
zuQPseL;N^mO4#N!RvM@|!)pQA0s9b(te{e^5#WMo)~Dhy{av3YjzVuq$o9CLC3fgP
zm`?b$kwhtCWJz!a=zp<)F5y6KGvK)S0|M^82I{awKQCLgn*%_!$^i`s%D({5cY9-f
ze?Su$VNCv)c?cWF|Kq|GRZj;~1_dK46&G6ud159;M&L;?CubL8E-prf|Jn8M_YS~+
zpZ||jWE0x9_INEQpO*$gIR=$P(PPOZA>R-|oswU@1Obqau=NO$g0UQ7Ml#ltXIXDo
zZ}1w{)(XnsMx0QI8Fc9kf?4jyc||Ibn5gr7LbbG5eZN<=kBvGi@^ioRP%#IJ-_md<
ziL1|edj>)3ybMrH9;m6N_QBw)YvGNy<&?|J^3~LCCT^i>*SsGlkOLZoDI^ssMq<cu
zJkyzEoC<>cxI`_*4rJLizhq`!Y*N(!hEDBpl0S?{>5g@ug&l>^(8k959g8_AuHx6X
zEJ;Qeg6X!MH?VB;0i!(~R-?7KOC%z-X#^I5@H@z#mpd<ryky*n36x1jW?o}3d^N?C
z4TY*^{A>6yVrw`u0u9wi``7e>Y*L6FQ+_OTB`VU|-<o9C)l!rtP2tOlO(=9p9WTW{
zZw_pgds;);e8kJjg&@sXzjh#5#7Zh)3z>eC#RMDE6sFaV4ak6%ttE8{#u)pS|C^d8
zMDUp6jkp<rAv5$H-Hlv5CqO-+M$sFdj78!d6L5|S8C7G!TrDi3-luMS`N}mw5vJ;Z
z(KtZ347Nth(kHZHZfOXHXex329kC`-On{#I`;DCAcgF=2Bb+l;CsTxG@87lbn9@Jc
z76Wcj<A<R{dNd&|hG`0Rg$ppoT=JFKnPDC?<mltU6K!eeifZd>9XY)p$<_i)Sz=Q}
z9E627&GMYSeQcj|B~^e@wFrWSD{_&)AlPL)gOclO8b>bUZ$?wpU*Nno4Z`eVf7&5T
zd*03|ETsR@TUxkVDYsO6aAxT$KC(vuv7apKQUm^NgYHe1Ej={*o(&)W9Lh&v-H)3Q
zpT^Uy3p=7*SebZJKYV*Lans`KO7~rtPt$T$|G0awzkLc(u=VbEmQnM~sL!_BJac5J
z;2%GWc-gP*%osav%=)#e?ce!&S7%Q7TvwGbat5FB@H0*>Z3FaYBZn<Zc4Q^yLUXFz
zpc+q$emmX-{bs}6$g>j@^7Vj@#k_F-(n8ng2N7qk5wVAxv#V=o4S=m2En~v0jP9G$
zmWo;ztH#9pxfg3?Vlrr{7n@aXLaNioo`L<E!Ibu~<jQ5poY|sLl*%t+>X;ZcdGbMS
zkTzZr_C;YB50A%kFE;x@bJw><uZP`wFO{vvBU&pzwpUG&U@e`g)u8_GG%D?tSQ8UI
zx(uz=-vm0!87T6y)7OmP!pt)^xCW<~P&<TaI>J{l^^B;-vfJregvc0L3|B&Z^K4Fi
zEFOPMu)hUd5ONb;tIK2vbIy{};1!Wilk!=Bx80q#^xm5<R86E8dgc0Q&JR4Vn%Wx9
ze`H28TaLi>3!&X2d@T~^RY?F-jC+z85@XRnwyY~=fD<m95c6eA^?i3-XHCAkX-XTX
zKh^mt$s~kEJIfG{%sX}!+!h~S3SI5j*HJLg+vK}|8(PHKSr<MHT{kvdBAZ&=_MB)R
z?#J1PrO)``>SN|_ZVGC4OWU20x=Znl=;eA%4(3zy*~eFbq>cKF213Bq3wZ}}TBfou
zlU}RC5AJ!~&nqV0CxVqBZ|hl?fsndjb~zM4xHA6>)cR|aE))!stRWYq)(lRa52;P}
z+QEEHyVMxRvhbx~m%ZqyA>CKHrdo5+QCgIeHUA-UA_9M(uX4hjV4o7fMK-KdUk0FK
zUfBNB=a3I?j*u*PBPQjhAQ&Di#(4C9#oU9K_!bsh&v7P%EW75p=L~5|h5mB`nIe@?
z(t(!X5;?9mSe)n*+O29Mo~-xB9?~_F;3UYmFOu-QUeIBnk(Yd4AbETnJZTMqv<*X(
zm%c@&C60uM)_wY!bb_j`lYSvt2{B@`^n3)AnYq6y8LzvY0MBfUdygQ^8`O{_))udC
zDNR*BE?R$FkVHz4PVxJR)tRU&`_tn|5G!Ny4w5M~(uS!Ug&rfnF4A9M2_u?2I}jLO
z1zp2LZrUgIn5z~QPt_UcU@9d}p!}pJwfKDtCcx|z4icydf^j)TUpXQey2~$yZqUPX
z<%HOdD-}fT#}rrT*8HWw+?LdTLU{)%1?RUVhM)m40wALcS5l8ldIw?M2;#?^@K^zx
zMT-Y?eS2F-l8*8hu5c}OpRJrulxWaKdP5bLRUNKFBe{a&>SBwF)}jz>)bhRZFD0%N
zVu(XI2!Y&MoNU%d1N%~I0Y8bT`-7{<0^yXCmU&3plbg{RVRn?Rht6@7`(^n>0Kbw!
z`!QlDe?VGsBWuljz$7dQvs0ZY%AG69dKm)qGwcY8^Pro5^bRphR;F*IE2n>)ybfux
zWbaOt>%&KL>?b%uu`?Yv>_aPOn6BnhE;E#W+!v9od2Yp!mP6KF^w@X51?xJpl_HSt
z<j`|xNPFl<xbx%pV)*(HlUBGivwU+h5*NC0P+$l_0Ay{=SAU7{{V|P1vqC3!+liUy
z8ll8VCIaD|om)JmIB!mM%7Q*~gNi1NcA<5t2<AY;Y)5Y%FG4H}*)zRwg4`L$qtA8r
zg)=#z1g%%5cK^ZRJO>@iD1o*xhL1UrPh6XsTSmS<&`8aW+lK3qwpM9faTyQXrjg5B
z(ny%q2cEhc9A}3mx3Z{QH%TQZCnl@Ab;%S2i^Iv>s9TBzzGmbLv$SJ!x4a`MDLIz=
zG0b7PVW|DE9#Yr=z?iP3;C9gldTUMOlNg`xtC?sUCX4VUiCMp7UkXZ}?>O|>7xxxj
z9rjriuJU=AYlI{Cc2s<IJ5!Gho_HZeM;uA=X*>r`|3JkZDiC6);>>yvgX*+gJZy3A
zsk8H$!)>GA=Lh<7qwM^Il62u3!L*N57~BMO7f5p!gu~Oc1YYY9RKy`y!HNX^xscs@
z0~@3u1C9uavLPTe6=XZN75!(IZY|ah{^)(aeuw-5&e_>KzFAi~q{OK?$l)~4<rT3C
z;EGg2>w*qd=TB8A{9CeB4`Lr%rw7F!_rJ~_IjF2ZLG^zN7^p?K-G7$yU-LJ%aZu!g
zyx8P*aZsqbfiwEKKMi)|e01Dw{`|bz^k7QWH(0njU5kD1;?xwpCESs*L#dT_Xz%<+
zHc7l>=P$R&^6|Fs<U~hbmHsj$omE=r#obaYAZw<0+;--9=g6NXuS7@aH6a<=@r`K6
zLK+o-vxCG431B||MSedvb7A7e;>*!#iPXR!?%O>-m`|`fuQxG)Dm6#K_LLuL9tEib
zpXf|nL1lj+z^#}ZlsN(_Dr)B{Z4@9>UC-3*{1HDwH4@ZEPv$sno%&O8Zs_5wql$uk
z1Ib?zZih_$Y;|#p*=scT2LDU=%6Yw$+Ulb&qIu37YYg4Bs_Q<e?vT>_i}w|G)m0jj
z)VSxtd*dF4caf&yuMaHwVBwb}QsN(LR7A1;61@}75QCH)7$ZozNwf5dc=y`+HC&5}
zL-w8`)tN07sO=J*!>=3u>nE4~2$)EEW<v<>^V0#3NnPo8BNarBHf=vwRMy(t-SubK
z83tC7qHkw*2=JY!;@hSeCjTIf{h@mOU*b~^rvG<pFBT3?=KrG@v_V(q7d{6fa0AKx
z%wTVLaZ`2Om+^a9hUzBpPI7*-9&W(^eD*(os^`n3f1m0dh^)<Rpi;-8zr-4`b$vMQ
zHTnB{eBADYgs5_iWkAx}8FX%dbZtK~IJ{3Er$c-f@4Xl^Xv)s!#DX-adRESm31f)x
z=Xo~}@K4%y0B_%h%M#Fm@N7qa277-a+)Q?KY?R*$PnHyVEs{;$qm8WZHq8m!18Mo`
z^Z9xpK;)s%uUA$Rz7syGop_{6>PoO=)<-F+!<Quxk{I;ouBh2B=uMH~#NPI6G7;h8
zChwg%_JX9i_h6z;v+sM5^(XZNvyoF0$~eH<{Nr^>A-gB#SC_B0i3Ni5q#HcRRGLyw
zysX6%3`Am1u8m%ZZC9`@Rs%SRMW$$h>1&d}Xwz#!<S!!(rI<@k%RMlT%W`{-KaI_X
zhCiq)SE&UU!Zo%Ps~tq!d$}OPWj|V}baqUey1(kw5$os-?%M+*Z@yE_MdGsVvs+nU
zkJXvk5IQ2&huF&zR>2N=DE9HU0rC;ha)L#&mrAZ4VrO`X$n*$uYx}CplEvBtV}JWL
zw)6YxFj-E1gKBMW#!*Xm)t)70j29~w?<x5W#t|;vl^6xj;=1DEQ#5MuTOHKCze@Y&
zY)Onn(lQlv4t8e<_^lxI_9yHQq)Qvlyr<fS!CGx7{PmaS4WnO;H8>I{@a#AoI_$-i
zNXZV(LFu5nE-PEv{`5qqB`NvNopeNo1E~onqop#K(o8+RJZDQ7;#t&dEHsrK;?UYG
zwq%(D8cF63NsMj#$oNHU-Ur{=@_Ka~Qx{oC$Mu=Mi>9$aMEnT%#Jk@d?^eW1K2$=d
zpFkOtx%2pT-Op<C+0g7if>1ZvpyH>y|5b<yK*Xk!M2b!*Ux+N(u$V+TU%MbhopNt-
zx}=aDgiQ3OW?XgU^Lx1sLX;ZF`D$8YF^^M>-x|6PiU4yJ!NzJFjufi~!^gR0n&G|!
zCZ!umQquv~E`5uP$0k$R2jLDi;s?`Ku&!##?nM1{l3|BiFwVU%m)9_&={R^D$!F0r
zk&+PMj~`B4n|Ks6l~$9qv)~~)h(GLEOtsPiL^|xAnlPZ8SjBQzOK3Tf`JnW+RiYx%
zc44o?JhNk$h+D|pQ({#kM=Iq_U>AzN#5!n%&8;Hl^>X+@1b}cu&)L1Mg>%%(ax$B;
z4NF4659ec46~U|gKB~)jl%z{}C{!7d!8Av|t(9a9_3DuMiutjn@5%Xn5!G7fCXqAY
ztmC#DS-RrY67$&_@CoOtv)(&a=cX_~rbB8w(LUNxSAtZzo0TEOs-=Y3E$n4{%ZZIv
z0NRlJ#V}b#UJyROLk4bMW^+P&lv=ePWFjFk0WIWo+v=1WPcI6VKN~2w@IhdjV#`%1
za1K{e?`c0We?-`~y#HKTFKX~1K+ASptj)%xSjIj_B{cI79`(k<MMZR`I4Hhu%psvq
zL1bR<WP{0|t{bB#lce}5Q&D6iD)M#Tg~g2t5Ip?@z@a)~Z9(N>EGKIWfso519$6#d
zrp<(CmVE9KuzNe43M)Bl)gt2tEd|cjyYK~dy?Rd+wTppJAN&gbr#>Z8Rf<R)_SoYy
zV~|Dvjt0+u1t*1*G(l_=Y8;Lg@)$pS9+%l8BVDj*duh8wK16(h9!HX!v(^#;^%-t7
zVToaVL1`HyWU<@gVriwwRXKOCwVGbwYblxpHBKv{B!!Y}B0N+L0yz-}$RFk2HjS&T
zMVStLv8M$a%NG;qE2GCe&!4t<vR^$Me~E;_3xmo!vT1-g7r5X)?Lx&<F<Wo3p+*kW
zYEr?8)d@M}!BCj=vw-_YhN<k6TW!D(%u0M79;}G;8fQlZj$6ibl{8>K9pLhnkX)Q1
z!i!6h$f}U~PVn-e7q)`g3t40AmLjKPx(2zxOpwEv>a|DS4<`x}BHgMbu+|lsk3F>#
z_Zemo(YBP7zQG@sNNK8NV*ijmIfI8)4>d=}5_wq@koge3iu>Iu=8T~1u~GwXplChc
zWciqh)}tBhpRO+f;^5YOzDu_vwfpr9Tu*fE3kl1_Q*Th$;TlV-0ry45)}HpREpA`Q
z>;~r`9v~Wj<2zNt`^S*&VS09HdQ*%V=S;=@4ew?-2K<%9SKXqLy&wr+Db%CIk$T&G
zUHrD=D*oG&9FCxp`6Vg0Z&@-bac8ezjq59={h2#{iYGw(L6J(fe!w#yExe_o(U5~-
z^6hKKG>ic+;X4Z&n<&_4ELk{X#SP%<rK;uT{i4nrJFHlbqQUF^lF*si?9)4?Pgpqm
z?D0q$Bw@SZb$`^y6~5^=09)Ii#>wG>>iZxwmiwIH(K(TiLEUYl<c5pC6sYu7%dPS0
zSgP!9Q$PNoE@nSLn*BiU;?HjB9Dx`2N0<RNr4mo!xSvV7Y}O>a`pil;$YHQoz*wVI
z?M1825WfyO<ob91=dG@hwkHOn&!sJ58VZ#4_s^w}>lRow3Kz7U%h9hcpKo1F_C$X9
zs&xrmd%W0I&rpBD)z^G(RqS-H@pnKu#5qF27>%8wN631=Sgx9Za0IKNW?8fSxfWDq
z64N5tR^&SnY*-II5SCyq&#_=aDBhL!c<t=K@C&A~YK)LQ%0jdMdA&v=D6AO3WIb&j
zCQpu8vWugNZ1h4o+W3<H2(HaME=*<GQ{>GE&p1U#K-cD)uP4fDYLq;*=^DgM_Z*GZ
zz26G10pH=<L?6r7oGIM0vNR>Rs-4dXy{ej90#f#70a`FQW>A#CJk^g~_`d9B{8Xk;
zXcD2Erj)Hb?#Zp$d1ams6GW@PC0LA|pJTnfBG!zJom)PsGAKI^_rXWyJ<%Sw1#rS-
z`7Ekvk8?m_$^BM^#*$lJg?9IlP?us_>gof<gf{xh4g7b@&1BGZ_{|KZy6BV8h$4Xj
z9S9SV0&`#q;r4^{-ujj4YU(nJM8Bt++Me-y>0x{}1(jh&gR!&DFr9H$1r2K@vynf1
zLoOi}He_KQJ02p+Tw351Q;c>45kUa7LmWL}>u^&<vEX{RLy}@9({xQN3`t1c#^P0o
z8tSWIxhqylprVdOm}&^|k8MPyuH1J)f+Y!MX7XpZpE*&9GSV~hw3sZEraqM2AJ{t3
zQ@qN}imb{^Bm;@HBWCSN?9z;70nzI1*l`xxNM(D~Sq~G1J_jB-3U1q!^e~MD?o7~1
zU$MS1P%0A~(BG~9PFgugR@rdkGL-&)gT_cO(1eDL@ydBjC7-SgQ)6;(A%1$%Zs%03
zw=|Kl4s|82{-V7y2`V94+>srsx1vS23K8X!@aLo1x^3JGa;eb*n6kl&|CR#1A_uXI
zr&W==?g3c`4|?I@8uu!B6YXPlm=?WBM3dxTKfF@7$MMFR$QutvY<h?qcj$US$!9!?
zI#2{dI)??7v4kAz1(_%&1DrtKv#9Io1P_9Q>1e)=*2`MmMK(FNYtbb=c+Z19!4srX
zwE#X#?_T^5I~^<K4d-f<mCxAq4v5|4kxu&ZoA=F&U5Q=L)W%CE5%1OuYn+;_&WXe9
z>dcZhHI}TClxGIhoZ=|dj1sS5;dUzew3O|j?3c=fI^i}064oy_oq=|an-GYU6`78V
zdP$Rm5^(IWWH|+woGWk#xyb(5OBF|=%gUf3l66L+eHQ)L)7hnP&gh9C=DwZ@m*gA=
zR6)Q~89Zv{+$BN+Tb1}S^C6-I1^izyNJPo$PL985UB|TA?CH#+e*NMX=?*Qn=-zH#
zYD3T4pV!KlO?3i2&^#?CC9#h>sk0`;<VH_7R#4DPLVwf9Ksi)!tZ2qi3_cthH_Em=
z>Z`3f)k(|v;|9FhmQcyDkL{$#n_3usTl6x<4ge!YQU)b6!5d1I>cG?HN7m57q5fg&
z&|-_9Mb;>ShsRqX?TxkQkhW8{Mw2~svdMHWorx&+PzL&B;<>Y9bdypa5Ku4JVoeu-
z+|dmNJ&=hhXvQ%i(Qc)Zy6~7#nmxyWk3ZK4W4l(nw)nHL!f<ibE?A)p0mXl&XMYnT
z(<@!nE;@9PxRy<D`lg<ob3Wh>OGjAd9#+PJ`m!O=?a>Q+Sbax=cBZArqCV6F=Hf-c
zZ|mes6D`-D!ryg6)*zEC8&Zh<jz3o(6vNFLNffPRB$g`7p|s7Nb_FMPL@ChIxbn}~
z3!MH?Wlw_&2Sm!c7H;}5{}Ai;pn$X(olvY?N`we+-DngHZB?S1NVOrIl8eOo9%ImF
zk(PBk#kjBAt2Xf?zKeBc0i&(L5{n3q0dfDqK>#7QeqBLirL&1ZyZ|koNK@Vsmo5Z{
zg`vQOnfnDy>6K-nIN(^2d&sF7T2_N=t}GF*-6fGL4~C#pqLPTM+V9MKG^wRtntFEh
z@y4VE-5ok4bHvDFP(DIG<JX?uA8s5o$rl3ClrLk{C`oCU1-<cXh!D|tuAtCe^B_(>
zf$gUzz@#3==~k0tFk;kYMe^s%Yd$$gL5f*i6XbP;JfvsCoxivzw{076fj67~gv@PJ
zo+i#yKxd`eY8vRzjI4ykRpk`kYrI$MwJe!HAB5bg&zph{xz<CG=kXHa%t?tmq}i#T
zKkT&zK6s6EfZoCYLI51LeU&+NnN4k%+I7=HbRmW=U&=3ky&ePyI`GT|Bzkwk4B*rx
zWZAds=zo9@S$i##bPrqd++y?gEF3)SbV5_0*6z@s1k@{wHJfrKTOCz20A83#r)*W}
zDq74M*HZ+si7Y*(2_V;+&E7q-tnG|e2icw&Gmp}8C=I#iJY$u}YuXy^PXnS^pw}CN
zb%(jSZs4z=HTaxMh+<bg3%TwnWsgw*1fI(~(<LXQH)?yq(W^T>|6shFA7tCt?tRIG
zk|;SB_Y6VB!NPKon?O!KZ&|O*G7tZuuHE2R_7dzSJo40H0f0$bG=3@J<$s(t{|03R
zPIp5OpDn4++)r6nbp~13R+VO-FOyqA&Q3?K8j_oweX>MEvsn8seegr3$7MpJj_ZfM
zbYRs8!o@pF7N<MLq6%bNrjJ$J`=8GP;F^ud2b=i+Iw$^rtl9i`U51I}|L{7P3QZY%
zQVw|FoOqll^sDqd95mQ73d8CeW(x)%#s(jR*yA4}X{-AzoVG@1Y;W=3&##hU2^oI!
zPOptT?N?b-xRjLbphbmGL}=52`35np78Cm*feVZs{<8j^k|p2xpu7~ND^48mFHg4G
z`7N(SLA|}UA4W*Xyef`Cy##LE3nR)Vc*`DR0f-D?b@?IG1d<cHVTh76+h+z>T~{W`
z2I1dd68{WVH9c({*|82{(@lq*T>T<Y$bro5NE6gDOvv>6VEI_~C0c3hycO(>Tg07v
zfmK4*um()?vb4cYDW`puR}23BCeY<4-#1LORa!X+V|_m?H_~&D$Ei)*T^dWe9BrKI
zR!?{?NjDYi#k%5=J0cf0j4Uttj1vw^Rv*>6@&2vz=somhG9qVu7)Ac<(3`HC?wd%s
zV_(wCiZG;mANeu3m4>cUN?0Dx{>A0+@`EmvyQ+$kCKPD{?>eZ6llBx0VE_q_+jeNX
z-Z$m8a#(TjdANv44xM~5SXPHT9|(CW&1WG(j-4g-oI8lmy`}Cg=MG4Jg%9X=zM`rC
z|Ltdi9ineuAfM^`SkF)|YuB`-@PS}|POP`DTJ!;3MSkF)S#uiQ0uJMsPs#s*M0j~+
zeh0kUK7$6;(7nD;iP8jKM|EBhSdi)R9G=2;A7k?e(qBb$GWWmXK$dcM5aDDIeoy&&
z5*N&-nIEtXxr@+WfT>=pOu4KII-<(bAAamN?0u;dmVI+x0-P(sn@qv<nuVheA=VN}
z&ImPj#AJnA-z(iqyU~>XDd}Pof{_`;9L>)+v4#ILoNvM*EX|({(53qjQoag|zrn9w
zPK>Cp$~i;dvLR;1LJO?(o$)LNHFe>ZYlc>%9*&lpWtw?rt<?0wZE8pAa1t<^$B%Ci
z&Q2Z^*^6#?r2L~|z~-hlUlX=X=dPB`vtyCH|8k3(Ug3FjG+Arh@*w?(_2qpnr1#{W
z!CgP>@(t{?=-YWoET#_g9wmdUgL+j%6>Rff8H@`swY%Vz>+N_LuH1f70)f)IxfNIu
z-<x*XcnPuZNlG)DxH$0of}fh!3VsWNai7Y|f<kzCwyuS8GFFNO_j%y1eh^Q!b|@!Z
zmNVjCRfK=znljqMrxMwk-Vgo=d@J0Tt1C!wRXDyWXp~gKQ(J5&kJ2>IM>2)XOP*VT
zXrtfBp&l!6%7bw7H&sL4#yXx@6(z#NpW|7AKc9;6fxBz$k}NrmGPc%x*n#BsWFLz{
zaMj8z9g)=WOe-bwX1_%4d~nE;qI>!3_5w|fXS^bNUv2zddG&@)-lK$6Q+yZ%!tySB
z^zaQNyQDH-`;aANYnT>wv{=r|EEBI-T+_*7@R}DBF{GC#SU;~MT|~1i1*8%N+~B0G
zVtHztk|(DM+N*`FSzE`0Oegii%C^TTqVe8eH(Mv#$9QK+$X(D?1mQ&{#bUt*Qu|El
ziL<tm`dW{8)uYZ`R&ug_DB3xil}BJkI{sH%?V%#auVTGoDHYdni-~)}d(e1==m$-^
z`sHx^F580wTMn(d*kvgSw<LyViO$wA-PF8B#Xz%&hdpg!3k+KtdR02r*e?`mhUSpd
zfj9Vv`@9RPSSxgQG#eonP#uxgRWv_{6WlLco#e;C+N|-_m8&VFlJt9ICzs9j$+;GL
zZ|c?eb-TD5vvhv!>@G|19J98(ZEDgAN$@24XO(_{MfvRor_3joiG8!)Kh;M-zYw+c
z+|hhc)`RzhRp$iIbrgQF2_lRH!J)7Zge#nb%wsntKx!%Yi5ytS9O;EsWsNpIlmnJw
zP=%Eum-|BWK=EktzuIZm|9g+`zqNHu;NrTeu?vHWtC7p!wRS06LvvFG4NDUj3u1QQ
za=WCdrMZO*F)Q$^u%(N$qN$UJy{&`2ovED*F(-qFy^Xz-ii4rCDTAn~o29X-vV`#e
ztjjZTF*E-s?bZLsy8J~Zj0d`D)}PF-GpbQKjx3Kn=*sZ$JvPp)@$KT`<880<q2jU1
zwy%@LeUX!UHK)Tg#gQ35ov?cmjNn6l2r{A)2gZ~+yHaN6&$Xb%No)NvSV&J0BT77N
zZEbyf^Sr%w5iTlgYgNg9?Re<i$|eT+|HG#`W^!_}v61~83W!ZupMt=L%lUFmVPWCm
z;9ycx(&OV}a4_U6$s;@nNC0Q>(aYlEqBzWLj<V0a!NbGDJ;NjF--5inJV2tKnwnZ>
zW+si;SG6GgA;VvwARwWD*{X_)s6;;(V6jWvkU{>N;zTF`GzbW>xJ=jKp_vm7MnF$u
zCMFW9D!XtH5Cx;Mdt=gUrxUEo#kt#nbI0ML<tE@`e7HpK^7}mh68L#4&I|Xvp?tg)
zQCJX=ZmyWbpH;7xlwXlxXJ>nXxXONo00BYP8<wD>qdPCn9-N&*2NvbfmMEGgQ3G50
zu5>b#)EXbGt)nAHJ?i2S!-w`7hVlmtSQMM&C+h9p)`C<y1%UTb4>kA#1b01=Xsx;Q
z%0NLv-52um^5_Brrs%)MLfJ0=76PLEJFe<x6FAgqz1N1u#!=8E_9+3TOCcfcsvnAW
zK-gB9uC%l?3qAxDUMd%buLE8{tU*qzl|&oo_2@rXM~6oU(OW5Fzb4>;fCxHN^n2kc
z4WK)fc>8$72>EgND$@e#t0{VEi_W~wt<NV}FTD-K3M5SH^+b{@f&~e6D4IZoiUS0M
zO!WlIvfI_iJLUrIFAV~k89wUo*`;NHqt94=ef9eH&_h_lJ}>Yghgj}wK`C;)Bq#`R
zA=C02RV*LXA8v|RFc87<nU$3hX#qe-)V2_4?mCc`JO$Y&e}9SV=@DG0`3wT7o?2kw
z>)Rpr_bH@~2KE3J9FU_XCqrPBCEAY7k-Oahm(G8~5B2){a3o-KQ+zU6$;U%6CMiSl
z!PeOv0R%+SVL#x^;cpvjK+;qdRl7W*0Efo4oz~ExMic>ny0|CKYiU`@a_a*DiBTK|
zT>`e^NSyb4mqT1pM0}kKWb|G^c2B|O^><ZXD#_gSe<IN;Bj8!m)^kno!Kh7$jywwN
zDK^=!5L~wpfOwM=@RupaFtlKmJ3>buih_Sh7s$=1jq8Ab#61eQZTwq7(jh@+!&b)#
z5{bJX04Cf53KFLMO2Febess5j^A~Uq0Nw^3*#!q8xSQZ-!wS#*6Uy}Wn<o(aG8qB(
zlOVF<Zy&^@9*o<pFlRuayU}=s0M``}zs`Y$K7a)27dj~-bU?uA0r2^S|LreyW>!I_
zo6Q6q$?6};k4T@`!>P7$fz37u{Oey{<kEJ2diuLh2@epeRD|5n)|T$`5(Gp^;?mZH
z_a7)}@70qrDm4hPc0u+6wRsE*Z}GthF;K3iNn2W4vi(0_Xsg<5J@3qbL@NMY`dhxo
zMG7p>nXH@Q_ule-io#(rI~-4){D=JXv$HsmI)a%)ME<;Ak!m-^f%@$R`EMApN9(h*
z!oMN~+&}{Ok>VcM?Kfq2<7EU&!B0YFT5rLCL*veZ|4X=6UQV;|l+6zBEuXuImzNh%
zr$DHh6YSf6j0m7d%5t9M($dl*i4gTbPLHNT`bQwq%N{uTX5t*wzp%>>|6v+$s2pV@
z6hxrY7m(JMwXIL~@7)0gpSQ_H8BVYpDLn4<e*u0%9ZWF-`;~M!PTA9#UR;DkGMWq9
z`;R0ebOV-M@Vymdgj<-x&_O_eMKEVVLgB&PWXd2Qf2)^61zk$z|K-tNP$RGZ$OmD6
zA=(<?mf@d{{-@GEHUTk~xA_mo|G<GQYs(n(L4Xi5@o6rS?Z5KyaPRa4KIRb9(Uq5$
z_5xLZyC?9Lm;lg}<YxEnavk`7X8{a}q8^GX+S<}m2uSVg@4puS0;=wn!&kaa32T=p
z;}38M7reYF)r2<I)_)Zp2;n)sM3EGjedywoHepM1z|!Gk2<SF}Mxw}pIzZwDjy?bi
ztsih-TWj@PAi=}Ww|=v-6Rj2moNP{&9@6RrzIu5u3h`0>6%7wYLte?|^k;&Db;_0X
zS>4NsymbxF^;aXms^l$=P0giXty-AVf8^^1!0h$7=`R~hFpQmPrpyVU>K>iy8uV5T
z?=2p(nR~swdv!)d^l1>x6po;PMoggf{8JquE`0z|L%!_OLoEE0cV-3E9d?ecxjqK2
z`R3998CPd5{@n7B#ic87q}iVA_Gv1i-qYh_MO+dg-v46A#JfCAZ=!xj3)>7FPUq;+
zvA7pRS6y-JJ{4v#J$*;q)Z-1C8$!JJ48YaZ6=?Ql{XtjK$Dzk{|I7GpKv*ng{N7Ph
z!Y!@j&zH=6;ZksSbhgW?LzedUU{q$(u9gH+jl(54I(#CIDsBErA|AK$l9I4S9W}Lr
zT?$VvAz@F&P1t{g-CG8z(FuDMa7vk3r#>_IRt#HGAk|i}A|<PKSJ3RLY~cIqEcQFA
zUm|<!vvGwc`k4rNrUoR*%^l?y7E;Xo&zLItBpDdW(hSaai)q%CX}lbEeu%fymBn=S
z8_t69w<zLM@Ez}!ts32`sVMbH3vic0SxfE}me{a%mJ=WTPa}SORFYbGJYy{wp6mQo
zoTe^2J>s=HZcmOh>qo#dyrD_!nzHVD&L2K1JXh};a`r@0A$kvavkInQ`H!w>gUJ~E
zzyG68NP|^SYn+atV#+AV+7z#LmTq;N<`fDUEP1LVzB3AOdhgPX>?O|?Wj{Ec(sH$`
z(xR7v&e&Zs&+)6afQ5sWP^6b)OmIR?78p>VK6xegUtSWN=t~sap;MISB~f+hCsIl1
zAd6!<8(Ria3(6vzHwfcHq@vobyo&Z85Wg|bh0Be(v(%OoyYUU{bCw_j-A*^a@vvu=
zy#|%+qrkJdes}P~E^8uH>umQY-*B+h%77@lY}(!t-fG@mX~}(FQ<$6=6LX(|L{{Xb
zDFoAnkPwqMW(Xos-I{NRV=wkNFr3fyenXA8=tz@e9J<@l9f_1yTc^}3;tLhLQN_1b
z_$`$GjXSud1?qUI{*cX@LeyLey;O;8F@S+=3z>UW3+oD!wA!cF`#E&6hTWI>)Euv!
zz?2EU{{Ew<)=iG#O71$hcf4tS`d`}k%Q1*Y?nj8XmJ{>;54$>fmnUzu_(@g<swS+H
zpai5G#tOJtylLiDQ?w>jq?&&NgA|^jsWX{VM8Ad61eQ8!fjK@Z(dKeuyZ>n-8mO$e
zffsguH@SndS5u_fHj}c-Bu!IeB|zd1C#nkL;!q?SG&q&C@h#;>KBZ3s#U<<iuDsk2
zv(Ipw<8+h5Yi`z228Bc@AS>N9Z2IDHA+_N`Tv1X&ax){8C~J6^#Zhn-H)smscqk;)
zYNVs57X|d2+dZ03#Ia{si~H;pL{;sTTr8ft(p*(yL8%M88!E=iO}eir{MiiJ*`Rf=
zNHiiXXhQ%|0Qf((G2Q^G%{zN?3T2+IX!glL99-Msc)llV38RT}(>D=JhY5;~81REk
zd?H^DZ(VidWI_`K$bVu0<bzkkNx{UK-g+S_{r8^jQn2U;vsINIOVynzsSRb3J|Y^P
zTPyANhlg*O<6KqR)j33bz8=DzgU^ru9LIJK_Ft@R^M5OZ<MB;ewDImnDk|@s3fFYb
z9rqBGanFtEaA2}Ba8EaXEiz3m+RNdTJ#E&oZG{TD9)sOo1N>z5#-srUA&k^RQh&qK
zTDOq){JxRAx{+=vB85awlIVA0N@qC`b>Fb$9BGrHMDmIGaanjRwE?6Z$$0J0<|kqh
zVC9oMSX=Hc{qu$dRcv90@-uMXf>sedv~r6rwOP^2HN90?8h8RBH2xBvPhP4(ZFLZ&
zzmgZ?U0tcSd~~25n1&I5a-i{t*F4y`&SG{<XT*1_4fnBmt@!@Tv6@v^*<II55gPU&
z7zz~RI;W41YS=N4JeTrk+&~hcr9s?AeMfODY`1T&w7^4EK)#)0N<xGV66{cjQy%~X
z#AEt}u-syIu4A{h2mZ<tj;eS!Z}uv^_m!N>*5J@xl+~PKoROLmDc($jwdw9>ynA!c
zQxF|(z`wrk_=azUlN4+yh1Bp8z1w8cSJVVr?q#T#xXs1)7%i1ip5zSkdh+c;fbKB~
zV}5b5_g{}Yg!x#0^@X<1LUX$!QEg+w1`lh*Q>oM(rNT+`9mk4!8S@*{m;D5Gqi?%U
zZ)ay`J3>GoecYV~OMua#&v0hDZ+|&_A1<?9q7O&Oz;13P2HI5?$5AZos3)!j?C&z=
z`n$C?t@Urve}tsOR|E6Hqn1h3?b0laJy-3WQ+GKUH}Qz_ifI5z!o$7%c2ryH=^jsT
z^<#qYJb`h!5R?!cXobXc_84WJ;g4<e*Ine&t2I~&vh{Q0G8PkiGtYOKJ19I9SEdzZ
zl-+?Aa2j(I;a^|QGF`?)(Irpp%64g=Z#loVG{@~}l<t-mXDi`?bfmddcs+cSIj}P3
z35W>MSp)yef`nNGV2o$X6A&QRmLACZiz?5zv|-arV`=x*)^3iMYAIbRNKa}5=q|_8
z3}-Do6+iQX;Qq_+p4ndy?6dsOnAIoj@(DAiU-dg)h!?D-^qkE&6NOX>2u;o~RW{0V
zNwW1vOy3~pLZtG}gh+g-IC!B!0-zrLdMN^O@nN|L$Gih0u|m5VK8)Q&QRSV|tneX8
zzG8fef;eYqYCe7}<`Nz^KB_>N|MrWyy0+$;;D^hh2aHI(RiiJ?OJJ@hFS%G$n^P})
zDjyXRci$Hrq&9#SnMC!!z4ZxAMMW2%ml7qSikodeHJ{iJ0sXJR*{LV)cvAM!xaMw8
zPQIOd>O8?2A^K0@mVbQ57L<^HqWw5NNkyG1FrC}PB<WBte|WTLer)OtG<P;@Y^%8a
zew=^lCgavTL5LoW^dAI?%TR!k`~=NMJ<)W3^Ndv~lqa1Up%!&Xj6oIyMc;YJdHpbe
zB)-)ZDgIyZKTr7tl+FO4$|m~VDnAjHeJPWfiGwsOlF(htDv?tl>p4J~Or6>3={B5|
z);CaY`T}&px$7?G9>~BM3ZY4XfvIT$LV<xOR6V9V@t&k(UfO#72HjNZPWfXFudFot
zeDeH-EE4KHPR<kwkoKY8NV7Y@AmN??c<J!DWV!}}4c&q2t~lUu+@-|9y5)c{Vae`T
z=G3hmYNc9m_~I{enor*ayBlinoeQC?SNm%ttSa>gD~z<yPSaTG=8Qw24i4}}^X=&R
z^5y>qo^ZAn7SPYEe=A}?<N)WD&EKWWy<*&g(K&F1uJmpxs7tITWrO4hRViHS3}*@B
zTL~_OERXK;g%241Rl3bd3a~!%Rz%(tf7zI|!MUh2*ym!A3Nk{B<Rd&Mq#!rF3(vO$
zMt*8YwF)SFpfBq%K*izQ0-!+X4h2H3c?yN&9W;v1Y71wukVZvUy04rH0lf{=j?jzo
z$T$R0VKPpH)L(0)7|4WtDO}5;!g{^<7wDqh%)Ceyb9v3h$*b1BB0*BPhTu{}ks!ft
z<gQ;6gLEqqswVlpf>jz101z&O&DgqX1?T3m3c`Sn#fzoD^Zar91_8kn7$h7W=`^R@
z=Ma;Rs;jDY1IcjgrRZ*-m-0v)yLZHaC4pZ!aX4PRDeFzc`Nmf)d3yNetA)`dAaLty
ziSVC3%iia-6c~K-sxGbIv19H1ly1xCV$bhg2qJN^Jiqzw4vtVId+PR*W;_YFH(4fA
zCe^3`itKJyX=&*ZaMHD?WAA%XkbE3!*bi{BCyfB2JPg=y%v(>AG#jg~ny;-4M(>GW
zIKTfcsQiSek(ZQjSLCU9oF!Vy<Ay{CPJG#N*?LCkeO^;vU;lR5W`Das66+NcjauG>
z!ZkmczydU@K-LT#oYLO14n#rv*wUOR8<z5<{beWv|9mld$=VOX1c5RXIZ2>zkVJmG
zSX-(x5ODJ;E-T}C*^k$wPsaQ^xzn%Gka<h*zt6$vkbcnHuj3fdZ{2V<Yd-dh@wFYd
zPl#_tdiAhRfyN#3is37l{;SM_Mz<$L+432jN}q1Pmquk8Rfs1&M1IeP3#Dq`Hnior
zB69uil~!Ez`LdPi=_lqmT%E<9GE(8~@A)Lt)#2lV<xbsdp<o)nLjfgb`2C^EzvIv6
z+kw4(%Ry3$^AY4$h5c4%QKs9fp8JNHs%ri_7jLn=ZF={vcp{q~+1OgajM6|xDv+Nt
z%1uCz=Mv&%P1GF%1sZC<>Hp_LDxP9wOW%2d&A{hI?0RF8?`gd+1PSU7KA~!en|$N1
zE!A;*su|1dgxkHtan2q9sMLmK)*db{_ZC1nMbXX!IiP?4tg4HU&#T^NnvmP&crud*
zaCP|ZcOSjM3S-oyx@hX*3Tpx2H80F?j)nrnw>rNI>){ch5u)cy<pU>(Krs!IQc+Rq
z^84ug^Xc8%VK$M*GRaB0pO59C_*A&&Nis$m3N*N{cp~NW`vq`8?@eNew?H$|+8KC-
zOHpxHw;p91TqL)$E%bu8gPR5vUNXe7Bsx4!hix?tJvNW)ZMk&zpmro)3zIqJ^<vHq
z^0_jB%Sd3D>{G1du4fqwmDrbm-%ygnr2E%WNdhIP*cuTiQz$uI)92?l+{A%xU;b7`
zxY+fHON3Z=YsyVRHaKvELu99DyMLSKMyhQfEV1(gUT+LE;*<BpLTT1kRzHDmpQalC
z$&tB-rh!j(rY%U~Z%cQ8NoYZacRE~L(*w}6v=NIElG;G=7q$T!Au*DmHkN;$Cd$6G
zR4hYE7*?i=kt`2`E6T?4gS|<FZ}>9la3>D7Sxsy=fYDdyH?GhgI1=+#5d44r%#GX}
z|7hvq5>|v9WYn;QxudbN;nJ=bK)PWbpeIyfq|CuU>8fFwB=z?_kV+viTKK~LA78A(
zuD3{*FXfo`)*ft9UnHCWSK^@dBR8b>AH!XAfUX-IC^}U|it=BB{>3xY>2_!6d*boo
zV0m?}!xlwFWzN9#uSGxs<A~SR8?FfsPF3^q(=+!j65*rruRicI1Tq?gkkxwb<w01Q
zE7|0GW26m^x`~vtS;kopeAV)-d%BKK<5>hxGU3HLYunykMzpJ#Zd%(M@XcTT1G)_)
z+dtY&th(5ybILah+mIOkj9hGO#Z|U(cNgCP<hQf_;`FBEYI5w5ffv<_T{5+LY~VHU
zX&k%d5#X99(OY+Eg@T|?NN1xD{5Jxv^rT&JF6JOdoLJ>9`g5|b^3e9Pj#xRFQM7aj
z6>4dQz)>B?DF3D_O#B41`a^fFG~WuE?L~B_`(6#K+x*@R{mrV@`e_m%dWx5;tJ(w`
zY1=};VAO+=3`5Sxi@XYG2Xm-os*uaZ4i?8#NA8_uJ#D@_FT@`1mA1d{b+!6;J}Z)L
z*orVED-T{dM8s>eV1PQ(=F0>E6Y%o0Ox_Te!FK3Z-Ar)-2hIzPgqnfPLGyo6_f}z9
zMO)Y?h=PP5(gGhS4FXcqAl=>Y2?EmHT>_$XBi&un-3`*+-Q5jm;@<n5oB!g!Jzn~-
z*0<K2Bi{ENW6V|QFj<Ao<7jDJr;)_e7MofBicDuva6N}S)_l1otC_m^r`G11e6GwB
z#UC<N(f;>Th8o<`cy^IMJ4xTvtzTsR&%GF-uhQzzI)Z+b|BXKQfvKKWZ<H>+9>%SL
z^_-eMrB_3aSl^G1cdnXBF)I6*@qOLL+WCI0iOV}Lt2l(|b3W&Cpm?j^wLf8R5wqG~
z7`9zOKg#~9;|!KpjAHOR0cTJUQ(bq9T|`ge(u=B^G-<6x{M2G+7p7Wb+(TfOg#bDn
zpFIDFm@CH@+B8)|Kdl#skUR6<gSKxKd(*zi=D4S8@Jm?aB)aPQ`mZwO2^_1%$e$6K
zBh-%PG=aig-8XunVG?hd?Yta-8kO`9|A3{JMP@Pk-dHV%g|~lVi5zVEmwW7ZFd^6z
z*(v;_`c=mrE7{7(k8&jL2=EzDOK-&>{p%}xGa_sX7g^a?Pfp~Cu*^q0M8~;>fkS9Y
zy(}3y9G5mQSjxFUgr03;PP<xFQ}ZRNUhPy=z}#6&PsCZRFGMMk+Cidt3o}Tor;bHM
z4)v0lPlo>6KNJTx%{ZgLYZ9F`?|_fto3Ofo_V!JSBz;xad}V%y7z(FdRTxr}1IrLv
z8@E8-b5?8{1h?-O(HH}lDX+;R@j<wU_#m*1bOOJ80Vt$FG=f5}Q!bV-+3uDD=J+9<
zSr*0T$MD<8N1=1#+|c`vOL+2p;)8zvd=?VVIVs2iLHfsZo?dEyZSpjC-BoZsz80aX
zrcOU7Zv0Qua}g{ev&csIQ`g?(!O%h4^fc}zJo38g{!D14*3T`4H#CsQN2Il<C=cWK
z8ZsvRrH)v*V9kniK7k4nc`J)sPktq!2LPvi6mw5za<mg8;CLAl>!;dqO-i(3ZlXGU
z-g}5;|GPf$=YKG)mX0)cb$)4|H^+)_?!@dayYR~G!g#3W0=3p*fNdkCQz*n|UYZ&+
z=3?{eT9lWRVUsVF^MA`h(mMP8PxUZ27P%w%GpA6WoQAXX_rXD~`ikZ8rRLv*g3WCL
z)L<2lCMt#Q5Pu1wmAd3fKE*7s3N5ciq_J7mb1H&TY6CBv>bR;$(}64^LV^e%<K8)m
zc5rU{fqm*Iv5sl7cykp!{a@e#^_zE(l>hT=kU(%B!;O9f{szl9F3!tH>@+lDcSf~-
z*IOS<DD>(<nVTcv3&39yqTT{ep!{DvklIh<QO5F#R>@%0u`;IeZ{82|x34+7T?SvB
z!*;Svj!dhZ@82A^GuBdm%?gQRB?NLefP84J-pc^BNOWv5tuns3C0n<N52!giQMMls
z0R;6NHOocTe+=8A%JK`*8(KiZ|9XGN>@751LjM1*rRg|7tm8^Z-Sv(l><eM0<r*u$
zG0YDe8T^S*g~j4e!`I8M0$0q;9|A5P^lm-=pI73dVzbtksQyh4RuulQ9tg@(x?jR@
zOAFoJ{n&=wabz>pg!Y~M@?sZv{wG6_a&T}=jRL0oWNTP+Qmy6;-vdXd?hLhZG7ms*
z_0dWi+}_&y=8tfxc!Z3~d~k83CzTqM!Vu2^gckrwFxbL@(hT0$aI>R&cwDGU$_kP0
zeu~;o`Np??rz%r1P+5X{Ep4UcoX%WNT34Bch)!NZPmS+CI`zDH6%{?G7@>7$0$>dP
zu7XRj>SYZmgjL~8<i_W48gn8+=GM`(^Rlp6lyfu@QK|}v8@WaQgv)mAsR6$Hn;2pm
zx+1KC{j(&Y*V9CAdon$6EywJZyLzI-f_h^_0^iB(_TyJIzB!jrZal&=1E3y!{q7jb
zrZ`$c-lQiay`;pu&&Wt)%sC<z(13|t4RXV}$;6e8hC1j3A`9lGG(`TVmj7C9Np`>a
zh2#TDhd;n6iZdIwMz##lIm1Ek2<tI>-$DqKR=CY-ny0am=X+v0R5_Fl7~ynTZzibv
zIevuY7VUDPMtV!OhR)Amu4ud@Ej_%+M<_Sj$zA=$e=G~Bypnb$QJ|RnN@Zo?Nk90=
zz}+p*J7aYAlC&kXt|%il-%Klu4$U}l=F_lutR_#cY(jFx?qrS4ROz#IXC&a~U=1S>
zqB6BYY)IAFygGuiqOESQ>2n0lG?rV&7>}pJD|Ry;#BUaGZZO9uC;JY(pK|Hf0vFlF
zoi6dFpTRp$=g*>4#9nZ4;}TSuB_c3o_Gi{!*B?BH3;XLislonB;<{tr=uz&=Ic6BD
zeLt2=T-CX{(<)g$F*9$q6t3;QAD)!j->9cI{oh~rdGGJ<11XhI=MF~NMG-Ki@2kQU
zjvuGV=-$mKwyN#($xV#-cDFh&v|sv?Zk=8OZv`TVCE5sbHp!=Xb3?<F)r{yV$$cQ<
z&#(A%@wDMPRjd8l0>;$GO?j^JrfQ@!{;qLM3}8o~@_30v1qTQB_KMGNV0rE=e-#d0
z6|RCbdGZG!)TM({M!)`GwvOLe7JutFx`@Nk!e+nsyYpX+gF#b&)ga9ArZnH(wk@d>
z*7w{sFJ4QDw9`anjV~ar67Ew&s*lF*N~x<Wr^Ul<YZ039|Dc;)CXvRj$RqV?*gkuT
zU00Cx?0lHhP+E@Jde&&>_|0}yeO*uZrBD(@I@!NS*zBDZ<8C@4QJwEzNjPMEr(+cZ
z#f?Q!=wqWvtBi`z61UW4a2Uc#YYMM{L=q|dzZVhg@<+Q5bEha(yxfjT8y=xy+lQ1>
z8fz?^(UoX@l#%&6t1~`;sL47!GLl9Tit$g}p_VPouHVVN_nor;oK7)D2q-c4Z*nrb
zct{1U1;=!US@in~r#^e>?sM<u{|OQ3ww`*KM1+dxEJ78hp2`!XdmlEHe|8XN4|}`R
zk_bCE{k<ga<H-~+%j$1U#4so1ZqP#K=J{tF>rV=gV3(@*;Pe2TdVIO(``n)rw4>^1
z`ihpyf@Wn>6j=Te?ACYnnVceIGUROkT@UI^ElWUHIaAhwtbS5t_gbCn_x5_h&zUSK
zi-;SVUtu%*69)w6&<ar7{AcGsKHS?rhjN0ao1KXI3XpZVqio>J028mVi-A@8)C+;V
z;&zP&8|S|`N<aQ1KFO-xgu&V&Kshfwqo36B06C+#+@h`FX_fs@juN_JwY}oCbs)AM
z@?EcHKWGBY5NISR1eRuU+nb?q>b}*U{0aA^g`yD%DiX95l7TZ>-)(_Jy>HIblvJg?
zX?V!WcbQQI0gFx=iJ;bOCnWS@*X)~la|IKc;XXm~ItC~)%+%}C&ZzIM{0*8}kng4M
z!^WCxoi4F?zf&ie>}84n1kR7wT7rX7^(@kHcSJ0Zir~F^8K`?1Sqg#|Wm|vA&60x;
z4vMN!G|?LKsE_Io4*V*AB+jK?Bb5iM@_HA7^0HbPvh}9lTPh)RnwxFf=32Qm7Pq1y
zS5T1}E90S4KZ;M3?`>AFKf;UcXI7T~fX+`jG*Hr~ZU_RG2^MLoM$G19rjAU=4;19=
zbIWI3n4@V;kTwptc!ydO31j%>o}Z_qq5dN+3~PL}u#<=fEqU1P@%eK8_!acVE)5d(
z1HYB-Z8ARL`=Cq)wqc$*$+|nNRir&h8rP6O$|a5#Zc$E-hVs~gu~osLee;3Hoc%BI
zibl?4TEinsmL=Ldpp;)cX)^!(3kd`h^FLxp?9;nVu!cu_LP)I}f{f-7lRpWX72EyV
z#V*U^h^w8{@7<(^=Y&(=`5)e3>-Fx0D_U+b_8<wsm!eXkK~zhIlE}^SD%-iM%+<2h
zQH_tws^fnO8=q_DzyKnbN-PaEHHbc+ZMhB2n_0L-;|Gd273N*#wb5t@YeAMzMBhI2
z>kU3DjV@mizEQV3o5G&Z5<8a~ox~f0M|%S*b2I;7gXb^cmP>CT8B&uK$hY+gn5=nk
zD*7En%xu)fn;MdIfq>lOM%p5o+cIiTCGV0PQ#}3^hA<I!|Co56ZvWr=B1#AgcWPKP
zoFBEJ(Q8!x{fVHq-6hN;_FcKSSVk2M$-i3emE3GX^hgL7ZpyIIi<QHTn+l#1QBJrS
zP|mRVCtb8M+HCwaNOIcP$uG{r=in0c!<l{R38>umC+GQ+cg0wrCpYWlB{y4eHmfw!
zF3u~jy`_g^emuzpI$TA?uYYpD&1~~sm*pLvBBDJDCFXj#T19sXeF1f8(7c(f^fYk{
zHH%yI{Xmoq&6$wn$EzrmzwyOY$W$-&B*EIaCMnu-N}|Ae`^crb%$J(|A0R0^?_K9`
z_^=W{-`m?8+4HVsB!tKO6D>ka&D(YfN16;_(dD$tm<<9AC&3T*_A#IQXzuyH4Mzr?
z%54ir6z$f+BfO(MPf2<E1hHOs$x>CZ{+yeG?nCzG>l}axanSVmFft;)KDf+xt8GOW
z)aBlLW}T1p%%F7$4b(t-+-PjxMiZDz(QW-KZyNFEGSBv-FJL{Z96@ON$yorJF23{z
z8s1?8{Q&jqH2IP1-OtCXkjJ5e7P)lc7mgT^!1t!8Ys+z{dGpE=+B3J*oKs4?Hx9{)
zH)d4h7!?hfkuvtbSZ7d*Lnc{Y8hInE6XqPJ13(u_Ea>d^8WKpSIDo0dVrs)e$1#>5
zKUUo}f1#_C-I8Ng(j*TX^OTkTI?33f@XxIBy$k0e_z!Zo9fOrZtu{9`U-$w|j-0H_
zZBNMAe?<lg*gSn*Rq>b6KXy@Y?a+E1wyfM}fMPXw+DlWMkh0{B#^Z^QSh;Q4&o(E;
z@1JiZCM{t(w`I>tBscacKu4l@ZC<K=TSoJiI0MMZ%xla466i<u_yaIAiBCaox;}hi
z=hV+9H|y#!_9>fIK7Lm9FX7q!oK}zE`&a9gqu%^7PHg{#fpVKs*nT$Wxdxp%L-s1J
zs(Tj|?~^yF?~UsW&<^r2i%|~FPZqFEF5^2xx#PJhb2k3-TSGsjcFjBGCdmt32Uja&
z972)QIJS=EX1QHHPIp+)fy_j^VE0y_W*<;09<(O~t=SvfY^Fe;V6C`6bETPl+&hE6
z(){~c4ui2vUA*qAS0*!iQ!Z<mH&bjwK;ingf-q=tbw;D{R*o7zXX2U%weiYxzvYID
z$c*FN$gr$6Bx(qn+zxfC`9#-$|Dj^f<XFgE`tNn{791n#<<OfFeS!`0($ebT3OXw+
z*U65S(rE63%22u4mOdI~@vs@Y{mKJ@PcM6TxF+njNABOr{Ws%$gxROvUn8j>?`IQ#
z>yrqfGB4M#Tx+36W2GJ3{KozxaEx3fE852C-2PC2gTc_utgxV<AV0sXNFpW-?N@>y
z+*75cCDhL<irHu)u+4vjo`5!?*_qbNh6bMNG#Jo<jFGI5nb4hoiOb9x15O&nkafh&
zJ9TctO;i<v`Y$85r4n$b-+#vPDaG^am-%5_TbMa#ER?eK<BhqR`Rz%Y6l|CYuZS^@
z<?KA@u~^?k+0ku<7pS4V4-I{8Vq#M9*Xket!Wi;?uH*~aV@z{h6mLHEj+NWxn#G8q
z_G71yE3f~OK=!_+GoxBpk+_diXOl<@Imy!sp@0|WuPJ;8m9;xOcMkFxcw$sf)oKa$
zh?v^4Svil8EjPBURY0Bb#o1~iQnp&&Yt|&KY>Mu4(e=00y_IyoU1L}ll|fmX@Rz;~
zl@$|>3#R;M2arICr=(?@fW}6;BN6e5&;g@KBaXI>QAnZBF?U@s<`%?`zWy!ND9avS
zXmp4L6N#`C1S2akC6ISqo502)T@CWBtr9-F3={yGy-T)QEMX_#9pV|qm2a1yu9(J-
z3#~YR)*Gwj`o-K(uLPBUCsee5VX4>ozGA!Gb3b+o1CE3fswuO~jD{3<YWYMkLE-po
z2)M~FW~mt@q}^*p*7z5_J#w>n{?QnUZ;HB&=2ZV0DDD0lL&e3YOna^$Au2T;1W4de
z^uLWp>r6v{dsbvev@E5C0zrSZ|4hk)2Dv&hDLK7zlm;_w24cP)7ChDpFQfY$=MXH*
zPid6E&-jdu_=QLRjfVjJCm;ZXlmGSG*<UtxcLrvc!rjfBeddtTrf^*0IHmFBLnF+|
zA0sk}%7~y-pY1<B1qozFi^H-k{O&BlDREqR^ugTLL5SjX>EfW{TxUZvX-X?Mk8F>!
zsF6WCAf-xdkvL-<g4z17q}S|SnYv18_9a+P^eiYT)^cvc_d{;+L9~3ZkXg39i=|!a
zl|V_;o5$0ZSbL!iDYSi`ePOrpoJ`OTNUvvnNAPd)vjpUK8T@>r+zXmZ=(^-P>gCjA
z<VJ>+ZVE%wDyyOc-pp-alT=(fUC&e|CqwpQLD6AnXD27do%=wT@Gx=wN#-d37rmOj
zl|OeShq9AazJOl-ypT0#8hl;5D%{k)|5a`_R`Yl^()vKu%;4y^C8mfKORKOI$$Zd-
z<Nvz@>d%osqJIFrW_P&eHZ4&q^qq!-yd%d-JXNVak69URM$&x)E5af4XTL4W!L{e=
zf_1~6(2(Aq{M&!NfUn;AkPyFzAo1o3;mFjvPqMS*-xTfL?YE{j{o<Oq=FU2nqa#a$
z>uZyp=OTjVy}o1?{b)JLag^Pn`j1hUr=djJeCY3gSMoAoB*WWISR=Q%c#zWM=A^r1
z(59NK1{#i;J3ltMmXZ^=BJ6)52Ym{rok%TDHry$9@d-(DSYHSAm))Oe`{*RxZrxTa
zT7Ge=%O356vc2F$NMjD4LMl#_9F&k=xzD^D6alVO-#sN6SyuFQ)wpm09V+rHv3yo;
z(@ozqYHfGvuJ6#2$FCZs6fZx8ti(0qPj~3ldf!7L3vrIvfe!15K*HU+i^{~1B%mXp
zZ9*EU!NagsQZ>A+{;>PZTuM&FYVAI153hrz5me2IQ?^spTE{2J*_LcxK4~1rT&?kM
z^mRV^wZnsSGU1@JWPwu|ZJ!ls4_UQ45cVT#m^RP2RM#36kl@Ds@9z~--j7A)ar(7?
za+a)=Gjh%en^8{fX=W=CP^bBEH`OhL=Ejq3uB5iZx4!xMNrFS+ss9bJB!r2Ampeq9
zB56sTn~8_ac&FRzRtI#@zCk@p8EPQ*y?7T)3e%p9kQ@EW@M`#hK!W1z`5wAHx5x`9
zr!lKB0mC0d%*kqmY1A`EE>x}1XD~y{e0&<HzpxsD-oLNqhOJ1z1x0^l6_u_V%A)Pe
zWLKROxt24UEU;)Re3h5)``ch)*!&*uZ|T*`MTKly5Qy-d@r48h*-4et19_v+0!!X(
z>TE-ytGcaE*KF}+yP?9u%!rAJ;qka$C9vBlqH3AK*Y*GVIinGEX_WV7MmVf>!*dMU
zHX>I2=O8~4Kwy?E{r&yv8n0HiN3#)#ny1l|9{zGeLfRB+m^VUI<!+3)7=GqtekuP*
z7RkA!Q?l**N6kPd%O%tem9ViL3+DBkrbUlYDb{(H!-g{osHSG4cH57o1>q8~y*k?l
z8QZCy(e`qJmy|_`ph$t|Q~p#uo$?*fC2~D#QIO%bYxaa4&5}Y^yXgQ%Sd*DdC55(X
z#43EOQ*tZ^1;r*u-v{ade}BcIRIpm^m-7xHA&^%*H=t@|lP(<8+u#4{vaSSJVQ(ho
zZX9HW*Wcv8+_t+z$GZ+4AhL&?+}xVAa0zLS?F<ZQww85kP#40p97^GHSa90x4kI!6
z`y1WWPbdKU6(NT$=ozLKLE9X3ALI7N57jlA?+dJ%$F<a13LU=`wkH8N>27#trq<=s
zylJs`rKLo-Yap37+1!yUiO20`ZGGL>c0Kf+Ks2pJ*870pKW&D2?r?|BhLQ{nKSs_%
z46BB&qttX;C-y-EdJFDV&ySFA3Hd-tnPM!5>L1JHbTVtvM0mo7O2p$jz`pr6j)i7U
zS?B$LkQ=5-tS(+*BvaRcu->SEGBM|M&;%<kD!_=Ikhw;;+n?wpYZlou$q#p@gMKav
zU_(Ji>&bB!kqUb~+ou5`ayCEGv@E}irO^4Dk+Y9<2yGy%>Y#IA&lCEWmow8s=1TSY
zrVWy5Jvenc11XXm`R|)gKr875DAFp4y{Y8zYi9!;?CwokjXRq@o+cvu!a-7Kx;mJw
zX*JaV%$2+(EUolOD=n1|KqxCMb-p~H)mSR&g1DTItGGmVBh%MgPV2=)MMnD6$=I$!
z#~fnn)A~10*?A*pD<}MK)O9*qzOsQ8d7rYv$DslF__Z00_pUkEXfx4a;m0fda;}ZN
z=`xXEg4o#DTpa>Gf>(;xAHc6qj!gJ>#zQ*W&BxgM0b+g~<9fC1=ka*E*yBy#ba%pe
zSl>psFUNOzHY!ubS${EO0@_dRZ+455T~B%>=o=JoKh&+Qt+jX~6l*nUW?BcXkdX$l
zM-QM0=+_OmgSpT+aPO~<A?Z)=Xw_4oR;Ew=RhZbRP_0ZDd|M)(^)O5Xat$Cq6m)<O
zdmDWF7KqERvKqpx%)T~*Cy-aXovJHh6;(IvNLe*5g_hWa-xjn&oedXzq6XlRC?%9`
zk}L-LhA~pZ>8ONRbHp)NPqMvc>*abS#}Wt3Cd!X|3ICv#9Yg1*(8mwDeJmQUHx?fs
zE~zn2CDGPDad$D2J3j-866*BX5QG4P+TWH@K6VuF;BgUOCVrKP|73P|b*x#;I_D-S
z7E#IEHL@WGb@MSX+DtI-Ml~(O<svuTrlbQaufyN93|=qjEz1sYisgfJ`pDT4&)p?6
zHEw^CUNt{g6bF6QZxnl-kS?D>{Enx`|AtMk$z{7Wv_D&wpAc(8z{*rrB`3&MU-5pQ
z+xbUSR8%=dB8KID=sfOlhfaX76)?Fs31eep?*x8)_)^O}M>u998p$3lN1$QzRi@c<
z0LS!-A!!8~n1H(l!Kl>vNB}}{ck0~oKslWLpZqmdVRVXeWkn(k0D@M@2Rgccd4(}(
zkBU6q7e-D^&1qaoh5KSkm)b8z9WmYoM9T7!3=(5!Q5Sli6=-=f?w>Fjn=T(If~SXv
zIq}j14Feby(f~V0c;JTWie3KlCU{deG=KI-skNtW_$y@<H%>tR+kb!E+ctT&JxX7H
z)QY3!Y6X_GxbepLXi0N#BlI`fONBT{DK*uAefs@Vg!^{^EZ6y%LW9d?5~qWC^>W=*
z?>mMUz?#$vd(8&T13{=;{jW_`n$41fMgj9Vw!CeA%@dckFp(@<fnMaRMHZ6@6bP5&
ze61aIRQl;E6eUY+ch?Nl2=?GkEaRO+dS{kjWsolJzu)`G%LS1EALNXo)v&)hT0UOw
zYC7)3Z@Jta&!0S5?^7V%GH-F1s<K!Ffr95|iv&xGjn<D8moUA$mp}qPI=V|x=5r;!
z85_gTUUQJ5fAT&&TsFxhamCI*9Mm0(8npO*I7yk^SN2F)Sy>rQ7ZG8NoMlrNwkq5*
z!`x1tDv1!C{>5izW(Jyk+T02)SV4?$Cqm=70yV06a+yTSR>T&<xNh&NMfckbv52iC
z>WBk?#>_!GgUdlp*MO|~C?c!BiJe9IPfJCUj!+8tV~(fj_Bvo_nUV$56eaX8i@^ZF
z^#F)WI`;PV%_Q}w1DsS;R3;Mzu^+Wt3#%6LtSw2@xbDwJRkYkH##85%iKf(0e>H-b
z_xil1nmY-%ef?jA`Cz@)9X87^r>ZLYD6x)jB>f3<+<Sd}eHoIMm9_DB-}I;=E1S~T
zF<oY$f!=FoAST)+9MZLVd7H=;YT{wiAsP?aF)1`;LYFcE#lxSxny$wk6v?jP^E{SI
z&AHi0dDdg^`ObF6cfoML-m$hy5tQ9@>W|rMWv(Jd@LN{oSl#VAIjD<$I`n4S#5t#=
zLdSN{mH9vzB)TgSCMKJuX0Oc5Oxs&4#EaAG)6K)F5^wWJ@GmL~apdD9h{#pTZGI+P
zS^F~;Chb4bC89qdIv{wnlj;+~jN5$B1&AYUWH2(z&#DX`_{B;%>!NF!R+BqvwMlV9
zWoc>n{ZUcff%D`2<%`@d!N-St(CRt|v=tigsrtbr?ir$*l$51bWco2kubItuHpAcc
z<@D<Bxi(ho3&L50;uRj4lXLUkS#VU|&w=Xr{N4*e215x#wB>Q0IC$&Y2--p(hDC^!
z*mGveUN?&dXO0S638cATs<4xr8jZH8&IO^fa6gCfv<!DUYC-V0-_^=bkiZYLk)s!o
z>?tW{iqv4GeG`_1`zMmgB`FmHzO7<U7-ZVw?RQ`pC7%uSiLf}IZLO^6V10?Jc-@>v
z&<C=t98pJ*t>jc1nm3a}`boQ|R}&P*v*vF+tHsnG?K=B?B(OlKa~Er>9P9%^3uv%y
zA#zmrIy$WI3AmIJya3qNg74|!Cj(2&RmfgBd<*j${b?p+sl^-YUaqw?_)bAa&YG$^
zuuvT%1uZT5u6njId}~bLf&m?8+c(nR-85unkHCA+TXhk--Ci`TM{6tqICOXCdM5`5
z^L`!N>Pb-p`9l$vhkGLwOze<feDZ(URjC}=<<sL02r_E&Ov??+Y$ZD(P5=`lBRmSO
zC4f3wPU}$uKTu*nna$)W<$pB6g!%fqYL?r4t~xu}og<-@f8c9~bF<x_fJ7Tas=42m
zl7TDgGwO1S9V<X6Ba?`hjAisCbp)vwd<o*Oq~D<J3z&9nTiIIRJ%`n+s8~^d2;Km=
zitu`#*(gZ2H5h4Kc{*rAx^K*?<oAJaX*`Pwc9gp}0vh>c7}C8v?}Iul3^w0)myKA1
z&R|0P1Vd`SSVbkJ@renJYp%I8Hk@;kdVpfBB&SN!t6@oC;2qK%!4lzVR2Yw!8;!!F
z5{SYBT&l0Gt}aTYz#bbGR${r_3iJ!h5M?OzuE?i%%n$}fmeJ*6k7<xA0`&tJ&mnR<
z3+q}%R4pUeWhW(%b{FpS0eN(q(kpmAXH}=9^o$cYeeN09ZHU{&n8)r9eDmt1@5nid
zuDAPD0t=}rDPSB1_VHZn@f_%fBS8?1ZCDsjK7&<?MB9-RKJO&w9jfF(oKow+Rq*`b
z)9nn)s@sS;dYCGUzxbr2f}l4+5bikfrpQRlea`5HrU(id)Xf$gO-C=Dw!I7Rn5O`9
ztrp{Z`g}L8eSwUf^8xJ)oKqD;CnYV7nne59&$@_$|7i~l5-G4Bu=D!0M>21Vg6yQO
zGQ;Jc59d6-Y@Z&C0q5MKc+&v3Wm3z-Et$uSTq;g5Gg;iG1R{651TX?Wk%herNChDM
zFrXrXb|)axKB0V~$E&W>LEcA?hfR;~hoOmyZa3T6yf<6vAhQv0+M6DkUVs1o9n1`U
zy*b+k+Z&7t+5{`tRcAi>n-Eksw{$;+ZQ3YzFy43ua)WTtTcKnc>U+spj{KDtKs^o@
z8sg^{>Ks77nVZ!&7-GH@j-Lcw!7%1kbt=W0Et|6n+0t&;>oL>U*FK0CJYXx9x{j}1
zn{LL7vm|5F7#TP2d|_rptOP!Bmm_ReDQ9GFo;w>23fk&eUL9#u-<vm_Pq^RT3^$b-
z4C-@tND6j%!y%wbN=WEdjgF2Ypx}CX!@(f(8yiz&d7o6V&RMqcK3)%c-hymw03K=Y
z0=OBg(MX1+NRuDL`KO?of&zxQ%|OUgP}ncoPBk;2m)=DL2)=oM(wU2!I2!=HB=IKD
zZRRb`Tg$hoZs3H4rfEN|r|45U3;OgxQyViXDh%mMuz?LbdHHJ=V&7n`7;Kw)mwhnS
zt{t`9&d0G-psCg$GqiXg?@pEY^Gyfd4LgXc1h3=)h*Gs--AXDSmE~8yxaIc_p$!n1
zoH;Q6MGY51z^>C)_Uee3yW#%!vTT+~|F0uR_EY;tqgf4SBNA+EFrmaBH0!T`?jGHg
zDbNA}c2>Or0G3XF($c=W)mg9iGU#^Z@NR(SNgmUwk4}+PxcN4slBv@CfbF{coPf^#
zjDvtGbK^-bO+m1<m{=?zB>>dB9n@^fZ;PDAm{C81!hl#caOkWE=GqdwUFDl)hebDs
zo!snF&d8XU5L^bGov~biy+6)>WW+~+d95)7>lG|<Qf2!S43e2QyE&il*c~nL+25vz
z@YLR4%vNyiJrN#yts;4m-+=KU0X%WTO3q)iKOJ<#p+x{X@<2A>S`B29=G-p_=>aT0
zh(x!bnI)bIY3fQCSf7!4Bbr;C;CJ(-ZZJ0ZnG7xQ4H36<q<O^et*;%zIa{A8DBdhO
zZ6+Qz-qO{%fvD(L*ZKw~_q3pFKpt`Z6PaH-<`W$t8|VOJ<R%U>B26;+H-Y^$T!Ib{
zbSIL8aPcSOov=KBk<Ed%^%$*2u-cEuAs)b(UR^GF@ZK)pU7xP!KI63C+nOw1j0YgQ
zi}%3_WY_WW@wY^=1GUA~EB@I@&f@Eoh4kRpm^V~8s3SG>SmW3%M6|Xyt!p;=<5~53
zBZ0#_MQmFnK#^cROMVu~%gD%3(R9*VZuKn<KRG$s@;h-WEiDDZc}=oPlv<QyFMn--
zeEhVMx%!9>M4Y1;=j*DS;t2sQa!@BpRhMR##N~|4Ou6lBlxpQ|Gevdvn>m|7sW|5C
zhRX$HC;Kx%W_agcCB`ogV!*sc7+O968{Tkt(x<cm$yp2it#pHD>YS5Vp))qj;7=2R
zIGR7|uCsLR4AS`y>?G`AD_Gad<j3n|jVkkojX0C3v@saE+O7A_<xX$#S<Ty%(O`(p
ze{;KB;29*OhF=u_eaBF8##FMc4S0$nK}S@V&-n`|+FRzIFQ6*Q%Elh%Kz+zyx4&)n
zg1f&tN9MUs<;=*=UOjvP1Gw4!`6MvCg?gtl@pVJjtKpq2Kj5hV`Q=+kR@c_Xy?`og
zQy@Th^K+k85uhshhf0}bo|a9awHDyrTj161K_Dea;&~1O!;zwJ51^2rq<%!mILCy7
z>RU^~K)6-tcZYO>-oo^4?QTz#s#&umwI@yhoGrA8do#hN(scTbCDF+TyrT#ip47b4
zjd%-tM=nk<`I_?Q&k-=y1Rgi%YJmiCR58otwk#HPLU31s6{w=1aA{FaPEKLrI9R{a
zYgMo@`M>ck`yC&yH@6wqE&NdMPVO%k{py&U>OruiHUe|uuj_`6FF-!}&8X+!dnZh-
z$q(^L$F(H%gwvpb4?#4+>v)A)O-Dmhwz55}?0!4DSUP(~`naz3Ak=VrD}?hUBcUAJ
z<@h;aLbi0G^Ss?;Zh^3A)Jw{Hg7NZJ#6m0=0t{4ke;NcRh%F;OJWZS^QydN#{<=No
z23~ir+iTF$tTr#QiFFc(Of4kCzK&@p2L_oXN&KugfmTrmM~F*9LsJ35*0GWxun|Ei
zG>^Muk?>%UZ&9lU`ui(RjUGYsF(H@Z(fOnn?^Dpx+yb6|^OtFGC<uhH2+FP?d{)ro
z{G?K;{(<oz2la1*hr4S5KU5`63Z$1zsowRclYeFvGI#JX^kFA4$R*jfGa^*CtBl9<
zD1UT>1$$gABdmwAA$}=RHImWS(*r6@_i|4W7%+bi*up-@(4ft`2yl)?`&qLV5l7%6
zcDi3ogxqvn6>ossS&U?eSvQ{t6LKDD!@<B9l-lh+Wt98drpLNo(vch#USXeKsdJzs
zUabAY(ha}>JhyvgYTMc2hG4S4SrDv|15hc@a6c}wHeE|~Z@-e^J|Byk@Ag_?goJDY
z*q|lmORZ8zz#>LY3Q8PNG<@ZQ>e`u$kyHv>Q2qn6490NRt)Wy9Hk<&pP^1Ni1HdnO
zRX(gm$czwfCCc}3gX!lKt_KiYS4L%cGuy5L@tEAt^R-?qZf3=sJ0CV&GLiwwhv4MM
z1q8RJ*F;5JpV%)Ir8lGoY~y7$gz;N9fKNGR3W(LpJPn|#*Z>;dd6Ex1aJ1?H&+|TH
za=SU>Jg9ElJulLzDgr(e_9Gt-=JPXt1WAtCUJlOnNQLhd@52!yA|hP4#Cd^q24o&N
zwbqThBu{nz8l^Z$o~)W$E=u*p@T}qPM4{qViI$PM`N4QT&Mq63fJ_!JX~p?6i(LJT
zXBf$r0HgWs1c215@?d|qHEc3j^y*^i8sw$p4xB~MVLm-QO&!w7Uvqi5SS<gUy`F+t
zq5~1KIzDywacS-|YIU>={`I}Po6H`7F9^9(4;LqW^jZzh6(^`dLPAYf9rSAwC4foX
z08{gqVS|CmVY1nfTpvCGh7ix|;SRj+XsIP3ApuNfC%E5igAV<5*og4(t>48XPF-GB
zbS<HKInLuamS2>VI>ej2;1G-t&BjOoX9vkm0kqMBMH}R~m2^VLdk!{R@p=|e)g=Il
z+bIZOV4MpOo)jdS-4?*WbLNtJLxDJ$%e#@p?eYVS=%I%q*{9B%QEJBPkLNBmB4I<&
zmZI1BrxQmma-MvUPP9;2NdTPy_D#J`h=ha`c}~n7%X_zscI5*H<H^y}(}M%Ggsqz&
z85zk}q0hTjJn-menREqVgqz1yPHy<g?;*YWIZ2-H;vpAo{u%($%*}SE;Z)e~&j5fJ
zGS6ebrwSnAv(97<4UO&n2nxoq05RYH+}>6Y_|H~2nM6+TRiD-af1EC_R*6oBACbE#
zD`4Cc#ae@JM!+Py&5TiY4U~>H1gfFovKqwtG#9XwIybzJIHfx(@X0?vVGyshO3ult
zW?UvJL-KG8s0eCipzlfb;uJX8@NKL4<80GC{<!B^0ucPRYOt`eS1?dvV7N<xHT2Qd
z6@W-&`*m<|(17<26qPj{mtQEKKpp1oY-KY!C4Ya^o~<-P7rR*YBWeV)oDzSH2I-z_
zL)a&7Fc1C~#G4o`_a0~bAY6u4{{><7sS;fTv$%zyeL&S4N)_1iTWbgWI$=KR-gVZj
zYB*j<#u230D^p<eW7?^6PqHD17)rKAZKU#UQfO)x=|B=mc3k-dU~vMsi+%TtJ^PVW
zJEbm7M{pp7k6z377a(HI*Vq_TbE<-dO8O=B12fxBJcFr;S#T!DXIEfdzzIXa&&y}}
zGtfk8I5gXRwn+E%-Fcg;l1fGd5B?}E48ZS;X@h)89p*Q^{o>%f10k1Dfr(i|fW6P!
z!JGvY+p^FlG&*kv?1v*@`hdjUz@c%yE?AnIn}bCv|6BibM2C!`q9PO|KpC!MAGOpu
z;b%or4MB`dOhHW?&VQ{|JHcrhBTOK`-+?6{u=E478)*2SP>4YA6a6NtN}95Gx4{Zp
zSgG~m#3`#97lh9|3#x_zhhVhbk_8WNLSPGk4{P@}EHd_oP@L4~jzi0893Vy&{P=qg
zxJ(Ehld#@_o+YKj=O&k+PgXV1f+{(n6o5o;VgMo>I1YiZt^w>a(D9zT9JOP}WV|m2
z@d1tN=(Ca%77s{Sp@06!9uRrlG0YLFCIrmQin@Tgtg0~37S%f!b#`_Z%Xhae<Dv5?
zVV}<VXBM1t0CK%`bFsp8&&GnHvho6;(M`AWher(Hhz~*i^PTav+*kOYp5H+a%+@Y%
za}7t>wmWvusa^16K?U*~c(%hA#}BuMWsls4^`{`kx0+kbR_=o&rRYRSF76ib9lTxp
zE0@ayLO_|G3@0q?7yzdyaYf2G>_sZ1w-BWo!ff~&xjM_3hPrq?$>qQac*WE{K*yIr
z0aLCy1+KlAEkgvQ0oX%UQE{xv;}HZFP&Ns8SBzVFj+r5PO#$#?3+$v^aUu93r3SUd
zwv#*|2T*^&FiJ`ClcWH6Z7M#{ti;5{dj|%x9%}&-1-b)!{VmvbqEK&C0=B0_JKD5-
zuzK#ydw0}Uwct!cPfx$pWNcz$QM;W5;yyi;l6BlI=8OF_fPfNBMJUzwb@Rb9XaP9%
zNk9v;oBUGswdL8hpT|`j(Ukdgz0)c126iAB0OJSaHn-h5fd5u|q1a8R836xUX?nbE
z0z5e{&lo6&GCbF7$q#4QAcMY;33*2qN?+7)G2^Kg$&(xcfF2=7<k{tzf{fbalUu+H
zA>^-$$}e+W0T?Cyq!zmGS4=`2m%Ns4fdPO#wZ5@Y>S@BhZhEmdt*56qH3%+B2=qsI
zdKD0X1Lvh>WQf=-zg8o}@_IbbdtCPcw`~HelAD`*<BShxn}GC{Gxv_3-r{tl-x3`8
z0^`Ev2x25q;D5~9*FY%}EWeBE1cp_(XAGnY_$Aqnh>-{fzH~?iQlkrye=dOz10aNA
zS`#d|wv5XWWT;}zhPZjU4<GJ2=-su4KcMG54*?86Jp9eTlMvJ3e6Dl<x(RIHoA*VZ
z!09)CX1tpo&Wkq(68p@v0Th3yzs;1I_b~whhYm_f;qd@!3p+vRPq&|cI*MYp$%9wI
z_ImO0esR{kF6jO)Ge$G+lbL!s*HtIJ1rTI}9xn3oi_|Mk@fHB?gb`LPHyj26CI=8E
z%LX_C%nP73;qHzAQGjw)YE6w=sPfAsa6}2n*@FN+Jn{R@wY9B<iG%q%4-BN_!R!w2
zVz5X5G8={D?-c0n(}_h-c=n?ZY<&Lo*w`7+AiBd0U%h&z=yVUN6$LptA4?0t;X7lp
zlm^jxj+O%n?4(lHvrjSsSY}5ckHfTMP($XLJUjr&o;Do^oCV<J8Zdg}0^IwvD_{@?
z0}0nd0%+$4^YP|);BAR+uYWV>t{uzdA$VNYEdrjb0>GXlkc$3{g6C;}#lpgJNC{pR
z{q`EzB)H?${>H`zF!0@96iL6C3{W|Uo)+={2mjLg9ub_eIVS&4=XWu({D0?n{r8+j
zrcX@l|5x80=!c4he5iWWI9R;3THn_FK9?mH<>NfJ*%T}1;n(I1s$X!t?Vnd99^7*(
zNXJKL0txbBsDGtZZ<A!2*W7MOA2hnpc(~u}2pSvsj()YeypAC}g6k<JzRnuvL?%XQ
zIXJbbBhtzgY@1A?CnH1Gist2XpR}NV9<0&KsdaRrgSB?y;|lA+<8zq#Wv_5l!8HW4
zyGg${ym)Bjd&cL8*%qGqru)0=S&G={nAZ3shWemI9v<HNO{0XsVM-VWkK28!IhE(%
z2@XPuoEvWMiR!eH8u5QshFdp>hj$vziy_q$3vlXQadU}>k7QTj_@<0>HD8SJA%`Z_
zSbcYlob%!t5sY$mVPJe2q8D9ki6DWGF)PUT;lx=w6JCz;BuWmuQD@vxIdi6LsKVvN
zOqqK5CMVq-qQBVU)tAl6Igwb`!ANEQ@$6V5J&epVsPDyLrWxbw!oMbr{-4d#JhNH<
zXxIO3;mqW@QKf{x;gHptru<T!w@S$2&SUK&$uuszi6PuEBjI3^DT<zYWPvn)cMtfl
z9D-J&YDN-^r*OrHK0B#V1ZJm@Ht*xD!q#wXR=B2TNl9!O&(uZ;Z{Pgl^>S5#>WTT~
z;?!AH;+q9u^T(O{sC{PjjEX)1gSeoy<nwGP^A;rU1~ab5cJp|bLKL6i{(?gvJp$++
zA1Q%fL{VRICBi$B^2;{LZ`Rp*8Y*vnl$mF0zh#qRoD;JAu6Vn5-k>g7LK*MYwthHn
zoiY<Qhtu?HV${j$_){o9yAs>WRbisxo?eaOp%#OYFA*B>2a2T^hKUTeJ32vKV=9o|
zh=E&qwjx-EC6*A(uJx;DH4X&l<?W++lYGZL?{l~Ii`MzC_q?1}$}b33{reu8EHcrZ
zMH01**|%x`@(M?iZJUI&ifY6Z6KI6HFIIlQnzDvYK3-4qdN{vJ+oX@rLRj;)V@BnE
zfqTzUQXG)8fdB6I2sL$zPqJHsMr2N$*oRxxlbIh`$uCUdpOMlAUbw|1J1Q;MnkWia
zTcv2Jsrx5sjn2mGYttghlymI$o+Z2)x*U$L<=gbN{kTI;CxZD+;?RNvPc+nCwLXGj
z^96cde6iT~ZU#YFZiPd=1A<VWsx=N6K5=9zK1X5j6{Cg9uoAMD@@&%I)JM<$uKi%E
zAXhjMPwCRjEuys}F0f8NAvN!^6OLZcOy}+p4bWaU9U=EUdHq|*lDb6$CsbaDIAD^Q
z@7qZ}-s=%E#pch$nVvmKwJdFKeFy|#f`sVQ&2Qd(4%J_z5`=ur;u!9A?9MA=#zz}y
zr_oGtbTe}~m+0B}O$~!kZy#LGMFU^>rI?<pYm-f`uSC~IR)Am4W?v95MFX+ySFL_f
z5Y5so@(t=calG$7y;twDNYhI)Afrg?&K>7@&YJ45HnT!)z015oqEyVe)%v~_eb7pK
z^x`iw=S*%7Q>c{feTlKi6n>#%#4mW(#Oe(;l%Dl}@b8x+Ktu^kBNA#%-p=$5*Wbf>
z&hpDzoPoM3>sZkpiHPsd=#GYxqQ2dh5jJkP0mUf_6I*z;6c%N%FIm0t3-!0}U=66d
zx3XJbj5i~c1v&>}bWXzHg=&>9L~6s=W75{Fv^ri!TKD;8lk-;WD`+ImkgiO%(pFkt
z=gxHI6(+ai?OcelFm=C6E`>@at5Bo#eX<jsOzfkf|6GmklWWR0v>tfwl#%#Ks1APN
zT$IfqI@gZSEx13|VwOFAk2xe<z~c|$+E@tGF06EEpEU?efPU7_{o&IrwP~~9+b&Mi
zE;2=es>~gP8|_)D&_u5>Hd*F;RU>L)&Y#{ICc#Y2a)nfOzsE>~;i^SRUwJWo3|5fW
zentxCu8MO&KdbZ2OR#+Y)JZx}``jm~im1|yk^c9~F^B8EY@G57wIsg3#2?!ilG=Du
z%kD?25b4>wy|!6wAoQ;(h8cS+x9#vK{Y_HG%V51m-)tr|%?MVuD|@EnnX0hyKl}Oz
zXR4-I%d|em%3!r}<_a%Bi6@F_wR$nZN{7sG{;(Fh(%#LkxVW2eZS*F%A#*ppAC2x!
zf3pt7TZa-u0*ctT&cP}ta%A3qmtWY3DhxJyg1224B{V;Q&snnZ@djAvrK&KzD*r;|
z@GP#zwdTvlSNxZ<wV8^#k4V9&a?nBeV!KJ-@xNwdTCzL+X8XB(FLc$T4^|XyMzVxH
zl(p-1kQZPfec?T}7-;A3k8WbHs;UkY`D)?PtVN!IHoKxmYf7bav0qNVx~OxX7Se<6
z|J4B5mx6P>=Y?$Ni8@8E`oRWraG$HCbP+{Y`!g9^x;?_I6Bo)3nQ}QYHuxbvp~O*0
zT*}V#9ohb7<-%Ay*@C6pB3ImNQH{3@VI#N7Xtahx<BD%=%f^RWc#WBr8;Tkx&*beE
z6#Zkf#I89cURo1B>LaVZau#F!>7s4^e5L5K(9TTm^S)G3C2|;4ywKulrJVK89cov@
z%`04k!qSyn^u}Wqi{c}&ey&nn*@ZOZJ&w{yE2bSsb;oZUn*8PqIw%%BxQ}-dV19(b
z6_=~dCAGF-iW_s=Mcbb($ZOkAtDAWHA)0(0<RihG^kGz}ul2eg8q1Oo(KQ*K=1;9i
zGDd3@uNPxi)WD>-dhT+&-{4YS5_h9`JsvLdme#^^JkR#LDmAe;&M#A{_q48j0O|0q
zT0-K3aaXgIO?<Rm8VWq7=9NNu*1^D~6U}7bdLgt&PHVrLQS<l7=1Y5<Bf8$ev|#@|
zly7zsstiIb*?SYnD8>Z@QL;f~On>u6GY9e4eiOM`2%>1m#K$l(N`Flz7Jg4AN4t>w
zy}4H+Qr98-To~R#JKj1aurCP?-Y3T1REmkmnt=Hl#j7TcJ_}ojJMj6Rz2St<G@V)`
zT0&Do>!pj%qt0JOW{_;ncGN0<jzTo&P5PRN4@PJab@g8BJzstfH+R0XS{_C4m$EQN
zhN^Q}ch&L(-+@I{Lekq}Xn;T&mq3rUA+5z9RSfH)*rOk#WEB<b4`20HH_llJ|5oc=
z5Nu+?^&{6DHPZyUH1FkWp{&N4H8~#bp|dM7KFHI_A_wfCmu<Qt7!|qHanVtn_D-Kq
zw*{4LqVP-DQQ^!W=iig)6Q#w+x+G}uu~RaytlAy28OX*kE{mx?z8sGb<O`=mVNI3I
zJWt!%-KNoXdYwY3_-A<DD6yg4WMyo4Y=%c(T1*jcOtC79o<5D;^L$KsGT1l|GkPop
zpZ^Dja-3B?pS6nIbChl~Qpb94;^N8_15A~`Ji%l>PgX}v=jB3|3`21r=LA!6%sR<W
zNUL<WLmZ#)dUjX0r6L%cQKaZv7V?|xsu50%NsLl^UGci5m%b67z+527;!zDAJ(3x#
z=`}gp@%9pCDW>*)K6XS9>Hyz(!JPkmjH+fp_6#v|tT0-(G|;Q}*j<70Js~9#oy-eD
z#6M+ty>7aBDwG*ZtzI)-`_vcejzT0IQF^6mJ=%9|uL2~6<Jzhn+8atuUttm%7IX~Y
zs#9ibn8-d@d8M*_;q{01L=<2^&I`^Pr&N(&jvc={uKTcQ!vNjF#9F$<H19UpJ(*S*
z$#zEdXk5Zu-kcNBCgjv6gphnN@u|3g80o=qBbjKDcUO86kTFjrzfGxS*C3=Sj6}3X
z!BekNim<j#-#|m@Dn*H1ex5bUsqw7e(0r{hsiZ+NXW9~;y6S6dU<$l=z9UWSB3x&`
z+Pd%RMu9bfL&tZ5*~(Z{`FzFmSdoZ0f-i|EfdbvLa2^allir&mNHS~UZtE)~ba--R
zv7sD$Qe9vgd_DHZtoX^tsU2ydFlLU>f7VqpRvHuJ<(A^iqtu~Lmb{#WiPy&-aWabX
z#j&Q57>2{gE37ICiNSZ#>0Mx9R3fTV?_FOt&vGLya+p^&7UK^8f*up5<@4HUS0krJ
zQ?!9$NgFiLY)0c&OZzDYmv^I$_bf-I-Gr@T1(p{lUOqPNZwseK)Ot4?%&~T(!aTE|
z?eJ-;zFzx${@98<Oxf~@?WD3&CW7X-!0!8iEiaGHYcJ!;*9MMUgM7nQX5GnOaLpa;
zYxl_wd=P`gG;r?~{RO@0M;&5}2=UY34p6k-S&?&M+=$;t!_r*W^w1xsr;;vb5go1^
zZ@86*jcoAD3_058Y`dF6{l!T%Cbd`nUpZ4(?KMPkrDd0Ymy@tteXx>9Tq(7Zpw%I`
zw6s|!;wo*GQn{j!zwCNQ5Dom|-f)#nBMaAo?EiTF(NUx4p5%SKZtm+qf?^@Z>=#<P
zyLbZ}16C7xE^szWJ@)RECHX`F^OI)n2!%dwo);fR&M;>38%TZqytYi8YFbfFuOqzP
zkDTHAMvr4%eLO4Twj<%+LTerm{efYa+Z8cVoVaveP+nYCY0BHCxNiJW_|Qj@$*SwI
zkn+*WO7sDv{+Uc4_E$DyZ;wAO&IF0HVYxd@Xj2Zq47^jKpytz6nw7{4#<t9_WO63R
zu1MX7y5qZCDe?svHT66yrzyC$yAsf-!|9uNrmX5oe3g?p{VK3!$kU3-@qIFmZ}R$%
zKbp2x20QJvWToeAIM!KXa$4PJUV(BLE6crF^Q&yMUItITH1ku;&x>qU*w@3OVtx1n
z15sQ$vYmvu?5Y88uU};|tm2v1ICo~u(D{~C{MlMs-@Bo5D=OXeVT)6D{k3Q3(3qV3
zfG>=QJ(RSNXg!!ow5!)se5Xp!P3tkiILqyyAk?xQJ!x%_-<z<TNpgsz93zqN*7=ZV
zou>pz?y-1dtR>Y%QShC%=QZ=Wco?Ogv+iyeA>zx!6yqMDs>cEOKVADq4(!Ls&1L2f
znR+#U${uiJe~Dl1KG%JyZfl_NOYyr58I*a1!+))n@P8tqfNy>zw$L#~;^O+>Vb9>K
zf0q9{5=u}PPEtdgTs(yKq5m@7*f@{0P(PpU)Tt~#z%&^{Yf0kn7GBI#Dvc<-wxSX)
zNO+!DbVbTDgX?~kRAzC3ZgKJ9%xI75=a%A7_F+>ZW_{?9ikRcpJQXEby&U+TWOt_+
zd6z5bl5Qq5JP+(c$k9GeBQ6#!f8ugFPvi6a*9(SMlvj^&9vD!g{T&TAI$TL-k-u+v
z@E9qL7^OWGhVoB{n%p0^rB;rxT=NfZFYfoXEL@*0IB92a-=AMr-L<|YlN_TkdoH<u
z*0m_p)No&;P5f8B{f_AFzS4)$5#6qv(&Jhw=>eYJ<Js4UmF>kBXO|N}T=%DEI9p4Q
z70PdXrk+zq$R%7xVLJTf7)S=;p`W-zR$9WGXkF(z^D;z((`nyhWBvSnwot>@{<*xI
z$#eGC^KQBJQc<<9l|e*W^&j}ps50Rgti*bmx_`i>T5FwZys{wt84ZX294~BQ-TtU4
zz3M`JS;?EeHRFi03AUhj=@Rb+?W4sp>sMLW<d{QhAHoLs7L$Z2iBir4OR=jA)Sm7#
zN{AxzXOoD9bZqFO|5&x7e{1Gaf4KHZO3RF-{Aym{H$=*jP(4yl#FFW$NQ!2gT``^$
zjjP2ppPZZ7FB-_-|H|Y~*BDSYl)qQ!j#z!~HT}Eh^8I+P!ClOZOSC$|hyibYJ4vL-
zudXUo5qaS;%Nl6~$!=Ato0rRko9Ak?cB-@^8LhQqoO?k@Gd|pPxk@fLFGcQ^OL=U3
zsFq)1RXF+|R0xS&?g@K7ZV7)-gZ1g=jzO0sDT<83i_t5mQ?w<joq8UQ7Lq>17XC|u
ziQ2Y&ptRZPFdUl!e#Onm<j**knZhz+O%p%O0GUvIABoaoS+-FcGz=#I`><?)YP4^M
z%O;U0(Cu(T))LcbBw(sd@VwDa^4<0H490ru7jv{vc462ph?xq=cyryi!gd*m=h{kH
zWvJsElN7No>}-_Q-?EOSj);%NW2yztL-fA$Wslrlq&i0te-D}8xFU@xt_;JmMT4J*
zxr?5BCMU&93^NK>*?+yYJWG))<MC~@`f6j-B1n`2?_G<#<$LaQb-lp$cA@l`xwzLG
zhs9sC60b)0;-rd5Dmh*jz9K3KBaB84mJa?MFcUEqBZse0!<VNj?9u5I&K?<I)qUtG
z5umdXB-=9Nh?G3!+n1%^-_s>NdU7ky2V>22y)wSEi+@{zD`$*+FiH)Pdz-stmk}DC
zY6em4{Xlk&cBdk!?|8z*(L@>}KyKMMkd)wSG3zMzSINLwWB?WV<yGfb`Q8DBE*6y+
zL_y-*5uaUInI4FC-1;AMBU*Nqg(4@z?ummLXhNmmXD!?t6itHfXGz~j=@r~(lUdE_
z8;>Tp`YelO7r#D6{3B4nNsExi<M+`&gHRx6FYO@rw;mV6SH+*%&QgB@1HvwMVqBzN
zhicf=tGq%Fv-7_+{s=R;Xl3<Woksf}d7}`P59#<CotZ(Qxyb0=#9lPi>jfnxVOzVP
z;fO#yLa(QoZ#O+oL-iic26~-!SEf@trNu2nC-H~pGxP)`znJh>C-sI&V)Y2F5)#Hy
zds)5Qrwih~F$Fq1DU%~1A5&$wR7_81-&AU5Fh`K&UkrAxLz@XtB%4w1G81Jsu75ew
zG$Yl^y4nr3ZpY56mK}c6BQ6)^yWCl*j}0>j{PVFijw?QlQ_>TQhjL*xj^+I$6MvPV
zly+oZo*1?J<&29CD!bpv|HIffMQ0Xm&30_twr$%sJGO17<K&B-bZpzUZQFKo`@gT}
zJe=pbAJ$lVjJ>PotXZX_2=}*H3l`J{IyBtX=3V7!(JO}^6Cy?vfy)OY#68T|1o;kC
z{3i6?E6h>rDdGUOeg4(+J<s#lcVgh!Gm}5jGex+FF>?xjD^&^f%sfqfGbl(c-bqGv
zGe6X%v!!e_h9kJx6{m=We9Oy^*P;o(P)rq=m<W985;YbG-J+NAQ)QTu7fXG(&V=F(
zbJ~J#x`S+zG8KatVS3tv#IWjvSUwZ?D_w%i*6GpQpR@!H)P^ZT!V+lRZr(V{&j*%0
zT+=CVVd4#9R^?_Ya(Jdw=u!B%$o2`GLyrZ7%+OAs%G}#+R^F2g09zTSwiD6E9?WhR
zK2t5kKG0}(GbS82Yiep&!m68a;_H%UnDSI>MmM_5bc!d`SSO6bS1=GrxNe11nGti<
zjO?C5_(cEvC39y*E?V}{m^M<XHGL3Q_a5^s)S1PW3-N`O=Q|#sm1kPg0WXD*<pU`X
zL391_d7x|E57rk~hCDw8i>cdCXoj&GQoezW<Pjey9Ev*ZtQSx}l*~{7R|L*EoJ`g-
z&-F49U|<>Mi1v+HVKFzE@2*LcD(Y~S9P*^pYtG(`#)8`kNwisV`M^?u_-M*vEq+Uc
z&jvD*w3|o7u03jDz(qq7DswC(#^@<U_Y@884qMk!E0{)IuGvv2U9^<|;$CbcZkV>x
z&%wM*(wH6t;<|cZM$uzT6e+V_SCZ_c%OY%}t24x3*<G!)enhHp`Bb_T^rTu-RFAM^
zb6>fI%r**mQ(W;VS8ETW5SUpoMDJ&+DyEJ~x7t>P#IxR}$r^YZ{bTFedAKZdKDzl#
zfjs;T56XSWjoHtW9)X1YVBW!kQj-ZQb`>`j=7jgKc4(@L5sanU?jU;zJNWpmhRi|O
zLE}&)>?OJzcCAlnPtwh^&;E<!8+SJ(rC;)iT0MI8`u^`BYwWGj<qg=RA>@sNaflxn
zj|sBb`c%r=oLf;@*{nj@M_EQsb_jmmwg!^(MIz_N)Y>XTlxhx(xkVo4aI!KQ8Rlwd
z336<Il~{`uqWd8NBe9vWwiJZ~#qL0!xhc8K*z^aF`f_tcTBn}alu6n6xX6a`I<eoN
zVJZH$lt~Ifu#WP&AG=0DJ%1oIL*?~br=1`wl`Jd6GF;db)RsfIOYqYN@YXJ7i)-+V
zw;snjJEu<KE@Iu4<^dKGAUnp9jEI44Rc#<oE1@p-9_>AdhK;@1{1FTW5_aH6+rTM9
zazo={;?EP2?^=g*nDTbz42hjS57IA`k$<LSrzDLjJS-B0C6)GFRV7ir3*;TLU#-R<
z;!~fzl3<U{B#mWc{U;yd0;5N#oyjC8AYFK*zsUaJll<aZGzE5<(Cp|@9!plk8&_1h
z51*N#ona`k$jYrVae~?^zmdSCMr6F$ss-=cs(Q9I7Z&NE@osn@VdOU~v_N<=UH#dA
z>2PNcdT@N1W~`=_W8|C4d&qDnLo)S|SF$Jgd#bEGzbz*`<>i2$#7Tc;SM9Tvt!k3i
z;_i+Oqhz}g>(!VgpFWOsfh5kcXNsv2DX7V8G|MBJGA`A*vq8>8WN~k%f4ieFg1?vG
z(3;>VdfHr&0rcuIg;Ze(d-c4%g2YMCV{^i^pOi4?^7lA;1~1g1Lq<AUnFh;Yvylrp
zH6THrRC1H#-~|j^6WoY(BSjbzk>!JZ+y{c)hUHS?86Hs29PQN5%_DpR&r>(s_n{BP
z!_@Hnx0@<};6tCqa#R(vC1_qpcFXwaPZ`Ox$TN2OIHMiL71!g@?Xi}up3`8<6bnAd
zV}1zu5QXuN1Tr@20AsYBwLfAm|AhJFHr^wqyX6T&@HYeIU2}lp;Jyy$^@)T>cKXaw
zdz&X<!E%ILN;`~K)bT)#3u82o=VV|&F}uElWkBlg?DhCTXF00QNYdQI--s>5dlyUG
z_Fd%EeSxL&>B|*0CDR<D;Dho<q-zHF3!6~Pe8b2Re|Z6CawWNs*8rC--QP?+6{Nyg
z6>FL&3Zng}pFQU5Y>3>CDFxNSKO-1kZ(iu?OIpTU)pc~xX+T>>&K&+aTY`)heqRIa
z=V!WrKKF*MMV<6_Q`o5KUMyslAlG4|>s>Si(nExoU<^f{-sN;fP0p&{sWNnOBV=E-
zzO)hp3FHDA3aUtp*M-aHk1WJ<86H<mGsl+e0j&mWBfgRQZBNz(zaL;bXf2d6a;lcA
zy_UCF_Sz_zUqd-sO0u>X_JT(KTLK19X?{SKk(<EWxHf_O1>AEkG6zN$TbM!ba@#;d
zLawisl=|16#ghr?7_Qol{U%V`dfu!kSl!B|YEsO8`Me66__S5{BAxW)%6ObX55+{c
z?WWh|dEzlO)C5w2pR&n77bDAC6hnHW4DL7ZP6*~}6UkM1EY0?gVm|%gOTPJ*!!Y|h
zwdeYfk8^D88!-MK#q!0eocgc%^dgcyd=5tU<e}HgX4ad)goE`h;{`qKz_A6pdlVkN
z`UWZo_Ys&alX-a9cOV7hx{GAd=d^I@)uyUFGvE`rSF$HXZYR4eDLD*4;t_wjlSUL4
zsOQsm>e_<v9*c5Yxx0cX-*#hgQ{(h1#`X?>wD*lVh{<*dP6PS?&(VaEqA4!z&ZWcG
zuv52qdgXQl`a6wX0N_YFo?*HC#jl!V1&8jSQ1L59^7g}*-uUZ^v*-ylg4wEy+%xj<
zVndTVy;4CtdiL5x8{2&hY1i^=Z<>hfI`2|ESh-`FVa$<T&=+rWuL*;_^d<J?_hku(
zLF6-v!CO|O`r~2?OX}8~-5EXUzVmhik<N2-kEi8y;&!&&*cbU1{7slJMwl1su#zh#
ze@SW2rX|~EQ)6e0jSGNIvWNm<sVr08-YS|E(a*ia``UVPwV#!$mkW+n6CUN#@Crw4
zR9e}hmQklMy27o}Ugc~#r?pSLYKQw@h$Eq$o)WvUeN7k_oNud~_TW+cm<tB=*ceFj
z?zJbzM)lYuF!2b%u2FJ0nmBAF2(&=s%c#H=$lIjUcS%&ubTT~Jc6Lc`iJ<;@CU$Mx
zM0mnYqM2dlwv1;#`t|vX%Unve`)HGO>nz~+>Ba+sHz~0kU;fnai!LxY=RU9UA4{Q7
z>-UQeGMKQfRfzL3<n3B&C6H`ODkrUIrcN}oa%T-O*ntRxPAfS#Kp{BTOP7~y<u!09
zI;kvak@cQeQHyOTYxURN|Is%AeFbHO|3+h3+5dyaGP7`U{!cWvN?XPmw+*iQLgP`1
z|BnptZiTecl7)>u8JDR{iw*sDr0=?_slr$f!s|Ktssa~(wD^Ktr_e4*`_rTC{K5=E
zjXi>ZuXsi$^s-0VcDH<(y~VdRtAUH#{V-*?nb(K!`$xm_?(4%i!8~??K+X@sr>#NG
zEU5l(yH6$tg&bMExL$&RpPhx9W_pzkCk(rEkl~YYtp|##M{(o*u?tukw(lbXdw<Va
zwJ#-&6RA3Jxs^g3CbT+To{1)(rLgy@`YJk_KS%Q?jpG6<Xq7ynsU;>7`7g=*&Tug#
zL;Vr_sS3wl6;5k*?Njx2<5F1yT_%2S?%LU8=pU*sV}p-R54RT!8v{KnVIb5cj_+45
z|9-Tyl(3AycU$KlM+8c05d3l4)CsJL(nZ%E!+$;wHduAvii)Kcv&h%Eti>`gvrRdz
zlj0gK%wOo}PeEQ5P7K-G@{Nb4M&bC;W4Y*R+fQ|`RI(#_m{_T_H1vWNkq6$by;vy1
z-3}qymZYuJpa>>zvczAg$38hT3Cnj6;u@u#L^U{`%*v;EZPSA4=Xf@$so9zBI|f&t
zH~!j$S5hs?N@J-u^3!N62Ln@#tKB(l9m-$@oTK%4oDuTJjMM34$53W2#~$@<U#RzJ
zWNLp3tQa}jSD07m9VN3Nb7yofnye`$)gRKN$tZ2vrz@O#HW8cJQt@!n2rpTi+Se^6
zb?UyY0_xE<Oq;DNZ<|artg&BDs+egfEVT5R#*k`$G-53G8gZ7kr8TNp39OGXdBmnw
z92U*tMdekmD6DOotFN*Ob?)5?j%D4by<pXqAH=T6D+>O~e!S))B=_1P<*`QwcmbD!
zZrh=7R)?;le9n{3ic*6I8_^KDU;avf+owmANyGnHk~Exvut?+Jh~9>X>%rUTqZ^TE
z4wfolAiKFc(u8)RViA(I++N>a#+9KLW(Ak%Vb4<Pe$peQK2OTK%(6%h<x-N7P9?1<
z^F5Bw891mqJO66-R8MdO9RGs@L>0vFM%}viqlTHcqZgM6Dj&e%h3GG2jvOygK^h5?
zfZlM5+wZ%Y71e7>iR7CK!e%QGbTvX6?serjdzirqwnx|M0S=%!eQUaied?#C@P1A2
zRB0Y}dzIy-_Qh**>Gjj6?xlA|)hhK1n+u;RyALeT>+or6NYYQIoDA{6JHRn58&=RB
z;V~y$f-;$&giZCy2O{@KC^(~C1%=2C$NkbQkS$bg*J)Vek#~w$(g|wkJ#0kC-Uw9A
zi_!Yq{B8EL6s*(KFcn$!7-drTruaz%sH!4LyK3Upy)w=d#)ie8af=K_FEP}e4e_-z
zb`=%?Y$`UDs3Nu?0JlZiaEnnoGEu)gu%pQw+0gA<o58daKZIafyNpOZYT-|QplPv&
zHWj8R=RrtPqWZY&fa!%;fGFCrp1FfsMBu`P=5yl<uF;R)RpV=+acxp%m)pSBq7=YD
zs|h8fPn3`?J8C%=cOUYBPROaMiY;8ar4m=VqYAD{{q(M@stR?nXq0KU_%1Kr-GPU@
zZR+4S2PJSJ0dT={HT@UtEFKbM%ChwrN$NUfS*$l~`~hG6{id0UCf>h6`*1vCB71Pm
z2vyJ}I$@9U%ZS@xkXSD-hu03WVYfCv4;&Oyl1?w%CJp@jY)dBZp9>tR<ywNQqo;;~
z2d_ZS_lNpijn!(gEp$ESpc&jK`;-LT5;>lql&0*VH`vLZy|xlPhvn@u4Rf7Jo8T#&
z{2UsboXUn|^yE;tJCMcVP0ykzIm?B!SgeP<%C7n8DA>8_iNgGj4y!YfL3WnlgpS;B
zDrTv#v4kz(kpy3L8<_qV#5QOACZbFnHtT{IZriTzx>y}8MS8WsA-1?5J}Tg7Xj{xI
zPw<jXmCiqbN<v&WAu2gCoIj?T+9fvThT{-uZx-~D^1Z5r{NWx_PSM54;wJWtPz8M6
zV6sPdk?#F)<`$7ksG#g*q-nMc`BZqV^#w^zaB^^WT5SFu;wlkAngSEJthfbbwH$R$
zJ^{bN$34?B>9R1PHf^GG4UYJ?<r&k|<M)>mvw~|`DR)INRQ<mlOpu7nlV}RWUWKAv
z84saATT9npPc`6NoMMkymm->eF9TD($Uv^L3_5^HUZ3HGdRqRmQys$3?Wf}bkN)$y
zMzskhMzzQ{mTH|&^<bTxUQK|EccO(2rrCb(Fm*()4embrCY}kuQYh;Lnk03E7JgAm
zD>9^n;hohx>|A#oy_fXpMQh!d+Qoh3qrjZnG-VYvN@b?WA$tsKI&}@F0jwcNM0&GJ
z@vGJ2`ZS^i_|km_FRY?mPg$f_iEfyZ8}b&}t#1N)@O*;Aw~Hrgw&PV0Mz-dl_|bay
ztZSZTl%-vQk|p)cM*Z?;@3bm-x_i#{<H{wkv<3n7WQo-{Q;pw&rsXiz2QmE5m&)Ou
zq}{_?3umaSTdL0t0`ZCj|A4pM;+!9K0dxrh0ZcKmNgH+?dZfw@7v@3pV3iNE2@AlR
z0`p*6L^lLB*NsU()3)tiS1>-Sm&qPggh0v(cm&g!c*R}h%oWHTDh^DQ30mS`EzGDA
zBUAW{Yp!TdJ%hvYR(YI_lU>w84P^j3_H8~`){8jk&U;tWaE;v)AB~M8r&&r;L~_b0
ze=?Sblbk$D))bb)(YM$axKMdQo<T9XeKj$~7ESalxS>R-62;*GiIk6Z^LpP0mH9h6
zM;OwRkb(E<2-A{DO4y~}`5iS#IdUZD#0`h*j!2|6Uml?+CqQUg6_fI*j<G~-RvzF(
z6TjEB`&;s5Nj#!sH}HjpaBAT1>;A?r7`Le4@}j&j`Unv`Z1!ASjWj8>#f?npwJGpr
zPV}oCteOM^Zfkxn{rpB=se|9o1tQDQzptFeN35?wG3<kC{xU#p0^d~)|K*<*%lT$u
z`1d!bgLHS!m*$7>g7a}<s+82(LQsL$hy}~-CS7k7uqX7@>}4QL$W#lPSq46U*rjIL
zv<)Vly+V}9d+_+z_L4>*VZW)KIi~i%U;&iH>*8#-8N`YA)go_1jSkh4&(*vjIhO-;
z<`_E<j%vq>+lnq%)i1$;ia2fZ4j&CFk*O+CheE_-Xu`0`?8N@<YG?=%dt5Vbbu+IQ
z&_&$5F~{Z|;bpjBsqT!&X)8sfzE615ukXj_TLFWt<KF*<(AYWt1EH}n{*S4`UASXT
zG%a`4HM<DByq7Y`(rWvXPRaICF3xX<gWZRC`%(ToF<o`A=A0r&f(e7aVy?w+;3D<E
zTW2Q}6jo+RZU}_+Ve2BX{N6sl`gegsd!qaE)G-Dy=YsV3{dN63i<7K9^!CW8(17*9
z`E@)r7<ND#7=yd|qq=-p07Y*>15te^L;f=l{Z?qX_bb1IT2DaJ`zK~tqoEN{j+s`I
z*HR7s?Hc@r`oh<+B?XFxuD%{}@X$C94R>FanoG40K=DbO_rK?)-aC0k@&M|FwgAX{
z;tuoQ%32=S6kWJ?N^=RQtG~`!uW!uw#O#!XW7Wy9%wqSh^gxW#SE(^fzM8Z4z5QuT
z*89Ret}B21E$PWRk1g?qU4~4H^V>MBzF#?cnsGq<U4r{UayLt=#&odppt71ZRBw9%
zCDl?&a}3@Z&%`JmYznH0cS1N|*PTuOo0noB9j#jIGZK~7a#J)`Jc4eqffv99z6zcE
zikoljSJ=kEWw3no@Z_yGF->s9o9-vCjCxqB$TF+Jq?ufrhlX~-h;w7~t_LjOkx%J3
zYm8CEZ}I)eT1YC}?t^XQ`TK8aAqEy%J))uM+aJ)O5+%d?W<OQ_665^{T^(^t(+KVH
z)395JqX_TcdyJZt%$0`x#LIpRt0aP5DM|=C$!n1MN>zJfDVWFE7fKZE>}sS~xK$??
zI<ZE5RGo?BGe~eNcU+)%<KTxRl{wJCyCGa!stZdEPvAnMg!gG~ITA1kU9-~RuUR?9
zO*ACuYy%E660%k<7LlIZo`9f6p5x@<4|L=F4Hm@lTp<4_Z*U0;=;{}2zs$)*)-;ZU
z2aqt+RNLSTuFx3ie5;nZXqv<Jh~p#&rD)*<!a=4Yn(9DuGaMFHOeYOFbr%_PxO>9z
zN%?#<<FW;8pOmFDS$Vef$s2-keOjey3nrMsq=P|7HQCZq3eC+~iGio*Ls>wsn|`0U
zh6U1HCmtg;WWOk%+V-W62#`w#@Ku{hxHTZA>*&|S+P=!Ch3<OL*3%?31tjlSun_Lk
zjXfU&K)22Ug%pEn7WKX@%O<$1uRf4hYM5p850=?^h%Nu`xm<}o{bqj-O57QOT#<U2
zp?}`Ta)Io7<HCqy5h1B)(z@Rvv);I0=5WJBj*Mp4@xfY>HG&z@PJBtpx589%q)Vm@
zH<<;ZgdLOhog1hg>=_1^jl$^jgfls<_!`MdH_4tT!xM*S3!SK7w?vC^BsAL5xz#$Q
zoHF#7amYoYknPfXjSy;avVhKzGg{#YL<)8qQ53*4eteb6`bKUS>U7(dvR41^oK}kj
z^T<ZoBJd#PGHENGgMg%SaEQXQgECHmQ3ZKITb4+GM9=gFq6{A@0n}kF??=onURAlb
z(Kq=i4IqgU)8TT4L}<+MLHXD;-OYo;euEmfjnUE6xE$%rEN}|IKcAuKWu2Zw&J*U*
zgj4{VBsYU;BzN|B8ErqpFh;E%&Jt(Id^3Rxbdt25npmUYu!t>QKV5u+os0EXIr(jk
z;m^31B{t?7&7eIeKa-WjH43rriFxE!gz>QwO3ej!<}mWFVMdSg*DJ|{0hO<FqQ-v;
zf<vNqRkE>Ftjo4K?Ie(p^H(L0Ktu#9DGlZ$a$Q;=2<#Zw+{;LrQ=?^T1f@_1vcGHu
zV6k9FlWNIDYLom~Ip$9-!Q14eCYgT~OKXHCv@N!A8eCr#xaJm9W><SHTTjR(7%3+V
z=6<57Bg57yQBFJ(W|4EJFPCyWsuK4|?$BSv6gX}Paqg{;{#u#pl^`u3)*04^#0KIN
zmNpQ|S;!~Ner*fNnrOP=p@O|#un}95%@vb#@)Q~AVb~Y3c%pSZOZDf(Tgviw7<!{K
z(#Dv5{O#A_cDYsfPTIRFx!(9VS@;nbXA;k9Ox^M0eS=VDBo^dlB?DKk7x;00`#K}w
zQ`Nm~fxYJ_Z&H@~9u{owHIz2sq=RRnkyX#@mFxCGPr#RAcv3*zH3NlWSjb?U>X75g
zPJN5PKwOqXM}hq+n@`g0MP5ybfFWW1y%23^!wiE;61|)gz6vdQd@3!GhMVWPd3|8T
zDwW~YL2uaP(UUoYD1lO@7YCCFcl;uxI(Qq{PP-6J$h{wpM_5kf-WDGM!|(9ohjj=*
zAz35wBw=&Rb4NIY6Wq6AuudY5gqI+(L?_MJupjHpRYHS`BePc}3)ir(n5rvlrxXfl
z;2TUl585;=(mTtDm<@(5(k)eRS<OMhwuR~h|MSQx3UZzP0Ec^F1NS`u+kkL8ObuKk
zp=^YQ6?z5RkWIE*<X$f>=tEL^YA)l&VI87@!;TLniuh>T&4D4B9+*z-(dt|d^PHj;
zsOE0q)%mwO%;JwMPwB~`yKR`kl_T0sKBr4dfSK9_r-S$s;Jy)YMzS>28O0{Rqd(lF
z>5{OTGojGw>hMtH8vUbE5W!eMsStjj*9mnJsSm<z*pNC>1Fao#8&A>prY8wM^gPKB
zihz)XW3~+tk0l*~B_HGFyK3_bvlQq?n$#BWGX9;up|RNT9l}s?rvxf?w-<aQ40I$w
zG$U)*1TwjT-Nt<H0hE+96&;p8;PXhs>y3l=6N~&K;Gl-_Nn$Nb|J(a5btaGfS1<%r
z0M=Z_7O*Ts7Uu=>Dzq>ZM%L^Wp_>tSzaeE09J**<8UFPY?Ht5k-~bp%#}O-$%n}c@
zmvXcnp)%!P!{-!g!1I2g{WPZ^j0$WIe@E5)ScB9!ebdNrDxc@_Pz?sH=z=IlJLUT>
zFpV&3ur545IGr#b;9@0ZU^z1(oChh`!Q+oZu=&|^X_SPM1<<=J5(eqCj!<W&=&ciL
z!q0YgD}H$oKVjUd3HOf<>iMYgyq23``^NwO63xmk#pJ(|AfdiDQVxZ4It~qA7oUyt
zE0sS|T$O!DN!zR43EOu2R4O#5(`(z3CcWvIZiAeLyhr!`y>hiLeUl6oYrAxNURqy1
z{Iqh*e*0N$o$8D2_$e*hy7JR%IBDR%X~`fi{~BEc(jd0?_}vbSk$EMEVG|eX#iP(t
zEX@9<P`T8QV)G21yx2?Q&c1JhbsIy<%26!MHTrUgKeoX<z3F>n^Jq>r*-J4HqWh{d
zj{HB1`u$>2-8j?=%n9puDzwS|UalJ)TvL#fYK|KXAv>VDt}_i+pXD=nG`$X}gmLga
z@_h$^`rg()H3X&Y)&(^4trhd!rr_2(Q5o+8;Fhc{uXd&8Q*XO^RF-A=NUmA_(Qa;C
zc6n$<sG^LVpkupCt7uF2l~Fd|k>TQF9OJUO35m@1dHsWKfkC^*m0>e_tEKHj06^4h
z%*LH_mRtLX5+n-zOuS`aO&q%cJo-X|ZN9ZO#FnYR_<ifELDkjzS=&0dNMDEL1*Y9R
z3kUH>!gYq*q6HfH3r#{*LN^xe4eO<3JKv$WbOd&fu!nDuMXq6W^@6Ax^+|c5I+O0K
z4aA&K^ErVp4}QN+1fx;==V@D^^~b9^e!7e}OQ1b|TZnHscyHNWdP@=c&QF(58KH8(
zrK+cpEZ-*YLa>M}_2$EO#(msQ#i`JzCy}F!b!Ud(fm|N(;Bea4S+6_OR}tSyu6swP
zYVk1nB^x0=TzAVqi}$d2L!Lf}t?432j?d@l$0T?@y0>{p05@9<Sy*tC0bEGoC7F@h
zy6(@deiFPyQ09*<bWYcg)j|Ll)VO=AC1YCo!W8fiebq%U=V%IQh+B9D7%Z`%bx)qw
zu5CwtNd}mE@M=W*;@~KQ6*?mrfIh=asNE@iprQL#Tju-d4X^0?0MwRIS`gB-r*v_K
z_x?Ca`fG$~XBRScP+0MetVm^;NbC30_z-Ij*2D~3dE1;9AYR6qI`<%H-=<~NBW)9o
zT{zbIX@KSf@vqSI2;p78E;d)az4Wlv=sd=~H+>aq5{KTXyc<5bcU1Q6V46ztvpynE
z7t8}y=$k;YIh+YfQjNNKra}Mx5`n~OH@A-e;qCg5hoLr)jUP#ec7o}1tU<#G&uc6F
z>h|~xb5(WSSTrSGx!nK)Ua?-ZR#VmYMwn_U4o#o?cL&rmxZ4|^%Za@iHaOFN;}RS!
z|H&no*#Ea~@0G4h-2ZiZXKG3X80#q`ynY1>wQt~qcsuWJ7$F3r@Idg4+QEOkh;$LC
z>8YtK#OQVk+RIgtMZQZak*cxgcFXy^zdC$>-d>B9k1<9W4nDqJfIfnh{D72*$0Pz7
zYWF|BjXYkUoCq(R$Px^-B|S1dwiHqD3rN>4g2+Dxo=6lBKBWfn3;G*;zHBl?55HD$
zf<e#ok&S2Zcb5Rs^U0k<u+p@i?C}5G_o93d@C!s=S`nqIF#5jT-F}GT3CQQafSAu~
zWqoG}vw%qb$Yo(w!6q0YH!Up?EZ%MTcl&-#r1pRm{Qh~;dn-&6bZ~n|>LF&Hh;dzb
zdE8<a%y0)j{$tH%X(c9sci%gIYwD%gKYMABxqQ>T#AKR`IAsVWt1W^z0V=0*+^0fY
z<=K5>YkXWTkF>11db`7$%)0Dfbmy7$b1rc8vY9OYW$pFK&F3Rs`geI<5yLuE)XT~^
zSz8z_4Om2ai7+#>@b@%l5exOS!yrx7m6uF`A*r;OA=Cuyzlc468IAQc-Bi0X%G@}r
zlZs#Lm4*Z6vNWBTR6>4%V}|8PXQZ$&2eIzY_dQ9>hj;-$`;6oXlH>&1`I!=qu$HIt
zZ<+$V&G9n#^DMXeowi6f((&4odt4ds3Ic^h-wbU1{lEAZ<@{ITK{had*8@e5t$=_c
zWVvV&b?eMLp~hz$Fj=&_oo3>sso%r!VVT7(oTG>n3r2DC3n;L-LVhhRNktp!?lM9N
z*uu`kR^o!^Q|X>@x@QPD7o*f5hL$$|fPFCx*AU)U=>z`wZWKiiiO!K>ztVP=J==bU
zdRe|D4I6SylhQLPxt^VQa-#Xk^-s||tG<tMH=GbVJORvhyHh7S%fZ1C%Z8U2X4V}m
zP_$hHL)G->H?_WaxE=ZK;`6uU!lD~zb@*tMykweZ-&f)uvH?lV7i+HRjU!rHxF#Jj
zcxA65YEW^JoxH)Q3SxRIPf_w*^}`@_a#g64++4BHlRbwhypY@KeUgk?j-uBE{DXdh
z3o8G7h^I`=s+yzmRVkX9#|_@W$((A?w>mIjWFX^p_NPAn2;OSDpq8D!S%`VUX`CV0
zS(>$ah0f9iL4{{l4bR?LJjG}f2r}C}2L084m=F7MJfbWc5$2Wt=6l)#)V!>6dh_sV
zO|czprZnX;kI(%Y-Syq4I#c~jVZ8GqKnQHKPEmUHmUCd47G<YZmLjLRBu<bs1n@Ix
zcQ-SDwQ-$3+JvdFQJ;5?7bvE1rQKmR3XV_+K92LSj4eF)@`I|sk-F!Ot&x7aBuKk1
zpp|B-P|DKIzZuUj$5%yIcysJKjTMqAg5+JbDVrT!5PI4xmvc?>oLa+=vG#laOl4Zm
zRv6C78&aAz+{uW+9%`YY;UY#AvrXdTwdr5YC1?mQ=d-cZ+Bi)a>IT142oP4XWk1J^
z9k2c2DEFD}vCE_V^vX_iY>f(rx82U~@*{+XnS^AS)}D=}EH!~#kcRl2xJ!Kl-L{RP
zPBmIwaqwpWNdE$XZVF=qgC}wtR++tC!Kw?gtsF%wOdFy5EylXPC%QbEEB^S(=|eK`
z^P!8E&`(a^IrS_s$f+ianxVyw`2BkF__Vhm*)*jN`m1#r^{go+^)jZ21OyJuyxwm0
zkR7~L40)_vElQA1=hEDx-K??TT2vNi106}3KWXd-M#6n%{xwb*%Is7SCxr>Ow5i1`
z77^QPf-kD`v6*<FB!c7s+6O18RXf;vkh)i6x1F|3LCcWfg`OIq2*_t=5v$K=cAqO`
zsw{F1d!zm8%FE&x=?y-)yRi3phxmG(KvkAl$j!fqV~(2vd`?xIga4KpKZi;z0rrYv
z%ms~FBUESx2%cebHP=woC9>o(SuBqMyf|^ss_i5QUsRDDr$l3{x{^DmC{E|EM!)<e
zi}4hT7S22qK7Iq)Lgs*IA$S}2;S`D+WsnqPbH30|BwTgsp!!b5Z-wrXACY98^n6O%
zi?`RVJw}ADs?ZjSVq<4x7<{hSJ|Yw}#&$dPwz_bL8+b0#A!*10q7vgSoL$OCkI5gw
z$Omj1G86T3g3@off|A-1m)4Z1tb}z?#Noz^9kX>^_q%YC;U#p)R814)&jIL`XW&<a
zaKDazwketpPVN469*3`dZHnpvQnvo<;~cBw1NUo)9d59on^M=}@fe!E)rTxo&VT3Z
zA&~3_pF{r|5uw`tlaLBm*|y<=JBToTTgguFK#7Ya{+p@BB31tiK?geeH%8N7=D`{m
zMtVt~4>5hSp7bhK*RX9yz*&%u!ADI2Y;96y4q)1iieo`PF2<M!1F+2hSVF7fHahCV
zS%JKB=W^pvx_%H{0&PKX$r&`E%ni`&2&#(Pzw@ulPZprcBskyL$5^C>WXB2_G>d{v
zQ@?pI=Z`Ox*{#od7OFjd9l16RA{(Thr&8+FBxjh6vbC3FI|<O{!nH-6%I=s`Kuydh
z>FnX=Lgeyn&C(K@kT8Eu-;vkrI<8qGXXaOM!zspatQX_gFAhHI@)WsZF2Pv}xpd!<
z&@sw_G;_cJl{pt26u*>1oQ=JeBoMPkxby33x;-;4SPvUT!ioasJzKzAvypkCFT(~s
z9y@PHn5xsox2|d`dz$Z7YK#3Znp^<qGVk>JzE+*6ac8V}na9F*uzLwqI43ofi51XH
z(=93i;eq-Xc`kFI-4lJ3c+6L_?qGhF_21lGTz|cPcu(lCYe+6S#{*6l=sdUM-JQ=-
zM$Nrn0l6P}{P1Joi|m|3?rYTZ)MgonU$Gle>33G|9}`#W0+{*PUS1vvIIHQWjoT7(
zaDkdCF9Hh?5c3ctV2wl`(cMomU~#ZqSJ-rEXj@NqOzULCUg;9WTWwqBu=>Xdr(X8!
zf8MFcAc7CnGcpWh#1DLbXhNl<;7z{DoObj@lQB7Uh0c!UEoCOdhw6yTj`YrWOi4(A
z8?gPZZ(1&I27uQQJEQgAY{3}7ME05@!SCz)y6nRl8UiMV78nZ;1rB@5FY~H5Xur$1
zYz_(yj{jM4$U2)kM+{;dCuceuWWQv!XHz%L+nPt-wZgKS(ER=J(^*rVl?#l1WT>jR
z*A|N{ujrzA%{Fo?06EtEJzT5<<NP=UGfD3HFH>Ez3aKW2qUw)Ew|1xP@~@GsDfv5m
z^noxBBCnU1x;4Y}hVzmszo;pG+B~?v2TYqyiCaJ9{0$SUHf%ZiCFM@eO{I!1?}J<A
zWg=wyFjV(INzR{wGBAVz8u6$$v&oh@Gg+}ebJM4ucCGJ~LGGS18-K)Nfy|gHQ|}C|
zsO}NA@cIeI!bT>b<C;4M+o&E6>eu4i;N}Kn3}guXGh?}atk4%<bXu6H9twF|$$G2x
zEZkaLx;RarQq(Qz4yV>Hn$S|myM*H<fOg^wBKG=ypR0w$g}*$jyY_T>=>MJh0+asV
znYEct^$}u#d~-)CU)YX(NxF~qo^!Kr@$`?ueW=?HUjh-xQ0~oPe~NvbH3fBoR#?Gz
z?!&Gtq%OUw))Lt+OlrE4c)Pzk3#6q3OxlkVM2NRola-5ZkKdI3ju#>8(Zn7h|Cu<V
zK1)@`M*3(fLMk;-0U(bU&=me@%-~6#rAhsR&fhU#$T`eE(%$LZzF=vir#Iv!WPM%M
z{*+Q!CP#A^!x8z!w;Zz2k5e5XPInZOb+0=4cS&;+`CWp*wN6*k!2n;yNd1oA2=!YG
zT#K5~`O2(H%Lb+%y|vHhuGGq}y#~vc8ove|Hqc1ABSFIRdRv5!E#|^h5IP;zt+44^
ze@hwc-0Pn`(H9ntF-Ex_+7%Ala?^E|bp!9{8kgOXY8ZRa9ahdEu)#e~U#&JB_Lb;(
zkAVp887ir}-y!j&uhybi?Y9wPyMJ9CLxz<x9S_5~{JLw#5|lBDh+C*uIXGeXc;+uO
zXJa|&**3zb%&1JkM9dg`H8zIsB$%5Tj(aXMM6W`hqHV~+;-lVG9$cgoouo(2b-)~%
z-CRLTc_@$iEJ$~C^uQI2mmp=aQ1_6Y2U|pT)*A@!k}t33ZF%LnDd-i+Nq<N2nmBDv
z_v4!&`@oXBmQyYw|ANYc*^F&#z=%d`ATEY%hfc|RA5*~?xC5!uvaY|B^`Z^;O~#9m
zW7?4`C2qQf<+bzgt9ZvbiXk1w{i$2)K+efqZun*9f=eB@H*2cP25V8Y<PI@OM*bfN
zGHceK|H@YWbH#$0g_+}j(xzPLb^tEek*7{*UO_Ox+wUsd+0+2n(G^*q?#!tLdK(e;
zwers|EqCCO(s`7Q*|NR)qCs-^S>gvSAlUq0Om}A^sShI)2se}z1C_|&#c?|nLKK`|
zXj$-9Y*<8T;!M6ixqffsTo&uojZ)J{mMo8J^ZH$WLchyGIp@grdocf{2Z64G`gXtG
zwCS~K4>dSIqtE)vrEuAH7a`$x$yrfMZ$Dqu5D0YLKK`?jzY=)g`rOzUnef*?W+-4|
zw!3_+xncMY5$&7pCuMamE$Q*g^z(o7v}+0-hYpIgu2O|a-VMnmy@<^hK+6Y!#t>VO
zcE$U|Axx9%DSv-Hjj;0E#}Qnxfbg+6GB-%iHX|l2L&^r^v>bzG5VN8IQ9}8x7oFtC
z1f+^0^c0%wc-snj&s}Q#R)bF$<_K4#F77kcYzURkfoo}a^bDm^odo;q_m-^iQ|v6D
zq*EyJ14q}B+Hrd}bC#Ra9H=KV=-JU59jp`f2I^v+I;oEG7#1z#-V+g{03M_@xM4Q7
zkd?8~b(RqnD9l_v$|ubL|25VTBv);iX#L8Ut$6$tb=>{)SEi~_KpIEXT-kn6MXz(6
z;6eUq28Ns5;^s@*M<!Oypu;&A`1G2CfbjgIC}Boim1zhLD^=El@{jZmn`MX1s(eJ?
zC*Hz$CsNJUjVlhr5w5c^UI>N+IPMR8=X1{rBm*`_FE6spHG3VMmsi7a^P2UZGgfg;
zTpbOEdap}#3<PYgv{w7DDFL+O2!hSdJ0==<J8lc1uY9mRX1pynrf2<LO<fKD=8@yw
zdfHn^9e@TvXcUV`NK({fU#W_^nVdL7K#TS(0{JUGc*ol4AaSmhyok%JRuy#}PJO6|
zBZNA$A0TsO>r4+Dfg`eX{erzK=3RCvBD-yERPX@MQoP2>w-R9|fjcAA5hp545X)(}
z&i4UnilZq(L^#d7+i{6RzTqkdrA|blyD{NJfj)LWH<lymO_l0za{skESfr=PU&MGs
zmut_DgP5G8Os6<Yr63a<%k?XR{MWm0WP|jUk9<2oEB>M5)?JwOFyzS(%=ihK=yToW
zq>`DByZzQ^e}kHGrD!xq#8ps<Eu(oea7SDmTA6}Xc&cv8P~hC8CPV6Go&)mzom_aH
zlYD63?sNVx>f2>qgnk*$Ah2=?dQq>cEHcKn!WlyVoxOSe%SYB`0qK}@9W~%4^YNQ`
zBlb<GlNcekDB0d|n(O<Qs}Qr8@;0Y8eOU{yP&oh1>`Kc=0uXkfqu)x9pQ!Nghh*YT
z(L=X~=;F!an(Bf8HAlN<XCbCarP}3HfuuAxM?Cx#7P{pJI$B&!)$2>NW&~VLd3<4m
z{y@2jyZf>ssT3Zo8BxG#r3!+KsBpZc5>KssQd;3fSiY?s_OSyVnGr@^!AzlU+i;V2
z6kh;fA7PT<N4gFoKFrAaxXN*2LhpSv{x`@7>lKcv?*wvci=lS3)&2!bQo}8FIzuuG
zm{P$L|5-F@U~lV*jJeBbD@q(>b%&gd4>OD`AudVw95uPMQiy;x4E-aovwz%}Emg0d
zA9yEDUx=rhZvr<#Ni=SZiE>oYUs{qzYh<eA=l1ATUjKWI_fZsLDdi?I=(oGqBpu-b
z<RJEi>=^kvj45t$e-3GoM>cc99KEQd$iOBO-1Plf^e>g)0{IBkme84Ma_XF5Ed5c5
z?LZ3_3~|55<E6I_yF}JlrN`8^=&iq2)x)SUEDfhW^|d?tco@|nTk4x}wt%eXl{8f#
zhNQt{r9NR(;y@?1K;iySgo4Qvh<2jNu2DBDSJfr4WI|yvEP{n_&Dyq6Rv!Mfz421E
zV<R`?2mmpnR3O}dKngA_5vn@MN7@Ss0e@ZhiODmDwkpQ8gsiX?E2mVK28L_LScThl
zBH8T-#QH_gT4a4q*b7#KSGc87-{d!&OQH!a#!ZG>#`2hzBt<S;de7RJ#%`OrF%D96
zqxrHL)O6Cm*r!w+VsQ+!a+$+^-<_EYbi+&eb~0a08SETpqceUeduHxw2d*L~S`C~e
zed#-L=i5kYQM&@E6F|cQ5g$X`Pva~Z>}^YG%~0CX28^-vLSz!iGAX%%UJs&pGCUy1
zaA|h9=VUF3D%aMcmtFrSY)#*}N_rm^S?Xj8ezzH^nP3gbn4S9B7_%YcS?^|id!zbA
zv1uE#cbWumQ4J4hwWW~7gx3Ge!(z-41?*+B9*JNmH6vo<?Th`s1?L3FpWNz2L|n{Z
zuHJP_?+mVOveI}w$y<9loXgpcRKm?O9Z2I#wrw<<s|o7Znbl|JBS!Hi%;Z;7tkqju
z^<Ps>6L*O&AEeHVXFoY(TY{ijb&0L76p%;Nq{b0Wt2DgVNkf7$!xt1*Li~b8eiJ4P
z$vhYt@D1w#0B5&*o06nXM?!~!_Qk0ZIf7jaxe(?tkAhZPvdGk0yLhPpjBYjcs8>0^
zFAtQ_!yMz+Ty+lKU1w$-&<tYJBNHAjn6tdE+pPrmt6dQIU{vWut_<%yA^dg)=z%7(
zGsDf|2Qi6CT@uoHzdct2HJ^o?f2x^g=FwWfLVFXhNIbfVX|UH4Av{Wt*N%$Sv3kZ@
z%;xp6VJ{s5^mp4LfB-nuu01vr;C)ir9@YVzTS>_w7NM4ORk>F5w=%VxxdA;RJ-=pC
z#MaJC-LBn{Y}*dnj)g{d@i{!idJh++-!JR}qW{V7-^n)8EOA9Tn`9|hmX#(Jb(K}D
z*Y<5<RqZaZKMYKNB}AMJG9h9O1Vq_g>2%baHcjgK-5DK+LSlxct5*ZxSfw=QgLPca
zv!kB~HKFu=R%P<1F|F|)v}tulSD?2zc4uY+op75WE979f=c5uenkl9hL_<?W1V2>O
z|3?0*c(B~URcbU9KaCqV`{LbJ>)O9IbmG_tAy=D3i#_WauW@;;WaVY8zaqB~?2K;f
zXBw0lLp*%W{Y)ZbAUm9HGVQS)w0q%!uN^2Ij_ZR@$p4<)COBv9fWk&Z5Y6&Eh2V0}
zTx}**4=`K)?(vI1h+oqdy4Kjrd~&_7+OB&TE1W-H)JyCapL5*O_R(S!eAxk;&-vW<
z7nlFV{@*|rC+mN-gqd0YM|sJ=3k1-#AWgl{97CXsb70q$)2TbQxvLdTKATaC)fWwN
z#3O%o<JiM;;=+d)CUd6Y0e)e}EGV$nI=$Nh77!S23>f@BR9Ut5ST(b{{jvjaNPob_
z1hDzEe*H$3HP|W1-7XXznBfs1EmHo}{JwmxLa2epeF-2DaPVjX{kO*G7a)+^|1!@I
zyIh7Amk@MspI(1R@z_YHmN)jb%f}m@+vDexknT&>2}Lkq{rP-#arBbv$TUnzRF&!X
z_Kl1!8K?+P!Prm3@^yD}wvyTtr|Mmb%cwe!S!J#mX!7}|%sXFLB`6XTdukSu;SWE`
zL6C>!*TabY!tI2Bp9L?;!z43SDRUzqXj6+Ykj3!!z3*W#0eF@s?R}D`wl@7OPWuaS
z)_USEYjI8~nBeYSElSq>HqPvW85m?`am+uY{_k4yKl{T3Z4LK!?<{U#`Xvbq{>A>_
zo_a)=nnr=O$RTzywONWDCoF5?!ZZw|ZyQoJhp<ZVdL`r7$4`=OYGgur>asd86}D>l
z%_KJiAGcN}wKIm6LQF1-$aZASN>kDWJ#}R7g^bub1ds)9c#jbL^}tO_Q;UWB&F+C3
z4hL*j0TLi5p=>s;g8j@zz_?%&2NHPtK#2-~hTImSy38XQMM<j0VF{7^(3F%2n!ThN
z#Ze%5im45FO1Fki_|Ru{i-POAhfFCOXWPI^=Ejkx?2p-rK>OcDRjpD2ROgl?V94lC
zv)%vX&aRtR2@E6kf%ikB<{n&(&+rKfx%fy&qn7%7{LJH)u~*vVBNALOq^PGIjX&zh
zI9wn@5fPs9X$zGTVWkMpeR2sCSO`6?-_=B*_1ND&*QQR8gQ(yHSl`bzpU{7i)BpI!
zF025SHqNmA&UF}g1ihH!?j?QTKWv=L5P81f+^u(2yLOHS6UG`V^hI%^HD5M-19v{x
zTtMqjL$yyNZ%Y5pe1KI|8u;OlMMjd~^-;zpgTsBY%0TOqmmC#1#-}Df9kSk2?wqPq
zFvDga-rc=CKek^*+$JDi(|xm*Fj0h>Si7!9l-eC^SH6jg9|Vs=zEQr=`S-?TktrC@
zxsU*u>RaQSL}69$6p6lz-p?xRJ#u~o9Ph<j*6m?xn&?|(L;s#$)FGds8YJml2s_VW
zj^#*l&BQKue9XxX;p(gh4W?1z3Y;4Xy{*2Y>qr6+4?X*V`iq?6DajQ|*Rac6nmYm1
z#$|Jhii-8pPI$oSFIxLqiUw)pxRUK{iv`Kbx5!jctsyA1K`8HA`Ts1<INkhK&FnOZ
zO%NKL6x0u6KzyE<(@gz+UvJ6~21@IF#4HlnJJmaQx%5?8G#CJw2~%aA%&|BhocTq%
zXg@+8`?{=$Dakg@g0?_@9R@=`G8yxnKGW-oW{c>PwlUf$Tx;K!SNP7ELWb<S+~&k%
zB9$e_3uBS`cdfis$&vheh6*ELi7Z<vHziS|iavkUG?U3`(O6S251I<3w2DS{c{Sqe
zT{^kx(*285r15|@WYk<ek9^k3-Ev6Gm#8K#<6UCHzP0r)HjAv!1MhEL+?@|(l@q$*
ze>q6;Dhwpnc+SJAXXUAU{_IVm4Uoc1-Vz$qgv9d{iZ@HRk(Oq5vS+qD{;9(`-b-Y0
z8BeFG5@-UeF_hCqL@M6K%-fT;4+{~w(eA6~;#3*aAJdkleU2e75*y3IH?H*FY{04}
ziSS-rbvKXfl<UILz{_H1sc~ZUVKsMYhxNr*rsR+EoTX%W+8bfNhiQ>OeXN(&=_X(3
zyC2W>Sb<@=(@57;E_s#T*sK*etY>*MtmSIEHVd>w--g)MAl3{Ms`=BD6X=~PY_%GK
z(J0L!Xhy9kGH{<xbe8F;-;+`J7U>IVgM^O@I18f5IvWf+m~)(R9=I`!f(kwv&W)s%
z4!}>H>mDUS{uyp?O){R_C0HakQZZi^wjbAwk=)cSZPB!qzd2?$YrJLeiElt8a&mWE
z?UNqo)0O#75+D<fqzH<;?y|~xPSOnxiHFHqmXV`pWoZoLE9Vuriy58dRII(p12u@~
zOlCzcT;iGT+c}zs<^n*TK9}vyKMxN*hE=l7CDnN%t1Ae{0}Y(7eVEO=GaQ($(<uXW
zVS_ECJhaei!9(QhoJ&u$B-~DAAe#)EF|oZVmV=<)-e{k6gp`|b8$-vpsbcy_Api|a
zRy4n4H#Tv52Vvo?MVs|R+v5>X9ZYlv>*CJ!;2V!Y;@#^5IGLX~2^=^lu3J8Sh3Z?V
zk~ovqbrk|#NR{tR-J~PprxhT3{I0|@FExF_xp5;b!FscCRS(`**B|a$;!-80Wy6#`
z(c>Uvlu>H~HbDL&4wAB8_&wyN*YP$ASB12^?wF1dyEz*e4?Vp7!Q25F&HPp!-<Iw^
z^lofOaxpGApW<oX6guUW;Zcm2KYwOp(4kjAZYhT&t|r;E5x!&v2Nmj~K4^x3++5gU
z)wD~LvWTx=35G!);63ick+?dFg%h3nnq@@CN!2@LugiNFQU1v3<PSGyn|$&G+)nIK
z5bE!dyd6dZKG}E&Y~57+5Wxi;Qi~^ZJB!COtR5Yotnk}JHbW+<Hk)PqV|@_?!>tZ4
z<{`-o=e0wq9k~1Wb7d~LkAiNNi~JgNeGLrYBUe9YkiF$p2EUd9Y(B}Dl<8bJtbUc>
zSyc|<lFO20n_Bzf1{Ju@2DXx%(-EU1D-!9Wz_(S3ca-N@vFKLdEoLPe)aMPTA*Der
z{rMR8_BpZx(nIQ07IIKP_Gx`7vebWUtL^zLZDDUwEldO(!rPFrSD0#VUdgRENnx=v
zD=>5e@epJD6O`CO0R_&cGB4Bg?qrCTz^F;h0hgw#MO}>m5!lqNbRow8paf@`Y2fL#
z{6F&02Fz&;a+m|k8Bf^M%8?E6h19oNK^QxQdCY9d;jQCJABMhRFE61@rU*XJ9*RzX
z=W5kgmoqr+x`(_K^Md+3aT2*{I(|q%<Ni-SW$&4VY_gHB*#=gq9peN<lp3F|GyQpb
z3K*Q8ElLMhOhqn(X-Tf0Y*NzhP|{0HiCTM6F*$N{Ys*9Zv%>?{^3Z9AxYbXVEgv$g
zr<V+aM$9pPF?{Y8LLm0ot@Vk+KzSYNs;9|ra&e1m$%Ux>k^VdVy7JlC&(o;#0l7J9
z0BOmdYa~yEixXBL8>aOkz`IzNQ|rtgBKE-+W!o;T=5s`Pd_5N#p$gT^O%xDVCSh1&
z_M423bs*5sA6cz-`5`k+F`Dxw;xG!~llCU&2d)7lwn2F7_M1V!GNbXoVLmR#|AP6L
zSpT=inhsCmVKdV7i^eSiZZ!Bm%vV*Fk>1foZQ9JCLRC*k*4-=p__E{?Oc;LXAj!2B
zFrihy2HGtyFn2g$4?@u0qbs2Q{Y>x3I7!WD^!n=htG(}x!J)nMU(j}13Zu^_rk~%D
zt_63s3*m|K{O{A2q6V)I;yz?B-P3>;25_P|86k+-o6H}9h0qeUP8#~aV3(Q*mXZM#
z?A`eXQtrH#{;R9hxLkb?Z2noSjE9!E)Ss?L6vck#TF%dlk*J<P<g`YkUtp8-oqj(v
zw-*yTFj12kEEoQBJ2xabDUSPPO~nm&y#C*z2tUP{0!Vc(Zm%Tq-?xWZTt{I)@IR#A
zono87MAJ>`Rw1o`NluYTyu@#HQC%VKZe&U*PkG7yl@^8!5#jKjkDo!<u93ERZ|HcW
zA+t!5!z7pJ0&k-fnp3Kw)0e)}Ic_v*WM`Qy7xB0N&Qp(Ik+m;>O*_<2JD@5^`~;vo
z0nOV;DtW8muzXgnTS*!H1V>?Jd1+HMB{jLip6j!5wB%CPHs#U;POA3sR#~_dYX!f%
z!f&!8WoM0qA(5Zgl{>w@xgXM~2yXtkn{Towb-XSqmcs->{ipM5f|HgQr)|18G%lI3
z-aaX4dNvVb0klsnupPQQv_6==5L5^ledyv4bNNX4YDj&vP{==daooy<c?DMY_6g9n
z7K%iPcbExd(#6PYra3zI1^9P51Bx@uN$%S?_%~thQYOVtG$w8ZHtrLKV$J$L(Y%_?
zs98|;TLH3axwblIk?8zceMii3^;e^IXr_F6G<Yq7X6h^BB!7p*8~M`LDA%|CA7kGX
zBucPt*|u%lwr$(CZQHhOoVHKf)@j?eyQkm#n3#yUANMz-A~LFKN9M}h3q%@-wB%#~
z<ZIS+n38vkdi=)q?$8#Rw-X~O2Ppm2<QZa`_ZYbYXffud2dw$&B&tv(L}9!WqPYFB
zosIrf)Cidw@;ShXrH5Hgi!h4V%UecRC&J?}%29^0B9y^r^Hxb+D>DZ6=Ib7+w@9*N
z#2Q0H;%B9{Q>ACQN|TBH)|}5tHX|OgIQ8zFcFAvcYc*p!qbmCvnZ?FYOMwDrVkwV&
zA+M4vc)<aiwdtOMGSeRL=|$gJwIq*&?~TZMYxP@1eU$ISdKn9e&0?L)D_~Mh2kR8$
zTM_jVO2a05I9IB1m|vP!9wSXx=yOrKjAYO!5O}mr-6>dBeEBLO>}8urAh=>=ZSZuX
zqhgnCvyPzLm_>Ka1<7HWpHtiU3z<Fln#kM?rfNvy{W;S%Ct}Y8(JhIkNWJxI^i9*I
zzT1q~7w_0^JadY2TGV5H;lTBZ(FZf3NgXv3=IJ`MUce7oLhLei>L=wgbP{P>@T^B!
zR(>5N==EXLo%aPAYx%mpMRdIJ+<-1hQ{y$E%OEAE=2zl~9GU0Oip`-4{<MyHqsmfJ
zpDfQgz7~mQ$&O4fij;*iBkDbo2x%zX+6fRTkohktV`kJiL518&1aY_Ppnu<N6&R-M
zA^Ts5ike9>hR(P;>54+-812ZM3KXc<K{^jnL*#b4e$3o^6u855WZimLoRv;w>*Fra
zsO6}tBuc~-x0TpH2bBh)xQ~@k7mSZ)Z~9=jQvH4WEu(!1TxKx~Hf|@Enk^!$is3eN
zSIfS6*?VNO!xlKXF7WuA+O^unUtp3&6d1;z@Q7P&+7e^v1AIs9_@|5x#y(f7xhNqX
z3<%Q*rp$?E-kjlDN$jIwsVaMR2;am4Go-6kRj~5wUmK>%`9-hPbnd<}>j=QZaCODN
z^K0j@&`P{p#SBV)G^w+ETs*N6xWT)DfLCP-ILrqTI7tr&Vuj|>V%nKs7!bO`yTL+X
zCciOh{Zod;2E}YT28fHANZHwqODKPi8GktYM<r{=5LdaTS#CYY`XI5GJ6_^0bbru#
z48?Ju-A#?BJ(^@`j0ArByU1xm;0JM!zDXV&j@=i@CVsI#RHmL8!y}1_ye#22X{4*<
zpmVYoy7;$ba&4T^61f6p?IP@T3PWS>8tLu+$J7x?hApG^N}~D%XMz@Nl}k#lWSpSd
z(iMfDl}LK&Ds`^EhbG>7)9x|51l+}wT~Z3Zu+Fvz_5BuqJ!=ohKdQ%=+V>36wkwNG
z%j@5wNAQArsvSj8N$DWnW#dI@=qgrxDM}?vTqxuu_R+|9{MqhVXp80zE{ZZEjL&`Z
zJE4wanZ8Z+r8m};M`%)u*fKw5Dl{52Sr(IKcsvDWKzv>ydCnjjaa1IgMU22Mr+iJ7
zY)Ch4uBedC%ViQT7N*p_^s5#7{?O6{F@Vavq}%@8MyhC3SY=06!&ZC-S}8ylB#(vu
zgWFm$1x<~~MnUlql&#)ua;QR#bOl>xv!yk*TuJ*Any*4Vo$9=LopmDhDVdCNR`&5=
zQ+mUvr{(VC252KM5s`Ub`STK?B)OfdPlLwkr$Q`^wwJ%;oAE-|C8tRad(z!}8fXD6
zX5-@4zp)1kOd0<e=s!=GVU?K4he-T^nd(@Abu+<4F=e!!eX0OQ7+kJ1Z+8vRu*J*9
zV{@0fLDv(Rzc=f=&ARN`eAI^OC{WtOS4Uvh2xN`b(v~WhMzFLqvJ-Sx#m{Ct-A0b}
zzsC_&yI0dbje>=|l`X}svekTG3dMOkF$hE5E_QFkEJ1xm%7*9Pc&W&0N}CMrnyVA3
zxYw?#jz54UMzD+pG6;y<Ur9MnL97C9k%AdgJB_XL<83w`3SfVX+S8yQTA?bY0fd$k
zhrO-ej^j!bfb3A$?HtrY(%eX98B>)OgKsUV@)=XRK%0p?2s8G`pE`hsZD*GGR#WU#
z>I%w;R-Ze|cz;_yA%Ck#SFgZ*Wk<KR<oY%ZusMua8>?J=2dr_$_FM0YgT4UjAi@|}
zJe2Aa>D3%HJFRcrM9=3GcZv6)rgCc-gZ~Zf_MQ*~QkLaMAC(WNs!zLpSybynv?wxX
z&BP9@!R|*naFlmHmDKA5D*G9!WVf=>&M<YLGy#?-3W1Q$9c-4_#%)KOUT4x1TN$(X
zPFT4o0iJd+{pHv`Zxq|urL7`7SFr1KI@6x8<}El;Rr~lb?jG7~w_{}x%&wLz9y#{*
zq2DO!)#jQMVrNX@6OFDhC^%=Y?*JEHQQCmvIeVyTT!`W4>>;l^1D7MdGe(Y{=!ImX
zjZAg7MP9a(QRfSRa(tmoV*2}ctyKe4p_46oBj(^t&WU-0YD@gblR17C<<Mrf-czkK
zM}P!UmtETbAZ=PtgS<5V`jcE^RzdbZVz>Vw*k|YX@66AybSC3|o1$AU>aXBggqp+X
za~1_x*S!L6%!^$F)tQoP&gkQ0ef-1a5>l-uQetfA5n8P(bV=pgqT9ENH{^cr&o19y
z?`O!jCK;*&V~eZy<bw47(-g(Ca%B=?fZcz5`g)v%K2ch9EraF+FPNM&mv(=8hZ#T$
z^(dC^=EFvb#3RuS;``-q%Aotz?rXoHaRhiv=5XrfEkMBKk<UZEw*;*ICTaTZTfPTf
z7j^pk@P6Gb(F-_Em#mcg`}^P1BY*%?%MSoBTgdx=zdUX%@<aVEeB+<rv4@qP<hWbX
zTHNx;>-(IL|0zu8M`q;key2|Gd-)&uCilbr75(~v<wy<TSl6QTTRmonFpTanpVbi@
zNO%>Cq7nz@-PEW%lt}<cY*7v>Axskz$h{x4_%5L9>huYJ!|)<sR?VlskOT2FggzJf
z6N`oQ_D^o<=Bd&>d}l_ZgK8A)LdyR##@T_-Q3I(&&7TB=N|X_yBK!}?L)<+Hf%_Uo
zEv%xuyH-SGI`!~?fiytv^CtL`X8^8C5^w?G>V9~PBy0Fh49&d3F2oBn0ksk#i}*3N
z0fiN2z&tMq`1T?vNJ`&N7kjg*J>~6rkwoq(|5sBnSgj)n9{@>?S5i@ciIqM7u<ksC
z$}{bF`Hh!`4a!gRi|C}sNydsn>ofLE*pjXXP~lxGzueeRSu-I?R-Gy3lZcf%c`}Q@
z9hKNXa<Z39qSll9e$!Tt6rb2c7=9jJo@);?e9Rz}GJ=;Lt+IUG@_3Mj%C|SaX5UPU
zbHa}riOlVz!SGCWfif5PFTLY^;U~>SuX8dWcr>mFGL!N+8I&G~JjR|C?&<TR+cDsJ
zOz-0*NE<K|gu#^Q7LwOQz!*huI!YNPl#%F_h~lFUgm~#T@m_Mqw)au6nYt8P2btPw
zwM~t|{A+AVNkw@m$uq?$idtGcOk?xGM_m5qkY}ihstw{aJpeUHnw)q%ZBpH3b!<j?
z?AhI!F=3bgOUSg{$3-8E@{s-2mF@pYE%n;ihtKaBG>?b25WVG6!;VMn-kTazdahR4
zKl0LEJBngd{}>0|SKGkQn}8CbN<)bcl;AXT88^=wqC@@1mp#Q$%Cb@-7J8PA_~(T=
zeS3C^vigv;P$tzA5&99;4AQv@UWQftThaNFRCD@H$&n=*r)}IvgN>SLR6`br+Sx@$
zF-O@V^T^Daxm4QGK*~{TD9bIqPD|5mWKs>wNX?VBd4@`|3cG)Hi*Vd<NSy?-P|qXg
zF31qgp9R|uv{mt~CPMe!X-`m=jFh#h^%1KqsMppMtk|tX2t9UCSBiean%HOXLZ@GT
zDCd*$6>=EolOdbejF~1MwZnO|S*sGC)S0bGH)FkrX2mRxy$5G-qf~wy;sE#HO%B6!
z!H=Z#Y3CTYSPkaLrb#v#>Y{sf#c|UJs#y*Q5_VZ^c|^*RMXd`*f}hg=`}1pw_wA(3
zaw4hB#$;AC!Y<y?surTs+iFM5<|MIo@O))|!HNA!H%R%A+na^*)%xsut3&1<kSewW
zLosHfHJ~5^;GZ$Hd(%G-=13U~pBQ&BC88ze@>rtm2iT?sQE<Iw2dZJt9q;b~h?BC~
zbs6M!TbGx2;m8BBAGY2B)I$&k%D8J|Nd2Kiv}9!kA1E7A(|j$_ZqIJms6EPhfqv5Q
z4IVRYFwU*UtG6YGA>a6}LX5n7TZ+G&WG!39XTK94Z}MpSWnUQ=;ZY28>;@0uSs1Cy
z?Jk8=2`x3gi1(sQYK^D^=a&sf#Btc<>&8VmE+~!c-i{G~AEw||90cPe6th~mRetx-
zu+7AL<~72fcLf%ti>(UM<?vL-gO%RgK`21_)zRJyrHkTXe~zg422pg#Q*Rmb@HtF}
zFf{GDuQ~Lnp6t;rj%0J#5B=(C+wPL&6u7=H+DVT3h1U9j>-J1*3ruNEpPk~#JF<t_
zPR`cOH*DI#<_R>BvP~y2VWmf<OOs-vllY~5e8C<^g{75$blBJyvUkm97W>;AEUueC
z23mk?kpe6gb29n=Mx9W_NF525CY&Jx-TM)y%;d3!lBinhOmUn{{qWc-hg<=Dzsq2H
z+wp9iG8`l$Y|^LP_Y6HS-%-8_w~gS-YsbtbZ8jA_tjEezBOM=6sd>Vsx05E0px;Su
z^{S!&sFem`*9JH*(~L5?*T9j7-PvqYQErmgAdjAzC?<K(9GN>SzhzT&pMNnX9;3y{
z^+q|=P)tF;Kl^ED@0~hHNJ-q?XAiVYH7_S;s!8@Yqp52l&jLJo0h~XT*jpxZPlc;v
z%UET(7U^indogRv^F+mt#9znThPQc=u;(rv8OB|kCJ8|trcZT@L2krdxx7{>+@mbW
zmI_8Y(_=f)!f7~1;Sn65zjogDE|i@`>gOi4At-X%YI2O@wE96ADLndU9_LVL_O5Lc
zbjxxvTw0Imb8%v0ThUBNHIa{JDR~#LWOsk_30gahi$52mbxSF}+(`qqUuM&!Vu?LI
z(2j3C{R&nz7*Z0XSgV^#%%raAn?p*|?cUqqj=v|QIKWWOyyPZ#58+J2Veto1bw?{d
zUXa`!MP^u*6?nxr`{x-0JxIh&Gt?Re8WofQFDNoEE5Ya39z{-x)s1WzE`X2BzX0Po
z*R67hN%!CkLyZ}IW-=Yb3uq}#pX&}dntcZjvqMG=h2<hKBf%%+fx;d}u;;-h=kS8q
ztd2JeEvWT}a4RE~h*4o3NW>4|XC_eJWS6oqnT4@Ee?65egq9BTrGM<w2w*)v{v};-
z=tna}#8~}}!$7-d*;%57<+ZiYfoD3f<eE+<J1q^hs#=geJC)AJN+pdR=CW>Q+Cm}I
zvMbzcXE|}ITRTYud#M#joQLGG0;O$xLIsEe^*@P^S@-Ma-J|Gwk30Bi4?J?1b7WBu
z0%uY&3fuiEBMN=A+;yBw`R9nkqQX_=rd>FmSXTP%rB8_@BOA<n93U_pVKrs=^*U1|
zNdbXd5_bTWTruO;O8$(ce1fNT^mEw4DsN%S@5|-K{8A31rDhIYHqfXy`!x7QnS9fZ
z!_aqm;XPBx(4~H2m)i<MSB-0`ZkI>0?dY?Uf8~xdePz>ETIX`=+h^mn(o2<H-EO|R
zHGQvhHWQG1)$V)gurZFfvlOjawjJw_>9!G0Det(!o_az@o++ER$sSOfYqEH>po$NF
z;ZNRI(+XxOGViSGo>IvFiTJiErMIx~5f!f@*W4t)7{@PV=FnQlOWAdwoy;7Tul}&B
z#M;~!A*l|VBf-W4l+$ErUxvOjx*zjNSD<u!9<!Z|85^0D!FXg=jpl>laoZR)2`mb^
z-8Ad^?Tg8B)h4$k#@JiFWI+Hq@hl4ogyG)o+osdD0>=D-r?O}hyG4_>%%ZI&?NgEV
zE==I^WPCZNboTi}2xN}qZb6w2!!ae$RaC}Tk8)3Cl`d?PP36tf#H#laxVEs4KjiZM
zy%&2Dqz%aE*X0Ga!((>^073~k@5MF}*B;7;i0CF7OJ+4Vv}ul`QO0o+ubz%u{zux=
z9<8Yr)rb->qm}uEt0$$SO$L7je%#W5dviI5)vSGMGmIN{159?!LYilvU75WA89kpf
zAn~$Zdm>fO0*VyZ??4vZcnt5bKWmj~Yh4;=uDe2J3u}nak;K$8Q*W_mN#-2Q=OV?-
z|L4`~onJFBCHg-=W)23X|Npv^js3r~YQECubUtiHoq3|hm;o8s^=Pju>y)ypEVsg4
zQ)}I}Smll+za@K~eHnxeLJ=&?&TCNz{2&Mv-~GVb4h)vlC+G6*@b>BWa6x3SG$L8_
z@p^MLdJWhoFZNI#mt+=8F<6{zI~$!%aSB}}))S~i@UZ-}z17aU24&U=;@usVd%@D6
zq36(p@z3A-QJGmy+oSZqMGE=CAyw|yk({n@U~0q}T!)e4@A-5PH-6yPMaidWjH0gC
zh@&SqAS>l44Ereae|;;iemfgD^nxW=zt>EiO@a1>YFYB4oq<5M`Y!5Iq&w(p)~hA`
zT>|}A!O^O1k_$sfm4wKE2@@e&0AhbglmI1Q+q>X+6g<#rzJ()7c^F^t6HAaFgQ4{h
ztaFVqNh0+c%1~N(kOdiRYHf445DILw0~XcdIao!qL}@LJt4%CYKVX?!S4#3`yic{@
z+yaFtdjtp)JF9p$v5eMg4Dr9e9F<{lxV5IMZ{uHrvf=&L4^<r`N%U{AI09x1Z{lbx
z)4v{g_w5owd#ga1@Oq-V4iLsX{=#I^6sI*!2ixzZ$psXNNGv`>+bj3fq8HFG%GeP6
z#_o8;m-03J1Kv5-83;i_Dw_Fi>cS2w=J{%)6~|Z;)z)g~iiL|SiR%<}Z){x{43Lnc
z;j=D;GP}>tG>7U@ooiiw1G@gSW1lSPp5Y_sCYnZ;b^6AO&;6-D#iK7*Ue7ej^pYrs
zv#7^SS8rJE6!Zn&QSl-VyUujRD=a5EBcpl5JF4<x5oiwCvu*75w<?-OPOWvm@{=g_
zmCU;+<5GP1l3>8BT<<&`>7=-+I_;EWlthWJIbo~4J;RGt*x=5+zAsl;<(AyL=4zwF
zH@YHWO?T&gK~iaeZS=Wxr65*uS*F^mY#~`1t4U+>n(10)oYjd{_Q{VzB0?We0<j&l
zjrH2EP1774R#U4hK=DfYAPv5*9cy-mwQ!)TMIB~(?@YZC{s}B~8R97fitlg*?NF61
z2%<3lLpEiC0#~}-N@jr{%U0W*hPbO}RQtu^$A2`x^t5CVx#3o>IOt-;?umfE{pL4p
zF%;_;F7I*Q8bQ)S=-qnDhm|jW6`3^e%p>Hp!}(07h1N{P=OXfQRV4oMRQmSv>GU!S
zs;dMD==1r|#{B}o;DL<MT<FNECfE1n`~Gx2Tf4;7`TVF@T9Pv#j#wqAP--ZWe#+H<
zO88F{ciV%;f#e_+%-Df!c^3`=fQ`ui^V!wv`twLyERgy7eTjwH5c5M5%6fpY*G?*$
zH`jv4F-Tmwh;{bXg+%dRz(Mzj|8>8*kM*10G6<XiN*Gx!nMUjZP!bghq9lvkP~YtC
z4eHf^;Z%hqlmBO=T;J{>enUfp*?T(uJ8fUM+)u{$xl1kazk*iI{v-pwCc<<%I=n_j
ztHj-`5aY*bB)8V+VCaIjVWJPURws6GWMUmC6g!A;ocng0W46`EEtSm|M9inY2Ij?k
znGXg*+<nGDF>MFq{g$EZe-`GRQWCnL%>6)IZL>ZWm?ub^Z4-d5aJ9PAI5|1h1b?6~
zjU@CyoBM~SO_uh=92<Ra!v)U+;(M`uFJk+_B{{%AlBkHHI^4njaT>6+BmNpOa@Z;v
zr~o}jWKKFT80>aRA>)KGM%?ZNO7hVb*Hjx&gQF6f#OBSSgh8RJGMp!;Xw|L4AtEuf
zrneZnv}pogkVPCawj$M#2t~VVdxqBHkXH!vfGNaO2S>1D%>7|HAlpySFyV|CTs+Hj
z<Okl!Mo3!w?zlUq8o2{`S_1Q5R1s+1b|IUORp?1BFeKmxG1Qqz;fo-kg=iO_JS$IO
z@Mb}Iao}IxC1G=RficXZ0n!xRq;roBIM#o-WZY>P_UZbh>#bCZ*!)d@lyeFbyi6db
z-P>S%0^n3auzdoF`fy=Lbe7~b$7VM39rEG$1K+Z@?Wz!rz>OcAWg-HlTiyZ6THi#~
zafK@pxt<fES?h_%Q4l@Yb5lU&W8f8&3m{d(X6+3AkkJCVnJ!iEU^J;-U<POhPZgV(
zSa*V^rMae{vrQIV-*Ta2)Wwry#_OqiJt(&>es<O{BP`-GrLmtzp{7Q@fE7>SJWLIc
z+2VL`GG4%O=Q@D4JEKPB^Mw+z>dYL;vhBYr_a{v8s$p1%WD4FW8nZ}+_v$9ltA<5q
z_8}>Zr?Hy~NpsvQs!N|aOj1u(Jq^-ae&N6lw1FEE8?03r0~Ya;F+@?8#k6TJb!k=Q
zu1-gvFFfUKbWPhM&XMUgW-!`QkGO*=n7F<M=%hxbl|tmBHx~pBxJS02fpCC#h!(0k
z0!&Zb0us0UFP|PqdSTi<wkmDP_`a+zszd<!4q6jhnwT@q!7#bAhvd#j0bY0saLb<M
z+y1k$D28M+Wpoyep1dL2*wta-N#=d#%1N8+W?g|&xz_bHF~{a2ToBh#-6%yhm{Ge=
zerSggnpsW(+#S`>Dm$4TlBDA)g`f;1%yFQ=Obp#*x8HH^2{L`UDPEs0R(SCa*+arW
zn@_b@7I|3Su{>w!GQ%Gps@nS1i+h+L{YvHK37z;8Lq0AZVTg)&%w4-x_~jem&n!Sb
z8HI3<J>!Rsd%E_WdGY{6XiiRd#5|g=E*F?O{DD6y`s*HY{cd;y){L)Ojm-tFaDCnj
zht>3`TT}LwvY_>Rv*+&jTv7hvYIsDi8}}1|2Ih^^OZA@QdWDB8_kY-E_uTKN`^Pb=
zs{RQp*Zfm7t2T<SN3X{0$lD38%D_8x&xnBQrxAG_#jufs^wme}n$UxPSlr7Eb2-4+
zrexEbbu&8pHOZ2zEsdEL;=jbj8q{42i`Ynz#{o**z?{XXcL$hpGpv<kn`|zr-&lj#
z&cf`<!J)geTKnP2K8;fzQH}D9L%NFE%M$(j=-C=J$DrBL*xGlp%=T-<m6Hv3LRNyr
zzP)J}339k;5p?#wy_OOd+_FmmSMmh9w+3XMkeXAXr_X0$QahEaD)UR&aA>K*{>Z=)
z>mK|~I~ZUrR;M1I>Gp1==Gox*U4~T0(BcU7BnLvP*yqE1{P?mp@!cXmT2GBw1z6L!
z`W$$-N8#L<L=Hy{$O9d|@_O}IEy5tr5q6ZzuD+)2t9tF(*}*@z<vjE@rh7KozTY`R
zkD}-6kjJM4!%;Zkz;tuxc@7hRR;kU$^3>!*y5A6hDU2EyW%>2EhY7ak?4jdvYa*?&
z4o>KGpkgftdeflyT8Zgi*WJJOmzgm%vMs3n@y!`f5c&(3onMyFMEy4qVajL>gSNgQ
zz=Zp(1$B&zw`q*bC+Dg)fk)eA$Y48VzOR|;9WoC<TqD2<9*-A+?-0O*Z^$!VclbjY
z<CMUC330qea|R5A5EXGVeg<;~FvQyiu}(s8Q&X7f6c;|;e^f`DGJdOg43V3P(i{dt
zMe@ym4<+ndubg2d@Dd>;k&>J0!CVD_BJ=0xiZaF#p>3Qo&B1Z1m3h+l&&O7rL5!O%
zHpA7xaqx2twpdAX-_8O0pzltYN@&0%!bbt6;*B`{NW315bJN&$_RS9f>%(8v{|I3J
zgAj&;<G&QbXm|W(jUYaK{0d=!_at->ad8EmvXC3|#13q0NG*>nU}}qZ)b}L_0;&F3
z{VQ3)ES69M90R&Pf>pr#c1;MUr!Tjs_wT^q@Ib7&f4{wY2_688%$M31Hiy@v+>7Dz
z`S5#vQkA;8<=O*P!U|3w-2v>o+8{3mdR7qcUfLjsuX<+3q7UtxJ<EGPns04BH!d^S
zw=Si*lj7}3BAXxbJJ0F;4WWOYjILQ!f35xZ^3j(5(>(<p05BlH^ZE55Z3R#q0{LE0
zggVnPi2uX$`F+1C79=5!xIq8;$utKeL?nw7vBNk2`NZz|3eNT&lq(Ap{&QTnqwG#u
z6dUf3{~h*7dg7rl=6IIV0z6ua$md0Z7b+<mdJym#5S@aS2~_S>&)4Vc#B`@#%6(WX
z1fycKLE2(r(>Uho;L8JoRsP1IT(?Pe!Ha+^dh>F`J>6fn6A>C~igJ48j=F>)4PH_c
zDMTdn;_M1cYYqDChor?uKq_ugSb!#%3uVq!?z0!s5~_^sW#-9<f{0Rmp)t;pbLw}T
zL*`{uLyMF=VxA(`MM`ija%neXM*v2kGQUPn0Hzmk7wnh#8AjKa>HYUm7oCNvD%~sJ
zukRrBqu12c@n|0D5N=Aho=4qqcL1eJVCrTAa(63c8(<Eju*i!c;54e4RbmtJwh++c
zY*P{_M`bFntBNw$SHjPRD!7^~X$qKeQ%yj}u$joV&X>vaPdNJf_V{_1=AVA6+EBK4
z$<CtM1}e6Tvh&b1@f8YzJn%heQlJbP4D94jKm26cc|F4(?^0|W8K<M^<Pg=shG||C
z;Ex1q?CjCR4`+$%p)WfhHivWxk^W*w@z}%)&IJgR8?%xyj)3N2H(JK!_Z;w>T(7R@
z5^jVN6X9o0cJE@g&ZRGjJ#FtHttMpYLH_t!tpO{Hwe0uHaleU$LLS-mys*g!xtyDk
z1GX8qlz?~2ciUn#xG8zipManp_&KoK-l5}o&}|Dt*CZ;L7+{uNSW1!Oi0b~Q#vx-#
z)+6pxb=hDom^`EC#ii*-yH>ZR(EGV*Hg+mZL5&;DUpKS09Vah^G7gsq?#3SMU|1X7
z$bJ0iTChN$K}qR7SLyzT_dy`zJTH9oO-;}n7FXy=JG0Vk`<p|IF4UN5q;A+4<0~}o
z7W|P{I~q2IR?<2DxV=_Ij~8a9kA%Hm>;l-`!V=tBK)fJ`KSw{U$6hHO8x46}O6&5b
z^*x=AFG_40CQf>HIB=NsK+PcD5!Y5lxk3|-X{ny&-1OdxLbGOOdr}^oRYwZ_Ffgdh
zCRyai+_O!(M<_?!<s87{#~2xleDyPq_=hh-O@Zxln8U@^R0yHNx64-KLJ{89RLS5h
zs{*)HdU2}F13_uVKtS2eYE%|55~Cu6^-&{y1UJX8U`bVnFtdt!G6Dy)K8M0~SO5D9
zNliWAu8~9@ZHw#7pNu7&?6(q*U4y&ObdM+1dp0#Iy8Lt7-oP6}j^y|7hfqE)@}nLA
zhdc0p-!~Y2+ooNG)`;K_wep7|k~VYkmMQ(^)l#@y&XI`6w$pYaCOjHQdoL#bPK<9U
z>_D6R2qx63oa*8tS!b}CfzM*p2H3iL(r_DAeg&p_+ttTEE;uWM6*dyy9zzd%xRMPG
zFRTS*?i8-KG_ogql&JVRt*+gOM7TFQv$(l9jQQE-Hz{3c^<^VwlsZxB6r#CroSkT6
zP<75B=GLRxRRvJ`P$g)UhfmcR1~a_vT^bnEy;j<m1cJGNiL|HdL$|7Ek3NhqLc5P2
z<~G*`AO+fT){<f>xf5W@*!<$5Vnt%u#ZEqH=5VHXV^IPQDb4e=uF_oH9j#TDguAIc
z<=GnIz=eFawBqTC+Ec~b`oWsi`cErntxLv5?|!A`ZaQ8uC^x23J7}2XX>;!hJMqqk
z;sizS>7L8Rn~0P&ULsXX2UbdB&gco`(P#S)Wz`d!kDxUeB!<Rm-!=m?MJgbRK|%=Z
z9_DH>9jr$QS%B%pNL?Txp~<oh-4h2-;go9oTd4x!tR04|P%ktQ`Et)gkFufl><JNT
zM!*njpu)B1HyVs}E;5XDEjody^=zCz^jS5rb;!;2mojnd8v;*pUJGwS)LSV{cy~K4
z3+J=1#q6YGYji5@abmRFe}X|eM3~!is~$H}8GrSQL8iV<o-o_xi>B4MP=%8$(oshO
z`Pw3E`A&}6(Tw&O1EbtH*b1`u&3Y-Fi?sfFW@cibUX+<NX?{^^);O_O%X><t)y8+4
zWiD3*)qV(G0j{1LKk*9o!&KA^{Lb43l-sb>jFxl)*^hq$Oqkc<*E<dlkWHL+^3Y%9
zA^?eI&y}1AH%Nuu6Dq<rB@^IHH#r2Kqz?Gg$+&gQ7ENgpq#<TU;dpgq4y^90WheXb
zHoV`4Y_^=>Qvdg5&0K#Eyrr%whS!8KxWC$-NG)w%#~`5Fl<m9_Y9h-&3kd`wM!PUS
zeFAWXMY`;<w$!<|rV#s1l`Ah3YV>@VC*tA42vG2K?zcITl#;NZm$XJW=qe?Tn%T;n
z#nu8tEL|dS5O~FLh}<LDXu4m@r*57|iIfj<_m1b3Qzt$5ukS9Eo4DSreBNw&0nEz_
zuNZHDr3=5lox4NmMJu4HCok9ktnc^}x~*V3s>69_^G)wtxg$f|N-<!C5?>?IP6g%b
z)SL5P9{i}ElxU)t8gKoao25lK2kp4!Zg;`$=@g0X!&tm{-_T-^->SbuR}nLA55P{;
zyBQr!bHy51tN$SJ4fJsTb9=;Id`$hr-g3jaP8Us@6%t!ULKW!>X!s7Dl<zAss<xVT
zub<_6^*k@w(J|`z<oyt9zH$&fx(NN^55!oHIH7P(rJCj2oDZX1oD}`>MHlky??C=z
z&AlIMy&;@Ddp3gBHj@!(TQnu$D(<w?7I0f#8e<9ScD17m<c-yC>xi|2o(FE-try`q
zpRIC7D9On}ILTL$@w`eAQ5=U|?fF)b1lh}JkaB&sv5Pe{>lUA?yvJS#QLH84zr@*c
zg+%9f{ZpZf4y@L;#L?cmBp8^3ihwq)A?1(SSswXg(^sFPTAQx=4;jTluICx!DsCGa
znaJ^+AM|>srUKJchl$x{R~*;5g^|&#g(RV?XD7j{EE?gV98R7gDrQbf4E;7t6@;Ra
zR#JEzZqeQ<+lJmgR__HGzqNK+HpWgJ+|hda#jW$cMim$$nx9zH+e*-P<`t0^)csPK
zGB)4cdK>p}a(!!9-iaq)yv&_KR<>v0?A07)-?Jy>066kGGpl#!!i=B4VqK;ii`zL(
zPNx|L95OWKtlRAN0w*SyGK@d9v#ZryiVgAE{GAY`ze;U>C7B2OB#_gQvM)hn*^1UE
z*9lES9d4SE@#g!g;97H%?-&I6z=N*dk-kc$&iP6iV#?K{OoixhCHOzf>q?5e8xT7h
z;<V1a%<<huN7fe>M;E88(%T<5&4*a9PzF<MjQeuZc5`@|eFbV#Mh!G5l7+Ss-rhh9
zXoZbs6l42mtMseI^wpp<?u|h$ubp15O+%I{dDMwmRmJR9ev7O&l@^uy?x~LLB{GN-
z!yj%Se~zhB8{Y)z_#@r+9qsCGSEU5)J9LlmN$!EaWw@G!8(!TRnf|qJd|w`Ch8VG@
zE&!G6`_t(L<9^`0i=+MBgOnB-vzJY2(Scn+4Up($b^RtBANqLVl0_ePFF2Z6#L3%q
zAdyh4I7Me}+`^$}^-C9?2fNRoAe%UA*8C3Nz#ILO|FzCNm5O#z0OLF(9*nd=^Oq^R
z^A#Zv+jb6Zo&0uFBYyscYI2s{=jk&R>Wv?n0R-hgfpSKU|L<0rk(1@WBhe0PO($)z
z!SvoJpMs{5LIm+MW6u8TTAMmtak0<LTQ5sLp3`}M2bEMvqTG0IT*fn5LwQ5|ga`mI
z+4@>LyIx#8`#LLUZ_ZT@{F{CKr49Zm@IY>1=d$DpaD>qLa<jDxuZ2Cj8cAQxN~E9H
z2j085he_xF&v6eLZf-C5Uc3kW`XT+U`*Ad{l9y#H_V-mVfmOe<96S~>`2#Q>%>DK0
z>meEKF7SromxAH{{!ytLuvj25dfj#-e-<dJ@Jo^hZtJ`r#ouZx&-V_As!^5Ld3{(U
zmAFx1;1l%wQg!cp7@Xb2G(P<v*X?^*JiV^JM$a>V<Dke5OdOAE3ve}8fl)L7hZv}6
zj=dm6lRA!Z+GC@#K&=d&ra}u)mEjh5DKo0k3QK5|6a}lNGJ~I$%YM94wT3E!%g{=e
zD%Bs7GS+;e;cu}Qr4+l*rxco<S|Occ5H9Er%&j0)%1UihLN*<@As$LIT_d9z|Ek@L
zv@A@QLSxhx5yFt97bDP>EIVMB`C~eGA;iUA?;R#zo$78z&R46NMU{&YIwAF2<W8Az
zg-8yBjM?Z6!ZDFl7Mqb6BBjPR$MpX+R4a}p_eNDlaVHWJbbHg-cDlYzKW+6+VEhqO
z8r;agV-y3-d}PdQorWv^1l}Q9#D`|vr0K1<Z0d_%RGh3!(o?@U_j=D#<dDO%4e-Ks
zpZH3=tbSMMwZdn=Wvzv~M1iea;=jV-dGopz)!NLXkbed_RFiLV-2p9gtIReWcdvE-
zs;6t9=66s@YC<(I&>3r@v-gYG0JIS@lTOB#VCbZ0=&|8uyM32^hHkx>xt7b>(3U06
z4Rn;kyT|qFwhcbMmr*A=v}bo>0KamY#HIK*q`y)}byNHlW*TD#DXc%X=jy?7+Ru2{
zx$T7SgcwwBE%r}42xIR&TL~=6h2z#9)`ea5k=L#?t!t1{Jj0hSI^no5GD}-KCr<w9
zfS;KAvh_rgDX7eJR6%kAF?E!;M_0l%+B08nWF?keum9d9vm2Jz)DY;+OF7x(&JjJ&
ztGKK=ygpaHxiVTChNZqNnbGd54Dnf8DtKeN=4}9Q@BVB2NaV_C@yWa^vDqG4BH%9}
zyc^^^vQuIfY*Gx&EjU|LU8Z)?#M|~iB>jC1a~dIm*U!crz?W>6OQfDg&IGIrm(eNA
z*fD5~n(dv)J}kHA<g*!5a0=T`XF3tE9%c?$-2{f>qO7KGJ5VA=UPiNg^3x#!!`H%F
zyC)ADW&*6w%${NxE4fy(cM|TKs?sAr8&{MjPUa8paXih6=A=fAKvaF?o<e2VDRuKu
z0{pAnl5{7XXAx^E%7w=k$DL%_D@NP~6=g}HZ?d-!X(;8sF6;}4WHn)$q?xda$v)p+
zeXJPhZSWn5I7j<x(%U5Nge9$LDu)RB(I<-Rw?>t-s|asmlUmms3YPU84yd)$0G<<u
zQ5a3Q;zLU_LSyb+W~>yc(9fwak?4BU)ZcI$4)=?3!g!s3n}6D(*qf|QP{|d!X*V94
z8RgPssndEidplg6uZ?c;!W*%SGUH_$p4l#eo2p<;z2Spc>u*Wr#AVTm{Fbf<6*6f=
z<2UZwjE_VYXT-`k7+X(@!(Li2RhcOhYo&ZHkCGU|c}*lpD6`wel8`+XCdzMFJ7BHV
zpv<u$@h%AAS$Vs2!bomB++OvIp=%ARvmvq<xR(mk{N#fxx-C;hc&S)~OILNb-W}hm
zAd-KbFS6B;?5n~}pCAy`tHPghf-<Am0aa^K7^5)`78Ed24`i-x#Fp>6we8WH<e?K|
z8|lANnH0SykHELOtiZ#RV*_YMcs(nKS^I(8X*%QtKTCn6JVn*ZOAiz{lgW$FS5`xn
zXrXsY>bw04tO)cTDfv-tEBeZqrc!iGHiQ05O57{%71|}&tW2lDB}C{7*o<`<k9y>D
zV7SBXt_%QrvAy4CS=;Uz2r2^jRzj-l=2J*e3hJ-l=iJ*%>jw^kLD%K>Na}}g`@nc-
zj8)+$0v9&sjCZB!1y49D{GkLFk}J4`qfbu$U<bU8v``bKNujA7y?i8tPHgj60i#l6
zrM}%q_t+TD)@+50D($isUJ;^E5ex$@5f8~|!&z}LLaqtQ^+D}24E^q=`efoNkcsSY
z$~9}jpJZH5lgc>;V0u?;r>>i_fct`6-yTlo9mvIWx%3_&-pVB0Muv@lF}K;I`m2k%
zZWR%l)R--U)fUaqyyDp1NXC-m>acN?yBByD?CgFurxg(@6?s%U<$t}abj1JVYe$Yg
zN@%<!jA@A!e1jFo+N59k>(N$o_LboPLt~A{v*4RL0m8TKdo(n)z&r9=RD~ucp1gMW
zFQ7g+9Q;vhq-YQKThggqzViNUQvQPx3&)2$C1-K>@N)z^zI=guB<N5gI4!uKWpavy
zMVlAx#Xj($OmUsKpsfR;i7GAm_60dH<Gx^H#V-7D1bT;9@U=+lm7fs^*})nZm>W6O
zJH2hrT!U;5JNXK~UzNomEEMA#Tz%Hvp*^`8T!59NsV!*jsr+wG|7C~G>h?b=8t-jS
zMT(7W&BY(kZft#_|FFxmGyFevOBGKCQ+jzLD`gj3dN~3{1_pXDODAU+0!|KgdKptY
za~BH&reA}a|4K%)tYv7A-GTUZt=~@%SO+?F1T#*ki34<)^p3R)2fL5bzQO>swoTT_
zkwjfzKmPUJyC>DU5@jUW3J;qA(o{{ziPP)JJQ$mv6w9n9#h6lbH|}{5tWJoLF9~W6
z#0=s@mR1tPu1eowf}~EEt)r6pFCn@+Ts>pnG&d=U5UnOD1I8247oD(_3-i%v>1|Dr
zs!H5Q!p9dK#<ZN-JBC(H{4WMoN+VdbVR}%nbXesNt1y&kDNv=>QLf>$6-A9{9rn{4
zbkPkEc3cs00||7cuflN(@5*hzL!>Jj%wU>I5EQsCx&hs#D&!A{O;(dv13rMaP^Gj%
zMFSn;v3FPrv#PqPb+jF^QD;pPc`iY@-tg<dd0IqeRE4$?Lw^Sskr0$?h`@xa2vgv>
z6{E0JiM0Vp@;cB@c*5gqNOJ5`gaenLPU;xQ6Cp{=wBl-jy%1ezok7d~ldci6u2S{x
zw1i{>m`wEM?7r-yo`0%=(T!P-a*U{dXkXD1YeT<ATNcyb)N1L#sElggTN9cNsqqME
zL~DQ~A-Nam!BF2XRRYo$`X)qqEBr9YIbaEa01-9FnM4XJz$t`E27p^q)g0qspbA4+
zoXiw1AP0H5tvm#BfsRLT(2PN^VN!w$#~=#;o~lEF%6g`Rpv6kX5PeZ#f`hT$)j!~r
zSeFE6+>C3M0j3}tWH`hmK`E#*<0j8hpdxB!qFP{6O{>LI5j8j{yqK?Jh#G&pAS6w0
zK5nwv^Xb=**Gzsr*&YU6+*W4daEi8Y+5N5c*k<nnGPvyY+RC>Op|ybqC_%7yQErMu
z`zXLJpm1+k!_~=iXI>UP?h8HLct0%J^W?GykMZmE&v$ON-M5C1^s~xx|NX@A$I!1g
z-Hpka-(dM#ZBLC9k9R#hUwJHipIEu@UGx2%oh|Zyt}W<Hm{5l|#0Ib>YXsI)F6r{5
zbj+5YoHV$wVE=pDhqx4LxO9D(<o^LpL(WE2nXpT)et0-w)q>RW5*QwsW{(n6&zymy
z7b~y5tFq5hFRrsTM;`n-(C2sM$8#&LV_oXw$dKcraSV_EA?)ZJ6I6V6c>pKKV&OOk
z)(D6&{s|x;g*|=`&>$wg{!XueUk)jk&sJ|bJ2TwCWv$o4VX<GA-@03%?7`}>Hhd|u
zCf_PW(&8V!cy{t)<+goFy+Hrb<b5A}ROXcej6gIW#o!;5a3CrYZ#l2jf37U0>vLCW
zioR|&Lx&jMFecFglnzO-{f(bTw5SMvpb0YdPRNlRMa4d(;Y`m;y63wBC+xe)q5rTH
zry&RO=wHLj5R-yNsV(b5$PvBY+@6`gTzqY_Iy*i5^Tu($%|6qGAQDDE<_<e6(suh{
z(10UkgXtsf5sbCam~CuEo6;6$suxK5=L4?6q%)9ocx!HKo{KgfY6Bpt97YIW{*Q6=
z`fee`MfsV%eBKB7?CJeWTXJ<g`PJo0X1Q#Hb-^++`UKg<XR&~ZfLd^qz7bu{7rj6K
z8*1IqYdXGWGEb;$MF3dC?Wgr_2J$`=$G;mOxIPfLW86xrH7%{QB1_oPa#7WyCmf7R
zJ(1GvByt}Cl}`jnoE7mOThJ6&weB=03UCDz+k$baRzGA}qin}_ck8esLS%g0AYmh|
zS$(|zoMU*E(VDV-TS{xYK_)8Sb<os6dxbO4Jjot?2t&f5Py%kkbU|}E$O(@+S7QTv
zT&nksyWUOyQN1K!;45yaj|q?u&o-RNu~N-(^hK+ljNIkp6a;$<p;q>Xt#)sJS0njy
z{A~Z(^`^B~;Sa;IVnAMl>2I&z;@q{3qN|(DIe|4QWm{Vd_hnf{n|yt?^2PDjX3j;c
zkC#NNT1GpDhjiM~??SC8wn-~$2@a3?*hCWTA~C?Vqy@)^DTDXH<I%Hh%ZCA<@NaBc
zcq;@Kjt&wQH?NczmLhpWP}=m=0(3}FkihG)BJQ0DAXA)tcF9=aT0Yh&^lIAkt*p!7
z<-GSsOdEx}o#5D@ng&G(>;}XfZu9&OdZBLOH5y8tu4T~4)3{Y@y>f*3Q#Vs&*;}cE
z=0G!wd9i2xMUh;(TMq%Chu`6P6HH5BSvsTWF|w@jLeOmq<mtvCSQN>n-0n!S+e>iG
z?|3_Na8LJy#68}!?y-Hvw!X0&34s1C+IK^qrEikiWdn=xKhh$ST9Fiy+2TvGe)893
zebL$Ej)n9jceI?W1+P;PC@!VzL*h#XUXvL87k;72^cz_oHRjSU<<4O7+#WcgAWjey
zhJ=qekZdT}r8U}LvuTqF6QXQ|_kk-x&O=UqC_#9C-E=3LXmgVE1C@(%cTv)Jn<>yu
z73rRbkG>p4W1tihXgc!1MH|R~B!iv@&k_@lDmCKnyj<}Gbp%#Am!lOB$d<-Ai}iL@
z@Cvl<B4~Q-^EMC^1UVa)>DVZIPDEbXSJ^KVx>yy|KeAHXDoGY}Ro9JGR5q^J!c(`*
zwn=H6J0T}O0S>R7>r}alJ1Ha15t_BcdD<<Q$5&V%c>3z<FVaxz>i;U%o~2jv-;!Bw
zJY=Jz1V2o66|kn;1Js(fOwb#mH>WuTul$i9@&$F>*$nI*xBNM@8Go;351MHy5xt9U
z^(Y=Ncv&V666vD`{tH)AD$8XlSfQVLNL}qM9&ADB?kZ^70&OISPxM*a9J9{838b8-
zezy@S*ZY9H5&UL22Lyv1)xsXw6|U`Ox-4uN?l38hKlwQ61j`<q)X{Jds^_mf0*Z(}
z@g$om&W!ix=ckKO8>6$Ga$|jNgWa@vZM;1@)+~QVZ)$z<?r3l;L-fkb=WFO1I5VJ5
z?B@Yk^pa#46^oE$Vjr?ZzyK2+I>Rng?$4xtd9rCJXfB9@8jT~ctchROcua{Z?Kr!z
zLNz|<gBU)j`arxr*X`$Ue9(ql-(TQR8zA%Q)i*wmXy8Fc0M6>uu{_K%fi@@$5+wb1
zlLvm1l+wEYfRH$T4fz-d?2W9TczFIV3Swkt{;yEbG&ZNxq1eI+wZZk!ND5VaCknAx
z0*({W;Z;PQS{j|WvBb>z+eGJ_z@8X;3)9?b9WgWY%>;?-i*<RIynB4TyM7OXyPPL5
zT-T~TR{|t$eu3#+)(jwp^jweMS?^B)tClC;kzsA*q2vA5<i=(n`RyMdyu-kcm$<?j
z@)YpH59LqY)zU4p*?r+4fj!~CnxjZ+T#G0MQ`|8-hZp{CKkstrHv)T%?tJ|lzdmbt
zgILv)j*q)Z-5diTQ9N!EQt^xL*N%^)YkuAkC{1eozVVZC&V(7Nc)Fm^Z}6=zlH{EU
zBdFdFeXaLW@#3zX&Kr<^G=q8<Aen*71~3ixiXeE<OVE(Q0+#Rt0yiZyMa@1q)MltM
z&}@~d;EIgam5juwNUK7^KwA}rUDaWce>4MTicm23e)H=QNevWXlu!%>t|X6n(B|V?
zyFri>VC_<T;)0hdl7N$dvlS-QAKW=XqSY8SsTwXoG;)EyEOu###Afy)VPHj5QI~iJ
zHZ10Fnnfc^PmX!Io)dh2r{=br@SuO9y(jAc(TqQE<p=f6(t1Ub@#4x7p<(JC$zC({
z8t^CTB0fs%{>C$%M;{d7Mbl+r-+k1Kzn6yJKsS$m93mW09#9EO&Qf1Xaj(u3ivI)V
zbz1vSd6w_!k8}%xe`;QGVom8OGl*F-)6O2l(`dW`78(e#G;MsSTnSJg+q~X1Jc;(A
z;M0XJ0VDFE*(UsFLUsSX^qB>VID&0zV<8NpfjYK=FfVXbyTFABd@f+Q^@WHYBgSXn
zF2HCpm<w6J^8yyytbUg5p|N(5f;ze`upB>0Wyy+?I@ZTPSw`{TY)A=x`b~`M%yR!0
zgbPR<k9GLBsf36l)L?zNCcap-7)MUC@_;!$yW8l7tlhhZhJ|C|UIsrhY<?^O44HU6
z<c7?gVLfJy!-a#Ij0CVY^waU%ASs`No%5fjXO!z0Gh#*N=v8nQRr6`hioElRW~={`
zQ6n@{SXww(&~_wZ6jjrqb;jP;#jN`pC&x}?B-cH0_5cB9f{A&UBB9|nkZH8W2Mih`
zJ1OTi#JGRxe$?o+9nyyAYC;$){ER)Kn%Ff=n@>-ms1p3p9a>MEsFQZ$BVrz+*SHi`
z-X51Bi6%Cy`P=c9w~JjgUn7#;T2DXdAe2zX$$(d#5dlSCO7&}2Y)bq1@H`oKc8<Ny
zkr-y^Y=_{ADGOhjN|dpgSlMg%`00uK<orA)DPavE^LB(4Lwz78Cuz35MB?e|P?vi1
zMG8v{ty3f;QdARfJY!b4)7VhKazI>~NPwZx!&`uy)}d8HD2E)%7D?(umr`N$92EaY
zV#@PvK$p(u&m@9yRJVNQC}fpEz=?AuSXafCnG<&En!$>4EUqO8_%G3NHaxBb0USt3
zmLAInS`n>Al^|{x#r7UI+T&nzG^sh|iLubo8=pBHNrw#PPEAQ3hqPK)(rBd-=F~Rh
zNHP$bmK5M&D#>A#rLfCYvAWPjT{Q#}^-i{$$VOq?UIEoGjk>9;ipmLNNN$28*yWW?
z6){brrlpULVC-R&>)3?ec!#GOxQ4Ju)j#vXb6j8$XP<JVVnXeP`Yh-8DGlOiT|hn6
z1upoXr99Be+>YvkVQkbi>3y8hk<TG(i-BWggMB*5xas;v+JuT-zm6st#l-BI`Z18K
z+-s7){_>~^iYDm`siGW-A(;)gb|8WHga)5-;J-w)1QOi5J$wXEY$>)P18Mhq`?iRH
z5GQ}bPE!z))~MPZUl5Xx=Yoe-j>^H`6$9-70>T~{=m2%mioN+6i?fJ4DlI>{bPf2C
z(l}O&RBL(Qd|24Tn*Wky1kZ@tAlG~vY?6_psM=+ZAS8F4UcrzTW}89Jut7?^B~$Mr
z@k$DsXco>s&D5k2)G@R64uNxcFjQJ{9Xl*#sY2s4tJLLR%-F3pbJ6~dI(!ylou$|z
zsx5QTrFO+<(K2@yBJ33p&elc@hJ({wi~!q7HsYyfuBB+q^Hu>OV~{%=q=4~Z*yT^W
zEktv{zlUEItMK%RHbM$bUts%Ikc3=sMt%$)omzCV%X+IvKb;G{QgxB0Ppe!U(C{zO
zD0D>JQk|t*o|P{M%5K0a>|qORUj@`o-bwUFh^{Hlun2DUgz#8xk}EQO*eP|vhgwiu
zO$)}F{Cc6YBd`FMD%e9>FkNgO7F73_l{^#*r;}jAy(vXU;?B6xya*tvZ5q=RsAHrq
zu5&)aJbE4DhzzGY?@IC~X-3x9)`q*YdOZ%Mr#&;rrtnJQ57ltvh6M9=1MW>TlV&B&
zjK-STj$`lM(x{2!Vhg@lwJvRf#{OEYTMc^)fYC{~IQI&-d4N_XMgti43buEE+NEL;
zlGixVR6`@XZW27hc{-f}ojSH4Ytg~)kqkN{7L1}n84!+fT!QQXo>4NV$5y1RpF|Oh
z=|bi_9VR2BTGkX{wBB>7L{Ql?(1Setk#L`v5}U}h#K$(auC0skROce2+(r)p!~2k@
zEJl4ch^wv{JOsCi9XAdFJ21sf{58}G)S-1xJ?(HW2KNwY;&i9NCLC!ths}G!Z9~)R
zx<nUC^5yBqcjW3i73J#;O_klCzT2&>rZaG>$}ESvvvh>?fHgGje=3j+Pj2elHp<SR
z+X&qIU5<sOrK@&Zd7-@{Xq~_B#U*b~J4ggHuwJRLn7|DyRi|Ev!C#q?KFt&;MDGNz
z`t^fp-U^ZIoAr-7Ey`Cz<0SPY@@9u`%iL{N`>@}!;UhU^iwK%nahb-XvZ9!KuL=pE
zPO*1!JMulhj7r$t@Ua!^|6%MMf<%eZHNac8ZQC|)*|uG`Y}>YN+qP}nwl&o=yLr9n
zO)gGEX2i+N6Td##aEMe6hCI`NeLyiYQFr{e6orZ9|E4JP%>N^2RzuQelNG-Ere+q8
zegdN#V8GYe5Noum%0@uOJVDaIjP~mThgaCHNwTU_f)0dQQZkO{{y=m(U?AjI?nDP}
zq-B7PP7pR&9RxI_r3PpgO6JE;F;6An*YKqG_Sx(8ag^g<d$gnYbkHwfwxOuQ-5p-9
z3^dZQZy_Kq_8h7Z>h$f<=fpj)Dcr^m8ZppADxO2TyBL7BQ>q+n5%Y54q$LI{yB*Fk
z<ox^7`SoKKmY-uf5wqKq@712d7a}6SN$MZR=c~i3`)d>Us6R);ci49it=OUnQ^at@
z=VKb2&+J5uk~%xrCsMgLUC!Qs?!e8XU$)+aD?By0g^M&jUM$B&{Hk~$Vzk|vHCkQ0
zDP~&pN)e!`Op^i705$wT>oPM8%Rs1lDe#!eWZdl_J>N_Gu?9j2^9wfM6^3xiEr}RO
zI68TDSzxQ6&*tc7FyBc@Z`zScxfO_~`u&>+{YsIDw~E>{ySj-vRc4mDLN9c`%t9gZ
z!Df<qzPNnjZ4urxSo>!K9h2KSvOtk734Vshwt$em3G+4<!VoeWu@+-BBfbJ!v}fxi
zX=+2+z?@dNB9y#FC6F`I9~`|ohKV7WUdX;yDo#@bVGBKtth+7gR?BN2Ftpl9i8dtK
zP-x)nz5MIYiQ{h<+q?FK@+{jt8pz1gTVsvC7)S}((i&lmXPE2by+w>!)R<Yz+Z0`B
zWm^o|giS8WYpF&?Z*5(YFPqk$kQ@DV?~c{1q;=|<RqcrvX&OrIbOobYG!dN2%J(;l
z3(POBQU=l@MB#~L?Hb~xQ%F8pa^vR5)mZvB_{b;F0?w%GKY{Ob;7wlp9)XCH5FfY)
zA0`MFpD=XGT4)|)J7zc;D6l#6ndaK8ik@zO59n81^;e(B+KIlIbDU-*w{;8QiRtE8
zWxHhQw}(?==e<x{SvO+O_(hL8fLqcO97FN5B9l?`Y6DnB_iB=9V=ZOF6v=MtMpME%
zoH)m!)-_W##yttMwZ2scVUNpYm#(Y|E0=P_%HsLhbt>tAXu|+0)23vU>c@RM9K-kY
zY6-o00uPGlyK88rC7bK!g4F}W3`=!qRNT_=Jf(^sPqp_Mu+)Zi<0GP?F6Nq*eRJ(f
zSY4iHsyL$`9&htt#Bx2R1#Rks<JlM^iL4*5G4lqy%G)U63B=Q=t)|j;$r3zx68BDw
z6`wq=IesD+i6hSl&an&0=H7fU-H$KR_L4qH`Wo$7Ch5)gWRpQ@tmdH7>r{{9X&ajd
zu!sswUM~1rw(UiZlW7$7Nu%j4WBjj(v~lJbz;gKAaYljD%YV}JFJ{*{%VE_7<46Fx
z@{T@W6@AAJt8x`g>Elclcu95}rD_^Vr9wqdR-Y(Du)P*BR;&p#<WF-W;xPia@*=*8
z8t^TO^JP5F(5Lc+L{oBK;AQvb#yurSB05}*S)Y10xldtJ@kc)qx%?o40q6iIXJq4G
zIb&_%X>r4I?`~A~YukEouw9ipTvnu9k~2Q2uz`Wx(58qp4fKZEOa!$%kn9}dmQ)z4
zA}*Y5IfB_qv5xkgXPg|pBJMW!Mx3=cT{UYepLj2X)W+Z=rcT(le8<i2;YRH*EO!J4
zJb#8eJB_rQ_q61PicThsw|5w>ZgyYZljv=N3xFD;i+g$LvHIjv62PT)s{&LtX8GW@
zr|2DbmFJvS!E~ael-iL}WV#S5wk&IG-4N#kXR!pArlmQv`}O5Op6w5<BXhu=2HhWQ
zxUg|jjp+%W1=Qo1c!Zg0V_i|6T>(Cp;Wt6vtKgkt(dPDBTQGZCHs^I0p^2{xhvyjD
za)*XeA8SoVAB;!J?WtAVb2dU2{2=lL>Sw|MCkp-6!n}~gvSiL9WSW_<L}V`_sguAo
zFxu_J7VgVt&=&je)uA3u_&NC>%)%~#nAxNRI8ua_gLo6#nBe|_)KXz)e>6^HTTJ_$
z=eN2w@*tZ7Kt5&99!(K03tM!ld2d;F9AH5FJ1AImDm3seuFdtFHxJhJt<F?60y#-p
ztNT4fb08w8Z4nGVBKm5uA>EVHVlr>kB-yEj1*yF!4SR7gW5_O3xAZ=Gk<I$($R_hT
zJR1mF^y{2hewwmMvx?**O(yc*IMmNC*`>KkMn1?y?)s31B_u3VKi-~L>G3-{CEid*
zyP|w4(6hu0$Ls0AKvq|^P<S_XVt_^-RNMhEyd`Ch3m6dbg3esTf`nU9Lob(1Y!PZJ
zOt)mx<j-{$v8)X4v_kUoKF1D><_+Q5goE-q;B0Im4G4wV9usVX1S((`%`x2p<x7Jl
zS@p%E)cU#M&a@bgP@q55ZbT&%tByDofzTYXjkd}Qb>rT0%B+DELWQfcV_;+A=zARD
z+$?Swv|%-h!sm~%mOf{s7{63fR6^ycfOut-u*zz~gdO=4`_+AzA-|f#spGe++hOwK
zR~dC<J!~MdY1abG?8Z-d<#kCY{p>PkB91u`We7F{YoU~mYI}d71Tv*CZe=$f!>6S>
z@!2tEM-(1u!;y@#l=^(lV*RX@14>lWKmX+4Tr}vZ281_GT1K$e-}*X5dp4;JvP7N?
zU{Z68MgU<%Ch~3lB~lS>hV7fb;WLv{K26F^9j?V0{0cG}NSjQOhYjbGl9I|bR?{Fp
z;{S2;K|HZDRn1J@qRiz>f_^wNEny{!Tp1Ms&PpyiZO=R*6i;Rb-DRBs4OM39h{$-5
z!GD9$l6ndM*L32)z_Hnw{<m~uN^>gypcSV3Me!9RSQ7D%PkDR$xU;P6(sul$m4&Rh
zGhN!>=lB&kazA9LG8*;fmS=p)WHESP{!h>R%e|KX9#0PrFQ2zFpj&upLbR{F4;!!%
zIO-qH2{s^{d<RlrF3%sk*P|s3yjS)+zh%H%gg4H+wC!$R$<2P8H(MKBLQZqIU!d5^
z$$d{Vmixl%A|4oItZq>KQ9)zmv|!%Y750ns+ttbqUyqvp&fQws!^#cakJlL)JUM!;
z>+4BO%Ke`{46*OszS~7upU;omjfox1Ac?hfrRvw04a6~Ysr$OvUu#Q`S2VuQG_1ce
zAwMwEK8^a9dwd~2w|^`G6Y7BU_4H3H0<ZcqGV5cr#Ndq#8Fpk0h)nx3VyhLMmW|HG
zU`dm)Gg4vvPWuN}8@T}%NE6apQ_(fUo#Y}*$uncfL3>S3;?dEqTV-vA07X0EHq^!N
ziYC;*!3Z3ZiAEh14#$5M#O}$bD8v3`lN-rp8w<Y^ui&q!Pumfv84*vdiGQ0uQI|R-
z<~_m>)Dnru_ULOdOZ1>Y6{%0_pwy_r8Y9(rOcdLY)Y)M$sZV;(FkXKRfr?^!9?qpF
z5m802_A{V{pikhakQLoQcq_4qS8ClKO}{3#qGXSF6!mzZT^-X)CCToz`$SmBF;1|S
zY}2&sF$SKaa+PZ)Gec$E!c${}u9A8%u;}bqmW?7coRA#b++7Q83mxzIZD*|pyr<&c
zY^gg@-+Q3;S0S_B4_);7c|<#qJyV|AJ6w}TdqoiOO*C19Ikc#2>&t1&%rjGw>2@)a
z*F>GWVTy~K$!I%Vj;`RtD|b{@D}(#fZZ1OjK{AO;<0cO9l^Lc6NqJ@{?$m6sWIzn(
z^-QGxXQOY3rbysE@v3Cc%orJNDURr}`+1UpT*Q5Y?nLen15S5G%<Tfy-<!!Q&&u!m
zS0|d0rAiuu@MIDv6qvir%S*2E6mT@dMjEKMmchKl0n{JX!H(`9h?aSKR$cxz_3xe7
zmMzZ(+&EIod~wE+D#JR-w0ip)U$aH!y^#va&<O`m%7Uf{*Z8%};by6gp{4YI=tlGN
z72*2IYSs}3sO%Wza*E^R)>UGsQ~NN%4d^jgAk0i`&*dB~t>ezj3K2wT2bneHO9IT7
z$nFxt3_<;|o;fv?BX{jX@TSPlq=c!xa>HaKgqD4zLn_@0Di<Z-mF@C})>kJnJ8)ZB
zJw6$)pf;IsJjSfk6Z4JD$L9&^ml`R+rKRm8{A`i3^z6UmhAA@>ZeZuez89X|iz`Ei
zZ$E3^7h8sFE98W+Y!hTKpSqlX&RXM>#*SLavbaPdH3cylM+Y>R#4r4Bt|t*wd@9`A
z31QyN>>?+f<b-f$+>=sbo3A);T%5GXdd=@~Qnzlp-_gj@k*}|xI1n&JwzAW|tg)y}
zg4)h~2S!&R=b`BIq*-*5f1MZ!=r|r0yCKw7{pfkasiR;eP}#(GdrEz#cB9_mCPnO)
z2*xBMTYMiVW%VU+1biChMx!F6(=s9bAg3sUQ8`@T5GaU61SC;kMp0nW<V$50K62^t
zf)$Qgf#szfl#3Goh!j|2hbR37%b5k0LpaM%s)#)-)(FUZD^yGbmV~^f5|z_bg5ZnS
zPfP7F%-$y3ajuqM!g1V8^F87*+zyGUby5*1nIxc|Y~>^>;zda2)S4L55Xh;xDJPj}
zkt^g;nna6cxs|WTlqvC)+oUB{<P8hMEtN=9$(T4RATpQQ&&0EoGD;_94WSwBHqAvm
zmbwgtY3*jap+vjV28VqojK+S{*(nQq3s|t5faB{NE!Zhxgyn{0i+V8u<%0+U65Wc@
zd6&JFEM6Ahhs(IJBe<1=z}nCX2eCb{oqqD9X;4N^6u#gL^^fwKMwX|W5fxf(8vf>O
zNa$$;Hn&luHrwd_V{nJFhtA}2w{J*{V)_f_f@TTN9@MNDq&cKTntLCxSIPlg_Upe_
zE#zBDd5}CSsyCeZ37vL7lob(}eN*dk)mp!Z5rW5*rm?0f12!G*H*2pVJrp)`+*xv~
z2YFfOE`dgCY5OwaNxW>9kAZh037!8h6iQC$y~6x7zJdl5v$cY##;ml!J0zHvVyiqa
zI!2Z`a9i(R+fnVVK%S<`^i4iYXMUWl$gZ#~FE4<~8F@`FrS=6_FJ;4>s(w=S2u`g8
z-}6_kgzO!|@8yZJa{j`MukxP#4ZQMR?M_LpsH7dX!jb1XM48)eDjP4X{52Cs$0x24
zPWg@l-RH||N<+uvKmQ+$3|dY=`Rky#!<g+)dFlA~T8$V>{vPr@$5Cy}9B_1fJ9hN`
zP#qP0fcu@hikk4HZKQ@PZpq4kIsr-Hw3l63T{sCeP)W+FXcHGjnF5=OW*D<AlG2C9
z#k^7@9!hvVnU4y#iL&2`wMmn%wHB>0WrcI2<_?uJGP_>LEiu>rDlRzP<!R|9Ii=l@
zjN_~yX4^o|Jg{1>zl6WPdcOY*|C+Lr6^n>-b~A=={lD{GJOzAw@(wO4pi(b8!4vxx
zXq0o%@(hv7{{Fb*(@mQHJlw*0P{^R2>U&edz;M*K-g-x*P*^Ky-p^1b8X(yg<%R>z
zjt1Xv7-iABp_vOt(JP5Ypk$wn#!}h?sk&RH`aqL$tSNs8igCXz4_7|=;ZwM)1fs^=
zOGc3%+b#=076Kotz1zD63k;JI*?s+PsSkqoZ`p9n@Lg#$yDWj`tG&Xv#@Yu#27IAz
zVi&d?LeaX781OiBcQ-&+bNzXP>IeU{P=-l-XF9UhPDmW%NfFZ)a_2;de#7$zdxQP}
zD|5F61$rZIDWMAeSopuR?PE_rts@8*EX~32Iu;KLL((;0yL5WSG;JGmTA-W~9srBv
z>t@Y%ONG0k;oSMOTmEAPa$NCBTd7yfe$^LP7_VTM7y0NI2pNhz@Et76{JW8bLFu>{
zLEsuOA3<&qpss)BdBinsDThuub+P%`C0Cq#@6oyUi1n+^H65}l$^AA<<LeQw?fjqt
z9`mRuXl`qQ^B(TB@~Jp#<Dt=xiC<*wp!s{Fi)_3yL!txSL*;q^<oS>ufs>!B{m5L=
zep3-#lgF}nYqSuBf^IluRB>XmXH4=doE>OSK(4{{lGeiy^+`4d%D(;41{7Q4^7z%p
zIkpXHzE7Z+mW8tTY_)yxQ%wzkw9!RpTlS!n!|Ff48@|B90-pX>{Nb6EZD8x5sH@{e
zvQ`JzSC4s%{~ET~<GQ+FK5U6bIP81Lc14E92)2T-wxoWukn90rZ$#@uPez6&<}JgG
z=#3%+e^f72X_fM-F9)@_sK0;tW#gj^f7OY;?ocQOD9T3KUw@nC)t%J1MJ&m{6B&)7
z6uP828*`0c4?1xrAk%}{jSmxb6S8@E__1^gyYImb;uO&Lr5oM)1xC&@yX{8xY<)au
z5kvPT1I;bAP!^TziZ4cq(KWhl#diUIdmn04F9shn&IymsGxzPc@l1aA4uL7NM$qp8
z`^_bLg%{^EZNf;7sn1HYTB}nu;k`2lP`d)R^blThf*d_9G3jg+`YnysR0J9ig+Vxe
z0qcoEG|8LNjn<SZTy@RNjv1-W)23GmXM2)89o?zjQT@v{gD`C&mtp5KKVTJ2!)yO7
z0scif|NlpR3~Y=H|I^5CNMkJex1ZKEqk3vKfX*6WWCsxWOL{1Vc||p~Zk83llstyl
zjh~q*5VA_tuz)+x)gixM61-QgBUb;UHo7~z-g~>Yoh!EOXW~%i{viivBjy#)pR?Z!
z4O~*<?m^b&lPIIE2GbIesIueue11JX=pv25b?g@&@h~NST7g8{d*S%5`A8~gYp;+Y
z@rT7jsvku0;Ed-q$j8efdcOO3IN?fLoN;QJ1~*$gg(Rju;`n6hrJr2{ZWJVakQatA
z%;;R~dPUa!hXd2dh7i<FvlVxwtr|mtAowrpetPLf;kgL2FAA*nQ}xN<btE&LYg2%i
z%!@)_8&jKo>+Y<&;`X%eJ|3KvlhWI3{F{eo=ZncPM_*tV)u(fF2umbkRtLpMc((cV
zN%ijFx&@nMDiC=&FUnCsaC~3&a@OK49yNoEn$?`f!KN;55Ry5_7*V9m1n?=7Ya}HA
zqSnxtw9b;D)-lb_{GImbQ1zM-ecLmU^iiYRk=T`HG)N0DEt|Ij9;i9(1xN^`W~x*I
zQ$H27wB#zZaDVc|*R~+OAR!MKssuElFBST$as5sLE3dIV?;z>9-<6Hub2Hj3@y{C!
zyxS(<uc^$bJkiQWuzD>s)UgRl`4iwuBufQ+)-O*}lg<fmFjinRnv2J5uRoSx+Hi0!
z@+&Ov>V-V0{x|CS3JO%J1f>gF1TIVIhCjUW^p78cyiy8!f6+$p**!3EUhGErOjOfF
zp{svtf8uhugQEMA)B>CZQ|PJdwS2_tX@V~Ua0WfWL$M@iO!^Y)0Z+@55kW%?Vb03H
zyEdy2w;EwVjsMB|9*G^lcC|5ve&#IHX@;CohV;8Alh8d!IC8<Mu*77GgFUs@xKI3k
z5&U5Yw<+*JR3FGH?<`kPkr1=2CWSNe+R|*CqPBK?c_g;)9Si`q;f{(s!0xS*Tb1uJ
zLp<OQ;V`Y-uHK2<%C6mgqWVQcle5uhsHS$QcVRSSiDEN(tQxTlG6F``(%zy1;xIlv
zCAtgI$D_1RWa-3GHVEjzWOC}Z5=(G8k&8v;AdCySYJLmD(il0Msedm$v$BKC@~gT<
z$VC*wltATls1np8SwS_e>i283dzBughH!ORIQ+uDi_-*12aR;dRxxWEQ?jVLrtjB@
zMzOW6R&1uNDb&_@nCgE;RtO_&-7%k?PRDnN^8vdMbum=@trA?5ex2Pk9rvk}i}>R*
zmL1g72(36|Bq+#Rwc4m0RdR|Owo$US{O9RM>i*gP6=6cw%kAbVAOHt<^<Mc_hpLlf
z8uM+b88m~>Zz?O%kzGDJW_${QZe7df;G+N#cAnb(1&~IlAd(O4_yeFJw>Am8pDwx1
zwHD{zH!h><NVi04W`i#bI(TtPK@B*Pddixs3>@%w>Km&NvBx@XjbAc$j3E6L$q%|p
z%#%?vTcF?`k2Dur${%OAQ12<%H|c*-YD9)G9Qb2rDonM+>99!EN3q1<kdnR9L8eUX
zP&i`OR(78Hw5_qzBqBC1X(iJ%R{4^=ZeA{a%*q!v5mjP4(^AiYLR7M1xUrG4*|B3n
zQ>BNy+S#B}8E0oE3}q@u)TTe8s(OW9fJ2*F>dQ)<Osh;m7c}5kdoLICB!dI{$ZLQ@
zs}dCj7`HcB@=@?D`8SiU7VXdY7lDcyyG+n2zcx%+QRGFjI8fHR0<RPm=!zasm4Xz3
z4|JMAd2)Y&0y9x?YImYdsQ<#TUOV&)nEch*QlIqNEAv#3zM~>*FK)M;<AfZV2-~$t
zOI4zv4EerRfCY1Uzv{9FR1_sO>ZFnN+rssTLmP=|_C=wnjrvWM&SB~!oX~?2cBdbb
zBJ||ZwP$n+p;ol}_T6ArxIf4KbS2qP3$~MIg?T1_Ber$Z*(-xgtMq5D^B)>o0<ELI
z0I@k_|7^BKV#Gl+Svw2oB2QPJnnd>!(*R{0q~wa#!NBTpC__Qj@%_|U9AwxK_t{5+
zNPi@erjFr7l<Z^t<a55`x*zoE%$OIY9;b!Q07YJ@iV7iTSKb{5WFq3M$|o$53N@9N
zMGx`gau>{*NH9w6?D7xa2Rq8{Y-O3bzjP{VhmLE!2B}pM+L~ukEqy=B)EZ}j*by9h
zp-6)ftVFF6<OLrJ#S4t{=!7d2!C!}ei&Hh4>2h-mn%VE<RD~|sW}_3x$CLY*{>GK0
z$tM}FmCDEy(}ss!d)N;g&2OId$iBp)%A7XYH`$I;t`+s~v~vML)aX<M<NQ@Li}7gp
zD%V|ct-j7G<B9bdsQm$8)KQoCuVL`NFo0Ou>HnuN*rXwu@E-=yhw3R@APK*JOd8J2
z>h$&W;j*($TIM`y%5`4b;|-c;sJoiQT%+Txgk&82D*`aDZ}QfRjB`!M!^U0<S`il&
zw%^e8l_=P!;Jt3_rdesjpMLVw?d^^lS%v()i^-H^1&VHJKCbSCy9B@yp4~w-Y(90E
zU*n{1-yFKVpKe`j=`)(_eBh&;dZRJbQ*(iG{Qoe3f*<aFPej1vIna*2qmMV&r$?%+
zZzbAc-N&~#I9(%P9`?Nj1#|iJ%Fb5Iz>TmFodzUY_Up}szm#E=ty#ci&5qFvTF+;H
zR+7t>Pu+#bTlD4k+4mt_C4{q}NMHEea|`eUY>*3Y1ddSbz2HBeo=}M?XvhO=Ne~L|
zsUZ<8hhA3FD1!7=^bOQ_gEeAqs2NsN72O7u*|*oRQ9a>7)RJtsuQuh0+3-!uoR5<G
z?EMcjqJlN4%lzF%&ODS%@Ic{J2y#M#B0B!NW5R7uPN@$j-93)RH9%||*?tNBbQWC!
zAxV80>nL!D_~al9f-*2-DO^Vzl-o9LkRQLd`6!hnjVv0Tf!4T1&e)7TH2{n*TZJp&
zdIdrE)Lf|*a@_+6B<{XR)jY_f3A4m1p*D;tl*B%Wk|!B*B1pN_g3i$5hFUYP5(%yJ
zC*3}hL19s@z5Z#q%J%*s*1;n6pN-a7w*H-C2)NyTXi$Ng)d;Oezhc-F=Py-TJBr;K
ze<+TbzIc_(BZmLnKM)j_<`w$AMUx+W2U@q>1i#JEP(8{a=Amy~XX(k|?<MgVtN{Ef
z2Po0EHlpRvT@`6CV{sBdUQEf#3xYZX*?GGK_S*p?omY|}X{4b^SwtKur67~!0VG>E
zsY?v=O!3fC(2N}-U`;DxvkQBD=KQeXc#6XY6`xYXJ8a2*pEUjLr{>hf#a@RViK(i2
zD5*pxro3f`QP}WO%#`uSESuQ6&fg=Fxjuo=V<Ut-wMNAcZ;l+rJemXZ%^H8-5pipK
z`Cx>XQ!4^&L^G(W7w=)HZ6q6?*Petn&oDZPg++M{PHocUD=-}d#&H#;tqU@LZc`xd
zmsdZ#)TU?m!UD6Q@(K18#0NLen1GPf_nY;kZRXz$+edV-gCDqPC^@fhx5J?>nV{)4
zXqHCaOhMM{=8CB`xmop&L@I!7>6gODH6L_~+%i@}om$=@AdA^BCybddYo#qT%UXcY
zF7|6|LeN6vQJAfHufzbNK@T$xrYUf%vDQ=+vZL~S2%eFkv<QYJh)d(NPT^PjLP@a8
zarr^1_y&p#)U*gE&@Yq>`quNjM|$<P$~`bBj?+YByAQ@L&st_bL8c=T8Lojzj}Z%|
z^!rW<KTy>dF{R~6P^UX}M+MzeOj_e)H79F6I&q7ndG#mQYv7+>+y<It@)RmB68SJ*
zMs0MqB+|^7lX+e~6p0V{2|BA@7T~0ybV`S>c3h$uC-Ez?>)duCB#q&!KkSY7E`}Az
zmT28sjvppaL&|VBU2KmTI^V-%dy`B~*;e*Sg~9z&d<61>Z^JM;IbEB8n4;ohb0g@F
zfd>9m*%!28z#&i(SY^LM7INALd%=@TA2!kiM@WuvuEl;0^1Svvjj-&pS(aE!9Br14
z9#P3#W?eysQrZPf4GjF}bEM|q@a_S0&Hf{B<1qfgmD+tXT;?^XN-C57a9q|?02Jd<
zP;Au{(#8-bh-sU0V6D{7f9QvKD{ZHJ_7dBYXtUE79Vo<48`lt(VyT>W$NUHpf*lqf
zyU&q)T9m_x*`YWZTa}PBpHUl8c)bL&$R>JX?kQm^?s6P^3`KL#%SNRbD=3m=PFC-i
zQ||=jrl@jLdwaMrWwUrYaGKT3`@`%o_NM`_^05m>7VNl-D1^&KJ{tIVTw{a7Gd)mX
zM<q&Vs4T%@LzPHg{7#PMUIrjZBdE2~<sApCzQj;Ly^`3d)}CC{f^A)j{`huW_KY8o
z=#)d%lXtVV4DOf)OAACNnA||YT`o(srJ%u|bEj(Ktc4Jq+(=(7!C_xQVzGMr$~@qN
zCz$KY9f$;+wKJ@4%WZqDdS|_}$yS1|{_O@;{^ZJEt-uUw-j?c_#xu2WmFf}z<CxUh
z)4+vmWHjWT>Ym*hF&$2+m3=B`eWt$bG)+>i<%skglq<dXCa-%Ck|{_`+HDshB};-1
zdA=7RdMzEx-{bjs-R6rHRvS{)VeUr4m+(^Nez{Q=XQUFwjvqKr9a`T17Ts9>Z;{8u
zz`*)HMYk$V$r$VwgkR7^>H#hET&2OCaSiCIwmHnJi7_02WlsDkxB*C*KK^u%?A;o1
zN5&S_5&ij^IckMkOiehogpPOH=!6Sf3>}*;A62{BuRCC8r)TG%p4d-1vG2Ll2oQLY
zzN4dsqY)2^x!aSt|JNR{?sMx-QFn*jmKzZF6=;Pm89zpK6mR!NcJGFfR^j3i6>lGU
zeCH&M#HU{jN+F2D4Vyb?=IJEm#>NO$7t*GG<re+N{m1^{CgcAz2<-SgT99@XfDE*E
z1t98VOxgK%esg^oV7I8C1goSpk6g|w=db?`HGL5bQWgIlc8PbQyC2_F67I(sEyamL
z#k(<NT>9qBa<`-aX~Z@k*vmmQYpBbl>C2}_Tw?p|q$-&275IC9s_T4@-*lhUh5NkA
zV!xHy$EEyQB0`>XrmocW_ji(4j;syd`%kW)n5R%Q-V4V2d?AyO{Dqe(duz1RuNiSG
zp#=bJ+8=BBQ#l882@5|N^LEexMGfd2TrK%yi2=6^E5n{oS3ho1!hROQT}AN45zWr&
z$(-PfMh7mRP7`}@i;P{kB!$r5|L`7+$C#@CZqST=M`b`fZz*iG-~uTo46x8BqmL_+
zbYoXIro<f(Z1zC3xwH;Zt0V-;E*TT#(l?7Q;G6t9IQdZ4`zszjP#W?$A&91maEj08
zgsb21*=%nc^{D^&_gT$5PpdYRb^3O}H7sSvUx01L^R+r1JuC}jvb8^VZ*<6d0vTHH
zX1{$MTh6rUGUn1~pz@C{Z4J>Lfa*PJ7N5}+GF)ChUj_~G)p5Cqba8}5r@?6aZ;uhd
z&!iT`*e6V^;U4YCbxWseWy{=RW<%_sKF#A|E-ph=i1GS7iysQ=R&=#t?G$%Lp&s7q
zlO^ZTnz_=&+F6S8<==xA4g^suo}@V?OMo=RyhW%NOt>(?mdDc?kuA3F@dxhLORz2X
z#P<BL=BVXjE#y@Z8?aDc@<hiDcL#4(8)(CqsOx#LgwqPR5Dw4Y_M?@1$@Hr4q$X6S
zvxXB=3Ce~A1fAlv@E_I5PQvY*eUmsXn#8ye+U;?l3C{DpzJ!r}gdLKvK}bbyfs4b|
zH1M1n3QN=ck~2QZ$A6W=+TsWpu?Z2&FE<SqDyd=7_4d%<yG8%DC>oHR8H{c91St{u
z->9N#Xx~-?BPtFW3<?I$D{;(_0A>3S8xYnV5y(BA?@q?A@_hBl2L4Xe1Uo{Hy<0f<
zQYVbr%hy=kt9O^sD;(Kc5%6E!C#OSp0i`UbNs%cZNJS`tY&IChMBp@)AJC7h49k6~
zm^@Ksr5X%KM<lq^AycBxJW!!^rg<7Na1p0l=Iq{)0>;;?($D4n_3W@#M^TqTZ9Z-}
zv2+ZJ0s#6*<jQZi%Dc4JLuTM?)aWm7-*Y%piB-2opNq~b;vz{6LrjLRa!hOSwlsp6
zanW9C=tyoYvY3?#_9~y0LC!q?Qa<}2?aaXp9Lg^#M=_gatqN?`?!^#j1n$zVmrH7o
z%SO~l2f^IO<&hXxpWOe-&nbKnOxQ%JFODB95;R<CqqC-ZxzuoM)8~+Qb~WqAkYGvV
ziM*toz4gADZ3oWU(rP(ZWiyHFoG(GVXI}=Tm>{ur(_(ILb^j)@KS3{bav~yOHaGf{
zZWICUQNA)6eMJLdUk&m(N6We)yyW<d*}>cqh|_V1APZ5L(UaW1^wgVdp_%YI@n=R%
z`1fmIm!d(oBokwSvwXe)aaAR+sf--AH)R%u_3RZdy@l>{mhJ@iU8wS%deJ}dmB-h^
zN8MC8&RE6sEY&48`p>p`If6rCzmoD{`~x;#n=rY(kWquOlbJMMz;mat^j=cjA$}TP
zVc!6YM^SK^P#YsFVRvN))D7u|!$!VC=W$OfwEb!O`-?VCtwXpUIxrb%Y=bMB(+A3h
z$>g#`@WG_E;IldQR2mVOA^ZiY=J6T^U9o5F(P~^1V-Gv3^MLuMGouaIqe_n@iWUKY
zRRuuXSS6qqb1UUfSx;-Tb!K`<Z|(46vzxPXQ7Q1Q#O4s6`2wHQO->OY{grj8yXn)C
z1IdD87|XtEnTWo|KHgCBd4H%O$IuH{poG~~nh=y+^~!iN9)PQUs5`CNXC|xoEEH@>
z_ThjQy)z^6NPEg1hWIxN3tB8<qzz6A0Z(#3R66#g?d~JJ<(9NZ=uvv3w!SFaQ^yv&
z&ISNfK)>69#bJ+PRBl~=`-&Q5mBt6qc9(#K4WE7L1O<B@o8WlgG}vgb>+x3a?FJC|
zqCfwY?HXu(?aX6<UeBaoNK$bnhBi<$TYYB{!tlougeVV}MZ;x<#pR*DY=y&YNu7-O
zv|Dj?p&yPvZRZoLX^1#Z&$M)fIC&V)m~7_k&2eNuLk3P6{7ENtK&eKZ|6#Dr3GGc|
z?C3F*274N`f?}jy#wGRv$Vwq4u-bDJA09mbfjU2+J~6cUhLxhN7Q?LLuV>DJ|8@jv
zk!MLrF`+7n+m$c`P-zfOz>>Viy~L^y3}2VEVVT7VIX(dLaD&F!WIEO|xP|AH)gbP0
zbf~ccttHDczjr`ri>|bC0kW;~F4uS`NF9K(d3Czf(w&u;TSf`&1v61nrl>C@0)Kl+
zdO<T}l@4?SR5f=Z4wn{m2V|8|g`LSHDG?N*YaQAd{^v8NsR7Oe6Sc7p6*(GaJD21d
zLv+I!5!eC$O^X&sz#<m9orzB#`a;$A%rjP1%$H$~xtnoYugzMM$XZ<795|qYBjd6a
zouy^K)~cGOQB2qYhmD45MklxZnifkLX-;r*l%sBgEvJXeuN2>3kyANB53DheVgE>h
ze8Dc&{@NIO@xFxc*{$=WRl(LK%Dzh8koQPMeVt<JqA!3`Fuj8-wtuuHPaKOay!l?D
z_Mz{iV$f7hS;AiX*u|XZ;8rJkC>UED>7d&s`biD>PZrw_Y9y?cI^v$-71(N*usBNq
zmH)PXICV1j-u_JYQKFUDlWZVV2G9T1ES~T5dOXQRQE%4g0Gk~*o5XHTX#lIjZyZVT
z%LSOmdCNTcoaZ{aVk@adO)|l*C{dAMh5xmfmSFMFP-5p-^WN#V<{cSWZ7w6DDPD^|
z;(casSwH{7dctWAjG|+eQ8}6IQQKi(uK#j{RXRzZWv=U1-Ind+)tvk!*v(<Lnp}~I
z8K@X!Ota_6bm3c%Yw3o^+Ur(b=e3&6aehYG>@0(R?a2a@cSo#-ONM@QK*GKoiZ5Lw
zO|NCDR|q;xuqsK9y^x%M{^|?<tmF@|+s6Q)HTwz(1V#~@kOk8IC&j^-%`MS@o$g<e
zdwc=1VcXz7F$3kLR=B`0V1<+DS2*$jm)!(}%dQve5I&ieS_3cp(Bs{$T`uQ1{rG)%
zkLtX9FoB4jt&~B3(kK%|@P$E%N;|g%J?Clc8~d^8Wf?G1JRWq#+kq{X|Gagz$p|ND
z;w6>v?@Qwt(MgBM66jF~$=*nzB8|?>KF4zq*@h(ET=^g+itjwTHOZLchPr1>VXJKC
zg40RAU!Vc*=seic;~J=w@ok_LUs1%#Y_%si?tLc4E8Ydt3KG|2VFWxm7^*ow2Tv`v
zRT7n=?ItdF?}|wtHj`@4e&1Su#An)!EY4o=Q`9?k23E5a?@q%@A!InWW?EbRLuq1g
zB?Ke~_o*165zOXlf`^)369@WKT9-H(Vn4FtKf3fQ;7qa8jlX5u_O$9q7Zl!Nl~=FU
z(iV&g1+Tk~`&6rVGJ)y5bv#+IPRA~2N`PI_1R~m^D|6LzyuT&_MIU2{8Fn;o=#f0D
z^LNSmfKs=}+=hYX>g1Tv*x_<RJ&^OO**@)Dw25pPoA~d>jG5&4{33{(FxRK9uTdP0
zW{$N%Kw#>f#P-TmR7HV^^NPVWova8<6|D%YYNoD7((cdPxU<CUz9k8Lh9=n4w~52Z
z04qV8eY^cn{pNdFaNdv=5}e0%{wgxkK$NND6{EFJ4qBnsJp6tC-wU-IbAY>GVfLM}
z^ISTd41~x<njzDtqJd+Kw$XF-5p{1qc4!H!Gv}lS5oa?g{F?=j5H-_=ELsf6`YsIb
zjFy_!A6rMhx6g5;>46B=5HyVz4b?sEHla^u{7B7fS=Jx<KOaBbNjW#V*56<z?x@cH
zt@2=H_&?(%CKks3t@6my`u{4A3)M+Gv|$CKJDe&oHLFg5c+K3uzU(8b;F0Kl1OWBk
zUfl)_9%9aHV|RI1YiZTA_1U^86(n?gx`#(}H@x2NFW=G`5ompsvMI8^s_NdcuQl<n
z^J@{R66{s9Io}>{Ck=^*wqAqAg^QReBPYH;8y5*G04}9yTIcVFQR6Qb{YAHLNk2L_
zl8er7NjmdLj-J!9$*r$4HHF8f`xFunPG2h}J>3azZ#O%MT|x*r9E#Oh*bpcKLl|VY
z8|}^kPFS`~J~FayKgW+lVW&>DG)~~4iwflPVefknAv?HDp0BqB#O_AtSJ%6Be|+p%
zAu*COe7q)S;6f^(G~ffTE0Gv!<Go%-3|++W!?v|ESqW>iQxEyl215)H$vhib#$Nsz
z{|%vvkcY=r9KYIga52`XuLUPoA>XQ=#Iu`Y1TS!$I_kLY*JBWT_?AR0?f*_5i?%{9
zCLS$~Afh0&9&^^ymqul^pxywa{hr)$YnGmI=`E+x?x=^9N3UOM0ISb04GI+C08vnk
z7++%w>P4_1?Yxn4vi#N%5fg9OIJVMg&HHw}Di0Aev(U^|DIXytTxWLb-!RJyuyJch
z@L+Fd@!DnuFL|!)5S+hNMRJBpuD4OZZO*eG;Q8Ecae~8A-Ba@+lhQkPZ)wC9<UfLa
z15PG*jYXM%4LV;pMMkJlYRIVh+7hAy+*t0jmT4SV7BeZq9a=d^Y!IU|&;s|tQsYgI
ztdR&edX+u=8lVd>0|?M#DmO2)5gsR^T$21O!qQ%bHt#}H7@k^08G+RrHNTG{3Lirt
zIm+qXIb|8P1$NiYk9pJo3hJq=ZjWj1xlB@WHbG({^U%&up0vAyNg3lt;^Wj32zP4)
z4njs+K}qWbP3m#@5<`cvoKdw$7ZIfEWW=xQFE5iTEfV2{Vo9+hR}wiLES;ZoB!3(O
zz{-{+S2~ab|G2d?#Y@PWH4pS&f506SOn$Z=8c;I}ELo2asiU_q<W<>aNZ==&w>BS8
z11Au?5Fji{Ym+kr&TKtifD4)p<}c^jo~m<)o}8m-QIy4iiWvaSjR7B>j#9U66_ZKX
zY;^n{Hs>=r2}#iG+8U{4buJZzMZiltKYp$tA`V}<e79La+_l!9CcskfYKx`HnTLb-
z%|-051x^d1uIPCJ3KRebb8i|H<Fz>4(fwIjf&kzl*D)VE`S4D+`-3g8M9lZp=DHcK
ztV&#Dh&%CTy?K2_NtchCm*QmipX@qMA>Zdf3xI2|o^$6CzGFT!gOXyE^~n^*E^Bk#
z#7+lD2&e`CFBrt7@UaB{3tK^(tC`SSNw*6YiISP)=LbG4n374;=}p9d##Wit6}QUT
z>6wKtDsH{65mo$4K<!kv8+AAmDFIp1$0gY5R7TcN!dW#fh%%N%xsA36=&EE(|99Jr
zuy|=mv~+3AhLn+I?HZ8rprI#ht^V62I&4t?g96?93RTJX2}vYr;!(1Nu+(wTjU0%5
z0nWEZ_g?O0alm~$qDI2dbLJqA(oSXPq{5PkH(O_`cS26)-|w=#I3z^3y2RB|6<-TQ
z$@?$=?WNuU46O3>`@cK<H1&beED&-r8vPgVV{|zp(NqX=n8#-fJ<e6(S#v^58cS2|
zu#bE}Jss%`nC2i*Wy>2Q+rdPoU?BT2(gHR>QL!vlu3)npYNvB8_o+hLw%qzf*6M(6
zFAOaJT8F{n15S*~n#8JJy^T~jR9X0K3DbY<HFJB-uynjet>O{_V9C<Cj{9%V_#M0^
zr!cg{O%83Gyzv}0VdOT#tKr(Kq5(6bTO9K7a{Kk1H`4xWX-<kF_(yiDAE=mzEU`*g
z=e?z<360W_;yr`Uyw7qf8;%`mwJu;nkf16z8X6vo)IByAVWJZWr%A+z|5NJ2(hmiK
zKj-Hw>QZ`a_+sTszo6>Z<-Y&GLn<Q0*IPm$%^FZt{nf)u8R7(=nBYd75Wi;>xlCa#
zU!{>K=YKJiBEmhG=0YycLS}de3VPCn-z_kdY|0yAdW~40Vh!dxTksSVHVtx=4TtoL
z-e@>N0)<oocMOk0Z&f^_V4&m`1@x?LeVkA*Mwu)vMqdxCm+zgmDEF3-pQa|iGGimt
z!bYnv*WSP{;&<u?HZcN8D@)90{~R;n`bzYVmlP6#5#8Wm$cG1(xA^PnZ4P9}d}Tv}
z7X4FJfF&glV^*U=eD$EL1emMqFeG_>VEy8dt$&$A|Gfcov2i|Tfx7NKM4F>jrCzp^
zwS<aTGW^~RLns`WyBQXRhxQW<uWTuu59+H9Y3_2J(VRYB``s)kL+WA)#ScW-WuZ9-
ziDChH?(*`7ph~qPw-J}4eRFGLGM{{*ogZ&@WT@kLqE`E&;~A<$5Se#*U))MBVAdY3
zT%^hGAi7`XC1Gu|HlqV)Mu3&=_Hj0N&%4$%(|jt7C{r0oA=Si`X~|cp>hNFrOu)ZC
z=&KnXhzOP_tc^&TJJK`&zj}ZfE(<K84|99LobM(TuaC4Mv40R=*$tZ^)j@JQNzd?J
zD3nA{ljNIf^S11tQ@RP$*))w~|K(e&!@_s%ZwEinECHj2jxQ<YfT{7948$0=;oXdK
zD0%&BDFCqkyZbWVte`b5;Wf84uDm8W?cn`Y%)8xMV-Gcxa8g`jVRO8mb=0SmUz9lM
zIVyc42M1o&!cGD>S?<f*p$RVqbOiKsa-jK#^e^*PiCN;+{BqUo10oZ?dkF$$k?9O0
z;EP6C@jD~B9w6(GYFVKfSEwn{9}|W>cdLSavGhfeno<!WDFWfvgQhGgbn?BmBuQGl
zcaUiNUj1@X3}zy6006}a5*8vbIZ;Ug89bXa;B#9uNeLRU#oUnyh2HY4r&7DFjqzyE
zC0gie6rrC#aW+X+z`<al@=$=pN2tvB<CZfd?nb-GX3H{?<>Fi{sC?@kmyL&v<|QO!
zg;~p5*L4d8G%%Nq#CaTCR*Rg!%S(k>$;1iH8rCN!C8ooML2<M!rPOO>`(X6)gl!fi
zM%l+W5SZ-Lt8xj@fd1jV6G}A}^lRey;!qho;lT!tJr;DBx*DnpvzhM`K~WoO%YQZo
z)BjaR6%`GvYk6U-vt}S2x^?ju75i>@+x=oPeIqpV768j+@D(|F;zVYfJq*3`X<%8+
zwWUjorDd|el5f;C?W)-)JMSn&*L0G5%t0KkLtAD>s;P|k<(mbKNYYX?7Y5&})D<rK
zg7dF=&G$TV4SH<^aqZ-HxFoLiZH#Dv8{{s|9#7zQaC0Kv*)zyj6p9vmE0gRLiH<V;
zZK?k98-Fc|xgGXV+6pKmI;Gwdj`eP>5C_EBUvE^XP1q!!Aj?)a%o?Qc=bR}$Ofqg(
z?cZ<Jw#nwHuMUc}DMO6LH}?nMf{BjLn0OTW-j9pQVk`U+LpS_aPB);H8>xF-zrRTU
zLTNd|s$OFSY(Fn1C(S&|GY@cPTi>WCtf6UP6llke1FW4;y0<BB`N9}tM?Lb#qq7J}
z6et>8yaZ$n|H&Rt<}B+ja{8Z~VN-p<US+8@2tzNspR6kuL}cyjIW+JK^Fj?^<+Pk)
z)TENU#1qOAxrq8;DHOnMsW@?W=a~E*x*7$+lbL2hWUMPZDvQH_F=W^Sei+~8Tv!|?
zg`*?N(4P<>1%`QZC(c@VUMLF^w%R$RM-~J0y<2j}h93rKT`6R<O<4fdO)*4gP%4tM
zV+CveTRE07@W6~k9jas&|6LJwml<C8!kY~c!A3y@ojB;-7En0+X9Ib|TPp+irK((c
zCP-jLq<>z**0G#!t0NY%uZeJ`wPE`&Wtg<T3OvC~1a=f+!}?MWk=#p40TkY_`*Ndu
z&eN@d%v1_?Z|gaGAv-{#+Q==c{uZ!4PAP#q28j+l9D@*yi;z}e)o4jX5ePwPY;mg~
zZkXkSY2GMdBMR!hFn7JL9FZGpic?*Yh0u7l9|4BxWRT^h%a+x_$u=vbZz?F`0RSua
zwQq;~`1fOVWh!-3zB$YZZ3x8|K{iX%;u)~G!qvs6A?$=CkxMLNxoBdz9EHaBfjwYP
z+<lG|3Kd~thM&r4S99y?4Ty4UryfRAx1pn(L>AZhoJc>u*sGyM+ZczRT&C*S^?U;b
zJpCcogSW<^qFE+D^ZgrbF3Er4voRgef$f%Ko>|7|L$St+#1CM<@TbWd`tWX!EzPze
zQJf{QCU4RA>lttiK}DH}xPv!o$)j|JB$l+ex2MKNgJ8`~I+_|+=EIv+Ry4EsW~!UN
zaG8Rpp3$09@|HbLN}5~*)Gv?`m*wEY)4rMo<Qh+!A4XY>20y!dFRE!;o3=0CpUe0-
zCjy_kXcwe|9T~P7t528dCqLT-mh4GN%eS`Pl}do>C{mJ%`6R5sN)lONcvY1G@UaWQ
z`7@&qPi9ir7_Fx!g`p<Q>(<D>C~r31rS`a~Z&a~aCCC(d$G?)BU#`}@H>KXSjSL4s
z_zB%{U3q;Wp%Vs<>5y%z6gFh)(#o2v)z{mNC0iDzwXdypJ(x2dr_3|l^Bg03N}Y@~
zH{Lmc5`A5rl$2Y%9EJjDKb{p_mY%$)!r_iu*vP687RPu#uHP*DdcvojpEPFuEa<GK
z3c}ux{-xA~T8YwfaUM-1VK+_u$VpoHZd^l|vh%Tw_Ar-+p1U!PKIK@c|NIP;xcq>%
zK$Rr^*P8JEVl9}M=>IpY#Sqq7G<MsSS9d<H6);gmI}|$fyaUE6o6TplI=&wgSMR#O
zvUqI|o<dQAlj9RZ27tSXhOm66xI86?C&uCJ_wnArR0o<BD4jn@O>j&OST7i&Pr#sn
zDHt4nX7JuY%U%f=Mbf|wkauLF#=*$J55Uj*hb9aF!LBb}YtIuXVyDV0z~@g~w`nW;
zhK%eIfx+y7c0cvq-&ov2(#%j0TlmSZ89blXxG+x&5z(J{yzi^GgQ*&}Obk<g-+;Y^
z7A$u?W^kOvT%dcK;2R%2-W|_fwFwQx5^%|$+nyU;P}X}x#1nG)%y|wAE<6873wAb0
z=@?Qw4dg80W9abV{Lslf!`m%MUt!C2g1z3lvJwbG<eY5!EBp`gDOT{Pez6K64LgvC
z2{oncA-d(8f(<cd(g}um4D<xzT_hDZp7$s;kq?Xc9m7RhjCM(8It+MBVm47N?LV%J
zBf4O~o_s=PLb)S)Av?kB)cop9gq$I@(X;<P3FJ3BKUPv^d_Esl;n}ZtfWuj75`!3^
zcZzEc$Rl55V024v@CAU4;~3$I5G`$mplU=O!?~NHJu$5#?*SWOu5MK^MF#uGm!@N`
z4~fililR{@z|AqHEWEA2n9QBxM%D8IyAhI^Q@Zy*xg)woR5ETbkWIe~Ti>`hJ;myq
zWZVED>;<_wKh@)Dz=3rBQtfafH~xW>Fl~#(wfbk=c+>(-x#!P>$pPJO7`yOSy-I>F
zirMpoU@(c{Fz+)Zre@UdE}Upcsj8L;nkc;x%BVE9kW0Mv`{ztE{RPCd%jR3g)mk30
zUyW^(^HCyTIjYs5vr(Vni})5gi%2Uvf#}WpvAJ_TK8Jz0^)Mulz<mFznej!xhkr2C
z#7zq~t-KAJ{u8`xgfJwDIkR^TSjbn75LQI|A=W@$sw}xaOcrBkWmE(nS#wI#Z=8ed
zu<g1U<Bn#S+|;9nJjq2~oj~9)dCsMANzo`f>RXLNRRbbo(oV)G{40!z63oEpS$X45
zUF<V5EJw@+t?5AmvmhpcNH;7TF_Oec82Jv5>4a2&CRLpG*jPik+lV|_v%#~U$2EQm
zXogISPF@qKi=t!F(F$3GY@1G;992BV{JFkJkC!m*VXaU#gE6}fDQF3mu&4S1RG8Kz
z{%<qB2}<0;q2HuR*T7>j6iCvC=F!y<Cz<sAPDxk{1kn^N{7iz({>$HNzxI^6OYTig
zpOi8Nm!<9km{!XZaRUNX5)jTg>tCA+nuWek#;9?^nns0Tk>`W(ug#GV^}{ymM~taP
zErIvcSP=r0)WKe(MPp_R&v%n~hn_J(HI-_nRMiMKW5rOV9^!CV9vz;tgCY#!aeNFi
zZW<J6tR_!VU<<uX^h6CEgV(s%Rei5Rot9)(8T5c`lSH|KZV_$sUpBT~B--$>CRyH{
z9oZ6hdL_6SPYSEK1w+blCwtMn&+v|BdsTSS0L?O=9?wqpmAp9#@zQdOv>lF98{u1{
zD7_)P#LY$=dVTyS6{alP{9$LOtl=sHCe2Y*XNTS_2h#}L5pkg;Z>CIyL+gt&lZNGb
zJ=VGj-2#f_THH*cncr=vzVL_!)UY|FY{Vo3V+b-%)HI8Dqm5!GM9wQ|ZjnvY$q+r#
z`>B<Bu4IFh4jE`2_L(+NO|dUmpRBTSnwMGq-{+poiagzc9hUAl&l;<dqPr=n1yutj
zAM}$O4GZddjzj(})8u3GqJbqN1d>CpYC)I@C*~8&;J)H;FjXZEi*9CYr8cKCqKi%c
z+omibcWUL6y!AJwAYc)331UNkCYUFfAEZQ_imgP<Ro3S{dttM6cg{nnm^C4P$|o1G
zb(tSyGab!W!5kY)(D-HTBDRsYw{(WBXqxQodx&-{b4u?6u60FXMDTp`p9&f`;^92`
za*AwfUh{oONR$hiYBjEnT%(pL<$fiMzz$OVcJ=<r2sLxy^2C~WptCYB{wf+&h6Y~z
zT}oJsjGY1mv5p_T;p%7&wid6cGI`%L!)<7#S9wvv*}v&nERMRAkhnHmV;JK46chpu
z{z*pC+eq~8n}!Boee}bdB3Fi8W?GbLet5hQ1~*Df#?L84Gi#p~lP}Jlq(A9W{MENT
zovO_`au?CkuC$dM=Oc3>UuTxltvGDg;g$+jaL&cGr3PxoOTLIi){$-h4`b)loLRU<
z>rN->*tTsu>Daby+qP|W@Wr<6FX-5|ZS={hI(2ULzSuutUA?uc#+-AE@njLK@<RxN
zEz$gPz|jqO7<_Hk507xv^<73(K(O{ig5JPc)?vb7xM1IM<uVw<&2IDK)JNniAxSa-
z;v)@bPM#vQqO0VTsN&C<#Tl-d#W7^Ny`h#%3-JT$O66LQHI~`2EOGk&NM;;kwj;Fd
zHnP~r)8IgHvoo=LA)G@))0ad5HxmpCIf?QaqJ#|Ca_1)*k0~D<_|4);b~umN%FXZ|
z{{Blp5v1GEf!55-9+;BI#wbSCI>Sl_s4Bjz(6QCK`Ao7+G>9}jdeE5g;dc*tdCAZ$
zu{6f`Em_fTGBg&R<2$FW4q$7#O^FfppNVBDWG_j+m9J+Qm^Uud6#@X(llZgfslihP
zH{fe$On6Wud=4gasx2jbEA1gxQI_sW%DFm3IIt-aI7NxN!IXd7xQN$J0?TEUyY7Pq
zAF9ZjY|F&ATQ>VBU{*Ty3+Cpn;u4|WZ>^SpG%5c4ET!yddE&TAK}C8V&5G`nVl&dj
z{<*D6O~?CtOXY4~s+f)*qeIVPwT$?Z6-)8!p%M2h48b3B?=fJv74of`R@i*Uzfkua
zX$XWYB*GD-@&Vn5%b+_3p1Gi!vI-yE>)$EmlVV97XWg_-k!XEP+Ww}^eUi0>2)~lR
z9k=(sO%du3bf#hDoVhT8IRz;umi*iyk8JUR5qHhcSRqyZ^HV1WG<)+>O~+&DmyDyP
zx-C5;Ws0JeW3do5Z}&JHNB>PnQ#nOJc(d}_Traoql9}Vqnxa)+qF<GJyD-}QN5Z{K
z;Cm^aHkZ;>bG$}KW@w`xUGJ-IL)mebXNa9XXXBlN&Cxhn13LvV9Z<$pPdeLF)c~nI
zXm@6kFssm`4F>=m<UGyZl>b%<mYe4<Xo!^*0cq2DT8%u>=arKzxMaN~9csg^ZjWBh
zb<nriBWo=vd;3%+`>(ogcB{BK-d6U701C_Osb)+6YQvn`PdDkBUxSn}F?;`dHgzCz
z)imu|5dR)u9)<qucsDT{I7Dh-8WuK>+nnX5S<ezK@Zu`rISRPS*r}p-ru|04$$l;j
zeDP&XPW&~j_6j9u8SKIfREnituY{VvFZp|~JIYqo0+uqMvU|?{e0pK~bCA^<oHJfI
zkbcG#&4onZJCcW6!0!gz?71;fhlf$%$L||lz^Kr%s@3SdRBF&PK*KWl;sPSpL4jq4
zd3i@Jfm84>DRVN}m8I^kOr~yKp_kBLWA6anY@9)zNiyOp^iRjA-Rv{w-=!VLFZY7Q
z$ua3zk03|$(O|+KHrAWAO}t4J$GSAZG(Sf>NEF(N2{QlWW7o+2O@<!;n^_J|2g3g;
zq2m;{VpmO_>2IT?S3ol(th-vPIIF*ck|?gY4_4~Mr)_nV@;E6_=I83(FAU*e(Oc%G
zFBcdjQ;vN+?W)V6Ga^8vhx~A`K&ge+-QZUQsg{``o!i3aRATvi1&j|f&e#;;At7ac
z?USM$vVuI)LwZ6FpI^tLgZ*|TYVd4|gShW9zR|o>4!&`Vg5ni(_bg<22Vp=5V+Zpu
z^CdUlH&2<-^@S<{Cs{U~qwaDdQiF!HTv!GRY7cYApShk;x>#O01|0(3&HRt=G@qt*
z7#;oFO5F9fcz<<}6zbP|!fL^o)~xZC7*k&B7kI*^)r9IU@bY5~%S|A5E7vaY{u0KJ
zFfd=Xb$8adw1OY>a$do;hQGFj&1dK~)^fZxx+i;bqK?X?53-2u(KGgFw^dV1mdY6S
z+xj4LpBng`PUK=hURw9#ffE#v+X7~<lGSiGf@}A(cb;Jyd*nf{M`!6>-~RrwumR1s
zThpDV8NVvs<%(4?xUK_AdCj~WH&<`x`A2DaQqM+u8d1M5>N<4h&2R1Qymo&0EL`hI
zLwDy6plSQmn;;q*C!YC4s`uAc-DM$?CmVma<$lIQflIe-A-7GJ+aQB;X6o@_Ndey!
zBaSM8mwc>r7`W+qRMnPaHe9)Qi$?L7v5TP}s%^>YEu~w7wty7Rv=y!RN<?XBc!6gI
z=#8PmlZL3-eA0VQTfXSLzM74<{r3L(hf(S;)jH<aVQIwj5JAS`vWs-R_ISenEIKri
zFC}c9c<1P~!H1`mNnW$$SGT1v>w|*hi|n<b&guZwV1d^g5uU`0Qj4-OnR1z>_PS#D
zZ2r1}_ijJXuUTk4EWNJTEiGciow5hf=q%nj(y>WjGhcMMA6feM!<wa-<iAZ!9RKOx
zV`gXk-$Gj7?sWSDcI2Mt|66(#KFk3!xMCNNN7yRq1LFiq&eIrvYn04)vAs>KDyq29
zxM&O7(6I_2KCh&RPTG8ya}&s=w=b2iU31X`gVA%f-{$TkApL@f44(dGF>=Cfu6sYP
z7o|+>y7mJl%Gvbr1t7ZCOCjJusCHs7c=<d)MQYLX0`l|c-AjP->tE`w)3`1^q|?Z?
zBBz39Ln!S(;DdbMJ}-}xiw~udbl)=UE?d)E=k-g~Fj0aXX5I9+<xe4s1W*2OLwcxh
zZvVUJ$?a*<0qP(VAhCZA%KQ6YXC>-GPE#VueFNc0I0cDEN}TDTbN$YCSrw(CZ!|+j
z|0JOAvO*@STH26alrH~h%#lSkD9N#%;c{fD^&p$B_?Od`%`zpjFR>M0dR<VOTxJoQ
z?&Is_z+67hf6wt4p*64wJnoukzz&rtBICjd5&L1uCfO4eeMN>sZaDgnxwr@^#D(}U
zqc({o2D<riK$A~1lq6H&nHl{R-GosO^Ab3Z5>k^sV5K8(!NnlC`&;Bx+p7ByRgw_&
zv7^y<>OvSVcH!JmvKeJ0ya7LFpwJGkqp;aTxTLk*M%vJ@TP`5mLE1XInZ5;rPC3D&
z9b1+x_-3y8KF~5~2n|<|tkxdOFAq&U5Chsq+7^j-=#1Qzl7@E3yWDuvRvC|rbmQA-
z?&Gc?KIkBfXMBdMY3*tgemjggeFj2Og%f&%`ulR)!B!-KE}n-rPFQ--BT4*;(!lZ7
zMyW{B?k}SUF32)-&N8D%a1AINQ{?g<O>-ZqYQ(^;2H(-b&{-v?6EA@?So>9vvDLKZ
z8(_JHHzAZb&Bc&-75Qv7%ECq3_8{;_pdy3ZLHhANNRY7Hpd*P;Oquf&)$DFmyDI8z
zc(?O8p-jeD-Wk;rDbX5cL#z~MNH4Xa(?OhX&h+Y0YLrB)K*xQGDp$Ja29wNM!sH}s
z@GkQ*$C+B%t5%jJYXf?m2YC#!Uh14WcN*m-88dVqU=$Bb+8QdC1#O~sA2vH$q){u0
zVEl4PY&W-4+0MSL+TveXBRk81E@tsNoMyNCVA?nHiD<HBiC5A3Cd>+$MHuwZEEZ$z
zn>^~7HRT{BYg{Q=vD<N(K2BjBxdX6i)Iqb0e+MwjiO`m2rB0<lN`I9gf3dgVHWjI=
zmyuyL6GuvHJ!men26E7+ezg?!i}+(-lTUgG?BhaI7YI1wsvMZ+S1vZzjz%8iHP9Z4
zK<;yYb!e`0m+!U>e%Fk9qn+ZJ!b<Xf)T3I|FQl0@RHS<{xQQ8jM?#uomS@gW>#TyV
zmA+Y3q-;$KjkEjknh13|bC;xC0SPb{m*)xyMTTt)A#H|-3t2+4Mk%3FK77k~pk2bQ
z`PV&~h(D6AvgZas9iC&kVDCeDb^B&mXnoJ*!iMfzw@sfZzFt_6OFKUOGatHTNID5N
zYs<S^`#A}P##Riw&6Zy#;5kYy<oK2Yznx14W9?ak)2;=o8Ib2TiUqSRKF-cg{R=-B
zC#-asmlXN8YD32qDTgy;@WRIQEAU2DaHbUEbRVlQ@Q3**|ItIFv!ok}4v#{3h^uKm
zxMA+nP3_j8(frevInzSt$%S0CDv#$>Jg??~M~x9Ob~c45o{?wgeTvi!_Sa!Vr5^$%
zAgJ>SPqcB9F)v$}4*TIdDH_@eoXtp;iIIHi<xbfzrH65zxYGtBgcBZ^ZqK=H28o~=
z$m9P!esA^n`2=m0e*Ww|g^IB}5f5nVo>!+MwqPwGc-SPHt^Xq#`y;1)PzX9l^rv4P
ztk0#~FJ)*8oQhcjOV&NPIfFMG{OPsZ$OMYAFb}xMeMYJK&oT&L#RRJ7tSQ#1_MXs8
zN$m@-Kx$7(<=1ra%R+C*eCyQ%U<Zrs>yQ|l<A%lICpxd5FdP+vA@OK69%dMc7i&`J
z?DwV(m7~N`*;eI@AV87~cS<efFj2|?Ur6uU#WpdQbqoFjThm*D@-n(uDK^_bvk?NZ
zimz$#xc2&CgY%yGdzsZ`RdtT<5_7UeemikxIauY5%%zNQRerf&Com;9-u#e>jlPem
zz@s^cntlVQOj`lIF+#qZa{SI{sM-Wo(1us*^Jv24`O%2js`0Dv$Y?9hD9tz`50w{w
z9bVcpp^P4P_7^qR2RCaHS7B<}5K0>|Rwh3{#A_}^DYmGHezYnM-VoSty;|ZpU&L#=
zS2ujAK9#ssF+Sgilji5^dJyIq%43cQ`r!PFiL#X=!vFZ6^=P-STJotQMb&Mxu^z2M
zF6{Yn*Ps@qAg>CRQFGW?rG*Rvv0OagV!yF`H5etwk(V^gnnB;9+av<*qwB+fq184d
z0iSd6<u^v;cEik5PfVpKN@g`37Ge%-{&+tl?v);IfZ53S?2$P1>>xFW^tvNk__pGB
zeqL`3ZU3)rhgrJJ`=u*a{6!~DTsOS=d(a?0Z&W$8@;-!Kjr~WRp#*JoY971aLnaQN
z710*b!SCK2@mQp-y`d~=U3=thHyjmP&6sRk7H5t=)obFEDL5QKxyA}RkW$&His-V8
zC(NU;H0Y%!A)ZZ#4ypwO;o|RqTq8;a#G1+S`t>W+Ed%P$Vb{sqD_99p&rCF%JR{14
z*tN)Hm-b+r!5tWjUnQ9e3Uvd?7C-P-n_;}ijTJt;WB06Yo?x8=XcHCm#QNuZDIKB_
zFscbcL_4rgy$ODGSo^zy5ymbn72E0Vc#HZj+Zy7Imj{?}{?ja{AN!1M_;|m^l`4J9
zxATsVC)7QIst@<-ZNuxcxoAi!K-47i{nX>Y$CK+vJ3743pe++R;{n`btW|028g`Ae
zRjE)_!O^}j|1WlkE!RK7mAU6Pf+8~0aF0orn>W;oop}J~PyQ{LNu!UwncW_?B}J{w
zNG8z2fF+MId$?~?&Fz=1bsr_Ry*0Dml+p2o&Ohw&aMqfBQN_&20%>?L3!z1<D|vga
z+2LEs9DuU)d(ahLWtUtk?+)e;U$;&Ful7Y{e)^AJWf?z=7@vQ!<u()rt`nxH&nN%0
z8a;usRF7@9qSc*o0o3<g^{?}<z(==&;GnxHts(?Xc@zlB8?cygmjsHc-_(18cj#kI
z<F~Y-v<&Xq_LaCIzXuyg1782?(bRDT?N9PSzxkL&6KBEnk1CANOkULK;CN6~VHNjs
z=$%e(k~l}_T-+cV24n5J6iCl+s<q`ptsg^o=9Zd_ny&Z3l5l6-2xbH#iFJxj=2do<
z+f<Ci6yX}D-8l0w@zAdSGCNhRMihW#iBwH}Zp=8H@PpkgsO&7*#$p&@xUuJP<e}Z>
zwj;uTBS4-iL`EO5VZ2wp;hxv92{_Q%r55MA20brRI9GU8f$kgC(2$t=;<Prv?UFeR
zX@S1W!S!)<+}gnUK#*>M&p3|`KwzS07=G}T(SFUTmWt4L%`f-<bUk%c^>ox|8b>d0
z|3~0xqhSAEW#@k`>X|thIsd2FNb`4WWWNpR|H@8BXH5F>JroO81MfA^3%cwcTXe_-
zljkHZ6NcA!swB}!1jUAft>E<*8MO1)#b7J`PyXA}b>QRv;V4hpICE)e>inpEavuTb
z7aA@DHlHR@1_+a{Prje$p$@i>?k#{5qAz`T27dc-9Rm?wJp!g@XdR;xtQsPcAu!*5
z*IN(cdg)Z%GZMA07uhtfwr&m@K9F1q32|%N@AKs}6y^Qrw&?BG*X`YQy4sH-+(yDu
ze?NaH0ZK&JoSYk3?u@UO=ZAHmzBzOzvpK!bK<_qnQrzwDIZU6{zypTfyHov_efc)b
zuIw+2bUy?j*ZrFOzszq?v1hphg~ILzW;IPH%uH3H#0Cqb09mMby`tyOR9af!-bAT^
z_f)-IchThle*JM=>hfyD3Z>LvGi`VBW6ebj)mVOC$qxWLCFb0ofP-)$ik82LrO!C0
zl={xIQ$g3zpMTcmqFYPZ$3r&AX)@h2&#=u$DKBnCkpSTL*(%Nnw_$cTW(Yg+NO4>M
zvkyTQ%jmHkP<l$%lCTG*xE_)ci)3{R*7lA$hL+B=Bbt&w$a;>|(7lBGR07u)b7-;j
zCB9mfkJ)v*tuVeAmYYJlo78z=vvun_Bb}VL01JEVI}CMjPkWZuToabjA+jFnS1FY%
zPx&=EDRyU7$sVCqboa{QK}rF?<L6cesWBz3ZlGgC%k<X<V>rCj(+qnQx0+<oZ;;c_
zPX8_vVvnO()`~lFSV(gkBY}@x>CbeR`9APevtQk9h9Jr7M`}ZDZkV@ePj3*mCYvTV
z)NQ{eNQE7;njU4^56%T3J48EGw-7&ztJpnddk8M0^2(&&ts1VE2fS6r1{D31e4fP|
ze{cPf(NmoT%s2wdKg^UfHR;5dV<Xc!!+_awK47&@g%9b{ZOD9!9Qczsr#7zsr=_P~
zd|3Mx5hR#6#NIt6ctKS@C3%)Zx_7Doc3mMa1C3*qhU>PFhcqaiidch<_BFC4Hdvi{
zy?)K&VLBr_?Z+54Qi+awwmEQ^5bPPq&n&Ax;vjJ?1s#@9$<qLI)aOau*+67=HO5XU
z&cZ++^9PyaD|#w16Zfu3VrP7<u295^Xo#uIM0#gz_5GTt)(mNNQqKzAJK=$TPw-ve
z2d`yd<6o#1{*BTk``Hy4x}9s&<Vv{QIo)g|aeOmlc?XJK=M>FtDlbrRjtO?;Hyh?B
z(Fb;0n#=1~Vs@c|iTMc@kedt+qw@k2>wRJpC06j*u|heCxAHsxf^-7_7FD!Yz?Bjj
zoa&qC#PqozWL##S-(k39hekkcG2v(m#$|3FOt!wVf4Z;mmQ{`M#VepAvzmR3NY9XK
z{T91ZR-ZVLyf)U)&~)T-XHzPTO>qQ-=ii+{jbg<V+v%4oSjO_TDRJR!p{iniwaMlr
zQ0}WxRrU*eJh2C9l+ig&QCL;eeurSSvDgs3GJ8P4XvSYU7JG6wp$Se!H+<7n(f0Z^
zY%pxA&8@Y8954V!Lyppk<MY1cqn%4A0o6v(gaUSYOMZ0^SwO_&u7rDS$o<hT@BhKi
zrtVilKX(5}j*-n+O&;5>dqc*}_u59;5!3+>goRx{{$X5ES~b3H&&7O0ChnK(cizRl
zJ-BUV-W1Jq`vZCzH79Hot@Q&g7nT!xG~0w~7TLl+G&Ds);ish__*fTp8T-FT?An@-
zk4u26kg)ol*`!Ei^YBj?Zh9GAotZm06Ni=L&h~=XAIg>|ZK&8r&zlPPl7WjU$1S>W
zKATFU;N*<aI4~JQ7o&@*N%^$p!wD((q|!f|Cq-w`0{wqgfb@nz%`3umowDUM3j9DQ
z8hsvIOTkTM=eYFq3}YXC1c_SNdd=HeR&^v(t5Z60bB*Eq%He_)4{6FbjmuiA;UcEy
zHIDP@m1h(eb{NHdJcH~4a6KvdLEf(NMu_(5g5Q=SdcsblIros<L0y<SSW>xX9wFht
zuFmMGUf+eJh7QvVxS%di(b7Imt2I?h4ycW5_yuW&(O>ZzEvH})@ry7OdC4l?ToTr(
zA5VfW=-r~UkHMJfP^?N^>aoOUHfdJBKxO>CxX=CgaKa~NzF|dmcPP}tbN(BH1o4!%
z(x~-(vehQlF88FMYAT&1rY$aZ0;0dfRNhw_2#zWYcF#F<j2E3nKldV2g&J0VsNjST
zxjwR%ArM(pl3PKe9gyM<EnEp$JRp}wjiV|;w@JhAr<SgikIk~$il^}Cd(wb;DF)kC
zSg-q9)v$Hbkh{v0xTJ#Co_IE7%OR<5FKXIBA;aJ_(5jpxUIamIwsYiYhFxmflu?3P
z{A`0|(v%|II@SZCT6pu*1DGd`WfvV<lt^RtWUXeZp6EhNH5@j8RVG?)?1)C22vvql
zYj*#_!%<Aa_xUIT41d%#p1q$^1_WGHBm<~Q(iZvQf2@7lhaO+=TO7X%zU$TQm^~@z
z!zOL!2;S{b6R4yP!fhKam1nT-2BIq$9MtQsBR3Yj{N}wAeIgb=Dgu{dy=Y$(v~G$#
zYxB~7_AbICwrj+06Ckjk>Iu#oF7Rax=BQM6e17!;8Thfa=9V7O%^K$CVk5@#CvDFn
z>K^Qx^$qm53F_BOmFt~c1|J^Ad9&QYd^0BOR+Qlmo%hqU0kgN)-_FVH2}54f-Yz*>
zNE{(DznRAcT-p0i7ZoKVCw?2qdbTschKJo0eY)VkaWvRg=)8xK`U`=@WqkNt&U&?6
zi26Kvv&TMeL``-Bop%#y0z^G9=j*zC%NgzlM+rQ|Y1hwd)1_uv$f(JL+H3MIV@i@{
zs<o!62MAweAKYq(YsEeHu<SS(2vjYAGpi&o*K7(W+1jbwHBc(+%}d6zL_(3M6Pu*2
zuaxg=63fGSk<^d^owMhB*)aX_mN_&=5XdR&z0xBgWV%k-*i*`;I+=`Pnp%}E7Ygai
zIv0wLCDo|>&ZIaI3}y3ifJiOqLEiBgL#+oM|1DcEQ!f9!S2wD6drVwEIeUSImciBu
zMO5OA2P&ooB_EqXky2p+CIN-6y$(+~ETLRBbqoZBRlXyoh^<fK`H4b=l0xH~vmEF=
zsJN2?T0SS-B=lkuQFU|CR@?fLFjY!3iCcB}G5NeSmHi3U^ZrgYiW134nIHhYQxh&8
zl1m|93%^oJIiKo64^_}OHT6yEj7F8+;HObzTKB#6UF*tIKIilllNJ(QAY6!fk!+*K
zjOCury%(w1@#y7Iann~$Uw=3<n3=$->bpV^gxSJ03sg43B1I$Q@qVv3t?>o$@iT-k
z0cRK6N*I<nxT!M#0g8LizAP>$2;MAf(K9A%bax_`{*K^akh@CArSmQ0`1|#9+UEj=
zfaiZwDR9qyOL&Pdd11*y0Mip_(bV~fo%0a~Qpg$-MY%5&QVP17f<Fyumq37+Qu7iF
z0+QcwF)biS3IY2c)xl-9x_3Cr%YjD&%;dG;wDb3~U{WERrfkm!L%>Q&;Pz)%iK<73
zyC?o{^GlKZDW6qWK7<FRQ5j+HKL|Ehs);i@T%L%XO%L6?6Bz%jm0%N`MDZWXT8lag
zeD&iIi85PsTidhvvfL*z|3Rc~3;j1FASdU40&iHD{@2=%{}+s&sXf_;?w|8Pmo#P@
zo0g#I<KW^3tApcIARdk|8Jg~&?wuJmPsVkd8B+qgoMz3gqDHTb(no9e@$9`hT-<#)
zx>-z~i#DBxq)F$vMi0cm?FAPn$sC66mxw%^8a?cUm?qmj4k4IoS9+;`xw^58dzRvS
zK#YGu8uhLey{aYk1MA5jbUwup&wfzj7lWP<Fqyy=C@4kK0nhqF{P$ZEhI4G=dZ6p?
zD6{4>Y~%WI#h?%h%Q$v-eR_NsfFR3owMz^$HXkng(_X{Z(u4|SvS9S=K38H%N>57S
zNccU2es@`>Z~tk$vwNj;`_x#yIUKyhU~CgxVZw)^p<KU4I*K;gdX;e_GfXkpd6%bf
zEHWX*YWJ-OVzC!%m>tRqh;$@2=0*v&lq~wa&WNzLIuulnnSp8(Z2d8}%QG46!AJ?}
zJj>5JenYAx{W9clr<d|o(ETu%;vI@_g@@$*3!2dgMdm_umsrk}R^affptjnFhbjdA
zNVEM_MDtvR3y9Jl2qQp2V)w9#Wv4qXP;o~?W64Xxzz-bZPaTT218-OZ3hb#B{S&KX
zv16!9T#$jO2&Ai;rmEIxhN(*9uYEG2tl4Nzkj%kMPp#wMl9udev>`W-W5(2EMV2xn
z1H{saBk@0jQ$0-XTV=gb{jGH=Y)>t<2x;gkaubTIX{pSvHhw5C5*qg`)Xe5>>4Ezw
zQsy<2q?(6ZD4(-1|H{Za96ZBbzmKU_ESb1fF_Gk!ZY&jhO4BaR(d9a&o=25L{tB60
zL$xa<_I%sws!q~9nKR>Hc$2i_#}2>c?l5eLYhag((UR{hF2J@C$<n?&&sN|#p^JB&
z%I2J5Ry(&DBgKecp^<D2qZM_~J2UBiztzMx+7z19K9ms{VlJQ!lQ6TvCd#|?j1Kt+
z)iz$1Fe|cyyxJ3Gw(p}?1=8X=Tc>3Cv}>z#h`F(x*@_MQB{saRa-_8MxnafA;<B`K
zqFPI`@0GW%SY6x9t|Uf*ubAabp<SScq`aa*+)OLHF|AXSEr_u)J@*rmMb~t&z@r=U
zG9wjjSM(hQU0p0Wi{2wZ)x)&OM5%b{{mPsEuyN@Oz&1%m%M3+cCD-jY?G`v?OP(gW
zJd8OS0xWr6Ht$R_2$<g~xJgp_-TQ92tyOI%Qb=&uwQ?TTflIp)0pogzSS_ciyCf&>
zJ(HunX;B7rxP`43+4{6h=U`;M_h7q$|GOF%y4=}?W=56Ld7!(%eOzvcFBDhgfk$>+
zMBBJveR|c~d0q<Z5GU5uuK;%A_sQOt5rD~H0vrifWHl8Lj5YuZlQT8uy<#DTSq(e=
z=n=sC;Q*NQ;l_ap@d=t7c8P%$==QaVIE7MN);k%zsHr<bHn6IT_P=DFTNzH-Tzbzz
zXh9)c%$eB`WNSegmhDyXwJ6tRC5m>iO~0fYHL(0JO8A(C{xmXi!(!_`aa`?SFizjf
zHi4uPJBxclGnG2egkrG>*I6ua>8l=@!LkkEfiS=mNtPJ{levD{)EM|fiKzlSFx-yd
zI!GF{$-B<3{yPbAoh55ZU~FxVS)EuH#J)@%p_@R0deFR}*fE0Zohvrq$Gf|U2RnNI
zm7{37+6q_8$F8*X!A_AayKOm<UD91RzCO3b^_DR?(3+>1IOeDElRY4h8;#=^|8m0}
zyHThh5Nfp=lb-nNvktUU-K(b$XP53Mq}z4Vx0_&twAj^Y+LWY?>V*irKOg{b`RQ1L
z5<96a;oo?8DVKzroKThQL9l6U6U^Pvt7|kQe8`~2v_@o#<ARktXrz3az;R^D%z6x$
zunFBAA}d6<76R#_ed5{|BlIln*Nf~<SJ<5Gyc*G2=YRQ!{lZ8OF3=r^LQCPgjdvR&
zJctGAV;w#cYrbY~XJL}?$t`B<GTb`E@`C6yfo4oB>=ahGXY05Sci+j{@+d(%+^FXK
z@nas%1%`JTqTj-wzL97KWTvXS^Cw`Ok=91w?$h=impI9}h4`z5c7J3j#ZE$IS&@UZ
zr1R@X<9>I*a*iD4Na{(r&L|J${+2D{+6|8joMmoUzYzlVkC_}8qlSYsA@*(ok{>^5
zK6F~(V56$EwFT&HbQ4<AjytsSoQ*vX(!2s_no3feB$TtaOq?Do93FfR$+9BGb(}(}
za-J;;a6o07OQrVbY~+hp(>StS#oVRA6>lsT>52*c_=eZY(#uozhIB4zJ>x6tjQJia
zQ@67$AFs;$<*mC%Q?wGT6{gmhaW?-9<TQ(v!*sC7qLrzAQ2&CfE-l(as3v_XaD7#6
z<ZER&NXfFRWn)^@YK}iZAy4sXO35+nm*WauD0vg%mXrx+0>N%l^=d{JfGO59q!XFH
zDxk7XXMQ%;xzgUNAqH6y!f<@v5IXYeFjjw5$0&DT<IcVM^R!QGn>)Nymrb_e;ueB2
z5juz@9Css$&of`UYh(I$w61O~DF<;Pybo*IC_iZe4TF@tU%CtFHf757XZklA!-x5J
zw9hh(Cz1W^clV48SHRzDLwk$`*=#DMQ^omc*J>$Sx;#TYavF*1cJ=i8tEOhCq;|ul
zD8D9!QYVSsk`DM(1e(pTn%Z1%@H&wM4`9QiY+9~Yb6T}@h1NU2T*8<#yu+F;UVe_V
zflIK!3BAb(g=P!nV#>K|mxw_YT?wx6sg$WKAq>EY*F;Qq3^_#i;6k+Wl!i&f$dyAl
z;^FO=sI<M7sH-{uqpi1<2N@%I4NhQvm#v&=VTFe?_ny}2;I4;*$VJC<vkDRi*m~WJ
zH37;C>agpYi9P<W8(v!NEYB5qVkH0Av39xX=XKSUm3t)x{hbqdVv^1dqIPcbbhEj4
zB9PuXh=PRQ3&}MWK<0Um*HuR)!o`jqB5j2G>hcteQZub}UW4pWznce>()&tr!-mPH
z4Z%N4DD%yP{!PA#rK>c{a~e?#vWpDw2W8H*u+D1Jcct=cgPYJIMTZwD@6fchw{q-#
zoe-v2M<budhxJ$dXaou^(DJuyVElS;^7AzEN})xI(KPZz{%93Cp4eqeIM_K5#xT7%
zbXt1LceXt`yRRQo2;#g80iaRiXydm18IvyeK?%f=^kP<bqJ^0Dk>yo|=SKRenF&9=
zmKT>^BC73^wFmi9v8bpgo_9}x<2E9^voAr+&!{xCs8ay)<d!XqomG{h#%$>Hp0N!=
zMU77e&eps_X5WnQ_5PtlW?#<r6RTbL?4?`k9R?Wg9xbD5;jR{wwC%K&$q?4H@IQ1R
zaQ*8vA23#2K9ASW*Iyq`_Cba#{zB?6G%>n;HvA>x@iF6CNx<+2TFxI`tO+uV2*z_3
zia$c~MDRLVx^&^|MXF=0vjsxT5Zxq3)^3?pR#uuFG}kLg11)Uncm<kbMVnG3b2L*)
zdBLeYZ9Ilvn5lV3%GqZCG-995TwU7mwT*Ugk~kd(=aC_NyW29Y8dQd<7LVFl2_Dci
z#nfK6=ugeFYtyalVU=FR!k=SIK=2+_yUHhLX_o6bFs)D_OX_^%_b~jh_Ci2bh2co_
z@F3~Dt59NdHfxuk|2tFsx&0zK_w!${IcKda(>gfApVi&bZtdQuy2MhE<UM)iajnl<
z0yq%?hvxt!s;NuA+>P^fw1F;EjuYb$o=W@>)X>W<*|QB@&LNP?7HQ}sa!w;|PNfkg
zP=jS>J{E{BssYIr69?c4&qNt2fI(|rMJE)fduK%66&O{DuQXB6fO@=-v%(_ph?bV!
zM!+@9Z7hRl0-nBbz!K_I4g{|`-AJH!6rP<);A?hm^3&~TZ1&{(4Rn|ppJqk9ZqI)P
zf=v9%jW04D-|NbQr*P0bT;8-=ez$%2OYj5s)5kr?+pzK=<EEh-pmE;1<nxr}4xS`)
zi;Wp)J!2eo<H7pcLOXI6J|4T~bhc^I%VaInEp7DD0CA`p$0RRb-E~%}LqE3g#uIXG
zI=qW*NURLM$};P$9!>2lHQe6<&9JKCXGPC(CDT8Jv5c5(#pl&VrC&eQg^Wp?slb=t
z%_mP4WMHk5WtBI8<j1Lpa)b2ge~&5|V0*&~z?6I?V>DB!C{Eln9~t>Lh8%XEC%}ZA
zH4zqE_qJ@bm3^LTU?{16I0js&244FH``Ev*5Z7iPby$Jp=Ui{6v36YP%I-E(t%}WD
zN$<NH9zEc0UT?LX!t{2xn)6@gIxYSy81-OQ+t_mw{emFj-BJ5*yLqnvw1lv*v;A+V
zPJoVdGARdA&rGd_9gm=mW_jz%D!kbmPuvY$258_VD9DN-gmeJH&GV}gSvjd{YQ-A=
z!;?$8Qs#tvNrSqMUt0uF9f83fLBEUI*D;yn;NbA`9cDMMub1=&z=k2tswtY@I+Jdn
z7m|Qpi_z7DD7AZd(>mJO-lKj{ACz}nc<x<FRjrQ9FtE+>m9s2r!6J}4;|PJ4H#hy0
zLGM(8atoysv4_<E^Xl>(B5@2ux$c9a_c5QpoA2YeKp>{s3VB-p>*+O+z(O?n?oP-J
zdP2|dYxM9vT;QQ|sp<|)4az#!QbY9T$NUPPOccRX)!v{b?#Rvwwhc398PYunIo#j7
znHd3{4*Rc{MWN0H!lbv&{=F<T(`IzU^iy>lMuqo>pL9Q6BGcb3>)nWH)SC=Am?dTG
zvgvEdoeIKHTSW%Zfl${f-S*Kb9|r~bxe|b^Kkb^up5da7J|f4r&m(D8B}m$S_qlB!
zcnG~CMra?0{*I@|L1H>zp06n9ll>$3T7x(&?-v%vj)}@QM7EtahQE{l?tan9DUnZ+
zh7#DrupyLeyIWweN*2|%=euy$!7QDu<_?)x)#mTO>Kz}iO>dtsAF!_E^6IY+Mx;x{
zrm?V3D78ID$9oR(b4fz?O4U8vIVz>sS10s}q<UvZ2C;wbQK00gYS#~VKs<F2Tc97o
z@ZfnkP4RDM90chlXXrObPtd%e_*<*IWfR#o1`W+1c4?}0K2+zbpPK6_)|F%NQzuAj
z_&N@VC(dYipchWf4?&j50;*#8M%Xdr9m(EZ?nqgNwj?dM##I}7>~qMk`#aNLxRurN
zLcw0SdkbgzW?qyiHFLKK@3M(y<7SbB#~$>zfrJ@_Q7os~Z!y5u7=aXdYDL3J<#gvm
z)!7?;<O@3Ar~C1_C>Q1%h;4<v9-iS(tVR#nw}3Hp#GvBEW`0WM<YzdZt<Pu2u;287
z8~s$<oczVC$tWO%#9=^4p?G1^8yQ>z99jbW?ZM^Vcf8*A*J__P#$Hn)g9PmmOs+&X
zI--AmxR$-lZcbuPa}e=;y7wxbSAidm+U~2i=@g12Nw_j;QwjgHW==H+vt|Z`E%xgW
z9>XorZur?iQV<R;OyfO5Z0`e&-G$6HxPo@;Se8PlIt46+3mF?m$<p4pARER-%sSED
zH~Ew2sdS8^kZ6RJkS}8@{|=URn|@XO$!!C!0$8~8m8;D;GM+keStQNL&s_)i3z2!)
zPai`RAQ1J977}H>1i%I9(ooo6JuQ;6>ayMGh>Wur=H$XmYL!IM%oQU^`{dOD$qE=m
zEzn6-^i+GQ6QvX3)6|CZ97L72O3OB%EtFsQ?@>e&99*`?!G75IDG8InzrTz306AB#
zPvp)j!cZ(US9Idi>C3<>AF{ADDYW_Ry%vJcgkM>N97Vkm@3sylkwCXeJbF>z;G^Hc
z<Ff~3zrmQrNF69xGjhNMbfXW|tr?r#g!q<`nU@SCu~+FHbxff$Qyxv80~UImn50<O
zG;lSO^fmKw9rgURb>v*?W~xzg$TY+ou;}L!ovC**aBz4?*fiswv<o+&8XUx?L(zED
zq)IrFYDCV&B4kPJ!&hL{G0(3bHiK0ZAp|s=<v!L!Sh2*4mlnq@-Au+6)Fa(^OQqzQ
zO+UU&;|NY0dGlHKY4yqLlU9){7o3!e6Qvb=W*l><_BWZE9<%?VCyA;8WPiJ|3HQFE
zYY>EU(lz|P^v<!X3YZI2l7&Ns^@@N5N%_-=M@$2c*uTS*hD&r8VnLxEF@TxK!DPQK
zLVC#_k>8{f37q>bWBv)t#CQ(-Cz<Ab`7ay>;^k_cdBQe-T#6u6RudzQDW`-$2*{IM
z6_+1(t8T^0_*$Gj!D!R?YNoXa24Db|+4|HfQjt>(`k<Ep13!yAV_a1G#dk3@0#67v
z4(10EpE20WJ!ZI&gD`t>B3h=%qWnHCB8@P_?d@;K9TUb}#F1UaT}X68#q$7<&AH}a
zFrB;p772K2o@f<?EbjzD8SwHep`YHR)0|sciw^v1(XmV)tZ6JZ+MeswxXqnMWX%J7
z$GVr~<#kfXoOmxkk6Z3q*P#PfK(Mt4u_>6*TLQ$jsEZGJBo%$5nDk3H&Ch?$WQq{u
z_%uH!))3V?X;c_zGeA9#!Wy-}>B;29{u#NePB6RBY?GEK2t|fylFVRklV&9HorWOH
z-vN@-?M!wH9D>)WaKYs|qDZBuf%4e>07Gi1`E@Xcu(N?4RwAT<=}U-1rs(){KvwK2
z<dShUxV=O4&QS*ALcffz!K*szNq!f<T1kSoiS^k+HCXg${_JuddI1b?On;;_6!<LG
zOjxEFG%H>eiV*x;Yb-v#EPPWDqi@&5Iz^mpp$&f`k|i9&_CYPRobK`=V^`zqf{7v`
z7ymwnsubu{I-2t|W=m^DsdKV;SexFm3()NrFQ1uGsy<ZlyAn}ReYBP*fwMg5Y4M{1
zpKoi8s6!GW2_M`l4B>I^{yyt0_Paew6MGk_Nz1!gH9xr2EVfY7%bB`B0;t=254(LW
z6(Y&6l5uc5Y^NDg*92vfZ4Pb@*t*I3x1j2CmNmBt9ErHT^<DVH!FJY9hfdh$RMFSi
zLTB7x7bY^muR+EiGlO|Z(|8_Q<{k6+zn^59&>Qq<T}n3i=%CSTYj$Tl%#RSfNjk}T
zl5|Ycq;)d6=+wzNVYPMQwzmv34-%mJwwF8*Xp`IYD{bYika}r#BM#DRuIfa7Zq|Iw
zz?bT%Of+`H-<<9cWs%^X@E^}hJ&Q+l)HbM3xw|Ju8z$;COJQ*FGA#dsxxLk>crIOr
zwAQ0;P0Xh0!Cj|+j(9CBwCrx0)9+z=q(7XVR>x~=nBJJ^KnARpAsXfZ+4`l_BWv5v
zwF#6*lF>oWY6{kEE5FC?hbe3!>GcMWF>T-KPFJ8~dShZ%XKM%lw4yn>X53$ArAhEZ
zDE+}$YJ5E_&!sb86q+k{JmxYex$d^O<rGAbM=TzYj6m@cy=#SK>UAALC=z6B1y(M)
z8IvtA3fsggj);*=4oi$|L%1;wUa(tEX?G+up&SkCIv*zT$ulTXXX)ke`s0xG*^#Ft
z>Z`Yw4H#-`>wGJ9H6vhKpr#$Fx}8xNu=^xem)(2gKDitJd8Ft0u?+}SSs&DNo`oY`
z4iLSuJGfcCUHo(8Yg0)e0QVSd-;vw;OiAy~Q-<;+NFw2UyMNiTcV!^>anw}W&%7jb
zX?oH8sI?+xSWV{Y<9BHXHgNUH>t1q8jnB};gO$t8?I+1<Y)~jRc!l?j_2#Wmg3}S_
z!-0@jSaumVFyE?mw^8NQgT35w!_91utPe5f$!?Aa3-LS#(}VIn_WtS}E6BP)J3Lzh
zf9H{B#pQ*`$67vq{@M^scOHWKBDIa(tyG@!FD<&Hn^jRpa6;ob<SWl-{R}ISY&*!u
zW{saYLlPOjw%yV6e9Mdldwis%E=C&eP%JnIvVzk4KEnVPVuMRzOX*LPqOQTl>H8@a
zV}zkVQA6>AVsK)qPao%1V5jn@XwnAV@sL#Mcu&z2aZ^7deBKQ7osPF7{)*ji`1NRs
z72M{Zd@8B{L?d43HA0_!C!kcIIc}<z2t?}-t&#R%@4A5gkh?>CS|7Upo*4VMizD6O
zz+6rZEV=5&qz}$Cj>00ep2f=T7zX63$iayu2EVkYmwwS8C&(C^K{nJ}D-bE7eZa!f
zpGY=&>B!OQ-jZxq{A`k(%rUli4r<euWN|JmQMwe|tuEyBcG~nO!XsCYdnpG)Xxzjg
z^K7|TK}X+fR}G>D-O1Ksy@PThT&m>S-e_?pA^|2Z@WT=Q?0do43Ga5nRjMC@CY+ZQ
zg8VVH{jqtKNpa#%K5%i=!S@|vnD)kNRp7+Ezz#feP7MrgPq@chq%1iI+m{Tlpf_2)
zi(+w**AxXlhOBx5>%25_O*^}xx9vl)(yS=lZPa_-{H<j`uD4DahBbIRWSjD)-~kL@
zVF{^riA%PVZKJ~s%^j(li7-2<Dx1%-s_S&Ka)O3H`E+%65d*vxc;DExDxS+OUxj_^
zU0%*x);Dr}LxLK1#h%o52?A9lQa5|b<XMGH&-Ln8HS)G$KIL<kPYr`FOV9Qcbg5N|
z-tJTKy*|&s;y;N7{6uNmNzc;)r=BXQF>%i&6vz2>S=#^bri)%FV5I2S6vT}sdDDs2
zBjS$o|Ew?6A2|G#8p7JcpY8uh-hchw8f!6nTeQXQMOO3u!AxsbLelRyBl%bXGRPjz
z`Q6Ef<PajPa`g&_gX<|$Az)zeE`^~{|589P&6a(^EreV|ZIPrarU1d#viEi*RGCFi
zN|3A81Sg11vw-#mVXCOu)aUQybt;JAxWr^5bZil;V>wlxF)YC^a4Goob9q&s5r^Bw
z-?(TzOd=0DJkTkOP)=1IWff!j6+#BcyrOf2++{+%2)-SibvIQ;*QoGyeZ*ZClQlGM
zbt0f0^^L(s5T|jFHH(#w^+oMhPk-?67i?t8uViFzuer;?xihkU^5h1Nk^LUpR8}6P
zOF5vexQw6oAnT*(40D1YNkfEunf)DEZ!NdbFA(`QDzSX4gW?i`NiEVXkeGD;kn;Of
zi-enLRrIfZianeP2Np|>fg#Hy1KGu!Sp(}eU<_RQF?g$kp^a}Ym%$;sJt+rv-1Ms#
zJ1JfJp7_xo9_UIR+Ox@E>00W$eGLtx>fL*mvL_Nmjvi;LJ0xL6-zwVv3@EYu%L;op
z|NiT&b<UXLEb^bTfC62t!S==&duHXiA2#;5gZ8WatDrowpV^bn`)Jpfnyy~$l^s^Y
zy8!CldueO<KTyw)NzeZ}lKvN#B0J0fmQv`}mbNGTZW%sTe^BI(02<w5S%XV7^8E6N
zt``0=T54hoHvKCSgu-8dKz~#A$yhH4p*r)FJo@pYLc?Z!H(d~Zd#k%=)0gwxLw^&N
zgT3X44~anfl{6%4{Y#vbcK+tc_2pJl%G9xYZ%&DM-|6M*)-sPElUM9748511OV)&k
zS{vjx!mT@Ax>@#zN;fGqZXUF0Bu$(PDc2kr+JG$d%=fC_odV`v2@I!qB>rppxV7`&
z#^V5{J-9Z{$K|=M&!0>%Z#FP=wea0mhQS$k|Euz8)kCO?q$TY6vMADrQV8qdofpc=
ziId5yznJV_ub1fso%f%uDEVN}Y2H#&ZbtgQ6$PC%T8boUippS1urvRWMor~%iH5-!
zM8p+_v{$LTl<Xp7mxS2@!mKefwpN_vhqoO~T;g7&!yWTx_9uOM=zD$fQpFtn3T?fZ
z(@hLf_Q3}e7Ri0OWwoTP4U^M|ySMC;)Ame~O+fWstfFeB;_&pM)8^aIygQ5n_uGWV
z<X!6x8k!=X)>`A;gc@-!LD#gVZ2oz*cURDBO|-5^Tki_XCtN%pF~hRF=Iu@D@K^dX
z3$m%Q>M7YWg}CkKFf}D1#ft_m?u;9}mjvuJ-6vDkAf%?vd53%>4+ya`eJv$mSj2vY
zpNCpPx;D_h*!(ITn_X0RRUR1|;Pb>SVle9zCSHdz3=%TilVWHeFdXQjL0whU$r$-1
zlgi9vU%0XXGK?@vsNQH0*=Yx!SMU-+N;hYFq$*7;vHX4AjDFy_v8pEQTo4-zOunhR
zkugl9<9LM=15-FJ`-zymxL7YC4^HK}6fsApqd(!R%y(==LxU2+Qg9ql*4U7OmSIw`
zvXQCWr1j@P%i^pK``3h5U?l-vDebJ6-v@ZCLi35%?&Ge-$4)2|e1DtLAMB!#XRiZV
zp<r5vF`yN(8SPTf5|v+tY~dHoNiKC1oOQt`m_s;hH}%O-e`C~KqK_ZxbJGYkKCkS>
zfhFV_@0-aPo_Q5?0$P<&6<|u5KKJfL1i-yx`1G#KLJe-E2x>&yhkFy$j;&HBB)M6(
z`*lB<ccft@SmiY}r4c*OMUQ1k-ThX)Kr`QFS>j2}7D3GxnYX}cRIgiR8mQojQZ&wg
zC@ZJ6n{bx-jY0b;ZMl;y5*n1}nUOn_4~vVK-}QfF0fg*Ay_zLCxuu;+p~+kLD7mC?
zXkW_c(+sE-@x`Z1XKO+y6=V5S)W0y2fM*-8$55t|5IbTqJt{jgn5u7nxA3RpG*J*y
zoVJgUqc*Yw`v(Jj?=_!$`7hq;VS-$E_JM;}#Cf64IMILj>xns^N^<u?8#4^+)ac=^
z76&)hTth%7+FySpxSZdHi=b$MZ~FVJ%u`9zz&iLymo`z;CnL#a`K7@zRjOLQEM`(i
z;b?|L0VZg_(C~Y$5!J@@@zS65#gcEb{4)7c1gnF#@!c}#e_2|M<6c_jCR$u{Omj4u
zYa*`tURIy?6I~@fWeq=pv<QW<lj>lCq+#sh%qLfzjtk_DrI78-$jf4ZtSzW8o&owP
z@*E6B=8QAhj1u`5xXMjYv*qbq;>4y0(4){MHyh*U+MvwmDAxeks%Z;oT->or<PHMs
zWygH>s&%w99yFdFG_G46flB3XmgpV-e3!9MD}2L!A-z$RohArc4DN`nW7Qm;p*}h`
z%5Gr~(rACStNx}D=4@H_an$c*B2~)?joN~9K?`Rdkfbi4QTX!a_-ib8(+|6114<Hg
z_SMiE#WG+dKRr!WEU)}kAAiG`5Vw{m|0?k=23HpohYoEV1_4IZID-q#05OY&1?s()
z)(%IR_&00*Qt02NxXOantTnwB_>l?*olwjajELlvItN0?E20QznNj26v=OMc+j;xp
zzQ~&bHb}Xap+6s0Rq&vw55mEW7C_2+pV~@#jW;BTBX>{&&3$_8nqr6zQt~BN6p?DP
zx*?AaYuz&CLblHt=#4c8*0qkw4CB$LZNntowS$cs&9o<1Y)LD!G0m;$hM58{*s9#)
z=>aPyKfL*7^cw~O)85R_$3ta$^BVLA1aj1%k!-`$m;CMEVf^hf$Lqd6Ud$2c&|3eJ
z9~21=fce0+DWkboBlIWGYk4{e;rKwp4EbmecRU810ho9ndgQf|G$x49lDnvJy@x_R
z$Lz}AfkN)2L|A9fZxm$_cjii~>cTGoJX}Ysb|QmJJ1?_)<SeqVd7IVAqH8{g7S*wF
z2N+G2ICn+A2C)TeRV_5vkIb$lBhA^(ENf}<xbZlFbk4e|N)RBCnwu(T?N029T(;t_
zgC<hCS|>V!TTBbe^dZO6wH{4B1L57MX%<`p`p@BF%N@4rv^MtX%c}3V*^Vf;4`<TX
z;oHmt?>D`QYXv|lZ5Q+QN*~cly8;5hHcy&t6P>ZFi<Y}4*;%a~Lk{=YZ7hSI(7V2H
zMYvG??1<2AJK*P964BWO?7s7x*%z=W5edBv@2uR{ecsk|aUk~5q~{?mU){vz05{^K
z4_C#5{qY^6>g>X@AyaWjzNnE5Yw^*NEdor@*OP+QEjq>GgyTga2pmNf<C?~2I6n&o
z_;JSP!FpBMc_4!E{5xZds+z4yPTh)-yNdivQLY}Cgm~ft?r*(f+0fW1xX(A2jTpCK
zY_{4n^J13hj<3_A@+sl&3RBym2fc|ZVV5N=JuZsNh%Qj~Lysm3Uc4s#oxj-|>Je7x
z?ci(^+0c0X^6&7gu=KTpbf~cUjD|k7lNY!s`a|}H1%Mva*)-!FJD2|Rt49O?=3#Be
zVQi39Zd%gDoSC98&s#jFXxt~9<$Ef}guN)WopDG1BUxRDF$0=Cy?Ll2hQZ8g<G}Lq
zk666pBz?jV6`Y@G!jE=hV$(pf3a&DBdirL?DXc(i+fQo+o;rtJ9SW8HQuK|>=^5l~
zukmPC{iXAPh{LA8a}Pf!RaMc%FEs1w0?lHk>K~*<V?|w@?AxKyAfe7Za?g2@Oz5}K
ztQ7!xZ|Iomd@eK5U4`gV@8@xIVTU8!jd-?Z4bM`QSb|9o#TYqy3g&21sHDV8UFx~2
z4_k_Siss$&{FP|R3`m=55E!r*Wp<1GVSDE)-7m(t#Tf0t@tmx3u8vNhDf^c$V^0Wm
zuz+^*HZz=<Q`weeJLig3v6T^TyVls!?bj9HA>Obr#@^~VcVo4;e&4&;APklaY-ITY
z|17F>XXhpRQjnJmDH}ykZB9ca!ec-?uJ~KCXuf67s@8crGQZ_h`&xo%N3uk#t^j(*
zm4(fy+Rx<sm!_=I=Tvltdd4-8B!Mr|@ZSAhcTC*N_g>bA^R^h=FRl;7_ihSEl5z5Z
zKSmq)hFpJEgu=#6D3HU1C+0DdIpZ4<kt200G`KSIF6^YU=b#u180(Z9Q+=qyTwc@T
zUup?TbXs%@KKN6%dtzGWJF{3kv3tw1I<$&WvKso<9HOBZ01kQcBOFgB*y*}MV$ZMs
z@5jmg|3#E?G5+VBgqfX%?SIWM{C`pEjW|V(t{*`An?>upxuqT+1UDb~|Iqf9v2g`W
z7HHzwF*C(%$ILKhj4@_rW@cuLnVFfHnVFfH8DnO@alT!>AG<$RPkNGdHP*eitNK*c
z>FS!E?p25wFx_6^ch8a~eU{}*lXAH7jAMdO1+DhwxC_s7`-k=Q{qu`dCH{1!Zr{O+
z^F+8u@NqBR8-61W_^dRai*vDu-Q0zv$CfKWRjl5`!%3gZ`{V9kTZmU<pFf`U<=nv6
z>dLIVZVwZywfLaqCWRv5HXKQz#W?20${v>Qf;cbC-00|>P4+0yiH)>P^?Z4Cd{|wV
z|8W>L!)@>NVN1#r_0<5xaZ8@D;^pn|BC(Z6s(3*fOuMMgZh=|a$M7|%__x7cAua`0
zJqcQX<`aNlAlpb2Q3~t*_1%oc<vQ_Fr~{Vb7fb%w)dF2uE>7sT+3A#d0}M~fB256i
z*iVqn>=s%+V|p@VBM3gPNdz>g+#EhtnyW*wa=K0CsHxY#r0Z{fUauM8sXtmi^0T!9
z9PA%1FTkg#;kLU}k$M7(o2YwQ_)+=C^nW8Q2;`G*NbcYp1%-rdl8hRnp2uAPm)-N3
zB{eDF{x;Kjt9@R6cU6mP9p8-Acq{}j>{pQ(%T43l57n5xmVQZIXD}2S*eKa|+A`{L
z!mZdQR$Cx*)Z_V<RY<(a8|PoRQQt|Dqo{J8%u0T_=3Rt}w|BZSnt4IMQ|R9OI<|Mn
z?UgW!(|1xQ&Y^5lPSpqBBZ`Iks9BX0HgDHaf#GN7zvxhxuaj1s;qT|1o?wPWi<MJz
zp|@3osE{?qN~y0bj`k5lKyJ8CIh1j;e!JV}(Z5W9>vvWM+@|m;i?zTVLI11#mbV)I
zO}sy9X|}u5(^Gy;!A`kN-B79y2lXbL<pkyLN8Tov3}ZvyFribCN!^Z$r{qb!5u;Qd
ztqAa%8{P9RgLK<&cLa6jk|Q0oAye}ZmW2XEPfFekRW2S=&BFba?$GCvExoIshPc)N
ziRD-1kkt*%x^meej{W|b@H-Bw6*z6;nGZbQX?sJLnt4{B2vaL#s5IuGWf~|mGnf68
z(-TbKDWx}F9#B?UVp$ioAXHj4gv#&GXN5?L=L2ZSu-XA)umHi*r~pqk(gb<Y)DH6k
zGny_ryCDr45gu{w#JvhU-bvXE()5eFqdlBOgx!=y(}LcJ-X#9t<*=@fKa34Gyx!}Y
zf3SE$FlxZu&hnoevJeY&QV_t45h8^j&XFAc4eqJO14zfchyg^4G)Ishl2TLVS~z*1
zEB4V-UGkc^0X9m8%cOmSKC%T-s)X37Zcl=$g+%ZMni{>~xL7161u>wBKOM@RvN|wE
z+tl6i&<5%uWTYzb_4%WlgypMnF+29dgOe55N=D0or#&hNhK-MDl-9h`iKCBlU;`3@
zc1^I^)pjYqk!PJ^$R9&KiHhXE-1&{%6xAv1)DmdIOONB&>}<I(UgiyWMwGH&m|gUl
z#W&Yc-e~_?{NB=_Nze1eu6rck5CUfX-Ar>(e}YczAe7=fhC+nH-v_hLNnT`)2%KeJ
zhISV}`WOfL#9PJ@Lv)MGIbkciS8q9T>3}WDcS|kf<IZGMH+U7#+D2t8ULglYL&d)~
zY&U2656P}txTu?Ee@;&apd--gO3q@Z_s35`k{38j5K9c}5V<b7wxe(pBpx<-TK4#A
zr~u~;&}D`v7i6bpCn7PXY@8(@Z-4voFpD{#P_f;FMdj}}ROI|M0oxnBIA2k)mA-yK
z7KqWxpQ7n15HPGQ%<of@#>^ElJ9Nk+;F%(u23d8MmlaT{f2?YUJC(UF;y$+|9+vSk
zu*Z?X#|{qVx0SLn2_K1Wnu}Zu7dxHKoDC(|Oi%<bfoD9O4}3i#5uK0{4G<rLm9IIu
zNy1X#kwpW$-;AKB(&`N8Na>XQwui`HUZY;9Kl1XFK1Bu<j(g!nX=#>_7L{kb_xsOv
zMfN4^<z@pv!vpC8Z|lh9YI5KUIPl1dKHfThF!bNC6DinfW@?5~*oo&gQ<P_*<!u6u
z3E8r8w-jp=`q5x1Wz9G;<s|;!mF)Er3#}IoUau)AgTnxDEGL|TCCuh}6fBlvrCDup
zNkiA?h2XdExDB&dg~07#Ts5HKR&0N>oN*_q1hHe&h^D!U+NiFH3ZXsV3MpAhl?b;W
zXg@pil)KB~V{8n-sBEhq{88V`NmR=3#k4WEc-=KGz}o0_zkv!B9VD|tsRqzM?C?h*
z7yT(slrY+o4Wn*NcK;YGK$;dr{VBCwGq5y`u5QT%>Yw;rKI^R{cS-#noe+!M{6;*{
zPu%d3r~jZg=L^0E&m<j%cE;aDsnt+#5S6|xTlUmf9+vh#Z<*+AlLB|>)c4W#l%Kxe
z3tu`->Q=}8LR}9`BGe#@C#k6fnPH8hTGdW@H7XH&rQWE6!q{I1u~H>28ZsVk?+b1$
z5J=FXf|Sut&4I)VJ)kK-9G=j=TV1R5i0yS;!1FRq9<KO;RBTgGY2Zi{cg&i)&%#gk
z@N8b*ke+{oVy_l4A|W`%+Hl43-uDLjuoK|c32`r6!%S8jk>YRqV*zwW7gzOuCC&O~
z7;dYid-_#B@CS0T6W;kRTC4IwYm8?*N)ISDk1{<z)Vu3yI7c3%GyA&X))3-zVTS1Q
z&^ShQb{;!`R)z+G5oI|J3#8Cjr!u*Y1liLnvqjQrcNl|!n%IYRWzyhm-vNs1j)mjK
z#FlwshpB!Sv{(z+xQQ7SDbr(Ubj6LIk*&ps&AzzEk0A>`AEE%Jp=})*i}GSPV|(s%
zuqb{;)Dw^Xiwdp!kSGE?=`s<R$&)N%{@y#ua%e|&(nEjIbl_%3YksMh0Cxc{{w4c%
zyy{QyA?W!ntIx)X#>;fR%HnoIZ)1&W`LF$w+jwh!gYL$Y`?MPl@_v12DI4Ac&Paz@
zg73+Jy9x*mEcT)#a>?^d0LgGWc1&MP+TT$~ZR_o2s)@V__&^SE&S;8ffCsYZvw<LG
zU;MN^U56qEq>emO0vxNAVDO0Bl_kqP67)*E5(Bs?Iin%pyqZk0%Zwkp<qjC|Tc@PI
zw%p7)+Y~>?^7OKqDuj*mPtzrV>+jA%*p6GqvDG`QU0g0_Lg)PNz&~1bJlF^Yz$Fm$
zgfDG3t)69ZFu9P`crq7q?Y+JlT0)L)E00+C*)6uFl|1Ds%=#q6HTmY*<M|khipNWF
z*VywV53qsJMIzFPiLuD+yApxc?%MSl7h{i$*TmyPP21A4wPA`M4%`etC~os?N6YS}
z$x7lQ>>c%1tv!i@YEH2BC`||;mvA)C#anE|BX2Z<j&)olmvt%>@2waz<Ob7{N;}x1
zAf=9gCf)}z<k+;cSCimeCP}Xe&CT5@xQ1AaGwd$%F+87uJt^+?3hpVEGUQHJXkkm)
zToOi?_AQUT<tqRWIWl_%{EEskUa`wHf9q=Y0*mSuC~K?ysQl7J{M-JXt1#A*#3rnE
zXjr2)@=3~j=v=C-rB|Kt@q{s>RB|nTGMgRDH5ogD%NgaRPM$-X*Y)EHj(S6MJ84~r
zoT_64viA7x!?AK%;(tv{|F>!>8^ix8F&zh1OV1imH(r#AINWSpo;<%q_+YFe<F_(v
zq6pWDre^)3oidz$fAb`QCytpB24KQR`Il7-;9k2QO31oDT5R<`zz;?SR+8p}54Isn
z!r^tofzYzPLq!BGz(_(E#{b~%faT$`qoJ-Kqyhf!pb6X4ZEbuMotRTK=o`)7>wXj7
z^Y}jEs~4tst(E)@lFs-*0bYNjoH3jJ1W9XD6W~t9{5I}SPrHHOlCy_YOOJq!=Z_5@
z-1nO)fL^)o+S7KT75ra+MPG7Sa2*wC$HzCk-b#32#dJa1<!ez>{Q#phB-|@w5qQB)
zDNH<Mni*ZL)FW(A-DIp0=K26n=LfUB)cf@tqZM&-R6`+*v)PY=1+=rE4yM2+m@O8>
zd8L$Iq)-qinRRpm#u-5Sm+T>@o~o!MqWxKC+)ng9-(a;OJmdFU03Ps%4Ln)ZP@a!k
zrq34Gf$$%go?(0|_-6zK!pijLcbGcd!f5XJHOP>;AiM19Qae4#BEw;fr~27S-M98i
zs4{dWC_z{U{+RP%P+y+-zt%Dgv}S+A>NX|PCj2&2@KIfUL_s&7%_?>Y)c7uIkUY=c
zf$mIV{o9UaL%5!#0SgTC>-zZ9_Z;mF<(K84?+0a|HtFJR_@+6CJFXf{5r0N~JX=pR
zvfSJcX!^`pw6L6QCa9d8dO|4={F1j(>Hpq(^$EplE&d75XAXn{0G{3?${cU%BI5u0
zsGRd)QTD=+nE8rt7i#xIg7YSk^>oK>U>{i?&!9$RSl>0{@I{aj!6))IWzNGwq#L10
zUUMV3cKh1F{z<{N2Oaj2A6AG$jfY0<D{;M0h3NH1xG;41#&?<|_du__3;OK?;5Q$i
zjb?;)<UudHPHs$pA;x4$tA=HaF`9<#<U48`0BZF;F@Ljv5LNIR)t*V9-;*{B?8RlJ
zOZ4#1Hn;L|*T6GDaw0xxh-|weB0yf)AF5lg_{$Dg>YhKZ4v!sHR&g5g8aU&A@)WXp
zf9exd%SLGNh#w9`K8Gf{z{5?wf1>d6JM8e|JP}PP&rg-s4Bhlno@YMN;DyI-0C1&~
z7-~IS06EKmTkshN`a~L%WqhpJo2ZFv34sjE>5?di=$eIPHavZ>RLnZXem#FWN6Vvf
z^Qj(`wN}(6&tdlrsCflsUAUnk9Ai`}1;h>eJ-J-4RRK5iB0~PF#GNle!bZU}!5b~|
zHvw9%YJW-oUB-x*WdC?jl&YDN6qPvgu>fhn%j13B_sE@?>@Kru#I(E`62z3f9DymS
z28aearkbQLT(YZ6hfBZKg9snWglN7zS5JzE?YfVww?VAV&dMDZO5k%lNLm}Z>cQ^X
zsxB4@$q#Tz<@RkhiK$ov^q|G0_)1|6a3KSuHM?2kvji<7PIUt}&5fQPJ188;ab_8A
z$NJ+ba}6x;@}hAFaaskx#*_%_j?8n}adFMC<uc}a!&T?=d}9gs7~3^?q$DmrdCsqH
zL^)k+l5rN1Fy^I<Wa=1)V!UiGk|m+(3&#XQ5QNpW0aSnq1PFoh73G#?mE8&lT&Yt#
zrv*j)e6DG#Tbh~0OooeHko8L6JhaL3vCTVi94nw||6^kK%2b*r2@)UwGcG?d;bRJt
zWYRN*-)5uSfA-OUt43vwpp@@#owmY?F-Z;}z3+~!gS#9fr{Ty~$sy6IG@&S6q*Sq7
zGHOyCXT<D4JXah`s8YsMZ4(|9hsn3gZa+e(pG^AXK?o(3y&R06x8;Z=)>&@|p<fnG
zLq2aIAaNm=D>#`c!bfCq!|FUmBS(lLfpxST-eD6-Nl!ZG!8GJ;{zp4}+Dgpufl8jc
zyQ(U7;B<cKX+5JpOTx}fwV?}|nUZXP-K<F>n~Bc#DP)aG2)5*4pdVBShQ71l$IsR1
z6|6so=QsNECqFWY+_?77p|L_Tp7Rmo<i&-PKw}KWnC^Esn-2P;=Dg>6$v5ReL{&&&
z>=V=soh8xnKHlos5n|*gbSkh<%7=fxx6p@p!){+htB9RkOP%bz+M<t#nYp=>DyI9+
zRL_~<fB!O5CN+*v%U@lpUAOk)XlgXT>p?YDhjG+-n!D6=wOxS;Mgxbh&H^=qjNxK4
zBq@@FCsmIXECK3Tim)zxL{dX#1(d5^s_aFfakeAS<_j*+5SlYK<;yJ4{nRqVC4`=H
zPP}kKDLhGDrFY{g^EZUJZIW~G2-LQsCSx}=&j<BbC>`w_TN4D;A?eOm$_!Cj(3rRC
zSfMM%(ZcbI_bB-whKi~@0ueHd4!c=Z#ocZhTF+28V4Rf+>hLl8ilXKGZhaR07#s5T
zL_&d_>O!Z3&^W33;y^=zcsqOThoYWm{s3#W+wb~Zil^&{l7QRzV>L>OxYHO`!>Ey&
zfW~%D&52rkBf|96V}Dd=bi*6f&`f!vDYz?`eP{Fr=Gp3iIE7@}l2Pc9R1J>nra_qP
z-n}^lKiPv07g;;0T5qr|?v*h4QPf?O?MY$S4Ef0->WMq~2JTU~X4g(lM>^-wsp)W1
zdne^>%#Z-Eazm6&w5Ry#S^^mRaygSBGu+X83(4=0awADIc02{0i@NyRcsAoiLr!}5
z8tSJRHpFub--m>nkO0Y~8wM&xvKrm5Y;BazkjQi2FXhaPNfQAkkTrER_cm?gF>OYk
zsuL~qF{-9m3HG`0)^IJG;_kIGDjA$1ZB2hxk<>F=OQtjJ4}~Jk)m2cK#KchDVnQ=U
z#*o$x%?bPkn%893qk_{D%+V&lY3L*dv0-j^U}>dIn6?RY<Hw)5<uT3d8xuu?*|06>
z+71qrBPQ{c%*(MO9}>%A>{V)*p*rrNKfx&e4(?2^riE=yk+_yGXL=kqHc){ieMnbV
zl1(}FSLPGqxK6g`w;iubrq9>d7lspTb|gc&28Py~`rP|ec2t=g3p#fN06WFmw&XIZ
zT>m%K92xmMwz|5l%Ole1!R}?Xw7i)eA$zVb!(&~&BX9#&4-^#QS{wLuY$sM}3u(Nf
zNyjYe`S7kQNz#oCrzK1kDGtp<T*-1SyR`|`&c$xy1`pe!`$7HsK68N9-#?b>rVD%T
zKPQG%b~1nZ9Y7j>+@U<mswl>lrfb%#s{LSBS`0Uxkrn>!xhu1EdZ!UAYy!7jQ(XE3
zV^19Wnvr^J#1vdYF$68M81`}di(J<{#$UUya0D{eT-0!7=$7P>JX)sXMT20!Elf|;
zDEn1uaFjJMnW~*$*mQNa@8-?tzW|*0&NzgqKp1;BwtrM#MTj`f((jatsIvk<;fg*y
zA`qd`llDy=?FSWq+hC^ys_nxD6$hP0f;rT1TwUAGU`9MU8m{;x@XsY;+7kNwM^1N0
zd|tu;WM-7k(BBB(Y@fVea5{p6p6D8#9*J^@+x`j)tUlWD1O=sjwf&x%Abygb$qxYs
zYj*+XNoDw^1}?`^_o0fEI1VG?FaPjYmzO~f$NY`o1G+l#+<`9rcM;+Po5?vCV7^&N
z0jGv}zPeY>VQEF4lug3P{qNni+<a^E$|Fn<!Bo+U_rCNBl_!I2#OXsOumC$qk<x)w
zcHbj~e}4_Qky(HH7<rVn{NFNK`v3W3S_UST|J%Oqz+DRM)_}Vdo+@3D%_lS&GiF3E
zC;;?^71)SX)??JqG+LSKxmgp}GduBqd4JOuwiZf&!?qP>Y<c0&#rpz01OTBoFtzt)
zd^1xS*3mwt3_F7d{tr9|a7FseQ8!X}fQj=v`|CaaAI2uQrZ|MLCaMga3s<Mfy(!=O
zCSxZn)ckuY8^nyrZ7)mo;=5ag3qANVJlRPQC1Vp2>m%fL1kD+nhrt^umY=~=O>#C~
z?j28#vlQBYFw+}*3oBW+{d6&xhr3UrhhV+l%pTouM%VZ+RKjRDw;%ZD^RWbvW`3;c
zC3e0M1bFT2N*RxBz<z{FbOs#g-mbuZV7&Xh3qm##exV7QZUITVE%FyU^xV5L0ZFs>
z1{6g=Pelzn2-5{xH8?v7PcqkY6+nNWeh==W34?rs?B^SVVjK5F8Sx5C!MZZhrd`A8
zF2r$VFuHoWb*IW&E_4eR&vOdA%@8}P=sNFzw%96|6F&N|u$FU_H&+^;KwYx-Ga8s8
znq4itzvpYyEL;4pIY#K>O4zm(Hqs)<u}VrM?&D8eP#~6cd~1IF_Fs?4M#Su6Z4=v&
zdIc<Y%<*jR#Do{mNut{=pl;IkpZ2?2Z23=-QFkJ0<_US&>ZNTlgr_q2A0GC^ANP#i
zAFu!AYtk%R_^#e0R`~ojGTpj%?rXg0SWT=ViHM&)+inS529^@B+D8yk=IDnPAJ;==
zAdmS~vsli1(@3hd6<+2*N7M^NNK&pP@FVHV2_$(m)UX_c$O*YdHkfc-^2Vu&syVSE
zCfp{15k23}Lb&2aHj!E9QjHcunORp;dNF?Hx)muaNqf^w{cQWiVnL|FkqK<|7|zpH
zDa8=@vmP?zU6_jjXtr0>?>2u&EAb9=;zNXf7tBi=mYvmz>mba%6e*?a7idzOBLvLf
zhR8p(?mL*?5>4)nCtV^lNX>o(wWn84?j9HlNaTdFU+4=z#SK$~NB|Tmmc{IajqZEk
zT-~@ATlvh!SK_8}>*eK;@$_pn8iVS8a;P3FR&m$rAOGGhvKroE+;$Lt4mDXIB4=R0
z;Gmz(7r&G;mZUyF6uzOES?Ob`%W5WK&m*^-)fm@9wZ1ylE4DwOtaBT)GDtx3FyT2i
z%&}}Pl$ICF)P{yfy10Z+<%tE@a(Fg!Q8vkif0+pj9c#fBdkLSBK1N)+EGzDu;@_6E
z&*+%-7E-b!1D64wys{?;M;j3uP!KJu()1)Wr*AwR%a3jaOWPt%@Vmw?D#5d`uXk8R
z2zfN(8T(c&t<E;eN|J#;8Jpt@BSIIBWueGIPY;=c5HY1KT1epxhBL~RxOKJCN#kpI
zFq<*;mXLDG+>yrsU9v+_`-T&0)X>22FKmB#DT!U@mn6n2$7oQ6z$>Qy%Poh+{cnML
ze@k*!KCGRj2G%^ooZMcDCl^Pim+c1ZooLJ3-{@}^L;EzttgV^;9(D|AMh=0Q#e6sj
z>US4eU?OfhG*|tK!>=FfPwEcas3~R85DG98L>L=`#%*lnMi(KIZT9iv$4Cm`w=Xw-
z|6Y~f7Vlx1A{2_<mh-x_D1sFh*(OR#*xj$m#EmA&;vA@&PlFDQv~2w}SI^`K(bH2G
zK%t<;9EqaAY#mBN2xb?XgE3F5;lLL}u!w)vyDmr*|D(R|tj_EK83*a78w{%%#v_WO
zx@rB<#X(2Y1g7qlt0)}30fHPeqhJd=Z7FBl>0etm=*g$^?6IxXSL^JSsaj6JVH#Tt
zMa$lXJciD}l)0qJiSH^8ziRnlttGz7IOuYq!n)jdk1<&x%R!SqlHd~aNKp(~|JaOB
zR?M5fPk2UYl23WI5)RP-8mD#RG^+B@i(=iDB_Y{^O<4LXlo_`s1|bDqS}1zST&ha)
zXIydDJlGJ^7$>ak$w;JOMZxRCOe`6Z?gtigPoX%&#(mxV9}1y80igy_Nde1^9iawa
zj3yjHlDis8=ZCb@N{WjorC+M{<YV2JgQT`{&aS2JG^oXxuE{ylplgAr`LJ%~?+Bvi
zABZBES*Wv7HtSZk`UW5zHlfxPakdynuArTp25h##5R2@A)eTuYHhDZ4OZl&q$jPjO
zOmn8KfuPa|3oa`B$#tpLXG%La0~e00S7ZKxsJD%0zGgm3z!F}_O@&fO!z(!|>3wF5
z(3Xl@8@kt9aq2+F|Lug>A^E2AN}oxTS}j9|jXZ97;K79geI$%2fil5=Y*r6ML7wM<
zi^8x$;W-Ske<p_m5(Y`19+gCYo>P=Wp^}~Yc1#T0fWg~)N#W1!QFLLv2S3F`1gygy
z9~$+xJ3%RoiRH}R=Gzkfc+nmPR`E+zHp5kk;;Uondv6zhYq>Dr^5N2m%8-F0nfk9&
zvlTp2EvYwiCI5YWy9rdm!`9*hccnKme`A{M@Y4xmR_B@73GtGkiIl9guLE+(^S>eN
z#GPSqDG0{W2JF{&tN~>+CB{9H3ch7<tRYU}ImJ8I9IrzIb+R__CD|)ALg)Ay5pxM1
zdXb}H(zE{a(IL|vll{(f_U@?Y;N25Z5yRs`D?YxXRjo)|LfiVF8o{{g{YNqaTG>iP
z*S+86`;YW8#g;VWJ(EX8@BfJGt05!RKM{)dW(t`R9@08&Ak~v|E6AgVhYB#AN{m|0
z7V#1_=dUoDLWgPSo3yoz74;XjS73Oo(V6aTCL}bMu^J##yx5Huw_YG|<_0D8#zV}P
zwI0<k>-i0rWwk3d;m<fL$xYvz-?;Br4J;Ro$M=TZS_#;=k5>gO57bN_%^sKo&MGk!
z4FgZQwb<f+z?eg80OA)+3AyrJm6B!1B6Uj|^@Yry(j!MjVhoOjF5+#kLBRZBT3=EB
z8axwJ=R{q^$K`vv!b+i<cNDBJ2vQ@At5X`#<^k41;jePLxZsa1n<oTHAPh?F|4@2y
zvXIXrt}ib2z^SJQk%s3V=bm;652c9O*Bal$_}!q2tb~S;;*8g2r`~5P0=O`kxlc7|
zBq+sVPG?49u0!f-v|a;GmNFvP(kdkf`0k^W%4Nw%ZVohQ7~HET5~OgCW;-lYr%!2M
zoP(FoD1t?mT5^6X58cn5*laK0tFX|D^N53M<Z4o&Dh%cUPm{_XVd`(1c}U85P)a|^
zO5#6NVw>MiF#xG;H%L^%yJ_~nVH!~#TyFz+pRwCGLngq0!kTFEFn*`Tz3UHxG+T-7
zsIvW?CiTpmK5^*x2ZCdD8(hjJwbk{;jD2Od|2mu6hd}0Ac1-lm_07t3LZ96o=#LRu
z$2ohFO(~p#-08HdWi(egUHOTb2}kxzaO>)D3tD8gY*nN6d)CaB?6yR@I|_xG1jY*+
z5wXca8BIZ!K4Frll$Loe2J{8XL$+s%8fd1PgT=V^3;ORw$dSvmWA&WWDIA9DA78^-
zHmU!&ypn<Gf1Dx9z{>i6tCY5@19!g$ZbSH@bZhgy+gb!Iv#ZSK0@uxZt>(%iRt#;q
zANL@NYyYs240T>tM8A=28N7LwK|$)eRJYjJ<$0G#YBQ6@N{Xl5m(2yS?E3lk{zYUP
zxC6F4k`=q#wJ0R`$?g5_?Wq!ZeebDnVl>hh+Q`jso-2p6Ss)S8Ms0I9=vFA<;Ngiq
zmy5S<xGbC&gPkWgzrOA$<xu?<C2Wt0GX9C`|A5N*ae96!71Q@|JRcy3!0ZZcR>NTh
zD#|(>QG=PM8#w7R!*!ls-;<nr|99C8SQ!l#TrS5d9$w6gm+}&+^Yeu^HBflYxWj4E
z4B#L%DwL8DC`>13btCq51Ur=C;b!rN#m@^>U*Z);>CR!Y5C#$F+p#)`ub>RK-u2*M
z6V|Nh5?MAX@7O@aZKvi_PF#PPCv8JR5aaiHe!4+Fe<NL-Qw47Wu8g6sv~waF|DLG!
z9G4e4=^72z5nh35-?lsqb8l<{GXbry8T7#ugGHK;Jk4L#kBs{#<~v?6->QwRr|Zj_
z{!Z=}!v(uk5o=@m@W>z8F!WEg<3ys_OH?WHDo(h4_(bo;&U;ozoaH>nB!FD&5mUoP
zFBS!c5onyZl=TGus20dbV+%%zQ=QO-?{RcKl<WGsSCG9)uL*e7r;+#-CK}w=QlNtk
z#>QB~Mj_@jgu4iY^B|majff`Ezd=@S|E9m)U*F$ulENi5=}XUBKc5fEJ@eO~f_ls#
z{SoSZf1W&kDD<+fxiLmT5hce7k^7a#r(NA-j6aT*m_PQ&p=|@g#L|-%%YRnfvv%XY
zKo*{pE0eY`^oT|cqtdlibv&r$m9~_w)v-zDgGQm_BqHL`x*RJX52BQU!Z1DMn5Vk-
zL@d}yrextm;;7A*hyt*)Pt`u}*)iyEnb@AuQts-A0&^TXnHDnDLB#o8sq7J}vdy)R
z7lWROL{WYdQxbdX*$*Q$Oy@7Sdbt=g<Mvd)2NDh9m~rX~F_`+QD$6-jNxYv6{80YH
z6wYjOU3&UpCUV^Vs%q;prP$Xq6208j+yqz*q|)OT;MhRw7KY4f8PEBO6M*e7E3O;o
zKo}jDNCPuJ1EGg{XkR5&H$#3;N5X-R;)g{7mLbQfU2KvIc9D-(u-QCEcPaS<mK)N6
zRbxczfG2O77ijaG-kx|ZsV=hS@g@w(k7>Nw?Gd#=gnh<UrOE0r=(743!MgLz-r4B#
ztECd^qo0KCOo$@%@;;N>RIWCZq$%I@&IGn#cz5{z@Uu60X-3~DhX~6m<to8?TBs9r
z6t6(tw_FDN5JG!M!v4IvyqFQ9ivZL@8@p|^>ye}ig=NMh9iqjnJ4L`y4KW+ju<?Zb
z=E0ji3wvBxyec}t_i9>RBB|oEDL>Qd6ktt?qMUT+Nmc&)uZ`lCr`E0~{kK2q^b>%c
z?x+;BR3+#TeGj*c;E$9>#^L)LV$U#BHj64rZ-&@xyOAnFf^3UaCdFU6D|zl#H%5qg
z<K<~ig(3B97>p3uhTVdrcX#OsOCWf#w57!*g4PV&oGHxRqmIXRzKXV(AqB=J{x?Ek
z6ca)Rld$SJsMB4(p=Sk2a-Nci!WZ(7?NMu%I3#tbnyFy^=geu=e_79M6`NFACDgG&
z@}MPXfGeg}rPXBV1o@`vDGAEe^Gjn?X1F|76o$Vb#A;__iPozmfNm9Bgx1Mc=;xod
zP&FO*qZI4rk6-UG77<JmcRnk}4-Vm{sg((pcvtR3$@i2zBC^WcrdSWnK75V5-$mH9
z{Q^fH;&>2l*RjIdM|Regy<0cRTBp)VgRQIn(}r!bVrn2erpJcWK1q;wIj%_lusQ#9
zMyUS!n5)P!&5#^EMoi5>9b|;a95u1PVy^_faW@a$8>=OSGeW;v(P+MoNq?bYH!nD8
zq6OZ2IA)sBZ6YsIWh@IfPG7uco2IWz2`9NRA?z1;&N10fp`!5^b6wrsE9M4sOxub2
z3d<@GS|M$oM`7w;G;39Q{-UI1cT8rk&ljh`FvbLaI-o66;E)1E$(&ZK;!G}qb#U_P
zEKm2>3J3)9hpqXScL<iU=PI@)s;t|&T`qqSf63UU#I?lmc}DX(5rFO1@MM+@dt|Wn
zG2f6&X7gmSOLL4AHkd8OO4-od?C;;qj}`v#@4YTgdH?l1XevH|fsk88_^o6ZNf=HI
zoM2_KT&>tiODT3x7A9mqIrG6`jTsiZCwT|P9TYA;jimhN`12_}2J|Pw3VRbKf#sFi
zVf<k|aM<jUtzrDB^`|}A0he~Dqjfe#dDhr)IEK@0m8HigCU<-u53zPRF?{GX;T+Z9
z>TB`4XzBrgHD)%OQf=aW<<g9`G97v0e9QB;##VVHBhx22F;H_IgOWAy8$@LlTM9*k
zfswa*=ZXC<W#&WYsnPa5qCz_0inxUu10&-r4m!lUx<<;$l&wxxH9*0NxQ5e^l$2=c
zl^aU~BQZU=Wo1K17g?AF-lfc!6JIE6oK;^QC|}lSD2s$Hg~Dx`uFur3P3Ft*RyH^u
zwddTSVGfP3Iu2%G`1L%}+JiHh81~#s4{?ts^akMr7LP3#Vyra_kB+H=U2s+(0xR$`
z$;aWp<}L0Y?5nVuo=?KUwzp4U=1d_uSJu`blK!!LVCT6;^Ee8uXsaP@EZi|>ydRR}
zafO{AgJ*VcS6Mu2z>_=Ln;cvDv68;brNBdSDLRQ1XnZdKKy0cgK-x+V+EygBY%sOY
zLlhM_MpvhRg`A5X6~Z$y5rzK|sQLhRXQOD52cH1<l}_lP$X+p@hRKrZMYBgn+)J9(
z)M{_0`A#HgxnQpO8<@TnRvkZZt%j^aPI$Pq`IUW_=(@FnW<zL<k&Q@t`pYi0yt)1F
zA<5rIaazXpWE&R79(?xnKlGZ@kw`S13Xjxx6LC=du59~vZWcYp^)q>%Vi4&vCop!{
z?eSX$ds8@OZPjcuCfIo+D|jIxN0Sp$FA>Cy(UfDdn`&or;TPZnnE}Ks`S%e&3$Ov$
zAta^a$;k5aYK$g*OYmG)*md@N^8-1_4twA%E(LmdVEzOhS~IFxt#s2!X$>S8uj00V
zRKyA>@7Qnz)46ht7ls0y`<3Zc7F|VNt3d(T;s_*u5jt;^1}KJ$g4POYB;Q45W;DN6
ze3vtQPw~aD+%U7v>QZ?eDBwAZm%k@tw239gx|*c(59OiTMZAU_o-AD0oP{OXS9*^h
zg+;`N4OQhC96iDSO9^2L>sv_|&Xer`^V>C8_C15|^cAg8J^8A%(9%A7^kCYoW;W@C
zFOB7<PUk!H$whxwqmEZRuJn#eM+Al?*jlHFmAvuz(1nR7-&j+$Zg}c`*xg9$(YyYD
zkAdO&KGC1fb_n{-_3JMq+qGsK*ss;|8+!(=r#RG?VYbv1hgnJynrGMJ?du=ZQBJ4~
zq*Mm7sg4P%5f@8@dohfMYP<jnmfGn?h6}`mp?DY)nfz?uh*uo#;51ksLrBw7)#|<`
zr75;frrjP)mW|cyKNDB<+@Xw@f4S5zVs`+V<0smQIeg6<yLkb&+cOt#k|*Jn+ZV?z
z?c>+<nUYoQbz=xcyLFVTC)P;D;mxTV)>kEmSx98!-d`MX<wXbA=$w>)4_(NegqhZJ
z0y3D5awCJ#<RkUt%%**rDJ9gxqI1fAjJ@Mb8mfzad<ZUsQ^ym2!&;B-+xwHQMj|Vr
z)9xA%u$$=o7CBhwLFhIz$@b!xV4(>GAy0RD7Pk8C14ivBxvy#97`WyV1Bh50KV-*C
za8V6|F4yqY|Cp}Sj=1Y|==rq-T{Hl-tA{MDv{_R<nb=tP@eT2Oef59K#{qQzBOj*+
z0GR(zD;LU@#Uim-V1W7f7(JUZBDpMl7oE0SEw7ttXge4es)S^y?#%~$z9xWh+}ZJb
z-H)?DC}$`V53dH-6VCOqfrIC~gA)@o?HUFyun_Ot5HxjWmch*R)|WUYI!-S5O?cJZ
zc$)-XRV>2ZVJBKC?h9CFm9L~2)@vNTH5*wlQJ0C2trcReY@PjbG1@)Z!V3&LqTnk6
z$v?r!B6c^WIxp>II=MT!&t2=IWmLv{c_v#E(`^xEUjBxPFOJi4AZC!9%4vE&onyE-
zKUzH;Hd#I^POdEW-q+3|<(+|7va*^j;0x+AeWgIC`XUSh$2DTD%Pi=voJH)D^2$nB
zMK{1?!(1U{-FCn?r^qI8*LM^`<}dXV{X9GEbT@eQI!q3&C=qzt&KcAQ+wA4ek>;MW
zt6!H5o=L2<uoV_cOoJGT{tkVI#1BCZ^MnCi@^D2O;%I7s{GwrSmSE^mzr8ps1I`Rl
z3oH7_cNz*gW*yLJG*-^CZXFm3zE0~9>5XZUr~qL^-cDqPWdj=s3gx=FGu_-Ku8Uc`
z%p15v@KRquMk)?k+%_G)u>sv>*JNsnN@bYV5@XLv1LEJb9a}KO5XN;O|GF&`D^P!;
zVjJbG8t`47cJmZbVW?N%ar$kU5E(mmTX(2Y=6wVTy{!S|9AO_Y*<hDQ!ttjdXR>S^
zv7*2pi4I&Cn*cL1$y)#OzhT!T6ibCaAo9|T75z8{#}gL6ra}F9P9HEQjZI?5c+s%Z
z$CM6h_pez6{RTA>qrxF84yx2$b|ncw-H^X%YWn#V_~mh#$i6rPrkI$!`>CghiJMB6
zMLM5-bI`F*SYIi6tv9L~%j>h#BVwrnr2}w7V)hl)g9Sp>vO?Z7CG7%%c1`=bpUcdi
zug`4u051uDq%c(2>_m0nI@(fvqK9lP^M#sswg$-j;i|^c{cw;*)?(LurBMz+!>>~2
zT1zOAS`6yPPtt4z$cZ_rw<xOB6L=zxhK3bt4go}Z)|?06#VLw<BY<R$C#9dAw7vZ*
zMFioyDXJKQ8UnkmIUXh6hY8cOAtRfni$5I8=cHrcChvr&t%n^t%R*&M@8&W3DYzDH
zY-uKPi}Q-);^(#LLYv(!PpGPt;ruz|w5IA}9v%Y>V3+c(D0pU@P&}8K4X1tcF#6Im
z*(;|>i7i@VtW9G26Kzmgf~;0BFf<0)YfHIiiV0ea`Aao)X!?ewEGbZeY4f$^JNkm!
zW-|97qJ*Ty9`_>u*a86jDnw)u6^o*F(<Qngf~l@Zl`BC5l-FV=jidmg<7A>j$C#Zv
zqK62Eo))--I5a{x92yIe@{M1W(ynoo8%10fKO&<xHSW6N<?>*#kPn7E(cG+mmn5DA
zH6*$*&6_o=+jhEIBBvJ&EgY6FjYxSEqjl*xB0DENV%*53?8?Tks}XO81w<dr8PBdd
z5yOSoIZ(A7Rp+)ko8k+<z=hVDyqU+D`KmeV)~P&)pk!s)vo$8Gj7G4dy-h>htc)K+
z@%qMiW+6938`NyV_H|-+Y)ou?85t2Q;Hvmo+$QlU<1@!q^MOBo$~J5>v@s>SJvV`(
zbM#C{*G3Fs8IHSH)wzVFJ6*_9Z&Q!l=YSxVM%n>KEJ3}e$R2rnm6SDrtNY#ICm&Cm
z-er~-LrKWr$a;t4G{-#*2l47tfIoE3)ixc+*O^bA=cbi!=HrIG+4cqfqw58sqrPWj
zpz-x`P$R7r#U#=L{g5G?-Di>YFn+c{jg$SP#_{z0eJ1^_{)I=nG^X``V_^n{|Cfd3
zU91ggm8En|4fO13<s5YFKmQl8&@nQgRWi}HH^yhCqoWlzFflT=|M!v4#NJNYz?R?2
z!rIExz|tO{m6qSi+{#wYT1U@-R=~i~M9)C>Hy^FIfu#}fZ2%)P6XQR6rxh|Ww>Pk*
z6*AYcHxMw;v(h)<=7zGfw>8kQfO1|rO9xk$KU}G|n|PWj7nZBmj%;^A4$q<2rK-k4
zW)cdob_#Ewa*L;sU>A<1kofiM7q9DO`vddX6W*lz;O1lzZ?ew)See6keCQTk?cuUn
z<%`Gt*;~iQI}0c;!vD9wISmyR6)unGxKr9u-RhQ>7HLh~;@sSCg1;=lB!j?|6O*6!
z6Sb(QsG8nu%FD}_$lK)j`M-sDuRx`(tgIkFSJALSbYB%w1FtPx_@KSH4Z!QEuO9+T
z<O=Xw1`WqY$7u?@{{Qn}Z!W}t%Z7o^{x8{doPC=s%#3wLhKf9%>POqrmWKE8yaZ=!
z&E9_C?VeBj?+4=%wzi6)=&6$_28^b04Rv)ZT|T0>kAplF9=FqKcxi`W*u{cg3>GN_
zZ|93zn;O+UrlV+un_dskH<P~{?gpJMx$T^!prJ>lM2e{>%}!2q{rswELcYNDTp>i7
zt}ZXfCn`N0$#ldk^pvYbdcKV7-I=Ch04|1xjNPA)$7ws;s3}XCygG{8+RiW5EsZ9>
zlac#l{pdp|!N!Swd#yQw*V7ZLsWlT7#WXWjJ^pKHZj%w4swU=jeWpVGW+5z`(dpID
z?h(`R9-GR-FO_yhJT$XPt!~V$<C~F}cS4%EhE2~G>)q<ndXo%zX9pfPY-w^E_(#|p
zq56xDA06kMM||R9Uw{143W;V1cAZ%wWwgPgwS@0_l@8|Z%|tp4I_>w3X5^PG-tRI6
zSy>(V`LU3x>l+h4EoI$|^={J0+#jnR;CYtD(@X?{&Q_Il?#=a9Y=1Vz%3a?a?ps(;
z{kNc7yb~rc@I$1<2)qZ4`jHgXl1ZLUCXeUza<%jdP1u49T1Ljm$C>GA;-<{%iVAA>
zmZ2Gj_a6U1EEab0v?8fYn_J`a%g#=x_gDU(dXkdVtU=0Oz`J>WXrlBP*4EaT0`Url
z5a5D@M76PrCl(v?dKu@}c&l!7B0<cqsL(t*odOyoTIT#B!|}}(qCT8Iup~5?d~{gF
zrb>7X92^mP3!fFD7do(|Wr>_To|$=4D&tnIWiH3dtAD$fX|cRE_x0gQn3lrymjxRI
zwZ-lA$d<Ov5P}0HZkd)<WiiIntqD1LW!Kk#ErIqv_^PV1^Yh=Jp!6)1HM3Y}%^Xdy
zs5P*Xll#BC7%3?-=Qga%h=}C2d(2#Rx=8?g$fVx7peS)}BQ8!KI!{RA>-Tv~y6?72
z-t~3C%Y&R$`n6hB*KH`R<z%Xn?arUdP%)@_wjzs3v(vq}Pb=&Vf1i_<>owL#c^Oc|
z#1^3P_S?dT*=!Xwc5;%+b<d#Vwd(qk{|VvgLT*%4+ra_cT*B|1fRzavn?bllUcWe_
z3C0$8wxi+Lor_FXu>bCebb7i5A|lvgg+G<5{z747PY9~w$jJ8E8yvLG3{JK3+3(2f
z$qd*eRv9f(QHM)Sr;|Z7HI8oHJ`h$2U>LMXTUwT9Y{WX8o)ulDa=%Rg{qX&LnX|<P
zUgx@FWVm>KI{5uwO^lv?pUZ8{c(Lu=OyJXLfPCTp&~I3)ryD}61((7+E-fn?53;bp
zT;0|NS#MAl5{kvhM0hyC6rG^RVcYlecb9J^cne?mEA}7l#|K$8tkFf5(}&G=eSJm}
z@hJ``8vLI*SuCWJ0!g>($;tC9mVYGEifAcI?+!D4XfJ2M!F4n~;odA57`S(P$gtWt
zW7YPa+37!^pg=r6RwydZqU7#4>x0w+y4QTE#oGF8N-E-i(O%_K-t0V+1CSZs-%$7W
zEyl+)733Ln8~8|8MiUbLLZc<Y;goppi*5z3+kO4yM<Vfn%d@lVyYZFwRS9wNr&TNJ
z@bKM|QZ;F{Y$b8=Zl7<sb+atyRq!xlyu8{##gLeKYnRWes$Tg6afL-oCNix`@=+CJ
zXM$EjB3L)!nVB1~fX9}m74aBFRM*;t{3=WPukh#yq@b~smWIFr2pL`Le|yeeL}UGx
zL>J|kmI>(vBm<ys5m<$0x=4eS`B@0i2*4P)k*7lAz{hvHF&qRsh10ohex{6?^;NxA
zfdBsJ=&Gnlv1U7_y4q?8#lNm#2S`bjbJ`7Psj8l*T3x8B4m?;_0Ok~f*oI{I_?Cej
zf`mi?1*HSBho+2<4GOOF*a?agTGVJ9y&Oewf=<IZHdG{Jk$qtuMaVF=P%a4!Of1X+
zh})IH&!GY?Bx^e~6pKa!D(AcPJcSv2<Sj1El;27S2{>CwIJHXijO&>)j@wJ#qG6uT
zCsMju+k`)%jA3X=3V>1Jq^gQ)3Vj#lhYj}qU@TdW(<~vn(~DUY{Bvmf2D`a1P*W9D
zn3QAF+TbYf;elleZ4>SH@je;OJTXDOJ8)c~W|Wwy6bRA-G+G}*FEDPk+A*n!^Pkt9
zm^U_(`5Zscd;YLtyR+@IIGjc|HI1CFMy{;dxPg5Rc(07^yu5GKdfd9YBQFnb*Jkfh
ze}ao_qH@>PR`5Kluy{0;@Gm-i9Fp<*R`-TuUmh-b%$KTm2O}FB8~yy~+@<2;73^+{
z;KPTHJN^Cdim2yCvG;~%0UC;mVwxMwJa~m7I_sS>=H}+MwzhV5cBZB$c_OhWsHi|%
zeD6iU(Wuk$YmbXN0ggNZ+ssCPAY<8tO-+v5+}VU>W$!N5jmXH$RwdHe9ZWAb+JRj0
z(E$ZD@^-H_8ckSl?~p5axM27PviCn`tTPEXU0A;BnHj69*eYpjTie{+?CR1UOQM&N
zk@@5q4?zJGgs+1U^VKz?2=%MUh$JB0ys~Z_r$7$(121iEDadK4y1)`VTWzx01*)l7
zuBxD}{@PFoRNyXGX|7VMv!tYC0JpTJrlz2v0C6W=R7}j}c0@*|8H;%d4yzzL7g<J4
zEpmlV9mN28QICL?)$UB$g2959mspoqnz_EVwl>k{aS+c7h~d{a{O_KZin_W_0kHQ#
zhdMpxLfkK%r7uE@^SsJWN#%Hd-0coROxfNB3j#X|uc@z3XSTou`hKSbFxFQaImASY
zOG<exO^yXDbm+Qw@PEVO>PH3E*X3wy1QC*w7W(>v)TqnK&ZM!~NoI1x0(s6$!Np~^
z+fNe-iS|O;(%k&F4kC%~D6g!f<oQ^#&}su)XhE{h`l9{y-tcyBpx67;VXja(yVpDC
zGCcO5Z7l^ij!>bAogE6GqbLHoBM$<ZU~9{3!`I=qX!7xIi--tn70sN4kKa#YXJ8D-
z<kl&bRUvhk{Yrvh0cWeJ3rmcXcd!=rbs(@t2=J1Kj`jye_C$&S3Q|lU!%y))@qT$t
zO-L##<)@4M+8Pi*k!&NQrK=6&DndfNJwG*v3<rdgy#1`?fxTF3CUJXc2_5390!Cm^
znP=+Cs!F3(XHrGxz(5Xw8m0#nn}~_Yp-jW9)imCLpP!=)+Q%=)<Z)`MO(gaf$bWWr
zW-=`QT)j`X`U41c9Qk)Ju)8W6P`Hq|Z80${AgRPhPg~oI3-u$A{2b+tsH@94QVxow
z^#kfJjRGU}O`EHo6v!KglRSOF(tjXXhoZ|Za_Z_-)zu+CrS$-Jgi6|3t_@G5sTUS0
z7RW{*+1e_#wG)etkwsNR`3%>Zmi+viTa%A&zxlE<q$P88#1Oi#;G6hP59nlJ;nY5#
zl7?XLFjQD|d^doJU=sFs9T2z<HDFl(={Y+3sf-Q+?%#Vg_dSCH{g*{nEOJF4FA$-e
zP(C1yp^TTo!PEX=k#u@-JRA|N^MOFNzT(>3j+P{Ie{aKH@bBXN%Je$QJhjXQ2KM<H
z9zJ*=5e7#1>nr;|76g+dAV{A3ezCMIt1$0tiZ9;P4WAGq+Ba7m;NW2KkNKcNfVl{8
zqy@t8$czBg%~miF-i=nva(QFI*P)x&I_imn((L(yBlYP<B9n(!fC&f!e&av_Fl91d
zJ)k;r!U7U>*3{@amjsZT2g0S%LScIe#k+WEN`QG>K%&<>1_Y1?x9|ai;Kz7goKf-f
zAwgu`+latsXysLw#RHQ{%_x9O=IbXeCo9_z%rJm{-1)4)4HF53vbkbu2}wj`J}k`i
zUzd^zu~H9#%A%qDrQ-zp$QjsrGct!mNw)WdBs@?KU%|vgh20^@K*YSuu1CkeXZE5U
zFCAB_d%+gKfXUo@NmG-~HXp#nWp!mm@dqxj&-VqYY8*ymO?u|opPapg-wE_+7X6@S
zAkl6BC+q(-aVfmLr4SKO@KLKn2R7vY3zVCwesE^S^y}B<|A@(@qTVf(0HBnmo&v{t
z>8bg-xoVf%-88m=_2hGefZsq1wIu*^@cGj2f7stggoI!kjy2U9MXIS`LH~O|T-wIQ
z513Tf8s-0c=2{gJ(tuFNxMOh$`QHQW_+q1v#d2T=XJkOopo-es0mMY|XJ<${&YvFq
zt6a;9mdxn)rveE1IY<PAxrkV5Apb047-HurkDqYr543BytdY@THa~oUB-3jZ)31Lo
zFPY8Ex?P_)Cj66ZD^p)zKANENdgZ5~r^oRRgJd4lbViZQbuQR{P*xPx)!D2t_Ua9~
z1HI@U2q+LwPX^&QrM*|shQN_h8b1y=k@!8^Z1?p)x2c9lS4)by0zO@<id0mz<OjF<
z>=fPie|-Zf92+wQram(>LjR1-Bd(+KBr97I5iv~12^^-Oihzmz+nW>ZfA0R$)WnF1
zDe>YkBm0kbWhEsEz-$5<`V)TyY<F4}HBw=*Qo-No-B-ZOo9w9C`XKCMI+y!j2OYHx
z4W$h98D(WDbew_N4+vaB0&_?(d14~VXU@U8Bmn`3TQ=YT(Q^e1yr@Wa_obKDr@;T1
z1=07vYz_r=(%-LOYO4BA<Q_Wcf9#-jzJBTnoCExG>`U*gurMJZAto4jI6l7N?z%b(
za&m<N3EFQ!Z~Hj|`5yt0stselYK`A|f$>B)yaFsiq~rga*osG`YviP*q5f;Da%7Xo
z4LBvxbOLGsoa-|1o}W{Zlg|T@`-ci!NC-ttoP79ajCBLo7_bVnyUbkf^#A&I_v0-S
zvd%>R=rk3q*UvBN$eND2TvS!HJU15_Xi>p5zvkxp-T)I;R-#YDh-9&eJl^g!4$3B>
zfttF@7#Q3DhY3|xnNM>B<AcM(=z2a6qkg`c;0FtPUTcV<(TWE0*|f23Ro-5|eJ%_e
z6BO{>b}F>ccGr5?PBcQlZD6JsMW9;MH8)B7Pi&OmV8qkVu!@7D<3F(@*zz*Ltj<&a
zZif2-%|t3bRa$Q0|JkEHwxoCkLhAB``TkEYn%>qnHpX4<Kqm?J`D|DkpV1(Mw{o2W
z=2K%O*XPhD2Y$Y5ZJ>ogz}Z^t$%7zS=711u(-svJIIOX15aufXz{UR4)t%`2NM|&R
zWCsKa#TCxc(UDrq(&f31rv~%$t`absrwb$b%*;qX<BkP<GSjQcG@c2$12*tqhfGp<
zHW?XFv=Y7jdd7HZE5m>@I!ema=y_m`0`6VGXyS3R<qqPLIpY1Eo*aP00Za_WmWh7g
z)`Nj#5Cg|sh0?Omuq9(#S{O*-T5WnTLh%3BC2x7T*_F!7pPeKN-oP$roJ?hd^SmE*
z8P;OdEBwa`B(yy8A6{M9K6S$YKR=Hh2%9c0HRD(G^Fv^Lw6s{+6Au0gM5L#$8%Sf2
zknxX?4Ejv05I%lhrDltKz|{AX?X$PFfwa)(Lh}4974=VX$KSx<19E-WC&^o1J>Ma)
zu~8OyP(%DHzUalLsndJ5(R@k;2hZfDj*pTiCoB5>PwJ9{gT1}S`%`%XGur0^?BK%#
zv7}f_%fvrRyLlmi;VF@lml-Af9jKP?MECBXN27(E(0>%}SgGpXm}mO*EU*;6NZfsR
ztO6O0b>Fh@-~`rJ9u7xS-?%l^KgoK{TIL0qWOa0?44HnKnCBO8)&(pxeAb10N4tU4
zbbn^^2?H4j>!S_>eei1I<hRde<NSck3M3a))NzpN`v0IPosn^Dwy@?0go1ZZTM6Cm
z-CYhk)LEbD53s(3SGKeiqf#68NoxpfMY+gtcZ*IKZi}4%f)pfsXD5Jwf9}(QTdcwh
zKqN$fIfEiUuw=AUyLS3|R;vc24l--(wYuTx&f49c&q5X%wr=h6$LpbkXSxc|0NwDh
zkx~c$4^LMeR^|3|4fKGtpa@celypghAYst0lpxX}-Jqm&N;gP{NJ<I_BHbX}-Q9e1
z^!~p8?(>|>d(OM}?3p!dt=UGi@-i|ev>2g6U^o?k-|2T@_5ghA)^$KSG*Z2Xdy*1=
zpWAY0e>GE8Yx3qQo2a3oAr=-EDJf}bY3bC|{dc<fRo2;F<QwXL<12;^(rRm~o@x|Y
zrRBU|KtTIKM_k?M1|KVH(bAH^`z=n+$7R@rsigPS-1EF;;e4tQ@haA2WnCS7=l-p@
zn2+!BkgpJrMmq0R@lE2hHrWS9nMvxQ12~X@EalR~4IylPR7!?dR-Ww4x7F9HK-);m
zR1gMj&A-R_6LCp%3Dn#gd>3J4Z2a!sJ5v*ri3874<DuKPFfZw$v0!_Y`^(74hywpx
zbdHvrjgZq|64lq#@m6<PHDLuSEB5fWAL+wFds^CZKiW*@PgIwBGL-8y`3X23RL}CX
zMKYeAtQNYR7Y32=iF^1BlKLN@s%ZKU`1#XEL@3fR=9}J8RZ>(e5rG~%yPzPbsOSeH
zCUKUcx;px5t(s@U#K$*IPN%V)#${&XCmVM7^}DG+q-19cG=JwqEp2j;D_AY=GTUBt
z)m1&@4E-|*orHFycb!c_mIVuEsl2>=X;Bf$D@-ISXx9XIbf3AMrV&3u$n9b_SZJcg
z>G0xmX{O14id>?~_WMNbIqJXv7rVW0A~X0}UK@^=r{}#I3N<xof+v~HXtD`8o_#MX
zyYugV8&Xmpd`R`>*6OWzQ9qBm^976f7cAe1L`>_2HEHQ{_4;W|A#3R}AfWU${fs!~
zoX3Tr6cOpFhi+cXza4@nq==Xw3fS29EB~UIKMYu*NJeRh$red0WtDOaSPkVt&wtw`
z(cArr-=YBtbYgxB1I~e8eXB$BU+VtIYtt&D#CZ)qBy6cKU^)QSNls4eU%)zKnvTW;
zH-Rjn*LjYngqzp?H2?Qe4*V2%4=Xq#As>m)E>4|**~H7Jx%oGjT&JSTZY>mxLIyfX
zYimwuh^(RMI{yDY?o(779VsuoR!6M_$FtsUY3{jD%6E-x-V(3~&5s`C7z{E{A{T)O
zF*>3OJq`~~>3?7Q$avKgeiZU+4T-I7q%_LfMUD=L{?*pL`8Hvd!kbVL$ose}48zlk
zEQtSGS!3W@fTclxl*SFmY_P6Q$HoTzzfWH(E7Kzb8Wy2R*lJfF8c53A2mN=6=Sm+v
zibY_^&dfHY!s;H0LmgS3iwO9)vO_z2l3Jy$M1TLTJ}j*H<Mq$L;(1vX{$;jXXf-PF
zt&c5{WHy_h2b9_>yQ&9&pSyM=U)@lO@xi}O0GVx#M<Uaf?={39lLMfX=jWx6yGe{S
zI8+XaVR($8L?z?jQ;ENPQTp9#gxr)g?xnT(9iSHeOEALT-mG<JmSq2aP>s{Ljdkk3
zgMHvWSXf?OD&-03aN&JOhaMz|#F$rs|Gmh{+8PndlkC(~Lip2|#(Kz(Xn1=ee{;ft
zyU5esg?3hCg%-EiY7XoUr#b)MM@*U+bEgSFVRIkFMlzLd?RSX$OOxoaoNcrW49K4X
z&jTRudXB%kGV<^5c3#IDit+JQB4e3{7@DgX{y$g#9*;N7En{bY{)=T=td%8EGVUqg
zzrXS;!9yDTrug4y$jwh%A5CCnnECf|JFg+}2gJ4`zl?_n`0Kit+Q^jrkb<}$P;j42
z0dg&AvF;LZee%!UvZIrPMCv8#>G3i$n%#)a4MOeVcwJJZUWdTH168f8t$m&>+zN0E
z8HV@>^KnOOAd>(4G%|K%*5#BR`Rfgd&z5OaUbflG4IBThJ3A@>oNl1#{tH5Gs|-2d
zXa9v+;>w@mUO*DB7#{!Icaz6e=2~7SCy)RCkL6Ae9B64hk$c<69UY~@WfFb!zdz0?
zC@`IR^5xq%>VGH0#hFc|phWy%@laS^5;Y%PK;k5fx#SGJIRM80JE+iNd;IU9^1F_y
zJf*z-FB5Iz6UB0AyOXlhtO69NhZQw7#rtnXIrO2!M!sMSv3g}c1uZSXzrP8F9)Qmv
zgzaq<7qlMr+p{)(y*Z+~{{nval-GKx-U}TYxkE+7Ym-_>Q7QirmDVF>b0Tip?DTY8
zWJLU~kBsaDLf`AZ6a~D-^mO9{a+_4K^z?^XT8;i4lsT`J)zP0#N|si}{~j9PPanh6
z_g~6owmv;bOLD74c8ih=Z$w4w8@%-Xz4!M5;Hv>k&&}OJt`{4^!`(89u^1=eUy6ED
z`1}>m*)K8@<U39i?;Uii*Noi!_l|Z-3Vo00vSY*K(U9x8ICXxSy*f1dUq&d9PC5|!
zhi7BDg!!sO^*XO$L)HtKvizErrQJ<F-_YQQ{7OLt1gP8beC^U2=ikEZL`Ax(^$*Pe
z)x};6390hGhejp}F-L{T=NFnI@yf@ksT!v)yS4Yi|8|$r<?>v>be$HSijR*~`Ta}2
zn*M*an11K5i&WrsRZT+EXBA@;6J~}y8A*oV00s=y7F!#s(o!i7Hot-HPl~drKxuy<
z{S|z=F}e9Xc)V)-vAM1D?Ay`D`khwNiF;{6->vDMJ*OM%X7MJvD$`|6mK34i^<p`2
zWJHxE4%M=|i}|H==dFK-v`g$Ljzu`QWv9F=FV95!dA9Q&ZBn;5aP*XWOL9pheCo;t
zC5Zx~_U-pb0IXadC*9X_b#Pje-G^XY>|lQdve>z!6IRbB7a(jNM)o{aU=M3~7F=bQ
z#KTKnx2>%o)#3pEs`A?tq)O%Mjli-uo7||@>!|TUm&%Z0j|lBK+UtwEev`IGLiO9v
z_@(FF{R4p+!Ql~#J^lSx$!^gj8XI;97#p0~{-PRT&*H%QBp0hbwhcbPGSt()6;>dC
z0M&<15F2<e<#`1zo&xn4gQkhJ0cFAYv3_!-aztQ<mDXe)jf@{VTMqrBSjbk6la<g$
zXJ<2CS)U3?PrYri)a?QC2$x0b#bVMVg>1p=cP=XtNilyEMxJp5U*1hU76T%}<8O!q
z-qS1g_aidkHhuRQI#;3k8Q}1#WgaIN_`T+7Kzw|AXa*kWAn6kSmoG#`h0V+u`%rtO
zR(m6b>_bxWh%a`#4(T2*UuRN%<2d8VydU@OozztFlqK|hMrZaK{}R)t^4qtkeVJc?
ziq6Slu1p8nAuA=V(`@Sv=gN-lhBYC(xKAvfOxK>L=eL!WyD8;n3JMY+38v}{SK8QK
zM?3;m&U#yi91~5I;W0O@Dsg(z7*Ov24wM#fK)c7gmPV5i5e;tV1Lwfq-(JfaisKy!
zc_u*;QeajRr6O)=c|#z4m*#LYQ~q$sSf4FMZ@jaUiGtiB#Nb(J>g}ec&mTP88-$4}
zyFUflQXo^CSh&aM&mAc#4&vfP5fhUx`s|kJgZfG1CoA*<MnIMxWKJS(F5u($7ymIg
zwE`+u_g(3d;a4U)_TrP3fj28bfTForeepI5alJe|(o>!e+<13Fsmlpy_CWT_mq4I&
zVq-%AXBF8O9YanBZr%TG%=+kJN)FNXl(4w1t(3~pX~G{qhhFm`CKZ5$AUhVkvI-t~
zYLsi3bp?wp#=~kcR#fZ`G9xJ;ajU8f-6hq0A}|gK3o18M6r7_a#K|O|oengofES!9
zO!rPWA`5bi0NrN&{J8^?rFcNMwXO731Yuh}M53o9;z6#Qn%pdTTt|jk=K$ya%D@x|
zUgAOhD}53lZ&voN%3XhEO3vvqHvTmuBZ&2_^%9BpxE(zQhxv-)QzL9>M!kN=h&0xx
zNw`3w7JUdUURFMf#n-endZvP(KOY92RDq^>dZ<GDSn{dn#6j;Da+=J!_VyJ!=j>m1
zNcfkiQ1I{%gKYTtsyY+Ulbg4GhwC7#nW8e5kX17o@|d?TzH#ECe_Zynh1T}c6>j(p
ztUH=H08UTn(pl>%B^wJ%X4Iy#{Sk|VxOiGd#+||S+Kow>&k>4W4seh)6N*W>iN#+y
za9vDF)Y&=aqwro~U1dk?yS?MZ@(fAVVGy%axOgd2KLo32^$(O{k`)l*gkY}L78?8%
zp$+iG^)8S#;EoQFeQc2`h>R6NpiW+cb^`z;l7V$|ROKU9mqwL(Nhu{ICA!u`;ja%@
zsxN<$VZRy24$n2t$;*qxX@B*IgV@AezlXfIv~+ipf*c~F?a9DU`SGc~Rvqa#sQDlP
z1JR0kf9=klMt65hp4=l=Q%z3&#Ryn-PrVhxTGgnVw}I98;DIqsZrF-{9XICu%B#Q;
zWTU18<Rvihy*)!9bb{b7;C7J`r}o-n`+h3`j)nOFOwzo&K-#Syhz<-eLhQK<vSzbb
zykC@#f@Ft`s7O1ISTHa<s$5P(tE=-t5`q?(&$eP|N%YrG{L6Ju>GtV2YWIx+5^Ibb
zsFp8|ly^N9qq;_{i;935UQnH%BC7<lXss;AxE`4NfEtL*wiEn#yxr>R@-JRYYzhi4
zyx-c=-js?X*48MXWnt0Z*0BQO=5OIE5_zaRE2f-QR;?^2=MmVk3wDKA{y)iDLSj5T
zxxoL^ZaeOn5<TT!nv39fEV0$lxS!W{M@jiH=;0p7H3^gk+RWaijqw{%=^f=e`2A|H
zsLoDuPoen*dKVZ610!+UMbFGD7A*9+;_~uJ(Y1GV-E9ohA9l~)(w-qVTvJV0>}o?o
zkJ0m<w6wQ77w3t1G%I-IbMvO~j%aw*_hfbwJmL#)?C&N1s2m%B{GiiW!!kAX@p|IO
zNP|!n5%;H&15rqdhZ_Q!wK5=hy?$-Fs!Vg;BN`CE#M@6@#A8)MF1vaXs!>%dOG_t7
zm){?Bf5BtJ8u>$2pk*EmS$cFxQDtZW^nOrKAj}~jnhw7y%)OrwCjX~u{hkm{RtheU
zX{<IdsCjJ~P0f)Z+4odKR#uGI*^8h0V4d&v$xt-a@9a>6G<~RHhvDtx`K@eyJ#T7i
z%%9*ol)dFq6=CVl7tiVVsWYORX1cC6NMM~DzXNUoAkU9V^(X*tDNiNi!)e&qEE~Ns
zLEyWb@n?oMF}<+1*XbY+{}raMiK&7DJ>YtfP2hMxyrx-+FUq0Y;5c(O^XB#I&z_zy
zo;@SYYeRQA&8G^F%JT~hyd#X&+ae0E&)Y{drz|jVos@JLVm65Qc!s{NBaVcTan{LU
z`lg;Gq?-R&G}R~xj{*=vK*$MT^f$2V?qD=MwtRVy^!CY!txZu9KV7Y>fM=|r(fMFO
z=>U}837?|-W2eMxYinWQ#k|OFb3QbPcpJ*r6Y8ihe8lnW`kh<TjhUA!R*Rka02-Ex
z9zmm*n|uE#WoL(|)iXYR=jOE1^XJd6Kg=oZXv<~c3jR|$Agr$LFe6&?_E$B(KzL4$
zQH+*aYoun7*LB2LBy$~f*x0K(B4}t{9yjUq^&)FG$qp9YXYmn_)w-Ml9ielXySHa8
zyoq?OR?bU!|Cja@0!_;FkBP0M<aXfwl-<R+UKVq}A(UyuAhXoxm@*%FtgHk>K`8`+
z(#GN8PD+?OcVinx?4lU3tq$RUYP_c(2p7T@_{6D=)Vh#@QYjUkY<E)P_T~*%c%8pC
z7BcVoVD{luq_|&sHPpxMvS(mQt?qUMt5&X~gM>q`Aj<&xj6B9K<sXIHR|X!toIYaE
zGRw{;jQLE3L2!0Rgy4Dkh2~*IeI>vA^XHU-9Sx1>(dr?osj0!5p?P_E&n!Yp9E#Jj
zmla?;B1XRbS%YrvWt2PMB76dsKc#I0J*=-<7Th;Tu$q73zkHVE={4V2$$#6wDP#H%
zbjTtKyxQ~09#4C7lm@}^v^8RDZ4C{}<@PTR?Z*GS4<DL`&nmk#P`RT+<)mEvd?a}D
z+V~Q9=Ou{osiOd0slDbklud|IK{P(a$b(t*D)8@sa1Z$n7WEOGS7wItEf-UV4rEgd
z(xcarC=Onpu_2+5ZJxALpsa?=!#v`L3H-};fHp9&Su@L&lo%Hly1@A|RG2hlH~D7_
zJVdoG|G0aYE7;hyftXxwCg5)4K*L;!VQ5&YnHYJOc+I!BFV`RnZKKHs_{{#Yygc6!
z8XA4EqtTMzJ)KNNwY5?x#%+99gM($DYgz;5f_J!aV;+PGaR~x6t!2+>kMukAc_yZK
z+V-EAOBosRhQL~~5LX4b(sXiTZ7sfcZ0td6={zn1I@g5`b{^L;i0g5G2~e(4o<8-m
z3R0e4NcMl(8b-AM)~L8rMkcfF!JyX_-6qfmXt{phoDS_gbIL=B@{p1uu?`>;SZVb4
z^1wwWhT?;V0(LFWQtO5e=#H5P2{EVDQb>t)-<Y<w#_sJW$sS4^jTZmv_(f9;g#P{<
z9Rkf=ack0b_X>&w0lKvGZ=gFd>XbkkDL0$I_#*2*{eX0!EGMUFnA_~wvJIFBVDkq+
z`=_l7(<4XlBp_4e2l7pwdMM`r#E<MBB!xa%;!#x_eZ*8ip}T4h6ni0ZD@$z}AS#h-
zFuJg|5({oU-<5qL`1+Souh&L>q(fASe&`fI-2p!W7klyRrI$2&#1A!UiT73uH((id
zkpbmkpN@BQa;Op!RSrD}{`O0)BS3v)OzH;W2pbgw@|JMbtU~is;EljfCt{nFOluVq
zL<C$m6!?3jCtpY8`?i<lE<hg#`Um`WU~yAaDos{)o__P{Ro~vS9TwTaO%w-w^!_W-
zvFXT~&&ww^k8TFmg5@ru>tN=(?(H*CUIyxX6w?jB_&o3YKBnqKf9;fgNK2f45_~_?
zX+!;r-|sZl_qNiZ5sEcdG|K5OXefyD3e147du8xi<bB8xz<t9n8RqrZK&c;)PT0XK
z{fr*{yS~4_75aEi4hp&Fnm<LrvoOp`22y!+w2mBo!l(Bk0HTw-onNA2?K~qP(KtBh
zay{$$Xth{XZkO>@+JW&slJZDl^)0oz*4pB)nk)|7bRQTz@IskU{)sPOK|ueN{$z@L
z=ySBH>sdu1$cPPI)YQ}#B4tyd$yO{J95NyzO-)ocIOoI|x|iJJ3kxn(b(w4LkYI(i
zcO-$Mj~eBah<GmYlbzkh&&~v6Hk0UPIQ)P9aAq;~aW74Uh|-bZNBCc}6u1jEqB0W;
zsQ6EoHgQl8T7Ib9=83+amE2K@NF6)S9z1XW#ODUw1Jnv&9keu^KaKq3vp7USRKXvZ
zb!0d?)mJbDJ7H6PA=E&(%SAOLfquiK%uRP6ky}zCC;Y&q{a!-00w&$tOUKjeMr}Aq
zSq6Y8Csb5zZGjiR##OWOO~=c--u^`G@WHl{<-(mKH|!!UFV?><`g03MqJN`j$*zxD
z!srU@?kwQd{>b0NCI$vtH6z}nd)HT*$5r7Di}27MZtJD1M)a-3;I9k==3cts3PMa6
z2(=e4rqd*hWs+Qfb+as(%iKsVb6$J{c8j0BCp6MWGtv$l%_3rAyJe=GVB`^Wj0m>$
z<wiVjZnj|88^ax&zY;PqE+;Fy0pwKwm5uS$iSt!YIkgK*7T_yE!~<{^;qPDcgz^@s
zP@r<d&wl-y1s8}h5tGMG;Nk>n=7+Ghq=zkF5@>^$Q!)?dTsA*EF+FY(1_G(o$b3_>
z-_O4Yhd_nRJn=U;$r{S9k#G+`{T#so9;(_|Ckl`wB0}X{Hq<{<RQOd_n?8AR;4@C{
zF|y<30j9f>&E~*9sTVm<QY!ACbD;7&_DhI}h`pr0MH5qB&(+$=w4GbHhqBFZ_1lS3
zk*bNxQ2%@zBLZ8w%9<_<)Pd4F3|QbKVOG-<PKyFLUD20*kZL6+`ezTW#ncCV8_oq!
z1Y#1nhXgnX*AX`q_xDYSh(bXyU{*KKGmdNdeB-^`sOG{OsICBE`!Q0rrz4Si9a0w8
z^>r{YL2JE+b&;Qw^V%Cr2^+QLCgR3Rh{uR{sNM>AxVUm)4yHw5V!lmJ-|`|K%2Sh-
ztuu~V1!pOE;#5=a5wFz(<NW2zz|QU}O4}C2s_KXq@G?;0)GM~W`eZ(lQxmlOT3Tua
zK?8vX9#7er#N3RG{iAIJWRY@np@uVnjEHAa>)SwNU(+SUZ8y5>Z)%>zVU^JXQ2-ri
zwb~6*Fd|@Oq$^6t=~Y-|zEn^^wDvwNEwRqY676nz-$gnU7AB65LVWhdbbJMAfaPv!
z+0uRWx*1YRl|g7s+FQ25I<adAhYsP9xc)K&L$MTc0elj4Yf&Jz@>R9>|0#P1vK&ka
z{C_?WVFCr3lV?)v$f+tXZ!;_I)6xP%EZs_Sa&@M0ed*6ne0?iI^CduA0{5Nb@4~v;
z)17W4UzZ4!wI78F`CVFk$3_DDelngJM(T`ew>o`Lr;X@QKs-+EobP_X?Cdn~^!P}(
z6WnYrCnP}a0FN|R$<N>GgQe8<JNpLW_zN_KKxp2?loxb*oSc`3R(j`}jZG0iZ&>a4
z_}e#F??$Kn_z@4Be-@^-<%^5e`D7$l$GE&IM93vw4!kk(+B+&LrM0nTiU%o{`$jBe
zn`dIt?^VEt1_^%B*(%EQXlj2c1vHFzZ49)r8{<wFWrZzBPR}2NT!RgDJZi1q)Vkho
z2YJ?wH=R}pB0e`B*WSfw@bv%o)t*&r3H%$Xc?3O}Sgxwr&rpatUz~y=4Wb@?$1oqP
zPTj>yt@C#(l?57M<+`U`Nm1rwODwvdD$8u~pHf#mjeai%-2sG)OpMOfiu}&hl+*7~
zKC5YwVj#M~?=vBVRMqFa431tLSHJZ^XJ6FCxz?DOuDCHW|A^{KPC9;P7M&(8wLlbX
zu+8CO=Z)&eVJL_y9xQPACh-mFf>{zyu{1p<aE8uuR#phiX6g8?8;I|KK_3yl!MUoW
zNO8}8hfup0?{gGa&y}6h$2i1m8N9re3=A^>Ig<oNhdUV8T@cgAl|>)uFM$nnp8Ler
zS1EFp)7iqJ5t<@sr@$Er5WaovndfgrTm1vyksr`pOdfcFB>K2#IvH9NlW6bw_yL5B
z9g+^_mxp%E#tRrUHX}CYkL=D*pH_v*9{~{}5qtuL<4>F)i-Y<6!eDFLs66HF#tnh`
z`nsIdnI>J_crkV-`&~BRs&}LTzf8Ise}C)q@=&IqGFQuw{Wt|`l6jrOmd<OF_aSwl
zAoTY7&5W)04jym~eg5cC2&)!$a~L3@50*enXTe7xHobGde}CI_D;E@;DyxCJ-(o2r
z$C(<(aqD<41Pk$|189_9^<!Bs0K|)=#^iJy9hK}7v9T<J(A_TAfX+()GEks_K0TA)
zirjW|mkz+NpzWQnClkDqa@<L`-`3d$S(j9xl^~BKkYzatYFk!)6j;YBUu;}xK|UWH
z)oQ;d^v<uo!lJTY)asng5Vn2o2JTfZb6fRqKdW<cyz0~%bQ>2JhX4ce!&;WwxRV#4
z_^}+1+BUa7(%yPgN-0sjx;i?si+WoQrgJznzu%U`oQmZ<v@c)bg4H^BjT+BKzyeTL
zPmhy~%Ol1+bDNQ54N}lhq0g3?%M(~GQ?RP*73<&s)g0;E{Y-w2``+EiI-6X)#g{+8
zQGz6TiDF^|K8ML;qWylVTo>UZQJI%F2&D%O)U_HbTxxUt#Y6%?jnKPb*S?Xo=6?28
zM4>OUFLRO*Y@%v>D=zo-k&i(IX#es>&TBOq^4#KLV`!+sf8hkA4e9|y!(3r#K{~#)
znUEb*vX2APE#mgp93GbuEp_#G>#x;>>hdPejIID*@1*7HI}gq!c+;`d^1Ez5x1g;B
zVqhsGGhTO1Jr6Y;`|hIOBw(U)GaIl8t{$V`VnqPAq^_X=B1L3~ERR)l60>nRQrOT^
zd_7VDdUkxg5_4hSPn1Rp@w?KACR>fzw&3||D3QUXW<K0>zmZ|Ja1W67`&CY_Fr>>O
ziWz}Pm|Cmr`yHrRK}{Jfi0(7!YHHt4PeY9W2L^@C5s+nI6S#@3TwE@5yOsoE`=UDD
zn+twB_>8R%OA=WuLa>~nivY_Q*o%URx20d?AdMKI^WfPeTj>6<wKAZ|E72;^2)6sS
zR-^NsHWxvr&;Bgap*uS(2b+g;4A*)`TF=FvczSq%F4=#jx7!Y=7Ele_zjF|QU-`Je
z-TmYSy5vwn>AKf3`>~Ea?*um40V$ZHfJ6Uum}cPp9ONZXX!=L?&UEf+&y!<fKobJ~
zOfdThxQ=nL77@duBhO(?g3VfVS4-&Y>MEs3EpIo7@CK9`z+qmcnD~EKva<{XEgGUw
zA?(QuecoSdtp~D$h!p1n)g4M^z-uNv_&qUnaCUGhIj~9hX2yF(Y7CD+jf2gDvov9R
zlM(3(WHtu@=7BzL2v}(~?Ipm!8i)h5A{6eYKD(vH4`B7-yrE**YC+8hi3%ziG*Y)p
zrh(C}y*Sx`Q%FS>Nbzk`^noO8FD&YU`czL`)afl0IB%~C3bKXH-f)3CzCfb@)|kbp
zgA*)W_zYk#1Iq8aTOG~N{l*Da=QnRUyHjfHwy_)12_h0So{lF~RNOi*uc#;&qHE-d
zj@X+y%LIFJs+b;BYOpB%{Y?kzK+1!#8c3SIzMeq$#9O~;7=}>50Q}&6t{A!GL<Yzh
zKt!g7)H$$6nEZ}1eM5}^!zoC59KhTIM+K*{`x5B@<%U`76TLvF<7NpYD$1jCrPDA(
z0|A^N!7PLI0YU)i!J)kGG?Zy6*@S=ke|DvX$r0z}7_qW^dR0BFE+R4yZXO6k_X_@7
zNRL|x@F##e;pD3npd6k#vQH2SQl^~%41++@ZSiYF4JRe`V9sz%<Q;h5`_Kh1oSQl?
z?l%gS{3|HWkdfAgWaKm_276cu6|RX)y)j4n9f71Kicx#r%-+2P)>AOeA(i)emxn0Y
z@~=niq4X~Dy0tnll<+vuy^)XrH~LKy7vQpxf-@rdhs!wv?3oSeKuXamV9=%sn&5Jg
z5+*wOK!Zan#Du?HL)oo;8?4}Oef*<`{n3T9OZ7(<_5~5I7;FJ<0)Ol8p9@M!LGFDi
z4-};5H6|JQSHQHuymOgZKa(z)s+tDlGtl(VpMNz*VU+|;JU{RDT?Q5s&KGJ=DZkk%
zSG{FJR8<axB@n6**!qz3EaR-@wXr^M6yJxeZhy%kdO(+53GmN}BvYM(KdwX2wp|~%
z@1E~ZWuqG&!f3_TG=beUS5U_64mrFhryGp1<PQmMhe_NZWD|8c+2ix`({_G!iF9ez
z*B>fDV;d%4;py25Ra8t2v&PWgeF{92cNyV6d#^o6VB;;e;Kl?#0Kh5a!zZ##`-`~9
z9Cq#6H#j<BVe$#Avz>rKdV6z_hG=|lC8g4Mo>`b1ko4J|bS-XVY)lfk&tjg4_EskC
zf#gL_R#qyQU&y#+4QIs>WJ!|su`?g4h|@QMt#-n<e(@aB#U5P;6YB#7F)?<MPHT1Z
z;Ww8-FO@5Fj?f86JL3&KrDe$*9`aQ!EOc}O|978HS>hE>wOz*bchrcuSl;*q1!(a<
z-wXs1D?3-Hvn|)d92QTur5}bAN&=E#eg`<?laI|c>Re|Dd`0*<{hp~Kl33v&{Xo?K
zEC+aRqO=YE%_g<v&HW8D1gHw&_BaQV=;ehUy~=ACtJu4!e*f>7eFIx4$3_BImo6CX
zz-(Rk*i3)}jx1QQUn?tP`n9*S7#bLqe|vER2IH&OBeJ%)4?6LK3^4^iAf1TdegOVJ
zmJU?MYP(Gb4HD^D509zjmgSzbGk^nh2pF$(wYL|>#*{e%Pd2jbX~*29_7p<+q%*3;
zXU)H=8XW9o3gt0QIHX1xqTC~VWk21}2r>MSO6hQYcWbBBz;xY{2e|@2m0tbHE4V<W
zN?*MF&k-VmYZ26?cYvA#KS5ezJ6brX=x=8$08pU1fk%uT<fFmefz5?C5G6q8Bnp`Q
z6a>v`3uc5uU$3G-5=}WY#3soS7%OUQLNYS50-e$06}Xi~pG%XwXc&mKMFPL)82yCI
zgk+6@J6einaAtS0@j<>i81QOOJ2T9+Zh+s8DBZVT$_+U?=JFMpU2JU92$(o#$L#Oj
zi{2e-Gt^WN5s8V2cvQeN^)|YsycjYLxDzd!%>j475Tdh7v8P_*;)Yd-AREsCxYgkW
ze}{wTWAZm0;o&B7glkXRyQ`~ELBWY1W_w^j5scc&X4|HjVZdHKX89z%?<gb*Co`)f
zG^%`kQp*Ck2YMBEN=mWq04;6p!0!%YU~ccf1{03f8|qI~RRuepkv{65jE(y52Ed#k
z6{kl_nYMkpsR<6dLd|!$I3WUN@f%JFqGF~_sn9s;FX9^wNCfp0394I73SvwA(p{Mm
z%TF{l&4aeSGZB+qAB!lx+B7a4qz3S@abv>zK7h;!_EjZLq%3;6ub^Jq`uZjPBg!<#
zWqEX`(}xj3nc6Tu0lbAZK;n0<leM=d-Q0rxl44)~{t^8AZ942I5g%~!C^zOEfqTjb
z4A3bpO#`VCR?i-|KjE_>t?_J^(XaKcO}FSBVKhzLg;)Tr<%FCs;&$(WY2>78h;*)o
zxwc||`xn>6cCay-o3qu_lX=K%@7|Sy2XmqN+I~~~&{##!#|ig|r5{&iR&Zd{EETiD
zie_+xclzR7k*KccZm+GfQt8=x4A@G<t{|tE`18j>;VYlyhl+c#Uw9u(`A!b+-drQx
z@P6*-`1$*BF{ttId=Q~EVWdxtY6jx2wqyb@HvkNn%z$-<at^&VP+8{}RYNfJ1Wgsd
zBfPqY4;4{mlH6Qh!mt~y__I{j+oq4}Kjn$w(A=Y=_3X(cXNZmdiA$!TF8@hl4;7o4
z{85jGAb4{z7ek{|BwUyTzI}sXp1{e{)!Kl7o25VM&SvxsV?yH8B!0~EvvZa*@O%&0
zy;Tw#Ka^e&18>~|_5f7-S#es^D-Y~J!~)SWlDRycF?~IK<YG5z$k1AN_{(JX6?2#n
z1R!QM=_<OP*>yVi+ycDtw8gh!tF@2pJ~a>HzkvLV@692bY33FEECGEm`sl-m{M(t;
zII-=Q8n-c7Xm{yCz>|C%ug$^b0h1TMn81zNDkq|vS+LyqKZPVkvakG;L}ZjeYb=9o
zp_B7f@xlb)3otm`RtEkQtf?TafL$E7uC*<$!u}oB{%6HNAZ0#D#ow8LQ#V=mL8WU^
zDYwwPtt0=-t)nB|kzxK!+tZn*A23(A<f7m8n3PTQJY2@C*Uk5+U$6I9CQY~O)^eY`
zTm>ou-XnBivl~D>vw1v>=B4s0s<=sdm|9reS<%UgO2M^(WBgw7b;JG$`0;Q=v&2re
zC468`>L!Nl2~01(a3y-SWuW&`gwaN_-%BnaiF>-rFnkv5y73`o{HlT1l7=NTCIET?
zdv_Zz`}`rVEc-AY6$NqMg4|}Aj@Lkx*2IGPwn&18+UVp|DZwDw9LAm0RKKi_blYg6
z=Ka<6OleNo6!GZRBd`t_s>qv$htqqmty%)|LP1PYr>Z3h-ZHP|3Opfras6pybdsxO
z0$4gK(OQSkPX(5_A|5g_rgD2o7#iA0N)~;2$N%nhq_phYx7X*V^h|1c77GD&H7@ed
z;1I+KwoV5G7#W%F?Ip_YikNeOXB&<u(oPU<;dAY5NsbImhbZb48gg>i$q<izM!{ea
zkzkLwyl{q0ofSAH%Iy}M>dOy@(%h7wM~B%_j=3~}mo;a|7EM?tPVd=*ic%mdMmz2v
z;ph)x<TOB>rKOQXK=)cFJ6#_wEp0gX5nL7|DPQ;ku?D`G?|}4hw)PAS-==*XU(%58
zeMa`De~<3V2`CG)RT4?Df~TNX^mA5JW?14ngZ-es{(FBt=n>^02e14TI$aygk{f|D
z)<1^338HpjF!Nh!B0HNR!%PJgv2~gh5T(wAF6^h4qSfWLKdLzGM0uF|WlI~GY)-g{
z&jvR6cNkVYd!Ozd+WZli2=p;blBuaU{;|&y5fO25=vaQ-2(V2x2M?WpHS^NxZlFL8
zF2_CQkxnpr7I2SOF(H<6g^;d%e7rA9f|k)#_X33z_{*Liv3^xLdLJ_WC9#&YizpVb
zk@5ml2y@_jPT=ZTTZ^r%bS$W*oQARc$sgq26C2>s@bl08<V&Y8yaA4f)Z5bwk&(!h
z$+yI_$j`}1Ih22@bwI+q_hCi+#@ewzd!UEg`5H8Pz<N3zRN&$mMx7s}U0)X#j+sPy
zg$X}=qf^L`Oyq|tFum2me1^N|#Q3kU<R^NCJ3b#$P0Y^^K)<=VD#T;kE}=PL+Z@2>
z873q6oyWOi0g7WB7hmrtS||)l*VM3+2{EPHF~dO8sCkfM>d8qNn6BsM{6a$b!Bl0s
zo&}QD!G?`i)nMd3iTmQ!X_9dQqtOPg;Ns`3hUvQQ-AVhILyyMv<mAMgH^KI)a43pt
z4}pi2^*#uLuYiTc=QsULc@nTMX03uxvr<c_3F0GLsI**W=HxMZyhow0<U7-ZCvP8G
z9;s+8C&1hmCCYo*57v82J?o?8m#0lJ1sP~~)ZDzhFx$~4tkVt2W?|X-ynhIXtg#V2
z%OY8MX6|_%_?@yADKO;dAJH9OizvZ->Yb&~e47axy4S!>CqKt-C3L?bK&!9mM9e+)
z`<TX2_P&^mX5B?opb3}-d)nmnN9yWkB*-)<XS(SX7DY^lfP)5LESnw|Be1{qEZFZ+
zYgeIDVb{d@2-Fsw)j}=ol<g}n;)aHtC$cb=1ujq#7pg|5IjFuct+2aD*1K`E(>7X?
z2~;g*(5|(gdHu}POQ7^%I@+)+Y*jq4HzP$F2Cb8Sxzj#-@!}CJ?Wf({{na72vu%m@
zvbsFaVg=8!a6i_Kj`jD;DfWj%tCHT0lp&$Q{n^7BLkb-SJDcAZA0U#9v7a{H>lCV~
z4(%A&Zweg1f6=dZ_wL~CjAlz+LfVi1y=6dCc+^X{UTCezy8-HwUXTS&Cq7_O=mDLg
z=aaiQC5QbHWJJF`KnITNA_`aMr*a)nO7>cC&A~gyX}lp9hmFOitXt-GHcRH<aOsp#
z9T?aFgOn2obBt<b@E_Dubzp->L3jmw^cNHJFL_nEsG9n^+`P#Oy06?XE-(~9W!S5O
zA%XU~cX3w;h~u4B2Ps!n5t#Ckp9e<1_le1DqIvOXF$vrOhMH@V*YM3_iy8m$`!KBl
z4hKaWn|%o~7e!^|=3^~={l&uClND2$L?Ji9^VM&cFQlalU<Nhi7Xbd`4ucQ&X<=bu
zhmNj#ddnZK`tk70(`s-TtEs9QYU36gM->MI@CpEK|J_3Si06Cs^nEQgHL^RO(+V%`
zfm)_h*aMTzV6{WDM3DW+@XvR;5j!rsv}&GocUM+cq$E>)u!n}rIvxnsiA?Ey3|{e?
z#)<2=m<yKy*E;tl`2Qq7G&;!1Z6Kb*%@8JIOCNe`)Ly*;R^|qyhEY#>VFR#KFsU2E
z!?%<RbKVthLi;cV1umw&)l6;}{;<g_nTIqZSw*U3a8f@MyMagJ40$%dmrh)$g|;J(
ztE<01Vt03Ayh;W+#R_>04B85rL`Fv9pi=1RDO>5O9PCLhIzRsvGqpV*Ws#8~v&;p=
zHFIXfLt@#9oxN`|f`IByPWD3#`!IlAh9mB>Megz23N<3~v7^H~^{TtyZ1dN_b_Rp7
zwYlIZ>>uLtpI&{1N`>IDw2=aT(fnrl*3^_S4B?XST4zN?5pa5(pX|4_x1%7$d_2Hr
zW6ckA*H|^7BS6K{kKpzSB>HmxN1y9*J9rh!S0L$mT`|3at0mJ&`Y+J%nVCDsst%hT
zT?dxp(9-wn3fm`Kz<y~mNd$N#U5TdcF-u47gl;4W1{BmqVf5wE`x{)Sj3Xm*U<>P(
z;B)NvDZWd3z0p&g+*6N&WIYrBBluymjv(CS4;B`O&}8eqqCFlGDFG)ToB<6prLplU
z*b~e~%htDNe{I?m!8jXI$-+kat)Svo?NmMr>=%C6`3c?a%rbL3q=~{kqS)%zzBR}t
z{j=%d|86nJ$`N<oQaT4h20$Y08FCEu^}G4^B8>C90fcs0tj92=^7hw?><c_39+KsY
ziBkFXgAWze0c9K7Gq3}0@3uMa`w`UopqWa(c=07cV+Kb1pzTwL5<ZP^t$0730#`-^
z>u~^woeIKpg(cNJ=z2eTT%%ce)zI(_Zd0IZ1Tt|IY*?=!&je*J_rXAKUCm((?<X(x
zjx4;nrq@}-{jjGvQP{*=>M#*^LeR+&^poVyXH>UK6Tz}T&uH-ED>(6)#WMxsOxWKo
zI&4fYE8Fi$BVg~6Z34yp7?N23SM!lMdpNH-y4u<`-Kr`}JkMqum;3%UJKQalUow>C
zeM&9TJffp_Jkcs4gZ*fr5M(VnyIwGXL}m-_M3_3l%#z<HhlfcCR7BLAz(5&l0eWV=
zht`;-EBLy>8aMIVfrwxA6&20}lF;%%5N;aCIgn68(~>Fk((o;`Ey)xZ_cOLv+S&WF
zZf)NZ7U>dZ+l4SiwsKgOTs3fs0N7~n^T->6ehCQD#3V5)g2S?D>-X<Fxyr8;B%U+M
zGc!MVDv%9sibZthhN6IgW@M)&((5zdRLzAF6~i6~9Er>KBgI0m=itmVF+nQCnEVD9
z?5=12oLET59L!$}KM3#(30G)4`}3;4-ptQi71R~vJqe%)`Ma~itth;LL~8PSaOc7q
zNKL^Qn~PEV>*bN66k<&*n(=h&K>+zbZ9)?Q9o}9OK{=Xg|D%&reJ-xUev!RfK-MSA
zEKY^=OO5TNmjNOB@gq<C9V+fzmF{0LK!6;qcXF9~%kR=V8&3eq|3bR6_2)yZe_x+f
zUHx2%Q)i5O{C7Aes-(HvP~56_a4>V4o)z3A0^K&~EJyZ6K`nQP<6kameVc3I<=&fw
z=5v{gteZxDKPvzl*~^*?tyMkKmoK##hOG}bib37rGPR6<+z}E&!<a9=ennpl4Qd<=
zEPPa}LVyeh*7vS>!=Jw_u#l8nB{a;y_JP;j*M9HF;(->%WPd;SiI0F@K0m36<_T{&
z`%^40m{>|lEw6&45MUVk^gt;Y%RiNxy||Lk?WLNa{nyw~ITP&0^e<o7LB&WGH+MO!
zV7}us{Gw%+!K~IQvkmM+iX{~lVO#~)7`~{8j`qaX7K82<pWx9eK(AeUlo}AHMnts&
zsDkj`PireoWzGAEC&D8+02)CU2m%)KWd8u999UiFHN53~Xwb3-h}c9E%2vQj0L*(i
zr*BgeCinpOc&O}lo3f4R=?nuv(G!wW#7H5Sgf<q)9gvy31gkh9ahLv9glqDD6%;T>
zw%*u*0m#&wfPY~=wsg|=;3jNOO1W>7&M0&^3+_9Ee|Y9=q?7kErgwK&iHXAiB&dOl
zI+jfbM|&IUfwD4MQ4v+O>h`p`1XlklZj_-@mbpfGXD0yF(zvK><0@Byxl5vN#b3WB
zJ{}(BgL<If{}Xp;$q+Ik4y(|`kvH%TFe?q#KVsfF;7XsjyX>S#|MtHLK7AZaKfO+l
z*MRGS%e#;q&<7ZUjYEC*d}B3^>Iw3mlNVC?4~<~L(g0@V;qD_GUoeuyvo$@@oQQ0(
zJUyTUonvFe5ZNd2FW6kF%E}fNk`-Hiu}KmvH{=!`PYsj*xjkF^a~0ZZR4b&#n=tL^
z<>_x4E?+y)M^ZY2E|$R$3mjJet~|D?V*l343_6Sjf8_%m_5wzdVW5{tnN>F&W&>np
zMTwD}CA1<a=+i9xrJevHz&{5!BlXj4{FxPRoQ3$22p-(%lbB8mF2KB(EQ&!Sq!}e!
zs^bM_mOJ{?TolV~F=G=)aCT!7%d{K)OYN40Wz$sd0rB9-HN#xPR#Tgv{-`ty<RF<i
zTYX3$Fek|KKKb@}at@KjrK^$s5jwHmt_eGGMxOdAv${RELC6kV6A}>Hjhw2CDO^|}
z*zHv?Gd2=G*h*!qXXF3)4Yuay71%|vKfAiVijS!QTbfqq+IcfV;A7&#Oqht981@S5
z@|}NoMNN8_!DOe67nm-=)CpGx7@A$I)sTNeUVxAi5%Gd3vFFJuaN9R-1jH>O;5!MK
zj3l#6!E9EVlXJs+F<a!X%-=UMvJ6nPMi=-)7;1e|qs>Lo_QVvFjPzHFFw#(Wys$A<
zZv=!pi>;so$Wyv`zzxGqMfcf6YiMXhw2wCYJM=)IeFFmk$CI(K%6Q=wUtTgIg=zcl
zRdWL0r!XNFq{?6mK8%T)4G<S0wN_Q6vp8b`7fr02!Ch`?KDp8U`7w_7W=0|mu|_Cj
zqPMy2M}ckz$L($Nl4P7-RzAso8Ty&58tx*@$f`Vf@)$EFJzWv5Gm=+vQ&(rW7l~Aa
zZnn13dtgz$o#nn)dnhI+PFqqEaJec39Mc+1P<Why3M0VZs}+|;TzW?e>6}ngqcIsj
z(P8rcZY2fkHTVxua;t$nZJ`C=U8spk`dHFKVpXAWYg`U4M*w9X)aW{kA~0AfyPVT-
zOp!rAh-|)V3>xvnLOSfo1lxIFX2YZ#ngf_s04N1FVcaPB`^aC~y4KdA{xAJP<?>)o
zg6hkAsg7qvGSV?RM{utcM%U)lyYXtm*6%;hP4&xlbo}f#U0KUl0E+iNN-YD+S#z_k
z&g5tMdM!7Xa=<ypcD-Fd#urXJy#?uOYN`t69?_h^F%cG3PQoUtAt~tq7j+r&f^5NW
z)4B;g4+u&yZx171$!=e)zqqB=db~$#+<giCJ9ue&69H3GQKU`<cBLH1zNCMltZaN@
zeVwRRHuFOz8YY6BJ$G~5$ra$)z6;W-45}@pdGN!<#(F+V!-1={3dI`5fHF5Crj@3V
zdCcuPHZ;J=$yI(ZUzd@o@@IrPi0I4yfQlnrG1HM`&Z`$74H0*FGPEBoSlR4w21Hag
z7Z?-l{;>JO6r7S0B~%5&54Yf!t(BD#4h{;#P1$qvF;pp-T$v^R)wvN9AO9Ir7Vs@&
zBsEWJnuU2;AS1&hHL`&VssLdVYLKFWQ_+uiPJAy_+x+1!F^9~&yjUQVfPV%Cw{I}$
z(Ica&51UY2u6x{TAjD)7x;(pEdb|G|2-B=Dcb<Q8KC57%OsfQi*=h_7=yEFD;K>3;
z0=!c~&e3mmZcDIel(gi@klkoVEikh_ETr?Ac1i^m0Zje|T5zeE8}PC3-*4m&HCh?P
zMJn4&{qQ$7biEwUjiXn80`_0E`$+Tu*cqYCd0UiPuO^%-4e<W&%)=cf9QEh5*iZ*z
zlA*ySlDU1=*F&_ifC;RyMX5UXiUNpgpfyHw2f7K&k?`vZK`x3pE=T0~z3>FH5nRX>
zRcB(hv9%|8xlA(z_ChSuy_HU}5f<mJCcQfDI+ytuTW}9d2^Z*N5>?(fuQbjtnRDov
z%~NSYu0A%iJ{oQMnmjQL3_G7(*WlxApxMB257+;s$SuJgCe|oGY@8my^Tj$sZokYv
zTp(xT20pBd#gn28K#7~f(`+zzR<sG_kw~}J@sN<}5s8QOvs&957o!o0nw(B$+ra4*
zGTv8|+t!&7!J?Y1Ss}yd(vYK(X5{De?eDaz6@p>cF3=DvMTpM2pLjVQLwU@0u(NtO
zJL?{h+AK`O1-75w>5LTI=N^(P`&s@pD?Mf--<p=;DioSbUB_#U=H>`cY(6UJ>P~Y_
zjwNu*P4;ticC&~}cls4q1jD%LV!)D?*f<5%qLaoeVVQ~A@tBL1LM<#kUB#}LUH7m}
z?n-U@37CT8rKLSQSeJk#$zl{?3#<hP#Yo+xf&|UV$^djm8?vj@OVI&d3*TGn4{ViT
zMg_)nyu{9c`vfoE<!a&P?H$3>wY!00b(a^Um~z(Eso6uoV~Wiq@1@&RR8_45i!d;{
zUZjidi@+u@G12v7RnHb0h{9p%L`d?<m^OhS)Vlg|e?MCWZjy`S?Jh4RWpM`RX=-Y0
zV2}vrWN=@W6a&Ltdo*JUY~lH7UB&I;_BFb5dY;U<%xj{rfT8yFr4knY1B?zFvLKS(
zltxo9OM>-J=zMPhbZeNf+#H#Agy0yj>H_!H?#_y=5-@mcoIURqpVK)5E(E{EQeCdf
zC|oI6d*6-t-May}u&fz;5yPjwEs@@}5jWlDNmAY;^UpJF`wTC%KnDs+bY{5V%PjN*
zw`0k{RQ*blv!&HfR4Jn^x0B!ddxmb13}H+J%tj?8ITKnr(PV+7xfK=DFupWHz}Z$=
z+0#D=q}LnJ<#K)x7x6+4iX+`?oN3bI_`zq>8vJ5zDg>WYQ&zG6!(HT|InMq1^%_xV
z@0Kgvb$G+;E4Z3B$WDKDs1({Cf4veTG^r@Zx>k4z)9J$F5^%o}M%OS48rpBQ+A(=$
z<+ic2S9*Gw``&|tXpbH-$0mTBg?bf2`(Qoy$$1cT0no0^&E=Ju*-$*e5g+I$jRVS_
z6J$pqYryLd7!CECv(L3x1*xe#d{xwfwE>?tH<P698ycM`KBe4&RvPAE;ARA*wy`Yl
zzuc#lcl^xbD$z|0(!9rpVS=UY$XZMm!u`zW=J^B6n`~@3Ja#K+la71PyFllePML6T
zuggs{iKnjx?!a>?bW46RT=KhUF-~|^+vw3G_c6x1?2p127+qT|DH-0o3Ew~?yRRWS
z?jkJg0!>PiR%E^@#nl-&B|BXp5Qlpy=Y37RIsn&;!2KgIi!SOtmq*QSoyA)k|C8uM
z#nHuar1kS)JXOH*b{q5IWQF}UM4I`TfP?R@Pk)-lAgAWn9}$`kBOFg12^u(RW?{N+
zCeKXrI?KsCq>9~zm~QsvLLd%=T*nSuKfh@xhmpIJbYbD|x+8E?8W+~U1ltNyS$!Wu
z2gXVD^3Dsb%Su)#64ut+h-#6)0-Mo0D!Yy1vXX$ziqlmu)`-VDJ7p}@Pdw2|ZIC!D
zM(2r?Y%q3`v=Ii(;TvX&=RNpv6S3I4+Gh}0?8z#yG2wF1+~f~t#fc*)xT&d=thTIR
z8*cj9WAZd~{I%gs&|O?i_P00O(c0dA2RqOW&URq!FQyb?H8f0&9>R3!_X=%}mBAr5
z7bmbZX3>3^zCkPbiMd}J?7Mg29xffE?#A6xLK=cf%b3q=lQRuBgAo#v2|h=?x(C#*
z<)I)O(E&DT-oc0|e*Xj4vu(_KnqM0{n{@i_8O}Y+2Nj^NFIQInFXjK&mI8<vfR0Sk
z0U4L{AzXIo$oA`vjg3FZ;`jA4*HX$%t(27lkLteHpTKw}OEtMS+BzR_3c$lb4XUWR
z#g4pXWNPa4UfMbI3E<n*s>IhdjCpeAINOl7N#&RgJi%1cU~|)F46=gbl@n$M2OEfi
zA6u7&HM>`g+0=#(1Q*ZP7flY!<nQ3$x0S%RA+|wj8*VPbnpH)#4gb>Ax4tL##f!T4
z3#I%Gd~^+*P$pWjbh~tK>1W$oqv~;uSL?4le^$$X*ff7f=UGEr_@O5Kac$|(@|Y00
zJ}>loU5t#}`vZo@<n}l}JhqabOuy&G!YULGWP8Y7KQmLAtCAE*VjLE)KG9W`pKtnr
zt)nCVPIsTOg4L3Jgje0=s8X&vj=hCBbxe#B5mBOU$Ww3auxf_;{HWLs9wL-fUz~d=
zs&1ffzzwJ+W#4SqD`l8=-|dgnt27*}?-Go$+u2Gl+kaTfR**3t`y@J*-6JU?VyUE<
z*CM;v5qG#n{sA#9e6?q+z1p5tRC(}6Gj(Xkf}#o+7--8)<G8dY!i-Ez!XpR4m5DpH
zYi{rc7>70i5)Vn@h=>58KkT^ntftfqcYifdB+l*5{m!&@q~}MS61ESigYm>i^phi%
z1G&}J+l{}%u`;A<TrQkh>Lz@@6~7#*=+Wq|yWwy#_p!z6$B(r?#rZh)E1Y<2{Cuuw
zzV7ZR;BgaWq`L_(xs^<=rIlB#?*VJJx9ika<1Lvc9{d<CTbZ6s3(1~u^oeo5(1?!@
z``*^?e3($|(MXSynA7o)wFnuJ?Q!~e_th}4ui4{}u3;}Pe}42xS*^^zq;TUlnHB}J
zOqJaxCK(wOW4=4a6d!3j9eInc?a}*j{jQznAkJm8RVtK3q7F~Vo2DkW=jSk_<L!HG
z5KMo!lAJk&FgzgOB7yJu`VhaHCON(Fk3-+l?^eQfm!yy4g<AKP@AhTHRDG`zo}x;L
zQB>kIHaelBBSfH7dmygD{v9kl0JiBwCdo`%&uCo*4l@o~nA`Eidpw#67|9~&BzsqX
zjqlyNVp73D*tsD|3W}Xf$?r9032Jtu0*xM5NIemJ@`{RFb_E?}rW0f5>J^^e3%Ak2
zj>#v<Hr7WUiHT88#~+K48gEU-#XWW)-wGR)Z0VMMNguB}{PUrA*sm^#NOJxbW}UbG
zzMpv+g%8$V2^x}zhTd?u7k%2e<bfc&HQfcl5ic;haya?4iccQ@W2Q6v^XGI)LX*l;
z9S`VpV|7ZsAcX8gn2(NY)=##z6crzQXJcTHd7|{XcI@}$Wj@XYacr#XX!$7z2MdVU
z_XXNfaqNw_@b3D*H93=i{d%&UA@arQzn?PfW@2(%B9Kk;8ST3Le6GqnQFmcso@<YA
zJ?e=#mb`JW7Ae`)@7$r^;HycKxIe$xhim`b40$xr2}T&z79yflUIvap$dL07%LD{y
z2Hn>WOE`Cnv~oBsh>h;~MWgYa>V+gj0&Am4w;QV6q2f$!YAWwW`N~R=fTTIDJ(?l4
z$>U#eIhG`bl%x(V3#%vc)xHXK8}zicmXC1OdpYyj=Aqf`->Da-P61__zT3zCjQgc#
z;-CduDG8smx`r0zmH{=HU}6>MU>&WWnY_Ggw#c8Tg-Bvg)RJ&n%;Zb|3h!Ia*Nq)q
zHXsy!za;x1+c02iLx_b%O?i-<df}4ck>!SZNTMeGwP3vluiv`H#u6d%_V(!w9#>3-
zd-+>j8o%~QJjA$Z{D&<Qk0uc{MRa#@R9x)Nt))&f{+5Rw>rMWhhvo+NTu!XX;Ss%|
zo}RtW!`nGH&wpBP6T&I*M4Vp1Md=aE6d>Rx?%Wu!6*`-0)<K)a3&*A0Ln%qBfZJEA
z`56Uk(^uX83_skMyu9*Jf_+)kiXE=!qM2=uK{YeEiA_<tmAqvCwW7kcpSCemrkGE|
zZT`h@gO_dk1Jo43z(7-vo8ve?1y+~%$@m5z3e@a`U#?Dv=A=yf{z!{)4m&+$pPR!d
zDQWOx$M;0wYS_Pj!r!9%&3Db(hR;oWVd20V!y9JYJr$Ay0$P8EcN{q-VcdMF&}ZIB
zyc!l}W(E5F;L1P&Du3*AJN&zY*|~gS%&f0{FiAzNr|UH>jxC2{ISUek?#Mi&rD~iS
zD)cH(uzU07!0%3{b9G~H@6DM*Mm2uk0@R_;Q&aJzqy+GW$6MQZe)qX3=I0F`enIY8
zax?-D?{c1R!?toA|IPs2BO)|y#E&3u?&_J@=XhO-^Ypa$H0wx@|5(*GH0(q&*@#Z{
zz(3bGkmTl9@07kpo{EmWBiIW5q~_mr$CtJ4s#<QQqm9`p*BaUx4u*?<o!61|rhU2)
z4@%#z*pLzxy$GYif)_$#R#fE3S6@yo*icm-yaFrz;TkE7YQL)JF6awb{_)WV>Og~*
z{8O#VCP+5t)!_{dyquf}wz}2n>-7a5&O4OIr)*wlWXylh)_n`@w-+oh`5>t-+aJg*
z6vZh)Dq-KJBk0OkS6+K*B|!z!_<nmZ$q6pAjQq^mJrKAuAkE7yeX}DSJ7l<}t*>QD
zii1Gg1G!5f1l85OcNxE}+Dkb*`^Lu$>*)=A=wF;^3W?)NI$YQH_gfjvw~FIz`-Xl0
z=;TDtWcWQcGI$!}uy+^!n2fIs^gpz0{`l`v%K4TSR!)5vrlzKCo@g1@G{md)JLA#P
zds3ZF6&UU=9;|Z{5}q3mQH+cXyV#F89ZrNOdqec>Ef4$}s~J@hYs&Fqd^sg7O7JNI
z0mdb7tbt|jpz&`_vk<4(Rg>DH8etCGai7w{g34-fBrBW8+AfHnSrxAL{kqExXJhA8
zUOwL3yx8ijh&@h^7ec7*{FKmm@NCmZCnJMZlr}#<?`mFGM_d;``|0iG`T0Od*TKZk
zt-!jyh~VHd6Fu0f_`74GqI2OKwi*G~ApSnKIBs<bHU1_zWV5pykB^W2n2&A2%8Vz^
z=0nvdPDfV*rR>O(ANxC*%Q*oufQ23uCdaTdLcP{!QN4bli+m%}GlnX=Q(o@slPwF@
zX~VvDK37?8+2N6ck_pa~{6JGf!ouV?6*jplW=xX@n3Ki73VmGe;M#NGO&A)y0hz7V
zBt}6>3Z4Jc(ly+}sbC>TM~5$==swtpwGkUF{FS~&$DyAeM3rWL#TC4shnb|-$zK}Z
zd%vt+n??Sj$v4G3nKB6J1V^VglW+cR*H*EC(S`ACAh7R-wta|a9R8e@)!>D`UU_CK
z$%Y{2>eOwzwYb=Yj<Q|Rml<a_VtNa?@0W4+cXmozTHHT6Ui*6>T;Xa@ieO>~4$Fs9
z)(6UMU75jzZ^tf7ZET8zI(OyLFoyE2Rx2Ht$Os=xmiyQh;n;J8*W%(wP|_Q}Kin{e
zZMu=VT)(uG`rxT`XA<_^gp-oeqRZ2|(pL=9Mn-VY`Bh5srO@ZO$z0W0h58`Ra={bu
zghRsQ3(|EZy{(WxLbThXpny)xHJedBt_47oH};T5MR!p9)ytFBnD>LM(k~vJuF$_k
zA?*--0uo?WmMV$RMY@XddA-LcCVpB2?69GX`1tycI62X)*PS}aqaedhx;imXrU_YR
z1(n(APDXky-%m_5jN@`5hic&NA#JXtVmDShE)iYtjg0-T`j1U57Yhu|`(!lv>?R0y
zo|PuFx932p!^kX|&_smf74QRthp@aPfMCI@;{Ql{>!>Qb?p;_B+<?d?1&J*sp_DYz
zA)V6QARrymNT+mnNw<V_BS=ekOG$TsYwP>|{`t-s<BW5jf85W|`@UDqHRGCdUDJlV
za)UhpzP-(jlbM;zb#8SE?mUnFD!Yg08MU9T#tLfdMq6`p*^Xu?u*ou;xH*Z?6K_pT
z{Xin(jakg{;i_y_cE}8#?;F-HoVII`jQ7;QD1X|cAcYUmziWkv<G^}>JNf5N%D+be
zGVEjTva>lNLM+QqBFJq|tvkmN4S^zlzjk}~F1o>Wd!gyB(><CpF_ni92Ulsz0Y(5z
zt97bq0x!1vt@@+;8Qc!GKSu<Y!9ni<XFWY)AdBrjkJj7VuW$J3HeBW8&=WHNRJOAc
z^=oRXJ3ijc*f{o^9k|B$5C{-p<ZJM_ZJMoJt*k6`y(rk;{+%i0;9HYP_&n<#>?#2P
zJ`mKPVMZnwasN~|O$-$HVhAkofGSLWN}u^ABne}ayF07N(k4kM;S?iKwMDGexqAIo
zm>LP`WO!KfbrIr<nuCL=EuabQ)~I!fycP$hf_>?5QeKe-%Rb~oK!Qs<q5@n9dVt0N
zQhKwe#rk&+?I!@&a_vTYVzobH#UMs$w6|9v8pfT(HR&%E51drZ@}<Fg2TO2%VH^kL
zw~|5~b^`<|wuJhn3@u~Ia*T=!b{D6UknU~<z=Bj*m~?%%6VdR}?)~!3J8qs)fJ})3
zJK{Su-%!>onH%62`d1&UH9$^Jb&A#3f6a1jSaZC5d1|N+_|I~)4JxZ01h4TTd$L9M
z0jvsQ!wV<=i``ipp0b(&Q9z{u64qGYEI!BR`4El9r11XdGF-plcte5;b8=3Yn2Us4
zBn*AIZC2=(x#Q~V4Zy|#-N+>8<ovbwC&7tZ&a=O>c00`Q_um*V8tTlFFnkO$<~o|R
z2MHC^DPUg;i;{tGMnCKv+Aje{;r;vPe;@HaUixNoY49V7g`K(Us~m8<k@Xx_s;W@6
zDbwJ)mvMMHfX^A#y*PyWcT67xx>ov=jC$wI<TT5xLKo{4mNdvf;{1cz6<kcSTAN(0
zaB$ErFM5%{`%$qU`(P3hngeda4<D$#VH7f;;$8cL$=ab`RMgdShXYa*KDoRFbaogn
zih|uEZ=Cz_4Lo)io&b~DPC-pArl#h6d9?2H_$d}E^w(Ai&{S({wiQL&lRC5h{=ZRA
zNI?j&Y;^Gj*c{}iOpFTenFfdQI(!04IieyWNW5W<_B4k3+MYJWAe0&kf8j~?{hEO~
zqp$Zl{M2XpM~?IX$LnVw;or&qX@P`<?0n5R@hy+aBmouXB##W*no`ZJ?zcD6o8V^~
zbFHi;t)Ni-+-}Pi9Kv-5z{a(uVE^Y=qmy-S)JwoEo_s$53LVVN&6BUB9UNFwW%gZe
z!-%!<=n>h1=zjjp#l)OOs6S5kI4Um>TwagcUu>~L@P1gm8|HrP7EfoFUsAGv%riDN
z=2l7fKs<y#x;LE}iHw{a;t7gXLbaW(SG&K60*-_W_U7hhK&d)!)ng^m?U7pQc3P;Y
zou;fY+y6|XRNLM0WRqNAj+v&RxERm^J!N309PoXL_{eOQSc($R+J%>BH1+s3&kF(G
z#mSln6pZ!l++y#?7Vy^TW9TME4{qG*tEzZEr2~g2MlrY8?TOXrrGc07xnKP*cWckw
z7s9ajL&)l}?k?f?HWS`-uQ^ccDQVS@MaKKJx%QpzL{iCc3xFdcO!T(NVS-i?m{FkC
zU#~QJ-fJr<Ic$#=0D7l;arzr@f#=a)+j?IrxXVH#rvY;6FFdSQujj#?@AGM+cknfr
z&-ngbTfBE_^xNc)ZY$VYMp8Y=&DLagtH%>QdV0>lVbW)s#({UYpFLQvzvc69B%dqs
zaQ1A#g`X+asOUZ;qo*sYa5p3#0r`iInHjp^Y>O;15%<NxOf@p^LMG5@xzc-ZdB<Pg
zzgLzY0ash`Hu!-tDeZItS9Ps{nlY1lTnCCMLJ5p|bPskvBz_^BcV(5~hfne$nBF%N
z4OX_JLN*{3v8qsdoxQcCgJ69Jz_4keqH_0<4^T|8+2?~RaZkX>C3A%9@?&EtTXJ5<
z5|BL=CZ3BUqsXYJij4TYXS@#|?*AV6w%n(6d69!%L<U541=*)fy|QyToKNtds1Evo
zQszkz2e~+qKhiZ?*I>}bIlC{du{y7)3WaN@&KO`Qm?ln-UEzo6KvxCY>~X8DohIUC
z_n0pzQSD<Hy5%K>w6sNs|5mf-Vjws{feaNDU6qxO_ZO7F5u)dw>h)=GoSz*3l4Z6P
zxQL5`c3tNc0}RVQuRiFzfl(}@f&aUqAN}xc$#G_6C{ve^D66UQYah=Z$zo;UzP5uB
zve$FHrG%ev9Ni&Ui^Xi?KSgrB)Eg}ebYGU^X;RY6+PW!V)ZCv9j}~Yj&U_0n3b0Iq
z57pHZ3JWKUjVu0oCRndtZFhc3Z`01q!|Phj_0j@U1&IX)U;%uY*4D;)eiBen#I(bQ
zE|$$RIeZY|i@?Uh*z<&sr>PRk*Pz>$DjL23*Hen^9pnzbD437xKLcYgent||k;t;8
zY-pI~J2u?9)aC#5+~ryU1g${fG}C?afiDgd@GSSIS0P|EqyDsg{`{%1P&o?Q>9G&1
zo;Of$mSrB>zrIEQ&k^p<Y>!M$u{l>JCeIT7lZaBaRhJE!Mq#A1#x;|nniU%d8id5h
zVFX4Aamrrc^>R_2o$uAud>R|o;ScabfB&A45Z_W#nUqvq;v^cpxvbHGw9YTDM&E%G
z2t*ad?+v39pr_&q>4JNI;s8PmAvGC9bes=}KN1#s7D{a1zPiHsTBZKPB+(LXBv16d
zaZ1L&{q<o4eir3nhK#CUdj0*cP)?Qt!l}AIAh==l6ZlAr>FdAYlc)gnPa^8-=`4uw
zArV1cTyjAKf0Eh9+lVD7@GuoQ>_h2Baw^{ow|{{w8A?J@x7+am@m@nV{1YfJ&88X9
z0%8(tN1z=j1)g*vfK*VF_79fAl>@~?wq$U}+b2jI4cH0^Z@BUWbz|mcX3>UFCj8<8
z1*_}pPYsYSfL!A0^c+5-UaxBber092k$=D7!*0dP=X5}NUfJmqfk6QHk3ihdJsexs
zT#1H0`1cE_03rd(1%C36U@6?wf7biRoFd8%0*1h1X_Y~r^d++mq;xF9tpOHNNSC~P
zg{`BEH&{<pS4R3$!<9~r-9?G-kiMS;AQr~L<|<=r1f3AMK1IPtgR<5YB8Y_4vAr!4
zgD~>75F(rE0y?IkLu>4SBg}NV<76u=BC_5SS#_5}58vcK)&|GR6wYHD2Y!2+EdjP<
zArTkibIs2X6kvEk5#@*LTHx|Erf9)Wqg-rOmS?I=Le%@P_ayaW$AIQ~FNJf0xXcTJ
z0^%I7br%;{wsko<+ThIP<{|-Q7pbbQp&>1;Q&_~6g%e1B%IyTei{`s0PUWwG0|b1w
z+L}t4bo4(Pb;ocCLgJqK)t<oG_iK{<(Kx?-!5SD5vn@r*XwdS3E*A)4e}wT+Mgs`c
z1uj1AJYqBLqj}Q`xpRB7gUJs@aHXLl!fT+`ZG20@sXR3)DGvA<*PeUa@T2<_8uet+
zedD7=>Su3~`j|ubX)n_X`3z|Qt~N27l!~%)OL6f-FlaAGuOBe#U>&<fs*r`_$zUK<
zp<t~>fc**khmlg<g9xboTBPZ(*z<@@0XJQP3xQR@o@S|SvTP9>>evK>t$`i~plM2c
zv_gkrB_2Z!JEIeP&0Bq5@Y271MOUJuHtkzUQoph%K%BbKa(jjk-fy?z9Y*Bb*`z7Q
zA0N#e81fn}HA@2RZ9Y(NNkd2tL(ArzBypA7J?SabBdW5}AK)*+!P_%sz%3#fHDMBm
z;|WUwdod9*R_}~3v<^>orKS5KNZ+PwQlr5rA`*`Z6jD_p>z&>hUfNRa)#Si!42A1~
z!Tn*A2M<=PfPL4b&%(l%`Q*awnrEBOT*cn?d8!Jo@i)#Ql?LFLd;gt*n%X!}%TJ%8
z;Xp=k6`(L)BHa)8@DTwsZn<!CVwN>NJ^%tma6sYO4Qo%JU6#*z<iiLxde-Y)n4)qs
z;ORz(I=-YY6&7eKZzyp>2@3jZh@TQE5661MzC(1HUMVQDv>!16&&j1c<Fm|L=T90x
zf#k>2JKX%?^#OQzu2dbkHvk`ySS18SGL8dypP9UYrBlwJ|HM0?j&BY3v4&89DD)@j
zwe>|Z0vF(Nk87ibjSX(`dSNFgQsl3M-7&N(jsdL<R9f0#sZE9vpBC6IFR$swxrBk=
z#X6o=i@ct2X^Y|kSQ9SpmvRQutk;e(P|C&qrUD1{$G6mS46dEA=RGZiM(3AB{;sG~
zL`Dg*fd2$FX4Y#F7+_-ow=nVo!fRfR`hcVI>(>=n!KSMMfWjal33=Yv!XE-kqtOur
zmR+2ufu;D);NDg-?&}oS?_P#32k|;?L+Q}H4k$=U^z>{}y^MvgYt)uNJBZU~W+<~H
z0!zO;fm^>O07yoKI<8$Fdyiy)xpEUJ50|aa%900EjPO|}!T=mzl>7VhE7JOvmA2EJ
z{mAC0&>A41CIgj!HYyf>|F!|JWj(zB+{B1^L!P6>g@$@~@Dlq8+IO}4WQ#n1hBLg}
z=7{z*yXAVY?*2UcD=y#&)Y2J(dma=6KEC9lu*moasD7Kl&5$}#pA$F2pd(!n%RT=A
zDr*nd7tmjzt2WOSRLA<kzmNpcX{1hJT4($B0ElkJ#)-lE**<>%@go(4=wbf;RF~U&
z$M7TX8wI0CRFPwJyRJd47PNmi=K}(waAH?9sIAQ>Spd;`x9J{MYn8{**}Zi-(EHRk
z0biw<6O!f45z)7AhvZ)it90^w_q>Ob95_Yr=heP?MbP9VPyJbpx%u+<Ac%P3Y7~qY
zsJ2FDwtQURBEmGQdgz%dAHXO;OmTZQYtSa{xTl#6PU>F+a7Yr}8y7jk@TyR#4ns@+
z3=mHUn3_J~#>Y6Yoa4auvLsd4n9V9Qb3glt0p9^q7zJokVqtZUx(;SA;ghoqC9uzK
z7%HLo;nOK+(2dXM8pp5+0YzL%Z#uF@Z8-kvm=6Zg-_0l8W1fuQszE8K;-hPBrjz9t
zd;k={=dfL14Ff3?o1ZbW_4V@q1dx%dK}4Vp<KIU-Jgg?m?r%I$$Ide1Rc3O;jd?z@
z4}Fr5*amee=IOX_j}GUIL_oY3-$a(Ao@}QLMA1MIQS`!S=}Unoa=NS(>j}TUb6-`_
zQL~*q6akx^zbhau%?H@%717|T_wG=16WCN>TG0c(8Gm{GVkl`JS4n9A_)Ao>yd^~)
zBlf`0Im<}Z!r~<phELt~>>@Hyq8N~+WaH6Q8dPvRIz`+(nE(h83#+?}s>S1s#J~nf
zbkxsWS$5Gb46TuqrJ3qaj}R8bb0OGR62OUenj;XmW?`$L)7mb#Nd5%9CV2$Bgfo(V
z&_5%ew<tWow`u{7*5>ArCyNLV6!>3&mBG%=l^j*}_w9Akc##&k@M~qNN7nIya_pD{
z^ztj3n$pL4I)W<zl}WVa;%R!1!1LX`VQ3^fT$t*>uvIDz>v<eV&5(D)A6NpY1}HV8
zCdhj24U37-b9a36eA^k^hDi&Rd1rFe^^`lHgVSMqEA;uZuC5-wB$RRP5BgZg)2BBN
zv$#lLz~}#M@+s(R0>r=|1k7lmsp#22YqVo#{)dyRQ&->T1HKZllXFjSnVT@!e=|4(
z(sJ`YIiAO#dbO>v@VCvXl$+bHcbo^HG{V;QvAn!r4|5WV5qCn0Ca+yC68woI2=LZC
zZ+_^cYiY(X!3@8_T`<jNBFZZ-&l`>qDfRuuA^se&oy%?Y^%HGAR#_hcOY@y0`Vm19
zKLNNF=MMpm?tl?&{f6X<>%`;%WA0X!d`(m!!kh^lg^qxo6-Gv3l`(Fy&;y-&2PHvn
zaFK(_qtVpJC>kHBIX_?KU6<MY9blB>sKyu&{=>sz9tTg;<P&DSeh;G%>$%PXh2Km}
zR|N&Q`gj-J-8D}M_vo}gG9*2Q+hXUITYzy3PR*c=AT^yX;pu2*-`FsL6bUM-fA|ZN
zAk(E@en1c820SCxiWdXI#XR3FQ)fYf5B`2L7!i|jfQ;4Ga!9WgyIrnd<cOpVwDGG^
zBgj^;u;7u3GGi)0EoXnb1OI3YY&k15?72Zox@=w{xDzl177D~g5d>a<iWr~uXZh;m
z<_819q1<BHL8|W`!v)0SG}g%0mQ#rbI5w^4i;K9or$x?C?&npY>spiD#_ez8V&Pz8
zGGiFG``h7RMgj*R`*#C?o-Yi<i7_Wjv9I+2#tj5mp7wx_uAj&CMp3c&v0EM*3$|0a
zip6qg5EUimn}0I_w|0MYz4UW|v9O0LC^su<FNlq|f@(d0wu6$C2xk^-o5Ll5giXc5
z;3Ly+#J4^EZ5Ql(0Q_Be;%j~UJ=Z~IW}UE~i;H|?%B8n4!s9ug>$A&QU_LDWZFzDJ
z$(O(4Ciy2KPMH+N3#c`XsvgVvx9qf|UdCisusLwYGFpLY`>$XDZ5>B{d$Zfs<pV4x
zkUxN$g9wB&8jw3hL`A(}xTeWY{QU6bFPQ9-n`s$n>AAGNTejd*o|BUn=tBSinE1E}
zu(}bgt)P00=CygoSKLA4Ri5_%+BjXxdgnNlj+X%*TRb>EP6nM7G6Z}cLt%{;vo$j_
z>K@JqU`eb%&Zl$U2R<`I{i|0XtbT&J`Q+K0TFad_h;Fr+K2VZv0*-HKKpY~+E?jHn
z<m~Jh5WtKdT3Gmsm9-j>ULKD40N3zlx+_e0&Sy?5xYLHRXR)=mx?(5=Me#jT_p7kK
z?`XNdeLLv-&i>N->(_MvWM6Iy(3Ko4dO~q^pjbfw!)bI=^KG?5K_3Ko;9HLy6o%J7
zvJxLGJC)bhJ9i@U0G+b_#C)=p$MeqZ(|v>aB)y&8agt|;<jY5ozQxnpU?@QAzvqDc
z#KM-uya<D*bb$Y{r%6aV@Wq&n3<hTn=b1%)CuHwRM5{#T7Tn(69>$(M^5N&tpFCrt
z@7@^=X9#rMeew1dR8mqB3RhGdLx;J2`~E#ISxH$LtB5TS|LO`8oSQdqoS*8;sHr8A
z?13x4A`%j^vJ6=Q-k5S054}#FEzEi+s+|m`YEu)u%<eNh+SuI-4i0{jP6lWMZrLU#
zCL|;zE-!i9PeudI-=V;x{tj}b(BD2p=o@o&86fWfGCGV0J6<3Q$7xp#l&j}`^P!Kh
z_hm+4KLl9oj~{Qaz#Oh0Rm^UVWOJEChQq2V#BVN5tZWW-C7;;sD3<{f0LKR}hT3(@
zLH%N6eBLSdm;m!bgwv&!uCqp3BQ%V8P{nsb#PVTyzTp!#@6qo8!aEm0{b@ACI6pJ<
zY~)<m$Y`{7amM_E)BCMfW>Q36`chx1bn$s!*D$msl%)nh1p#<E-o`xtPFDX2;fsjI
zh*v7tr=v-@cYexm>_@iO787#>^ead>($r+-S|fnAi4)^rqGDMA*B%%PfOJ28JOt)2
zOFyvGdh;6uoCf3m^36eptR(^VTK1fPhK4+{gIiM0I?Z=o61J570N`6{1Aslah9L;d
z8Rq$4V*?O?pbSW&)ETI^h>f*s^0<Y0?ehU`@C45r<Q1bB-v~2fB0)W_OWfr!&X9%d
zyIOq=3qkkXU8+3aN1JVpjWeK@12o0bfP(6T1NUV7C_WIi35kfjOhUm(f$Z!&_kk>S
zN&rt<)hPnsdRdT5WpXlWwGHG`DfBKky->MBy56vP(2!?)-ntlS5E51+c^?2mNYv*)
zvNZobba3O`25^0V?tgZT=)d#0>{3y?R2CGZL12&rK1A)uLz1t#YkIpMDF0z+<?~o@
zx0iy5^XS95<)YW=5(v|PYx4qN+nHWotxq6U`}UyH;61oGPdWDNpw9;h$==lTc0!rf
zA!IW+8bU}o+cVbM_0ou+8{{+L;qk>`?((uTa$`j1sD^Bs8)yNiYYF?IZzHEM><I-K
zRp$%q2WG(MjEA5AJCgI<aSn9!fu}D4Rs(EHJiQ+aYc=V?Bc#h6294i2S^6)VuT?V>
z0|OzwAub?K0YcX?)Dc`RC>8?3LT}$_RMb!3)iwx}3G`3XX`Z9uX~OctL+a5n{G$q*
zy3V1ZZsyZ%CPP4dZBGI!+&3Nrf&(j}p)xsGd6&u0gO3?mmiwF0z25m#9z7ZOZelE8
z&8MU*0aU!>zuJW@1j@L3jO}`s*f{X>o|4t;|7nB12v;F}1?sam8i2`0|FkFsA`uYh
zP#>G(oFb&BYJGG5761^rq8)P^%|Y}5!whb*$G<>=I0Wv@?+FMhE1&+ERt9mJp^3`j
zR(Ky^pCy0P1=56|&WmpOF;0A|=sR3`RcleDarHl8I)VV-2YIQ}VzoGM6BR@d^hn$n
z%d1!YGc(KY-@D|=@X*m&Y#yf~A~w!8&znp!YCU~A1d3b^e_%DVM%RF5%Ud8fu3yzk
zoYec-X}F*ESvne7z2XtTZa|g|XAA?1{-U(OCc7bCg@B<b1;GBHp%R42odd}{fp^as
zYN(h6f>YBdO}#)0)8Bst1e5{jNH}<t*)<2clRH4GiwXYv^-7LRI9?yN_PDt@ZF%Xy
z?9^-$r15~2blg81p9m|jmIh(m@tF+OJplbRTVQoo=`Hp3tKeiCUX|NzxZ}m+1kFQJ
zqrA#WEqOoDgIl8eCtLiOi?s*?!WS?KhI?ic1o;=j@rAID0lkKRt9o{}1yH&9c4~tZ
zOy4u;(SBJAFUceZAt<%cos~9&G*?wEgG79%e>_7#(4SYEAWd+lD<?My&WB-Z=oNq#
z&Pn)}>Og@3F@RJ-Sr9nJAZaeIA%3O{(qP`+dSYVq!9F0b#s3Q-JyEUNBDuDaTUwSu
zBIh>(c-b+4{x0vX8-T=SFah8YnS$*ANHnO00db#B(^WA*R;pvEp#+32H&^Bwg_cWx
z;9UnK2t!;TrUa0nr1+LrBR<p857Ep-04@U`=?jrKys8#}2Tm9y2bw^3MZ<YPhmn0y
zd<7h-d^$g#gc!vMQzMVh;8hR2HUI^W_1T}6!Sbn!jy(5i$FVpG7{wzPII%C%?Qsx6
z+WR5GBC)+okyl#?h%!OAg^KJ%)$brNB89l!=IxzCqhLu$*aBKJ{{3RQ-@tEfr-(Y>
zg$H3<0WL;llX!6!?6$~Qr8hu`s?9N}kwD=C0%Eg@Z@@;0#s($G+#c$SX=~G&y$&-p
znWUz9PDm)44G<)5WQ>dELCQj4m}BB5sASxZC_?(5$!pk<zNt`@AL(f)duH*_;vozX
z*^STC(s*XJ1*F)F0b2gZ2hU9>1Ulfm4{k~N_pbw|_u`Mo$%DUh8R7ysGbqjB`k1hH
zw#!ROT6dEK%c^nny^#mj=Q&8OHXMSCR?P!2#<cy2%M4I|6C6A~RYB=_drHff4(>W2
z1K-_z;yVDkYOGQpL$l~w!9wB@0aRmdfL6KK+7VPOz>hry1E@Y{=NaJnN&UX3PyU~O
z$|CFuATi)^hT>=D6>Q<*Ck()~o1S1^?cTyJmuZxml$1^0y*Fg!*^<EL)+Yc&cx<p5
z4+?pH_CCKkn_UIXJ7Kcjq88U^1?8X|$J={3l02N(xsvKX1J}L)EBV1(6kHwQwM^;#
zDoKih@-@GN1oR)h{GH4N{LqtfGwT2RV~+>oGW0Fhq)q^ne))2B^+!6`8TTVPg|V7i
z9Gv0vz0((*QN||mW5@Ch2P88HU|*c%Nm#2_u1>|2lt6y%7WlF-?w*g9##uG)0Jb#-
z5dcss0G*J?tMxcL`=5uBdjLjDtF_JHbTGhbuGzc(X}?fjL%XLZ|7IOehbaOOKUgos
zGkyNgyhA7r=~s7CWpdeqgGE-mpTFR_+Fb7F?-oJ)&TL`vQFf@xs4oSS$C#Rn9xy?C
z|N9gkgU;{mofeXkyY&0UI;V=;^*+6}HqQ3xZNZq{-U8CEBv<6Iq^~Fh|8%6b`h0i-
zYI&Br@@qQXJ3r@KViRIf(W@Jt|9FuJD$7n!(RvF+fNVa;?N2o)r!uiVS{tpkD!Q_>
z`tPLPl+fAe0|)_Z@}kEz1Jz)HUIoxgU!?UVB%G~=M)BOIKFa9p7l87G%VP^5Iq1Zw
z$WQuK`^r%MBRP5e7Ku`9Bfg+oo}uLeUQ_|<@j%!BE<|&3<u$q}b?cUQp5TGh;bL9%
z$??zvfRxtGuXbyH7{2Y->FfB<6be){-Zrr^GA8$@1p=N5_WSF_GLYS`FbD`w(;fra
zUyv!kK9f(v$S*Q~8LfHL4=#uRd?hBD8Qk-lUAP1MovEK;m@r|D&)~?*Ue46iRQ7xv
z8F_Pgveg><TB|#p{T6W0mlnI34-iZODM?7|;`K<0O#=FEmTYaYsJ@;P^t2zO-@3V6
z0%2Wbp0IHxpkVa;EtL@<YNoASf|2&<zw(J--ltp6#PkZ3q_MDs^15;8j=#cVvLJe^
zXfm8Z7)_N7qI`gX+ydmK=VoMrZDf{<hn9AKyr|L}!3B8w;ro?QG$u+)b}};3z&E+H
z(OE=$UF6Gmj|OU8tbpNo&P)qBSP-&Fo$r)cu(ICvARgS^q8-Kf-uF`eXArPzi@b^o
zI-Vvb7M96g8Z{O6;C$;v-g{uP04~*X_0#s`LpI10gCP9c%JVwpMv@*oK63iyxuk?d
zxMs&K8Sky<{XW6D=SF`8C7(>LD+J2h=yLiz@iua6Bbj|a=;gc2n*WN)SdHwJZe$(l
zr|uuOWbwVMa}00Z*8X>~!)GLz{wA3a5J8(Kh&%)&;1Q8QcP%Kw|I;s(Vjic}?x!j$
zDlsY;jqcZHpm($|F`_RZ-2my~Z4%^@(^Fz1A~}daL~Lxe=_uuxKn}OlpQ$$lU;K|y
zy*D>EFON5}(GdheY9b@!fdd9&o^1NW$cO+jBH+2714IUL5Ku9Q2<j=3|5nic=hwpJ
z<wiQFmMKH@_~x)P5ltma`QV|hl@%W-Ivf+=@9FJDLPX3!_tyOpMr`OC`W6`ha1;QV
zj=;WebPNo5G*HWhMrY7Qy6{FsT-<b}GpLaNL4OKo^obdGS`2g_0K5g@=|58yAD(!d
zKLK0@D@Pv>YaaY>|I5GvjLlTCX`0n$u>}P*g(IGVLPB`7s*D~t7doD|)tQ+Ok(SBh
z0MkfP(re}7!0>QU){BtZ+H=s8k5VRu<I|odNLJ(chJ*XHC0Y%RzjDL@smOH0r?ULi
zBnYywr@Fd&r%wR%g$V1FxRKF%1gX&6++1)-2)OKWO)eUiLFs0=*zDfj-Oa?r1c(UU
zv0gkeMG*F@tE(d=CCzE4r>D<HY3Ugp^td?C^Y8#n48Oa-5C&n(V~?zL24NrQ-jiy%
zfLesPzTVzm9|+Xa(o#h2A|xsbi<??548{YNx+M$U`?Ii$yga7EX9!LI-R(7l=J(d7
z{$*=$878FGYB7uUAOpY6_!|fp2m1TvP<h++E3mMz5FR}O%|)xKtK~74qZ1OGTAnN~
zFCUOBPFI<Z0(y<vme{B$m%AIzp9s)kOn6go(1Z~m?Xg6gC;=g%O0nu2z+6|LkQe12
z>W%yY#JDdShIuW#AeTL9Jg)TGjZ=5K@j4!fadG~Ytv)N>K|#G>KHvb!p*~#x^M{s!
zVR2%@|9v3I0f`PtblX!xHfshG6BBIc#N=c)7_cfplivVr^^fUGxmT}W-=EdCEJw)8
zXMg?rH8=P7Qk#$K`QA%BYfVwlo}r=p4ZeH8L<ag6iHp}-&IczQU^3;CU=s7VRMpn9
z*8vWAIcyI-bMx0!RPhN3s>hfqM;)U20E-I_?(qX1*+It#sU#K@R?Iemr3>_ggoF&r
z+qQ)TZFO~*&EZTlGcza8pq7s_RRsphqobqp*bm1Y@9%EG#(f7ZB6=dpA?}9bg-Shi
z5h-eH!4Dqr%l>cW@&D>CNDS}zY#X!hQH?AO?d)xh^eo`Ntn|%LnK_wRNf=4sKe@RX
zq>L<0>_3n&u`<0O`THM(i0MarBU=WMk9zh-!bS#GhDN-+sQ-C|(@c|wWH@;(X7f<l
z=IWl_ciLtYd<2#x`3LDK19L>_`mPMxh-?UJ`+30wpZLbf*N4dqlC_j>RIO|;4lI^B
z7SpsB(wjBE642Ta6yq;=zIgD$`sN9d6#<Xx17ixb{l$QE&m~R#M(u<5DUw0)UHj;K
zr(0^?9hlbLPh^Ia1&A1sv_CI0oN#@4VzqNvzyU%1b&o+~*Z3g^i9U$({liA+)k)uV
zxGZ<@P$K(DEbo<$)xNAJ?`iaiI@9{iLD|0S3BS1Rub_v%dg=+wWsClc>30j!1N|J$
z9W1wPR-W#J^;|mSX=&tXbND5M++O58W{lm;;^89Pu^XD?2=%Gc@kKsruMnubpNz{d
ze<wp7S)UyJw(Q%`&MH6Yw2b{F!i;%gBR+C6425p6Tm12atg%pv?xIwqv34&Ck|@qW
zmMqM370<7z%p&IA$lYTm1=*Wp0L6ahP0;IP9Cg<6trN*%25hhQzG7F{pUw72<Pt+3
z{;o^4UVcjLT-9~nc9I=W_30aTzB<~CFzmCMGop%=scN^_oZ4HXgmdFEdxC^v?DCUr
z-{UB?vNO~TVj=T1F_Yi1a%#Cb(%EZ$yTO!Cujn^v=xCDOqupu`ozYdkQ_P;;>8YtZ
zc91}>dkqL*-WV262Ug04un}3{X^*}dQ6TtwQ&=U){IuE87Kf$ytz|6TyQ%V@##M-@
z23FbXG#(9Tutx=U(P~Zfs`^#4j`ro9eOu3o*l{U>hm|4yvW#e>TzcZaKdC7C;^Mn>
zX)6*j8cv6!(=SDh$+pCmV~!1n^2@8zgpu8BlL;$w$w?(^@1P5!bu%(%k=N%t1wVhX
zyA-~og;up&r9JtGpP6ID^?1fR1E<iyEzN}XGYvD+y<=`ZfpWe>s}=cXqNvIwrs5A2
zi2(8Z`8+I6!3BkvzBXuOsju{&eyhP#N!$z+78~Z%C5<vTD%s^_6{r09Qixf`r<M*D
z+{O3$;kE(UMt=sbfhmfD@o<=>;Pk*lRo6B&^-~B!d77-@``|3%uGg`6+y#jh)Yv@_
zZAXUZF5e2FtQO#McH!A7rpx==REyQe{0@JIIIrOlMLb@<z@7W7hQizEwySy=U3R=D
zTHj0zZ|JaDrjm$}wu^&oq<7+LXkdbD^)qyWH^uwKINH7TT<B4ZjlziUr(+nKVkC^S
zsU>STd`jFnaUSJFNNMm0ti2UivdWU^ElbIGV5@A9BS?M}twdAD|KXjSK}XorketMV
zskQiPoPi3YjB1mKgyuCR$CA2870V{AY)q@yA<>-aRZqC{clEPZ(UYKuSZ%-b6)05)
zUsjO|TapW3QxJ~C3+KTwgK2M4QOh+0g|dhjNC+6jpT++Ke87`o*y2leBWLTVF0CXV
zhZ_kr%AfJXY;B}){K(qw!}Xb*n@y{9n{2$@O8>x<uHx*{<%`E7dqg+Z!{ia0w-!dU
zjwRWTy`kOCf@(6>d^)<J#QT}x_lYm&Fz#Z?BzX|8hUD~6_{W!}`sLw`-Wx7FAsh5p
zkvA1#I~puDg-{2@K<(Fu1*a8N{oN+q?^~2DvLrJR-;~u=tL2xo9Lk&w2=*`Z1hTvM
z&B`;KpyIh0^vYZ7>Pi!H_^9RotSI^*IQeTeVNEJcZ(>+s7CoVD=^a|=c>aTO{CQ)?
z5cV7-8^`ZqQlA@C)H6n&f0$O7Pc!Pt{?w1NfIc~yDy6kMoOgA*S)0;&wL&S$Y=|by
z$-@Opd#^lu*piu-{3{?#aPzl9{5}8G=P;>Y8t>x*a|t$=fkI^!#n-a|R}sPmn^IAL
zO*7`Zo77^-kF`PeiAIJ8G7efyTUxR<0i}8ROJBDV7xXud#e3b>ex@mkQ|hmrZiNpG
zE%Du6&zj-4S7L=&PB)521&^xdAPQ7RnqS^t&3aC#WeWY^|3pqz^|5L)Zd7*14GUk8
zyhWdg2ErwqO-uJQZfg2Og^6bR-JD&uMSJ)U{w<c=M>XMIic`)_ut%Rd*}F#gX5M#)
za6}qD<n|ERSu%U~qd{4Fjf960QYS@U-c6iKuOwID9O0O5dszJAsBqn*kbOq^DdGWk
z;gDK!+s}x>v*hT(H`=W-6cUBY2z&vHCKIv^a#th+>7L6SWfupcL{fvREqa^h{Q*!<
zwEK+lIg-#{i0uSg)!O9!ke}&p6dGb5wuJcBjIp?0Dg@IQP4s5DIL(i0jjL|EOHx?A
zp;K1sqm7NIHohDSlkmVxveA>}wxEkEV3CLvc*~vqG$sd4y2o{7;AkPyGEQZf*UNRd
zFiX-=_nGo|T%}-mL*KRXUU?^>SjO<og)6xL9SW`G1M+-&Lf?9vlY-7HZ{7(Ajc1|6
z?UW`RYI@YX^mlYw^GEbyiO;!_>F9-;BZn)M2$NI>y2a&_n6I2~C5CZ%aK=rO`x1W0
z&%W3E<n4;%>+sR<Sk;aPS#x>hTv7)QQw`SuD%rIsG=3_>^?>_ReUT>3LZZXCQdrf)
z--l{e8K%U{+T6&}km3)o3O0z)V&rX_u>Ct$EL(M{JAXEa=DLm!9GT6$7!41SN_e#-
z^yZ6Hmf31<Zb{`SXF>N5+TV>q+C3jtns~xZ4s5d;zP<>snQqni(suR6Sg?tSeut-i
z%v7@abG@arm1%~<VtaT1MaPH;HdJuC;*G$$LL?8Hcz2TAOe*6Ge}hNHoW=;@UeC=f
z&BSw+1L>sWEu3&=&hH`ndiZQhxCYw?_)ZnPkoaS@x_P5QuF%2@R2T33dbaFr4}@Jx
z1uL!Y{WYQkq(E@_MZ=BqQ@a~}*wX|(DKgk2W_*7+s}G0t2VDpk?quHsn_jfMXuo3=
zA)Gpnu$|)xTa5o06&JQ)^DrD`bqP0mZU`q~!}h&U6CIwil3vBP8*bk8rzvv@k=&hG
zo2qpmp{y*=&DaX-g*KCjx)oh1)-w!^1nMhc#B^d`7Pq=|^=v)oU5LV;rqHbzu8R_U
zWRISjR6sOO__i%3TVbXBICj`FgYqk@ZK+n6c}Mw*`o5Uqw<)t`p9i7Ll|*?w;=~d^
zzQ=x)+$go`6eX5oz2&T3$Q!s2_pFC$SfmY|HYhnb&gVYDW7gQ^yf3TmdFO5@RF&Gp
zX}YWLj!{c2yrSDDqbqHiLwr=hXdYb<>{jn;3Vr7XbN@y`s6mYy;LUjZD(>xv)E5>f
zPUB2BcBz?BXpQ^NVe`^92q8}nGX`+>6N|&n`evc1quQ?t^dvuQy+B==Zgwn};mUUz
z3tx)NGaVR+>UwlZ_svgw$)2~Icy2$Q;SV0CB4%>xK;-cPW@p^+<G|e|a!I`V=Q-r|
zT3kZ2(x>P}IN|lRzIZC8n!g;aU8g(@6iV)eqpx2)w6_udU%81a%>O?(@y#1%j{hS!
zv07EyYLyM6=~`tI>YGR!pu?}*>i8$ab#|h*D{W>^n*e88^4?=nEQVIW_9JHl@&}2v
z1DgvLHzNO}{im!9MY}eqPS(GKqq#+AiCS1g(c2lY34E-`zM+IAubLmP2lYITp$o>~
z48S@}k-I+iSw(!Zka2KDii+oIB9A}B<-Wf=Gt#_Mo&2nAJ<nT;+U$8^iavthyPea?
z)jV9>^XIJL+qX8SHHCN>Dfsb%#~V8mT%B*UEyx!TlapU~vM(y9LwsS0QopEfn)?O1
zb)&wZUar4p^~^lR4%Lzw@GRNdVSLJZ&*G`;ZQj}|_v8=1qX#sBz})sH(i!5i78`{#
zXXlqRBy&NR7qPvDIvq8narmSnq)wjC^pWXyh|8OLA25)1BclpGL1~vj*V^rC5Df_5
z3E(Htq{l1DD4|Av2~}tj#b39qstK~c4P_H~YPYgMA0tx4`y}0K$Q94SPj-k>{chUV
zWt9eDjxdWGqp;7R{&k<sOltmz-ndG0yRDEeOXkYa%JC}^u-uEd3%*ilXLyH{`aMIx
zev2sjVM_Bj%>G)OwGd^a^D}>{D#Kp+>HJk$`9@aIh`X3hoZuXF!RJ%?y~r{31}ly+
zgrG=n>ac;G5AuDb3i#t8%Fc>>!y*u#suczKgsiuPEG4n_${m~1o}C51dz4E8;^c6|
zBNZy9a5eWz-sb2H(gbsvCKu|N&W|xy&m=4y`ip4q|8#9QYs1z`*3o#n@J!mJ%!O5X
zm)H%(ju!2$2X&KWr_a$dQL_{wLW+Ft!rJK$VtmSq66iVdp}^pK^mEkHXFLxI<GZQY
zXzF;S=AQ4?2d-4<P8}Stx++xP+}R$@s&%l9_4LFK9w#ll)#cRGxc&)k7tOK=VGONW
zf>eZCMwe(3IVZeWXZ}v$JG|THGKwO=_sSwRdY#BumqW^gv)>M*%eTjlSDeF-*v)3Q
zu36(xxF3fI>M5Gkw>#Ca?#i-~CgYOpbvuo4gZGqneK&K@l;4a`TwN+<)#rR@MB!_f
zsOr{mef>I6Zta3EBsS(tdeq8B-r!&-tz8Z>UD`|}YuKEj<z8Pg6deUoF$cyvEb}&U
z^2Mv;^lII^A40jDG`17Vnkl}?YrL~O#~V=@oNL&;lSP6b+~)Xt{Y|kb*AWS?9s5+i
zZ+b4t>=l)*JR7nW^hLHwBjH^PM=o%%A*fzOW10_7Ba0$!UfI`8wK0u1<+hC94+w3T
zj!kuvyg6~2iu1o!<ULiLo*xL~Zx>4Vqdv13-qWMUNq3RdA%<(po{f}BdFXN1)5Y7~
zWk1Dtp7ZRb8i#15Jv$_H2jkZT>VaTS1*%t%J?Tq%`YogMi|V{lzds$zm1H*j)HF30
zys?@xUq0<hX9=ds*}38~BhsvPtCH5-r<G?bUdNyf+2+o&NZ%AXZ+(+L5TcPrkFQ&+
zDT`#0-Q(4Fb7k<<NJK^&TUgE<$Zp8AD#|GOp`(+1d|0buST=MYR;rUu6@7MkPj~aI
z{ePvufU5eRFjvvp+K53_R^QCXz@9<DLEj$!|Kb*UCPoY@riLH@W@BVz5Hm70`C$L|
zr=Y35ot%-akd=kCm8FrTJqZVckkv;kTLo)910x1uBS%vMqxYhM{|%p6SeTia|4+i_
zKdBH^1=E@HbC+}1#rf3K5qV74{iZMnBG-NDSFfgA8K}by@V1C^g2ZAek&&sJBxAJ^
z#Yti!M6gJjE_l!yO3lAgP*L^Xw%9t8A`bT6&kfzW?%Aas-SHqJB6#tWVEn)Q@wO1a
z7=eMvU+<6wGciiZosePP0VD_I=JlWU7hZVYWRAVatgCze-@nL-i+jfSkOLV3+9q^>
z0schyZUsNM5dpU3LjwN+_3|NEB0&Z}+JwNL|5yK%$hYRvB}D4M-&a^;%%HEmzyJe7
zfOz=~JCfXghJTU+`QN`n1mkTLSR$z+?&1480CWUM<}Ux;Lz<sLG&&^3rb|Z8YW%4x
zUj#5(G(PvIuiB{@3(5Tgv0~ZWoV4OqXt<FZNYNUPuEh{3an3t5hv@j~y$*_Io9v&}
zYFM$ptvO!1outWoYWi3?^m=fU`eXQToXMM%O?v^;*n~E34n*)h(f}cJNN=xD0Y7zd
z;d<@{`CfU-%~mZDHP!yOc`7;GoFslzjoEPJ_Vd}D37BX~mNr}7qQl2Zr*``jtJSe%
zj!qg&33|R3qFW^SM@sG6!sszo*20EK<ylGhxiQHJUpQv}5VqN(e5^dlSahE1o+hs)
zaNLT1wuRYUTtv`Flv(33YHANClVxZpGbVa)PC`~hu3Uo<SN^UN5#Ta(=n)DM7ACX9
zu2i*bt1_Q>%0)lkzYruAa91*GjcNnIb&=v)G`^3Y%OMirlyC3sbBH{;X#YvjscCO1
zt(=zskz${f-;bh@y67?eOS&l^#sR0L87!hchC+VBfd8z4PrwrCtM5*UegnJ=a+m>(
zUZ_PiE0)_6S0yF|@%;(tvYchVeD}rPXNnc@ab%@sR29BtY3@C-S<Otgc!7-@-tvcR
zhalr8rVVy%5f(c1zk`PrHU?lRKVj#LH{G9gI&mD(ByJzZtu6|albN@wJmF8^_j?bq
z8TQfy*SuCT@vacU4RM0x52pyy0$=^;Huz)~56dU0L>Lc^VTMkn=^-cWcVPtr(X_^}
z<#5VJh$`1`evU-D@!)5#>-g~}Z2z)CjZfY#BRK95_j~hNh@YUikdUMv=}Y>sr5z1J
z^JM2^<M!TY%=&sz76~a<U9Qvhn(a><Bp01I*|%EwmN%kvXL}tz8P3PMIo^lg<Z(Cs
z{g!LfVzWgkKs|FYiD`YgSZz=;j(g<S5uZJ^q7L6o4>f3e$WEQl!keZlH!Swmw%k#I
z7BdDj={75Z(_)C{o+a3EfWQITX;0ZnY7O?3lZac9An4^yJ8E1spY#1GkSi2&WgyK$
zEqaq!7lm8Wy+6oBF#aGT8tdy<2Z-LyH;4y714`7_!EF$ptZ|)t7sx})qqP8JB`!zA
zG+tuoYsai6_qaS00e*gQDJdnhS2XOknT3TY%Tc|5<WPF`59OUz6*>f4{E(BQM?Xgd
zv<4iQd`tl(NN8^#`o6Ol7VhPBm%6(<zc%<WL#K5mEX%NVb;=?7<IllCZC95fuTJQ@
z`^=z$AaN6Xp29^(n6oOg6?<XQ^d4z{mM!v||LjH}yZAeO4@;WSg<8FH-1dCd{FKb?
z<XY|!Oy%+Bc!?20VKVV|NF75ALt`vY$8}onuAG+mLt1Wk7Y>#uKtJ@bLT*F?lZ}-8
z{GU-+ufBA*bp))~zFS|nSB$7JxSFasINfpY{_;)8jfnN9@vi!Q2%b)+os1YsRIlw|
z5mrz}Rq*a}5l#!Sl*X^dXEHNNl7mGoYV$&m|0%cfkUxD4CY%r!SfkgJ3bQ%nG|Tjx
zPE3AUaK4J%aG+T&XwNI}=O@wVd6WHxxJboSv;G8iTn8{(R#$(lP%R97K@?kIwfIX&
zsBbp0**Y`X-=C&XFP`prcryI#__au9yu{a|5FtGWo>}s$lD;?g@8;{Vr3kI;Nf#u)
zVnoB0+?w@)+}%NRR?Mv7Va~0w)^4Q_&}Su)G$p4efQ1>XceZGU$&lmaGNj{hb=Se(
z8|`WIqeqg#!PL9wSV~hus9$k$zn$+>k?%nXjmLqoM!l59IB|Cm+1^fC#|g-XeRNNh
zXwdh1OFFJl`+kV9*7v2@R0&4je?MLmRb{K8Lm@x(yq2z%$rE~;%Ola^x8)lE{pR4{
zLnUL+zrFy(8wlI<tjX7}k?I<LW_zTmgYKu>^F9z7joK?yv!qP*i&n1<*|armWGT+B
z%<{sxb%L>S$#h0rR_8JSNja8!v*C|kKCJ>h%S(EnD2h!lF761s`pJ9APX|*w$BUZe
znwTw9h=1;e!<u^g$e;LO9USbiR#oNr1_v7{l0b<*ym2aDUubG8Ev*Af@7$Rg^TP_|
zi&cXexpKB=@@>}A8#4GYB0_|6kyiMo>kKm56lr+`a##w@N6w<KH=fFJr@!gg`tjn$
z&DG9ok&1z!P+tRE4AKflL1on8h0Vg={Lt0On?&CEt@FjzU!&$+cllMTFo}c#WN&1@
z&Warxs{KqX?df>(2Y;=DPk?Q)+J*PTL1%YIEpql`lrqsEmyxO6oOzquWV(k-x$W1y
zAn{i}#}T~nxnqpNg=nRvUurcx<FZJ~214=*E1j3_YL6&_-ig{W=bI)1+=vhpZX4NI
zx-=(23m<}y2rMmo9#G-f=>L+h_<W#J@!K#WZJ;!^l*a{sAZ=G)7lV*ca_LZbiS#j9
z<dlpICn5XQ*&ov?=VJv#VRX^6A=P&jOrzb&Rbpkvcd$4+{=7IM2zV_p=CMa|WXnc4
z!=-B2Kdfj|zfHOi%yW`Ui-eZ*dfdE|lcVuzLxSSIWJ?CEfm<9)*b%bAWQLD7*3+eC
z<AVZjZfv*TDU4m3ay$9NtAQnd2ocy`&CCepZ`lz`_+IA}T+^4*r*J)&!ZkTQ%ve_*
zfdE^PU?G4g1p*zpH9xzy)HZx_qMqm)xGkWFJJCNQ-*AI5ORo4+DSnv+E|y|Xd2nJH
zvls0Olj=)!z2nB$>=az?DojTeR3<g+7<>_IrYy!M*^?8$0-hvxqn|=t4gE9fN&Qyn
z*80B)?N5RXCP%HNs_7nrrZ`}5g)5mkem+;1onHfje|)S=w|BAJUj(uZPHmkgrc=IV
zl*HG=!%nRRZ<jNcnU?#@oyp3)*g#>C)R!y1(objb>S*d(1u(2Ja>igswi4U2&*Pw4
zxGW-87G)s~j);hm_+H21C!Lvs!X$E1(xcs>6{&B>N9K9Tu6>;NkV^TdZdfXf@6zh4
zf5@liN0Cdy&p`yLE45xN8;^%Jix8E;!Q)$P$XSVv`&HGZc~QgXfS-f``P`vF1?ts5
z%@1@XnZL2HB{S@ZMO8IRm1x-{oXj>Gx8#>u_=D9)7et4kuZZ5<939<$l6YKVJu+qE
zmYaFK%1-<21-IfQM}Hsr{=%)4P{&v+Nlm7}*;Z<Df>!NseA8uZVW9%CVuHV37>Ux6
zT)!@>suP~nR2+--t=~hceQngTV#Ch?+s^`o$&UX_=Pxu#hY_I^l^XY^`J<?v63UL>
zREF#l-&L1H|2x~Q0>w{SZT0U;3Ew?ZRy`jX(XqhBRSLSd2qi&Z-v&yf$%edH_62sQ
zFzUOr2AKox4fYQ@#c9nY7Eyjolm37Aewrso)TlAdlSRX>>lTwLHFQ<(>=Bd{oi9*g
zJ%#TcYvdrDZ)M%aD&;=6%A-&1S6O|^)Z{;9*2AYN@d~5a0Zp2?mR!J~`t)WoI_q_-
zP+&({L9V~Ufo^_oZeo!5_GH|Lov`mg(>d>4PSrOMyJG)7{-THPD6YKV<!Pw<VCy)k
zJ3Hsj+X)BL7FeJVYiLI$)6CiK!w!EnB&67%%?6ZMwF5G#--B$A(!NImZwy`V&g+I8
zS|Jhm#{B8g6}>ByzE1hHWTNad9}Yzb6dSvUO4j4MS*(h49ms2%t}{u^%}iv~?c~#P
z-6KHw0^4-CUC;VK0(oJfxwsgO1GzmmU~Th{?{vU7*0_fL<`_Z+1ekaIYNi-PMWxF;
zZ<8SGSln;?6nMRgvZH2ao|oF*=e_=|ZP25mqnUb#A}_DD3WFaWS6ftke0;!bhr+PZ
z@_tB9Nbi07Wi#INQ=iERIVv{5&^Dyh@*sjV?&!O8aAzMV2Y0V>9LE&s<yy-4)!X0y
z5!^Bd(yGrxdm`b)YWXR+yO2uVcuoBocsd}Ux0e;0>g}H?OtEYmLr;(QG_zbO93JGW
zUau`p2oNAG_VYO)13%Z|$p;P(HT0MpJ1u-6dghbTfebcV6D~Jm@3;_5*8bA=q}A(L
zX6E`4hxnUV0Z!+#-|<=(nq`C>7ZtBwy)qt1Vm0iE@Oy$+IjuwN6F|*rwb0miKLJwJ
zy@pIW1;W`Q7bKc42|F<Cx4B<F>$yrz+k^47_jg=ee99J?{xeTFHbouZYT!?Wk!Kn?
z^xVc2P19K{Z}iYJ^Q&r!_29a#ovp2QlM5^0nvsmBci*)DKh)~&w{AWAhKC~rGx@>h
zj!NS?b@_jpPtK~?PAq*HJ@^<ZFy-2#TSdeNV>1jiGBm8?9*mPTV$rT3ky&U*3qnF)
ziS~H`Zh7py@VI!m7vf!hSU$%<;}R+p)xF(%Ta#kU>xBHiIH01!x6OwRP_f02{&Jg(
znM=}fWQ8@+<+8T9%WH}sjVx(1FqAEK1U!NW9G#U@7I4_EG#h>ympz}UDtB5UMW%wW
zvDMIP{I1aV%j14V`7Y|{;)Dwtgw0m%+lc@<B<JDb0dzX<?(W3IsUSh$-iQ*R$z!M-
zsdM8<s$o#KPbq{8LwRJA-w4anKe#&mPAWvJpdfDh7Ll)|(Rjgfo6~?sB{G$NX-V`p
zy&DvRsH>_vxwwEjagWv1Rsp$B+&FA@+PiOh$5nrGojN&qJOxYtEdUb}Hz~=uySu>F
zh1U;NXohts>CzSFJ0p#C`4bRg)^BDCM>&mBN!vNJerWRUaXY|Rk)z?Jz)JjP(f~G$
zw#DH_U2ql5q9WzdXSn0Q_v(+{a<&SJ54$MZRFp0@%ANqxtM~xTQZdOs&)z61Y3ZYS
zFB-DQoe3iH9j#~<y(UX;xJwXK8*uXHXtUrN`>GU$^H_{F>aznzz)$dfvL@^0bzhXT
zeD0W0(i}m|j?NO7ZbxmAhw#TPMOnFVSbvR9Xb+C5S{~2RnLQ$u0%zF~s-<area;@g
z*z-JRBEFzAC^=z5JArN(GzdfYUR%jjeed#vr)&{5WaNik)T=i7K%v+}MwY5ZE1gl`
zIUjNP=BKH*94dB%Nn%kxO9yoDM2Nj}LN4z;x~DAe+-sfYLn6d-mT)^WgPjv%l@`m=
z)vT=(iNZ<ahOK5M`^Pbx$$PI?AHdD{bC<ySvgcfu#wS{)P1_H@%%IW>5ET{%SS>_=
z0YcGY_LUPr;1-C@W)p4uZw)1uG8NS~j<`LSUR}3c|I%(k@+3i5TIlfh?tb{FANV9u
zJMtjh1F>_tCc<Bx&gXO{CwD~eIoKT7ObGj9vuf?Di{6=;<fN)BLJ02jb_g{`>j6cS
zqvKG3C;e3r^Y7w8tY8uxjoW-g$B&Dz+{fxtir<zat;o%D)6#u?N?^FWsz5!NEv8<#
z#-p-Rs}oUfI7AB%7{-jC$?07&VKchppN4#jA_YmM(0n-*6`+z5`L|bzU7d=(=ZyL&
zx6oN2OukCh|Ma$><QTtu?bgSYh=GPY{AbVJw_*U5ea^^PC!#P)%Mebx|EQv=JqVk<
z*HOpy?`KTU(?ptyE6$RBFLSZ42$%Dsb&@(=<av8bK%Q1cESfRD=#+LBN+~a3C3_A>
zC2oA0I&nm9(~6#F*OU^uT(9>rF@jK>`$iS`UD=|uKTZ4E!@C%Y`IShwBjGMqglD;;
z<OjF)gs0s?;eQsfdyJ3UgaU;Q4(tHaKE389WlJgzub8U57LT+f<uMI>bL2#Sw)!_c
zuEVavGrc?1tHE(Xwn2t@&=)8ceqHxm>3o6~_L3~p++(_{>KF;`gWmTvpcNC;j_Cc`
zG75o>kCg*M2fi??|Lt<Md%D;4{c-^BE7M5WZ7KObftHaDEjE|qbv>^XV7FS3!U*}g
zL<;6Tg^7!&aDB!)Fl5!aq#8x(xLZhG_4=9CLOKFmINhho7LixJKOJd6gPgN}2`xqh
z#<;CRrgx~!>B8FlZE6S-sv4ccj*Q$OK@;95BtQ73Y>u*ZeWcuP(LSEN`?f;_;+KL_
zW3K>=fF(UV(6Dc;`%CkC8bBN(2H322@d%*vF~M(ngdu;wN2@I&a2mS1{v$Lmw3f%?
z1}UB4@!wCDic`n*C$4U<#rO;?=OF0$gz>I-&JQk-z?Vbdeosz9b?{-Qcye~>wV7Fx
zv&*y3e?Q2sq%N^DJBj1ddwMTO8m`l_Wk35(5H1HtvQ*>E_YSvl-_D2lUU2@(C71mA
zZ;+yf#&I?FTiE?<IdLqt6QvyePD`kNt^C$^b@C{nmgLd2&NmV1RF@{&#u@*ve7TTr
zlo%eJ%SZF%plQr5v}eG<%5@jrz=aFMHZ6?~e}l0WkAbU&a5qEC_svbjLnktz^)R?&
zN&w>z3+uqESNI9U@?-oZi3I~^sBmAM^xjR#yLLRM>zWXH6N#+@f}pPjrvD#L-yKNx
z|NUQ*E9u&ju02DM5<;>{R7O^26teeTam(IYvdOh)BzteN$=;Hcy?u|~pWpBQd++PM
zUa#l#JRj%rc${-6|E-L+XM=+`+9Ms_IEd3Fl{mHyMlM?X+Ybqx0#nx|Q04LSC_)Mq
z564V8I^BOSnpN1RvDcsvh4zrTyZR?miHOB`rlAVwzw2;3V=w*tv-6iO;?ue$gzb7S
zwd<t!J<gwCQUw`}4C-#(kJ%vVcl4TibQZ@?G?o6_=;<2XF8^(PCTI65*X`72NXXCr
z-B!bs(UOD7>IDD4BK&`2Fl@(7cXr_qCLVf|#07r7G<xqjV<<|H;-T3@${NQZHa`j;
z#Y(oJp$aACSDHNKpb%*p2BU@5$b8lF?XlA-EPt!7soe4Kfu^m6wBMT@!@S#S$KL_K
zpYIbLnDjimyLPSV^!xowSvaeHTIwH2VL2%{Jo~KH>NvFg;0y69iu;uMU8tdYulh$L
zf>?qd`LGEJ<czZN-Q2c-To^=<OH<lvex0;DPw@tBpW(7+8ZE3RY2J^(lNcDtwwPU!
zwNqfedIo@-pte4@b(qg)Z%rR>iTGZu(Zb51v-~-egQ+vT;=>fEF?oD{^CaP(gV(j9
z!h+Td=L=X&BBAKQtijINZwlo>!e*1H%C7U5tSEfu&9q$}$Wy&S9IljfM&#2$#E`wZ
zTJ&#K5fc1x;)M=>i+KF^SsoQRwkgR`T|U1tT^w$eozX7dl?o2+5|`)wt0?Y<-deh=
z|8C2Ip4@M;H1C#sjY3OL%x={;X#z<;L0I#$OrP*-fLyxqB<R;3EjwpA0k!eLz6JHa
zhd3kUc1+~9qVgT`J${?&zZ*GwZb-88r=j@Fgha2|%=n6r$(fnKEt8+LR|$g7&ul`$
zhFI3{LCs2kIc-dEiuC2WA3V)K44$7Iip1w|(1L{bM5~|uRuU5RetKT<0<|d0!)w34
zE6%;p#c=YL&oXPnl~{XzK$%mCY?5rc(cdAfklqjMKfQ+K&JPQ|)Q;}If2U+&@$uFx
zYPJn*)p`;my>(;LR8iWCh=^EsXOY<uH3&i8H_lYEeMAkYc)PlOJ-qbqQT!cuj|ikP
zii+3Fs<BFpzn3dudlTylAErWdMGV_m(D`al{)qqwxmy!Hi&*`;|8A+eqoa3s^{U*|
z6?keh>T!`Z%7oH2BD&#;{<8O8#cI#&o=*O$H|MLne0A)Ort{^0_asF}8L+k0-SnJU
z#(lA-;}(73=tjV`TaRl*oPA@%#Zc7+{8Pd$B)APmgZ5A<rOw>`Rpfjni5VB``X!|6
zyJN{Drlo!Qy`)vX>8-%(;f;l4>pVEBpnW;q@ZzXD`>~Js4=eqt!5r6dt;aVl`5<(A
zrx9ZFnz?4GAc0G3b;{+&*Pu=QfR>`o2)CFl)64B*;k|oDD^Rm=?#lmysMtJlSnHpW
zKFl{oj3lzkjq@vyqRd3yi`GwxSjZVwoh!hALHwuj^ek!bT<vTLc>sy~zvmbs2mc#^
z9nKnwULTF3j{7=LMa_WA=iMwSOKU7>U8MQvOX^PbgAf_1%(cnM<5jhX|K1u~#+NuN
zYikeEuG&2}(;;agGklB7s0Ie?vI@_*DEQB=(KJ1|h1{yuu6jN`1OEaQrXpEZ)$Mdd
zg_e+gZi&W9wzp1%iH`TiS|H$ArYl)#rYFzZe9Gl}+;~VR1MV4y2b4Sgw}RQq7rJa-
zNM$OuBniJ(4j>8Byr;Yvc83?25A)b3!~Yi(;ysSi>d{l@U|H7JKek)H85R6@MXcVT
zpej4gD+DNU`t|CG68&p%YnOZ^-}T>3nY1}Re%c)P$!v6iGeq*=IWBv7dq$$MEh3(8
zIx)JAHgJ-amErM+iwH7yZ3&y5ZIkCu&$`{_W?%8yW>dv>BL40EfqhyI^6IAq3`w#c
z2qb;MxY)81&B@#YQn*su{=qyT$<HnfkIv+@=_FqDWq*&xqlQS65$e^KE};9afBGZZ
z`-9CNnZhv-x~30kJR;uPK|pUA+0R6EsdKqjRO%>ed!Ue@ooxOrvyx+G&aj`JI@$xq
zKipe<C!d|foY2TZM++FAWBblx0^mH<O7`RUNv{06O1#(UX``y@>Hah=aeYHTw%xTl
zd_M16*t7SDupRidG+)<2v|CeGmUzMa-$g7HpIxyT;}cM~kYytbC0{WJkaCy2gw$a$
z6(zS=aWr*2z3%q;vlJy)O+`uMwSU)0^%6~*^P<6b5X|Ov*~dYQZpvGlj#`Tj`(p8V
zU&QX)s<su_?tV<>f8*^<@?CPlk0_+PHUZ?#(Fv2@>_~c^_gWt>-m4LD@PiJ?xL>Od
ze%;(%cXM4);j;(+JoOj#+Ap%XRSj=mB%ag+5sNA40CW={(3>7_o`~GUuC5w?CMh|B
zk$ceDiM0Vrf#f%#3V7p363JEVB8+V@2oQq^&1yu-9&;6Kj1t1Av)_}G>U2fEv|i^r
zNE06NPQ9zlM4-8>_*bIl-0K4mL|ZN*bn$DmlON!^x)|7HDEx4(>|P!9nWCFB4}NGL
z9j77e^k6Wy&h8^x(jSo<-seeZStD{o%)E}-RKAOMVs<K+$1XQZlJOz|ofo>lw*HqY
zR6vvO&n2cts0le-1^Fl$$;WyRH&CP7=h<14lO!WdVfZ?1mVXuW0&Jk@l4LjC&g=+g
zceVstL>PBxgAZ@((=C5Mk6%A~aSM*=m(B~NlO;e2O03pn{qH+3UfjdG+!qo<wV0^D
zm>W~=MG|$<9f4X=a?4Bp%x7?q<>ASwnbU5zHN4$NTKNB7U4&QPSAHGAPx-Sbal-&9
zy4$^I7uFAGjYvpl0MZnlwXes;ziA<C-?yj#_cBbWb2V4VfcsHH(hubD4-!XS0Qh7p
z0<zEVQkUmm>jb`!te7`jLb(3DifcmsJy$k&Zr9yIs4=`Uj6&ISMr2w|T||`A-ilH#
zT@!!GoUEri06cI{e*>^V@U(u`yBvv*jc;vqh=@P#rLSBMG`|3S)(-M8M84<8_?I7N
z<J95e^l|)sH)bbj=V(C6YzPXZx@u?HR?CCv$-}P7oU~SDVvwRGB)xw9IvH6K=Ld9Q
zN@BR>ui%QI$83kEiBZi#TMON8jEbH&k-_z#kIGU%bb8R%gwuC1Mq_HqsUTIa>yG9V
zci$Kq7$*t2u@4E$%XmSDjj38_L0BUa<Lz3HwE2>Hyq<?_oA>p^FL*A^;O-S%FJky7
zsmiq&u-pXy{;AMW>lO}`3)#Ku{QO`^#u7Idm|FU~FZ~jtJdatLpqF0gN|BgXxr>0@
zQx@ER+d`k<Y@dO8)~8-$Hc^2#JH9YNYePv<?9Y6A)FmAJ3rO2NR#!QmB0Jw({(%)`
zNlDNyCtUKK$N80)SEfNb;x%oUO+ZUN_eu}&GRrW0Z!%IiKi3+jbp+*lL%S;@BwJ>U
zjg3Cy^zQ`{Fg$FZ_za3RG@hDSQVUWTz9GE&Z&8$a4dpeDS0`kXqQcGX;S_%uKOZi+
zw?3J`XF>1Rc=+q4wT%rFNK|c4s+JhPc5ra;llOVf+`hBU%cM)xWb(ePV}0sKtD1TH
zkG_&t_XUChsVPXq%AjKuFSk^~rguO<fX3PQ(MubaH68O%VtnO%7!{5Za<q1FIpLkS
zv?o!0o5)99$U!<ia%!}&kzkfc=O@AWiRvtMYvILlI-{`;NmDinZ?9sL+x6>F+V!3;
z6t^#-R-_b)%_XZ>-<H|xe3!PkYcuuu>YeLrw?XlB_gdduySBUCVEstcj?Y`X^4;LO
zzv4U{1Xr#=|G<L;sP5-KQ`5>4b4P1MsEAHjML14;e^p)V;`AYH-spTO;|hxR_HMfE
zW-Ivcv9ji7$AA4)n1C0Z5FBj2Q+*k=lKp-t`(`nx?YpXDL`1kO4^OA6D)xE6Ahh=`
z)(hb68X7rO2PxFqj)01c#C*d~$H{A`W0FDJ=1#KhDqfv^7sr1~xoDx&2}a6fsiTZ{
z9YiLq$oMb|CSI|@0}ouIf0ixfVBRF^<d?k$ymfwCtCWJnq2fC1*AZ!gk03r1`QW1z
zMoaoizOC&CjP>)zok$2(68-}~=l5V7B~E`bI=Ob`Q}iRFKaUWTX9o)2+S_o9#+y9t
zcr(y3(zuJazQQTbnp5d`;~s_jf46R|?BcTZ;o<@XjRG@@v?7Q3<u490-+so859qV9
zo3x|@X8fZe8kdEKU<E!6@<gAn$Ej(AuKtir>;lFgC7wBIp2;If#D_68@p7#FSaEiE
z)WvB{ODctPt&o)?^~3+Z@abbXl#!;SqKvON4<x$4=ao#n=I6A^kAuTUBLZemPW+-c
z9I(!BousBdNZ0VQ?dH3Wp_Xt|5jUHyte6AV+$^C2s>fpH>z$w=)#4=!S=o%wPUkL&
zdKv3|^XQbfC83stKCYulM%_LDsfKJD6rSwd&v}_cdswi!*#gq9k2tZ-F5$UbL2HGM
zg^@5K3hgn|_BrO)dYb%HqG2gjhNL<qn_*>6QD=33XF0>|eqe*#2MJ%o>6jhP9~nk5
zBz*|^%5?QD#e<qxR>)qnU%X38)$7a5v{>%n&9q+mFSd%FUW5gG?AJPXvfX#-)<R>r
zF@Fn}>&H+=bJ<U$FOA+wL9VQWlhy9!dRn1A^0ejgKTPSsdOKjdo%7BxjdBT!R7s)q
zc6rWn>63v-bX}(E%zgmK1#3WL?0$gu1enJZ|9xqst&aYrBSunpk(L;U1I{dX-;O3v
z>!w3o2Xvk6j1~l9JJ$!=qg&@P{{L4Dn*$%4t&5UPy$LH5Shc%3mAL$ksx9LvZd_%<
zDoILKP2j>N!MS%tc<5)6QbYtk*;L{Nl@G)d!xc$f)G=+OG_Ap3*IPl8owIi|4jCC!
z@&EUEqL{JdByBduhPOO1R)iV0trtxlADPSQ{z7t=e>agyTRt?U4wsrgJ}xSFOsf>Y
z_2D1*;SLv7QV63##0+WziBTgm@Lm9(j!B!dKYoeo?ncXGNhi(!E)eA|vJxlY`#M#X
z8YTYe5*v-e{vbxxA~@SFz;!X{ITFxzBJ@*?xN-?pOFDa2vCg<Zlr%C5H{eLi6$H$D
z8Oi1}s_pc#Bq#MYq3xMBeMqV@=i3!!-w(%^U{&m2N7t^7^m%tHcZiYjTjaO)qTx8;
z!}OaxmBg4haj3}M4r5?uBm{OuF71DRzc4h<tY@K9hRnvA-y%3c`H5;O0;Mof!a(M0
zs7$5p<k`CH9mV-{@<+sl1H}2#OOTe~#6B`%p6jgLBferWh9gHM!E{N6ichk7{_$Wr
zb3Mb$&vU^suMz+I`0kIi-qVv#AMvl578@r=O$?;~Iu*D#6ynii#TCm2s^hQ4w@G^5
ze~A|gU?h%EH75X11uk`)cv;!C*=BZ2ZNmeL`~!n`v(sudB56hf#Cle+0KsH(gZj?>
zV&@%0&i_n1$LgM*t&82&3eS+fxA5ztZP@soM8EM5+b=Pnl%S0X>~;+Oll5>!{w?n>
zM!)Vir=}WSj{k7Xw}0BrQ+iOw_MIvT&nn*}xuFov9mD#!!1ZLu!kFXY|6ZQl>l-F5
zgwl<V?uW1=z?k!+$cW#ZaJGbD)Q6ack<i=i@VAIn(f?a}Jv43&(4Qe^*eaMEQFi4u
z;AkvFMDpsfQS!&9n2w-`cufJTulN2lCAsF_eQRuGbyFc4EnoG4OXu6iaXeIm+o(2$
z5ZGB-W>tWAVUfw4YSI3;E{0#gG^ik|tbH=&xi*nh$ZFK;6-_>TT!xUM{_IQsvoXo2
z<^pJoW@X#ukZ1WHd>-#V*y8+~U*TE7-(tC$z8rfA@psHNmEK}?Q>;^lVct6X)7A0N
zFr)vhQ5$c4#6#ULK|kO3Y~3QRrUt8_muf@=OXHZ?yB%|^*5^x3bk`@CAyxa|(@}c!
zrgf%q&7P0M!RkEddFBQeTe_q7jRhbdm>J%wqjV5TXAV;X*6Dv&Ac0Ps(~HX4NC?7e
zPU?*%>B>|{DaP}JgT||6{qf(w;$N%};KKlgT@`@)<*HGqC7hjp$2CQ}`r~%rAcD`k
zzNFp~`M8o)T<n0VW#ZAF{lC|P-5#Hx=>=VJe$aH9nuP^@5uxVsw)H3vZg|jEoRJ$)
zs<vW|I)U}+v9xsi(Eq+5f(n!st1tfQ`6+Cx_5+%AQb}e43)P@Js$B%9CpXnx)!+F*
zDtA5yY}f}7T`cIYM~wj=v|;s#!ko~D&ms&H4=*i<HP;;u(>M8ehwwNczg$<8`v>Im
zuR;P8#v%+Ft2{3<bdR$5Bhv+h1wB#B6N+j%SziehG>TG_tA(9>ek9mlMnrYS%=%`%
zs`EsHdum$^$wp$|)+JIn!L#YYVjQ?L8^rd#?!mwCzETJsRG97;t6$@9dC_(5#^Uov
zeDh=pCI4OUTc!VfPEMWMX&>Q#uWEE>bH3b%go?DhJv5n*)xeYqU+or|k0QfHcobUx
zPM%5LK|k+sah%=y=O+P$?*IEcC%dubd1gctU3fR)2yq=CBUQ$N<kK@rdbVdY`di$W
zFP^W96wI5M{#&7!)6$2M0}CV>Ye#*wn`)cHc;Owmqw&InRi2{CHTxeoX0N8Eg^+WY
zqyJq%EEOyn)IyHvdXjp7G(HKh=U;}YM67|tl<wlK#=_NxW)hIeu>K1vtT+9sP1s4L
z%<WC1EeOfbAW2?8kfjY`!awtur()4Xny_#eTCS>^od*WlPN8tAv?4uz1Wx4C);8B{
zV&eDEVo3r%Of)SiE)h1r#SMzmxT&ea9JP~)EX=>uOB*r2s==opn5DqhxcFPUN#vH6
zERTJeA^r`;hsXypk4{r5n!m0Jx~LW#DHEP=VN4e_9`=k}im|->M)?h1Jy_B)-^3=e
z9$rESM`20|OziP}{9Q)>ZnO*yT=P7?<AD*etA5*^u?aH=_yu!s(1ET-3V7y(Z$oUZ
zea;pn>$N>T_O)BX|M>5-gJR5K=@|w}@*aP7(hBBiPa-e%Jt2GYf|UpXL?_UAp_`ZS
z3<R{b&3KJBKK=KVdaK3cAJvyK6xd7*XNYh+c+H?9ny05n@!~}$YU+coE`CVZ!NsFh
zNr~2k08b-C301UJuJ;|Gk@ywhV;?-Yv9{(QB9f-9-8eF`1EwRQZ}YB|&YH3>v+CH?
z+UjacrtfaqTIuwX4cO|uz0w}#1dd5}ujSv`A|h{rT)9uVc=H28nA*=#!e7qLctp{r
zW8T6RGmRxCr+(MV0e``uN5$^=CvFGrt3c;@5tPP18aL=kZe~}qH#9;lE#s70t%5g(
zu>WT}c(p?Qj&e4<*zOXrBPP1m{l#aj`m8K6@~k&)4PGv^#y{$lAvoPFbwVT5lFxG0
z7SUbUl_TAGd<-dpQu8ehB$`Tcqng4sA{}dz!w0bcSo2p?(}%he&T!B7g@PCX6Y(v>
z(;%9~0+r)rdMR=7%XL3<)UG@Y{Ae;fCE!%j4vH}hd9VYeE8*b)kxeu>qX?r-Ctntq
zc5m)+f(2}0KEy|y8HeXr^PYr*Adyc)S4Xzf-(;+~D>EEAYe2zo|9%Sq<Hx&J66|Jh
zJ84zt&To}-ue_==r1Mufq0ST2SM)~{>)nGxRa@d4*62%DSBB*tII-ZVPHBUMP?fTE
zw0vWtG8GtQ=9kv{V9`o&<zDHjf8WdK_06b~=pxYq;kEj@2|fKG7!5l6F5nyT5sk62
zaX<jm@Gm`T0di>Kw<UaLzy};SED%xAwuj%ckQTjqL@_w1hQ|<P`la=JO$5lu{NYHV
zJGaS#O@o#&Dx!z5o#VDvg5Zq;)3|QXFx}3MPFznq2(Qb&-aA^iSXjt&+Px2q#oMBT
z{k?pn*>`*>ZA;4lFh67}YQosVH-B~P7*fI^s#G2M*<=NQuO(b)W20L{?otGoUK}Pq
zbGSq-#R4o*|3XegKbTFWXt`37*+xT3Ez=ggi3V{UFm8U!<wg1_vKFM&rzPK`eG@nS
za;}YRb<{jIAP-+VwU@i2{89a(4V&nzKI7Blr@Ol_j8UsHPI|(v@*L%}n#sMq$xZ#8
zP4G-It=H{%_K`2)KlE@$y=k9ZAy}B8^%vf`fE~`ugthH@0#C4CBsneuLKP@~9=^=Z
zeq*vS=EhBFibrFIB;1(~5NEH@-i)9;V_4U~W;9mDN*{9Y39GiocRq9bs{hm%v_~%+
zSb9PilmzDzdQwc-*(m;ZAf)ZevW6!<?3%3JZwj!#ecMf8CQ0*8e@5W1Ul*4Is;VRa
z{^aDu>D`F!d`c+Zpg>PBf7dNPH}{||A{u;>vQ#h9t%OgB`hNMjD*r~jgiBp3-+q0^
zov&3GOiw;REP&(Pz3}$(tE%|93wv8wb%$~jfy|~~4!EfgW7VJo#T5tzX=(Rk{8dN5
zgXBZP8Qv~-(3zso^?j9BdVK7(Xj-A@Od5F48j%B^*m$jbua4@gi1E6Go%V2?9LWK~
zyL(YMdAC)<g9C$TqhFp<_D26tE{PEe7npu@-!OFPsm9ISGm}rAkwGYfzKq(i%+6X{
zGIDv@q<~OuCznn*tR^kV`B*uhrP&w-fl0Gg=JYO?x620Yd?gzm;6QJ70tZ1kC8f%+
zS6w|R)3DReHqg+;5c%+-2*fB(ttzAU!AHH?ptFFB698S>79rYZ;JbQHl?&a{53U$N
zZQU207309+X>|Uy@0~eWX9^80rE+MmrOP_XVgo)L>G9|~+Y<M^KOLQ61StDQCZA`8
zPc7UIt<%v6F%U(-npZ&cwa_MOav<w&+Y+PRAHR6vGfP-qouuU2!%|a;>V_gA-@)=`
zs`DUsQ$`faVp^291v+l$AB9S$f^*ewuOf_v0gmX)iq7}+_^9{@S604+#c6N`9etr1
z3S&9I9E<bZGT%^pt;#mgca`6g(Vt6ar__rZ3sKG-sb-Gem8h7f!9kCvc!ek&TqimQ
zbZp}9ETw^6j9|YD!8GOTecSBuc%km>oO`{!lB@SZH>R|8nZ!Vi7Os|i_M*wM+Gjzp
zHeXvwx6c-LLW$kx-AphX_CdhDRp(=Q9q0qH#d5#GkHV(56&FonWY3xyd;mSZ^%<V1
z?5&v;!k&i=8cq)_b%R&a)X9|Q3(prJghdQlaqxT3{ZZT2X$VA4lwU&Hssk0>bSaLH
ziDR_O^I8=~45&kFYS$O|IJ|<#tMClLa=Q<<$mem(_A4s;aQnKt#p@H&a^Jlm*4{dL
zkV0Vv5odH~fNuEKo`PeAg>2h(1(^{x*0qKPMflP6>gMsPoNYhcb1_4g9n1aq@7Evf
zc@09vf~cRcD?29Z5tF<c7<q>9-64#40A}J_jr7StGlwp&SQQ{T+|k^me|n1YR`Wi3
zmCOPFO`9hGGIxPGxh?m5^zu~jr_6LfzCWea1Yi;&19+GPNQTLl=nnZ@jy{u<$%UWv
zzUi)AylHwl_f#z7MLiiJO;iptl4IVq?<N8u+d))RFvH-RgIb@<NNfNx+=FTvkNzEi
z7AJ6oc|CPiR17P08t$c!=Ifbm4Ra>@{(RVq%emBGs?JZ!gff?}Wf$LQ*y75fiT<4A
z&3>D6S{gwxI-+lI;5hqa63$AN=lR069AnMt5D;GL3W|^a0xux|{#OCgo<fL(wL)(d
z%$RiBls8)E8B``z_+x!3rRvvQS>XmEQ(+jqWuMVvgJ3TM<rAd)IKIF1w27$&1&U1P
z47I^T#Z#$=YJ>!+Vilv%A9MJ~oHs=yhlhvlQJe`pM&zGF<?xUXgafiC9m&s3nn%lJ
zcc_W55Vc~&J||#o1=H*$gev|`J($F7CDZKz4YLP!H%5wj!E`jQFp+r$u@_jeYfDA_
zM}#fHarGAu0YOmp@jR!Ejg3@|hys4sXy*$T^jS&Y&!4^NK|bQi@<A|O6JUi^5$*lr
z<L%#)3fnuYMn8;{gR_is@e}r@$LayyV5T~|a+TBg8luq=5&i=wxdS=EK5AmJiWm-H
zZ31&3$9TO7ZU1&Lmto=DO7YSuxX+Sa3o1h+UN?;^m-_dg*IWXs&|bMzk<`h^+iXsX
zW)HcVP!;CWX$N=LOTHz#qsPM)q*POzeYznzGZ=z|YJFT9cB4E!!Jj{|<6{%=jS$;w
zCNOJgcD}tQPqWfUtMmCaA#ob#UE}Kv3}j=EaHNriD$WYBBz<j0dWP&g2<Vw|8WprX
zRDJp_H8u6JTF)h|L5`8h-m$T7S}b9$6ba_fJ4Rc>s8kgshn+BvbxB?vN>-Me(CbsP
zGkSVBJ*agsr>dmHV1E7@n51LJ{j-NF1!S3K^9u@yir&Tv9qF8%>aTo$2>Ec_p@fgP
z2Rm{mi8@=u;`@$nCh<-^6Y=k<+DKj9`xxEFeA}>`Mn?Gc+sUgT83$pj(uRF{=Z|WE
z3LbUIpN^Q8H0cCZTj{H07{p{%+TN;>R8q0j<+@CF>FKKXb`3t=#^t4B0WrVvT7@AI
zuW(YfF%5vMcU-KWx4i{(G#j*k<K7q9-=$7__p~PIDFZH1yCwsiOMhu|Ji|6wQyVDF
z7}}9=)-t%OMNmYDAi8F;)C5ydLPJ9d5ey6r;F+r`!S=1=AI{&h-0G7|YZo>!NQE;E
zShrh)V&r}_gtCQZTDZ$GDIe~0dBYfUIp1CxI_ih~^DxvzjJXtJj?$@zhM77;Ia>!G
z$RI7%!MW;X%Ax+<eDEo;UnAe#Kc+}KES$5(YV;sk_5wbYp`v^;{nhphh=_Jr-zH`W
z-z?MgA<(c=2{}h=Xmnx`ROc*qw=3FA-uLJC{(`#SGBdk|4LRoBoM#R%UU48Y4i_Mw
zfz;V-X><O4|A{9G2}Z#R!Ob`UpL_ZbQ<lxnP8Vxyb#rn=H}V;L{TR;4Y1yBcO$$~t
zBzTJ;V}2B4eFe2r$e2wIuTck2?3tClF4LuN90T^hkf?^7^OgK4-hScv90YhU2WkG;
zX1kT4>agz2)=;0F?bxsIO5U)a4msZ@LLb6!jfnu1lRk+++A2O*qiHh`OY52jr4>$e
z%;V9E$eB(t_s1wAIg2G(R<;wpZd->9S_xg#y0<7VjHL*n^WS9A7c;W1sO&5wYM!>H
zTDm`1ok3_Rh-J?J8%Z~Nz5eL$#JakPuGl?5){;upmV9-IiM`|H2|LTW74LiAvS$rt
z{+*JMv)Es_Rt8RppFb^k{>r{YF@~u%1(ovNlRmPH!$qz9YQzi3o3xi2vHC7fPj$CW
z_3|JKHj-_yeo!f?+U+l75v#Ff0{rus-RLLAGx<~=-3d`y5nh{9Zxj_tGO?=*j*;g1
zeoVVRNOGoq*n;Z=0$veN;X!t^?BRtuyo$@(Gl*Oo>3a_~93Z-U{%k*$^gKMQ59Sfa
zs499`&b)DTods<SW40TxcL8J}*V<9AF--a~*&7LH5C$~;?rRnw*)zF_)cH*p#g2!^
zHeSF$I1O#VIH6Z7KF)RgaBWoPZEW4H+Tr06ixPKtrb*YE;6+US^1a>KK9ieYT}v27
zrs1{-2iI4h;3UZs6Ui0WnZek;n|Z49^UI&5(_s!Djn`_5*Z_$~;e2AHi_+QY0#FK*
zE-dExMZ~<lSKbajQ@Q=!P+*)W;+GQf;9wyb!Ro7(;=c&p{1mzcMHE0BEg!59#PW&n
z_<j9K8Gwu+K}gVF@f~grqp>v~xj6MkzmdyTyxcHDhyQqaAUcdC;yWWYvaKErdQV(V
zmcW_`yYoJ`6diHQRH*vw2FwaqsAgmt!W4OpFTT;x*h-cN;x%!E3DZ3hRbc${$-tlI
zx%L(*B{8$ys)lWr^|#OV9X)uqq(tYbW{{}xB(I&FPCY#XZx=B9XjM}yt3SVN#pPYk
zBC`orxR!Q?tTM0hxi<<5^FZqpP^U@?6zyF!&wn*jA(f?OC(2~?9Z1xIx6?Hu-xS}z
z4NKRrh5Vw@DHinz6&bE@J3^|K+<1m?el>)_WTMgti=eziBt&u(G88dS72BjlQ#JkA
z+=fyCt_9lH>#9ZFS0g&6@o|(Lu5-Z}eogFi|4v|H*3B`fY60v53bkN7u5m2~Qciz%
zrcwFTS3%?VAJTI*7T-u*eha7MpTPs!42_YjEGWz~HH}Jqgd*+x48<8zs(g560+ykG
zTFt7<*x!;}K&^Oqi!VTK0e3Bvu=eh4m<X+G4T=A3aQF4~ku76@j64Z;BUH#_=ba7h
zbI}!Y7maoe`H=6KXF6#Qu1O=gS6wm|{T2DVQhMKXp_FM7M5hw*{MyMJ9LPA-7To2V
zM~dWVs3ifDH-qtAqhi=)^nn}lDej<602-dT(U4<GwxTAC>za7`AO#D-H!KoDzBaE*
zmSKzRZp)<M`{~iSa0#I(!_R*Wya)fGJ{h3(SkwYOc&RjL!!56<2=o((#K~;rTt$he
zCHP6-S#BRM3u7g=F?vQKXYzLE{_UwZPe0;9W!9b(|Jc0-wbS#|0zhhwNCeUH0N+Es
z?k~RuEmH%Mw7Hi>(O$j*_0iF6FJ2VE>#r@ywIJM~m%Sxqmun;!f^WKTwG9{3bP;JQ
z-fz|P+)2x(WjY`r2NS|J^x;O!w3fn?C*w(CNgxLq86m`<D!Y_#Umz&(5o1Y5MKccT
zBJn;EGo+ASyY|ZH5B}>_o!0g{xoR$?eES;D_yp^gzKu^>D!<+gg2V-qE~#P(&bf^<
zTd@y#kF75^`xo%seU>JxS}P{lK-V^U{Z0i;Y{@MQy{Kv9js{cE6_OP=LPAzUkz)-j
zhbD{e=t%;zgzN7rjSM5|OHH<hhJtXAY*D^`8R;)201*_rI9-sUjH7&npl;(qW+-%R
z&Qk+Yt^4`&jyasU4v?Bn*XPU^Tvhzz$GBNF3BRWZg<D^g4dC|-?sq{Px^hK)iuRO=
zRO<M|?>eU~ysnwu)rLHwi=fJrc|o!WNFBHwhAcIQ<$4kK$Hsb*Uz4;Cn`!=9QmTxu
zT$kXtq2b$)W84Z9|MLAA8}ZvQThd(RyoscyEYi162OfCwJ}9BzxJK~?rglDk^X8cy
zPS-^Qlyd=B1h#6Rfu>~i`6*E94;+hE!Awa;W?2!|;kfN@+la-tZ_=|am^x`C-;Dxu
z#Qy0?w>z4Tl<n^Jwy5LY1AZv+F}QV2ut6IIN%kc~yiEM@2Do*i`)(nVH=ui)3vltn
zDjpR<EyaJoPTF82U8Us>z|-Pkne4c*F{m^R9zWo#YDB)^Dmr7<j&_EcAuo^l{IzvM
zo5WnUgqql=Bj#Ql@G-IsSo#lhq49739yQ<3x&fZ+v7;^cVo}U~EH$-e-cgF;s;TK+
z{1wR_Nf`!_@tRjr3faTPRb>u`AQeIJA2q20+ZO4?^C;CFAF;b`W@-|5G2U9;`r%%0
znsTMnO$b;+e{YBV;Do}yNFzeX*~m3YO}Oopw@|0c1LY!Uj^*VIr})gcdLtu1%snoL
z(JQ~m;8?jpJ;RSIQ}2O23K1YZEu<zGi5JlU1f0?P=y<^!f7hr6??ts9xtd**fd`<t
z$}6G!V>B^zA)(wZ3>*O;M{j(}{9Jz(k8+d-3-V9-iE=SW5A!;n&xGG1`1b8m^z&#`
z4_#8Yj#&badE%$Qm$FACuee8os$M~bMR=43KI-_h{f*1u?4<KL8(|1G9htE^-oNS`
zAHO`mk(#>9V^ksM%Ec{J%bWqfJMuU?TNC~n>n!&UUdUrwTg$c^5Bm-4aVB}qPK%DX
zc2VMuftV7xjLvX&guzgDEQb*7+3}I!WlWxtPADEK+~{F|HwGq5HUl2{@na!b{UY^@
z7OB*qU*9IS7KGT?GB5E~v4by3$7z_0{hg}bLVBQfr$|Pz;go#F<FaP%)(|xDvz_pq
zI$uKvH*BUhVTfR`!pqASVO0{E%DbJ-{O*hU`ohZjuo(B~%1}*h?k&FW_z|n%5CEBU
zW22nsTuKWUQ_9(NJUMI^TjyFxjP7C#)W}<DY70V-Z6E(iL(rZ%C$~xA818h(PaDbS
z%P-P=6f|3}1LGwy(^m=HvM?^ebY6B!&zaYkdN$xldBjl;xhzkdr9??d3@EZWuV>>v
zALbB=L(VF@8Mr<AXliOm{9e9c;pWHVt+K+RKj3q29Pl`h`s^lqZ3+BYt`#S6Bs4dZ
z)QT|?FBjlb8~gC>|J{DPu_0~2#P#6iDp-J|+p8H;;_fccvfN1#Iife;FuLIj`em@J
z-?J~QTBAOJxywV{+4;d9u_a&MQM=N@T!Wz58QiHPVxK3oD`Pjjm-o8$pM3D5@spY#
z`%V;e*a3`9RJg3bz8ZxUXlab|4PYjT%GwVf`!(+c4?zr=pi4N>i}N9Hw^H&M+mNk+
zL;wQVT^yq);5o`lOq3_xjk;3_wXjpuLkh23`CwOr0>tvN0+VUD+`q2VQ47wI&+cUA
z8|@lZ?BvglK$n|TGB>}Gi6>0+_cH&nk<_yck^NY~1G;62sns!BFE=$tVm0Y@nuv`_
z6}2mZ9i1jW2d+Iam<Al;>6uFiD1siBo96H$qMFv{BM2<{Kd3jm?ZD%ck))luA>g#X
zN4P^C$+<@X=F7=y6*&fNFbIV~y;4cn&;I-6Htg=rB<@~XQZc>3!G#ij)WAW#HA@(X
zyA}$kzeApTAty5f3v#p2uK*>Ix?;j|w}8^-EIB0${+ZnR1dquiHP<`uv@9c2yIlzn
zJ_Us<!n$qsy6u7>icL)~g6myLDUay%e5K&B{yuf4@#mL><glF$3+Z|X^b3yCX<7FY
zwB_aH+wE_&-ly&6XZU~!`8?qfGbV6<W|nb^HeMpd0!fWeB{<B@m5(iWtgfi|g0(`f
zerOOGJgt)%4gLWpd<{Ci{O}Xi7Bu=BJJPCei|))?3eIRW&TAEwR57I6*`n`_2M<i_
zdckUBW^AnJ%YLqM`-uDc?g15XN8NrTBY>d4RxCq-l1c*_u)ZRlfUe0bTl0GbbN{@A
zqlsA8B^Cg>#C()9N)=skTxaXEm1Nw7bCGrHt=k@fGIcV?U0ArEc_x6)V*24+Q)}DU
zOrqamuKNAwZQ*Jns7rXx-2Lkz0EBYbma@VkFH8CI=V&>zFs3_lQ&ZRJM>sS8*3hW9
zVk;U7+`jD*aMMLOd!xa=(&dZ2_hAzh!b92AN3UwdA4UjNHs;NHqf(*PosE|DUn<Os
zJNeoEw!6j(u-sY_050$^60YzKqc$@CE}2YCEgA}1W{n{FG{_XPSG^m1q0|-%B3+Im
zb03|Px)46^-+#V7DbuvNrchIx_j{J5Y?7KNQ;C|MT~O$%tbrgTM1V#<c{EbFa_|qo
zTj=7lzKC><Y+!;ST#0MZ6J7GH$LH&&y86KHIvEi0>L~5)_oF7G$3}+X`|jnAmcZny
zysjMVzb{}h1I2n52}Tv=MTP9{;O+bw%-sQ-1gc$e5_FsSV!BZQ6yLtV8A}!h>;?Y9
z9+?XX#rKJqpJ<j5qibhHB^-vUdN&hAzN^`zW;V)0C4WWoKBb}vUlyoZjxuy=S@gSv
zB?T-iimL`i)`E>ia%5E#ir$HRY4r-!)&%drH4zq*H#UQ^#nY{Z^7dd{dw{r3R=gMR
z$*dZQrdHB^k1yYeJT%XDF(`)=;Eu`^^C%N-_|Q3lk}mUf_xD?%`Uu!Akn#19Nh6hH
z`iMhX*a&+I=V_bnG56D*C4PBp66bDDH{_H`;Ut+<45tXMO(p~NU^tp@C0s`*6Y|9v
zZize6qqpL`@NtHKL<Axgu#MT|%ZB;@mkvHwGTbbj+fJO3n8yn3!Dt;-676kO4*lmC
znbT)4Ov;}+!i%;eSFiIqFX#d-N3)VLC3RN2{{eD#)(3Op%l%e5R*b?+xPYG*u@C7E
z!rBJyw8K3(h{V0><abiljkic!jqxa!l;P1)^F$fDQjXYf;LOqRXJq)4+L?`h@c@>X
zX(zaGCMAdqNYkN`UL54iD*$z{v!4W&?CjnV$CcG5Xm5Kzl-r2Bq2uGLgYe?;U*w(k
zMVnkX8>X6?VySa&`lxD=uYp<Qe;*735khFzH8V-6_1^UdP-wd>D}4MSr8<=;m>Cbc
zBEOY-b5N<BUH7Z$zp9gD0Zj!(?cz#_gXwuR8VcgItBN3aZpZFMo3N$cM}<p4rN^zg
zn2a9*ePW6OKzV!L;vRF6$)zq4o>_sK60cY$5tOT(Xi`HxgWHU_7wop5*Q3X`>tnr~
zt0;B^0SP#6a~W2j`7B0E++BEcvU~MQwDkb4{c8(Zm;fVW!#+pT0aoJCS0$h!z2oLr
zcBST~7_<4~3~}t$a+&)wi5Et-^jjzDL5JaiC4obpvxgdi936^A1_p2!%g}_uV~Q7S
zRZgbsjJ62ob4~Gn{wkn}ogG<NS${KV^3yRoZP@rUO*&6D?o(!tsqL%zIc+&2-G0{C
z_NbJ<lYX9j*?DGFP?f7CTm+J9Vd3&XUp=VD&_@C4wC=w342yZKYI0lccP_&Nl(Cw&
z(~4t+K}F%ro6$NfnRv0&T1}vfK`dr|VY$?0BJW@inDlFg!NTk-4D)qShgs;)ph^{&
z2qdXYygw^vdRCGbh*_hhB+byYVf;b=4c6O=Gu()qNXwafe@f#oBVCg}zZSrr7jF^v
zm4E-9->?o$$r@^Pq6^d$<x%&Iqsf*vvbvdMfAEpwHfU;IhOYh|^v%7SJbsPJyaZX9
zpZ)#iWq<RUCpH4KyKy6vE0*jgq-xFvLniL%Gew4fS$buZWPj4?l#+ARFx00*J#v56
z`pQ>5?rm9)$a7K7D6*(<<tE0{Vi?GfKT36w`hp1R`KdDEj{+{rQIe7S#fv`JTEB<7
z@p&Q}8Vry55;;(*Un+)HZeBzkb1sVMwEn;ecH5w(3`ohN33zWmUDqB+kQ5RD<%;it
z&Wm6d68Lt+W=?oB>5i2BzM8fBf57}xHu+^ma)K+^UjYM;mF2D;2{bnnZQMK1_{}GI
zXhKsxmD3i@OoHsqRpq9N7d@>Fg1vY2h)=)!iOok(?!UqHM0Z|qzUz+H$8uM>Ce&<&
z3Y9c|JKL{apDi1=w(<o%+7q<7kzet}^nJ7-_+9EIVM&9Gz~&>&!Ou$On~B32wYkss
z_VaVT;X&S(+lpx=DU_lk1U$}=w_QRH19x2J=*<T6lZDR9Bhu8Kw^4fu#<z@?oB!@*
zfeJkh=#Oz$FZ^yv+v4udU=DV}NtFwXrAC7W3i+Rnw-`Q{A|0Npepilo4z@xjpwKoH
zhdJ5tP%DYJ*bA_Afjgj;1Wjsgny#%hOUWV0IHiQJK@n9V<0jzDmklc(gjJ|uibxY(
z|LTSLDg7*?%$eU27IX;1DDz5%s|I^|n(ROf+GO-6P1R-8vC5aFZScqhyP4m?K_5+t
zw1IGps{(BB5hK5Kv)Fbz6ekpayQFYr8z$eNt*18*JaUg>Jb&NFn4re*uhll4Vt15D
zDOFuJ-kcL~9p*n=1nzPMWtUiZUqnC)Av|Y7PZ?zI`a^IcK$iJQR%2={9-0!=XS||{
z`(in4?7AmZVZ*Ks`~-$k!ODz`!i$Ni|0eIu)1VTlQ$cDV8=-2gLyRo~?co4xgIdTR
zKve#>&g@)0k(<pQ&``UxQRwH6rpYoLzhWwnG?(@H$OiRenX99h9Cx?@JvbP$WwcY{
zV2ArtXy=%|Vl*c{uK^|6k&V*Zc+)N&rI}G<+!<&OHv0GPqy2wq3ZEJK>kPMNC1&^_
zroC<;m~zzX@+0KkrE(%CHITkcACuS_Z=<FXx$$-|ot=FkZ_w~|k-NB`j%H}FgbYIh
zYRL0ibJ~KH_!&42^t;57H}!b*{Byvru-t^{csr8k&@%5!PtW%9i~U_|%m8@q7rP{&
zdNX4&|J^4mLPa646=g;njL#*W7t|;}9Y37TOkY~ERG>)SG~ZfCfAr`RRm_Q(<vR2-
zfel?Fgg-;55mHkVTIqy7f-W5bLp!x-QdHMzSlM}Ng6aAdKrl|*T&fkr<Nfz@vi)&F
zzN(yCL;mHjr^;h9Z>9*%WtFH4N1Q<dxpvbT)N!!s?rpK3Hr?@prXz4PzlcM4w+SE{
zbqrt;R60F_1oU(^hyWl1q+2j>8Ro_xI9pg-4}7}H2}j8KP`B}96(OmY`67uh%J^s1
z@F6!D6s#}tuEu5yx}(`5>tmizfSv}d6fLdkB=yv~)^TgqsCN)RMOH&)tbqkuHV<`m
z%aB!A^Uvdt35n>~GUXL{IZqSYq&z_o`pgrL*%mv=I@${ZH(I|*8=BZ$mY&XEmEr3>
z#g1ZA%Y4m0D3&{y_%NCLsGWKcJhHXc(S?1Ge$WU?I&69R7lwt%`3Alu+-J_>%g)et
zaJo({e5wPq5l}orPVq^uEu~S9WWgkV#TSp*C!i%FtZ(Dc#su00y?dwo__0r(SxAu;
z(sAUM9_7vSlImQ|fP-Iu#>y~YwE8MzCpvyxVdG`YrmWi0?;JIsrBaYPf1QyY7A@^S
z^8K9;>!G?FQ5v)yio3velWs155xhRo2!ndyUQR3uQa{eGTp!SE3H1(wJ?&BJT?uad
zreDS()`C5urhze{cFMy>!w-B1sw$EqWIv%n3;5?(;Hd+T2Bfuv!#jUJS}+(SMkY*B
z(8%E9u}P${)&;wZaCmIyYKQ=bp0aGbtk4X3B5adWnA>p2F$qdZOHe|;YzH|O9k!@*
zpx2XaPHBnl1;;apgiAI++9QUO?(`v;ldmsJL3PYG+Z=+!9q32&=<qtoQGKT+e%Bd<
zX$wzel7#&<AwEHCBy1`qS<9;aze<d_D{L`6Zu(u;nk5}#((PwIlVrEcfnUXJkwZz#
zT}rYcHZ7^Qts;?XTsvRxnr~D)dfJeJC4tRK7|+rkF+Z=!WsRBlyk?Se9gft;S4al9
z*{a8#_k_+}Bi@+rP_QyEbiRM3s)4!@cKxC(N7o$BnEcp*U+7lGukO6zCgxe_(_}4c
zjb*j?14(0g((Oihda}L;bnfGC!{^O`q(GW1w`9WUS)Q&pR+vl@<z!03kt0^r#*FrT
zOKT+@>CwtBb{;=nx}klw2XDYGVaPhpSInB*`r-9BicFHXZ?CKpKGk>q-J5lCl#ufF
z+!uF9v2-c<aU54Ib%DGxF9jX7OEDC?V~Q1$H5M%DYN2P>XBW*22w^271*!V3_8JkJ
zS=zRHtfR$~^kTPKpj2;w8<Q-;uMZ~|Yiosrm{Qm`RP`!&Y_7JI+O08~{@Nxblg0Ro
zq>&_d6zQiG###6n{s<*DR8W9~a1Fc@-dIbIp%q2q@fp}eaKFRet%6!PzyRl+kv2<)
zq#%QACByb!hk|WrkhQyN<lA*|CkpG5^!xkBU6Yu`83s2(b|Imw&|Rq~CL8=pKg`ee
zn_Xa|9Bk1feU~^PaXZs#Fw(7_Wm85fbr0WxVbVy~Z037*O)mOVi9amN!77EhcXqG`
zT3iasFVx)0Gv;pnl?ZB}Ccx@NL!K+A9U@r_=><K}GE4UBS)n%{93c5TKjW~l=iL-a
ztj`DmLJy2XFB))VtCy4e{q$^m*ndfc>0Bp;?o(6`Nga=D>3`tCkqxO}^|w2R(MwB5
zyPpzY5{XvFYkGs2-5h0CUv)dBO3IL}bzF$xz^83`_osv&IJm|tV*zP?>b>?ef(9Y+
zj?!Y~a4(b3K0^S-2o-O{#K6AsB9j}WnM<8ZMO!<v!LBOevJSiCWGlfFc%x)Ns~to9
zwrOvFsY>i?K`d;X>DA5$UEuSaYxM>gdbr2C-ZOhK8HDAeuU`#V-E7?g@)q<&l$3;|
zdc0FHGx6N>+DAgq6~>Z~IGI|`E`H$UAdAO<jc4ie@-hy??lsi=0uUpu$K}!`pN$sE
zvfg4#-ajBqQIQy@006R`B)jxEHl3kVg1%yMDVPSKCOQiRglx*M0(^XW1_vifOh$o>
zoolEC2)QB4a0zjDcVjUjKtLctnga%X^kQL2txd{2X{<Xw2$FUD4L&Uh10dCd6VA?}
zRbr*3+iYcjqINS<flY$tmHKmNaYZGyg3<ihb44}3<{Fu!+tFwx7_pRdx5dMsUN(a`
zsi9%Dx3|*LYVf#UY@Si7+WQt}>%;L#IWI-b!n)R%sHdk*)-O3ewaJU)K`Sqaxpa3@
zeh|blvVAN+OC(7!iIIGle5V<34>(bh5+xd|VqR0?;lu9ODdg|#(emk4wY9Ykq81Vq
z5;{3KaoSmqh>m8pLYA16-Zy44B>uONM2GGE)CS_#k`DL-A}%#$0`a8;D$#&20FMNH
z8$OL`rFqgQ*kytKh8b#Jy-2AdppXj+Owfy?&BoRy?M@09`3%(dbxW(8p`n=!^lXGW
zgbUA~%X6yR`)u;-5I($yU0!{<gFH-yO$HtEd|ndN0_x-9iTD`2*GLx^7a;Z^mWD<~
zToEVv+0p8Es50*v0!L6tNCfNNjNvXSf%8cMRop13I3N(DnAV)0n4q0O5t7d;@%Pw0
z*xDpv(FF7&1OVSR2KGp9Zul{uHy#bNF7)aB93LNFRdtH!H5vb*xPY4bIzNyjlxxk8
z>USXkAF78GW7$_|R}m8`4+aur{V0cZO=7kX7`imjyQx*u@pj?4ez=D$+v)Fg1h^Yb
zwhyNfV%u8VqLWbbzc8x|GTy4Np&MQC34}?a@t;AmldiCtS1+9gQD1xN^;mLqqb0~U
zQ(V`;#l=eB>(F_qepg#!iRAY(P7sY*P}kXz7EJ<3eD6IBg;0SU$vu2v{gk8g_dopG
zjKJ{yV;-Ko>%A)PUgGaN1G$eJjUM1tklHP{CttmU@{9pU?d5ZeA6U#Z{QUWIH*a=!
z7C}JGVTCLl7sFhC{sC>+QakAubRM*YS-Pij<v6J-35Dd6_WoYgUD>7iRy}uUzUT*$
z2Jfv-rs>3Um5O01cQfb)2o|2QvdUW=sZ-|VPpqXbUk(Ti9C37{0~^i`^a^wQc0F@2
zB_%6AMkg`V<c5ve%Xt5#@zHnd6K_7{ePClj7OJB$ixI{PZxSB4$jhI8yI?6IQhRo~
ze&(8EIW#ap^}u+zH%LKM_1D-%9txEQpM^}D+$kw3KZ{T}s>sTYLIwJ#UhFED9<hgU
zH^>wX-)tDJsYikN*}N8Oh+SFJ^CRiNP@jeUuZW0uZfDNcp3+ezA&x65qm@pVOicQ?
zxp6tjZ~n4uR#cOQ>Nj$pYZ~LU^dZTZJ{ZcqxnIP%dR5`9<*a+c;R8=Z+8KBtN4f6_
z%m!W~bmH{ZXfC*({bs1AV6Py^W~H(xwwF{~_Hi(h-dl?3;>F*UOczC;(Go&GR=3x-
z8cIZTsbh@md#$r*y0J0mb^gh22eUPyr>bfsc=HxkQZUFztHQqL*vi@2c_>$-db@3F
zvD@rd@vUQFn4N04NUdbd-$y~}vy(_Uh)%TmQvp0*K-D(|Q`3V=r*-8VM=ag`xz@d|
zFAfKrx_4t=e4qD8D)jSbgeiI&<%3yT!hUPx0g)UI_+)>I^bhb7XeBQna)-4F*>V&s
zbM`HEUsd=@O}e$;h$i=;cs~D-yf0m*+HPHQtDBmd8u+!4Lx-Nw8*28iKZ)<>w}>1#
zsmS#trs*Cobh_n{FIavGPEPm|Y3v#v#f#ba7<_QR<LP-<_{6$ejkg|vGqvzbvm+&o
zgRFuO8j@E!x@6TRv6D+ZB{2Hc2<W1er()yd_fnlAuI|nUVbufvwlkNl!J03edeiqA
z-SC|<RSc#vVUgltW5M}mK3#W{*HkhFtz#jlB}C+hc6mz0$4Uxwk&I%~XFoi|q9?`D
zQXbmnbSwLn74>?nI1Ga7&O$4-uv^XWcL&iH;gjEZAI*cPo9ewLV;*$2*3?ji9_;_k
zkL69FdG!3#>f(YDN&W{z#fXRinEe^XJdXZbQ;Y6?vv>nPxq|~ire%G<v6tJmoqQ4u
z;8jO!%>C&{yO+hvK7UacnG={&$l$EUC6s0DV{e;ls$FJ(fKQ^ICV<UBZnQR52D5^x
zcntfyh0i+Pai_J5h`c6q*p^xmU~3g=Y3=WqVo9GPVW~UkxbM!a@lK8{4Rt3wI5vf`
zXhp|1mzAYGeG<Ap_Ur=Y;BcY%O!cYvd&+=_Ffp%-R*xUAL9fKGUkmiRD#m^c5R?cT
z5678RSN2qv_Y4=<wnvuLBZo&>DqUZ&u`{&Mk^3}s$Is`fA7Cd|Tj^hzTA!W2gJ%qR
zV@G0Q#ZBI1P&+9f_8-!meZR8v+x=2EiYm;<fhqIz_WX6u^D5KGEy_4}J+0X$Us5@k
zgz)(J`A0@ZOeRO^Zy-j@CSFBRHB_Z=w{>t>*}fd97N*!FY4ZwUST~(mF?vUO$zNI~
zZ^pbzB}HVsr!?<?#_i*ly}ZEcv-Bd`%IO?!1e760^R&Ny)rXDO3N0zM^Ac%4*@v91
z?#f~&%C>SHLM~G<I=h~vj{TzZYm|AOT1PlDTp&v!cp)T-S)+KV=Xc-q+S-lRukT}N
z`ec^4-1tF!UoUMTF+cOG{ZcQ1sABwG6>cQA#-*&lv8Jk&el|`s`(r*T-_KwJym-I*
zblJq<#@WKPv=rqQW#yuOdkZ&M+SW-;x!ENC-O5rAl)uxs;=4W0FSRRqe)S0jPIPp|
zr&4s#!LEtVmr*(etC*N#78&mD1h>mJOH;K<T(a(T7yB2uUgBP5S6IHWtWd@I=XaT>
zXWLFcrk`JBok+gysL;?W=^ZM&N-upwcPRuNnvoxT@qphZ^X=Z+6Ua0sDq}B@i7yI$
zq<o<~WZUGAxOlN_lb>k&@4Kl&-8D)|W>6mUH2&xsQeI_^nu+q4NzmI_H@k>^yez0H
zZ{Wp+`%H?>ZJTv<tV|hjo+!`X)lyzhg6c~&t8EdI!!cuHhV!@g`g1t9>_y+l{FIIO
zjE-=+GgK}thIGySdZXk-OM6m|J|dDkPb<hU{x0ODZ)%j<|1hP_(9IQjJ;fm;Hzc15
z`BL4_&ZhIdERBYL^Ef?W?QE8Y*S!VQ^tFz-tDQEvfq^fq?@N%c65>@JAAg2Puu1sB
z=2IhepMoxY@BBvHzhIvK!STioMR~z#e^hu82geLKrz^)IZv;2u6MZk2iwOUVZ#U!L
zp*UhH<6<MdT)g%@@<Au;8L~q!6VyFm{+c?V^Wq0}Q*+en@bqL(Y|qL_S4oLaoR&jU
zicZ1)V`wq9{rly4)LH02MBR}P7Z3jB)+ni}3f_}Qs>6w*YXt6%vM+IP7GA$@gduBN
zY0~4K$Zw|$u{R;5zOA|qFH?Wxcx&-;V`IbTSENUl`q-=2d|_ZIC4Y6EHGhk1to%c>
z^fq3>XMC*W*Y~t&bsp$U(iIvS4lCz9F*Jzj`aHj~{CUv!HRDx+GIed(v<sb8!8rAv
z<L<=>`gS*DQ`3`2q}7XEmC*C|G)S0I)CXE>AC<@`rhQd{mP7WAdbY(>5AJV#9;*LT
zdI8Vu*Bmqi2{gyO;K1;7;_@-zcO5<Qt0i-DKL;{@zACL+S&X{WNiEBIRvLK+@mrt6
zdy9-BiEp@=a<*lD*bX<VSAL=3ZIW!@>IBtr*9cN~nXaGj)-5t-X0s#?s4V<0Lu2)!
zCr1tHZMak;7fUs)AGUYirrN+RzYDg2j8~kr8IMNX4;`r$JhVJ$og;mHarXIYYYzeg
zTe*6NtA&SaYFs)f%4yt;RgYZ&ku2<(Pb+gB+i!VdEh%rY$>bbz`!xzX-Y<4@12ebf
zr<D))`B{lWp61OhhJ~PW{I@3cqwWl&UA#Vcc%QD8Cc+u^@(b>tpF>ug+ri?QE}M)b
z6xb@@!CM|RGrJ)E!5>EBjn&b2jS2|Rx#O`${XV`&A3`y4S(eA37s`$Np~40iuk9@t
zUrn}{SN!eQn-mnQbK&^;rkef+e`lJ(8Vfd<-Jxntw7dn77{d<xS<+GT|HsmG$79`o
z|45;78=>qK*?Z4Kkx??TG9!=dy$RWSugs9W_sq=R6f(1A-emJT`u_fVUe8PSeSfac
zb)D;+_c+JY6b9NM(>^_lGX9A=-oPe7#P3)_aIXTNUm_V_pzB2!eL}4Y8F304sj2H_
zwmy>RkI*N)B$gU0n!1FKcn9LQoVyk<(i%|DQP5C)ZosHIvWm)nu&!*1om<{jTwYBj
zi(|Dj(I#7mO9)_b65`GqZlt7<^9u_R3(GNwS>GYD?-FZHOCH~M{5F>&e1ckq+$^4D
zQWYH@wMiB}nGL6!(5m}&iA*iNTz{q7bh2_g90XbPd-i{JuYQMFa+POy#6Nv{N7oVC
z*8Dfgb5p;KUvX`1LGSw<)ACL3iqYG@%Is>{xI1z0H+omgn+{v7M_qxk@nZd6`+1N!
z02~?^_%LrtPm7F?t@w0nIx;MO88s%&U&u}H{F|rIM=nRxpYvG$=OZ`OZWm8hvpu{-
zBj2rtJ)|dOZoUqF1RV99*%(8^;k0|E+NnQ=If#y@vwB&8oSt^bq3f{E@pU8We8!G<
zw*K{u=J_wT1Nsn*>I&%)Z+#A@a}Dj69M&Z&9CFDf=0b3oycijY(~3M#Y3Z`jM4|%$
zC2~IH-n_gx>UY@m;B+E(_IImQ_wfm~S=BZ+NfH-|m=-DDU!r`>5X5QsC6QYglV4iQ
zOnY?=(uM`Ey|3UQ^|@OtbfrT4{#Xhoqk>BPr|aeY4M`hh&5NjQkT5!*JE~1rk9_gl
z7>YCIjt!AVy%(FdDIV1|kbU0_JFDaRisz}`ra^Dc>>+zVj7n3oLtl}L<Tc$T=jz<P
zII})br~|Dj+`tyIf2PZB_x0xcws`E!zK8s1ZdNq6C^U#XPeq57K@aXWgiN!?CGu})
zh$x(*PsxpMHApk%GzR;-q!8{e3{TPdt#W@l;AcVf1ak?DO^;dpqZhxxa;uY7YlGKl
zpzT#rEkQv0!ost8$Rd3=C!AUaSuFgYV5=06p{{3SbQujceL?{_YbSKhx`nn>#j)DJ
zXce9mE0u5@boy^`63RSvlmp$QsP>Nrubb;S?1(tiE1~Q|3F~{n#T9gMUcr?-P*&nU
zI$lX8dLEi(=xWA-E}|S_4>?#=6x00t#%8_Uy0yzlrVEII14CF;G6b};T3ZrH5aU0#
zBD~FV9OtQDyV3k5o6IeZxPmLQ&}=c_zTV~!v2obWy#2;NbK%l#yPs|K0}S~2b7Xnh
ze{1OlF>$JmuC-FBqANJTa_^Z|pDwVQ{aU-ZypST|$-&{X7SCA)l3;ARIS5zx?)~yO
zAGU;u0~nT*lc9=gdCagxu4VW>lFnQv^pv|H)LNl$f$%UP%uR~$mNfE3e0S82+tG~t
zL;vj=5(Croi5U?Z;RcNTzu!D{Sy)C#zqUF2*@XtrwNlt$ShdVA&dW#BC8&6LF$V@1
zJUyKtW)`~J`H=nzhPm%x%G;jk$Iv}m!CFRk>*eJUQ#}J&i1$Lz@9*pUdc-YfocIn2
z3EXR(--Oow8=A>}X(w#YJT@()dAX^97w+BghkF0Gl48slFB5*Ig5t0NuC+oQhs#oz
zjhR_FBt%d&lM%h}rSvK3A1vNGFx9X#?<XFF_;z)^%Cx%!`CbsA$myY~McPu477CY<
z=to@U^DRJa4-aR6D?eNC({wxIm4Ex*m}u@&z&*lSN5#p>JK}XPT+mmSgwQylRqsQs
zE_zN??D|e>y{;z{DdJlOsu8|8!d-ihmQHbJ&I5b??8NB%PIEI4B%1O`t_1~)@7~D<
zT*ah*7Ep5=uR2Xk*5H9<4v&gguR7*i(p~LXzC%V9rN=L*`*FF_<n<E~Ud<cp&W9f!
zg(KE(58fOodK>Vzyz}njkt{N^WqzWA)@~>*dlIH55DF{3eY?33fzlyctevft{mXC8
z{cwtRN%!nlR{P=RYs>lhhlSc(m>vAWH-E>5k!OEZ<zcHv$#T}Cx=9vgqvNahh!pjq
zgmxW%seR!C${n~XzERpt^NQUKs>z|Z&i`5y;oXmS_tVla6JWlT9~{Y+DTnD)lsfMf
z7q1eLH2dV~tt(cChkFGFGaeVGq%hr6`YLZD=q4ykPrvi~B%$8nej3f_e+9*%j<4W$
z8vRb%S<QlrIi1Qkc&lm`$}ww&Ixp}2V=F=BWl*Ia{V{GZ2-plE8J_3zZZR&R7)*A#
zH7Fr4*U9)dXP`nNLRyaQhv%xjeZcoHKXE^=cPL{dJUzEDpj9IAd&6PT++Ql1$xD<>
z!m<CwAW;6Lk9KQx)(V_f{HQiaetG$*5R%4pxtrEl<fb5E=(xOSfn+fiTdA?3?1tGR
zMKhsl?6J`ycBlPENaX6-VR?<^!eJbt=o5s!qC%mV^TFj(_k(}VtF}VOh|`rDTm=?~
zzu1<L<=Huc>x-IOX0b``YdP%n#>R4a`FH6<6F(eCM9&B2uH4B4yFLqGJM17nwPZGX
zz(`phy`4WwcMxy38@gQA!hqqmd)ag#))($gCOS<?SYK~+BN&ValT9<t7CJUsTGLFb
z)JjTB%?FQnnLfe*TtmNEj6B&KZZRn;d^V-8;A<cfpx?B|cqW5%sDJV#jSBzwoU080
z;b_M>P>)Vqh(lc?@wV4i9Lo^@_>F=>#OJrT_|Zldsc7TBLdh=g64qR9a;}ri)DeL%
zuKB}w)VuFb3kw!+-b8Y888iMJPiYmP*JlqxXCoZv$w@LMTh`qoJAx43)q-FUY5Xw7
z{rF)Gs0098T?PYOvmlw0#LZ%-6>ZE}gaU>W3Blny(Pf<=>%DYMEp==G`ceC?-<S*j
zXQ%Jma4<S~A@`J38eOvRs7Lr2?=QK8Z%Ks;H@lsM1j#av1rKS!<rcl>TWOzHmx$Ho
zT~nj@?K6yGk9ge&0TLSF!vO2sA2-K)E1cw+jwd0x+L@(%NGm8n>*qt@We4rJN#aTQ
z)g0#P_p9?i{HGx$p!RMF4SPvf=+^IX7ERZ5(&O#=q$Iwx=g%XjYaea0H4D!)i>WOh
zVPUg*>7KZb+euS>+gpvnuvSRf!r2gi_#ot=)%O8b+)#h0bg~T3p|BcKJ~()5XKQ)`
z9WJVIu#HHCHVC=ZjY1UycZxWfEga|uJj#&<wdYS!2|iHc<;J$mbMy8D;D3DNSU;i&
z(o%UvMJHSCTMEm|F}M1>-0JOFHpX(ZV+tn<kzR`CBbnKTg_m}f$Z-k}?oX-EW<m5L
zwkYDsQ2P>#E1@L~ZB|A@1Lnm~1VL}h$VvTAT+FVPV(g8gpHE)+lb+|Lq1D(jpkf@i
zVOLX<CVnL~{<=NqA*Sag@p$m$SKZ-luQi)Ltx>6|JRRo|9ztm0DEJT&?_zXaoU;c=
zF6QNR7BEjG@!>n7?6!_^|6T0(p&)=NnfTxpS`T<ixO6{)t<sl$=yWu-$n|Us>zJsg
zPu#m;o&%I6;8NU;=Uf(TXc7r?LcUGBLL&!Io7V!}|6Y<$h==Rlhw^YDZxts9AW<rE
zR%naduP#m?`A8qKv6S9F=tC22a#6j_LmATf!N@6SOTm-Ww9d@3FAw{v-MM|`7t+it
zz2`X{bQnJ!64}tVMwb7=ar=FK@?#M(8>*o`iBWo!X6fbj?MK+{%#?>wt6N)9lx6H5
z7$XQzU-lK0m$z=O*tt{uG)2KTQP3nH;Fw1##E{{2Z2unDLQ=W9w{K{LO#Mbf^Cizo
z=mzJ{0{qqTX9&2A#vknD45HpoUeX;;PdX&0f&@`T^}EkVv5UiRxN#tdydfE-O9Qh3
z&iABhl>egZm#DHS&0F_AhU>sHEMBbL*iYe9{E%C}eMY_f^Y}U9uBl((VwxJ|qxFLM
z2MXcdTX}EnwM{V`L)=!&tG?GeaS?CS04eZyP0!Tq;wAHNfh3zYJ|OoD9r>tL=7;sR
z3>nK@lamL4C+J_|6p}r?xZKtI0QVM@$t$Zr(){S1tV=kS<P%!ND|rgPkTTC%1|Cuq
zo()K0qESN@O<&Jz>R4y=Y95L{Oibl&-dk&=LK6dXVyc1ls><y*1gjhTQSAZTbb%jl
z>PH&-ZHWCgVsfF4s=xeZIy48jEYaVbbgtI@_vx9?y44*?e)h5)GVHV}fbn^#y$Zgr
z);aF|9_V|$vz07jmEo;6H9Gy&&FW~|qZq};MpmbK;4adyYpW#N7p9jxLWvKK{H_H}
zfj_+;2E?h7hi#{IZ_1j<Z_=a7KG<Gp8tiORQtKLmD9f>EjuaUNB`Sp&`9y+ClO*+C
zJcoS=PwvhKgGaC#!>&XecM}4-HTx(QtJ@$?9M2ayJF5QgF!t2}4lN6UlliT&6xR4;
zMW}U&!RdSLm!*P7WTHA_4NvnUUf&CQ?0T4|Ceha3%}^Va<fFff>{ZRAre~*;O<K)D
znfkkFw0EL)9xwLE4ETQqaPOoF9qBnc6FB=v3Zf^`O^}xd#$`?98`(#DFYTF}L|a#G
z2|?-~LT*-Xn&5(D*?@W=F_BKe($2jXxmw?4{?p^{ANVhw7HmB|pZj$Dx|j5-Lil$&
zplTl)0T$6XrJY1X@<ot07D{n$FLrLNG1<y|4UgA|49_1Sh0|mKq@rx^!EgDs&Q6Q;
zs-C{+*1Fw()$%U$*6wNxS(w{$wOgFD24r>1`Rh)53^ZZuC4q|jvb~PuNp@J%=^6ZL
z*&b;e^mqb4hn*qQN;h|VER=*GRj=Le{=OXkF+ky=yRrPY&^uN3`%8Bxx5V1t!@HWD
zodm;Ul^}`ON?4hJql`>PFcEgPT<zRR0%5p!m$><J(P&<M&$?G9v#$|~beFA<mB#aS
z_yy_OC`m8PE{mlfi2~{YXsw`)@qe-{w<|0At!?rRPZ{169Xw?JwlDito5LfTzO6&Q
z6oP+tb~gln>!dgM0y&YS1%j^m6B`_f6MBf3li%1FRic{O-rt2|0qJ;uce)32L6sRY
z3Ba>JS&e2lP0shh$pu;(WlL69m$C6;j8R~IG1{3;n{DiObuAl~h(;sDpBN*+R!6<c
zxL(?^5XmGLspoEfPjgB52HcVF=fcYCpDf%~8ymHW*=A%1WWvi&j>Y+^sli0G?M-nw
ztn0?Tut4fCQl}r^zFi<IpAAVY&8zw<5*Ag=VGVO&@~d&0-3K)bG%;Z<Qr4XA>-*Ep
zwR!_BiAg*+HGo)Xf|c&MZ@CO)^9KHhQ0PJ-+j+RA!qz50+5#7B(yJL+H@w?RX?fNq
z>-GVq`L(&VYw?drKi9^@X=P5ieII0#l4ygbPV>!X&zWZ{nRI7`ALJIJ%VbR)t?4n>
z_g<l)4nyXf7?=KIgwqF(#Ov;RuU-`=@(-hgRw_R5{fSuaY{{OcjU|(;f3!4TU?y4r
zTd~gVdp83SW6RRgRqaENf)5#wem1a@_o{aI8%RLTX?=UQ(dboLQ<El_5%gAazexOa
z@+FT5@h)Yh_5~SfWON04yFEQ0B9@5M6eIXa{iENy>z~Pf1-q(j2|4W0-Fv8e(rtPL
zM;;ncKK^nRH<;q)w51Up{8_+F(6uM}3=8W@+rb$59$;+lK8o>y-9EV~*e-;uu9E3v
z>3$G`$19eKwi!8eKH9(NYB{O*x0g}ZaJ#;#DW$2$t6WE>PG@!(97Jb%>Fva|yY*kb
z9)h(*nPaiozQO#o@#D#teIkU<LI{a`{>R#y%nbO*K^41W!nRg|#>mg!cmG;OsaU3%
zIUEHyx%CGHK4eKf_&cyY`u($9>_SbQivvw}?53FnKx$*T4{nGPZ+Xy#S62jIT^y~(
zi+zY(SYKDEtE+;7P9-DvX9-!4r#C8N2)CS$a?l&uftIpoE&?yq87Ca5sih$pr1pgq
zlF57G9zF)YfU2!vz9t_{?hn*@`&;J+qZf&ZuPl(Db2B`%_voWI?6yQmSwh5wXg}0C
z3_Gp)-D>c&LpvAlJSEb%b*V^Jt(a3pfO0<Mf*a#k%o!0$npg(+u;aF3N_Xbs_+Fhc
zh%h<I#)1BLT-F(m0HC{GCtu8o5dDC4A!(#i002Eu=og^R<Vt&pKk{KHXTY-LJnnu$
zoE77!10@l3;V7vHwm31c<8fJ2*c@QXI#4+|TZQQetc8UH4}OF~sPph(kN%|7BOxN9
z+DF0DXDOImX}uhUd7pHDKW|vVZbkQyM<@q<IP0Z#iS^*}?NrN{b)m;EKS2pIxlbgj
z!)=k$OX1P=n+x!OZ2%k<Czb#Ek4m``icPZyI%+R&qXCINipqXVbiYi<zh-`o3`-m?
zrTI@{bWjcx$|yWXjJ%9zPrgw8veVuf`**|ewA*&F-T}QxZGj4}`}g1cyAQN=nrA1I
z=+aJwJT6rLiqJ`prV~u$v{|{?M`1UxTT#)o(6$3Z651Cqh9rP1ZW|+dtQd6%J-wi)
z(lry+O^O~Lh{-D}j+Xb6Q*;{}Q|=85FDC?nOj;f@!}FIbx>>>U_Bcnx!TtvQ*H;d$
z=dY3X{8k#Rz6^p=OHp0$z5tlC@k{wLBYIDKX#ab?D$rG8Njp{fJo1q2LEIP-7rCn<
zIpCf6_+2|BOKxXgo(*_cCvASY0CWS#u~neG*FgW?vG*@_;mnN4O%dh$h>0mj<?Y(w
zOKBf(s4yy&vU_G7+V%I~6liG6y@Y50z}~r&r|^+0>?EUAe9X&|f&%wR&Lc2?jE*o)
zU#ae{1KJ#X;!QMLFgDhcE>VaL_m7?N`vZj-e)^TY?z;S%^^>(QW&WS@7a7(&1crmD
zB_*w~K{$|B<X~CM2FD^(igm8W+P=KIX-$6ts3A=IX|kp*`?9yZ8dq*jJIm)uG^?+s
z`JIQ|G8IC!Jr^AndcZgF>~fus6|Vn6EBL$agt6#&C{7~L?{F~XU(GV@eWzG}-P*l6
z6(9TIUF+bRnD8gph#vsklasbkbIh8I(oytL^zFVH2zf}{d9Y<dPNCyVo<|3WdCwBo
z=TG;zr8G#F8f4it%U(?)rJ(ruEt~=1(#5Iz&^q(U&dF?}_}{-Twww%~J!^vkX=>{I
zH_m`Kypj_CHm7m+Hcf&9)HQqnbl1{1tV(9NqU%l_Y@Y{`E-mB7$;}mvok*mQ0ZMkd
zXS8-{WtDYj5(x!We}6M4gZMj|tBvt4ND^PAB8rQ5#(W9@bAMZl7jV7CdDhlzK)?w&
z?oP(7=sm%XiSX8M`p9W`Q;3czIf5GfmNK!rLyY3;D_icY6PefP3c@C_$hcVDaBt9Y
zjdxynzbzUX;Uo`lG>!^WUODykueVtxTfu&R9M^7O4S1?%$+ya?Sa0&Y+vo|{=RH5j
zq%<>o#b49Z*QBPvr>dzfwED<bKY5zW<5E@rXTs|$FF52HTU|u13XpnM9@dMx^~JFr
ztql>Ro0%;4JO&suUMzmW45O|sf@V-Er0m3Ci{L;3s?in8?suy9&9a0E=3!uLP0pYz
zN756fTHA^p$gSm5T-kW5lx2IPWmF8BZ5#%Fmb9Ejq0|D{QdLES!_JWVw$1`Q^>n75
z^ZG3o_q0%9I!h_z6NxB)Kul<e=KZq66p0W;7jsRMgZn;>-r{m>+{)v4;qQFQguA;y
z;0%(_f}Y~BBUQwQFtWQ_vors}rG0R~sNSGetgW*odc7o%o%b@NCE7A2esF#9`?Q#r
zQ{>70>tpJx=dTP}#5eM2{&Yz$x}ss8hE`RrH#o80nX}Q=X?BNsIX(&(hiiK0unB2t
z<%k*|B(WR=7osDGd3{VSYoTI>L`b`y(1)Cn20g!cYUE{N1{LSxzfxg)&>x8_IOU+5
zc6W$KXyT>kj}EuvA6a>-MUt1zR}O})Xy{Vk4G4qSMysJ&0I|S6EBsVPKV_(7wrs90
zp&Shf-BQpX#aA9mij7Sh8Nwm5Dp{$bX=_)1f&5w&3f%umZ``xH4@3oRR<h2IsvwsJ
z&IYJC)F5nf*NOV-1lQP4$isf)cZCoy`*ztpSV88oxiOw%t(Jj0onFB&Eo2@XG|(IE
zE%)Qe)N2Xq!%@URvWN?+p5@ts4pS6@#C~;u86&_#c1I0YYaHWPvutTk?8?byMvyH{
z5xF=I2)xrNw=6<P*8tyUQSu=?Kn<E#)hU4U09Ck`NfS9s%OUoIWYQnvbHuOt{^9R(
zCU})YqBu3@pxsP&xc&}7tB3SpVj_VcsnKw3Wo1>!Y%fQN)99vYy)3g7&2Ut2&-1qV
zGWV;|%HtqG6n{YeSkU9DO!uH=&MhQ5*IUjO%RRV+1qFG_&hC(0Ka^VSCf@Ly?@A&F
zUU>#Cv#60}?4S+`;h&6@lt=fKi6BO->D{x_myz+|CZ=Me_OjiaNLm@O_{Hxc9j00w
z*v$R7D(sa#$ycVVoYHq&c=vO!w|%m3KD%XF)*o}0cKe@B<8+x3cB;4=6TC4?X*r%M
zzdStQTXwH^-W!i+U(1Vh!k{g;b-RB(7>iGVs<yc#)HQ)Rp@IC{5iAaHRTlt6K*wh?
zG4hqwxY3j-X@KnAFPII?aOcVG3qoYA*ge?WrWst$G9d%8vB@8n(9g3y-0AbXT=Omv
zz+fb=kXHVYCV3JS`v%-$BYzh;^-CWzW&?+hb?+A_2(G@1nTJBb4tFW2T%i#;z`AJI
zA7}DZUcK+Po)|7&pXgt^1_DjMUBmAt_-mX8Yc}hP2~|rH+&c>NcAS;;G+#a(?+V3>
zb&2P8^|NScPRj*yk?9FV$Nr26$b8Z4e}Y=^6ai%9zf_d@ga$7l{dv8^jvG2(Z*N~_
zqpSM)fiWTb0Gw!Yprb`aCvkBfQ3h2M5gjkbeJyL!oP-=xA@szXG&;7Sp-fko;V>()
zgv|cAz9&v(?GzhnVK_E=?~haipYM#amp4#6z4vvrOOnef<$u+3J;}Bxwa_W4JQV7q
zSW3euzzM%g?|N2+Ci2LC<Bv$LNoZ9af=z~yG&7mI&}LsH@H{tH&*Vd5*CJ_XMRd$F
z67DXUYzq#6MW!Vj;BZ)C;uF5XpbrjVd4^5E3I(We5V5s_Icb5-ZboV;WtY@{^*f}(
z$A5;>ah|63AOBGOq^Pc*tHQ&(4~*NJX+H6Hm%6^P*+WBg_;imHh9x=z|JLgr8YW>E
z7D*L%elRt4MXrRz#XU&sj}=6pj0hMb@C!&4yT-sX%bHHAkyEu?-+CCepOau?vn2vP
zvVv``@Z9;b;l^!ivtKdZ<UJa{Dug2k?%<Z=3j9<w;5@MaQHMV+&E>&Yv6z^c7s=3d
zmKYe~#nnN$8sR=`7t;n<zAWkJY$VGcxFEetN_4}=V$IiGN9UT^q_7DnOKaq=o-}@M
z9(1Nc7lR(%Pei&5I1Srj4xfUw5za!Jd+yHh=}It}$=7d&_Cg9e<+lJ{HzOJ^v{@Cs
zhsDx<E{^C?PDx3C8x!v01BF?vL7GfuizD<141u3{20h{%W!``Ib})zUX>1C)HvI?b
z;_fcBp@16T6r!&2jCWh@*RAWDQv9xWKqa}z^$eK?F21%5KeCt{&u0@_9d4-?dqhiV
zXr#inYeQqXO0O=ZRtJ*vb8_aZBnZBT<ZY!`Vp0kwU!Fu|XG2>ku|>wP&&hL}n_{u2
z@4daO!AcO>C!xGLonF`7$NV1lvhoKOXL<@1t&$>s+wx_u@$vEYIWyxz^BWZ#Gl@?n
zB!qfeYHe0$kHgGo8`Ra+CsVK$&34qS`uyWhlvFyP)z{0HtV$Vw-ztp`RpwLqDALUI
zlZy~rpX@VT@4{cUrw2hVqDb@dH-{zE^Es+>9OFgnGvAs4ryIbialgUGbs(thU1aY8
zHV)Et44^NEtvd>Jr_i@R41nu(a~rVP0p^J*EV&>c(v5~&?cy*G_tD!T9FgeUo*$_N
z<wxgBFNE0`2Y1kYc<eePVsY3#7@0`w9rz*-gOke?3?8Yg{&*8ef_5J~fKxVEIM@3q
z(2K4b@^8=~2?_?>cmC<l<J-WH&~$>#m5lIasdPfF#4tf~ORRhMCJHql<KF>U4vz7X
zwfHTqh17$nCYz<M>y#$OmQv)ua1}+x+6vzPOp<Es@+q!A{BqF}jqANIzs(A*GyUCw
z4HC!&%IB)eGU%;KvOvq#ps^;q`c4SCqQf=okM?vKTloph*Q{<VLW{85(ij79eq<`K
zsVV8J5#9CYQDuy(14mF@ddlTYU;5u;dE@UPnimAt3aECz-UsEel~u1WV#Jy+2oJ_x
zIs=#~9jncsE7nGoG^HF3dd$A(ti29b&tMX-hD6)aa<Be@LSm);TSlk8yp_p!;of&v
zbsJUHmIs2`yHg|a=$?Ff{*K`gcd~ALu!}w$Ic?=I$j)xLljBFflx9mn@N8m)(`>b_
zacx1Bc4bdC+w8-=(9@B-;ojCOl&NyGVcIV~p99}p^JPv7N1?l$0CGwUw2YVLMsPts
z{%X)?xs=wrycQ3ROFNeV>oh+dm3}P09x^jpUTcK`QnU!$HyARE{^4R-S@Io}$pebe
zs>`Jc8TB|Jc@j0K<7!W;?Qgy3_36mHJf?94MPIRYT`+kSi}o8T<M^hu&%UIO@X?v7
zQZ060@9ux^cU%2gIs{UwoKEq*qrqg(Wp|^h8xOS_jyc}*TZdDkqO!g{UOQN~JW#L>
zkTjJ2J6<nla;5@mHk*=(PKkxPxKHOGN=k63DYhzyc*B4j=nLU;+3;kM1q7)Ji;+`Q
z*rjX`Mi|#J!KetdV!9umpV{aK2c4vVTLSbR2x~k}nbjRd$4f|ZYh4`D7|9;^2Z2k6
z_Du7(Hxk}YQT}iHE{Q*PyTsL*cq$bExZsNNCAU}hiVVzNUI}Uz<-><9MS3m7yLjMF
zOi(*yra}N+-7=FOWadh|fU@@t*m(A~2ye78;1WSMaxy}RC2}5_V-hAV@-s!18hv?C
zTw!wLB2epNcMs!xNL=$A{(HoV353<1J3Vdpq!4&Y=m_C6xPjL6&=X_N%0~o*+P&=i
ze*}^$W1#1=O|$oD;UDAL+HM0$pt3XMFA3a_S*l3y&X5k62PEwJRvrFS-Timo+n4Wv
z%g(`8?m6}}jrgr(Zha!Q^2V>eu+J8B-&rc3_$mcXT@4LE<B~k`<c}YJ0c2~xrrpYA
zq%C7P!+gEpQ+fuAJ%^`OgMNJROO;_})p7W&cb<0P;d%XQd~BbOw%(_Xc>334U6Ie_
z%#lF!iL-Ojb-6+&ct;XkEtOVesQ+HBm$T);f113SAIYoI1<{KY8)s747(GcJ5u=Mo
zcau0yKHHx=X?^PJ>x&uEsjzS}I#DAx%W!JI`02&DHE0N0VPP13c+C$4T=~&pJskx)
z<bU`__6Q!R!f~<+zw}ez(sDT2Iq!~Qdzp$*J!vWj3FBn>RwA4EaE4SEd6u3p4SLL%
zpSizb3K0>W;zt;)#u26?XoQZmQjogB5y+uOl(o((s**(@Q*3IV0!s&~p_m;ZL&L+F
zdXk849d-4Z9RK0+uN;0f;qxr)9<Ye7CDC#%bdXaT2<PK~yyBnArUEoA6x8Sh9pbn_
zO2WDeaT6#J5pT6dLD*S9=eF7zvbNmYY<+|i-3fTYzh5Nlhww3-+YTeI!hg&flUe>$
zI$s>Yh>I@1ycx<T8Wg*lGMKXU=XX>gotxkH`;amH-bp>{i2U5~V~E`|5{HXaAm^=D
zwvd5;WvUMRik-5IB2Qs=md$4c1|FfL%9QAve?*!qM9Y|!-@cuOA0Nl0oZ57sOdgzz
z8R?!sHY|XKcvBeZ%87ElV^9h1D?yJpCwVLq5j2(U#;0N?411|G=mcHj?s}LIq>2Yn
zm{^j<t~QSM*zI~#x27+|+&5lj=9^3>bfaCbN#HnCUm)Xog1y<;DwA!7XH$b`fg+Ie
z1mQ&=HPO4ejTs?*@5^^6fb4oH1e0NyS!Cq?x>p#$%OFYqw<MDy+H*}SdU^sIrd}Uq
zXhPIF<p;L=p2KG^Q(go=E38Z+{d+Q=zX(M8IG4xUw~8oE!n0R>-24_PK?c>by6~a{
zXxNIG$&y3+Q!*9t0&<wbV46e3^}lBlt}=FFo-|n<xS=$vR@o2j^MpYLc(~Ub5TN`}
z+QIheYlw=JD3zA!K?&*w&hGZUxi(7f_RjjOOKJMowPX-FYvkd~ebUf0O};N9irG^u
z*dd)G6TCrTS4YhELD8WOg8lab*Pnt5yLhDhtCf1flt8yTGDX-qE6r>6M4k$rpjTYs
z<TNToJc4sZk0SLC&{1Q;9*|@{53cEbe|H~gLdn-;i^~8<M`2yK!InY4?Y2eb>6n)_
z@e3oLiZ{g)2*%Ieaz@QPZa)m_+5pM3vy^T<_blMIXSV&0xXM?}V&DCCPYN4z5fqp^
z(?8CCuOnjO;@Rjm>r(X=7z?prlMaaM$!qPUJuA~pN$Dl9-}<1YGVXt(q}9I-1g7i*
ziQC@ZM6iWuv6GXP-VeB@v67uECTH7iW>#xMU|1kM-z`@9g4M)kSWyl9watof`?>K1
z(2T&L2Fb*<e*X^NA8<XdHif+<-N~0|{LOD9AJ7izuHlQk5#EyXx{XHv0kFb3H4Xcy
zuy9WA@}qR-6lll>K4YeeOsmuX`Ezh9exXbJsnbJZ4SvuuywtuWZ@CvF`=aeY0F!8B
zhS@qCgA|{qBoLh?avNiHPLnWzhg(im$6j0niSCOAaK;Uk)d4rhEe-tZrR)%XTeBOH
ztd)`HuE-ImDrjMp992A49ZdQN#s|jKvvO$N%xcQ#;GiuWNX+m(iHcL5`Qm4V<nvbk
z;0DTu;bHviE0fVFK3;n3$KYZV1C8=?K+%~TbYeSVA?7dW{x1J-LuQ>#$`NRj*rTv~
zn@eyK8UFVWxz{|H`T+V3YiqMQox#c=@P^uGWy$C2kF*a(-?xpj_q=Kj+%UAx=YgB>
z`hv;SJN2?owB@_op=)Z&urhddv<!)CNXU8YtJxY{5fGjtQAw{CXMC5PwFm@ycL~m7
z!{f9hiOs+DGQ<)XM0mH)Q|4@RZD~wDR{W)FX!s78(L!yU4D*vOUsq4N$)e@oVV54q
z>7<xXulW7%O|K3*SlPdPoq$cBXYx@77-k~Eq%QGR%W$#;wY+N0KoZvjdb7h1uliEw
z=g)CzG*W|Vno_=gc}Fwp^yqrWbBF{7Kf)&aFmu9`5kOh`M6GF?|DvH%Ztj+VG%3Oz
zC!nh?|Ja9iWQtr_5GbS%DPrNbfrCZ8(|q7{eol;q74>G2{M~h=9)m{8bvrw$`KYO(
z?h|<Y7~GPGyt@Fi<TUAO*=#mdR83SD3w&xwTn%e<8(AI2aLoX{s>>1Y+4(l!^;?Q`
zUR#O1?2c~jq0g3m(p@26rtNz}K$`oy+<!q67vyw4V3Ai-^G3E;xgY7*GXD`XH#bJd
zEelw5z(6Z2$>J@rAi7Gcf#>rzh?<OzV|vbNLEn}6x|a%Xc~={@r4^2(mW?u_O#xu^
zm;1d+JcFr#Q18r9{2F*bAxerG^X_b?@%rk*n{3NhrSUjpk$xeDpQ<=p@maFsv-u5i
z1YU@DNXu&wjtMqj0~=zgFXapOf#NDk*L(W4eCotGU7^7=;S5wl*RPsjx8%gyDp!ty
z%_x9ajh3sLlj690iww@v?|Jlee~9~hEX7etL?TnX0<}+*kw%xUalFq_b=D9KfNq=5
zIR87)DA*u?dedY)A(^os;xRO+)`b*_@c#6B-Vghp&4Vec-veZCW})S7d5DX#im1o0
zUn7ubnH*{dR6PTjcE<uTt-$N;Vvh}3$JL3E)bc6+_=AF6ld$en)YnUtdjj+YMJ(iH
zlFCd@l9JCNAhl7PKG)-{1qrmQC6i=1Ad8L8383Qr7w)Gd9-@Qv)_pz0irSqp45w!-
zBsoBIDXqs6NBLxWTo9ct;ObW&^WWVC9EjHveoqCRe4gl2t4Oq=`QE=?m8`$JkXJ&^
z!q|bn(EgQ2i;}UCBqzcfn)46T+dh2Yh9k>Q(+)yjvsVR~tta;%sNNZWyUK>`c$oK4
zAwDu=Vy=`LnqszPX70Cm7y4f@BoY#iskekG_f;o!iR)YAC<_|-wHIEyX8*hT5HBX2
zq9FBsWqp)6*R=opV@=KJg0v?ytJF^rjK$FT(ec|2S+6;1`pj_!y27}QR+}zyn5I$@
z7+7dTdkmc%4`v+r)R~)zR|d0Wac|$=R{WLr<sD}`Mu2kA^`a_bgG9;OUHs?81D+C@
z)h&Z~5#@PNN_5O#NGT-XY_o0ec8~+R4*2&tyT{0AK7iQSidpBwfM?%ERAee_T&JHs
zK+c=yhK6EsIB56%R#LrsDP)-UbLI&GFV>sVch5VIrdsG!_42Y-r!YX7&Ewd07z~Cl
zUv`)MtN~~hmO(}){@>an6&~T47$;OYzmyop$5$n*(JnpdLr?com6b%`C3sW6f2Z_9
zngMdpgv3rLHS^3*Z1xsS)>CjVBx7Uupi+v98UJ^n2d<4K2HcW=f)mXN-nvWoEi<^j
z3ip;=q0CuX5t-@^P38tc^6@~j!-p6#s8vK<h^s$EGo#%M)@$2#72f%`<p{A+ygojC
zx@BLu{3S9$bJ-)G()t?WEy!geuTxf-^6YISWO2sFQ%n;*0ooPjB|wY+@HUr!BC!%D
zAG+(vQ06F2%~L2!LwMiNaMt&$-jbCGnxTs1B*hYCslZR$zw?1u7X^lAI>le!fqObj
z^rYNEXE&Cm$=XFosJ{L3f(2+I4p?KErG+%{H-M2)|6zx`V%`13mD5TXmL=8usfN7L
z3pH_z)lI+$!FQei;_}JzU9!*JOFc+=V7u36rkPob9-g6EE2s<qYMPmAp?b04(knke
zhs8Jl$N;BD;`$h2WTckoHQ82+6dRP^0u|No{Krp}X*_bqoLN|u0#<Rig3m8(czEO&
z7iCI?DA8kDC7AM$fs7%j&+m`w<Q5I0OZ0=|F;9u_4Yx__MzPL`9FPH9CPj!78vH<D
zMw_f8^xS%SnGW-9!ZX6X(e`%xNqPX)B&bRCio63NM-Y^t9YAu7Pl`7DGGXGE<?%+Z
zq9S8wXr+?hxq!#zIF$cSByOz?ZvJ@mB)!^!*Z$5DWs8WZc96VROjqVjamEl%W4Gil
z*c>8~2*$hUx3Xg7X(8>HX7e^k;aW|vI7R3kZm5)2RjEJoq1Lo0_ha8(8~uKs&zF}5
z3HFgCfFe4LUN7k;Rm8u_O)41eyimi2?cFMlb5S91@k=#3^yp|%tLc|SB@#GskX|V-
zJbonREKT<Vq)|s*5v!mtcR{xfFCi~_d(!@0YIc=1ab@VzNvW`sB;xuv;i5mgM`zs5
zsm+=j@<Z4uvex}ViC};NAqDz7k-I_sfCzZEfF!5|0*G)5ju3`*csx1m^bhX|nN%^}
z{D|%QmG@JiBEp-oQ(U2ND=JX_>$Ib1kn9GG>16yYc1@F>=JTN-{rU-^N-fp(G|?h^
za0Z|9kz7Ngv$|2ZI{N#T9aa|7e&V43?wrI@nzfYz9NBDB-Nwts>es;RJh`LN4{hh}
z!>XIOykNMs5~e^eYQ1y<_tdjz_x{jziDx*ifezF(q2BecOuH%b__UVWUZ|p@KdCY+
z!i(7Z3t)+acR%SY&JQH#b+>>;2>kFYtaLpUCp!B>1FWxQd$|P@y+q)arpKOk;;6-^
zRGU+)v&4qGyX^W?1QA|2CZX=eTPuFpO4VjcEy#@8+HoNJ05Q$#N(I<VUj9Cic7K?=
z#1|c5@eenRsBAFG_{R*dT@9lCF)L*pOKF_e;Yks%M$}Qfiv~TGS7K*V0+wT#Jo54<
zK`?V4i^<FdJcHg3fL+2|rt)rCpBh)l7NU7?lf#4<b>~y2#(+2-?VXm_R;|%kI6KH@
z)Nv=K&cdsoP%d@bcBM}?tC!;25y%_kSZuyC89{K_g~sI<%WOXksy9C6L=*<&zzb(X
zPcqWDAFb@V?<$FA{MXbdrkjF3d!k_PS4m6KM->$p;ZIAz^OFWnLXiTDfVhv#F_>`-
zY!ec4s+vsqEGm9?4jIP`z?+H$%>%PvuFdjxZ>n&vJ~EF?VBdB7mtT7WL-7T<sQYKQ
z#uwKU0^HCAa@4Tv?YK@i^tth*rSlmQ8XJvSZj>sz-~C9ctc?g1c-zyiBU<a{B~u9V
zwUq1Jq^v}xP)0Fb6iX$@v$IqX<ffGl7C5PFt>m9bm81P2$Iegpd^|&j6a7Zf{gl9W
zxo4BgTER&GL%-SS6?yZUv2Uk4lb8|5J08uK<EVkhC(xS}6daZMMmRA=e(VUHhT}ar
zyDBOnKLg;P?cua~_QtGC5{PDfh67u;bwBGb9Q!di_U?S9b&|C{x9^plZ^+xD4W>$k
zjhz1Fc^vT}pdkv+5#viVSTLY_jSHwOE-(OQ!3G#MU$Nn(Z=5W3m{z|lY33htcyY;a
zS<rAld--<byHVPrBQRc0$kpyBIgm{MF|Esc@95}=>-w!Xjt0NxKR{WqM3(kGd0cGr
z7&o%<(fVbUXS7_!+=0z#%Ha7VJUKZzCZ=N*8RO$k`R1qQb@E($CA#HtY}uvyHJ#4%
z^7z=`Xu~iMUsct<Q-V08^1tCPljX^|%9XcK0MsCljpc+~F{&r24DPQsER32mVe68p
zJ9qAE41f8)yU->TLOc$wz-wyw4HqbLtXDh2s_#!tKX$VVe67DSIQmsxLn9k9!VTyO
zSNaoZ&NK>u6o^72z5PH=1)SomJ2!V8ltR=6>MmvvN8tE1jDv)Eiv42XBDFd{-jivE
z>Wl1!(Jeln2ktNBGJ&IK)Vi#x%LY>fc!?Ngy(0qsA|jT6U4eukYI__ose@gcN*k+@
zV#>B1T8LOLGh}AaK9QQLq^GBY5?7|ZC3|~!$!WbIS>F@(J!t%sTIT6rU-I)mLO~-a
zQ>$v_f=R69z{i4&Tt?>Bdwv;v@~NMXNcndQxOP-l*CYVa?qyZvfB;+g0Xle`{oOhn
zu^3Disx}_Y-o2t16x5)^!O^Sk^$>V%a<bOe%o>A<7*LHpuXQ`?hpSMg`b~tbh^^HN
z@Fj3?1*znpNJn_@AMCCTE4=$%jn@>S_V(S9WI_WN81kIS^9Z2x@o@IU!Rg^<EL*+Z
z=0sNnjfqLK<M3rHi{|Zn5802z68S8&U%q^K6Gb^rNljR;nV0sgGsIuS<$T%^=<r*(
z_7Cfz3ncWao9XYRX}`3URy<;73KD$h`}@x%wzYK+$(RJ1vEPTqg}#KBC@2Y$2TIO%
zld)X1r)QhJ%1B3*H*fGf9zRn|EP2UH=u(m0+S(dcTv}E}%xNGE@H)vf)r!;bdk%_Q
zF@009$r`i8sn1CG?%?W?4!pk$U11W%=6t%9dyI1o7t$5zeK#K7k)X#ejTL_}EFtY(
zq=jY7{l;(FKwfJy(k-T0ivW3?w?UQiz%2LdXy<xHrX+Rc1t+x)j&kh|@m3}zYw~h(
zn0m;s4GnsrF#;7Zp3SyG$NOvjT8-7;(^K=TEV(|*Xl3C?5g%x4jJ;lED`uN#rdBF<
zPL@M=&aSyR7&&~2OyDviz5K#9o{qMfmJ@*`|HG}mA1oN{eW^=Jcj#3?2SFGVb6n4!
z;0)dCinM~fT8D-?)N{3-BJkRw>oezcIkl#&6&}|ML8f@@bLCkO@koIa@AIV&0|UBv
zLqTf>Zm(^<PEp}94vPn1TabVL{X5Wk?<x=IW(A(X#O1dT#7QOXI@`#E&dp5mrkW=^
z)1*f%Z%zcM+3n&8(0lp=2Ab@xdN5z<`-8?mN{q4UVXCVz+UPy3ce+cKs{=ygNq$%Y
z^c2{ju!GD%kLt_Cx;Si3`Cg^_usQR3#Q!@z$q$_y`l+omNb~ga3B_m7TmW$_<E#>_
zFVfeY4xrh++f5!iq3}D#+A1n!9(q)n*gVAr{HVLH5afx6a42)i0K*V(*KOLV9&;~G
zZbg3Y%{Ol42KAxJX`bu=xtAT7M?e7uev2x>DslK`osseUh#}Lal!sUaaM44_6R<IE
zV$8mjb_d=g@&2|}*onaF+Q1$P;*i5$wpSoScpa-QftdobzEwy@01?gY3;{Gvz!8Hc
z2h!}w$R=w#Po1|KobT>#)s)7wDezrCV86J+TCs17+Z&{S>&a>ua1YVYx}kC2Qn|#4
zfxYcZMi+Yp5CN(n_yn+jmz^{6d>qqHo$a?YO&k}+Y5rHp`(jqp)adV#mUTOvayYHL
z(xZ~lnZqg*6m60-LrP^Y?tOaxoCBDSQ~_jJx~CsR=;lm-Vfy>@r|BOM5f&NDsDL=7
z@(`A3pA?Ow=Y#f#4Rnj@aEVK9-hDQ$z>luJ(JM%mlUYq_14k#PHb{5i=|r5PvxDav
zsnHF$^i)*d0r|0Ok>5<fQ^w?=&754VR9F-!j>vnAo^+>zuXe|I<0QrLQT;hCh9tNt
z0NwzVGR$6~|I`B=aQ8veM<-wRfJ8G8ew^mU!TK%~e;OAAsmgCSZA`{5P4p3}{gA@J
z0E;9AXoBK4c(Q_?>MyCkzJeY0xWVJ<0(^G>M>E>%HJ8EoqfvBhMaX+*s``sFKLx;N
ziu3UN0knxex^B>;!}%hV`&SxBH;EzNVM2KIkc$RwRvPr5pG{6G=bDDUG`gndw@jK0
zx_^VBBVH_UmUb3Id(d>smvLTJ2{YYqV-x`k^Rtp^9=W<n+5qpOe6B(2DEzv41D&;2
zc^FA~t!mhGwRvO2F)ufig6B@xS*2Dz@%AsG@Xv44efvL_5NT`AmVkIMJ3GMV8EMN`
z4k<*>fCT+bJxNLGi@{SXeb(ZQ`H*ZjG+D+W!KJh*>s@bY<<bk+fliww-jjnr#;%K!
zYG-lcM(q6B+Uo4BpTy3gL^s)VFe`QrBUMf-EBg<uZrH>I%L@_WKP%lLyvvZ5<9UyN
z6g{hS&0a^es88o1=1_?<EV{%iS15xqCrS4|Z&khtyL*=i{9?rXLom|8hJL%RPyZ&W
zbyQ>zkpG}S1W93OX%?`p&{zQ2{7_P-BR>lB_09Sj5XnV}19zqed{v6z_-oZ46F>O;
zK0b(sO|qmXE!_fd>Fu`1;BD^XPf@-kIc<x$UE*MgP|<t>>BG!J{K@k<fa1Y<4@nOq
zK8b%=?`MgSg47eoOVA$z-SRTod2e3)j_zwpoP}Z4&U@I10lzjuJ+GVRAIbAvg$y`r
zhdffn*pB(Pvo4@7L0v>rM3_AlI&h0;2tJWiAVq|zQLu9o4X!hgG@MReAg0&S?^L&e
zYYIW?IC0L!R-a>Vx{~M3CmCXTa7O@a3=KQnh2@y^!SWsN^5}?I@d8qXyGaH1w?SwO
zGV>vcaT7nNra@%c&qXvL=Jb*{d(am1NgSw;TX;_qLX4OSFc`42El-q)6`yz55}*Zo
z>{QWy#d#I(F7b8e_!RAN0FLc>wUp=7YL8^PgIN@?GoHb^%&VTg93`uyiL)8qtG{{J
zV_B+3iXY8qmN@E<>j3Q$RJtBHc8N1MtG#(M*WBy_)^A9jIGrTMJva4v(J#ai)acU*
zVgNipN+vjW<s^U0)3@u7G2(Tr2q*o`n>}moNeIv^*a>p1_{YzEEfTjQD9nfEkMNSO
z|D{l(7!-t1FYj(^RQB(FLfZg`3P_sPz>rip(C$NyS4C!B`&2~aN0duLuZM_Aoi28&
z+%wH<(RKI!*X<@g!=BrEhCJl@ppN9M-EI$HPg^)Wd^Oj|B!L}JoX6M+R&&-tKDy1K
zyxIjyLX6@2OO)0nTqjB<t1lAjUQ6e<d?^?pf}mes;!naqWDU(J;KnlK6Ci=?O`o&Y
zS!~)_p8rWg31ao`Gx)(-F=9M?J28E~0#&Hq*C^2c-Rn*fn|lV5bnw|LFlP9D!VBq~
z;yIPEZq5QpqN59?H9Y>zd2Xfiy*63**9_z)eU=L5=7(UyhFHm_xk8SoNZE159PW+B
zL>uZV;i9_n#Cs#0-u5Q!is)yT+*-U}>I37{Kl+hyHK?0b&9`_*gL4tw5!V7@<)SQc
zgiQ9gBotf`LRyw{EoIFxh(mA$ds91&h5=a=*csXLf9G4Rtf+VbJuT9NLD%gDy$_E4
z$$R9PpZ|_h|BA;=XfS!}3q<&(G*ZsbWEUyi)&iS`q~x|BG6cp?M1%|Eb(<6Y{|%;m
z1CjkBwLw&_1_>Gakl?Ym8)6@%<gLIDKq`soVQH)Fo`ikisL2YlJjtvHa5e{%IyuZg
z=`A<XfvO|JZ-l~*+q&_APVi3p(c&K(fCAt#J$#X6L0#|Oa3otTbM~cHMMo-Ho!TCQ
zT?3j1L)R*pXW~sR^Q<`dCN)poF=Fmbik9!c#|-ST6itvxs7TS<_L8DNel-i|Sp4~O
z1Ta&8m|lKfvqZkPeNTra5-m4rarc+sj$?i){b?-UzhBEJ+~%XglV#(eOjgIRvR8_w
zd+FNPg3(c|6LrJ?;!G1J!?<mX1;xilzS^#Rg58|yC48EE@&nJM&!hAlwK~*i>-x7p
zlJ*U%ZOFU$(L;FU>oEBaQbxd<u9e;(Bgr)`15&u!!5%7f5mOa)^@{sUr034eL8CL$
zVS}Ca*NcY7+SakCk^g;(Bfn;6+^zd*9|a~*YbMY$1uh0~o({(E&A0kI06OUw>qlu)
zPfXBqnjB~i82=gAsIKXmx|!1YzYpSVQ(q{}#O>V>c^d9b-wje;?FM>5SN{J%WC0==
zz-X(gU-b-R=cf}Wf4Be}B|PLG5;q2$HS3Kk<Ws?$(^RlorO)rJK8RKLmk6Fy;^!K1
zeLIhc^@CK=Xf+yIA_F8iHpZ=9J+Qew^9lTsnC8;T)Whq{piRDY_bs4P!nOT04&yI+
zqOqV<G5xoz@C<AGTQZ9TG@33m{U_n6Uv#uZ;$`QXOaTrJmHJIe0NHH!7B`_6NggAC
z@jD-%d*IuFV-Jq7bFj5~6x${S+%=9DZTt1CjA^KZ?0Sy}_FI-d!JwtW_dQ3w1GPsT
zIqKQRkKe@8sMvAyI<QC@LW1>E#MtPWl<1J@2grlIU*(T+M7=<TsoVrYN>bkDmdBrk
z-OW1wB9qkx9{lg=KQ$@s5XuqPHngZVu6%AekN64@EI(g0u}lZ_X@M5H<73>QB>QjR
zv!5Tf->>>3QO-y?dd@Pn^Z0l?KzDMEp;^lI$TEzBsRyU@n~7$NtffZ=(JWJ%dDTf|
zNrC@%jdIAnRl^4mMl@=X896VtH1Jo#^hG{>;1F);`Xy|>I`9z`(Km>i&$d^<>;v+l
z6j8G6`EY#F&#8$TFTRu-oZ^0Z$`PZ1P3h}^_Iz`*9_3i~mHX-XC0LYMS#tr+jeXF5
zrzbBD@zx?St(*1<<6Pl8AxQ{e0izD>q^uVOv|zO7QO`g{LbFNQ4JX*MaMkizJl~lm
z!;nU01~1n6QL_#Och?=YT794Ipc4jfs#Y_(-I2Lhy~}e21@dp))Lzli%80Pmr_W!%
z9;Fd$&ORY>1Od9QFQK!OA<DX(+^lxZF7XZYB<O7cBdjLpF6|qk_YkCEH@-uz2=73l
zdjF*G#>=s;A`WY6tkL48_m4XaZA$;WUdc@%U@|~3p5B~)ZkHJ^*0V_EM|0!bGbtrk
zHydz?gAo;&)=CeLU0<wO0KOqJ1J3X>_&mr#t%1U^fWo7N<l2oF9Jg9%g-I?>Ta))O
z2<E_v?6BQ{JQ@rMV`E~<e}K_6JJ5kjrR)KLO94~H4FPR&AQV!foP|RF7}puAjleck
zMxbu8R>;d}?#sS6&ngmZ1%p21)m`E|JrwEwdJA`asG|Mi7w8WU!2bL~%O)Cp=Kv%8
z3vbGU=L?SvU(*$SCF3J37r6ebz3`++rKSaZcnR4iY4L1N{(Y9+0LAN@`5Qy>J)i6f
zRQKQyV~)aMEMw_9x6_{wD>ubX3i`bMw|@eg+T4?#o;WF;bc&%Z=OII=R<NCaDmHvx
z>`8tAY}5}85kecMwd(`ze}W94BuZo@rYJ2KXvt5$zkVI+mnn5wb2q*ndfX3hd-m_!
z`%-z^*_hLdgcYl7#qOnWw7u@xrV>j%*Fj}mlePTcLEb+ki|{}B76ys!O4S$PR}^}e
zX?`OGD9A3Av*L*7%r{X3)kyz6bo=>w&Zn617$H^#_tTh*6%mNJ%G9NE+0ho~-kIfI
zz!`tg1!oH^^vWxv_B$@J<jqyqtIyo67~n4-ReZ|u{rDs6?q3D>*)dzbPj8EEb^MER
zc@lG;!nroP)L49Ju$<u60Aza@O}^r?+XjdpL)oL7cLI3zD68OqTPyIKhWfWai=Cq#
z27gS;B1|N@h$$n=dT@5oXFrnbUlJAW2^Yz0kmtxMXke=7jf!+BYDC_Kzl)Z0RU`t~
zelMI>1-<CLYl!jxmSE$M%-ovSu44h{0*JS;#0pK1=_H3X5?Q^8DnILOFc;)Xz=DFH
zDgz!%6Y4O#aa>^vAJS-VSP&q@00oB!W2Zo87$y_>iMZZHD%(O&agnca3|j>^_9f6h
zr{E){2!!zi!0SpdJ@^1y{i%?7pM3BCK9_Rg5gw0kXY|GvYeHoqJa_36e|U(jp@I*a
z39KlzqkM64(_^-5u4dPZE=h%s$@<cj*zRygfqX*1vD<@Cs*fTDc3|S&ES<giju;^}
z*w(`m_^?-U47h`3%`_&>4tl2Qw*wp7P?1u#!uSftl@7lcl=uLe)^M29Y&Vs}h9z4{
z+lRzNq4wyGalCtfze}J-FD6ooLD*FAYMoeG%j%)V-K`v|RV_hh|ECanRI+!tXO`@y
z8lWXpV84+XoBndc=ld(z5hD0DHCUI$j8&f*$Sba)G00rn6ELy<1!R9Wsh)uE%RFce
z*+l{}6Iv>6{Vx?1J<0Pdgzy9S_l)~>u3}hbAQ#3N=gR;U8@zP+Qv_SsFJS=^cHA!d
zO(t0m4~%6hwy!noYhfDNoNhLS<L@A4{NIlnFW>Q3BVGBOHge0Q>-P`NKJq(^YisFw
zRT3OJ-$LawQ%gU-lZOmnZ?PcwZk+DF5L&v^CH?Ebb?U|FB9&~Kut6P~v|A{1AH_2#
z@R5_G_^NC--9xW7?QYWOm1=m4_rEVH-Ko&#v~&td=YZYB=b@*(I!{}WCypIcO196i
z%$~w9cPs_ciCwmH=6ZdKGfrz9A5vv;O6wZPtf4W}O&ov*KusW|PNjd@^@Z%tv(FCe
z<3Ij@AclIgC?lC&4>ZHWOQ}EOuD8=4AvARTle*u1QsR*+{_$eaYKj_{t^5Y#5iES5
zA2UO8{g=+6h@?oQIP>%G|GQ0pI%HzEd${#0bDC%#8j?(GI&7tuLgr*Hw6q+#LF&d>
zmHHTK^sC$qMwX8EoGex0ZLR<19r|+-wV$ZXKbXV1ZX|wr=62W<yX}{{kWgjOCxtCk
zJXLUh`gkzufezhzLC!Cy(sfNSrx}#M>x~)Voij-LL6=9=LqhBw3N5B_ds$DTD&#L3
z$#wA6ecFGgqcupvdPL&DizEH+jS5-A@Dh6&HY`v0jXT)bsqA`B<5;zv4>u|yKTS*|
zUP?=ZbKLDpstlTV;~EsFUy0D7<-<;mL~s!C<eHuk?aDrXK!V*-qo)I$6KL3^P-Q>?
zr@T@1VXCq(xxUVG1Rl)Z5ZIY(HY!!4rKP<~$o#X`=CL)-;cJ{JEYXDn<OI%doE4zT
z6|vtM)uCB(?FW=UqNe)x{WnbRXd#;7_yYR6z|)PGjjiltZXeuRZ27zQi%ZD<ee@e>
zy~%am-8w@f%(>}fQmm@s>TA2>H&Bkr^MqcaJbCm<Oo&dskaY`(BOcHa-M5DXN$}3-
zm3Z+F6R^F}EVBo?<B~ql{QWcW6H&@WgO0iO+|lv4%@5no))v&K;odoa9e*AqtIAnJ
z7Xu(=(={Gpjat+hc?{3yN$>#-1N*7rvp#{^pBpj)^nnn;DtXBjHe-lLA7SwwA3YYi
zcAHv(>ck!8)@{uH|Bn3qee3TN%D_yYqjDxR*l$gBM?HR($jv}Q^ILY>^T&Cv4>WMR
zyuMDwbCqAe9{w1a;Ma%8NV8AY|9|X#Wl&w)5@r%0xO)gL!6mr62X}|y?(Uud!GgQH
zyGtOr9^5^+yU$MUP44^gX5Oos`7u*9r;1{qeb(M*pYGLP%j)jc-~5mwyc_g2r++`x
zZIoE*A=d#=spAW=thlQx%)aH=r-egA5@27&QxKJ8vS)I)CIi}Y8*fywIb+l5X_s6=
zxV+zq8w*EWX@VM<bIrq*Q&_UuUA~zo<e*pt&&Dh90omZ)E8<Zx)3Jktg9{5uJ`w3H
zAP~szVrN_xF)L?_P*MEB!w$$40;<HAgTjIInxBp&kf&!ld6uPZ-<Pctw|7`J>yYHM
z*wE8g2>6sh0-s=q`Sc*$)dHH88Et!kqO3IP)U>1x9eYb{MP9f}p54S*hLt7Um%?ip
z|D}Rj4@J-6aHoM9yOr8A9ciZ2Roto^pehJhM39fiFPis;_tJ1aHr2O5dq+`h`kFBl
zYJu>|umo~S6dD?O#q%214v7>J5wJmIgOXF=13vOxKq7!fRTXg?wHnWfRzEi#=<Ysf
z0a=YevWt&GAh!{psT!99h2c6K4Zc)9c`)_u!%IynP}0)F9jKA>g$<v^3;b0?dKL4_
zgxiktHB+n?@$p(DAL%Q9O~RsTKY#yB-OW=MVPWAxATnP05yB2hm=E#$TPu6dnKMG5
z5K{f&fzZf)Owa;AR9#L(Jv{~oV6#ZD`^)`$pgq~%p4lb=47U&>P>)FfL@xw@gP%%Y
z03mA(H~=DOz|t@I&x2oj@Xt^?q<e%Ph)F(pK=~=4_pfk=@X_3Us;(&iS!a50*&V`P
zj@ieR&IkU?;9tf|0a_BX`E>8xce!YBpoq7s3dM6L3=R$sNZh|0!!A{;CSzsA1WvJ7
z&c`Grh42*MK-SjQ4h{^Q$U2GmLn4lhj*8-Q5Nl~^4W&z)oo)656*oJA#!T*42P);-
znTUP-Kh@q(KkVlt1KQ))8&9Pdom$OKq!iU|R|nWADA56X0e2U>8~gjAJPsm)L2$Ot
z&OdL9i;HV)Y<!W)m_)DVmzs*(=^SmMLY%_m>9*S8iyEu8)a;DU>-8X{_--;=G@??!
zHw!TpIjpxQ4BH>LPf}8nta3JDYE)d@^v-xTblX>8fvATFEl>am{-eSCFaJgJ0eo1h
zvrJA))0U8cX&RSSQc<y9Y_J8s7yy_LA3g{Pl8gbc;Cwg_0Z&In^lB;c191B^2K}B4
ze*YiC$(VR};Xz^<K-19(ptHc;d$wKR*98^iI~~mGnVQbknyaKpXFjR}y*g|nC35E%
z7R*gdD&{OMa<Vxbb|OE2?)P~K#>K@o2Hy`32E9`UK;`w((*4zf_4(P^*w^sA-HH6d
z!a_qsL!bm}RZm_0jayb7XnI-%p-w?bNeNho)8QSFV0AiBWo2cp5q|##hZO<XFM^-b
z!O>9|pJNvgkWGTSF+5>W(d)gbVncq<rw7*>2=;>6$~oSr(}>MYgBpQXa^x|<iW(zh
zHKAb8*scSU5Ggtq*3Nw*Pa?q^_F1pn+go@XmV*uN2wcuO<59ey*OQ;#J_v_m94<9u
zmPUR6dN?y04ed;qE^Gm%WkkqtvZCj!3<u{&H;<1a-o|`DMn=}1>3u$$uLX9;Wi){W
z^7T<*0{3UiXJ%#~-(=yW%4umW0c9yEDJf$xwdi&gk6E(hKu2I5GoYyGnbY&`!u|Rv
zlLkmRhYg7@)u^Y{Y>4by%fLyU8A;^?2p`T<keTp-%)fIC1cHHr^1;`2`JCpVWWPPa
z?R=D?N*x{+_OKR=-@yvd-OR+~#R@<kVAicc`ttdtIf#ut5Y*>?NEFDG$9_fuRb&Ho
zHvy;K_7tg+kz>`yGQ7OJHNx<Qy1MiQfi`+`S5TTsOs1d+-Y2LH$d+xf_m&HFrG{Ny
zu?C!!V#oZ?RFII6HMxRC#l`-<-7C*e*T-%5W!^v^fw;IGXqYL&wG*5``7Z-WgIigM
z|7n8$7jyVO3@LbC!F!AH0(e2)J<p@F00*n6qZpg4FosqJ_6~N2dX_(rtUsE;FtRhT
z6VMa<Jmlh{`(S8g>|jFho{gT7;Ma>z#MHvU(2h>TLeIfa$WY(fz>tRr=I>j$%$I5m
zTD@h(SlXv@qtGJ=ins2u1B-rmQH9e>)vM;(BjRIA{}qnjx7`aNUAU@9wNxn7sknGA
z`ue;5sNwj-#a6eBbvfR$j_0*V(N{z1uju*L=`sO0Z=X?mg|lViK4FXJU+c*5F4@w;
z@gUxikK}guu3HCNd$z@Ho`5fH2ySZ>lk?I+@C3BucO9Q;#!T;)HAE1LdB=<XxCwYi
zV~gF>O&+jsFv{EhycS}7+3udb)Xzu?&Z^ObQ;fw#-SlqwOzg}fl=oIkUZ=Gmzo3Aj
z8Q8lhXJtSD8C_ACas?hgHEj%0EoA$i<r(3cyvTj?V>U!61_`sjq!M;*6IEET))R*5
zQ$Z?BQgFV7%N~!p+?HYk$|kw@zKBeGsU%~AfpZEgmuGs)u#I37s<Awd@rflDX8sD2
z-_Kqj+UT%>L_hk%FhbT%n@bg-qTSFB&APNiYtwAEd*wbqHML1I$b8JTKE3t48cI&V
zExH-~0rkD1ac?|Dc)a?K`kS9@9+%#+a-}@G&Rff^P=|_z!P=$N`l~7_+u@R2?W9fr
zzQkh|&(gm4REy!|Dv8wSsLDr1RSo1{Hx!C)!sAL5rP^{%A9k*p{52nNH*!M1=%ohb
zTQqJAk~{LGZE01<O_-4<HQI+<3)a14{M`KhZXjIHLA-r;>Q<Q|mKeoYbNGl3ul&;S
zLP}Fh>0N3q^Hz~grbd-`X!WKuC;4`rJs(t%11-BxN?YtQCJBuGji1<-beQ#(GdXTg
zi~thNsZui~@<-hyEVO{)n^@02WXRa*b<Y7LbOv6YYguJSTt`RJEt6PEsNDh2Q%bEb
z5R%i|Rf>3*w&OH6EFu*y%MsWqB2dUh``=i|t(4yv;S_SIjOCd%?WBo(NK@lVj-(!K
z+YTOfK|PtAWL_ynXw==Rl2AVgb>1%;iDpuiUlG6M&>nb^W)Zrtr{sM;KQBUc2%%tJ
z0mDyU21kQ23mvMSh)gY^8vQ}B+&Aq7>uxxc<ql_bWI93eZT|*ZU!d}k4Qs*ZrEjFZ
z0Nnd5wNnV0kDl`v`pn(Nl-CJzgAHHM8-qi(c0*+?v`3-XHqw8*{*LRx3KJ_4+0`Zi
zuGFCIa^xuf>XdBD9A{E~w|HyH+_PXSQk^;3`Tdk_UWt{s%a>5=8BH}(Uq#XM`PcPb
z*TqtC>Nlbl=cnb><YfsUuiRQ~o1~jL1x<kai(i=P0&^B=WWP#8biFypp4Sd`H-rmE
zibnXr>rGGQy-30?&!_iJB=B0d3o0+R!05Jm(iuIb2qG3-nH-OApmwt|aQFt?Ixbhz
zCe_ra6S;n@2QG&!9`_B14FV161&$sekF;5$Ut`n!xHFP;Tb1frPy!c1PQI!=j^(oI
zx~rN+k1rZAVejpafXW^UC;xC$Ye8o9dNp-343v7+aVA?TzXDAav+#i-VaqhKt`8f+
zUrQ-tB$+m|tjQ%lAAcG`xi6c2_iQa!3i`-2EKa1Q=3cOq$oj3I%tj10;jp&->C{Nw
zJcb|5gvljz+`PK8s!(rRjk&Tg$ysd9D{a4AVd@2Om<Dr^@`FyHI^lSKa)eY^y;2mt
zn#4-+*}f(^jr`-D5_FDIZ9!|7MLdKGr&MD3kB%%B88CAQH~Sc&;zap6#FE-%A_PkP
zC`TeX>%8iWF%GsC)LRcmMDvrlUEwG4kW6&LWr=)F9Xsfb6Qro2{R6M05u#3Twp*__
zOxAjG(V3NCy5zRKa?IUg@>gMV^59q-8GQMDGlju9(H}l*M*y2(f1BTrDA@RJRZC(1
z37ggBdP!5zUl>(jCy^z7)6|zPm+8F4DzD#AXG3|QYC_RjczLoo_d34pFbu@qeBr4D
zIfj9yl1_`8QQ~1{z_v65@j%Hc3E`IHb`rt|Z(-T`#3W;5EDxm~<vJO_`bqhCJiCCf
zQz`*m4)u)4h45qC0ajhO8uxJN#1z_@*3EK`&bBKeJ+hI_qUcHCZ52eCn121cd22ME
zpzL=X+RCW;^wg=A<wzDd%au9y)wuFUegUI!vT0bSoeJf7uT2nS0+~6nBB6!4?}MEJ
znRE0)J184{Cd^jOQ10p!;|`T_#&BG)^>fLM_xD{5<-)};@>BanZ=AI$WCi+@wfB@d
zM*ODJe^kF2jmexUiCMkgI{o%;1{Pn5rNfc#q*syuCgoJF^H|cB?Ie7&V#G?OzM2GD
zbF^t3|3Me;CB+o%@FOkhL&o=@m2)nm@;Wo*2jg|;7>-dT1y%N>9`MAyJof5oE(TP?
z*{|a=XO26z7hgT$L3t3|{d3bkq%tY3{v*1i%3tIKhPTi7Fjio3hrkey4|T`l3lY%}
zPnbIfLqplZn4+Ppl5sx|8)h6$OV2U-)@Du|h;W&Nia}CRt4KJw<6f2dT33$`)(521
zaoZefOU+QRcp7?smUmyeQ#`WY!1L2ETON4|7Q#l6wmwY$ar*Q&n7|VIvI`$g<W+ni
zF=XJIVCj7O)gjR3xo6U7a2TqDwX};KqE)<Q-Q(<2;c)Y<WO<&+gcCa|N1cBq9ZJzW
zhSNBGywY~I1a-qfPWdzE3Tb4fOE#HAkJg~+r@iD1$vlhK4Nvstvp$*-MavQ|KnE$r
zvtA(~qgEOepP64llVGwMK_t89-Sm0MPGb`#<#<8*ulebCCu4k{t%M4NW@r47m4#(Q
zjJ)^~*HikzIsu*)1JNoOvbrTN*7Z&o@r4ikF$^q`bUcT4kd{G3pY&b=Vo%JA<VmAh
z_SsW>zTTI;{sp#7H!{+~F)zQ4Ha-n%GJv@uy;`KaTKh2#>hu!mg9>hwAU6>4s%#so
z?UE$r$nh!)QaLbS7a$RZ_&z*-g!AzN%{Z8L+Wi{MSS4yO75D1qz+|xAwG3G|v{`%8
zI=F<DjGy@QK4s6Rozkke@2y<X*v040VMq-P@=z8&bE>n!m*Q}cI<KSR3&+_KhU_aU
z2v=xlMJ1Cz7RQE;_&mt+I-O|-Lu=821WGwN$waHlIrva9k*JwaCE#(;7cz{aWEjTT
z$<D+#SL8M??~9+Af`eayFDyPaP2`uW@IDU7Ja-J^OYef7*Ryphom3<5W_Ejv+w%oo
zIE5SG=4KN{Wda{r6|Q=(F={_10kyuOQjV%K)7SF&2pMk{I|rwP9A?w!I~!nKTUJf{
zhpPN-b;uP(yYmiYvYIl6%|$tO;Gn6-->ihTv?MC1qkjJQ?o2Nv%HY)uR7mCl^9{{*
zIDdF_<~ReUg5F>Z7Im_7ESmiw_jl}9YA6kBs40X{5fV92kar7TKK`i7_3;+Dly{&<
z*5wc2v$Y{=Kf4PTewCUM(<%(<Ar%1UW#$5=f^dH?=tAdlDvl98GdSYuidcU3w5Gq@
z;&7-ob7TqHf4KI%zPmDs4LV>AO54&jw?Cs~VKaK3%r%>zbGzH|^5|`-J8emkXwpgA
zU1YHp;%zBK(6J8I(LyKh@eg!ufD|UsROw*b?d}g4CbhjgOwS+Xd(s<wV0nw(*F_O}
zg7_3-z1vK6eAa0>y5>1%vjhgU6EQ(a^p?69tI@ux9xpfCSb8Q~9PGR4TT8a<I4*~h
ztIQE1pOq@+v)PnhYP$79L$_=woWfq@{eFFQmZz?QEQ{>TzL|69y;}>gxG0-ZQw`oM
z3z-F)?ff)h=L~f0=ai;vVu>Y;VmF7Q#tTSM^oy{0=Z!%Bx2Nsw2{bw6p>8*?jUMoc
zFR5YvjRDQ}{{24&w4$qxA)U-eGX)1rI%xt1dV0WswzGF2U}t;($GE0vW&5v;>nU|h
zYvOw3rz4%ti2mRWyPFXrdK)@NH@?uf2UL-h^axcg+vM-e;Z0N7DCSp|fxVjMFV&~a
zwC*)mrSWWygS(8i+1?HaNeDor_HTea*I*hX(o2{ayLLn2zSj#>^8HMnBL0DG&Cke)
z%1;k(1L6jO4-47T3r%={OBq_;wGX9W^F|LzD9BKqSPZf&2GzhLkI6x<8|`M%jvD%G
zME5nUl&3*0?lL@uf}E^8S8$hwxN%5;^5>bIl(8gUlpu$Xwj;2XaK-Dd3Ob6|Xlil;
zD7ffP!c|~eUkc=5g(pC~%fv<tTVvi*9rr=BqyE0}($E!eT}pXC&`04x#D{0ZEDg-9
zIiWX0_EU-EtjB6?&MY31VeGJnxaxq3C<LWHPD@4785TJz7F-EvzRzArR+<r9cJZ!H
zU6|Nni0a#>wBfxdRvh(UD72z(f$m4CTZp{(SbQUMF%)5qeX=jTwwcdK;Rm+!k>V{Q
zV^Fl^@}Oe2a&upVD2We1Sv%yVoW#A~;_p-r8s4>pfy9Qy$2t*;#H3J&>^<_$kL-2R
z9}10(k79yuL2~`z|Iwc0J(21(0YBe~MS_rA0MevyPmzEk@Rh5=NLhTjLqCMXPAtOI
zkSPp8xR;G7eLk{&fWiiUAvrGIk?t~t^vaJbPLGKkp)`QcPi8Ye5)-Vw6FJ5s&+mmz
zcML!ngc6w0^?aK^gQ`5JcP*px6`;)+9-u?%`Vz%Q4uYD9P<-z@NYOL{6<9t4C&tt?
zTs%KXH0xqhX5L1;FOqr*YH0W(j#?we0(9+9w{=fXIe1gK8CxB4%*BMyP;b8aft|sL
z+R%f782eAx^iK7(zsL;S)!dvM72K#UPY=$8M6~O~M|!J$oLAEt*@M`Q%#H5d%4$lC
z82tXe^ki&hAhdr+%F%t`s|$-R$k@d@oHdb3Mx`ZfxQM#1VCw=hpWV2!irv_ktvQ+7
z9`haRHnaErzQL6+nsw2wF#BkALG?P__xtrj3Yzw}@f&wBezADk{yvf53MR#dd<mUp
z?*+1FUV&snj*BCdPNh8rVf4yt&ijcfVk6W8tq~D%kl+*q#A91I`;pinqnP>>l}SxK
zMWJz!PU5Hu?0sA62Ud1F%3j86Hj88!pM<PKVaKqAoME~&DVsfZpQ2W~q1?d2pOS2F
zq#1nhmYMg_KX~R}EU~dHWr0(-Ts(rg#ukI<zsGixz<!wEgg@S{k$@DP?dqps2B$XX
z)K%jK{<Zm;*JNcGlN^4LJs{eIBt%PNtv+a$opFT74?bX}wYtK|SHAy}4SAIuT+P{{
zHb&WOUR)8I5oZVLRB_xih?pPX6emo>tzijWzf8Tb<t{W5yZ7ex8hviRZ7eIt^rj3M
z6n{J)ZhWhv@_pOYl_A}OqJ8Mr5-|6+=oAiKLtfYC;JJwqH@2u#nNbX6K$5GkOip>c
zPofR6WljzYHFacHoBVp3@9N4<hY4#_A7e0T*Zi@3uy#mUo5EDi{P0JOws^3VKI^-v
zuQpfP!#;<7)>{ET@IruXoQG-+{CChE_rymqPJJi&)jA}=2ttYMW${iY;meswB3x?q
zHO%NO!lHBdUp|71m~KpfyKv@t4%hPEr!~ImSy@0>@;G{oWWqH0`lCfFm;D$|N7-Bu
zw3e}#72h3O3Ic7s*ohW&T`DxeG#nw6e<zL@Gtq1xUn_2B+UqTsy)m?(^9_2UiJY5Y
zma3nq*9))zyx+u$#@J9Rh1eHeJ&MAJRb>mvpOVs1nz`IGI}>HsX}QxJ)+tq3ES3tP
znui_<=7BEzUVG(E1^>k3yTTe$wwzeFwN%<+3c*jTe06y=w$$B~;YOQaT%hcx*T4$C
z^a9ZzrfzI-Utnfm6BLbusdKz})6Rj{ZG}j5wZ{)E%h0OjiByHvNW{sNe|A(Mn(tNo
z{2H2y)LVu0O$4}+wsus*w7nL|QStWUo643}5u?^uXX&&9-#6`8R0TnDKX|B-2b7n+
zYHJqlCt!N#D|lltr-YQx2mR2jX);~ze*~jt*$K(1neP%ZLt4e7`8h1U)U=F81N-se
z@k%m$bH^D?89t~PX5PhA0ACummTAOCr;>JAd)kU_;SSw6LX4rXS)X)?_qzqXOsh2i
z?wW$E%~P&wo?%#BVp81Na-$A&)p25>X`&ONt%BgOWuBD^O(4I~{FIX9Qd!cu_Wbnj
zxheHzk!W3VF@z`1Y9676$edMQQ0q05gna7knq=nVmv5L-&HDMnwN+u?FoGBdW<paB
zZ|Y_4AH&A!50`tywDbe`duEXdrGoNsZ+Z>`9kkqB?>+4;r8G)iaN4Fw_7tl<Y#Y6q
z2kZ0q&Ek$_><<TK@~PJ4giETG^JYyPIob?h=_;lg<uA&n8q2e;2WN}>4;0u_rOPn8
zVyizvLIUj%#h%uD8I@nhU?Dgr)w_A4c~gyUW#J9;=zmnxUPOBa(lADfSLi}c%j>t5
zE_0U(bv4~c<oDu&Y03Iq3-MAmed>=p#Xncv#Ls=Ue0?8@u}%!~InoR!vkW3Jvvdaj
zn!l4tlhX3^Gh~$Gs=r|UQdNpEkGZ;qh%o=6m;XgR&-moJx=z=+kmbbJ?8RbSup6(3
zwFk_ccTn!$-yyx8PDe3xJ72P`JgImzBUrp<ubA8F$lLGO4BJdTHnxPcI*zU5!(V3p
zP<+2nmKPVK+tkG4eH41))KK?8^BhLW^KHeVzN*%fa-pUPuca7R@)`;betxN6=;;1$
z>=1V5e-p@~e^38k*&3zloL?7MkdBU3Dr!RgD?8Si?HjShozKFW*=AfdNh^~zky&r=
z)qGxvQH#svDDLDCcOu^R@Q^pBrylU|zFooNNo97Db^()iwC3f=z=?kH%L$!!K;Tn_
zUs`a#J`|TQ+_ld^lqh8OeEWEHVVQ={?4te>@~-1NTd;4xHVbj=SmB|)L{cXCsf>dJ
z95?NKRgR42F5liKadN(yPF`>C`!fQ@+w_qhXYGX5lC<n44m2UX+*A``0ft>p!ctr7
zFUS~F*LNotM29=Xm`BhTQ3b5sftAShCq@Ki{po9zhvtg6h=`AmRV-PGd+)t^z5x5h
zi@KN{?K~l*xrGH-ZC}3+4nw(i!PSwH>BWr@Hy3Xrnk8SoxlDWn%-xe?Fie!!#Atg*
zqu0SMV9Zd5N{~Acpe-G5XI?z7?FzL(c`-MUgqvWm$Gf>oDxK=s*Wzx`acH0&Xc7uU
z&f;2v1xmMXe46NU^FwPCgD(=`T@!?R2wvp(Vb+S^GnXg2aLfB`e-O6Dmdia$fGWbF
zil$V{RthO7>XEz`a~Xz_&lBr2NMu$KGeRHc^flOklZwt%EvzaelN%Ri3M_PC1;x!m
z8Pbe-e%cRV+ad@vDe(W~p1In1^xnEeU--jKeg_yxrBnY0F7ekbnf2Dj_dm2W-V3dY
zNwC~o2?(L_*mR&JUT`srz~XXw3#46BMH9k@)XaUWsIMY2<3cmMQrWxHzg3z_oN8ZZ
zKou4jPeUon8uAjZ6?2wVgoQFni(211+|!8Zj=ZAUb$CaLEyES$7D*bT&q2fj<O$J$
zrR<n-m)PT*BG-B>m3zVxiS?R-*W!<lqgSAI)6+Q&aF&3z-M#i-7xU*&EP1lEVod0^
zz|t9bTa~#6Rh!~a#-ma7Nr4AeW<PJcMOK|TFBqcUfc18>Mryfcj>$BiFUx>)8AUYY
zK|`p+H5_bfVEKGX?U>7$VH$eNLCY~o3M?#h7K$xS8v)B5w;saaTOylbv-&8db$gb&
z!;3s}9Ert@@6eTvXP4)8A9reBu&yZ&5?WIA#Ay=b+gGL(npWb1IP}XT-k_@^leZf-
z4{Tsm>##sD3G~7<(z#vp4@SmW5a4`hcKHSc2X__69^$}CUU7p$%nZlw@Y$Jfrb8>A
z0OHkQz&m!5B0HhSOFyLoaP*e9&V*|PM&AllWt&CpggFaZd7DmaIA7KHm?QgGJ~z=}
z(uvxIlIi@ou2c3&tq9M>sv}P4xSH84G?HB?Q~<%gJB*#Akgfb=7;63!+`juuNIkp(
zw&3lHvs7Jey#esGA)!{zGh0JBSz-H9HtJNTY3r||O4Ckz4qUwaTH>j%QM)==6tY(A
zmN8@VJTl_O=Z_FOX+RQCMhi!u9=<1Q<On|xE)QHAW*T$%>%bigXKYIDc-S}Gq(sNr
z-wgzPvHkXD7`ZsaDx;Rf0wusVa|}0ngT*{?-~Od%*N%HzR8Y6yt7S1_QZ^ruC>2zf
z7mwxk?ll|<yK#vn^x1;*+@mmS``Ee(%yVw;u_Mmh<HRE}jw4k^8VQ=O9JoH+%faw<
z5;bc?PChJL@L1vn!m^yO58OyS!m9G7^lw?2-(Z%&`k-J)t;W>2+dn($nn55qDu;<z
zuQIUAz2T<2W|x$nxNtRRMC@gkv28*D6E)pYo&PeY%zcD$lSJgp#kgqjQG>ATf>O|#
zwu{yk-FbydgdR$oja$)TO)XhHWX6_HS(}&OWa~xG<cZ9_Xtw_%0+tAbry@Qgl8Iyc
zRRWFrRa{Dwa+^$|dyU=qhEQU3VaN)*j>f3q*nH|#zK6`Tp#di&6!$83wgEh_KwSyL
zH*Cd1RAY$8`1g=0$({t^kv`mK3%#OSi;&UBy`eS1eDB7EyicmEhq78R@um!h1Zo57
z;`6C7!8E)Zc6Gsh+DWT@32LU^no?BvL%-1Qy!KJwrb<T91tk!bGOo*K7GfqpN<_Vs
z`_Y+Z7X8eASaRFWlJ@Z(Ej(fHi*&A2%}0)~JFK+J6iZZ!yvmJr8#%mQ8lN|i9+?D&
zpG-9qZS0(h$sQB6SK=Tm?^}gFPZQ_X5lg4KR`ox+zp<guZ1u`ROY~0;XN*M`S?V1B
z_%fp_^7P2L24gpl#n`q7g8{xy7j(2ZZl~KJX(&o7-|)$RrGA6niJExuV~9zR?|R-h
z9T+I9Z#N#I8sca%avvJldtBx?IgZ!Ra;XGG6tT@L+(e-Z!PSVtv`PqaE9NV{Giuk?
zXHUGM6DwnKNO&tUQ6UtF<M;GUQ4_3?51!7Tdqz^*8FLK|UYj<wLaQ}fN|o%3hfd1@
z!l7BrM3g{Oc#hjrk-K=lu*O&E<~xkt3eRD~-I`)s7a2cEQ3#1A1Kfrj8~G8z#$@%N
zrdg@OW>kyI@Wq@PTtRgc7>wSHmSoUH(AkRVcu@?bo{d<W%ypDTXcJkz(<P0adDQe~
zkIM=P>$^bgOPG-^xv6c5=4KbT+qgCND2dPnJD%@mUE{O1dMEv|aho6gDza6Iu2|);
zL5Fl<^c~+L<E0W9?66wXqRbOWkuxx->mc>SiyTWoIwy*RdW4FFnAUu<aW&{-=cFCv
zQ&6MsGSEJ$ig?cYis498Qd4XVO$Eqmq3SZy+Wj#UwCh-_tv96{nDi|#O<QS3!Go=b
zV&jSQyZb!KeeX+}qaXjyVqjtVCvx%k(jNl@GxL9EF{nyfFR&rC-K(55xI+zM+ugi?
z0jqn9xwJrQM|GuW;SPhb`^@WbgGNCi{T2s`Ci#u$`BP8`3U45aj#F>`QhqhdH*uL&
zS*}N<4pua8zjZ0eZ!dv{OI+Nc*6FbVgZq7h<hjgry(`>0rq3Cj)!=VkI-feR$<N-n
zzVuFCKc<XN%k#iQgys((g-|;UiAIPCPY8R}f~fa=JfXv>G*mtVd<n>cBt|*v=BdSf
zP8^bO$y&#yndhNO|D;B(+~a&=#@J!~;`!*9jMcli39qOsSXu*y!gY-l!uV2I3|V~S
zB{zj|TSLj5A*#Rrt7ALePhj2N;@?tSZIM0+TIJc{%kwQu5K&U4zE?8-AmVl>`!%Y<
zv%4)UKb6FfG>xO>R8e=g!XreUqM;x)Bv_%co4Glwee9UzIAVOhsX9bAPGL@TqskbW
zH7=JQmDpxh;oCJPji`A}<GfCd^8A!wf2F<sTpXGZqoWf+qcVIIxgT5{%}yq}y{3s<
z$J55eHmNz29ay$nBBSV4mcU^yA{*s{ms%#Fso<=<$u}NX8Bz)%X1x6x`OmUpIDyoN
zjf?0D*zOG4ZPZ>2m@Or<9d4iB`b}O+V8P3~qxmSKsGsiho)Xtxf0|dwfpk(-lVqA?
ze%JLCdG(7TPfe4yk~YKqiT>fnj7;V1t+5ftSR73N>H8wzr=%K&l<eG!Ws?_<VJ+&a
zbK2H%GboD^3c~{+>f@a8;uE1Z8FU@f53&1&)8N)0zO#_9_$fY#)xmSNS~6pYTTYfL
zB<h!Zn>=l7I4f7YZKK9?&O{^B`+C|4wVfYkdL455BXY2gZbvOE&f2l~a`DrALt8yD
zb61R7RiAK-K0if`-1t=EF0fnk064p13E^w4LBrP1!Idruxej?}kM!Zo5jj0X@rWh%
z3WJQLg)za0G@*E#u9$PI!-a8Q1@g*<kQnvHCv!|b-3asQ^Uu_3%}(LJ`s{PgRm>Pu
z&NjMNQ!oj8shEyqXD(kjfGus$mg$j0j>xGMgUv5sBUy}wUTpSZurSLsOA0^Ab>)V6
z{!%ljXP_gdi7%wLQE1o9v1C8L{NuZL76Us~!>(?a$#O&D$FwG|0FUD-xRxvk|873+
z8Ru3=d|kIA=rn%VMU}*~-uAp$bf=s#iSsWarGs0q<k8)OQcJZV=xXbR@w9>}cFpEu
z+BoX<2j`awTsdgQX%#&Wov(y=weg2LjB$_S?>eX(_TLCackZAAi_0I)5yL^BRVFQO
zSHFfi9ckZUjW3S`w=v*nx4CcX``(D~i0m>6^K=;aoLac>8Y8_$#c7?v=IO)1sY_un
z6N{tG=W<r)hoFML;Z$0Lx<Yu>@THCG)Lo<D>ml(hr$sx>(TLa{jdgj5taiwvM^Z_C
z+}U#5nNLa$7pi-oGt<4k_d(9Y<(Dk3_S-qRL%rq#dzs7Vli9+u{A;$kNyrgA@>E}j
zoQgN)D*Py?r%ppCJC-ZtV!eTk0hzcMceLMWzdQOA7ZzugvQB(?_IrG7VfPsJ?*NUN
z<sU%%6^c~_VzGw$4s;5RA02*%X2mV_j1B2jObvi62NobYD`seFY~t|iP{7o|Ue?e~
z(Av_*+RD(%fq;!p(AvV<PQgY`-;hqo(8*NaP+nBvcNmw6m5qtzzsZp}Sc1@0y1s|B
zk?Vd1E!TZyW!%iJPn%L(m^`sN$-a~VyRG@gK_Xr?LA78u<csesC%w=dBB<9P^iSzs
z8zavTw6B}wAGqkWTaLJ|C$C>!51$V&YQFaG)TPIQ```KZFlA$BC-LFK!wl^;P#UGO
z`DM7jzkg)JOBf~@sOOKa%Aqq|J-z@^UR=;6LO!-};7g369nRN=e)@DXL+cC_q^oSI
z96deW1IbJ7!d{_POHG#{F-*k7{-iVp!0s(eDngJsgyhsy+5QfZ2=b6|F$nK>2yk=!
zCw)nC{oy#oAa@BIb;WlH{1HI>9Adb?J_AOA`1kmLkAQ){F8ej^*Bg#8VA!8eeqQt6
zz3GeoXDl_vU-t#h{QJ{;9si6*_-FM0Bc~I1zjbx30J{MKdA&gLpT&CjXIYUgql4K>
zAj$V|vTy=Oo%%)EUpxo|bpm@VTP`-7ehY*hA0IzlsDB3faT{{tb@UF#)6uEd^v%w)
zk&*`f#UNFA4%LQ1EF>ggOJ8HIzM&ze_wQdtRkZ>g9f8+hoCB!W`Rz_HEF@&ULQgOh
zgXYIjqCX7-1H<<a^uhlAC|sb@CVxCzbmAxX3c(~LAYl#1N7A?#njH7vFq;9bd)Gj@
zlCxF7cG{?@sE&~C?)Pn;_dG{7_V)H95oF#X`O>NHpZdilBqU^HMqMC?NJ+PVo(0(D
zA|zyFZA7F}Vq&`Duxcdl-sR`!w!)A~Nl6LoJr7eE85%wg5=f_V)Ji3@xYz*MjO*V+
zi%n<f+8^HAZT9uz1_}ZV2jrjOYV+aG`G5Ei0gdqA;~VhY*a<5tDi+G;Zen5~r15%d
z*Lk0GLb9;1lwjrp^OK_@$D!gRmWqq(EjBk^PJTYH16Mw`+ePFd7$m}5MkBaxcBg}c
z%*?g#p_u<d_j1jNJSi%Tx{}ID+g(sDkQ?d@G~SuR!^E5^RCuSLFk!bfFr31E4sD$A
zM+ji|tu6S+GXK2S+4|?Rzoq401Z2|p7tekP2*NMc{I}Bbk5E<R{Bt3|sNW*_?@#}v
z#827%&DP(K^#4!)N17Dk{RMA7NyYyc#{l3l_=R@AL<IXU+4_&D_^(X=IZyt9z5nnA
zfac$04FEv?4Uzx$^e-U)d-N~({%iFA6{o3h%J-v)MMu(k%QqyAjEo!{E`WB`|0tcn
zsm?-${5Lo_^%iqg60ww)^EKkq(s758mGZd~xEu}zvYFq2Er)+m;5S01ayl=!xGKQR
zRvV9=ZuAab)i1jZ_Vm0)L_t9r8XD>#`op{moA}fE6)zk%8v?wXoSdEAU4Fhf9stsG
zydPtK5g7LugnFBrPZia=oUSjnNQ_65N*2*crEy`S@f>vkxyA18?!k}P02u$_`RX7#
z(EkKThrQj&j@aA<;4tjVmnoIQgM+*4<L3r_DwVQD06ojRdm;%1tscKdrKE(lmur;s
z0V&O3pKnW5h@+#UwWFfR$jGX40;$V@?CkR&!z!6MKo=bHgLYTpzWZZ>PtOLg;on?5
z-5(RbjH1D$)g)K1DR2WrCHru5Hb|#NH{cO@OHfS`7jvt-Mx`PxJu;RlxZ<{7IyCgb
z(#=FquhM*mR;h{I{>MMF7FMUx>t-z&NC|dxaT)3hohnutbi%<&HX2U)sdBDYyvr2}
zK>&IGg+Jw*Vi#VZfb#Nki{6N-g@yaa>lJ(;%lP(uJKg5>F9>r4VDCf;;m`3O(D^-U
z|7WJe?-~BfuK15{{v+Ta{DXtPg!KO~;L!=(h~BL<c}w<oe(%NXHS8Y$WA6k$#9uv?
z{&*VZ9NqOe{J}ic*|Eb1hU3r7VE)I3>J-rFnH-v&d8K2{TkSc&#Ke{AXrgSpzupH4
zYwr5bv))ozppO-Gl_2mydcv^!{f!g2svuDjO*2iU`Xa-F0cy6U44e4BAp-cVNxUA!
z9`F`$W(_ktT>0mv_kp*piGQHK2o*%uaIzy{N_D{2tl#q}pked3>Om+M6G!a9Z~5vd
zYP^vc!}z%#<(wD`4VUCm7?wNmH$>?KMz5ph9S+r5ziuteJ-N7?K-Z#rsynQIe$H+6
zF3uAj-g*~r-iw&GC}n!I;;~lw=gQsHLHF~SBTe(Y8AZ!uK{8cSx7-L7kCeyW=OLW0
zw*-u=D<1EwqW(agN8zzf?uqPrrYKRt>un|dzsyVBOls4!!_6jbN2#|`JkYl!+Yu?f
zha!-FfiQ<Exuo0O-7d@sI%R|M4$35{cRe;o-!D}R>J-qFua%-db;_zM1uS0fzkWCM
z1?ki-7OFQD4{x9vyV`Gz_YW-3xd|{u9{V_3qz59yav&*~IFQB!(K4NypKZk98>UAp
z5PtE%S@N`r|J!i>X=*UBa4R2TeT_>b%iiG6tU-Cj4~7G_;@>v(`vGq#V-$1gAGMnh
z0B-)JPW;OS#1)S6YM@NUJMv#5!F_@mYZLmH@k~2nkx~9~xT72#6Tu+<*MCV8#$B|8
zW5K?*O0C}?%Y6-<qOeX%c0pO}_vNZPl5gC|6!WEI=V3U9k1qOL>;5Q^PVefR-mekw
z7A3~!u+ove*XKXCZ#`^Vr6fBWq>YHk?jWyqdi;LIXIF8^a;(LUeBU?$pyOThQ)DFJ
z^9cOd@5gxxCT)%PET75JzHHg|<>iuUN}cShmm7uPHmr)z?@3BvzTF;Mc6H4k0iFMz
zOS`ME8uGE<h{Z+IpgZs9u9NSWr_|iM1_NnwEhL9OCH+kvV0arfDTIghkZadaKU@L#
zM92w;<S49PU9)BS8t9+9d24}g!fxEOFDYfpHCLLwo~#%la_E`JF4T~7{`P2aA4VmM
z>Y($MBq!?7F}R|w-_4wq=$ROsv`*Y5D#-mU3IK$)Qx!`0Vx<Ci2&2KQi>aY8-8Czm
z00{e=d~K_whwVAf3$V-6b<nT6>wmOQS&O$nQ1xv;MA<)%Ts}Yn?2*<W{DXTp>+xu1
zTgO;nFrtCsAY^!`a5&sKQ>T1UB61m^4d$O&idm6Um{|B8xquqccl1l;evh(Ep$*sV
z&GU1t;Y#_g+yvnwe_D@mw<xI)>58<~`i#L{+b*IvP}n!`Wy<-_*MVYxO*A(za8j$q
z)=VC!#SN2n6Fd4RU3H_R5>Z5Oo%D0JeeBZe=f1o@=7>&H^k0(OQlv2jNRhNQ=k4U9
z_uO4*rN6CE%2}e*yM5`IjI!Z}<SlFYKUDRgkO--<>Kh>JJ5wJvQR*P$&9vFxN`>H$
z{4E<Nfi#8lWDTrN<#nYeg$!O@H&qCR`z}+TM?c}W8eQB@o^#}<qOq8+OnLs13pxvN
zM~<Eyhh$ffwZ3+8#+HMmzG$wWgIHfArIpgi5?g*6hJ*wo(ci|{&*=ohuYWLE`<Q8o
z8dLWz4c3ikjodOhLot%*Sb<_Fg5aN#s-W0>6{qBxLmQCZVb79^Q+@gr&w+p{KT-_M
zv*6kVWtTF-#J@Z~1hwgUX?GA0eC_$r!RIA?18XJ<3lM5}Qj@;fQ#b=Ei;T(U!p^e4
zJG{r67oB5_nH-|yzh$G%hjdC_6$u;}z!!kExApIqB%=4moH5$6Evf#YVU*;Xcx08{
zV(I(2L-<4Zch>18=X=wk8t3>`hwCiZ{N9Hn%yU_ms2^HV3nnas{!mt!*0gUINndp1
zZu#=q@Oe~c0zhv-g64u3%vta7E@-T@-mW-n&g<=&yurq&fQoY@H*XzuyG)!qx$xL?
z#fACp<3s8U0hqGWaJk_y4@2stwXD=r7B)<a<N5{*SM^(7_lZ*1n*MBHMSA^{#er^^
z=Q5mIb&LIN*?Ci34b84+hsGMgcIQ~EAZyyf9{{@f7{|mx=mzKkv*e5#bf?4xLqNVG
za%7P-9o^7bnhqT0D7vwIR`>07RFH@8|M31<NIa@-D}4ga`^gsDR)*WDqw1xbvYY<Q
z7p3~QVZ5q~O%1?wNQIGiG1k@hi3jJ0&ZGx$GBY)^t+q`uj^{}IUvK5TbY+g(PS#)l
zwH7K|jq^25F4Z9Fy6m@1C&q=sul?P5z0+4EOODzz(8=$1O5VO)y|BFC&y?0!L)7r1
zw|#H=V+D?m&^RnC;vee-7RwobksH@r-I(}<!~SuIm`BSVkL!%Yx$4Lvd8X7?#y>y<
z+HVyg$CraI;bgc|M{!Asd&p~1&o++Q%myBOCrg3Hdj@`u-gVGKsFS8tT<#kdxx2A8
ziczM*ZQ9SZHRyPsIR9}fFi2iVFxQ*KgJ?~WJ&e~l`9$-T?PU(|Y`!{h?)IYrfZ$#G
zM5U_;5Jo2_9iwxaQJwfwvN1V&r4z^58cv<3Z*u%%ID_@rc*Xmz-<e`A5|nb9w%R=S
zJ<V?ZA<MZK44|O#fCAxnGHIGMCgTTl)sEWpHKr4Q#jT~O3DlkbWxZ6aM#X-ea@RXH
zJDQ@+(MF1PquDbwv7Yr(z6a)?Z_)wFaUU0$x&LPkGM+WXjFaI;YxM@z1olKiR8426
z)OkZv8pcv7T^zQXttL7`+^b2Vo8cbo6M33Z_1YFQLaIqMkDpcrTST1{J!^R>(q61H
z1>#S8CPzvZc+91|;}378XjfW3K*Fk7EVs;oKvkaiSGkh$(^l#1_P{P7vE1C;ymvd<
zWKu~PH-9My)sqI>bCGA#NB*^2obBE+99aHH52+>g+TI5|0Yaq7YLNMlm4p5-o;U9J
z=o}Kt{Lys+nPaXI)dZhnPUch1r5=-<(k9fDvCzP6#n9_1k7+$vfiMIyW_E;B#IScG
z-Hi8b&mt=yWm%S04Pw1Db2j?>fhl8UMttg+TY@%GVPR9;?w<Kt^WdPMfsG(|oHEr)
zzrPHf^unAz7=Q)ATqBEfZ+B<oU^OCDp#$S4XfE@1YoSa|6~WuppOsv<&mh+kOcr3a
z@sq7Lbptf%02AV%P(HzyPB3(}^}v119zaFpOyU}}WltMtz-Gampxx6vm7gb75LHOw
zzh~dT+zq~<e~`E#wvI-?2p_XO8HlBFb8|Cu?cIj>OKb2bi0du-7=nf2hEX*u#Q`)p
zIfSqV>y&rOorD3sBPPeypRD^fL6INzdiuH$F}cPv@TBas_18Ce%#rR_&_@&Qw$`So
z0kMb1wZV<)0UocgA^625g_Tm-Qd}BR6dD=Kd;3JcFcFt=wVv5ti$lWZ`rDKOuHK#k
z;CCSMKAhP0P8Y~!%cS#UKZOBdGMe-Bu$UOM2ZHxHzG5;m?XKrWKZ=oO%5}VjCES2e
zr%Fx^5E4x<y(=kD8XwQhzNg+(m@3^X>{giNvDm60nh1H(Yh~X&bzbCrdiT9IRqt%N
z|72%1)_*S!2b+~+WKVUz&O!1y?lUW@N}8$~s4iy#kJS}D9gbM}OGlyxL`{gy7n%0d
zr)O4s#zsrs1tYKF5t3B#*4TL(vBrzs7|Dp{3tuY8p|`_4I**Ok0dz>3ZWg<!5<vND
z2mr%@=@d^ozrtHj$<MkwY*@IsrJfv)dy}31ki18;z%&55vdKJoi-?JdUGZX%y?p*^
z(EiGocaPKg$e{hr%Tt}G=)BzA$L(u+r^DGxhJ8nt`J*cMPjENQw7$F`WI_pUgRwON
zLHDke`p0Fu1|8;WGgMX?a6geNkv^Dwi*%5uwlj%grD<gZm-ymvL-uFUS`B%t{Emh@
z?aT6&w(xxCqZL!uYTPP;(Z{rgtf8RRrvo<y(Fs7ySua1P>l96L7S1*ndQY(aX!8x#
zZgsuv7iPn9t_ah@PVu<?yd|@Y%CXq)t>fuNx4*n}?|gn77nOKg^?bG5?2HJc)tBE`
zPe1LlB_t%M=7=FT(8=&VKb+%x-S}Q+m0%jSnyeLIIxZXsy)~8^dgB;+uDuO%%1y7}
zJFE^`vTBAIL2Nb&(t%cqbDw2s{H%jYVDiJ@ieTGUG4iSDa0c((y@UF&Ab3Ho9^ZNX
zR&*QC-`bX_L&F;HFrBrRTJKc$+h6%1+ah-<!&4p0<L#!+TZ(+iSrEsVpczUX(zoj7
zs)h23^7?KWvN7T~50kD=$Dwx~z|-mJ5M-K>19&(6ZsYL<+Gyi>Qpp`1e75;O&<|kD
zF9m>+bj42-Na<I(TS=@OF&54f*;Bh#9K>tov9du#!$@-{QFG_en~U3!vdqNIFUwNK
zTaTx)tn<*efhX!19uIha%8csVR%D}#Ey&m$8hm=t3n+aIfb#|Jhe3;RqGC#QVGgqA
zAAO#oESIV%Q&|GQydKIV%DW|#Ws-$IweFiA8@c_AN$}xw+oN+%94Fmzjz67|X1mwV
zXfIIvN~2tYg$IYrp;%`<SCy;TXb03wTLW>DcCW{@zZ~^d$?Xh@HREDl1xEhLjZ60O
zV$+xldeeg$RZ;i@8|vGH5uJ}C`0OfH+O7uTFHy^~tV?+v?-dUTS(+_F)X>j4c5?UT
z$SS0Q)o^RDb@<mqYjp^0-L1#n{W2_sY$w%#T_N-*tr#P+W%Koi0NZOtMnqCClfMFq
zGa)t>J@@lQ@3Xgul!e6s5a<P}w`ujMeyNwhetz&(dPgQlC}-Jv6ACoCq&sG2C#EZo
z)m-k1GUuf%Wnr+aS2a*9=hn(2jVX8Ag>e^vwumc>>tk}(T81diJ50&rj)i-D!M=08
z2t^br%>r){Qbo;Od4VjwwU2uT4#UrGC7<tXq9}kS-FCQM_lkpjyKqab*w-11q{#i_
zHCp(J_DHeFPdlIb0|H<U%{uA>^CR$Ecr`~|;+#m%tc?{|!CBaeq86k5`-y6hU|fs~
z6$7iak6<x&=|$e|+)<rUu~Kf%Y@?rKU_`t(^#*uusci5!z+ZgL@_K5qn48s49+$&y
z*hC+4;}hFbM0P`1gLbIfuHxakfT!R6>eX4gtk&nJLY(wfg7tlPon*EohqMXy=NPNs
zu8s~kcrFK%cYUIL;r3AK@LD`an@p~3baPZDaec~){7@4X`G-C+B=Bv#w0l4=P2LX6
z%9F&hiK_)yZLowX$h2*M5H#JM#&7rHfVrryT?=;zw{+di`}4u>2B&rK#MgM=L^#M6
z3C5?YUo}v){c(Drpd?HvVmJhI0^mL0m@UAZQdZSfT^agjkT@rgSW_A)4Ry@FfTHE9
zqgP=pgbf!EFVE8{U02>Z3~Yk!6zvgR_B=V+P%K5za<WTQjO1bULlMw(v2b}}IJ?;y
zlYV?0WiP0M#*^Txc?3D(!kpF}v{fbl9T8adMqjabFxvlK%7LSG-8jO|CAwgfH(~js
zyB3uUfsE8bvH(42tRhcTikphtz{9LIQ(`TXDtOOj$`vq|)WL1l4_C?wW}S>Q4!biU
z^69DGnGek~-0rgg#vxTj5YIfeknIxPjHlX6@zF+{>s|BPnmNPlI_U5f`nD|d@0OVL
zQ)eYp-s0Z;NUSV2b9}#jMLPA(W9}?@tW+eLsm%S+vUnIE%5c(4!z%$%Sc5(wCk6%>
z^S&8%JSQfgaX?k(Y>?I=>M*xdTI&;hGC*R2RciHj$E>F)D`oZ^;Y~ld_w>J4b}pxd
zCK%zedSnCexFDdwn*+a_bTXXbM&;#q#c944th&7mN`i!EZ?Hd)A=aJK-O_xrG&!Vr
z*aHAvs<Cs=yGC_fO3T1V>3MMb^EfJy{3jWAd*gFJGcsLRl3NDiA>(O7wp7@To$>cY
zhy;cUP+tF7WR;wYdzCl*n5@%ouayhK$!kaYX*;80`ynmk<l~+&jUQl;b4Y@oTM_mG
zBnYd%#lQJG_C<Yx!1EsTeu7$)S`nP|%`M9-`xUDP<s~A^z}O|q-sP}Z8TW<5I-jaG
znMaLlYpG=(?1&Zh6AWp)y*mooS$lpNUG?eAlD1l&s(f!E9k9-^{4|wTyOAxMA3%)2
z?i?ejoU66UJLkFmM}{mLz=B1=_(m=5Jp{(T1mgyVgM2q_WKXOO((~PMLdg~2Tcn_|
zh)cQ3${RSXmIsN#i%q&pZ=<k~8a(o(`}53N?HSAuwPU{*f;A59q<UQpMHF%G^1YXX
zV7>eaki7CQYWE+U+wRYf)~CPQY8jOGF+lKZo*r8gc@+7B$W$k?^8$+@p5XtAFY$Az
zs)Eg4s!vMF<53iX+)R{8iZqlS=0-jxy^Qu+d1_WHPGU=f)8Xc#;w;Y85hTC@A~U3j
ziDcT`NX~pqSz11YK7{OfQLS>O%^=6tJ>?|r!Ys+}?YN!!9Dr01@y5b{P-monFSAm|
z*fJbusUHVN`_TYA1>ot8Do)=Ay7cnf#qNG^66R~`w(Gs+kTP_fTc%fj912;c6x2|D
zVuX&$ljz`s95KS$jpt|5vY)yEg`Rj`xcz)3&)dnL8qQ#*u_h_;J8;qYfb2j*{%qG|
zu0K5zk%S2T5})TG*AFcs3R>VLSDLqjS5GUPPl0jn3~Ph6vh;L56|X0FAOaa!9K4Ge
z*$(y^w0tcX`P*-}HZ-s*unitpJ!*^MrVAh-OvHR<d13v-A5fvPs0PrPvT<O)0}(e>
zkUUdS2OXF+ftkhvQ7?#YUwHuT77%dp7Sa>~OMfk#hfV=#^~BW+&xueeDkBqX)uAF)
zT?c|H7F}oF?Y209w|?<ScU09uuhXHt_AE`#sXc_kv>E4terUb|mZ*!aruk8=<N3ur
z!^1c}>B+1QHl5=Td)SKjm2#e%k!iUyHYd{4sSmjNHl^DMJmJ4Fs2w$;gne@EQ-p2p
z_R!<O{V|<aJD{IKO))dm!ixA%nC!O1Zfl+m_vru!9KD)Cxhq@5HhEox&I%XheNKWS
zL&dD4wn~ceL(u7qO2GTburb)_liFq_)aZ_Te`P2!8*@XliCReTox29w!hY)qI@j24
zR(9@*%bBfOq)KQ;1BMsrTeP3HHZYIF?y>n$!DV2Q1LA#Z6Ti-t-q({^C7K-QQn#ud
zbRF_)-mU4_Rz&`7-M1Ye5SiZ_ZLo4o@T`n9aRb<b&GHFXHx6(Yoq~Y~2(j~mi&b{h
z7GU$6i^{JC9Mia!*J$k;y&U$*pq94fzI3x3oj<x?Qc2Y&?O)xg&5pX>m6`xFVDYXT
zs*w@^Sy_)2w~z;8R?3_{=&?wK>$*g;MBYO3pWR=p0uGHWTSD=UIx_YaarHMns$e}C
zAY`!<V6yoU(24Nc464Z!;u?fPM{Lc+<S~IHJgRpuJ0O~A9yAIjz5-!_P7S%p+RuN7
zDsCb;8qO{TfYAHCg;140;+`?L(i7=NUtu6L1t_MX>dwARqdkcRK%Wx?-%%XAJ9M4Q
zjR_DChrTSmB%B$G%rp}h)E5X<M-Oyv7yx{9!i<uh;voq${V^|P_e~ujP{Y=;676*Y
zOIM3v0S|5~%r<K$GN_(bYje|K2zE+o*H79TuLuZeh(_$C@r?PgsN46UkThniF1AQ5
zz%Z#nKWb{B^8QY?+|+Q<0L1xGVOkO0Hk$g7fZ$(rj|5uLD%9Vh+>>-Co1?YH@1>#U
z6y5I{KW77#m#AqQuWQhLhyrE<5dD^i%H-1n7QF%El#)$_zJi5s;I+gmvXhv&<Zf#N
z>ufZIc)ix*YpWNzQU1uXR2-!yTIEfgw1>bjq&0m5k4;f39>3(i8QCV`UX0qb@utUv
ztG=a*$~@!YQ2}16ZN;Mvbo}Nc`zcyri3(N<_XDr`6EcDji%KL<B89VpI(Q8b+kzs+
zZMWs3Ri9T@g8+-~ctm>(<Y@c?zNZNUBEJYoV&7<K+;Qk}BWa@*>$|7MV}A03#tdTs
zq0fj5$TII(?&v=bgi}P=gC}CPMOI;RV);`>Kt=Qgh;6!^Scy!ADL<7G48~qi0TSPd
zk@3<`7NuO-sXPn~v4MEEcu%`FNrO4ZXo|b2OS2zlVb~1v=`}Tu4cl%-;a7?Q*wqtI
zg;Ey@m=b2+7(dc8;U<QQ%l}~NEW?7#y01@n3Wx|uN{51UcXu~PNOyOGgmkx*G}7H7
zB_Q1`-QB#0apw8IALsJR0Oy>&_gcTT@2;^pDL<yer!<-6Y+~d6L20Bka;xy@)|D(!
zF#Zvj?+Xc)sX>`P!^QMW{*_N<u_GAP5Wzkhn-J%8wj|^9d@Y~_<o~ox-`Ih#c*rf_
zwBd8TZm(T=esL9Ya$n!Es=kIS+?frU|Fc=f4{7X6tgFH#Q1aDPB6Z^!BKVCgcSMiU
zvL;bd$-2tcxJs7f;7ZEgr)&eDMa=)Qgnw@rt6(8o6d8*YH1*b<{?K^W^cI*43Iz8d
z*-L3-stxAqPFK58X+JH;Z)Co<<d?jaZF?BI{1!-+=huKr{6&kf)ud*aj`P39@JR?B
z^LkX&{(R}{)#DnExD!=R(W<VH741EEk*uKR^Nb7rnExaBgrE%5cI{!dXU$&4e9&mn
z)Q9&SGxSD>P#eru?R|)N`A83;@!L!!263suBR5@%1LA6KXu6UzwgNrLJM;m(|206I
zP?&CJ_#pf-H2wJ~NsiM_0#$(YJH9io+X4Bvt{_|c1IOiSf^7imVVp-C3x7T>^qx7e
zt4=`jzg)efVlfp}j0l`?uBSKcrI(rEUKJmU@mKm+IqdZ56fnc`^yr-Dv<eyt$yN>l
z-DV65>47=ksC{=Ms{hJ}mJghqcyrC7Neu$2MsvI~^@cKnwM0b8hGzfJZ|Qy*mwC59
z58s42-CBBWI##`fg`NrttYr%J$sT$vJMVs^_W5(d!@@8v9P)&V*1e;qD&%q!tOeX=
zIXoRx9n0;39Q_k;FAA6CfaqQgX5JPgq-uw%gk5>DaqF|D)v?`lOW4o9w%FELwXYhX
zxLy%r;>Oss!K!uAmxXK8Kka&3awmGCXwbZg(Bk`F0r5G^{>P!YRs%vaCw7;B?XlrZ
zl>f>YmXyHKz&Nel4`4SCYxPVgq^bE5e`!kf#`zQ~9U7=H(>Xdky3kJ$!j|=$`z~rf
zz9zodS|tWY@BpV!mDr$XlIM~ZYcSq{0sgl<X`d_C-X$mp&tm*^34>zSqtvYn@@}y9
zMahL8p(MMqoBJOyuu3U9lUzWdnekQgl#jW6iA8*Dx=@qnB6SR`n;%kIUx;+rb;sIu
zy?jfV7=c9iHj92&RGuBkd;j&EPh|l&T)i8G$1%v4f0jc(Y(+NOe6#o|)ln;3&DHN-
zn_U@WlRIFY;J4z}0A2Q}#@pxQE*nDv%M`!W#F6>QWOA{DOh5^<CqunSvz)J4*S$}A
zh6&OYRh#wQg$d0ok+~pcWT>?fRVF$kTb)hk#$;#A#y3m?-vZClTfD1ktjyF?65@g~
znn(7QItm54$^vsG#A;dMR`5f~_C7d&cnJeaRYJ)Q>si*l^#4^3)8={BCydXHAzNsg
zyy9;{bP#Cw<eh=>Fh`heL_M!K@;#gW#d~4|)R@~HX)G^bi@m6OYfF!ikXC=pi9u&r
zA+1YjPa_O&wO%G><|mu<Ag>49t0WC=6VZ_cFXO8;0h?5g)YXmCt8kBAT;YJ6Rkf}d
z4u4rER_1;!{5w|Jwv18-kA}BK{(Yj5S#_M)*t;k+{|U?L>;!b;_iYbXOwPv6x=Yk@
zN$KiPIY^3yg;<J&O$TGFX0$DKSHY62v+>on+!hzwqtyqcb*?*(17_M#uSk;O?>`XJ
z@CS=~e!mRw^^Sy0XG)TiWo42i2o-+e_HiHs`~JbYu`)l;N@9Q1OqR0QmWw<Ot6q9&
zV7Lnvq759SmChd%;o^SzVW)=y%~i*(%3Eth(CQ1El^4iW1@!+KC{GUXwb;B^{(Tdk
zXt$=A+!!vz6mK^R%lL}0NsCGkRTqc)ORv_JuBwlgTU@GvbOyRmdTsizYq`<2Ea;}?
znMR5;<VEG&1Rk&S|2)nX6{L#QQh3YQc7ZvV#Hz;j`(Nj}J}-94E@~b44k!9DL#dz%
zZ`Zzh>l`h?>!%-e!}I#~!#Vqk!I~+lO13_x){^dV9eMG8PobVU{B)Mi6L<(6H@^gn
ziPth6@=g5$b3_f<BMj4q#*fml3%xPDtr2n-W*m+rR=<IP{@2acbg}G}YoKYZHX`6e
zVRB(5dcm|Ix*1u=Jnx*oh*_k8H_v2zFt|qHmpVjIjKb$ww$ipth(|!>X?TcVUu``o
z**s=aNKXy!REd{D$=5Ll(GkgtDlKDTgeu?z6zr$_=T2qlP;xPHHKJg*@orrpG2ncC
z@zqz~>*$ujL#OiOj3ujNE2*vv*Y7_|@KFRl&YIFM7Fdk=it5o4hLwEjV~Pj~@v~fI
z;R#MuT%eP?V1z}eOa1+f3TUbTM=#Ooi!bi!Polbe2*=)1Lg>P#k)ucAEy%I&p$(y+
zZJGw7>}Y+?d59kum_F4UDhl?^@`oaW0~N&0SH}70O$wD1O;v`A(wS&R%gMvw9&O)@
z2A@|pfQ9%#`MPF<z%5%<5DEux**gZZ@-J$H+1eeDNB{fXOT#-+(X9-bF|DC@2#|7L
zaBM>xr{1wW8K+eZV^R^cntXUv2ZQ1qOdX{E&2$TNP=Fc=T+pxkR1z@@=ITh|P-f^(
zaAf$<9|#bjUX#c2zoa2qR!&qwIt)7MK-S2d$e0V6$DhOR5=*eph<!fLGarl*5j`i*
z%Hb7HvcG&HTwqolzRyo{am-}eD%3GcG2q%z?py8Y22;%LJ{-~(L1wq-zif)0BD{<4
zQc<}-IkrEm=c!NQ;h=xIO#Y}_MKv)MOIEwhlMWR@n|TAV>xZ3fT#NQm4Lj;hQ55I6
z*Wy@WSRH})hn~AP;FW_s`8mjE&5>)oywYYr0X(>jO~)=3eWmv$AAvZr-%vxwMDl{o
z*J!ju6OS)6H_u<WFmLQV+={AgaW+atrDCEkjiDM`)L?Ub(!-NjVhgLl0-W<`{qLWx
z5?ZU_0C4!HW%1;NUzGxfg^6FTt`z8s8Y<wJS>G)pxeDfa)bdyaR8l<3RM}Te9PrsZ
zG}d^Tk-_Hzx76BXNf~^zr_3ytl5|mthWl*lGnJv9Xv*>U?}j{?et|*wMK@dc%l=kE
zN$$YK``mlItOw3vv)1`JvnVCRv=_Hiw+-?DP?-rmKLe&h@+8nvMhg<plgiq+wRQsf
zh8;Me4~MI{XtAuf%6K59Q3+8fty7YTKoO@i(Luoun*gW*-_`tMU)Xqi32VZ??&U}Z
zx;gx|2&J$TVRmC&Bk+d%?QS;z0B<Q`KaxMz8NtBZq&UAMp9vZ))v3WTr=CzvBt3wV
zD3C4qlMXJdKfOr}zvwKG4i16{F4_A^=j74#;<_lP`!$!61&l(Yt%1sgZvV!!@*LW1
zPD)CxfCjZ0BY5$6(n5i8g$mWD0IfjKU?hlVI#VZHu>{*3vnR9jXF%<uNoJIw=@(%w
zxcw5yDgY+%Bzi{tZwNfnza*N~D|_0tT{B1k!XTGFv@4|8sY{xhwsxZ%EFxiOCkYU8
ziK)3+we4GX4<{i}+bZ~)-*kXK<cY;kTA2zMRdd$~3>ZGGhvzOUTnp#tF!8J9m<F8h
zG#~Y?DuF(CP}}jAAkYQaD5TonFK_V<@uI$af&O`Oc#|q#J~WP?)!ngs7FB#ZnnuuE
zy?uLzRJ+NTU&=Vv%7bhhF*o<c*UR3+ALDxU4`C&A+;x>Xa@~z$n@|fthO{>`4J~yB
zZaqMEBn?d=*Y_i<PpvJs$~m&!-yjU!UDikN|7%OHRC9(d8)$4XF!8>SV@schkn|-c
z2b_V&VXk{LFQmZDyAUtI)!sAwJ=ZBemK8a?hAmwyt07G#vA2stna}?5kX`HT&@NbW
z7qUpxsJaU#_?RL|^p_jjRr16w&8c=Oty3EZao-7pSx>CK+T?b?uh_oe?!a|~POTE{
z1}!~5NvW`E_zS!_eA3SW7YT*55oo9is`(Xbx^Ex6ZwpDX607Kg@TD3OAdn^x8N~6A
zj22_AR-c}ZtE~Zv5x`cwV#Cq{Gmf^3rimmZ8=_CP^sOEF@90u#%XQrsHPC)~Q#2KB
zPA&X`yp&i>X&x!3`zoz%n)PN!7z{7Bz#d&Bxv(||7Mv|oq5b?3983)QrEV#TgSw~Z
z`_On@hQ2<dKMP=bH+Q2t2#7_I<PBvU+)32&Z-S6X;X|#z?cOAE)?NZ?4Jk|~nj&HG
zgq7^11B#2|+<M^n^V_xMpUIAIKjQR$@4WQ9W1-+I%{JJKtI$o@wNGU4W+Lv(dxtMW
zW;@F-jyn1+D4%kOmdgj5SK5Fx|A(1-OJL-N#|}Ch?~rdb@)>%5KBZK4fE6&pxOSbk
zju;7r#~9&B$glH~r~2UdP8p?&q3wJ_@xUlk4s}7|5&bN?TOhixZ)sco^nLX0;a3*{
zlHnq{$oMVXH-S1RX+S^2>v~rkI$^$+566odeF({DNrcHsy6!Dn@3u&sn$CKn`f-aR
zV}95&{NY#nFY-_E*j2xVSR38SLc~dOj#HT}^89fjTI;u4LBBdun3?6_T0Ybj>Cydb
zg$OYv%xdYkTOUEp<)Z_ia6%W*1Cq7I%E{ZZ+PrUVT44U^H`GJV1g5>p=Za(>ZwV+k
zAB@1s7NJldtJ~P`u6B5`q_x~!mVP~0^BsuBP6osYzrksH?!TP|!<1-!rJhzEZ5TcA
z$q$;FMiyqDbyom!pP)^<gOcqkKTsZ@lA98-DdgU;G6Us?QfCXt?dmnWH+`fFuP8nx
zzbjduK(1oBJ6%(!W(BPUZRF2%FQ8F(wBJNRBGf??p)?0`CMxRr>OMy~dMzEOskYD!
zoJ(@pm2UhFPOTC9d{tRR{YM;7PyO+g^a=w7@!xabo4z4nQh!E2{{FSpZIOTp+2YtM
z^NQ;gnZRCqUi+NttF$*-Ztlj~3Qh8oz4dnk)<PaaB^j!x$}-TDmYKc>f;q(3onz>>
zrSsnP6;nPcy=_Bm`0WZN$=_*$lV}>0`AuMQq8mQD-;T$pD(<6G86*47xZ9ODZi4R0
zSEEY3Y~wT8bi{{BR$mu(7)s@)Uo+cjoT+On4!8~*Uiib*K>3ntz}rcOMv4(%vmExw
zh{5?<*{z29Q%U`);#Y|CyT@$o7GRp>edSy<`44R5Wg)m?CpCtagDtIj<~z8@@M^wH
zR9MJQZ%jPtiJOA=t8XxRd&e<NqA3I0ksV<GMV&PpK!e$20{$ySdxgi?q|=wBUQ{RY
z)=LA(`3I8gDnO_8HCs9L!zh-~PXkRNU@3kWuDlKZVTgR#4jK|m1(WL`{I(EU@77fd
zZSnWNHHj+LPC%XZ)#&<|<fc<BT|r9sRal%J|BgkJ;@b?H3_7uM@?_5Rnl8}jR{7h-
z`C6m^$(PaKEJ%G6)Y>Qd{3?qig@N)N>&=)u72)4F6H3axtk^a3T9GhJOC-1i4AiDK
zV%f)nyRiSwLQxR_C@S`1S!Z6$Cwsd9sVK*zN?25XJktej2XB8?yct0d5xqu2w1CwN
zPYLhOzItzX_6Gxa95p~^sw^n>Q1KMAccJ(b0fzi{#oL<SLH}H(Oyt}xC1`N8vgBK_
zSPH&CW3DaPTmxtz-?wOU!SjL=JF}#ElA<q1rLUFhVMO8UGA3X48v@DTPsSiEPL3N(
z)_{Hp9(|6{UCM#YV*7>Wip7k2cPrc}wYHjs`1)lxc3#RD2i9li7wYlKFPAdB#EVW_
z<D5Q$?tGya*P^Wh@sse#wF|}v2PXV_1A-+h^55XEU2vG3nwavBMFx)WJ0Id5A>DD{
zHOqff4MEf6A<*5>dvS$`P3pz${4NuZ2TQP%-EzU+Bpe8?HuK)#c>7Gd>7?DUsrh7`
z=npl$2L*Ssu=h(ZQz9hMrmqA>AlEbI!$^)T_%e@M{1!&QlBmtx6=+(O(q>4GOhN+`
zOC)yHw%;MD*g0h1LM47@gvQXhY|2}d>i$z{Y_u(o1&_s%SEM)IJnM1?b5*pWodcd&
z*NbjrqOYct3&nmW#8e!XI!8^eI5rtRd3>h29)c;z=cm608-$2w`U-*erAQHuxE36P
zK0qV;Y#{gf;EudzJ6z&lDn2{w;v;vZorwADBlVYl3}PaZZp%PXof|RNTW;pX(DpKM
zl9-QO5t_FuAjSre4PMq$fDqgj>Y>)}eA3exfT(RZjo`4VIG1?xUbHl|-+5ng>N-u-
z@#QJ6AmRoIGi6btK@xq8Oxy=u6>Y$^Wf*RTs<O@(NTj@FDX}){2dxR|<FXr@Rg4M*
z#_R$U|H`9znPn)1&?u=sebF6gvH%zo6G2fyzyr+8*g1=l7Y>-Z(>}PoY{^}ES7p-2
z1*{Wp=cT=f>L4A2IO)p7PxRW`A@qDJ^j_de6sHj&K?_m(1)Vx5z4+9F{+{#@lSoDZ
zQNd?D8}T2>ud<ReIgO1d$uAP}T*Lk&A~$eDhkjYm_E6q~!h~cWoKUDrib}VojHZ#}
z6VGSQKps$OGa4pfTb5(f9p6k{MS03dRd?Q_h$FCnd~zDK=QbU8$+e1z<dQro;B>5e
z$<qg$7Oj@bk(N+HTjZP;lk8*ER>yF@C~~yM-M|nlppAA_;F#8rj|)&3a;$7b|EhMT
z{5q#4Ls}ll*0uAnYqXzH6oiX*W?}%577`5pmINI{!^1jnWqaL}?6kP)#O%Uq!&PDh
zP##wms^=Ryu)cV`g>gUa|4&~^SBkcIA7e;l9VKT&Zg;(*RzJe~;qEqLPmcEzxXpeF
ze#V3YUo_a@G;#ohmVNg7G)fb;1}Z#3o0hQMnScPGypTvSh*0eN6v#(#MCif1pwpre
zT09tNY2WtRDW@h*bfISSQdNP`0s>8UZf`w)`;R~RHIfFJCa))eAo6yABMI7n0A`x?
zZS<BVyB7O_-JGp+023N1qg9U@>d0%!>Xi>XLhwQ<o3AP3BVGiq2KL1OmFp{3F4;O6
zW1jBl@QJ?*N=N^CL_|apSHYV40)DJTZ4w2Q@LUO<#F4oI0><AC5BYmQXzhCo0XXqk
zKot;O|CnRc5805Um8%dweOq0W%$W+b3>AlpmuYX(ycr1#;%gZ#K!bX?-vpdvfF%S7
z8hz;-CTD~e1y4>=9X&lZ=@E1;P$pV|NNS33tqEV3$odTTJ~K`z$k>_(Rb5#UKW7PH
zJCE3W7~^`mrLRg8GT$W>sD(*<j%n8M^!zR!5=ZtV60=$nagC3|iR3Jmh37^4;QI|C
zl+|+b9X32il{1<macIVgZoheZe^I|qK$B*ZW?~g>E%5Uf_&m%2F$+QP9W$9?Ns<nx
zz4f2TN#DQ;Bjt;7h@@7|2TPodj@IuUB=H~9g*1$UD$TBoSEI!@l{MTynt|ahdK-{|
zKs>pyisu+GB7hki0wWc`&l1DrgLh<THOC6bUqKNK=y&Ho2;jU~r!ozvJ2k)U{>%@?
zNR8e$6oisL3dsFQd44jUY5Z~<m0<cq2ycCP#hLD^E&`yotI(?GK$(+~X>YW3TYJ!q
zc*1uUmS?#$Z_7APo&gAkhm^bW>_<(wSnQXT`)40l^gWtSf?oOEi@QN2v}eLFWjy~c
zVF(KI_p}-uQSHbMLVq=*TyYcR$Hf2Q6Zdi$HOp2{piMY7D9^|Eyt;xDU>v_dMn*$2
z;GX1Uty?G~{y~2Sq7n*~WH2;=EL+RflAWsOCZR3WEd>`POg6MF$qU`=^*6yL<^Ed5
z`4lUGU?bZv9)wDVdHfx$epbHXPEorZus`2~@?m&#ziP}|37<H6Sh<o(m$t>JYzB=#
z;B)Ne8=9)3g>|K}WY-HEb7~f?!dq~z0`U1?6c5#G&O=ZJMp1mB2-D`h0CUy@FsQ`M
z`1H;=ssugzhrVhtT22a7&bL@!wCwhu=QX-5npCENH+u=5JfZ<h`z<NM*BHIrd+e9*
zj(QcqNQMMEZQVof)xP5)F7q#|Eq_`CPPh|mrW_C7-e@T$Q8LKN0B6?7-?!*5um0mx
z59Ok)AW}f&Zek&j5`l$Df6P)z>Yy#R%*R7UW(YumXw0l9CJI4{Czua2h5i;nq)Oov
z=ifg61XbVq<tWPRN6ZpZC$GJ=G4Wr;jVj$zpq@3{=GREsV#uxznj}R=$PL|)6)p}r
z+c!9BNVI1y0hcUA&B}G#)+h*cjfXr9SD^Q{F<O$x9QsvKFT6y#sH6o`6U~y4$9s82
z*e+BL*n#8g+LQ{jJ1yDy_o(L3VorAa(Tg+nCqa2ncwc#)AJi8hFf8##PXV$hPI>Z-
z+@)r+)M970+SR3c)B-Pq@1iw9(>Pt`z=B*Rn*E@?#g*ru4*R-`a~pR@4KTg{)YpV5
zX;8TV{f*oMy`q@91+{cr+=7LH0QmpXVusQnw`c3~cSKzJ#mR25`=y8}<P)f*F^Dnz
zkZauglN`075e8zUMReWqY#h#n6`bNdO+yF%d(Nqgh*}*XHJv0nKNJ*QkRIi~0kl1p
zD^2a}8z6ZIsQ41>i!UOk?Cp_pk+*}ZfT$oOssYAP&O<v#1eaGeo)`>em%Vt6?i-XO
zSqT2@d`qI~8fa!ftg@PpV-+AzdwcMv4hKl8ZILbLt`?u7=Y5>`#H|QtUaKXssm>uO
z*pmMc_b?ncgcUfB<ciUN%KY5*g!Fe1JQi`rcu!^y$J?!zrIC?V55vKf3ONXeeQc<T
znuB5Klr{ZOga8GFhl94VGQH=ZD^?XCEz{)d&r0swg^B_GTq67LN&@Ceu;%$RP9%VP
z-yyr{(dc=`G$R;1X<~x^?fuh?XF81PT+lrA=Z%_>J@#vY<DCzMSTBaq6Fq$ksfs@5
zevY7psqkgm7RQk8fo>dK=yo_6el>>=s*I?6--<^REYHGKGnqjTo^*!v38C^TjWNN%
z3dTmA+0|Hdn?cCJ6F}aP6v+oG&2IYtu=nsR=qyD;pTH9n`i}1yg=ayPU(8=F+pb(n
z&1XGJc*kC#rG7&E8$ww@KFpRh(05!pBcZY*JA>oR|CK(3MoH*>nkimE)2zaq>ZvS(
zf$y_$O4b7zbUhRz1D@;Hn{R+v6aK2!RTs!Pp<^ze5aZ=G-r@-;<PB}W&zbFht10d5
z+_@__ike4%Bv*=~s{);dZ;(**yN}4uswT&EZtSLCHTaqP8731Fj<1_#mNQzM6@!sf
z?JE2YJ>ahrITT`8gxK=JMh`&M_0`&`SvQS%f;r~%-On9=FG}s&K;|sO^i6S_Xa^m}
z4qsd(>)n{q>196%MG;4ez=hQua(I6LHCe#wru|>sc1Qm)xylep`&nN;Be|BoS(!X9
zx!wqA=~If{GZ?Ix_lEVc%>k2McEE@ycOnYr6jg1iQuwac&iePEOKKkR@92(FEeWA|
z*O(<@@l~|2#Xu?x(wvkrc5|tx87o$wYAcn_jdXa;`AQPCaG!(PGL|lfmbOnwNW0ZZ
z-(nV0udfg|+nr!M;lHJRdRNNTjJEHbF|D#|Vx0IJq?*>5)@;!o&_)Cfi|P<1@2f6w
zv0-L4%uRL|`Wjt{;xCXghN55dtr7}RAY4h!CxJxGR`sK49w)X)oSWH1o#21th`D&0
z{+{pDvw0SneO%~uA1Fk+Mw%!3Oa$X}0{@D-TI-I*gZ{V{$Wz~85ewc0=9Si1?fkY7
z9W+IN9CojF`s<~GlnOPER7^|?M=FVj;^wk*35~bJW2o}?zIeW^$nRqSuGBr{<(8)r
zDGMX@QHDd`R0(s=T5h$m)jLtB64yxX@9Q+3|6w&(+ISS(HUsaHmJr8sH)@%KvEnai
zRD{9oGaE%tXofXY0B--DTp8{!<C-5>81cKK%k@tTIC&#s!28=gVio%LHGg;IGY?Y>
zek{M0^*U~*%by2gtbycc8gJeJwk2M);&mjtxja=3HBO!zOB~R)$>d;!h#2nV1G~OU
z)W=o^(8UD`$rh$3hK<f;)>^ntGF7qJN7*J}9zvy8wY5Z!?O<o?b2-1OWRYixuRjiC
zl1ug4aju?|xWad8V^N>AQx!9-D&0^;B+qE{HtE_#1VeOBFL-883g~&;*x77I3GqHJ
z`<OjZ#uN7Idu829Qf!-2xZjH1^5F>>s7<X<5eEU!>|PD~ul>(h`fQ<b(*{i+X(yck
zGMP)m<x*#{kl7ofACq;LfQ#LzQd<7<tS&2(Fd|0uM_!MzOUX0ryTvhp;OOwlK|hCK
z@16JhwVax_`RYeZJC?4V-@+TQZu-&mx-#wVV4!SEXq&JC!3$V7Qsv@3ohyUsq!zR9
zM0M{rRXo;Im_ISwT^e{5)k4-JEN(R4zRfVVRfAtcUy;FYuv}r(^4Y0vz^z-CfqqZ*
zI~GgI_SB#xxL=8c@hPqD;naC1glB%p6vS92W>MezpK@CH_c_O453VW6ed*Eq?21oP
z<h=gNgAF)z3>G^7WhB#|Gc%7Npz1ZjIe8r}lajMvXQQ>hC>J@dp3EYtXv}cZTBMrH
zVPZ+i2_QKY*f0iNj$Ch6;OFa9C)1b0&=`>quhr~INGFruoov*`vEKvfipAJsZ8~o<
zLiM;jM@|T!Mk%)VDTMkWta&=}5+8c5fP*6_y<f!emXYO~B69H@ypgdP<t(HdEt<ef
zM3~77)b`nHR4VRQJ{SlepWMVY&5)NSDySPK+W0)m4?A(})WphpONoOL?DxCCcMo51
zwu`p|I_P+~MLTZj?;HsJl-DY5Mo-;ad-S=z@q|jzD*d(`>hC-!kAcZkvy(GNvF$J_
zW<{IsW5ny2{DKu$!vJ;*2}+5%-2}{*9TL4ZuNXgPj_VfEvQ_jM<&i1_EV#DSTq?+j
zW6PesQ7=A=4*ECHt!Lq{$e2g^unR;AUMeIaJ1VHE(i|!4zAWzG_0PTzhi{PB;R<^=
zS$X(zq~=sH$JJ1O`OxWceGyz?>?i01>t-3>zYq<ZYYKe7_yddK^3K^~AHku>WH2k#
z41Gv=CRI5bwQ~Lrd&fae$Bkql!n5D>oio_K5FnM3#iZEQ%?<Q?{(v7u@@GZzuM(L<
zlF5+9%~D4}0^}WCT%jK;J87J+Ul$2&2=_j7T>V^3L=v0*2gSQFaX9#>XiXY2_@~n2
z_z#s|TrxmxVb^V1Y2Se%>WHPqT-jEogRsUj8VDhDO7{I`B6jTd<xdrazxKof@B<QL
z9@1FPt-i^ec+cDmE&uS!4}NJAZdrPyL-aNlDEOD!qe^%o+u96&zKn5Gu5XhN$IgO@
zTRqgr;~xvSvui8+d|AE(-r-4G`3hukmIl4Ag<>FAOum?K8rI7dri;RDxs1v+zX#&|
zvS@p%2g?)m<1s?PhhwvBay!)@(eHs{Y0Cj7k#8`Lz7&epK&l1mCXd8gyoyDJw956d
zdJ0yrWJR8igj4?F3I0#A*+66d;=Pt~_;qx_x1hZdk6d#^+$E*lc{9-p`cNlv*ddlG
zHpTrq8f$&4<IY(08tNrEEuSu4fYMz5<fjn_$g{A)LOaTv1^i~UlqW0A7+{Vm9r0Se
zSKf*fQ2nOEp8TIT5q^}77IcFd&nn=mFxOU7S7v*ce}Iv<dtRJ79ii%@0Qo0mkeBXA
zu?>{sjgRs#BxZs7-u!EiVm9T1!Gf6mY*boX#Ef>x0Tn+ccXe>$w{o#`t&B$pvWd#s
zw*s`*B((%;3(K(eT`+Wk?rlAc36i=2c<tr;s(3{ott#dQx|-cm3g+ZP@wago(A^k@
z$AJQoV|%$en_V|>PE$E=UN{MaO30)J>+O_LnJrcUl^hIpsfPogxvMQB9kiH7MMQ3~
z$}YZF`LrIBM+T;zQx4Jbe>Rc;40)orc)TS2l6}WgnBnZ8*G6E5Sz*S)j?wgMKsI1q
zaq=vNxd{?o*-vaa3hjyLpS}TtGN&Zi<d0E?1V=6%v8h_q@|ZXMryAkcu#8Zo(8R>s
zvslfAfli2dY6Jf5<0T)^O_vM>qQH>}UCfSPx*7UlOuYmEF6dCw3{HOCU6DU0$xen{
z41q=RocngBKzZKDH&j81AY~KoEkqf*`i27rp^%lgcngKY90B(=IW9uj#bC^saWn0c
zZ~O>)WxOk9kga>pOo6<NMxq+$C=}C9LRdGjg=%*0DXd6ul;LAU)&%Ehl!y_U!6@X`
z=_C4|wMv9{y#PnV+H5&N|1WE0b$ggHTn$K0X0d8ehbDjcp1MS&v=}ngysooPPJBi`
zaO-K7`YK|=^uYvYAk&<3aI9dnYZP`M4~O`AjMkR88=^DJvG3k#zJkdw;@m98=y>(p
zP0QlM_ZNz-H<Py&9nle;NxJWu4o60ZE5U`pZUJ)TLNM^<zPG-&=AXN((^+7(`uV}Q
z2$|N)&*^%1m5Nvrl@r7}J4yH~9|Z<jfJGgBYt~md2fM8WTNg`v*xJN)<<tC&ds2W9
zNly{x@72T|2mh6YE>MK_j}YYy(nfd+%m62LoP+6vH@fjY>(=Z9V~o&CV`1GT>`%ZP
ztJ>dSVoy#?appPT97b&K7|+HD@D&I80;GV#UI_IP5FEoLL;cyotMO54FQv@RBrp1{
zl@uwx%AK3}mWtArdG{(fVv4pskAPa!rI->$GTZt1d7|R({_th_%T0!XHOY0hh@@uN
z>@_;)8GCOjYu)fSO!5orGe6$h(G$}^6we*ii7`=0(UQ7@mKCuHgariBK$yj!{o7#_
z+wohGzZqCh1Av;rpn|KJ!skSKCQxw8QEG)XQHNvl6!Xw`(?d+iJ3<*G6Wa-O;kAp;
z@rqjLa(`fLov^)z{nGpL!DXvm3xMh1=3u&a)KsBWB=EM)4Os>pjX6~V0ok|oEPWW=
z;|F`{n(|n!C>%;L8F9QHI_tpwyBVUCqSF0rh8l8a&%`<fiYw8Al++YWFmsuLSVhnS
z4;i;Y`qj(ULglf!#L79dd=YuLL9!p+nb8dA(%)M01y@1Fx%Vxv?#cU0wRt4a^WVF7
zPirw)HpV*_`2zIs3rvo0!?x6Iq0B$2^R`MfQb2{QFkgaHi3Qu5((9}y=jMYNe2}?u
z%$k1oFdJ{&LrfM%5TZ~+7>O7j<PQYG;6d(u@K>6dj12VPOj4&cJ(MPKGb9nlr_m|_
z937f3C@P6(7Kh6VU9%rSGs#{DV89vjbz;f+B#?lHkp%6HNQ!^2WD&B6WKi{IQ|Lr=
zgb1WSRU*0@{(~1}sGdx4@+l#|bW=J|%_z#l2PJ+`U%szojc$~%d(jc*0ihD<`T9om
zHnEc7cV9hb1oR25Y<b5+gbv%wP-p>-Wek#5A-{?}UbTyM<RV`YiqLxTTmNFKpo<nF
z1|dCu_<T<6qWu30dv~_K0Y(A3TM@BHc}VGV655s1?o7BgY=mT?)2cnKbd@1%)x+@0
zpI$?sip>TOw(`|0LIgA;J0gBo)0=FQ$6uuDWMBx7!CV_?Nd>7&H2_SnJewd2Ckp2E
zDF8M|FcbXLqPo<0f3mSYD96CKmt22eo-n3^g+yGxi@%_T2b;wPpUR|m`GIHjo&GFD
zAG_D`Ibo8sezRi)q;rD-P@SH{eJtY}Eo{vQPQgQJ$K3E9TPf6XAE-5APE?-_r_*(E
zS?ySlt$mr-cT40dthdK2t@O1Bm@FqNIYPF+9g(~Nd^|oTE(!G3+qT7X`a{VF8%S)@
z_TWd8VUt|~M_4=gd5Yxrd6a)M>7DIgh*!9>+;wTh%Um=EW8LJ@sXVJs@c$wI^Or_6
zchrh9#Lj)ZUjnrd$o7Z24*~f^>J91yQoO9>MhMr)JXkbI9Y=Y}Aze_0@P_(<+)|uR
z>+6KrYql%wwmbYAsy0Jj&{{?v8Ai?rCbe8{4|$~=p@n@r$24!&m)g|tVAH#5k6Xvi
zC1<`2p=}6!aHTv;>az14`zZR%N&zi4h>P3pzzD#Zv`wn<gtqgprKMK!Y=fxWzXtk?
z0q{Z?5GAyf`t*{0-qAabiMD9HJ0Uic^H7ee$EywR+ysE>b8N>&Xq!i`&SDk+l8leJ
zXyshZ3>(U!!@WYELx47c3nSne8-VUvV@}<(<q0x5l@;UZvP@}6&WR!$M0e}gap29q
z>y|-Q3%^GIPL;uIzwr^hH~JJ+zz(ZV6bdG_E`XnG8>jrrDi&J4YEKs@6N6}`Eh{%f
zwql${ceN<sHPE^IHFrovM8I0<TK}r5u^Ljix&eLueLPT|rbWJnc+u(58=r&&dCnLI
zu8$o&-s*f$+n)6Z^<S031$o3nj<U+{v5$nVZWit0nHXhs=SgM#@wq^I{F9~xNmoe~
zM|cd$#~3ZeZtv%67JSB4k6jd57?A^6LAnEnQH9-(-JDw1F7Xm=y6V=58^|ed9O<Y2
zZBq0^;a&7*dC1IzhkKLbU)6@JI<UXO!10yyrVrxBfbeQv-J?u@4s*|~>NV-f&Ty0L
z3(f%LQci~p7C}E!dbo&9Whb(TS>v2+UH7u}>SbGqa(TQCHS;GCMI$4opdw-fbMTUO
zB|fv6_wUfY(6QG+SdyIheV(blTR0>o9wzSInEc~%!eV>Vp_w`0FUkILD=E?eeI(%z
zC0(JGBoZPFpzf-Gn{1?b_GKtw9kjVSS7Uw%E8V?7&RCb}He6XUmQ5-?R5_XOYHzOk
z7_ZjawB%$wR;L^8k>Eg0Io_~JVTTXCH8L;@f-V_t-mg!e=ji^|r^4BYx|1&5PFb`A
zp;=PEK{xl9EdK}5rOOToqS*dt-5!f-hHlV^7H}r%{l$L<2co#y`&?5_hYv?&envqi
zdTjLhGJN;(02lU#h5zyGv<h|4!^-*tu(zg8ilKe;#y3G`?fDhmax?NIpZWDm3rWBn
z_IaPc7Xr`dGkp-HXO{y8=gjg<=PUI-DKLnEMHoa9oKP)B?~&zA^{t4lJZEOX+zxU<
zLaJynH{btoX)6WSY?bjcx{UQzV9Fx%O!t<Gyrv3RF#aW=jjI)pq+;O(oLaDwV>161
z;_ZsiCPFgEyl`E2KZ4X21uM_g=n}>}Saaf8cgt+44V-O|ma}CZ5H_#NKWPP91Ug9q
zzT;a%B+**vIfFMJWQeCpsdDn=!~dece-miOD&@F3F3yC!^!a0opAsgwoaOUJ-!FXk
zMIdE%2jN4&{(k|WD$yngehwhOT6j@G6dMu0X-kFu0R5p`jT#!F6KhO*Lv>93^3uP4
z&tqDyC4q`bGL0BSTC|YKVM!?~GJ0<Oq`=w|8<QpB;!uaR`##P=&cuXG4&%C)M{olT
z*DZyr&80?|x*BM@iY;%>+*dZ3L@vQA9Q`fFMo3Cc3KQ!|YQ+o~Pe6Hs6j*?0ah_X;
zL8{kXKq|WGJxpj(O!p6Vjrf4^shba4N{;RaiIR)#kCeT;sh^1WH>^x9Okz`?nb$SV
z;gvNu-?)^w+00=eo%HS!Q&sHBgY_dIq`$MX1M-~pH8qxCrz2QO1%5Z4sz`<V$%_0i
z4#-n-o6}b9sq85$hIPd45PvuB)BtgX$pn&U2Y<vSJ{|e9iUfXsb-uI%{a_C*y*umM
zyeOnMrR=q9U}P)_A1*<VOk25*T=$^g4T>Z7c!31?LzYvsKtt36l>Atwo%EFH$g|aK
zxOv5@Kw*GX4`64wmFx{f7<DN04{8qtko&!uE(4iLuqKGAv`b0XkaR`_6U6pGdLg;V
zkjAt-5;4er#Bl1%EC`(BSt8;gRz<_uI|zVFKl**cu)(d5mOi%1F`K(rWI6Rp6^y}S
zFrUwl_3t~y+MW})9r&Rrw;=YHvE?ZGveXL)8?c!Qs-w@Vr;MLJMtpg=$i>0o26h1$
z4<@!cAG`(?2o`;V?z+bsf2PhCi-K?R87>EY3s_Wezi4tA0{S(`L3ps+hZGEuGgu7#
zq$_Wk6E1WY@Lg}jZjpl&nLnSc+X36!gZ?m|H1yhP-iF`yq49zkf|vBDIm7AhqOqcz
zvGU`p`S;`PE)XSKR|YFm+z8JV$7%m;_%L8aMFQlVl$dquqMTOPIe8IMaqxVuPX(5&
zQU%0CAZRR_2h~4=J78>_FY*(&ztyX4)1hePM~PXA*)=5$#n<Ek0!FuPs`IjQyE>(5
zIX}(x!X^)=jhzMzw3HUFD#yO}E+52b3WvHA)da$>Ao<#UZ8<$bChp4b={$>oQO^rK
z^)87@z_LTIf^EZWB$H=xVF7R(bo%0HEBFq}z8Z6NUEn^=Q+%_c_1xt-*BV2|CXI{b
zco&mhoBW#;bh_RE(igZ)Z*cosGs*(`0(7@&1hrR?OjEt3I$kEV`*HS^ux%Z@70|jB
zux(QyNW+hwgiD|`YNg8t9Qkaef$|BkT|ijHX^goA8U#5?9NPPgro*^~H-TBAXLJ49
zjEq6^XEoY2NPA57{r(nXL#9v3AoNzluBgq$%c;>V9%}h}_wT%15bL2@6g$)Hf=<(6
z1!=?vEX$jZi@3dLz4oC%J^{n=3GWJ=Ew*IfqU}?xcOANA)K4mwLuW9U^bwt4XXn!0
zRUA$WU>pe@<~IETfjafRHSiJS+uK_jHok}R@vST$S@<1&0l~k&3~Tr^mxzA3Re*Hf
zpbK+CBQ?W0(VIWm{`fk*)ly!E(}h4D5OWeEpBr0JMZ@ssdA2_*;eV_BQE?`3p`~^j
zK+F00C4VwXxn?<L`x@tlE$0{78-a679#HO~lOO^C=8+IZiqc^%C<&br@rN#+Z+nAm
z7dDzxZFHACe268o6|o1d#YVZR?ttmM0X6S*_>Z3~eT61m9o)z5C<8u8iqMLRX`hGL
z-I3_?e(Z%AC>6UANeWY~7qBT_uB?FxvV~0O__3-73oF2>VfvH0gv}#gE)8o4$jr9Z
z;{&iy6HGsEjz+-r6#bM5RCsuS&xN3a*;$by1YZ;f8$uPbIN90H59Y7zxYE+n@bU0c
z?(sm<y!Ap()ZU)yvCX;_%5J?kGCn>&X)iNMn+HmX7vYhZo=2vxl)VlrE3v#;C(Bb`
zQE_Zt_tx!fM6dKwa4Y2S`_J5~uBFrKfeTu3hXU)2;za|n{bnF^UxtsyOg|Gl*g1Gp
zb~CH|`U4M{j`ORwkL*;CRf!CVG@^aQ<{ndtKWu^g=)n9yQT`>u2J|X>vgHBD4OtEG
z4XTi3U@=29KlR?>Y<#8dclo$BiZ+=VX8n>uSYDXl#Z<CT@G*NkGLzMF)+);-yx)rd
zXdvSz{C;Wh-hgslPBldr_SNKBsoI#VSr^u6yl8I=d3m{*Ztd<w$d<O7r*W2u*M?oz
z4=)PqbSjQ@u|ZzNg6#gZ9uH8vgp$yCFnyqm1>W01&l}barZcZkM^+r_wgSuc^&!h~
z(EPleHT;r8%gFRtc|ZOecT(jovFf#Lq~(XUsJqU$6ckJDENft02%F84Cyr!lVq$RD
zW13oE%Tq!CSg<;_8SuPFp)ZDvsNNxs)#~qNRHOGp@OrbJ7<f})GZ)ZdfrG#QzWrw$
zK2nx8<85D7gL5wDO*MCd!V5A%F=fXxs<Wt8@Sz38qfdsu+<OKfOOBra8&MiYt)weQ
ztGSrlRPz&2mxaeXECK?punW;xQAVHUM#0>~TxX;Sy^y_{x~P1O@>8^k&Dp1*zTu_J
z3Te5rA#_74`N7=8h|KMG&E4G>FoiF7YQN!L3sE2w?72^hG`4}rO%(@6ch@>H54tNa
zxzk)Xx6Af!4oIcPsSYzZH20B8R!zl5V8uIS*q)ldnPknFCx6fZK*Kc%Fg2(Gxs)pO
zNI(ydA*Tj?*z2s~iRs<*8dybm-gwD)DoiXaGPxKO@JaV--FUQ@<eBDE4(!)~wsY$r
zu8y_ogWe%?$cFpA%+8DO!qYdvwGR0S^7aTp5Fh8hn3b8-`0EXfiK~i$-p{{qK;6E<
zB%j<^NDb{dV|7k~*Xmq@<}X>c!eRXQekbG9*0_V)YXbvXE8xC1cgA08U0@K8{<{N6
zwJ_Be5-`N|f4Hy-%Tv+tp4K$`1rx*<^nmTjtv<&8{U{^oNSsoxAKsY3df{_pxEoMj
zYldoqmr#w}1ibK#Nv*{xq0j8-9*H`fI_>s<%`CJmi2<z#ymL6(G|B)cdU~;v4}c-n
zbWK2nuh1S6e>`64BomF`FCrr&Lq$cc&~6k2|G;-!F?~kZ00a`xiNTetGTeIqcb<<z
zk<?rTE%ov#%a&8C8jU{AX&^)B0+6u;?SvmNF|Xdpa5mj<*h{GmW+HoBq8`%b(qu!>
zJr_K~=`SjYH`Qa~X~lVw@06{A7CgO~$%{8tQMK1J8^U4@I9}DFT9!yhTyC{wFu4ng
zfb<WUv2?3W%9`pAEw~yOo*AXv_yM#AA>#*xyD@vpnd_ZmIaq=HVEEmX$(v8dlMdc;
zBrmQ!&z61xwZ7~84`t$D6HXNz%-9IHHek^~48Aw$9tGs_vu!7K6)+rt63~Zqq9;7C
zi2YpN?BKDmusqJU4UCM8z|KqkPJhaM3Gj>X*<Q(<;X(5t2<&=LyBf?6I6P~To8gAU
z$9<zk)oje$7x=Lfo#;Yku{f18HCP8>aG-zSk<4Q{L8q82`fH<)L!O8&YM=nBS)CBY
z#ogty3QSDF>_J@sS0(ygfE1XdEqlg$A0ulKvp*0Q>%}(v34Vjg-g=Wl8-Vm@5**?{
z<sA*AFA3bgB~uP9!LE}9E3k`dF0LdYe+Hh>AUuLIFu2|FP9zbcm)_438|DWeh>i7c
zf;C*H?U9NCuQ3bnNeY*xp0ua{ZC7b4EIT|t9SoAv$9#?$L@NND$i`3?r;Lj!5hFzx
ztg8hixKlka_(+5tm-Z|VKccbtiU3OEh|crVM5;pYWpHfF>K=&dIAzRF4)sgB#&#l)
z35cljx<{X7XS(ndMZmrvg)f?FoNy5Dj3A7h<S?%xG+Z$QR+c$SL05Xh5)hEJ?aYHx
z=?5x#f1|QhZ(P4H)<G#ewY_<oYesJ<-n~L!X%tY#=qwBMNTYS>j!DrPjQLYeCzhzH
z$#9uDpsc;p(aLWpx4<s1Y@0)>CBH!Sl+|sx<?@KjM9<<ka(4P>Z^>cgcFBX|=q~-J
zs|%0o6NErsw0*r&>M#8`8&*#;D}jVub!>k-s=+QS<iex$FIb|^h!Qzket59vBgK)w
zl#UAgyMWrU-4f#Lu@9CJ1Pb2Yu~l~i6*<`jSizl7qXIp7WpTC-Vw*ud@NFPAH_e&N
zQYs*ttJ^E|Q7}yXr5_r9_57MO2pY(vC{r88XkhmQ$OZ5woEUM!IJ-{CFfvQfnJDcm
zuN76O`n=6p7RVVI%XDX~1;=GWua7^YJHv*jL1ntRoHH_JPDF72Sai;LjYxf>e7vV#
zg6GIR_IA}wqp(UIr*Y=EwmYUH(SbSytAtCcv2fR+4?czPO|nO-lo&_@*3TT(szrs<
zsk<MlU6oo}+Gx*~*~>(wl1+W_{SCB2<e)9Z;R~&$qs#93U3`EI<2#d-%Gna(3CQMS
zP&1$XyPU!Ejt~YfRvpoxFZrv|*Dazfbc(rQt8f%+*&YID3urs5nK=XWOzvo|urbJ-
zSgp6($E#4W1)k$)MYT5e4ljJYiEepL+rVm|gzs$hr*4F&?x`HzzrUg8+L{*8)+`yK
z7EgR8m!(KoPK%WkbKS5Q)p#K)pI@2WXs^^>&<8U@o@KT%FHSL3wx$bPh}si(NbaOQ
z+Ig1!N#&||2Bx1bk$A9VMH8|aYLtqaHL4k>um<^8dY7&VeY~bPtbm-cIKh2u#5}t!
z72Te;^Turql~kH9>u$su!=29Y83g<<*JGMbe*Y2G*XD!^U>Tw!e#RqiVVrDz`r-?;
zy43OFJTD3D)%<`ZumwylO9z!@Ded76h9BlLy5>lwG0{X7+W@S(u(_8BNL-cv3xz3q
z%*<-c`Fd_91#nTvPd&e|5_T*OBIbsypvgK$w+8t}-1RWAKMI*9`-pnyg)niaq67p#
zdC(ug?o^JB4r9FPf5kX*`|ztTDpyakIWWEao6PT+J<-=(uUruK?f+~hTy3>3cQiv1
zC*F1f)o+Mrh=k<@&pwhE2eu2<sx#CuR->)bQ^DSFnUE-*gfT#1uZheGIH6b{jd}yW
z?6kM|G`k&0&LmtjY;YwA)<av*AW`p@0LA$)0k%Tx@@{V*=abT~KezVTr~PEVByi$&
z4Xb~O=8wA9*DAjKq{04y6WiDO`z~@n!G+a1pe`7;N~O1^ZA98%XAXFJW-;Dr7cW(o
zF+mFJi`}DV@+Lp|T{3*E!*pe;43)uDq8y)SF6wob+(Jl4oalZ5wv|Ld^99?fho%p@
zv>#Q^C&zkd;zWc-LAZgnCTk7ia00A~vZE-4l{p(x;%4)VrWHdYTx3MC<Ja>e*82V?
zMqhrjea)A}KqSsqD@dGq|6qS>e1NxBAxcr7Fq|MrV(BFVg7beV2hTrfyYdJykQ(5G
zfwe@Hl~VFaH6CVhb+Uq8EQZq&oYlMW_Ec|~+sXQ0x46H<-fsWeK8Dk!2#N4n-K=no
zezPXJIX44}i-awZc%3bht@1qvDYE*@4#QXH?G?-2^OYBKWfp9Qi}JXVHn94|${TKz
z2@1oNBU28VhKQqhP0m#IO>Kj>1FkueA?{#C9aggLM$=_he_aLDNn#N!UR!Wvpl8WT
zVhN?%(O#kI^|s7<I6r;dO$$cgI`m%F1W#0;=0Ml+<J06rACh0yT5W*BT1L<GNy`8C
z0~TXDYfR*qD$R~0dcmQdVE60rRJhyHU+C_JI!W?1^v_dB855i`jmj_Odm7gdToXNY
zpxU$HS=RveZiJIv9jCWlBgAAC0}>Mw$tWe`)E7CCk-ZIKFshe`eoGO}gC7abC#WkK
z|2?Q9ee1`H8e3Ig=L?c$!1QWRU$$8j+xtx?$<b<US{UPl9n*UU7W{U*mO3dBk?Al^
z7$R<O2(9Pj)K6qNI^|x`>OPxI6SzY=_c^wnr*G|H*>oxR?{PJ9GHc(&Dmm2}p8O=m
zBAYDav%H6-R2|D9XN||tQ`g|~{MPgE^C~G*^S>?j$G^C)ucLnM3LifmAs-(*n@Ur~
z(&f|VJAiML-}%(tGr#)OIH<_EoZ^w{qB$dR5Fdh#g4QLeeVMa0aiHYEaPWCTE(Pyh
zb*%N{$xVPNs$k}%m0f*d%9UoP#(VsHW>~szv}XF9;HOVS8K@JLnug20Ol&Z32->{W
zmo@b^U80NNzJJcM*&8XtMG_G;CaxVa`-7}SGcJUp{$Bs2|A#rcyIBOM<*#=KE?QRm
zyIwO|hhIL{Gr4|Yt<#wc#u{I0{9zY!xEkU&2Z@2ORxfa%QTu1=+XP~2W7?{izoyzN
zM<(%4=k~-iP72eYKTtOKt8Rfkx9{P7mM=Z-U=NPCTXJ>S4YwlTmQV*c687*HaBE1A
zY1frEjn*VK=;RB4*wmO(gjVlgifV@W2(~F8XIMec`R80B&xcDgeM^lMBFsYm-D^z{
zUW3-Z0)NJbz(r)DZ}{%x0RluI42G?;ZJ90kp}0y^v9IWk|EfOB%L~i%<5z`{;mwSP
zUl_M~lOvh5gs<K%=HE~as~)y_e|RFpJA}J=UANm(Sv{o3sI7YSjX~C|d#f^P0ue3~
z1=|4f>TFqhJ0dq5Tg^uI$M`zp&sD-iX@s<FQYt6^xh6sRP(#124lKXe=u7o6=%Zsm
zM-Kf<^?<*4vMU!R?vt~_e31<M{?**zC{<rtG;KdB*bwY|Hn)1ZWMBY)AxlQ?b-saV
z+j%<nB|oH$5H2knJufbdL;Kzj*691J6*QG`eDzXiD>Nw-dFz~$8!nY4bA2QVRDkEe
zUbLFDIf^1*zK-c;Aew7YgVjkMmqcIYgbl@)dCu%w`#mS!X4)OKym6vYV32zs6xGU0
z)cv?owl5>644}h7Hy<hmdZY*j*H5MBhbnAq-gb@f>?Z}EbhtjHGqECC0xZtWYU8Z8
z3$pxeh)5?-?ZW?ihua<HLj-XIw96;)g9UR^(lqRv*YYzYxegp+G4;u5BL+k98$J?e
z<5hly5PsEU@=X!lRHtEf-(;-t$h1qLlz3cPe+2BfdVY*Q!6aO4O$CT!Q5tI}sXAz`
z670H73hlyc+sZfQgWXl-b+ag4_VxK|r7l9ouN3PCWbRY<TUyQrRrGc>cKY!Ku5H+9
z@%ZO^e!~p@w0X^lE!g^-Bv(zi>|G7B@LEBF-Pn{wpWI!=b_EYXi|YJnu^rVR#wFs#
zu^fgl9bYTKlj9Uj5XhTO6jr_%U7rrhi~G--exilP6rc<s)^Qry75~YN?oO&PUOy$4
z2=mf7h?)A6(eZnJxN33%3y#A4^E|K#=A<hUIlJtrsx5l2`-W&>;anXx$=ygIS4hmT
zQ5qVuy8M&2m`UMR+pX1%s7)W`PpTPbY{P>S;=YKM33d!hzpCe|@wUp4q;zsL@o>Fa
z2!zqJIo|P-y-D53it(9DQITWbyUR{qApY@Pa)!@>$$VFz=6==Zr>)q#X=iIKP^k!$
zT(b9D6EC~!n0yXvwWgz+FAxvooEc@j!e5Fph{Dxhtf^+yAFAJf|LFX~NHn5B%Z~9{
zKs<xkTX}l|=f6k#zJeeSxe}c^VFHKL%%6-W@w{Vt<%B3o04{Uxpf&(kad67R%w~pM
zsh$icteucCPG%xi13CF+HT9I8!OxF%KjQJWnD1VEBysFE_BoiX<*y9$z}c?8>Duhq
zq6|<Vc0K8R-U*?Pl^)FWxRpg_ICviPyyNDV3IA%H)O?^?y7_6w1`+EcL{+#tPUnuF
zTFVB`;f3AGPj-StDxn4G5w-sO7xpS-Tx5CWE9iGWZ-{+_*KQtLu59>PBDrzRr3zh(
ze&2y5Ic`4wa<GP38a8poPW}O^Xk$}2kH6Gpl_;s$#PiGlcbNAOl5BA=yfb!*8TKEN
zuGIKG3c&2(i<`66FSgO%MPHFDLQ5EIb;AghLbn_Bm}NUNofv^$O`X?%QM|v!SHJzw
z1BF!z45lP!JU95J;Y9EdM}o<{P4q}s{e48nhhyvG@}p*^@A@G@gb)JYDU4Y#`12{(
zO73a7)SsQ!$bjibfMAYAzIAxB7w)MJD0;xVg%qHMs#M+NteGm|DPdLjL-VNESJB*&
z_9-lt(h%YC=mvuy8D+x9fxN&Dag3hkKo^Lj^4qo|>_>Y4)63ltX>OP>#{KPi|N9}&
z*0(Qm8?K)kcEs$CSK~qk<rsekO@p^Cj2L3h-GG+ix^HpAWU;bVatmsnr;aHH+WxJ+
z!KSszp@DVv&rNXKUe?p2#dg4}?+{Gjxh6PM@z|a9yl{ToRS_t}zYh-|3pqM*h~p|h
zTg#;XV>L_@H^GD+v`C3|G-R##mGwyFI5lZ2FyIymMuJ&?fHNt`_jl26cDD49kTX7P
zB)^rF$$;;1N}mh!kkC1)e)tuq1XQl}@cuvc-YP8XE&3J)r359UM7ktJL>eTe5fG3@
z8U*PEDM3I)X+#<Xq(iz9q*Lkckdp5DCc54Ge{sIJI?s1;_RW4IUf%b&)|zY1F~%HI
zbDt>T(j2j;ip=Z9?9`)Z{~;<*Nm)r1B--iEB1KP=b0+AqTV<aj(PB1{HnfAyn1IRo
z8(%sC6_EEBEQ}l2Jz?LE{l_rhlYGf44y*2VUXgW68mTS@-fWCIKxK_ja8B>9Nzgcv
zcb`~;r>e7A&Dd?C_qFdmT98S%kiUv)nj-0|OD;b8S)0mN$hk(&qstRzOK7RZP2hQ5
z@zxkIQ+S;w0rRRdfuru@gR8%!7mboTJ(zM;`iWByxs1l};(C6D{D>T9nqg>+$P!&v
z(rK=cwK<vX!7ip}H9EmFbOG4Li*b4Lyw1w>&gEgCDns^OVcvfeNvAqnER0_ujv3gZ
zR^R`l|LY(^re%IwO@RqXD`-~=(t1lg;Q}Y`JVBuC{Uu02B_vLCW&|TNrdIHufQa$t
z02rKFH&)`P$M!v^h6ZB{mdl#rvh!Z+WxcPNEvo*$_DDL5R81^8X3QbSD~?3w(?`(}
z^@Z5g;Zf5X_y>$%ubl9cINcg-WiC!+IkuOYzlnf|&d0L$W<g|TW6Z-pwpB-{dF%BY
zBwUxVLT;LY?e-ho@Y{AgXSV85g<*Yd?%m)WOkJ;R<)AMWC8oXNI%L6>AJ29g=o^Oe
z!H$IvDLXnvS^ZaewL#-3jvCt?`*#}jcnY@-y@LVymVr}cy^ipfqDYxSKV07kyYxj_
zUM^o5=F&S-&b=3J?Vqk7fyDq<Ay*bh(FaNYUiUBC7%t=mJcWxcI6KpY(Xk^CRf-*a
ztu+5Q&FQi@Vt*H+GDSJ6lch7GB}kCvKT_4Cl&Uu(SAEGZR~0qbS*H$CW3G*<;t>l2
zLz8}<f}KT-*cl9Zx6_t6C<nR)sRb2p89GWvMDPR~&wu0XeT9j3MPT{R?{P663JC^+
zw@&`a?J3-h*A5h<H@R|9K8B>UIlp)-aQ+xfcooaPFjAFCiSf?siw<5F=4Hs0|8ibc
zyOH&v!)LCqEj?65-=BV?n6&RxSV}*eN5!^`dC6{nZ+$i@b-cMil}o(31HwZx9jQ8R
zCGW$5huK%_8^;3F)Ly9EJw^CxDd8XV!S6m17TYLNv5UzuUoRW##bzm*FVsvTvN_k0
z(p*+LMIRI?zH#CnOLqR-=UUtF(cPZ&P=NstX&`6m2+#7(<nJU0mrE|9QY5(`RGwOE
zAMVr;6vl}t`+LSp*KbkQh_I-9Efjhm&4Ca%zEV745coI*MZoz~cQ{7RIvEP~J((+b
z1;&8thy9wNc}yak%3uoFSwq~ek;3PMGR9QW9i^C$?d`TN`oF5U9J{w3gqJ?P(#Or>
zaq~IG&&nzMyf$VTD0W3<iOMhiRM>)ySSHK4H#YexKHAf@_5Z+o0oc~nU%I!M?A5*_
zdPv3V1~)x4-ycpY0Fc8OceC`m8jq0igPdtaVeChPTFsfVWla+bU%{QO?b<YMJq`#$
z{)(_S-#<C@Mw5Ow#ZES~4jHQD!}2uFcWte$A;u3NyPr9$*Z_Bq^TLIwF)XVNui<Ik
z(S1|-b7yZFj4^5Ip6@>u91?-V$$3~s5P?uXT4c7Wxh$hiV=uJw`udl0?JbkXqIV}l
zjnnkW$?(eV<~5+|t)BEpwNQXw?WXve(uW9J#;21NZnS06w$l4A*T2X+r|*-v#qi*S
zk)WM82~Iz2RyQ@K_GzBBUeIm_`H{4kLVP>63sa%s>2)+99|MRDINP_s@3i*?VerlJ
zZwbm_AqTt(R=}bQx7^j7l)wFT8Wm+Q>PN%h#QePI>(jILE+$@#4Z+&{nv-<D1HBDz
zmG<}2Q`EJE^Kjie{EEtQLC+gdq<g+->DN)aXZfPUA!9d|n<Tw>6IFIczB7tARc=O3
zUqLj{Z<wEvR@&^qm@JN=3fZ$aW=nTmc<lzGo_}k90r`@(NfIl_g`(Sq73)mqAA6!C
z%3Yzy)4vdEZ>|41_cb^&Aev1{?9|H8s=qP+S~32zG+(48PnFTmNLNYx&$J!i(MILy
zPA>CLyFUs>Ecy(jf~8p^1#TSU-D^A3ySvaD+P9IuO7pSZ?Lfs<QaezMDV%ptdr&o(
zsbBHUBWGYWt+1dk<Pt;pXh(`fviNHejE*&s1hscf?JgG8r(6uZ74_0bov3DzSWZ4v
zF_&aNo)Ey_B&~K-s3GNr06iaOW+%a&(HUN~jbW%KdFzYej&GM5Vw`)$vCP8Y!aFw(
zlL6($z{Jq}v(tT6G}J4#;IOXmf0Fm>feov0sqb}6DcaJo_F6<HlxagDVng%!J~WK?
zr0<_@_oEB13o@Sy?4HP&MnPa~DpmZ#Z9CBl)_5UU9~@W$wYvlb$mw)vpL=@?<q;HV
zCTWfS=qYF@s@;!{4A!;jissms^M2x72@Q-n8*ZgK=4%9_pU|95l5(qMc<~bCc;CBU
zX(yCz`@8T0bk;NLCB>?cW`Lu*K-b!{33nSO)2Wa6!WE=hN~-02j$Hjk!={w->FXb%
zzM|N{hIUk~F>@5mea%U%D5nx1gl%ef;xx_rq19SP!Rb6XN<;<A@v2u!jFze-8SW)3
zOb9oqQS!;EX>0=w&77qm7kdyKKA3D2-!9!!bCJn)<{6OO{$bu`?4HHF4J|Tx5Lz23
z8qoBd*$zadk;qUUHHhFnB1Is<)9+asyBU8|ySd^zX0Bteg!G#-oO_Qad5PMzvnz5t
z$2*Hf)eP*`v9Q)nUWLuph@spFO}lTQg1*(AQcb%ieO-s|l;<HxG5_ebYMC#AMN^zc
zzWgAA%v{`YN1NCSoC|Y^c`!_3>!fr_&tERsy!`uc*?mV0R3AB%OtjCFk2#7T#MK(`
z7B&Df@}a<8g%0iG?L&4UMPow~<(De!ikNA#q>t;BAhIFfMW-^tJ+Z!{)-wK$BzAh<
zh=Rl5rNJiR<IS+qL#0?0bZ;O$=3^7&7i&0d*p@3ZvNa45*!9=ntdI@N=*Cu96f2!)
ze)%=dOM6z_DoAPWef{$dvmN#Mt4UKi7bwaDyE#V*7caF5zy|iwHd_3ahw&Mn>%DdD
zA0@M13s;5eMnCSGF1isIh_tkhr43%o;*y-Y3BxchKR)XcFTJTQ#6Lq$Dy=*$EhT~3
z)<#J!TkQtz2oKy3zI2ueZ!W8#>F$8nxRd@hC1-a+6cMlFVBs$G=H5{l4WUda3cANg
zEi)JP3~m}>>&9ao?hM<+2X^B{Pl&~Z)1px$gRbhz+;((CUwhq|@pbnKHmdw9%r{5C
zjoxRt4>?T(r0eYpV|`O?t`fz-e;X!!$9^s?HuKVMQfWsYiPa1&DIN<EGuKaU-dh!7
z(k66jdScDSim}9F&~m*b@B8;=Y^%AFh!-1O9{B-QxXIsllISLja}1(pWH1LEj0F@|
z&~x{S?L>dFzjMjHq!!@B*ORPI0~8~8?HEb@I>n}7*1Rj_4@&t5n!2+MFN($%cZ?E6
zrBdlee<&AIH_O7I&+!t8uWT?Eb-+F+gX6F5!MT^yTV&-b@FvM5jq8r#JK2dKJE2*}
z^X8)`s}hLJ%duWGI07K<E~(J5Yel<1lQi{2P+{0W3d{Pa?UOtn%LAMU4{Y9}q~f-S
zXv8$(%u&P{+4g9#eeg<p$4`Jm{z~++9{;A(Q7yytNkyo}qvJ#>>T5Kr@hqwjo0ypK
z=OefnKZzcn*(WPBR~}^}@`qBG#GhlS&m@<Nnhly8TD?vrAH|oWi)Wcr7QaLIuj6cR
zI2>qKp0mPn3dqD~CjkHWdgbQXt$l+*$|f2qqB5Jfam=SyA+sxYr51KbwTMa-*SWn;
zpHv{5z38`i!o__BJNddJ##(z-#0JT*y5Ftb*(YLm2pY4YZf8@t{;Z)6y1TZlj=PuD
z^bRX$dNlmyJSBK%g|lnx-Jf)t%o%IqJt7E*N7F<$#`D_q5+4s;UIS0=p*VG=&-XP*
zDsE-#B@a>FGt>*DB79}g;vO`8%m4iOxdfz<&ezXHQbf0I1_kA)+k3@AymmGTNj#60
zm&Wr5{dfMB4uNncuJS54nRA#7(l?LXZrx0vxVK8S@^VOm4f)PI1AT;FRGrm{_Ce1|
zTgodY6so}fNkWHfPKe@ws9rYSs6{Yb(d;(OykjG1b;NUI$wsfNS-Pt=YPS(t<5%iu
zD}#G6dhdDggVz#?DDE2V$;#PYDw0!Q<0hU}-G1;wGWc+r@3UZ6^?g3`u7cCP!~wx`
zD-YVhE9VlYvBz9@Ej3B-e!ZRW<$_!&kwsX$;G01}OGqEI9jFW!vLZyzl&6*2%daI2
z@2celr+oYMpv|NJTPja^DF9QEPHmvZEO6xR`4&QgmCHYQYmoIO)Nt01HRIVS%|#$H
zhhjDA=se+^hgu$J+U}D$8$EwKHPk4Zd2Uu26Mvj7oaxVV4MAo;;v}ESwQX`pqHjg-
zAsMn3KQ>9Y;qQ&*Wd2jllmY!tdj-A0OO*KQY^4wU8dm6&ymiIO-u9-EO7L6nA+Rx*
z43RNo3Xn2iGl1Jt;TCkE%i)>C_Lv5B;*3O-;5BU&B>AwjG-Wcn%of(N<=rOG|0ee;
ze(_PKu%kJo4*ggPv+k@8Z0k@rE-^f$e}cZewZ&ef!~emnBWKQB6qSgaWAjncIE4ia
zQZwOw`M2ZGs<Xa`!9=g!3;!zF#nu`Y4;w>oXV-Q0o*IdRVyW4Y`6BFgFlYbOnp2{v
z`n;+|#U!z}>qY<>eWMu&XAngAeoCvL2Wk+4ZGt}qniFOVspIGp&BKk5==O(gV91rQ
z=hNDWuGwvQP5+KnY>sS&Wg4n_L_f0<%e)OMPBCR<_#TEj{!OHmu#^vp&w{SH7o8K&
z>u+mIm&YUgJkXw?t9kHwJ;rZAR0XTv5AyIRxUJW2o+P?Fza#WI7h9p3y6L#i@(dqy
z_B%vmNvWquB|9vcc-;@#RmWKG5P~u2@6gQC06hYrdsXa8iNW*^`%G!^T0zk7j2>51
zvShljuY<xH(|4F4voBu4=_@KHQLSV7aad_l1!t}L&g1jtBf)2fUnmZlnTiKhN1u(}
zp0>8W!8~i-I+S{!jy<|sEIQ0|@{7yS-JOoRT^dDZPp$L5gCB^Z<QlrB9!GUn!px`e
zz=E=-&_ElsXhuS`!M&`L`ctfF($lP!C-|s)+|!7NFgm4KPUTUq5a!_z?_i3YcVj4L
z_-?s-2;{H#c;`!2w^`<WQaImvS36%*fcr9%TDnEF`Wqw{lkWC}QY_uVnvSM^x3O{l
zj{eycv{6XI=J~OMgG63_X!T~2!kt!V29k*nly&_-z80+02`AF~q*p{o7TiPR=+8r)
zQT_Hs!Q`_li7P;Prpu-ffR0S&%SH@*zPZ<&zc}<&3KNeF!DwgZufWb!s$D*!X%cPl
zz3`maUM*X#i#!37+Ro*pYRs1$J63*=3aekZM9U2UobPz3v{y<)MlA)sR&^FaW8JUf
zzZ`+FXRF(b?wo)dPmm&vM23;dNpQ&Zp64qI%)Wt~f}eQG*!Zzgm`+QtVkkO#;3(eR
zb!Cknarqd(EJSqegQ`i-9Jd%LuCT(|3tlVkze@3RtLpepXjPUOjYKk2&u?pO2tg&S
z@xTinNS1&@i{^YhbBeXvoyAmvkA0*&iE8}Icr|5+u#{P=p|lamJ=DBgXg`>pH1*w=
zZ8P3y%+u58mzWIt6>;#tKwwKXZl|djbR{Kl{OIufgFRUzFS=;-9gU4b@qI@3-mRMo
z;hi|aloV`i^Y427NnNGqK1}!3+ZrgzP#P-ZL{x-q-u&1c>lI%AB{dTZeZWpi9uINk
z6A8xJ=tfBkj-Jtwn0DIZA6+M0tDRX_+aY0?M=>QEdi95rvkURD<=Z`T*@&`ZNY7<b
z-l@+a9l;J;Xd7Pz6su|C?P_T`F%n`-YXTtm7y=;(zv@bLJ~Q2}urwTu*IfcG!<#|v
z=tsB&z7~_DW)#GY5TPysUQN^eXr7@k`>ohg_uHv*HL<*$hB>x950s=Rd{6N*6N~kq
zwtPpilT_}S=&HK6jJaGuUXtkDf{10U-c$9g+2>tigNGtcqgks7kk$iVHHpga@N#Ix
zbd;h$_tlbDd_vN6QgHAgP?NDz1yF93RRgB^EZrx(g6-o(BbO?%-#OE!g`EPn`NMP$
zw;!gQcJs$Lq_{<Ov+#OmK@CW3nJ>8VpW5as_kVtFs87SjVB|#@D58+F*cmframd2t
z-OFk`^TSuj<a@g?xERV9e5`0$%=-4Ba@{H1(j6v$-8C-rx(8S`Mf$gFX`{Fl1~&Ei
zRBpty9o@x{=V~tfeoqi82Xi#4ST-7DHjb;oLg>0EpQr+@gZE*&tb4f&6<Zeu7fS}3
zwD{>}+cz>9*PmlQWd<gJdw)W|Av%d57t`e>VptW`cW=twnJ_v@PBTi>#$SPj2<81>
zHdS*qiQ)|gAM?jjgHff&MbQpRgNQNTKYlmolC4Y3+xI~rAtiKVw_$z73^nPi{tymP
zp`PvzzR*DYVtO?9te48c9dG*1id?sUt~%&QG)jO>#hH9dC2KMV9k_5r7ZT3QpDI6i
z#*cobQ~`jBu8GI;{AmF+u+5e>xd$J6ux|LkJ&JH=z5rzqC3mUxLtEG-V{8oK9mvwu
zlx~;(R9~g62=`+OFOy@@)4U=ke(Wc^Idr#<%I?%`K-GB}@;4{#{-<VJG$CzNCTozV
z@H&0%OON*HsNV&e@S1PoA|N+KoQHs5sgd6QGZq&VG-JyYdFy8%4n${5589u7rNpHa
zqm_X0fFj^Es$+qc9Pw*@tHqF}_jUruC&f=8F|WD@5p!Yez)DFmA-3?H!=vpU3=M+U
z0wa@Jk#8!=XeTCLb$lNXotX<65CvZ>iiy_+*T7*CNZiw)!<nZvvGD|s8ss%(ll7K4
zURP(z+lzgo4<tkLDFenFZHR9^G%DbDNqst{FNeXSFuVRJhZ(o;gmh1o$cCbXE0I9h
z8I?ZSdbEZeaB3npOhMvV%O)w1zB8)vbs=32sJ+Z|<9;S_Dx-5Xo4W~k%X5^mk9AvM
zAQ@OSQ~4q(VxzXHLNT4Tjd}7)1(@L|NO2itE5nsLid1k`saT7bdJ~M|JA%~su2dLO
zc8CmX^+*$SD-YsbHo<=B0qLQT9z%;_!xmt)GNVz-;;1F`-1c1xMkEh9M7YW4Kdvv9
zd{ZJiZibZn-3NkuzIQR^misyr2-PzB(&vB-DUBVI<KyJU&RKclAp7&y@A391q@4_y
z#(c&@b|~AUq*cUis+8Fxqf$5e)m}QTgAAqwZ}1=wYt1i*=n4|tLABO%4Y{agpm0dI
z{`it?bxa$?r?SXdd~$cj(Hqm-y2#fP<M1Wb%p+(nhCX?wvu$RK8=8FAb=!WLm~nDh
zF9@Z*&v2NjEL!onxf_sNqnE2*!pppdumMJ{Q+P$!Os>os%Dr7l7yu*v1r&x5WM4Mn
zZ)L=_O}yOq9b`9hNQnnVXblywm6)V)nxta>o|#==Hd0SL6ilJ2m`<mVCOU~1Ns?By
z4b+POU<=a$LPklf9o|Z%^7Rd|2UeFPA=BJLr-=>TjYIy;`xl394%zRQxTu1!I&R9u
z1bbHi!SB_uAyUg^sajoa=aR{UhYq?MqffLPj!>Zj0}@~w;Qb>rRp~pfGG(MpMR*}F
zI=Ko^9f}+KX3Q%}-K^{)<YL-kQJ-(&xfi5$J=4bCX!N)f(~oVvst1R>0PRP?wP&|U
zd2gQ_a8KWNSv5@SnbrwD&sDq?d2>%1ER_DhB!|W%vNf^CW2IAM4_(+zS84VXgI^p*
zW!EwrYsxYrI$X)GUAiDUyahGHyBu%SyGzB;X2pFZSIG7$=KZTF9KmdD6-HmS2znLq
z2IO}^P2}Gi@`v0fO!d-%qCyZhSM14tMx7|$`h%DG=O%lR!4)BON7v}?9}n2r5>B^}
zu1(162~s}(hG}Hx5Qo1T%2&+hAO_j+yR~;hrJWC7b$%%>>W#1OoM2wCe7hrH!rx2=
z27w_+)>D!0!^DM>M%cGHPHiRd(AsS-K8Qrr<f&?hyzL9&S@>tX`R2M4b|%qwW>1uR
zo?`h@djcHpou1bG2mNgqY*n4SmjE8rDYy(Y+k9T${FvCSx@<bR=hEPiy($Y%H4$oS
zg$UkZZ@@;%Z=yZqxf9tWqPMfD$na~gHwcDN4%M)Vc9~0+EGy^<{Xy5{FLDeY-DKQ(
z%<acl`(TbV^`%Kz&oq)>4jI*5>yYe*B*tG)nVr=2agCIT0^QWqiRSx6qHpQXZ25v-
zK~cQ)3uEE*p*Qr%Jy8Q@hP$%%{drHUC(0Y{rO;K0S@SJ<zQ-utxEsN|BYAR{cC8KG
zqos%zMe!0JdH>9KpAtdEfqNjfDi%E3PBz;_h3=H0zmh9uYB6b~_6G-0zf%m2rw2LD
zrmAYkDb_2?vw2a;3+d8k-j+;$=MWA3B^7MD`Pv|}%XpDA5fup%D<-t&JD^<=(|^A9
z{CZJ;(5+!ZIMzRKF{PQ}0VmPFc!1RnT$o~yw7g$V=wA3Qe&SGO&WHOu3yE{IG|AjI
zLi`N?!<Qr7!)wZXkSz!0i!C+YNVe=~P?<gecHJ}GVT@T#I^aB4s>4_!$9nEqU7Clz
z5w;{OAI~%O1(jqL?v%d+=%-LnpQ69Kw|;{mWSla#FO#Xi8QjFTDP|l+fPer<$)cuw
z;QQpI+w(foR+7RU&nP)`s551#qS$v$-{fvqP%u4WF|HdT%RuY{Fdb%6o`G?^xngJO
zaJq0JmBiI@a20qZA%uN`>OE|sAJZGJx3Zrp|G?koTer=kv8C~F-8+OGP<ZI%9!q~#
z_vdIOnxT-Jj=EciFng(UiKUvQ?%zQQ9Ol8p;3i0N+{LK7Jvm;-!(TFH;O`{*77DX~
zov1CtLNSFEq>?UUeMN&1hK_hmeBp?i2+_<O{QN)XQWEu&lZM@cSES^(aO(K2Y!o6n
zHJG^`*X(|<)2LlNu}t@k56Pq)V_elLg{W?3fUia^MR2Z%Ow?4lK3=N33UUiKcHpi;
z7%R47o5^cD>_IFHgS>1pd!_32qr}dMj=-7YhWQOy5?PuKQ33jJeCg7s^=VfI$-j6Y
z-SYMFPU?I{gT~15kpP(w)v=i;=V6%&eMI=UgUisQ8f^&)DcfzSpC>z$5-4vKr(W7`
zbc-6Hzk?G+t#^sIRSOPBs<3m6c3BJo?r0IO3tX?bz<Wbwk_tPhqgl=A1MY??-9K|l
zT%JSiu8J?Z6Q+A?g)8;9D!fOGAhv=O8fdzTe(|>9Ly;y`SE%28;MkeYHpUre7&t>;
zsif*%CyCI@2vj3HqWOu-4~3`t9fJmRR>>aejqCv(ckT;OYZc&-TbXp9B*mC;n0vcz
zKJ-je_3hL?`Ed$Ub=jq)WkmKztz7Cr;^PHJBkL)D7R3ZHF)ybk(oDWAH9ePsYP|N-
zwvwBi@>NkFrCaII6BO)PGeJZrK4kx1G#K8@qk*32lVsSZPTV{l|D@%;i}=ino^$>)
zBw-{Enh3}K$22V#B>r9UwDF(*MqPB$Mh=pj3uUcX48T_jS{~zA$I0-d7o9`$4Xsq(
zMml-Nwrk?ZQ5I39k+$^6i=55oaHHbmNtf~Q=gp{JB~rj)GGZBpxuqw}t5;1T?=XF6
z$E(?VV%ll?LRUApc<FUmoyE!H#kH#ztv5g&w}r4uJ#}h)#+8%>%0N4;9jrx9h5pa<
z=QFav1%~E%Ly@7f^d-L5z!g}hL5X}FoDvZSBo}fmz9?5r^?^yImyx@DrNeJZs2gL_
zLS<}{K>gGIS_TD00!N_|gn`C^({{5UI`fb?owtEB+2a<)0oW6^Ku#!D8!cfFY=$dk
z>a!bnGVt#PC-Mdsk=2fChx=!;F&CpRC@>0x65jV;yF&Gp-GMz4gD;HZu-DVY)9`-l
zXo&TzTfz!veRr(lnen_M!I@MyQFd1WR3plhQiOFG^BE~a!41Ov9|EqVc~ZUUw2CIj
z^8kME=QYgfl()Jo6UuQkwxa8g@LQz@2HDJ}Pm%_IQDzs88PM7BylQxSKh|2yu)fan
z=&7b*xxr6ev}X>ydv%iLb;YD7_QR_b4&ER?rX`jBV8hY}WwnP)^&3ne+N7x%>bPua
z|KS`Cr+A$~F#wt@-UpN3545sVlu|AnHE|`vA*2&KbAA8R$55bja<u+RcRcsvM%0al
zkR)NQYNUw>W-D=l=)-(Cw$2nKuW8q1u8>9xVP{j~1U?RK67&_pFmgpJN1zQHC9wF}
zry!<a>A$bli~CE`VN#~%Ds-A$`Oqz6%V)p2Qcr&OP|`A_c5STcsKMC%#TmqkPg+$5
zPIX)-54i=bo4yiC-4*i(^8M`?VGD7XCd?QAom5^73b%zN6k)q44dM&U1Mp%g2sR3J
z4BL2;$$555Go%l=ih80v$?T|&$D3_>-4UX3ddqll?5Z(c7<SIo39@)(qT9~mmp4Gd
zLe7x9>M<muB8??besv<Edn1}lPMBuj@&5X30@KS~i#;1=QBBcQk4>CS#5)%86qtq(
zkWpP3(7EELlp8DGj5`f{P}Zc&$YpXHIk=2%^vkV#)JDIWHnV3|Cc5hegf0DN&+2=;
z>aybhO=OOnM^GHWtP)bbwm}Td1qPO;(16g-L7(ToHB;AptBa3vg2hW2EPU8iU4)+H
z%DfEy@=rkB;f1pNu1PM?U3FjhVa1qwW0oY?(jVQ8OsBA#3Gm&-Sbhd%9khl}^A(L7
z@j$yVGm<Z$`>m&?!`M{Qp7Y6YbS$~@No|?%g#)5Kv97t@3Nk7h6R#0wO!GRYPbrmC
zSV6yR*sKjdjU&y9m!`4ScfLwz75mN!Ax*L(mWnTbUPB6Hz(sDn+C?cB*Uiyq7kT&J
z&t+x+Scoq2u9rws-wV@w&Oe6pZ4d*mP%!*bAz|qTilBZj2hY0zARoQQM3uQ9o;+gS
zh8{PH_mYUOVW@OJzCKT{P!4mEAVL?^VCAINv6b#-z$OVtvgy56Vz^3tHT8Y#g!E_>
zNx3Zr=e=cx-th@7w<lNg)`KR_X3^S}?)nZr$djMCuZ>Eo!0oa+zhti|G>JYn5U@t;
zEsy%`V4G#3r0m&+<L(nT2Z>zLv|QZICh&;5jN{WU95%tW%i*>bNAIt+^?-?;vX>sG
zEe5F$2r~4=lNRq~cw2q6nk)k83~Ah|&c}M?(m>s;JzQvgKUQBzX<|@6%9rh4b;&cA
zp>4^9uSBKmrRAES$QUlQQ!ttLgIa4_aOIwbF@@MlO}|lkRDFG+oOHTiM%jK&?;O?`
zqmxoDx>PPzuY9x|;l;MRs-h&?UZK_PNksa&@}4JEn{Y*@zJj=OzGd=-$_$zQfOno&
z9@t`PhnQBXv^>_;>YWik`x0*3Gs(fVZNOyz1}~j);d%Y?(iaR!h%j!@6Y5(+7f&*N
z4=)mI=jD-~^`v%KcG2|&(Mc5aFDGvm;beY?M7K@l`weqpBpS<yQTrfCMMO;*46bxs
z)i6{v1Z7cy+tKm@fFoXhzho*6Gz-wY!hX?2gYI$*aM2JN{cM(3+DUq#zi*)JM%4G#
zcEMj#8Aw@yu~VaVP%>TVUbjN667P+1&@*>kt2-UdfRO+ZQP18sq!LT16J}k_`z!l$
z6Glz@Awd7AaEIMay8ghpj_&T=i-V=3pDbv7(H5I+<np>1_BXBSk{+@wOY5KnI)O}&
zX;n*QOkG~2d8!mbCfe*5I58NegxAe)Y?d$P#59=)azNMLM3L*Ys+)>@1kQ@l)u#1Z
z$ZQRCWU7wY*OQ#ICbu9Rb*6lH!$%%<Ref1X9-SsY5`PQf<P>-#?r*@ZB_$lIzc_KP
zjy@ot&%Bm|RHzYmck+IToWlEp9lZhmtYuk-$d<oF@X$)qhX#lT>NpYCac-7;oDhMr
z;Fo3X@+4GEkT^xyuoyhIdZU%&<;(W|<(OBg^ry1abUEz-zn?d-v60ZcgD;#<7qae!
z%1~{5Yu`Ao!9efg@0^a`z|pBdUUB(e1bBLh@!+VZg(q#4w2%_ts|`@-3R7-I#(sh_
zsPWn~`F`*OC-fH6qUM7^irWqvmB=Si!$vEIfB_aJA~pP`&dPI)1meHQpyx8hTXJk1
ziZWEhL8z|Zuf4nN%!LL?BIZ$=AE$hyAD~HcijvC$Ui_GZjs<EzDn9L=L&#B|Ogn`g
zP+ZepgY}ZISn=8V&3-ZLoSz!^?{F-hzTa`4zWr--IRA?R8dH7rTMEqAxG(t1at|o2
zrZWb=1K5!ne<L||bX5Typ*}wT@`+QdC!N9IJyjKwuTIYCEHX~@-n2?p#k#%g{KQN9
zd}th5p&`8ZD3bbxTXao2Jul>MX;I(<6V;|gf@vQLCHG&Wx+mK3bhvPHFi^87@8X@v
z)SYib*TEk`ly*qE%Y>`N_YCO%jfd5$O*QDfaa4n1*1m}^&GSBpT%??f<-E<|En);J
z;9ZwD5-wVM_nc)@y?)$Hw(_@@_-WUk!;p{?lL7H|2DJ8KcQ!;^dj2R^xyAdBP_hht
zqr0TH1ji$~34wTrEI4o6S?i@^ZfFh8{{H3`hLI-k)r?6`H*#C3a%vYE2qx-&(glY|
z)KT~j2T;&sT-^AVRk}#?k-vz$nbFrECxl&*AXAR9%b=$!vVvZO5>{`Uw}8-oo_%5z
z&?i7QZz-e@_9*o=DS<tLr=a;U;uUEfAFCnHY$|8|8E(aD{B%`G?d0P5>Nt;zQ27*K
zN%iPt<nJQi*u)!&d-YzpXryc9<-i4?y=JjJd{`Tu0UxQm%nNYW2-vPXGNn`Ms<~at
z=D;SVuh#wrj~)u_;j<*Z>gIsue*9VrUvFIq)}ifs0gn6h6o*HJJ=w0xXJZMNZ|)}>
zSgQctYDUPyS92ZNCDeoD-|$r7^X5`uR9i0qd>c9pS}E=_E~a$wcFs(2s)2ru5HhMC
zqQ{(I?&(4YD2O~={0B;@iTx_61qyl?i*~&m=5UUXmtH{vK(cDS|EAUjlnPkQq_3%m
zlfgOZO$2W}){<w&^}_g(kA~l~gc-d7z69_0t<Ak}5Dcz)Up9VtydB_0afpchG^{|h
z@Nyw)xMD+ZN}kaD?E*9~I3La@k5jm-qldn|B20vA32HpEgg3LEkh7s>ckk}Tg&2zf
zxU&k{R3LQDQFX-r76$o^_2gV<hak25j>s_3o)I}a(ds}QN@=FrH+T(Zed8~GjIN4y
zjI&^U0*)fz1Cpkd>eW%F{O}ZsZ^hr7fO%U|Ct<Ck$-~eTlU=Fm>*{>%=r-O9((jdX
z-a$Y?rp3$5<p4aHHu7!_^(9%gt2I$NTZ)J{$G#QGP0~U-M=aSmxI#y3ilr^3MQ_L5
z8g5W|=3i}08iS1VDo2AZBaQV1$Ire_7u$iB1G0&1&Z67fQz0L*9g-Dpd2pN>;n)jK
zpR-pPKj?o&o91(!_mG^|`QV3kx;0&@u~A$HY$XLyl;77xGLriG#mkLATEfBnguO`O
z8D!=Hn5WLIE0jVjuFZSt(Oj%08wI1RD7KLf41pFhe39bqD&1)$^9`jiwbtjsVVYG6
z78NQpIkr1><$liSw?WLl()Ujo_#>UAvE#nj`Sn|YuFjSA=q*2rZ?9vqkx++(Jeg^r
z8#E~ajXC>Z_Q|KCmZcXul@X3@on7WO^RwxmgV*#0L4R}MStxeu{ME(=lm1YKl7372
zA0A$q_fVQ-mv0u;UL{feu7fWE)r+U94eLm$%DmsYkscw@hYP6T7mphGJ_IfCDxk^P
zq)S$lbt7hSOeZz_7aJWH^G_}o&Mwi&+D_qnN~>V0%BHu2q}JD%-zI?}Afkdsf_VSt
z)6*Bz6S|O}jiz8^UN(T}!PIzf+M?iYOP+LE%$U&R;<O5nRag9AMzNeggo7pfyAmsJ
z$q^WRK@ij`3MM_m+Zj~!Wuj1nr^;b7SlEosLWNz2j+P;&aK(z7HSHhR5AtGsVV7eP
z^r9PRpNa?KLa{Qn5Ie3&hCa#$fmog3hI>iw;)P65w^e`H09P?r1P%w_#msuDR{{nK
zIKG7(Na;Awdo}MjPZz?p)w<g{M$q~xX7N^kto>IcA&Jf=yPYBNDHdudmdaj6*5--|
zGC@ZKfv3!j9d-mZdI`Ki2jAcO1djG1H){9>&DEq5lOZ!g@2eDri)%52BM&~(uPUpE
zyS|d97;p^E#q=IdUSN@N$u=5iP4$2_dx{=VlMeq-Tkd8p;v_4BoAOwI0ujl#*B86(
zb4++vOXOH`pQGGrSv5IL!;gVD78Z`Y%5=78WkY)RDumI|69seEfjf*a<`25#Jq=jz
z8tCMN1Ctv2NtHlYm02pcDfplIxUL7ah}27iz&Q4%8DX<Yqf9(SfA~nKxPOrNTmKws
zKB~u$+#bRCGZ1a@i^sa^jmssU1ATQaS{{L?uAHuX_tufBk_w9LOsA1=oG@=tU+tz>
zx;KJ%W<Fv9$~Q~ix;Airsk(`D!~DALU|1{dcWkB%yw5P{aOK%k?1+>)G+)4UX3DwC
zybKU~Jn4X^yCNLVTYpUtw_p?b!SmmEGz5|<M$`*Ngts6Wt+ZS1rv=URd+DI84Q)@c
zesui$QoAuKYNmDNQJX{q;SO)#oQP<Ow`2IK;oJcL6A0m}6iiqUN2hkmy}>Iynf-P4
zVK4j$r=2_MRqBlzVk$vjr?fcf`a%#w3z2P0f4p^*^omqHZ2^;+HeQ7$^^|yTkn#$;
z@ZJ6V{%aceTRgKV4|yM1Le#&=@GPZ$YUR%oq?MD))X5n_2hof}Q|dRE2~1VzV@!dw
zq7{W2;ZZ$d)Q3zpP}~{ww+dN4g*Z3nVi77zX%x;YT$kd}t}H=t<d;v1Sz&%KY<7Sm
z9$Ixn`vKm|uPV?q;0kRh=Rm#j9V)Y;T@AWKAG&8RYYMnaW)ZO)zl`r!&BzWWX!Ge$
zpwWa~N-UqIKkz+4befOkem%$1l4De{v)R3a6ET*|+hv?_Rpxj1Q)dUJ8G~$fyc&#c
z?EM16bEt3|WiU@jLcE39Exz3X%t77XM6X0FG_X*y9Jpjr)HIjQneuPg*m(|`<99b7
zQT)9222H4NWH|r6a=RdP9o(EiS>zg*^(v}?j$;4%azR0oSUs?7W<^7gwhn>IDM2l_
zVc1R<nlwaBxp)Q#U8&!Uo=amrohSbm|If_al?EjR7%V*QXgM~|*S1broDVYluqv3%
z9WpK@b1T&#_o_y=gLg01L-sp(cv%d@-%amViGlsMxT831(%E=<PyuMh9&v^9eE@b}
z*&+Y9`<)Y@T28#nyFio`@I2!v;CPPveeu!k<+w!(Wthznf_Yn7JDEV3=a)A%#<REv
zC{2t=?f7VTADQ?u{Mmztv!MwI9$T|w@xDYvL{*OKZvz67ee(H^Q4S*Ikj#?arEt9~
z9oIXeiY1rf_H)Z!{glV(+lmx8f`3(Br7D+Z13cwOqNgu2Lp#R14@;CyvOcFr=ZdCj
z)2f;Bq#FSWp(f#HcLbOZqi;LHMDQ;Q05j$t)P{vuap>U!TLJft4-P6>q)g??U!#R&
zGnx4DoY06@WYK)@X?|~`cthmI_y<?UtZn1AotNy2;n-Q~(J6lZWPKZ4e4-eXKD(dn
z<1?v{JalCzb<0gw`sRBjXp9@vaIL2JK#Vid#rkt?i3jo!dLokh6@4#VxJKEM!^SD#
zs5Q*Na3aWmB`d?b+<tBmZSi(7(CTDAEGk7$`yfK}{)RL(AqW6%5nnU&#y1B&vVckN
zxykU?bVU0Tx1%LR)eJ>$I{hF{_i&G0P*WN5#q$4IdpGyckPvHd6{vZ6u)=-#LxOx{
zjLGLM4J~c-$B$kzyobj0$~farE>nR+SiEZv*yzy6iM8#e3#vXyZ-K+pD{Dpa&m)T2
zq7bzVfkouJW0{cgV15x)RSG#rYm1{!Cr}4&9M>|1UEvT#{RTC)A?e|QR(nMlChd+o
z{n_rTr2cEr<09_=j?<BKaCw0aQxWw;#AWL?D^ox3$GEt-dpZKx4DzU0Wc;SRG`xqE
zqL$Y|Oovn&q$Ri=cU*7okL;7szU%n`dAqCC<d+^M5;>4F?o0ObM##+Q=5I%K*^7|x
zT(%Rd(4D*lrMfT^YTH^R_^BIm(D-D1RLfY*I`q7xNY)Q1r=ayZs8K;i&)&=)_%b-n
zK-c4cVexk8kJ4(;jydLVKb_w?wp(?OOTa8lg=?nvDJB`8qpZBVygRSOXtA&l%Gp5a
z4PwsS&CTIc6N)OgZQiS%T8E&mdH>4R0dwu0taWoYk+bW@<U_ZG!aT>b!9AdT)6HW)
zX_wE8qZn#Jt2Z?U({uF#NVT2FCu^MNsKV+2KfFS9W!AE}E#@);F?$f4Mt5713wRth
zy}bvqltEAPQ-BU`>id;DmwtnbDK-)d%LyGXy{3V2b-w;%TXP;JsV#K-@HxA39x-D-
z!qMn5jkM>xW`Z2^5YstBB6Ogzxhs|{asIueyu68&o%(dk5p3HX5vi>?5?}Xl3Ad-s
zV&dZa-7I-p72B$qe*XT%w{K_3TbP>{oa?OD={LQ79Lst3v*#A4xmxq;yT-;x3LisP
zulOEYPB;PP)>kHn`m7e!O$q)}R4mkX>8QBdWd8O&7STO<)p#vr%h8?}SYd#zZ{E7)
zpz)o7=dO!MryT#Us#L4dF7}+Go#sTn=Rc#rHxdV^I|VwpvaAGj0XJ25IIOQ;e%F<l
zlc-EDj(@_{k_q!6rM-9M)6FeilA+G7HT_<(G|`$G8;lErEw0&xfwY#$BNrbk?*yIG
z*RNl{tH#;d+F~|k2~16;NlGOAW5QC~Moo3qofdOeqmr}g`((*?o6Xs#z@zPjZ}&)C
zj^}~}Juc1;s_z{ii{#=$7?8|minBB9wp&3xcs7EO;4!itD2vt4P=L#twWxj<xHt_}
z`YvN=2?h`K+KcUTVC}`p(&eNs*8Jvu-hKb5;Nk|oZ&-`&3Zt3q`|ogvVW7?7fVI?F
z|6VX1Ne{2rK3uy6=Lf{K`S**&vk@jJ3%`()EF5hhYy=jjR^Mws`D8WV{sgbo?Pz<f
z%ql7!VvlBlE+VO$J0en=oI-jY>vZdmc=tGIl_u1+1x7PdrDrI9Sua&ky?gJ9q*?d5
zq-n4+GPJjLw|;F#a9vq56x~IMrrlJJpUxTm!esy!m2F(qi3#t?IFB4U^6x3W26|RR
z*3h>nk?tSE{i%kHteTp$;V%Xl_>8Yn1q6DUog3$P9aeHQO3WHwUyEc?^^hA`p02}!
z5=#S#mf#Vy8cxM5kEZV<e=pXBpr79R@7;-I{CA(pyyWrcn|rmu(g7&l!}GcOlPuNf
z^fEjjjzQX>P8Y9pvjQm}1X&nJ`4g4<fJt=lMp;1fr$}<iGZo&ViY05V1??#(;$!wl
zR6M}Z@=B6IJ4dIm&CX(%4#&*&G)0HaR@}pv`(B7wJq~C5hlhuk)?Q)Oes(^0X0zq%
zw+xOSry#8}t!BpJBmW%q{)6Ma8yio-yBphLa?54qhGM+SBn}Myn=fiza%ok<)&k9D
zn|5w=NjqPZgmn2_Z_rX4T~P0zW|>x`qDhHA)B5VQ$X)s_FWS&+h7PklMHzJ>s85j=
zGBs2)MB(ytS<a`gcDcBN&8(k$y@|k$SB?1_FR+vH69Rkr^zWpDq$|Tyspv6aa)tJI
z-eCvGanuzri#*+K+s!hII@6Z>;RTZx6L*rjvUzLjY+1opK-YF+BTE~k>rVH_^jF@=
zphu{<g0yD7thH{lQDi*^fFDrVruJzpLSw_FmI?{7@&2+OLz%dz?qeI)Dt?P9SDQeU
zpD_@pveENWM-1@EG@xZce9${w_GNV8Zy%lRFT2f^u56I|$@s{s&zARm0S5$dqN#o5
zc4>~7gAux#xpAG$YdLJ6w#DLhR9E79N)<0!`?|us)1}Z=gp@bYc#zV#CiXVW!y8C3
z8%i>%Lrr+tb?UNNn21Aa_$O20>A18J=JA%lXsVWJzcoRb@@hIqw5YH<B^6f#>FGAQ
zCVscHUTN|uVfgJ7sza>q=s!Uk1kX!J3Tj;`8jz|8Swm5g@3?f)3B`)3SC$25xCL~)
zR(s3%{_KCg-hkXf2=n<&;@?t@m4o%|CmAt=mvad+uEz100slyZE_<c6IzH%JU%vrD
z$)ld_B-LNjPQ<kjY}k1cthKbM-kaI4;FeJK;ncP^Eb;u*9k*j3O~t$f9O|nVlEyK$
zZOsDYw38d*CXuI!`#yqX8nbKR`4|)B5z=!UWroJOp)JhKTKeC&G<6Se2B5^F=8L9_
zTI-Ts#dufh;W2_V{qUH?Ej-!cR`?#g`X6ROYc0vHyhtuSsUP1_XIy*ixsv8nLLBlZ
z?aBQh`g5<}@$4W4WqVE3{@^4=qz5kWUhm<?{k@TXIbvrRS=_pc4lVjwKE4X&^tL}_
zC_hqf*8^dc12b39b+VQy{#}b7!WQNg1x+6LHQfzLQt2M(C(D3_T9;ln?KRv}@z_!{
zIp60yMECi-263xN?-ul-qhH(h&XFu!uiu6+{e2YaU=&+ufREIm@dJ23zb{pg?fe3N
za}A^c-g-Lugy*Z5a{YhyKt-EyWP=THi;<XuoLc%wCzI1={T%89q0<y@l|&VG`h8Fx
zFQiBH3IDBLbexc&_skBQ4%!UMuc}-}2plbHGAIC{&iC|a+<Ll6x&D!FbAn_#a&@P{
zJN2e~pG}wizM5WFf0+7CDH=Z6Pl~^paz4kr<MW{=?E&Ic$?t^;j2|S4JD^xvaUFPE
z9-ky;J^l6-oUc}ui?!MeX+2Dl|M_uB<yc=7MVrrcf8zsaEQ6zHaPQr!d=c@+G5HV{
zx^Mhdz9l?p%J`=|_<;#{H0?c^nE_2z2kcWW%vtu^N9>HII&41;737Wxeh6PCNxwUH
zuJhwSYm6lDDHeofduqMZt8wn?45!cQC7|_OR54zP{5;xcW&Jw{Xil<S=9mBdG*(jl
z=W{l7MrR<+g3DkC+JGORI;<c$#DG`c-={aliL~JO{E^%1(!9+?o{tvaKE{bq{#^@~
zF5!5;`%u<2)(@nsn#9kYLld<*U0@4l%j+PXa5r(2iY7S>h`?&UL)G*c0Z7byqtVFl
zN_D>1Hmwnu)m>YS9|TF52Kp;q27kp(mMHhYpMM|13(Py?a&M&G|4>Im4}K<?wdixO
zG37M61U8#m!=I=Q_bT)4lx{Hm`)>N`>u!{IDA4adT#LQr>OhP1`&bCH2seHVxi5|>
z4jgK0u2K5G-#L#2oo4Y<Ato3+meXbY-%3{u3tPk|l|nc4pL7(S^eQR;t(OYva;#<l
z#$LErR;GtTzD)VM<tsglo#0Hod-Cpo{zmo74a-tr_frYWla!Qy|EjUWW?~nYFzVk@
z_3};r29E!Pm;W7Kc6++jb!-KF;H%8QEUlOKgu>tcXG^V;#=SJR1ic$@J^i%U1rW*!
z<7A@#`&=oc{mGkYAdC7JYq<hS6VyDCLq#xc`MY!M{C(blMFTXdUe-@C7GnQ?-cDr9
z_Vw=pns-)yx7W9Z{U%%NYJ5g{>wmAp-xaMGuFtt)@a!#&yZL7e=>|DkgD_)^#j;QR
z7B&vTjsFoCChI+6diMX;2ML!iEBhUG&W9={Gf<sN<dG-;yOxWts3#}~xhqee%1~mn
z7jhlLl?nmwpYld)W7?+$%k}QM(V{VMy;KR}jS_I}1W&P_=MSd;_ab*tOYri7pIx~9
zZrHlS*;0KfdW4fVL0n8cLi}EhZ}#ZL<@5XVfb0?>psE=d--xfgoDuYAE9vGrDdghz
zXa1f<`{((<n{&9`e0uj1;Q!|xjnC`)Ba_<SZuV*b+3Y(^qRYW*|I~>SHfP5UIyzHe
zx_W+ove6PuK2%%my0`KpMM6_U;|?8Nod7D>48-3+3?`o@hqw4=XKXsOgiwm2+_*iz
zyv*aWZRqZPR#3pI{%c{u$k=#sX~`eoY#__`#r6x^&9n(Oe9UKZb>n7o+0gS>i~-J3
zCGlhXj~cH1{Tw?%J{r>=$II#GXD4&@;dJs|x{{J;WLM*l_g3#|SF&^Q9D?L`QsyO)
zXl^sj$oTcpvO28`vgfj;l;po}>2dRrj_28W*%ae)??MM8^-A{|z|fNTfw8Qmt%83;
z8TP*qQ|GDB`;!shpgRRQj{<Hkb;m1_XWvuT)I0!t%=~?w3fl!N!PA1C#QPqnhkW@D
z!otFYW|&5vx{A97i#-JCUB=9ll)bn?P$xP~ewLJxsRHNB_-dShfB<F&nHW)OT3Y0v
zOTJ-~)h@V~CD!X|0PD4@9PztYK7IPs@32{}Rbkr^buTI%2M0$@*xliHcUi}6s|m`j
zXEm-~w#$h)1O)qs4k!X>)erK`tW^8*KR7a4m|KxVV;>#f#W?-dFf?s5`DS-_mtRrw
zv|10!I?y8XJ8x3li@>{vhkkM=_^??+Ok6@jHAn3}E31NbSV+ibH{V9;=g;kZGvEBd
zezK<m?lKL!VPnhKUx8IGR?%H_A`Zk{%0g%t+<#On2|6Eb8`;^}>FDUhFXU>JY;SKj
zT%FgE#6S2o5!*r1KX}d9;`dCa%Z;Q<W>a9&lYoMP^7!IivZ#OiHYuND1mn{zojMPP
zm3|^t-LH20A-JTSqpXW}0VVxsV{&1GFR28(&0;6fUK$i_1KFzI%9pCbaY<#XqBuG2
zbL0p;6u3S54p)hWE&;8}L(<8Id4AGbvFCHyCiFKV{r%>jT<QOO`5%9+e?+^N|N0#-
zAc(?c@cW^m{JubW|8!bis{j9L3x7Y#-#>SmI#T>YH2;0U-?`}jr+)ya^Y>N$lrz6u
z3_|{wU;caB{1@#m-^IU|(Es`JpDgSD?U`I!ME?owx_JS)nq{0;ljHsU#em=d;3F?1
z6Taa5Jzqx<7dQN}eAlT<_??wpUN0Y+Ncr85+3{rL<?HO16;JlpC4xv8I<}ImrfLuG
zSdEu|ao(B*Jmug0Q$MaP4!4Hm`shr3JfG8Bg1fhgh!{Gm?3ZWf=O6v<C;5F?bEA^(
zYL+4)Bg4_Q8YwhVvt8_r_D06Y8kdb>m+Hd(Z=HFVX6NRXObEdny|7`SBkE*#xi3{J
z9Nf_p1>A4TfFt+|L&MZx_1)ch5NSM1JPi!yoY;Sq+swTiB7(qsc+%c_H8dtB27+}c
z7ZoKXWpD(ZXYfA=kd_<(UVXj23d3A3k6KTpN%+AOd2@}5g+<QfF(F};-wk5pSgfF1
z_q~rOc7zvwI>Xcy7kw(K-uYvaGHMiOsHLQ*qk_LH1i$t5_5a>=MZOj>mu0o}%oTw~
zz#AZXzv5RvA5hh~ahn}&XlA0yiI7QU4V-}g$CFcKF)>+!*oeodoB_bj<=RL~LkCe?
z8!OJ0nAljEIii0Ta$4!1gZf9<{QGc15H$RKZZog`KE!`k@vpN5obSJlXP0sI_bdJP
zW4~AC--G%4bNoG}|9AiRCyD$0?Z0_9oXG$6m;e73kUzd8Jzb2og548KIH@KmD5$DB
zRISi!&=Ld-L?3)wYC4#sPFBHY4=KaSK(<SAYWSVU8fBK+&*+!m;xhpp=#+Jk$Ttk9
z33l+3FV|pi?>3|vT^O_zl@6=^<Q^X-dY@gr>cekql%bT~ovD<*Eh6QAorn$gSG9to
zmDL6gwYb8=|CZs>iRRCGu+iM+vgG$XKc%)}94)n2U0ZXb!D`HSlCpUaNhtkRFz@r@
z*hSy4Uz0VOIyx7UlW2B!J79;u;2SpEaumU!1b>L`aZG=<YCe}sBix=OGZcBc@UfmB
zfBppf%3*cTA_nyn<Y7;<l-n?9b5--7!3l(O5X)|=SZLVx*@as6@nd2Rvj{?IjZDQf
z!1Lg>;d%VdS&EB_3<}$HS#MoO@u~lR`d_QkeV6Td$g{vzyzIr0+*qs!oDz6oc^V}v
z1$y5sC#$0|q*qo~$7<bQMq~h(44IE=luo66re>L?Gis&%@=XXy7}Bngi83)|3%DQK
zua7>hI^AC%n{wN^G5Lz%u7>$UB?n;yJ0oM3Qo7vX8YHNZCn%Fz@VjMG9%GQa?yU~J
zhyvf_LaFdOuu_SvdJS-}?1ZHV?`rIBwUDPgIyCN%ejuTp5E6131l?-M0V9N8v=x>L
zCVMzqVP_IiEEmsvXbmW~;nCU2fhRzS@bEY+M%ja}=NzjbKjOBTZ7g{)^nqFZQ+_@<
z5mA7zsi~>BxHtv@^VjXknwMDQ0{ny#csfNPAt4Y9fD*#HtD#J)QSGv`h)XB;E;u-m
z*TL-Uc+alBpBOT~6p7#jHY0Qah4t<G2F-5UZFHMd+}!1mkxEHP@l<}4PvFmdlH!+^
zMk{ccVAfompH!?_?SQ#`X<1o5JmBEq;L6Rlk)ozR!l`0YT#q{88>iE4$h4P9@Fq10
z?P$3Tbsql;mhGufQxj_4>iEP2lX6A~#&<!xMcL=ie*$(ZmniUFuo?314qGj^&6CsZ
zg^qIZnVlMlXcou$SJ0>FK*Eviv6-CV=~N!KSa?lG;A~e(Q&aQEMFmpJ=y%Iob1iAo
zk?kBxRzjk=KCtxPrU=`a<hN4#o0@1P2tiZ;yDm97wUVB>_uKe^YJjrjDuwmwN_PJI
z{r(K)Ohp!^jqk~#P<aHSe*L77ECPYBu-dpQCR-)9-G~RRP=x=hm6cm~nfVA+2P!5>
zD-r#q+q8tFvvYY-(E|_d^v|F5uN7W1lwU!I{1aZLhhw3MqX<p_F{k<0IbRIIQu7f-
z_e>z^JUl!qdh6|$x)*v9b^4uSTm<>kgUx=<%)FX<u+rx=<*xd&im#0K4OLY|MVo=)
z?nd>F+^>w%5a->WINMVdc1yE1YHDh5pa!B)uHOpb3)3+yfs3-<>Xw$ve9G&+LlU?u
zDVex=c=jORu-1%nJ8w>BF@G`XA#*^6$Wv-DcAJ=3Q%$Ya;igUn5QiDC916l`x!ar)
z_V)IJVJqY2E%N#GKB(FFCWmB>EB#XLCGv>^gM))so>H(eRg4Hni?v&XPx~33e0qJ{
zUtgd6^57TGao|IT73ICsFz1=|_4rj%^B<qYgoM2KJvOHi^AeYO5_>p&<5~5t*oP1~
z!+Ge>ctX9G_9R98Juw#xVFb>K!`ko{xXS6{-^F<Buc^BQ9p0c|Iy^baE}OfR&lvL4
z{R6E~Lbn<0c8(W=aa>jjZVU5+uV3$bjE<4K7|QF?=)G5Ws#t~D?#mRgg4{lq^^|Gv
zshgl(W63Vu;)zCts(D&ROsFZJKB42$Ck=XBR+^&zm|Z_~%uhY~IFZbHQv=2adL}R(
ze~ph!OlY6+q;^r;y22TvAX3iG6%`fDewzKc_PB{qr}N`|YA*M}XV0E3E-oI1%SuXq
zU3$Y*!0cpWv-~3kb>TQ)zcqvs?DdOi77y*heTVY2nbZoBM)MdXTHsCg7Mm)812}t;
zdJg0ar{|4?g^Fp?dU|>lbAA0btkl#g)-w%I4~D11b}1<??)-t=*h99*d$c(N@8TPT
zlqFZxIyi_MwI_098zV(6!^7N!#cN@StB0f>5WpO+)n<#m#j8J1Ig7`=b?alSz(^qB
zy~U2G4G3p&pkdt_;j|eY!6lOWSoBst-zMo;sYPsY1F!Wo!lRI?H#Rmlu#eTev7V2Y
z+W?!<iR>E>QR4i2ccP#t^@CsI6+7%jU3z;f1McqbTem!oAkHXn$jy%nB`bV|Tx?}|
zc}CdPwUU{f+j<&i3~a5y2@Rc_>tU53+<~oS)E;pS6VoF`T~#grS&i2IQUDW;KrN@m
zD24FXt4Dkfoi}yiGGNgv_qt~WKVaK}j7zTLi;gb`Z{9POzML8IxU^G^vCDw7lyPTq
z91RPLaP<knX+nLE?cUN2kC@p9?}LJxt%y(G(8^meZ$`eDtmg5s$G@vlEFH<1kdUw<
zBP=ZZ`Sa()|BJb|jH;`NwgrO)3l1S6xVyUscXtWy4#8c6Lm;@jySoH;cbDMq&fBr?
zyFL2ece{JM(c_+9=ZpjU?5bV0)?9PWsybdDm2s`3+7YTU94<8F8~$Rg;~fD6RF(V^
z=HZ6O`1pLPf`yD%o?ar<rwJR75o*2BS$Y6CH8eE5J=~uv`1U}J1*Gz6kMR8z7)VG+
z)c0w6b(YIhIBaa}?2gC9R0V1X@6gbuf$@&|R_u$P-<D6*28Md0or(VY`mK#k`2z<)
zwdA?mLUGF_5|33_<fU8De+06{gbl9QQj;T1L^Cu{&_{DT8{UNtHxLbQq>=sali~j^
zQT~7U?pgH_=MI?|6UM+&|FfN~fsVz`cUHQlFbr(WY`C<zKi_h4(nuIs8rc~GE8bb?
zaDTmMgiOrs3~XtH%ysMx1Pt`7^bNSVVg6jgX`w}(>+?lD((x083bDV<9D>`h6C=UA
zl3}U%#_@KcIJM*o{_NI?5eO(Qv>d*mJej`OmMsV;>>+oPm;Ftdl%uC}gImj^9no>}
z`D;uEtu{wogz85KDwdKxOs`7t<=KeiIF^#}fgOCDnM4t8saDs9#7PQobZFK=NU5@L
zyg`%!$Spv!|Fku+nClt5lkNRoeWv&a3K5oZFuNeE^^`k6{Oq8oG?giqpP4mRw{f>=
zK3`10_poNEG#%IB(3}Bg*mKedRn?;=p6B_e=kr-53yUH42FNMziN*&47>VOYs?9y#
z)2SH@h7Oc!uc4G0URk&3v*?V<&=DvYAzJ#gCT&yj_pwM9uB|7;1Z8Sr@-s3;ids>2
zF?{b^<7Y9-L|(__WCu-Ai@Q=XZxt=oQ@8a?(KVbCyZ5ACaRu|w4s`KA#P(x(|Hv~N
z3yClPD3Uw9646k>C~W^j+SI_7q{bG^nV;TtwHEZz8kS+|Lm18=b7i`ZM57RUCe7>C
zAXe4x5#=rm-j>!gnFuygZOfXKeH&hL^5Oe42Ut0I$@^v;S2MJ^_uThABFV+(kPM7G
z*f>EX5Y*-tEP=C^&)O9D+Lxk>p@D0tK5XrQfkJG_(%3DFfg;6U4TZp~v<R}I0;|S3
zjX!3vEwIDK@GhBR8^wG+hPrTq<^hK+rMu*PzY=V4Dt2M66W<&6&ZH<bfj5%}eUXTE
ze0Fr9CZaAR+wc9*8<KUzn^7BN3jqf7F~1aYs*EwehAeX*emtZ~d^kUHMZPn3{sa?H
zr5@~IA!4zsAH>s6H5l|Mfy3=a&A#Koc!TrL?)QONm{bu@*;y{Zq9o|7y;R)ZLhZ9u
zc0rXw<PXocptYL)d9qN}omToZ3>M4!SvnDfhRKsRk-VENyW64fCT~~r`|G(Ep(o6J
ztfa^#zvtO-7Db)u@#FaF=<OgM_s`uV^eJq9A<MZ~qyAx;s8kwWbW{bMPbqItfJBRN
zCA(vG?3rm<=VFrufx`xZ5Ki9ad?YEWPM-Una`Q8V3vHaUmAW*s8GCthua}07-6x6Y
zzNluwDuMoxGK5@0F~RONY<JVq!5jarkQUm-{PQ?1`x^x3>#9DXgJ%}e=ltQ_<JIL=
zJ^b2CW%h5E7#(OQWHTv<Bsi4|y#5(qb!R)tnXe^`!}%M_lnjMwOIDhK^tY%92Xc_S
zWbH=2CHJ6EJjm(|eD?~{5E~Wm!u&#?fP$92oTbVu`n{~mEq-NT)%R-V83y)dfwR9*
z+rWPuuBh=4RKo#QovN;>mqNr?CiQ4s1f1?euq^@ImtG|5;jKgD$wLPNCFO|?3SMUe
z@KlLgz80nR1+L-;Yn!hmH(YBT^{nMvc{QK{{;%R?4}Em*-$dqqG>#=v>9C9sWG#Gb
zxrt<+IXrEITYe%fPCxL0*-i5+)qjCPsZ=J(8eY0hR*^{fil>J|H>3C55r-c@1q<an
zEH+agynQTNmQI$;$Z1vx<1Z)=t7GFq-GN9J!~8*KZgCk6RRTvZ0$x)1pybKGdwK0M
z-?uSGQR7ZdMjcI%{x<#9?*!8H#la8Bb-Ut2*QHIf&OGVJ%QbryAL7uO3nFGlRf#8@
z&tujE(~@MIacn_aOP#GMY3*0v-NPkS59?m!h6WuswT*VS=sBDqetHs$ckHYHWw=&Y
zW24Pd4&|@h|1#u+rFpIW;(2=fMF&E2x1jU$8bscl5}){iy%(r6sZihRND`mIG`U6<
zCS)jHaIf=iy=Q@|isZ;+EB%3o3>9yco5yuF#NNh&?U?t?&`!-DZsMR6$JCotYPgqa
zpXyI~cDRpGT?%dNUh$||9hJqYe0cl}%Id7tv*^Ecp3HBZjFVRPxlmTDV!ZUBx60?+
zK|`PX3<{gv+hr$p;{nNoL=1GuC%@J?*{A!bOY&zE43c|!WashZ3ti|If+WiHFt!7a
zk4j#(Qjy*#sI%Bg6(z~<roX$_)a?gZEYlA|Z*r7LSkoyDM|6k`DN!%c2u&GOIOKkL
zm6PV%xs3uGft5PZAFq?qT}5w6CQdp3*pQt2-YE8Nq?n#7mPy5vO0lu(LAovL(lv%V
zVsUNY@!<1kKDQ74?~>o*#km|<8osWyUiosE9yi3nq$k;O+|@OaZ!sMVJ-tF)SxKZ^
zYwT=}A~k&%B>H@Lw&SX#5B>wIS_zXr{o$y;@Re%*x7VIctGM4b=8Fr^d>}_ZW{~`7
z9~FC0f~dTS_-Mhpf?5~O|1O-Rp%+0d5S>`nuY|aznKwFOkZ}SQ)9@rRLFgjPqM=fD
zS_Hv0Zgn}5S0^`Iy{OwA_RYh!o?jd3srBQRP;0mjuT~TWpld3ODx)sXg7l@!zLsR-
zMS|N1;$`zw9i?V}FgTB-bh~!eODtyXBjjF_tI+$*W&HxB{FOlAUX=|$N>scRS2*Fr
zIB5aJtayaE09rW@W?_;ttCl1iV<Q*BJkJo?Eumri*_{?=;F<ti$O#5$Co`um{n5%>
zfl;;;_`8^}>7XZm4R82nWSSQ$tMmNeFnH7@wIg>5uom?09k^ckF!NhcJdMz@CF_l(
zqhrJsV>RgYl1un~#0U*%W2SNYhEIk>Z>{_ywBhdpO|$Oae*QWXX0XgSK^DU4$L7Tx
zoWd2w{%UY2fF$`xAhXT8)~Y$2pDiY=p63c+d)0je&0`3k9<a%IC?2FUw=@!pFA?IL
z6Zn}hYmVy_jT0$mwFaUV4<EEAERZ%RkU4<uTTyJBtU_}rd9S4aqZ<m-lTe>Sgargc
z0?qU7QA=oohK?9ME5Ad3c|6h!`I~imR1cKoxHtB99`jK1#~CFbiIv+Npp3gSI16>N
z5YTF_TNeWMl__3&h+uWK=iW)dujc0QMddYYSNYG{Kp41D<B8hP)D*zfRBk36h@5P~
zjQYIAa8teX5U<Hsv%R`8)2*2u<~EImlJvaFUg>U75^~QeDw9ef{X`?5VxNtM$HJbe
zT4cufo%AG$<%LeBw9@h!a?6Iw@;?JhEG+*wutdi|_x}tmm8kSv60jh)98latN>shI
z^uleAD&e(FYOll1!(|{LIDLG<>#k|)SEB6y^0B6i*vz>&#^s)EZ*TGVw$YUnMOw-k
z3_fob2c-{|YBk+pBC#KYRGhY@RT9{<T!><m)VGE>YJICNE{}~9(Mc6Vntttc?cIP2
z-+SCP=`+tyaSln3Z_|(=U>^8X2LmYph{Wg*)5nR4;MwxrIw|N@;uFTRne}pW-srk|
zHJ~y9!zAU|cC-OYEVE^^k<K}Gd62f(d+aK04;=&PEv9ZZ1VcXNgCjljTJTKqOE4*H
zkxFb0K#1g7Zbw*ECF7z}UPa54O<5MnR!0HL4nBYkLa7L~Gud*oV=KF*BFFHWZlVPh
z+Ow?ETqVkpTuGu98JY!KgFx~@7(Yzktc|LLE?-7pUNCD_d{_hFtzBfTl9wngi9*pC
zA?b3|*0k4@Z_y`HEW0vw>3<AkEk{b*ta_71wT*q5tPCPdy<?H!rqQiVZ&R&36iZq8
zwRYIWDiRV2b#d%-y|!Q=FT~En^!T^k9}X*9Xs#@$+ZHRKVjDlS2{LQBaj=VfHfAaD
zrrLL8&V>5w-^P3#fOy7N5Xr)q1hcB@S7%h{si#s<SsnQx@n#v!Dqm0i%e$c0-X9Lx
z&*{#@(XWf%>%2I@yHJ=QstPj3e+baMHxr@G!KNq7Xj2yRdT4fgYddhbCH~s!vvwP4
zo1THqN08m)DPCm1M<?A=Hy%N07-KsF@-HZ9&zAeT!!V(DA6J~22A}b=_cn)*+ZQOz
zcx=1PDD!2(6+n7kVW14|dAR8!_JX)SokIr;sUol>8!eLQEPO$Fl7|)HpFpQE5+yGo
z>=$Krc9Lkme(q|^b;EPx4W*Zzy)vEbkJR0_(c>n<c{hNzej_)Dy!e1}p;jo38cOBt
zm~ZZigiJwrsGL{X8T~F#950M$eMBUg0o2^KqHAaI3PDpe`$gvEz`ZPI;5b=_Tbz63
z2N;h9NmJst83nc4#r=?d`TLUAMNlK6uG=|_mXm1TuPb#p1<->@tB%a3#A)}@aS?4T
zqJ~b-EVcRu%Qw7R0Rx9R9aLAQJ<^&9YlJ-()io6kb%Ag$)L%}{Gjdm;-xA|^!=9xu
z<{T;qHNFc;lgv6I?U7%4%u@Q0jU|};5a(k29Ys<F9dhj&=Pax#Gy4oD0H0T^GYec3
zbK)ZnZ)QaIosoQ7He@ii7Z{S3m~va7BYwi=@&-D6Gb-JDW59fa?Yjg~o!6X!$(Hc6
zt4v+S*jDe5F}q>QvC9V?QLX|vFsFOw^I|F4U11<wDY|QGNKo^<dDETH@Ks&Y$|RAS
z^6>fMEvUYrBU!XTQPd`z_?D^hqB49cnlnT%O_u*sKQuDZmi(ToQRTUyr1G<Hyim<~
z!U%n`-Lhqk0Tc{YG2a2`ez8^KvGd5%L`US%LbmL5v+-LzH8IU`(O9h<>DwSHZJ0@q
zBbw)^Hie->F@{w`<J<&V=Fe&7cHip2XDCm^V2w!*(6<%GcOaNZ->tUS^$!lVP{8&#
zyTVqycuq`PP6{sX;R*^PbrNbndg!E=Tf90aEFTmvup?|;d7;Y4#T=o=TxQBBmDyeE
zOLjpStP*|RbQ>@KI#iZ)ln87hW$GMj>iH_*t-o*uevdAeL#I{<*Ay2ScV2#!Z)^*-
zwrHL{*!B0R#J!m&MRe4MF|86GMeoDBGNl&Tk;JlI3U7!9xhMF&<ADVuS+FNaqS*4v
z|1{Xr{~K%-oUILLRHSuH4fO12<n49sets6U&@nQgQ8v*BYF1`iS{e}p6C-20UvK$L
z>^{pF*z#LhSX)^dSlZ#T((qfETiME6>*yKK2pBk+=o!cf^Zl)OnV1=XI{4@7zo~c+
zmmySQj6Oh~&Qtk#f2wM-?@6Y+89ok|+_7U{-kut^c%34LudU;#8xl<Vn%rR{`uG^2
zPafJhkC*2`kiZ0<c+k;j*SV~g13K?)coGu&fkPYrkG;-wySusw0qVD%IxaT$I<mQ3
zEQg+p>xlqbDMN-<PUt@M?PH3ezP{U=Nc^FQ9sa!-F@V=vUS5Vhw1tL-#_3n5W@2I@
zBKkq#jNaYb>q!{d)7?!sflv(i{oM)Kap8E&6(WAiMDGCc=%<qfkcNcBAXR^<MpNFA
z6TEMKYfis6*ZQp(|0^5)G}`{v*8f$Pe(JLSs2Tt52ZDbE;rD-Cjev|K@>dR-;012z
z&j)teU$(N}f+uiue|`r(^p`d6x61j`qyOuh0<eECjRmyt?~S(qbLn4g{r~IIRp=}%
zb!Dyhg@6(PtQ%0Rx5kT(Y)Ig8bpXsLz~P4L`}@D!`Li`1U*Q3h$VFEW3UHn;K=KO#
z^p(fsnQ#Rb!dg~D#3(=-GBGmZ<KzGHpgIsZgdiaUQn_3=cx`QM0b=fB_S{OFrzZsk
z#qUR`oo#o@|3u{C<b=uN(G1v5!o$PoHum;Z0lq#*j?Cc}PzMZ0l7WL2f&Tq{6B%62
z=h*<02B;`6hgGA^EiDy4CUV4M$Q7d;X&D(Yna$@y;LGH4V4`}(UZSzNT)KW30Rt~6
zEG+B+N=!gNz{$yJsaMv=*Ow<s#Mj5?-5#V8A3uLqRu<0}5rFnZH0RZn&5%K*4Ft5G
zlcfd*Y{09M4hS8#3pM)vQG1hlZM&qHwtx*L-!ph=^2;9)7BT^W9>=XMI3Pecf!?6&
z!#j9D98xgHVYAInO~pPC{HFm<HVq98cv-!^0A&o&>X#?0TIM(mhC{`wRk$O72>bI_
zfBX>qx5Z92B`Ot}fHIDOfdOc0Sln*)ZEbCH*ocT@S;7dClEVWr6oA4)9mnvm(u%Xs
z!A!IJA3lHnGm-wC1^>OD;_vAE_kVs5;y<JF|N4RcpIKn{_1{_G`|m8k(*K`23x2id
zS77-cK6eYSe?5|)5vLD0boI}asQmMofyhGnTe|-@?EObk{l9qN|F0DGf5RU9fBV4g
zfyw<pGwfYFJG20r$z-1Nw411uRIAPAm$EXNfBN<|gT3DE&K3?X3<zmJ@@fK9xQvVp
zT@HO<j&2*Jxhnx4GeE2RH=^TY`<9lL0tPYsb`Y@FLrE;s<}c@iG*_3Gr?{k~fawmF
z@<05sL#XHT-A+86-YG6Vex^tyKBA0y8mIF)j_2cV#r;>Ho3mN~oF2z>6$A`La@Bz7
ze=aW`M-Aam!(I*?g$)Rf>`&4SHk<#Nqi#LW{Q2nYcE<qyH3)}BtF<^I1CIyTTZ&e<
zD-az^9r%}UERO5DJiyLE#?2#cY)tyxy20av45-6hfC-xv@S)&{W_NUSgkFMC06G)U
z9vuGc@$qq}THG=yXV}-gk0cBZhjV~1LjZ!L4uo|jU}fRm>$j9Dr=18q<(knyW1*pF
zX!PA$0D7EUK{C*&I4zHa<mBQ+Y*RBcKoru99T^}0Y4ZiVqCoz8Jl(jzn$@CEDmFGS
zs39)GWVaJBpPigEo6H>!LLqs+KPXFNG9xA?{tw-!sUdRdAx{Q)3EnQAnSH7;onGtk
zH5yL50`xnLk>6wg=PCz}A^+dM^Pd8O|6*qR9kKsujQe#H|KXee91#2$GvogZAecay
z;mx6^V&(i1ZWMKJ)g>Rdk)oknhZ(+n=vQ{zl6R`Np?VXo9uYl$a_IjD#q76>D26=Q
zKip0^F+n;mEg~VJpCX@apvYTthb}sE^38H@Ew+KSrXEcic2y<1H><;6oLpF}<)`@)
zQ^F*)G9DUjmP6#;!X)JQ;`sJZwpVGwwVqqj_j$Gtovu=$NY$?^{OLBVcY7R8les;0
zHCBt06-Fu&@rP>*AI!w}V>l0;RX!oMX%pHZQ7CoNr#WOb?_b}Xun=5AE|U&Y&84~>
zX{voR{(?^x5hoQ9*n>Advj{5uyrp<Dm~I36ZN&9gko)Nv%ZAVZ4&PxfIhwNgRJWUL
z{8<0IeUIEo^c6wuy_SUAMuB)_&w+Qe<75YSi7>eIC(tY8Q4Q~)1gjH%N&>R?&N*5U
z{a#Fa6IA~M`)V(;!?m&Xs;jhl&8m`X;dvB?`*ly1kcGka@iqaB5^2`;SZw6TlbguM
zNbAdH>7w;W;t%wbG5<8$qyyDJw{rEl)9}U2wNjO2)WzMmhbFjoB^GFv&;rLUFN`%<
zz$n7=_UO>#-PeCXU)qZ^w-ZK-X}}bxvS<7BbCw}O8H={fx6#>~P-*wxldVt%!S=NV
zIi)|?6-ZL8M^iNGfo8j${Ao&^?oFWxFU$1fca@t40YD$rF3%G$z%*IyV#nQwgGnQV
zT$-?u&p*Nc&g1~Jym$GbwpJq`m^mbN7~9E<g3eR&l5^A7K=%Qj)D5&-K2vWf_d}9^
z=lFMiJWE!m39EnFxMzmy_T5f()}UGSgf=7$L~Hcg^wPfXhJv%s{BA0JJI(0j@kTNU
zW3|1oV+GPxu@^yzP_Vkjs_E#s))AA>HrVkZ1Low_Kb?LV6TDHNH8miq$wjPbG3P!!
zmKvEiYD7ALEzDWkP`Q2UzB$S8Jo!MxM#aXk=ti%*$w8BG5vI$6^=C^yU{qgRLZ6^x
zRvAjSx8pP?!NaR-Cmwo`P;zJ2d^$)C=dyaeI)udAfI4y~sE6qg@M8HJMgm^V$n6oW
z>MQ1#*2xxicwdZuWlJOS7c4jQM8t);TpOZIF_ar=1;b`*8O?P&+P$Tk_DG)b-@qOi
zfAb;q#M}-Ck?nY&BcJ?fG|418Gl#qzgx4bq6O@V!HA`-CS8VGNZhAUGSik~?vwkO7
z;C5fA_67)N4>CSqHe(>XNvvmG=_evkJj-<d!Q8Rtr!C8SpIPDUKiSbW8Ry$W_3tQQ
zV4#mGTY*#8KVH^Hl*_I^5{{{WZotWyp(Cc{rseR5$;qVUc;|VW%vkZ}&yx;_V4bL5
zJ2fsUY5hcUdlipq(LzOq%V=e8SpC_U<!Vh~oVb4PIv4wc3ulPO%s}Zh{*It-+7rS}
zoysyk8CAg;`#XbHIyq0cf38P2QG7r8bx*tnxb>8*fZj6p)m*LOv?V%Xb~d%pa7h2m
z7(}SpDT;>sh5hfz2(P7|xz+&&@K-D30wTyK6r|IjKDnM4@;^Qd2QD9GOE_`hEYSN`
zkX4vMy;XyBSA&zjK=_4H0bAfHoj{?4MEJ5pmuI^~8Z!Yt+zfg1<LW&TeiZ%!TeX?t
z;u|P-C)Rv!MJTMU7clN4p_Z%3zu}qEJTjNt?d@}tzw)Mc$2U)DPItDK*p<N{Q_^B>
zg^Wmhf|@;k*bHCU+;Ak%`9FaL>rsYaSVXalqX9hpem-yd6u4Ov{~I^zO!leAs0IyQ
zwRo({13I=JJ^X+1Fpg$+K`QsU=YkW<dsIeCo$oevZj!I25m7CC3N8GB2RQ{P?Y<B(
z5|QKdH)*t1iA=^o3u^$;;Kq`TQqEHngwV^@#^o@3hM05##?^c`a+7oMq}lJu9Mhnd
zj%P&TPnlW1MxfZ!)muF83*Otd8e0E}gdWwzlw%3D1sxS9yOxfY`qX3Ysm>P6Q~0^%
z?%J>>EPZTXxe#&NcW+%*GYm3UxaV&}-Bt-C8=hFry$h-zS1)t=XSD(s{M=Cz=b7b!
z4Y`;<P1{X8i$b87WZ2S9Xj|M`y}h%u{o7;+L=&Mxe9O@Sud>#gBU)mYX?O#LL-BeE
z>G##{GT#&>+i^VJh(eiEG<JKHfOYB9&`<?%AsD6cFHZL|v2vFwmihPc0<jW^&UbX?
zet${`Z4iM(?6>btDm$9G$I|Iwk)<?;Y~TWy$eOWiA7T(CB&97ZX758Jwesb5ABP%3
zJrbC(ol~NVPg&mqounTt^_-IS_fp(~iT7(JKzJ+@(h8trROUQ$4jTk%9mV3Ar5^`1
zEvv?NQ_+pz$6NsAeS*C&v)=o@uI*x+?f8xNZoxP3&sEmc!x;$(`mbEZy1M&7BB5JF
zPr}>l+liSaO-qWW?Gz^umkrNFGQQ<a7U8QnVZM3jCe-cTcWVLvJ>Dle>d;Jt#7vh;
zQ9QVaA8ziL9jZ_~+`gIhTlkEVjRS+jnI;1|<ErE!j)k=BpzvX!z!=XOQM4N978(oj
zzvRR12!SLle8kx2Q{eQI+a#WEWlJ6xum$#^vIorHdN+WcsP@>y*k+nqfLTR=WS|G~
zC2Ey}m4{Zq-)1BUB;NN?*rnouzj^6YTSvLArDoKf4J@nm$XOtULTp<)|8e@Mv#JcY
zY)Iw{$rmGG8HIqt8njj#_?{!q<JFfY(0H|fE_NUYL|0)`aC2%V(AKL)j{4XZy*H%E
z>x}%8jJT}JCN2?hT($x-kGB}8t?H1I1bWTtZzLNwRbS2ivx%x{BzX)p3C(LauEn~G
z+ju5d`Y71TC(Ia3B4SXqLH@PRde`Ig(?f;Juq`(yTB<IlU`4NmEQL0D+v`6@;6{x3
zjVU>lC~0cmJq{;eoNOG)8DqUkXyD`<&Co9WT`>cvDMb(wkg@4s!LLefg1#OW1(Nqo
zq}QE*j=7k*1Gr$2L~u0D2Q9`@eT_XJ74B{sS737?9{!js+Dr`Jd%x#D)*-Mnc#D(O
z{@n4GV7i@%!hTN>=5NK^<#}$^An`>>n+TsQv&4Oo{jEI^^j#6J*K!_c*s26NMLM}Y
z-b;up5MIwu<%PK?9ZClKcXV6^8mG+=L{YC&ms9+WQbR}l^rV!!d*o$3V4WWHN}Lxo
zgpv3qNp{-EmCf*G>{}K6(lWdIn~CV)s|mjE-*<fT2$|B7PjA!xI~m!~{H<hektqrY
zc9Dy!y~e(y6f6-t^`AtUs;`R=Hg8^#V<?*IPm<D7TW_x?NN1%MrZ}AKb;mu?wDCjg
zF=L(r@QXn4ugUlHQ>!=*H!Kj_F%dy?1BBidicGWj<l86pA1%jE6RYSMz`^Mi`1}kG
zHB_cASa)@9YNTX|NUJfv1*sGH&fTl65Gm)AfB3J*JQ04p9-4C#E_|0ODFZiMyimKy
z3w#R(i)m)waV1vm122tmISUk2wCuhIeV8f&l~Z`b?Cg8bGmE9`EQpuitMP!a8J%t2
zfI;h2YL#a*;bu%z8s}OUiS1;UezM(K$7fk!*}UeOv3PHpmzFXD>!Um9Nvfghy6;^8
z<dkM#9iI0ALmJOgnu@=>>@go}s87VM74vlZ#)B)b>*TY?-XYNm^TgWi+IH=*;!Z-t
zu49Fdj(>L^`i||?`RBnZ$52&kQ{IqiD!B<cAVPxqNJXSfg_Daw>eXR)9F$2ovAp|@
z4AzAAwS|Pf_lG=dja(iuF<H(AagMBBjSwj2(9%F<l12vu>z5A~`8W6l>~~*kpQLL_
z%Du288g9Cp=Dz~D(edg}lh=fIv}gQipPvkGubznkh;X6!d<ny!4PpZlW{zRMI9AfE
zbC*LcKAp<7X;1usXVfwhjC83Jr1SEKerlDZZOvD%R+j+!S>`?3!J-XtKH}$-)2rWL
zq9dZO!_HtSRo-Kn1j3Dr<7_DAhGpU@;@a$;8RMsEMxrMkPss=Z?6`3ny<NlnMw_de
zX(BuEwFkD7&xt*WvtVVg;1n2&Q(R@PyJvV#_l)}m82%9J>2mKG4S79iX#O&+GGd++
z6Ppg07e}1j*1Hj+J6tOVB7*k2z=5yR3owGf)moEz*;+1KHhODVW@0BOn2y#MZ=LSv
z_f|(AOLeln-CO4_1r;X?-I;I5O0-<2rxlSQ2deuaiPLT$1iabaRk+!&^`)H_{ecVY
z@ok{ToX3kMnB2CjZU*g+ANIo1KP(<wep>_=RM$BPYe`9C+Y!n5c)FJ5J7?_BFOiFA
z^vb7&`8o<|Nb`KLW6na@tw~BosO`b=4_5XlhbNCYlcH{2CVu_q5pVB%7#}j(MBU-%
z#Pv}gT<W!)-tZNM#z3jad10eL^->5;iVXG4o&gL~wCvN;Y)w&4B1BX{nV%Tz{PmBJ
zgSY50IW&z*t?YrGB0yMF<h-JIV}<f(RKCKJQ0_V5f&c*v7?Nyss*(}B@15sSo3DY9
zgi;i=o{3bQBW<kQlCylheArl8#F-GPRlJ%vlXgA7WYkdrVM-zVs(!<zn>RyEeZ6d?
zO^Lo>s;3n$m2bp}>(r%#BmeXq(nr*i1*SjgcR>uvLkbxQcFc41FrV$!@p)AV{+0&L
zL^J{vr$CFA|Bm(120pEHpc+#;DqrNVslke2(UA9y7qkQDe`VTYYm%+-eM{MXS*2s8
zd4#wfMgAU0@V1rf8a3khJLx1`dfuAa?<A4ZrlrL}S=_2;{Smz$s22XpV5M)fedQp>
zM~S=Hu_-xEH2<{zBc`7h<S2kqDW`oJHAf>TUPU*)wQtF4QAB!Nt$hK(s!<yb@j&m8
z6yXy&Mt$7_F~Y_S`7UORO2=_o3aumDj2-_fL?pQHSX*CiYpz9)DU7KN7(J#g`yq&^
zJOjW!1!@##LChWNO<#%}lKLaRKU&ig5TU)K2~d7g%wK82DfAV~IR`imVAqn>avk3V
zd0?@>M!j`LLV~KD9plf8e##I03&DC|kycs3gfUcH(VZifzZ{+Z;@8Gl`!qo=6hgQK
z?tEZcS{^hpeVRti4_ankUD1%M+v3~tZ%Bp#wzWQk1wtPc3YET(a2jSzQ?USvu%qVU
zd-!MTOGuFbE=8ndPhIX*Q{(0G3`u@{f71{IZWvF$i~L$>L&F+0zLr8mMQ<z7mZ(fg
znPcXQi~cl<_sZu%XZ?<lF#})U4`%mIu8iMz<0Sd!Nu?A5c4>f{ABd$YYIFydegF(f
zWo<1si=_rY0U>h#YEg>=VnLec%ZXPyBR=<wxW2?z;3tBQ>%dOgz=nKZa79{Uf!!h7
za@x`nq27*@P4u)%9*4Q#x~2iyzSxIw0S3Z97hMv<M+CYPxl?Vi=6G*TzL=I2u?r8#
z4W2DW2V2!!nCMwIA;TFL{Rq7tw{Dv|xM<3eEEtXN9aJVk-v7wx4S>PL2~{!tz^ZQv
zQdV#ox#J6%Y}YgPahZQEua%OL+W5I!3b4#o_$LN-lg5&0b)d1nb!$8RmaRGtL*D%{
zyALMd1S6cYG}r_!TW@4EIZ_20#;mwlR$nmE0scrccr97)b77)V3x>5~O^Nm3G8I}V
zIqqzTWDuI`0BTQ8R{?%}R4OPJ$x+&6$z)w2Dsc{xBp2&6#e(=z@XGiK5c61l_jf{<
z?K!doEL$zvUVP7#utJyZpxO$)(~50u+Q^_XK?2t15L7b!UJja~Ul4nh1M(XeHhB1R
zcim)t{W1w`eH2{e_xf>HA+vWTQ<eEdKYLHh;t2*SmX4G_`RjIg`tCb|IyCNJCqKsW
z0t@-M>;)yL6SWdc#nnP?+tCb;0Qh-2fokaQ1Hky|SPFU|;S=&s7ZLX_H2tA;S~5T!
zV}zAJ#BD0Tu;DEXsTptF#swK!_)z^mavNQPb+rX2#@D(8&=1ZdTwGiLC%wyf1t6Z_
zu-om1dJd#{J|`rn5#i&5wgqlQq<8i7^Z@#vCngam=OaN7BPC_AHZ`i_gX~&0PxE3Z
zRNZna>Idc1$S+^Mczixh%48QbVjsqAbT`HdVMp-A(S4W#ljl)AgSI|ws<1*yBGz2j
zq+#!zU=Uv)XIeEokL|1PJ3eTNo2FS@X?3@jIQ4#j6g!BktvAL??I!80gKX<g;9#|b
z3WW)k41BsV)a7d`KAPw$9p3!fD|;sE&LSFj9e(ODo>O+OIZZ!ve>~<`w5$x)vFzx$
zgKVOpfcnhid7(_%a~GvN*<piy1AEkj{!)fN=hLjqOKqI9y#_2JQ1H}fb-VL}M$xo0
z2L_nu!x77a=Kx@IpS-C02>B=-^^Ao2C`&?2oa>1Ru>lB_z|!QM8=@#2SBO!S?PXh)
z=%>=^uIxjOA#VDI_#L{Mu=|}WtWu8V64!Q|y}E}Smp;il<`@>rm)ZHYFDICqvz5o+
z<zc<GAv$6DVB#7m59%6lqUhU<+c2e(5mBs3Zo~Iu1O1@h%}(W-IPtN1NJvNmah{@*
z?|@>aJ~JNfVeXqAFcgO^l6ncBb|lea-ub7fFV;N+bE-byGt~1KPXo4IX(Sy}s`ci1
zW$DHbzM!DsJc2AZI(rIJ^SXYkUPCHs*Q2*-9?^|v=wu~28Wmjqt1@NZ0uK|UFQ9_~
zrt*3kkH_Pt6OX4W<(ZI5&FHI4|7J}_Dq6}zBR#idA$1EyV&XPqmJo`!eKzo}*s*o-
zn)Fm=GR$!hrdl=cG-&e;@0&w%lgg5+mpU7AN)FXPc&SiI7!4Bhai2C1ah~cf15^vM
zu#r)rY1ofqIn*0VPKfJ(*jDl_grm$z)oP@I?c*3%yZe4ef#Pr>{_?k`;XWclJOS4n
z2efbGsgbEPa2>~upP*Zi>puheC2l5!nzcrIcO=8KlM!(bJ*m}c9p}L}Z)G@O+YCW-
zCZx5{mlRlFpxNY5Vx0%@^#DEplgjGgJ4mM|XXcw-Y3j#yQ|@Vg>{iH;5gMbJIa%ai
zEez*I3^SXAmvSAyZAFjS-P--o`Vv0Ov*QjQM^Tv>_)*53^ZJI){C%@t`I(@eb9fwp
z=>VwSY)m{lRE+fARQ;n4*^^u!T^WQXn9qXY!l|+;MYS0d0CqiV224@R%_r-DD&q)q
z!2ri9Uu}!|XR2&wCEYHS#j1UY8`Kh^`s8l=jH>7ad0UN*<asKw?s?;p|8t)+V9${@
zfd20eM(gzL|3(sWG%_**EQv~Gwon6>WCS4Kqz=CS#dDvGb%bVb86Q}zNJ0is(E<hJ
z=!3V<cPn^z0wbUA;hNzsTa03tv?eCAa*}X|)-0E$3tE}<{CgiZlk~E^pq0r=)!iBf
zZKwG!sdv#P$qYXa3(ZtrER3X;yht!@4vODB_uYsry|%<5MMV986wX;xYufv+VR0yg
zRN>&=g&a)P%{SQ`T1W6Rii;ywW{1J`n~mj%4EB5(Qci_|7T1^Xva^c|U=>_INJt3%
ztfk?vr{cJw4KTlkoge_G$=|A#aeD3<#_$SWQcseAg}^z4-2Z#Xj6(@iD3f{VLO%AK
zlj0^2c(6Dz6_CvJR&y&&4DLqd!!KoTvw&*&t4Zn&8$yXveW?sr%)2+nuEW<gNCLd<
z^9vbFy7TUyj_^_j2Uv-){^^)31v4B|1btv6JO}p75%9qmjMbKN=86o2h3HSz+ir_G
zkpNH@EL%8hZFXfV!}ysv%gb5gBk`yZ#+}^&f?WRCsda@70Q84hRRFSu@L!&FJZUbi
z`nqwu+RGbL(^PYq7#pO7l!D(E<+JHqWl56NnFueaggV}qPNQ5DEK>2AsmgToCLDbe
z0IC@n{tr0EGzH3)r3TBhLO?A(!pSNBtE7C|B*mrH9=ql>Rcxvs8%yE0ah;bYEfaW|
z2y0*sJl{OS0r0^7-W<b2wB>|_aL?^kB8EZh8JEk=?{ZUH$C~!+(x7c+6&%3*dlMzI
z+(D+j@c@nS9Dd&fq|aNTh?8$*sQt@!d3z($DIt4r!Tfn`0MHj!$L!I1_SxBkK_ECA
znWVs2!RGj^;k<v;Iju9NZfPFhDo8Bc)OfPInb&!cBDO3nJs;Ru%1|^Mb=$0eoL3gB
zac4ahGw)VOsCxsj?$;kb{k^BPCVSc++x#YK@u~92;Y}_5T>2+ig<=``3<6?vOSKKA
z>UgPwGfLy-_=v!L_2-d<?4eQ=s3SvJHZ@Vn8_1z!`ing2J(;ObJ)0)o^PK|a2}usu
z3u{af{|)CH0lA3Q2VPlle@hR``QE<GEPng_lgH08mMG2d-6p#i1;h9}Jx7IGdIWbD
z+8qC)Q8#7y+`9ljD%vX(4tj?wd7*`*aLQCq-RJi-S1Uo=al=?vY=z9p1@=h|I1X)5
zvN>}1Ia@Gt#&hk3yg#4v?UpHj!uO<BmJjoc_W*h=zz>`(e13ufZz?{>WVbP1AyXhl
zs|p_)(5$-yFuoVlrECf>>g?0{E<%_EMw@X$UfqxM6=$vnNo!!Ax9Q=cv@7<-vBq@(
zoDqV9rD2*a+F!(%t@JwnRx9CwBKPGzB2rp>*-dQAO`ya&Q1@iAUjhju(?`J)MPsX3
zMB!_2EuRhQMEm)dAKvMv?rSMuIj&0fL&D?cxR2JVV~f;5c?o^fRb`m;cfKOK5Gj~R
zYF&(yaM{m|=v+}A1c4(Or2^Z)=T_tje--xF@vZk;qI>;9jw@jTCr~yCiGy_s>+F)x
zA4kQ1Pe`#MF8Nz+D+hKZ5|Ik3T-sQ+v%W<xy6lSwt!Ej98ZE1H#{v#wYosH@%ij5t
zdP^8r#4o6hmyKfKlb||?Rv@>FbD-5-JwQUiZPMD<U`$bMTUXg&>Kmqa6E5h%*lF@w
zu7QTxLsqEm4XuUQDl;dYMlKA;PCas$nGCD;;47J3v>!Ra+8;H%7NRb#W_hn(^q(`v
zemtM?A2pd{Ubo`^DN-SJ(&fnPAps6k|9)cj1LE0RaXZ`U+JxHBbW$p%x>uihQf@_8
z@E2grRHx2E&QPW^?y2v+Dhn(oZ>Mg7^84&|utaNv5TO&t=R1fx<g3YGAvB=Q6TXhd
zxx;|^!Yyf@sK8tj%CHLGkdja{=2j9@b6D0kT-y*gQWch;XO8V&F4+s`>zYHlq9^>P
z&t73l*tSqN`Vu#-k`dm7+DTj3XHU;6(Qwr%93SAz*WX_1Hb^2y;#Yxr0#A@X6VN!s
z_hOI_ZUEh?+d&d#nEBe`E1qeRsq`ERuMnd15+I^Xf9%qFis<#^+;HBCIezpR^#^n!
z>reV+eRL2$^&5*8_-+Nh&(Ep73l24ph3@tY7LOnP`UhxxqilQS?I0yN?f1CC;OBrU
zN2D4$SR`kLKSkG`a(`{qA*>hU>m`u6GSzCLY;sHja>*vphK~}#c=LLt=$-o*&km=(
ziq79oL&Tq}2!yM4l=G?jRsutPRztO_QT3`Wd`%qd^1|7vP*FLc!u@MnFwDe*&dp!C
z{CQ`Sh?EjwsvWEF=*+I0ogK#<Rhlzfw!-NGQOy<UN+#UhZ=xEBx+Q2<v7Sq#M2Wsm
z&9oj7y@&%-r>@)r1bjMIK>jBIgYU1PE~|{IGM<lrE~bp5EECvOp3K{jCo8?`GS$Gk
zyJNrFAgKIs<}R40+E8-JmGj5Bu-fQ5&{aIon*<lx=-Wfo9D!uf*h2PT+bGN#VSY9T
z3J!w9{U``QhE0It2kO-PV<fgdn$=51Ag`}Hr4v=GI~q*hHhnCAVwWpRD-l|)N!?dr
z8Uz*KH0xt)8<;gbkng}9<&@gY0qR_>*iu&ePTNIgE2j1dv*$+sh>?O6br9q?rM4z~
zq}@pGFsSo0Megk8K=`%ht*^;Hc`gDQ?JJPkxg7d|Ble@OiHnJ2RG9g>dbt8+ds^{r
zcA=`Gz7GXW8D(vzigV|Aw_R{aOT}B&59Q*DxrK6p{HF?Lt|l{MDL0O-m&^(mlx_k;
zoSaoYk=3C7OT@eTM`|7A-=lry>8oC^85zRyEV&pJxxJnjHjl>H$uTkshu?7AU$p1T
zzk6Mipx;%asdvx%r?CHuDPh;JKWk>XR_;tzhl5_rF;uZFV?n^(^COj^V6VW}@r|YV
z<Q&BU?0%3enr|&aqK9_lZM7DVJl;RicV7li*A;5<m36%@WB7J0O`fa%h7w>rAa@cR
z-LO9%w&n1UCLc3CZB2gbds~r{tkWRpl0$r)MoKPXgOD=oa2<Bm4OT<VUWmE^+d}eo
zOA%mWwzm&n!vq{-E&$rJhAHM4ct?mdO^<xL+MoahH<L%k1s@y%+;~#=*s%2!urz2r
zeIubw`eB?7zKN&&Z^UhN4r;x^)xVe1hXLpz2v2)`iwo%?<)mvi-bmV;ISo2<M2>dK
zrUQpU4vo_mucf1|Rgx_#pxhI)Cxnn<wSH)?d~LAES`Ff(YZH8$e=1$VQ`4;Ld^*k>
z`DTQZ9>rNJ5=}687u7_2`jz2yRWXUc!_`#TT*kqD>8nA4zN{!&A`qf;VZxYtBQHgO
zlf{JJ2G)8A3nl0&v0k++92V#l+YsyuW#U%=AOiY<IDgwNvjh>}adiv}HQ1j}=olN?
zy2Fy=!QuQe+c0kERBp?_LYE0r9>-af;l-Tme7(i>_v|~B6mwD0p97UIGvwEuIU1B{
z=0lkV>Pv?ud96gqet7PDtt;$G8YEcCd-MHCAYXHS2#sHo(1gCpTov4sX-j#Qyy85U
z6Z7GVM1DI<1>Zn?l16FNDwQfFXdnEYL&M<QyB|@exQAj`tWdcUpm6)zP{knNaazq-
zAHU0E%{qYl)Z=5Cx8=I?AADhT8%m{fdfTv7>e>VlOR9^R*A*na3Mays!K$8mPfMGh
zc^wQU2Qo`en@8^s(UP3@Ky|?Yj{qU5hP?$^X~-L!FOzS^Rf5vb!_yOe`)7}xS#U5s
zZ<iqNuycRuvmLV(?+b#Bla3~qj0es2z<ji>P66l>J4TfuhR59xper?N(+;zxP;&oy
zB!w(mVQ>Q&JXBH%DRJz^&&EuXrbPDQ_y%A^r%)UM0>W?R>A=l{$V5k7QXp8oDt%-j
z(v|E4$9tH=0NBJo@EUP_^Q!$|^E%8osc#k$IK*R-qjYc1f}dr}a&E4@v4cfwRe`yH
z(__`?VsN_<RzEPqKT#-g^E$lKyKnn^f`kE3A5!?6-WI9jPB@cnuV)%u<tWlxLoZ3M
zASu`biZv6Uvm$>g^~gWb|2-SsQXCKaLt0aMAj0q2t|}b24@Rn5E>@%nWisV^0nngE
ztk#zN_@l;36v(0nUMnA50WnC5FXR9%0BLZ9$~BUpuldwJTC!6FOhL6l{S!iCxCX*X
z$(hPzM=O~AS=g1m3$PpZD>y{ID5u(31FCGNb}UgXtU-2rs(pN>A|pkF?cm~aU=Xia
zU8ar$%)LpXF}~)jfV^DnWO8~LXYV&m;Ac2|hN%H0haTl0NH?E%(vl-S4!n<5cXO#Q
zR`)w}eo01x-#|ZF)-_&Fc=Gx!ajiaCjZKmmAC|SINs4iP7F@YqIG~1xxf(fW!Sfld
zYJf7K`XmDQU2tmw0zN+4(}=ER%@qX7hkQ~zmxetk|GMiFzzbIN#HqIx>#7-inG6N+
zkXV=(Mew`&_xgq<A3+Sfa8qOGc~df-#2EY6LEf~Wxxoe**(`6=YY-#}$-_$_QTax6
z;_!DOH3`{$nnlyyAY_0a{rZ832*zF<Zh=l@QspxseE9jkAa1#STYy23Dlt-}adniG
zmEz_ii!RgA1S4@gb2x)K|A-N|_-dnmIw8QQa2Png!MCDG`nDP))%K=8;K}ii-`JTF
z&~tU#;&C|wm+eRyo_Nx7tSX;zxvL{F;g+~h;jE+;0{D{eW3PVHekU@DwbNW=V`pQR
z&ylpgEnV+hGhec>b@I&3!#9~=p6!8umGct&{HioVv9o@v-r@S_mFhT}U?j8<v7i!j
z!J9Sqk^?`kNVl*BE|`Z97`d^Tk&I89ieVXe$oyNOaK#^{fFus^Rfs{$<rlu?km)e!
z-}@Ew9(r~@j$Uzjr`I$w;SZ$V>(rqSE0H5I+A#=BrN0-YLzV-C*Yx*w)oW8ei+FVx
zQabThN@>?wXr2!qr=EGg9VdX~mLU?NTGiXspMFZDG!T#J<6;3Rn6##>-X>>8;l>3n
zXrB*M2vNM*pN_B=a9$qjKc3z&wE2=`e=1xyR6r7_+{QX;d0F>ZilQyeb(W;kqas6T
z*4F-E2%*GXbEmK<@en;*0LX~|&*m>rWv3Z<5mJ1*^NnZI!&H4rNNyfbc`sUb)-mWi
z20h|(#{C<+iE-lJyzM>tfRlu=2;hA-Ut6cQsJRrAIr}nt!>vgV8x!Zr=FVh!sj1&B
z;<cC|9cUN0+PUz_K$SbFu&y{=x?f<JSKwMgDA5qMv6GSxPB;iA^nFi^P~_cwxxci@
z|Cy=%Nf&LB-fXe6pJi$!Wok-JB@t28r%3&L*R;p?1VKBNZybG_!6M+-f`2leR2zdV
zJ?Kyyq%SHaRG|mFvpR1bNUdaAgGJFWR~#f4FGDW_cLM>P@+%tOkvqZ-o=(?vNvR0R
zAb?Kaj3XlHOUMv*<@GqKOlDL_>Sg;8IE3vs70}pDv-NH-y=hc6Gbk~E%nQmsRGnhL
zoUr)BqLGuA6%GAtvyCA0@RKxMHq=ym4u4eFwBo}aZ9RA{eJ{n(WfankglgH^a_9d#
zHz(jp|EJ;x?TyzkTJ>{3N1%|#Y4$BZNE4OUze!Boa^bq+((3Fu#L}yjkKgohpJ%}z
zz602;^M$j__8P#X5_o0Gs{@E|rm9n*K7f+opfPE48s3qPD2ig7<m_bl;ei<_Rq%Mf
zA$6rp(U#d9_S22?8+u=Wt}z*o#^=YA(gs(<7Q6-TN-OCr|Jrf2GPel8js|Rc-h96c
zPAE?qr8df9pfa0lSl#v*Hcx`+hFpej$Wmi~ZuMrZl!DWP^Iq!^hGbA|97dh9(nyHM
z1neHC*{k$X@|uk+e^h^uc6jnE-A>QC=-fWdHmwq}Qwz3yPhOq}#2_G(gfDX_adJN$
z+6iU_0}5=>@hhUi-8knk2#d%-5MJZJDTKNH1Jd;Z0R4*Z#eJ{ARWO|Se(<q8IfK#p
z>N(&6s4aTqYMQRT^^ufYb%i78zKyeInfY|ua@w6sQyp4kZF_bT;@X2Q;F|$3)qR+}
zd7McJed_9^mhThOxi>K_p%ZW`W!v-AuwY=OU!!eniAIF-V!UHt6_E&*`n9;?Q3OkY
zeA6N2)RLRgE?m<^zz!S1<(mh6Lv)+QGr5t>$aJsbWBtK1HB^jv%1TIZK0McfyHYh8
z5G@8!xr3Z0fI1a0eKZ!I_~c;=a(*ZBLb4T&Oap8>fq2NPICrf%pTu904FAZK5Ecpo
zixk0^4iyP)UoE>}*qOM`so+QG#eQx1T0{h$okk&>!WzuWYPHkdX$|d%KdOp?EI`c%
z`JV{dQ$N#rbpz<YO;&5{v8J77?xsp~UA&n7<qkLVm;c$r51v@7;pnBi&YRy;qUGg=
z0xMn`zEJK~3<_Vqs>QG<cnv5?Qjt>IQtl*k5k(ESLR^6C$M5h)R8l%jG-$7~R^0w#
ze_?w8+Z(%YVS&9mM!FKHykvpn0#4qx4?O0{OMWV91SrmRKLE7yE#zC|E(3tl&9pf`
zo2k{96VNg_Sr47-{VaoU!k^QhMx$yzmIHAX;BVDG)fDP$h$?vd2%6uS9w_6>Qq{K&
z1dO^Zga)eCN!t_rTJUn3WlIY%uln1zfK74(A=Npo^RwM4P}76eAt)qqU|{1-2aCIB
zB-Qa-7KPuUe8n3j&y_dtDn&B&BnEW3z6J#pQb3-1*fl4#(e7SY#xM-B*`>`7Hi;2^
zAQ2=Mc2J$odH8ZHGCZ&sS7fY$!~#9lWu7^4%70-Rt4tZ$;r)G-#UnLc1d-w>n3-c*
z#73m<YB_ix6^+XX0Hy^7vanD-hcI8zVjHMCWdS)0jpM|*91``~_E2%SI8pd3B%*m^
zVZl5RgZbm}spI(_Ai?s=#oMj$Co&w3sa_rZiSRZ6lU8`^C?VjsagahN49Sno_pv`t
zy)~L_XfT^@4o9j$4k{*lp2I#wk+F~!fYx_35Kw@4^wX$WP?(q+Nj<{`s@aF*Y5$D2
zJH$>8D|NLvK_i*)#cl{7l6zaQ;7UF)Yb|SB26%JuJHQv1sVXnX3nzYngU-S&NU&g-
zX6C37$xt--HZYO@(RTt0?s1(n9d1tt;F#%s3Q;#!UujggD!MCzD=vbAZGd9TGlC8{
zW1c<}3r=~Hd}>Ycjrh|LBv#U}pj}pG;|*PZ>9r5;$*t*{3Nha~N0Nz!9zj?|{BSJF
z=J^t8uPx$|V}N`1XFwS3!U5<8yPcX8Z|!TQlaK3w0qdjAhbCs=*!8jlO2n=NZ*5=@
zuClp9f`h;zi{iq@e6<oABtOFRJ86eiP0i<EvSXmMQo-gPh%OS`2dqGIFecE@ZcM7v
z;Mp6z1P4O(M%l^pS&ME)Rg61%ysC7a*FX@lf>>qWC}<4?BeiS(5x?5z{`lg%+PV=W
zJ>rc2lidVN(_`@DsrL|Js%x{K2gG{?n!os#a+jWyi}>)3eUW|!2K3jw(H&1x#uj~Y
zL{z3U&xS`L(qsqg#E+*^eJLjqZe_I_QAXJ$4k%1tV?rrFh_@U8<H<~AK15*e3D&BI
z<bw`JKu{K7@;qLj15@^v1nSY42^#y6ho!6ZgL)7`Un2XQGhbFQg|CHz<m^4KmQ0DK
zv)9qsCnm&yyct-xQLHM|eR=7NnkEihY=`f@3f^21P+#ZB2banAA7Qw}p=IErJQsWz
z`?W}nM*$Mp5SKub*@+1|@B{$(b+%G9N>#Z<iNPQ)x`P#M5qL$wHf>WI!GhS*fp#bI
z2hzMM6c0);KiXLC9uiguF=9&q$6&4CQT*0B=jP_6MB;OSY*#sdqZxLjA8!UlQ9Llr
zqGu?$98TuCss-jqj*kpaB@YdOL-Wk|X_Dn{d`kNTzX645Fc7`{#dc%cg|gECVUf0I
zu#q}lGTg2|fvDR)*tZX0azGqM;ka_3>#Al_U*-M?8=$e%jK|^~CsvWDdq;{Yw+dLI
z7HC|<(eh|YBo&DXMSa2Z9>X2^sGG)rH&p}@MN0G1gooXDV*@Y))j6rvxXIXit^8yJ
z;mZLyFkSn7DMUI{`qN+|+8YWl!f8A|AX+dDnTarHe=rHG<x_-Fq!5pE@FoP9BtTT`
z#_&0sspc1A(>{1L4g3BRBXX4~KbIR7cZ-VHOl8j-e11PW^<#5U`P{H2E;+^{!Co;v
zQ7Imdu;o)G;|JU|QjBNDFh&S}uv$3Q^DU^%STY4cMAzfyI{BK#p#xc{;Yi~UkOy0k
z`H~~GRk)Ui@1u`v)-T+vU!HXFUqk=qt=D=t(}Cd$Q0-6LX|ZlUuGM)OtZBIj+yT9)
z&c~`B)d8q$G23BTqgnt7BG@XF2ALhvm%n<rnl~ybEvfHnUv&nI@zw(X1g}UIU{3s(
ziyA74RSK*WwQz5TGpvG3N!9rBz#U$9+=h1@p`=KQ?Z<NNxJzBBts%pIZvv!N80t0~
z;DD#l-nrJprAqQ#;K;IcM3RP0&z;+En<Bek$@xrG+MbRbG#Xp@T}`4aI$_38gj~U$
zHO`y-a9&s+AWJI&2Q=U=Vp&;%vcBP2qM?R^i=38Xr!8o7wjY0&*n3WR9^_P`xKrwk
zafJ#MJbNTafGB_Rax`hFaJir2+yL`CDfrsRBz#59(AxnB)uDwJnqknDF5c#VZg$66
zupH*zO-~$1qK}Bk4Ngt^+f%-b=-}M}Y=<bQsZK7Lfa6sK>^a-pHP$}D345F|c7kh5
z6_`yyY^4ZX+?Rlxph8{pA`yUoxB3lJz7TA2Dx`=jcfEImvugX5sEazrMbjn02coAS
ztjXXArx43txhh#YfK-=`sDxi6DDVSrkRZtSszymHRBDB$L`-Y3(!t#HF#1V;U^lKK
z3_w%fd<{*o6Fq?O+zIRXsZ*Jdlp)$F<NR5UUxL<R%P?DP{aN^-0*OX}8j3iz(eLJC
z^EYv98iC)qgccJOEn{MfmYfUU_A%e>n~B7hi>x9{BsjFv*Q|aE|FQMBZ*BR^cLgy(
z*^02rq+uVBpPgTrtGJFTWnj|{3Wbk@S;>#FUB0;jk4WayfJ=Zvu@lQcIXm{hc>1Qm
zyrQk!*lg@HYSP%Y)u?e}+fEwWjT+mwZQHhO-`#W0f1mo=>~F0(#~heT%A5#&IBN*{
z8emc)H&WffWf{G3t?w6=Ix}uL4>8Um1*`mvkP=SVlCT(^nc136{bsrN9SEzOKqi9L
zZi^15Nw`G2)&Ob3E(oK7x@@hyPc(Ij4VAzE--f%P(I#4TOK{2G^CkVCBJ~IirIMN3
zBXyvfP+L#}%xm7$@OZ7u$I3~w{|#<+Tp|_l-N51VZGMK^#8rM2)>G0jBH%#qZ78z?
zwimFF#f%ZfdM|l-wT+@S!x-z(b!6w#9|zAfe%!+!dL73L6cRrHu5Nv*mH7VjT*qH1
zO-TcnYLuRb<J`#|@cqr81np;rXvp%+k2T=T@NNa3Z*qCzCE+f*IOv2>{J_H}S_5EA
zl9N_dXJT>U5RNF^vYXc7SlNPbUX1+H)oyQ&9~KBOWL8CqrUSew`^E~T@%BLbw((VT
zz~3;d+w}1v{Md(sO?uUWE3qlWznUvqEFJuC-tk^?X73?^2fy130BxmhhWv;a9%wd~
zl>oD9SWI%)Uu&4E!={>kvNwLnpm6%lVfgY3Q#Sw!+1FNLp5-+40Cb;IEmPm*+ByUe
zJ{<Akd2>kp@k!_0dAZHE#iV}-_*uOm|IoeaROoBu@1G>5P^19-(&=qv=B^-)Ka<9Q
zhCU0?OmDeMW7wcByh)coAcFS9_4uOyVHMGuZ7Yz#&|Izmm^+yL^FkgA8@7!+Q{yEU
z^<#w}_N|+vSeC)Uu&*dRR|Jx3!pvbFiruGl_g=)uieP2r)F{s+uTwE@(*phNo03%L
zPn>op+yUGyx1JAJd(G|6%YeU6K!gM$Qs)krne4wUX;TSWRMd3nF25N8Q7QkbTxKty
z%~}YcRbW>Dk!P<#s8g*IFf46{<-q3;J<&aBQ`)F$4=RhlQ5O2b|Df%17R(l~-Mfh8
zSl9{c2kb2XLz3#P-@(!w@q>{Ez+bpbq%eVfvPG(%fj^bP-a+%1z=$VY(^1no>&NT#
z64lC7_CjZo<>&z+ZDSiHP5y)}SM}Gd8rq}^W2s-;<<q9(Aoki9Yca<O&_D??J2rqQ
z8G=JStMi||A}yn@dY%HjZfL&jn)3+7MH-Db$M<)LH-IhP()ak36>j!Ftq3r{VDR}0
zqG3Ker^gItUk_kK#@6^2@~CJKdOD&I79(nipDdavWlj}&i|f&FHzK&<eBrIPX8sEM
zHPzb6k16*t!diD;clKNstUGuc`I=PZWA0@(wVR!}Rz;iALLPH+MY|SK#^52!R~}$`
zt&J8>ehd-tKxyk&yIOX?CrwfN(#{}cs`ozH_7g>ahQLcwoL~;ty1u=~Ug{<4RzDtW
z)T=IF(vXnkCDNX2kf6n%X(~Vv|8DrG7379Wn5Op!Kd)HbRyGG*?C({tifxm0U-8}#
zrr77lOCD{{zP_}Naqxg2=#vV$w&w@v)Q6nNuYG6w1@u=tk3rua6NSk?m7D+U{l@1T
ziyaXEE#JKM45Z4a_4aXq(;Ub)0q*_gGq<g2(B?{G-CtQp60j~lbJCS0(Q~WLE~~(R
zlB}-m@XxK(FU;pUeM#~I3Bs<2)}@iBv-(ppgb1)bAz7TLEp)q(xd4XEMTSby2!IAA
zsxb&rD?x7ErV`waNf!puS-jUe)trBfdf+>C!qkO@zF)t~&O2Hr>Bx&1%Z}WV+o*XM
zSfM*@C3Y(_QLij9Q!XZ9DLqiaJ*z6<jX!Guya48&h4txG)g4A&drD0e@i9=pe5oXu
zMGs}jCM0xGi`86RdS#18*cD>`sZ(nMkHMXTu#0PIN(?aO$)@AHd{~uie@uZNcXW~#
zdHMGgV0INrDz&l4_)mg$>vczW#tnJX8qL3u&dp35hRkco;>GF$8gBn&3J{aJ!_TC5
z4urnWup^R20^_bAC3bM+M07eew#w7(S>`{kqPAI1^R?mT6$javAiW-7G@r$Rah)*{
zl<2@mk{HItm?-_d@=P6F2s4jI;swI*KcXi<#Ofu_p_}FajHYXFpX{m(-q6k-M*>P!
znjWVQU=XF@BF9bYO+5ZFOO2!aCAhA^B0WBX#qn^;ykH*6lEBAA{2V86>O^A>a$G=u
zuPRcNb)eSu_^fi(|A6!{zgfUb-_tf&peoP`9zP^chK3>pPmGUJ#V*C&)?E9e)ph<$
zGJ0s6^7&(+x5SmA1Swj=2n;VYU|3fqa%(?+4f`*-3E%F2oE;0qh5n6){GN2t!2uKj
zehGjH?4QXHhS1%!sP4DfRrZP%1wHzLhB|!aC)$G?c7ubOQ|-x+I3_1OQ>P_TDnOdJ
z<KF+=*t`id0w_!SRU*ajR{`NdB#fEgkU)ugaFuI3xE|u6u8$KaTXKrp@WIhTWJ5(e
zs9qX}lDB|?tnVw^1)G5E1O)ID^fzxX6cnN)6n`91B5Xzp6n&_E$ocxR@W^#`M{)b_
zkp+uXAD!i&aW>fM%zFg2CJ2P7+D!e+04$_s>JC~!w9D2>_wJW|c^;UI{plx!k5mW@
z*ci-Chl7{vyu`uL6oAb?zDezk8=b-Qhr8>X=F{i@eT`bcr^!MhLz5VPW$xYcR0kL(
zzPy~Zbixy=w`!UIF3BCV4}kq|OWW@11(yJ)PW>sSVtT0;od!Qle1_L*L-Wg`qs2``
zF<=Sp84G>Y%eygx*crRi#S_rpL|)UkVBx$QKA4xg<L}UiAX8&h&6CX&*eGdTQc;aY
z&uL4XvUO}eaSywlVi9)io7jt2Wfq#A;KVTcAWXpot2<fp8>|-$(n-MXPN++Fx$!&E
z><`l(n*rjD;yjZ~2MK-T?x^`_`zkj>{MGtGv%gL%>0S$98V73qdH{QCHjj4XoA8yW
z=r-EXqC?mmHN|3>-T#C%Ko)MkAL_ik>q5QNsneg3h8dR!e3U!0c#`;h&->?i*48j`
z2$SziKm={Uy{sItXe>;t5B+mOf;tYI6D)RTgR7O9Y2Ruu@!C#JWD`(|P{3?er2=>N
zuswaM5Wt~kMXiF(-K)^y7b#$UfY($sQyHfla5l{~mVC~JX4-OM(<))7<0IZ1LfD0l
z6eRd&v2@pKQ|&dc6s%BS$wu*W(nb0YN#R{9Z?<&I(KO{qxaHpHr&y%P8^#kOd=rT(
zvnuc+eHs8%kI9^j;)!cNRfMk4L12F@P!|TPo7l+~?dfIZL+VsD{%rhe-9nXVvF!!I
zKueL(n+Kf#6{M=JI4Kbk=IS4<xv+<I5AfRbqFf0<afl`BFb>|kfYB3~TcpAfD5+#R
z`fM10cZk-Y9PsD9>%Z$O>FY^}_J#o(4WXAk%XiKaz=f5Cl1y5|9Ge813n0bOlOzmu
zl1uPhfHU~rd@pSZ^TbtgPPqc1G^_N_Vu}Edev-D>ohkW<xDM&pnf4+ogL})fKAYg)
z1CZde%sonL=TRQq0gfD;H#z|b-u)q+01c|!c*LyqtFXyG7+fUP-S&HrU{o0wG!;tU
zgll=9d1tmlwinxdv7UBrgkW;5_6+VtPLeU;2kJt_^EI>ev=!_a@B77=bw#`SKk`3t
ztlQf}B0Rz^?v{8Kbv`?%T%N^pf4@umcM@^DVVZ)dVPKWnO1D{z#8`37>&FE!Y76G1
z3bJdi)YDXa8&{wh5F{SAG}@E`MErb`JUw^nu*3yw#GtNVmL=Oq82P|K4!9MP@VjU)
zKGdSpXba9>d*ZTeWnnTM5=6`T7K}R)Bu%N#cwM`xD8YZz7mTs-Jz*^7pDF6=_1H^Y
z{2vu&IiZ)tPkK+i^lV-V{drP=*bmsmmzJqlPMwFW(mDah#}Z{kug*TQxOdlwK4h?n
zsJ_kM#(yc-WS~-5=af&aZxXn`Xvap9B#*yJzXXw!((-zSCIH@Ro2Ce+|I!qg_RzC$
ze*P4652uXJ<4IvCFu2+b_{|<<d}GQ~BKjP9n-Gjrbb~vnd)Xe^?98RJ6lA#V9xQ6U
zK{5NQ;G071?ELKVA-B?AthG$^*p7f;D&-D0!y+H(SX|&XPIubkgAm(|qr(&>5#i$l
zU<@dd`?3>IEPM~vtCI;2?pSF!&)65l4bCROa_|fC3@F8vg0ViU1M~TTWHlNubGYoH
zSW<w@F{Ek750Dz!Gcp#XsoMcp)qm{BCFxcRfRT5SpOJ{_+$FeESS5s>vGyq0#AFWT
zpWPmC&FZ_)jFnk4FO7e+@M=SoMKq!P>gVlh57Y`jLqA}=98hqG8W-Fu!LySFL_0{u
zPZI+}nXM@z|8JikxDNgQ4f+ASmH1;jTDyG~kWeVY$8S=av$jJHQR{g##dx*69lS0~
zq<=_P7$`s|6YJvsH@ye83kI-oF+IAuZTtBytovQ@KP<R?F&GE$N2sZHcXo^kk|A-f
zpXmogLWcQ45g&?;wa?@BGg+$4eFp1Q7R-02pq6*Oab>b-uVW{PnX4@3&}GapEOgBS
zcK&VRw4Nlu+Qi7GkevN5$H4K+Q9V>P^1P*^^slN_$}EdvAe*WckhAI7_;50`&x_DC
zR%s`{;hW~IPZ#E{F8EUv{eHTpGgS_day%Dx6>%qR1=JJeJ}X@c+W07jjlb4!)Q6bo
zlk^{|mzzL|mo~D!)bz1uBzY<<6`OsXz?I=LMfd;~@X_xwFsY*<M+L_u?iDwbLLyDY
z>G~zURnBuII%-ZQuK4G{Y-<wIzqJ}q6F@>|_ZSGmb?M#Cz4R_=HB#rRo>U#`DMJ9h
zGr~rfHR(;!>EY=~xe6gYQ~jeXD>qRFW{Ytc!`Pv9V)z+y*T#>Pm;J-Ua84e)N^?)%
zBQAL#j!6$Ex=)b^J#ipv+Z)fy=?t`Gi9UCmv&hqjr73XDu3%Gt;g+ymk#(vkVR^ui
z|8KCVP+kWLui$fUNHD>oBl@>Rw%H)-ko=|}c0-hj8PI+J%8p!%7`Pw^mmY1mD<DeY
zv-=eBC7y^*LN&n*W~sHJXgm#N6C_K`N6csh-cba^<Fb|(DJQ6A>JL=`^TB!oJE!X&
zz$*q3m4|Is31oW(P&{{k^wEG;is(hb3|#V`uUkrNK#%_KczGOVF|mCGE+&wNw`={e
zw?t$Nn1MWlh=9AmYyoR;$O7Q_iQU1o&Iw}$m<9mVJt{(g`eI`sZEdO24{A#|#}tRs
zWo044W^F;iglY6mE_hQsp7cG0zEUKzQo^70dD8PS^6A3ks2n<_XAdN{YTABLGutvI
zCjUd7QMs$@G2akufK2w6)FrYnI;0M`S&S7ZZD1p^jqKebJSbQzt`?-)$$;z4gdm;V
zeQ3+$v;3X5QjJ8f4~IpT<(lB<_V@jWk!_T{1|X9uu~{X2;0?ruAgTTsE!91CYvb!5
z8&he=l=`ZECGb%?H4NVbtGF*5&E%NsU)@(m{h5!_5xvZPzz<VEP3m~=#T_l@^+eOT
z+3pwvq-?rf_H}zuRe{i759Z(Rxx4nuvhqy;)Ls;pk008T@RYJhHeH8nPI%3bs08p?
z3!1yQ7z!R1C%qp>7E0!GRq??{iay_oh3{g_5kQZUBRgTS7R?NG0A2h5Kh;0&6bQqb
z;!4w6*xK&zC#Zr|`0&RThtkw!I|=&)3db<$!#gJ$UoCj>ndU?a<Hc$L*LVAR@|*W>
zlm`I>j4~^Ayg{U8PLZ+Duri#MugU?|(LF?<*2S|?)leU3U7sNOMD<aOid)tLONhf+
zGTHA=9>gK50^d_cCE@_@+bB-FFm5;ky4kwLo4I+)V!+DIXPC@1F~?0|R(0pW$H0Gs
zliiPt-E~=2TU4FW?|L7t`di^6H4dX&LUMx{mBCxo`~Dx*A5VM1Z9_^ir+TAACZ8bQ
zw%bana=u7>x7TWhAyr}}^S`v~fD_JxS?d)>n70Jo0yzl1i}-n-Z8M4;532k?_~56p
zENkZG626k7T%DsQ${5Kup$_mPMce2ebO2LHlTED>6CBj!4xI0Pq$eRhR{O^x04qe_
zDZTaD-VG~Mt(rIxDX>!;<5*eFauCyjJZ`y;qBa+<QBaQjVU|XLsi_4_-jR#!J5DhR
z)GX6N=4qqUK{6;8Nvt%dC%Z2towy1FVN-o=EU1#GRF4@?&e&+#hcUZC8vj!VHn{)B
zqBfoe0Q(P6TORDa|E`6bqMaR0DU_FAC7__c6$2X$Q652D6mpLU&S^=P05-<V*`k##
zeZjfJL>*Z(Mi(w9CO%=Fp8NJnh7nf?q|igKak799kdx$GQs~?hNl0gz7K9$i#yhPP
z#Ge3fP+nLj&8x`<b~*`Qr4MUEM$VH`NY5D}%J249Psn+1xWj93_<N;b7^|;W007tt
z_jKj)l;CsRATQfSxP1NIaYR2MA9%Rfe4(+kYQVp^?q+oe#H<KCTxCq@#5DK(0E_SY
z=srELu%HAc>_i^;cXw@KG6gBj#hz*iFiQQ~+E&$tFXMGaz;FfaneTq!^cx$<eF78m
zqQI4?racXCVu|sDz68@CBSZK}#2XW)g{DEG!>#1O&6K4Kf8`D)mV*$^eTeuXw7nGb
zgmOllDB1HhnuIVkYzi<;SpjxqKRrlk7iq>GEs$^J^p)S7xF9K)q6nl;lH%#&>#qV>
z<WlsOD^6R(F4%88Evjl>UoLWnYYIeVW9~TKuKkqY3y3nJA^{J>mguUTjkxiG2$w$D
zN60#<eJH`OuF`6OB86!g`CVP7btbo;pH?-LI3d-~OYFD`gqPVJ`?!o=N3Idx|1E91
z^0z}N*2x=&GW(4^=<jw3lG$!BR;q}dflW7nJfispBlMr(8%ZQ$9}sCi_+yD*P)7dY
zb$OFDf+a@GP0V$nwus&H@G2u}2IdYRn0*4<%6i+-ZVT}L=2#V+UtDSrH~oo`9s~Ec
zoaSHnwCmg5;p!J%HwI#jvEPL92TF1IFJJz!(mvc`4wRY8@st{skf(WR>y&7ISqWS3
zb9t6_Ea}$LLTn4evI1OG(i1c|bp|L6yQ0Dkfg#T8n}Kn1yd27JltZH*In7T^U3_QD
zZn7rx4MO$nD5&wO7HA?3;dyE_d<h<~QOE<MFp}ipy}4k(IuqPlVy>l5Krz=Z|L;0C
zf;U}!nGW~G%E7&?(C}F>A)Vto0N+n;d`=yd$Y#oVX}@X(KtL1C0Sa{q%x{MyyggRf
zFbKEYA|wo@xCHj^(Vi%7$;afg(7wpJBpnNzzmTUvM5J32flyhvxy;`u;xkaBPCR$j
zAIm^G@}#N&yv@RT_1EezO_`?1F{COIJVh;eNz;%^tpz9Fze`HmQrdG_VH)zwe8kj=
z&R@8RZTi0sk@#k{b!@+(-3o5kcUcIayxMfMtFvAK`GW5JlT(N3Pm+p|)|T_nHa~Qg
zFoVYR^@7ByWv#~14MwQSF=|<+IQ!RXz!V7FWyl3$;QedByGz3WorxT7$}5D6d3^$D
zlYsw9)S+#I)BKOyQcu99gN834NK*l{QFC|7Mp%HnN?3(!T-yL6NKDrbDyvE{V+B<L
zw7*!%SdL*Jwk?ulVn5u+KA<Wi_C4TxzRn3y6Ga0J0e>4ub9!*P8k9Z|1kfj1i~j2-
zCZqQ<d?S~)4>q2ckWiBH69ZmZm&iBk3JcjbF6WZ+VbH&u8)~t-jM?xIWquL5avQJ$
zqr;<*P!Ad8{vAV9XAsgnYLmC*f7@f22>L|hA>{>NUQS%D5!^qi!oIOlvrnG{SXi;S
z{Xa+qXz3LAty{CkB1=LJrfOY}4x~o||Kbk_kaBUzGML3-SyyF?%V#he&6<)hCz3E>
zw7{4$E8;NtPLO4L+?=m}!!iXW-bbdJWjo_5Ir%zfSnf<VFsjN>LHwAHii6K1WKI={
z|5n(el;<0%!uqmsyh>4;BJq%{(*O9vc3+w-y%n|2+z{l3s7A5>hdOM>cJJ5cLX@#n
zUUy{o$ihQ;YNd5W%Fes{$)X;2P$;fO8;!oftg<X>Y=}mIZPx<m42Q_<2x_R}Z5p;X
zGHC1=oxS$mg10!1<S;<Byt#YlULm<Q0)c>kE7rC&T^M+Iz8elO0UTA_`ej?(E9yJ@
z-bweEH|G4i`|2Xq@uWAEK1fqyb>ayG1tn0TzWUd;8W*-qk5862X<q{a=y~NzFX`Wz
zv4-UMzGHOZf%s>+-u*d$rR*<LpeYjLKu`s{$tz1@Cr{Zzmc{7lN-YIVtWd(tf96K2
z_4db@a1HFJ&(F+pFMxRUmjXgXF}T}z8*`dzd7A~l0TcF+@ONEnAX@fkor@e$ukf2K
z^31~?SCcoZq7y?gKNwLk@{kCTg!I-rv3dS22VnVJ4U+@7!e}Y?4SmYVZ~kTSPTO0J
zin>G)0sd20>79W8^B+HNv)JwW4eJt9IN!(tZF65d?Vs}zSZ#8ZQ>Ld5R0k{@EEq)-
zqjbQoHe&NrfYW@k2p|sdG}uMxpvYt6X`R$pfL0cpew?ug>33>f(W{)-&_lCWh02NB
z!4k<Tnow*$OL*m{?o@s<B7E6jI}^|9PHub8m-_OPvD}Y33|68$Kd4%aHe5pjgC8(#
zX}Gde&&K1y*@5@EAc<#$-#TtiX~{QjFF?s_dKg(PmG71?fA)FWhMscqI&(cgN)~;m
zgjmS=cmTwqA^O0RHgnEIH1ABc-`3451Ay<$o8v|JBKH_Ao!*0;0N9UH9@F-JfB9(l
zT};SUvGg0DIojWD<7NEIjRQ2t92-f}^KqXX@mT^TK2Sqh6olm~kT5Xvf{FNv5?!gE
zQ~)yp;L<F$R`I&&9)`=`vFkH10~PeI>2*Q2=`7%#p>*B0T?Ym8%K#`J>ElEpcXHaN
z)*5Hc;D-0c2c5k(m{vg?vzug$8O1iB|AU1Hak=@l6=7jS0~prnW!VfzDmPrr>TFe_
z-O(}3S79$`=qb<lSq0H$V>MLxDjs|7+a#y|TitWX0gO3NB$nK6K8AO~0D3Y?I+7cR
z-2)lruiQZ)MsA|e8UhsV`8!uB)D{_@?ux9jKn!4r#O~|-cO_z&)=uuXb$H-T;A~A6
zo)7I=Ji<->-$;as9}ynz#1rCulS|WRNIA3d7VxwNMdW^XzemgNXKk8n`VI&*s$9{f
z(WQyiD&VJ3H~Iq0+ygYcjp!P(ybQ_-J%Cj*eu`?kry?*y7tot@Wfz^at`#Z)^59$^
z3Kn;U+Lky7)DL6&>uq$ccL=PZNG4`10v{1;W_>D;iqKD+e+h7eP{yYm)d=iI)r~Aq
zK+=Y3i!vfu6GObk#SOx30-fG)SvWllbD+-4r*jO0S1Tk)2?9Z7Jp%S&zy!a!pQxfP
z0^0}#PXLYbCsx*FRfB)i4iFni{yb;9RfL|W^rZ4TMx7Qkz=S}`%p@fI7Zm?5uxF<U
zGx_)pZUPPDs4jDNbrO02O#d_ZUq*_I8xVbPqgjDFYHsDX1zz)Jf8tiY1-S_yBZ?cF
z#M-~y_l^cclfa>K<8pg*d+mNaL1qkG;K7Tu*c>4NNQzJH`h(}S5OC1+nOSY(YI0Sa
zhA@)`(v<7qnia22DmiJhx>c~@fAmfhK}0Qm0tY)Jv_4j7fuYoQu5vT+H{e-P9?R=b
zI3URUc7f>wXHxBds{|-`n7oErP1Up7g@G18DUhGP@&kO6)paBzfJV><v2Jb^m^FU)
z$4nbP{&5igFSp;Y>Lm$c1TYH6svPxY|HoC;;oNPaF_ywB0WLg_e`|1+l!d}#C@w$t
zR{Wutnn>e~RN4Sr(tuH5l6T0FiJjMb_`EY4-mmLh&>%~VtsE0{hV!#;kb9$!gN>E<
z{*+dSy}^tJxUGW7e-jtB=c%G9Cr-I`TaQ{eJr`{S7G|ptZ?zeL^mW3K1}PN62q|%%
zFS7g08^aP&a<SU7pth;x*p<49T*?|=vhh#*pFnHOT0SU2d}wrT6ypGFQt{u-?xl~L
zPKrn<p+-sjdoyAtj;rHbCI1_|PQvcA;dw|}g#&V$uh&n}fR7R&C~l94M+#O`QHbeM
zZZn<5f?VU;R=ZLEhvFZ=J+FRhu-M+woBi&#vEY3<>wie0<umZ$d{A5uku&o=!TrS`
zzkY;sylmspX>^!qWPD11n}mCKa0sO%c1}qCNZjw=N%}c4mSQIR1AG)GvDX^NfqrcT
zB|<Sd%cusVB{`QriMJLOn<}e6U&|KN@ihVV#PZ87@d$3BmC+)c)WgP-W{RJfcnQ3y
zRA4RawAdv<`+@;&9WGGu*T}_Ku0k1oW_T=vW+1F3nC-NF{rH)tpZPBwvzEsJtJ}}(
zQ-aw*tD>W$CDr&7@KWmHhipFT5?z=8D|ZC&VZ!ba6+|xOw=8}y|N6KT`q@Q-)+bH&
z;x}RyD67J&6T}DPBTwY>5`n{#Lus%B=rz|w3}`lggkFCX{P%>|;}%JTGdcBnKiRED
zhlzsU4g=}GFl09c!LVP*>e;2y)W3hZnrHv&H`y~bwCfit+ot2|rGM4})71CP9xF3K
z@7e)&1$O(YtwF~`u5&j4+jCBR7Xg|Dle912ja8KuQrqZ$)US}yW6%}CeQ4bn7@YqU
zz-rAFmzd3aWIsOeEon}3SUhVH6t_~K3s|#S9x|ArdoJLc&iq=Y4)X~D68ORc2}!CM
zw|v3pIR*$-y@(bmrYky)gJO&>h;)&{zqzi_tfAP07$9S5RU69y)4J5oQm-7+sBGG}
z3Ttn&-Mw%$5^JN?hLIH(J`=D-ZBR4n#v)`8$O1J%H--D&=Da!Hd|OL++XSN&`HlEq
zFH^tAd33XrWuB`0V9iYB!sEpq2@fpni$*K?U#jO7b+Ok3*!2Pt`L68rr|!)b2q=02
zYR%7>la`Y(^4lyz`M;mu4-?>JVY%?Vpm>;leqd~8#19(=mtpPwn9>oUDQdYNp)($r
zvLa!`qL2zGqSeITLm(j_e}>(Wv}1kRa2nP&6YEDB9X2CRZnW1j*Loy*U9y`@aCUTp
z1~L<vFx~zh8EcWU#-H|%Y|i<)A*#J1;2wJV1(qPRs;-K)Z~??yTdu7Lel3oL01h+6
zECf_+O1<uhDB2|pB^W$vR4YkX)63?O_aUa2fU0#`TwL5xM1DfBEwG@saA{zv!QBS3
za0=_gdFC|WUEW!MFa&Bt;Z4yNUjRp+HLug{Q%|Lrp(o@nr2iY1^of)rBm&;5jCJk_
zCHt0{x{<m|??o9hA5!u*12{+P)g2gG!p~X?r5U{U512sg!@b=&BUz(ntgA@;vUY}`
za3PA)kTbux*^kUm+Hr>%dLC#uB(`5r#(13D+ldyJ<W~y3_&iu%7eDu?eA6ElFOt_7
z&0rv|$)NeJO!sq8CZ88@%ey<S2M3f`Fn$)0-BiAFeTZ%AFZOR01XPGv!9?+LEztH@
zRowon3bHG>T}2Fp31wf`<7&aP=H=o0mT|0Fj4Bv|IgrI_36jMUx4d5D!JJ^s3NM%t
zK&<ffh3mxGkcpb=_u|h4J17y^BwMUI`0QXV#TLc*x16jppBe_Ah%nGF-S%E#==ar^
z9=M(?LW$tc(;t<+PJa&=5XBko%u1}nO(DSHf4~h?%cyia4p4*UV7%pHGp$bA3`Rry
zTIApqh0aEggqc^a6P~a#KI+(MZ43$iqGqAKy2SsB^3)q@RNPtoEkaZ2<>?vA&fpe*
z4}MZG!{4%*(0l!)BB5)dXpC~-VuTzy_e;lSOwECSn-qcD@acBezLqr4b2B9b1ZWlk
zANL8)@HO<{GUR97n|<_VhV&#|`GXokG9fIA4=Z|_l7z2%Bozux%Y>q$0@I=yUEVK;
z`h)X2oIgUb3NT83#PY9$`TSLbm&HUY(7RozgAC%j@IBBnBI9PxvmrjFzH58Scdf3p
zE;<{tA(BM_2Ge}!b*(ny$KUeq@+5@7{Js&Tw|((|lp6W)<>jz;U1ew_!$JXls#2nt
zK*mB`^?A)Rh-<-h?AP?ygpwJc-0kp^d6A!LaGSVYIjt*Xu*N0}0)-JX5j#ve;Gb|s
zxUs%j5z$BFLY;!pUM|hX%L{ktdt{pbZBD<dy^h#Y+t3p<8NOh*${#9=-N|NE=&pQ?
z<#7A|za8i86GBGg!PkRdFwV0VhONH|4Rte8H&2nuytU^j=Ahu<kv~<vJr1qzH&$X+
zL0-ZMcWK}l33)b>T4>L$*6=`bL{nEMtSpK*zD(Z6<e}(3Dm#6LNWtLo+-ErDRMS@P
z7_-7ziJglT@nS?|?DFGd>M%%CiFqKt2i+1Krt#@17?B+w2p-A7&7odXE#qtt+A)~b
zv5mAdtDNAIBaLNF60XfEbcts-B<5^Znh?AT=N$_uu{W62RhmC7u`J>9`LgA25ApY-
z<p%32_@(ptBQYj*N~#^>!35Lj<LH^xIpepd``>v6gdo_wf)y}tZc`UEFLTxsJ&>b{
zdiPTdYw7`fg5E%43_7@Gl)Pj4it?*;&cBB{y=swKo*iYY8HdC=u1Gid=zj<LX8%yo
z=GNwXBRH}%IEv3pcpZtGJ>@)!d=M1MK0q9Y5<(d<y$Ef7mwB~%Sdl@#T@7u-L5AEP
z$O4fUjdFEGtigwvvgGLH00EPWNuE+Mq)FZ9kRA}BAm<Mpl3)|$c{5Xomw*cq$rI1=
z>M|<W9ZTE6mk_NBeaPn0v}mBX#YZde7fHMhul!x=C>#3cN|DR;kL&DtVXhrqEI|e)
z(wBSY({K6IqJ3w1i?BAWCoboCtFS12AYB&a0<>9EpUwOdUd-DR$5K1!rXf~C0(C(|
z;1$uo>_U2!A?s)iDEy)J$M*W_`?9&FnO^j<vJ_K1BGTHe*Munh0x^7uy}tYOH0}`3
z>AYB%KUG^uk!E_?J@QO;SJv+x99S<^2}((sKV5AD@4x=5ULCY>81w`ioiZ1K(U-N+
znQpsZ5!Julc8-2p<jzAyF9)rG90wkQk3%Ch;{M=+cg=g{yC)xq)FaR7p=rotSd6F0
zs6HxY$z}b%bQN;<_mU@5mTOWd#?p`5%VmO_5jj$T7ZpwkUU6S$-`0}4UWH|Nm}a6(
zdq916E^8%CQ38PhuPflFY@$LMbr?ikEa})0<PS{zQDdRPqyk&6Z4E+3Nn~gj-H+@@
z<J71X0t_&8A2x&tc?pA$<yVtUY@I)r!*0Xz5@0K7#IA4M*Hv{42!wm&;bi$pc~*{G
zy}!kn=Zy{h0(cO9z*XG#Sl_LMz{5bkQ@&3J7@%eOHD`ha?ph8iKt%>Q#3aL~H}RdT
zCVaBx(EO`dU`D{^3*bFtV7)ds@DB6oLiz0MtQ&C1EM4tvRXQlSXN+c=X6o*O=DbOv
zhbcIopZ0)jVnLr;tW{@YPVb<@*5wJ3ZCfxaV+%?K)<&tg40Z_VlO!4fnH^(HG0oRd
zxFnZ+$P*n%Jd&C82>~%Vm<A2jC4SW5SK}AzjMh!BD=MVgbDTYE9zAbr9@9^}t81X5
z%^3NaPC>{$Z&QQ@T3s70NhGwVwpvZh8n_swLXHipd0xMYCP3B5{VY|+SD(T}oesk<
zXs~MU3@>}NGkdmU`33`uW|txpkz~#uHXs!_tzjreCLivchnSZ=JB^CO2~-5yA=E|m
z;CO?bDrLEt9WIcj-*S-NAhV7|H>^gWntJPTX^6YJ7j~0s$<NOhzjZD;zXPYMW=lr1
z-u!;68{C~6cu^TxT9o0>1C7Vht*k;=o`pff>->3%)=l%Di#t>gYi|{J9}jebp;FOr
zSfGLOis;~IU{)dXD5I^3TZ@F9ygwZ>H~DI&hEs-izE*l@Vnu*3wdJ<)X1Xe-kM+o0
zL7p*p81VGC_x;T^knxx8HX39YEPha0HLU}6Oetm6IcSI|8iC^N#fV{xH#X>$ud)c@
zUfI!4(dYJb7GTSn-4JW{?W};(5@i(yZW9Ri4pWf`PLiRF;gaJm2mG68U~n-q@IOl4
z&z&dI6W&je($^*t^W)W=w8EH0$rY}%HkUk++Tqd#;eay<FvsQ}o&;Y1tBR~0FlBeO
z_W9kX2|{paf7(wA3bVu2R${R{#%){0R7va<-V;8IH=04=d5kQR;jg0uJQpqryoI5@
zhR<QEJDN!QpF8O`7M;nTUbo;jcSXx4vxf1XkEZ6KCu4=qH|(kSDlmkTWjk_B)ic71
z6HY;xI<x}U4HME06n6RCIJ!6tEWEfVxtyo(SzlV7c4vqxOB^ac?JuyXw2FIbdcsEF
zOe=u91*M6QI1}fGJ=b15N(TB9!#7ADOhvU&K>rq}nm)pJoSuJ5U)SP(p(vCH)`T!K
z``vZ4NP#@t{{ve3yA+qH3y7mT?xSVNN<4PU=ge<KKi6}P&Ry|IqOUKM5bO^=CD_cl
z5Wsdhl7>esK{X9WDZ3?NjppiZ;XlHZdnr?>U$fF+A#>MYDAV>?N}JkT9|PB$_^5eQ
z1U|*`#9;(tM(`BW_@Qx#o?}b-;voF~eU}9~UteF~84<lrb5m2AUrkp|3H|;3qN1X!
zZ(3rJxCx(((i$4LADuUyzJC%D5|Wdj`w`gL*~ga{3ntI3ER2%4=!Yl93)Tts2wgQ!
zv=_@ikLKp)7QGy&F1J)B%fWP7^s=_W3L)zVGEtF}bGItH%5BMC;C5rZ^!Q1^OC_hJ
z$`hIrpSYr(n%$az>iG{sWkJLw%m6Y6g!24o5f*tbff*~)H+JZD;cG2#_xE3YoVK95
zl4COX#++s8#@sNJ6qgxN%&-KoDtCxL-HY@YFMuxt&H2NC3kscmAAlYZUqLyHJqn`C
z8|j{-bkop;?nDJ7al^H-x{4|lbWm9b0>p<5A|mEVDx&r7>q>-VQww|LMl)jyt=vqW
zy*gecm@6DLhXb;Tk#^d9tok(`>!be=ZvV1P!R|^ZqDFTxK5R~(ftsl)r~yF7Zs0a#
zAH-J5E?1#R;Vk5%`yTn#yerQgL#xq43q2k1gpLzYzGf@$B{O^P^V;HdWeV_Kaa`Oj
zEl+$+ZLM8RW{r5vW9N>pp&^+nGVq9r?QDT;WhsyS-Vm@zsQs-~p#RGKN^{oVNDl3b
z$!A*Ko=fX(i^Y#}REZq-m+RwWq}kKSGL6qR)AUUamv4A)$s2!?`ot<D|Frx;_qGhU
zhcVue%cpWw8Lvsk)}io?RV1a@tIw&b%Vt^Lh0*3k@Ib)<<zNYfqAH@$m3lccx*p{d
z^D-AoD-Nkre=CBNa@W+v5K14n*X>bazSfIRePB70+(0w18eG^z`}t(O5$e`O6i>6Y
zG}weZVX+vfCGRljr_>ZTO^}P5_kH1Ltv<t|@zwGh^&wfSWV}eIJ#B0z@xGJL?m9xN
zR!FnJsa9sYr}NM!%F9X7I_fulOiQtgFA-i-);F)47)~eMVP$FYPS834gF!=gY{5i!
z?S6Gi?Mex}8cy^5d)S5$LP=@q?QISZ^r^_tigs1YO*a|)tC5i)hEoC#HMwcHxtdm+
zLD8>&AtRal>|x|*6x+#lvz>(zS8wQU0)#+>7zOMAz;qQZ8JS1sc{459X5=~C%}Op7
zU(Vb9P*}@9jr%9F&HIZFF*7xGZy!<aJQZs)Xz8(VFpIL4hKDn{#}q4DkOrdYz96PQ
zWy3duAKivf8VCWVMl+Y+nHOZTuDbHU>9*;}z(c)i(~W0I<YxTQ!|_N_7f}K9KSDQg
zVKNgv1J9<S2WIE*xc4{O;MxP^9c&PmGeYu0d!&7856~?8qPPgSPn`zjz&z(xn;OOQ
z1_!rv6}H-&x}}+raDE?L3;pY-c{T3$C&1#goibq|A&q)-3?iaQhnT~xEq+lUshOP3
zFNc;MpH4`ZV}gw^>^vvK&+7n2LJ2zLFR~-s8v6B(qGNgbRb1!2-SS@~(+vFhM~e)T
zAWSb%!q-XH@*VSyxyFAT)x@-Y7XYo=njH1EzTFtedsh1LMN!v~6njMD>wCtFugLmc
zD`?UYd^-c2w6J6RJtYOh%BCInu!mn@R$4Otu2SVfO{ijAcA~H^;anbT{!dU+!yECU
z#;jP~!xrXK7{8(*#0Dd_4+L#y0^=_vj0b6d(ttAC<@hLm5#Q)%FCg&Snf-{af)>f@
zDfciMLeLr%j@`o=ijkhb6G1sd2Q^tVnJC6zc^^YvqoLaVCNRmbqr+PnXc88Zl9I`M
zUM)pMMF9Z;xw!_stZqoZ!|};*D4F?J=aLP#y!*Sm0t6xj5QVnOu^NYtbyME!cz=6T
z%~FkpmS7}d`RnTv5#DRvw>HdMdK@q3*NoJJm4sPbFRESmp!?H831j+N2w)cp8go+?
zcUHXh3!yN{p10sOfj?yP*)FS;(mL;0d53@e0+oOgc@SMeq=ZNSF>Qr*-uo$$>CAj4
zCXt_z-^G}Vmhjis#3Xk@zYcs1+7{2`o3sKnksh)efnQKC9Z?Q?7xtH!NVM+b;QT;A
zOKbkl_6MUp*qPPPF@Ae%$qQ(~ucLx*+C9<vd;1>WXKFh}{`1$n6kB$YMjdW74s?gr
zmF+3N>d0&=zK}f#*FR6-_kp)X8slZ5RAq@$r1G!}LJL}@!*F-2q?+?%z$v^IG&DtV
z6oINq;7Lt7r?fwx9jKg?OT=zFb$xcBw^E(gt)ONMx%pxi$sgq2jNEikc$qO_hJbnz
z*XwH!2>6AXg;g_}H^CcpDd8yLIOE23r~R2yY|JN&B`nh*YvjwpgK}VcBa6clE(LFj
zNDouR02P#wNIB0lV&*VBZ7#3Fj6W7FhQ$tcuyZ}ivQ-8J4{Cvs6l@d=i!_uqvU62Y
zz?8>E!6UXsWVJS}LdgRV{HfFOHr_<HXC-xY4Sdb`FKw^(&judo=N@1g;*PHG<GdQ@
zvy4$Vvc-;DW}7thJMzvm@;>A?c!$KXRAemy^n$>2RrK{O<6$ZlJ2(Nz`18aq_mLw=
zzt49H%w$Ca%Q2@E>1)FR&L5?Av!$4!<#QUmQXX{|V_f#Xq2Z{cJyh%#Bu-<ebmX&I
z1Y6wl-RuwSEpbT~MEzQ?f)<>2AZhGR5o+-lHrx0m`P&laVj+~X``O&A{CpB-;(EIF
zWRf|R!*$^)xh?M)E(D{(k}`B=5atYD1K}tp-!}%@SIrtuji<^h6*kAUQ@f8qF~MO1
z6_OQzp%OhFqY8_G@7u;@^5&OCa&*Bc(XV8;ubzd?$2d5_2)n_w^hijbEuYFj#Rxpy
zV(;l5Y4f||UAgX6Mc@4W)Q#s&gan2|RtBRdFEgDyJxGKPO+kU53IgeJw=tbqLK|v<
z*Upxxk~kAV!b$ANdK3oD*90<6mPg#Dv1w}m1>}AHQ1z$-sD*!*Ql*59yNxSLnDMQ+
z8xghQ%H{8azRUMh{DF~1G)MySemr-$M>2k>Mfz20T5e7Awv^q2IML9WNCg4S>xiue
zOFpjddM_-mW&<t}9=H#EUr}n8z!MhLJDF3!vg`^tc&O+@)v+|vdopS;|6u&_&0<9R
z0{ux3OvPf9@pB1lJMEB%OdHNTsdn0ziG;rivu59BkP{(m?yhf}t}T?<2`a{hKTrPe
zOEh+M=ip`A2`3r7Wk1HYW#IL6Q&+1JTS2_hlgy9~EJc7q4-yirs?^eET<v8&-tEsa
z%T4&;=30t@lIV(jTmNBIztSLY1&X?;Ug-dTk7-y1u9D>?rWP5qkYxNeQksD48qoPl
zcVRHIXkLfoZOMZTiU9B?+8)}dlzqNg*;P3VdtU>^a~VvZT95WfVvxA|S^7Xo!>$!~
z9}6iXHVkSFvX3>2>X+=DrMb(U9q<py_9#4U9^VCSeL)zPp~cmrkM!oi^Ma;?m5#_c
zs5t70F?Itm39WhGEa%iI^}i(88{0p@Cl8Bt4RnqnWfD-YtNy4<UDC+Cr$6-8Q!uq%
z1rZm_`ON^MqP)N4+BY&L=W_KTmf`!~TXTA#t1M}DpcmI97!F^#_Uziu65v0rKE1UF
z_lObWT6ZXUfJTfw>Ddn=>KSmz?-;p)$ch<QvZ#y`mu9b&X-Wig7d*=!V{cip;8}gD
zPo+hXz6lqN;T4hW={dTd7<Y^~$OKD_U${$u#d7>ac0r?`-p@Ezovcxd#6?gOR<mNp
zFoB8<TNKPoP{%#ym<Btb_l+fKpLR6GOlBtId*o&Q8u(3O3!P{uioXCG*|#Z#AKL;m
z(J^&O=Y>qqZWw{YKa}9iDHwc>DF=^|rVY^oKp>8i=e~f#*>T!^7APsJlQSpCP9y{f
zRo(XrRgt7u6MUOk(ZEc7vVI0I^9}wU=a9oy?T)V5UM*G}dx7;^DGBvquJc@%n*tpi
zsGyG}aDeMi&6hX_&3-uBJqUjKqx=e6M1M_wu&A8RVCob$Ea$H!L{~coyrqHi#vl-1
z7ZeJ2k8*R(oE?L95ln?q3<2;&JFoGD1docnhYkB~@?2E~q=Djqyo2WV@CBI-WLG@n
zUh=R)JaZlA#SzCDsN?Uu7t|E%;#I{ktPzg6(#8U(3>L)J4F_IYLr4f<fFiCfFbW|@
z+B|K<!no3KaVIEm2_1>}(nJFxc!l~BRfDiA(sZf~F#<HZEI3f3phXR<8Ms-#p`;KU
zDLelphK60HgJ8ISL;SzF;rB9rS=m^%eRx80ro|_faU-rZ5gGafC!@HcxP9^}(-ex$
z*Nd0bd+5cQuv)fEqqi+2Ek2O}jw7b5>^SX0U@i6)7X22~N@P#u1KY9I?pD?eIz2#b
zIqG>#CW5NhD-mw_y+0tIRr9lu|CEx0qZRP;?F{Joo5Z-YvNL7(B%!GK=g9f%K@ZF+
zs;TqYbafrJQ^sf;Kup8<?QrePyv($Ey4Y!bZ#i<=7OJYtLAShj@A>5!9E>Dm$wIDY
z25z=R*t@HZaw>0u=~s6M?4XtpBq<?5DYUGArF!h0{ZOyGX}!F}R?{qR;E17+P)Nq(
zi<goV9Ud0{hR0+UV>iTLeN>Z^1%3A8zaQiV|IMCL4Kp*1<1r~OzjY*ThF><io>NUh
zxtqf8a}iNqWeaT=Jhs0qe=b4exWhjAm5JfUa|#pZ<~_p&W3L|6X*@J66p~+p>zRK>
za3#LBtNqtJqBxwY6$uxBgkddQ6fhZg$@naBufX82$RFTP@-Z^C?HKiZtB0)zos+jJ
z+Ls?e^o@+!i#aG6r|JaH$L;${wh7Gluz@0C6fqi38qWnIGVUSDgr#Y{Y3j)C8WAni
z4y2D8##pT>!GCAZC{yycIM?hsADEy9(wjv)V?3Vpy0}2V{GP0cExt=D<r<>Jv&9cY
zLoBEqp&yW;wBw_sx#(-t=W5gFopvY?^^-;Dq_)U5Gd=3m4*okM|L)0CpgqH7o=w@A
zAHMXW-0_uis=v4PheOHxTu+VA#gG2dFn|Y>I+mby(t@iK43s>Q2mpXgtq+qw4fM1j
z1)eoz31SUQrL)F4Ogft7HxOYZf`Tszlts9?*@XDu)6Op!0cHN!X^x(MHkCRP&1u;^
zte|{|K|rG>UeYfq*3Zr6(6aP?G1RSqgp-()EDbPp?Lbbz+d8}qJSjGWK+?gRK2wTG
z*H$V;$z|KJPg!u7C-%~e=*TslS<R(#^wVN~#WX~7L~tPM0b9G*k(t72-=fjp>*!jA
zE=KO+K3hI*ObR4khG=dO%0lr14U9r2kc=5=9omH(G#>2GhE_33&FowMb7@x{g%^5A
z{NLNt$%3@==($2a@?@p+h44L8u7M4D7H;P1D$~`Hq%}a8Sw2oESAboRdK!W%ya%AI
zu3a+^7$VcnAS6erM?ta9E{oRfau>y_VDCVWhZur``E@~SinLV0DXB)50pX+J(Na=d
zE4&p7J&K<#!>>y~E*Ubl->_fch4ECV6pd?(RKaQbuE(+H+rxxNC&1M)Dpfk8q1k^K
zI!2xX2-M-zSgazEL46(o+|KA`DOOJp<{@JR(m+geY=AA>0CbBSz5Q2H0`o{c5Vx}5
zWSZMR2N_*U%*V_VvH=TxdKH6Npb1TWcTl&~U*|9S$07fxV21DL+S$6kg(Mt%)P2Tz
z+B&U3H&YsZ4H;zR){~%&XN$p{M>#K=RX8j_67~72{5L+cZpDCApmxLnBjgHz_Ynm0
z^~W%&sC%<mDY#*9@y1}Z;TECvE}C>Gg8kehcp>PEdOIx<yj++DM>Y#YGsw1rg)I40
zCKQ8OZ2mc(TFKjG&%_i6#e$7MnWz}h((-UG4+(>bjM?eb`-v`~`AZZ3)kSMbloH~;
zA#?Kr_4sSSSE%g)@?D^QPZd)0s@S)JS1z;GS;GE+6HCn^)P@j5e}r*qgD?jRgla@s
zYSbkch}OX(`*!-l&g0G7km;G2cwQQ|2$i-C^otna=kKNu^jFR|N*=+q1Lv<W>EpIV
z>8Bcl%{vA31pBe4D8@bwhxr3}$u1IQ*R<!qe$k+Yx|pzajN*ea-?V`Jwv*a?;46tO
zw2;jdM42+p#|nti?nQa$e5=)&GXg&9!_UL=&NS$7qi}`znKP4ho3lxBRIFdRAS@-1
zr0$q5l;iBQ^Qo5M)D3=3g%cCo-xtqQJyT_cR^?nB4;Jbuuf7T}rVxwjl(SkSMmSHM
z2t4)boD*B=TK81L0-YA34{~b_EB#l10hPBkz}?lY0-$oafBB5u9?2!h@7DW*-uC#S
zaLp!vI*du5Y@d~77<C!qI#;Mm<!Lfc@VDN}Wy&?FD#m7HQ76^^ubLzrYM2C%`X>dz
zi|f=gAEVeq_qHg%KwgQ9hg4vuA)<zW+r(B>R06_ur9vtX-bPHlx)T@PO(C_8L=~?X
zBQ#6_FpMCxD(2+{+%h)#8)3s7I_J_M=~;M$uYd>&(tQ!ME|?iUNk@3Y4-o<8H>tzd
zd3W{&shi>NQ>M50K*3$0KMhp%sIgDaXmt{7CAPVk-oRx?C;0OC&lEk^E^0m|c&5~2
z4zzVWWV9VI!J<-aRnvJkTq|NNU;eu2YXN>C*WWgeUV`eNj&wz`Kdud@S9cqOE22+H
zRqNC}QsV1jD>&<efc;m!Kxzth1_Jm2K9gBs+|c>MBuRACuMUF0{{OD}fp3`DcbA@I
zu>X*mmZcp-$gMMIkAnP=a6!FX`Ii{%8H!Anhwa+9NA`$sh?jU6(BcG9oVw`4I1q~h
z^8+Ir38LRMC#=8jL&8+p>eyY)2%=<{tE^VdRaqUDIDDu2K-+Wp%zrmE%zEznnkY<J
zy~Pr22GafWwHU>v=ogxm=5}V8A2!BM?0i>U`Jzpyj@vh#iO-qwtX19cZXe=wFx~Mz
zuAdmOtO<}R0>Qw1Y@86)<MPM-55hf5zjb>Pq^wk-q$DgTsULC0mkN+PI*)ZV^bS-%
zM>ojCaTW-_y5R7hls?IKEX=1h)1&qN|J(`yQe-khQ!{B7lyP3^56E3+a$TiMo7z(n
zezek`Q6MEMCpL>W?RHzB>URsFfZTW=L__n{;mIyo-RY;n&)F0y2|=5ZcL}Y?w*)I7
z*W;O6$W>w}V?qA$LDn@&oTce1%~cun!g{P<FeQjbfdI=Pa1;TZQhS86bez3}s>*4_
zETJ`3A5~n;|HAcUZ{25s>(u3d&<_-QlM#ybCHX#XE&&1*P`?=2B1~?shXaHJY~Pdb
zP4^U5NSU^Mk;aaFO6CQZ7^;k{N3Wa%?n6Ef-a(u67JC2kv`O_G8cZw(5o6fK3D!!*
ztyrT|$`{Z;6hnN2%&kdF{>A%DiciRO$8<v8HDd4>y*;NL6xjK<hS6wUaE;C8nmYXd
zdIrBL5|bT26b-OVAg75avw+9zndK-@7iqy~2?Ur8`<NYkmO)U01GlU@*HK*J&MAm)
z_yZaN$enEpQ1_YE`~>VEgezktDEk!gXnMY*U`1?o$cU8r-|o-$YXQU~80MS;EKIo^
zi6nO->qeSmX|o!lq$j6%r7hRc`zCSya(bhKJ^66E+KTJkuZa>T6TGKHpH2O>gh~Tk
zFkj4D>lyyL4d6h?3d7&{Y?dxn;26?JsiP$xGOiZq#Bhe|g(?BDFjce66@orTqMg?i
zI-%oghM~<RWkjql7>TgxbohG<aDp79)~`VOq8d3OO8O}BD&{5OU9t?dY>sm@xm_Nw
zPnxxr{vlajB&e)hT<{>n0r~B+F~e%>|ElX`wEx-J!Yh3!=pt@Y9EOX)*mywdSw>k7
z&6X~}>s1*y^jcC-`P-viMe%_xw%|EK(@YslD8i@-tD7EX{5}100VV(vI7?7qz@L2u
za9Vz;Q}hA@{&Ua(<^&i*QwW~G98f#`8vPyn9kP2e$=PY1!{k8|Vut17A7{x9`cY}k
zYa*?oGEZ3KVo*}YQmB6F`w|qRK21KrWL|9&Jc;y>XX{r-j3#rOvzZUvOF$;IgNx4x
zBj@AI@T(#}AkT>83TMpi-d5@Tsya-uBw4A1H_paH!6d?!wvJr*JL}Y^H#_D0V%c{T
zpq!Ho=pC#AiGI%zpW94^n9y%*+y4Kmy2c@4)}6du{)mJLH{~Nzhj2rbkLCu)-8<8x
z?Sd`>NBHWxV!++X;Q^diyM|qRY5Vz#()wvr=25$aeiN*2tgNb>68(Z*89h-w32aJk
z7;=ik*c_CT7;pk&n5)=bJf!WMsfP|);-}f=OxsK!sUq+h!KQ57E{I66>jmr=HDn?S
z?!@j$DdrRk@)^N!tE)8-PRUX^1@Dl36t-F(Tl`p5P})!^yAE-%C=p|kn!~1tISqWL
zo9W?ulLg@)QFS=WlXXM_(dPka5D2h+mq3R^k)f98{Q<B>Zy*A-t6mMWJ%8V3-mV*V
zJC&ZIR1UTRM};Uo|2#wVi5LmQBo?MXq|S+bt*}kccHJmpcy`#_!|DE4ZvavIw;=wQ
z!*wVyyExp<1&4hxKr;vTtfC}?sVOP^jh+$CH~3F(ef4~dEsv4sSkh-8d<FQaIsqf!
zQ9q#3fRGMRoBlthzA_-H?R%SUhLny$x<MKw1qr3QOHx`&Iz>WCN*Y1Bq(Qo+r9-;A
zyWXw$ckloG4l{Gk*?aA^o;u%O_I<iZI)h!ZEL6S!APDitU@2jk->S#a(^_c=YLbOP
z9FODtOjREu9`-hLOl36vr7-Hdwf@$ebWN-FPuH@qUKc!*|5T%9vGCyWFcp}H1vj=u
zD2iG~pYQvMQZJk370xyeln$|U+id)mgzn09|E4RItDljFu?N^e`P-KRSPsa!%Erm}
z+(+Ry0VhKz)_mg*739u>OHY?7iUq5~Du`VtcVA-%StpFd01}y!-hFu+kp!k0=WOe`
z^E9nxrZ2I+Ko=jEu9z9sv&;eQzn1;WUs~<lg7*33AEtX6r-H!-icqlEP>w#l@4o?m
zDt9^^%z1r%U1hW3wz$|LdeM}WI}J800|RrW-*jtR46`&-5ffE~0}JcPS?GoFTqBNe
z$8IR05*$ijav!gzl|qfo2<#{-nzGu9Ds!v+QMan%xakl~Wfx@vFX$djIBXSWr&y*q
z$5;7Tpn&uSfQmk5!)P=&JkKvW_Z|;ObD=_zBzcy1$J__phJk0KXe5CLg4-@0aaSLM
zp(%aX4NrEkMB%+myjmpzpFMkc8uwlZNPfCQTZkwW)(rZDYlP$)zPEcLvih}QVT|Wl
zE+1;3u_JY1>ZM*MI;zWV-#(nJ{8v@{0%<Rwu1aeNc|GCfmcvNjL9Gu6mjR4jU8S!L
zCl;rDar<jEorUTg;V0@(;7-34E3{Q6+XRe$7<JzR&4QU`NbI7d3~psRrIFyn5x;!&
z4#GBeAd)sl`*O^em^brW^`;VoaK8A6HwvTVRY*)(;tV`h*`zt>40}K7qErN;&x*v;
zB`~iL5TIGu`|v8ZRkHx}_(@}4m=Ja~_p)-3qsTauHKsIMp@);#^yO^Cn&7X`r-(l~
z<778>Yooll$4iTT{^)g^*cAk-9+9ut#%z6AG4JG@5D@4-9~wv@j8=bP`aEt0u!BjB
zUMZ@ea<{n4TQTgG#L-JjE&*!0AfF^gl>bVLy?~awROsiE;j5MjM;K|W4vJ%eV?C51
zVNSQ4^WIra3b|+uSC(6fT#-&wLongi&Mq^Dz%~nox=nA703;ld^(UJf0`iyV^?5#i
zXn{tZ>7Y6L3g=LBF`#3<qoSN&W@Sic*!Xp1d{wm#>Wpv&WD=1`_RV#SrL>T0L_}oe
z(CpsnS}+wL!B_c50I1#6LTKNW<wPqGdKsM7a&t$LLe6Z#7GnCdyl#CjfwuHjCNixH
z?XPhle&U32v~^5H=>>vc?xAqfdyFU6Ui(F#qFTRhnkqKLwlcVTT>PAAX#Or}XTyy?
zyMpgvCZ^jU2geX06D5`c1JZ)u@2hyx7jImu_$-Gs`2U9niLMz*S(}fJ>dT%8El@e)
zQZ$Auu<e+S?7zP_)toq(Ku;wcx#^z$Bea3lT|U3s%}gIvr}glpS(15Goe&rD5h`<A
z-*4X9_T+&1$lZb*Au}quh|>;HH0}Q9Q4Lj!1f1A=?K|8TxD(ij<CAFy$YOywfa`NZ
zb))+BA&Mw`VNozrBSdaTCU$i}bX@gHny6y>DY+Z%g&f#D?sMXKFJ5b4fY;*YP1Q`<
zVI2pj@#FBq@GE^xnCZt&BWBHF_Ak|@_m+{+d+Fs22)NGqX;ag0m^t-S#g6=!A;_0M
zTEDE!N{9#cRwdQ&qB*WRHka_|*q_X*MTHp6izbGJeHr%;Bp@_+XnpMZuaX!#mLK({
zO+TO;?o2nG%83&E0jBj(R{AO(S?YP)`iI$d7pqJ0f)U#Mru{_OA5mRV-c-O7e$atP
zRW4<!CgGvCfx{cfr-ES+^V`4Kz0xh3nnr-l^(TO*>YKH!0)41$*+g@oEAmnSP`3c>
z1=Y+)Zz9y4Gmahq7aZ<rIg<*N<VUc{OVqB%^qNSx2!J^wP{ID%SPMWRVAL>GB!-#Y
ztFEtAws&y2`+od8&eukO)FmM%L$2NP{z8SrHd8}J76oR0r|C)l&fDK^ap(C0^c&?2
zLGB1?j>2z!(Zdzl4!$m$#osbzX6V1ZEfmb?0@JW_y;cfTNPN}-|C!_H%QABnWOHA4
z^W(pBaP;O<-X?UHLJJ@JuDMAwZpbZNapvPuU2|HmSJ$%DvisT&;x?xbUZ4D>{_l`9
zFj=;y0orZw#X!9eUbGqv>`q|l*_JsXeyvjP*6gtqV+A+G_Osk)2yKH4DZd!GhPBBr
zZRYB2Gh=^}cRRvNe+^FKBC`-9=25XkSXpAtgzwyIsS~B+G6(XPBb3p2R1jCIzfb%N
zb^|_BO#Fq|XN=(N$Vd|xPBbI#-_}>|bH~$xlShSVnXc)l&(8wd{IrSkO#Gg{WyV*<
z>o9u_sCY9j>!fuKiL7ICQ$22fj`Xw7@LGhuUY*lu>QZ&}R0ss!++@M(X6rnTqm5-g
zIiyMdb+w$x{di-zoA=38&Qr*tN%}x|IA_JHh9PD|-e=i#`fJlry{xq}P;<Xw)(O02
zgJuntFBy%`WB^nY|0kJBtr}Vu#sVl03$k}TJ&&@Jc-@tX;wX~X$~ZHQKFHtfdWfLE
z)0@2-20cmrq?%E&Lzh;U{W_>iUIs)!>G&<5SR>*Cum(P{QgOiCK^s&2BoKQfu!`>U
zGx*T}l}XrT$Bk706~KsxJ$Gear)bzJT#RowR9&ErlEG54557fnLglWv4WPt$9TsN&
z?RxjxFccy{D-a|R=Aud(pazY8oA(&V1Y}a?>hC}wcXi!Xf*nIal|?Q7GuY7ct(1r-
zr<v(V;r%)zG36oB#=i%S;?Fj?4ASrSYnAe8q8~k>Z}B4fL$gs;bS5PlPd^<avWR`1
z_xbTQ1fM$QB<0Yao4ijUdSzSwYlROaJ?wdxAVf^`nQG>a?M6ltRwT^FOpIj2J#acE
zuk(jXM7Z+pD1;&&7t|&2BZjp}yZ`+1bNL&f(Vp<8&8t-ksx1pSb;6eb@Vv`vvU5<w
z8skamOBCHt_@w^lFqn=Zqcfxn&-UJqCXNybga~DmxF#c}hzA2f6fD`>#c`y_U*~bF
zs-4wc-!bp5C2yCg&ma@8oniPqi7r~>{YAVp?86v-N2e8IcolfFXt;`}?*>PC!vjs`
z-)mpWTZX3FyuN77>JJO-5PjY9Xh764i-sDU9wkiNT~M2IlrCnlPk@w!<(C9Ux5cU?
zgU<fgc$c@-qHcB#6kSR&zuzl2L4{klG9SeOm~0fIf)FVn_~D7;E&Kj3Lmv#<Ogt;P
zA{)kQ6ns^}Y)1LU+*!pgl(y`8g?Dprt3k&~aDtgMhBCfoIvVnP1qkB|)D;Kp&VDl#
zE1CM<(YM5^Q1AYj>?jEdFARqZo?0nee#+vg|E$LfLeiMwN|{`ms9Xm0YOqRZJ@S*I
zq{n<K?^cv4n*!&fxF}5NCE325euTWi+q0T}S2egL_|?NFM+ohi!|VotJ{FD7@&AJX
z(#hUPd>ecaFuTukGuXsVCPOQA9FzqG$(3)fdDNSB{nqo`?PWU7B(<~s&Pp3ZB@;ri
zI>K+1AQlJ|(y)ABgsDU)`4yc1fq^L`Z$jJ}?St#G<EQ6si9g0H=u7Ys>2l7JrqdOS
z752Gj{o8Hul>&?i1KHL_BPr}#>M(=5Jof#&N^weQh@^F*5Q1UHssgxI_Ik6av&gcB
z@>NqG01;PJhyTLN`77WXiaw<AJA7D;$i9q}Ab&+F_u>XNt_}h^(E+FAV8>Ww+M$5w
zA_nTWU8YzL+CoFTTVlt`<dlM=2-^-t%W&I!qV5GuWxc#kChcZ(p57JzUwPm~K2^Ns
z${Mv%Y*A!ae3s$ISoB;&U2WoyIwUXT6{*aNbNocYcTy{#Bj}-OuO<?fAZ8`Bx>+}v
zF8;hf2QekE7$mC!bPR)HHxfU~V3Rfb6w2!dML7tf3EuiWPBkpVpql)7&uqG=8|$mi
zlgRJH@SS80ULR%aOP|73Z~$Ag-iV{o(-9;_>GPC)ZYq?Q=8_H4b=lb^3%XfdTLyBh
zGKgECF_B#$mM$f4d%(2dU1|l4ENC{e*nTz3mZSQi)CBn9N8z7iT~Wdy|2qG3G=TZP
z{(YhrJ_Wi+w?5G+s*=_X!$#NIr=LZaiS&8CegEI}Qy>?~^^acec1IcL1ElzY+)+6R
zh<y9=6JZc!bH#ZBlz!a1s+yoYLvN6(^GAVW==>h75WF7TiVT@qs}8;YxN9@Ssw-TJ
z#*8xToFt1~tQ=f8_@z2=hxHeE9TLk|p)DSr0n}u&hR&QfKpI0|%<@__85|{PE<#u}
zL>Q={xeoY+%c`zH^z90nJev*{&f*15u;Kz-^|C$m{%%W}irMK~<6AX2DAR+Pq$I~5
z&-{>5O+*oczSY@giozSJ<M?3(@u2hmAaKL<Z~$i#lwAalp)$-`gE?lEs`Th|Z&PYr
z3cH>N{_?jKapAu392Psg_S?+x;k=grYptjq%kQMDSBgsp^ehz~LUzKHP~_D_5&W^9
z-wjJ;P#dCV9szBRaxXhC2~lL)K6eUgx;b=Sr;;bXt$_le38h`Z|8?2cP3B|UUaE~F
zt0Nd`0Z6oUK!2Fg%8n6AU>l&98Sx?sKIe0E0JiLN6t+8Ihwawk1<|oP`pRdj;=L@L
zf230=KNW7$7}82z@_G2~7=>6vZ9jH`Gm~ywA!(f-J!Da#0dswBDu*4?$+)C<akGz@
z2zW@mu{RM>Ti>f+?j~RH(s_TOX_+Pe8I6WtqxANda$fewWE56RGjt^|^R?szT=R|@
zPEz$Duw8}J`-+tC{DF-B@@=A+$mm}7OPk&`F=NPw2XcGyH6*sM)w9EI@;SCLN8Zrw
z(U5xJW4U6(+S5~lH+C>d&!S;}dBZ0=t28S#`=h{VY;vb00~~|^Pyd+frB8hRb3I!3
zDc4h~&NuRJ+Ga$$y<wioaM8(@7WM)_Ft;st?fFMRj1Prh@;&6A?8JMG9^+)b?8BLZ
zLh*^Ysjtfe)2r&bc-rIXc!9ZqSQ|x!jm8V%_3Be2j;-*7o249r29{1nF;*4B)>89K
zIu8Wq&9o%xw_yNg030ZM-+929Wx!d>*A8D3%zsKhz{aBn#m)7%47-mz5)yxh-z|Cn
za^AXhU3|4I>u4KXOi{ew`$2*!%Bc*j%+$C<?#1^aC~%e%!ux0Qs9s_)3v!K8+hH}S
z^C~?reiT{{Zu1zSF(qmyxJ@3ueyeZJKji0Efm&^x;H|N#^OUfKS~JpbTw6+8X!!8W
z#`7XZR2lIMiJM<c8oVEPIgqnvJIZ_D5)q-~v&Q%M5q6}oiT8uIz;|$XJl7$CJ&|bv
zW?-abE@0iL{0c$=fm(RPgGTCC`1_C7-!>Bg={U1o34iX0U&T}|GZYmkl&yg8V<Q+B
zBt4}CFPdxr_py2{3tWW?^j`H2&m>e_EcJZ!mBJ=-b0)ibT5hvpccKqVH+kR5JV@#p
z*M<<MW@2Zed0)ypt`n3O{6dx!)*vp~&%|5)=E3P85lj`qoRBPt-%zsZ+c%%gtN>Yl
z+cRDM4fKw^C5n<#Z>k0@kIVoBW+~{^N3>ccB=%F{vk6SZsga8ex7vcM9VrZ}DbqXF
z0}>Gu6HdwF<gZB|h?Jyw#T%2xLkKU$twcPpBZ8rfD;0gu-V<JLSP%ZJs{Q#EPoeMm
zG?`zv8J8H<xSgJYp0YMhTxP3I;!=)LDq*nrEw^(%6KlUpW#m46hTG`nySo3|)Gr*%
zw`S+-V8fFJmg3_n8wmrT6njQ(F@Dr#{Qc%bFVhe$y-aotOkz}HJ8r~b$iR^X$euu!
zWdCHpain2PvEhljq7jo%@g#vFO`kmI*XdCobk!3coiFk{881&xPX`Z^xpMj9>Bjc*
z#Cw*b*^u;|1D(VB!tc^=c+{<DwZ1tLHnDx(`I?ZG@Oh8!oO2tVL%gF#%~s%GnqMwq
zydxj^<IG(7<@MHSzHGRzq?#1;z;d~d>hVR%gzjn_)nX<BmYyKYdfma9pjXvz+iQ^i
z&>JR=C`XX&n6S#?bO57hsE(WnKPDe2j~c9Yx>6ZlP3ZN3Cb%3&Rg{-J>;(lapiA#M
zC^OuZ8UF0@Kg+>UKuZhZt_xFJ$Ln@kdIxd%0!)L@-*9+vun@T*{4Y2H?H&k6u=z~~
zV%TAh^osS0F*3JAm!Sg@LEP|yJu{J{Hk|IieoFbF#4R~pu78Rr4HR)vC{+o49Z6H`
z*1F!qB^nS#I%H!Bd<Z#nyS26T*6)B$LBp-&h_Q16BZL^_TrZJNe?r?rxqoJbTm;^~
zJq2B117f<Y9t7Pno4}UrsL0&NQM6H|FSuj6!8*%izi@U3dxD|v$Vvh2fD7^v_rQM!
z5haa1iIt)LA>y>2`{k-^OG#a?9YKYte2XgvNJ#$247d3#(|>E$Di?l#ogOx@r#MC(
zF0n|octdp+#`R5Lc^W~CLxhh47dINt5!wt;4VqQ*S=bU+K+x5+3o0iWB(a&YjJ>WV
zCTciZlaST_2{6_y+7t%9w&kZxulZFAWf}G|$6#$EtI5&v!OW}3Z{Tb0tvgur+F9va
zx2VTMZzrd!UBI!JDNw$Ati=b6a3m*@JhMSE6JOVy=0rN-G~?13CvYHy_Ru+*2Eqpb
z1@Xbr%<=U|)|Q9m-O&@PIdx{k4Ieu`j(;1sW<rF+pe~43T&~eg`nxQb?!7|#=qWJl
zV;jn=B|Cg3Deu}1O$1IcHZ<fUESNoBcRLt_AH>d{7bn^VRtpR%%J|8}B^tWiU9)IZ
zq@Kj*pu+(6lRAM}5ALg5j`OUz(s4lWMi{0$eg0%X3d<7a(_L6Uk*C4+R#H8s?@QmH
zYcgvtBb_M|*Pt<pJKw0sT5u9N5h{_i0HX)#<wfXm7OILE0YeS~rf9}3u?q(WZ?2nd
z&r9AP2(<A=!p!x_$D%ne%nV>Z^EoW7fFUGfQ+VKCj^<F$X}`(jcBucj;l$v2t4On6
zjQRUEs@AADpOiTYo1px_t5i>u($yEzbV^B1tA8|)5jixP3sMyEj)Uj++jGL@@FJ0-
z{6iG$x8sC@kaGAIK)!QK0hm|ms{h(hl1EZ#+x4j~RhAN;Od>ApcUmZ~5wHSP?QsYE
z%083Mq*p>@Ow>-=e*B&BDj&39jLednO{FwjY$7eaYx*j<-AQR3Y2wyrLpdd7t{gd4
zhXih{ep1+&AxqhSBK_6;%QpMPLkLku+m!U9xiZlg;%jl}Ro0f$mM^ni?<ml{WiDHI
zlKwZ0{J867XOO#>UG5r}H#`1$_&G|Xh+)?^@ln%2+h^nvdYj+F<+Z<ET{5m4gQJwW
zG^RY6?H9G^<(u8hG)?r*c|fT@i#VA0wIok<0@z_{r3K{-IZ9viG1?G`<`wIs(g3t$
z^`RreuMS2GJLHUSR}=uv+_RvyE-NI*gTV*FT>pjB(^F;urEXnZ?fbue5g4aKo(<=1
z`5{3?Goprg{$)01w0u2%CYjLAvjYqOI0C(E1E0(HPXvOyE_AT&!1AOM81^Qhl>vP$
zf&nUX-;5K??<T+{0DaPtnR@cbO3qZ_q*BY+6Q!r_eEMMF!}Z_Q&m~kOT?HM#9rK+k
zW-?B%J?s#d;o2%#x9iu&Uy!Ne9CA%9|5`4V|JgTAAH4ErQmv*;)=)-fO9%#05CY9K
zdbT@M{I+zK%q0?m3|20#lo2+7rqhG?LZT8Mx1_un8@p2)%rmnKmB9^r7L!;E5q)9F
zZYC5*oq}|YH*>uxIC9r?Re@D4s&^2lL&t7Y45N-Om{oAdI!`cm?56brt_U&y+Q)#J
z^$c%KHB=U3`KZ+x$beqIA&oZ&x*@QsI(`-|DoGy>P9z_;)KQ9cCe^yzw^<&q<o;~s
zcl4fq^8=2SiVrpj(VBi*A)hPiC@<$l^AE$~RId%3vBH(u;YXkUCh}S#cA%Dr^1<}p
zI$+bX(U*o8M|~@O)z8!q^DLPO!Fi`RH8ukd)6AhC<_5iL$M(0U-!)7!Gz~bB2HTL&
zJ6ZIUpIrYsmF^WjVL(nFA^c&}ec(?Hbzd5Vg%X8eqqS+Ebhys$x3Wy)z=!M|&4`~N
zAX~)`0a>i&;##DvmUR-)DFhd1-E0T*P{0DXwSPWwmHML+27K=eLm&>WP*!vl1m?R>
z<rLz0SxRMbRpg27LVdbmALTstya85|g40|R+W<yHMm=4s(pUi0_@^FP!s_G46MAC^
z#?j~H6^q6l-G7K3>1AI}`0u)r5(-by%;#^TzrjV-U9LgH+kxXZ5Fgu3=^b2+gx&kt
z>mXhgbQRijnz^^Nmw^8VwWpA-PaYc!U**KjNd<d~005N4Z#Yq;mp%QNf(XVi#u3J0
zz&Keoy*Ca2fb1Z{e6H~>Z9^g(wk}=qc^kDD0JsAzd0A5&E%?3GkBW9U>@3Ns6M3Uw
z-r~dN0!nIr+a=DNyx89y0h!ZJl<giCoBf?G67T%0`kqKUEy>l)q<wGQ$zSYSQX@cg
zyHwDPiK&MvxPDaS5RtL+a~xH?0@W#TuoI|Z7{>-!(Zy3eJxeaCG0P5Lh4jgF8?bPR
zzUr?jtlC3w`2H`pTTb=usW7vwkMv2OclA_Lizu=(g~fBbZ*4*Ach|p%^>mHXXp0z_
z^Y|?WesR8F2^nxmg8hb6O<TZR4^Af6?`T>to}#m>Wim|SMasj}w2!tQO*{t;)oZ68
zY<p<(5Mf-0oNqd>NZ<t+8T3_gHSq27CQ83~yaJ<&%KMSPxZ>z>4U49s{FWI{E)Fy}
zTm(i>;8^}{#Ke;h1`pGtu#>0wqtLQ0ee?*6oOQs_Qc1Mb-cq39Tj=W<=xKH5v_|v9
z2OBls-96K%<nE#+(FOjdy4LrwQ*=qnNt~eSNOfg>nfsJZ>Ig;E{itf<AsnpD=&WB`
z6OE9&iy!+9M-f;q7_ENUn++RyIbQPqciep!gXzXJIB5{$JRBB>3APCsG11dq-c{o7
zHph4aVU3tMs7yO<;o1tT<qv`@8EGeSh;CQnQ?=y<Am74}Q?k_0Bqdl3TxA?>H}iR5
z4EL`r_0ackd@1+IgZQ$YY9uhbn79nBPhV*L3O^dcEjz3b;-SCXyyDa6Emo3eJbdtS
zp6b%|X0I48V2S-3PR`HiF+~NpiNmmt%J9lIYOjVx`^O?{QS?lQ%q}q;bsU}44k=?{
zY#JVYPh*Ty3{t*aq+ko)6&)Ug0EVlW5oKK^nTL&AQiyNd!<L}~3*(?lby&Q9^zPdI
zPJiR_zw6-$QxTO<^!YkkV_CC?P$MPsF$QvIP<)U$T4{*`@La)2tC`1%-T#jL-Ifs8
zRskUQgEO-wAgpr&ORY84KEfW1&0<vb^Tt1-94Q$x&6!@a&4NiXp^g>S;tMTv-|H}g
zqZsXhf`y82OqJ<)ad(UBkIF+LnV7TX^*ZltGDXfdG!9F|CQyoSgd;p&3yKs3>!3Gs
z!~(3cS6yjRGGWlP2s?9RxGKY~3K&F?`*F5J99Wr{(F&P3sG?&sc1n0XZRsXcF+@Wn
zJ&4au5D5{HdKj1X{!CBw2)m4-??b_XWhwNWl}r>PQ`o+9aRhVJwl|HKgtG5RY!#BM
zQFFu9MMse1+a#!!?6%<%`Ov>jLK)*5><G%HhBH|kZ>T301TFv&=ip#^=@y~o((!eK
zaiGBU%tnSD8(ZZ&fbBiXUljYZ1NE^)msqv911u+kcttMxiKrw4f`%RaJkl-0uRvfF
zAwKNwVt+>cj9UPotdo}(jpk7js-mS$#BjGfj+II803%X?J;2o04$Z=>4`s@fotb#{
zs5e~5G}(Zx{>|}4ppyV>Frf)sreXdPd?Lg_k@z>x7%sd4V(Ja4N@=Hoc#4j{&Ko@Z
zG5ohh6B-hlCgoj5cL|m#<mqojsHkLH`0B8W2kz&D9?nIh|4+@fVI-xavGUtcWs=-0
z8~N<kfrEwB2B<LoBWhcT(<gu)+WxtQo^E4`cr74J>qovU^PBD*7x)>eEQl^w_;a9e
zk6=>z+eP%L;(zV*1*M@CNgH*}x(3mi!5>;E{;_MbbG0i|DO45<md!7gbzB#tf~Zp>
zI`;)(0}Fy%Z|bFnrr7Hj5q`8sIPxK-qLn^j_xQ<B@40@@<C!j?0qr3p4?Jz4@b!)x
zWN`4R?%DYu^Xh9yc9rzk<gay5J(LYk10-0khmqUVrq#boyP)UYk++;<Z#Iy18}7{V
zbgrDHpMG^Qpmn8pR^3Mvd*k@tQCz#tVRw0i5sIjrS6-|HVkhcvPk<`-kI?=AUg>4D
zz4ReDKzvM3?iu;E-}b$6dQ?~z`XeHcxYxarS0lm_aX@x9O<gRaSMZmqx_Ewj{V9la
za3Up}ozsO=M)U}02|lTgSKK)VWFkZ^pam1A8(8MfMr`QUTlD?4xY~Q#5~k~Y<eFb;
z3U*>gYC<uw5L%Q$P6qE4Y8!?jk$o>A2OxD+yeUHD5MLfeq<?u!d@q*NGc$~U%`(37
z2V`w_9lpr^@E==lU-r4@9>0h-b!GEEXKf604bk`ck$HI3K;ot@p&3wJ(NvhOjaay*
zzXr?*6!t<Cnf|HD4Ta92waS;SOamIRY9%YG;(<oQ6QJbrGmQ2GRCR-N!;3UVLsb!6
z2LvJic97P{*un}yz&RomRGXA^{W>pc$xINS1}@J>@qQ$6)`iD4AZ`fmtVh7fHsB32
znUM%IyS%x$x>yzU2?hq?(a8FXIynCDfv>Nv8;g82(F#=+e+~|s*#=nFguBv*lZUrR
z*8U}E|9Dzh*Irm@-%#uaq3$kNpWpU?G#VgaY(*P^$A=RHq$C_Wu&CMQjHAfN^}PzH
z0=gAQ&D(5jl+T>;;__(olXE$>(Xk`INl^6>-w%5Xn_8FZ;sw&zXJ)=I+_y<*oOQMi
zSAEOC*W=GbXB?z(3skpMq*?Q`);0PdUI1cW1TtQYu~yJ^TdQ}E?H``BNklI=DJP|7
z6k}&Xp)-E%pb6hO_6@N^T%3`-P}8H<Yr9nNNUKvVvyTu7({)Rk;+>?^^5T1HerLTI
z)8R^~e_PY<^54FD#rTAn9S=}OC#xCp*TMEKM)zb9gOO%qOBJ*~_-|j{rBq_6si})i
z!fFLcfTRTX9QgOp#QX+JYo8r}>A8(>yKnsb0#JJ0f@(PK8PoVuM$Y%k>QaK+ZUJ%d
zLk@dho7XBHD*UcMIcC{G4B;w%QDQZSArqsXx=yf+1d-d}Q1H!6pTg^2#p-!hS->RU
zTZnQkyX}08t6@Y#gK}P@Nva7G_p7;JaGp4U0ErlgrO$>8zk)ifWCsk5_P}o)E04z%
zYZF|3Iy9x-JFx=59&_OV*n2>fo+|1;bs<hB>-)E$8d4&^qU`F#(oHY0IfiV1-r0m!
z5vA|z;lsgFmq;sAq>M-Js_H1CVWPN)=7;(TfdfGaNtZ!909w6Ax{A(N)nkuEj@SUJ
z3xx}sYyXVi<<jXNXuB|ndy3mJ)}dKf*6MSKatYMZ{vUuE+JU7U1?C%JE<o{3J>z?r
zK1!1D^f*E+r-3<X8P^ns(_jAv-*ne%oQsOyUueUEJ;T_>9>W#$myV}}m%^Rr0x!#-
zFo`IzvbWjC(Af~WF}0H~isYm;<l$I&mze(T9DL--1}+;^T0)j?9ij>L;e)Y-$_iD{
zA}ffcm{}jMFIm{(+7T)+H8j)%vesxsCVZGEwj<aBN8RoRZ$RRVQ`&jH(w*%mQW6Al
z8)kiVkA(YT&@kCt^6byaHBEpD)%3};cG6OQlCy_{?)~R`c+aq-(0$-KfY}0SEP!Gi
z1w6;pr`D5FlkEKr_Y_A3ht%@jW7j>AF*u?#DQ@-O-HV6}P9d*-hl@s2V|^LU+&zmh
zvvN6JmZ;3`AE)B-{AKKzVX-i_sE5n1^O3TLpRIo%Set!vV0^sx4lnQyiDNbYNbvgX
z9W7-%5}rAWO%sfQ`>BR_@GCa<@4aebr$gnNY~XRCb*Am)?#k>UVdS*>4Mq{8p!9wT
zSlx;Gt#?o;witGCexa792Wu6LGbIz&cM>>!Ia1j`w*)s<E41;e$YbY_ZJ-aU#kF^t
z*hRB+!n>kqb|PXx@Ul{I2HhwTjDaQPAf$#ULf8%Wu$<xU3|mrw(mVaBPo*aqNR^D#
zQ_c5p&Z9~Si=WSp3)CDo;>(lT_@e(?Hhn6|iU-(VYx^=-wOs&^=F)1@g{*UXZlUtq
zwm8gNAcM5!p2gs%wDep|z`X7vC{^h6Z%MrkrlG0F66;xRgRr>%+Lu`H?VE}d+IG=d
z{~osMVn<=JpcQ=W963e?Sf_h7PHj#_Ly@izKG6?KMOnaegUfG9^)Zs89JPlo2{{Em
zvN)LS8Fj1+FT%W5SqH!#!N#@Dnq0Bs#e0jtgZsA!zzi}1_(`WJ9-?Vw9O1BA`Y>I+
zy9K^Pw@HwtT&u=(bbDvWmpYO6T?-Y{o%SJ+e?X`^UIDH2qzj-uKAYD$Y(aW(Asb-n
zLm&jj-d~~uRU%Q`fk)kL=ka%o>pcm+4rFevJ`{?=?L2WPIBHHqu?Lcv2oj&T@$Esr
z`~jlUk9rO$SHF%3jyHUfs~n3`6cTaHnP$OG@`LiE$0Z^Vn93+Q=wmdDfDp+Ic;Hr~
z5<m@ztBJzpfDKmGR*vhzMvGGpp6Hi5Cj3pa84g5%AzF~EQ0X#y16%I7%<76TBz;AH
zm_phov`|g^)rG`#ecZcTrv9vK(vK3{64*Iun9{~(t`uD%KZUk$OMdzPLmw<ck;CnR
zy8C`^PNzRWOD4G3<rw|V`3GPSvd+%}&crf6fYMJ5tsi0-a@((98EOrFe-#$wbIF{C
z5yuUU4g&ZotZ$3q6*&TpVPHm+uQ}Bfh_InamkvUnL9?Q@&{Oygw80<==LIzg_(33P
zRn4ylM}qWp1d$Q1QS3a8>!5|P_VO=JX{#@oB;u7&YCoCXHvkD}R|XWJ7||a&3O=No
z%-X#NyRjQ+)q^o5sl?0wpa}wd%jQvV2#G&f53R;eKczz$K+=Sf8jx!{x&gv#S#sxe
zHNUXW<>U{(e4WVYR`4GZ=Tk!VCbskEPxxTU!nT|JKpbM6XnUiK3wc{4^_V;$w0WSF
zEwl}|$_HsyfJTL4QJp~-BNs&#G)L*X>7NP*NW_x-eg<8IDxnyoAkvpoL(W68VFIWK
zUg%NOFI_Jx6wRmQ2rZ$I*yPPXKGLKtf$8N7bZ#PCb~4qVL*hv1cQxCMzC`Q`or2n@
zC_vVU?j2@5mX<q~!vJ~n)hk0Myqom}3lN=#-vdaMt%xNXb+lHBDhGfx%1e-IN)wl-
z3XdVGqJHHzxp^LFtav(m4O*=lA8}*xPa2NnUOtykt;8&wMKbxfy}3gJUZbC#gB10J
zEOZa>H9DfmZidssz~;UMBkRI+f`4q+Kf|i@@sWV`aT>vJt=W(4GjH5R_Y4x<4bi2e
z$JW|SZ?)@@ZFE4fKk$0OvmyasGdek1*#3{YvOHr&B`iwuGJg}`7Ll*bSm-F1mY$C!
zQpd`ssItl3KvP#Q7OgapkT#jSwtZ0&k{%`EOZ|Y9<PI8zy_LOEqyY&M42<b^W#9n?
z<dPp0+CPg>&rocO<gK1YpE%FP#(rQnihgtHHSpe*K8phk?{g9(M>Weu4GDweC2-+k
zl~p!W@b8L$?>Hy?XQn8N2#iORV%&w$4J{7c1bCA`2;vF15w^vWf!!e*C^wTdqBZCd
ztfBxSKt<cFOy1L6c+tCSF>4P{MZnd<r~t!6Ss^_~y=Ia=5xkS^j_k@baMx4KfWRs-
zHnKUuX*wYJ@Myk;NR~PoKDmFUGpXhZj5i0fL<GT6qZBv@Ps*5Lz2DVM90f%n%p###
z?QapyajwTd>S(^_yFMVx69E1S{_It($akV)gOTyP(XJ3TpYvLI?0t^wgU2)*98ro9
z_HI8sUuG1Q_ta>_JwW<E)YBS~_?-NohUbda;Sc%Qen|?nc)k(wN9XTGc1oXQzS}q^
zu6P26T`jL)fLIW%24`$xb2#%%mQszRNKaK99h?aOUJP|f9cKMW?&(}?p1IMYP+8&i
z?Y0kiD3vmtMpUimfivRL;5YgtMkNVvmk*j6)e;K!i~JRy5wA>uQQI8`sR6N^<upKY
zgN1pscugZC#q+MO2XvCx8rQ(+ol?LH|9jVwHN?3K3x*&Vz)~Ywuk;edo>%-)xgAsx
zgAzI#Ru~R)Y08+PP{IH;hc?);+4%nS78!&3DAYLbU06lre-^M0H5p4YDGr~Ja!v5A
zv~>#w1bj&Cd1m<#TmZ*g9$?vACt>6ZkbJ81B<=}tx(ff$I5k|^DzJZdz5x<w@RcjS
zJFS93Mpew~VHb2jZRl><3KZ>ruK-s))umZ~cjts&O~7B1V$64od{I_NJ-5Wpk0Lxh
zmQNq38F*mC`K`*G8g17f`b2`tO)w6)H6B&e^<uv{a)UF`25<Aj9oc_^(6Ye5YYTW2
zvkk(lg)4Xc#>hYb^TI?=;jtemEWuzOd#%hwC-|evlFNpj!7;%W?C^~vLO&wKmobig
z%dInx_5u2Lxc9(NQ#EGSQ^%FmYvKv`xf5?c7zDJR!Pq9BCvt3BxXC@%c47eiBlh%c
z$hsk?IA@{zb)#J+Hditf5p7rTt>@D{gvtnM-BjpR;q2#oAjAcYmlOZe(QmwWKwF0u
z+#Z1L54cpW&{acj?hN+-TCZcW;=(>&r!dmSoHgTd>HRzkVJ!<7>sjqQnze)1KNYD`
zp!w1>F*Ua~A)SGRD=%~E6P?OR%;(ZXgg~YPYyj{r@7hh13o2;IYDvnkS?{SE5i;!M
z&*2=wVJ6nGXZdlx-l44~0EV#Z?kfX#kE}EeA#N#bn}SK)0Vh;9SWIPqDPsj$1;vhZ
z1o53FDhANKOSKTC?0e7Ml;u3IdW$Kw($!7bv%B&tiBGYaeGdi7h#sJTa9<zNCI81r
z1^6;U?!V#)A9dN(>guPgyQY`tkoSo&_#msIQrTig*dNtDhByX<Cb&y_lw+ql*uddy
zOQ%K7@>P>E^VxlEe-nT`fo&#577yGH{!D60Eu|FcLvQGVg7lUeje_UU`l6s&%&@VL
zXOsxuXbaU-R#*uxlZO~<;+g@Wo7uyGIjyutQxesPRdYuDZB>zRHo;=7vlcc1OoJ>3
zIu4%5F(EpTYDjhI>Fb$oB_N$BEh)D%S5Ii_^;V{rU0kPd?+;b_uXtPYYkf&etlFxL
z!+b1ay1}1;U0{8Leu2!Oj#^!eNvcVLsJ73B$yl}%J$1tfKTt=MqHW&bRiRazsE{+F
z+MdiJI29}^ZPp+~Nc^}5*>5_n%3#uCDr4#CryK7+k733};y-9ECr#~grT+F39H7SV
z*ph~T9_GLW-9^f<-G}HuWLX9pzberJFUU<KoSmwAoAJt6+xY&^X7)GZqP)EDy#4=v
zvLL@!dUPy@oqRV{hg)yv8KXe2l(CL@GJtE=v<yEtD+OH1?B){C5PbwHm$8og5UY=j
zg+M5#ObUckrJ+`w{iYJJ5@KQww#}zrJt2`m>MyMOaWG5cg*dIT)G}xxKz$*7%asuR
zCn@jOGntSyK=xAp@Qo-kke-PkvEv~VB85B~Z3K3j=n|z?^Bd!SA;qA7yoNwysoB~V
zdmFWi#5VN(_$n@qH#IO`JszY7)aSA_BBIS%Ri`-ENC2xB@IHa;9&9&|=!orBg^T25
zle*6j-plE;1>R&LM#Ve>llF@@`}WKE+w5QFdBX~N*W&?d<t{)FlKlPma%DDvArJ9e
zQrvA!Bt1F-G0jhwQKM!ru;6;vQ56EX;y@*k3l!*_Nj39lGV<9!oaG}6w!^WjHnq!$
z5@C-v-g)t)o5x`O!xt1D7aw0_;{O%IbXt1)&Fj-mI3d6RmFfQ=z{5NJ^sanK0{Z#o
z$R;%#rv6StTq?@L`wI<qThUs50kL|jkVVBNtR~*W&gTTdN&zL@I<SgNQ|dC8pQ<QO
zY3qPunlEn@7?=@Nc$x|g$mMpK{NsVQ5TlwbOb~y_tmsmMAo#W4!O02;JMe3?f{fKY
zheN|M6lk2YugzgIb=@nTBBqAr{HZ~diXQ=w2wwDH{toRO9H%~Iy^qEt$#r_->IWA2
zd+mbf0%>{hzPZnlTjU@MO`f&(i_Pur?W{UgO9xn}s4>yeNrEociER2c_KOdLP8?Iq
zHh^|<bhwh<muhVf#=9H5<bzUk@Rn1Bq=2-~(%HsJpUoMCN`E3TB?;*OT&7<LRv==n
z2ITyaTZ(Wrl}B#Am-7j0b&?x>TR~UB!cfnUV`-C_li#of_<FdKtR6Zhd7Zaa1Y_8w
zmYmq&4am)@Ho+`tQ42zS@%C%dhdA5H7QEMwIdM`wwCszgCMVr|4c({Ce~d*(;cNv3
z1#SnwQ86&^mzLrfRk?Y-WIo;A-d<T*X<g^*Y3u77=%7Pgos8?E_YGLk!mq$?BzCJ@
zz7Pk|gA&VNd5KX#!8v1LuADoe&&-WZ`zNaA`r9?PxE$b6X51Vz#Cqj{O+m>fXHe!~
z0wNC~HxUSuLCX)d3I(zRRiUF-YUXOdhb|!(ks$&Fafz-o0|4a*OBnCygc{;Q7c86L
zJOTxN)6)4bnolUFfw|?c+jLmtiIYH5Obitb4T9w8-1+Vl<#YExnJ4O2+B*cnG}q+r
zG(Huo*`aO3kz9TyKYqIHe#e?5Ccrp*Gy2HLhpU3W15zZgS;|;6>k47?CuItL0Xn&?
zP}2x57$i!;?UVsRX0U)KPM_}we@dY^Mnz|*f=1EG;b2gT<;B!#%LVTNrZFfB&Kgy5
zrE17&5FpkE@c#$FJ!v2m=qUN1sixsB{T?0T;oY*4nc0D(A~wgWpU--~UOMFN3$WCi
z3%EZ%c;n#UT<phCOHUSSpEj5xW0Jgi^Tzr35Y9#WniKt@-l5-noUQK^EjD3a!N3i$
zlKCfi1gTRjJ=#hEAV21~rml{tR&U=Lq#6Rfhp+xa_>WV)ae@325S5bIPQuKWzRZ!G
zU%Uk@HrRnp5gs!OmGXV^Uz36P5Q!kMo(<2krugd~aug+vGp$ZhErmG`Mn#M;4TlJ}
zB}5+WTP@#;9ZUcqGfhY(@icG`?53N_tFY}e-aLmza6~hWij-%U>B;vq&z^f}!>(51
z3Xr>!yRuim=A}a`YgT#`9o^4*cEM)PJDFox9Qv}-^Iltfrq&_%$}#fueVnilLw9gU
z2zzM}4Gj%wHN?fmO-xMoba{Aq#(ni~^&6atMZB)!TKc;`k)Dd0%f5OAk0xqxTR)7q
zLT@XxkR6|qlJba(uCS9>*#d%q8LyaO3=(QtD=R9VadGVydok{)fC$MSLx*2$^EwpZ
z`Be9F;bd2IR;H06%%_KYdxz>s-aVOnrJrEFX`a2|vv_ZmKWv&&l?dvmR9Y}S!HNR_
zcQ<c`s!pBG%*amW>KC_4eB(qL7IrHWh5OF^nQuVX>y!SF>X<)e!#C^U+}DqPbG8Og
zCt5ZpoOh!r3NGZag}Hus<#P>FCqysK=kE1GPvZ|I6&00}mQN@9sa75yf|-x3saCG8
zyb2ep7bu7~;`|>i3N-bDh2B!Nwrvh)prfKL6GEEin>;VeS{`m#2f`DC<v&u^+<o)!
zpr2eVmgnGSOsalD(d1c0acg8G0X;ecQML@iBN#oze##&(7Vzf9CD93kl4dVmMt7@5
zz1KnE7;&_byGcg<IqH=`L^;dG!kqesuOM4Nx+($|p%V!R3!iTHO!`LOWV*uagO)@2
zaq3;m1?M)`qy~p1@2oicxuO4~wLw*34!)rqh#~{1P<(L4)8}xa?0v)GC#PQxI~y4k
zYE<(SG)wieYN=mFl-Jfi?g(h%k`T}N8mFILEBfBU0=^P$8qM}_l2G8(<L(^5A1O?H
z_ka+BeE~2o_<%Knwy#I+s@*I^Unpi?X3GOJNoJlH_2$=M<%Cv}l*|GfFFP-{zypae
z2@rt5zXlS>Fa;17^O~V_YR32PpNcR;HZm|<htuOIA6dd*t!%euFb=obCj{1Mf+qv&
z-b{`xo{g5)c&B@(Ym3jlLWW3_+u?%Ue8a@M14Crw@X`;ReLo2eFm4o^5%*JO^-xs4
zw^$tF7-Vcd0PU0*5f`<!s0*M)oMa50onq{w*?8m182gj@07kpyw5;k^(d2sTk~an3
zu6~Gn5TN)cIkF9;oB`HMZb$A<fPr{cp^UBbB<H45oIs~jFEm^nI~IgA`8<%ZpoS~N
zF2-=!4SFJlEdTS%%Q8VEE5#B%D=X`4R7l>ZPlJW(MLv)BmseL;9y(WO{C^T|`B4`>
zT@^Pc->m)|an1)6aWZt?#h7sa(sQS$`(;1dP<J(nF%O(%rKeGkLNwaid+9H$+|n|#
zz5LTj&@Hk93`cs?r(?3Ucr?ahHe$l7lI+yQ(h(pkidM^$;{4sJk(7tV4V%2zeEo$1
z<!}0cJorD|_oDgrs6dkGHVL1zw)X2gU2=78by?Imh;qW(<DyVxZIn^nNeu=+2{wf7
z?5V83>Bu&@eL*JgJ~_}PWGMik1=t=8L)v>IdO>;vh%GrCSmgro!61=Av9+ymv;xd2
z)6J*PqLMT=2#%bWhU5MLfFcq&ox&)8<UZ%zx5c{i^}Hlsl)S%k)VnV<gq2DNr+R^a
zej|l;2Ls#k5&E<eO3Ikf#PlqVQ6;QK>4R5`m~yV44YS5eGac-i7nx?$xmJRs&fk8>
zem=U5>)^@_wzsfkE4vI79erhaIz)TfI6S{@x3T<Xa+CA(3i;l=xot|UhsULpQcCF)
z4%Tbi(DP0sm!r>y&lwFJ843HY73)e}z27WQFB&||8H;8eXH_SRR|x2Izv0CE&7%ap
z%Ad+!Cy1+@q37Q{Vp%SBwmxXl`mU8+Y(45Bzson(!bqJjr!k?e(Q}V{nfoBWUPhLH
zh|Cq0_Q{5ereT<-!{v$K2tKkSHu7P_*1+*tqP3uV5^ikxQ40@a@=gElE95tFsss-W
z<#z!mA?;W*#?Re_mpv9jO^WGTo~?W^YN${&8-`E6_Of~R%oaP2Hze$MK7m;mA(+>B
zp}7)^^KeqF+CcxX$U~oRQJ=Ezb~SAxuze_d^xX(HF2Ve5c5e;k8w%k-2zQq_S!dgi
z*K0{D-GfmFBF+4BJ;jVHjtKUIs|~N}H@6mq44_wwT#tjE<g_VT)C`|5ukLzIRg&!A
z+}hDD#Nx2DzhlEhmUj$o_=&Dw<~&t)Z!qBKoXH4TfkGC{+KnQ}d6~_ZxL;|wa%Mzi
z9(^k_aME6Wiv66-r!ENjnUs^#OC>>=iJS5-UX`b_LnPsEf4^?D5_Vku_DS8VXsrof
zs#%$~$eOKGyDLv_xJ3BW0J81R&#+;0qM4$O*?@*DAA0lb{Cyj^3UM;8a!4{)#n;70
zEuo#$6+;<?K6+%)_p_VZ8jDp-v;6s;<Yc#F)7I!k`lAmnp$OZVBq{fylQC@4TG64G
z??!6nhHcGS#lvZBhjF%Jk6)_e|L+k!elT(>+(?UQu=1gGk$-RXZfmkg*Y@Ix9BLzu
zySC@dM$1Rz<g@t;6(@)N;Oa?~KKwA>Z=6DX<F3Do7ul=?KjByM@17lQiK+hil>^9a
z3e<QTr3?iYCQXVQhsIpBhk8Z&7=y)(_$V3|1<%hn3%+e;3Y&Txy<VW)-!_VTNy=yE
zBl&ruei4m3rT--SLm>ahK>jN+GVU&VsT$JcrRu_*^bd=&qAA`1qs{FHi!QYymSY~8
z^TE^T>fS|>5HkJNO~Y>TxMsfPn!!O2-TbqrQ1Gsu*+KE|9@OS<4^Sn%rwDvQ*#BOy
zQVe9Hv%5$aOF^eG3nAl!80(=wXF$8HX6=j-m2%yBUxEbH7WQge^i_1D5JFdexzZvT
zOGO|-pz>jh{<)mnyqf6u^nX{@oki-)B0Em#wdi8}f~p{q>&RwM)L>^(L(%6zx)fIP
zV;fKVLsswx98aaJ^!GDv<p`?$$gBxfelK>+?@f?>Z(3UjqyDT?l?3vE&<J(DL-v=L
z-)`b^-qpL`uh1%@e!}u@J4B2Urf3(p$}j@2UoO1X==Yq)fDlZ!6?18RT_7#m=6`y6
z_5R;AN#D&O-?{6p!h2^FVqqqq#S?HMB3?OT(^Q6y9KMuEbQ21jNrV*EwzRME1Bv#Z
zJ3<U0QCEg*8)p@r*$<V)_c*v52+uqVvW28DIwE7zy|_GVpwy(SO=n2&4iWXLJDpVs
zaUIE#AMxHRa@wZ)a6u6GD@&C9&qMnd=<2}XmXY~(G0p|n)rWD?K#w`;?Ssp^ZIe29
zS4?ac9Z_9J-E#iD5M!_^P(;|~FYXrT!WktfmDi4!ETly%n|XuWNe*BShFLuh3J(ec
z1uchqV}~&CL`}bFXXN?~qyO1?AOcJ%36Gfc*V%pC`Nd16S+d#lAEhTqDx6~dw`yJZ
z+eH10;;JMimO1iMI@J6QQ({vSrK68bv#!ctB#<RI43Sy8$?j(l)jxT5%WEkGp$BBb
z%Y$N-HrDn##;DTghUW(*O2FvGNMp$0S_+0~_Z;b>-Brd1PAksZQ=vtdm`f}<eAk8X
z8D;D$mk|@$+lT0eF2AJa&zPF_S${vtvD^=`A-Ov~cuUW|cmjhHmy4euAz#>y)V9LU
zkp2=pNTyr0g}YmG>Gy;`A?6f9MejZ9a>sfHmQKU&8kfK&LLmvUfdZU@=Z}~&2G#Qq
zWv#jWDKGE%C0;7^Egy9V*Ie{^4mLO9H&-8*SP%*`(^S`Qz4%2I#}^VgFLYrr6~mqa
zG0(sXbv;$9rAVc4A8^D%mO}npN1%{o-18XQ57WLhT6ak-xjsTS>L(>?zVE$vy@RFR
zjhJU-4jOZr@pkK44@N7#8-_G6mhb5?V3S*h1@dn(K$2C*c*vz`*sf-Aa6oM~_}FCB
zR90>Y+3+-aO`N9`vGVQJdN-@iCbk9m#^e>~PCO5CS36fjPtcWOB0pQesEBw<hdB*0
zd@15uQ}=z>bojLai@{wx>MdPRcgXDP3^yACV&{?6xL6MRZ-wGhLDvuF-a1x#Q51iF
zLB2b9e5pzPV=LVr@yD<B*HX4aa_96=43wsgO+ykEtKwStmc0jlq`C>bMMDe#B?9e`
z>=M&dp<m`iH^cM$6=S%Q$?CUIewn`oi=`^iywY{;_cYZQxpP!3uf8YquZvf(0qr1q
za{buI)%D`5At5Rn)(7%99J1?GnQ&8zmnNJ<s=NovJO>5mNv0TMT8qJf{~R^#g83<p
z91WM!2{wT$hFb3*GbOVgwS9Y}lp=nsP9jP+ZV(9BBv5fbKZJLP&i%7xv>1@_yr--M
zKXK<VzmL?Sq~|g|43cY(_psStc1KAbD44q_(!cFKOH1|UJTr)D7Y8>KY=AcMar+zP
z`Qi-O#QZ%%4c9+=z{7Ue>9l`c*d`YP3yyUgSwZr_Q%fz8^YN5K_TP&&!u5Ur)%y3C
z`C)7TdDQ~FC3tcX^sAenlOYtzk<`y1gj-Gi|GSs~np+7O_YS)+7G;7R=*(o#OGIod
z70dTjlN{Ha)i#x5_CMr4p1o>QL=XJme>@VKN;-EIHQ~BG+9)5$*4mJz*14_JHLcp*
zu%Q?nJw#siFX0Udx(GCEcJ}hVP^sN-v1R(rr~SWs;r^KPS!Mi@4{ElW9$*+{01GiU
zH`~yie8Z?q*)XuWnceV76D;PMTccmi{h|N+&f89tfzX5XFC6yk`*7f&h#>^~9(b<g
z@DHxC5wH`i$hik7OnXGw$W~zWehfs6uuC?4{a7$+LV(;K=4VmbiE%@78QVj8iT|jE
z^uOEGsCl-y7iPmZo|-^)|1l8Dpj9<78yvj*vwSZ7inK1zD9LjE_5d4?NJ1Vj*GD_@
z!(#K6mI9wBk-+b09+neD&|alD{_}*B_`pW;&(ylmQS59(K9v>qpu3bw(g*}s&;ALo
zVFmi9^f->tZJZ7KxhK5Hv#)!^h3HpD?3MMqKfaHt+1*?rLk;|UDt-(Mmiz5)KTng<
zIOaIwlW|~b!E+c9VZ&F2vWVO%?%evP-&v7tF|XYy!8&b9K!7B83~W>hOLOmw;#X!W
zZ3ead?~ib0FRr71o0>lTdfG&Y{ED5xRAI->bD{)E%7yqd&e^W5Q1j5gcSDFkR!;fH
z#US~I-_-nCOmHL-^_tSwoY0-7?_iZpx3EA)|2BKyIw4=!()f^&_g6CoXoan|YYSfr
zo6K978=|~dGW*}{V2G_)MwUm8y7lE8^d}>G!f$;$vCoUw6xc)v8bTI+OSy6Sn^JY2
ztwON3>PkJz%bxu2<3AabsCU?XAjP@#O@bGwJ6LJxX{49~Mf80b_Mc7?u`$7JTj0;Z
zAnI*()ql91drdsSX|`Pb&!!Qq4@@9a1?v`+_s30Y^0z<od4hu_%Hzlenw?aIJkp8a
z>D)`6b<rlI5u*KFEn904x|`BZ+d}nFWCp?C58|Qi{y&fj(Asd_lmuy$3ze~sRZ`B)
zAmo2nLnfDyW!*){PW%C#t%Nwm-p8Jwb+&Frd%Ew3Vg!OcwdFd+KMSxfMdt=eznQ~Q
zsRiLup95B)f6ic8RDMfdX>j8!7|Z>z-1w2d4Q8ogDOAsBYRy6x?wKU}{oPv7lxqyR
z|M{jd!K4Z{XyI;M0Q)!ce(_!JfcID_jI@6m)lkDyoELl3#+4OA1}-1}y^vH$ausTQ
z&&LlR2=7|>qXXY#MKIzAOF4`<>_^q1qPH9Rus?k6>gM<a-xU7OW*pki1VvHua3c7I
z9kTZ4Y!)K(nKoaUMSmOofp#ymEB5&G=hgzHRO3cMqs?al#;)ZBsecv>y)(zHmrQ5N
zZx-!f(h~=fFD<Z1U*ao7#OAG|H};pUgaTQ*$&gvY|88Js3FPeg=jcA}7?9a!DpO|@
zh7#*(R@beFfzOF#P5-Q{kr}c9ekZ|`Pi<0l6#@|Rw1#5x2kire*6mlHmPyRwy8aeL
z!NfrG)A_e^Pec@UGim3d8`kNM8wzwWM|~^%Q<F<`D*pL(c}^VETblAcI8wJSQ}6_n
zvVYMHTxvFr3WGkLGqSm)^X9+5YZ%U{(ibJ%9~OuSoS9*iR_KDH#Cr;k%=AC0M8*Gi
ziI7ONKQ(yAd=m3%Xem=bK)uBy<o}WPmcel~%eJ7HEoNqBW{a6wvcMM0VrFJ$W@eVf
z%*<plGc#-2{?0uwX6~7IcYaO8Oh-q@j@`8>D|6+_tlE{W_k@n=L{z7s<Hw`kyW4*)
zJab5JeW+7XnjQUlifXl|!hUIo+zaVrwryqRR>^;A<=pL#Gvsf?(h%*RRAEsY!$C&6
zyR`}7bDT<r^SFl(I@j)fb3-?KYLI<8v-?+XY^Sr}i!Q(7uaDJpddvNQ$FAB4`!uu#
z=>9<K)Hol+pvA{2J+8<fCk#(L*#qcl0It<#)m<*~|I)R*t*M_enHtDYuJH%>*U?rW
zfKZTRH|4b&_$kDa5n}&!J1Qyu<n;hIoN?4ecl&B_sSJp_7ExiqXr5Kl99@z?@ESCj
z<+Vd_IKDa~Y=yL>ice`E0NMV-7D_B2BRNRs=^&Qr&t-#jSKStLwRm{B9?MWL^chsH
zE18eG3*6t4AsrPWp5l_cXimgCkE=f2Oa$cuBmMu0C<3`s8V@BZUdKJqpS+bYU0AL@
zptv&t5;|*GhctMR|Kk13x6ayi7M=_K+4a9}Ih5bLzw0genY=IkbhBbl+3YE6`|*=3
zYCGq5hU=Th1qb1YGZIwM`AbVp@<O-ZU;dnPgBy`vL+E(g>2VqN0NwePoeYSB02c#b
zPUuECiSA{->uKMf>MFp_L_byk*Tr@ZimkrJIJdUTPaEsyf&wt8Qd+l#?(u1AX$|?H
zk~JV85vcTkH8<!`at2$6?P-Jk{1p9>YNr7A_{ZOzaLX;^lv~BgA21wV*uPQyrMiUh
z0NetEX6$<SsIKn2tL%`2P%!Qm=S*IgPt2z|REpVkw(<?>CNA&vw!0lH@L9Bxus#2!
z9q^%{q1TU(*_^gJ6`J+*Ixnk66QCd<gV6++`}=^!>Gg==i>#g{;h-&bSPty*_ylvG
z)E9BbLuU#rU_-n<ch{^xBVTwut!v0u72LexA!mI5a2BVKwzk);9{*$pJ${412)q4J
zaPakYs}}u%u$5ot)w&&8yg5Jmg7j0hJs`j2Zl4^yhd&2M_xC_~ikz}ls*aR(eNks|
zuw44pQPE*>`+avJWX~u)k2_k4?TA{;``CAgHdWIph3#MRYMAO(>*NDe7x{R%&ty)Y
z0O<kb8#gBTK*RPI)b-XIEx%pkp?1;Ktxrev6^GpdE+m-$`>1i^e4vmah%9gfs<ZKi
zZToxL_9Vz}!%1#?vLBsYt@m3Y<Q==k3=?nr8Dv9-hnoKX(l&SFz2Vq=@#slBdQlPH
zx2wJzq2;+b``!R(vaKAAwRaoNll2*dH9;I0BQWJWV&$fbxu_JHAg{Cs0&@B%?#J{?
z1XC7T_`JVSfHbL4<k_2kerQbvKsFp!YcbYi&aUq6kH>R|x?P_G7<v|)ot~FpS<Ge&
zB3?=9=<4P|ni)DLDStX}XXY^Gljh0KJuc7YKlEZ-DPz-}uZrDY=Ki>E6KR{0Sw@U0
znl<D-BBh|9AR{y46bB59m`W)_$_fej32b7e6%`bO)IGd~3vD3VUM1^!QDy512?^~D
zg{(5PU-eFA@)Qy<Xg3q$6K?VV(v5=t5bLLpfIHL>XgQw(0|!sNz!9#HTd&Z&8RRP>
z79EGW9!&{L&Zszx*rX?Wszf+tdw4vtetY_Pd&+`gn9-gRZ^@XL8gfs6&CbXre&^N^
z&x%&g!K$h;J39*p;m^W~21=>8nB_fVB`&(r?pkj=k(SPG8CPujYb6-pjgsm5qD6t1
zm$yhZU5$8E^wKK+jYMg^z3YO6z+Nk8y)hgRbFlbNQP91!7hUdMwPvzeBP@+;99_qn
z7bjY_c`vQI>HYZq+i}cEu@QYX;XKa*gYH~JL}b45+tTLdrm-8K=8<}Z29Z_BbuuP(
z7N>0<7+$;QA<S88wC6&Vr}83X3JZ+wlTi;|1Ezgl^_tA@E8(?%2I5sOMBO%Jl_gu~
z?01g$Zs>aciI?saKv)O}(f;)Dv?T7??|PF!7__waRDjM)&PC(%0SZ4Xc)h>0G&jRx
zGH&sA<(k4vjl7&Ioz-ZTLFnw7(2jAs)DiaexpSG0Fk$tv0yI7j-FiAI1?RVwsm4ST
zxHpKPU@p+ii>dr-zVI!N-yd@IC-j-z+N%84G^t6(dRGk#YW8cz#|P-g+WU~+fbG?V
zdHD=XyPf=$(jlH~ajb2Y9yy(Ma;u}B@zwL0^wPjrgXTd@B)Si*stqZAH)_74rG`s{
z<E1mXl*?+oT{1COr0PfL@{goe%`QMBgwmm2eE!0P!S8>q>&?VP<GRxs%C5JKfhxz%
zNJUNu@n1Kz`~6zk1M(Al4l6>ln7-HoM&?cvg=2#U!0*>OT&xk)0U3=5luzPK2lXx<
zZsTrNrk(Il2g%u4Yh`D!P(dR(LkqKv-*Epr9jB>0BrGyy2R`4wpzlwp^MCsOrOE)d
z{9kPTPXzT(7zao__5Twa{W;)YsVzXee<15myMLVP-**4XmH^KEkD>k{;Qu(Rze$XL
zrLq7`|8bpv@8SP5r_96uBh)_+9`iTR@aMsx|1Z1#$w2&V_U|6x{!csq{qO&F<NtqM
zx0Nd#tkqz8J<c*UU!ghI*GIy|rI|)=b$7Z{4ZsL``ae~m{;ip!+%D%UY`H1A)#{}x
zc-#({O!@(=I%`LB#nCPlmz!OmrNvock&%Bg+JD&!bz6Q{mv@`fd4FU?M8wVhm|lyW
zXxc`hbh2nT`u2YrL1HTux69Sm!9+SB%i3-*1kXgq%6c~=PLSC0rw|LT+`sLCTTS1<
zfMXA%LzjV%keE0RFx*sCRRx$m(Cl=cmSVTlA1akdJ<_#3#?S?Tyu%JJ??h@<{5mN>
zt)w|A7F%1}o!i6%Fi9C183_pq`zX_2S`=o~{Lugx0B~j+8Qj8idwYg|T!7;6H)43*
zF0XZxE)j9@RbIR6oq^AnrLG%11b71ejBCPO!9o9sIMY!Vh2MBn8C=Sp9<C#46j?4p
zrj)Xm8y#)Y(Fp&gt_rRl50{$+=;-9ed0Leh3+2g#@@XGEkOXk}yc)E@{{H^i9ycQ;
zC6s>>Uw@1_=f(}NgyKS}@?q@|ez|T(YjAL|TycES(9jSpN}!M9>0%;$*57>lAB@-_
z|9AKJ{fWQ-00ZeiyZ-+E-RJYK$Q%GL|7qvH|NVck@i&?DH|hURn&B@1{Rad3e;=q+
zs{HfCBDrmM0$AE>%GIlN>kLC#SUX&<cK`%2({R1W_4Yun&hTiqFkCF(HmGQ>SpM^R
zh#)ifxJsuDKmr2Y5Y~DC!<yh=V4Quhr~CW&$HmzIT@en3|7E?oAsN(xMy+BKN9G?M
zKDJP<o|#(;;GP_t9=mX|fbm-{k0+b?zQ2M3^zT1C^GD1gf3%XrV%cU=*KD=1GCO;d
zi()v1$)Nvm3@a@RK^;IpGIGsqY;0QW_9wZ-PZlfN0H$GLO070&GL>}>CtEC^y-}{t
z;Qk?6sdkD8A4axpUfr=cu3M_x;Rbl#G%mptz}*zf;hC7OG~4bTW_xd7rM0%Us#R&X
zq}uPLnq&be9Y7G3s^w~ZL0>X+&HN$0h-kWh3@~)P#r2RA1t1DR{ih$>vhftAe6zVC
zfF@clRb}2c9Rrkj(`7djP+7|JX*-yTiVDW`<nnSeR~T}pL^horK;III%T>w%LJU>U
z0Te82vRYRXm`h1X0Ze0hyjX8DU#eOJ)O(y0kHjX^^-ZGthR^LF0W1rsP{r+dLVBPB
zAnXawI}?_xjgk0#tp-C;090C>1DOB$$%z(V?3qfr2s2U)oOQj<XpHBwodi47`5=wW
zeD7GRvQe+gOaJ4~e4?F!a2Qj!kGE$4vkWf+K%4#nIVq{cTz!50*O#OGWL>vX&3a=w
zJN*S_G@{=9vE=-`yhf{aEr7ge)T=^5M@B{fZ~~BJe#*tc!NK-+u20d9Wux<@?k}wl
zz!L!QZLApGdM0mRzEs)3z~FE^_4cH^ntE#%2ZLT4^Q+ez0F(ycer56h=nZh*^i+4L
z^?ZE@L!&uy;0~@bV^7!fdNyKUU|_oB<>TA*xRnj`WG|LYKiJ)!D^bh`^wJ1QwO-f9
zys|tVr;R{G7c^TxOf*Q^BmjOY0>Z^!A1^Va@m&M8D{{gDqcy6(HCkv!VzZ(8c$u0W
zb$Y%MlaK%=i;?q3yuUss7li|we2oqXK^&uG^6u;HtvMDHr6g1nwBF=eh)$upIh3Q7
zLv!j$e%x@s3}@(!QJc#0xRE^8mm!ZY{kc#U)^Zm#J-yOXQ9*nCjO<xRFd%gdkS|_q
z4(2TX=i9B2uvrPsXxMh{4S%9OrQvx__q9I)e~c9@8{pu=nofYirL&sFH51jioYfBP
z?CusGf5B#b`gpxWV~0S>7qS(ua1VFB>G6jwJO+h?--AJYh}MASa=j_Y&d%0olarA_
zRqi(xipbi7R4Gx|o-LFPJEloW1t|O{z&QKQt-IbE0TR2>0eC5m4!2*#c<r{kgCx<s
z#HF^clMgoHU(2`E^*qTIP4)=%yx$#9C)TAL6;yOv9ZDU(MH~^h9T6FwHO9on$CC_p
z0f;$DlfR!Iz@8{J*}iuJ8NqoClsv4p*z4no(Yh9BQdMBwGEq@Ux^?G?hOe!y`IH*k
zzuX;P-tIkJ@7{nW(Q3{%TCP&7mW`)#$ZN@kHLU?OVIOk7L{V5*oocMb_11W!h#hBj
zploS6S9l-KWXE&4RGH#=129Dn@JW6zf`*3PJj6Kt(H7~iuUZU<)NZ0XJ3H6goSlz2
z9x}M>_a>y`FaTPr((X_9tx>zpsnTKfJ7dWHA%XW>Zc3eUkqi>Ig!&f1KBVKxW%t?!
zBd~Ulk4t@`ggT<|csg%JY4S217XNtt@Mp?J-~vdG_xq)r15PC+B+!n>Tag%(+nT+-
zJ%}&Z6?$E|Zex<^<OLem@-$<B$U=UjNH_?t)+^$OEQz!ZBhMNjuYLD{*Eyo3!lf!5
z_pe<dVqzO(Juxvc-fz1EHycv`Gy8(WKAyX|tPgl?OMp;aUEQp_vkvcUt>qSA!Fvlt
zr?vt4COv-O`izv9jT?Yzg3=5fq?<gZ(>YULKeZ+8od&|t0PurGAX%wKbC9nAB0Yz}
zpu3hYj<DyZVC(Fh&2E{-%7{V0VWzL5GM^vKPff72H~qX9FAsnP@~%y*O)lC-0Ku7k
zDRvQFEK{YSq?Ew2pBwCHXlNJ$$P6%NYBt*Q<?j0C#_q^Z|ChL~9CqYjakkeBpy$Gz
zZPuD?0VtGcMex<D9l$wYIP0XfbTCf0nV6UWFI4EhAAbF$+Gx6Qn*>CQ39Kf4eSHL8
zPuq8Q4)wpaYxM^<y<TP0`0e&2PJcjAorgRwZ}$Zk=GVw&a+8U~G!l!6i>sF?2+AO_
z4A0Hg1K=KplE?kQ;#%-^H~MSb1t|&1?n9;lJfOEOFE1Ia!sRt5(%AMZQKE1ZrBHdD
z&X!T+d5Nr!?%m8@x_v>#BQS|BIvkGUFn7bFbA^{!G2y289@ieuSFJaHt1kndB+!$0
z09k3glihM91lX^c$Wu-}n_mYIBG%L}UW=UrC<`-sivGhwiQ?|#rgt)bdR{YY2f(~$
zS5|yMA^Y8;Eb?{__D164?(@PmvXRR7Z%!gL*wAk80Gv*LU96(NH-4RX!wx{y5NUJ(
zz*?<bJPAO{{Wf;%jW+=G40+6Y!ql?1v|nSj-e}KK-J6NgsDUD&Q7t1S*SchPT<aeg
z(Bw$$QXu5_(YGNwDAm}GePrTth{EMkYICw=y42RzPG~wx0mRJr9tlk<PcJWFqx$?#
zr;D<>Beq8lfVoPorR@Zaq@)Fs@#J|G3U&aF24hN{R6E80T7hX4B3hN=dDFSx-{1cS
zm&fNu6KTeV1d-ea^=l>caEH?YFh}8vWGY2s<YDYO#BD`#m|FaQ=N|u)%lzN(3rqTh
z%neq>g)y=+w0E#G`eFHJ%Ub^@3=0=0D<LD{pG{s~1{otO69-d5W+p~f!oME|F*6GX
zBRd8$iysa~B1Q(*hDQASF#kG*%gVX7to<rGV#ldEW0P@_SWWki`CvZQYAs9rx8hDC
zyPxU0X~OQuo0mKO#G;{KU`TZtFNEPqA!N0$58FS0!<@aH+Fu)5yf-{t;+C<D(J*KD
zyia{Px9%`FJxf$gfKrmC`MfVX*X-z7J1&j`Go&RMgolzg79Pa>3R$+IbZa`8ppHGL
zj=&w>wmvVXU@KMM)YiaZXV=NcGIT#=K(*b{4Z)A>U6#E(KiapCAAC1{+<iQ?KfAM>
z`-qW)J9T+(K20ft#H%@Aqt3y%dVd^sT6+D~JfgEVen)zLcNj028Hd@m^m-m`+dLkd
za;S4?eC5b^@pQ)9cwcyz!WhPCjDYNm%Q#PhyOm=!(P8FEQJ7_lx^O@VbS?s)bNWqf
zb1*+|;gbUHhq<Ix$`_&thst62jY)hErT=N@Jo2DdIKZb6iZiUY`mI)m!7qaNh!TWg
z_R!P=7J=jQ%PhwelV|>%CK%uTjtS%PA+#3)ex1YFRJZAB)74PVin7J=)a|4Haq9_x
zukBt7AJm1u(rG)uc`=gYC@LiAXSDHCo|pk9>+M?J4NnShpg7M_2C(4!6_33sug6CC
z$FSs!l0XtMS3SHLv}EN-l>}5EYrBOxKpz7#SQen|MDOY2A*l)_L#Csk90g!v$^zNI
z9p<{2DL)~NK}c}xwb@}J5+Zmw{j_gX90r7^fKr%OcqkadLZVct#6lTHGgQ}0BkH@E
z!<UCOg&1RYpd!RsJ*Zw5MvhY`ITUNv!H#Cji@&!e&{CB|6Pc+=2Ap6T>Yt(osz&Nd
z&wh!}HqicZftIgEr6*J(#x+mf9QzKxlMYMYaobbz72k>`1ui39>~Wd_kBn}E!oG$$
zRc*)(OL)}mAZg(U3v(v}x;uQ>=OxMz*Dx<ZGpWBDce!-kPMqpAMWN75AmLyRjz1p%
z>b7Wl9KHswxD4A8k;M=GdG1BCeSJnk6n*ZCE7-XQJ-Bi&<RbI7!{sea%kO?jf~=UF
zi43MuIbZ|I{6UV-xv5KK6y6OH<@k5>%Z+H0-siQlDc}9ISwV1)ZT!M7@w2+x!>W%|
zyVPioW;l9O69u+o+!qU9UOhpq+-+SyYO*=+>C-Z{C)JR1@IkOlf2dqr<^+aq*J25M
zV^UZzj3^gDWwjje5edRR@Mw&pkFoNOTPU`rMeHX!i@+8{1!hzy=ds%^x@~YpA|!p*
z9uEV;)S+BgumQ%GlkFJ7Xzf_Jb5I}CRF;caQ=p_6q?1cTxz}vS$mA_MVi$o3z)TpE
zS=7){d9?m*eU!&)vC9E3qp59NpJq1$bE6M<fA3J_f>&4<H%^X~YmCR}s5IIRf^QM_
zTnr5i26#?LM-uIXw!NW3Xr%LyJCnO7L`Ek6$&|7wo(KLsDAs$gY9pXmMEjjNnJz$=
zO@w)x#dc#B!flYL@?JDd{T^?G6Ka>T!!)@dhOJ1qI<Ikj`t@a?9BUlB;}IsnY{rz4
zL`D(Icv@G}@n?^FIn#8CSCFH0x>R!h!2Q+pdQCU~p(JU~i7}q0=jKdoeN&dfq`oD;
zm{2PFrup@S=Zq)>1)8qG{FcbGHv|=3d}ahp$Al>%TK&z<Y43gJC?{-Q%}=cn0bf+n
zYdgn*d=H>FH!>6GTsKV4aLwrOcFy@qsak(HJ~7pB)Nf+yB#-DyBh3S)+DZt!19vg>
zVO{Lnz2I#9(iozruL;+Qp?$K-)Ui=;oD0pAD<(0U$t#DQ#|vuXm%285aKC2+!m~0@
z{JBe|?lsxS=onky5^E0^N?(L5IUbC$G%tQ}#?nqn%o~AB3&OH*OZidl4z?Tq{?+P^
zRs921+|o>$3*vitd37?n0WuxvzHTmNx?Hy<$nq|gMQ3^Af^2kj?Xy)v8nPP|StPNN
zgyb%Iv67M}GA%`NL|WY!3dugODcWgtX)Z;Q{j#=X+w_pphRD?bV|_1%?%cGwV`##T
z47+VjLOirpR)(8dizc4ELYhEPKjP?!6DZ>tU`6jHL5L1mnsAh|YZtxsg$P*J1S?DJ
z;obP3g(LSQmCX;SEO4GJ*X=2qej~ekN>D{7ISl@xZg>v1D2&h)sq-h6U%zjAbvKi;
zRb9K=*Fio>y?&kuXZgm3+6S;J)cFxDC098v0PKWDbhi3DK^F6S$bseS=IH9@fz{D&
zjeS#g!Nh#$a1UIiXQn({gmtHdZqw3EzpIX87Oe@vAeIPDk&w8e6N9E%*`?l~HS@_S
zyH|V?)oo)}wTJg6@GfdzK6e<FL&$oxHmg7|BfvF0LUdmFbd0?R2&1~T5_S)ibF`_Z
zy|`^QEGG|M2uY|0^Mg`_poFl;<ZVmRxW-1h-s5RHAvROqtWZl{co#w7SdQ~(Q?AP<
zJ*rG=!WNRrC+dEORA1Mk;xrU1Q}0sK5{qU63H<4Bmd4E1y0jz0<|fuiXHpvp(*ix=
zn>dc?d5vfCM+v3DGnW&tyP=GGw3mWdrn@WZtj#mj>j$}_B#2%kzGaN^FLdioQbdfl
zetD*Bt*}p~U;Qg5EHVQey_t!&@lom3U*i_gXK9PfeG^HG?Wds0zyTq%&0h!uj3#@f
z)#6TGldo|^&=v=+1u&3yDia(AdyQCIh5_pWltno}2=b1uutV*9NWXcjroS-fVIJ(Z
zwLE)my*|61+h^R@ej5x3FYACN%`j$sE}T$1DQv|V+h9@*W6R{!900nHet|lq0*^@5
ziV}oaY|k2E3YW|+Y5g^*I_VB$xA$3xwc6Yhiqd!z!1F|OHS+2`@blN>e`FzC|AB?D
zv;2=($U1<9?2#jOysA?+6;K<S<vf?Hq*1%$TB}->J~bvPchdhBwj0@$ClT;Bj1(CE
zK6OASos<I}eK7>edbK^}i;#VRuXn-EE6s*!j1+;X`+@0$Z}S41Vf7Oi4rmTJD4C(}
zFggQPc8CNo%%19)(ZTXKBG?@|NvRtY&!(9JM&FI<2&k3+<x{zpi6f|Z=SKtK`jBbr
z+_p0I*6b=NjewaTzL)Fsb=1%fXeJTQ=at>#<9Z!P4x2FB+sppsraE#;gT6qF3>Mql
z?d#Rhx_o?~VJgdM_0eT<s^l1F>bzl`Ie>+H1R`|J^bt_iIygN!%e)Lw9FFJWd|-Q#
zzw$z^fdEhb63YOFACAFczDT)?2HAS@Xol=b>>^b`&GjGZ>p1YcpcF!d)+|sHaEk~i
zyjhwj4r5ey2rBl`RLjSbqr}@C)lnz~(FlpC;_(ekqCLjaNHTBTlZPfXBaz?+@@K{;
z_A>@i7=6OsFynLBolBtUJ?)ty3ZCO}aoEdu_G)#SF&@R++@GVw$Sya%$E15bJ&c*^
zLVE(PV~6OxNn1YH-6mmDGislt-o6~EMUEf2KL3~zb~7(bXXP;IkxkGSczA6NWtNew
z-yzp@aeXU(U_?r7(R)H2!XsMHE(TE(Vk8=9)Bzs%)md5hO`Y$LdB8Q^W_FhF?V%F2
zyi364<UG}i-$J-s1|P!Xk{6A4KsXxDR*qkQD7tj?_#QQMjmq{xhY3x(kCj~vBS!`u
zs6z#}e*=Dl&+>y7TzDHNt_C8Ru@gT$fH7L6jv$hcweVTv7s6B+ZZ?;}H2wOBPnzoe
z>Sb5B=DZuatmXk%=QI7~(^%{hKGABd_)iJU-=T#9O!KRx3!0#;>ee{y4T5ubWVG~H
zf_1VQ)}*kDet4PRwMII6bd8JpRGRDe9I{f|D_rQ9#Nw4lkg2)JWvfHUl4izYOk^xE
z$oL_=BBB=uMv9oJkYwPk@p{JX&PY>WdEI!FWDpy4pdjcX_bXPSVt$H}Vxua$-^0}W
z07O5~K^e(dzGzoElDNk*aH;VU*q5g%hpy}rS9GLTCwa%zU|YAct%d}^8E6V2QCOAM
z5z3odUO~N9s3l7I`=)A2rdb9IfsI3F>=-3EIIjvNb>nu*vbmo3TJEZ_h%gL=IV_n{
zeHC>J4lh}=vZltBuP}wH73C5oV7K?QUu&KCW|B4dQ?!X*w8}=c3gXIcaNSGG`0yKd
z;ru|(g;gD#>bToqPC5_5(1G>@XC8oKXl}@tgTE(uanpMqopa)12s*?#k9{v)vYneg
z;XCXKJIOyIorr8XCsip{Y;v|m-bdXf)atFMW;d;N1eQf1)gxf7G{jYv9yYqkkwOny
zhvmD(kT>wLc6x7nK8R5>kxx)QR-ZW6yQ0BMa3ttr_OlJ<VB%PI9AWIBa~U6v%a!N|
z2>~;#@5S#sHEPek)3k}79y_~EI6$X!feG>zCXk=$7C3RvAOVc(1G96T^i&yem(hS4
zY*>U<h!n$4$)czkz@3WD>dYClh;`DnI2xD9#JWS*TsAH)%kfgUP(ud0s-8jw*W>a)
z2A|vh!mA>V;D4#fsFBHOi#4lwC_^U<$eDN*fC?J<mfpTnO2;kFQ!TGxso2XJS7{O`
zg<<X#=ofy?`2BcSUz??Li@nHZ#k@5#$3Za7Sx|u+_v@oYYHy25n&D)}-FGYln>0I5
zPRG1uBk;>IFvPsOFqW_9{0dieIQ<?3n{{DSMAp^^#T|>_tV|&VRU}Ymq{`UVHV~$=
zD$=qhRb4yQZurbAnkDBq<z@3OQHq(^+55|-gA6vamWS$sOpm$9U_qnWM|W+3Wavc+
z>|kVu_*m;kXNB{p)xrwuvuZ58fl2YYC68tt16W$CvsNu?%C+en&%wMX!lg*P?48v9
z8&Gn<c1Qu{1k-^ATFu7Ot32VYE}9E^F33J4nTvxWEHQx)emUW<tcV&#6Zb177}fVW
zi)QH&UvGEuQoNXLRIx3sR{X}U2%TS>;W*~6`vUpr9N$)TKP|Nbzvvg2m!U@mpoN)l
zy(XbN9vYvU_}cfFy2{SNO4lunRDZ<J@~Qm#;h8*Lj*qfOxD*L(%;gjXO%GLjKFucB
z=$tF^1GQdBCmQ)xj=RA?v0DTD*HSY6FT=&H`=e{TB4RKYkRFdfWu(`kI8nEenv#NK
z-XUE#Gv^u0rR4mmi$40WQUAsl|CFW_Wx(2zY<(&MdSH?(kY18lnQa}Out}SdrAdx{
z&&@@tWOaeP&5R}bo2b&sNs81I)MHumYkT18emR}5&2d{QPfZ&uIUQ1~3cj#q#n#SK
zrO4>q)|NI!eRy+xnO;!AO<e<mM^&rhl=8oyWMxRA*pNUcnGts?Ld=Sa!$oBFbI2Jy
zGEY(foyScdl5h^yAPDPBL{k&wn{?SSB<TJQ)0~<^__}4xilVKfSM_e)Mdv2Vop>>`
z76JY^9mS=nm25akUQZ*sYM!T3*C3c?VC?=B=Onue+c#!8NBs~FT3U)+{G-hJeP~_<
z5~+1SqRE6d*6HCJ11;@!un_&p4*`1n?Ph{Ztfk+xlMW;2(=yKZc_hZ=_S=$ggye<p
zs_}r;Cpf;}y^YYz{X%dS;18SQHIYa@#C>hAAsh63QnHe=jR=Pi1<%A!&ydST|E#=-
zD@KdoS)g}KV-aQ?l)XV0GhYt(fG;%w8<;;%e+f^j5XF__pU|YpCZ)7l7~E1Rc#vIH
z;G0yHDnocohfGVPdvSt2FEl&saIn<qBo{aOf|(wFUcBiLQFg~HiET7nCxwnl&(9Y1
zUG?rHRi5D-dlClcvm;Sm$7s4;*`_-1S)Vjv67}j6*ga<l>^~70E|&jJU|2ZV|3?Jo
zTze#Hj~%&VNqv}ugHe{H8AXCr!5q{x-Dw>z+t@;HE%<os?E{P>RBOZ#t-c(>Q=w3a
zSa{E%KU_Ra)8WZHr>{7@?J!+0s~MA(37Hiq>k}8qTHy^Am4TYe6ey1qZb}bOCm?}2
za0n0IPigXc+QsThIqRz(qXvl8wa=|M9bwWT976Z4`y*fC?JuL|F{=)6EBP7bfxg6f
zftLDEHJ~v+&QHMDM$q5^!)jRr;d0)*+7i33CEAmA=ptrKJWn6Z6C9VW<9R}9MO(hI
zW|cm|;0nl?q(v<wOFCke9NW5b0ZrfT_Qhy>0F>wR{zy0ahYWMxM42=zh|{IOpgm4H
ztuV)Wi_M{YZfI48d4_qT^^K0l(L;_seMl}1p=WUBLhyId5~!5=3$5>ia=(`?H4^l<
zu|0d^GP3o{J;$PgefYq_I}beqmim2JM}#1#$YiklAf&+XgQWRAgo8Ep-fzq>yw0GK
z4$0n9$7`I=yJV}7h7J*1>Q5Wn&ScSYZVzY19}G-t6~oL6F>|5BvGp&P=-Kb7_zAiM
zZ!B@z(D*o$TIDxoULab?lppLEGZ-5!#z)6gP2(Ofibr<wR5_z>#cv$qg|QGf1eE!M
z^d~c%t&v87bV1)#?l={|5Df>>!q9w?ZK)#InM>S1e$EFFYF{Chtk&NxY6JarPy60m
z<t?JiFRQXuUG$x;22HEe1G`JZ{lOAWkwrJwB&mJ*apZRG_bVjG_Ib2$qbP(}_Mq0t
zf#pYlWAu$lYaJDG`+KPoEiv!!LfZghtgzxt{!O}>yyEsR3eusu8`~1qF|vDy)yB^z
zb2ywNYbi6Y-+k+4-yMF_yKZ<Hhc(<IO3qV@lU?l2?X2$JPIxay=4N~+eLFpkY@_G=
zKDP`Mrco7S?G5L>@n&zvs)R-+wac~ziYNuDK2_xrDm7par7AUw6$~`ng6iKZ9~Km^
zw5^)oP)b2cv6faS>ejR`+!Ku!D532IcJa30lb_RfH{@*Pv_E#IE%s$bg)SSGc$)A^
z)10AhoOH$_5Xp1-?5i+VH|Mt`v99)JaS+uAp9On@Rd-@wlv^0$PUsm8$sg$A)>`a+
z`^iIe&ebw`$P2!sM;cjy5>%OkS8=!VoCl?=31aq2vXb@;j!zBl1zgzv>M)`u=yWsR
zkW!I<(I)5PSmyRVxZz!%Km0a%S|4uzimBZ#Kg{2x`Q?`G+ct3l(I}~QobO0xq<31+
z?U#}sWYE#nw{vIiHymrP!QLqAkJE53JObh1u3$a9EAT#uT<0J(#c^D$B2bdWWm6|+
zq87505_%5B-^|pZN)OAlFJXz&ZNZadp#|8HI29?%*m1UL?dN2ztkzHsI&vudb#d$J
z%Ei)xHu(2WncNeK*{+wISrgp^2qiTj`_|dlsA?$Y6N4^d<A(RmL1r>)(&Fey_7BHw
zjfb{c?la>gq|t_pner1i<nIpGX8NAaSK0!dd9Z8<jy?W1JH<u}7|bs5t4!ND{Geh3
zf*0c+VnKzuOz}xqW)~8`gy8))x;grU=;N~&-~whcuf{>~xj&I0t!jJRwJllCT}<{k
zl8i0L(<}Rb1+MQW=$OD?=fp#qSR37cpZ_G;<5Z6yk1Qwc-jbzrG#tY#Ss<&`lCHnQ
z_@ck~l>K>4q=vMrny?$G8JW#LuIf$Iup+$0*+9Aht=|OtC?P0N`{g3m%+4j7*Rj!%
zoZ^`}A0o6hKPotlM->S6{98<BH*yB!C<oLLU-3Jt-(dlrUe*Ii`36KhpcuL=2y2@_
z&o2#y3r}s>=OiMwN}cj*T+<6>U>a;vSv@Wvh?H3Ol#VFYEhqjFDO=nP>xgO*`ecw6
zc%KeuBOHWIkF2C#nxj-PJRiC4)RhI`7(Pn&Vd0wE&Qrv%M-PuFS&KAba6wX{u}jS=
ziVfm3KOF2%9UU-;cNdYX?wDfl7tklxX>yGH*v~-InM5omEgMw_*%k)MQDSLOQf|Rd
z^kunv_UKPlxaDstG1|o;bJvI58_0z^W7$eDTBN7gD6`zE>M&|j!>@)};*4F_Xti+J
zks{2#`qQMw5YMWu9O+Z@RsR--$tv!xD2zU&5)B9}uh9ABTmXcKK=m?Be?pk=*ru_M
z{FUpp#^u+pz3y2DsH#>XoRbO_hM&a5=JLrruv_(@0M~nj&{F#J?YZOj{&_>SF@oe|
zC`fI%p?x1{T$?3>qRB7jDV<!o)Ty6ucey7-SW<}yBj(zapL*TWLmrTVd5@A%5lG&r
zqp)G;ncNWJnhce#^oBM~W?ZYycK410vSK+2N4;trMIu+kR_jTwM87+zN92SY!pq?$
z8>6b3U1&;4t8IxA8{DsLZBlYRS{b2Pl1weDI!ISG0P!7lL3));Zvzp8#Ua@Ys=&n%
zD%G81+&7~3_?^~lzkGj8Pul|dlLD$)=Y3RVt4Gv!rW+6Yx=y9lebYtT{ffat?cew%
zLtupdsAs}SN=VYbAGMt<FZF;q@aGlvc2JLG{M4(vbiYNm@J%j;IF}42+1sG-on0gI
z6JV^Z7!QB*;6ErOD_n|GVkb^4NLXr4NLcaT;%wGHO!Slph4wT1<SZ#0z%@s*2XuC~
zm_PeDD2|PDJxF88;b8VUZg9pYbU93#B^HK}ExQ%M^G0}}YFyqK)&z;m8P<1ykjq`H
z1f-l+?wl9$w}&cpLRP_cM_)u*M8qQQ=8x=+mZITQB|WNab(V0B7Fqhl$$8!*aFba<
zz#3`f!qAwn2EyO>gSxNOsP*=p4_xf7)K`F(`2mzs=||M`P+|h0sE2!Pd51K3iHL%i
z&Y=b27eIy9<<G~_*Uty+yO?)%aQ2Jc)wD{#=>Smg>@MtK96gTO^o3>{{YO6ut|$R~
zawN~;Fq+=Nr>t_>^LLu9LBIYBf1G$z$&clI^{z2%(Lel<7r+mF7(I`6@d-n=PAo5X
zZ$C+flRlD_5#7r4eabhtBQAfgEWIVcdNFybx2oJuV`zs3Ns168=Vd@uR@Z&4?jRWg
zGNZW7sZMI3)b4gz$!zk{<d#b|VLlRNkT&<yatqYaw!D=7f=G4V=B=YO6!`;)dcibE
zCKGdlM!Q6G74)3ZOI2Y==SfyMuc~vYqD8peyQA*ujUZ^L(Ct66vj3D;V*a19GTG=q
zgIkx?3f5ny8FB5Nz`<ZG*nb0cpsjz4QC_{9#a)s9^a{q6ENX9*m?=63cbnFtiAQ}l
zKDf3Ok%sq%w|nw<+<v}Tk00!Tw(c(7eqIH-3oPg&JkDeD2SrrtyL$6{y3v&~I(`y>
zlqzMzi<;!~OnYs@=1ug0<lg$I6viAg7loXDHhA-<PX3AWp~hE$?%?vZk@h%*3{L`Y
zwLq=AfA{0`cFNiAjDMQKFgrWz@}qN=z(+I}4Zh2>D||`Oe_HnodPEIykoN~>=lVNy
z6XP%1z-W`Z$J;zf21AQkOu@t22?o9WD<02h7PQ44gwLY9&0_^Zkf-hHZS|ktRveHU
zpuh}tzZ!wE+>8rwD!vE}Rb#r`I((wHH#ks#-RSzIBnk)ZDm;wcJJfH_haXT%)DEW!
zdCc9aL9zmR{&cP~W-CcLe{I_?x3py-<aUvu)?319C*RkgTpX8)`VGCL2!8jyC~3$`
zBwZgf`kKBr1ID0h+v_3&CtN1;t$*Muw0a<_2!56=N4|`>ba~B!_t2`4{4h(?ia+w!
zOeiqjn<)%hkcXYek;;bQeOdeL$;BkYXA<(MnZ;-HyZk-HGC_+ZZkOBg{p6^jud_Uq
zKT;3Z?&EqtFg@+fuDtaiwM%nYA@vyc*Y*)mYUfu`Ml3wAFBhm&deM|3!1EtJ+XzJ&
ze^jrhNeEFWw1LJZ$R+lhg0N^BHXQVYPluA>A5K#mEbtbAsA<6xo}Bn6HiBe+4MNKn
zyOi&@TJon+^q@V}74zZtYkKatZ9~4fF=HNiekxA%P%;@(pmP_g--p4^O;FHsynzzU
zMqJFNxc&}yXq%RHGeRmW`jd2Rh8OAMMB*nt0LW(9M~~&2fnf}mK&jG4+{d?+S2047
zvxXbTXGPJs$<;#(W_{DX=-3cSMglq65tf2u+1%v7mUc$+ql}4oOKuv79S@@z8xOne
z_DRvTO$?(R+gMz&(zLVk$|(Ail*%|)--jBnR=x&YSFE-@1wcz_%)_c37wWA*Dgh1G
z;hwg-F=r)J;v-C3U)U)Ih%_nxx&{{aqz_HXPdixiE22MhVVr--EyQJ`$Y2&vwtKaR
zz|b766i}4?@_ci0wQ?fk%yk64K-07a<Ve#b8W~B`gpy~)m-dAL^JO{{x>wWP=i&us
z@<9--TWE3Cvs;Lv-C;=@(=ir4-s-y=SGYsU(+{f~GEtRQ$$BnWtGv0{+~{Oe3f>oK
z3gNV16#*Cxg!={LQJ70%A_}{*3JrTbr-d<z@Hu<3wcooaZ-Q(>8ju}q)!IcdTn0SU
zxl!EZPAi*{4R(uD)E1d{t9=H1XQbt){BUHkH9qF#87o6gQ8?U-EIP+9pfJ#7+_w(L
zCK5-#2-n|WvWf6K=^OTQ5$g@ETc#CNm3ESGH@N*Y=a+{WDO~Q{3N+B`bub|^k(pEO
zp-Pj4%LL(nAyfn5)TtaxGopi;egocoW~Q%4wc`nGkWr*Xhnf+>t79@D^3kwPBKwjW
z?w|3q0O-&vLYaY%KRlX{noKLom@M#stx?I;eZ0!vSFW`d!Cz?;?dd1PIn7d<xE2ti
zhQvLY@_rz5ZkkLr)mk_7`{!bI`tO(edT}zz^HJ-<k_QQmXvUGzdIWJv8df1k!+SK3
z2@)M~1<0g29mM&bf{D0v#P)vlTu<GUC6%UT(Go9jg*GNwxgE`#M_c4*kVGr!`p&W(
z&v|fGJo65IQ*OHHw4|8T)>H{guU7jG7b#$`72`b5&X(-PV(WF8d8<!9J*>rf&$rRI
zp!f&!I6s|i6gg*uXs)Xa1=;!#Z0KQ=OwYa5)9V+~$z@rWq~AMK6MwR9H)*-i#h8c?
z>uB)?=Y2HYMH+}sI#Xr|y)NFhdm5qgkJ`>ss)dA&?xGxI<LC(<-`q|m;$+GRD`3>!
zhnFXLd0eQIgPs*m(SyUWagu{6z!u%QL8`yLT;aO>SdTQUSrOL~P^ja)^v`kF<FYX?
zEyvls*Ch>fW%@P+n<%$<26rk~B&1#MFuGGBaA!DgX=wZBi0OK6z!Ak<L`OxmlchVP
z_7|QAiw1C$R7=6@tUu{nl3OEzIHjgi-pn3-{j9kbZ#_N6a<8zh8J_3(Pe9U!nM+Tx
zg*55f+XhIdi8L!C^0r&m+p`ae@-z85bgp5ivH7wLREmc%*w|05Q8FCXpB>}IJyXC+
zzYb;}-KIMhQjsRA-H{r?bGb>U6<qBPA{x+WmZNR?t)s%;^9RkTFa@CZ?Q4A*GTYcs
zcQFp?4fuhhm#AnRR1CyZ<8Thz7B~NFJilrZ-WtO7uv#VwOwL~!8D@+N`w&YT@wzOo
zbjX)6C6HcHd0C&?81WXHect_M?_1AoR&TlF2>}ATLC+7`1FLW%l6~9?VnyH$tUhE{
zTwVon38vu*@4=AT{xzvZ4{Q?pbYD0RKLbr_r+ibP^*VTVNNcm;zDymQYuK^e4`A`-
zekei~d=7LpkNgRykqV}h;-Q{%@aGz+jDrOaFLn8Xx$S%eJi0^}z$JldA7}@}jn(Ag
zXK_CiN9bK6E6--na!;eGHN+l?FmA9d`<^#v9E8+zZ6oKSl!3j!1Cnjtb|L;a4%Kf3
zHO~taT%xqeY!kGxd{nsEaK_fSp%b!GiqtHLM)E%!U3%h4!!P->M9^ls3pamG^A|Z*
z@P=nq&3QtB_@BHf`IFpQQa{>0l5pC;GH_6KCf3#PKc`e~%%6m(wquB$;<z&??}w3}
zx@YbU(^#nBo(tAtyGN<oV5Hf}VFoXG)C=Ao>9!NOU*3Loa*1U4@7#oyk(u$oWsF!D
zIsRuuUJJOAal6gey4BN=*bp%qB;2UEN{QV1R``#W`f>)*3|an%JC{5tU;?xS&1x;4
z3X-CJp&}QsP+rf6fqI=C?Oxu`W1#Z`%*BC;$(LPvU>jRc>Nf4BcLOjza3N!lU+KF%
z_O-F!+IPfl;e80gyYReDPcabT#Y1qs`tQS4iK8Gx34FUQ@4T@wvKISvTwxKry(43|
zyi5zwaG}VBkkUG<ySzS+1I6#ay~1~1-(SbN^j5$z&r`2=ZdM5x9E5$pZ|#ZNK`7~b
zybbL?CU+S_K{J~~d2Mf6qe%>L`m#cF(fPjZt9w23Y`j53`a36o2IP8A+f24UBm-7b
z3%YWZw4sRETEncQ@q?XUZGu{3$@vD_bHDo{%4v0y{dWE`U?XWLGHYt+6Z9l#_mebH
zY>=|OiZ+r|jS}UOn5F?sY+51^O#RUXuR0WoJ^HK#Mf$oansIK<mR>*g<{E`)hVz)j
zO5cNgO<tpYF2!8nf=Qd)5Cr+Dq}$pfY<Czg$;;Wa<I7y25y}xI88`yX$n+N0j+;-U
z^bT;uw@2J6!ws?;<q}J<+5pXYaFW#e53J<^f_C7wpTjV7KMG8ugp>AgYD4EYk^8J;
z&Da@feu^T%DySlgftp9ZL0%KPdY2b+hNvZ+QPf!nQTR3T=H*x=y_*XeqJEmx5lFJg
z8V`WiD>W7fu1d>0jtlTySYU%Sm6MamyeWL|iZY)g)|uhe!BUG;mpj@pQ}MAO5JNui
z)4|Ag7|AP$<^D~Ob>J9^DU!Za(C_Nynb^bH?Ar=zQcb*#ZsQhVSWX*GMA>5s!IN0a
zW?R%u?|5ocYP%K7CYgrsY60G)`>i&Q6sZYr1|8FICB^|<-HcfzOC7hAaYXH$@0Qp-
zyrxr}*&q7QSI>WSLH<rOy&{vF*TBqpK6K?+433o1H%aju9HJo%qHaqT{+Bq-giSYA
zF3&>Z=z6fy<|scWrI0u$dTf_cyj1F<;T_+gv24TF*7C33=}ioi5Z0?Zzh-Y|Fykg{
zzG{K>@>B>yx=IpPoGu&5l^5E>eWeouWs>{Wc$WlS+x(M~^myZ&G-F`rCM;bjKOfT-
zf0}_N7!hzlWoT034O5+~ul^{f!wg=JnkbLFCjz`6bQ7H+bSy~x8-7xt8{-e2tKR&@
zodPt}eC+sVAAFc;E-29zj;%0xv9D^n1KHP6GDK{e$QiMl+1v<JsP!8SERY15%l9w7
zZWM_Ft~yOObHtP8@UApVHBxy%eWa&3q8EmEV8}F8=#n+xpjierv7%{yP_zt-%OFXw
zYowd-&csq4$K*3f9+Sv_LZ`&PI2o}*b}{?}&Ea5MY$!+CX*rDNTuJ?vRgK-TaVvRi
zn<ekQy6u>DxO(;waZz#yY$YFGt)X#CQG)wkJ=o11V_n%L+!S!_C0&Msi*JmY^69we
zN+8siB9?o)t+MMX4^y&Pl$ka?jHt-XHsBtJ&V%aK^UJlU@aN$t?@|$&@qE#@PlTNn
z1yIVH@>rhrUsw=?7?M{8ZS3$VOzF|KmN$y_P1F>;gEtZq{X1}_aJkB(I;82Yk>oJ}
zh#I+#lG_{lq^_T&-sQAz3T9f!3{T(ofk5AXr5f|O{ptdL|M7v|$4kcDE?3E-Q!l$}
zir#EY=PS<vV+a?aOAK@m0%5I@;H2+}F}gD{nn?8f*ZESTQoOLF$SU+F@XCT;+~CCW
z{ngp3MnZsw56zCx_m?W7EfkO0V(M|Mb^^~%YLH3|3mh2Eo|~uJfF)CMQJ(jfci)aW
znoKA&H|g|xVR2lNBuPS;7OBSfMO3PHL(}$#Nec+3&MgP?OsopAPfqYA^y*L7o=U+Z
z0~jmmayh&T4b)_(yQ?Xd5JA^P3ZyN(zM)ptYH8v)mwEDQS}m$brjvl94@6*xMVdg(
zlo|q3dLhXoMuDHkT%-)Dq`bpp5BlZvh8Ex2joswJ+~n*Gb0}B%a*-&>v56VO7CDQy
z7N-41o#P^9b=mp~4#uzaqH>jCY<V~~K~*fzTNggO(dSykuN%+x*#o6hk0g0iS(Syw
zsUs=fbz?29&a}{C2ESLMF<#+!;gHb^4=&T8QmMFYn2Y7A9YJH4kYuqSwc@0(Wn=I$
zsONlxT&N_QBMNqKbu2ovD*thjz{M0HBjI&@`t5de?*LOU(aN%Qf=MY~BE5m92)NCd
z0k$ks4WkBKl}dU>gu}60>w4^og7UlL{vmS(Q%J5Uku*OZ7}Y?AHDsTP@rM@lo^b<X
z8)9R+8GY^mri7q4ejR!)nfXwoTPj;j??8VmNduFN+fg87S9m74YN3F5V12okT83t~
zQTP*uLVl}~0w@vVwq-cllEE4lV}fXrIE$=!9Wrj25rSpgm4uPd*@i$~kk;~>pnK#S
zE4&y}RJet0%68BqTuc|sngo8%<XV=^7QTAeB2kPMBfh$p4%I%>4?h?nwQ6gf$&y%;
z5-GZded=WR$no=)J#ZGokkjXYvfDygM7CQP?vS}}ICXD}`SyMHKwp7UkPm3R=BL%J
z`f&Sn^;N=RCncAcrq#Y%!=4q2X|oZ0^ruudGY9eov5vm-W{-eR<OQKLNxkSBz!7ke
z59}P(egSG};ab{D{+9lhFLt&?HQK6XQriI1m;~{9F3h%$b&w0u{2i~XC)P7vCoTq6
zHM0?!im{*@%yM0d^k={|3QZ_6#+Se$Gfn~(3&gS;DTzql>@d*!(<*X}nZ6R`#A8Bb
zEfl3hv7?7{4#`6OCIg1P207C)IOC`<^Qvw5?9z&;BTnFX;StWm#-J+Q4YZ;?=E-fv
zmJLI#(x)j{D2wllo)+mNs+RTkTE@ZW&8!W%Fq7yT*eBw5Y|i4>U#w)S9EUB;1WD4M
zCz-z$5dI9Sv+Q$XG4e4K9EN|ej2oJd7Cz(V=C?+)9t;%%-Pm@RX7xqN^_tn?w@jxr
zN9lETcxGKKm{GQMTw|}Kn~JU$i;<bhB(9TbZ!e@77(=y2S>mQ779L?7(sTO~K*lRJ
zvYXFpTox*BWVX-%F5RG>rpaa%^SGKXyN#7I&l+<eB{$1#V#MPnp8(5j5obz>2W^kG
zA~207^1)hSt*&Hme_KY2z-1oZH!cmKmV1bThw&Wcig#Q0&D{W}Y2hu1S{K&Vn2c$_
zw6uj{SviFxJn^=O@Ir-LN<qd?d!#=B9M?qO>3w2$^Kqy-ZKL%ph4RiTq*kRlj!%u5
zZvXe&b$5lj`7?@J&j;8CP%<26Mu{O;fu^V|p)wuM!ZM0fELL!05^@O@a8)|{Jrbw<
zwP4smIgLEAVwnK<Liywq_gY#Ei@5Zcgjd_Q=`{cIj<b%_;%cG~+K+i1VD~%<u$91_
z0V#0cGLF*8vw)|FO9qY{BswR}?8Mpyg=CgST%_}4KW(p)!>ky;-vyLzrP$reLYr2=
zWOY(bB3Wj`lE<k{Rq5Vy##OP);&b58<C3q($Gz?z?=EmH;dtysf}S{wMK^=*!#gd`
zMb6eH*f{>aMLQYD&@ibi9~Ob-x%7>SP020kSWzm$@p<f=Q)sPtMU_Q(I%AR>RxE9o
zr0jf$MrQ5~D7%S4^g9ZqDI}2))U1dIoMWh&^DGvO@jg^U&Y7eHHSt{Cre<8{P|8fj
zsosz54N7V^gA<DT(W#2`DHs&LUez-2YV+!{<y2hgV2NHb#Imm1*S0$0PqD@zwk2X9
zb`I_><t-rH=f>TII4X2Tw0SH-u+nBmJ}!r&S6vbIN~QFVsJK_gSq=2F)+WDp=!j}W
zc)c}CoBzO3Dp}#%ESsJ9&bzB%DGwvh57R}Mj<w5H+aX9;vSx0`bE@4#pOoUX)7)%k
z^7C}9%)k6dfD*WMH63yrZ9fuHT{I?#$=KF?^3&C84{3oqk<V^tMh%MQI*z#g!sSL%
zXC7Zjn)VnjkXQ**KgI?qEgCYZWAOAChz0F({gD4Os%oXH3fv7EV^mQbvjweRYDTn%
za63M#$}UZk?E;8i{&oa%*^(!|g!{3~3z{->T)9(<TpN0(Uwv{A#lyuDUpY*J*tm%)
z9Gc;|-KjA0E^_u>4^Ure=s$%MES&!>oM2|+V*VcvKucTivLkjLs+Tm;Pup<yaUfz!
zw5Nl+nWwMn%Ud=jU*!Io#($MtY#Ey;OF?C4b0)D9@^+9uGn7jTm<xY>F*ngSH}NV?
z#M4&+;=ew(1aAsi+|_NCua*h?T}$BcuJYAmi0kJI@3gobJ(wrld*fZDQxga+9`NIx
z?|NlyOcaI$LC&W8r?<DZY2}V?I|6z@n{*PF=UXg9^A-CD%yjK{C%r7M&h66I46bwU
zm+F0Qf^4gg7EO3D^Pc9~^~+7GH+U)^=x{vKgMf?9&)=RM&qijBFp1!PVdj<daMW7?
z*`%jY&|Ne_g{G)s+;nEkVL{bU)RO**tOBKTpI+Jd*69zWwUKwPm%IFh&iJx&!*jY-
zjOPy%_-1|lJ9Z6+0-KD_{9$l}%suzWSfbM!fEi*Pp)<hj8B?>DB&M(f+CZ<j7#r_`
zuU<t4;Fd2OxX+fHUoJl`yl$sW&a^-_S4E9+tFR96W=jDt3ZyCYGSRO)p#<^?e^(1}
ztnWuJPI6K|E=<`8ilBku6e0esX<NYcBz0x}Ey9WPdt0)=EeO-Q;4oSWHJ)HLi-a{`
zhF*TA<q-*QYWEu$$k#z@Z+wE<*+O~Zwx$63dsuXnUV#Hd6oTC!8yQly+X$eC{+ad@
zD7E7A{|9656r)+Rg^PCC)?c=5+qP}nt}ffQZQHh8)n(i4I-Q%`a~}47*vXvhWxcG-
zteJdcj1T5dxL=xZKBEPE%`rAd3h}TM?XwOR0ca?xPzUk+M?AOH94j*w^Lf@;%I>^f
zli5;#d{{31@m*j}$d)#Q`I5tmOTIKlldSw}A4N2!ZC%LnemNVYxv}4cz=Z1H>__D=
z>@9ZRx5M)mG*`rjUd9(B`^wXeRqV`n5UOWSRTwy(X~sp$h0%GSDrLuDZ0uv%VDQS>
z$agsan=%-Zr8#7@{wa&L1j)X|Csj=3G)~s#+6>4Z_&k)qRYLLXjwmI2MCb3$iYk_o
z$sX8wPG<=1c)s-ay-`62Y${sw-q-<kbaT@dkR82>UqLvX0#)KkfD@Epw%Iy^HzY$X
zGiTGRMdgd44S`560m}%D%>DWxU?*gO4<7V`1>q@ikO%i*I5*Db&+QIZ4Re*@5H?v0
zbv@f&kg2ec%b$UnN=A~jAp3XABXV@l6&5`=hB*YA82%L;HnPzx_Z@gIOc!r#o{Cyz
z6-!~Bj1GT<`)-RYuoeW#iZx55{>nG~Zp6{#Er-<ru*z?dP{Jl%NW}WcMTBHZLewc3
z(I)lDR)4gU+K=g>Q#U4d<I3Vm?vzZJA0!MOG21q<zAvNr^X3y!jj(Nu`NQxaOWJ~I
zJ*P8i+*2@Z8H|uBi)9Z|MX=;Zm3#5F3o)vS6ShVNEJad}UQ3PZYm--_2l}u}T?#(u
zYMX<?QIDUKMaRoAz&Y`HZR(9)j-?IEh+wXQvq#BLx8O<Eb#%6B%HivjHd&Hn>dcUY
zz*fGvgbH#&Z?VBL^*1l6*H*LcPitgt>Q@qBlkXv>T?R_xlzj_B8F*vT%#~Q>nYR}Q
z?%E5J&SW5+P=!=nIGXGsDUPt_I7^<VGq#EvUQkGfwXf?1boQc8LN}=lbFw)m4>G~r
z)Eu}ZL@uGuG)n%d9vPMVgPG8Vw*WZBvSq^iIQDDznyBzoz^wO3PO@Ff>pKlToT}uM
zlxt2su+^-aY1~EK9n@j^cT#D~OCGE|!BJWrl5|LNOy<WeFcQzklLjS4<Cd9H!70%y
zvZe9MCM|V&lJF%fkLsU|MweF#_-S8sUCZRrcXD1RdKgH0$}Cff$N#><#8WK4mi!mm
zh(VS~MFJ4}oF;uWzCJiR(y`*!a7DoH<`LH+>nTP;PLRvp)L;73$a~AgA(EDxM%GG|
z2!YRggXL6PB&05|%vD+=TJYA*I?rcJ<$}qjwhACq^9DdRYZ@3=N&OBmH`00-H2SF3
zuLhx*JkWP+(|e&%6bR*`Yo8YdwZh?aR&8X6g-bFNnm9T?WhH}rPNUR+Hl#y*%`okr
zt97ZW-MvC(*e(v(Od2C;cn@cc>Ml9qyYWqQ7^cRU1|6;!8Z^&*7tK^%D&4$8<;dhv
zHs>cK$U=JE2CDpIFsogXyKL(h10{_&(L3+_OfC~s24U|RWwbwrRBaom`PSdx25-PA
zR#cwh=MW`BEk&^LC^Pkgd}li)pYNE;q4;AgT&y{z45T{RwyG!bjd&!GyZ&apor_#4
zAq=c~^AAx)E61}{BWItxk+#><5S%`L-GM$CzgU{Le;t-ROPKM1I)*Nd3h$+-D(EdH
zoIjYCS_=xJrvTOd?XyK&<mq1#F!^<&_}m7RE>}p)mXnu@1Yw0LNGP1*)PvRNsvZy(
z*%3kAk7-t4sd5*1;`0~aKY@Z-_SuA|H+0b5>qu7pY7~mgHjA`bKV)UrOl2m%#O3}i
z!3=pp@t8^1^II{RVKek7gv<l;n!SmYPACMh8YMiJd55e0079FQQupJyPm5_W@=-dh
z+#Sgd)?tHYziM<ywTpS*j8`@ZbIRr^2W3`MGvRDoHq&A}OF&uJZyQu1ax%YS(pkfW
zz#bKOsyz@jmu%xjs6dBt1)kF-8m98eH$)ko*$Y<&S_0G53wj@D4*tA;#Go$2?K$hc
z&z42ujgrewy*88Ug|_>U;tf`3<1&7vvHRrSHy&0W?L;lOExWseUbte9`(*Bwx98x=
z=9@*}i#O_&nm0&pJ+$fv;}@O^tUqiZcwLX7%oipib~D&TFvJaaU2Kk$I=Wloh?-zr
zUM(IaiO{J~_<6I>zy5Jq#+M1dhb)3E&=7nHIWgybVoO2%;wzuab1W8-?MSrVt#2CP
zdoumKSBQd^;di@N$Opj>s~s5I1vugrrp2w5=QqFGZhJ%TfU-oCS|8B>H$E}SFJLyT
zZHC`;*|FN9y1Zi=9sCV*nogwH4e%(qgJgwB?;B<HT^c#yzf?1(`ETfkh5bL#4HNtS
zs8ZCHj@xE~?Y*sY_&3N^6^-x$iUKkt|2N2O+0mX?)=xW`*ZFvr?y=oCT#Ig_ogW(3
z(=McVI3$mB@_6+g>xx5|t3&wHD{%$e1RZ^*_fMjz=kS5S!Mj|^{7XjqOpo6^FP8%&
z({Jucc($ZCufSVc-PJ4EpaLxW9HLA-D1K8~kU#%3>W8}xrbpF>dR8#dMP8DrEYIFD
zK$<?O_*V;<_s98j7ewyzzcfW~OiLA)5<H1c!e_C>$}_!pZ*MOVL_LTcmgRS-MW5vG
zZy#+i!{I8m#!Srl>+!Gd`6(vr`s)&zRUyH}^k%`vOeae+dKZ7n8kfu@8U97>ir+YY
zemJK&)?=VdYtA@?KQt>$UPEX&{XT}o7k37eZw5KY&Qy!M@6yBhz{Bz2*q*rVUF-|}
z{KR(2nRpSHL(ZT`<>KKOTwYj>rHnKF#jW(k!xdw#=RfmZ6yuX<n|(8A;x0A?z{d%v
zn3u>6R$-ckaHshcWFo0z%08RjYLjSM46=x8tOhzhNR5iE@|}m6<BN|@jdZ=E7mt-#
zj^Vw(Ov-foDSUC5ubmk)hf2XfW&De5^T2>;wgOF1OmeLkG{F)#gFk$h$r`5uE@RJ)
zsh6<lHEi$d5agN?pzAbTB;#GGV*w3Sbm89o*n@Q8hcaYxjEl*`cGXqg60fR%uteo1
zoeWNphe8i3drN*J1sZKK_M#e1NZ-v?{*-npWB|$SAJ1>N#x0s+E<;{VKlU5e1h%#X
z)tJTE6w7PW5Mum|1?WoWPtenFkt5XW?xmttHhqsoO~YLmkELLG_UIiz<891xStPkA
zHh|Q)+R{D{&A+&uG823(Y9>VUjChRXnxQ`zV;+;`{!k(kh~Lv{>sc09GZ3`@X)n&h
zP9Gg1<(G7Z#u82?Nj#1y`cSHz?qu18yMj-5a$;#pP6R6s*E1rnrH#MLua3vO5i9!V
z`KXEPG_th(Rx<LvL<(wUKUUtom54>$i=oy=Um<!VMsE?H93jZ9?vYo0-i)zFkDt}Y
z*Uk6E32kx=>fHSJ_Wl^;{~I}-qzlIDQswn{?WQg4R=eWcSJnYi6oWMEJDxa9LzG5F
zJ%4@+PR8m@XRD|Jvr~VDCG2-r5>&kVbKbAgfh7Ek+)%{x)8+CEwM{n=l70rDN?TuC
z)vXo5gSO`nihK&piYBU(_1es&X%3w#EK+rmVB(#doZ!UhDF<Jve&tv;g>JhK5u@ww
zmBaH7KFI;heBDwaYeFe$56g3Xbg_%N4%Vh?Yg6K8W4rCcNxfJc7N3f4t&(8Rk2ZWe
z1JH_e)J{`S0t+c$W?~A^QqLXe=u(u8eAWnma4TjjDDPnYg;=ULmyjU0w#|M$Tz4SI
zGw7bnV(EFU(iH3jmJkvFH_6m^(xjdthtwPL?rB_*L_s3{@8K@}vt#$=bT`o*{etsT
zCfM7*XMRt7`iL@w_@LAW4B05t+(x>WiR)BCk{BAsP2sMY1<KLVtn{e{7b?{5=xe#b
zq)=_Btc`bt(VW0<Is>JfScVXE=T^!!*V*-oX!ryKStflq(rg9a6c^gvEt<nz0Yt~G
z=rec$tqJvozoH^rV@6Zs`Q`27QK+ZL=3FBnMd+vGMTf^{_G`|Yyi6~AX|7La(eQ>B
zDd3WvFeDFaOvq`klWRK2HweW~_WtUG;iMxSl%8!+8ReKQI+``}gkBa45AKFv%tmbT
z)bN&_zDodkKe4}SiwAv{wHVM@r!%tCYGsL52i0O!MSBrrB}Y2(>pi$;STG=GoU1>F
zPF#kfwjhfUj@_XiF*gTmzitU#6-i2&WJFlb0<ln*6bqDBg;sARS9`2)8dW~ELdY_2
z=`q!}4LEX^kfow=m}zQ3P(JU7HnI4Ymh|w?L7M|AwF{lAf>XRSXB_+tH-;J`bqplh
ziG?Wl4sLyWi1W(ysB2Dy9IB(sK^c7X=Pa}lg;KWge9yGX=|;@Z=3;#nv2^P^WD-UA
za)Qi?y*A7FhRuFsiPY$awfwHltkZ^g!S)*{2(_Uf5rk(=)pTg%aJlweh*sV~g!Zhq
zmE$#asd1*#^+db86}K8tFu}r|TM~aNX>AQo+WD)~ACV|Cv}lBo?6RqL1Kw>RU&Zhn
zV@tTd$C0_bRB|BsZt3}yZf#D&-S|Q!nYo8{SvDP-NW84=U7>FvVY!7sSIuC{!iN&#
zT=0Eh-@a1&1Ptd<QOigs3%)d-U?Q85PaVLui5U5D^8JzJUI8QLexEl8>+SPgt3Rz3
z8}YrK69<JlP5chH5n%#f#XX-oM*<ytr{`mOQ{@Sv16v>kxy~IA{0(ZWSIn8tLSpMm
z3lD4BVOO|{vApIhq7>I9RsH6|BkoBv8pjkO9TQYFMyyWE6aFNK7w0j&t&W*>=;e^8
z-f0(mnZHe<3cZO9^GshfSuxb;eF9R$rnrF_qEy&69?ffAYclE=S2NwtFK`NyHzxBJ
zU<QiWYX@L9em6Qz9J07G46p2xO6mL4jIH7HT6_`%#%9Jl)<}W@>opdXLi#g61B1gL
z@QFX&sP)kXrH0?i@JhMnvr1}eoz5MZO>mp?wDda8^?~`cF2T=E;LlemBZH1mfDerg
zzA7uSSirgYmy>o~YK-R*plvx8+-+8wT#`b6eqv1RvtDORwhG?=oJ<{hEi>HbTF9Uc
z{a{sH19|Pf4NRaiT^0uv+$aNJ#OE2KJN!ItGhmPQTxWC>9wMO;S@iGdZRO0vt67X<
zK^C25nvep$et7~8G?B516O3P$q4GK0+$2n|SgDxmv1)XLfWPmHJ*G!{=<Qqn)rv8g
z+R(gp3ZsF{_=g`{S}1|!nXIMa8%2`=G0t+kpo{L83B@x~ONC^1>iD`|kk_Y;1tY7p
zYxYDAaT;vQd*7K$TA}ivOlo)EZ#->x-%nzDwnE!=YaAhuSOtie8EQFia~6AYT;&Pj
zh&f>8$?vX8*-+;1pWnNT5MCp#muE$6f5vgU8n{1d%R2tTv8sM|{x`hzPdfP@t`2M*
z4F6*{=>L}^SlF>KjGQ2E2M7VRxUlVQt-CV&WH$IZ;*X}ie?#3QGm$Dzo48`$38qj*
z6bB;F<Kb6?$o0v)d^o*+xxJo~7%UD;l)gW`Yz^Omj`T$z2$GV|02Bjd?e)v@@F|X=
z+swBEln5P`+U@Vu`F3H<>Os8+V|aJ14H~-l+8MX~9e>K4dRt-5-;?YG*TdGCMrKl$
zqoLQxS_kWCn&SGpza7Mj-eLVUpFiKPZ~EM);2mlpQI`7pb~_FU07#+o0$_|{a{hn0
zzTIBywuP)RkPlz(ZqqcALKSS3C`0--Tize(T|N+Uzkpz1tV@613-xs!rK+2k@R#J*
z`D$?O09@Hrq{q|yGRTswY<3Y2Wr0J-=l8jom5X#`0bK#J@$EpJs0^S1@Suk;__t`L
z6`;AMbpxiVVP<xTgUHY?k$l&0&;blK{!O~G;ihIAH$ZukL9~^uc_F$8RCtDG{lmaK
zq<Bf~^gF$=ej$6>{yJfeLSXn^KOM2pE<Ck(JxjVR$`u!bgX%)0{D+*_lU=1cy8e_r
zqjzb!N)*;ZJ%mJhIwREQZ`Q(78wy&wE5C=xrUDx(Dl<L@*bAtO3S36rXFU#aDxGjZ
z$rHN_tbv8fII=PE;DV1V@oEp0cCpQ=*E06-EF%XhFRY<AV=)>gQ^~vFHXF$RU54?6
zyWV+n0q^3~tBW*>HJH=g?>qc`asTp@{&_T<u>QUGok{rwozE-Z0wn-7OobzoE`YRX
zl1=49&;jV-z`qUgwCHyG=DUQ`4qsW4&EEXIeuIP1s~s^oIC`79Vsl(vJ7PDas&x4z
zvSYMI&QTdspdRSeBSQrjsYDnhk4`n)@ne``&!AOCl2G7S0=arb;zCekWAviVIW@GJ
zIIGE40bZ1lR1GPLUd{vB$`EZpx|#xr2^mIAH0CJINX5QVbllw7cFATE$b_}aYRRF7
zgSUw?mK|&h*^+1HnU)i*r)oCTP5ZRspqdMlh$iK(i;DS!y2lp;ec=Vc2B9c`%s{ca
zHQ(M|#KNd5*ct1TQ!ePRk*l-r#R1Ngw{9~zP4~92XR<Oz*gZ3Jum+ZKpKuu_@*~Fr
zB((}xN;+P+&xZ@{CZhdw05oRzWz`2hP0ch!M-Ad3R=Xb*Go5?OxG%UIg6SmDzfk~2
z<$kxvwSWKpFp=EvW23X1#q+}fm|fc=dC*pK^OVB5+isfHlFbg(fAv!w9bL~Br<I*V
zqZ|C`O?ONvt6t2ux*gRH_lEByNw($e(8(x`!x8;!-cqSGzjwAAvEiWoqPHjrLf6AI
z2DFsP*hyOj%-R&l6sIR0ZU>)dYETB8_us##Scs~yfwsnl=H4S_5KVrA32n6LOq}2D
z#%`(1>j;ZN0|d*1U(j-_pQS>b2O6VaQ+;vk^nsU+6F*6k@zuvbjkih$NO3_jQLK(D
zW6#-MsBo7ku&oZaOopmQ1veZ-wQS0=V4Zwnr76YI74S_wtNW_B=}`4>Vn}Hup_V$$
zQ?7dKqkuZIm>&lcJNJ9nv!hRCw(0nMl2ZPzmn2N;#;_=@aYzyGXza)V;RW9S(Dy?f
zEVCN;)m~g)UB-e~C=wfJ23e5CzC4P?lGM<D`E;xu<JQHWwJmTD-FlcVtQNYL69KPN
zu{K@^EfCxSbNI(jdVUQ}bgNBPZO==1y2cbGujOfGVN%;DI16*gYn<bQUGH@_5LNcA
zaqqGSQpMhv0s(hNy~pm_9fhfz_U^=d5U9dePhO#@<VQ^_#-v}we#4i#AkG*wlz;(|
z@;`0v$||}vweS)yJx(-z?!ND~R94HmW4c-lzV_)6oVTNF#6{vZPY_Iv<;xZX=3ijf
zzCOF#kJ%1dE*3PZONf!2xoclP+Q$yNOM9evKQ}PCQ-;A|^SV=lDG<fH8@N+qvs&5u
zWQpahaFi={B^hwHcqq&m10u8w+rJ5{y}xyL90_5#e4=482&I0$kF?XmHB_NO)a%fo
zcN+c-pu#mJFs2t`O_YD9KI-F1cFF(7$y*L4Q^(N5T3Et8wtaHAJi!gZ)7V1~W16s!
z!@K28?`u0nweYfV1z}x@>_%`d`uY5BKJ-UPegphqtKehcH@`=5^lWi*%dMqKw7S10
zX(VO#7;G+(z0gS8J7^QNmWiIX*mYWXyz99S!QA4Y?#Dke|4H$>C+x&n<WY$hef+F{
z73yIAVigz*(>{7|<mCKKjsq4>cuHZl-~l`fAryHoe;U(N#P270#@B@RTa;zCJ#Cdf
z?_JY-Q_xm+V6tl_EElQArDEgr)4IUnlJA2D)lguHE+vkKSMxfwCB7)`?-no}jov}-
zMGL||xy>GL8F8uSo@>dqlFB1h`b6Vbp$Ff=sH}IBfq6MJ!FF{Y9_5tug)g;$=Be(Y
z8@J#QhbWk%nMu{CB)fjQ1AaifJSog~_c7)iL+)()F9`la#&cK_TjI<3&qR$tTLFnW
zAqo@w@}6M9ZYq;0D!?!?5ctY_j^K$}fMZ-EX7LnnzY8tQ=qq-nKtjS4C(>~hk+{IX
zx(_KyOSsm23}l6MC2}X#;Gn;L@$6rx;el}oMw|lEN?5YExqBPHEsD#7;wd0J-;;8Z
zfb#b;1cDVMvx~EDN_!Ju;!u7e8rkVdBR^sR`UN`2pQ5+Ash(wD$A0M#QaG!8^!YSx
zzlslK9&hHoKvCp<^MxKL2ZW+j&En>VL@N~2<d$#meQ8rdMKLPAF=ZO_bdONLdm3k`
znp1$Kz_)7e$|G`}^@uR*a??RnfV?_z<(sGXqHWnJq9J6mYFu`&YFys3)-i%Ay|^kT
zg^g_Rey86NO$}>4<EFO?&~gJLpJc<hy1cCy3^gwOHx2umbkAakok+@@Vip9}ai5;Z
zp{Jq(G(A8U?`3B#F}b*9ou#TLDroyX(qS;5D@u_Mzd5;xq6~ShSc3Sn!K+=VLIrPK
zfG61=;m98D?Z<d$ZCs*WM6!D7Kta4neTe~Qw$yOp2cim@H7&e7wq>rC?UM@_701I3
z*0nPs=}nGb#Xd+V+Vk&)e6L)kxs`R4e9qv*^gYgkBbRv)pn*ktS_a*^QCgZJErF0y
z(RVuqAibuXQQ4L-W#u+54m>&uz?z+A3O~OnxQd<$ZbB7gKAgcCkdM%p4T%krbzY5;
z-kUltPD~^2^W&0a-v2e+(M9WF$g>yCkn4-p>l2FI4#y1Qdy&ZH`iqM2FbQbp_i{#u
zguFRH2qB+MGX7a*(tY-B;QdxmMoX~zT!{M{r8ph!Vai!;G;2$EtU8Cd@;tnU5P*Fu
z4H}G^XiW+4X=sF*-*3X>v_IU?YC+I1)1S!HG7$!+`@?RApAiFlB~`%~BStJ+$Fv5>
z2vc9H%GzQMLmz{ZS)M~R^j3z*ij4Z?cVKwR)PpqOss+uwC%3~2&O(56A#;XmOz_r)
ztjCG8G~(o>F)s~~EgEvf9(bR}XAEd&HMr^EQnn-haz**aS|Em*qxGy|+bDIhj<l(s
zI>Qu`S>=xiwgzDABxLFKw5{t0WXCZR`oE!AcBcP8vCNFD{~L<!(Uy$UVTbR1tt&73
zoNtt{%4vo<qw0b?LX=_h1rSV7GP8jL@Dr!Lz6(~#&udoLe9CJp)sZC?b@-I5JU@IA
z%i`SA(z@E_isjTe;Lzpi^W^&ZCH}%B;$!(X{u)Hg*6Ypo^G!yE>!bTpQbN9;dN3cm
z{rD2t^8n?!hnYoC6U37rr{9}fyX%RLYI6NR%Oe7Nh|6d^QU_QHD&<e82taI}{?YgA
zPSX5RR5MO7h={BfWE*Lua`zJ`dXF$0%q&^q@;BSvbRm9=czD(>aDY49EK7)-L3P0@
z&~SRj5zAnLp29(+2;pR!`#sDVyJ%^5q(%zmb2wfWYhMB|hp^0h-$~0TtyaoIn)GtP
zQhgswDIe>mlX&FTdhe+0Ry{#KRsN?~9FHRf^A!6`H?NYJAkCdr^V5X+t+i<B7l&0r
zJ;lP+(`wDI0BzP~uOJa-`brEYu?=O@l|`qSAaL!`jT)&R%aZfgs4J8I`<U1}B1n{8
zQOP3xepEVDg*;}2b<>hnu-4*dAj{?<(Fjen)8^3nZ-fCT1-}z~ZW+THY)19vmb$iQ
zJjO2;?VU(k5nTPsRzmNH=`M_C#cqDBx~W#FBz()Rv;E?Lk}TT>dK7A9%{-Au1i1vi
z3V4eqLEWZ=%<|^`r|Tv+a`NW3&{UxULRu0fIC@h>s;m#zEE_gYB|2{7BaGHV;lE$Y
zm=Ai8Y~3s1r-rl;CDrpe`6^G&egRiF7eg6^)?Gs6u@1@-)5*>f*_?`hby)|9tysWo
z_`*%3a_E%qBtAHV(Sq9!eOoo`12Kc!;)dD~=-vHifdkmcDDI!e(z@<~h!BTIg<|t6
zm-K)Bc}stK7H3||Ku;(K!~4G}KSke6&~s{fO>mROEOc&B#wUcm*bqvk<y3ZNvn+JE
z!BRO}nP@lwD)v7`7|FyZsWvgZ)y>8vmohGt*Cra&vz=7!;6ty7iA8<vSoTW)>GWk$
z?I8t;N(SxQ*rb8U=HAh*FDXZV*%4`IGHggWbJtoZ4OU2@d}~gO{qw{+oN@+1DJ}LS
z3ABhMkdh~mJJzI@q>WB>%PzTMqXz~0T4I1PiS5yRU6?SG`Bu(WLb_W?&rC9nv15u>
zh(qzn3_wBgpGC&v@!SOg$JL#LJ&z`}CiQv52ATh?aV<~^B#FID6jY;`r<l+*kv^}j
zCN;&&hhgJ0*lPID>Z|y$WS+PIsU7wPCc><w4NRzX<vedXii@9rtdpPFp3*r$pQKV|
z$uEU%G<DA8XgZu$613gSK9R6|wy}5O%k%uLBJ@s4S79Ss3Ph#6>U}XkL~1>%F3@tp
zo7xorOw1k#6=Cz4aOe}#^xpQNs5cUdf9%t{C%(UNB<(ekCc$Nxa#8H_c<wj#4Jn~<
z)mGrj4<9S$AuadXq=CD7M{|&$m-E8^n$<`MeYec~(LDoJ^9wum5O8ls$G_92#sUYl
zPx=z5(I%5p5*5gzygOjlUZ9}J&|&y#8sb12B8Jg5EvUu13ybvW$ZK)6S7k-8-4Z)K
z(FGRe2}+^!%6hD!9g8mQhzrGUS|Ghpy=8i2zeCeyJm)z(ol7mgg_=_u;m|lh{`jSJ
zW{18a%I7Nk#+ACh*(&OtIKUF(yX7O%R~d`VwySIMaQ(&+SgO=CdZPlAvCg&6W7rDq
ztb0)!IU~hmV^z)~rN=b6pEtocO=t@^Wn_Y0#YNgYpqG`LkV0Qg&+rX40w}u?<1^eC
zVU|Y(JNjkQ)^*)iZkDV7Gy82tsQ&^OYO)^>nE`-~0yBe>jgl=QVnL||kg;^;jvwJM
z#tMG77;v&sX+f(kA)2gL{(A>`tb|&e(TGwl5YjJ}bEjKR)ye`&y=9`Z6<SId{LETT
z_czheF)K<P5q~NS&wfKlMGjz^+U5X8Xa&Er8K;$As2)jd{p@_LnXZN%bzf>7Tw{J`
z%ln!syRu%b?L^DOE@>>i6iK5UbI#h6wVfhlnA0P=A%zK~6f>zOEsO<99s~W3?tChA
zXX$La#95O+b%Tz*r$|sP7ua*Q-?LXgAq{&QMkDVO?$l(;W>6sik81NV>1J2BtA^oL
z)S2$({CH6)acM@Y&Hgdpm_k==hsEi3IvXA*y~X(`r^*dd)q2^j5dEl`!Wh7=m|oT{
zO1|H0I%LOLz-WK>rww-)v!Eq5HgFa<;!u#zD0@@3!qI0#14IC9SUZ}Pz;vYDXF{V!
z*94d<`pTnwidh3&nppwHa>ioUs_VP})0G;Byao(XE_wGLSa}2D%Q(;mRla_e09*78
zJBH3LQ#`?DGeHWj)6wRu6zfo0-gZB#rb`Dr{i9c*kTIv&lLhmG#!-h{1iwR9dsnwI
z{V03&g9Yrj3dfw8ypj!CuH~@Orj(CNNc{vQcAKi{szMMCMO~WLVTwVlXcXHoHHq}j
z;La$<B~0P#{OL|dT@@LC7v6x~W!e__AL}jyVBM0$s#wa^8yuh$^v^AG4bT#c<~<jg
zf(<A%voO<NO-)wJ4256?*<1<x>!uL%EjU<!qKJ71(#wrJW_St0s>`RM)5t(^OGAL`
z-}~d6wLiz_`d>?N1xI%`PkkJBFpu$kG9S0+`0o5=v)|A{GN_s=|K^)xcgJ$Sj|!?$
z{H7PCaiBhjHMN`x1>wm<zMR8g9PeIe06|W^6gLn90-b(RF`rYPIGNO$D{M9_7`BV!
zkE^jrShGyMSrZ=eQ`hK-At2&d=GQF&5-=n2aqIQ4u#+0bN5P_Jn6~6BMh28MdMAyA
zptHWi$0Oyz*M*dCVJUo*kfo)$E{GY~op=(p9_i|^X&HTHhhV<L;0w-dM;hXhj$Q+J
zTWo+*4$DKeu)_clHUKEhau+4{dJqRRNV*)ZxAuqx(Hjb;VKi^EvQ9b#&I;oOIR5Z+
zeKUidG8jJfJxRV$g>h3-+k?W$xiCuBHBbQYWI?H5I7Pw3ZTlD5f*tj5ab=U`=|t`t
z6$!zToP*Zaq}Sc(#_KEikYfY><DG91txvNtk9I9{`b)CE4nMy}yWh1vn$BzdJ_&oi
zU4HT-@^Cqy&0%Sy^W;vo#tY$ddhbyMky9B<OUFuU6cvzH^y}FU-y*o;H;GSJaJ{4-
zCl`{eV+m5@7#Ny2JSK|v6pEp1RN_ysOCiU|LO(%bvSJyYb`=fuHYSn80zxt!G>>(K
zyVO0S*5UsRo`x?ql+)k}?1anvEE_zTz@y!MrLhf-*K;HvqnsJPlm{)PheRo5ooeYs
z0|gB_w)W`4yJ=Z}Nl=paWYzf+1VPtUk6GjQ7o<+Vw8TNes>6Rhq5;E-N1Iux2yf0e
z&sQ^@;NOrW^EV*v@>*i;^*6wC9KyOZ*n9jDVbqreP@Kr7@1&?mG1TLLpPH@8+OUd_
z_sqCy-uB<lchlccg9h{V%?6Rs2yCng86KPxagXI;$cTbRm$g!b4VEs$dk5<Vt5zVH
zw?a;#K+k2$oBY_!vb1X`{&hnhBLNX}#JgtGY;~;@@64J7LK+KN(}7*VmU>96Jtu?s
zJ@$2-X={z#v~^n;uqNGE1SGeY`?Yp8GeZ)aVN77%Hc^{#Xz>bR=M>b+NR%q^;vyFo
zYv;$Z!1|L{D#MG%g4d5c5H*CfFz$PLkXQyF4bx0vCm~7DTHn$DPntqVcB3;4&ORWC
zZR&j`jb7UgW|2ayWOU*(QF+z95gf&imK|C>1ruKCzqGHv{?qpusJtfOxk2#8T)U0f
zRqV(~Cc7T`Jc^FU0RMje_v9GuoOfpjhoWFxYLs}(Q86A=zt=s7Yj~V0^sdqNt!By=
zQJRlVUs0gE8r#3CR}YA{1Q!g}hQ3aW7=-lQFQW$PboE91j+M*}LabLF1o|VgtJ>6>
z1j^LboWypbJnj=2f)%CFUH*RM)g{gv21vLY6957`c|ei!9K?lmm1{!*Gax3wL}&Z@
z!w63a&q}ORA$Bn^rMFLMk-fJjgWYb?Xn|?=Sbp1QS-?%4v$@e?$@0lRqoxY)n^fBg
zc%r5A@~$?okAu_U^l0NoY1pSANXTD971;w@d#BA<sZaCm#FfJD<;wAmA<GE*WzU_E
z*;FTdlaNzKd=E!*cx9moXJRhrZBnI1N{*ooNgL(!{KksL8EQ=n?d3F;0hxl7D{N#A
zW%znL%8wd-8x8b>00%rA@t?H`GbbzS|Eo<@JRMBw<&CVAU2N&)2pAa{{`n?5IlB;W
zax(m1Ph~a+_W!B-zM^GokKKXv_eOspZjD~YNn#c@N;p7gQ3vG7!7>|t4Qq73aajaI
z+=&!3?e9MV$;u>sW1>WA4$NNKhA!Rvm4jX`VTRQNUAv01tK>aDgm^)E0Drcuf%;Wp
z>Ym0t8wvT_NX!|DILdg$W)6*91<qb<C!mt|aL;I&OcaAs&6+;GeMg$ktWmwAUBf5|
zfGidw62EM~_<_=$N`)vg6UYeqU4g|U=AW{u4DQq5j)sq%a;4`aQXsjZDD|RP7|Ybg
zURi`J1Exe^Y8_@z7=dJ*@4aXsv4B=#Spkfwd8krl86>OZf>udba~7}(KtoV!%|#1P
zYS5#?I||)MR;Dn7^^TEt<pjV^u!)KK2NF74Ku1AGTqQQPMA3;`+VAN_zWicNWIlmH
z&O}8@a-b~06EI3ca|VM#kU@%HP623jT|fg-iow*9P{1m=l&o6=?O@2UVU<GQva4*Q
zSnY&q;Di+eh#L<n*u+rBDtJrIILL+L<o9S+^#_-fYWRNzifVS5w5GtS4)Pwa#^S7P
z7GIDg8blArug4Dx1aB%+$TU(Ck!YCFv0-&4DsA-~{MzuEH7mKc6ywg~k=U@j%B#$%
zjR7L1QWXaKR4!j5*$y%y7+6i&vLNNWW*iDr>ENZU-DD9mQa84`9#*yb6E(0e7tOe*
znPOX7;16hH4dP$Mqi}%?YLx7!S=}VVdO-@3#l#u}VE}?l-66H3A4s~3kew!D!o^|j
zgTFDoU6(F^1<rEtx%+<Isrt!@>l{m(aM|y~FKuEk+DKqU7~9eBI^99h>0gWON-J6#
zL4q9tVVL)BVrmaVsJD=*DD&`{$JOiB({T64mfv-cQ_snx8(orgU;jQlcK-an*)ekD
z`*B|D!KL5T?S3g+LfG*)D`IcD|6KUY=;;B>$e|PeY;EM>@pARgwKA~2Tak4PXB2rA
zTszXyp)YN`9og*W$dMl|*}1r}@#V$&@pAI#$@1U8)05<NyeFx{xT*MvsZdezk#ESw
zDIA`C9XxN(jgjMjxxBP`Gvm6~l`XyM!t#xh=stye3L2`1XQ<eN_}UI{^%1;aNOLRv
zF@i+ZqID2|h*r=8KqI`jdtc3e8@?Vstu%XI^k&6jxzO*5-9q?7!-O0Tz;@~XeYS6B
z^Wo~87}?mkuYp6X?H<I?e8gy`5`gse6O=w1W)9pjm0sh`$dxq@J0&k8PRkr~p_fBv
zn2ad;`#_J6!H-90P(=5_o-0cZ6MavU@niYec9UT`C{fa<ApN^X+HG#Lot2h;oUXht
zpyPmujy~RwM(A~B?>HQJiX1TJB>fXi`-j#$s^*du*bevX0hp4Z>=>mvJoVGz{!H=J
zh>e#Qhc>noLq5;7P7EJoOr7RvWMT_Rg-J7-EUKgA4sU*H!N-e{8+Ycj+OEwHs|Pdh
zi{X2nCM<!cc#&~8Z(r6r<zqi|gF@Xrs3+vFwoU%F8Hkgy+pYO*Eyei>s9k!Kt?PuR
zY?@5hS>lqp&bd^OU`(>X$GFIA&8YQcXQ3FyWm=y8Pv~C2Yo5OaP3;{SYav+j>rv9C
zPMgf`1k-yIP0|`I^YN8(Cq{~1Q=fHehvdc<R?&FoN3l=G$A=2jEtE&yY?0W2_blN%
z`fP*k0pSbMFl}Op`@DtFBxqb@y@`~yC&Uw^m!X_!LL~z|W9`G-4WOUS$TAtm5(iYT
z(!7OoJXd5@fI%Fw<6op-J1LcD9jV~{oX1THcvbO8Kf!%f#)j-%%Z2zEd~VgHqk?;6
zP$u7D$7>OVf)i@fRI@F3rX@MX@-t*9w<yCmXv>Fjm4RebZKc6PSZlOUi;fuKaTi2h
zH3fdZi$<Y*`KJbSI&m$l*K(6}EPj!MP)?~9=8Zm->pkn7e(A9NirQso&1#NKi(gS|
zCP)Z2l|>Y;_Li0Qp5~^N_Cdp(1vQT}A)F9iFcTyXBoibTq$}kx1Elbr;~6JB?^wYp
z#}iIiZqfFRdm)n#I2v(P9h52*fX5g(G9XY%^A%U{m?|QP<jA`rC7IXc2KvQdN?jUp
zd>3!ShZvG^g^@CdgXB|r3Qxe~C7hr<XG6PHO6FHcWXknlF7I*Bk3j;2g3;)Dl#3lk
z5~y&U&OrqxPoU4OfPK9u1M~rgXQ7d;$W+bX%}R+$y^A5tU|Ml_fKvru#&4Vc{s?4s
zf*Po%Lc&YFHxofFMnt4+x~)J4CBaal@b1zF1$0G)wmWE%iSl|R6Sm*&<Vp2j%^c+`
z<RZhMp}76wPH?@cu889mROxV1NV2MhcX%v7Satc&)QI)c4VQ2PpfONV;}U<s&X81I
z@}b0NfrO`Rz%8|r%NcFHu@WlLeF0encetV^SOO$s4W@dHKq9f}PzeOlnFUr639l$*
zG>%6)zGRxD@Xo5yPxTSTzkVByk#Ehf|2ojSn2k2jRft0%h>UNeEtDm=86iU?W#j;t
zZB@&`MKh^`l_k&<=?npZX%|CR=})oMhxGBpCuBC{TXz$zjI;HYLh3QnK2txaDO_19
zYF&6?4S$=puID7tCC-iDm*;Uldkv(EFpo_eW`^To{I?c%D)gbvunrJqNzk@>M7*YI
zGtrhnFlG3wR-lYJS}m-wn!i)S3c1mRSYRO);s?zwQ3A_VR9N$!L<ry<bW5zkvyQZD
z(i^KW?dZ5M?Rt#bh0&5ismS0|W$>tSc$z0ZHNsG<aGbQE`;VGg1r*4}pc54?-B3Ea
zREZ&}%_d%0V+2cWj%FAiSbLq|M8MK2ThHOQYX5O}-4R@%D(1@{0|=e(L_LYaHTsR4
zkwt}$1qC%<_I1<agxsbHye4e9ih-4zo9_gg1PQ`R4yK@517HKAFK50G1*2)Ii+(LY
zNE1k*cioE*;;|S~Ca9=W09;xF=Iw><DU0eQlHRfPiY9+AX|?aSOSkVkVcJb1D*u1W
zeG<`qvffK6)HTBo|1oxoVm$NuOX;CRV?Dy)fWUA<S)z?bU~q$>MbmKKVLUnHB~4n<
zwSX+)MIZ39@149h1x!>qXrn!6kqGQ#gkYwZR7Y4V5Mbebw;e-3R8)e?KDBoo{Lph%
zAT^~a7<j#&VfhQq2OfY^%t3T>7YpcpLxqO~bJoNiY_KFO&gI{x%qE@PN1R>du+JTQ
zza5z{alGGOV?TU9xA`*U;j}cDb_<2lS2B?`K?K^7jqVVkP*OonZd*ICIPk~s!_U)p
zu^e-}s{H&!e%vnHgF8yAr`=`pbz;BM3R<x<sFYpOr>6L`jkQfBL@(d06ElzO7BT;Q
z^UNfKApM6YndN&SfLl!etYh6!%poUwU4Efmeq-D<(j;D}_#L<~fdyUiv)p%OjDi}G
z0G@yon=!kWn{%xBvt7HIG9Se)P;P)os*hs|lC8sRE9_62clX<S9>1<TNJJigfMe0h
zgGD^<SD-Ad_Fw$cIo+)P1g0zu9RICp&i+4w&wo_eqgJHexjF^AE{RmLdszb>0r$CS
zU_Q`@Cibl|5;j;9gnLqdepR1{Vv3gK9r*YlF{=m~wFbQEkMaE_dXzjq{+vD@-$%L?
zlZ+*Uk*DKRs5_9czW)1?g(NTovB9U~%cmjc)4@I0QE5jyQa$uM+}-CXawvH1ekgj6
zcN4VY7o|P0H=lw(9lB>Vhw6^dsKx$r$((x2i@<3jC>6h2LmsE(^?SV!B7FO~RbJcZ
z<$wGxz!(V9ne0A3*T_HeNf^$DO=y-T<@&yzKi<#lLWD}EGm6N+K0QSiu}*?X6ZyY}
z%l!AoZtEMK>^|fzK3ZUh%KT;hhk$C4+@c`{k;%90kxr5fcHc$Z$O=l#b>0t2+?>on
zqmtU?&^R7D$+2aofieN?Vwa=C^UcAj657mh!$}cfqD4RyEY#0<w_M+=Z-QHC@}WtN
zl6hK~UQrmCiKC?-a5LdvdPZ^cVk7n@dcQ9yd)DrGeIdd}am5xbm!*qkUj>3VJiI4b
zaYKjAwA7{({RylEFx-xBLkx4n*ICX;Tj61Ei$iB7!Ri+9sa9C|6rvB*u5}rqM>1Fs
zB5Qyi_q;m2tA&|ou7m3^Px-XNhAV(EUBS|xY?hdHPf_VqyzCd37wpq^LtZs0zhZKS
zoki8jR_?Y#uNr(*&nJsEp0>*N%9hx*Tc)<>-BZS}qpPc?O1tC_%cT;Pz1Yxvg>@jN
zNU#Gl-iFc<hRy%}hIpjLN(!_TG%EnCjpX3dZc1nu6iClC5_!noG7r!+N~<r7u}b%Q
zC%<I5hs2I_w`@MYSP8By`oArWEepqZ1FjZa-iF4!e7kPe+PL>xjf0HgGyeMGYD<ms
z!s2SHU$C6tEPbnR_S9UFyxzNWw$=$vR1|vE?jE~s^h2|!7*}If)YAM>-*u>M%;t%$
ziec06Zyh<??#uhd;VL|b9=QKF*n&^FZFC~18ZbAw!qum8qq8-BL(uo9h5Fw8!)@oV
z-1*N<uzI$(ai{%_6jNN{dWJf&W|OQNZQTvZ8=}7gX{@Dv42~+>CfCt?^5hAc?Lf(<
zu+eovxs+nB6t$-f5)(6PS;#&&3Cw0~Y4iTL0+gr4Yp4|-Z8=q=<uRGp<OwLpnzUha
ze0FMU;+q$ep(ZM0N)F1&HWk@Gh7`5wv@FWEaw)FgadQS;)*mX>-ZGn%(IvjJX&60K
znYAiDx_jCx&k~z;7Nv;wPPyJg3f3AWfX;)@Z=R@O?;Flnz|k(1GNbChXmJL`Zr>X}
zxJ&7y6wQ^iG+=(OoT;hUvs_kWUlC)I1axERi4yQtBu};Vny>fkv<mOX?^CK|Y!gJU
z$w2oMSF4x~C@#k}4Z8Jy40Eat<pKNHrehM2E&!ClbZV2##K;q3GUD`@T0zcDowMv1
z{q;YZ5EfeKlxW784M3!VY4ku3u`glSiv>F^fsvH577Qwy^G$JDk%-$5$Ni{rV8Iy(
zYGNg^;8KY&&5ww)P_JEiN$eeUIWU3l4GnmRRbZ~pA*(>afNuXzd5yZgtx`wX73dls
zLlykY`n0fJ$lLP9P0<twWzlUBztQ(j<^XF&S<6upnwFzP*ZMtmlocJ~7blR~jN|r+
zEDAIOCz2UsZFgx5qdxj^AXtkkLWW>M+YmDuLrQ}WBf6TCZK!Ji5#*we{GBUB<*zBw
zv)-0SjX6WZ;4y*|Jj!Sl3CXJzF|N{VT*;&dBe~F^2$=5)v6b%k+`!CJVs@lVfnKnC
zX~~R}a*h{f!+M_368EeAbKqea>%^M9mgMz#-3Mv33VRGis<{CAQg1n2_X)}vLjY7N
zvn#*98YA&GJnj`^qv~Hr*;I{L{@LgREM}`ZWvMvRbs{Pu)8V3Deg9e~kEEqU(@OLN
zn+(_6w%U;CJ$t%5elWkiFsp7VZs3#FCoo^9^=Pr|!BY<O0+9i<?O#7z(i}4~$8;S+
zP(yDX5zWMi)_pBGL6BuO_M$oyZ5Gya<41>T!dmMmG@#$GYq0SaOqF!7YX4HN6B&$p
z?b&TxagrVC8yevu4F&qSdeMKV2uSLqEW2<-f^Y8fW#>eIr@cPUc6uNmsptUiouQr;
zh0qRl&<mXBRoyG^wq4>%y;~pNQQ?uQT+Ae0qUH;s7esnBgz79VpE7z2h;+4boqxoI
zn`f1985ZNQt;Y2i<~_?oztG@v+%3*r6olF5j#%OL18Z+KJNadd^eY8T?q4VrT(P_2
z)BlA#r)qiyQ$p1!2&t`yAb)fK!}$7DF%>F$spVS(lM<8u+i1KxL(DC~7DugXd;Jc<
zqN{IHq=U&3hY!-nM(4l}usz+*xe4v{NH;m2i>$rO&fMi04d`#6nAyB34S!B;0-2C<
z%&5#snbR->;GA-!NGk1!>FHNXjX9EE=M&{1+>tf5kA7*MTJ(WKSK_jG*P5Y<4^q*@
zptk$UC+0n1<QTfB?-qMX@yaIRoT}1P8<y&OsVn%-whQ5D5C_?fPMQJuoi!h2EW<r>
zow;MY;y2g!yl`A8IF~)o_V0(*UU&JX)=ZWjvr}jT8SGAWV>{Gq%*Ciu=2B@ROi#F4
zK@Bq)D?_X5-qz!x*JRY*PR~WVPQlM*+)&vGTia+NaYx{y{~6!4cqZ?%U{gocq_Wzu
zIyhzc`m}!|EeR2m2U!-l(01-aRwg8YMt>t-t|Iu$WfVImjX9~#N3ZIF(rrO^L7s?A
zlKeb4dZWVLZOacTPh^U{v${6e#W426_MJb_9vKeDGt&&@{BwTZw@qT6AVnDr>sqQk
zoQ@R}6a>hrb5FKaz?3o;(d{`d+`>uQ-fl?JkPgs!?LZ5`z}R@gIwC;!KrHP5slZq>
zR#_VC=9uvYd$V?Op>N}^TsqiqfE_iku6Tn5a|dQ3lO19x-K*~L6czf$4UK#|>gAYA
zg9*g-=!(CYi6?28(WrW1!o-R~9jEnFw!W5+4nV}G0x|Dj89-+Y4#sh4;<LZ{fI1os
z?ZzQ915@g;2_t?$_6~DT+AnF;V`4*2m7;(?3*i54pUH9}^yVDT8YPG}^%=^VEs#U-
zT*4F1vG)8#dakEUxUW8}ogR?d9*h)uC`yv8f}$0P5isIL3^K-pmu(dhc38zdj%kJU
zWZy8$VxN|@R6c~XhweHns<xacW~H^Za!yGGaO$$phw1m(MHx$S|BdHJCxe6+KXxRe
z`)uK{)5eoG*Sd=2p_UV$;x&U7GN^7gs#{u6;(}t~Ml^~Gfu62RR-R#48w9HTkjf~^
zF3>HC)!LnQ<A~4P4_tgu3khnm2<hp^HAOJaF0)W=YO^f?hSN3-T&P!X?j%YSjHP_Q
zL4ismm8O1YOkgNs)oUZN#jYgSVzUx0Mx!SxY8B5&S~P&3XC!aq)sY;{;Wny7^=7c;
zoXjdsYdxN^yHLf7Fx{!tS`$Kd6Z(*)U$3@-q-u4`(6OnjihR(P1HqhL+Nx*B<p`mH
z3h0Db$*5y?BhEb)ws0)-(5MJ(d-kt=l1Om5khlhDr}c6XW6`FV{Sd33od)DWE9VqJ
zJv-{cac5lWs~U>jI@{tq{06}hkDdlymN3B?uITl1vwChfp!w%z9_DYG%1uD|-o5gV
zl(fbo?YS@`!^xvTvNo5(gNZ<qy{NI1C7pQ~-U7=S*AOg6YEn#W97W(Z?PS&BoVqko
zCJ#RB>>EnBL>Yd-^8!9t@XO74T}suOwdcHEb5Vs{y9UK#ZWioMAzovCe7wUBge28X
znoGI2^iTbMrZZ~ypbkBAEJmvZJ(%w%CpC3Dz+!S{b1C+kveyUltz~3j>#*rw!C_EY
zHpWv`Lrd&twkRq2kaz;BC`!U|y`Oe#J|qEnrY2W^Hmf;J&>^yi=?D{oXU33@BcTG-
zK}*Kxw-7ae^YH~wtl<DYdTgS;Zt<sza!ghBOP&EQKGD)b?m@RLg0UM#)Dr5!*&UJq
zPjr|i@VsLM`Y-bpR6t9RMm9c2-Rm9)qtYsbt2K+{B^M59uqDPIO3X8SJUNgOZV)wj
z`uSBgniZj(7D&A9#GfU-tCGzfp8mMfAz&E4w@E?)F7}WNnKiC2Yxn(qL-G?QqLmo8
zr(qlmua5kO02F878oa3`u;a1Z^rE6SC&ts)>~v+7$SJUq6clHdcKwZB)|`wkpm<V^
zp?)I^Mm#QDHypTcKA!wro>Ac*n69f^K;;ii(gS|)e=S@7o5GKi@qg->|Ks{NU;Qgv
zF4gY<Y0@5Na8c5%s)@Oys=cHOGxwAczW3bg7(d@E0uDfuj!QD|sOA#&X>l<(2r&t^
ze%3aRheK1JLuu5Ghq8!w>)k(_0lW1d7}a_%iyprYAv0g?XXfNr3zrT>5n59v5bE-R
za*rP*5M4ldE`id-=M+X!E`UDXS$%IukIPDWqUJP#N6LscE)vAN>XS_>QRf@(yxxz$
z%OJF;f1I9sBgnX1x=p$sZI%aZKL1>~pD)m>hegHx%Ixp;I|_RkhtLv;2C2n26pTe0
z7OvkVZYMIrG;!TAsBWq3icTx!V3`$>7rZE2CrIQgsUt%jneQN%ZY_ix6|J`YPAlxb
zD9E%=--nD1HvruKd)g4J%KsYf-{TPrg}gFE6H~ek91CJd>~+Eyh9j=cl-h&qRM`6X
z<{}FAVhkk?N31YidJ`RybPWQzr2wuW8oa>fc1J`K-ByTAL?Y_8qAFUc2n|CtBqF|7
zh-RoTWr9MU(sLFvfJH>|MbsM+na9j3n?aHx>EVlIC2O{-M#K)aOg<7f!<;N=075V=
zzbEEqUVcLo03wua<5GQ&sE@=HVFayBtdGGQwqQ7hix^d2pvH!hLUXLKkPkFcbpRE8
z-msA54lI=88RC;uaAEXHM&#wdpm`Jvd=MkQ5d=s|*R8}?8Mst2l!Ag;THyvJxR{0W
zqsE9Bh`@^ypEna^ixjP7RAF#fWIhS~W?$4|S^Ws&u;OJ(C?93CDQ-zGvopPX{lgo0
z2xpK!Q9nZ5&%h0WqI<+y9~Ik!U<GMY%(h5F<ARu$T;|isIyzE!8boK`uxc|CDXM7X
zUT7kg%k4%!dB#y}|F>dWS&4iST~jl^^!nwCHAYGAwdL3~Qnt|T{ORy$YwL3e|M4j!
zT_Jj=uaEbim9C)V1>eOE#Z2Cht)K7Pr^-e82e<-xjk8i2y@Z8TK2@4xG#7bWcD<C&
zwX`>HjOib}(<(iSh)bEE+@;}5miC8YK0?D{O3TnEzNvJDW|ERh)u*6txfXropIktF
zG}sxY*r{lB3V>pjX9P=#!R=QU%0h%1e^YOM-YL@vb)FQm|KAa|p9O9Ha|NZw|HIfj
z1c?%E+qz}jwr$&X)hgSzZQHhOW0h^&w*A(A5jWz-J-cyw8QI82<{vpZzVY-2@2zIl
zFl#r<uhgxBP|arCu;uP=<v$sd97IDnt_Z~W&}u&)jbWKdY&5DadX*re36&eG?m+&G
zuY*)#yI$pbU7ENVqwK7M_AncT$ig-XkwVNQG~1s*6Z;aV2)yFO9qW9ObfXJ3M=vis
znzpvI0e}wf9#z?v>SmYsnvX<4?{O%v-7SkK_V_s=5Y1!T=-?Idn%2TgFjt>endu;$
zaZ=)wvlkr|r-=7xsAf0ycSF>>e(rs)>lHBV3(O0teLj22Ys5>>c5n83@e=6JJ4`59
z@B$ie^{{P(aE=Smp4RQGrQv}~IcNEz`heP8#e<x=<#WodI-;amABIjeW;&d55M!D(
z@@ShdosriP>@Z~{gfO=ir(#+bd9R*WGI6^zId#nvLBc0Wxl~1o;E4%7lo`#g06}x=
zs~YIVn&X!j&+(^ox{wGThMCz*q}03F@e&d=ypjY1sp`jVIbGA1s*at~7lt^A8#Vjv
zP!nL=)zphF86;OeH@ePXVbGLS1Kj@Ank8Adi>;W3>)q4ZE^1y+$&Rj`86Dfv#7)BE
z_Z)`2P$OVd3y*#ueKfP~Cuwt<z3ZSf;Pzkq?lOZl|1_-~dwh3|=s4aiSzo8fgT_!(
z_6eYd3DgEJ$m5VRTv65~?;Fx<-t&SW!C`=uS?v#)R_)F4&l%wcK*pA%eLe)_EQV%U
zzVJPr{zLG7dCa}A_DRVt`NO6F?pBe*ljqNof8WoB1@18Cf;~T?UZMZIBYE%XWrt5!
zA!Y@cJ1yt*TVqrbl!VGL@7!1r|7!faJ8eO}!a<Kg<!Tr<NrrQR7(^>AqPq{)<o2s$
ziX0WxfXUA_HFVM4dVppG^i4+Zx<1?|W{4zputM#i*{I`?9r0-H&e-I8;knD{l{={k
z`ha86k+ZLLAW4|?_+-@~E!&8kxCW0@MlEGKv*)(ywMmIC!_!%_D%%&=YMg27J(s&=
zTPyuFASdF8Z`+AH1{vRA%Sn%3xrCaJT1mR+t(Dj}$(93*k#NEKG=NE*e;J1S5+<n2
zhdY}FbLg(Z!b@d{Z<67fvYmUGeJMM>l|q{0_zRM>jqYyni#OS7mXzFWwFHOza?7KS
z?G9?*`$iNmrJmk7Nu?`y*)H8zG0mK-=JV-ut7LbJ)x3?~$$W}5{=;-oMrhZnzn2l*
zQDYlI74k1nigBF}(<^Zm3uRlDgA3A4631)a4phfqS7=rj!A?fuu2$JIVKQU(xlbj3
zHZE!t6)k&-+?h0B5j%i08P0l5k*xjMygt?>>fbe0CIx7u;dK`HWqHuovkTC&Z=*Qm
zftTy<{-;Y)mjhtwW2co!cDj<Tjg7jLe&2d0u&t$*`&`G$Om|$x7=|lGqlli?hHbTz
zZi%y0m(Ln8D5<qC6eRU#EEmih%+%^w8=Ja|zDeEq<=H3KEvA|)PZ`=nlS!8QX`%RQ
zhZ3nZC(qiH4bhSaisHX&uzZ(v0mT)p%QTajF0vvJ%zns4ouE}f_^{&Lq*YS<4XSRo
zvy}u~$a8+B82D%ni;Y9FI_>aPKr!yLzFq1mC}gM;KA37P9@?XgxvveCkG?khA(^*_
zWjjZTh@XM%5jt*fq0LYnSY}~)-17q3#af1{m$Q*4_FaU^wR~}Z20c!lmN*S}k>lSd
zJkG_{z{_RzgiUuS6=vXnM-k}p@-B4lv^StyG>IHPZLGCE(!O|xo4LKKmf6*5L-ysO
z?CP<m2e)<z62z%xBkA`t(6s3saceT7OWTHHpA58sFlpJ8&++fzA$Az1Y1-&T>tvxw
zz8ggCTuiFjOZ5`MDoEfaJzYStJwgTb%-WuPJH)!4Mi%?x>-(o-*W8jUxjMxc9j%Z3
zARHV959QwzeG6_+(|h3*H-7u(z}e7zIb)_64jiMP)5kT;u6_Y@h5DxcTLSfu_4NNr
z{+L;q{&#)f9_-0joVNR3z4;D{1e#&-d(>+1b<At1d%|#}Z5(_cQb!nd!7H4H&x(U8
zNw%^wFIQv0nhb-a;Y!0Ejma{6GQOTZZr^UN&qDpiDGP(+r~5~eJJ8Wz&{4r6a%vD}
zDH;4enO$B-I#@n&Z-1wfyN0*z{W3ltihcV~_rKyecr@WGdXnBCzWnpIJoMEo_ch*7
z2+4jj@tk}ti%_)?rIK-6-oN*o+_`h;Z<Lj@r*UrlyU!pPgqRrYJ|B+a^a7Bj+1b%1
zSXWPz{AIIl_)x(x8O-9lcfYU1@hc4E|6sB=UVPucdcILfDQ=6uf0yV@u=&1OzERoR
zLDm@Y0H}>y^?;|y$GaaPuV93zXF1-!3D^uxNup$U;8pEUT;SQl(~&6*{{d&Q5ryXy
zH3_Y!d4S}J$P_3PG&%Cs2c*B2&f7-NH}LRa$oG}_rkPkG;Ug1ADn2+P5d^SC83pzi
zN8e(DnWlz0HC4Of^@(K!;cEvi_7QIj*Hm#VMQZMTG|6=Kc<oJNy6uYtJ@!4@wcrH*
z)65-=cFKbn<V|W#AUkx5yn@K1v*5r*BB|jiIi>9Rb6)T|>a8BQ$o{TApn9(Eo}t_W
z8E@n1{NW5)S$A@jlefT)<u~DVe>z;-8I-V^6<G)az2JrV8oPGzdRiCcaDWV{#feXv
zkxBy6-%iKn$ZQxId?Y*-j1X*uH4AbMLL0ZqBDedUDm*$Xaj}UmfNQv;dF?sh3}m^m
z`*p)t<9`noM4buR*HOniVuISMySn0{&T)cpURO0J#dtHU5Y~owsZ1)<nRyGh&*Q^6
zOM5X@r-1NBvQDn=8LVTY5|U~<p!xToV+QbZFE6O@acbiJtG9g}gQi&p^e8*NJh1n4
z@8(Rd8L{Dl*yv`Z=WuvTr*7R^)zvEXF{sHZhC<e#5C<<o>u7DnY7KBNI>&3_BfRER
zb8iBWVSagz%Wss@9ndqZfiMFe>M!<$1W+Dc?}`}aMS70=Yd_z&uP}K(beOd7v{J+R
z*!u#3t=^%$5e7&h^&y%cR~ha)MDP<vy?&lh<~ynM#(Eg<%o_pCEuAjQCvvQ37#s-=
zS~#3~js|N5SQ8uwL(J*VZxwjc>JwTZSy0#yQH2G~kBG}hst-x|Z*KWv)!)n;K6y8M
zq(3c;BMTMW6)e73(vG})mHGrUrB0}3lktP}3zbu-c8vCX?3(jhV3cpH`SC)dG@4MX
zsB<AFJWBI9VF+rW@nH{U{Jx>^xHTL~!6>*STw*psH3EpJ5nTy`kFWB23!Q9DQ}%cu
zQk?lfyyjtz*21}{Jj_^SifB2dr|av?o5q-`j^H(x1r1!_0BxTPs%Ijdc&2~(tWCj8
ziM)}SJ!9(91pvi|gPBW$WrEnI(D5XsrnF@ign(S7#y8II2v~X4DlkhNo>q?70iI;}
z)oz~!)jz&PDzpRg;S;@)Nxc}yi`c2Wgc1tOnP=zu`bf=9MzfhB-_T?<sEJEA)OKR|
zcBdiD-)J9W0hx1Mo1F@!)fJ9Ggr<c{*#n>{pFuYd)fmzV7qg89Lvp#ik=unNy7y>|
zn8mGS?fQG~Ae@1dPMx_OF?Lx3WceKXtCdokL9-RaCFbO9r0TkTr!2hRS^Kp;78Nm1
zg^1-rfofi&R3$SPaF!(_YoXy;pz+LlFIBAlyVGdrCUrf-4YtY>6G`5kPsA266+)^5
z<K=T#-u~VaoO<}ggDl<^6wu}weADsL36VnrBy`}GV>xi536N~qI~pCuW-ml)Pw79=
z-=-zT%M53}#IcsvypfilRfBRZ!BxsBTpB9GH_;%^i%oUQw2TF?oOWnH+XOmVI=772
z7`K>xax`u<fySzR?kaQBimJh@ny2`{f|2cDu~x1Hir<~lFqa{RHBE&AWt?&3xIk+U
z<_%%~SzkY1*RCW85<LxGTkWO46g_$oX#MjtRYhQFi81{qtzls~7nKG=t!#%hLztwy
zM}w-5&_3jmqXYwy0lHY1I68GNWhT$eE6QGUjEO=sfKp9eSwv2;Boq*Dw#oEw;0n)1
zTI*Nwb{Kg(W!Gfp|JchpUHxUy8V+bxPE%6jok;{;lnb|E@}1cI$ee&@V+yT7<Zq5n
zj4oeOHm+34o8)v|)X?j&O-habtV2#!q(kwNA0yWIE9J{E5t@K3_E*fXT6Yrc6{Ef7
zt;a(ygO-XRR79<#G@F>QX?fugX|kH#VLb2V52?Zd63Tflw0z1C?&`gWyv;Hn4{J)+
zh`A>B)=<6<O6B%Z)~9NBwPCN;{b;uk{2vaPes}TlU4_h*i2FbT0v5m%Z?(%*ETa)=
z9A%|8>B^XD>>4GU^p!>((g!hHi3D&{lzmfmhL7f>S@bbMIpCW@k8*|(QvO;ZGTzcw
zv1aLB#HzBQxe!QjuVS$0Xp3XbhIa?m&Uux+J)zMcorKC4O3oJz%eo97sb*=>!z8>k
zX}qC`K&&tK-EXWN`h%S*$+KH@+D?!umBnvEa2ILGG>K4vnRg@s`o+*-7KM@2UuZ!M
zW+iar34*}FFq(IcIiP-xZ)4Um-M54r>&2M5^|(}J!hvg7B(np4Gt6>OuO&^oR73oo
z_qZwZWX3ufgXW3@K2EQa1I9!@;!5$3tqvUyROu6lf$P(xJDCF0=lB_%JP|@slMB<T
z(`GvgEJ&*AyJkD)h47f6@J*T*RmzV*x(&LWXX`IGL%?eqcD;#|+eh12Nx9?lVO=0N
zn+A7q5kkoXFaO)nVyR{~r88f)gwa>4I+$fPH)%tLqZBFI5;0D&nb>kkQeU~iAUu-f
z_=TNmh@0=1;m3sKq)aX#w0`+0p1bUIk=LPib<5F)37=3!_X}}GM&=ZkmQm4;qq4%~
zIzmOjks$AZ8!GS(p|Zac0JH;%%5|f&hSh}Iq0DPU#%A~8pEYwyyK>QpSj_dRl&B|T
z&^IMCcC)n<+%jh&Lr}h~xc>Wu_I?4(kJh783)leop7_VSmSI^v@=rE@=5oHJ;P0c+
zr3l6Oz0#v*P@~i*WjQs*XkQN{mG-s!n$Aeq->QkseGBoc#nqlf<4FQE=UnFxteecn
zIo#&P@xGz9ib(!?wvY8pu9V7UoOw2#uAF=JoTk59cJ({ZL`az%I`VUyfjcV&wAD7f
zavu~|TftE-_r@l*F3XHy1&=4rSB80V>S~_7QD-|v7IERcxSLU8<QqPBchtD)W`n36
zPe@idFOZo=n^T=4<J;FH1}K8HRRF7c&Lp$=gqBwrAJ$$9sQPD8PG55?px(vj^cA`o
zu!qN_Lpr<Y)j%@27E;MLW;3>3i)`DK8+}fpwam>H)QR({x&_zP;D_0GdmCFiPO~c>
znyv{`W;%b@t75fgJ`9iC2qeYfHd8KU41$l6Y*=&A+g?`wT~FPvqJT-hI)CfQ{re#)
zc=~Qi&Jtw)WBEYO#yxIdutc-Ie(l@?^lyJHlf-DKMr+VWoOl_XmqllMfR2Auj)kHo
z)I{ZUzV)-MyY2b2bzeR=g!#3#W<X<#alR*c=xqT|XL|e@x$tl8@xx_y{E);%o6X_Z
z7NM$73`5h4i%cE-El;&gpn{uEYZ!+)dDu0$*EgUxmwGf11DH=p(+;emmbZ|8i@(+r
zrtKRxY4Z(CGWHI`XqLbC?rW{f;`Fv#`udjm$R}XB-IuVLRCXEm+aR}eK`hgCn)YOO
zYp0P+drD-J?h8RlMh2aPTQ}l{#l2PUrv0+cjYaKJtLhaNpMlf0Z1@oJa(%c{2xQ3f
za#lPo=1S-$QUwBXMdrcNaS=YbHkE`GxPx#27dYc650!EistWo=pBm6D%V1gA!Bk14
zb1rQL1ZQ?k8*U{1!z7vVPC|#Wssyb%%@K!FSrkS6_f-_;Mn}~dQM0zbDXz^ZcCY8U
zq@cP$f$qcSGPK=8BG|QllZa$hKLjPui64%O(YIxu6PZt!)tyn;@}xGL3{e(Hr`fQM
z*QED}W3FX{ZfpCGYYr)4QJN#XbTJ5-_o$n!AI)A!16uXG+WGURenDa9ts$fi&tX<5
zu_LM~DQVK4n}Y~tf6X^uECq^%CB9{di2r>g_L46O0qqKBpqc;ao|_{Gi6w#NLMUnd
zhBhG8jw^fn3NO5C_XH!)GVcVKKca}+2>Kg@|IyRmHR??&Y*r?ao)I}V%g!WWMvJv_
zHO1v6v(N}NYK_;T5wZZab;p)b0aoc&%yLu6w-@{^yYolnr3VJ{$H{@@NmxlIClh4r
zl)Fm~FH3h_cwbZZ8;6&48;Qk`Ko9wyxKuHt)ixYmd?d)|fbJ_|qE4RskL4c1z*?l|
zEo{evcfvF-w`OXHC9VCe(r{L~f519zDSFx%F1EY2t~4g)kbdgQYBA=nFJ?aiXu3M;
zMyRI(xyVg)4N!|7Pt~1P>>;)(o{;U2%dZUn(qMC6DUxrx(Gkn}BYZ6-WxAaytIQ}m
z<Ly6-jWa}1p-Sml0n=+c7}oE~B9=_edZz+4B2)WQ0UMR6+^;%%KS@(o_d4aSukU_w
zHM>RswP5?7#9q$-gR~>~w?S_0Y~n~SW^LeXB4T1>XKcd92j%4KXkuUk<-XCAE^E8V
zfDrP^2TG6N8p`|y3<TVY3_ey{?k4!hW>L1uoV~|4j7U_v5qCUNl>o*h!MBtT4qqJP
zSvkN{(+0ju-plaVrT#32tD7nrEdCe5aZ%$Kv4vPAjQ!aWUOzjz1_g{D@!*(C!e-Si
zSkQA%4662)$LL&M3~J(;&d2tXb+@%+*%}Kp@>LYGxHuOtAhDD>WK|D-mYQNvyrMyw
zA5X3C&med`nK?{#Z%>U9FtR0-Sb-?+QOD1hLS5CYYGDc(?n26!--QJjY~gu!(gyi-
zT3vpx^kI{8DaO)n>7~{M9M?v<Zv}lJyv_k&2nUCmO&o{Vbq?ZvIVG#fK?EYS0rjk1
z$w|+Wug@vL{ju}v5n5D72&`k^OFUP`Cvocx^Rt3RBo>G7UW1suM%A$=p*CB1VpY5M
z0O{HhynEUJ13AaB6HqAaU$AQa7juQqiCQJ=zwNN^I&cdp)u#OaVlWi3a|UbV50mq{
z6`9_-5IKg9=}4Ay9hfoA6|#}}I;(e^4gaDNJ8AZz!%XliOJ-A^XE0g9NI3L7N!fZ>
zswvc*6V6&$YXpvIoLg`Z<3@Z(hAr-GjLW6FNgh+KEw3GsrR0#trX`~S?Lac=9EJ=d
z`i=UygT_-SXLeIzG5BUT1l=buOIQ9<SHwOe_9TQY0^(6fDr7tw!LoC?!aAz7d}$MY
zlbz$*s~xga<pK#tZ80(1yIb#UO<&f*ZF;va(aAc;&ZxEU3@F-wb<uomMX4+34of%A
z+&6poBV+X5GpCz6m|gso2GT5B<l=?A;+KNVj*B+IOE*z#k^-5CX0YYisq<#CsdT~&
z1(U>aW@9(Yu@|ya-#djELRYo`P0(7<%Sf!q9m(AZ(3=68dqP*vxDQ(%;tg9FR@+XW
zU5(c0q?%Ur<kfC!5Zs<TyZuHoEywiE;c;5!40lKtM_p5%Suiqd#Te6M#8FCym8TSW
zg@uk%ZxeQy?#n@Jl#F9CH_Y5FS?21S>$mk|r(nWg26cYrw$Kiq5r(2xUlLhXu28j?
zt+Co<3~UX~SqX5A+sRfl90mkF-XnxT`96~6yTx;Or@PEX<!3mh?AW6Xr{sX6<h3O5
zmTC^EnJ||&9QNyM#cWO22w@5=PG~!|=iVf?JnZA*Lz2|>=d(~0nDOl@`J~8VU=Pwr
zC)1~I0#oXKT3?dzy3}+ICReONjgjjNQmtybm2^n7rgPhk0t|T`GCB2~+Dzst9m=fp
ztoN@`tU*%xcLDs&F)LEC5xPk?f4c**pbz|<`Ed0A@^4^d_`lR8GA6cW&gKM6OpO0K
z<@+$!WbEeHJ)d4d5qJQ%>h?0dK7l}@8O#xykzCfa33?YNr^HM0mrrG0wXI^YG4_~2
zDAM|~FFI(LAI{$0JLJXl*U!@_J2sqovdd1_;O8y)5qbYFVmH(b9*((*KfVv|*EOgB
zHvrcoKA`%bUoJm*@5Z6HaDe>Y``hblpURNjxxmpXg4QLf$3-|MVbA?;M8bJ+Jii_l
zAlaDSKqsh+N-=A7qs%<@_8rk6(E~m%{{tuzrIyFZ?b7?~SryDRUB?8Bip8QCIY!r#
zmpUf1vpyb;XU{}>+eG`CI<snXH0k_^2={(Vfo=#91vN#^!}(IhM*bk_uuDay$h%0=
zup5-1M1ou%(`)`)03^7=Nax<$1<~0c{uTA5e6P2!C6c8O>fAKa1$#jiYHp&znQ|(%
zlfrkS(X*mpIJ4T&?J0YFWO?124Hv)aba%?F#%u3o?>en>3>!aXLjHZpn;d6?KAD!{
zbDF=^k0H<`?sUsmP^+y6c7W>6Aw<kcwUvK}st&T}u}gcw<-djA+?>Gmx2#-QHxE}%
zQr@nu1T*=+c3Kk>YMoDOHm&_q6@nWfo^fG72Bh5;m@|#niMTc>oC1?c<62^J&lL;i
z841B)ZjM0_6F}eLTXRmCRU7r9iKQVA)p>mEdHCgLH&tn*zo@^IyIQpkKc7YOMN%E!
z3j?OPBe!FVO}E&iE>H(DeQz3zNk1yivu6Y8#F|w5^LS=TyrL}6Ve8&T2l*G;+c8!T
zh!q_JnU*KP&(CW~Qu3eP`24EMB55QtX|s!MtFL8#{L{T=Li$;A`>S>G=YW_hWahoo
zH&S8zzTE##%qpgp6)B29qy5KdUSmfXx1MDdVn7dmc9ZL!@4rSuyD~(9S&kNJB~p#Y
zBUYBf6vxaO*(56JbyLqI^Qng1!h%Ne=Q=6N%NiLRq!D>AWm@rv1xA7Iw}E!nSV~;g
zV4RWH-}JZZX|3r2lvSLj!_y6<<d~g|mMQaURjg-KQ<%WRNgd9WdcR`2<Jd1Df!$O+
zjD>C*kFJA)7fG^EoUa1o8&}&p$QyDk)kfSVEhWiH4W<y4Tw{}HPS^tJ_)}LRepVr}
z@>TglJ8a{U%uLPIB+h!H5(-ZSdZHE+4=z0&Zp=ug)+r51TU@NFAAu#tELuMk3t8`L
zzKo8@Y^kCe)E))IW#xw}DML3`iK}^4$W9YoE5}*+)QOFmO)`f^wJWm<KO`it_2JD}
z1Ldx36*}l8^R*fIvZmAV8e*z4(W?`=eEWl}uWOuabo%O#0N*m_foOE4QIxK@ziS<|
z3!R_d7Rx*3s@kPXy5<!dKu^-EPRCWdJ>f9+8irw(buP;$rsY%8%VrP8$~>ISvGoV$
zji|pi8#EXv?y$p`5Qmz-ZTf2$zmX!kXH~;p$lLYl{b*ZR+iD3n*o_7$Eq=^S*eVRS
zZGu&5;B=&}5f7;~GvdJ0STfKCm0|{wO-LSgfx2fqL>qMj^zuS!b6Ryf^NzC4BH-82
zLJWU!S~L33g^w)DabX|v3nP!f&fxp_{vf#7Yum}z6GODsuP}$78fpOT4>kn8)x-XE
z>1OG{zw=0<pH>`++3#QV!$RzXGHkyNCmzacjZm;unG*h+q<VHnxX(RC)V*^uX=qCY
z-uv^}z3g`9UyjpAfr7G8(&fVHw#GHvmX5^+lTCT^fEe&OXfe8{hzJJe5Si8SNn@*S
z&abrI#6g9B4bf=k2-dJqs6jzxBc}LrW#&E>L7ZU9FV+BdkWQce_JPS_gZ!XrAxpeO
zKg-1He26&+XC>aUvnLuvbVm&FXy`UbPsS@r;odvhz(Aq1`N7NEU?hqM2NvJuPRz$E
zusx2AVbbg;VTV-g5n1LEpFSE4cvL1fT-`}5kCx1nPK~_m{4jg!Ws3B)RYAA7I7d&{
zGm)w^hj2L$n_k(stpw%RuMtf=)~Yl|CJ#kCmL^-0e+;lEAb^h&QaM_;sp*g!gko0H
zzu5|=lnZNmWW()8k~>HC+!WK3=eXaw5U5Ia5tck+BLLtnVu7_bT%`49!}EO+VbPf{
z@M{Wpn~m_Ofh&qQztu<0iN>QDUQ^s^Qe%nVFekZoZgif@bCD}^6HMqe+=XQ|KL1*H
zr9n5qcZ?$zWk}p+5YEQF@vP6)6LyB@%xp6`gJIS+t5V|IGF|0<kxcMrOT+6Ua!-^l
zZB;Mj=%}?j)^hN@(N@=H=~KU#4JzM+HIrhO%#<c=M$)$DQpyhl;RB$alW2@kU>5_3
zW-3d|Zx%qWd=d~Z2xqUM2QA?;pRa$vyyIKPNsAR?!v!!gkFMyfG8gs~>z?aP$p+jZ
zpA&w=!R2d)!BBB7jnEMV-tj3`^`9S?6^1J<xLd6^L7%_?qd682ja!ubz(r%xBS%m0
z1rFGK;hYb)TTrUQ+yHA|&gyWM0A1BXy_Oza=er3~bkSG}y}4ZI<XcalA(!tGi&==L
z8M#>dte%E6>RPi*m3r-pH2gtcjTlj`sP><1&2ExjXIn76YWkgEezi57DEbB3UXDpG
zD-#Fd6GAOsipR?Djv)WDqeRt<K@N8kq%91{gOuZ#=jxX8U_N@QONWXbj*5nVxnpEV
zAG;#@$+G;ug?QYESi^>`?`Ts|G#9v)v~ZY~v@p1%Mf$8RJ!}^a?I-t@syfZGTIwtG
zPcIeje%)-a3;&)RKw4XA5Mb&<7sk_XW1beh+17#;-kla}l(>=JkW?1}T#Ih5XbC#~
z;J&OYf;732yzsLRD*rTR1b$%nh4(S!_T?#lrtxPOYSv7OZxdsjn8jf+(g%Z2(Regr
zPXSH~*kN2ETEuXiVIzvssM)EvQ%w>+6>Z5e>AT>@^)oK!nCv};?QaXj$8?+R<mKD@
zD%hF3HZ%B5vBK*zL+6JXc>cct@(U=whX?=P@*PIT|2e$k{9hV1*Jw}1lD5Ha&;EH@
z2pSk8dXC+Mb&fLw8XB{W7qzxSBG3RC-+2F2m_<^hs;E5_oxQH5z7SM8;!L(YTIA0=
zTEyS&@$q@RMlZ?Bz5~PI^Z6b{o<mRk0hNTW@o)tAOXdFYaa)#`TOuOOZ}(JGq7vzg
z^L=x^*zK*dR{=WL20)LW6MDJ@*Ecsy-7>Hi>-(#6R8Hzbm_oX)FKi9p`qTaO@GO68
zoN?BsN5cn-NTazui<E+d_cw`EEnm?hb@%Jo2j%-E>QP<|q^x_Nh>&&+i#8il1EeZE
zE4Hu&fNx}WVT{O$k@?!(da%S-2i_>To-G?J@D+c)!>*V9uaj57vk(@Y%m;~!wCo4l
zU1MfVV{GzY@gFapyqpR}P6WZjq!y)kXEYBD%if4NuI>D{C<{(`tOjr0-<~pXSfTN`
zCjmZ_F?m)>kzWD+``5QA(LR?3P><q)mQ(@=6PC-*tkvhtd8oK%pa(Aala)eL2$<A3
zC3IPCM=BxYq|m4@nJhcr_)<<n`-;9Aa3Wy}jM0S4ntU3K*^02GlQGfeeaoPlIs%vx
z4Lc50fp)=p{u*#dO^l}^AQYO>0t||1+|rjxPXc6il8E?U3M*UA0M@@3g2CDS)h>J0
z0FxnXCr<<d!07L`9h6o~4M-sYM#@S`b9NrgSdW2>faZN$f%(z#c<S;!fOI2e0I_?c
z_saS*<Kk%m@n*<*>2E#Mf<Z9leM@CzN8}?%cnfX!uk<o%Y5B;04B%NCTrnu0E;X~O
zOj#^pFHby03t0gRP1KKbtW#-bY~%)3g(#WCdnNZiwhm95VwPBpvt@nG%PLEtJov<x
zdaMi16>w3d%YAr{(Ug)(@L0~*gE`_$>*x<0s%d$iS}FRx9~c?rHm!VjMy+y<+47kg
zQc81x+e+kuSUF}1t4UW?@Y1CzeJ|%lYnJDIsreuT*j7mv0Zf@<qC}|xQGH1k@Q9R_
zwz}lF_Ms<-@7*gIXgpB2*X6Azg^lficyfu`O%*hmq$*ou*}VJ%8ohKA@II4)_FSBY
zQQUa)7p^|nQ-9c9ZCSZ&HkTY&ZC1RiYLhET1!abDt%BrO<6kYsi9otqPD`^k1%rz6
zLyK^y${Ka`)D#E*!cV<hO#hW?EF|i#+$w5M7=c@0KF75eDOShHjTQs4w1?(ULxlR-
zIM*@<;5Cq)CT|r4j?B*#q<b(3%9r>@V_yMN0|RvtCQ>`QK(_ioZ19H(jcX(&i1hcc
zho|(n&usZ%nc-?93LzC18I4a6Qs`seqq|vt-bQl>2t=ZS)HaR_EWGB`8-fE{?vNzH
z;o@Ow(=A@toh3I7b{eGct0^k42%8@vN);sx@Ym6p@UtCYzDAv{bMRpe1VSUdXOb!n
z|0)sLUOR9~aYGfmW2=4EQLegDqKi9Px9EH^a!jMm9nC1=v48rVc#saELyd$bZifPt
zKP)*r17BO(^Z$n<whhu4ePee5=PH108>e%lAdL^KCG)>If0VnKQ_2^z^-!=Q)l&=i
z08D3LjGc>L0WzUW2vci5_TJUl6hps}B3yFE3ANL;^m%Ul+5}-LJ6EMK#Qd21UVKWU
z>OF&B%NrfXAQN5+=DUI%JtF6QaHqyf*>+d>ntkyS;9e0QmO?k;5rQ!ftLBYc003eh
zo4Kbr5TPfsuWIQyb{DzQ@!vHZjQ(<j@~Cp1E8xI^;Xur=nfbguS`6Kq$A2oAUnu;;
zTFT@@pCU-p`quxk`_Vzj|K~8-(ccQEi3@%WIXu^KZOeArU^Qe~LO7n@c(OlXCD9OB
zPdW;9?!@Y2y-Y+w*n-tU7AB7?^vu)aAmY*z)5KONcM!@|l}0urebU-eeTl_VP`SM_
z!>gtkVjbXX99#>INoS_U;cAaB=AX^PD?oM8?gno6-k#A(IOF7+0QVN*E@{`E08TOj
zsq!G!t>r<X(#AX=c6U`y87$Z%tSe*uDr9FYa7~R!ffxu+$3sPMl1pbS{vrfQ1Y+pG
z$K;xj!V)6#EvZ`rXbOQRXEds`n24s=)#*EyQuZ(~pxRJ~T<RNc6N=_~LCx2nN!O5$
zBFJIA$=6nRa~(Kq&?c%ug}hrMpUL4#0JLU-zi-Mm(}b*G7o%MuY2m@934S%NOb)*&
zw;iEPwf-R&-F}lB?K7Qs>du}*-Sl0Wv<>>LOR)UUa{4g0-%1X8VsZWj_$IURz$R__
z_>4iR_r}JyAypGys0cI4n^$)dImLK3t7V&pHBsu?)n!>}VQ~Cv-m59m(w~H?=uzOh
z-&9?8nDCY)ib@i0a`R5tKokarP5J08Q|&Jf6RJl#sWHk$5S19B-lOLj{DwQToJ!?b
z6dQN2Epi$-VF=VT))=;rxac_yc!rHS=XazfDoox{$1w?1p61wCpG)(akA_ARF$(<j
z4TD;_$_GKFIG|(bDSW$dQd>d>@YbJ77fk#Y>b0UbC|OH2J@t%Au6PXr&mfO9thhXd
ze<#&4T#_0+^z{03TWHcu#=vp!RBMXD1q~@55c%2D<Nk>A{~7=%_duYO8oGnHRdIyh
z%LTfk1P9gss5!&~MGMq>Cx2gqR0YHIm$0=V(@<_?Eeh1rcn}(J5+xU@P><@!%w{<&
z?;$P^lNdh<S3;{1RD!FbL!U6cbzIO<IVP413-GnM94HYiu5ep@5)<K>oKJEU6-n?x
z+(N__`Z(Y{jA1=G`pS$X6B}`nW8XRcs0JvJ85VnQb!Ju_MtX~?Fjbq2_YtW!XrI?p
zHiHX6$K8;kz9x3vg?0F7)BLk6yFN6bN#Qwg+bNX>iF?C+woUHYIZjiTgP%WwZ<Csd
zgeVbJ6P79KuBZFNQK2^7f`Q$_#PX(VY=~oinp+eZiroH&Y@6rhxip*8?8DvT!EbA^
zgWFij7q@7m8=;IA*X@65E^xj7LC9A1qKHU~sYO)H*g<R9Rzw2loo>(7`IFwT-~3pD
z6*Uz@`cX?%HTZb%><MS{9*q?~k(j5>kE)FIz9sp7Lp11)mUAS-?&}CaQB!H^KDmSV
zxY%pRpvygmrLBikD3H&yxEcGPQS6N97$UtxH9zzjcrTI~(-@ZwN8v(6_k8){nMi!X
z3)f6p>>-LEA2Z`bHGifVNKsiQ_r>K@dW9Esi1Gfbs_b%iKIfs1Uu)2sAmwf``dmV5
zE2s5kZ<Z21?VLp^wR_w<@CYmu{Fp8utqtm%bI>kSAnPVDV)U5I(5ca`9%KihBtn2Y
zfCVVo3Qc!LuOS4U1jVjDalRgRX~l*p@Bu-R!$7wcfI;!t1H#rYJ*}HhgEwzP{?HE#
zTj-PD@pKa!+A1VoziegbBwQ3O!|H%qZs@V~7;#NUq40H=8s)tfmW{FLnQ;A6NfpOK
zV)(uLEEgo08DZ*kdsj#rE&G)evDLmVDJ{OI<mi<rswrL_gU~tOc4F%M_gBW)sVo4%
zNk;KM7le33B9OG<VGrD}tI>i;9TMR3_Kly<h6f=vf@9Wq{H$#K!9QQU@Qo7dOzxSz
z55a@?C&olBvFGl|8;YkAmau6Xie?WIf4|DJ0d8-V%!sc!+OU^2U@g}e&9bQItS!Ax
zaJD=wiJf}=H;-7|k#-#PMPcI6{86lQ>f1Q#3riiOsk8i-azciQADhsxoof%JN!<&c
zqI3js=3)-5>EFc<!r?L510|Z_F;^AYg!vh790ftZeWv7kl6T{;1!}phdvRS{|HgWL
z^7VB!$%SC4Kn2S@biLvaXco>t?LUZrUUq%m8-2A)HX}CQP1N4HKC_Gddnwmd+qPZ}
zS%8Avr%6qpYCfRuYEh@`ZY+Jn%cRH8ushPRkj9)kJ4_X*-wJfE`O;_aEk!xRpwZwQ
zx^fJbTvw0012flnJVS%-48n4|@=m~}h!*TvkrIgv{+&T4jJ=!17BD$fb?mm(`9KBB
zl3vfPLK+v#kQsL5#y>CTbz$J$MunDshCi2uXh<DcOT7{1ggcO;NQ^`(H1&L1b4U>r
z6^c>oE?ZaGYlG_rotZKoOiO5L8x&<HYg(rD2S#w4FQ|m_Hn#%t?l0z))j*;b`gS>7
zql4o4JD-BB>j((QWH%uzWS~MIa4>o6T!;Y2uFz!}HoA(DuLHnY8oW}nw*7R*0{b85
z-^r9_Y6dVhm3Zo91m_%S@+N$!H@(gYTDfqRm+aLa`CnIPnNXk88eZJ1h%1|FCCVY6
zGi5^gZ<Z%BZY@@aC91sYRZZiU<#8xeTldB1`E{I>x)|5j>2&_MA7^qU>>muxEkvw&
z=Z6>mC<ogF1Q16Ew<Pcra*)PM3Gn?A@dY5rb^xWTp~am^2AVhyRQJ2%ywQPDBbgQa
zOyElt@?Q?DV%~b2MBL&1H*yu1RCQVFL>);sasjoQcxtaFGJd?yWBW3<+8Vw*NA({#
zziD@Yd+c2Lk2{QFM5u}WRDuqLX7;zZp5o!g7$$ylvMMBa_{<gCT2Dh{cPuj}=^GbM
z3ws1DhZ{q4J^a!(k<uup)*SDU+(Z+pJsu(?#Hs@wv2+)Q`eU|bltb1=G(Vx2eiJ!6
z`O~)U6YQhdIh63tYMa?fS~HF_^P?wV(d4u7wl`<!>@yH!id<z^GBDGFIqKz*Y+!eg
z9Aoy>Gm6&4PryZ-BkVQ{2mCIlx~?;({_Yh0h_sTTx^NSd-$Qo!e#m8~Uhi4TIMzMO
zB-aKkJz-s9ywg>f&u7F_qw(7{zphFdo7Wn9;!1N)KY0?dtT+~)u2SkaylXePRfg)_
zSteKKk5Nps)~%g+uNQp!=N8~`57mRhZa{<|nT%~W5W?N(v!Zk}F3mW*d~*xD0xO1>
zJ^Kq<hi9|J<~c$SRX*1Hm%s?k@izegqH4}<x{ACahWQF^Z~{uI9F7SYqI8Z~QKK~4
z2ZN<JGKv%rAG*eRa4!}pwEa8MHzK;IFvwksaAY~}ci}*WylY9GqXBHA10VrGisa&S
zs+CvX4sK}d*+b(GF{igd!yFpL<PMEv$=jMgSgm6{93HuUVv=<5_dTNt<z^bijAb7e
zcQ;K~(0{Yi>@Q108~UD0#S(cgaDDt}_>T}xJ+8e4e4jYiTA!Fsfws#br<+)D-9(`>
z)4q8#$UY~mxJ%+ZqjP^NDSqa((vOa)VY~4(=SqQfJHA~w^-cAG#-UDsn?m`^CJ#uT
z`|c|9^7?u|@tN+&LjP-x{y!Bu%>M+?|0{*gw6?Sz4(I<*<K@>Ey#?3DuyV5oc9LFc
z8^8fumwiE?cQu87d`(-bB2iQQ<D!ba29R^3CXFJQfby(p!uN&ocIWBo$?fB*=xINN
zy?=D~;)8Gp9QF$uCT0aj4$>qAV|P1P_nIvYT}SS2F_RLQ7Bq{u^Y8E>;dw&5sfLGf
zmks+|kn8O|L->%q1kIlEz<B~e*3v}7vg!L7JWld{y}V3Dn2GWCVDP(}1~ys7N?)=7
zXQoJ}4?wsBdH`=>f-{>R$qjaE8bSAta;~uG#~7%^VfD`^7#*=MAU$XKA;9fSEB}O$
z<2^}#!FoDeq2{AH8z`6w?wHxyPj!^U)I@wP@IMPoNF?;}Y+2ydpBfe5kM3h=X>xtE
z#N{q3e3YyzZ2Bd0iQJfE!*U*qwo=eOtU8c5>n=MTOAw_zaBGTZFG4(%Ae()>3ucrr
zs%dHq*Rz(~CZ&{Uc$5^2U<G8ZnxIzlCBhMt4m@@j*5YkYdJAWzulcJ@d2%<<r@ed8
zEJdM<^aO@hR>ygKN2=aAz&f7@NryD%|KTCq9$EGsZ=Q{CLZzj#M9~C^4bf;|=l}D=
z35})$zN<UlcG|Xxj|><CB_ql0?Zv+juL)V&jRO>3GkAB(A9h<tpKOUO(hY!qfaYD|
zL^v=|_oQE9ER|l!7ZXLO7(P`Udh^`J?*~(Se0e!m2jV**4=Fm{Wv}Q(4Q&Xqp~b7m
zUw48423w*+lK*Hq?VX=v@LZd_LQxu+Cv$mH#u}?pM0Rn~Bbz1~FjgC#Lle2)N5MPX
zB6{wHuaTUql%Q{*_%S1n79wLsPY<Q|`nS*?uK7|7E$E&z3>qd{e$`29X<J!WGrOAP
zuULd2r)J7Gh|%)d6h?q)2ooPe&EG<W(`_NdiL{VHYiq;_8t7x5#a^V%eO4eLx?*PD
z0#d}rsOOy5x9CI}f&5#@2;UVQ>3L-Bhk>Jp;zztyc$n1ZP($M~G7$PtmWBY2G`1B+
z;(w8M??jwZH@2zxHGo5!B4OHY^#L)GU7cUZ!30q97Pqq@6_VPHG9n1gz^=lEVUb{(
z*5fG>61~-OlN+KtX<7FH-6#VvM#*nIFHIqm`e)&l?|`1SMC%Q@z#Z6+5M|oLkFW>+
zZStCg^2d?e2Hmd%H*Q#u?t$Yzt@kyt&wP;P&$90Ok4$kH3H+EAPhFWi#q^j5bV2y3
zzN#)Xr}zsl(}TxNaW|qJ)17U}`5gc|5ezE-z)O5Wd^rDb_G;X?=@8h01>DTBGZrm@
zK2})S13#GljY~LAE(7PQ>6=x*DF#SRjmupvOEoFg!_zrGfA8=0=564B|L*|=#}cyz
zNc)%92b35cVSwV`?9;aJ+<{uVV10kea!}gT9RTh`|HhpShlRR9LA0HwpCNVeny&OZ
zOsqoSV64eM6O)I3iIP>ut=Xn}J62gZlPcRCw$ixKJ#{pF(H(W~#%&38^vA0G2VZut
ziVqm*@7^|-?dcLF!2t)UW=bj<bd^3Y%;W_GoQz>X3u-f)lt?BEg`DR}fB&+AS=Hav
zRI$AUkY=cm1(h9(&p@E?YDrQLY=ATX#Q6ds&5|)o*-Bh~Kd;$C-~RH|6W;j_WF<z}
zuzl`pT47w^nsDNP&1u1g4zF_^p>3m32p7jL_r$7vO}JExlY_OIB*eo0zi9;qjN_i{
zP@g7EfdouW^?5F!^XkhKU@IHJm1&^io#FcTK0|E~*)XZj{O_fgWk~hbU^t(MxWU^X
z^+0~D1#)BzIMx~cTM@d&hvMO!xt+`$nWl7muH1i*^0D|Q!6sTd=JaJ6K@X8Z;ySzR
z_I4FGdLdq`5HO56L{h%)4+y3~-&FF85M*bmXet{dy8}hh37xMgWr4+6yd;~9HR5d>
zWP*@Rb;c0-3o#-PtAq5dYz9naw5vW%$AjyhV<eC@g@U%qb~h)4@C>A7^K0m)Fr}K-
zQq*c3^1Wv@K3M9kQkPu5j-^Bn2v_thp}l8N{W<K!j;c;ZJxof>M}8EmrnOVhm#?0S
zGRitUidg?}WV)3EZ`O4t-Su!!*VY0(E|&tgJ&mbWC_wOS7*-IA_Hb^i&`Rl$YyLTW
z8$R~#->F=ypUp@Pu7PYb+0yL+GK<9AE^ZmW_{TVlBic&)h%{#{WnN9ta^m0kEUBh`
z?v8hvhp1+8q#^sT4$>&pux>vMlW$Q6RzT;fnyL8rTzR8o>V>A~wCNvKJ8}!Xih#-h
z!*6xI&U&pC9}C+i9QNL3jUTVGuH){PT1jj6rF_=Hc{!F%U)bjgEHvUuD<r@_7xwd6
zv(KB+by&OZQI|;JLOnIPKsnk)EQj60Rnf%DvnYCw)yt;agF~l2&)lHng<(VSOSc5q
z&iRW)&)mmiQTxL_kpVYgx7d|w0iUHsnKF3*Ue!m2-)OfaW4Prad=Bi)h=0m?fp6Y=
z4<%1rY;quX`Po@dl@7y4SpqmYW_es_w<~z4cRfe4KEbKPQDkjFRHQrIp~#jL+)U-l
z>Z!Hp9&|!^G1TXk`F9aj$S~0Mo`fiUb5qcd42U>3yfhW6+9z1M$J{luB*=#emqaYb
zaTJP>W)88Gw9xvK&Lh#nngQjpcy9XF7dBGs!W}{QPG1%TI8K!(tJ?6!wzJ>&2n|@e
z$-Z#!K!6wBJ*ef+Dr!6f4_y1u5oU%B$C9)M`KZqzqP@gfz*~X1-OD-}KaVI%u1UNy
z#RHLTWhiP#5tp~zC2JnmR1qja7xfpClSaRP{&Z(5XooXEMft_P>?ju2<;tA+<mR=x
zo2_D3J2j}(BpdvVEmLOwkmhuwsoi}Cl9JlbS)UStf)}(&%+ROZxk$`VNC$h1UG#Y4
zNxHHh-0O}e@%xT}|MkcB!z3c(`a<pMU9~^}`Qk3+qYfq%5&a7kOL2o_z$bx5KrU%4
zj^rKy^)PuUX?-XI_BQ-JX>8CDRNzcs>IUQr1Ulqmg<n+m!YLrMY0x&+3;mM85BZ00
zIoGuH6hm0tfV+y)a*O^p5)Ze%9Rj~tye{2(Z=CG*w#a(fN(isj!kGP(8Y|$Mb4`mH
zL^v5Zmo(`RG@=4C0I3^!IeJM<^~%IU6d7!%<BdR7y&e&ttdap-wIS;iRmVJDbP|3c
z1W;;vBqhXl`1G-)N7d!Z)JqiC-uPJ8C!Xu><{H9o%2TpUBZ+!+lch;35OrE&jJEiP
zU|K?=A-MaDWt8>bP1b3N>J?l`vRbkRTED+joD^N;euWVmU{7$f0%XzXIKosLF)pD3
zmk>;Rk^~>8lt`&~n$5Fi+HA{o){i(M(gih5cW}dT=K2-wYaL8zd4p!7YTNuJ@l>qW
zIGYK7d1hi;f&%}o78pf0nxJqT4EmzIPtUN{Xbf2N4|ar9OY^nSsC706J_*k$61Tay
zY-~t-1eWa87`7HNOR<I(pF=Up(LgH^RgK=+Ymn1|ZSiCv=ILY%dt_1N*xdd6LU*rP
zmMTqI!m9>anRfI=znVgf-GP7&Ct0jMuy#H3aHzng05QRz8~-8Da4n_vR|GVLIJVW0
zhE$oU@X#H;H~(be#W|&x0#X7aU823ehDkt}kj7{m=BA8&`60;f>X$Zd1+CX|#IF-m
z*x`e?qoWRd2CT0SV$ka~mngRZ=$`EYxUH=6?FII4LqXzAL(&!~3N}*0*zo$O9_H)j
zCRTB_fIHYGV*uAcO|b{O4NV`Nw;l_mYiBifp?Z47%~Wt+-riIXCITTh)H66G<wJ<9
z+%=y7Q<mEF#Pcl!g%@(!-htPyPX!Bz8&8Ht(^Tua?KwQpQ_>(v2HC-I*AQFj2|#`e
z-o%x^8Ou+H_GJ3~>W-8yZO^26{W%pt*;!37EZRZ!95Vhe72iTF0=*)h*eT49ud_pr
z41<Q^PGtrDNKOI<4MAdq9G*oF@LAyId+S0tYB$_i9(|gc=$(27%m4M%f=I;N?_wzZ
z)Uth-8!34zBVW+HJjYAP?Ghmy2Ss4lZIWK#Te|X++oMih1Wa-yV$g%-{PuryLzvi=
z`-f1K^CKs8YR=80Cp-0!<iT0U)Zz=p8*M#3dLM|(CGfM{{{Fya<=2}>)|rTVT!tZ9
zu&f^sMI7<6B{Kr-f#wN+iG#`hl(8RO*&}vrl4KdS=>`=Q$*TAM21^j~m3Jc?{UgP6
z^SlL>S0&WCVL}>78QZ|C_^;UcBMD*vG~2WJbaB^<fhuWsX$KHe#R0&rB518bwesDP
zqiC76fA54G3diBYmk_jcczFCw#;EvCLnN@1w=4Gx%twz1_}>bifAh!xS5e2r#=`Nx
zDtMOvLsDb=Pr-ACk5NN$0SN<Det{mpIrBn0M!7z2a$4VUwW}wUYMt@WRS7;eVA4jb
zh$MMWOxb$%b7lPc<NWFVF^YXL18LV^adYSe{v^0SFX3q+dkXN6UC!t0+~=F3q|s^X
zY(a^5?)Y?dZcC?!z|#Wo0cHPwz~X(qS;5{}yCgR3`t+6ksVGfEIIpz(pJ61M-p|A4
zD7`$BJjxl|^7r%Md7r#@<%)WMl3U-edn7(fpg<j0NFdt$_w((|=ORBiR7fKVAK(1#
zCyu09$RE1IuV?o4Z$W&&f)Nbw*S@8vQHJQNpT+Meur;Eq5SieFHLMkgNx6tUVKVMS
z28`gPeU6A6<|ZliJ~`B4h|nb2iKBpINlmLMkqn85WWhok5rPAa3>e6bcqMZs1mrWh
z$YNDr{5(kw7>jVmj-2z&5mzzYUUf`Wu|%wMD_`h$Jgvk`3oXs_xkVXTBA?Z^aA(*d
zHL97j%-xZ?wg<GQI~~)HbFW-f-pNy49qJi573cVMjT76kDG<TWQ#u<iDZ^7RqDHQA
zX6*3WCESq-DtaK8kC6Cd?2_a0F^(^;3AxL+?4#7@){-jT^O{rV+oet4q4yLDRM=Qc
zF&6ZcW{ykU1!E^kFm}u42(l$n*kw-_(MZ>PNRH8pp}6W@#j=LzAd{CWW^snUpxaYC
zwfrpI_$7*OkBQ>7ucqnEFyA=Ht3JD$z4L!=u_m~b<&-FvjL-QMhUS%~ylYCXw4&zQ
zq+CTwEvoOg^O2NTu`F}{cv#LsnPJ)FtB&f%yabV6MwX~1nk837&28P4uja7gm7sB1
z4Ln_w*?shWx)&8fDg{jAA76gR%?DqOJg6pJGLB*+19<>i&b?j^7h~35vQUhBO<an>
zlls$uU9#Tn5=99nnVcvtd<rTp`^wYmETk;B9a+(e?{@De#=NE(zMGQ1r0yQ=I4%Jz
zlWg|^c2`*MKg8B->_9=mSLyzP{)^0=5R@re0U}|`Z6*wQ^I5oXd;q1S+PQ<r@G<+6
z{uaH#&G8RPzd#1c0!X&Gy{uMaTbVZ{VL7m`NbQFMug93EfYYQdiKf2^%lmz=H?Z$A
zMT2qPmLf)DZ)1wri)+(|8s~7f<n^8bTo4GbCzFF=Fqx=!=ntn@p|ExN)b(tbpv#}G
ztQLduKDuaoyi3Albicp+uwbDukqpVPi5p7ws*5CgB?KLg=@|7i3Ug$FCeljIs*@)-
zv6>xnP1QIOzRtSuT_jf~HZg-929uI!w~TON1|K@zQ}lz%<v7SfU6xbpg9io5?Z{AY
z6*TDJnkF?d+0hY$3T>L&KBWFRug??wkaWuIu5b}aNGi(YlpIvc_#mgt$TakLy!LT`
zr$QC6@5LM#WeBEAYMF;MO^NC3j4k=DfawTvw9<*0X<D%pPIuPY0c4Z_Z?gp4sl?Tl
zaRnQ&T*Y6R6BsBH;{$fj)hi;&1s&CzJhi+6_5`NVw3M4d38iK1EXTfN&+JZ%=AT8K
zjJ4Yl%#UV;CY!Y+EHEhUc0l&d)fp(ucTUx-pQ4Wj`2QGtr|!(6tqV65Cl%YQ;EiqD
zwo|cf+qP}nwr$&XDmdBaT%3z<U+f>ST3hR4w$<n8qxT2gT*Y*#)0+5@hzkvMStPgY
zFC$(Gb8V~1B6y&@Vk!=J>1_ZjZxPoC#HB=vS^etzt7!2=q!1z(7mv0GRnacejkqli
zX`PMnI>es5k4(1|1h6vLONXS-h}N3hw6*vowCvt6Zo6aZVD}aBn1+>)LcLT%T%tA2
zpJPIE=;x4Lh6?E%QmpNEaqP+fVYJAuI^RWU0hLlfy&;%kIYUWgx-Tn9`n$Zd5Q0~!
zn*T;)Hhr(b-n-bv{5N%lo=x!c+ODkPEqYu8=27YaUkBbt)p`P+;%F8cDGa{^%<TgI
zG&TPC-~h(XPJuQioVh1Y>)hMTiDxRv+3j&FF?hfGjLATMvf!7b^D@TJA)P@dUQn7?
zDyBgx#NXoNbYQZ7u9JmB31IgwYi$L8o~K{}q@a3a`=tI%R(h!wgJ_mIcw1vz6q~A?
zdj#4g_VYXqfEwK<&_O^sB;p1Zq@z5AB9C>=l3cC*m2#8{9@O#luCn?L1*MLL5uc_U
ziA_s%;{g(<KW(fko3D;lY;+bVlYU<}cid0}Wp6hK(y^bj{=M1`5%%E*Dj+}ra-M(y
zIVQk~Zw@9$nlBbZUZ+GrC?}E?Dgb3xD}bk-glZVrPj2rdjta4&kY&pSb6w!bGs0;L
zAj`%WGEE3SCFFw6K+#WFfsZ<gt%cem8+0J&_wAv~=9K!oGl{b1Q^W(vRt%Ij=Bf}|
z1?A?k^IGjMdwAGX<IRig@5nD0`11@f_#+v-PezzTR5-7~Tm-rT#bbqopI_1ylDWK$
zm`@7Fw%cX=7s<T<4IT<pgJwt&gn~kmENwU$wYcp(ROhHbvU2wF&CV{CWS}5jf#%jx
z)g>U7$N{HwJ0*Z8vlFxmtc(R)=(xY7iVKz;a{_uXGG@L+S?R!~P^NDGmq}*Qk<6p8
zai<h=-%+DE7rVc)F#KpKnjd*qIq^BjNS(FbgDw4u5qr|gzmt&UM>wqU8Ai-)s~Wuv
z+r%i~_r{%H;0fA?D7q@};OZ9;#t>NIzeHFTnFS)6pzFp(PYe{;jS(l@zWV7C;w66a
zJ%5%hOETH5UWWE8gCB?xR#rnLLu%jT5}LNF{l`&_9hl{ir?rCX<BX06L19ksfDV$=
z?%el5Io3EtZ0~^DS0s2m8;X$`X8SpI6Vaz@!UxqXoDp{DrV{e-y<YDYPnZ^pL0eLX
zgL^7K?X)auUyg(~_^LPQOao})cZ3&yv9ioH+Cf(@nX<|r!k2(H<T%q1if89EX6|-=
zD)W6Y+&1mhQ4-5h9?nlZIZmBwK$ZB#c<XdY{$vMsMc!brz34(7G|W2X4#tA;xKbB1
z?Idh1C}lKDsvg}q4djZ>N(ok+!s<*}zp^NV^5X*KqxATuX5u<s@8Mi80=O&_1vveD
zbCfik1APQ*J}Jv1OdPLbd8^F{-|9tHc&I2-G@fOPYEfn!pj|9GF6|h?Z&`Y7DpB<^
zP`PMaR{u$rkKx)==D}R~JgDm%@b{jPBUaVHjikWgDK4Gmx;2<QI8ku*sF!_DC#q0(
zr3em>GF+uEaq;Ce(D5a})2S$=BR0$f>&&3<Cc+ZSo1U;h+cx{p^CjC=Zmu}9VempJ
zE4oWtzw@~ni7<*FRD8fi4s8u7MUQuDFtd`hV|1WW3%V6+wY+Gqe&{21N9ZzK*~SzY
zK=n|a?8X-)5XMLGI61;aT?f#4GT%RSPNZ90ubLmbg`srk502b1e7(p`LRQ}2kJ5-h
z0;4LzcLwm5%3+wW4a|%{VtkW~vLbn^*OoAR?w~C=dvP3&v3Gd|hyn%EwN%)QK#*oW
z;9-l?UT9FE&uK9i_gMr{K~U+j%pFE!A(Nuj0+~w);Z@fklR9(wPPWw>LY>TY^qECt
z!vmOZW=Pi#pYt>UQpQ;}Pg0a=IHey`-+_JV&Z}^iMVYu~LuKn4-aenM@4Uvw%|QQ6
zr855C-UkL&R>uEHrA}dwMPsu*SJ#x{o2E}3fI~)!*}H&rv!7}13hst+L%itivrpd4
zcx*-}D<nx&=>e-s3dhCR?2O*-{p$@$*}8Lj*;DV+OIzq4-+S61xP^%J`A_8o#D6Lu
zU}|gzZak)o`c6XOZL3ju(0VvO7GvNg*fI%kRSEI%sY1On$#U-4cX7lYZvO$Q2GSt?
zDFZCJ?wE8_A74jkYjT{q2q|tAnQ~-nI(nSB%h%b2lwoyuoF75R`w|lZ5bhpMat2zq
zPJoF5jwk0>KmIX*ET|1+HiAC9(0aZ?e7!=EQQic92bb9x;r}=Y-UG17K%B!MdO@S_
zlD=;QX)Zi|RRoRqkpVq;gA$XGhQr#y6hHmG5ZH}lP>ZD2k;}wZ0+?b|2+hWLekN7|
z(zB&dzF39wj6n08Q3cJWi^PO&>!Kvkf+1X=cYU_vJ*SVqqo6;Qf`xr~0ayxm>N<fv
zw_Zw1zM<h1t8^BH&cZO7Z%qW?)Tv4PSw}$W;wK2L#+m=9AbtfIjsEr|K^_*3!v9O?
zBXMKe*l#B-31`-BlP$=q9du^A!^e8o5w#<_@FcNqV>}Q1;Kyijm=y`+YnSC#lN~OT
z9Z&OKnz2v(8Q(>|`go`w(n+H>8HbT7+WTj!*-Z##YUR)=hv>rZL|&&BVPs5-n^Vy)
zp_`p@{G468^;{kBIt8qO(D^uN=R{Va7;dNg7cf@X#F3%qB2k&zz@c45lMyGgFr`kM
zrM3m2&cHh3IR1a%BFF%4S4M@51X<#1%6P?Hv1J|9szcxhzSE$Pp3&A#AT5!X*_{B$
zQwmZZN}PNHa@JxgGD!E<%H&}n5i<sz17=N|Hzw6K?tTt_n{c~;Egob~0RB=@Ojo^2
z(%PPm0;B0jo!%B6YI+#GiYJ5AM4hc3ol#Q0d0GH*Y%+i67`8E<1wOEv)V;~yUgsd~
z0-SH5kiXWH+h|szIKged$BG1K<fz`u`Idr1$H?uq@iEw;ahW(^po(4?)~2aSGQ>ts
zXhN@A5LQyIHKD52M^KY^>|m?S;S*|znb`|p;9K45M#ab8`UBZT@+#&G8k5+ySqOa5
zkSC53^Ivj=;V=x=x62oEd}lHgVOk_YJ!Pv!;L~nxSk8cTnuzf4IK;hHth61?^J^P~
zk!!&QparvMxN%AndU;DtLzNwku;HCb1xr5OE9TnN{}TQFBOxMVhBg|}8L0&fs`&Ns
z!Q)8W)4iD02wlE%^~Y^xN1`E0FM9(@r`sG!9InWl8Ba#P|IlMpVDey(3WRx2cK5o@
z)<LJb!rdvCjj^?I0UO3kP0P|^$xj%130K2OkL$r!vEtesfs%e9`i)bVpDtef!n_Vk
z@xoG$kf_+j-jQCuAhk$=0!M)@BQgwU4Y_TkaIk)OMy_+eWRFl^3w417xw}`kSN3U}
zn>8;@Bm+EZoXmWts_+LTxE;9xJOypnd@sWUBEZlz<w`P^A+Je<iQ#fi8qke4D1l#$
zk{oc0*EUfk$FrrTKGQqFIb;{~l#-MJ;wF<A-{PzGp^ba20CeTdg>Fq<jc)9(n?Y`5
zL%Pr^X_h9TLtSK6B|E9lP>d8q7{@Dkoz^L&zh!vA4rLVWGA`Mkg?1m0SMv`_A-_t)
z`W<)XkWxWRVH|41T3QihT+qUGhDq2=XzAibRuM%`b$NS#sT#ZviBTALm=hO3k=tGU
z^*yQBtk~TvZ$CM`LuN`T+@;*WEWZ~0^HFLr?oMR!Bu-3Q#_=ur%|dnk<@}q5b1VL8
zXM+fpU~kskXjkAPjV&TmPiloj=7|;0a1tA;evS?Vk8;^~_$YzaqAc%=fG}J?f|;W1
z=*3YkYf;!<14Spvt57=dLHr;da-sS$Q*0{|0teDq+HIfPx(;G~-C|>;iY`4YsEg-$
z<IC%HPBgn1oW238F!hx*0yPc&d1`2KR`PU~G60=bLq+ASlSGL`hGQUX9~>?Pqje2w
zs$BPxBq0Dk&qH*y-yTW+xJ^Vb+6_Xj(+laJQ>l$~56ze$uS6WdLgQNET(m<+rM7Rh
zSW3XCJlRZnm)r<WXnY~`$0pQdmf-JyZG}vKS)`?iKX5MZ-PxEJF4^Us3A0C*niSP#
z9eSs$+BR(>=jZp9Q@ET{JUoX;9Hq3IPn&B(62o%=i>R>Ls_oJu@g)-i5>xwoBO2>>
zbZ4T)>@!g4v8SPS_%>ZZ*jA1gR~WY$N%ixQD}tcyN`+gIj`p%_TK46hMTlw3!sfRL
zp4hZYLLdOSq@&&YN&>ggPkE^z<iWpT`R2;WZq3YAzat~MS?<X2jB(=ea<F(rFs$IY
z#M&zDF#S!R81~g=U?y(V1>uP`;*FVq+BUNpMc;`rn7Q@to|4u4eX#nLp1Rd~i(HEp
zQluLBC`kFFh{S^_4+~%L*@_)ZbXnYWy=2l(W;G4XGV{Hg!>luvOKChtKM$>h|CFyz
z3gyFys`}vT8Pe*k;IFsPUxC#)dgC~?S6X^nch9l9K&F)PEGF_8f+337oH@xdrZ1ci
zLOPd-HYk)#Kvl1PjVc%=M@nO9b8BE0A?nIlwI^g#?hC5%-i6(DLNVZ3lkybLq2AG-
zaT^bDBL{*T3z27iW6_DbVG0!2|5nc*W}=5NI1B!UbxiwRnTw);eTYM<)B*qf^>vj}
z;{zEy@g=BO_ci=SQZM=~SqJslRzY);?w~_IG+7iU^<=ykC>lyhToWzXuVw(QR;I6i
zj=6x>bdk}Pz^H-Eus)ge7&bjG5`@39z^n)VO0Tcn5;2y>bQYUm+~N#%<wW`MCn=U)
zvS024*I_eo?XX|YGORW(h&xR9!IW(MI%%s+hoBOc+~6jLHB6EVimcixOwGo)KCfBA
zfI>oE7^}e!+>^8gJ>^!I#|ckZ2}1~vy4Zfl=y_^76u17_Z!QP}T^kPNlp(emg;TxL
z(_Q3rU@RZ}Krq`E!16lewzxIRF@`#&6%_YVZeyjzmP9$?6}G6M`k!hWE~fuf+feA|
zs4{|NshB%a(-qLNVUc1C2z6fDot<(m#2T*@I265*EE<`xVSiY93|>z}*8jaGa&xMF
ziuGde8WVl4{>?|v-DZnZBQou3u_qP<1;<^n=(+q?sr_3;&0u8Co{4l_tdVFYqk{QY
zgO+)&%d3<2I{Rl;Izvg1GJ3(+mj#BbB>iz76a(e8Ii&D6qv%`<(xX*I>#NQVjqZA?
zWRVk~VFv6j5rto-q&evMsYrLBuJhhM<X7=w;V)}CJx#ZVwpfJT+-&&h-^q@FbZwJ$
zd!+*m0me&L&Ocy1CdKE5rOl%FgF?Oy_sO{{=eN`XQL8;2snYxhc`>oypg}U(a@bLg
zmJ<R$S+M5kN~Nxtd_^#9r3Ug%F3mmz5+*+pDlM{S)8H$o(~GpMRaO!HUdX)*APv%>
zY`zaRTJEs7@Vie-(;?D@F~Xi~Ht7>WWV5iad|fb<shk9BN6>ON;3$JvDZ@-@G_#XJ
zLzMgY>n7sXn-G(k6P(~iaNOuDIn{rt)I2Af(SwC%<+H4*!6{NJiUHWv1~8Th`m-cw
zQ6h73MAPr00Op2z(Ui>t*hoi|o6pcV_F%6}GtGYwlTst=2EH4ek>(6#Y*;sWIowSx
zVy1w1Wo7?Ot{&(pNt2aNGLh@=U!>U|?f+FG|KFrN0~6c-D3N~#xy1h;N1B%oOMDU8
zlu55+17y;w!RxpEIBA<9+!%?EPuWYsw(|2|c>AOYs_=Iqg+if)y%;~Z9i8eNTHPB0
zaOp_HboiVeco?7D*D-P)CHvV(L@>+O5BiS}34rQq^(tr)&??h5cIS8tZ7+-h-A7Vg
z3#~@&3g5`hDvj6xF7EcuzK2*C6*;-*3dO}~TKLud^Tf#uIu@9btI^Wa!NXBaijI;Q
zAkO~vX(Ymf#ut&!lP(#OB)G>1+uP?&)i@SI@$V+Cw|B;FX(Lgv#mN1M?&&T5#~mk9
z0P_HE_M_-^Fq{wE_b}Wt2DSjV0iV(pk^zV9->4|d-@l=f<@8?~R>{-$h0=_wBatXQ
znsriO1hw-e_Lvsi133ms2yDcu7-HHFk9Bfs#1_$n?VkP0=`NB2r=(8m@KNoJF5?lW
z))K&JN%9tt`wHSEnQ;8T3AcX?^cN{nx?4Ex9mrArF;xV_MAISssL((0=7uzoNl*x9
zoqZ4*Jnk_3#E>E$Tf5VYy&~jr5uLzbErW8dNGL#o8Z7++*@OGFlay+7f{_Ta{r(R9
zzM@o_vXBMUBeDV6bYfP-cp*nReFIhKB0Ny=js#Kz`B_4|Sed~tBq_E@_QVV!qP3Bc
zjp3_DsE?owiAd9f`AeB1$FGpfvA_QIWkUUOZdgToVrzC!yV6=*2vNlbmI||ILs@(B
z%fQPNBkb3Jl<E+7OmH!<H8Q)eEkm(?2TbpL2(jXPDDqxCv5pM%lwIfsj=E6XBD|~`
zeq3k!7*b#_+m-y)|BTjhgS5>55QkDO`oXuy#c=8%D{K<eu4Gnt#F$$o<p^|oUUW-?
zk7%h8G2avz<ihx{0JyGKt?7jqi2&edDqA>fgk1b$eoAi#H@v~uW3oHpnUhFzv-JCo
z;vmlShitYuwi9$!C~1ToD&Z5-ueUH6WKn^3rPhe@W?ud3;#X`Ze<-GtiiQ?Ai%Z%3
zYXHPZItN{mj~tl3s-!27XIGE2L482mhp|x+jRZAo5@Dtm6OlLD`*MWc7X2>z;Rne8
z$Xu;t-9YOue#h7L@M5>Us~csbhcyYoeymrJ2-yPFPzIfY1xdUJiujI-6s=dIBD|l9
zc&>1a4b>ekn&S@`vq=WqCW+wG51Ho-%%b5ML-Blctg)rYvfOOtN$wVAqilZdSqH`d
zu^haD{?5F@xCU^sEXCOC01@3eF8*oVn`9a{W<tN(_=c+2_x2ZcpD^pxFvOFQH21m(
z#)B9*aY&e<hPHpBfO2WH^KPqE1>_)c+mYK|hCCGg@{~yLOihZ$J_K23skGDPTn<ZH
zFe9t+W0Qm4R66)!gers8s1gDw9;A!XmBit&V_2s7*D8nxO{MA9->~h|oEbEOSOX4g
ziZ$mq-2YZBmX8tNaRk3P#JQrp2W(M4lixAkCFNdGJ97IDFivK3lW*gRjCZ`txSW_Y
zzPW00{>@1g&p)$4q|6EFz3xn)n^3B+;(-k&#;9*O6s8Oe&(Py9C`GaTG<-QqdiB4v
zL)J)med8~D7DUXaQa3GjeZ9Q<QAp-WBr3ub#b`*5*bZsDUmNHhFD(x1QG=V9NiITI
z)@xXl_s*(I|MFZW2TFoBT+?NWqvV<*Z^u)O%rEfvm3PF!*F)hRF)BDk<h4QZ`koM>
z4=LT;>2c(g+K&ru&qY00iwZTD`GbFPYd3I|+OB`WK25zfjKq%Bok`DBH9M({ObVT&
zjPpc8J-~c3ui$ZW%cVJ!Joh|)T5sftV5O5=Ve1_dDel36`8KJ==x=5t+l4$3r$8R2
zLqnE7mc!an=Sh+mIo1{v&vV{ui|$R5xyhAJaaq3jmRE)<#vj=1lph`T(u@+11ZF?D
z7q~WbS$Qyez2J<lSR1{b^EhE!dnWU~RmPjDK)4(Gj0#j{zKbT~>k!0O3GJ&H>g7g`
z1sMch6)sFm2-M^$-*+wUNA!pD8EZa%+i|}izLYh~^sIJH>!r&QRh!}m5^GakNWDE7
zbd1xn?l>NokXDPr&Tg;^gxrol36jV@4Vv2{Z`mm!3^+3NFqAAIj@XMNkW=5`#P2p{
z4*M;5fhbKuG!xWMJut@-IM3gvhG+V4SP}+SbpSahoMv7kpVF1j^uwVdY@7QFwP2ki
zncsHmA!wpGPWKFI84Ot<THmmClp(2#KG$PQ+Aof%GGRy4FCVOj{giVb^>Toy0&zNW
zmpX?#xHg9}HH4LF)TY=sLU_)GIjoAXWiZiO$T$^EoY@Px*@kFByci>TzRNWv7G(_V
z1O$_=XW6iC%K$O`Tdb+cXXxW@do`T?bA6`;O~{Th`&PZ^0n;D1Z>RXA0l?9cOiq06
zPlwO3`GeU-Kykt`iR5j+j&5D^$u*jK)|<Bv+<T$Cb~1DmxJ@LqM9w_GB3~4<&xXs{
z9-s%_X{4IG;_W_a;6OB~0o^VK#e;_lZtE)ypD+PKxMK>jkZx2E1qmG#mVhais$zP}
zj7XuQ6ZEF6aFKB2y4*^9gQdgD^$E`&!cCI-VCRmQa1-@Xnt3(#;kvO|Dbc8^zT!ac
zkX9Bx#MG|lvlAgve%5{ieVt#7)=C&IOpmV4H>7aw0W1B;9O?Q9=g1zpE@v^L)HW2?
zh*=JzyRi3{Ut&@m^ykn<>#wBbGjwunf}XFmdrY2FVwRF<Ewjv&Z-g;$2TdeJawryb
zI)Dj~B*~Uzo!+bt=Tc_v&KMFID<p`a@kN=Jjoiqd9acK2*TpEwuLd!4qBb?v$>G^|
zVZt;nb*#}MYO7i=B8HHy)}g$n*uJpciM6Tx!v#GDeM?h@eknft`voGjp7@XEsT1f4
znjH$V+Bk4mnX$x>Pk5>LJ@~ukpQXL^sYRAk_<%;>?u|FP5Aj_Pnr=4;=JQ$!pLl(+
z-bGT!)q`rv&6<eA?2f2IqQv=kw5;l!CvFNKj2zU>Egggqvb|}rx=JP1Q{brWT?=Dh
zUDB6LedITmU3id2;LeGMMAvCuLktj4`kUL&@9ReE%zy#4s>xp>m=Y6;{JlD8*avu>
z&{bceNe3E;I@9>?{n9|E(5wRqm9T(BIN6`67q&@_A~QsEc#p~xCq<$Sa43cHFdeR;
z#1ETeq%DSg;!+*G-FJl}q1*`ev)7_!&BB4-gq%)xqFR!^*DWF5F<QjB{t)1XT0{y!
zWkrK=FmsRnyWdzTwwcdCvjGp8GQS;ZNwXZK#z#|f#MM2Q!&J;`1`XV4x&LQWAX){!
z%}|aKwUWMLU9AtR<`nqrck1Pl_kR<BO#fv@%1FoXzs*QBIe(mdr0EmYQ+Svuq6k{O
z*<QyVG_Z*aYuetZ%l07)8`;-CnsE5I#Krhh>s{a~_dn%oHx%B7_x~A^=IB=I((-yq
z;J`FMM`Cn;VfgCUzN52oFPhT<Nl9XKd&Bf}+tD!nT#{oQ0A~!38}r)QIwWdXzaG(c
z@68dm``PZqS2&{(e=^72LpP>>;&v4lT4XqR@f<s=Q+B$&-S*<<k}0iX5u+xPIxzTM
z{)|JP$~5eow4!hy<YaRS5~JMtvD}4oL@LK-lkX#ji5d|lX}WWIU*+gXLVI_{-$mh1
zEL^$;Mk{gCHe&#QYi_B<#xtM$-#2Eabn$g*zcVU;P!q_5cYHcu(aHV9DZyta#koJ;
za`s|Ud>WM`8#D<T9IBGaC*d+Ze?e4@Ug{<IC!uR>qOwBz36Xw3oV?_@-OuzVN-&&R
zawhCf1h4esurhXU8d}z9F-+?7R2u4ONv3JJ#=%=)jHh|OGM|R-BuA#1d6Gddc+FLg
zo09T38<QAV?8U1i2MMKFvdEa6z4({-8U<TQ{C^u>#VI|rdNm(C!+E^-OrEbdx3o8`
z*n!k;(PWjW^Y3{2Jdnqg5r$+)5*)*GxMXz+jL}jF9h)SQ)z?arFV>bo$P!3)^A!V^
z02I*OgWlmn0|Tv9gxzs19p;xY>2@}W5J;?oF|rK!oc?_Dh}8H~?pelMPS~9J!9UJN
zyy|&d@jEM}_VH$oG0WuYIjPDthE;{~X9qJoFJqfC8^Mz*bkf-`IB8XXrw@_9UFD9g
zw^<BXqSF`=y<&U2QI5kX;J5*n*_124ulffC=)?P1aDV6gT0Enukc-45QEpf|$Ev6a
zS0BQuv9I%HVXRix6x<2OzEf>x0HlPyZVnHN4`-rbmoX3*Q($t0MaYI(><<fYC}N!H
zN|mj=ILn$#rl+y2$0ygEU@FLEVpVTos;;B{>gPO?vNGvFQA+Hr^fFBGpJqV0zSD;0
z7y|82%XCg1l$oYt&>wPWM=6w1PWzrAEz-S(4Z?!fA<R*%jpJ~Do!uyWd}t<YK$9}}
zjGs&=(g?Fo>u}qMRc9xiDhnFXN8)V2hzt5>_`*U_;FgF>r8sTc20ERKFu4i^9(Ef7
z@2KlN%r^q*;5rIy1uQNF<HZbi|GJq(?BE|-JA(0?XK2+YA~u0=W?$=iY68m3CG}@>
zLkW#Bry+3BmXrvA*v;QeD_^;f0r`pQ;An}gS(ts`Wx}3RW979<rh)jFcEtLGXxFOk
zATRJWymV(Dnbj=I+%~?*5{Wb$Ne*?GPinmRkgD*6%Xm4f>cV8Zu4;nL-$X%*OF2rt
zvBlAUuhAcy+TwsKTbH@t#12i0af9HEO%o5*^5h$<YiVf{aJ_R%KA=`8fxpbJ0MAD;
z*aE~9UB8nLR>JO>J)i^J=@Rvo>M579jl`TMeNQ5$wGnQy035N`75<L@T~3A69m@S+
zs;H<2RBRuKzhqe}bH#I-+G9aobCAiQ#UHVKfB%Wp+9>`nC@|tnugtEA{cFOsrZVk=
z>9l7ONpHpwN{j>lfJ&|<NVrR7hfzrLmE^W5juPnhT1<jTPI`@0NGaX~k-<4ZT2FXI
z$*5^LLa$YV-W4CThKTXZVTp#ICcl{5TpGRh&Ex5lr>hVL+fncL)^L#`t&)rtB}m!&
zQuy(t+!IGfXv(OWonHAOD_ZQL6~1nOgfxtRq?;9N_-Lu58TbSyw*Kc8>)(&y$em6^
zd|Xqv-kyxL<ym_OsRf3*BP%nv%bT6z_tZE?D{z~xUl=i+%CZvM@%(`WOGAwX1SMYn
zbt;?rqDHh@@{~^YYy~WTF(lhRrGAs^3LAzPCsd!m0a0E@%NGiBjAIVjOvwErE8<v)
z9-bE=+GqV>v*cIGN)sp3J6;YcC3@;<q_H7MoIuJqj?V9BGKH4AG6lXiL8NfkcnduK
zJX^RcQX)vyrJ3TsNk(5G;kAGgHk~+4zjGl%>YYzt;HE$_H_wP8XuFQ_4|Y(K_VHK2
zD$V`uKj~piFPN9Vq@BpO^{Tjw`^#Qe+8Ii=4BFBVdZ_q~m>s_86b7Kg3Fp%jgWICJ
z-^TX@6k7OGt_f?LLr62KX~#sw`L||28MrN@q#Wt73n)*#*4w7<1>XIk!?scEXftnL
z$fdv-7__YD<6yvk)-M;3ESBTNQ$$_loWt6O0dn}yQ6yIvKgkX#gFa&2D#xR&8XRB8
zcYY|97%nBbiR|eu)O9P=z+pd{h7GOK-G=3U_ucaE?05?8jT}XlL9;03=rZmCFEq#z
zX=mS=4b5GCxij?w(kSbvE6{fPfl(08`!mOWnvwx+%sd3s@G(A7<!bvHAUo2*C(0;0
zHgx7eU(42VLHWFNc(EMZ@Bl!|B;VY)xCj=S6!Et_cUB<Se1L?xMe!T?@|vnUu}PJi
z@D#^9!2Mm!igo>Ar__9;gn7jom+|xZI^-R67`Kot&4M}Xoz7Rs^pXB?kjtXoxBx<)
z9j<TEf#z}R5gDDWN}TxoA_V-}ND?O#Q@2_L83^-G*SkXlc!0M$l_9B)b!q;(&r)Rx
z0XBd++8SknQ1^Lo2ZR3BP)4>HL*yWAUxTzXb`Lglvo4N_vB47g%N;6}bC@-_<rg2=
zM&uF1U0GIyXflCNug}Q}pe^(z8#n)h(9~GY{!Im}*EZ&aBx6<@bur!rQASaVc(5$}
zV;f%p69fs77Mi|s{NgJX_K~y$BMQ~iVRXmz<uVba!NhtX=UiUoxLP%9zyN3dZ6XC8
z3?;`HPDvV(s=(-OT(605T=`*cXIzyKOJ8Jn?Nxwi3&s!~YiV)dz2o2yD19ciKSg`u
zx-vUp@ZUUsNH@)ac>gY2fTafiQ>%Z#gQNzpGi@-!9`<a`rp^x$Wjx%NX~<oA(7-ts
zIpF0Ut3@}ax#2WQ)n(#m3K=!h$fIh}@ko}sF?Ofr*4Ue94?j3*;qPVL<yYW>!Jb?e
z+=k5NF6IE+m9#4#jftfUkM|w1!H{b;YMEjBtW&ft*t)^lH0uM>!XiKWU$xZ#9i}ld
z(EX2~Canp3gErW$f0U0PQ?a`p<3|(|RxV4I^JiO&m85Bw^AVGzJ@~PpZg1`I+G7oJ
zlI=wNKZFJ#$A^g?+K>CIv-g8D8+-+ADhCa<){oZ%q94ck8#)?3s#}dPHE#d@;N)n<
zWV&znBAj-pE%CMfb#1T1`;91Q01?;+b^j<rd7nKh0N%4}9|Q8>_5pipAWqC*GQgqh
zPQR1x(cR<i>Sr}F69e^?ru^jvU-sMUT%B2qZtC;sB}aA-f&?r9j%ImJ(c$euH-$I*
zONZQWJhSZ<>id(0*bLS|4`fW&=OD#rKc;6_?`-SH$L@5d9;WBx<Q`s2FQ(Rj3n`gg
zrAIhb*VpYN>{38La;EjZC4p@;?U(T;i*jGu3iq0+luRzYok62@Bvce+qtJS~+aH->
zl*F1@45swo6ttzfA+o+Xj&$)Zp8nv4uHNH24`lkmaU9W+#)?<9^hrb=4H=tpS!k2~
zuYdN;rhf7kXn5{nA~TBm8@hL>JRv45zZDB&%-G0d7cy_egzSuAxipBdwAYF3hFYhA
z=P+bYVi9h&C(=-bDG!h-*VC&`SlC+T!rOJ~<_1F=mNMDiQ_E!{97UCP(+=(vlt@Sp
zx^am0dwi0*lh-`nISNS}<RnXq+6qYaQWeh=8bNyl1NY?_t=K5Txeoc&GmqUw(e@IZ
z-xCb|^i|c^M+J5O5er;5V*>OiSK7Y&Do7CYDDli+D8c7_>%Th7srQx1{YZDEoJdj;
zZHo_ImE54Q-i1?hwPZ{HM9M|;W0)p<z+^}t4${Qg)xpCDicje(VGlh1b%{wtB!o3Y
zS+lJ~fopdSVvN8VDx_RVHMHx)x)bTD`=bjBt<J#GfY+fXoGO%^iS_bG8-K$!e#ji2
zm}PYZBUCTy9=Cmg1#i^7X!v!g`4jxPU!zsvpX9w5zNbYCGhC6#MQz1N&8MmPC4m>j
zE%#|`sOm^|WP`+GCK0kx;zE#qFArrTwRQu6E(3#oj!h=Rsr80c2#8a_EvZousOUMu
zAp1#qhfyWEuz*Qpj%pRpcb4h|Rsd2E<R9j;Pb>P(+YGg}2mvmD<ReU3j?D2lSmY$l
zDH`LoHcb=}WJV97?%d%RVX~bY{F0jgh`jottDU!1JJ+P|cvl}eLw;-Y8T+ce`}&4_
z$q7f{=+{v4(-YC1-mKn1dCk;xsJZjAyh1C(pBL?<=o;tKd)o5)DE;GN2M<TAJ@={{
zJFH$Doi#rNbYIXMO<aE(3rTPG;IdlU6|o7}d%KGJGEWmpto}VWa}Q%VTup3rJVgVz
z0W?nr{M@31w&_&TB2{n{erIQDVek%1!v#S7+Zs*wyq4loMy$`2gO)*kH5_^7veRIa
zrYe6fx^RucZh@UOl)2r~3RDp2AYTU%AQgI4=r8c6NiX@p!p+X|)cUUly3;>cc@=C-
z6EDM?iYz)Sb<T4Uy)J5#U%qvFF!0`~ZXFj;Mi{QAfFTEFTvC72wYNF?h4n8#WD_qX
zET^@liJBH1QC;zW+d^XDHaqP$3+s80p<->P79sSQbS&ubob_N|_L%|V10>$z#>-r9
zGNVfmZvB4ArKsTYQ@$<PJ>GoHDoyd@<THU-<S`?r^+|W5T46A%_lBYLc?9t>hEz(P
z{rn$QQS-j_A{mM@^m5(XrkkrtN+tfQ_q_7yQwcXaXyM_E#7?sEq1CRLq9Gr}_?y?p
zYvG<fS-25t(WoGtx+j-vA=#XCJu0hmAEmBULcG<2Vt?YM$X2w<3^C5CLw(WsFf}(T
zV$tO<JKPi^2pIK!$I7h;#>%GbU4<U%9p*6nLh`Q4Of{ztl}9lYn#Xgm?mc+8=Dl^t
z$5=P+?WzK#ON4ri@N^cg`~-a*+j9{A5x14xCyyVL`3P`bg`_QEBrN6ASF4}O))V1)
zaW(6q8Q;EMlHXfcBplkGmPj;?Xi#bHT>o8T&<LmWM`MX(xVTs`a;VSYAb=ycJ*ETz
ziEIGWZF_t`5hQ4n&Nrf@Ofe}*af_O~OaM+7(WTGEJeh-s9;<U2gHFi+A?&SFh+r+g
zT3`YAk4ZrqJ?FehA$$}abTB--W}UHq`Xh&Tnm3Puet;1xic8W~G*5tn4H6e9$|809
zUkG8zt5ebSU;e7RdnQA0j;3X>NE;k(WZhW?6ar$@R%?qJpL6L5d5^(o6zGm5@D|h-
z;1cjolSSiJvsjP*hAN>BunuMjo|Y}#)6A1a{AEyHvR?m;ks#Ow6Qgt<*`;;M5Qzu^
zeIOlYcIjWWDZr@!s(`C>F7p}zQSJs+cdUeV@sk8v1Y(tm(G_Ai56)#qcG^AeDNN6v
z3KM(AumI~qrtXfOQWGNjWKCVV%t^N|O5@!+TK=WXsdIjWP(0JSm8^7-bavnd>tzKx
zSe)CF=0GgM$Re5GGHZ_6mcP-~Wypu-2~0)tB1fm@(YfhqS{(dmi<1R>v|z7V4-`;g
zfJRGI+zDmcQgc=9cBBfbGMGvSuS(mEm)h}cNr98#x*!uHclU~=R`nB$xas<PXGBkx
z)FI_m&FRqsFneOx#CM{ZQG>H%0@3Pv*^l9qt!z;xkYD{}n>I(t;eNXaBpB?KiOT!0
zcjKLhcrymfv+Z46<n$r)?uCw(%osGUPseAk5=B{w?YbXINCV)(%gH4$U8>lj`eX%C
z;c*@FTvu#=PJ9aBG~b)LgP$}|Nyp;Q5@;Cpnm{0@>-1t5;0SmV+gONQBu3C02rYt@
zWAwwQp+dMM+34l2Yk}=FID7JLHMNZek*#SPsyIqPxTJ=aTd-+&=^SFKb5Q>|#MRgb
z<^EP$BJp3%4qNh5ySEmIA6)^W=&E^7m5<!;OUV*c!W!U<c5f8!*B*(@J{Dy3s;K5e
ziHKiv0`Dt9TSNO83z(;E4+mJlqD9Jh?`p~IQ3{dgHUA5OJo1IlZaB@m3PX<xYWr4i
z1T!VB#@0Z5qm!17JFlHH-q>l3aH0CATuy{z?2Xa&tdBe{-H|PK7L*G1T_~rNx&;sI
zTO04Owh-NFI&CUL(QQ2HBj>d1XJ!<fB;#zGX($o{ixYBFyVbKMHK0HWr$OIXlF!hF
z<BC}$bbww2Jf4<RCWQALW?JCJfEL>>%%=Tg2dzQpLru2iEKvUoT$d~zf1gbPBv9@a
z&R(TNvk=Ef_3IMhd60+*_-V-greaVE8LBN9NfO}`4{oztppvY6mAgxqVC7DF`S+Fp
zekB8wxD(v}9(TU1V^E_Rgmax8Zy;+|<Rm=wZS82SHhMAN3yFqy^CtKdf*<{bAf!A6
zCl*%1v_OPSFrT2qU0cCiiC{iZzZkf*D-)P1+SA%4or>kN6>f2)U7qJsfWI<eirq@t
z&)@P}u1*$T=1~l=+yuJvoLyC~`Sr^S#(c>+apfID-Qen!eB?=_^43)%VvOBK#T2&>
zH}7eJH~(TMQW{pN%iGJ;Oqt9;Jq0D7wJPN|^g!98?aqbD=3SSy<K!K>uY+O^u!GE=
zm2BMYcG@Uxs4q1u$WTp2R;tea2`*M)|E0mK{>qoeJ5y>MC4HsOHY?vn>XZx03WxkU
zW5s8!_eLDv<$^}>r@cWDQ0_wZCer4^p|j)T;EY4uU%9Jte;aHiviyOO6Kxk7cRc;0
zZm?q}g~tM$#63NfRA=m_G^Ab;`gWQx9RNpis$0TqvY3afpc@*~@&`s>_E3PgnU=#%
zmcq=Zg~geQs_Y(I4RPnXJPJM~tkXasrZYm#{^?^#%CD;6K+1egfkyuM3sqONN6H<f
zeg8!9qMpL+Ei5P-wB3gIX-}l)iogL)yfzcz6q~X8V(s2`4Ix?M)-3Z1p|W6r^*$vf
zlnFvw{eU%|-b^73AdUG5THX<$y`=WOfyb(aKL!WVyLo?7NxAjamB_*royEYcH22~+
zovi1gwB^W3hi`%_{1ZOE;c7JzT2j!S`l>Gyx0cytAbgy^;;Ynsnm8@YvV-5IkKU^&
zX$^L{=U=Px`2xoaTtBGIV{SdswoKr){>Omg3wmk!jrN@&0O4{@zU|;62|e=%#@HTW
zG^MZOh|f0q&Zv}Mob@Sm<3=rH0^MI0v>O_#2(gZ6gR*N3<fdJlO5kLAjnw~;ag6r+
zj34n4w&ZWL{1tR&xiWqHq-^hA<%OWlgmhwKb=#OM%XGe{Pu<D&D&nu)J4>a3ON%Ls
zqLc%LiIyAivaQQR4<}t19=z$T(2JcNAFfaQlq`ki|8|{W{=cmj3`|V_<8HtYHtmq*
zdRuo7A452?6kX0Vx1!j>cavqKg`i|hnuarEilhhs{Ihn#xvc{HcjSgqT0VKVxW((^
z$Y6;L?MX~ZPBjwQKqWA7am)@`E_&t{a0!1EDM)W);Mu{y$BQtByfco4pv)g~Ox{lN
z_4!i-s3ZcP%`crEj6lUeWYzX<ND)`IcQ5e=$ap#1GPCN1=>d>E2zQA4r<Rx;U00%?
zkk{1b=fRN=EpeHICXuqt_xsnrZZ0TkRVvW_=ghZMPtQj|?r14{!x#GJXS1&&_EaSu
z-bqc*-Pz9Ly$Nvgw2Tkf?2QpF&!@{f8Fg*YDgz#ZB++se;c<fAbbDZPzT7;0^tCq%
zCt$(ftRotg&B0vgXMYyh40%PPOeRANTtK|PDwX(9Jwpr`MI2i^0L2Xaulw^sA$9m8
zBxp)WsgV9DbwH2zoYpHVMb}*cpN({C;G@JlA!f~$_+xNRemS^6JT%MX`C@hG9(fPT
zV6HFbFZg5-3erlF!*ILndK{qN6QSsjr_!s~`aW)G_b-o9BEOR<d5@_5I*#43dyRlr
zO29j0dZa(L{3^0Kc_#~DkBLo0BFHd*5gkxG=yaoJb^g(IlxN+m@JZ+*Pj@sbZQ-I-
zn_Oe0i^-Kqx4MT!nNr=p=oVcjo+;>*`*ng@s1cZ`RieAhm)2GI+eMob%twt6+aEnz
zohHZ7NrB;7g6fIEtZMqtvNjTBa;1!N#e{JoQfs$aA!9{Hp_v2m*2u?z)Vdyznla+f
z0|Ioh`w$u07t&YO$V^u-Z06zaB~0)RUV`a7LG|b2#X<^T#1VkAjbE3YRuU(m-rJxz
zQcww)`i;GU#g3`s7n7<oj4jzM-u;PeCcw8L>GSpBH;4Cj)|R$w!;a4Yit0*Y>lChE
zQp$V{#Jma0T`vI;m-gtdJj3|Lv)~8Ple#(i+v94anKb{|oqhcOcW0@iz?-_O4)cKG
zVA1gE_M7zP5ut0=aN74xb70^e%q7A7?e5ZmpLJmVNjMw%N;tDA)W>xB;eI0apf?Ju
z|3KaQe3&KngU~_63_i~P{UN0co6<(Ux$zg^$AkK-Dt?$spS%{up!}*io=BjQ1aQ45
zl`H(0-R1a?g^T)GdzSrpvhs%Lv`Vq<ia~G7AnK1^5Hfu>l#T{?#Kz-G&6wa^@*8_}
z$Q0?tQ^s*hZQmNS7NcJg#&a85xN@)v3t=9>=j+1&OT_z4MOA^}bf-3cS1^H+-$E}s
z<Ht8FPe8{H`?BV!EUX-hO8u35RvVm`c}HgZqSe|)wW<x<kqe+ZhYjMTpo?z0m=!Zd
z;p)~FH6Kh&&bf1TDOeRxvzK}xLzONC!>lVswV14F;18-z&F+E-TT6%6VmKNgjA0y0
z?PT?M;b8#OmGUzw)jF<p-U(v+UA<1TvV|v`c{=U^hXWYQl?*rQ(P&84*4s`78m%tX
zv5Ena+$J+($gs=S2U#9GQX9+Jab_kwMhh?bt7pTI)ExgpQH-h6XV8<ARqAdF<!ErB
zkRHv+XWkU<(J?G~J3zFd^9I6OfN9)}|CGjcv*?ch6oA-e)H9}iXpApw)L;M``FX*<
zePF#dT|HNt5U@I6nlUeoCsJ!=Cr`joS{*VFYtJBaAB9iaa#y<&k026sV$G~RuPurw
z05R*~r)`J@!6bTkS`#O^%-mw~y0V66CCcya-WJZ#A&JL*^D<@}{M+6wA^4EsAsd$j
zqzbX6ui=zf8O-pxgvhQpgo3*`&V-n4tT+7aT`Hk^*FoXT!XOv|`3y*B?1wInG)-qz
zI65uR>47}>m({mP|APRAW+8(i5{g`}0hXDF+kkMx?+O~qV<wjBEiEVJ|9wxDFBoP6
zBBGjYnl;N;KSvlO5_R1X8CJf=LBSgL%C=rY+2X9g=Kmxo)c-5YqPP-n40hd3k#`De
zx{j-;3-l#*&&WEm$pL&jjfvKU$1zVKi7#u?u^2*v+@^z(yEoMcsCW%qL+C6oI38zF
zBr}V7fAaP$pXCL(rR$QO`5%p!t@x@XaZR<;YSWv-15Hj4RO>bouD%wn;>tXd3cYs2
zTv7?@S~2%NOb$+&<*hQsav_NqejL|UuZ@#V5>kAkFG6&KR`dBFK2BrTt!YDD`-Znf
zGW!AnNvc))VCu4Hwi~85V5<;~XhUyHbk`=c;Z$D115+799&2JPX0=|!nlB{4eXjra
zmxUnz0sO+G4DUG)68=F#)b`>E!`gHJPCf0zLm{WKcmCB|M6>-Ljj{V*b%7-2o9W(!
zMi}Ii@l4LHAVO1tbqAOE+;NOm&q6$><~7P3NiUw3U+8b;mTN1$N5WnAQAzGU)S@iE
zSgG8!scfgol`diqQf^*%D(OqLB7VibAa{|0>UsAhnqTSRZtMjUNv-e(T<I3d&+x~M
zIDQEyqhG>HH(o~=1?Hc5uUL112i5ftR4M!&pKk##ZLH>X`r(yJnkTEBhdr9KS|-BZ
zbT59-N)>nYIg+>in9hC)V>h3U%Fg#9q~t@I;e}sx^fhPpJT7}PDhSDr-L3ybh8omY
zS(pQs7z;f7In+iX{jP~`*!9o)O`w|pLXv)_a9s+FV`X*89gm|@LbwxF%|s^Icqy1V
z47SX^s(CK3@FApvabL@6WyH+?W^&`-U0FxM!X^nm2bOq-sBEjc$8@Jo0#&+<eoS}Y
zWwk9Xt04YX>35Qew<n*|_FtZWUP(a859egxaT%%F7z=}LXkVk(5_M@FwE!K6tYvlO
zUNHPd#We}X$&VOalH}<juy|hKJG6xDaDZMz89yxt=)GNT%^)6SnNbw)8$r~VakRI0
zjj|+Uw?WkHa2uXx>h?Twj#rTOJcDIj+q29aaS$DGaJ*P4ek$AJWgK*~GpEUJ?g#|5
zo#-G>@f!RUR%Rk2R=M&yCTGx|d{ACEgcf<!nY8SR*SZAbWN^d@Yb{zLH|2aq_}(n+
zzLT4>x`_+W1at~HWS0E?;_p>kRp6487x4NJ3Mv*5J6Po+?ZzZt?&(Lc@RHPe<K=Q1
z?}*b?I2gQZqGcw4TwSCkWX)I7cr;FtI=^K$>t*>P7$>-4Xfr>F8;~b;5;dXt<j{|{
zmB?XUCK8!X%}3|#5%JT6;)&mcZz9<7ZMSiT7;Y&2)zcC99rnaFUZub9=h$*uXn8r$
zmAEsUWA6tszeSwb{vPQ|mDEhcYa-pw)S<z2w<m*lXv3|@sq){19m{|5{25sOxAcCO
z2B*VCGt%`ZrBb#cM&IAN&O29;_$o&66vk}vIPs(mLxsXtv+JE1;K2a3_<FIql3X+S
z-s_f|A%*+f6MnpIuFP$`?v7j@wmn!jC)a-mK#xI%Jp^ZY%n%R+RQ*4Das#xo3KWtn
zVKmevMB33`FOIA;F_76RKwiIZKTesrU@sPfs)?)Q?QlPRrvI6e#KoGL+0+tNL*&Ch
z;+~%9?s$8~VxGF7AHeIR?ov(XHOl<XSC>aZm(FCCU#M|JbP=3^A*2gj2I6)I8}PW-
zGzeT${tDGZw!Vt!j1$d23zp1b{4~(oGYzVH(^N)C>H;UOHbEtk-4WSL5zFG4sOOb$
zOv=GqR#xA%5}Oa7sLgSOON<eY4|Iq;WY7Z9Rm~2jzZZgHTsQyjcZzMTj#)Jn@xc@~
z;|M?5Gn<dccaEbK=9Iql*HLRn@4(5C-j|st&-rpl12@!eySH*3&ym?m=wDbOQ{Vr9
zV#Li@$sTz=FBO9AG!)--{Z{IR!s8anb^7%Fn?X@B)scJ6U?4O_d5&zH*F2HwAf@I6
zNu*}<#6~*Hx#xfF_<dG0;}(&&g(L^sEqp{hJ!LhA1g8Co_UZ|nnyiy+qSgKV0lDn`
zF5B6*r^4&SyXpPvNeyONIBTMm6MR_A&kGw?Q=GO?pI1Yn>_OQ(bK&?)6=<((zqS0|
z>N(bwS2SjVMKvj{s@)QwqDm{_x7WK*9k9A6XvnrBCtpI9IEqmqutx+rkRjV?Di?_%
z$<tYSrNw*2cM?#_fz!>zQ54V_XWLI1do@82bgKl7y~l5<%0o72lZw$rzNZYH8<~rl
z@yzM+@g>)ib$tITrKbnwHhI8aXAU$^M!yk4^l2iB4s`mNP|Vn6k+34Qr9M?L4=*n%
zy=L3_L<~;Sz{BZ-q)A|zI)OnvNtMhk^MV#_POQ=GltYH<Ozu+^P875c^Ij9A+H4Dy
zA<ja2p<)u*UbUNfU(s0MI;V^<-p-mZyM(Oki}PQ&D+x>-QBK;}g4#O+*r-B4x7{R5
zTBD%!1ymTVG~dp+6oTWGTK=A90Q$xx%IKQLEdx_MigsN}R159p!o|kGcX3}H7=_pC
z)lJB&7O|Y4B81?ZLVWP|+vn}bfNbV@YsCxSajX)aT6nS{u@@a2`J*MHC|WN}KaAh=
zhpCYg$uxSS@m~R{)X-4d>a^EXu$9KjueXyK5RdyO$$~(M*bF;~+xf**;lzIMbsAz=
zBk=OAgqpiagu_ZWrYhlqw$fNZ73P<MN!b!aXC(Dm%Q=6zenoCt;Weuj0ok*H##smX
zCRW2tBynPh{V>HB!^oajj3%<}!1A!K&8KRf->A7izagva^7lphi=z3^-MaaVn{m~R
z$y`2jFN~)yE8{Xe$&og_(}{!FVCw}2r^j>#)|K%Ktwl+*qe>YkI)9n>f`(O<9Fs(z
z(G$!RY+@ft7}FZ4U^g;w7WwAFA3cx)S1>l*Dy>i!xOMnD5sFWiJVia+EX^m$h#J?X
zw98N_Z$wFzb6u92MKjAar=Ce^+6XXU&Wx8&K+p=6?ZquU=eIZ(hGTYLWL+N&EE)nk
z02nNYOET9qk`GV!Jh6%Gy3WPH%iy6AO)-I5a^s=J$StVCKUts0OL`Lw=fu!##M>?9
zAPxh1lZ~(RV0dOY5hh5Wi>_ldW+qe(N1+wWo_xn)%~u~#JvMWjnKY#J39#B<m3;#D
zl4Pu~5u!x+6p9A>RNx`b@Rt01#t{z}l~aHh=Z&cT^SeNZsjUFB0U&D)MLfK`g&{z}
zGVJ+aW^m~{i}Dq@E`X;v4k?NnTs3UabcA(9xhRmAKKHehYE*&l8EWruzcU}i5ll}o
zepxWd9So_R3DMI@>o)V3|CY&JVMpOU+W;Dm)(TS>=9|>0bWdLV@o21mcfFV&_N118
z9hZJ?w<WRjgiQiFZY2GJ8;nkbWO@6h2F+u*k_jd;E#UgS3e~H5xgSEN(A0(eeCX^2
z=&G?6jx=|sg*D(!uNUCTg8Z=NBH{P6fov&bA9<kQKb1GFLQ4XR2epQ?9UE&bNRUXg
zKffJVH0Qv0y1X5!|JNiotp0uJz7rrN`DkliJlclhL<(s_>W=eN%J@#-s>Awg!Go+i
z18fCSN6u5G(qq#QJ@SM))fDT)^BbIWcwWZ2FV(}gNj(<22kLdtxG&Wxv1g7-tf^PF
z2(znBeTdVy);-^}_M!8v7vDIl-oTH*%GBhA>l}R?Ey!=VAn66c=-!f4-D`sHR}|9l
zD?ElW!F->(Pj8_=A`eSfp?4U*f@}mXq5#73z$)1qhTYO=LGMLby@Nb9E{A@Zk$%OU
z6Ua_p`H)~jxikTeL36miI68wtb<Tn9c^xsYRS5^5N*~k;RU4zB%+^%o^RnQximMq9
zG(&;)O+otq!`VAPXVz>DgB=?kJL%Zg6Wg}^#I}=;J9g4BI<}p3Y;<hfHvYWt#sAIB
zotZncR;{znI;YOAt*Y94*Qs6KxJlX@2_b$eX1|A7!I?;NT>SNhpaJhZ#5#6@=d2-W
zGH9z)fGQT-g}cN<o9Kj-5~eQJZHhU+GbULP1_s}#gK#_RkuR%UK*8q(*;vf_Tzsr}
zMKZJ%f_csk`&q^AK9EQF&7HvnTE+A0OM8p<rNeO>8p@@fzf}pJ<zqXci5RaC-v+``
z%d>DXXp6b#2GUQ(jX|ccX3H*q$zkHt{n3nPdMd=oug=M5L~Za(f4z5mpduD7#5Fo3
zx=2X|-04~p_SgL`3{e*I3cb6Q_7qz@+wPZwsW1qbU;V^#y}<uVpBp({f9PA{DCI@M
zya~^Mk$T>RtOYiGJ`V!BFY%oq)>?F<V-r*GO^<t<QF66A^yI;OR!$ngd8emJGDOg!
zi?4ExOR#`%#=^iq_3m+~P9F~>rm&6MXcSc@WLyW|V8WoyPe*nCXXIn+0=5U^u=(j1
z`mXo=AUjv7$b3vd96inzt0ykwj(fUktV5}LLfg?}8EU{XLp18psiT#Ll|3l=Rz^5o
zbSqPaII4b@Ij^9TU;HpCzjR4%hrcc!*~FWB(uV#9!8-&T^1Ii6w3Pls2#k^Kzgy97
zNaJ-{Wk(M2dV@_?`Uc+is3kHozM4rVze=Wf&@5#X!{YFjL-O4RqYny(ASTDWj`K<c
zIif!hbFu$>?aSiu`S|$J(|#L80XM~vfA9H|Fq96_6N5zO-<gh7r+EF&*LkIWB=b|e
zn^Pri$Lw76sXH~R3%r8cQ~lmrw{VY~+(Pw>euWP?czglO?NBTQsU)9xp7h@QzW+Y@
z8*iPzt0Qh$k*Xa*zZdwtadME}Wmjmc9SB<ea&{xzT!juTy7&##bH?W>|Hb}hX7(@@
z281PU9!Hr8WH^Tvm!ccdj6CgaZdHCpKYv5WDe3xPbkNrE;D94=l)BlnNKH`<d}K!J
z+C3GHFtw_17fjB_XlP6;E5Xib4YIT6kV{(C0y*s#LanyPa4Ss~K2Y4pt8c?h3^pnj
zx_iIvq4p&cZsWRlMEODf<M3cQ`t<$S7Y1yBQHIC4n-N;ca=7|b!Toyp)S!ow;nzI_
zN6mCy4oOkU_%+MA%o%++DmYI6K*N^^QLzTjiwgJ|8L!S>W+w90?m-n_SmF^kR(a6`
z(z%rDdN$-=YR8Kz;a3)Kt%=13+)0<J{A3=ruv@C}TmHo;+TXi=PBZ)LbX*#AviUIJ
zDK{y>BvQ|)56WZ|IT`h`Xaq<RLNB=RG`_5D3ii32@>#MHKv-saq#92vRyB^17n7|b
zz#h4XC~8|J)pkH3Fi*ftignUq&j$43agPTLB|^M<H$%m<3-`Fw$cn0;mxI+VT|pv`
z;w$-Pm-MC~!%B;QJ;hH-(Wk%J*)ypPbvPTfE6k(QwugX8;dj=`X)t?yrE2OO15|+5
z&Rszw*>Dzo&-6Fnh7X5Do&iU>mTW>9GM&T~QHdORK~|JItc~~AEr@MUG^^V|Ai@IM
z;}e*IJsT|yhVqLNxs@VL=e0xrjz#_O4O9$EwN@xc73YfPhxL8CpvVAn3)0t;U<Weg
zozCP+<zv2EI1ACfO48b;8z6B;(*u(uCh;;&cU!2?E~mUnx>%#vI?=X~`)Vk3?qLK<
zoU*Kl-yB!%+70Aa+?^dh5j<$_*V8WtX%g$xsahoU@qZ&JLrCa0tmV-~oPkqdCW=J;
zus$XXs0^x|dxAv87vAMlVAaQ0Rd7ZS)vc^zkeF%o+Q`W>N?i>NCl=hKp*GDyF>%Z)
zs&!mxT6=*+kIZ!>q}~$OghNNyRvy;+vqC;)dV#A#6f){l?2oQzB-)Rl($nSZtEq__
z!|@4;*m&p@5^Il@s$y<hxU@=fmY^wF()c=qvLau|xNb3JRWOCF;uRY@Z+2V7m}%4J
zCqhiF=fc69#)8U<faHzEZ7NlSkqUj~zeqPS-dqh0Q_ig>c5EdzIu__)c3V032n_on
z=tM}20Wrkof$JMeM{_DdLX&^x`Y9l1_|z9<CL)0tUmfWPF=8QAQZ)u0-6pSf;yehz
zYg$4l=4aA@X6&(LVBG>DM^RDn+a~t01lC1)8h4rc_rX}!pt?K9TQ?lU&5;f#F55p_
z{^#~ptg7OHS4Z~Xi-a3dmUnwoNS-HeAfycCv=PycP}O{Rjzk<1E`d<?G3q!4ueXmG
zN*+s`MJl^h8Onw6?-|$rZ&b7Wn!qbYcSEO4`|vc&(b?sr98Y;lf;oS@_OSIB#6#Jd
z8`kS2C7S(Gvq4VMH)U)6YyYFI<}k<$B|jE)06pEUXC<bag}D>a`isUx=Yj3D<eCA+
zBlgWC95(0R57pu8(Zxchg%%`5kR4zC{+Jp`+ja?$EjlgT;V8F^X{EV|vuwATaHl%e
zNOUo|(pP)8$&ZbmH6vq0kF2HE{oXZR4X+oU=4N-Q58l4fm9_%x-jk9qD#|K~%@7C(
zo4wlX?JXXWb~A|}L6p&t>~deB5_4tp%8A-o_pI6A>E~lAl+!GMho=&tjrEvhvDt3<
z!fQ^Rz8d_F_}Trj^Q62#!}9Ytk5(9m#yGWzkh+)wvDqeX_fvBYDnkG`wFbM7ap5Yi
z^#FjVOU5p5`#v7*ivorlaZ`5fl*VD*GHLe@^SP0#j5PLHNxWDaU1|E!Y3i-|vN9f8
z%jkR#yJGtuJBuqV=ifJ^teh9Wd2kceHfPPWPRT)`#S_%{=z2u#@meWV)nr{%q|6VB
zbS>rX2rtg@gZ=8+qg<dfzv7Hx9Pq>tY(`{8oJ=utSDKSd(%lNtG`QGCS#G>9q-q;(
z8PgHPB<Fb|bg?&ZcJ)ZoU@iHWk1@<s%UCuZq&&1%URSS5J&NMyR)UV|ZtJEcmsVca
ze2ocH9-@iXwFr)6tRz?+Xa}3r21~7<hOWt~Q)MZsm0Ql)QBrcV2(A~rsaQf2sr}GY
zLm&f<Shzj;Da*i+l%@BRw%reu9*2W@By&n}By*aBW#moikmW$(I30v(C3W9%*s+mI
z^72NBpyTRj7`Zf|!MO3Pj5R+|f7c`;S3nZ&uqLE4Qo(`pGQq(!vIRbVyu@asVXrVQ
zsG-u<6HrQ15L|ob?q}&<mc(dn!?no{<X|1ChXlz`I;t^967CWMP4)R!IooZy=_q$s
zN?1w8yzKcM){;3Jqpbx0BErLu=V-tE9uh&Zfq^FcrXe~qV~ysE9rpK;s&)ny0wUPn
zh^<8_t82&Kf`~i`At3u}PFr&WwYlSgx_bT_?KF|xxmImG=3Wl!8rB(QiQcYm>St|V
zXF4R)GOMbx?cSkatvjXulR?4pA3`qxw*M|ud;{*NEoswLWz}K6GjdA#HVX|BQ50*)
zgTe1~2$a4nstI~L6QryD;f<n4>-e{fK?1RgBU4_D3!}Tst<m417Z-YbYaQIVpW7{@
zu(0@i2*7zaZ`JKvOXduM3~>4{k4`RcP!DpBZiEPj0mMG=Z+1teh%uD3709<LVZHfr
z+f<c-s6JU12#+qeA57hUq=dQ^QOjp0sxX7}vHW;nuaA21<s&<}Vc0nLS<RR1Ey%!Z
zA?eUr22AM543kA<J0WQ=<Zz(IGX%>l%>zt8&vKH0{veO<y}|+2PPhJG31VM^jEM48
z&Oe^3<-kF$aL9JPeBITD>09(bmQ~@3y$~S>tIL2HMLQHjdHB#ifRB~@1|1Y+UbAPt
zPK{<?3%N9!-euA;6UsUfBY2*lOgj^v)12=EBq|{BXYMwb2NwS%a<D=k1IoTN%PW&&
zisM~OqcNlG<YX9zoYMKfHkMG;ZeW;H6^dD`e$?brie5W`8y@D`VrvtDveQ=lcy-H!
z#!jE-jk8AdoPYDD6{of?e$Dju&154x`dr$4zrLiLvRNwownGvY((3|Ac2rbgqVACp
zP!9PSlJ=e(6XrCPg}YkoCBbs)vV{~0`R1XH@R-$?bN810+TSPrbkC>9cX~59FoW};
z($Cw}lUMuxEdv!ZNp!!Qhk~A3R-JD0M+Cx<oSm{5GubM={Fm@h0c43k>Ss>ige5K3
zC9xF6F6%=+X!QbJm38F!EIMxOHqBpt2AmgLdw_Tpv(OFMo!AP*X*SD^alZ+$oN@g*
zvkxM41qekoJb>$Z^|S|YnkDzN#wR-f-LR3CinpT$4%2L$M?Hf~H>=5%H7DDLKv?U(
zD?#+D!zkYMH`mb(wbBu(a^Y~V&)a}B7&JmH0pZNyPe9CMay3Mlg9!-<(DxZ$dhVXu
z_uM=BSAigH1IsHVOLj$JGP}$sMVZ8N5B6C<=T(Rzv?o?5=YHU6#dg^&QVysn1zIl^
z=O3PE#nRyxfObMHndK)BBI^jBGfoo{5Jd6LNKBemVjV#`Vd7W>H*hGmZAeg&z>&DR
zP-X<gkkG~L9r>UWDnJyuh(qL^^4aw%3)3$OUf+R7BaeRI?7nfKOS#3{vJn~Ok6_4*
zM2}!raavBhVbmj>%AJ%6PSka(;WgEE!dlt+*iJiPP;Sh2dsDsgoT<!|gt971!;ECB
z1+hSQxr_Yh^tda68_8yTY`gta6PG03+Uf~x-jV10<sEK?hDOW7i+PmR)x}<qa|{w;
zmu{m~MGPgy*4}b0i!ZR-R`mGem9@Fxqifa2_Hbpf+>82E#0}j~3OUP;Wzwv3$03Rv
z(J`Ptkd&o>nEms&^6++UWhd<Wly9faMG`Qsr^0-53NWbit?5l*bi&DF;uv>`!6mm;
z*UDWn${x>ncT}<=<3%Q-;#=CALSk|QY=YX%m-yAg2O-xYRZo)nKP4(s<l40_mz?oX
zv(3^G@WiwR16D|%t^iy?bd`SC52%-!pom{pk|sRK^9lo&IN44dK>+Pyd!Z!4wD6zX
z%ScHxbw;YFdq!7a#MFZp!n(pJ+AP~gXIks`q<%~C5EC#sh6lZ8OzU@Ebawc(zD$~w
z2}CvIB0tAexxIh?g#u3pA7Cj!<?#5H-FwiM?OA_Bnh5J2@GT#&$FIu|yOo=8rlt}d
zO$JcVL!Pg+nh#A=qEFJB+=_Sd8Pr#s_%)dP0Zu%QPt0uO3{U0rfaMWbQ(j1FMb+t6
z7T6e{;KeeDbG4<FCY8V}`3nDD3^prvi<W>c_sEhAuq<zz`=)|7Xdfpmk<C}@48aVP
zp8mdY(=wrS#&|V)mq@l@bUWDaUaVvA?u;gZHQcO{H?>+qMcG@Qb~!ej&-#!R19-eS
zTu6%`AB*}%U#TzQROdMwf1vU%c$9DI{b((#&>f<7o0@W}EoG|Pg+*a4G82O4{OU|9
zBYgXG3OL><Kck}m*`z$C7<Tj&XueIR!R{iVG9FNpUsR>*Q%p!C!&-Lvw#e1&x9?W9
z|0^E!YYVN7>L!s?qSd)~Mkybw=}r!;^0#EVlT@YnkuloCNv3JTQ0urAWp<<NtIIQ^
zvxeQE_SMmZ0PdhF{GKyBB{!R%pL(r>nnxfHH(&lpUw9YY=;9T0r2l{jG8E`IDAgDc
zi+!eJKC&HM$9#O^791{?J&pTOW&}{E*e$&7embMv621?RGtZWMswaQe^f}1S?%?86
zm-2TUEB=uhy)@hga_-Fw3Yi+9$Fm>;a@NwSiDh1M76vwWvMnHef$=@3x7px&X_Wg2
zBE9MX%JNjB>Q8*F%BgUE!CADYnldal*NXA#JepdETRWV*YUnU(bpFPHi&qjkUZ}d4
z?nQtZelbHWS2_3n3kKI|bHv$W*Wn$Yuj!U#!A$$aNMq=#wZKspa!9bF*Im@~)@+4w
z?n1~Cd@!R1wS{e+t)?@jylJ$ym@HkSrslw*!RYqq-laRAjTv^sBI;3C|1@?-7j)`W
zSLK`V#=XocFqcQnFUyL8TKD+9){>qQZpuDH?0j$Zq1xa;#7Wj5Z(;z)<K*|O(H?`%
zN^GkF>4C!zrQtW{aiQm}dg?XcRQJi8Vbrogq2k@59p1cDt!0T$!hYQ>cjOH{ef{TB
zyRS>qz8Tp>T&b*fle<f7>%BMm&8(89r9p$Kki180M`*H0Qv|~%y9|U^XBLCGkJ7Of
z(^Y{U8Ch<$Z93<*r1V;kjw0x8Pgj9vX5V?)tBd<b(V$}>3K>N$!Htb{V#X%dOY>?<
z92ki;YGg_Z<zeWnCGk!$N&)HS#Px*y`=LC8;W2m7-flLFT~QrPLYLhR=jpBC$395u
z52T$ZKD>4RKVCk~v!IzsgifyMF%wh~fx^_+0~{I9Qtjj8Wa0dkU6%-a`Kb;MPb!`6
zZl$o-Z{_&ugFH;z{(+W#BIl_geh%i&8n}<J6p~0&PCeNiIYaZohq<rKG)n58c*%QT
z=`?s*J|tJw63FU|YgxaKESA+6;~_<({)!ZaEWIu%t-a+`fOn+Vq}v&%*cL4X%4oX-
z_TNrCd7fMa?o^E7Hca9YaFKl8HW?JTqmpvH>8(z{cEqxLDO*V&`mu=~7e;a|EdSsI
zMH-bK#E#Q)pvQnbIhZ(POv#xw0_Vq0C`8$K_@Y1~?cfp_-S?{}3H$sV-l~yoF4yNY
zH=+T0?x7o-+gD!s*|+&&%5FpYgmV;L;e#-t%=%f$dnvc;_;o#n4~!TSbTh*iPz{Ws
zmlOk1iH*WGGOR)(y4iv<l=gzdDjYz2cB_@pK*e7zD)gvMNdYFpZwkbH7Q?gQ?w!cW
zJ1+BwtfC5~urDBWeh%v;*3Q_y8k&m&6Vo3ouH{rId6%4i@4-l2)4fRbi%l$kK?VA7
z_<XNkk9AGQ2z@F75Pes_93hwp{Q)N^%U^?qFfFU=6i;`@K9ssgT<T5m<?h_oc^3!=
zmDSJvoK12{5n8D%@ELsAf3USvo?wKA?j+-^Q1-OCtK&|`Gq)%B(zFL};O@}95__I`
zlbk)5)WpLrSClTNHn>@y8%1-Ff(QU>TN+4p+=oRKLc{00cuPG&Xci(1or|AdOCrOO
zVo-vt;rW|1(zjdPwQ;wJ@!>4G;+lg@P+V?sROknen;^_!r;g!0EDbbW^S`CtO3M)V
zbsycRJ{FjEV)_=eC{JTbn+Q+)(gN2w;#mhTt@4iXql{8Zei7)4e{HE(e?m`YW-2`%
zwG3f}*_=Bu(w=U2UpSerpZFNLOCOqk+O|5KZI9`S!ln+pBFh{aN7I30QZ#v~g4IZ+
z(ra0!Sdf0`qF3FSX>RtIZ(~{>t6+zR@J-J+@XG?oy62p5hz>p&4h9C2JJPAmV(S;D
zt0zD&j90~SoNXKZlz}}wfA*tgTdes{hW%&God0u1bQVTdz<;%mx<Omoj<f}#^F{sk
zk5jeSuVVFitDH8o_Mt44K7j!1dt{TkZg6aac<JjKVP5IdN+s2%GvUvDx^eM*d?HGg
z9*={91isDO9jiXxYR=U?7S-<VPj_Ed|B4p+THJPnLg#N@Uf*5L<R_XxI&C9KmZ163
zeq213WD}rqo`bByUfvqO5GeC|?@z|P7#t3MWV7Hum|!Sj7v2qozW;vh4w;;pWZ+*P
z=26_FV4a*~aQo;?msho-64>6d)4vb`qe33p0-1tCI=;;ABJ}Ne@c}AoNN@G{y*`9S
z7oo;`;21wzfK6umCWq=bUx}ZWw{yR8xW3`EN^dprzn)w|?K5aw<1;o+V^wflcf(o5
z{`ATfQtn~j9JY$>d%P4zF&3YcM-qcHVI|48@x;wq4t9&o0D=GQ1B>75cbFiN7bX*c
z>w*E+MKPAwOA8CVEn&~0;f!px7fkvA%gJsqpC)>+4r6+jnovTW(Y*r20AR##|4Ki`
zm<CTGT&6RB>@@GZ2y-+pz2ToI@#V&kP+}NXsnOE*H;MT<#XGn;T&iG#kGs0zJ;Wm|
zxUHe63x3}p5XLcqr~E>B0|x$s3r~tZ4&fS*8YNJkcGSKR=-PA7Fkyx$%|L9?T504H
z_$$E6>Vl9r3mPsm39QPa9CAJej(4zznRzw^NA;HrrY34vMCf{Wt3~jbsQcEwwDf7W
zWtfxk(GCA3H<$Y3@yGe6PNoem{b3pZbu9$<S}i6gG7WE0gx%fan5FmtBkxLB1J9Rm
z%WWlwb*B5_yZyUJhiQ6Iv<oBp<$T6pr7b3Sc{CT9?KS~e>QjkD^`jUUWS<<h#(X=+
zfhWUJU{5m_88lJI)K_}(@5GNq$gv5hT8A<pWNOOCI0P+HTjGZZ=2lJOd9YmcU>LKs
zq#i(B?&f_uZ8Y!FZuPmtElxvIYey<ogN<NY9tcRt$~zC*C{s*QC-S%n=ZKx9@^mL8
zb#b5f=Mw4r^L5cSyWKuP9x3hjgTfixPI^imL;}h#JFzg@I(nnDa$9`={FoGT88~7~
z@~hg!;{@s|zGSBu%C!0<0m=<!tFf9dKlXQT#2=k>dhdxM1Rb5eJw!8?ymvWHwoG>j
zNz5QmwM~5e;51xK1)DTD%l)Z|Xw+RDi$6wEEBEkL4@>a*9ho}fithReSo<e`&h#Hv
zHp9%^|IskpAIaiysQ758bM?ql(1yEbbQEw=QJzon8pX3KB)#3KbA_V#%w<S7oZXHd
z?ybG6nSrg##&XQ)q)b~d>7iy^VG?K-n6(`xwlQ@)h?P+C=p0AURX~h1l$vqlEiY+g
zE?ealJTnsRbTWZ{u&zls-IRy)gu4RG!;XzdejewP^D<j$S84atZ~ii`!y&3Q2x14l
zfkm2AH*&AmdU4Wg3jxOxC~OXa;#}MEu@M@_DQV$KkGV~YUJ0#G;+c8u+%xO*wqMl2
zMW3=v<`zM1#JrY;w(pj!HvoNM6@X8rlNFOW8x}p6s>-r+%v}320-J(xZd9(2);HY!
z{Hs|Mrq0P(-XeqeTC<&_Wj*YB)qvkA-k8&V1eL|Gk7GNE_GD&6o3mY#y2FuZ+@VJ?
zp1FX8^P1(w$IqnRWgrU0(3nK$$M5O-2Sybd=kO(6fwp>n7vF}J?_#d39`h5?7;kp9
zPgqO0z>{*wVsLwNA*_+2f>hu5){_oM8cMWe%nc>=^qrirxkc&)aZBpomZ?EO6LDJo
z2@n2cC_Pv|TuC>g#J{1)mF_7f?;+p-ERR!arB{FdX;ZO;G?byv4X;mbpU`#p&ul2B
znlG}oN+t&iUh`lWRa;!xp&JQzsE5S|>X)#zTMCEgR-ITh{yKCw^<*ClLDPi(@@C{3
z9!DZuVn)PnR3BmF%n|l=tdKHkM?|azU2#Hm;UW5fW{HYDyb&2aMu!~y1FCgLcg{$(
z5B}Q_qsO4OLn+Ft-%@PNU52cJeBC!=vmmvpN!n2j-$m3es{G)Z!hY#qyn<Sg^P2|O
z@Tw3;GGxfvDDq{c3syzZc9=b>Yu*uE^5s9LEPtX+vVXWnP>~;VSSNz9nVC|W;v$Gc
z+8x@uP#@SVQSnRe=y9KUcGYCEw?8aXfev<@b8*{PxGr08*Oi^3bzqjtzw2Q`sVn<T
z^i=&=#xh|DhP0-r2l}l;Cx%e<xaK|d0cgEI;DO@I$*@vrX|vaqG!7LeRY$E!rxBKb
zu3EKmlj%g_kSh*{!v5hU-`Chcvr(2KwGiW1Rv7M3Z!mtF;Cgp--;f}qofZOtP?<34
zFO>^L2J-9%H%pdxCiN;CEi!qQ-&9PWZ&V~xEv4C#f_1)D>o$~@G4<a!#tUSAqn{fw
zDCUGG9<5hj;72r^KMYD{4D=v-L16xq!!W^;vxK6+s}}^}@x7n{+Tor34eB`P%vI%C
z;YzCVs~`w<loosWY*WSs#?Z`)#}R`f!HVK?xEq?;8OI!krG6>Tb$A+Hn?hz0*noyo
zhBCVwRbK0|Ex8pR&mY^z%TzIW_hwr%OsjHZdTJm^iRjtpnA`$oO$M@9@%EMhy8Nry
znQhjxYF4~v&8i4Tz^`k~fIhjE<;AB`6X6-h^R6hL3`oHVS2PyFCN(odiW9EgMKFGU
zriU&Vq;!J26}O|Swwj&YidiSi^$3#V@8zn6PCKb6<>N^uv>OeYp__A|x;a1Z^h33y
zwk5CucLF;*7P8Sfod^rLtm2k<NOvd2+Y@<Jh~A3m#RUE;d$BY0cqe8@6(_uMx0rBJ
zZOGtA5#@&2B6HvFfTz*D=#YqsW{u7|jsDO3I&L9w`6g&l<@8?GxH~y=4ia!v4()^y
zjI&9M<z3k#&8kv=5{fJ97YLTeEOQHO{on@WT+bGVU495<=Z$EUsMWf2rB8Tk`{8d8
z9>CaM&APl4MK`6X>cosI24+N0wJPllnyqVKf+<We3y;BCYZR`Ng%v}r2YZqY?VPOz
zHzV^F@g8OA-w_VbA%@{$LIOyTA^~BKGn|PgD42z%!AdMKNArby9$a+&54fH29dev?
zwhoIDtBj<F)}qi_v1|F&v~-4`I=qWa#(c6IJ&wuRgCN6Oj1=}f!d5E?$h8=AJ8d<I
zziI9h?4uHeN`r&cC14g|z;ww9qfLr3v{>02Ly0Jc(=ZZDQ$nAvZFnN(G#xp{LbwjA
zVP}w(#s660IyY-9MqOk6@Z{kbwTPeQonzQ(aWIME6gJ>w+oM%Z^+o2$NVy)$`<}Hs
z%2=<M#b&@hR5=ymufbPYwEA`B)iuRuLZ+f_(y3um^(hDCv=!+$wtdTMdjYpKOtLf2
z{bjwpTjE#b$RhlLKbc7y@iyJn(x=6{A>@~i_Uuz3<eU@F0db9_8|<vz(4s8ynWbM&
z*SzJQju}&K6d+3vxYQwz!{gJBe}f=;xFm5=(s%<nmP!!MC<}y)Eh^xD1^#Fv^;<83
zqZ^jfHJ5Zf7Ls<%q_0o3D}qB<Ak)o>sAvlsrp2C+DNIkZr6!*nrEeTYO%syk`Y~Bk
zhONfW%J!r#>$|g}GG4qPWFxSEi(0AIM0GaFwjcSEFUef<bH)~F;$XhEUT#UV&EmF|
z{KaO>drf6L8wf6I^M~e2c87wm5iq}i*$|J%qI{E(RY{by5&`g_<<S0kO_MLMf<=zN
z`Q}po+|(qMT(t$CK(2JP>^N%E+bj_Hg^gqhsAJ7;abZrSZ=;jTS|$8z4y;y*82%tW
zoSh*D(7X?t7N{06&CzPE-L{opRWI+Xczh-gI+1|i_d2$vR4}Fc+AhN}7ck3JRk6#&
zZ3&?9olDlDcbZy0s_$te_ssHrOn$j&P`BGpA;zcGt#jDBx1+P*=%n1>*HQwp0$-l6
zx5S1_4-P5Q+a<)kA{<TM%4`x*w%sXsPlOY4ZRx1kJKzw7-aFUv#x)9lq#3<r5m0?e
z6V=<J*Z1_wXy$;h=1v9K7y91unrqH|{!Hb`;#l^ZQ#F*C3)tAn5JyMca>kb;LV)OD
zUn<)i>|<rG+Q^mQna{<~-?jQihyt{>`A=>G;Q#72Ftf0*{+Dh;OEd`p<?e)LI)f=~
zd{Z_@(f=1U8{_&!;=<=4vPCt&>ub(VFYFjSNMqXjRB%d265#pxAW0;yejsi@ae4c+
z__lj@Ix(3Wk(#r6zp{Pd%Xx%Fv!e!>f63#7C*%*Y)29T&R?+AMC{0{VKQ3*ydB62_
z8^HW7`cjo$@6XVQ*S6!*%{-RWJK3awB%KE{f<E6c(t7QG50d@e<Lzdr2@}WPw^N!0
zi?~9~8?j(IAZHw}K;R@kc+_W5jcoG;LK_vaWVSPr@k)-`2~Pe>*8r)Hcg`U>(ktye
z$H#Dn<J_uwIUhn@&9h7?%4gX~M^Z8|6|1RAC&YDksJ61Em)zC7TKN0gN=E-X?6GuT
zOrVQ=0fXp?CGF@cR;%ZqXojs3xHx0Cj7W(JFTXd|d1-T(dUIyhr^oM0ecGGtkbGiY
zsWyTISz)iZJ-2Q;A9hETR4o%D6k@17+6E>j{O<lnfn;h-WR6J4YA9bQc{Wo%UR(EB
z<3YJPD>6-wMe=@kH_M-Yc^%<#&nVwf5$3UNY-y^RYkG)D`RJVWrLG|AN!!Zi*INba
z{%+q}joSBx(D&Wz@nny%VUFh<>o=le@KA*27c?&_$bNt#vy1s)d!9(nkK5~;7lpN_
zx06{Y5tuxS;6KEl2RkYmQ>`0>cWwnN-Z{Wy)g^{3Q-^`}8e<^T$gpE<P)E$FCY=XE
zM)YT{=52x3+qk!m-<Ys|Z8uB`ndV-8^KNpUr?EHH_!H@Gzy7;ahUr2*5+|?3lGfe2
z1V5xRibaM=ED>n&YfZ<@>854!C6PEMO)a^d;aO5s>+{eP<>-W;nAEt#qHb6<Xl`n$
zV$$H{w$;Ih5iQ1;E5BDGt0J3sb1C7#Qj~(+Qw0VPz7~RDlW2+(p+;_6{-Pgeh-sqf
z{-e$@x}TnuY;KY6z`f{F5jNk}W~<|A-pY3|C^78SEKfMl-;FY1KtVsj8P{q!O;&0q
z62Teaa8?SCp9kuE1i|y|$LPLiFbi13mC>>cTLkxSyJ2UowQh@qYXZxpjq&$&=8ho3
zc6w;w`8)Q72@guN<6GKJx*6scxy=*cHi_6G=}j;uuye&`3$nLp96G)`$XmskTIg&Y
zQM*mhnSI&#=9R5W@32n}U6mqD&xvik<4Kj4io<H`1<5~2+nq>uXg#st#D8@8d58io
z#(p#P%ZA3SxMe7!ACg})#8QZ1AR3?-1xmaEyAZ4u7xclD0yUn5+>nr?2K6TzXVYD)
zCY~~lRa;qv9OeQ5ha)u=m}btY|D}V#{PNk$9-a#jwjvYU5<gzd)ED%Nc7qTB%{H82
zfojDgm;tD)Tqwpee^FQj!}~BR`o*o^#c059YuLi1E!4o+9D)L4EFm-*Mo2K%K^ALZ
z8mnbvP0DM`Ion@xD@FgxSSQeq`A?L4APl@)<pe;+UNm(I0`-~^buEk|)!>vABb9V4
zH2}gTHJH907v8|S=hX`Zj);+JDFDJJHQ2TfSH>W`FeJMVd@m>^p$CsV5BHMmYmWx#
ziA7g7v?2D_FA%8rl&Ei?z^3>NvwOj9vx0UZP*ZwkoJrOVgq~hq5m9lP$kotfbdXs7
zM&sya1t7CosOPB3=)gTD!m*7&9yIB2!1xW`b}JnsNt*$R{jo0uFUHli_D_^&Ai}xR
z8o=6K7VR(u!67xub{5yVnrS0b7*bh8awhg0Qo<iN&p^0KstY`30TiH(aEKPE@C9i;
z?j0oVY(Yl?Mn(W^o~Uq$9%=9S=hM$LuFXpXEE`lN_M<=8r>YI3YyZT2HgZ<C0kFmr
zgEvONb&ZQ!kz;RFw5!hbhU5uFK<J}fuXfNh5kA$oza`l)5P7C`M<juzQCebj)CG#1
zoBYCkf+XoF>T$yQnym4MzT37eGBq>O`Lk-!i}z>MRb;EQjYc($g6;2|(^1|LU&UDQ
zo+{oH;a=o0GgMnTe`|2*7RNBTLXX>rNhHHDR5_~br0P`cBymdW8V7<>9J;srQm%T9
z&Pi5iSxR#MN|?8prb9m_!e!-qIK6#C@vZkeL7k>->cJfEMos|#rG+`IkOx=~ykySu
zMtAewAZs*SCVH@l;v%q*RCjr7Skv$UA5A>jA#Kt?b9rJJkg&ix)d&Hi)StN1Vxk)}
zXB73Occs!_$eR&b;pAz&WCY^)NhD;#{L$vGa~KOs>FWzrj;lv!M=dT%>lGP@q@7sE
z<|lthAgYgDj;}M*D~Tcsfdke_Vx3w?-_IJcuLVHc4<Co6$Xas5^smSRlf5>i!du0l
zn$G6K<}XmQwDAy<@i%`^p(VL*b7b8yM9L#$n3C;h@;D4H6a{$cNy4{c*;D--9#3_A
zJ<I}eLvk{l4ga`^_X77Yuz9lhW0fJyM0s26AMmRL^e}%h0LLPcUKQOlU~gxds5C6`
z8KfF-SEIJnp>{QdTa&=F8|NS(NODI>BBBE;WfsI)&wq%Lq?u15s|lZqFeQ5%Ux!(p
zPL;TiM5%{Tyh3{@D58czFZ?n#wyu`zIOUKw#Q?<2=@~F>R)9=FF8SK~3L&Wh?AW?+
z&@=}+sN^)0!TRa#Mvh@ct`A_u&6r*KC@LYB>tN@8tB}hzNlfJxEJ1g!W;<MYuO_%1
zP6~FYvX9_8zftv3GIGO3lP>MdA<o1uP4sn!2DKyj&~lXDxfIZV*BXbkYsWzI3g5JC
zzlx@K6tw3C!t8WZmicSI6C{1T2dB7GY+zn}KFs>6ac^8T!hj379{kS9Bo%lwycA3U
zlvxv19axvBW(~5qKV8KAmN<X(k{6bxC>RvfI-tnJ4o1v~ucnq#$b5dtW_xQTm&Gt(
zI`W1$@UX3<A;o4rGWr?@E6aI_*AXeTNpt78c{0&(*|ADB?|U_CzqK8=I{(0TxsjSG
z;-JxzWTpY>tW}v+tY1ZYuBS$@@e5>6#QaSrIRFUO+*VC-`YzSG!y3gecqoI?^;lK!
zUFSyB_n4y?pt*AB@ZIpID*Cf8dV-}xV1S3*xbn=KS^Mhl0RPo>UOP@|Wk>8r$*w78
z0C23sL-AWte;%%*d-lsW&Q|xk_xG!7zr0Z``2S#QF>(O@cZgQS!yd>WZ)B<LY{MW&
z#Kg$RAZFp{<V*zMWMPm2+L}3=6EOk)%kb@-)`~420MX|`{m{XcP$iG_m>COe2*}I*
zn;y9j+rSi_#Gb1`t}4?uzb$h7^$oAIbVik`>BWdUo#&UVsv@x{fzmJPbrCf&57fYG
zwRfB(o_BJk35*^p`QYC%F+%?Qz;6NJKqV>a7NKgXKX@Jql5tm&;y=;9CbW;`1XKfy
zhK$%LtA^relM#xEyaCj(q9y~8=CJ4jKWNKKDQkE^lhA;2*YHkG;X-k>eR`B-<#n_w
z3dvVJU-fElh_FOx%S3?@w9JTT!W6{$)>4Ql&{n&YkVHip-QJf#1)VI2)Z(F9H2`KL
zPwz23EFsyRH%4YP@1Wee7Fg*}Fyvx+D17lDaY~b7%Pw57s#0PmfGI5Q4zKBW+s5+e
z6<F_5UPNyUDq%b-hoGc_#t4gG2p9|I3uilYje<cmX@n<jb_VTnjg)N)rc}G2LRb_P
zQCRpWD7eegGxI_qEIc+=?P%Wj$#6{-5<&rpILKbdp<0fC6^+M5YPo*lMbHCUW`hZV
z8Xjrt-_}x7Uv*%SyUPO@a6?TmlVpC&L-&J&C<|jGZT&#j^}tXWYW%F%hzO6Jh}i@I
zNybR(P74YZ3t}fuVGpq?05c()eW>d}9>(DBk3fQX(=8vTSs$)}{qiLk`A2sOwonq5
zlv=+GG@=kGnQ<Q$3PYheEL%JkZfG`2Tsegy**$2IYApn0*r-Xrd?Fn&+XyvuULS-n
zbYmX?|Ls=!?i#*ZChTls&aUg@yLlv+Vee@I2FJK<Rn=rn%*B@TfaS?cMyV%LR~Cbu
z9X)x2BeS4t6@6Nv-dLhd=*Q#itM{@FYeU|B_ptZ8zyOSaWIRA*1cn_XXwUVFc!G!^
zwgkqP)nyPz2v^%SkkQ{RV5G5lZ5;OgY=zo%iGs)LB0pzr&-x%?S{4zD%hh)A2-T}F
zZiX$^GE}SdvQ+Gj<z>x&a`>&!E^Yy%q9)Tk=V{5J8jSQ8MxV}%lo(bbDP!JsyR!9o
zvi-Oc42iCqcn4o0Oan&Uw$<3>t<o`H|D?!Cx-~Osk4r$vpP?C~w4c$28g&1*by44-
z66S~JnSD4mMY?<KXbRn46!wk{6`BMAL4?+CKKc{MU)?(;1h#S)Xs}tGMRCujP5{Pr
z0djb}WX*1exjo@+wpR`1^R8)bSSn|?!N&Z&;OtsZdT`VK;;D$iaWKbUreBI_HjAjY
zkgyMbP?}8A7l)zqE%f9IRxQV}J6BZ}Rk^R$8;>7Lm0l@faw<#ar1R`+)6Px<=1n8#
zs9WF9VJ-NZsJ<meRK)m=FPwpRKWancSCUBxK{Hq(<v2+v9AC8@=2_me+Q@qK&($mk
z-g<+yMpGDb=PzxxmcQOd|8C({+XRT}mTQ1R$dMCPh;1363V{PADav%655xn9%Cda)
z+C|`v4i^@>hPnv(SjWe&2TNNs*a>Ynj)q^2@3KYTe3?2*t1y^x$wQ;={HnehiSt^W
zm6|$oDQg`4@DvmG^3NK#VbPais7#sbzauN&?;Ql*Rb~l3UZ1_-cy5fP%)}!is<GkF
z736~v<(9(fNgG~(IVvq~<lV`9*m<&bd@VnG+OCaPynCfjT6pf{p+T7JI4Ya0meL0y
z+jpfWuPVr1nmCH}G<Gj{=9I8WciQ1?*c_N`1mO1sAAaqn3_eeX3>&Z@B8<KLHQvsD
z6xE_-mO~~0-(MsVlTH(pYxr`%F+2iRbYXL?$V>5d<C>uCR_!~{K3~bg@Dz$kIS)`F
zE!hpDrRxwr`#zu_51!s!BJ|Yex_&Q>)*{u+KR~tmbq6_l8!^LLQkwe(+#fsWngsoq
zN$=3c{Kff2107d?r>*W)PDPs{9sL3)l)zerl|u`or4`%@rD)3xy%`X&IV7ewBnD--
zcP(=ed*0jr6!{cbz%8Up@8W2G80hqa=TutJIc-(qaksZ0Ap_<Xsb}2g2>KbmjHq_e
z%=E>t4RHxq(jxMgy##X0V+TYWAygbksP&p%?+?Nq5#%GCM_c>tLC&kK*vMYfOo#Rh
z=IDb^GrVjGmI?6JnPQe-Z%vLD@WSH!uutYgfNMnjq?m6S!dqJwgT!Z~_t$I!2SK}-
z3CR)OP(_0o4@!BP`)v_a1Pcb<3J3!p2&j$tGK&?^AF<~t3Mi5?mCO1g{t~-(Rd2df
zt;c>3YPS>bZjLLgGfR6bK$dKsj?A$tJN}7vy?K^DPv0`_;r_s%ikxZI_bbMriGSN>
zjE<Ru>%uDbt4r-_tK>)YGF(@qxM_ZFm@psNkmW<%>7~JQc86Fl=z^~oHojs#>)D^=
z?lrC35am<jg4=P%cqF2+6E?vjna(BLsnV6QeLrP>wRO5Q!wfgx9SDrYbrZ6UiJ<dD
zrr7X{T92FYr2D8a<fbSp=as1kqwqvAKBdp#vjDp|{e3jGfPB&OBVQzoAEelmP#WoF
zqzTXr^&lGanTvZ|g)nTbXqHZ3Okd}ZOSUHmlv|9UPf55EX^|JFIzipF?&xU`71a6L
z^iQk|&OG6H`m?^_urHWhNu;dc@`T31w|5{-NBa)??Ym5R*MjcDt|@&<XN@Ee3}!O6
z+yB7~U}j|e-)6wSrg|op|8DndgT}b+4kvu)3)L;CqUeuVBn5EuRnKa#bV1udUIf#9
z6<V$fKaG)OYw6nbUq=LZ+{~g%g#2;!iH~%NkJFhGYrO^voJTMj7zg^GkHOpe3OweL
zoL?*u!fwhQ--{C)y)z$$3L>{guj&rC@sS{Tq`!1Qa1!A4?2r~gb!KUZi35DkXLyoL
z&_^_~aIM3xe?NcU5afJ!dT&!6RwN6Iin&tz@B+Bm>7KEpT&D#*UM=@ynhW4m1<#Xy
zFMqsIem(A-w_GT8H&97z5P_f`vJ9+dmq6W+$w6iS_lO&U+qt8vO}Ihvn18)FXet)#
z^wE5E<pFJ+FINTWB|ZaAsYsX2oVj+yZKk1rUurzb6xo-yapNQ?pN@fnC<Fu90PObV
z_9@>+W&`HKW)?XK_RQnS5ZjLJ?)9`8o-AKIN`{;Q#mowvD!airuBHfnZI3IiT!_nT
zN2YP^k}3O;%YAtCQL*V0%Tf6tIY`N-gefoShV5WD60OUDPdOkhO#F78{Vj=r0KI>-
zHe{&t-Kvr2C6YA`I}}L-slY|Ssw%JO7Yh&i^ki`j)>6213DfjwdmRY6jc~!|f!Yl=
z!Y1V+2e^)YC?W;-lm5lCoI+?@Q^OZ=8JBu$D<I?UKi1ivX{K1!sViHOrBMVm9L0qs
zL@6oXi-Q!HdQLpB0+wl&G+A<f{anW?EKoeUEs#<_470#{X}2R`Wh(It83ha#j55*n
zxSa2UD^p7dlJM^>j6DKo9@C|iqoSDO>lUo7C01)_z<faQ`I1*}N~NT|hIwP*&2hS{
zL4Q|xOUa1&L~R&<_a~RWxXcmy5s;kRF`_w7iy<Dv=O5mHQeMe+Ve+CwL}M|=isiwS
z^1B~$C?|uc@^XGBV}|<Tgg4WyMz+YJ<Brj+*U$>LV}U~#J=?d<73fDPVV2TJ0yKuD
zDO~|_>sWo&Yplpk(QBdmMY3^7l^gfu?sb!kE=5{Q$Hum`51FR~O+H=vgriwX;*eUC
z)qn}!Yk~d7rMs$)qwHDC%E7oAm!d3H;)68&@D25#lr$5!N@|5bZ&~U>IBr1gh<c$W
zW>eL&vnXlkF%f6b2(`=P<%h5biRkL~qV0~q67HJ$%4s9?V*L~0GUWC+I^HZB+S+(|
zDVwoCi2#SiPywjuLW-B;bi4&C#?h0nh=8O#Wh}0ks5?lUu6`ZHJ{wk>y41j|dkGRi
zx3pW`$G{HXbN6Mm1t1MF`|4t~qdB5Yf&Wl?_4dk0P$QGD!G&8B1x0@}!9e)#IVZ^T
zfjg-kzq%0e9BO@9UCU0;7mKN<WIu0>FAvh^l=ZWmd`?zQ0xHl~*XPueRY<gDp-A;D
zQ>nW^xaN1%+jWbbw-P^x+7;~>FFnwkl_x8DocT>jtw<(~%<-G)`+nS^N$Q7<*c0c{
zXdHnlQWoq<OHQw6keQW4yB^n0xnVrvQs=c9_!J!l=FXdkja6cr_h{p>f{Y_|#nc|x
z=d{r>#s<gn0jAVJ+x8#EFDAhM*pv3R8>8{*#{i9;8I)a&od0rbBy9}MfDGyuCZA3W
z`=@6k0kkkPcP3)}ycM!=c2WR33ftM(+t~tboryRZgzc>D9F^@2je!gzKvxT6ppv-I
zzg-|!0P8<CyM4Mx*3Ljj1~F?xXP^ks*v<sV#|P`=><BcpfpuR#TZ3_zS=jlwGO3j#
zXG!ph35%6QiGxnd`8FS~hOEMNKb^PCU}=|>Tsl2HeQab`Tw0(yx*(4}x7=iTl$VS$
z+o@dtal;{hFwC)fyGmqgyPkc(la+4l$b7K=^Cr#7bbr0;gFFaX2>RRq?WGprb$?n5
z92JYgVRP8->$zK~)b+X<=U8d7<2ULHyE&NDKI!^+f63zYay_24c)H#@<^6<rpGBoy
zI9sfkFPF(xQc@Cp(&loaR;E_rakIa-yE|K^KB?q!xhcrNP^GT#<FZ_D(b?H4ok9<U
zLJ@-gBVan7rn=a@GZ@p>(t^+BEMcwR;W_tdvo6(&S|j1G9$pO+Hm!<{5mxYhR{y5u
zs33DotyXUSR->3NX4m!BxLB;+RGG@JMeTOBT(EGm<nl>g{lVsJiL%vfvEf^d&&z!v
z6cPrmgM-5-0Zl3xXlR_AoEZS8{b&TaboL6*yJI>y;KTWvpCye^;cL;n)7}XBv-w4f
zD-E8=v-wMl$4%iGXKHqKc6_|i8xseI+wM@D5eWdG`I=JT`0@T09UZ+rwOZ4C^|`zU
z1o~Xd^SzwbeKxf*t0odtso&+RJo6V{Wh=yQeA*rV2D<aLw<KrE6;G$lwW6!*L*MTW
z_@p?7(EloD&1yA&+OVKkr&6g|C$=h9(*3aRRi)izyPPGW-|l|Z-RVgcjP_p4=u<zV
zu^;#shJOK*1yG*LNJxb3?c{tmb#;09xD&%qTb5~*W!_R2&_clLxk)6?`y3YN>WSR-
z^ch2*UTXKKTf4iOE|8{~Gui0$0nP|sNfayO3JD1Kr>Cbc9<H|375JX7wtNc92$b^O
z>U6Otd8ACeYQ937gNccWg@uKg`FgWEkaOr#h|pn*0BQS^0lK<t@zZsr5eap5?jtMe
zgl3ajXQ;@uOWol==1LdFDAN)3I=wygbMbgRXP@<rwA+Oq&zd&ESLb9uCGkn{QKleX
zAM_`ljef{*u&~}>0UOKCvbFB+IMN5Xj?cG8bZ#Td4OVtltF10pOk9(xOe7?29VSrO
zFXpmRccCMNfe%!iVP9{*LOE1}AKk0O_)mp}!%&GzsqB?DsXs%8AQCWxLpdNrF$M5#
z=a}N1PPJIAe7lfZFFVz%y7x2ug1AkDXaw%l_ld?|B{C<b@o4kzWI?{{M>FqVFSm)r
zro_iN7n}6`+&V)4oYE=x(LurE1qE`3eU`7ciY$n(eywI@@2I5WI~gC5Cv$B`h|6h#
zgVFd-j;^k;(V?MR15vo2wXt&N_&Rxveg&qrGa7sX&e7#=_WOAMT@C%Tws(4(hr#w2
z4)ytbd=J^3F?ZM`0HR$)@3s?VUVG%fNcjI+&epn>i{#D{CUS*C{|%vG*TC94z|Y>@
zzD45acIAUcizC@T$oxde6fd$^!D7gaUxHorZc99wRx{uqWd8rT0{gmqLy;c53<bL0
zYxmxe_&$kCbhiIgt39Lmy6flR#ADeNakcR7zgbTUgu<lLLc>9)5M%inzf>vTpX&a}
z4oRqN&*PHV@E`Fpm0A^^?k`X~e~AP!xihfju9hqv=W!VEvzjWRTE&KxuWmgSl;6`H
zX8w|Pyo#bT7sX!;{kx2FHruXosk<Q!lc_1<;o~#tw*FIsft<zIP`n*)du}PINPo-w
zU+nn1q5u43vx_S=HDjBIP<(K#_3q!yW!o{Z-V;&1>BeE3`KPp|;xD>Bd`E4b&ed$I
zHX5uj{{{V%**tFN83}bxDtn2<1w9pxULDO>M>9qL7IVW^j=;G2eKKLm7@seNT*mKz
z((-Sj{*O1^%-fq=y`j9adKFNz@wAoa=YMNx;Ah`p+u6K7TglYXnQ>~B!eFY7pd$ST
z8lQSU&RNXy`OkvF<Q&4^8Wzj<Z<43Q@w^of-yga@=8l}N&G5_AyT1Jo^fS2KY@-HR
z;$`pB8<_Y?s1esuq5AP%9b6t5J*6>f+URwf{r{yn#RyRRpW&Rql#623LDGo)iKEk{
zx_{uusUe1UdxzvsStPBd$Nv*WgWht!T+{U+r>H6W^}OdFM0Ydq<%ABG>Af5TOELFz
zuwECW=utyJ)z#G@m=azcn1rhvq!y(x{0lpSVZ20r%!*X4B4@m@v9U;bJ09yTy`R%Z
z4K4p{G%OIuNnx&)@?T&4KM4DOy!zzGzclJ|%=#xN3BrGi>_6R*{98SkK>y<K#gD*l
z!@bwu{2pUvlW;q8lY&F6e?uMTJj6tsuD;3XWa&zva?#>P#Zjw7!+YILkY9z*`dhHk
z(Omzk-W?Apw$!b6#>?RGs9gj*0_c)O<mcxvi>DxPH3LVJPft%DA0MCZP8jq%SGD-<
z*4xwAtjNj8I-ZZq{NC;`KSw^qZG$V1*B|$6glDr3&YdzP>$D72!W4M*f>3jn5zB3X
zDGd7eS6A5lzOR4s#rvH#sunIby7Xu5RvO2x*{d{awqGu~VxpolnSrqX1SeS@kD|l$
z4Nz7f0~1}Ek)iadSeM@qin4uC>a<RDsr5TwKWCwjk7u!mb6&6edX<SRUTy#Y8|srk
zY)6~tYi;VK!$~yi;SQeyGIx~1LEPR1n~cKcC{?d|AEBu_{WMQLjf%LqxMVk;&q=9J
zF4MAQ{L@C!(a}jH`n#93xia$QO7&B-p%9G3Q|Wbrz+uvV3T`$&ka#qO;o;^cB0QY=
zr;OT5UVc6%t%ku8pX+IAU|?^Fa?$zYiO=&b6He3~9B{4G<)8t*2C8o~@Fp$VeLH)>
zQ)+53t0PNw&mdGzH!&@Z#Y|307adV8ELzp_cq&&o%aj2f(evf}lSL&BQZ-3XyLL&|
zrCM+Mb!L-lmZgqLwbW1v@ydy5C=8=ktV&x3Izs3Lh4WSVRrB51e(#zkY)JWBP*4as
zzh)ElbGV>{dwP=6N_E>BwV;yHaQJ=OwV)zAoY$C)IO$APb8kryn-r`MCNhm)hT=#B
z1qEr;E74Yr;yIo6Wkzj$KS#n$Ik&gxJMPBI^R|mdn-$j$et`R>0TW@VA>WX9>{Cv~
zIF#`=>K?vK*~V+#B%kr;Qw8ap?$_7XzCMwY7ytmk(DhooXg*Uo{z)BOlD_X#LF(Ss
zJCWGtXSc|);d;F6MH)0?#cs5(k~z9v#tU46f~r`mF%H$M!u$w7+wl9~Wn`SH#cMSg
zjt>ii{=?0LU_D#>r`oXB9{KN{F~yD_8585?;PikHl}JUtA<$8T4Fz{l(X@X3@PNs}
zXJ>cu{`R7+#@65UY1y%*kLNLSbHDTi!+lP^e|13fbi{#yfl@6@wp2~b4&ifmql{0Z
z1r{{?L-b$WL=b^fWBV_B|6fav8ou)b>k;obftcOT?d`&g=>oX_#oAkdWwotq!-B%A
zh?I0H-Q6iD-Q6K2-Q9?Sw9*aI-JR0X(%s$N;XhbQ-TQod@Bf_d{QtVVuJtOsbIviw
z6ZdmJ_Zai{*A^bySq*1OTDbV!eEjgi#}AQ!wxb!iDNe;!jD&xF*j%9|*Y7(vLpBz=
zBdOZ!CGJ{ErTY8%e%{zJAmplGNapzb4jUKW@J3v7d7F%4?GJ+4=B!3>h?~Ob0`lrQ
z0j~NJNGcW=z|!(a@P=g>J#UxIqFUQk_<a}tI$N3D{}jTy-_68Z^2CKxsjIDcT<=zi
zNfxM-2nfn^)DXB`e36s)L%=hcD?Qz98af$7pX4a8DDvFnyA9aX)9UO$I?q*%?;J?v
z${HCl)$7Qe(u`-b?D&;_)E$Z)X0L*4;rG}%wKLoJMs$YKMa3i=ZjO4oRB4FgT`msi
zXl$k_Sx!&Kwfz)>!p(9|Q$(ToUy4cIf`0+%IZV=jX_w%nhpvgdLW!rrAv{Yf!i<uY
z<~S{s_gB2<$8kJ}b&L|-Tkjp?32-d79)B4ox!S;(=LA_E-V$LwJ&K0hjq+c1%89&D
zI7HyRzN75QG-28Q6O{kuL*-`5<am)yfpF<P^w=nOxr5^WlMCXke{Ae3XK01QY;;9b
zp9qoC5&W5Cn)?`w`d4oJXAD+M&CdR=KVc-llbFtOWV_*Kpm?@i6kfhciLOz`n=X<u
zwSw9@9Eo4IFzP`#+n?87U!iucQD;L;>zb~j=EB6q#g!=xTM@R9i|5{Zd-P8;ES=2z
z#6GU8C@{|Lrl;B`Kk(n!?!qiPyKdX#LE%eV0ZvYv+u!hDB*X{!yseG3WWNc8DlRS_
zA|B^Iz|J$s_1SKogXYOSl>e7m`rrIZHWWnj6hHI?bI#=laV!cFa$<aZ;O}c^CtjvB
z$7nPexJW6->YhCL-rM^-gW^(`ZKwWkZuoDX+K33@Y;ShDbp=neAxYnQ7zY8D{Wttz
z-ChVI5kE=NqJ(uoTn3q9r-z@9k8l37NEbqsbpGO7IWw`K^(6xz%w&*2rK&9Fd4DZ`
zxhjP3a^<;l1HXrkqm9F4qq{2wyAi$fyYh{(`*bO4ep8**C7we8iroN;3KFh~)cp(H
zH6LAW<vQAcUz5PPmojBPIAr~!bps7OJRxCJF7uIK$0iSVp!I=P!9SKZYRK)!PTD_=
z&3vZ$J6t4XhM`!Jj~XU28s6_&t=!QGedoe^b5Z90&Vx*T%hBztyH;oEj?OD96|q*<
zuyNwtsTXYI#JIn#)_k(rWYjd~=KWugW0Ei3M$QJy@t@$&RJBizF^fV$ppw*|)%m~u
zP|t>-Rc*C+@)P#AugDxDfZ}^)bv3rn&~N%rDUt0A0a+U!_VxZ;!@To)BE?Uf``IqH
zhJb8w>vcfhWgFtp(5HPl_|b=5g%z*fZZk|OE1uW=${m}~Vy61#-J37n+v7#rH8wE!
zx7wd;=J$bxi3*M6URPPRXyxwX1y+UF#*Y{ty;Ac?kU|n7oyfVmuuyEjJ?6^G{&=(2
zVK;Ytn}>&oh={02tHH#5Do%p5Y;q=5FmT%cuW?r(bc}ot)5d7$$8~#wP-!|Ym1P^o
z9i2k1n{s`*o-A&zSq^v%%ZAf29s*oW`>l~`>*a7V>44n^*Yg1XuM;Yu>Qzyh0FbFj
zvo0z!lEv-9>X%84)ccrES1Q)gUJXn0%_>7+UXLMu1j8!PZ4Y=aa45V#x-J+>z_+)w
zWOK04On4f~cXt7Nm1bJo?NIO|9NImclnF;r3i98-8(VKh-;1OW#d@|jkQ@qTD-Hp0
zGNb&2B=}fLMOrj#?zh+H7(d~5NTrj^a-_w@<5^6j46etuJprNzGuWi}bieW1LQ5-2
zE#f}q`~Xm#*){&w$Ne>tK>&Tr6<kd(C#D<&zfIqdjpyVYa=7*45I~)93EtnKjpz;i
zRm)SlnQCi2076K^MO9vG0sMK7sA~F3^g7$j!P8`%SlQd!9$%4pczA4&6>$Bu6fKfh
zj`UgX2$nYy<EB#NxG<8U%E*kO)x=I#QlI1h`AsJyvmYd%?&bv>SuTei8IuCFU^_Gf
z-ZB1qGxj1z7#NuKiX@bi5)+d+O2{%e_cm?|U_uFddjv`r`;8wjrOTq>APw&4`}0L3
zaf3a|9-7R1UwxM-ibf(lt+FX04}DrA;#{savqm|yM@E9Mm_|4+`q<{lD@bv77|-Op
z20njiR|*adRW|9?k=sp?lP9r~$;m;dQOmZMsW+23E$0NJOOpq58FQJ*Xl_df&tcp1
zCB^Oes>ZaJnW<6T0M%AkS3ho6#V`PX>IeFoPiH8>I8M;u=(Z?&!=T&BcjSz*pgE6R
z+@1L*zU|bb?Non%|6a)c;<~<#t*z(9q90*oYNp5S)%`4&wRUxh_Fho)Ra(qmwPR@C
zT#pIAb3NZP&*>c)kQNhL`YZC|SRYi)dpy^xaDo*@$6#}3AVNbh`Nr?lypTe=mt@Fg
zPft&81@eL*rC>0o<-Va6QoY08%;m;l+C3sZIVtCUz-l>1t)-K%)!+g=698~p-<S6`
zOYIKPoaI>|(6qo$V$U{*@Njc?e|I|wK?vvqd-Z?L!Xso!C+`7BpCJ|rl(^BfnkEH5
z0UsqVFTcFJ3~*9mIr;BU$nMmHm%!)gajR}^Z!ak>o~KP-VKO?jeM+NV>E@;P53uE9
zL(j$K`u+3Eb~i%ozXHk+e7v2}JY@wj^R7?>UJhN~`+6_xqx~T4Gwhw2n&Ghy^uIi;
z*>z6`;I#JHsGB&gS8vL-E`12tvq>Q4On>K^^O<H+$vEN>3(6*16ME<AL?3+8eM25$
zv7|0Csv<gsfPB?5!_~x6z0ME-+JEKU6g)t<Mf!CKA4%{ob%66!aN*(Msn@PwsH&<$
z5O9B&=B}}b0Eyb$o_}j@55%~>zINQ3aUM(+oOe6!ySY48FE>I)MlR85e$O`VCQnY`
z;KZP9_4~WH<)#qR(Oedjk!-?9%{mA7qb{=c4&P@uZTSahbeMktfXTP!;VkK=YiB!?
z^64Ivr3PABT2Z?xDJii`@BaK!RBy9Vfto*86i7Cu)mzVixzON2q!ZX*;-YS$<s^f8
zanPvT3>0@-xYoY{QyVBJV-gd0E$AI^6X{fO@$9RTFRt!DQUKhpz(=@w_jAp~u9VPF
zX$y<O30~);^s=(!Ov(5^^{x2MUwO{%pp(E89uX0vti{XAtGb%q($cbo#=(Z4-}_?z
zp1~Ck#TQfV0ipHp6jSa95d>VpK{JHlWNTEs7YY<Ym-kT*5PygcW`FZT<pz*@E6ZgQ
z@sN~8EwKz#RcUR8?4>;WJKOw-!b8=IsB0#7o<f&R(9i&!=U++XuR8#W!vP>T^RatA
zj0cdltKE^d4i2<7j99;i&fmvnRC6Myr8SpV)Oa;@pTlH^ngO9rLYg(FMM<XEpm#gH
zFYz}zEr)>=Y1AyOtPmrR0|ev5h5%T`bfrA}uxNYJ8S{<j(hk2y1_Pm)Pn)ka#{ng*
zCg;S-W;98>$@voGZ%eE!qm4^QxB;aZShuN9KJy;e1siETb|_(C_rdtUgFSnEo9MZz
zSQ3;_(1Ui3%83rVnF>>7H@BN|<KYkjz7QXFP(^5fTobAG43J@!R+dh0VD86MPk>Nc
zTN`)to_OFfcBGuJ<w>bLB(Uc0cLfw>0O|Q&yD<%xBTnGY0$$?(qC64ilp(Y5Nr~FY
zWS9Qi2NMw=K0La=CO)s56NvP`K$}W8LhaTw{x8<h?4VdG1SA7D!rW^}`wu`b6sgfN
zA!B2k<N%)MT`2xNg$ejbBH@so@P5FEfhRY@*iJp$NNL_7^5h(obC7|*prxf9;5nc9
zgiMr=K&w{%-H4g@Tb>js+~yj5mY{ZamQYafDO3{@62%X4jPFNzqYS}eHd@FY?$S<b
z9h9vy9WNZRJ^*}&Vc8`pHLBppl^_885@-4?p)rDT^_J7=^2mVAvEgR#2oa8|IpQxs
z|8ulCKcBi{JJn=9i>ky&I0P4YPBk<6BVN7L;`?m6+nyhQbDYZe0cB;9O_oFq{o+|S
zAPSPWol))slleUVp`21sp0@j`A~J-{Y~lqjx6{F26RjpkLPAq9rx#mLi%x-9wPWLI
zAXAflbK`dZ@a66hK}JT#t<gNk&7lky(=ioX)bQq7`)%~&8V%gGX2uO2Ik73v%%W(s
zt*^pU9yf=b_w#OqfnYIN+O9+Y=FJ<>JLuJCwBMi8G$OBUayd2mCvzr00v&<1wKeQu
zAqWJ*##RL)iN!?m;ywiX*KbjK`|QDn*EMgvJ&B@NdQn?lyR5UQ0eTLW(~o}@y;1tS
zMkgkNjyHg{(-&K|c&#~m0xIrbF4?xSYkKrF>SivsI%Bf4D2VHxjOyjA1~BU-%HO}^
zwtx_j8OME^y5cN?ZI5MkbOhY~X1q6!rAMCdmlrIg<l4)EXBrc{xBCC>9|K%FYwITg
zLE;0Ol=5L)tmCdOi@&Yiq=Wz6yLDj3V?Lx^5&88V?Y$5HMYav;#lk29!pnL>@VNds
z0YSdcxrrt-5z%XI9b!?mc3@XvK-XIfOYMQ-B!5t>lQgbE9j!|f9x9P7j{hHa!DQ&)
zj-w~@o-oiNMgBGS3U)tq!s>gI$=Quee?hiiPUX}Ygu~W9C`;x%=_re9Ukkz!9K^iy
zKIZR#{(D_gqHxWAr0_dwh;<HLGLqDtlQ%R2923K!H&PgQPuI5d&k_6`f!w?hNXf`>
z+#iOQsde91y)J23ke(19GU+7L)>Gh{$^~kQGBSS`O4*_i#ezT=@}j^wJ;<}n2qAOg
zfr=#@IbLdNc*y=&A2y4bjp3}&z`(#$bN)oB=;&x*xg@`>De-cFh{snv$+W#mBjek*
zSGe_l2%rX&fiHk=BoUAL)I=rp6_j3Z=v;Oie+si|P$Zp$qd&%r;w3U)zV=1^D>a3O
z4+97Lvmz`z3txzYL9uMnRR0l4W(4q<SicQLrWGNH&r>UFm`1HUTciO=hFzr>=?~h-
z4nDtMXdHa}yIIG1Ra1xkxx|PFR2&>Cq1n0>iJALmVIq%f<$*{e_ypQM+I0?p)8O~<
z*PDwRzw718m%3-lg&H;3EG7!sGW<d1b8~YztY+2M#<Oky&jn2ZjxD<(-!Yvi{)(iu
z6CQv@h0e70%N}%|;EuS)38dw{Q7$5rf1ie;p)o7g$?p?gbYr|B_Vd~Y<TMY!iU1y)
zz-8mBFut>X+b3j@9~)dB0rZ+26Qfe+V4mm0WZ0W6m(dO2PrE~wRAQ3z+V9L&Uv=Ev
z>)_rdLN0H@J2u(g*MN#EAuet^lEeNiQEEN?>>=g(jT@tG{q0WKz+Nc~!>M|WjRAP(
z0t7e{a3k~4moHvC_%)jZ3$;d@>?#FnEFc=<uvq}>fvcC3m)}3)o`x<3kUTKB1t9?P
z*2AOG;QE{QLtT+3o7HYSPxqUPL&-SiOUthspaNWL%5X4@qJ)6j4lg6q7vR6{=7bqm
zF-iB+QRQ6R6GS*!lP%B^GXbs>DQ7xLkekunjL3HbAU3}bB+sUJHXJV^POZ``mJ-rR
zNJ-9Of4C$ViQtRR?Ie>^k)AuMz`^c%W|omT!ikxzB4|C3=Lo(Eda8;AJ^+`?n)J=;
z*nJM^@5+(HX@4j41XsZ%S1gj!1P}cG5ketg!K@u~>|`VeBjPB%!CR;!zwx(sx|!sv
z-3M<gc}dX1)q;k6gkF+e+}y+>AOO5Mx9sul`FvJZR-m83n>TF{6!Kz0_cjR{(@slm
z{)G1wWPdW>e5w1%hU}qVKj=1hqQfHKVxpkPohOeJX?p@I#JHouvfs+Jkuk~}8TtJ3
zwF1Y5N}!UMq`x;0Z$Tv-s7fN^I6d#`DIq>YsQ{bjjHp#$Z|43*?#s#FUY#qi>cl(j
zM#iaCTfLPNdgu&!-6IHSHh{LWZGO;^+X2)rRzO0W)y<_+CaTWh4hZMp6%y*XTayRF
z=qLw>6M$eJ<*YWJX3ZG;@Z#vWdBP8y)pSfEha#=2h!PU4)hH%OLQ0xz4<;21G1uPe
z5FBG%kYoR2eqQ_6WTUWWU&Iutfe8oGehS0F0XkW$l5Cf5pr(5NvIO*snS6YF>^Wxv
zz}Q_|L&wIZd_;ZMn(#9Y6c;RJ69*OQ{vjb2Q{|iQVV=HexKIa>pmCM8MWaD$pcY`N
z{6G^MO@#uVP@Xf)?bOOj+O3t9m8HUHH(_Zzet>0)><8ES;yc^gR2m1tu8AII8E#ZM
zt$L?(syrYxaNkWvb+G#81+_a}1*8O+#43LW!#4hSd-}sgvo2|;GpB3uI#*8_ZhT+B
zgWk%1dVx!Ah(vA;R}=`i`*}0gP&_UNfUtEbuRyu~ch7MtrH?avcc2R--XH?LxvS-w
z8r#Ig#Gj3b$Uf@;s}}HC%r}Die=lm-1>+~fAN<)oWcQ=JebqAl$sa2D)<LFd-1X;k
zP@RGQZkzs2mf$Hza(xE4k0th+{$e)nlVLHuM`+St`SDkv{XLfd&0izz=vi1s75BeH
zj!T!ab8h`3#4hXT>Zaz+y|_!zM@9f1fd41JG9lz@-t!>;$9<^*-<EioxDt6ssnGSO
zh5;RxLKmO9W@6`6$He}enC%z;$c!GEBRxG&e_!|C=K10iJXF_|{oZ|`C<x5s;lz~Z
z5ij-G)pDC}-3b20rh&fl+g4wWExW}MiZDTL+op@9KshTJ+~{gQY_X(DLFUz+rQDGS
zK-jRw5OZy47lzFZ|Gw*fjZczRqhxIZc}RtX@q|IKY)vRgKcGPl#;95EWWCu7$s6hU
z!fh_D;gSY;yp~_t9AI%wWsaA;Y5prBkujG>AG0?~cT(ntiPetp+ZHZq(Q`V~%LsFM
zzgDXGdPT>|$SK1lHjGNSgazl(qmKY`^GImrjtTN`_x;NCM#~7GT>RI${jXDP>T<HN
zh=-bV&l|R7unke+c>iWsjT{jwO-3o%DZegnxWx|L)RefSfey3ICKo0K2HBa{ihr9R
zGMiSF1>Rq#mlEUyC<wqRgL;D(38JT~3$N7}h7pA#*b5oYSN*#+0Z@IwOWC_OZ=of@
znmgaO0sk&5qe^}19CmH&6zo&!XMy|ufoh;f0ou!>1%X>UAcFqR?O=>&OFE#`dPt;4
zGGL^fB`M8@DE-Hb`q=L|viTn(T})FMEauw1zgxWk84E*H|JzylbD{tC!=jGPC1zKV
zx{~Qoe_!h#@))%J+l`*fF~}aBa8;<C>hS*l35o5Gc1`V~TTkNl!>56a@qHP~Fa8kK
zD%i*TK8Ih|5=oR_G(fk&!}_>$VB+EJt}RXi_Pc9oBfIsNtBeUojtTDeH^ZS(R#(?%
zt=mOXFG&}vRlxoB0Ga>lSpbf^_)w$?R)wS;8;F@KcedmGz7+0A5P<6PMs|@a`B97{
z#ctHUBgsD=?Dy0EGOhpnzkb9+trNfd5G&PB`{x>Q)(17~Nzco3*xzPKop+HX|HGR*
z_yI<+UGHxPt@(wy23NkDgVra{Nl~3M@1<5gh`cfE)=Qwg^!;tC0A04<BErFKfv}m(
z>kbVK4HBeL4xjs#oxZ+4$d7^z(b40eN%<!zRje<7;_;rKoxD4nw$#+n05FWOa^7h<
z7}R*`TYk@QqW-znuzL;xd<z9`dGQ43FM@|r*<v%F(r1kQbw$8O2k`8#EpFmqQ>eDR
zKKyyN|Fe8C1+$94{bUe8Q#Fr#jhdr+=VR(=0M;F>tN@s4!Sl2R`mfPkMb``&0B5aE
z=jFB+U2<_z$*EXKGV^kA&E`rg<m5r;Rbe&7Q>R!?URSMPteL>$f<LFCJPr%wZ$^Uv
zIr)dM@CvIs(_~2sCR>4YjUwGfFZOj*-lWUPgD#TnYKpeZ2Ed>ZYU257l5@>*q2-CZ
z?zM9A{4cle5tE_`B^jA;0YEGM(U~TA_c}5;)0W5kDOO$l13G6pphNtNr38R1=RS3V
zM4Mx81(XzG1@{1FHe-%IBU5YvkeQ2%i}<7jhaV!TLpWzaU!*YC-)Yu$cZvnaPNrTz
zSG8;na91>Zq{PH3d>qtTP8j<-a&mG_o{T!*9^5-yIT3K2ncIQ^5=cXKZbjR+*1cMw
z@9rfC$c!Y_0&Wa-;`fY;0?s}Drzas13=$lR11z0H0{dWTgY&WBm>^(AH5EYnM~F^w
zI~Bm%cbX%x<AQp*?@kjh6D5!bwspq@+o_N$?%`U?+_+#n5c+ONC2^}(+W-O4GaS}(
zv!;_F@g<6&U*;9(=~)cVNxs+Q<MZ0k&~T5B!|kH@)b%-_+R+eSg6>R#det9x%w&L{
zA*T0w9x#z}N80aO5P2_jT0T5+3;mt-Sn6qAqW>i2{~C_}W4swj@=2vgUa$$vnm?p>
z**<mo^P4n%5LYR!m(Ev-pSnk9XJ`LJ&YqoGBWnbgxcF0C3Q-dvfBE$l&<mM;))C=r
zu7(&#ET20zV1Ir<-Gfl;iV0{<4P7+P@yT|yUvb2IieK9HGCdBcKMce!+pwYEGx8AV
z|LR!Es3U9z8-h9h<I><m<yy*y9w;5dKyB;y$33wiuLT4WG&Mk_D29HR1!&xVlu=ms
zSpet>qJLe+$r2C#>C@w<c`q)%#8aNCRpSKxxdG^<&d>9u%cZU;(_=f}x4He!^ek)t
zDO&y9(pe261s(2fpddgDNBkY#K%r}AZ6D1o7(d7+bMlvZ=6>ov|9^u+&)6?OEmZ65
zWVm`7PW?U}gsXpqwN`$HS2fkOPlbsC`@5Wh%;cZO^<SIqQIZg!_wQ$(bsS8fb^-cJ
zyHx)Vz)NAh=Z8^48}4O1_ihHbgFlPTvJ7a_d&1*C{u!zCvB39rpmQhprHe(8qX6p0
zzb6qarweFUzC_57ux5+V(A4~M(R)i`K&ZtDvcMNYOskSWu0nN$|MN8_5cuXnXHqsV
zq${Fh)|c@YJ<!G<%<=s{DF}YKi;h1Ye5klFk5QOJSNHb11^qv0l`j6|&!0{b3K>bh
zN|ot4{L!}kr*-}_mj>PZ*0?0nDU@*If1ZSfV`F_i%ho1}+dG$hNt#Zj_;*=q{eNW2
z|CgEj|K&#`LPDYVv#Ekl|13}}Am0c8yIApb!C_}&5P<yp`ucliOicWJ?*~YY{|t>$
z6=V${WH7mYjL&ZM4O9|XOonT{vAR{%s-=HPQSaC6U=I(qs_H2O^bW|q%FLRM`?+Fc
zWB**lppzeJ9#oms(*SXS?#Ms=pm}0q;_`T1y%NY}0VVQRsi~fsl?4=3m#AWWaV*=n
z>V#>_@;K$Wqp5)s{!#_upFgLJL{#S;BC-IPMUB>R0K=h}WJX3tWMGcXwdH8<-sidc
zVe?37xcY$2G)tM7l?D_Nc1E=w(TvP)<$%1g><WGP$Xr8W5)uj)wC6bL9BG5QO^OYZ
z;Z@2?UA={<dDV4*B<Sf>jI38?iQ`39YYdZ<U$T$(6ydm>MN^>71+56N0zri!Nz<9k
z??w76c7<OO9bM&R6-|_K<OD(UT)U~e0>8`XG(ti&d$@O+PbDYE>PxjkSLRS$<@poJ
ztdR((0B0CqQC5-wN1f_#;yEn)DF@}kNuSMfHQZQUM##y-h8ImdDXq(MV1n|NlgG!$
zM~=)BXI0$&;?TZRc(mGc{Yd!xK5dz&2m|7ilDf8V!P1=KiQNAFelbaBmuoWK1e!?&
zHzCmI_mh)vY5tz>M;tF^J~O87?C7r=YAKR_v&VP|gH~3zhsp5DU%SK|0(v4Q?qBLX
ze|nUJgKmILr#h&b8kLK*^xzD-!bqqpBEh=59kydou_T_)dumVPX#hqehSan!G}+^7
z6QGRp*-pT61-$MGA1ovax&WNH*?cpQ?2T)=_SLFkA|V~H0qMS+AtTVEf$|rapZ_kc
zW-a&a$?(o|AdNR519vkAM4+3S+wN3FT54**r35tiN@~{)V85p;&DEPc8Vr;~e(wNK
zu?0jk%u{qXV9h{J`a>TAs@zH-%mR;bPmAc0`O+kSo|nhj<U7DPO#l@aDVGyK@EPLK
z0SG2$AGNtjfi|KMDCh}=kAZv#HF37mJRbJS3D|dq>G&OJ`h5&FTtp<euc7CzFSd)C
z0z!aHw9sl5C{sb>;2J<kR8-UnS4b!>2cb|5puz`-TTa0^W<f^-^e;`DG@lC>01WdC
zo7ri}pA4WxFj-`~!O4IMP$@I)1N?YYTpTt5f#=nU@$RX0PZW)}w>Q|o0MvIrnr=D5
z=54aiI}1g_NqJmO4vvq@0ujx6R=|TeI67kR0L|Y=h;-WEJw61SFDwKSNIPv$knDb@
zs;sOG6aWpbj;5>L4<QK&30ptIL1vgZrNY`*N9LKTwoa?~)os(Pwc8xY37;T8lIv_F
zBt=g3>5FH>ko)xdvY%^TW$WY7(!$l%Rn&q6b!9j}q1*r&1C}IuVh@m~>{{4Kmvi;b
zObiT%t%!UhDovpEFPkHm;W+QE+}gPZ`n#x))hgeCO%iEB&-!|o?t6By11$(zwRT;c
z?nqi2cD0~CHtBIqzP7L1eQ<L;fMG48QWWcL07U(|uW{R25LnHpPFAC}E6osb*^3qP
zl+LC*!FG#e0Cq}m0h<x1N9ICP*`B5>an{Z~Scr$p)a?J9Os#UfKz)6Gzi6@K1dthn
zgM)$SG(##)tY#An%gaRg8M-!KM+FDLwI9RLOan1*IM7Q1eFMvI9U%gG2)%6m9H;`S
z8qN<LAkSc1a~1Q(z-E%g78rBfPC|sW)zxR1^xsNn>l`gzlgHJ|019l-^1QvWHw|v!
z!XzM|tJRsjyWbkXu^yx;+zSO5VQ_Fguib`B&x*46C9o_aLP8KorocV`D0eBTkkyR1
zs3RagP@((6;d$q_xDZ97-u>RE)%TgWwDkBlsD@eo!<4@MekoDWCyyT&Z?n86xi@#9
zif91bW8kJxnouwRlsVqK0f)zf!UQzBwctUXKQFUd6zDl?_JOn?5Ndgv3^W7DqXF3Y
zk|rFA01rRX0YMBPr$d+l1qjWsUvV*gMOJUc?akfwesgm(;F}1zovchua$^YnF*4#D
zW$Ig=JbOlSRPxB2om{`Fkr0o`P(oin{S5bJBP~={PcKU{o;2xeB&8zdn!<qtDnl!p
z^s8;`99CRvH^~67N5a|J`L<y5-sOQ7byy_}h$JXAibE)M=k(o`1;jak@}xJG39!hG
z?*{M}8!u$15dw^hQnVjZr8?Rbe-Fg9el?K9lg9vQUo6vlpecfkhGwvjhYQ+xy1Gld
z^VG$l_u7=-BUw~`BzC;kSFx*OVPVna0S<o30SyONCnr$xd@D8i5)r{{zoqbu;0;5O
zG?-qwPzx}&$|qn#hNdhT{_575v6?PC3nWDF{rFK;dHU%17$`w@YaK1?4}~NodhzC?
zs~jO9X916P0lHtC6_G=k^6h=E83xepeFq#E^rz0Qwu`FFCWi(F29CODgAIWGg@S@2
zoK(^TY)mpU%RfV}Ld55eiH>FiSr`G2^OH}t?Yflt$@WTDxHQjcgwrjB;fb!^#*eR_
z^R=8o7qTipNBBy%L-~NZK*El{m#9sKf)$-ki<%OIzX=fa81y<G6SRLn2G!;CsWmD@
z%1#oO#hSd`ceUUZLm3h(9B<pgb5vAjj~l1!-ISP#01j6Nhssk1W7W_14aRl{AJ}IA
z3AFic<R>~DfG|%0+V`Nk%Z@F{uxbnJMWa&s1WEs_+Z{n6rmQpas*P#@1eZpz#Rr2<
zYeZ|G7HFXt{RjdOLU7om#FJ4QK*1Ry5DOwEOec^<_2|z_m4U`Yi+f3Z36<8wr)cvf
zFF3d=3vh;Lrz~!Di+i~T^cm<_WAnOQfVL^{IMZ>wo6W4`0B%wmcW#FrmBsD>AU?GO
z>Eb``0{i`g?>v1#k)*kptN?Nwusu-HMqIXoA*e`u8*~AW$dZUn%_xhx-^JyoR&E69
z$it}uco1Vi`UA4gsPJ;sWDqzB$Iot;0_gxaaoC+|*CP3Vq$`YYb#bBM$?b7dJ!vK-
zE?zcv0|HU%(V0AFp3T8Ml?$RV;E;!Akvvy{BxPy7B@!%AwZkrU=O-iqksOYT^8%_s
zRoc~JAf5oRWy_B9YrIy`Kz~24&B~`$R`9s?0EIa?oHe2yQrWKc9-f?_R-rbT=9DM6
zD?erj=H}_?>8G9pv>l6OM?%8FV5^xCke3Lh$=>X*CwtOnW&r)PdZn4DhzJtPQW5vr
zgq~pGtPq&D88Qfrx~!}$@KTg%V2s$H4+iubAW<d(oiqpm+wJi{I+O6_3o;H)-H{Tj
z#f*xK%!~TxHVt6!T$I>TP>P@0I64x5oC<~sN^&9(@fdQ_m?#J6hYue@L%+#DvH}VG
z`h4GWGb0+zpsud&8!mxXpaTvh03=7p7s3xnHp#fCwpGJbrGQ`t4F1y|;6?7&G#IoR
z<agpSU^|(i=1H`+2l8-Ni*gi)_0oqJ)&SAuQJ`~dY8mEPOfk~GySW_0W*|VdeI0`O
z^26urm6X?C*}1sZS65YwI>=>kSWF^fhIvj$<Uy&7&Q|OZr_j;d6;2iyXFQx)K#j@}
zXa?x*(PqHVpj*6Q-`Lntxe!wU4geT3Tnb=d=Ux11gGR_(egq!XZkMH}O|Xdg)P}Yo
zL~RbHb$UTPeDv|-I}=6(HB!g9DnW_Ee1tQw-Q|-Hh&X~KCoVj~0Z0Ph-EE4$uj^AI
zGnEor_JRkID)0fh3xlQU_AGu#8v}v&=gl!YKs0S*xHCXl@WKqJYsTV${twOJ0Dub#
zF)_*;D$uD%XPdas*oJ^tO6|C^c!2<{wglr|x1q}#5B~8QY+FbM8(4s6f(>tg`>L&}
zt_E#sQy?7z(v!4o3UYEJ7O+1NiK!Dr50yE6v)S4TnH?Y#;|9z^_9ck;!ok=ih9Q4F
zO%xtbWt~GB8X9Plll=n%u1<!fpK_5~nGI(KgPnxg5|WY?3UIVk3fV81Q6m8~D|4F4
zk|uC;bleuU26>?A?#dLG7<kGxGE+53BY~T908{K|CIWF*phFw+Apm1LivNulckA2{
zV010K{x@4-8R=R6WeYze8#Dd?=+wjpb@A}!>SvATMF-Yr^tjNEv=!ww4O~0aE2@ZX
zr*|m2EX)<xJ~t5a@h!xs5{oNUo&TUwPkpsO^Z@z=wRBVB%4TPDPgS%un;;og0&HNG
zWmA$Dfyf<Vm?gG*>4O$>cK2(KtIINQ)cSslCvk?{&AZF8<vfqu>~DvU7r@Dh1T@X-
z%m|h@j%y5<9f#|t9-+dFPjcX!wu2Fuy*D57`5bT4&3m|fi3P7_d_vG_aC3G#-Q%ex
zR9R6a@Rj#=`C8oR72F3KCmZr0zT3m1;UYRuEdR3@b{&ay1~qAVsIFIoU6^zVP7(BP
zzoZ*Ze!3*>cyHnF$bX}Dd){f8wb-xiX=X;%deba|wfg8IXPgK-AA%AS>qQ4iv&4A$
zBV?OncGslHZ#C9JT5E!5_2@|>L40eHOyN_c9?)ioS0smdt-+@0><=}#nz7}e*e-)#
zJvlx%4P&44JE0aSik7J{d=z9L%b)n_d(gKPsnA!8AM#%qY(v91bIWJ6d9J+Y3#XmQ
zr%u7B|3p-w=wVqz5DdTewe`xKr_p2h(hHWz^@Fa~viDN7ptCAI)Lcl~^=YcD)_`~i
zzkz~zV=@bIVZjsA@8>J6A}K5_Q&?n=ne<#g<sR!6;-*A3ABd^&Y2PfdOz?N5xXTs)
z6cd}AUo_^Mt=Bk`7E6_}sU&ZnCk>A|C|g@a-LSIln8yk7ZNO1wW5Dhe3JXq_SjW7n
zT{LPB`4)^?^-LJ;-NRK{L*Z0S7q+2l;&zw7{Opgg^F7!Gt(mp58TJmSJ*LAe6)i?A
za<frYGr>|(rT&u(+${;Jwb^oNq^ki#l6y0Q%x~zpm-!AN)14=R#+Z?x!1y6AQ(Ic8
ziwHT*i3kN~!v;)$R|vp<uNczfp~>prm$o7yDk2qO(|Vji!o6~vYmB&M#|xh=KvJIA
zMHhVzl_4X-7Ot(0AN|3J<))V&{l>|is2YJMoCjM>D1O75Hy^dTpIFU5$SIwkohiP0
zX?*SRazK-L#uFJQE3QkzV#2ncPbcyP^<N28d6^NF2V#5hVa7|oh}3c2c-?rhM&r7}
zD1Vkm0#7|xf7PJFK5l(BzCqY&WJQ7Vpi#h4Q-T%M`-Rf>%gI-br2%VZX);DiYfE+^
zd~_>x`O2SIKi^zpIA4+R-S|WN8upv4H_kbOxo@vesy_FdAvu^`NlX-{Dc<3~b=I9-
zydnHzPF#K@5ayZCCe>g5cHJtOlFDzKXtEAV{h&X<*p^b)Y(1&8OtC@ha>=ZR7S4Xg
z1Z9*sE?cO5D_aIf<qNvkfrTAa-M&QCQJ#obGZ{_c(ksQHeDn3WG|P!P3qP#62k+0t
zSCDQKy$@O#%dj~p&nlFjo6s0g4Ngv4C#Vv(p+t^-xI86#`Qy#5cy`5;@wX+c3rptk
z{z;E8gkP(T(g>ZFC`HqbY@}GQvEXPc6~EV%f3p){Fdh+<-ZCL&hnQCEU8w)i=1O!!
zECeZkD^Fe#OH#_CpJ9*+Zj<Q@N>4RyQRBJDvK2gxn|c{an`NrA%GR|J<4=LF&IoZU
z9JtWAQnWE=<CQdm#B>}c8jnkrG+yRMu9Eu+xe%mx3eJ0tWSU<$RMnhks$a5%EFX}y
z)9drcC_OU2#*H%$AwWx&Hl1mETOc&}Q7b4CycdvBVKe7xV%5nqJuQyYcJ(n|=TZ=A
zWn|O)pR{OmhJ@-|-i-9Ka6ht+!<;;Hha9{(lfP6)bhZ^hA1gt<ou2SNl4WLnm5Hjz
z4KY|URmSd}*nJ(18?!0&^hwR9)Gl(^^Yz+3Hg2E9Px$EW59J~WL)TyBCRow<D^Ev`
zL6=!3@zFx_z2&hV{zQ-4&(NC^{`A4<oo?<=7SH)Qdt6`Lhq@ArowhGJK4-KpBZViy
z&g%b!|AYwMoWPJCv%-PogWNdkDMtG|G{^{T&ekJ0$_6X)t>SUhlYA4he82X2FU8cy
zDfO;Dj{EJNHg#uJuTb!Z9ue24eO$RqprH_I*%mdA9^RF>4ecrO4i|@Nm+uwiK*1GJ
zu#H_68SOfM+Zg}BEz_9}VTbp9xFI>L{=?dp0oHTVgyFzNBO2@<{Z6awFK}MPqsntb
zjPjO{#UOUxOi_nDU(PaeSL}0rk8N6P26c*XIB9ob12{h-5AYABJeZWqwpT6~-=$Ai
z)J04qKgTEu5wG_q&M;c!5HUe|#o~p4Y$blTKYGRN&6jEA*FO2)|3leDRJgyN=nhxV
z6a7B-Qd2*M<(z?#5UzIh<TD4ks9uAll*Ns{5DEN-Tijhpk8E@%?x^;#JM}qL-^PGm
zFGr)cVL(~z$h<Jaxyw_>UA*a2-494u`V$^w2<WyJG5#BSGZA7IG2zi~s`FkLo~=D|
z{TSCLJ@cW$OrhBf<)f&$->y1G)Vqp|Mpc0cbwowX@Ut}+wu6N%NSZ3o57)Pkugdg-
zH8oE;H+g#`V|#dDiexgFPR*QHlOMh{;CkD%a@p_g(3;<GXL0<#R;R~y$TpP|g^^(W
z%1i+#7O8LfX==ph?y`Ao4{J2D!unULA3M}3!$vJ<D`%-(awy~z^|lYN7zocdGiSSC
zrdhz7d}YI5iMmtKpfT6cc8omhv>k>gQ4`MTs8Zj9;aGGSI@O@om8v9GC|nkxO9pTE
zGen6L!yFwYyY|vJHRk?aWa$`Ekoh4RxoHrgOk&*F0*MRJp3414g^C{EazANdsfh~C
zSKe8t=(83a=8+;{@?ocw@Y9QfCW1Yhmxuj%%pu|vCQXr(>gtcKYSWcQKcT;}Y0p-F
zX)(x<qKm!_H4wr&!9wmOb0Ed|^<r@!w@06OA|dENsmAzGn~@6r4LT!b7r*<(Ev?sA
zcbPOC)ljqe{(i~lUrj`hVb^JK8x$-{C#=|9D)DH%OkaGgl$Cxca^Bwl<B{bp`CeRF
zbR1gg70l0EyFq=;hJqefp~KeKQ_pKB;?@f&VV+Vd*1Jq^+GKNKyvA<kAN;n^fY2F}
z7&paU5g={Uzc6C_!maOmF*>6$Wj7h$YYh59hee7^<mk!TJieE_K-byy!|{k*OMh(}
z8|PuqYUi1l2{?w{`mr0!OZUp|w|3iwnmcDpACUu+CZ%CXsbBl&Dw3bUB!p$+UNpe2
z7a76kZa0m8ZXxFhNy2^;W~7;i$aafZc7{?^z8UJ)eQp%?RN2O3j<{0_H-x1N%4Y5H
z#hs*$SAsBM0|}aY42bGpnl;XHC1?UTMpe?;d$=vTR0z21mf{v3y1Vg8E3znDkw`6}
z&^yTrxW$$G0g4t$tVlz}w<uc!v(2auPt4Eq^t)Y`8<DAOOtOh5R*7E>B(R>?^ZL@i
z2(826$^7=7+3ty9rlFczG({xWI2NKp?Tnq2Aa*AO9>u`c;Y>ySpx$%Yo##j<#%_G2
z9_mk?F%dMvs*20T4kqlzpGqBGd$Uk!i3o8Ksw0<{;u_RiPuZS7tV60bo`h78uA^XI
z5p%h3LwBH`%puxMt{CgMFk!uV(QMq-bs)H0xPY1QYK|LLbYG#KIZC@c(;~(h%fN#%
zwo}X2D4&c3<(Us`7FPo-S^B$-&>%bawBd!Dj+ExhU_9)E@iR5{Fg^NYa)Ua3@uEeK
z><VRC@vHDTHce%@4CLj+oHOO<_@esG8v@?~Hq(EHfEoVv5Gy7I*8kaiL>q8>!icJN
zd9?>SCtH6!deXRbs31BO+0*D;rYPEeNa0_l5y{hrp?t@?P<FWyRZC)yz(Z3Z@}Mp7
z!6$rP_oR!H+u_6EVYYYv;cu0^k%zOi$x;Z|Z(*bmE8^fE45MU9YiA~Fb6~_35@~*~
ziLk@EGTSVi(;`;EhB~48HXX?C6ujIPTlg>-*#p;|H^Z4Z?3GoHap7=(DC9k2$7|>7
z)!|TSZsaNXjAxARzvKv9!}{(8ydZo_Oss)y#C<)#0nQ^7WZxA=F*9g|cOksN{rTp?
za_DEsDo0z**5~Au5ZhMx+Jgw8NMtB7C88UnNTakbH!H$d_PMVU2*|nL?TBSv7JmKl
zdeEC-KA|E8*H+CaKZC@Z>2;_{Blpilai?g*E>Wg;>59A5eMp;RL@}wZF{kQrrTOS`
zk2}!Lpi=|`Ugdrl4+uOA@Ji82`*x_JLb$-<A%yYH;Bi4UbiAPuE{f~7GRLy-D3ad6
zS_~=ZJcf}Bs}$1GA);;Ev-4ydvZy{{CQCtRABhYa=IABpG~^zAM<synWu^@Jv58_0
z!!K*4M8P~cQHm{SwrcD%J!TRrJ3GW4y^xCq$(=Oges(jnz^P=So-%#}Z`boC3;!9(
zq3%kEsEZCPWm!(l+r%V33-{63LN%&IcSd3vabIGQxRjIhF6M6&OFH%PMA-0+nvCQx
zyFPmvK2=XxJ4T#@-6a%w8>8TxNuJ4&J8-N`7gz8C9HJs>t542ZMiPhdWVzSdJj+eF
zzBN@EnH<Yzx_GHp8hPV8yb+I=gu0o(I)D13mo4ano^*P$zE4b}OQ>CGVcK*yUAm6+
z=1MmR8y=#~*qJXv6U)m<cS93dqdgtBjvD7WvQ<A3;1FHelYPo4n>kBcAv49mtQ)Gv
zEYF%N*cZIcS11xVe8)(!9P`B{UQ8q`#rTVc{F5GdtRIX~272S)^NnD{v1tz&&4m~?
zkw?q!N=45(pJCfcW6Y`x?#+H6>L4G|eaXt;`$6K9G2CQ8aXcmw%P=^(;Ui{RDQm1j
z--~%HB*n4|!8dOxFVhgdG<^3EMPIbzTCNJT6BV|zXKHjkCk%3K%-?$a-guPZdfJgL
z<@7lDrms!?^31&cwpbF2GF!Y`r;_R+bPaQrI99s4)j)LQ*A7C)PHWZY%xMj!iAPoD
zaWc}AnNLoGz=wf}tS6_voyGz%_M*>sib*141(;rrZb`89_AXL#3iH+&B@cADk&_fZ
z$&g}TmXtQYkQ@C<DE+ps|E(kZ(i$?a(Qe8#x((d19RdD+D`JtN41tj&N4_sK#ehc>
zQg*E`BU@4WN-*-$S8@H~u1lkcP5B^PVrnWfxft(=S0iB*CN=r59!I^zav|$gMau23
zU0WrH`?!)lt8gi$+WVs1_lq|RsitiNfiWqY?bl+jvCm{fByrt^78$twa03(z+lB@&
z=?u^+`+rh@CsLY>=u_2C;#+&pTO=`%S6}Kgy&9?6VN-aF$sAdF8jyCak=>)uN^aJY
z3QfmE%=>N}dEV<GPPZ}#CX0TsJWEh5=JSg@M!G&VDcBy`fK)jyaB}acPZ(2j96V=F
z2Q#jOLDES<RPuJGeLX>fd}5!4wFJdFeBJ0e#SD0>sQbmqNwPOo%MS!8<XgT*?n?q0
z=<t;n<Q=+l5=}o%uTD3!@J%RVtUHY1<nIvOuWpY=crIgdJww84tf%G`O)V14kjT<M
zOXZa^ebkj?Oca6<JhX`Q(!V3p??!}zd{29|ff5}<;K;=sS=Q2Kcw*Y?*)kMQMzHRZ
z8k!P>neBaE_M;Xq{j&&H;l{g9z99?miD<^#>B;CLs*&2zOIHZ6IUNc@j&KlcTDRQ7
zVnx$lP`@%ya=Y<RRQ^zeAK*<iR8dG-^Oc5^5Z9n|0~Sv-q1s2O3ntuc;ggl<aT~QD
z0hjmWo1%!u6E>qS18Z_86~c&Ka-QOK1adRxYZp6)G!Nbr`5&O;YZn#~8MGDRsd1|o
zDv~00g<eM+lEbjjP&B+B(4(;>?(>b}jp@rjG7MV%Ot|l2(8;s!0G~n}-hcLad?M4E
z!{TbtCWVNT=8J%-CpiiN|A|LuI}MY@IVMz8B6<d$1txWwCkxd_i;8}$SBAN|RZ>h6
zM^m-5&j-FnY1gaiidCYR8)dANq-4t->vxG&>nrIXf38;=(?Wi+y5!<jDMkllALq9_
zmu%o(t?5R?`)%6Y9mN__?0C#oNklK<g^lT^HbxH>;a}~+e1ub}-yIx;chI8aUYC;k
zICuV7n6?PkeSyWQ%xufiEJ_Ty%zDLyV|1#V-=t>)d!uePDE;}%mHgi9H+S^wehg8`
zt;Hd8a|4btY%$?U$2r0u97VVA@JzKBMH_JNq{^gh>BEl=E?@FA@pl&J+L>qcT<v(g
z*e)h856E*Vp%$(UGtG>^I!Ho~@|Z2i7t+^)G6+rQM4LnPx-5o)$Q|18IrAdOxP0Iy
z=2m}HBGxqTp@Y<mH|@=wAj(AEhj;DGe0Y&^j@onWoenU(4Gq3Yr&#`15wr5eV-sef
zyZ7GM`P)+3@5>2VHYk+M&~H7+VS^+5uItNBturm&CvHc6H9(ZyemMKOe($TnC?0Zh
zK(pn(w0l{yMYd`Wd_b~=aZ@Kj)|;C34zVXMNa|-dAY4L=#7qdT)UG^fbSvLpZ4+@5
zv6I3CeJ0q(k~f)sThqyOWv4O)wS<?1yjg-%{;rQ{ts!@~V%Vf~)xD;*W`99!s^ick
zl|@Ge)pmO5IWl3LeqVD{x;}ln1L2m=MSe#}QP|x_sjMow6;?K4Xoif9PwLiJ9OfI*
zHlsYGVSctrN&d%j{dL`(c_d@KuJiOn8s>Z(tWk2tou7044%j!_)$X><hs5oNLfD}k
zv#jviTx|?tcr~{l%*k6xlFWFwk;BYGO|ew^x-o19iv0i{=C_2)LD#V!5Yx|9CZm{0
zYp>KX_pMYsOFsU@yhN2HX{8PMQiCLz?w#nN9Wl>t{}(xY<-}BCm06A<6v4X)a-Jfo
znW1nY^>~&9@jx@Y;7|HcXBHmeU!lK!KhPMZRitI9%IbJyeI8Z1MU4X97%bUmz{V0i
zJ3`$c*+}$mgXT>RRfCM}Tj`a3XaB4rh-;6vxSu`Eejjav`YBv+vnCl0g`WZ&JMC0-
z&4&W!HHk(%P2B}9N<oTkOQ?Nn)YapVr20v(jyvDkCu#_TtPJ@dw0T-n<qqYa!|~|*
ztOj_8AJ;(_n9Lv@7<_MeJL_K7j9XRku3bNpPc!VHQZ{y0=xdyg*Y(nz&(M_^-<~K+
z(c+0boc&r$N8lvhpFVo)U6ms73D=7ApdZQ5jWwn0#2G4`P{4`(_0LNT-rI1Gn=i1}
zWr=((Kj>hdCExMfP9|q-LrF{*L4thUur#21$@ojM#dY2xHD30QFFN>4h{W<S3@<UN
zpybH&%=|zuPIX_<BN^g_<E*K@o#3@h`daI1L=@#L<o}qK4R<BD+Vv@6KTJMRv<<g?
zLRspPzM%!HiRy|S)9a!q0jSdpnv~@jZ9;i3i?<jjSn@XeY+Jbu-#;g3IU&sS{rY-1
z=-AV2#+V#at8_Ci$%Y;U?5Zzuwa+qtsP2sRZYS#Tg{3(Gp#@W*gAu_{uc&*C`$j6!
zr+6sLP90uX7m0_`^2YV6Dv`rVmM&8QRMhZU?$;s8%f?2Cs`+mU8TxmH_)@l_hPUvK
z8gq5T+?t@BDWLFQ@RgItJT3N#vBW5PiD)U*UbB3`oq>$kR<V2>A#~!F$MwkqobJQd
zf<5CkB%$i6E(0U76YzonM*-@}eLd-!HQk3`JK2yls75If7&%!RKIOoObA=MMi0$B-
za65}k3N0Ju4*c7Nsf5)io_Qx_vZO{9S_PO^97>XKGf0~@OA|_J1qqHkM$8h<Lkjr(
z=?cZsmV&(5Lop=By2nM?z7NED@bS&-kgUV#bP%?m7h;YI>=ZadH8YI0-3lvmH0SM9
z$L>(=JkVbHtnH8YqRZ~<Dn!$-<r#<V!*i=;!(2Ms)=Nj^Ln;Oz@s9QLPH3jEZI!pa
za~(nP{*5h9r<i2sm#5DKCMsi*73{IL9?IhdODORlFY!G&D5CggsyMM1Hg}idw`UN!
z2ou=j4ByLSo9eFSIjU#o+ud~G@b23U;=^8Lubn=(JI6;ELs|>oS{>u{gyoQTbD!$I
zv&?om+(GsPe?tG;QstkH-T!soE+Z2g>;H(XA64(NnPEk0*rvL8=sP$en-NFZCt?$6
zX>Dh_8qwC+=64Y`FRiCtjUX(M{?Lz!c;eOhw#NpnuiXXjkHz8H%hkPQCJY}^q4$QZ
zJylQ#u<g<*PO@G650Tp6I$bijHDpClaQ8bs7USrAz3k}o&iU$Fv(D3Rbq}%-)xNnK
zbtWA;ePivdXx)##z^#0P96{waM!7xpup3eji5|A3OWrraHR7YpfZ@k-Rv?#Ei4dT;
z3!8|hzr+xWm?--uD>6OP?9o<J^PbF~2vsTx?%fk74<q%27@WLlN*{f1*3Ty=-m39J
zR-!!V?>>H?Kd5Ei3vlJ^spJpYp5)nF5s%$373%Dqc*laCMR7Xhl$JRBR_k4nx4#Sa
z!aFk2ffLS3%(g6?ryLFSu$-YNBkq{O#h<G}y7ffR6;4NI;$sHCIe-7ogi170yn-cw
zHtkgK%C$g>AYruL@GhXU<J!_^&z8MjQLiq$tbbSWGlu6V`jPU7XDZiOgr!=p=oKa|
zr%IpZ7V`~Lg9j<4rb7o{GgBjt#Xf9Ao+Fl;$h0T=6mR|z{`oTOyk&vm4RJ{~Oa_<x
zmIqJ6%q<V>2U|i^9FL#Y>lTj@KBOtgwArxS+}K_?oQu;{WRexpD~fgdCV4G=lhw2E
zUn4c;hI8PakkcWmUPt;2PNbiNDkNw2R)U%2h<m@J^U03jVt4)gYu!->5{90aQt_?R
zBy%4FslOr7?nXPFJbl|bAMdZ%Oa=5>?|h(=A2xh4*bN;WK>4vo`VHC+no$lZyR^GF
ze&kg&?5EMm45y<3KU`S<+SkQ7lozk*GpsVdRHKA&%-m=`_h;e`XO@@b!sa2_+BdEa
z`bxSV`%No>8jdA~{ynpGajZX)iJNh`x&8cmB|Q=X7}71Z74DG_d@u7zNYN_9<)1uT
zZ?OC3jqF=8GYAz9OJ-Nu3sHyexAn~sPk5O@5`fn4M81QGWUPA>M-C~J!NELDA@YoQ
zvME8T$|z=8=yu6G;0-SsFi8CIC0q?j$p@R6)teycblxwcl$HqQ4QtU$XmdpRwNwTH
zVNUDZn@r&p>ewXXZ`QfxNaduUjktwtGYFB6*(IppG}va1kR*_4-nJ+(5D!2449yVJ
zV*nXMQfz$wY<$(GWY&(hYPw|T0*VYa{&B=`mQ<=$XeJ_ENe$y_Zt$D#4dzb*NP5>D
ziCa(+RK6)ztSPn8a7-TQs8FNvDZO+4@iy-r;?ralVveVEO^khw)FpZ!i<b~nca~)z
z*DpknOk-s&kO&kti{R)Es%VDVTIez($3UMfyEf$SY}Jc15vp32@LGwPrb7y@59b$z
zTj+K9`!8e(VUzp~jYBNCh1<f|u4FZLx_rm=zAA3?pi$ZCSFUj=%Z%?(Cs3y`TxE1?
zpq%8m#wMOWA_?fxYWX6o<4vx2g<lfxv_tA3^lVT;a<f#XCNe2P=*u_AcODky#vpRs
z<scPxB`GJ88JBe<t{>c=jtLrl*2G?;8L`T^BZfpeJ8)GG?RMu1t!`}67Ig4pku&A=
zn=mYszGOQZx4klQ^M5AJv;G`G^@!%s&i4I_V%_wNhvCXptT0L_E$~oEHZ1r(pHyI1
zX3U<OSScFK300=J2hbC$SDC2daGumq)Wy?gpp(7JeVpbt>auPkP)8V~hdpw-v@{`I
zkQyf^n5ZBj?G&^2xYv)%xGdIDXk^+a<2}z=tc{UUfzxJQR}_|bvDh|RdY!zcVM!%w
zW%OgWA>stamJ&tFupY_6t^#6-PyXEB+N^%a7jk@&&y+YUp`qVW^y;^lOmXm3h*ctl
zR8eEQ>J4+W=lMd|-qRCp&(PDjxR|Jn(s53S)X9sbbii308exRq7~8(*=scnL2;Qk3
z=C22TB48wlHGRwHY)AHNn4GV#uN`VkmA&vTM_u=oZA+}nM@*5)FE~7jYH8-B+A-%F
zcnku0IaDMF>&gQmH3Gc~QmFzatZD)=1E1Y%2cLXp_>!-pQDH{2oq7pNqFXydWsUWv
zHUDR)b58hzcjKdX)>9XR0nw^-U!I6J$v?60;v!d;#4T~!5n?EN^@Ye2UbA1!zWi8E
zKYf?v=}3vA;EuQ?Y7mP^r`?7g=|IWN?(O=}UbP@uxR-mJDz9Ide%3^nX|VsY{lKPo
z55_|q=S@Gp_Op`*Ap;{5xgI2)wisJ(#2@Z3pyE`7s?A@!>eYOUC26T#6{uia=2_$*
zJvUz82`FVOL~&?fZ;QnviLVor_|EibTyb`mc7qn3D`#FdYOCVNn!=Q4I8u<Tj9tA&
zWWS)l-G9X9<fKMTFNn(ASZPLKXjJvU0bZ66iUU75A)R3Hd-9a*w(`Ms^~1L2l+($8
zd1*@lsHCr)(t9>v>2*8nG<1qlN*H%-NRTsLFJCiVxKlq5r{}&7J}GJ%TtuwD#KN&P
z_4RZcCa59cykWW7svJ~$v}{zb<GY%vf+7JCq7HLMQhy^-HRJWjrs7jD(eNwg?*=6c
z60<#;H`@<%a=T^OOWA3-2rR!y=(>KU)ePO!h99fe^3|>CQ!19hoR`D<@DYXG>CU3~
z^Z!HJTYyEmwrj%+Yb*v)Iz&ZjBqW9w6%de;?h*y0yIDFWAt)svtsoLgNrNK{NOyO4
zqkz<RJ*cdA@BJQo@Bcf#e;q825oU&Y?&rR*Jg@Vl&gitvB4vAYse2mxDez6>?ZskO
zwpK|tryh&LC#*RAub>RiQ#`oMUFyoV`unFbFT1gyzEqLQ@UD($aQ=RSUa_rCaYarU
zjpXotVG<{L5_h@J4-bXB{`ys1QBhWd`TT$|2@!Q1{4o=@FHhdCoMQVz^|hafWWrK$
zmcQ#b3B%}{%BzMvLv8oZSkwB}{QgsOcJj8T$h3&l2c+VXs=aDi$O<h#KYV;#V1i47
zN&3*DGZtSF?_LU56?lF^_{fPVKF#t%oRM);@xWbHjc0)lXR(h5Z)j{-ed^S3Sjg=<
zPR4d8s?q>Q`BJBpnvIXFZxi0`@OtVB1rx=o<hsHKM#go;J$G1dHo{8}9u|FnPMv&<
zc{e|OWAnZD6(*{u3p=?>yKA`({(sGXUZr~Y&<1DK<&&>f{<)8lku3DeZhl{qSLFQ`
z?TH)7Z!RrA4O+EM58*$^`0QLs-w&Mw-K8>b29C3FZ1QoiPdIOk>uaI=CvRR~Ut8<U
z_uLTN$+wedW@wU6ZFw%5DqQyL>hOKK<Hx56tu>#j84kQJN(~EqanE7Sn?Ut_x6!|g
z%DDe4go2BQ`+pae<!H(VQ1gQdf_g38C$-?4B7Z+7aCf5R$<1~@VP_txQ^3R*wzGZv
z<4ei20mV)~IM4c@z4-n4ThS`Jb*^tiE#r%st{8KgRGUL;gd78?!6Q|r=kl8y4G11J
zaTXP!_37TJJc&6d#`LazOXISTSBq5&&X#CU?eD(!3&M$B)887RH{Uk~)EJ+w6w!Rt
z`lQlr<hr-P#+T)*3nj%K{ogF!lk`6_a`~>`z!8}%XZ@K=>*mZ8F@63x0^Xx@t$#_n
zSU!pICK}+^cS~kxLQ{~KooqTu66i;g>A>M&DieFE;ec!WoffPx!SM6bi)Ck;8_c*5
zXx6T-tW-9(ZS5r2A1wY$xaklZsm+l(o`<QNZ#RjVvgY3<A1IA;<+vt|e*Ciay0m1V
z{o(2hahG5FPSb>ut4Rr+5Bm0TV04k3-)p?SyKtzq^Sk41`lsGc#J+0jo~k{NU_gSN
zjW7^k8j5N#N;%rBcXM@t-1eP`Ox8@EW5%prv|E14&e3a+dDTOU@wG*dJYV&3{Jd~B
zOJRIrGR!HLm;2EhK7W+*a%ZD#h;_B4vH8e3jF9AJYSXBxV*a+kW6$PV?ybv`LgZsy
z6!B*_(aCM!j2LW>;J8qoaef5KGOfvecRQQ4lA^RLb;uc3gRj3$A7DGh(2(bPj3e&|
z(@@cFTl&yrUZ2O*1odPcjYX$8$!-Y<ScprX989_((tgOEI3RMa<nO14hL*5{ffs}A
zu3$JVg(y$z;=WMGsz0<%X`0apoZu3EM?`=XK5nvgjAAp(Kxc|FSKE_-@ZRIKpV6y3
zR>U21rsAo(<nC5dGr9GwHz}9KeF>R`=D&nqyf<<>DkS#p;K_d3z|9j0!Gr7>)HfdF
zUcx^uA}wFle|>IPMR2^RmPoYYM>gTRkHH;1OEuXna=c0@@nT+wH#TV{#5}bwh=gg<
zGc#(ep)w?m?uvWWQ=Q;6berZ53MM$Gz4^$w`|8l~vn-XQmo@rN-Xiin;8ofC<1;at
zheh;p;#8g=mStK-m&W<|MYn!@7+YMm9$D6x@0|HO*{x<me7oP8HsP+n-tj|;$G&zl
zoYE4j@p!-(;QoW;%?3^My3PkKSBJn+y8tfP$KTH}^J!iRekb%WE67II#t9oS_^mhf
zSZx`}<x8!1o-nvnW7>|K+v2HyC;Fi0@PW|#?qBX4`f#(e;Q*D-r7B-*!j_xveO1pN
z1>8cf_8C^shG)I`cvWxpnKs%fitLg)k)!9r<-<$^jjPv|>k2JLD32|E?l|(vlv??~
zgG0HUWX&3)B<dXFIVP1y`jc#j@5PO5r2g<)Bn&caVHJ2<jDM{?9);zVt9Um}(<E;7
zSCl4C;iM`9^Wg*GpBx!BXkK?+Kk?P!eWVXf<SLzAR_r2umTeXl-~7<z*lj;1dG5Dv
ze>H^}chpYXKS+E-e!iYey#VKQcc(9LX3(G@*CW--0cY-?S`l-UX`=dvi=5mcp8GEu
zZC_cxb_jgvW@ul+XzgD<-}B{cLq?^m!42Z4({Hj%&nvF;&|7}w(qUrK2ngMH=5Api
z<Nxzc`@KwJms1t__CFVnE3%R(a^#l{>7FJ%mLh*IvEfiLd4F_mYJX(vf+D%;z$Lzs
z?7=xF!3z;jRKl}!6IScw1J1bzc@+Cerwthd%2rB7v#v=8-J&feA-S9xV||I>Kr8zp
z=0l9E+}_Fhqgop^y_am0EcBf}Rpr!$s#z6ieNpDtD>pK9ZH}ViNF?Gj6a9MebDLIp
zUf0_J6&;4wS&Lf|dHE{7o@*`Q=AlN3GhPID?ouW)eUlAPxlF>Sm0KTJP5JU}Yo?}J
zwE&4_W#<`B!Zf$B;P=zXr?lMy4JAk~b2(7*lbQEw5my@Yi&UzJ^-;;+5ZvK69AW?V
zmCgJ~a`}%{)9+&MP3d_5t$pYI2O{_Xw@L&*7q7rxI{!baM4((q+IDtrPd7USSvV>4
z6+aD(zIyufX_<2lM{k}x<r7CSB`0y7#fK@DiS|2+$wh*ggy1DBlg~K~CUx<5F2n>=
z<je<dzg-vW*e-O=ydre&!Kbbwu?ANchBE%3s@i(jIf1UI?Gd8?9$ILZ1)~tPfV*V?
z`Xl3tlT$wdmk6YaIe>u$E)gO3AE~a}Fu}m@vN#AljX)t6ksx*S{M!%rfI|T&(g5Bq
zqo5F$UJ8I%U~F0hytOmTaRRV3F$6}S8fU)$8c-++2xvgjd3+&2ppYodPX-TK_mm<E
z3U*-7!Ds_W=78-C)DcK)OiU1Z<|}Q_mJ+(8;M12!(~ul2e*KzaB@FPIFoT?^VcPum
zIfYn+8bW_d<^SY5e4_Flnm7E`dTMH+8#OTTEBbRX!RWlGs25D^B!&R<Q?gP*1%OgX
z`7z!*TQKM1_-Yh@qPtNlkAWEJCHBkn!m2kQhB5c#qgtA>)xbh8JG-dY*1CH8O3d!o
zM5wdz_JrEnP)gcBXLoo3fVsQ<cYDf++P%*A4OxFypG#i`94E`Y4IqwT8E}W`haOva
z(qrHk;`P>4QCS7lsPE(KY@Q|k#1NoqwYRZ36&5{w1c2ACVq!q`v1oI?aM&$h^20Tl
z`zjNEKh9_c%kP_Z2bQ9_I!eg-C&fW`1sx`9Id{><#KE&>OF(Y8RTh6m+u;@eZsYqf
zJ-4Y)=gzLKccW7sW1ET?J9CXYerZw6OCDEe=966-wbtA(uqF+$Te~&}#_^`U1(Hi3
zV0`~*U>@+=M1Idyg=;DTpMk80>X^Hwx$t_|y_k)Sjc_wE<Lg`{XHEc$*&T+5<XC}Q
zlEXaJauFG^o(BT_Bi*W8&0(mZY1iRvW%MmnIrZjURq-Z#g;EKQVQTaJ#X=WZoVv$r
zmfM}1dn|jLgUL>yB1wpDvsvl1>SLJLz2@jXu9bX@F`D5wXoT^%clfWks+gD{>mN(b
z&c{)#6YC0*qGjD^B1*n%g3ZLn7dWi-@gtlLKPolz{jFT-3f5-ZwyV@ha&=e*jA76r
z#0C@`>gGMK+R{<=wrFECRL3Ro>pV<M-&XA!fqc&_ryK64SYnqvmpsYtW@QWq6D_B~
zU+CtjZn<>5VxT#RNzXXFL{nntGy;QA*`J93mR&q{2DCzm??;k0xC7Am0y`S*?=oEM
zU9*`X9TVl$V#->HBeOhCMM5}v`$v3{E2RfJ8eb>m0w=+18`Z1ct&pKvbi@l-Q3N@#
zb}iI^b+$LNy{+Z8BAP(z+*wtnDd_kF`#BV-&?O}#DE%d2W+|gF*IK{jr-U!r#`^j*
zg4lJACHHVO7#2!~5vZ!FDi|OEqI5W8K+%yMY0;fsRo23KiBZ4*TKve*&>ty}JWrLq
ziIA7~**p~l2zJ%oP=(&@?IX_yuG!X{72GD*_b|kw`^8+7jcg4xM%jg!qNfi@wBBz0
zVC9>>jY5Z#1zcvmulcj{k;02AZSyiUunW01i*X&LD!qo}>ijgo+sC1k_euQ~%+EyZ
z%Ch-R_#m@G**h>}G;9DPbN$`=M_?i{#q`nPoS8eUdZqY);@%YNDR%3P#i26ov^y}r
zLWa_x8A{G0BC72P41MWX!zhorT;PUhuS}W%O0&41nAaK~GluU9i3#h@8x)Ewu#zS5
zvwsm6iX<<d+neIHpXN8c`01JS2UcsNRp1B^_%+%25bHE+;HFl;sNT+O)pb27tYqWp
z-pp*^{y^|)8kMTCnlJ^;s*j~Re+Ht<OK4|ptvUG93X{Ub`^^Of<_1dXfIZcUM%Z<E
z6!;D9v6{dEO)R_$%rUQ~#|<`LL~Y*kkX^cV!v94N{F>Kx5k^Em7>YE=MLfbf4GsdI
z`N6O8D}rIw+AccXR~m$Yi{mCsbmdt(<1`LO8yil~)}!d|SgdR)Ta`t>QbwNRE@P!1
zB^G&Yq;-jAsa%GUZOvx&E$1Q_fxSpt8;iSsLP}oMqnF_x|29ld?GFn7=Leq(1YCd?
z)VgG&qFdx8n^o~2l3|SLJ@kBMMp>KOnJ5kQ{;lg53LIwknMX9nXl4~V1(<ntGjhze
z&!zPOGx80c%;CGT(VB`HDtlx3ZH+kC4huQ24x-Q+9;tlj2yUCfs|#N~HRcs8-#vPi
zmd}C$t(J)O+IevBZYs&9MbnYet(WyRHb9m=)lw<Msk6bh`P7e!_c-bpy89{i45R$&
zD)%$AW<QVC8X}V;BQ!ZSN3O!E*&pz@Ro=eEtv3@3Mo`|eTd9A^C$u-u)7F*?t(J)$
z5yhPQ_-3&x!w1uuu3NYos!Q5n=dm4nO`7H+#*_{%x-lA*Ht(&Ifk`o$0QLiQGZgPx
zLBqfzI%?6imI(QJQd`bkWikBlEZUQ$4wtv8gAX8?<($=|DPl*DJ5nclG-^4Pi32&D
zhAK@qABR>Otp`lY)6&r9L?3arcd|`U=0Esn8oTiY?HfTya}AGz9M`7N)`I=v-hcbR
zxd!)aQiI<)YM5?Wr#BlWRQ7hXojT1{X_vjEsIW|ie+(xvigj`j2ta$Y=Ql3VkEQAL
zxF05LtHi~AdjBd(U){qnDG6r{P6GK8H|Kq@FS&9eS1FV@?-WU#>3eKk7lF|bM`kjI
zHCm*-TqN!*@?2e@R!l>#WqN7^n1!EVM=uMmPYMlGu_k#q5}>SovGSMG+s7r?X*|Y4
z)UTOeVu{xjO3{SbF&m)tM<#tJzOzQ>z4_YQy(A4UhtcQ7yEHTcc5kh1-VR?hSJA?P
z?3C&Y>;pq3<IuU#-1w#|+Bud1SUVK(-=b_ONELM4YK-W14dAT8!ny>Z8Ah{_wG}{n
z0X6E27rdSi>Bs7H7pSSuKF0c>3(Q4Q>wMM9acY*`ZTHnL$Y&Q(V(HoHf%x$pELmzE
z7RH*Isj+EwzE5tH4;q*NyO7ZN0{OBxFWxODE-kvv<#Yg1$leTpe-ZW|icrS7D`&0z
z@iJLR@J_q7byqa@Ib85I{qWw?)b61oNBIM8Rms;ZT(lN_5C@~EW#ai*U|v6XVT6<@
z;wf$o<_dqa``{vN8!IUk%VZ9*0_$4ZA=F6FZF|8`7PMv2!7_x|nMTEilfR?8$aF4s
zD>IY(mw}=IqPt@M2kjNu`t%5(a#nwVg->wZNcs0EUX;fGk=N%mHM{ji!{0Bt=rjv-
z(I|-p66Dw(D2vz50bBU%mqwLnpZPv9jE-&Ac4A4`KqldDs)f_PU%a=}8niUCAg84!
z&VdgI{ASvc0HenkQxqB|e?E&R?fBdiU@uh$vN7AhL7>EfA+7GeKQZCmZ<~CcQnFDv
zk5+5Mt*$yO&H2zVGqr<6)=(@h=5=teN=p~MTNB1zxzML+fLI;ofm7NjNfx9@LqfHM
zxF}%@8)#{1S;e)vHu?FTq{%*r0{-B%df(4%R*kU)zh6`veugCIj-?(C`p<d#MUAlc
zJtY`xNc^&W-YoQ$8u{<vKlc=R|9#QJs*hv2I%8ph@F+U18BIn`E>vhnq91s!pi3*q
z{1>GGd@5rU5J}TTcBw2Kp03;UJ5EGYxwmu|F)mMONQf5Rx4zkW^>UevM6>5%h|q+A
zhiA%4JYqpWS&SVcsJHZzEc}%SSUFhFhgfGk4rWa&+wW_^K`Yvjs_>`a8g8c%wi*1}
zH?o}(bGv_a=lr~0KSraE21r9~3{)Nt)2FFP>UnvIBD)q>I2H>w;_VYl>n@!4?&bh*
zlx_uE54-h4Qs-hKwAyu)a)g5-n^luD&)z2A1f@DXlK<kt{*WT<b64K{O^CC%+YNG@
z8dVdPT5A?k5|$R1nRf4<p|)-8lDlHs&f;2lul!<hfY`4iRE{ehi%l@DOH!u`9#FqG
zI-_qU^8NGpAk6Q9r8A!44ap7cHX~a&4SW6~Vrg8iptT~&(|Ayx2Okm=#QCgUwXto9
z7ILoF*=SSO&o@u4_jRN2fV~Wd&Op_6v3RZD-hNG`XhvN|52e2wJ3SH`7Lc*}*qL<U
z!JtOV)aq2rFH6!0VLS{;Mi$C1rR#@Vx@q*DYMZA8;{YBSxwyGbagNe+_u<ySr#Q+5
z-!DYF#r*w#9q~Nl@k^aq_lVHdv?ovcv$>8t;uS)c7x~m)U^8rOMa7AI809r|w6y*>
zd;XEwr~SD-uLa_Rg8Ae6VMyKmJ{qXF1vy!iCK;@y{kspNP~b2`2um(oY2g(pfjCGf
z^N#c-48>`(_D?le@n^_GUw47mrwK*|jL#su%aAhqb$Cd~V=#*;Nm!PZ_}p28yLS^g
znUai?i@=w)%n+(}i*wL<!}jX0czl8!J{_EVSYy1n1ZrCX3=CbUk&P~xG?rF=D;Q0S
zjXih%d>o87!tfj4M{h`ZfK*cX_U(!QI%Lcm2xY6CKS^hi$rTVSh}gyzZ5%n|z1HX4
zRw*9h3S=a>p0Poog||G3RxcW+A}6xc+tK2q@`q2R`V9x_IWF$*3gO(yY^E<cN4qHx
zWG7c>34vD!A#QZwh1pA(hXetOR1eg3`h>F6+|A2ERgqaKJs^XE7^_7-!Mr<Fr|<Fm
z50jEp#hg7$dp``l1L84gG#awP!l{~P&z=Df&sDZn(8+j1(o)Ue?{Z!E1H2Yo4UEUl
zP|dvjasGm#QVv*os;%pLZ~8-U9Y*fg4@to`^M96X8HR>g%N+P0Nc3u99DSl8hK`cb
zb-b1`LllJkVD3IK0-}2XA)#MHpPz=&CCk;UqB}V0?<-?<{G1jQew4_bd4>&n1GGe&
zKYzlg=!}D4mS;vr22cy~veOAUrGn!E0#%i*AHZth?&c<WGZPr;vdS5hyRHA`MFnL>
zf(pnrEgE87S-iN}7|eY7i5s@l)ZTButtAX(4w8@G`QYUWi98`ydWS#%l8Xghvi+l{
z*(j~>DKBapd+lw=HEmLq@kr2HJn$eW+S&lgL$i@1O5&qw`qp@)bt`SK73{zQ7POn|
z)H}TGu__icLXM=xMze}Osh%gHq4>zRLRSlt_FImq8WEQZ7lcoG`8c0F6XmU<5}d8k
zC*H&??b}8htmaes$_M+L*{peZFi?>c1>CIX*xQo@A&(cgAA=Yjo@4bHG!0ph5q@dk
z1mh7^8j`XzW*Hy3zC>W7$LNG1E-vuF4mD*o8-b<<De5sZi|0#<hh6B0!_vu$7O+u_
zkVu|@fZ%?6DshcbAQSpo@?D7PVPqJ1Rh4?bLIMTSO+{-TVc7iYm;~>mr9oh(rFP7m
zj&(sdxP0QS#9C=7G@Y+mAmVdt;wPmf0^)`aOU+c5i44IyLG7ZUmrLhHaD~)M6teO=
z1F8ccJ&+XbCd1}1*Zm!cTlmu;mD^>G?ps;|V~*>Yjm`3nC-IEYPG#SfKIGhr6XN1v
zp-~A(IOnpHEqXr7B{V(r#%EI01A}{H;&llk`m-r8j(n=@3y=)FpuDojIc}YRHR{6Q
zlx)WpC#YkW77=c6oJ<YySRm?2&l(Ik5#onLO}V9o>qH`;f^kU=)G%&}N$-)su+9$>
z?udSKfur%M+1M@oy#!EuIP7e>ez*Q(t`P@lE%0kMBH#Dy*sp)?^T>b)77<{&d`~^?
z>2T74{5&%u$LW2o7mXYm5;1&BkN(n%k@-7*wEweLWWMr3rFY&ACzV&2@LRVe{_+3A
z25ojWtCLOfvg%t~jkSCtpv1_AM1K*TvHv_r)q#^!u&KM~J)z+1z`%2Yu4DXtn}_lW
z3x5%0?)gvS8`FHZi;ijQtr=4Jll}HZvEYNb5#tP(@gJPe9z!#zCBiuTY;&5_<_ig*
zxyQ_7g2mq3Gy1g0Ls)yr$<lNS)wH&zi#FQ%v^844i|uX%WV?L(MBMK46zzRg%XMqr
zc4@BNXQIyg7j^Gqx)24$>9ZjbL7@ZE{SKl8Fn1kDHgmE>+W+`yNR=yXZw*~Gcru*9
z(e#k`Jfe!l?zp%(jA=EnS%^6N>_0<ZtzFYxNi`ZGlr=1HYtB#72hZ$(A{t0S%xslx
z10;1(tYp;KG6_a{_ICG8CwGs}MYkQ?`b#8F_-<dnK*HbispiY@{Ms-6q<%+~r=K{X
z&k?2{D#JV?kZ3d;srwhY-2BE`D8>amUUQ+etBZ>_zgec`5h*3-@9gYgOr08gIIb?e
zN4weP^x}Gdtg(>sFYw-;A1ofxQ)YYNqYxs5*-0o{uD&NKe9@UxFyqDWU3b5zUtC@w
zgrgNhp2qIn2*KoG+F#pxe^0@KR4cSkIwngl0EuK6XnGNPdbJ2_fbjA&%(xH#6UR2U
zLr)+5W_cR4bZ+dcVa!$DeRzIkOzzV%)PY6}-U%>i9D~JkG4A|P7R`I)$xUDEwQE#l
zgd_!3+Ag<6d<&ERaHjeo-VLKsg?>`emcz$kJ`@|)f6goPeHcGEzonJc<=+A5aIA=s
zDd+oI-|^AWU&*%`0jf`=qGUJ0<FrMqw*DsSac3;{1ym}o!1g1uv0vN2rY&u>pg^Ii
zaM37-kBxmtzPX4+9S({lK4tIMeIoqY`?3WO{`oV51CFpR@9~S(k~DXI$4YylBjQi?
zzS?O0e0+{Sms2i!pc*wUVP*A)FTGq`$O%L=8p{UnjXkNzsQ(w9L?|PGjB0h!zm*dl
z^mfK@3tldKJS@b0h!qF~{%Bkw;pX;%hsqJX3F~OjDC<FBP2BT6$T2Y;Kw>At^pO3i
zZU0yQ`oQHnN+K+}^OJv|ME^tcOZ?XTe?)(5A<t}``Hss-7NC_KtE-{VNs_1C5A%hb
z8f$7~|3r&_u6=(I{x2S|4udpAz`CczQ#R`dO~>TNaQoM(;;iU^iSFO=Nj+4VI9=q)
zio*fdcA>cl>Du`V0@^986R~<9iDbV~@Z1k6g52$PR^plc8wLfkh2k|glmt`es}9e+
zq@i}&Ru`8XO3u$nZl>1KB{4lHMdx`jR;l-?JbjX;AL{Ymmix0x-W?F(H2GlOw^153
zQzwl%zv4$pRU%=_^k7|sZ@&Nv)M0p2%agZaX!kFh6~keUMHg_N{k6%m4dF&zAGX(G
z=^yF7zQCHUiTIJ>*r!SNRJ78E>Ff9}vWSS#B0FfV3|F$&H?Z1|!e3&u=Bf&B*6Ws}
zY27?LO49#vJzPpFY6!S_AX0+Lfq8GA^0FY4X!9Y~-xOkyNhG#QuPqajL}81b>$xqD
zR`E?j4xm)*4cza8mJWXc{JAP|2ycm{zejuTxbEN8|IZJyiBTZ)=KR59KO>myM^qYu
zgM*)?-1cHb@|a&=h0dJ({2hILp1`gGJWLU~h{#fIpE#ZI)xL4v(cJJ4-VL5)BEN)1
z4K88Uf%;iMQ1CeOydsh7_i5E7dNwt^e7iAV!oz5Y&^mNb-k@VGty5{542KX9F%Xi8
zt?EVR_2dnK1rcXHJP6zqpy+(wZK|il<h-t#S+ZmTajD9kJEK5Hb!(=ky4pc7YVy$W
z6DP)Q3>!Gwz_DgRYCZGw^XXbdc<SL}%xsCt$Q^Of;{h!h;PT7$%!2UCtnEV^aJO9u
zbzbOaKk~%L@cVQdu)JT(fMS*Jh!~KjAXi-Ei~dmSdz|W5Hh~M=kcf-qv-ElXoR|a&
zTXSc=$AhSkvy+o3WOMkR5RHjIGa+#dNMOWRSb*ep@iT^L@k7Ulnb#(TV`*NSugLXW
z2=;7@Vgv|_wB<eTUH*ZHqH@i<gWwOo^gtn$=c}?suUxUmJv^K^M(4c+EKExfK>KK)
z@>G(Re$I=CG_ljn_7&#(Myl!Dhg8OP0d$+w1m3rq7h}JYPF!>>c>qjYh>{#mXd2Pd
zyf<*e%5@Whb}~QF<!|%|)mDjTp?nfFQKB^HfUv}2>b5c>)^wu|7MwJf6O^i;uDq?k
z^!<Au&}4O$W&*VkHl~ypHN#&k#QK~<Yn)^?dwft?LF5wqd}n?NFN%KTHjth$A>L5K
zzHxAbF^xosE+p`79j;!tAS$g|S|%-Fco4{C*xA*b6}9q>E78V^uX#}`5jnwRAmg2*
zS9lsf{6zN%2yLbJK3`F_x0!BDMk3yKA$UCA96jy%4nI5;)smsI#LJ>&M)>vX*J@fG
zRTQ*(?xuTyd1$3&%k82TU}FD4c=Wcd4G_DGc?qe)%dgz42kWCnLqhU)*ud2EJretJ
zkp+^a2r;G2c6^NS!FoNwH5>V>)rmLJ$@n^cKm*NuqfWlloR%Ys-AMzFD=M-pewWLo
zh{r*E5GGXf0WHG{tmlrTsImlO_OnpJ0f~)}l5eXh3^g#2cwFF3hYAn0r@Xf|&Jr#y
zUP%MZmN1aYf&22B+(h_2<}3-?XL#3%Jd=zp=4{yO3+(e<SxKOS&}$!%=KsrjflQDe
zijs)?Y-)B#6$vTnb9ANAJB=2QCerR7hpYoaX=X`}L=L!n%jIVjo*fqN0ewz6R_J+L
zy7ah|XgHP^1-w7@{!cq%HB&~@=Dyex#1UhdoNa*44gKOp`%d!^a7Z9$+qL|z#8cyp
z+o>6o!B!7CfQ0t=Cqz5UizDb9R2!RSz9W2fu#+Hl-78d;i<6>6R!blqwN-4wQP=O$
zLl4RcGM&8$5?`cUvp}o_nl>d+R~Z`{^8{U=aiqL_*UioC+O@CH085ke*{5RdH$#Q7
zX*SO`fMC|0yLb75-U62km46~A^*CqT->1ed(nh4@5K<n9!}qFLP1Zz_E6=x7qm1Zu
za`vQ-O;IcHK$-$HXoYExX<(f`kHk(+j4dpm8HK<YG3|j3Lr92kk4kIsg>oTFrFdF>
zA$SoV=Nel(TVyqG$L~Th{C;pIl)7N+Ad$$RuCwCE(7Tl-^&;&T7*jA3NJ$9ibHhSy
ztlfN==~wL^Wa#3t(<|{Hr?wZ;s9hZXy0o$~)06wrMJ?ZUq~elnC<z&vk5nu4FhRnB
z#mm&rF0-1NX5a9pWMd)Od(CP15MSNgeA`wj5}ijuT0LO!$B`i=xM)E~8BWl-x$OHL
zsb<dX?sgF!J?a7y63F#6*3-0_^29sKAQj2Dr72j0wsIVl&5xoUyOGW4yG>nEXxQC7
zYiKKdS+9s<&t27uXAnV*P|SEaI9IUF7d}o(Ly@VZq@;>UEAR<1)1Gv5cPED(I0i&$
z_^K*1=$`GxC*J72h@L4ENT3@V6}eb=sQj0fmX2V>*%%oa=>+U<Y%M_hhznSQ@5hAx
z(^Wt%B!~gklM4@;lKX}q0iDa>17K5H0gguJT;ztQ+M#`7hR5TSJ;Gwb^kWk}ocHpE
zUIoay-~uh$K7{CNu;U9$vZ=THde7}UD53r07YIFp79X0LkByD37J6XAce#;Iy!Yj_
z5IdC+CL<#Qb&*|*jln@(+r{m*0R?$kbqlkp@8y^F=*yjJ#FuW`cWizW^S;3|5Xi)_
zu>A2!v#H0!3nacGo?9RrXkCYE5Z$~b{DXv_`V{5Iwh*Bj2?+_ofvb3<g)fV(ttvDi
zq+<khoQDN#eY;;^&oJEjX0Co=_lr=|ZK}`dTkcfI-ilpqQ`{l8^$EM@Rdv%>{cvHP
zS$Ua428NC6%EgPL?KC=2&B%(Bw9VS}d^_d&HqsI*7-_2&KKQOVcPR;pq1Y+URj@_8
z=<Z-FyPH_N#+#OO6cRc0>!P5eIfl3}wQ{7wlaFO~Rhy8oZ48F?vhN8xR7xP$@`Rwq
zX6h+&^^WR|QONd~QHObaa-W$A*}SQ^eh+kMB6zBBxTbDnjg%`YFR(@*Q|GW-eOwq*
zZ#2?UKX&Zc%0vSQLu1ft0w{YbUdpXyDIfe1{auhw8#2hTLxdw9lH#?WRJ`R6Eq%I_
zF0>8}?i4>Lm3cXP3Qvxwm*J+jHumle<f=N^rb=xQdrA0V6p&(Xq=xLb;g&`B*p!RT
zpvN7G6qZyyl+5zUbOn)1g~}Lc*>;zB3hn*tAy!^q{_GKbytCu;dAunPVrd{>L3&)k
zrxhuB_nL6+Emjop$RTvr*;kORM`ni%luw>xL#7Pij={ohZ*8$u;$92-lUP6N&g=Q)
zgPDw9q7d9XvvF67<7Z+sO{*)lFP*36QtnFQCZr{T)&tGtp-l6fXs~CXSkDY{L&Wrw
zh?39jAna_Anp_x8k{^U@Y7zW7C;}72y<8rrK6RTh;VfE>ABBI+;X^8R?z+PcsePj^
zjr7LdV(bN$I<QZJhKtHubD-i1l#nL((dNQM4kZt&g=PntJVnIN{r%5Mnrdq9g9NXU
zEA)5nq7XJcE@6_gLp0gcD0~*xy?f;N>Zb#pG%;_UVA~qSg`&wxkHWTqf^}Ux4aq$z
z6Fn6NH>A9c@LDZe0TUY;8fvuisUNoj;wr^KR}dHm0sRr47MyQCfpB~x<!YXFsw-N~
zXGp3eabgZYrx`K`8PHXSs3*`!R!w?vm=$9xgf98su8rJ^Ghuyk)`}YCeU6Jz*oj}!
z1p4yFSP;w#u~`He86+MA3b}XsiW+(ow_SZ@()pK%q->xwIT^S1;#O7tlm?5nmI5^M
zY%d{_X#R#dBx^Vho{F9xYLsQm9!Z%IR9S=iHC9!ew->Pa`T0Nyw+(cXpsa0er+D&6
zB#+sv;NbF^J0H@pfamb)tFR>XJ@G{T3(kBpcD`X^1-uzh?P?u^ikfS3tuLbWu@3B#
z>xU#XBFTBz4?`Lcw0qO@^FS?&C_P8Xca(a2$Hn>j)_}12pIR}-bPeqTVuV@m-+yaq
zSt;cYadfn7ipT5j1_~jDYHL82PK40vtlqG`>zQd#57kDjl1rf-aI=Y7&?Em~q9sDV
z&baNsu{=II&ER=+&rI|o-LraRIhMGOu$CyvIt-E#9qWnNQ%S&SKbs7tD?~J?y5C{r
zok?{=hy8`1?m<SrcO>A4C&<_q^Y)&Q@+UX?R>btHJpRWr?wiQ}`GIK`8U*_q@zsR>
z%8^d3ZqtbK+54L7!~r^^!@=(M*58qLi_d=UM1eDC5@L?0j8<uAYT9SA{!}gCva%9y
zL|2YxjSp{qPdfsR*%~$AzEGSUxhHegor{J0V36Zz1N&On`nOsbd^q^&UfJkBe2?wJ
zxIkv2%D<<u1r>oKH!7IDWDJI`2km*r#68@P|F5@(_8x2z2!niLs5lA^`_Aw6Ocv6y
z;F!Ce&g?Z<r5KaW8tczNsa%4Uinpapjy|onzFsYyylwg<vO?bmkV{VKcWJD;B=FHw
z%Zt%BG_a>yVg#h^o?)Fv1`&E_$O@p0JrblKDquIdcU3!mu1!c6CBgWH%;(`^9Y4by
zwgQK}gd+hdUOTQY4go9ixY!9F*6g=$N1@%Y7ncZ)#;%*-^Awi!E<fOs7L3~KezDde
zF{!WY?Jv1@V=_&;zVx6xXMi&Tj#DJ8fF1Lv>J~CS2!fvFq-&kV0h*}@&0smN-SVjP
zKi<=65sI|HUfBuZ6W;B7tVXFx%`ZVsBuEr;LQXDJY$5!~)!;Zpf`*tC_4}sVHkRy)
zcD|JkYHVzTt8OkD2h5ZL<SV|ug<jikhxxrA#0SpoXaW0*B1sEn(+dTdCidS2BJfp?
zNXPdnG%&b+o;Yy=61&q+C3NDBV26O&^22(?z1r93Kl8aR-?3K*h29hAEr2i(7)7DS
ziQL)6CunoWe&Q<4_=#ZtCfO64R$aKxS`R!mj(6b-D2IO1D)wH4ZYI=SSeTgJLn#~5
zIOxhTij3#J0dCDpjEo>!(l|RPwzCw_+S&?5{^w(WVLM61vxfz-zALP;-WV*NrB)Rg
zl$DaFVO1wC^)f=<hyURjYNW3`z7~`ZO05v^YU9|S#SH+5AVj_<PKt^`w!=4CmVvR@
zPm+GmH`(ZAU-%U@y{CAf4)O5eg)H{;ItaOUmL}xZ>(Ghs-m$~M%0_6Vn@9n2rnJ-p
z;IC$QkARCHvecVXigbEOQTEjQ#SG6UDOG4lF3bA*7C5QfD%vpo%C6wPk)jHn7MR)3
zig^|JV2y6Ev)7xwP8voOt7=1r6;lLXv-Hh^P?-vihv{fp^KdW^FNzT~QGdPOu<Qsg
z>iPxfJ%G%DQ=v-}{&{5<<V`qWOTD?*G`MSGk}NQ37DzVGh_ge<wc#L8?9vecoh-=M
zB8%g&%|=l;DLf8J+P_|a(3|#&n=KL221DkB0ZPc_r`PJl%}l^^Lo)+;pXGa+b>dSw
zTc85^UWlo0_SaKQZO!0cX(S3y+GyFFhgE1v5|Zw>+n-PYKh4U^tGd<r=+PrMBigAO
zkaF1B+1=<M2HIed0rM%MrDtFm2c#Vtx&n`9*UT&zPCO$v2|Yv~Z3+=L2Sr82pwi}4
z%`1kD`K5?hBGuMbngEk2hVl6+fuz_ZUKHMhiTyI$w0RxwTOJA5wU=Zw#h$4Gri)8K
zYCe}mSkBi)V5171g2;eYxIF^<y~io$%Dora7@^#kjwpY?gYl|^Mr3nU0oL|8b8}q@
zaJ)NG<z|czE)vtf#Y-nU!t#%Y-+POnhKv%w!bOUL>hW7V@W888VkHhjy7dlz1>(`a
zIyLPn>I%Qskj=SKh4u}|gXG_S{eC{yA&cMP)0Vih{U=D8ekINE_f;Y3O`VU$Db62-
zR5v<0M2y57{!CBK{iNyGfV~b7K18ere)e(9q$V^OmK?Y$<y2|>{QQE++^G*jo2Sn%
zKPz=QHro!gF+QM=1Ph_(=cm!pS<$7kag&pWu0ys5@G_v^9`<<+c*~TO`!+V8PAP8V
z-|*Lhz8O%&F3rX~_xFcfY7Vj<yxUUN4SuT+#N*~XRO1ri<i+gZ0N4_T4dA}<hLb}b
zDjiPPaLiV(umhEX$Vz;zFE<B^NF*fAe8mZXlRS}V7)!AXMAjw4)Z}G9?(O=t;*=63
zLC3H~Wd$f_kCW|L1U|v_fu&e@C)um0IxYO3p?)VVE9(xt=XLtWLmpvStncKF-JFnE
zo9kgbTNQ8YG2+i(uIPaGAw^8POC;^6Ba$-khb=BHGAk#(E``|iC8!ccxwO?G0<i#l
z3@;@KhF0Z(gSK_Q)E+<vUH};<B>HrGXa3k#oX}jXWCGU~sk8GM2nbSC|6?=tXlZy%
z6`TC`>$@5v3yfv|zJ&Z={oQ}_69ZM}Mn((^w0Vjo-tnH={U`MT^oJ+s6Fk{TN)BsV
z%{mIA-!3$^PYA*R#6<mvFM~0m*ZSPIm#_y$W^(^PNNBR%QTR|&F^ryrn^xMd1`n8F
zC^n}mB|NDvdM}x<=Rg0f0OCFV!&mBQ^v@?ZeA93BYpt@kxJ<KFif5GhY&LE`SI-_>
z`&A)B>^h$yU`9qqXITWLU0bNJeZoanUE)hvZY^i^q9tDgiOzfJ$Qi)+`D_l7tH~;R
z(fv-PKe-7q@v#=9a@9E1MAq{!hSJ`;bxQ?s2%Rnf;_}8*mmXD1ycKSg4k#?7H?bE0
zh0GeHb=yKSpZQ1glfN4wVeg8-KD?M=5;J(y)!xQ|M$iG;4u9M!yvDghd^h%#trBZ}
z_wylDfIV5*2*a<TbBX|#_P!9f1i&{I%WwVp$$wmlKTk8CkFB^K7TChVUK%7JTgQ=|
z=d@oHm}?UM_<~4&9*IL{cy|aZuptYvkP<HG4GgP~tV|*s9%>JP5EVg|nGr5WDz=&I
z8l&X*+cz&1V85dA+K)&@F?b3RpZ(yeI|B=@oV++_nEh)xuSwI;eyY_Sj)fvddf|6h
zrKmmR$L~!6nT-T_Yr(*oW$g;yneWfCQb5)C*N%ye3P44@p?0QYC^gZsn%KEa<-C8z
zP$k;?i2o11FUkusg-XOQd*>#YBA}01&VN3D0Mprq_=eztZOQB-5(od;*>e~sw`&rg
z4qO=N>+OXNwEuBF$zUj6zDdJ8m^wo<H9UM<0DVOA((%4g_GG2KP#w8BxI$hIEi+Ki
z!67Xa_KWH0{Xh7xC#JguI~@C`Au&EWyMLlhC(v%w2{U}Rka`S+o5KXZB-UWZX;@B3
zK>hfS;rO))|L?!MNYVNQ1gvo%VeU`Ak|jsP@BEL_nvj#C12dIE@$j4=z4ElL{=YVe
zz}o-rg4K0FgvR=@2vM5=3nx>s1D+B}J5k{E6~wmpo64qPd+<AG=R@#BE9#NgWJ)9A
zmTk5TRlghhP48sMGxR=ofZjpJ{?+`LK%j{b4BXM$0W19`Dhjk3O$DjIhhIqpR_t3!
zO2qpM>$&S`1DJuY;%9t8()Bf*aI<Xy7&!{=N!ayEpmGGU3N~+xo|nx`PVOCD(gtWI
z0c^{9UgPwcGa#aE*|H3ZemB%@0O*UiCQ+A?l&rY&YIdM>w^?5Rs`a@@G25P+H~(T4
zT<|MpgEsLD6EoYS%lLM1e+nQ)5tOa5FlXE+E@lS38ZXKygEn0J@O<+yn1y5v2I3qu
z+wCL%X+ow5jd@)NOhKsHa$aUuE$D*YM7vpUMFf^ecs($?Q%tcA2gN+Ii~?+PHGs6i
z&$7Sp_b1h0a+->7IYC4e2JuiQWRkz0K`qF{UV?;Aj(b%E`*)m5ed1Gh@dN?MJ*Ksp
zeaoR8p08uYwPwv0m25w~e}5D4@!5zZ7Q_&~){_Q%4tp{j2^?ftC7ve2Zy|64jmf?6
zs%a{m72l$Q1Zi~*4G=O4n<CE?g|g~cedY&jv&_=!tv;L`qkJ1sdAn4(J;Lx!4g#@!
zW>!{1jF|OLZf0gPDC#7RVNpaMgkHaT#l8G-51(rb<#7P;v@+0=7Onz(0Eo>~o@0k`
z^gm=AT9?TvLYO4I$<QAf_EI;4!&gAk5MtVBQBN(wYcX8LPt05aJ&Cw`w?Y8okEMd&
zr%7d4iT`?oz<*MHOm5YB%QMhbX<UN!igAS>@xcNpMoJVQ&1mbU_{V0NG|iw+i)0jo
zbB=)>RY)p)O>}GVx@|bno0D#bTz`rVAK+Kg(GfN(KU%EE;qZ?;6GLi2>rNXm*U`R3
z_7!Rz@-d=OYfXRhlQmT&0sCKnfqR4sA*uD>32m-ILVKTS&qzW#>tBFYD<Y6j^Am=g
z_|IG7%!QDqIK}xN(Rqr*a{pq8ch~`h{1ecr@#<mxY4w~?uhO<}tShVOzW4!)O~`iM
z@2?icVdwQk<H=7$ImS2gGlZTX@%6K45)`5L9f@_vzj}DJGbpUF^4_S3V+8rWqEmM<
z_S&QDE%{{zL6JWn@Lx=3gb<TP70cu$ZwWcc6sv!=@`ZFnWTHG0J7`~k2~hrtE6)A#
zuKT9o*8{#GK;R71cK#I2h0@|bo+kqvAhweS*9N~s>KDHXe29~YjS^S~#f(1~lD%{P
zpWgI;{wLJ9P7)D8bG4R(lG*wp-+wfcoG|>+(o<^i&BayLvKtUZzfBUX$-h(_)Xt+G
zA4P7+LhRA~6@uO<_a*Js;$?=s{n+Q<gA4xgf=GPRl~{>%XV@<!pWU~AeL|*3kDiH&
zD>8F@0;%MFK&IB?s&oQ2gd0CmRC~)g>-aKeR@VKU2}lCGi_I6mka+i({P)yzfrEdq
zypR07#*%*n0{KrHI&~b~?qGM?ph`PSk4W3&$A0@A79J5#KvhQUxxjqe2Wt2JIXl*<
zJ<!yDdb5B1r;05n+xhEDXOo5v_AdoiluCgE@BW7Bz*i`2?yt%J^j`n^zW)y%<FxA`
zJA{hE{s)+=#X+v}ePN&!Dlah5l_aAXa8_h^WaOV`RN(-I?JKox-yp&$Xs@q70A-c^
z3sa**LF<qG?<?#Lke+gI5dA5TeENxD8jJ*?akC9*i?^Bb>~B`U=tJiVnnSNc5lN(i
zhakJcss<pKXX#(GUx6A_qW3!h7w*I<5guO%%3U3gqBkCilzH{))#;$n1Pz*)D30t9
zvdUKZoA;Z=GcXuI@dBa?1aV5=d)8j5w_EF<rXahNMl!%uqHCL#EZ5!PBO6HOtLxKX
zm8d}zr6mR_v=*CJ*D9HW4}SR6^MfYo7BDmi8em2ycV~t_O?EfewWn$GU*OWsz$A;$
zblZ-hBmUw~HZ1I|-yW4)4TL`Bn9sp$GdYFoae0eg;*r^z@==1KxR*YWX)UQ)&IYfu
zHIiB+I?Q4<bXe&XbTpfdin*BBOBFLZId~X&cvQ65+@=r!es$_Yy(7D)%-*dt6DS??
ziIlU_Iagp}w&YD%ZI-M-6RiQo!;CHQ6SCH-sT!&LM?m{|Yk`2~484;^#%&9n(79S8
zy053XNqk4_wFLwPLWIzwJHxq(Y-B&RX?T_#t_ichRQI7zULgd?qbQF5qLyveFpVaS
zAwPev)|5Y~ukkeB>4uaVTAp3SC5#f=55gBRH<_H9raTVX9Ve~NbV>D*O3miRU@%M!
znzmVxCrnkxn2`JPozKlNvNiQ60B!JW4dJ&f+K$R>Kb8rD^=hW3riOEsQ&W2~Q{uKX
z(MOLyyIFr*w5E%_T-n!dbC52$`7PEc!~2ehg@r}_@ZXK8KZ@;zzGs<st*vHYY~y9G
zJgH={;>$6UT3E)Xr>74~KQyvcF@K?!Vreb{0E|peltjZ10E0$^B3bO^GHCI$sCn&d
zOk96*(pCNy?924?d@%u{KTYR*^J$)bfE0b)=F|rxGB%A*M)>0|u;0Mgnr<R})z=f}
zLHy|+57q8D5QF(3r{_MGlbxN7;339f;?R2V%L5ujajcCIR3Xiw2X*b*H4s_?fgf<4
z-?$K!vNStrumBYF=SqVZzyI-Lf8=KDkxmf=MI{^Kk<hvwH0^YO0i4eeSn}R)*EqQd
z=c@oDU0}pYzrf*n!W}G9<$^&|D8_liw3>9m4k#@F5@Km<V`X=H!P`FAf)nWtfioyS
zrj=pz81uY&_a`g^#E5GUtl0&BaPxxTv%I_<`N?EcnJEu~*PuBQSqEY4D2Ud~%xHu5
z6AYY2c3Ag(uIuS3eEaqy2L}f|{Ros8XjruitA#~GR#sNPkJ6HnA^RT}GgS(rU$5T0
z@z~v7^KB`x=;D46k~Rj#Dv~~9Gcy=um6IUKcVDw=`3a8?A3h95M2wp|Gd0!kpz6CT
z5K}|RF*3axTK8*HEl-V&vhrP)h9OD?QBCp_C_rtlLSx1EF@U#^qwx0XFGA$@kR%YT
znrXhl_)8*!Ss83Qq+8F4pK9nqr_95ns7uNSJJp-Na*#-Pu`<Tq)UCLmCwrjOyA%`G
zyx9=QLcXMO5Ae*1V^GLj-`KFdWg$$4_K9w(tE=-MDu8a-2{Hj0KI$bMJamv?j?`%5
zIb-xG<LfIyG|;~iasOfS%z_ULB56klop-;PoZn2{OjafkUCr^nP-v5bW<C@>?`!SA
zFr>=7z;%ljq}BY4qt3gBw8PlRiBGEP?iHAa8y&sFb!}^;!TGz)g4*=LLMdS9(J#|X
zApD1~6u@Jz&G*UmjX+@~A}GlCG7}pc&&JIUrsY+<yQ;&`ABvYEz3zpUCAC`w@W8Na
zM)T?z$i{PXb7N#ahl4)R(b<yCd7+jtklAq?eYx9rLj;3?)f5|N>e6}vMwJPX;RCJC
zTg78A7i?#5Uw0U=wyx93Nf<>{{^o|aokgx$dLJiVtFWU!>K20W&4vzpOBhGd;x$M`
zV2xe_T&&TQCt`=&R0hVGVfIC?29Pw4j*fl#_Kc|NyI1bFvz_rqI!kF<IkhvOrNiSq
zuLG-82Ok+9+n>Evd!?ZloCZzMhq2h3uwCk6?zhBC_(FBbCfI@<6zIaQ+Q1Pa;|Be#
z@bR1Yn~HvjTFuvt%r0R7jaqhRbLk>dIj|{7Ye4V|F!)7t?>-Q1;Ku|UCL4FXIK#s_
zpx=(;${0II!5RDZ?F|e9ocjSas%$%0XrpsthF>2+rTX!sM^O3jIcTDc$ONgS)P4Il
zxu$;0TJVhVMrvv*jPpECx(w$8@DHOJU0UT`l+YG{V?6+qrXc>cLT4l8NAdF2s|(TI
zKE&}Gc=@-}<1ld)f5H<4`#xI_)Hi)vs)b_S4}Hk`Ubt=pW*T<k3S9ltPfO}r2(a&l
z1-P!;0N5F)uc_HJ-IgMt$oCpLcF$|f0fUUdb^GHt;M^mqr5p7zVlhg>f9%PBH>W;Y
zZvvJ|$@>yLJu<Qc7@7QVYrtR=v5OvAgM9=O3Zkyd7o|c#QA$<X&@fId#q&pl-&qks
zzzN`Rh9D)xzf}QneW-6scND>_I$ZcV%&H4~K+Iwww!goBoFjlvY<_GkBs|<qZw2}s
zz9Gv2p1ay;J!l`oy#@=IlGAre44ITjq&AJTA1;3mcGUg_2uZ0#U;z_UWh59S&&H-(
zkPV&^Ix5%r{UwzZ6>XrduPg`3KvD4Dn>Q!DB|U1it&7KohR(C#*CNy+=1W))dBET?
z^rp?X!e?hKrgUA6jg6sQa@Kv83vSF(DXUuQsj}w3|Na|<lNc^tdd@qX1Y2VWIyY|+
zfu@kbSOZo;!R}QhTaJ?T;nAufP_oK{m4f+_CyySz7Q5{MSr~N2pwW7Nr#4Y85;;Af
zV^tS^r82wuQ3n)tdwYAM-RJYcDx~R`>Uo55Hf>RA0A?9<v{Fe9Aj*+&V%%Xq4^|)A
zyp|cXaYG<vS{?>z+t4lIYe)01qkS@ZlEqd4b+HDc9vXb0@#i|0X;M?}eeA<6{r=jb
zNtnzpw^@EnuE+Ti*`gGDKH@Q5;`;jfQ#J}1CcaD<Bw3;2I!j`Fy&h4r5*jrilm^)>
zKy`q(;cwYJ-U>Vg=PDfFJk2z}^JU^Jjaa=)<AHR&BA`9_@#PJ_3(z7kbNZB*r%^yW
z6bn*@@bkL+)z#G?=Y~;++dUtMjgF3Hhy$fL+o7+ASG*B;pAk>Td!+<|gGV6gl~H7(
zP44r<CgP&R$+*g|2w?y@cLJrWtIG~&lj`>;i^b2}xkXIR?JHJm+4*T`K@ek7O_DJV
zJ>$$DEsvo}u=v-WG!-ujaB!E7YL|HAyUJ;5hPUFN6Uq<T*+1C<ewEDB2}KFly3(ER
z92pf`9aj;psN1Rt#w0#0OdWb#09!-kHtVBBdMYcWrehoHy1HBeMi>c*;C%Ze4SN_&
z&qxHDWi*=SIHT4=M+wle3nY<(j?>_U$|IoUX9Z<`qtRzrxOj^$oY;=XE<ofti@g{G
z1%F04_rDZmvYGA9N%{Ekqeggiw3QZf9ss3JQ~87d0wt%a1zcNUetx*V&lz%ZXq5~8
zmb0SI1MHXN$u;o$YA@$>@@&K{q)wiOmPv&YD8v1C$N4SQ_%4B(Db@LZN7iwL>mRa?
zKNOi<T$ukAMP{*<EcDM1GNxHN&g+4H(Oo&)a8u?@cyIU5_Nin`J2Q0MJMYp#d+GV>
zuZrgk=}p2;43<%U8eO$~)iUvPWv1*YxqlGpi!UFhzRw)my1HC?ZX+?S>cG(iL$__w
z?VlIlaeG8<Tw~4wIs@m*nANTKxTRyDW7QBqHpCpI%;6EW?st4*ee+UjC1W_SF%5jL
zc)Ids^}!T@=<bi7W8!zkb4A69MYpyUZ>Z^=xX4eL^5{l{*Y3{ZjJ4fQES*>Q+s1ym
z-SUKiGTT$pra$lh92-g!5n)=O6ys7d|9+JX!_AjDbuBb}lG4JZVlsrm!zWkeZ;6w%
z`ppsy*JdPg&+(X!ANU&Ab%yrhafi@S#xC{6GrqG!6kA&6L#Z)-P1iKlw+Bz#Fw72a
zWcq1R(VpG#Q?6sFmila)^=-0EByY`PJXo=wQ<b^s>jnEC$&NANtl6JYQ+`CRlvuau
za=w@3xt=5`aUi*WzR2kzHooMR(AWE?yOXhx^%p;W?eUm9`Z3j@=(|Z4ebKOjWA-@t
zmQdY{=g@;kZz=Qph@&^+L}XT)PgF5Gk17<;kPP(|EYf^D=jhpGc|ze$x5g9fW4d)^
zCt3%tV&j8<s}TgE2sF+RuG9WiF9~#B2BQjHp63{HkH)T-3}heB<oME{g_-;N=wpcp
ztoX?-W|K*y!_~Hb@iJfji-I*cI>)A;;7bVZ6wOJ4K?S9@DAG@PO`l1ht=qSMH9;M+
zjT{O1{PjyrypZ~-*1tJrPlfV+QGKI$BJHl6L7n!@)ok)!H`<eT0w2UnIeh=iJXHU$
zMxBv^eMi2?Ut*B&Yp1n*E`G$PzcIh%-P<-x*}wRi-*9N36H*Pw_zQVvUphi*#9JtE
z`S_{%xu#XJuH;4%?W=+m-_BvHB+?1&s;|y6oLas3S`l?8NsB(o$=fEMLV`12XY|&6
zyKs#uzT-!d4l7yvMaBES=_(fJq)0fK$U&Z_-sO0zw=!sfFhe=731iiOuHbxTV2A%k
zVSyE2Xb4RmSmgEa+ns3)CJrxD$5~MiK0;C05ju|;JnFdnb7^8Vg!rUN4?!=%$g2v@
zZl}LcRR%AsGV0A4Wl#EkO|7!e8YMU&7a_8A`bqjD*>lS8cemF02%Wulgs)#EsV5S4
zakKt@C1x!8fYgIgfh%K5;>Rg9m}G7h3e~Qb`G%$3l$PpQdo*z@E9kDWZpA0!R;Iq1
zm4ufcm@X>}DLvq0X78oX)3%inG|azWC~bdSg>%zDnyTXf+x#UR&DXr%1Urw$?c3<`
z*-l*J{Yl@TNfx_q9LMQi>^n>2t#50~Y#cau!@^K`eC7Tl-HY#whq~=dju_5m1ihGf
z(KvN;buB1QiCgL`w(!;4PEnrbK0@<BgT`CEC-CzvmwH`i+K#qs=3Vp3vgqK<iVpS;
z<DI``8+q_Zrdla}S5m6fa3Wctg)-s_@%T~Q*^TkT>bEY5P7K@pc=akko>5^)x+}2a
zjlY?~>-k__Lzl?s4CbzdWYNi)*Rq*i-=*p&qjn}<Twn1Do(x_lynAuUHP+G0L?)Rt
zNv5Z$Nyd1Z`ZcdX3Jsyp<-VVuW@VRiw1aPla(4)<=Y<k~>@YFG$oNsOoLf7lR#rHZ
ziDqS(>-p;UF8#gZnL88tg*SstoLlKD)3`13Y=@sXU}z%_3wA7LUM=hMNj<2ydZ^FE
zX{jpU*rW1WiuGQa!rs0QE)pCs5FO96({T%W;QML@u<Eq(Jq*!Y?3uDBZxH4c6qSBd
z4Lp#?MVn{&{kUPB+#*@mLy6TSz0KJ}FM{4IQxqn&Q>yJgT;iLX=)XUh`Oee#s|JMv
z%Jp{Pk(9osZ8rto4$EuIhlOAKe5~jtAHlU$An=M4CEbyYcM7{>?vdcq7Zr6`L-W#2
z_TXotsKb4&M_9?-O4;4{f6o0?q`e+!@kZ~PS9sr)PcKz?fc|Y80p9|KL&?VuJM(5S
zM_aZP1){_6&s@~LT1V*C=s}h8^RMJ8jzXrW{!^UHQbaV@r_?{u@oZBsyu@ifaC_jE
z@&O^dV&7RfuFOi(c+<79OqT=*uR_1AN{+e@%^m1U3mOS|FsGH0?4a%JIOTuYGKuuy
zIZD^GmalIjF57>6&8jLI<#gNiUCZ~#k*$q!d0#PH{;ui4_`fIaj%Q{Z5GV*?%ILE1
zoBeDzMSVwyc78yp^PsIA|Br9DvovsSw{(a?EjilU8LA4Udzmy}A6MMvw-=#^o^dQa
zskfD}HSMzNMEp_wo}=y6;!38qS9wi@;i;E?#*DiU23rFUyF=ASK81Jfx%TE3ykB)Y
z`q?Q3qe3=%-}NW+)I?$0E+=;g*<ST9V;Jmoa-vzjsmYKTe8~N0of9r3F1s1yxb-QG
zo>ZUFIecZr;~w=fdG>k@%V}DgmJBxS$8Hy_xs2#@-{v*e5+3t1G_+!w-KvRP|LpBm
zz8;w!dc8zB+uf_3yQAUUSe`HC5mzs5y*~XL8+QE~JGYr#F0ha3*0zOLS1d$*vsoRy
zDG@y(qWo2JO)QC9rMB>#Fr#%fqgak-<;{f8t`_LIYq76v8Pc}|3@eule!gYSiojF~
zP?(?#QV2!wSc^Irt<*DX%QrABt68MoO1LKesWGc)bb!SzYUOHnxOL*tG^@Xy+~rA{
zDATqNQtSzX8^qC{w?mRV@0-(Uc%D>~jP6>YQ|h-+)f5)9U>KS84!4)ERK7fQaWab~
zLjGmV(`YI*OV7Eb0pCD}cne9ciMO)UmsAT_{heMV3@ftlXj1bA#Za?G%T!oxeAjMI
zJ$7Y5L+xU3kMEFNc-8%(tGTiPeQk`XJISvKpLi2g$hCj?w=v1Tq3!1XUlK&7X-iv=
z=8-~7GHFWB`E=C1_f*o~AI#Mc+DE>4S#w7|{8pp?>UsOBfrgQin?_xfw5=?%IwiRr
zDl239C-g<dbo$SF<q=$dqIt<rr{7EaQ1RgE$H7&_6#P*f&$F?)&E=I76SrF&<4y>m
zVrv?$Wt|<}ju@Ifip{<DCYsPE3?tt^DcC|WRoS#~KSGM$>=-&?TySp7MSSLjmcsiM
z4r?*d#X}AYJ3oIc#ymZ#DQNJ1Hs1dZP1|z70+HD6*D(F5v49iz24|NCI9ChlF27bb
z-&kAjcp7upBe0%$+lNe*osNV+Dqb$!<-DQV^hG~!sa9FV_0wmi6+$mGFOhM)J7^~2
zKu>WrR`*cW`+JRtrJl4mE<E$ZdPxLEVV-ymeo3|UZU6E%LxLzhy8dRNjRR$T`2EHI
zL)kk9i4Jv(wr$(CZQHiB+qP}&wr$(CZQHipz5Dg4^Xk>DI$za2FMqOvN+tP`%r)1V
zW7O&tD663r6Mi})8#_M`JI@-Mk|wGxR%CI8V>b4|0;HRnWB7CaHDgh*Xs`ca#yD-r
z02j`O2C71JuZxN*kWoXv%fry{V0c^p2ihgM3ev?Abw-=`4ANE9$&>L|`OQ5uh|{GI
zRVme;?qB*nhco?d;OFHjhR7ogzz-<bLdC&?>=r=nXw~$i$Tz?!S8urW=b6i{bT1iM
zM&E&9K(<dC8N-F*&A~6(8PDdgw7a@Zl7mc})VGI_KRlTZzD!?TuaxI50>dkZ6gFAD
z;~o>Ce?frSfj&^^w<-)ZK7jybhdP`NIS7T6)z&-`#*6yAyEc&PRtez*r6Dy4Sc4<L
z<d_XnAEQ(w9GPKXULb9cI8-`BUfRv7?cX#JJ<*ztwL(%xeFZ^mE}|JHA`nuNBYf>&
zVZBEDWQ%<h2^b8a<Fm*aCt}iI;SDKc_z+c*=Y>8I7!22Nut@NRTV*(ebHqFmE@I-}
zsH2hX{(OVryiez`>3eN`sb`@y5K|KRo2Pyu48}7)ma@(h|D;mtRcHgGJVkA(Ubqqi
z|Gd?C!D)*-6U2y^P{}B<VoGRkEd*e&ZE=W1eO00ts9k(pka^Eg^EIkXMN2a0eYC$S
z1$y4RXlzfa<+c&(6v-CbMy0F*n4NiFC>HcaOpNF|saGZ}5?G{R^ZA&0hVwG!&NcCp
z!EcmA15q=n89gOU9wZ$s>M(`IG(CVm4LCVx4IT7LCcfgEL$5^JkGr`WJ_8-8_mXSF
zLmOyzCZh2S_aeg?Z0aMnuAqjSA`bMJ9HrTelhSS|>F{~b%ytf05)$`Wr?rToy)zcK
zV!}mFXI-O@(V2C#?aWOy76fh<Au|ah!BqA`h3L2+G6v>&pcWHj!wZIzTfdQYyI_ie
z+-TT?@o-q1?}}^A3fev9d`o$!<YK^s%jHhx`L}t(F{BYah2$JdhZ~tOfjAXFqDY`>
z1}yF{Y&Co%QMcrP<J$6MbZWe2CcuL&(fL8MJ_VD664h8>%}_e^kzRymi)Qb<+O#mX
zOSX=uC8d~V(lfKF%P<Xls`nu6BYB|!jZx$Cb9H-{>&lf?B)_?Ge-5j{oy~X2!KmQ;
zZC+)|2hvis_m$LcWy-}bvV1$vtZ;HWt7^HHT!B6uKXoJ%CFAg|Xp_)TL5PwRH3u6p
zXzUiLIqN7fbS~OUl#LwmivJAdW#Np4>y3Oa8f+~X*VB602)q{jTs8*8gG?S}N_&~W
zc(Wj6>b~kUIk)8=ah$Irmm&LyI_Gfftj@+gt}S{vtX#xuFn*5~qdV^+MG2y~d}m+9
z&$0bWj?egWI`u-6Bw|ky9O}5b{u)+B&zCS`hwVGLNF`C5y+$hKJ?vh~SZ)k;XX?@E
z&`K$l<H63BlJd{rd4bjy)-kBW9*z3+D9wV+{*0R#ZLf_`oe-XqOxisn>7?xg)@6q&
z1=a59OZ`mb=@hu|QCIcGd)*9N@#YNN+$N*h=pD8lZQaMS1AOJj-YJ!6^vM?lvohBn
z7TKe2)1Up_jw@!zr!n8a)kD-I+QpL_0jz15ZMW?J_f+mUn>WcjtpY18WU-Yc3f^)!
zot`BDJe>V%Nil0vi%6REetdTEX_90=>KFv!XRj|0fG0vfJhMK@v!vUJRNda522|U0
zC&OTD#N+1UQi?YF-!9C$)o$i))BFl*G!(g7K{Z&0XW5(AB4pVlWJ7}m3A}FJ4w5jx
z=(0Fm;rZE!cbzSMjs-t3d%{suj3lKJ(?S-8A$6X%)XU^e0Bwu>`34K=br~^o(0s<q
z^$L<`nBlTjEZn;PQ=-Xu*Mt&Jrpoj#GKzUF+1@HtJhFQd>2Hy)>3K@>O4JBEbH_Ab
z<~v&>k7?J@{&34tG6_D5xkZ&Cu#E<5!-X&Uy6d_w(Lq{f<iU7jZq<8%xqC*n3cKc4
zv)Fjm_VtzqdotsAe@KM3?TVoXzxFkVtiuAetzyb~(N#>m)sfv1c9Qz-DF~)wxB_ME
z8SO3`tIXcmW%&^_Y2l;gHzlWLX*csj>+Ikk-9(B+;wXBAdG_!x=qDR9YINr#82$I}
z+(|ol{^-gRIk>~EPum~(Z>X3quLn|}+<gvF&YI@cda?;fIm@ufqF^9PJ4gC=8rtkQ
zq-u3w*}mbnjXh)H$eY2Q$aLh)3q@^gWG5Io+ZQ_9qy5V$1JR%6aQPS?LBc=qgS0rq
z8kcZGyj2htcG5#aPaI!Kcf>f5z1j$QjH1e$;h7hHwE>iUjJopn{_OJ?Oe_vl7(Y@C
z=W=72DD0^hq}ETQuU-vOgrJ2(Mdny)=-L~8M@2<J(Qnx~Ih%||k<wi63nT|MTyH!(
zKLgdK>(y+08PBt_Rbg`vk<Lhr|H=ZGQvYpY{$$6Voq2zKp_F1Y@(wc^+3+9Vf>2K{
zI69>Omq@;BSM~DpQa`_)mYU^HA6`1E|B(zBUamQ?AQOXvlsQ+7(50wuNb3hL5)c~$
zdt2;ICOVBcJZUGuR4JiSj5$&bQ2L?$KF+}DVg=o-nWQHTWWQK^A$FoUT9sop3S}??
zO&}@8(zxd`3E{VV!gu=qsalhw195|im^|a&Am!&|C)IXG=1gRY*{HYkdP>`J*ps_i
z>^##Kf~Rvrt8+eb9C4uh_B3aj>rM=4rzw3grWIag+`qS~Bl&*2wlIAtN{o8joK8~S
z1QlfNxX<4A&9(#ow-{;LlB50MIq>J?PJwCAE3jvqQ8pqUB^L}<E4%JV%Q>Tf%p>mE
znc#*?6ryz1lWOStdd?58e5)-xEduBp-E78J;6VbcvJ2kIbV?RiP$>ci*y?kT)4H&%
zs>WhjzdB5L^#x$^2-yE)hYhxRvS!piRPdUiiLZf^zT?H`)%ya2YWZ@BEabNkv&b}a
z^bfP2!90p!-%EP38Vu#v2XT|L#>bnBZO;IOV=peP=wC2)GmWCT$n7Y4kI?O4QBR8?
zS4FJU51d%Pj0s(i2XQbkwA-W*!dRN{qlwBF);;i#_xjYirOB2`Uv{L{Z9Enj#{63&
zn;e-FUkT;Q)2vK(GnI;ucd+Z1ck=~N-%N7y4W>(($rWx-a4^i$iDitF1#)-Q+ZP<1
zxGU$@aopZ3hNMw^|6_jT<&vQ1XrbwDj6?0eNcEfJGq{-mrcsxsUG&}sPoosuL)?5Q
zV@%e)%REjZOEQo+QzZCHRX_4WKLEjLG%Nop=4D~~pLsYo2B!a29<EqR+Kz-B?l<Or
zn&;CXR$Q_gWz>I-0zPW_TdzSdZV?26cZKuP^!b?TiB?>pq)Hst(pf2TPnhr_A*uJT
zYiu&st@hR}cZWqU>rRZbt?S1@@kwam59xV6D|%^CotExS?;l@}1SA?Cz4IU?OIdvJ
zUEXi@*;p+Pto*ZTrP)}xWo0~YyC;u)MugG%SF8NOc$Sc*6!RchhpvM6<k+8FKd+y2
zIW$4>@+x6LCzf(^F=3wOa3_omYzy0`dq^Hp+gT_ZZ_%rG5hlW(68P}Y@U(xuL;Fl{
ztsW|#(uFYJY#rY0$LuO8;Q_vCbkjdD_io;iq0<7E6RjLvL#@AHOcMBc(fLPZ&E&o*
z=w3c6Zz`ltwu9oaoTceT%5Nx9Y<A0(lhFxQB~si3xoKx?C)4vyoz(~K_0nnWXPGUv
zCl|p2zM6ZeUy$k@ON-_(ePxohuPj`>BM7vRXIadgVw>@77oB3tlZ}UxbDhbN_JN{U
zx}~W03Q4<b!k=VT33)Vi>YcdUmSq|(6YNvB3iI^$B2|sqm7KE!>ERF4eVkmdf7I_U
zted?t@_eha(cUSRKRh$4@RGNAF?5x=G)HMRHOoXcE#{L}^&8ryqi5>x=To(4T(eJ}
z(@qM3>rY@J`3x#{e+}%E`$UmHCz#S_mO&7lVsWB#bI(jDUxBcSj{&pKQiie4BB@zQ
zAM&wy7BNlXtZn~{kgkfgOsz?aEQlm}dp6ob5^2<pJ+_0<N3CRs-~7BPNSw7p$ayWs
zo1*~4NC;>cF^!M5N*S@O=OOB=&Vu2M$$k8b6*}J{D^4aF3sB;dacK~9|DFEtF*7Wg
z4^C!T6Hgos^hDxk8vUiigC88WLw(E7$InA$b$e<`(u%aPJ|D$wTZUs34&3U8l5ow7
z*AHPaS~hNb-Qp;Ay5kZis4c#<-O$d)X*Us+&wC$22R>z89QE*I&(L3s;;u}M0kH32
z`+)1a5jXxEk(-fGXMzVS1X{&5<-w{&lO(bLv>|#amC`>+pW>TB?j5FAw&0s>rgkhj
z6CS%l`gO2gqxJF=FO6Am&n-dkk}9y4^1?GD6}Ssj4;3W*SM}5uZryy5V>gPdQTyc+
zAV}Q<ZA!XLD`z?yya-@h2tqjhyV^Rlucgm2tZO?3q5M-cVZV4IKl*^}iL@*(R^wWg
z-kQ{SlYze*li$e*S;1U`ZO-zLUbF!!rN+f4oS&=?A%}5&g}$#mdg{`hEDDBnQ=n62
zvoAeU^Qe+99-`WSFwIXjPu153F!t>SI=8>7?Y4i5sR!%{?wKlYk>h<LE?_s_jaHHp
zcA<1H{-QUe#J@h+odmd$j6b&Dn#5Y#GTp@hu-p0s$szK@Vrr94fn5!G;Eq;kZK`?6
zqHSC*K1zsI#(q2M6(2r!Z|ht6>r}gB*MTjP$elZaB)t1$0TO-&s#3yTz@Wd0s<2%i
zb;%AAsviRT9U&&cMELy_#X%HGcr#Iv07qpSDrCFlF3|Ec2I!<r&T{`CkRbvJx#f9|
z3W9bU*3-P$9R0?_JM$tR)q+m}=S+~2odV~{z3JlL<LF{5Cj+s-rRrJA^S@V4_NPF!
zY>6g?V%&kcJ2q_LJ1CXe697kK2-op_TYaN!dr$qxR^>9#Ze)~x+HEG5e#=(X<1<T7
z+zSak@=_K3&Wo0%238x=Bm=c%!#{}3Sq<vMxtrGC<I{!a=q1{it6Heh7bnr2*|F(E
zL32$K5@$?h+RWi8QhCA9SQ|Ot<q#xL1UwO1(B(76NVJfzRA&@1Op*v^s^Shwl05ec
zt!IL&(e>Rb;OQU|PF{(40-AbumG!s|DpVj+f6bSaFz}DzMcbKc@=zvnV5}!NBam!I
zPYC=lXuFm&t$TWQzfWz8ft0}Vu<ris-uk}F>iL$Y^6Qp~!2NsJWqJApEf@vz`F=e<
z?d(0xxtDX09qx6Tf9hfVEi+cb{d7Zt=b_Vj$S_hRd39QIDZwtNgI+n03o$w~%)xR|
z2z=5cp3m3R&M?iCADlCsmbY?>B5aE|*EFiLy50Y+U0qsg{zX3XSd6q3Rr)Z_l3XI~
z*bYWRnFdre|0##ES7^@^{S{;yOSs9r+4h1?U{_%OZJo;z=<@SHZug<Buz5UVo25Q%
zv&QfW0MbLcC~gA<oX8#r-Kz$O6%k4U(wE@eM1<*wl5FoZdK}~lEJh`(=cG+18xv#b
zPQBV3Ed1piH*ZG*lcQ|ptO_E>glg4#COFA*$hI>^6@i~maZF?bL*^?P$q_NpO6sh_
zwLF;;u7IlO^X6Y^_;RlIx!y2xVy1WC<_D%twCmW!x2OUt4I?n%s~8aFatAta0;O)v
zvBRGtp17})$JPs%h6S;;PVj#-LdJRwpRd}xA8LSyj%2RuCS5vmDZB((Os*~-wE5bc
z6w@!%fNjKBHC~}z;$$@)Zj)qhS!(5@n}zlre2%vND1Vy?pGoAnD?NEh#>YX~d=FUh
zs2U)HBEf;nt}P#63P^G$S=Q$O!KKa3mBEmw3rj(uuoOWG^7_z)DPh9HL1u#v=$xpy
zu0|l-^c{z7aRhjfEu|<FTtdJv(Cu&g&ne;@z5I2BoN<UaG?uPQL*7cT6eK)8L+$IN
z%PX60x{jkS4dxrtKAtj=hI|35`yiRu2SbT7a<kN@knE=;AHO*V(I+@9>{jB;MtH#-
zLV%$KRHQJV(tAtuQrW^&c$JBe9v`gMQ5IZM#IrjF1dE2tiXc*mCb8FW!L;#A(Eg%&
zA)imL#%oD2RYk88b_oA9*yW<JL~MA+hi2QmJxlO6(%iha`1pq%Gf?6`13aWBqWr-)
z8pA5?mOfYiMds?Fx3PU_EsoWfW*>9=1t*cUou|l^a5x2%XnSqfeNyIRa?lcRTN8Fv
zA4r^|i5K>$rCOkUDd*N8NcesXP-26XuN5nnG{t+pc`7z_$`p2ZTw12zdk5;mIJ8l#
zy9EqkhBk7vSfp|>MQfkI^p-5G@0VPXd+bDCl2=#E5OX|iB@l3>hLE3dP*YdeD-2pt
zgDVK86?bcW`9)-*B;z1(O%NY}h;d}t=(i}>QDfFV_S=rC+qrr_SD3HZR<giXStt;T
z?dq^tjlQaPZ~S?TqBoI@qXp*|>6XmYB3KICdW+7fJz-CGEX<QMdXwSdr)hAu$qqRr
zc~e4qY3mp-;!g6thPe(SLUci_&6>nDa20~BPI-F=Xwcs8cY;bOBiKdtFk*+CaH)I%
z`}K%Gv3j(<sWC<-*LQ*<H>NQe4c2T06MHkOOr?)%Fb<wP(}$kn5&jw9Nb!!L3yeeY
z(>C1+U>JdZ>L`3mF5xh*^h7KJ+82536xiSt^!l#Lh+KB>LaEF)Pfu>{A^d<I4nGV1
zw}HpDSzE)vt*3ONvhWPUE?n1H^tYM09(?Q?XzVik_~YmDbNdId-$Lc-KgFG_|A(yM
z{~iBjW@h?trvkzMMXJ!OJsz_)g4uIPnU)D6QoNBWg5iG@1@nd`$pFN`M_O+{%r%sH
zbpN<IQ5$iE_wsZ!2R;7r+8m+k{b6#j=w_EvYscrU<motty?^xdd@4ByEVmoDC~HMV
z4$>$k^QTvq-}gvI>x1TLK8souU;KynJ0t8>${maFtQ-OUHX9B9k>fp_k2yNt0q6S<
zG)^=LijHO=EZfk1;2rb;`PQKPpI*N~=t6uxO@2_cswS-VBryhyVdHA!t;wO|&EB5K
zJdgsi$w1|uC<3{Sam$%33W8oW1TUC^^!XC-_*n%)+mIW<9+fyF%`8*3&`jg{7|MYi
z#rNwD8aW2OKpx`AMi~#r3B%og@2j!5Lt^rHQm8o4{jw_(rXjO^KiyLD4#Q*E5F!kO
zs`U!QMloX)`Dv~l<d5u*y>7R~PP2DqMJ6+EcyE&~o)Kpj^Fd!?Lq#MFS?KzwZ~@I`
ziYIUyI;Cdg^TT}wy>!XZO68eFMAFhVVzg8q*>U7T4d5U9+g?UH|H}Jx<r!0~(Pg9=
zrFbF4dq;RAF#RZP%%!k17XFqj3HY^_G#OHam`jmvoLoL-@rB;p%O@#Ch0i$!7f|Cu
z&7_$k<@_Bq!%0@dJO9nkfRa)cGNFs1NB}NzPoQ?(Q*?AHCd|2qW26MmyEa{m(es59
zM^iSfMEBIAsu!5x*KDvsm<C@oee39eOc&h}`nFvNWJs5^@sc@SfpU2UU}Wb-q^kgm
zO$ORZMJA6x;BJrbcF7l3NM%YEE<_3eF2+Z`c{ZyR76iJmXw*n^qn`veWL0nBXQ*vi
zZ(R8=-&X{MkC+q_Y-}IoVX+uXxr<PN&?{+@HJfBAtUOMEh^N01Fu^91;2Kl^vwFG0
zfY)U)lo2pCUBRC@0x+UE%mhllgr&5!CQCMU80nO>aVpxlnKHT3LcN&ZnkL{>+Qj`+
zM7_YKUX{SQR-~8Ps9Zqz=%VCOn7V(N>^)_qbXXXHed17}FVRr=t~_%#t@pUIz$@5l
z!{Oyi>|DU)#^lDsMkRT`Rsy0p>53{1GdNnKYqgRb0f|gTJU-F&F+VVUZ%&#{{CXyC
zwxjbL(myz0N)G|o>5w6&0zH!n9!+R;;u?l-uI|uEexpF2R$*M<v+>W)chMsst3uH~
zt=yte@)#<6Vo@z1Sj%{SG5c}y$Vt5#)hK~0HF4!aY?TE%K4bC3JY<M`EUOEO#YHW{
z6m(lx)jIYJ{WQ@om$3GT_L}m3ldO9PrcB<1&jyuQ?uz-B40|4%8H>&Q#Iu~cP^K%l
zb+aKG#;Q$zZF*H>JjK&<5*Y_oxZZ)G%R1Xn&n(kSYw1k;@(?7-EHcfOJ|TC&k<Zc`
z<#WN?dwc<7CC$2!DL1q@=v7UkUvrP*RE=gS?qmf7Rh{r7bC^c&#hGaerMttT<oKdk
zm}enpBwFdDTELfTW`oMdGs5rgz91g+Pwizp#Jnw~y)oD6;Q(Xr{<L%d;`}nx?-oQ+
zNx6SzaksZS9KH#Or8qtjArvzrZ1;!r`|E_Bcq2iG;}B#}=BvQ3_lF7AjW4x0ypPhc
zt+^x<AF7`+>#4{juU~r-Bcj-fZpV#q(-FueOFYd*U+9)`R(oYzHNVX5z02(ki~B>+
z`wEy;j<FzUcz@obJiQCe+wsyra=sDJMMdvhitGBudq=4iDS(nvjFI#<$r*V>qha~4
z!r9<NzKYJ?nnn6SX=K{p1f>$RocOuwo-{R4ThvUNNpB29IHlHpF-Fuc4lP`*V^_5-
zGr<hRPJ|#(>#}^oAmovonbTW3@bFT@8`lLYxZxooW<?$C_UdHonykz_W<Zi=*~h#9
z%ZHoAf8Oh;6tDKR88nmEp6VWyR(;N4r|P8_h)_r>Wp5swhSJ%fZp5a(`HQX~@-8MA
zyj0y-MB)lHe@VChMp~-(GZDHde|{J?!>8S-fBN5{F7xU8id_$$ttxi1%P16Ai;i{c
zT=SMlW7d{6;aBBI5zT3XPe}JGN&Dt_mY|dL>ubze7E#2kC+aX?<hXvR;^7g)%ci&9
z5f&T7M6pxpbghMxsJ85V$uT`TMolnZZ~u+w$Gxm5+IMhXTJ}txE=rt|@K`Zr;Xbgm
zYkaoNEN_}C)0;Pq&^C?o8Zp&|7UeBlJI|MV@DB1Sw_iQ6PO{b>;8Fgo3ggrKDDVyS
zHO!K~jrUkH$yfLt`Gqb1Sd`89G&wEp!8@%D7B%X@+ry1$m<prZBG=AcO~b!aD|1^m
ziSlD+q7=nwFhy;tJZd~OC~0GPliqwGkx2{MtrJ$Q$2CpTstZdu32)PWcErNKm8(cR
zqG8!GQN1#QzV+>xWKrD^A$e^dPVkC&>RzkCB0FZX=IrC;5fe_KRXyq~zo<sLY*a*M
zd9~)cKrxfFi7n(Bia>9jD0R3v?{2EK2JgEK8rXEVlmZY68mMLvgQf=Z4fV#W)1EIA
zcan2B0c#<cp`dO{6)`=-H~;wL7pok)Wsfj}?{~1sf42Y1!eM#AZ*kTt7AF6b(KA$>
zZ-%f4!bGSxrOfRM%9u+6ZQL8>Dr}&V$7$Y%JIpx~9F6<8Ek-SFF4P?UufykbF$0=0
ztSBxDDv?Pi+zq8GsY#`b4il8gg3<or1Y3pzUrCh1v8OB?8hxvPr9WypJ#$4nz~uwa
zznVL2fc8Os84Q<xa{y(4G0-)*JMS4B05(;IX;7`?Kg{~&vpo7)Ze8si>g8Xd1Js7f
z0NrR&me}fgMb_`r%;#m|B}}c|h>{SMNJ|9n&C<AslRog;`72WZb$XhJL5u@=!Ju_h
znE~K3R{c$Z7dUNwN_>}j6rId*m~TS2i$PkncYK47ynfJaB)&9X;{QzJO!!#hmMxfN
z+2RsK<mR3ph{tFJ67j5$`p8V{<{KIl!ESgpbb|2I<^lk#Hy!O_72#9<K$TE=0l;ha
zAw8na1cjQ-)qzq?`sXWwl+y<TBulP0tjv(Jm~X%k-a`4xpzDU%Z6FeKa>ijYcJhhB
zSo8%&18JxWBEj?}g5$$=LZlJq(A*Up1W@@8Gs?baH=D&R`R}mY?GZA9|5b8fl|9$#
zo0EY1TP@%W_^K(0i6`P}!qn?BCcih5_EiT3lE=_7;+TenCZS2ugkOOKwg@xp^%y2-
z$r_vsOp5^#I}y@ZwzGEIx6l(bmfi1)A>V@D@rc&=2m>y4M_|ob1W42JT2J+o2LW8X
z0qcRK3rK9W&f~?f0JBy{?+`TfIHxw%S@zJE8KkID48)<Z=6K31matUc-_Q-%Y<Y(X
zdZFES%373D7tTgu2)KgIFao8>l5`g8A=fMkYeFTEzK~~VP0T}50lF~Fmw!Q6z*A<A
znNSAYoFQ8ReQ_LMp^?)nFb4NL^t>WYUmZ;ML_4>`s>TSngka;f`>FuEv{(IWj~Sp~
z*udLlx@&>K#qzAd^vvK(FlAtUahg;}M4KzW<}NZi{JWPP+KdekX$iWr6(`PEO(U#x
zg_;k(3y;9~sDSb<7zlwr-J>I6!(8SA2HvStc+n-)oQ}|f0Y|iWa|+N|w5D|?L=1Kl
z|IuAoo`0>UYk&{r(Mt*1DGzSAs=t)IT9$uxFvfUM7hCn33&&mKwOp_6HeoG#Z)>0K
zV$P)27KT|UcIQ}TX_h|!9XusK2^U8q24p&x#lh$W1g#F3v(dZCdiEW${@C&Q3E;MB
z&~XvQQVb@=CFxh>Yn682Q~qPSd)0B>3G0!U*11&o`glXT8YZtL6AFqhp9MqBw-Idn
zymyU1x#CeBD@YQshIa&iE2kJhd8bJQ;g~SN2FoK2LLj%`Z!x7w^t%Jf`O&PLU`&!~
zE+(fqkbt~gK;1P3Dln@s@OoI13itPbg6OQ0U|ga~o`2SVpOC{N4Jyc#T)2FB!z-YR
z-fIT<`7zN3d9HGRYxTlj$nly%RDdefeZ)Ha#&<L{uSrOIcqc~5LlMM%#EI`lRN670
zi(ql+&bPCaD?@QAzED(`ZAD09@~5u6Hj^r)1>BF9qrU|WDi(E?(J$<jxUDbIGj@}a
z<nV!zHrtiR?bLIK^oAbi({c5E!IY!YXoBwfK@TBDEP2(wj=N0?fo-h>o5#x>yQihQ
z^8Dokk&%$)Q12@CFJjk`dQ)q*Q#kSPBX>4*Pt2hAY0D$L33_*x=M=|?&#8NbaD-mN
zuj&N5MmLGVBF)jOn?_w;cBp&sKJFX>)J&+gwIrMD%EdOzIB?<FUwz%tGip$ho##-$
z*0B;&n`fLyUsf;Ig|!g;%oC0HGpOvef5Uv<=CkGyE9tDe&Jb6wp#>S06^QPQnmpRt
z>a7)5hlhjKMfTHqylG>erLGrY=dM)Os9Bvn<)HJ!(L|q}z<Orwr9-zNiDd~zg4L11
zc8so9Mg9jo;)EjNKX#}7!_*ZHHunEj8h=<vI`OClwr93>Zk`V&buUvju_j}*rP4c+
znLSlrvMX8T1>sLZE`A&cDPls&^Tl|6s}~6rNIyJ*p4Ph0rSa?H;N<6EY9(7XdViq$
z_UA3`9&Gqe;85g@E1m!c?DpsD%_=!%^587AZ&Z@j;ppM5|L)5}8Q~tpuXa9F+-|Qt
zN?HAj=MW=#WdBy(6Hc56vK*HWTUX`~@LX8pcQo*ppI^dICE{#h_G$R%<Ky4mE%;0e
ztep92N3XA+1bY-%e$Yh$+RV^RY_D4{|5v*5^=ZP}-@0@BujkR$7~S^}5Oh)~IW99B
zVvaZp*+owD5V_WfJ(;A1g8`a7JDnm?Kn#>%)pM=JHTqS6#w1x3{gNb`{H!uelxE2!
z)%5<7be3=0N$YoC3ZrA`cWpiTSh*5diLde1DS7@{p>!gyELfn|@)_=8bu~hn1m$n?
z$lCV6gG);7XcXYmDr4eU-zH$h;?u(Cfexf2q$ebq<<;pfN{it7J~QC7W7%<nE@$K&
zm3Y72&4iv^sM?j6g_HpgV<2a6-m461qxNN7Mm>yculC|}1hWy82~5Ni0uo|knz3a_
zdt^(=(8e^(x&##WSYt`GYj6Fu3A;Q9s&fYNH@05Yy%3;DakD+UusS705&RjQC^jLB
zW>;1&2;{FVOZ!vt%wh<HX)fKW5n0XOB*X+gKWzr*arT`rQ>`3Wn14<`-WY-^M~7;O
zT#zg}c^pq3)dpnHCyN@YXocX!TH0}26&;=%3{I)Lwr-QPD?n`pIn#VTkM`8PD~y&z
z=P*Vf%QTRA79vu}aE`>sR;2sj6DpLoZ}T6}K-9N%g{H?!NW~)H)ay~eDhg_vAx1SO
zHkEbmOR+Sz!5pn#fAk<`pwbQ?`aj4vTEjNE#)`_Q5J8sHC-gdYI|ZcC*VSs=lZ4z7
zDbb2L{!NNNx^jG;mf-C@$--4*MzP<qf_YA@3DUI5b4v0rLN`iXJ5o$t6XDR!;g$=5
zkde?GrOil8X^?I36!;hh+v^>x2;bhzRwccuC#;EC`R=s+*q*B}If|WX(`R>+bOBrW
z&RTEg9BOa}dRzwA#_OJVuapzObj}_)Ox;)#Fsj8DpqAo2D4=jwgJPE<jh;ddl-}HZ
z+#x+|R4|P$v9DM!AF+lXdVRZb3wONu+t>&z;&|KCBA^1%H`E~pJ#nSbZ(gBA%0LQS
z5)Z4p#;@#M@62#=^XGmKbGP3t4WH_-sY!fof*A0jA{fL25GTX~4TE>2B1GJAbzCQn
zCd}c<S@}-54&eMGk=j_arJ0NP(B?r$k(Ni8z2X>a*+pewB%dR*#4I6X6?uy63pC58
zz9p6CH`3h^WQ{xcqI4%Fdu6M)f>p45pRk*>>)PC=3ffzQG&ZF5N<MK_qlG?UANIVu
z@kveQE+%K)u-Ei^h8yfI9x7~csj2sg%7p02_4^aHsfd`ARJ+5}@8=x)q((~!Kte6L
z&+YKt@smcf1**oSgK{2{TyYj1jMD>dg5=+#WrJ^tHA(ZtU!CJ?Q&zvQtBX1NoaPgt
zf8|jM#GhFfy2@WW{P!P>e^y`@8Vo9IOjKN3Pu{(cnXs&M;uoWnSEz-%_yD0lw*1{I
zPVq>`sAiv*QY)2pMV9;4+5&J2R){S}ReaxWW1k>$Du_IBwUEnr!3yB;1?^IE-bVS)
z5u*}MO$CGUb-2fKeMQpW@cr?0<{)@lc(`b-;}g1Qq7W+Dq0i}+eduOecy%MoMCDCB
zS<PGaMaH9BM&T|LQSxJ$l0YV)+;C^N?dSuT7a>8e(D^LW+(cip;EKY7o79`IX<;NV
zVX6xp6PO!>@r)&J$U^iYkY!!q6J4bES4pGfDmCP*eKk#K);wH4Y$_$(z88;>Q#eiQ
zV;zzAGSmDzp7Z`J=^H^=d(Up$WUh+qdpD5^L|?;XKPPA01A{*5n9(6&{IK=~tluyN
z;kc7wR_u^PwM{$(`BvH7U>}$o-%n--WHL-n>riM;f@oi=O=tq{k>}9jxg&g+qSQn|
zW6-B@qtsv=Yg2JJU0l;1)TQs+AkE@Ugrq(IXNrunRm=G664zk`I0g9d=CqXBXFqBU
zO7Ut_Y8@p$jJfS0QyzpVt396Qi2ehDC?~JvE&LM1ya@VIe27z&#EVa)u52I+d7Y0*
zwtNpRA-E(S#y+VE6T!i0<}-)-ghB<|E-s>^orQkZGoj6(8x+rsU(~GQI!p6jpHVKy
ztCNf=nx)_zrldQXr2<)F6S^J34Y8a?YQ$`+WW4E!&>4qfCR&%t?ZoF&EhNloBQ)O%
zez9T~DNU+}io|x*i<r%0bLTyhshoV{vI>eV`fHU7nyHxaUc^ZXvwpS%rf^^M!b&2~
zALh<}J-if(Nz4$%zHtI%h#HI!5+7dH9v7G?uG}cC8s%yX;^q$b3}Ne$d6XK9Gi(gA
z4uc(wfXxVyHO5f7aZKmr3$WId4b@BF4OUZ*GHa?cz=|5Qc9z-J16e~^Q?5MjFP6&1
zMy@Wqv+01*((Od0g&2LpOp{%-3~MX?8lZ6_!_lJ!6IvJ?>WRUV;2S!9(M{9pbj5gI
zOYDoi3^M7U>n}WZfKpQk()U>RAgi=CA1mmKd??oMn;QD%Eryu#MDzD=i##}tLeM`1
z8>^Wp`B6YE)bgq5hh-u(DpZy!N;CeNzd7C_@oJ6fDIMgOHE9F_9VdD2+;rR`q0j-!
z_@tI<xTZ^(V`XEh4%-~W``1ahwP8G`A1*GIL0YvO0mW0pfNJIAmZ8hZVi7>ukOAAY
z^ic~k@7d~;?^~zYVO;HarbKv{i)J~cN5=!f#fKFNz-ixk3uP66u;hhxBN{K$C}G_k
z)!hJxc4+r@@<Wi+N%9b0F`7^^93el8jM~xSldwAIAI1@_y&<p5<yp`?w8l;+yv<2S
z3RF3X%o~Jl3w1+n8V|aZM&llO&Gn^}z<A9UD!O?%0#axe3a@!MNpotsYXU=*IRP1m
z(_#}rbMsrZ>k^cdO?-{366r%7e7&m@`O{*YxaaWeGKQpP<}V3mZmD{OUvOeNit1Xs
z(nWAjNd4vABA=vgR^mv3--K1xzaK!}xT=sl$!-w@RL__}=$O?79O3eNtw6ZuyBz6Z
z-!aG6XrV&aOv(2z6AGA9DD6LM@S~pC3zAsCcdSpaVl>#h+~JmVBAUyud4h#b-05-8
zCvkH*;*z~qy;~(0e1B~aWH;SdpGcxx16R+iN|6+OPQIc#!7THMpx@dRHnnTPgB}$6
z<iVX`7uTOKk3A($@wX0nmBrp6Dz0KWaJoxI+c-L%Ypd=o8&<Lkn64|WTwpJmfOOo+
zlZGg(M=2^f^4c3<++kNz#-s@|6>%qKD`<Z!GT(s0R<=H1YtFIS9o&+q8ybi6OQtzy
zM_-WxXSy|`9!5UoX+apM%$hhP(!FA0?!~(qqa*^=CwO+~8^yKO<T_b#C1okHCN3@3
zyC9F9c44S6`Ps%Ht6{SaMyp|UT=KSW4<XtA3Q@N;GF)cbo=PfPg$0;sP|8LN7Djtg
zO&>l;pLVM_J!q$=Ey6{x54mS{J*u*Lk8=Ov!t4lV-1xJ|RPIScjv|=KUTW~~*e*y9
zQYN;Jd7)-GcW_%Ge`fmuvn*LFllc)7V0`=e^pqK(%AO7{T#-FdV;r+!bkJu3Q?klI
zppUQj^GXHWd%#M$GsM%X7Qn}wBe<ZNM2!kAp;^jDsbr8CdxYhOH5$LWSl_Oc`b)O#
z7d1PNt#&mHz-bAe-UWzY`xJWWo*m-<6kYd$Z`2F_@}B~5Hs=31kA{==e_79_Egien
zhWK9rI6g+kFnd5h0rT}a=*!7a>o6V=xhtG&+&0d`tFkwWW<indvk{L#tQLxD6?$b4
z(YoUY+xU5E*y=1ie$0l99tgCaY;LcdKRn4l@bKsvx&I2l7f<mQ?c|lH!gAHWFesZx
zVn=TQYhT_Kh5PXQ<pFc>y8}uW<Z<sG1ASC3!gk@Gd9T7rid^4pof)6Oi{K^T4<2uG
z{e0q*6)6j(|16qxd3}D}jSh~rG)nnr^nAHXT#E(G3Ccp@i*bLy&z@fnmc~=8N&+l}
zJI`#S%Lx<|OdaW<aPd-c)sHob=D<#vD$-R>bPs!lGi|IAMKVZ~e1AS}vNU}?r_zOn
z15Fu{yruC4nd%I!niS+)=W|&JR^Z1d)=U_6+pEPEl4N2in+Re~<xeVzZ!am#0eS0~
zpH*A#f$0UIi_x<b5IwljaFOQ#ZuZ?Cs0W}!T}hzAQXTIbR#KvhJ{j~EX5BA%UJ}UM
zp?e|ovFA%<*Va5u?k=j%zn-$>Pf54S)}opXUEaUXOO^$hJo?gfYs5;2td2i<62hga
z?z3hni(`0A3V(Njhmlcwn<!N~nMVnoMqT2Zu%j9DgC&iBN7XUlpLY)ocqgGuG}6R`
zfGy(Qh1SiN?C15f7hzE^OU<j$jeRQ%jo;wa=C%L^523J6bTZlqqo`waM#LMbUZB(E
z>W}=@{~CC!<LghTlwf_aT|DhmX-)wHtnI70)KtSVFw`j0q(xv;$oR#bu?pkS?Y*KV
zOhA<r#D878aef2i|BJ|Gjd|$ZdKSkPliMxrE#~nJwHFAJ4i*HdmPbrUL}DD0YDui)
zf^x2#l)O5k8HTr$OC08#k8zeaMD4SPNyLZ(D4qGF1l>9!q_q&}UN~|pPxYi)oK}LL
z5D1fkFH7|)uRcGgjs}R9%D_qY_0?-^s~wN%YOJ(937V!*x4Ml|O%Mk2O6=8<GCW1;
zuY|tIpk$gEz7k-|q5`NI<ypJ`$uESJ0+@bb9Bmf287Q%7l$~A1S;3M44<*kzQpWm1
zek8}C1P-R9bG*F@I6n`#NSqQc>I^M{OFk5Swvc1o?rnON7xBEr+6L|=#zpA?n#NZh
zt$AynMx$V^6MG49Z9-0K$Z*wL-wZgaO$RqfYS6WEH)A=h+1lSy7wBM66<U6tTeR^6
z&XKqgi{2^EpiHq_l5`cuq1fR%G~R=z!uJ)VX5nghabSK2I$MR3b=<0mk5Ihu-B%4&
zFi+P%AHPKS$&l-uiUz)TTrki5TR*f5S4$o5ybPR#sMVZho7RoCcQr?f&%3C;d0dgN
zim+vt7~RT~#<IJ^gXXt+mCLat!#D2#tz>(U!ZS+y{vfZb7x-6ZW{sD7G3h#+)XsY5
zNx}d^tHC3^y2&U;D(t>Zl!12&5SHOla3M>}UYrzR80scGY9fG0(#ph%%c6VQMvoFt
zQtMqs8?JfJv1tDEfHcQ*y^m@x0`Q|umYKC<{!`6%U4ySa{bWrCFvm>8`yuPkWSwN}
zf~GI;89_rl@O$FcLJct&CajBv#LOMcO*z3i4Z15x`voK!k;qE^KK-_Is^KFWbS!CZ
z`i|6n62g4=`Z7zOQC+q;UrMh^t;h*mdh~+v3^7g^kI_5NzUw7K%{07-E%dNIs@7<*
z;lDJ5xGyFXm4~>1C{dEcOi&6drR@uVE@{z-kzN%FUhlWr)>^;!2e+#261yK+-Fg7Q
zF=&V&-z4p@@P$1uh`<RS1U8XKp1%~$!Z3lXYD;??k=6T?Px?vdc{cjhyUJE5WEjP3
zlbF^XS8HJUvO~fJYg!gvw_WjV7sAHrSlH)uisiQ@+TOW4QmyBMx8>A5&$Z|Kmrr-)
zmdjkn%Bcj<`)7A{vXuld?p4w0rrOFPMi}@K)boY%*3C>F+J#aSE9OZI5cGh(tP8TH
zmIjW>$O}YJOCqcFKEm)pRci~W0>RM`N66S7sF)S5k)fZj^!yiIkCy3YJU}Z0BOT}S
z1W{~!tIeeXZr5_%oENqZ%_HfY38&1~1cAc785q){Eu(Z>rQ1;b1xWqxa3N8Xh%$&1
z;Rr6uLt>NyOlwC)yzy)j8iM*P(0?_B(~@XuN_<;LLp<uHcEj?j43FYCk!XqtQa@YB
zZ8S$!6?C*{9n7gQ_V2jJt>G#&aiWBROb;ba*4ns~qmuL<k^wP#bgVgCsG^|gPmJVW
zWT0{U+2P8%8dFHG-@Ph;O=zw#n@?tfwp|baf%%JNbMBh9{c|lg$K87iDHg<R13=*T
zII6}9W3=Wss-4&urD#>ar&S_q0Q#;X#S8X16zGjmE}a3ese%`_;aehexX{3vOkg@`
zex^1vj%-%C8=h+2-u5b7tYGqhMiA`NlCPN+G6P078`*{|z*;Wjc4|Ms!3BK84ZtEv
za*bqXC{ib@toB>LTbk`Rh0P#ae`bE<ZLD9C@i}>^!X;%g9m-J)kS0bKwKK9)Bf9Z0
z-!|%ZfAvANS_o`rtNs&oOA(NCJ9l@2A|~Q+Y$+!evGzx+YT%?{6`jVxl}>N!*r?Z5
z?Pl0Yb?tyfz!0||ay#V_?4Uw}LXVn0C)-`BXzXeClujL*AiPf|?On2MtoG*8E~%K%
z+6eIQ#CDjGGTe9VN-^>E96zV1uUml`9WoAVCq8IB9Zc|Z=+t7|Syc>5J4!Mer{PAY
zPS`T+T=Bl0-kVb#h>SBZQE=Gt)vdChh|uMFlQia$-yU8I6Ag;#VY$s&8_gt-m?Cib
z=v(wY`SF>x)lIqR?#w(1ny18YpW~_5w`hy-AhDOehAw)K_@1pps_JDFtfu6eZx6p-
z>pZ2v_0bX~jga;bg~J6?u8bg3{}sGKp`%GsoO*4o7Ae8Bm)UMHxcq;UQL|cYzPc4L
z-yBjIh*U9=7$|qE2wK}c6@S}p!>g%$??R?D)yj{%lqMm6)Hs~FqM>)(@SGjQBveM;
zUM^YT2D40iZ(JJg5&%1UysDGq`e{7zSF*w4)vXytJ1>EsqUO@yc<`U9!}eoPmD=rN
z=fuh#*+_&RJ7Y{msSD<T<J}H5DKi<@`X2MAzZ&thW5r?F)+L;(zq)=?Q{4OC9CIV}
zP+ZhZw=xD5*MAzFG&2{^+Hm!63<0mB?Ww9X`i6AT@|d-V119BbNLI0*J9B7dxD<&-
zXU55+i<pCZhS1Eh(wpxrWbXlSz<OQ5gwD6UOfUOTpDzrGv=-+vNQ+R<@LE&lu7@@m
z{39^dKIM>igSO$JM**Rodwxx@!g0Vx+I0%04QjMQWo-(*)gh%PQ0&3vwZW>94)jsw
z&opZSN85%Kp%hJlpo^bDRL9E!fv)QN0$lNI86R!w+%8;NZ76%+gq48w3rLS+6<t`B
z*fREXG{);(vKgi!uF&e7H9%O&vkM!{b>#2%8z`Nhj>75xN-Iw??_ePP?R5lWALy@W
zT5UQDI{RdK8@Lp{AS3adon0_Adu;FtiUBz9J{HmWZVZJo<&eX?GVJr*!;Vuw3WEb4
z(7Iy%AGxBsZyWy!y-<*KIDC^@?lUIAMA?#VU_7J`DOP>b_fg=6L{ytHcjxUKgNcBi
zCf}59WeEG=_&zgKxH-&tPV<O<r|Q>c#!3Y(S}6bIh!+#;)5ZV$_q*OhhG@m{=}sGx
zo6$ijmG9B>Q^Mj2g=~pQ_YE<<MEyozu%OrV%Xcl?T^wzUt^R&+*?m8*6y++V3<Jcn
zZP+g+aqXZ@aM9a7e0%8GZe*)d8nEbH6LP{#-~K2RP!JTEV;7FCB#s)DLcZb!hDo_r
z3;P^REeq~B*Z&Ryd2v7`sCDg<wuZDA&nq6nVe=13d`i?*7M_0dFffo<oqJUBJJN{b
z`c`|<DQeg1Tl9Pc3>cT~t!z&k;JX#AxL{e{HZX1bhd-I5K=mJE`~R6-XJlt$`Cld1
zyEUa_|C?1Yg|8u+REDmPakp`q*G;;nJ+C}}gn}~u!=IE;Haf=aTfG^?%Zw!IND*-~
z8o&4|DVNJ$dz&dw9-WB1<FCB<*$Qbcw7@Tc)nw`jFkqRs<0s>1U96JgTl*wT*<6xa
z=q0b-=rz9=4rGp||9XJXk}j8h)=G|rdVcn=@A7YghJ=lUO&#F^@ICke_~&-+ZLb{(
zHJ7CcF4TcyGZ7`3BcQnaUr}`A(Qxbp5t5lVHM0ZdtPmb634AAt#_*IT;Xf<!8KoVr
z_+}T%`fw>Gi0VA|YYt-PP^(-*Xp9+VMBOx(pRC~#9N5Y`46!pT5|3Xc2H>vRJEPAD
zhuEDRa8T&2*H7%yN{Om)!xI+Ga!yn?S_!uAP#(NEk8lzChh!s@r~BvEK97&qR?+uD
z^G|>H*&KjaA+Awn3+oo-d;jJ1*z`nMugez2HHoFYiM#do$PbRGzq}KSpB6WdOnKvx
z;DTgTDGCvy-467n_+P=gv=)Tu|FT--@X$Do5d6)u3SkRtN#x=yv25hw5aq|MUihgC
zo1gP}%EyRja1+o}Gl-NI99kfRJQ``D-M|Y?CjKz3Ssz)sy-aXVwNtfwD@7x(oW*Ow
z+sXWMZpE94%$pj+)YpWs1#PomM_Zq;9@fq>Uk-{mZbs2sPX{k|*0X&?L*a{(DTr`C
z!v&0;Rm=!1cd;*d+IM|@ks~;c4-kfgq{R8Sxf}9wef8|MGa%_HPjB}=PyxP(I?#Jn
z?1$cnQ&jI0eVtc=q<vo3!<m7{wZj6*ul6eG)u!&Nh=0uJ<<^3CE#iB0I~sdPDU|Li
z*&;NHFPXx5l*x>aC$HwEsaD;zEzyRyTXu<*zS3Ul`!q_35ff+cxkgCjbL4267s_Pz
z8bQ`&@^O94go}s~mwk`lWG0S%GA91>jAQK;CMq^VHb+<*q|7tH;I`_xT)iLYG&tr!
z$?wxglr<gQS^%lZI95>+5-w_B81gkVwA~qJN9Qb0lc*$MU+7AmFQ^C^0w#8mpe|)B
zhaIV#K6q^6DW1NIs2h3paqZikjXINk>7%o<35@z_<~jP7D>ReOEfA?n_@{geCsH;f
z#u5@gayl1#pP=a7GtqLc>+N(zY%~&;@{h&Cq8-0P?IC|Xfk_9w*+nmk0eyCpN(h!8
zqz9#Fd1<$;h@ypdjLy1|AXPiuG1N{o>=a}c^*@p9yAV5f0y`6?ZC0GJBDDVX&^A~4
zvZt1IkwG8_F;tO$2=G7LNk?2Fp+1rb>IW&}&p1XTy7l^vRD}d3t=Vj-;VRoybmhcU
z`-MbibrvP<(<L5s#JZU9t;9>HEtpD-#LURu<`mrk#dB;%#g<`)V8i2Zo=kqL(v<|a
z3NrrN=5}JrNOsIdK_%;$N-R1qU=UIg%c&ldWA)Vca<~H~*fW7Jo80r8xYpbRK8DhU
z;3L}YJtu7uc`2gYXAvdmiN-gW>)@(WpBo<Jm=c|sFD|I9z)qY?iputexa+NaLItn9
zTDgkDIsQ&pXm!_fC6;H6A!kz2$<q71x#wHe70oH@>t_rE?J@f`qLdAZ(Q?a}Gi<xc
zC(WiUpB2UR4&u^G9j)*J{d&~iWw3F9m&e|4^^<dfmplkn#~q*2ZgkFI7FKcT(BF#e
zi>&o>u$;*VNduQswsP&;({U&~rCT(vMM))sKj75~0gp~WGH?tpo^g_|qQRGI`<APq
zO{7Eq%)_0@oaDSWQcgFkIu>$1*U|OH-8<1s3*1*wRK1o(i!cwmyhie%!g*11{ZF{C
zQvX^#@=??>;|Ckr5(Pi{cvBj!uqSM)3`8w+E7z2I|Ags_2=F@U48vhsz$~3R30T~Z
z=I)<z%)k`~xE#rYT1-8V15Bx3FPixukwx?(2z+)R^y1yAhr)M?*)9Sgpe?+>F;Isv
zi3S_wH^$eZcZF-G1gTh9H5yY;aHzR~ik48rdi!e0-+>zGNzM$r<<3!*o5lD8alKOP
zn9`{<Vcz*{F%^?>^MWK3zBmNpW7LMqBM5c`yeUCd7~E+&SLgAI&w)GiGB%NNoVB_-
z%OcDK2fV#Id<O9O$F%FkbapR*QDnCnD5rD);J+He52V=>I&3{;tR_tA{MES;aO%S{
z|AYU}o7s$|sT;ps0xYyN5_DXDRT-V&1vV7zG%NSrHt3dvDyc4OV6#aDD`cTCWtJ-a
zQB3LLr*~0~{y}R-vVZ*8RA!v>`DVmac~Y)^c_4kvMIr5zNEfY<dI2ramvyRG@6YPv
zCwk;he$#*a6aF97Vk#c?CiEKehL$Eq&h*MIhR*-BEooz5W<sxSVeD*9z{bEpFJWR~
zX6{VD^4k=$aQ-#CISSj^*xT8f*g6w%&<opH+c_%R8yK0;i<r1t7?~)E3(^07G$<DK
z{~o{ozsv@{&IZ>|KK(a(d)?K#`r^7~;$`=0Yv*RS+Pk=^WmAc}p=G5lzFk<xfuP_J
z0D~Gws{FF!MeKD8Yapy4=?}wn_anrjBx+3Z%XI#QGYmR8A|)p7a`U}sKjyr4+XDjp
z|6BTC&d<&`91ez#Y7rxcA(A!(DdzygMFI)R(AP2k0%HXtjZaK?{xUVm{igf+048*=
z^H7#%#>U3x<`53@^g>4kaF!<b_6RKC?g;$B_XPpq2ZH`n>!P@@P?!QmuRigL@ROWh
z%N+p^<c@#`f@L6xm6i3Eu6_JVm9$@!Wslz7-P-z^psxKdoB7YLNL^M&2IZ5MmUcYH
z@8@&(zZi+Sy0|<L@Ie1{39Re>k`l6C*~Kq}j+&><dcEP-$e0)%8VW|#^J`UpfB6M4
zQc+Wnzx*QJ&}g+le(w$dFBm*P%u7Me!s6T=K0(c6lD0KYMe(2guZ`nZkz;%~7Ek;O
z0!+EJwzKQKAHnwW^4b@4N9g-~Ed4a_a6r4ZeXwEO;$M@DR*yI5y2dXX^;cGjN>AKw
zASl@1%=30pS<4o6cySqZ7VzKi!#@XaYI3skx@8juetK#OjZQ~`Erm|kbFU9@_#Zf)
zJHapIe1IJW7WR8{_utDIDW2b#!!IYcpMP<EofBdSR!Z<XE|<ufO`FngS$7T5*rTPm
z991!!&gcK(=_>=GY`TVTi<MkJkX}GSx|;=DLO>*>MFByiJC_!g6lsu7>5v9#mhO=5
z?(T;7T7ACn4?nneW=_qSIW^}q70y;>etr;-e{Q2r_3^;RMSvPvjW?_8-*ly~&-d?v
zj%1zfz`LYXR`b8$lZ91tOpTTS$LXCQFVbUVnu|`EgURezYYldZ!Wx=uV#BV16LoPX
zC>-414m=fm0X)1XtEd<wcC%A90JJvj^h-*b1zzA}o`3)GqZ=R+n2-PR!1OS+_)BqH
zL)~8J%9R7>>D7B}J$bIKjS8%NnQ?7}psn}V0jvPS6rP03tgNpa85%kQ?N1_t4-Tv^
z?|?`B>(A$0f%c`q(}YfQ&PIxg;K{D#!#&`AI5V^DqoX6>6;*rSnZj|G4=9E+9W7W_
z70;GcDxABuTulUBku}(QZyc=PQmiXt^kCxYz{B(nBEs*UQ!@u?kj^C8RqocAsC0Wx
zMTw{(uaz^E8AU`oS8ZcLJ_C>SebY+I&OX1|E@Ia#{wgDXcy)7g1H4Vxa5iODQ&Xb>
zcJ~4#I`st@=Un>_W!K$VR_K-@&-<c?sOX&K6B{i{(id3ZyoV)9?Z4K>>FmC_O}?+R
zFzFyuIWbeQ&;R2XmARAmc1}0arGRs|CbKg?pFXWIKjJ>=c&mNvxr>XE6iTL)zJdI&
zr7&*88x1~iag1h`e(MghPh6AEeofVRu+Zz?5p%}(bW24*YEC;&o%Lq}cY)kc;oD(G
z8c%6=rwp&1mje;w_UZ@z{`7awXSLh?tvmK`ToDfF(=4E{O!qBUgUfBf)hW^G%eDTD
zm|qq-vzsT?8c~=_SYU;IrN)d3>tdlw#XVvDsulOqTw$h_2AVx#tEs5%5gpPAc{k}^
zhRC;ynxMMA#2598u#=_(s@NHXvX|<w#-W}4&=tNICQhTuBs)3R{9nW4jt>l<@<JIX
zAd8jYXY*=4=eW7SI(5RDif%ELKNA^3&BOd{=fjD@V#0}-w|@@2hEp7I`h$iZwIF~B
z?IRcGn6gH(hz~nGzDaI$9+F})W%JH$CX_TR)xYQ<1ejlrBI9V?Yd30xUUIWvDcEpD
z(td4q5>Hkg^(RP4szh4wLjxehkjI`}sgh#}V?)-oRUYm;QnwZPruuQnv@`_Uc1VrI
zvu~0C(9ckT5%$f31!lIn;N;#J-JieV(larRb?<BrLt@`Cy-WB)8lMpJ`wjz~4)Oyb
z8A`)@uBM{~&DH+s4ogUdP(k=at6B!XoHi{`VZh3u42|%YVsUXcrLrD}Y`b|uPAg9e
zSRjY9Pj8Pr0#%O!ud5>Ou@ImK&`$W3-6Y0siNbU2F<W*e;l><QCgD|gnSslWxdxy+
z{49!iB5|^kj+aYGdfG>Oc_S1tiiY^;Ol)(9g7|$HN9Vut^GUb<aw7G4GxH|RSpGBt
z`}q8;FZUNlxhY1|ti5_M!!Ok3o_%}G3Q_%7m$0DVqwQG(f7$p9#>4RCbr^r`ZV)u(
z>0it$<ULQxmbdxZd6OB48MYG+zL>tZYG!Y@%N{l&bWmfUA9c2F7RC$MY)|R!tu#~l
zSQjb#I+kMPr^k;<z4e8^@^CDn<;Z!RKTkXk5S+j8^U-*|*=**;_X!Gxl(io#X`P0f
z$U~xi0;d-rp>fr<rZ;W1K~67>w^L{BIPb1>+7GB)KEcCs56IBi`d#W5Bkr=sW^-+B
zk-g1t)0?Aq3<%~`p_;S2Q6$6if2s<&3r#vmrKJt`^2n23o11GzmoEn!c!q;(`ND_Q
zx`>W51GWe=*47JbM!i6rV~?f@FImwiYuUv*i#V$g1709x`DsVZ?+pn+clb#Ns*F=o
zKxG(mSw?N*DI*rZV1*@^zhCmp7d>;O>?gbhP*3mZ<w14QiN5ZIGhXz}w_<U*l2by$
zjrO<bkc<ytbK=bFpzA7jpM<$*);F&kpH*GOL5U4+UaxdfDkpyb)=Z-nB-r9+IQ|xH
zZTQiw?<Yy>+3Ci6R{7e*=gvOD%?OR19bBCI#Fi`9SBrgjPjoXz)v=$xJ2rZhyLCL^
z-a$s}12P$XL?*p2O6u%<n)3rW*OXPK?R=NgF%6|O6yk4+?L0JWd<I6N@D$z}3C-;X
zeQWN-jHQ)^Tf5YM88{t>{t*D<*VIh9CN)$Yi8N>E81q5C8?QGcDV=J}-_#<4PJ4tn
zFhff<_i7q)N*_Wp)WB}EZuzo1b6vVBWAmmeadm<{a%>%*SCoy&0rQagLr)M{&ni5f
zN2sh;fn~U}=fS~3lu(zA86><h$qPHO$T`CXLVI%j`>LN=UgS)cb_%X1w+Bm>%u*S7
zGvoAASM>0c`aUjGqwSClKh~Q-FBF25SUPrGL&i1ta#VQF4!c$_^a|+kUJRv+hZuk-
z6_e?T!nIRf<C+nXR1^fyR{PBRXo8Qayx4Kc9jC1)ERJdAg1Dk{D(|Y6n5>B9{tLwL
zjSW!ajYXSE#&P^dz>9l>$gD)?3xC=dFGM5ABCx_&0x<+5Zo&=1X=9>#O1=%_fd6QF
zzR^0CS%%6`b){CvG^8C{eN0wY{f?AFU%>Ww(u&BPHO9$ndr56y)u9IuQx`e#NZC&;
zGgm#ovyiy;l6MAZzhdjgv2h{)nTUKHjkss^H}gPB-B;FyX2jO`yqXyV`&@;Sj<*oX
zAO<13e}+17H?|C{ZPw2*8)Wk9hl&;|zc#1XY3{&8#}+fc?nI>Pe&laKoE)XDzbNvy
zBk;^|<G~4EUXt{e`)Tub)z*!!S)7wJTTxPK{wX*_#$eW_@aN>bP;>!`2T{uun|a&i
z#U<6@dTp%X+8nuCZUS{cubs;Qsc2A@cHIYz?E_^AMbG<n%bo?8i}T>_@WP*0LXzJ*
zhYH!WIZ6Em4$;KQ8ZSwhUoYbdwpYe5y7PpLxrBA|hVl#{e2OztlN3^-$5lCgw#xl>
zi}-JDCB>8T0l}R~geg3X>-q27LHA&jxkcD1eDJjXcf5d0yAo|Jl+?fTuq*7@vNSN~
zFCk9_QBrBHAhRp2btLl`w3LmoGG`K2eo95n$*j|{xqR%rNvgeOv{l=I-}ErzKNntk
zXb&?`xYhXEvY(C78?-t6Xj`~GdMGLl<`o^<PJKm#PcBC4e?yn9`n^3j_Q;rg_b}D1
z?`I8m%osFp?axZn#tuwe;az3O@JmjUccdCi@;5`Z3s$G>bq+gh7wyO{9uQc$-7>UA
z6g};L%H3rFVWi8r$ufFkeC&5wvioJ6B6zrK-82ov&VEIudz@33Vks?S2gc9aHggnW
zeHw5)6eaFtwMCOVuX}qWOKPmln>i$UK=maN$B_rS+!qBfr6c@TT1s1IH{XzP;xmoV
zc9YG|+?t+0<?e#}T>83pvv$+|%p-i(Dh;)>i*_FNZn%ZTWW$ch<o`3v;>IQgu!k{n
z3hns;;6dRtcU^O7qPw|X&w|@I>s@$4Wa!5R*#%TP%qrOmrxUH6hd@(5@gTJJ=O;UM
zT85jhny(vOdt0We**ny4M)~jW2K6KU>tDE7fORwE6#!ZQ0?7yk_3c0?ewaF}?k}Fb
z;~<qD?S=Qzz^%2gaIh(CZ)aO2G+2HNzSzD$<}scui$%kRWPJBRrd{sMcN@?gP*FfT
zi7*u=M6z}rY}1uvmyF4eNO@xxe)EQIotD)sYq59o6yceKfzV%sZ^YL2!(u8Bh%M;a
zME{U;akKymK|0-<yOF6$bnqV}g5r^O$4-hLC_5fA!1u4z-1Q9@+!&L;kglq+$M|7G
zQ2NAzP3_J5<EN0s2g<oS8RK5J{XMSNxeZwy9;7@=FggVU8d$lDR3mBRh8!M-11Z=o
z6j?cV0ToEN|3w@uv2yXo=ZVeIMOt3aGvH{maq)e_V5U~!{21JS5lew6JQH9w+0HO-
zm^mCyE$lH{V!n<6126(d*Sh}AB1{NgoLbm9c&G+f+T0gx`$F!l@}d23i#Iew{5309
zMj)^e$#xrBHjndaOh?^YE3#lJLI+gQ6Mjj;TSk9T%*D>V>c95|<zA*TvMuKj2iu*-
zOY0}chD)VZLue&4d^QC<BXoZhoj5h<la?pzjfV7kfpTe&+?EWh+)?x<RwFE#Kb3$h
zX;X?YJa4avreqDdVH{m;eD{(PhwCz-u?yQH%juHLbq7p3I-h*vnBtPP6-ol3H6Sq9
zvA*Kr<(FdIf9-*${pLSuK+;UoHt=Zy4Arnt=WBoYfi@SAP58PZ=K~8cEegddYpRUu
z!C^74c;3(jSf-Nd`?{7*&u~V)GUrIS;Z(x6Ur2f2N<Z-e{lrAi`$nDSn-C+*$@>NC
zmxRy&b>J0+<f-<WBn3?9i$tS8_}O9r!*e!SY;G%l<)TAWUBPaAEz0v&jL-mjbTE^z
zg5$VO_I{S=xfYqOD7F?7D+`7I@y34(>7OTexQWnTgjJk3jpfpV9(s6zhL{#%Uxo%&
zH-w9Ljz=P8=*fn+FV`Q3;W#xmA(*6?$hs^INM8CDop4`{+-e+IMJ^v+c#Ff8c14Pv
zz0=-xy1;RR2_UL3+(hR)9d3v0KKQ0H7aiF8(Jg1gdUmsis{BCFA(~9`dHI6)l%>t*
z6CHa!Vs^UM35{{@ZB@_IvLzN_y!zzz+r}$=a9r%?5LH_c1BB0J{>w(`x1L(f+9*+!
zcq_ui->FK)=@ra_=Rkqi&wpFphz$N=Vr}pf$$X-_0NWBIZ&;Pw(E*?#f~dB|2rV0x
zc5K#)DF^LiG0~&BiRB9A$290ZgCuf_?C#hJl6;UD@K`Nbra8|H5;@=bZhg$!yHB#D
zcwUH_i@5yCARP}H;0M?$0TMdd8Uk8-ES9Uo9uGC))_$F@x5g4SKkUs@u1avhkOu{O
zXMiAfG<jC{IorW;<w*cDeW4fkD%*NYi^(jB2p;BgcI~#EXn_e%XKewD>BRS3otLx9
z#9?^_=Lk@)^uXsd4Wu9(xA5=Cw-K;;17HiXU7>^Xkcf3rbVjf@Xebe7828hXd^BUk
zQczj?(2N)<oBw-A0AwRYEoqMje`)PerL52O6rA_uQ8qt!LfHbW+!$qu4=bW6VoLNg
zpKC7y_`%OsBl?x$xE6oMokYL}9<S<<_<%;+^!(RJuM-wvJf?p~$)bHp<o(fnfR)+E
z36&9PcT_w<zgmRxnEm}KgEqTdy<Mn!*ndEE8apY<6fO&__xGEM)Qi;uLDui8U7Z+!
zd5Pxl4`?^PH`srj35l<{HlDY<lLz2bR_u5M)Cd|W<^OXS49lQtq&$~0YC-fKd~$H=
zqxO$cD*E?HMIKFCH5LoNJ91q8)5Zl86W?E#hIKa(5@O&Zf*5=UMQ*E{B+b7b_515q
zQD7vo4_PuVSKpLwLJ&Orx8p}F{`yfCMk0{IEjHCmx3~a&H*#(#Gh))a|HkPM-Az7K
zdkI*Yr-?lde~xc3M#%sV$O9<6ek0&oBiw)k3^9lW4H#-}U9(9%@&CJ+&Px*efCWn-
z2_y)KAov`uZ%%xHX)5p+#DlA@Px^o<0|m!-kdgWpNK^b9qtG1YPV96{Vk(LW$sk_d
z(3YmTe?|BI=X=ZQ2O`--(kvz@w(d#s{_p?XB<0wvmVy$K9l$zv9LA%Y87(>MD96wO
zG7t~vyr`61fjMjE(IV^{=Yg)J)7(Q|ioehcP`hp8?x~Y8iQxgOx$j$jwB1n}{YTtq
zbch%IR@wf;a*HrDSL?<OqB-BaoSlEOqDR2;Kkf06eG?-92G-l9`1}C<tZV`3^A%K7
z5P36Xpp^LT&Ea}8LUkR%w*PXZEv@-~U*DoBf^4d$mG7$$sM?4H>f<I8eKElQq=1T~
zH`fP5vX2Dr3vAmg(n<di3^CeJ{cj(9`kmSb#JE;`YbT%he8l-L(^4N!rhYx%HnFkf
zh2~&Gcq&~>6$e92|IJsIh?RV@zL~S@#|3b-*PJrClQ!q<9|l2PR9413B)vu6pb8|=
zNkHPIIi>WpKgdW{QTKz(o%wzP%54$PVZ7Hdoz_1owE$x>S$BYxc%oTQLK3k()HtoG
zt@tXPK>#61BCzfZ8QF5wmpPjds>-w4uAPf?<?euDfIY+h6(DGnJ&$)i5!wIa9yKRD
zEyA#&F6qov6fNG_dbK;cDD*SXm*?n~((S-YKf9r!!Ow-Gh;DxHkQX%vn$LI^birxk
zc0pqIQBt~>sND8s@+TBwPG2HskVDJ}I){S=cwS!K5&I#~t|LjnM!exZM0oEhCT8Tm
zmAUzd5}1BGdqcPpQ3;r4nBfL+`NTFc_s~Sf6SbXkp<|!0anvS^Pr{17IB*oqx#!q$
z(J9Z$S_As?>}_l)!Nc10wa-0=W)R`gi!Q$?({@`BDlaP-ukjiY`%frM7b`>O_w_H|
zIH{olqHSu;HPNyCvg?zH2o%nG1~~PAz7K5@OUQAc1I_8U?tGmycS7Ho68MV9b-UU#
z1O-#Be-w_DY<b=6d5+%IWPZwN*<CSp-!p)+yaFW9y}n_G)VA=!KM4Fl1E7tOo}ONe
z=WWgJsOab~Mn;~eqZ+M#%DMK<_r&=4wt(+*RuIS4_1a5J!nXGT?_#WF{`%$Ue@bSI
z#*Ytq*C|hh58iy$duKQ4!yZt4OD}eCU@q$2u?#eYp&r@Fbi%T@S!1Q3C0SXG8eY^C
z5*C)Vi@n-z)8~8wp+$3B@B-1_u!yF3ZWmVlh>H5iOzC#q$A8U^T(mZh4iqp*<U7#z
zn{Vm_6(ykcSHC9Bq1v7Cvk$dEc>_&Nx!}D)*{(m>KoUJ})?E>FKvS5c1I7sSyJgDz
z&c43U*5mXaT^(4DJYj$w7D?$DU6$Q}bZ`VTC*70q?4Kb0^#|@yp5gd_0?(7~)QlPu
zNZoUwb4AsRt(@BEs-WZ6)>goUjEjtrA_YZ~G{jkwfxt8!)VGWsEI$4d;Mz0I&E}`F
z|E*s@Xu48W>Ff8l%_6lQ_Pb#Klaj_^K~7h8Z8C-q2L}*9>#BJGwF{BS$;koeaWCNX
z6hbv{I$n?Yp<`E}AMbhqKl$}J+%pxJHl`L9$Ko|xLdU*{C-!swA<lCrd56P_{y((*
z`Qv~xQ96!5>aZ0?ATeVV{j_lWEx3Y5C8o!<!1&v3CDUh7YxnHDcSD_njUcp}HueL<
z`>)sDGh!j-@oz|TzRu5@e?H|w36>5Qy!G>^BAR*}z-thOfaNQ^q<~G2<AZ?$<ETA_
z+tPF(!y+XJy%8w*(%eUD{8}sFtrHS-Be;GOca_JM_9=@r^8FN)9<TqS0iXkN{0xfo
zQ!Klvpfa3jcRjoPA5wR@{~(qN&*26W7N4u<2**C_F4vs(3YeYh>we}$fuPJ~yds*q
zE?N828^mbPc#a=KE?n-81{~+x+QBiSwXBmc)XaKapDeg%aj+m6BM6A@1%XTIu2~|e
zsB%uX<Ja~(llpx0vfDi976EF?xA|iBmY*8ADausgVQ;|6*8&C5#}+ATkQF)FXIlGr
zgQ-=|K98*$)$sG4U<HkMBnI~9m)1L=$jTjo>~9z{C@y&f8F2*H)Ws+f`By43gH6Ak
z{-PgOWJj8;4RfKqqnYw|U#Gt|V2XqSMg~4`4JeB*8$27;#84^Gp=2#|;#ITZqvpe0
z0OouYfee8}V7+;t1Rw$*Ton7FO=x@_75%V8N5-{G!DeoCh~@49scC(_Nl)lW`4u|~
zhgYur_7oh}nY)WfgsN@p-6XvN=2XMCH!!_=aO5%s2VBkDbkFVq)6fk4Su$2jYH1c-
zziJ08AY+%ox^t(?YLWmP;sl?~&mT)S!KO9dFk-!F)%sGOA1>>snBr_NDbzyu34ksv
zz1?y_$hs-GUmCb4w0HXAaZgLraiLG$nZMDK{x%@-z>**IeXi*Iyy>f_v`??1{2*&z
z=$<)|f{_lF;U}xpomZ|XANn7D67xNBd9(!KO}DTEDhH89<H6QF8=jl@KfE10R-Rv5
z)oo$;ec|#Mn%ycqYX+`MkPIuiai=tA)cr)3tmro2mfQHEjNO-;x+a^PDC~w*53BOB
zPRPTLn*a$AoUUuZ<oyRQuvB-hfd-{591q*`ER*WTx5Aw{J2_}DN4xN^PA~^5)DzEV
zC$jD(vA*G%2pG646)R*y^&x&mQOXRKjE=&$2Jc<te)w|^n@NA$MumGas3ZK&0E%s<
z02EV=_?S|#e}hBXQ2GYRc@~>_>#rrcm&5W0I<p7dxgZ%|xHfo<SAOC7V}O-kOG?Ei
zZ%Z3rbTj#eW6}U%Pe?-s*EsmTJaPF@k2)~R<<PdLUPl4~5G<etI3I3;>QAAf(k&YL
zj67$>iL5DI7M){L9A~3M0$TN`AW*s4f-40m)VOyFu*JRr7o(Z|3Sw;C>db{`LnGI<
zQ))b2;TDEPux7IJPIKp5OKQ~dr>533H8!M|K6{IjuqS7m+1PF|X^YJD1PV3oR)oLh
z&1jt4uNYxg!YEGVIPBYyL$`8aAS|%5-<PA3M_!HTs2>acPEz<W&>k&iVnnOWs{1)6
z8%6jCKuGQ&JfZv?7_(pfioPiauwd8}yX0Es=wi}rdaf<kM8)-vKLB9Q`<m_Qc<?*M
zkb)cV1$6v$BoDfIgsm%;Y26JtnsugPEu{rnSg}O0P<YCqF(<S~z}NtY6gn4+<qd7F
zhYr+KF+L){CXF&bc)woG8nHghpCVR;4jfvW;aEagq^@Q&dM$`38>$Wdo{XQy7kSLS
zE8UNcdW^KLr^}flE+2rr3mZ`q0373l1;93r$+wZl0g&{X4|QB;9XwFe!cd0^z0@w>
zUW<zV9`GshzO^PPOeTxJ0tbMDNswe_lf!uFXCP`&yT?A=FF_pR3ntbl;gSYG5wdhQ
zNg)P(RT2BqNDc8aU14xAN;4%J6xZ4CknAi2#9E-;aArARJ&|nDa9||hBdkgCC@rm5
zo)(YCBaC%h>05ba1jvx8G?%fM@mhWw2n<$UhgF5PbW02#Ffu&xH>QyTStNSbRyT;Y
zUu@?LlR1;408{q+UEp<`kolcEHLU{})9w*-j=zNu;G=<&^r&7hTF_sm#pCIzi0SPR
z5|?IoM`QV;Jbs9WQgX}_fYt<b2;m*S-W#6ti$z#BhqVh$tsn?n_kHH80-|?uuC%;2
z9UP70w-5J@6mI;TwlEv0=Y1}=Ui!ibXdRCxw8-Wnc1~J<?so`4@WDYSEm_g^NpE)}
z8Wt>%wg_Va44ur<Ul=?$ana%fK>2&a1XRS-A>-cl?hXv^m|2TFD_7iTrkYPW^s;=?
z^(6_LhUhh}=963Z`ghqLp`4d%S_mVQo)qj#4VfL3^$fnlIH2MUW4l7wO>bY34jp7G
zV-zQsb_QU)6*_X%${_+Ml1EQcKvK`uN#O%pL~bvsh4Ty$@yoych*2*w!(8B0W;I45
zinm<&GP(=5JZ|WtE|J-dTGVP5Gui;4yi@Tbkh_%<*$;W@2^L_caOTc2j#;8Y9X*ld
zl|kl$bCcBhEMm{;#EYG*H=2r@NxP$e0kp2{`UK-$A0IA-Jpe$^iQ#f=A=d?mSiizd
zsD8!rLWjnYCWI!7rXljf{_=m6nxXoyK?CH@ikxsXK!T0XooZR<xdI+!{s&fW`jb<}
ztWRE%_71bH1vrwnJ}PSej4mf{TZrR(j*0qpG?rm>q=B{MaU`()z9maHUHcaqq1Wmp
z6iWN+(!A>L^bIdpV!vD<0puysnd|qcpA}*6cjd1jwY~<`A}54J+_*=f+XwZ;Etfz1
z=CH8+kc(I2-WjE5SGg&RfDJ0AVqsBm7!#h&fB2hR@)0D_kt?<<(Ew0z3svV(vFxTi
z)vQ$l8lp<AgJC<PIS*h@WgkD(D<3}e`APX>-H3+tV4*w`X13Og4;at{<9~kp-x9;W
zaInc|*hPpXIDG|~9DFuBx%wo#c8QXm>iD-#lk}+-5CWqp8}+0HDwI-RP=$+h?pgbZ
z9g`OAnRrNg<3bF+g1TM<-StMgDmdxqD@!z=VnX4!2G%`s)YD4z(GQ_HLKs4(mgf$X
zxSSRY;;_vFdQ2os83-OG3aL9-M{);XK?O{>G`oudTvQq7B?WHqw_EFem>IX6W8w85
z$pfw;XQdFggSSzm&#{udKt*v_mZIDSl<mbR;z{mr?BYX4>*cZXISLHBT<nm8gXY^<
zP#Fm@zNGlO@5^aJaYjcad>Byp=ny?oFdd==i}?v$NowcK=?&FHh;!dqYq-^RrVsMZ
z0Ts}TDqDaV8xS63U<AypLE`c8n+Fh8S+Jen$^)_Nuy}jF(dUk#_DONY*%}yNu4x1j
z8&2~3XYNb2m09$y7!Y!ik&K70=>duTX3Z4hJaW%+Y}!Mj=#~3Sm(l78_|FXDD8;VJ
ztG~;CzB>Rz^7jI=e>Y(Qbk0!<$hF#m?bqEB^DGAnH)N-a9ZVw9Rv`+BI05SW)Gg!-
z4qu?K2XX1Uj9b-HMRFAZLoxxI-`T(DdVwQFPpwTv6F#VJqKg3;Apq<AhQilcy1hTP
z%A1MEY|Pd8K3hZwRLl4m8=saf0VBWBXSpZMhL|2YC4Wd+pgAx|luA_iHzc>#9<t%T
zb;~3RKvfC|Y9rjME}h3t__zy_ecCV`6q&hOd_lbZ2%;5)5sP|9tXn9OdukJj7x>$1
z_6JftEBHSjP;`*Ox9bjStdH8_>jYw)@^OWPkP%6+4lx-ta>bMH^dLc&dNCla(m8C<
z;&U(G3`*sVi!(*eZExehA1I%^9$3e+s#sul8iK=>Bt#fqPK|N5w-R!Wh=adf*IbR8
zQ(Bs#)ChVB^Ea2?0WW8HwE<HGjyE|o=YX{|g>ZgPNK%`KC&79`<>hW*buLST+H+ws
z56u-<9&;BYsjJHI#bf&UM}GtyGr?z6{=ySASKr@OuRmRYVZ)SEW~v+$z7HhQV}Y51
zPJToP-_AV7$7`9|#B*(kOgYj>@<T1+QyT1edb<1c=W`cp9=eg)xphmF1*uWYFtc|K
zC-UR{QyDT8I3n<I^?t`6#5;>P1VNQwl!Vp)a0yEwm@2dB@x}8?#Qh427GHn0Z{d-Q
zeQSYbmhn>iN%qX+Jh%~DX;_4KmzdCk;v0u)HV9oLp~`xRn9J-rG>4#)J}H5+yL^-m
z{)lF~_etL_Q?&^xWW*k9Nc%i%$4-K{VhrJ~Q@Nu(jZ*$0lJs!^WmN{sPp8OlA4~75
z@d}dzo>GYU{zl>bESlbk7PwpTreVJ7uqT#d6ap1+EjXe{KCCj*JSGTy%@>Q=&7{VQ
z02bba(2f+AZEoyH&{hm1eD;R@CtttS^m+(M6dAal)>`r9C~&@##yBnT2O?PqRF9ys
zr30)qxwzrL3OHpS0L~FPnce(;3b+1I{N`5Ip8vTlaRnb;F5WcD#ik(BjWTKsu>}R`
zF0rqSX(0%`@5+VbdVt>VB+9_LR`X$?CXK~IaDRkmf(C{JA4h>jduE`$LZ0I+13V1<
z@{@t%%>%Hca)QW&ni8SO4}&jlq6Tx<{)|qAB)aR5W6UE(=K6Oo7LwDC6Z>C__T4@O
z_xpw99efJ%ZxhcdRLpaYrwpf3MR~=r4nApXf0%qNjG9dVHoQmqi*YF_bPxOF)AXc0
z1@Hc(jd$(T$Koz~rTOyxl}$@0X~(y(6dlFIoa3y+p66dRBg(oH%ypTczV9jad#@x^
z*Cn31MYpy9LxU^PM^^k29iA5p_|%NxK3KLKzH=~i6S}|yoUl7J6M)yKsr=B^kI!uY
zH?5oO2{h*t@UHLJ_7CK};#h>yT@TqC*HO0A)plWa_B_rzq=5<q85{)PG>u~AN1`+!
zH*zqnJO44X6AS!00n}Rak}CEfHSgBgwlu-mrhc;3W98V;;UOj@QIAWLjMDrt%s3}O
z&tU+i1?f`LBU%vO4?elQ(%N}P`S=+`HHb24rzQYb`c{C=lUBG(W%vkKuw~8aUR|fD
z2rK`SNc|s&HE~CNLZh`sm?6BlE`!=L;gL5GjhG3gO`?^}Yz~$1JOjo&s@Z-f9f^eT
zwJ=2lXD|3XOG$r&3Fc}>w6rXpsf|6Pb~%b0oZ;MAn~F0m&O12)OPp3e4Pe^X##LKA
zRHku#G&_05r`^F?<DmFGWMI-(zp1HUV9CC!R{pb(GggtmV9Q<Z;V19g>FPUXV)YLf
z-r=Wf-dC^Dc_~3)t`>w(odwR9Hiv-q5xhX@=H+TO>GwmmOi2n=hV;>6LO1bt8`XcF
zQJx#ZX+#OAo(1JEsg!*mP`51E(fkr!Hd|eh_ynwwG?#V!{Vke^xDUwur5xIm868Il
zkI!0~4ug}*pkWq?s&ei=?z<TCNr6+M6p{DID(`*U<rKl~<i~+v83|~%`1_>6(KK%H
zs_vfDe)gOgeg>69qUdjaQ{o{jlpdK2oo0O=SIQ4O7?>z6jpWqAoF*SchVZUoe|Y;w
z@dIB*yR78HM4}|NK%D)x3pOLo)37(}=9LTUYc?*VZxYJ%&6+<l&#-hWKp6P{6b{YK
zm4ErLza>^Peutc@DKs3QT$MhHO5*G0I_$DBO*(AEavdAdHUFY^^HqB5-tE!A_>Xkf
z;XXtO1bxWxLBKlg$zevT3k0;?;;4ZKRxgwMmi?<vRN%$7tQ|AwB}2!?rBFvjOFe7A
zoOzkm9aT8RJDonq<*^DwNpC{Yy~MMu!y+$Gwnj<<mv`?v`tP?NtoZ3T>dkiN=6Hoz
z6RJk{rgUs%s(PDKo7?K>TdM_`YifUKwrU&|4R8MNe#OEiAJQ<-yr{N6Sd7ntP@n;?
z8EF{q0o~AxUeniFi+XlWma@8NWNve+Ym!T|RF{_3j|s%3P3>%P_ml=Fvx-$)r{Zs(
zm2Z9-{ZT8ZvBzgvB&1fq4A1>3;c9bMVIa*9PKG+$&eENno)TW0nVO<HZx4vw&$PBm
z+U`~>sO};SI_`VU-gmd6F)yE;C|?G01J9KNP6)A{;S~|<OB8*y$0<Lc3hl7^V8I&Q
zjjQ~mkD>(v*=edD^>E=e_*r%yuRXT1EG4z$hgQ`m_?n%7!n25Z&d(qyV3!u_%wARc
zaWsVA>b+6-)F*MiPKgKmSC2kztRVsK(z8cXp^M9QhYhq|uQ;scppPCQfS%y|bMVji
zcekd8RN1WT6Z(*pAZ|j_<J9IUg}A^_S1FNg3K-vOw(nHQrYBTs_L15a+RWlixWQo=
zp&1_GHPN)#v|S6Qo>TIR4S{koIJ&F?H(L{#5)g6sYJS`x*>ny4&njiJzs*KvBmRrv
zDK5(OaCw-S0&;hEFPhlg+PVT#+j^?X%F6bu*CF9DlJ%fS3bEd!I|!lTE<|-AEW?N0
z#aC>VP&Ps`@U3@FUJn8t1dV7ZD;JEV<0S$$!O08GF}EmI6nSkD#%>YrlXo%f6lXkU
zc1^DDD*$;p!eej!3j9$%#M+-vO9M4tECr=_-c3$S#P+OqJ6Tdl48g6cSAXQ?F?|#1
zG+}>Q8p}U2{`D@B9-@k?oOoHPvC*PalGQb&&uE^527WF4TKa>p>1V563X1=hzbFya
zWXHpcE8p|XI>jR0Pb3^Z$KZfPG(2pni=gja(^*NE@Ft^UlTEd(rmkuk!#}Bo@i!`>
zN4gC4A%HYUqk&`xJG<+f;{l+dt%<X+cBOe-T3Xr?@A<r^r|nT|Cgt<zXQ7@q%h7Um
zS6lf&ujNOcSwDY%{&Ynv=wW>~Gydq6Y;OSpJ#{(NtwhRKQ2i3tXYX2<TD-;W=H){m
zKQQV@y^B+j(N(+rr9F`Ftg0OP^ZoD!_Sbg#w369d^10r*Jg4s@Hyxtt&cVl5i4p@u
z<nH-&M<f4*_Vj9XuksjV6HHC2?ugD+PkG<aO&fr7>)+f&_4oIi)Te2(WW5f+4_xVb
znR0S+LOTugX`icity;>RwrUg+6Z34oj}8w%{3FM{8Y&U4+E4#%(Sab`z`7XBk2%aq
zNv`N7btm26bq=T7Z1cgv@W(YWJwk^3YVfNtDyKFsI2(Gv0gFLvvOn!}#BYAzpHENr
z{JV}2^vHB?@xB=9#I^Q^x+ve>zO_eh_>A{Cz+nfQ#TIcjdH1*L*Ys-xDV|3wvZXda
z^Xp7G_K-h2dwX}=y7%3NWw}uH_@yHgzmD+@!OeY`px|k=+>NwF&NGr3{!XcKwOyF_
zu24ed86>Oyn;74J?YlCcM}7goI1|3v4vXe(G}f*3jGMf-{hnM@@qy8Jyi@k2-xG;%
zWVcge@jX^2+g4xR9SoHpr7AI)Tht3|v9Po6S58~@^h4@Z3Q9|lfY5=lC8VCsZD*oH
zQS@>}Yd;kOj8n`E=rZiHk@R)ohTYv|q=rW}x0qDP$r8}FEzzsCg)|VUI`Mh9CdF9w
zecvrhM7ipl3<pEHr%dXKFj187z?qW!uXb&rI_T1u3SYNYITiJ-fVZyWi*)bjfc)A!
zpwW8WK{KYL+uGXNw`?GbcfCL9x}D7G1|SI&8(Y>NL@rxOX37U2OO<6PLw3#2&8mea
z;d+(M`L#I(plKT9yU?@eySQz!3n$jEfjvhy;`bf<Ypxky?gm1&WX>mcy6tEt&ThlE
z`KQ&;@voLN(MllGY|Y^%+x$kyuEzM`;BFm5F20(Y8X+MeEq~oo?%v*BL_|avyAsnZ
z;?YhwRS*ay)H7&dngd!Laj3ur&+R3heg5?7vlH1;kv%v|a+Q|+VDKci(|`z{8!b_h
zr(-XV%CoC7&sBP~u$GqWxxHjOAiCTG#yfu&!{WzP!eldd+a@zZaH4C2g*85eH5%0B
z=lS5QhB%MXzwFyRKDNXhG%~FLt(KoUZA+p>&R7R}etMr!IsfqIM=&2xWsFt%@Qkr4
z-_zNTHQwi{t|TMQmnZh;D)u@d-lw3|7@V@;<U-N1w}V<Rn@@z<BGcoV9~*|OG-dMt
z_=8?zDkwUHwLx`Wt~nY8fL84fT>`Q1H=}Es*o$-CyRu(1jUzIMOmy6{iga7rnVHa<
zI*q<E77!@8{TEf(B0LiIve8~9aY7e2>(4;EwW*mU$ldA&n3no>CY@(|V3EboSqaBq
z>{Ar3N0HN#^i<7vUt1A@DINQyBQD-GjOLC1h<hq0lA<NOC4F1;Pwf!|MbhkRj%84%
zHPU)b_(@@i%rRar*FIvqfayodSz`Cd*g@)RxOGkuHJpcprTV|v4P=l%HSr6+x&gBh
zzb-YYg9Y<|%BSB2i4>2mW~W#Wk2Vj7icZJ#ZR|Ep%_-FP50!S1K~n65PHoe;gli=A
zJdT^zYmZdXYS<MlX0NVSa|?YRkvf*)13@b;XoG@7GFXpu8euEBq_<r@;pyDguth<t
zoCqXdeD`T8pZLk6t)TsaA`Si@R+0$(vOwmY!DW|k3>$?4Z%j;JoBaFSqv7PFeAe#M
z@62v9*}@gPP#I6~kzfH;kaJ=s^r;*f6kZ}t`e9^qNG?c-(}M&v)3vy+AnjAaquB;S
z=i3deiN#uW_6LeZTG{Ggot-^WOtW3}HDVE6!?VVQx-hf)x78{ifTXtbPg&~`1nZ%n
zm<x{<+U9FM_h>@b2@k(((3TN7DA$1e{o8i)3;Z9yIQ3xnGN!6x>lWFa@LHIk#s1Um
zs8OA`TVRg+r;K1@0U5-2{Ayq#vxNv~yfyUPR?=frNua4P2$NmJa;|()a(h23(_cQH
z#{1&-ZGr%hC6r2GGwsa^*|O^LDl8vF5K-9ZNjBsRN9w$Lt34tMm`f{`Vjq;>o?ddQ
zYuLaW5K<?UA8voRZ<lO5JC~B1CEC2WHzuj=5vGT|$aJ`WO$pX1AdFZHktu7Yv08(;
zRJ*KG{>&CRCyA-8EjH^WnV~=;p4A18J9ULnimOfSz@C&4Uf!SLcyu!N^9Z_VzA#kA
z%V|b%cDOI!s;Ie4yeLU6XB?&3{S7<FQ_~3|O8jsVcb%Rh()a?Lq=J)F^pxr{6)pPV
zEHo#TZY-DckM;}vkAn{vkj%J8QwY|zQ@%{VDX{=2^H?Dh7+~&apmvjK1)_2!3<EyK
z+@7lqMjsT_BpYUPCn#v!uKPbPc0hC^8VwSFx(s0#-fJACDHo-DAQnzFZDdlqe4nP@
zgwTk5r%~AyDi|b`+l+W3s1_%jWO$5|h-8Zx4(|<FaiZ<H9&sNMKli2Grvc02$Z)}O
zzrTAg2zc-Q^dXwP`T-X)9uTd9%iEGk$CUh@7JmPd>$lO@pVd-;em;WVRE`>D{K*aJ
zijt9Xptz(h<PMxEZF6FjJ-J(N*pLoom#r}(Yh+JrtxuS4-SI;O677#ah3p??l<lXh
zHzCF|(+dwfn*Iyc8IdJ~&_<b8<`<gOkCoLw+NGVP37LL&!7Y0(N+NeY=ISX2B@T0U
zVVR9IRg*xuumLX>&j^LVp_0>Q4AUo20Vz|Vc)}ndUI(WpgwKUUa@Qi!`U~P8PXng{
z76R4+b^?w7n8G$fWx{g8p4up-?rw{t3SJ7KvD;NIhVCp>(c#uL<!27z1xgJ$DvxT_
zdPIoVa5vm9F8JoZ%saEtZqA)UFsrU_t(hF9-=^7_IT4zQq6jsHXt~jez$!J<SE5@$
z#trq8Xpe0EDv?Th^MU762qU302e0s5R7J8zjqE##IvJK%YN$ZE{NX2^CicTKt2EIj
zL|KMsST@l!mu#o*hmB3m^jhFejc_2-1Bi2_gW9|s0F$}KH2obV;DKsd1v5Q&cvG0R
znXUl1i~twPlS{AsC37H;9Pn-2P_VOZorE}9zqozw;0-MEA6Vxtnr|I$kiSrV^E(p2
z>qVtIvN@kj6jSo`?*agypHQ1Eo9jx*1Ma=)S?!M_jPasDLa4~bR}gx{A}j){G8&Xl
z59F?(hB_A=7}dQPY54vGfbJpO`sDOM{%3O?Dc-LjM#8~~bOm;2f00GOnw&_@JP6sp
zk|OlyMsqSDJTTxlqvGaLchjOjgM>)vd@x&8*>y1iq4*C%8(Twm<ujE|+JxAuSOjiB
zV3#(Px}0Z!7YSP+TKi<HKsov!L{&KMbQSZ4jqUT0SSR8yg38Oto0$yx5VLpK3$XYw
ze#>{+dtM%Z3+gdohl@;%&sKDv3adT>@_q?;$333mT8Bzo#%atxiOUmx0(g=iKpEEM
zV)k7oDRR&csAi3sx3Hi}l+M5XI+VLifb1mHG>O3M$F}(Io-M#gUp(u<=pU;gt9)KN
z64DtpqX77PfYc`2xBZ*fzJjR$oys$-3ivl(xo@OtDPW^=H6B27?7>$)6DijGsk{6~
zoHK{NzAV<B@rnt=?3^){hcKZJ<S`cQFJSW2?gB|tc0UY!GJN*X3TZ`D)%BnLrB763
zBt_$-357_bf{B#g++1cf_HIeF0}f~AwOf$-?<ltgrPJTq(}g$^cDt^+C1#o{nhnUG
zodOvX!kI71>T=8c0pp)gf%WCdCyP#`mO=%-U&KG=te)6ccy5XhQ~%V<cJ#lV(%zob
z=~q&BR%yC?(-ROOR);|({G<Mh#&v)N=pF5)TxA!eGzCf)t^MKSG+!&x(0V=+!5tN2
zp3$4H!x%{YmtjiC7cu)+uC*Iy+EcZ{-A-(K>+;tzjY_n?EH+>436p(rVMyK3sM3!R
z7qyr^a`H!dPTA>Hn$dcE+}DvzK9XU2NpG>ag#hugb3!HYTWT8K(*jAee}$cle0*h{
zZg{uWrzYNaiiGi8WR;P2C`Hz~1nNtxMQa<C9VQtg8Sm=Oo)spUg&zpy3Ne;q#ypeg
zdK(Ya)&L@_gtXzW|7tC*<&L*loC81SuL$NHaU<8|U{RzFd=Lm)Wz4uC*uHg~JtByJ
zM^30Di`$|B6>k&^V|Fb#qrvok(ZdQ@|Co+Zhp^~_Vws#urMHOC1XtxYhJT}D)pkE3
zQ1<wCT2@9V8DzKPOdI4jRVkG$9+BPO{-aw(W^#m03yFB6fQ6l%MEs8l?Ds{zzYBTD
zxZ(|qxtL%2_NVRbEE@jccg|KN_9w79V9;98RJaisZ-~hVQ=F*87yD(+IJy48p1zJp
zu=WWggNhSNOqI;7_?J`n|6{>}+Mi!adX!1sgA$|cZwX1Y*xZoT&hFB1^<amQ4eZeQ
z2{m#K2FMbSYE?t`kGjZ17W>;D!_)GSR(K|LtU@!9oWA(B862IJXpjsv9UA_b9KMTJ
z?pDr}|MvbC<l}yM4M*tnb?WVeZ8`hKMa(DP8KZD2bF5XwEpCBWdwwhKU-=hE{=9zn
z#7}r=k=bX8#J47%vmoAAXxyIy*u#wSSY&WoPAraseYP!Bv=-1=RiHC0(a(!EuXZ~I
zC>El3wPvO!yNgZfNq=PCg-1lR`hjv4u?9#gzi6rT_<i{+8c11}h3~x^77E*=9yf~(
zj({|Cmu6*VpBk9UJ%0-PX4In2f@4c2_}<HJ&)f3v#$Kp!S?E8@qU$$>a4|){u{*z0
z()`ge(*4|pLu{f12cn8uK`jqKWBnPJCiC#`KCE+jiQfEh+zOtk!wYd!);IPZB1O}v
zTzl>@l;x9UC<@K#ZyhX76K?!psvd8na3K?$jK(tZCAn5Y6mj#;)pK)QGTYxk@WoEw
zSRkJGjj#ScSl}VQcd<jP)yKm_S$Qw&C7JXb-Jd=ELNXmOkZs?BFg3_=XK>m1rf27Z
znf?GJ$6l^sql}>vTS7<0Dz44$<`rq(bZrh<H{ny$9TM9r?7YYVxxlo?L#Wdd9u}4i
zIrSfYT|v%Q8>-!4XH%qAq@XPO8)liRf~}R}LXF*+-SG0txS~;yMmRM1D=LQrIx+O;
zBKVlAg4{7DK`LjLdN<d>0b9a10e?!^$N1~kY3j{Ysy+iFz65~5h%Gomz-{lWZ{a`x
zrO(fCg1*@|7MkSz$@@lRT?4&<bCLr2&Tu^df!>m#)?#j=zk+I71VT3>_he@6PrZo!
z`~v#(dzI;jw$Dy>WtH?sM(63+Kq9#<L13EX*T0ffF@o^dKKFTD{m!k|1dW-yEV8_C
zDO@=K(@Jr|S%--N#+?xeLcdY)*cmdVlu-Ym1rkRk4QOp3?Yw?uiv%->EX=H&+%m_i
zK2|xWx+JS$7oN#R??c7L5{XGGGh)?)K!Z|1)?lh5aBQ>c`w!DIfd#q^KG^5fK39i2
zGpdea5tzT(*pm`_#SX^zsTW^D4z+R+{=+34G#$U)Xm{sT`qRSPgk@Bli*~^baI1S|
zGkL$28^YgQg9%~}u_mL*%b*m9k}H%UvP;zN0~b$I=9;m@&UbO`-tw{~#y4KVR9%x1
zs!C0;&lzbT!oaku2l>1e6hH#?BCOptta$3Msx*RyNs@!7sWTHTdmAQh|H-WRmwg=8
zBf|~AN?B*qXQTIPf!a?2OzgA<L+nSfgG8l~X@1v}{_l6jroQnv5Xe0FKAAv5__dmC
za~tNJbFZu)JL9*3I`<JG8>H6V-P-|qp(^x|oUXY@6X!4`<?g=O#EAjCpR&M}hv-?$
zgu0@{U$xw608R!wVDh-~O<GE5Yx$^tIpr1<gf0$hS^R=)#E7T9f4`86!6OSKaM|h`
zPmFl~Wo6t3>|u5V<NUqBNN$mLFYvP2O#@G}8c)uOl=l+&J~<XQy^nBzOL=4W*}BCW
z7Um8i0((?w&OAQ`c5JNQNk|=wLXHEhgQH-@A5QfyO+4VC<^2x0w7gV?><4Kgj8XD}
zr){T;90@0pKUe^U;NgK?NO*W(zTgDHSujHvr3Wd)(zL`@_1;f8msaj&Dg_34k2zP;
zvd>Mrlo7+MPvXpJb@)a2u$GT`TTm#&|KaNfX2L!8H;s_MH&+Mg;J+yP?7}Z<JdGOS
zdvkL)(-%1Np!c|d6GFb*%TDJfkN%h4VT+JD7Rj`*r<{`>Dq)v)dTUljHIzHKDSw(}
z)7BOIrpRBlC1enIhY*3@Yfwv)uE3=9jZH?LlO}~u0H*}-z^szjTPBlkIeCW>bUs%Y
zo1~-sWQ7iku4!>i13sZugz|<#5C%Wp$Oz7hH}R{m`tpP98*y5$l_P*oIogoAnfy)z
z$$-sY*W!KiPdPn<W5c09M0lcI5-vqz^}4XulF=vG?E6d#f&r-UkeetF6lEkS9DT|H
zD0?6{)Y0;bjV@3x=kf#f8hekg42$epHt9`7a@<!fr&(oOxz`bn8G!nEp+vd}%YA!R
z=Xub-q9jQOsQ*PI^F{#+tvD99Cwu3O6u3w1&*_iQg<oHDqs0{*b;rh5s2d1M(i=fe
zwEocDQ0{a9ULfh^wW5SW)1(BT%5!X0*nPuVd|_w20fc=bvD50p9WUt`e?q8zQK}7>
zVI09~;hipz9xU(X6I2O_>NXg-^4-YvrAhuf)x%8JFx(wgBA5S`4>q@vtbs`OD^T#q
zrb*N-5vVDD17!dXmfP@%*Z>mL{wWxX2XOaI!hT8*0KybLHl_|_ch<X}Ckr{`)z*F}
znQv=ryTcC*3_M8v0p}>KH2*ogbvrhV%UR=Wt;ysu!@l#$h2RA${@c9@<e%zR+hzs^
zc7=-F2{`S|tgRVd-U;vh{(O0~y4`MVZ~r&@xOc3*qM^p@W1~PaDF+KO=iQXEwYJ+}
zj{prSQ<CZHKqHi4>Hcp2J^+^xVo;(ve!Gu~-f7OwRx=(P?6V!8cL*Iz#@57zK;9av
z^&ZOp+OdQYqqIzB-EqJ0*w`2b7$Ruf+TG2>&ktm&=YbNc6Dbz6J)JINZ+1yr>r8o0
z7m8JHVyz}h3W)#iP_=JWRp)ii#{%RbH;YCo?tqWA%x1a0cCSWGSNC4)o|&G0t>Jp_
zgI%D*_E=^{#?$+^PEJ+0ISI^f(&`(ntjd@7f2EjXapY}my>erCMTL8EcGkZjc89-v
zZ}?U0VGaTOZy9`cw*TKUc&5yyH}bZ&M?h71LR_5lsR~e9&Oc)<@nO2Prxtfx<ZR+|
z<#JA5Uc?>Ak~*Uj|2Jh~0`$<#iz0{1c`eG>x=KSyKQ3B)^i8kq>&B&RxxE9~yj>HC
z|DwG=VYR~Qt*bx&A<w#V!HDQMIkArz#v|0(enLwJ^}K`j2bo?u59tMtyr40!*iKi|
zA&bzA<$wLCKj^3}-QS(2&u$4<B$24i`Q0c?k`G`Z@QT8ru+%b|To~<PR)Qsv1M3aq
zC7LY$JtaE0+c`IyOs2opjfX3LxS%-n;2s;})dnhmCld;wa10dndj<|?D@G<!Q&SI!
zoX>PeGdJFzO$*wsX=-Y|O@9Xu8$%;saT4|^<9l4}2siM1O+!Vxf;^aN7Iur!Ol0%H
z0c2L~jjXKfoxsXD_*<KP1X@jPE$H!M9$sGgWQAtLiDWm$MkS`%N6LxxZQfX&@5+NL
zU>qDw$Wd^4ZwK&52yioVLYXpqJDcTCKNW&aN6+!2#?lFZ>cDyFOHYCw|5VgD0F<Kp
z<}_5?1}H9H;p)HA<HgKOO@m<NcGdR0O#7Tl^!Gf1<G#hyudkGt3*HtF0Qr9qdWR9e
z4GXb85KW2Ye$o{{3~Zrh#3!bMV`iDN4lT+N6FjQT)l>qWSws-cjTmMN$-BS15P3=>
z!)iQqs&8?!Ux^usee@0k>mXH8YD`-cJ?YWyQ&d({Gh!jW^R<AUiTre3Jl|C?sggiu
zT**5Oz$G<+yAMebh>73<(n3($-5UE*_thgHbnTW<d2!<0t`pn8CByLK%GDZ{Gas{f
z_Y{>03&8d80fv;M8|?%0xjm@}SOmVpIfW7i&v~k=bNk1X-#QPzp|a~1VFxc)s(|-k
z9)F+3(dAh_2j1G4MC?%S&59oGWo`!G|DEfVSM2Px3EtpuwgBR=6ya%E*gFjSvxRKx
zyLril?+pLof05p9Q`Mru$FM5$4x{`uRpQ-Ny=ckp!vX{<>+gW-H&uZi-lNhnQ0Oeq
zd|Ip1YsSH0#O@;zrz`r$>Z?>hMItbdso;%wcTZKzIm7)KXVgOxkqcPZJ^0-}U2%}A
zJ{(UNQ$6ofp||6B;YCg8j`7CJ4^qx5<1+Zt^i`|0H6V?^h0QQ;S>%*_j?h+3IG5{<
zmnFn})nR@!0r$N|K<D4<-_xCs4LF!LK&mejxPfS;kqgbN4^>eCefnCjieT5#C31Qi
z045NmXwV_DLMAA{@=EjX)Q@eUPz@JnpbFm<0ouM6wmg^pR>=PP?Z1y0aCcT(x*hXw
ze^l8;72@Go>ysB$e-ZO$eu_O~xL-pIF{lKAjU>gE$f(0onNbBXaNLih;=YeujOO!L
z(fvVjTgaucD_}>cyk8AK6#)Re6pu&;CkS}}nL_xJ0Xu~EUxU$+Pf>mfl-y`%s@iVs
zxcZU+WkSajo}xGJ<~hP8zbpO2ls=7Wc5z|S=}UKmhmW!n`<I<H;0Oq;kV1`LfLUwz
zQmkjUk|RJQ2vvGlq!`>)xpQTE=+VCeB(bJ6$35hp@DQjk;U1JnckO1EWFh=le1aq>
zvAYz4%77pS|FvlcS6$PtR*Wi?;;{d{mg9c(`eURe!I<*|I5QJH50XcUF(ZFTga1)V
z<L6A=m=cDm9<Um>2X*WJj9?D!74AX@UZN%h=^IDaKJt_(?+~<0)Fm9SQ)Agaa~HCD
zfp~Eu?pCYLMb^l2Jw?S!P?ZgULn&+4Fq(gu^bO1bD=?F>Ks3L1jAF4sx1aacc)0p#
zrKw6?o82`d<RkGvLo6KxyrUd%f}NSlB82Aqu0kZ6`@{XyFNO9NcSAMXL_uk*KSY=l
z<EtK*;=JU0qH2G7@N&Zh#j7U7Co~4LEZ(T*UgKr-!0~|Du}eNSCu6#X>q}ZIo@F}=
zdI&8atpYi&Y}S!FiuVgJ4sMJ;6yN(Bq<w(=MS?U*^+9CWfZ04nm)3HhmgRWE5$*oE
z3bYgkiAY*uJbF0!cAoMdQii=dC>nl*ji+uXX7kH_Gx|UF?kXy-U|Se<a0%}24#6e3
zLvVM8;0{58ySux)yF0<%-QC^ocDDTI{^RV2eIM@_^iz*!G+nD!^<1;6XMKx-Z<;&i
zHhNaZ!zGqGZxrrdM{gvlrM63BtUhxs-}HRDPlqPRtL`+mdHeS=|5rp&{92o4D&;E9
zKUc{1&!f1^Qr86xbz8U4g#}>vh`{HW)PFV>uw`@9sDQ;N&S(2}KY?!^x^N###@bHF
z&E;R81z!@nG+I|+n@>-J%iVsYz1qxMU>Y-%Dxch6NA|B%#hneLVaz&%H74QB&CK<5
zBt`TG98M>9<ZpxroUFLX5K%zsuC@E`IuTbJz<YmA1kW0AWUl{>NPiboj}48O%8N@;
zlAQkMQ_-I!&4txwQcc5C<T281TZ%K6{^wwple;C&ODMq3xZ3|_Cx1nqqpn}Z>Zyj5
zn!jCUvK{&;$0rI!jMLu7+Qmwxt^Y=(zvGmHel2;vKm%G2tehl!1?kW8QvQ>u`8^$)
zFR!{orH$1(&LBs^Dvo^LIR;7&piu7b0JzM;s&;rXJ7y}EDn&_IZgN_(Q(<R2g)vq$
zY$&LDe$=M~hWl3=Oode>w(tfT!Jjqa&$}~mjvriS33~q3(NX(h<8}LQsdn6|s@)wM
z1(_qgrZWHRK)_|JfM;fTR$>cTBSGm0uKqt469H9eK9u8itbOv3GL|04FdAW{-v2&p
z3d?u1M(U~Jl!HHs=`u~>#q|kG{{1)KA0ezQ=aS|Mlwg6E&kq(Ui2weaA~i|Fg8GvL
zA1^wO#R9taERd^zp98D`2V-?W?u>I;^0wXq6#ai*4{6?con72vq6TxEnB;C)P3h<H
zL;UB7g*T&!ov35|O<Cqtv34ASeemzWX2dikX&xIVpyYm`sUa<s`330z{JSv{Ku5|$
z>#VThJ(LUI{;Q7(Kx`gffA=KMS13Va`s2c{+(JF-|06^AyL!y{McaDDkEvX%6lJ}^
z)Hbw-{_nN+FW)qXBLLFCp~*6{?3cfZyYV=k42Ov}@i(3RJFX<nj*!Rd933cw`liV(
zS_~M~5gGpdM*4~j)3amwD#)^z)ds}se=aij_>S%z0~$dAw(fNZwDKNvZ1s=p@IM+6
z^Llf?1L&h16vy<^r!mA(z&3(^4l6Q){l{Ycg8Hlc$>s6Zm4ubqe&L{hADJ&E@67U}
zysB-3_5?_6(LWcYxA^ySlT4-Lp#4!#Uo4-Sxd?qg{xyg&BlQbdgel$hM`~cY<Nsda
z{vK%2w?Sj|*hGOpPy2=ARek!DJ^x(D@q?%b1$<M&m5vyf21ksSL1I;dp#BvGW~l(3
zv&`=4&YJVy3AiswiN`;i3rKZ}&mza_?(+@1o?aaRRl$A`h!D;;A(pJ{>`NmfrOnNl
zfGDqha2ugk!H@Hg814sAI5?dzq2qSdr{no{2auTn(*{6v#*G918}poPj(`6Ax$61x
z#)TJY!O6+l38-lJ*LXxQiG+AKUwav$>!g5t48XQBF*W^XSnl!Vli+Hy+Y{&J=I-kS
zB;=uwRy`^Od3k>`OhA%cQIYVD2^UBVb_|}M<p{`$BL@U=W<&q4Jh=XcJzxUz{QMje
z2w<G7)zz2%D#L<<n=d+E=EH~y3Hcw5!@|OzUtawA+@pR@0@N^DS-M^FobYg103VhE
zc;x0HLIBb=ODzQjrn?1XRQU^jz#g!>lZBDSKK(1erZ`0v6}Fw5oR9a1PBj`dX|N(h
z)>J8vQGk6#%Ejf@?E@C%>q6Lh2@3;*0C9KybdV`3BsBR-ASfV^;c`rwEJ-%;7&KUs
zZ}6{E37K@jA1rizJwivP(J(qZd>k!`S6yA*JWu_vuMdH41_YEQ?ZWJ=qNpf26_r^`
z;=e~t*AIB<=)=Rq&X2c?NSWa=whhUc;Q{Yw=i4b!>y{%jzyyKoantt$aI+pjTkICJ
zoYveQ)&T8m2H;;m{_e*ch+N4dudlC%jg?nZTlxaPy#ST!koABmOv)SrC@@fr$(R51
zI2;`u42q15jEER7Rw$(33>>aH*xkk9cKHRUHlU@Y4OzNtZD~0jkkilzRV|L_?(V*t
zqNSpe=ZeqFtSl{c1yEzFDy3#MwY6s-VfraFeRxhzPN3pPK|(pJtD45f2US&7S1$k^
zYe_7RkBAN}6!qog<OI;A_oqut9r^Si)6<#m_iEVKsQ@3#<9dj^2zz!C;Is`AN1iU%
zhrYf#13a(|&ky&fI_Ko%WTy@HhYSf66chqws&D=pR?WhQXS=&UpWXq*EPy%d&Yxi9
zJg@!n^p*AX2tad-iYSB;-5$3B0UP$h+&;GZLd6P6He@^)h-AeN7OM^Z)^d2h-(<q`
zJnFu_1BOQD(|)QDbe_)hQDG)Z_+CZpIgz2Bg@vR##1rO3()Y`!SI&zzh4rPF4e-1n
z9rG))EKL%F!A1yhaI5w^<tG`MZGK_iFBF%X03NPl{R1$$#%H$^>S(uEViXh<{1bxp
zBhW%$AFzTjB#wYQM>x?cHqEx~i;Ih!$7yD6PP_sru=0Fa*}%fW8f0bI1XQ4v4Q2pD
zx8bDkQQC=3GFO*?G9A(i(Yq`R`6a8iE0HHV!aO0sl+D=6%F5VyVR12NvuN2jFF7`D
zb$PiZCr6+<^Y#1a*jT)ECbw$>31tu}Sk(JHU>7UdTUC#5=j(N7aj~h5O_lom?G#eQ
zrLL~-W++`pk@Mp?=S5RXOH6p<hR4d-n3QeLE?_EcAPG9C;&I-5fCA5Ewc2F4!m0=p
zqSkD;2dI3a+aPOZe`2Ze{CEeHMFr<kMMp;mMIICIx}XDCx}0{GJ2NCpbaeC@1p)+s
zBj+ini3TWXaR+oXA~p14qc#4}A5?h2+_l-^L^&pa>*Mw2qVwbZ@v9P_10I+C{^$U{
zXPI-;&!HHDfr#kvzdJb@q8%(P&tTbByONLyEnkZZGvv86TOG?QDrl*|<1sKW+HR*M
zNJ!hBi(|=UB0luM6+<E;|ITQc7#RiB*-Upr=C~>!f5RyH2@xf;S>qi=S9pZHGX4g9
z<+!=H5J&-g1p0b~L(u@kE;^h%!&!hKtpHF91ej3AWGWked3t&RbWZDbS}(ee8(;&j
zcL+bGXe1tKr^AxPz2C(C?(VANie(TIOI<@l40BZPR@B+cc`F?$sc^@}^z?P@1dpJb
zqLj1q;~Sv&%29SwtveV14U40=B%p-M;p*#E(f-{xGIG+hn6Ci{yiengkB^>jCpuxl
zUl<6UIIjCxepquEkK@417Q$S|`jYIZgbAZS`-fP%yT3KAyL|PQ$KA0;sj0614H*A>
zE9UcMOMt}^|LPSDp4Ur3YHI2~;&TWyXY1`6S5doIMA1Jw)@EjIx<KHg{@}>vbol}v
zPOG+oZkIo$q{cP?L2`XvyT;PYjKlNYZJKECzQy5$)pGey9)Yuq%ew1Dd#Q3ck^6PR
zHw{_pB5y_d827)yzEVek$0YMt=LKMq2QZpn&s#5IG6w)VXoKR9zq>tNnpd`q<h%ij
z(N-N6v_eX&0eg<~cC=c49RW_aqz61a3=H6l?*9textE{;;1*F`ow?Rmo9)r*v|o23
zd1K&?3o|`p9=iB~^YZedG7bO{mk<+SX|dKg((MBf&+yzXyXQ((Vrj$}Y*I${IsoNV
zfWdk6H;POO!ZTcSs81h@?E{u`f<yQX$znc?ca3)kBdY;bpYk|YS64r=z2zxMUZf=?
zh^HcggO4sQ@IO!};XFM(^+kl0ffXT+3_?RgD;L!?G_X-n00g0mwsr(UF|hp5fD`}q
z3-NATLPCtAoT{p`uqzZiJUk>M!Jhx&|5r&95Vo@{UwJ<^=mFb$t$y17vNP1P{QJyW
z-wc|Wk&T6bp5X6OF0Ox*p$sg{EdLeCU9UD~wfY^VV?yB(EGG!*4gwU2F;=p=y*?Fy
zayzYzfQR1KRCeY`K8e<l-;;1V?^c_H<nVapn5?}W{(9K5apPId_~ytj4AK2-VuLda
zPO3BODBDTvi|o3<!|Swr^UShyOT@42`T1$oLmv5)b88x*m@mg*O`TAXPVH&zGk1K9
z0egq<Bp1;n#gp?htwZDq5YzC{^{bxl@6MC@Man}r-`MUt3lDlH1T9$$irnnh8WTaB
zvcL>;j|k+=TXZ~~HsCYO%6nB)wB;f^56>4OD_pMCa1xo>rCl#pM{NUW2y|RB<Mt}n
zZGXht@mVHGo!;YY<)ERXJ{6eQ?$U|U{!yqLi^jfM-7vRJhE6yIeHz-I$cTFCrNC31
z&c+kIAtE(1>0UK?2w!Lu`XR9^?LJ~)u#0EEFk(RU@yzr4h2Nq)kwqh5sZ~K^Bz6K{
z`OQL8zB|AF-twIVFP_YJ)*64|G@--j{Py+Fg4KqG_SXCtHYQJm29cwn@L#+ZAXqTa
zmK5?RI_<VgrF-PqY_5!80e2{LvgV8~QyhGX3@p7o!4r$T2SD7B$ky$6CZ%#{IpY0u
z2KxNvD`<v8#eNO**DGbL{l&Uf)f}QXtgFjK^Z`bC(mpd3uU>B->*&AE0==NL{U^gf
zs~~{{|CCsQB=x8yV+u@=u?O_{J9AVrjkT8fz!tRyOO*1R)*GT)z@P+7nk}9gjzC(t
zxoO`rnB_HMoA#3(8xm#F_ai`v(FJ+E`4JHEsHk3fEA!lPLG<Q~W=NuPNYI9bivYuL
z01b;Jq<dQy>@#J7Z35rh3Tg%zL$y7b)ELgklITI0c2y;6$;#YAL{y6M7aqMp_haVM
z7-ZfLq(xAf=x6|{SYv^ElO-%^Q;S@mEyX0guE~(>vyy%v^q&9J-pS1=oxgrbO1$O=
zgf<t>8}&&lpJEnO-fC;nRqRqtlQzhMIjQ;U1PkoQvkB#jy`tf1j@!l)Z0EA+AnG#F
z<hDjw`SFo3h@#cfr|5<dyIDOb2TxXpLTZy}jPVxronbPyM?+f8N79)LG|2#WO{V@4
zY%-=@v{W*KNeYggx#@zMMGK-c#L81Jn%Z914EX?N@f>a$IH4<7k~kmO#IWl)q=Y_O
zB<+~C-Jjr8hn+mf^w6<Ez2kB%yOf!o*EXtkpd8v5Y+&-$EUi@}-g;cj3GMYHp?7eq
zZC!!X7se#DumPdV^d&{Toc$X4S9WbmJqtu}d)G`JA8FCS!eOnuO~e#hx4!vkJNlb}
zL$<AcUd}Ao*`fx+E&Ctt9Esu4D5iC`-Ce^~z$80LVW#1%p(v9)^dbS#w>*kyJkT{Z
z@cw(4W#JM49F!jS`-B8I)j=?qbBwXbc^~@HnC(zjw?n8iEu>ZnGEhCeY2t}&1c8Z#
zR$$H=@0U0Z#VLA5@zGx479Cm4w3<V;y~N6vmrxGg7<%%V55uWMJ&d)tnYSVT3KbLu
zt2k6_Fj{T-GEQ-JD2n!FIsaP9LX}n}i@H%?pE%l5(mE2AlQI7m7PLLC`53#=740pg
z3LXwq%4M$@sw(+pt|4w6Jv;X3k#b$5v`H9D-_nYV8R`zZIJMqXG~2>}@G8>s-AQv@
zje&uH8kUrzLE_#7Vl~@+$c@C3Zak?a$B`KuHls<+g;R-TX;l`5)yH4KUzj#vkeK8<
zCzeOltep5YX2usp=gYFrf@9wHeD3=&4k^)CPHbmtqza^%jF9i(`V`F~YiI2TC+rxS
zvNxzBdQ${QX6MUE^W4g8WF<b%iCc3Gf~~lk9DgR}IGkCt*G`BJg>I1Lkp|6v5z`zN
z?H`x`v81JRf@i~IeQV`~&zMmFkFJ~j{W1{BvbNJ1Tl6xk>Z#D*f(fyL5c;Qbe)G1A
z>*w^$1g1yK78j%ysGPg5v^Y%cm(pJK$xucgQ1S%LQUbl%sePjS13j*=t!l14I<+)`
z>jk<S#naHzKhX}E<Z2H|^->$0bGMA+ltJGN7@Sx|Fj2A72Shdmf}JH(3-B<<5eKF?
zQ47dbU-1QOYyWXE`Cpe6#{apjd<xh9Y3lSTTz?AJpThO0aQ!J<e+t*1!ZrM-aQ!J<
ze+t*1!u6+c{V8033fG^)^`~$R_<{dZxc(HbKZR?E%1`0?Q@H*Vu0MtAPvQDgxc(Hb
zKZWa0;rdg!{uHi1h3ik@`ct_66s|vo>rdhOQ@H*Vu0MtAPvQDgxc(HbKZWa0;rdg!
z{uHi1h3ik@`ct_66s|vo>rdhOQ@H*Vu0MtAPvQDgxc(HbKZWa0;rdg!{uHi1h3ik@
z`ct_66s|vo>rdhOQ@H*Vu0MtAPvQDgxc)yR^`~(C52hWS@l&||6s|vo>rdhOQ@H*V
zu0MtAPvQFiZ^AYE|4^iohE~SEOb8g6zSIAA!nL|YEO8@j$3*2JZxu8CW~q37iQGj*
zshdO%TcV8Esl+{k&h-Nb51)@k19OTNJ~MH(*q1*$#9ta#-xqi9Z${2eMr5qPS?PR~
zw5@eKH&A$@9#zC}emUNVLmAKB->uxA2#cRx+V{f$Fop5p@_>Kb(B{QNwF&4c6xZVY
zEt|+F)46qkdT_`L-+AGHJ1&GY{aq7VOY#PFTIGk#*}_K0`!!K%mb_46M}g-}C7q7W
z`wL^}SsLPW;92H|_B05+cv2SIo)ELu%gx5h1)XPVv0Xl>zj*Yx8YhLY*enXc9fhQe
zVw}D$lUHg_yBOaV5F2!pY1i1w#A9(PqSw#mOx};hg6PhV)d#(BdWO-oI-J>QBTmws
zTZm@a<c48=gG9>&m35=vCtPX^g-=^4r?CZM%j4p%&>z(Io|d?VV5+<tqS)T&BJ+gM
zYY+lP=$9`(dDTQlt{=~-tMC339^K7i{vRVuoh-C=FBA8XKi?pJfh_Y*&Gm@7M^>#e
z*F?lykn2gIAZ_1}tWzc{6JgkA(r{SRF|D=YRroc42Ww>i`s!*ZFS`T5E%fw^y(TDp
zzcP1W`dU9;KjWsNRcLKt6ID<He{tqLu807%*;@8Q6YTU%7vrGIyWM4<)dn(WS=H0B
zs7vbYEe?5(Y%zFJh;l-C%jEZX@7J}X?A6OU3!$>rJ4ar@EtX5=4H;ghJG9J5yRqD&
zvG<^ET$SvuoLEqf?VUJb9yLH}>L#GI;`r4u3;P9*sb23x3XVW3<45DooCD7g8pIOS
zcVt*#eJ%|J$hTXgUNkV1Z)?*%^H7*8a<bBUC}AcVd2V6@4@#VjvZ$ogo8=iO)=D&q
zS~>LYo66vRj(n1^D;*6>-jZ^^SR0RMm5uz4^$}Lwa;nf1OYGh9x2eH2jN))7m;12R
zGTI}k=<L5NvyBu5@Dp*fTb|aQo($2Bw$@9&n)4;-JTuu(r0?GT<YX(m22qKbRvCAN
z>St<4C`?O8Ml#7u#p_>VH8k{yc;d%vR~olYGq}(2Z8Z5724mmA=x;Lj6D&)R4;C*c
z#E|nbDvaf^-zxx`c8O7?oc4AW-i;|;1>6OfLZ>w`j-a9|dC#QwTs>1FR#Znc(`t(j
zLY9=XA4=S5oEwK`xI5q2(V-o?n-mheovHMDASBTs-a*X^xwjf{GFfZ6W|rNqe9u-C
z1#w27zwwlb>#1>yJJ?NT026WJgD!@1V6*dg{l#GkX}1BDjmftkxH;ag2jfI@xa~Y#
z!T77ZM)-%1-uF*!6(sWhejz}R0>wsKX(PpVA#rPXDqQza)63Iz=fefj3MNAm<J9<r
ztQC_JkIL2dH|1Dc2k$H|vB7eF9te$!BPJPU?y3qbn2ZTIE;+MR-+vb}BzmCj+CNC#
zwY=_bjtPrQhnU^mZ=o{}Tg6%IwsQ1QLCicEO<0SOh!q-#9llx>+TGOVOh7pH#L0=*
z{S?)f#6cw?DvkxBg$ZDq=$@IT^g*xg8#dT2Q9fs%vfvG(elgU<IW;D2)X~11G`}v(
zzhkU_nKu{&KO?a`P>)S>6fb5VB*IsyM=6$198^<{FlvfO(=7NYpK3cLxnvmMSIyiZ
zJ}~Z}!h))<<M#;b(uF#tDqeARq0rc42ccS0RtRpTX!jL%@4RbJw?89n^|)H0H=cDh
zHri3MnX(z$K0aX&lG*-8`RX(MaKsE%+h`}5ig<(=wGiWp!*`)n%pSxHne60+;yE`3
zQkCb(6UDhjf+&y+pN1B>iLO9O+cY2vQTuHApR|_AC|wz_b)%q0?s??!mT}OL*C)w6
zxn+yRSa^=zeEDH9?-4N#HKiXhWs9{tR+clcvN1ol&{a*Hb3kf1!Sv%T7k$-8kJAMn
zs`J+fm8LrSOgC-4c~Eurf*F_j@?VFQmG%qv6=QB?bD-@^d2k%U2N+>tM|@x63Ogiq
z)r~r;&j-h8ggS~`msC}7NJw5#)A%atRc06J^vgzh52ul|X1Wz)Wtf=7znotqzh#Mn
z%8vziQJ5cDefwt3rwdI_!0e#s7{DuUr(x{DRAJb*bab@URL4ywfzv<Y3fB|TgQR}6
z`knrxlGrbeQ*0Gl=i0ZRSo%PA8Ujt&?mAaG9{*gJGr<WC`)+93Oz+4M4((WqV~AxI
zu`R>@wT;NV0T^%mTLlXmR7E|pGoiGpGP@;*(IjQl10E5!gBCVx8F2{us#AzOZu;c<
zo0rcEH)+-VW2RD9(V}WB7KjaI^-`DG(UnlM=r8!8Z0@`amNO9XV0b+Eh@qM1g=G2t
z4my`bL-<#c2DdfFr*Gy7Nbo_idzg$U>}*^#1G;l}Ar#_JG-|XuD>yU|^TC_OBcvd!
z_Mvrp>CXbk6&k9TQBBD~#%GCZ$NZsdXr&Z0a7;RDl@#QhN1L=X5a+~NxevEl7ARQw
zQkpSul0|G&uFdh%=xHMHS12-1Y-{2xcBSdomF^wc9fbt2(5T2p6|nfl3<Eb#4u53R
zW%#hw#zmwoS8g72A``wj&>Ul`SH0qqRE0gU8d@aJ*Ko%jYsZh=JF~0~D-Aa@=;U3u
zgVyP-Z(#x1hc+9Ux(ALzBzXJ`KMY=4M88);RIF`Z_);(AqWfijXQn;!{`4!fx}Hb&
z_7?Q&fzi}MM^+Yc^E=TOoU4-~bR?LJh~=@k`jqvVDAQso&!73<!)!tb{F(0g7-pNo
zP7<`;8T8c+J7#eSrNy$3Ge<o!Ftw{Md(@Z|#I>*9oTCqE_7=z>A>VZgpW0r3nx|I;
z<0kTIHOss%RNPV^%L4hGTYn_ImB8fhI#P5Fgvw-e?Oo_BuO`$?j0NEMzqnP_isPtK
zffs5$Epncp(V7KqSJ4>o;!Qz+5mbE>MoibfvK@W{L*$m5CojLnDxa~iPf_QhiyT07
zh&m{%zu^B~xxaLl$1dhHT`GMIQfUy4N@LdDU?BR0k7(eG#K$}$ag$LGe6n!mR&1PI
z-!bzIWGu0k6nnot4nL;S{ELSa?D=U*%c)n5-}MMw%=O52b1taNu19~tNQQ!YMH^u|
zV)ht!WOM3+H~%gRL`0kGFm_MvIL%y)z52GP)yaN57V!D`hWIxhjvq`8P13QiznTL)
zILoDZqTT&C()Vo~etaOJe8Bf=M-Ts}dyjl){ok1xI~&`7m6^3;4_OhnUR6{c;?v_l
zfYFByV!3*)k|k+G3DOTr!ZGo$j}UfdZp&+itjU|(!VJbN{;X!57r)vY=t-OiT&N7(
zSdZ7nxe@u^?Dp0H>D;x|xk+y_w(X~@9K3f@xmP)nCh@gW#o2T`-Y3y_0&jh6Bo+zJ
z2-2DA&q(adHv2^E#U)1%4(37kfr1(Y6A-8h+B`V+wj0o;uvenX)5-!bzB7MZt?tgi
zqey4FJNK(cmI?N^WVw+>oxtxE@w&q!!W08_3nTa+svY<pvPIkqZC?s(%)>F0h`!s1
z)6IhyZ0C-myan5XNKWz<fp6E14>G}y9Cz}1(EQw`%fwz6XC(aQPXngs>rkawBuuYO
z07uL%1UnIzOaMF(KH}^{({%q60xz2B3jdrwtP8HQ9in3x*F3q(ZaX5lwd+LH5!ZBi
zx88Qt2eSMMrY=_?^cCziygg$;8l(}_pMrhQmM<rtlIvV_+_sCFOf$*xhoWf5MJ^Wx
z+O1Ht3&=221#%?wVZGgVrr|juo}sc>vX0*FY3qd>ny#ydjB`HcfLc9#GN0Vq9lZsc
zMOZVR4z&PzaI{swNCf!UY{e8YQgSQ=r^p(b+rxXUipyu8vV3?NtT<aHxY6xG2RZ5y
zX8O&mGE|=JtRuK>IAp(JCV?`tLzs1VdQl~%{xyd|0Y;-tuR22@I6{9rs@HZ9UIX2f
z3_=+)yGRtatHz9O^O%wjmq6$#Z}WINpGVU*lJ6sF_V`7OAsR+utyzr@j=#3_d+;JJ
zIi!1o+eU<QzQz!^B9@Fsrf1WN2zOr0$W$b=GW}5qe0dp(-P^7)NUB0f!F)kf#)-yb
zU5rZF&0x10+C_Y$$+Eo0BWiPr6A>c%x+^`PeM+b@YNzHPvv<ibOvQ-N=0;?!6%dI6
zg2cRCOA5jD0Qv&!hVbHMhv(uV88s^4<%dK;t9AiXi4YPwJsA(D0qp$int|naJAb`B
zq)?V1Gn`xcbxMo)V3J>yaHOsp6gHpusF=OCp{rxglRtwuyE&ae$C^JB(32k+R-JK1
zYjQrRC0D`(0}yvbO|zYWU!2fIZTXLE?;_C>oEB^ppj_vW`ZM~-%&``MAKlD;Wj|Rk
z%({N?<c1PBN;@@cW$&QX%4)Lqz0RBw`m(`jry{xa{rM;xAuHW^3X2nu9o?DJ1Npg%
zGmh0a^h@2)szCLq3pCSa-y^-JC{Z7YwS;(Lo<eQ*frUZkgNNik9Z0RzUbEyC1IUpF
z<M<u<Rkb?WDy*P9<a9@hbVg!pl;tI5>ye#`u0Xf{8=oXR15<aoP^|`vSk{m6()mdD
zzDApd&A^OrPlw`f#h0t(qnUqho(aS5=q(j5+TCwX#BIYCdmx+7-$Oj=c^>F`L}+w$
z;+)S9FJW^d5(J8;@@3~ptDAM_Fk_MXO18Wa#dE%X)rKwf$%lgC)!c!8ZA3RL@x|DY
zn*C0jG{amKl=X{>46Dx84FlERV%<zi4r8x))Z&c5LiL%8xVv%#3l6Gav@Ak~eLP)6
zgU?SUfS_CldpP!7mmGSSLCkPAOCF4pTBYGIJ?QHy1^t;2c~t-}Uf&@nmJHIK*y?xa
zK$A^*Ipdz<hM6pw`Udpa<Y3p9`HrH&`oP2PRF-6F%L3wtqn+VXKZs~0ZxD$7Gj@8+
z2+IHzR9sZ{slydQzFzP&6!rbbr=d$o5nP?HQj^1jajDQ2rNf{RR)Wk^<O+t_Ya<_+
z@oLD2`e;;z3~d6#%KmkGWo3!MuU#xiTSZJy)d;Rk-n9wKkYA<Mq|D^tuBcFB#o3)z
z>|F$83@lWe)Feg@tw1!4kRPJd30;4o=a!gb<52OTfh=Jh*?h0c5?@{0ob2~1RK%a$
zB#7E`sR$kd>)VOg=9MpCu_@X!Jl{u<fdnDcP9msC>}LojtgW}+$IET#P_mI>>Y*s`
zDda4fq^xZ%L8D4D24W>e7+ZO3O6Z63Ud1BMr#`GG2}U$Y*oYcCSVoS&DtxIIq5M&!
z4r+cLNI{w>b{?Jb4KyuTbT*7y2(FOcF99a=<T^o)-ZS=Ur`iwMa^vFb=u2DPP!-<o
z8qJ5+Hhr0QGwU50iYyXQK;Xvv+s;d7$D0<t>SbmVuq63mftUR?wP$^mKWJ<1>*xk9
z#8{jjVw5w*SM9IUBIu}a3v=GqBm5ErevBR1Z<wNj7M#BIR-v)&j<)*V8zNNTh`P9a
zXCC*yL5eGca*EYK8$@w6tRDV%O;jG?mG7L={z@O3a2tueiNj-An{@TAn?hyu3iRJ{
zZjLb20|_h@s#mp}5OE?URq@Xt%Y9?i@FhGLm~FFytCtk{RP=50h%o}cpb2ylx)ttV
z!w}RNC7^JQNCxc*?fZpuP(o`knXepJcMiAUR~_3wWNdLk6?a)I?LobQ4zhKNcVI<q
zs?%x~>{!%Y48#~8;rz;jtP=KN*B|DrOs5A%k$tTBMPa0#iB+xR$OH+vHKw?GH2<uF
zW0?1)thfSZty~u$)~{3XhReIfJv++k1yX}SS{qW61eU_CALsQIoT+>9C>)goCKC&Y
zQ^N%8?yM{-f{#h~1(OfVqfA<5M<;GEg-T(8Tu@hb<x`7SK(S%UyG72puB)_$O}Jnc
zp<WW0BNno0_w??wTmcUhN>yKa9=6<pw@9S$Zy+f#<EAFZMaw;B|8S1WPiP5LVK=B3
zcjsG7wqlK0^a|3%H1&9iNolV{6cedlp|U*TR^g)*DKypS^!b-42H42uQ<01cGcj+F
zPL3U0%-_mZ?b&QNZRyWYIKyhwPHM>;N*4^{Bm{63Sd&cb!ayoa1yh005(_tK?l&3V
z4cS-<duuQthu5{)45QB4LxkBlnBh1NCheyO^q|LQ+qg$anbHy4jRP@^IY>bp4I9j#
z%o!S%K5uulMbrUHOIJ^YHC3A=l){71&A+PM;F!OcJr$fzyw``#h_61)0#DdSG>;}Q
zXqlxm1k+TdGiDBNC6~y0<O9hXPx(4@MA3btFmb{>5njC195q5~w`S=in_g=&y3*`N
z6_$KE3UEQyiC>V!IHxfj(p8&`xZH%6OEaiqL|``K_ag5tE0r5wo!W~D>}=mAHUR5u
zI+CVFC^}t3#V>YidAoZIf%oRpGHuUo)+!M~c~B?}Agx<*y+f*D_-2$=Ig_~ZY#Ads
z`0b<}IykM;J=O$m_eNaqnW|DQBL3d}Pu85eYKaoQ-kf#4(OpDtX$RrfG8QN*@P*rB
zaYRXg)7|0T<idG;ISz(R33JQ#gm0$Vg<VN|siRY&FUR~6s@)ZHz0YMaW>Ho8024L1
zO{j*s!>O~zk0m2zvf88jRQ_zzG--Z4vU0gQqv-D!(gdWpZawDoujv+?3sHv_uw~<g
z2jJ>db)sXMp?EdOMYu~?l#1h#v=%y}#X=>&W!3L9Z&t4kS?J8^%`PO)qtXqF_R&@c
z{!D7T`OfL`-0tG(2Z3~5+`%I-(LW)=dLv#!I79e6y^ZUEVt9hZLPHE(p%8D>8Hl?<
z?uZPcb5Q=g&nnM+UftZ!RFZXLeJ04CmN7!6UALPG;ahE!O1U^V+ZS!^rdXdRrGZu0
zBKCk0BG8Gs{O(F+5l^;@%bS+{s2N3{>~0C;@D?dB7@>hB>#@nd*I+tw1^x(&esmqK
z7Lv&6s*o}6A^4F-62$#JsfEeD?r4oCZ?H28R&2Mg`0gP3oCrwos8x<3|Cm@C<PknT
z**A39_pIJ>?hInCR1)0XYfB}aGTl;{Ri%(bS0wR{TbXU?9~c+TA3GB&N`Kd@3AWX3
zziQMm!_pVT?xCr7;M%^>8y#Jm8N>Ze$h@n5@9i!DedQ*aG5XH?Yj-FD{%{aE2>x^s
zdC<vm7pjhv<65!*!D$z4cVoF3!f_WQ>0`bmDP?T|<hEK86VF+mAjgvCmNVt*7MC&l
z14zU*aOpp#`|SU(dqVwdA(n}eo%O#?_i_H2?jwx31A!9eMi1~=vATV=bEeln(1rLt
z4#@V2)K){*xKmL;&gAR|_0O=`-JYs1_hx!W76ReP^U(*|5T$8#I_ZErHZTA0Y~NZ2
zknK-=<w~*7-?Y7K43xQfE%FP*1nxlHews49iaQuW=7LBf-r;(D2|Cdw^V%HlS=JRy
zqx;eUd3kuy3AnE&h^P>+hkI6#nqiWP(bN6O)&v-thx+s>#JJ+6(bEM|pdO-R!+dw)
zw1D3po8;5w1^&jz^%;D(Znx()bcYF9IHbpue`N?*wvVdEcB|pExG7pU@MbCcS#lLP
zS)9Ma=E`+ezJUPdM)-SJuD5Ar{wW%Dcbt>0{!$h;)pwyFOO%I~C3xksV>GCv5D~Xf
zSsP0?|2sn^w{V#c?Sh^XPg?km#K4L8O#W`($GV6W7Dwm;e}x=lQ_PqIZxKkXMMNAn
zBZE5fIw`|>jdCyM2^@-G^<%5}MaF4F&44e;-m(EyW<TA0f&Au2W6BZ(grlj(xMb8R
zqso)(&n~PxZPwsb?}}eWZSedRC9zw%fPFE8K=?>GSv=;XZCRy7n3MvI4L22hpD^j!
zi8n8ZJ@8^o<J}h|E&QM$-LyGg8M7n9rzg<uA<}{wo`KI%%9NXOa82BcES79^{6W*S
z9)WMMZ3O#;o%V<Vr6(e%uKr97emet~-l9Pk?bb5YSxuAkZ=GPQSm3EVB1nw(La;Sj
z=8d?iSK9Zgorw^W$CuVFo1$n3(<!73|3b@+*h}*AW66A7Vx>+cpk<sLOk21V%B*Jy
z?YO=|Q9UncC$^|mbK8(q9)<n&O3ic^9yH_|n+#}Ewx3jX4br<gW}#k{^Jz{W82<$H
zM3a(|ImeOtMFxkI?Vt*JC#B1j1jjYG*wQ}v$6Bu`Rvk4uB{OnFPXAc4a0L4#UJVtM
z8dBWQHodUc1m8@GaI|KiPc#fDcea;X4Ej~^a^`r0Pa6+Pvh(bp+xDVTH*Y1CrY4O&
z5BD3zSkwwfLz!6vwau3yW&352Dcebm$9q+$$?*LG2{nWwkOmEO1;U7qltc1HNgJKt
zN3wBXGbPc%;tDoZe`2@ZTd_12!aahA^VRwp&e~5J2|l)l6{zU<R%VFVW-@-`*{Tc}
zXNPcL>L6s!<rv#NzZj4n@Xd{;E6w35X9*aR%U!YScU5x-kZEN8D6627_Ys6q@I`>F
zKu|Ebh&N;PV*p!3l9gjZ?%GQ-tC<cT=PF)BL>+x7LGSYk_(eknMAwDuF$$LnhVAb0
z`Y5zc0Qt&p%Ju!XQCiSDr$@AcPBY*1fv=j!hr!ePWwO9oXqpg3Bbwb{hAhw?kRQAk
zOmkRi^*IYNN^(ZjfWF^RQf}Ob$A{|RU}8e07ui2%!~@PM#^yij2soye(3xIcUfU$(
zGEOO2FTcuy3YbY5T|hrVYlfiExS1WpL(`u0aH9`}GfQ$Pu$M@n<(i|#`fOw;Yry$o
z?8q33Ff`{OUk<IXVF*K#sB|72%4*A|;N%Z{#iLB5X$`<atYM-k6}IJan5%B)z~LWA
z5Jg-F5gk!`j_wclC_FSnc=>vh+%pBV?WMO|{e3}}LM)fuFPx#E>9qe1EA_=<rJ`Si
z7)2pxuiTI=yh`S=(X>XWyfEsQ7fJ0=F4ePKn}AeW*I0p(C_1p?S}72<9f}+IV>>+s
zjpB&N^c=$$orPp1)elneB2@lkBq-`<TjoD(PqVYk{KBGf!_Xu(=zW^s4MC=)xg0jq
zoD7;CjHu=Z*u%x_Sz{By$;WIk^on(DgeqDezeSs3%V!a4R+8m98iMm<C?_igzZfkm
z#HiWIh?(t=dc-IyEl?Ll(n?K%DatLgx0J*;_2Hxr{gLUW+xM!PvOY1ts$uvc)^ZH9
z;iw=(o9sj6$JfSDD~GN?n{GfT*%oeES}lew=tykx66vLFtz?otoDkFeL)Q#{7gLm`
zb7p0-VQ?eUMf`-T&Mhr=Bu;Z3U%U$IW@>jAxq#(WjayylNn5`ITRlpmdnVa|lO_r-
zy_ct)(*)ghW-KFQK^CJcE{99jxVh+;Ol-+3!X0dEI!t!OVnxwcJmcXquF?RP1p^I`
zMW|+_ukCm{BKaKVq!b`5bv37oq-6q~VSy~D1YC+TmVx^#fxnZUf3CSo4*;JyRIvhG
z%d<4?NQ+6Zx>{Ce|GW)Us0MoCVv3sC(6BVwOOn6GErA72R>V%bYMRXxp%%aq>G<Qu
zajCRgEs))m>D2d4VPiAohGr1+S_0jTNKq`DW&!*8*+o!5YIBza;tC7DhXETp_DdhA
z1IE%;l48mbMq7|%#GEYcR>Uk?&j5*9S&h2LQ4-6Q>L12Eb$6pyud~|bt3+%w;Xc2w
zC>%vX&;bILvta>X+*^t#y_~<QWvO9ia!3sd6t(ly?V92AStJQM4BoWr4OVS)-xG$J
za3#P6uC(9Yz#GO{K;SjL)fCJpUZJ$7Lq|Q@*fwkN@~Y#xd(@oQw{R&1Jtzco!-j0U
z;Yk(j3`1QH4~;8ffhZ9>;{eAoi!JYV;RHqGbfpHziSWzgY)gk65_B?tYwOU<B8r-*
z>pMJsunvvJ1w0!(t>)kp?jT7^w~~d9W9m+97_aa$VF5|x;dg5X$uYi^L|ALt=YSGR
z$g6E+MmZLzKB;!<2=<X_v9xOqh=_uEiw%gVMOD5`7%5a6UlUvTrkA8|WR)N}cdw;>
zA?8z(#!PZNFIirvU%dHC$pno^;*nKQ<`fHytv9A#QdNv0#RR;CJ`7`3596r$khxY`
zHfU8F`Bc)0fzVuPB2_*5iSkuCP5H`5QVK=M^U|vJ{cga2QNBYRJv?Ng?9fU>S>7f#
zf?Z32|L!Uv7q}M`9z|YUu&*IAVn9fGS}CP6J`f#`X7Gzg&yiTJV62st5Ey$3_g>vM
z<;|M;-s%q`e-cZ39RA?GiU1TrQRGU)-QTx=5JlIVf%3IL)p`wAwMNsEfa^IAyAaE+
zt}OS2YHynmXvsyPTKLY6&L<3P>(UPKP6)e<Hk7GsfLvncN#8NQDJ11Be1CA6P%77-
zCV=P3=PYvO)u~=e-8~w|=_9!(Q?l9J4prPrQ&eCpHv5?(BlA@<TYUDZfHQdEEgyDP
zTmimP+F(D$Zc6Yxa@*>KyMbG3D7hxxxq8DIqAUp^mV(z&fCJ`~03C+$D=in>w6Z`=
zp_ObJH<F|xFg+~vof|GQ<uCEuTW5o^v^cBP9AZm_wMHrb*5lXe&MDu?(rk;2D=`p4
zvO&k0uZo;ps08O?MTwA`hN*s*p%_1SVSa|r#|;fsdwl;vmYC=$!%U##K%gctSWk{6
z?~uN~-MjrPc-%S9DYk#g&5{o`)34wMNv+)yyPoJBpIUzY{=OB0@#C=fcdzB2<gRzt
zgN%guKl{tlc3j=t<CD^iv#eJUSMA|HDY=oKxv_>oualqC#yvis(@rT&ayhjg(XUvY
zOuxbs>+EA?Ojnu0m-b(xeRC}Cj;T+|Q{Udn&N$a<CALiK*U0V~YP8rks7(~`7$Ec5
zns3_CV7s!P3AeDbgvZyQ0)@6T)y1T2Hp~LHUhPd}kb{a4;GO9WbmQb8tvURrKKjFA
z7TdHJyYAA^Oj!nDqt-vu!PlZ`uA2?~F^Ak|qu4hxfVR!(tWg6S_O+A?(#35n>z(a@
za6?}0E0|Kt&5R7|wkb2bRz1uHOfo9IRsXq8cL*7j;Ll3xa|cNHf-Nxrs2LoF9m7y4
zbSe|e)Pp?`pO)czY_00|Ub!XOoIR9)1Mz&m2CZ<q))!Dqh^aS{-1Uv)Ew*>)1qGH>
znWFFKlhLHg)~4L0*&zH@o`ml#raR_rddF$f5COlUGC%x#^X#YT=?d}5PtkE33;na)
z#m!PskGr`1f?aC`x|LeS?(WqHy=g86TSD*&5tI_$FDYAdxHZL(s@Gpy>l>x7wptH(
ze%-s`p&x+g*6b)(b*}~iW*8X~#YLg|F_X9l^5RKF*T{bsI0?Y3dV(de0WVM)t>#s;
zX*t0k!Rbeh2C*ZD!V1p0_mh{E5{f(_PPRNu^x3nkcH$Y2UO<h5U%>^r{t6y&i2uXM
zM(i~|b9#}e99epxVa~SLaJ%VF@Rg`(!#3*(gvmw9qhZH)nJbMhgbi5sQ9;bqs`mS3
z(Uk{f_i+3xUPkX(0TN3%(da^YZOFJ9D3z&QZfj?V%mueWqH7R{(_7V_0sP<qSdh-W
z$KWrkK}aSCJG<$jvB_<ma2i*#R;#l9SAUH6xb9Zx+7PpzoOuf{tVFxY%R~~lD@tBP
zbQ3mYi4S@uiPjcc=SbhfJZ_pizK!H&MhU&c-}2_^|Hs+V_y7Il$amKNu54uuYbXM@
z`Ru(T$0LhHy$F^bYi8}9DJm6Y!v_bL_#Wq!fQfzgK~RiDLSCu8&l2JXOq@<)2odFZ
zPF^BisP=kh25*K3JEv-n__|Q5+iM%fYu1%#t}Xo|DORoz)5w7Jh|ao+B?!;Nqu2cS
z7V_1FU+4Q(cd{P>cqi7)BSLqZsm*=Z1ai+JaG+=BB$uBM$}(75svzwp_*H6Hr046&
z0B<@oLx~dH-o!OiXZwru^BZ$5ZWjge==IS`vD+8pI0;8vC>7`td`}&oOrB9yf3!`M
zr;A9s5$0=SBuz$&@3=k>zMEkxA_t^?Wtva3pvFCQStmlQ81C{E<lRZyk%&3K=wJqI
zB)_nPZ6|u}5#y?hAzvQAfHvq*u@IZmUav3m;iz{m@pD3dpvh(Z5`gxfnxXr6y?$Dd
z@nnK7d@J?J>5sc3E8N3w=TJ<$vwaf3;Lm0;+~qf9CQv2V3i$%$oCD*gF0D%}|4@wm
zrY=(J13K>*Tblj{co4LOa6emDQ#5#e!9%tQ2`S!GFSiYqhPknA+bu&$dt|1QiDLZO
zlHFZf8vjQLj~@ae3F!&fdqC2c`#dw0q@avdv4lVi*~4ORRVsF~y4ANSk9O)J0VSVa
zR5H~1?oPjyvSiT~hZ}@#eHbp1GN~O1yQKLLV*Nbr--@{)sEDvBMTAGJP?esZ&muP+
zn`IAc*1b2nWeS_U*9goa(W(X#;UGL`;IhERCfC(Sk6%EtoaPpihBKU`Afr|T^9>5i
zPxQ1F15^FI8P*Ii+Gt^n%&9>$xy<@Q5a4Pcy_uMeqhAlX-d%n%=rD=WSgVUz41q;G
zN}winZ_pqxg3$Lmqm@|NJkxv!Q}Okq&>2oME|6zTbPvr1J728=x<n6xW%C1<J1P(<
z-Cnw!1TGfPO(dDkwl}iWFj;?__BRRih}ETLFF?W)z6e&4NfTBtEt}DIWT81B(p%uK
z(aa3!n&8SBpfuPX^<@CLjTmz5uQskJePW4y8({R2Q~!N<4b7=80K?GKF5YJ_2sa;`
zn@^T+6@Ex9oGJ{gWRPL8Eh$s@LZ$PK$wkAISjLh^Lo@KLSUqB9?;??^xEG;2&0w}p
zdAiC~-)t@hlFq!z>}A!$uC>9ec2dK>yeiE`ln2~`+<5!a4)xGMxwsUV{G+~jnx|GX
z)u~4zR`o^wETXO2Zd0XLr9$9FBXuiaZaI#;9U=bg8M0w)zfp|?6xjp}Sfv@|RJi5D
zS<9V-cHCU3=-1Z*r#7OGfGgGsl5=)-m{pag#0=sFh6W>TjaaUcZ1tUy=(x7{#zGde
zIunK&vqwtqaK|s0p1E;ZL?2im2$2lfYxRc;#3s5y{MP+=N97)f?JP~y9U6OQtEq%4
z^aZyj0^Qv>3e3AtmKA%bZIEMy@!IwJY+}uN5Z8Gv-xi%?Kxn+W$B|B(+I9oSeeEz-
z3YJLcP|vBsHA12Hf9o*uA)#3=;$p|P0iSTZ(dYJmZ^%3dJqf@#VKv!aMf8l9R%fTa
zg`KXo(&|Oq&natbq$+HSj$dj#3}T@PQ!^-HHaX*h(~96`i!^tvy*gBO@|(LkF%ZL7
zz7S!50)AZ}lb@T?@-#(R(!pB&_73b4A)_p^-J)HLx39WNVH;;^AJ}N7l)#RYqv<QH
zB^AitbKjh=hq1A*DS%ZnUBN*5pk$C#(%(T=cHd!*NtlC68tkw}gFi1m6*_lraHT3;
zY57HB83_A1q|uf?SD#5QU-=evz4{^_+<diGt6t<g`9!8jY?x$xbs~i<!7B=XWR>Q)
zFUpHKZ5$@uw-|)}aq{IslZ(3Rcx5aSmLIkKAwDN!RZj0e=;_btO!6TRCqw<i&^n|0
zEB3YI<r%apJgPWK?@(@Cp|adz7Kb>~Lpfa>X)TDEx|LzMoK!PJM(hZp1uttef>KUs
zSjt=@8Uiv3w-5aE$Xi%w&Et=^`EgDTv48vO`J-@0kW}}28XNl+RvxK%RL=i;dDW@~
z!=`_cK+)AHy&|@}kElEOnYv{Hsl|0jG%hVT{!~O?{Wcnq_CR$MWSW3whZnw<c!tha
zE)eX>0zV^!iGgLu$icijKQt|m47`SLbLqJdWuN4t-+%ToX}E&_hHp6-zveQ|n@0$N
z6&HNeAAlq1=_FV|+_<o&ihi$SIoKS9N;Zm(ZKrxLPn0|o!xS%47fqqDkll=Mt%#7d
zxL#%%f;EWKD~1)+Lj6Ghi(|7v8)L+<2h*`3JIp_}6LGnjTJxtwm^uG2q*$O#hVOkj
zb&$LjbUtNMXGwWtwqrlplLS!C(N>*caT{5yDcEYrs`VOJrG-zjG<(ub^`>pQ@@r8%
z0hZcIRl9Z|=)k?u{J_bPxV5xhjtntWdwz3ARfdAL&f)~AXs^Ak0^QiU5B1l+rE|&E
z#kFsi<KuhUGyA7Cclnon&A9d{@V`ya?!h)@!ErRQ(ltHYVka>JZRISwh~Vj2F@81f
zF<jT1m%@-T+H4I+DOCn*;GU@SspDl}u~juh3@i6G(NZBYHwMmlz1fsXY9v>%z_^wp
zsKv^1M|t|D?!iNh(R0WzM5k0vS_XF~VW(<LrZnW@oYoaD+`cp%#HMIqhnwe?MfcxX
zF>G*E$PSsW#E9HAg<F|A@3eZvbIY2y@h$8x+|;p{F$IVe%Z|JybcW`%L8a&mPUKsB
zm&E$^4ecnYF%JE_El*dn$H&fItXpUlzgx34KTASy)7*h?;H?*OLw?ZZBf+PFjnG1p
z%7lISYDF-X2OMwxw+#?G8&ktjOKMm}t$X|df|3`*!l6~@aelsiymc4M*=ty}rX;26
zNPfJF7-`u2gqw5OE&TwVhNs3A?gOK01hG+T<_@e^&nyp6Yx-t(COo<Wy6X}_KC?m^
z(jN4gxv|6kg?^Xm@eJJyFhpez>>DkX^MeIvBX4A3t>K+3VSG^>?y@{<Z~i(N6bx~1
zALe#mS_=e}TfIhl>+O{Q<<D;__ozvzRJQ$&gLL4<S+HPF3pSB=64R!e8Lsly0|aU;
zv5_j9TBekv7?qV>M=XPNa-0?By|vX;{2k@KM8=>_YHXE;dCcrOlt=7FYBJa7qO@RD
zvRII|1)G{q^V`;6H8D(>kW%xnBP_RHeoN*DjbKv_vQJvRR$oM_-agH}W_!dBVPh`0
zW%$`b)UOk_h3jZwlVh(X4tK?uSFJaS(Rj|cE3rH#qJ{XyH@vj1pitJpKStJkNj>{=
zN}HL!)(4~U2T(}%T*=gI_th4y&|tjv2W?^1t(DUsH>nw2B~T{BnE`Kz@*jrA9s#<=
z&q)*(q^_=B){absZCf;V870^1yGcbd<f|%og74f0zkfSCnK+pG2YlPV>zKz?F^wb+
zmP25br@vfFbSVF(G|`Xnd*(_DL!)lxGAMEdo^g*=okk;;aRyIG*A9Klw0y72k(_#U
ze!Jwfxe_Xb%!vGFOa6?09^ZqC*$-Hi(hgX#%((kYrq$J*{hrq39=SP5mL5=UvN$V;
zR;a>KI32*+(-1PbIr8Q6eThGDm-Q07Xm;C#+j*QreI9oBo|SMAari5Ni{(SSo|b&n
zwh@?0sNr(dZhh=7Wqi2N69xI@9HVE@`kx$Irr^j!DYuLBpiZ%1iotWxJ%k(Wjcs*g
z2H1#kD~iBoG6(L)`Rv#V8P7k|Hmiu6BhJ6FoenYcIcZye;Sz9a!zWmKZ){sV1kttn
zjTt7sZ`;*VZfJr2?(mTM-7ekk>c*uNlU$+fl=#zSyhVbMQ>BDX(+`nJ_2u9L2<$Xm
z`afjeO!RF3JM(6xXaBE$r>NJEuqI}Q>v&O_#Sc=S)!@^$*p2`>Er)1^-LHadaD`9r
z=7PI<NOX~{EV(A8g|Hb`7DIUo&Z7|Do{0>Wscb!HTuaPVj5U<@MV(oBfVu=5_~@#h
zJRAb1C)RU$N%C;t)x>z^+7Y#-CH0{3aC)45fq;TD|Iz`r$g2t#S)B2G`vB@BNEW^m
z-;T2^n52n$aUL807-Sj(KjUR%{k?rBd5TW-jBKE;QH0jkv`qg;Y*XA|w1_yg>$KMd
z`6)lxIkRI?6nXiK0`ASg-T`6#Y+D_l^qhT>ocOPj!i;Hb%aK&socox-i$IEQcDL~P
z(ITJ9I#+rgX7m?`En+#Oet#s-e1zsc1={c)f>y37iVm`4TZEoYgW?UybO}i5jF*YG
zJ`aoHeB{2{hJ3VLv39d|biE=2^K~wV_X@FVl7TStxqA#lIE*Ca8iGnR%<RPW*+}9u
zih-~XfsF&8$XUqd0QBe=uW80o^_;|FzWtH<v$)6*1_e;kWZM7~pgBm`;c=1RQJQP~
zq9u-V6GySVNmc1$991LBBe95<Y;`x_wMTICFO|7j4!-SYl4nTu3+qfrRA1Fd)mgi^
z$WQ|Y?8_AWQ)**pIpn_jGR#LP;&`X)Mt<Q_P9~>fXZzEvg}Y5B3L!xNNtYa@|3!}{
zhdesYD|`z3UI-~t7%Oq|1cs!YFi*>Y*w2!FF6ei5tRWGUMr2ad)lVx|V*HuIKT}9g
zzAayMVnvPDK@0KlDtjvv8`m<KS`jD;ckEm~7X8u1fkUXtOlg>a3p(F9-8bB^hZ=4)
za6qgkryN5daTt_h8v@i$11?{dzgG#)2L7q<G1}YE;VsDCO<WjmLV@war2Kkl#60e7
zQqs$tLFq`^>8CbvBfO3}z^?ZHF!oMCq6FL4ZrirISKGF2+qP}nwr$&7ZQHhO{p<Yq
z-iRA<_QQUuhs?@)%BcKe%rR!h@tnO^ug(@wiDL})hhz5jY@TbQ8I++qP3!=7<v#G#
zW4qgxvSzZ|Yqmr1MJe`YpPmY{2=g*5Fic_;{?&c+(qB`;qtD)-UybFJHKuZ{u_M5D
z=h2g8toI#3#~)d7$C52T9<DuFTJwxn$-(xD=g^M)x{Ml8wjZU=Gbq<<T*^CD@Q2{F
z1PLfjom7^En66Hzfalwp$-SG}#0(x(3zJn6fmx(GQu8u>wmy@wr=b&1pvwZI<)JU_
z2(x%fp3M-|<LZo<>occkV3_I7#Mf@cfM)W2W8^dD_p#|fO7IyJmf-k=17K~E4=5<}
z$1F}%cQXc)QBL%W0xh4d$U&8cW_TGvq{cHGOA~2Pnqj&*#$B?__ZGbl-N1%pnpPKr
ze{^VFLpcI5NxhatL-9zKucgCiUbmK=wd8`+hLxmytQdccQHI0Z!Cp&>r7$~9HVWSk
zviC(;^`Qni#2!`F*%lxvSLn>(k;nO^h_t{;zYoXnu!hDe@}?m(!L7{UZBqV_W@ePv
zb?V`0#IT$U5plnxrE1t!QBPQS;u&n5MUbSg(jU>&^-4L5IXzwNd^il-fkrxL?+wf(
z7KbdCmRNETl>j!O!!6d6l&O9$gq;C*twD!kneet3w<ttP=xo7zjXx*=9}>YU2MA@?
z=0hlM=Bcs*Hb9YOVHzMdF74u0JD_;0a5>d2qw?)ZluAh@Cc3puaI4@3rXAVHrtdQ3
z%->hR@=3fb1>QZPh1XEjVGNxdDa%k}vaG5^6XPMJ&$T<$r6ldXZAZ`8wC0~+9ywMc
z=~sp6#-+$j<A3PPuG`oeV>#3iCta@O!L_<*H1=&2Z{XHyB$V>C&47%qc8OxDcmB7t
z4&CSv`Zkp8^`VS9xP*mf7;(}Y!cc+fm<05?^dOiQXlPNeTDt#Kk0Jl`i%+!ZrM8O;
zWIaJdP%Jv&uuGEgum!oWlFU&}l5yKWv4i=>e`%IeM;pCmNZ7APJ`$f?{zSM+EspmZ
zN5-u~?(*6e&o6wC)52;u^i&-lI2;ooqSZWMPMk6pDPo!icMWlx=X;ytGB5RtnMM+)
zW?~-7{ViElkb(S2xR4Bvh~4zZwo0Svac!!WK-o^DA>z66O73OLq)L@XQG!Qd8_P_H
z!}Iu-HkD&2*@xp+oUzZ>R`OacnR;+^&Tkcoq^qMA<+YQ(PoTP~2;~^KyDST=M>jCz
z-l>0~ULi+mVrhAP@(3JcG#lo(H)!mPB!BM?m_&%RK8)r+uq;(<ZJXRx+Jf>=n7u~W
zYgc|T3<8l5$S-S01tQAJ`a?h;?9y1O2ggQ4FTVv5fSI@XXU~4HjpPZ)jSGlpxCQ~^
zeG3jSXc2m;dExOI;t8cqQ2)pY7=#)4#-U|Debux|KklFGU>h=sRJVB^*f{lxq#x9a
z8<KmVN*?WP8|MEvTy%wgXg8K$&VdR<l($WYA7g~7`NS1~Gf-=s&%T53XF(x9t!;h0
zkbn80@%G#3^)wh&uS^{qQvCMmX%;220a2BeL5&=B5S4w~=lyW>)vxCl9E0_Zt9dVJ
zupV2Ui*G@$rXd5}@gu{pD_&{Y7U-md*^&`wb+G+{3RY_1vso#eQ2aV`A+KP`=m;;7
zr^SIA&Sa<n{34%%5D#crQS$rS8akqCLnOJP9?=(+vcu*r;<`fR4U_lUt}?RbhwZvo
zuFDQoj_#h0RNtYa8Gwu)R8<g~0<{3p{W9ebd-rh2uA~K(IR(moyooqzRJ|rJ>Ox;f
z8zFTHGmsME0wI*L@N9!OsK3Aph=guD({+Omdnjr$5@RLg`p5S-5!+`#jBb`0;Uj_~
zB$QcHSiR*60xEZIuoeeeW|h*lCn*zd*7W#Cy1<28s=}-kD=UHsTCWUjl*CdP;bmLY
zKA=7)<~64G2#i!v6q<A){vU1C*vdo*FtJ7nv5#P>nJ8}Y^Q>y+iTPlmX74GL_e`fU
z<qGA}>|tsf_mMAt5{seo?y$@}3h;UvCfO1l4%eJ=oG|Tfo6BY7o24*a_SBr+1&g`b
z=LL!_dwtHs!wcJBN}eiGw!}~IQ<(z3H9V6@=OyXEfJc$P;KH5XpjgM%$SxY<WF{4(
z&Mz(N3iGb*2K2eHG8*tOJKFYT12mi!CYcweB`#KewGlH6_4E(wY%Gi1Z|^cAe}<Z@
z;SCrHvvE4wjpJO`q%<Comx<H9gjqh0$bIuZEIma1%uAn?@Zx{~AU2C0o&??;Gq<u+
z6;iYaD(^u}n2tY)9(L$e;*w}_6VL(F7}K?~P3U()6CM^mu?Vojl6;5V4&kJn_IMwW
zc2A5h#3)J7mCjQ)$s<}V*x;mi66}nNIfnz9Z{-wX)Qv4BO?55}b$Wq;>V3C7r^6Fe
zqGLx4P|@La8}I8Ob^J@#kK)~&^Cn$Y81IuTOKZ$T-vaWo_3&KgDM~GAE497rj|2G!
z5^G`ST*#5ib(`$BS)$w}Czqp8zFlRdB#wnxo(Y4vJW7b<_{i#9imMQuxA&KpW2I^Z
zsBwfWXhW1OPJr3Hnv9vINmtV$>SCXN@%>m5{u6#O(*G~~WME@q{;%+JSVJa`hzz0U
zQ|$>K1FZZ&s!_9%=P<=_($ZEzR>4)yYwYJ&2#P#E>*i&2BMFE5WzPAYTvvsMA8r?%
z!+RZ<&$Z61e*n@hd2wj8O<*2a!Y`#*)C#ODxjsKM4u1&Wi=w4uHi4I%1i%O9`{l`I
z1YJJYliag%>Fq%N+KJDnbT;7QY%R_=?~Ug|f|OX@lWh&*ld7=}$Y<;AbkvS(oWBHu
z^(R{I+wT^?%g>d^nqJi$H?+_9>Hrvor2i6-8RGQ%?(uei^X7HuYKclWL9MuDWU9N6
zxC08@D#-th+zWKKnE42_ID>E+j-d!$yl$C~uMb{la+MC2-Mparn#PLdQRq4*QFWNw
zvx~d;4;|Kr4a^|}rk?ds_~LflELT8%dWYgRr-oP^B}Q_*wZjTy3r4uc4&4a=W)b8U
zZo?=CV$j{%#+)0`B*eyFK9eI2=$J20I7b^57E1wVo$5*hJKVLZiyOpC?HJ^he*GA9
z;b%rcO=PX=(F9W>>&CetdD^EH<|g`Klsl;3PBlu?XsX*vH9}RafjN57$B%Yw2d1zS
zG$XulRijvOjsf1-p1<Jg0W)e>x!>MIJ4n}1Aq{zmS4=~zhCVy*d=Ss;{<Gmv?=U~3
z{CRX|WPdwA3`%N=Ky;JGcy#NT6Lq6o$A|TW`$LYKzc`Z`*pXPo%&qB|*`2&Hij2t0
zAn;JZSsQN`jiz<#L15iZgH@!Z;PgBskt_Uhk^U^{2co6=qxVDWyFDiDpi#E(G^SSV
z7nPvPZuGKMvy;ylGhuEA%wimXy%1GC->r~Bm<<BuouFK5PRQ)q(J;1h-x#&;Clj7+
zX0AMQwFP734X@YDhl+Jj?Li}vg)BBL<cdcod57TmpNbT#!BgOZYMNOi-*|X8Fd}W>
z-~;2MkIXEnU`-w&#gAl-IJVORk~K5-3|}$<tTzkTE<bwE&=;ZWcGi^0X%rd<B<gy|
zR&2M)%bmq4F{cIMh`SoqG@U?sknHfcfmiwol+@huDl3<I3?&CTX}v5Xfo7HrS9$Nn
z8!eFiex|ajB^o$CF=_XQjFD4f+l%~esqCxA$Oj@1RDh?TCMKqU@d;?Uq}G{N)xK6&
ze)Q^&p7s$`Hk#f#)S%ZCrhn|@k$o_`2kbl(*`m^*DfaC88{t0Q{=L^zCD%w%PI&Ma
z?9ekJr~tDv*oV6UI)#gBh^!b+tZW0P=0udULBlu)u9^(pK&U~RRDGPrb*2TpYK|Y!
z#`xr|Y6gtcy{)zdi6EuG;o-d189&O1RyjnWK#=q$Vo50JZ$1vK-iU!P$o+Y2eXFhU
zHk<F}i%oHJ$T7jqgb{Z*RE6JSh5jRzF;kiHN`_MdvN1e`A09S=S0%o>j!Cxc<ZdC(
z&`DXB?2&5%akNUAzrb*oL~SJ}U6Cl#ajHQ|+q@m)m#bp|s-J=uku%!>(;N{y+CTiT
z?~%#_DZJh7;GJs%jzUJmD=SxT9h_(YP$jWHeF(x@U6+&8mm7oMO+SD*l~M&fr$F4k
z9=jql>pFto?`{(BctRl+`t)~&j6o{o17G?jmby%;lrE8BJSIB^sTW!e`=6XF-L=7f
zmlAr~MyD5TE;r%nXwAdteLRV9WGQb}vXxt<@F04gChvtN{}`ll!yurx6L2I{6PX7$
z%_N~cem>y1Gj@ERrCZ59iMeGETApMBna~64gf$vBQ3`trVaZpK04<@zqx36PZqM;a
zDN$Eb+p0|3uxyUevb{TnZ>T-Eh@X~&p?7R!>SUwwykX-lnD~WJda7iSGiBjD5PdpW
zX&>iUvS&4UtsrFg=3^wYF8ddFaB#B%bg%^Bfp&km%6S)El>U=407H;C@GvP#QqG*7
zkYzrEXHiHF;tG3wUN$!R0NdQoQDR%BH_rdlRg!2-ZI7ZVpkas9g2Y8Or%JALK}(Br
zwLq^~Txlr$`N?~)nRYY=QI{sw7bzG1&{_Oog3f*eNA}#5m9*xNl9#B0e-B8PgIe-R
zuhF`Bj`=h2m=i%fTh=8d&TdK{zmvU4nuTNS01=yq5O0@V`-5!C^guI<hLU5Kr}S41
zZisLISeV*A!WQ$|R5x4a3DsE5sAi*wEcOvQY+T~D<HJ_dM(!%gLs_o7J7pi8DygC}
zhRjDKGKd$?yeiqe)BIc04P)~?eyIeQZ#qmtz<h--A0x{eogESTOncftAw{T<S!gxZ
zwlY*>cH%rXAlWyp@G6}ax3`&7%dSy2CP$k~>bPpTyIQcy_xi52)mj^!Z{QLjwK@Ox
zw2Pg;m-9EP8+sriS^3-i8<ohf&+qC7l+)v5>pz^64F8;?^aOT>mQdW>bjlv~CUkO!
zmP*bxbg~2t^z?M17LHEN1RU&4bpNl_l!1eT;lDaVmo+zSu{n@@X4UQ-UjA4o4qp;L
z#80wg>!V@+)gWjJH5hY@s}WYkuKDL!*6S-<i>z@KQ>gO6oDw=9>=H+l(@__hxF`-m
zQY|BHmKnq<2`P}qFbLe2Q;2Wf3i>5v*bjlufEcuz27$bFltWVJB3sGlq6ovY5iT49
zi9kU`Ry8+)QG!c8?Px*ABA!JMSmy>n0nHsE7e8ex32}{d4*-}r;hj1qiZY;=g2)(L
z9RtD^%4fu<2)K5@Pw>~Kj*7?+7ejcFx()}Kz^Vr9km;%&hDb9brHWSLPQ^iZpr-IJ
zP!ye}^f$qPtP<d$gP<`)m^Ybg^e_iN^?(!?WxXGv`Ouz1L3j-n36LE7+%f|N9d2Tm
z#(9Jwkfn&T*H_I1wU1ZzKwE0iWm1M48f%ot^GQv21XSLM4<q%yF_H<U=Dy`%&&?=<
z((v4|AWWqs-&anBM`l0vz~_Z1WAYw=GHdhzS%Z9+Ml7dx3LoScod~00&-*hVpGN~M
z>HJ4alsUiA1jIXL!YJY_SSpkaGwCi;ovvQ2j#>`KAgc5mOW>V53bl8cVrpv>#%H25
zlLD0TlFq<d{w0Eh^{D(&EjorK4_V-RDouDU8g>9cP?d7P;C~)QK+tf!bU+OJAW3kk
z3SlY`1Q#B#&`_H|s3H*dnDXfy>Ni6Ne}r#>phL!5CG3k4FcmQp!zttZYs6|a(9)T`
zY5iP)fCrPzW`*x7Gu58=myI={0;BrShr1*elb;3CTi+*FPo@o+EF=pyV}mT`xxG|z
z5I)JC0%i{S8@RAmSph0z$D3ivrJTygVD%yUri^m==ysNupX#k+Bi-uYBjJx%|K)p}
zAXL2H9xH<}PDzTjn2!rJ8i(4bdg#)YD|1sPS&!^j`zD&!k4&DQ!#TKt)2A3lRx!r4
zFOS_%R>e3%@5|%*zR^1aD4W%*PaY3%8O>3RgGpN=kaC*_JnY77hRT8{`aj<OW!#RF
zNBeq=KbPxAXy^lh_j*45WZtl5;$**yB^7<cs^jUu-|}x9TmBItj|V(?Qxou<lt%RP
zFUK^>(0p!)E~?)$$jE|`<vMQ=lhd^PJGh-q>rpWj&*2a#0V)Ew=j?lSf#w+cgmAG_
zI(p!b9G2eOu%E1IHH;IyrJfZ&Yu)@aNFgBW)*j&UcGkTrUaB;!>x^MU#|d}e1E+TB
zrJE$J{#f8scWCRzKlN0Uf{9uay}fU6!WAH?vCeQWHfqWAbgDd*&K2yt;6uM?PalOI
zC9*6lZ1wN29SHOB<~ct_9GRx#qq^e3>!SVD<gV2%2)xNV^09alo}B^-iOq;ofn=sH
z%Q;=Ijw#LVtB6x8i=BJ7x_;nvf8hCH!BmAdY;d3VYWKb?hc4sP^3_2;5%Oj9;^8{X
zm7wCIYhUHAA&0gYl{9{EbyS%vGwH{FhMkkNu`demOMiU{zZ6G+GEzQgTOQwd-Ud3E
zc~_Ji$*e}$ClX1Cx*CD#S71b3!VY=$!^xA!f?lP>Q9~QGd-+hE>qE=8%8dmDaPYAB
zC7S0oj>!S_VI@!#OW%nY$XZoeX_oZfm_9Wtz=EDOS*W3#Gxd*e9*s3=Qe}8D1YFz@
zmL-A&p;eNW<fBBa1xtwCp<mC)@Gr*G9C`C-X5><nJUY6wsrTrY&%3+3j~N}ra7um(
zj3^|MMS$5kDD$T<`y)I>GsEteWx0296>Mj5Tgymtstt*d*V$myccFn-VmZ)iuoK0d
z+gBn$+B>{=rX#q8KaTL(em85gZU59~U>e?vXP8IxW}lxARRDhI&)%F3o#WoL4BWx$
zqZcW=LX^pbd%-8oJ11O~AAqfFVqupS$F913A!e=6*SQQi8Y$zRew0;07cHUVQx&Yi
zO~4n<!v6gvJgNwbUo}E_FbxZ^qu;`4Pl5CM9pp;??!U-N!Z(rimCrc+EF)6n9A7Ps
zvvl9It9khDcy%M!5$uD5VFp{c3*x*RVCaN})&mXeQg$w|voJ_52@yy|`$#>BPa8Q4
zwXY<vaJp`bN$MXTz92%|`RsJRBGaFT1OF+7i#jibvw|Q0uA(&}#y2RI=ma?xYhRkV
zA4*(u5RR}{l~Jg#J|VXzje683PL4W=qLoLjCVf0`WU(`q%ZlpE&Mq({??e@J<|6dI
zIt?eS2$Ph?BOWaC?M)Ls_LAUJ9i%U>T*nSu8y$$V2Uf&hMLgeh5q3PRO@;A4stzbV
zP@_Y6)kNllm9)G5?iyN$QTe%YUQRZXD_cHZt~}=))@>vbdi6FzVpk_M;7QX6G>jj=
zru+Ceo5D@ihtj9^e4$FR+e`S}FUN*aYt*i$2#z&qF@QM@AlnF`HS(iVTPzW>$trir
z*1}_>n`M<%_L6axTH)6G?Xx*=L$nEJ$di2d=7zOE#MD1ei9CaHoyff3?*51V)8M}J
zLmstmSZ;hhzedy&t2(rX!pMMNweleCe49cb&B})bfI}Nbh+QTzAan|n!9@K0771hg
zNf%54-UG9^g<w_G@GB<a%timm?2Bc;579um{7B2qSAHPX<i{hNXq-=-o>j8$ImgzT
z<AwWp?a@?4t*mLDEuWPjW*;BB#w1v@7>N3n^03OS@8{uO^Y1S(O^`Lle*gw1rvIHk
z`PT^OzpLxiX#T6~bRhn#>ztVrVbE$&_p*ipNZ`RIYN7Fhh`%(CNx#&GfdIbTmc7xo
zu5I6F;2UdPnD*f7sYvDa<{X_*22tOxwuRyA#_&<lM7zx7?)2uO=p;O;Nsx83kWmSk
zOQiX7eIIH_)+-*Q=#rPrKRG`2mAN_n6wfc>Ubz0eU)P1yqwXPYN7pmR<`0RsBNe0f
zE?kPzX<>AHtpoI+^jMoAf&h(GoM8}xt8kDP=AShWK9YGRJ*BLKoS9UFkgkw6HcN|*
zt6Om+q8H&vYj6=+FUoWoHVenAmyZ!9vh;R~{gIx>Dgr&wx#<WG{lZZ-P(jVqh~O=<
z!tqumxQ}uBmupg-1HG10G&kxgal<Cg{*e-3=zCa93#%l^N77vn36#fKDKbEz$i&r4
z)Rgc?@{1{l7W}9us9eHNTPcEoN?qu@c{Wa6{o=sfVHr{@dpDKUnmHB+eAX6RT4jP3
zGl)Q~Gi@m3vA})GSqZ9G0fH?-j5W-{SN(t@@%UPrDbKdwJ9U*^Z$@26t{Zq?P9Jd_
zcJ7TDb^Zz6^)f(=5-B(<)at>esRE77Vb1ZI4E7$pd;D;9t^60!I<PbcA)P&?lqw1q
zX%x+}6hHaV4D7Q&T_k+qUu3ad*`Xj1SE#uXQ%0~pj=I=ksw<rGq}E{ic7D)AHra!q
z>CY;e;V?0te@#)cUZTX;s`4u0q#EKmJ~3&Zkf-_+hc!DPU+0rjN`0<dbmWq+)|10K
zAhkmNp?=xKSV}gPGYJsPQrdCl$g0sVF1=dcXR=znk)Rzs>0f)YpQDn|0mvpqp>E)!
zK0z{lLXjxgpTjsVt1Uj}dgc2U%VJq8bC@3pcGDUc3i6`7UKdJ)#E-&ec7A#>7@qke
zh+BdV5r$SS11qcoc85dN)vE+2k;iemHF?8xe;z+2@d|Tq=KZoEj+)=<(rUW)JgJtT
zE{dgT%IUXM4U&3C9I=11tWtUUu@EW+Mcy&jn9}vV|Jef3j6>4;-FlU3hoDYJ#u~Rd
zoV{o;%zmZxxtgopK3_2DuI_qf5`cx$@$u5WSao*=KfuGj+C9@_|FJ_LC6u`6h7QyQ
z6k9>a`?K-w>HT`2?{x%_B0FS}u8-$0ba~1gp1C!FyCC_sURfQypO#{cQFuIKVEbFx
zk5str+esN?A&A5uTAN5C)m?HFZW6+0(fv<?1~5<P#lVr93Ja;gDv}lv;4#t-bT2hi
zWQg>aKo_ct3a|9!#(F9?K+vk`L_&5|Uj{H5DTZ~d@38Zr+{nUjaUao0c|~a2cfap|
zMk}D!0iWV4B9R)6z|Oc9E_U5aUcr}y!d9M8%M9&hv(zJt>bpS@vuBvIKr$_X*GIg$
zF@mqp0MSFMAo`CUYrrjDtxhT^vbqdiX#tO_-IZQ>a)TR%cg=&+_1q%YmbCSsqMQXE
z7#W}bdQ9PwF6vwCbS_&CDvbbb$0rrVzB4Y7Bug?aicU%a^D=9Szt%;_4&1jg!$`f-
zBsN(JAB!ZjCQ}cfpO28$cdgblo(AU2GM^lC#8%p)!k{(&ZByuORxqu8m5->yH-CZ^
z<e2$PW0TIyw%BrfFFH7(jI-(U)8*3R(YM(1LV0M)W~D^)^rOsrv<yeIG#IO`PY81Z
z04RzY^#)+Suy0cV&o#4?D|%F8y%KV+O}Ty|n7ne{pxbz}WQ_n~>M5g<f%OiFt(^^`
z@br4SHk(V3-KdJYbOgja+K$_#S7s&|;kyh5nq6DBhssWK9c);br90R2d>qda=Pm>{
zN$<`*vh`IPm`KQ5?ssH~oAz817HzjaUV3KzlfZGVa;<7`T_Ld>l&p<AW-mj-_iv#$
z@z_n#S>9G|3>-qM{oF~MMKPuYwR>Wc>ICy0cv!#TAd6xX40&w6pPT^#99H0gY~W5u
z`}k6jKa54}H?r=7(ZrYKIOvQsit+qv#IC-=Q4#F9fwHull?N9}_Q&)t5YsCPOUQ7S
z^utH>T06+ajY6Twwf()1Hj&}~x<c>6T*yGXl;O=&LO??c{7A%Zq2bG`SBxs{who68
zFc+{NLlRUdFrSu6GegM<gOU1eq(U=X!^fth2Fa5WH&LoszkIe%mzR0^7qK=Gd@Y#v
z@%aF~zMYo6=ChmdKV2kNUbeW*D@x63A(1nhM=r{mkFrZH>LWdllK<(!>}e&W=^()0
zU)Zs!P$eeD-Y4I4Ri?NMCXDwJ<}aH}-~#Nzg!Gy;0Tk7@$^!{jht_Hh|3ToF(Kjwd
z9cYTa2AP=B(WbPE*{t_hv${NVw$sy3>H11kJ5D&K0yBmVsl853l9!_7FG};>enL{*
zLli7SRxOwBe(LkH>x6`C9`j<e;fAjk(!a<%haBA_*leKH3_Stbggd$6y1ce9261A}
z9e^l#w)dnvE|*Qbp@;(*b!<kOOD^X0t_u#WFK{Rd=K88eBvpNY6k=rOzO8<UHH~*{
zf>HFP3xXS1cphcE-bs_VgSlK~7zHD6V@H^Udj2>>2j#C_8agoKMP_7sGSfgV!&%4a
zlg4kSoU2FxBgzgS+-Rx3xwT!vtbk!_D~s}0WP_bFMYa=T+LscO$xgV26_#C`2E90T
z4xEc{sq%Neu1G=y+Y*ez@qUCX-T%FGvxHDB(M#;s6RJ;zB9~>3=0kF5f1HzmJbmsA
zeI7+P_Mpn^k~)H=XuN-N+UN~Eg>KQf{$VVx0e%;5T$qZl=Q;y^5pEt-klBq$PmvW5
zV<{+#cRI!ez^y?O%0*ZJ*%HnahZv>cTD71}jXH!d0er@JtnC2gQAB2^#Mq{}PIw|F
zgF+n*y57^7r~JDW+ZxN)j+v4<VV3;oHahh4-eQZqPB~EV3kNoiw!7&KmaR-L>pD#z
z%CF2r^H{DspDW#^!+V$ddm-hUc2v=Vw{F;gOTI?ml4s&tKr*mi0G7-(NSq35DIAZl
znW+vDDhL34-CWWMs)HgcMS>CPM6RJIIB=`?Cb|k`#S_Ka3vCHxmAIvf!8x^w;A%E-
zBPd>0W?9Psw~Y>Z<Q633UZ9mzfBbEehN<OnqM9F$RbV#wkSHgpOL=}#Ie;&o(q)M6
zpf&LxEd(rEJ?<Ca0fKc~r&JZ?f)^l!_A*^Hb)utm3_<yKL+;9Ok%ZFnOK0b_E$kQg
zBINS!KO)+HFugOf{C5%IX{^b_t=9Wqy@4Wa3ITI~7%`2E%8Ue6)I1~)JKDSh@s!Fl
z1ii0M3x9W~u|FKz9N@HOkokINZ2e(%!1%m8I=Z<#JW?zerijrPy1g*`z}Y`QWpP;G
z!2Qj^`^TGilU_3+rKf^%2ilUJFyF0DBa2#)siyrYc<enaP&;amXNa>;+C(rej2}MQ
za=XR(;Ait7qmTnU#F^kTd}VrizMcmOXHPT9P~!mG`7wqdbc&w)Yt<7?CHkT2HfG2^
zWD!rIg}}4fR0oj2a-qGH#bN9UvxnFbiY4FGe_MX(s!I@``eO9()pQQsKEUvSRBm%y
zaB-Y8%~hmy#D{Psw6yfcbn<Vecm2Ypj`A~@77Uw&HA<8Jg!((6rhReI%(JBt(C>#`
zkQib@TPgxC)4pqI652i#f?CAe5`UjuGNRd}gGot&2$5*%Mt4LE&)raM8op058l<KI
zG)w)cj~oyqmo7K3`RV*6>@XfN=sIc>c_C?=@=z6r!sz&Ij0Qn**!}a)K-*!nhJ>TE
z^#X>F{F8{P=YB}?tSem0Omw*Qw_ugK_FwQqh;CyqoYON%;oLZXr&;u(Ng#6!x=woK
zNCADC#R&^m6ZMgV?Bx6cw~(|p;qnS$8#1M=KNd`n91}NO-k^D;h|p!8ro1UHB?rFv
zw_6L9H16Jk`~Fxb;nWXjx;bo_vx7Qo6uGZeoG?}!JNp8}mxaGj+gtqi-NA#mKqgt#
z@{K8`@X@IuS1K{!t^=Sd4P5U72NX17kDzk?G}mhG8BO%=6ig#X^P&+FkanL#52%|&
zMM;xUvB@{eh&V&fB)uMaI6aD7q53q^Ny`wFbU3_$Sdm*numZy{c7oTlP&+UgwT}T5
znaDF%yR_VLqlk+U*bGDSjG~i>Nnl1ni*5hV8;Ow8WF_XZ9Gs|S?=^!R)Ws8;LM)v@
zd2VV1<Ki7A6!?KKaVbffUc$euPT;qeD3kylQ-(~6Cfy*hv9jdynI=oqm0bxuWL)6f
zG#8IUR+hY@9use#W2B|x2>q(+zz}$qCTAEa30lPieJSuI*48B$zX?~iVW`aD=U-I&
z_Hc61kIY&^%3wNTKW{hp%RjNd7*)V3xOwRI7ps^1y9hjML1@GRR^tG{_xs1?tk*l%
zqwcUcL)*%l5CwqDMFkj3#=|g{we}&$(@@)`@ckyEdLK>ym|XeQ--PH^b7hj3P$EN5
zvj>5gpOIy0jM}&hO1>^cSrQ=hfh@fI{h?zb4YlkqyjS>L!Idpqx2wbokXZS2+HBrQ
zNpA3>AtEgprvf`XYJdirqtJVD`rOXv5;2o<^J8H`#;C_ST@%`Y)p+%=WAc<{xfA2R
zxGI?y6QpOVVcn{&f{IvC-6pl1kRJc*T5PJ>LHF=OJ~!y`Q>dJIUJ{u~S*)Pxf&jXM
zVRS_05tBeK_}B`%kn*~7dR%G^aO7+2=tH?RD18Z8^J}9FUXCeOkjDzk#D#Br`Z47-
zmb?N?NwQ={b;U$riU81PXL-<?mxoBJgtZeP!Zo6-ZjyQX`5!3Xu8_nj2e@BnXwkrw
z66@g(B5rHq_8|ALJ&ms>kR&O9=Zzhj4bdp;P*g#ojD&TOb?P+be*ci$GiMMr=|4~D
zr|w|;O5I`O+0UI&EXaa5=a_}F3Q!qyB)q8)UMXzi^1;T6CbKQHpF6c-^RLcpC)Scy
zV46=&$D^A_r8LQ^_}S04N*(6xG$aM7AiuiGJ+2DMMKWhb<rni+;(^_Akqc9}LePZg
z^`60KQu`8vh|Td*BNP-bw2+Bbq3r8&zBz4<vRVXFzd>-V$-#IdPz7YDmTI?(qGB2G
zIdcsHfQe8%vQetZ=VOpGZg*jX8#n>GhG#o8<Axz<uFl$KP4egNYr{vA?=^4|BmX?}
zsjt_%kOWGtb}8~(^MfU_5<v2b9@zC7A3nf@pn(nV50iD?Z!sgjg&hm)`*HM-S|fvD
zzc75CvQxP2sfJl$Rp|*){_p%hsfvk+)_?&EAx34>1geUn!x;>YnAA4;LwL4l#M2^n
z`IX{G_I&D`WW7P-QpkMxifc<jJUz=>Zg?eB;K(x8zwEgwU~zw?M#z+U{!oqK_mp8>
zcF+pMsXvI12U`gLO@PfXfwhyfb*(7~Zra22?O#%!S&h#|KKnY-HC6~7$!(0hUJ2pi
z6cwoi<<=bS#%Be|Ze-A9T$56Q6U|U0O><@SYSBzo;Yl^tQr(J%eZELpo=;Bgig!aH
zR{F33|L#=8cD>!ewaM4FRp?7(g(yCduU!G_A^+wRqtcZ_UV`;pwH*4FK`esZn{A&K
zN^qbGPw4;%Sl+mi0Q)F58a5#^s6cQl!Owqq|LNeQqbqxJTYT~vl@lb)GFDy){j#~g
zvY2Nb{gN}6Fv+GQb7e8IqKe~?Vh`a01Anm?OTb|G^RahQA0G)RdBI5?#~Nm2FfTA&
z+A9F{!<v}BFp6$<h<AJ3<-AgCcTUlis2Y(d@nE)n<x|r_y%{k<HPAjkEmI_U4BXt3
z+I>7H&R~;2%DXxfIZE*Fos$loA&4pkWKFeZ0T518dqKCAhiwy*H(id=aYSyCQGm}2
z$pPH;1fJe{6y*>leGdXp+mQNLT6^)W{>$W)J#S-I_>VM;gJrUb*~`VE@2{PtA_4+<
z7C6=*^|A##YWDY>1tJ2Ywqs>^=^MwCEOB42QS1o^*%CDJmlBbc_)W$hu{rLovL8G*
zM}Dl(J>7jpS#S+nNjv7&4CvA<onQCCIYfKsP*kl^+*C0pYQGG0uY8r371Ha?#oC6^
z&xyi_?_DF2!>q<92g0nw8yx>R7%8=Te=cy`zrZfsIreTCH#iP%O?HcluQ2$_ZSJKF
zi4*F&PoHxn(crlxpEt3zJiR$)<1>2QG#Njc^bN-cn`g3G#7-1$OI@xHKC4|q-|-yF
zL{L5iUFSzNpm+TPT4Ip3+)mIptwe?t^S8Sw%fH}zAUlHpiH?~6gMh{I-}Xz?8c#fG
zhwWKaI|C0T?dK^hB(yxg2=D0FZj0bp&+TZ3&yDiiB@kg0k(Z{)SO-sOOhy3+gr~}I
z{BTNKTrF$6mGk>l_H-P?wrBVGH3Iq&T+oYun#&vqgs9U0_%Zsp){ty=?wKzsQI4b)
z%j4Oeo}`1~j`4Tv+&e}u+xK9IrTr?t<XN1fo8diuJ<kyA^rw|_1}cnBHpc<!)YbF-
z{kcq<5(M5haQJ=rKJSy$wzjI4SCsqxx<~1og9?6NFF;~0IIr`Q^W*;RQ6x_;xw}XH
z{_YfA(Ab|BN$mTkE%TdC>s!!A{d($SeiAbwD*KiGB>>Y3?un16Cqxai3O)`de2X8~
z6~{mgUfO4m$Vt>dS<x$lQUn<qC;hh*^(CxgEc`|M1&hL45d*@_^d*2&Na@Xik_!A-
z07v<~Uml`TASD@n2O(fS;LLH#y$8J0@=Ss~O4xSOQ0hzcRd;;z3|i;Jfom!+n<Urg
zS|A6(+95&3B(6Ys6P5vI0rIrg;KGdNq0pZe_Awp<5djggA<-epj>zdbacU0B5cm?B
z=y~fprw<9C=y|28vJ}>x+;r3P+1-pz+gLhiksCqMZ(HSO+&G+~W*5^3d1965EPzDb
z*pFzSn`_~oWXbKd^C28c^7+4P1|Dd9iNFQ1wUU$H`cOebPiw}P6sDa!J5Vkz8nq?B
zm?jyrTyz@aJJJas4H2Wx(Z&lFTHhn8<Osv!pVkH-n(vMfn|30Z#2-+x_Hxmk7~W~2
z(=@~^NKH@R;&p`U$92zMQoh@;V&*^tSnte>4cQ-BgNa02H%g1$f+PIS7WMpOul}W{
zRRwChv>2_5$dKcySCoCMvm_Dq&&FI3^0_8nnKFW*+;s=3M<Z5&V!YYV*=On{{qxc!
z#aL4KB)q1R?DQ@z4dlm{)`ItlbYmomJJuUv%uOr*XCe%`kXRut(n-OfppY|82f83$
zeg!c*=#TS2J!BFqRV#VZnt-SvYy@)&40}?_8n*^HVj=y}L^OCoO3lI=R<Kii97MI^
zJLTmd*Ysgqw-R2<evat{A~%Cp4CCBZB$~0Pb;4EBahA1dtpB0*@aNt&a6fR7BC|;{
zg^otbP$0y@{d{E-(Jb32gm(^K6>_KmMPyQ{6Nq{WJYW&{;eml`uSvlJQf6EftXaDY
zRN2=MzdC37MLxstZ!ZKBy&pQ6-Ab|}a5;+!WEg`^#wN5VdT(n!ee(-CAKUyQBf7Jo
z_c2Js(qtYQSX0@ZllII@IEij+15+Q{=&oT9_}0T%EBR&v7M1y4I}HWp*O(lmB$3ua
zG6@K;dqgpN#X<7RXfEhPH4&zOjC_&Xqp?j!|34>m<(VZk4ltVm=vwT?9HnLZ8kf*B
z$s}PM7m5)?Vdf;OBf0T!Y$XM2r)T^jfdJ((?mOwsuM-5y(@_v-KvAVB3l5rq1Xc98
zu>KXi0<mN=wsX<~n>>u*A7>OkRTY#bjYAlG+%8Tsg2F;vDnI=Q<tZ_e+M4;G{pylZ
zG6S+~77-<@Wy<U#ZM(J*^2Ff+-TQkTDUi@ye00>LA9LhIffW0Z`SWypEhjRu(Q+LJ
z=g~*}Oi}x6;PS;(6-G+g`$%BHiL#0jeY^D4!=?d(z(+B6D->Cw(ND`~`wa@G4lsd$
z3+M-$9$xhtr=7%RvHfFQ%JdCzji%zcCC<;P2c?%2e)nrEugz4O<r70k_*A82I~yQA
zGc{o)_uElWB^JpaLXW5XR8U!kSN-mXU~6+u#Eu+qVwWzG20$_dm^9<ye?8=+Z(<|~
zTqGq$vR5wKvxhx3XC4kUABMp667xbTTfJ0<dd01k0bR8RGjt)y9(=ppIq=WP>(ozP
zIAMloIbcR-VL7y~Xym6SDBv8~g-{K_qtXY*ktt|7Adfk>#^_ch8DezB`(6BeyRoKv
zkVIYd<5-gP8NWchgp(RsM?8$<B%zhtOQ~)!qE5;BggbOgyo8jCJ4L{J!phlaxSRB3
zKbF(X+ryJ4ef2HUQ`;oVcDHfpROp{vlJau*ydyF)*@4PBb>i!W8@M=imsErGJG8-Z
z=`>~3m2p!xdFVJ%z#7WD)n=o1AUfI8u(|DGGqd;liywD@v3*t^PfmcsHabsfVt`pc
zge?>_0k-Uq+RbMTQXYHlEEEV4^`f5X^9u_M$reA(gR`<L9yI;6C>WI6B39Ue+qS^<
ztTl8UcteVmlJj+0=~z1-j`Z;6#njL&qV+5k)1~Zok5GJU6XK+*WPeaN-L-fu7TioU
zY;^NzHubx#sRX1RAIUAEv)E4Qv=ms};anv;WBea$TWZ4q1NIw-8hfF-SVN3}BqG3y
z+I2A~WlSsRp8uZ5G^Ngs^lEUqgW`vk+(lNW%b6t`d6F^5Sc0e3)9#A3)C|VYz+}74
zsA$lX(AY5a=5#g&wbOFD7bwJJGMk{&*=(iebR%z;0|r0mI$^W4I4Z?Wx}KJ*$}VH3
zDG2#fl3Z`Q_~fkFWW36?!vOcF)16%yv$x$P&e_t9sB+TSChYI(@+^ngb=c5wC)q4+
zo8`x%%m<IXx38Xd+BU8Z^J`=>>nOMe@y&+$7&2NpPg7B*dPp?8?wGIMYX&l_1f1lK
zTFvLz@7?-!qqz{oc};dq6rs3IpvG(u7|E^{f;D>lI?XW9%C1(wsBb25b|scvJZSIJ
z<A2#8k@!h$Mjl`D{Pog(*NU<n3p$iB=aMum^QlnZtITtbL?=PL{!tvQU1EBi1<gDx
zKj%NeNP0Awawc}c>u@1<IkEp1zo-Dt0cci@?cx(-E!+EWn5k;z^0K=_$!D(KsMXs=
z;Y6Z-Y>*i-Nd`Od{>hYneIWyf-+-CrAT{MGZ_a7GJpD5c?qd<;p}{O<u}G++7{&G&
z&|FMiJ{(V6KlMHOL`I|ec%-$8V<uf_>99I5+PX5FQyCKX5Xu~s#PP_1N2})4{}j76
zyq6SZ{3Ys>?Y)p1vr-o&1n`-b?>yegQqF_@A=U%uqQa^3_5JF?ZXu@8?m?feNouxV
zoW5-{M4HFH_k|CZwIyKsB`TuM8`ht;hhy!C-mdhlyL6<Z@bd$}f@qXR1Pw1JGf7-1
zT9Zd}A^D-rVe;A(lhmCx!@&zf)*=6IE&DhZp7wcHCU#o~K`)S)4uY2-04cMe&WtNJ
zgfvYONNKHv{AaaHmj1CCW`u_3TBwhQqfHV88)!N=C2g|Sp`%OUkW8ufRaMq=OVE?|
z-QiqbkjJFmp**eFcrDy)p)A*SAi<^5+IN4NZNnJHasUMR6A=~y3gBJo&yWDnXR%{)
zD0k)Oo&bhY)-NBKNDSA(dDwvE(~v%SABEYj<@(304{q?{$L0q^yL>)g6SXXq*JNQd
za?j7}l9{rn=S@yd)8{Xs)BptUf5glGpP}miyC{W~{=e%u(W51mbjS+V`&xI0k3lPd
z<gM(|QF)!4Te-!`)ut@FE<?)|HcjHkM*ynmE#7Ke^_t(RRS0l56hr_qxrk5Z@$usL
z{__wAZzC(~4pJt&>&FMk6aR!xyxr{X2ry{*4>=t^j#VU0&O7H!ikC8J#ddjo4&I}{
z<cRhp(X;#5f$NspJD0$9)JhFzW_lk!o#m+hFJJ~LGAb52Ae?n(`+ff`Mj00b@D_Z0
zeS9DL(e<pYxbHV}e!m=&@X})izg!68Ph2kM{c?Q&Jj@FIQA!;gwR?Ki5?X-9l+cvQ
zqfPueR`d8^<N1mJyWErWD=FYRu%`eDUG$UtCH!a%T?ZbZt`V`;Cv``)B*Y7jjU!Ee
z&Poo|7YO{@{)kRRlrHX{pPQQw-wUCSES7|{#}liJ$cdYR!1svTDWFV7P-VfoR{$rN
zdG}ZPmLQz`?33(IcrN*RAOR*?qC$aDG%sW1$OAGLPmV%T?6jKOYVZ;36<+Qk&s8#j
zOI9_~fLWzf;&qX|9IY!e2%@dxbT1iKXU{7MZV{hHz<@>%N;EaWdFNk3JqqMErADv6
z?1qNijIlZtTyX|P`M=X7af4U(wY^y|WIG;ZWWKB!&iLzqKA1VUtg(aRBuU)4$5qk^
zyJymfHjIs4446SBVWoTVZ%8s$x{omGwa%l#5mIoN5k|QBI}_r5VCWXwGn&1PNA9YT
zCe!S)%`PK0C(YrjjmGQ*uBvyQ;4OcfL~7*Ym|#Q_+cE~3qck4Rb*6|=V?dOM8J*^8
z&J1&3g&~EQF0JNueZ*#7Ou;0`eD(`_?P30j=3rEJqHhR|s4xbw?D`7Q(Kk}~yPdyi
zYO*Cg0EC-$*&7gM^mIA6>qwNuJw)_B1$KTv4i`6HQ?__Vj^t_MweyskrBOa)pU){J
z#&Nm!)lNanB7-1+ruo^<mP`{#zA+nyJlwS{f6aRz7<*{>E>ybA2<xh|Kiz8rOT_h1
z@Z*8%bAt>axnl)}t_cc|Q&ksY3u(b7mX5g5TL8-;5p(MzkSF~XytBCY{HR{a$M%x(
zPQd3#f;^hhN4fxEIlwbCh~mVJV)AkHtC&}yJJCua&a`t=NX_>H*C}Q0C;-ny#o8p-
z>TQ+wRE-EaOvbxNR%alpz>JSR3y(Ie06pg}fWs|r$teP2!so>eVsM|()&0gKM3ca(
z^WFT*PV3Fd8yZQ5XnapwJ={uxjwnV}|IpJhk9Td`S_}MBeDlkyj|f{)LM_&@{@OoQ
zW&j<P@Uuz#qfs7axzzo&b|EV5O2)s+#kmgYugtyNI4bnu!VvAEye*bPt(4Dj!tLUN
z@kfTLKW#+=LxC_FB5D+|=~qjQ`wEyw)95{sjlrc;2z^k=wt`qll{v?dca(>5IZ<|O
z6`5i}j{`Cq>~3wnUFErGPC6%d>r4yeH?bonMjAeup;>Kh_DZzfLpYJ<xg=E){70cM
zdcN^mek7KRpif~enAcgU0|H4GGqh3Y2s88VIj|g@X-v`$>2)2+h^LO5c-5uOrZa?g
zlI+6}JJ46H$fp+Y(LAF>*TJcDLRo{kLr$)Vhhk#S5L^NxPD8){lIldRhWzX{Zkv$2
zwq!J;v-L@_LjtY~iadj|%_uV+MpwqvnKvyU5Q9OxMv!X{TQ-OqdNij*G=%vOqVPDm
z%@84j6#_Rz)PkN94}*@i^}1g_#8Gnt5p%^8Y!kQO8$v^`$khcUe$h0{%<%Of*EHjm
zXw(Kq7JVyj)YrR<EkxGcwQj^TX~UVbKP>tif1J1Abb!qluLk?~*ID?7HTA@rm7Mxi
z7STeKNgCMg>8j_8jS@PA*=-S3k&pgW(<EC`K$==}ynW%;K{#(#3l!Ui>s3$Fg(N0R
zvd!MNvm+_QDG1sF9zKy#V(1~N8Z$>JoGu*mtGi=vC^lZQ>*ak3?l2woEzp7TPC?6e
z8fT^ST;l0tcAW@@?Wn=#JQQ^Hlc*xr`Yc$>ZWgy6@@B)RrJpyZ#$%RFfCH9+#0S)<
z085E;L|WUfhkwB%Sv-_>9P}uNQSmR0d0zSr-3G>s@~hG;Yxa(U!V5Zav*E>WgtXIg
z7X_OrxYamDUm`Lf*kX~HqGWY;BCQBvC}p4zp6|QEfMDGF5&<7MY<UB;UB<J_4w;EO
zevPv&tX9G-qqfPV+|Ooa-fswyX+!HcnRe0Z(HMKPY~?C~6m3}RCeANt)p8}*a0glv
z>Cq9@)@KuFF<)_qQxsqF+nq*;3h{cb8v0jo-k7^Gu9eYZqaU!B?<H`ySKl<$C_xHN
z%F9k&1DQL{4ANePB}gW`3IOrAurU_ZNiVYJ$}<64c~F*nwBj<b%XV@wtOU}Kb8Veb
zl>N46GhbbK?3x@05rpl1_h{mPfD#N#*CRo|_UUm#K=99By3=Nfla+u{G{ppYBK{57
z3JFS)G%tOddvl3n;r6K7p@yPKolhsy`;==?w&NIzbEXQgPGl(s>F&rJ2Vm@t)F%dI
zw=^?)kKC|xM$W$-Ni-DRnPuA)($8p0HgzdaTdB7tr$EqzEE2Qd<uT`-%nTY>jcPbj
z#zWgmAZd}o*V_kcKfBbjQ@Eu+ZSiKJn7+$<rZt`D&piaBOzj56V}U(|Iek#hZ^znC
z335{tXom&xf-}S-2btiwOxe%DG1sALN^KivVRih8KjzwPE25ymsd?nW?&uhTB_Bw=
z6Q8}f#?ujj0owl-PP(y0+9VFr7a>UdNuIfES5t8+c)^s*q2kq6VDJX*t@Y+}vmEuZ
z`^r6R;KT<Rl=0eGnf+#Wd}S^i$G*eVT9+}*@)B;G*&x#Fv}4JmrDsj9w>Z<m6$jkd
zvE}ZuC|R@wGq-qtl52sbY3N;T!h8SMai92p53-6!yHZ=P%k-*(q{kLU)(xdirEG{w
zXJi@0XHj>C`A-|6-;@lC_Gs^_C7!=xN(U-RWieS~i3<&R+FdiN6Gs<5Q(s9}?yWrP
zJ-3@uS%Qlir?Sp@{ZdM;%61B`grXXJEzN54cW>(J1sVHzLei{-DV37~cOGZ`@a$^|
zkV&eQg2Q687BPBc0r!n9L-XCl_B|C-c@pOc`qgE(nclg8+n=8|qAukh8^EGJi{6sB
z`--Rx2Xxd+cUh2nkyTI3$+lk~eZeVJrn)IUlXW1<kDE{$s0dq*1Ma>|_XE_E#%>yX
z+L}Lx2$!vARz&v~T`k&=-Q0ai%nY*+MU_n<9qY}t>Gx+;mn{NK#8JFOsiWvz;)hmV
za;O6{C(sM>nYLkhNA1mE%Qo5T^`ro2*xj`p`gkmzKv}a-%w0?{_qLfo#+9p#zR88P
z<;$Nb5Hl?)^b}5oNq*#v_tT(ruare&Oi(SsZGQFfW~qMq%=c6mP9gYeT@SM^i33xI
z6Mj+Ya^c!1tTRZzrwYrpmIn@S^fy6|WF}&Mz?HR4?*EB8S^k5l`QKj8H>^GFc*u&>
z`$_Qz6a=ts{-xZZO)~kSeB9dPBz2s;ebnf*rv30?5f322ETNu^an7$)*#GnX0XzpA
zxU!9(bJNS)<Lmcv4s3I(LU35{`Qv5w20rZ<a}pKHX&3Ei-g(Q{sh8UVHabd$uWBXC
z%k2l}U7AK`4%9|(#yd5MZa4i{jf;ObFv>AmPq)K!`uUCpgdYk&>PYA>o7kQcA#pHS
zx9jJ-uzpta`1xA)%NKLPB@b5NgqPg9^p88w)AU?k@qg}5WvQYIKCA3Lh*;Ptkz9T+
zf4-mUNJ*)wgZl+7Vj&xHSaGkaOz>^o-+@Gte~s69yznv+;AF;5Phg5n+V=m_ZluwL
z0=#Yse$&|R(>PDqNL%t*V%u?;ZSpnZS9q*-vslPrK{S>gaKdPQUY=il&Mg-Vlp7i&
z#6YdI5C+XSA<+2Y2{s|=Mox%0$-C>wo98>hp}x^8^1}6!<Uuv^&|pH{iRUp7m>-hK
zRENwKtE2NJ1s?HQG-5)D&Wo-gN1{a+t2mD-zQeTkkD=&Y&>1xA{`n~KtWesaSu(;(
zeS@GN4(}0pXPogB9m#a29*TR5=#q>?>1ux)4=6DEGq&L?EPWQ%L3<-KYt$4x0mKF4
z9b0t2)DI_N;BVA`d=Vvg|Ff18oC{qKT({Oc@4ng(C$GVt?k4DSfBuAV<@zuazrYN-
z!&YykUCsCZ3_t&@_1v81h?+3Rj)d7<k&6x^;6&xf>v?nXfpz(G7nvtF941c!PqVrX
zP&tl1$X2!`$-R!;bZ5C86)`$MeN_bLNNTrmp>Y-T>YL#QZOiUBXaci?{<H+(nuWwp
zHLO8CgK~8exqg|*$oZU7x(Y)LqP#f)a~%*e;!XX}qEoyuoi`jVspy>Bu|mOe;5gVX
zcDYl}5A`oqY5Ef54nE%(#v9bHIdW;+>1ewp^uHV#@ZgVj!@*vw=s<nAUrgHf4nT}-
zy$VJMrzU{#C%ci5ueA)OZ}YrA7IoECLA&Kl;Z99^=l}D)cDI~AcU{?jtcix;l!w?Y
zp~qj;cv%lM8uqt}MqqiHB7E5Lh|3Lj$=In%&1bPCyP1%$v}_>0TxX{-h3epa3XbaN
z?Ztd(yas^NTR@E1ZzyNhvev6}F#|f_1-oJc%@39`paw+`^+)ZEeCdIMAeoeLVus{q
z2VkKqRHOfgvA2$@V+r0x@r^sdEx0>uJV1iG6WrYi?iwIyaCZ$B+}+(FxVr>*-kfvp
zTko!Qf9tLL*WS!-rmK6ZzWS=ByV`P?s8<}e$26KRC8P%;?Kno_;~}#-dHj`TZM9OE
z=5B5Mpb@QXPARGfW$2sJ2PparYfMfn45Y8A=rR2c{lPks0+_w-ylBf?5L^fZx8LUj
zPTqn%nv6s%B}l@zj*Li*e$Y3i!9=CO<g0AP>}{nXG*bh&I*9c6-d(QgoTHRvP>r~y
zNl>X7TOKb#j;F3}P5|$D{F=!>RU$hTB<jYBq#-+0I84D-FsgNEW9psHDN>BX!1qBo
zS16e(rf))bLYRR2*BG%y0<}fZZ|vxPg_KPa{q$fnkH&fm$#YZDURdaDZx)qKr7x2W
z<5NhK*Rv!foI>`_M?BNPzk7!xo@sa0O=HGNTkZwwGS)U4(`8P$T0PQz@5PBanQSEq
z?Pymp_E=1dCk(z=Ene(B;)%{~a4eV9a_fIZX~5$!82u&dP&BLYx5CT>ZaK<Xy2Wmm
z-qIhHu?)TB`MPuMLROT#zE94wLqh4PH~Z}+A4{WSkI7nz!>8WaGB#s+(PF>ujLh?k
zt9OKrF>2g!E`vyb5VAAW>kY<OnF--@i-=^AY%s}hux-lmWC*3%N14vykv;B6L2Z_O
zk7nC)hcbKRJF{ky5Z?+3^i2qdu+Yu98{~GI-Q#uz^Rx!oTURQ~W+JO}2WCIdhjNAu
zGu@G5#}x$AeJ0{>*{TBk`>aqXqQLbIJJ1{mr&+23W(Rp|Ix=z_{?C<#W$>rhn&|wc
zJ&m)#m5#{A3jyXkxFqx?qCnX$Cb~Y@>AG+e)3B{`e;mZc_?T!sk3D$cbzZ(20=}HA
z{DwV3>EDQhXjr2Q{M-{CNY)W2<0Ba_zqR7Kcqz{i#XB#N5cV`JMc)y)TaQdufr%%u
zI&-@t!fs<*2-XN^eKt6{7Ian!HwALN7D$3e;eo2CXLmH@>8Rywk0xi`J}gJ|qi`kK
zONc-imN3%Dqu^W(%?uZBuC8tk$=o0D_`_kXAhzww;p+h|EhLS|R>_cI;arU%rMY%B
zbb}Vg@+jk&|CKqj8+3^^^A7n^5NdphrELCi>$=w$%Z@mo$cqTOQ^(1M3iHQtK9{^R
zXzD=QhcBrzZE|+N+g|Fgr}^a0{A_~G^8H7gh&e-723sFXVXO7H-*V(Pr#!ebW@8!U
z4Au_r;(7}9uA8c1vFV&D#O%D6@N{CX%EvjSPSaJ%Ph9ejUf<Uh;ujdbXBbR2$P;Hc
zj2H?DzDg8qj_3yRpM#?qkIT!8G03&t8@|h~5ojE@VdD&qDO@BJOg44mK@cY_%3E=o
z%?q){w&Zf6Ik5?{q|QFtejF7)HwMe{&C$8h4PSfF?wl~C`W`jzzI^@LT1X}66U2rl
zRxAADu;~N0@{BrOv)cFHe{)|?BG{ng!n}A$Ii7ml)aMft!#y&88(3s<T3Mz;J}&XF
z?}u*-q%&3?Uf0`d>l6(Qdn8myT%FS(6mWqraG!*^!=cFOzv5-&5Ev}kLRnQtTN?JD
zK+_fZyfRAhncBzV;RI?)dTWQkL*jVESIi=XX(+;|2oF?Kd`^k5N3mvm3OllyjR=dz
zo#7Qsc0tuR$~g6Gx4nc=WVLN>`F2@S$ORYkI=igMi-_V>^O%O9ENSC%-1b0ro-^xf
zPgXX>HH5lGm5SMtE~B-G=F(Fc6~N6}?qK;-C!sYUbAUEw9r9?ISN{t_9`g4U<na@H
z(L9~X@oWB{Q4*JDHjd4#oU23mZ#T5ix52%h&ZO8+$BBC#P1DcHVcOISZHgL%zj4lG
z{S3c3*DZNJyC)~6n%5h>E~I&!G|cw#2OV7>5NN}47ny#7R+d_=my8JCp*790=V6`<
zR=(Ws9-$&%;uKQI3Iy})qLyZSq=B_7K($2paz##ss-aps1P$#eh4;EX@}7B61G6yL
zTE6IHolEFbEnC+U$bTb3w%3i=ukC|g0TqmsF_7watF&{KE%AZ^#`kFcADP+zkPqZx
z=lH+cMA=U`98J`CE2OU^@kkc2yR!o&3zu@FMEI93?i+`9gfYb@m+vf(uL9>XacjDQ
z(0hu<M^;y7{B#|!HGj_r!;-7evL|x5dl?Za8ml_(1ARAdsy1(wcg%tf2>SjxHhGld
zKza5WSd2&HPa@{0{nS?+L&jKvPUG1(E0N!4TpLZ|QFM@CP(qVWI_v8RiYF0?p$5xK
zpOa7X{gp)%czwDp24#KUAG3Tfe#h(dd3(Cd4vT|50;`DIzdzh<`YDp(X*LIZL+-Zq
zdvkt%-MXMyQWsT^4Q^Wao*7ITK$B~EhWC)27yLk{JX%L?0L9%6-gvu-ZROGT^<EMY
z9rd3w*ZMhzJq#hmE%{sIoo5&+ZvCguJY73arZ=K$K72l2uG0(&B)lSyh)RLrm$a>e
z;+QnAOjrS6IaN8%5rmM0cLdG(cj$?!_n*xt>*#LT6Va2q)QM)%hxmue%#IPnJPb!N
z6?j?$s3S*bH~7PfBZ4L@XWf{IpyYQ1%@!Vg<|+HQX$~WdW7bf_xKph&hz|KEF+PKV
zGj%UqMmTSNtIag>w9HbUtZldeJ(3+OZ=q-nD6v|sk6IXl&Z{~2{%czc)AMZ{?#Wl@
zIICj}x+KM;4sd2wNNmN!{AH`t37NC4hB*>_ZZBuldQj2qvH{Jrup!E0EyU<7;`sTY
zihnOY>fsni{kAvnR?_?J8s>9RD1){eL5EEdOjQBe_P|75PGfICIUqmuZ1t22;0d|;
z)KES&plBx;ZWTPF-*fz18fhQwk^4Ec2PbsD+qXr9Hk1qRGR;_>5X?Hhr`sZozT9LQ
zdV)cyAIxo-g58vV*6e7nU28E^!tQ%t*K@FQv!^p>Ol0JtYl|<)|McM4#;bzjeqbn0
zF#coth%rSM>pU=v^@RS2Lxa3Uolv635VeKp2(9g&mSc8<xCIiXY~m)W1(H)X(oGS?
zVjSjibLdKd!pu#SCRP1XP?o!S&R=P>r?7w}5&*F3rQLLjKOz0n-3hKc#}@`QGs?2X
zmF8V-5#XAuG>Z$8aquWD)<y~bW;|pVaKA*~F~nx$$B6!~*`aC`TUH~AgZ<1uOy#%X
zzZsJBKw%A>Sv;Wb=Thdsx%RcOqhZ@jCg9Ges2Q()Iv@W*1gv?%;4^b!G7#jL!p7ot
zD;5o6G+KKc$Dkc$tTM$YD4Ht1N`^Dgo*w%AJt~-!X>Y!T=FGKT@lp}$NalBP0d!*H
zh#uAJ<HunJ<t&!Gq%^IQ%G`6r0<pu3D0+j`-)05&YI!El=NewyEY<S1nL+8`!FSLB
zu}AsmWc-T{Xx^O?2O)FbV3h$Ys?qcnzABh6y!!4VmBucE`IUL<a1Ox@hdCimKR7$I
zZLMs#g$u2Zak9X3gO)P3>;ATh7(}H}PK(5K9!OZ(=B^dA;|{7C&%G3u(d+PnqIw7d
zRKU*qUsvII+yVTjzjhOxNIB!BvW^8+gcg?^B(<g91P5S^n;}V$v<C<9omIcXPlcuS
zKBu0r=^5MQ*b~2*jgc244iXujN8xCaw+^-$vERY=Zhz6s{0oZy1G$psvU!pq`;;;E
z++2#j9^OfEjq`SA(y>#`I@t?iEYG6-@RLHyK0%s*@>DX%oII7YBWv0*{kV`y$SOBM
z`|IrIx<=e%DxaIBBq15s)3NA=Q7XGZ9r>}jgP!xxhi9fi+i3yxref7CpAm{6yQ5|m
zwrGgqDf%jd-J#n%QVVoou{mC~Rx(ZHLyi3DqI6SO=P%A{kc8uY8wIy9dISzfo77dP
zmyr#OfH`u8J0Bbp5{z`C9b1m%FMjRBOM`3}ie7@HM;*DS!;t3PW>XU@nMlP){+J}y
z*e7Rad9w-2C?zkEO8gl<56B~-cXnZIxRlQc#6A1yzJPGYh0Jg`!^X5`yPWrlx8U%H
zl%^kCTs~L9ciRgg(Xq@fh=*Ggi2Q>wg!cDewJA@n7-{=j(4R7CLn3KTUyG(0osc&}
z@yXRyGV7%I#*s%VBM0tEgi*-<ltV1LWJf+bn30jlh4?`otfjL&gINvsHEEc-O~_M!
zqipftAB!PVqD--{pJo?pD$4edtngm6qN`n$TEGd6T2V)PNdP+o<#s9?WxxvRZkWsh
zCqCY{0kpx@{i{60)<0NR==x8EB6Ad`y(d`4Xz_nIHnCEW<2kgj{3KVkuZ7Gk?4%*%
znwDXL3IJZzbi9fSzkQDEUOwwIJZcOKwSQmty~A~?glJG@b?>6ILuA8+xQ;4w|BBhI
zjxJeWo?S-@=SX21A#L1*oXFU_1%Ee_NwHP#m&5Eo)|#|t?>I*g<#Q~VT_>Az(^jCZ
z`8ZqHsbHD8;b(P$IxfeHZ<?rS?)e4x9`P)VnUkX2_h%vztYf}0x~4Dw=?I%RTH6-4
zzKw;o<zW^C>1bh0OpXq!SKCqV#9s0m-rJG!LJQiKKlQw~VZSvoB<?d(_(;{RZErTC
zcqU9+a3zazoA*1Dp+Y#uXl0K=pONyJ<!JiJA}hOZvO9Wj*(a4y+{pE4hUFiEpp9|{
zHKDSRSljY2&!Yo~-O7@ouLlY1R!11#HyLDt8ov@av@J`A3_T+ilnW$xs)F2)Lbkj~
z=8>Ng#gXG<9Gq};XjvSPiiVY9Iu**AI(6yw;aYSnCfi|^sVC!SNj}P!em7O?)3Mpg
zsM5wc{^k{`4hC~AOi5CmC@mkO&A?SGo9vz8wfsz6RYE|rKe4<_TXIuUA<Tgp1m7{0
zOt`s5CjVJDIeH$hVzrj})|RSD+@8$fy*}T{E@rX+X~U0+M{TB~Qf=H|M{-h7?D6P_
zYb|o-G%Hets50!(2`xFplCtCuZ(BX;=2{z$Z>CF<zDW+9--)^R(V1<9!{*6XC-g05
z5lfw1F3EtHMKcKB#16i~iR#Lh97Ti=BLv(Op@A!PQ;7GsAB<Vo;89jgdeWhtW`h-`
zdl>DC(^d0VF0(Cr2Dd_s+*GHcRE!C+dszBY(?%B(Wr(0_EdTBnWbFF1!u}v|Q?=lv
zk!~AbG%;XnFl3jav*!(zR5Slt!Cr8Awkjd5^Q%0rdiVS1NR2`Z<IB?Av}CpFNm~Rb
zj%Fs%oS;wfI;ZG9<XVeZbVj06N=j`RT=uT!U3hmeR)OzK=El=y?uV^>WwNiO$w<A$
zr87~TraF^1goS;g(4*jWztM)W>Ke7HUx!D&AuY=qd0>pOUpIYth96D5ON(o%j8;@l
ztP=ZDw{MGUlhX4onCMbna32F4?X*LMU+%<KMJf($!aDOPOdEGtEW;XRiRIbFWSH2~
zqpyWum@fRIlv5#B_GmWe5RjCie<x%Ly)f8MU8AHSkT##ddEi5Hg%6pwZdD2EU}6`#
zn+b%5o(W%}3Vk7`szu~xD6M9MEI@YG&t{&HNKc&45=zH3!{`<edqgUbJoKE68t`vL
z)Srx}+NNj!j5TR(PApP-d&?+2@VaE^$8;(4xul36?k*H+&1pXPDU~$rsc2nEX?#3Y
zMP(<U?Q2B45@$_UYwQ<3au__ZM=*b5Jr{0~8gEgbJ>5nrQQ&A=-cXAj&iPJ6s&Ys>
z_BjO_&+B5@Sh2_I@aBNn_boCWKV_*GVUQOa*8fC!x!L}|iX#8-3vf2}|GV?3`?bfN
zR=JUZROAhmp(NDi<c~p)b-9ex)jv;3!=<)PmZ;-}Rr-n_sp8_BX3~}ghyEnezIZ=K
zLh5L`ZtySrc!Jx7+Wa6-e>oKVI@jy`7GW>>R!*tUIj2<^3_0KacGw{xU5Y|VBcvz|
zA;jqC>RnXqmO~VyXf8B2%O82{$Q_OL+B3*I(nVuUvUBZ?!buu|*6Y-^z%IG%c$|vn
z{q*$oC?cNUI6<mU`sRZ-&hwQvXF`*SQu+QMCX7Yj^d5KPWmO7t{oDJkm|yUgw0qup
z_IXU$Mp;TV>j4o`Aun~Gs}Xof!-RBNuzgk<b3@b>ZL$|V&=7tZoMHIcRU3`DL#1GA
zh+<AS&rL{l((wo4j)d(F#ev1M%nb#yTY3y!g$VRf30-p~sN(h3p~2_yYRVsXo#VSS
zn6rM!zlP}Z?xdK7dXJ1AUKK&eA{bKIZNPtOAu8cM$iBNc_}8{|MDEmw!yfd@Ll$I_
zx#WVJzsUs<sYqhnV%@ZPixQ$rL2(6_z>x{rT-10wjC$FH7O?)dTryGBX1OFizx%?F
z$RQ7zB9aYV3GuAsg_wl<nX>4~l~1>~lcpm`kVL+}M&M1kIqM39kqjC_Cc?UIKAcMI
zlm-&VLKLqUX_P1kJ+2tv>l`izWHtnz5k~RgoN~g?MS1-5i{J*RHK<L6l2R_95$6;S
zx9-2onRYR2N8k%ecq&8Abz5VJb)?W7CfAUs-PuQ>CE^`o4=U{FDjhwGh8kz}Sj*8?
zWkGqNHG<Y_!}YjZ(MP(=Nt{#BZF)KFQ^Z>Y{x&pMkjJ29R4CnLF`{b*4uP_54Z|GG
z>0#*7-()Vq$Did(7zk%YUD*sv7`SmAY#0Chle6&Ys_AJg9{&nPAVemJg#N)XYvR^L
z8%P-{mYOsR`ZmRQAqXo%1!{BbFx<jA%A4sLzg46-oIK@qT|m5JYj|h)k406@;q`f>
z1`WJuvu!PK4Im5i4G3GprU^VG@&$o;oRj`+HyZU+ocw_5e0>p92qYf*Q8h;}^q6Y@
zZwB(@pA01LXJHqz?^jD2BPhGr<LM1I+8kR8d*T5bkWX4vLkHbe2}Bk|$v3uJ42(Ja
zsze+Qr)*Clx*DqZ-ic+<!wza2DZU{HoNB}+QnnBY)xDZ{zq(H{n@M`Fm~^_)D_Aq)
z$Jl<UlIm=KA3$bA&MCcXq6st;MhXZH$p}d#fC-rVZp+39WFX<1X+m9vk+|TkoiSC9
z(dU$D4_1tTMB^{9^YGk_Nio-@qYRkDuP}u>X`O6n8nWN(hXmPx3KwFRdf=ZwbChHJ
zc|=zl1avsa7bx!gF5A+>Hh*36g6OFbz`IwSOeh+R-*JZ9(Onq$IdEW@Zr>(|OH`?J
zUXfU%w18e1WzNIN&&+eX#~!4$pu`c?=RS0b`Y=kJhi5KN3KOKM?W3xWV1A;B)zN-I
z#2Oo`;x}s2{bTzLI?+cqr^&$Iu~qZ_l`+?RctAa=LusIpnfj(CXxsF-Xz0lxYcTFy
z3U=}k`<cn)w8R6S&tiA9WRKSpe&Z=AHRm3B!_l}^Hld}?V%d)4s{u$(#MxftbT-@#
zSHrWD_Kqp!pvRo_*{A40e)wd1`x<?{d!!o#OM{cmIW5j{O!pMJJ&!~zA&ug=hmRqO
zFP|M@|7i0al6&WwQD?B4|HP#jBiMQF^`-TB04pb!&pyaGqBw+;nz5tSGM5ri<GRK`
zhZj>%;;SX*dJ9mbkJ460%mjHP>HEveI*x{KHpUX)7FXg2i_~dg16P$na7g=~iIyG*
z3rs{qh2S$EwlLe-KS=hvrA#5!)ls#26W+I64a1i|uLixR=ZM_&dh0j>nPK+9hkgBQ
zBIIpdI*7iHGUUBGL^*l}t|71;et(@3H3s1I24!Gx*DDI#fBwqapwRq^LaLQ=ikqkQ
za%DCEO&qDg!MnJGIp6mt+Ph*fKZ5A{B~h7WYXm92#sN>Gg=A4@r>d)dY6^Z)=a@<d
zAvp7?yz<aT@Q3}C(BdlQ2B#wuihZBP&uUM@dn#`wBz`qA4NiFdDErDsjZJL}i(q@^
zTg;A(R88S?4x+<6p$6(;-a=_xxkVWCk*AQ*#!3XfU9*rWCP@{iqDSPN7`iK>t>bnw
z-iKNKWZhC;Tq=f1new#XNTkl+xb|8F|9;b#JN1;vzi7-w8lEq3q?$%Xn0}D`4R5^d
z;;YOyj@BO*=P4$Fl=HE?mnbVPEl-Vm`^)p+@Kg<EbZ*@!-z@calpS}Wa+oWap~O+N
zlOr2hU!=wIB$CGO8&NvPAEX;FZFXQt`#0qfZ&wx1Y~fg;y3T6xV?Fj|K-WCdD|EJx
zMHyGR$(@Y@iGP0=g_wLKFOIIO3;9$R^I%Dhql~JSW-mckd0dEd&=0fT`&}kPI@fgd
zPi}A803BIo^$a7WUz6VUrT8CqoGM(R*|5-}BAn7#aQ$6{9X{oDNOsC?!b87LrZ4^R
zeq^6yozBTZL#Rd%xUu@LMkiOy`E}nnD@dm5T52x|<n+ivLt}^x7FfI7fpEVilyz0%
zmHXLRT>(MxpKMdv*Ky?KFU|85$@G;FPGzX#n9~U;sMgzLMt7+6>MwfBsB|0%VeqZQ
z3>6_+1igVjExj7<nRj!Gjh=nKhbxOJynl9h{jkM88HGrBuxlraFT?NyGwaM%Pj+Rs
z(;aiLo=v4WF}=@7m{40vZb7G3-d~N$OzBs&QvhL7@mZp`!zQ(NTgw6yr}4LW2F_7A
ze`F+$7V=NcT}TjnK^yRjxv6CkIB7Kg3-`|&Z_Y1%`d@$f|BvK^>;E7pY`kp$chyVU
zzmCVR{4a7ciVpL@Q(Mbn?Z(5=T)(DGU0gR@$jh1B;VTmUTUIiGF+Cmnl|l*PVT%Gj
zc4qBf-i;J(u(3BxqhO-kpgaBiOdI4ac1>f^acFY@nILw!?(33GTF)K55=UqmH8g&G
z(Y0}R1WE3I26hZ81}R*+PsgNt?75Dx9}c$;z6j`x^`bQKCSgTQOJUvzncLkRjjj|2
zO^93L=su-(>~<zM_I84JVF)us4_c~(QIFurX_1Q}ONjc<4>r~sdoPrasXjsCk)37K
z!+t_lo$GLBrk8Dm^JgTq^<cf>!>jn-nl0?>d(oFDEF^#DsXP&sFl3;I>UT`6%IR8^
z4LKWho?6IFySv3zSC`OaS~@kEgxR|OdnMfR6ZJ}P?>A54ueO}KIJx$`Y0BFR7Qw#4
zhYA~?m!C`&8p<0dCndxRL{EwMz6ky3xv^ZIW=-{o;4@AP0pE`G&Rgnm`y2|aZX=@I
z_gc?Ieuf*L&E^9YoG$ZYi%(+Hs+x;^Fv{Q33Nfb>l9_>~pIJo6<oHTXk#s&N@T*Wt
z>zSFRb9UrpNo(}4b;?V2=ixa^$749G5%QX!&mw6lze>d5wZc7(L|}c+VFqG%jbiza
zHN3r9XiejRm0~0(l_DvE7e~B0DaYt135Q7v*W@Ipk<5Cg@s@Rl$)6e@NVqg+JKW5U
zNRVZ@kJrQ|s+vK(U0>jI&EFXx@OWXM8m`zR=vXtUdO5BIre%rX76((1nJ&2u7-eBP
zr(r2VqL@Dq*nQjD(2n>Od7^vVVSr~yEPfQpMMOtG`Zq4TkWtPWh6#`LX&QW|WV0)Q
zHgca&e_LDt<s5XA89UgwcY_7w4_~m=vZ~{rkoUIZIbv5quIq%*H&?nzF?lF!Bc|j*
zpwl!Ckn_j7AH<r`*K>HS%0AnQ=wjE?w|bcyc?P1yLQRkz8K|KrIkVANi*%7r3`>@(
zZM&QU@`o`<(4(77O24_>^&lEiNxsO3_v}9KxnhePeYXW(A7`JmU3dDsk#BTFyZzGh
z;CK}LWpKc|X=L#ua|$S*!Tbib-mm*N8W+4*j4V)TaL^VwWiuW#(v<2sawmQ6SfrP`
zXZKLzU!~*6Ox`ARj-jCiX^|$tfLFx15Ilf~_^P@k0rz*#KWNeap^fDYgXRgBVZ#yj
zv#6=J^+&|lgsl?RzfCdWC8g}vg9y5m64(;N{CQ-U8b?@2@(d6~uXq;YqPf{0nEXC*
zZ#xyqP&?JnV)xA^PRBAh21fr-ij)4Zi5S~u8p{zCpi@oG%=(v)3^(Z{^Al8$vPB1!
z>=gY#JO)EOyk%7r7o2l<=+5}BA8xf*P_nI&sd)J_n!k+-!QB=(%L8LNzqedCII^|b
zXU~kN;i_D^*q5sO@fDS9#$+vnaK4{GU(dy-8YZUB%XPOG!$e~f#D~F@1YCr!C?H&Q
z&^ncwPGj?jYNvk#+ivYL)%mpk@rKQqv$;90Y&eXv`}60d@oCi|fnzzb@~mMzrS1f`
zkh^B55(i7gIb^QeL#4Gz836LzF$ZqVbDBscLga|RWQC)h?FIg#$78}%TMeFj45el|
zH#>iqkwNf;<Zyj0N~8}FMtemy#ZK*H^jiirZ@L4ngmx!WIvauRT_t3PcR}FMrW2oS
zPx49)1J(Qwk{Sjt{mw@}lA+4f0%#V*b7?}96zpa@uDAdNb@j`IHUYT<!vhYO^yPO?
zLci<fj99%$qc+aK>}97rL4|c8&l}SA_3e?u^;r3PNona-elHmk8GV-5)A|S({+~jT
z&QND%+OpqA6lfw{Z|X^*XUCBuV;OaRBb~NAz=tD5-}ru-?B={U<GepgyoOurG_3Ak
zCy1rnbq=8Y#vF7wn<dWJL|k-TGwM7;*u+#Y>Ay<OcC9b@>&ve(M_xX!n(ZHhUY`D)
z!sOm0{uS9~&Z5IoSlA9x5@CjX(gnZ#?KI`ZXJzE>KMhhzQzl?Ke-tuCQxM)H(z&VZ
zGR%y4%FtcfQFdv|6fC^)D%u*#QZ^z@w34wmbqdhPn46?z>J@{&q7W!<xJ;PQ?oyE!
z=v9dlF)rxAwR?dOI`G$pjKSm7%Ph>=OPDT*s{CYaAH9p&ho|hxhc{#&J?3ed-$eP<
zN9tgrhmuBGRK-?mvg-N!`h{AGg|becwCz1MwMS`k{u-^UQvhXF88$U~0M`6vbO-|O
z7NV$kSkGt4bz0fa7p@^vooIm?3sj~DKBfpuIX`$%Mli`O^BFPP?>m`Xv_0Z+15a#-
zwOBka`~)qT$l<PDM@27Gis6q4zeODSNMwGbv}KXLpJ!=1#N(9t>e#DQ|7DV2zV~&E
z{~J%9ip3G}F#&nAh%N=?9qVwjS^!m*gmMo%r4FG{-Gyo){1HkSHo0TIbX<OGt2Cs6
z&Gc&^{TB->#X|USLbZrP-2k!tyraFa0kw$ObutuO%oDYd^&9>R1ewRkoG<x|!EW@r
z(pxBG46t|KLy&q_3>TJWqrM~j9ZfZXFtu&V)7i2(HM4Klzvg=wUD@&zzT)NnpPV_*
z|5@XkofZ6lx6~deMHWNBh2C+XapSnpI-urTD^^h82I@xP<k|E`Z{23i`9k*hnr_9k
z@yvzKvM;xo<D7o+#95FpZnjuTUY~TLDvU)hAuHn&j!3?WlqLrh{}x#A#hMyK_ABt{
z*7j-ZpN*WEM>ERwUy}1Dv(EP)3c9dh$6UJq4qLk09Uk`JwwSGpk!Jds`F>TA#GNv=
z%=V|<k^c-Tb{BrTc{&sMJ);amC_J$8uo1S{d(dkU$VCfLtiO)xsLOypvJ-s6nsPL$
zUmO#5F(9Y??oReDX&rU+0v(@GWa-R;%~s==mzj3TX%a)ef_%cvSu-!}=bILqc+m;$
zRZ`OBKbIA<ZLe0;W(SE|aU7D<nWde3M)Uo3&dj!a(jPo%>bZCF{>?5vlaWH`Ud(lz
zTA}M~)~xbl0G^HsR+?1o=8=pj_@>QQGH!G%8?C_R>zsFSn7;RQ(cH~j44y>M^LdYw
z!j~l#inD0wPbJt3Cj<tA6VlNFLb<n-QT!;aE)#ZKLj1JC40F>yaqJc1jaLWQ-_a6B
z*kR`qB@fES4sXykX4c`yelD2ix{R6cw;+Mox_>@?RG1m5UBx_9X_e4rN}8zn(z3_o
zW}?+F+w#>Ug2xAHqF&RYP81U_R8!J9h@oJP%K;~oY#n0`zrii{G{COPV(0VaooJou
z&&S;_Zugbbxh*$+xgnU_p=i7i4f`P&4R~!1c$MAU2gmR~F`uwIk{>8T(nZt#4Nnkz
z*}5M?j&#>bEeP(b%9)6;D9so2nH#v~T1=9Wd^O3`Tq<I#JSmy<)7p|sNA#TGNKUi^
z2Kk?#V3NGmv-GZOim{8U7X}<uiy#<B>i-H8FY%kjrqLRqoH9<!W$i<!+|z+13U`)>
zh}H`Q7hqYaglS^N5@%RLc5ZfLILpL!E33Q-zXR#+!UYPys#-5QR;5^qWsTogwCndS
zTU;hz2UW9b5(?3V*yKjeFmr70^Jg0g%ul{JRQ5ft|D>C>RPN74Fa!_q*Ma-z!OU}9
zeyU@=Trb!#OVWvT49X??DtvA@FGO<A9qE`w!jv)%zcT~L3PPt9c$^AEu@_3>WHO_d
zp2%0iaUUVG2Wg3%;lOMQwN-n{l_hmW*IG5C3U|*tKjqho7QQViyW8S&weFs((jO4$
zPOO}2xxo;|rx?rA9+<WgMspe{FE1TPibYexidW8AJv&uEr;jA`rFE%tI-VOQMVl}|
zJ!FlF@_L`%%;0>!9Qyj%$SpqOIzFI?O@tXcZHBj>K{)0drq{M}7o)5G=E$;O@ugYS
zeusud)<M1En<JhVq~@QS8l5BJJu8L}Y9vu|QO2!fVb+KRYHf{re-2BQW(R#fCTEP2
zlyeDAH=>vpF`-a7e{77XoczhPVv;r07{#kWbC%C+r;TRwM`^*zoWG3GrY@h6pR?H4
z%v$zHyUau=A+%kOvi>#5wHyQzMVGvvS<Fn*JmNTG`Ia7ItHtG}7Gn%CF~%qImghG5
ztaJS--e(rtBjy8!)<K?BSb(TLU#O}CAFn0ijocviH9ass;@ZBzdq&$ztgCtYt*D$f
zHqRGz+rO%Db6+xyW}4JS193v1UVX0WK5rx90y_nL8XjD4v=x?*58~4dQx{uRG@XW@
z_uH5;_EIx#J3rOBi4L<`*}lvS8gw~Q-WCx{wfl7JpafNUUPjl?3h@Yzmg0Z(k<g|6
zCXR8!s((0~GWxY%$)Zk=(2V0Xs&-a-Z6N+<k0jBn7gDP{5(Jl1f1VpsD@kyoS(QNE
zf~sIuj=|+E@q;O20xo5^eQH@*@;h|i^!;}|iCPZPR6gfLg0JZ*^QllT3$it`v7@Ip
zGBnE*3C+ECL{$vV6pAaRLp$~So6&4+G+2u=j^3VZl6yGIL=D$7&GkFW>|GL-d};eR
zBFlwd#$2eb>vd<9JmoFBNd@!2?4hqj3V$r4igyt6Uo(Wx6JK>t7-lSfWAz#ow&yQD
zJ&vw#ZZ=Z$h)<8Wt+o~$R8ceFUUT#b_bl7vAk~UgLGR9+Ka2ZP9{e-Ob)o0k+XSds
z#2%T!I1X9hI=LMJQWjM<wMp2#?`T0(U2IZ`&fzO&Yn(>w#0fA{lc<_SDv=)~9j+Uc
z8!d6*KTLh|T~%dL3__nRR22=;@g#loY^=zLS9dQK-&3@wmpBqrG<WPjEC16^j>~!T
zXKe~aFbdVI%5DUy*bt2}L0uJP0Ue{W*U=LTBTfrc55kyhR&g$H{V1x%pm-t}tBCq}
z>W!XqVb7iJlx%-(N3b@F0#^6XEa3X#-0T`0G?TZoAL<j^(WQvRg*dShLN}%UnSuc0
zL2^%3(Az}lK>hI-B}SQXLb@G{b#QHaiQ5W$iA<nBF~~++42MLAWoNcQ9b1RBJ*ION
z)h>gd6SccvIc%1Jf&<m;)s2<MO2M-@K(4hV^KU6#fsLoI<P_W%PGz8pAiAU-N<;Er
zHSzG(8yq-vxSq{&8`2*vD_c#OM#I>Bw^Mq<)pIxEtgE)g5&Ew^z8uCQDz4S+QAhOc
zGjzH$2F|+Oi7XK8MiaQke_n^C5(Elv_+xX`e$`a>8mIB$R=uW;8qAueDR&^gy?+Y5
zQ~gbK`YDuJ6V52d8T|@;d3m>T5o2rpLFz~DOHuHx8xC7UT&#~?lWqpxP0zW~8reAO
z2R9S<eE(y|rDBcOkl+u9w}KInWK5W0SRyqkjoI%<83dM$P2olt56Kj2j|{^f#Gd5D
z^9T1>NYXaYUXbzvNXz~ww!!_s9ccG&rF2c8RJy6L3yaD(BbR?Fr%Txynwzp{Sem$4
zkO4dXSR_p?%`IHWIDxOCmM+eUrcPq^whs1orgko5U=}fZ8+#`e2Sa01md~cHmd2*a
z5~BZam30m-ZqEN`3{ZjH#>Lc$Mcl^F#q_hOvAv0@kPxD?i<7CLEu#CM(@c;K_8{&c
zWX=5tV-_|m#&EwV0xs$_3r+$r)&UY@8Kd5HFKn+Wm(O3Yl|&f7qhN@9Qo^nwZ+8Q4
zj;z0LjoD7*_a?5j-#@R0O|tVeGJ0LKUyPlm<2Mt9h~_HNVk7?hq5Ajg<|c#7Iw?J!
zaN`|AUw`9hwphQ(K_u!RVc5iBLaki=@3|c%4NdQ2i}4_8okK#>-o*u)-p=7+NhQ`q
zwy@vd?S6M-r9#YJRveie`_9)7^x0s?p02LV4ZrujXFs3)G5iOK2-Ah)Y+tJ!#rZNd
z;m6&d%(^YIX`I6o6Y`XprK(Ail!_I)t*$Lk#pmnoUM~-ShlYk`d|5w!gsRN&?x$5P
zg^`CDsC7A-fzn0BMn})NTKJAZk+V87f*75dk+JDM98Y!Jl9k3~9hkj+A4tyxdiMJz
z9z_u6Ua9kNzu|WRT(~>qzYKc1Jy~q?aNb^XNCqS*m>~@&<T4$>)_qmsFrNrUYIWR!
zb3!lV((Uvm{!uIuLqfmjygx>s_1o=24~M5zqlzRUEHt#I>i&EUx{BBJ6we90pH40V
zjPdK{a;x`oV@3hB!{b`K`EtHoW41^x$ccg?KN6S4=^$GaHLtGf*RNmFiS%kEisW2e
zYRxZ^EFFid%^YO*b0tb6dsBii#@5q$A)%EZ-tRNiATd$V9Gph;NfP+BsR4#d)v4Xu
znwsSfpVq|0`Bt|J)%B?8=)3D{8A(b1ovyB~e2Exf>ImEDczN+h$iY~LM8du=T>;QS
zLPDn?4tDm5EWuXmMP>;J35Q-N=vE-V<!!am;ls?#{5Z%u+AER=h}^5KDDt@Luq+}X
zLSLcJ!ot#o!0&b*obGV4P;q){bGKGfLi-@8tc*z{?0yNs{IJ;_<iLQ~h6m=*;a<HI
z3q$qr^ekI-_Ve?D2k%B+t4K0_C=Vucc6M%x|4F0J;r=)86Kc<RV`F0&pF`N}%*+(A
zVt<)Jwvd6aMx~CS%=Y#+G78GUL?*xYgTtrX-xZ?w==5ST@h@rd@lc^<0x4)B(9&Il
zTMuuq&(HT40G)n$a18KEOJN9xV%509*d>#;CerzBg0RH<e2Oe;;py2T&EPnaKsRM<
zbkg;=&EvX2nNHh$XCO)u^?{Jbjz}Bj((`<k_iD&sZzOTV&vf7iK63ABn@6qh>yyov
zdA8375|YZ7;w>X&p$`xJ_124Xqur~ktJHF7U9Fp;1(fgC-PP6AmAb8EI<jYf>eUPo
z-M$UQl8<~VQpiR%R}oSeO=41;xZX?F@9=$1t59inG7a)J>P1?e*K7AIZR7X2GQmL=
z-N{w2_-YdZsOz3zQ;;am617f)4Zb#%Y9OY=E-}ArS2r+uOO2A&<A467y`;TAZqAje
z<iey?1%c4ixflK39wqiJ^g6to<umzgLPAjp*IV7t02M+7X*b%TB*CrK0g4>Z?0Piw
zr9VFR@qSi8SY#_k_|5(Hcs`j$zh^Q_(5noLv}F+afoU^XBAO_mVdL@oAlFM03?E18
z4)TK~2%EaF{!>SXZMV{Za_Ze32<vvS4x3KOZG(p$jn5(e__6ct$(pP_EDQwW2OtOv
zA-8J9*Lu&}<BN?>4GNG*t@&iAoB2eB79J=Jl^BuA@BNj>WdUNL3!9&x--5gx27S{K
z_`!cpad2&|-T7dGJ{&ZeEqt-o%EO7b`JubN>s{XH%V0F|vMcOM=kBD)#foljMg~zJ
z??cCCh15^#Zn+?i)n;e&>;3UsvvK8p1K=TVF;{5TKplVby9N*-tOuAKkrn7nD79d_
z)h5UyC={j!7#otW)ipIWXrw~Hjdy1&5+mJ7k`ze-M2vw!=vX-@2)PiRf$Hk&yDC%i
z#^r)OPjy$Iv*lXQ1YGCa<3TnG0cVM7DJB3Iu&9BBkbjm$4--{O6mtQ{v>7f{DbD$m
zBNCJ#Mtg$}X?wYRcxYjY8c0<uFY*}-f1pZ&-0ZLgWqDhnRU67QHZ~Rvms?ZAb~6$2
z5VPs0^qBJ%?7rCr>5ltd9}g7JxjE6#U9go!%Rz9C=3Q;nAHl}P*6%Af)ny=q=p`g+
z#jMj*U9EzFTIPK;mGZb9$&!Ok`$Il}UUV0OyosDY7ImD4m1G-x%7IwWlanqin%t1+
zdASGtuxy@9CoJL*M~MwcM&%$+yWH;8c$F9%%J%Q5(o#}N5*lq+K2-Ok?!_qy9X6Dx
zYIphyCS(hED5nM)+|kd&gFvW6B|0f7DUwE99%%!JWjF!!c|${zrUs>vLEPNjBjcMd
zM8PELfb&X_U_sg?yJ=G$$u*T3kcn*KX1{^CueLxHO|AxjacW6X%6Mq9`)j=wUCh@}
zVcGsbCk+h^7ev+R<CR916vR|5B2rS|fLLDV{n*B?izSI*00zRW{9gV_)DW4p#(tin
z4(<{1d8$@Np+*S~Nvd|L2TXt*r1WPi$7xV{8EGk+B-NwFpgVAnG!l8$cD+q2HE5;L
zo+UQOqzI1T@GF$n$zoNNaw_tMp!?;gWz(FCe~g!2SRCft$%0%vk$}q~`qABo0g;AQ
z(D<C7*PQ^M6Q5AiQd3J4NCn&#3qc^=9Rw`uPlSYBWiJZj={!{ysBmy_qK1HWM?CN#
zfD$jzraRx?koJ<2k`A#;^1d+QduQBk%oZ#BZ0y?D(ujZMG8@AKH`LcNB?XN@6Pvg|
zE&VByBfL5R{78=cVx_JcGX((g|JwKUWcuM^6={ORmoIVpxzJGJ+KuS8f9j-CgC2Vs
z(`oXR?QCoSIJ<_bx&t6{SEDJkNG`oc|6;lJV7ouE(PkMcf!yz~ysAT)1r(ri$;@xa
zDh%rOh#?U~{A^YOdqK|+4ks}Yb$sg}xj|f3XRh-ND4v=QRW!a60R+~tWWpX2QBY8z
zC1N%BR6{f-Y;OcO3Pn3RyJ$f5zV2dd@oacQ4+7?_HZk4eb<g~QC@5kv!fi1H5uKO~
zg+0*yk}tJ6cj{kORaG?tk5(Rm!|)NLq_b(z8;a6xfrW(?`6iaON#>2{tJ`Ro>6E!x
zrC**Iv^jUS(x^ew&S)fv4}nq_AQ44iYGMN8J5n@>feIj$T%nN1)zBTt>;4?2v#g-4
z2SKJs<S<y2T!$V|YF<-QQxO#u0Zu`|c9~>m+m!~>jgJ)9Q&^NmLw83by&^NDs&^K7
z1qGC+XnEisk$6VW$1Bq_XNRdAk<0R{U2ua}p&n|ywo<t|)18N@)PkCJvs2-fm|#!0
z<IX^-D#i89GQY}y=vO*Nj&_m23qSY>N^o*^-rAW;R`sZiAo^UYVbU=J@T}aLK~z(8
zpZC{$gRb4J&6}gy(MgeND+6l)n$?P|n_F8cKuSbgWDig;f7-q7U$B;HYipNX2;>}M
z1(;RO`2t2`Q;Cna$vAx)?Ke<O0)WiEz)(`T1G=x8yt~TzVU1t11G?vT`;88@BFtf%
z@R%4HDyFY>{(Qs3!#0_hmkv4Uh#)`@lnq?v-Y`@F2YU9k)(k}xX_!sj7YhDstE;V{
zePUD)8SpGhirT;ey{sl9<N)b`MIA9=XV-23o-O(pQX%h$!)XeDS{(MwRcP}tGs`;@
zQEDuesWE`1BqiY?Fe)qL(z*AxHcyvprLxe7{PW{d_q$3n`Q7L>)VIi9(x>3`stp2a
zYn=@YHUVG3$jEpMJnLLhSb$2HqC)KpJ)d&G!&m9Gx48fPh;OGRFa9wO@(qKX?h0^)
z=;%Yjg|a-qXN&>L#bdMN@De1+t!rb$nCb&iG|eV*fX?vPt@=;(6`cWCu!nmTi9sg}
znZS@mf*BGLQW_G}-Q8{Ci4jb}YDr2;3J<$ky_qs3f{hE7^fUxKIJB6Wyd4b8&BaDG
z@yHYms8E^iFU8(BjiFzK?9Pf>Qz*f{%-*c;?2hzAvQkpRo$#_ja41AXL`yN~s6E7Z
zyiSzbVT_W6ZP;p}qEK2P{_}^^`NP4|iFzbcDjp#y+#2MZoHx%Gof<;WFF0L%fQ~Z9
zNdy<*q`fn-`$C-w0xOZnL;@E9$Nrmot;x~Y!fD5oowgpn1#OCr&Y=aL>=dvE^ewfK
z0Dr|w_U9kfn<I$~fXlD1se#O=k_rDUEiWIcQ2OI?RCp{p?CQmm+n*$Hfbz&y#556O
z=wLECTgZptN#BH69(|4+DG3uWrmlBQ8)`I8>kWE3Iyxzj=GvD)=Q5zkZyJk!2l5$+
zpT89_t~gjdRWST3|6>@eDCGBrg@vh|Q9rW>ToM6y1Hj0vt!0u}al$`4NYOW{J`Mz`
zR^IhT;yU01UOY&z`0AS|SIMu{ABLSBP@QDhWuF4P0=uiPi>0qVfcf2{e&Y`g4j$ZF
zpkwd5`8k+apa)9Tj>3!mROkmO(PAeTj*0vmzQhLo8^!=n^`Flf&|16K;@B86?r-t`
zFTnX<anAqWM?!?_`D$uhTy8iAL)G;d5*uk>mG%Cb-2NJ@wdC8|TfbeE7Kw~_c^A^Y
zm&0rkppCUqZsAOw?u{)@^78U_cFdiXT4J_%;|PnZ)O-MW#dzjT;`sb??e`S@WuTU+
zxp`#2(GCADnG-rRyc4j!g!j8B>PHZpKN)D69vH*__QQ~A0KS)(*Ym?eNdgE4CM7+6
zqt#8JnU#Uz1mHKWuC4%G8QBB4QDb$rgPR-6(_dOAbl&ff1@Kao@jzJcRZB~^-BXP!
z{^9vKNO--|Pru)Tn3%YmbZ2`zHfTs>*I?Pi(o(Bbxe(AN*Nb)C<`;m)M-lS8om6%H
zX|T1nvKsLv&gA>{bX1&OB%7+->FeXV5uj2m&t)-XFp<Fv7;Tn-2m61W+8hH>1Hd1P
z6|&cX03@22pPGs)?#k!+j>~ce<^@B*;%KI@fR%~q42X5`1l%tRq<(fm1_cGh{VNK7
zLToHV_u207&k-W*@Xr9~U0+=V?f|hLy;@mtSQr-g?qo6M<{^nu3u;}0=pSN*%c2*&
zv%9;?_w5)44(=l;E(q__NDe?pYrG+oSKGaK@Pi>6Js&Q~iGpYVTFpgIAJJ9g_x{#L
zip#76{&oyt93Z-;S*gQRZ!a`K_GwjD2NsQdS9dJA*xl)}7-D;SJK)5SKtNamJ81HL
z!e%^Rv(|#cODll-M8&F25<ZqS<bLz#P~`<S8$PMf>B<YjZKc7MkZ85ZG2i2Q-^K$l
zCm_(<Gyoh1Ku>5wva+(Ks2F)cg>(q`ZY)I`0@@QAX{};;E5pF%qV%d@23IRyg`^E8
z4@<K<5wE>dZMjsoVm<8_r}tZ2TwU`;mE~S==qORJzSqrR_Zy$@%L5R^_KE-jED+Pl
zAm-{?85lswb(d>YiTcG>2-vT;H9H@CxC;|{y_p3p?!7!VKR;hnPFh+T^ICeUF|c(o
zx!fm}veJZ(n-zE8diOTwO)LC!o^x`W-@DKa<VL3-+8|)f8{XF_&q?9o!~5guTRFyq
z(Ew7I^T&sVhEAcfT-qp-PINu30i>CNM?OlE6*wf&yHKZ|r5WoC1^X8$@yyhQ4)qOe
zbK-i}jtH`~vGKe+9e(38><Q+wUg#1jmd^yjJPE}7ygdIPOH0ej%E}7uMl%K$&FX)Y
zEaa8k;`cukI}(S_Ve?Z%8a63u_BcafQBlv@&geX|Z!Cp@-^(<Y#@_op6ATb3r*YdJ
zZ*G3_t*@>=1cHX=b081VVAu->KAg(!4MD<{G}P3@vrPWrE_%5E#B0pz$`8_51)7ig
zj!Eg#=bvZo4n97s>c!*<w|r^Q<&_8t0^E<r`S$C>PD~gYDfK^5%f}D^3An80V*Ml;
z0MbtF^*ZITT4#wB9f@eEqx$hDkI<Jh*FU;PPWtO@vKNsN6}1xz1X;tme^UY9++Z=?
zGcyy<<ml`SJ!t56w$b_icV}?8dpci2DmNxZ0f-)DgqZ20y?-7XC3-IZ&ZtU|`G6ol
zIO0cQ+p#@K%5;DqZ_|K@N=>Azc4dFJn6AA2Ie%^5ejvkBq}Jm|%&}&EqH|}#^FJR{
zs$M}LB^)Fq16Y)Pn>*y7HV{y<8Nrm+0(57qH>?*D<{DV7**81;lL*+{U%ZRQON)o9
z=gmU#wz8X`Ia00$?O~3ywY3G1N%C=db$~4Zrss2ej6+mhR1|0*zXJe9G>Kpu@gotJ
z754tH#UQ<tnEuq;=6VR#`y3e$!ry$txF6NyDQ+I@5KUSDrNCnaVj(yO2Oy?V?(lxR
z0-yp&SF{4mgqoVVfR&jUmGyp6-!A}KbmwR0NS*a!<wAva?_{>`6N0>2U$z+PRa{VS
z)vW8OGD}>?7XhYnfY<}{I=e77Gn1H3|F;S7)-|)6^!et9!}rBq84n9<wnT{nViiC+
ztNF6Mv$H6oxoEZT#y#*L<Hm8u$6`;t0-~F62~;PUxVX5%#NGXU@e%w`M<=JR0G+F-
zsBrr8$;IUc5VPwdOUS3po8(~=06T>Yu+mK}&Mw&n7>^=e<0Y2d*n%1LM>n#^_%T35
zM&8`-&vh3_0enV#gEvvV8!Ra)0lc$V>&<el`In3jyd`?&u|?FFsNz-WidPEtZUsc%
zsS!kT*5ts_)oxpb!<<714TK5O-O|{g<9bADiO*oXW*gOE;oDL{pCOV=DhCd6z;nP3
zbfX?vP%OOFBlCVI`ivU?<`H}}=|QUl{>=O{=~7Isf(}$m3B_X!pq-*5Q5#_|`ecP|
z`av7u0LYwTX|4J-XQ&cP6Nl*f*m?Oa<uA>C0CzzQ4^p|9LQd^$zanV=UoAa%OnOA5
zlx?JrX-rg57aGNY-s6JWwhYoV^>o?dce1<4fWL=C)N57_&`~4u*AolgR$Tb%W(Gt)
zy8kC@Z}(0&r`Wy>&2jVpx~cEGQBQY9<D|5*Fj<__r62CUo{7vJJS6YN_n;FB$np`C
znCM_cyk+1z1)Lz(Ni4^^w1M;QSg(SDCgV$2)Eh8qXp4gXk!DCFkCLCGr<+kYCVtKw
z*Ob6fjQ+pAL%O>|)`7n7wcneW{&87El+um@lHioBBjxy~hALVR&Bj0T85bn5X;3M<
zen~kV+l=4;-;i($TQo6Hi(2^;3f=<9M!K8z{u3WW<l-&i(~ubSxu^~X``^13&^Wr0
z5y6lp6};_39IKR*lK(3fiG#zXS{ivR-DZB2(tl&T4S8mEVbB{4qAKm23lp+f!l1=!
z!>0h6dmk8lrdH|xIB6Kg*k_23{O^s&su(-9gM*}8&8SL+M}In9y0iX!eU}f!Ll5T=
zs-q@%`QHFI=Rkr5+J8y(wd0>UJrq&1Jc9n6M&5Ce!#5w-B-9AL{S_vHf2UCS-RJp-
z1FP?QUi=(sp8LY)hY9%gAOB3Gp7WGoyrG83*zmmo>A!6;!}HgtE8NG;k_&t<WD<h(
zYW|yrSt268$mn&(8)mX{l>b^v{oFoOZzH4uS{7d?1c?m)*RqvJb|g-tCT|ulo09F{
zOVv|;Ite0Rym})*uXy<7c=6)G{}r}*RK&H+elil4CEy>m{s=0?qc++?Z^2<v5%^>E
zXX=Ali{`_>x_Xryq{|{1A%FJY=zo1oj{2x8N-d!}rfakty6?h_nGgD}VM35E0;gC}
zSv|Ng_T(Sh6&Lg?Z&8knD4~cFTNjHc0<2-HT_cnNWIT9>R8Zf(Wt$DjwyPYrDEueB
z+G#_o3&VQK{g*}WY!Sj#rU}WL?tdps$VrhhZm~XplI6Ug^}m5>_Mhnj>h*l?G0g@2
zlNQtp2Ds^vNVL=Cu5}vu?1YkX-!+f|`KK%TttG1#oC~ClJZsUF_=m+tW+3KpQXD6`
zi8c<^`)PBTES2h%Y3|Q39?RyXJeQ}?vt#g7ag8vzu-oSIkBKCyOL1!-skcG8MN}RU
z$n%0v37pVFh*UX1{0!6w(M{yz2?>8G?Pu`f$INW#q9@`{&_Mmuxs<npc9T$m4Aphg
zLao^jtg|Q5(_9MexjRf)c2eKmyMdA9T!J51qq*6K%Uf#;EFy4<=3@#NWg{!;SW$~~
zz-3G92K>MfG`|+=u@`Iq^~_*s*^k3b2=ZH#K+OYyYj^X9m$E5pPx+Ch)F-1Oamm6T
zZOul7kun#iPrV}I^x<DkMYlC4$_Ih23@j5}UpZ^CIW^MrtnIu+d7D;o|DnfaQ9-Gg
z9!&j+IL_?_14%b{h*c6vP$8#3BNfj2<F~nkF)%^3Q0K8^A8IYFq!yq>fG4ZAwsrmm
znA6AJFN-=`GU>mxvxi*%8&DV60!LIkg?M655Z5Kf#A-zuTfG{c)K^d0jqr?_$wNj>
zl1LW(JDz*LD*Hp*204fegW<RL+EG)>>qMUQ+u4`)U%S!&DQTqI#;6Oc^S#Qk)?|EY
zk~!rm-q8z4S<A6K102_umq&r0q3oS4$qFb~pG3-=?1!)f>cM{i!8G>p2)d{H|6%IA
z1F7ue{_&Ph$Pw9&kdYO#kD0wQLuAXIhpdb^Mn+bm5F#RE6tX*_WUr9OI!M_x?BDyi
zpXd4h?*Hzbab2J5GhXl4cps%WGUd@M|4dB(q3H1gBS$FeJ4HEgXnJK04n_^!+AM!|
zq&PE{sjru;{z58e@W!2%6m1vE^N69>*)K9<n9cIZL!wc`N~IkEOlq~WGVk90$Y{C!
z(qf77<B#=l!ZK#{@+?kF<#^U<_LOqx0prnp-*!EF3~vzGbi!iMz41mKJ>i=yD7z;=
z9zIe(b5Ul4u++LZz1m!0@Oo`luf7$I$)9{EPk5Nl)458V9zn4-y;mDg)TRh+?vNiU
zS~DWl!)&uHDq&$oJ(S3|6XQLPv`6d>JexCmDoEM1ZiLU}w25SAHBELFE&GvSDXAVi
zHgfTJn@BV2UAwc)7jro+8GVV*_L17J@fZ))5~F7!=~b9C2X*hdNSwSceUjngT;Qx<
zhrD(h$t`<^Z@E16oO{!<B234VBMiWRK}gvKZ<M7m8=ddpK8-XRsRD=X{*!i7n=t~#
z9whg%Ri)b<93Pg%PU@|}-is58OV)V6s{90bwcegdncokSU~xvAa6aav;`QwO88YIg
zUQqXyzh^IeYLg)yWY;9OB5&IG;#p_!qGly3^Mpi4?@sKuhU=mGNOaM26I@O4lL-Z4
z#3f@~xMZ$U;T-Al!Tes{*E*K@b!w_*^3(1k{{3{N-OOkaGoH2Jn6%|!Ba%L<69`+s
zeeKbR?Pef?kZA@3E}w?V{D)FVe=}5O1z#w?f+K9F!zjda>z%s%7w0b~54LUQOdKk*
zO4a726iBc(P|?2kLoP%`ClyiG`Mlth()^!GqS#u?Y0G2xqEHr5`ij^kY$p|2sK%19
z_R@qh6GM&klnxS;#9_Ssl93c%3h{`EJgMk$P++aa4iiRKvGTI}-=KPGkFl=SpVB(d
z-M?+12(yz<0P%9-Im_NRP172=`jwb&-59Jfn?`!mM;D~f>_ic@Ej8l<1Yvs58}H9h
z!%mrg82Ke2cZ_x+Q}g%|Zlkn%9vi<Y+~3sU+N<w(nCg6GG7u11Fxiwoit65MQg^uL
z_LTo$WOXujJcc^PD3h90Xo9yC`Bp6E8<H?ctcb35oGoC8!e2z{=%ejN?Do%cw(mF>
z-+J$Y{K$M;zjcyO+u|sBcD64mkAJCd{VCB{kS*1#_-gC;sRZ=kF3~z+x}FPNy;V-L
zk&aDibFBIu6tSf1WD4{>x`?Jgg82%n``wD6$usEHw5+9rXCXnG_#$OB3l({9mUc=|
z8~34~{In?;W3bA<w0k^W;pP=gNRpBdukBm#z=^?7Ng-(cht5AGu6-fdXy1&{<$P!%
z9=0HSMuxDyzARAGK7;d-Hi6p6;UV+g@7gDYo>>x(b2xPtI`L#WgXf=S>eXwlPje%}
zg4s=!%bCcvj^BCrRd@7g?6VkQoys2vhb+g}rh?4ADed4f&yu_((wF=arS*19jdsMh
za!{v55$Vij^dQ@RaKpDtJMxS5OR2}6e!|D6&J)XJeZf1Ij0YT$8bgoRXvy;9LfWDT
z@WVNt8qInV9^Z#9(h_0)O6H=id&gfygH*8-vP-sLBlR+gN^oe}NX9N*T6BCNS)t@{
zaJ&?f8C8sb#(k41uaG=_+ezw=kx2f3<G#~@BRllb_35}7(_8*D5f%Sc@{!UjK1)Bs
zvTLeTpk&^enmr#`lxc&W5)udvgZ1!J7=d#em(aZ;`S(WQmdgSY!XP|o=|*vvDs81O
zl5psVsSiDJA{oiph&<c$W;jx_N)L`bSTTkJ57i{yDw>~3M4t{wXq>akDx=q4lxncd
z`R^)m54@L<oSIRzMlZuIf0dFo`%6mzy_*TxmVo{%%Z1;wI{qxuBA%l3+6I$jS57Dp
z02^#RVx)1&RHz-^Tl&bfI_@~aHC{zL!r<-XUHyqTy==@KL%als_IQYOEA8L>o`Ah3
zBaT|HMC-xR&qDfJExaimUmxdgUwKDZB0p!R4_N>Q{7Q*n8|@ZAeq1rXJ{r<EN4OAp
zW^Jz7Dh~7PO4RGGA7CnyWsWD6D5Ak4H*r+wOqT?vM|w7wT`WuX-7S&Q_fbF<!qjbF
z-H$fOkeI!D=%=W^#CGVLK$&p+cv{gSWi@hF4JJ*qKGI`nDG&zKj3?R0_u~kM$L-Pw
za&O~JOnXbN>`eWB_%dN)?(*^1F(DQwUeXH<tzsNyw3<DWk~Q*-j#Owzkbe{j6N}%d
zqU29ZcHn&IsCn|HOh4>Z3dpo@?B$rogsVMC8}T#!y-PteEDb+xhwF(i6c7A29_oLj
zmK4Zj#W!2-9Wn)X%6&BNO#W|a;^Z4fh5kG}^`XFJffR8up^Wu8_e=_hBrLJ~%FFFb
z=d?f7-@A6ykj6Tke)JB8pB15r!>v)1Qo57ljVz;R;h5!Ad2*OVn?ZwAC%gCX2Hc%v
zsak`LV(n~`FV9iKy=*-s|F#O@Fyok?{^s*>9Xm&77MuU(-QAwXX{+(yLvMxs?Ksv-
zPoX=ckX590FD-=nj#dLhcXOGM1a{7f82xTAU9z`>G3(p4Gz7|y&R5ph=sgo*77tZk
z-U?$hZ|JkOJoM9UDwb0G1d|blBm%3i@_k=~iK+?pvF;4{s<+_c7h9+WHjP!X2$RTJ
z(&)jdRhnyx4!h=^@?95>*V<)deZ&CYFF-gptMh0~4gh~2>lLb(Y(Z~H2sP)EgINlu
zXt{d0D_akZt`<u|l~81iLKTeF2S%2sVn^W@l{%}0+0qko<L)Un6F^b*1ztjJUgrOo
zzvF22y0~PcH(Sa~2N4@Wvj~`p-dz*mHFQ_!=L>GrF)(PTo>d}P0sgK{g>K?MuipZ~
zP=by(#n_Nfn>nl1b3Fsyi#y-fC31Ns<C8(jg9i^Jyk^^gnyhu7e0AF$*uz~WL#wXI
z$w}3Cx@*AQxVX6Bb8py0N^r)zw>R9iKlHnI`@hP2Q!6GWCV^yStTJCAk_4)7UH3ci
zgYBc1Z#+1FMcf7F#m-DLGtkKunE6hi18BVvNmyl1PgNG}%EMme)*hDy2{=p-wJ+;0
z1RH0YMwXSWl(!s8Iu93tq|K8)T4KBg4v4n3#-=89Yh7L4o($2Im6fn45F<dl<HP9g
zRbKYUMt+U;n~d0BtNEW|9A7z7dP~@KOb~M(74h+<M&$7D@T;38Q4#$IljF4>Yj}JF
zitjX{MMBm5;~#3(^y@Q1h<3fTF($n`dFQiDiIN{a)DA&5hm=XuQr16w<1lo!Ac0x#
zd##5hKKGBU@^`WHw)M1!oky!L|M)~)QF)<?)3!pLVe=TsYX_#v1X<%Wv`{-0vbUHM
z8Wa?y27C=UsFHmlcH?iI?%x2wjP{pHv+o$LqHyqfE7jqIEX<NpYSl|QKpvVZQ=lS3
zW`SkRHSWkDyPwE}Bs+gTVu1^zi>M5@zN{6Ua+eV=^ySkofgw<)?1hb@f*?(~^yi-Z
z<>$9L@wLHEz~}GCYjZP@uz-b;pDEBx2VS@<#BLx@S=hdtYni~{Ess^s9Z_XS2VkNy
zs4iZ7gyZD^9*2XYvgJneQ|<jbyJfDu7T%!2hzsz^pt&*abQ#>c+t%nZg~c!#IfQYU
z>HPHciTTO;yNp71=RV-cs=}ts8`VZPW8>q&RHde=$-5pwO2c<T^)M$Tm#HL8iahih
zKYxPd>z98_aN`z%`>M#4tTg$o`n&>@@^CW+1qBQSLodrUhj)J8f5yDg-z&@m7)9hh
z1yWjC8tB2!G~lTlDiZIT*0>U9|9W{L(<y)P0OwQL81?Yoe&5jV8o4L)p_vOju7`@f
zf~~*EtHRIOy1MS~?|4hBj#vMBD(s*Zpr)o4NkYw@5`K>f{6xanu5B3TTZ>l8nVePj
z$kM8v)s8**(bZX)xjs^R(femO`~f2$T>!JUZEmw`e7qi18P9d#Q$gsgni>X1?ER7_
z{M;%^Bvf74PBJImb0-7ch5M)2<%FG>3>!7l<3q@U^4#RNb_o@aLT^rmE={Ze(F=6q
zsZYvxDS2&b`}Dd;cDT$AwHboP20qHS7Q?ohcKhC+<(-90k5tWYELOdRdbcrncl(;C
zDC~c4K)H#;0*!81ZXv_N11N_F`KMV)*zS+a+%~8D1fE5tm5m1ZnN+DzPSapE@)K@s
z%)~Kukn>Sp0I-w>`ufv!l32^24I--t)=%{ERj!)VoJSWuGNEV*y*3&9x5r=}6WZ=R
ze$;Rug)eE|-_-P!oKLX~J*Wm#4d@D|#valOM;-VE1qWw}xsm*SUJy9_KJ0RD463^i
zmtG9pokrU1Ki3<U!s6&B3-+3{+Y*%tYersM3)-HS*9B~CZ6%%vvI<Zi0rd=QtmEiw
zrtN<j;vTC&;4z+{{JaM0I!G5aU|Ps1U6$s==cd^b(OsI{Mt7Li?Hh^ATZJ0qb92vV
z8~oUwKeCu^iTZBi?Cd<itcqM%oSu#LA$#g|Nk73-m=@h3IKaqBTf_R!YUEMa`_0Yy
zE>KLP^n!y>)YLaib)ZIcrdId;wlugJa;bBkcB*T*${siYD=1z)buaIEL=F5_i<%ql
zmPb~VNU4<5_o;Cx%v+%&QNu+yUt~!^>Gs=4kH0op87(mn&;|fxD)z?nVS|ikX=-%|
z^86UV0cTB$h86;HFs+C+q%Ao$^cHzWZ!P=p3Bnme(}{aN%053;XyAe7>e7n4y-%wB
zC60_h(ZSR>5&HI;?^pf0Anz0N(?3;@p|}+?st|8f(cyhj(R<^YPt@LMw<ywn!Z(l9
zWzCHJe93hJ%}W2*9AIh!^BrP-J?Y&k3a+_fI4qK}6Eo8)a=Uh(NtqF}g!ff#)d(?L
z9Df^3=Y979(F507?)1Lx7bxjVfM{3|q(YCgvieY5h`z?;joA!zJ8t4HZh7~kR>C<#
ziF32GBV<+4VHIUra{Dk7F`xUv?D!PC|2le}lDXL7N^jcc9N}Gd(ZN1~Kf5pCOo%W^
ziKy-XMkTFlAxElnTZ*p-UmigPc*Fb{@lZT$d0LWoTj&7K@+gz3I!J_Aw(dE&J38c<
zK9czMm7Bcz+Bv^0sYf8OfNQx~T({{QOLpth6``Oy*HxfG7SEcS>`lqdpC^!KKLw4{
z9?kw>A0d16xJ!!1mx@f|7)9*N4ti(kPEOux&F@-3BIf=I<O$cfKcm$@Ax{`TA-%Je
zAB*N~Q@b*=3sLbMN(AKb>gR$M6TX9#K9N6@4VBua%0n}9iAP(LP&D2}@zO;x;9FTc
z^Di#jJ1^NJ9N$fm3lZnQ7rzqIxJTaqwtss?V2jhQs%=RPU>?0cbq<r1j!Nxf1KLZB
zyQ}vyhrY^<kv{RN3&Il(Rs>*DoAFEeP4N}~^=?B6ocvRfz?C(hfTP7`0#8M2qrANv
z)5Zyk)C+cwboW-XPe$E1QkoM8Ail{oOO4AdSP^yksNzPPx}@psr34gTwUQ2Ve^$iZ
zi+CY=z#JWsCKL^BH0DzR3o_gRIr7oLrrF7!>)}N2<#~BaKzMWG4!e(KWo7M<U;VMo
z2$1(@yo72cxd5gDuOU&HeI`r9m~n;(P>&9puC_0=cQ)W}R%%6}T(ge?(m56j?dN#a
zwb|I%@GLL4=coZUZRol4!&#M5$f3&5ph8K>P7J&no<-G0xwZ0(Pw62MQ%MW0BKlt$
zDMptKoM&Gi^BQaD4~>sYUcE|&?x2V)6jouT8F_8R8uBO$i{krTTWT^Uci1l^fA;t)
zp4h#4?B4z>G&A*GgdP&v01O)|;%LZJj{6tGa>#ZEE``?~8aRj_9zQWg-ju%8mzvG@
z1}?=Oxa^*>YrV>i5HO)b`FMJsLn2j>n;nLB@r~{Dl3anThEXWKGR$}BI}W_Ub?ET(
zEFJRM7{T8uX!Eg|Lbj&G*lMA}`b5-lDjZEtR#yEWT)>eEBI?f0&b9O8X^VpnNQ;NL
zbL=@|aqU8A0&yo<9T$<2>WzqGxSoWSp0^r*IK;DL_AOf%`FdR$>8t^+PY_6$r2Rd?
zc6DC12rP1-a!Py6o_Wcs!GgH;bm9dzPxI2k4iTopXG2Vi7qg_6k#T7w!@%eo;I1%^
zA@I^BVZ-rOA1CA9Ci@-=a-ZC%2%Xwh`UvynEaH|rJEez*hp_OQoY2EM`<{%57ywnz
zM4TuwNm3f!$;LGVy7#4E2X2sO=og9kVCWv|^(|$P{px*|gmNW<LEHnAk0O#J1$(P`
z6&Ue7rWg3~y;N*;4jZDH$7&YLjUflWjbRGpvchMM@5Liiw(GPWeM-U>=ufQfu=m<c
zG_=VhC8eZj&D_Ar+9CfOlVtv#bs$KF`>AK#F}tQKZr%#OQRkT?HM)Z-cpi>9K1DmY
zC=C>oMs2p`ejK0hLl>jUpdE~yT#HQ!SdiU7N9vH5#|SRoWqCH>?Mci{@B6U6QIqb&
z@r{G|<K3mVg4$YIp!b|!b{Q({0r@(SS?+LkQmXem^8r3rN4h7e*m`TKbBi!(=3B2x
z{;+G%9y@u4Q0T|9eed_$d$uk6*ZVoDkroS2eZ1Pm!Om{`i?L--j@;4y&X$*#SK~Xm
z-OspN**+2YzI#V^5}@eeBJid%31^;WC4EA@jZICZT3zmjT{PyV8fE^{$NtR1<P-sF
z#p_|}P__~WhJO6`5kM~3VPoRrNc}Hkk}{8~)Gn9XY6{ym+R)>zian<1?uoRzW?dGT
z$Fr=M5Q!-98W#08-Scxym6Vh$EG)c?`27%D)h!}fclR+dw^Q*GzJ1x#=pt?~cYx6s
zOmAc;iDTcMH~|ip-g@(%>>?#)5_TfFhZk1-cfGu2asn$e^MglH_*aIEc-br{*WmA2
z?SFpy^GMWLXOPNauef`#dpAMw6uQW%#Oej5O#wxrqB`-n-K$5J^5N}510O!v0#KX^
z{%a04T{1E<@FYw7t=@)xtWB14Y-c<F?KQlWI=i;|1Hpr*e5F^lW9Q!8Y-?+ikdWAi
za-yG$L#x!J(gx-TY-~()bPrHef(ZF!&8Ul1LUri=*|K$O4r@j4N~Mnjq%?GhBbP^l
z@i{rny_wSq3i|X_SE?%uBvjZ{fpG6fuvZcVM($U|zI!L3s`?4Mw=b3BXr|-AI;(&r
z$;AFInK#&u5fi=f{$bFn=Y_GqV0(h;eCm)v%vH_wIqXf0U<$(ME}FhJF|c{t@N--P
zW(u8O_kt69&4Wu*bICW^IXPisgBm^!?!0gn?rjn%)e=xtVhG#AgGXbp*u|tTKPr46
zE^~EenTYg0_I>lg-rq!e;X)8uBZ2`k68%}3Z1@Wrv(VG{%)hGT5?sljGqdurd9P>A
z!v@kxOl;*%P8aiJ&gf()RBORIoAZ+meuzKxjEo;n?gxV9?U_R_3!>VJdRd>mZVih~
z;YLVh0|B6G_iTkh)cHJ0q90tVeSL{<mWRxc7J&k`-y#NsxNSRdHjnKz|N3awdB%4c
z&CJaW6=>2bDk`pj!eJ2}U#r@OhAuW+t_FE^$X_*4ZcToC%ovi-av|<n!q%?~&V!RD
zPc~ZRe8+gUkx6pZ|I|B4+RKKBBd*9hFSc<VwUgFm<aM|o;Ft;+&lesl;QG1gUV-D!
z%lkqj8Fhre)KJ%*RMH`zdWaD;VlQ`8zxdJO<+To+{?XO;Q-@v+i{2AFlov1d^z>ZO
z&KQPj?}QC_wp0=?k!j(3FOuMMX-Y2xmsn3-{#P`*)`!nAD?0TJE#C=beSN(o%@uqu
z@=oDcaA}JRj!Xe>b&V{BDYV3Oy@#~q#8%DQw^`tiQNLJb@Uj;|0Z>yFf1B)2A|N=b
z#W}d!Z#O<+i}(7rfdkOR?^q2eAb;vI^$whW`Y5tgED^=0P&ZTlqjg*}s!NGpx;hrT
zEhq1TB{p)4b2|*{WUMnt#5}R=d!{p0oDsJA1S}*97=-Q5Z|_ZzQebqA>uD8pki#|9
z4%C}LC(tH_L*pAP0BVN$zuzMAoPhuY9M52<^=OZ!R?qhVpLB=3wYT70FOgW8%R}ZK
zvN06P-0_?O8|bMIui+ShF9_LN5r_YZ4Lc9^@~@z>0;2<QmVZeP^pmiWF%*j`WhhPB
zk&wN3TBDPE_tO^PTfh};0)F?_W-+(%?hMhckr6Rz>D#rID2y(tQS#2Z<h619P{X23
zF2(fq3mDzGa1+9$1D~O-EnE{LDCkF@A6)_9_CbNK+*&c6W?Rz=w*H$VY~=M7f6|r!
zj$1f?;sG<qS)|(7-*#Hd?x(_^a5z>@&M2w>zJ-;fQKxpcdO}2bfU+hSV0ZRqHa2iG
zk2DA0TUT-n2jo?BF;*3CiB^ww|G<JEHQaxGM`Y^#WEicH*J3Y|)f5zc678~3LX?G@
z|6x5`-%O508}zOLQGM});B-78;^Dj0|6DN{D`{A+DWYf}nyh%a19yr;{Xv3>L`3N~
z0d38}+7sa@ya2-cgTfE39P_Q+hCvoj-VDOJjloh-P}tksgDWF4+X_M@aE|oOXL9-@
zWerMS8h+N#O1etec8Xt*_;`K}lCdm3)%*!R%T5`Sgl%2OWtUB1>N}bu>Xbh^gU_`;
z0-w(u@Ex#2W{SGhfXM<(EZ#?l=d!Vd-+iR*_RDMr(~_`62`)<IIZCwi{tG>{B`m)_
zQ(esxXZ0uW!AqETC7~{J*U|qt$<Rgeg!&(lltH0Z-Q8a_iAKg3#Sq?iHlwaMMaAD^
z2MkrcC;gWrnLV(_U!^mXsQ4Vh?5*Rq(8K-t$c`^xz%u0u3c+pPL6_6YD6T-=b7+$r
zGtz;`jLteg;POE~H@Yo+0sX`>CM{EqUA@WmGh4|yGBN-W`qj=~z_lT2`!OCIRNgCb
zu<v~0xlDVcaQ<DCG7T|NozV9u*O<}oVBqH4iP1%O$m%o@iehP2n?UM_=GhlkS}Rd9
z_DYoik?^Lvnj?`X*W(=W9~{y6TnjHj=id@G?|&z9UPl!<UtM@czZHj-etaKP=+A}~
z9oXFD$bZ2q8{|ZQ+NY+=wnSEfj%+N<l-1NafCd6wWDln`geaPV_ly<^Gfm2f;>`}Z
z&4)5%v^>|p;Y;An7C^d_<YM!%nl(ct7+vMcZ}us4Df8ZnCyX}WfQR%4Jslkw#uWee
z{S8~pHXqE9jrVo?d-YoVgcpPmV2y$rSyO)t7SpX;x8P7Gtq;w4R)Eb*Sku^6JAnrK
zWIwIQC#$bqoT_b>C2H#G1iS0n_olG>Op-oh;7LnmLHv%uE>m6z{wou%+Ybz{F1D(w
zir}qPcV?ickBo?5L5P-R>iLXY1xtC|_&gGzjO5|nIX~HOb!Rz5)|6=a{m_s%&|$y`
zmYX;H0jLAaUfmm30m5A;J60w?*+`oF%ffKp2Od!x7OXC{SHmF00|jgUTJ4XwV87rn
zUqPY&S$ar{GuDT(-^WKa;h)|7_wS#G+c?#D7MzP<*cUl-@^vtr2rDMagT>};@!Gz(
zBOL|*6sD@$cdH^9Qcd9Qc{z+NzGU2>4NNf{<UfH)y1DSdg*~t>qh}%iX3`E8Zkf*H
z9hj10QB=|hgRF!Ln%V9PX9lcx&GETsIVfZDH2H3<d}3EaUJAM)Y!{fe;6?b530VV5
z8KKqN)A=|3{r!CeZUA#4k0i;)I=R=hrO@uXAWq_j8R3c2BCA}8FOl=Nzqo~GNoMB{
z=PPBZb9}<bc2m#@$IXhUwxtGH<Q!tuUH0rr>n1s{@zSEUb#m#2ZkDxk7vLDcc0+*t
zCf@?%l1!l-m%&>toEEp~mJ(ZzdQ!3YR7<Xh4qfDd+W!`JY(S{5VaBX0(mhrwXVOtn
z{S&7uT}@JKXqqytTa;f?a3A)dN)o;I<~H4`qi*3(tmwm0O2j$XulfPzQW>n?;)Ooy
zn;p2ucN9OAX-6#N^j@D-0hbooCh^P5w_U5Dx`33<c{G<5LARy2S7X^kY16s`dTZ)B
z+VLpEz9VRri+jcJ8lHt~=AS90_~3-=jtljBc}>l(j@h?fb1L}U3iK032Ce0t!Ok6y
zuRdp?qz7>m!R7nx`=1xXWODdi5z%X{+l!6wDm@>tvu2E<_-5FFkEuX&C6KY98mnE;
zMi5Q;gXw%46_nmf7i?nkZA>5a_@*;f6<IhXp-4L-TUy+t_tusgQfgN;3*pdWLlAZ2
z8Ve^8kN%i#N-0I97sIhn91o^kZGeb-_(wWFL-_EyD%HCbJxTN-O?+Pmc6;TctZ5hf
zHb<$L1TCr1McO4+7^Susidw1)b`L0)is@Z)uxzRcB0@q8l0HB3l;b>)<jd4Qc4X!h
zV)JOH(&wGtWdnPEmnDWykN2aHjrJshP+*rB;RFGx##l3EI>fODI)5}>U}+l!X78tO
zG06db`mradQOQjve7tjhFZ)nhl6B7_;!?Mw1L!vhVr4j0HHPona@3WS?UDHEZX5v0
z)2Ffb@1KCvM*B{@6cd&Cg26`)w(Y0bs^#vE0yedG7gQSuCXT5X)0)^0GJ_v95oyV=
zXn==9GD(`4nL*{`#6$dbv(w0Dd^#>wNS0*O0q7mS=Roi5N4HQwDh$<KzU~%E@$Hi9
zZxVEoXu!RD04%?M`Et`mUKEW+OPY#yne#l3ij0g=x*R+`{rV+*QZ4_T&=a|RsQ7Q9
z>HMWDOW`QK=j@c~j*gBYnd%xEu`x07NK6~g_Z15T*LzjJZDIWH{v76>Z|nyfFC85n
zWTNCB?&QIIv%4a5u=}gLw5Kmav;=lRs2L9q_tw5Xtbst$AndLH=~|g(yk;Y&(PJB^
zo}^0tPxb17UGiQ&T$6zzpFM9GMX3mwbe&`u6zuQnDv@@0=euIs5^?~P|G*muo%v0u
z4MZT{1%(`x8h#jz&A~>W+|ij(*TEN`@1Hz9I5?Pum8^|TM~qTzS=f0qql@%p0=FFl
zhz)Nbn}dq$*<L{k0f<R`0|WXfdI&ba163nfdI9X7y8yeOh0-s>6R4mIs={{8V-JKo
zH*P|>^sCHx%5u$?<m~(?fR@iH5z|P}HttAZb8@EUyVCuZ$DVh~Taig5Jbd_Y|HCC6
z7@p6ao#As3pOBxsxQ?KXLNe6{{&vRa_Lj|l))Th*ZdN)8mM~h+b#U<VgfU1)fdt^d
z_!=ya(b)d}ez>YrDD@q3xqZ*|Y59}h>wWhCeH1#>iQOE~K?>=s5A$54<ic3UP|!hM
zp}3);Atpv0sOwjT#itPKF7&;pI}8?bYis>r)o@Tso|~$)Rb^p8#b1&Mgr)SN|I3$Z
zaC~;iKQn7rj62=dm}p`S{=oZKIG9gat`6i!Cnl0}FN1-zsYwPMviGMfIhkVr4QN2l
z5EI<Pf!R^OqVZkDP>t)?O^6zWkSMEc@)htHTrPJ~6t=tc@v%$gxdXga+7bK!uZ@`Y
zW8wHh0TbhsXU;R$M-T!kt24BgK_@?{s~2?S8P7~ZR0j8Z%o4JR#7rs@KmLb~w}x!V
z``QZaS`e$qb?pyUq1^d7z5|71Ugk(zTdR&)7E~14rx8d(zFA`iU}b~Pl799OB(hp{
z{#lB-juE()P35;9Q`x~{6pW0dvdHgbw_U!~T;0P#UDxyET7Jb49Df&0-<VTpNO;-Q
zXHYRQJ$oI%gzUb2VZZfSZPrT-5Cycfv=$%R*w}R5i#n~jWa`X6V)|~n>Q-d^U|24~
zZ>E)Konbu{?4#J6Q)Qj+abTtn-dm|%F2BgBWdi0{)cvWjzm@pNmR4j}k>lb14kQbI
zykh@h`)VHf!7@Vky@2qDW2uD49UZ>?*@3xz5_f0Apdjho=jWOEN?DEH-Zy|^f9I10
zgg}XJPyjg{VrfrT2<r<VC3X&Y_gki>DVqe&|CJ%}ccSQ1w6?#FHa;;cOTV1nd*_(b
z1#oO%%JrvpsBQ|CW=(c<&S0v4bM1aylkvjB0<4b{I%lkUvSk>@A59A(k;sDiTv#PQ
zlzg>>3fN)u$8&JY<Dawsfae6$rrk^LnOWrI=dK8BZQ{zO!(87p&}xbliR@j&>K^_`
zW)KsDTnGfgx3;#hbMOg>iKlnJfO2PgXw9Gu)rLvEykFnIfL;=qV>{_W?puL1D#$mW
zuP_>=<^6_@_43R=hLcK-t#}BcyX!;XlZ2Wlo<Z~{Fe!EyD8o<t%;(-SF-0~9@6nvl
z7<}^_H&J1&ILQA@*a4|2PD&aFBg2Zg>rTC=Gs&*9>g~yk(RK0q+|ls^D1M&#`&Xx@
zON{Sib?p4xuyyp5^<SR`tpU=_D<fr$0V+pO%|kf`n<dT+{L$9m@*oQr?W~h6{k7_j
zCR|MHxwXcH>6B>josMP>ZsusONf4BTD~~QfK~nyVbOdwa<fpKYtnVzK(1Ac7BZQ3g
zaR8X5iyjCAZE;<gk|C=<G6K8zcj@7&BbauVeRj39wOe+7@-BSHzqu|U!MIJ+)Yu4D
zad#81PS`YTJ~(;U;C24KR)Gl*+Myey15$HZv9}J_Kadlg^_>~fCyEGq9pFR2)Ge-G
zh33z%^BTWC&meN{un42iwcidiUY>NMt&Kf&(U1<6+JVm*2RBOzg|pgvzc{kW?5=8=
zkTtvMCdF061&FAeJbm^=3km6|X$@tMrZeoKdV?@bU&$5L$XQ4eisOxGY16|_Bf{lQ
zEyYS)g%LK|=yR&Szu(0a2gUFsFulVJfT)F(+qj5hc<+;Ir^oYP8lT`perG<!tSfMt
zD;9oriZNsgNZ9>|Q~Gt4UG{~AG;92Wbjv&UI?IUwa@kY2JPT1=-k(hS+9B`u6Sm$K
z(3@b7I)DB=58T>;sILSOx`>~wK}FbZWViUr$PzW;&Qqu4xKKC}KSdBf#%56yu54gR
zHg+svtR5Vz&x$Hao$yv@YPJ+cvjzG7{@eSU+A?pxZ^vTk{rHonFVy-yx4LW*@=WGl
zjk~<_ScNt0dYn4hT7_(^1aRVe#UQ3wm5&jay$t*AC&{5F-zFtpfVvi+Tfe<#E6KGx
zM;%k&TWpMqpOmMa>DzARv_hUea|Y&b=ll1$mMtNnq5UBLlD2^N+!w;@IYFE1FjDG2
zMKQWXsM0%+7;Em|K<bZbPyp5PrARM-Uvm%U=!r&hFoT3~*UD%jlyzE?i+{b{{Qwg`
z9Tc?!wx_6MsBf5<$PSBAT7qX<BVjN6)ZSiDP;hs7jN*C>ii!tO5I@m>8|5k)OV`LJ
zMT#y`L`tI30Q?D`JDE^O>M~kM9XC9}G|YzN%$d|_RoGmTQE=W_gelyRd*w+>BOH!<
ze;0dRL6f+m&C3xT@?l<rXjNobtKa7{HtRo4`A(rto=i+mIs_BCK)_~>)>H@87;wQp
zL_NTQH1OgZz>V7ickNIABu3DCxev!yvh(4a@A11L`Qbc)9M4V{Sp45$<~1blzrOsB
zlQ1ObwfkIz@9|}-8?@#@CC>1;px{m>`77P$3XeFo$GM6@bcKZQCz$mFkDBxA#thMC
zVsuGGy?38Ob#zNs9LScV!{W%u>bs2BA-Th2RQh9P=J{>^m2puZBMe?{8kxfYTcK$W
zm64dQJlX8?Ui|^{3d$b@&Y_E#ph`}dBoE)4WkMu20^Fa>;`j;uTnP4R8_J7=GQGsy
zv(#Mw6riwhg2HjDPflPb7q~D+1_oF!=@F5ZAv<^AiHx_96AGaArYF+K@iGf-SJ}Nt
z0n7!C6G&gbLVBU-oUY+}_QBJHs-M5qP0}zUjn5z1DOXzd)_t;Gpk==bwgY%t(S$Hp
ziktUcE_|+UX)6a`A~t$-fP<}SAHZ`pca$}7AX9*mVTL~eI{XzuT^bI;1W#MNilZhz
z7em8HM6*FM3@mNdvuEgk?Z=hPpYpA!#2U8=O$Xut^O=15tz@icT;YvprUe?**dTDb
z7;G^^aDcdWxWyuE9nJMfrTpL-@&b{}Fyi>Z0f*Hg-!wlxvlxK7-LxYSOyfC30b{5A
zRHy*N1E4^FzGN1*kFTrSfplaE072-ufwu}s<$WYB0ou5`a)}&HaUIo-%p+Y>B^cx2
zlP{q$N&z|mnwq^le?5{$_I?XY;V<~*3{j78x~O8k{Eq+#X8gf60hHI}d0;c<?Hf{6
zQ9k~_ftyOH*)eo@L#=r6s#uXYgQwj67lnm2&R>24pVJ`^0G?#3k>Z7uy$>}GP&*W~
zy0>p<g@xfDhdu4T(EW64vG2+}r15kwF)+lU_}-kY0}AhI;+>3lR@UCS8QG{hN4i_)
z8YveFuAqYct<Q8Mv*CV2{UrkVl=)jt%C0vaqNw6;?%$P5#M=7ZzJk)KnhJv}SGr6g
z@vNmq4pk0&0c<ZXq*+-{cAh?UDhYeXz@Ly1AP-lev5^3oKMlJi&ZN%uK<=Gk5=HrS
za6MFNvyNBY0SRm!mF3Bsh{gVyc)4V~@C&L%JxHmQK`xDOn1MDX3t2dbCqPACVp!aY
zYxqQqeEk=UUy9}GB0jtkJ-8=jKI#ukRWE^KyBA6YrV^HhH^09xmOGx|mq$9Q!gRJ~
zCWSpJ%`VocNOuN^Csd<c<8*F(a;>x?Liw4(9fo@9QwvtT-&pE@?{8kya!)`C&=GXs
z!ka<AQfS2`nR7|IKCS*Ulj||5-TFZv6|n!W67tPaCcrQ@UQ&Na1}=(4N4`G2_b4sd
z>bM1BiAz+@ZM+H)L_30I)UvnQ0G%-w?mo1&0d!7DkRHMpt!Q6LBWom3H%2KZL?AIT
z%Z>OH&@VrrOfLFlMZgd#gwE@@<1_ViV*Nt!9Lt4%*mGfGfJz47fi_i*Jq052DKFd!
zW$M`M2MPVj{21WQ=F2<e%eXl>?f~BmSx6=k$9{szW&`N*I22zdr8oX@8L%C>c4V(f
zuNorlWj*t;3ifY^P#Ywv(b3V*pFhW9PotpEM3(1;?N7|1TF^5F&E=0AsqBek{J8Md
zM>BBfIV^tyg3`_q3FCBxDGS^iQSLP%p;y2p133={^8RKA*dWEz*%_Ls!AdZZO~iiA
zZlVcTgWFAkqEa4yZ4W5;5Ricuw0aMPWCi-%L;DmvcSrt&+mhk_VW!!MK&0P<4f`a2
zNJxmQx~HcnKtw2aI&l=HtKRFpS29>_5A_QN%vjZsv9{lb$`)zK9^ENiVnhI-fGFPk
zeN>jX$9Hh95Ew|SdXpOps@ja#8uD^Qt^yC|lpL^4NX~JSlaPdOeoABplNra9e+DbT
zgjhIJ4v0?6tB{wM^Q$YhXB)3pO-&7KT9<V=OpvEIP2W0J+Oz}tUI9#;dcNwLp_C_O
z!VlQgY(6AoGmM;_Up;?5v$go@ThPlFFDia@vmi8If6WGzmDs>wNp9eevU`zI^koiq
zbAf~3N!ZKDMWBN~ToppIhX8g*OK;8n{5cLWRiN;G`dYdkv;167F%@%Ybr2LqZY~E4
z0y)h9&o+hQ)f-V5nQu@Qsr>pDx|M;wOdQ-zR=!us&~mqRF89Z08DODz)h{r^H{~WJ
zCBYfU<FU61&XZTmx``;N7dQRe^}$%fu1*LQoWmXf=IMvB2+BA)pC7uUq@<2P<6~p-
zjN-`g9Ee1F8*>Q>QKC&sT`gf70T1U@5tfKgxCmPx#?`kaS0b4<VrkyWK|2-lAMh!C
zV9N4oCO&52W@sS#fq%t55bNH~8yEzqH4*q?I&vkRW@f_E{^G|Em7!{`*sjp2xm*zr
z*4Ax@s)`yB?i)F12nrl4JV?=&0JTSb%g-UpVF^T@d{|gmNV|Hi%dPcrtj7@j2XZl2
zNRW(Z?3r7kv>JS4B|=&UZiXlNE0apzC#35=zWW&Kt3BXPr@D@G#_=Q|43NE>sf9e0
z9|t|8AOQo0(618bE%MzRoj#e!dnH-hi0R~?mOqKgc=u}lM5|zT+Ep{>k<zE3TK3>F
z13|*9J>hyMu0VG1Z3!qQBI3Z?(wtE#LOlaBSB?85$%+o%>NCgxNLu{XrW*W|ln8Hy
zOGA+_2mu6S_33KVb!gRdA^<S9x0hGI_K)PvCt6BMXE3^BMwPo1Zkis*q&pLOuI=|`
zz#T`r3$z3@!?}^*=H~YG&*JK2Lo$|QcwTTkX>ZtqZ@?%!KS|;V$hZ$J`Ye&~uxo%o
z6LQCQg1^!g5-MB>@-v@>m<P-JAL*<>(aD+y^y*}TszC}+N`e3$1T{e`hu!~u14w<2
zcrxeryuk(I`>PS)&ER6geD<9EJ=;;c+l_<Mn-F@&{wHBv+YuvUt0}k3yWVcpn8rw0
z2R=R`J_ibliZgT}t(9?caksA}9s<VzV2~PYnr~l}oNCti@;zP?=P5Y%onv+r)UG9U
zSZh#_7n#-aZ>SQ^%OtlOmO%oxIaF=T773_fh1;jT@9Zs>O;*CO3z+Ba+eT$3oX}Sc
z-fCd{wSWCVQ;AY$t%4w6)*W$eiof-Fi>UaRn5R;H?x1}B0k7!qpEtW{zwjde_!Xef
z*5(to*Kvsn(L_J{ZYg9q;n)L;2FMwoAN@11|FK6xy>t?(;6VXXzF9)GY;?8Si22t!
zEOz$6C?Hmj58$PIZC;Ufx&u%>Y37HxoxJec*dz1YWOkP1<Th0p=~^vSn5qQd>CF-v
z2~%J)f(dp>_&@Mbe8%m45Jd)l?B`%{{ucY@?Kl(FIfOwk(+g6UZ)?Lj4~(Crq@<Ai
z@8^mkQ?U?~atajc$+b9O1acdGVEO!c!)D{==4Oy1VQ_zcGM9WIx7cjhRjY6UW!F}x
z$NPygv@c+b3f*z*cp3d(`SG)j_-nvx`A_OJM?lAVJ;l=71ZXol*C$fGZQ`suA@OA1
zFAjO=n!&mFJ#W>Nl;v_V7aw2g1CoAzLb#SNHRT6arXTOkoFu)F<lb(!N?#3X5;X<I
z%<QbfeazfPvUCM7pPDWRI7MmEUWxhR5qUy_F4PfiF5NjPUw%0|78<g7SXsTlVIb)>
zs{{=*Qh9yWzE2JhQvuUk(vi}y?O$R#^sjGKp0i60j&TL(ZZz#@R=ePKclYy^9I8K#
zm(V6}o9mQ5ZT&?oS<MPhp}ExKV@xF2`-ZfmH4H8$WBCw3XAM9fieo87h23FvHkTvw
zDA#O*6C5+5&r?%Vb8}g$0{?WS3t&koFcs&BUII4MNDG*QJis7Ea3z)u^vJfrrdWSM
z{wt(Ot6<|r6C2a*Sl}1k+ZhjIX$g~N(-;hAdC~g5kEotakc4<AS?7ssK@}nwe4pUW
zLL-9VXuk(3Ub`Fyd}CldYWVF7{o<CEmXeZRupXc(NrxAScLcUhXy?kG?dN|Us!g){
zuV-Mt3Ir{UY`J_C)+cuO5050=yiUiuVVBEzH(Dv~;oIkQKR_$YqKkrNYxls2^f-qY
zC?IdpOS8x-P&3mz1odhMDpOU687x@keX*aN183Iu)miF~&<>|Bnpp%g_-Dx&qz&~U
zHx377O{{lplF%PcCQE{G=k*)BH`pl9VGUt_2|kz;&2yg;Gcz;G&1zo*6w#G%9tO`j
z#qJJ6Ptw=v@;7BC|8>bj1S<0)e*!pWD3LT*;fyS6RY(q!J>Z|WiXEbuWuy1Yg3@OY
zJ_+PMhwdk=$!dm@G1nE>rJ#cn-YP7%+m6Qsf2)+BZAJCkH$ezy3F0UWt$Vay`{C6W
zufvYg6L)c&BpG|ECU#PX8=>9AjToA6g@EDcGU1~K3p6u9nzLddqVoMUvI0G|RFO1A
zJtv245OqtNIGRv29XP_6)&u;R&>f^O6WpkO9Te~=YAx0F4mlubbwFx*!5s3>8U{16
zsljQe5Qte6+T=Rh$=5S{OjLLMOS%s913GaN{Z{o>!&Q3!xdfGv;WU8;$kkw#e2v1t
zHzA<*6tc-=#lWH2T?!7Xs;w=DFg`}GMLN&X)L2H!tHv<BHvA>(%uAzmKZ0#BypqJX
zz^3a$r#z@7py>am!RJO14Bg<4q5AXs(W6JtpR>aIIVbWKB#*)jYiq;PQCc_RO=UBB
zkw~)t@})|)6HZp>qXdkg)L0_Fh?fs6Rz+tZd<IZ$cfV}%iW*z=WZYm3$|{<>`y8hE
zi-z>2)^Rh3A8u^~gXGq)PrTbsaSb-ILVIUe3GvI^1ql;?KJ}4B(D$taw=cP!1(M53
z>0a?v5Z-Egl?XP$Z#M`IHxQ{3<5IEB8>!gYF+t^gFDQByPnC#q!(XW@8C95rS%-GY
zrkN01HDiYr6~6d9>v>BBLU@-d<>p)O7nL0Bk1@I#8ET`qO~p?FAU=mU{fN%jHkAsw
z9_@$HYO=eLvuEZuviKkx*42VBTf5}L%2<>$r+B8lkpL599;tCK{Y=4TcP+E<@ey^k
zwZ;H-c3H+kwGU6@DdQUkU6y={yn%Cu{$4X08vkN%3uXPg`^Vc7E*-)NP)?^}FBhot
z6SR`9jyw8ILbhN&B1Etimd{uQ1t^#2oU0@#6H3Uc&c_2i2wu|=s{MSUgaWe-`+{GO
zAsnsQ4g!GV@aDC1wBH-So4gI@&(%<Z2&265H5E&C1KwaOb_3uG=SLQVUZmHBHv9wj
zn1T7wf~Y3`*~zJ1)wdETWclgrWal04`OrJcEwt?sUU(age;SDo+b9H&KdOAlaTVv_
zAg7^}IJrS<o*=vV(BkPeFv~zu=GLO=5HHJD>IN{V(?$Zr`>&Z9;yG~D?3<dosO3K!
z5vBi5Yl3gKJO^4>)z42e$kh*j5Ew<jwdFg09aYrLiP%CEYI;{MLLD5Cg=s_q(A-H9
zPbioq6kq<4Xj#+^NBr+=uUoNF*bhHU+*EjU!yTe&fp|zqEhS;yXWkZ%bXf2$=x4M2
z?`y{z^T3g@$aWkA08?=zm~Y{wwjP7>0|m1BHzpxF0<YydUoyt5gqB#F`?6<XUiw2a
zLL3}EdiAqbHVKVYHX*;_sG<yNp*IaY36c*yV%#-<Eav`FAcj1$;Za!BTu0{lwBsgq
zii7_&@y(2vU5AOCjBTxm<se_m9qN6_EED*-aZN+~mya~;&`YscOUZ%r^d%R1uZF>4
z2rrJbrWTgrM(CUgrE&l|;v$mst2;Dy&;!?lER+^dmqI&;9vbK{B1z~w6cbGP$_)%d
zBO^6G#$Ib913v*WW)Pe8Sq%nX2;jfF8?(wI5bikAY)y=dJ1r+FEPU_y7&SJ!){Dig
zFMmsf#LNVoDffcNFiC#(0bBL}R}KN{0ZT3FHV(m#)Ks%ux2j#gV$u}uE9-|qyAa?b
zfYx1thD^*5O;DAPn7br;Lt=Vbo3ImT5)2U!ndP$oO~zcjqGk;po!%|Jbs9<&u(Sa0
zpqw12o;ONu)^AF#`iRcvvqxmr$^(E62FkPXVuVyogskq?#>O=<F`5{q2rQdQyl(ON
zBM-o+GSZuz9oMU&T?J3A@!fWuR!n0WSx-_}0wfF39exleIWR3)hmNWT(2%ZUqyFx%
z*f9)<nR+Gv7AWVZv5ATgB_vZl1xH-PFiE3<&=|8o<1vflMdY8KU%FabXF-H1-B64K
z=T-;Ry$euwxI?Q%$;I58=X#QsJa4r2KphKd)L1Ok^_A7t!H4^3GYPBr_dys<nu8W7
zBgF;?bkW6<f3v9h>8->^g9V?xV$DrCSrI$Df}7wvyJ}K#e?M?{d&&AkRL@i4#Y~T8
zr;$>6r(Cd10b<WMIA=&`F+?I|4NfZ{Mk|$AJ^N6i+^nqfJEOAcn3y~6<Q+XTFF>Jt
z-{|dM(AK6DVqOt@GgTMmnzOy~?Jt}bS)jl`h6fs!DUg5|916k;X8-Wu9tfwcNpF2u
z#*YrxLN!yj!RJvuVV$f7P~-)<kzND?uh>j*`u;|by7pxhVPiL-Ep>usG76b>=b7#3
zdLK^WI8JUh@2yG+alOg|uPi~7^z;Oy9@HgO^WJhr72aZyHGodHM?5|tZB}NS>Fu~A
zr$faUBG?6ag$|Kx;zYiu^fbG9Ij&sU`W61TEYEP(bx4GTCHFYn6u=njvph!de!M`v
zvv+hnx3Ydpt{TojkRlFBUO-<}xZnroFzW+ts3VJ`W+}8VBcD9*3o3jl1u;IX?dt7{
zLZYktKu-z<C1Ft_qM}4oVeXn?YvW(PLYYg3Jazgs&}JWHbqJP6T$hg&e|>kfeMfu~
z_q^-%4_<mG+N+sR+Fq}EM<7GNc9EhWAW#modFEZwpE8Bc4)v0T@JNZ!3(?suld<Ab
zq@R=nUII@9qk1#IUg^6y{{cBKXoHM$A=2Eg&|?Vt1@AP!r$0D2xZ5EgqxJnCm0NaQ
zq;OW#2g2f1XDdsYl7Dk?mtb-7Y$OPs*s}vQQ&o4shYZfQDV`MXADoSMS4bZ05B$19
zfz&XVO~&pI0)q*OrQB+#ybJWv^k9=j5loN`Yt1fTmjS)6*yGL}<+IX+expCgX}s*O
zdamc$to$u}$)Ue$L)ee`if`Y(4b6sjV7|bOG6Owp0UawJzW@9g+(OX)aytv{oqMpe
z{8bp@SY?+*p7k2I&+A^-uqgfN3h`$jE=-bTY+Ri8kxAPy_!hxCe0O~sM@q|IL>GGf
z`iWG_=AftAs4vurXv=h(cO%S%^=Sb8lVC}m=*q<vX9<vKfa~bZ8!6}y>{JW|q6f83
z3Kycqhvz=lYfp<9;XMW71@xq}idVB35bEP3>uHDLgMW&!P3yWq_1Md^`cW78$D8{;
z7TP^PbT4NQ4wdnl>6$A(y4%sblnKYLg)RL=$vc#aZLh#6=>xQS4WS#vVQn&&YsIqy
zYoqY1SX}7;_sWV;ZM4LO?@4kS@UFv=vRx0`vQrp#`Gm5YnHSAy1H>c<S8&eA>mPc4
zddh;xQ}A~c5oCG_;i!Qgr#I?AyFJ`U!5kU=*CFMkNxM<|n<tN(i6Wl8Wqb+W{(1xq
zB!y_Y$60y=JhZK2^)8f`kiWp^o+!1c6L1UhxTpXA&6D_h>Br?5v_%%_!38J2iS4lX
z4{3qZ$D}<AjNqykRPS+oQ<<=RKKcgtr0rdiX@IQ!i0!-6_>TI&O9eaC-c=`xx_~@~
z@9?^dz;kjv#H`;E0tdDY$BoXJPycgDovMhn9(7EpcablVSZEB<fTGEh7;XxZq3ea(
zLXMdm^Z)OjYWY*K+1H3Xbx}J}kXfAbIJgGI&L8v>Afu%TWVFh)pt}FNf`RrJ%(|2Q
zNt|j5ouUNY8>izglgR8~Ufa>S#~Ww=Z&U|n&(BZq#h=Pc1k*Wu?xrzVU%XNKpC69T
z6rP|@bSL`nK)-);xTXNB$4wOYMhh;~TGG&BMB5(PI^goCj~!~l3WO~7*$??<gv9h^
z#Ev`l1+RF%AFYRxE1NWRK4;`vA7ws-XB&=X++WK5aXsBN<bN<ghhK@>#UX8fmvJhF
zKl!7FHz2d74FQ$mUU|pRYA>*?`St}Y#2=4t(wY%Pvbuw-sFeASO{QyR4i}Ib1S*cJ
z+IwAvU9@%AcQ%45`zgusRpT3Op5eCV_mi~bLd^Z)gKDsu9Ktfgy|^jqA8+LiJh<1i
zh>oGf<|I#}4L4%%HjFZ{sxkDIA&_HSzcGT`suu0B_Rd6{*N(p*z~h|~Ptunt6sB4i
z$!Y4wilUNf29R2uj@`pLMroC6Qp#rh-`8gaDCYNjT#9^c_c+L&d@^=)kMJNYkiJdV
zh3#&f={lZ;FunJTy!f7AJkuPWqqOeI!T>bg1;?H~8-kgmf9Ao@ky(*z0^0v~sWk|F
zB>!*jRIJnJGBMOc%W$cpn54-qmvM7$o&WbbBe}dV$F4aA+Bco=<{$~!<Vgh|N(&jw
z0H$_Rm1Tyt|94MMesNs&{WheP!x(}6{L&Zpf}gXwlP;1p1K|BzA2IyzN~g+fHr+bq
zzZpkjbgBHIPs;#?unhT;A!{1Kb~T93e=a=#`<$-Lo}Hg={~L)Byn-i$@cJ<@wXULq
zGJEMl1f^{?y?kgA|KGKaiS4ZXP@vuEBO_-en0lb!iasgAcz5vZSO=rNMdtqv>Audj
zd%P@_x=4*#Hdx$dg#cLR7jsjS#pSfg)-X^0Yd!@wT1V<-$s>;;()nCIwn&)TyU<wU
z)jrCHIDhD@Ts3XnoqaLrnRSe0N{T#GFOFXHCMo-%))s*UDHhs{boSx&VqXsO1{DN}
zQv)SXS6_e0xCD45OCKU=Lk@4P=zis>j^;|EzDaftf^&~k!;u>s8yq6GuL=HJ&=R@x
z!a>cH#a2n9;;9cD?&7%5V|4vd(7W$9L;__s%@~C2Ix4I`5WH7a+?opFGiA=^T55?9
zsqoZIOV;1ql9G_P{yvLC=}WP};xVBn<QQ#3oWFLP{S=+#btFv(&hN-$P5OPFWV6`o
zB09!!YL^ETwyH$bZQ&0-rdXBg*~XA*=f#Nnw$YH%X$n#(w_P$cTqdICeQ=6ogf979
z_&E}j>pw$0Bz^_`Z8^8hzuKCyB`YoM=dji!b?<N1&NE@ZRW1K_BhAP`odp(eT{q>H
zw3u%q0c-Sc9A5-+e}AdhnG*`!J>aKZ7P8GZeX_R5SAMSfGD`u{Rh=k1q!8jMb{s$b
z?97g5WS{TL#U_W+t}{FOjftz*;4jYJ>$Bf}^OH#rS-f=VuX&2?mM44*JyBaHV}TfA
zb7tCLl)eoo{#W4%{HYo(v6&SV(f#iBh`9CfCwI%cVb<+Zv#{og3m<U*v`O)r8{c)U
zZ_{`uq5sMCcx04@zYL1dnjYeVk4-{%p!j@{tHkMvCO_6PlmBM-`Mv8%pKZdVz+Vg#
zv8#Ube@1*f>I>a6;U5Use$zQ!C8sYMOT2f;f^6RY)(8Lk>XSS#^Uk@aU#y=VUu_cF
zgi7L_kYO+G^Wl!C;L;BoxV=xc!i3xNTzAh4vT-}mcG58rasSpWZ@xGUXaYLFzPr;!
z6B>F{T6(mmtfQsX=sBZcWktBge-^!S{@SvRIBqWC93AO5=@L)gO=-KJ&XbTxJ}v{H
z53H$%yCcv`&=-Eh;OcE?F*ptEyD);Z7gu}|W6S6-sYqJWgBmtRa<a4gKs*+jj032r
z6L-&wj8ukzlh?D25$4eS=jG)d(2BXn%?!q7y(w$)N@&2)TOK}UWb>vRr)8|)Xu0WJ
zZdsWdw3`K;+5<>vJG+Q<@HM@xP2)EYgnUi7<w-E$OPF4gLZLBj&`S`Q6J()7XMAm!
zYDd<CP8%4)t$t{sN9K@jv<W>u+i+mA1q2M1F~az)m6Tn;j>)Xff573Bwl)j=#RWz<
z5ip&2rTyUd2LS6{TxFR6Iq(YjdKv<M^PrT{WJDe9?Cb&qcVRgk?fw#2BA&?{GZ2}U
zc$}Ya3q}%Xd3|3|32ay$php18x+<aO&=9;Q=B+D(43ta#`+oHM)9mB|vTycC8c9Sj
zv%}Sf?Be~LZ_m0@<3r8Obad9BQ*A&N2ob2DL^tIr0}L|nY3IGZRH@MiSg+7-qsQ2o
z;;Je4JFoB*B1IOR@}~#_EZBM~WS}$wgP*M}h?;=#E<vS+n&}c~QThy>>mi&4g!6E@
zrQE-K!At<z@CtVb8AGHA${8i%>u0k-xBK?>>%&Vj3>-l3O4L9n==-)Hxji7L=QWj;
zxX&)&KlfhHTF7NUEeG6|Rj?1yjq-KPM4P-rNqfq~x+6-q#Xrl<wMLSHLr>?%jY0!u
zWo4O;G%mtwnes3jn19fCSMsrgF?H(hdIBt8%TWEt*n=Qp!@CHWqsiq}R1)W^-6sR#
zrz`Z4;kv*SQp$07RX|XwW6$nu?Qj^{pT^qo&|Kd@{#t!sGnA#z6{BN6gsJ)Y`30&E
zIbyN9ySovMP@CCgQtB_{ZldA?AKz>}n98a4&+fW+4I#&l3?!1{FO)=lvh2RUc$t}9
z01Ne{y}LWIMGBIDumFl;YFceV1HY4tk`~D9s5Vor3D_FWTb0+=-QNriTy>fe_>Lnu
ze!&@}j(%{a46I~ajlL@kE2ohoMS6;;@ZrRMa7oQUhsUqqzOn7W2#!PEj#}cu6T**m
zoH1bagp`n0qlk#e3r6N{`FZp^33h|eOAJr2%;ym^oDGM)Srkv6?@y&9b+;Mb($YG=
zi31$1rejj;Zcf;p6x|}1+|kchIbh3)Z}XqE+HS<X!U*nGcXO+WCQ8ukB!uX(JF2Lt
zK<nEy40OpQMQ(l|mt$E2oh@hCt))ofC)5s^;vSd8?!O^|gsa80Im~ZRNORxg;GY@j
z=(vK2BE5v5(kZNz?3U+1KVYtEPM_#L3`+Wp+F$e3UHJfgR|lZO?e?}0;gU$*l&fnM
z|7Ei;j%E}T!k+fby_^T}<#l!{_CBG5x64SWx?I|HZa)|V;;4Dkpv5gRjZ1G9T74RR
zhi2?RJHoH2%$e7>-#@y~>2=}Ai)UPzgQEbN`omQ0#{KQU2@h=6?eVy2Ykuy}b`wV2
zO~(R7;y#c^bZV4d*q+b4o+{ZH%97UB)^(agIJU;d#tci0$QY+j!80|v9*k#Q%oQRt
z#N)GJoeK{>mJY?C9o^i>T#kO0lwY{eD)9d?_SR8V_g}XtA`Q|Y(%ncS-6f!Oqeyon
zf`D|2bW4YVqI5}03rLBylyrl1-o@{I$GLZm^PW4-^N+}Q#@_Dl{(M%<HP@Wd(o2;-
znAo(A?yjy=buNE_Z)693)!7T+<%qbi&BL6u#m^6hlo;&ORW?Drm(ULja-_Q~QqN!r
zt*A;|RRII1@o;Ys!8V0mKTtGIet-)Vv*)|hK<_U0ji->z`~QlH*VA{B;_iIoFsgF~
ztaNsE)-)EVlKqYmz~;G)`oyxuX*}G_a>djRI(`t!OOvh+5nYF1tE|NL=K5;HiWtq~
zP(@Zv4L5apZLRL*rj9Fd<L`pU1Vgsai#s|xwk^l8nhwJ81YzvIz}~~56wPt1T6L}h
zApBP!90Kr{xEmsf^ITp*0Z6<CK0823TMO=&vJgy7Wo2v7r?d;{u)0l;>A1J|p(g7?
zkC5KW_Zs9;`hwEUgibzD9)d+#;X&!wgey5b<VLTQ#ihG^{JshNue_Bsd_{lu?)~|5
zrD}Cm9UoP4mA4umIWKaZxda9EW)94-i65dGHc`Z0UTmv4ObNid+<tP_d^;Cc>>7=S
zI)2|g(jh!t<jO(z#&hE`#-8<Ld{oBs{#RDZ@Tc3aKQjpSKj73FHod;BWp*|<-^bCk
z)JN4!b$46AH)d^U-Ngh)vy58!N^cg5+7d$4zXJU!oSkGVrTo}Hs&j;=54)2ve<PHU
zoMQo)>h)7x@mom}R_KN;;rY6isR;@$#6!j(s(fHj<FJzR)D2Ww2d1U~B|E&mL%<z`
zIkm*~saZN|KHz6SvCibNh=h&thzRs4)c?Hy)QLxLcj;5s-zs9$H5r1Z!b7?u<S<CP
z<<@hX2E%?5$Si|~N_O`^I3+L&&Yi@v#n^O8NQev!41mrPEL7x+-TWg~dy<ig#!Wz2
zx$pKI*<c?7Edq2aOc9yaKxLJ$y&rfVvTd;9VrXS$^<VGoqdM>%?~RfRmr2ZLc%1*l
zrxvIu9s##-Uqe)v3=c0ZIXP4oXN%zOr-b-;>#tigg5mif%Yt2_ls(w#At~7kiKG0^
zAd%BT73kr>C^%V|&P~k=d*{;7@mcm|T2|r9+cbuZ{b2^nb78aaa;^^(wNgV9#9<9U
zZG>A)i*jS65azL%w|5ZtOMP1LNYJy0utx-_#u+n(_^7<b3{8jAALA7o&o1T+V+Dzs
z32RY;fu=Z<uYdz|N?>up+@6_nWp4jg+?K)ElL=K!RJ81ByLQe43IL0~<iof5$eZv0
zfUvR0a;spY`2j1AkDk5&;xf#mI~9QXE(nU(LPe~oouk#t(md5p)BWG2SLfT6KNi=k
zt#g{}3L_QH=Y&h_f@BYeA)0$3l?#+Muw{Ml<TkiDa9}07InVMhQ#v0X#dCIg4?SbR
z?rIPz-oi`pwiJy)j)5_~KAb~%$NKvEh*2DJ!cB1m2VmOurUdcX2-rZL1qBWffw59!
z;L3QXtBp=ilPTb#8dlpbFE7i9hzJSY;-^AI_^4FPJtKWtdJVzeQ~A33Q?(8&FLK?u
z7UN5W$!GpzU~jCiqhmL_Wo^SX&eun{USs&r4edCeAIMdb+bVrXnatU_SuM{G7H6f{
zsM%NEsN_4&@T9;VWE7bC5Ea~G5XcGp63$Lffg~qvj16qVf%lJaL?$LDbMG;^M*|!R
zH=~LvoyU_(Q1M^ouq^S=M7bFUJ9}yns|#)%QJPNa-}MmF|2)JO9xX+Fi!?DELGKeU
zz1M3c*!UR$$UVzx+Y?y7%)1>#tLLNq<n6x=Y-(l$ePo~Sy-Hl1-^{5*Dz-9wqaG2I
z#z_#T%({#HSV7cHJUuQpRt+T9flq{o3YzG}mq2~1qLiH_C+9H_2fk1ba{rS|;Us&T
zrJgvxD*4{;U?F1<pHHU$&*4DGZdy>i3m31l&a0G0P4~I>e(7>}fa>!(Q(?edX|Nj1
zGF@k2K?M&T-3E_sz`h=zSV3v>d(6MK5An5_+NW}*dLm}1J+F+C1dBZGfk1qR2Y5RP
zD*i$b46+;RaT|g4$7m96T4QWPL%6JY(xBy3QFyyou{IQNE_bp$&2%mY=aFuVum4H^
z_ZdM?c8zP%r`6jwHfbSdP=RNzP-{qgIzk=H^fW^d76YN4Zk@HZ>;wlT*t&IVHH0&C
z8+?(6SG$*_F7_1*cjgO1=Mp(P8(v3x2%f@LC%~=ipoEkKDST^gU?957Eij0fm>2*>
zIZKccf?m8)NYJDWiM+!0H!EBtsFUJ2a*;PD%IS_D2)kRu7OPUq!zVJ1u?TJB@U$yH
zjR!Q{`%}4tgo;E_FIwJ}en2)SOuh}Y>9A_kO$++tcTtVtnE`vS^U9Z6Rd`U~9-GDO
zvr8_F$q2eGhbY5wPUEuoqrwGHJXcqq*eYQ|Ce^H9J2fFe5YgHI>FkXSQan7VkekbY
zI2Jz64?rw|RTOTlm~&!W+`{s5H87pF?m9*`;7mMV04|ogy84Km=e{eT^04d#vCen(
zw<UISLeRdGdEwkcQ&%>c&t*`47HyJiJ%(N=XUeupVc`38Z?*0Nbk$|${Q;^^6xb__
z&auSrnd@TZzyoR*IKE&SLMiT3cYq2t>5uQm?Mp$i(-bP4vJlE{yG6)@T5e$(LF&El
z><3l6<j8$Q^HmpIhhSNES7<D89D;;P4X|B+2pqAFaoa<d)z{7x3H<3}|Np+k&UTwK
z=dVN&YBn}kjT1k0!QuvXWfu1X(dio&Y?O{(g`UI{74JBmREXyP)ux`YifQWj<X=9v
zTHalxCV(&eecmXM8TR!@LBZod$U<xX9)wuK5rUuuX*nvZDUP$vu1Y(kRqXY(+Tx3@
ze|jtPVZe_H4vwQ!07scy65k^F69%DE)@PRUEW0Lk<nOh^|MPwu84x~AOiYGhPQ##_
zY`#1@-9_%x;vjd5>2*K(M>cDem1l*01|*HOu>S8jF+nn+SnHsIFDu&tou3W_yO|Xd
zKPJ@9s_pFUr-Q6o|9N;Xm4PG$qbhK<s`vWGNSX|d0}pETFC8Pl2Ki5EwHV}4&%{Cm
zQ%T<>2%Q2E8cb1)jg3XaOf4-#NwmGNYb5BYX#V1->({|Qk{yHm)~*f90sZ^?r*6)0
zI3dr{Ie3H0!u$8Ux38M<HqNc=FAb3($*KUfsjh5lkFjq@5?x_M;GE?9g1iwnTc8%`
z>;p+An2{(y#-3H~&zt_q@Tjnp{(c1<Z_Laf=Sz!SZIIRK@kUK|QCG0Av8jjNcuHPL
z4`exJzRA0};+k0fq*#|X{DpCCRmX7Dl%Zhfhg|A>R$J@ZESQUkn(d&kL>`qsd$C{g
z=IlB|WsAST?U&z;eR}UB_hrr&jKSTJ%Z2{y{Ve4Qo)HNwY)V1r9LN=m%aFE!h@Ck-
z0C|j`i;K(W^9O?QIH5@@?|*qOBBP_<<Z4P&QczKq8orE<h#3BY@!5(fz3b`%KTnB@
z>MZXoTo7PjD66RWtY*R53<5*QoOZE>+}zc@ddeH3C8wbugv;9->jTEW^^}*}ue~Wr
zsheblbd^VmKt%;;@-mI&0LVjq?lxTyLl<DJ4bHLBGuHRk@ZBA(YW=8>T!OM1L?tP%
zhveNFS<(OVXQx;=6ikWi`aHFk9k8A;yzUkedSSuUiE(-or!ZBnNJRSekzbhnlUl3e
zh?Z1_{3M^?g0)Q$L;=@5@8hWFaO~oAY_txpU)X-wdeY@r-f0H64zmarCj^Csn=1|F
z(=Kis+*K!x^ZVHu4~{Nhejm*L*`lz|IoIbEuf03cr=<MT=St&uOTh#Qe>7`H%hKw;
z6Mpt;!+Un#e=-LOWMuEba>As2ZmtE^y+@A350>ZsYX19eV8#RfU9Ep$qGs>laGzZ-
z2Oo9r?RDk9AI>5F)L?JV9sS}a98~zHp)2_`c@C7?e-oN+<f5)i;f~cKz*&WqHj<0M
z#K1ry@y4i@UGv+<PS{5nL4miKPKn;|&`=rd{dlSD6$xQOWIu$5`|ahh`v5D*)*X0~
zd_lWi$N>x^<%%i9gNL_8RM(B%e*EhnsI*?Lxxzlt^p78p!&swMzCsEV)-ut{&&rxb
z-F;PDT>Rt58CK00IV$8BD<`K?nG`l4LB|C-UUl^FfYzIPQ5%er++AGa%2R`soSk81
z>NX}47YJd$_V&JoJA{PRwzm4;bpLm}xeX+5{~Kt!4SUI2&@+P6ga>IXd`@ry%G_B?
zP!*Tb%8?-wc=V`y@S_aT4lrJz%lIu=1_u}938~`et}g!#0<tjNAt%7g``MuS6SjT`
zsQrPP%9gtlfF!k5Rin8W<FIf8(=J=`aR2R=*6$F5!7=3^5oKv%!3EN{=C^!c^MVPN
zLKv&h;hMV9lBRGWEI|~xB!UmtfY?ZUZCm?)bR$5q8HnQ`uy%RBBoB9aXXnjg2Rg;4
zJ&4F4Y4}@EO{OV?ofy1j%Ju#ae<;_2Iv*W{J{YY~&wc|MK8Vl3cgFE7jE{rk9ZIE*
z7nbOT^UbyIsG5n#3=pet>p~FJ!1$K~OLDJ0J%LxOZ4ho&z6$SIV65r(8bl!h0=qd|
zrBcVIqT+0tf|nz5l9QG7eEphI#6#s<ud1cx3cyJFY~c4IiK$DHg4x3U#yI3LKR`&O
zuMFBCV7nS1Hgo#MK}UUDD(Vgvj3(%S)8Q_I(FDL>P&YP=ikeRnV7C_xW(Xxre`_Vx
z(9jqGbwtP6#$vTMp#Cw+#Lli_a0*kr)4$*tO0<IB;<}9C(d@CEPRyE-o*s!Komm(H
zWjwL2;S_r~3I{8z(zhFsQZ(rTTJm_?qAyVQ58twtFlNWcOD|7I8AK-wePmcVmecFC
z22e@+lNO$8I+`6SCsgQL^YcR>Z{yT_7K$-k@n6fEKxb5kRW*p4@gA%ZmXkw6xJSca
zs3CqH$yf`Vp1UYmr{Kc}ds|WuSiywD1sZlp$YD3IqJ&A*)!FF_#v2m^7eGtajAVKO
z)J#VU*!F>T02bj`F&1>-`7heO#a!Fme7h<1`fv?!75M(1^)pZ3VjUcYK)olPqZk^t
z4uN6lCcNW4D)B~OMFH3)FCw`N&GxDL=$+(S@~H_r`0#<@oH=^l&j6>9`v8gE`zpB-
zR*lqup$*92f|5D9sfJOo8(@XIBZYl{s4t|Nj(7s}-WV()U^8M4dk6-fmtdq?Qd)|$
z8)Tx`$E1>q5EJxv!LY8OLDR(3vku$~v5KMUgxy8lzAm5x<UmIQZFtx-4&!E{d?Io1
z&SQ<TL!L`X{`)OLAs6Z;_K9Ht^I*dKj$8=eV{>;`br}jpYF@a3yT3oD$lt|5_y({D
zEXXjt(J@*DkWoVjF@l)GU}9=2kK9;ciYI~LRz^Uc_7cyL&A>7v;YwK>yUp6|0}7W}
zx}-QaH{N{#=)#l5u6DuAQ(iTozOJs-uv+tRu(>n0z0cvWf(G_Ac)W1Em_nKNm*~X>
z{8jlL+I5g@;uU}vxLxuhez2%1adns=zljyyDtLz<tRCHkUWMNIbspb?3}D85D4|o(
zLE<WHe~_kiRoh5d)YkU)`gtq<-rHZUOl3$2ank3of!Zn!*iVpy&v&rB`4s?>2OMt*
zd&gkAN}r}dkCMF6_o7CcrXOKEb}<aT9(A<H6tsnTU-MarQzMt^`O>a;-e<tdSz?g0
ze1|8vKN6n#`&Lm70Szz4GvauCI3JaNJ^X1ouLg-dofo@tvnamGPi>ODP6lCc8GVYq
zl0l03#0ZEc#Rmw7P)|Qy4qpU*G&5qi7vHyI)J3;pvE|iO-A`I92FPYWdbB3)!N+~4
zhKAu+V%#D=;-G&_E^eJxEGAnv1-q2CX_=XsP$BnT7qX(_;pH|-UoZ++?99{$rA8VE
zSwj!#k0yX?uaYhP4Ak2=pNqz5+wjpKW4}50!9=SJhQqS2Y5TIY5d=7o12WqcdupTz
zSVetKxk!mQv;>(@@jCCgfg>zl%Kd1y5<TwWqkt~+#f}K`%lT?W0_SQV8Wm)2MhQ|4
ziyXD0bq530QZ}tGmD>z9l9WP<h&Dn=^U&Mw=dP$Hi`^TA9qHDG$*8xiRTrRc*p%K6
z909a%1bTjiCpw`<2+5q5G+?7R7tf$vrLT>QO*@Lm&1GO&z%*JJP@u|A8%mES*^Ls&
zY+kph7p#qz$M<aEl`AeciqS4K2zj{1wy+FTbv6I}u9%pZ=Q1D@FS=;cLO8=O)qe#x
z<YE(^)0P>Rn<6KnqM*Q{OaS5q<dwG{%v$~L02!$bW#kUueOkmYiH)&%4Y`0I*o9I|
z#BF+HHx_lC?Ca|bI5fj?z=KwjvH(vTxT|@}cXSQ=HccK2p`mu%B3!?etC-BsigLp@
zse|RA3hqQ{#%sP$v`8<ggL$Iw>RJW`VcitkEJPv}MR|Liv2fp^j-&1A`(!=HD+8Hq
zy+9b1Yk4OhGDnz=tf5&IlcCj!e5~FYV+{;(^G)aHKKErWpl&AAeF?p?vF{O4gPvpH
zy;f0sd;3^gn$wTcg86#(fbiho<1vt{nw5;IB9oQ*m(B|MWtEoiNVugNWg7xIFj_?V
z_1x!Zssy;XzoV2r+;*v`uKwM|XDlQXPF)Yt57smQq9X8($vK0vTYlOf^Q!<gJ~=s<
z;p)3kYaq}gPSp><7A^%~{SH6@!Z*Qp2>Me*c3@rQ3NerPLN%Q)Vo}?t8supe4J|pM
zm``KrcM$0w$bOa^!){4<z76;47V#%5qcx4RzKRthCPwAu<z0A{7o=D<u_waExrjS#
zN=8OzM#TyeCY|L|t2uR3r`ARqwp8=4!uLDGTTy?OM=jaGP0k}%uWkF{_XMpTU6I3a
zv2_<<5tdoG{>K!<dm=#B4r5l&L!#C0pa+cJ6fi2Qj;bxwvu~9B_~RVH!lct6d_aUb
zm521jSX#Obkl|Zc&E|H#q!vz|A@2KatvH9+8z8?^VRMu~+0jty;&lj!6Z%u8b_qd?
zwy4O+-~SY4p-Cep%DuSd-k0%M8GW+3I_TqsHiS5cJXUshs`_n>Tx6uA!FdyI%B@T9
zoi*zp`*O@(mc=wAyhkq(2q53!CG*0vkdjJAD7$wKO9*qw1Sf_90Sl#l*UU^tvmD_0
z9pV!o5Xfd(X&v}>M-U9{$hDlc5g+ltT2zYMGPLih2y4xa8+MOERge*<4dFcML06U}
z_GeW8Cbd?ulHRC2vn6>-ze8MpKpe<7ardS1&^?`v>YNTFGV8R?9J&H5gXdas=;3|h
zKX`^GUS3|7-SPf#Z^POaQj~+@ZbW-4xU%K@Vk?Cp5r07vK*6D8|7xzKMObnII7DLT
z+x@RS&DBFqF*<Fdqgqo|7q^~a{&$AOkBm+3GSaZ$IivIXJb}9%(RS*Ud?4*E6-Opb
zmwIc)WJ6Y$E7A!_TA+g_X*F~~ERGl<LGu-BL<(8buS#BYg`ccFB)T?xUxKNJ7i5b&
z#2V%IOB^-sL|rCebMiI@lD}AR)>y2DMP`VX5mGRN248bkKH*Cq$UED=6InD1Lc%$Q
zdxuMLlqL84aS^9s;80UmMy)I<!_>7$opW?>kjEK99AiMHGYbh`9VS6G{h}2(-07J5
zLJKJBYCsUSPDQOjeueQiXiw_y-Mh5Nrk3`apD{~7QA_#zBRX2z`Jh8AG*R_aOA{Ad
z^!uzT&)5JJ2U5S*&cXuCWjU0ge5~k{s0kljSkiyN^e1-rE#6cCWYS1cV2-7o*X?c6
zndcc(UF(6L>Be8zSFfBv{opMB<3}c+gg;6QG+$7toip*&Z#6t-0;GGaOicp?$Xi0H
ztNEXqASqj?m9K&p1;|QB|5dG`J52%D$~6xfK9jre1TbKO5j(-d-CcG=y;y_6L=kqc
zLK#Y)H<8}R;b=g~0HP6S;a7nF7OmsGqDB@c9Y9lPmxtM`94WMTt-rf06oNTJH4z!O
zoc?Q>97>U3d0Bt|a6TX!-fjU0(>fEs)2RdAXdOGY6T>$#yu`N0Cg`{F#Lbxcpb5>>
zHV$PQN-{i4ZM=nD&>9EI+i4)o(4l~ftIHkKz)j~ga#n+|xF+q+OIshx-~)E7F#GKq
zqzM&l69CnQFylMsTSK`)W0W+qXh>$(lNAha9PQC)bM$ZR+YQh!dua*KVc|`JwmmQt
z@|M16dax5}M{6E5k+7sF@=k}uhGHXPBZ<?Pl8i40?H;lcu_b0y5W_EIq-GcG0MB6Z
zA}(50VRW@XS%h(L;kyo3wvj9jMWO}C2J*>l7&lJi3FTnd8J%L3A>p01qTu3d_RRpg
zdQKbu7;OGDJ19qaU^guLnF<5I^7m34!O0v_fwm{YKfn}oVm^QV?5@U~$CBC&aDUbD
zUXJyPKlJ$8-=OG7NFX4+XTyga=+Azt0i?>56*b8P5Wy*HA$GK*eV?q8nx%w>i5YY!
z-}JO}=tsmA-phOwnvR6$;46ofn^D2_4&$9a)IHrq-|Z@p*4u}nd!SRU1F!c(Q3zl%
zG|f34jBG>d3yahmZ6VTq>|J{CLs+!;5vYiCB04z(jZI*22Ti(4M^&$6lob&Xv4}jE
z6tNPZM~zu}_subm1BHSB4ZoRhFWLGjpO>sFD48UWU}zo<8Rf7blzz9KsOO6efBRNu
zB9uycAx#8{bx)+sa3dxim-J!nO%ll`3}rnqzhe)eH0-lMm&>)V{6F?e|3e>zaj?I?
zz#pN<&hQd^-5x!vof%WxJ3c<{Pi`YfNmtX*h+j2iVNdKgwb7(jY^uFIQb|d_0YQd2
zR8dG${#**If3>=)RaSj~wNi`}ZYXS_(tzay_o4?d$K`}J+Q?`|WUe#i6PVjpS06!(
zP~0&-p3+>>hR*@*5G;dA332GUgZ#C9koAZ!6qoAEzV2fD{MA6Tty+TNLCl^M06sFl
zrx7o8Xl!9D&((G-%iT8y8>+Xuf$^Zgr^xil0=q)507hi*)6(Kv0Dm|JkQipPL~J^v
zE2H70%Q<aPJ0Kf7ag!e-uE9K4@5Y3&PfU7G|M;*TT9I1;?}-Iyae-DmGCmGVmBb<7
zsM$(`SW&Ts?3M<eTCGwcZ{&be3UyQ@I#&n}FE5Dt#CI<#g1;j)TO;|<!?{l1K->m#
zzrlP*Sdzi}MZ4^S$)8)0B-As)gNk~Ol%69ZJX|t997TeUPs_xrj2a;Om|2iqA9<*B
zimfy&w9F!TceU5-A;0i(W_C#JJSSg7#%uBpU6hrAgcxALIsjQS5OCa=lkk)oSz~mN
zf8Nr53U(rv@WO^h!~3Ol)K{qaEP4qO-^!twOy5Qyz=O)|_4V{_o87G$Q1Liz1HLq<
zq58kFNty*&18A+tqy6!fm6f^G%h<j1c7b0oR18s%$E*wMv`!M_Z-#-&04<HmFyjDY
z<_0l;+9RjuP%o|P3G%zeYS)n`S`3Qc%F?e(<3m>(I+>3fY5@9(sa2kKpuOo_=<n|j
z;d${H{SyE~LQLS{JRp~P3{Ts-7;gjs7tr80%#%fL44L^gFz~cD4$P((f~=p(Db0|R
zlNazIS%a!uOznIWfS0#r7b&n5<`FWUmROPz5rv!sb6UDr%7?R&kN_K6Y67AQKzriY
z-=60HYNU7j?e%|eZIw)XJGmRc8r(qr3^*OYYXBY!REIDf3fXNfh124CerTh|Uxp@q
zwEk-~jOvY~-8?uerU1f0%ya`YV$hPJBqxs=GlNv-%$%X`jGJv|Z!ZdD4cJ*-v<F`O
zcdLkq570c6w*y*)=M7t0BvkL@gTM~W2#hPnf8cKtAR{9OC;*BkWG~3hZV$=`T(0FL
zc@U?bBL*T+Ek2GE>W+F~W@(9%aC~wC$6W;bz8x#fH|{yQ0wJ6lWfrMuL2o_&gEMZT
zKS-sl1yx2|xC0_z#DzExV}o3sdFl-Hd)Tz}UFAJ@igB^uA`mTnP7H#;{k;zk=PzY8
zHnu9Vs=5uRDvMf*jOtyBxZGk33CMkpw;uG`#7XU8$1MdSV|Fw=e*8EmCWqeVBXr-U
zcaS6Tfhj<9{L0-u`0v{SRVnK`L!pX7Xz`KPRGzp>)?d9(#T)h5Sy5?<x*FgZ=DOip
zqSG-oAon0Ju50M$%fFLuW)wurCpsZ;K}!vk-MwWD0F&qL7J^c9B_f*a{@1nGJ36s1
zmAdoz?8a1rGojb$N&|COxmOrIPsFg>N{<c?i{ei`K*9O@WC3{rI0vHQBwNfKNiIdP
z6HNpE1!fBq2wP=05RFoU*fZy;4>%KJphQn@QsJt(flgQxzZ$#+9%S!P6gVwT+>%Rf
zVAdOfZ|=v8_blJ}5(;ooDfR<0Hs}BWq);VYHBa59=wc74R>@juJv<!ErbPMEvj!M+
zYp;<bz4k{Bw(WWCPw5#pYWzVp8~c|W!eerz`hw3H(-RZnM=23H{%947>IcAF0-WFy
zc0+?(TG?FS6+F<fX)%(<%%Ph!eYOjI^WUYBSCCRZQ=%@sSM+4zExcAm#+HUTXhAPR
zdN_u)eJdn87LU&I31gEaz7^`vJ+6Boo*1YAcf7R*aK7Q2PmmScUUlF%RcsYD6Ixe$
zpExGYfJt()m}^&W?<e@KVUC&!iaX+YKne0pk;lYEP2n(Pj4s3<mFh^qu<HlPPCA@c
zcWgv+7_3)p<#LvdR9N(pF5^1OlOwGZbi;Doz1H~TRw-G(K7evC+HS1~T<9*5XW~30
z0)z5I<Prh1NUVu_%QET#xo^V-cN%U-7kb>ZfuSKmDUqOJEgDSi!o)h2cF+peiD{ZG
z2MV|0j?(r?b>uNCjD<P?6U;s=BxKnGKS^*6ZlJ?@r6MQ~X~YK!-mQsyrX=yFqa*k~
zcPcb2jP44i84^(N8v=d5ZyO5+n<d;_TqF{28a+253G}*^!H0(S3!d&^cBugBcNg;t
zX!t-hsa-?}{xQR*2#Fv98s~w#@CB5X6vObjQT2hX;7^$GZ02o?+bVX{-0<*zlk-q#
zC$g3DlP5%}GSpw_QGbn$%qzixA=;h&_7c;|47&oh3*pWH7**B|6_!Q>D78Q(R&WC5
zSk(T7d#w~BGRs0}{%sZ=T)b#rg;V6g^Xo+%FZWiKs12|R-rP2y&D7Y0%Mj-&xLzQ5
z9PjPDz^;g7L<QWbNyHE^m2ey7B-SI~{R4<-xFz7aq)2iMNe>`{VCVaXE*>6T<+jd4
z<Ky?I;gWbk=VWfa1jxZq8S@V*{o@|(;CzrI@&?!!pfHdJ20W;b+}tK12)KQ8ev?py
zR8WEi#RUDIHpcHgag@-U@Cpi&TqBlF);du=$)nY5GQM3zfe9t7>O3;C1j+<V_0_V1
z;P^RP2`b9U%KG?R0+_g#MprGwto;;Q8U{=wFxy`+)<S2Pco4x07l%(sF_81uUmTcX
zqgWI)(D(-3Pe6@>4!+Y@I7txgb}F2K7&tzZU#;Yd%MUVjf?~<Ys0zuwgE$pc{Zb3a
zJ-(o(zzr+F!R8nZ6yiRqSyOZ%IjWK+sxO9pOIZ;8W(Z9ed^s$G@VTo<ikK+CeWx0R
zNbg)@LkUvT(jH?1OUhr`JVZ#fjof@a+5Z!O4aSfmMJEet=0jonwTLdO9dftSs6Kw$
zo|jz*56@JBg!TGkVIk4dtqXO)CgAGDkM?i5SPPDPN=#16OF17p$<3hx(;3DSEc*EI
z7H^{el77wZ{D4YU^C=uFz_GOn4*}Jh8_;p2__(VM!+_ek$Q5ym|Npw}Ti(sxg%i;N
z3=2ndH+NSH6NlTMoJ?&o_;?@jQFBq>{wONSsbb-1<!(*Q%OxO0{h$9h<!$WUEnGR}
z?M>V*WG&2`%q_&lG5-4{Jg0sdDaR}cU^kzX7D?dB;U6K9-qX^3buV^$V1T2?=2b%9
zfQj3WIUEM%3R%N9Np@wZsxRx!<7p+tWtH1G{fthR`}!_wQ?hq+9wrNMlr;KWT+!~K
zM&DpWKYG`!N0WNT<?Q(CV3TciWd0~RWsw|#^R1uA<*1WHfU}_V%pSgzgg(Rb&n})n
zM4s2m$~5W!OV8@|rryW*d{FSWa47&knn1U43)be|cr84(X3Z^+q{*bs^}G8no0MgP
znn>^RgXkN#1?=ZNnhUpP5rH$ODOmVn4H$(wCWJ~K)LCEtbf?u;ZzfVo!>HqGktUG;
zw(vM)=C_#3N=R?@hoPPYJC?tq@v?E1b_#a9v${2E+3$77f{U&4ERv<3qmp5_EYL>^
z&@pKLq@4)IAql@UITf?LCviqm%>2Omjd}XRVe%Ga^Nx@93-lFRJU_%WSPvb%p5OZ$
zXDZo?L?h3$V=j+o{tijl%aPF)P4$%+qNrQ1x~&nXM4>;+HWSapqxm0M8^Sq?VQ1Dk
zyiu<L2W5rMip(qycv)@uUXP1;8a698nhdd0`ABIS#MJdpP@3;0;m8@>oBB(V(di#P
z{2n9h4ueI3GVizFJdY|D&&CgH`xH5OgNcbWV+^Wud0I2HMI6E(winzv*BY(q(~TQx
zM&}O~BIRkd<J14Lvld*L^E(4wzs-RAsn(rOh+56wUjjS!?B-qV?n-NwKXxR%Tm2LT
zdjs=pkvUhQ;?sw-|McG9eX!&bP+8WVO0rWbgByO*_Npj5=CRbr2G)?H==?zQRx;!m
zH^=$=#h+|Z7Fef>+K;(Lnj&Hy551R<V{<hHjW(TnvC^464z%a1e2Zzx?T1}4xP-Po
zU1X?`=3WWEHcyj`UTVK<+v-i&^C)n2O{^><T&qGup}-V(pb_0hz{iJXPQ0b*(#e#&
znBkw1@%7Oa!)1;nmn6;|v$VLygH^wWEiIQp@lPT$c%uioD6YS^Twnf`-tl=t6uG*V
zG(+;`eGjYPH`a5W?mBxFlqWw9O$Yv@D=n@z5L6~6SxG-(w&QK%0CV3+!^r0gSqAUy
zhvyfF{Zv_6-^p{9=hJuKGYM)+iJfH^Rj&+;m&?XK<!xM(l~wGdddDirEB@JdBf!>e
zbu;BFO#!OqmF%-vZS(Yw=1v4^(YwJoxg?7b3kOM_^*>v}%?nz~tu+}O7}53&$-L+n
zN%EVHv0X&vr;EpTem8+`kBE}~!I|k;xL{;d=mXt*p6Cposp`z5ZF9~DWHa1-_m-JE
zN_;eW<{Py{(==p`%cBls@Vdl8MEXQ^G?Sb<nc{Ccc57K{-wx}qEb7L*<?k0UULv~q
ziH1U#o|E|6QY&(Snx`&HcJGipr`G#T8ujZ!xd+HbvJq)NGYt+)kj$I|wCKYKn^?Ph
z(ijTcz5GcTu^7hqyRQ%^UGQh|Q4keVg%S3W=bP+AS&%hJ4yRB91Lk?xY;F=;xtG@x
z(N3Nr5iRzrekh107^OveCB19*T+G1CQpem`uZpeNn7_;9@wNv^H%+5dKZ(`b6D$-_
zdX``W1cJw_?exqg&qZUudkGP;HYr?^+-v<J*cWGLJGi%$-2ULJ3bW(h@-e#a9|^Mc
zv|2|TCx$d_(>CvNo=ML8D<pOw4T$@8WR9dZicp6lidDlIe<srEMcEB+<(kpiQ5NdZ
zC`c7Q%&Gcx<L-#kIKI$uaDDVrp?2&iUR}~>+hqG<r$TkCD--GE<2oN59**V21M~e#
zWV>f)zrF=Jd(q<w4LWlupnLGB8#u=f(-7@n>66x(7)+y&im%c&4oOkD9|p7{;-}US
zFog0eF_k51$$Dm!>9x9|c-X#u+Jza-%EVLrX`9c3m+j)k?>K}n38FtU>k;m@BAL=p
zcQ18P>z?&WA-D=Tzavrlt8_A*Eb<M*@3Gyv4sjZbdjz?&bj8u~c;~3d;TWpT##7tH
z%`T6e*PO3(-#++eK1Wqh*fCl%$Vy{iOBPru@?06SuV5(QfvB2_^3M{j#6>zO_7sd@
zQ>MrOQ+!4Pa*T~{lQe9J;_*4iW9Kgho(^&Reu+h1?e}o4hyC>Lm@SLA>5aX*bh*_{
zm9V`+?eYE6O#BteT*U$d1G{@a>U`%FN~)zVWS_`wbJU0u_~h-lR99#Fq1l@v@?4i^
zoTK^aZq9S3bS5QM9owfk6pPPU{9u3TJej=s;Gr_}*E)NVbFYWWd`vpeYT%{5v8&HV
z_SNJo&*_oB(ns7NN@2Jhbkh$BMOaA3U~Bn!ZEYOh@!;Z3Qpn3X{7{XhU5Ag}Yl3W>
zB-V3&e(J(Phf-!~pNt<RDw(~h6HqDs@KH*4S{23UAyNc=Q?o*(l-5qtVv|>hRaf{o
zJ2??|BeYd)u7sQT?@W(4sT;Y8T-z*+{~;odm`E?}&&oZYeaf6Y)zCu|zPRd%9kBO7
z^*ws$Q?!xfG<5dp<A);#nSonw>)VMMq?M@|yLWc{q>^Wi@IO)b^(!l|w&Uz|SIy96
zoE<FM=Y78N(R`G8a7FJqJ&gO<QWme8{tB0SJ*?xo+<F-MA}wD*)U{vu_7{4)@+Jz3
zX};R3@$N!$PsBGPWNd9;!m_o#t7m?5YZ#v8&e+XB@z|~*F-TABMaL7X7Gz2se77d#
zggYvVJN(N|RH>X`SQA}EUYEvKgC64t7S(u4rn(VV{uG5j=6jC3zTIicJCv9P<MwUg
zqm{Z1G7ZtPJ<g=eM>_*4G<0hE8xFl0RW?7cl^DD$x?S}fAF?jv`QMi_i22buj-dPI
z?{nS^0c)iNcFh8Qq+H6J5#F%7j3pmcJFd>uw#Qf7XWw}F8HLFl--#;y9jOv?5zyOw
z^{(MW<9$8@iP&HA!j;-#j-c5#(Lz3+CQ1X<P=?oe1ZE7|Q>;$i;{E6r7OB4uu-uld
zhcz;cR}if){eLyl(|jmwUZF8GU;mOX#kykBA3olGfr(z2bQNOm{<Yz6vd*I#+T)(M
z-@_U^zi$S3*p1fLjEoNqbxjacYgYoK8<|TAt3GhGy~-;iG)N9vZLKS7dUl_|PpW!-
zzcpHu(-eha=5+we_DF6;W}`C0mwdF8=W^NSpWmKOaRzm1ePGhTrD!S;&hB;Y*&9&~
zZSOv~;j%?9BpTIdUac3(ZCqS2ch^ca)Y4l^_cGQEDDSQOU0&He?mDT7R!gLsTk)ml
zWrEd7?61Uv<;(62M~#sry`SiEI`?pXqpzdM?cpwrZ3fqCk|RpU9SQlV1lw7<<SIo+
z%m@C6bXB6)Bsr_^8CaO(r7twWPYR^>yLxAQ{Z8T<?Cwf0iNfA++Mh1Br1*uJF|!4`
za8sb?{}rj0O*J;3nG=hpcx=7rBG&xAL;i>JBy)j2<6Fj~^pSVn*1|`+p575h_lb+^
zwEX|!G+=F(y2o>FPpnrbIJTF6ww+VCp~n%!Fx_n8Gu`st-!T7vvF6(8EC24%5M1z*
zLn1TRL!mO9@AjM#@3KPh&n+s%2J;>^F8J4%t#!Yix#s%*=dxVjLaAIY`iFy<7OMdH
zu(&%j_P^lF4q>MDfKQ!wUEV$ymfcXp*4kb((09Bpna-+89cZdnd0i-PP|Km{t}`*=
zrC~YaD4=QOl<RuQr(c5~%ww;Gp{2UqesOY`&cCZs&ef<FeIv`u|LVh)UmJbV6U&Qn
zrhSK|!LruIEwe<j-@-fjORx8R(30##t6n`caQpXuZ*yC8e!-f1XxL@gYtA8Xh9`T1
zdXWFHdw%M-wxj;9RT1ZS5@tft2TwT?&^|titD4<7N2}^UtQtv=>OEz*UcSE^Y>Z>@
z0INggtCO0_%<MW7S(Sw=pEv?~-lhw+!${s`HZkf*9BNzXxTzxNN)dW=XF!#MS>DDM
z6Y9TFq7pHrG1}RDgUC<%=w5uN_H*n1V53_o=jeL&GO?gp(Bzaawt!Nt?`v3`yJ#4*
zkLLJJuDf(~3-9~b?p5`rwu5%@`*IvH;dzhgK9(G$>kRb!so;9aB?!#?%PbuKTkP^H
z^6qzekLVwZgbh}?Bh%EvtA%EV=3D<HCA7>r8YtHi!f<`;ZCkZJuTS2r1^F!Itvkuv
z7@Pf%t%T73)=Kbj3kv*SwGzJ#f4a^HU^QQ`93onx;m;4MSX$Y4AFDVz`1UQ03e@w3
zt&EIwD#jVb#p@V+87;;!_(Feh_v{Yh9d=!hIg!bxS+V9nzMjJq`U_|!8?((XTXE?9
z0+-&03S&}7h3;%CoGj-j_U`TXpehyfQzf^Ei2quD9`u)ZuLCt>hoh}qN<@4VlS&o+
zuIiYN%GFtxjt>G=s7s>;54DSk=X`r+bF*)Ag3VB?6-h?++|G5gU6_L;I^WletL|r>
z{-nVkpLlg}r*>Ae|I{`ZV@S&<w}m)eo(Ut{ymRhF>ESfCI6;-T?0;AN>W@VEdJL_+
zhu_PVa({*wOy)>qD&;Lm>>><}|Im-Iq^2+<u3v{I(eWxB(huYZ5@s1pSnFX=lofEx
zCYh8=v5Z7LY2)=2U>86#m!Z}XE-$Ah!w&KGc^C81n=VKQ?>+jtbD=bq0JiyGIz=nt
z6HJcwm#smn^<i&T*c|V?c9!%@=G0f>Yg}RJ4gCJ(NX9At?lCJb`|C-i`M~FB!P?HT
zrs2VfZ%(@ucr5&6872?K;;2f0QYr=J#<O-Jl`K5d=@BNfU$K=!We(xrpr)Y4cwE~^
zc20AIUNyZSf-q`u!q_(}dWCF|bpPo_ko9i^MPuU1^!zq6%e3ryW#Lp8l~B<a2LdRM
z2HgXsoZDz9J90D-t5~sCQYILqBXK&n1XOMg`_rpf6jKk|+3Kf7yBT)EWoX=fY4~t-
z^sWV&{Ir;;U(3JemY4qH<<r<&c~v>zrGnN9fz7Sv{LNK_dw<m0@81s{A-!OrHBNHM
z+O@hgXTxKmmT^q-HQZ<~8tyWR4_&6f9;+lA!RPE9G1>p>IYOx58hOW_T*|2E=?TW)
zgpYj6nT-KtoPM1#qQTD@|B8m{JP0GDyzsE|QX%0fkCT5eTh}V5NAFdP%|RK;^OrP0
zEXBqsc1bKzAazzVfj<4YuG_yJH!)uRdBzuFN%iWsCjaUVG9H%EquN`p(ETlp8Q^73
z$9PS(o>qXMjhHq;7bCa2OXE;3!_hivwX9|#IA`4+>%e1R-PopW^3=w+W+G~=M6R!B
zwU(aIsA=wy@f=sYNkBKj2H%qdt?WqxUwj$9Th?gnSAC9&cX%EJ6Wwb2bOHO;F_Uk6
zN%>uC5BC3gkI$SP9f>+Scm@yAvP~9*7xRayZ&^)uc}UV`CsE<PHx_og8p<FlQb-v1
zWD~X`RO;c_H$mU>aL?#{qp5F2tyxO<)237(4}5~B2cOd7ma7$N*Y2%rkPJSvSSu}3
zDB0m!47<<Cbgzr#jwRnU*TuY6Ub}J$!rWcM$65R`gDi~ET@(kHD3}l3;$>`tSw7$W
zs^S;;Buop{xnV&o@WlMlKQ+v{cSeO13Pm2Hm@~979fA`o4;>h5<9rYrz6ATOIjE(a
zzU*nzI=UWYjIMp~<z*W$_qz@;ub=65^fNtd7l_FDE}Z+e{8kN~{H0Zy;|7a(jRN={
zi`DYa7L#OLUOjX#S*S)&6u1ZijqdB5xUvL^BK4G0hI!G+L<IZXs&b~5>Lt2U2l}lJ
zr+42vI9+o#?H68YVL$%vsEsu_PoPGE`K9WaZy@oDE5)`i;!+`rxpy-2`IfPYL`7Qh
zar-{X>SiGScq8uO`a4rOp|2;m$1PgS=V|{~XT1Jj{vLDx%;)rO1I-tozHjsXu>FLJ
zH#gQroczU(#rA7RA_m1Xj-u4NcMWU|=2fTfB=|nTA*}stX<Rlf8T7<(9C5{%cgsCA
zmHnx=4t~<}Lc+bFRQ-npFM3JLyj}#|>^v!n6Ha;Yk-diEZ_gp><C~66e4eQ@uZ)9I
zvQ?_Uuim?n_aCTlM#VnY?#qwlul{Va$#(DFJh@$g*@B0TX!g|$xv#yUKO<d1l$R|i
z#_qFb`<!I)*%dR!vj85+x?Z)D{X_lIWF%`b_t^Tes-+dLFFff6!G`Q~Yz^`mY?y@_
zeTkP+HsOwJ>ilZ~mk%_T6%ZR8ygRl@UzhpH92<&hRetzb+eUqnh+mn!XcxI*#<!>U
zijTU{yj&#<k=Iy~JPfa%i8b0wX=r%yBfGL4Zg0a)Pwk3DPMokAhEl$oLG>3a-?hF&
ztRzB%3o$IAi*r(!n&wFbyf2q!x@%{zSLiXK`xPo&x~wlG^KHb4+%1y}?oVw$)<CZ7
zQ0KUkkF@>ao|gM*A31OW0hx6oE*HfmAn(IVy$Hwoq9otRD1*uSdKlNgi^TLl;ic$(
zGMwwxpu4DGsjlEGGrz~V@>(_7wu2#=^bF@89X~yq%V6ACxH6r?l4^&!CPQZZBrXPQ
z)!Ey26i)IT=n^Sg-uDbKAbI(GKD&$KVl8M~&7}V;atNU%E3*Ch>QU!!?aCw~WMmD#
z?|A$z<t>MEnz-#h`D5@#6v$t{v??m@*I}SX!g<Y8kfG>C=gx;niqZG#;a*};xBO}E
zqWZD)%hm4pSdXl+-Z@FGB9>{PSQGNK8<u}Q>Q!HQP)G4Ihb!qs<C(5SFB#{uW8`QC
zQQY`6VR5@%xQBEcS>>r6bJq*Gk(ez%jfGrMvwrLJ3vKG}m2HC;I9B9chaPWVPS2`a
z^G^-nWxZCNu4xL|?`PR+*``owVV@}*-?4?v^f#opA;?1wQ7CuoEUeq>QAVxI;ObO8
zsg4&`^$#=2UGAw&P~DbT@#rGv!#kF8TXT(K9KyJFo@w1yGTLEz``enQWNxvHZC@Wk
zeU09rLGb>8;gIY^Ajh^;`h2Q__iLtfXEy%wv~uJpTsHowR}vWhboU)*BN#80p4eO3
zOA*lga=xa1qR25c^(rqS=|+5VV8Y^wAm^ym&FsDsZr!WA(C(ME6H4P23=_Dfl!qsh
zsDV8#1RSDYezjRPCtpvH{q2((eNlMYQhLvG+3#CT1%I#A`+qI&`?($eQ(Ys-|3B0<
z0zChhbxmLNQVsT>6Wvo(16^7<F@LGM+DYpOlMJ=5q;!5{+V8N(KY1gO$T3NL_0{zr
z7iNmL{YJQO0rSYn+RC8(BZ+1siRah<qHFe9WA}%Ad~C2<hA!>+B-}?m%@SpP`S>`z
zxhminZ2sz_{{`>t;PTbHNK;cDI=hLZPg<?6)6b7_<i>jg;~Zuak9NhEk1w$TINMXq
zGSG4e@cYs*jrX@}W3r#S>rCzaiQ37AcX}OXE8@&bGJSb{#L!e`O^9NRM#SfLQ+jrA
z;5OGhZ%Z^{e0c$$(TPv}2loFkwHo{V`bl+pz(m|dlI~|PT>9APvts#~-#_E7akKa!
zllW>kS)v~*4{>=swNjtIq^HBlNF}jr5aIrw*mJtD{d-;3PWgWH+AoUVds+1U4E9R`
z>~f<LkHgU)TbU<-bFSmFlbL3w`VoS)k=Y6J3HRN8?!3=DcTBWquI@%T=mje7&#6`*
zA<MhYMKD-@WZ1E^Z^s{FZ9g3>4?gP6yLrmK;9;=n*d@RoP^7{K|EI8p?<xH9)$K29
z*mH@j^}3OXWCAiZx2ZdlExjI_>os8;CO_uz|ESCm)-y^lx76~1LX-Lp(UO6EM&w$r
zEj#szClgMhy$l|qo?#=em)Ry;>=hFePUYV603~fw&3BfkwDBW;P0p-R+ZWd{9_;)*
zT+U^^`qqIugtd8JO>pv6UfJ9jBzjWVlqY{SQG3Wml*PjB`&B@qaQ68<DUrWMSwX^Y
zXAXSE>eGJw%dnUq`Fc^zGW#nr^n>_YCN%OqMZ3z)f=7ZQOM0B{KX~{#d=m_@VIOJF
z=vz4l-;<^q!d6_0=?`3apWkJCkr;c-*1#i-jSziBZvE3N+!xa&HJ1(-Np9+qtPZjh
zcHN(JEB!AMx#5+x%2x(h<z$AWnkHi%0>`cglry2>zrH(#<K8%WnEwq8ua-2Vedw4=
zU4JjmBB+96<{L+W)Oge#a=|kq{qA*gV!I`uMx7kNpyh^rE1|oru3YO9U#?P1s}Sj?
zV>aZ{sE}K0M9CiuNwaI<w+gcru+$J0<`!4om64t@Aj9xr=kWY6W*Sk4-jwQECI08f
z<73gL#sWt9fkNHG>`Qi0Y0jSr8FHK!7h$jh?Yzi;kVW+RUa8%={Vvf}w?fmn#IW}t
zPmitq6d9aNVb4K~rt}0%9;RPx=PjZeO$!mT@7J0r51x|+1#)?+>U|iO$G#rUR4(1R
zIW9VRp0(C=?Y#Bm>|xuym`lhkLhtmi#&PEm?(LR-v2}k&cT8Q%QGuW1Os`!?y5}7t
z#hA!#4lM$|s`VcQ<LsD6%H4RpX+&in=#Q!l$LRSc!}m(QxPe2fyYr1@u}4nH`Y%hj
zXSwOTTmDwwm1}bt3X=-ef0loWQnYLEgiq<Xk6#@o>biH24Bi=J{NmXoV{erEUUOo8
z@YsI1CG+<s+MmrV2d^)Gl+SdiBP_Tl7yg8r(RX#3TGyADE@LcbjL8e8?LQz|l)-L}
z*{@H$B>9W-vo5<X`=xVakw--m;*e_%8AYZF_ydnS&9Cj7C{PVQ?auh4Cs%%4mDD!G
zQ+U2*EuP=DO(uA;tYY=5JL6f9>)0sf1kWZ({)l@`B_@USLQv;q)!cylaoNUJw|Ll%
zk29I%qsK&rGKnGxXcH?bpA8YFsXE6`P`MsAG@A+^tbe=0x_sWezT%72w|6x-=O?iE
z12B;;)aUz(wg?8~Ry`v$+pnG;W^UEgJ32WGv5AS4kV$F{qfWl(u9jh5tuOu_x+*PS
zk`;yL<%=5p!@7Jn3@;z$X$S4JVbzFk+(P-f*Xul{<)hx`IJG|6PV4NtZ`;w|c0b?~
zU^V+9ON@0q66`>ZKSF=l^H=;sn6_wj?6-QY(NmM>@ABL^nySuNfAl}oSGyCMc{X_B
zrc1}c@~rN2Va%@a(OqFa%_v=1W8c6qdaHMzYAnf^{^fnF)_vZ(Yc{zS89A@8jVtl6
z<3E-mlEzrNMN!#nBO*cF4^^rp99n)S7+WG5ze}uJ|0z$)zO8^tA!uGF=boN}^I_@5
zU4kZ^eni+pn;3MEm=MY~5Ol<UW11mL%qS%5p`T#eksO^V{!LN$xjNE;sCIftvFkjS
zeff&R?*@uG#O&CY2xX_9$aRE;k!0gC`NOn-6_7M%O`6@wMyg00Xev7MitAVn2Tdj-
zYwGWMcVoo!-xm{;HliouQ0j11QhFV2$+_GnugE!s7dSF8p5>nT*6&qAw4|STp2Zwq
z__nN}U`ia0s_zcA$^t`oddPaH)Z{$V>AwQ6r?FSDkNylWe36$3sv>pqb7!Sb4*u{H
z`=3zkaTczM5U+f8gN9;DjA~rH#Ymi;)TrRFllUsqY9hBU^H6QE`B!yniXr+(iUZ0t
z(Y0e9nJ!*cf$L*b!~ZmtV#kSTuBtFT^i|0^#b|MPeNV~%Q}m)&w#1M_iwhyAwOfRh
zgv*Gpod6@1)HDB0amXN>@d#<`)BY_s<KV;cgU;X{;a2qJw~|xL@b~K==aVE3q1`_p
zfxnt|>PWOMPV`L8XIpzXh}PfRy(DJyOW__%u>M=sPbEm_@)isQR7E97BR`UU;|q(0
zX7xopa3@?b>W%eOXApaE6}eCGi)an~=)I(P?i+$|UxRu<g6Iz^H6qVwF4CBy$yJ}6
z6u3Uk$yOUg?|2vZG<s(tazM{RWwk|0j_I{<ZXQ868Bu2J^PIbr1Hz)kc?4@rgiMGo
z^?8O_MrR3mvgsdD)r82hwHnY5dDn{ZvT@VCQQVyz^ck}kEO&8HK)0+~$W42P->*P~
zm6aAu$K3wC5aqd2+c1kt!?UD4KJ=DXDVsliHvW#CwpdWG@_6rz-SqvmF5Nf&rAIse
zKh`h*J1)b`Ey(+SRlj`EdlEA(h}E>Jdx|4Vr!;yW7fF#PA<ui7wze}X(m^$-^YDgB
z=|=%O+s@LzoN~H8)uq35EHRXCuwS26Ol9kED!i3Z8<yf~U++dZl>B8p#JBzj7a_5h
zD`PgyPr!go)lBtnbxpE-&V{0G?K8vhDnx9)7Y*2G?=30rxfQs)z?wm6Svav7ASIP-
zCwR&uZ$)o_+8|nUlXY!iBVh2=>7%w&$&=@~HSH%Szq^%?6_>r5a&5fM%;e@R@XwTR
zjGLPY9+cjcUVq*EATnvk@&3#EKNxWjvDNDd8=ujRoEMoYNt=6&#r=x@s@O3fup6O5
z|K5m^kT=(&OZ9V~c8pU_>1RZ<rKDu;wo}F(%zzYWhfJO5%Q!Ww;3yoLzkf+ylzEqv
zRxaavL>czndFefCSV_@6eAkzdzLxFtatfsw{slMpK<?L`3pq=quck;}U&szQNC~ux
z<i5A}`1aa3kFMtdgZF|;<|McHE@~<751kM1wC6`&yi=D~uzX$m=fKuZD@X08Zc5_!
ztEhI%52p(?Nz*J!oj4{O#xfSH-0iJw!9u@Yi@EBbMA0c#e(cVuN>7V<r{UVzwi;!9
zz`a-<)vv?JZ94I(XSl+3?o(vdn5?^W5nfUT)+%|&Pf<UGrt3RwZjt%O8Pi>5(Y(fA
zu6y;$c-U!;<Bkj%vTZ;o;}0imE|rzssK}ycNs<-?2Q%5v7YIZJT$krgr*Ur=$+cK3
zd#e*|d(?K`T#<iEbDD9D{~%n@^vY2uOj6UO?fYp#2{!(pMJnBdwrDo2>T^#ac|BR#
zCqMg+!s{2b_(XDxU;KDFFVT59*<ULkVakTG_)+I*ct~`GW0Tli#)SGEnerbC+eXc;
zw1D(!K9%fA(c08;L+gUiy!4NsQSCg^E254Nw4Unr*se}XiX1*HmW^T@Ffi_M`u-6c
z)6ux!^H`djY`8F|GDL%NCNXg(i6HsC+pM8(b_sRYSbF@MKa`hCJ=UGO-SuNSLb{ka
z4W6m<LN*oTc)leyi&`5K*^P|*FHM#5XhpWV0(+R-eoyNad=tEL->f@id!nm^%e<>i
zd9_>NL160Omu5!B-lFrVCo_bqpA>LDK15;_F?sIT<0SUCmw$@Ak;3ZC)UeA<<gc;S
zx6Lf&<bGPU<4@<}ZyA)XCCAOb{w%6ciuK9-o~B){uAfgS#dJfp@`piqb37=1H1*T<
zC|?gSHLZ&7<zwu)lGV*naq=saE!iFI{p&Frr%gcIa?T#Nj}$PeoUy0Om63LAjI^Gr
zM9fTa6e2rB%eX$Yc%<SN=}?)bgCxB`j()KEbqC4$?kCiwwf>QO%kNX7Sxg+V=R6D{
zLPTcUb0e6_zIYTy#l`+wH#rI=qvAAMq}n{yH#8?U*VOe%snOx)>w246J?-Rslp%8G
zN*GNR=i6^w7&fB0J^cND?W4~8nMYB%ez+F(?3z{ph47qr@c7lRIXkP_lilA}Le{a(
zkIHf8sWi%(m&hY9y;HDj8xp3|HtVJ8<|tw!OLqQvnI<eA(%0K=Pj4+%F8@jU@rRQ{
zj_ZEFeTk=oH@>H2OW6V&|6{P@`G1==EiY#aPCZpqTMILHPR*C5?zew`>|kPL!KrIw
z4onvTE-p?*3mYqIcWOTPQQF4cP0hkp#>v6i$<e~mom!An#>w8vRnytT%z{(a!o$YQ
zLPJ6NfAM4l9`OtQ-|%F%GEub%b?<lI0mwHHdHQ3h?32n)YkdA160=8L_;y=elv0F-
zTc*22EP4JII<o3gNV0M^ks7iE%uPXwAz$WxpQPFo6XSY>{mHzL@D^;Jo=mH)JZYZ$
zeDau{nGhpHI#=z#{`kZer$|UdbOhGwK$)$o;ydbvb>mAg<%^*Z32mWMGCSP(%~odn
z|M2zZ(NwqJ8*oLPOeaGKN9K9VJcVPPl9|ke5Hck)6vs^FIb-G_QxuVTCW>STnIc7*
zDpS1s==uKMwcbD8XRYu0uIE`CpL2iid*6Fs``Xt9b3FBfK3ZDuKy#PR?6qFV2<Xe$
zctwD^gYF9poA#jd8g`(W`<_+U2oDbrCb7V|vNR>B1!gH>To3+2X}6&_5I&cEU}RSV
zOdi95tY&XtGO!{pEKKeW4q<^5eI(soT`EmRAaG#<PM2Un%I*ZC2t?%C={s1fmSU`d
zWRZX$=v4D-OXpaGT#_5c&Ig1%1)pMSX`sd)ftoW2)WxgHJotg~aRL8izP}Z+@Bqwg
zpj?<-ARBn*B&AZN&6dH;(C`n84GY}6|MlJ3x$T`DrW3E9bF{VZK@CYIm@h=6qICP$
zFP2t!pmo8J?<<%Dg~4H%z{|39gYWFP#e<c>1tu;`8uN32%5&h6{H%`c&e3u|S>Hle
z1ljwG;@$1QYbNB?22w51epT0I3)Lx0npJa7j`lsOzk!v`N4-!Oi6^8Q92~rzg8@g8
zI8`>(hmRjaTS0Op(MME7#6Vd&ta%VjN!u}-KY_%{;o%NKPW^&m-@-f}g1QJZXf?=Q
zz|MStQ%v=<{Ctr6B9y5C38C2D#DMbL+%u2GKz2>|uOkS_I=WW_mswy9*z|kgtTWGJ
zX>DUu@Xf~Py!a$&7tNckxx-A?jZI#kXhMXl==+o7KQe2AQIU}_w4pp~3N+|llbD#8
zI5C)UN{?bf2%BJ~X^vWeH=FW6PDv?rw^c6h9q7}{yU(|srKF^^v~{Qkvir*FYLo)w
z&6_t%Vy7l2WqH(JFW5Odzj<^LKu1THK2lUr&~~T)-7@$-!k+U2t?jj#8=$oO<}XYl
zFMK=~D0}b%sBfFM8YK5cfx-;@?#*$7=f>nljH8Hz^mM#lFo$(L0$L(GNXJC{kNvPQ
z@a?$t;OvXA!IXWYc^b6lnfhYQtjO^2I3LmL*J*jvGBXK#wJO?q<_}*tHK{*XfqBv2
zV0a(b3B$1={S5u&k#|wJf!DDk)|<RXwxnleEy>;7+ysgOz{nnW>-=TFF$J4Jl~3f8
zG=8{|5AELrhZpDQ!@vNiv~<fZZF?$kByp7*fD!Aenoff2>fA64bE3#wuOFRA*W9$v
zQOm)3wYj}5@o<-Yrh97Y*7U<d@ryjdu8Hs@v&}xbJ<jEA?V!^nPtyqVKci+D)-`*e
zQF-@s3yep7ho4<5upr2m@(@U}j62s&;I|A^-=vmqIGzj7M~Pg)q?-WWBkR&R5&adv
z$AhAu{g|$|S=rd&wy^pm>s;t*39k}HixW9ofMg-gQ()=~&%~(ah&WR-3*?-C6x;<=
z#=ug+)wcbvGFlXB_t&rVU7+=USy>UtNv5J(2Wch)j@dW9lym0-+OO?Jm6md>DKZ$>
z+ThvrynXvbC@uoFJ~NTq;vrBz$HvEPhXbQDvLy5~(&m4~VTY7qQgF>gRzJnfyuoY@
zr4}2SE7pzf54&}(OY+Ji?(5PQ>w94^xi9T43{R(%!AL4Dht}EwOn|46efv;~<HF((
z{vIyZ{?<gpY)<DMc(1uS945IKos9~V9Q`S`lKz7-6f@sE%6kHOPjWmp87L|6MMt$+
ztnjp5bIXBKtKtz?`b|AY!#JkxFT|QbLXx1?P7_duQ$o5c!z>2<juoeefY4Yt&M`=U
zzca!py@MI?w$`?|Tmz2Ic3`DqS^MxVw8F464$qD2L7A&f3`{-FJz-M8>W9EI+`L8~
z29)_;%v-?0nUB!A@fT*=&ah}Ld>9+U!TAl2tswe#u_IlU$_EuZrZSFnUD9T>#}hyW
zH-YO3s*>*VW!&d7+vD^&S|hkm!?qFlnGEuwvW)zi2jXX%@Ivx#;J;P1{0`wE!gs)R
zbL^%}h>7eQTE}(Z?iNHv^b7h?aE``}g8Ce5n1(M!HyRlPTKBBitK>s5-QC?<JGjM|
zmtfVt2ct^u!)INArJcNB2{A9%Y#$<-7=kdnE~3)3u_N=sTiM*JM$%W<(irtj4}X1U
zyPkBx^9%-;CRD-I#ySU#eU{DOh`ajr^}}CiySae>Sh4TIUa1gbVM(hTrje^#TTdV$
zM|?66`;se%grER~SYqRlA(PGU+@!|`@(#$3oxsa$0Utn9P7aOx#J)qfPh1tWAXyVr
z)kX+F$h$hX{hZ7X5J=?m5V$<=x(!mG3|SYCRSuSI#V)>8axlpdav&|`4I1Q+fGLyY
zMoG{t%kw02f&g5=%5i;-i<cKUiQkaai;g(jLU*7yt?lOoB5X|0)H#0=VZFx2Rty7v
z@6v1!vfBQtW>KgyuN||nx3m$ZXR~ym*;#V^EHxn-CiDx&Tw+AnVuxPD0@Hu6bfQZh
zrN~6KyH>-g8D;&pvT`8;0$I|JfOwMF-b0uoZRacabk{cKQjla@p$%t^03!-F6)`Lj
zafW)6wPea55#vf|$~4dPrgy$rR#c4lZ990JsyM*ppl&LVtHYF@EqBr*o9n@xdz<0|
z{`}{K&al*BQg<z^fi&ly+nu)=`ss->hwFl3R8ra!HMzNucNt+kLCu3$mfn9FtQF|D
zAo!$c4uMN{7noA_xe`yijTWPCkdo$!a|r5EZ{tW05{mk9Be2rHSy$-jZq69=d^!BR
zL9hnb`Le~P^xg!hi>zcU6LioPo+?D8k4c{#dB`#(-odJzD}fBoW3poyS`iSxLPMio
zl&lteF{vMIAJkv)=<lxy+YUuzm^~WZVDdrv9qnCXhVxUOsZN4KUX?OpeFKE^IL(K4
zWu0h&9_1dA+)%<^V9?vT$5Al`7xzkyVg>deocw7$`SoMg<X{1&aS6oxIGjN@-;=CL
zgPfeaJsrkNLSM!Xnb@Gf%3M7IV_7)5YWms}kF<TrMw)=q38s|^5qW3X{gGRLf3_TI
zuDp;r+QcbHY-HD;r@!L0#y9!;{@wJ%{sX2WDuR}q5?(_mT(VaS;8r%uzyRkuwb2+P
z59zZUHn2V&0*r5dfEd-ybW|`SEoWlfx#XMn;xKTKnteuI%;+KMGzj$A8oR<VeD@pt
zDJsE?XqP_+bYAySb#-+~YYwmpdzh}E|B{UsHd&YTpzW*)^<p0#tg+c&O3+AI$t|!6
zk9+J6(R1{dk9V{dKFA4lUK0vi;3@HDLWr}Nn3(Lj<%Jq7q*GB*xng}d)LMUiy4M;{
zg6n{k!{hTGhT$@X_b|n~@&Xh#;*~Bf<Z5W7fwi2hH5k0Ziqwla6u}Eme-;`(vYI%@
zoSxYGga)xT<TMJeCSIp6mz(EStwII{JH*&XonH1<qa5qKGgvVT10ByakU_a!a3cuU
zS*M8`gE~iC5B`Fsx3_n?^K}^b+2*5qzXzcXoP3)>w<3#9LFsj&d;@HV#ajYg$bn_c
zwfkhjZeNsNl@KCWSjNn9i7{dPzQ3Ikj93x<#0+bDL6NwAcZ;MSfg`iPb9TrPE5~4f
zLf}pvBSP2+oVXY}jmVX>F)O53RMR)mi;WsMS}(*^iUtj?<X;dEFFS_-C!%(yd0ac6
zpk{A@chu*8V|n|aOGh%Mt0u-i#=yiu<-=@t0OL_GRIcTbcq_ifb)jRxH7@PJTK5{a
zM;v$N1PGA0oTT)ly|ZmbHR1k;t5^96b$9c5@(bPsj^{RQ<}JY@($)^V8mAbSEVMe4
z9-PyaWMo<bJ*3f}L72Gx%pn2!@df%)TmhW3_zU#6-Z2Os+_FU`l&;zIT@676D?QOT
z(Fk!HKNHTJc15s6S*&m9lZ__RAiK(7yqNl$zzG<qq%O>39gI1qrlv-V&*S38HM$B9
zqE%Jth0mY_Pl`2=0`kE&5}CVw1kK~1Jp|f8+{Sg`LsyiM?polH8ElYUS@{F5IQ7==
zi;EqEQ{eDav^x_E38)-n!_~xR5-8?uc0r`k4J62%T{`}|{$ixIw$^4in2W#0h^xOH
zd|lLjtLF<n#eks(jC_T>`u6cT0WU`^=3NXItzn|jZdxE-2by1dZBuKLmpS0qxZ$!z
zwd@|{^uy6H<$E&?=mX<Iah1>=n>gGY#uqLF+(MJfmY$w<VCA_xKCn!`)Mp!Wro%*w
zINs;6EfO|aBR%ZzT27SI#VgpERYSuG$cp0e_#r!GDz>$=OCWzVpFCgIM$0c=8>%94
zCF+|iv7dsnaxqM5zMH-taMVy($ddm17A;aujZqWg5hP=~U=sYUs0sAyV3IL9Iy!AW
zE_s<giOr_X4HO7qG%FNkE$hqcs%2r3qfrIpmN-Y~+rge;a-9mSoFNc8%<)5C!C<@V
zXn1&dZf-7U4uC)yfpj1D1vZou#Ok`_+-Vq&QLk2l(Q+8|TZ>9OzXW4MK?!tX=Z@!K
z#=W6|5Yeyq8UpY-1hDnQbUB=qn;(%F-NS=}4s!Z?^VCC04qjqL#@k{SkGs1>QmM`|
zUsLI~cUW_yTc&@2YjC4TK#ic8JTvK*0R4XTxfl&!JKP^iL3Jhc2o^flmtee8JLD~(
zze2Cqch>c>VY&)6HpcB#aQ36OmkZ^0Ev2a73o2QvTh3`wr~a^sTc9T+wJ?VH?vOfv
zQslX_{qL^EXJ=<?+>!q9x$Rd0kvHsko0SM51Y{~~H!`|3*pp+}mNqtSytIppG47mE
z92Zjwy*L<6t7c%EVy+&7S}vMv@h4~lyS(7~`ugJHJL&EB^YF-siO<|nktN*A*flmQ
z3PVs^3p5D_ilU!sl<4SipIJ$J1O^`qLh1F|T0WSr(Lhm#^P}kpG|S7NI4*GQUHE*I
z!Xsilf2HGO4v%?XG^14DfFJGl6P0LX@5@@^sf~Pk!wb3s`J9D2Ll-OQC@Bpk4B3gu
zuGuMV=BJ)S?RZO7JMzn|IU0fE$=(`>DP?75K7K54y)YQlHPj~Gu5gxaXxrQ&P!i(T
znMSA`ZXOW`;Lvf8Z9x!O;BmN@oTR{hp0c{O7CeuNb{{bCT0G36Z`53K&liBRW3uwV
zN})y0VS)*X<fI)Zx=Qa;?s+x@r2s{CJy|FE*I)c#42+?gz4mkvXSh2_LQGuy{*CX0
zWqK&Y84y$NK_u!^N=atv`g#e}nHqMWB=fEfTcC@FV+-}9fp3u3L#4=lT{iUpD);{i
z`2WBEGt#oMS^%Y~B+sqUPqihrU33LXR0dVO<uO!tx-gs!Ri<u&ReqLb==d~q(sLL$
zRSR=AtQd%nFR$f<e`FRb%Oa*`F9=qg1smk3p_fdNI%srYGp*k{O|dkqKv~(KqkOEr
z!VteZ@dhqDEu7tJoqgy!vQvA4a8wr|6XLyQ9Hy)SM3u2ZOD`5~gY*=Pqw`v{3@vRH
zjJ;`Y1_?9}Av^sXS8`_32u4Om@Flh37H4AW1l=XbJ6LB&Mt;JL3R1G85g>92gLN<#
z%Ka$SWSO2?#9b>}!fSmG3>QJC^b)_}3Lc1Pfhb-YD3^k@F6eX(`?Z5wHE3Tt!2Byn
z0LjW@pHJ{e`&<Bwla+wz8wm4`eu`ICR+e7`(RQr(1V~>9Hf7$Mtj>ek$$7VVEzrB8
zkI?cD>G~XS=nVQfknx{;Adh@z#e|SCfXe|2wDPK!gx+?TC4O$vn*G5H?6Yid-^S?$
zu|2qBAb7g2udPizQzVeii3O_>>rgH}OGJALBnd)@Ku@d#b}vY3vGu9nya`ti)_e=q
z*^eFY?(b)CJ^K6SbAP{}-4MU9u;S@oEob#X;VQ`Kl4JxgTM!UL)!gpwXa`mEfc-C&
zD=2ySP7qqtPr#f%7x7E|{qP-pZ|AL5j%A7U=9lLw?p<1(=zG55a^B+A1NZvYP3@}k
zVp{3nOu~m{8lkghvp$M`$=lLYNSL<#4YQ)lAQAN%RyE-g%?3;hJ`pd5{fsqFz<{{S
z$M*I>=GP4kvJ&obv=!d0ky*DD7{62{5{yuo;9f4%M&BXfo}|{&I~MoNj`B<F{?^eR
z9Zkr?^Bmx&?Vqq0T279A#rQ=<bHqJZJ*J^n^?->%fhrUY&0R~&YBRNbJ{E^tTU*`J
z-w!eETaQj6`Bv`#n#5~K$~`EPKwXYdfZPvBM8t9egOT5{H4p}_3y{hQ+V6w+4^`U*
z1iW-}b8}`G)U1XafBVNa%*1sTUlE-Fl*zz=!h$3yg(J3isXP_}f4NV{r+a(TbeV&U
zMvb>_^e;VfR}vg#p}iX*JMgH*=}A>fYA;-hdh?bKA4=>~I%dP=@)!pUX#riiR(yYL
z^b5$?+nVpgn*$?kV}`;2<h}SCPAzsno1@ng1I#3=bJz;0e^#Wx;CzMSco{hbh0^i;
z2M^e0IpBx@uEx9*50@2om>UV^$}13VNwI3%Rp!@y9_)kbSHRu^qKBwGgp8EbJCfLv
zMMfs4SUM=uU$|=XQ<n9j>Bh(o0e*7Yxc|*4p$7Zce_pB_41V;UMERhY>vem$@^mC%
z(~s8cmLwhQ?djE82-R!Va;Vt+;Hx+uTt?@~Rdx;!4pm(-U;q5)*>fC@lfSRh8;pz%
zkMDnJ_@bO;u~emIyH_V~m*PmoHMkoEP+fw<A_&$eUcDmrb-bL?Llee|RWA{u#3UsX
zyC8qWiU75i)E+91c_&DSoKUP}N2ob<dp0rBKcISG=#2k5Id-4xD!IOQhJ4b%`KZJN
zODih|K?IFsHpthp^}-B3LvMR~dyd~W-lR-vNXIq3i*DJ`j>C!lZZ(W3zf~WSVzQE4
z8!<7E;DH+>aS33o_$A17KRh={<#vgTjSX8`UbY$iB>v!sse5fiL_g?u#9;<rCjRA9
z7=4oAU_Nm01HESLPVi>z$mv(QRdLLYO|GrzqhmktLIX(XJQC18q6m<>gUW-87WsP@
zcgT76HTeutc+V!uGzRith;H`)V-mSH$Nlr5ubq0;@?!`l7_)!r9wS0u4Jzt$#bfW^
zQ-R$VD(*<0<ppFa*a6TMKOabH(hG?pQTR6Xf=$c`f->loXpq7ndb`itp&<~6#t9hk
zoTW!3-(fT3BPal!T@@Z5cIH=U%fq_)5X@1Y_wUyyk(SyhDJaBE*(=Pa0Sbt_tXfqc
z#yH#4V+WNM461)Ir4u(=LJifupht0E{H=s6OTtPWyors_J&U|JIbCx*Z1<t8<&*q;
zrp;l9xet}lpe+wBf_X)12Kw;V9odc@>Yf;!2%UUazZ}liLH@bBmrRXtMCZetcUCM)
z?=A!4ZYE}pZGr^3-GLK_$plkR2Cv%rv(EW%cS)||<<k~iui^+8*lL{EUCc2jUS3{2
zz&d1A8>_+8oolri>pue=gDFk?^9emH<yU^NMF~bR|8iPu9T6axGFhP%ot~Vb=I*$^
z3r;<j!pXI_uP)Hbh$Bypkl@$W^X+M#V_rPAyf=D%T1se3cbP1FgmR*`C~AFfIRY%7
zPJgS;jdLNFUy}BJzja}@z*#8jj;HLV<~dqkC-)SKf`&WQ9O364eB}wmT`FUMa&O$A
zqwMoVM7Ltd`ejP^nB0{q)pg05;V=>Ex({<JW9C}%Mx#HZLY4}zbPuHjsy|v$s9L+s
zgS`*om%TXMj<)lDiZ}R`EHiadU_Y=AtweLqf%0qkr=WoX!&?u?9B9U>3!Z>IDEvw}
zleX1W2SV12lgce4H2jBT8@4ZXhXYP*2Fki`*o{)Jla}SfwG)ql!qLirbHQ@7XbqkU
z*3!{S=UB{g&8?G@tJvt-ql{s+e+I)pE3k1&8IBZJe%n=x|NZE87@Ko`x-{?7)0Avw
zV_RCTTZxlPWwJMjXkcyhgpm}AVWh~St3ryWUqio|Z0&v!gY9#&q+5nYS<U_rl0|r8
zLnE*k_j0aW7lE&wSei>CxIe*{Flg6AhK-6C^X<FcE?64U6Ti(@YyXqWQnXyf?EgL<
z*M|gYmp%2*cB|ZyZMI4erK~6M!=AS_lFH<VvX#VswwC|@etboy0S%VVepgQaKtM%S
zu#JJ&4wD@netD{XI*%@Qg3ycNtnwdt)~Crw34dzqcLlLDk6G&f{Q|c?bH%bWW~e3j
zO_iZ7!$S9!gDsI&T)7%*8WC_)Dse;&HM(WEbbJr^_ay?7vSewVm-^fO$r@o#_`s8{
zNtXTh{R`XB&qZ(hpP=RHMV#PIX^o4e`ERi=SkC!Vp8n-MZ7+9*gU@ec6_V*cH8j<2
zHGJ#uy&e~w&^A47a5T~Q-}A}ZPXB!Nbo1czzvZQ~Ws_GEk(nknEoEq>xwmHdEFLH4
z<QgXUQ?|}L(u=I5>AyGLPFVgV7!$&Gbh_}CUp5r%cJc@O6xdUaexekM)lF^D7gLX$
znX6we-6c776Vt00jY%Vc*Lhv}^i3weN_^X1_vH8Y=do#jYI}0Lo*b{3Pu*F}QZR|n
z$y}hk_+hRZyG8YvEC+rldXmlUP5j$330{nE_TPtmjs3dlm(29>c$347CO`*WC@s+W
z^qI$d`Me8P{#!_mIRT8TR$P4W>4xTVV9VBdO1CAwJ^0I`@XRc&&NbPcy@U&1o+%Ep
zum8PJPSc3HY{baOZI6Ekm1V=mxPL`ioz8#PP?A<vQcqH!as7Ac#fRIXskj+dlh{pH
z=EmRdi8~u@&aS28hTTSvR}qDxGWf%LUy~p2pHt0GV4chVfG>TVPv9(d)tlkpdBjg`
zXb6|(Z+E&(S&g+)<?Pbb3?36PP94;ru@3m*Kb@JmYBHDTb4wu2s5dmHXN<TLdxG>M
zF&#4+ukn7$B>r1@mW_f{;b|pFQiE>)`*cO`7P>EZ9ZBt5y?T39*Hp8C>%*1K$~a6=
z^fKZ1w^OOH>I!l(-gNHO7sT0Bp0J@b=A;q-j!KVa+S}ONCn>|_2WBnJDprnv%AC^h
z6%r_rb%!4|o~KEohN@@P23X9!&tP|{6M3tTs_4S(tN+4fkERCfn+(q?V+%*aiyNjJ
zkMh@t$IIPzgFhCYyL*=(q3AO;RB#?`eD3tOx6Ch)_^M03{C6r0?Ab=2(#@)td>r29
z2uYY05wfw@bD>J5-p~we;*~8?4K<TSe4cGmASjeCn%R<Gwj^?|-kkEccBgu2^6y23
zc)6<{uPEyi%fGHKR54^=5l+T%G*QIKahz;9!-0r9A6WSQhWE(Ezs*NJ-5;LtHdX|e
zbDqX5@$=3ttEy5Z_i8cTpkYf3aV68PJn=%fKbI}DCv2_nPZ?@rLug<7_C~?c+t@NW
z-1ve*rVZOS`=8VK$%mi4S3%l|)0SzzZI&x6C};*-*Yi1e*fmb`1uq0z)Ty%~{IZQN
z>?Mo~{^k2sfYeQV^)5{&xy1RZOs-RmeT9;(@~U#@F$f4ReZ<;Kf+Qj=75nq`@)S=p
z^Qysro8M1z|5?G46=D{h4{rucIb@1$lK=Wy<gHM>ltNPMeJ&gX>G72CZv56N6=rIX
z%7+`h_QebAvp@1>7~UwB9}Rd#Ze)smqKsa2&N9A;Oj)fBWAPUPW+L@xmHUN}!XwpR
z-1s;8)O1j9n!Xpr(EyU+4XstMSweyOFX+XBO4%Xk8o{kR0dmnGZ5Fo&`QUKcwa)f-
zbOr7@BVs%RGxTu0<IbOdFGt)<U@YwAdkSf-%?I%j1<qXLAtp>jO&UmifrO>hOa2z9
zbVAA%8}?k+!9IYR@4bK44SqC2dl*3O)|ZxWx^6YNE`ZAjz3qHkAb{E*Mn_+RW08B@
z3zUXTOK6X?H`Tu*{$(m+cSZTw0pCTkZ@pqyx4%*YvVEb^>cM5pA2%d*C<=3?OX8Y9
zcx<q*FTA&0Y;gpDeDIcnMB$6&w-(#3-;tl2vA175W_!KsfW<lgehar&4{dsuS6BmH
zpR1KKO@LaS9kmw~>T*<~$c<uuU4~2K?y1m-!)k8{+=0d~D3X@;p@-f1eGXC7t?XZT
zPQzc)9kw@BV<OYXl#I4B?<FIU(gIbVXF@U@Ckw$^{Sy6o?ZwwJWz}aho9`r~r5U9Q
zzNoF$(bbK|KG>*CV#Ge4hv2`}O!w01hfT;~C&1}r_K0azuM@C(!lsP^`hiBT$ft8*
zs0uWawE}fRbb*KWU$C*AR!vtwy3A|m#t25zdw%gf)GrAUisyfl8bmXnCy}`ZXeG)f
z8N<j>^Ts=I0o9JR3*4(Y4YW^Qsb!jt;FbqYl(ESXi8xUqSN2Wwo95O}p1ySgq_1ac
zib}lC4I48Za*i^#`p}o%t2HFuO5sh6Po!P4dKDNNFK&JMn?>v93WnEYn)~fUP_(r5
zb#-8SyP8YS6rpr)QUC{Ax~Dj>-Bx=UY_wFhTNO8)=!8en_~^(>^V?wi;;SlqOXnBb
z&*TiE?qaXFuAAOBxQV~rKK+ArS^Q7w+PN4_qC=m%BJC`s)YS4c!mi`DzK4n(F*pJB
z90sB++&;yJW>>F4#T<tz+^~572lB7!dHYHH)%nA=>APy@?kb0{A1Oh$dKZM>YEZ(^
z0E4yR_FFJ~dj^RO_jVqe<q7qK#Y~FW-?Bu<O|;4js)cV(wLUrl`L=#v%0tC?3~A}t
zAL+lFz*!Q^z_X;({x-ageOE1$g3&PTosnEka?1yiHc))RhEMLF3w!3pPdCk_B2fwq
z2-|fV7kI24zYy#nfF$wyoXzARqy4^?Rf<kWBj7Z5pcTLEbqk+>;1fu?xyLO*`4N!a
z<LX)uHF+Fn)GL86PKYzY|1Z1#>2e4>eO#Pl8iw5&9LN>ZSEVa1*z_y6$bGZ!!-}%C
zkW+|9buPWnpa!+k;hO(01lSvH{Db@aW-SNv7u<<MP)RQn16ZEYZ6HbVthh2UjU#K!
zZ@TKqy3R>Igv+EzEX~IbajiS3%@_MJ;@<BpOOYCxOoZm>Jq~^DjV#J43ao!ol_@(R
zkNilF7~`1$yVA~VFe--()b0|2>TTEPCSoF!1<v6?E5@T8GV}cJkjKYi9Ktl#@IQXl
z0+32vg_%@+T5Oa`U}8}L7RInLmVl_@%>%pxTK8%@@wD2VD#su9u{SAPzxP_m4V`zG
zgw_XAq<mB264*xmNc{J6bq*4<!J>l61h5*AAu28Xz%PHfPeScw+jTSYuV}Rw`s#l_
z*^r_1yWBku5HfeIPO&s?=F+?Bl8T)`fq}pNJDE%<dOV$0IE^S!7WBBW9)ikBN`M*<
zqHEIdR_IA0gys(6if$jUKOuPSRQ>1vS?ZUI>a=<EZnFv!JPZO2tWOL5S^@*npce)I
zK)Dr-T4u#&_jM1f$M4OYo<XCQ*XBio7oM9(zN;p$>?r=TR@+XbTW<Q-AlF+eJ2iFU
z`0rsnW@dT*a6W*JP{5({GIpshZr{uowApoz^4PbZj5It@eQm%s_Dkp?%DuWryzgdA
z{?pALY4G(K5NHCrcT@tpKeyrIE&dGmTlQFQlUh%XJciZ2s8`oSBW}cm;*KFIR`*77
zD`Pj8R9@FScNT*OmjoTWuz~wB%7dBtD?<Uh1E5e!apiZc%<*KIK9`x9nX~f-bkF@j
zYa0qrD1V_7kcnyDeEbdJi@k(}LorBg&rl7-9W*cX%MKO_wtHEMIEOM@mSGrBc6%%X
z*bG=Ng0WBl+&<7-g1)K_)TL*`p>MwX8)`>)Fqvte>tYpLe+ee0eeGz`HA1SWJB&h3
z-IOlcBk(UfpZCx2ue(+Qfd_P^^i@<;)?QiPz6}NaXOP|_uhCexeB#P~5>In&uz~<$
zK6{J@|L-iL-h=D7cW>={UWs}Ss}5m=r73{U>UX}Oi+*eF8ESu!W8K#D-rr^sA_q5V
z{e^vc#NA-bPbt398{TFWaOPt?T;=Jr7z7tzo`FRor6gyqICyedtoxotO(&ZD=&T38
z+wNR4H8rks>CY9H>6OLOe8*zaY=1Z_FVja30-D6KF+8UEUi}(;nGmj%g&Pu*{;L)V
z34g!dYPbg$4gp|;y}Y~(L){=u`BeC>K|iA*GwB96oV?3LgFySAKY#fAJGPd3SJfS6
zB_w`YgG#u@B2Zc4F`V7)>w~=xB_=s{`_R8g->FS60dwh~Wq^l5olQdG*yIhsh#)_;
z2c!oO^^3<8-e(KsrNNe@%nF?b%{rHkZm84zEO)|sx>W&!0A}!QRp0;uN`MX^v^AK@
zM+d%O+1ch0L_tVOr(gXnE(qkS1JB0w6=-1=7ZmhmZf-8fuaW|J;SNwIm}#ZYi?XZ%
zHr9tuUbHM)GKU10?7?#G!!=OIkM0Fz%>{72U<^2lfoUJ>8g#anmjPG|{1nvsxxp<T
z?A@<kz1oMSy75|BQt`%7FdWfTdg1KQN&#AifkCT8Om13Q8lbOdtUaOSqxJ!8V?ciq
znpAb$6AqtS-jtVfiC+a2Hip(6liy*86Was~na*~^rK@`0{->Ly^P%9i;f}blKD4mD
zr>D>d2e2Hwg0!=>?avhS*<OZ6Ot|ej2l9SQh!b4~?t~9qcol!Sb+PgHOM}t-(zW7O
z+5)u@?<@m>DKswt9aFl)TdrKLyenAmBe1?mdaw;a1X90?$9=A@;1BiKtfUGU3QX9f
z-vL~e?O#=8I~M$L18rr)s;h!_gyO8IOaa@0XevQCDBWFevE~Go7R61?K0ZERL=QRs
z^UgfaDa(<&=ip4JGiw~&A?0xNUN{!+YyZa1$S7xG^0a0F2ur`ePu!gA$2L%1c%V}{
z{{H>@_V%;Vqkv2LoONboVgg5~F*7v;64j?02a=!LAC$E|_YDo%teW>XQNDK=;>*$h
z%h=!V<5*Z`Yb5MXfr9py0pLA^2*V1ah4pnOTif&cT4xwlMfv$Rfzs9sa-m_r^P36e
z25Z36E1Kr_*Y`|PQl<>Cpn@vQ#-=QM<JtE@>yCEJO@c<dNpAMR&o7v(e<wB1m?osb
zu1rZW2q}+nWEXni=O?8nkp2;ylY(L+OBD@;DIs;X?Z*qib_2ydwyNtF=P20=IqOVn
zuJ6@YTT5OXC9S?*661dVzOaD6;?BP9@6@1A2@dWjho4VoX4i<Tb8;-e`fS<E6#_Ff
z8f|C5^e=7g)B6v<Emc^v81f%v<e0&0^|m!~F4!5pH>qm4U{qo<*XYsP)m0#HXZq78
zkXwF!Byk6RKb@QcMDfh)?Bk`_62eqQ#Q{`ST&L8hMG7=Mj@Yh^$9yXR#Jm!0M}}UV
ziGibJ6X*&ucQ>=8R7YuPY4h^J{yvLNk$xs<Q{;mO(TI1UsRDP51nO$~>+!ZoXny|!
zyuWNH9h&qe)t2b}RSjVLK!d^c49^t&rv&Mgp`jtrQ<QDIarjt~(UNK=b^mgmU;pOy
zuaM(BeE1Mfdmnmy!LlPVJ{V(t)}pgHmH-8^bk6}?2To3_jJ=$k=I~*^_q*^<Z+u0C
z$Rl`YuRp8<Zi_1+Gqsp2eY$BR3YR>J`NjrB+!2LqaLsw#pNP0D;Gk?UAXH^EPyYc2
zQoR)JTh{yNmoHyJ^pv`K3AGGo<=qA5wRfc8BwFE0AU8jeY7u~E31MVy`-|tjbFQ)t
z=i6eAd_g8Roi_tkQOId>6WC{hA|fsyE2hA4=8GAfc%L*P(BG4FCo(>>BeKVp*f9@W
zYvN1)f>#}#X+7-1`3KNsVx~nsf+jgOL;-gVY8cbuXyHn?O*x?~hOy%yS!N}OYzaIo
z@4IguUtRYgwsCaCzroML;SajeATOD$`F?zyR?3^dZk@B4TIi2?i(h(nHt*OUo#W#9
z;F6LOh%Vb^XRr?gb*|^;!&$mM^b?YKRz}8i)S<QyZ-K7<KHVxc;?Wz$+kH+*3do{b
z8fnRlNsW9f8X8KJ#8hxM>Y)WyP{4fi<_PpVz(rzYco;eT;1*r5+e`65G*r&l-Y-9Y
z0k;d%onP9*^t-&JGjsz?v#((9VTslprR@h2%qGVlW?n}C=s)OLP!b-8S@)=h;qX~v
zkDsJ)vtUrcBJ>V~a^=`$i<ENbn5cr5CQkxOtqbXalv;f6Sg9&|bnAQm3M6t9z-MY5
z$+R#28e&fZS9J2p9&*ENTDo|*$?;XpRq>K|6gQ!d3RuzH16%lp9`uO9Yo~Xc+&OI6
zdtwy%JlABz*Bo0-=_g|LZ_Y#{+e!R@47H)B-ND>*YJOB%4sN&ae|Iro<UgnKvv!6w
z-omD(j_&*;tJFkgUi)J1#NCjwURK||IcAG#r?Q4HCSwKeFS9nfUJuCNflKgFLvJ=z
zFtm+X;5jbV-A=XuIYsK=e*tfrKjgwk<S#5|L_Y;1KB1XB=8NrmIr1ZX%^*iYhG7#c
zUh}WFbzU+*&tay($D%69U$D4&m%$EQBN(~QEX+r7OUM_+5XoePVbA@f6N9b#@O>Je
zRcsku)`l&<gyyuXMd4{_kr{jMsRR;#oE}!~0A8|&w47~hd8KEIEXpx5sTH&X0Hqs%
zuqfuYP;PuyCC<0vlBO|Fb$A?uDf*JxBaK*y0Xa?3jy?F%^z-47;b}4Xt8F_$>(T<e
z7G*JSj?m#a2M<m9%kZn_0CT)MOWFDfE_UesfG{&Of3UVQJ11Pr^miaL#`FZPvbAED
zr=y+Z)f!<sQ6MaQx9F?S%>_-96(}ZNPCZ&=f|)0n92F0=0$AdY6d-#VG6isMw;DIa
zT|;BeFHbYxwlB*?RKE4S=xDNKu^Nd@jkt^Dg{pry`jZDm+&}__R`xbOJNrfWvPl$V
z=UyE-(zsXSlSU}adjdCj(jXoL@oVhfQZLg>)_N;RDk5et=fXKS=;hNiy&jXqsrWvR
zvm5Q&58EHi>97Fxs2w{0Di0n0bIRph0kJEzv}OzGP--h7Ym4}JM>Vw%*<oKUKC(V(
zg2#9{MV`XB9((C^W2686msIH0U?CHp5`vwt$7$ugrA>cI?LwWMoH#w0EG#~QGV(G#
z!Rs82RH6X)Pz!ol?3RtGe3mCJ#jEEnaEMxmCJqRh<_1kS`*<v!;EVU&K;jGPI<$qo
z9fM4f*VVP?l9@=uX5i^rvSB`<F_L7HKQC@?XD4D269-8-z09@H|DbN$zLAlUPoJLl
zE-@hV12H~+syLExR%C9R<_E8{7!=hGN{NyeSu69EkspN+QP1SpSewH%;xPm7AZ3E^
z3GEe7BDQjI0TX~c3XKnXNaHsJf;KIU%-ARIS}eFRrW&tvW%~?vtAS&3NA|mcr(a|o
z*Zny&7a**dn0%5sIeGzYWMjI5;$l(zcl2kTR8{dw>ITJQq90lm3)=i{CC0`jsfddO
zJ47Efi3#>N2iw6O9jJR?d!`MGHaY1HzR3_zv0Jj5e)H;V@Sh>%a&9d-bV={r5e5BK
z0B(}_NoF>I>KCxRj0eD*;!1eE`nkKsTX52<G$aY2#w@dCk?D{|0434{+EqZTjRsT)
zE-i2fupcdmh>017=#H&xEaz|xU(by1^Sj5{>VAHGzdSFG;oT<4O!uKF>b~JYiv6SV
zbp6Sj)%j@@F}f^0CESSC{gtP50ecLBKq@4HbL2SJEN*i_ybq0RY5K0W(8F$I%3h@k
z{zsiqu<supG9i+A0#!yw<A>uKw+!Hw?4A1v0jT25y(#4*-gS54O+>)z22ax~O>n6w
zX!YJYhyARThUw8{=^v}(9}e|oIw7eW9v+7FIS}Sq*J%JLgO0VEU)&9|D2Ox0hL??K
zh?*SLrrv~K_!!`=Yb&m@Se9Jr!^*;fejUkOsf?^0YG0314qsc@H&;e-%0o^#Z^2_+
zmnP<hwVBTA;HWg8xw>pQF4U9d6<J9{{LG5W^k|Qr+US#0QpqNhnsJKw#kmRSJwr0w
z-;kG{p5D>|Oj&n;Scd98D~c`<Z34<H4}KLFunl!Da9aS&A|SEpM1|#Bp)@#>KD^A!
zdPXvZ4YCs@e6;z3KTuweJ|L#(=;~@}&Ze$|F~VL^ax(b`J@B$5;sENie!K;-c%J}b
z?51-;l)w2N10uNq3{8QXK5q$WNZTw>oLDu|u%iQT-jbbZ9IqfAguT;zA5e)6(4w`$
zL4YFITW%PR=w)K)!Rix9D`B5}G1p_ck7V3UX_<Zyr6K90^j!3txVzR`cZd;+n))FO
zZ@`v3N2D6u2^Efv|G{PP5<P!?Qlmr9f{R2ZTdfXpl<@LwdwN5^SGC{Y1<C#SSYVKr
z{U%^U+*+X*tb2RY`w7ZM7h+|DOAzgKwZg!9n#rx5z0+zrqQ2lt|C^1ynw)NL0c;bM
z{=?_+vjhSOa2O379lwEwIv8YczZLy#fA?;p_g!E&bpkb1Z)3rp6tTJw2!+Q_0&J>g
z<@K<+pTG9OQ+_UeVevf}Vd3H`zcVuTXG#JLtoljZ<DQ~4Oa-%DwUTn8P<o*P0s@(t
znNTzD0@qt`{SbJpAXj{E_H|*cNa;1y5Ty!rGVaIFPzLN>?E5LO$4;u|B|hDqFbrf>
z^ABOixXwVV;t`iax`W9mYv-EXu3X?lpPbj1DVc}8fGUCwz)n*FckZl0N<RKbjJmUZ
zEFSJlwmX5?EP^yD@zrpF$3{>&>9a&yq?{YRuny^>G(5IE1T@rgZbO=K>QM#ixmUf`
z{r=6rhkrm4xYX;C@@=unk>ugHDOyIkP6NMd<oih%KAH=*Bm~w&UHpw@nf@v2@`@c)
z$0zC0C<TYBi2Tp*+ooTO92H=<0oM+<+2c(AeedzgHfe+|I4~rhWFetyXa=J&0!m(b
zL}Nc1Jzyg72@uug+=R>E#%dZ5bv!mYPM?vGxBG`4k#2318N^hs1bh4zpM-*pi013n
zcBm3rO!C?-SV@4Zi=^mP)MMb~rJ$r^5s)@&ah(m;QFEwSezXcnPQ4iRGEx7KFzcxE
zo4fL%bEX+`+gl1?4ET)rM^BzS0r}?#@hUbB4pX^k9P+h1MPzM1dTx+j;{EqM>g8M=
z@kuaYfmQMC)3&x|bk|yVy`REMBji5616X0C;kTATA&D)2%KHoSd(?YOYk?8b(ZNeV
zsfr7`&<l`9AG*@*w5`!Y_3~33Moq`qn3}j6N|1S5Aa{s?h2u93^)V-8Oe^3}hSDUb
z5um_GVsO|~;l6=c%@CSj28g>^RS$m5+=94zbmSL`Y50tb5KsI0;HgW8uZxVk))YPw
zk+7)fy^j@C4Rd|*Oi)bi`?U;6qwcR-D#aI6@|cNSN#TKXxS_Vnc(iT$4tJRc?eAVh
zo&?rR5~u==Kx@Gggd)J%z75Mj(ce2Ck8yI}%$l&3nD}1C;vQF@^9#Hdn3SSRApZvB
z28f(Mba;163AO0Y?+yG0TMy}sd}kGO9&R$88;`=gy8@g+7Og5E^$hhiL9Va|o{aj-
zlnSuE?K397g*fMVd|?wQwE#uF+5_GQkai#=`0z_0i{bV!onR=k$)$G+)hk7Ar0U&!
z8cKajhMhEml$3P-qlOyDb3@oZje{-j2}loLGbJRaiPY7RsVN_*vZ3bEB7qe#{PO1F
zcn0R!0nWqUu_IGy*bK=ojpo|=_NL0QsES)iyR1()qs6iPq}=VxSFSX`NdkToc+cEO
z<6nRoj8Ss!<=l??&9iVVjVXu2L*aP{q_s^4?K0QT&jZn7V|`tPuLS6PSttgCcwal=
zr{_Za^MV%Kyh+Ri1Oxzh0V0tpai0^O#_Vd$fqNWV(v{3YK^wmW039P+TMm76TncCf
ztB%L{heb|dPcJ87+O+_x;O>xw;O}2=t>nsIh9&Vs(;Lu9YBP0ezD!decoHOg6ZTwe
zuez%0*-x0I-QC^G*ccIq84{RW6y>XF&`Ow@oqfxtV`-UNb{+NW%lrsHY3Um`6&-EZ
zzFAKyR^^PfO|P_q6Z(gc|IRaX4#QkJo>IqlpI5^XyK<;mar^Jpp!$madpq4Tb!3Hh
zVuNmcW=>!H6c{Kg;_s7TJwsIqFJ9c-+R{F~HCPde$LR_h#%WJOPCuf7A#OO1%FU$h
zmZY|MD8Qh2<pEr43XyucNUyK&tRPnqh|tv31bbdi(rLJWSrKGIO3>0Hf8QF|t(XqJ
zH(s_gQs6LBn4(?6giy3rZQ7~R-PMU!d9$v-Ln!(ACk0#zJA8^}i9mhcb9i@<h@I;Q
z-R<jBIo-k)Zd%-zi4$W)sbXL?fdbgp2(xTlQXU`(fA?dBc2-z^45~R%HkC+3e2|Z{
zEHSGj%BuLPk)|Qa*Fswt30HZbI+(<2uwG|k%GZzvi1R7t*^#Nl+#9cR6qIUT<^g{j
z(Zd?MGa$-KRsYx*mqg@foj#pO95r35A3IS+R%(6`D)^JI(`2=0>(@q)RluwCm|Mu)
z<Akx@vahVoCU(|td3s0#9V~{^Wc2Jo1}}|Y3Wouoq&7k4uv?AJ{BI%QZuC2Fvw*S=
z^4&grdZQ%>Oc^51K)GTk--qK`;TA`ClHv91r#b8#okJ^5j^u}r9edlEtp348_v^^?
zN2)>1Zp7yGjM$LHbMcEKtHn71K^euBl`e3#W9OfGA(`y{1^0rbw%$b|e77#bD1xNd
z9@U5B67nnF4dF{(&!Pz_r5`Ea+`X5^NW!s=H_>;1u5JT_kPoEnJA&~bU4kLW%vv^<
z(;5-z1onJk<4WATT8?kqi>W+Aqvsp~ysk}bC{R%iSI;R#MrG0zll$799!T_Q!}F89
z#Cp%-xx$Yq(}VEgxb$x<3X<C1p&>tONi~$;rO|NQ6Ml@b3<O%~9Q+S@ma1xiYCskg
zhcR8kvL&b~7Z<hRFG7pSI=-OvdKSY{I=d3~Gd|DWFS^Yh_V#kN$!~F()|+{5A_`{|
zU=u)KGnm1M7ZOo8$HK^!*EzaQ$_NTdNEa67L+qhi{Chr7CEDxR9wD2_P<~*2PQ%?Z
z*tr%5C#Ut*)zvRwx?MLRd6AHSN=LkG6PuVUSq|U7YyIix4+$%`KphL@@xv&%Qc>J;
z*k_5l-Jybxul^?L7QYzSEr6v$gNxvoZ{P5Hft^Z-HNDFV0h{tj)beG-i%cXO47!;%
zSeGJHG_a2jh>c8oGw4#;PQTcTVLJNricsPX^nn!u7+&Jeg@lC-y7Kez5RGDjg7g-!
zp|_`fNKn$=LOG6-f^z*U%%4J%;{wQ5`=k5d?z(JAzC)!UsR(URp&m^=w$4Rs^Jn`;
z)JBpbrZ{33L&{oOj)7HaBX&{x;)cRH<rivf--pXfpWD6K&+N@Za?VU}Platj@<XO1
zO~;sVe<;E$GW!wFD50cwDdDP&Td{WW%aQo-<k1<sM}DqgL<!mO=I$;tbY8pR1q3BQ
z5?I#92aU@om{P=zh!9G}Fl)E=pkuf^fZhIhqP)9Pn*c1cpAB9h8rk-dBdQ@w9&nwu
z5|oH)+j*AYJ}8aw@bFNTqO!m85eUk_n01|j=~!|+-bj>9B8Efqgpp$SW@Ptb(0MQm
zI{MjqVvrh^Pl3u@QA{P`^<n>TMN&J9u=Nu9h@sl6x!={wrKrbMONUhVGIi+fp6GP&
z{Co0*DCt$%&cD|S{3xKYROoT@8pVZ#bhicK#A6~CEj4QtTA8c8Ttkg+uPfXa$Xs$%
zm)E>?&%oSeTqv8+Mp8uNCBCB2<#hn+zBHY|1_Iz@X-1psrG<5$Cs^F_2HX~St0R4V
zma8#-(tvvHQhZulTLYJcZ^QqGhuo+o)9<fVNvd<B=3vCPfK<Kk%a<?XU~>Z0RPbP1
zhWXmUn_!!9;YDmTR1S$QU~g|}2~7Y*+YCU&aTwtt=xqJ==B6tvDQWC$uN@$b-)(*e
zpexD5V;se|lxB_Yg}W;^qT{iMJRlMCWx#g@Se}py>npn!CkDYs_QipF^@K3Q2iFpn
z*$svBf<wUA!ns=o&<kfxQ9BX=peGS$o|;5`P0bd-;R{!!2&DBTPTJZbuNq_qynVdN
z;`7K)|H4J+d;pmpjr_=sP;5SD=gdt(Ign_?Em9R+CTMXP=AV}2__JZ#CU$oKcnG6c
zMCm!VA$J4PdmG)E8XywGXqm1qGSm(LoBK|CYf?|XiE70Y-kEeM+$yRx;Szqe?#fS3
z|JY5`m8t|^&dd`E5125c7lzqz8aVxWD>^XQ0_dfU(_R!`O9M<>rlXQd?*NS(u8hPr
z+90n*K!0K7`p{Nh63^G2Gb6i0jH6t&dX{ppO@h~3XgADr!*v|I1eXF@F{wa4<Rv7V
zjsxhiUeC}Wx@@^RvXDDQNjXoet(miwQ=iRz9b5ksm64J0`{!qiho$s3XY7so?~{RV
z58<#ZRJzlE*gGJ<t*xD1J97sba?B(s$8R4V4Ss@`4D3;}s|XV~aqNXNGp3NLbpO{1
zjO`AAunoXJ;D(|C0@g=ckQf5t)oOU{)2pibUx&umy$hQpiAdH~S0`XL-6!>Y4h-(_
zE`>5MFj$}Gp(~e0>@%u*AmF`vHDB3KSfCo-d95)J_5s5>Fbp9kn3{<=(o=IFSJ}iw
zl7S(wgA2^7&V5T9Y6Ya7)9f}FowKAr#{Y3a2pW5#OE5Ktxp{5X^XJ^`++53$i30d~
zL%Zf_5^yXIHPLFCngPYqY5<h>^=XZEzmUhV_jUcL0~h;xNu^SmacJI^stlZ<w_rbQ
zw6X-p1_l(%G=+*wO4!}s0)3W-n%ca1FCZ%w4k2v#G>?G4zYGP8tt>6)zkgRJ1uLXV
z=m-6*ouAcWvYjFJ69HT5e~k%A$tDRy*NY=;#7-N1Bg6XIlG>7;$0Ouw6#BN_X8K+(
zK-y?yF57G`8!u8ZtslLbnvigA4H9K?^92|UAC>o=hpQ{)mF);PNg7m{-)MF}+S}52
z2pch7jpKLJM-|^V${#n^rT#8UREWIDrxLUSIRCll0T9J6LHE@}HUyiExSo053XNaE
zH9-H}!I=sJ?VotGd|YhoB{Ds=HwTK!b9sU`$9R_=v$H<H-SJsrv>%77H{AO?_HAsS
z8_a+f6adA{S~vw`ygl&TA@hE&VJckWBXB2(MHH-zN=;Hq0UQYwRh16b#|{>)83rm*
z)$Qf>1YzkV6u7?7!r$PoZm?9wGNZ@ksKBJu1uViM<u<?zmxa#mypZFE@X1TA$}q}=
zJDS69bx!^jh#kO*G2IYLGi85Y?{f+5%ys<Tv<7&rPwt2T-+-vX--S-n(%Q%G-Q1kH
zvN9qLH{cc28?9hO1}E%skL8+`v=8u_zQWB5?+Gb90=qU|fsKuAlh2(nI{>moc)=Jl
z2of>^0tvD^*qWuu%Nb^Fy#2*H<NMHFVreIvtitz;<&8|MnfQJAk4M*FFJG7_*|voZ
z0~!!pFmpU_3`=|T_iwu03ef1=^;cmo-?HsY0vIzn+uM5|{&sfN0SmQ~+%j4OPxgCT
z46wR{Ml~HpI;{iT7QeZ%kWgCKGb7N@=hfm131yfcX=}^1Xbk|Rh|;58K>-1fq^N=c
z&>zr>h>neAi<l49P^6rk1`Y|#6O(A3`F4#iIwmF?THfHjnAc)F`~hg4l31E~yk?|V
zO~+yo?X6q4u(MxRn6tdWLVi;LZ2CAYDC~s+*~^`S`4Jd5>ndPlVF4pS0#mrM8%G1l
z_)Hu4UFT(8bx7SUHpKS26{k5I=~zZpBul3VDRI4_-uq<7#Zmo|(={$$=XSCl;PJ3v
z(3>{${gLGXE4JbWa0lQGzu{s4aN@>|8(1ue7&oAFzEc}!_ih4PK}NijkJ!E_>w{D1
zSWE^~4p%8B`=4j@=*uG=&mdm#(XoBZUiBb~BKVk;lhX=k6Pmn<lY9UH&b+#pV=a1T
z!U;LNoi})aQi|BII_ra8==pRGmcaTjycf5o{mufw2{aZviS@jvo6>H*i$VT@OXYT9
z(VQOqv3w5p%Be)ss$?GRnnuH}m7i`&Jek?1`{*78p#R6q2*C=`v5T)q=pavs0pd9~
z5082=pUKM#i6zX;Lhdhh2w>mU&k?;$q#mp`%B)NOf!5s{qJ+Qr0az@wW{SnZZyttr
zpO+MXIT4xG=&#q|q3d3}kVhJnNKZh?<UgWF?ZhUWY-O%M&p}2yfdx^qEsY)ze%ldi
z>m6VKSh;CK8%C}T05S740I2i(P&L017eBsSD5h|%#eQ4C4eqa~?eY|-l@xy@Okxyx
z9@TYqb-^xWM1Y>h@^it6w*;ePP_~ujPUf7TAq~g61VgJIFg-vDN_2~nFL3e1*>9F-
z1G}Tu_oDFW)_fN90M9NT(6+8CZ&z}`-mo*Qb_pGO_4?@OD3pVdn27JJfYI!@SHmZ}
zySqorlt4}d(<1^YqlOm#P?PY3o(8#}7@+k3FI-v?@1^qs;(enM&fT9TJ?<|0!5B!-
z94D(8+~agnHX{1k^Rj0k>BBbPKsMvswyRzwhSv?sYyw1!Y{7z;4ilB;)Qj-T7xyo@
z*mENO{{8zPKntLE3SQF&UA6hA@Wqe8{R{Bmh1@g?jis-ZS0bQf{3X8kRC438*%IGr
z#=jqNtxlGohK19nDGoA1;CM_VFkQd?23n-ouHnUFOy_eq%_y?2y!>&An|IO>O3B?v
zg+Z~J_r=nT*-M?}B^8-!yJu?1J_-yejQ2@3Mg5&AQ)w+x!xQg2hdA$mcM?k6wHTy0
zH9h?jCObUu@1CO2KGQpyUXjPkob%yX)#+9>k8>LA{Q5b|icRO~^%M$4%}%kn?t5#;
z)}_9wu<bIWu3yIvC2FMCy39%deMDKg7dj21jvwr|m;1p}vk&caPzL>7=YntouMYQ6
zf}m#m=iY;aUgz#kh#{yz{le}qw1j@RR<mz&)Ew32ufVY^0#GtB&>KPkXp+QtLB4Ya
zzklU%U;F5``af{<^Om?38}}Wtzg@=E?k6vAKUs#&1;g-~aKAwEQRr(2Een`=>BF(!
zrxJNu(_nc6P##gyM;2S&dz%c^v8r?scEMvX4rA-s!Uqt?YD*lGSBG6F^oam#oJ+kh
z7>55sN4%pQ%3NqKv~_hYg9J^=3AuY5kcS^FYKku0pNZ-|e>VfX_@=m}L0AU(2UBEN
zssnVdb?d|O@lo8DRj6~|^150OE^f`{WGD9LB@WyQ*3LJ+q}XEh8oq!1n$)__>K;dc
z8ro@5i<UXuc7ylHtPTU^bB5qXd;9hh)E(fFCn+HTY5k{x0k!d~F`TZOLtIdn<-d>}
zqLQ@Y>$JW`W>`i)`IHF-{c#|a;^X7PtIEV^u<%c85K>T9tih53ZH0g;sudDDs8#yn
zx}dp%+bZl{U3c&~SJ1{QB)s7^TZL`ISMn^jxCXoCG>h9!kH=M7zn-$PP^*^1^}3#e
zYrd(~4J8~bEG&($U&9FAqI9z@=L#)R#q+zJH#z4kv)K7Mfhw5vp0D5l1@aw`WiG)I
zc6M~Qwy;D0?*8$aCwaSl5HCSI$JOi%Ed8`uxYx5y!YJ>3JT7Gjuv=B)-DtRvd;)Dx
z-(t_2v*)X~mUK{QP}H3Tv|7NrEBxB+p0tdNGAGlhj4!z=-20!!r#T)9Yj&X}wY@@j
zt~Mmd(c6wJX1wJ&L`nJnt|$bIs<9Jo*G<gFNDQ&MYjp>V?+ET|!ZZ@h+r0!_l89R9
z@uR>{;N3L7DIC1;RC05l(Pauxl1g*-Rnp~iK7a0dH<8VRQDRAC8rL^Pfp{l!4nE3r
za~@4gn+5wzP=@QuGH^xCga9j`?6<393Wn?lH1=Nf%2iI`5m2qLb;F%h#H6s_2>rAf
zA!g=fSld3di#z|6^;KF*l>iKEC~9CUK|-Zg{O9X?`SX8nm!Xk%msP(_kJ@6armRmt
zXt>2u73XeFP07|w?FhOdzqHf==oxwV)1bGIAO?V`z$}?~NhwMLI>z3C{A@Vzm_pF?
zO6D?5(WC>eO@tr@RY0@~5a#?P-gC(5h+08z3j8$Z269P(XS^sOLBX+!3FuILhoLzt
zB=jKBjIq7)NcUh2;8;Moqaeo~ez<es6vbs_@nl^gBfD*H|8;dW3}y2V$fROe%*|q#
z><k(4i#TL+&g3K+KO(bB$M{LN!nhzTKIS&mWT6;nWxSJ$j*Nu;8iZlg#!mInw%epx
zcMdWE(-rG4h(uwvm43iRf~pSMQh+@IQoj$ZCn>ed#KdiG$O}bdR00RZ`;C6Q*sqf5
z(%8Zr-E&L-{{sk!ic*g1+i;mfi%*=Lvo9abY!KVI81(j2`qQU(^$UP>14kz6D)c-A
z!NnE}`x&Jrw&Q-JC}4Ao!^j4LSMhA4hjJ(e;6DTcR}J`)ZD_O|{s#G!6m1lN$h}k$
z88M0nK1IJRayyt5^8zrR2MK9Uy`Fs_Z1s@2`NveUh*N$P+Pc`X{_~>~e*=RQ$a5Dg
z0UR&z`!oB1a~s&(11s2OYh})C{h25NJT-98+{+k_+use61q>z_i@`t?#JMg$*j`;-
z%?bw0&eOX9XF0yhZEtW}f<p&RyH$)?&!DX+W_0;)kio@E&)cxyb8FwPBq&nNj<u_|
z7wAM2)6>SnIVMkBU8`xb7+!3YhQIK7)=WtGCP1@SVdxPhWx1ODA3&e?pzZ{^Ssi%i
zZcadr4ddR*3JN8|fL7><Abj#-*J5s`oLIWGuWSJMZW;)aL64yh#8#`<+TnCj5J#Y*
z2g$x0wT96j`9I}bslY~wTQR;4md>k8Wr-%yY#5YH4(5DdwkO`kR2EzTK21;}4W|o+
zp}-Mc{^m_SZf-)yJh;8#CYM9H5CVlLbe6((N;4@5QV@n0mSuZ4n1MmjXNwfrx=(J-
zrhrVT0!!%oaOE#I0IMHpglxTlV?ZfS)GL1Rt;`Yu;HlQp;e%e!3J6yZXP=SbcpVq&
zG}`81VC)#4eP`t@4IIUJY_psM*|O6hJHxR0nuvnZX%5-;3ZiI#Sib`_W~2j-#1dNo
zK9BKucm03d!sWUwc8<<cL<+>SIYY5wbZt;6lqE3x52q**BVPFu#voz|PQNMenSeVD
zR_xADaR7z}H6HUD;ORgwMXx@rQvkvCIX~;!;NM`>D>7Rfqd4lD&+ar|0&um{7Aco!
z{K~~l7pV0r1+_qS(3(OGwRi^`5cr$HrMK?|#>b(NC6M6)&ZoKvSmlBRoH||)ei&u@
z4{Kd%$YN*e6z4VQ`Y!17s?^x!)(Unz?pbPS#c`el^`UJ&_;=QMUq5DVBa0$P`FT<Q
zU0+|H=)L3F6($CT_Rdau)j%7Vkx=gZTHF4!691j%<s7Sc{<WNbp=z(R)-vo>2e1>=
z?Ah0S4U!*00$D4b7T}2UD_l#4o0plLeBLs)8~}nEev;{k4nfnsBN8HpdaFo<UlKUb
zuKj-)`wFP4)~;Ox*o1Cc6t*--NrM|{6p&EqZYdRzl8}@R>6Vm`knZjV36W5c4v~-$
z0VVF-e*b^R9pnCYjC;lz<9N<tFV<e`edjypGoR;~fctL2Y<qy~`-fDek!;C8r2ZJB
zO4j5V4v3Mn-OHB;dwV{<FxOX82#C3B9Efxa*un=^$J){}RirCs=#E#Usg%x3mH3zx
zbaiz#oJKtDjnZ`t7bP4h4Tn<ocrV(YL*K%KAnF0u_C(neSv0fma6<lAu6nVzNtmzK
zUjS89xABCY-|8PI-O<<ZS;t-6^NFqLl^lV-JHW?$HZy2jx&8mF73LRbF?Xe9c&|}#
z%rNou$67s$kB%l`ljTh&u@Cg^qL<@ygVfw6Gx6WrYuRsnu*5<wsQPgRy5@z29C@y3
zcVEi?_k*f67uh9HZWD$ll?OA>RsT1~OIX=GAO=WYu6B=SqI_xd94;ZiB?D5!RU?4!
z7AZ)g;Gm2D3C=@5x3@jlKQRHs@avc3^8AP{JnZ+h|9RLbQkfJ{<M9`Z^yI0e?Vc^)
z0myUSoO+E~=0J3HiTt2o?}3eN9zY}FG6HbSI*%kYtYBlQzZAr5S-j&#9U72C>F#FD
z1&=-=OIs-1M+CK4=%e$>F<S`~!(bCwRabWm)v}r31!<)HYu}&FS<HNS&^?9I?uO^l
zjxiiw5Y)wCd&01I9c^t4^%tZku(<^J`QHQwFEAr&aDYK<ID=(*3xAkf?G@?A9ym(N
zpvwrboXnniAsp|mlAirkixhp=9AN3+?tqK_{WbO&hT)oFU|`BSHiK9*t7M3G$A@*K
zTV$e#I)0e61B1;DHG72LFRcG){afh!($W&0Y_B9IsAoBSYGr<$wZPSGbr>5PD-Ac=
zEy*PPq`vEz2N(sLWm5lYh9-%=;8EWV&vzHL?(g0(*T9F@nb%*}<&hBS3ywDw#1HN)
zh?ZRoC2-<(Umd*R;o)KH<cdQW>%KF8aT1OG7RpJbccSz!0V4QB^1x^d6f2;KTh6G2
zR-F3XW=japWo0~=43WeAm7(i`Qxv4Ez#oPjj&02kwl#DylT=ZIi+BgXOF`s>V%~zW
zxo#3j<k-@tDR0oYAMLzTqz5X)cj!I0wz;=VJDyD)pF-@3Spb~lpKTN=rNA>&3yU6B
zBsF0cgKsJq0cIO7q@-krjRoaWH1DYH&J^GxZr{EQkQ*$HONcpF{r6EgB|6Z*xr<uX
z(W^3;w0yFoln40uLO=em_hapok4XsBo8GrkBA{;r<rgIL(6#4{SDCK1hN1y1w!qxv
z62jrz5DB8M$qagL;1Jlt7s+KS@yKrHsXg6mqnlKocM0La$_rIRX*ieYe(<!>Udjjz
z>^UL)G&V*B)(!5g;HiX3%|+4*5*h53@*)1&M}y$h0~_<qH<y?B-_b{1Wd*jFS+huc
zTkt;=QsF-!WLOS{&8DR@ZwIfp-~^6dc_7Qxd2~wd(acFAP281#3$YN%z1a-KL%it1
zJXyp3xHlfz)cne!SC;DWmVwbRSo%dub~yc*UIW8V+>z%9qZ|8^A^}Yo)*39AQxUVn
z_T}4r;y>b`-Jvsi(BO|UC8MBl2IeeD_sID8K_(_MZEdT3vI=l|uTdvHhOidic+VoF
zn2;<pRc*c~P=n}2Q9HZw%JsObh58Jr%BrgIPoK(5OTCxin0lfo0O9=QoC@WOorjDj
z*cb*&3;|QUAdtv5^mx9?xsn8Xo2jp6@G`gO3m_I`-B14reEClLHK1e#SEv^o$^0T2
z#8nT&WD9A5Pck2FQK{<a%}%qG?auSgJ4R5E*p$)ZGyARGFiKGa6b-ae2HUa^VuBp-
z94d&|dnMzA{sX*Fuu2HC8p09IkF}T|_SPDsCdCk5fXI}hFvvCWX7Q3x9w+M8Lr~99
zrEz*{GTw)u2NL5b^`n}KUP*snBRH#J(>@0avOyC_(z+d>g*n3GI7S>qj=5CI*1gNm
zhj-aberL%FN99&#Im+uU>;s3r>lsuT(D{boAqa%)^O8tk%3(>MEAAEoAqR7lf<lQP
zBC5<P``kE3Mvq`8L$ls&?^T4;B>obdxVIjqD6Lrdpwy@l*z0nRhC{dci2r@cl2p#3
z*WE*|byGINiT-h`n2wn~UopP!cW~`ee)K3DnA_`^osI(0yjTcYvp-E5bFay$O@~xy
zcOB~ap$e7!hiOPO9|}J!-VONZ*<uN9qs~8P!!Yug&`ha3hnw2Lph$-X)qkLqA76Li
z_4T&lcXf6Q+m%Tn+4LYG03bxG#j`0Q;_<wfKL)?K2A#;7)yl6y|2`|}-~}g&bHw2h
z2E6$deU$<K(oZ@Bc`=;S(CKM@yK4$FwVxRq+d)|lUO2qk4(p$C0PbnU1KN<{A8-*&
zRZx56=mBH^wATDj$rSc6;~H}i=C>W1nZK$B=TK-i4yOpkq^A#O3oFRI;%B6zGls7M
zfSJDz4*`lSh_MLuG&gU9ipRc`SN<!y>MoTGEn*-_ivxoWLewaO8Vj--*AvaS#-QZn
z<kzn+qA@pz)MRfm9xI{2OeGgTJIVr-hM1U`@pOQz1tt25|H~$xJLGj@i2dybqP6~J
z$17!-JScuaLFzj;sM=-lX#|(Piorp}#Z9HUROtBZqQP^<eh-?W&?Bz59L>Z4@C^65
zw7#Gk&oETeg;ompDMzT|niE~WLCL9bpFZr4RlE??lla9LL$V)w+3<!&Mi{~;vSbPK
zeEL%&;|^YAkk~%9y%XRIUaxsOe|jZBS2}r$(ob^hKw|F5ROAsOiRAb%Q0QGj<fKMf
zc7>2%khO0#?Td!onI+lxi73!=o_Y0y9_@Is=}IQmt(!;51OXrawLd{K1=1kBRCaHN
ziv_(ca2kR=$t$6?$Dk7eT=WV84pMgj#OCLX{s}mGx<Za78ybd!$b*@~3E|1bSy?|u
z<0@9x)Oa2Ii2H>Hh>+qOvXUmQzwNV>oHbRw6*tHUQH7tkw(pL4QmrvSUT3*Fm;_FO
znsMStB=%6(+Sv6v=+~TrP0vay^mQlml(V4Au=g{Fn8DmL^4+_IIj@1DH82J1$Q{fE
z7qKFGY*ru#Jn)E3PA2$W*h=s|efsgOcgMAmPeRL18l#^`Eb~$)r6v3knSBKd>9jEc
z27|6%C%dwMh_$Ns34!I;!R{_yj}pT;IKw#$ShAfREk(SclJXa1WnF)gK0ahrzhd(u
z!GxJuC*NBttR|*lZu^(ukKDr&hc?o8u5qcU`u>C=UFtLCSU&<G6#!@AN~nrXO4#=8
zyvb~dcqPt6;Q$1^ZNJDqtOPiL>TMUKk|^>uOJ;$6+kavNEn|p0H41%`lL;lXuP^V7
z%Z|LlCQzQozH$C)%}W!y^@D@UAL0t)jq{odXt6MrQ=v9ro_oCK2FPR~t!Hw_I<zp)
z$Hu_{qQ2oQb^@pVT>!4TuR<nualALeLnD=Iv^zuns3jO0<Ra9Mm$tI|<=MWFy{Bb<
zLy(3WCYiUdx!?;fV=j^}f7AF~dH(!4-`F4kJ_BzAdLxUir-2h!y;r;sRJw%V2Ozn4
z)?)dm@1V2A#rSOMW1ie;rjjNWZhvyXZIWQ%*4~qnn!ipN44yFz>tKa+^RA<kK}((?
z+ttEvt>Nd73a*BoX0ANYQ@h^c{@qPn;D*79(a{u_P8%-7w0M;I9zHZ-(T?AVP~J?=
z;JpJ@S=pTdzf-?=e0~q>5M%xZYxR%nCv(E<nAb)zfddXI7+Z^+<ehFmdJX;n-w_-u
zkUo4&zO&+*Qd8p$>Z!g2Ors`SF6O}>5ZZ#8<Gl!S=<qgPp~t!S=X=O_hMk9$zn-oc
zqIj%c&N#0FXaq*b^I{mW6?6^cByUnvBX7%>k8^pKgS!-|eC6{=&IiE3xnRkuRcds0
zay(xF^txE*e=!U<O`3v<sq5dZz^X|1_<du^&NR3any}EGYH(ct1V+H%mbwG!?(#C=
z(U6=xgp>-bVdisrczEO&jowLoI=Zq&<`GFRgVd;g!=^nGyIzU_vjFgo)GQLk4G9SW
zkF3O)7}6Nm+PBHcR5$K`bSn5)`F&m_?uw9_{Nog~v6-2=r?)C5swP6OI?O0Nu3)<h
z+g_>u8QYefsX3)xA1KkMzI@?v7!Lx*q?NUGQBjeq>HwebGkd`JKxzpXpE(&qNGrTt
zp1ESh)ld@2CI`M(d=^7~IQlOa`&iWfg`!b+4m%!VE}Am>w%TVa1%S=i`gTWtIaYLH
zLfe-%(!Uy}U(A+P9*zPg9)OFKf>s#DS00OoaWjj(QkW3SwXxm9m3vI;sTN?f22xU<
zpP;8m)uM$jKOyVy<44XA920`Shc=no{MnFQRP@liIuF!=@{1Wkni9_Zd^DkO60ybk
zZd``WfwFV|{7o<sJHjg=FTclPUy6c<24LsW-@hOB_~7S)ecLzr7fs}-a_Jy%>R$o~
z@2OjwQ;0zrDIG#QGsx3Da2xkY139d#)8ErC00shuWTfv25Cm!5dfXvPQgTH(x+wsC
z$$5T%AcQ0g6`PAu0%92*I=};A<>T820ZtNlCIfz&0^$LXI)fk18=Rk_mOhv}jrebD
z&a49o%>@wWl9Q}of<u8?%zX+76wkKB0p5@W?)hnOXimCb1m;7*zBi<!>#G4a+gIeV
z-(g_9#>ZFhUzgM?$t1@}0?I%Tt-&J1!or&O{FSl@dTI*&^ye0@|MOXd;fR^9cx`W3
zdkBhy3h@-r5zFh8@zD)INR#B`UOT1d=H{}AlL9I9Ut4dt925zVP8gA+wKzD!2i;G3
z`3h1dYZ43uOM|@=q?NK89nzSZaUVx;oZn?Tzz-)hUf`sQNZBsAiC@yA@%~wqv2;`W
zCxZY+e+`$f4JKW&vk-cq3Y?sz_(vay!t;&;q!^I+ftLGHlx@Jqq9-zI$G?QqHOw|G
zMJKYDXj+huk4o67u-_&l0Fxg%yqFrzw{#xDyf&JFHVy+PufS0)%B2rDh7qKf$<3x=
z%L8Py1To{|f1Wz`-%mW0H6v@3keh1xdyWz;Xs&(ZO<L3V{e0+QOxo{(7SX?F0)5ik
zPk#<K?jFu04-5!NjvG|yCzRPHBu%t@EJfj!)J-PjNe7T0TzYaLCU2+G<8e4aSI7+w
zoJ*9w#P(oQnz&vXXtTZk6Z51sUi+TE_$ktwdAC{6MzRYB=88&6f4+`=mA;$g37xVs
z52XUP)F$N<&CRCJDRwpuHHl3utNZuBQm)g7Hb_3oz>2nZ;IMyY@Z3{@G4`7o#qi&q
zotad~bT+>^tU3<?bNqPYK`pn``zQscv$5Z!=%mI!k@3G;*3&{nx9GxRl%0tY<ehEV
zO$v!wO$5Jh6%rv<h9n6d8zYm*otcC>+oaf*j4wD0`$1iBr-K<uf{5#G^DA;{7x`^0
z=|{ep;VyKvSyeeR8$r~0PVpkUv#p)&v|V{ME&rA?{Onl$U18yAFgVxMeVi=#5samB
zzDuCgh!9H?r*s_cR0n>gQm9xi1O)=$q4(F_^wV?%DPo_k?`e|lH2plioziGXaPZU7
zOz9+EGd8t6B?t%dacsuZt*x|-jDBdXdZ8oxW7}2BFTrJ)n*qmwU&ABP-^pIOvC$!~
zb%`c?#W#L7sJfVuX=Y}Yla;mf;|E(5?&^<~6>Q|&q$Fa5!X3nu)vF}e7CcgAwCIri
zQN9V8{=rqxZ}#_dT$l^;<&AE1`#6CUmNqK;hpr);4Mj}6phfZ(_<=y>-xC5`jLRS>
z-K+n~Eq?WC7zD`PSVQ1gX@l81qg^SP1NW;}FS;e4I>o;gq);;+U(Pr%l*;`cyMdia
zWv4VT*5TtPv~!a=-pHSVLu&*sSypfhT4&zg7wXKgczEW&0r|$PADG!n#q=8_Bna|0
z$)^);tCpjDe@QfgeC?NwzWz<={|k5Q8%@AJpN?j6J7d4)Cqx)w&7qA1mqk7760ZiQ
zEpW%6E2(&=T%?i{;_K_HSQi?q@TJ@g0EA%QoH$<^gPh*y_=bzLc1nfTg#JA^A0Lrd
z-itu9WG#0*_ZQ-HABI)(QPL7P6Xa}aiQ-;o0M?O^l7dNHEhKb+b?a(t=l5zdBiVd>
z-P!&ou*>|{?Io+W<0fqpGGUAG*?~*XhaTQTcbelMbRHJ68Qk$fG|kISsH3IDc<a`O
z|N8vT^uRNR%O)ho{%961B+e}Y22=h+vmDcFzXm%c%fx0Y63vfCMRR}0nm4nz;d_ap
zOo3fL3Skfq#0uanU<40;`a311ZMpe=6?5x${r>XotDcg-dZe#vF<dO*OdYqF7;*0P
z=8x9*5849trtBwmc1m&5C^alFyV-xz9KT%o2D2^+5OWIlj)d?TPH74gdBw#HEqCLQ
zW#13MzW6IfFg*cyEWjn)2S35X9?tQ1x9yaOQSJ^)KHRihL6<T2Pl2o#C70>tN5|i0
z9UL5i`&ZGs!#C%?;*o;RQ$mJQ-6@8iX1b13DQsCNoM@5pCdlUIS1REZ6eNgXU!})f
z#(sm*5_0c`6j5aCVBKuII5#4k<LztN8f7aaEDY2J{S$Tz3kw7S;pXO+0D}0*;wIo`
zJl+y@-7|%@uX^@FTFh%%UrY6ub`d2chIP}mX7r2d>G4c()xdUNh>LpdOP?U$^O*Z2
z))q{PzkV_67YVUMJ$SE}lvz|{ZDuCmeq`(5un&wgn8dKM{QIs(zKS&X?Sh*wQ&$_n
z=b(kwYdq>(*JYrmw`~RZ?BA+-nv3#pPH36IWAAKRbg(_X`A+p0VpwYUvJ7gu|JAXg
zOR014!(h$YM3kegEjKqe(Xhhm(h?~(^;y}eoC1KT4GwFM@a&0R_NM#hu^}ax{#})|
zN08Uq&EZ=pyde%|MAX@hI{eg3*$yUuQ9ipJuY|?;8g6gvpCsf9MuWQRX+0!8kv?Zz
zgmH{(V!Rj=X=vK7IGjbVWT>$fX;-X54@~T(H*AMG`<3YE=pf8?ko&8N8@e-lpln#E
z*X*v{OnPDiLZ<V}kU`6$20yTuwLkIwil~?+_tddmG@1<u_oWWD_BLi4DL>VnmSXBJ
zmO4?2b~;RqjQ5w>1O(Kj(YP@B0^^60wBGVHJ2+SldP?e@%R#;&^G1t9u7lrrrEETM
zApmTe2DAR0kK5DGQh=UIDmqcKS`+r;`3@KUP>`=4-n2R1#UG~|pnWtONZ7Y9#y^|<
zV=B1uayc~wG+;a+=!G5L)YSCDmK(@Z3JM>fM7{Mm?fynTenmwEu~BkT(o;)IjGJg@
zo62C@(N)#6q*~VdH<Jm5B+Z6!TAuVE;T!G!DokNJS1j}pR@U+3BSm1#<?uKlczaGF
z-)7Js&ws>XZBH(1#y(9Z-=Skv$6x>P19+cm?(M*$&m%+7MLq)RQBmtOm>#$4sH&23
zeNg`ZEj^xb9-+kydi0U~_;usB;o7xM$webO^!MxZZz6fblsn`0f_$Zzqkl!anhiEe
z7AP&4gvS7v^AE;ChDI|^C^1IZ))D*?fnJ9J=L7|}N3(kR<JixuDq_)E{pjlrmU%=@
z@0s%T9VAg|tjKC%h8gj%ClV@^@D-AdI3D9ouf<+ME?R8OHZ=eX0}NeiK-U1z4N|Y#
z1}{`b+S|8~Cr@3xjALnOX>=J8ON$tn`WH#Q7N?iumZUK6F(EzEP*>z&Q>!`l%X9p|
zU+9#i<2l~Hu-{+F=O%v<;b7lmYEWxy42Kr9Wu^g$wzA?l@9OFRM)%`j-%d$0Tctg^
zXU~QssP7a}aKxb>pg4FD7BfULy;pv7-t7(YRSXU%xPh3%F<`xkQi}|Nj|ds<Y_Ts0
zcFv@am(|fC0o|-%5~@n|C+0O!mL|-DkDS;@JIxS40FT7P74us?cmxd`RZMVRhYq&I
zL_rXKC0ij*mSy-FeZ+IuBelPg_v;XwEWNTnJ}JbRoJ3#kQR2a+#fk9qK_Qkn`O}jV
zFbV?B7)QmPxxM`!Mw96SK_Y$8XpR6*FrVqJ%<FbU{kV0vIl`Lk3PRMhI8j%?7pKLm
zB_?JJs#n*2rl<!Hm8)OOiGhW8Y%Ce(csWK5ULm{|kYk;;E9Ah<%-lUQ(+IjbU{F$u
zzH$UM0vLu4I{{9b4H6^}$pgf}fIMSIsLrifqVY(d2j)_2aKb)c65^`ERcJ-c>D)yJ
z%{dMx@&Gmlgt#QMoe3b$8{;1ik9MDIQi7f8%+*Je;H(NAQEUWc7s0;rt47_vRD(+-
z&Rl;|gpOb@5K{678LBJ?xJwTm9UXxj9KJ3oLcFVOQd{UNLfpPISJR1SSP|tHjVx4~
ziEUwOZv%nN4N6M-uggrpxq0>H7whwJ*9AZytgH!~eFHEr^~`m4_S)*+L$S(gGTV{E
z9LI-|+4Ek1rI@}%rvP>l$Evz|ynphZm=?L(A^BaofQ*8#YKmD7<!4ZQomFx77rkN*
zIx!V>_=D?EGeMu==<v@E*1nC1n4rvWvlVMCz@)9H_zP7wW=%0U>{R^}dWV3yF<%8J
zcu|h~b(%IW-mQu2Im!BbE`g_eqIQ8k8KLFwAMAVVpkzRgl;@i}-Dio(;ooQqQA@Pj
zSy9oQ&?7tHGTwJ|r(Ya|6vU~(E7gMP)$h$D(1pSi`OhzkoC|&+T=V#wkjv*aPh+~E
zmbyFQwe|HA1?r*c{eblXOah)H<RvVUC`Ke(PbnV*HH+hSol&7$EA36GjdpwPS!0r9
zi{uE=%a0cffwH^kIu3yskPRbE|Dx-#fVEEcFL<@nHzw$PPQ`R(L=F^r|E>o5oPU{%
zY5X`;=XS_>qGo=rErgsMb4K>x*_`xBH|I>U@06@U*SfoE436<foIbTk#Xm~ecI6gc
zUuxt5`bQw)dCPcK=YA+X50C~h-YvvzGH{RA&PDQVbd$!Kev4}2qqREISn|R19>Y0S
z8GUmX6QhE!JN1tq_$Zut=!~PamI+h&EW*pm>|l8sATx<Ll6lSKN8x<_cd5R>3H-4;
z-+rH23E}z&1W?|%QT1ZZ6H0#Y;sMhJBUtH~nVF>nb8|F7{2&J#$+#$(lA4Nnwp}-z
zQckf}qmzga6|N$Ht9U`Iau(wp-r#iGJs0Ry7DFj*oS-L$d@JGxUqI1juy+isG&NoX
zsWTTQhqb-e^p@b*o~Zr6(b?Is-f<n8RiK7Vh>Pp$2bRjW`}H1w1u-Flwj<uNH#}|j
zZ<fNNE!I94)o1A5<&BL75Q$$Oj$)977vl16s->}sfFcsJWxB*4x3Aui+POhpDD%;u
z9<9Yf4F5HRd$Tal);O(Wg+Z;Loxg<gU&d>*<zh@c`yKaRFw!ayI9=^_U5$V<l?Xo4
zDY;@BDNYodiprJwh>dyKY(j}Yn)OGQYB2wEG2@NN>e=AQqrKm8lmZkX^R#d0ph&fv
zsRd`0(Vm{bh?@@;6datLWrfc6pFhH^u=diWI}a(gC&la4pTT}f-CAA!y&7iQz<2j9
zC08B0G*~}3y71g`RuR}_n->CgUjO@B|Kg(CwS4|(Z<+QRP<;r2XPnDJZEM5n$%%=8
z837z(_bFQr`mb=%m^1)KVGitX0EY1bQzdl37d*Ie7O#$Y`7Ddm&Xe7HuR`S2YZH%Y
zzgiO#5>_`s=z`Sf_|Kp61b5CV5D|yoqWx$H)h4(D-NJ<A;kF1qG~PhUfoe&q_BLvB
z<lW^I;@9k43pNAkB|WnsJg?|^jd>n*R8CIrWcxi+{u2cn3num>-Q5A=%rA_Md4Nnh
z66v-L&{G+PH#_mZtEUHQa#?BVRB<l@MVWjxN;0zLt*x+$2v{Pq6bSKbCOpp?OUw0n
zqwb|vpU0QpwA?xKlCP-Po@+6LRL|#EU!~=!4`w|@8n;|+q~|zWtd#)yEWPNT?;WJF
z)b$iQ!z-itoi#M+5f|2h15g$f7nhTp3#yvK4npzLemFDz0t25=tpGWF>IjCh0m~D@
zIicP}p0ck+IKYCfbC7_+VT(nCSqt_HMjA5?#*x1Zcp#y5=5)m+C8Jp7`nDgLoU9Av
z+2ed4a2Q|v0r})&gzd|hMId~JZa-%I;&wa|ZqO{1i@03S*Y^6BtfSMmZeGjon#JN4
z?M;_J4ni#OX`+FQ=iuOA)CU6-1H;~ehHjNMJn8|XP#uAwwyxJ;IG17b3@Azy5!H4i
zSjCv-r-wcg<Gr6q;M_q5YeKBAEp#oUv(0xtBRm`_KHf4^;Q_5qef@8cWtSyjJ}`4t
z*se#<Jm#m5I}{iL@mzoFSWslwj$l+EXlC@1*c-9dFO&n712Vwnm3a^_Xe|J;7j?=L
z1H5fi2_d&Utm1xu-y3u_{+@}c`;bD2{|*h9;ruMpTH-M*S5Xmpfxa=ePL?apknG4e
zCd8ti7S_;;%@of*f`E``W$0m%&++{JukWXAvd-Uxlv-Ol4!@DiGMDptJ7eaQokf0J
zSq_iCNF<4wany(c#~xw!UpqVBVby(Tg?6N?N2D|+klHWIH#CTacv@=qp*Z%Y7aq2p
z{!XsNUd3?Z?(g&ke#E>q*z<(>!GkxRm>&)ke>^N!L2(SN`8dP8A?RvZHrU(es$QN4
zX$IK%<F~f~m8Is{c!=o7U%-ooJ%NcD@N1Gc4BGLzdIUXJQP3%P3QAR=hKPFnaje_{
z`ZvZkV1NizRxlzE{hbELd;jk4&gL7$q`&|3njy6|e2R46d3^u{z>czCmmAp7c5Z3n
z2xD^S4%L9_wWbDu9R?%@uUKhs)AXt%2;E7uKN*A^@~@U(LPCPC?`3xwR+s$keIO_>
zyY?@LyX<&lHcEqWZo0GiP~t0l&HQCrM{xQFZp%&rgoeZ2-36}KA``e!bESsS`hyPo
z_%sN*FMwHFLRxy^;_QI=56rI&-FTH2=vPL+24UTw1QRO^J05*)P}91&wVLy(Us5^<
zeiO&4oXX>EGL4hQSJCjVjL7&5;~vf`V)Mp`Fh;lt<SAglON>~3^}<FE^TvaHnarl5
zYf8Q_h(q}=FK&tn%$<@scH97$>@<E775`7fRS8JF;iTyj?R?5OUdW9Ju{uJDr5@F#
zxG%oz|M!o4lr7AuwP8jU2qT`DnBWjnTbY`gnwcHIYidNawl9bQZvsPw=!-dTxb+g6
zXF!Mnv63lxH-O;+CW>Q@@%`q*;d_{y=p7v`gXRLzTA`N(D8R?k7{Vg*ZKWeuBE)#A
zFUW<!`}_n{hYJi41`}3YDIKAY(cRq*p#t2;z+Mbe&B4A@*YVa?BiQ<YqE}Q^VZ*ii
z7v=}Z8T^mgNixSXA@^ldgXwG|V`B}Tr>~5RXi*PQ`2dSNn*wM7mw3H`XfZ>_fyi;}
z^Ji0j*y(WWL)R!_kO!^_r-sEe9KX<NqZE9UE(UsnUVJdo2V9AWoE)nEONb{FC-fXn
zzuBa3zYRSR3Zi{hUH)q+LOfm!39=tAXhT}LfduI&91^bU)sQ%qUin^W@aZrK>^eYG
zU{n1gA1ZGkX<rEE(#Jp8{Nex!?c(JTn}plCf}o0Mf{G*8-yd60qgYP}nl7H#6N73i
zy&)xT^37&8-yZHd4^88f;9OJA5*K3^5K)FC3;JFH_T)j4=OAXenDfRL(yNuRp5vpV
z41ZQMLiMo_ec_)g-;|K#X`&KO2b#&gkOICl8S4inlD`Ol`sK~7Ep;~ki<5Vqw+9Vj
zL=}S{5e*S3sVYSLSTec;G#K74twL!E8R<ttJ^=wUg~k<ih~$7AJP#r>I#e*8q{t9Y
zYWbezVaRF#H0hGh^3or;btz;XVh5Ty&{jA-zISg)RTeYJOI0<dVPW#>k0=`QbFQJM
zLav!_T0<eBrf)3Eens9|C20M1CjpQn5JWX8Mj+6=e??&XU<PAaacGzO6>)dLB4p&t
zEkn~3g3VRPgZ1XM%=ZWMglt2G1=AWY>%uaf4n00AfDfUF*O|!l@595%@$u8oQ9Ihg
z!ou4D`K+vXx)#yw<m{-bivBAE^d0faPEOT>ycul85Nv6AVFH>77tH9h|M8hM=?PF%
zJdXFwszrg!L(Cv!v;+;>-it!)#1>+}<)1%9<ox7gd|^m&{!?8lMQ<o8rNx<5?nZu}
zJzBk6w;kUuG9L~k_r&i&#sqGQX5L?4d!EfZa(LP$pFa1Jjx8DXuZYOk-ho=wCtKwy
zG?bjSXZma=b`AufEb1n`*BJ?pS#6s4Yz%-GCm6{rDypxb5c=@_GoZbhoBz&_$gtx*
zZ{T5zu;aymn_A!k2Q6akX0%Qk00kIDs1hnRRBY`4k>3snsFH+)1Z*le@!D!Z{Z`%%
z*x`Papm!r=j^l;esxVY(Qx?*g?gl_?-rjjZjY1W=Bj4xbcs#QWf5yhb@?BNC;#q%3
z#}aJOXKKcG?$iv2`Cb?SrB<Z_Ofu>C>jV?LFhe`Bqtx}bx3}G&l8jUMSnyXrb999k
zaKZbruI%^k&n$WqCc7dZU6MI&QMN;2RYBMBy$ppb$fpX9(Zji>Ha4DHUj<&~Tf@xI
z_wUVzEb^fv?tXd_p^QQVUb(TCZ){(X4@p8EaV_-jW<d5PP~Qmdd_)nMjKl~tbWrub
z@JY-X>l=u?Ne_`R32IDCE{KVda3(_$eE<LWH#~eoLa;+D_rC<@Vg)dS?VT8CB#hxv
zjpwyP{Q(_I^@OkVENs9|BA!Kq+US6CTZ<o*r!ylXO2nGYfN+9ZP>rFZSS?=#8kOTY
z)e!RY0BE7qpDk(rQveTG+J)hQj!x5K`l!w6nxD~^me<$ofXb=xgF#bM)7jZspn%&G
z&BetfP_P3iQs@pR4Ndwm9lZzeoq%6J&)HB>u@BuJ`KHv^SWbh+T8m+7_)$QCfTgPA
zcZdFI=3vx)dWdM?>hwSU;}53S*CR^`tlz){3v=@{7=Hs}ePP2cE_HNte4$k_8=`^2
z9@y<W?|An;LAeDwrO)5K>6^z7!SG)IPQNvK=;~qIQsB+QB%;$blyjFRfGr@;gX+)h
zSr=E8OVd{{FoC`d^sXCScBR}!3^IutY`vwPho<iV!v}mDfh{a<w))cp#{3m81B-iy
zG^hn`YR%Ri3?Pr@VF6~gLkK@YrA0@U7{x5wc|aDn+yF(<)MJJ;XJYZdnwmQz#Gda4
zU$u)|vS(7IhAEiByN8DfLPu~q0jU~j*o%9Sk&%ruIkfMBy{SS<NWK#7=Bz%Sm8{*g
z+=bBoQRCkH-tMlXKTrWPzJGn6o7*wB>o8EMZ_83k?rBq-SS!NJyb4MxTbbQImud@Y
ztMj;v`CK(FLdIcfV>6V-3GV{QKG@qUVGJ)E#)BDr(6~L+1L@S_3hW2SR!-q-vgZcO
zYXZ@Crw&u`4&On><mos3hW<Syz97}_pX4<tHw}LM8bfu48lnXzIBEdLaV}4p0Sswi
z&}F8c)rtKi=pESA3w}u#cdnS5o6ouLUv2mdyaPlHW~tPG1t;h<0Nnr+lsoK!`+#zj
z`cL^^CX<f`Mt+UuX#2!OHB^b937Dto>FUBZG=|dEiYt&7xiRvt10V?{W3f4|6EZFc
z@!vp40wn|cyI}T*?LQ5wo~fUuQ0O}1z_epC^AQMSoci@ULA|gmrlzioU-irYdrX$x
zXiFXN2ElzyR+9e~@+WvZ)RXX);2GJ^WCR>>!kGwDURSJnFDcuoVUk8E_BR^=_|Km{
z-PMr<0(zj^o<f*qdKL=jig3@%fDe1#w%|cvX|S^egjqXmGELn|WmsY73xPBvFkDVI
z;dX0J?F0-Z=D%snspk&cMk9QRoE&uVG(z~h-LTHG+G{r-koMN$_qIU{?jxh8Z8Nj5
z<|^HT`9Ed$*9DU&FIoAhvsiGF9eKi$0@O_YPL!XAPp?%294ufe(Ix|Knk)TrpaXKS
z*CYh>4y6=_6^whlUL-0!1!Ggd9Th)pOZjSNq^nS(^C_~^-cnTS1YY#DPU=Dw6}(=P
zU{|pZ+yl~!xgoYp;4q>UjXNzN5@FqW9anm)OZE}Wy)>+t!J2G@@s^um5M^LuuJfbq
zH@h%WhDn$AjhZdehro3J!ejR}%X{n?o<=ptrD$w^Zz~+n6BLW)qtCA^k+Xg$j|L?>
z2Q5<olGFN8ke}bTm<Jbo=V9I?V|xdE^A1}z4*MH<U0o_Q6+M5@W`=LzxWkCg*giLI
z`em^5su&^b9*jgxLco&O(?eCmcGH70*!>!c)wB4cuvY2o)o`;6++bzni4}jH-Ovb2
z2jBx0`t$SfNP54vZp7x6{@d|29x9iD(RLc90Z9lQ+1!i8iJ*<RtzRE&z0^+szP$_U
z4M4FyZnVT`I6fXHsmCqQMvFBo9l>&2aOCiREw}T8sJHBW$LG&vX9M`e7c<wD9Ieme
zd?l!qScYVXs~;xu`djyOcgM4;;yk?i7Hwsln+T@>*u=*<L=h`KN)t&Epa^2Aog}=@
zZ3H}{P3Lq_LL*odFJ%YpX0(IX)QJKlM`0-+lu-98FHcK&ex#`OE{VIX`y1-5O(seQ
z!Q^XFV=zFsR7%gib`46CI1Tk1Ncxx9H;)u*=?DUeXsxG8Uz0<2a#Q;@!NV>Y{}IQ7
zALAj#wgwnwHzcurfZFg-4cAH7&hBEX2DZVY#iL}uPZhLtc}^OS;)XaoC)o!=HgOQ#
z&)VV3h6V?>32+!Tiwu`zA|814myf`V=nxAcKo6zU6wVIPrDi9A3JCjrDa5WpI{s?;
z&L)l~H8HW27?$TrnB^7nZ$(jr2;&%NyUmOKhEy5CO*@W{1vq`3#X}lVp6?ggWKd`C
zPfnjAla}X>bXxz^)!u%|#{aRbXFvYmU+W-U8BQgh+f^=ZrJ(MG8oC%vndNKSZQ1?4
z5B?>3!f!d^{Sz}fcVoamf2YcF)TBd@h!*%k^j50~kZ4+ghuE^7|01mk-${ROm2n@3
z0{EX_Mz|V&2b?0R=*bfRUl);AjUFU@skwrHJpRpQa0^yOk~Iu|YN)R-$Cv=LKJtEf
z#L7$Nxznk~n(m8xW|yy9^~_%+^yMbQYpJWp9)sX<Q4K7aTh9F^v*LJk7I0>2woA)o
z27c&&wOm+K<mzyd?puTFKC{VFwZw=7u$iKGOR<^@#&txwLSARbw~i;hLgrkgQhZx^
z!TfW&^x`b(8JuPD$*t;LSO0=3ix!p&J?efHF%n2tkxlo-2AHW3s;>^WKh9nHZZ+F@
zgDOv`<-qB1Tc2r=A~nJHM5RJ4bOGmfGL(vIwDaLVgD&rExmZvvl1^M;Q|n!i6w4|z
z<600~WF`<|Ef<m52Z~~8J4Cw<!88%qOFj2dU4-yrm<V#^{9oQX`5tsfD><MP3$@(f
zV#8~{{J^dQIiR`v>I*MQ#CG1hu1?4+Sn%TQ%Yqkv!OE<VTIMRuf^*xq>w_ZX>dI2G
zfmaY+HD8t0Cp^e#Xd)<~)arMaZ6wxArXh5M@oOyg?X@;R)>9vE;?k`|GbtPym`(r?
z4zuZR5Zu;UfbvZ%@H%a84|OS$$MbTmngq11%ZmS|d;SUjX5%~3|L&oTQpk5Q#zhZG
zT=}kvoD)ywM0}da;EdR=7rdjaZ&HL9GF7+ZdL4<F;RTw!Ter2hH?jT-EmtjLdwX6h
zs&}j>%A0ZU*`YjKIue;ec7v|#i88bO4k}nVgEx+49&Mk}e%^=rTx8y21$-^82{Ys?
z?d|V>P|m^;)H+1B81ulyEkT{&7o{3EdE%sgmahy;(rTf&UjCd-J!svzYNIfP3qZu(
zj<6p<msvP8jO~^z)9N9hyrUNUva56t5S|L^>sFu+4Wz+H&!r#76Id;%cBUo;U2_RZ
zh>B{f?+3S*vNoJYr6*Abt*!outCXEj*siP0F3@DV1JfB1XtMMYMEISWdOkdG+#U{M
z6iQQ;OjhjTOYsY|l%}3p{oUO-rL8y|*$_<&y)FxJRBQA3#3vCJU|dHs%c8{*qs<VV
zHb9F%CFb~j1e#YJDlM?b_mYz@90IhO$dml`Et{jz9yoYaS^@MyLeV}78M(WMhxTk>
zo-fm97#0rsrBTm?6^;`LCMZ|heOUNed$D83Jk;2#T=&gS{#a2}+kuwP$)8vl1PC<K
zr72*8RS^{XElqW1KJ0v%_DwGEBsOtYfQ?k3QPltH=?LxN#$?1z-;B4TSTZ>uJ`CWD
zx8}uq$DIh_))auN^)2+zE9;|K2&yzJTcGq+Oa~c}R+~|eMwbv}ZftDK!BBcV+!`sJ
zo=;Pd1qs^BBw~};-a|78zXz^tVFC^^lzp)Kn?A(OhmnJPwAqdIc1r=^S;%pXSgckM
z&`O(W%(=ITz<I-;LMZzM<|m+tuKD<lLh}EVAu5t5?56OwD5_^cS8J*Gl)>!>vZCQc
zp37RUXD33!poCqp<ngwjUjjo03xDuzJEBJq!=^SG#P=X$5fh`1>+~Q0{J9fnAZ<jc
zwT|eN-u^N(2M5R0_&smT$;#~kbnSU$OU~lrqE7+~Mhpnb>aLzgEK-(PVD;2}mIa})
z+}{0eq0d48-s{2=!xkC3y(rLiES7ihxSn69G@ZinWEjy)y6wO^0<1uz54H19L)`n(
zNso`afc2IaOccSmf$o!$?{QlvP>{hgpGy3yJ<r5`2^O8ck1MT|IiVQ%t@+0RAqIxE
zl29WU79v#J2u8Bc0j%V9aX}H}+1vlrv&}NOSiR^7z}RQf?aRKMI2)o+VF*Z>s67Et
z`oSY@%cvM(6xo+C=v_}SXu6Liz`WaLUZUXrpr2!X`a#p9CK&Vo8Pd_tm1U$e=q(St
zFJ7M41AxZt^g(1`AY$g)!(WgXf{>z9rz>SflGyT@t$)~?H!d~QtTYRs=x8odj%X|u
zgvc5)i2CIlB`Dh3g9zY=Lr-Eg%1JsfBb?rjg?8`$#|2Aw#hU}_yE&26^i3o&cOme0
zB7(_8Ymw_|wHUwYH_p93K}{Z1xdAxiVh9ZBX>A86ZFm|#O*xb3_aIG}TI9%&5cWYD
zC6nfWY^rzfkB<LzcHXa?%1a)c5?W?1y;>Q#sZFl(j>vZdC(EB{hw5HqGFie`7&=)q
zC{%Y(4m!92=`trzSamG(x||l**Zcm?foLkcKP&iJk2soo<aOIzlbfb{nwJ2ZG_G|*
zZF^w_F&z)*&1VEu*ABS=8O`twq;Dl9Op_I>0ulzMY?{As_l-CF-0&pd6-}0T6P+;V
z&BDlQnYk<7$wxk4KyTbeF6O+Aa=rHdsKx%L0!@tw7)U~Q@9y1rXa*j3F0`5<p1ncq
zn_nt6mrT>P8KWLxe}<?yKicn(e0%?`s9W`dI4=(ma7+I}j5Vb$RW~a;?mmtU=6%a&
z@d(fvsE94Ctv?a~;1<r+fZ+-I!i-+OJrFU*H*HQ7B@h0%V*40Ho3_Pw%h}!&v_*nV
zpP(Y#)yCu!&jq6%$ZG)MpiZ_17${Jxf|ZR`zF5@=gsAyZNQ5)eqqF%T+bDn~``o??
zdK%Ek0Hb0GCp63@s?uWJ6Bo)WD!Q$jmxLKQvCBvWJZel{BV<^0RjET^ckn6HjxcP6
z?D2Kvg8eCQr58y=MMW2B^Ruz`?`8LhyaN9dVDEb0(EOnBP=7+=7BX7i&li$%(gbg0
z%lVnGq8)IQco7vj`uj<0n4@Fy>3e&73#LDx-;~SA$sG?5;|Y2cxM3YV2^FRY^97q-
zzH&i3i*-2cV9YucVL}Kvk0bc@&E64_e}=j#fdG$L*8v4ij9oldOnQs7a?<GD1b4HL
z@NiILjE{`;PU{n%`^bh)ibR8xWTwkofqb>Mk&(1%>rn)^f+He^*VKgSHgb1fQknU^
ziTjxR5XG2(36UfRW!NPw*Yh(`_KV+u7|X6%e1#NS5@6Lqi%Bek3%FW8fhcCjhJuna
z5N6>*B}VtsqzBW)yJd)02Y40Ybh!QlX-w7l{*UQhK(~gR0Gndeo(d&hWVAY#el!f#
zSveuF9&mC)S|z=k0fiPn{YA<J#U6wG0_5U$NHMQ7l*KB+2*8X8u3fWTr{>HaNG+9J
z0B?csY9`d168kZ5SOGxF@qE==wb!p)rw-H#AF~FSlBMk;fE=9$7aSqR1YM3&ZWmU%
zlx{L$&E1Ar*h61wU{f#n3`kuwtuSbKD*;|~b#ntlpYqF|f4A<~X<nIG13&zB&ijdZ
z3!tA2@?}(?r>%6^HJNQ^c=vql{YQ+){1l{KBj3JZtRaBiBI<zJLCOMi+zmiZuz>Nd
zpe_3Wl#WS5ucnt98ZM-IR}ec@4*(gEKmixDaddc?FrM(Q1%ZMj6y+UlM;xYE0Ez;a
z24-t6JpZyp3QRZr0BW@D1%cNc<pwkKwnTljA_?E_7_-00;x$1^0dTlXpl>}>8+H5p
z1dQ`(=3gjjIRmMeK$0{HbIVJ#rsn1vQ0rHp0vHJKwY~avE;w3%TQ6l>K~t0Hbm-M1
zvZ+ddXO4ctW8NjXid_mA?Tu<mWcXUecgT0?^?vQ{UOcmvz=(1oz~1`~6r46==r^x!
zZeF=#ah1T3kFxF=>?~+`gWiP8?=9%@&tZZD+yME)Jgg9-Fsmp%sIsg=h1J1|!<)nf
zYi*vZ<)r=4=Mu}p1{M#^$>&d7Fc2PC=k83r-IC2dbXc^PR0<K~dfl_&+OgLJc?BCK
zVz&+eaI2>kKtdOI`UxvCZXx678eOz(XPHo4s9DAMslELq3bDH{c)wFJy1>r1H<E6L
zY@+W=i<ei1{LF%o-Qs=N+B}HYx2fHslTBxNg^<vK2@1%;p7l$YFL&J`Yn5#Lb$Gaj
zNK8&X-{~N<qACvL3#E9^BOt(pj8^mW-MOU2pzBOX?U%h7{Qx}4Fk_$%hS~sa+C6NS
z)S{s0x^dT$#YgT8lfK~!eA)9SDpx?iPxM_hlZf)n1i<Ikl(`4<;<zF#WKDjg$rDhp
zLf?QS7z?diW`Zqmite=`&g8YkCsL^R8Fc+ZT5?)iIGZt-$x~SI2SN}cuR3__uU>sq
z2LzD#_;`~DAm`BiF82?+jvyr^J$57Rq<=@64qashR3VU!93fADI2<4xff6Elv9_`M
z+Lu=>XosYad0IxmMXj&WentL2apNwZw$dc*h{_XzecbqAAr`*X`vDqYzH0fkl64xw
z=^4D!R<Uf&F~Gc_dm^Ped)kY3IY09_!>Bnj{Y2@|US}s@PZj!NY)Lg_P|_WL7FMv1
z-o<Kz(kqBBx1|`fVc^P?(JnWw!8<8zz(M>jvczg{LPuBSA;z5Z;*JT<m4+g{I(vf%
z8MQtk7V6zkV8C=HLQJaX0tkkc8*G%Fe!$%?_}H!iUbDDGC1^E|GQC4qt7HQGk7^jf
z24}m${bX1GA+|*3ga)O=dfaMYHSsw0T`8*b)}obrk~zjs7JGRffS0n{1{Ww+q*rM-
z{4OrKa)n(1zS9e0CuoAD6WE3dJK363Q6)Ft0U(wU*x<0j0ZRDd;dBA{dd3frRP1Fy
zyS)<lnNLh;Fu&rUMJ&96s)_6;s4cGAMP}op?uTHhrii%my5Yr~O?Kv5qOkVR&RP-p
zBG4rPhqCVKBe^_SC=`LdVCh&^S~|~ue{F^KS$J<9G!pIwB3Jx_Tt+#(S3#$KG8q6k
zvC2(kGfZuVfC~^^ljAxG`#NCtpy*LDa7WUsb+_S?9dl*}Vz14EN5j0~Vs)lM0w+9+
zJt16#-9TJu=m3x{rmLPKg#DIwroZ=MjI~ELFFW;iMaRUnbpmy%tYO-JWge;yNqWFV
z!P5i`hTwtIeqf_TFp<cC^a%cTTj?i9(>l>TWJe;*ViTa)rS3c?X50K)rJ%?3dtPQX
z8Gvh*V26>x!8fkI7CHg}Pnu@{7L}~+K*a9;{yq61+#Nv1Z1+JL3wjF0)Vo!d$3|e{
z29H?k-o1tBqY<b8C{lb~G;`$Rh^~X_L58(<xv5+nn>xgm6^GJ6aVaC5(3@ymjIF#W
zbzE3TNO@(Y>0_$0TONS3bW-1C;XoK~&3PQ&6eiR&f@YVI%|#(x;7?RT2e+vSBEA(_
z*diL&%QXD}xK##0C9UXQnL!eS9CW<4N1=yD`x)M`kB^TjvcYAS-tAcgUKxvXNV`3Z
z7EJKZ+B|0UtHoZU#U!?1Hu&3#$Q||?JZm;~5Q)&zUWi;IEg*a7&CxF2nTgS4xe1pB
z>FufVjV}*NGVx<Vln$;9ZUmBmU@77+`9qh#&`JV^fVhCZ;#ogHwa;mh0&<MejabNX
zJnPQK)+~>L%y+eq?l00}@-#I`z>Op9nbMLT+F|%wf6)SOXVd>@F#Z2X;M7OXFJ`VN
ze#1AjF?DdXH#4@z{ON_sGkktNA%0qJTFf6sM7R{qY%CltX?gki(X{{m;*xo4<!ENl
zC1YjmXm;Q1$qQ36F){rAx`fNuCaS*hO~LJ1!-cws-XfyPnSvRk66%K=7wR5<s-exb
zN^lMAGxHRWRdjuKTk}it@_g|#U1rwpY_0mgS!3UvA5TW)4*$H?kdf(YR39-XoPl_I
z=JP&z_T@zhSAuUnliRr#MPa!c@)XakZ|lEqmi^<%mCkWr=VFOt+d!`Oj)=Bwf@|(Z
zobkNE&hUxYw~7GHi~-UV(>FtrmN)gj*EgS>?4HJ$XI9CJd3~hWYvH1KeC`$vb8ccE
zPMz)@u!>6%^vjHC)5P8xIX%`r{5$5wICU}Uy>sk<*Z1jkFUQ3p!ojZW;LxvbYvy`b
z4-3178Xhnn^-C-22rs_Py2-HhiI^g?c3!dh&h*yWA?}Ar9@bWD(=|7i2aQgC*mwkQ
zsbiLgPYR7Ldp_c|FE-RXJG{)K?YHt_+psU9zg*td;MtA)^K03_N%<mCNGBKR>3Rj<
z`uddZI<bZnKJk@&qn9`C4I}-N77iXqb!;O0j9agLN4f-Go!w(n>&s<RT6&dNAuR4b
zdCfYAZItA7*LytDK#v(Fl`Q3Vm+_QaO7=g!@~1=kv?%&fetmHx7nPk~PnM+4s-|va
z=Et?6bRjNzT~9gukT{ayZ`UfvQ+u*jn%e?|Z0od<*L!@};u#{p%}S2RWT_KH-+zzo
zGfn-zF8ThDkbJz^&uDCIcM0Ncj~}u&(WpLDyjs?e;YxLbWaIITg15D%v9iM7gUyi$
z!KlNIyY#dK@$$x@<4%&YPFfzQ{-HukAJOa%8CmkmgzFbI)}Mqc?BV>f3fy%2HvOF`
zX(87!7oXMqX1k<oIbE*sJ)$qEFG<=1gs;!tCs&mfR(rkq@{h@KSn{<D0}Y#p0feR@
z3T#KCnFVfFJ`rPI!mIYZIJP+y{~|AuyC4r?@q+nJ4K)eX+imhA>`La`Swk^bQ7q?y
zlj4gaSDjiE4sWwPq2RwzywMzn|3phkkA-ho@#D*7DW1!}pD|(QYlqMXn2VNBR$RAi
zU{gp(-&ZC0t54eXM4|jzq)u`0ZKGzFf^{(973sBa<kEht)n@%okGsu<{K?yOlef39
zr7sn$)E9+1Sk}DUoxS;$OEGk|BrE(AO;~hwyw{LHD^uMo+upx%omeGOxCAFZA4>D^
zu-4F`D+(++Y)A@!_t#<Lg%+A3?%RH8xh?2_i<a3epFf6ZPD9%?Oq~;5cQ+vhq1I_G
zTx}N^WX0K4sU0&qS2&c&j+45rW|6(z>DEv%lvKPSv?i$&xtZnOb7d-EV=6sklv#xc
zx0PCBoNg)k@(j1-W&C`-q?UQkzcnV&@@f}mBW{G@<C5%~<GIXqe9V6aoMcqrBZ@|%
zoK3tO#>;XSO`X&$cdQNaDBk23CC#W25t(IW5pE<ZJ>}kNx6szw4OQcL7C0uKW1qPz
z#Vwz!jjvBa+r;%0O{zl2a`}<Ur)c!^mTb2%^+G21zw4&@q=^U98dGtm)#WQ2{EQZY
zV%d~s+TGU~$(aV4T8d2b@rSIH@K!GVih4NZt<W3Fxthco@=>3xVZJ@-7d?*Im=w9r
zy0lGsM#6{Nil&X?rV0Wl<?76|!I3{gUn1&SR2Hg(yuX%CQARUeJ@@XS%=M8Bt}`ps
z^DN#e7$sn`C-<X$t0mih<?DSh?nfPiQ}<~UKL7Ceu&HP1N6Gkx_T8-VV2R7QzlFHX
zHcpd#pF+J;J;RcNRV4w-+P5gW82fGxI`PjplfIMblMLWFt33(3Fr~C_s{SB!9MPnF
z8z-lr$Sq6fd+@mQ))e`b)4BnkSGBRxI62c@nqf8yZ>bZ<zuw3DAthDB8|v<_tEBeO
ziKS-s%T{|%V%{4C-hE2}e}#VN8cIh+H}O1ne$p~A&iO4hH-VGo=O)U4A%U5^3ipY+
z$I*y{M7^CQtwzK8J9*TApQz)GU8MI+NCQv$rA$5j(|S+qI4o_={o~)e{pGxh67_A|
z!Om@^`4*NWG_0b$A5}`J31S%=sjs5YGK&s;uZz_0I1EdC`VG%`pW=0nx~$z__ZUOO
z*DQ|?)lVbyp|>bjDD*^Fj*RlTQS3!J839Gb=XPzwW#32|Gq%hTugR;rf|4RQb4i<$
zZOjWMRL!$a=~YD+MviGG?`F&}%%HzNtKY@%VqAHsb5kmKW9hb=Uw0PgC53^T>S>aL
zu75k$O#X~9qVG~vOvJ3K8*On|C~j$@7DPX$_PKryQ;v|#pXko;7Bk+fyK7;FQgPRP
zqrGDC9;rQk*uyV$%`~dH<U+REv~UdF*{_Oi&z!)yu|?F|GsAe#o2B2kTfRlOhHT)>
zKafwkNX2V8bE!}`i{v<mdK0%@Swog-*ptXvW%XrQk98@TEo&K>Z7!e3Z3par*4yOr
z4+Q*DbIH`ldJDCl*stUMx^iiF;o|Ue*9^~^T#IJ2ro{ng%t5v>_n72^lVANKBqeX$
z&*OeF891a$IEu!9(oLR{_80x6za!IaOB`8zg1S{{c+0Kqw8?rvS58u9`{PZtf3Ji}
z6`Ky~c^BT3*EInQ9RX{a%q+X2!p81ziCqnA^@3cCz>Z|m?`3$pI$PwImF}udezBId
zC)jaPw#sZM6)~P~=PfO#0fub{n!Uv-=`DLDFIKA#EG7A5`KMOx7`_TVD$qeixpiA_
zV%z2BGM2_)_h?m>PjK&1Eal15OL-|38m<51$_{rAk(Masj+Hsi1D8McOSk1!UPiyD
z6A`*%%qx4pPCaLhjTM!v5Fot&xvu$GhpuOBdZEea=3Lj82Cel5o>Sd*)_o4^S1Tt9
zTNGQV;lJ|=wXeutR`Gc4<c5C5_xV@SyHS}o^lZi3?i9uKIxX_VM}vHyeKaIE#E2f<
zyf!&h?04dH$wb_rgO2*A%mioG<)_P!x_!DSM<@1OCNlKOFLpn?`jvm6V}I$=zEsxG
zaJa&iU;;9}?RSrpwo_<+$`WIJw!f4?qrsyxrOBOS|7G&DWIV}lad@m}-<f1@V>?|9
z7uWeuVHeUvW#4?%vEE(tSeKH~!KG+1A=Q}Rn8A~b_gdR6FS7Q(_CnHvZn^I#8XkDZ
zBHGaMx9?kQ=O>nkc5P{>kMak~dfjW2c*7u9Cwm^?bNP+;S#LVZ)+O@S{_bj}fajS0
z!8d`6`J<OEOW@O=b#oylbp4m2Y=n4%2fuAB$C`Q(y|(zGC}cVG(n=xn>*oG%edD^a
zfqM<!eJpo`^6ySyW#<V6-8!EsktxrlM0j?YvFdVr;#9v>YcKYy|BQ$K>t$9GlQ^sL
z8Nt&Xx4r)%yW{5;_`kC|ZhoQvuk5Z$TOoq#&XwlBnu9c2Mbw`OxUWpj9%ME#P3exR
zjxLjt^_`2yMzaKe8gvO`79AW4r*;oMdC+fvT;IF0W$P*~GIxc;0P91e!E6f${w>Bc
zbrM9XexpkJjjBd3-7~a&{!K9ocM16?JG)0-tu>;eH^^=M>ld>gJ*NH~Rn#Fnw^K)Q
z747PreTp{c7sICsbj+OKRc?RzBZVdH%>De7R?d<&Q{nvm<jKJw&m+%7bz5HCn?}bQ
z({vBLmD{nx-r_}soF2Iy+aK|(uHfL~Uu8_KaN>@X`C-YomdX9J;CgRnt2e&38jg~X
zMArrKvZLztuqS)*<T}dblGjbWBBFE4?&CegS7KpCdiP`+)Uf(e7`bP?VPDX2^6k2n
z{nNzsVr|vw@$Snc8L^g^KPCt)J}T-Dtk*WS6@{FchIHWOeos-dVJfNYdWu8!^|eA4
z(Ww4CE#;}QTDwc`mQHnNIt_0=d-?0jgnZ$27G3<cY>IzxKZMYRyDxXS_@FYJp8hmx
zIQ*cJT%)W0M}pRn5td)lUcWuBb=7(VBV1m^y;-{MM%ANWmQ(rP56b64nCPycoOx)k
zzjIH?|M**lWS9I}uEf-}*=QweQs*C3yH+xMH;Y=cB)R{35>Y3;U;q8MySDuBZyVDG
z#?GuekHpS@f1FxbN%wQ9+#c({n#NX=JH)~8?y3Tr_xH_e6@M8nfpS(Yl8a)Alsj=m
zqVL@8RdiD|G%6T=-0P=y)gV23N2;xB9xF8WL$P{V)>Ywh2EiSzN3P5lP4VL1euzd1
zjtGxdn@PvW9piVl9!kbNQfJkkXPP_=yGI*XAmRC5wFi%MXDaMv+rqSTekl8kp;Fq>
zjQdJ96WnH#gIzBMV#6;PywVza_cJS!Zb*$M=Df1;;(fJOOeH^WrQW``mTdhqd26H2
z-+c5fZiw@DtQYe?LMkopveSC?(KYT-kShJOocLeVy=7QeTl6l9iJ}4`4T4HZ3DQWn
zgd$xMBHhv@sDLymoq{0UEg?vENOyy@bjKY(z`gf9=l?wC+>dvE*$8XRHDk;g;~np7
zkJdZ5!SV^0Sk1RIxLEH)b+*3Y3pLxj)<!?pwU%A49G%#%XNTYrUjDczo7o!Dv+c#!
z6J2OQe8kh9f3)wmE581V+})4jwf}1`)3q4w(3AxxODeX=dt{Z{_99iiZw)mc&uI<k
zk>ej+Ff>^9_bKH{DH9J-Wxmbg$)KvjC4+T<iQf~M7fqr*#;LTHRhBuF7?rAd_@wRA
zqrH}010TG%OS{@{yY^9^1m+XWMd@@*4M`+~wBPX3mR?IFZL<}-SbW!m{knN4sWOX4
z#(mk#pGB`d6gIl3;L^9CV6D(F_3c{mbO=ZCT(-Q6{Dst&akJO6H?<~vXJzf=0(ZXo
zy`09r@J2Sh#po5mgz(nA_Scx2@i<R#%w!0k5UstBUPm8#ukB$NGv<jqb16=45$PGT
zWc5p#&t${xEtICOa+rPv5>H#yJoV|cQ}7IKiWZ5hxZ_!yXzwN1sX>A<uw*7b`%2+)
zWM;1Qa;4W{S|KUxqek@h)P<VdNv+V27ffQ~hHylK3oa+#5V#X}ZMlN+P51C~bX<JK
zo0JsU%j$|O_I*R6v&$TN70RpHKJQy-QK{{uh<7DEd;D(uEaDeJFvVQcjs1M^{w%eS
z(}L0m3;m`-%Z)PSn?pJ3vkj>V7r(zSi;oT^-fl0q(Fy*#6vSPHbnO9dUKP@Gys_Kl
z3cG&(`zQCGHQ&u#*^uCgW6*uQNZeRw*HcHNIxinld(5S_Xw&xIPCfMnM%FM*WNCv|
z5Qehb01NRr-bYXR(e_fW>6HfMIqGUG_0(jgqI8?BD&-p+Vc&b+Zco@JMAsrG%aIn=
zukf%uzvdOy^SOtlcMV^<(@iKP&|8q*)6^mMO&C|zyAuJUjm!&<MFN3w)~`x56Ll@G
z#u!BOMhobE{kXrK$>MuMpfbd2h=y%4UP)xYR=M6g>O$r4)epfY&xX(~OVlbUaK-z?
z(|)y^9=!d0>1#G!#8M{R4{N>tiPGhO!@cspU{pzt%s$JM+%B2-*uq(Ce<qHJG*j+9
z$WLe3#7Qym^@`RdGG<Aq7UPbmad;Ir&>lX=gy~&U7yL+fkaLZtXI6VATxnHWWXG{7
zQ1Vx5WqsfAGs2a9n}J|)5D9Xkj0lg}NFOC3dA`5Ld-IdZB|{2I&w#m19MW46B@1%e
z_55Ofw~WL)YDrZx*q&d1y-70?X1@AxgN}fsfi&79<11mH4DCu43a5_6iRC@k9L}qm
zTY-xEmtq()0`m*Xv&$}_-Sjlqq36O;sj!F-n(S-YenX%=bgl7qW8NbqvW<vr16%um
z*NghD7}HVYwgW1r?`$;4Mz9C6Z5)s=h>{idQ#sd_Kc&QRHz@0Nm9!2^?ayyvJ)ip}
zE2DWoP&CXpAC;>BeNnE9zAN>&kb0`huP@=8S&nW!Eu$_e<DECm@vYLcGu!N!m)Gi!
zJ2s~kWYf6{Ik*)p$zz6|y!qo6U~fI<PPM8>l=9O%kE_`6B;$`3>761}X;VV2iD8p#
zQypuy^&GQ@Docc}N?szFJd>HMYlRH>vr?N@_3<7H-DZr;?JmFUYeHM3y!m+W^cSb2
zu}}sjJu<RvXu74N(ckq=R}%H{@BF$yG}SggOOrKZ@oF^u_v3ZKhUlV?RL#4KmnmCH
z6qUD-?sYT`Dq+zkZY6Z;ZarN{`0&c6BHw@S1zySz`Ng_Vd7WuQ>B{}|18U-K(G~49
zs>eI(ls}O#+<)%wx8yKjyONOgnIvv2{mqUDwl`Tuk-g@F`r4(U3g-)@AAhC&561~c
zc9wsS6S8(@+H{JNng-ghE$L*eG%XST7BzmQt4*h%uLW)-$R444s;#f9XL*kae&pA;
zw0NOyE?{bGW@@5sVtJ2^PQcX2)Lh2w)oX1!L2YaO*V@t|{C}NLj~Q9nAOGJtp_Zd>
zIQng!oao|4Ruj%*;#0YjO1nN4OdvE^f98hjB_`5SbmwzAA=Lwc{%l-E3qr=<zK?Yx
zeSM<|tu;hm7cN?!@bs;i?l@QU<ZI;}lz(Y=7x9P8m=RmM9sALkgYuO{hH#|+tN&sr
zcjw~$W#wq6D!t-9Db@-s)jVH%kr^~{kwp}+`w)#+t8)PqZ`Ds(PWVd=>Wq7$K0x_u
zU3?z4BEfHL(a{*<m@c7n<V^8#Hh6U2<Xx)qLteW_BF2c$aUY$7D4i1oz6^jVU}xAu
z5?r$Cb~C}CX4YH;`5!}=_gQE-AQu24cEjKZ243!HPCE>=(^637mn#7v3-SPuK^^w-
z<Fx_Zn7d4RpX@5Kvi?Ndj8f9rfkzWkT){CNxdTkduYtG>Gy)*P^&|WDtv1hEm-Bip
z?lVG~KA>`j%4l8L40~xdwxifBzSvG)&jA{-iRo#e<u^UP!D85J1P|GM(m~Rsu+F}P
zR#I9DU;=mv+vWZH_p{5fv9Z8O<mJn3*lW2w*kgVOv8vy#k?zR)SMOPCA``5RL#P(?
z<pAs3q7I3f)J#m>Wwh3SA@%YFg0Icxp}utz$b{Gjwli@DUH~rG*WyAvIk<EFJ4?kE
z-LG;dC0(WT$I_!%X%EQAW>BD41@<5-D-uZLwx`HMOMEeWJUkGJp)oNr+L91#qRFy@
zTr?F&N|v%8KLFCG!#>+c5pxKMSEkuSAZp#Z??Ry7{g8`;F72mQPsJ$5w=z542|NZU
z*`+4~IX*rf0%V1Nw4UTR*jm)JYei7vJf^z@QJ4q4K0rG#djzR?gSpx_vb-Z2gC%Uz
zLr55yBQAJE_g8vx;^N@+!LUIfyY}S$9%ctborfL%{0J8U0M09jmCsH=>)3SL2-k{K
zAkY7<&F~u+?eq&{#4mB<x5ucu5XOQkKh1alJIe{q2C+)N83aMwggkCF{{}{b2P+jP
zfR|1J&^-~zi(d6N;zgs1nCDg>;1Ny0d)M!I*XY2WVbXM`s2+F2@mH+(psL}_MYq$@
z<7}|%xw&8Ld7opN{WH~HOU3VU(KRn1Py5!~cRFPCF18)9G4alSKCN44^D`pl+~FXq
zaj7Hx)p9X#%3M_b--k#BJ5`=Hb?G4n8u3;b$)Ijt8S53ai)Zg#z56S4`{9>2|9;9s
zu;QjnIlCY7d#EVdB%h1lTtL@C+5zTi=dIl%Vw&|>DU0P&o6llK%$1x=@aY-Bmwry~
z&-(Z55$uPF_|2qp@u%%QEk-Y|r_(JB|DQHl2s-RGSL)lM!<Zo6ne*@)%t+@>`#Tk5
zAB${uku=#*tlm4DrFtE=RdM6X?RdZbE%deDT%pBa>@vfD3XRVz*k1h4D?i8=ZSHsf
z(*emOgL(-;UnKu_8gu~i(R7FSf9n-9l(0*h{C9ECYq@<im%T1<ndmZ1FHR5STTkLP
zo~N23Q~&3QElFwjrBDC#dl(7+=M*bt|9?ACFH<Wuk3a5j&8z+NxFjj_-klBlX_=}l
z4fTvn|LJAakrbn+vi~Xb38moPW0e049sB2n1NZ+9%szQ)we=+5`C!MA;9Dy>Xt+rH
zZ7f)b&qbNHbe{iwmLtwe{wcfLKP~CFz5HFxujSwV3M1LKd{@Sf_n)tNRyG~#1^+F=
zmpyJwB$=34{+})mBPk(Qvl?>yxAAimmKg`a;s1F>e=kv(n(RM~H6WK+7TdecAAvBt
z)q6zgxFtrP{MCB@Q|c3jr%G?-&t?!<6b!z?tEQmWDE}G!A8pAe1HS)$SZ|IjBQoZ{
z?T_L=%DnUMU|R_iWCciP{yJa&WBEk#cXaF|y@dbMWePi#L`ky$scT70vZYOQEFKu{
z1+>XZl-B}r`z@%P*Uq*Td`m$|9sOlTt^d3Mvmv6H6lVWJt$%FDxVPaAJANzg{W}*C
zdk~n}-7?kDC?90Xf1e5D4?-*h|9R$!O!@6%EI;Fm23KChCWdbzXrtCaH1weBCME-7
z-#T3fzHw$0vl4y#<oS1MrA9RB7Y9-Nf`?9@0kc1FoSDKyeE-t<=ifSo=8~%+>Ua-f
z<}k1^#Bzw0fb#4YVE4_hdtkX>y5x2fv01_lWg=B|(2JLV>in$@;|+<nFBtz6+ke}5
zjx1pYQ9Jl>zC5WY5Lut7)K64)&dt;Pk?hgCih}2H7fl`m0OP&6+TtMH1j!<b){vGl
zl8FmAX!<{6kSztF%R$fu1d0%Kb#+eLJ3a$xG+Y^-XT_$I(x?>DJb1uwo8gKKjHil|
zPvztQFgyb)K-ikV2>~j6YvVlH!r%8rlM#IfmCcr(n4DbKMVp<O!Q#;YNX;a@(4M-x
zx7@_75Ler5re3mS#8f(~oHZIK0YlasBC8%?i3y6=y`lkJjmEY4X(mkfi=9qe6ng@f
z27om9`T47k#{tdhlN{b4kx~65TqpL^`~Ecz?QIP}2<^2KV`24x_%9@W=lsy9`~WB&
zWk0(KHylW?lODV@I5Y(3fFsr8aThf1@6t~o(y4G$z`hr&enwiSZTTb+mukD|QqG-H
zz6KEW5f`UESPLMPqe8>h1)M`W*G^PFaH+n9C_e2$jP0bS+YOxclDTIcZ6bb6B}Yj<
z>o>z;dMToX)Ve^2;xp>_^m}-F1Eluji?aR@?P7ohL58L6D!JKE!s$|cI%?@Rf%OgY
z@}L8BK^sQ1+R>y10@y`H1LlCUgj^jsDXa|S)dTT!rHi`#e8TGLs?FHjB1pAhwN>P_
zy_9%Yfrlfm<9x7gI_^%mo76rP6}`;?Ov#{w6bp1PhrfAF20&J5W*kE1*YA*f906Jq
z5f>&ZV`m{xk-XAyZg$4FP$EXA9&Op$y>mO{6Om;MU$7IyyYi1|!bI*PKb^?`nkEII
ztP^+2xy_u;6Sv0;iw9T{@A<Eu?~eBL%iCjNUvv=XBA5fFLnpxjDcdNbf6WxJAL=bK
zo4@oE{*TSOugLaNz=rY9wAf6e*KT_3b~f~<TjoA#a`mPERCv?U&$>_f+&E63po&Q8
zHK$aF`#b+&x{a_l-Jk!@6b(`v%A$Fa@NWww=93<7Eo+>AXq*S3wHFf_?&EWlQ&E2)
zI&N!#=6vQIiog_U0hKm+yzIO@Z(rY9m*5aMF61SBJ6{wI-HDa}dZOfqSyJN?95p*?
zF8$@ak=aYT`}^fU0I1EL3+xh*(QBK{u%Y#?j8f@*xzKw(UdgaNe|u)gjPvRFit<7}
zT1x$z^!WroCIQp{6J*QT*z9Cy)62<Gzq)9f?IWmTu#*|6f3^r5lhpa$RRIGoeem+-
zOJN_}$DoD*JgsX7lYEQ((pzb1s=<FBWvpRXg=7z))tv&A908M_hP5V~QNIATA7<zb
zAc1T<lo3me_}{TRCYYR>;^O24lHi*4jwL}(Fd^rsd-wUz+puf;@ZrOLSY!Z$z0v>L
ze_&{6ijTVLAbo%M(LX&b6fy&~hc*D|58IO`C_4rET{^&X1_ELLF@#TgRKyi*hMd<j
zBpPx_07xzQJrU%y0KVV=C~-*4G3iSXTpO=QTBDkuWcEE@A@G3j-^$F)cFLC(1JaCj
zFWRmWuUWoZ3jKCI;4~z_Jv2X7Q2|MbfYX5f+yny2EY(r~BVwT`Q;re-XPO;~D2@^n
zGH7`KtHIt;B^-|jAXU>^RUQ*<l<>muf@2l!bGv8tB^3pLNfhYL0PdGfy?deI5mS%q
zeadsotBmB63IdE%1P_nio&*<B{-eSD28ukU^VhX)#^Cw@NlV6dQnDRDf(nRGK%oIy
z&s8*K9J*yGCvUmG-Qfn+C<ZuT?MM-<eSl4A3Pd70Olt9$ALEQYUms|00D*&zg+-dt
zO0~<$#}0Aj&|3`W`^gXz+OZ4@^<mwen&<cLaE$YDDyG=>tB<O`XZ>=|8YR5H(w?gy
z;8QHl$VB;1$lV_1k;t7_)7d*M%@7u7qPLsBE+n`hXZXqK$`+xNtzFKMX{bT%U91G$
z9XlHvsNv21*crV6NuaLR$15YpB{Hpl|2eNfW$vfo;F*buUug<-uu-Ajee%7u6lgT2
z_!fx>0vojxg3QlHbDW1a9l18OqqbHk9o_vF0X8P4D{@9C#U-s1uafg`7EbUn!c_(E
zIqC8lgRpk4#cnaogstyMotvA+0Hi6ct*+Wme->F<0`!kzk>j2@r```EjwCKMslRPb
z<w{Onfb_sRe`0l@jR6vu^27CJxcW`*WRN1-Kob8`wmt8I2`AqYCPR+Pmi_9GX@&8x
zI5!}_2B0b2i=$&=)CVwry0AX0-&)UU{yXvn;|mK5)6<)POmk9xV%s5R$MO9G_cO`U
zV&oh;R{s2Sk#8!wY}J5zzLdqvxgeWHpg4Gc>W6a-VhyFH*x73mX+ELl*G&3!cD9Qu
zBZxah|9WoLhvFdMr|z&^kW07++a5Uu#S~v!HLKLyC%lPgi{Wl4EMDm#1gg#6IXVi0
zRM=?Bcf+rc>q;W-I;UIoaRKsOLc*1Ym#nR=p`P<g)b47Xf0v$cKulJs=s>a}jKSv}
zRLZf?4Hef@6H-h;&bY|AOSdL29YgcG=$dCVKe_c!7u)R4040soQvLwhe|Cph%Yl2u
zVtwa@!U1&s_`de+_27S2d2)ilZoZ+v(+VzIS~w94>7P*knjl=-EVlv$nwHPI!l>E>
zvTC@wxgjOO>ZD?vk6^sIy2;1vyhIKCbfCimIfh~94*7wgbkt$;^TnLT7o>4Emj*Tf
zB@YG^Om%_NR=k0-n@7l2_3L?;;k+!d6iq)sC1iVX{mxI9A}{X9ndRTie`kZRhwz-v
zSovGqj*oA6zLwufKc9Vv;@rFAmsnX@VS!eG6-fSea6<9%qrc}jbUZwLc1laY3PSGO
z4^VpqvbdJLxI(csW^>EJ^X;Oe1(bA_nvjU{@yM@I^8{6Mg?S*lHCq}@`HJl-cJ5YG
zuZ{Qlp)orORGnTbDk{QxYAt%Q`Ib;nKQSI&F=#|U77U!ONz{9JxIWAAo*&ytM@L5?
zyYwN<T0;PYYdg9^S{dNXfj1<zmdE*XNJuYC07%>8+-PCcBmtOqnKKY`Lv`OS7xcYU
zRjoMQH{?`&b#vf;mZ!qsg=yT6V`VpQ9j@%Nr*0MX5D~Vw`}wJ<j~+d$<=U;qQRf6d
z(5x{c$(`rqrP0e$>kK#xLZ&bF(AwG>q~HJo%{E(ems$9&o&Vn>CL6eQz#hR_L$<A}
zrR8x1q(s`z$rwaS!^ImB%!VyPF!(^tj6ndrHx(Og9JKWaE!F1c8)))3e`gx_%8S8G
z<n#cudJ-5CLhfh-xxE+%K=4^}3{>Ee+ZvEdORN6T)fLk6c#UXOOEy1?5akdu0xvkg
zH$%4HL7r?3aAIXSX9ET~!7?{54{Mj>Nk%AY;%E~x>Glx#0ZREkX|qGamg;Yqj=CbZ
zX6ZWZ&nJ;>-?#uU&S!@gOa@&bi<;kgcDbXY12!FxTWDzhkfy2P_r}5k5a`PKXb+E(
zAOrmEGMXz&R%jb&m}gdxJ!bzl-X(CQX!U~UQ6V@?Owb{F3s|a}oXTEIw9;M&Wj$#~
zDE?XnhYrZUqUGj3F8xz(VQGoEuqWa#Q{@5xejq?%HH9-kF=Vi63M}dOWk%10i4eb)
zO-Hv=&97n$OSBp=qvqk_TPoufS<+a}@<yqtvH*P*GHyN|kU^ZNah|FK%1uPy&MpJ`
z^2b^ofhMR8s`k@TQi>Y=o}RuoR99O&QHdh(OX3lA!iK`%8A9ga<Ad-=At?AE$>VUD
z;y{evYC-lD*Vhc)n78sqVUiboJw4x8TOTg~k{ygH=N0>ma6=9_G(mhA-MbR@!O`(o
zYZ`W|lP210l;^WL*3HAi10%))$78HyiJ}pX)wY-dG@C64lF?7L<Zf$Ln`ft{LPEeH
zBtSqWB#kHrQM!hFj2xUX&sQF0aHNTvnVES;!&&{MdN+9^4Sml>P-f%#0n~K55|EB?
zVHxL()atMvpx`?@;J7%@14z958V*h<r1sR}Ou_L9M(xMJO6gkcvs@baw<e)ymV0$)
z9Kh<PX7vd$sX)#98rmS~fEz5ge~+gE4)iQ6y-C+}b$<iP)d406uor>;Y7y53PDLlq
zJOL#u;j?d&ICOjn4OIZJCz%8iS_m=!2YzrioSu`Z8Evx(Py9nRLxmY{oHuO&6og^H
z+X6Bu>jT7ss3RDg%L#Y>(>p0>uqgoF56pL<H_rx+B%2W{1xTKYj*5aQ{}I_ZtYB?n
z9K1oNciKvr%&<WNWxg5YJg7Uax%J<<dv_$~bv+Cx6g?Y}?i|a_6?3|L-&(FbJahy`
z)Q#VBZpb`)(|0zG0UrzrILCYThPE9cxCJU}yEc9JWT&T$L9VbbUrcm$ZNQV8a|C(l
z$}0t=hYz=bfPxThklsnu;3VU$8UzC_lQwV6%~5XN#2i9UlYwNA5|^g`D%=Ji5aqu@
z!IZf~_s=2l)*@deq@){{OKAWRNnV}>;JZJbvI%6s6=`^wYh^7KYpqF5+bLh{bX^eM
z9)~*~4J$II`E<h#y0TRO)mQAzM$Qgho8%kPS!B=H(K*|heU~c5ZV;s-k@j@k23xWy
z#Gm%5i$iv7>^*XF9dIB6OZ48}UhsVw!^e)6ihGCWSLKgev8ky*SY!cJ8a9H=>3N{<
zf?T;87vXVBzh0vP(v;I7HF<#sS=YsoNQk(I0y93SY(WYPTuFe>lf`~hk5dob-ZbPW
z@N7$EjYWoQxtmE#=x53{FdjET$RYx_!_W2p$g-pJQ`t%-5JAF8D=aJw<G{k)9I@=t
zbLw4Q*On2Wn;(`tJ8k*S@(D>nc3M8zHO2v?!NO4hR=TxVzA&>F;Z85l>?hvtMK+`*
z0WA6toT0y82BIcTI8YBoA`zuOeXeDY(btfAw%UH&KoB*zQh0m0eeF;OCYF{YA8YE5
zF)_qM`F17yjqh>J4!^7s$iRDKRa8Op@FC0*&Kq$+aDsWd#loMcX>GGVG*UA4>;ybg
zO*-xFc!w8qJx!nBgSK@`a<hV|XC3$OR5TB=Y8Xu+Nqq(EDO0&%0XzVRR1r=+Ce&|;
zOVZiO#Oi}w`==9F$bj`_BNP~<)72}E5vO~&hr^CE!FS2LB8$*Wjl>$xMI-l!fWrp5
zg2Qh8Tgu!Z18h`~OMpG3>E9m{5#fdG`%yRQiCLrs9O%$X`8Xd+P&ncc6Vnx5x_FVM
zFXS_+=VBx79Z_bu(Nr@s!TIfdzwbDooLnt_M{~2d5hNrf3VCDCI&=V!fSN6=3D{`(
zpTa4+a}F^7K3vK1r;J1vkpv1EL4ZT=DYF2@1b%@lSFYH0ywk#(BaUTL|Jdw{!Qm#c
zq;ZKTNZSr5D-h8J=mwe{LQMWBCjHLkXGmzH!lSn+yISo|k7(H!(Bzd@$BjuRyn4Z*
zOL3dyuP?yPBy?RoU+lC+&+N^>ivT;Ua*1V1{|fA~Nzy-Egvr^SRa|n5JAKE*g$UG)
z$L4;;Lnbe2A$X72N-$SrE9d9wOrOfax2|dyah;x93iBTEEf<KABRUr08Gjw8msrSc
z-XY`*N8NkYDq9*lI*FyF2VgoJRtw#*&_G-WXbNT~YrSpJTfm>fExjLbb}mla*3i>S
z2F_v-`G89acvUR=RBb>(!;=n(%nBf#iSyb;ImQu%{otar_NpN~Rgi8U2iFuJq{w;6
zDjYHc+v4^6V9+}-e{9h&b6;O@adLjkNPc?KPVaLO51~l`OM)Hn<S=7;dKzvJlHm+j
z&^~Z`39-H&j7o<tfL<3TB4jq`z9Y#gHE06|apo1CYuJYY-N-D5wdeI++|qy*0t73X
zJIfd0{sWXGoY%0426u5{$VksGUbZ0e7-?F35Sk980dRb?MW-=hPqKPRmxh0O#bTM{
zgQypnX~?!JIPrjjiSt_R`~+Q=ufkaq@z7B57;1iCs@czzab6?qx8pOccpf5m+8Co6
z7i+lGxnkTLCU2*+zJv0&iHIB0=^tve41E<uL$mIjHff6}9l~5qjU1I;yF!TT3L;EV
zN1FDtTPqkdJms|A?+`A`aN-OI>72fF`fn+(N9D_*-%ic(^cUAlk_(hb)SF=3!x8>o
zJ{Jbdr1fWd-0Uyl-iUC)2Idt<cq_az|00a}_e0Z*)q9dx8>Yc?i@2rgN#vSmT5be_
ze_D|5{8ueHY@bt@mB<p-xbMRtAigCPIZ+`J^qCJu|8LQGc(6gZTjiQQ$nuju|6D#E
zYJy?wkD2s~55AsVsF?H^%kHF?mG~=&oL{<b4c=5B?NT+|^3J$@<Gl38krQ*;M=P$P
zj=5jXE;${^t>aGo%Xv!j^na^xKa#IU;!;tgjQ`rg!}AyUBJXpJbq}kIMps3AI)5{^
z72Qm&bi;<-#b4@gWef1QpQPVJn|{mxx2ISHz8Wk{&^Z&d5x604o_t)mgoI0KH=D0D
z<^mtkWYMTSXkyu&t_2?>?=z2C{{FL`s3rRp`Uv-z3!Ds{$KNA!AGr(<zuY)R_+?;G
zI9@IqBLc!3NPC&$8#4L$S}5irXL#*mJmL$tx!B%J)&$&e?e2yRA&SR&lv5T2gMh%K
z%B67RrkZXyiy@otM`#6o95U#oAZ99@S9NrCBW?B%4qzo1HTPlD0J%!rj@ifw7lEdq
z#*W_+{Y4*w>~>sZ|DxXvSm@z8@$C6?Yz`b+<?bX&;^xPYybN0;_3~{<0%U<O;%(<u
zV1iSlj)qiy!xs}lkGF@*43AiG%527;c*<!l@h<iHeHw=^!*Q6VIk?x8Ul%ChGW!(^
zjg)}Rj`je22D2+GYJ-yClSqe*tyW^!j~}TLNJ64eWZAU+k2;wXGW<E|#K`l#eC~RT
zq~2C*ZUsr;HO(VMeXiXRn=vlL{Rhr>cOg$Rf5~~Z64`CuR<-+H7G`usDqQKZe1C6B
zsl}35)7LD<+1c66`c_p{i4fUMd()Mp55Y+Q7=o^$qGAt$Mi*=wp>*nXE_M1JlW8&<
zsq}we8Ks2q!d;^=7nt?x;64}e!Tkkohg@1^ZdORR{{@UcIB@EMixn^+)!ibxfHtVb
zxFp`gBl>2Ef$H?-VVCgmaPU)<mzP65sn;Tw;nD&+L&SIQzMT^J=~6g&^H|g%uSfbN
z>e&keE+<5sdeyC~hCcQ?%6nCEr<Q`4U>!PPSSf{Dd<Zi;o$AIPAv(sD{MTN8w$ihr
zVEvs!C3r1Voc<{1lf`%NW**5_mkvf|EPvu@Ooz{frMe2Ypty+e%X^gYGhZ-FDLB`~
z2X~@ovQ+1vP3vx*{>V6WW1U-l_p}F3eH%;$j8P5U5f05a;-v0(o#G5&#b6Pr!pb;$
zVkhnKfJ6X30r5^P!CM1}Ihw~2EW+9xnt6xu>}iHLVWX=;jBo7)iybl?8N3;F!%joD
zpKdejG1}Ipz#6C0_7zdjvqA=Mx_1^}kFJNl^7?QSstQqFcvOq=Wx0tx1Fw^1p5q$+
zBQ2;LtX#s~9<BEdcVvQ`E{%aNKvQ@7bR!e)4r|S36>|37+=%yrmG3k{<db|GG3}YO
z+=p9%d;K%R@?U&}ZQVQ$i51UR{MBi^ztzyc$|wD0yC1tase&*x{`OfBEj|}uHQ5~6
zg=5ri)*I}P691qrYCS9BZVf&f`-5=Pc-%d$2E7g=gk7I0h)$tBR3OzvMl=Q8r~hX7
zl|%@^(f*j}(UtS<hgixzqRRp-)t|+cC7p{8gxT4ikyxKbtSaBN#$x@DSlBOJDijxO
z>-t-+2II@mn;fjnn*rr0GUjCemJC;H_?EB2JaI{TsONbtG#LeT4XdrGY>pq%ab+kU
z6o~%q)dV~fKY{tdcgE(6%(?ZH1uw|Xzw6P89*}u!;J$akgrt=Go*3odp=7GA_&uj>
zJoh|dz(8O4dxR1%OnlffK0x0-5{bl&`<~uSCulJ!mf6<|CIwp*z*Ec&i*27BEP0qz
z|F)q#kY+xw?@e5XFIo}A8Kb4o%i9&yEX97nu5~(88sGbC4b1w4jBR6&FC2n<D|j_?
z=ik1@;x+%WEf>~VrLr<XN5`Ha|7A##{-4p5#r4t|w0}-;JGpx&pl(@aJ5jIbNFMt?
zuS?pO>p%ZCsO@rGS$#qz6_*EwGl<v-BP!oHFDBD02fjGk>O^K~fH<(H0@FW#)Ju-g
zm-J$ph;_R1$DO4qkE92swDL|@TmCuXCKy?W@rqi<C|fL(G9}j>g)>2B$PQ&<&A$)G
zcxbxqj^0n`!QUeywlhz5lJ-r<;OIhZx}7&aPT78BnF)Ik;(n{_8_tKIVBPNKKF?@<
zgwIj`YV~LH*8I)<Cl6k{+3p?_K4g#>h}k#A@%*=RdG~J-+X)6S#638kTbv2HHdH)9
zDN;%i!b=i}E&C}Wu1|B{GsDTAHRPbuz9z5|x~NA_66%8XF8!wzmg?fPK`%C|7PW*{
z{1MOgWoZa8ebm+SLKFkz38)-;?B7`iX>G}-L@%DC2QFrQx@r_yfhtoN(`SwX^*x=i
z33xt2^6&MV*_XZJ#VTmcqp>=FEKcVf1VMyq?Ita?dw)10CW|K(QOWx-VqP$@a5rdw
zfjgngzPNY2>dWyuM`e>2+mQr;5Xt?-)2SY$Lyw-KWOpD_Kpfxw%OPlcae9#YoBPoW
zE7}~zkgHnfx;BR9AYv{-kvZz0C6J{H>?pQ+CNB|a)z1WcnrgTQ@zN5?J(ibOqs}|^
zyKjzoDkX!A%3JxPO%Dd!(%~u<X9p=7T=k)>7GG(bmmO`YwLYiq=4HfvxV^;RRms#k
z6$HQqJI6G~dV>)k2x-0Erec&#F8ADbtGt0}u6<sIhdIA7BT$nJD{4~%1%hETmK~=f
zn!m9aYUMa#W=0WbT#f3R%rd)*+J%5=p_fW&hRVtq`4zG&Mczqy&X#<$pS<)qvw4k(
zSk=!Ic`+-+L_(T`v4GR_^vrh8TA)TPHte~!p-S64#e|2|<fpsl^;`RPY%s2-Y=As*
zv=-cB79O#F?%h^+v>|)i-GD_HkuWHyBvs|X(iQ&h6pV21ZcwLzUvA`Z`V~j;4(j}3
zeZiW6S%mdOs&83&6T-F;H~?01+A+(`+}X(FW88#t=cjf~p`BgoCM&cYhvc3oUdy-N
zdnqa$qkqkO62J27SB>UI>gfp(0-7vDrRfWy&#?@vt*$RfN-fm82gB%zur8)GC^shz
zQg5PhZjU_hjwU9Q))7giZci?KWI&OZFWz^$!b9-Q#kbx`h@n=?P}A-qwx>mWHyRo>
z-SRJ<v0ra~4C^Jx9<k2*^_Yj=S`WOiPVErvkGdGQ=_6D0rK{?ZA%9jM*4Yk{@X2#T
z=d)LD0KXl-BlTqW9}?Ld9Y;HVCr_mNA{*YBnnD9QzK7WroLcTq%;pnKaOq1(lB*!y
z;!Slzl~gV-CnrM~?0YykMUpBoL^t;EcNRQrT*y_llK%!#`1J+uqxjE1SQ?l6<`=~d
z5(4s)Ec=xAzWVM{@Tm=Fz;V{`ZsH+sLe`^NG$M&E-*EF**};o}r;9&8DALIDeqx5k
z-+kU!7~Q4r5^W<LZ7=aQOV5fDJHX#4;EyT=t14NKB6-hCa%L4;E5)Qgj#Rv%vo#ts
zQ{#>~_*-lV`=oQBNjH9ci}_7bX6Jh&5(W+txE#XfH^#}Dr&#b1nhbfMe~gjgWX0xe
zp1fXnq1HH*!7Ot4ZM@SL|JB4P^OQkTX#u*Uf`Lq@{p&pKOC6Gz?H6MaLD14W3x|4E
zLQy5$e79(FlN7DUf9oq<ArEfeLM$wZuLv#$7?e1UU^)vq6I)1n&BlC-Sudd%hF#Uk
zTF{O|(M~&y;lYavG5>6MyKpmP(Lq!$<mq!z{-eJE2$elg<J?b6-{0>0huz(m={}K<
zjSMVM#o4O&Z|ud$oWuWN@=m;J^88!YOA6XY0-^k|^^Xi;Ie51b@nY08XCI?;sYCSs
ziOPHKyZ3g+sBAX6##W39g!}YbwF5>lQoba0=Qn&)Wtr1Ox!%KFD#N387^IZg<+q4d
zFc);3Jz2?0w%FV$Fkd3wK5O79B$koi*HjJ59o&~Gwh2gRLX@>SS8tA(RJ|bxv>fx4
z`=*qr{&Aj#(3w|V2ZaZ9;A?5Fdqrk<Y43u<XDqRdf|m-Z=xO~hLvEy5Z}*_h6$$Yu
zrfb!rm6xkKe`@@?AK!YdkJyec|F{$`(kIq1jVrd^%c^F54&N6V4D?bvz^VY)&^H9Y
zYJ{pG)XG@(d#Vb}0~yxA70eNn#XFW=wA=|Qf84K)ygT8F8#1fn8}XrA>80J1eP`05
z7HR2Vt(+CP+4i8?OT8+4QXv{^-#V{nzjJG)VE+|%yH^aBzlK`17c!>!xTH6c`k^pm
zb@1}_JveKf0+L1+x&kd56JK_PX1-xB(Q5sZfYR<&@J+Jnq&IuAPg1G`*iiFs{g70x
z{tG~@tC3g7c_t}I3`)<J&^YPAe>Xj?1=|h+&4=6M0!%lyTRjR0uQ0GZAwVJL08e^9
zWQc^(F6j_a2uVSh93nhs;+!DvUELyk=z-x-OClNoo+@@(`Z!<k{7ax>NeBzq0r#KL
z5?ukrUMCHI0?dO}F?&43wxhE0Yj*ahoRp{L=H@`9DUOQO0!MIgM+m@AKrK>Y^XEA#
zmQO{+iHszd_G@R#S3eMlm2!l?0C-na2P?O?w?T(9<5&Zz&+QUOIK$vbkBrp-bX<$w
zDNu@pprGsS*n2{%Fhp2)vDX)V0+uYV8R-H4P;Kh({flsG@Zw3bhnNH4R#XNnp=b|)
z7@W8Ob!@LoI`aFidy9O%F7mCFF31q)k9~Mj`bZ6^h)h?XBWHO}@<Fr?9c{2*T}-a0
zbfu)Zt4qcR`SN8ZP{6#J2;op=xSSvd-2&8VJ+ZSRlI%TyuI;d0)(7q=i4gdNyW#YU
zO3uv=0f=$HPavOJTU(p3R6jQE8Yor^l&L_NGz#uz@H(3g<j!0bv3GDl4he_$A@jQl
zEMc`H1T$!q_@Tdm#0vx;Sk03)TuEw5QkLq0#-mqR-pwlxc#f582_pVCXmhYTL~(#=
z0+c^6gaB#U6bv36J9~R@_U7-lP5w4i{ciR?)`S*lJwj|Hzi{dgfCO$20EfX-4WzvR
zof$%c%9DaM&7pMOM?jQCSI$UB=bwJe6_E(7@Z=PKDXnry?XiX}pr%=oY%jG7*DR7J
zH^7?S0tk21evuYDG5~ABc*=g?J{COhV0D1(H3BZ0dLO*kqjZTmI;{~Q?SH^+9h`1L
z3->_apnhC`03P5l8y#)!0^^}=0PZ>hfC+fgf%=H|MbW3#RkP-&J4*vyZEeq&ZXW=K
zs~=dIn~MN`(YT-<5E;1$!R*ONV8jcal-v+3d|>c+RI>qCGbpZ(mN_>CQN}F@O?VH+
zFdJfDh+u>us5RJ7KX&^=!Ph|ejV^9ycC>G`rqy2-9S|V>!`{kbsmcY=z7Jx|x5i@>
z27qcB2s;=|5q@ysd%Txyt_k8wa7~S|Q?Y%)#27xzuch={LIRTj;#;&Ue6OB7?w>V&
zyrB8xwG+#uV;giix<a8758c_}o6WM8J<$#;#r5mjAbjg|a%7`PX$o|f5OHY9@<T!M
znwfy%2TfWCZCy{|<ophdZh*F)Y+nPS$REIFv|+8Gp#hHF>#0$5K+OfQL!-6WxoHS(
zL8#!RoLp70ThMB(;)<vIV^eRbKV3=wDRYxlanrQ*9_wTI`geYL<~T0p)YGVQA8mGl
z50yM*eHNl*fJcwqdy-S-0x+u`#&%AUFO1&5Zr!*+5~Jd6HVjjB3()aix)l&dVnU2v
z!3^0PxX>Y*qx5L#M$sq0)Bv)4Bhuvqg0~+cR+@uA+(YcR%Gd1ly?_X_{c+7x?@AVD
zO5qX5XCn}4xdo27brL|*-8p$6$O<$xFfmlhffg^hw3H(!h#FL#O+geFOsPd3J8Mjm
zd(YMI_Lw=%^@D>0Z*5!96YnpF7+h^?O&JeFjz&kn++Yeu28nE^H5`806YKy|$cSDA
z#}0Yi`f4n$&g0IIfa}+;p<gnkvghFCyq<H|0N51QiyqG{&B2#LqrN@&tGc<_Td~O>
z=wn&+JEtL-<-9in$TJ{dF^mv?F(U{jfeM0DEJRMbuQs+8o(c8AY?Zk7J48lNu^$AW
zflB6Z#s&aE;Hwl8o0;HyKZkc_32{TM1e)u#^~2D;;F}N(jfhZ=7rwQzM*v=sOy~}2
zX^fnpJ8|eWW~1X9lt5ujP;{A!r>1KPeB3X~Y^98xVKqZ&w*Ns8HG!Ox5O+}&1Sb&w
z+kBj;dVgZcrJ3J+IX7`!j7PEKk0C|C(-(N`q}D^)vptp}&miuOL>T>77#Se#3bg$~
zZ@r{=k&*&Ph-<n%Kh*<joiiNZ>z>O{84xfZ0xy;laQ-Qc@B^0&6YM3hihDiR>rEd3
zYcU*+iHaHqDHSxHLo>h$LPchb2i7$Jt_NJ9l;yt@Le=&m+*q6M?TmT5bT3GMH`H;x
zks?l@#)1=|06DLYWfO7rKWF2bKAwhvI4skQ#D`7PSsvIJZgd7QeF?MxKv`a<{X*nm
zXZI$Rckgpx;MAJ)<FQH7M~^@Y&+dV<6e`5NFIyQlHZ-88Ex|lE1PID46cnoCjBnpa
zb7U~D<b)zeB+{yuFflV1?6yJht+=$*yPy%&R+T)5sXzijsZ5cEH36$xV2x7?00@&x
zF*6!c>j1opGZS34XnOR&t*(E6p?Dj(ufWGR4ZrxP38@jVJfte+C4%oA0i&{<qX$NR
zjG~F-q{~;Ynr{gqAtfIkZgxply+z|({fP)ol>tJ92asl9Dc7-%Bi-+^n`?nT+!F^0
ziHfnZ60r$Fg@Z{`c8@nVEhlOw9bmr=0jrs$zX2-rWGk({$QO`e*C8%%hK%e^pq>s~
zeQU6ArYjeuLNy>l!M#66;(e>O4#Fi<b8~tMf)D};p;6MLD#1yzRa_(4@}3(hWtF;w
z#H%gP#RP#@2!n54t)GUV02l&<wgE~2!HeD5NYk4AHDA_IdgmQPJRQW?HaC5IeSJMV
z&<SQ$oCrbj78rdISuGm35rp{0&|D+X^6(TI_WgrXsQ2r^8;P~j%I9J9d=QqliNkxW
z>AE_ieta*>`yvuA83iMw3>)Z!zF}pX%_tg81zx!|E^wsY;Q_QQZ6ScK9)-9AOPW1I
z|7Jci+oB^OrC$`10(Ce9AeDe{qgzJO3bbT)zX0y(8d_TnkGc^KCZ^-@-fzHy-f@66
ziWy>s8@CqM?ja#5UV$A6SkkRMWw^YjW@mfBt2pTZAY0e1va&K72c~D<KEO3=p##`D
zATM@-UjC`Py0e2L?EnkD_!7>IAFvzcJceA3H@pyWnNc}8`s17?n}UIZv-4XdlG831
z^e;N%`kmmc_2<Wq8pc(mcVcY_T+KWJN(jVIFrOij2Y8YbVAVbWz+J5kpcudrjJw?M
znFK;MfJQO1v?K&=(^^kh2XS~3QWt=S4B=S@HNTA_S6(=U03kG`6aAAX5m8_V<_|(l
zVgH4u29FS5>w<W8o1<`Y%A5c?5V_Jji>F^686GybuxK5_@Gb{8wxUxu98~~f$Lvf^
zNB4*4WKR;Tqvja)o<K$<Bw2INRt08qNJzpI+S{<HeIn(!(;oA^y}jMv)AP1>ImA#I
zF8`ek|Lgzh$xqOP1O&hR4@qcDj~V|>LOW+HVq~Ic{J+X+pLxk1s8W2p?Y9E`OXR9a
z?o>q?A!)|c!Is?I3}2%^<nrUs4tGY7@D--+mN7M6#bo#A9osGWg;a~?Ot?GI+`Ko`
z-2O+5m~caw&2D%5fbS;`;V~8=k+OvuKYd!w#L!N4mPlKITOXcUghj&m%y^!=&Bngy
zbu^NiD{7NlfnAgRzAmAT?tNGE=M;4P*@(YTUZ=Z2cf;~S5U2EqHy3TMZ|x5isjIP}
zJR{40&KN60+Wqqvl5@fX=S!BiR(rX*uP76a(sKK^x1{+@<@>&FqPaWC`qX=B=}wzh
z`xw<*oS3=_tw~kNFky~1ub}+2JQQPekr?~7&+;L$+)@oYuI6Da&yp>U68&ZxT4!?h
ze~4jwZDv&kf7ZV&hN1h?%*TTzHDZo4)iIu+=uVrbMZ23!E|s)+H{HmXQ|z!7l?5`v
z)Nl)<@Xu*GYP_9XTc-dDePyH1UltzVC=bLd<USK5V7KF_Ewqj<u=Lt^*w_B2j+m3N
zQmG*SV|r)H-ivA3t9yEzl~?vK_47q4DC}ql`9l>&Sy~*6`ZBRXQevsbv~Z?%y`PH?
z#aaE@l|EE6mg#LIwb>c&{pA*zrN*nC<YD{g!G5PuaM+d;*7j2?<WlE0kDniWW6{UT
zKUIvD|2VWk!HvhfQ%QWhcSPp&HCO{>vi|xk(&X{s&cse=g-oi)lc+agRT75~z}c&R
z&$;JihR^ohRmHGn9HX*Qa$n<vPG|26`egL$n^l{?ZRoI5D9vabg_0ZF!yhQ3BjG&D
zr`yhKQOgOr5RhC}o#0`)@+61Z{iWu$$}m)_oA>DrWG*b}E3-8Q*`?x>2#NmWdqj5o
zfb(to9L^5oCLU#7{>zjU;bpvBf;J|NzB}5)lubJ(&-SoYFanh>1V$ywOgcrFP|+{9
zFj^5%QVsDkNaZQvC>=j~Y%@P;e7$qK6T9P;vg>`K%c65+qThmgF!t5nzmePTB<d_R
znop~Hm1i)K)vBC=S%D{>U}B^mix<=rzr^A<{Y|3LqQLa^>j<%wyl>AqL}%(g&;`eo
zeVaf%$<y9f`hHQQB#f6Xuvf9FaF6Ute7C|kjVKiiP8|wmLjL_DtxI%deZ{ul%}X#3
zZi|L^u_dS6+kD&VQKtV|T3;uJ(^S#|FN<qe&PVNOlz_VJO#-AmT<->w0diTtH14|2
zUxmTwZ{O!VH%{YY?+t!MT`xo3z>r17JbPmVt1CkSpXNb?oWXX@ulex1G{5riisf&9
z6YSThP_Ei|uFf%Jl=Mz*vusVmBc#$xm`mA9r@aNw?CB$!OZiJu>zf7`<a@^HpQGBQ
zlcMngt|L8Xpw6Poc9nUn`g<{ObaY;H@S#O_hRNW<Zbs|V__@;HfG^llyemq9LF(ks
zz7<F;(Cah2XfQAIok=vtETVC3#IXvDBHR3ag=?zGF>9A<Ca0fwn}wPnBz*_fynfbY
z?d#a8Y+BuxQs1{KW6eb62X?h8i^oVIzy6E{Gh3SUdxe{L$z9hk((Y~1AC}OLSnr=O
z%tiY1%@CV9V4cJ_$%%-^hG1orw1=mB?@9TFhYm|Dg~5%VaRuwF@1^l+>zZ};I_d5E
zsZe|d1ZnLsf0?)Fxb*f2I3nvOJ*jRIcsD4bJR*`7;-!>{Y<jpkW2$lim5f|QSsIzV
zIy2(#!uPi5Dy?{H)JM_x6m&^hgB}|sc%VD7f7j?UHKKJBf4U&uW-<98A9s&kydYT!
z|2|<;V5#HR6T^;s&dw`Lig+@1s}n=Cp`yRT-3$u)){@#I9|dyLFETSXr#9#2mRm?W
znGZ_s24G<~Q>$HR|B!E(fvbO!uP)b+<{DmKa0{MsRD5&N*FBuRXd_xO!rOz&U*#zL
zelmY__6;U5a^X}UD!kpKXgYD%i#Oax)O*a)xK&xd@{RH{>+xc&r4I5GIYO1nq(ON#
zgCqX6%F8jyzkEElOSnm>;=^wGbw)EzT4TH3yOr1e%9G1LDA-!?Re+>Tajg2v*F8VQ
z7xg2l+krlzqXt}>3Pu^XXT*=c2iblOdnR4F{xUX6$YxD!Zm{XGiecs(A#5{Ds%o99
zRWn#KUwh3mV!CumAIqX;$PGAmDLm(Idn5hD!2GFMV(Crx?#lIk8;RLyDnCVo1@DxU
zvAUHd*<eMZ2HjiP<qI~6V#&AI10|)IT4|>umswlV-LGR)^S{SZAAJ_ZF&I~!+_C*y
zhYUaJ6*?_5K4%-%A@&2@&xg|1LbqL}RYG+vE-BBy{@nAdA;zq^%;2rj(Ko6excCFi
zza9zh>2uiN)+5On5bLC3rx_JQsD$*YTQxjh7E!K?Wa|~c$_`8^82LH!oIQ`@MQ75=
zLld?D-djVjf|v8wF3xBhZXz2ed{Av1FWT1sggVcZx%cAAtn6Zfn^D$B#jQ^@HX;mp
zKO6*6X;HNl6by%!?+p@-_LWZ=B-98!k-L|>vgA&t(DGA{A#IVvqa2xgcbAj%b~5vy
zUMVhllWU6#&E0RfvMiK_>PH;XFV8jezO_%kEF|{A`aqHLE?!2ybI#YB@>=rwbsu@}
zTP=x(I!o`-bIvziy?;^W1@j%T0JMb4XP9JqsBd^5YQAEM?xt?MjEQHtIizD@U&>DN
z6_a5ZUB7?9q7>P3_*qi%R?6}U`szi}yQ)J~`71BcJ8sB!V+QKBxI4O;&nQ<4+hmZg
z6oy<O#y1;gzSq)d$#(VUw@1<{T2e$DIERMXxgj}kv{KMonkogKjSVIB4qhkckCml<
zh|)PrOCLqr&dQ?vS!>t#NzZrl7autee!glpm{wMt9+g5Zs8qWnz46*ldeG}GMkM9(
z?}R)Hk+wD!G5VzY#XFB<VxsiI7em{(=%Twrt^I`wZJN-PqkOOYytdt(!f(xL&+Shp
zrI7iHS#3GU_GMk~`j6|pdJXRfC9d=0>ZA{qD&Bj>{)*{ZX?0+hg7)<kuCByUw+RjL
zW^J3HZ$ZZ)-f2U+w%yeS#)1{vjbk+d@}VqljWJWg80=Q%?NeO_@<%y06SO!JNX7Xm
zIGD_q-v#PKX=2TUt(g22^yB5*9t<UUY~i5R^Zq*5oBMVJejgb|@tGSf+uy8gF^H{%
zPlcG*G6Xk0bo2QodiBpG2kU72QdQ1Na`D##ZS9pf(_V$LhJUIUYwP0+_-sB|I&2!y
z#_)TXao4_joOw+Zb*jHKDxR${&`!0{2A7Md3wKH<K5Mi<_MH|Imn8kW0HokH2*lK?
zIPK$=F(VngC|dQSj%vP(=20vvw~9<f)|Hu%J|S6Gda$NDo$Jmcn21&t^LY{V*vEzE
zqO*rB$1bCv(`DytV|NejqD(7^t856I>nhu;iBlapZ3lkd>#r4CyG)Y&-9awzRyIk|
z)hpMC{PXByb7PP7qbsoNTyp1qt_t;a2j6*F`=Oxpdd^G1N5chshFHNb8Hem!#ZJQF
z%%^ja7vQrOx8SqM{_xpuPg{;yZuo5X^|8AzoTE%HSixtz>nqzYSv{yZdM1q37a7(%
zUnJh`-l`<t{agRT_1{l~#C{xne?X?qC3!m|yl+MMnONM=&4R53rOQ{wXSd&NS1yE`
z4O~#B>Z(VgohZ4LueFMdM2lrgpw~Eu{4QjD<mw>Bs!(i#BkTK+n@;#0&-sl)vlO{_
zOfFx}BN!mT<a*Ebp~?3!$9~kh?sZSzYQVraPO5H!?2OkIwVA3FXL|R}FsZ3k23|y|
z5RM(`m;{zX*49OJ^E;(YZ!$?5=(u*Z^E9#7aI7tGw2qF)RN1)c2VB;q;?Vfl-p`|Y
zpuOg~SS>Pf$l{9TLd>`Nc~PI>C-JQEs>8JXfa{-EOB)@LVH|4NBk6|3^5sDVt`_~q
zW!!DW?|9y9n@oJyKkzO2Wj<RLwopeg)!LwqX67M5?8-*cceeo#+4=G;`kFPbF5%I&
z1@v-`><zZD4;-$4SXO;pV%#|eJ#6>ijDGy$e(Ob!`zlLXq#Fvi|Hh*^0o$PwnHXa<
zqj9vB9qDbuBa!iTDA?IULNc@R!z(_?m%9Y>OiL$aCHKR02Cs=dR+lL+b-wTRS~^U$
zt-bk%6#ne=QoG@LW2YS2g+Fl)>Sf{=HIm*@6pLwUEZeJ_1Ylllz0Cb)W7TfL8s*_K
zL!-L$l)H0fXhjV%-cR`%XZ^XIrjNwO7P}&4sks9EmmUdLMc`={Ufa0z4V(Cr5y@@m
zF4R(wA<>tVN@YPP9X~{|<qheBR%Yc0E;#UBGe?p(EnUw!B(f_D9Xn}2xj&Wojks%T
zpAf6TG_cuZiT1=Ff9iO-;COY#N1!B9GUc+X4afM<x<|<mQXb|8NuiM_{0H*Iw>?$b
zjwya5o>aUq;ZSU?sD6_2*`mn1NPyxex!SI-2x>RS8jkV1fU=S4x6N)9SJO}$<*u&n
zO7+q5)pP{Ji);?%wU5{vMb_VVN-<F%R&V(9fi-=*eh=sU9ftrr36@{9v&Aa@#v5$O
zxK2KAqAp@?y1aRmR>wZGD_U1s*H2qlEaPq5rP#pF&lx~tl3=E4e>qjS!6A1l=wsu-
z$%TNN;|u@O#f*Xef4i75Ff*Us%kKRjsAS)MY89*DMR)zPfq8T>e7R!py6Xj-KCB<K
z)8t<+b6pwOsZdl3L>J_gElEk4|KcbSQ?rTb_nPfxOG-_|w}-yOq-+|FZm(tqg-CVo
z;8WP&4PzD%X$WtP{ZJm4*c&cOI(=Dm(W^PTna}?H^X8kTWN6KAZ4*E7;#t_seWAEd
z(N8uyS{M4WIlgb7?c<n#aMiE=xwj}qNJ$YRcS^BE7T>9?a@X4z{m9B`H=G3K*~ATn
z?<NzAg@7uwDz!TK)Ghs5o=g-!!|2|DJZ8_@#_`W6q{j(~Hd$@-%~vgik)l<8FI1x?
z@2LeatKhI+hz}hQiaDt%tlqhGYe(1hf9gznwtqYG|46mMK+pOQ*9!d1{{NY5g^87(
zfdMg6|A%2(8nJQo-8tEH_I&F4G+F6Gcd}A^YN52UN=9euK%9YptgKv`P6jGdZ)uKr
zYqBz}Sc)`1NPKuI-hTE8J^bj0NzCGQsY%u!E1l<U#pZQW$L<96i(60n>@pUo?Ply$
zPmYlPS69IQ-~aub6lD6`iJcTnEcHQ|4?%)bV7Sx9@yBuVmk5A=CB?-Vm1n@C2(a%x
zkk^{yI&1Nqry@4e9D)S}fNS7SZhx)pi9rS#*}%+lr+5-n3B^KaD35I*#AQix_KGDh
z%?+f@H{cP6R3mU3;+y0_NGJ>3G!Sc;!KtMNcc<9M_wV09b)G6KOQp!K?aTx7PT+R)
zMWnY7$e>Zo1}oqlm4$@RQBXwyEcbA!jU=rdgm}Ou1Sn`3$~%OFG{>y$?8N{jHW12j
z=t?;IycxQt5)uS`DI|MAi`6378bWq}(@RW+9!O1j1qf>p^K(&VB_(xzbmjDwcP{Ag
z-I10rTth=cDx>D0^b{Txq?My7CME_B-CGtl`#XOmvo4-K{PVVl_Jw!c42+Bb>$`6n
z2EIkeX~`~kgOqO2DL1elfxM!wqO-bW)1B5OC@2Ux_nS+#0Xqq*UQr{~z%ZUPU<FzT
zVms%>`XP!n0llmBPI0<M6>VPLpOC&_q({ZTAZ?Na$|06?XJsyVAVx$V1O_6!Ga$z(
zx3~bjWhu)DwWgKx2Hiz;<Oq0B#qwtS2-z7ZZh@|+ny~>y(KhMNi*Z8~GYfAyIXSI|
z48`)ffYL`)K0D+F6%;^pq@(e?4L45P01#>5*KbxaIX>J2jv0%7NYt_j?giHHd9AYt
z`>*8CAa!&l#W<xl?%{LUTed|W0WD5&FCv)s1n=x2Yd$az%R8Wbii`WuwD<}N=68$V
z&)ZbgvNhg!j)O(a>7G3}O?X-w03?}kjKPDq+zA#J7<?5Z6C`*poj!B#_i3NS#m6(F
zMFT;tzrTNx_!PMRFm7=}YSJe7{}BI~`<-?eWFkj+7lJc1Phm<S^JO^wt156s&i}tY
z7dV@<RRs~vIqb7n+3ugV9W){J3ts}+KI8&-oR@0GQj?3sD+ycyol7O){!CIidu7}e
zQK>If^6AL~RzMr{zZ3-_)7i&nE5q#eSI?er`-rIBJhUx>wr^~Vhl?xv%Z1ZFz012F
zas|nLu+TVCrw{b>?vI?A3KE4NVuEc#=AEAP2ta*R)VUr%egj@uFvM#&Za{u6XiRND
z$kk?LxESIjkW~c%9f);VK}sCRqkvRK76i|Q+|Ro4r_O0hK%cUEsRUs0IpzD!bQMwc
z2HH(QU#rjRqQQvhh$IMN8En@9YE~=F9@4D7^1C@YI<_noLE<6cb%{8vY3`z)H7gQf
zfm|?GpXUh>v76TvCK7=c*nZ34nIU8|LbeYu86~8h75+)!v;vS!(zz4}6Rug*MaT>D
zvx?zJ33K$#v+_#sL0?qCG6acVW%LK&{>LC9ItJy?q;KC8H-sU|+qYB#PX6q@Y=m{1
zQ#?6)@B0I&xGQ9ldXD^5%OE2s2Tc=RquZKUibW<P2!WrVAVG>7z-K>eTB~ylh8QAJ
z&%SUf<?GD;0P&T?6x`dlUu}Fwux51)!ZL>U!Vnz%2z_HvH-NC>9g1>LL>L&Bm6h#=
ztTqr#rwPH-uDRS_XiOlEi>SkRh==yPi(&{e5bFZ~+g0%zh<Ber5EA0=WXF5sm6vqY
z^k`m*43$_JmOJjP!yiFgh>(S>*9s!}LWU7o__(-iuq;Dp%)wS6DKXLOR#&$3P2%|g
zpc~T(MPww}EcL%y_|e(fiGoe@FjT$FHm6mD8)UJIQ-E$G&y2>T5b`2YK)H8f93)wc
z0ggRj{DLM1Qhq$KL9`nH(8Pj5LU<vNrw>FEWAOdvVCtl@5zxF;sygjRRskAu)|y;S
zF=%Lo)6p0E1<Ls9J_x(^_bckEK>hC$g2M3Fn0hfjBLB;R1HyrTs3>?M3|VLdk2Px`
zAOOPA>E(dp9;i4uj^xIHwGI*12Pq80*=-2H0qYg@NkH)~LX4nhpUZ6gCqyD^ru$^W
zIczdhY#@~5`!W||Ni4*xh`286Rvm#Uy~k{5u>CS`lmOT{m+k5}=#DguJUkB1NGmF}
z`O|yzEl4>dvPhs+@git%Sa6c~LHwimCnSVv9L4EeKb-<EXWUIRmG((OW=4TtV1Ijg
zad|nzQsKpm7yd+C<cBd3{hw)%QMW^|k_`MvpZe}rhtcak$2(IL;RYapfnmRX^lp*B
zpodbLfFf8g&jCBPig0sv?I-(P%LV(2%4k!~e7mMoD`Z>Onui#I90eT%L#N&ZfaM#>
zYVnKK%PsIV+xv{Rz^b%&B?huzxjdNw!+HxECtDRYI0SDNWJ_}Z7SMZ=D+v2{A;V>e
zr_i<a(Y!(o&;P^So5y3le($4C?FN;iGL$HiWF8_zkrHJlGN)3Mh@ycannXgz%u0sL
z84}4*k(o&5A@i7I$dKW8-RiSH-}61^yw3l}KYPFS=6T-ldsypQ*SgkPPW##X{jyz5
zvZ9bVddX-A-3UtjGLS6I%EzI_q>`7{dy#juc6gq0fAY#$x5BkDdg-}Q+2*E=|3sQ{
zo=7&D$38fln3e}T&#lIotbThr+{B*51lYd1T16EVOe%^sOhH?*plgZ6#l<f%z`4>V
zPsHrh-D-n=nQ6=__T%R^%zPDFMEh>4?bby#kF9$@mHH2TjX&S>`X20&YM`)O`DXtm
z_mdw=TUL6lAho3mcg8nzI|{ahz82DM;UhyMCr9K_IVJk`;#G3497v8{=Fajl$ytbf
zE?wwDW?b9G3sMLhNF82~;UQlov@xE;eEZ6wUvSH?Exa`Ek(x|Q$cAgF*heNnzu-XW
zTaPh^qHpmC%W{meO12Y7Q4+i>kF3%jRMAGbp*HW@<lqU-LY&&D;c~_F_F9u0jTrp>
z^?>mB4VK}r@%7=33nfT`;U{yev0j#kuK#ZdR6?gR<JMicctk7Q;a6Z&)br=hvtEDJ
z{UCYl*s=cl<Zx3K6bv+u8oEYCMj`<p?~bw7RmvS`I~T{1_WP>No2#f~tJo6)K}_je
z2x@9;$$&2v`rDX#+vAO06us$H{}N^!f1y2YgxRTQrm11el`M{yUfSZh<Dg~kWgOGm
z1$4R#E&O!jds9@Kb7lYQe^NFboE!Z(+}7aEv1CLq-!4cES&_mk#_lhB?3j!7saRYz
zeO|}@&up)i&=SaxXPrC8usRGZv2d8ZEK(&>RpQ=t^Ve`}#%J7~@dvA^w>)FOB<3~j
zsTh-HNA-w~%R4a8Y<_wD<7a^hDR7Tk3A!5mEh%S@numqto-Nu?;nx!q>6s;!lIgpk
zmt`F_jbiU^%nLLb)IJQ~dHi=8YF!w%?GUpJ8T-Bm9r9qXQ0HfuYo8Tx7)EPmASN6K
z*R)wE?f#OtP<7Mwhhyw9mG@wdov1=NeZj%A#>ObnC_fNNMgU!W%D|(|tHZ#g)Nlxy
z({RkvEsOtznmm=xS2Lfkz$*9~ys~9Ex;+({VIlsYZACf1n#}9!b1rw>HRZG)$?7Zj
zg@t=-Z>PFP{aHQ$5~h2fOP?dvyi-2}dR>p(;_w;!#a>&S?xJy4lO0Zh5F)~6c4FYg
zyD=m|gYA^7P~O2tbJX7c-OMitQg|MTTD7%!6qWxuJ>s0kdX=ua{`+Szf)NP!NZ-4O
z8l_T}+vu6dyO4!6v*wA)Fex5SmkWpbU}~<1poYJ6s}U)O|2Pj@UCgq8QdpBWjjPr(
zof}HELaw@Y%=oaRjVHzel3p07(bSgS@M#qUXM<fI=uQ1&o<rGScD)syq=J&YF@|m5
z3YA1*lTw$cbl5b^kRkv9=zuibIF*;H|4jY);ZqKiUmX{xii5GJJX7m!2710ddw*iP
z2XCE@vhw4xQnK_sR=G3L)ngDv`K<1QiSI2^wWGRQ;psA#{`b(C6{M!R=`y4w3~$Pb
z)KkPcW>_RwI+hzHOYZLkRZMuqEuU{Jvt~{)Dws|(%o@Xr52}2`9NnFsnE{U)FKUh-
z@rIq$OW1kvSPnA1stp>@z!{4TwdG_7F{AAH+UlIAy+J24ESWjc7w53%NsBNBbIb@#
z)JxkfE-vm#j~C!<<KX94O}vdb1;^?3&qBte%(cvtnV9NcD%XsH#5jS4h2K9q{ZPff
zB~cjTy0YFBt`+^V01)77jDwwCFO=Q$_x7(g*3;hUFUMPE3c19lOOJ9-><~(5iA0Z_
zVl6+20ON8lad-D*K9J{CI*%eeW0hEz3m+3T(AFSYn}&+&Ri@cNyRls`I~V;>mnxxV
z)dJs-0e#82vvDn)kI=6~vR?)DQ@ir2Az1QDTzyyPrTQV8m!&yVi9ey{0j&Bp(rS=<
znRk`OJUqt9Bj)gL*YCw9Rej^;Jml!Oi1CZD0&uW3<&zTtf@Q16=cn3(%^WdcxYX{m
zR&?0<7PCLTuVs4*@75GvV?7DZw!?OcuV+hvigp`uoyvU|%MDPcO%n|rgQS3nv7Y1?
z!_+`DHOPqDA#JalWT~=vR9?RHgfsP4@1fAkz@I2!+-K3Co+BFs>7?N5oZsfjIi};A
zJ48dd=y-M2Jx&@Z43DEH1j?xaW)^|o3QrwjZ)}<i^$)v;j+hLH37Celt#h+bP+qzi
z2V=44gjVm42N<)>q7jYy&R>H&rgZ;6lBz?zLkErZ3Q~t9GjVxp+Z`&3i!Y!ff^8H=
zSjm@}haGx#@hi6!BU1|RJTRREr<li3)9ZLc2dK^%;uCaJ4kzIX__oXrLz)qWmZks-
z2_?5T?XB2YblSEn(?V*ak%Q`6(&@B(u!X4wv!;LzFV?>Fisj_u8CWx>@TI(D1d72E
zl`O`MP4O=bZ~j@+1GHp8Ex8z1i;0M6W%U$d3kuKd@eG0d!t>E1Vf_VQcP-CI#BK6B
z?tWW3D2|F2Uoy58wdOgV%IeA92^|?u4ud(=Wk<HB0xt~B*g;!?s(p0fc?sOqlcFDc
zJnD$tQD)B$_pXHYWx)FEb*YAEJ(dbMyM2G4jl<=_1$HJ3hWBIqfHUQLQs?p0GxQaS
zi$=B`EBAhPyJX0fff;f$*HSFHy1FijpVl$_v$8Ztd0}OnFMGdt_wcBDtn~iwN(i|O
zUOaG|?pU4GjSDnkc%dM5;;FB1cURYoUu{YPIlhU5&~kY1U-Z*4N{RNj-p241Lbz+w
zk)0N)fZzCj1{4IA7WVh_oB;GjB(eSA?E`!FemS(-blknGK)yO!Wm{I#mQr}x#(!c`
z#;Hewt#Y}Tl8;9{I-pz%1jH}My9m-jsB`R&tE}*s+zUsxUP!Sn{56f12)|&Rfy7F=
z8Dc!KMmXjlXOHg6{QT#_ep@;wEq6C>7y)`{12p>f?Hh*Jp^UE|?m^y-ta_#Q1&B^6
z*SYKZFW`3ZOtnEqgts#GF8+#^u;)`s0p5E0w5eb+JbpVlt8u^CBPgiu%b4~PWyX@!
zcX~fvfg0y^YQ)%)lXhjVcb_i`74P`QT@~szm4qmbM{f-#GvXcaxTOzuz{C8MD^~t(
zOv6|>Sf88+WQD_6TyOC8{Hit39WL~RJGPwtbDmh)iLtpRvG^_M1^^+HwvX`G+j2Ub
zW&a%RAQ6R;96)2eQ#$1a08?4BXs(r(T}6EP#&9L(W_VS1>DU%u01!2(mnLNN^@A+u
z!vh0jXXqy&IcqUa1UkPu+@O(k;+f1H)LT}{?oQlGlVr{L38&OG4`Byfl7WQN*1e`%
z7*rJ%i72_Ji!^%){Pd$WJI2mQ6%8n$MMK}7Kq$}r{9ZdQ-kw~stM@CSh?6(|yvb=g
zax?ggPOPf$JiG^;i*DYBv}y0;oOux+EKF7e^w(p_Lg$A*T`_iivOn*L>!!L$UM{gm
zI<BwJ7I~hn@F^7y_G`#3*aW?8rQd>>{p=j9PPs)j3b2&`j3KQui~?SVvi9=ttiet9
zZVfjG^|YGTzI-0024G<!l9k}NxG=U%>|$;4+}ftM*W&ec%kN+$WX-Ea47Ge|0=lTn
z`cLRa8Fw*(XQJm9%S<gOC;%QW{PYRO)*eJw=_Mg8_)QfRm9LI)Up4Kp4&B>|AY<HV
z@D&25T{jlCS-m@>k{GQLEZJXD56_xD#%%;)FCxs>Fi;kuCd_BhrYGp^Hnh~8UaUv2
zHuSV;f*v?!kH`ZzMu7!jA8EU>g71BC@zj>`^3J+P$GSoa^lJKFTml%}z2iDyTw&V1
z0P4Z>BG-H(P*00-<#82=v>P7$0os(_jb%Iid$<j+T&9nWjUDsZAi{)xa6j3Wdl--#
z{07)KYbshT+|r>H1NVTM%f7;&TuiC!D)Q`kXtHkg2Ig1wMH-fetRC66>loRY;MH-e
zu7<Jo<4lDOk55k0?{x&P6>(7NT+-Xm-=bq;c1T{mfMJDr8R2UJnGl)H4B9+cOK)s!
z%$t#LKr~n5g**s3(?=Lbe)3WljxB%sA8aS+sjHoYDD2a{_)Lf28TO&7`_2_VF(kIe
zx)1~Ly4&_H4wW*UuMDYHeh#6o+iMus7(L!!O=1YV0Y;6lTGF7gFd`Qi&&s;R&Hzy{
z5470Oc03F9wxDqV)fN3Y^rR`+<v7vrbyP)5huLs%!NBp{vjz~IApkX3DFj1<e`)St
z?{yA<NOPNjPEttxpD>HUyYo+&_48XPnsDY|7$}_Obu0-kjHQf>)6F@&O+_LcKaWmr
z@~Km~R}3>YxfINK?aza}wXWsbMc@zX#Npncz$+L7t;skE{tO}Gt-0wd<{s*5&SD;$
z*ZG(YY`<r&h5hc0uLS>9^X1F;-~64-i5Wk7dQ7<gscihaw+tK0N>3t)vb|j4;_mME
zv<+P-ZA_L9BjOfdGIN_BB{vi`!syLa5@Ak^l0<EKm2WwOz`W=y2XN#1={^T%a@m{y
zqHUMBrdZC;%Sqyy&X6#oUYTdc{>;y$D;d@2N&??94O#s6O7uv=xgE&t#X!T-c(pmr
zMFMmj0jx8CuMZ>M?-?YaP$4&1W*@7z&YqrY7@Ldl1rY^yu&MM`GI(Pml+4OKOg2b6
zk?Bj#rByn+J<g%U^q_lFLHQ2x%cLpjy|yF-z19y2xxn$ij<l4&iEoJWXq|!8K;CC?
zELLFntWn)zT`>C=EvN8FKJkF=wZ@643Mf0FpJ8FvuIB6Il>(n9bY~sip40Pd>E}m5
zb`RzqUT?*W&N+h7QMag*hCL51JlQdAgy%w$xWveJOga~tzl<TmI)B-afJ4MVuDjBz
z>(2$N=lz4tbb^)T4W6Rs5rnX@IekBVgePt*G8^gX>FPRLF|J<V?PU%;`svUPbXOXH
zXETGdv3D<d2W;v_e-jaT>+B*J)u6Kt@9g$|sgYWx)7sdEVuZ<#N$|i%pmf9?#3xiR
zq&s?MI-<zX@~=S!8=S(c(|tIFcO1fcz8&2l=_mhq=YW7cYE8pNj;<~)Fr_ex4Dp!B
z3pcElc>JY08_V_&+qZ4okJ|(tWp_%DE+W_Gk_n&T4h+ZZa6`+Lr?wb&z`4OhYwnH2
z4Z4i*d|Q+2<Hsm+(E|5uS{25!V~17B90JB}PhN%~C4nVAf<jl23N}KL)6+z+76xsm
z0lh<>_Z?>UkII9OEOA&sEPp<J7Cc4wpsX`>if#|Oj9B~KXtO(7RYisvqb}Z;SFdK4
zo@vu2;WHmrXjC71Sc_)2GvDKMk}K1%2EWGu<X1Nj7@xb8d#M^O07G=sUVj#ec@Dt-
zw&;)Fw-qPwu*KPj+EQ7$95piNZeyj=0Ym5HL~+xhWhOf|V1*t?``X1-0}eXo8alv~
zF~mRKE~TH;`Ptxouy<HkE@BR`HrtQw?d_#Kjp=5ZN0}A0wf*haC~uLr7bjaah06uC
zZr&IJ(a9<*J}Tsf7c^e}m`PO8#`6WRPw8rAW@d}rP#CR>^VI9s*~I(_k4ANu7AmlX
zr3EELacUwBsU~iPV_sffLHDw*4tR$x!4Z@n@wAKci;9W@9aIEyOf3RW@L38&UVU)}
zNT6?|$PtQltfqeYg$@{r+h-&%>=k);B}7z-zVcwE@i{_@FirjXwU0hWYyzPG&%?+P
zMINJfnfCdh685%N!gTbEb#B3>?C}4~TbeC7Kb={cq+SmhM4%bEJ^0I~@heuYG%+<D
z%pz+y4D{<~#XL1uCUE}E1AvY)S<|1Zei$qFjKSeu^-e;$(#F-AXlAClZZG<aZu}9S
zbtR->hefu6bjZcaXwdl!YI2_Ezu|gyPia`xD)Oc5`^xyGrbdL*W5@03_xAgZ_xhI6
zY!S@={`3!Zw3)W-IGpm?x9cs@q_4*mcE>W;9s|-*B-G6fCcfT|n7~GygYFG;-9Bom
z&38lg84qeCrlL!RCf}(zEiRrP$x$(2w%nN|j*s)S!|gFL|GW+=;Ok3`du9WJ_kz8=
zs?kt+d}@R6AZnwuF6JOY9~(m4iCDk$+QEZ)xJJ76x^r+qJ!+g6UB^B8{!($<*!@jh
zbp5y4Onf7~Cau$uP7*q`fZ#ZYeWTLxzNM%z=$6bpD6_^(hi+eeux0u915g-qb91$0
z1cfz)DX6J=y13Ya0rLCZuo~-aV$wK&k{UJ&E-X4Ko@N9OmGHP1kRdVIFCaiS*WNCw
z6AWI@;?3#N=ic{50Uh)$IN8}>z#NT5BAE7G&HslgG}!5HvKA~*Z{~Xt!S2EyXz?E+
zYVKcXiYPb1&Ltz3dkGs{1wZ{dq(vMp`^DBDNk5%?Ib7PI_5+DgTMv>(c00eT*yrOQ
z?#EHHA1g*dNKyLLf%j6^27Qr<8GF?QYWRiN-Mz?u)h3<XKWS;<{2B?K-3##uC>erb
zj{^_9KunBL(!JdO^)5?*ODHpR{|$zF=&LKBuQ`phq_s#}QPI(HW;F)}gmY4^ho07F
zuZOop*+*?)S5FVef>~U*+QIqAwzB!yn6S6jJ&#%sY9LiPCjJTY448#hgx{Wx1w=>q
zS+4wadkdlkv+ZXLaTroEcPj_Hy3!?oWkm3uFpp6JdOqu$0Khm`pkhR8P6w3`!otGk
z?|uWy`3^BY0`O_ivMKrx#4@EN5bLm^@nw|Xa5M3$Ka)9r+`I6BW}2zWj9_nNXfTL1
zU4Kh(ll!msv(n7+B)>B@&?F1lRE!wO-9O9rXT+A>QOc)Je?wd}yx@wI*0}iN*!?$q
zjzG{4W}xQ=m)6TCGt;k%to`1%rk(dJe6w!t+Kbf@g&@wQXS#iWsgQRR0H-jxsGXo&
z-Jh&Rnxd>^;5WQ5*jk^U$1Zjv^&Cn}jxp;03G)8q!|-{qIW;5`SV5<xs5sb=^1_}P
zckc{8KOngWYuy++8UhqbBAtYc8OYO9X>Ih#qpn-OJ`{u{TK@gYTbvgV5~>dR1o-tb
zW3<JlYU{{0*v%8|N#K`&&E43gyKp{P-UwQuUT&QI9fAW}+iBdq7?t4N&d)$6J6qL)
zct%4eH@4}K>>svC57(_viU4>Uv^s=32Kj_%xZDJ`m$H`xi~Ydt<;NNJKiZ^BzO197
z62Cd|vez*!HML+66(`!kaq7?SPS&3YJ@>~=K3k$|qOufud8DM~!3yfs(855NrkA7K
zMiaGeGre+1%T%@Ec9w0tJ@@nIHq#e=K8N6y9@-y`#jV_veDL<1t_7>{s<<gF3F~f`
zd(tmvm71<I!mdAM*V^&QJPIt5@U@dBmmeUX&LcF3aBkPGT`!Z*gNZVDqx$Z`!Z8$P
z0Jxh*&tvly`6SL2)22)!Yr!fhc@sGqP;`F57jY*GgBc|{{<tt1If68lY(N+8<mOh)
z;!?*UfywW=Fb7fqH+FbvC_eYrv17$3EV<~=wRCld!3EeL>Z$^kMG^xptxy*!&n!f8
zI3&q9-_x3N^Qf$%Pj*|RPb);$`)cy!I2-8K$@)_|NBK_W+N1Q64|M*t+iWRHKKR<y
zc{q`@?tAv2G0{LQo4Er;pTQz#$I+^Y6e}Vti+`Mm)6#H$hQPY>2aF0`t5M{!jXQRP
z-(9&5m0`}YL)`axESfV^%xr*|ofm^cLJnX`yikHXjVx~!96*58!M%HJ&@X;4NLW#s
zuQ`nC2lZdOK))0=R&W^W#UX+>ns;|d8Mz1<0ANf?YWEZxG4`W6H!XYTl1tBe2@)X-
z1&5`{XHf6za?ruSf#ipfQX;9Rg*&%?qn@rHL_j;C0YSZrUlt-cOgZ7KOdMON0~r($
zu!X0+2R8z#PP$X>wI`x^(}ACm0~x^i{2L5RAKFsv4^kih<>CGNFFm(wyIlL(WS)$+
zvw^NRdO8oEU6cJ1#_#3Jm*D8;fp~yQuTmW-!Z}$AHA)&98UWPMAg-<7#)q%8wDc|F
zKWwX7h&ZS&_wsV(zmRv1mAYFU<w{^VroK0GE}&@Xb=RZIE%=H*Ib4#I`2X=wu|Kp4
zq;d*P^*>TFqtT|a1vaVmZ*thlVLf;;8VaKhXu|7%6ZDi{C!QZ(8IkO?4SoF2%Klxa
z(?_w@9209Z-y-3Y61<dBpio{V!wh^uj=)3KTE{MbMih2G_wHzgSo^Dge<hos2X|Qp
zuH)T6k`rsbb;GMh@Hans@&tLd*}p|ZPOW64t(iVRYjZ4|QfYyXRl8XKZq4VTKiYD>
zUO3nB;Q^4SS~x{D9{W6a*P%D?JfggT&3}JyLxh5CfaEo9hvnii<hZNKP|?uDsQvxF
z;&f(IPD-rH{yw`?Wkf{8F<D9o^BStZzhS)+zv09FpWl#4`W(q-k0dW1`hdt9Dx_7O
zFMXKIc_ASok|cshgTw+zstw#8KzQE^PwA#yhg)EYs<}}-9S$&ua#*<^%P5Ae239$8
zOv}HKsNu-+<rIn_kt19kg{ee|DQtY7P$Fm>luz1?b-Nn=gt6jK@~2ze4^I8<M#{HA
z@~mH=hf$Eh;|dd9fhTgAG+Tvt2?~C`9L$GU2Qk$n20re*Jw~n`9@>LwLwf>wB|8UT
z)vLQH`RkuJQ^eUMmO+28mMreK5~8_~hLdLC)upASWwnZfs0E?<FQgaoO)>uO0E1w$
z@aD0;YtXri7#~A)(|;Y&kTY<c;<RlVWV8YvY%N(tnI(~m4Fo6123$z0Q4QY5$IoA>
z<%c>{QBhKLjkKZq5PA!Lj-pWBRHMz=Br_`Yc2zn}ta;*gjJ(()5kRn1{C{eST0X&6
zafR<e&)>u0lq?R6>R4#Xy8d0CoG6+lrH5BwpagKA-N%WaKlPxKndC#hpsc=LCv7On
zC?_a)B_(rn4K0O}Al6N^6w2s1qNfnUfQ;7c8&7YPffNEIB%b|!0`2zGhLCdbgHUMc
zZ_dk*9o)X8km-eU7s#3Les^NkiWP>Lm)|=_?__81xnAnerv#N!3ym3-Op?!LT1Z=t
zE&bJ!g6RKMh?E(QcG7MlD_9x5oRFb@$PNL=TRo&1P0L7)s4A01du~osEsn>)8H$s^
zVDlRRe1WCe=V!~v9@v5>gp3P)?tahAlo%vUU*YBB<Kw)vVcS0COW*NVSA{A0d}cMY
zyVb&!!zr^3ug#y%O<uzZnOH-igxw>mn%z)`K~XX0AAHbp0WI~J_mR|M)a!5GPD79M
z?GsL-^@H8Z*x1?cNuxSamdL=Q96)Rag(O32()X?rHb>v=JBm3Ye3nrxAB(Jzas2Or
zvqaVYKmYvG!}kXHX#$Sy%AmkeXEIDNUO7)qxol<4bXPoO_pV)^Ih|xn)=&XQ?j&~-
zi3{j=L`=)KRLC`d{;Zyz4%OGWsP_;Be5s;PrlJ*OZ7~dCU~~Qz+*aHTa`i`<mo^pJ
zxuA!Y)4zWg6%{o{{RHfDT?PE5H;eaK0|$MSY56lxWFNg&;P@QayriK*#z>U$5*a0y
z{0W2K5KiBRI<5v^rFOaF!%7?-7Kpigmi_`{0_8K2K^?fif#MXxx_;T`Kijc%QG@@t
z5~{&jWJP{8Iq>;DONt*AP|;~aP@9EWE~Bt=65$+dh4_CL?f<HbvOetN%zFD1%+amJ
zl(P@#T8Tu7Y~c1k@-&zeJL-`AH$0$;Vtt7HVvdB|Z?H2Cud#GW=$3Rzc%6$GhRA!N
z#2CUVC+6ATKk{&$hdq{F>W`ERw7*EUB|d)u>aO9#s<L~uUVa81Pzt9=EE%#t*I)om
z{yq=C1ul4#d(+ss6Sk2NR)v9qVf*%6Fn$1`KLD>#<BaCzU~U^ogXQ(xv>~AYcnX)i
z_GrUuiY*;h_4m?lFC8kC5OIyPA%4Bog4scE&DPuF{kSxp$u9<BS_aHnLm3n)|Kl>6
zp%^_V=L$erW88LTW<vlSXy*3*PDJxyWZJ+PDnY(=z<u}nLv!zB!zme!xenQL>umrV
zmX`02NDvfT5X%@s&kWH?cus&nbN%tLG5yTTXD(jkKDnMUsvF5hxz5JPdH{mCrA-PX
zrqm9y0x-|rcrPeIgihx>Arg8A-57+JA$!kztQR|vA#?tw;RkS^R5v6a^V(@`4v=eF
zjGY@nQ&+4L$NFW(Ylu0z433f_vjdF$7Kj4j`zpG1SFT=-PA&7eWZD_HqSh+wmb6<`
z|7y2d=wS>tELjeoo&GB`bAY1o<>yRH^mw^SQK=0edFg5pJF?yJfU0Ab+VB|&MhWCa
z1w5ihey|&JdH8?yClbu`9tW9*W6<8K8S(n$Q^{uE@r=~rq6KR2JMEE;dJC}Z2q7#O
zzfFsn)pMxQ09`N+T>d8WCq|L`P4Zue>P~tzUg7)4>34M`8=4T@kzSS%qaf3OZWIp7
zDS{7l{e4*ddw#i?bmT3T3BpbW=!(;GYfWQxL%{|M-Q64<CZH^gXP)xtsOp)(uf_}w
z_rpKI%DlDb$-qf)8Mvf7hMgK~g9ItZqEXoGdrVXmJ(8zwoS@w;#;Z3li5#<xLob%c
z&qUiV;l@v)YJ6)BZTlZBR<k|>=VzSsUZBV?ric{)iwl>(nF(O27p7#U5Gu-y3cTzK
zH3|=ogD38Xlc<Y{#6O^PJ#`thfC1{Rn*C4Flqd2PuA&_KmMw{<o`i*|);E?Bs=+-S
zF*FJ70>_~31z8YxU?;)dg@2@4QU5oomMOv^Oia`X+2J>`cJ10tJmw!32S7YQQR-J;
zXeN{$^8qFh+o8|mf>6bMi;yvT`XkxxFE9do*xo2*TJ)NlV(AH+vk%)(0M2QI1Vo8*
z3+}}oP52KqYQU}E4z4shuBnMEINyyQ=#BwpT!LP)v5mI*%=Gk@oMN8M2OpGJ?W(f{
zSGT^1nVvo>a1&8p?NI@;e0<Y;)2*k|)w+9g$&xp`dfpVM9X+-NsSx$>Wm*wOTp^Yy
zE(t~?r9Cwc_-DcKR3gcQGD-`3KYhkuK-EdKd&Pb9y8zo-!%3lNf)<7hk*sgIqmCE}
z(t(aPZd(~@%1HD(%^8U;9})kh<<5d^?Rg7weBYvA5QaSgqxKw_QS$l4v896H{^joC
zjgF3v9EYqB`9ur<1elCBy-&}lKY{IFFksmKT>NXC*cK6e6WoBoEv`4-j4f@utq0!C
z%b9X+9ZaT^1*BpIAZMkF^#rdXn<&awNUcHo{h+3wQ_55T8o;2{73JXv4;t{{y1B}E
zCC()yiIZgyHU&5$AW;nI(mYd3my;iY`WWn<!9}lLwTk;`G5Xq^%^d|D5PcmQ^OrLp
z(3$B?O{zN7;&5Wy`20gc5-fWG2GdOgm~k9!gWj~?4#s9?$j}mFj%-^G!=H6}3=zxI
zn+JJb{7%%Z2N%J1!zDEt(JqqT$42@<HPt&`$69fU9S6pDJ9w$2Dsv-*vgRrgft|>4
z;?{pHe}bVVg3siu;xG#!!Vx_C8xa3)Jv1zmk=M;|)63uRhF?NBbw2u)z3H3jf<5GL
zZ9nGYeER1$(QhRpc^=b<63GVc#e=Q=^I*zbp;YP@0Rf240Y6`ikQPS>e&I(kLX>pI
zfM0mw#{3*=@Uc%{A?s)k>-F*Sdfry3bmr&mvxZSJphIT>9E+%$YCCjNu!rlb$}*xn
zvvL4dx#!bx>3>$CBMlT3dtU&(n7meUqvtV=?m_?9IF4Hr{bv>gLB%iWuv}+8c$ja6
zp9kLeLsJGjVsGo%{r!0UZg3QATf~<jF)LM_{$9*t@V>#9;@1lDs<*CyM$a0L{HP1r
z5mBzQP+|v2JG_lGs0zhDkRpXlSFwXjGuq`ublp4}_HNO1$A0SA`)mvyheVxAPXifG
zhDF2l3tq9@pnCj{f+S3DfH~_iW)KJ{z14+*N9Dd(4a+FsQsB804goZ>LZmq032gzc
zt0uITzT1i%F!F3<A`3(wf?rXD=m@kWnp2Qm;vUk7GOSP2L!zO(j-)MnKD(_yWV13u
zIZ)V`3vIei!<xbqHiH9$nZp@NW`IwBB5vX`h-4=MX`PX6pR|N*jORgz`s%<%CmKEz
zJJEymi^gxga#~Ft+rsI%LU53#KaZi=$uob({ds}jOG~;HODbo*oTRfzprT8Cfwrs$
zLK(FH?t;|Qr9>8_y5Dy6**5zSq=rG=Ae%|_p@Z%}LI$aPe0)To>OprI@R9SzJB*RA
zzK48NTIwMU3nWv(+uJ3u2C)bkA2Box(Vqvwf|z0+WM=@kGc;Yhn&=$jJsx;B(9_YK
zi&Ml~er-$AjPa-z4m!GM(Ue}&>7^B?rN9T9ucleJbd|(Y7JZy40spS?V_pA;%JjJ6
zCPeacSJ-c?YZK1GYp6q@zB^%0TVW&w_ZA~r3tg?5$0*w5mz)Vk!?n-pTp4sj@gTcw
zfhC|2<UbziGN<G{c<Sxdfm8@jKyGEfzh8eT&gG4I*zk+oF?6eO)hT@{U=^nbWb*ai
z07Q?~!gx)eA_)2l=gt7#9@$`k(v31L1#-u1`$BJYOYM`E4wGJ{ReEU;fseX46qPO}
zcEwr{y=wyeu-{BTY@B2}3g$6r{|$oMzTOJjiVjmYVgKafCZ9h2%-;6;fZfNFQ@HW_
z(tZ5==mC-F>8~LVNLal&Ip)h7lDTdTYiXoL6zu0-EdMJHk_%~l>|fvV<`VYS*@QGg
zdYWST0^L%4<Ehci>4Fe>$muH9^=I%&6KkFkHnQ7_4#*-JKpa`u8sO@xKynUhfgj_Q
zNg+Y@=QH5PZ{o?(r0v!2Qwrtl`y|F7F*Ms?g&r^xRzQj=L8fwyLOC`!;s`gJw~@_R
zm)Y4mSQ<wC3jf|ERgkaa@5+AnE)YT7J4V43@^1eG&b*r&4ns)49N2?G2gD{%fKf3&
zBw5IoOK1WsKi)KcXdngKRhkS%ym&KB1~O341tLh-LSkf0HxUeLM1+t1{me}9$$ch|
zV5aN%b>tvvOrH4`?VrJU**kU<3k%;b-5I`u{@}OCKhP41^=S+2@R>mTRv8UnDVQp2
z>(s)JTQEDVWegT|(qb#++0OOL1POl5Uj;$I5#F*F(MT_B;V_Nv=Lh+R7Dm-o7&~Wz
z)d6G2Wu(!22G0dp>E-suZ}aa(SWtLY-Si6@JWo9_5Cst&c=Vgdh{w$G?d~}oHteG&
zouX7mGCS!ZPo8`;^J!cYQ~Lib<Z)F|%S~-T^7=&~k=O5-BV2oh(4N@PI52HPjlvX}
z-7#z~$(>k4Ty9j(EFC_`^TTlXn#kdkH5PUmLGPXgxS%p$F7{KI8at$SB7^Z8;HG*!
zpE{h+QUix?k3eq~AaBU4c(BfB7qSPF-A+IKB+9r4lFkF?H)PLhZ8S*ori*?FO(XE=
zV}H8rlsPWA>hmc-F&(Ki?+16mr%!`R&y;(+EVvIz)c(!3!Ya$}in{huD40vAbhIEf
zfLnfJ+1MUnwl4=<Um$o#Z4RJ;vYHxWb^+1{UjyfW07F3{RzKXO{T|_pyR^Z7ls+BE
z5TVObWlJ-H@%Y|i5FlpBS8CIlH*M<tl~ykmT?%d=K}P)cGbq0(xDYh^uC1+YtVrV4
zuLRNWWhWjn3M8E`qflfabi_WZ7w7#WPqtcdX{+60jgje!X@rQT$qbG7u+32&Q@{R|
z(L7^M>22&xUA}THQ5NyOCC!@#E`98)RK^5Tem*y!7tuXnWwM7NuMNa{2Lw!mLjorX
z>4|`T`mveBH78*DG&MC<zLI2F&j-l2U-c&4!(H$)clGfN!)>5GnPn|klFYe&*>+33
z2MJ*;WO$X-$hsAz3c`v-6%i!HodV)N=Uj>R#dac(8F?rKDFjb)wrdws*|k^%4308P
zdP73#iXw^Y!vQDt(S^w)+N4^^QsV5%8*wh*wQ~1|$YULN+iVjN2V47*q;Sv_5ylrD
zE#D6W=%f)Oavun~hI3SA=}AyDfhRd&7e`xm_psgERn#UyOHfWnq^nJDKq9^RedgWl
zV7#fnXLxu!&T1-TB2fOkwbXIDS8q@54}_A<QCEca1bqTj>GR)PcP*wASD~j}+|SkI
zLdx6ez_$P&)`>{S=_hdP4w%!r*?}<|Z(+g&ABe1QA^^VvPS+U@K@AQ;y#*syahMn!
zJI*(}I&-nzpGT)Mw+O@gPJpd{k#+@%WUwU4s;bxTB{9q=%v^<Ud#zOVC-rGvwrtrl
zbX&ZVH!wJ;Nf>fi#Rby&$*u#9_5H#}M6UL$k6)__k$5pv44o%KyGDq;=y{Hhn{&GK
zS5X4kDh(sq^3S>oT*LK)OIY%)u>F=UCPF7%Oy1u_Q!DlC`m(aJ+*lE3Mv)Wnl@MV_
z-JZ)7y^(5dedf%WAT`dq(3AQLXy+xuhhVsnukU%ViXRRRqM`U5XJCR&*SfjuFT6t;
z!RuZhNDnlUzDCHclQO^qJXesVJufu7B20FUncLBe*!KR-6sJ~bX(WvGl<C*QXUh6m
zo~<xp8A?2vUDxT54+4y*t`_+w!yCUt(aNz`76Z<j`1zo@HTp6J^W0*+{cXDzWD(`T
zV#zm0JHw+DOB<gah72juY!S((r{)zQ5}H;1%P75o;Z?hGBfln{3?@7C;y+)D5~$OP
zNThUcm&U8IY$NGk5K)Qtk;h=ZaY_0fHUB^Zv==kHotoY68t>n~4{TUcLA35%TqRfn
z16BH&n(rWIP2zNk21jiQ2_8L*8Jul$rL7F1)l*q|hLa!wMON6d60kVS>hEXs*T`5~
zlNwK;p>uFsD$sUZN)~f>bNf=>hp?=9%Esi}x#8Aq9px-sXx+SM%v4;2c;;>CFF=A*
z6-TF`g0OB&=<peWT(a-eDN2h~+{EBlC8=8EZ?{PmgPW<yn#2`nFfhuyonsgUlzuaV
zn(}Iu8P%u=?u*lElBH7ub(g|RVkScb+`Vk<rWmo?Fyug#FJ>)Qp01QDgK}I20KPl^
za4#}XMhUUte<Wu{uMO15*>p=dYCY;tLTOXUCI3AdouC-PM%NcyidF<Z>mh|G(Py?+
zsEng<En%z46-#d-%aJGRQU|gt%o)2fDndjVTdy;cq*F91RkBd<azHyC2rTeAqT#Q|
z14HTn2F}2~9U!HIkC9vJ*h89a|Eg(p7rhFD)~;WFt?Bz4K<S`=D(!+nq1UEooqoSu
z?f1UULCq+vF~j&W_MJNu3|r$MNr{kiD=n+*zdo3WXVGSYi+vl*ut%MhpWj3JPIk>M
zR+k@EWRU=!mlmncu^?5M)!&2>$1p><xv;E^hnKhf0y|FKI%;YyrAIsT`Erv@rBNHE
zQKjbaKQuL>^#9S+#5xo4OOR0MnR`SxHy9fuQgzvJZaSstHs2NCfz6w3U=HB@4K#I!
zPvhz{+-OBYw!|bsF9T{wzxa3PTq*BHqF*OJ5i_bzL;3LhlMcwPjlFS!b42j;4RpxH
zGH+z{D`OAymQ?@@c6-p5N7A9IrFxobMc!F0Enm}!pr$GoZYe1Tq)V*c{dCUD%kx_o
z2WYA)dL5kVXXfV(`)q2y!QO~U><8rUDpMFy^&^s*8+V$QBs0!=k5K~2@zEG*&!=U)
zVA_n!&}Py<J|Jit0raP(YG}W$flLt00tY8+X3Ljd%P1LaWG$hsT@ynv3pXgEma{|<
zbLuv%h_#)S)eoe8yE}inh2qanAk8>BHYS|si@{?fT2kcn>o3xV!vuzjU)jmV_Fn4_
zSeRMx)rqb>A15kXV<So;E&_EIas#g98gigO-~#_5vfSOqKmZxjo8L|!j`9a5n;N?m
zb)3#Y5dKhPMWz(FoBI&Rkw|@h36{W(2W_1u-DQ(^-aNs75c7gLXLymo!NDTE5@Vz2
zwr?NN3JnhnLow~487dlP@zc7%`z9n1DyvO2s2@FSgd}2Gl{AmoSFoDVH<Hvyxk72h
zb&OW*+yL>eM3y4Js=Gg`kunU25zF-&Fa*aK8rj=xau5cI)!>+{)2$`q1r?!L@czxj
zsR;T81oRIL@ywQ1RVBmfGK5R!IQ+hz^-zva=uYT&SyZlE1eFssR86EdA>Y*z#<J~r
zE^bl5#4K{*Ns!pC1+D7?_<!xFYvPZ!d!hopXu><qz=t*yT6Qx%PkT(4$b~tC$RbBm
zQCIYWqi2rSeEZ@Ayg*!!k{`E;=}mDQLz#>VHMH`c+f~BzLoaNi1rsJ8jy8)&czJH%
z2>!qkT;~h`@Y3)Ny2h&UF;t?I%_!&X0S=4}?M~mpCnY7db7v>Xg8z!g_&3hEoM;rr
z7SHQND{q}))q!C0-2y6bBBP>Qbr$BQPF=C4OMblLpb=jNQZ^OinqVD^Tdq6D)<Df}
zq7DjB+`Nw0tzyIW1G?K4Q(BZyc!h6d|0OwI@hDl|TKgLedj;Kfgvd_AH_??*LDtCt
zHyt_%!p1vI!UzK7k4O>E_>xp>jzB6NUg!;^;}dfPh|M=K-vcxUotYKR6W(juWg~k?
z+9ERZ#v(I3Y;1?@w#>uib&nYzo&5C+2=&!i6}X(9((J=?ywO>aLrI3laaTuL^r%Zg
zdtXrL!0jp9AAs3(2x(O4{l^p0-m~IO3o=}<N5LOl%zZ`N_&G*1s>?cv%L&8%WHaQs
zu$5wABqGuRktOGOBK9~3JNw7EDf^l84S)4ABdayRogMHHLL1h;;YCQ{X0iss&6H{m
zBY6@#9O2v*(XyK#6ZepK8#`}-N{+tn*31GejqP8xAWBChG0p*)MxuxvpFqJ^Jk@z~
z0o9y8)igQKRhDyd;YsI#+dZZ9Zbv(x!HDjIK_B7d;Tg*63nNGNe?2HDsv0W@<3vt?
z#}ox)eWFhmQVDi;_V2fMA<tsR{HazdGvqzeXWg>jTeIKTUO|ml=MUpXCVw6NNhE@g
zBz=V#V(+?wip)|iTr{_u=q)OH3k$+X0o=?MQef%ko(Ft)aR$-HHa;6@etQP<Xt!*!
z9jxUt-_T6rxzr)V6(vsvc3??432H>W`jQGdX(_HloAI!kTQI$_w_oS1w6*K7oH;o;
zJ)FVf{mUy5PO!Z34LPubD2vCcADJzpxZ5pn;zxV?tIN>n^n3t-3cFs#uM@}(`2n<@
z@2m%YVenw|Vc9VUCkK1`-l}jbl`e2=TINz{ed=641siODSxZqesQYl7>0#udS$a?l
zK|ni+_3t<4iLdS_e$weAZJ9NKz7}hOZ@Ew}<`>j21W`-%LwFMx505s8C1Fi_)->Wa
z6wSB986C>lpCTZ!xTidHzfQ#qug&k#IhLkpf}}Li7&#J0d~gx=(*en9D7c*yId<*J
z{Apsy7{m=MskrmRU-1jeFoI9ePySZ(oX)>dEA%?|CmMqnpC9~>QpXmpb&cb7#zj1t
zCA<?q{<&g^s)ItD1uxd_nO%`&4tfK`HL&y-P<+k*^HnrZH;002FGq1v@GM+MPcaYU
z+THU%4sP1em=&WO8!=_@V;4%%GsiAmLwn&#Bw{b7hr$yecM7F>{>k$R<|=+9d7o3t
z^@+3!hehMoOO2e%g?$W_u!$}-C^vK29B=J#ip~F(?F$&ccyqlu6KrGSrMl-h7#6HC
zY)JHh)?tiNkfC+tBy=~duIgP~e}Fw$-M<PIYPkE<eBa#3$;t1vvABQ1s8LX692mdn
zryGfYr6xgfQn(TMrDt;vXfGShq`_~ViyGaD7$4f~FUSs}OSj0=-onEeG+mEEO4yML
zHeH}lY(TpBO(;XuY+?X4KavM9dF>Js7K9)vW`IN8PQvE~mK5IeGQ4Gc8_Vv_oDjIJ
zm;-1ExC}8VQr8OM-E}erMMaCyMI2lRhk+?V|Mhx8!}6j^zL_a;%eHOthTSLgZcDZH
zTcDn|us#?f4cCK($D(3ln&W%^)zG*~oGmhCX+q-nCW46dxh>$qmki~-N@3Kto@0fx
z++nQeRJMkOMrQ|;<+SdT+FGrijxOUam1D=AWg{0w6Uiq20hbXpZ8P!JlS5CrG(Qy;
zak8>L1OJGGc{o%}M~<xcxqAOn`p#2sDOrmyxs#m9q+EM(61^HAli*%7nRq{i#OF24
zM86fu)=&l_l_1lZt(WCbyi`AflYdpGaw&zyNs`|;Zz74&bJTy1)EO3{SOao+5d5)=
z*P#BMT}z?dApA=V+hd$q>6AOubrV=I#y#r!vk$)j+H!y|ODc0ka9w=zEj%F62mH+>
zqjmDf9TxNR^P-*^F)#3z2N8z?lz-uD^hSGYJnVr+9dI2uo;7zes^MVbPPpL$Qw{+2
zpk>GE9rmYxK5TJup_0U7Ovy4GR)xIg4Kl6Hr$QjCzBM^0ik7{oBuBb}<Um)hV-Vrh
zM*SJC+h_NEOIj2ZjUlU@j%41rfh7XW0S*d<WbrZ-P(0p}nbK?U1!enb0R0$@ePJz1
z`%q~ry=TucO>!d+7)WFja8~=0P?(gk$4gx?VB}3RTKd9nK`(V{ViScjqUhsH$~a~V
zn8y4VAK3N)-kZjeDkNmWgY;4J@M3S8oR8BhRDP2MQj;ax6x8=7Z*8@P1k3>f_Sw~V
zN7N|_*3hP!V3UHZRJ7YD0G~O4>eh0Yzc^xYZV7-n-CcSLkSta?lm748SFK!GfxHMG
zpL-?b0^hxypFqsdg%U!qJaI6qHM8Qsz+E6O#IJ)>v?=Fe6pn-ss5;65IRII3b{&X+
zkT(XY+3EidDhE1=^=zp1i)0H(K~MaO<hLvbYy}PY=D>GYPwtd^w2PDTOA3Rlz)OV7
zFy|y(&i_a<?f78<@%+U3w0muBEyQ;vKNoSDn?I`(t1s^_a7wjZe<`g^ewEVw2+7eh
zGb&iH2sLVgeS%ip2SRW9T!}i~Zh4qS#rB}5fq`7@^RDD*Ed6{|L8WhUax&9%76?cj
zj4zP#mF`6dY*YD8Er50nYW{_SpD#J47s8Bk%<K98BZHz`m1Vjs7nDsSwKzAvPwWJ$
z-Lhgq4?P$Weemz67oWsyKlSrNi3IyYGUi@1*kA~P%a&TGo0VyT50jIy^wTX_LtCB+
zPuR;wC1KY+fy_d8^v#L>Lp0l98eeM0Y#{lgrJrYc$O9mUQwpFdqC*S$t+|akx|(<e
zLmtl{WkL7vZ+_b)j}AUDY#>B@q&4P%`GR<;K-s^_dX0rCRhjwydpv&98M0ke5E0hv
z&yV;&Doh`cap>P%h`eoF)?He32(dB8;rg1s`(~KxzIG*vfy;k?S*Fs<`Lu7x$;bPw
z+AM6$opoq|1dnl+QE#RN>+t7?EW?&-eU&UK<%_`PX<4LedE?KU_YIZ|(9$o3(3(*>
zBcmxQv>0M&Fn}2c(XUCn?yISaK2GXyEKmNs@F910ukV0GdZBh7BAGIwMOZboCxP|^
ztC2f9$a7N{OV*HcOAIONKR>D@&3bXJ4n46&Ry;=sbZgyg7^aNZ<M0hYj~A0z;jf==
z*G4`Qtg7F?8+ch2wUCi)7Z7b7+5F4%@#FO8p$LrogaiTcZtBA8F_AcC=_fxBS{+he
z(7}Qv0zF`xL}n{O*WrDewrw*&!19rY3iT(-S&P)0Xh|yTzi*^sVUd8bEoM}y`1X=B
zizF#FwLr=ZOV5_W7!R~95;yechfxUby@vK&`|aO}yT?cxV<;YCyMvx<U>wfo5_A0L
zCkfIMf@o-dK);s`j<DFv5%C_jUSD$xv?(ChMK!!S@2T978E8NFr%F(+57FMAAEz%n
z*B|P_L9Jb}2+NC3KQLd)C!ny_eP?@xq_oE{bNU0Mf}-vS!U{j`--q__xstN~JxE!w
zo==>g{X_-R);8ytU=kL8w=Y6e)Z-06tcOM^HdRtF47Z>Tgm^a<>6=B*oZlXColT3<
z{5fmZGH4CwfR3P7;%(S;ON=Rd)8@0oeAnG`j=JYeHJA^-Y@i>`S@&qCi%a@HebQ;S
zlSIuA{<HH#$K}fM-;?z@&eqezIp=?HwM#v+ItV@UC(S6zUr8aQe1i$qT2O5u6;&s$
zt%5>-`}XY!Ba_h<0NUcHMV^Iv>>_qRadr*3&=dAus3}Wlj*XxkQuMq61H1GS?J0c!
zKID9$mApF4358DNuDQSvj*N>wIXMp4SpV_J|9)-l4#zn*e>RAU?<$8$J3vYNwYz11
z@l%0z=rsy&tv~=yn<q0Xp|9$_e91<s`+%Bpa&p4$1E%UlA`Q8PCQz`b%fQfLqELXt
zQRpiMX%4;LJaWcpkZ}$D5x@<g<a-dG3#>}De!XAiX4j2K^+xQG9sTud_hwhs7@}`R
z_bllfKUiU+kwaI}J*AEq`DYnHt$anKosrObvw>svxJj$%7#X1jhaf!%5GNco2V^&3
zPEaY(hdd6x_zlo4T&MuBHB2;oaT|od2)~hl8t&?{`8C{#|Ar1g!jh6fTb9WjNRoD3
zXsNHi0T2d$YXN1$D4CeS;t|Axb|!6Y(z0=S`ucCPvgo;vM`AxvzYf~s3-lvk9d``e
z*N_x;oF2WS4WCh~?mQ|X8EOgK13DGoZw{P@u>c?sm<I~R;o@pixjX8yL`@$)95!T9
zaO3AoK7gV~#btQ8WA2cKf;PMczm6pE6xtu5lmWwD?}7b<=28%7`B9WksknqhPl-<q
zX0ie<n8CM#zdv6dkb*tHqU7PYB-ejgOVLu0%!SSl>VtWR2Yw)+K>&tdu)1_9lr1MK
zD@&oM8#@P4?^XN{<kuIbf3kmHNs*C^pnk7T=h>nZldhl3B1;}^nw#8i)Y!m{L2b5(
zOtnA%$)?6eiB^Gg3Pat%)%93==o?UlE`{+Q;4;m4DZP6{L?&>{5kn7R`V15WPf>D4
zZTg`Gf`#XY0rZfaEr3IXIjs18l0tc`%q#oq_3J2VHc}zX&TkFx7t^?hZY33@?vb}}
z-n~>~FtyzPbS>6}38`P;9O~YWU%$SF_>|0tMy=HMuCC1+Hy*mTT(DwX7<x6F2b{;(
z=z+2-w%J=?;5=H3;C$e=As_$b_B1rd?4eT_d!+cV2sNfK1J=+_@Isx5_{OHYxTCkX
z7iz}*u2&Py@OuZ5R{Ll529U#_>nN1S*ZKJ}Db&xgcjv}lqtcZA)?ths(MvPivv$c3
z(cr#Ayntx62up$7&;T+t`AQ0y{G^9s5e!PfaYR||@i?tm1YD$q0QLq1U?P?%P}l^U
zul-DB84ZmmEL>J*=9W>#AmkAc2_rz&Nz^A!E$s2iCmv8JlQwkr)1xFJlH3<P{AxD~
zNp!SfsgTlw#)tHkgY$i&Is}$t6h|0h9$%LJ$SMX}S~AoKsSt!}<nmzo5!9egq5`|t
zH!>2|$xaFikhI{P&xd%w2Il7&J1k8MHJd0%Y61S!kv6&&xdEQUX0W0uP$dXy=!eJ^
zq&ql3vdd~}RcISPjR$~q1`-XZNy+dhtkw+n8EOpQL|y5e2QvVqxYMT)4B_7pk+<S#
z!HUrF8|vEU_nPz2L8idUw*mDA*CbgSC^GtAh9(yl7I4awb&}EAdU#q>!|%--kXdQ)
zeYfGr1{wm499^z%LZ_#B2;7bp%^>hf1I8vT&`j?n&$>lg%>bLgBWz`0kZ;Fbb^;_X
z^jl>@bt=2^%8eV0jC>@Zv0%c!L=aobcMH82NnY*RkUv%>`R(x8vu8E2p@_CTu+q3w
z=voLU@8RyQ@-z>V4fEhqu_jpDK!`sk`XWy^UuI-{iaSV#p#gherKUL3qv2wxJqDDp
z;kK0=*drk>j+}B7q#wBb5{N;N6dhO$M4%4wHM-txleV`-9>F_rC=`bR^eN_M*+6d^
z_Ys<LD!Hv6aN|pM1>y$bp!utYyxuB@nQdM><@U%xf_=~230NAc@JO8bl6L3A`uox^
ze!U1a1D-O=Zrt={EAKJs93!-Z@IkbQ3!8|IC0-vdCA^3#+Hnj*I*aE(vyx{YDaTUs
z^Yan6KyOx5TFNc4Ivkea#$c_&tUoM5@WPpW5=Lk`1lH2*lfpunEof>88QMXEp*P0B
zVcaHDmcv^QuO-6hrY3#E5L`m_G&6N4Xq%$}#4&2_84d)74yp8J3!v?l81O0fCC}lv
z_R2Nru7>82C~ZkYHmxKf>xq90l(HZWu3fmW=OmSkFcMt4lC`!sZ{GgN|2^jX55VRs
zyjE0~|B+A6zYL%@khIT$x4=zv_(=tR_7}DVmG;M2#5A#<xeyP+#Gm-^@nhl95e}#h
zKM!lJf3*BQfLygKg`yrslx9#|!}?=RQHkAH8q{$(>6thrgHv8qR8%AwN9GXR;_t$E
z`?|V1q~{EPWz^MaL)XegwB<VNw`w~uAhK8D^(u@yBYIuasl}(!TOfu4VzslgbNJ~%
zi@h8{uPpqd|15=KPcOIiu1NUB!Kys$j!w?PH1Ht;CXEBbS8x<6euYEq_Q-4wh~P07
zg;KjqTG}Y6owOz`Jjoss8VZVW7XPn3=ax>|m7Lr2H{lE+m6;qJo$hRo$Xi_a4QrJ0
zexsAxRmoN?2-;Ygn41R|RDUcfQJ4(^eu0ej{^uh&l!PD^K6eZq7ks3bqq)vim|9@G
z`>QB2!e$v5njxsJ?*N0w8%V>3<<<WA>zr%4^%sk(=j2Y&li~=FtGTuu?@)DssYpL3
zC%+C{y}$b{wh4y~27kKH?GAakpFcaI!kx(69dl$)P%Cx}lonJ?uo-CK;$7gqc>v7}
zfX<&zgc=Uek~RX#H{$@5FWsCOi`rL43#HA@Wah|42T)X14n0^)cz!O2K;sg&PyzF-
zzNJGqOG@ZqYSC*$;ncNQ?ZhS+6^JYpjZe`rFk~Qj$B9Avqg`GX!Q*i%2gpP^UGKm8
z7B_{kv##v20GN?`CqSd6K6ETSfXPCzeaDLxhK2hE2PJJsPjAShP<E+kIFAlD?)|!0
z0`sqsd4CV83t_4dz`h0Of=R9l5*>n$4}*h_N0dEA%D_CUs-Uot9wt$>3uBhfGK03N
z1M;s1-?HmEPijWSFG!HnFH|E?cHvn`@}1~%2J^e?C;u9=J56v2@L}rSl$Q<QFbJQy
z>(M)uPm@N6#6rfLEHI%3^%ejJbhR|Rp!p>%0<Nn<D)eO<OuVxF?Rk#QzNzWy%$ODU
zsot87l|p&80fnEiTd@%4)fmtdvxaBK2Z89tw^-Fgy1srr*(WRj+YiHlw4BR2bsxA?
znCiW*Hx@v4knXKP4trBKq8O3rB^6pqP$)MC|IpY3UFH2%8<=0l3a>De61>t`5NTer
z*$<8diX`_bhPBvS^*bn!0T@KX#eSQzFl?5RQo*k4H;-^*{uc7XX;?soVSCcHvF+*F
zm-$F;0nCkOy$=I+MJ=1zjLL*O!`7{&-z_BPw=Hk5o2_CNvowSS(OX7g)%0bX7;e;D
z=Qs&(LxwG45EcQ?-W+96z@m(_pAXla*2jzC<GJM&r>&SFhlN_~^OxDTZy$g)2?+&`
z?h7J~N3iCYT&Xae$mX#A(6Zism<8O^g%&eY*aReEKUvBR^ng&Pxj&udy<(vKs5?`k
z93ap+eID5-4%4G&a4yNwz#m0aZMU_vL;KP@!k^TLWzV;l0-7eAZCaYr#BtZ`?BAfH
z@7=vi#eEWH5=;6H`o8e&8Zw9uMq?x0X@2zM`JGiisBpdoLYiFbp0!+ocd?qdIIewJ
zJMp3$+2vQPUcH%)LTTcaRkuZ`2geMa)(+`S++2Bi`5Ga*rEg~j@T^6j-@3Ye)n)AP
z7vGWaMr#tPQaUu505~#5wBx!@8!i-HTw#r%-p!obK+QE*CT*rJ7P+Hcoc<S!Oen2x
z+Sze5T?STy(Ai;rR`k#^iij#eVoFN2>el7n3K~x8S=J`PV2TQed;xy4riQ=z`xQhr
zK*wCig%6=^pQ5pqJ9ZOK>g`L>v_Y#W04+{Qi*@we>Bfa?N4^H-!eaw#vsE`|qUQ}R
z05uH_CkMsJltf?lS)})0@rs>t%RXQ~N&s+tAR4UAzF}o<K6Jlc2ID00if;gmyuDq~
z88#eeH#_@Q1miBuAnoewJF4i!g<dPa(DEXnU!CuGWHHXOUE>|QtXor-&|xJcBy4IO
z0zNGi*+Y>@BjQAVw#!Cq0###pv|iG#1nm-#ix>>qY@#&mxJH9%qBx-|SFQ5)wiE5g
zU9m>$`PCslUS1Rs&vh2AJHEW83>OQTg{MaZL?N-Zhov;UxbGB#B@zonOmza{X2H7c
zTu!+_YeF?0g3wcwI9gZ0DZtbcoPrPLEHzA4`A*Afj6HqJN&8g*7)&hObg2bYAkN?&
zUkbZtDg5{T{rfqwSFoip0g;i*v0#Aww<4xCq8t~zz2XrHCGk1+vNamyD+WlNTEvf$
ztC(8`Do$WT>me$KKqis$sA{ov)NtI730>CJkn3<Of*sS-Lf|mF+V0E}#uRRFjX?`S
z2y8aUz9TCrpEs1!KKZju<Uc|r%v)W?1e)1#W{hKFtTLV(h-9-vl{Hczb|u13>K?WL
zx^dpjzK-PtQ+OoO7+?+l{8+;*ZWYWI%M68*eioj^mrW*mkIc*CQELM!l+!z*ECOKb
z>F)L&y(o{v0oqso2=^~!h)}vG+VP1AB`i`mC%)a$1wze2;HPX`{5m}y6%;(l0o7DF
zMtPu$j_|0X+r%yC!pY*_iJ*QCm~buv0UY0EM!BON%xs{ol;0z6L5J9<qif}X05N7)
zMw;mgga=5Xj>_;pF?xsswV_$$E<Edw$jdJPzOHG%d&bJ@AS@$lYw^)9!y>3Vu%hQ8
zR3t-5jW{vVr;oSgIkrPgqNG%Ie}4zD@u8!l{a~SR0Q$&t-+%a!^JZ~H;9e3_7auJH
z5+<E1M;i=~S_Bc{MMIHky?giW9@+fGrRzR{Dl;i~t~FDi`(YP#b+^o1(6+aq!(rK+
z%ly8r?e>TJYrtY1SrM5Jz>n^`Rcwq34Y)rDur^!n5rZHZLDASU%F8V#RO)oBS|Grb
z5nzx~PBapi{cxUzOjCuS@}gOonptfOH0d~0$2YdjTL7s{NUYW<(TI2Nd=9oB&8_tG
z^f>IZ=TUl5vYPTt+n3FW1SQHv;IZ)$iJ2lD{r%!#&q7s?O4D-8$|U6E%p=s8yeTz9
zIdv6w4PmZzaiI>{T_<bBdPg)jG&L!<V!~mVJ%th{g5+B%t%%#wXCA@Xg5~Gs<)vff
zfDW`^j5>TgYqc>T@o!wiv1QUnlk1V|YsM^Eq3h5$6tXd%qT8|KASeM0(Un__)l2Ke
znvn;UhPn^dXUrF|w3u+U2Y}V1bz$I8Fzkx*3Egf;SB=64RAORpfG;*~$SsBt0A~Er
zLf7D+;^MMRlOS3UDD82692(jRv36|xp)p*0o{r}4Ck+je$B`AMP$r*K9}2!IdGgNk
zm1Of+Sq3I&u_;YWP2Bv4YSG?6Q1FTU6J@w;1b~XVA8Ve=u5SBhoeN@1Z&YgEx01@d
zJulhu%^YpIJhlk!?6@TY5WbO~UQU`q5hWkK^sx@;%u5Oht~4SydIoXlfz@twXw7`&
zgBUv<x0Z{J{qn`%dHXX+e^SF?zl9&5JJ6q_cr{lqG!D|ioOw+PDEhwt4@G@*3N@mE
zpg`40gfT#3#Ip*W%(a(*i+i+g1930#gKJiE;qIXwea!tgQBa|)h<{E6oD{We(G?Pb
zz3=Plk`+IOag3vYPn<0(q_GM4ba&J$xtUM_Q*vy+rb)+&vZu#5dxE`35^V3q`1yTB
z=6LJ2ZA$oCynbd@7FUwV<Vfp&5J;Dw9YPd_ct`883oJld5hL~O;fesE7Y!9S5QH_{
zu|o|@er*dH;F?e|9U@%H0Ee(}Dp1{{b$URp;^<M?k?#clP6Pd0<kaGIQ#x@R)~pUn
zg=fMEDP(EJD|u7be1JX17=XahR(Y%_dGckYwCQJomzHj|JuEIax@dKNIE>KqRL)H~
zaluT;^}vUmtZr{<X(61r_~*NsnVI)x^IKbq9{&ElbPy%g2zc%rn6`qU2cym{v-Pfg
zB}d`%`#7v#IC%1xI}6t6=;#!3Jiyf&xGMSKP*Y>(l^0OocYaNQn4NX&%VyyIO}W!D
zl2dD_sdq7Nc@L~^VPR1)0v0VAMSecil>8QSEm-;IQAtscp~B_s*U=?#TS-0@NLQ2|
z3l&Cbp{?GC(V@iU%a?CruqQ>{@61<yBE5TQXlPP4>l+w=n|OSVLg}@{o$nk*@X0Q$
zvDvI5<e=kig8u}Wa*Pgwn=X8ifNTvsfam4YB49?rQff?mgkuGs54-$K7<9p-vW)B3
zufy-}+PbA3t`edY`l2>qKn3%EJ}jdIOs?ScXG00t1LfLt@XVbxJus0tdNNHYmma}%
zgU*^mFxUXGRI5LM1{Fb5_!KdJF3ny<U%USK+5i&^tY4@~{1F_%6Y3fzcu=q!bUb&!
zRHCitrVRjUSQhQ2Wl;tT=4Wb9D0dHh_INiNhUr~u1|%<#4g??})kV}=u;DAwg~xb`
z|4&y1g;fJeK~wM;vAQ%`Nrp4%ONEmms9D3iJNJ^i@QSa`NL=-T_<7X8&}Z)ufM-4T
zWCZljw~-P4>#9eO7UAT3QI|I%TB6grK%zZV%X;5&)kP!WPWNMQLg=Q-x|QM-7(so=
z0XGn4l@tD{rWo`M{sNW-VZ{+)!Htltav}NNAeWesP?vO4Nmx)5ZKpI^r2Ao9m6fSW
z@*{D%ctK<sALUbjMyW?Pf73_8D<+qMzphO529A%97ypi{GzIs(>mQ#@3=F+!J9AxY
z-8whqFY%aeM#(=grMmOid+dh7s^|f5&|vXsv(3SqMLO=nb4oBFjb=g>w{hFHV1ZN8
z<X9XthkM6>^-jM#pwn_u_<rv6UD!a}8M>!Nz-_9+f{bX!g)64rLiHOO%VVpec@eB?
ztb_~PhtdIWoCp?#+3AApo5wj4BMK)=kbEzOTe(LHR61E+gGQen?f9o8-xrjs)|Jg?
z04)lZe_`Cb8B<Ys3QY~+64<u}<f1DktdU!e7}fz`g#%CAsNu_Yz<&CPa=BLo<vHvI
zv2gD%wdc0h)oH(D`V5SYo-*Ed27<+}M0D={0EgqUGFn;ktzwaG$`{3>-%u;}gt}Z#
zS2rLgCZ^#bE@iPL2%t3Ct#>6!94n&_X+ff&pPzU~hGnbdzJ31ZPV)mXo;`Dh?N<l5
z!NQ^<A8KWFBqBS1-MM$~Blzc|!h!;TMxavYZ>7F?(S^1c3acomjrPb2-XQ1oNgc{n
z&zYHV87(9GYiP(^lK+m^AVawtNrxQQmC^E&Z@KpgSYN*mc3@XYKC_&hoTypNk>M&(
zBZzEw;>EX;HUVX{w-;v43kwN=NC9y|Y|g550Aq7=^D7yo+!uZse~$FFMs@h=^-M>2
z1zBpUPy8IjIJ<fTQr^_%Ixsfi8Vij=LEPd>KY_h@j5fu0g;$K9eB`o{3Y)VNXO_Aj
z`2yVfPQ696-ga4OdI?<nBO@ux7IEV{uD~vy5wlLstJRl<s@6lxoMP2i?9~O@ei!Z(
zqc2+kk=?nv(YBSYpg1i9*YwDImyz_kuMJr?F%g^2U_lG+UKIm8D9om5c!v{+y@!S+
z<sr1XQ%hC>A7i}rE}RMof#L#QZPps^djv<+d_WY3UN(`o-+5DHM3FduX3O~j?%Z|d
zzy*LO<A>?5Uy~*_DKaYP7XWz9^=lj<G}!(}Fo&jW;2=Pk;1L;@&O19&(4|45@#0kr
z$U*~>@A^j+Y_n@a3F`X+FOKpaV}vS;?$!%D3$Bps%E8Usw;O@7NO=RB-TQx-`toop
z+wc2_2&FQYN*&WN%NR-=ky*wfrNNM>6iE~rl1!B`37LmR5g95?$k1e_D2aq7ijerN
z+xz)m*YB_Qdf$4_InVvv_rCXDYwfjxsgY-1jKGt_`6PON53Tf@o$&*@R&)0+Qea#J
zalz9suBe)P9R`f`cjB9<z=|%cRj#r{#aeqiyQ9jBr3H9+tf1)|?Zo7h1(z_L<Zgzf
z?w8hUyvdi9YOFWRhdnNY=g42vfbX{$7{qh7x}zV+@m$g?c8KC4SoM#=FMd?hW(=Y#
zuEni#l9B~b?>m;kW{oaD1?|Y-e6kBbyL-oNGdkeIt6TG}9}L{Vhhe_G7%nDc6*V3|
zK7Qop;$p}ugaDqzUXAJ@Hip&Aa=tLnxPSkC=M?x89a)NMF%S%?K>i79m=KbGZo+-!
z877C7Lrz=7Sd8LI)@K$Z5Gj4^V)8^pM9`D9&PVh8b2hUI#4ikUE)QH@IER9b72qqD
zvA(($z$Nh0bL2g99>gZ;J9`tr1l*}Z?K0B%KLdJOw=@_550af&&Spaxnzuf$mX{c3
zTw5E*sp5i-jl^b560<-MBAJVj$#w%xk<wN0oL|7B=#RfsQo?8c@pZbkx$K(+fciwn
zIPwB_f0|v4J_r^GpL>u#!%Kr}vC07=u(LaQ;zZ!|Z>jlDLxy7{7*cP=M?;x(jy!lS
z&eetu8zSGa<6ND$%h16@9lYxszH$eVvR5;oAq7F_%beX3<C`3V_ah(LxtbgG1B{nN
z@LHsRlb{<V*3|Sle2J=a&tAN^4CPbIN+kTFKYw!Ncc6&lUaSItjSeKn{u$wswks|{
zk6ixee(Q&v-3GF6l+w1^qRH4(su`yEj)=bM4P2(Yq!eDU9sPVN>5+q9QGbaD9O#A?
zt2&^fr>6%#5fQ8lN0Hy*TO$p6Amia${1bZ;>cU7}Co~6Eql4T9r-VMAEZhX55q=C4
zP>zF0D53y`FF&V`P~qt06v28er(*=<q=?TgRNCB&f0BB*dPD)nSIGJ1KPOTt6m?sy
z(eyka@a!#W)5(nEpl-1XB3LIs%sC7ZO^$Ew9~AyHft+USf|{)ZLqqjxQjC)0BilWr
z`Jg>lQAx0)DEbI;SX@zKFUDaz|9>C40A@TSt_(!9wYN{8o~mZ9<N3ZG0v`cKok$@V
zZ#Fa-f8UDOLM~|m#y-^QeZhYSV?HuYe&h~x!+_7zBYW^w+RkpLVDfs404AWuiFEw}
z-8cTXCm7ZJ{{8MXbQljm2>83ch{lQ(41a@u)d4cMRJ)*1+?If)YQ!5L3kheX_7(%|
z!d1YHo1&~dl%%ob1v?^{;Fs9Gdy2*X6>i$<CsrsKA$f9qV%sz_j&4^+gDXmK#-q}I
zjSZNbIB{b1!YY?IbZ_emhDbrom`BnTU6kNazQQ+<e34Tu(0dHi5u}zT<DQD;sA>S(
zQ7sSkuCk8{Q`~3BTE0({uA1-fnJH5N7xujQVvkdtMtj`Ww(aUR<R`uiW<z@vYCuqB
z<P8w=O=J!8P~=CP+6N|e5NTpHvwssFta?FVS(zFh#Izj}q8cUAXA~EqXo)K{DD%rW
z2~rV*pNJ$^pf?Iy$OJ_v{ZXNE>}p{Fe@k3K*kS`$ajQNVSr+6R85MN{n(^`j827T4
z9JF{;`oOWA9ZkW?Ebrh1^h0hXtNl1reK(Cg#bf*<my4(zY>zhmxdmzS^7}7=)7G&V
zRw2KRJaGhDe12<Dd#_Y0v6z^cG;l@hTL$$G3~V96?*(>Hm(c>l^9auowr4=;ZlxPO
z#l4*yZ!nsx+}yRKh}z?V#qA3&1FDB4L%}RTIA{LGPzZ;Spr_~wBwDzGvE8@fH+@2i
zcriO#P+9Bb?{K@A0@P0EN~wETgS`lHyP7$N0fM`xks08r82WqMu+t`Cs|SQ=z6Kn}
z_z0=10Wg*fG7aKGy<e!Fpt)%ixSj}B4yd-wze=Olk>I7dpjRbr(<vW66&GAEl9J)%
zKuC&-0pdQZ#Nsr6zRr|GSXx?Is+oI9dRt_gN)X=&<ih3piI;9VGoM9z9PxK~8>EUG
z%^2<ouLhbwi9rf>d!8FIu>|22=aE}s3*oUDiy)N*Ua_Ow2!^BSc0qhm;o*|dZENO%
z4q3p=9L`#d4o=Swbv8COpZm;KKadd$91$T53RyKBXlNV-nNK`yX+@z^J@5o;3{)_P
z&#wGvrg|>uzYC^8HmJlCU{j`f6TE^EJq9jCJ2i(fpyjM-Q0u8*pU<bHq-Yjns>upL
zK_gj;;?0rN#h9+8$vLd^!6_W?Z+}CA(N!ySrVH@$>dG>P(b%K=!1w|Dq?LmEVW)xP
zny^tHpcc5t2v&t~Fe~9~?}1Uqb5M_2EW`G${_l!K`NN&x3XOdX$6hXAX4nJqY_@pA
z1cGNpy%>KAzqAG%NF9p;-^1N|K(P6RcXf8c+vE&+7vV~wB?3Z1X)?C*^Z7+CZnmhh
zWkk?M4`L!P@M=f!31YnG-ys+36o2sWVLgk&O<Ys&-o9<fe$momUD(097!nD|1}<zJ
z1A17lB;)7!xH$HIRXh6ca3a>d&Id9KoK&ue3L8M)e^E(BMn+p(TdC<L;6c!*_5Yrr
za-Bb~E4s-Em05^X2xwUl$4XQdF6?e%@b8ml6hQPy!qN0EaD=w;9D9#Ke{!d+EQyV%
zUF!7i=D|dA&2dz1dUVE02Ng-QUu(8|4ZeCQlgRYn(<;Ms<-bP_3Ze`Q4E(gvsJ1?t
za&v|}uWyWujdIS?K_%WN=m=X598%I@%eCwQ01X`KAP%bp;qIz{6jQ5tMw~MXC4b*5
zX<3_~5?)?jYgo<rRU6IioL=tObOi{8nSbyYQb2r+HNtE>n*&Nm!1X9F_PTG@)m3(B
z&=U;VM*<OJ7@E1A_Vx|GhG25Dna;RE$h(>?U$N=y^{k2M-QmM>CxX!O8_yAgI_y6E
zwhvs3Nq)quf=su8i!*#a7s*21J?20B_3O;^_(3J6=6NNMCwQrqD_0uR-y$ILR*Va+
zUJY<UhaSZ34Ry-Ra%^}y36L)l8mpB;QKH5fxdUr5@b?+PKX5!AcVG}yx_Ua8fn-xk
zR6TI^f%<QlDKO|OflP7~4S{wwS7M4Lcq4%Qu`lqA=O!Ns6grHDpFdIAb*sh;)ao&)
z1op_t%io66*_i(o3G+2>Vd4Uh)$#G`E^6)ENl=Hl2mu?6?eM@=@&`Q^_*nf1bW!w{
zt{|Vd1fRH>fG*d{oa=c!wHyHKCR6&qFyR1z_0}aSE&v$&ENwTiV~?1k+UeV0hfY=6
z|GmG;<58wn0wn3k*^TH9p+^4{7=y@@53kx$PNaQIb0IMI&V-%g(6)({5eWZNUmp%p
zF#+$?#xAVY-imd%))m}5EVYOW?*1nD*P{!UTu!@l=ML7Z@!1`0c;_rfvUx660*8>`
zeLTr$Zb8(;2wwC9$W#%eFC%yXo0wnT<{@vCycv;e4NpWa8i4p9Cc=IjqM64Cw5&+g
zbCOWBN0t4;3os&Io-L_+pe1ifve*Y~5h(^RzeIE)NxjMkl`hWvm~ZQn4L)BD>DLJX
zGFosSf<kNt6H`-D5GMVjnBnFCe&sT)fy*k4>>I0fcqP!LVCxJG8PORVR)PW4T!W$?
ziI)L=SX9VuVEd&Dj}dx?4gz)PGus5YFZqa-HJ9=egvI8wxSaQ(gsP{AJjj_jCmWc9
ztE;P@q0jg3WK)~ZINN=eCos4Se^v-+81RpuO?%;<bh0FdmSIPKeKu1vvd0ulz>@QK
zjVh3NsAxo9=i?aBd1*bwc34&y)5@5V3yhE{_D=wF2-=!2ia&$Ec+Z1UAgbp7w?3GH
zmkb&<h)fN{PwhPjw$v=u9PM2UI-~N`--*LGkORS8=+RNKJWp~5gC4x%IF7xGpu4Gd
z<pZu`j1S{gfseM|z2ZA}GB|}fP|>2ACl#%fMYjGD#1Ok8kzk7Mhz^0VPH!(SoJ*nP
zf-_}hW!=Rj>p}X=`&iEu&9rN-qE5v`JUaH`HMlnwM@21mVXs#Nw=vq3h8K(wPwi=Z
z{n|X$d_Il^ML=EvUM%+9Aa3kBJywJbAt9lG{X(kle92VaCtyQJUdPze6!tNb!{L5)
z8rCbCnwqunxe|Q3TwInhIXVYYAqp^{VHM&><F9KaOC!#o$0qT{=FnM_iToxiDoT|1
z7y-O0%nlzq1U-ilJ;;ctWi<@}Y1B8fQ+pSoMkSoGc`jSv%qZItUBji|&cMq;p*u6A
zdJM$g>G5HHZYJvI7J%OO?%f;nmzFccxsigbZ~5|vh_0f%cLgAi!Kts%L4{*T$1<E8
zzwX=n-yx=&Ex|B=M+2fS_u7o;*FMZW`gIl87=%uB#tC0mapeRis0Dd>d94J2i*5rD
zxG^kll?j8^pwA%aQerYS!xkIxw8y()Ry>He%3o-TXcucr@!G(zb~iUGrND|PBM%q<
z@*r8utm^zZT+^6g!tWWe|5d8gXCENj=5t`WPK<tlRIJE*$jbd86J=-vC<T?J<`F=Y
zXYg?)#QHd51Y98f^9`htUe-k9%b;3HAfg|A4AOiDw%v*@U}{h|#RvxjIIY{GEBJH}
z+)iZD#?u%IGo=k<B{9=daM+~WR--6Bgm;-&mw{j-YOny1|J;yhvgZ0cmaPmOp#zRH
zR6KZ4Yqa75@%L?Drefrg+k-mV@8*{Lb9346i=cpF9*yBf2OW*$`MZLdC1Q8sYb<G4
zH44H(h$wQ{&^!mbx8)MX%uJ^3;JU;wZU=TqLaw3#Qq`n)O+7sh`7N;=dyA`+t|Gdk
z!xigDxtZAyUW7%v{ZH4)=up@1_<{sIrl%6<!5kdqxuo<JT1sdZ!nvs+gd-zm(Lz_&
z7f+!YL%i*0x(}lsCB=<^=AhQ|5(S2R*_5)LK<$9c!0!cdyd~dZklzAwT@tG(7zY89
zbd8!-SWHY7h3K<5a-BuOLJ4Q^V%Wm)`+3Sj9q7WuhyHp+FraAdbY{L>49Kz3=^idX
z7Dpd^iR9HpB3Q?J`-5K4Iw$lUzGY#C;0>o7?S^(?>!x-oj*)`F;EQUN&M5L10Oruu
zDmJZy56zFN6}}jr30XY4l}(njQU_ujKS1?4jn(Nx8>rmv{luD-{y7&>#_JaxeIu|F
z{y?@Dx!f5+cQXRXGdK=xhSpR`>@^&>6TU11Mp$yF%c?OJ<0?dlm_yCUC@tsO-N+=|
zj2L_HB`+Z{tmbHf!W=(5zh$U8>oZz8<01q(mMnn|AU@>?)&oyl1$IzPP0g==K2fEp
zurXk%c*2NNivyd0J!}7Ik<^_OOJ-${!zuAh?_75s^MMKr*ZmWqGKA1(8aMa^wIL7t
zUIEB4BrkRYQbC_~d_)Dwf4dR4e2^gHi_Vh1*E7Vp5f>)j(=|RG9uc{(hYufC_D6X3
zLV6$Y-PJ1Wr;*ZPE-km_9HX#8F;){rb&9b#C-*FT>NFR;N>&CUahIqQJ*Q6MhEh^o
zjtGBg6%rdC7Us)QpaP@V-b1_#mW)<|>H}=C)b3&PDs)oO|2>n^otzxawku5|o3792
zY`v?KQwwB)c(Q3_icy2NDf}vcRB*hHcOjMd81*;fX{%=7tY-4SAZ2O+1!n)~Uu=C|
zc8bRO`{X|WHON)dg>!*V#n18s1b_tSfSfY|9Er-a(@0fuvFG2^fY1)ET*zp)d{{^o
zSfLW{-%*Id3%{Cs^z5GI>)Y&q$3+aKVDe_SyN8FzhO`T&l(L9cb22n7>y?-(e>n&;
zQAY1KJPXQRVsJS@LmdU55*aveF){+u7e57d{bX*Q1Nu3kvt8+v#7Y&I0tqhfAzt;!
z!bK5bA-5RgdXF+R`EnKR-3Bud5Nga!FB%D`agY_y14y4SCJ_rPWxpZmjSwH~K&#IC
zmoLpyT-Z^i&0WWqgT5+px?Rflb~iTGmv0o+W(Sjj$@Qe&Rf*|^3HEFG>cjecY5@?F
zn9zf|q%5;SH6zb)3T~!RLH)AKq|*_aiPk+CmjvP=46C1c2rY@>4yLDbvqw%`Lv`t_
z5DG)Y_ozFrE`yv0@OuOAF67kgED9U@E~x{G+h?+xm!xwQ4^gaj`7Tm0z{<w4dBLcb
zorQ`Cg$8fyWd&itp@N*%XxN{6gnD$>85edb8OEeD#2;bqQsBN(zGH5-(juUz592(&
zy%Su^AalbB^??T%UEUhpfH6JqCQhxFQz<G*26e1|a8P26FBvFJhZG4t1PGBtIK;WI
z6eV}z^a|`kz}Yt1hjYuRkMRJ91n)guEhL1%_V42}HhmTfFgEgvtF1S0-9kSg?mGkY
zjKn?nf~1VS+WBl)*xSy|y7S|R1jsD%s^ylFE^C18XGtF>+~Ln;YG`N#d@DV$MQ}5C
z>bA5Q`0iI%LnrpTR67&E<iaEB$b?CQdx$591v1&uYCAi};~enBTh=r}z7Jr?+3`xX
z?Q1~BdU|@O2=j>GWZB@6s;Mv}hs6-=8V+^{HtPePu=*KWVjcDqL@Xqcxn6~zq0z{*
z{Kda+H<<z%ji6oeX2=_7Bu=83d@1(tamX4x9Zr%wO3@74>rx6R?)}*Fg&@IDokG-w
zgov>o(sY0gCFLxnU6Rj#%{OGZRg})y#J8PAM>~TcM3Z=bW+9~@N`^t`vZqcl?<e+3
zW8na#G=nea0S0P#34xmkwy`Y`!AY@NTxuwoXk?f(aFTt6+9pv(Zp_AAgG~^X;#r4D
zlxe@2{N%`5q)1_&Oq9YppdE=4#s7t~_W5x31xW>OI7cN+EPNM^VstQewmn4p{oL?$
zb7f|tfD{}>9oiF=lmR*-p>G+W^Q+<}0>_ht0YzKg-4ip}Ue|iJv=k9aPL!G2>jThZ
zVCWFkBv^)0yvoRR`bk^4>Z)7G*jkTlqLX0X`yFvaOpDJLCnXep8PWoa^P6$ft$yvR
zF3aCxXNNY`8&d9Dm^K4*vHsP9Nax_-u%wwp4v65LWhv^#E9j<g-MVF6ktjm7jE;UX
zz&-3|@uTbRIa5<J5`Pxr*_H=WpT8r__l*Y=ssvH9>Fy6;#QC$Ee?tl%-(ZgR42!OV
z6)2M33vlXD%v*nO$D4;E*qUr8NG{;A5Py9D*(7GDI5S_qkBYwW6DdO_xZ>ew@u|OG
zOzjOqOg)Q5FdnXvj5qBf7ihxPe4sWbqzkva#w|9Uu6GL+njB+N&;NvTgALqW0b%_P
z1lLGX(|)5{s8Cs$3xX!SO2CEih8t;MwsZFeL^FUG&ddc2oW=gko&rlzycV&20JfVq
zD7bWF<sL!1uIWElIGLZ3Zsg+J9rV)(%&E2i-weEONE^IQQ9)DE_9X?Hg+|tL!IU&6
zD~fy8PeZbyq$orzEmHDO=i$4b;V}j>zxDU$v<zW$#e{-uQ{S*5uh+{P>a``~f~~e;
zF(?t!qlGrEcPU$jsH1y?jk;V4%~iN{3R<d?DQT{?Yu9O6l%SObpAHn%s|70fXa;rM
z7huOhY3i}<jv~h`i*6eh?ksy{v6cbk=JCk4p)IIK(8i_v&$(vf#W=_Sw*Eeb6UUpL
zbnMtMsI;&Gy2PTT>D#w&W;D8pme%{4nk^nzaO&bXGFB3CyTKpe{_|lGX)N4k{v9kd
z+^C_kMNe<VPbW$Fij;Aoqf*H*%gF1gGDZCR_jUr>Q7DA89|}QHiUU!1^<j`SvDR7B
z1A&q_7q2I3uhG;#04w<ZTJSQm7l}h7?#JF^Xly1Lg%H=bxfQi{()G+-*!%KK$s2Y?
z_kh#^C1mv2bZ-XX(gSSNeeA({G><Ve**t^O_L?40+hGAV98&kt;9B8;2xCACx<WMI
zYkI-FiE~kmEbk3z<+pNkaOa_j*!pVV3n;1;GrtP5IRSeubF&YkVK8G+Y3V=5RhTp5
zgr>W`M;O<X#C8sWUV8odcHrDsTM1)L4{}FtO#<$M?&r^OLUf?h8PIp@xGfZXAhfXX
z2w{}<KR}}3R?vx*qxz2}2o!Q&hFCL6JOyzD35$p$E)y|%Gq^WRPWqk+TvJneLRn7N
za0Dj@0$dD1T?_Eta8DO>G@~D6028ETG)*y4`v-BS_xXJXj#8&*1}16Z#|Nl_QjuG?
z;nqS%VkLt&1zXxXke36&+X40mXNSYM1}UiI#Zl}Bo^UBihsu&}$>yJ8TymG;77TrD
zzW(6SkV^1#1htCBjISW1UW*91u}+ronIH?s0ujpc)e|LQ#V2-Ns|}aPi%QNX?AXT8
za*9U0V}p<{{t=p_?7M&IG=mD=y@`WT*hs*B&7?$0%MeDZH|r<xMzQT#^-^b0SwF@(
z_glR=5re&D)QE<PL#*fQxpPsk6R{gFn6`d{ycJt!`Lc>`9gBxhO4YsG1aoej+wh#k
zAcB)x<5=yk^fnq*ZGO~`465l^8Q_roCN}F`!9+Uj$GQ;?-k-Xe;F9He?3hWX3KC8H
z4O@RtP-D3LtJkf9n2-dh-2gNs&7DE)h;ki=@YSHmv8HO`+t%;jzZ1e^Q(~+DWHx3j
zao+*Z$iqeF)?sgNmE-~|SRbH+6<$;+Kle5Yv9zrsCVbuCk;J%A0HK;#x^!uu7S&v_
zGi%Qjo*0y^<aTZpktQt{^Cfi+vmJt(L;5BR?6o(v0V>$i0y8vXAP{kS6t52Px@O%~
z?LaLjsI~Ry(F#Y@Ai`Y7@;gQ_&M-lkAIVwm{_XB#xOR#Ha+SH2jP^87g><imHj&14
ze6=qSfzwD~phUv4xP!z3VR>tx0{N}+CqqakK1k=H9!tRZ`8c{(Y*~o}v2O&OY-qcY
zJpB(Rz4<rql_P>7O^dj+G~hrtIYOoYr1^-Cj>8_B>w~%L;lo5-OtKYOQhcVkU2a;i
za55FGTVf}lGywZ2lht4x0<A&c)HyvhRigjecU_<GRJE|J)eg>CN9JQ{Kw`g{CXF`n
zSA~_9<R_A5na)1+JOan^MV<g%V>L$>9%;|aZ2<05o+o5vAP(A{J3grLTqzW!Bi65`
z?o?TWc{E{s#Kiie_JH$tSF}SQL6N8=4}9O;p0n;(Qf{vJyWSLhL%Y;%Yoo?E1dN_?
zPto=3x6d@dWCocudVZ{BGx^cSC#J4`*|1H=|5OLw4M*tbpnVZX`1+PIi#4apXu0}A
zZ-ip}e9MqGuvp%X1)wUDW)+h^Xq}okG!*&}GARAY0-%NaZNv8V<CqgI>A%}J4H|Sh
zJ^2F}r8atum_u(+boB<((vACZp#BCd<X9ZrG1Ks{tE+^(Vb#A0f?(4BIUw#F1mcm^
zQC@}W5uPm}+nCmzRFiV#<`Nekj;R(3MG-JD+!nQweWjR|3zpRi=jOciO*Ui9JORI1
zIL(2|bh>*LFQ*4C!TmO28Ila}<Yzrk|G;I2(MvzKeaPw<LC;6N`p{>itx)-5KU^;V
zyD9i`AtYhg%26h<NQuSfmfS)%fo5aprw8{SpQZ2Q=TduLNN}s1Y_z(0hC|z(U~0p7
zaAvQoD-z%;=)}-wGt}Z!4rcAs(vXD_CkDyXZu~?!B3LQsAzbss*JP={Ug$HCNY<}~
z3uA5G4DnG}H~O*_YApOqZoId-xp|Q`3V3udDkEsoP3$@9TH2x6;HdVO^7$x_N_ou0
zQQHr2+|v|xu%-H)ZoF>aaKgP(Kvx$t%nY!i0?(Hp&RmM<jQv~=_#<m)uNRw6O7-DI
z&?4h5R5=Lvgiz`^(Bd40T};bdC(1a|q4assw2(Qp720+)<O?;vew~OW4b}PWm|Mu8
zeZGOjr*Ba|%4MlZ;o#ulzHE9r=W6vB-<J-UiQ1X>fWA=(x>*B{a9}39;Q~kTwP2y-
z%X~p=aj;M&Lr7!#q6VQ3%;9({t5s0}F4~qU<SXt^kS_uq=b$Kaf@Oi}F(aDRVozuv
zloYK2?=0_FgAPWM(TnzIpXxKV1rlv&$675x8d`N}LtDC;%TR5wWU+SZyvf6j5_x_N
z!SJaJRqU-}2n6|OP<SRc4XXW93PrJYSB4QIgmI6xwfeI^A>V+4Y+ozB_9ZaQ{w@s3
zk)AsB`B`gthgj~}=g<OJ&19q8F-efjEbpx85$7*J!qA<8ghV{%cORG<wQ5`;?0SpK
z&tyPqyw!eJ{rT?%>!LEGfOM?G9aT=nyn6d$#M~l5cxLj`#k>iQmnD?WJ+A@dqI1p-
z3BV6`Mhj6vM#Se*a~p2G7{&`5XXhM%0Nn|wD~$)1tybP2-RIbWy;|SZ9Z{Byyp%)i
zAv!1ZJi-BP{<q$D$3DN5QJcnslJ(ZjZ)wAb(A0n6f&P@PyROU%Z4}pBB<RBASFVui
zpE!0bUVA52{bkV&Vlib3x?%=Fc0UUf1jD$62K=rQ@W-+0MhgyMgBhlVytSEqj5U##
z_knwQNb1WT070C-DI8MS9Wa#tHV~$MYC<;o0MiLB6$16`uRP2O$Q{47=Gr4vFjx?^
z=i1c?ja=2dt05XxmP<u3fTBfa;u5wEbb(6Y*!z7*r;{YE0p{QdpIXGuUI$E}^#XEk
zZ0u_@Zb*P5qa4T$=X?xepA8i{xJis?Lz_p9M{rD)nwY=S4s3^u+kie&09r5=q#wR>
z#IU=khj_$R1zvEwBwUb<Dixmpxbh-}@gep&b_ocm8L0oqI8}YV+5r^>HUN0~2qtOj
zoW9cMnAEz>u(KN=kg>-b6j3~el|;AKy$Bd?>hc3Dc;<u1xxBJhxre|AvK-m69k{CJ
zfnZYISP!%ZHUTRG-sj=&!agj?h(TXg9`K<)uOj$@M5C;hl7xs=a}$8_5XOrhJUvd1
zQ%qY_pQ9~KZA3>m!jH|NmY5hWB}J{#Z_04|5#c)DyEJ4!WIB2tzwwfI*r|f|*vQqk
zttgucoSK)bG1h}eR^_MFLG_Cs_Wg>3AfJ@;pvHYgL;@)nQ0&tt`*9m!Q3+&Sba`Xi
zRtyyrFS|xk_?#d0%a5ctEybF)qjVb3$FbUoHLXO8pezngU$i#(!s9b;h$F;FJwW?4
z^tCnbRM6C7ZSV@O;VzK*ITdKy<Z;fLZGC;q%kTLSa-ugLtgWQHj~SR@I~yAEKCcvL
zNa7jM7t3BB;KhQDK%nJ#-(brY)XfExGnGg^`}RG8cF=u#0ly173Jufe1vShpiQjk&
z)CFHBNX3_=?)7J)nj3K{YVJ><-f<2wUx>kj<(<R^$I&tvWnkms4d^svR;<E?t~XOa
z108b{JJ~b9jyk{N`9yiq(1^PM=s@^$e|SlxZOzYn)a2d3)ru?|^OSFPCPB&%LzIL1
zf)vdGojtp*#ue{>iIo(?m>}e`eUUXNNL+60dC9YG4>9Rvar7tjX3AcO(OI&%5v`Q4
zBo#d{cffJA8G=JLx5PB$I$>weUi<M{<*ZUD8dh!qP2qjG`vVNF6o)~yS=6&rhRdN;
zE!12KXZ{4n&w8K%e72P&;gB7EgUqX9M{dh%JRJO>>dfu{68`Vuh~GoMk2dH*yu4`;
zTsu++up4JV2yUCr&9ey0n8tk$MhDH<A`>^DaQrfUGHey~xDI%(h`By0Byp~pA9IVi
z_}0y#Q2Hz+c%67*R1lSx39U9sJMeS`$9zy;`9z3CPKKBreh0CD+QlLncmS*f8qS>r
z{{Rf}#^jGQ)q~qNz>5J$8}{MsH!$hjJu%eyk-%#qF%as`)l5|MAv95rnf^x8JrBDW
zDgJkTw(9!(NW}x+)7O3zYYIiB7w_csG*7rbuAD{!H~8#%;H!v-yt1pxzpJUq(qC*A
za<FZG(r<#lXf23U4EzhlZ*>X2>-P5&NRQK`tm&RhDxf;A6`~$r4&xLQG1o=Pq#XPh
z;4UtTUFjs$Ca&^zZ^!l-s73m6grm9+1(|wPjxw{-{O;L-!ePJ%&J0TUZg;?_yGCcr
zQE?pnxXd;|m;fj0k8G`3*^7~uME<ae+uGU1$4%`uMRuH&pNUk4H@ueU91+a=tll&=
zjS?YnHXWB+;yX1=%wutMCaZJq^6$WUJyBdYE)JF6vF`$4lK^~Sm<u9C2o62He!${J
ze|7>`VCDV$s~&Oc9K?6x%tvjK14*SB!9poqA)x=U7?tI`;g?WL-nJ`p>?^KrXmCf5
zjaW9>zOgBtU0unHXmUOGNk$POj6|ACOI9c_JpKTi;6|ij0Om!*Y)$e*{0H298#z|s
z>r(J_Xg7%phPPBNj$yPO6~v7c*7|A?tQ;7$EBvsby5(EL8UexzNxz|fKzT@76@Wak
zb3b}}8ML=#`;bG{{T;QRFt$&yR={+4%SjA~0Ns>-jRhPV6iNU~^Tv;~9s;_3@?jOS
zU&lgHSlig}o#BQ?m&g-eYZ2)&*Wt+Ua5!sPb!))~oV&iD83|9#y`@W*i1Rw}9Aw~u
zP5=AW6ivV4?(a~Ah;mU1CveI052ESuE*Z^uMgxe|y8|xlk<2WI-^ap@<|a&2F8g8k
zL#3<BL&fBwSQR%X-gd%VOJF4b>-TTI<-$O0aAVR1OHcltxYgMWL>eL}Qf`>9=g`|o
zUL)!hmD#OLRFFzBw7)~qMp}2!K$yPq#MTtc6?Zv_OMKsgBUb`;60JTFi^xmCPVYql
z3zc&Xw0GUFr5oQLv^aY7j@`L2x1EVZ5gi>#3Q##PbjOqw^B}yc|A<$c!YDbjk==fM
z;H<xA(jg?Nzm5ZAk&uuWwYpcEUJd0UnhCVzPMbjTiI<48i&ko*=l~RX(PmKp!$mM%
z6Ek@qr888-6%_73oY04c@MoWp+KFX*AlN!)YZEbdv>U_)UXy4K5##zswhe}^*?Igy
z>C44VUA~5}jVjkV^(8jiwNcY(>o5jW$W!3zz=j7*!G57MeZ|g7<#2oc6Md<}k|l)n
zS_Tw$`SdNT+y=ZnBmME;Cm(ix2b)3Xb;2mQ`>RhwfWV<=ZQwvIDkKvFK4caZj-m#%
z>CcGK#nu6mGYF7?*<8XYj7${gfsZCB>riPE?GX6|TfG`U-C`5~v975o01#7lA(Y^S
z_MOZ|-h>Af%jWL;C5<{woXhDgT0vG<dGi?mw|91e59t1mwIX|m@OGGdcSijpd>$l-
z>U{Q3pbI<NUSi_i-ed&t3!FtEAjvRng~GjB3uD)wfR=Ft51M-S?p@8E0~6!s_QwDZ
zBhhI1=HcR^iUQ2Bi4VX=#ezFW5@TTzZNKyp0qU}ZaF-M#d^UXBUF>Xx*LwFCnVAqz
zBpK*zH;`3=dY`FK4hSt$Et9uOxKE(VSM)>8ag~z=UzwJcrr7}>gtH8rBRJ4#*!9YW
zRmcz&3e{PveIl@lNn%7Bb-(Bb!6*9PD>XX&a>9k$Ta{;?F8`_*3_ga;D;9+}-{#Z-
zc84mWdAr!UH0$txccJnC3FZuKyV!2ii9dfNxQ?g5|Fz^x+|CCm48JB6m;uQ3Qd5V~
zMeqvVl;i^fC%8(Yzcr#o<XhL>=grMlTG8Z0Y&vzxj>kAkiDeO2oT#LvB!Z&jB>F6b
zp}S=QJ5ZDyApof&gocq(k;v?d4YKw}yh{?CF>^Is`Tu{c>D1IzME(mDLMWu^m-KSC
z2r{#>*Ncfm!6I^7Ik%JznnlhUk{>yuXf>l<2LwvG=H(?2u;_9>4fXFHI(&HG6)jDG
zJ6w^}CpQ;}{BNscH+5%RW&lEf&yWbTc75HOVEy7~^b`=q61(~O6hIq?1ToJ?0phg_
z>y3aw!&B&#-@o5z{Q_(zyIue~*NGo*_+kzCMgb#;Yf|!5mbu!mSPUG!;UkpwN&g+_
zdvJZO=z+=#e97KvxU^cpUemMhJ323mNQ1ov2-wz{5`1cdc0+S>T+5|(_Cyiy1XKnG
z)kKyGYHf^{xTyIEEvYdK8#03o^(MD1H!t0kCU73%pZ|XKyrTOu29wawtxMa-9qvQ!
zTi~i`0_&0xieiy3fd;lr1#o{`?pBu3=O{i9Pp4h0QIuh$2jhH4PQl4K{9zOmbxp!d
za;%@kJ7&vx<a3}^c6sap1cHl-cQGKr>TLs{-s-C_WBujM{rLP@|9=sB!}06g-%)kr
zUQqDF*5c3qic<joh{wbW{oMm7T*gU|4n_EiJ?w@30CFMxVHs`)a()-UY=}qPxK6nd
z{XQnN>!`+;Kw?wj3Bky`JT}HKU%M|30BjQCG=_@BJR^90@BudL$wTNz@VZ~--WG}1
zn}BvV-PA%bkJISIax|)HO%i^9c_tViz(>n5g+K>>>(z9e;RV=DR=VA9J#n^JvX-Lh
zM&Ls%>DRESs$S&IgTaosCD4**La+74h?*iL!@`dEi1~B|M$qzQr;tM<MKo>4<%b@I
zfiw5e<A|C$eBcHog-A&LlxhbvhuWavBGT0wmqFOaL8u(#NZ8`R1Yp<P;fv6R+xX;(
zjz>Ghgrn19giDf@r$Yr<SA7}4oc!mtuz#V?r*1@D+Wj4Ci_}fQ3=jp;s&T)2`qU9A
zSalpTAWq&6%{|ALLxKASead5|gigWx55IK>JS?JEjwzvBIEa-Ii{p1NnfpmSF@z}0
zyJCf*EEifBcYGH{;e@0n@{ffDU!f9C;`>QKtfp1B4lxd>eT^Db!n_TNlf>@tpft$~
z*8y?bxBgBlEnnZgfh#8hHeF{XK^fxWsIk_D83Ork2G;A^LF@A<P$cEe0Us8Ge1u_b
zdd#+7+yq|B>SIOOf>)#?=ykJ@ZX7Pv*sU*>R@5OAs<T!_KZRV`S7`kJmzcf;rh>af
zVrw;WE#XV8hC4{02!o-LV-Y_(3+?tXQ&zpCTy>2HF3%mlV<$pzSV?QSr_<)`!N`6Z
z596pAF{5CMxQTn1JN(K~Hl_%kJsV3=xgu7_^MD{p8w$UB6jDm4?~UlH$iYEpj+f5w
zV(3WYouwnzImN*Q=<@Q9Owf$t=IFncgVzd7t^7<-|Gf4fkEr*!J^^J<=w>-|`-W1t
z4kHW9xM<l|<d$Ur7$Xc+OO+$pgGQk`?j%x)x$7jDHgASR1~0WH3|`eI0SDJ(zpl^k
zfDVuGV&w_0nAOnFCcG;lWTvA(dioMlNH-o!7QOGR<98b4^OrA@WnUr6CEs$cZ_5^<
zWu5#fT(lw-vS=XsLxpQMom63Ecl!t|9yK4|-zg|N^bmLY$*nor?tp3=FdlHossqsR
zd!_h{w4Y~0oBJ;!EkNL+c6xi4s3(J*feh}k-dA`(K>e?!_pYl8y+}M!Oi~S8ap7tz
zDmJnVa^Qvp-vCVITUJRrp`hZe_r}mYzlCVPpm8zANiWOAX%w@(;iE)X>=6RB1`qq}
z+EXWlme)*7UYs*iV`rD#C*!kk7yr04-G%13_qi(fn;VXM_t6&ab37B=8PQ`~WT|7p
zr^AtZD%j=EuaxI!o*!8=xbc2eqw&{QRaXOh_OeZ=4yJ0j?ti86*aMCs5^P`5oL$F4
zF;*f>xg9(hZ}uAv>f9#dJb^2j>9p1olzSszzQnL<iHWfwVg#uZ^=Xyolfhr8e(wW`
zkG?&QQ=(5RJ%SANx71N@Z#^3IbpjP!Wy3*tcM-Jvxy8iAIqu)D|K4oN^YXB8<t4%!
zH}15xUCztpZVw%p#*qHDZDfui!UJ_$MpZTO!4(VugpAk+Fsz)z0LJuk8-Cm9KdBI~
z2rBMS2G#iu<-B(y#;U7gyA~V+UqQ|S+fyV+`m(Besc!7i9PDUuue%k7RQa6j)WU@e
z?e=iv^P^r5-nf3fIq4#R6qM9Hq7nx`rNwXXBqphR=(k=LOx<}0*~J%W2$U>P#8{O1
zYF<5aW`Q4XDKL@T++4%F&n&F0vB5~dFV8%}_jLA`<$(c+))a#3sqV!K>><jc0_99g
z#|40UEX<D`vC7UTAa3GLZmj|>H6{z~0=``U6B3p>d~k(RGN}IHc94U0SM9(d-qAHP
zlaZ1V?uaffDiWUify6<9DRcd(?*sJ6#FS>BK=ch?NKuKa044=|{Q2vbE{&3f@4>3d
zD*S(*6}z0RB%CDSvNEj)2QsE)>P{!(%+U1rPoK_^uX=;TvVK*5SXdak(ZX{D!moq{
zaj+lHDk!LHXo%Dl#o?NNj0n(L&9>)}HPg5#f%|B9a`*s=EtjE#Tjw3&01>ENaPQc0
z-p6EB!wB56vG3}xU%RGwsU`S|sQdgLN`xP=yQZO*9$Hw%5eq3x()zwf2}%qUY*DN%
zEN{kp-o9NK+?oifN3(<@zz%knpqe{>IZ=LuLxZ`>xdD;~Zj-}2sq?>*aMr+=3;3X&
zy#SiAr6USb7{p=$lsDC;+uRg@l}+o(?tTDV;SLbTN93~@X(mQSifU>StNI~H6uD$y
zdpS%2^T29Z$b6_Ao5tIF9&O$F!gAZT01znThw7kVJ-2W<zKZ-FbWa#{{6VCk#Z(r9
z^qAHIYs^W$073do!|lw>C!a>aM>6Cvl!7=+OiiNYg}<mZ0zgBo#AY+IB}dfSD)8H#
zj^R{H-s<mn9y>Yu@9zm`=EvKyfpTR2A|1lgax7l_ER2sD3))qbi6uqQ&$u+|tBE|@
z>+w504U$hWYl#&tTnbD&#xoE@^$%s#Mq`hW6SM*@%U`0vD6=p#?G;~oPWSk$p$7QY
z$~aruLZnc?`{eH3Ym}Jw=0be(Ml}G@c>eP%?%)5^USPA&RQE9pi(3%eVY0Tk+*@oH
zLGq71Yz!zY^I&C2$Yx7RZl!B7E-o(8($c?&e4+9qABS+&tlHDj0X+lNng^AB?U*cv
zXX(oxzmby@S|o8cE{+EgMdL|L6)ZDC5deeGS|dBHq<DPThI~FQ0B^G+?KcIPukC~^
zC)fNrq9JyOjNpnDZ&d#wryxJHeQ%WAwHiC$wY*$8p}k?Ba9B3uyfJ?L`IBGBjm0&f
zQ8$Srb7kH}b#=U|fv;ap)0aY!j_=UIh`fG<8Qfopc^N*)jL>m??ocR+Fi{DTQc^{C
z?i8r61K<u+b~D{d1nmnucc$D_M=H!9r}`s~&2QGtCMF0YIKLC1Ix`MdCM7M$Z^oO#
zZ$^hBe)C__YjFE>tMBT6GE~sp*eh3zKV;nQ#o@EdX+XS07;6Pgjc_Hh<eC=}EU&Q(
zSFXIpO~kzzUxV45KQFANMsTe~-~J4U)_``L$oYQ{DXd?wkN8SlY;1plAi}8(vbV6Y
zaYin4B&uKY^i1c64}s_OfKi@6sfETeEPKM*N$udC+e?-#L9Y+_f|(C<_qYB4SBcWM
zvH$aDX-Ua9Z8ONQ>vB9`KOeK{*zg|2_eLVix%~+hsSXWh66c?PA_)6e$8r44l9Gg;
z4;v83ckGx73^c_mReE95Wze3W@UAazLd3dxQx>%=1>cIS!ll6UNr*ZIn-DK~=@A1f
zr6--8m$jrWKUwlvtKTOS-02qbC5Iu3>_7Y(`7xY#i)VKpQ%}XS;3nBV8^pkz&NRkt
zn`gaWzQCU_yie@4ujZNx7WrI6<`tXK>ws^qGg*L*pA3?)<@f^>#q4@3EwYuTb6^#L
z2(cbUkM!p~sD#N5(8q{_q(i<-I62`vDr?J;aOn~#!=-{sp`&_xf%;;4K=e1?g20yz
z4bxzmupdDvY_qYcyQl&$?e*&g&Kh6k7ZC7)ErBfm{N0052H&w$`}a4XW2TNJ$P6Xf
z(1!0HKL#^$o&Y>6(%(SsL~qrl+r87%)15$SGxV`+G3(lJE0HQmE1R}wb=H9c2ViJ6
zJ7JAG12v9>_5nvnr2otL<oquEcbQ&V1Wp&rOM4A$w{C>F-H>gsAog)(2)6FZV8b)0
zaN<7g`nsF#F&*e#rpzQNJM2uxOTW3<hZim^)wsHZgCq8x%8FxLU%!2;nW0^Le$2;*
zu8nq;^Af#Y-|>!NLgIYwao-I*)NSy$zqjl|Wrk3wiq$fmn2w8!i-}Ny>xr$6P0s5^
zlt$Py2kmz5Y$-j3WPqI3FboYr4jBG*1i^K;M@z3)g}uGK^GYC2eSr^3N?K9*&G_!!
zOKZ5CdjI6Ow8j$yGqZ0<K@iCp9u<|9&CAb1dW~J)&d$Lhyj*n)tvMSh>+fr@;$y>m
zW|L34k(sH}@f}@_&bX|3BO@cBJJWwZdR1Xxh$h<@tjucZ?C?V@Xy%5@q`H$U^X5%X
zakE#faBIrVf}YUs<05<)>gA{^krI!dm5AuIAhm|xZUU0bBC-iu!Y%1nP$?pp@$p5I
zu0tv8ri$<0#nUoMKaKj@yuMU4d?9t+Vq(###3d@4eO+GH#@^oD-Q9Q%i%bS{O-oC>
zEI&ZEFYhtJFrP^7vU@<YlN5)x$NSItI37zBji*|wCr_U2tXUHz1pdvy_V9I99Ck`|
zAH+KSa%Ap>>;C->esnB=xU#30S6nZR^jj_K>+KDCy0BHf=KcGtvMZnpQTv>aFA~Pu
z+_+(~XDOzEq0banLUQN74xs|!Og8-1^<+Cc^_Aff(mOO1!GMoW4+41I)`Y66S=53G
zS+0jp?#wmgIsiCKc$wpY1L6M?f~_np-avC#d}!~~L5iAEA*S^93jc*~X5U7KhFE1i
z{P1%O%Gu%c>8j#lrS&=#RiJ??N^Gb7wry>I2x*tJKd@vnS8UKZi(9vPaRP2ZWo6~!
zLpL4n3n0U+TjeFgGF%yiq+&1brcM@}E3pS7VKo8-YsU$ghgMqMxOR=7pP#gux&MZa
z)9btD04V41fw}8WyrGLQ12S5I9VD&61B;I+cCg?27j;5BNv;VmQouWFla#-Xh0(5h
zs59nvD|G5Qv}U@9NQjmWLO*aA{8^sdu5$btMzb;S3fd&)<cQtAFRQ&Iwi7D9*stm+
zT^u!a)KAZ-_p!c3S<A=IU;jrOGH`8jE^X^(?HCM-nJLDNFg}CYT3Tz84O6nRM6P$3
zrDp>$#08Sv4#UJs{UdKnIB;#+B%x~tT|4n<?~v~}Ma(V=NueQ=Fe%oBTZ7_Nn&G{4
z<~!t5cWlN`oVoE4Tk=iB0as@aD{RlW*n>>p;b1sqqg6VDk<$WeiZ3lCkgl^Z)9wSh
z(OFsmfGWZi@0&IKH7S6)YA4VxnUlj(ZxbFJ?c6<a@xp~mUoEhHuB#h9d|1n|;LQE|
z8+RDmmwBK#J&$X<Uk^}h0LPp5c6k{YT{=~K5V}6^^H)_53NXiR?awOpeY1ep0$shC
ztt$f3nv3Vi*8%%w^SfUR+8Zw@{OmonOe?J=eUi(FyXs9xmPgxm?}o{Q$)$MMdiE~m
zp@LE~D)(V#_HXfC?0mcueH^=j8#h|oG&D6=CmZIYHi-}TfnSP+6m)TN((Q18F%Ao&
zfE>hI=02BxkB<X9<64YVb)T~y^a(s`BS$?n2CzpBSA<?rgT_ZsAQ9{;z&c7}6#jfo
z-{w#+aI*(RZwQ!HaY#a}N5e`?M5}vK7HdE;)ak+4+?M<I=V5a@B%#9};(ltj8%Sqe
zd*X0Aiovx-F)V62o#;NVHW=x;EAffrDMB%q68re4=gTnIvAE>s&a$Osr!PQa<uxWs
z?_E$Q2_x-F`DG-ZG$i99Vxz(lZFYD=_){r@%MYzLc*r1aJ|Ur7IMVOaHNE=T9rtl{
zvk24a`o+4Q>Z`YH+g96z#DPmGd<+~QK$*%C;E|zc&yuHW|2sV{)3wtV6|fRhjE2o%
zx<9X!y84P_!@DSiY#TF{QNuB5X+jQBX{o8m3LdYWp8Dkq_NZQ~8So-k6b;)RcO%c*
zVMOQX=TO?JFrHdeU>nOwOE<gdJ2?EqK0??IBBiXU+AJixn`x91k=1$2ZMU*AHa9n*
zH%Gr7+-Ya`CjSuBns}x6`UeJ(lQor2q8T4Ia<4fo#s8p3=RxAPzP`=JlVhSY+`C!!
zfic0MhVFdx)1N!hVkZ`LJ~q~J^JZp8NVm$!T`%|O>{(VEHVr^snFl@RX8&=A|AsCz
z+k6LD2E5Tj8fc1lINR3RIyN?jG(D(#(!VM`p4U&8_)@K2%2v<b=T(7R-+4O%LhR%L
zQT^z>nmVva-%@AM_R+Ma`DSOtGE7Z$ge}&kIUoz<_t$dc8u#X3%E_rmZ~lQI*7@0A
z+|$m?G+m+<@7M#$FLvZl36OyGr!4^`DP(Wwt987b04~hfryw<T_3`7!$1JeVwF77M
z>S!16>+mu01FkS^M~-c{kP51im6IFVABe_cb{0k$X@&VZF*iB7AlXoR%|+KkhfLGe
zkqqEFW#GqOCms!Vb>f*N=F|I_X^ifSZaaWD+)<K{Znu?kY(9Ea;nO}nfP1kD;p<>z
z3k|L__Z6*x4P-HkP~f~dT$P->!_N<`cQO2cTQf$EpeExEze#hQUGLGoLgpWHT_i+Y
zUVj)by4a|gPUH%3vq2pXtP?ducrH(LSrrg3nbAz3V1njc5DtWeW*Q>`HQm{#Tg(S<
zOeP$GJKF&4bbU6^(?gGEa!e99$4njD_b48mMPvXQ>aKr)cY|(vr6;ILZ&zSHB8jBp
zAUSMmX%W4zd%0~}1r}`hEz3vnP}r88$&1B){kQ$Sto(g_)yXs6Kuoo5OyQ&~Qb=Iv
z*oCVoI-1j2L>GCuUb@zN)S&&t;WdAM1eZ7+a<iDnD73DqxNW{y0MtfbHt|NS85jDy
z8r55|&tnw~KO@gYveD1^5>PQYV%y)sO|+*Mx08#jD^BZB2i)5#KSRr~GPv#%WS7nP
z=GNBA>(;@0!%)NPzBzX1|AOOn;FS23=r-RwS)Yz5uGcvLI>!0=$D`_~RO!pwK-cVQ
z>NRI!VTloM=<HM%7M@={b044e6>`0Jx>$~&1kwrk*A3Xd4!u}>(<X;4Th3~oK_4rF
zPBlk24#b_xfNYGk8rtw9^egY@?27mX`T4nLu{A%?a^wVScA;$sOUf<^ZA6z`wse3e
zjy$OBxPM*XnZz$<=jBbz{yUXmI;jf@p?Kpq7%zAoJ4Wh>zffCW^7w{0cY#6WbR$z8
z+5m&mrvqNG5ibNayu<={;JbeYf7S&dNL}VAVh+ikwm~EkddLJHbmi(*`KUc&8#ktO
zFSJ*>e*HQ&Ene*OHxY{5fp~u*+BjHA#>U3|p3rbzaFj|$9%VTcJ1CDctEuUaQ;rG}
zvuP6j*MoPV(J)2wc3gVqg^8aZbv^S2Ews^R91nc4W8X>}ZlU#uR_rJNt%nQWVk>5G
zlk5M`w`(`}vFGh=mc49kO}Krh)s8c{PNa58JzBJIyh?G!{;XSuC`CAluTy}F%=;2&
zMs{INA*AnhQ#qTZq@}OV%}M+9y#d}0j&=Lk2oSR;lZN=JZF8S-K0#K*u?YgUX}=g}
z3X|3v<uvRKt0zFH(7k=>1#vM(3ePj$m5{&(J<_l%&?x%wB%~Xy+y&`qnCJrfd-i=g
zb_WtGVzmZbTl%zH&sK^0*ROv7X%u{l!p*`4PnGxJLHCOn(H_Gve-1UJa@w(kqjeWq
z%Xi(5h~NlVLuMx6!P1~6!{z$r?MXnA&bUCuVZAQ=qs)5FAGnx+e<4zRNDF_DbX@d2
z7l?z2!l?hB-*7rIw`=hGdso)e-R*t+xZqkiMDvgDgsWA9Zy)VRdL;nmF<R@-In=O{
z$#J6-6Su=X0odcd6qw!{zG&CaGl+47?w=%ZefWl1h>`^jTH4z&wIS`+t#9ZI1cl&&
zOnSmp5>)ApUv5JHF7p>jOxeQCg*z7)j!WGZOBEHBv0uO>gS-KlJt=nj3ezEXFYN);
zpm@Hs0`j{IZMOvk{6qSM=c832`l4Mr_76X1a#iQ8Bf7+~cb3^h@X>70-n^tt!t9FP
zY2z(h#DM^OSq9gHIBgC1IbWYWi%r%)y`{FcHgR3%g2aqUCSYj4Rd_9}O^DWdatHc-
zBIicFe%)1HxcBCt@tv(P3K)X2@~Iy>Y*h;~=M-33&w^dgc!BVzZ*06sV?!QBL&NXJ
zFcSjXm~CWQY@MD@ZeCv8k57lYidhjYQXV|e^co++*&zlcqqsY7AZu~{1FAex_Ag8-
zFZ3_>K6;dE&>Hw8=(~P4q&8yZ-u{T;PEVgAv%mg(kd=)sUab-X)4;!;i<5#kJj3(*
z@If~>MS+O7j{*a0=9*htvh|q~XBMHGQ!!l#cnHybNt-mszXn!NH<p)~$t@@t;kbaJ
zji4n4s&2a{+XSygt-SbL7b@)i!1ek^bj8KR>-6%>Qq^O%GQnnc@~vXL{R#!u>U;@t
zab)e3qNjQ)Msr84-VglX?xAl1?yb8DcW)Zo_`S18_r+@^pPDSpElC?yMtbEs6Fu3&
zhC@mm?Bq{m-?%}1JQGYKECa>?BeB1)2NIEri~x5gd1oISggYV@EnS*tUX43PdWs*D
zH4;q^(@u09_4JH=mDY+j!Fuh)l_jVDPAoagMll5VmoHzd;T2kx$ny+2JQrV(P5p}(
zt#96B$P6LziHW%Bbm;u@HEYldn8SAQIqEGz2VJ=Oql=iTt{(h_q`3a{D|89ed^~D5
zq59#0-?J44!~p-r>%5$t8J|)A2sIsF0E$khKHwLfo}CdC|7A`btvnkxcz^kc-q|tN
zFWYWOuix76@@1N$(?X4w90+6M_UQcZdv5nPN<Xa~XToG^gAH()eKRj@mP%?<lu|!H
zEF!rI4o3Tw`@7Cgnml!_*sX<L8HVVYNAPojj6c@d(9p33Q-;fG-(|uOP=`)Yj`o$_
zQeC0z-J1K6JAeIq1?}9lM|nsNCXY=mUbJY1)6v>RhNyy2u=XzlS&=p7WVKcUcFxET
zo$?*POkT4e(btEyQgV0Y(|7N79{nhFo#I@n5<bm@WCoiD$>*3Ju&DB}Js)98@bB`K
zD=W`EgF;`qjUOI!d67E|4GFd7Jm7L#J_%!0EXJa&llR_JrGiW|<kb+*cvOWiRNe-P
zKt>k!2hPCC&uc-<RgcRR6&enk@TNmeJ3r#toS!xNMsZNX7Kp!JD&1UM2#{V)tP^Ti
zRy0?1H>5iq3i<hN<4g_=b!rP9z^N6X=tJqj<*EvM?(7W)bI@A{)ba#))>9nN5ThLk
z86@I!MrGxOPlMQ5tGT&@m8J$U7sIIP<u;@1s9!*i^Z(ghXMov!7yf-ag2Zg`csx#J
z{CGJoEf`{$3lR5~EIzcJ-wFv~zHY5^e2d;oRcOF#5agvlCnP2$U<6?HS+<zX#CGW5
zcO)4x92*s`E<q<Q503o48^A-{NvyhnYxR0T!JBXS#l>R?wL7NF{4}ptRcZMZb)s!#
z5~fTu>Q09|+})=i4J*U$?jC)Qg!Fz)N{WJ2XYz+xe|!74(038XQEhKSL&Jg@(XOs8
z{?|gCO(TZJ#_n%2n0C<NAP^dYZoR{hHT8e|n0e!dOv%Q~j10!OBpPp?b~ZG;dza#W
zXY2Rb1=R7g*Yo>p#qCDIT>Jw9>T5NmrD4b1S^-W~X?iQx;n~4da1)^DUB1ftBMH5C
z8lFoZj300?yMUvMzG$81|HO!;FJnVYYp;WWK|b_kxR{r5aKO-Dv9Xh?h}r{n|H(U-
z(;FHav(pzLOXx4t4g{VVxwZ<PJR7}EVQ4^J9({)e^3te2{Ol#Zn(t8tpqE)tW^LRt
zLy+VvKw|#>gsa&(|LpK-npMz-6L8Ec7J!IpY4RN-9i8CJOyPjjsCQyG7(FLvXDSNr
zt6+T}K|nj%xN8AbcM(uJF5kGrpCMzln+d?BXmol9Y|Uy%{1Qnd_{3vf^he$>lY^IT
zlbVf?95Htdg+YtjKEOkPe}AFDg@2<gaDgH>F%$W`TgYz5i!HLizgTSDdI1J&d$mK`
zu^}=tGSI5kaPclG^=VPyMo44o1Ts8gQ+&|tT4n6S8^h7RUIj%$XV}6FkFK|Gx6j^l
zMTG?$>L3RxChEZIrRTb}!j#a$<MYSfAbN$)=?~Jkwg~d`i<hhJ+rOW_D2@-QlhyX^
zd9N%1$X5GtS$xE^{!aS^4li8$3n&$}MrP+j>K=}cae!96SSgFYr0)zmS}&g!0AA~%
z{zKvHxJFJxV~3g9`PuT6lvO+n3v>xxeb5`2fYQG7nHhgCAqUTi9See1jv=%ax0RNa
zNvO#PW4dU>FAp$QBcoIPD0Yj_qRw{p^i+&`RNPBv2cO>;nCgGH@F^%Ti<q1iXimj0
zbIgLpw)JnoyyA?E-ZJfU%v3455DK{cuhI-Wpz506Bc$SP<dXx)Zn=9maa9doz!rH@
z=hd4xF_oRTinfi{fkn<<mUa8~)k~McN|=Tt6BGFz*1p3;om}NkckI^R5!-ESq9Y;@
z9FJpkZ1C-9-0lEekeNW^d>3ARZ+ny!6gttqqWaUoo1&GGNrUU7A^4b@`XzCH{t5by
zxNJF_jTd~n9mAw%lHO%rTCzs7yze~osJ0inZ>Re`h=s{TXf8AR29I=?>&%{5dM-XT
zR(M3vtD;+2H!Q~Gs&V4<MM(@^QpIG(03hZzYPzj=&p?HUNh?8>?S?)xP^%OA(%{$6
zpN!Ac@jhV9&TU$reJo1*Zgh9=@)~>s<Vch^=&gZAq9MflGXEaoWT0fP&##PblfVhR
zl$sjN$;5b%hX2r_A6GRWunUjU*@0FQR8bqC{S1^Nd}7!55$E{$5|EdFew~z%@FvQ{
zy62Q9on`@vf%KvU<G&C(TaO^=gE8u+he2`Be?a``d?G3oFJyQmMjSmRUGop4b|-E9
z;guB)Tt-{A)O2-K*jKLYmsJKE(f&vmxhL9{bUO54pJvL~TL#7>riBYNBtRqgE<#zl
zqo+L}Q-%+txJuDRwfU4I?l)Y8Q@@9J<5(yS&mZqMxpNu6=Ug5*wJW){lVu;|YmTSy
zIF{{FDk6y%t*9Q4{pHtvkV1;uZD>Cm9z;k*M}|8{Y57H&w{Dru9Yhr7;o)&sNv^Ke
zgCkV&jP;Jp$EcOE0jWT9Nm~80=t3$7-t5IxA3zpuo%iA$o**silcidVk?F^^r4E~l
z@cJ~?*EhAbi4RVQ;4OHgHKw=?Ue3qFwO0!YVq(zGNbx8^L<9^c`3bsKZxQuXVAZO|
z&c!r!b-OiH>KC6YD=5$&`7#)fk7y^RrXD(S#Qe0ewe>HY7fHiIW4{K?p$eLQ$gFsz
z2@HwSNj`LG0#}WWj0Ak)uQi7=3dr+!XcFMD#sgY#6QL#j%&Iz`=#xZ&d$qE%B2}!f
zg~>1kvEUg%8<x%u<F<;syXxm=A;#gX=}5>vvPSkIz99Z`FCKx<NsQbK!fPxC<6G~j
zFMr|Owga7aGSzpAJ+H2gUZr;1@hp=LdLb|-ZEZ0twH+rmBgy^R@JM^HytH&=Q?ISk
z@r24Mi;fTojJ_c6UONQc*`Y(SBX=V&T$n&kdd&}6NKDvEKVtov-bX^|l$6&8mg}kD
z+Bsfrf<33$KV@vu#2yiyfE{C=5n*AqSoJqqP)B}6MNU*<63j7}3yQSxH;%=XzY9>c
zw6?VkeNYEboyr$vPjo)pr|>p~dHpPQ?K<S)u{He<^!6X?xf$Qoy}xjE4O`7g&${CM
zqZ)tNDG=X|YZ=UfV#o7^R162g36d82yA6=1PPu(SDk*68@Z7oQeRCbqZ^BxLG#LWZ
ztX;NBP!Nk|A=se1-9~BdhTL&2myPoBTP-bPXHQi>a?s^pF<q)?8N~;pe(Ys4USFNZ
zHOwMf<z#doBfm?)xNO7FLO=&^F|*z>X$$JW>yOagZMirw8sTs1hRd`FD@_;WH;&QJ
z^^Uu)=FTG8C@uBZKS0a^Dyw?y&Yd@;h&=)AZNlCU$>19}svM0EeyAcq-KZ%SWf$~%
z@FL9d6xUUim)o7@S)t*T!}6yLKq8vffxF1r?ttFv{B#54#AI;r>=kc_cQ*t|Dx>kU
zc%ASp8&=6-w1KHDsi6OPYHI42Ptd;9dp|+vt3Y;&*b-|)igMoU=$M6{?DsAv6F&jn
zGNl?5FJ8Zv*pqyxxcKQRZ>qExN_g>+S@35u5xdSe5bLJ&0K9|iQu?x#symn&=No?|
z-UV5G`hYzcWgmz%HxIx@?@{%-)tJlvN%bz6wYixo>)Vy?(PmtzOVqbqhHydN?oLo;
zH2`$dvVnjusVO@F%-n06E-a)+wNv8aywLuD0Z|HQy_lXvTT|RndHJsI;Bf0tlDPuI
z+F3Lv%BBxuB_vP)F?lw0g0i{->Kr1h;r=7<?w?pU;l)~D1~d)WZnjJh3f7p2)w}uX
z!iZm19ZSaNoV>i|`@B|=&CL!z??0-JyAC%8_uEosMa3=*n0jBj07&oeFP6nREp2V*
z0l(j10fMX!0D93vn-`)PgDE)Z+y53`-f&#Hg01N0qnMbKly|osHWo|&`}JB|60017
z)Bw-eaqsX>%lM^N)Q+ItD=9wS`dpwhW>f)RsZ&k(*|(Z;-pkWdcw;&^03EtlT~kwR
z_a4*MkOJ!GAWn^*t(v%auKNT2+5s<=qM)!(W%dBKgzfH+h7K7R_(>zCNc<VH-o70Z
zk@Fs$rImAp>@s^oWfE#(N%t3#kEnZjAGy@U2s=M<dh+Z(f@y6DnM;$U&P2TdZTR)S
z)Y4n;T1JKf+`xbzo7BhY5b^Td1IXn+ax%BZAVXJP@1jc|laiOekjmyDw{FXVpbR95
zvA4b>4Bq%>Ytd7I4Qan)2gkQ@w*v>DP_-64kh++PnSB+3%(yG$+lNPcmLA)CDC7;g
zR1%gzjV8Q98WK+TzA~6pgv`j{Dsj>AMunL_z4>>Nm5mLWBQ%Zqs5%XIb|VL4(6#_X
zLNAD*pjj?!_Y+{a+uueQY<<zTGBYz<G`E=&kd%=TmGD}*VudsCK7on>=s|bWkjEjN
z9G8}oa)T|EhE#_B2hjjnO`zZU>;wgr6ztj>UfwhD%v7@!s6yiW`K%Nzq*JS#l-KKA
zRy(pL@?2{6hGqSmj_T?@yJGix4pmz8D_wto&5g{O=x_3iGR1WOAV<=cqNRhLAaN*7
zzV^mqu&Pk}G*A6`bn4Fq|5cY(&h@1-+uR3+8255BNLHm-SCn=8_wV=i1Hh=`;UVwH
zRG|O<{d?ydxDemlPv!#iT!Sb8wcnc^1z@=%qNA&?K^co%1L)=2V$vzzdgMfai@5u)
z7Y8b#v$x#*CXE(V4RON3zA}JOS1l)UU_`-!>1}m|`b+!~gDpxZ#%<hu`v4sEFAWfL
zpZS7DPY3~q=0r%p_1qZR$5MBJ%>#0d^{$M>os^iABr2eRj9a7lV;WRaN=yavVD0<k
zYJd!0f7vFgjq`yxG`5VkL{O<8Csf^lqKRYnq4_;a$UxRfOUJ#IgtgWRx%Kz%#vTJO
z5}i0x)4dv&lqj?hLy-p^#;dxz$#1Vuoje)$2zJ-!T#V4(7C#e$A)KJK4VV^C?K11u
zp&($cm|S)yGxH<TQs)C*P@O(S#AyjcNVeR*U0|yU<G3Jt-769YfkMY}SYS**-Jf`9
z!(oSYeG9mV`YoYIruX~})b;WjgB!BmLsv%^7c${|%D-_rmGc~sUVYv*FpQ{Iyi#h3
z#k>lH@liY?gK+k(6WNE?_lid!{JUsb4WH{Ox>xDW$cGO%HOA$5tx`F_H6?Uz$*Qmu
z!ypTw2S8N@k(S=3jl>Eb-T0oKMK1zMVT}{B#~BUD>OE8&Y%_VC2fR+4FrZrqOG!mM
zxZeql`V~7nex!z4{+ACOPm<fY{am2w{{5u-S+V$F3s?y7d5sMXN872430y1DMTb8i
z4)(8KqLqH+zyZUBO6cO?;~_bQqq`lbZ2a_|-6Eh<q@I|?A!z?U_TDoP>;4TNzO0Od
zQdDLc2}w$bj6%{dLKKDE4V6$tW=0fcuaXg!L}q1Xi)2)0c1Rgzl#%ruANBh`ub+4S
z7tj5wTe`0A_cP9MoX2^bL9tRYGHLUJ=px<FNn`NiObiSeY}VnD6_jR7TwiBb$ltPQ
z)3c-`S$TPMZD>C!!e9&b+1)8AZam--K75E9c(_Fz?r?YN1x7b3_`3cksK?SrQRZ)F
zXYZQe&<48r*i7yowg{9e01r%~>beoPa3X3(GdEGRx9}=BC`-J0l9J*KmbD{`sx`Y8
zi4voIKe?oCT317lWHa%_o3DUZHn{1ks(Q)t?UIt(tVFu%-?_OW`15Z{$FL0}BU`{~
z>MVD+LVr3*n^MezbOEjO{)La%x4Z<B1oS7pr$)EiBl_|=$nyJzY^5YJ^{op<(6X7G
zAIJMTSR>B7chz1{7QNtbX+0M{>tM1id@90KrcvGIw9{^Q8Kmm6w<+46ke8Q7a_P&r
z*B&B_1{i#Rup*-S(13{`r>UaiKf_gA(w2HWr`{4&w^ESDMp#u_Y5%$yQ9HHuJf84}
znK@%n5D5M`0OFRpJ2NFyLR#7mpeky4Y`;f{fkhi@qiNug?mZCwoxcZh2IYe&?-1~h
zNg6GQ7StX<v*QyJRU<;sjAt#&T~5t^6ERV6>JJcr6UPGj#~on3xA5@<Lm?mS=tqC=
zv)N4*cMBfu73}UM$Hj%|)8(bDwDBOtRjJ146xzVSWNK&&sYbJY?eEun|3l~4F(U8<
zL8jzgriqF`O3Hs+f~BeU5x;-`K5(9WL2#I)1qx|5+GQDd+H{WOJ%2t7Ut%=d`Alb9
zZZls~I*>SaHELyQl+^^!r4J1aVW2KehPyT*I^pfR(A)vvoJ@!;tM+}rVPR>BDXv1b
zZe!Rvs6cK@ek}*S8!+#lZ0coep7H9{*@7<~CY^LLD#g#Q*-D6uAI>af1n4Yg72ub(
z`vQSlFS_S-7CdW4>tAqp9`r$hR3CVThb{RV%e}mK9NzhTZ7tf@i~m&n1LhA-rIH0^
zos}iP#g^=-BFxVeQ8S7;qO-rI;Lm;O;xxmOK#j6^E+)3}#uCzD(Dl!xIB2UNA#ZAu
z2JkY-NlC6j?_ibvtsz%ab`GYdk5VUk;2k$}II90xZKN{)j0$1XQ&(5eDT)>jR(4e5
z;5LueeRz2R>Nx$^e}X*W>Z-`Zi>lPn38+_BI1Iag2=?*uiHL}(ImnOeyQ%0+R@P1?
zM&&TDoC8u+^*Gl?DGz|G0NDMT)S0tqZw)k1mA?khKFam_FA$c1iwSY=5|G=o2O<s3
z=QdCl+0N7@czdgzII+<@er9Gya<>Y&Y0BcZNH`CJPA9&Gh$1ibt^4E*Gb2{y3HlT6
z=;1ecz*Dqxnc6BVZ!tT9XfScLeig-{4&MRbM)hvXE@^2{R1V(#-Y>Z!`L}Qasn=eh
zVuQe<Lfhaa6#fBznBVM%<O65hSYA#J3XL{!_0im(=&;8B8*&tBppM)$DR_OopwY>b
zXD;K!8)|4Uv?W40u4+*BxVF9hE7X<+xI3yt&jCX<GxmVmRK?N^oPns~OT4P)kbooC
zJH$FHPq}f1(R+XEy(65Rh%1F(A5C|Nnt=VvlgLc^;#W}jq|e)O;jV9#+qz#%E8sWR
zABx^e%N!Le;eAkNR~F}g0#h99J&5!dhtw7jc7J#rF!0PbM(djB!EwcfU-$U{gckac
zw^?6YNlESp{=F_g|D#`G^;P<?JZnsLxr72Nd^^xg(%jXc&%l9P5Fr+hhOSb;X2IjQ
z!Q~>X`X@H-*tzplak2ZKP>iVrGiJk5MExLCstzDsDVP&{Ezd_dfhXy8N!<<I+#mTd
zdTV1S%nu_^=vEOCW?qFERQX#E-yuMU`j$I21DiSA-=!MouKB6sUaouWSoPD_5HU8S
z=83a~K9Jx^G72EQ&k1we!1@W_k@7?uLZpbgkzk2XC}StV8-0-zdk{%SfCByU5O{sU
zAJGgljzTM|5e4rr$9**Ho^rjofwp!eC`}+e;j~7ePKTjW$eAX3NUr$2!jBez=^yO^
zm*N0#yLKNXNeoFcyf@Tp_Q@nCd>b07gA@s9K~;Bl{r7IIln^@5EU2|_!%hU>JxsXR
zay)uX0IN1kGY&ZJ<5nI()AGLN2w)UhBLnle1n`7pwAT6Qm$>qcOBHpV6z$vZ-@k_$
z4?gl1W#R#a5G0+yzkhlQBnxsS2*|9U2J-kZX;X#>NDE@rq1b0#5x^Ay!=p$T9&Ugm
z$P{dhS;t$RzYpDdEcpg4Z@>XN9n35eb(B(Dr<5$R`SLqta@8uk^uI%}9n_q-aKxZ1
zP$d4;z9H!-B0f_U6{^fNp%D=+*%yvUy?~!XLsvX-j-;C_OpKCv`{0b&R%#U2*51A3
z0BPwNXGL==(;T-Y#oSxzK|vo-B{Q5ol50%#$u?Z1`;>I39v3O&`xXj==(cUZ54X3O
zqd-9YX_QKI8X@uxU2C9$C?g5U*J1sjGOe8<9{=i<eFYV=%4cs}(9G`tqfKu=gl7-G
z(tGmc$)(XoD)+Sj0b&sN=EiP4sH5|ttBaTG<L#?WNT>v^he5V~zc_^%0Ot+?X-P@8
z40j7$1)`wM@f2u2D(||Qns-pmY)i|>tc3g`sn~j@t6vX8Fwkr2lC-om{7aGB(!ETK
zszjkbGoGZKzYHlmocf?bPWlmJ>n|rtbR&cF3kU#cWn^nB#oK#Z3X-~+g&0T5e(D#U
z`$n~wV`vf0&d69T{T*XVO!4pZxbCHtUjw`?RSXaTJW$Nq_=uWBwWSnvy)CHSJT{&N
z$R8rli?hO+L22mw11BDs`XV48VBUEF;U}D7_dcgwxMOUM|J}Q4o2V#Ya%8$djauW#
z5jAlN|I5OBAezCc;i;+HY7CyzQnm&~MSaIiHp^$NcYGqDcpfP(OjX&}vyA&uK4Yh<
znLZ+>QuxnP2~`afIiq;GxMc8YRbS-+qOr~S*aAlh49WY)nA*ZyFwS+A+#gZ%zea=W
z(cLRiOh=H373Z%-Y5~^Qkyj?b<`0NtG#gGC=^C`9xgLyC1TtM-MQ7hQ!OC81Liu^J
z5ebtaMf5XY=6L*y^A5A@$t%yOH2|Ug(z}+@7k?S#mww0BHbZwXA5YXWtrE8-o`#v<
z2Q=oy<Ydh{ivmot17G9&G?H`y0fEe&Sy1CGr>3|ip^Pgo{vI=^8xv{R*+c$@SNzW{
zFH}QG<bYy?*@8zIr6}zFeM>VlmEH0{8GPxISH{*~zjn=HB-y{|m`aR}3u`j{z7_^M
zyWc3A^M?aTt$!c`M#i!8ZaHwVAWoFr8ll)iwk@WM_EjlZ6`(d$-9f$<eB~SJ`jA%M
zkU?>`03e9atQ$NQoqCqM>pxwg(FI1nh5p?n`v=kDX-}VW^6=a>RJJ0_;L*=vmmDGL
z%(hZ54ZtZ>hJz{)0IbSzcgC5G|HErxCwt?Uh<W^f78c^C-$XliduEh&VJ~xyB|NB2
zGcb!3X5`UzCW;gydjEpoAk=!xs8iyS(Qgh`x`<`wD?~+ze#lkUTV?dpa@__8)Xbop
za@$au1#JVBHNB)H^Ke5FqsMl3a~`?C{d#T*7+z|jhW_(?;vR@$Q2h}clIBVte1%~K
zmzY)*Ya(CLyB?Y{p?7ZGF-|IF4xNV^GNpG)OIL(jQ}|g}e6EhP?m(sQZ#;txH8v*Z
zUW(;fEV+%1+2OlyW!Ltzo#Cs;!=ue^6Gze}*b)O=XwF@j_a$mhHLLP_SJL8&wjqF4
z95oYqj}j9tP=f<ulk@ZC-BMDYGygk;1Qlx#f=4x*6Qr2lMF<;w`2>g_s8X`6r1%Dh
zh7$RAh$Z^nus4NH)}LP&3jUnDD9e46&@qllV!pHg7nK*JT}QvTp+OQ9%naqbdw?Yx
z4Iw@ySKd&WSKEadfF2Ov1-4`v4-41l0I+xM%7$a#m*0?=m#4?a6>^ZO)eJzC|MYM=
z7eA#c8#Wp=#lUS3P!nx>ts+G!MvhNoS)2s;?pkBQvuAS1o{>{;O8OiT6VnB;tchz-
zaiYy|kIupSw@JX%V?r*b7;OW-y|-=bepD2|HjD%rMkhvyQ0&l6)4B`#=xX|&R<yY0
zbArL~<EU4vAXL$%4m82en--j9Ter^JkfPiSz7a6r;L&y{qT$TyiHM7Lpr8*RP4&p0
z^PjVcn$NRvTBD|T6hSfJccK88*h+j=p~JYx#?T<4*o~RkB58yQoBU3rTrz%ZQ+hjM
zl`1K%-+PmI8O(#;1OFl#6xr8P6BChl_amke5TZm0<TUoefnhh&%Ww)dTnvUTrT_;;
zZW_6~mR3E;8;m@k!9zrO_!|ilR%vSD*<w3B-ru7_IRE0s=0t%9n0NbQgsB1?Nk?$>
z2#48-0p9UJP|!^a_b`Gcd_4Ml4=$yT;UfAS8ex6Bj!r!OLjnrt#d$fb$~v?%fsA3H
z#=^t|vTw04fum=m^>lQ`kzV`K;Wo_9!}Yedw;wCxh_7eCyxXt!v3#7IJ}HT{K*Kj|
zPlp=KiVTXF%i?@ndAXPP*fObKdk&@$RI&0_(be*gsOLaOiVz=cJU<RXk?-R?<exy3
z9&!hS;3QrLu8~N~KY#;*?p`&6Q|LT_2^+bJ@y8RlHofwq36w!Vhe|Wj)3q;6bwF)=
z#z)Ww16@+<_MfKNahaoa%hs*crKKL?Lx}fLPFIuC-U41wh2iRK94b^^G*%cA!uYuP
zL9qGEpV^~eh+?CE)9iU189AUt!b{W|TapC`u_Lo?kF0E5DK+IJFlotGUWbkz1+bce
zQFuF%uj|lo02+k^3RY%p*4aX60oAR#*5IFjRD9_S1wecT_uoHaC-`MS%#E9p%Q_HD
zWEw)LF*G!ke}m~wu)%gTw-Q3W`7=4Bs|Y&%y|vHd{S?r%$9tbf;{gg#J9%e&O`yIi
zQyFdz!jpuE2s&y~^iZ@z`5Zyf?fci_($ao@{~i?>=$>LxLv*DklTuT8z&J5<r<2;b
za|2gCsG&6M3CbIt7yf|A87dwPkCszIpNY2eizo}V+B}&8HJps<Low_;XD6aCaG0mJ
zP$kkfNq#XgrUc4(oGjHT<9lC7Z5oBY1vCjpqamtnutPt(i}=!hzY|f6RRXEKvMi1`
ziVcAmKqo_y#XnpWNm#XeC{eHu3R-QX>chfwM{H}=3$g&be54V}_wjB(d_1VuWgOjD
zi!cF}H;09~F%{>?e6KYnPz&qhzi=9SBn;JQCH!<KU}>Wna?ctdN%j~)01+aOU^~>~
zz*p|CPCcI1MH(xZtCCQRM)y&O+Wn#D^YgRp)$V$ZkSlF!l`?26-L)Z54Y|92g&$PT
z{S1d_1K!F5AM4w&=g6T$7crF7p?5>vQ3cF6>!5jPxf|B;4h)AY+O8ktSXtByiij{|
zP^w;l)ol6A3<FCnKi<s97zgkp`Ny>F;OnfcGvep2A({}Rk@B&;Pr?F@uLz>umKJ%c
z-Phju^Xu7*7ZOIsfD!LSc&WT1HzdPu0&?{kd<G9O&gWZ8T5Lm1L2~jC@^(OXUZE7r
zyFD&hAR*y)OsuHarE2X3D-ihE6;zgz&d~#xKuKtQv*lSdv6ZzB?lUprxhq!{?&U>W
z$@}u_*RPF%tE$}_Z{QVf-lQ@*4;G30)}8!fFcs@vDgdkYj^F$RN_+O}*XI<zJc|RS
zgs4p1m{EHr((Mgm{<wt*9s{+OZE3;AWm{@--nn;Z=$&D4xU>ERKxDv&4``1dz)3O_
zb{qm59MU4#3oqDF(zHO4x!8=1_=?7g#0~&q7!^=#czq@NIpsf&PEI>-_=7j8Cax<Y
zEPNIn!-!h<&CQ}+(G6@MUFxJw0F3+>_K%K?pvo4+OsU#d88d<R!c~T?x)0m~NgTN1
zR;W8hMn&o9&%wY}X}ad+fRt&mL0eaM4@ls_R9mN^(Omf^8ws?E60JoARuMQ*xuMI-
zi~jD7UnXaXVK^$#bVgO`BLwzs0C6H2+c!54AT<xZ=xYE)FCR@dfp)1l$N-Q*;_mN2
zcUsk}47>o04IYQd1^SREX!vtk2|_U(!u=8vekuK{0^U9V5zJ`)+O=x|LkBTy5V=B0
z`^3jL_)$OBSYe?7Fc5{rD)URGro=@G(tLc9lLJE32Y=T-oaZ3_!!xk^P<+Pu_L^G4
zYPvxb>(>%hat-Rq>(d6!cB)V)nt;&GlnK+*(}OZ=0l*aN2o_cXako8j@$pc@WKP*9
zJP1xVXbB0I#7|^1jLK=mBku}&utUE<BPt}s(?a<x!Y860o^i(%`BgC0_i@Cwh(9I6
zKPg%FP=bTDp<Dt(xC^bid$@_~1Xd!;&Ei2lQ>G}IyYhB;IEOti6(tUPhtWlMY(6n&
zfcOHu97-vbsf?ATB`qy7^733pD&8PcOETR?=kVF^n|g3fW|!FH&~6&48)yPH1OUl%
z*@ZBbI6EqEL@@mu=7b(9IIgL#5@GRS$3S}w{E6RB!d)0C%w63iLSZB$uK;EBmo)>-
zMJ`9t#dt!{SlSj^0dN4#7ei^s2>LMStFXPf5hGU-7Jm6{%a)CP3-z1d(Usa9PJcn$
zz+g8locb)jxLiJje;7n&w6|7^w4?7=eCt*fS#2Lkv(*~@Q9pee@b#RAdb3S#2O^@|
zH)?$Md`DJX8grhaKY>8K(zXNw2&7-D0oXyc5FNY%V*Xo(6ifEl`Xbbc3JVKKL1FdH
zd-v*~VE`_A&^kSWKl=-rBNirfM2k2UfC0>yO!k2^-lP{4aQ9MHjyWhG8CPfgi9-Px
zr}X9?8)REp`b_2BXdRW1V5v^O&;z*^Z_o)3HA+1Nst0N~1Cw&Jl_mUY^~P+dkifq1
zlm6ii)1o$S8RP8Ha)911sb+QosDB|_Cau_FY;Othv9l&g7#4e2W5tjvBCRvVPm+>+
zW%AvLetu1Bb`q>gYZ8Mc$`esN%^JWxkD{qTHU);U6CR9_M+GGzt%rN0g@sparPBL&
zR;tSu+Q$`QkVOr{$k%OsFX6TH%Bi)lLPrJChR)zC#do^G|H{kf|MS0e`cPk}cMyJ@
zm`H{a3evnB0TPHSvc>OFcsL>O1~I1-U>ZGqeMfo_{$S!WPjz)Q9_Q?@Ets#Deu{gz
z&~*WT^@n|1q{Ku-{5MLAN=dERO2u#=L=2<T2xauNwcXLn<66_y6qez>drF2j$zFl?
zCn7ZHxNt~+S3Z5Z`_rdMnByuY!Ej6^O*^fm{^8NBz*cisOZwRQX`JVdT1jbXm&vcI
zN~t)29X|v}mM{tlHI7$m?)UBCEKmwpZ#$V`PIpQ*h!0@redwg%zlL$|yLW350|c(2
zTqYrp)ew>cgplsCnuiVn(ywV9VB>m4trvl`5B&)!P5R!jNBg1=P(My|eu*W83qKq;
zj{FF#LE=#EBptX_06PAeMA^u{Mwu5ySEA6WvxqrJxWc!r<Ald%=ZU-J`0%lzK^C5+
zG-y9o4Ymmh-I_hk-<ng~fi><ZPAkWKL(2A*svVUjuZ*Dxs2?@NCcdN57uA-gP$JZB
z0d^S%i{bJGKx{THLG(FN%}F4*k>b(D+S@0dHda@E$E{&tN`Bbm)PwvCzFA|1Wff9>
z?0~5365WGb_j$HFjzYMuSBW<*BOW}E2T4KQ>|$4eT+^hg?1cEvotU-Sq?aXN8S)ar
zx9+xe9%~A;CX^^<*3W4v)<(3&VHsbob333gP}#^E3~~Uf>|Hh9_pPQjT>NkhaR73H
zJJxYGP|VUWF?pCl?}M2k!BleB5Q$b^EJgvLQH}5VXj~p@0>ERUZCtKfTUS+AtH>$?
z)Ptk|vb}YCn9&l=z|36UaSn(uy%mGIwqBbu9TmlxgUb596nmiS9qDGCNxbrpaRd<1
zSCm$1%fEqu9w-SVQ6t#(eX`oykYYYP991fE1}#L9CUg_x!op7gMf%_P0IY7PNPv5W
z-B_Xp48siD)xXnFuP&{mFV;pi1OW!AhVTSKg|GYV+buAyDmrKf^WmcAU#q`c8?GTj
zjbgmv`3D$sNC&BnnyW8NqcDI<DYJ^oXaFBwyW5c?wYVIcxo7qR6k5~K`xxOmyetEK
zzcE;VK(-U?(N{J%H`jM`yM{2@2Jk8#pH5_GN$7dJ?Sr^xGW_7a10v{$#UD`qtw>$7
zCj__hO#B%RGA{)6EEL;?l=g>Fb!;JPQ;{o<rppg?byna+=e$Yhdg0%M3|CN4P}N%b
z02UTglR6Px1g&d~hv7komlV@e$X){zE~O%E$P&N;GFnztbap-jn#$+S1S0BM=a;4^
z+v3nN&?l2**`s9~fRgFK!)%|BgQ3&L`<?*^Fc{?%ODY!zjFlDc6X>%)%cd=u%qeYg
z5;MecP7qu*rlWqrEb~cO6Lnb?YyG=0dW09*s~o0!-~g(wQ99x9$B1@TvrzU!y>7Nn
zoeuPQ!kyt}J78%!g!YE@Y8oJPy?y%>)_PVsjQA#~=ipp<2ylW(0F!&{2uEdw$QE0c
z0aX8)A8nvrd)smx?HHIX1}h%V%7Xd<w7M1%z|hhy#|5B=dX!Zn(FaT}z&Eo^F}4U8
zQ00^AP~0cNALpaQgUAS00$qC-QnoWv)Oub0yr|a#z}B&awh!t4>8Dxa1h<8j?G0R(
z4d~j4QE>FI*-^DE6pZhkzGOWXRdsZ9P#Q%HPm*K^nnPk9wcJeV3-%}smejsM5Oodx
zrd~8*2>8zcE1-)|#H~QXZxwx<EtWqkJ>9gC-^8R7tx%A;$*R`Z)&|kQ46Me9n@Rw&
zS}=p-z*wp~6DawJ-QMP`?`L~QLy6m<x|2a<&6?}!5#mOVa4Cn{Zo(j<%4SVZu~0KG
z2=C#DT?<H}fNiCREO#B6+G!)MqOqe4;dLOZdn9N-1o~;6LX#!WG0xuCM_xolwZUoy
zvg)N}XJ;dO2GV)_+xn;Q*r-XSoZh~&&s^CNG6~CH+&4Mfs8I{yCF|D>OgPkf*T5lG
zHQ@I{y%xK^{_%+e2k5t#Q-4Q%21ygsQ1)bfh|bv@Jjggfm6VKpWTh2Ob7KGZ?@)C0
zb#CZq+MOGUOJ^D`)(9R@G~twf?oNff1U@PL=Gz%#W)W7~FE|k(zsL(_g42q)1kEqz
zw6y0}P^nR$2QLBa1u)JL>_Tz+PTnuL=}lsR7r;CPU$t~)2EX2o8>sR<HyuBZ78rPf
zAXfKsWZ7tb-ITL)A2{O}=jZ3#f6Uf4vtg^KUI>PI9_gof8HYAwetvgLUoRH7jk_=9
zTyEcO@FUvWW=EVtmFLLFB^@UtgWD*408juwgl%gaPAQ`!IC=Sq9RQQYbFPGcK6&!w
zkU3GkS(mPU{(Mpcr#YtBkk-%uZqe=AyD?$0XX+tvT+qi2J-FgdDBSp!1a#Wt;a%KN
z)SI|xs{+tl0wzE1`lD*P3&jJr0o-z%v+W*z8y<FYayn;WVQS2L$$O&MfScthpdgRq
zCrfnF_oy^JnOuv}UZ^h7t|V)<@5Puy0T5h(*M?)>#Y}J#AcM}Jy;GAVJvfQe9E>o`
zOhvZA7BruOFaR)+S67q&y?g44l%}NY>}!A&JmjNvQ^#>X(5Z0Cw>1@rm#l0KVwm%`
zDxcvfkq8hGqHk3i$Btb^_c;ogIK8d(11h{XP9wyEz@1t4CH&SAEv<6};+hL$(z>Xx
z{=W(d>qD|g&*8t-R5oEjT1H&oP_->?984!P1H7bYQnn0vx5B_f?2C6cy$k0Xy13%M
zRuAE}@0ezay9I0oy$uWod!?mcqWuNw(dNwtd6?dYFyTYLM<^iwDMGEe7mUIPxZ0H-
z`RCSi(!BHJK0#o$IFN&Uvv~9|J&~|5Sno$hBIl)Zq`t4B9Auh->M;svy2qJxLR!$S
zdH3ZTl`h{Ko5DZBnFH<T-)x|H<j6xsm%IiOSZTf`lp<AWwLihne*&emWDlXHf=b5X
z<LPTi=4X1ODe8uXg^8VVM@Qr3B;FqIIJ}NFKiR_ywKBbDl~_Wm0hObY66L@x%v}TD
z_Q}c1hs3^o^{Vo!+w$EUDI2gwt$PkL!jXdib?O(N3{Y$Ikp4#3)rIMmsP&?3?eA+S
zuw~0zbYFluJ9h0l3J_X4RF=msT=-7H#oxcZf0*ULbc)~JK8)41g2sLJ(uFHmmN14!
zmG-0f?%np#vGrF|IFEo2lVnb<S{TBetgNUoGiH8C!A4u>8dr{O!Gw2Bp56G4i9TRf
z8tWREsqfHU#E8-w$k4uje;CLK>ZllfosAyy3L5tb6hI)@#rQBDDF2D8eX>bBh|N$8
zYX0<zU7-RSh?uD?e;37WvzS;7hSYuAy9X%_1YgYlQ`=CO$ilVz@cAVG|8)Sm)KvM!
z5KFuPEnX)-2wjPNsHQcwdxL`tYxiV@jdZpigH;&ioo9g~WM~pbA*;DR?4C7g|9HUL
zx`%D~X;<R@Ib)g`LeG8jd5vzdn7AmNx&Il3?HT~%crI@dQBf2>6cM$ru8syMgv<@K
zUxpY1gy+z2D;Q+f9z=&6qML1&nThZdD68oPGa_i^$g5zkc>jI`n)nL5tXprsudb?s
z%+Y^Ht-ZoWQQX6b62xehvnD{TpUYk%oh|I_J0xyx^q$~nMb!2Fo00zSTo6Qn2XA#`
zvuyM?rK;tR)2I?j)@NO{Y81M;NeAim9u(P+o)8~jHP)V-4FeYM3AwHclx_@uKU|e0
zhDt2R;%J)V*z?i?tn2fvREh6krM)F0wRY^<WsG4qsA=g29|fv}=NN8(o!0OPeZ!~*
z8ywz3WfTz``*V1>9TkoEfdYsR!eacHY*|BeH2AWho7U9zqt}2D_VNL#OWzFx{e_ku
zKVRj@hJ^f|S^BEYQv*F_orIkzi-2TA0MEm_XxL7lJ^KWe*#^!Q(!MXj&S4gK7DO(D
zYn*vs-H{k#l4f)Fiog?S<=|3u&S_ZmM{x;k5)?Rsf`1_(Z9_E+t>0cTnb5wz)cuC1
zp2htLf&e6CV%N6M+=+0KvFeC?@Ib7s1d3o4vZhE+5z|XbtUxeUO^)%Lp8jXroAwwa
zQTd$5kcNXC3omBJU*Ccd{7J`geD>5oemoZSa0KYuXq*lS3qy7<ig+?!KLr*BWYgk}
zvL8K%-eMBk8L$u#<m?;dIgGhx5P?yl(kOzp0IckHL;4Z;JAgpSpRcyVtlWD;dMi60
zyfGhoUR0zc5man|9f0fp)|!#fFrj6~1CEt&iQZe35*SQ`w(i}_(`R1nvL>)N_KJsg
zFPw#cpJ{^fvSDz0WWPeQ@d^+G_aupn72=-(uYy$brIBVn7akY9dEXQB%?%9=H8m?Q
zUTjRo!XoBHBj$;3a)4ma$qWKtEZW;HIU>XLT`l0H$etK+ZhmT%%OjaU?l>4^21U}O
zxs?^~+O>MRx*jRYybxd|(ek6NulF$pZ{I$TNT20SnGzEa=HY#ji&#u6GB$=0RgZn@
z>MIgV@=(S5LbuD)ee<R$Xq{mk&&I+lT6QY`f}Lb1yhHIyP6?Ey`x0rE%VG$X{E}VZ
zvVd$MFfQc{dN~4sCnhE)EII<bF`c_RJy7+4pd0MMy5BGI(Nonli^)PIx^mK^s6!+1
z$7gUht;B_Sdz*cWqzr9;1F|O7ZAIW=?XvlzYIEuj&kpnQCDp9V%wPu_!46TXQ2N{o
zv<qMyc`(l);xm8s?{HzhoXe+!Hp$8ifk#r`6--+D%vv`%iL*8E?HgQ%NcN#;=owh`
z;*7bu^$RLXJu|ajD3_@g4j)CbL{!eKj9Ri0bpqoF$zR3vkZS_c4;FOAEC7<h7DM@;
zLx0%r+BJkOC?T+0=of;s2l$LMZHms`S=fqCyOp9`hqxm*CkJhqw?C^5Bj<dic<64Q
zYVB}5eE##eILd2H<ySvKlzG-bApy$vYhGXR=Qx<TuokhlM>OYc(60dW8E_YYR@fv<
z$FHOmYZ)b$YXV)|LzafPlRjC@+81Y>T``MI=%TDP`d6Jmi&WzKMs?fL1XNH(G<c$b
zu2T(ns0gtLzbiE9FIUT=C9bL1z&7OW&_*;0FwH41#Nz~l!wG>h%!_^(+hoq8M3AX@
zjJHv>)B7|1(~X4GT4XMXn>nE1ZoazLrKxL_9RvB`;{10D?%8u9bHC1h$;x<wB1)1N
z?6#;Q#Cy*qG{I6V&6+{T1E6_QLV|#0-5b=xoBR*Gy2C@!^LsZPl&HOQ5dbSraM{F0
zH>7Mj+fp&@6(%()3FQC~8=&RnnM3I>0gwcE^opsec6I72gbNa<3TCjd-^6eK!~sCC
z3f4$!e+d;1668LSUyP@4C<D-1o4jCFuyAM?_y{xw3V;0vQah~FQaK=wvR-T4g2W}<
zf?Qn6z*HZc({dj)cd{AmpeFZVd<&Wf@if)v579HP=O3qf7SoRsHmRo#b$&&O%+bMt
zge6bcDQouk_hSyCu4jqnEhoMGI{@MUk~6IgoB=N#2YnD4)fwv1A|UZ1)I7hAs$Uxs
zGR7BxLD=x${;ItZct1C+1S-;A!g>hkm6eqd`lxM-G$QjU0XTw%uVzm9QdL)X3iG%3
z9yhTrZ3JuJnf-0cxW(h^aPJ~=q|5&94njqN2P{a}ZnF;j_)&M>7rx%xYuQ8jlHn<j
zgp)KS<^6}u(A<=Yiy9_gn*qO#4yb5t%J*MCI^~nEqOz@59j67&1<Wb}iUWqL#$l1<
zF!bY)0)d!Yex#)<U@Ju5=}yuVAU9UHa>4q1-+glNFx?qvd}Q<QXP!EGv<~<TUacJ<
zV)xgtLqj(P>v-OOr35O2G=fy6wzk&0IgQLmP@=lI8*oas{kFHu@4-$q1iA`Oeo6Q}
z(iZZ8Nk`zc7xQv3%?)dul$a%Ru8w=0RqbUVI0|c`YY4sW0@{d><dr$n&MdX&PfI?0
zS7_6-YG-9szn<pju#f6Gj|4Wy@1tLPQ1qkUrH@gML=IjH))d(jd+71`h*ytv^+w+9
z$YeaZ{|xuw*ka7^!{HMxfA+3FbN2jKN0gWHq_UbXJGWf_VdwU9&S+Y6&1Skz=wp)s
z8!pYBxZcop>_H?oY7-h2HH+RO@bL?<zCxOzRdQBnRJbuwz2%s$t{@ki(8rJt+nQ~2
z(Q5)^=f}!0acMUc(sMpnx*^b7`qNT`V}{@-B$}_oR|GQR2%H5rP5k0sTQy^ipKpn_
zF5N_zh(yuRVJ_y`n*ROKtqVd7ON-M*lg*6?C9iL9aOP&|<k#IS6XW<bDmr!oYQpaU
z!CbWx+`E_bMJY{pT%B&#Ky~0s(ZW=>>s;+RuSe(MWwaD$EfAPY=Y|GGM?<z`5*0F{
z9zZuEI#mk0`k&U`-;Mu8nAm@d_xX{f3yH8rzzSs4<TJXFPk<AYEUE&FmKJ{IM)^KR
zyaAU3#^U5a#Obf+JjYAFL$JqXb|3;%7P}EIxJZ0Z4+Sg^7m-#tb${?weDQy0&%PP2
z<a!!&!}6y1W(XPExy6?6()V!oW!KrRk_0wJ%!5B){I=V>IcR!)ShZPbYOAj4*R!fJ
z_9Lx@FVj&P1Fi1os7|wQRgx{}lNra$_BsE-<Oe{K*FLCmj$ZE0O8Wg4hNR2B{W9L(
ze#9mF3|=`FF9B{3>G($W{$uAx_^PM)p(m?i4!HieY5PmI7tOEves%NSgLSbB41z*J
zAl4`0YN7@|i;=~5zSMXePd)Lw*N0DWte6HjKMhYS@pQ9*%Y<}`@no0NPtgDNe=fec
z8$<)_#-Q8k-p?1)I@3lg4u|pceNm-)v)l=%vsViqRdqetz;n}GPFEy+uD<2D-tnU>
z_sW}D;g4-UE_3YA%|7Kg5Qe5Yn`(YxG}gYEpQs5MKZ#*3wrwHYF26nitmQ?aa!ahU
z?pn^Zo0s}A*bGx$<I%SPzgxe+d<qwb7r#ha)w$x7=f~gjfwyoYar?X8hsW~$ew_$f
zZGw~cvtY80^R)B051*@u9^aTRv-Hwr?2<EnJ*d6huEnGU2NgQQ1B<7RW8??kc<--<
z{b&y=yv|*$!twpEt?R%siB5;UyH7M{Y6cIA=n=_+N*Z#|Vov(`dpyc}2<(VSDo-U*
zmBlJ|AbeqCoy$zGPumuFfSn)T{#cxf8arD6EB`W#yBysO{%GZajOWNmR()4m%OH<^
ztJz?TlU1Ug^OxIO6z4{=NLZ2V)`A=ri!1};1CnGikQb_Ob4}-Bu6;ZRC<1`X^yP<C
z+}eN#?}mPiFt$BdacAA6<Y{c-OZ0`AxXunhXL>RrK=DtV{7=jf<o~$o0MiuBFPf)M
zo&f8c|D>hl@*{U}^Yp!oX@wBS3V+Mz=4{1#x!vVRB6w1-JMG)C_+M;}&#Y9t5$|zh
zOR3W0IO{Q?dGATVwYwodV5he|Xz-yah&+0k7S%~U<t?cvw^V&w5wyk$8L3_0U6~(0
z4PO9~SemVOgGmT3JIiwp7rJJ9QNhI)i3r9o$AIcN{VaqZEMnJE*ug4&27_e5fBS?y
zfXgb!tlW3?_JXle)NetRa_s4<E5J<y?pXIg=I33RE(a_#xrs;-V1oNcuU^Ah`1$$e
z@YVXggc)2-;kF2(I+ge&X4eJXsjIqk&becX`k3$>dFGzn@5W=8TsbfHB4@MINI~B{
zd8|c@#?pMle%8Hzz9k^}j+M3f0&j8p#`nh>Ez|FJ)+>ycv*YMPU8cWqx=S9tMBvGt
zx>_p@nvOg+PN35XF+AE&CT5T6Khqs}<R`Ha0*hG1HM!q&evC!bR8_U64aQG>f7Nca
zQ@^w%c1yhVjj<B?l|w=dK{qw^u3W8?8}yaLQ7Z3R;c)!4i(OZxt&^A<T2|T+TK;@9
zSJy;EjjD|^b{{Q#9jHb?`DQRH{P5AEiZI2fKsn((?7ubi<_*6<mqtV2ru}3bIgAN7
zW2k&QkTfer2iMZaL)(u4W=2g6Z|c?`bO2{Y8pteuS6=rn@|=9wXCQH;YX_PLFfb45
zW0NuJg4mQJGM&F*!*Gf^)aaQ4dQ7$jDNeSY38)>Yyy&fsb6Br1`5rZpH)EwM@h`9b
z7oUE6-il*pyQ@Y2*<>$IIjw#1O)z2BgAKSl3Nm(6ZRZZyN%J|6-DF7MQq>%>g3pDK
zHlkt~wQNqodRHNW&67}O0Ve~xGQV(lkaSLFe>y4#>$Un}mFD5YDL{&++PR@W(8u#o
zStkSeJr>Lgki+Y)8^hk5r-;h2mzUSR*4l8`Y(yq4Cf7|8A1q)P88Hbt_S%tvh?`PR
zPAmUf0asA;Yz<4D0;9J9P(zhMS~DmG=sbMC^=b6x(&a;*!ku7nWwH<U#c4S8_2z|K
zG3x&KR;0capkt%9N#IeO%TUZKVmoR9Y5xLn^KenCK|4MT;|KH=CgRfojro4aB@m~>
zsx&|L>}gCMP9s=YwpbNDyGEVa?nRM^lY-=I<3tU8dRh;gzW)0etP#@Y6D@Ri!eM29
z7Tc}GzfZ?Q|H4BX5o3(dTIWdNpP}cQ{jJ!xt`-8K1v@g!SAhYZj}SDW5@ox|of-~5
z7c=}oI4;75L4@oZz6(cbxbfLl>lgzFHthKtc?RSx``=B+S<(8`b}kmIB_c<)*PK!w
z4*(aH6*bXc8nN>@Y)1ioqY!w#%ki8AEsDe~;JJXu{RWl*2TFD@`szkx&v7cAqyk|7
z`2uXh*+hZMP03AaoF>=4KaL=tVegG!>k*KEj~?qhQLXWi_xp(=j7CHZi%?B1deiP`
zo5~IJ*8-=?*|j1?dy8vBw6y_#Od&WZ`t|Y9bm#SMIIXKCbeyUym=PFo;SeDlnCQ6F
zc!(|ecv{SBdpwituMY=&9_*Rz4-r_Gw9<zx62wkz|9pSSS;h7PAAOC{-SYzIX+S>%
z{1ud77^>oS>1qHnN<p0n*;dh)J7N-rzdrg&<Rt-%z&qY;Oefx-eCO^<t?(|y8o|8c
zJdPq3MMUu;Zzx$KRt}K{@D#<R`AJCkFQT0nCOd*eLFxg}w}RM%5&fStyl~@aG?vsa
zy&1VBUU`0gvOefz3Mhzj^PBk``-9+DKwBAJ+|B#)t~yS{<kWtg_Eep;);Tyt6ZvbN
z8n@HoQHYJD9I#oMA5=1dWV5@2J(h;s>ztb3Ix!Nx&9_vP@1C_DgJE-m*I`|HMH~(2
zjT;apxJ-8<*~`a4_9}8})H$GbwAy7ZzlfI1<Z+yl!w3$}m;2Au$tPx<ZaI%LO^h})
zg?_@KA~RNBoPlxzhejU}-NLbf`62p|Rj0>mWx5@F^<}DV9=eSa7i(S0NGI==-yeL)
zm^@;RG#I6xvzE!nPdN%H+3nb&rO164YYHgi%#2a|#TU6>F^9Bmp$c)Dcgt!phr;Ld
z-8DRQOiEzj3W}75@XU`Cb)C8#2XV1;eP_{!`RIAtn>^jenoc9x$bB)8c}PJ<5j+;~
z6;bO^lRk;TiV!@%1m%QzJI}S3G(nLqKvY0)KK}~%3@0^2sujHlSy??W(-($C?26rv
zab-pBT?hbAIOxP7TRH7I`yb0&l*8t6fS8x|kW-U8>jwk7PH}jwjJ3W@OCD_^)&f}B
zs`~f4m*(8s6G7sG@{xJlrc~|tQiPkXHA}fOU;JSCqZWQ+$gILdmB3s(X?nmMo(;xn
zD5(EJh=%ZuT%Q@UaC~cpcf%@O?XBbR>^wH*tLe|rBSd_i3*~Q$&3UlLOMZP&Ird2P
zHf)=l+pzf;FK1Z>`yU-dAZ_f*0Io|G1f^YKB7=i<5x|Xk^ziZ-`hwWqKGtnlvF`ZD
zPgVtpNRKV5;h;-TMr$EWR)S&h_`X0ELX>Lrbst2jv^%=fT%Mn;i*`hey@GT2;q@*?
zrEO9dJ|>zvKEg`Uy*wUl6)3B9yJ|ZTlVa`Y8BH%^X~*!Qi&PW}+Z6*00dGEeW$(hT
zPsqiPBzLbnZj6F}AIt?N>-FaUl-F?&4hccqe$ZTA<jf{M766p5$RLUNIR{+)!9_B~
ze5dmdT!7d~zp$NBQW%28r#Jx9`Z<knChA-a<MwO3H@?Ng+<Fd2^2NF>I${uQ=JkX;
z=o<;NoDcAUTHvPE&Tn|IlUZkvSZ#D0e0Uu&swb!A$ao6u(Dt0TAim;tA65m|ht5pY
zY!}l@vW`=t!dnWq;E1@VBZIH-<8;IDXA+}NiS9>6oDZ8$Dy1-J2e@tEONAzkeZocc
z;o^*zE+%^|#X8Hzq19u}?~nZe0)8G}3EIb&hu3s}OiTUwGn|&-I(>C%rWmLC77NdP
zpdVdIN-|n|13mE-yPb}p+%1I#GOMFkLI|re2Eoki`bq&#8AHirorFsK(yO;+Kq)<5
z-iEbzjP<q(oKGw#(aEfxB13&3tX0EHoE1~I%gC5OkwkRvi1)_uSZl-^tm`FvlApi8
zKh+0Hbq@y;z(HZyBbc@YtiX{lHeo)NXoOUg{yXa!mRP#1%T_s_AB5v&kePU0w6G2I
zf-Cfcv&i(2B0dTiC8aTo>NPDpl<>~M3*k+u?;%<}AN$q~o=;`|WcK-t`pUA15oAAg
zy**=^%>hb_-#&g@QGHVLFnmcX#|VGbteqSlIZy^z2Z>QovqnHEWAmjP>dBC*Ne7t?
zZWv(5tp0J6|B}7jpJmYuBKTRQMf>tnn-n@VSVu9VM)io{=xD-7Q1=4+QQ1w-BxRj$
zGl3af$$84bZ03WfF91_WihWXh{}egee{ZkC^aUxsIPr4m`_tIEO{>BsOg0*Ofb=`q
z;Q$lw;fZ8xBAWHh*ZUVeJm22s8YIuXU$bE1of!SK`s&VsUbWqLaBKG=^L|ff5oIlf
zaa9M7Bp7xWf|{!rb%v=+7I(wiIrMsSZgy({d`iGF7~fGvs^2}HtoKM5Sal{~uRU_B
zYd^ad5Qnrclx07Swta^*-lSxH4(pc5$8-@D>ZwW<(v&oBgCNjJo+u*qUR-^_?8<>R
zIPkp!)3b&u4s^YUO0U{o^BlRNc{o~j$NarWPXY9lBe;FRfJb$6H1T&1WTpFtlse$E
zmiU~&H%H<`yiQ*1$g!Df`I*yqk2*RT=c5U&*(Ps1)}(B9fhTP{9Yw<LPa;n818@i{
zB@PIB+d7=s6iibRgjk?X(JgZ2_5w-SH-NEU+_MT?(7gSY*GYUxCDrq5jMEPrHGKEX
z#o@op1SCf%fEmcLVC>Hqg}t}99w~<1@9ZHmi?0N(1AD@y!X~t!3Wh2uF2sMx`Cpin
z<SqUg7Km4b5%qi?TX;|hdFGV8CINm~Jx4~<H|#v?mG)IEtst|W?<?M8<-MPc&qAIj
za0yOt1%&JCVX{`^@L~U7G`P3rFe`4rLSg|;z^}QYxR{5)4)rD+5z?4;lb+WB=_M4y
z*3P<IUw-_R!{$^$tP}C3op7aei=yv|2}!O+w7_V$2?J)HSA|)^jgN(q1bHG1N!yJp
zPw`c&ON8Dae4Dm!l^cv^=QhAaA&A_32=*}njYK6|ktW+?&-$f=aGlnla@qzmb{I@b
zQ?lL1gIh?35AR6MFb*Z5X8~7jx5|fkGebh*I`r)6lfzLH@S`LWLYp4yc8fT<7&b8k
z2r?}a7CDZgsp5$yHlA?BQE_C|`b}Pa_KN-VIxqacpZ%tYLK|(AOEQ-tgMw_3f2hBK
zpNiPa%lF&T69K@F@pR&%S!-%-i{X3t6dxmrCtGd&M796;CFBdJG-?60AN+*zXcX=K
zA@gHGRX!ee@Lp<1egolY$G2f1j!rF<A74K*sY3g(h*Qlo4HnfG$I-AX;3lr0+z1yj
z<uH%y$ZOn<bkgR^c9k5&5~WdNu$@0Aw?s*r?VAae|5@;<i2yJ7CS{yx*+{h3oW?4Q
z>0QQAU)jI&Il{Vt6DqCk;=8;_m;9kH(RFVugMK+6`ki(d!{sVF@CIR$lforLhP@Hj
zP-c+Yw(WAP(m}ACOYbU8zTb(rV}nQ*WVgN;qm1$6(^3%znpDwG9iKW^SM~!@7WIW(
zcrBf65W;3~=q2$gJlLNqfq)cXQY4Cf-J*|v;Mt#UOHS4{#1_6@0pxJ~QBM*-$tYDK
z5m}>H<7Z^&*#<uO4E7Mv*)(jg?2nv2QG=Ao!g@DkZX^lBW{@ZWq-@>NpX4qNeQ|h&
z9KVc3UflA=6cV5a))7$OQpPS?XDuMar2FWAe_B>6yk*Nuhxe_hJWs7SpfZ?+=%$dT
z0}E{}42GX&>?p8=Tc;DaGawvi7BFc=h><wQhYlZB5ag`eGo13e+f^P~l$!WX_dHn^
z$PfHC%@~67eL^w)G&$Ny;>j@2H?v6*Fe&_EQ?gJ`Mlb$;&rni=hnN%cVO5FgE|ydI
z_Lq@Zl|}_Vc=~v=o7xP?su-LUu6x4pW52XCG-NH59>TLKEzTr_;_E1OHpY%{Thp@o
z=5#j~Ue3vf;-;S9-$v$`B@ZB*-+4<Q7XcwXpR@JMO&LU_#CMr4enL?WG)N+WSdUUY
zZW_X8*3->+_fa4^KA-piI*pAZ2CA(+x%&sq?wAoUmx=a`uRDA2zD^xBZ+U``gEJM>
zz46}}ki|HTjf=n1*3nK1&gvj*OA;p_AXDN1+?4uD=TApuDy#*8)LU1TNFm1ZoDBF3
z+54Fz^g76uq90smH+CMc=*@bO^W{#*eABsD7yFqMa+0rHbhh^l;(dmh>iJJk!!@Ja
zpaft7XqN-(6YwhKNTGm>{Rug`=R!b$Xa{bPE+Q04W5swJ%4|*#8nQf!T|wZr(PI)4
z7AB~=@<x&eCTzqE_kc69k@&~M06W1}Sm+PYQgq=j{M~#KM0r`+r6d4lE?1oDrZMbA
zW*drwkSpKeY%Zucryv&Wg(c|CLLm5^b2m$V36gQh`q}E_l2ih@=Abp1B8xhwAueAM
zyAMhefA1uBI{^c^cYD0TgSENHryz-5t5JjOK~(3FYR&*a58iCYD@+Yo>^T6CB*6t(
zLA*g}?24bA`O}J8Xy<j+TgES9&$+-D>*P1AZ7dD9N@uuj{EN7FQT9{tti~}Y^p_?@
z<R17D12{2w4HhO&;+{U<E(vgggJS<Xv9xRt(-^_}7~rc(ZJmIn;+0Rh!pJ>3@uMN$
z)BUYG4vGxsHsqv$)=)P)gPiWxryfL(HMNJ>tgu#NMz~Atv$7*At^gi%{?U!VBNHiz
z_epY7k%db9d+!i{WNTs=<q%T?>-&$cMB!izaTq%6*I->m1H;js#MLwbHU_Q2ZUO-z
z!eo7MsXsuD+qvz(2(?kAkOxJNaIbr_)cx0Q8X0e1g}_H{=r=u?6188RV6l3qL+75Z
zPo%|*Ci(dV0)r4167qhka?0g*6LK>{6-lk&a|9S^?Dz~NkIC*!D0}O@FaZ9B_==i$
zXKBDDGM>*zfSfIBVFiSmbC)1&aTjKM<xhpl;k2SVjs<p}_=(@=B9eAEo<E1W06`aX
zp5bn3FygzlIp0Qt_JDp4pF!oiz0)04EvfvGTdD(a%hXc&Qw%dUFV)`TmXk_uLn-mL
zy*N?tvbI4wl>7SbZ34kI2iv+*S-k?ZyTe%%c9>29XAEfGqqtWpTzhX^Yn=On!2jQi
zo;lcf3TOpqeV1H0lZRvQJdlz7je2g7EmLd<igpNzNZDm1C%E^ik9ZCNeOdFS94UBP
zpBX4B_wMJ`9uX$5?c9vUXj<}BX#er~)Pn$}`sy*@)V@RCkhpC7x05?vQKdlE=5<cV
zD)0M~kt;@IUHOP(xo=qG_5c`*ujM3=c8;O-SDTH_4V;}M^19xv@&lpArrK<wfMVvu
z!x63f!dR|GYrJh0$yRjorv`GL;R$(sM?e32v6Z*+(B4lKS&+)WeC?b&X~6bAr2j3Z
zFqF=Q1O^7yuMmhbLULpM-~g{<KXN&*yugz{DG~OQ$p{_X(NR$we&1;?tbGgm=niK?
zxUi|V89=UOPmQm4+Pydm1A{4;jOR*$+eQyQ*?gPlrN<Z?)S-X-cnd+0qUBIu*zr@0
zZ!VHU>#6`KQm1X5_|bt{taE~W>MV?%xuQQ7batAGcpmUwh)5Rj3{VDGy3YS~(PW09
zqUdEz+Rl6{kT4a?2b@Te3(b0NGocPue4RbKBxa9hWK|ebj(Gq;%AisAf~95hKYPzi
z$Wada1AKAHyjau8wN+a7G0%G70e_74fU@xZp}dTc(0^Cx2_lZ-c*kb|6=xK0<)@W$
zGr)fO&0h!NNw+w8;PS3USQ?iYTZ9}wTF<oO_hDCf0}2o^mmR3PFD*70j>104gL+^m
zmAbZH=&<+HxzN0fXs-LcIUdb;2miN51ZGzFsRK(#s<8i_jU3}s89R_>b+U3PxjGBJ
z^f-ftp7gnMb)P<deAMOq8OXMSuUY@~F(h6?xB<g7Bf*Cxq7+?j++3Q6qg{FWWsO9>
z_+THN1X<(;#m9&#oa9Le%D|V4b&lo=_>(e&<Y7Lz^0LfKCD#?zZI1?K8J_`@yY(Os
zcJ}dKQ*cJP@to)MjQR;NhOY-Fz9XSyeM$zVmUd$j5s98Xw^?OI;8p^q_=2-vc0j!_
zboPqd+(`+Urol8jU@YF-JK%OAC&x)XtylC5WrvZ`o-qk^T(0&()~f?y$5*^X<Y&TE
zaa9Hmr4q4~&(2X;@Y?<ILG#Qqx|kG8W~cs*7IKmfVSd@qudAWoUxu<tQdtVI5KmDD
zx^*lGXbO_!lIsvP3|#n~|07S#iG7jyengr<X&ug!wIyb-;CDX@TEGOEpA<B<E}?5X
zJ_?k&knG@bw}2bqZ@p<I`8CT^ukJ)+FsPzWo_cAi@4%TdUm$l}W<8hvY4s)KT2s$S
zYmFX!n~$3AKbO9KX8Nq#r3v8W?F*92i}o`mtL}53x;l8?h`{eS|A%sJZy#tn#xQ3<
zr_L^o6Ph%9nDq^yg9*G^agqRBO!|`_!nfqg&A>!5YfWqs7ADq1j+?J5EJ^Q~uAGc(
zlm2l!ddypW_xQuA#h+GJnvr)}veXx-Vw$}X3g!thcah|U{JRD^GSE)}&^=agU}mz*
zY}p`mBWA-~?scvn{7yyV*SlS0oYYLGeNJbdc1%aEApavrAo?cCP%j>yyAzJm`~eAL
z006I%mrn*Eq)0=IHe;(a)8jd(AY^Fc`|CoG(vQ(9Ot0Ma?H3}kY+TC-b(!_>efLnb
zI<WV)HPTkp2NZm=T;ZrRlu;EaJ*So_^OaMka{~fn-*e-~IH3d62aL@BGy7TCT_=TB
zVw<UVK=s^uGf}FVsK5nKUvxtkS3Y0tLThQf5AR0BVWjW1W_Kb>zrzaEzi|$#d01u!
zR4n(~BfruJCn&+?UMNC%R<ad#xwUTx40xUEXQ4+Z{=-jPHM!<_6UT`dq3x#>qFtd6
ziNyY@AS1y~z(Yq+aT%h6K^BlDvjBVl(J3~*4}fuz5B+cJ0~dmN(m~?*IciFFp6Olc
z>#mJ;kBR?4kNSiX?0ok^4`L-b=dqvzycb&jY=*q*UZ<xK$cqRj<fL#G?vlinS5ECh
zT4wxe2?-_b7KlGynI1I4fOu<U`v381!(bpC`>6I|)ekFFJYc{SES5;gFhF5KW9p*u
z{~UQ9?#&L%K5wM3Qa*jem2OpLRh%(xoHA+)C>A%)k(}s6e2;hS06v2J`qEe@;2Uyp
zkhx4ZgRi9!tSyKhHW8<^c+`}XP|xZ;xCpz6uiZHx<ENqMjQnS=X%XNg@2RLi^=^wH
z3%_A*P&Z1x_Wfg9_nt4l0zk8`M(mf#kqKJtui!kbfa<IL&nHs^*LdAwd+1WpOz}g^
z#H4&lmVgl2dE5{O>O{msNC4}5*Dfz+GivOeFb|!BUr><MC8fpfF*k;uIkn$)Mj!>a
zsSu!wSGERDBdZ5J)n-me{^!BQpuBUU@mwyXh%@S2ghq%`S6h9S#UW^Q%$wl;G8-Hc
zSh|W*&DZ$<UPEzc1*xGTy6`?d_|Own2p5?Df1&Y?GoWbcu=9fm^M3a4V;LJ5G?P>T
zSM$ITetN#ON~77IGM9=_Msd1m0iO{>V^uJ&J}xebB;9thRt9Ok*bq$@LH;3t2~74O
zx^iNtmVT$XNdpWeVDDx1#go8eaZqx>^aNn!H#5BA>J;n%uw5B6P8PkAywiUYVIPWK
zf`4NR<3?EE&8@Wp048gv9K%UUL4Onv=e9K0jc{&pa%mp^s7<Tf87Nv?5IAJNd9`5N
z?51c|ZhtfLNRmSiC<;#bD87e#_Da9*6X-{n_TieJ20#vpq0Nr^>D%(RRgM*116`ve
z|5e$~z9f@;d*m&oz>U*0@&YGkB(vt|p6`icPfoUNo^D3J^}mRK^8$fI94Zqms)I@8
za?psFAZ}-X^3L}CsH7Cl`Pty2wE_1S;XZos;A)A6RbSsC2xNb`IWRP|5v;+&%j2Bi
zJy2^s7QfFsDGOI51I*EX_wUg+;7ES=hu14QJ6Ihh;dy@#>N^%}^Q2NRp9*yxRM~{u
zx`(<7^9^rWN!Eil2=tiq&0EL{UHj4D2vBHF;W{iQP_g%2(qh)YbKXw*h5a*4JB&GD
z@gywWZjq`SUi)2b?%#JH`Q6NDUboA%NxvURumB%mJNrDYxZqrv)8&;YI05y36wV6J
zH2w58<fB`3(o7U_4Qw5q!v3Zg5`qBl*GeChf{K)J42hVkOIq;Go!y2faN@`Kz){h=
zN!@801LX1rtXA(S;!1#}!rlc8r6Bj<{ohJ_0d8>jen)GX+r#C9&~1X4t?CWn(Iaun
zGLh$6PB~p}OmNIEvJ;B%NNvgt+j{IQP>z4c+`AiPkBf~-tAA5UQ5@qK{0tzi?CF6M
z?d?H3Ud?7X1{XCTQ4X1$C;?|iSTm06eHI8&2MSAjUbdV9N~yH)>zE#&w8NO`zXvM)
ziQBuk-_#K~*8mOO7X$=Vn?51WdyNEAhrAOW*roTB5S%1<eoCwG05U^i@Ba5dvFm7O
zwN=qC>l$#u9*c+0Y!zp`2p{&8`MNmp!lQ~@kft_;NJz-ahd=ybOUieOF&MC!EArh5
zMCA&!bm?eQjJsZs5+X89TKa3Av)4eaxRoD)s`g-32`boohGIy}rujEK<Ftk8*mGWt
z0j2N=*0m|X?vbBrA>jmKg2q{D>b#QqL+A@Ve3-erhy?V`uVzUo)ya-6-1_fc&K!x9
z>>YO|nT3NUv1!1E021R`MX*xxO7>YH2|}T5YimT7*8N>uE$Mne*NOO-aXJExW6`=z
zh%XaV$})zDh#-<pyFIOIw!KzO<$oWu_vAke<*Shj8gkmr{p~!78E@>fo4u4ykR&mN
z9kX@Uud5U58kOe|kiuU@UwnB-WnPKj_BXPYrEb{w*$a}kSH)X}82@e8<`O0kXG#ah
znG~*Q1xMRPrc$JDF@w#}V~9b0!73H(lK-Iv#sBA54}<@QF(TcaB=`k}IS~I0z6JGs
z_0$8foBRJgDX-z*Cmpf#M7ZO3f2zLSlZF8Ndh%Oh959~w$-jU_A>D(YkMg>rb$A5`
zW}?Sma7)iFA@LqLE&$&RE%Ix60<kVrQU88D`rofdj;sKvxlwvHHZ*h>2)@|-pOYOZ
zMkCWn+~>`tvyCJ$|9kwhe~<r?i(AT~YEwRPqIcwMmmO8bXB#1yzOIykj0mYennagj
z06qh@ewv<F>&rUHad)bU(Ze~xuOMaN#jgp#2Llkne5WELWpBB!zUK{40<8r9{R^u5
z?%+66a<)}k%S48SNdriRALI#U#Ty+wbSUKC*U0?)n#s<d1AH!e2tyT8ctRyi{@mUY
z^S?*Y`u97mojtfQe8>iIKKgA(;}El;8iUT0!oRDtgB+jrlPL3aAqGZU1f1H%fR*%!
zhp_*{zz&OW6aQ{Mfz&D2JZ;wBgrPx*p9a2AE#~-uLXicC0{P@I&77S8i{c~<I0g79
zopJvIHO>2J-2X59^Cof?wehI^G;v&w|1ZW%Y52eQ&$!H1+RDrOoaEP3QAJMyZ|*NZ
z_kZV~_mS)U|DTAz)r0>pCo3+gr2>XxA+|Ib4IU_A<N#$|@}b!pO!M_1%wMV7N&i*g
zWZmiL0NmeOM}T^7w`AJV_aPK3zaH}c$_iYPrtXYXQzF4!t}2w*YS!nFzVlnvM(!x{
zqByOsURA{J-oAPH(y|R^385e*l0C;t>NP3@F_DoDh*H{R|0w(Qdmda<d^v~)z_n$`
zwngOe599^#ed$F<<px4cZ^WY?nVOoC(TkcnlhrtoTzJDZn50DEpC$rEU3+`7>J{19
zDZZ=V$??}ECa$Ognv;Xu-vxfy+QZ_Fkf0J1J9L=Be-<&9T+Roeq#DZj1$SdiW=z93
zOJ-Q3@PT1E<-y3NeE4?HBe^IifJD_@QaMm~o{j?X4k~v;G@I+dF!r2sL;QY+xa~kL
zbeS_eS&Iq*Hx-q8D+%mZyq<$m21o1|MkQgIC&qPUpv+zpT8f7i6Q-~^fzZF+wci@e
zLVq811nC*0FgsG?myVNX@lBD4v=VxrbOU{-BJ~^FW)PK3rFsFttYM@0>I2qX>2|3H
zvFuv18m3nubGNpu!4x(VpO_(PIa~_%kLL3C7j8hHB6(LkL4DPBeOmWcd+>YcUoWCU
zvtOA!-(E6i1SVSYt~JW+c&qYP1O(($GS0j9EE2y#$@jaze+HmI1uez=B-D)&6*#++
zzo-J<YjT#4hT<#JyO$ZbS`@ToF=>gSYzVv-U80gV$9{d%B1`K@K(%?cFEH%OOUvMd
zV94W{C(=Kpw0LD%0<f5_Z-7?p@++4fl^@ME9qPFw_JR$s%tu_dB(n%8*r25T-JL1;
z5uexFXTBu_;0IscBDjWnzGqPJ-*v`gygF%>D-^klyq>!Htw@Tp@plA9lIvxWSJIIk
z_N{vVJ_XsA<OOS#`EKJuAAm8oWj!-G)`6GoTV6pB+9V6S3HgJ3O)VJrtnwr2rV5f!
z%HKh@2P%NvhdHBxpWw8PKLaPLB(Mkc&gysXmQSniIrLXa0Kb$pk%Sfx-gX889Ufz6
zZiB|SyvDdvfZil^w2_E`WJq=)6@;o4G@f7zolwSi@W&uu6O0rA?<m2YT)~}Z4@*jl
z0C6~R;y&3XW#<!_76%=&JW$*`jpH7(fQh^Wo3C5}5hf0DPJFelci0?|RMQ6Z`s>#<
zvg+4?1JIXw0_kTYH@!R_czD9SJ))7Db7sFjz>b4+EaTA&{k2jgO3Um1g?Q)Elj|42
zBEo4Sj84#1SE(}~lfhlY*T^8WU=*fDEm(vfn}8ZPt*000gh!R}xBM^A=>Vd_o=9$o
zk&L(a$SU*$7L0k`0b-ZC{3i2}D3VHULl%~8R<;65OpR(3<1HqHKdJ8c2<3U%;gl?o
z3pWo*<f<xz=Yk!i+wTp);0hR1WY73{!I)Bj`a~p$11>k-gMm<JP`vZ_PzpLNet-bA
z1RO|D#+yI@!PkX}a2G+*jd2At3KckWco9dSdsr(bVcNy<05@MoN_sx{u+r-#*NPcj
zLbt;5_01TK1)*)zCa6ZUu~SU16Ql|_t>jr`hB?dcb15IiA<t*-sApr9(`5+Sd}|dt
z5Io%REy~R3G8Mx{O0JW{93#B$)K}~}lTzLc4AHWgDNlwLAYBLr9ms&Fu*m%MU^KF<
zU`6J*cKJ#55dp6frUE9q1xsGX>(F!4bW9uXi%LE2A9gkdn6I#)AZ-W*<^3mkE;&HT
z-9Q96{P|AZ_%1|FvU}@Jj1Y^PPt$w$5!fvDru8ZC47&ZkD<yK2mGLqBQ`}A?87i~r
z{639)YTE~Vh4~g6&e`dF`_$#P6`w`zmX&dKs{IO<x#}*$I#69D)~oq*%jqs}PB<t4
z0wVKZb_(EOf0(!}{4P=ck{EIn#to~!0Z;?c^U<Cm6e6r~e-{cYt6d|<u`GWt)G82e
z8Lyl!0Kb~7*P$72O)#->F|eGMWRd841Q#J4!F3!-BP@sYNA;5RE`XW1yklye$OS<;
z<OrQ79TRx!xnX3CM>I4fAK??+4c9D|n*tgGO+iAKqRi`%Owc0p!-!o+PcT#ZPtVk9
z{-3R&P(<?HfF!qnA5eZVJk5kKq~j0~m*ji_Qn>uQB6$Sn4V}fNaj;=<HIRhW%-iq~
z*et|~s=#Inkt?V~Efv$S;+O?4v%zM@30=y7ELoh$4O~=D>LLI(;CSFR^D6;P0BE|7
zS(40O6egdu7H3sdWNdp*4>k5?;tI0B5=sL43IHDbeU1|#c^w159aroIiYvK$MSzC9
zWu$HZD_Xu3cfvs&;TIVt2njlF@jgtdz!yBt)bSy-<l5W`2Xa9B*ady7bhH!^5mW`4
zZf!s|29(0>?`Paed-F%Mps*MW!mU_9m4io)5K4&T$MsP5T0SH?AJFq2m!ve+dWL+6
zAt@uQ;39(T<OqnYK_iL{VeY9dpmK~46)_~F(K1#A%r;S8UXd{;Ac3B?ikmi>0d~W_
zu1)!8yVl#z_wg;^F5-jc0ZFrmPX4{oJ<9799IPe#NdpW-6_y{w46UMhU{S_J$3pJ}
zKcX*1jjoJqLm|%1@@esIgERKn_Xf!bAS?WcoB!NHI|Iov!s}RG5vDBMY&!3G<Y!|1
zl;WE^%}TvYdjPTDMlE3f^3z&lHpwZ6oj2jKjbUhL&H)4l4$54~8?bjOa`{Pm8*$#^
zywZ>)K?;LBi5^6SJ$etou40WS6egk_ua9D!NH0hPqQdwl(gZUm5J>pC%PY_H0M|f3
zq>Fvc8f4Enyu~-eqoN$kS0AAD3<OWrX!&8%38~ZbOChQh)Oa24Wv{>>hk6-`l{ggm
zssKn?pgD)GVh&&oK3Y~5-tFm_x;p&$o4=2{17wA`OWU!pD1MR0J%W?q5w8J+;qT)H
z96Kr}B+`zvV2R_%@&qy%^!xuGs?G$i#=YzNyH2r-vMD5FZ=w(p9U<AuP=*FmhKP=t
zV;(YZNzyS?W)e{dCmGH$i-a<dM`n>ZNv4o`zw7GU_x(Qa=XpQRb9C+N`VZ^BhTr<F
z^^d{gqpL8rDwTCPbf$Vj?E>gB=(&l_=>25x;81>B42eNSRcvYYioBbD>K(i|#i24R
z6o&n2LoR+rbrF@Ki%C0nP;TPtm@UbBW_<c9Db0vFQU7)0N(P`1d=C!@-al-Kj58Q2
zJ<5Aw^Q_+wJ!~K+r{K;T`~v@7#Nc;Xk>WmrFb?V06Q8MVW4y^z!J}}wC_+<r{S57a
zBWXyxK-1APD>@5hNfDFztBIRiI^KY&Q|85~N<4RLnm(YJ+Y#8=iky@*VpS_f2_R&y
z=A*Olfzs<tD8i3XcF*!89VZvHGN&a6&WdXOpF5jW!$TOYsOE+5g<P}WQ?(`a2%u+}
z?m~@`aXKxdpBOErhTJ6Tcoq7ar5xFXiVL^$KQ7#<;C>sSm0V?@*5v>UC6(H_A)d|-
z*~(L=r{WXdufuVPD`#2Drebfhc>i(bCQgJ9DVwtBTQ9$Mcz61o4?sqX>#Tt*E@p(E
zORi*#y&t_EfKjbfTl{If-l190kvF?`EAzHoPNK}T{!H{QlHoXx9B|B#!#&AwQ#Xpb
zQ>ki=BF>)@I?#Xzj=A#BhJWPIym@n*eUJbHL};Jrk3slY7;>DWMu$C_i6^$LQn@C;
zamx)d4x`oF-c6^0EjDR9%8EM1Jpy1JaxFGGIyL01xw4GEjKuBEX|fcL*<L1dkYXNT
zAlV#x0Lekw$@0DobfL>}yb5DJGosbQslA#_6+#3mS2-S*;lMm=8}(`Acm`Ib1MeC7
zD>U(fXDbCEQWj-F!L43?rlz4#0U<^?@`x@Frr!|N%@ea{h=*(VB}XW#<^^Q5gK%Qf
z-y^n`<(i?7>0PNfQjYyVuFWPGH=`xD>}r}+gbCr!<^|aEzmKqsBmHN|3E=K)1hZ}q
z^%>?7&4uov=u$o4VVBy|di&3+RLjkuC$>JNY)Ku_98y0eC`tt3=t!1IJz?mlGnLx2
zA^yGp>(#mZ$E$O6MEh7o<<p!H;q^ZV1I@walIRW=Nu@HaOI-&Um=4!};*WEh{SSiU
zw8|;pNOBFzanF5>V_tLBeM|+EQuOndOw!c8%@U8~AAPF`O5uBIYAi3jv#i}l%{Q#u
zkzF|*y&d!!mq3|s<_gGu4!AMTzrZ}I$Fc?-#CZkLGTpWThG7Sh6>WRruvp*2<9YPg
z*tP6Or{4HY3m5c&J)-EQLZ`-!B~#NO;bGeR`K)bJilFv3+5f*|H}j*=_^Qa|F3J;#
zTi{NAoy~n(zblb}c0vj^e@jLRl^jT5r-v4;9%w2~@O8^EQFhqfyUelk523|gXIc62
z)FtF}aiW@;p-Z6Kz%Ra*^_~AZ6WVgOxG=#cmy6uG>3Lwte3ETWjwz%}LU?7!8IE3^
zijyJM9iIDRx+2gI=x#L}&Z&P;8YQ0ya7F}a6`}E;c6*5fsd)aGC&)CK4xg6Pbv2e%
z$Ff~=Lm@0oxMEs8NUP;b(X$1}_3!)j_029wW4}Vk66L%$VGH*TI;Qv_+E!8};^j$d
zRV3c&?$K#uk4%0ZcKr%nKotEo`59?+3-dRyynJ{^_7npBngzTZ1{Duey0kEC7NM;S
zG9GYyah;>Z0Dfh^r7`2Jg1=7@MWRBq4Hg9#e`;U)vArT(t;Z65`FA};MU#we+WQk<
z%bnrY$VPDClqh@11mR?g6eNAZNyIkl_k1T&fgKEsxeA<&DP*}@!R4<M^$Mq#E{}3p
zZ2lqRFr;7#d18cGrONApkGHTc-OV5I;f%k9Rbo?Y-=IOtz(c+wT0>?~;_zIjLu*uJ
zn19nif^0?uo6@Cr?FN)RfKOiW!b?2y3Tl<667?Y)hNwk5D<a|eZ*<`A(Vp^FS%`(g
zO09E7YL*n;l8!T;v}Iz`zI(E95LuhZ^8ZF5<vc;Kl~bTT_9U!ASo#rM1sUcF%IzF%
z)VKyS$>j4=mFn()9KDJ2a2(|=b&&lE6NY`FszEu8!z5Yd??SVLc^obo$!cnZTG?3Q
zr26W4g?1n?P+06Yb_h$Webk+NfsG|sRlwU-j@c6>O8L8#5CEN_nZSh?AoaDEMxI+<
zTX9Gy5<8YVTe$DM9d40+&Kp4EW{zzj7s+8axE&%-Pn}~+$2ISdz4t(vpWc)@fO*Y6
zlCV+Sx2Ja@)>T`T^<Pb;^`Nxbj&iiq*G56bm8t68$1;BZ2AffYvKR*>oor*<=RXQj
z!d?zPY|x4qa%uM}L%qtCAhjN4zJdX(tmo64jKkVG^S`=Rdu0=7)oImhviwy=i$3jz
zhq};MGNgQ@fI3W*UlOa-?@ETLjxvhl67QU;^!n-sS0YIjA#D&z;~`5wk>XJTun*g*
zKI<?#2(?3dM`jy&7#jY<p*H}x8fp1N#_IpLcgxD$JJnr}lA>n?+cr=kr?~G=&*Q??
zus+Lj;GS2vQW?gHZl{W)SwAQ{(PDTs?UB6d)@Sjt7`P#olyW;8r`nXPlvz#I3);08
zsdj&BGC)3zHT_@LPAVUjS5#sHYpyy&+TFKD55(Ny2qs|}(Vg^{5^cixhW5Ma3`>HE
zF+5yG6CoRy`VI~g?j4jDsv>_E$F6;BDfxLhpVQaEyoigdwRA)LoI)Q_TiU;W;-jn0
zbKp<W(*b*N$_@q4(8kS`w$yrg$53%pcK5X~&w&cU5}tZAhbq;a#jWOldtP06M_=*L
za~g$`AY%_wTQNgukd@B)0^1!^X0p5Kk=)oVCLX{A$eea0Sqk$AKM3imzu5DQ=A-}P
z);*!V4av}7QLwr2kL2YP#Tjz-w-xQGS*M{m?;W3cfS2aoWW#JLZcRB}*j0|S_U}I)
zlO4id>BqaXUpXs=76X<kLgCc^;Ha4L1;k?7F^w}}KVt-#S$!&l4y;ubmS+F>=(h3g
zx%tKCv%fxl7@t|Kf4A74W=Xp{%|5urI?42%_3EypkKEY$FeP(<cc*|Ow>l;D_`|;K
zqMn&9BVK%uzkl`JjPZM~p0?o<jkxFYN}nBl*!1$`7n9;gzBm$h{w>{%uTaZq9p9B(
zW`eYQ*nSz56EDlw_Q&(yJv?@Zu}qfA)YENVqUMLoS(?1=*WU+Byevf<7g(>C6lTr5
z)}j2Ok-ea8?v=rNTdo~EYjo)8`*`FFSPhtZvlcn@#-jx7iWOgONj^c26VXrT=^q#o
z-ZNyrB&JXId<k`ddx%u3Di0~P&<uXF#xdY253ge3n{t(mqx+u}CGIwnk!+QUk<jGX
zK#XRLWQNdh<NK~|I%pP=lZ&kTB&nG_ee@@e2xE_K%OP(}sc`5BS8%!Yd?osD{>dNt
z1$DmyPhHN;jOHsTD7U8WN7|4eGnKD|UAS~fc!FpCcz`?63nMYJIkbp^++8qop&aj4
zrrZ&GGq|8iXKCKUO9o5b1F6-VZ?`eW&dP)W;CY<6c%RBMHKTnm@k4?2ZqQ|Pw=&v&
z#h4?c$lQ5?Fw+=c?nA=k8hd6+1X1?LahPy9i++bH#zk2ylws#im};S0Z-WD~*4)oT
z6lpg&doVO?V-(05)TN92^+p5#&I1%3rYxHPwF{qrN6}?6kOACzyOG|_@ne_avxp6}
zj{=ShUwg28?Wxy5GFdrICrN0B9=#0%Rr!hr?3%1%W<l&1ig8vl<Ee7O-5)jzI$Qu7
zQTb5u#@16tuqI^dma#}%$FczBYfaRLp1X}lCCjfVl@il}kiQ|>G1|;Z5RxkS9i<pV
zr)bm%<Z8QOF9u(-v~B6l*uKd{iKm~wB(P9E;a<~an0IK6fdH{&EV7u+W}rI<_s?%`
zB>kc13y~v<>($P`ag~q*dZ@Ss`PYJ5yKxs`imet|{4g5b03XHD`8S$VIkBY0606tk
z6cs))m~8hN<)LH>7;{X=0VCYJsmRh^t{|)#Z=g=OvzE7HB#>DK{*0)eN{yk&MB#m6
zv~;UfUbH9ZNKaYVjE#6S)y?~*4q2c)0L!&e@12_Xg%&o=^konDZfdh=Z3={|>L`nG
z!sK<g*gfdj-+F!?DF|Rg{(tD}??etV+h0AXJ{(f@6J9f$z+|KYS57DKh35Zw*NQdH
z=uYY+edRtN3=WCPP;{N`DcJh?!IZ02`GiwDN{w=|`U#Zv`~3zYs83Jxgn9L*wFV!g
z_1j>^wNI`Rkp+4A8}_(t)t;O#U#FsTAPm=2IK3Ac%Q;%#*N%_<>-YP8wY(-{A->7t
zwYv$8R_cMqCV=SWG^DTeJ@mP<d8faVl{PoMw7Gq-eB`Sfb5KC{#&meuBB#o+pP<Cf
z;+Qu^9;{mMZv}l@F^Pm)<{Zu9$iolv`3e&1ZwPC(Q)o1N3cv|K(b73UKteV>3~?(h
z_>>25_JEE#^5)1B@T{yG_Ow6V5sS5le=`3E${VM*R(t}IEVEW+JR(U`WpEs|0+V<Z
z2hedM(P%}#8xX;R<Q!<EkAk_gFP&u?@S)IVG$tsp|C3@3YJoN?A4-CIc+me8kN}Ho
z2u;O*5We-x(U50Wl|4bMAF(!xH*tZOzW5aa9J;G73}NL9E%@?ou;S4EZHAmu$%d{#
zh|0g+iupucv2Fsg6oei0j;TwiCd#Kc^TLH)=#|j)>ExPFuK3xbcui=Bd)%hQTQfFN
z8o0<|9KjD=y4T8EzWol|0d|#R`Wl|G9Qg@S<^5#QQo;`og~+p*xxrC@M)TMZRwL3&
z&X(VOL6qgV7K%5R-tGF^lw-bHIan|pb4JtT`Yc|ZgYu56_|jq1iXiZd>XCZpWCFsR
z_4i3$DKb&=nhT*qD-VDMCIdaNJLPC2_n8##?bGvxOb2}9_D>AIr2lb^HkcxH@t5OT
zx9F$+3G$)5;0hJg&kO3&lCsX=d06UDC{#B{KaCVO$ZqDaQ1d&WqdWa78)2wI*iM=s
z{a{pIzDbdvZ|Jug-<_@-sWQ@}Jkxzq+gDo_V&T^RJn!6pe=g~gy?y_@2M|%O9jj`6
zhlJFi#uorPE5h|*1!oth3V_>WK+EIBwVH>7zv4Ivr8TU&%LcaHuDw)ZQ3~G<P98_>
z_}h8}{Ed`%RrlzST_0E&OqxUXkj8!mXK;nG3eK-;GH^Pc`PN@l-fAg%_seYxn*3xw
zP2s+iU;#XP!@I4&&u`s<HF`o>VB>=h5R0U3Hyom)Z_jW>s+agtg14q7W;2wHJ!5qQ
zgI<bjnlPXKwI2XB0x><Fv-c4+GJjR#;S-y*;gB+RlVLqQ{~Hl|E*E_}3jG4=UBt`D
zq&3}WNKoP3Me+b0%VT?<ewi7j$=SJWIU|eZOU6gARu$ekI4E(`doxeW7PZCj$!aM}
zEU7KRD8z7j7$ZMb(gx?K!t<^yB<)ht;Of@ETi=LhDd;(9MdMWi@%TcJXzxi!Sg#~Z
zFk8q<XJFmJK@?JH7f0LVdH4j<JHr2wo2-d;g@bx1Cy!WAE{lIy--s-Q2nXp*saROF
zHNy<ksG<v7L+d_5L7|#kc6lK2pDfolk8Kt19+|RZ94pf^Xx^SyAl;4najtCIOdXXl
z=ZOjbZrHh5OM&-4vk3au0Jy&``^UC9;Z|D@-xzmm=XJv3By*MKH))tM)lb7v6H#N)
zPq&7KO{B9)%HR_N5;@Qk^e~+^%E+U(P7Cv=)?Z-$crcXVxTBc5X|iDYgDru5-dDJU
zdM2~(h3jcL7I|M>YjUI&i=^6Gw~>Pyaa(PVPtOuM<I1i4YBm^3Q!`~IkNh-}8t9Bl
zZH{lvHL=QG<z9!<vu<tD^V@;JO}tQBEnv=-{Uu<quVw%o=HJd+%H|{pBiq?QEk4YD
zwq@tQL-`cUm=fD=1`fl&u2S~f0r|cf%htKk?1WDxjxM4uG$mhA9+{iKXx_b_yq~lP
zU62+jCx>(`tiPB)tV~2CQ_~#Os3`I}F(StOC<VhFZ_@JX)U+>OO>*#Ss@tmcA?xYK
zh*2;5OR#W1<DBhT!0AhjRQ6l(HL5lZ%2qbK_1ceC;iL#NEI#9M-k};3T!yWTyO?9<
z`RWV?x<ONwA^aZzxw0wvbAG=bnMXqS9D^Hfs)~NaHcDa8T;Ebov=euMt2hS!(6(^m
z`unR?SwlIzP*{&ETk@sh536tv5RO$~SN=E*Gy&*smGZ~^GoTWVJTSQywRDjxRqRzs
z)Y76<`S!4S8?C~5Pml*c1Ags~&A$K4Jr%`Y$e4#ra8OHvYa+k958P5V`&LqBoj>**
z4SO8w)$?*wseImJxu+`U&F<a2u!&Rw{!XV77nQ2vRv?SYR_RE=K4zhr6y}dgK%YgK
zs<hJ*ckDn{-BXsWa+re}1w%*g1WZ`Z%Ycr9fG?C`n*25DddQtN`#_f{@=wt7_nV+8
zmgK_+wK#(=G)GPh1!n0Kdn6jEqAx%q_oqBS`ElN4Dt#lWErAlg#ZjeNM*#_?PgszS
z%<vRZl~fzHX1_0Xb3T!)Y*jRdV8@Of7mbKyY2{<igCV2QAf)6c&d@f02*gNB=B!JX
zKHaaVs&Y%}Aa(~jAF%cIC&yayX7}FWM!%?Hdjj}F5H~k6%LiroQo%x1<vMAN%fqzS
zC7?i><|P=ZvR;Dr{Canx{WW@>rd?BMD^Gj>=wbJAv8SYR#gq5=b?$twa+g{|_m^Li
z0uaED^mT3VFxFSJKT0;hnff;8mMU$-6P(7=>#u?y=C@g+DGeJrpWUS1Y4zg~s^~~8
zPJcjvg3X`757xcZ7muaK`o!7~_x;u-?K1c5;bZ+RL7Mx=)`YFYj3<ppiX?JK8%OD0
zwe>3f(^-;~9~d*?#OF2LwO1a3+nhE|X9Spb`+bqN(yMB?@rOLvp%trX&^CA{-a-FH
z$EB0n(?Lr9OhM)@|Aa8`jOaFX$EXQUw|Cc${}YE4&U8-oJn9J=@x!U)n9NQljie8p
z%YwsI1N22$O8HVdR}*jRRu3k6r5aM_e4EU`w)XNebWAW{V74=f+9=;KYRyY918x6g
zo!--5<mDeW*E;CHC2*1knhDC^NMC>o)+=6olC*`(g3tq;;vq(IhYgNM&n_J;5<oNf
z_I}d6rY2Z+d8-oJH5ShZ<W9dUz{?Q2oiAtgurPWGSypARHO}}m(%6&Y34DhvDWN@m
z=o288PXmnuzmc{ZAb{4N3NJeE*6C{rI#slY64VcPk2o-~D|mqepEC?XQ&8-%c+tyF
zlFK%;c9TKWW)O=74oPS2<S=yKGNhqarGLZyzEbX9<X39u5T~_r>_?9T*a&lq`e0;6
z;^ZZdBUa+q?Pc<}kHfQ&{{RDTDqqvOnnczSBuK9^tlAo(`tS_bxi>2pxg|n1DFop^
zZR|<zl^sadZjES0kJt731(A+gA>6ZGYWfxhhK*grDbqL5%w}<-H!wo1UyNDC;EqN)
z8^)%@fTlF9>>3UNo@EjuOa^jW?2h49VuZ>xpPDK8V0KMc6UTm)O~bL_IGR1XaEypI
zZfUk`an!0*!zM5dkh;UR#@3N6G)Mpoojd{tmMrwTW*r)6S_rE8LtgqE*CLQFSnuFJ
zIsu$$9Yp5hCPj4DmSZ6dgLtpLqqPPw%7I17yAHZ&i=4!@?$sdrvGN3Q=|b0Vu!+I^
zO9}ZbQrnAyX9z6>P4%U-g;B^joEnjbb!=WX?fyg<qo=UG-B>kpF|y=m!4Qdy2l6=Y
zDiNv7b{5~0(vkEQJ2@{GObUHisj}#|*}USSCj)1BmIgM6_T%@e$vGr?k#w8G>{*y;
zI4%X@TI5chBtu*l6mOAlwuPJP?`0?!nao?pX!IS8h7!54LzxL2Q8+ZijISsC``6q<
zEFjrl7EneTy_ZtgWd1hnH|nSP@A>12ISSrVCewqZ8#15MR6=RgA98k=k@lvQ9J4Q6
z$bah}b-JbTZU1xXdVOzV)BWU&a;o2f>d{olDQr52E6Zf7r))tZ$&&;fkZMAg1;_tz
z)b7DcA%&XH*5H_}G0S~Oy63{0CN<{`7f!zneiJcOEfRRFX(w6fD~#n9ZTfyk1qtv&
zAQZ&nyy~fErF<=nwsw;*9p#N=LpOBS-r?-APDE;kydD&L7OjZ~4r7nUimU`69@i(o
z`|FNCqCe`lXmNm`nFW|3fBe>g*WnITmG3s)snhosIm}%tt@rLY0n)^GA2Q#xf||s-
z$nRvwtr6GIE9*~{L^*1MV3{!f$Tf_gik?muMw4YNGb^~7WNrh*;&<@80R;5c%xEx@
zh!w7fyZ7>zM(a#_hP?*1b+61Le%7w;+Ql*tGN{bsr<BqlB7%EcW$NS<+t|=k+SWKX
zorXq2OUVrY<peLitr;@^&)~F^I6}&dnvk@!F#2)_?vlW@I}w>65y;5($Oz_)_r<fR
znkH-<Tfv>8OHB7y4$!~4_m_f0_@Nw08oMkQCoQ5d%d#{FyT`ltXMQQP(qLDp-Kxv3
z%t?w+wZ4N2Vtm>td^hH$D6kM``kCU`_w+K%WuYF>kCKHdY>IM$kF@@j9coQw9Z5^j
z@^_qQufEnSujy7p>hyO5IHxn$(q$1ISHzAQus0a~!%o&3sB01sXLBa-P#U@{P|L7N
z(9v_>&{L?$wdcW60kTZ}_+}m}k9iY$+3Ia!v>!*Ljr`3RwQ;u}&O@p}2%5x$ktx#R
zNd)Fms8<6`wE%rTQu-YIknXJEcI<_G<|OYt5g=<=z6?R`<7H63KW?Y4Y%Uo~Xs%^~
z#2o*)yt^JDh+GE111_}4!47mL(93`mj%U!O_~>Ek`8kyS{_ML{D|h#vB;RmCM8XG7
zIMGLBdmfVLkGcSL2B(Rh&$@cczdz#NvjH9!nue30=|WCw-fXf|3n#pu2^>1y^MXt{
z(u<p{dV3k7iL*rJn{SU$9V^Dd@);Utk%>m5Rk(T+<gck!&uC=9-~X28$^2tgH56zD
zvI7}Q5EYvJ4fEV3vo`MveS}2OnO-r>C|MUh20<0yk#%~y*w(n=coYXBgMCq%6EgI=
zrSTq4S)Aw+y?ZY&?dm~P;pdzVCOt2%pQQSD%LUPVPdD>>p-Nd)({K}Y6Hdn$7DfTm
z8Y&t^CwFcg9<d%SO?PioSt@sbFRiP{#Ekc7AVZOLfS23r#FfK5Z}egW%7|VO-&^9S
zG#!M$VK|%=fjO7IbzvqT_US8Ef==|cSddGKB@One;O^Q0>LouCxc2qZcH*#-)#tEo
z8A3lvS0seoeO9=AH76|z-R|;u33F%zAlXQyKV>KXArq<+!1=*gyXYD=>A$}t4@P>8
zQ75WJ{!Cn<X+7B>)iN%+Z=w$rphIgnZ1`~3aoPTi2MIj<dizGv+Kw#aEM>kd;p>M}
zZ}#Bim4T(vdj2+E`P&~mb+@QBg397#zYOPFcHI6Tu{&k^gDe&#TJu7Xc5HjbMtmXD
z$2VF`&h=g&-BJb{dx}J6N`$`WQ4&XIemPdAL-e(JdTIO15w%~8AQ0O*TE!aJOl3%2
zMT~svZH*{k?JkCQ++pCkXuM&Tqq_(OnFs`<N+A4LTjQ);AZc0pDw)9H+amNXAU4ND
zFkAw>wCjT=omEtCh}l0H8l>{~rvqq-g8wWzvRI!oH#>dGagf~$;U=G9HpsQj<v#*I
z01;_RuuReu!m3Jdj3Mb@)Qq?B1PT@>UM?grnHAoV1)zTZ1WGl0ON72kKS-M#f9j6}
zQfke%nXnRu3`C^LVba}aemjhPlR0K}v{+C!yMu(MyLHirl_lF~>p55{CV6#)$G!X(
z_x^ejY=k38IDu2+Eefp?-Yw9oOV%R>e&rGsneXR!+GM3Hou7s7j@;+&xQotY+S@;G
zU~DGNBFL5VQzalZV&bLd=XdIcIkz0ybXDng-%X9)%<ajayiM`9N#0ky`d#<{>vN;6
z!o$Se+n09@|M}EJHxG#deAFTFi7i3-o{Xc(Zk9`ynS<I?Os2OjEw$ooJuO>zh_u29
z6Kw-<nbp-oH>0kvMow<y2|_KoT}!?fyTsP>%I^=4{zZ3_Z*#*6@Zloc%*H!ce1bRr
zUYUy`t-^^E0vj~93efvMX7rnEJ&izf-hlDBl^(d#W~>k-v(5o1$$s{Xvh{50*1^$H
zs5z;IYkPy+qivStQ#)sIrO9H8nhjR2JA(A5XOXR#2uIK*OH7sCTtXt-^m#;>{`cU*
z%A7Y3u}Gjj)fn}@>-2Y;n1|a~1o|{ARw#sqTdN?}OGxCN3eT0-ScMyl<7VI3H9Y$m
zh-3PKwT%Pwk^VJV9wO<9dysUhp@+p<uv(&cyh9`ixAnZ_PHJaB_K&e;KCOE3Kij%g
zV%Uk6Wvec9Swrp@I+Z5tGS1PbKg{YP-Bda>Xnh^ueu$B92}LR!j#0OMX`=hmG_82D
zXpAE~ehMR9`o6TpIc^GAD7?W&Ym}|V3)c)t%N#Z#PMeV}c;I`hV|Y{&B5h`zFw2Hl
z83aT-mO56q9D=^XKRZ~;G&Iul3eSZC-&I+5?DIaN9`<j3@dt5_60P^>Zc)EWXx|{<
z799tkv@9W}T;t=rXTV>L`srWYJ%IP2>e&m|f)Mex=1Z2Dr<1%m*~>y#3f<Oaom@UI
ztfg2hVeeFHC5NKtdRtFlw+@l@WnxJt&Ih(dXqr&_yq8bAWs?dZy;MCU%Zxv(!ft6M
zr<-|RLRjGx8?lmE??`%8@^uxq-OA%-%x1Q>PlMKSb_sH2NO8959?x@4hJt5Noq3at
z4)9p@4I^fM?3W)j{qAs4hJ<X9DfBifH}XA{(<gX0j8e`V-bI8U7g_e6zdEQP9L~cf
zkSgp~rF{~;V>qK8Y;ub2hAC3{>DWr9aeJq(<8>4?@UPvaDU|AmWQiZ9`Z7wwk=9>m
zS2L{2!Shz(d{9~|T{~$_2VeRv(^l$$%~u~0q5d$kJwxmEM(4PO2falxO#D?^J`H_w
z4T?A_11}Lr-!Mq>W36A<xQ3rt;?}{9Mq4ndUwHm_7Fln35m^uW^}^TGu1?l<Jc}p9
zNww-u^Wd0}tOP`OFrUId@yhI+Vdj}n2dOb6I@2H2l%JeU*5y9reASLyAT27EwoWtv
z1b(@t*&p-M%rmBLrz^DPrrq;NfLgNk9bU}`zm@iO?>|I!hJ6|PG9R5T_zL3Fy9I&!
zlUpAMR#s-syGzm^^l<pPw#pLU<i-yKW)!#*zQ8-$`2yp7De8|8^(Rqb#txD-;|W*l
zE}IBZIHF!78i@jVHAIV>GRU)N^y+b;fyq7V<G?z#-bW;cT7~AJiarr(>&d<k-+$50
zPh&|Vb2fii=@L?MCs4ulJoZ6F;xw!rlTGlolL;2IcMW&X26PdE&}$3s_i|EX>BOwR
zkfrQxyzf=%h1~hG+jMXvM8hp#x$>$s1fkVJU5}s8=GKf^Cz_`pO6W|lZS(}&smr}4
z`JK24+%z*CST7`NNu+VQ@I9>%iU$9QD=MQ&sZ#U+>W-WLUECeus3(DRytqLaX!4^y
zyJ9H4Lo31Cn#l>6&C?Gk_HbHR>nNdL`jOR6{d>xG<|XU$##3}UvYb1WB|m;t#W|`S
z;1vqL)1)Rv;A<sLG2x-ToZ`;i|E3vhy?wbEEV-P=Y-gN%-koLtp0KaRzMZW>%FkB8
z$$y-d#tPtJkp|e+eg7Rih9M!)Os7H+?l)QVo*_tTB_@{hpsr+n?x2qQjsl3luJiuT
zU0(g=6`d`5du1j7$Ix?J=XP`Y&`Y(%$%ik6p3^$q(iEN{MzIR;^8T=GJ!-iFZymCN
zV?SOHlRc8i0IT1Z+fm5Krs0@0OtAr{Bl2j6h2=WMX{Gs6axHHxI=fzu_^DrNgg+@B
zjz?=@;!nvVcb{qRAmhpWgJ4w(TgIxJ*5A+t*LSH~hr_Tnd;PkM2so<Fc8-5Oou$>j
zwdqb>b-mJr8=YLC+--`X_vMDO&#IY;5t>0r>N2+uQs{hP`y>?B;wd}Jz$_(trQ6*q
zpKw|2OZ%jk#=g9pWg(U-Q{)L!<84i02&Mg%cW?Z7rLD8;Hi^12&0NFhzJYPvY%A9Z
zlK=1K^dk`o)%TWOly18DFdB!)j;*|Dxwz}wM_9{QYAq#ZuHoaP{^zvW7Xf`%DsdDY
z*kp-k+|q1)1=h}J55rgK)od{;O|*b^F$Y^x+n5BCk4La`8T<VN%#b;9=d7j)VNY;|
zSvvW%p3TaODE_RnM<?akW7I?0s-YPsv#}{5S&x_-^$dbF=nvwa)}}^djC$sbd!u)B
za+)YCqTUsBTMY#ylxo^Q4*Mh90TRtxj;Js)o1Y#G*4uA?b@Vqy97tT|HtcZHWQ!bJ
z{mVaUo}fLL-G<HOf7^o+#>v~YEQwCj65|Y+zljFTW7Nlv;Ru9T<w+iS#KVxR2${k+
z9q~6~zm=@p1E@%W0|D^GXLo848sB#mU6H%oHr-B&7Hql^<qVHKf&JqXvZ@`sb+j0Y
z1U4*OIOEl|p?6jkqgh{k|ARKEf)zeB$)TA=kG}$}owT&X`2dz#c)?hesdEi~C~b#N
zn@y)ya9HdzJT>CWvOt~S|JEKIh*FE}C09Pry*oW^aUtZjxG`R%a?%$;6X>g#Tm|t!
zvd)C|v#mpx&!s6|4TrmjY4)MWZ~KP*ywue-)2W=E{bWJ;FJE7n(r4N;<HmOUQ1Aoa
zzdi_=Q2hGbPMibqGV8rMY4a4aq1~u!+J(ae@>FF`Ag6+h(k<CGUCOPAwIUb!!1KZ*
zj9KXiQ{9Iol=-_6WLcAc4MnL#$cU_NO&Tp~r|vkt#gpR@-WGMY--Yn(;{ByFn_;k7
zNBR83C=a~J-u<?}n-{71z83gITF=dHE84%?`$s3bt#aJ=<HKpNv6!%WT|^8U5nHB2
zq+aA^>UtyYCESjxyuS+$Cp@}tIGnje6+%4hl`~!~Hiz7Y`8yx-b$fXLc)n82U!4vP
z`Svxf^dW>Xk}Uful$c#y#D~Gzw%<#Ycp|czYj`RELYMA;!&Mp`o);h#YtwL!W)Jzs
zN?e_AxB~4%r^Z(36lE=)8+uK-Bx<u`jBR?;tK{S}R{cO)Tewd@#7!a`231G{*Rg6H
zF+K)w2-(LeljcZe8x-;#FAQ>-2<nP%sG*gORJmt?#25CuEREV+r>J{4)f@3JVf~Jr
zeQlK&Iq{-n(#sBP$cbAXEU#Kxwt2M$M0^C*3=P{3Xi7<qoWEDnrTw<VzO|cw9ZZ4b
zzRIF64cQmT%EKZFtu_D{6aXn4S9}_-2bAO-sypF<U2nI>k{DRh;@7%tFZ7pX$*eR+
zt)t(=`|qXFI_ck+8Gj%aGQHy@fzgG|G4#bJdj_;NhaIQeHBj<55b34SqK<#DbbEq2
z5!kA?$yhZ|BsWv8HLHxo(BeWUUXd2ACSP6?I7`5i=>S)(x0iM{4y^f#Xut_MhMI0;
z)WiHl6oRD*)EIT%WzLiB@h7|F7=uXYe-#Pz_`Z|(ZG9{1`DBvxRE{<qilCIC1geZ|
zKJsV8DLbr`ue>ZRS(n;H&T_^=;1!HO_{$_n#(QjyOMB@J87*;7@TjIOE={`@s$OR9
zY!M0cnHMA9oK%;H6~y?WiK&UV8{Bk9li5!-Mo)u@x9CZftkV!E(<zF`Ti=1uG@UpW
zn9XblZqwfv?ryu5U0}SxZAXMY{9jl_)Bcu5-3I>hnT0t}MSANE;n&Ff4Sb##Flt=r
z0+KM?y<}^AjHYjb0vHa*a`!hbldDO|p@Bv@P*_)**@r(d^Q8XwfwA`NqH_i-6K^L`
z!UcH~@Hbt2`YTgSfO!S9B#?;D7!2(=D3J}KHRRN$FKuq8t(>eYNPB3mCOswGqYuOm
z{&=@sD7i0@+=h3M^3-)EIGOUM@)gvuF@*dx$<xB<zkfE9YX@?wPXG$JrRlj$7VJZJ
z3*A|TWoMkuy`aO5%`%oZKlh9M0+bmj!biaJi}pz-71Y2PB6$K%CGN-avF%mX*8Xj2
zd2h=x>V^_{8j67QZlhye!>tnyswG#H7ZJ6tlJHgUxvtH$F%vJlK-XKJ$f59_stkE_
z$6Feqmak0wO4!eb=-N-3LG_^_Gu+`@+ci89?7`=h^0H;HTz2acH$VUpEXU9>AtHJV
z(hQ@!p;lja|8Gk_sIn3f<}vDIFL@w*TSRmls1Un{80A1~SZiiqx9vHbN<bK6)aZt|
z?y>#wKhRQQaIcUp>S3Yl6bI@C2t(vC+;~Z*^5OMEEOO{>+uO^#FO816Q93J&XSM@O
ze2m(MwTO*S*20CgP~0n%ssUa$ttTgTk6mGGY;U1_bj~eJ?4vRe0Uk*p`pZ6=_fqhS
z6Xg;-_j8|uXpCc(4a%!L=+qix>pA(DfHqg<_9qFhy`9FWv)<8oIzVqo9D3g7nNe&{
zw)(avcD0PiM~EK1yu7;#kJheIJ2H0rit}Mt4+8KUX|z?DO~xr=CK!&#`uQ%Y$2xfY
z*fUZ$+K~etZ9*r!O<-5tUE|_oN$LX^>f<qk=KJ@A&+h@#NPisT7`uJyvA9y5{Qd?<
zWyzbTLT<WtaTjqQcIzm;Qo%E%uT^o$_S3fMq(0fFTBw4A9DCg8DywWg3zu{-(lu`U
zmVjQ^5{R;NfDWqjjwWV7ekL7bWDO!0&=2L|NLR?Ya#KY-RaP(qo}d3xLeRb^B6^PJ
zZw>nCp(T@$Q+fNdG7g)~)kLJ5VWW~9y?=1DbYSwl+w*3`D`lkn!1tDo8{tOUO29;_
zD_djqu=SnZZj5LU$QyD==88Gju177Rk@|XhkAk{CeB!1biX+jDvN0|qFQt%1cIH5E
zk1jj=?c3^Ly!E0;O2#?4UgTrztK`!kuvpj;GL~ZrFpiXL$6d!XoU@9#CbrxgVadu3
zF8v-z9+V;c{Qij=l{C_N=!!8=8o|flz3|g^l{JHu!}2T0V$88`enfO{?(8Koafb^{
zhT((!_z@omWIzw)cGb<4qmeijQ@C4azaYq31F4+r-Sco5K800i>XsYrW!~uw9WP%W
zG5otMA2{b-hPUaBi6Dx5h<LU>Ui%}!^;R}c8EjX<yHff`RI1Z-Wye)3jXi1F1reb=
zb-zsW*HPDoe7$e_;^8lq+@JjKt-#UMFO-+>atp!-$nV3Zv>o9ZZjTmz#SLAuYqn$5
zOPX#pPEXmWeT?M+OnLOONA$;hSg&8F>ZjgqelFCIUBrKVoV~p_m61XSIa>Q87U<05
z0TwzlzOPu)VM!6uuMja>$$j8m=<^f)<K}W<LA+&TQ#pQ`nvlJyH(v2P<2Ak4q$Xv?
z!{y(kma6mzv0zF>bU!#+XfEi?75>psBfB~#Mz^vfF+#A(TDcumZUm}OHJS|9tj^k_
zc{abf>n}mf+o2!fpO9Q6E^+Z%Vw;(5s(7aw{3*B;Cn5GK>6qOS(OV=2FR<e7tFQBq
z<r@k%mlp3p$u=Kc9C7m4j$gJa>N4XJbd@^KUiZSS<&vt;*uQHtq$T%lO%=t!HQb7a
zJw|QjfLAVsVx?~h$fXI^f%+2(SKHOoOj9Ifge3mf^@yY5r-)F_WQ7?$9#DpnEKwOu
zr1I{HJQ7o9*_XWAXV9#Qk1wG(l6%A&maC9ME4~{oi~~lBh6#DWpjMn-m5i%Ok;;Q-
z9tUha)0a4EKfS%<Ct9qzYj{Q>&y4J)S7UtBq^&e;mUB-c48o;@aVFgaxz<3;9@;9t
zVP*wIBi4b_a0IR&ZuU!l|Ni>6^JyiY_i|v)Jt=y!|6<dD)IvP6Tg~T2Q|c@lP9Um%
za;NVPV^3+|9z5-KA3X5o&u`E5dMuqK#rZ;=$lZH*blh7MiYpfHk*QO@*aaZ~G2{#X
zU~V2@Zo!~#vxjJ~<|m(bz?DYtAG+FQHjQnQb@oxv(FEi%^q!J#<`Dk!6hG|k6*&<1
zMOxOiF3zAcx|r2P=Dj(U2+gz?>qpXcP_QI?rXccAlnn2vI$c@HFBd^C*$F9^htqh7
z4w?uI+Q>0`y&qT{WaGFq=Sz_dhoU1w5`Ecnd_z&pVIYquo@}Upuqf7Nj(IhWxS8`y
zzKu-t_{kL|$%uPGXio^w#K?o`6^aI#MLM7M)kjF?P&abbwV@(ld6%WvBEB?=@G~97
z=;eYgokMEp*_c^Y8xlcLLUFNwe4b)DR=v3P^mVpe;3s9a)(xx7+nx`m29y_F_m<{q
za7qPrhlG(|h*0Iyxo!LPmUA}=pXzDG5^9sPfsgJ3Y6>IeePH)S+@whG>`Yk&JaV}~
zu00p4;en1<MfX%rQYQGEw9an++qqBgAD6<x84qFuwDZwpf%LB(^=3BLxI6%4XVd8P
zKaW6XtBY%+K}SDPj#G11_K~8YoD#WyCPmKh7?6hfGo7?1i#)!fjGdrJ+pWzamwYcj
zcn(6CaJGc2c%&H>u3kj7A2>x9e#9OpIU?7t0uRtX*7J{_7@VJoW@po^3!>kXN1%_T
z(-gp<+}}r?(U(=de3$%FoWP1oiNj`(xhP2wS_Gme@=G4dJwnHW|7`v#l})PsXVxQn
zBjbku`b)hO#pFI^E$<e8`?#&-?}Ji+2i`UFNYWFMscS#}zJSk5i%LQSn3bal+7CYR
z3gvDDT6KPqexk4i!b~fm;Kv<|Hj#CV&H!NkPC4drE*_A3{;2F?>Qc(x@p<YK=^#rD
z7${T<gu~C$%l^>#L6&uHaMfC2;h*4Z(%s=jJJQIW`Q|%vHxDt=wOsvqd^Nz?WwxH~
z6$ab3T45Tlra=7}uEN@5p9|qa#IK!Dd%*NXeJ`-ZZ)ss(e{IkWSg<>T&L0NggAn}c
zM$$#xhhNbF*P<TnGd4Rfi<vc)qe<`^k4O;jOK&=wnjPQ}&rEwV6xM|tCmIHRd2?H?
zKwAeO2w?HC(2Y9CmB+f=puYavj1?gvW7Lxxw;U-}+kg;7j66%Bj=xu=)wckBWC3np
z`EHJ_TaqQ&**=z-V=kw2FISmxZth~bI5C}{_Y?2V{n^qM{MQ%sYLE3U^$(UiXzvg&
z21uI|{$d$?y^Y8wtv@iOGilr-@Xg_88`{DvXv-x@SGh2NGOKS~&@&VKMQ{?d_bZPg
zHnmChUVo4}?)Wo?4$grVIwcvQ&*UzWIzb02)^O?A;Dff*oa0o#pyq_*0XR`{dpA|5
z+s{Um>5pTetah}UVK0c<kL4ZFCLcPxr{e1B-78$Kfw`_a>1i-W|M=ssL%v#-t~-&C
zgd^5EE%GV5yJ7x>PU?RIPW693GqOBt3(Q=~3KE@M!$YkW8hc(t{5{8hAt2N`Tr4&Y
zx<OVTe_o&P;}32QT??A^aBtL;KKeXm9p*`bxpH7_qMc!nu70QlZ8LJWWkds=Y0+<X
zkTY60?D&AYRiz*5oh|@Ti<=(Ha}gt$-N{1t2<KjUka|?r)D2(LNt#i9dH=CJS0y}c
zvYS+9{=AXrjt93-z{*sfF+zHN-r74r)j`aeqb9-Z6nt#J&f8oCMwu^{s3dY!XQr(g
z0tZ0YoJ_+5Jp@g5Q0kD=K>E+#jihpJ?w9swG2qa%o@TqR!$uFZ(9M&Br#Ep8FRDD?
zZ27e*<cg!7&meI<da-5gFVl~mk&JD}2&g@!BwA>OSwOyc@a%Vg32zcz68H6TG<7^k
z*(>V&1O%tq1X}YL41{`XM(vZmnJ2gIk`--DaI@HM)=tI+_OTnL#iLS>zCe!H@q}-p
zWVo83_J%Cty@opD*Y^*-e>86=CTKodci)_52@|T%QAt%m<$(tA!?Z#GLStkgW;MhF
z65khdli>Ziu5%&xH5i|oU{8edf3@gcI%==``n%i|x`}Jb0?Egci;-?3^uwI+;PCZk
z4XTUo&~!pmNIfD~brhi~1~Fw0?@q+&QR$o2TndJ}AeVKDj<cxw1vB*$EZ+OUDlQB>
z^5ZI5*UQwyS!X^YMPnsIIkDJzaur?0t1C&dq(t~@&HpxNhLYQWS?rU)yRd|jNZzyy
zTT34cJXrDIFzi=<l)Gp|WsoSlH@9Bnq0IwDt?txM^SyIKtGfqek@LCfOd7#f{CRsG
zBv~phrk|@ZQdmYaO9%D^9Q)TCihm)^g*myz5)S>RO4G!GF5Ku)NM>_weEeYl6<sWJ
zz9JW=?8r8=*{<t_^_FmIuT5M`%m?;)k+TuZ>Vv&QtC7)dh6yF3%T<$7t2Q3El9J}a
zlU1kCd_?;FQlmoutS96eCD1Z%xK49ORyfrU*h%(>;xp-MCJKYb+q<YAc$CuaaPW-M
z<UYIq;kulgq&q2Jyh;X)$S--~*Uf0$dUF!euhsb(>Xskq<v3B!a-BK;oMjj-UfB@Q
zyNA3^EB!g4*jf{YMR>qTp#(s>A`joZiJ%zydy92ENk=6zw}G^uM2MgN?3Y0JEO8r@
z6&X6tkT`-!|JMQDJ>>YQY4ru6<}Q9qF!xRXzHEXz3}n`oYj2)C-h`R&3q2nH{y1>%
z|2!`@dT$J$BKKzSElE*0u*wOi22rSddiI>dZR#t@4eLE5)-w7S>C*AaYBe&YR)eN5
zEkEs=iT6JH)2j~zI=gZE<pzXjk9_g6mV11Eg$aFxsv&sI!@oX8r?spd#^T~77ndIn
zLMr#1zyzEHnO__tDN^rUvPEts`kzZGqFxHNU9zQPzms!j9&eX;ZE4+Nd%MY*x(~nV
zx(oF}mFmZ`_40Dmu-s_&&JC70)~okSdIa-ku&G<b1&qHpWy6go)PMy01$#f6reH$m
z1@}CS#p?B~$u^%XES^_XKU`D#b(01Eho-quS`Xqj1T_4sHcxWY4u82fRz}V#nu8(k
z`ny@9r}-5z<B8ov1gcei8IAXN+V!FotFO#fSJNDS?34uy8@K2ywBCQfQLhuVf^kcm
zt*bwV?*_LGAQ+->vVgwFv!KWcu8$j=m_B^>%GJc#i*toIE1rg3yYDD?iM2Ph99?B~
z&9`(vN+w8=+61nE!&?79faDI#d+;U_^x-4Mf)Tv>T%V{u8{8+mm)^UB_z<Ygb1Uk!
zA9Bhd7aigtwWegI@Z_HYSVOZ2CY1|a<vnUs(!6k|y0z71V`;rd$wcsIr2nUIL{Xdc
zxW9oSi2nNc4QfFg4@C1Z!p+JpYEY3VY3;$4@wT1|q_2*mBS}t$_!%u2#33j~_Y02$
zt$mylQD3h8AVh7}sHVSM<V_6x4OH(qX;A>uLq=5>=h={{pRxZ76Cg17Oj`+>Gz=c~
zD&huFeh3~q%?&4!n5)Qrau}0ti?;p#Oc(!?PAqg)e#mjBz4SiHdgC}qsQpoo<_D5V
z{2M1fVbUX77ojt*kV?i+y5ZYjI*?=yRWk3V=E03Yu1C3s&#C(1*8|cmrhy490Gqe;
zmfKjFG4V<<=#JEV^==w=9rFTIE|q$u&wk)8BZc<wj8{EvdyYWx7*m?!_BT$81Tyu%
z4j%}~U<pNer5tUFC1ygY9}h_l_o23^%e0Dv0O;W|-^V4ArqJwjJi#8}N@5THjvYQr
zJ6E5RJ;lfRiDmDGpZdkF#g{vH{9k}waAI<~PDC06Vg2Pss^fJU*Mj{l4mZARD@rVB
zvh5sn_?ltq`c)5`i<|)#e-Em7CwJcLzpcWv?~`?7iiH619O?s5%P#h(VG>oY<sYw%
zYs>N9+yB4|i4X1C&El%EP+)D4t!Y<nCwp4xMkqU99Z$gCHQdHuuZ=?Cw5BTrKKX0T
zIcD-Bjl+5oP`c;OrPkxHXuW}l<i2yUKYk$F<gKNF?(|zpFP48mL$4rh{J_H3L7h8*
z+cP-Ga)$t^fPIJrhg|se#ETow2d|?71qJ*2&Fz=z9?6|mTg2j^Bxn<_aXSJ}q6Oy!
zM{!VIW%#9a5ALx{>CZ*6RoCZ|o-6DFmLhAxn#eFAE4~Fq8U$1YY4!QswL&o@5o<B`
zGGCSJw+M@u4BGGB^>poLWcBkW8TydQO0VFYM*ximkpEKxC9-fo<?6Lb$upTl!JQ{4
zSULRS`J)TWE&k-iR7BN%zhgItEfi@XsIs-)DVl^*iRk$FrW3a|?&sykyn9%GJVU7f
z$rt2W%5I}!ynTCzrs<>Cexcq)3f!=dQd|bRfTi`owIfMhcz83}seMq5Uw{i7d-wN!
z=~N+=7%T|kxbfA5-uIi&g9|#nPA4|}AB>{hE_PzBjG-C#LD2%bh%1SHaLSgIuS<)c
zdB6s=<VZt@I3LT~k6S3L%lCnyZ5gN^JwVYqBk1$M-LrDF*!~(xnoeWX%{AXQZZ)V>
zDmEngF|e^$rFOTbo0yJ5uMrpgq_jhP4*(3Yf9>;9&lJu(bPXT_5^pKZagJiigK>!@
zs$%PJJleMEUUR!<L-ZN`t!c_ZN1u^4i7!GkHxc+<0U`vnnoq7zuD#;YrUc51XSR1z
zTbz|0k}FXjr<GC-tW>dWY}IX}@l-;cUAo-Oz!~L6Ds+p~G%4JU4+tbZW1))zyHP;v
z$W{c2KS5e3Xe=dgp+PC>B|i^E$94$h@QIG4RFgR5F-0hzdl%)NtkuIZ>B{mCsN8xH
z9Y>A&bMT3KHr|&%L|c*X1R&qbPrhO@X3$KbW$R?JsgLGYm+Qt+IZ|lC|8!Yi>l<>3
zQo5d9@6n}=t(@J)vtX0fcT}LM-0R5D&WLC~;<qG&zAl-mn;e0Xb2TTpvn}+=8BXk6
z$?b{Z_=M++*1St_uKGPJ!MW!5rwJ2c#tfX<aQ|<|S)<5SQ7Dkd%X?f{zWAYJ!+<xT
zsMz97p+}U|^n=p6cgnAo9<-<%iLZ5D5%?xA2`O4!^0o%n0vhSkAV9cassOO(I@TI*
z7q+&`sNO4Vnm-kn3?Gd8idw5LE~z#3S_hEM$vE(GCo$c|1X%T%(>kVicJnJ2F0>&m
z4D*`=K7g`(UUk)mXA-uG-9nv+>Ud1AOZ+GEWs7$Px&Mh$mRS=;mrx->TaU3|=$)<2
z*uc$7+|iqihvy%Y*5__AHclRbAjy@g;0V(tYd`gRC#>O$3r%`RuTU}LY%?Q24>DAY
ztL$*Tv18TgOQqr1lV1G~Pt4(oTaD{iJ?BM@5uz$Ektm5Z1ILfX5aS4KoBYdHm@RSq
zf}ZlB_*5e6XolMs4}7|gQ3r^Sn66M#QeZp!6x?6#_8A^qox(4Uwk!UaT$=-=Fb7J{
zcn&b>pZ8=f8Y*#VW$L5nkXz#dKson}=Mx)`GFUY+EjhNtBnObEDRD)?lDD)!sNSh?
z+2VqWVwNkWU4nbC=KR{3xOAS*AF*HI!7osN2L{_G|HdUIW7OBDUGG#~@Gt#Acv^u4
zb?wa6(Omg7Z%QFWMn~UX&Vqz=`PsbRC?HR1<*@kVh@#6`@Fyp2^t!QG>;Cal>G?-<
z7yI-{th8lKMdPl~PR1jQo<*gY*DdECYjnx+@Y|zrH_R9`D8T5KD!r|J!iK0WnOQZm
zQ8%*bTdw{(P4jxKEzg8MZg+Ll%P&*%Ypkm=tLKcXKT1yj(qq$`_+1;;@7;AJaOAMV
z3&-3^`AG`1MroFZU6^lNyV=v7Ztp~a5o$Ppj~_x%6q4K-(H}@{B6Qh+4z2G1VHGr-
ze?t@HDG`$OF}c*O+Hc%g)6>@TL0vOpM*q_wIuNUa4yoN%`x8v!AwB_cZoiDxSy<(B
zmfrJ=au!di<b5?O%O~8q^<OhIVVynDLuckrg7<3kGRxIYz=7Ehvg96IN=ZA%c8w}J
zyi%I}e4XTg3UO6w_j=o=gVY@Kd5~_VhEnp7WXi9=oponX7Lh(>4M)1)#terz!!>d#
zW@`UXgbEkvI?jkY*Ff{Xj^m=8ZWVC7dNW{X>AfguO&aF{X9!>QBX)jmN#NRhZPi$n
zzKM6qMUC-x$O{!u7+WDt3pmNIh@w2`lm<Px*QfV?5x14=^ZE77moV+rjVe6~!N>V-
zYwl<3xi8XQi}CNb5uw-@9pOI}AA5m`iQ`~ItUrOj^diD!z|ytj*zjh@?>$Fy$X0MC
zBle>dS75XI%tE{;p-yiN_8#+I8#nEr{y8t&@%f4mO=?u4>8>w<%}+~mEWriGPWC&w
zkam+R9=zYe<)iO#gw!V&0!?lkm=Kw$v;IRc1=+<_CZpAz9E^;*N<-oHC}iof@zwH;
z0L)m!e)(HGc+HgeTcmA{+)xt|G~GL=c6eM|TxRXON`R<W%twwKDFF$<ajo|AidjL)
z$Kmy^J<~j>$}?S2(#nlzThLD+?dQ7*XIGy+fFctF6G!q2*OA9<Jp0V7y8mR-S`94{
z9I-4Fq&u(wWazZc7P|2<CgC7hqpTcz9NBSL9BDj_vH1=22^gBL>GfdY=4ERKP4CY?
zaHN+`Xy>qKhK6%&JsK3~p!a&RqZTz>68T>wCobO4B&5n?gjK$E$Ud7A$>2;;n_U0-
z{jJ<^mcOziO;rLuIV>?bF@^p+UuoU?)u%4G!F(Wtby~kF3no8Bzgf%=1kUNbF|Ud;
z)UB*&C!$595lOq*!2g9@Cq>%VFV5}5_KoA)ux=uP_(pY8_6v?ecp?!338qPitS{CK
zN{8vZi#rM|(}l{{f05+`Js`&<0C`labLvn(x14Zl^;WHn>0k6AdiCnh=vBPf`@Hig
z@z+(iXG`T@2#!0|Iw!s0@NvWLg|%6@;W3xfKN+<Dz=5CF3@#3=s5GA4VjX@{!(4S1
z3DNdi{r=3Xtp2b8A;Pzyzv29!Z-Z%I(jMDPls_Qjb2jWfV(CIan;$O$!O2C54+^H-
z1LQ4lZm-btIb=R6v^smeOP$(3_KKqf$b2tpVPfG6=fuR?%NGxsC=QA&B}%!NJ64ZT
zup`jgzKU0#RR2yztbqT^v2<)mDF)=sLMv~_8=9vLB!ujGT5!ks150bR^4)0N+096n
zu!=@|WEg=r=^6xS4#!v;kg-2=S((Ob?jIMsO=vKfYA%py@7uM*0m&Apy%9z5^)nu&
z>++>5KwsA^?jFv{!(9=Kv=+kI$uBzV_fx=g<xWT|3^1{oINI~-cer7;Lx-<CgmC3v
zq8^#6=%A$oNxORhhF!fw*LSn^94dIJv2kMjlO2!6Yo$-zL#$V6(fRh$>K&WraJ)_B
zL)P~@M7p2*PTgIF!ExNz99rQ8%UhEpL?s1C0liGWVUT^ntw3sGx-@&)hUPuJN-Vz>
z6b^Yn3#^G)*spKz2q8l#L*(~|clNem#U>Ix`RECHZs;tA2Kwj?h4m<UV$xy@1p~W7
zKmxus?nYu(O4H&Jnp1hNFPC-Y%5wKp65(#5+fCkP+~V~_+*g_gV+=$$a{oog4X^O`
z{oXA&xckg@Y^PkofmL)TcwqnDu=Il{Xn5bndnmBO_Lkx4E5(~h>Tc_q)2WJvlCj-A
zwj95UsU=#L%($Q+;GK~FgGv^c?zmx@juNdK{1E?-(){&5z7myMzo@b0bp4pFVzF>@
zw->A%{pk3&D-m6AU*tLzG3;=N=E6S(V%bJW^R4N^!WU?`3Z=trT!uydn$-n)0RX~x
zx<w4&zx5H935`}Sg3af^Iu(*lcqn=g$d1TlFzyVkOS@<FS}J#SXBvl&`<e&=6BE5C
zndYE7FhVi2QqURJc6M^lb6(@$xq%xAbSG25so(MCq*-b=bV95NLg8qYvuM(2bz4!7
zW@`P=@Xc`>yS7!w92@o<XX`sZvpyNmpj~Nz5adM;tY@g-uYJz{;y=0lkR)B+(z16*
zz29FSj2W$Y!#;saM_%6W*O}VkFw)+S>BB)AkQmDZZiK2Xv>2^61>vgTpv@kMNxWe_
z*zPel*nFl<r*AqVBgf;t2Q`nUJ@2D;3!emb9%?sV6SI1C9~NL@JBjQ+s8)?6MOd1s
z%XL1URaZHB>sy8w%RRsQ&fUdPtFKjk^S)j68Bwpxd&_jf5$oXJb9Ve<vf<a?$GEGW
z2NgxebMiR{>y)8SxuKc<0@d_2b@Y~kp3}~Tk{z2lL-XRwrkcUOOH`7$3xE7NLqFhc
z8+P7gu_C!H)qT8dT8vowYZiG<2i7eo=}puJT!_MaE&R$~Pu_#FTrK$xU`TU2!(?CE
zbV~#*!m{?R9w@i&;cG7>>2hNYs=Qsym)0;V_IztabIJUBkU?rD|4zE^+%NuDe7v#~
z01VFEOQemAq89gb?0YduR~#H~VsfmJenHRBvxmv&<OExZv;CkC4pNi-3n#K0wv|*Y
z>vT7Zp*Ts;k9VN<Un{FsrAz2(?DUK0b56VOPb$PN*3^}~2OE^LIf|UY@?*P5vSeC%
zf=(?fZ@|907atH!&ueg_{_f2T#`$=uE3_Xpz}VQ|yW#PW2jI{J?<2`tdF1aZe4(X1
zbZK)tRSCXDq^z>Hp7HH(8D4tv<EYWBmM*Cv>VsU7S8_U(V(znpZgcGGwpG(V;Q#)N
zqiue*KWP3SkNdhf!xOy&n-qFB6qhC^Sc5Zo6|Jy$7Y3#CbPcU+D?c;R2D(`!JV=Ri
zhUZy4HUuV1xpL&w)uYu_G}N;dL`|a6!LYmYfq*~s`HSj$?>;P13p>?h=0xkHGryYi
zc!U2vPAgrkSbCZkaA9@PF3i;-hDtYBh8luCjw?Jq&+e{QjHr?&+ue8Qj^jb(m|cWK
zu&g}8oqe_lq}!r2ykbd>;C2x0%<kMw^s!>3PKBBrQ87VtzvY;OPcCsA%~Ev-rEYp?
zv>N*3DuXJc$zOp>@#tO`Ojt||NCS!b%V(j4s-?WdyZhBnRrkX0wqeQ|o0CcD>NNf_
zhu~tC=TQ@rEs=a%uI-i<!=-U?bk9lYyB|?PWEM3Rv}$R2L*(R*5sGBQdfHnIh0gmt
zXSbz;7Ld+^^%}Igu<^N=g-d}L_wROnKv0%ri&=&TIoe&7TNMGNRN%t=EIE_uY4u3b
zj)&8%PmfB}kw$3pSK459`+8Dw)CcJrKT};Oo!yEq;;-6fg?PWKJ*KT%A3$B&>O(EV
z$((o7V-<!4Ji4I)RP4x4E3aN$^%>V=)GNL}RE%csO$`QrTBdGs_+HlHIrb6}-Siqs
z$UE;X+NrHY2XP$jHx;gokix5Q2PY7VYd`e0E}V(_fagC5nJTz?3BZ0lLFBh!sf?>d
z7t5U8+%x44tA~xL6Vp;Cw~ZI2WqT{6H4FCEdm{I;n(h6faO}UY?y6Tsb!y$exgD29
z<T`ERybF~DD)Gz;<=Y1UTpS{1?A#PCQRD}e|CY(I5gH$*pl}*pi0l{9z}dB0mK^%}
zH#PQJ)!C%tJvLqJl51V;`vd-NJ1A7#ohh#*Zpnuz^ytJwJ((?T@?h7DSFdatj)6OS
z@*H@%ciFPp-O#j|`bzV|Snh_WyU<U|j2*mG{NLx!^Y1ok8aQZLpPAVnMeSMj=CARf
z#_Z6^oq(cRv@?^Y+j`m?2}Ab!GyQ`Lx1%Qc`d-1p4+g|ce_xxnsdHiYpZn!hQ)1+y
zZWcMLZufo8{>pOy1<<&CqDbNvo7HEkq6tb4qm35b^^i0*B&3n<stnyCWvY~=xzU!X
zr;Ei<f*GVNq_;=l(kYWCgVFtox!g<4gggAlxLUQiIzf;-FtTOkA8Yjwdb*WuZ=!)1
zi{2}}p1nnjT01vBclYjH`f$!i^?3k{5FdM-wLICv&fsE=kU}0cpA&$q&dK!a@{39z
z+x0|C6Xa{(fe`48tD;wR_mV1Z<s4ZguZ_5Na_(&Y{9(WH=Qs&~zVu3B!L#Bcq(eHJ
z4}GPIuh%FLnVvgz)SJ|EAG+Rc($*5ws4G$A;r%BI2sP0*UC7<Hf_*XZB>D)buAc?w
zyCs8Mg~D;ra#Lzg)Tq-QQdz3{jfUS4X;A6Az9FUIW})Mj&>CsxjKyr9GM1hED%zl;
ziA=-Thyonl2G0T~L8Czix`tJS;I}0)NpG8uzO549O90EGRiPPl-oHa9Pb~l66VkC<
zEJDxQ#G~tODf%w<x3h4Q98D#`Q@CAzLEnqD$49G$gA$tV!mTGK%&hA*CMBeASOgSv
zX0)0_Z`ZAroRwD%wztS3FY3L<&(2pr3bGgC5Za#{gsN*7v@$RbK(=4hirPMU8=7{d
znDF{rFxC6zJ|^l(^p=^eYN@>bwfBvd;W%vVzO1k5I>ugkeIJ+Yx>`h_C`(AXKh?7D
zj3l`=0Po7i7CA`W8d;j>T_z{#I#$s@CJAV#wr*SY)RIIU&BL0VZl5aK>=-ECn0;5}
zlg7#uqq}Yup}+&PT=c$P*W8xj$X+-0#$4H(q|vz8S;{faK00l8{x1w3H^4i4;1VM0
z!`X`;Am+Fq^)r<HzYts3Vkn4<&)15sSj>Lm?;DIEVlsZzgUL9$_9X{s==nDkDPorf
zp|+kyMq?BcAG99#C#-C{mMPoyX%NY_C2?hH`ofWpzFwhZZO>TAy_?<V8`VFreDAyT
z(e?GR0Xy}adwim?XUGeuUsOm`jB@CbvXrM9q>okuR=mW7RLNZ^zKqJ4#vh%HAN7}E
z4OfQc5|b+8z~+};Tp2Nu?2$wIj_)5gmpbuC59N{lv4+v=nAYi#{&oTzY<bZ2pmLyI
zmmR3pCLE{-N2JlKN*kEF^rnlS$a)@VBImv55Jk+q#5;YOsEC=WE(3#=)9B2xrL+~Z
zi%*sQt+UEInwE{|AcmU5B<d)lj<kYTX)6189VTA8@Mfi4gTiZXj8tAb(2RpYoVVzu
zDK$;ESbY5!IAO!%WdR)<mkoclcC_@s7?9Tj)`LBDb9Y#Y1){YsDa$qdtg>7$no`)d
zzem}L%dHn1A*9k9bG~8~K%5i07<!hSm?;<xfQ&y^wKphJt}}dNgvM8&&H3XZ?oVp}
zF^UK+(Z)rjvTIU;bL$b_6SB4it+Lw3HXV|RIKW=Y_&TP-;M=bVCs3ASY=ro9Yp(pT
zn{uf+Gj$mzGxgRCXQm|aBSsnlbwn>`3ez3qBCY{G^(;F}trfuzbHrL;aQY}8v;iCS
zX{ZR;g*Xu~?W7>8Oq14^30QW0{x9MCFLrR^vZ8-V^ksW$zZ{i|&o*G|gz6k_8VV{i
zozK(4R?V(id*Uf08QV~0Y>r85B7i9Bhz?Or@AEbPVK<%V?r5#D5@ydO+q$Xjyy<I6
zcg=!(B>?H&o(@9)nWf9}M~|vIT1~pC;g>@oRCF`wdH{)D)cb~_tTM_z)G{eLV&a8{
zJCqMS?e%~Drib{x<fq#%>TN%isWVlSED*#f`ukN6yezQ^dX1)je_E{4DT`$+z&){~
z^^YaD(&m?j>~W7Zq{1z;u!MFITL$e#?S26ds7>E@C_6_sZJ@Gg8NEG6s{wc|&VLiE
z@fLzt3qD9do-*8yillgV2PK~7lc<Z6h^5U&*RcwI5P4OSlmF+zenJZQ@2CU#ztCuK
zz;`WL;Gzh+cRz%VOSO_BtBN^NMaeU)LX~J`Q)4>a&;ZBPp3w)lMR888_YzE0>t>rS
zOK)W4tuGU`zDxf->>(1=z=)gkx%jo;{HQ*YuXQMohNu#*yzg{XVkUc|p_n>bm}`#J
za$1HHNHt!rocnaeQraJed$Il1Xf=tysN1Oo>*CR=qAtp%D0P(?0QCIT=wBgN@*c-{
z+0UuOP5-3)yFXof2i(wzr(F>KoE+$_c>U8<F?{wXS$d&|4muj7i1F4aP6ZgbjBt%I
z!iE`}BAU(qGEP*u-lr`LU(6HORijINkV})YMKhnINWMTeq!jEPRUicS9-3nApsfX6
zfam_*r)rZDWx)j#;C6CDmHoTRgs$u@)gYj~W5{vD=vbqrODkUHqpP9^DXPNWLy@E^
zuSu#Cf#J)1vyOKiyG9PomX<g{fS{ZGltv-xR%<{ey1q3s%Ty%3BN7i*G{s}l88hiY
zxz@n?tvblu>3nzKK&;$kYGR`OpcvKIZn~QqLNb8dn^T!{L%E8}yt`;C`wGesCN~^b
z79DKqPTkuU%KpCX(1lw`b3hF~#lO3@iJ0K*yneSK-%YSh2gy>bxnpCVWeMWB^`uET
zdne_EH-z6o`_o!KW=^9Vuao)IyzBsFKl&E>cmdNPY4(+sM_n#BlsRkuFNCIz_9h)R
zC0Iy&SV<qnN=7f0ywRp48553Z9cAlY|AD2pb1y0_d2AA1Z@C+Ieso7gYxp!m;lWKZ
zaXyJ`<@}gw2l?5;loeWnlfvdDmL0O5ZM%E9QK!Loh{?1~nly=vkiIpGtu%lsO_2RL
zitOi3qkSd&0oK!B`Seru4jpN_`tCy^5{*1qDY{7cZdPN=zGbHJy5UV@MNxXUQQnYP
z&9}@r*l&=|fA7OPG1xL+&W8YHIESR4vD6T_qCWgzNJn$Uf3uGv5CwEC_4!t9sS*D8
zjjW9kbhiq<EWt}8Py@TXBz%TOd&!a)A{8)M*jN6zT#6%Ml+Q??45Y}>kIm$+iy_Hz
zwaw2(Owefi2a%=RR|ibM_ZIWld#)$&Lle&dfg)+teXPo77-{}B7w#qu8CO)!DE2I|
z%N24&L{^{QKMo$tM^x5pYqS{gVB#4IJ0s1gJ1Ltwt4*Gr{(`&6|H~Z!v4=eK!uGB#
zJ<jw7h=x+ZQnxA4X#wyV7ec0?GMkv`vr)D5qdN4&uZNLVOYhRy>+qxyprB6&HjM?B
zg1>>h_gZPJL3l=UR0=B8EoxD8telZ18fEX}(5~GH0tv$8*(D!i;1~3%WL9$IKrB3p
zVoYC4Cw*&zs6E+xBc8msHPP!?X8XOgRv)tUIw47>NiNq07vmV#J+z`+X%Mj=3B3ls
zaoJFeQ1ao<a2~TLd~j{3pE(`)@!ULTllgbFD&xsGE_WZ2Dac8NM&}CcCFn^Xc(wTu
zoQu-QjzdkEanEYY90Gt|#Mr;3S%;;S-#e<B%BZzC_;nAt)3WeHboCJ<M{dUB${C9?
zc>iRqq3|`n$+8Ffn#Z^HSDECpeuXqXulU%-*2Z;A?;Pi=nVD_?+v#iD!@L|&bxW-_
z2y(r)>eG!knEpFG@S9H6<_NKS(5|8Imt#H#!evPrYII>WfzL@!c}%e5>Bgtu+LkZ-
zMKa0U@EvlcSN82aFIgR-^fsQ!iF0+*A8~<F^l-X;=%8mRbtbyb{x&XoLC`iB8IyXu
zKr)$FpDgFKRQzhN5C{!34lmRk&vzSNHr1uHU|=V!eqF!2IbhM8!?EOQ209x%wts~C
z^Kg38qTn7TWqa>?{c6zV&*Zy1z4?ZfZRxCUeVvidnXU;dz1^{F(O+BtymVV`W{B$E
z4^J><rqKm4)^h9A=Gx(4Ag04xS{n?t2*{hpv<M_UT06KM0=1OKv5xXM9@e-+VJ825
zuI|jMLABe>nmVTsSA=wG;Pj#PV{`qD*X-niw(1TwEKJKrmkDFM-hrj0KhFl<hN@Dp
zkKB1Pz()IgE-o-_I2|XJ*Ew3Px)oG5(*70U%(|Z)MR@t34+M9g8!h#%<=*Ls<_?a@
zN6I#LzZ<U2nO;jJgv{S1uxw`daE>3Z{p)#d?m@7aXa<p06mMa!)m9goPjY&{em6ML
zl4E}hzuRRf@peMza^AnQxh3O|)pgu^z3kmjFV24PhT^wpyBl#zcvH+h%u=gm6{GLb
zZ1z|;-aB(yIc0KHf^#q3MrwiD&;*+Gf41amhYG(W{>{>M8NaECzrV-ft~ZU8{Wk%?
zO|h)w;qEe&(t>UWo9V~kV4t6Uf7<Cj8!XEtYCpXLILM24*pRARcI<CB<}LmocW?ez
zbN<JFCnZ9evV^G7N+~;u$TD`JLG~p3)`ChQg%GKbecxrvR*_O!vhT`ThZ3VrmWtGM
zf1cCK=X>3*>-q<-`Dt!5r*q!#*YaE*&&TueY&UsZyQypCLBX!pP0{-Ezi;gG1IBp2
z!v=t6Yfc|0y&s8)>G+W=wHqw0{O!Z@rU;asb!3Sl=X_-Hg(me>zjA_<xk382Xb_y$
zcb5Qe>|I%H#4CC#eL_cjoUEatG!eBvOTUp*Ti}R<xae7sD%?|wAxO-a99TbU<HjdM
z7je{aX9qXwU{bZhwJaGq6nB681nQga7ro%|^WmqRup@8Qnjdv+udeQ;yv2cQzubKW
z&#dq9#oUz(EQAoT{@~mG&+qd3Fbfa-K3Yr_hevT-2P>GDH>Wl4C3b_e+@io(qoc<6
z8H0QFv=J+)-GN>PsV-eR#qiL(<Py?9;#x4QG&Yp>5(ou1OJe53-S_B@NfbKlH*<D@
zii(ewj$`*G`t`zV%T+BGg-|HSeaD9!8++j&R1SQoQ)^tvsazyyzVp2ChDN{0qEF(~
z;&$WNC(Uno9ct;c2fHV1JrY`Ds^Ac?xsI(aEahV|upa8it3$PHhus%7f$L|3>peZt
zwpz9Lz-}$ACnj`ReFesf9cNs?Dcb5pPjlc(T0a;;@IhXKk>-$Hg*i;*_4qz9^DtBj
z+N2+)^jF~=NwPbtvNF4Uve;cbMX}#y<F4z+pC#Ry@cS+%QYb<xn<YA4Ip*iI=`s-z
z83xVP@uYf!4?zyQJ<gnZ5vgDz>_hft&+BQnx4Tbc(Nig(MvQy#Z1kj)wF6Uxf65VC
zJ5}q?qZUVpY?NK~OZmwkA|mBMaGR5#16ivsw>Ye;kdOp!D0Hpe;wvm+t&QsUbkwfi
z7dAA`$6kzZa<(+sP~<h~)~selB7ZAXiJf*|DWR=DS^Xw0ICG{Kwxb}IreP7bVKKY0
z!fNL86E7fe9MlSx%Hae5{=O{CB?{PhKG%V?Q$9&M{sxz@3l2t!{Y3HI$K|B+0%^Yz
z?(eTT>&d~u!0|nIZ#7XsLaEiG<OqK(CXSNs4{CcEp}|_ZHUYaNB$aU}mw$x!)`^!a
zhH9fg@;}_D1|~5OGhbJ&r-R5$?ye2cd*JGd=dh|KqKTr2hLai@6?HK+MWl(gB+$;X
z=)3+<IyQFNKws!ReBPrsfWMc|=SjD!>*KMTeCizD7V{FNw*$2U)vFAd0T<)tYn$3N
z(=@Z!Or!rvkpMc><TZ#iz}KzcnR$761l@@+@Zx8-<j`0R-bLN%1Qn|jNw}nj6?J8R
zFj-6Yl4blwJQHWHz#hN<E$>D9Zu`*7?>G4ryy+R-@zN9_&!Bo5;Rfa!@#_Ggpkl&L
zUra^X`Y3PB$wh#M-^V`sO)AqWJ7`>+ad~KoJbd5R?=q5<D?S@`Uwqz%d#Jqg<Axns
zrKypvUVQBf%eKn?<DSa$>F$-5_V)fjHZkfX-xHTF9J-@Z>qU7Zd2Wee6Q)YTkISx-
z*WMe#P=yW&@;DE`rKokC>Y%n?KHPY{F;UwPRz7%_kkzoIbLr*HL7yHF5T$5+(%APM
zh;D!SS~zFl6*FBO`W5bmU`2po^V9w<K$2p+Picl`;A%qi2ETNA{ri9#q8{$?wV>$y
z6^e}x<)wLhdw)IM{$plFR3CzyrC5k&4QZ3`1`GEJrJJIESaUTs`){;c^)s(e2eW1N
z1?jfVekpq5Bg_lhkEZeG32;!6sAE4R7CUff?R7H5?XQk2oJXbx-H<YVXj{AP0hi`u
z$re-_willfvLV>U#&MV9z3q*}6y13ZS{XeuI_lL${G<djZ`;AjkEnYXV(tg$TsAl5
z$`4Y9T<17yl`)Z`Jl?z-Tp3z;R-``S_lc7l6~Dw740dTzD;>L1XHRZp5p?49;@m0t
zKwAs#mRho^={(DKExWFO15l|ww&T__8qSffO-c##kh%TSpbnVDQUObBCrwP49}CC5
zC^;UE+b%X+=06pVY8SM<qpdiCEhm9BFS5WfAC|x=T;yWy)o=#vE`rl-P8S3NJ>fwJ
z_qDQ)Vtf)LkB{{ky|8v?{Mf7&E#`2#BWm6GGsIADaA)dM-QmZuXqO1h{L<%A&1%@^
z$opO*c`q3{HDoE!<@rr)%Y>+FkTH)PS(^^h-uS$5UtGG*?~E3J$DaE;Iy*H|)1>oe
z<G>kMtubGfHmmViwic+tV75Za277w1keMh}e!S)f=X}mV^MW|_C)z3Yk~*GS%Cfew
zx2JLDlET<~dTe6eur{6747Q;FI(D5^H!6Jhd-UB>DrL%&KWj?M(5Hm=SpBA-pSzwq
zh}mIM-nd~k6+2o7CpG#Ixp&$<<%OPVW@2&xXCLzy<>@y*KaK{C7Y-dDeDK31_jlTr
zr3Vls+&G?3o8R(12f5QR#mj!Z%3MJ0<PQ$s>R)%CUBxmti!WcHKQzt%!HCI!Lp334
zfhsdE{nVS0qbGEvp$t1eRBLlMii<g&eo47Vdp@(~4i^phz^#b2QCs^+jBtk2R*CoJ
z?CFm$SiZllkOo5!y(DW{5xs=mO*_GzpiQ?h<#Q8Mn`8#FNB9^x`vH;DzrN&~OumM#
z(YKiO?3xc;7_D6+>3D*uu_@(McVOX|KlBnl5zC&a9q8t+MJ7@^ZDShXpn?BlFIOQm
zphOn)^fknDmkZLe;pFrnsT@v3%7#|<VA0=6LjHhZFwOUpCtQaH6)z$Cq|G_RVWd#{
z{P>ze)XC`T<v7~ziWMg4oN(gJe!r?;zoVlrN7Eug*hmT+2DG@w{B4hs^0=__h0f}A
zd#p~y^(qAYj4HV+k=z-$AKv#maugwQ9vP{h2{%$k-F!=L9s7EO_WTkG_}6xA)*WWB
zsGgd5@i8=Qo*-g~sSp)6&zxN0IVd8EXjX6JdrQbYi3b^)-nAww?+MdSHLSQ5N!002
z3Y3ygFq07Xcz(;_6xY4xZGq4{PLa--9tK{H%$qNf$F&-_0BuhEUhX!%YT2$I+A^_8
z!C8=}4~dC@r!qSpZAFv6)}dYT53M-p8BHvmMObCz(D)cG6{eU?(Q#h0Mfqgf#P$ZO
zsnrbS>`;@2ZL#9?Nm^g(<;;3~(1W&%4~t%4T7tG+$DQGep0Ee6FYjy8dERT-VN~#<
zU7APKtFu`;U2fds|6k}aEH6*8yJ~(TT<A?Jjjz5OJ=Q7Mmg+U?9EAn#ZhMTNlUwnN
z@6vYuN?jr)efOf5Ggrp9Dw*+QD+lbOi3p8PKP4--GNDb0U4r7pz{!#B1M1P->+*0m
z%(xxxrVjM=OX77n9&3MprOr-(TNqk8Q4q%yDeyA$9#5%OP5V~Wey@~CQ2r;mv&6`D
z3|{f79D{$Ch{RI*pKjb1Oisoq);>{2??<ilZiE@?wlAo^y4$~p^!K<Q;x6vfr!#fM
z*03!>+S#koO9><?tA?C@$(5zeNHc$%HQsgCuX-zrPPoG|fMX9^+|JI9AXJ3X^K9ba
zh3~hVpH6(2(NvO}U)9Z$s~$2D>4>8IuJN5SvK=%J5QWI(@60*vq1U*?fN@M`B8>A_
zsNOP2V8W7b8R%p-NIXaY{Zo#UFy40O;g-K{dlLZ9l7)XyMAJAio9iIEaXa=Iu_?xL
zk<YEBM3Vg5z;Cnt5DS>{rqQ<#>~YZj^_!${?KYK)v|fonp2&8T6VZ3=Z9#R<L|1L?
zJFRSKMxAHxMWe$qjFsOVp<s`i#v7MC?S}qe+LwMSZ<Xg05tJThK*X;o4K?FDFp_oG
zP9f8onp9aSL*|9wD>{VrR{o{k8NsNUXYSnE?(wkQ<;_tj%WV3XuD4)zLDPKhhrRtl
zOLXk$o{8BYAw}t1tDT!^2Aq+(fWlocc{LG{Q8;uc+d}^fb6$9H;Owm{iQcZS#p@0)
zdn#@79=jTi_i&u-;6zxHIvOk{GuBj1ZL&<bHdHJi5eOdU4g88}ZQWz;1$f<AgAy^9
zV3hWWrdJT6GV89DK%-I6(r$E{(5tb{DF$BA^{!X=6VHT40(jcn93%f?(-S`h>ba>d
zseTXR`P8-!T1EeTj0vmdyEfC?=^0l9Q3R)mE0MS+(ow%b>ExhfXLwCTBMIBSwMOV@
zFo*z|WvY(-iNY|n3kr#!nnOS%RtrAr1y(TD_SkO<g&c~-Z#RavNV($)+tA5yzL1-)
z&+?@q>q56-UdoNrdrfvwJg)}qChI+egrn%8K<gv#Oolx8u|+(h=F;894!V;@@C;rT
z{?t_Kxx9FH{7=XVvl_ak40ulL(cj4_^U$41bVFq1P~lGB-E+-ZS^ohZB2u@Ygf7I*
z%g##IMQRT1TPe*NigsAab>XFM4KaR=Sg;~mI$oA&pq}@({b(=@pL@Lq!({jQ{@3BH
zfG5O5wk6V7Cuo|#y4h{H9u~<W)f`cfm=iUwILWl!aK+WOHx19Fx+$LYs}5LYeK*&R
z-i?&eqN!+9js8MA@9zZakIYxd0aE-?Ob&pjaP3t8vPXI8mnQm_GsR@v)jGw4D+*2_
zgDA_BB-U9mh}=6i)lD+UxD07t$qE@mb1$G=YU7*K!PCVUWOw&|wz&3W!A4%U_)gl>
zu=pk6*o_Y+1o`W)RFG9d6J3DX?&ZbCdOeq`Heoz{RH<(!Z1r|vf)LJtqvLnSecH59
zz@p>l&?KVyT;$(Y*?xL@Y@?34nz6Ol{%Kjn4^OPGt=QuhJ{va8%6JdB6(kp7&60mY
z<IYK}V2^HGY#;`G<??q^<=kydlk&u;0$N`zD+j@AAjZ>A%%(GO_FFfplEe+>OKN}J
zn@_Jwmw;!ql6<GrXR-(t_Bq2_J+*FW-MV!?$yYuc1DfptFh{zU2|HAw7*5wsO2VKv
z_iFX%u8@Yc-=oR9+)hhl!D6V*%C%iHCpkznqCn@}*K*r3PN~Tj89eP5N?l{NqMZ!$
zTyA8}`6otH3`=UW1HZn>O3ACbKL>f=P6H;xuzU?j<Mw6=LJIYL2lyr@g%5bX92auk
zCZ7Cw%ISUu>$bxS7+P)ZmS*6SLtFS0m$1Na+F&V{+upi-^Bd?$ZtG5@#2(vHR{W%d
z)Vj?DvioA|nxj%nc4~-J_Pw+1rN{N1o>E(auOL^p-Dj($wfZPok;}@+M*B{)TItuM
z7$l1YGJNHk20!X|I!IUQ1w^&4#i^^F!p7JsJ5YwQ>-hLEzoK@+L53G82JzabJp!9~
z(|5NrbbaIYhp%oIWw|0VA%wTBp}G(>V))%V+Z&;p%6Vp|F?nW-FFJ`d_<SGOP}!3z
zgfhD2H0=#UP&+HlfJ{VbDNpAT7kfoz36aKzXAK8O)lN`oyN-V02$rQ@+(#!(x<7~p
zufLs_b#-~!7HRdFqWj049`RX{_BNtaHmahd*1l30pv|n7U9F~NFa<@rKE;!*{yr*2
zff-TF2;DXi26*H4_DoEXs)Gc?rim+P($kt?Snx$X({gLS6?d@1r?19-L@mVH<QM02
zef`zl$Jq%lwee^;Au4-at7*8FB&7vk=;X|qw3L)NBG<&0yfnlN!CgI@F<Ct^0;A%}
z(JxrvmMD{$;38uRcN{z;5EtF6A3yR!f}@a>+)oNHP3;;ti#uAj>A~H*(lwpea*p4|
zy`<*R9xIar#vD<*eqU)zBKkhh4-U7lERHZ)CrIA8NQLvdpcx5QoxBh<ym6cO#+STt
zv6d=Gn~3GvZnjc78#`+$==HNX<AGc|iE3)9NqNugW>Fh<4LhHjAAA)~TB`Kv4g23u
zEg_dxG+#!#Q`}M%jXd<eu-;{4*Lf#AN(bDQ#;eLHG@G>u*!ozYB<cgJNfAPKmX*;H
zHzVrlMFD7<Edf1m=;B4btcZC;WE{}kLdFr$Y;yggt^?Rivxc=ZKV<(YOctd<c6S^m
zGh-WlaCwALmHI)qZqnx*FiU58XBOL}a>FC~d0iAZ=+BKvy)N--lj)}rq#1bb;6>sM
zttCr9ap<MVakk#N<8OKsJ9Q$<`3cTdc%q1ElhDpx6m-f?rEv{%p&;H#e2d2|i3rz<
z=%bICD!j2sCb4e${1fr%$NOvFfTSrDo>Yf&Ce2;Ec(L_KV}n7-J$yzLV{AhYb$UvW
zrFR}8k@Ol7gy{A=NEUN{*~o46Wzh16MJ0d$i$+Z&(xqatw(dHhSFq>&6GdV@a5x1Q
zy?KuLP%M*vFIzdu2rg*>1RkY@2x<yey}YC{Nvx3bs{6bI6_5rl-owV_MG}AJow+MK
zHj<gw?F5)5_&bihuSpYM1=%Fbnrng8s&j5$#d*F@2)!aUUMOV*j4&>egU^q6jI6kK
zZf#;yN;mVem%+Uvh>!t-ML(B0#F}0~g)4r;w6Hgy>i_)~HUX8A>}C_<wJfh&(?uQ|
zDNqysM5p4Zeb1OPD))+?m`h>ZdIh2WB2)7uh+L@~dRS)vaLhs#WX*(yM6n;<gpWPr
z>NqF!IO}h3zofoepqFK2^>KSe_WQu0gBh)zl22r>wK`RE8n$8%F`|*22(prP-D(i=
zMLtmUU4F;rvUO}WCevteK+2lOP&*)mf1pS7DY1Gp9cE#?+{K^^>SbxabNPl-vjg2G
zD;-X$T<6R4rxOwGf%X9^C$ti7x2B_)uR%E(q%Dto3bPQRZz!?z>vXJiSJo}V#DIg<
z<V8FM+@1dPnn=F`&=x3q99EYDSLvwL>?tq?k-Z_k?#Y=Y*cIb-f^Ff&2np@{irv>o
zMN<Dp5aUv6hvi#c!!@cOzIt48Q`iJDjGeH|-fAP=6TsKwA^ns&h5rI!`|kzni<UtD
zsCj(=^n4cS4R@fajCAz1i<x7F(Np4tT`uKD<vH|ph=ipL#koa7G^a=Ri6Eo?`-d0I
z*^fz*rZsY8z}2x}Bw(*aa|o{gO5RjP=a6iEzjX$G1Q>KzuiwbNrkZIeXL}-=P?WKz
zclrAiNG<|6xPdb6ZQShL#Dd^uxvx8fyq?jc=69(9vPXHqA&-kGZ%xk5&SoG=IqR{j
z^Zd8s6n5>J&})7NX89DIxAb*hue+(9{LC)lxzac%x+NsrF6*40K#iV6XB1ZaE6(b}
z#*THUqFxkIPDao~px_QQm9e8#A@u-S$INj&4c5{`!_p4_&OgRuq~%{>vX0kpJ6C-Z
z^#%FQ6jXH`<#e5|_21v0jRW*_i}oY5d~B=?8s(F>_Df9vEFq~p$Cj@hUiGyt`!o+K
zA_|(Nzqbz_jDG?TwgJ-57%{P`sKJj}R{dCg8hJTA+b?fImZM^Hu$tLD;wJtv4Fz%-
zOaleoHQUuKHC11(7MMUJru@fG_1)M$Vp!!r*@RBYe@O!=Q2t9eVTJNvjBQoF$57<|
zUwl!|R@?qR>&k(=v$8_dPKHMz6Y^&538FuWx*}*I#oTFH)s!`)gX>6IN#$ikS=obr
zShByno0jI@ABbYnSD24d*4+5-w=HI;g4X;+tZ$+p4uD(CF{CVAtC3EH=)M9Rqdgr%
z$x#^9lc>CIq*6L-M4(rij=HKOIg%#z>AwGPXx8Hv8u>n>JV*b#LE;(y6Jgb^9~zeP
z*<NL#?e7NhImh@mKm>dHUnGNdC%vM{E^Yk?wS(Q*Ol1+;OeSPe{OtDW|Ghpu9&f5F
z@r`&)zD5Wl<76-Yoyx<t$UQwTGjT$y`X(bzgtFE=u!+oZzVrwbdcwWMw9*TKH7-)v
zNB_5vG1hsI@2N+?Rb5&&284=>L?B>>OJH&8zi-+dV<q4Q(9BlyGl%g5h)B98*4<l<
zLr^`C)Bm&Fiy>BR%kKSq;<r;a9tm^>Fw>Dz@c;f_+O}x6GVJtjA@-YFW7wW0sO7Bq
zcbnWu`ntlbDm_jsN$6mgclL=50^I@XRA0Rn1)hG>8e<;To<5FVUBBatVWLsJ``<5Y
zB|uqq!~iF=?(%U|g9dD`VHx}Hw$%5=z*E_Tzdi5u$O&{V^Na{yPze3sH$(!Cg*9$(
zUhpyuVK!>Iv{W}~ym#k+FF%0Xu@RKaNh4L0kJdj0{!W_d9_W?w?-H6(zlPje-)AK8
zwWIsR$ACGl>kM0(;i#spY#uvkKoQR&z}G!RgzKqM{%`pk%ks}o*Kul#=WgmVSKvl>
zC6jIa|M3ki-%|hz8{9U^dW_yAsa!;Z2gDuMZqQ0pr+D7~ZPpF^{MyfVpCOad=Ys<=
z!^0G{oVEWhc{_<hit4{dZ6NgWaYBoJ=Awzbs$mVw(f_X7Q~Lf!j-P&twAi}cI6KAz
zuws5w{@s>UZr!{WN~_p$q1*7DZ~uMILvwht<m`7;uPCWphCYzay`o}X!hF2O6vK#c
zn)qCh)uO$8Qg;><?_P8CLQFX9_Rn1ci*KkPs_InFrSJscKLHNfqF3?6YySKwQTvN$
zw}~F5%JOAN+KzdbT8?+({7OUT60Ko6syqbN$KLu0#N(pL?4$u%XvwpYqw5+ul%wtN
z56x4O=TwReRaRt5`W`S?qKE5W>?C)Ao3iqAJTTwFNdp+zsh}bf8hw#^r&xGw1(2t7
zic1zpI5*RM?`jrKkjPeLq$`<FO~Bk31iloRQqMVhn+F`FTsP|%8z@sxUCJjRyGkOu
zgwm?tp!?f9T6LRy4E+>^MKi~h5#x?9vFTE@(L$L62nrb&De{mO7X2(5UKVu?9*%q?
zlLI#VpBI=z1Y3kWmvhC(=4d60J`S<@=QZA8H3jT;?umK^`L;AD_Se$<?N7r6o!{-(
zH2$Z$+OX!`sCg;u`a@rmnrpyqa-uw#=sF=6$jz<qHf6oK1&l;s;9co?6wLa?MPfWz
zA*Ze}83^(8qbyZ(tTZlhhCS1zddfz$dOpc<u2e^69@%vfRIA3<cdr|fry13M@}Xxy
z)lz9+dmD2l7#u{ZeNtT@+tb%M7`$v=#+g6&^;!a*eIiZ!eH%@vA6EJ6`{!7+7}aeB
zu5WXC&4iISie_((8i<-tk-xC2FaKPAWbih<p<DQS1<jC$j6x?oJWeTi^gI)!?*A-U
zTtm0H;VW~QzZ~qS*L(t!HXGLO2YnGegR#w52TR0GHa!bHvA@YDykC8l)nur6UJ|?*
z(B79WvxT?O*Qv4PsbNNN#$!ni1=sBsEL?4c?K1Gq)b!@CXUmKpZQH#00=3;0wtk##
zpHJ1Dwzk<vxisbuH{`O1qStSiJW(6T*xAB^z~<&?sdeuf*L&D2YVR?rQgdS_asPQS
z^u2Ykw;%Vb`eh|LMMdFsgh{^-ou+ZHMJl9OR6Jqcl(~YkS@e>VcAHYPb9Upo1jm^+
z&%z;25Y46>8LE{$E4%-Ll+2UoSxcbY>jl$vHN5wvymr5SO8dh;A!fzxkdjcBf8_9c
z+nyeM`YkP@#VzfC5;Oc;PfT!Wr1sxq90ND#(Rtqi?oHXXyU<}~7f>(S8$eUmi&a0-
z&`ZE5%kW0U!!LB<`Wt>9%oMTlw{YJ(9(2$4+!NmwrbOD^j||ECi4MuPDPNd-l`mC>
zZGWSfHCGmD>@xrm@A*S(n|C2NwLTNjX3Fg)&<tCtW#hk;1(pLKOc(IU@wz5bb1t@h
zO+DV_{T^gTry3|vIsM;uJOmN(`oc15%wuBo?Y)O^mqW}X|0eTXIY_dTveC`Q>kVGb
zYBfxJ(h|yCXXYY&y!Ngb07L8oP1JoMi22+x3EMb4c)(|=_4}9MfVI7E0TFH-S3yB*
zL-T5jHdmdczZ^AozuFCbuTI6bs3_4V-qle$hz(~~l5%VsDZAxY@7f_nTnxZ0XUqTf
zg}#UXlsZPiXKHf&+Ptaxb~X)r9O#8)3SU>90MiqmNnAK!+edUPV5SyQs4?88tr><O
z5An6LZo@5dP6Mm!)Qt6L)A<DKuhPw3hW;f&=2N!VPPlZw&6e!90S#aXA6@#thwW-L
z>cml4Sq~4wwZ654Mx(MpMGvrK1!n>t+A6}BLNs#Q3gyOv|8wV8GEa8<uua;6E}qIq
z1K{40?puBGMg=g;W&GAuj|?+Kze3Fe`x)v>)llY<q_(aE^U0reK^V~m7bFXrd|O(u
zV`87&>1kMH8Xn(T(#5Nd9%2p=zXrCc6ExaJ(A}sFJ9~Bn)o54iJ5z<`Qd|mqz3&_C
ziP**?mJ2J$`E*CSQ2cwXZA31eP-5NSFCutpE9X>^N=_i}0lldJRrxky0P|#hDSC>Q
zpQyHECU!5mE;<w4(gk;ZHfu+RnUVf}0FBztf)~a3z&Kq0f2@`b2g13)<}rlLdj4Oa
zGf$cqbyLSyIl5~}{S{xP>T&eKU&2qRx2>7fNcx`misxDTG@d+k$gTwEm#()Pf4<FS
z_Z&UD=3$GNDPF;iXD|Qs%yBvA??Y4}b4YE}-npvG_{KX@Z(ZO_@lK#t|1aW8-VA@K
zFcX%C&ll0f{JD=Q*T-6~9$Laycx1@1helUkCg?{|tDBi`+7TX_$LJdEUr#uxpv$~g
zx(65^Ri<2Lr4mc)#)J2+Gc<M}yp-bNnmMe9kHNQpo>`TPUaDb&FR*C#6;mrKt9D*e
zl2`p)J{LFauC<U|_K)`G#`HP<8-V!yqR6#|QB7~&I!f}tdR+)oZXMdU_YPTUF=TuF
zK`t_6;CzpIM)oz8e3(9><M4XkViMVu(?(bfd;dtUd5e1u=CZx-Qg-~KyjjaRWZ#ZE
z8;2Pnpm4QC^SJHNB#6Xw6q-8tM&90avHlnLr&J>Qtm;Ej%f&Wm+3>AH9DR2_7w@PI
zGyeC-ewGv(eH%`^N(yV}?@8lY{LFZ{dv6DMIdxxa2lRNS)l)4s<B>v?iI{mE&|{2s
z$9_85b*QwD3br^vgY8Hls^=So9d~*+b-COn%_reRryP5pJlKz+RKyB*f1gHY>9EU(
z`#s~8D7jRg(P=gK_6UHXzP}kO>El<4%e*&kDSwO0sa;aYETvJSl&{;8V@<W3-zR_%
zJya|etnqVkue_{uuBJ>7@>c0T@KS!xvww^gl7Stlq7(3?Ejfg%kt%zX)><+eRUsPm
zP3z49rq^7b7!Tjd>ztk-BMCXMVjC1>x*F4mf9!<P^FGGtetlbSqIE~0sjefuss42I
zWk>J+BG$Y);ThW%JI;{_d?K(p{#^CG|JI267ksTL<La{tX=0Zd;`wpXeK}vjk}=NJ
z^mP)2#+I+ddp_M*5;2z`tX0(Z>36sOajc9Qx|1SmJRkw_Yk`d)QI@p(wOqDPC{ibk
z$Y3UJ+Bh5>=Osod;G*uk9i?h{;Z>$~e_%Nakd}W8*ydW8g+zFZ7A<<45)m?9e9)I&
zm1EW^FGwzucIyUP1s%QMGOIZ~5(hstYyM_iJCS>PGjLGL(=l3K4;0-xifYD)KdfGJ
z7wHQ=ckc8leqCX4vs(R?|NUS$cChBf)FVUXoFs?(4GL@8*(P#AG1tXZ>4;mEVcJY0
zJnD6)8hpW}RUu8T2aBml+plcuO?NssQpj-stNL*atnS&iE-nh%pIM*K?14m6dm6t0
z-$?Yz8JM}R=c6i=JzF~s`AC5K-}Dh=V;p|d`R?QL0n`vGMz0g&#QAdyP5X<FY2}CM
zH}PHz{C3u5)<iUoIsOc&f>60B{Ypz4Xc)^lPheYd(Sp277u@T`L1Uc2p=Lcj8k0Ii
z*!v<g)5OS<mTBd!l{rt&0F55cuhLf^QB%3P(<yjyCo^diaXVH9w721D0AT#gBGYHk
z%~;njL%M-c&P(`m!=*b#uOgWslP-ENY1OmbWG>)(-TDz#OWYV&9f62D^e9TQcY8}v
z_zXa_#gnzD=AU<&RY3T5@B=H*hg(Umfjs-M)fNY8eBCYvXVfj26Oj=SfBec6Jc=;m
zrNTUg4EI#+!2L?BQ#1RRLNeZ*P0nst;>j-_(Ea&`@(N@i4U(RA-i;8x?`gcp1px|1
zUj`uOnhfCQ`cGS8XmJ3wN3(z$g_Nuoxg7Q>e2d^f1?_bo)%RLfd9$;3Zr$3(G%><8
zx9Q6#qU}grGiZhK3M?WQ>|<_Ssq0_Te7^Ip9!nFveUetyY0)A!&1`t&?X3Yv{))fb
zVf%na0~=o&;bz`);DGDjT5nrBaNViMUJ*@Nwv0FJ8{wHWDeR7?&94&s4SQ}*>HfUX
zaKY;bZ*$Aar!?Q`{iD;w*-Os;APoLNs2%ournp8!r(fK)=w^rL9^74YxSPl4eI|>q
zQ0&{0rC=aZ+T^9h9DuZ#T&`k@xmwuD_BCNKXk`x<?HF*bIv1yKZy}!Ua?-&UulAGZ
zBy4m;!Y+p!xr2uc=}xSa`d|#>5-67*ZFWfHGUwc;P;2yQ*KneNX5A7n-AO;cDj}9!
z^vi>Ws2SFtUQbOcTPN+n-^Bf#<~>`t`nsA#wHo;kZMyeiDfb8=__&t9Ir`w6?3MSx
z4yKmj1JQOH*zAxD(=swLnsb|-|F*h$B2g)J)5urGuzxL%#BW8z=(_2%M?<^azds~o
zH5<~#tc!hxE5HI2Ka*3enao)321}<#ESa%*ol_iDJo?I3>^TS728?HO*MX9XDo&O?
z?5*4EQL2PL-F^U~-+zBV99h<I(53SY0;eeIsAmPfuIBd?b(2wqB06&t`N8HlUo~uY
zQ_iw>5LB{uH(zZ$JIT}Vj=l~pt>tsptXX66(WCREF$67<`g7P7!ZJdV9v5t9Zuofr
z=!;(?ehkV&l{4?`ov@?Ri!6}aeb`m=zfF3yuxAAU4fXSnS3iiHAQ^s1wC*j+RG36a
za2q~0?|0KntU|Rfg2ehJpM`&rY8WNi3kJyksjeWcEnBu!Ovib<R10|I)56eK_p!Wm
zWaT145beyyB6#YzEKf##-W2epQ1|ZJTESW(1O2?EhxJZJHN$Xw;|=?(7t`ay9=7_W
zyLQ6{Bj|+9jXJ`Rt6pgVHs|jp2guSGux@R`)=eDRRA79(Dl-5KMq~Q2rk~aAh(Ku6
z`{|1c-%fzDSmZ3VdCh$a=tdRATatuIbLpMNWd?~!VQOR9=vS8(&Mv)wwBLeeju-Q`
zL?=%{6FIy@Yt`u^f+wDI&T{Dztk%WYr)TrlTG>=qDTlpV{zOa3YB0*+2vpy2+aa)r
zot(A$h5{ZW-Tp|zU3xj9zf{n*5*KETeOcYsgFIYB-HMz(l-7x(3R9R7)BEkwA4Hd{
zVn5-H!#*km>|U;AZfa`yDCceR5exUiX@2punr5|8E{}c#(n`YErLM1Xe*Y6Zhp4_w
z(a}LR@!1EPN!sEe149+PZ}LM^lls?6v?qP2KB*90#N~e$;YoOyUB5RILtB)_a05G*
zn_|>|VRA35UnIG@@42NE$@+aOg12bqJLj#DC^=bJSQz<|2z8%orFJu|vBuSEcEc4&
z<##=k&(}wk^pi+9uFF*TH5U0|qttWIcI^6b;S=#qIPV{$MLDeK!Sp19H5)eAjy(Q!
zc_yrtP6lPuW`oXz`Skwd5aoYwYoEUaAP5f0_1u!SbogtcwCcKc__X%z+Rf3-GO&|f
zu&CtyC%U=0{kRy7+BB<3k8N@B(Wg(JVse{hd_)%L?)w8_ca*pCHUZ6+B@YdjramT=
zIhW)>nZ3GgPgc>pTlUpvqad$|{&`X8E@J-pQC&M(Bkr6qb;G;+M`b)q^9pGo`scy)
z!fDB7<-k%E1T%9rvofk4<uYCVsZHBDI8@MukmjIXh;lhqR?c6;MC+WuJlp4#3VqYy
z!#I6VKF>W0GEE&;<zZ)jG}6*B-1_@7B$w=pvg(DAt&q4b9Zo3bk^VD7RJ4%Qa7geJ
zpriSxjZo1M4Rq6LJ`o!Ux@`0zWHi~%?|z1*3TiCSw_a#p^;FuAodL>*to-y%$jXB^
zyr>MPSalRvBb2WG{c9$Q+E>D)J|5$hcPWr+VN-{7{K8x=*tunbki72knvkkU7dot;
z5K{g+hBXx-YsT&tLh=M&*>dGNl%dF9QwL#pL`0P`d^y2{FAy|+Jr&Up%86yPbd%sF
z(ZJ*Vu*Yw*PU{f|`xkw>#p=cm4}5=Jd7R2#V^x}uuiXaBATKo!6K@e0n|pd-|NgA6
z?a&J4(hd7DeCqh|<HLe)kw=0nbos^ga)(=hV>wPPU-V|vrcEL`^!AE<$vHr_d-i+x
zi@5txdoMEJJ9jo_(u<bddoAM;u56f^boB)xAyvQ*2qWc<ZmVdNg!w3rSax-I1|lq+
zE3ANslC!czqH`(ciij(_D*~xS{?;<H0`oSXlMBp%NZ}8m8({CN$7)ye=z0HEEO=dM
zX=(L6dv{&A5_Ed{{Le%Jt$&4v^e}($b{u8)c;|&mn=0)fwQb^zOWL}FQ320xdLPP8
z2{5Spvl}H&h7o3vG#byWi1kpU<gMQk9zG1&T|Via)aN7*NUTAGFpkJOukv2SLaJD2
zlXm}p!#GR#q#I@T<}Xzi&~k1;*WAM_)V%2q<)NGMj>v7hlA@{83IT5NyD~sJB*C6Z
zcnC~D1Ho)^`;q95<W(9TI(TsBqPu$f`Vfrv{N8g}<gd|Oy<JiA8KmY{5-P%=?Uh6c
z6`wV_80TvE;SqG_+@`^`Rf!L0uREy+H@f*Zqtki66#|c}R2s~|>>}#*aN_Pbth&2o
z>-HtxM1-q)QUTh#`*vrO{okutPJjInr?ZTLj3=mmae4?3a?WR6*4IDIF7u~QzaD9}
zqCJtAU*bQ(T_}Li>mcU-5V@=bVwwMx{oRsEq`<8lWUk#h;8wyDzohpa2v&5YDgE}_
z!6KnZyrA>FJ^15;w_z_{b<fa>jEp>6xp_Q8l>^n(cD@{KBSwf32<s}RLrephzFYBA
zNZwgjzf-{O^07Kkxg_iNaKhh>5oahu$)Fy>cd45aRVAEkl)*bPIUH4}#RS@&h5n;R
zInsB{wQJW%NEaZ`FI}6^3C>pTm^NEoC(TZqV|l+Ky^U>;ZCA#?jo5vqX_wHtYCV^B
zVj@D;{r2Fgc8NQdbl@7ls@~|?s>1AAYL&T_y9Sg)%VXO6BkUK)+URCP#qx0-7%MEb
zT#*MRTk$5YV}!ZDvV2PZ$B+3^TmKybkuV--4@fd_=un5m^p@oc-DR{*r=bt)?Cd;u
zHtvA3tQvKYDaq&EQ&!Yc%!lkS0ba6J;xkn7PpYLwZQm|Z%v8YMi0VVZ4&U72WC2y;
z*wmIpRR{DqfR!M9_$j#wL?K8{8~U{9S5k?D*+{TgIa~t1@V|7McZ=gpNfe1GyNxLK
zg39^t-@k7P`;g+1gxuWuAJtS3*HffWl50l+Oo~TmQI_1z0pS*{`>b)qYR|OyUZoEc
zaNd@UQs4{B%*>iSD!l^T7%p+ai*vd3xP8S@BSXB5t9Ilc+q1-VL~ZnqYmw2GWB)|y
zTRyhBqN}H;CrIYqrFrZ!BMoL}65CLoXnO)(!T6gdb_L^+HrL;$ro9(iE0eP@`<9&2
zx3)GjFln?DpivI7;coTV()+nMw(nR-(Ud_ga5Nh%m+)rNoPh&x^WQ4~Z&+HN9}ShS
zThAZ3qB2@P5_$02n!`K@S40SlqRHye&^vehi(v$KbFjh?v88SmUzpWwy=3*!O0KY`
z04R6f-?S0{l$zczOn80|t_S7xX<%x#LzZ2595bcgW<pUx3`sNRJw5v9w9eVcjT>bc
zk!ZA2Rix~>I^P<GG4IfrA{Lbibo0MNdjhU4ra*IVcXxX$t2;wAyGTde{+C4hx?gAI
zkiWh;J2iBmTH86;G@*+VPAy&7X(FbLb^vb-1u28g){}&PvTRk7ctpS<4;C`1LKu9M
z{?2ZwrXwA4AD>ZC4Ne?86M&yM?OV0g19&9SBRGwQ^okWLQU{+RR$BOGVfuQ28Enx0
zmHCuzDA-7g@v{eR^XOEg%fU&J+poHOrCQiaLKjfEOGSe=(@m@cT@?Wl!@!L8@Z+iR
zU4c*$(s=O}W_xfzC|-V_?>I4l5t3}sB-U1U*=LS;SDA;?$)T0Hhx9Dhw`5CyJy=%J
zo)!qqRs6#C%!lDc2^@sskvce@+bTM1*!r@NtdB$~GvLXEY$)H6Sw4*z>h0fklnber
zwP_sw3X?5&;^?+T%={sN_<_;u*W!=(uuxp|7n(_!KvKMK`;mhOMb!&J&&g3r7QxN{
zT63;5Xv+WSgYxyEp(5%m9)T-0YherbsZ;TUt^M`ulPL^4qDR;n2{Q5O<a*^Fcd6AP
zA`MjW5fOgxAt!lS2e{$5mA=s%r?CI>CYo@ZkG2*DFJ|`TG5UD<XJ|O|ekzafstSiS
zYQ`{?2CUn{*!|8^VQ31n%{h1MJdAt@B>Up@2hI1XesJR1vu7cC$>5FPu*2;8VtkIs
zz4pq{!euW8LM0ROeYhUzEuNz9;lqcurD<L5YE0VQy8%dgV(%6nef3^?_jtlyIi0DY
zNZE$<TK2H+XzGSEPv^b7FvwE*i(x5r@305=g2VP=YT9_d>tpn2N2N@~<G9*?{L9+;
z@Ll4jABCBx^^||%C(a$E^N)L{<U#|x0)5ya&-juYh|vgL>~knvOx&!eq;+`svdanZ
zzTUrk=e1!waI$DFt=XSWVDUr_^1Rom$4ZE=`h55+jv}$!4!o0p!fTeD^2((B8AT+P
zi#|S7?9Kh70$tBjE-}4YC?7^=b)Ob6%|&m;d&4Hwg^@{}-@m@U%<~kQZAn9?0RyCt
zgJ6plY+Hym)nXw@ArbIlw<)@9i=hiUOzV76nm>fgR&a6|NX-(~nbe1W>JMfWKSQh^
zF=9m6C(j?FWpEt^bl~6h7^xwCnLO^+lI9A!8J<w-HB+)!EQitRJrO(py7znD$YxUF
z_&q_>_$GpPAE?1$zxVMF9G~%~dJ#$rkMRfSjxo$C(E6PgOxW?+Ib5c5C`$xO4#3Ic
zEIh`pV6;{V%@7#)&XQOAB|3Z+>tu9R=<?jNdodf)%+#d?)XHj*k$gPRVB_}fi}4Y;
zOa1+TFt=Eb*{?epXP+qJ9|~JjP5p9vvlSUA|Ez9|M?OVOJ#3~2WoJ-|{Dvilw{eV6
z8#YggMb7&gdi2*+A6~@K5*^=Oy?TY&c|Zfh-yYQs*%B3n8h%o6K@Y0bk5tb*kZ~>N
zK_lWbdkfGW+WT#@Z%v|5!g~)+SYP#-e~4ZZvFrs?9(YFZ%;<6CWbAtP?j7ZWtoNIh
zzq!9J>_y^X?>hb)H*N&SX`4DR>^eoeaoh*P%cHm>boE$cf!JvWX_q6aYjbm5Ebm15
zrC!LCSh&nG=jW!*C?Wtle?};;Y+%4_kUW_w)b>sO+bh2~oYxdx1dQh>dX3L62G4JW
zzk_U**{4MW$Xka$|NOJt)7InT&gc&LpaBtmDBtI`^VB8!C?VsRR6RosAP<RJ?$(^=
zM~A=1oHhv2pH2`ZGRk4iPV+RFm4?a{$l4?|ODbI6gFR1Q`sX+--_X=G2-Et(kG^^R
z`bwXq%<>PNi5nWJK3{(<BtiyYG-&>o{F?b1rh02urw(XO=!GF)Rgwz+J!E&cF^%Q9
z1Dyq{<QcU$%%v%(jJjRR)LXGcLg0tFUAA>}yH<0md{AxZSz4%pK3aA=bCRh;eAOn!
z-&EPh5hi%Q9u|#DB_@5R{OnG{-@i&8v^^pd<7E5m92vz9QF`#s+bQONvcPra&UM*`
zX`9D2Gt_^Hx5$5Ud9-2i8W>AmZceunO?&+7FqAFT)Uq^grKQEOie5V%J~9II!4UH#
zfo`2gj7YJU)^ti1Cm_yY^gLsiRaQ`7Fz*&QWbohs$i$oeT(%v;CJ{%fKb-c3=_kh>
zsru}Bo;&k*fsbOcL>c(r19i$?@N!GP?#NtG@`~6&qNUsN#cGC)$N(=%E<7cJ!SjQe
z`rEiTiC8zD{6TRg78gt3^nrHD58s@@!eCyP`l^j%{3zVP9=GeYrbM-b;?jx|1z8R5
zt-op)4qPNU!*j3dcbfYOgK&4dK3gCqGal7A!D5Tm=y_bcs)tyK%SKPHUAG0@88Cn#
z+KD=kX5d(Hy0r@*z{I67X{G8J`kA(FeUcPW;rW@dU-A6W)YTK^9Z)<-@&C>txiIZs
z6V?6}%XlIpXdT_LIFq=5GOwwt7vY=d5%~LN67v1v`8Fwct#l+iO%{La%L^@`d$qf^
zic{H{0nG^^d&p{EO!`~3XGQbkbE1koF7#QWCzWXHYmsF1>4?cD>|wjQb?i7&_4>WV
z7^hBo#-iTc-F;j`w+Ea)Zb0+5uQ>c0cke!>+^wGTu<AOxf3xRUQk&W_voLtDr*sOx
zc7C&9qqTC~?E=X52$Yg7JN`<Co)Nh+;|JzslkFsj0KLhKw?C`=n1*Qb$bHB?3Mv|D
zU0*}GF&HQi)V+s|O#vu<tm^w-fFjD2_o7e1G6Aolb{OWmvx9>}MrG{=tvaA6JVw4S
zP<<X2NcLclL+O(}74_Hj*|e{xv~i_G?<Kq@bUvs2!k)_;tmz{Xmd$rEs(mHq-+7h#
zM`%;Jr&sdU7={TA<qP+^(pMXjV)BGsW<I!jL+!&r6P0CqDr=oyotJs3x$?GK%HYC=
zi6#D=4`ltd##p9}aCXMc8#t_K!{MoVwy9(m9IU!G4ze<&@tZW|AQDJw3^H+WSV|zP
zS{2P}_zns4i<+Kf)RcK`LlJ^BH#e^lePGUuv%rjz%D}?1CXm=4h+-2W4>8UI4XJrL
z0=r7H*(BxD8($QLt>V!PLLkC=v)b8!DIrAz<AGf}C=Zca23fL92d#r2YkkT0i`y9E
zSXi;okB_tp&{L2bZC3u2GL)r&q3hHKy-|zWhPPoGrcQqizCn+@w(12lya_u)oQ&w9
zch>KcAOJ+dEK#rCyxFX_tY*DIj5KZBs8ORjy&$7yZ=C$6E%9PQ7Rf!W-Vpgjph%I1
zL7mhvd5E|SjU;^<riB8xVFEPnShQ?86P-iVD#a+Eztm1z1A}hQbB*G8GzQnIRV!Kh
zI4Kk^!d>N*-Z1Gi<%lk8tG}j?`*IUepegEl2~Aej0_=aucH`Tdt?Pbn*%j{Ln5_QH
zh$clgbs0!vMF<<CIx*ia0a2vP--!xqg1N^6NKj*;Ub%K8ryV1xyK0^O;;?J)5h>D8
za}R6ss>TzYS(QEJ@^(F+4j3}zB>8Tee}tw?>g>LJ${%a`NNnpmhAdDREUKNm^qMYP
zMdv{qka^edvWO=fGIXfUs8Q@vKnFYdvU+_*6@#?GHLdJbX2mG|E)%7baD~%@?&K=l
z<s)jt8PXk1kv7@)97Q)mqK(!b35ei|{+5dy4reSP<*L`KC}9BHicI6*Or)8ielr~`
zD<wLQ>hCip=Qk~vx}{8Tc0DtO3g@?P3&v!{{h~3x3A=~1tZMhlCFM-+-=4cGOcQgD
z77Ru_#-zt&ui1|euGyfnNW~{f`{g*B+@o|hQNN3);8R6YYo1R0d0e$n6$>E4X{O#K
zQzIE0%SS*_G38CIFL5~X@KCB{O4?@UJ*_nVA)YQ>W9WHd{3H}Wq&iVcUO)Ua@c=Qq
z)|9&#`h$lL-&VIs7WB{h3dfhJ0>hbnM%Y$qV83Wi>H3F6g^x9chA|WP+Rd9<(WeQr
zRO?vQ2htE`F1Xq>)DC!*smnKhAe^n!(6v{u6lmysRljG{jg~`2?RsdPR3DP%4zXjd
z7_*M0Ktp|%zeOOCv}_svA?Q)g)GZSZ*V-Atv+GV;bq2qnT14m1g2hb=luJKL-69jL
z4Y7viC$3@sXDMUYoxKR}P)e`6DvEgbxyI1Vpnsa7ZQHi}LZ<UrxnW&K%4Et!H4(nu
ze)ZT{ic5qXnxAD1L)9nD$ZPL2_tbT~G``MOKQfB`@YF4TlEe2>J#4l=tu`r*j050m
z@Ff>GN_Yp2Xn`YCD`PRnR1bh_Xq%&wuXkKtu56)dIL*tenDj|?$A%u}6RT`{rd6CY
z4>w}&$UY8Z@=F0TIx27J(8E|KO_SRM>9y6%`f<L3a>l?xWMyWnUf2+qJ_SIw&3=g_
zIxH`r0+yy2w1U!fv~uY^yhORfU?|xGrI;(*dnS5;)M<uFWE&)#tY(o+xDOoGu#N?I
z@{wBEoe?#dSlVs{z|3IP>-Ld)Qk$s3g9kIe^n!VKCQ0Zs+Jjt<ldnHjEvv*zl%*T4
zGpqb|q{mG4vL`U?npr>NBC;bhRgc?6@chgsQ?W_gdsdt$tl+$9MSrJ|QMLA-(p6l?
zR0L;hW=|)v^*`e(zIIKYsx@1|Pu8~?-^HLmlOTHOr#)b*Xrsg#GazVJt$d1fDGzsN
zrFrQN%L61iNIrXoIglEvRb9cKXdOu7wYAmDGRYK!rpoY_r%(S@9klne@mGph`vwz-
z>ZeuWV#Bv?H8}AabasqtK}s)C{tx;y8wKBph=_f)`mBsTb*Yt00{%o#wH#qBBKl)T
z_zaj-bge$22-#Y>Bw#s7FRPYpqJSqwt=B;f-p?lhD20^hk_2e>MD;Bbg){G3?gE+7
z(fHB=@LQJ@Af$R#|5v%*M#5a?{|ZSjz~lf>mg+JXUPcMhJfQXg3VE~CADMOUeg(OK
z<W?0>ua2qCMxHdjxEA?RZmO;R=%e&~o(U)57~v=fBXG)SqA}N^jM({;9F3bu3YXk#
zX#f5h34Z64*63=C%USmUr49cQ@J?g@Bf54l!%a@)NdKp5Gj6&;f=f;l@@qu@KKiRi
z;6rtyz9P!sQkg?~SVQn_w$9N1`#)wi`<w>->h5x%{=7j>PU)(b>`IhJhqKOKVIW<J
z=Jl?SEE(b|Pkf|o@V9C^K4MAD3_6*cnH@TOc$WI`lN|2k{md{6G?o{+`Y4`OsNl&{
zh!8ZE&mcl7(G2|=RNJO%9lHvJXVrSl)fwY@`9kh=9d2Q23X^Jr>ihc({CGZGGB5qr
zivX`!Ql01D>8(GcTF8UK0B3IB=2n}xZ97)$S|Bd0#-PjCN!1<ecN`oyd20+Uo<Tk5
z0RCG99hv|{q*`YaF33sOyQSo7+UhZ)SgdHe3F$(WWi&3NNvO&0wt$|K^jFsfVD7uN
zeyvd@;MS@uQ7na~nl;w(pUFZVYPj8CP_>?QQ4)e2s-O2O*|4nK#GuVYV2hxh*1}y0
z(yS2YsA|zE6fu)J1~fXE^v}C>`}TIgZ|E8#m4=Z8MemM2LCcoMNW5s#6AluHxsV0-
zxjiO5Y-U|Vtom1rKUCKz0f0ktwi@fG)kK-cxPxG#k93N>Y^I+F7`I5O|0*d=ZyiN%
zmQBKD$?sKuq7lTHCBB#97Uho};?^7ALT9YLzP^)vFNkLaoQ~G6)^@O<gY!j@OIvAF
zvagfg2R(AbcWA7s%k9Foewx|Os%=|O_<!wgtm}dMPre~&iDor94#n@NX4L&FG=04H
zdgHdPpqbzyb{~%|Cr{$`oWfPpbW%z~*h?TcwCT%-^(g0RhBj`xo2_#_Cy{a0!h({G
z);Mroxzv3wW6p2fyqTMLcrS5D$feGYNF}vkFIRL~R~=9QLHR#t)<Nz|H|XB>EBmUR
zW<Pf$d4?yLeNdkZ<O?4X_lj6DMz7-8&3P0Yf&1*5k0Z#ZnZ9<@d~opwZ%;t^d8E1^
z30$dFxodHS&rKJT>73^9+u5`b8QHgcCKDSFUElH4OV%p9u}ZA%s|5sos>cs0p)GsG
z`~{N34EToD*^ZQrm4p^aKvMmFnv%6Y@T{ST9AxvTQ;iO!L>Y$}PuQsHXXTP74~wCc
z&U2r;-_@#HnFu{>DDZ=M^vBZOc-5IrtIs1CbKNnma;);CWfch)<$5$D-g~;^{+B=!
z^&2(vah<*B^Yi?pdvb+S=h;}aI&;=9;-_>weFUpCR#^&PQpLKf)>hZjsc-tSSr@Yf
zSS_PFH;rW4g57=Na{4Qub$Th-XvSWqYX19}Y^rAO>x{pvt(VW(Mo#)KUfS@^M#c-;
z1^hGyGicx5`PB^B=oOtmi_1`+5*;(WH9)METP>Mv+E4#173v<{W)W&!G;eEcVqyZ6
zl5v6+sxniXs5Gq5y3WQuTS4*L)XZ$9&pKY@j$YII@b+ELC1QfAR=t??LG%4J3dSc+
zSPzL4P%EIOXz5Q~Qx|N@^UN5kUo%y<p|>AJf*?;fOqAwiX-n|*R6NllTY7VD`i6F@
zSpQ-koUzubRkFl+)V{Gs1m%o1UlYUgM7!RruaFaZ``S8^MB|LJ;zCusxC!6Lqe$~-
zcu*b=?ifYfuqRrWV99N}w4K5oO1>csLu0u+SNENtpXp#=Arpb>9lQ!bXax6e=$#y*
z$A%VB!k10H(Rxy1!oQ!M@u-D?1x1d5;X&r@+kd4cqSmrY^D>Z`He1^kZ=rp0en3Ei
zu9Grf)XAaXQR~JUX=^c#X=P<K8|Z|Cy&LH#0l?r5aif?gcb1%8tGiD<BO_7xc;hDX
z2veFa=CG^Wj@Cl5AFXrOfVi~vq8*?bJt;oHNe~w@qwP$>ImO&7fCwss^%`FdAlPQm
z_%p7{&0|^>POk^cTVC|Kn+wI6;CtP%b*lY%O_}eH=FKYQw3&{|`CTUk=I2H7Lq010
z{@!njF;#}?2b0-C`rU;4{K#}vvquxwo#hhG0%^W<{UAS}(ex)EzrJ!K(kQC>B^$Rj
zGwT~}aFLj_bLBF0)QG?Hq?jda>GH$h&T#w=A2|};Xo`5vIqy*Js;RPpBq8Tr!$BhZ
zI%=O4(p=2V<}~FG3)@&zYRZH<-_{7hwC2S^v>R2H$IBmxHP}(c|7uQ_uyNdL!(kLH
zAkBp8w-=_L>-)DbdQ_)2{tD+b**J#&gP?Had`}YU+6`@So(9%!TRq{m));pYPf_(@
zlncxaOeHm{rvnKJY;`G*B4S_dV)pT|<Z1u2$6o$~u0f9P;MiK$Tuy%9&Bn$ipvFFs
zdc4GF^BUBny7LDZrK$&)31go+bu_%q+DE-_w?dH`a!0Q^QF5a>Pv~A6s`6^F{AaD;
zYq(<VUKv<5@12g+o2=Z|R1?DOnIW_`S}QM2_$%}y8rIRH+i;*`O$wfqI1QzBda<8;
z7}uq9aVAJe)sIth7}lnL-ot&t#@WNDLKtq5K`TR|W|S~6OQY#|h}~6-P#0pDf$1V9
z!q@YP6XNbpNAdCQYZl2=qce`nN-k&Cs@gyIv~bt((vTkIj`h4mTraGBO^TrVtRL9~
z{5nf;Ik4)T0AI_($LqZ3qTKSVA+f^QK9`H~?{r<1XIb?!^u4p1BvI!5&;P=2F-~<^
z!fv*#^-YVG(_b+P@&2PzfRM>Ge#=0Cs=wl_8r9JQ{%+a-{AFMOYWwu%c#YpZm-3I@
zP@|n3Dnhv28R=BLU3ZX}jI4SjL0;MG+0O@u&qxP0$s00;Xqq$J?AN)v25pRQHeMkG
zn`{FQ)EG5ho#){Ed!#6{BpZLhCFq;JWY+D^tIs9>m#E*`={rWv%)%nKORz`6sd-sX
zF9FU3&3(ztysDdshOj{$J?^|!|MSt9tuu~5Zy*bP3m(wGy)esb)fsHA?XS-*ZU*uI
zb{QwA+ip=Brk?X8$7$1_j@9kPkLWy_Y<HuAI`((Moekgogby&*29f?XDKQEvg5GDw
zkQ!N3t&}id9o?wuX{=ck(LbBo?1Q#t(tw3=qgy0u%XFbXGfdSRuni=j)%?4qZwi+-
zn)3o$POqv(!xA}kGo$<Sq3O%MEg}GBJ~(I&<HP|(&3l_r)@B10ztML#wcPkC23RsP
zcK)7=ZQo1Ly;A{IqJQ_ATrws29ZqkiUQ^fwe;66<`8B>L<r=CAB96$narb`+1(3k2
z;zw+faNj`ujx-w#ArSc6a;9~yWfO0JEt{y`ZyaE_@2G?YpGFLEyMb>V4N2fh*bvrP
zci9W|M@J=zFqcVvs{5dG-1p17Wa}@SpGKxpc}$$(dV}Nj^DHoSa7YBeo;Rm~Kgl5h
zP*C09Y^o|3H2j;mBzVj_7yz1Uy5psq(TW8*<Ew9hv5n0sQVL?hD+}7rqyVJ2cSV{n
z!~q2c%LtD)gA9>tz(urfJ(vD7`d((@r%GFYT+Th8L{LUA*i(oG*=r|Dg6w;FdZ9s;
zXLtrA=^9naU9CBy9&JK|pasORzo9kA^hyvIp$6-%v2?Bn+H3Dp3;@`m%2`rZZHw>A
zpTA6B!%d_eg3OkIm#b{B;cdcjgLV3qMIk+Es>kG0tf(6m0Q6n8PB}D4Y2R2woKRoG
zZ5-r+AhXOX{ytT1ep}lqf3+EEJ@WV_<9e@f+l{QtgdVLtEm_0p+gLg_4eQ@wxdGVd
z71t|C0GO&<6oP7KLlUCF_33pzQR~_nC7kh_RI=)^*OA=U(7!Ha$$G0t!al4!$@?48
z&J3mLG5m|_P4Jp`SPhNkE*+_#yT*|GurmMte<~xv``$5kqE*dn$MM7(q2PhrtA4zM
zb<^rC`sJi9$DQ^#B)cVn2dXk{o%saU#E4YmgL<C$xd(jMHAdH<O;6Qp6-l3TQAjlJ
zHA;W=00C}B{e;_9wP3acU{<57w^}jaaDx8o?j)mYLqlti`iYHE#b`@MVt<3;-$9Es
zt9flVk$O9r^lR2ga-rO^(vd*#&TYo#p^j@@&tLK%*KI)j{ZzNmb~hrouIFI^IXfEv
zke(Tr4^SOcDSTQeR@&5{HCZK5wUhqpwo)%Sa~I*67OI=l=-9Di0<A9aq<FTyr}cnA
zg9eeQJ};&Cr~+_`k7BR+{}=@FE-AR3xe7rB%1+h)|Hxdzx;54j!Z*}#yG2wKtoe4Y
zV|CS`NbqNUf~!rD%x1Yu@y#9;ZC$b$ARnr-VQ(o=emftoO!(Wr2c@htsss|Xsp{c;
zpuNgfT_;oC9&5Ozf+(ltYW<gMdGwK_@2){%+<P5I<4&U7>GBdP&AL7<R5zs`vvr%-
zy0H%KVgvqMy5FqxsXUaZcD~;%gs2>}tGPLq_ErukP(8EBW;&3;a9kgEnReG;i1ACg
zSPrV=&}9*UXC3qPq6JE#T|x6gjnGTLzE$5iU1H5MG2q({o(=X-z~u(_eFJ$(_00#)
zaj=W(Iw9bHv{oMxn1@-XVI?tC6<RV&x>`T1Cj!>j*q_dBHMh=<i;LsNRUPFEu$(kZ
zZjwROQh)RrXr*?PH=kCWx3KG&qTq-r0RiPeqaBx9%HeXg>%r_{)$`qKLDVD0?x`e}
z+k19^|1rFD$`Gp3vsJcf$^rxtYdGJOkVE~EkVBGtp_Ox7xnEafc69M{YvVh3;J}xO
zOD>z&TPRJ%u773%XjlEfJWBooElk@0*bky|2NRYO+lK1kMHS%U>4{U&I(L)A0>CDR
zAg^{*k$_N}H&-DM<5AtSN-m87N#wpf*4W=4CVH@000RJ2p@yL?0gltaf!gNbqBCe-
zueAt&gF=COOi{hY`=j)kCTsBHQ9(V=e+08T<4e_Tt3VrKgFOtd;8+#0ZK2nU2OtCY
z<$nW^;l--3r%3>@;KH>(gewu~;v_+!0}2X}X{kz>3Y3|k&KO*tWI@JzAVfKznPVH5
zrLNj<{mF#hiq7q^DHDMCV1Y#L+-Z2?1^YTt1<Pp66j!hw9&e~|_@4EKcI{lwKu@8|
zC`7efwc4SDUDQ)V@BluaLGE@<mQ&YIWzRjmUe}*W%?&8>8>HB{d<N(7cI`xOAbIdv
z4kpb|+j}}Z<-J^wN$E##jp`)r943QpOV{6fe{(~+M)3|Bb_5y!{B2KV`4QD`4|vG>
zxwd<O5U`_hQaoX4P~WleMgKz?H)gHx{I=$#Rfv|1A7WhDf5;H3I__tB0nW@*?ZZYB
zGA6m#A2w3Ye56slfd=9lciK}EBR^B!Qj2@|t>)`2H%3Ov@a9_CT2wztK<h;1c1$0C
zK($}(NS#mik7$v+Ng6~RY7_@U^jR^#9Wh%~fnsBx)c6i>;`B|V^kMEXi=YJUOwHQo
zfD@{IDigdBYC*tQt861Xhn#X4F3(jfX(pvq*HL2*_K%=0JWKt^KvTV$Frv<wo?Eu-
zcLP-!%-M|-p}12Ihkc;d&Q|!`MWyO1U>5aOZD;*U)NO*-Or@kN<?>oP^+W*pjKSw;
z(Yfl%-^qIVlM(=cP;h(>ocPle38btvDX`7t6b;pyITxT#%&MMQ`g<{zhTFGqYcErm
z9-gyFEtIu3kK3fD-N@dzP6TxrqTgkkc!eTO<p>w$ucTRA@+oXS*cL<GsP))`wRRc;
zLBij0JEKPIpQU;f<4OT#35|jpPs7J|T)y!=RM9#-%95W8RF4ucl@(C5{QzO$Xsw+y
z2yuXJXPlAVrz0x3E0a+22*)hh^c;l>kiD~4YspL(%}|>^|CG6uD#Ml*0t-Svu?}g{
z#D^L~cd|*qOj=e2IFG6-)E)vXN`dw7wkPX5!{os+SsL;1Yl3q|L6H5=0%nh+S%KBb
z8uG{^C_64Y7Zb07k<xORrfoF1BdjM1BL`|3P9hbsgDVug5_5S0Andlf+>UH-rOYqo
zg1eoBOUB%tF<D(Mk(;G9G6sgN!Yo_WL4cs!ThSx#u>Ad0j;7-KGe95J%E<J}#ufc@
zq>4pNrD2yvc<j?8PG{l<Zn6YY>Z&?1mhA{fU~jwL_XB66`c^=pfbA|Ki&7byKDz*&
zL;CH4F1DSzB1k$1RPV$wVRAz!QWeAgfjbs0%Y6PE9#Niu>hc3%Ud<EjT-@B=F=^Pn
zn!MZor$oWB&J3y!?bG7y)!`qJr1b6hJ^$5c<q<VLE~Y^dlS`z3td`+18goK=xX5T&
zFkw~ob-1wiC6Y1U1F{5m5v5s-wDJT|5IwURdghxXrGH*s7w@8-wB!;20x!Qsv-l$g
zDD6pU?H7C;rdp+~Xs;Q?dW0MZf1lb1@B+_TRW5E1wJlSunRXK+qe=qJTk0XP+*W1z
zpo`hSO{Ksws&{g`OSA+LENILOXzj#~t*KE{M#s(uap>t}j8GM52S`rpYxojj4P_B?
zFXPr5PqgLcs46q7N0RF(85n&lNh5Bl$8*zd5=Mx`A8tKOc{;A0;a^3wRq^N};j#Pj
zGlImm9^1ji1az9TFhW&Sy*rNim4d8%V)|ZR^x*8rWCgPSqy$z0QTdbyR8M7TEWmp6
z@ZIoWac7a5XBXyWv`rA@nSIJDG;A;PVokc4&t-|HO*?E|-$QbH-w8XRcq^Ax_SJG2
zz@LoQDLV5FGq+`BP#(VkQ2`4rQg+HyvR92I7I_Lt4T2~1O8zF~0t*YR*%*t-7_BN2
z!gC<H;q+>{&XUE4jeNV8<5!GuR7+>`fKK_#Dod74RURkpIOMHwT)(av{Dk75(5lZ4
z^O!?!$LZDC;}AzZMfJh7A9PQxx_JaIvf!7G_G-HHAuC!YR0``8b6!QQ@c^y~2{{Ve
zAd}&T*4HjV*c5DA(LIU!BPi4G8uGTr;Smv!NOFoSFF=xH)kdsxfz&Y%c(=%58SJoA
zlaeMjmQMy60%l2HZc)z*99i=6k*l^o#n=_Sok&HSt-GV#RgYoRCysN0d`~Rq9-`tu
zsj=H#^h@f!eIP88DL3jU@2i^56i7^Tu+`KZZNK1SIPpTAwk9|Gg;<!G38zcB&ViFI
zu;<d3*YAWt$R-i1+g%I9v8^h}=c?1ikf~be>_;TzBe7mZdx7rfFVtH;qXf`j7Sw;?
zkJISOnyG#@EKRhNa6Lr>Rsdus=dW8mk=o^)(r=CzLC91m)1yA4X|@wU?2cOm@C#eG
z@&5a;KAe74)@|IbkLuA5{9uBKK)1Q41TQVge|4nJaCd|Y!9U!W^;LanN(agX#HcHe
zoCf)rIcwJOcG=6g8`65|`O}qpSJi$sk|}?9h?Mpj6dZ**)42E?dqx7jvg8$kDS|<N
zsW#&MNAfX*{TBEd_|uAbQg*2L2%(eA#87naL~j$Ibro+ZZX5!BWXh=~FTe)9LGEsc
zXNnjqd#G5_5iSU|1l2R_G^P+ClSh2p2+fOQ(z16&3}qb7`pS~W^y&Z%IH{gNUvx1g
zg5EG44@+<Pp*m&B%V>(_maHmyK<Pll=BvI}L)e%f_lFMsRr&qX6AR(+wG<^jzOiWD
z7;X_<0ZS$K(Bm<@ik#A|FXvzyQ203AWv3p^Lbqdj13vhIJyw0AJNXKrl-;z8blva`
zhq~+>%u&3*#No5D!{t&JIkIX;ZnIK5H*KoF`!yHt11a~gRU4;JuO&^n=lOH*|BdK3
zR-{KD^!qMt+NhD#z*1{&(U;NQm0PaZZ-}=0ZB#jFW@aDDmK`s~TxmUJU0$8DO;=0~
zuhTBLMT*Pe$%EZ*b?tj}+le7J92WI{-e>j9w>8_Fojr8$@uGje`oDRb`@=pqy=3IB
zq0vLT7c7g*+4*^K;;HnnmnK$@PxqU!g*?<Zk1d%?0Pe25nLK{0)!Z@$x&K{kUqn<1
z1v1=l(@G}Q`1ttT3J1liQJAEBPe?I+iRs#TW%s|eP9{_Up$_#LcIn;w_w=o{Xgr6I
z?l+-ue<f#=Klytw2+fn{&l^7Y{LiOPwTwP|_z-EhX)PKiE-o(pqjb%OY~Qqd_fW5@
zZ{OSEazNQxT#9>8EPxGV1ye1v2((TZGJosg<MebeNc&q#JH}sJC}R6bp0eK@SPF~4
zRO`AokHlB@uknl`&VtX+TSqE)o6pYbW4`eowMvrQb~*Ocybc~TEe>w6{t(d0z(qIp
zm@C(%ZQF?Qss{<|DuqP8%}hV(`Sa%#-^}cW^<A)F0pwiM-l%=dAbK}dmIwHYr!UD~
zfYz+GReS#QDf6^j*_!lR!%!1I>!wYb46L!?%9Y<#j&A8zef{4$49^$`^UcgTxnaYG
zJJ$sU;Ac*9S|Y0*IC+vrff3sV|FyL0(|g<b{m2`Gz)Xi?Xq<&a_qTMjaoRwhrw$!5
zYiuFReGKsbzJ6eQd^}iNyH1^=2J196Fc6hiE344@BR%mN{L1D!x&W`EN2i}Z-=X;5
zCEEN7JrijpNw776ELlvmL-vtbi!mKkU5mrkf<lIbglsgbda9#cX)$)4J9n4Q4JshX
zB#9<HM)vE~D=0cTIxH;A-PE8-6G#qS?d^BeuiCMmBfl2BvG3eD7H)k`_clJbFAk7-
z-_7Ib(T1~k@9y1V8&d(O&COT+?~fY(@y8c3q3m6~J9X^%{l^c)A1(TB9^v6p#4c}&
z%{+PH1gU4gf0s*5ecRv<g@yAcPmb_$c=7ysCriuzmOZ0dtp~k(@#<A>{_G`7eq#v|
zP?G!iN7SwnjliIHcXbW-nHLXkiyT+&aqprpb6Ek3%;s0lnl%g1vJ*@C-ydn~HENWb
zoqh1wv28xP<EcS#3?AIR-D1$JT(WivQqr7oWLO}_wN<O8c0A7S>ojn-;$0pW&z<{P
zQi9~K>7ZE(q?oCMq{Ujn?e{9mavY(I`uQu%<D}gLFVoA0I*c1P&fniZ$?xkb>yd+r
z%pDyaxpDVmf6bXYmjYDEXVa>m35@JQ0|H0#>#o(EI&_c`?*NxWqFRLk$I<#}n&};*
zX+)DBpn!}0zt6TQimW=<zvk)j95x6C8(zO2ulK#$n*_fnw{8s>J-TzI!z@4}S6Abk
zlm9(6w^JbVz9140dj1FV74qJ{|5}_gNN?nBq8ohLG4~}$(9z>T+ihDn{@<H>`zgnj
zr;9#tU}Isxg^L$|AbFtEyj$8RyQ%rbId?dSo!Yi-IU_#{0Jld#sRsa7ywdCQsGD}+
zk)>uPG^8PuYD?>kRl9bP<2Rp*i|b-*dm5mm(I!7OnJy>5%q{=@tivYonX6Z?a*}7)
zKLI`sK)~6p_UZp(@4e%>?!&j?mJ*c;Ny&;bQ(9z`WK?ADtOl~lRx+X_dvCJ$9#JBz
zkP<41j3VNzLdbo5uItiu-M{<!<N4$H=YGAO{<%hczn}4bpYQWLj^jK|E<ZHo{90U`
z!>|o``Rqg$3`^)}YvX2Nxn;L{&0Kfj=`A2Ai0YeDQx2_IBamiTjkWLVN~=0}A)XLQ
z0mYQtwGH*?n>vq`2M@MylUt3jDhkNE*REYlWRr?SODf<^N=nL4jsu=JeJoLeCb<2K
zLK8}`0%G0kckdRF#T#$==fx!MVia|)0G+pY@7@vl&6h7<MlcybP7H1D*|*^2(MPt3
zwn@6^a_pk&iV81p@1SF=OW{;&ON$^{P0(+Kwz_PkNc2uMH#V9ema3ZWT|r)lHU2#2
zWImQstNzLDk&%)27)4I8u?^#>zjx@Lj}d2o5EJ8q3&n-K>+8Fgk%-p{2~2HmEm0S{
zx-81b$e?T82s}Q)QIwTAB=u!z7C*!yM;-<7+HU#PD{37_>H(V7w67@5$8^iqt;h!v
zZ{L6X_%S0R1Hj9Vs)6_K--9Az60{j6h9p5WgOKKb(&GB{Wng~j*Y}wXT75IBa;~nf
zii(}c?cw`=T;*X($}2n^d@HQ)z60Z_yYWrWTT2;^ojCpkkzV2z!Ev6D?CsgJC(Hc?
z<c%QIK2%tzSn1#U3xy>vb?U>TiBF$ifno=<9>9tPT~t;!1<zbaI{?!OiHX%=Yf4K?
zKXRHyN}hgiPb^Zoef!)1yxzRI=E8*w@I!*@F%;?I>gweglRa<Vl$Dp$dFoWz$xePK
z_~&t~NpyvA60+yEj*gC&mP|a?J?trIX#~;-w$c7`XA)8ne3QMsy)HN`G>D_Ud+^hz
z;%l!CczCRUNi{+A#Sy-hf&%bqOn&}1RP&D>Jt9&g<jd97)v${*K#&HaOE>@{GN}mG
zhzmk)ib4;AQ%I<JQ7nG{6A7eActbd7T%4VWs_xmdGg8;1Vq%cT5I2WIq_D8?`(ITP
zaT94Af`SWJd4L=UH+BRMB$QmE&bOkXG*Ni!>gXV=nnzwx>p0_CyBawXg&(;sp5W&v
z0bs^qMVg-8Eoftm1D1G+$B(xeq$T3^K_S!g8{f&!o&cMUFi*$G*zUag=GV}wUA{bu
zRvWx39JJ2BHVIG(OcE$a1QAB?F)1mZ;KL5s_8rkA8z+SR^I9B%fDe9wE)s%tC`E>9
z(<WbpZ}=0sqH;ra5_|N@{fLMjJOn98$#%R${6cSUFJLm#>ZV+Md;eGcbNWf+vuE)F
z*7aCwPoF-0`LYm83ux6Z=mQXcBawPph9SXqc5xxZ1ne%Wcc;RS^?@jkPLNj|8dyL8
zyd#5xSn}pk=^~O1z5}%ybY}I(xdIRpQA|WeM~~|AXCUlhi$8zyqLGOr0d3696L2ah
z!rCKs+?aeT&A<9SGd39WM+F83Vs$y*62EZa8x+U2a5qF&kvxp$h-Ln+$VA?W-ptHw
z@EwHac&pr8qD~~Q(ACo+ajO35Q>aAtJAHcm>eVb7k;qhK<>W4!+rCCnMpG!)_}jj|
zpo^<l`nn%P6k=dQ+a?H#<z!_!OAMWzf8jhLP2D&K1h0tVpC7&X3xO1jn2#e#Ut9ZK
zQ<KaO@|l^Ls;a7BTCc~C#clpV^6<kqqcQmDhMgTv$C9>&hL?}euH*O7WMxb?2*FP(
z9P9u7V;_o=7+l*-NnK&#3{Fu09bEkU7a5wKJ@fhg=XM@KN6W^yZNsKbbV9_qFB)qc
zZ!&*&fT9demEe)}$mN1J5Mh#)cqT1}SHHycAS<gRB2kSi(M*7C8Qc&H%W2lP#L>aO
z`cvcV$B%cvd81b2nwX!@L`|(+Pt4H>l_@SOJJ7QFLm7k*w``HeOAtx&j1EYENmmh-
z*#G^32mn(t(b4`9REX&q6{AFIM$-rBCgQ^Wjz1q00RTW&PkefMnj@?ky<`tI!APCO
zY4xZS55NLByG`-L;lr3^V4u0+hOI62It;<^68;l8^neELcUo(Jp2IZSrI#r(NbR;#
zQ62pEOX!d)wNTW9#5=qCTGpuc0UTm$+k`jP($ey$`nj29j%a~*h`#>^e^t}OrDj!t
z#g<dszO=k7`=3uh3^_)EhtIaC)is;QMM*^!!NatB_ilvokiW{-zdoP{qOh}S+rLgu
z{^tYU<L$9?lOxYkf8>1eVsJu20@iz7orj{L;@Pv!7%)kNAOz1hydT0(48}OXP&RMb
z@*V{*F8}P`1t-41ln`D8({9k$*YC~-ur#)Zi;yI5UM)(*c`=t$gTJE=@s3rsysQiv
z)l0nNB8SAZG-5Ww)YSR<)tgYH#EpeN|EZ6OvW`hWFX?vg=J0!%nOTV>>-lqqi(83X
z!k?RaV8m2u+`fWH%D!16<;+&=68w0G@!x-q@#$@-sED@3vd^w=Znr%>59!rJLdn<a
z4Z$dvrUW@9CAUMdk)ffSP1fon7eCjJnanuX(r9@(Ilo|iA<_Vu2F=!kL=ggtvqj8D
zjwme@uFLA|?w;%|`~-N^>#AIy{Q}8Ewh_;2SM><OR`_Wbq;vbS{0R>P5tlcJasvZm
zJUlp2a(7Fxva%vkXb@qfrKR1uvj~U>&O=d`p9)Uvp{BN4`R&GN@Xu{XmwK$Y@552<
zwv00`?C@)>F+yx{I^_8YtZ-*1r$-(uwf^a+;3S4U;4~XD=V-XRk`PSmOPDgj3Nz6R
z&ltI)UmKo9Z7*+vuv3vhXjm96)qNZOD03vgc$@9-Dv9=|ZQImDUc*uXt#@8kIBBS>
z>>)9Oz=EzR2erBc<tZpmiwg^Pw8CaQNaim>2&H%sbno2=Jipt1qSiEQa+;quKpjjw
z8-NduL7>XcHbeh24b$7+UYakyonaa%v+%5Ng$JV2<B(`#!PwZ^(z1KkQSaQDoSX~=
z{HM{;%ciDWnMJTG@yzRj?8v$WQFB-I6O;=<0<Y0`{^j&YV3^^gI5K4%q;T<}KWPyw
z5K!-dAgAg#`sE8}K4Ag{v~RRJwglBv4K0j#XsA9SA9wQlXUSzu0q=$KEVyGUmRM=8
zHNHDV;Bq-y8j%9*?6|dQ<HqrciMG1Z;$mg%Hzi!4I=mfs_C9``*!jp1OYS6wqFeLs
z+ZTsw+?8_ErcHO{_t4X4I{J8?R#6Exu6^6l5je%&jAcNx-W~A*+d$PdEG}+xboAj_
zBgA?Nx0!>yyf0AW6z*2)q~u%>twVeb_7DQP@n<1qVRNhVFp(2<<WQp^@I%A&IHUz*
z)$EJdku-Fec`n!zu@@<d<4&Ru&Uy{E2!i~pFvOcA42Z%*Ta2M<;aWIzCYbl$1|=mW
zJN;TLIOYo=9QyEVNxZWlt+@E%n8!1+7FLrPtzULyIPQ?3$E~`$x}|KSR!vP!v*iE8
zKgE}LbX&J>Z9Mt_`~J$2q4#wj`i1%D(d-g0>=bNSh=UQyQtMq!U@Q+$Wj{n#3VT0m
zI>K<WvC5Z<DO`aW1}DM;!PsX*hQxrA@oRM0GvPWI*cr?Cya!I{nA&+DwYjT{ghPvr
z_HJNcgyp^i+gs_rQ?vkuj3dMNqnZga0(yWl3K)X`-bdrGhmKBc@(>CvWN;(k9cMd)
z>;KtPDnXHvpNEIpq8_|{^@{r$eI`nxA!^+Vai`K!t+2pp*|B)#Wnhvh-$jWmO|5VW
zK}n@=qrL^=;ZQKpn6^1@Z&wlbB@Op&`Mo&MzSPzLdgA5b8J6+KW>9q+1hmRQ>cdIz
z>5uMX?cgP%i4>1?dxX+g?@~}ugG4-xgB|PqGEJ1iv;F_1bK-uGM59xwr?IiIx%q74
zx!06H9XX`Eh!5n$_R8V-Ix!(1NIb&#WcVu-6H0{^a&rEuo5H@{42_M2DUy`+AhJk5
zk9@rc(A3L*2<B)}I~x{1ZRK}`gc5=X(&$HYrR)F@6-4xF($N&73LRR{CoCc|oA;ai
z=^6KRMYXl=NS4^B<`GC8SGNnzFv@H&%Y}{?m4IOtvd9)<QUw71h8+~k=ZYMES>p`*
zes_4g7x;6@A63Yd+7@5gv5z3`u>Q`heB9Y-1vgJ5aSP#rckZAZ@HE^@aCX0b{o)$`
zjpETR;k3DVPF`N#gIJKOEn<Y*WEg`WuVDX2#9*g3U<h(b;^giCDn0WKseJ3sj#!E{
zk<Q<YB)#$9$B3Ho?hWoe!j99>r+{Q(ix(CKl&16)wgX9L^AB%f5KvTdhXRLI&<J!q
zrbb6$3YNdjUOFjMlh`K>x&@%x;D82eg(A;OmX`T7a`5!R4Y8MEBaGo)eEMVyYb6++
zeCsQ2YHQ1eFu;EuYB-YacdYa4*RPMbe9ZpwN%2e^w-|$_W38c)(MSR~4$XOEjLBOF
z*9UD?u^;1(9+cl(da-Unb2G{V?vkx|FLjVoAc3O2yA5miLbvPc^UuH-zgbBA`)l1?
zPM`KNqzuIYcsGI;gI4a+DBDz7SXo&mse<fG9lDSr7Mhxxikw&Q)G4l?x31xwrq`Xo
zr#6aikT?upLe?IIp`_W!=;%~$;nTNowWP>SojO&o3jeOm48~o%5F83{I%IuPR(u!-
zkochRX-NfOv5mecwP{2Fn4havMG!jJUN-F|M{y^xg?!5vl|pu8H5l(i(qd#}RGr-M
zPeP?=13?fz-oNEBPCVM<k|3pMy8{9OD0OclbY?(onA}P<5m^^Os0VE6GGeLq`C$kd
zHO|ms99i7FBw?KP;)UuXXVge+=fb`{%J-l9Ht?;#dJe`&JL@*E*d-4IBEO_~887Cc
zx7o*e!7k5LqUrHtW)>ES83RYhir^bka35f{c#e)yHZremTxWz3Zw}?-Gk+l8<07pF
zHxB9G1SXW7yE2`xUd^be;<EKaGt&?k)VEB+n_s?siNv%7MfA75e1qS<>7P+w0sYcI
zw+Bb=xIKFd&2AjIV|l8fF)@pf-U3d*8TT8sd_9xo16E#R6^)U;O0ZXeS|r7RSj>zO
z;?k=bZN~ggwB8;o%j!DJCr+p#dz^pY)pc<}@@6mA@rVoyImBdYUzk=C1QP=YP}v1H
z7yu6&8qfOLS$qB63F|OyDEAB(Y%mQBN(928(${j3pTCIb?Sr}PUf4$2di~wqs$CH8
zU*XRZQ~3)W7znrtovcQuY%8gUx`9$W*n2M(71e?|m_qeO>=?=OpeGB9Nls4A#BFss
zhH@^vzdW-|0_Hinu5C92k*6T&OPW<yRNP5t0_La{KgS)?g6^DeqoZoj1Z)k{)Y77$
zZA%t$qLNT_JFy5{VB^M(LsbAYvsM%ykrqD25kB$@g^fd-29bf0+ycCQw`xn<+-pRa
zA{S>TXy3tEgtyEOTV`&3L8nKiao|5?#!XUCa4_otvveLn2<a@w?c15-&whQC9>`S+
zv|)dqJ%C=DidSeQ18#L8R^TPh+(ti>bLU<FenfsakA00^9+TGvhD@H!AuTX2;m+f>
z8^JYZe>|Kg5C_11H_JSLAfBlM3#@T5G1?zU$gLy;@9sTtO!nd^LMadgaLr-<*S8yT
z-G6o%-4nmpdF~NeJ$zdGkbE>Ay@kO9QbAdB81SXz242@&&1r5bZhniJ1j?+w{!14x
zcA;bdh7)k_o=U7X@MhZ9O&c~qMiU`LcXpe^?i1#b{<K#x@<mfr(PSm6HFcds-tF7B
zjT!aaV+-^0rjXdrh7X`gp~4AUvGw4R>fq#LF8MqZrkRVYQSzfylo+s(1Jf`@&guyW
z3{+?6=DL9xBxxr5hQ|~ebk`hv$qQt!Po79%kv3M2V7GU~38^<@0E1fPh!_Sqkf^^C
zs!gp4n4Dl+?(E?|c5Dd?J-azEw;6#?H5GMM{Qyv6pqz>};nr9{sNNe16Qe7e_kug2
z1f%A3OjlQTgp!?hVKrV%KLO$-QA6e&v`qqw`?e!1;U=GDUblaML+{CxCrLDx2Tz@H
z1`_!5=g;Wq=qv>uKrh?&9Bc0A;5%?&^!BhAs@Sen2-#^=1R#}$DGCA-s1Q1JM6MSo
zbh(x_RPVN<)uCu<9O@IroULFx*#-iF0m?C3SRpr7eg9a_4?q~_y7RFDN|S%mG1}8M
zHbwK>y}<g=Wc~uK(@5=yj8-j84Hh&Fr2r)~w8}q?jHC_mbrDVLqoWRR=VQ%{4E4tD
zlYcupH6=Q>@&gbee0_3IH6r&3qQNWwiM~D5<4Rweva0Oz6b?|w@1J=&jO$)tB(TKG
z*2R-wzsl8iS))pr5hB~MW5@W9)sU4Tfax5E;y3FQ8T*af!;ct;d_(v0(h^1P50q~r
zA|lqggn_`D$>#DzzB^X^@Yp5<z9{x+(<veA&Qh{NO$wrQ=Eo0OlBV>ttUdM?4h}z&
zg_U1_=YZ@7sQs|^0T3BW2c+?`aiGb&UxWLm^mFK7GW%B-R^m{k(2YFD1BZ_tQ~6Pa
zFcAEF=!~>9trj6;3-%l6ieBIG;zcg-=h|0oNZbVht%KK)Zma9O+`>W*Of32V6V4_K
zlLJXG48%^<c2of!Z((~f#(OK}=70V2AD}@A(1fTe$M_Se;gGsNs=a)@Ke2!!4af12
z)OP<8`gT+r4b*-h{j$C$ilSA^IyZSLCba=YF2%1~Ri&k+)zynw81l6(u3uVR$m^lL
zdRCCb5EsK2rEoyAi~tj1yA24y+ImM_TJVDhiT{3&|0e`yLYnbYAu%N-#gWw-qPeWs
zzhL0oP+C({BRAmy{z3Ees2D=r)YO!LfdQ>?ncz1Ta<c2r%8i*uW1~MXU<j>7em$au
zm{;owWJY%O_Owrslw>;mxgmdKS-&}cvXFbCqVdm$(7CFU3kutz7uv^;ABUsNKvz{9
z79o;e^+wwb7#J&tlATKQ9gpCKB<#U0ZE~+F4JGAK|7Ho;iSJ$jz6iM`SEGDIRh55;
z)lkFcQkT`huEGFUv>hExdKFd@KdSor`fzd&jbm|hmAoM=U{oQ1(X6_`7SeJ)UKmie
zhLn_4_mvoLVazhgUAhu8%jkI+WT}tI$Y-5GbV_D}Y9C5DL>MPkMIg8FzIE%Cqgb-!
zjWb!1f9{YX8Rv_Cl=%Afw^{djGWSL3;rt6LcEd*G&MQTI%U@Lr4xR8L0V_K+AkTfp
z%uEQ*mcVdq+E!3Svt1|OyxBlQ?S6Y$TYEb(6oDpvZ%kAap&2+3686g*32x(@5JuhE
zd#v87tGKMJ?9k@k<lo@{5mScz8c>naXra5J3Fm`SV?)E~F#5KgP*>0X{0U1HZ=Y7s
zTq454tp;UI5Xc70*SRhP2WbkPQb)_x03^AM7Gmw~?RY_4cPTL)OG`uJj^!*?e%R=0
za8vvZ?>yFcF(B~Qg0c|V?+|nejn)XjhaoNip<Eq{k5JUYFpG^kw-wL-1)^?-#+h?_
z5c%hAP?KNLP>XIy^`%-O^SSThK-pQMmLJm2#+y#k-ktgB&t80AQC&^9xP8MD)M5?Y
zZf<Urx%}8o)?xf{%DiW>y|H@$^6-8*rdJ7~B2-!)wRof0*e=WkiZj4%&4J!VdQOrj
z9F&nOd#eSt7}u|{vbE)Y#wY`}aQnb(q@=ojf3lL1M8MzJd(gh#RdItFW2myBvr}<o
z#Mag}H#e7d1a_2529{ztY?+x8@#To$tp}spt{RGAU*G^T(opkkr*(b<$cG%<^2oPu
z`G&bz04*^pulI#xXc%v0CYG;?_a6{~9=LA+A~HIpYBF|U$_7ikJCEJ%Iee|P6UaD<
zVE^7;jie()3*^;cK@N@#o;+^B4HC-A#rICM<hQogLAy5x_G6+of(qP1igP21`!=)9
z9hZah1nkFEX7)uzM_)EEQO*4GonPPu^T3gx81Bkrx?yG)2XcS!a3+BL$u+oWu&~@(
z=fS!Tj*k<j4lQi6ICNTvY=Wh>L+kjYKxM8yti>3|7pXa`&W(8<B;!aa<du4<+-PLO
z64!N=sjtW?`m;eUvh;N<;cZ9<FpK(hBQJCs#7!ZvOE2J))q0-;`D4OxPDE;w+gHTK
zc%&%f>MaOqaOT(@KK`K+`!H;76n1SX&biPYV1qd|K6vneRriR1fJV`Q<X^4Q|4HwN
z7A`p(^7@tKI_+rk9ieoLfLP!%e_JYZ;5=jq#@DDVtdLci94q8&h)`Ji4G+yN@}*Pn
z@q(|KnNjLqyMBEi9bH67NQzwU-{!a6=XWU<hI=AZ8Z(0g$ohAodRDm1<vF$%?)GI~
z8dxd2ptN*A6E)_%r35>2j&y;}GIUmevgZ0Bl%<@0|4iTX_wPJ+56y!<WAkOMabJ!O
z%6t>|4ja(g3bF$98w?I&ekeE@7kt}n?Kfsu7}rorhhcHje(^{4O@JGBiQcSpf0iz-
z#QF2>=mKmkzK_NkPCX7(A<_>BK3aPe2PG;HmDm|)z+xBcQs8?Kz;BCs@}i}cp<yy=
znRzZO0Wi!@GHc^pU`i~>2YB@dvIH1TO=6kY+1dh6)CJx|GeA^V8TspUYh5*M{IF~P
z-LfBVKWH%qaq)e2mY5v~jFCu4qN>9u2I}g3eVafVLaVq8tq<f_nLC#tl=UYkrX7&l
zZ-GtKwievKKZO9!8%2D)yh8igY=JknXwRecOH3r}Jki8Np?m4twWv8sMBF%#Or2zP
zm=b^Gb(YdtE4pdrXU=q>t>5e5XXG}JCZVXq5eI&Sdt4e6^&!9g!e6ER6(#;8L`)69
z1vKwPb>jhL?DlB5vty|LB?vgKl_gZ-P{<)G4D|pMX%Ar|f0YasRUm(~L0MVZp|D+C
z+j>=ls;h;h)OAm(BNw&B%VQxdQ_y8cn=ppz6eak)qLVz1ITW5c_a;sW3i^HeJTbxD
z=a<1Mwwhau-oeYp|5eA=AqR(`1&223SAuE;&kYv~mL={b!qs%pc2FBD3*Q6k5icZj
zx0r~C8+MI9{aK)`ptn)K4^lNA#IMPNim7%9eV!7Y4CP~l93?^i047GRudSu^_U+pf
zrmsH}3db7~*eqUka0B5;eFQK9e}(mh9@R^qv*gDppi*HD#F(6_ngJAjhu~3UbLp1j
zTMr;kjnz{K^bJXP%Zm4yX$SwD5`MZhjxRra(EEMV+{nmiP+wQKAE!Giml+^$!<)k1
zA(`|(qL2FpI!voFibRcY$^?7nnCA}pKlrQ0S_?b75|GxyTA)AlA36WQND>V}4+Z-B
z``^F+8~hdqzsizl-3OYHEpUqojpx|bfU<Y}ct7ll_vY>6LulVs&JX=L#8#p9|No!=
z-#&q`3+Coup^jZzS`ro(MiGb#TnrX~gzqo5BjVPAS|vhddHEa!q54ICm=^XBG`rEz
zyw24=8Q|HSseA={4dv;OKZ0gaLBaM~xYIFS+wjlTCU!ZJhW+l|y_=bN?F0L5tmkr!
zh(^k@lr$}?pm48d3TuH_%`2`5S!{o4yz#3y<-`w!EX>dE-n;iM9XV=gbofj_lETKu
z#vvyP@D=DO#%Y8`6qc3ksy%n^9K81*B@clGKg+rLoW>6c3)kQz4nDaJ37I}3pkf#f
zqs9|52o_8NR*A`ZadG=N|I%LLcf|YQD=&$7UgLTg_6aygB#8r(3dPztcNuhoGiR@U
zaz;M14=5Y%99%(K@8;^N5{RTMKRcTnl+TteY~_D`FcU&<0(f~|=j|<h_3BlnPY~6}
zs%?i;m%@K&O*}t1OVC|dL{J!jKy!Bz>;C<X*5>BsXZcn?%e1wzQJOR)4(zQk1p!cD
z#}#b8|1hjHKq53nt?cc2Wgmmx<B-HNiQDprMsxrpga--gmylkywsOfnC&Z}9z+CRI
z{rM?9xWfTcY;4<vWDKrN>|3faJ;&)7IOyP2F~H5Y3&d%-NcSREuz1)2S}fK;cyDX`
z`vcBS%fCv2#5pbiik$|s!5#=4b3auDaHBt(f}745?h!srwgF-KqF>FWC+nOZ1FjFg
zvU<m_&ticLuQ5TU4z4vq&g$gJlPs5zw>BSL{gH9P$0(R?yAKtUp4Jm&C4|lF^t9D~
zzNE0UG%SZ1{RekTR)0$|tiQrqs>n4lJ3A&eR^jsM|IcV0?MUF$l)y;BHP=?z!bE>F
zsGW?-o(v2m{O1#3vZ$~>5NE-I|B#^g<`xo)4haW`6aV2%g{(f_#0Px%5dZH7G{U)6
zsmPP)Rq4)#I(q?sAoW^Y+(bs^?d8P)Oc_!U6la2j(q76J%9J`NGa=~0y2p#Z?e*uK
z?IO@oO^N9rKU5$6LK&NvRD_c$$ZGYrUe_a=W1n3jaNmpec$0ua@Q&HdHlf%B#DaQ}
z?QaKS{6^_vm!IY><|_dA{OQKADsel(SCKm71cPLet$X!TLo<PjLL&)S`U*7Xs6|i;
z?-R61g(R}P{Kk=Odo=zfPp|0_VE0Ssxv6szs}EN3A!@vnC=zfCvM-WpYHRaBtH;ZG
zt}znYvar*Ck*AdZgFMweEiJ8EVZ{PnWoNKt6oLnK^3BvRZ`h_2=6=FQ38@%RqI~G9
zaakCol?lU;d2Q^hB!8|TZ@rM`3s@&5y_4@Bxs<cLMS1>NRJGX{p{oFC3Z@7U-9qH#
z<lBE^ZRn7iQMaM^+e|@$>^gV*3UF3XIcZ&^MR)(%4~d>Gz)1in96NSw&#|io(AEQq
ztuX;w4!Y+!5<L7)+m(9+;WIAx5(zUp<T*vv9{2Sv;BTDdeSCcUdv-vU0beDZt4olC
zfj^mvCBEuz|64$Lccl=HU*7iK8P5ad6)LRJ#fvSl*g`+IKivY#Nw7t)UcL-!K<@H`
zMB2$XZCGCUm35UtqXcc`YwX<F+e?gTUk_;dh#?`Dgp?G&&`3u3AVdFrQC$n#lZLzN
zu)u0;M*M?GflOjZYceZjkVPtfR{aa9<j42|!;^$H&raVx*l%$er_sQJmf_NplHo5|
zzze30Uo0mei=6<WDRz>dpKH9R9Z)FH(%jE?kX*lrJo2==yA;lBLv*TB=wx_(B0_hS
zQa5dQ`cx`~1D#L_4<3krkdubw2~7y>=ktq;ZLwC9Xt8*KVrOQg<>tnHqogzhN;gD_
z0GgJbPM7imgkvzv;Mkaq`mSMQG5%x`_{Mz#R^nZC@X9YJFIQ(%ZoHnk0k2dk_KGeT
zV~|&VCs`@MDde2%?dzkxmXiAC)rvnVG3z<sSbCcnwV_i1pfw~lj9>%iCVDQ!8bE3S
z>IQe9Qw|oXl}auAe<ME^rd-LMU>|MSvSm2t9D3e#K6|>}Man5xSRGVZxd6hCBES^`
zR*8UXK);}-2}8={K(FlV-hWq@_@f%PMYY08|DBRi=&mL~3wppP4Tg-;Hdjel3jQ?2
zUO-(c{kDw*n3+y308A`3?k;JxZSvCZB><_|C%R<Pr%ywm)T%!R?y`bZk7cNq`Wg@0
zX%HuW+kQ&G(sK~<y55C>7uT)>&=IJ7dHC;>m}C&NIbC?trPgc{Z~+j+-})wN?0{d~
zKm$wc7<`v@?QaLR*dT#DfWfILHU9H8mnC3kqSALvb_(S`kg4Du{OGMZbu1s<142SV
zEMfqBP=fH302>~o<(ve~Mrs}kOTpUU<mAyy<NE%6_TNj`2eW)_KtLBR@Sz$#b?O%a
zBq9ZN?Igs?ps42XaDtPI4g=_$VROYpxO;=YRN;W~*Ry=;pDp{r*b|vX9zX2ZA=i@3
z2lmSXO+$^DLgPD%%r0Ad`>?Sq$j0kEkO(g?FNb|6Vd2tp=Bj3g2RDO7zZlAJ#L$a$
zSwKjLn_>}UBA<@KsxmVp1^B+{)vHkA2dIR`5S5DXddtg~e<JuA$`E728edZMuaTJ^
zZlP}ZDT7T9KV{C(yS?w+xr1KNtEQ%{lzciBojpD3N^_X0lzf35Tv|$c`a!mI>>biR
z_lF8mtV=flYm=!~x-FJf$as-2>>@@ZH#AZ#&(t|$s`Ju<oLu;y%{0yd9f>=6J>feL
zHN`C=vJ8hKAR0JatfbAfjg3=XEi@&5@BLvo0n;i0Eoc!`zWzc~mj7@d19g3IadBW^
z2VN}@wHnyX1K~h*$4vpz>=MXs$sa&YP}?ss&Tk2;p}Bh+^2ZR5oe<M_^ra?p5r7<P
zM?E>`=j&ui9<+^BuoSy0zi}Hw28WnQD-An)+o4rWmc1R`#KMM;I7+ymOrqX|NTLn&
z7szSrJ5t`J259ja4qNNuvbZx|5-LU&YjIRiwCgDvp>H*~emys79m`jM5C4Nx<X0g@
zLy59oxj8qPZ#afV6=+!^7o>IS#??p+7s7)WjP@}w=v<X2xJ(W)q<x<LVSgT<sVSf_
zQp=!cS-uh_#BXT0aj>%5ZgHCbn+S&GP;{#x0LW72>r`H>kAy-jqyb3h#YexepHRLn
z!eR;zjZ!<F06m#$0)bBq1HwX=JN&WjdIovKgM^E@^TP=8^#@U+lBpIJ6<s{`tfhr>
zfFDbmc23FJxr%fdE)Xi&YUh+ji1wrm<38zsn9jt9C^%D+l3v0s+1}jM&h9&?hsnuF
z+Qr#1AX)m>dtx6vxCC*Fg<hty=r@Xb-1gla2Tz`yLpg^i3E&4V@O9o|UK=B(m+x>1
zsSFD(k*kNUs2>OcCU|asm>2_7UNgMK^XAQ)H71f$QjI1mQ6+)*?$M~DSEZ#~<3HDe
zmAXGcQ5WV-dyWgSvKpczir_j-z7c{I|5qezZ~5pOP>clo+TXYl;YE9~VfuH0R3qZg
zZ->CnM#-bYGUp*G!|6Vv{TsufluJ^Poh|_2HSaxaeH(pX>Q|3qgB{2#!h9IlyZ?Dh
z<0x04f<W%?FNPQ50Ue#UX_mqPA&=iRK&Egu9#^@1?HZ>_4s-c5<ob=_i#BL*;tbQD
zhf-L=<`daDWpy%I;|x%<xP&`=EZ=_vn-8}f#8$Z@&nRfaW*LQd=TG(VO;67SG>Mnc
zgP<L1M(pA3X#OW3dL0oH6N60+DLA+%Q`2$a=a9ekK*2dky%0MNv?qdB`{&xgv&2L?
zhyWY>s9%6)j-@~s4cv<+5hk!aqeDWJt&4b9l&X3G#A|Q`Z32<(7aJL=(Y?`@jDZW#
z#Mhqugwj%b)Z+&(#4q9W$C-j7Vk!JIexJm<3uS<XD^L-YYqwov_`QKDMO;}uUdJJH
zQG*N|Y{CGTF`PvWVv=eMsoYoltA3z8Q#pRCJj@YF%_0D14a+Rb;mDvES-)-_lqcMm
zvVg2Mjt#dWE6i8%ly^7J2Y_h*te}c249}<aq(HFQ`-&NK+>&_vFz{<e7lvP1ogWd1
zshJ#jFm;9d-^Zl_nd`7hALLNP*fc2SaUNiCf~4@*?8CVRJgqz@hK7mh+t}FH$BzW-
zF=TFq;EW*Sf^7tKr+@4ul2TbMV$gQixQwhZR625E&fHCkf=;AV0A<PE#71&1PFw_x
zxpSZ$cdhEjmv<2d2I=0a7RHV6q&~7%1J436VWhv=1Yn6jKayGAl<?CsGT7c-Ku3y_
zm9|SsJq0I>L>+K_hJ4v<d{PYtGF<d?h53!_FwR}oHxk)zq}Y)l0`-B@FwRuyqC)D{
zrXF@*ng>(nIS!{I37rapWQ<KL1P5)+3pw_GuI2@-?nD=R2m{p<oUQHcWd#L_&#AGn
z!kv$T=uuhPTG@FXCq9-t;X|fdh87yu$~Yv#5pu?smX@-{Eb?JC4O!n5F}IK=!H)SQ
ziU&!m9lM1c3o2T%T2w17{?Lc<e-1A`*w<{`8h~>YV&#jnM+hQTco2(L{t^AQZQJA?
zdj9zS9bio{uqB9xGAy&apUF?jYilo}m`>ZN^O;6=8mG2u>yGSNsS#{<(yNx)os3T<
zv(YmI+X_E3JT_V_`Q4W!&o`8!Xj8WM^WCMWYnCy5&*{dY;7I6o4dY=Q@GI?|585IS
zn1_XmPoduvCvp12hiaZGluqIAbpbQUv)e#Z3Js4yQDLDqP_U7`dE=rgD07j<x`Ra^
z!WiC(>%|2EiLRLf*2q@QKufFonHGsTyQV=)^;OkB=eG(ON*Mo4Nf7m2{ksl)3!J6R
z$<6KW@5ipnYR&sf8w+{b@DzGjwYY-NgRU++j3=wsYOT@YBL_kq)fH`d0s2<hy)i!E
zV68qji|UjeBIA%2YHW@_OFwH}%bO!vOQ$vMQ9bh5U2s8Za^uEaLBZi4x^h69L*J`h
zx^yY?G+~~~MncTv8WG5Gm4YR55o#L$XOt_v8L5s3k34~FTg7h$&0=@VoSdA({%mqp
z38X3w#oh?<yp110eBj3O{F^?3y7U0)FkTUQ7#`v+kH~K*kqXcxLa=THK#QRV&kX%k
zaB&;$j2St25Z(+8h;JHi5Pj{E`{77*_1ZPR0mffQC)$qwb9CP%(Etkq+I!#o%lpb%
z$jjv>l3_{``uv5c>wJOTH~9l0al%eEUwaF&hO;`TI--sesOET4D7J)S3or5rF?Onn
z$w}VL)yT5tmdI6VlSpWu=ZenC&rh)2k0gYB;xs`D9wYrB1^+DE&~&7dEw?J=fD}c%
zv8Mw5xmnLa%$kmsn2nA-y>q2rQxFPv43YBk^;Msq+6(ba>Il2oPxZ5ydRW?U1avVV
zEj4LI-_|247y-j=y+1+3CA=Ka135BtRc9w${E!wf)_>n&t1^-&pH|`sX1k9=27ox)
zRAz9POXdrqfVMWKrVg>G=tGZ|rR6vD+GV)6;gAQ#PV2yFfac)fjS<2RBO)T$9tH+p
zP<1)<+ARZan3bK6a_=*WWDUK_MG9?XVR6Xad{%c4NA>C{)A|a2m#Xh8rInRmLF_xg
z47lS6#w3K=SzB{{uL`d0Lx2h!(*?x@(TXO}X*_atF#S#JB~^9w_v?1auy2V(X$$jP
zYXvj236wMHeVbA}w}heT12qYnV%5Ux!J0jL_RKqvc>{REwhU6Q9{PB6OfV5@VHfwm
z*8_Ir86_=8ALA8B<$?vTjHuu9_YYR6+EsEw=^@Zwl#rtEhg`pYJ?%kXw9Q^j`qhqR
zg06a4Zw7B5L<GhwL=``K&$Ty+1gQeY`D29Jani_atp*JQujRc2qi6zoHD!SCUyP#X
zvorr@9u5$Y@WwCiMavk8gp+4iL{C?jGXm8na}Rj1iHnN^mPC~sc7YP*&?8Bc-&lNE
zOlOxtXQ1N8k?%<X(;RIi4h+PEj4j`&gjCps*8RXz>*N~1@QMmKP!-TQkalSU&#2IV
zP>WOms2iQK{v-hZhyI(C5tEqMiZjtuqo|jLnsA7Oph7#W2Mxocnwsbn*R?uJte}4P
zmUDM-D1+9k!W)%ni?%rqHmnbSey~O_7E``M--OoQJ5pQsoeCXXe&XAT0=v-wId515
zp=p4~x}ue6ROPM2hy&|F>~zjRhCO?@4)MKrm-tU&{UbOulmKLsl9JNXd%)X+X5rxT
zK+ENfd6FXci4$eX{jYHoI=Z^--_*!))9NpyPhqBuHKq6EyLCuVuz8CNsTc<J7yNId
z!wF8I6{*9DqcG3RsF-l4RevKjLio=4La2&2KzB&zK?yO+VIiF6+ws&^rBT>hD20}n
zl&J6vxBkiZe_$sIEeMgr_Fp;4+k{^ex8L|G25URNpgcS-&JB%8=xA#l`|uCF_;L^Z
zu3ZP(|4LW_-lOwsX%<ih7=z6TfO{Sy#?>E$mOzOHq)GZP&qo|lhs(I@6~CaasZj=<
zT>*`@<kRzTM|mGs|4B`ecE+g*MuDwQvB>>zRJ=t-6I^dX_YanL^KB+dJ_IaeV@d4D
zxp?9~5Dq4;TGcetI2yytIy*Z7GGXR*is5J-q%{i`#-T{ku%`w?kYf3@i8)R~Gv0;@
z>GwclBJ+Yan+>8(a810TW@e)z(dE_2Nwh(7c@xEYd;r?zwoy^>WfVAmW5#Y73BoXo
zbdImcd}^%zv)tk9m7J6`gVPcsKK4e>+qZ#0(-wOJ8sXEc;rQ_k3ILfDb4O=qwdYGR
z($e%uXa|_jo<F}!6iH`+t2DD^U2~xDoqzx$S)XUDwx3Y5|M@`aZW61Q!~4$8!vGeX
zT7VNn0q7r72<&%cgtl!f@V|NwN-ih(HZ*D;1p18DF{LD4&Xm=_^!h%aK-X<ngp+9)
zCW^`s?4xZA0uUB?+K1r3SWM+kcM|Qb{QUfiO!#VqO%KAe+N%z7pV;!6rBZL4^>@tB
zmZsBUY{{N-NQ!9MBt=SEa&d;E4IjfwtoIVibjjO70s<VZ$Z$^$@B`F&P}jnG`1tWh
z)}MU1UL5)pnqJEz!*o3W)N3AkhhR$_$51H|%}?py!(Rplpx3R{LaV~_+vHI~tYd~k
zkig}ZjvzeOcoVt(At*3(yMHFAM<fNu#SueWoAsrzkZfvAAZ|{87|HJ>AKH`xxSk+o
z%+B}$3HZTp)m+Gh%L)sF=nnJqR{=-!|L;Y$LpqStThh?j2u&F`G_1hbfnX72@r67T
z*X2AX_=1&dwknMy`QUcv%RzBjc<`R1atg}n2@IVHrQWx188trv$<5@($A2KqWt5d?
z&&Vq%<TT{7WxQ1MLj<;@Fg7)<f&_fVe2K`M03C=vll(kI@Q{qI1aqpI4ozyS7+T_%
zLG_}CiPq&lkkzEK|Giw8>D@0cFGq`eaFi;RbJ6dlYU~O6bCl@Z(cwMC3W@~)n$tUd
zNDrX;Xz|4Y04Hr-YmHun%b=LPe$~;G$#LEqhJ=FCy^!y_5h`W$egI!;=sAvzpG4yN
zUa`RsM9D%mmrebC9v#f2Q<&kQ^dQ=FGYr~}{Xq$6cI<cw^HGqh3-`~EoFu_i0%yb@
zNZ{jX!fqp9#J3PWC;U%y=3C|nE$7N2j7K>+?kpYr1=f;xA%_{w@92;)x^m^9(~HxN
zoiG8Q^zc|AxJ{vXP!*`5N3;W45_AR&ucIhJnMNRd!I%HVD91T)*#odiM+=IEMlT91
z91qPczT<a5Q>;K;?(w?=B@!4XPFMHeOHK>Jwt&G%xw|25_AoF2HjrAFjBzUgckV#*
z5jvgUsSNE+8Z9hQM7X%PPMy-)eKc)%nZo{6gkMfcQSkw=#MiHnecMqEks~s4QojwD
zYuP^ycB)=@Pli;nA3t82Q3o6f%}|KPakTW={2{oyXCTb<3miZP@55;WXmZf!1KnWi
zJeRO=tj=Gs{Parc4c`-6Je6q9m?<epZr|c5&cAMd!8+=KXJYH9FVWxH?s<SEll_{M
z+koIyUx(hznl(e8tEOh3+Ju|T6=!iJ{Z$P!ht_`6uS*T}`m5*Zrs6Ye*LbSvhf06>
z1+7Vr8e1?B<?uIrcOvZXA5W~xt83q6qM0|5^|xt6gB^p+yRx}S=8jo|-R-k~k9u3<
z`EJfF`eA5S5l@f(-%fh})ZYKs{o^r{<9f}47&i(N3u9Y58xw=;#J^b@no=A)#(#qO
zFf;M5!oqwCCKgxht}-7zCdkkH*B`z!*UarqZ1~QY8`zngHZig^HW3x2_~#=W-@agI
zyT8ADMpQrP6OZRek%sHW^=tP}UR|4D=wexRLH4Ws&+Yob9T#T$*kYE0FW*}8#`N{M
zT@x>|3|}sNZ}Lm171pg4b*oAeroT+P+xpk9#m%3|(migylFB{5@#ZzJS3Q>L`g#I%
z(V}#}&bIIPV)WTkeqelQ$CEDdimpxVOaWdIrVMTk0irz{q~ZhIWoN(8Q~X-vcI=6@
zq!pFlv$eTo<4f;)nBc%$)ugz$h5Y=pPN~pi@^5El4A?}1EU&2ij+s{JA-!%m7_1iL
zb*pv2%g<9-PS9EI7)Nr)t4}iX0~cS8g<l$9cl17ufw89ki^I{+*shk7Hd1fQ-(1F3
zAY#$LvfxxNBz{O>lY(%%Xn%=-;Alx_Q8*{Xi;Cl0o8nh~vYdMoa=ATG^##po!=ePk
zNVVPRuCmI{qHkY3y#F*`#N)kzb(+JH=Yq`Ue~47Fh8dkHofloHXr;~c+gr~#=+E|<
zL#HF`X39jS>G3m<jCuVMqNy9s)r~(tP{J&Gx5z(hekJ;&X~dg|yd~2&KTdEQZDRQZ
zO^uhYfA}X}J)007QNw5hFKaTJ<fA!bxo7)~UHXDq54bv~DBM!Hc26#N+iu<B;bqsw
zt~@$BiOHml!j(HRg)7J3jq#xal=0H2U(-hI$m0=rX)ZJP_v)Fpu{_+6S9QO>xvua4
zN4vPfuZ<zPogPBN^1lTfKM5aQJa<9x%r#?%kQWcr`G1o+&R89~GR!A^<>>QUOsB2l
z%I{47biBL4{v+?s^Yk}F$4pPI<H}TUDElmRHR2d=@kxg44Nc9q59ThNIV0uvyJkwl
z%c%Z2o0hdHtyE^z_;-W0HqMq*FDU`jmdE2!Z-Ty_zFsac{icO-Bh`%;MZ*!x(f7m2
z?6_)GT#k5Zc1a(0s^L_EC1TIj>KUi&)ibhpn2ewB1j@VHN2@M={CLDE+E&)=Qde7!
z8(GA&T|*%vYJ1h=HF<e=>@Q)VBD<2|lN9Q6+FT=1K>MXR>w@~O>F1R`TR8%A&S~nk
zzS`HoZl34Vw%w^M1IL%vF5#5Ohx#TD@5P*Z-1xjOyej8&SJ{RXMWg3|(;tpS-me#X
zG+f1%_jbeP^t!=-Ga4B*rF-=&y7$J{oUwi2oI~-R-%gq?G_$<lkRgt=XzzJ+tL50U
zodPMl>W}wCc5h7Qe!6pnF2Cff*7EQ7;^&`;x}M)g{kqwJ;!9*&gUq$2HRjgb@r)Ow
z<F7xrDtn`lY<5{}mdcD#YxkvLlNfrB<$|g^GaufRDX$TARgn`}ksHp>AA7*rom3ID
z)#}pcO#bH6auo9K&Awjsw<;;=D)E1)sL9Tt-Xhs0d`P?I&K?cjP1iO#Fl)bV_;u=1
zl+C`y-=x8m?L(TmMeSFlvm&}TPOuf;FG^mYv;X{2-=Etjt`5*<__9g7^qu)|E?hdQ
z_ABj&Z;y|prw4woZt@v=_h2U{d*qvkPcBI@@gC`pav1eltMTFi<8kE*)somF?{?iB
z9&dhny!d1)xj+4d*vrqRs96<19F%3%<4^5&`F@d{w6Alh^!l36_xsg(E%s6-U20zp
zyp?k)&Nf)Ucj!(|jj2OD?SvUeV~Mxa(97mjA1TsIRM8RNVWSOFDL&>YZYBvqs<tls
zl@|lH9H+L52<kq1V~?f?nOE<16LGE4?#lQ&7MY>owb87c42^D=N1GoxS8=E-$=<6l
zxa=dlomcPOWuM1)_HQ-SRy$K>=&U(el((K?ztNHS$YloQT4oa}>w85)yw4RUw|9#+
z=8yjL&%Zm=ae4Dl)ssP<qcLGp8nQd?D=Sa_a=0&HMjqE;*1IlOzSD(#$=)bDaZ5=8
z4Fx;9)|Ao8ZuOJ09LIvV6zETLiuaSqX9Ul*(7ds2IJM2<XcB3()YF0Uf~{Ipk8*Et
z*0^U<X2t{dhoh-NwOZfmYVGPi)QT><9N_%UvQ663W98?te&<wEr$2S4Lw2(~+)!VA
zyCC-S(uC9&o1Z?``|4BWm><i^k=Qs()Rd3gwa4EnN!1)l5L*$jzu=zL&Up6jOUX?d
z(=!s@q0g%<c(xU(=D!)zGP$Ke^JBZswyU2?`aL;6C9sUq*zR=s+B%-1z;(I#(MuCO
zv-AAfC1QJZp2v`m$a5x2y3$dI?2p%>UQXQ@yEgvj%xRC=)Q4h5lJv2ojn~NEbB!H3
z8(423c)RQlUtjd-hx+B9h;)tDKehzDNzR-PxBBSI&ZBrSlFvEzV_C*+$*x`EmtVce
zu)a6Xrc<G>vf#mA9&^yEEj{Mb{upYSeVj+0@~R#*eg9Lj#<;JZ>#a}sz4fGN3hL8M
zqjDkpv^vh)%shK>W|8{2P=lY2@&Su&caHi96&z^S8j-ZlR;p~;Ldm%lT6835vcA2u
zf`&hprsj4tU#=Ol)h;#eFWTCrUz=_(Y`wCMeS>6A`$@|@`J)14D)u+sn;0ctuF36_
z@INSZB*I->mvZcsaMLe}J&~UtGWAEVSW^U;oX?>5wz;=Hjq#F>*spCaYvqH4yf2Bc
zRV2hXwSDDKnyaQhe_;4B!<Z$Nc**nWO3}VW7SALNoypfMt@F)3!$K1eru*0JbQo9c
zd&c;NMPhTv5J_Z2s&9wpT<lK16szQ8sm%|r@Y{-1GNyA_`eoDnaNo9%Ed71$?Su(t
zS~1_uyG^1WCk$hx11Ey>Uv2q*uSmvCm!u~WM)^p#Man+q+l+n7(H{>OXY7v*>8B4z
zUI|vXSpTkVY^^rwcuJ*Qiw;XQQ-RfU`fBB{4c_cs(Kpo6zB>QR3?`qHo!Y0;qQ3RZ
zY`mdddv<yQd$GQi(~+*JmlaPx33@0FpCFHR5-u{9FaGf4Q|W`~=JRPC>-G2=%|&V{
z^*TCTBL}}1^vJvCq^R({-Wi;4Dl;(hBRFuJ`N`86jSGIx0*n(QN8i*km_5pOwPCrG
z;OZ~rq9H1{O*?)0P<pR&u;{0y!ith@ffK^{@|D?QHH9z49rt$gypq(Od|6vTCD0vd
z-rfFXqM|kRMAL)Bs;(#erR<!#s+Xi@L(E)Ec5cy>QZf5cDRI;&+FO}hO}5ukSzhP!
z`_5B6e%Cd%!M|^fG+C~Sj<Y)C(Z7+`$@^rQI{Qh%bH~nC^gLqQsxB`h7ggG4uv4Zg
zGxr3|**(psuBkUeHz>W;RzGQLoi^wjb;@p!=ArM6TNbDX?#(NY&vt7{^vIW1E16G<
zO9>1d{y_c6nqugN$t{caEXm&m6R&veEzTHj+V^~BYu)SqmuVgso-IcXL?1fU&yXAW
zFn}YEZsz93fiCB6#^>e?0mpKcTDspm2@fd?ktK^U&a4ekrb@WaJjXd7s*p`F^I9Vg
zynt2LY^BeKeV0ojdE(NCKPK)IDIZH^sGiBoo_pucCZRd0oZYdx-QL4m`0dpbiF!Pq
z$(%+97oU9TWo3%q(3R+Ku=QE%_4$JZHBO{gUlp^o8&dp*gs%?0e0<yVP?1vmcaFM!
zPSxbr93f`$J+z;wttApv&RMt@*~s0BZ!uO9Ssq>fUe{=&$|1kY-<a<6r9){mNjItL
z(&7VZ^BPWy4=B0?_ER+mT)i~ve)!lmwz&`%1-o|zv1xUin8UW#-WC5%+AFqvTkp5{
zZ8zyfF@w+LMO(y{C%zZYS?;~bNw?*u>YY~gHM1d>E;gILZjel({4#8?UFgXEyGOh7
zZnKx#W%*W*KfSg2qCE?m6Koe(`g~^f`P~kE`J_D7!}&TtvhS+Nj#vFl?}MqmJlpo~
zv7>1?pTbfgH(+%1k@Zx?{FhhnIF+LJ<i3!9HzjgBKU5&2QZ`A?x12rEf+bU{BI&oP
zq^t7r#B9y?svloyU)^>r_J;c2Gv^MsaaB8})c2gOaIk-O^YF;|w=|o^0AL2D(k{GR
zJDSQn&=O+#rCy?DAuFrQLvx_A&DP|UrNMWK+#as9+@DvT+`n6KZR@}>df~q4$MHkD
z5!|fd6TB-=c`uyRYJ2|XIo)Aaov0Om3!ZT(@hzgL#!b|3Hyidaj)?f^byo<l+*q!n
z2^6Uo`K247Cb{(>S*}oht7;!dvyN=RO*%nc%RHCsT#d~quNJ7VTvq353(HAz-aEi^
z=-r_}@AD@`TD`NQ&zd~ujM6qbx!Z%`+TJY`)jy`?HfgjxE<Nf#BAI(a#506t&(+Jh
zCl@?z6|`R+ZF0HmeO_9Bg4V*6txes@yp}?Fz;VkB@wgn<?1H@6+-sv|UmTtr8M~{0
z<@c?yG`SZ|>feygZI+f!`9V;kk7GxgQXa1?r`~inUqWMH*Z|F=fs)d4zH?Wyiavk1
zqII&j_>9@RG;Zyg7oS_h<LLB*domvlke06x*!gtx&P<)Kdu!?R;nqP;b=td`UdA2U
zpD)_P|Ge~Wz=J{OO;yVIV!Nk5KeTRQ7x$@ucW8E6z1a0osoH$+tKO0nfpwBco9z21
zR9MMA>paXUp7nN5cMR^A=(jlVe1`<bE4{b2w_VP+cOAKUVPDR%?axF5ES|CI$)C*4
zzZ;NJ+t(xeZYg1hnwFvz^*iTRpUue*Nj=x4tGfJqVv(-8i&NA0#m<3Om8vJT0)lhe
zeiy06{ERD(;yZfm!?APpdfWQ-V&3KYzGHlJ<nczPZEL^BF}dHiWI4o;lai$I=6Q*s
z#EsaM0SDIsEuG?qYW1Owdw;k^E#1g`T3Qk_cf5Trqap3B*1;6-Qw)Md7Q<Rw^Hv@P
zOnbDNjHLfRReVQ}|DP4#@e{}YUsZgiI)WkG<vZHv)DMSxdRrem6w}BR?7WyoQ_=d9
z`yiKeSy}@{PJ_olz~0YKZt4VYyIk3@#?Afeu|uS%`J3G5e>p7t{OL<Ywy7<4ml|X3
zvd#8+&rkYK3OhgFruC#}-!|Ybv}8A|v&Z$&{F*VjL8BQD--KV^^Sqw!f!<f(k@ohc
zJVsH{`UJCwYY&>;uM`RA^b364#JrD>ohq~-zQU{HRmXt4>nF{tw-S;Umq?S7VwNiA
zDO;5pwyn%h^46?jkF+5%(@J@7t6Q0vZhNQC<t-U?l<xP_I{hnbxd!Loij&e476udM
zNisP(H9|`#)s}OM27Nu4e*0{(*x_(brk>*5S9Pxi3I1=c2i~2%Izqo`#cAgaxyx)v
zEzYT0jUH#;duwFFy1RQ*eyriy5c-<pn%QWO<3_bLYp%MwKi;&K?6YgY$2A*#e<pqX
zZsB-SWyz;tomYs%NI=8MrQmB7uTPXwtxR0pUnX(>*ZaE7DpJ++2SQo5uYE^aGr}aN
zed~CG;8T8K4f%>`Hb;kW24+pS?-@rd=g+!au5qLgNnp3W;ww?~ab*8KpP$c#)`bQ6
z+&X+*QTI-|+l;`kANx0OyF4eaWZNQb{w3s|fOXm#aV-won68`KImmev_)`uJ6?>h|
zEz6*^^s0J&GR>mF?P0Pl6FI+(e~%7ht3sNd9eMCut=6OA=k%kUWx79vs@-xZyD5=r
zJazko59#JqK<N&*k1CUo##6srC!27k6<y+dsq6h@jcc*J>(i%u1sojg(=2%C1<wv;
z&TsTTmb_5+MTbK@i#1H#kKv(Bvl`>DZPah?H>OK2E2;^k=kJV}%!kRXvmC2i4DJVP
z>&;4<kWb!~bh>+vjIPp+m+utm*7);xZ}&bl3Fv+;XHe0zhh(Mq<K$59VV&zAKFoJj
z2j}e=JT;;BD<!jZz1V=qD>9?~Og+x&wehDrHK#=c=L`4BzrD&rv1?<h*SE2^tty)@
z>SqO|e1FMGANb{B7VSkQcV5fn9Hvi>>bVA09KpwE`kNgNIW`sXg)s)GW|o|$iDy+R
zvkX3!s=mZFX{XcQbbI~7>yyh#Cf_?`imk_QaI9~RslD&M&Gq-R!SE9!E#|FTj(j-v
zP<)GP9lQ`T{p>eAtx>g5I9ia^xig02*>%sU8$*libYqc0UTq1r9(2!K=bEMse5pb{
z^l{x$Ix3O+xYo#z*`xj3s8n<)hmgZb!_m$ckDGG(5^_YQ=tXYCT%BK+BCw&4qHp7l
z-N8&Tw>3>EyWMJD1|F_CVS33$hH8+rRJDKFc%x$Ip}5)~N$X!9cpj^^YyJMnCiMox
zCD#<%!&BWG+g~`(IdrBz;b~wyb;R{~b#P;jVo9f#TZmwX50}!WUG%l|rjJy*luTuB
zPEFSK@V~CT&w|J?bRkPPTRu3_i0M6-@3s2k3)MBbgH7M7yf2y96l7hu@y|PFNtNvT
zSy807!#r0!hfICwLXgN5Mcvm|UB6@HH=ZiZOAoo5^==)#&ru^TE+fm5rC#<p5jB6`
z$^P5QIrN(jzT2u$wuOGZuL9qX#jY~e!!usBKYrg-)|Ps>nKv`ndL%l;jxLyDzH5OL
zn`_b5xi6PtVH^3`Z~1XF>}p#-d|gmE-D|4a{NvE=_LHwFV{4iv)PlPjQ-OzHiym6T
zQ2hS<{?judrlR+?z8AJs>YemR(3KR@H;sO4G3uv6<>So0_EX%2gJo1F4!6qey~H#p
z-D%rq7i{phX6yX!k}8SDS(nFjjk>q^HqLdq#n>z-MqO3;P${*Tq?{VA8{aLIxpCxv
z)BA<i!*{PKY+iS=Bt(y?sq#>q{_r09ZTxA}{2kZlZ7M=_j%2cheR;;rUHOc&E-}}0
zR5tw7^SD>0l;rkT$SEhAoQ&U0J=5Wx^*$gmUXbhZ=420Rrr-%GwdgF9K^N{dT~faO
z3h9(8nQY#6uhZHu%3POE|C+ipEL#7d=ZjCdTQ^%M4;$<>b=xZRlt;POV2ypuV9&XA
zj4O#|)T%>fzN2P6DdW-JwTI^4Gnwx^Un(T<@X=PI35R`~p)D>C3+T?Fd|yZzU-olI
z+f4Q5EPanTx7f~S)CtGQTIfwBIp2T2Eh%?^=X&Uny{^h(-dJg62W9J#FNT_2ELjoG
zV@awfE>VV*b=<KFV%eh}GE%}n^5K&QcW=3ce6f=Gh492P{wE^_NS|rbd&jr6MFzaF
zjl24Q{$Y#jTh%8YYp5G8KiK|G=fuVh@5buBXm9ax&~R?0AIbT-{V~sr1hdb@(a*_`
z`c4R><ixjAZb&M=l9k1+d$um@$`;YkSuUn~#SA-qZ0tp^-QbzD9}C=bS)TbrNHXP%
zr*EWuCv`$Trlp(>Z#>od>Eag}`aO|1?3Hu^qb`3)j`E^p`23{njgsv-9*4}m90}yN
z-|*jcOzVnjd2=DIc&jFlZjJu6{b7b5%qzOXf`;bIw9ok1Zx7$KTQro_lWWTCEyIqX
z+AF<1!gDUEY`@sgIo7!{?GG`3#2AuX@?`kj`|D2v^M~KrQK=Y=rakErQB2}lzsZC;
z^uePm>jP6{6<yfr&xgJF{8;9E;S+l1%3ou1`c$dIuz`~_u-|k=+$&|+Sid6iZc_lW
zmdk^+b&KcqM0up#&z#^iE?_VSx82GWO%th2zI^AZL2_R0y*QoZgdI`LWL)n_Gi_sA
z?~%x_aj>toU2k+CfBW9hZ%tzh>n=$a8JLZkKRIeXO;PriKc(9zzcr@HG$4Cy!H97*
z?Za5Y$?&IM%^kNIZryVCrlfEeGoj!-UTibrZ$@iyTWQOl#BT1xr!Kx-FaLyNa8cmQ
zX6^E4pJ&Y}&ukuG+`s*B(e3Vg1!E&t0$Lmztvl|^e%=@BQ_4NNPSEu8`=IqU36;-B
z6b-81S`7DY${`I-ud7*8;MH|L@X*aPSEh|TCe<6iQubbce6dzUu|D-n)rPN>!_7DQ
zLyYQPg(m*0*&>)R<#);A673NGURtADs$-FAqqG|PIIo}oup?(@ZJ<$D!4I<>(T=i9
zVbhQJ);g|T@%$<Bn~eD+_02I8PL1VIPPujBeSC-gZAGt75_eHWdtviH(-FDe_@3Qv
zn2E<RC{MknH!$=35rbuJ;~7u-F$a@dOLcFlOr7T4^};D0?bEcAl^eCWY4E{PykuS8
zk>!ocM<4IwoYcr#7cuR|P~kPX^&`H2O~mx%t!x`BTeFX%eDZ6jB*SOje@07vdS%#r
z_yWJX`n^t%uj`8W4@D>m?uzW*qx(_+R$oxp{Uol-#O28M>b;31E{A8(<%;CA@JjIZ
zuQODibhcSo#BCSmm6upfS5QjfH>i8e;rxsJi}-~xHHCeSqW9Zxjo1ggxT3ksc$@H-
zRJr_vsvWJ04^@=unT@E!x!2%7sPANMdH6kO<f})h!P;|ahL7FW3dqQv+?CBS(6m;r
z_rp!%O`hF6z@pxNi{r47ezLRdo?ZMp7QP+nBOHUaGFvqAC&DYmb{Khf*HEUg9oUq&
zZ~M@~+pY+w>nxEW<0~}>SxfxOUCGPi&IM7#7nZ&0PYQA0zu~QPzR+CEX_9emQ`ik9
z(Glf+Q#seaFBnWoF|0d2@b>M&KDO4JdW%G-EqdpVx=w{PI+fYG2<J`LaL746Os!C+
zVW(}cIN>QnZx<}{r3op+y*8H-@{g}Cb2g0}5v-rtlRg!;zk%CC-NtKWe~wXl{PlGa
zE{;M8Ps^-&H&Mmh?{bW#?ys2im2}FT&8=yi6v}jcZxpU}<bN^tjzOXX&9>;+w)fb!
zZQHhOYmaT)wrv}GY}=lBbI!dt?u+-nALmbZRCje%M0Hix%F4`@Yqn0d{l1>)niCYS
zvNwuDO4&qcb^nG2oeHh$Fo>mTP-&RYsbL1k)s`?f?<osV42Uh&5XnLY`HPFZ;`ac|
zO7!I!AB)Q>yBfHRP#P=jEki~J$r{xL9Mq<=XH%2f)GCP^cghM@RGwiYSrM%=<p*n`
zkY>salptd)gj(`X^hbh6wcL~C=<~{HUfgyiOCGOuIEY5i<~p0sI^MlMTt7T9(kmy@
zfU#=Eg}rK+d<l<M&!zuulsL1elNP7j@fw;PI`;f&^I)-^w}HV&B+a<wo_b0Rba*h#
zjoti`d0JFm8EGG;Ec#caGMY5+E@KMmwB><uv1#Q(b}6H@?FV(6DhHV3B<(Q%e5h=;
z$1tKF_^OyKcAedOKyi<u;+C?$^Owc5@jY8M?gZ5yW=$5ibN_09nMJJ4+dP}J6U)g<
z9||L*YXL9m>!?Iv-J8-J`axyc0;cXmG>or)Z(Q+d6M|sRe>E_$VRafJwUbLwtE97$
z2qMVo87?yQ?*iBf!S!&?PF<~}-iy#qjMXCN-0`nU+sNQQg4uX$+-zv;PyP_uE&Fd{
z9so<fGl}8J`wSMj)7tFwF%fSYMUg4aF|8rlX+Mp+6uICiP>!2zVlw@ovrK+c(&Fxs
zrW^MG<^7qT?Z+P=`gYg$y$WnTN5$-oqrZ>dYe6552a88eS8q|iZ98mPHz%2mx85({
z1h2<t2pPxN)rBR74mD{nY#^+Bb^W1{v|fs~zE$~~$*9P)F6cSZ-PiSZJ@PP7#V0MH
zO3zoL;;MMEYQ{4M@MqS3pGBRPOdNaJa(X*HVl6-6$sZOe&=L<^ttp;9p5x@7OF}x3
zfvy8sW91*uz~mi5ylt*Zz9;VpPf4rNl~_C<RKV#Xrdc%rFm5R!ku$cgDka<Q=jYCB
zgMF<VhK7z>mrkRH4x}U;DiDwElDsqh9<QK*?kheXlDwN0=rIUUeV4Tu%NKjBdScs~
zmC|SzwKLXvO1^uAI^I`XI)398X#NgaZ-3C4;@JL^>d4IS|5F`*U1^#AN2=qn)}-wv
zE6mmj)hDnTDv1I-K15tPF=&TXrp~;IUgow)=h;oF=X%P{eVYzDlhGuKhcIqJ!aye<
z=eEy>^N08Qu~5YfeWbqd%bOSYqyM7buCtPn89-XfoR9C7ujjG0d55ht#Y}P0<eZz-
zuL>s;Sd3<03a#6N0XiGx>KqQ=aV-(z$?(q6E%#EC2YtFgJPO(0O>7Cigr1iAOHSss
zFJUp0^{n?L_I2^~a8_Ho4pL2WFyh+l?Q;*G1VUPU%FjgR^73N0btAX4su&8Of$!h8
z5=P|jk8BxH#6|j52oWdPF=kGDgbG;LWspUS9t9UdFG4tx+xv4`!Nupjv5|)Z%yK4j
z3zYzjG*2JhkOEP}B<>eYoZ1RA4#?wi!K1+pzMW!aL4g@aIp(o62AXtNV|)si;8d15
z%D|QC7ZXewfD`|`6L~>kP_6x4JhrvOGv4GNylW%7gG)Pv%CxR?B*;A4wtxvS9@_~y
zS7)vvRk2yrgl3EyEwRkmRt0`Z12kfXXgt)vd&*3VXPJ0iD);T6ZYR)28LWvc-{=W#
zo4(g!JAFeX3-JgUG`qGZl7Y-jQZbL3fcSAVKMnNGVJ2jk(laa0oN`D0t1ElFQ1N`m
zU`d&E`Ux)97GrOJt$m$tnF-ND!UU?E>Wk3dfH<v?)OSN9#ppYYid-&5nXk;BRz;A+
zx~Wg<WE1v*?1in<+yAA)hxe$RGuMwSpXRh#lf3cgV1W`7xpd}6vg&JL*S*@~Y5*oC
zb?AVXx*iTY;U5<y78I2YuF5ZQ>HxkcCG2!R7LfbUEOLE4-<h%VT!B3No?rBX0PW~X
zEbkS^cy1apYX3?b7k#flG3P}h^PN*7$}J25e&hTN32_-B;BE$f`{MGE>#3Hup(leM
zA0F4Y_(CXX$>421AZ$3lOa$zt#Q)`%k5z<>2)_w>a<#8D#~83);L1DM>6&(IY)FvU
zJMWngFXJAdk7DLt!xeOS48^Dr`h#p&=sqb$sW~LbXn=8}Ck4ka0eQ?6F=yh?VSN5U
z9(((>xwyF0!ICPEm9CuVth^ovoi$C8R9GY#&Pb%<M5`dBIZKj+dKlYqNZ*KudjW_u
zGwg+VCB&)d?gDh7TYjCk9Pf=BRjG9l7&CgGbHD-5X*zecq+>3^OLws{Aff|uMs73I
zN%B*`z>1=3rK9ZRL68NzSDo_8YL$H$D8XAlgDwrbBF@jpz*=>tm6!`%rsWl~y2;d5
zD|SFP#<>-H5xns5P?Y-ev}Rm(j&|OgBXgkJYP#6=^uv?sdZuft^taaI(z?RgqtCj6
z*mkO9q9d`f{JhiWEPtv^#aq&<qIXWZ%hA?1!qb;kEl^j4rBRg_LStDSniA?tGh%c#
zP?AtdGP_(5F7jV_dfwFHK1*{RVFI)0V>ABA%-;-pr6+=pN8|d>m@Juq1Hux&QNMjJ
z?I9SF@L8B{R#Ob_;~dMJ+z`e6oW{k=IRlZJ&j^wQe7<IfVrt&$vk&>n@r+f`>P=o#
z>>dE-*x|ud(6DC~P)SNiY8as}^g3-mC$GiWuMYS@Q!O@nVpB02+JQKtWVm)?uGp1(
z^2@=xwG|hb;zIU;D;5mN(li%Ox)3rP7xa*;7F&IDj>QQcEFt4DSjf~ViA6Fjp)@?s
zW~qt#aHGa<Y<A1hrp_@`*kLP4gOD2XHY`HTuTQzM`n5w?;(11*Jwo#4Jjed(nR652
z*uYNs5mPTVdu4|54lpqgPj~9REXOq;dDhG>z~U&5aTa;XnLO&ioQOFtZ`VID)6cfM
z<c7MAurUL6Rgs}2JR7o|@RO#Qt#v5Mk>^as`v!;~6-IQbwUIu%A<1<y?cY-Y-ah?x
zBX$qI0gmk>p5Y(h-e1fEBisG<iGy%xogrE@3Mo=FqAt4e2k+BG4dpErDhuB!Ejp4L
z4fSY8w02c2S2hr^Ftl7`1Wg8(_PTr39(IKZP4%q`%{}fSOtmAb-L_RDD`;=Y($b@`
zb#K}u!lC06EM}X%G2ifViehddiv9F}$dmiT7J5V^q%Uu_JdzJqMS_|oRq3bpdr-|H
zqN_7^k02~APxJQM>#T$ej<(ey>N?)p0Tyr08{-f9g6{>Wsk1>xGS78Hl|)9#TN+G<
zeQg<N&Kj30T2PD?*U7y#y(VLg3M+PhgCImkarhp0$%S>JAL8-3XFFFCaW8fwSSOnx
zA%7W4MoV1;v!aPjT^Dm0;}M`G;m<(^opM37MQ*ba7}x#;cZ$JJnjyu;nU`GG50Mm<
z!l-JdHQ931o~;I03P5Xv_}azYXRTjaxqU#j#(L@O7dju;Sn9ZbBtV0s`$W%-+Ksv>
z24b=-pAF4mzE201$45q`wt`s+?PUB@#gJ((UOR&z5zcz<#o?{6)jR2w`AB9FHxtK+
zG9Gvudd=8B<;l&@;EON{0k^rgs7pp^=APCftI5LA<UVh72eW=|tCMq)M9X^*%zI?#
zH7#PYFyoX}<|{rUU4QBdgh0Qe#rKJ5xQ_^TV*qka11{FbccyuoN_QDfPf?fR7Jz@#
z_MuV8;#Jd=8gdQSRAL~p443L3A}@iJ);zh}$SpYmDoSu>lmYK0F}RYo8vDRtXPVy|
zVAh$}c0{h#fr@14TB79@U;J{b)s?!f>uU?{mKT}G6*zB@#^q-)dy*{qp8w+sn|=|N
z^6?a!z8uM1pzhifYu4|2aC*M?^DQq8<uV^2LxvGpt+buE#D&(8^hBy!&D{1{rBM*z
zP%0IP4VSLeVA)0VQZ)&XAs<$=tMK>Tx?-&P$f`qcipSPeGX>MFt1U!GgmwBiDaAUs
zuhk^PV!@>PD^e<ivFK8UwoY7;wg<XuI`-&q-r0f~^Ys-u^On5C5BntE?p@?wrRh<z
zO5KwOk29q>EG%ET(jUun2Iim8X;mY^FA3>4x_iuK-~Rx9m>AjqFZfaRus5NTGqhB4
zwxN^7XP~G5^^|gSa>nOiXZ}B4RE+ej4F4lww5(}kw;_u3eXUpD=VQUQ>FD<SB@ZH=
z$ceri*x?Xyk!^qxH8o^utPpo9^?q}cUR)KHkE`$~wGIGjTE2JP#r<&U$hu-qmsBST
zo(~IM1o}oDmuIX_KY%eBh)N2g8Z%(&YuIEj+qfI-_K7UAC&ca_C)}hl3Q^sJR#lZO
z<H1->XoXGH6hcZ2N^w}OAHJ+1Dae&f0uMQ{qAraTR_8z5rGU&g*g0Q6tu9~MC%S<z
z{zfe#KHiuFVv$G!P0<Gpd@*jGvxX_4Y9b-f&=ukom}N${i{p_>GIe6)i{2QF!weFY
zaOB1VS=dkvms8RA>@g%m=DX?@0AUgyR6(?IJJ*=f9sE%T^kXi(#y;zCr+|@L2-qsk
z%&liXKs<q@Dz!%GDryo-2(!Q|<qj**hn2<BfO9uiA`&AdjBg!-3ZK2-R;Q)gm_#o%
zHj88mxyB}|5%z|;ZNB<fOI9P41xdgVnmsIrybdMU2w+z&h0rUBkUp_KA;CW29$T_)
ztgKPIRceh{WWXv}{eD2H2^<Duwrz+A7}7yxH}BXT5*otG?jCyXA%>(nI5t8?AjT&`
z!;z4De|VR|`|u5655OLeUy?BH;7uMfKmjZq-+`GX?ho~mS=z7(&mnPB<zDD6@GQ)Y
z&|#U@D|*7rEWMMXe=?}4prkcV11>5$`T{UTFW{r_L;#2?5fOkQKZ^sWJaim2AMj=2
zEE5qI!|A70u<+h)=${BSKkAD8GjCnC;k)IH5V4<(8NRUnQBMzD8ab46-E6cUYF)Hh
zS!XVH9hnT|>aM69t6@oF=~8lqg}|{1drfZCw2wA-h2HYWll{&^YyNg5<|?pb)U})3
z-Pvqv03q?vp&A+hV`Ilw4Sou9p|U0rdmxOQg^|D$vM~FBsoKb$?+3jijP_+iUzW`5
z?Xo--{|j=u1CjtHFmp?~XsiMW@|9W50(aHf^P{_hEaic^2J)DRlJXQMfvek6x#`G*
z5^Ku*4e-#XuKv%^sSL`Bt^or5^6{og+ycw~v<p3cWqh^Kpx@xiLK7a2nvBH#(>H^=
zlJ})7TavRfZ@vCJ52fi$)CU_@z;X3uwMmI)OUR5r-4!#XN3l<(S=_NWe}zlk{leMO
zzaJcJ7SyhK;!^q8$F+bv<6FTXN_E0Eqo7`i9<fmM%`g?J`o*{9$SKRyq=+Q!uFeF&
zjfWSfFB45QX93gn^m-3_cXXJi4WD8Q-1%&1YqlpOd^EbWR8J<)LeB>ic0`d7T;4ul
z*g0R3CV(%aL`iE#n)C5tAtl##=8h}MQx_duC?|#uS+kQqKPuFh{!|{VoyV#u>C|wl
zGTujR3*w<S^g=%1R-@(OX${)ageDNz&D+WpHZ}dLxCVdF(#O*V!1u&m^xS>J$lLeS
zf6d~*2mqfKZp1_Ab^4OZ(f@)bM*!~r2F30f2;nSgG3;&myj@?9@Mz2acO9)e5?qa3
zE-TvN*lhRDIya#T3z@aEk{>TC<HdFBMW?k;+l{MbfaWhOjeL@WKOQ&euY!WZC2Sv>
z*TI{!NzUCaHn`}MoN#e49zBq_@t4dESQdsA!K`;p;nYf-CFyb`v3K=ru9PAF<-B&p
zaQjL7#Wd;S8S%OM<DRpuXVoET)ncrDX6JUxAbM$=hnB%DRw4J|XtJT>$p0_-&-25^
zjzh*s1|6Lfc>j*3{T%;Nt`7@rOKM0nEY|{pHydm}H_+>K<f+iP7p|xWZm554Ve_}P
zxGLpCDhJ9Cub2oEss3F&hozm^$5>cl93nUY(Q~}PEz6Wo6?oXBjrl91A#TJttZWDH
za1rltkpfC$56Lh}=%&HRwi#C)cOD`_V}+X3W-~5MtPFg3=2#}4BGd(*Vuvkh@e)*v
zONk9ZEog^SZg^FScaX!D8R6(x)Z@tolMc4$mf8BooYuSaQ$f0JN@HWo&ZDiWU)S)G
zPiH3{c*5ZZ9bo3iQy5zeD0`Xk#>GvCem3lg7kVP2LZ|-qwgH$Xu>sxcW{(I=w(P%B
zdQ!5)bGCFf6}Su-#f#3`8sI<UudLGd=zySPGOS>bE=yY8jtaE?sO1N+Qa2;vumYNK
z5oBl1C^I0K{9OS>szkfoR40wSsSnG_H32-;kwhV5mG4~{zhpD&EJqs}D$ZDm-c(Iq
zrKgtF@OAURt7lrQvs_#vJ<s<CZcvpnv6#(xgHi8HO!wJplb)K<Mm&FtT6OzaB=R&0
zvz5mX-L*$BUPi}F){Kwip6-1ZV)o>($|TqS1=9x*T)Jnrl)Wx5vWfBie3Y(Yb48kJ
zs9;nB6z+FfPtba^_;jz<SJ$0ouaUnkIJTA?EWGSoGcUavZn%|_Esr|kINaBJj<H?0
zq!19j@!^}plNN)&j(jydKMXr}30*1GzKpQ-5?Sudovf2F-RC756Mmn^ASI;~N~|Aa
z>zO>))>zW=$?{2~Uy9dw;-U6xC!esNNRye})EvJp_jaqNL`F8gZ)rFwWa~_q{!lr5
znpZ>%w%(hIr8;Q!$0y*uIU(+3=Y*Iu-kb?I?Q0i*A&B(wtn3NY7;*X}U#Uu-G*m7w
z!?ahXyI#3#b*K42_C3bFn@*WA{j7!pg6Lnueugs@`Mg3u+upcy)nyw;^?J6dz8G;7
z3Bfee3N)z+(+D)V#dslu$WksiIy+DvOG!q4{S(bfac~58*hClV%91NNS@u@UHlWj$
z<!e6r@e|R${P_>CgPEEA|H6*{?V!!b%=ABV(7w{-ial&YoH<b=ZwfkU7~6-&fSKE3
z^64|twR2?GHvXW!1MB%U*Q=mNC@fWy$e6(~^CXNJF00uoHhdH&ZpqQD-q7jYU~!Rk
z0hV^$^t$oW5r2V7z*nE)_}5`|-0*ei=@l!d)KKY=7q2*Tyzea5dVCc0FG94<gWBog
z3|O(m(7Kdfq?qq!Hr873@+?Ns=1n4H=Or0SGQ*kE+}ic}{Je>fmT19M3kf(ghTg9$
ziM)%*EkzZSiZ{Vj8<@im{wzWmo!+0}&bA#fc%bbVkh>98P~$4?lSz{(x|;ALV0*x(
zpLQ_+`H1g2?$g2f3#{fa!Xo+gbZsEZr>SXBWmUY&jI44@t&p}_r>aYis|n>UM_zaN
zN4nfTox#{vV_#|jxbZN~IJ|HwbN>U7hFJ2JbB?}4FS<C|ofssI74A{H8LG-KkQbuK
z5)yf?(j7w(<p@N9`L}+)-08C_N4;oe@FV4H3<e_5A$V;NsA!bk(V0md^aulwiI%)K
zcrwr4sZYzuHzX-`j1k?krMDkllDFWNT1IeF4o#D7_n)b1&{|-u_yieW=UQiX?3GaX
z43jJF*_STWS;#SAF00Khz*%FDi8koz6#^#>cRc1qhd`03*2BoKvS5pTx)|8AX`fs3
zF6xb-cp&Ba66+o6)XD7uJhU50XM2teQ{JiDq8P!uag0Gm+w|Ec&FM&*F8jw4Kih`|
zbCS8!eGMgDK@;c~!qo9~iUYjp*lud8yLcSlF#Y7=^~_|j0Q%$CUm_S68i2TCk+^(R
zMr;kxb(T&UL^46h_X;p2yd(^bRv#oHS&rDdS-I)0(U;U8X~a~`V1O+?->zF|;(F%4
z=yinV8((j=*DtBP&B_;-ctpibD`^U@;1c&nW@l;|e8Lk$#0z!zC$oGV$K$P(&_W37
z=f6Q%4;)(Y%E9l(TMEJ3&AbJ-`6T3g$Ej?Zr@N^svu8NG$Ii*-KHb<m9tMF#<z`T%
z2mUzQDEQl6i>2CPA_y8F@01dTvTiTSOUmZhBw0{@Z*YF&bvapzGF3F!p(jpgXgoA>
z|0-hmyP}bKBqw<9=M{ZTjB{LamL+QLwX`*2E0hB_V58c*?CmNcbfdj<5-AF5Wy$4~
zYgMI44valjOGL1AywQGm$g=%?kQ_)iEIVPYF$_ISSEB`$msKO;Ozx&8QXz3~I><|x
zszY)-8SEmk;R+R*ls;n3OOgWje_NA`T%bRl%)G1JP%S@n-%jD7B=N^_U@n#5Y*n^3
z<ssFYW<Or=6Aue3Y|4ZtX8Nb-nYj_j+i_D@Nc#7HyngnSy;|Q95h3z@O|XyTBKAsr
zL4~$z)S{JwUO8iReb+{lzXUTmG*cO2be;rl3N>YNg)q90Qg_=cH>PnbWg6nw{w=+S
zIO%XzMA^SlfyB@pSb$}FM60IV3RtIz_|W*`Fr0xZpQ*khLW2zu@a$V<&2I&&U^NLQ
z+uNT-i#=y0u&fAZ$D4U2d9aQYG^eObN%UT{1GB>PdS?r8Lzgs^7pZY&NatU--AYB8
zm{W%y1yM%!nf#P!GB%-@@X_qz2DV9&Us1DQ3SJJCs<pgI(rR>eQYE>q3Z3;t2*RQT
zChO(jf$v*{kne~cXnSUZM#z~$H(f>gxM4w%03Fp5Wye+BaUf-98WC|goeFq!1fc0=
z%u$hKZa`VRcJN(O-te^4$^0$bT(($ueWR!@Hp}nEPf++0G@ob}I}C!Ki*xSRO>z@x
z1THYMg^;*fuT(?|)Xt(vr4WX<X=Zkw*JeabYVF0QSS0&}NENCJnLre<$A_8r`}rlc
zz)V2`XS8W?)0KSWw>XC8P)-5&zi@8<PBn5ATpmEK{8!#u+E8Ok`u@Z%ft=Aft^gF!
z76$EsP-jVkk!F-7Yr1|P;a&wD)`kQvCyeA)6RZ0O;`v0_B`H)CWtHJK#h@!$p_#<t
zOxNLJ>hwG8*cc>IE4qFv8x)4vZg6ER?<_u%ZW7Pcnzp%UswP5ta?7{DlRNaTya}Iy
z9j^nr4OXu85H{rSk4IjrX0X8!&OL+v7V^MEniPx^CC-J}#_GEWtrix%Zey)_u~*^^
ztlBmcEm*%-?(YU0mQSBRn`jDNLoHbNm$ia~g`7Cw)6A6!!92Y(U$aOA5}smUuC>ad
z7%8K2LB?=Eq(7?<k3Z4|gE4uA1`ii@4IgCPpXp#uwqJHQ?}b%Foqle9CyZu6^k)=E
zC7XZiMAcoBSj%Ql$Ki;h;JPbb65QjM=9a8aT^fo8^))*xO|T$Lm=VcF$wv|I>QrlR
zp}KwQ6DG}ZGjK_g*AHSxGQGPumi_=d<OAqY9{t1Uzu!#Ot}=%UX4A#fW;QYuH8EAC
zw8);lG{>2bRV%LoKNS<zPV&U~f~}U^<`|w`smLGOGc4P!j%%RNvgs?KZKH7v*cm^4
zmKa8nMD8i(k}(sFovsC5B%KJIR~OSzD7>!nfSZVT_hekrA<s-C&txM1BW46_B4=0R
zgEWEn7p4u`#0$y93+ial+&a%X==L-xxM6W=H&chaGm*SgiTsb~Z|#G>FW(a6THDqw
zo$}9Q3eTnTyip@y6Q^aJ5ORWR@;1P>=o5Ry6MHDXz5e6Z7~Q|-z%MKUTDSpvb3{6v
zq=<_RtPiL^m#RMTuL_hWxCDChL^`~nh=VVz5BN1Vt~kkq$-$n{#@Uex^kf6Q(PL=g
zSGtEd7JzQc?AO=mDYTxij?=n)T|Ex~@303!55P9dKaC6W(pd67nZtdf^GaYI(`Hf8
z_~xtqSsP{QI<r{5B*QFht>Jm`e1wT;?TF100xUHQ72{R=C~_jFm&E|ibw*g0*W#A%
zUi~0`Kez>SYTQmw%yNd<Qq1BTA15Q3K9%Df-y;|3unoK67*im%MQ@tm@r_GdVtHsD
z)m^su22$(~X7f;YahP?g;JD<Qp8g|&M=-S`vnxC~jrUF);XPwPS^?VKcM-E?$GZ_s
zZ^aDVx#NvhPRi@-tG7W&;f}O;N+CbM(^Ls-JWH82yR#gl$0p%MT6{x#hi{g&0QlwP
z+6*Y4*YdNoqFY4}JJ_4ADvR?i5zQk`Ggr;_ytnK6e97J1$jC?g#aWTF&i{oY9fSn-
zDHN9x6qmsihllEeD}Pc>?W4GO9k3wd4x!+Mpx^~IKiE(!WSKp8OHAaV{LTkb#syKv
z1$=5SP9`Gw@WF?}3qbpy&>RcX{~yh9Fw*~z4vMa{r(>yFVS8`%_O_Ao<m8Z^e$TQ0
zZEt}eu9(^h$l5GO?c}#V?$EKkV@MEr6=8Y>fqJGj^AIU4#s1>Etm05V+|#{{#ka4s
z3+%8UQc=cf_?zB87`%C}sR~`?hXPEZD0PD$FN1WmNsfP4G3f*-0Ccf>u|JbW>H1(=
z^wGQi-9@5Za#)3;`wd<onTHjs&gS!N1IM5bN)g~ra=0Yb;SScP@qE93FCxQAGkqGb
zom_k@`sG-*peC}Dxx9|PkiGB&5b3`PKn@h~eBN#By_EKnK*ckXjQ9-p?m#7k$crgb
zd@jWt()BJ3^4;yp{llo@_;j;;-(o2p+6VaI{fv6FfNL~>U{aBs05UyPQj8v=0el%7
zq$Vi_GC+mM<^M-M0DxAj7Cg8$s`cpjSx^R?-qT@+(+e4l)cNjPs`Yieg12dX4w{EV
zoj-7$_61d;i1hi5uUsr3SplumiKx;Ey*%BhC!ms7K=d`5N6Czs*D23qBmC6#zn-J|
z(@)|BQ<4)loe4g5Aud1ADznY{f~w3wg8#O{;mf%3_uPc2B2}l%Hs`nTDJ0rmRvp=@
zaAWMM=mhlx>JC#e-AzL5EtYDCRk%(f-}sUJ2FaLZNO{97f~H8dCMC!yZesW5w57r3
zIf1ax*I|l*B_qZ3e^O<0$hqQ2UPIeTa%k7(80e7Lo<-sm6v`{SKiRx;<zjhz%52T9
zB@5K%IgD5_@THHOc4y|&PAvu!U2o}isd;UuY}p8R2U<^+m<|%dkc%BLZE1vP%292U
z6%u;%b$P-40f>_Der;gG*}b(@F3Kzvalk^s!6+4iwP_W%jPl7D<VC6qE>t@K+~ESV
zbMvj{ELk!LsQ`qGa`Ond<{9b^;AGEF(+ewA?5OOtm7_1(=r~G*IPI@ySjBBj(F-Fn
zOfu|=Pvp`tv2o;%qQMQ&5p*(19P4J&=nYO<0ple-pg7<o=vlklMhv`dHsOsc8)l^0
zavHak|KK80%-e?~ZA-(FZZ+kdwdmGU{QF73pu226BpJNA6wUy^37fl!Li*;v`)~<B
zV1@wD5De+(0^#Xbm<8EV&yxKO;JrnEso7tGXZOE#%eeMKl?G*6F6H|PDqx6UTa>L4
zQ=VLdNajE&oaCfITN4*Mqb)39+q5+P`4{2Pgsvq6_4`8|^`XOIqNc>2@CPQ2&c!+X
zL+aduQzN0}gmRQ?<CHE)WQy3pYi|$%mxUSOMcvA=ff6#vd)g22N0Ly|hKru*A4Vzt
zyF*|c-Y*)pZD>E(eU?(oJG+8hD9HS@@)G3gez3%F7yaqT?1{kxkr8N>r}66jEWv}W
zIT?2-C14BFY@;vw=4DHeD*dPQ^aE-q-uO@{$Lzbd(IE=(3yUmo;AR()Ea4xTFhuF!
zKUlz?&)V16+HAHyjp0JcHxvW6wLQ})ZV&CfQsfLg@yd^sx5|hFOTQkA>mEtQ|L{OS
zY#T4%V7bMfx<c~rqhxyrkEx#F;`dkC(1#E!S!jDK$05*}Ogdf$_55ITggc=}X)7-8
zQ~z{|j{9HZlHBZzS!iUKvg(J{w_6Ma4^OB2$n(mpO9Y=yf=6Jy%U2BR=a|f+dtu=9
zzOgb)@O?Ztl9uobgxu8aIWk>LU092GizHywINjyp6wwFfQJp6gZ4yYll@qsP-l0lZ
zO162(Z!4K^nw9r-m97&=JpJz%`>c5E+~Nrl$ip2V<!%s)ev|^XVBR8e1EG7dh($x$
ztk<I<G45>Q3*M9xwxYjX><V>{iYv+&P)FfX*~JxoDJN`3y+#nX!<YUclCTX2V{lFC
z7Dy<SBN?9#s})ap^BFKA9*3FNjKR&di#vQ%l5k|&r%EXD&Ar8t=+GuoG7@{nCiZyy
z?`ATHG5Z;w4|PH%9*Z+@HD1do0q0<<#UdK^oRbzY+qq~?ZnKK}l4_v<XbpAzAyZ}D
zo;E}J*N*mQ?$T@d#A$?DTqiJd7WQoKn7<Olc#klJN>)zpW3u>WJq&VZCq15@fEgmQ
z+Ul6PEl{vN#+1aT{!*8av*(s|p??02GTJ3=iI&jDTJZ99*0XbvHQK5$9h;jlL!hKA
zDP<-_C@$q*=B8#m(lxDNClewu&>&SSWZiuxY~-it&0;JwAZ4zchj>D*xmTf7TbnRb
z-dES7Ux>z#>s9`A8Lv~h=a>>gy;9Ffj=f3?n=SPyU*a1tzSgs+Io6PQzxaqTUtba$
zOFw`C^jaE+%B}QF0ZVN#sE3x_B2Q)BksLMD<>4cOEM#B0j8f7%9`G^!xCDd&fP&e~
z9)Nq*qbufpC0pc_Z}gq+RyvyTKAEUcbWrI(uK;qGre$V#<(=^022j!pDNvJ(mm&CI
zR!!icexy~^dgBl%0qmHrjvJQ#H<idGA!YTbtm3|Qp$b4;-_!(@ieGIEJxE?!urN8!
zh=-&zL6juO)s3o<-X-1nJhiyOVM(|^FO`XAwWNEjlHMxQY*-HNNkKk~c#>ElOjE<F
zcSI+s3=&?GgGftk@w}YuP#Jzd-1WxLQu$tYFLXZP4`~?H<)ot_;?iAn=}FI}ON46K
zM2S~H7||hGY1v&L{5L55nV6-*S<M3?^%~!Yfxujj9K~^&(7yB<amr;{qYM(%u)U)-
zBT0F6`S9J>qBA3jyQ*Ab&9gQ`xvZhHOoAmJm8D!bu2N$1vdV^)-Ii55T6fdhI3~{O
zBkM0yDMy{FTvz(D+-*#Lb0cCjBQR?D44^v(0lb;=yke1EGSDbc;OT;;+MY)Qpc)Pv
z^whol7n0VTlhq8b`7A=b21fpXTiRBI7SQG?w{rQw&Bo~nJ;jMDCL#FJj%J$)fecli
z{c=v<ag#$(?5?C$W4w)J!+^`W^G!oV%NVfA&GSRw<WBp{c$bBg$qmJ)K9eD)S}BKf
zHfyZOKgQVchP;;~o4<N)!BBmWCiimG0&c32TE1p~<Q1;S1>6h=N<W!x-O35D3b!q#
zU5nk>lVa7X7hgFmbuxFlFpeoO_A|&<V6YYGro#`t`!jL>@Fj_}jg+cQtkWRapyQl$
z>=rzXv)R!V)I1eTL$|3I3?YHvOgy7!Pzv}*nq-&fbRCy)D#_B&GW!1f-bolVFUQ^B
z5>F3cf<O(ceXX-88q0u_{8~7bOdlUf*MBag*;dk7nZ&lM`Mi)kU{6p9%thMT?dt6i
zyWY>b-@@DtUq~;cdibiInO8?=D&`8L^sX%6+;7M|Qv25zaluYo3=UH0n-%%U{4t{D
zSut4s%*T?1$(V0W&BYb62YTb{?pWgMsk+8HJDPQCt-l`H^69!wjAC~?nuvD=8I>W!
z34*X{FkG41b%u^*#P1kc3uBS{)!P|6nb>U19Y`Q8J-3Chl@h?3VwbKFM?RpiQ`fh$
z?AJM$jZD5W`wmJGo=I;jo6KNdlaOUz2O{(mgqakoDpg|h>!j9<rfL2PFv4Eo5Z}ym
zh=-V-;CsX>thR26Y&rX^cpZTMee8c|C$83E=jB>WhiQveHSq18%1n9#I8xDa(rsl4
z51+TFUXL99Q7xB#c6^ty>LZVT=$*#$!rosgE4)#w3hLhfs*Ugp4q_Mrr)j!unp;6C
ze~Sx9d10MP*L2Qk?h?zrnmS%%6WunLbF^)>Z`JE~YWd990y7HBXuKPa0c;=_4Gp@!
z<QZ+E-RCPy++!3PQFXb-;LXZZ$u!_{g~@EUt<kw++%tT1DN$K8`e21|OYHmCseC!J
z$sDf|K~B|)vm6)kc!kc3v*wcg?<`umlr6!$>cp>L^(cBNyIY~9>>WpBJAW44WPO=U
zcst)N3}&o{;2y5<#?0<XD)J!q{J{_6hz9gOqW1rn>G%Jf!LqQk|Bq1nk&aaCAv5~c
zODzOf9LB+m5LR<iFeb5U*fxRcpIF#Ne^|diYAV_s672~W4M9CvXzAzWGf(xkS5AL=
zBp$9`V%&DRVeII3yp=ngrm**nU!Jwd_UYU|)ox#lS+bD>HAp&I<+kK*I8s@4g}Rte
zB!?%5bA3LUYj2Wn@8B3lq`G>XG`m7S_DT3gtg_bUYG-<K4-ecVN2G=(M_dDwK9>}`
z<b<i`d%r)Qzo2tDY4%I~XW{MW{CYN>H%4RUzHSyL&%CNZlyqHFph=sqzTcZau8u;j
zpvz5WKF>;DZ`>kDc@F9&jBpnxve!S)%(lOf{N8-JG(dO0e0rW1&s_1+?(#nhKNx_t
zu|<8n5TB;2Sxoco4AS;$8d>RG-j_vo1an9D&Iq%ZxqQ^@&&u=$^>&8mN716gcZM(*
zOMfqW3WJm_RlGx;2i8Ni9I;5B8`)ki9c<0iGZ+mLwMv<(J#sv!TF)td?6nn@-^s?J
z88cXaN|#y|e~3JHa9nJOryqNCX-MAbu3c1r-Ww)KzjD{Zg4&`^oy#qhEt0;55D%CQ
znYNtXAMqE>RLipu>$VJBo|Lozo04wdCtoZ!AO4j&T|Hjm<cL*~fv`PI^tpMy_Xgpe
zpZ)yPYAv-eNq6<FUsx_;;n}MwuLfkk5P&uqkk;gm_VVVb+^s-ZOK|>dcjz|fF6=ya
z*xYM6=g_E1xvUy92NxD^T!>-`ixAr92YsS1YeH3+7g~396PJ*?DzP~($sLv0mg{Zx
zXP800e?{BSQ6NfA3pO30buahScT?4PB3`1jb|Q&>xtb-g&d!FZQ4j8NRAVBYwMSnD
z*(Pz?&PY=Ax^P!y6kx1WQFLP}jS`pq%|k<36W*+K(`bs9NQRVn#~}FdK4N4!B1=_`
zJ0LrqvV_Og%<8y?>5#?I9<`T&(f8m=VMdBoT#J?al(PI%F2`V+`h|6^dZTfw^Tq+k
z(M(_lT|UG4b6kj)ZP3qgu<Fo+?1?z)^Y)~c7&a66kGvu2b)O0e0)$E!AqhXwKUWq+
z0Fn@*E&IQVQ?w|Ez#w4=yBG)z@_}H1$8lf{arl7%)yM+!fe=8_A|L|c(BiZ{E#ecH
zs2HpHltC@&8H=>^W>>C*+YH#?xAvwpY0zVwK<<<`#v8a)b&W9ypiQoxjV-EP$L(7G
z)=A+UEQsZcx@LJI4U&v>y%f;~(b*}ld(r4oC_=I5#V5Pq*H9|cYYq!a6ezf30MS`M
zSK*YlDvyr$@BnRV)7KU0p&|LEcZomP<84aRr<a)OvCi7C@`GFo9i}fWI5T4GVKZxe
zzX32Kma4Q&M5v7^Kf9P49r&3P;WHijG;EfQ+8V_xN)2-p+hTO1&1WPGauaiG-4r-1
z>N_1JR|S}84`ElCwM3~xY&)ay^iRLOniI3wGZLHBqOg6~DOi}S@_9M-s1lKLGP2qf
zi#o3??x9qfo3RT5f}~*P<pVLm1A9gPdq`h#|JRU`riK9m2}95m3&SMk1>id@@S`CK
zBkTeZLP(;7{Gckln?Ba!m<{0Sq!nfxb8n&@lZ%sb?H&Vc;`p~tdvB`9ry)aCMUkK!
zb8lKD$W)zQ@pPJiB2WD9<<$K9f?4bTtfu3J$(6I(lyAueRi@lv9EOxCbvbdL=1056
zMnn#k2$MTm<egULE7Wzo7hXH}@?x`vIVd;5kAdvcwUFtsNKf9HzeeYz05S)4^7#e&
z(tryguBd`~=bW>Z)@_tl8mi{}s6b19@VUa)28_`~58uW9SuySW<+(?;oZw(TRiTXR
z7iq#)bDT`-DN(C2VHu>{iATA@-?3B+-T4~HcVoi3c}djy{RM-fDi3v&!jx%F&0bcd
z=N92LW*ZvHg6q>R6t(Fot;&nms3w+_8&w4>ygqVcPPpSD6BH%Y99fYWYYjcTqo|Zl
z+W8KEa3fYjn?RBX$QTGo2>(4ndsq=cNW$=Ud6E9B=;-n{4B_ER7{DGV41rt#Sl~kx
zSVJ5^z<(+npIiV0Fo!TOPuSr5mrA)M-9M3cftIiBQ*;elyau56FRsMp%9Ltc_e;+n
zq@tA{k(ExgzG!FW{febt(NM>;O>A@S(&8$&t*U0a%+9m;heeC-IAy_`lh=k2n5oH_
zDURabcK@ZCKuJ6Jbj;3)nlveG6h?>mac)N=0u<~$25e}Z`N<*~kFG6bWO%@c63;Pp
zO!A%qIPkfEQufaQv5N+MVrYk4f6l6=KJMTmB5Gwex~6+EOeu8>axcQ0J%>$$l5>hR
zGqUS~Ny3cAMgCbK?mgea`(;~B!H`o1taKmRGD`)WN5gAnc4jR#F%RIu9=2OM`IJWn
zF#r8NB96***v)OvB5o1A^wQHvONW_QZ4=ds{#4vJD~GuUk`XI%0T`(Ma-zUI5JZqH
zl;7EkAYfW4Jp=$q2=Whu5KMB8KTg5|ANp5#+yefNI+PP2D#3>-B|rOkAbSUz80U1?
z0`xVElfQzz8WRczT%(ej(-mO2Yh@f0t6P9J<MFg^LVT^}>^f&<np?=@r@O+9Ijysl
zQjQl=$AmU|aYV?3P}=3;FSq&^-D|ke#>3vqorbGtlYajzc1`7qj)gT>6Az_tPIP)U
z>aYexK1U&HRy}-WE95S)HL&uYyZeh|f|J{J$dLdf(6(frnjNXUJ;Kf)Fm+AH{4<f}
zc;Q0=4<Lqq32+fK%~Vf-BRMR^gnj9wMet!jXg`-cr757j39i(Y7hQYV^%owQ2zZQ_
zgA<TE3$&Qi!YhYO^=4JrZhBlpQX=b)){-@eiKW`F4}(I5ANy{$KB1`iK*fnkk*TXy
zP`tQCem)}=^Q76yxCW`&WPKETB$XC@27%}~xqvjdzuNy2z^^I)OCDDW|CjI>yTn3Z
z;P;0B6palFKw?HXw8t6~{1w1WzXJGoz%M!eOL!bNTHecMLeki!Egk<`Kr^7jw_>{Y
z=!$aU-mPY=-nL3*K@(*<vw7ApInQ=024^z*V4L{->8)nwBGP0au`r=E`e+^Jo@S-i
zkdfIYc6wK4^)5ktQ}-;Bb|<Y<p-#?*rAYgB3%ob&VwuB(v`(I4rv;go4cnE*p3iCp
z<6PtOL(we4Ahi42=Nz{N#RJdsYkVR0ZkF{|yU<?bd&$)v;G5D{i&a`zu`u?RMl{Xb
zBtrEhVwhLrd1k(iZpwd|B2WHW#8Tlb$1ls06^D$hXEb1_VY@ZOl*-jruj3SNbbpkj
zH4=bN#mLA9VfY3vjfM#P6`l6{VgI#!)2}H1d#}=`2*dbY+<+^E26+ImK(6etK%h`U
z-=jniSN?j87OqH;34g7y2o~!H(4OC2-Xc}&ReH88->KBRm#?RD+ckZ@fo*OE|D(14
z*JtqmOUK5@%EJ0TLYm!LlL?okNLw#zr{Ea`9?0_wh$=3fID1>}o+q~xOwS{326mqx
zviJf(62)fIf;u#i2~)&+I0!esH|{Tk8&?-sdhq3(lUZO?9kn~0edstpK%=3eZfO3P
zkXx%S*C+5ZfBP>&QunnbyQsf!Zq@X*N!YCc-MV(~ZBe~p*7r#8+_%C*I@BKozV6ad
z+yQmL&fE#y^MbE-(P}2?^!R!l3_FhI)Q96__YlOE0-)5pl-x_1c?I6r4}Qj|{ET&3
zVrp#nEIbGmy`E^ARYEkkoPIVH_p(ROurgE)h{>z3(=**mC)g>UVm=MzsuzFKBhEiC
z3NS~<97TA2@=NFl&9zJ-Vi#l;Z>&5cX$Vendq}bw?E*CiXW^%}PF6Q^iA<~x(iB=g
zo<6kP^NH!tgilC#pSJXVE_}T{+@xI0t>NbWwJ~U_jez0M$h(Mw8!>rM2F;Tj>uQH<
z{GFc6?e)FMAmf9L62k8_pNAFcFO`wNLIGZ1aZrqqoVT~_=Vy|0_ap0x5&`3P?PI@F
z;laF5CQ-?kv1@WnK9jkxAxYeBM)7dWi02J9L(`6rTWm`i@fLnbyM&C<xSHIw--kK&
z0?D*^U1JVaq(>+z0cH$F?^14~jJ7VwZdSc}wb2{|OFL<?-Pj=g{%O}?y+}jvh?kLx
zy&(p^xM+#6`?IJ~LchN1tltaA;~8n?4B^SfvqTEKbDHjlmlujQsrBLTaTZ_#^hlYB
zD9A*$RL)A2q@SjuwiYEH;$a835jB@#Q8SZ?95jw~U`QmCmb$yK*q}aGWplOPqk~q8
zsViq<VU+e9!U0MqTpywsZj=y#<A$H2VcLXq$B8}-1I(0g`3rI}mL<#02sI;wWiEIa
zq7CJx&cu}$r>NTT$6MHo(iZ45a)EV02$xi8A|s^F0V>t;+%Hr!65?Vg%;Z>$aU56D
z6XtkRNJeEO^erM+(-(yUDwBs2H)eymPSWwhd&#yvM(yM4@aM?)4U(MI7jhBqEG1Qr
zO3pYDp2{e#-m=3z(LM&lN(U8U_tlBY(~k*$^G-k2l!-A^<b#A-)v6--w7pxNn61A%
z+PGYQLX@Y)86JHm*}J;?IzByVKkEq8CkNM@H5))$!2Wr;%ZOenUZ>Pf9W#)6%?gAv
zFCuvrmzt$7PkQUpbp~NU>TT=>wez~a5QLB7Sa0t3)F%cc#J3B?$JSg7mD+*UUI3zK
z@x;Cs2GoH_3rF#WCOiXaIV6h<Aizpio$<)(DV9%VR|OdN*V`6p3qMuopYt3Nd!9nu
z-4(!q--*2;j~yS4MB9OE@4py)jRBhyF3?9&Aae9~G?orPKydC3YnLuSd9nC7GCwOB
zWeagS)QqDYPWczIgvKEc`8NOJ`Y6Q^J&I2tcLsEe4%VC4fmUX<J9my%Wr8oG>Pk8<
z28%xG^!hXOsS(bx1De72Gtl|7Kl!KL!JqRxMy5B_!Tm+-C6_)EI=DUzA~<>!FlYV<
z7p-^(c{xYWeq27py0|WZ?y3&P|HOEnz03BDw^>aeNKJV{U)jM|UaJe6mKO<1{loX(
zV2~i7a*39I12ix?MOJh%QaN38e%ZDht*MwA4Q`*g<L^%58JaH%sUn#kb7usTfq|*3
z4gA>XslMvWrfm#5GXczjU~V;}HaBL#5KrQqnoh9S$rwiUrs)MOHd50Y;x;*81>+}-
z6itOyt2*>K5blsqd`BNcxl}x{4IDra3fPmc>^7<-<<cTpipy$ftk}et#b}i4+(4t3
zq)xy5!N=`XX*PR3>(J%OT;Os0K$WRpk%zR?LZu}#A$CQ6!c~SO1EZ?k1JdJcqi*Pj
z@1rqe9{hP3iOqG&dpm1!DooDF!`FJ!_U?VNwHxR`wxspi0<+DpHa>6ip;xndX4xwZ
z-Qh@~m7BdeQ=;|SoEEB^v_V{Rz})6wi+o7Rz<)G-EI@v)bK-H8bv6w3X3?`tYQS5p
zuge(QZK*NWBC+<(l4DG~7&NUQA=&;S6;Vg_!~HG=Eb}@`phE&46-vfj9B`SIpxBoU
zCtn5g^RZl?XPPV-zFMXMy>fH3(T^A*azvRr%k#;X)TsptJ&{~~>mKchO3%c9r<N5=
zko<us8;%@kCtq1Z3nA0E<5H|AiT|0nU0v1qQfVkKOsIlUN*4cL|A^qw{;Pl*N!rLC
z?3t>F>L@@TKlgzuNWh@WRo88@F^{Y!y@8HW%A~JDMlVGcKj@HBsAy7!W9iX0B>f&2
zr}jXFQ!N#TR5u}Ke2F1zaP5%46Sbkpyg$iFEUsXhc*Vddpvql$blSkk2*DOlEgRrn
znZHUexEw0zg1v~hPub}o9@uf9^NgCa*DO5?&Qx}NVRu7{p3Z1gTU=W;2do$)Dkc$U
zrG*Y7JoYGTG)|o(pf|~YD?-@{l_2GC*01U5pD+WC++s;Fdm}jQARvM{_ALIPU}YOu
z<`W;aiqn+uTzgN~i<JrIG{%qcNF|w2CPc;U6~2RsM_JA(aPfVR4YJUKnk(N5jjm!C
zjky$$ltL3bnJ2lm@4eE$$9wMy5)*lb6rbwH7A%9Ad=fxLo0|>9QatL;6}bZEeAR$E
zYFRD4e=Dj+J+U{j_QnhX<tm#=Q$Uv+c!)q2+@Th~SRcY3W><b?L)wtUjGQxVbtx2x
zi_&CPgVEZ-p?-bXpbHAKh9AFHa0OMSGr65v?37a~@(%Q3P13M<(0;rN)oe+cikDRY
zFpaqbXvNM%;H|zQFdbLu!n|@C<2P@spyg4?6)FPgzWHYWXt%-}<YVP;m~p@o=lT~9
zA0THWJ};Ks#MoA|*!uk&AlC9pvIeIZLPXv!<YNoI-;soF;~g>doBFxepz1<mFIx8o
zmN}WN7#`1(BF&b(-7E?HAIKImTJ1mZ322hJ+WR`0OgKx0)qfkhEoQ;Yz*M$8ZZM%L
z(^CuOjOYBTf(MUy#l6S35J0Fzi;*f+V5%=;tm&@K9>PqS{nO>-7XS}&yFdr?fJLta
zgdwzoD)cJ=GZ(pvB}6Qw;{JSl7UjF2dZ8#92;IG*Q^;6x{LsJ!?=Pt3-bHuwx^oAt
zst8M++$d6S_^8$&>3b%p%}TOWk5&#{ncbg8Mupr*Nt^`D$y}&M#ag9yi)E?Z3yNit
zxtI5fdJQk2aVG+ync5)_>&Fv0rD?5621u>P7nH!!n+n0Z#9hWRPLa~C^OSs;E5G<o
z;*%s@Axn*#W7C>O^WetYXyzJT8uzdqaR3(@HOO3~H9fpEnuf-QiVSF?HHbxNun+Fw
z&_~5sdF0;&le$A5qeS1Z&?A|xPl6t-i85r-<H%f6M<wx>s$P4TL1VwE*4(i$kZU_N
zga=fj*mn9<n;P~npO27vYRb6x&!3N&Kxf0@gu)Xo0d~*j-?-pt^f#sLi$|MUzkrih
zvO}JB9Xd*~jRFIZWV=Mk&tNh!K7OwX0|YMQWIUn;Qocga5@Yq3e(_})D0XGnko|#X
z?*}PU@z?u^gf8_1fQnLu`%mYhtp9;Ln4bQB7WUGuB^685ir6z-dwMWsV9py+TSdsf
zFhi{m0)|wJ45FKBSX=LaCg3gp?aRxXc?n*8C+m1>Q7(dZF<em_!Q=hzvA@{a`S8(M
zn}aHo!a-A&we6P$;=B7w#g2!?qd}0SVCebd@&4V{&fGTB7UooJui@?5c5zUM1V=Ok
zn##_ri4<yzL;DtWkzg@5pCObJYbLpm)CP0vPT`Rk?68kRj_2q8^MpGqeY2Ijz4LWZ
z({RJRs4G1pIU*tF_u>8W1@HZi$mh%BP|}t8jU4I=HwgNmDfg4s)BSNEPqxiK1+1`j
zChSy`4-C6qUlVCLAhyXeG*GK8<Q@2W7&8!Za*~%&$Y?S}@fpGeHhGUJiK01?N8;DS
z3J+_{+$1_jeUzW4+iPVWPirB)8)2>etg6_rA3s!X9|dhxJe?~ykI~(bZ(NXv-j9<W
z-+-DbO&LBbv4M@Yn1i{gH*T1rVMJzxlDMK0sgn>V;J)Xe;hs&Ji0iH2%=9uUM$`o_
z<fED|O=(?@X}07()`ON0aDPvtiDS{nDLN29X{M_<@1F+?76CfOSbTY*&hI!}_p>s|
zmvBt!J5($<HReXH+xF)5XaX1g-llavCsnW?)p})wdZZ~)i}NtDdi2EaS)A^8|GSIO
z{FG%eXdn{uS{A`M+%V4tM_sR40HNTr6`U%yCbh5-(hUBaNWe4~7KC`4_^#9ZM$|qN
zN_ky+4Z<`ZfaqKTVa-)yr7SBJ+|WeGaNxlnfolSQ9Peq+t4bNIvA>_Yf%snXqg&8A
zHMJ5ueN1EHTeCUq|GQg3d-GhpisyDW!TZ~$R55%f{}9NggkQyX_Y#1xQJ|$Rf^Qz6
zSvqU2)z8ztUPZvjhmEiIc5`n_4kRD;E^{0&Kc20w+WYdo9%Z+N)Qx46o`fBB!fDm#
z1HWe<c;WDcWUuBoYB61W>~B9A5^PjIjB&b!Ah)&FAItZ23Az;LMLRK`J=Y)I8GDsx
zSES$+)e2h6v7+ZnEK(HN47X0APA8blA}hur$@yXV2q9&87FxN+Sco|rf$x*Gzg!ZC
zb`O#z`PayCxo1O9+8P8k*umYZ3l_3di_|K4eltzkG`H#|?|ZhBbnDE?L3V9&y&eA#
zW9Jl{Nz|?3*tTuknAo<HiEZ1qZQD*JHor`4+s5SNoZItX{C8c|yJ~lJ->lVpz0boI
z<iwnkM}aRo>^z<nq3`Bs@mL;@%nd8>z%*B>N~N=>YBnRLkU(DWq8L&nO?$73U7=}v
z+(6`_LZuN|1rhbYZk<kYp?sF(G$EErdTv@;Lx}J`T&cO>x~Zo<TjfgqG2(1p-qBUV
z!`wN~Tk;on?$G(k=?qlp86pK|x;skfH@R1yZ;Ky-y88LN0K9_58M<93;baAk*1Mb4
zV;yd-fO#y%B<L69M-sXWq93qC<#`?6+ff`13L{tlG;f_Id+e9s_}}*r6<_U7qP4LX
zKZj^Uu7Ov)ld90@&64yFisec2v{XAW0`n8ejQzhoH&YplEQRGg2EjLZ7pqOAD81s0
zS&-DP-0562<sA*>46bgN&IeWsGk$DJ&eDWUv*$PuIjb3q5=GN8v1>(Zu@pK)aoUb9
zFnFc8buH_MH1Bg_70ngAvQZ+<%J`l5t~wjHH6ppr;>9vmo<^pw2iWbdCsTO6Fob`D
zqwyAkQzqbhWG)Sl%5z;B;jt<vPK+PtRyPkk5lU|s^dB~9yg%3`t($U>%Wd&utku31
zc5vX)cR0g)K0@@~+#hS?zQ8tzp3L%UE|mXz0KAHAB9kD+t~%aW(q)4&7vtn~^DbUT
z3GIaOq(sUlI(V~k93@jrCN`Vly?fKR&=1rPNYT?(sR8TYbEirCW}Sd!no?aa-+i4q
zY!!_>X`5W$vRY9YQL=4{*D)Z2z}La6t$iIt&*(FesUh%T+`MFzMshN^he_G~e{8;B
z<9bN8I_L60{*J~#5jvmbT)(gbJIvLy_cb;mm|XQ%7LQM8wb~D->7X~ooOe<**8;H8
zQAFU?BS;*H6_wbW9ofU@aM5~=C##g!6w|1Er>#TqQ*fZHDK=B{{6N^JD4NxwJP!@z
z_r+)Nn5pa4q?zLN7xBAd4V7t;+GRQES@n2u-)R2UP1P~6u4DI9?Q%YN$w3;vXurHh
zx#~;ri-*ThjeTK+>w5izhcu7kV>LOs41B$2Jxy4*|3hcF{zy2eA=$k+NrAm`FFhBC
zdDc-XPOVE}MXxz`APPE{hs&h{n?5O(UF1ZpZK$d<%HCt1(7rpKx<g)Cz-X+%Ij%l0
z#6|F!8dKkDIEl=60ywF?)JUV{5+>krw5*$`KrT?1=-uB~kc-8X=pv!H@uaI#ubyrR
zrAcX;_Q?DMIj#TU2VeXyQTYL&w@ub6*e5uMi*KCx>BV8TpZ~_>9w?EG!qyU?<TcY3
zyi_uzs9>PJ1*B>vL3igY1r*5%BF;HwKDGrj*{)ghJoOk#uoWC8^$n~x0li=>F74k`
z?%gj1P7}Cc6x5qDtfTIXJWEEgI^73=FDh1ZnIfTh+wOcH5zjQZIfs##UHzWCAq<P=
zn(4ThasECDm!KxM^DN5*s-p{QV~vERKx#yHh?sgh`+^LK4o~DnTKPq9cO=ppcIJf#
zKRIjRsDNMuPs%EVnUT~^CF9}IjP{k%*ks6BmiW;{%}Ff-l4o_oq{1p?R!Vm{Rr`c?
z39YzZ`vNTRMic{0IGmwpoFN+9Qi)c~@dhUW{fBvxMCcge^{ZS#L`*s$)#0<HcHDLV
z=q^Fyo!Xxe_Fzv=m(EoXMcfhG=ukL=*f=UU?r|eVtG!p+gL3e>EcHN+GGZX;i^iUS
zptp?nSxMSh;Ve|LKc*N2yG*igcp;TTbF^P_O`OHJ-DQhd*A(x(dUr8TC<A#Lw69h8
z=|<^NY*P5q%}t~Nj>6*Y%d`d!j{G1^9k(kPb9(W&b;ljet@9bi@;lgxqVx=$U7h&e
zuvcS-s7urHzk_)Z;e~Fvyn}h+2-%iLY9{Rc#hbyIGM%qyimg6uSVJV=%b@c)#P*zQ
z$=E2VzpT}LC$0G!qugSU7fL@u!yL?YR_g{wwp9{;W95)MCNKKoQfs9dP13&!0jLhT
zm)N#B16v*p<pv^=t#4EL+gq)bb01G$@If%)y2xWWIDgg)8?KQ?*pN1+`Icd7O)!6f
zYcphf+YjyDpgs7Eq41h#u6i*iy+mIH{H}z%RKE~fdp=zzVqC&(ox5E3(*}r47cdgz
zw*T$S|M0M#tG&Tm{4PnF$1SIB73oAQVkNra4aDvX@8jPJ`&*JGD+9^dGF?XU`ESQq
zGw7<MWdx20C%dcf#UD!q?vnc2r+pj}t&^HSe*h=q5$1Yn7QZ|>;S*f9+t>k)y~J78
zT_(m++`$5S5z(Q2s{Da<b~R8-GYFs!IRpmAhw5;5wX7r@`fNh^;eNm#@UK;M3kUmK
zcfh<5cEo#XGi5(T9SHMF)Davs6oZ=H^Ro%A9LRfmkffcQ%Lq8!=8CC}J`S(!e5&zn
ze%qEny@d49XdN`ExZuESgqDD4|AS@k8d(4Rs`KT%I?f)2aipluyl?6^7v{>f7#&1x
z@w~0pr(KSaFnz&H)JR)86X734>+okTvjeWn6f+ou?b30T^UBuVy-gEAjeX)n)xA<e
zR~471;Tnkw)f^?igF9&#_)_!;_h{XCIx6o3qjxQh^IV7JV6SrHD?lq2aU$KOOeX1s
z?&@{>g&eYORCQh_Z6M-Lpg9y~dlLhQtET0SVchiWw`+EK^YLLC=Rx>ei?e`2O?g0>
z<1GUD6R_nLUbCEcw#@^B0#5a_JVLg6B8tXm7=D2g!aDTdkV)p}4O(2Ho-ILQ7DcI=
z9il8}@9#N6e=m!>-;#q2zx;;@?LOH0lRb7HAu8*Pj-{{Yfyq<sDRrq$)wrNw7{`VJ
zXZ((0ljwILWmabG2QPLUTIz1Mkr@Cs9<f)Ig+7zafs2iq4=a74ZRrCd08z$e6yv`{
z;iMaN?9F%IQs3YXi&yFYsSV&{`~N?!xw-$lHeg0aE*`HPx%Wb2+A;JOTQr@YRyDAh
zL^p&JlcBj0X@s-1jR;f@(ZBC-z9uP-iGNv%NPof?JiNQce>eX4E=lMW`tuIYvN;v%
z?fbkIe4GYxoz5TMry_5Ihx^Iy=ChJXQPcrvPwjvmf>KqCd`cfp>DoKTSNnK-XTS1^
z1Ng3ekrVg0vqn!u!l$4)_V@vFbHZl6Jj+nTJ!58hMd{H*8v?of?|<(noGi5bRu=fM
zF)C~(!H8_XC2?b-1OYpv81V4Nhvlzq0**=$S$9vL3|2N&QzFjB=eSoEj$zh*$qybY
zpS*bsV<k5ljOw{TQ0iKC^wl}YHU;g=LHi2xAA5}U%@G1x$T`ntsCUu{&q$8KotBk|
zKRPN%RKCxXd_l^!H$4=HA#M85WWXrD&*YGZn;0p*mW5caIN^5RkSd~}(S-d=8fZ3*
zL=~prwqBgNx}-HR$j)T-+oPpFMr}ya5S{;4erRa{l2>u>HEvL?q4|fFAutVUrW$o|
z(+DrM(+4e!5ErUfCE*Uu%dKnL2BK`VTyvJSp@1nTr#bo-)@&~)FZE^Cp<PP&)HMsV
zTg#<);*cpyb3un=?GoUcA;Aba<w{Jz1X+7!hNMh0R=g3I&~rO?+OjZ(K*M$S@h~T0
zz&>NHu9MUgV^nAm)^w6h`9|R&n8-Q(3M5BweZ<LPr934BIqZnfOcB%u#LsJ0)`2c2
z{m+k<pmBR2cUirB2YB(K-$ziS9R%pL!C}wA&61?U;a#0Ii?Sy6{Iv*$f0jy@iFd#C
zo@er$8*P!xE=i8-TKxeUNs+6aNjM3i7I~UnfL>vJ8p&L)b2y*cjPpU2P0o|_1Sa`G
zA|;0f)$eD_jM%J~%ZufqHLmr9WLo@`IPPU39aw`y6d6b%?G^J+9^t><3o;umu3i@u
z&Kt$1;OvL1LC48tbn8Frwm7s*<{&Muh1cR2uyrOUCe7BFm4}|kS$DR{jf{Wmjx@5I
z+uv4)EH+Q3fhKX9XOFdyn!`>0shH=Q)!4IdTQESv5ks9->hQK%a0c-MK_@U}&|-5l
ztjCH{WD#MjP>{bk*CWSuYMAe6u#u*r#h)XYo<eaz7<|5)y<WUM&G@Iw1Fe7x<ovta
z$bJP4=Cr5}cqg&&e*ZcFFxvOwO4w2}*qo6ZF!7Tlha4*|N+E?*nUplMbX`24rKw|R
zQ`KwvjsiX9hj?ryL!AA72VIO+8Fv-8ZkZ3*+q1Q0ZE|?-iNUS<w7s`S)!nHs66R|z
zbX;!I2#%8$-!~j<p9zawNO<>Tq42YU!^x*u$;sQUWw&6U2@75BscYo~$-ldj%8rn1
ze8&10l{)s2)qmU=50$kb`P#Qk-c=A%%3=LFpg%A4GxNP@_&qAifug`ZOU4MP{Kv83
zOK37ACQ_=uL}5aBkFU^1{27cwJuWpx4Kz(&?XRZNjASm-lAGtI>2R)#W=tMahBL*9
zS(qQ&#%`x=6I?s%l?VY0!$C(g^7xzG71AHe-xcj7gl=c}q-o>Nymz;t%8kme>!isw
znp9D|H!50eixaVM1XBK*uRof4_l61MWVI8$MpyH#OJo3m*GA9naq(@n)QFq$V$Gnk
z?L)EmLYdW{-u<47ud(j4n$Y<hf^$5Q$uhYo=P}tTQ+y0l+|JXINpqHThP5jx>M4rN
zoZ?lL+1&`7s70fXWa7x?(|DsFi%#otnl3tjZvF8Fv-{N*p?Sn|%h-HtUJ*L|q$_4C
zEidTxihsjYl$c1uHH-^JViDx|QtWw1iIzUbsdrp%ufi2bm&lQ{AoD49533uV<38Tc
zn-ZyJ!hDV_Pq^r)(e1*fX`v!O?$<7FWPXWLxcZ%>Pwvu;I`kzGbT!hJ=7SYeSqJQv
z&yBrRBvS-ULL*L-s<N+6iN{S5Y{z-Jp8ZYF2BFZbqP)9y%kkDcJ{<O(_!jMC-Kq9I
z`g<8ov$~-fy8Ang39~YQE9*E3{#Z}uQ(ev_Z{yis?fEt{WBoXhsxvS|R$f!>AyB`D
z!YE<FIJ9oe+qx{vx*dB+ds|6e=QW(r)Ea0-puqsq5G)OT7e^lqO!E#?;PkI_07S?y
ze>IGvV2D~&(Kbx(2whV!#}Xl`r?=(t*a*=GU0bC-o+?S9Ke^bMpFHCs!%`AD8NPe~
zgivbrW2|4O7|88^55diwy3&(4fn&_*(LSClh!^DBp%zDywclS@$gkpV{4K*{K>~^1
zl@2teF`A5=hHSFD`HwiqC|lOJRF*|)<DExD3UuXTmlocxC=?gUsqx&X`>XfXt^8~)
z#@Pr~JT0$%mY$2Wt|ehj@cE3)l2#XN@#6N~0<F~6j;SZqasW6Rm@XQ6_4{hs<Q+HC
z=;F^klm1sZtmoM)6l1kxaX%W3Q}tLsL@Q27bS;mfYq0?GpG)KLj2x|ei(>FH_5>wH
zo0Wz;oB(IDD~}uLye+52iKs%aU&+*F;{683I4Ig1ev$bI@>w}d#SvKu=YDID{1Ruf
zrOABbKVTgp1n(~b2fM4GmAff9gQxlzz)E|!j3j$Kf%unezYVx8In$PrF(RNE4NpX>
zD5T}z=wlU#G7L%r%(zdeFTs%p*N^UET*=S0&@><dGs`2>fHO`mHDTbFG<1@gLq=?h
zO@f9i&o8%9rZ*#h45?dYd3X-B#e*OBO<@4>wfWnfA$?BV*1ML#RWca<L$1N)DY8^R
ztgq^<#<>Ts6jbl@0CNX$_QO*2rg|F{C}^eFU#D8CMV2?l{SscTm1P0lE~slN@GuJS
zE@p5B3wIHK{nT|Ovi5<`*D2;UC?Jb^Zn{*6PbHdOql?3KDK;bilibUPb?^?k0c7fl
zIZUV0w><uYCozRO3V-!(e<;jp`_I{vaOW2-G|VF-7_0POMJ-NKI69F>@Uob6A=v_6
zIqX1sVOL7m0OiSY8T?-ApXtrBA+YTh6TWvn>)!<okgHVSx73g|=Rmvxb7o9~`OozB
zM8bU9d1}|NqqxcnA8GD)Y%81AZyxc1gL-L@<KA-aTHC?Xjr}Y4lDo>4u2c4OkklCx
zFWpx>@Sf@KOxgU)Fj{T`d`EqZmsx(vdVjX0Jy~8v?Jwl%-mBU!fIb>Qdj=J@beC;s
zI3Ih~>)b1`))^G_uf)f~fRf{AECKI)?ge_RD^2D{U3v05Yi`abxnG8FUCUnUOtH#4
zTh~>NfCYK)nyhYcy{|{?a)@fZ8~W%S<!q`v8|)jEQJWfF0W;Zj`k`aT(6j1~9D^Hy
zl=iztqImOrU@@%M&GftpR9>cox(M184+_w2f0TN9v=Ya|+Tz!<O+ODU#-$ATj#S&{
zrD?tKIn~yjal~fAobmXss-7pi(fr3R$GXpl(6G?E;!3#ot#BOEs><+<ZdtRqGpB3&
z<r=k`J+?Htw|tY>Ybo<sk9a_=wiICZkIeEVOMbIs-P6Ui*_PKR6e*_lzZ_iXyCfJ@
zLUEE={3|Rgw*$99mvo~gBV%_oEF&iWmaVf^?XBGTP0)r7-oeJGl97n~M$)i-y(yyt
z9gg8qj7EcshDKsT;hHN#TM@5Xb<RUi-aEq3DW-SK5E;Qm{ymmcMLTmXPHb>{Q~WMU
zTX`y8s#ZgdiTp)pZaw#yjp@!rRuGHp-)!8vHmZ{Oy9H;PbM}-SR!d4u+c%tBM-3h6
z=Shs(=(OV&PN?71>Zf@=gp5q!oms&T>@9vpWxTo=ep7Dt6kgF;f8wgK>+H8uckuHP
z@F71t`qbj~(Ad^l^dvr<J_=I|RdK_gH=5+2_xylbU4<XmnERxES&XK5MA{C;8rI?P
z;IGhAPE>JAJv||eZM95!A_EOxx#;B3Zq)j!U99z9@ZW+Mm38<u+qwfKJO}k&lqJvM
zQcsi_Psp?W`?kdB`PFvG{=p?W+UP%Piw`S!j>*G3_~d<SviO|~nEBV_jO_F&-g+yD
zbc5&pAB%tP|5yC8GXJ-H3f;~m(oW>3589u4WyDuh)&*?B&TiN%X?;UeX0fiAxh2le
zAhJ0?R8JJ$bfSV90XH>u+{Z(b45cJpkm2$E;eUr#%EJ>dMQVPB8T9$}GrfX_`eTPO
zwsv8SMqpgtKONj(xJ!42T|lIls6utX`LsD`-0Xlg&j)So{l3|WG1;khbGU=leIXO+
z!~NptN8Eu%zzmiN#v9{)N^(h}*zx~(KZ>O&rFLBRf8_k^pQ~^8@84s9Duq6|zJDRK
z|A(l|L_*gO#Sqc>et#m!*x&yJ9$iasf~^^0RjX=DjIcjoq0LM3nO_u&rYREXA@!{_
zfbT`QSLgB0`>2|=N=qW7`}Tu9U^vb0|HtP#uDRbaXn)w!s9<RepSY{LwC+84F;p)l
zi1B^%bk7r|&=LHX+*ezWfi7`*mveDW>#JjFIIywN@80#DTc*M#*N+g&U`ep*Q;<!{
zrmvC=rYcfdFTU5$YMVBQ8nIM9T*5-nXN(F?>NdAYG$7Wg+WWkjqW}gKQ=RQ+Idd|(
zR6K+V(*g4NFO^IrulW4l(4!$8KLV9SWC~u($Kwy&J9@dt>~xUc_p-_pV8^FCZ-P67
z)iLeQzBuH(#EOXIqU2ZRcqJHdSBUm!*;Kw~w0~GHmrnZTT&y9it?jHi$dy~V=97l&
zVgD2mB~DHgV5K^CHNCK4Y8K2R6=8{qDz+&|o`_&kX27%*kc^`408x<u86s_r{D<9;
zz#)V<!oo<l=bAf!^aiuvX5$^8?<VFhn9MFFzwJpYLFS+m{U6+Z2T9n=><m<qC5}HC
z*!JS%TSfvWagAI0B6Kk{!TH;SK2rt}*M&>*P#vm(;Tb=rZx2K|7A75eW77jm)=Pd;
z1=w_4u63s7{8J>hj`d(hHu%h7j4jzh53Yzr2bf{`uka`t6ESUUzCaBVh(|DzY3fO&
zTClXPU=?99S06NGsgmEXU#vq8h#(l@R(a>^?Rc4d4=xd8403sqH0#TW6Qf1I6VP-?
zT`%S_J7eBuzeuiJQZEBUC1TEPOGnAAWOmQV1Ta=uHwQgXa(f4<8ZArsNUn`ru$j5`
z`0H@05<lRE1{58BhqsB)g)`?OF~waCI+YUxVVr|yIDi7<o0G`%{K=F>DV;0(>0mgF
zXMfw$vCv)uDgt!36BPg+NK>Y%+uAAkK`a_EuhXI%xXe4*vI@0Nh=##1Mbv`0c(qHm
zF2h(Py$4<hJlgney9;dsQHUmLUNFBD!#R+GbeRy4kGK{ZY`8H=B+p?)#WC$p;myGy
zmrz7pDLqa1<qSm?a-_nL@ifUhVGx$Q(5|Qv);PvglvQdLpDf(W^6!&_uvVTfp)PB~
z6ZvqTf3`Dw%>1U~6&dYMraT=_xq#=sFPQqphMz1U$r(mU@W)Ym9cZ@*{=$BFnH#!f
zyd-<A(l-<8;z#emR3Z@mu(^uyP%*`Av!vY-m;zo9cUOnG@>iB?`5sUlCINq6PK&d1
zP_4Gd>d*qXi!4c;r?a8uhY#VmoUDouXk8*reP8uq<y<Eg=k!@unP2n!$;v65bNqn0
z#QRnx*^%WI6((9BlgDhf1aUh<_Ht8zNAGFneBBD-OBQvyH<H}$Q(A@>lRWe@C%Y-Z
zH;+<w_$Zq}0N@M4MrV}f>KO*1JkdK&WD+^Gab&aR>R2(9km_R`HmtO{Bb^k>@}3!i
zn;p9$j<MvaNY>QBjL>Z2h_-ad-7aAEi^xyyQ%Zj~Q7@+Ao+D9FuDK~=y}sc*gBGY0
zDBKLT@#eT6_zq>sek^emvZyi;L9NrGJ-Xq(d6`%Fom*<)tO>m2(?v;KL$egM>gOPq
zQk2l9J^?Ky=wKFKVkMy^{ShtmI`TyHbdI?#4e%>cGC05bv!(GLRnHv?Y6>m3rOjbY
zi3Qjz(p@KtBh|fik2TlCq_>JWXUWjY(91y&H?AYU2UrGaKs3x{+7glwlz6Rj!o3Xh
za7pUoMdx2cJS`r@c@P)zE`GC8s9!c8xG2(E8!|QZYQNzhBKWPa)Xp%shU|@k&7!;2
zqBXQI`<4P8L}(CK)%TT4Q|Cp26bhG{RXo`b5oNiQxv-RN*@df^AE0OfrcrYurGvtJ
z$&oFo9c*h3u+>PMENei}DV6G~v~i`+4K(N|3Vx(__(`h!yq*qSl5QeicCwhZ7hM?k
zWI=LI0UhB^g#?N~Jot9}EPExB!gP`N6*Q4&@|($^AJLWfU>_$B4*O^q7KrKgS3`=k
z_xaI<J+QrqBHUAMcp}`<-bMZ`h#_J-s5iw&wGhRT%>44#2Zcjc@l11XYx5eUgeTl&
zmQ|^N7j97H+<%6<{AwUFDxQ?;WG+L!E7h&@l`fF$WTr#N8;8(cF8Dp;GJk_3JkJF-
zqS0P7M0tMf2UJlv?Le0HpnEc;usmG`D>L|VBb4_1bt%e5@!ZSnV7nYyuL7@jB-}RF
zyipwUz?h#L!H!vxwj$w8tplw||IJZz19E}wk>|>TZ`M0(Tv6_&A3m*SdqKP`gBMEH
zICO_~9m#_iZ}d%<!y(nU-1%=p?B?TB<@B<_-wCg6`-NC}m|tq;^Sn_JXVKV`oH2$-
zxjHtrXCAMG`r<2rANj8rM?E1xoAYop&8?8Neg^;Mw}bg#-{()P;Z%@)I~SiATq5F3
zg?Llo<TG23?9e$@%Eb_(O)AFwm7UWg!rL)}f<^;{f0S%rO8O#<-(8<GgbEu-aI$z{
zD#E}<v_WU!Wo{Qqt5P8-3`uXzByLutcn)oPn6TUKFjIQGL1Uk3eDi20fj%<F4>h!x
z6giaFPb8?zlnQJS3<3mhqhZN=vn{-96EGnkrx-c(sg;P9DZyA(Jif`J{4#6m+_p2F
zUheg(`##1<f@40O|1?cDa(Jh@_xW-y;n&#JZ0}T0AUSFUZAx7)oKMQR+?t;75nX`X
zNEg{D(|X_o^d2lh5`B!CXM$_%0F^82I!C{z!)ixAT8wC`1G!%(+F1O&RfmCeWV5s1
z&@Ot#RW>oE)zGR|nWqg@j$q@*;!37ku^rvAIbuXoXq|?;0aZi2I}KGzA-s1uNk8IO
z)9<-EUM;gtOk9EIy@Z7F3HtXuy7tox)TfKaWnftvxt2WX-y>|avXnx`x>=k_oL@iR
zI}G0>CQ-gExV}2bP?JV8I@R{}DY-H4ZM#3PQg>5%N|XZNx@B3hhhU2(J<N5$9u@>%
zoCrV!zpO%uga}R$YlUDvdtrW*4~BQQ88YMxgQ#cLd{2#wp-ZdliD&o?M?w#%b$rG#
z)qI05p2q_18KsIrotjBf@(JLvN6S`hc8`E}|GmXE9a{Ox-d5Phw=x^sAj{7@4+9Mf
zSk5gC5kk<{v0KM(LU6YmGFwKPnTxyr^6LXm-Q(6K44m0Jk@$%obBj)3MVE!D{4%N|
zu)+kXrN7~|<=#AlH_qa+?;p-+fT^Gfr_gDNx^M;uv{t$2z_8ypLTf|5DP*k;@n)=R
z5c8KAs1$k23EC9F11_*=dWIj@QBGE4uH07=W>^0Sn0Y5(_PLk?)cSTVvW;-<Ud2aF
z56S!PEb#`+eN^zHkjmh3g7fQ+DAnZgF;}MbEwAVfc)Z#X!)#mkK(RsdnpZT;r|93K
zu%dfRe2-`^u(D}r0Za&hw#q1>`m48OxgGe69wS%+<M?ziju_g(D!@+jNJD9sb0b+B
zffvh$;$QY)Q-?bzDxtEr_%nTcPo&xqBIX^`HTj!E1_8q5am#a9+-2{fY9n0x*SHa^
zQ}#Y%mP<c%*W|*R_G6^*GnSc_sr$&|2FYGRvuF)Mi}dNodk^tAYv>lGn3aALG{0qs
zu)wLvLzl>0J3Nk;E?hmRKK%nx8Ys**>WAZrO}?xqI8lCMYHoCsH#g!;%HM{|T`h@O
zq{A&rxhm8Oy?f8oxRZ*0d|Tmqd4x~%T|2fhpZfbAW5fZDKS{G3mHWo1>sf!j=oy$N
zQx>{DBoYkbx%$L5KJ5HFQdrj5iTjrAetssKDOb!)ETFz&v9v_$wiF86@y0{)uFF#T
z@5{$n(F0AzzF+&!`dgWC*c`iE25hT;$5&oQzO53>W{4Z?D%EoK)+5YDn9-&FhH&iH
zpwL&+fPd7mwdVf|tKCY?DEROPPDuZZ{7;<-*MDHEaIta!SF`h1x>IqNZOB`58jpy&
zS@Hh|%A?=u3~-k+Gxs<e-acg5m-7=GDW^(8A(w-VNTlofTT3CMnyXSn)3o&X{1Lm|
z*UKlM@8^N(02r1md%HSl#~mW1>SK+fOzp88jX=44zWaK9pe@ziy9i_~P=)fb{It2H
z+UbKbYXHx_9og=)Fsv8j^|^ERkyC=`v3$b$IN%<*8@Vl2j62HVk)n}Kw(s}(d=x$?
zz$Vag{CvN=uTS?{1xk^mu<Q5pcN7dT6vgs87(|hJyNdVU^z(j`!AKNU`MpW$^WBzU
z2_=>(P0eRT`EX)zdUf-9ps=c`>HN`d|9@l^Y44AdD|@Q^G%pUgC)v~_r$Tf3yArXz
zG^iJq`R0<8eSXC^kB_1X09gSAnH8+!hH_W)=2?1~0-O%(5{s{bc?k?Z*|~a@kGl8-
z&*UM#)~M0xA@~fYm6nB05fAmMf`wUuBQ><2U(;wNxwa#konn$S&Wo3;7qbxKEkS(K
zVkTfF9vZz~r){1m)@!8s4=7B5us<W9qz~PVZujH}aGQB1-gGjYX3|HU=_i-uPU4Ld
z2;2k|Qut|Y1I-3pC}=Ii%kJ_n4}pUc#59#h!5RND;q>6aO{sfpzClB(3xmes$kV>2
zPWw4O#kE2z#soD1IDB~qhjaFWa5<rz{qyA#|A_RKf_iAXJ>gCu#Y_<TD9@X{e{=-a
z>#fz45$apso5NE(1uQO~tF0S?3`+9ga$eq0-m|@`GV-Mmu_@IIaKW07DMEt4?6jqU
znqAB^w`j$x<H9xPc;~@2ctH|j_$0`pU$2tPQfu~pEIckgpozT!sTXsG(l}*9Ed{g^
zr)YP$aLaU$nZ@6FF2Eg|g~Hm8Qmi*tB5|_0+gA8|kr~Jzbj@H|3SW9d8<ZT029Em#
zW-^wJmO-r;XER3r=7E^tH@(14Uiu`co42IUr^Psrg0f1Zh>pH+?Ms;{=&~5-tF)l#
zp7~rQ06$W5UeN2|j=u@u<Pp7Hxg9m!w<dZJpY4%?;GP+e^8f`$7X4&c@l??ls(GL=
zBISLEu}kc8C;B<!nc=XPZd!NUmc63KJD#!hf$UKt7VHS)%tD!QCLt`>SHq#09$m|=
zgxwzh9o2N#ed>0%dI}=k$bbzq_MmwZOP8>{8%v|9so*s4#k+AgIq$UY7sOe9hkq+h
z1M*&U1Vz2&7?<xlg^Qe+d+bqovR7E3m1C|xNJNEhcT8P*qNd6mxs*dKOsI#nT+_`0
z&~L#6q&(~*%x27ry*M4SwSmt58rQAggTwl5p?`YIr7A$9t+TS{!i6dVyK~exi%q47
zagoD372vZlJ$t+Mn`eD*v|#Ahse;6pgVxgGq10^7sM{axJ=Qa8DpEFPLl#pT8En*A
zA9}nxD&r^yKn9Cb9D~>FaEq3#otr&1pxFbUEbd0tjIt=JXU3kADX)A?*8)1~kdD?I
zO*MqhOu^XaiN%xBs3g)5&%(^94;_GUH})r8R;DuAES}>u5kHvIH{E!QPcObQ|2R|L
zn0FTrHu8==$0%u)`^>LKW%Lr(Oxd{?A1K|LJl=J(0_4y#m*vQWtovVeWPw}R&5P?~
zier<2R3=~^fl5R~-Ur^43aR=*o;ACTyEx7fVwu}{N3-_AMC8SG_Gh8iNenes2Xp&C
zpQSjXMw7hb&l}sw0mi-GqjZKKWrXt?04cQ5hYLFja)^gIw(OyXCVr{yTMC?8>M9@5
zB~72$!m+;;7cE|An#?<y@%;Jx3&ezdVVvSfG)4ostH-edqqwbe5@)X46-P;XIQFu{
za&Dl<BV;Stk}iUWC>K4RHj*1t*??CBk78kYp9um=3;BxF@14WX2OFonlc4}}7P7^a
zTUq8oQGGbH&gH@?MjET6Unf~5-b%!G71aaEgxJwo=j)i*)nvM2i7i^s0($w&6TeoF
zSmv!==&6mo>M*RX1uxjurWd-Ft8&&3_%3k@y<RGR(%Z316yv{WD3{3!djs)Nm~hl+
zX3#|KPns)rlhuPZyyU7sQ*Rahtr_W3NY<Ch3k@TUNYp~qQC=FLZcs^~3o>qJH>UeI
z6vmRs1k&N(czRWb+A9WgXd(C?eh`m}dxz1>9(3Y|3m#fV83Dx;_WqxK)w7>ZI?E6z
zC<m=(e{$5t+8SGd3?fx`^jQ%~K1gxmy8jX$TvhF3089oCPBhz2VHslk#TbLcnZ1^=
zT$vwNbbB$ug_-9}RW|COjF|U;*%gP8BRxbyi1}$E<-VM(;8%B+YtIgKRz6WO&Q?8n
zc{D1-6n0ZT39OX}l`yq;k+ggsNDXqR#b>#)B91wYnDVnY8S$}xLs{k#>L?!IojIVt
ze9n|93^G(D(SA6n9qz5fiIFipMsU5sY>`L(o8>oaS`y<4?t0zZTFeWDHH+M>S%~#Z
zONG3MyY`)$Oy_KdzFc<!x0su5<9flSfddXGd-BdN)b6-Y<GWU<=(c*z-OpsPg%@0i
z!KdGgCs}Bu_MUvP>DDGjqe$CloqcYI$R%|yZ3bEymJc#^puc~g%6F@iblq{%@6cI#
z-+BReI&tM6OhDzPDZpvCsEEjg2>m_s9Jah|Xg4+_xv4q$uua}1nzix=f8g=rlP$#L
z%(4VcjtG|Philkao0b2PpmL3=!7l8raspnK2XL+qTWM6~(^X-pLmpFgW-)hvSz>IT
z?_`WF&fFhwQl)a@bC=C?AXUAd?b9WWDT7YWB*8!h%!N(i89RR3hOv0`^?0-ho=O+W
zkUNs+jg;Wn2eYqeEJTXL*phzo`6?g+$(0Msndu5CJFjqAx{jTKs2HCwJDJ7^SojI+
z1Las4@{Wx&o3=uE{aueQ&0R-~7dtQ+dE0KZn7&{NyonjPJI^{CXq1PObeAk3EwlJ#
zFo7T9-Lt^4;@<xL))CDdgDmp4Ie5RNibx5{$N;wY{N{*>iFK#^4|~o-b9gqf^Aw%I
zYdMS*gD9$387{?U6%&e0kfShE(!4glV5G`ulACakTYqH?|7gYIaN#8CQTDfu^^kBq
zJ)zYjq7VkO&AZx6bmCU2Z2uR~)|jqoC)T8j{zzIdqI{=AY;eOhUkYf{ZlYdm?rb7W
zLAp+H<tMrA-x5fXuOhSlfX8hP^Oqu-)^c>h7|X5<G0Zz@cK@3fKNqj5E7<L5-?rzV
z4<Z5h5su+$tHJv9FY&>T@J;K86&ELiaIb81tN2F{vcSP}1^|WR42Q=JO+&{dgYMjg
ze?YZTh~QU05J?>z08i^m9&^-HI5rzYZLeD>^!3jLl7=hRibzJ@@)wW_eWb^~et1#Y
z)`1AQx-~1%`Cp+=bkHK+wsp6fV`|r#1KGf~^M-_V<xhft)PN!Co=J35DXg|D@w-@@
zu~rgaHVTA&a(q!&KAr=-j(wN(75;a3G^#@qO>zqL^_qdMqfhmgRjO^fA)c#Dv@Rvp
zng+rBx$+qcBf?4Z1*`x#Vn8Q^X@cmSKP@wxK3ClRLB^f>@Sn)cSU631&&RdhS3U?(
z8!<_xx&d{dV<8QkhPM>p1C0*@!`5<QtY1?P-}ZVy^q9knZM<nbScE@TIOmy@Wr}X`
z(2u>yJ*5A8?o>+>ZXdJWtA+?)+8Fto)!E0q4ph?shm7xdsQV`G^~Q-0xsh*#HBSka
zYe@R%s=<7u()G9&&PR&RJEf((f%EHRgzpoKv0cm=3~$Z0)uE^l5kKP1r4J3yO1S;F
zYwynVREg?!>1Yp8^?#5;;6S<msoduN55xbg9RD4dx>aYwiM12H?}Pdg<k-QWO1CO$
z{aJN}e!@R{l6qpA+<9F0{vB8<0Bp@>ttmCQ$0eBv==U3t^8Mq7RGyz-*FTT<v#nJX
zq=m`?1wlVvY7gSQ@}%4OZWTpok79Yje_UVhB&E-{or7c2mT*2GpZI(2yn?vsMj`9F
z#j|q*RhS$I0`K?&xQh)Zy9N#de5^PEFaiXpWVNPchlxKSxUVrr1(eJ*lDALuf9&5c
z&j)9M3e8aZ<b}1|UkCB!pMyB$vP+bVVnCVSpDSnW<^9vDok2w);8HiaxM(G07BhK0
z3WJ9l6+}qmLGuXy7hwLpJ`{VLR8xh&20<#H59W3h0LlITHccJHGtf@{{#wvw_jN(q
z&aA(sFmnE@Q@CDXIZtXa*$u__Dh^-P!6<E98(}G-DF5*Z{uG0j`UN4Tx+i+)TI*0R
znq)+fORwEmpK2N0K+%%eWEb2UAZ(Pjb<SZ5zR{z1=SL3dQS{lfSaKx56IHUs$CDin
z=vc_Cci<#s(y2f*n}jT@E=6Ci^mO5vDk6L#4+dyMUjWQK=jFj{HTXJ0e0mPHH~Sga
zsT(yOk!fe4Dbs7P48`o8KnXrBO%XaqG~28sXCkgOGA0(}&Jbfx1fr0lAMK1&btY|F
zkQpvpR1f(bs18Em8y&3{3>jUUoos0D812`O{LJt5srn6$#xrYfg=UHgcV%g~6N%F0
zQV?^Y_3ZKV;II$;+Axo(#(ktX*EI`c|IsN>{4kIB*0%Dt=Azq$s}wHAB!Kfv{RnnX
zCn6Fq%{`@*B(v3f92CC4J-p@J&Os0Qq7B_JTTze4*N9az8&sAi*Pva5Lfn4!$BHIu
zU|pZS1fW-M4JStfkqM7$-mUac0R;Ds7=JMtM}Mak`XBBkTmq|oeLQjk#c?vVAc#|~
zDr53Dm|*@n(;7h&O#hdk3OIHSv`v}z`Zb`&B6;6VMoR?~CW&TSm{$Pia7*=cczPI6
zF|#bOzE<U<<pUm8Klr8<?59toPyeriH_le1Xr(<I&l=)2qCGT=-C2gFT&#sCbtU0=
z2uy569I|3TaQfJZCnowRpI%ktO(EGBDNVhA`;XX=F{_6z>KL$uvRqyd2IM17igyEF
z_mlhm;p7oWyisw5mukm{A3bUbi7{{63#UmC9_DAT*gAbb8nx>W**os&PMPdSL0VVl
zAVLPOQsUbn+x?WzqPN#blgOwcIk`{k-lGOH2?B6Mw;xY+1{2*Hn{3tjR%9sWn-Kdj
zU15=em#ny;!Jje##v6V-?Om_!PqDms=8Wn}nxLTm1A#L$1)u4p&W5;M{`DTt(&$^H
zVMmXeGD?@EZyy1cNjCy%53qZScZ=+qz`H^D!&>i!Y1vJj#F=nM*elchB#25ro-G`0
zEirF*4Wyk66z67`Jv-WKNUx5QPHcP@(lBilei^mZwY=T}B{w<F(?0Ez^z@g*r#!;$
zyM8``i*f}kEHE{8hViy;=7npK#6E9tDE2h1yj6aH@;$XxHVdw_|3q=2?H_%+dg>)x
zd8%R~XNx4GPH4UOyYfn&Yo$1eu(HmP&Uwjn!(-b#l`=;x0JTcK5`|PzlB+ZB(Z^1!
zUa1FeiBD<o=7M1`yL}Y<Zp!q&W0cpbr*6`U!Taa}t*Yt9e&>M$lfFU`&m_xr-=%n(
z9uip44cUy}r4lpk{FMQ|yP2_l9L6LFroWY05noqKH4!cjpclZZp1*0#5;a{IsiSum
z6tw{XDbj3-%N?otLrUSXS9v2+C~zuPTrj0@zo8gd2@-Oav;t+f6vrRQ79W&Q)TFWn
zDv@_Be3)tU_j11jizEZLS3lvSph5Q9XG3_QP(m{KxXdhE=em*eb)wBkLyj#nUv~g~
z8hwd<pay-L$mncovv-|9L9drQ-7S~``K`!G>;OhZ{ScQqrA4jt!7<lmqG>ya&P#=p
z@eRPV6HZ>;U~=N_e+0T;Sk>XIyFcrTq&q@R%?<%~E?52SQ{+yOlUj;Yo7i}apNPsp
zm(lFxtp$3a4Xs!oi4Dw!=qUt~Us5+yaD=;!)}#&gO9GJAo3E2qLX`W}4D)MQ$hKoG
zyBU6eoR0>U;OJ29%lGb{$;(x%dX|~%e9@MDPVIPo8_sym3{jsl9`ShGDi?)<Yx`lm
zyy8h@d*%m2_e(j^fYx{vuh6c#k<okawzmFR$|zTK{Cf(63f8QH#1E>LTf&}|qMOKu
zHDcwp`GWZ0E$F}3`K?&w<@2%*<b%z}hfNl+OKzocmeoZ~*1DoWn-BG>v2{N?BI;7N
z&#^gTA=)R(1VenWvxa}^?wswA0}v6D<tp~A6;-fKj3on`uqz85<2z$MGh(un^rKq>
zGYmo1ej{RCzhvdYQo*Lhx=`#p#$i}8efF`|m^EA>SiUEGZQR558JxV@@VnxB=X^_$
zXP@LBaRE7uOqw520d`$Uk;0{HDzu(tCR<r9>OA*w8U*P*hvYPmq-lFQmVpq`SXh<`
zbxr=!wza0*so50*f4W_-y|lVj>djpvjmi!r5lvx@JJoM(l0?uJNCJ(itIlmxhwpDz
zTEE@8j@93*TM(N<+$^nX(%B@PadK8(-a6ssEjn3AVi&3x%c1yI0`Y>>|6zgYbry$J
z>?@zUUWP<_F^7uw8W*a1`?ud)F8_os2&ryq-5km*d~PiAg8WO1wMvzjx*Ln!FptKs
z_^`1##s=USZvsaWB%$MpET9kjMhlYoIC@6Nvk5-O%{QRt3db~DIBxf`DK=qFT*+nK
zQU$8}(l!W^ZkK1xWjN-c@zk1X{BW#?)E&c$`6GHL*n8f)-6(gP(t27THP99qcVMIl
zO@XcoXH}wjgDeqt8;M#)#)jJgH~HJix+rrF%Vn9bl7)F(sV$WsTaYm@;gu8=g+2My
zCmrfaI?O=&t`lN_y^XsqszPIJ>bY_)&(anE0eu^7!UmEM)7-kH<!L<8rf_u^a*yLM
zu-^c?&TXA;=dv!{3M`HbE#+Bm(LG^^o4UjFOv8E*U&U3!t)_IVEXJEAbw)3CO?KKk
z4)Qxc2*PD7zyrDIpGLs9wL!n$LQXd~lu3+6#_ER;KBQIB6AQ-@=9xh6;~$<w78vyU
zmKgD#5YZQs9)#$kT9uQ;X@r5Z_A0jnpv%l$Wz*haT!$y82~OO_QavaJxt5?`x}Xa#
zN`YS8xayioy(hJ8H265uGCZ8h`ddWCnzqHOxiIH7m|cWM5V!BE8+ljz;9D^3n&4^p
z17rlKFvd%rRP_fL^%jGPC(G^OsS>z*v&&6bqUW~bqs0<gdh)anu9t7OPznjyGo@K-
z!lAuI$97>-2G?e2WjdS(Ixr$1RFgE&oJQQHTHNY8`<;@sLv5|T6{^?q4cXc|R5LBA
zB-rlZ{z`cScCLu<ntWTQK!<L=C3K>|yw1WJdHrgH1H2Vy{UiqraNVmY6Ym6ieH~Zs
z$~|`vEl&C8rWPj>%Lc?6h6xQZWe@rMSSCjzjeB_@_U=EC6h9WTV|70!86kFKcaCCn
znLmgA41e8a#1nh>P70nR_ng~)2V=5SdcVbWu_2Ucc^BvvRrqEqu;CmW=BVHjL&KIG
zYg~ipKyBD(Qrl#m$Z%#ZsxG53jXsF+H+3j}8*#lB9FY2VTMEZ;-6A@B6s*q%Rcvg<
zQMF?WcEa+CUQzyT)JPU-!aGN+^2<oradR4nP<oU4^)(x?7gX8<{rt3=T^NuY&?8Y4
zZ?Psyo3zN<UU-sPgyELjpmVZcVaGN6dshHqEZw)5Ze9A#Zedm(PL&^pX@%~opz0uD
zAE=_zsEI{`f4efTo0%p82-&OhBo`VN{v9s9Vb2YL{D*4eP^Wf0t?M<iuzPK)sWgX4
zf<tJgw}NPlUq5t;B9^DeLmDg0_}d#dS1M0(vG1+6$|+;r2imCx#!p6`&|p}R#^Bdx
z>qJbhLU+u_uzGd7^|u$MML?4#K2<oBxXMN^^ajkT+wXa^1#a4-l)avyhLjy8+_;Cu
z4td-m92UJ)3pO>~wWgY~`);3jp{8Vj8+jHX-$QY?`jY;r{eWDLfmZd^dIz~mJzg_w
z=AO5ok=1E$6Vv8XhvS6S*l%+X7K2qA{^{*lQxxeJ7Cj~lPEkC2+9gM=lSKqfXrwbb
z2Go@0oGdFmlKE=y&5Eu~2jY=hKHH0#%_0wO6H1&j_Xo(v_mR9s;eK@t-WoCn+&^SD
z)ogG47%~PghWNQvMO&_f;dh0Kv^xg2pB@+!t5Upv{K4Y3q!ycM2K<b3WeMjiM5n^D
z*Iw|%%dWf@_JkeX>-P@Z&A1wh7Fu=3!Xt&D{(Cz)LInHQ>oNU~XGh0@{<!QdV;HJl
zAx!sqP)ep*3Jl>;jkBl4#CTzjzY3*j-H@4ENOe}4^YEDUF{5))9xq$xlptS9FO?@M
z-FrU&;nrPluzZ6SwdXkfhklEliT(c#TCuRR{&y*hANjHqT|4~N1?@W|gWT;5&0keZ
zad+h-J(Me5eey=m&S;N=9Ztz0o4&3SYSRTktzW<{u>(*7nt`9e>wSCr!Pu07c_qgG
zvVqd#cQF1y;`s+2$Q5J4{)q`WIvBg3qCFYhbsQFPWFRNNsK?KH+RFwDxBml;$mP=v
zqhI}Ui)79xPbPw2<5_xhmxty7L;!L0!v;z)k<8-%Im>u}r7Y!spwJCd`CfU6Pyg<H
z)ML_SLYnIL>k0Xg21%Bu10nS4{k+RRYESQf5tdQG6Ex!a-2F?~&$yNAj}(?i8n%g5
ziwhS>+F->z%8E@u4#&P<+Gr2+QPwzjUVFBqU=FEbxmd72FN1m`MFJ8-MJc6G;j3hR
zl_APhky=69R?vtYYg~GVXj>1^acm*ll7Ul4D~W1%i6zZ<$%h{%r)3vrA{I-d-$CG6
zEpi&5vF#A%RXqvm##mD)eUC#phZkWg6#*~dnB86TU_c22P)3+8j6xkN-N<LJ@_I*j
zfq9Q9f#Amw3P0QkHA0M&%z~nVcsZ^gGo-|2a$zl1I!I(MiU~nz;Hg?UX`%j|yvYO;
z!PD&5e{d$n%mYaQn`n3eRlF%dPaFXEDFHDXxH&-WLSkQ88Z+hyTV#~6pU0lC{vDBS
zDb9=h_3%B$y6$bZn<=m%C7Y&O^Qxs=oNGLeFt^m5x{9;RFt;laXK5dsbN7?#mjGbq
zHJT^}nlH8CvMVT<I5C_%My?@&g(%Nz53jyU&k`!}KoQNNF0Lq8(M8@3a`Qw@*oT)-
znpd*hmhkf1&U@p2k#8-jGNwnG{G}ZmEK!*xDsh*KcechX@m|DZVNP;6_q}vn{E8*Z
zU(7z&G(oDQv0`jNVsE{`W%F7<2wzYUFWRZ_dkoRjiRS;xyNy}tnuP+nE4!pVk9mRb
zpVbGN(LkS?E2hcytyN45Q7Z61sc_GvILJiZZ^JQ29t_+sgNG#Z#fLptY6=w{Qgm6B
zkf)l7VfGaiK<1`KOSZCT3|UxezeH>kXI!gHtU3t_sg%2tu+d99rkO&oM@O{LL>Wb-
z-Nf84>@<qn07jJaOd|f$@P}K3;1#7_$DE4qa|^ut!wR(GKe6-NkCH*A#oD5A2SP4w
zPY*C$&Q$^1I1EWemmp|(^5#cf;EW9!WnIMYpOlNqF^!|1Lz6?_*UIMUE=&GE+6&;>
zfbNW-?W=iDD~7gC-OQ*vw^7QU+|u2>O%)IUArO*5VF&^Ckf^##w2~(qs;gZA%bnmp
zZ<k?EXRejosf2z(7sAaMV7~yS4@qRht4MRQByAQdDS!`aTDnyil~ilNJ(}3AfEh$?
zc4mjT^WBP5kb};%AT|GU1~0D=4>{3%7LL4o6*2r_ucn;bm)7cdAUUE1B+zbbopDs^
z7Av4tm=^`l`&2mkVR&3?JUtPbFcVNx9mjidS&#x#zBq{6Yoro3Z_WC5xMoO7X}>Os
z1TAG2UIjA=RKU$ZS%s<=ixK+9&&ky^DvqrlVgPgeS7hhF!@;ya;;puBhB1@Qa(;w}
zb9t3as7oX@?$$1Uk`gbZsn4j&DOpQrv&s{c={}v%h9vHMy|6MC8c-8zOFlFb8KT7v
zz6k4SE92RSTGo`IE)fqgaqnUISoNf$BdH!OId%<eI<oFg?7k_WL)Fr))`2~LQVHf_
zRH)QA4#(r4yu8CY%Xv}&Wk8K6exP5``AqJ5Rj0@zdvH~juGHaUj5aFu8Z9+rC;1AX
z4xRS0M9_!`nvGf$uAjU4V$H80&-${PEr>d>alYLtoHy)@=U}tohVrBMkF02g(Dw#j
z6OsWfx!2*HSuiN*v@v+0upL4Z2bb+~WE)G68B<-hWVZ2#p9Q?V^TNgmygVq$FhB}Z
z7%t`w5cxS)E8clTiZjUcJUIJ83Om4{8eF4qorXsp49*9-*49c%%zGw0%Q@5V>we}o
zLsxG`V-L15ann8~2`*3sAu4x6mLoCgF%P;vf5Hu|jSK@_Ig5vR5*G9BeZojq$E@X2
zXA8#I#P4f^_fFcce`~zgOh1-~pIV>X;L_-L&9Di0k?*g_CSDaj?;mnng{w3Uj<+=#
zvtAq`Pwj-%WfndG?s2zTJ|q)w-+qem=l2~<6JBaW2rWxEmEAt7pIiKtXtG@n{s~v1
zTZAU1%d(k_QH{373WKhve{AniVer!|U>@}&>y5Nay^g7(F08;kN!P=!;1-{xXN&3c
z*uv)`+LT#phq>S#@o`P5g_*9FC(Pwh(LHA`eI8`NEGA0adrbyj4xB#P@E#=_y__#b
z=*_aNChNokI<1rV-OeLj_c`j{MjyOQ`O23GSayvxk`ymmTw1lst6Z0zyVtQ%Q?aSm
z3S#8?z(xwPc%Ohw<>vht=q7=rva*$}iK6(*rXF8>fZ#z{B{!B{a<)_mZwR~Oombfy
z16`vBHP?LahEo`w-B-R&Ong}A;4L>~G(M?21HtA>0)S^(JUzE!urRhaD}e($-(XJ;
zkkq4lyyMWiyX@aZU@>Av<?aVIGU=2&C>xqvBp`@XOm_zHCrZVR2H<kP4-b=Q%h=%F
zcQSN?OhH^2kG2n!AZ9epw#-k`cF90_Tb>OOQmV(b2*GCBX|nA9cDJm^aP7dCb4)|;
zR4Cv44{>h+6~?lzi4q`aLU4C?cXzj-!67)o-AT~k?jGFT9fG^NyA#}b&E7lb&dlBK
z&OWo=yvxd3bgHYW|El`F-{`I?%a2=xX?D?)G#9{|<;0cX_cLB;)NX*jKLqvXNAh_N
zmeMkfoBtU`f-_5!@~s%xTgP;(*65k5L`$VuVo})X$_?Z&2GSwuNI=HV_@-OS?8BS2
z2;(WrPXX-sC>QwtLs%4{=jJjVmM7APBm~44uBngIEPg|*HnM}I%F6;pAq^nJHFhl@
zX*v{W5OtxQ3>p+wln?e(%LxjTj?}&+ZZ-z-DY3OK2uo}DXBy}A9!8oFXdhFnYpF*Z
zkW77%{=5?2;4*XpW=j&bA2e09|HF;+G2eYZTNIMVa>6c6>7Z5N@fw)~GDBrTr9wne
zo(BR+mngqC54AYc!E)Ho`zk38q93mJtz{x;Gjl;R#V=E?fDUJf>5+9BW+N<OkWu;6
zT`qclCOv<i<tVkU%`b$opOWn=SOR8|0ud>2f_0Ja2Fu*<*mxf~%wv1sU60-#gipNF
zd2V`SGC!-wpCjNGpjVmIK84<awFPUmLTm~eo9GpJN*2*9NQ8P3GPv&E`0lzq7>cFD
z8uCuC|GOwDy<myrn;Q$_{*Pqr=JV^|%H0^Sfru#AP90kl&5P_Yt$r*&rZmjMsV-0}
zKM`0+SG}8A&euk#f#P~u5^}UJ`c9<<*<I*Sp;0>du!r(;x5FG=dmsP9^C3~Iu4cy~
zEmyL8q)Vk@veB<`LA}=VsdhoAZTb6%38dGXjZ^m4zjGKF82_`w$VSihkKTtF*66n+
zZU8>Zr8<BVW)g=$<0t1=rg=}#R_A~<)PH@u!Rt)>>;XAToam;a(iNZ-tB}~COCNXR
zjokKl*h$ycyws}wd`>alO&_fLbNetHb_|dG30#(qMf)AdBqFU2A8lKU4K1@4TO;zW
z!i3>oQ;uA7t7Q2l_|m@jVbRAk6&|tp+w8+MgZ!d#t9_3~vbS3N@uVEGP$wk&#KYQN
zFZWk|fX`^R;B|%fWQC_g_u{gjmr-M|G&i{F)qKlG4lS*2^JX|@DxUAf<!*DktYvC}
zAGMEP^UKd3{`P*eplmK+6YwA3kTJka34=)enN70;3*4Fm?B&K!^CvwRp&D=D+PWne
zkHw*7_36LG^R<W$r=GS&h=$3T86ZzWxQ6-$Vv_R}no4tUv55s+TF^oBJ%V<-;KN-g
zoMyl5cHP5%8~Pq8(gOSTaG7nNHrsF@NgJksXKqeh0b(LNL*FjIo_Ye=1-e0)EL2~e
zr!DE!bef^Gw9~%r96T%)mdYyZk;r6hsem~`lj-NUy9o&GTD=m!>E|t~4EM8%=>v_G
zC~q4sbpp>D)yiWf^fw-YR~#6*Ce4%mywWv#`O<Kc)Orb3@4M9Qc@gdunADNz&g)k1
zw(%}|b@ui4kG9*^$MFo&J@0O|t$05aL)C8CXqQ8EwBaY{fZNTGRQN!}gxT9Zh?q}F
zgH*xAcPkv&&JAh}Iq+9dNw72G>&qC$O+TI$axnio%9>b3v7!h%p%7a+*P_4E7%8eb
zvK~bXwCcZs=wilIk@lQprDDA~m8amC-wDHzdHxUtJrJb4(_x^W6U7sSLq6Dy>FOx{
zq2HN1hxjT|yYT~s>9k1zS)DnveHe#+ngKIy;+RZhj|MZmH9;>bsC1y5Wz~Sqv1;*W
z3EsKgPj6Fom1HrC`Y4rLF5z_MV5Admm|TZLL&!D3ecoK|gJ09X;luKSpeIM%yZAb4
z&U~xaRgAtgM%HqH?V~|StJG}fll+DT+-Da%F_e1Zv<lYRudW`Ve{UF3{NuW<)o0vz
zA$LpW3q8ByL=0#jJ_WiP$vJ=c2>L3LA$Jghkjy@Ax<vF(QIaQW+2A>P!jRA5;u3k|
zVA<Kxm(k)9WiJd06B2>hTht&W%;ivyU&0mos;zs^m=rK-Fh_bvt7W%P&NJQ1Re!P$
zBTr+19dO*LaY*pqF#Tdm3rr7RH0s__vVXT>XI5VMQ!REWC_Oq<>`bizJnV!*#Hh!r
zPH2DgLbC$T#$Yq;Cm-^-?daW?ha4Bkhy6EI)bnM!I5=e;b-@9aCwW^R-6>!>b?>*h
zW;V))XJ))X-qwX1)Gia&lGxFy&;>8UIIv!Zwvbl~H4SKywC4B`vmxy5`zTz>JZM{4
zLnkX_qNA#tw2krJ4xI0<wWx-P^?yX~GaW-Mh{#?LRmX5pvy4PkFq1{w)<RkRC~@(%
zNE&2s8VOy)o}y25CQ8B2dradH6~VpUz#y0LEly#+o$LxU&u~cq@rW3A4Hk%^W`R}g
zyiAcZs~Y_q?YCQ$QB0)TLS1}Hu{_Z#Jps5UN!7StzhKi*UrhWACQ`t?05P%I(Hr~>
z>RDn%E53@4>N-`kW#CW|vWfI8j2mh4H3uLp#e`9WL+Aa(I9?9TC$ITDHUf)n^!&&z
z?5uB%r2h1y*#3+Rn-y*g(}K5-!N)A#R$$Ki>eCoDPFB^jwK^Y!)l(x4L2@1VDKnM$
z8uNCE8p^*`B6>H6Jc_H!Y}K(>6t=?}R<%@-A1+FwxZ9$B^*04AdzxG&{1nRN^xz=g
zTprD-FM+?c^*!G~ZWaXJE$^mT4W8#PHnYI?CqqHLZ3I>xLA9ZMXdF|Wv5eX9$}Ua^
zKG;VwgFtRBoI|%6;V%%4cm^Xh=b)?~@J^Yhv_1TkBw#4CM`Nl?88yOGVC?7nXf1Ui
zsM}b0cD|aB@Fu7S@n{r$c{9%2qDfceKLYmw*m@&G6z07h2lJ4d+$eOD2UX>#U(WaB
zfzHu2l7yD&*XSARbagcm2kI;zoDFGL$P?F))~oEd^yC!9*?Y5b*O3SU`@1SiBUZU;
zXHr*DqB0WK&!C(ZXg+5dVJE7si1V(XM5WlrKYUcl?I3V!RUOeULJg<AboDzVmgHF$
zRmd6s8Bo2?@sU=8gjmJo#zv@=27VmdJwLNv<7bZvK}CJ@s2AZe>H<Ry$|0-Xagj`J
zdDBheSBM<_w}l(S6Z_CB)HrZcv^+?I%P2z{6fXM>6dy<{_wY&gR73Al+Y2pZ=#nc)
zPvRIfr|%n7a~`#K)_5jkS62d3t)@gIuU{B1+INy&Ac%<uu+$C-#n2vTKS9X1iHKr3
z?;7iWzoA+%N@~F4pzg~+>>(qNK8V9V+Z_`N>uBn03oe5X6HvUZ8-f<i^8uaD<}`g)
zb2vMvC8f5r&cdMF6ALOB!b4}SIrOs>G_9_`w_Hhr`_yA3MtiEn!rl42$IhS?XHzL8
z9C#Z0AYkW&uAE04kCcYN8S@KAu}S!0MHu9)ZqoxokX9jo+%a=i_nE^2f&gWk_sW-5
zh^*lg9g!Uea(jic;+<4Z{@C7VqZoE_STMK-)P@acd=UMHr*wLPAel$lN0R$#z9;;$
z@~Ag9!?Fk1W##@9FgqvvcT3g@a^{m_P`Z-g)#)Q5#&@9Nr^UweIuxQ2nMj<*&)A5^
zlZq#O(<}R(Vgfe1^-j6UEgiR?FZ8>X1V8b~R&rlP`4%@JD2l8`{UFwLu9|fzsk#GU
zkUkMCR6L*q%QaGQTEfTH7XQrferoBiHNu01UX18gA0<Q5E()!4P0swSuO6m$Qfg75
zYp`tWzStD+ORhpwuk<lEM4rXE9;QP{i!a!6gR*d680bkR$v|9H$~T-w=bNu|tjeuZ
zhx^1S^Yptx6xgBk@rHpS7^PzO2`~o6-Y)#{Q$eTM0X&~qN(REB@Al^H=%$E7;3X_D
z(BAI2hB>E6WjkWSKOvs`;_9nKs;s$i(P!FN3*u4rf-2au$>33uGaXGHGhFl)570yB
zE!4!`?$ye363<Aab~9J7E$}>rtXNPsv}luFSq@iw3RBkmK9|@gSc-+Hi#=KqL&Ymf
zCBN&eF2^NbRV#cHgy@_PZ$8e>yyTY;)8=)OQdHP-{#-X~oV^RNCK^d^+Cy1cSB4^?
z_jU7=k7l^kd&i=eRV4$;j%qAZ52KaHBnhRWrwbeoGEn-Ah&W4p?8XI~_ZHtm{nFxa
zF2&Vl)iLbKy6VS&euHVALMo@h$qKTDIk&99Ny)+1n|zO56s{KQ)r8T-9jmM7bz=A<
zp4;cBOQK?`FsRWXI7jAEmOOZRDX|hFUOEvp-mh3U2P^Z5M=uiEhRu6akCAJ!gqJb8
zub`ykym44`7ah0XljNHdXL7~vae1G)1Yi(G#ZVVyl<$=tya}Y~kvOzoswx78gSIDy
zVuK$nw~D(C3%k^1$J0hgO@`r@F%~EHQ{1V}KKe#mD^l6Rt$jVC>ARjcZci3}!<l+?
z-eb>MKHh2%zKDJ<OW-Onm+D0Rg^jtSC?x`Ja))+U8Y4D{yFb*vHGSGVgD~gAG5<~a
zt7ixHj(uO?ZynsfIU4<+I=Iss!<N9Or>%z+_iu*?f?vMs71M;H%`KX{Ya9_I7#Mxn
z#Bbu*x>6KDohNS6Fq60@0PCZv=>J^tiBAWG!}gxHe}A`Y@^F1Jm^B)$gYv`VVV%S}
zY@k={Pf=sCvKVW1nKG+Ms~ZLHu2Ce<kQnvN-p!k@au3#d%I&xZ-|?=jcSC_VaUD|o
z%+3IFr=!UYH{1gAEbDADLKZwixSyfs_Uz%14zgcb3IWF>r}y^o;&6Sms~+4qMqXA%
zXBF<U(}cwSs|2KGO(L*2&E@f#VYd2i6ua=Jd!N<kd*1gnZI$p@!UU4L#mBx-eQyS)
z8_1bk1_YVO>4Sq=JU`9d^`pjEz+<Wz4a_h)AQ!#At5aU4PS^HU-=20|FV>c~myRUw
zKIc(f>Nw3ch>@Of>VeIAQtL0a2<yx<ODjsY%0gA?wYd<0ojR(#^Xuf+5ubef*!n|m
zV9ooh`15sB%JVlmuZ|H;|NXXtLyc5XS-#x1NXJwH><D~>9=4SfLTWO224PaJ=ynnq
z(~WK7Ya1hd<d5@{1~vyKVvYTrtjus)un|2HH2epCJt#P$vwFIUNT(^pVnpK<iM^se
z^?3r2pRun%C(8Pm3jzl1<Uhdn%O=EmY^k_DDD_Y(wi3(_WV#Gwz4we4s|{9*NtKP0
zY1B^Ib?gs>dPaR}NchPcO;GvFJWmsiVPPkr4vOiFxRZ`_eeWWv^&zlP<b105JJW7T
zo<yiZWb^d;WLulp!~WrFZ2Oy+OoTl=c_I|st2i&t6tH)N$Kj8QQr~PCWhI5*DWkIs
zh|8I2^-TMhO?G62t1S5EUP0D~>UqC0fym2oJLtRIjT?A_P+Vfzh-aa%XO>1#l5Z=}
z_8lfXC!M`D<VQNNV^4f3`$72rnFXhN!9)n=%lf3t{)}&d??+r4A)>Jb--WW!4<eZ3
zQiz?dU#i9ypgP|gGt<&Qg*z7HNjB2fRx`uW%7_=qlgn&#)LKF5e)@t6E%s55C<%!%
z?!9Kkrz(@(Q=L_3O}s9cR4zY+flqA9YGT?^;v}kOU6ks;%RY%b1)lC>am{HI;$NRI
zdGMmkFiT*crXgeu)&{4};?Z2$$}R^j_cK=OW{x68OfEu|$RPqgbyvRmJcwA0#?S-v
zIUUNGmwXgN=@8NoI+PM|h{;nM_oTh|;b<n>jm#-r54_^U163CWqqUaWThT)wj~~TE
z-}bhDiv4E;qbjtZO9}(*g6qn_<>fLr3HrKw`^85(^*8G_m#)p42i+Z2wsD|Dmsund
z1odHP7YesS>!^4x!`MoYC{$T#YC*NOg9LGVB&NHbj%nXItM0f<OV!G@u5hS+YMFDf
zzcb4P4MsxzQM2>tm%VhIFE7Mm{Y^Dqj2QpT`~1F2nS<$%3TpcHo=6ZSHnG|%H>BgS
z4(}AnIC8F-#MO2Wn|CUVzP64+tXE>r?BfP2jHzr5kHE~)vk_f07Mf2p$c=xmRc2?(
za{ngG{Fc~a{7FGR5Yd*m;y$<EC!`J(KKO%1%U~x{lix7qVma_dptUpTRbRutVtoG8
zM}I;Cbg>FQ6dP_<$0Nk4W@YRFH$8;ycNd@<CN=ebt;7QTpz#4TxE@$H6ugl&Ga*6z
z4PKR`!&PxMM&K(rspg}PuxW`Gj4Q<hSZ{(O_SOO%(4@TmNUrShPLRt$mW~4K<ot?M
zlF}Mf?^p7sz6jWb$$j}~7xtCI?3wC`M|nPhp(bSvEC^n2$P`7;33$zt*+>?hca`P#
z&Y(2z!eNubA+V_s%>T2?UG!Alm%AD^v;gjRQv#Ogs>lQ87nM|2Lm2wkNYi4P18g9m
zB4}c=5LXFq8q(gV?qq&+c5a<hK`+@2O>vHvp=(NA;-MR$K}<!*`i{<OhRJ!Eo`>8J
zzIx#9kcD*dgQUV0O6XTV&9iqKX`}*r*3qbAw(buUA~Mn6oTcU9>qn^SueKo@gou)P
zDN=ed1g55#iZk<OBO3ra)|t!dY6ic>x^PsBRJ@wA@w@X){35JRBAh7$(h8BTgTf_n
zwQ(RPNNQp6J@B~XM3XLq$vdxFyC+#`LjjSi1f<{FHGD%>!UoJ=;`DOin{ODuXO4+@
zG@6Oxn6n5h&4H|yB0W1J=z#d?67=_LQom$I{r)GT6fI|ZoK`P`QpQ4xQ~g;p%21kP
zkNvUFCU#*V^{B{{5Je_tJEmf^13dcO8k>~i!J39u89}*M%{w)t>GR}RWBjPbto%8j
zh1>C5O}ga8G{8;aIjI}8I9D8gcD5U4ES+Nepf<Hg6-*stVOx|!(1MX2xU5uAN+0!{
zgC!R@(vZQq;42?>vcukzj-E9UL!uiKS)TWu?jBTQ-Q}c!?Yzqyvf*&%piAS#_utYB
zKxjmd(^6WY)f1aEP1o-G2**`}$8C1(I}@B58CUbJ9!^m(eYvkOB}4sQe4E{1>@Ht{
zsIQKeHk`6d9PQl>A5lk%dB*{AP|r#of!B70NJHcS#JMWEV2Wxo;U|{!U?y)u)N0N=
zrKUV+JOLFk^|Kz&8IkwMFt=h~jLFQ@xU~jhs3@U~GZAgOHLu{MZ#9R9cTHJ1Zk7u)
zbZ9B8FySLK>>H!_RpAc$brjRq;wEXccwmd|nauk-G8XLi@<;rpI&r3=bel$Eb@@z2
zS=_?cCq<S|JP^rep-dC|1K)T#5(n*_O%d7*cD+Y)7Dt4cah8kI*I1DnEGXB@I&*A5
zZyO!f!EYJ1R&2f_9|`qEdDAu(yh7K4HN*vOk638Te0K-KG)n$CL)h)T49UWMtEEi0
z1|yf-mW;;zRWM;TsGt@Y5__HRVNJb~xJsN6EjcOavU{3PqO$RgWkMtjG1dcGgoqB4
zQXvs(S<(re7Op5ZPThUvfeL!2^P%oh%clcSJOycO+V7i;y4ElQGzZjdnkN)7siMTk
zi-PfiT#>f4P;khW9q-<OK?G?;?F`CqK3c|^&@eR)x<K3S_lHbL9HJ(ifCk5y$d1#g
zd^(tQ&2YD2hJ?L^a?5hzLDdjGPkMJakFveUMqkuVSQ~$QO6WSP@C33tDq*Daqbyx6
zi@mZaH)AT*-T|uP9CCm`ODbyx)Rt=e0RrsvH>^uIPHpQ3E4f-$?+Lv)s!Ar5SZjii
zn9^_M-FRk0`|+mzotKuddfmE>{g&_bu+h+9#UsXtjICLRxEhs1>h^cK`46{U3%sS@
zp2bnI>F6{eASlu78LLI^Hop-?FX};Hb9Rx}{#r@YJq(tp`4(e$RiOrpn2sn(CVcRi
z+NIOrl;F-Yjwk9=tT>*E@khO(YP~DdP;CuW@s-D}UwxPclnjBh>V!K;T4sI8keKGC
z3T~v9^XQi)ux<?-%HE)F2SyD<-@j|Hmlig(DJiVkb6!-a*M%F;G-h5D&fu&bUpLvR
zfVi(6J8`N`o|V?EqD(+tgITswL3K7HTej&)e<m|VS8jgHSUgQA{*iQW2nY6zXVs=d
zH*D8rAYZgBb8klxm`dtnP;!3i$nSdJ%!r@QaN#xQ+F!SZN!IA0f%=5n=$H1@otp^?
zoVEmSGKHv;bCfJzqNQ<iqC64tT;j!=kY%;;9`DiMYWVLQT*m)=?j!>nJ@Y@x;7+T{
z*e$WXYdI|&!=sl10p~N~=Gt7IOYAQxu}Vx?$c^6_)4acdmCVl0aa3KEb<ZG)o*=q~
z0u7hiKRA_C#ru#q+eJrbK$*ww({*t5<?TeqHl0a9wI&CMA#~xa)%31d$+#qk1cN$~
zgpKX#_CPNrD4krp8VojG%D3JXJB2fm0S>){tg7&Rj{>3?P=(?(aABNCb3r#wPnReA
zv3`>h@6bBnyM>!a*V`L;xDIuNg}Z}rtAIIOC}fx7_Co&J`)AdwtFteG3uJvh@YyAX
zt@u#Y<woE<<Uum;1*k=_2z(9cMInq0bcYIL`*hP!wmIS)J<K5T%p*o>bg8xu+^V;J
z?9_#SmKal|sZjaWJ%Jd~0_%}`vhUrQtbd-}?uS@VGaiW!wl8O~DH1w2N5Eld(jFEj
z{T<8NV4H%W<Gqx>qDMH-aQ9TQxY)CzhV{qEDv;2jGHIj!ps$3~V%r(mAJ9P?`OVX7
zMI{!$`q@zjPJXe3wDnuLD%7Zojt!J=tZ@hm9kfhEXsozz<f6M#Re`C1xe~m46jAk(
zTqJXe%RXqk@3*p`YrN09=b=qT(4><^wfWfa_2#RC--LsBIS#m@{DSWtJQ9tNeA*sw
zaq(q4$;t8<NTq?{LL>^lrcQ)Hntrp?%~JJr7#*xo8#Z3?2m<p=WwkoamtPDtX?7Q-
z9NX&FceDW+ypC{OX9Nw}VncpDH8FHMM!py59I5k3c4`R>P~HVTdUR`B@?WGkh<@RB
zY<CbO(NYqh+Xo5naJ1N;5RQ|Z217cfGs6kIT}u_8%$A>gYntbeKPE2Z|Mhd!>Bb5=
z)nEypUbmY|?@1R7q>{<dTdFCI!p2`=38M$pUcJLxHP+L^^vw4KGZL)(M6&ArZ0$9k
zo!QYD<I;-DHXH&(Ke@=}r%#2O<8SxrgoR|x5n)+=ai|zoGCX<N5r&=34T%UnTT?>0
zMX@TBB%1~*G9;dKSBH!b{{Zhw0_Vi~oMfPD9Mkr}O^WGVO}~YnSB^g7)xI;C@+fiW
zJHBs)bku5@2e8>Fg{{J_Zpu29m7#@n@*UaKy|dFTLfz~MnHEzmk%m-Q2Q<?>QgL-7
zmUO3vL9xF)mF<&Y8HO1*eWX9miLGFbL*g#E#oi1yV5~2>bX(#X>I!2xU-L;X)8Q^c
z8CDk4w9BS&?6dk}Y^3c5K5Wl~mv^eD<)(g$jKEu|tJN79g*x>Y$vLX#V-sa1#$Cy*
z)sY=UU(Qe2VVSpM{mI+`ucx+1Q~l%{4Uy&ZeULS|Cj@mP(|ho4Y?laPl2e4oj_up~
zegpY8#p$Bji1Y71H4{s6Y1mdEnb0soS35ov<fF}CBWFyKTxbPH%C`@GY+cHrz1emZ
z)Z1}^={HHZ=9D9_5>ZY}*WFeARQY*V8NX!CC03kAt3NZoYJ{hj+)qkYwZ#&$hgv{S
zw|@aA;Yw06kh2yHiG$qKu6=-%i$17=+jnAE9~FMO*#uHldLZGLP50+`o@#2@kbi<n
zV!Qs_zPsON&vUE#mkpX5oxuW(OzalI1&LM!mVhuMHr>F@Z@5exDyfQ?ms(2cg}rg%
zcD{VpQF2b4rU(Ru#Vy>X5}T!WRjErYy*gCU76TC_R&}_XTAJC~C)<JJ4WEYQ=QPow
z100oglVKScVtVaFPglgTdWFe8T0z=q;>#)Zc57huZ-P27d6<{l%=unp*H%6a|2UXW
zJvq=lo$TBkNfR}G(6X=1*7t*VoS5&^#&Fs1Jv{1+Lg8%USI?}H%C6&ECs%98Ez1k1
zO1IVoj<l|Tbd2Y0a(6>YeUBS^r7>VGy}E2(T7@7mgIgjYRz*Onjv9S<y*kCriA6H2
zcOx_GhP3dC5KXCs#lvXqsqFk);$2AS<L%AnQ9S34)x%93sQmm}XX9^UX=GUydH7NJ
z{+F7wmg=i*ITWNLgSO8-JKY1*C4sh81i5b=&!|;jM1Nq_-)%7mv{>68gA1M3)yCWV
zJ0Fs%CTqf}7noq}_e<c{Mm<sRDQ6knnDLc_<E=z<&6NBi_Spkvz{~V?U*&F4@&p}#
z(<A!@O$^#yBeu&T;*yh-xi?v&XF<T3qPz!@6^+0(Pe6@`(?MBg<2H;Su15a_byL&V
z)KO%*r#XBr!VtHy>Yc0QhwJkR*wC`?U-H<cX3_Y=s)s$<i(sQi>eJ6ToNPjm@v>7C
zv^<7+-n-X4>>DafX%8(v?OO||H5~**H9$?q`DI`>^^X#7UwLhIMehEc)noe4-|;ii
zv;VUf-D=c1><sD<M((J3<_k5wt(j?)yD4p}m13ESSSKkfX~bZP#h-61vESQouD=f;
z;Y7FHSwGl3M3U2a_85=H!_{u(cE34TLsUK|T^Nklp@o2CerCbMgYXLM$zL1V9xRJb
z$E~S-!$)`frX;mX?&m4=%R`=3sswwN<}dxU)=*vCdG5zK)kn<-+%x1ZeVy>pl<x2j
zxpt4rn#LpKdrLYwKCin)pA%uPn}bObigbNzt{@?BzX!x_Lsfnl9))9~6F5Ifyn!;_
zje%3w9*qg8l_C&D%)9=(N-TxbBJj_5IlV|41|Wf=_?HOFvTtLcG#Bl@N+1YFk(OGx
zwO(Sl8y4D67Hn4YN<OSVxAv{M42Af?aAulJb~{Rr``hVj?-{RU@=o65S;^o{mO-{k
z$D8MMeBh-dp^Hri-pf(cY&6++7k~2-P<*`2Hg)E?)r*HKjz+6y7|pS_(HHRR?dM$j
z7Y_oH2rR6IF$o&MlA%Odax^VNDUb-LZxPC(Db&6RtPdbKXv$MA?T8ZZ?;@EywwKQo
zYF;Lw6B>!4cCb6cFyu-@Vk=keCf9}M1gKO66+G^j(x=8Ns6woKX;-tguF1@qDUWYt
z%Auyh@_rtHZU~!p$*cc@orb_vD3<FdgahI0cbS#9bsj)KYzEa=jnt^T83RrS%vK>e
z%=Ps*=7)pz#Vmz%XRNz@Ag<hOc}M{(wpCIKh`om^I(VuCwU5+0%(PcPO+_zhWCf!)
z3A~ka+S*^b(lntQ4d4U+(ds<V$4!FqE`khZQXKzX+d*SnIqb>hGVe@Y27i+&xkgBj
zWn-lMIK>;wJ(T|T+4sW<`-|6UvrO6cFLz=*@o%%Dps50`W9JwBNq5>EohulqQqLQ|
z^bX?vVEZ;;B1Y9he6dTLI9qCke<!m4OX3GI31ZbX{zldmihXlN)#_R|$re`=(W;=j
zu?g6*a)@Xt>q)nQ`SN0}KyvlI@C+nzKsG+^SMIkP;^<FJIr#LxpZfZ(>JQmu^uXUX
z7-fTO#fWZvPMEg>3wS4O+q-k(nuM>JCSQfnGn4r-Gp1VtU)I?&FJYri_9@OKel@9Z
zFJbAfl<B=&H6a?`Pt}})$2h%I4MdTQps88YP{vs)>Oy5`cTda>JkRo8{8*uHu6H+G
z=Z6>Z0Wa9-0q{XGTF<`@g)l&jk#`I~(d2*c?z93n^*KMgahSD~Qa9B~k-PI*EhGar
zWjgYcm@%p6)O984^zHb1y1qf<FY!g<()h#``TcYtwv*i(XtqO3DbPvbOV)X-33Jea
z&-I~_MMkx!Uk2RKmcO1X+&PFzsZrIffU@uruve+)K7Uw~flS%wi(+fYI`%hAcdLA{
zsTR71k=X(>)p4{B!!Uc4ZvcZZP2CWipmnkrP!V~n@gnE>LlHvg2=biBN%L;~TCn?e
zg=S^`E68pPY}L`((%aE7Mg;<<pwv1Aefoq5k;LygrrTTh7BEtCdva+7?dCBt8-qqo
zF|ihwbjG4!U8Jj76*4T}Z};t<98%Vd=D<lOr%&Z=rc|$}^PVV6GB9grls?_lBps)N
z;Hv4xlOA8&_&-lXbtYMtG;>u|CzzR%{d9yUwx3mhz>b`C6GZneSug-;T|77EFQYVy
zx6th=sK($&mAq(2<j8=jBeSrGQK)_n4xSIe(G9V>Su?|nAVLSPU>$xG{#td66o%r(
zazV&*n*TXBTHC;J#?@q#d6q)P+dn6mVfWtGr4^w_z+XDM^Q*^E{&+1bqejsO6iz!E
z!C!=&8AkI}`2K3RHKXQGTbrK~HIf6zzBw53Zq^{yfcXtVXsrUr75ReHU>ax9^T>Yv
z(9^FV3V9Ua>}|5WTqzgnUK#f>El7WOn616KMDv{pPUPOGIw^w#DofJE{ZMTC8n8{@
zaZGW$?E-asvC>qY%!z(j^w~;(V_(M(H>RV0Nk&(nzwPFnwdoCM5=80K{Evg(_i4^x
zACp6g_8!kjXZ>kqP9T!Tua%oCCygxRYL%yAB#I5V*4?Wcn2HD>r)wC~Vypwq7OHlx
zz9c`W?^@}>hcJ$xE)^D)LXrRQIN{9*n-f!+P)i!5Am>tE25+*R-62)^qR4^PyBE|d
z3OoNf*=`OAagsk-D2h{~qsG~~<)Zl8S;Il2uZ)@3xsiHobh`-GCb{7`vs&Mt3&mpe
z9LdPL$Z-h5BF;Um%FljT+7nTP{w8>0hv}Xb^;+SV`Ub~1m(|k5{7=51;yQ{HT#}3n
zkfmB5ZDjiIZ{hsMi|Nhn#Gnq~H5n6`Mt-I$Gh>gb3iM_K@k=i!T=hVadVT9MgO+Rc
z;l^FFC4E05bL3y8TpK*#uNGH|k91r((-A{t?Bt@cG*nLN$x>qRz_j}Sd%+j3M@qG8
zw7DUgv~dYFQ1dc6iOgmRZji(%>fbaRUW|_QhR~{V6k(=6Y(c#x8gdytQRw>s+xbE(
zQ4KDcg@NjEzw%|O>F6kRqH|iN(awe2$|;T9{ZZ2Z(nRPMc$&yW^)&CZ69aCNZs6+3
zIj$*HI&4kLAi0%)9h;YScGM43>1MFD&@I3&-uQ5Nnh(I}E@{vIl-!N#+9jseCzjV^
z9d+aK-Z{rC7?~!aInz%u`$1hHfg2Z9G)<LyOiP95IZ%8DG0hWG#}QlFAGwx-m{?zX
zdz<rSED*cO!b-CJeokUxT+9zJb@l<&T5WpZ_q^m+O!kAm@mhs=@bFpQ%t>U#93QDv
z1|9?3Dmh@DyC905_GyY8<ub%NB_v!mG~_(>L>30w)pWg8pR=M$FG5bwZS_~wdv~n)
zJJ}5<2lfMh{d)9CJN$wmqEj$VuR5DIdsQIp$u1A<c?%Eu4Mk-uRkjfeJtsDuZFZBE
z-Xuu<+W=&`bLJdHXuuo-#Ias$JYzHZqSKcAk&*vKl-!1A`b<mFIPXY?IO0e!DkjhE
z#>g=GEVZY$+C=G19?m6B=$1~^t&^dM2A9Q=q3`v!Hr0nGh?@-4)0pr?<28rlLI3D+
zV4M7mF58JS)<q;u7ZT&z)%0|COv7@nm>kOoe+ZQ6fG?`Yj%sbj+l9Eh^0afFZ&vLz
zjdC^O6ERop5*x^lof89d6Wx@Dgz*oo_GZ^<Sn_-acuw|+jTy8ib?=xRd<fuso3z`t
zwGYm$K*1w}EGV~V;ENRruyAF~yip!mmu_g?%j1h}%inBycq?M2%w0^Lt}dA`2`P+s
z$A+lCota7J-@*9EUfHb&bz`vXFsr5C*S9QTK|n8zN7X1m1{s`T@XNSP_<%6Kxo>xp
zuSwje$=7x{vf2Guh)O>gwV1IX@ivcdc?$F-gFc5{xwG6GsYevFOWV#|c4W2EpS)Oa
z$Z&t1X53l0{7vAR+)E1urd9C-YnvEzUl<G)mzwvTS|96nE8q7Bi(4svBL253p2M`&
z2!7|J!<vn~+RZ67hbf8nUi;>iih(UAJu)`UGmXvaiSe?<+KAxu+;eSk=PGM+Eiew}
zonqT34`D=`PVAL4`6(0Kw;9~*D^L`H+S&2N*AbNU<>ZuQD9)=|D>{(x*1fNjLy!_G
z;bH>6j(iB_NGEyMl_q`}op1+B$mb!leyz-2t+8!`(kF7AtRZR$R;eGVdH!%L^h3Za
zdOM155^;`JLUg}5<@_#pe*t+s3Y(VV^gS^}6;daBt9IJ2=(c_y(hKr3zt~0E?){7q
zA090?El!^Uw&&DFd(;?hgbb@H{i@Ck;`qnyYl@M410T)Ltv{BRc?+5)B93rb$rQex
z`?nXs%Z>=}yIHN~;0rU?-{P*-N5oE-PwWhj)>%LbwwN#{Xo&UQz#ETS7!boC_-g-L
z-G=7o3?(%498k4bf8W#-z|pU4MGiGg2HrrpJR>`TuF|4un~XJ<Kel0Yi|D4vwT&`)
zr1C7IVL$fP|8z<xGHJ$^@KcCm&j5)N6lzSBM5(LL{#)(sjEO|n9oihROHcS!DGkA!
zHkftC^H1)i$owmESnIxcwq&xmQCv2i@k;?n6pZ?(ReGTqHsv>SfmdI%ziE0m{;Km)
zBtH?SZ9IN3pGs&}Z^(AJ__bs5@^UxW^#YP@KsNDrVFmMlK3bcBo$a566(#CYR*UTL
ztq-cVAAGmpt@7*QJ8pcwnIWv}PD5Qx^A>q{`M^aSfQT1`GssI`No;#-v%(az^s+Sg
zG`sxE<7T_~as}F|ef{~V7kAGmyA6Lohs6yNzu5cfq4&AWm|}dwAzMnS09h-{)2%VN
zD;iXeXiXRmJ1xb3*9=?BIrs1<^JT}pahlvI^2A$vwI4|EhS>&*3;Sm`59wU}8Eq+{
zt8RMURc6o?4(jmB-k%UqWXMCZgXK*GGkHpWcyv0+`=5~htg~lxsrIDET4EYibsvWk
zGI?(!kj}V#x%|uS5d!)b%Zj?b?=+K@X)Rm;St#_!Y^Nb|W>M|^O7CGEq#1TH!gDat
zG?aA4=|KN7c_DYznW?-k#h9duNi2;$<;!ogLameNF1Z+@EcI^4rp1XX#M-#<FrtbE
zNaHexco@4O;<NTlxbk8>329oxHP9tnEb7Q}XWaB2(RUH`MO@}MQJs8?Hn0X0-voAm
zEeI?kU5KQ#e$8%MjYoyp_bb5Z3VCBsGroapPT}CkMx5`_v45QA|9q8>Y$&e=VLcMh
zfujDNn`<;JQDLT$dZL(kK=<)pVKkXlE8k7KLk9iRa3M!?m*MyY_28E#XoBG6bW|-N
zRE9EzW1i1pL8`x67Y>IW2CUIg_`g|K3AGI8R|%RZ9dSC?2-eoGp7hr$$fP)I7pd8W
zx|FuAq3$9S#y<$;JQ&`HBmd|b(09CpucY%792F4Sp^w6DgJQjZc79mkX=yq2d{51~
z0Q-6EIXztKF(YtufXJ`%=k4mtiPur<mD1t&y>HF6)mv#;W$&Xj8&)(hz}pQv_RzC)
z^(NC-`EHhzPj!q_;C+9rO)Eypq$G<H9S4E-m7iDy2}#;8$(DiFkCkcuoY=gSHDNgu
zo1P2mb?--y?%aQDSq<l22|;@g#SEsY4r0O}kKo!&*C`gzxz=uUuEy4N#t#kkj(sT4
znmkZY;tB+tl9GCj5Y7Q7K#<@GU)IH}1ry5;TKj8aH0_)@GEUA2x#f>O|H+Ti_ZccT
zJjdUsFuc>BF0XNFVoDQS*cM`LE#|20uagv&ys?EgnoI;4deosuIa0&4nO!F83=24@
z<lkz;jyfGxWM0A-)F2C3<%uV~Gl2{$8?=>=`|<WNODzi$PMydaF-~C-V-CJCmuPq`
zA&B^NK)};y7hFcP^x=4^%Wl(`<*61VzRs`&WYLOh|1@LxOOkoDJe0$@s&T`35VA3Q
zf6`LzSCbfdgXCqiFK;T&GH`fy%|pOA@RK;Of@iQxk{kygW^+fPt->WFqgnOCx8b-%
zLcPr%zo8lNbncuj$@CG1i3s#VNXF*Y`L5&(>>RMB{wQeoLhRm=kzpWT@66EUFm6R}
zSG`G8*nGGttVSlyhdfahp~JT^Evuhn`?M=CgdK}To*IXE?;Ud*GjRjWLrL^*ACu&=
z&9PHJd^Hb`Vstdx>$tCQLTcFQ{&qY3L+AOu&d|NfQr(+#MYW>>b(&bP_(d6Vk7diw
z*_4H-s=)QHi_#Jr)|PGM^<^yI+XET}vJ2znm5;Bci>Qa=Yb3o2UnE&lnjn>f{oA-o
zT_dTg_k3iQK}`ladNSu5otVJJ6iqpqVKLD@@f*ggra|nekep(VS<#@SB2L`3s)HB!
z#Xg_dB2at$Y*?=<hOXOY;t<*=FWY{n4vTl+VfM9yf*6v_y}xbRA0LDN)9C}sthWI@
zp$W^H&<wn44Z_BggQa+J452QyMAtNQHSHDs`9%RE8u`^ulr?3#2yt{*_C^@TSo0lk
z9DK(6yGf1$MTY>Jf~I=H_LO9YDV1Nc6ww@(@mc}cCsHVVG1cGF4%CzQ;ie_Hj&9&M
z<j;jP-D9xT&Cs^WYgQ*p#(tSbb8`jlDmyXs7pCfMU<_j!uIC7cj4LPH=80#zT=1k#
zf;ViTBM6vq9YyhYxJ^_YsY@{~DrB6qsPxwDhAaf}Yd=BYCgP`HIIY2i+s1l^_Yj&2
zW|a>M3>{>4baQ5txGsBfORk7yC2d<Fu(5yBRR^U$?paMWd$!ucRahW%ThUPfJAA9P
z?3=K>O60;nVPPVC$~0|zpMOW6EnIo<(?<D{W?JrN*ev8g`jboWI_g_;+^kIbFYU+M
z<!B2{xJcA)&k6^QS*O!%MSi{ltkz;F+o+Cm70j-6b#@Ujksmp>Z6V(aaeG1sOFAtr
zb+-=Xy4Z`q-4*#5+q&mS6pK$EXRa*xafFLQU|M-RT5eqB9kWvmXHdHJNDe%oCDYf>
zF~^NJ<ajpRA>i>3V9`=mVpj4Wgo&`>2qUcq595Y1HD*U%h$6y>r#<4dj>@~LKhwHv
zPU*X9wU;l$#+NG9NnX)`r~mpH?ZTpji$J1eZV`UQiw7i|i*q8pg$<8eNFXVYF<jU0
zVhim!-POvF9^PUzP=i{?*(u<;u1C}&=Zj$EIVSh>7Hd65V0L+=x)#)_T+Ka0&TVGs
z{+)4V`X659QgX32r28cI+00Pio=(x>v;FJ;B`oxe4e3-(4eU+uS?THN#0^c2P3-ZR
zflEPCdpmhUTOlh8Yb#4bOM85FIw31_D_cctJ$*wuVM9k#eM1E?K{{zeOJkrl2h%_N
z2I6?xr8(3A^!8U1a6E1)i=Yy6qNR<HWVJX2fiQ{aAvt0JX*9{c@|_G4LAHpRWIwr?
zXKsGA39*y}x!e2=_3Y_`t@g;_GxN|-{l=7=U597(U5?$fle=r4)td|Mkrt;_UWq8;
z|JQ#LUMKxz>6*3X^9-)%Ep9iM>seN#X*|%kg>4BB!&0YSxa-^9H@pug4<xzK(1|hw
zOX)U!aML9kH(ek2xJ98Xmm8;{@H~>)tYq6&wT>S-C1$y!ruL`%Xs<Y{TN`}#4Frv%
zJ@2;TQ@NZlA>jrp?bsok<fJfbsVWCUM{UP^w=&pG=GQZ%#o>Z{;&9*w{-yvQlH;!`
zwh@2T`3FToKG<9MkThJ_$dDX;0&wb%>?VSk;BbF-krc7<R~OI}|8CLROR-EcD-Lru
zm*Y_am(xiQ3f1cn@F8Wy;nLb`tX5kM`lBQyB)S4Vgj>1qWd^{YP%71!O@HnV(z5L*
z!D6%A$}kM@!Dc4la28}WoyfieHcYkdc&mn!!s&2ue|4xY=7)iS@qzmc<W>Bw4Cww2
zr24N~Dv(t(HD6|HEf#IZD!T;Wk|MCB`61aXmp%cTLfCBf;2gk?^1VFjb@_YPWLUM{
z`TsRnYk}))rCLpxK+Q&n!!CEd`AS1MzNbq}$j1AFA~rE4eCMl!X+TGoXf`mu(YaFd
z%ExK&8D3m6;7p8|&jud({CF`M$PFrxuYbIDvfKnLHm)Yuqt~Fz$EMXVNJBwRT<%|C
zkT*y3cAKGWtY%Xty<s>3FM_XgPL34i)9av^D~-WuAl@B>vRLy4t@QD%>w}*D?no*u
zDmA<HT06l*GP5b-e+;%#zz5=qLCS7t2pTKh{k*>zE8X>g;<dXz0ieH)*Xd&<;2{6~
zfd5ngOxmAS{#^K1A^t(pD-f`&kF4cif%6CCpiSX{NB)VOS3JG;N}B!e7cqAN(EbB+
zh5*)IAt(R$1=!4R*fFmUA%^T0P<w@}t^9vcU=#6Y&ELNV;Q+Z`n*gKu>(>9(BDtxV
zGMzT(UmH53)25TT!>L^Hub_LS^;+P<(R>w^N^wF`QWB9^Sn;Cqa8hhs+)TM%*ZKC~
zs_h`nzoeAwn<-Yi=zfN?E&rG$>hW+LibEMk{&6EKocE8V{evn$SdB+gis3Am>KQmV
zTmTGLB;x=KoXCOw7jVFBtXG;{){TIlBcpU)k0drL&-}6ywJM`B?Uo*FlZ9$iHzKo}
zljZfNN>wqfxe5tDXVLJuLTW7Ae_{2SzStebLvmQ>km)m@3&Y{6wOUQ<BeIspNn!b-
zlE&?_!Itj45m;2(cp$IiwqTk!U8H)t0Ose@OM_4YjGpDojQRM-I<x6QaQA8B)OdtH
zNE9=y<B&x<yHli4WGz~iPkT4gm+v*NOlrVtzTD{WWwumi3jwNEvUj9odoZ4igv4U8
zmd3Hl?dAlB)81Np%9MOpr0T_GTvSyj-G1@Q$JT?UmHNH&A{p^r5v1d06Oeq(2HR*d
zxv=G8<)WF3QNB@uR~Q5Mk!$W0(0O@&I&IiZi;0P`-kvv354FH!x7N*(jO_`-0hl<;
zXgGk~L-YskU$xO=t>IwYiCWN8u`oFk(?w3aw%KH^Ogybt9IcjXg7F%_ia!V#i2a93
z1r*RLQ2>Jc`&X}2`>XNq%RR0BKV%1J_R6r=;y+b4aP_LKU(5caCUEtejek~o1vxnY
zJ1C7;6#=lJ@)iF;BY<xB0FwZ_11L<O%`3Nn{%zJWJmU>_x8rqWxoQNf<EfIgfq8rF
zbIoR57XV8X=2Gm`>0%%4^7F^yA=AG^!exVFbzvXqzOL~5r$YWQ0zjbG0RR&IAwM9`
z>!bb{5^(FaE8x~2!v$`=j`q(9kpFA^L4OV3=FidpSr=&jXXU?>gws73l_rT^uQP^1
z(eY@m<F9$u6=VUlEntUUjTB%f0L1Au*pmD?wGlpnit6eMKLe5UXAft~_;lMr{tyay
z5K`ViER|dWJpt`<1cA`+#gR@i8cA_kaoMf0SiJrbOnWd_p^tth#B-ivH?qG_bGQc9
za<`N0Mrbx$s+G)Yv5=Kz`FQ)whp0p%nlzH@Q?>M+iH<|fB`lF+`Mmyq4PA&MF_C-Q
zPcvtSs_!_Dn}d0(JY);1JfDAlH_^FD;jn`M7~3+sb9#8ZSZ4*WBvQ3@XlO|PK)23v
znYQ_?6HVJW)9G{tFt)b`MP+bTeUh=1ES5{RGhhy<D;{FPelOhs-Ig=6Xpg60?ikgP
zO~wccrNRhFqLa9h8+^uF4+?h;g^$+WmMBb8in52iJy>;DDCGvqAyl=nDttom#3?ih
z2UM3cgbpM)W{cI3){9zeCUF{y41{lSc6d$?D5oc@?jAU?NzDNz1&GJLULj-+7`^`f
z{vwr<pyWA#@Vg@6t3m1N3<am&<g-1LXd)+THrCpg-YOgiOUT<ua0|%TLnpf;SbPy#
z2P7YP*UPmyE>PaVI{TJ2ZsItR)wpsFSXJ(d#ePXSZ8jl9!<Lqm4Du3XW}homruvWd
z8l-FECL+?G%Kzwbf7zeT_cB$W_y>Uilu75y5r$%ORNC#1gma(w0b2UQh4{aka{Rvf
zUHyMy>0iwJ?<gf2`U}VZJN;ndvH8DY(4YM4epMC!BmMpw^WSQl|DfxiRDNywUvcpN
zpp!M|CBW?TgyZp<jAb~jda%7?{}(qW0i#K!TB_OR`P6jMSOj=T!1`_fArzb~y^YyI
z^*)Z%(t0!bpDpMaM!)Cg=8gcq8esKK+g_fpj~9P0sdNe|^(T1jV#Pe!!<iB+W)t}a
z+ik#nVIOoT7pWWrJYEHR4>)Y9ubjbRHJ{7LD*ZGMGr!vL!*D?Cp|B`dC*cW4de}x`
zRlLaFaro#|H%!w7*YkEAO2@%fbh!_QyT#R&)k`O!%a=x@miT-CH7qtP)knwUN?Ac+
zB16FEGh12XLE!@+m;8rPg`9!+Y)jUw-|lZt+n9cI+Ux;-s^dXEmo|3DwX!M?-6Gem
zmkJq49<kO6O{bLPIwBxRHg-c4?UQ0m+5Hw{Z)b=3b1Pu_FLGso(`yVDtBgl0?9O)6
zz2M+R(s;NWFDiO*zf5EYUi=sd=A^@H@hKn^anD!YxZLdOXnV$9N6L*W+i%V%EHfHT
zDr>n`KXPqyJU(fCu>4I1q`bOS&nDqOM1W2Wn_*l*XmqX4zdrx#cHw-L@y+p~)7e@_
zz+J?<->U|V`}^vDHFp1vH2PEM41}G(!1F)Z)W1P_b*s&pQ`z7XPt=c6?Jla{>Ytky
zySZWWSi&ML@cjMUrcc$#;GYh{?@9TMwtucF1b)xb|7;%rukr0)dPHdku#yb+dt*QV
zmGYNq?f@?q3&(Z2*m=50fBCI>bbnj{PJ@ww0khT0zbXoU0bI)c{rzv9SeEtQem56B
zj1<G&2XIk;j)zmMkR$0l<N<jwTedp#eWBIk;T7Vd>gm1rs}7?c2-+jXk+>B;H_aZl
z1-9d!(JpebgC}s;&zy3RKL7)yIG_TklkmBY$)dc|p%n`N>VRX$g;dYuc8$#i-Uh^R
zg~~+`{?AyDMSvSvq2JpIiOFHRRr(!<&aTV<eG?&#@yg>Lp?R^DQhbb}Z0KA^?I<Z4
zCBnKBI&HrTjFA)M=}VhL!-d#mAuQ@9@UWf^t@)-BjXF!f)m8<q11zilt}bB80hiu+
z|9N6!LYC{e8VLVW(RW4c-a@|A@(!@+Heg0_C8%E?nZ&1VHTzYy`SiaH1{8*3a10=S
zD%y19)p6Zwt5XIL<*qt!hE^F3|5lUqZ@4;m$pA0y)e)MqG8;+Z0DPF=cmH=7{pT3_
zYkB^%B!5xhPhI}+1NPtQ>_s1vD)lMgDzeselJqx8!aqQAH7CTRL>2KzLeWtaZ~(6M
zz+b|x*{rP%cPU7wVjLIIeySZ{TXp|ye*U|M`+veuPH&Y`O%8*;2)?I%xfa(e^>W=e
zzqKQ@GO*U_b(TDUdrGU-xHX*Ybbq=3TN{wdss+>mwyJK)7WQ|GK()nM3n2OOFJgkS
z0;<{jEv)Cu(@lUYKM>=THlGh%LmteOEY^NcVltMknmJCYv%p&?VKF`dEEWu8Rgq5t
z;PFRl!^D(uyImhyjMo58@zvq1mDHN!>B<uyKfQQNTpaM7)La*LuItr-R+A%+dj>QT
z=}}<*+3K)|dT4g7&yVKp9eeIRfnFc%M=k6WHmjAiJ%HeD<+^g9@rj8p#R!_^x&j3X
zIocjq)2oNOq=0?bIrGl_*!JcRz>T?leA-2&)%*nH;Lvn{vw0rx{sm62GE5e;7(TCO
zKJ2~PMN5E|e>9=DxxZetYU@*||2HD0S6=;h<L1AVA^Zy@Y)t9pirc9No|eG$dn{Jr
zrh{i7RkSL$u+bgN#KiQQpwj&s<WA!gcakdK52C8t`+4x($va7$wg+MX0uesKb&-gT
zEHN1k{kBK+96FKND|T7#Be`DtGc97(kUn>lvZ2??XuJ<UfXMw1FYy0QRq+4KLdc;*
ziiMy@o;H(WZ|Pp!(AUUv0j}T!yYrWTFl=r|LnEUwOqN22!<m~UyU~dpDK9k0CWphg
z%_C9lkhQ2hct|1}dXg*}NR9*#?BY}#0?2SJFi5y!rGg2-0|Xo(^<3#hzySPSv;=%O
zAo*AKxA_J@0B`nsAhOKP&IU3V-If4<T7&}Neh-xNyAPo22Y^Jy;|`J>x6QP&D$YTN
zNDwl`r}DLaNJ_QJEfNIfOhLaG+ufaFuP?swTZB2bGfgC&WYRPaqGne;2)ue>cC=Cy
z9xBykocmjfc2kx(fZrBVkfdIs=PT!|;93>4f%3y+nxdO(r#{0aM$UNwFN$h9>^!;H
zP0@B;0io>xF>m))Xp)?43**Y05kao6Ge@@2&2Gx7iz>@~db$h>PV})Y*W@mmaY6Xb
zH<r<Mw_lOAaUh1mPT|)i&~w9ps+^o$+t%D=UE9;~90$AC^W)Ui6cB{Y4X?Dg)tQWM
z$bY~E@|=K=j>|X3G|GLUEi9A7Yzl;7r~CP@jx-R4g?4>p)$8;f&l1B5$g<iS%Y<0p
z!{YONOkmJQbO8s{p#2f{QoW77t?hZG;oxUBAjdVABNguvMNA?MaN-CEe(NhepY9P!
zB;7B^#d?IJV%Ue1Sg6jN_H*LnfDi%5pFseHCKk($S3VzjKb7l#%Zzya3xJ5Wk<kv|
z$36fQ>-N%QSP{eXV0OPfb6RmBJJfnjw0=h*YJEJH5^|CUved6x+TGFgbcanKd%WUt
zvkc_shEq704zGdCQS12tRax8feOUlchepL`5%f4<KzNrQs{wyM1dEN<T&K;`W9C@L
z>&YP@Vlq#T%^p&w`2+7Q{<X{LtJCu9%&b3(L?1{L13V4qd%it%GagOz7z<!)z1;|=
zC*XTHViIy{MpoGr^n*d6QD>&0pfE|el9{SDodn|gW%rBWP)rurU)^Z>i-77@YjwYq
zP34TG)w<u$OK1K1>!jso1;`lydX@dqiu*P22C5%S7onZ-JY7y&wOq~2U7xl-lmfZS
z5_RCM>B_+o&ckAm{5K&iF6Uc>B0;_lvw({R7!){6X7XAOAXldcQU+L*I2w&OYBfe`
znTq58hq<?as%rh-MoqRss7OkvNS7iijVOYMgmg%6Kt&p9*eWF@Eg;<>Ae|D@9n#(1
z-ErsEBb@L1|L(ZIG4B1wxMw&V9E7#s^{)BO`OIfNb1nc@7b%FSkhz-+KNjIL8+mr;
zfPv#JKsB*X27kS$qc>2nea=x<Z5l9B*LtVQVB{0mQ)e#}Z1m{JFDm64sCCIf;<feV
zV1L)2R+8#@wWJvL`Y%E9d+)32L`UJo0<N_3v^$M~c}_nPSPyNRCAbC1`wC&%T|8nZ
z)@~QCv;I<m$N_mHETSk0y9hWJaD=Y&?kI?az&xmxobN2tSSVO+>?$$0!@vp!D*ZHJ
zn8xtWTc$Rp_UF}#Z*E9t?(eMs7;&Lj?*(%syjG}UyG_Z+I8T1KC74?Tj3%U33O`3H
za5mvw+e`#%^MFe!erJxSP-Jaku2S_Khm<1|PXF?c(sYYoPY4`xtouq*9|u9?VmF?!
zpH*YW!e!a8?KJ$ALdpfd3YQ-TWDR8$#%yY$-h0o)#5{PZNC=Bo8#x7-br9@q5X@7}
z3gK35cEtuAHlNNhypk&%EuN;C^HG5!isM2j+7wlatkViVxw*M*rF4a`Xg=%uaZXmN
zKESY+Hm)4i&)*fr7F*muvXu-l6}3OBf|^~L2_LZCoi|7#C?;}box5${@0@F=R-6MS
zwT!T*SiYlCNtiR*Spv@;m~Gy<c~;<?A?-^;VPSH*cPd-y?%e^NX;{s6_W?qK`(SId
z{hRcUgPqwjo^9CvYn&KjLPGNJWamzTjmdOuepWx*U)twN9$@dsUetQeY@yFbl!Nwp
zPinP&4AlYUJn&@Q8F3Mu@tcFF!!aK*WiSYRQsu|~Sq5?dH|CWKEi3c@60t7j05(a&
z1{2Fr(`GoJ9KO_>5DF_@BCGMEmj!-J{Q<f4Qi(8gp~>KZ%AzT#;wp;iXoxtwBOkNp
zT5qqwEpo3lKZYZx+nqiRSqzzJaoMmMxEWv?2(0R(g?!RO-@dTE)!N!RO1)faYtAO>
ztfT_i<;|Nngf(!yWmA;74=4eJ)$MN%>QANPi(lMLQ%mOgIDF4i-t1cG+=a^hl{)6M
zAUJ9l1W~&iValI52|r)b5Y|kh68VZrc>CuRX>pYu(2nHy-^y8|(xFf`T)Q+}Hrtkf
zi>fEm4!&bJ!XITuh#<EZ^MYKy^1G8m4A^=as{X?6Bspr&@U0bLJDf}6{3Hi%rxSfW
zSvsSNrIGtB!ps%l4<S)wkp3o>%c9$p*;ViMZro3HXHKT0lr!ICx`nxTZ^a8MW#k;%
z0Mw>kQ6%b6bpYqGGkS39Q`#)Ls`WtcWe4t6;0+E^5Gqw(c5voA78P*718g*X9KJ*a
zk1q|it(1-w$~}OSeyy;KZEGsZ!?A+hlzUDH<3Rn8p({x~W9$4EgWxT2pV#uu)d=!o
z=YjsK%};}53WB4%(jsJqK)jUsrLJ715x>5XW)I(X7=N`f*Nt(z1I~Tfz#I^*pUutj
z$b%KK-c;~XrRg75=~R)?^j$NpEcJt8D%AW~fDWqjXU6f^2A@}YVOt;7_NY4*oHaU`
zWh2BNOO?glb(`ATUUkKz-<ae*++MCC$B|XnKJf^A-Z!Nm8$Ums)@h6Xm1nFSMW7FX
zPHnApK$Yz|wyZh@{szefd<H^ODg+LhHPmmN_*`wYCR)XKv3d_z{EGomaI2ZiCdsLt
z&=3h`yYsw4=X>6S2cjoc?eSY_9IMM}-yI*%4Jqwa$(BVl1Zsy8VV`CO*A8|-_%kU^
z=cC=SnwyR4XbyIYfE(dum1`QLCY5Zn7Qe*fL_^+mBoWC}8{D4SR3j4vV|w%Si#qks
z&H2MVU~&t0*QQrfQ2X=;L;XL@Jb=iAVhNhG*)z86$vLcgoq?^f7x&7}P|Xi}w+o7}
zO*_Uk!!%S$?BL@vssgg0G^$m7sIs?-$snJ8tG?2Jgj;{0AW3DV@|ynI)Gv*$RJG-V
zM3v$cjmcpTRrA6W9O5)P9=@xUjI%g+j#erb<9?}+iq=|$58S2yLuFlb$Pw$>L{$r=
zkf6+9kqtdbJUl^2Di&K}>gogUdPMnqLOG1bW@k%<R*2(czUll|U*EPw*%XKZ2wcv0
zGg9zqQ)9q4!toyozeRibHyp683+1-TH(MmT?|8b*j*JBPGV6_Whl;Y9`0#f-+q-*k
zTCmr*;B%@_lU>ESU|z#@IPW^4&|A(YBr%4)(0gcR091SkAcU^PPuSoDC`srhz<#*c
z?%i7l#{Bakw*!!XS7{}Xumxa|ShpMy!Bz`-!VvHZ%ncy~mNlfN0YhkrRptjk<thZ%
z*C^oI3E7R1d%XJ~B(T@E+MUly$(w+C@*Nu+gK*Up!ujCfeM8WLoVVR9);PmDumCQN
zN=)o$-jVNavqHv~y1$4zw3<7Y2B@aOm4N(?dO5@!D$!uMH1&%2N^?0Rh)ZP@tt~Bj
z#rqp*tN@{uSt0Y)rBnfTVll4eayYPLp>&@%wnjawLfQ9nJ_Za6l}bW@$bhqO@<*47
zJElk&4>y<MbW04m`OpW0#cTQnFTnetBgX^N`K-=~8Xs-o?uP457eNUOORN$0EEC(2
zB0bdx5Zdm>VWG2u%}iUuvK180TT{nF%m$n}&AT=ID;DRW4A|B3!)$Tu3p{wo)Agub
zEPIV62$~g#vB>(8UcDcT2k*5*EvVkgL)u%kk4k2EiJD8R8rkH5YpGpCPzwFxApCr$
zY-2LiN^kI>cngi<XeDC{%)(XGfd?rYR_-gI4zp%C$chs5z*hX|x6BBbO*@c9bL2se
z{@_9BMP@KuijB<)I;nUAz(m^bmXdP@TzJ_?=o}f-cr);<CLAsjR4(p=17N2Sa_K86
zu|qh`r}3Kh0iFU3d$V|^1VQe`$p+Pyt<KNnya(5l(GUvquBz{_*YQ#b60cvsMjJ&h
z+^Q`=0C1jn6Ya=tx77A@G!!A~UjB}-C|gfL<yr%50pLO6d+W*U5T5MI%X6+aRI9N;
zda$}?J{0M_0sVKk_Xr^6&xgRgRR~6V$V^Id?~PP0kg6kFg>d^OT(HcJk{gd5?cE_r
zy7psMRvg}2nj$S&$e(S22xXS6T*v_c{4mB2(1d0#4^ZY?g=vfKcHYrZhD#0v%7g%A
z090dO5{g4bFEs}6;c}}7a0XnH!N$N}q4RM-vU**q!x>FH5WRQRc^j7$Nztg<c&z#k
zME<2rG=Sa0+26!=EX^?kSn(X{z`Gmsz(z2t6cwyS!iSmMmRab_1LW2QXAomiofr4A
zJ^*s~<GSCbOw(*!a+Ux*&Dc2-&|m083An2_;XVX)Z`~KhLn{RMaCwLhUH$}gQd3~q
zo2zgGha$sIm%!JC+|_Obs?1|00cas&Y`Rh2xI;z7RMiq73Je1V-%|WfUC*@n;^9fG
z0ygabB!FW5pRN8ZFlOYxYV|J+s`2i<f9c`kD&F2niiZz~h1=VtgjU=Vjf7cpq(p^+
zg{_}5ZSS+*zY>lAFqIAQiP&53q(`CMmsS;@;36*4!7y9%$K3M6Pvrv!ibY+RpB_kg
zixGcKk|+)TI#4W9%p-i>vC7No%M~xD@x{*`Yb)pdMmgJ@oh-MJS(e`yMo;Q<iOM_L
zrcJOPTJbpje7t;{q3U$?b>|dX%+c|jP4X~gXy!=Nx6dhyVuxJ1aml|foYmUd@0H9A
ze{tUA_6LmN;IM;z<DH$gR))w=UrHVH_qNABKG$$JP_?U9LHV?jTQ3V15fug<xaQaf
zp#()yV#P~1nP)g}M5!Vx55B0fe^}yLzGZruZ_1(V!1i(oG?2ReV3ZabeocS>%@@fe
z`4+-RKdm%@lQ-(9hULvW7`C^^eY`H2d|bDaBWInxXu*?id)Xn8FRIO2rTH=G+O%g;
z#sWWgf=p(Mdtm!iByrNy^A~5-j3%df<QBM87#;4VwwARkJk4Ku+x<ODRPaJm_GTWI
zSnkPF1qN5hu}Z6Ii1}~q&+&St^8{sX#E)iAWjB5$Plz{pV%I{SC7X}ixs*<v>x(No
z8)9(soWt$W^ol);YUIMn=P$yx%ud8!c$A-{>d+tgN&IZ;rT54SGD{soW&w}v^v>P?
zY#LeDk-*Ssq8Z~xb*I(IKu_jrST~|FJ=N9PW1NSkt?8xdSihjLd3(-l?+2xW*xiGC
zlFv#i**Wu6ds2kuTyA+di(Dt&M%rKpf6sm7=jW3uVHil>sOZ`1WMVyy70u)@FX)sf
zSQj#>BlL87^Sm)K<b;0XC<)TorIC*Ji4yl5ybmOHlqoOkx)_-cpOOb@i+%IYD#<uo
zOyk!}_wTGMsv9Dc9tKNR4i!`Q`Z;IbJV~5|eRfC^3s(ovN9Sto`m>>zkIs&XU!J+z
zD`ep-aGE71Nu9hFc}9WP-#So<Txs5Sn=QYLkl6fP?W2v7M4>6lvh%UzCR@gzF=LHR
z+K$ugT{$zdIz*4zp6v6O;F~@fykdfEIKac{mU7X=*Zc=Am_DyARr)&5JtoX*7U=ho
zYvnlX*r2$C6ZPu4qi^V(9ih0Vy{<g2iLqZ$$}-uCbb)xsL!o&a>g?h1d#OTA5jH6&
zVjp@|I_k`~$mWh*$#!TgYkwEeq`S})CpGiSVmxZRarNB|g0$T+zn5m`Py6psJIzI?
zrx_ONi~BW%^!qKZv$A#jip;5t^;oLa2cyEwu(gA}(bMMCerr(I4CX|sTix-;`eY>h
zHT|+?Se3u`$#+zYqYsJ0Me6A{hTjIKwsZtvY!HtV$RS^wXVV#bo$N4)6;!7icqSKj
zZ0h4R5?W21-S@ZXlS1hUX<ofbKz7V8sCjmsH#E<}a^avnw=3Qzz;O0iJmHs#+GJ(A
z*W;bONEGio>nDs75}u`t-TQfp{DY3Nmt1oslXR{r_l9JB4iRV&#4>H)X%MN*%y=sM
zLwy)~H#jV#PWIKEnDIelQ~86^+P-@0`$**NLuuRL^|RA?^%eLTtkQNqcim2S<EY$6
zDfNZ%&$9hYoIi+WDUE{8xN8aA8zmmzYxbfSC$*W4L#Za3&60Ya_}hNdNyM{QIANaM
z?ovv3FT2%6txh<oYB4j?^kcP+M{+-7q_Nhkv)ds(zSWL%eQy=+*p<Az694Vi68nkS
zy?x&lHE9m};T^7tt*;76nXep@Djyd}K8ovcEV7dzA7g$&#}}TmTi0U7l$_6A5?&WF
zG55Grxn*KDdn2m?m9Fm~l}5+)$R*8V*+cb1zf*4w?<p&l#mSW!Yp(=Gq1-uJhmcfX
z*YUHj@!3&(yPcWrtE>K+B9-SV<m)D6`Sg@^YRnX6L=v;jMrZenVbr)TWy1R60OhX;
z@nCaZ%9dRI6njHG`WLoE)sA6y;sc@F#i3grywne=bBTF^T4%@E4Vh--Pg+f`v2zJG
z%G-Ufd1K4++`iPQepNAKczb_oVw^G=V=+>)xBmJ<f2B;xn}N=^r|}$ouF<l|W{s|v
zHPdHmTLx^b>`pjRE$k*y{!~P66>hu!I>i@eK`a{asDjhC_&wDyD-RFq{9;O=%<8=9
z)DK1e$$E2cEqEEBSMdpi%;JPN!76V3W0{ai?;Tl66*`PxDT)a(4gx8!?dTL}8QR~~
z*gq1B-z0T3JcX5h-ND8>%B#zyyXphJ(QDR9hD9mmk$S-pD^-qr?i4|Nhc2VO8I|Tm
zKJ80iiIbuUI`Tv0Xx+%tQ|VITz8nnQ2s4%r6PLxru}~fEta)QDk|R%xeNA*bL{8|^
zo7aSm(iIUsZp{_}F5NHH#;oFt2C8(dm9UhV*vztUQalj}G0beUI9kfz*8=XA<$qk2
z*6+E>J%_wk5Q{-yC-iId2P)8Sl%aY1uFU<6Z;7(s%=27E?+lXLC{ZU4-@7O0uo-@D
zp}f}oq`70QEYn4LJPssYDcx0HjGqyMg&xcaA6z8JQgP^vPBA3O1wMmWd}0oQ^F1Qt
zBf2O>3u_T98IC6#Hgd1&#fPP3;=Tn+v9)8xs#|_n@9dT$R*|(Mh@C3JNfH0)sl1$y
z$1FSTI}j}wC&KZ;ER|SU(he^g{+0gv7k=#|QM^PAO^+hWoD%5`d5M(Ae^!@pwVQlj
zxeV=5{AlwRR^L)y(Dl!3RuM85m9DO=#W2w%A6A(Bfi<d=-%wjIlz)*H=cDiLEAgm{
z`R=p@w{*>(1&3VWKZs(-N{L@5KdB+rf7aKp)S5MRcs)3)_kPvkNlf|*hyR2(?)}f<
z@ON0*Z~u3AL%z<a_x@FQo6y#2J_8l=dUGy~8kp6hGpj_Em^mFQ&l+@K%lz=fO}>`&
ztKAnr-8x6|GI;llrc*J3)4FFq%#7Tl-0lJ2&(Wnln-ex_>jX)r97J624bPOVdJNZv
zWKCtxXQgHF_Y@2u3TyWfAHykz!m$>3%^iQ#u<4q&nbHtiFZL@1rr6l`JPeWy{`3mh
zyXbTX!uoK1d$kcif-mKE9uN28LI1is=cRt1YeoBAF$B+5X@-Wn-j@$gs+4W^>tjXS
zRK#rwIk5f8hoONRZEU%Iex**9b1#9JcUIL!ZHpOa-^)=E_uU7SBk$Eg(bw4FtcsKT
zTC9n|>V+Q#A5XVGz00pFk)4wHkY>}Vh1W)u)$ny%fiZ=q4_3pgD=w8rdk)+o!_Ug6
z_OBUI@_F_BFn+fl+GtSm#@FTW{>!un&ky?Z@-4-*b8oU-8g|dUsomvdHaRl`zjNW&
z_2FxV4wepE3on%l<L<uXW366eVfso)H|O%o|530{5>@({CgzfkF)hv^p2EVLy~3Ap
zm5fhv4cK+kC%Tkt<LTZEzxQp2tb`{x&;i4mS4iTrCjovIQAax8?Tg~iFnOY>v%XfT
z-}{Pr#m;iOjXN11Vc*E_Vq^H8m_9j9fz%{bM3JKuhTHSDzL#%W)ZP=sW90fo(XiR|
z%DYD^rcno<^F>0Cr0Em!WIK8m;bGxIZ@(uS<K6Bg&n&_!g$Q?7tVky17&nk%d7|qi
zTyNa*1zm?Oe4udSN~kFJDGhM_Ix@A!(bqG&kI=@I__lFH;f>sx=TY;+mKW&qkoP=u
zPVwo8DB<D~?cUr@3y|qam2zgozc7@UgNuhpvAHn2K)hn#!6o~x)AyiiWv=H9>y^5;
z+f)qavIm?`Rdvg|l<U70!+GLdRNJZ5TFpB7b8oi(4XXq>j|PkZCDwdWnk-_ZC)}=w
zIXf~>z4)f%Za3bym+<lop3KA_d`lR7JHO8OrzF<2j(#NA%`IXwY_!;)+JBYC$c4{b
zKi74>%qcSRw(>PA|J5f#VX2;H45alW6mQs5s&poYlwM=}q8#UT&K*m*%(!A7heFq#
zj_JEO=eMgb#J;XP`z9yIE;uWdwQiL&GUTT7p-9-z{{rSSqVU;=4_@*KKNx+QUm~TU
z<-A)H_L@rivnqiwekRXGnqtyLt+<luf>z_QU8d6L_*L<nShAW;LQK=Xg@gn>n;u^l
zYk67+Brt{NCP_Y3(Cav_WPYBJqd4*XNx2T;NPHaQ!iyYV>H)S0e`GEPok!rs1!XZ$
zjAhp{?dDsVUGwJkavtG##;z0wc`422ekZ>%+J1xnK~*|Y!g|>_heWwlrdf5!XQtX3
z?B{J!a!y+kJR5X)xfO0$*n}df^FC78o;gDia)TuM{^kZ4eUAppPbU*AB@%~rbkwsn
zbtG};PgH4S>S%H++-GR6dq`We<!YAw&ivs|6T)eHwI2xv{I};uK7Pe=m}RFipbr{1
zP;_Yer5~}xt}eqCoNnf~Ej48J<gVzcK>DWaFVnrj7YKbTMyl4YCdZ9j5?x;rxS8{d
zKk0Q?ZQ%AzZ^`5sGH$`1`V-Q_tG{dxLIn;<w)@|8CkTXWzfnBjmtjb@^Go`Ue71ac
z-u3pRjFcDhh2Lyy^Pf~myh*^e(5V{3OeiypifP+mq(*S=>{y08;xi^mBCrc06H|_I
z%amX9b<=;T8D$W$=EPV2;Heww3f2`3r+~LK`_g__<hl29G;H(ro%!y!Vhr3n+jmP>
z{5A=`w_cpP8?CD}G8=`|j+}6|Jv1(FxOKNO$9Cm_UuaigL<~uI@1@S$D+1Ko%Uayb
zUYCBp8d4^FxJ$gkkrvg#>P28g?%I>CCWbBPPk8l1^jU%~r-oa6PX!8+o@UHQO@~o_
zl?}p8QQ6Kn*2P?qQDEz){6w7W9a4XJ)@iRI?JZZ?yQl~0R52(IuCSV~{JzCi1tted
z`yYFi!a~gsPkSP{?EhnY_@8h0zH|S-eA`^Ie2pH-gG+@?vYGl<K1Rq_jC^^H(OeyR
zok#QX$DiIfd>_ZI+*qQ0U(8dHRTHYr`Urdeb*g-CqO5c@gHtgEyLApH_uN1z9+nIn
z7QT}cFJcJC`|xays{juEL&CFVWvff(o>#DVscoM=i`h!sTiB8%-@HJw_}<EPC^zSf
zlNLFz`NSm8Y<Sc?Cx#}S3)?4cF=v-QvAi15|7jHN5k|%{hcwOKD?)XNgh&sZ^M13n
zZ?S1X&9JpqShQqx`Jh09dgSJ~%E8vguJ*X2+2NdR(r#srx{iw5{574+Pv<eHGjjwQ
zPA0hFHtwE^Zjrcx_}-0U*~BM#!h?YFPB8zp>x5Zy>(~329h2?T)FZkRy>F}c6NYz4
zy1stbF>?9wO0|R1j`)}hrgX?y#+kU!A2S=Q?j%WHGb^xsrc%np@vg&8^Ii(?&U5+V
z220KTnbl4bdGoj>Tc)Bar5AdlX3i|ap~9IZ3U1G|Y|i77n<C!ecE7w(6SlW@r})(T
zN4_UqQN$zXIZvM=kyR~u(t#)G@v_FUzKk~NgwY$K+HcRTu9ruN(c&1&)=-KV*Rt#}
zot8O~tjsxaeZhru>9)F0fxc!{y{WRq3T2P42Gz|KLe_6TqD&paS|;6|FuzP>oZy%a
z(Tdu%ao!a&8y0JMEpY4Rp^?MP9mk7v7w>aKDQthsXn5Uzi~HnV-Y{L(zE`PQy{z(O
z_>xwH$pW7T@rqww6JD-sV-B>@nEO65?NU+N{3CXxrE`0SR^=j(lsJ8nTOXz^bq$?E
zo^Y2J(eh@0HLq4RgF?%BjW?qeU#GY^I3G7RpKF$W-!gE9(&w~rf|j|u;ef1?-gCkC
zcT?{BR)0;Q%m1eRX!O@%R_v7v*o7RK_KE3Tmi`HEzxlTjoXPJhmwJ5du0)>a^_rQM
z?{}UM*AYn7obWBoZjU#L;*{-u(o26cjPmqKbGG+|Ycr~|3Db1ve|XpKQe&lIdqkJ{
z`;pYiL5Z7L(Yp?d9TR`tqeCmuK+1E|bxmN_KC3IV*>aj#n3m(NWb-$o7}lGKx{QM+
zTesQFhp~ufu*pNKJnS3&gQ$7(`oELPi+ev5Z#~!YK9%5f<b5iuUq8pHLuQsA$*4%$
z<axz97R28pF^;rnVtj{pcf3ILzOr-R%6+m$s(=bRxyJrEzPC)L-d@u2jP5J@vZWPm
z@Qn;vHqAYqIUe76w}>=emB7(jcf3K|YtppOP2-jA`*Lohq{7&nvZF6I{Vv32uTG|m
z@ppXS@C!BI^*?RqW#{*%q44BDa{9ULp-ErTO942IRQUaa3yghxMn440g9dGD?MJMf
zFAvbgah`QN@4)v#ljGY(Vg*i}j+-hMZS*&mP1_|CxAT#fULLbDBo}aS>BHC<bo7fF
zyu-CA^KNXD`SqPF+iz<z#JWQ1K9oT&9CbqY>=jDcQOT!V@>I)N5*eTA&X$U<Zk8hz
z&kA^5sg4z|>U(nI(-mv_gMw>0HwM}BkeAOqa5Ip!*}LV$yEW>>cLAnL>G$n$y(*$a
zq!l}bZMR&5cetqGhny&#B`Rtz@FfscC!g4rR;O#|)QJoY(C<*Y^Nj5xt_+W1R*H7x
ziXVG@#4L+;o<SsPHBfP8Cj!;0^X;JpV{EhGb8-7V21}m3JhhJlvWccoS%RX2WO{`U
zrcQiTB3z^}-rTngCKlUX81W^qB8)w4md9*{c+@+vJaf6E?UxbzOjCon;b@u>*M@H_
z`L$Ig)R~^QMKtI3Pw->fS&6KkJ$dOWO+=+=M!U89<@oRiKYm$bZ9Zn&6h%gDD-<D1
z%WE$NF<)q~&GcqoESlPudPRsE`ZoTZq`TD<keksOwmd9dt)BM~w2mHgatPe$*%Z-;
zxLLA9lVKCw+wS4Q?59VU)t=*r|I^_VvAvgGo*{ilEf!hzg$lJjX7?Xc$d@tsW5kHG
z+pn)~UT(%rs<j#GeD^SGvNGA~m95<3y1+%j=FSp<p0^)UGgdVFFWe<g$&{!rneG@<
zK~iEXW`CqX4lf2_u{cb}ThWVMS<lxmfBrDwneOEW5*Okm6g7l1u0H45%k`2R8PNJ(
z^5ND4?7?e>iMC!Qrwm49S&Y0&m$a`Zq&&Of@!&)*okOx8m4onPZg8p08wyk?R>aJc
z?1wGEMxCw&!=~T#i;9%CQv422?BXgs`cJ`%mHqflSO15*NAIxSxpV(_mc%UZ`mKqk
z5wpNsH4{yKO$~jdCJ)aAV-q7yHQfv5J@ftNt)Fim94-sp^!{WY?Nv_Z9P=oe=6lSW
z#KMiM)6#7cyWbO%6<)@~XJ&u1p;qWrj^;l@;7pmV<Lr4!KKgQ`y!{sHpz)n;>x~{d
zAH#`t+p%FoBO@)A?VaGJRqLkUK#R3cmmZzFVgKL%RbAHv)gI7>=ooJalXIE>L~R<c
zEDb-#_%Kqlx?Cfg?}cY2RAj``nstYbfJsH3$og$N^Dxy)*7`$p0RfCBiD8F(P^{dl
z;;kQMhL%*O!~Q0|CD5&CkkEv3k%vXIN2TECQmF*`Q@|DGf0(r1n~K{-hB#Yy=UqL7
zQj+kmk?SRVE(nDCq-s+|{*z6SvAVRjAhu6-XArbxB;|Y`|Hl`<zj+l0d*nKHdg93S
z(TB@J*WrJD&|zPJ>lj`_bW0n0igUpw*s8y~*x8Si@R5|4Xj9SC(~rEX`*0d@^C=TM
zlwF~$qVg+5CEvuE$HNYOn%LN2b)rF3RJ2_sQ3_~&jypSBTRS`X`S~rAa8*Nb4Q$X5
zct9f?a;=)2ik*Gi{lP$Z(<u7Gv9FLsAAMC0pNIbQ5q(wkpQi@!>%ae;ho2p_Lx1xO
z!q9X1_sR$wq44Qr8&+Tcy!G{EkX~6%AL@Mg@BwN!*cUHe#KPJyn`edshHmfu{d=IF
z`E_`of`Z}+GfU_WuU{e6&6^|B%Ko9Dw!0g;n%kglf?7@KpB0pF)xgtz#}N_~6twFD
zjV{Z0``<w}<_7YbCVkXsC6<-duK(7Fk9Ky#{!IJrvMSDYD~lj5Kx!h9NNCD4`XK~Y
z0|5xZdN0(L*B1IY>FI@)D9qF~G)C8MEiW(E?&Rj?qJ@t%FCt}<6`*p}SX1*te+49u
zejqPWE%`X}6-q6R_V&LY684CyGW*snsF7k6V`5^s3neDl#8R%Zw4IMV*Bk$W!VW7v
zBV%!}SU*)l?<RfNV)*m&v23xFV&lor9JF`MpMvJHqB-I*R`y^qs@TJ$=`PCQk~~P7
z)_b9NnhDAYs7d;xq&X@|;5qZy^kQHUW0G;n4L1#8J$1q%A4pbwV1oeJXCD;A{yd~r
zlO?`4MiUjd0=u=MpAR1+9B=|Fl9US-hGR|Vd!V}IJB~V74Fmm9BGlDD2c>Bj5J0H8
z3vvV5c=3|DI$<rv6~s5~s~{Ppc~e~pt$?1+-t`>wb8}qW+=G_0jNK1yrxE1V;wy$6
zW(!GObb1?8zslO%rL~@ec44~}yRx#fb_0rqU$8)8#{!}XJc8w?3-Ym<)S|&%Kg?3z
zUiT2xLPC+)2MczRYBe2HRyC(R!$d>5<n`!_zeD8}x{^1oNdgz9519fBZQ%tgGpL-t
zh!fneny+umglsz7fnZF?#3b`zd(4%9xu>V+nXB-#XDknws=X*!LXGryuxra`#KRvx
zdUOhH%p>7>pfjXjsGQ1vwoW`bTlq;wK8*m2#CL#aMrdSY1i3%m8ux@~TMV8Q3N!!*
zvR~$VUC*|J;ZZ>Sobuv}7cWlr<rz;Rw?Q9~ZoR+r?qo%0r!4z2O&%z7TUYr|sKd(?
z0$kQ4Y5Mg6chb?;f=6{mUA`N{b8(~vP-f>+4eX{RJa1f5Qj*R<==|zJ!^*jkkPx%E
zt{Zh#9_0R<mwh!Y=DO0LJTJ8%Xd7bWy}`c9w|yur&ucQ}*;l-m=3=)q`2<h#Y!Otd
zX(ii09BFB84y^z$7rJX|gbtp+3#BhJS{bVWk<BzGO@e&J1nlmggWAwg;<j4l<{ZAV
z-kYX+8WHzAaHYYg`l`V8FAy6c?^Ba17e+oNk$UZO4T|!MAO?WipRa|=l23I&^`5Yt
ziHQkofHGWFSx!;0?n`Ylw5qIw3b<}(bHGX;n#^X}XdfUJ`|_f!>2V*AOU1F#NIuex
zxg9O02#<NPK5+FEc5V48C_ZL#SWs63#+DuyNSZm7m-CM|`8t1d^J8PGW3l;yNZ7l<
zQs*<B{4sq`B5<s>v=zJP@L+e)W|Q>KQ|t~Qu-Ur(ZgwQvB8+H$cB3+neKR8zr7TtB
zj7y!Almrrm1C(ZW(LOw^c1a7PG7!Bu^ZU_B&m{VTGVwf<X07j?+{ZSECyIF%P?v2}
zq+KlmnLuKvcI~Z&K<tU8FqOhNB*?!(Zy-qmE%D#djkZ_Dv4~m4^>6x!XdZ0IlCG_+
zv=l#`SVna|=H=zx_R%Z~^!8Ry-S=0q?T&biknaLDrh|inCA<6cxA0_{zjt?c*W8De
z2_K=DJ-qbmZgjn@x;<-7<hDP5do@;0#3~2a=a(-V6G8eemMm5BKTMNnhTfk%?F(-n
z_n`OZV^S{21u+)J0}of%jI^{n=A}B|*`ahTTa5P%o1EvZ!Q&SKCgNWXLCi3DpS-|)
z85L)Gmal@%22oa8+SlLTUZ)Q|8rh%==p*3*dpfANz4ICbf?&>AGIWzqxVgBHXE1Hs
zb`5=nB5%h9t9=3OnH*I?LPEkxu@z7S)?lQE>rq7M8r0DIeiYZSi5vJLCVx;nfBIy;
zw&hJY?p?jKWFRZsIW_gFq6Fl4jPmLH6XKxhkP=-6?`v*lg+sz7zUI~K^W56keRqA9
znx__~FR@AmK2gp$fs(yY-_@b9UO4&w%ql!IG*1PLKuQR=>EoB4w#O03zR;5;Y+G@c
zG&4IJ+CpO9(r#WiBfNG^XDED%;qV^lI(?niK#SxeU+WOnWFoE=i;ay9J^r48k%u!F
z&O?-UWi&0I?5(~44_q<p&Aqo>x3_)#E^0X9kQcq}PKPjP4UQ%|JDcTY?FLJk?ypFP
z=kA)7bT|_0qRu52F9KyaIXH-4_PhKO1uMI;oc4Lt1FKo|`q<VV3BGs@8Xp%skhmtZ
zT5Ix2KVCqBstMfZ(&lCuw^g>$_$0(d0!D>4@Fbu{eHbmCeHH)gv-O)|+cKHDt+DQV
z$`O3>w5Lp`TgFE{c=17$8L*UTSNi4o2r`I_+cKETLPEY6`ZxBseYtLJGp?|m<y`NJ
z>RW~GIzLS>9XM*WD-b5iO#(xR3N9#rVgT*L!<1me0DK?>!EdrDOg`lxSfcH%9^4vs
zBG1<Ax2*QO0oqq-efnP4_bYKVdoS{y3G#NPsB|8-PSgjs7Nbyhy8lG`=CRsz8>_&E
z@p%n#64ss+ToON>w{2Xj?H^Ipn~Q@fQVHHWH?RJQLhyI1;qAM3C2bPIy>Y2U3;I-U
z+!eITi2aKo&i!LXe{n|m<ttaNM4}uXo0@Kb-<oI$iYvZ?K*J00-c+*&FeL&^Lso?`
z)W@e4g561wPN*?4F@Yoo7r@GIe3Hx&0K%WqQIPzSx0k?y$75E88G>jnZ9!q7q3557
z{idDo1}zN@&GqXK-x+{(1uRGE3u<@%fv|8L^yn78MCi1C!{e_e*M#8sXSpe?EpL3q
zo;RC;td5uDfnIRFJFFt&u~VqUDuBc<;9yiN*+ILz&-)C3MqDQ&T3g9)<V^Amx}6X3
z`nQKl>=uhBwn1aDh#Tc5Hewca6|uO#nR_2z6gP<@Qp@M$jR$I*Rk{NSJF}(R6GCxd
z{Yu5Q{XZeLpllH1@3|Nl<^ne_k&wJhl1qDGdHK%WtN?Itn%*zHLqkKoy-V4|xh!~i
z&JTW;B6-e99^qc{diO(d0y_<EpzN|gu{Ep5FBa`a5*!?y;wp>Ph6kGRp9?kjAzu3B
zJhNTfy4UvQ%NNjtSU68UVb6+5tNsvNLOp-Q#c4G%n81e7Xu5@R-8lf}e-!O5`~V$`
z@A^tgK)kh7322fF3JQ#Z(*D7{hT_+L@iNotCIi|48A=eBh0(9bC+$u46@VYnC^uY2
zczJEQb;gPGDQIa3n3cI;3`yU2&ys)s9pLkyAN&J{u_gXD<jZF&O>*lPE=QQ7&C^o}
z4{^TYHxQls#`ZK!`_kX<POkzo{ez)-#D0HM#?fxg;y0GfWC4_mi#DEUzv1f@Wm~^3
zBpDNsdgX8A{aTal>GEx!-<X%ajL*gCH^SvpAc9r7{YLC9wFaIeYSw=*d==*p8b)8)
zoRIwg{iV@}`k&?fzB{(3=ym*FoS`S&$)AlFx&C|6e~!WL)%`Kt-^)O+?Gobmt^7Fw
zzt{YK;=*v4(NBX`f#eqK8+_Q6|MA6y8c3n=h_;Il7oc|m+L-{sSVPM}x|`kK#ok*C
z_xJCF)^NXZkVMPL$w3Q4*}y;%bVMHAy?8J7zz;LAxYj59{50;>4Gp@Wpi3{%sKVCK
znFaG4I@%B{PC^0#YHZ~nrXpXUxtXaBaKG=kn;WLQOq2idzKL)I`1xtRew_!?2?RaF
zTsYX+s)^+xdBR7Ch4Hu_EiP++PZd%cI+-LkK(^Z2lb|A_r3d%N%mXdY$Jf*4B*evS
zjTXY?6F+cwb(NF>eQj%JXD1uZ3GiyaZ94V~EHpYMX835WtA<OUxT?&%jq&$qsM=A`
z2s1J^7Tr1xeZSw6NT6XFT09K?KKt+q(!CK^=A`6gXveu@v4U(bffOmR^Yigiz%7P^
zD8-9M5Z%?{vb1)OpHY4LHt9J|%FDcuNB4E_kYfiF??h~XL#SpUDReykr@8Q}9)M_8
zI4o@tpn1KCU7K`lPj7a&O8^8zI2BhAfAuPUBwWfj2@<J0ji-P9bC6ZZAgPOl)Cn5n
z0D|~MMQK1T#$N{FU?+4Q_1AWd&G(1E$@6Xv*+EJ$ZNj0b*b9N++GHT#Upx9L#lphk
zb|{Q?=IG=8s`0P4P#u^6G55)nCqWe$F}P@Vy!^!>=zM{s5k3|WMdBgPu@$W@Qfp{y
zcgBf9vRYqYi3yplCpx{ZPrOA%RSP5FZr;QpKaEa?(RRH`$Ii}9Mn(pi(H5s!ONO?X
znVA{b8??$WDP%rXI)tcQ8`9g`3*9)6epdW(6L1izw)W6NkZS#z$?@?wN;fZEx#I2X
zd&}l62#QTjP0{e*4|eSAMf|=TeH9oW<3Yv^(p;CLTM%h3Sz9xKKoBA-8~*TQ)8ORU
zg=E+w2!1{c^&mBcKC;@{`1p9((Kxgf?qB#rHb@#bH#IetmPUep9|G0F?(W_;=Xo5Q
zK^TMsB5!yYF(!BOwjm0%V<AJ$0mhR=tAHm&@2mti;0D1<n3$M*L7R6$?FcZV?y3=l
zU9fgp^AI+}$ovnGh_58zi)jN+TOl+NyIFUUfT<v{q7O8*Of{4MADx|@-A-5dOpfA`
zvps^AuNB@N!r)ELcXwzJq|ibK=n)T<5ZR<6Vq}!w9K{v2x#<%aI6XQ_;ZkD+b`JXN
zh=b~mjtx*R*CW{Vw0XcElOuUvy_tMc5njwoE;YKi$ZI+$Cm+CVWjb2v?&k*)<toCM
zUh1#OOsH)v^eci2A9B1fIJC8ktJ-=%F3bZqEw8AEiH?5TlUY+za)Xt1?XXiq!kYmG
ze2C0s0E0kov-U!7^Zh@JO$nD9+%A*<USJSxGBnc0f?)Lf_wS(N9LP0{+KB>XJhYj1
zS5%Bc>(eF5BmW$Jo+B4J;lf|42P6JojL#VR$gH>^Lx2=>7bXxuU%!kDKK5S^%}S2Y
zt`8W7<Ilqgy;LR8T<zlaFZ0E|@KId_RGlD*&IEldWSiMKosFfXj}F^W$3NzS&11gw
z6K)lCcz`hr$)7hb%64?l^4Qwpc}Y^)L4JN)PQSj15>54>1B<HiI6U-`erUbJ>HrKY
z(-FptB0CBt=%WvvMwV7qf`WpHIso9H#a!i^6zyw~;}7q}c64&$WMdnv^G^Xt05CgA
zHYGGUIoaL40Xp8_9&brWK?BH{oiK@3Iu!LSEG>bpTAT0Pgj`NwcmD5MStp17@N0&*
zaf|_FfxQtXxqbfp*~O*w=)iq7ONG20a;FD!dEj=STg@N@^RJiD<qQZ7)l*Y5oNB&d
zw|?i&9gdI%G21gYjyBy|gw)N|6?ZYTh>q&sJ?Rc|PE*X|d+sU$UhCSmAdqSkB@Je1
zd$5@UAMgi2z_E#+pG1K6cWP#ajxa$720$o+q8=7QG<ouP;S@eRfWv?b)7M{wRC;@7
zM?qfR*)8&T4XjR`II+>u&%d%JK)R0yJzR=83(y&fd*zDTM(=;{eq^z5A@C#*<Sa(Y
z&qJ%=@ro=E{e69*w)^2pX_=YjN1HS;Jyhy&vA@6n*iL0gU#26SCr9Pv<>4diC7Hot
zr>CY)b!Si??M~Y_JK!*E$g3I~zk|uyth0dl8C)0O-eXVR5Js@Rx_Wma4(l%}^o%^h
zf6hp7!xxRuA6=5wY4*vxyfeppJWv9%@T-L{p}`BVqq4Fxu!JSY3*X*Dw>fZeaglME
zzdv%44b3I?_Gdywp?mxF(apbd#Ku4%T<rka9$!G~UEvh8Ji!$8C1|_wrI%%*rbZln
z`q$p|Nbt&FTR*I~(+$Z^9G%nK7@e*IMK-<F$1!|@<jBqF0@H_XGXOLTBNl8AwkP@c
z_yPk1kB&k!33_0TCG;rxX=9L*lKMga4;{at;OJ2RY~bulkc6c;0qYkR6H}7|qxe|J
zP0Y+vB;&A;wyQ1M4hBASbaX&AdsnX5YEv%|4>-5ufRDv#4piv!0{BdfS->-l|9d=F
z)zXShPM(L8h1GGmJ9wC#nwkqu567Rdv1-R{xpwJ7oJf!<WM0s!(ACuip_!AD^MOuB
z{l7qrg(5HDjR)q!L?38x+&Fo>i#4-T5XLxkI+EVMe-EzL$<dK`GUUl$(}h6$CxU;n
z1Hk_29~_J3zyBkkrvd7`MAp~S({s4YiM}o(A_8K|pPfZx@>N3vplQHnrM`fzW0P{&
z|1k?d;NU`Ge1sA>PoSHQaDF%u2MtynbY|VZe}5&#&rce<J-EW~j_fABeHIDG4njl{
zOuGU940EQk7c>5amr#W;9qn&^P(4fa4<ZBjuQeO)+YAA42Ye?$19a0!kbaSnsA%P}
z2@k0Prv^O~CLAFdU^Y29iHV5}b;Z<Tp~nXlycon7Jv}`j&O4kEfDGp0`4@KS#g2`M
zk(Za>2M!-&7KfaNA)_=tp62)#)DWX1BjrCJ8@`qVoq=BK!^6Wz#wcEfJy47qEGba~
z$`E7LpHWfk^=pgS4#wl(2TukRv!13VdT>{Bb2GH-Kt;s>2u?U2O8F-LE5f6vPlSDD
z*=^-^c5rwCy|wPygTsGabrbHjYtT{D19bX#g{)^<4yfsDFE1|u(!}5_VTdKa^nNWM
z4d3o@=rlS8w^V~-RCn(D>hHG!cn>#{0u4Yw&B2yw?%N@7fYa*uuEcpG9l9iW_m)l{
z8*r#yBs3hyrU!rdA`|}b1F+GxD;>#-g#`utkL3r_G@Rf+R*oaE6F<C3fQNt)e5i?f
zB6+eq>h?R%c(AuJGBSCE#Fzh~@t&tJ(@rt-RRn|Ce9t==C^qBq^5six61Js!MerAO
z?d@5eK9JjL!;cXuJeQD=z^I}8I0V&!+J9x|1y)Qiui3R$HVsCTn9#@%13FC^VU})u
zIH49R7N#8W=<N921xmQhYY;1UTG8xy!D8S8%=71=FQ@)gVRc05X~qdKE}W-Kd8or8
zl9s#JSH(g-{;&$I%%N?SLz%OCVz@y#GHl>E6Sstojf|8}FteWvyRZ9|E4tWORaF(8
zV*!KMeh(ls^s>n|`76lDm6w+rDEqF;1{1`XvoY6NyzqgEMdaiut<#@L*jLOJkl>C1
zWI(v>P$9y{pIBI6I#s&@1KOaL;N9Es_izWudO1fYCeA`Kqq0SY{6cNa4UvvY3{COu
z#Y0~dx<3`bFM13M)E$l;AZ~%E7O*Z5AFPH2DGTYhHie!(73U1H23KrkmvHm1D;A-)
zwY7x~eh6sLjQ!x+y*F+A%#{`={-T6>MMTgTpmuat4?($l78rv6kSr&2n2%L^0W_C^
z5g9Ngr}pd5pFiRLj$LU3XD%FkIC|0$#sR~DW0sX^4C5_^?osBA9_yoz*dk`18XsQ=
zRt>P$^;@_2-z`D^5<rVe_iO(?bW~JSE@qc6UrsMT9on{+047hEJz8713@JK7#M;5w
zL+^Fyf3G_77(G19GsDAothWs0&tYLnlakzj8{q2dYRFS|>aTTL1VD3ZU0VvULeP~I
z1N#p|8kA@lbWWW<?Y?^Aug49@fhOUI`U4X&b$xwa==mFc$>8$0Dk=FC`t-raqV|@n
zfI%a&@F@ES42H8Kjy@?UARqub^6x>c1F9D$1CVf;r`{pI_4oO#-d0mnLt9Bl$Ge%f
z>8`X<&cy=Aw6rxeGAvfLj&?0x7XhXT_9F`og@IQNRETa-*7P-Zce7JaoIZ*rZ@Nmh
zw#*^25B)r8w+=`-{a|Z(^xwOR_Jj%w0s_PbX4KTwu695nA3NI~TiaCVnF5}VRgT?u
z&pfKz{_ovVOacVD7>2BLBLdg7vA&MZYOFb-OaPoAl$D?)MRdH3mJ+aynFIcR`6p7}
zuX!uf$6>*Om+0u`USVTHE*HD^FQ2`NJ&^wn+B<EZxZ-D}rVec%J>Q85RmfiE=jXv9
z(1lBpU7O>N$tEp;g5mSivdId5P97(Z9)ld$h^o0c+aI?FBvdn3lPi$?Z=kdOU$N2T
zHZ5%eL@{`rSQjo_IQZ9vCrB}IaF}JklPpp{{>yKz?(Ojbeq?4&LH9~?n9z)D9z&-d
z)7v+0@J-l&Td66l`kTy@6g(3Ua0EaPCFp;Q8RFms;54Xh0?EOkbneeC0+ohFWEdFu
zfN0^sK^1azANdN+vH@H4HOLO4oi()o{lma*7dRJO_-j+YIX=Tb>A6odkS+f{;HyxS
z{)ZTY$OdQCC_)(~M0P{do1kD75Ff|bbk*9{00<|@n0_Ab{`3%}9`88%A(uPeb|Mb*
z;nK5)w^>>UHIHtU%dKQR5+*D|X#B2HTvS+Sqos9>oLI|dpP*A<2yCNVCyub88s>Kb
z=-jS>L`u{9Rpx)n0{?Y;tXg1?t(zXlW|cLg3RSFNBb>!Vfq!lj{jAA2Os1e4^zs>~
z?jN%S4OJz;<ZN$m%gf1`o0~Jd%oX3cauga_5eN-+b-*S%nwrwlPrU$Q5KWG+{`Q=|
zqV3Ys(xB~doz+KZd?w3ai5;At%)^K8-QC?G=wixRTIMhRLx^OK4MFknZEb;OuVtpQ
zwLCjs8Hf@X8AXPSO-vZX!{3XWLy}!!2~;4ED!;t#AiU=2tT<2S`lRqvl2~g2Cl9R5
ziA=D|PcQ$Mdxcincc`mJ2Ps|}0TVAXmj1ON7Ap6S{uwPu`0q*4tv$$8%B!l5gANww
zR>OlqG|AZAZP5}fv@;78OG)6%yeWmzW25NkF8tdEKy&mwj7@-5L#0((RzOgYfI)5v
zn3wcDC_h0>h4}cVw+g0U&Pnv*5X?gD$<|F|7|b;+tE~kdi3l+*`;XN-+d*qCF}G#r
zPI6)4{_<#*v2iLeGqP*mUS1`C1s~SQq-(3ILExTtl7oVJA;So@C<El(+tXvo<s26`
z2_UK^Ghl0Li`82gT0(sS0;>ODj^z&&R}uPQJS*y(x2-nb{-}%{Q#I80F=~!Lo(;q=
zbYZD(Y`jm4Kmd#}nG)Az2L}pd4-Cj2Eug8IuLmM&!mb|CfaKYN$qStcl7jL)?`C}M
z>>leWVgL0+DerDlP}qI^h#oNq<EP|vO{Tq*&^{|8V->Ot80y{&1t@5!9Wyoldnk%J
z<g?!Rg(-t_B_avZVSRL&;F~n`-1~)wem+X;&`<gwHd!L`Z+!n_Q$W-meNm1s3zYxE
zjQ=OW%>QlKV0snvw@3h;d)N%{ON`Ge=*4h>Y$ryJ_Q#_}=GlS!C^;3wRL!hs%%(V>
z8<{#5H~!J1@L>^nNb~6Q`x@sb5QvN<t}+M{c=*55=ZiwnVRsy;S6v(s@{`$~+5BCe
zg5|k$Ztd&GwkTMoWSTH<rY|tQnu|YVCp6zXyBIxz_`})%w|`foHfKjvpFMkKPfN+b
z@Z9m-Kaw27W;UWbGm_W*uYA@_46OFgQsV2D{8IyfGJv~JNtxixUTl!)EX|mbV66Rr
z`*DAiKmVuye$|LgMNQ55BcH&)gvE<l(F&o{dn$I-JZYZfc`e`U1D9TW6V{OBU-AmW
zzhz2A9cX61VIV7dp_gBB`^}QGpO|E;1=km{b||XR$EoK>D2hxS*0o&uy*o?q%u>S@
zU$=JOGZAqdvr6<kpEoMsI!1zd?oPxZ^Plp`fBkn!A1*#V{#zF4tbh06Cgt%K<5TCT
zpDmna-?@L^fq6GwUy&{MPvHfMnZ%5rnHKT`@5D;DpN#nb*WUks_iVegNzhdjQsgsQ
z@vgH+cKHhX{_FmMY-`<J6#C(Yh@3py>nQfFR87wQ_v9jzsKL@ovZLR5OWHO(WK5d5
zqU&szXCsf#<YL!k*j$>-fIXp5Ur}gyJH-falNW^`WS8%jM=F4!hh1$lYc7pDi#=h<
z65+{-%~Qn}l1KpmcS(9qgrwdFsikFUX$k($0Dg5~U;w~j#UFtgADS+JIg^TcMw*)J
ztgLMf4fCLeIeJc7BWOf}z$6T3;02fuBY=Qy{(HDdSaA^mcDU3^&#d>t6sNht;=@42
z@A4UyfF$Lai;9Zel?7bBtY?4!bpvImOtlhYOsA%l6XeWAIQun(!PWOpNJ$!lVpVBW
z{xla^qB>skH`WIsRT?j%-K5<`sn*Od5HuWqPv8o$E@FJQFs{;3Xw#oO+GI26pYn!L
z*+4&wPO|@-H8<`~lM%hnbZ@N*nTPK31;F30*DkP_ExhzGN`(>{q<b?tAdP?w-D=_s
zuqZ$ufBo(s9v-gBI02%Rb5wcA-fUeEiXlmi>Odd@nlA0i;lOTyMv&1EIrXJOxd)_h
z0c9%0czBu;*f6o72ny%5MlcQ9!Jz_zl7$yqw0XCt#)mp2IQR-Kt|t5K+u~X!Zs+0K
z8MQEj_n9xvuguqypces6#NF$l@T_Nqq^GACbqL8Q(S^UN&k0YcY77kx0owYe!vvy1
zAlU5ufRBNVM$?1OpS#Jsi#>f>4dj&Hcw3v)K7G+zN?bm?k^%xEdWe2+(Ct^WP+i#>
zy<Qf1Fa@^&#D0H{e%Qe-Zf!tyb@f&?P2l4O5^hT=QW8k`fpwk2QF!#|W4;Yk$>!wK
zQ8mh1R6qQrN<lXQreU={BfXywjBYzrzFF?xU4MX$h2<B?VYR_9|FyuSuh6O!G<0#T
zpn0Lui9Ux;anbWuc6Ll>+Jxk3ucKeO0aPu7d?QoS#`$%>9@n{VKwi!E=A^95L*{ky
zDoyj45R?m}k`9|t2NY#cXljCV1U;YXYn=F3KS=vHupvH6cIfxzzUZhkfZTK&bgOZe
zXQZKI0%M8L$s5ZvCv4Id>u$g`QGh0DhvFB1-9ShIQ!9lKDFRlT&ETQq6O1?zIiiP|
zEe(B`sL}?~VTBqrX}<;sRSu-l?;VhkK~JOXHyiXjDa9!Pmxn6WAq-o|&Cjp>bxI_t
zV69q|_|m2DMf7C<=xuub{JcCNC;xHa`OPgY>3V^K6jUf=$7}?>nzgkx&(0Ku4T4@O
zerfFsS{(p<1k1A5$*b6`i>g3pV#(B=Nd@p;3o4oo-RajY-N@~@Fk4H=@iWrXkrp6H
zdGwy0Ef>r!PK*Kci{sPoG4luv3?9opfhv~wkovaoJs1Cs0J=gwwFrg)*6s#Yd;J7G
zUb3gz_KK#JmDRYqE&Ch;D5|!f2Ra+xY~|=iL`6nQA)p+puBo{WKT3PeWVS;WU6gJH
zdj!-HKgoo&+Lu&QaDVOuC<n;r*hN)f?~o|pGCQcuX~twS3kwT(tNH0yxe-xRWdd~v
zdM$V%iZMUGeUCp72KpKrQK7$90K#!uq%%~rT7dGTwifOKKzjV+I5h;QaL^Sxn0`OY
zNwZ2UPkScbDa0s~_H?A)eaQN=x-!4Q)Uuk)2WSt+#>T;+!2QX4gdL#R$~>S;U!NC7
z_SX@CGx7<L(k4y@p5NZRJcpdPStvXwNs4*#qOqaji=6Gbj3&#qsT&tzNL3(@O};`J
zSe~Sq!|QvB*>7J%X_kSIjV5^xfS!K@y;6Q=dN16DH0lvIoVErri*h~{6%`@`WQT#H
zP&%urMe_{6z0KTb(F5_5j{^`6bgQJAMGU%wZ39B1F>#P~N}OJCsw<y_kyWy!z7}bi
zUxDp(3ra`83_3pS$Ts5$78RLqqGd#P)cJvG*2I}<iE#yl3sXjJ(EE6MBdM>D`02t~
zllft)EAL4W4=YN@NtR0+fe)J+7^9SlA~B&gewgK6hne=&>n5y)NGf#P%f=XzsF@wW
z<$?kNZ_EvlC<fW@!A8E{7sGo}##A46bi#XtevMvqN1kZ6XogHC8tcsu^EN}z8dYE_
zeYaEUO`d&WNYJ}$G*%5F<hn-?D~fj)IM*FpHe$AeYPN#8tvVVTMYS3*D$p`1B7Av#
ztmLGmwu>!D`~*={=GT*_G2k&>u=XGe+wj%8nek|xp%tjWZ*Mb<#-9aB((67;xJe0=
zE|dvO^j9Rj{ro76^MU?+=Hws6lJJmy$~!NDl}0{AdEy3MeDE&`0y6HRIg#vQ7^CS}
zH3^im*W=A14~ie5Bo1_f$vf<pYw?AJg$#Y5i@@OL&Z1P@+1(8iV}w_b2uwJp8-S`2
z=KsYrDt!+3jk7|k0exBfET%uOJ<$XNPgziTwDSXAy|J+|{d$NB20tI)A*P$-U62&)
zi_Oz2<*xzbIDVb)GotPEE%q6`-t6S`=2($y?=--P0~g!!v!;$i8OzcehVE5|()2R`
zPtZ|!qj2SFG0BXTr6uGftG`C;sA4*4WlQ13Ey}N&H@2z|;y;VQxrbNqtgLC`K!!*_
z?>7!LMJ7-W%72%PYsBH*Q{~CJwOR!1Ew80@V{3oEQr;K2FChCZ(jVlJ$3I(sD#lzU
zN$v%puvMa92mu5H1nc?g_gEIKW|MP#HM=u>>06PXkOs(4ncP$mNT609Z<NY7O0lXr
zzU=^q8D_HO)ZSk#+ziF^Vz7s5B|AH<7!A~lBMpYZ>tD1lD4z-=knZ^*&Zg7Bz)qI<
zBFfRoJS;GfmRm*Zvyu74!~~>$9IHg+`>T<lfwllQiM#{@<<SG|)~~v+sQW+x1YL!|
zCWF~OunIWoWIWcnZg%zSBqStlI)_NI`QhOkhiHW+*td%fi^|5z3VM8SrwWRZ=GDP=
z2;A^}osKzpxk50Drb|wa8Q7y=$kPc+n9%6Sw+rL@qRpeVeCHp45F`5OW%<R$MHQfu
z8&vq$rk2;+(Np-LCV?KM2}p^Se*6_*KoIC#=k!`W0#DuMu8HaJgouPln&Z9&9_R%d
z!HcY`WMzU&0ufF%*NfsXVfkk$7|q!S62tm8)?-W3`#zeXG)_r?C_paEx@98T{TxFu
zKQOzAkRD{Am)ejdAQs~sze#itNPf=|FZ?)EUubYJ0|nKn1~xO9@8ep)6rusQ`96QO
z-I=Cd=U+w(6$9(SYxpLBX<O63f{#ySM={$zkus0e)MQjHFb9P+i}+56pcV?MyZ)m>
z`wxgb?F?Rv8(mEUUa{xi|5L9BAtMCNUQj>4?MX^X;*fD|`74iQL1BE{V-@}4iv3pY
zq7hq&^R*)8C`N!GsC!U5T(<UxS0KQ>n?v;jHJR1xLnQ)*Q5yO5=sq>5<iOCW_RX&l
z0&q#%J)$hN_V)JX=I23>$@!)&J`RPRn9N&f)Ev&2UI>-%i9siyOOL9Bn}BmP;ZOqm
zhdSE^jFm%=*8~|JK+>5XGy5?}ZSCH!u20c3Um<uh$)!G&F9Id`N>NtR@A|e_3NO%b
zP{>?dSm1%LfT-?=>ZGb2NC$_n->~;2FkXVr1ewm~!E7%&2!lVsxK%Z4>jIw7fcnm#
zKVQ4+@GTFj%f``GY6$eqQz+rUP&0O@@x5;0KOnuW3Z>6S=PdBB!B*uutk>IR)BAvU
zfY+%oDCGk*7LlO~qar6Jo|&CZv<e1f1;e_Y<_$xp4~5NJ^gw0;`a^!Er3WOk)5y74
z{RoMSJ`kgX7<YIVyY{)b9;L_dFQ3&&sI*`#Ryx}O3{?JN2iqpuZvn><YAZ1E6uhy*
z@-h@jK&6wNm4)X4at^DtsTUxJ={AS(<w1t}I0M*hycg6FQ0#@uPTjNh?GmV((ec8B
z*XK*4XgNv~TdrmueqMXvU2U@Ql|9g=Jvz4?gpGbZ9E^7bfQSaY5b}V<1S${@0j?av
zc8cC0>^<6VhFuBr@u4t&3{1tiVxC`LP<Vn~3M!=DHH4thc0Yfw5{eza)l9JQvcuuF
zFnE!Yn2r|@0ZeCnaiNEY#|qJpH?g82bU7~0&K6Iw0kWj20ITQSd!ZISksvGVfJ>0%
zmi?ubQfzRw4<N`hFfcGNi5|8P7^XrX-RXrfUGb1t?dW`AR$<yS7%6v_oebybh5=tl
zP_rZR6`};N#z*T6|1aX+IxNfW+xs;^kq`w01Vp7lI;Et$yHO;iyHOuh1QBUzq(h`Z
zNfGJpZt0X15IEye!L|2Z=j`8m{(0BswJsER?)#qij4{Xjem~<4^sX@{`J-Kd10MuL
zcb$4lAR2`aUCfSPGI%~*4GDMMLPS6;4}Tg@Tbk!LUSw-9MIw-OpBx=~`R`}|K^MeL
zhIqT~sKLR()2YDW+m*sJGBhOM{0Q^zr#MJ{Av@XS+uY7;;{u?ag^tKiLqo&FWCLTw
zb_3d(hTo$C5d_>>cJOVGk&)fHC8+)rUi{DR5Krnr1P!?^T9u3f2-cy)K!|Ls8**9D
z=R0x8d&sQ9THYwa2IsR^DWsi$tWCt(Tf?w~^%5l7c>@sLAHG6>dgHd}D=0&i2P+>)
z)Zk1DWb<&$gl_4B>`kx2mX?+v#)q5b^%zMJu(h>SCO?vn-S`<ltOC)7UFl8alXu+v
zDiE=$p`}%9+BGvD$JbQ_r#4O7%vE0nU<Sx`0~(vnp$&M1;tuCo_?@UJC`3Osu#Ck&
z|94|r{bJ}5Dzl!p?7unr*#BpdW@cu_|Ex%Ji<XofV9=mQ^BNX@)r0TR-%Uh=mYcuz
zb{9*_T1-jVnb05Xk32gSbV}qqwH#If4%*M0{<`AQQ@5jbMA_qR$Xa2==_+u@I-iwg
zv*IU=zd?3zUV<abl<0BJg_fbVnW5m2#!iO>DXEv@;fbrUd@U2$Rb3Z2yUxE5OPxsP
zcul`0*~fK%4v{Z6?^5!@mp;DfimbdO^OXaXO``tloni(o=8Jbka(;&LFL;%g@&4S9
zXQ(nEM`!o?={~+sC=joP5r1`Xai&kNCE&%FS+^s-MB9aGueqV@os^OdHfknu#e{<1
zw}LN8CGtyh*)M#HqnGR!kQ9<+DbSZ`WGs-Ek!18;#q~SbT$rHA!#uNjDniB2`-8@^
zsH8;7m)T1R?R0jBSHl&D-z68!$zEEF7kyz|L1WPW#v!5keBsB9t&NY4JZRSMV#ym<
zxm$#@-fCznbY7yg@Hf#7Lg{VFxR-0(y<nuE$8amJQA7K}(x;8D<aJ4m=?bPJiWMXz
zS<3G*waxhqpITE454bnF3iQR!T%yu$-(du#{ufK*{dc#XRy~?Q(5<t26zE4<z97UP
zS4N3iMM*1M8yp$HXeWKQN=#Tk{P^s^G*{`<+f}Eon`~(g&)X4Pe`y=Vrc3(FIt(jV
zd)qd*iLl!CLobK=Wq}e`iqBg<uU8%ulUAtuY?Px%UweYCdCT%U^IC{l`g8nx@w;J}
z3Js6WdB3A~c*2>JYowtqC6dFl@h*Aclv$D)xei{+g-cudj&3ns`%K9@O0u<CSFyZP
z(+fB5EH>cq(0`v_T(7Y<LD9P%aEGGQii;4{R$KLp3hA{w3(mW*9bIA@=S!lo$1~iy
z&Z|w@xK`<6kax}6UF$MHQxnC%koAS-5@V3$5nuSsw=w7MGm(ny)>lpLqnR<M`xtyr
zyu7IsMc3?e@cxt*!e8?GCG_!reyd=egm_6qh6$+xiTT8s?}p`zp&Kec^DZ!bGCu!<
zXR=9_igC+i<#B;_1wO8D{F^kH(?e98YN<WGYkR&0xc5IEV!SqddPUVBFvr3&NO97!
zp}`)_6`yRaBspC^l^9(Q-_4PWL-PsE<w&y(vqQgbKMPh(fkZqs)RLje-nn!c$FpCt
zD`okV{40KJF`8fHeZS@`r;9+id~kLj(F&2%UyQrdXqJ9XlE~!SURk}?DaM=&c#C6r
zlu4Hs9RxnpV6WH>_=#9unj(vCY(cXTeN`{~BUN{{M6@{TDD28-N4!jI4q@-pmM_ij
z6_!x6ExW$_m`u@@<zi7E81JCUIlIsisDDtinQLDobJI`%ZMIRPH`+6svv>58{kM{$
zxxbl(edOox8ak71{d)fuZsT^+6P!15o0nbsk^<e94+PJpQ+{PcFTvJAB}_(*t&&Jo
zi^4Psu%A*cmSF8b6%LEZ81B=`2=z4>DC%*dlU0<vzm8emOsorqr8)0KCvA6^ClT_j
zRQ*UoJUBo#TH#fB!(QsY{iW%{_12i2(n7znlCRu19pcn>HXKMP-d)KbY^SWXKIOvC
zf7^azmop}HRqgc!pJ;Ths{)?;^illtbquU;qh`IP+T)g4^L9&jW3bRE9`l;|JV{rN
z?=Lw`cioK0T|9@_yXx`|-8G@+5a~|WG}Dh4r6@2`YW-13Zp&VZzTRL;t}+zU6;^lU
zlbOELw}Zf5A@^*<0+%DV$s96&<~n<hmBMeO!uNNb1|F@*5uY#c981qw`_y|Ov(RA(
z?e5xPFW!#2FCzmfKJJ%D=d`m{wFHSp&nXtorlx8{uT}bNOB}Kb-_wlT#1*8E5uNSP
zmKWl++PvZB#y3-Rd4ckx{FQ_0QD=KyKf%4KGHs^YAD3$Q;*yF+K1iqE*|*}p<z}}k
zyS}0BFXR?p&O7XkW{{isz&PcU9F0Z)Y-05@$`6rF3Jf7CM4vRV+ei6q$|dN-&n2cO
z6}PJh6%E=bHjp-9wRU=1vQ_nanhxq@iRp1YwP>r-4i9mlSC=l5tJ5!UMN=cDm>zkn
zSR`}j**kRVEpN*(Hfqf`fo5N6!fGQ!7t+J()YNe2KAB&1Q^nOpK|5RMrSdM>ABdct
zwTeD2K{tDh##=lx&;+pu9o{c?1t!Rl*h!A<jxxDi$$#76{ied8<V)FbrKgNiiDC%K
z5*{u6^S84fhXbFDZ;Kxl+t!7?q^`Vjdz9KHzJrOxF6KVBsAXoO?lXd`emWitsx9oc
z8KFaEX7&+vYDR&}(Nxr>RK=8(31{-w4nxGx<4k1AR2ASnNO}0RKwvV43j0Sikz=AT
zfAAHxY*R@!sxxn8RJGBVx~#~#%Xi~ni+`WNv!~p4En#^=TJ~k3p+QzW(8^S+?V1K^
zX~jKW2ZnV<+M<yGZ|T+x<U0N&#1^8(-_l*zD?DBMEa%^HxN^@J=lMVC-VJ@sbJr?v
zI=Hs1rQb3rZnZ~_8Y3cN<qaM#xz1<n{J1Z74wTudg!xqh4*9o~lFi*Vm5LqmQ{pl%
zmzpf<^ywuk;@Ajz_3Mh>8y^&1)exm3ZDY1Px-0m7f5oxu_3#6qh#-lXs{Igb-(aIT
z{K1r+A8idzt?Alp^BZp^eU)f#ND~I@t@E_d-d1IpFG?ZhoWL>9op|8mV%d}<njT>F
z{;jrurP7(xS06BLyM^ZxeR^9y#ngjIB#cXYbG1Pr-kc|KGzRfFNw9Lh-Gu3@bi=)i
zSu6G?=k)9Gw-Y5y4k9V+$r0^Ct&Vs?&Gy<)RrU~?1jeQxRxs|oUi~1u5*jfxY~Vk&
zT=9kK3ESqwOd&GEv>2-Cj)Sv%_IO;g!_=3fF589aU&X+n)DV$XmWcP;_?}A;T4G+K
z%(5`VqUbhjcGcb@*lnTkZccpQf<M(z=C)z;rsMhg>f7GK=XJOOY(?}`owNsNL<_&{
zo?Sb-&YH^3@l8}}f0QYk`0z&frUcR24UF+eE&m$k{0m)&`9GxV6spVE4RNBE@2kyW
zd7l<vdqgo6(eKYaq~&C{xLHuTY{(L`*;!c`fU_kcQG}a{8kr}~_xSVeQ@)X_7;HD3
zUd)ahG>A{B<SaRC{LH$Ke%Ui(t#hE3h1U3^i>GM?&t=mWj`yzKGHB<CJ?mOp$wI`4
z2rwzD-1vf(=wh<x?MR28x3lu@eI0jV2Y$<GN;8>@7Y)o6E06Ymnu>{-Dt08-b@v>_
z#yVw}x?XA}5aDC>s$A-PjIT;D_dLe8mW}_&qGFk^FU{cW`R7mSxgJrXJb2o3ZcZVB
zwVO~phwFNW!BlTTpW)SaBi>i&R=qqutui>gJc%~1DctNm!z%N>L#{|P0^cbA-D~Ov
z{4}Apm}fd!miwE!vRRS61vIHl5|3Ul6Wv{2B_g+?U|3fgxIpf2vG*=7upz2LwBU1h
z&Z07hkoNADYnRERo8t8SMpqNAo)_!!8F1kFn4MR$E$gznsA5mYt~rU5IIB?EtWhNS
zKAS7k|I3pl&Y_#=`xqM9pEu4V<<iV3txeOi9qBJ!ouO#`*~p^jO5Odl?`LT1*Ym@V
zl>)5`*A?Q<ricpJ)8(A?W>a`#Hgxktw{$*QZS;#8rq@NaylyKRrk4^-GSzpU1Y)1l
z5cRGSm=r5ej}>y^3;8@Fw$a5MuKBJdC;cLqt6jN^-nFF~f%Ab;JfCnUC0EY$zVx`{
zvM6LL(wcv|qo#&0BYJyD$pnj<#OPYqEt9W&MD1u#NEqIC`g)YtrzpQYpZYoasr~hW
zQsJoTop))|iemvuF7KMHr}-UNDD$0*Qsll(N}*Yk8}@Hk_G9nmR-bm`r#oNXG_Ro^
zV#F)e#(qn9F=`{<A^!SaYlf1RWov_v@_QZslCY3x@;n2sMK(g~EnyB8X2hF~o`dhT
z7~>V*Co?fD;+0)t$(J+g?q5vDccvQ+4->Pm_c`k^A<j6cDH^#NJl@Nkh=a$XeE+=P
z)*db%MQwSF=oQ?B3_^#(wDrcDy>T5kmrWKW2z+yR`OPz`i!9Y%q%D~jTkvVpn>?tu
zDbp0F)BV}Zk8%@>T;4cJY2cN`l)t%J39TXRHw~Lvj4<2X2A^OK17TF7plocGh~>Ol
z-hu}lP!r`WN+#!KGk;kyG1`909LGQvwIhp1Fmdy(`MobkrV>WIF1PaC&PME{ic8sO
z1jpx1YoeZY3mhxUFPsbT4q_G|7dGjcAY=<A-R(`$u;HYkAWIvasM~S)fpfr1#3#~>
zIeR7Aq@9cWsCVKHMkl{VZoGPNx=ixAw6kGnP%61=^2<!xZgW*@Nxv114NQ&vhnr|F
zcODVGSnysP*sb*<G<<gN&QKUjvU8?AV-s5g&HOnv<-`{>lcv*;gep|*U-NA6CU81A
zJU1=Q2?{3{F*WuQ8qR*xXKVLzzGsPEv3%l<vbetUp0L8wg*(3Gc^;n%<wSnoQDl>S
zM(I@8eJJ$RLUHK!CmLmjOKgkAWm$}7?Cp2`M5<RUzeIJU=LRG@u8&|Ca|cMSd>|WH
zv+?neXNs`!?H}ng4t3Cap_bn_NwUIy>ripHx<*lu)6Gw#gw%o8Qg^G|@1?-!p8|6Z
zsXTQ&qx|0%)k}Xk%yy=h+jG(KqO2Y{Gv)5C8QAw6@?0I-Iyzf=+w;x8OpdIqZ2yhs
zBKN>d$=-rNmXwK+5vVJ+cJ`!PoXiZ;hL*<mCZx>FY@GiQwz5Z4ONodB&vPfcU5RG@
za-pSIc-B{~*Th%2TLn*RZPSY=RJT82s@_zVlgl|e@D#JFk`a20V!m<8E=hH9n4ef;
zIhsZA;;i|V-pRi6?{p}F7(cxTVHB_nxTZQwayyzsno}U;;vK3hS9Q-V#y;`6MOep6
z_JCw5RGgamzPD;1wWQ#qYCE?pgPbI*9mX$3*6~8x<Zj|<v?-s%efz{o(5=Pe^i%?-
zw9C8T``Cyxp<9pVL>czP@q^VH_2Z-NUlhc7?z4ZZdy}3YBR=7LfRBv=Q^4H4Wa;}A
z53f*38{%flNmem`-Ke_I`*8mD`)SqyKLPc|y~#L?kbV7?Z?~Mu&~^JsCr-a*tZCuV
zd!VEh%OLRVO~2s9jFe_XON6tciKH15I=#ZpOXnM@a!5nZX)J%l^2}pb5|xjQiNE2J
zCw*T2&2^laOXj4VOB!ufpSMD&l+cttuc_{6UUkO5U3Q=C!Yg<7(;qLb_HQOhI7{x1
zkhzar-$LJ3nC8Q3WYRlJBN9u{TrG$tbju{`b2TBht#lB^1r~3*SWWpgd@9N_cZ?)%
z{rG(4Guche-m{Wo$r0Hi_h`H$MbgAVdqhk`Ec%sI5sle*$rB8ft}>J5T@}#4)*z98
z`<iD*z;_PU@60IEIis1&KRb-knu8t`i3cX3-wM}XOv>lhPfoHsck^8EMIM&(f$yQc
zYv?J|dAaHoZLXXTr9SJ6cyU=ihajN(5p|&F_EbPkCU2b;Rjy9i(dv=Dz*||r26vuO
zuS24<!r$`<)wWvg8R?!)+a&Z8vdU=4d26tG?e_mH=|YcGoD-=7(I?{;iuucPy3!R-
z-0!)PTk?46n)mcrr<!m3W~p7TvE0qou%>n7myv4r(tf&}L`~W9>NNQbby1TBTlR}-
z^CwKhTVgJ(gg>=Ch7Blmt5+MArVa?L)~QPMi&a_Q<MuQ*%^i?ilqQ}L#TFcrz{=<U
zpyP}2yzFgr!s*I0TP7WkQQ4-%ur>mh-v-#<*a!rVMrUV9zqRw^D_wRX4rvBA?+?6b
z;k-XFCN?=mU8uO7M+$`&3{bNx9%v71JR0jDZ0gM_q}P?E=y{Mp9>zP=M1EoJfUS#1
z9EV?ROF?6Hx7+i(*7p+Umnb*`s7LqnUucJGFXDH08oXqw63019Mf*zG^hhI@h2Pw{
zqBya^^!o$e{Tu?7;0LKc+O{7Z?jI(ZrsR9!iEstZCU$j%;0CIAXj}V*9x?Ji_rQvD
zWsH!?)Y!%HIfyu#W6$X)Ec6!f>qrwOsT#SnR!VX^Hz6^%`Fy>FeL)k-_h6N?$$Ds+
zPWsl#u35rW4o{7_7Kss*Km5ioYD|k<9HjJpb(27W)g?K(cGA#4NSnmjB3X6j-4~O@
z51362CVbrZG8Bd$j9Hn9yE&gOBy)>+Dh4&(TMXGG#Qd&H5syXVYPu=(vW$f%q7Sut
z!Tc_L+qjQoKkEK@{Y>1fWk&O7E3+LLc0xMX-OIxDHbmS49t_Ut>DMRTwIy7hRMQ|@
zedoH=E0L=fgVS<V6HCd7SK6bun$7Kbfw+{opQng7F1DJm<`vUtUZx|7YTg>y2YBjq
zDa>oNeY6D5N|!!POr5>H7Ug<xN`b9$Eg|%)E(>m{)@qSclEPz6gENv%?N_lH^aZc<
z(h%e#iaMG;S6=R?-g*4gY@F7yxoY!j$hGUE*jmXCC+}G1DBf2xt-5f1(+$l`=$#>&
zf*VG>J)b5QPe(WC8y&A?WRl_92K5nP@UZheU@=~$U)bIK#B-gMl0}20p-R40&d<!3
zhh}j~u4c%L6W>ZU>K1XUac~}2_JWd2XSBcD;K0lUol9~Q_4+q-et3S-TqOHR{%Q)1
zEUT+opA8F(pzN(`U@zY&%Pr}rWsgrskwI4c`su5i%NNY-gL)HU)-vT!lQ|RbJ15?f
zEVp(NnAkOS(CyV2Of=)9J#@}3E!<rRU%t_b61=NVmn!D99fI3Bvt%2HC4l|@t5|mD
z5XozP5;n>Wp_jf{Kl14V7CKB_G&hFSZ{(6TVtM3Wycbl^CU_UaIEqA&eIq494DtOv
zzI5tThxO++;~Ezi=f|(Z%(c%aX)R|cI;k06mFT)_USgRfd(Y1G=lbLm*#z;9X-$OO
z8><g&CK}Va-GtZNtoSlp+0NWh*%|aUkXWnd$hJ+VACTZX>s_vPsxl%#4B<3=|NGF&
zit1{8;h=a;-K5mTPo#|QoQLvPDn{u>m&_0MZGWs6el8SQnm)a_amtD;u9r`cr%3ip
z_NCU++C=sk=kA5=@6PuxWySRyy|7=(SlLgSTv!TJQYbZ~i8PdGyLQ*Zw(~qC#w+n!
z$roJpobSC}GbAS|9tv#a;kOuV*s6MW7iTc~xYql267okrLgVy#6-ITsf=K}p?rpEz
zXcm|qvT~F5T4-E`3kyG8{^wbQk@I?XrDtR%uZ!0(5z*11(GQ&srCY|v7WN$zDE1vA
zQur^L6$9)Be^AXmxVwY?Fb%5&|ERmY#__e{kJ<uMm|~vL*M4X>#9(PUd$yu-*@5S>
z0>22Qyrw}QTYvKdCBK6wV!F;f1~vzuFZw-|rw?{k&QfyfMaQh04!!;!FFZ53*RNL4
zZ1t3{uCtWlv#)CV%<ov&UNcbd_+2}bzlu|-TB`D0(w(0+zo1isFKSL_?5gD4<>8Xr
z)N>}P6GNHb)t-J>_{n|ARVF5Y_~3N3v-YGH8xKKqWJit?n}*%l4<79L&mQriNH7f>
zC0$g!xo1T(AN_)<Vy#8`l1MiH56c@?w-cgO)S{mjx7JO_#U!jEL}Z<EskWX>wDARh
z^c3nN*(%1Yc#u>jC@sJ7=o3bfN2SY;)AQatdT|3kCnU>S7w}JORsPhPa_YO+_XxxO
za6UTys>2R$_ai+4IZHi(ip-M5<hvt}lk966B-{%QjA=bAUDEqdO7MzwJ`W^kma>!R
z4zKC7HvjBUFA+0xtUM%LIPa(Hk&a*fL9+7ED@Xp*@3cJ52zsXP9i4u<EBUWjZ+4dd
z7CWg!@MLIU&!FV+z~05$kU_%YzOf;Ln(0G`oY*0p5;rt8HnAsVg)aq7?d{|ZZH25X
ztgS2!E$vA;8HB9Nt!$O7?;9902pc+@8W<{y2_DB?Oe~!2zj@*RDG4t^m<;>kt)O@x
z;WYIJ)E|=5vkJQD3kx)T)ibsayH$%nnq?J)4DOV7EBY$m3hNF^JJ7sBCfsXe%IYWl
zlfUm%%#nT2N2*^eyyzEoP8D{o4H4NaW3#)laH>8bbgYW{Pl0%C{BY<T1MHa+J%bg*
zaKX@gdLXhs&bU`uJdg9*V^Jz9P?NOfbRVD&sCkysD+#aP8JxWsOzcHTyAUb;DPn%X
zB$~GGb}ve_=FW!c)d<HMTrq(HazDa<Laj9<C9nH{Fu%+0_3IMSE{3c08~nPLpMv-Z
zby;5C6Dj~2tJBQrMz`%@GMDcQdBG=NtsL(fx;@w!;&KpumoQbTjW2Qr!G>UusDHSq
zd^mP=$YEs$nqE*Mn((0R-|3?(1?-YgjT#77f1_Z2!9kGU|K%S%VaS|F@REp6WSD<`
z0e<E;>H>b|*C+cRc9H13&FO}QsnoEY(OX>^0YwDM0U+f5jlP?V*CYnfvN3pq7Akqt
zl@eFHQX(DvH)er<HdOfN#UK8UnZp{?kY9xw_E~E3+pA+^MHX5oddC0u#o%vX2NG1W
zeR4NR9w+w-&-Y*cw6g|pQw2m!fan2EJzMMRf1{${5tNmnHlQBJ6a7WIKos_Lh5&@)
zetuGLAPqJeC#XhAlmxaR<n@(h-~bbd1ME;?>?skC-OMW>Ab?;l1U_;UlpDuC$l}CE
zNzSyfK#(Xi@RDiIi<h0fx4PN@wAhW+;7JGVMlAPE@u=9C`-Y4FCX&X66cc#um%PQ#
zQr!o=)CB1P5Asz+W+4VHb~y1ZCoPrNMy$NDjKg*4(SJ3wv=kj0nrsfjhAs3WP*oNf
zc_Po9yY2!FZWTP>+B@xF6dLcQ!tDIT9E{iCdTX;A!i24y!Y_Ls-#j#@5KOMrWl&ub
zw0msM?7X}OW}XUj?liyjmL7ZdBh_FgA?EiaLY<#QHe|awafl_W;m-OGD}I*2kTQ$=
zD;`3Xpc3f-EuD|g1vn)xvuXxuVZ<)^&6^XZ#Nc0Ao9jqC;>K%jMY>^arcjP=X7{cM
zp@fXg9Z<l$5o_3A0(~Rpw{V~YraB-^r*OQg*a(hj00AKV+kc>mc&-3&$mlWK@C=+t
zzjeEjTq>--9Vwht1d;t4#nq)HcYrNWgH{J*XSVS2kflSRLfQeS0GjkMIK}yYAsji?
zsgN+R&zl@-wue7=ZCnQmOoJ8q&AO=wlxe2n939F7rh&>0EGIZ%OY;1!n{CI*vzwK@
zaU)J}4E^{e|Fc7Gpr71lMR!9hNDe?oraanqNCtLgY}baNi{pRZC?P+HlR)Cyz1je(
z!0%8$eL-z<JeZ7$DHk-1fJIx|-bNmdj_Sp{5e0>w?GX!<h^-UYb3tB@ZWwuvYtoT`
zL~^V%a&V~a-9K#!jyu4rxNyV4fHYL(>&`V0Gp&w}CW89~5$CK0tDIb0-TTwWZ}RQS
zfB>+4_5wW40qGoqqi;T5UY$me4z91SyLMoTrq&R><{`mxX}%Er^Id&6H}qh7`c_cx
zLRl>(5;(GUady@Q{=>jvKFBspZodF+2}o(^tP&gcp!Kg16Z`)e($7e4nu7WOlr>#k
zKj4LIFZRh%OC1>jECBxb+H8DUx2qDz^c}0{^qS|b;a5J)S#=^kcOldOt)QNUo&oBc
zcKs;1Sf2S<J@6G8F{T1~1*j;B;0i=4B-hy^!7&70QSg6XXEo8cRstW(@9}CNdd_^T
zq8w+F*d4SFP=PKjigMLI-hm5T8GzV3N)AUFa}JM?RbV%Xd0aS}-(H~UdU9O=C3_kQ
zuI+*KFsU65+u>}VkcSrWb!fL{Jj~?Nr%&LM<nlT!TW93t><8tnzKY=Ub(YP%R#3Ig
zlFJd9ccru(gj3r7d=nI88;A1-q?gMCVI?tFvX(L{xcL%1Vi|45Ua1h3?W{67u*#;A
ztVV<&wsA8+zc(RL*6E&2!zu_8UtV6`WEKu<`Mbs*s4z-BK;P)zclq*Vc`#$;B5_qm
zY9Jf|D-2?IxM!Bb!}DQEN6KpJ`-&sdq47en_oRT-2&6XDP^}E)QRuOu*xzS&$yX0V
zk_DbVJ@i`w`N9n^L?pn(aDW-5hsvTCo0OVLNltDUbSCJ{n>BPP5sHysF7bVy`wuEf
z5;@;Gaw!x0cg>HCrp%0kut@hSZA(J~H4ZA+b;d!mwNI5@v6m6Tzx(ZK`VA-qjO6XP
zJlp`y1yEmf=VlQUwzi_86$kq}H~8G0M~#6|nmQW}M#Sn3e4P^Dc3zWo`RaoWM$~}a
zM9@Wr7|oVZ%^v)+&9?gCNy6(d+G?{0`y*gg`$2Hr-Bj7h)pSp#5fm-!Yim>{Uo-L+
zMSyOo1Jme_Y~P=4_5GhfO$k10Juqk@5Qr6`^jG^2H!05pUlK})+JaIDEYm@`MMVC`
zBNWedv&~PkxEG2x&-;RmNEPbwf@>Y8Af%CK80ulh+;uoSbOi$_lDKDJcAk$MCL3R0
z-ylH{O@lP}t2TkhqNo3vKf^7@`gS$mwAzypXxb(pBd9ufUgu__B<iQiKit0^6}OBN
zGs9%O84+vX@vTgbOT_>w00{vF5>I2H=q?2?%j#};NlQz^Z1DS!vJ%KBpszs1Q-G#S
z!4d{ezL^+UlK7x{>O9DNKv%o&dBzX1x$#?e&tQeDI}6S^es_DT2#_ncM6lH9KY#P)
z^0jN>-@ZMhhEdT6y2`u96Ltjm4>5D2F9SG<GNw+<=%Uul^cy=nY9vMvA98YW%~$N~
z6r@85Qd=+f22hbeiR=fA)qvJ;D7KOj7jHLVMbWE0p21jYQO{Qy7%-3!J5)J$&0R|8
zIPA0nfe}d6H9sxSg5+moG8uhxeFoNE6gx0ALhq5LMm={<8I~IT=OE^kjbIk>Is|>(
zaVx<h*`U_-%6B}GnvZTkKXb6ysuD=J)U1zs&0$_ZX)`i5rkuWp;9kZbHr|JZbXm<u
z+7iL1H%|h^Fcl-QkEcd+V)wWEYmWL~D%HdYj7#!X9J{N23`c7tvV!jG!!3edUOPiM
zy{z0-SomxI&no~+WHg2)>+IRHp?RMpcTm`1=^zHPyJBl&W2adm3jd>Lf12Gy7c){6
zw2H!?LA(xlU7FQ=s6V?)Q~B*+5|N#~`mG^2bRAwT-NN+wzM&y4YLJu1gB2KsL-V-p
zzE%lJfGUZVu!ZI5O8EyF5IW+~sUt0D)b*#7(u;F)E?m6m@?(w?Kps56wJiXh7qvUQ
zOO!{S|12Qz5e=ZkS?tN&+3CAI+Q4WHIv@~A_o%eKWfDROJzAc8@Bn5?PW{i1!IAJy
z=Xm|UjCzTDSP$n<P}MSkgaHDH@2{<dhxi};_Lq|g2^#z#e>#7m-+mcm3XteJGa`W3
zj9yW^2V9~jf0`=o0cv%1^?`0I`W$~w)1%eK?m`8*#Dfk%75w%nA>&E+PeC!Ui9dyF
zenrWzf2}w5QH1|rQSCqP<`3Wy#?b!@gZW?X{hz=8pMCv1iu?7m|NiB_U**3w(w|0L
zM#6BKpxAI*%>JsE6S&Hu&j3C+3s9|+1jpg7{1zrqY@6fw^{4^PrkmLVQnAlKmH}$X
z@6iwcUvON5rHas>0$`!@J|GI@gDIPOb)>WpoRn;AxF?@gg9$ObAEMZDB*<1>{cQ+P
z%qWngH|oV!u=M}A`Har0_wVU0UGf8~|BC=w-@&0FDgThGSFhGof&BQ-4uBLhV}GLx
zXosrVx*kCB1S>lf1y`U!Qa1R&A#ySUW(-w<ybQwhOHbFW07$Gj0F}F~tgNQYk=U4D
zQ=XC()UejnBs^}s_+3sg##&5i<0FvJY-F^Wv=;{g;cyxi$;6Hj20h($aQF8EGV)|k
zpt3s7f3zzH0!=7|Ef>WuGdtcI!|i~iyd6JQPZdbrfOG_gi`j|9N5T%6q9P&<Yk)92
zbJ9RtV~LgE%zT1?a;l16NZQx&yo}3D#T}rUsZt@9y*dtW=&Jw$?%&Ubwt#T25U6G)
z@#EwAH39>PNhvTB*A5^p&cBAov$sG90SPEnWzL1c1_mSGa^qkCZJQ7W$NNuw#c6^$
zcB2ev3J{rTB?w>i7d0?2*Z|rS!EEr!?u;T7UfI-<!tp2nLhSs$M%0@(Z$K;vwPD>V
zyf980bb&Y;RE?sEHrCc@8u>;Cl_-Cocqcfwu#lM38jv?e_brepH{>H%eUa4Gjz&ij
z5)!&(a~%mrKz<mWs;!+J@MB;JU3XILgg2?CrUq?Jqmn_VRV1wjJKC|PMgU=tdhT1C
zzIJ?(_H+C<hQR><8zvy$>@`2SfQIH95D@OU1YR)^D*_7P$mMM9S7g}2$Our>eCXY<
z7dkeP60r)PLfvx+f3iAEU`3ia?Bl5(MDlt)_8qC4W4MdKax7^GR&H>}Dk&<S>>6cS
zkbaCT)gg%bz-;j2fK9F<2q-C519^rpQ6hIHt}GhvAjd2rco^-B+#4Rh?_gU-KwDkj
z+}SybTGS{oQGn&*;RP=6I>3&PwhXyyaUj5Gw4RCQk`J4S0hE@wz;!T0d#b4gO)24^
z3f1cE#e55gO3NQVat^+W2j9~UK^R&<={wGT4>x|~UVn;T(}r<=m;rJKm?LhHzk<eR
z&!1QLsAbG;6i7%*<G6;%Z611pi3$t|<h~Lv0C@pYOE0gXfq@-3?`sjSPzkV@Ngndl
zJGXbY`uPG2l_yOKz#){<Z%LoQZlTz$M>4(4bt{8Yz5O+kVSyYmIhXCAZZha^!C<rq
z!vWi?C72F$=VbuFo12?UN=gEAQvc8pDk|!cv7T>0_&Ti9503yiIg7-aw-cW9hX*j(
zu%7=Wfd}{;fLIv7+6l~Vi1m={`Vl3tw`W4!3Mg3U&$N;!<0Ud40UO&<&3mu%PeY>O
z*WSK`>-einm5Hsbm{?f+AjbyXFYKC5UMM*CA<q*6^T1B7KA0JikI8)Uu^FQR0`g0b
z2J)tXDg^#8u#q&vEbypM^T6up&r$;uS7#M$rYI>>!MB9p4PjL|Y{1%V`u^G5pap3W
zRS=ZmcVFLdx`@QK{O<oj!NE19FmUm^^~#(T78vx{MN|Dxnzak!Nx(wD^Us+<6zAdY
z4j4=-ovSAUv4oh%W3K=<IDxlB2th3m(rK7C2KxH}8-BFeCHK$yA_^ivKv$Om4s>N@
z1;oJMjVgu7sRIJJKQreqMBcCAD%cDk1sjP6@-DDB1JLwINABTPhaf{Dn1VnVDA_+k
z)pRHB68!nDaxoH=rJ1!2v=Edn;0+P_&s_9+`s;llJIWfl10(pT%9`9%0K4Mb9e39O
z?RZej*C62?np#@fnuU$=hcFpWjsvRZWJc~e*wu%HCORhO{B4~Uz`{UG3mBS*(HKbT
zC2~$w_X!k;17M(P=GOYU8Q5jO_XH&Z>Zu4%-e_d?4k}ojKoq_(HKhbo0$7^B4{;L`
zFeYBae}aSG+!f^S=-*6%E)3HNN~m#bw2hO~?>rB*ka9T!SjaAw{8VgUNj_OQ0r3Vy
zR$E8s?`<!P!wfTgqo%Vn4-yY6uIcKz396aPec3lo-l8Tk`h^Q?Za6^SgqO2BZwSao
zaG(5_EYyLE0!%#yhl8Y;oyor!&GnSpKTmFQ3z8PgL1#{46%))pKQpvr<nv_w0<km@
zG&-ArKLp^ZnTZJ-kdU9e(v6?)*09q76v%Hg@9)PknnO}cMH68w`zzUmY!IM$;a|Q?
ziYY2Ae6k`(_G~4<o@xX(7n0%h%k%RQQBi&;>yKrhADD3+A(TaeR{ybl;UgABK<MsY
z_HQHhQ+R9ZD8%5fJ309Ilke!~2C`=x8U1~c2n09?vdgz4*Bjzt<EEIVnl<Quo#SL=
ztE}on;uRbbu(yi^!YUREGxI|ppac#ABKG9T9e$w+4Gm#5&>;3;9M-2XBIy%<ULW*H
zB$BhN)OzBnfCm6G|E0$c_V)wuqdoo*ti0`Jg)j}BSeuo1W)ZM7<Dx<50V`n(?A0Ix
z1|%A;yLTbf>E*nRgq$F+m1DgVS3K(4G#<u$M#<cq?xbr(xCb0U7ze+MGy=h<9V%`3
z+Zms%)#r`_+T-8&HLv{73@p6TUsuwTfBN((cutamCxQ6q(grgj%#CW<x;oUQA3qWy
zGkbp;q1~N`$9nu8%#J`eJ89pRc7b_oSf?LaSll_%)YOD3_N}~?C!Hc=>u!X%w>LaB
zu!@>{dyByG?YVp6Oqbv4^Rv}G-Y)Epo;$w%_<_4rrR`AuWpQ>E%m<|b;9t$mu8{o4
zh8B1#NE`02jO-LFFVoZ0OG|WyyF1|A`rGv72yk55*w}Dck3Z&<hOI%CN+#aPRZ57Z
zrKIu;3Vt5$uVO3+2?{<`#wQ{9y-R?Z1a=sSiHS9(5T9P}hJ=R%vRuW<kXBh5ha3fT
zE!h1V$yY+xf#4QMh<|&P7au2N48fuSC^~SJ{D7_x?(ggTo@H5C7Qp84|GP|GVgfG?
z(oxnCH`8WPI0y_p-a4Sk9yk8iO0gTbxP=geK}Hjpl(M4F<O-J{y8@hEj{9G{A@$&s
zvFDajSy}mt9^6&G;x^;}PHWt60L$p7k`mAL21>|cfH6t|iXs42CYcXbD@YClJ{9kl
zp38LTF9O$}#>r?CP+$O#JT>qTf$z5jy-yK#1I4fjH=&w_pabbQ7CX+#bEZx1`}rcE
zPnW|O2?4u%)_#pI#QX5@Pb>WY^g0`PVglxe^+;222-vxSVnsHh1UX5u*F+6Y>Cl*f
zIovs7f$_%!&fDiDwhX&iHC1oEhck<NVxL2z;vwJxgbaVRVzH;6BGKl^ckJa2@0SOz
zuO`P=!(}rQZSL-E15;(0R+$@&S$@vRjRf_A%`jdds!RD<Tb=~#GT2($d67{t!OnE4
zzP!9VP|vO{ieTc}c84|>(2Y1VzW5dn(OA<>UN}yHl{xNyI(NF0!;}CrxY}t^uc_M1
ztjvIG?KqEptwr=8*S}l3rk0i~EaT*;6UY)CGJ8=Mzy<2&><q>&Ej1R%)q<f?-l;-4
zvqw}^6#O<wy!$ldtwgvO%oJxu01X5icGv(Mwn`#Bw{Ql6T<+DN{6II9csUoc4)$_z
zE$g&Tp2Rb>%x~-+9Y?`q#gZ)`>cZT+0qfm?F^uexRKSmELqQdYPDAK*_8?(!9DowB
zRxLh+g#5(({BE-rxrf?kJzxZ8E)IN~052EDbKm2Sm!Sl30sy&kb8?))fd)S10ZoQ~
z&fJUoi+@*|`kmvhda}yH7xGz{vEb%}=>#V658&QZ1Zr?4xNuL#NDuc%$SVM}q5HkY
z1|XHF0X3lb2_h#jCh-h8ZJdlYy<4bIEMwGdwLCgHIzFB>1_7ViW5NpAk&}ZR<y-^|
z^+{yjHw1Q!@9mKL*Amy0ixu9*hh%p)0-7;+uPNNW!*uG;85P~vkDoq4+%Aa*Ckl5_
zp+n9?hAVn3=WkE}w2?ye12FS$)Lhf!<mO)O&xM$o1o=c({|G)lf+y_dt3y_=+SXP}
za6;F9QpY(CD5%L;SXju&zQOSo7CO57Yd4(ZnTT{J_~a>A4N)QG2@}N-B<bLt0MvRi
zSdNcrM)Y_-r~3-jgl@Rg#>NbYzyKcL2I(6M3k&c#LVLVCUUkzfQSI#PAV7j7I<3%`
zqsr@F>zex8J35$<l(#=k3yWLr*RLk5kx&5#zZ-_Pl!v-H{_%~)`~I3Db8<MHTGuMM
znV1xa$~+J5o_xE6|HcwDG8k{sX((kPAF@Ur2l0elkW|)KHwJWoQr<%e;1B{uvA!B+
zP!5a1?%G<T!2;8hfGNY?8b)Cx<=;?+hB(B7V)EiiC{us%@b98i1uWp^q<?NVF)7H%
zpg_|{C@7+t71nnAaP7lH=HRZ6i5Y<`a|IlpM*L&@3bVhAOz8b_XlAs3Y(^(rZ)3*t
zc@Um_1`_UOHRcJt&XfC{nb!xX0xZW6mLSPfYkhy0yip`D6Atfe0ZI3J)UmdWAz`vR
zYisHdYl8Vs=!o#7Q#x=%F%{aW3b1nHO|m0d#mH~by$4zrMDM>C$j6HS5C07!aO=Zn
zH|g5nZG{l(!NEZjoIU_h1b8g}xGV>%j<)uXgX0MA+YSX9?eXwF8CW(s@<H&Q^1YMh
z-E}`1!_YZk^>uuyJJTAo4^C28#b9H1a;Vx4cN7h1Z<x`Kk8(IoVliP#QcCO?E^)YY
z(uU~7Wwz74oX#uJOpa%dGYrp=!Qa%>6fC^Znvh0;9qGw^wXC^zKoAY(Mymc1AYC@F
zE}w%)8Xz*H2;oGFlWfn)p5^#yp%H=S7I0&Ar}Cr|rKzGK=g^FCYk*rFznhH&z&9gv
z+2^aEC;Sa91CTVx^{uTQvqA`o!G{^>dGX|Vh=?GH127l_4|D(rg+}q;*mC*xU>Wuh
zJDR~hkwu1rZIR<kRUO&QoA=?c4$fL3pVg@id!yPhlwUt<W@`-rRd48Gf7wR*OX;xW
z#tHoj39vK&fUGS%CIt>qnXdJQ{puFV!3ap}<XQ06+YeK(OJWc;a{!{5&|mb_-uJp2
zLYMvy9!=u+zy+Q0`Wt9vVIu3cG#sHFk6h9nmFj;q)$hjm-~V4#M+uV?1;yV?PYtFg
zEDvxbX>D(hEZpRI089tK%qYAQd<H-PfRO*q?R8c`1^|hygo7&NGOU8+w*$-_&?c!v
zN0mqWl}@8BC%sF)W>$p%$M{(fROG=q0GgBCXIqIOqCf(Lk8hR5CI?L9ah`kWEAQaQ
z52{uOU_x4uLy!R&&L`on8LEee2LRX;+~y`Hui)dqU4055n}Xb2n|&781p{qZLP<^W
ziXtvDzan{imz<4_jh_CsBNY`mUE6Y$g@4M_RILo{cks!T&>VI6A{u7pc!mg}Zk5$g
zV+ax5D&Tkq)zT=c6Rzc62m5$b2;v6lF?_IZhuA{rf&3AJPRUmng}<2<6N;&-)r$QO
zFn>$#LQY|Od;9#B?eXZ;TZra%vV>_HY_~xC4g3P@XA(0TTTe%a&BT{ya58K@QnCey
z6qhuK{;V!=2n3we-33EPGDso=Rpfahbm-T_+;DDBeB5*2>}O1n9!>q8$;#4?QxM34
zA-`hA|I?iWa=iXYW1I|~-(N)fAPV&!eN*fh5vEAB#kG*)<jXExc5f52$<Zk%Y%4-R
z6{F1L06fsIE(r@w0o{Y-2u}%?sSF8e-Wd(fPrTWM76+RKHC44o!~3sj(?=?N<oN%*
zdF4-tv4i_GI*X4B@jk@1J+jcEe07kGIK1EXG3fc~f4i0c^{*9lm=xsX&mt9Hlkn2W
z{7suizarW9>XpsrR8BX(6&z||<C1zZ(Gya}y&_Ef9%Fb{N9<xY1vA-kmptj4|98Kz
zb7TaBbH_m8TOsPJ5#~!vS?0bGt=sAYgU~xwlbeQcoSX;M6f$fzYN|x_DtFz7#0>8L
zjvam|aytW(ntK*4GJ8te(2TD>PxX4&0Li7|J&Ngg!*tt7?9#Yxm43Ri_niIKGGnvD
z!pGSUhEhb04oyU>mQa_gbKu=&nR?=#=M=N!ALCPX8jVCYy@<8<JM2cu{tw%xst%CB
zU3RBa0wH&CjdtT?9*bup{%P7Tfx4=i!Z@JfBJ*hV@7t3`-SMw5p>dn~sQ-U>x&I$C
z<NuvsSz>vO0ABH!8R4#Hm*dG(-sbM#i=k&%xjuA|==M&IzGd_zWyP9@xHIU`i+(VQ
zb2zpf4#fjbmIO?`S^=q5N66_uRLG-4{w+%imx#{Vn11yc;#nxnn6C&}QXuYzlk&Ef
z7T9IGmmvl<t)}z^f{VOqk-y-dkBW)sd|72hg@`T$cyA@8rAR0iFdkrY63JwMa(pp3
zje$&2EvyhU({Qu_N6N6}I$0?qUm}1nfPdFMO%cj~QY$9aLssH%Xz;ZP_C?ZAwsI#&
zI)WL{GObRnkYIrF30w;`Lqnz08{z{rx~do@cow6T?Wu})UVOl}H*tx(Q88cHhf^d*
zk@<Y&PNOn=pnqUPg5=dKEf472O}0IE9#tCN9ZI~TG>|8-1O4@bPUYUAj$sCG;=bKz
zY>=T}#eEuorLUziwS@~CnnO!`BcAJvq)O$*`#e*vN}nnWdyFUwHa;rnu31{$9Wqd=
z(1b&O&5Fm!%uo{1(opd#JhMF(<P79u?uM>L0{{aE2{Fmxkk&#;V}ex?gY^7E$P7S~
zl1z-&2&da#0B9M!IR;ESFUW!y2nhfo;s9c@=7~^PT45x^6=0JPt|%)h2}jA0cbb;E
zg~Ym%A&c?CHWJEgCM~V5G9>B`!aw_HAhU|uW$Til<AEe9gkB~{n8HjLT#}g2{Rhw_
zQmKzi|GpLaQeA!a%o!3M=lD#dbPTr9=roWz*nyNRh^p%A#q-VL0AU9mf_6G21m_P0
z4sa^)Vm;x-KehZ4p8|CAxncyX-%5I>S~g^pnE^_7bm-kl(aGG&*C{!Z(-sZDvp!Bm
zRT{|3ZeG8`&aTSgrK}uMmjCU5e3hBSMXMn;2QCkpJ!EX@RjCOnU}yv+2F@6cyuxA@
z;0VmV@0-XsxPOSVK}`V5zoh??H{>)(;D{W!J&*?sl{}pbizIXvXwOPe<q1IbJiTvO
zngH&Bnv&D^#l^*;Ac5vl1VrOBTa`c8{aB@-y(8Hz_VyODGh+qk-^2FPe%TtzvL$;$
zuU2GeQGuiiTvgBM+%%NHgBJ|RDsjzu$bDywSr9_msA5Z<WE=aPxi@Ds;4T*lWk_hY
zV22rK6j*Bt$rwnv0*hODj1S&OP&hJo4;!kcO37)^R?vikglpouRoP99w8cOaED~)<
zQ38?#STRi<XJXi0RsuQZ;gXxg?_qase~1<c>FCaCDHjL=Ag*cTK^9{CFiJ#7dGqEw
zI5@oFzV@U70S9JH+{SSCkdRDAcTYO~>}(HrKLGs^k_4xK&`6Wb%f+R!$2vcdR0fGl
zH9ZeDsMgyEZx#1Ub5#{2YG!$J<ssEv#v)Y=yrZMF4*%E5r=jm<?FGgYVZoN*OZ+c1
z;eqrw@$tewU8VX`V70&*jxAFqkt0aSM%|w`f8`9kA5(@r#D_CeVPSPm%|nP|8`0s!
zHv+N-IM2}um>~}|^6659-jVWK&D+%z&^SPtQyi;;k<qlTfQbc_>_R2yD?tA74I`5W
z@|9&oT^XauQ_}-D_W-O`iRU~Jgdp@TVbO$y2C^LJQ}kdc;G~p{EKH!8rP^mt<7f>c
zKLGK5^w=+|Krnv@l)QP*sh?Ubs*|*20GTEo8t9I+sytAAUJiAYfV%`q*mh#!hE_W`
z+nb(%$^)T)FmNULijbh|jxnbAwknnr9O$>H^B}GjdAfENBrr2b4jdq$X`ew<217)u
zxDJn6QP_z_Itwt5YH+TVtUn%aZ-}h@n}S69uNRfOY@{4Mc7>oX<HouV0uV)=nDAzq
zYhS*89oC4{{J_ymN6YPHb`ZKz7npX560A~98P0FZEXxwM@j`Ym+~c0^M|5|{p4-JB
znCZhrVI~f8R2mh4{dpjQYUGnV0RydDj4ulS>L5ee#GM8i8R$mHg51f>e0;j2!cP;K
z+hFL!(GuNxuBQtyb}R>A(t#8RVl;}Uu3Q1LB!nsc!}W0N2guqQ8}<6~yCGtuQFb%U
z9X(Kaa}#8}G?qjz{K(EFHou3tut&><|7H)4UHIJX-vzS9TM`uj2yGsvr+i~>Pyhvi
zEa_BR$=sPP|L&<}ERe9!@fq)-WoZ`fL9JH%OOnM5F_jSoZE*erX%EA0^PCOnEW(i@
zjS27ItY_sZD}l(ubAGcHI0+^uB05(AiUT)E<6DVH7970Dl;(U;y$~qTj)df0qfQ5z
zBlLy&@&-EzRn@Rp^Uw^Gyk$h84J1^kI?U`Ln34+);V$(gK!fYVCE+Vsl9>VarM%5P
zlFej0x3WXE-W!}KKb@w+50yj;^7DhgK!!9)z&^F;AyV!a)mBk(zZ!+$y5t0K@jz7A
z<uGr1F4=wf_A`k8sY4~5j8(@vExOY>ps!-kKu#WHipiwBm8?D>!6U%OHHKu#Uge+?
z%;&JLP{6$M&H0|wV=}Q3keMl>!KSLa1-8%bOZ`lDt=G$5s}&zc0C2LNTfsFS4gu4{
zg<Pj^dvbLHS!k@-YOGo;Tt-qaRr_Kku)E#~j4a<?71P--fE55HM;8V+^#YSX_4E)Z
zxAyTt4O;RrfTds>jcltRo2q~f8ps^%VnKG>Sol>Dq)zTo(;`IVbnd+Ym)`dpN&qn0
zFpY+_0C?6T4mBn9AG6^!OHyN0qcIflR-K%k!V8wcASi&tR2tNwPCjN`RX#n)_|-t-
zKC5{H-mm0U3FMdp*`^|&JfGc+G?b=rP-D2fPLv2Whv!G1eSrgohVSYK6pWDSeJBex
zzhU@|>V+|MT9`wb<F=s?=HBjX9OS{(3GYLtRp=&tF$oG4=FG6uLndN>@!+~gC?MmG
z4>iXN5U;j?8|Kh=*uXgme62$ZplQRjj<f%;Y-wVm1NNUFsJVoeapXP={T_f{2T=PD
ziiWIAPs3zNJ1s+IP!45DB@<yEEbd4$U>vdwC5aSMqzTeNFXIH0wz3y?N3~d=n$ukS
z1f)wmru;dORlJoOJDxIfK=*bjmAD3n`bkk5gYxPjwSlV1brcM@02l^|hmND(PJq8b
zGJ|1C_~206QEGk+)K^eL13TaB2A6wJ`9lKA#+!?<5I~;h(b{4(a}?|AO``=qgZE@0
z-jNS>`~l4x;`h*g0EEv=Je@W3H;O4KC?x$P_TN^AqfWm;!hC}%<-!tEU}ayF#er(}
z5_j!PS-I+wMYIoZL8ddbHqcdPHhTq0goC#ZwGKMv_jm(d9fFvctS)%(7ZGkg{u|W;
zK)u=b{cm202L+%FHig9x9uN*Cxd2D5<k7GQd)mw!)3b42t^4K`V;_;_+4C;!^g&3o
zUI1`A>mwF8#Ifkf=z;b1sMV<$lH4Ho2Cyk9Z5hOqaQc>3+NZ6l89AP|&z0dmwYc~Q
z3Jt=%2#~TFk8N0snzXo}1aZVzZLR3m&IFSCHEcKChzq)Yh$w)bb-1W^w9t|G;uqE`
zeC30yT<Fr0`J+dV0+rSBK`a8DiFO)*r_?)H9%TrCSv6^SK@tZx^Q}ES4zQ5G$paJ}
zVt-rJ)pZpW$h$mI2vB47_CTmAVus#&4_T-LNLDxM;C;jBoc4TIg2cNqb@JvT;ykt^
zl9Qpv*&Ubx;39X9pr{fgkZ*dIT|k9kEolZ|c_FrtnY!z^^bs=Y09{O1_LSiRjWNgp
zH>f}styg3ok%z3kC>26=GZbnD!9WUPN>P4|&ol{zaKpdAvC(E+sQGBwt^`p?U_2f9
z(ArwtZyKD_)Gzc_TT>NiU*?Bh000XtIS@1E*-VLk|BlpZw8a(Y=3)oHec3}w6d9{D
zLs;keWA2(VRH0+Q0`UaMj{$pWn%?cf_XCpMv~7SY<{MX787<!@!$j7T4l4n|WAz!n
z8^5mR)m6n%@#7Xm!xR)o5d8w2Ew8Ak0#?4s_~hhicqE679mr@vf%+1?mn|(Vp}Fsn
z1rX)Kf(kE2mh5h22fGw)B*4rTRyj8{brmUmg6^m5jo3pTvzw&x&|-q1OvmdGbm-KT
zxtUF{k9n950<vr0#;e8n_f>%3rlBg2tSgZgv$I3=3*^SaJvTdh4W0xP)lgjfKuYgx
z6`Zy$w&^7Zzl7nq(Dz=qT(j3265OCrx6ILigcmTYo2&~VKx=FZpF*MlD_7$WXAyhY
zsBnIguBQaSFTGC5Tp#2M#d<GeV}DN|cK<Oui)fsERT5WLhoqi~+U3AR42i<MRB}*y
zfQTSe;oB53Xn8}7LFDya`8Ny>A0MA;5$ppvzX)=92skguhl+8q-hp08TYDIWhanQ{
zNYlnQDj)&EB#?cO;Naj864KD|y+km3y2A_vi{3RnJXJ^>!;}W&s%2;>Y9j%f?$v7i
zL&$E!@s9$1Ms86N0EyQjB#v!I27i4JKEmKJ(t;T>m=m9X00^ja5bqBT444*bVqpcq
zObB(+2<iT!;2?B_m5OH7Aqu6TrKN{@m&~Pk7Lk_$3+=UnUr0a&0g%3J!Xc;wEMlwq
zyLbCxUX9R0dCMt&;pR8PHV6pl`6ZQ=l}C*?4n~q;=V2J=ZY3{tyjA(HM7VF$jYnrq
z0-D_ag#y6F@ee24iPN8j<v()zmur3snqt4)GE0l^il{j=S36CZX_iFrsWD=?VaeZ}
zeDt&I&wa}`rzBYKN$@N^P7ptX<<Cb#_+sgB?%~dI*DfUHaC7XwT%WezIdn^%LnP^u
z&nL6RC`qczkL-u9q#9EfCND}~le~B@Jn<-YWo8pQB=yRIVEA1R`?Of&>A6m|68ej0
z<y!et_ck$Eb$sciI0YUN1kObzxHiV|ReJ4^1-Hk2sCCkPJ5E<XR)l$e!sum{_{;96
z?^M70w=+q;Bn!W2PdVT(q33Zp$X|0Mh(734_C|J@rF(67#<?Wg1f|txu7;akPT{dr
z96dX0v3K4xeVKfCiKYIrh9>_OVl(#r`TOnmbl6!*H+k9?($v{C(OSwVo|8pzrK^|r
zdCNuCch(tm+^Q{h6t2dekK9XF#8^^L`XnGkIdas8Ajy|Ch+3B<_kXFUm*<`l6eqRh
z`m&CJv#8enrkB;eU`GD<rR5Skz52+v`+_U;7x(r(;?mqt38-7;V{IymlpDHf+YyPn
zMM}BbO6c)sNQbrJHSC0IXNONjm$|rKEtFF<UXtv?SjFFGnv*ZbRG@q6|AfeDEHAZq
zC$%yI)8O8_4u?v27vCF~Lr!ILpO=lVTUi||8H=l!&3Q?a^hR0i?GK@&!^1=aD^GEv
zIoz6<0vvsjVQPsgpO_v=@$f0kB$^-MXs=9mQ>~Jmh5V&`1Wt9RzN9J~ew`*EBYX`#
zcUA_MtKi-TS5c?rO$;kS^CCf#oQF4R0yVJeI(@#N>f+Nd&DFTzJ11z(hfnZh;APfa
zVvTlb(^Ag3qsw@URC#?Kr_~a#;``xU>?dU*6Hz#pf|naVr&&jt36sUdGDJL7T>JQL
zCnY&ls`If!x61h#!h<sB4IvBHsdD+#VrK<zF-M!^QL?Zo>JB~bzVW^tZ<^dS&o;iI
zrD!TBQ|ELNIU1{PB8|zmniNBHfr^(?e-+oeDk8%dSC*d>t3Db-$kVUky^Uob2)?O%
zZT>-libnO<bL;-=b;6EVIfto_-rQ;I6HYYA)WxhOlM)(2JVznSzQ=OT!+$V5v+zTQ
z?>&P-G~x)-=ybR7yw?dt`#imxkd=2^daIjMRUt=Ud)K~!r*~U>%9~vDVcHJCH!45l
zIGQ0n0~?V@)#o?K{H{eiMGWbeGsy2*Jxo5*s?aZw_v`a9oh8{1CJI5t+uk{A7Q3b&
z>HF=wd00FBV5P9f=fiB&y?YO~6rZ#4y>Usr_jdoY*&{^ixjvhS{)f(2!x?2N>DR>9
zi}E&gZ#b%Qs_mH7{aE&o&lGIF$XMidtt{SDx@)o5nBwwn#Cr8md8A|)*}cr6NPp}S
z21)8d6hc(YMq0^&(vFL6$*7lKOsCkrM`iB1?egp{fi`tjMxaRD#pSZS8DYWe6Rc`W
zt=(Ko-9&cOHllZkXpCMFV<r@S9?<L;kPv^lcSL#9a9=Cv&6#(h-en}g-i!yC`x4>`
zXGMps@Oc`Gu9-Lz;Wu`!xoY2fZ_>U~5>FjtGVh=nMI3(d3$qHr<;y%`?w^m=8z)?}
zf&#C`<&E%G+a%DTm<&=h%4&s;GM)+8ti}1!Yst~p;oBLd&D=F;gT^X3CrM7{uYvlO
z!W&(I<?H^EiU_OH*tSClQ(nZbu9X<O-Sw#Ld*3c|9aMBVrkut2!nqMIeM(bYk4rz;
zqB!8{WtT6?6Kq-mUJu`&qCC$@zlc4B&&g2sLb9T+j>*SQu->gdz#u;)@4*Aff>*I&
zDywwu+h4t`Bnk`PSzEopv<RfAmKv3cSpKdta$c|p@6x&TI8u)%Ydvkk(KnuJ;t*wN
z;yVdxYb-`#-6*uvYrx_dVhN*rX4sN##UU#Ez@I*I{6!RZxN`Ry>+<?Xqv2*t?1knp
zHe#~)0ygr?>G+qZWVe`J$-POr<k=q+y=*eSE2}Z~W$5Ch>=db=TkJL($EwZ~Y?H-u
z+Q=6OD0;N@xd?}xjdfdE_C{Vpr%`m<#sIf$*4zi99V4~+Chydb4|Q(7wwIj#@X*`z
z!kUL&>>)>iPy=fJ<(po6bftRpQ%=g9GXh7OB9*UN>Q~xU>e4c~KeMK<KJk}ND!u+W
zFVTAAW|iz^6v0aR9KuPa22x8SGmq<G&DYWE%IC3|3{IVK`r2_{l3?y`9d2S7uK2ZG
zpFC5CuPL;ZUJ|SKRVgs2@?~*H1A7X~^>#KzbZ#H^-*4!-9KdDe9V+a4Jvv;h;cTk)
z<C)Z|S5hMKxfCTYvua&d$z#r}^)x(ft{_({*&I=Mr4_9sZ@nT#Toa^u{>cX;2@6x{
zF}dJ-WauoBbghrkhdwM{B{h{Q>XdZlpL@vj@yeC=TKq4>^GPY@I_>Ya*sGgLGG7yY
z+@j>(`{e^y3*-3-pG*0Il6XEIG}RV!FT1`8M&7>Dg64s4s$%MBTBR=*N!L*5KXoTd
zW~%F4vM?8!zg#c@{g^K+AIZ|a(dykvL)?@|MPdG|pjJ%k2#i^tG*n?yB8TtzzDcY-
z!$GIFG=pS<+l!e-l-s1}zEFD$PFZMYbBAOvgt1%7O}h|+EvW}j<--#jreS4C-M1Z2
z&OC7)oRp(neI9G!^y11^EO$t9m8|A6bPTrV5+5u+L}IGu^oG4AQy_UQKabkh%%){b
z-4@_1wwhw0)?By0p7f~tJEJ^BPz|w=_=LW?>#~>Y*GU<Ji03zJNfrk_*yrL8aVAg=
z87SzskIzk=S(1(QwIh8rAy6mWi1Xliy}w?44w0f_mwESL-%4~?p2nkv?G=97gE!n$
zZ%5Xu-$d3oi1$$h1g>Og#-fOA)=~PVHcw%MS)n~ZFnUJU6JK^5ph9;M>!;2XCdFm>
z%D1$qG+XU?;cZQOqjW7c-YWwt`SG+R(q;7Uc&S7I+YYTObbkKI{||X@0aaz&t&OTE
zC?Fsb(kLJ;(j_QJcXxM7BMl-d-Q5k+t)vK|lyrx*(hbsQqHlcr-|yM`+vkjboN>nQ
zSVNa9p7lJ>UGu)?bzRe$E1KKrd#D>>bwZiIl-)}X_=iwhhR#Bjb5*=tO#|J}l|^MH
z`*sBi>R`8}QI5X4kH=0Tdu>0<j`lY%4%>eWgE;;c^#2hCu`{y&j)F-48{vJBl4Qs@
zDF!&jSzDm%;_3Vhy(doKBwtxA_P&xw(M(#l^8CE-ve84S#5LkKma!q96WB2_&ZshY
zrE@xV?H#oQlQ7a1W4<AqH1#mtiAT-7=Y@tgu<*8lB&gK8)$bL1D<^)~)NLLcDB)*t
z8RcdydT+B$NrrAW3r_IN_dCI>(sp0_F5!F68kkG^379|e2>qTCQyih&v+ed?qFP52
zZHyqtLo;VsRQQWSePd4rN2Ao$bSW2)zS;hmeD>x3Z%gXc>fgwR`ND}_Swx9Q(3o~I
zzT>&4DZ;ftClKA`au3Z<BCx_#`%AM-!}Cv38(l*rI2~L!W22sCgke+?Qu+{VSgjQV
zW9gVnX2mwe#FO@^i1B1RNnd{cme|NeuCm0-a~<c}rl0bQAE_TdkK$k}cv>(n4+Z`_
znoufT`5a?{@-70Sg>jktp(_2e!?(R0E_$o9ufo|67{>e#x*AefOlsUU>7D1)O;#Vo
z8@-P%A>C@x`xG3gU_s~ldW^PKFg%=#MowabeyO#_KI<T`>Y?WT_iH5OR+n4xvzg}J
zc2B7Au9dPKrLFNcm}YJ_un^NLG>{{I#fYVU7}$!oT3WU*Gst1_P1OpOSfkyrvHNNE
z@kFo_X(h@F*Gb>W1-*U*Cx=z%=<F(C*{YG+By0w!{Cm}7=o67g-v)`o(W<JMvK6&o
zj2FAl-C0pO4!>*RTEv9g5-(+boULVfpZqlTI7O>^XfHaKP<@@QPSN}6OhFl$1>Vs9
z`{UK*n;D)wxMRcgBpfsemgLPHtD96R)B9J=nKw9_)X<hhhX(z10=6%syfgE)-YVFg
zcEPu*Z0}9nog16vy%{5&Ki<!&66Ju#_H*mw7N=Q&u6!f<W7O%-i<O>|X{A^+Q%d!_
z3}El8y=b?DJCo!g*V9{EL&9mMMiS)~{pBFXy0LL0>Er8jg7ZtX1QTQbI5ZyqJv0<9
zye<AWdR;tZ;`lx2{+l7UHi@L5Fx>Fj#zN-_78b^3Sq1WYR4<OOnqcB8GD&BTHc1ze
z(WgwmRUbR@Nyj4gs=c3<I!p9Okur-vK6h#Ai~4>6a%7dMbDL=6u4aBKJ<d)y&y|7#
z<^<*l-SYEeyZyM?2+IEswfSG2Fp#?e+J;P>u1w<wR1`X_#f_hs(BMkznO|7wz2UnE
z8)Dqs(i%e}j;HMu{=rm-b0LobK_v2tpUR@S_%sMM`rvZ_BiI@VKCXKNDTz2vi|Mn&
zh|2O$f@gNi?=bv>r%BVuZ8p0&-jUjmBxy&R7}wU^M+V*lrWlkL3r#2E5TO!<+^0Vp
z{7)tvhzwXnkY|MNU?5R*!5jRdzyOf5hc{k^LzX`C#|78vg0=ALwO^!*U!VI$f`IQX
z8$l(+Nq-&i5`Y>zJ%uJl03t?`U*s+^gE%e}>K8?<D4gX1gs-Zo$h`O=n^2R_O<>TF
zjOL~<DlPpx-+Do+_~UzU8W(5#+fakzSDqcH*r6Ba^GjVQa&eBcq|UV$biCN;N;mxk
zXXh2b@8?K=e_DVH%>&>ViYWrU-L*pY@;KI|Ex4OF&y|3=N0=P}z+GduM@MHTm>X<#
zgj`^1&o@(Dn5n1$Tor^9xwtP$yxb%ATTEov6Gl7Z51faZfMNxlhokYFUc(=Qiy-!g
zlD^Lm8NuS<Szo4~iyd%KfFAYxYT&+YR&qUu8UT|xaoKu$dcb_mod^sG`Br8au=>Sp
z9nc5<LD1^}$12}72C{X4`*;8%2HScweSZFr0K;W!deE?g{L9k90%wtuo&5mh<&RG`
zeI6-*Ru;Z7?*(~aPzYU6(9uFkL_0ufBb~^R_6Yc|!vGEu5fc|ZpfU8uBO-D?J#tuT
zg<4nI^)A1~h@sQd+SW4m9j@_k&f*z#s8Dwra+jfri55Wa2m5@}6TodDgLjdGkphW6
zV0_deJs>!H{J0S~`2gIA)^2A_Krv3xk8Pv0H>J33x0RzI>s}Pxd6WXEHOQN?GBY2o
z=m1UyOejO)WFf%wp?D$%JEV1^tRjUDpZ+{P?kI;c4qlI{K@tSb`>Tb|3w;0;xi4LA
z+JD~^>n3E=39+yg04`T8rWW@10j5v;82~oGNi^I$0EigW0dFtCNqYeL&D0Af`H*gz
zlh(prgKY1}!K#DoV3YTuEIy!K50Yun@bU24m}Qf#0Xv&-`<>IrBs@tN7E}}f;{sX~
z9z)+;4i4LSk-bY{>(E3OiNS07`8l@&TcIEsRA_QmQPpf{X@b7=<}K8sf%0Wxl^05J
zt=Ux29YZ~`!86=5tEoC_C>NWMj|2ssXy5$K*Te4|Ux*$Rp~;lAmQj3Xb*!yRT)ghs
z0|^e$Qm9-{r%BoO7{F|}RS|d_6lg>F$_$X?9X*G982{6!+OyLD6Bakvus|78!pJQK
zG^mN4i_}3Bg?sOJ*8c|DnIsbQgS#E|AR&-U;;IBiPP?JBjZCy_JGbstI$>d9)h)kp
zy`ZBaDb8)^G-Z&9xS&DN20<D?Moe8Pc$v`T*|XxhRVovZqR^mSppj4y!F`BtDHkZ$
zko%U|C%bpAuC6X}CH?nsxr%h&;KQ$#l>>!r13MKU=K*vu>1}N%bI1>fh`gCiE$9HL
zem^+Sybm}EI{^}AI}xxkFMwN)GsqbGXF^T_dbQPXQ4)Vbrw51;!Vm;XAy)&k@*AFj
ziG?FYa?_ys#{&8Itx{op8I1x_<lI6m^mf`9Eg&%m{vz{=^e%b(06zafb<Q(zJlKUf
z!7pRH9Bm3<<7g>^_b~|xv@ddzzF;ztkv+9}2wM34gN<*+bVmcAk39bpMpjmYRWcx?
zO5g6k4T`Eh{hE48^rE^&D6{Jel}jHd3-JgM3miTMY0YvKircoQ?6(NA{yf`QwOiYr
zzZ*dGmaUA?aU^>z+=@rXcWWP$d}_M48}plnN{;VTVqyrDs351@{KqtBv#H2Q#ZKQ5
z<sOlsV#MM3hvqfi2B{JJWKzEtG884A8&z3(G;}xc=AH-vUCk1}e5lsS4FWy5R;@-4
z9v})FRC2Zf#U0r#X$+E3XQvyU8>8jI2P*@SG{m7PFk(S%NJ~L+?H3JB@U>63gY{2g
zqNx#)^+Uc6V<N1N(4cko^`QTGkSm$OJK5C4HwLuWf$jt9{ur17wndTJp!yp?t&ru%
zCnThsf)|4z$$ThONqP7Uh329_850**4e(%>bLfB3n_=0=WkrILV{_~4MaPp2AnV=!
z2ARCW<Kz4wkZj@LMY*e90(4dRK>jsT`HfWm<ZU4to1XS|u|ruYsf#)=*8?t;IHBpy
zd^fqMYPyoR4@Qka6EO&WIzadneLtEmh_Ya}1?NIcObnPi?)Lof5|pZ=JA!#vpc^<B
zfY)MM%Zx49AfLwkURU6>eku{NFf-F`RwkMOM}j{8Bjq8<E;Acj1Ek+qpN&^Z)<q;i
zX0QU(Dv;R?hVFa|4|>@IjK@!WOX)DdLD4_Ja{Z&0V<Lak<}NOl8VRW7AQ{`Apsa(G
zAQTk83n}*kurUyu11viV>2TTy9oEEDmZ4nPr-YH8^dS+4g#-pI<pA`*dGiMPGsvgc
z3Qw|r(~NJR;VfU&`vQZEl;fAce<&y@fP#0LP)uqsDIFw(?aEp(vt1MflHP7~?dt87
zB@7G-0;{2uh;vU(1#1unui7=X6Z~4oq1m5JHu1RbgV;gMS#v}5j6IY3{AzeOXfc7<
z4?@UXZPhB^Cx8$~#Su~gAWx$<Zvz<tNOmAG1O_Q^*UG(o!ht{yk}RA*I}^UWOc_*|
zQtg4#Z*R<jqP_yT@1qP*`H+!i+#(aor7t@&pLkblD`*b@{LIWuda}ty&UCQ-3L41O
zp+1~Y{TqL9&3Zgd3MVOcUpiSM=T$$YrFwl_*J$ruF0P~Z-IDgO7EFNlW1#!evRS%E
zWhEx39FAYvONDQz;jytCAd}fM($Y#^XV;-^-WwHe{Z)Rq7w?@<TU#5-HwzG#Y>9F4
z5gUK-x(t~{v!Q}GFs1_d9)wyI`+pBn8T9b@c(v)3;XnJZ2?v_lMb%Dt{8K0bO!Pk@
zn~O;4BCO&NnezAX_VxzHkS&@QTf4h-{?g;X-xwQ<{~gDq-$F8Uw#~h6bbfXU!o1i1
zMgyMg{=eVb9wMSk2lU^4NBf#Q*3-X2vi}@@U4&V%WL|`Gzr!WCb$^^AjF(@FDThe!
zZ{=+t`d?2m{|M|ZzVkb<yST33f!#%j^(!cZ5Uc5rBVSBx7XuIuc=7lh6~b$O1dYF6
z{(Tbv^ik@Ikdx4$;_sUSUHkWo|8Cf~JpmLsu+-3pmIES~q?ni-zF}1gr(HG+sD#?*
z!Hep5SeU8>1-=V`)dE_N0DpgEboBFu812Y?Sm!{BZ3Ynna9e<K0A1sElgThJTa1)I
zt_jNoGDldGPz8?c7_47d7#ZK&E?lrDe+QLExa{oJBO@cA+-lRahE@~Phq7A0DI3s-
zg{>T>3Ao7LaqAufEDsPF0I3FM=c=+YwFbBMEbSi9@dx!Gv4)L@cQ{cCQ$q}uY<hsy
z%K?kipn|Y|5WI@&LN#-oBpDN{Zw*ONQ5Ru8#AoMLyVH>=o;_b3XSD4oeFIZ}Oj&`m
zlvb1vNNHg&p}Y^}A)yo-E%$|Ltv@@+C9t@-_~eJ+3ZK6I95z=N-?512&U6#P(z$SE
z$(I`l@B{N#wanr?o;ypW@bWyiH)+lOG&WjPpv6I9&mwr>xC2Qs<>#9A5RtZffDZxP
z4MSlLJ0~aQ5F1-tU|8H97JUDpodUloz*ZAtfTNg_k`f>qQXPZ$5eQ4NRzZ^y8y81G
zPX5IM>Ujatgzm*(Z9dSLnU&SnR}9ny;A2!&S3icjfPK6`H~=2M<H}H1YwMyQuHc^=
zlq8SNH4rO7fwnK;0b_YlQ4A{LfpOe<`0-y`m#1idZgHTOlK<LBTnvSO-0h(C-GDLj
zyO007qvv9T|Jn=vhhF~=BL((KzsDd9)<1^~lmP!@48nN*bG-d^3cuElKfZG@ar`?U
z{bLLCPxH=SxB1^4TA=y(-+wqyr~hRG`g^DPU+g+>TnxKEZbp*VUr+yPi2q_!`X6S+
zf1Qy2Zis((287}N?>9}V6%M@`i=ciSmjf88(7<;uX6GjM&@v@#$*JQy`&!wt(Wz30
zivjf4n;1LD*g-}bY-*ADNnIaT5H6<5KNJSPXYyY??q8Gxe^?K|=l)Xb{92gc(-)6F
z+#KMt|L^T3|MkGX5C3oft$nD&4HaFjIv^5+Fef@R6x!zZFXw(kgY0)72+o?DpIY8P
zM|ZNe{sBY>AXUJ2^e@>{ZZVn|#nI6b8W4aSE<8T$xxtcQl>{ivFm=EN_>aTA@o`z2
z2KvD(bP{6FU2?s1JG7mw)P=|oEbRXNQBSnw$DUK5{ET+3LGup;gDUwD)1iV<(AH?p
zr<Rs(xg}i)VF$ilf5GI$L+(pKM&=_$#R5KpzhM+$EG4rq1_Vk5RF-S1)?17HFe!A1
zL?oj7zUN(#rq#c3d@&a~2r2~154qH!f$vBb1ar%};Hn0sJ?C{Y5V6G2scy?BLB-ym
z4QE2poJ`H%D18@TyKc{)2EL3@$mt|n$sq22f7eHT_&hghZPwpe+mlnyoG42WiATLy
zvv!~Lw1Izy*2Wt4%1~YrVY?qVUm)!K@)fuQ5XJ(Zuz+}ReZ9e`>(4n?19zc@YFKb@
zwdM!x_i=cG=oV;{i|DwWot**DfDM3@xVZd9*a~dqzg39Do*ps>HMyC<Qi5$XL_X#~
za@NtA4p{g7t#`MPc}(~Sf-E3!IZ&R!P9JuUg2KX$^GtMqnqD+0-~6xp7KaGs?|t<j
z1M#9u|4VQEFMH^(kN*C`|Epo}uS4ShoabK#{?c^(mv{LeCKwE)rN0IFf6Xla=289{
zvi~}>{4I9+hbimt17A$g|EE1<w230c>VQKE(Hly|_l1QEGmue7LIbnkFu?!;((&$+
zzL62g+bQhqc0dW6zXrUds?yR6V?dDaX=o5$EFl+<R4uM|09=A6)hJjNRg{#BT!4(<
z=jX4#czk|RzWTYf6%Ml?n9u*`h;rFEx`3(@IqeLLn_t}oV*#;@fxOtC=@d|G^7r67
zwSGuV-8M0C(&Bf!j06H!Fn-8?yuH7F!93TV*8$i!!XZ2^&hL+Dqr`;sIZ&5C>L3Z_
za^G`84S5K;^B?aW9E7}i^L(8Mv=Ff0*HWa^cHdKsKD!TMYYGaA0IX=he0l)a1`NP<
zN(2I4n3|H9U3%6WzL^_yH=fjP_BK}z$u44BR4nRnb^{((>>=6sOiOEuRzQYWM#&A-
z-DcuIcLK%dKmxEzRjMrTh6Pp8%ZLqlzPN<$rJfZc8S+uigB(h_ni%vL#-^s!b`7>(
zd3j5KLjUby;fV%D1YoRi4zxzKIFdnbuz3Z7)_u@$A|k@ZcUKVXY<`anl}sWa_!+Y&
z<qkh)&~5;0DJNiafkVV(o%8KKyL0Hqe-MuT*|z_W)_40Wf4262@{0cNhTI>M>HlC3
z{(E-#*Kz&-!SkQUL?`Lm8*BIhalY7p-hgR>k*{=qB@D(&4aq6lY-IZKe_mk|dkMvZ
z<OLH&KKa>`Z$^5r{)|lif3X1M9-y6p5`%T+Ij6a4d9hlCv*hPZG~TgM{%_|bwc869
za8T%U>KP}%9GSDcDR+X#i_;35UQ13n5VM(;IaW<~zZy~fweU9a@i>%B-m#S2@bHNr
za3o@*7UEXl?D4y3!oMe804g^Dt$xiC4bqi``FWGKF~F5psT<VESc+k(INY_IRr^zy
zer`mgoUdABX4w&+JT@xPNe{y{Cr=9m4xaQOt?Qbv&5i9G^S=#b{b?e2)}^9cWVG9b
zC)tH*NZ_DVX%?03vQ*y9i}HNKT{7)<&szsOT*^ON?oVXlUswPC`SYiCn-^qGy7or-
zAvy5`->f)O#eCcNF^kSFLbZKclg9$w37UfAa@mF~d)2h6?c}LZkAL+lom;lnFnKL#
z)CV=xXQf$AT+Z2E^!Y4gW@cAt!Z^KC>eHNFQO;VpWH4YDUUlXeO#FK=2!7>5IWQV~
z`?*+}Q}N;OKnq0z#~l_933=(fxPS=RgSfGyamB5_3<&<XqNeYQAYOpX32kL$Ny$Sy
zWfDR<e>9h2YnCIO$lexx{IOLHgp-qt_QpWcjs#>rVT3b*b@^>lSEkq`rj!<k3Kr3B
zfkyA=&&8iUH&>Pq7R}z79{*)mWV0z4$7UICvywK3e?rprq3PaoYD|XZhSz$$ix|CG
z?O$=(A8W_&736>a-wkiUT1T>11@L9@26h#7pT?E>*(i-ben&GCf|7weH#|bR6W6MX
zKtQJXz4Z(~@V24S{2(AT!11leM&fqZFjUAXlwFa<AG0vS+Cc$3_Gh@uvh`+9zQL}F
z%YH?BJev^h>;@7oTv|g}*9aDpwk7zsf+#8KBDX}iIybj7M7WAuSQxh0E9Ha6W}laq
zr#icHOaAwr4yf<}&iVym9}vWQNqKpu{QQ5-x~bkAU?%+$_IJ7L;EIhwK=8L?7b%)l
zJWC!Bcb6r>1sNdR@5Ii44}sTg)PSF_ui&6ER)GJ0*3=&L!eUNR;XBK&*<c6Oy+|>o
zJmvBl@eH9Z=vxIFYqL7Vc{$_>;tTEDEi8MHW{P=g-<rkcrP~T$mFK8yO7c*kD?1(F
ze;vb@+DlLw<1jHJ4(-ZK)XY=ss@N-dJCLW4CpS180WO!r%hRy7xA5*6fmi_ybF}O?
zn|zSw>}`RDZgAQKD_3_SXM_ZD2b3)T@si&Tyjn6d$&b8e1Q<?zL5oOIJlF)e82_sF
zGe}w&fLF@e*xp6Ln=x>Hetu+ZtVp$7XdhPd)%kgH{62bk3zZUV)vIBV2mPSE{Y1Gj
zIuT^RTYB!9f7spK1rVzr`_GIyDkb2{7v9{kg&d5A^&I4xA$J({K#5l)49s#5*ph%T
zNOxx^%ZK2x5?!D5&*I=Nh0EIoTCjnAI0FcWE_^H@UGu7+lgS8Yc^<?<%OQYA-)}Yo
zGgQbLn9TtX+JIEFcK0n`&^<hUGX~44I#p9}(d%2Qhn#Kx=VI`Naxn$t8Mz$9L0L#2
z$mzkxF6uoP8OWNt0hJvRzL`%i>^uND+na{q1&r@>L5A3Px=$d_b%=5U8JPyKbqJ86
z62Pef<O3HXGq6yiqo*H~3knP@P^(;qwAnE1&mb=1ar@bm>S78#qmKY!VOjOeend3F
zJK-t7+NTY=@D&F~pMg10BP@7<hd06!Nz>P%A+~u_kMmQfEKQ?ANT{|=FHP1~gX<-e
zQ`_{`=4Pljd9x^<!Yk%bLL#CIv&Y@=&K<FgKoXD?RUQnKLub5h8gmB0nmFW~ZhqHv
z1H(GdE9~v<C4rA41f2$0)N2rg!p&HRdk7{ipq9vGfpjK7)oLw~%=uv9xRva={q6pP
z2StE=f~4=Q57Q>t3-K6wg2a>w-;k1lp&N2Vi(H17`C9cyV26A&C&plZe|I<1Zu<(F
z7C41}2m>q><UPv!=ZD>%%O2uBQ{3P~rkJ_~O$kUmfTB@l3^~%E>s5R}_XR^vDhteF
z)WEW=nX?cR2^w21RhDqzl(f8tG+h4rf}9UlYU0CO$}E=)xe>cbPwI!<P<eZK`F`ly
z?H)=>N}#Tb?5C%$07Vt`XNd&1Lbz&))%k8pZf;1XjKQuyn%QoUkFz7ZG;=kD+o=d_
z&E<T6Q$elxHga3AApAn05dPVk%WQR|RL&d(uHu)Gvij!(HJy6D!%WhQ{+Q|*?s!CH
z+6)I3EiR|x8-M>Ax5Ki&=R1p%kn@0Wd?7m!DF0>5dM$6jUMjcT9nj~fwyr0H7qv-@
z!LAl;OEr<e&eec(2aH?AbEtzr6bHbwEPbeA=0xuk7tycC_sGZ~wL?pkrL{Z(as83a
zGrz<lAo(jh8dy>RhdU!tACOkY>WIfA3yK7zy$&U?V_#l~2eDV&XVB@~?+_v&AlM6r
znc(>6T*%GL7#KwWB|H%Scm=K9)EG?GA<=mu3*r{bptCb5%~*xbg9}Dxa0g!Xh=VwK
z1++p`l&IU<uoJn^I<{vTfYBi_nb?K>u;bZQi079uvszU8=J(OqV0QQ&0me~fW#vs&
zRD}zVO{!}Y?)Qw-0M}>_*ek<~+)H&4<|~0Y0I^T}>)|29JQpsa&HBK-=(}_KHd%nc
zUM%>X(qw!AROkI7X;_<BHeItD0>UB6D45uSG0?n4o!5oHvpqNB0!mI56BUJS5VGDU
zlzj@~XFPs}Hq3YMq#a2dg(G(nMkplL!A;A56|DMB;0MKO>C&?_H3bNyK(-C0n3=m0
zewY*k%9W6In#dBz9fd{r;&4VDs@w!>3@};oNnslFDyN<XFf_<~#hIV@0K>yFMzEqR
zp6#Uw5$xn7N43k1f<U;tN~c*f(BCi5vzQUWD`!*)&LVFadM;#Q*#*O=f(8Q5z}XE0
z2??qA<ck0~89fJw%9!ELy!+tG6dN1+$m2xCOeuR8mLhD9+x^F1{1?IOOV%Y1)-fiV
zwy&V;MJ9xD4|h@SjhX_|epH06dV49kn?TCoZiWVrh_rN!sHpqFTq>VWBqI(!{utyd
zvlu~?{3=>P8*xMya-r_~4*OtKH}F<*Mf5Bm6I8egIlJIq5BAnmPh48NyBQ!YJ}NCC
zL8nv@n^-E&*LF|sG_^V;B*eBAs<I@2@{1-2a#pW)^sXKrzX!AAw45+mcCb8LV2y{P
zfvkoqKNd<C(2$Zo;hdVBbO+Iw^fp6uJDt<?hs#fYWPwVoG6p*KC)QC&%=Olfy@$(%
zd@eNysGC`yO=(aMLdJ1V@nL<4)oWBDE`_g1k?!E;xl~Pgcy1dfaSXHp_we)fJk+}j
zvr(utoGI*~S}vQ@-dsT))X}I3;7B$(6_L`8_o4>N*1<uM_iaR27})Gw!*fOKlLgq$
zbh4(hjO6xh`En!4tOLlXW==#diA%|VQ7g?cm>3@H?<+mvMkQb?02@4sjd=zE0L@U@
zHcnt!fQj)4!c4JaxD}b0hJk}cn(7_~UiF!-n)Z-%g1nEd{}k*SnGD;(TVai<y5oiq
z6{#kTxP@Uf^GjgKQ1WX5%m<oLAX#{Vh9}D%1~`+2fz`q8t_8TTx&-Y)&k4Vn+Kt$n
zoeAb)BMs6r-jmH{cm{H=3kxPOSI8!*SD4(br_4zOWu-V|#KtaEWv68Bw509QH*<!{
z%E}<2KYvmOX3k6~qD$0kva+(!@R34Gt)LU83Id&fMDJjQsr-`4;S#iJY+sx98;Ri>
zl1WgGO#%W5`4~AlIeT<)z3qel8W=HmZlu=2ti{j&qU?Ng<|!v9giD7|Ajb`G*DJIc
z5Hu8l)GL@h@#{sC0t9OK9OlobIK4HG!_UHk!yP9>8rv}Z0|Fu(eQNH}@Q@l`ms5oZ
zZ}9XJ7))o)W*%fp+uGXNZ`f|l3Yx|UhNfXB?fUGtL82X!w`t=ZU<tky3ZN*CipfPC
zk)p_C1K$5J7Z0RRScNAHLV1iN|3qr!_|LK@MpmYOkMFDal?UzPL)2^2^*Ni@tWR;d
zqhu(aVEJ(p==u5`7psKc<Eneq5!Uc>`o`55Zx@pBhYzx@(Bgzxo1jnEr9Co4wns0~
zF(~ZOvL3&k_jPvENFd8VZz@mFYQDm#q1ya4QC9pXoco#WI<vX+k7K#o<BpOn4`o)I
zJ`3KwVm=e<DZ7Jbdo3wcP&t$=f`)|c%ln(p9IIFZ{^=??nf|73{w3~W`r|VH8;a$i
zT9*|eJ8Hu*_2RXHH%GWScPfW%yrwjMD(%YLk#rI9zt6L~CoY<q+1t6g+FBHGXC#Rz
zNMP&PmypPXUET2hS`^XY@y`M52W%3lz7A8hOIxQ`e+ta9i!i(>kBAU6@eARVY}69X
zON?B^c`PKBzcshz^dY~#Vc`BW>IcfyVo%qTxHsHsFBm<7-c@Sdc^1pj7ol{Q@1EKE
z*3>O(%ljGzP5G>1_l0KeDX?u=8&Y9a3JukLmaUaaNj3^wt$eY|ohj=icx7r^BpDaQ
z5=Ed5tNm3gT~OR{&Bti$MVbB1d^CZX?35@@F~a+-jfHN<QcDfNQ@a|5!Gs}K)eIH0
z){z%F(=u{8V-k85lT6Xc)oZ30k?d3Q1{G^e8N{FT85Djgqnx&=BNtaLoQ*)ROBC83
zE6A`o@J;f;)lzmUh|e1%58#jH?B!b?O@CHgY1r|CgdRmzKdj!%!Vr;@u)DLo>~uwo
z;HZlQt&6#j+4*sXd>&3fpU3CL(3VkgtCics6)_Yu!<a@>GV3nSM#HJ>?R6YtN@j2R
z;*XQ{-FCehJoO<9-LP9^$m94O%RAJMrjL2tYkoe`op)hh*mSt-8{8yyF6FG^Rd!!W
zj^k3GzMMdiXFqP@&&3V&4NsRNqN@*%DJ7#KZng8H_UG8HJmkmIV0X_vCbf8j<BTC|
z_S!}m`8Mmjni8}}lDI<7m&aaN?dszPeY~2XP_5$A(=^vo!^ht8h}_>H^5A#`ok`g;
zfKJ|!Y?6Yef^BlvMJvjxKd13~-n-@QrP)*attV=gQim)QU-&t=PSpEpt|MqYEgt2<
z>b=v7G$gkB<EDB>Qk{d$4cx?DDqRfX{(xkfB^u@sd)pn1K0V||EyFf$Ptkn}IOLKB
zkOM?N-)B>>l_g!nxmo>9I>lrXNn@|57KbY)wc_EBWK65-jHW~d2Y0a{PyE+%&Xuky
zF8>rmN~yzB0yc~$dXntmtb5~zx7TM3d@o%$@4A7kTcw1RwwWa)SMVlix0~q>*%Pa5
z&p92O-F;o(r#hH(BFbZP`H!x9+4T0hsp;)9rXwSezO%khi%hSZNuq4?fEt<c{;3?{
zAg8zuPDkB#;Dd0INp`Wc4~ZR?1NU-$@z|NvB<kc_N78fe>b+Ai-cF0UEk#e4i_IBQ
zIQFC5ov=hZ=BDYscV$tHgoG3QZIZc86Y58cwL9~SV><DV6pMwaRd@M4i*U1h^mN{z
zYan3TQ8Lsnrbv!9?;?&HT?!-fR=OgA@P<bQ{c78KfZYLx2In~KeN2t5tM77IWLng(
zShtT~X1f*CkNTFyh&eDHq0yDWzsamK);F>nm613MWZ5w~*XTIzs)s!=%C>aO%5hj%
zF^T$4_~|y*E817{4y?F%<Yc&v2&~Km^`@zhRyy|TDkdjAHs7-_b;*@RVq@X^;mc7|
zv|7@l_j_oFq@H7SA#%hIQX$zYk&P5vwLhzmXvWYVoff7c47{^HY-^%8C0Kf;l4X2W
zOLzZ3Zo@8RSEOkVuOXpMrpA)m-1M7R>?u)(1@+<`doxnQn_R4S6jJ-O<sVDlt19Ey
zlbOu7w59ji`i2+M@v`c=x^#gh>V2hri$Y7(Aw0QOrSUs~D?f=3q33p*C2(=&TaL{t
zm2DzUV(e@t4YGYbR2aXSW%-uS{lf<rElVG3KD|++S1$}5c99A#G5DU7?_gF2q4AmI
zNflT=XVZV^p}LpnAXz;*)-ih3b+AT=jQmOu<q+$mgs3NjD7q*w$9|A55w!`{RK1@o
zWjYd{yNX)lv1=a5Y}>mlJ3u_Il=rRUwt{R>xIQB(O9M)?hbE`2zI_+-0byevy=n&A
zY#A@^ry;ElbECDIIm^<q&<VfIr1xJ#He)7QY6QrpTU1p;E=y6rAF~^7zm8|7US~fb
zUUhW<Q;WQEmO1u@rZl}Z0dGF@xte3W^m6W8zE5!GaqBti*FA1DM2*p~?<&g&(a!O-
zaxXgFrR#hMjmg)Yt9K@13~u*5XCRD!*5juI6%PB2qNw?D-}OA4xYejq;{K(p3hSGZ
z(~r~?Tp~GNWlCm$5p*#oMfne?7&#1{5*rM}Lg{X3<OacNOC9a$(W74&?1iu>+V3Vc
z2M)R-r!Mv!%O6hL=N3Eqt`!Cp>GVhp?S$V$q<lU4JoVj7lSyq0rKjP(F`4^_os^e6
znIVO;56*L+=a*VX&A-cUJKl~>X-<)g&F$4{4u0xoR$jPimUie)6@-8^+jU>0O*^i7
zRHT@@svIRZYRHmhiAA`uVm$HXY4uz_M~T)?<QodB{I`73^=sX%Q=eGtM7Q4^9J`Di
z5T1YfG#}|O%MGpnx?jC{SLWo?>^fQ-3Cg_gF>Uo*T%?tliK5)i5g!HF9&YD8y{y~T
zh8gr|QM_OE{dYS*PKL{yXNjYmw}wx=HeULNQP{3=3_Zwt`#?OL!kj8xhJC55V#!6=
ze03-5T7h3l%S!KXdoPjb<(;RCQZIr&$aAU(9J%_DoUBrCV~u3#)uks}ufNjE?xJe>
zIkxs_D$shg&Te+wsM^Py|7L(KH>I8alq1m!!J^6D%Zf*w*B(d4p>T3O94H^XSDzIv
z+bgBg#AryIG7(ctPdDfg?5EdF%}QK3@YzJciiw!Eu6U+2PJwWp&Z*`!(2hgXhne{M
z%?#g+1s}><j*jB*U2a8R&n3eQ>pdA2kPrj(R7#pC1sVi=ZoB5YK2rwgvg@3yN<<ke
z@?E6cotIsvI}bAyWzUf|Z@%e^rWup<W67D(!M59DkvUf*Oho%OQ2n$?hXw!VHNMTn
zX%xL_&KaWXk5cM<y;8hs?h=vOTh0y(M|q8?FOl3-%Ff?DN@!}oW2gQ7SSTv*n#r2^
z3o%Jy{Nk(RxA2yJqy&($Mn-Vh4n-)W6~!Y%>C2kvj}i5+w;jmN*0XL{-yz@2Bx^kw
zqTd?8swGl?etuc9E=A!VVjs4L?Eg9T;b7$aw}kCo>f#}U?{7976z4vESe9Gu)kNCA
z7I)2d=VLYLYu{VkNZenn&<>B&$d^C8UH)t_Ug~Y*lt=EL`6VPXGrA*2@IfPCFbVpG
zR;6m)b(}giUJtEH+9D?;<ZN8E&o9le(2+FKd%8x-ah`C`ri96vpX~hPna*){>A@&L
znkM+ben}dA>QdFT&j9kVtoELN_xe(km(Bg4P1BbCr^_fgIgOqtD}0zfasn|8+307x
z+gqE*-sLL=n5ogy-_Y_VO)$7}1!p=;VjiEBy79VA3ydpZzK%RE&^Y^9RuZi(sx6z|
zkZCt0b#%o~jOEJ{4E!4r4V|vk^z9ZhMs^<`4+OsuZFMTX%tnpx7hj#2{t5e`NN@TO
zs@Y)nW65p~pQzo#<fE)t`lizEK@4ktWQRw)(*}L6YW*x?JTV*Zc;ke1!*}-@3nKHb
zMadm1;=EKVMJX?-FBeQI8ptZX^-99P4!f7?ov6w)-`+NI5ABj~4C3p7a}KDEha2Gu
z*Gx`c57aReQ(w2#v=*IiuJ!ahJc*HR^!}iGE~?pEBqQ@i_fUzr@D~0}b$ack(Hq~B
z98uo9iDyAUjthKJ>0LO}Z~M?BDb@0hT4VMQmBd4_n?B{N#{;EO4_Oq6F+UBdeLv=6
zQz|;bQq67^auHK7v2bbGa!_!Q@SL+-^UpYYPsq#g8a-2SC2?nj;;L_S;Qp!)qq^EP
z-4A4)sPTo8V>fPR>?zXrX5g(>1aDUfQCL&=UI|40NXEOA)iq>LMDH_EdBao@xiZgV
zd!LBsp=xQ?D3uFei;WW1n>GRO3UwBsp&tp&HB9Gfzv|}yemN4SwPvA-dZ4VWf=O+(
z#BY6|e!pm4c~Z(4H^1D16PL`Z-(3aK(O&~yvjkn47Qq|KW*XVhT4G78<g(^eUiQl*
z4B<)3UDH#wH#=9~UVHB-yh4T5io>SHn?hpPqlCICJEYH1@!(ou3=6Z(riMEG$1=|7
zP1&O(Vg9xOe6hzp!h=mbm+Ku`B+(dYt*~6eg5&Knl+)5*`EkBUj8eG9fTf0*kwoDo
zFhggorZ)4CceB#Q_k|7lp<gbK_HFKXtY8<sqel#Wm+3D(y{649I(FAD(#7RE7w@9M
zm&}&z23AXl*Vwd-5_f-+pGb1xR}50L6sf&_Xx>w$uVLPb7Vqrb7rn?(V-@|uJ^S#I
zGpaaKAi-`PGAm9m`x9O3MtYeaXj)HZv^yq(X7`_>Phht_#1sm2Rd;^N%J-$`_Kn4d
zW0?~y0W-_}KX`X}1CWoLMcCspMJe8O8WYuR`^<c1)ln=h5E;^yiaMQ^{6uVir9LV*
zi<nw9gE(kQLb8eMMpz!(UGZ;$6l+8`o?qp4!7AIU;u-9`j%C&@q#5B)G%B(4nYFlZ
z0r~Noeqqh1ydxt8;;lMk@$$=#c8(1Q$YM88;&0KK&l@L_nM!;r8rDJ!En25Y+d?j9
z(vg_dV${gGoX@z>;%}TB*rU6ErXr-Lr!YCAf<V}H=k+eJV(tsFr);-U+)R)qrAHkU
z@^BgOFoUu}lhrjXEp>;gag^UyN+>Ej74o!~vBeiX)w@Ym_2MDDWsZj0*drf#nGPuj
zC6?0Olnx0g;b()n@1q~R40~PFy*aXb&6(}$HC8Wd(~~%*`AYJuEMJS>aXG5wnOiso
z^6Qa*2rtb>EV8RaQI321ypf>uL+dJ{lxezaq+@Nh(#H(Z?)t2I`DAa3ycsN>YS~#a
zH_fsduCDBl(BynmxiRZkcr~c}_LueI7LkQ_S9UAw42Jhxbeu2M)Jb`eY!~lDO?JJg
zDpEpOni+FQa?#cp#!KEnY*WG2^Xz_qC>2|P5y7+B_FQ2$dvV>!Y`Hd0&Bs^}Pj+Hf
z*JLzT_lIPGx{Xs{Esp$H)D#2#TbAK#YMWQ=bfw%W`|y3EA`--p=D5b0OA8_A*onUU
zMWi+=?p4A2hj)r7exf)SO59v7JhOPlY4D0;Oz#~RE$TNi+N)9fpOjo&)zxwxI?Y>h
z2PJBJ-(|PaC`LQ<ja;1^G;){L3pKSm6fCa1U)fw`r&73Jw<ef|sIpIdzmchbL|)tU
zN_G^sUe{_<pf$VegDjk0RBUnQ_`b3}epA)nTTS*+4_IhyL{!rkOy$!mypeZ}o{*&*
zFD{D-^5|lT*CAtn%xChPE#201GUzLPbJ>;c#)CXdsdK$Mw`cARRwf3Q;ULrRb|tRF
zkKi1+v~CwI36yt<a!hhkiGFxnq^kPr>Wt;Y2)av88TO2BY(H~H<cz-<n|l5(JNdG&
zm51xSPGtQBdgsn?)~^P5Qia!CvMh{h-tstr%Y86jfmP-Aqf;rrSC{7>Y`2x!<Y!W8
zvWK3v2*pULl5BEfF2qM0+*MNdmTVh-8L20MV86S7&$j&T7FnM^DR*6v_g)n0?9WW~
zglBme2#$*y35NyjAMZ4OPkcTlu})Q*ze(SgtNociS&ceba^NI&Mr*5cT;i!&C?}cZ
zlSzTcjjlOKnAy*!22GayFutiW<zwD)^)AqgZt(H&$Wuke#<ivNp;PxPp7XH&cC$Pu
zzv{Sxk)Mlsz12vd#7LlGB*Sy@!|kefFF*JP`|sGiFVfFR?5H1&BjxfI;7Gyyv=*xV
zsGE69ULD)I;=2NV6?SlSd1rDIQqdZV+>S=;SUug;_H#)!BiWLk3IC=D`q60bOEH<8
zk%hH~L#S~ggvhOo%Nnz8P7!(DpYG4=sG``mQ;8?dSFfb31=6;kw)6Si;%YT#Ut$SN
zu*YBJV$S3puZ&H2(!%3tDTN~tCm6OTcB<$iDQuNRR-2z8TC+$(L8;HnZ@);7;Y2KH
z68NYOaZ$#zSG6|*;l4{2x*OgeVVfgD4Fb!RPs07g9cNUufx~VCyet_#hvA`g#!<@C
zJMWhYTWQ2=3F55wA1I;~3}#535;mjdb(k?DGs+YTb;wJ4?&j{=-IYJ!Dx~S;coh;!
zjLLnlkZYMNer<%Yt{5{R7{zvMxyE|E!Uu;YwbHb|3$;J!C&xOfl+~kG+a;IPjd43o
zb@is@wbYEhU%NzNFWfKjgpsgxWAzJO&~i<fgJJrcJI3tKtC`m~QW}$Z4x~rDw3dwC
z&JK*OrfLyHEG4+CzL%_wM-hMPgWXiHE648jtbRYI;5IT{;{lSCz{}SpHXG7esB9mJ
zHclD+9gUwZzans%PQ29eTu60onCL6QMq#zo+u(!ue7TqMy}lSv)bi&RY?@^qnipYw
z6la|#kT_E^t>|+e-><8xO<H(1d15C1;IP_~A?qbmf#tJtY=%)HquWbEWI53Y;~jCd
zF)<$t#9~$hlh{73jZYBP-PS@}U*s~tJ-KW3An1+UWPg{j|AGBZ+%^S!itEPvLj$xJ
zMHkT&jL0`l6S66Rv)vw=@ujXaH7qNqJ32(VNOK#>Dlsb&Q4Lf64tyv*$@1K;l_!{A
zy|+xfW{$Q;i$mm6UWZTbUOoJNyZ-G6_s8Vz<4L>jy`ao`c|1oG2EVR&b$ky>>jKH6
zuLlZ;tMOy!8}|eng=}|1B>FU6n=?X@!q&ac-%wqRJ^Ep?;TO;RiCgHd&>=sTUj0G3
z4A)f=C+}}BYE%2-OZ@h&aovh&In*EDTz?^H?!wpRiR?rjuyFrCXo3~7gSR_xk~8nk
zl9|!FChk6T`RHAO)LmwTczd(My)vY(2%1UuH{)vE>MY|{7y=53BMWXV#IL6f`!`~j
zJyoOg40mPaR4(bhCTpI1|806uY2WY3;p0y-db$4;P%|<9=YX2^A?v>tP%jXOKVz=m
zTGvjSQcGR@jMs4asaItc^6t=w>hv7aHtN@l=OkR@ue&u(Db(AB(_dFQzMr?AVklae
ziSc=KvbMClwYyfZ+jQgEhnbnV?+868qMm$xsjNsyqy;aUKNr9DY!LB_F1oF$CYJRB
zkmaRe!W8W1@t1cIf86}AHENoA%Ow=ql(_p7Gtc;3IijV5#?0*7IlDxZ<)6hNjgLiz
zWc+pJEKlN6rXDAsow8(@2V#!BZVqYW{e-LLEklg%rF5s_^*0Z$lat*LJc~(*zW9^R
zUz(pa5@v^y#Om{5e$K3HMI}TKwHQD$_mSCBT#WUI=))L{&mBV1Y$18+C$?gOB!l$E
zH7n@#t+Aiuqs@y&@ryLf8~!c!l4VRSP6+m&TxZJ}@qFHh;yWSS$A0U?pN~x+PA+{3
zQ$ezarJtf3&%Y;3enC22|IT6Bn%8yy@rN6ijBb+%kP#MbGUgl~D5|A?9BdG{)-Fop
z^kh)!s$4eS+o|kUuMbgQ=9#3{?HOo4vGiGLD|SCmyB?H&we8{I#}kxipN(m~C7rc-
zW^3PF(N<Ya<Nf4hqou%_VqN@`RF&-Tc76u_`xnyoS8us!<VoY@yyN24uhv({e<=!8
z(1^*H+H6|B>ywk#b1{oVxXY<w;$ODGdu8Fh6n8^%17&N&Mhel{hz3vm`Y2B?&5bhe
zpo8N9@=({Sg3n^7#@(n9W1c9lBAZ2R-m1OV{}DM{I1KY)nw2_#JMx+vD^*8<oSW_|
z<u^S=>sZ1uTDX{S_CrN49Zl%GdiUJ9%{lJeN@Vu1&(2)KLA-giqkWDU(}!7KfA!j3
zrU!@es<!Si!6?=1eGF~23ZHFWU(>c<`IOFw(vRwbhg$NZq?q4Nf)_htwW8jComltg
ztKjAdNo;BrmO4Z#|L?j=#G1QW**{A~<?*SFl%fjNgnyWR*=Mr;SV2TUJo<9@_@*7d
zaLnp#wU#To(vK)zQHtv-4(=m6Np~BW$oCQBP;MNzvF6k@70%C&kI#>55il$h>8qdW
zZ3#5Ap5LoP!m|)QD9j#UI6ReTd){tPUxbNQkjim>4}Gr6M{AH9EnJW`UTA~!1IN^E
zeM`x`HM(Z<kaVIO-LzG%y+u_Bdl)^gK^lmX?=Y@%b;c0)t+}<aRhTkYTsCy3x8r48
z58f(abWcOsQDj;%NTU(%H>0@Q-Nf=;r6ZJSjXncACvw7AAf{TS^&~f&@cUg1^fNCH
z{f1P>r>0V5@yYY{x^hw6UB$*TL5u8kcwHNJjf`JdSG{_;mrNn^fW8>pEy|3466JQH
zqPif>R!E7d^P*sX#y!PjJq=1F#e|!5jiYF3&FVV2;<#E~2vH>4i=AgH*Q&ziqc0gv
zFtgBvXEiQM%mxB0Es9XW_P*Lw_Q#-R9pW_Z7XGkBjOzI<%qMyG(ni;_<aNe5v!s^F
zHFw`(5SB2sr8S9{_xthjI5+sEBlO==F&Y*dbKJDhRUW~Tzi+0lfF0n?XW5vET;r@~
zL~1NZyDEVj#!1B9@nb_}cMM5a-BI|bPlsY5&Apya0T!H8?RtShw7D*_*%?n%nO&V0
zL(2AFjbtXK*p}8kuu0sZd!_JPf3$A<k>*CxkQTPsqG8cO@AHt-kJ%lx8MX`c_?AV+
zhH7Hhwy_--i`5(++#jy7%zc@DDbLs>;X&pWPy6y|?yh*i`;+~`MVo?%1xZGNn9jwr
zF8lJk-}gE{eYNkOs1sjqHzdFtwpz)w``ES|Ty;e!%&?DgQ!_ZT?97o<%vD=#?rLdd
zUl8tHhw}&1k^Ca`a<VZ=X32DR2PM1N;YNlUZXLz)m!Bmxu$`ziXfrxW*7!Yhp4wzt
zK8t<5bD~E({dguS_QkE@S?aDtp1>&ejC)JN-xtCYIqFI+r$&PmuV1M@Po7jMOuS@^
zdtk5MfmY_I|2nK;t95eNIVQ_6OVj<4O9G1NZ9{$hRn%LmxLd5>Xge@Rjr$(F{%~Dk
z@nEs1q62}0g<nB4No!X%D*lJ%@@u*-wUW%;;+wbYD~2>poyXHYE1{&gyh|gwkAWRY
zdR+IE)b%yOZ2`Nx7*oBQukK*ce+|vSak(9bGt~4*-I9nR@u%AQvo(3Xrw<Iq$X(j}
zG<WhoBC7d_Y;bCB^1D`locVZ$@7r%)wd%6anI6D9<LDMeE^T4;Nm=mPwG*rcaY{K`
z^EK&@&+d$JjHKq?jlAKG7}|wC%3*7xEn!dDe~g8QCa^|y_c%CSl#YI*Sd*N6S7E=F
z<!NKir-|Ze3(={XsOU6qQk9`Sf%RgnKxN9(lo4&DGQ_%YPuz8hXG}Pv4O@w$->o#c
zK88PPeNAW57)&u*z&Ds;J*;;$VC+zBa@6zGPIHn&Zb6-!MiwW_TO!SDfWmBUL!HO)
zoT>81%hDs4{7gAnGZjwx9ybH4r5{~XT(4vzL*Dd#j%Rvi5yovanWm<kO2MQ(j%O_P
zO`3INC5^${YST0<xe~o0CrfbAn`%u-{d#>G*Kk$D{I;SX7nx_U?3(}2cO9&^Wkay6
zmL<{#Bdb_)*x##lE?6k*tThcLfBcMG#4@amNB&^cS&^~8#jNYWJ2vqXhukSSq4bSU
z6f$?z?6M7Be);53*Z=7T-6S5}W_f{~`i!%s_O|G~pRT1r<@y8$Y7U~b96z*Q;yk%p
zhwD2nXJ@CDr6#<r%R|yb{d9!Pu{k*7ds0Smt2oWjSe#YQQ9WC@QY3<W9}C$R3YVeS
z`J5UyDHIY1#A>XOBdhBokL(^vjhU5`s^Xi@Z_+;*_FGrOS6(Zs5YZUsHRG9^$jzHc
zZ)PQW<5t#tsa+;pS2+k#^ij+ix+x3ktKGJNxwlm}W)bqMZfr{pvUN=rjR^%Y8SEPr
zUspim+<C6wdCOIPk64Ky4gFrbL*|SewJ=2(O8zK&l#pPs)PnR`U&_m|hkY@YnRo9O
z{G^pUV0_BSLf^@Eb}5yxvq9<mX&n00^%s}q#%~}*rkLSqmeCW@9{phcthsEg@SQPm
zbwqLZ1BQ}>`rGSDh$2&;ypkUG919HHygNMlt=QvQJEi3HbA4<JTfU=O>=K=*65_KC
zl#$rG+%4x3S8!O~?BabGMjjErte0a_tHtO|7$o83TlDO;D|UX~7fCx@-WN&1#l+E6
zJ}C3=*8<H8XC^A=ofNg#Y!EBOUgzvs%{5_>v@RsOb`gDUE_ZjUiQ^(14^!sBthK&g
zr8|haG+@R}JR|*mn^BXfKDN@6KEG0kqn?W=fJFS-&g=P#<d4P5=n|&FUg(`B!XLPA
ztZ^LnQMZuEiJVcswt4!_1D{dXN#BlWs@mS-D6s4#yy_H5W8%5cKZS+N{}2{3!;8QE
z{5Lt~E;aFIlQp-dQnb?^>%3xiLuln&i{N<Tbfb>8=G}FCr!cJg>VUITw6?o~VSUc3
zH;XPGiePb`a4;w0Qsh*NN~NH87r#bl5R;aAj)0akym1=>!*f2l4Auw)45g;sBfF#V
zj7XCumuzwIJS@%Nb)M7sIiD-iJYG|m>MwKfa@-u?N0+<1H`nl!zkB~@DxW%$@H>v>
zpO;K`w>FABc|;JY1WUHq&UU3!w9p$#T;3^(dSm81dyKy2t3(?4Srlo8Ptx;5|H$KK
zpM0U6D1P`k9_Q&6n*+UeoGu?`GX#ZncM!wh<x>6%3XNjVsIt-o&DAP+dj(hdROq&w
z170iot6{E4O816dXOU%TVGD}WtCNnIQI$w$-AmIJy_a#{RL|2Y;LW_a(7nWT)+^`Y
z;@TBGeJC_7q%U@O-iq+bF_hyElD;WNT)dn%C|!E~0x7D7{Y8<tOOvWW*t*fJsOaJp
z_5~ll8@NJ;S6g@GI9#JFuwq9zrXN*ubieHr<mr<#S=b-Q2%havc^VR(yFZi>>=G8@
z=hfc$(yq5}x?o#I$>gJJ(o{i0taxkmkL%ZJI@6?tWqzvM!K5z;kXOgS!Acmv!l-hD
z7iVdPt%PO5vFMZ|wuW3vpl-w@{a)B*r}F@X@Kuyt>Xw0XR*${2;>e>4k3xU$l52&x
z*-$M+Cu9O=BC(lHJq#-D=eK(~$ZYEybho~U;WpQ_YIjYuP(ipuZH9<-4Tri}*r+Q=
zN>ZIvD{m_AvYKws8D_DO2A1GRFTMj-`EnWl>w6-LF~KSwjTJ8i#cS1xD!tb7GBheN
zr|YhB=pAAsM+Z@F6MxKcD_-pqSCSR*I=y#EDMo`(d01Vube=n~&OhE939+&N%eRk2
zL8L4-vUd@E-KDM5qHr(}dGsRU9orpjeA%9C<&)eIO81hk{e;tcXXerB`vh+PgI1gs
zrAx#cD2yUBxg|)|>ThnfeCco!xFcjfbE~)~S=0WK?A5%30dEQ@pc;7|f=1W0R)9dX
z#Z!jOaJ_{z%e5T4T@C{=@z{{IMe0`Fq<jLGzvqL1va8kQYz{3O)l@9Sp+Abgs!kOn
zjk#QQ<%*_gVmzqh;jd8g`V_or<dy5|(Bq(Fj<=az<S^yR!@%N!jFppGVeXukEu!7U
zYxJPBS1=FnRHUdY>miGwZk?gjLb5H(8cBlA;4qGkWmIC-Jbm9Tdrm>Wq?tW+)Sz6z
z`v))X=OkWdOe%EKcc-Jl%t@C#^b&J;;;b=KjI>w4d5f2kPA9i&SBmGUn;74_ulktG
zM*dl!24$~>eC3F+?6uP=pmk3VmtQiT+FX3|K8e-4m!kO+eVIV<ei3uAe_Ei=Cx2^<
z&cL@FqKl8rXuF1lWWTb!3ABth4WxgmQpiK=DM`&JBUVQq{oMMppo$_Pj`P@0e|kUT
zY)`J&a$>Z_eG1qzu@7T5lfU<5-T6Xbh8M-fb@}Vw!0@m<R-Teom7_sXlZr?FPktq=
z>UG8Ie#i*n29ag-#2lD~$%WpzX7q@GY%>KaeCWID19Y47FTc<`4nz-@DD*PXE7coe
zyH9P`;eWFj(xXi)kE$K4sKaS`=}jqJj+amIN?xd#EpaMq7Ax!a^zaC)dgg-;s&41%
z`$kjh<@}OI#jnG$@8}jBQ&7+vhVuX5LHR-SiO44`c-z+`bHhDiU7)s&qfLTYzSkfz
zAlh!}2JQwo`CGea!Cu=QXYAr$Grp#-4=Y3u<xz!)Kh%FLRPpAHdNRvZ;FdWc8<<u9
zTu{2qZbpc@U1T$wiEiuj0t=!UW{t_Ai|-VJvEZkfLMwa8@=15RKL3W1F<EZA%v~J{
zYp=ava;p1{4{7^$);ZDT=3gLsmdezq%CdTg=`|GyrLTP4+Df!g=`ftxwQC;_A~mDE
z;_PL$J|pi%K!tD1Y^!opr*1u`9WgV<Y$NAKfi)KSzQ+0UeTnjw)30mZQ^-sP0TY;q
zWji<%p2MMO%2wU}vcfhZ8t*a4Qj{z9iB7~v`xU=(c(6p+PK7>xX?m;0zzCQ$Qmn2r
zU!9eBaE=pe7`uh885go6%v^OR8o_^iSYm!nrNNPdOjyOFQ4M7(NK|#-x8>y(#G5=k
z{X<_Km0xC;OTL-Ohm2h)EsXcNATexc=<uf-v%3t|6|-;s4a`~mF%>mr(JH3KY_?X0
ztNQXdHzb7&69Q=%t|wSlzvZT{m-A|)f3;I7CoRdPN<BRB;9yU(s3DOk_HGWB{Mk{O
z@nJYlr`6Y3e%&+;^Ft;vrp_i!j&`fYK`OuKl&He=F6>KZ;bfM1(ZPj0tp$kzGCdwI
zWR@w)f@|-OPo~DoB*x2pP<-25rCN+V5RX6+#6!U6n&y7gIdkB*+Vq_o?^@Yyn$Yp%
z4XlK_iz7qXCZeWZ7=^nt&5hq0&KJg?VR1`7JNF}_S|E>lHJy`kl1sH##U%0g=V&X>
zTm3rh%rcksHd&XYl@DW^mRvtnx@034y9*!dnUj9NFU3p5)WqeEz7e1O;;vkxx^n}G
zrM{qIjz}tP;7_c)&5@rXa~-)q?-cL9x6hJ2hXig;b4+`8>JJ9r)V`w9+vz^?J!KUv
zX+dYPR<T4)VIM6js!)<@kyk9|F<oY;j^13E8C}+cTG7*e0u{)<eF|v~YXqo2hT>Jj
z_rhO%!C0aCC~qsQcB6BsbL{YJT$=bmD1(>1G43k@sqgjHz98j$o}VL@aa)UP(AT??
z0#aPL-kJs9d^R~nx8NK=X<0oIe*f}MCf1+Z{3MG}9`ED1FMYh@e$B}UC}E71GpQ8e
z?TR>7&!@{Y)-|7B7ymAE_}yGGIcP|2G;?krBS~Ds*m5#)Fm6ct05#>w{ha<}JnM=9
z&MRz<tTA>SGG7C;N1{S5DT~FbLl)}}_X_c8l8tZQT|@_>(zUl$<NLS2M>-GrvRWpy
zGMn!-W*f5abD;-~Q`c^H$`U#gw%eX8%O>Uv+lY0^&A#Z{-x_zp+ma(B+&<)Y`q0YQ
zdWcX}uFIIkTzXbT-gvz|`Rl+nsqb@=AFci`#@@lXv#{OTjMK4g+qP}n_{FwuyJOq7
zZFFoq>DcMXIrE*GshYQHPSySkcJ12FdhWIEbroMmK5&`E1*J7ej0t}myK+J$R;!J0
zs>`R4G+C2mn@zK(_eI}jw%qv-nqN5lC!xjqAHsYrKQ8zGMre&-PuY=l+||~d9iAA%
zPf+&)iA0daLyp^`_u4xBg5a4oBL8u;Cr-;G(FtZsZ6=h>;l{UT$S=Q0b?6zQb@{L1
z=eiIS0-{lND9h4M>!J%WlD=uAf>&tesf>i^=62H!Zc6^7`9Z!HqyRlS{|w-)p560#
zXA~dB7jR0R-Qm$<=YzrI9z1IO3ISbxBBAJq&=?xmw+Fgr&;H)uQ}WD#%3wJ5hh{$S
zuiYQE97)!SFrZWEw_TiZ`CuMBlnx+#^nL$2eYSn^3)FR!R*w9|>HGR<hsNh<>j|X{
z7qv~)K~$28BcfY`1btG$!{U{(jj<ywc^Bh{i<YA%5ES&~hzdy;vXBai9I=rplGYf7
zX6M1gxu^O>zpX;X>k?|vCNGd0(`Y&`Y{Cb**=EDw@K^;)i6Uj55-l6!5@<A4Q;Z8O
z*%lXyQ{&z-rHE<CO&hS5L?gn)F3L!)i$ZoJ>6HQIssh)y)iEakc#J7!NJM&@$J%A_
z+;l;3%TOk}kV?ZzdeKc49N%6QlzbsfKxySX?Yl%8ekkZ+qjpl}$wh#LJ+*FayoHI>
zQ#f<w?OVm^we6Vil64|6Si6qbHTHp2U5Ul=H`#^;NzW>}_5id??*(|YE~`oo2>l~-
zwo$G68$5I~QrP3VSWsb!DO6HjJg+`&NI7HWJGN<hE&Ef-IZNZkTj1)*6POO4zCwqv
zkqz1B@^}&=b3;mAtvH<!q5;~=X6<l0lW7Y!+jQlD>|mji^R%&T^Ce3=*br(+l_#qZ
zOtj+mFZzMp(<)=$sM`X}HT`W4ornV&bku0*T)VYF6l!rC@U&s1!#mKvhY@;|@kTIA
z7caB{-(oT%mCTqFIjvCLCe2biGcwwTVu~3DO#l=d`^~Sf<@;yV93>-XW-<)`&JOFP
zWNP-@mPCO#rezFx)%BAqH&r&fFPxT|v&DUXc#Aei0qJ>#lPWK8)&WnJ#+6ZKN6@I!
z0i!;F^|cpUIQcJtj;_%)#S|3IB$^5xKov)i@10hQ)Y=3K)l+riz1p#)_7t(_Z(}J7
zrUzrIlLl@zm^}}td0~aJAFhdJX;+-=WH5Q1a_vl4XlA*=twRhmNLS+uAhRQ7pS4d<
zj!JQ`z5S?pU_>EF(5$l23UtJb1?c#Mv19pSBvnGWa(;p}h3<gHb($1~1tKTN-*ZDS
z=MVj-IDRJdgCsOo@`V(?Z8^Vr)4mdWdUNRvo41}wvlS|)O1Ye_8fE4iSkx8!UtvyG
zg!bk<(cs6G7&tPB%C`OR$7IX|D}evdq)R}1OPvPw!(msI`lDcLFCEs{%4HO0s5}ha
zGBxyvR;rq#xrDcXK1m8JjP?b4gSy*qiYSGh<#BdaJLhO37~E@w>@sPd%w2n!nUCh;
ziDvvDjw$~%PFZV<TDfan!>gu<i|ug#9-ps=^gb2mw(d`!TEfecTutykpSk!^XyS#A
zV;li*E6r-+BjSX=#v{}~Aaay2m;LEx;TuX17Cs#Bba7$rONFj&oNqBjk!V63sumH5
zmlgvrH3YP(Y^afC(&9c*iXOH;#ZP8xJ~4gY9kL~A|1{kU(zkyo9YBxiUZ$3S167bz
z(B<f6*gy<w&eB{9xVWv$VT(Bnykg{yF=Ff%j5w>wwE~vbNgJ$lOnMQpDnUw-WA3;-
z#QuJ3bSMg1BaoUVQL6~`3e9S*=Iie%cl{UPDt^~h*i(7S{1`wwPcd&4|A6$s{Htf{
z%+l1=!9IdF<4aASBvqk$^28zZ$?teL!N`y~842T)r&^|uHydCovg}|MFxHzC)&a#`
z(&ljL3=}lceIezuVC*HM@dj*%a(LWR*PCDqh^?0gIjV*NaCu&8&fyqEqBLnTf!Gg(
z7&Mui+A8QBlL1v2b(<Q&Y5`0T4NPkfPqt^c#zX5GY-(}rg%`T`W%zoxa8)6&si$a1
zn(0`bk$auKTb{XE@$=UjW$ErrH2e00>vlV=Dqd|NZmZrksR{)73r^=q?3=FlavBzb
zPUila$vzn9LDzF>K5uM5lhJ^C$zJCQxj1F(NIU*vQ+Sd;N0ER(w^q9KMMTeYk!I`G
z@ebos*6mW-fMaK!&*T?SY{a^?>R`Dzdz%ykd!$6p>D+-*G#wZIm$R0_;pBd8Dm(!9
zK~cMO16YAos7(=;Ll(49&E&unM(M@;(;US)F4x6$sSD}Du%`A>vtY>GC!PY<cA=Z8
zqMSniV<%0_+e&`0)U3d3I#JP2bP5Dh<7fK`i%8#=%{Mxk1K57)7)Z@gRGe==qDOW6
zDxeJctLqwR$3k|$;MZAT5x@Perl+g5s5+%-RaS22u)`j})YE7^FVZ-%K)@yPVekDA
zG@&6zjmC;wj=b*E!|z`j*q=y`sYrZIEvlnHCX>r$yU_}G^<u?8oYzw!ET{&L32^7{
z+w`_<$77vn-d8Vyk4yxE_AmC6u%4Q$&bV9B$Y!C-asx&R{JHBbM}3v)+6($S#yw4X
z+it|&wF*&o8nR#%v}83c^L~o6g}Aiy@a=7zrZ?5Ljah(NH*9xDKM)d10{h;De0G?t
z%QvdyT$JC19=OWRn`*`alq%DDO20!hy{hPHQEB*GQ``z>bvQR?vx{V*YD9KeTD>|m
zhap$$)c-y13iPzjRQI6Fq6j0HN;}McYc^r;5O^nN!@H4<&Av%wLSLf!r*8ZurY-|S
z!zv%Qp6pXh$i$S9mA5oEC6viN9%w(dAf%j0+#s?|(<Q=|P%&MBWW0=_@$!yR7{r-L
zw?Q3$Kly$OLuIIT{oAd&0#>qgPBydl<9Cy|*<2b=uI5CgZ$}bIN3|Q3o<BNi-ydn)
zxe(_qsS7Rihs}_8gB-$;BrCi?cGNhx<#%s9@0AOpkgXB8+8@s|pVTP)`V?g@R=IR%
z$4=u_J|tKJP=-2+yoLFW0^Nn>ZNMe$zW%Yx@V!vsnbGmN8w&G4c)p)oWj&U`5sw+7
z7t4~vFXTw_|M>HLLSUohz{tUerUS~$(3$P1=*CwTch*Ie$Jah(e2%pxpX1xEO#4Pa
zKYEw^p9B@#|F`p<nU$UMe<i4vai-#M{##HHAkGj+GxQ@8$MQiqJs8^w!3X|NK=4a7
zk(NDc8(YfFyNM}3F>^+u$AI?oe5$w%A$)3Tf2L`Ks{D*F<NWvm>`{2hV4n4C8x<&}
zjj_ik>-%2a`taPj0Kv+Z&KKk>?(te)0w&S?XKdvK>c9;FhFt(@3pZ2cJj25Gv4jN7
zcOTVXB`=W+^ZmG*E<oS{{`YIj2y7guY<}~6sconZMedXFcpdTwdTjSTmlBSq`Oaa9
z#tX3_taVzH=T{VI|NV6QkjMW8hf4Rsp|FfXeMmTRuH)(Qh^sezZteRl3KGH^C=JFX
zPfra1l}iY<-CJ@IL^*T8=vt`b%|sPvE!D9RTK^%r;LgKA!FHSAq(1aM2ZJp8h>lhS
z4k&cX=uO2p>#N_9_`d^wOILj};qfbk$EQ}#dJ|Yk$Bc5nKZiitDWRbX6E>Vv(ul0k
zxR7EAr=m7HYlwUMZ00KsOE-rYsRojc$zL~zHO07iOd(#A@r)<<#2;#$G44r|$Vuin
zMXqpx6V*oum)I5j{rp&J>%4)fd7t-9_l+~8T^Lf3Ac&Y8*D*75byU&1yyEV-puQyM
zx4gDlq2e-;X-m62ua8%Olo7h%G<mwSlLjKPp215&EN7=Ql-f_WCT%ya0|SSB+?Et9
z9ZNiSG;P%-Og32?Kcc`EV-}@f;J_j~^~eTb<~q_BA7Z`I7Dc$q>F<g&9Lud+=IX#=
zNg-0f7)@rP)ihPp;S}{vm*cQfhasif<}2(pF8_Jzp*uXo@~Owjy~WumopvC`_=JV=
zrXftQ)X3K$1RLP72e1G=tW+%0{=(a{1s)i{*5$>>1Z48hIM+Ku^#o`t+{{w*YGvy_
zV|<5BrjN$+GOXlWed6R(j}eNiopHCV8s3??nu{8?VW3km^N;-IqOh;2szr##G3{Vw
z%JY%Sk0tjmoB0*V-`5Krvq<w-s~(0hgPM580K-iU%mLk79A&SP1OLor#B}<?Kz3B1
z?6~=2DLzL9gGHiu8dKPegxqnWLR7rhm_q{hQA*US5P4|}#yEs}v@rQeLj_pRw7A#*
zMm+C+a@AtHAF8}R)ZE#a2!Ax3u0QMa*XHO965d#F_AMtsym)dAyM%i+x<u3AwNo>Z
zFMv}xKhIrcAz>V-S}~?+_6xJJ%vkwma3$%&6Tn`NMmK0xKWO)=1AU5IJ^LBSONN^g
zZFDKhe!pCO4zELN&O-xhQ;B3#wXm33euTyP{m#gn82aXlc2`91KVBLbF0*D$*-Hp2
zbaL0$r;vw0FeXQHVe0FYFyNm}FRK;vsqCCz94Wlv9YycmLQ)YQH|PsFS%3t+-?a{G
zYCR#22ayS&uX7iLuBUXucXMHrMbe2W)EIY|%!B<lE{4JG_XgjI;lb%S-wpJ*p<=)`
z-$cV@b)do`FtM6FZhJx)0>FYV^T$#Qx_ugr?}0ZJcH@y?Cf2(<BbJ?G?P$E(HT@&M
zrMmtL@pAO#H0E3Fv7fwDhEDWE>!X`Zc}w)bHV9qu(Cn7w%uoFuL-Vv`hdXej*56Jd
z5r*d<14<(ohVMJZ2KAP_U=Z^NO?M`)9GY=sI6y!Al;(u6mfSvt_wsPIzF>A>x9Dr|
zTjYLrI1g`v`Kley%P$@h)2(+OIBfl6+z+bN4+IDH6va+~9hT>2JnVQ@d<q}~&?6g`
zMZ-FGk5d<tJe<|al1&ekX7|OHe}~@`39N_rL&ENqe&D?7Mf2aD%(h|VIzz)|LZAv1
zxqfXfqqjbt*8hl2Xo|7>E4RbD+$q0<QuUO$A<TmD;FI#^S9_o<r;Vug$1zkwt|X^;
zrFYHR`jk-&n(@^u&XTPX$dZ?+&P5DV6#R`f=aN|Hs`}QlxjYi}YEf5tlB(}Fje%NF
zh+WG8xP8~V{C?7(kjXgToZ`A_x;Lb-<+=sTWTJ`}8tgJG^8(lB*t(^q-{lF)T}ExS
zw2eBm^vpL>RNoX*Jc!50mUlqF-5<TIF`?V!x@5+r1-UMJFOtL6v6K8M5qzcwRSK92
z71`sF;g<J4>FH3UG<#!stItLp2$lmhTWoq*ZdW<I7Thaf0+7JzV|?Zj)W4Ud{@SH2
zGkHV`73djxr`94I@nznod_t$uSW;uAJs91${=Nk%fb18lTYL{uL%fqTK9x*SFFwSk
z9-3}h+J}EG*akIICW`60v}MPZwmo}5S+i)-m<{Dwx6tw&$BHS;?WD=mG@Ky5>Fh6#
z@<;kQ2<QpRso&#%$>DThL)eN+Q-8tGLWZ}kV$_iLBUEb1w0^y5EkJ~3nvv#f=p(ND
zd!i+->rBu}XJ2<wjox}hC?UUB8}zR4Zuio-GJ&UJA!5`yR9^&z@Am0SEuI?EcJ&oE
z$(`>hq7}Y=vIu~I-8^*AEOOChLi{_4R-tjUI<_tr={DmP3wbQJp^~4*^W6ZgKk8^8
zRn&WHrN4TiuS0Yyeu)T)&9q~O#|)3c(FSyi*KBJ?)>-HJ3l0h#NmM&Lp^BF2y+bP8
zln*<T5N5N#(6v04UOW6`#91QYzfT<h`V}G+r=8-d(TSx#NS*K)jSa%Rd4y)JcmEA;
z(;DW{t3~eN1u6@7z-_|xjTO`SN?BUL?t<VSypCmfMv@mnNTdYkeu4`^;t!#7;r6Cr
zNpl%6vbef!$1pJtdD|z<jfe|A$1a)JXkk5^iW*X`7TwbHUGE{U;=Q^C`-UkqdU01C
zPb$wD`!Vku_BPam{`8f!j+)Yzju=}N*BqT~EQ7F>0_TF+Yro{U)T}RZIZgdrj&t4t
zyf5+qpVe&k54l<AzKJ*0h&}H3-`+BVjW4FxOgb9UXyhv5{kgju#-m-_N&zt$FwU~7
zB%xaY)*7_oxwbUrMmXw7x9t{a7TauQ-Q@Cg8R@-MW$Z&K(hLI2>@BBUz}J<IQ{*A@
zG#XV7@J}~<nM@IJk=8mgUbW|}kb77pDu(vdh_}`pQb&T<9%~9`O#4oLZh+v7_7XV5
zjvd6vE-&fqxwg+CD3j!z2%4Am>FZGPoNArTl!8!abmN9eTZn4#${ByPVy~>#^ohD_
z^$@Lhy3Mo6_t4ZQ9hXhlQ16bAv;6DpqFx9+OWR3%dOc)_9IK2_TRcy}MvV8$cXXI{
z66OUfHdU8aXUy~ya%lvBbA}UwN?@AJb+@fdtxmC3hm{jjOKeR?9<iHU2CdZSN|qSb
zw6Y|pp~02u+Z+_^Q_#~UIISwwXZ?&~^Huk{KXB3E)zp7dd>sE@#mB_;KbW8YAFp!z
z$E!SIhwYuK`%4g^PAbM1U>Gn>s*~SF??mH8-RZp{^74;By2nPQm1VB@8bAAYB3aa@
z<Xy?x{$_r-fm`_McG;XML>X`J(EigU(7o`K!925trvuO=GHl&$_%1<<IL(|8ei2(D
zuY9lf+tC+jU^*6|z2FYe6+t#IRw6;B;z?@9tCrv9S1_=N+<yhwzCep_KLqssx6cRj
z)dxoF?u!0B-FJR?uQpl3QPwklzWHJpV1kI@L*5e?VEKKr`D}l!sNlUUHj@4`;`@AM
zcOfx=HWY#}+|WLOV4Sx^unsPKr3N+6Z81mMh?qa2jPDPg|Mhm>^kY(r!#r-Om<A?;
z%ahdq7Ufb=A-~u9NBW$FWaM9LQ<ERUXZ8oq;Lrw;;yOQ93QoXGSXv+^?Lozx6ZS}&
zJ4z*Tlk3rn`X+@x`=@;}=*>HcfxqDDnX50}=|1yQ4I@rekkUVCYi3*)BSHxL)Vr4W
ztpyJ(jhFs!Q>=lyYuGIH=~mE1!{0VHAo&gunpz;Sn96@aiCm^P9xlL%b>&6FBxYE|
zgCw2wvo7(1kkuJ0ElB6Q{s^s_-hUhy!)7j{YQdmP9^OrnLb-YCbeNJr;6c4*Jpfej
zczz>pX_$!6xob7@8<M=DA5f8t<WV-LoSP~c6FR0@jC7sI`{U7#5jAc1mcgSoDpmF;
zr@GJgTlQGK9vjF}O7Zo?0EX;8T4FrYDTpIJSb%Uov448Y$2hJnem*i-X(?$UekIJA
z%UW0qm}j<P8Tm2iFMa?q&N|T(zd2N>^UNU!{LZk#rViw-w}XjmR6s_(AH4+M`r>8=
z4^~L-S(k9!1~p1--+Gey4XIs9Hs<l}LhhTmcbdW<AG)KNix{pL;O}C-wkp8hj!lpH
zE6e<Cf|Fp*gCOrf5d|9QnhWaVQlwf^sU+;Jx|~q};dQqBly^1(-N%oaoi%e&El1z`
zbl$DFp!}MkxL_xxyZBM&?yM{>E0WQPeB|Z|v3{aUh;MBo4|J_IUs{&b9-dOxef?*c
zmyF>?9RsRW1N@k`#xa*U;~-m*Lc~8B*-BO0se9^Kg05hk!OTy?Kvo@7%R!nQBNQ2g
zjw~GcrL|2{E3XOLoC?fgYhmId!RPZEc%ioR-<=#d;hYR?hT|6c;XRoMH+ooAC%fzl
z50l4IQ_Ay5M(YV=JQH<Kjj$n^2_G31KXrI7Zg3^HDT7*lO5nZ6fpQY9c%KacCuy3)
zhJB=Nv!TolTvb3Tg_j7WqhMUYE>SBIrD8)G`g4jiK!<NORlS-$wz-v6<*Gp$2W=V-
zFO$@Sr;0Q8BIZEn2DlcuYJ6`{+xqY*BXOh_dW5m6CX?IJ`WqjAH7&a&|4wowL|Osd
zvjQz?=DfH*75TyCG%WiSVfUxORo?EK1e-(olm_K*Tx?j}0Uizq-Z4sSmfbTXn;oUc
z$m|x<T(0qbcDl(pa!Al^g)8PLq584dENWI_8)7}7WvI@T1MS#v7~}$8zsW@fA|y?4
z5x(P>`CL*Cw>11MA#b{omR6mVqR}DDnd-2ySoSp;pJ+Mb`bP^4P4G<JLl<4UY+BO@
z_IbJ7&!^)}oucNH3OuoZr|TMxd)f>>9)8_v0-+e$XW8#xLBt_!-^Bb4l@XdHAdAQF
z)^wyMDFO>6rUH^9(aL*up%lbLAo6&fcih6cL*GOv%O&+vt)|jEi+^w#{B`b@UF9hQ
z#0_2X!M(h&l4CfM|4ou6EE<87X4k?Fu8He4d}zV^SrImqP3(}$u)U|&{O~kcHCYL$
zSv=Az5(S`o*)_58oaCwuy40~!6^2Fdd2O9JS*<B(Yg-@Wr9LD;9tU5#>!L>Be1;J&
zcvW1t6nC_Ywb_hgD$0lFUbJ|ow;?OZ=D8DLf)l?=km?_x7dHp+gk}D`a+JEb`^e#{
z(?e>JSfjgp9kTnzrc18&C-sS@B!E!MyLDX5AAt@QMuOjmTCTT2#@WbPDnnkk$j;P)
zIGMO`_t5a9X^plx!(-=8A(ywOzhU^|<Zd$y=Zi;<!lF+Kd1mg1T>!GK@X<PzRh)pb
zt3T)^j9tZ4r=|NNrFt-9f^E0<q@6fCeF(B2d<tyu5$fZCq_;UXer_Yp!>AZ)UdMPc
zhaKc|4LlR!m4TaP#Of<u*UFYORJI)?c`P`C>`RekKpUGKqD8pu7+zAF$DMZ2J=fr4
zgSk^Rwu$9d#%YwY$|LGLshfXE(d3ZHpbk$TUgrooWGDw_q<YvcQJthDIcYh=2V3+H
zxm*86iRFZX^Up3~uM+9p{O;<0!3W0xi)i?cQ>ri0T#9|{c6q07dyoTm+YLgtQ4BoG
zy>CI)IK(!<KhKHz$xqmedoe7JCZ*Fa)?u_h5{25nDI`1`RUq|d`O28#bR%ndMd_p#
ziHl{TZ}pm(S(8ewdeHl4+cT=ac9l2wY>||9Rcf)EAF_}!7E0~?Gb#lmwF}wqOQQnT
zrds|~NTmX<|AV_PkV_Sq#{CSCPZ73>?EgouplwUylE+Zj<qxDv@-6otj&e43uKy>E
z{@+wG1!F4}7h48-LMBGWpQGH#*@cjcgM;CJYA@K?{&!5{Wo>JFoere08-s!Pb&%uZ
zX@^%zbcjZ&9=-f?2uEm}OE&bV8q%G$G{Y&m_gn9<#nfv_CsQpq9l}6~^zNg52YkRd
z4x1T<?jO{gU*G0#s+CX+j&?!Z1vVgD3(urHX3PbhqU0%i$}!KR9jc4)x%D)OHB_C{
za*7(x{fI}Dir&Ua8`fyZCFEif*>FlyaiplpjYVtosRxoOIDOoc1Mh0U#?4XGAt~6B
zl?Acrgd<1B&=Zb8<r$K=Dh-ZR+$S5P&Eo3GjV8-MSo8fA8y8lty47q(!nLJ8g<3I&
zHB|K|b~Pc(>-T?Q9f<#cL3y?0S%~#bkL03Ji2Wdap)1%`G3%t8)z7-jho&|0f0rT`
z0q55cU~_Jaw^A&g)I+rRO4{U@cn_6McalH}m#)CIJA*P{wr{|bpUHvh(5UAVtuu3S
zl~{ev=P!<e*~MDbtjR3Lp!GYTI=F$k6-dG1YLwZkC=WLgxBkiiXI~ygOqR|vFygS)
zbs3np-MUIb6(JreF*l%!l96iowPq;Y3ZNQXY}{9f!mwJBf~JR5QA&cYX!T2IrxIGA
zwo<HT<6;Mzq*qWYK5!aEO(>iynE~b{V+ER552P|WBxKy|g$mg^QK*&N#Iq5OT-XT_
z!(fYsu+gxQOPH@qFodwEmPJd5szSA0uNdSTCxz8!V~y06>xiAMO*?0CY>AoyrbHEH
zo|-t3DsYu0A8r5X@!FmOZEqT^+I8vsJj{bHGNV2Dym)+caL2&YeKB_7?pz~4hzpai
z&b!n{AWz;O={P)}6Y&52*)D)38x~iJ@GO3c0^1YLR}5s|Ht;g-k40@SFuly$oVLAv
zlb*Nx#n0QQ;m3^<jU{DHSH~bO@i8<Z^zweJ|Kj#)7K9&+FodCEv%TpbA1?S{^ef>s
zdH?Nbu~6E(!<(fWS5FTYj-hN|9KQz}KM&Rq)0eA0a%RNoSftxOki&L`Ij2QK=A=WL
zlGJ|Tpw$lD+w3vuDfNmDo-YV}pUc6iN2g?4-zv~E$_4>FoS<2{#Qma?_JY~Ac&dA2
zxc6qhN(tKgA!SVc5YEikEM`Gj93Fvjw9z8QjKqZ28p}kBbyRZ%G9*vVi(hoNP1Q%_
z?S*aNH`?wWL%2)pdT&2J{!DmO05+e6%RoPSD<g_D1FJuG_Mzpi`-jFY`SWER#c7`I
z%h1_@hsVMRT==#@XI>`;B<t55M60u)vaifL6gsT@&~PuI9ZAVY!c)EX4dbWZ%i-B|
zx%YXl-a$$7LI(yx-{y8tL5v*vX||6r*fBJsqzm$zvn}j!(@E%rjObdquAmSnN;}Nu
zkWfI7C;1gOtD+DkCc8CMQa-S36HpKcFCXzk(r`v1dgv0ms!|d{z5^e0;$G})4amS=
zTF5XKmL}JpMs6@;*>8vip2D?Q_|*x*+69=ZXY5ImkDVN?SAPB9!rvP-u~9p_G_i(r
zJ&hOBkk+@kE$(w0U4R-DA~{Eei6_~;-+~Y1=b>vHcSAimVHY$BM1`+5ys1ST!Vy}$
z5iioWr2+m399DmQ5#k~&6Cq+~n)+P?OTp>Yk1aoO)Pm~P?(jpRybzz%F!9HFX5zWD
z7SOsE+#w=xwSiBAdC{UWNX(bGpi2&79cx7$HB#2Ms3W>HBcZJ8nU$G|3EG?MCAj{w
ziU(4oj~cdE3bm6}N$?5RM;8$8EUH{XrpZ=1qhdEOxb4(~j7;<W1A4MGA}BqoUe?KV
zvgs%l7fJS!KW_lkU4R-liQXd>MDFM6UbP4N81`5{cY@E5oCAt^n#6dXe^cpWmj1p`
z<gNU;rrNPkU^*E<yBUjp9!^-v!dqX5_wSu)wtvYyaae0;TZ*!OK~-dJ-!)NXZ=3Ab
zS^tKhYPW9%XpcXua!fq?%nV#RSfwR9Go*x501^+79QS<Pl*y>Pcp@X0Iqy)!gplna
zEyraeMEZ%x=T4%zUOU0SQu}dP?tzI2%7Y&N*dC1@Ow4);x~Afi5Mqi1gL~)=_p>EC
zjO!wWn+R?z+rNQ@cSYXLz&pFwyISiPmG#dgpysXMC6rOl!g?bpK?}SN#1Tiw4dQS`
z=`I}%vuN7@YFG+hFn5B+9GGX-1p!U*Zp8YffvC8<MQJ|N-eRT1Jj^u?$UIgI<5H5k
z5htC+7JToVm_yiqC)q1S2eNr<X0PoCgIKUo+ozfR#n7}uoz$c&0(vd!H#G%BJ{Z+)
zKrRH~%T*g@QV_?9YxMuw3Z&M{-z14)#31#E=oXdb*ajGX?H<*(JNTKdg5b29b|-cf
z*HM(b$esJDc<rXxzM4bzVnzz_gDTmLfx%jtu@R<=wT-K4?M-(OSE<2i4YJF!$bSh(
z$Gg3OPc>eGN&Q{WL>Hu#Eon?r%8S?&G=tGfr3>Ed$tD&wbylzcFip4Xb@0K4?IBK}
z=7WH2!-$`@tN`_KUoSxSIzco<Nr4X49%LsL15Xcke5T}7ufJ)=20uW1l`~Wpt)djO
zYDZ%|;E->zbIXtj6gU6&hHc#>>FTeUy5Ui_P7K%gTcbN(O%z!oPb3M&Nh}iRi|Wb$
zvP^x3)Y>30we<HZ@$FXH7FGntpdGo{+vhG{;qvB;7$W6I-gWgE*?b_Hp@dcfVX_hv
z?kF8^Z7!CSWs5Er<TC4e0^`^gA1#ZJsuTLPVvf^@493aSo;Da+znSSu;~LZ-YB=JI
zn9mLT(FhF&4$j1og0zKnf@a2Md)EyzfQK}x1dgTLRUhszu;BKq-Js<c#HJt=h#Z(H
zk4LX4f-SFHTGJuX)h({*A1XY3@IvN5kZ9F}zyR#lzl{%B#}CAVTo<-VNTW^d#$9vp
z>JU|k8cL%l9_`N7Xxk5OdovWkr?>WgjsNEO<;f4H1YEcOZu#@SxiGi>I+nm7;Qm<d
z@#^`vSvR*6Pd^KhY|sZfn{#<q>xnm89{ti(zDH+Ry^u^RBLEN3y%hfJP4vIp;Hh!F
z&M_}PN$i_<JyPIPll{M+5{&oagpdF8Fn;y!|Lt`X|3xEK7T9X2N?+e)%!CkP1uDXW
zkEi$Ev~c`4SpYl7wQQ|?O3IxTGtWih_6L}1W&M44boIEgbrZ?~P<VD#5@uw~B>9!h
zwZS36Z?R6jw{$mj=cH)A3$tEP#;%WjWhWWfy7;y5?dQt=JT~O-S_hFFFAz`R4QJ3+
zK#%f<d)*3wR_j}dUZ4Y}DF%I@*oJ2I?bgXUZev;!$alt;K>gjQVnUF38~^yz+V?%Q
zJaV=zwdsmEu#Dy7I^U^Lw>|RrYkd#bpa0*p0}i5CHwlJPs~L|UtQ#>cJ#hil_L4cU
zA1n#p+rV)u0ltCjEydx@y$A1^qkOMy+CVo0hOl4sNN>8S{jS`GrKH_1?4WuYu77p<
z%T1P#l#+BQSHxS+3^XEGX8^V&sF)^<MaLvNyxsa{+wVo28Tf#D>%Zvxdl(OWh#SZh
z4iHCI`Ls?|mG3JU3YO~_3RZ39k9}^f_XOH?)Cd1T#<Fw$zu>z6BUM>BS^rla<TmzH
z9B#+;j)9WEd?V!r6fE{#^ET*_`Or^ENq<C6FvWkBlzb^7NK>n+3Aj!2UjgxL8%v=|
z-PI&5y-3M3gr72myJP%GM%f8aSxUD@nFsv|&m+%gP5}RkX7u&U`F<uXX?*l7jZZ^P
z<d5_7erGQu2`a-9R0zlg%8ru<6|x{yTIgoh(6g{nkbv8`XdfI=D>KRd@gQFq{Q1U(
z_r@5)6A5XQ2-nUpFZbIb$$MojpM{`dcVz!%A!`OObIblLp90Q#2*2d|mWJK`y?tLZ
z82kecO8erFPfbRj6#Tnnjn7TPTFLlDDZ-FV;e!$(7;-s*qOGvzi51@`5BamW#M%q&
zFKURmVO_Ez4Hmew&>YnbntDyr34PS#_b5nSZky0v!8)uoth_QxRCB84h$--uQMOp=
zFBfjVw6Bs_)-X9jrEfkwu(+GaD`eqN%$OsKWqKY{ot|cfoXu(uy78bB7*z~lLk=fp
zTg1w~uuBDmx%jX|I50s#+`atTO&IE@5vugUHUeollT;?Bra(P6hicn3FqrDRZN4b^
znDvOuSF|uGD{>{ZML^PMR5{eRZk9b=ACX5YW;JyVvylO09ptRoSPKg4_+@e@G24%g
zZk2U`gAB7^?brq~YS9D%HNINt*J<FPKja51=2s0dD*-(cjg?HWVZGbhAWtIsdVu(l
zGo8x>&(|r20OSoVWhw8qF;+*l78cGqmAgk(15rrFDR#(=3vQi96P^opPel3*xQkCg
zq^wC=pm0*kI;(@jSX7c$B;Y~f|AbGQIdy5Ch71Nz#I-7$W_;kAWMaKIp)IPIM$=Ab
zH_{SpJPXh2b*jxGtLvF^@3I(CRdQ!jaW{_j1GtASe7Fio!8PPRl=He7y7kTHznhUs
zCkpmiC{l}29noN?UDN(aQTg3C)i*Iiuv9n0X=chVrz*)~9*k#CEr|k8WhoeiYU`QT
z$E_1jOzJV1=&iERwHE@rhF1b(IkiDdyUBDFtN6MTtvby(r6N&Q9n;UH6n`X|;yv+5
zYob@<eD~8cmZ~r-?OgDllI*W~))n^zD6kkLCLu2h-I~P61tQ8#QmE(W8L)e*i6)NL
z?>&unSfq|ADKuj%c%(7gP6WH57;YP%RwsMkg&#eu^*g3}y8<Bj*^a|E=vpFsT<2G>
zDEQ<^t`0^!L?hV`@p&d$vV1V!^$#|JnWN@_*q4rg-^3~Hjkv<^{S_W8$=3PZ*4ZhD
zQmm-O&NrkR3~`~QMXGw1l>_(zQ9LeD3u2k(^HLFWq_o$05NedQZBYRt0MvoE&T|$n
zaj1>%6VZ<>7P6bcaK;-`bX@dhznDIq%34!I_4Sb(v1KY5d||AvKgGSdfIp1RaS4Cd
zxESLggGAOzST`1>wWw|ef2tNjf{sdKO`OeCd~+Sm>Ks*tS?A6`^al((i>olbxqgS1
zS~gNpD_*?0GvAgt(kD5AJEvl51u)Y>MUZ%>XIa)7#_y5#%(K&Wys>TImRFPQy2z!~
zf|*HATVrdONmxRBIaTZQW^W1asC8BYI*ZkzHjwL}_x~`)(~||RG`Tov<}A(7e@*iA
z+ABw9YUdN2oEt-{VCl%O5#2rZ{vD2VZYo!cEk^d-M`AR(AcJls<BycrEWUQwj()tv
zhc+UJ+JV49Bk}8jC4si?qr((|cGxM6KqWXbkH{pXEYQvH`*-?y{nt`dM>QUmW;y#V
zFL7PaJYlYc0-e;dyqUey==7S(%BzTigevkuO~jR|UGtQr4H_ax0ljTlveRxM22*45
zH!Tg``H_F*C)SZOO&--(iHkyPMr!tg8i0qcI@poD_DydQMoM>ZOnAvQaVM$lnG&|7
z^kpi!_~aMga9AvD*DPV4{gls|+CH6s7ZDdjkhWp8@<x=lIysHMd=ae(j^>XU?z|*Y
zJI;uXDo<ys34`mNj!#WGce@=Feb50u68p@sz!CiDH3qmysbAEH^~0aFhpb^zlCCd8
zgU3w%Tsy3U)FF(15FmTeQX^C^@F3}473F%2QrSE|4bK6QripPRDXM3z6{wEGM1cI$
z9=svncf&<k49cvk2CunWn2Q-QRI9vhz!!L+b}4dGkn0@7HZU>pbq9p>@ohKUiA89^
zw;V_UXD<)utujHr=so3o2pgtZvyK;U+W&g`gb7+CFkk}9b9KZ7QEmSh^Q`5UrU=R2
z4L&HHG8VzJ>KxZ-$Jm~#oxsB)q2LWx0%VN8@k1udJIl0n?~wip+t@q$gRt}tn;pba
zixYhd2_Mn{ZSB)P)dW?X6~j=s@_yS2r;f3Q+>ZpQiuy~O@$Ze=KOY??>$R>Pwgx<a
zaqWboPOiEBMLlH^e~F#tQ`I=(j=GN4#j>JFTRV2db+v26yv(~K84Q$4mdF0>$;{k5
z`7pC<P`+Wi&>Ei9TrvC^&Vd7nMy!XJMV!f3tkdp7>zu3xtvahFe*I#0=vUG@NB==+
zXEc=AuB-=@84V?D;@MYrxq}!w#%3D*-><(Ew%5l8gAK0VDYHJ_^$3WhQY})xTuZul
zkksV)T+6_6;)=`Glu!XrtQ3}|SPRFqtdv4Uhb<-F$5lBl>rR&Gg_9zltCSNn<a2|U
zS`5r++shV#HjgkW1~An2MXj`+_tw7JrSWwRG^dK^JQ#SBnBI7k300X)MGCI(9Rs%4
z24S{2^ZJq+JwqgJ9g81bRatr@Ye8v8$2$J!3sX*d41GftuGC+(8X2spl4P@jmvt9G
zI4B0;+E+V$g!x(;Y#;EYG-9YNXCk*AJ(pnIZTYk8-ZHDLZHK3N8Y}oAN>&Y<YhrV_
zm-@^ybE<i9dj7FA&nrgvp%|t)y49DTk_y0{7iXP9wOJq@3uGT0yz=K$2poFr{BNZ+
z@ky)Qre1B;ilZv_J%bLy4}XrCI?`{WL-p>K%uY_GP?v*`+B4nWihTWnq*L)?eEp?;
z4?etJn!n^K0NRd22_GBZd)f8ZZe6{2i^_RjRyfU~oH1)B_O?&QVlcrEL#CD0Xe)=g
z8j!O4xDV<8N$bk`u?ZaxNKh7p#ABrZoh#Nq(zdNA*cb-`S-BNe#&)|*wD3bmcAq%6
zs4j&5ZT-(vH>iY|NC#8(pd1K0RO7EnwR#SY`Whz6B!C(H897yu6ubtzN(VcR7}n3L
zlao$DL@^(nPJe;Rs6lC8F(!65MQM*~bem54_LRHoDdjEijhb8BL}oON$($lA)0vP{
zRLLng%n`AW^PrlZDl~^4(PtKh{)Jo<q$0aIl`jGry_2T@pa3}-S^u9w=KoQEER5{`
zyI6k@&VMUFegmZrv%nZy=|3=-tjyaiCWIU~y>w2$V%E6Ugxs9RZ^M(y0cuYsy!xPg
z+apPoQTB<Ay1n#Hp8oEiE-yD?hX)}Xdw=fFr$8Qs7Y!1?n#~=72CX#teGUA46O^^y
zanFO5t>g$K`uKdahI$|rNdE8r$t@nRp!wiGQKNU7Pe1#U4)Fo#hp-p`yqiPKU*SI;
zXPY^uASJ)pzhev@mLAHwc5lp`<Q#-)ntZ<fFbwh`3VO7=k=bec{yD!ny%>pWrzf$B
zWW0v^eSN$BCM5W6AcWlaU$q*aAX2i!BO8P+KB;eJ7JXqJVHEEW?|!(?FE=AY<m|SQ
zb{P#7Wid>StlyqoTRKp*7_5&C;FZuR?Y1u3u3}Yp!YC-vHLS}C*$GF1{Ik=|&@s4_
z{>UZ}znIL$ULv$6j8<J)3;!D7S8JFe?ycG1xixm9@2S%yM%g7D=Lq%~E2tWY3}cNQ
zq<ns*oU<Z17;jXQ5DPLWx>^Q2HlrCC>i%#EM^ZtIMq{{YQT-|4yNA)-HL<}OTu0I0
z0IjduXydp`C@KU|_sLXd7vAGzFCXTJF~7<+x{RJUZm6F$7GWr^tuV@L*~m*oP9c+n
zyA_J`3kIbF8d+BK(8zZznhFP4@hPB?-Idrl45MK2J!$XPRQv$)J=5??g28KUm<2J6
zh}1}FXiLW<Gi=a4N=oE;PB7MJ<!8<m%<k2Sfhf?^yzdY+x($$u87J5Dgh?J@I6N?t
zbmMomDg9<8qMNzN%$lWx=G2Q;W3!P76^^e=)9b*E$9v+?#dL{+0{7a9rKjyD9h^Ed
zORr?VRHJMhNjMnML)<%|8^yoxvy8cskWx#}JWK~3gw`R`)Eky$;X@xi;b5gL@DAVX
zVUV8vr2ZfcaC<a;+DjQSYX&cfXaNBZ*;&7$X}i2%^6Ct}-`@z1T+{3R#fnf-5wZ*a
zoqGASI9tbwpW^UqHC>Rzh9p+Z4T1>;F2aMOM4{Pxa0;JH)ZqB3kcw6`&Y~D~O;({k
zIOuN_grsH_jX~31`B(vkQL~L~{_mK%riYzR@tc{tPfXr-2W*MfKkb`dT6cnss!?i;
zaF5?ekTU;T+o3@Ql#<pQTLYR=2Edy`X$BJEZv;QS)CtY(c9cbh&`dxKWAeMTx19rB
zKIjPt)fiC~-*WePd@`P;6Do_2euK=gn0nJ!QZl-;gm2!??RxHLWIpx&=>Hob%fO?|
z8!5H%MwHqV>!dH+;}hQQnA{-RFznY!>5)Qj(5!^{B!aZK;_kMO)(!*>mXJ8u(F__)
z=8w-fq@(d}j#MO`@RUYK?TnbRkccyb_^X)-Q0hlLR?*G3&oiWgn0}BJ-YDl6Yzdm$
zqe<QLe4*Rz%km#HAwPF;NoZgwhG8UKF?W>qAJFI#rBt1s4<7tR3XnN9v|({P+5yj^
z+&A7(dm|HJMx(F6m<;)oDp9<xH@Phh?_c#l{VM?T?Am!8v4od!s)6}3&|^b{)NZ0h
z^uQ*s`eEnFbW<#e+sf!?;?5(-Mju{*NM>HS0@qBmw*n#8Y=*(rd>;DaGE}F)NT2F?
z69gse1q|Ir6_#l&6D30Yf-EaCv&u3sc|#j5U>RknEBF#y1~I!{<5szrs<`zWy$z*-
z=mUVwMZu<+PGWL?igK9Wv9JF0x#lpFS)xu{ZAS{7#A>#>o6A4_QKRhmt+tS(s0%3W
zV$l*$i(*?3UHjT2@0~G+T;p2y&Lba`M-x6yTdVfs7#nSk+w3dr+_ol|=s_wD6B_Xx
z_aIk=pYJk6oI@>Mu<RgeV_L~_bHFwhXAF#)OHenK!LNBCc`Ms)?3iP(p;VN#)Cz}J
zUaeMTHlS@1Gb-L=!sxxqG(Q-fiUV7ts$1Mm%yX<Y<A{6Fq7DH1a<a(vYqcRsgZt7!
zJWtPNfI3ifj=YG*0YR>J6b(bg*wz~xFU243YmZ}FenTO^jZkkP-o?izbpE2t)9GwB
zk7*KOt~3e6d)|2(ZrQ{Mta^~;x;>-N4nGoYY9nJRqf&oFkh9U2hB@bm%0JA~r0bZY
z)#fR~-+Ic+?-@5o8e{vnkH*Zza=<+bU2I~{mS+<NZxD3lgW7Z_a*%iH&k;pOoG{C4
zy}mmeRcv_bVThV@SNO2A58P<%fMi>er>z7wWNt80+aBSPC+4VlQLq+YMLlj$<~>y?
z*@|$`Q=^{ckFC<lw+W1@uC`f*x>6CBNs1!%BSI^jF6;00p%%OC0;jzKor}l4l-wYO
zx}(~R%RBtjdpo7)Iy?JHAAe{`?7|Ri4^CUR^{Z6}(V#zT-$rnHygO2>w%qXL=^`sd
zrN6zciOh_d#b+IBV4yguNmOZpVw#j?e{~9|tDcjUZApyfk)b@wYg+DTl83z>eM&E~
zb)ECAPA(QBH%K?FGkJCKs(=Hl&sFuX&#{v32e7m{p@&nHevInpN3pSBo#+#05L%Q}
zbp|#v^ui}xb_Q-N#hgqfmMDGEoJRy*atCc$WH9!M&CO~jm1HvAvhh+Pw7{P%tODTL
z+~Fu`i9O3LM|xU;Jz>h!v-@go1vT@2_f8%3QF<Rz^$fP9e1$A*(OcUg$Y(s8+KzW6
zl4#a+oo>^+$!2;YZ9G#f@CveH7h~rD0a|BuNUu@?s5X^WFgWq+W>sb$BDme?kuRKL
zd?!79LRNdLt=WFJ%HC{TpK0r!JhR>_BDjD>AM@U<vA4<Q#`{#Qs8iOTYL-OL%O9?f
ztgXbV7F^2ql7<F{D?PJxQNPzP&#)Bv?7M5OgAo~kMtV=?O)w@D7XF&W63E0(Xz8!t
z75L$)XUipI+RG~)VQg_8EA(wuW_Oipr?8RJnbZCjv>Z)M);AzHJh4iD9%XLjKUOy~
zwePWWo-|o}QR(q}7&;M~fXHESqr@N9su-dDD_UV(S+7)a6|X|pGT+(KOrR5w8~=1O
zI2y^h6x$^7w}t54oZYcT2-m9Ss8DVi;ttdXR#+vs)Oca>W#a1_A6mm1hYI6r_OEQO
zxr#tF#-xgJ{$7y2E`PlJ)zuUJKXmuJ=Kc4<flZ6R=w(>fT+0UQIdT8*cO|9<yjIMQ
z!$;*(yW-xoogQQD!Y)0*chEC&2H1mTpPh;lR*pxBdECwZo!qhxC#0WwkNZHw=|JBK
zATYYAF4fbCtz;d%gS(%kliNAWC7YurqH95PqYKTs_q6C}FSB_rrOnC}R<vTff;IdT
z+s3fgIlWgh$z|xCQ?=i<B1wUbx;8J50c2@i*pg4>D_J@_6Pdl=vq5!l%4|f<^Plk{
z1xB8{$y9!$xSK{@9H@&$V$8_AHB7;`6A`Q4KA}A#?)q9-W55z|z-66A-wlv+;*+bG
zNys^vuRb3TEmHhLb)M5Jw@fdwKuJm_6jhMKvF6|MPx)1*zTZ3m-dvMSYCD&&@^10E
zLS7DA=*I@nkrmG>4iW0u((OswnIhv4*?n_+tK={vgwidoh0pyDzioO2J+ZbjWp*p0
zE^o;4P|KDMC!fe%?JUNRzvFOhzvXs=Wmt@B@)ej_SSy1MVJ0l??4xpXLg9i*?@Q21
ze*~ZhcyEu&?MKR_Z>~T4r4dK6L?t@ZA>UZBKpE)pRP6A`#|9-;uyG(&;_cBEkZtkp
zuUv^~+g;devB;}Cd6lNI>W#9{?T#nVoQzC^Emd&1!PY+?J{0Zl^Yh$2z67oAsAO%^
z_EZ0urM7dfHf;S}sOD42j~iAk7E#?aleGa@nz&`X1bi-GE*-E(t9)VG`f)-?4(4W0
zBRm7ii$<4^tdTN;T{K##D+jDl3?Y_~gWenpg9XtIU4;li-H*KMxdu#{1vz=HGHz|B
zvKUd9WW_edTA?cTri*o675JSrQl)d*5Zl+aKUgajF^{eysC+3*zjp8S#l*vvp3oGt
zO0^Yp=@84DxDI$3Ax^};?gEidk8Q70|9T6uEvukG{u!((;t_10UA_k$s>M;bnAj`d
z=9RS76?&RSylc|c^qshsHjzRGENV)!(N{HM52C@AH8orB@%K%%#($s+X2i7r=_0}W
zA4bVsEdRT8mjAj){GW9e%*+uEV554L#xX%Vcpn%iP0R-8dW`iMqP{$0tq~1(8Eh>R
z(8khr(}o=eWv92+@#Dk&w}<`hX!dA~fgp{k(>j<vn8<&@!$JTI1j6jF<A?ZT0tu?r
zp*a@6m=x{9vBO>9zU#x_umBdIU7%iI51+pf4qybu>YVW>l^8y5@dywxES8}PRrmyp
zx*lGir>A{*q(qcrF6VQPYy!ToKCcgo3(L|T#r;@_dxmfKhY9+a=OK4?3&2#YQ$JMK
zUVQ!;#Y=tU^#5+Wes7f}@uvtAi-b*{)#724)!BChoEFZXs~Kf@GL(oF!XewrzY0B5
zVn+D9ARD3-HVtC81X)a386oR`8i>1N3c27EQ}QD72n8X<b8q4hT}L5D8#hym2OluY
zDtWnuR(^hE?;!c9GhFw@>@qGMj&(N)Sqvw;RYwK;ED`PN^wOc~Z)h?$L}Z%LBGRC3
z?!K82B-^Qx3hX^1@a%C!$U3u(V|ax-RL>C-p+AY0zy4Nzfq0L_5dWY)N5kVoibs$K
znUS{8$`b!`2Y|Y);0DC-YPWcdB?h1{X_0Fd*i?4{UY_D@J#`ax<3f+jKNBfez;~-*
zl6rTW<Wa+k4KAW67Uxhdsk^GXXYW)RSE}VAWqRY@n*qK~!wncKgMRp0!0~4>u^@xo
zGIyc%r0{(!Gg@$i2QNn_yuadm-!<B~a8ez_cNb=65bfQaQHo*g>@s~rjkhm$Z2C2^
zuhP6#S~D!8V%K3&1T1+w+NPI67E}g=5j4#Wn#EISvt9K0Qu|s+;vLfFK^AvLYWCp=
zt1ecWh;r9d*J1<BU(6=1)jH+uS$U_f93bG{%I;dqC1wqx>xcjegUW8Oxr{8$@J`*Y
z;2t=#&Zg~|b5}8vMlqKuMBmi<)-hP*xnn=S?M!D1Y;~j9T}qXI+F1R^%d3;TkO6`T
z`5{P|M!$Bc-{Kz9SWhY}rjH6*@i-uYxVSjr@uHZiU_U$dhT~;&V}V>m1s?jy5MUJ1
zxoAT{`w0UL*}i2CxVk^)+Uojz-~J1fVxJk+0A@lzT>qIbLFOWO`<a6eCyoi$77#Hm
z8BA&Dxu>c(I%kj;tfZfxA%gRn3sJ=QOR*&tbIRqfB{t}7n8gQJE!?91?#Dos;&y)i
z9A|MESm;Wrgzj9`qRIQ?950&2mp7M>9(-Oz-U?M-DVP?JMEJkfSExZj=y5y+Pf0GK
z9)whig(CH<nCHjV3QTF`HD>Bx@lVP6Kc<8^{`uJnQOF3QM|mN8?CoEpLNtj=FJr;@
zdJDRgHYu*ab{)>G@9+EI72hE>g05ST|EV1pB@vE+QlFKxS(9b6<eobjPVAPDN@*&&
zlh<iw10J9gpu>|6Z4DRSDyzaulD2f2OUGg^qsAlQy{S?&8<wA)VV$>;=loPHpi)gT
zrO^7#)^)XPaoY-4>90J2bZ=`p<>$n<P?v&+ES2L{qvb`=u4+J5%sX20d(O<@fjQKy
ztrA|^aV#}(;#KjY^Pnj#!{>36*+gL+0=1HGx+~q-98IN_wflTAEsVpK4*PqIS-uGK
zaUQ&Y=3Q#kVxRb`{0P5@DF`asTefO>%ga#4|K0M2I?@i|vD-u~v6zzxzw6*FQbxv4
zMP`(Fj>D`b6Y)Y<y-)C>3w*$Jd7>gu1!k7nsrQGm5x5DlUHxnxfO+iLG}08g4?&Mn
zqA2Y`@|D4Y#=v(TC;xP<De=flL++0ayT<f~8(B~wOCCOu#U<(3(gMsP=?6bFi~g?B
zOxhU+dIRO;By%)B?xyh=Jen=i7B=pgB_#5kaXM4AGiy$GAWv0<<36nj>%5}ExA3aY
zh}r&&#Xv6mZMHMZmRgmg7AmOduLHLASgkLMWU9}}$F)2#w`{%R0&q{lQTKt?M$PwV
zCe5EFEDaesG7qbQ(%TI=fz(;SZ$E>sOTZgG<M9ZX?FMiKQSDi3?RPPH{SmJe1uPsZ
zAr>o*X&qPd<k|*LbXT`?2n!xwn#u(wxN}wZ|Harl1@{teVV|+Jvtv8?$HtCr+qP}n
zww)c@wr$%sc9O|ARZ}xn=Tyz<i>|)t>Z-ozcddTb`#isprQq-SG11&>A|_IX>Q~Ga
zq6VOb$$<9INya3GfWEq)MVwo;0jx?0>j_4;`rvQ^lhO0|S;-#5spRrl;yFpbb?&$#
z_SA*$qgUigrqHmYE}>o6z0891B5Y(zugCFcQzChtUT^!AN;Zy4>H?c=yJ@>(o1Dzr
zb$)J~!E~p+ifoGm!2OW*iCX);ya01PuVDIx^1R;?g<(|v*H#=+Be|;nPJzp`^<V7}
zo+4=4%ZlX^1xhKmX}sPmBN)mOlpg?^Vw!4~cM1I!I}~1;cbnvLj1C#ClfCKWZzw^P
zFsut%={WTLE`gmaghbnhHGAi}H9nV$Q9S7e*`sO4NzMGeDY-;`EL0tN@gWK&EyI_P
z<RnCa;Pgyu@&y+IceIGrwDr+Hwvx8Z=j0iP+Eb|Qm2ooY^v(YAqV1J>)&@dei_fbv
zO3=x-YK!r|u7AJ8bxm1vQhI_g;z1VYt~rmLh%*6Bn&~SYD58>I*IbhmkEL~s==mts
zBWosud%^Wm4?nJhnm9XT%)!b(?0p<n)D;tq5(#ZCuQDl>loeF*cdjTc{TBVRU6W>L
zAOKyut0`v7<l4c<m2{PUIr@|440j_A@`4|zP#!;OGM0<6m%)@5Ek^x3amc-rlh%k_
zKmS=ovLsv+NWyRNHA*V>@<(WKLA=$Lc<UyHxb=~)C0$3AlBAm`Ie_bB7Y8Z))jp`P
zG1qT<eJJlncKxFrZMAgGkj8=uA@ji|9lk0$>x0a{OwpvkC|(`2J9CbNlBMZgW!jvY
zroLiMIUbYPOKr3liD+8~Lfp67kcR=Y5GG??^0-P*LdC6z&kotJYfTCAX1<O^cq%QU
zQY+C8A)gL8Fpy`X#vw@s4FL*M@zeD@hjU-$;wju^G^W57Y&y#<UDf{U;mMC7_eD$N
z4^=j#QHxB$wrOmqNDF<-orXv14|1-sU1Eh#CX)Ea7fQ>l8%E@w9RC(2Q2{x<ANTT3
z&LPp3H_a(r+UR?Vs`&8cMEfzAFq!Z|bZ}W+Ri<99t5r48v*El)9$Oo=V!?4Yu2Nu0
z3BQ)TPd(|SLBq-}m8Cj3y53OVlkXL;RD+5x`#nV0glV>Di_nxDKd$v$=Iiu=&+Zqy
z^ERNq7j;-GLhjcD>`H%eGfPXB`rUJ$B4cjj4$!pBL>(NL?${4D4g?eI;F|TWGFyjY
zTEan9l*M#h0oUR+e0%hXuXRIfIukWh_vktEA{rF;)?vmrspu5P?`GSPP0yBOgtVQN
z6bABBe%+_j9m|XDz`V4;1<i+@s&FRgWFfALu!B=+`BtOmTcb=@<y{w9z152#v3iuY
zAO!J~n)STjc3#KWsB^T~<7^enFO?`?kSHh&ujZ(TXW`%cP71Pp972sS9}@23ldOuf
z8MdsjKHa#4JBa~?Pb`K<Z~>;#HQVa?<2Dt%;$ZvTJVD$647Y?{+n9?qb7T)yrRkAt
zBGZ0%u@fU6%>Q*QW$+P}CIk}YDk+y<`Ge;{3kQ|vsyD|$YmO@H+YrYZubzD&Tu;23
zM&6RqCAgti5;y(~?m!Mw&+5Ox47=Fs$h_0khNNP<-F=QZ^Mq4*=7xZ+_&idPHV?&a
z1bq>KK1Rj4eZbuv*<=E0OBVv01ihL~tcVKKsa&ahEhn28^uvv>_Q_<x!Qmw+=qlxf
zA79c|&f7|}xBrZc9Uh#A$y_%9=k2cF6rt6!rDKi~VV(6=hPB=>a$$dNH+)%=1!E9S
z)nsdb&z;t4oV)s+K?~37y}wMp@-$-hsoc9a!kOus&AxVV9h4-|{T(jPDLdYS6?nbP
zYE^vaRZ)h9Q!UFCitiYW@zZS_T-~}9-h6+@SEK#6KK`d=rzfyAuz=y>`hRyX|L_X`
z@8kc+q;!QnVRP7ex2>DMS<ffc@Jcp|w4Aa9;gIQX%a71UaDZbH;QgbVD|(@b%Pa71
za&xW`Kp2U8si0nHTkv$kd%3M~cVAy4D!`|3z^lO3<U;l$#C-c37S08N>xuN=zB|8e
z<T&czdJIb1(~|0-<>GEXxsmtnLG|MG;`!Z8Q}&{H?J4ZGa;f0lZbiAmB+2WCz&8wc
z(4*mv^z?i=k1+XhGY?aSc%qrvr5u9fqVyNrBThY_=Hx%N|DxVz5vW$e^Vj<;csU@i
ze?u?&##mY%J;0)4(GRPyjL1y&Y*4}anDT7>;he_*u|98s1y#aaUX47`Zv{C@)dm*%
zY@YOongYn(IWr%jutaF95oQnjSU^XYJ63^Q0{ndkAE(kjct613e~4pL@k15v)sUMo
znE^&=7b-g34qS)+7kRHs-<dS;{BK`cYsDbfK5`NIYp^~xUC91t#bL+5k7A?k4WDc@
zL>iEh4X^%pe~ryAFO49c!4HffaE#-o3&J*<AvTK4T(HTqQW*tVEWLi}zoqq7tA!Yn
z7ilT)o(5hFJ-Z@Ols&tgY9}(q7`vxj6750Ki<t4g5!39n(QYVGz>gQg7OP(2G@r?j
z(gh+z%BWomdVfM%;sVqO(U4HAk;^3Jf8}8$dab5kqO|q-05iunwvvypau|y(FV{|V
z=HZQ(2_E-^JDJr(-6mbaKaizm*?FFA=jgif(#6OTMsfl|Wp!Qi1;zD{XK<x41+Vda
z-cD}aU-ETXv00|sr^UaS@qOH}+f!oK7(HW{PlnC7Ih}XikSypy<Ybic-+mh$l?y|!
z!lbD$Oo~pR#>Yy<wZ272LXS5p0L#3i_vJQuSzR_Wn9&LKFy3#JCyQ1XP(KzPWsvZa
z@sj9K;>|O?qonqe+mRc4^$_<`dI^{vK7$_x+Wxk3Kb!*0(QN0IcZHZHG{F`<C|DK6
zNojfqtS8UxD;o1!yuH?sc`87BHXn95SJ_A06s1U6EYzep(}0kKzSH${Sem~@8G+pR
zgGBqLS*xYQoJLCgfo<#C3sy)K%<D3Rkn-<6NmeeIOp@W+*4(kBOc6Ihq<av#mj_iv
z<$RJm0oNA$bBR@!Z3xi?w}0xD6HKQJ85!P@&_h<Uv!h|sxs&oLkX=pA|4UYlR1%9Q
zfgoN#dKIl8CXaOTNv?pkp&UM#NMDLoeCRZ*5E({bEn0l959DuuJ!TyM5FM6RIVx;4
z@cB2$P?fkseyt7pb?nkZ_}2?8bsmHYE(Es3JBX&+IFg0aa&|l$bI@l*JIC2c6ekJU
zrW4Cf&0r(a<U{&f<laLP2F@<lHuix%_~(763%mK7Belj;@d{9qP{Cj0)Jo8XVx;J$
zFjY~LPc9qU9)Y{CxS5#(36$3)1E{)`J$vzi8gxXU<K!IC5Me>Opmw+Q9r-+t_xx!q
zJ}zQ2|L2VhB_~`Abkaon7Hdrn-Zjv1k4dmT`I(l3ahenyN#rw!YA;?)ZU`I)x#hw?
z50(gP$l?1D@5m{*#OobKS02OkF~C{7kqwiTX`TU+fsK@)G^tfL2=zz(1~>p0@QB(W
zt`~mwV8+bOE)3F4;m0~Wp`7uJkg&!%*?lntK0YdOesU-vM$hWU7W+uav_-srnkK7}
zJLd9QUA%OMmSR!k1Hdc;D&chBgt$Sr@Wn{#)xY$iFku1N*p#_K{;eX}>(YOHisOWf
zck<^O0Ml0=M_r>&f;mchLTSWS7nZnDK#XR*1Ypds!)N3HWck7F-VH;d>S~q*!qz|X
zI3{04OHS67{>a(Q{2)EGV8@#HvsBLXqv>*|Dh@I~J)wIe3}ix`gI}pb?)8aTT4C1#
z%^d7c1435}0-D5!dN=u_qXJzMmr@UFy3097%H$-mQiWgb(yNM1J<RyW<eu50iXpb5
zMFN|=@WJxY(nh0^hUJh&NCHN6UF>z1!bb*Hn&4uQ51oD}rE#xbpJ_kKb#Ky||K;*x
z0MZ;RDp&V(H-$!6Z&sC9WGr&ov`D!uOJ#DoLq!{}b^2|s+SYMlgA<Ol$i7^CM>KKb
z!(eBll1Lc06TxAwB3VgY=L_T&#7JWcA%xbt3d-7UV<QM$^lx!!%OkY91B%J+X#808
zd^1jB4Cb@W@Q6Sg-B^0z?L~<{nd3qBEh!%hn<8X>Y_xuVhhs&F9APlVH#MG}P`*3b
zMT5o{Nfg9vkK#$}&#v-0mQR|SoLkN(i40L8$~s04WDR52c1Z>2D^VfoZ-fps;rahO
zAl-rfWcl?7w!vEIi;A@q8d$!-;~yHxj(w+r$0*9z!rIJc`9E!ld;Dp5gWj2LHN#YN
z=EkW|kgFQ?_&VhJ9pEv!HWYJ1v*lOE(Dr~U$)=!IGDRcF_8FmSR9PLiMuK6|`i0A=
zdNtpN&U;;G%W+EV{Z*T?>!BN9MQtA|)8}1%qq+mvhWMuGFcrl(BKFj6H3aqH_#5qH
z_{o0Ez2D&1{tiCMd*7rc_Ba&H1gHbe<n8B!u8)`tqM-j00E5(wGxxp}pVoL<I(}@D
zveRIAjaLprmVK=9_S+b<6G?yLoBXy~YB&6rpMT2r6OrUr?sH#zlQPg1JK8tAU3MuH
zR)VNS4>Y_xCUqoyxd<T^%_1@?`*m;^%PrWj`_1TZQt0dsvDF8$wXa}r&U7KYbG3B(
zCHEXsVh7K}NzeVWQ4jaHDeY9r;{!1pr>ENoPKN1v3%={lX6R)|L^H=jNm4Z_bwE`m
zPlYF5o!hhZ_HyWjHcD@DMbv9L-&~n+_rtx{o=y-a>Y>=x<JB5!3Q0iMFbcOnHH*T*
zKds{<cbn(?^$h>CAHlXK1QUbKpb^fkS$CmgnSBbD!)xp&Cf|lX6fI2@xx7zrmJ$eZ
zc5wdId=(G{!DsNOt=SW#GLxjZ*D=n>A@UauG6g<^EP$|5O~bs0&YGBCxh0GpmA+q7
zjdTXShCSCLc*U;ih8e~5({j}t&jlZbyK9iQ6MJF`i^CsKo6o8!E+NfWW*R%~OBUKP
z3XfeyQ&23QXKRasi~yf><R@#~5<uO=45j$HT9``i_twWVJcv^(S;8S4%mU$pc^Zp9
zfKZV|zM04OJ)^X1uPF66C#6lWTTcSpE3UO58V)ZJaTKRQ*`GN}MDg$~u4k!O370xT
zvhg7!%sG+>3O@v5Fz8&n?`dG|u&-6#(iy_L{7);F^7d?8=wmDu^JaOakrRVZ@_rOz
z(6odd@OYz3t=iMGF@q9L8I)GfLW|TkU<@v`qLro1U1&e5@h2l{XDLeiAdFwL0rv(a
z6}vd4Q%0O3_0Z;8zHL2E>ZF8qtXz$Q?jnf9*2fag4Lc{r3PgEaTJ<`=3t+wN!(own
zrnf&h+2WM&h+T4R`4*v{ilMS3TxEWMHqDjev*I*5_V1oE6|0!mw=2%#-ThK1Zipet
zt48-Zk&6F_$1ZXQa`N}7^nIt-f$X&awdw3idAIn9peFMfj|sECZvW){1vRC)6ahXl
zZxUsKt<3r(c(Agov9zU1aYG491&ox%!_J-Woa@^cKISv~f1^*cG5>!odjHSP76uO1
z|EZDP`T@q^bX{HRGITZZanF4KV??h1w)NXaK77s(B{G{ARCxUGINhNmjYlp{UkmM$
znS)rZ6eX4?8ER=YcxMei8`xSn>2MXbCgX*q>dN5p%_iwYy6r)sKfOFG&dN?%IKf|_
zLlmHZ%q4%tAZ!eY4ebH-+8F3c0X7kHBi@Tn3iY%liTW|X!yNBvIR}}#wzAE$C#t+Q
zp4;Ufk00Hp&iHo!4%_W0gjdXpR_M3ZKj};dhU~AtiuNThtNjpM+oTpygWd590v4!*
zERq($f(Qmjk<LJ6ggL=DsLWteF}8%KuNvUilYMY&Rn_RjHbl~i%e-|;Yf2*S95ge8
z$vF-mpqXs^vln{+(zAoUUdb))pWyOL&wL40p)`yRs0)`IsgHC{&WIBAm`4O3_yRb_
z-4@sex0b$)z>Mu>3XKWS-vb8U598nYD#pc8^#P?zeMJjbGK7K-V&RlY5=Dz{-Oao4
zIg20t{?wiA3aj!nwVclVWMymRs6B~s&km}u)+jlDct2r{=fl+1l~n%bq6D5tN?;|^
zuY3_`dr?#_&M(M|2$=$+09e{%S`h@hWt0!{yG5P^N|^*<hQZe!z!M#)#uXpG596O&
zcOsOZrZk@F;el0x^2ivlU%{-O9<kMupH52=J47ML5R`^qqKrYdF%un}u-%@~tUsWf
zZyQi8HpRO|H3}qw=2t@Z&fXrjYFrCR)Gkb5cXWL=i_t|U)IpbYrtzZQY<)Akx4m)l
zi9!s}YK#7v3h$}2-9zJwqVR>;O|Q$Kk=8C{jEN)06iRfs?@&vi(IEkmLM|T)stJOn
zuv?3Uc}C?(_a}Z$oVpVG`2=5XQfFfkQqAotn+7)6>r_*8NMPvg{%xIY6yZ+k>BH5D
z<b;BT8V`#~$}FR@+&feWb=ZzaaT-eC2_rI-`k*UZjkbGEpn!sms?gE!FX5~+Hp<R(
z+Y5r|0t3bFI~tQ?2%Ch-KN_tzGzZDZ9ui<Trd1!D*smNU<ZoORJ$}r|IPQkH&@-_$
z<gN<C*eDUerKP=zz>xfW|Ln@4nqid*c4O)Uuv5UbFD4WSM*Y)r;jS5;_{))ILCgYy
z%%NtE7hXHiOFZEcyA^}1GTB^7thO<?jCDgOAYJmj0)6bAc5kFsV_3V;Q_4)8*kC{H
z{<<tOo<`B-`7&dx`~7%S;}8P=EkV_*Kz@Ng@L0$duq|RLoYj5c2!qx%mzYINkmY2l
zg2XrSgIM-TzmB?ZSlqo?KYd|_`Q=OxAo?5G;CoA7XVV)36$(}tGw-wWMxejcJj$%h
zJg*oKV7WkCS*m=-;Qh)Z9jxq|d)+PLiciH`7~*?-#RaB}eVNn{KeoY;GLJO4*&k>}
z$W{R9zlL{cw9L=av5Oq5l*Vm@dB$#6BS9HW?_dt4_X`ZG;Ktr7^q>EHe;iSg-YIdV
zV;*4bQ_{#n-|FLI0#WM)Ud3B)2Yk9q$%A!S`b7m8nP*_!2A{FM3JIlj!D@bLB(NEA
z%Es9G@_>2WU2*^-&#!+ZJYUVM%XZMGGmRfmg}Mg<5nbKLFdW1ls9q7Z7CsdY0gR0O
zXkHi$+|0&x$(r`pBXUjzVf{hTzz|{Z!gGAHeqcW7#tA9xvw9|Il+Prs+t@M;p6wp?
z7%c|;o@GYU(4u<iQHd2?#2G;#=ARe8ShuSI8Mvf{vjM=fevT_?AF}ZGZ-VdH*Kl>R
zxvx=y$nm*^iUOF&dVvJtcv=Y%QaQ|b!##7!I_eohM<m61qwzp{v~0!QGi`Cu4x{pJ
zUd0|iC3evv&}mz>p4q$(agY#PbI9>WccMQ*L)c^?fQOWD^W%cFG+(JtW;pO`O=>%S
z>#5a4QcJB$^|}F=nQ_c?`+N#*g}nQp@XFUm7xmgG69lAsxYurew5$h|BUMiQBVYh7
z+yOmStGcAE_7m@c8G}UfM!pim<!{Ph8fEApJ0S{5+scXcB+~R%6kqswgISVSl)A`+
zv;JR2ZZ?NQfT8*tMQT|4tb_{O_>x3nsRXkFLK9U~)%~@Pel}zE=Y7~|=o(clFigS<
z@WGHRLc`*9<g<8DLmaW1>9l4jy8);d3D<(GZ2lz(f?@{=V^-e9KXYgoeT729?Xy+f
zGriPdbTs6c^6Xqmi3F?gb6-gg3I0g*<C-anB4H_KOH2(N-DWQ)fRiJZU&K#1)~jXV
zl@0URTpdQc7we$W64Jhycvc2ws#?a`eS`K|($29s4z0;dTRhK>-Lip&kMfT1(QzzR
zR=_e<GkQ&}lSDf!%P6vgFduGxhWAf1c71{uvo!>heKl}}fw}-YNT^d;r1z92L-gP5
zmPe`x@&>JEo!38_ME|j^g*JC^U<oE%xM}y|mX@btjgZ)VD%wA$47|)-08eQ@qbAPZ
zC$v9_tB*sp-mQ}LanwaNd)GDw^cmdk^_bUy4B)N5aPoQ^2eBoOm)B1535xz|_ZwKZ
zlJ_59M?Uzo6KEE@6YGpgl0`<aGp_{J131zcgbCHj=lVxLvP&?fDJ<~b?P?r0v$Tf0
z()Gggdgq^WM+vVEEu`8o8&jvvMoJC$ccrg1){m7*-T6CM1lzY*Mt)w$SVl#cJs$9k
z1dczKY2I5bP<*~3%rlas>UVMk3H{p&TpnShsivZPYBT}Lp3@=hHKoc!Y2$;cvl$j0
zKsqHJWxd)R@v<jx44381?EzbnNx1otpyevFE^W)Sq6uoex^&BS5yM0Fyyg&JTA*{&
zI#n|I&eG2CgGz{qRADZfFOBJ_hrER-VVM_65-b?#zRg)r9b_CGlS9@Mnm}eA!V}oz
z*g3{MI|Q~B4LIyH)jO)8S{4|DW2utG@(pYb!r|T=gZl=<g>KvS3=BJK2o1U()U=bg
z9D-8WSCWNu@G(Ojn=cLnhR}Ip?<FJ1`<=@!?polqD?m#E{x6H@j9Rd4njHp-r<oX}
z4SGZEaKzY0z~F@H7a)|lDX;QJ9b1nTR1SY8PZ>Oi$zxYQJPrgvK+MHR9d?aSZ@|ms
zKNa!$0L{eUvSjt^5)|r^<2nS`c|!Fiec^)*;zH(owf<X1OPpnbQ-is63kP8+#88-$
zGk{8D1!lbBo)!S79O16QK8od7ZA|fPh};B$tfA<MreLz1Y^AYTW~uR{m8SJkg4cRe
z#8CfHz`Z%nT3ch1VHnKwN&xTO-`*iyW~lDOZ066eSYL|p3Wg2IMe2l*JJOy)(o$D_
ztyU-W?xIVD=uy=1PHf6J2P*JAVfeX3IR<MZbSF1T;2V<1ICXHHnvTcDKUJz*o7t9z
zaMYAKxJg^x6fhFY%<|Z_e?6994izu>t(oNl84iMsQO6=6S&m~qoE?b@C2R9$D>akl
z3sf_}>yM0gcvqB9)M%EQBC<9ryY3Wt4o*dgIqnL+ghS}}$s0Q}&l%h-*yhzlUW(fX
zQ#-S1#lt|n;51#!s+UyJ@1r*GYR9wn$Lw=aa_aodJI6RV8Z`pFD^Kdw|L)9gJ$!~_
zuUvO>rj8aT2I126((9_UfS})jTOMovNmMw=(}WXmsldA%ed2yCyU%xuK95#lqzN~c
zIm6~&#7{NTfud2gK<}L;0W67>`u**~T-u3w2I^Mq&7=E^082|w>VhG3B=K}W-6{9k
zPTKXpC3J;@V{@93?&a&dESz8ctkYdKFfVwHf3V@CMkbI1Oj$R8{G6aOv>pV-k$fBD
z46^KB-B%yu0-s`SZyitaYF%V-`@NuRPE^V2W0tvz`MGmh+vu0pDIHo2R^P;o?xHq-
zndzd&h|;{Ci&dxUthOGhA@{;b9hgzfg3hWYAvPS$bS090BEAKG-i3A?ka^4)Nfg{{
za#NxX+hxBHb#dCFXwcBqfNx!KN*g>f4BKJ7*;~8a8{eV+u7+&N@~xBsHXwp+I-up!
zGIf+9cggG<@vhkGc8c{Ib6A@RklAK6YU4XTH>ogK$He%}Gy58ylg{8q*>K99#9=Re
ziA=bZ-f$TC(sNv~>vdNZnmdNN+9ay798Y{BUCI8VYC4AG%&$L<2}mEgm6`|6zuKR}
zbarjE!xxkQXAN}(c}rJ}2iM^ESeySFh%nf;uhd{ucnRcFBH|hv4LJSraZCI2VfPWm
z&u=+9zYx=%Zk&qtnqEV(FtIcV-;Bz1XqV)2ZCv$N18XJW$nAfIw4WP&gRZQ{*X9q|
zs7BJ6r&%|<m+CIxr~I3z<0SL@2A0B&^ZIX+Jj;Kw<QdqR|2M`_mpZ2{4ja<-6IF8(
z%p4O*Xe{3^f@~7dvw|{71@OA~e&GFaUfzMD%!v)oF^Pzdu8kuu_8HsOS}z^FH#guj
zk8g*&z3|NdiMYCDYPwvsL-O7oC>3_WNMHcUZ6V!64I9ur(5dH_8IZ!ix_=vY$u>2~
zE}bCGum5M>uV`Qs<fO5I=}1J?xRl;@-%@0Y?6`kH__e(EqKIsSY*~zMB4Bou#N-#J
zgEr0j_sXV*bSzWGzHyH?h#bS`jT@fP+f6{NmA?Ss5aQ_{E8dv{lWY~`CWf;2KC4)L
z5m9O2qsTm+5XT!TL#~=D6#eFbZnh>qDKDr7_^c|ZL^Q7MRAt_?Nr9B#xP*rKpnsqG
zSZ#6K)>VUkS#ywgOt6HD;#rx7F1C7hZ*o2QOA`tRp<f#Nz9<(w+&C{YP(s1At+_i+
z%c<NWR4iIzRHuJ1-=`ZVPwl)*`-gV8UA)6F9Ie%P8&b77-0Gx&QPcjZ<I}M&346xR
zPA>WzwMm5>)%Zq3IM!-ZC|Nc2m9-Tny9PM<8mHcTYm=m0Lc_g9#02|5kP&u%lKNCW
z?<tZDW$<`RAIS*JCVjnZ2+B6OFlODIf0;%jk?+i-5Q~ZQPQ#ClJ6wt+<(Ne06seqM
z%rT_^ou*T8sRN;rfRC(BDj@#wrs5Om#~0Z$cvM=MmI)wpr@K@q1Kw6tfMX2LmbjBq
zGxY%U)Rnq8D{jh0_2ryzR1<*XG#0T1*s&c#i(@oFhssGBM8crb>u`8>c<d+74vrPb
z*#W4>42XoGw5qc^5O$r2cN{4Y(nfM02jlzBCFrP!d_3P)+!4%*vu^xS&ZQvX0_s60
znekB_m_L4?iOw`tUH^<%&?JCIYBDzrx~ZlxhW)RCuQFpJ1N`C-N6PT8VHL9?2qQJe
z@vE!n(%kCac2XY%8W9_%!{%9q949~r<O5NU?IxZr-wq2X0nU9*a4lOM{KUT?gO^HZ
z#z9q}x^%{R)p-WDm0wQ<GClOKx6ZDinU#-3vacZ84q1LO$V3D=nRDCwNxNOcGdCPG
zi}i=tlnVngE>(QC7uNqFpP)*c^YSZ(uajoA1(p=>yd%|#EH#-YHxS72I6*Be|4kNP
zqd>p^Qv8gVLSk<ophlv_5A}_uL;toC{v}Es_|Ds{6$D9WhdKZ=Cd*Bu`>n;J{g2CG
zN)VEKBcU7W2ZWWnH|>Ze#${LS$FS%sV+1z9BBnl8Z{fNVM9fKmc0&C+kT?Jkh0pS4
zb}0HV!oqzV*R5#Z!c}Q;xCeO>VS~=btJxn7j3;DqkaAQ$8kIa>8L#VDaF~2%qUf9#
z<(Sq=rFL^#-ETc2s)mHc!<}nA_je>#2HQh*wjq3t2bnJ%3oz17gaM%b4w-U@0eX;j
zDBYU0CRlqZVq>NFefx_`MnkNdMPZMtZ_QO?rB#AqW(hP2=119{sNt#jcaD}ml9CVW
zyn!&EhXY$M?kKS9J$h6@cjTclepd3ST8tPBs~Ac7r~0@5j@hw7vjofo{K+rhtnszl
zwhu3^3{+6z2^GA`Dw(LilucO6RW*McXF_9WvT=)(#EMazOMs+ps2t9n|2T%psNkKn
zo(`~n9kiF0<nzcJFaXIOoac3|SgmAb1cdydHRMc_WxP)wf%V3_(I`omy!hqU{P~ab
zy1<QK+#e<Ljc@PF3{P_8Dq-?iuDrsFiemPc$C7&ocNCzOzok7j`Q?p1zv(n1{;uuJ
zQXdi|_N|CR@7#H#Zi4Enr|Bg`Fs4or8`Dy{j586FD_Q%eXv}ArxET@*1lA|fbH62#
zxRcm;*m5%0qikFMK_bmaQ%IdUT{vdHkWvh(8y$P5aM_7J>BwV_y~Y^%VhdKeuQch}
zxfIPdE>;9jVbO<Y-E@RHxt%$wu@&xzqJ#?}Mn>2TiTc#dt39mS#QcttbK7mDv91Xn
z9=<K37M%uVrmd(#8{Ti#ke0B*Nbq+Jj}<RnAmcWcbrN~~6OyYh+nizitVKc(rBUqS
znj^d6756?)U6$h5KLOvOSF9cgV>bm<(w~NazRf6;N4j_6nZm`@MKmkx5uDGc<}pHn
z&pw=zuwFhPO}ci?Syh<K$877DFvj`waZ3FvZ&u&gXcWD?i|MRz^zrIYmi3%Yowmdq
z;1IzII9wm0PltJ9y=TdSF<$5_?I4Abf}FiMyv{j@fV+<m<0=)uE}C-<g3v9#ugMyC
z0S-8t>tDw5)ZMl-neAa4EjY}W&~7tzut-rA@pAx`2X543g3$1l9sy#v91T5-!+(33
zr#K3-<fC3igD|byhM&2;hL1||JZY0>=mp~NAnzgdETdx@U|liC`4@JS@7V#_?45q8
zDOl6KPPPoof|i=?<^L`*qOyYWGUpV*sBv55fm4X$$NMZ6j8ho`=|v}}qGfvk%~>83
z#=p~JB-{P<-^%Zg8}2_Q;*6~SWBJvPjK$_a=>AY$#5WU8Dnm!Y*w`|MT-6NdhCil<
zBI^n@s$=q8?de(kkXN?ZP`>MPNv8HL5{od3A1ZAL|H*`w{Xq<phbQW*`2SkCc!54e
zl-cPum#Y{9qNmaK_~iI}j(5yIdl14u)RK6|Xm5V3h(ScAo&u7>XXn8H!-~HyI~r*e
znY(f9MEiPATD~{{bpW0wMk1q#b@=!Blna@YhRd-4fjKvmCGH4zl}fdz&hQoC7$XFw
z8>2ACy?&1v-&@MHzb29qHJAqS&NB66n6o#Rlo$I4=5+RyvH-b&D1R4FEt+iy6`Kol
zH4;-TQ^s1wbd*~<+U0QN)uZt@0^7xaxaJlB6UA|5eC*4E=t@@;+H;%$))(Zm(2@)J
z4Jn~F9|4WqKv3HAoz;!s9aSa{$g<Z!?m@}@^2rQ|uNfQ1A*VXfp!!L%v}5GXY6J-p
zFw)*H6(N|BhXeUz`D#~R8y1#zsT>`q?w&RnTuMUu=yhDv>%(^6RL5C`X~sYV+ig{B
zl)c`|t~vdh2duzcs@*TSf94(EO(nZ$Ms^LD!IS6KrsJ(A@5QtY{1=vyUu@*M%Z7v0
z5sxO1S(2PZIPPRti<i6$p_JgKy_}{vxd%68DN17=jnH@;7sKr<b5G!MuDaWeZQeH1
z4l;pKE?+PFvjAOl*Hl&J5o>ie@3gJ-FgilEp3ZhU&J7*$xs<=rj`lWBQ??c`lK+`;
zcu~_<DoTB{hhg0zc%*G>wbgd7oW_-&54U!YwyB%6A2K$bq~{bOr#njDWM&l7l({e@
zCZ#BAPgbvQP#=Y2s+;AujOE}8N}u5%Xe1=6tU{KaS8QL1lGz(so(-f8E;%W9dmt}_
z4&|Ueio-Z`F%0`95*I(gFbH*!rsxXI0OPL?aCHg?-H#zO;qxIMglo7#F+hns=hG&f
zk<|L=V7&ReeSRKoEHgPv>-a>GU?{)z%f{|I;y0ay>PwH1e%yh}RemvVaYw^5L^*a7
zTb_1)m!gABEGdwVrvkXe@u+k(XFzb!F^tktn!Jq=cpC*Ie_zB`=GdN#rcfbf3zMSO
zua%tRGJ5uI&P+pD=L(Kzv(!w(u)6x;YntQL2cS~#NqfGbhhz3p{XmWaqkxGyyp_YI
zQf?T%{%y`yc%_zg!&sMLyqacKBpzXGD8KftSt#hHOu`%$M|{F~XuvS)ka4k6r;?}=
z>JhwqdTo&aH2SS8YI)*hYkHl(&URfxRDD3Y1{)6Dd$Pa|Q3q~pMq|pM!Hw7A&-=u#
z`?+Z2VIF=e;g#7se;&lV(pLNOfjcfg#X7WZu@*DHEh#8;l2mg~$pZ|`XCw~&Fo;sm
zW7o?2tdIzZ6iavn+Dla8EkW02Wu-aZ+~ZE+lZuV7BtE6iN%&EV(<^>=OC%x)R^o2S
zs50ytl?&A71-M$1%@rT;s~XrWv&p}BBbU;RW)nMG=E<dQ{B@j`x(e1h9Y4t`o63Tc
zj2T#kb!e9AvH)&TBoDRf+oA{^5AV){__+U5;h?U22;-g9LzeO44wGo!<%ClA&Gx3R
zA?WA~rH1UER-BE?^Y7&9gSY#Wf6vn6qei!nT*3~8o%1uQrJL1=ADZ)X6g1}EN<5vF
zUUUQwfHUV|Z6=|kDcUU!i4r1ERDX4g$a{&Qu~TTa9QvMhrwE(uLfx=0KVe7qUWhx{
zsZ6PF8;t6EB0sO;D|zui>5KimzIpPu@mzOd@5AX;;qZL)Wi$@nu8TxvoU!kft2%-e
z5jaGyx?U)YkV+lXnoR>@>`g8D@s*gG$!LNHU3{^8b3v?Ok4s1mGru;;XXFkQ7|t`A
zNZcE_M9V4_sbZUpaExWVS;zMy9j$w=erC!R%mauJ@PobJm$UnQcnM9lsFSR1C0cam
z?gFSN^|2an7VvLFn=#bqQPBBg1xitOYiywwmFaEf@^NJFE1nTtiy)5yxIDl<ARl0$
zel~xI3semM^yYHOnl}jB0Z(Ck&Z>uW>B;gHhGQ`Ac-@KOfhydhG`s{GPvi9FZ|dW4
zLYhBC%$L!#X?;7!Xc0dJW^gEz#nx5eGvZIRam9EDbU}5}IDw#uPF)qmJd>2X`Vv>K
z(pCW3gau5f|4{OQs)NAJBZf`t?&TL7P1}yAo^#|lSe<LUKn0)YSTv9%JYHRM=q|-^
zK8gH^DNrbqqe!CDpv$i#50!PLl%z39$?Z{zG(36s7Na#5q;iHur_WaX938DkLz+qL
zJxEZPr#y*8Q8-3BuiXMecb51AV17ms+?higSzh%RGzWz0Y=am-)ty&AUQ7KUmHku~
z^E$TTQJ-F2lTv|}VZ1J{iHF``t%;v}Ple@PlTWc%YW$oFR#p!g8>qT2dt|tXX`imP
z;mnifWqwZ-%l0ZRTR!cSesztfkAiWf)**BW9zl0gk&$m`DVgay&b(^E-EAm^V4)X0
zv7XiUP}I9>BrrBF)6;TxhGfHxYshEH!ESx*C1ih%Qtf#7+(3d-#ft~86N`P*>%=Z+
zA?WOOq>SeXCY}$0fikv(^LE%Gp2kzCcGD)kNy*9!n9`<?ejH^J>?nG6@{SJAu?@0k
znr3t905evi5Ls+`4#?|Fg|pnWrHzL|wd87gT!?mRqA|(}m<=mhIM_Kj1Mosg{McEG
zGEc5DN_*8cyEIXE{IAruOHE;qhZgn_3E#=Q6%%mrpgnh)gnc}c6ktNYw=PuJ%(~Dw
zaFb}^3lm#XmpvPGJvSkz_o7PX<*!v=y{+Y!YF9W<BYapQx}25@yJ{-1L3HPoWiKxO
zO0=J=T5+ok=rmNfItr^!{Z^`<%5D-x5YcQ)efIcy&OF4lD@-N#Ikj^e1cUoNu>58W
zg*Oj7sL}*kWG~6C6e;$kGhch4<sL@DYj#wF=m+>*4|(m7OMNtAn1Lq|1A&b@Fd=Cw
z=zbeIQo-jlZ1bNv{)&z4(|y>c3zj8?F0Jyilzp?x(l1W0*t~B@!y%u!|JEY0{U?Q)
zfrFX;f10%pX^h#Bu)%eGP@O|o{t}jl)nnFcxdL{W=xre=B}iPrVf^-mYFg?q*RjAn
z$*Z<W7E;)J6l%KuI&*q?+WWEpOmMDGSMcphzua|!`uyw=V|M*>&gJI-wKWoZ9ZOyn
zeem2rWf46(_3+Xw`-htd3=g=Om7UuIb~Etq@b>RCp-;|)YhriL%G>p$S(ASn1-*3M
zqKvQ4zvsM)THS4c6Tkc;LMn;#Hf4$_H2~A5HyWUEY&LV$VjMGw6V_HBtxy`z`rKoH
znLAv>;coN(rE}6wy+*UeL3CtQ(@{kAkwB5;3=kwtlcAcRA<%zGRQ+@FnX-(h7oS)+
zE=<mW+1Uh<zXp3@C$D}&>P)6kMhd?$mR8xbPzXIX{x;<~FMrV_`{UEcSvDyL_o@@f
z)&!8%h<i-mf5dO)kP}O5OE*beTNY*>Y18pH*`?d+L~>EXeAx3ImuyAqYKk%6z6tI9
zT{$MDxz|J0JBELM8>Uv6E6h?}rfP~in&#o2Yx&y@Q^i(QtgDG}K^)5~TNteK?oP-3
zxy^c;a+LelDSQ-g;VrDT@Sa5ePw}Xp2FoL7asrFW-J$#L>#Y1m74y=Y!vKdWx`iQ6
zY{07PI)Uz)%kbeVg&{H8LUu`+cUhG~q`qm@fLCj=hww;US{nW2ci!J?*~DPeC%)hN
z2B+|j^?ziC+?0HCEs<%uXu?e*7zfV{ZlLQ{&XKd$5)f|>Dqf$S?{06bu8Il+ptf{*
z@#bM8^Ww50naRa*+jD(gckq2ujViR8=DA9};%K|s&fZj7CdxRpnhsjCiX`-2NL+-Y
zVdx+CG@Rp9RaGtdNOrQ;Pf)6kqf<$Sj;W_u*fCMG*GNo1M)zeshV>7!Kuk%D3(;|L
z-87x?4$ZR0`fzck-#1G`Je52ob5Zy(8w9c|nP(P(ImNsf7sO{;HYjr$wcFAWJYXrY
z#Hs&=r6`SXY@MNa4HDm4H_0pV)YHPa=mk40BOMAE6oG4)*$`p4i^3)l(3Gk+j_1X%
zYu;QAV-3+i@^z_|5|zdrJPJsKT`R#B0?#+p&H}E>E!bZW@zd{$OtT<0qTIJlw!)g0
z8BBhaOBUcg2ciLwNuSYJH?KUngZu-xFq``8U0ud<p8zsND<hwO-O`Foe>9uFGqG^K
zK#UPBSp(?VrU`<Fsmg}tUkT2m62OV0i8karMA-BO8;Tx-rna4QA<GTx_hHWO$2c_W
zGA;#0ZOXmVMyPJ4lb|`jiXWOzK#)u2avp<C=^d*IEO|-{qlfQD`=?U7CK)@w2}B9&
z0y_~FhxI<u3V-gYeXy1)Z5afg-}k}Vp*PK+Qn87(IP{(+hfUywfs5-<gKek;IKpEN
z&r0=9D->@A)JgM#N^3ZHwcbJwH{`hyWS}J3=7}@BC(afbEA4%f7mfiIvdtVME)kg1
zW1L;enu;*PjS<n8L395?-TKTehrF1}%1fDv&)UTGa*`bui#)lUqKerkO0EMV2r4o0
zZUn*;Q{tO~(It@>{4jlrQZJS~M6OWcdv@lbI=?aEvQ{)P;3@Ct3D8)nkl@K3lc?0U
zzkfx~^HX|4);5zVf9cU<vvCWVZ|6PMZpGgKyJo4xzgwJrgSOIO=)|fAFhCNRSMUdi
zNs7G#ombML+Q%$x(E;zC78tBU+N=ib%K*`qX@m5m`b{iZLH3JfWsL3^AJ71$CC;pd
z+CkWV;v%*9tNockKGy4D0Gk?;&GI``S!gEBu)K-xh^54J3>G1dlBnDlF{H`@mpRu7
zQ0K{}49r!;Vqu^>h^3B(iGU1H=YNApLoMWmp2WIAZI^{;P%Po>yG&Z`DxR~RtwX~P
z{>Pa|3}O5R#6(Jge4ZjuOz_&?AqRlc>5VCd0#GNg*w(vNt+c{MZ1`i<oa$DZBC46z
zj1aTv#Qq+66XKA;A=l%U**rmhVy`>=jak&JbX<2B_b>kK=i@QsDyF4RFzQIQF)qHi
z^&*?G0I;Y23Jh*Tp|L#@!lIQ-aDk4?5Et&E38j$<Arv!*yNp1M%FD_g;-RG@Psm7j
z|M)vn>w>W`Bf||p5+%!J6Jb5>E1D6ANhcC=-$Pi=618ZmVQuI|*ma+jBw2S2Ci_GM
zH+AN7<6bSrG<fl_Uo@{#UU@XtINpAPkuqHy<+*Sa+ceO1C>_kYCfU^{p7O2n^au7s
z8d-fJlfWh}j0|~=a?U@2kdo0Q(8z{bT>$&xkr}Tpi`4~B+mZmYFf86vbD(Tk$EeB2
zN$7`NkE;Y>nYz(=SC!-bhtg@-V?*1f>_WzuV^PrhwaRTMY^9>nuQ7J%lkz1MX_c2f
zJ;YMg`-~%HZ+$6O6Q^5MOs{NLK&#xb^4I()5n-fppL>X1?QlK1!3FyTNL!S+jhgDX
zeP}Rx$tBt{G8M9|`Wv49mW2KbzFJ>igcu|m5ZF-m5Ci{}h`m(Qwmm1!+j>RILh9>w
ztgq&pfI>TBh@e0un(6YU4h|T-7pp7P!^KxYelFP87R$f|`jWoV0XA)={rkzd786G-
z?Wm}7@2@;A)h~Taf#s8ppL|Rk^)$BsTH*gw1Yuy~_}@g3Ese3516HJ;3jYW?wWlxy
zZiZ32<>I%=cuO-17@agnNdM+7YE?|?l&~jrj4LiFTUVPftb4n4-W{OJ{B(afE7GH%
zETDH__r(P3+I^&Ahqu6qLXx6l$m)Ub>2aiG(e`Zjb1Xh(e6%`?*<d53o<fSFt@CCg
zFcWI4Qn@+9F8Uzkk^J_O*txjEwgY69@-c&V{r316d0T;(?RPkC^nGmksyT3fcqu{0
zg=gF~d41Rz?g1iAt547YUD!m}`NrzD^0^p8Gs)@>=N%2b(G9gZbga)gFygLe*&p=?
zNEFwfU1vDb6~5?$JP}(5F^4w-zg0mZrs2)5Oe==US_bxvFd~adpopI$kc(LqyUYIq
zyAvfQ@yV}b`K|GsM2bJA{1XeC^Xt=!B|Z>hRF>Tw5F=i)LDpq1OAnXi-kW2@B6t{#
z6=aIi8Y9B9qyNC5Dn4Eixrsz&+Dj7NrN=Rml_mF}($8^`-E5oj@p^|<UU>3wyDo8-
z#^h1}d}H+pBy5bSwxEb|%OxBag*+6rKLQew!2xrULo{6-Db#0@LYk%vLGHPExz5;b
zlh!rKnt{p0aK<!YJyyI_UeY*vnjyFNkfAnWQjqk)6_>=%vY6ZJ!90IljPoGDM1Vgn
zBE@jS6jQ^;Oru1K&nq@#_A?#c-{HyeN-BtPOwO%vc(f7pJ7b?hC*3UlNfg6YjQcR<
z^?3xpvbYV+2n8f)V~=&0E05eaEm$HKagd6~*72A~2Sq;B`mnsD7!^~^^lB^gX&UId
zZS2oBD^O6b8s%>*<Ehh?LutTk5DD%{7_*CT&8qcfMi!wa2u8tthe?-2K1@swAw%nx
z)j$Qg!UkPH4WjZRxeF6{w+MNlhR=$Z@JEyMLs|z_tV01(&0$qZ=3OP3O$#{|U18yf
zqqH~=fpQI^AbZF{bEf1loN|7P2CYwEoSQiqP&!&Cg1Mm=wNL(0vf${|jNwR(?&(L_
zBqhBzp>Vq(Az+@<{WE|ATq5%gc1f!)ZBLR^{87_M;)X$Pnd|I$?v8S-cx=4PhxkCl
z(dY?1DhfzHs3jV`p{Rzs=NxC_^I`iCCdgeho<2tp{1<Z}5Vs@W0eaXmM9{OM$MJt|
z1G@K&nBe!3y|aTqd`XN2XJUyQ_ziYgN)F@()XSX-Yqkb3*)(GH-2_ERm9l4f@)Op{
z1ADhiH!C!0lJB;e7=xh$`wq&*Og%0I`o5)p?E$|reym&R^`wvKG=RG?jmKbCw<b-3
zX{<pr7+l{nx?0wc4uQsb;MqL%ZpC&hz{LRtwB>;2@a2oqSrSciFV8Ny6=K_d+Nf@|
z5=}i$8!Fe{w3KW1#;(k2qe;u7397<3OoTBKDTlSb#u+Rg|B8(UfE16eqSsgYgmg=G
zW5@zIBw)hqqzr&pxMgLrXNb7ykWBqbM`A_+NQd?<jG|N>u`#iCmL_cPkIFEbP-Qby
z%IDOpPy6^Egf-jTvBo|RG(tkdzw6<)O>~6xv#7j}X?$>}Q?`9JJ6BXr;fWVH9~EX#
zEV8b+ko2pYP~domo6xI33Ed;Gll`@j)48`6*MtGBF8#Vh5JrkxUI>JO1{00;7{e1L
zG`0q)Y!*79sQ#RiY%k^8*HT|7WDSudQjJ0qyhxK&v{D@nvQ^Qk@09Muktry7I-N|5
zEFDayB{(E=DF5^du4$JgT^OYb(&a@2ixG;`m~-b}<K2v&46+oGHf8amO6%9)$fI63
z*%qKM;;XNREhj}W_os=Eq<OR@;iM(CbL+vU%>5-w^1>j6qDp_@l6{F?mzKafD`{bz
zlL!}sHJX=Djt&U<`jSU*NwqkA;ZQ(_t8uF0KP5|uEf=I?8Q-sQT@&b5#ZKz-x3Q>#
zcDAeByGICNPTK41Y7LyT*Iw;AE!&#wQ%f-(yGt);M}E1eiMX-AXflAmq|mAKmsJzQ
z3IX{V$RG|jS^dY^`(1b}7So75_p+Q-vfXR2-T#Z|???`>S=TQWd6PO``z}KI87W7g
zx5AZ7a^}T=^z*}Pn6gL*S5~lu-7`A?VAIN~lzTqwtD@~IIEW_65N%m%6?f?EYJUMc
zHO3=#+QPs1COBubzNV`{)J>5c_Q1oU?#c5_SB9Ms*0h9?OS?9xadP=!ebWo>!dR8v
z!}%a>n5Gvr8&b}OY+3Mx{9mNOIb^AgNP6xjvBI>YWGRue8^ep6-$};ubI-on#&>}D
z@Ss$#>B1PJt!jnpS!}R7AzYqIh@L9boZ8oN7iadunKLox;q_f7SAO8))77pTdq|8+
zeAiX9u*L&>v98qwplb<qskyI)b7x*!BIRn2ojKZx&c}-92pl#-wmaK4q4eXf^9@os
zr%LX>6;t;A{9R^b`k{mUFCM$09}pNZ8*=xB>LWgVXzV^LQcx2vC!fvlSG2=g3?man
zrpW9|{coQLBh)byMVpCv8Mwwp>G&suy9!ZFT3xiZe_n2G9A0j$wQj#E?>*e#Iw}6~
z@0TX9pI+`KBTttd9c|oi$V*(^di25;F+<SBzWn56oK2VcF`>*B%edm9#H6`79E({}
z`|zfCW?$!;cX4j{FH(}hh$*f;dcS`jr3RAd7l+YqmDJpwogLkN(nr`BqfGbTpKTC+
z0*OHXehfGawzqiuv2;J|9f&nECbpJ%skQxkGsYD0A-xa|U;c^8=%~+0FjG*=itI1S
zjd<E62hdo+ma*ia5~%aTYcPg-tzbMrF;;8w?QgHbr>eE81;4cFE~DG1l-8lPAkOeh
zAHDDkb>&;dV}TWv!3x(3*D!bQqfMj<+4cd9t4rNygkgYxAd92z)gphio1*(S*d>a5
zBvvF!+xx-&S6ch6jQadV6Gq&PDklrduYQ2{S5=}XjFAoe7*4o6_{TIV=@V4(@{hzr
zZJ&0#Z%!vgMA3}a$)a>EmBJ?h@ip~<x;lJw{&5zD*zP8|ooFf-_RmtL4^C|^3{YIZ
zu&<8mG5!X}n@~;=zj0kdjz%8QWAU3?CX~8juVIi!L1Pc;RbvPQM{9mvzJBC6qMg4I
z_@V20v7}qe#rKWc^qy$Tg=z#MqMl<M56k(3G{s7}v)o@$&5q1g-Kmwm`kWpdq$AL8
zy4EMKH~`t1r))!77Z~4|Gk?DKTH4Q?(Q;TM-{<2JY@fguY}lA3iV!To3T&uHK5SP5
zwh7EAPJ$dfiYkvDqbLxeUM#Y>;V}@$O3dC$zy4_%%VN{d!Pv={*Wy<{Tk+UU?NOoG
zEQ$JaeZE<p?mOw{HRHF7=yf$oE6(q3!h1XAxpf4Y4enn&vSL^fY%1Ka2wP_Ok_<}x
zxjt`-9mD2s{7pFcMS>P{$(vT>G`hkSF3Ql&8r8`xU{xfqq6cNHkAvylu(i~tc)C!t
zM8EXPN-giVJU?YYe~a-bU`CijMcqm7SJW6gUeSr3(T%_1b2u8H2w#MJ_4GT(lq~;6
zLa;V0wOQoHV=!L)4)5H}WUP28P6e~T7Y@{ljKeVXr<rd_q&$%)8xAKE`-lO?d}xR=
z^$}%3F$ZyQU=C|X!1fq|BZ;0(Qei7~fck+h=2Xt2FUfl2=%1w?ke)>bf`)^Vq1)Ct
z3{Ed520M#`a^~lhzWeKK6?0Vv|B)|5fVeE17!<A%>F3QymjpD_+v6}UAM8%>P0MLK
zCxoGbVG=%f;iO^5#8O#CG#d5_IYl17?Amc@KRAmi{ysxBJna>mS*g$<&bRF)n_o9`
z_$ug^|2}+Tz5l(80_)@#zj%6hDj8CX*f>PBDw#286B~&!qAIxc?*OR!Vi-g=I&L@*
zxd|Ab#gfW2oepv;19a=~^GRhVOc4`J1I6R{jMjW+gNNb?wg9}l`Cd#F`053eQpB!D
zW<G}KWveLu^qf4ivFw0<Spt*9lKuH*A=V>=`9a17y7QGkLo@Ze40TmyTt&h51p{gQ
zrs&jF9u_ZpdqyClcNKEgwnh7F-WDAZ$+QSaGlT^Xv|GnrgEhgA<iM9aWi@qIa^0+l
zX|xuaY|e!WJD~>2&U$>dYyoitMTMgF<0F=3*5aqpW!6jI@*s?D`-a@SP9MtOs1HKn
z>^bm>k8z*rQ)e970BZlgp7uNcTHU_AQgt{KJ<jX&zZZ|QUzmwEhX<N|%{Qa?dym&}
zO(j*7q`3scB>rEFy;G2A;j*n+wzbN(ZM$lfZJVoX+qP}nwr$(CZk_0kz7c))!+xDF
z^I;<Y%#q_8%+HlJ^?3r%f>|^7OB&)qd%2+6q9wc%szUDA8LGsxH@EFnu;PCg-aSz7
zrcr=D?(p5PU;c9TXb|Y1lbf6;b5zP1s76=NloNre?Qr&Mf<P_G+_>4sS-Jz+_ar*+
zjIVSVM}NL8Njxo;&Ge01Jd6r9Pn6zmPHay5@>ajCM9ZIQf)c+hsl3RaR@-_N6BT8v
zFpTz8!=I=H;Mv9G6o_tev|ZZPjdU+R;2I?8W@VjS<T@iEJr}#J-~$@}JsQ%lV$NQc
zgS(f;$y(7QTUXoD;04oqTo#YOt>V0)8LHruk#u649wjSCZ1C!<;5^qK<-C-;PU&~Q
zec9S~Q*{3@oH*~5b?a=C==I{GJ-yg+u!LM|{=o{U@KyhW+CsL5-K3;&De4lqFN{iI
z(J-!fSw8rb(&RJ$Lf~Plj1oFdHlDSe91cCMo`UMv-xB+&Ww5>p9Bxx^u4+pvlh4$7
znEEv+Lz`l4i+)SaBun}(oAv*UR1=zQGjG|Q1wvb(d7Tn3wJJ<&d;~qJwDCC0-cLAT
z-E~P5g373|Jcr_k`>J2Wm?4(=&K}sT5n_^W*NQO9ZJQ&)Hu%Uzh;tBeZ{Fi9@W)}1
zr>w$%N%V=b($HOpN7}xLh2K~O2vZSSm44xE<5N@w!Lxu%+lI7SbV%T<a;8nx?2m7A
zrTX*M?NOQPKb)S#Z2iE$V*Vyk@1n<b6u+dQ{l*T%Y7s5M<cUTyRFJz7OttUa5-(mC
zG~T?CnEM@!a5k@u(;ZY>(*eCcO}DK8aoxo-0tp}3Jzzr=Jp6!r>;4<@(YbFfAWKER
zyze1_J;2PoxNoD+rLPn1;qrW)NtyY1eU#-UnMnC5Muj*M)I=S)kHU=#9T#r$EHVw@
zC|W*cAe0Os2L&XU-Urc91Lvgy?={lFz-phXF=vK?$aQLA2S>;uWel(hH5NzNAiw)q
zhgT`1jG%TzCE_8WY-Huxch^b6U0C9KP||v;Il6V@@T`1}1p8rJ6=!2ALCWov?~-s6
z=kg1Y0?Z(C-+W;z=*qXIBn5@b19wEgwZj*{v8x!r7cZfQC!i_r6;f{0|BM6#(~gjU
zFceu=045aA=v0VPk1*f(DzR;Zt4ND1|I|L+q_752YuOZm0p=uC;9#iLhcaf)Aa!``
z<iqf=BCE*W<~^YF#Gv{>ZY~Z3^3>vd3(t3vUWl=6j#lVT{yE#Kn?`p?SMD4V<WQ=}
z1awW8PPpK@W{3zBD_`XNyf?k2>ICkgM6DLa{rHo!X^4+)eX32Lc+OM}RcRhqrq_Ja
zE_!mCRA#@ObD$nUtZLa(b$wjpT>;bN@h`k+rak-sna&D}_o4q-ehB&}i_cPnf5!*V
zCB&unKc36}H~Zi>p!L6#F_tu@euo^t0j-L!prTX?h2FjbzGIh@KwVA@E%_z+9VI-D
zJ3n%g$)7|Vby=+qr+pHY;&>F{g?#eMTQ4#$y13If@HyKsPBLy_AzL3UaEI_TIk`o1
z_9p;|EVErdr{Avgm{nS)ZQ_=OhxYdy6KpxSu&!r-iFE8doIvIbdqz`=xC8fJ^FPC{
zA2cy7C1Uds1cVTzqdvQ`KHWdhxefJACd+Y1Z;Df?vgjo7`0?U;e|*HIw5*EqiH5q#
z{~Yqpr|7(YuqU@ghINs{9FWIQ7#^T!-5juMy6>NKi%1bWQ&wR!g1y|wZ~oq40%v$c
z#T(;#lf8v_;&CPS;UU*HbE<8Ub1aZramI>Z6%sZ$S#C_cS-p8cSW+ub))qI}Gjx=q
zK)Orem?cKzncG0=3jcMJD*%f1UFh>&VXLc~?{mR5EL965___uU{tlioH&V>b@@TNa
z=X=Y`<F!k43a%t6ncxfj;`-#J+Ttyph}<kKq&e9QO(56Ueo-AvB-EHNk%y?t?HMpI
zt;E!$Ikb=jtxc2~SDe=4Tay)4gl9rb!Dsp22}yC8vQ~|#Kqs1C9rGgh+8UFmP@q$I
zf_jhTqS0nv*~1_<+p&=6BoxV|<4Q~bkfmMu8QEMW*#8Xww+WfnC=R_$G+w|@Z=$a6
zH1Jc-1*55kqpMoqIGeLTH*|`jnT-_;Xrj&?NAp1FBD}#IoM|CKqI6?t3GFoFA+GQ*
z3|@b5e7v&4Q!l?G8q02yQf+=^a2*;fuxfj31v_(#z4=Q!ZHZ)0w_&=USatBQ&+_AP
z1c{u}V_x}~uvZ29G8uq+I;%d`co`xrjowv!0#JA(dZL#kvXlD&^sa!Ud%$Kj#8!!6
zzuHWOg<poBEf*EsLU2qs&Z5$$cODm{9@{~if|Ukw!AL`OBgCU%v!Dh3W48u8!{l|C
zR>Ne9`6|1|f6$|~D$;q;CimQ!Cwo&dp#AL=%|K5vS~&vuRacv&@3>6<=6wH_ZR*dP
zmsMN!>f<hmV(qv(dGtt_vXIfLHRDIYKshNDWvyT27Wj9KwOZ#OjrdIfKAUgdkqjCb
zIk5<n7K23QsZDO1w{=~l^;*O4@N9{^V=hac^-ur%pArr7>#KHs$B$jiEgjk52T}r>
zcLzwn1@Y#qt4Kv=<-Uz{yWzm$`^F!L@iBB$8@NJo)P29i20@*?OywK+{LIQIsOXr+
z6YRK|wSkP%P1_ZNvbI2v(o!}kEi(k$g8WdVhh_oIl6)?vW!wx5Yt#nKrZqU>FlY78
z-{7&xQDZb~Lpg;|b14U|kwU#hu#T{dEj&`<4hx)5KYay)WO{Qv1hx<fl09GgGQ=C`
z1So4BDQ8}8TgU{Kj4SvAD2A&bm4QOmDI9%+gYaPKj7Z>q;pPTcGl5b;>}*oH1SC`E
zpg5Epc#IpAG|?XqLDCcwi$N?Q(m%MAjk+I3ez`FH<x|f1CMP+-{$02X`{dZ^@?DU(
zJyV$V|DaOgnE}mfDy^+|tJi%EL`7CsccU8#{`w{GcuHV7J6QXB{V{-4-Y8lRysI~+
zXC{IV;veN@y<Yk_KrJq*n7cO{Wf223kMMA^Rb8wV<=|m_OzA`<dttUQSa^zen3^!^
zkg;ihDmjkLx2-z>>uk+5DsAUGvKQECKcY9Dwd2MNuPWoF=b?6=*4Qv_<|g2k9tZ1M
zJQ?Gj@q$P*`YD#GRh?iQpl^bidm=(JObU{#iK8eI$i_j^=5el*LC}g?#KzNBkK6Va
z%5g(9D;v~DPb`R|57sBsiO!Nr6XK{&cU`^QwbZG-cPk2hV>A4$#6QaE&9p<J^CHn(
zELjna83}LeY+$D{OboI6bI1L=Yayz-f->+q9KhZ4kH77z)b3Py+3cD~gTNz(;com$
zS|U{XVudcnlB<7OEo2>YiTarOL=cB87mA-nLD`T~(3B_)=decOMBnND+3*LOm{<v+
zT7zIm!w^VzytR4QC@|0MJr*+oP)^K#6=U1)sR-j*(#d@JCBWkEX@FqQNW3Qcb9xh1
zSqm-|XAKam85V2VNk{0PaIJs?rgP6%T-denUzo*LhhSx!TK!s<3W197fPhw@OsHq4
z(=7S$ze1<u&=scZK@Ac*_Jt3hcX5aGrpf2b+{%}xK_i;2hFRh}Pl5X&qJ2+c?SgLX
z6tc!s*nfpnC!r+aHn-dm{-Y0<UQ5y7*qcBJYM#5%V9*ZPw@vY5CN+LdlNO`$B)}#e
zGUywJrOX8D6WCU7K^36IO*iQ91v`?8EH{aH=KYN+)?RTbg9m12@@$!LTjSc9uaVzL
zbVh!mkJqCMbx?Qw&k6)Jw<+Z>;0)-q0-(q6CiF>*To_V8%~5@W2Jlxt>M(Ih1H(2h
zD1!)~V{e#u)vDF+8WvHnZ`UcU7K49lw4Uh9mb!~dL?)uzY*Ui^sK$;eEE}?H6NM7i
z2Q0c7?zW(Rw=?Z53pX$dq*%LGo_&0Pr#{Xy`co?1cmJBp+{Zs;fY^A+R(&Uno%Ce%
zCEEXo8qQAtA2_5;tpD9?W$AxrEC27dW~c_r69|wY=QVNfc6;200erx)^Gu${|Ff-G
zNL;Q|ZvPodmI}MOY9fJxhx8U#pWQ9%?!L{_dCAM2m*pQ0=bxt2EzJ+l`~@y8d{YdN
z%iaHY@9_T7wLH9V;iscwDeMmVn{BXlb%Q_{1)!;0+XIaj>TiB?*FA0b=uMsj>*HPi
z@bx`y0UXAGAe(C%`~2hjwjracFE-{`8@~K!4586ESciE-G>+7V$gS#VMhaFcNOF$*
z7c?9>qeM;jx4uniW-N`BHo0z@=?C<w>>gzspURY$I8U1DEWLA*&Jys6qQ<TiiHXhR
zWAT_RE5ihuqGr!WcwWi+VEmPP#%Uc%_f_e$DBd*XZsXzh>dyG(V?1tJ1hpxv`+M^a
zI3kxG)7>g*#qRweXQ5^5B+hyZMZ-K@@tcF^>&^?4T=$$-Hr=`HdAn_h7F=3$M`jxX
zH8ZZuh%Zibf1<p~Z^-8J7DYdyV98NubJT9rF1>UAG$>}tB>5=GM?DJk=x@spuzYU}
zmG2fWjiaQeZ&EO^Kks4_j?6<(=d=HO4b5|XuAt!Futl9ow6$?^2PocO?oq1tpvV*U
zo5B<PGW1Ovo|eN=F0PwfIj4?1GDA5uRZi>&qkM7f14+XyB0+Vo+KJrn2}?Tit@U7;
z6eoInO}Bgfqo(0L;-r4b?B<H?Yt0mYN!sn%7GSZLe@~GA<|+~u9xzU*zTClX+l>Y|
z$Nx6co(%FVo7M=cPiMQ)MirdV>^DC+hT+Qi$s|DQ$X}ucP;31-A8%9O=qG%j6RNTB
zRJNktQ3ORof(Y$zNxvPugbnk^y3tRKplm(vP$ZaxeBFiJChrsQIP$tL8@}xA?DTxy
zKdH(wU-5A##>X<kDP(#l*X+g<mU(4*EBTiIoF)uIqI|t1lZ*Q;&myF8DR#?nSprGy
z8pwY485v+#F?@Yy0tn$!Rya12LOEC)sXvOzqO2V&*V{;=d=G1#Ykf;T7^C@2wL9+^
zlTW={uX|-U=Ttalz3|S)qQhn}aat2-bQo*r9->0XIisN=$*I_lp_3&`_&tv>*w@$*
zVM6^z5_L%_`4LillvJ~E<tlJa1qpWs3L%$w$2<}xqc@UBZd64#?y{}B19WQUVYZ^X
zeSbcgX1!(=^SImu&EcRIu3nXJjd(Yuse*B!tTpF);?Nd=(uY8O(dTfg8((C?!kD$q
zD3Zy@Usf+%KD-^{@5JxOXc|iKHK3AP#M^2$^vC&&CeCK_tU#qVin-9~r|#O+vYeTQ
zSMPGLhv~vZ3}BzS8G#)#!0Pn{!D;wdal4D_kjF!89>0QankfOEyls@DQyA(<lUi5T
z6zCBU#CUBMbfwh2C0rH7Gx{%V3w|$Fy)2sOrJN5kn$%qjmcViaRrgYLPIPO(uaI)4
zB{C47jA|@-qt*LtZSyx}HlV6tLv4kM0LKhhmr^>c1vBz)FfvzXO(xaX8QSY-z8S0r
z7&2&;by`TTB(wO!gYGD4HLNcdmI^{e2F$m16+x9;9r&T(A{QsUu~HsQ!-N!*FWYtF
z`r9p8zg3B%s#->edMIYlFw>$>n0+VWq!~3p?5I}ll2ZolHnSpH`}(889wEpSW+D(l
zp=ikWaN{ktTQ8ZROard%jtO-Pwu=E}>DqO-GX4RCC3*I(Vq7UKA+15&J$c`sp`MH?
zQ4)&AZxqTsMI-SoXqC+bi@rd17r_8;Ys6*8@+yoZB63F;I}k10omWUmzI}a;m%I#X
zG<+l>K^1CK-~pmbRMnR2&eVT6D*t2ixMr_12Z+;l1mrp4E{g<bkN81Mum`($G@s?i
ziHM{tVfCnjHK?pQ?TtE4<t>6gJy!kMj`J*&g(yr}_&Kh-!W?eYGw0D*C1sK^4V%YQ
z-lb?JQ>&ZuTK{t})Yoo{+KM7NE&4BThhKrj4sXD2L}PkG8dWyw!NTd;RH!s4$0jk{
z=ukq_BOADTn%w#p*UOlD{%^hC2bXVG*auc1UG0zVg1uVR{LL|=c6icg@^h=xjqfyh
z>Ic>Q+bL{jPf2-u6p)-XH~44r9qPa$$Hjp6xid)IZN#wpcvsU488MAa{fX1E<__=F
zZX*gQR_1b<RQ8~>k~1{5Avcmj{~S`uH8eW{$^{;LBbgw8fzsHZ-I1j&ft3At?D_yr
z;NI)`_7M_LpPYAxafie1D{Dz)nlt=Mgpr3G_&O#kb)=xqyx+6D-GJ5MzppMZi)iDs
zO6OCNGf7qqiO4IYU^>AGRL#}SO3kHj5)|xr=!qj#y+p~58e6ZOV4G@A_gp`@<fnkr
z>K<(`_KtGq=q?*k6TFV|w=z)1mf5Zq{7Dk~iK04?_KFxToH%|8vrflHCD0ESj_cuh
zq0XbJrQW+5w@q*Yw;<}osxZsHu|y+$_JcZQThTwt;3#rI)p7Q#3(rZJT&6eO=F2t{
z4Xamdz;4*qHnAn7EsKY>Bb%;QG0aN9k2H<#cbZgJlO3-wqaR^)c{s4oxNbXSMTXd_
z$Ia<1<XdzL0#{CxNF5XE7n)fUdYk^%Y9Q77D;%!YfqO9y0}H%5c%{A;;-)ICfs;37
z)$TF;pkWJCoAh$9*tT;4cS>w-O!mCX0cCvQ=%PPsJm__%yHLnG72rgyy4CL1@TByo
zI(A%J0ej^;xU|m|3^Je%^=st;_JCrI;|ORN>~MVTt&zXkzrm;@vMUaiZT5$*#PAy5
z6#lwbR9hRm_LaGP8ICQHg57|8dl?T5K=+V-Nwzv4ZgYrMzRTris=RJO=Djc1xK}@4
z@VvdE;W*coJ4RMcXVq$m0Z_X&->9RogtWRBlT(G7Mztg(J*11cft~y4X}ii0`a@?l
z2wV!&S%9n*7Y|cyYdaXeP~+~sR=tkCZJr0ksyNzxI%Oo#ftHVZ*=>`t9>?P^k$BDa
zC@i4(!8-G}k4_5j74I`*e$I{IaC7|@H-7aGPl?*q6X;lml~oay-KiYYBw7rA;6ysV
z*x6={!CD$BC^R1m9$CR1oJYHO(=teEw&@;>v5NePo>+Y);dpbP6lNH)HSjgQ2(t-)
zx8Iu+Jc_=jNwRGSG$;9bwdyC9IjDVdrf_N*ZfB<6e*kx%({=vCX^fSF@&9^fl-&Lq
z)5#iGC^}lx$>1~4)Bn;W?HwHPIoKKgKhh*K(|;E|wWP6OgWZPc^`d&qFb4c<qOs}i
z@4e<+2N4Gtcpe}P#>KwDr#BFXc7tK}$x*yKi#trKv$5u2g(+>%NHM58U!<^aPc4aW
zXD`wK@qbn-3hFK65UR~~Z$Mzbdo_w5ET}VQ#N4v9KMT8UNuJMkEJH7O#EpVMYImeb
zEo?%IFP1cvgO@PG72~2k$Px<ST~AU{7SoNqYzhP!lC+f_6$-6}n$oPoR0tS^D|5G4
zq$HkPb&sey-Dw@{&`?Y`!|5)MbV*Dfibow+G=By`A}dE)P?1>VPV5(6)Fcnu3p=O~
z<{E~&T|$zG!3}gyxL1Qt6Va8Dpk)xN>~1xT26M`@c;sh`KNXNG^{6BYjz{AWMTV0O
zQ)pTJQ<Pnss|N07uuA8yBrbX<gVeIjAl5R-%nAtn5AOp1k_;aDX;nS?j?=w#$Q3;t
z$Ar+VZ##q%%KRMyzVrZx8vF`f@m~OQqJiZ6MQSBkB-Yhxe5jy5vN$?U4fgyjI_V%=
z=fHEs9C`o&>_H!{8D22bU;4MyBmz6Ld=mY$IyKj78B%0Pwd2@Z8FtczWF<pyWzVKO
zU-rT}E2+Y3P=`_yDzKv2ev2rapoq0YU{_ljZr!>x*2^$s4Lk?^e-{IBgw(4-Tw)>r
zJUTC8G4;U$<1k_bW`P?FUjE)A!R4XPM+|GJg|$FP2>F!c)rK&4Zg})(4;uYbOW6aI
zo7A5n<NK_2f7;nusW`Fy?$Y`4^lnzqvE~(4Lk)b7Sc-TGd5HOAyMkA|8ma1|`}2G~
zC3}*iBL>~4v>K`S*uy_##^UsOcX{cF8M^;=wm?<`uEU>wYk*|sZP>xa2#wi)CNsdW
ze||LLW4(W7do%r=pu(4>``O%$o}Rd{q^shi{q@-{15dYX%T}{D)An-Pl*~05%0C6l
zMX|_;%BB64O8+^rM0Zj%tvWN0R|?oDY1kMv;#_g=;~DM)ui$Sn1IxutHTQQ4WQC(V
zu!WyFXkS5+p?7c~skAn5fCE9Vt^Witn8V+H&7sUoHf&rl2v7huJ7N}WdSaj{RNYVC
z_s=Ap7I>fuFsJ>B)+(MvMENR!u%#s@YcYedp9-vo{sx`|bYZxZii<|9Of!2qC}W{h
zVXP#NuisfX+6;m=5m%4e*L#eQsQWaU8pI&)$LsgkS)59^3U+7~*7x>iJe0ARV4GDM
z0OrjMozD07@fBy3oNTZEl!=#J^7`ZJiEM_hj?K2!a+wUB3;<K=K_KCk?cF{|s;=e1
z8K6m7RPmcEfXJw_v47msL&QT1-AcyqqneRXsaeipEm3{mV?tS&=M$?Fe|Hh9+r{TX
zao8eE+@O9XJ}o>N<i~~4$>nT2&BfuE$JnF=&jX}3)*H<ZPzy=IR#zTS&S)BjcBSY{
zm~xi%Z8~BkAJQkWF#ZQcp%W<H3t$BKNO;~ROk2gmWzPpGM~&oIF&hmx2sOJPfE>FR
z5%B4^z@rsm6JJ-Kg|;iiv3hx6;g<#KFNR4`D{-qg9=Fc`Dt9TP<3|qP&mh*e4D4&y
z)CWgjd6dh_)=~-9`{xbQ*A5ho7dX;V4#;1#PGBPh7ygP@ZvGiUzV^j~e~g<am)If_
zlFlM4h!vU{+6&DU%?QmAGVyLQfHWEVvV&zDyK-H!x?2{8H+P}qp@Gl%-xsK$WL#<S
zJJevWs5s(zOfik2(2G#HNY<Xwt|-iS1erEJGzc>#Ml9jjCVxCLx~5Wy_}n-+5LU<e
zqGVZU<@s=g95c{heyi^i6R<@9wurnx5bKLGfO%6d+OoWHF1Z<TBLo%2n!M4pq@JRl
zzq=Bo{w+*RbUzYms)Me7j$eUjl{$t)K5Nx@Uuga`_&J}#HgpPq5kM!Tl^8Y=_P$iU
zhRqZldvQUcWvQzrvKVT>yK&epInltRhPcz@_DzWhzhf7ySm(9{^kJ5L#Zw00oAzG1
z6@0bCHSo$8lvWGP?>mrMMbE^S9+zoCeAB_kRJZ_hw)ltKnk5u{RZj@;&k6XyD8yPL
z^2f(I>_*^{gI4|J{(|PD0*zKeahxWU@%7lExq+l#0vM5Ki)1ZWuvZ@n@v4hx@daq<
z8noIIBdYVwY`*AM@Znr8Xgn07GT76`0uW~_714s9D%EqawcppjUM1H)707Fn7It->
z{6W?X(?(k8R6Tyl%ngQcU|%o?TrsDn0Q!rS?5qObd|_=Sc^oW&3ex@BA@8zz!l(=s
zN)>rM`ubNS9RR`rY&gbk7HZ#n$PYQ-)+oFRVE`Y3Rj*~B03fV)d75scuu<?rwnYe4
z%*$vfe*AE+C{Me~f4fwOKi3{<g%@6okSh~Xf=nqlTmrb+$PfUf^f3qk41|5uBNQ!<
z?3WumE)`JsCfCR==Z)eZ?8d-h*O3+&8z8wN5V?|QevD6XiTuh$pxu~KNGO=}<g4EF
z=NfP1aDLaiTnZ5xv6{0*mQ|!TF#%V$PU6j(7Y)~KyboSly49_`GmfFAoP*G_1pDzs
zw1E0;xz5+6s{SBINCcZd0EiS_{=gi5umEmufUDL`#u0yqz{q;fF~tb)?eaIIpW)yA
z;^5#UMC1ob=Ly9w$&*3A<oerrstFJwcrZ>s{D{iwQVF;-j=*+%=67o^e73`>;FTbl
z^$Mdj8!zMKb+-+U(ZtYee;EB-XqKx*^6(qdWlu)QSm4?`zx{#coIXVM^Vf=*>JliR
zZ39@s-(@y<4>t3xdP;}QHb!wyOl5Sd<F7Fn>hTurLsGjp0phS~?|wS?7u#*Yd*GX;
zQ__)}$6(Cs0SfigII7VCA!EUEkWs$K4MZ>nGw{G=Vma+U^M(8c%i$HYW7R-iskWj^
z=EHwhYK13Y)$EJ<NRkJy0=(H?q}%bd+-r6f+F<Wyy<kAR#+Q#eAcZ#Ig}S8C%jcXT
z)}*hEKdzUu<WU!13@g8cXc9KCE|3~JwG5U&kzumD{qJTKDGK#!;q&9=my3^CUS5Mx
zAlUN_d+tFf1h%3RPtLEYy9Y5?$G5|mCMD%sBI{x(<++-pN)0q4u~Lb>RDRO8NinT^
zAff{hPHa^{cjo#kFLmwPwg&ntwZ6~Aykx?8J<?}HYG&Bfy9cG09UmEKt2GM$_EnEG
zn7TJQwpDQz+|!mO?SqRe?5}_R!>I3&uCVqiXS10bmX+wg;{fVyXrXw&>=igweQQ&E
zcuhkVx85YqV<zPnWXO&_u&L_*0f3VW|4&wFcKZJ<pfNJBGyYeLO-q^*F@|Ia-7ht+
zKDiiB<Y;5{MjabON6DqO@-p(ICBHtPpICoz?&Ows9-^b&qMST(i?8zBx7(NAkQ%y`
zm=rpfGOz(^f5d^27Qj<*nVa6Tyy3q8u}Qzan!bJww#`1a&;JiNybH#w<uU9;6*5`W
zH=K^W2NU#BxV<5{gdP0sfz|uZ_Xp4P%8BX_D7&nm8PFG4$^#x=Pv`9@wW-WU&3D1q
z#n-{nPD-+q@K~bLK+aF+D_$Z9X%#xZS+F4d|7;3(eXhq(McAb?braxq|L`!&dPmGV
zWEPGziB=YZX9|~tRpBIjm&m)HM>!O3QLZ-;QjpVumO8&3+2|+d?-znwQN*XRL*)xI
zD*D}vM^Vg}x<g;8%7%VPsat|+G$w-9yp9({8LabjX*he%0@owUbOgSPE;Ng?9hJc-
z&T}hfvxsD_g?RdC6|*?GYQ|Dbfp6Uj-Yb5-uB$N8SO2Qa3C^}=l%s&1KtMz15)yOr
zFtyDE;*w%)k~ntrzWt?$TpgMO*bc=7$5qBM@1Ne@OWcETE<MBgZV&}&3?_=W;N%)c
zkLQgQyaID|+nrbr9Ky2_CR(zHC(8TO2jxr{^Tq_<xhgdlI7vd`f}u07h1bAZlHs#R
zPsJ9I*sx5Upy5ji@%h0BADRdKW?C(yF8a-$I;8Ui7sUMuZ+W_$U7*qBWJM2!vm`8v
zM$hWVkNXK}AF`1?9%cDsCO7-!S2@ve`)|W9uN(OvQD>}^MRv?xEu};s{9u{Hgs}vz
zlzP+ROI#HpfZ;ygB65+DMl=5G4Z9}|aGZUO@v@VgutQ{~S;6n3%jRwr4byf^ViyiU
z?>iYBb-j_jIcCS9(HLK?Su=gBbS{knhLRA~C$l{MBrECu3d7fr^(U$|95v~M@#NnK
zPjn!!N#LJOaOe*KGnjX(JT^q1;d7G-2d7@l{dC@eW3rMmQe#OuSRf6`tK?#}{3doE
z`GJvkPNd?ZPC`s^x~!xGZfVx0c{MJLydoZd<aZe0@DFvh?*ku+trj9q<8xLiD8OUJ
z8pZg=8!s`W9{(jSyyk|V#!$|lxdx0%Xq+wIOMHd_AjRiY*gef_B@RcEAsUv|;1y0T
zvVLwo>|Xq>=wg`VE|OV*!BgDV*k0A5fOVpIIVp-|hc}#Ham^P@eL$6|(jp?M`Iv63
zG#EZk$4CGtbK5zvC=Y$@Ym_dUy2L)9RjHa3(<azRB5G967GYO;{<JZCCkY2)cK8<y
zOE6kfo57MUg9c3(d9Fbm-#_brS1t$gCm&i;rg-$=MAI;TZd7PefSZIf=uMZAEQOuy
zU%V=1B-L;!Kh_}&ht(coQH4a2qHx0;5A7VeRebL!6zsdqXOSJHSmQK*l%C3=USXo`
zcB<r4!PrbGUo;HTY}uJ&o6qlq#5BLzS4!nHQVbb`q-TpO4~1`bX@jVB*A%>%STzZN
zhO}BD_Yxts`dfzlHmZf_S8|n9#}%Ex0>4^F>+9x3Kmd}eu++p->O}WP$yIGZVQpN~
zAEd0kNh~SDikV|*gCgLF-E)G-0Q*`gZLw=Y*<e$bzX3%UU@JFJA4~zEjTXfPAC#$a
z&)qR4N~0vzULYrqv`r~pLe}Yk*up+~n84ArAhW-qssF&sT|EuK_~o&sW+FO|L>ae!
z<E{>OI=)F<CMkr0#>v-7+wO0UNlkV6+{|2(OdhEwp^Eob?>r%PSlhp(coCq^y&K4y
z$}vHWdp25^@qFB=BkCMdaSDSJSP2>lic4A&WYw|amwk%uD5~2hIqNdDv*vp1SPzXL
z*Qu^_k%T8uT)!L>FL<gXh!NClYULt}1{)*~{tL|m8KvwgP~D)wz(f49+?F!Vj?E`p
zT@WRy4tE>&lC?PAJ^-3cX>WWl`yC~E*u0?2w1$ARri!T7#L5~lVhTeg0RQvormeqy
zY1hKPF=8vyC{#uHAyZnuhVP<!Z<SZ2-4ktEr4&GoTjg2y+E3c$79A{i?SK!UOJGx^
z&`mhv2`v3ILN0U2qh7wC#){)e^SB!UAY+Na&I;7#DKM?F${zbe@*8oF6k7qqiqf<N
zLsqMrF$6M9!gKS1E!NO^FfpAO>KUVkMZ>QD7>+JJufC`m2&OcZH+$`iGdP^2dm4xK
zkU&)vQ=Juy!-)>E(pd+~S`E>pN<v}rfx_)pp~uhX_SAfEAe!{gQe{6~<@T`;l<ZkP
z2qlo~5f0Xr-k&iF>;0s${hVwxX`5@q)@i=H=-=pd5M}8177ZCnvpqrTev>rUVg%48
zHl!a@MN*L(LZ>eWX(tIOCmH9d86sAr)X9;}7wo6(!1eOV<wKoK=oueHN$Er?U>Ri{
zcwo5FUCeSMbR>yy{+aX`C()I3*HLm&7!)5hV^YBG(UqJW?#U_)bAjBMd1&*VcF@%)
zz$Yu1bJ0I=69@>Q%ql2c2;jGYCP~;cU_9SdbEDRo&N8?U-R1bNhcJZAhGTr`0V>KK
z-#Zr<CK%{!sUap%s9b%1e9BnzBNyac85^n2ZR+!Ptx`AQ@TbjPjEAkrSCNnO@{}i8
z*q?Hv%(^spdx%!NiXF5_s}Tq4kMGK;DaCAOO^7!y^K*He3}uK6afmpn$~MJsqOTyy
zVS3b0XxG0BQvp>NJ(>W<Rt?Y!Au2f8b7c=v%+a4kbrvsb9^<O=p~G=y<iYF@ZtFPm
zxYKO2jPHWpHz=wjN=9t>)tl!HD7S7M7{^W2*`rR11345nilRKM7!JoPbjgxdx}N_?
zlY7<8YV_a$Bs$$Kh|$6(Wr@*FRUVt^TPUAg@}5x*kauiSjmbT5Dud}qo70SkKscor
zsqD0nTQgO|4ati`olqzhmj9|vHb5Zp{=yzP85MHWUy&c1zR8d^O(k0(mgKy5ye>j+
z2fQv!?_^siU}b5-Pf~8nC;<+kEU%BZU*{%8WSccM6yDd~PD0invxj0@w@2N7%vg<`
zcKzvcRlQp}+9h_hk5RjmFGzX3GyZ@up4FZW@;uB_{F4AUPWw5ZmD7S@87UHQqVcDh
z!8M4eG`}Cp$cBdIb9Avc>IF(uuZ|0%l7Zj1bB6+u@CH=bzIW5NvY6BcieS&T5b(SY
zKO1c%XgYN52x8trs@U?um@Mu%PMAOQ$u_vx{12lzc834W!)9P+{O^R(tzTjE*TMCR
zu{nhg+=DTF1porBYtYfN)jm{tRC0W247vW{b9_Z0k+;uWb)35rC>D|*__ZSEl%4K<
z<c7?23|w5_8JYbXXJW_J>4C*(YxML_mY4_tmRs(<eK&mllCW&F$2ladii(>UbD<-;
z!Gm#C2UMc{_0IvNG3eV)DY^w4b06QJ_vDN}IREft5(q;QA|K*nKl^lgubt&rs`L%l
z*qQyjKem5(td!dTBtbHNz8U_pjRiF3h8yNIMhwpJf!%TLX)?Glu?jvrIC;rSd()VG
zz&|y&O#5@{hbC@b`wPLcHGTx4pDJYu3nEroF#pkW8Vt?gx<$I#4NadB$NK%EGpWSa
z!HOorYZ{H{_}AKEB%E{E!)qj*^<(ciEPL~TIv8svUgMuks?pZ!!^c}|$N7^*)G8Sh
zM8^8ryBAm6*97guBP+aXL9N4h;i#ta%&J-%O|XCVUGx!EqtPQ>m5jWeTMx(Q1!>Mt
z;Fs6>ha%NpU7_BmMuk*h^cz!!$)4BtOAQx+QaZAujOP^qTa!jPxb%zDWZGGoG`9g1
zxPvlgj0v<;M_}A2al6@#q|y+eCH)c~7VM<{9NXEWi@XTuN;_dDI@xUM)3yb>+e^8W
z{Npzsb99s>Vpzij&M<9mb%QcJQ!PaBFbB!Td~p#839w&&{^e}fNn3fj!W-0IyYg5Z
zCapBx4ptpRr+*ZLsJQJg?O}y=TIj@w<v=l_pyKewe@CQoF#axCtxBe3W?iVLa=GDN
zN~Y{LrCv#-oa{4jv1B;P`l#^<%NYrE5S>eZAp}58=Pcadx<OR>w;5t8;1>aMfhiE#
z;fSjS2yBy2HIWl2FDo@hYxSqDOHE&T0yp4?NQR8x*4Bn8nOF%_g7k&z832upvraB{
z&W|fX>S}8)KRjv~!Ah5F6*Yl^7NZTj3E_BguNYg5@VDGdY4|z(*vUwdl7sy49S;WT
z5QOLkBkSrDuq#s)nN#a&T6Zi8Gh$#r$z<0&x65B}Hq4QdHEyr8wy!uMvwuBbsV-oJ
zWsM0CeF1TpIlYl{yt*1;$cEo$Yp<#$Ch?5@BdTi|NuH;3tDHOfcC21Qd~&s3;5<Ne
zbCH?hn&lkGO|?*!N{Wk^@%lMMTDPb>i3t+jhb;jH>#V;{;lf$v6!GM)!}C{yiezZy
z)e<>D{jj)jbfi>W<r~d~eko03Foe@cjVJhNoZL&Kx<F+HVwtAC|4D_7tZdx$N<dU8
zBdTO#`**C9GMyAxW#5n>k7?M&e!-qiM)vHHy5^Lc8VHzTRbfeIo~1$pnTWxkR)(~U
zro85D@ED;b_B`)Vcs+VFm3mXLR>KmalA3D@YNRPe%JX-L19dCg6cv%pNyp^6nJn^>
zRI6%BIj4Qi_)s<2%w?>)uw_gYyknokM{l#XVt7J92mEiCgf~s^`24}7han(`6<Tb=
zGRT`(Sa%Zl^GZxNGN7n2bK{M$zkG6kkt$)zhm1SPz4L{@K)L0KIfzac^OT#OI0<Q!
zuR@9qJD4!BadC^8(z%kt^mfrefvCO)PfWUIn5yX=(&^`>H#ZapUKgMhgy37TP#%}!
zFYN8{AC8|5Nw>Q2vV2Fo4{$v7=ir+QLEDtiHYp&!Ibw0kpH#C^o^|)|N<w|^^iXiP
zB}HbWaEC_HmKfp-Zsjx_Fe5D2>{IV*>wKNOYuCxlhifGw*00z54uNi_CVUzcT)~dN
zVip8J;?8Poa~YCqDk`>Xc4Cx!hBlSYT1KokqRSrCNO}s0wuQQX$toLJw7x<>L63}M
zd;+9>S;Mtf&Nsw6llc-iv+#8Pje6#{UIq747si$hoITDk-mw*&ZEt1z&LeKNwOT?B
zh{bM2cA6yieuEDdIY9rXKm4zD^8W{wF|e}ycinfFhC~e(8_e$?Udoncwj{JrrI-Hp
zauL+s^43N~RwO|YZt90G5tD{#esUF8j4ouYF?OHu9@(116XWpqWAD@X^&bg~+8(nS
z_BJnkFHZaoe%st>9}t8PeWw@BH>YG+a1E2Lq$FWCbq|;K`2w`x0b=evXvv=nU(L}R
z7@P&{shPi?Be&kt4^w9%5+HcL6pmfq*3573j|2P#gQ8o~&S0^Rr;Uxh9-$6LA%)z*
zy_}km**pjk4(!A~aAvR1H@9!Mrhl7m=b5EiY;}W+j}#TMfmO1ZMK2`X@YM?}=!c=9
zwB_CO{54+lMd`;D_N6CAD=qX*@TzsT4+|^hjF`xt<Rzyplq+?H&Q%1q<m|SzI^wPV
zb>*HCkxUJ%j>_Vdfx!`xbsxH&*9?bvvzAMNdGWVFzl}l|YtI#K%^v>7vbV8oiXfCq
zWh@fMnXoOU5JNsF)~rpj)bK~u2cdarEr?@qofG*`C~rGnLKldgI}7ks7Bzz5L^t-k
zDh><g8$3v+GNl^Md#Sahy_~NcBKNUX2Zi$zU^88fd6+1P;YcYQ(q)*Q4JJ|O$EWXj
zxrUs2R2?@7ZO1GAs~c0T4$H4<Ont<%iud9)qmSNRHeX&M_lUPKnK_Ls#R&nUK+%n^
zSEbe9OY&;CB^9O-Pg5anqEuWbT4`Kf8QU0sF4A4(a@ooWdkNpKHUYg~mlg%0LAj2%
zOZ3}c5)f@}o2?}^d=^$-xWI1`L|LYFO-@P?e@O)iCrW#yVR=nm?qC|Dn4Za$-XI2F
zXZq8ba&%3W%`mYw&6Dn?Zpir?d18>l^IMC|VH~g&|AxPRCsu~h6l%#fF{3#6)6bL}
z@Y*U(lv6&NT1JVopHN-KG8JGHIN5#{6d`R3${MqwbPToz!B<OPod%4@#cDVy{ry{a
z*Y{PPIupSwuI(wOjxv9NcN#!l0I-1^z=o27VG4eHjgQAUozY$52|F}Fomw?^?|lLd
zsxmv!eLqPN;Lrs?$q^+LO$3@D+L<|Lvo3D9Dc@CB>tRY=8D?xPvTTCJxdZr}j}#AI
z3A>Vg`-jJn#z(e#IyR;APxfQD)(x~<AD!!UP`jrYnWYBhjG%UWG((kTUUgjLv=7WY
z$yMNRHhDqb>nEBU%=}STuK(&At3`oD8mRc;LAa?lD|8fU_ig2177zfiE8OK&W%O9+
zYD;8cc5x{MZ%+lD5bKV9xO<*q*(0FhVuO1+P$Ip|3$uuj9sKrMjVhdDKb1vWGUh?B
z4hN>X`MUb=$WITXamiW0r`|PljJ-mTkiVRSa2GffKT5KxM@@IKDo@Ak&r`p$H&$qV
zaW+*%8Osu*LlDDogM`1D*d$e$lW>Q&m?y!XS?D!nbUx=l3Rp)2?ih*_IQnJ^Cc!4i
zMWd)O{H9h>$PP?MQvxc1u^Dg6chL^TPQF10JIwZW-$9xfM2!@*?JDgdJ2G^lg8LXp
zM=!1Rzk##`lePt5kCnGR67y9ZY?9c>E)CI2DBDrMnMkFd`TNF2_`rN1;JJi*9$=m6
zdgAbgSv~BJ->JjpU;(iD&AS@pz!L<Mf=sDs6`?ZZbOjxc<G7=8ib&UTq!p8lZoGQ@
zdd&7!8oWHYx9L>#TM+Hl>}1@urBRwr(jF_FV~oDr315hTjNWzjYN<|tl=r6w7ct8<
zIbzWLxH@8A-W=G1@0+stSK2(3!7~5kdZ?6WhEts$VxaWj{q-|13+Qe$&X}wJ9>X?S
z59cZ}SFaPlHNP!7M)BE{{WDC4w;PYAD3b-do>~MssoOAMc+FQ6W0dxQo#h-O;o%@z
z;x<UKEH|%8A(rfTY;%x4_ygyjuF?PkgP0_2vv-4d0Fi1z#Q}6GLCHk2dl}LIB~=Gu
z-+dqjHGGeoHU5oVLGNZ4I=5$w4j}6ngdOD!<d0yt%+If;Id^JVKTI;ksmjBYnA&RP
z+frB;QrQZ6im97Ead50qy>~grhwQ4=6w3-!Wed2Re4H1>1_j3JuU4Ur)92cPS$g^N
zP}3t^ShV<#0ynw7tmr+iXycGJc1xb}mN;5ud3d~uf}lMzx+Qms-DY|ILipPZp#khf
zq0X;0?$z}|uA$fOmX;p{M<J~C>E+FFl5!hmMDk-w;#?r(V{o)(^Z7TSxO}pmbJfP5
zz%fVE&k*_eS>$+DzvGosArk$k3dZukmp2(W{>vU^>i_mAtsJQnBey|dqK39exh)_Z
zO{aMf>5sJ(?N@nUZ(ro`ND_bR6I{En_>)<VJMl6R2RcroU0N!x2F?$n@~0U~f|9ld
z6v=LZM@n`9Ft^kRQswhKT+h7Q4z)0~G_?lX6k4gfxi+6nssaa~1kR8$nn(ac@39JK
zk)wyb{`5@Tev$5FOr^J?@X7&51M&)5dwac_4<*#Dz*nfS6QMe6yTs5{(BYtMJAa*{
znvQdAeQiAHk)nKaPFE5@BT1PbNWHM3d4FEKosEp_fB>)a=~LW<O3-v(=<F9P*TSAO
ze`IFk4Ag*Nq=qamV$7Z(K{M_HxL&Fd)3jQu^7Db0njVUe1F1t6XAbrI2>^~Wf`kJz
zKYVBbH7F59=#&UV-lCKeZ{9+DMxx>lDD$2a^}ZKq@;Fs=7%5vv)~W4hDq*@9;>;x~
zgA(Z@3I-W7iF1IJ1Sr;rw`g5pn3dfCB*Ih^gTmr8gmNvxZh~bo70z-6@KY&{=6c%h
zF*%z#f*L_6Of`aM>x%~<<P)m1X}ErVzsBEY`#EWD5@j{<Cnm6Nt*fb}0hRIEHDyR`
zHgeL)xMAHV#=3AGD;+Z=X|nL9ZaD=}%a=WsE^W;v^OMY_N7o7`Tl51-9z<YD(JOY;
zLm6S%v8@6)i567HMmT}OiCp^IK{%H1qxEqCcJ+Xgk&$vBBF#)BBflA4Su6LM9EnVa
zMB0{tKoO2t2K?aj<=MoGRH1-THNted<2oQo(#jE9|JBTQP6$Jh#ML0fO=Hu;e!|@y
zrJ3++pR#4O91sc$$$ljT*2b;=BnDci9b>ENh#$mOt@8Z3p1UFQI<)bqA*l**@@RV^
z)$kKbyNyJBua9#kTU04VHL1;vaesclj-Tc{N99<XfH>o5GmGM<Q=>G4i{w9{lxU%{
z^MEVZWH?A)m5R&s@=7?YE2wemDS5V+n;#Oz*_Zj7*z;5m8d<T|k}(&;Ve?okhIjJj
z=l-gkEW3sWneV=?a2SexiOvJ4JAHK!#b6RXrN(Yh_tsF6Qe>m2-3Z*qZ}2BEKBH_}
z)Wbk08j<ahrd3aD(?6hy7A6fy^yXJj571xavZQzgm#;LymPk)<h%RA|KS98sV<12+
z=(rR(5xOJH9W|*85YR(%$zWzh8h#NC<|!o?L-XipN*c!d?hq$;M;iL2rB=uv4ryq7
z>{00Z?}L!;*T4J=8ly`-RMcV$)C#&AU_|x{u1A%nwkSkcm3+lKdNXK~^8brPMp5^i
zPw^NroJiv#Gn0q1N=X<jDA*8Z8bh8=6h)|^+|NKE%J@`qWF4W$ChO4-b5C=e#0+*-
zYF1Z(Z7dI(MO`8lM9n5?-w3p&R3w9zPF*eY_NPs|$eP$z#oPfkM}cWyx-@3bd4oGo
z&wI^*(QS(O-FC74RwxsIxOWngnX$)3wxT21g0lCt@!grq(+oEw8_GV1RA@$D%LeBC
zvfhkf7iYXqC2*f}^;vO|&X!xig(CgA<mHK1K(HhXF`u+`$BtM?K9<qvv0pXzqs|ru
z0@#nc2C{&Mq6iaBGn_$@Mn%zUhSibf-<DJSX)Oo*x`KG6XB5Lg=TgudEAwoaC1@L8
z|1QhQEHbY6o%sNV(i)G&h-1{71Z#%QIfi~0ee+3-&IpvkJEtbnS9Nh-jRoiJSs;5O
zPeR?~#W{}n=I6eN<cd)L&AYc}O(m{6yUWsX*Ct5on7w;=nrOTMBTQ%0vOMSJQKi+$
zB<H-j$yK)Wjuu#v7{;SD#^B{mC+}Yj;>iB=3*_a#<b8pG5~Gc7Dmw_C)?)pdNp1a_
z2rdyM0){SJ9r4k8;=%4ZNl?sqGuea1yzA%-)0g11npAd5=JN6-X80Jxkk4=8dvW<I
zO&MEhLey=8LBI>FC#3P9q3b}2MLEwbWKyK_FOSZ7xPway<x0rx5-IZfBR7Dz&Bx`N
z<!xokMEv0-zBhKO@nRp>ns{QokPX}o<r_)TkrI1uo8WB1@nA#(w2Uu6cXJr1fug43
zrcuGOR2+PY%!66nqKl|`i3S9MJuemSmSxMGZUD8!W%c4FD>o@C5(fn&w#Y~`iq585
zi7IrjiFQ&RcY_+4)JNDMWElzABVa>?hIc84c$g3*l^i!2i`H%j8H>^`XBiC(ES){4
zGSH1b#dhzfSiI;p{?x5f7LN&@Gse(fN10fKBBc^-rmZcu2B%r6sK<pu`HA909L`?#
z?t3~tVju22rL%*?g=R$@`Zr<O33_S41DbyudmGJPw>_w-)Yu(){ze_=<!9BuSdnMf
zi>@TXFP*wSrCBuoi&u9zM$%^TG7YK@!%@%O(TM`_-?C%SjFQ=npS^EYY$kvmVvA8;
zNOFGLmHBs@&C$9M4el}o>3maPv9I)$7~*2)?q%oPGWd)N96>XEGw9;zeG(hjz4*3!
zmNg=UPYII-zi5UuKm;uvb@y<;hrY5ZWONXgw!nb(4KLMm{I4<{9U54+<8>3<cs)-b
zwo2o`qh`o~5i7daE8vp{FX1Sk{hO}u2Y{Vp|36iAR+j(Wm$Ls-wf|e(1r{5^FYe+J
z4`E>B78G5Oh74{t19&GK7plM+-fpfi<onCplQyM;YAiV)U9HMFIr9$(o~h{Zz(jAz
zOvcy2@jzF8-~SZ#frAJ9TmJ*ae+W^yl?apM4BXvtJzM{1VXJv)_q8ar&~|XH-CLCV
z^#gM~!?u>=0Yu)i<k{S!hrejOQuaQ)vTdC;_Z41y0l9-9wVZqSzT5ZI)u%flbqA_^
z-XB}vn$MKk!kNa7ZXeqr>BN7_FZL)C+2ZZj%%0oBtV2~jdvPKBRLBjTkl0&%)(J=E
zNYtGBK6l@qJZ9vbRFLYJ+hd#qtRbW*awdNtq=CU<hzA)O2hs@RX*Ln}XcGuoYb4>x
zV3ho9gQxHl8A_=x2i8Df9FOT3nfuaQa%x>ZUX_Wvudj8#e$i{96X*4O!8h@Q_!`?Q
zvQC3u%oCb`Tq0+lCr96jVj0q~MNXkPk0fapiyc+A!{h`;%PvXRVb{mj@+|_dJ6=va
zd=wonaD|cdUyfZN3YG@@;{k+7joo^(E1@!MB2!b9P2;ad%;A{Qoxu6)j{4$p`ughY
z)33bTnspp78eecle~sCEuV?Gm?=oLyEiSxlt6#N|ufBWA%9>PmE<t_Xy8QwwO25W}
zDY|gulBEpC7cwqvqHMnn!)ddEX@Uf!xUPh2A|beSBH&$G)qAaFp2`I032AV|Y5=Pb
zLxX{_9NaXlMMoDjv#ALJ*GJ&rHAnX0Ck)_6gTFZNptW`aJ~4gtaVzv-!mvd7l`AVV
zJ8D#ILMenDk%iRORWss{3K3B>BP8GfIg<v$fsJsf<sc%%g@~<PVyaUGS`MCKI2Kq~
z7c;^3A4MZ<ht2Q7KrtcY+b%tEY4mO{47B`<Pv*c({a6nHCXeV9A-MW<yOQ!l&s$K<
z_!K$-yuiXfXg*mE75MRh5nB#N<bTAjXZ$5~{1Z@8?m&v9l8JWB6WHOR&Gu0A?y64~
z&)V2rKyn!Rf6EcSn!%VP0|ypgvN?drD(L+oRGp7a3ppfBFC2G0GXzv_;=#+4=5XEA
z3~MkHe07HV<r|7gAO$%{-QjMe@loJ3^M+(WxR8P=WKshCC`O=FTyRc^mbBU4C6}`9
z14)=wtFJlBOn<MoBvf#5j~xKq{`T-112;@H&QyG~2sOBwa&XFt>r541bbI{IZ5qlN
zg*D65oiE>p(i}tL>k%qRP%JgfPo<M`>)5nsxwU+(La3PZ;hfzo$MtDtFll8<U5PLL
zENb|DS;oaG^c923y&4kySZ&b$P9t~zdRun9K+9XFmVPWLY8{q_Pr|2day<IZgbF=)
zafKr$Ee5D@)4E9Ju^(nmeKgy4+!dD^;nsR0F)00r3r0W+o>=s^4+vrbzxWIIAK17^
z=>WdkhTvVr&PA{ju+8ST+nI?6I$0mRzp!*a!Fj$|0F}2kuF^5==s3)c11q1G=-Wt+
zdf8OS8d_5L=T=#2kg@|{B~XRHwPNQh{7;eC5^GY?_@)fCP-fyHA-#195uG(Ob!IH#
zp{&X*{ffKsC&6NqbPfH6t?MU@C=K73tTO}01Rly*$U;aZh$?)8B}5wc<$A<G_H5ZS
z#6~ZxCb5ZK`!vr3P)$uONF)OaNR+T=sYm`m+(zb0$AadUw8XcR*idDg9`E6JoSFg{
zGdjgQ$3Z5VmE6M5Hutx$UuP^c9(|QFQ#^NTP+^b48TwRplk-CX47djCG}&)%t{B?6
z-HExB!)pJBv3Cm6B;2|%%XWQb+qTtZ+qP}nwr$()vTbzPwtD)+T+BrL=i=PtZAN5d
zWIpd&Yp<Qy1kl9;jqt|Ac1_aVT4)GQ&Lsj{(r7)LJ#+MM3A*Ea?WmB`kQ|PJ$Mu%X
zjzk@e`Xr`8-=Bk7la4Z%0alT^Yl8@G^~@aPARlorFV$;Us>*zt&!qZhV3z$WbTS(r
z^&Q?;toXcLDIA@PdYpw1{Qyw-n#sPRWRbJlk}KIov;A(-R-&X4Zy>&?iT2r9pNa6Z
z{AAjg{J3uTs{|zLuJx}5)DX3`bTNj$_OkJ_W!oIvcjqO2d)?C#XNy>KPeGf3Oln>;
zHr)1FxAl`qvOnLgEel^;kAq1T8eOm$dX8)!6PPe?bZ^*2gMYn!W<sw30W#}XX`_k_
zbWXzdYTVhIK6uBFJj^;e(c6^p34oCvww3PpI)sbsBY~BI>Vg!Lq!9w+HN>`Usaokh
zXYr_MuWWRL@E1AXNaH()NM)tw&sT|<o};JF?Dmz4(Q)PIZw@fIrgo?7cQb6l8A`5?
z$~`ejl)bs^D$AeE7dd+oCX~%yDYELJrt?s?sjj0aZs#(o{#rluvVh+~uH$Uc@n138
zcK2ewuPedg*fXWO*V0hG!shXL!ZON17QE}y#c2T8>sCt((tJT>c;3ROngu>?5k+~o
z+x?L4z3tPlWj%%NI>^42e|;14j}rK;P2zIH>}!5j{ezUJ@33C6<;vhEg0lyK8%lT`
z8Dt>-uDdC|y~1q4UQu_7)+9(%a@T5g^B~0k1^z~DNBeI-W{&?W7&Ei{%W(ZKKj#0;
zN^bwN1aIr-2^gpT!087xRvfNFcuDVTFNhejo)b}JJ<a#;{gE!Av~)jKja?0(jg<Do
z^~0SFPQm3zbn@Wt@#g=!i(vhaXw1gzj~(Ej@I^q1fz77@WQY`YT0cD3u0%b&#?vPu
z^{bCofT!>B5I*P-Tkt*TFqUi|<w-oay)*XY)7MMw&pT5)#~4l$jE|s!%)Xx>fxqX^
zIbm}{ugPfA{!g@t5Q@>nIG#ObrZlVZnZl&6Wo~|_J-kmpj`h1rum;LC`88Vq`-9kW
zJ5&ZRL)$8JVH<g}^`J%(#hJ>7b;3|Lo^=OnnApj9xWXnEn}m=L*D|M%dDuxQLB59@
z9_<W02uuTCXmm)WXw+t;rZN$$R)C`~gAkJh#0XTB0yKiivX$?{th`3fkHVe|)f}lv
zG@dl3KL3pvH5U!w-9K|h@jDhJ$&{vd+Y}8(>P-W^cvT!NgT6SHx-|ZrBjUR@7GnG{
z!f+L_B(!kKDVzGVYFQ4t-pm3h)7O5eVfebz+uMWRC%e{nQ4jMxEWJMb4wMMWMCDG3
z@-qCkA3h&}z-wKm29UH>;;&U^nU~@-9kw8iANfxipO6S(6IwEW3Q;<wUHasWj$fgq
zJTqfu<`XBKAXj^hsG*(!+o%~9Ri+x({1LsUAS1w0wuGw}Tp^+9w9(MCj}7f(pFP?K
z$@82j4hg4S@|2o;fYvK`!g(`Ur5y%EK;n+-m6PJv8+dO~4>h0pWOj~g^$tVVSd_lH
z|6pAJxfnM-NvM_n#t>(rvOH~dAkW<w0)%?5EjWLdSeSzqZYoaj{GN(6E^8*U3BSie
z!9I&kt-E}dD!rViSLphDC`-e;R%0n<#ff=uFcIHI7i6e;;Uk_WMvk(UEbFH@c)paf
z257B?y7vDa8bVU*=EGP8I6e+_^~W)xr3Kcrp!x9VcknnL<u*XLnbe8c6c3NDTW4(F
zRHuRnjlM3{kP1mz8m`2|;>nU|lHmTy<CRC=(-4VlMn0{gTcE-20x;)Ak3eted1y$8
z%8RW>b>LiR*JB>3Vy(YPrL~C^hLc>NyziBN5sqGSeIQzr2ldi@;|_+LcIS|cC$W@H
z)2!JjZR#Z9#f)kR5OwDWmUrFPO3o`!WbhpQ&Y7HKp%WfQa`wfZnk(7w?4x2<5gX%o
zCI^TC-|DA_QqIBHnoCA73)4K3OF{*rpwHw*yF_&uqbw5TOA2>w+NXpRobq&V#6foJ
z)yF{cw7?z+;gj))`04#=>A72-5pxhd3ztJ;s6uFj{bnaZu|^q$yZ=Uu5cANR<vWO~
zmRoSF*K)cg-1cNz+fP+3<Fqjt=0>&ZhA2?Uc{qTB)|;W1YRli$8>bITg|#Ysu15sS
zd-Z)v77$)`6@tHJtRjEt(8!xPwfpcL?WwIfZ(_vTcH(p7Tc4&mS^y`Vt~T&tUVG>r
z5;0PE^0l=Mp-#^~gM1*UVZ|C$cNC{zBDLEs*xmch!VGO-0;Sn@oRZFV_PMqExvgKa
zlEW-1^ljC8pN^f}4o$|%mveYoG$QQFZjL;rNg#tM@S-*3AO*ukm<<wwGTRNexZ4^)
z%Ma7Y2A=2P`{D~Rf*OMw{S5>RJpieQDDd_Ez7)^U@6^HQ^F!0JZ;;eD(EV>0#xYV9
zd@X+8kvc_8S`5sP^pR#OuY|0ZjoPk$7%0_LV}x7loiydsm8oRqCbL+5T6OHctl|y-
zKx}AfJ%g<wa)ld`ILjo`E{K#Jj$Ws)C7RywN?&@__QmU;J4;R<%*>~8f2Lg{a81G_
zRI3Ikx=dqaajXJbbmkt@Qau`52yS$0z4EYRO{JWGX00#_29fw+I<lDnX=G!x5G9ox
z-**_`9Su}6E|Qo?!fcCwg=omkwUL`2Uk_~UpVPH@S>8jPz@b4cSayI*FiW%sBc-Ap
zC;sfJ^i49C{;5n0raDF*HflZwx_X2Z(Kkj`;<Cow<w3uW48pb;lofEWXSFUOvHK$P
z3gKy~D7{fZElSNus`E-2$!fN~H{#Ums`}GmBC0mElM)`>l(Kd&LJpyENzWb?KH(W#
zbOd$ht<aft1s>eZpw>RF)|sJsFc)$U;Ib=jDZhR~A5l%zK7@LV-AUA;aLz3*SDqWa
zoZZxk<WNe0tlVH!G;B9!MUxLZ6Y~vI%HbEW^zIT=9eA`en|OyVj@dg5p)UzCU!CoN
zSngJ89e$Jr^aJ*}8Ie^;&aX-h|JA0XSlIjIM2hy>buOszw(`!MH0bcP3%BjIE=Tke
z$ejrv7v*$y)4^p^<#BdEYVm!9UpnPQQ4<-93dUCwRuwr}W}!e>Th<W_k?RrAlSv8^
zJ7SV~bC}JYOT`9-4kiuISX`Peahe72JS>9Oe2gDj#Fle`TCOx*W(ZuX2MR0eb@eHM
zx8+n6<=?{bK+rtgz((2qN~rr_l1)rOAP2WlId{WBsr$dIA^RE;#zNiFaL7tHxBEkw
zMay>X%fSnS(ubjC-6Nq|;%!6*K!$XBXp9=Omp#3VdO&fWwC_?w$VreUr}2yOvsFQ!
z$=~$eRZDL{m4EQJTJpBo*v7s;&(3|#SxwK0uBXIFull$i;FUHu(x(h(*AY`ufw)HK
zyI<T<JLP@$)2-=pDMlJaJ`;1dZh08>{y^2`_YsTqEO*lFjt`RU1t*$AQdi!}ZmG!+
zb<V2DN1*FcShTJCZIjPOPy9WT!V;Kmv7bxoVwg88=f{ZgcH(ydu{720t``uZ0tWOP
zY?mjHwnFW8;eJW1BEE#49mq{hIOlg<sK1n@Apg$fB!sRIKR>R{`=|KD>}w_=SYH1Q
z^-vLeZf&o=jBF13l-qca$dp$V0QXwUX|LBVo-tWRGB8@(z{ZG^?NRGBS+fVFjtm0*
zK=3fdP>{qCFwRlC;rGu<gqEt{>oZ!PZAfoCl?z&-o4re&mDzQ4d6qdM_O)Oa<;dMW
zEzS9MT!<15xn@STNb3I_rR86(uWXXs-pd6WqJ1vXdrX}S%@fs><JDQaeFs#AaYhHO
zVT!b3Ep}{W-Tno-?d<vU--wFq{~8#~tp8(8_6YVw9S+CUO??JILs95v0A%r;oL;~#
zO3y!_LjDg>b>V(|mZDqX#9m``*F%i56eo5>mPjc2yf@t)f8V_Ict4gHG)`L@9zT3}
zBEJP5_V2G-JQ;-=h(#X1Z#^u5o-XdY{)XjLhhd1GL%98DP#-cx1b#;e?*$K}n<R+w
z8aHO$l|R=EJ1ytk#eEqly#odr0+8#3=jZ!x-^*NOQ(U9Z2};?SrC`ItofApX;Jw!&
zc>5d$BK-o20OT#|j=7us%Y^|wmH2uL^Ps?#Kfz^<iXPBM&RI+l_eN5pHDoepu~g<J
z!<q-AD{ImD6E$D1#s~URlx2`ITK|rKKH+@<*(uPqE_;G83Me=>WXRnNi`OTbV&v0J
zAbNf+921&k6h0V%&4Y<7x2HQPFd;nhC(#UbKFigTd`V%NnAANYcQCI{Vywatq3a?1
z6L6T}2|srm)dE7kH>IV^V6A(MVH&Fj93NOtd@ei&<j)4$Dj#MnJ;l^cPNKH~9@0r^
zM-~&?vc4~7BTcqQXB7WnTIE@Z_e?a?$6*LFbp3tza^v}W8il(oa1x@y@4FR>SOBJ6
zK_RROfsoekr^|oKA7k)H>OATw^RdKRsk!#R#G^H*h|$5M<-SUe)EYJd6&=@RwV{19
z&6b6RTZY*}b8(GLJh@8AC!8~_mxKn!ea?`y*Gl*f-ozw>$t9fDQ(XNd7H(66>0KR(
z=e&#_rxWF#SvE!7ODccO);!cJm`BMLy&tOeOAyfLMn}Lf<ugFNqFOz!<rL@H8H8TK
z`WS?JJ%49GrWmIgcD8e&e^h2|d75f+;y}01XEE)VFo$P0B5@+PtvDqfVcz>_J+1(Y
zzM8RGemPy*-OpPpZeMasH3xhU*6v$~dpkfi#)f*u$Aw4uvzG;mAL+Vc%*NW1D7%oi
zVdO%EKOzXi2MhSin!STM0@iWBEs5NSS`9@~J8hhj=oK`R<V4jZk<jJJ-`cSqwV;oC
zs>CYXEc8+4S~#erqjJWARmJ}ssxTDIu!4M1({Y?dH<4j?@EHwVv{akbMzmZkBRqv%
zk#^1T0`{2cB{nVeqojwdZ*!M1n@G_tf)uxrSMHF0V0<aN$T*#NG-Xk=k9R~L@AkYK
zt0Y}(q=jwH_V=BN7R$8UW<1#eXw=_rkwnAH^CXrd_WPGlBSA87P!~vJI7-$fOo$#a
zIH<*jgp?k;X=a07CNr#p-f`<FF^kCTLempAN57PM{FzeOFpmir@!0VWlU34TNo!(t
zd+RhbUggJ?^md_a=gY*rq<(7qqjuNU?K5QgU(g#)%RZP*y;35xzb~|}mi}x$@!nVS
zy6QY3b$R+kUXf+N(K8MN1}=WXXmL*KRe3lp%v(r;a{r}kV&GcI8@x5&)cuy)wcnp~
zT||RlR!&WJ+mV6q==PXl{ZUIgiMYCVnIibtN6*{Y@(fsQayXUekk<Ab8CZo=O#F#V
z30{cNZ`qSp!+KTTJ+@_Ofk%7Wubml{ZS@K5fmg|l9&;<bv?<LlWnm{P$G)W3z9zlt
z&^(c2)@5m6_p@Y`Ev>Fu=z4^qspTKWsoy2EDy;FwZg&V&)Sp?M;Mm}nHYJzX@@*D(
zx#)4-F(<)S#Tm(s-mlYy59&KVl&a<yz$7_TF6(*SP+xHD<{e~FcAky=>lhi0@VaR)
z`z#=u%XYw9PUGB9z_AluCdY=rT{*LsN~BcHst%iA?iSSCr_yDg*5x$Kn)$qp_?xf9
z4ZP{}y%5uSzPLufbmk7@N^Q-f&#RAj*sPU#W7duQ5wq96Eo9$Qq+k0>FfJ7!wK$2A
zElL|p0qGYGr|dQ8vLlk@pulbQTqyRDG6y~BBvtplyfLw3MY_ohPP1uOnaDXq>rV1J
zqg%G43x^^Jrv;ej;2WX8eI#=^4ESO)g2>yTb+!Q#hbx^-2^V6cKXrOEMNh;{28(4x
zWu>T3U@3W#<#Oy7qHou5%{kixuO*{8i0fEy4e$Zp>S6jSL`j286dD<;xJVnNNR#bX
z)XnD&mefG}VviF)-@i95AHU%Hv^zlw%VfrG70ApiP>p>vBO#9JiPplzpML%R8^xOh
z=r~vJN7IM}RVf&e86AWL2dw%I18S%D{c!4(fd4y4SikQJudApuEUD0as{JVnBxw?>
z*+=#kK~s2*GUIe|ASjWC8|h1CSZcW36tV*aCN!?S%@4kofBCb-215MSKx7Q8j~(y@
ztY=APqrW9qR7o811t1>M4oDgcQ^%Dr9<uJ(biary_S1IDPu~frRkSUMrGb_Y=pEdX
zDY6I{qF#_n&4M#hQsPdEY_I6Vsio^LRXiXzov#6I9WcVjilH#3;Z>F%=3pbD4HA@f
zA=u(LV!Vr@V?#ICB#Fj%WvBxtmgbP`;P?MX8#6RPyH2`+p4F3(8EYMcv|k_?IdFXX
z-kh1UEV)_E?g16$airk9-5{4?xx@}2oJ5osQ97{I&-`>~75t@f^9LqSMcu)dlLa|T
zSXZcBOn8fXJL$a@)6UZ~eV8tQ`@$S)%L%+D36B!TdsC591=UUJZqz_6{|4j3FaX8O
zqTnm9Fn^3o=#+O-j3Zp3U=Ji9ODmOQypY&|Exm9@D(3M)CbWZOkPxP5nc&<&VlyFF
zl3Lcef}F~pC7x{gN}8<1S?o{F^8TM=mj1EQzkjM0)u8PUfwMTN!^V-tws)rrr8#*E
z0j)1vqO#|txW|cpycx-(GNh^fa7mLcPf5gi9?cUMW;xM3j=SV2Sndn`DPm=Q!<0)F
zF}h@_)tzyo-b`sD#2j^0M#4MG-TF>ZmsMi+e2Wc@y}(Uc^IpMSVVOss^c%c{t7s1%
zb@Z!*OCJCYb;Q;<MHvr4{Kq|)s0SreEEO(ms`|=b=wBtXo;uo^J7vuVbk8;lzO7H9
zE-6uGg2OtU4@%;w-BoFuGgegB<Hzl)c&xb%(ouWM>V?)+ncX;5%(YVya$74qRbDXi
z-Cq2R?QM)OCcah|G{>BbE9w#iVY)S%Hp(Qy|1O6qfUU2-ryZxCMtc;bsL5Ge{1=2;
zgU>6*l;+_{0vE3Pj!F`@8d}BF+vT$ydbFtM@CM^HP2o{PIt6~}LhMx_0l?zW)4P)f
zVeFrGsBmkT1QVm9YOZHH1R8}YHb{Vvwi!*fkKIi#N~$duWxp<^MX6(zmpL{EE^Prx
zsU(VGpnR)`a^chZAbXGMxorw;t2TiUbR066J*)|U1sAqwtQD4OY&;Xm)iUEYk&VaR
zwVXbG5U=LGN3O1>rE%3Cm~pg)>A!;D|4rL7v;WiE{wHZodm`>%xkc}q`Vk^vgscNN
zcBn#n2gE0No?|#X(7+NQ=?@mc*Js>*9x07A9z1OUcXfuzFGkEM*rbE0_D}`__c{A+
z_j(xgsT_1QdV770KnBhqG+dlmz5lpgjXmEQKR>@aS>HTI5gqCPezbn>Z_DAmC<xwA
z5(JLx+DPj~UxO9ptYU|cnElcpKUt&M!}uM5JW4`Y!!ysfv8_EfUg#}pGg`)@d@>SA
zATrSKB+Ots?p6!m1X9S3FSYV^Y_bLFz6J7lP>V4ztleI2rj?|}7@U<UV$jex=;gS`
zNu4z%3A3KtN{eM+|D7X-QWK^4FWao-2$eJMgvcl77~8eUGhEP8x$zG5Rnr+F`G#yI
zn!<%L2(8JImGLA#F{fj=B}UX785JB9on8Tp*X1loKhzg0W(470F*6}3MVXOOHI(O0
z%)H<FeH_CUCbfZemgj>3gA`ih+INf-gU~T1+Y`Jr7m_4TpE7v)x5F})Pg7eGK>6rv
z2UGDox3C|ZV`QA$aV-D&$aaUA{Y&R)b2dZJCK<pyo23af9CrKre|}7*>w^y){w=KY
z0dA4}v-3dmodPEknl|<Ngp<*T5Dt6dxL^pZLrBpmg0F$yN6V9d-f6wo;#aOVHxQUI
zV?dVSl*VlaA*-AgNV&TD`__9)qlB9FfV2toDY~Mh@d<V8RPia9|N3?tP5is<s*k=~
zUb1=<)3`C(S;^9;Xc>s576xL-2w=yWolVL(L-wjkvs+}GV_|!$-Oa2crNAjNMjK|=
z3a~Ss?JUi7w-eUN6t<%-a=oT_nx8?_j!LvD<MV1T!aN)QCAw(0?C||_EROvOFYarT
zm-z2PJ0h5EY9GK7pDd<D^lr%ke_82$vs0Flz?{S&M$aYGy1Rlb=$4-*mcv=^o4Z~f
zol4};cY$L1<k_qkB$&^k3Qd6|m?v_Po9e&|MW&px1gI9yfkQ@8;)N9qP^z#oE}8LV
zvKw1qn=qqZmN(x7pt7(P5on1<rOf~0Oh*+v_tlY9BC!SnScIK9R?s-fmoxeoP04Br
zs2@w#W42eRhxN!*VZ|s;GSU0WGg0(1?%CK!ibNe~zo@CZnc1(n74?2wYr0Ew8SasW
zUXdYCY5d|wjfj8^{YY}Yb5{7G0;X%uEQ4j6wkVMKcK~3K#<)7?(#)ExO`tn%1@#`h
z01=chj&{>+K|EJU_e5o5QPQdxE6i~6ElJ@km<&-Vno)0Hjw$JT5NYuU!G@}dHt8E8
zI>`MdDM$p<Ql87XxBW#tJk1V_rP<!ldjtJhkq&mM$S;)henJg`UG#@hfq4P~TY%^y
zWI%8n{a<;#W~(Au>)8r-D`hQkZ$cjhA@_QQ7=wFh1$1eE4atVf^$xY)0uET;L8TT1
zb6ie*;~Y#7haawO>+h<{)FTTsZC<R8*R&*~ZB1#(&9uSIFQM~tl+ed%99syjp25C2
z8^RHJdrGi{L*rLi<Hg(Hu0c+6UIorV3}h=wS~{%4Ac3DwpB>+e$}-GwU>Z`kk7Cc5
zrRub~%BW}DnM53sjV@?v%zNA*|5yw0RYF*T2M{&W30DW%I%cKcF?L8bt^j8#ny}X_
z6>w7XAJpIat*lgHnEx@TO(#xO8^{u9K6b4VKye(nxH3<rlg@jj<}Ogx)dG8k^TF&e
zsz{)!NbCOH4dcvbMzzY>%wO0!&XY@lgzh}^WnK{;gA7-%w@!)phpq3FqF`x`4e-yC
z)(E}I<4`QI8h{5ps{|_%C5?sbS>JhOr)#8ZBE>JmuLk3bjDxxsgX7E`r&zHH)A4gd
zQDDeiET~AF3Isk^mBK?$><r+<WYj&JY=07pFv~i9=9D+lgyg&lZ`p5U3Z0u=t;f4L
zL#XPK)SH1s%Q;{-|AiB~TYLM38TbsP-n@opx=jWOA+cF~7COHrd<IWs27Q5>172WK
zzPA51C;wu*p)$q11ys?dFo}FiI3%|GD1>%rytPPseQio5gL*H!)jyC$E#92EVy6VI
z;Y3oS5YS*Gh!*#v<PTE;6uB|QkEC1!y_;bN=CLfrpo^h+nHN;#+gHcZ58g8$mVVti
z8*C!RoeT1QeV<4Ig>qmrkjeG<O(?-NRici_8W4Aoi$1T&(~3C(*0t-tE;#o(T!#Fl
z)w2VWGC~`<&@qt|q)vzdVyvCIk;M1yHAki<>aY2tBF19+oH^~vff4(zn31f5p9+LF
zq_2vIlq#^EBzIEKE2&~iz*t(L97)%5^r7GS2}D*~#^L#DwvLrU_M5zMC%Gb|IQN!c
z5!x~%?WDYfI)@Y0wUGe1Y861B@z*V7Zn<CgHecnhlr#_a4aHhN-Wt#4UfgEsGsAsn
zAw(RSwE(D}`^ctZPq~wfs;&He+@vBG#4G|pJ|(Ua{br5qOu)v+zeDe~64$q2x7A$j
zQ~Y8F=<p02ak<+OP9Xi_AB_zUcVQ4Zf$N9o)r6s?9Pr0_kJN-qn;+tdFfvb{c;&0&
zJ|^SWr1ww9HzTPT>}QFN^04!l2su=I{hb*NUk;OeU9<C%5PYK~=(@iT2$6(L`ATe6
zl5gyNPm<FUnpHS@9sr{OC=BCvEv+wZ!J6e<dWS!K_~f%dO#nUF&Jy~rDBh6x^w_*o
z9-e|M=>{78fg#A%sy4FuW2SG@m5wzTr`UbN+L4$kFnE<lEoRXITw%WbZCM+30>^IW
zbiCU0tlIYN;fZ>gJ_8$u9+-@|uJ7rNh#@v8Sc)rX&B<3PjZkoTl%HMv%sFCA+p<*u
z!l{1QW-&EaRPsU${!Bl1*%mXC&hWr=+?bXiS+-($RrTCi3h_-AK>gVQphc;%k0Pg<
zVyc(MLGnzUJ}GP&=-BB%Y#BSI*E^KP-lVzS`ZiwZnOVagtiSo~Cce2sx`dW{tz9O}
zbnXrOxdrU-;}!t6nM_c&`6-*Td&1>&<+$r#!ON=jgj?S$OwEM6p$O_(%`d@Xvx4LI
zr_-lH(AsDBG2%6^%?unw`q)veRIuAe6o^)NTd$3H5Z81PZ)E(8`<07MWz4>EdPk57
z?j5{O2bZuXtFw*Z<~1ZucI}1H*_MI9`I^^n*f&{4`E}+Hnt|0W>zH1g^%_1l*rN(U
zbTmX%&>7mUEJT%+od!*;c$ZaxFoK14h3>>0Sv2U9YQau6cPt^ZCyPuQaZAY9%0`LN
z)_OYh+C*l3xXN!i-(gFSa|HLiH2GGe6EG~5zWoa!WGnrd*EainOSJ!kKYt`@{Nvn}
ze`j6W#cJyyylGN_<I1|O3|1RWHAe~nZ#O^T+NHb}HkUv(wMEkioHGi+_IBK<gQE1b
zk$_@pO9%p0(9!>v`s93apnf;<mJisiPXO4{v#9|z)~rAD(SfFvF>iJ^nz;mC0=}-N
zS5Ni0___AmnnCTfoo*Wy?ByXqN`0Ief+v<4j047NrOGd49@47l&1J67fX*(qiF0)-
zGdFgYIts$NfBN}Exys$?VB>u@2w!Z)#Uo`nrdaD)t!lGVOs(^kY(I@St2e86b$$r&
zR6iL13&L_S{ug(ck>!7D!@t&&j^F;*hJRasN1!1}`V58+YGURN^0aDT|Nl)BB~fap
z^*%nhB%||r@RCfCj->NDR)p6F<MZ=f=@andPQdAb!{UdpF@WBI&g2((&TkBYQi?P7
z_~H0@FJj~Ad-uhx{_^qs!GE)Jj|U4h4@iV{i{Ax`SJ>wAN^+;&Lp<gwyA}7`yT*?d
zi--hS0y_aUeEqy+K-3DkCUf4UJkcj(F@(&!b?cvU!b~51<B+!GlQu`S{Y(@`v<$iR
z#&_kOF}0+f_G`2mm2-Zy$7&DfUK$-Dxu0zQHzeBg;PVO*{|n>o#yqLMFS9a9=e)aS
zTBdrkheEEIqj35}TQl*TG9~j_GV5D$*TAw?@`=itMq0}=!tOyQ(>~XtYg*zKH{fL`
z|2gQ6bgyvJT<2a$jkk{CktNgUAt>}2j~dURrSEJhitOPcrl!Fn-PO?Oc`=h!(pz+h
z85R4}r{*^LJBgMc4tUiljEi!yFCD&mPX&)4WYX)1h})F%MTPu5ne|P_#4LTE6IUG#
z=aw%5V)6CRgvht1d9Q#7Icbk8Z%N$`A~{a(B3wr$K>GV_f~LycEcd!vU#C_?#BBy;
zq8R=}2jFg!mlZ+x5T8h?;}Y3bCE!7iB`KPoq}1hm;n&;4|JCZCW?j>S6fV%mR{|L(
zh%61hNm)MppWxfna6O-<q_>u(U%8J4pU+=z7m|Fkc=4I6`B_b#slx(IJ~IbEs-ton
zVCytBvYc@Gp;p6Da=~I3*+!n}62@9=s!!R)G%4qqBw7IHecQzC6u@~!RjjuM$@!ND
z@!Ljgd=S(|ex^DEBk}(wik^cbk5GP1>%0k2_ykPTy5_|HAjK@&xz2QdVP^p`eFYia
z2HUkGx&t;T{n@cqH!+Ad<;&NA%SUHekt#lhTE)F5>PSTqB5ahKYOsVD(I4SpP31S2
z?6er=M3KKwS)&{1%ZnjYvKN=2h_))bqtco@5PDZFe78p~T{<6x%o^Bb@wCG>CGbDk
z8k<yGXZi`z>AiyP-Xn!VTxC}pBkJ~Bs&8JB5zkNjSqe6BaT&|Gy<mAqG&1Q+#``V}
z&72dsVCX;^c`AXAYL{xbpD_P62nhA?Y+9HL{?rIpUm0kPjUot_xhjFONE$dukxzK6
zpteciD;165khVW2$avuz{}OwYc-h`M2F$hQr9iR-`zjDxBvNc`L|{ypNF?YU8EP6$
z)2kk@E@`h2C8O$JRzQE+-jg)IxZu)?Q0Kc*Y`nY{X)1pS)MmN9LseK_->a!cBjtSa
zo=)mPd22y=3rDbBx!yR~7-A`7l;Q?v+HN|MK)KEsHsYGr;-*g4KNpjzdhp|U${YrQ
z9O{B9Zi*Nb>cr<Bs{<$<Xw~Ct;~#xo=#a`Vv9?19phILycYC+-0v*Kf#S*~65XH@i
z^dXZ*H_`wj^K+~10d@+B@lt$y&%OP^Pj+o!+qK?(&^|KFKQeNeW0!7FbHD@ZJiAM%
zCxCXZp`1+|1(&Cg_ba-*`Fp-Y=v;IT4<Rl0fQiSyX!CPlU-NSc3K1P>AK!;SlMOv*
z7CZqUe7Sm$Fk7h~j(pWCTLcg_aGau>IfJa(ofg&O>l~fiu-M01dw}UkyH)#)f05>+
zd*x@p8ccaca@G|G4RA5|fejFifVr-P4se6Wiuo~jU|N$m!K*90_R6>eJQ{brW0urz
zyB8kS8>q}~=v$r6%MQDmZP#V`+HEp^4wYM}+FJd<T^;L7N;jZW>L{i7ODekx$cH@3
zAbkawOERY6a3fIIZkf-eM|e@LDSTh`Ug^ww{iQJOniP=R3imOMk)0I|Ug~J~XZYKc
zSBy`weOw{fJ<PxdG#?R_DX_4H2`UMa;kZ(0$$={`V&X8=oJ;iRF8*Gmq$a=2MZMK`
zr4JE%wn+O5F4P34KYxoJcjWOWI(cpRJWRZK89=Vk0~tZW>1}GsPMgmkeMwOXeh;`0
z{PlRoeAhst?REVMSPI=*Mso>gLE0=3*2!zxwj|nyCTbg2A{{F@%X6K`Gg2O=gXi)Y
zGrE6Yxbk`<=wWWr03G0tIL34#dTM5g*s-5HtaKF%{@dt@dsuRb?TI9<tv#{jyEyCF
zb7ZSCvvs%YiL>;={|)=eLTYEbJeSc%lmr5Lo7pDMwVu1k3Z^puB&+GDirs8vw@53J
zo$09%(PDs}Ha#SoN`(_vKwFNkRV@QtYL~+FnVd|m41XFu<SX83-Sp+?UWICVG>OFW
z{ixd6hL;*)J9lXK5zMIGBej>Ct)%)~!bAdgZisVk(rJ;}m54vX+jRy+>s)xhF0yp&
zm9nc|qiJ3<tf<Uao67gQvi{LTolxXG!OAK(%&uu_V^%Oq%`@x+uYDU!!p<yn(phJ!
zvfnN`Rc4Edt66tRfa`+21$|bt{~<JVXUGbJ)<U5tsSnmo6L^Qu=*FFOkxIwU)PCpy
z`GGf%FLB?=mXVKWLhV7V8=d3b_!<x2P<kP(XYWQyatp8{MNoFfy%e75M5(nSE~{5s
z>sD}*D^ee>x7GCV_)(2AVq}QLMsm8+_1tu+S#ox<&Mc+dZ)@Y$#yF~c3rj@Z@CQYm
z^7tX7-Re*8&zQh1qy<O5GDI4&-&*>9UY1}ghclJC)sU{eAb+|>l(kD(e`|1|9hsBA
zr|%n%?9fiae`5!x|6lK`jH#Wuiv=MwGbii+B!BH_|G#Aur~fINAmj1@2@L#o2{@~b
z-j2$PvXkd)`TKXCh|5Q+(Ny{p#9M)O>BxaIXN|+;?P=|Id4B#bf1aNgB%9_*OM}zL
z8$0w@;GzCt<HFel%phFE`Qhz(JH~A3;3+)82p16lzsUE$X=E0taK4Zt@tnL`ly#6F
zll{<9i?5ueZn)c9*4>;uwDcj!f7=cBd57}-{=OYYqBiGvqUsHgy<LCZKg}(TcfF(q
zE-w-I3xpB?pdjSr<d_f3{r~3l@qKfn&?!6OCjMCPe}AsAw)l{3c*Zpyj#}(Y$T<Ki
zvSRMxhU6vPE?`fw*Qy+5U`z{w;C{7_=I`celvZ%w1mRN^0d=NHK$DdbAODK29Qb#v
z-ni9D_!i#4B#*(RlHR>n_0k@RK{K}3xC?}@Aj4Isn*F`xny8mr-Xs&)EA$_4@CaOn
zSO#Ezt<`6j;3-i3d&DZ=Y^iIuhSl+aKQYHxkG=w&>YVMoH$M*^akaP+GG~~V<F+;>
zOS3a!tD(JJ@o8F{LkV!<7gdk})>-go3Mo34(hM8r<*Hwj1h=znq+j%9y4<vNT*qeh
z#MZgCo2Pk=vFSVJhcm)y^%B^`8uzDJHBF6~4#PWgG%znft=O>2|6p5J`ZBX=CD;(g
zm0mirhW{-GDX>91MrtsKwa;4mg|qy2uqY_kD)~eB40gj$e&M&w1$vMYV>{>ayt}L!
z{;tR;7g$U534w7Qx%Nvf9qf*kT12iPSz@;P!BftGv0mgI75=KCbw6?a+Y$apY8n#h
zz1j)8Xt8doLfXEBaB<yslRhRYoE3o=9-U7l{Amgzs}j_KuwCd8b*8AtN^HIKxAfc&
zF7!xC=Rm^lbOL`E5dz$6(V@iNcw*ncf|i^!-NI5Ci<SqUB}VG;6KyQCeOQZL2>Pe6
z*WWyV_*9!d`zYpO6ZNbtPi#!_sqoyd<dmQk)>$)ce0s>;0m(8n?@|2JEtyhxB1`^3
zOFTj6@ReDRU10-uQ$G>C0!LOuap9)w4zgmkk#K;W+8A17C98cNk91v}y;@@Fls1Q&
zuUeAB7LQZ1dP-8`$gc5#Qr%hctaZc#Vx(`0==lV8I@R;sEbXIUZDtk?T?w!ud?HJ>
zKb2helW1mF{j$h0{tjkxr%WI2%&}>L{6Z-&d)<V3th5rH3%ZW|zAT09Y);zqNgi)3
zTvysFla?R_!w>pv#NuLTFEvTVn$tFmRI@;{Q>bNMqaQ9B>Qg~CGYG3cNGe%jbPQvW
z<^!+i+Gawk$Y|}IjOHNi5AQ(pe3Zu{@|#TXFS)D0H=6<vLVg@d{Xuh)TgX2wEeI#6
z{*3%MW;>B{5LSqT=7%sZhs}Wpz;8As9R&HYfNN17_80zioAwPnxxn;2JM+cDzCouv
z`GmM7xctzz?LyF5mfj+e&*cH0utwWPObuN)hlzmUA)nVl_G3k~Z^XdO<=%>GZ2Dx$
zVf^4JfE9Lq>NXs48lwn2;~~&M@LR6q7AU^At6ewNriDB7CZ(Y2QE3p6!ZQtdKoTH~
z<D9HCv0xj>vS@QgmV28#-FfTQNLbAy8DzmJb=q2XT|}xPWHPeVvgWY$Yq6wq*@<nV
z<fua-emnbvt{3^qv);0qJR<5qC#Ztg>*llF(mj`Ca{N?}jsO}EhVDytZ0AA2DM(j0
z`ggon`%t)k(vHW4j>Jbc{qn0(K18qDLE7=@z3U_wzC^i8^Epb5ZNe)dTtBD=T{s<}
zSKHe5u{)5ac0%@iE4Z&ysBMj`+N>qm2C@#r7}YI{&-dc1>ma+Ji?OGnM%(?Ijemod
zX9EKyWDhAhgXdx$g@ZO-8z2zwmgw$&<J*Q*-3_nD7_6xm7XsDvG%T2w@aN4AxT!?U
zKSpzhL;)Q$hG&}~q6s1jvZ&@CW2CS>`5{e$gbH{$mhgH<5!$8aKgd*o+w&-C6w<L|
z$iM*(Cz!9s#U;W5KH(pTS3upQ2;ERF#&a*If;#=ny`=Ijy>>$lcb%z}cT7$BT7h%b
zXol~HVtQ`}+_1LnbJl=SR|%Nx@4jYjbJwgx^kFq0J4c8I^8?QA#vitH9#6b&f<cp?
zZ3GB`2CKzsI!#d%QSJAw3%N#ja@6y0xh_!{Uh_k)e@kaUKV2+J|68X{BQ)h)Gb#JU
zew)8y7y`jDOKp(T**4y0%D){Q`e;c;ptF2Q5zE#~|JdU>v;bKv<KVr3bWg3U1^2sc
z1#UQQN1LPM6MPukZ8`c3Mp)iKekH+<7!ItzMmmum>nXQdAJu(#r~rG%QAw(_>oDXf
zP_v#LCQ`}^Bh;l&QzF0)GWL#>NT}jX!Z9aB5M)WsJ{d(Kgs*|PM1^4%S!L!MKL2K`
zlgitSV<KArT#MdA9!e8zeZhM)p_S@5H~}GshKXmgsS<@E)p&B*C(EjlX%cj&+o#y0
zDypAZ)+GIV!9GbVM%nO6riFP%Yljx?^X5*UecfA8DEf0-ClW~oKCZbZto(>i<vM8x
zK6~B1jEPIQ%ZUC^@nc(slv_^<+`Yo@D(V%WJ6Oa$K~N-T*Ee@vR_mVs`>(!n`l3NZ
zz?6IbX3g~ztxgZegC`X)1<ZA61t^2KWDL+|W#Vxy^eKDa()#o32=+^P4G2j!)5##$
zCF7;Lgw~o_hk!L+g{_RX>*flmaN@tIHam<$s55;aTwh_gCyo$e`9o*Srtr{0+#tPk
zMUF7SbdiJ}F0ja3)SzD-PQ^OvaEEJB-`|E%gRVSJytEfFL==;JX3n>yT3&UcK|d}e
zxi$(mbLP@+{;W`%6{5~;xTs%7wd%+36M>JOwEW}pQHZuvO%8<h00cKN)J3vmCg6JT
z{gb=!B>F82_jZU<tBzeIr=&G0ISz|uAu<wRdOIKf42!ShwugmTK0<nh@16i@9UBG;
zJeTfucY27hsnqituG7ZnNy}|Hjjo~l1M^OkOa|q?)#l?D)_TW>Iz2SoiVfGzMD*CG
ze%8||D^F(0C{AjCZ=OKj52#EVN$P(kr~jh5u(5OfPvo@n&p3SSK<;~0J_Dc5r$y%|
zFdT4mM`UaV^M!SSF*{)mpZ+9JB8`rw^h~V%l0dN22D~@2uadXs{A3*c^S|CdugB6y
zC!p<9g3mu+B^Tf&{ZmVXtjWt$8nk5h{N?_7q@&RK&OQlIwU)&f@8$K)+!v&lW5K)^
z$3cciiMeoUea~9M`RYH|NA&l5^zz{=1i>H&GNN4<%Z>T_{rw=%)SwEA6*ye;cz*wW
zxouB$X%X`{F8BX_N4GCf6ZXy!EL0ej_5Wu7{&{IfZnGn|l>IS2dRik`VtjB<(@P-b
zv#e!eIj52&9+}uAbm>Vg4JS~-K0uGOCZ#e+3CeQ(mwl*K{gqMz$wHh2RSGOjoM|XL
zYXL%%dn`}v&n|qVZektU#7G>Rv!H6u16ZweY}1<;0yUvX>QKp-S@KQXO{L%t+`Oim
z<0|43zuci0DGo*O`-Nj7!4o#bkP*X)(@6$g{VTz)G1z$aya)3Y^K>>ENP0FpO|M|V
zx=;r*v31gq*lN;ZHO)UlEkTddvVQTSBTP_;_YY`7Q?(evB(wr7@<e^QC|Cj5W@to_
zH0y;ST96npwiuuZxQK~++XoP1IR+w&?Bh0co&cejj6nTp_O$#%x@$(#yO1tYSu1$D
zCd*81E&J`+b1&Nkweg&sf~Z#Ys_p(~a)&ms=?9I=CIZww7abhenB%P`EYGi1c1of?
z<Y4S1JLD=Z)^nN+lY*AY9jNok?~|NutrvYP4Vcr)&fxt=vF~!;EE~YkXBZqxkPYfr
z1Nn<!Zok5`Rxe#wF={!8_D6Fi&4aJNr$<gdaWSB$*(n$q7(>ZJsGGDq<!@}n<pp>o
z$c(U`_gOldF$k12E`Q~o;CtsQY|32vIgJx`;_C{-*s)=<`tnqx_phEJy|}UAE52oI
zz~Lds%faxS!9%jFy<mhQby151rifIu*V~5ZBgS&PT1Z7E+4o5!07nLCHUwGdG{;)z
zo-zg&(kw4x>1r-do9z%iLpx<_X;65}XU%lz#-egDx#}T1_9<c&_JbCQ>X3=!F+sI^
zc@f5VI|AL+epRZL9B<fLvk#U_TFSWTx?M<(o7?O9W^(eJHuH&sXjaVYw+FjJ;iH7X
z(kHf8(d^Ti=}{ONiDpKtS7Wv{)hq;lohDDxUL;?NV|xAfTSOnsHZ=jk=RL3xnC;(e
zpR@G<P18eX0rmu3FO91prWuZE@i?7YLi3h_%O-uw2)VH~*7e5)mpSvlWZ{AgJucqF
z&M%fNsQBM`<2sF)4MU%Im=7}`HnbgZ;IsS7ak@poj;mywdW^<`L}*|A7}MZuIh=G6
zJ~>kr%xu3t+w4VOxPDbGINBzfA!TiID|sZDrQweNEp}L`E7Z}Cffm3~MUffrSNg6d
zrqA>%_Q}<O5(Q&TsCtoh#tI5JRbJGj48t6IK7ZVTV_4(ec_Tt7C>xf2LZFwnLe_5c
zXxEzCwtI<b-e^$ZC~q845&cn}2($5w3()Y_6sw?N*4mUbC1MIXU2t}HdNWalEoE}9
zq~W$P&W<JbPxP*m5pBB>w#TMwY2i{R*|)Bmj@&FQ-ic$GoISTOd<ca7{Tm}cE6S%!
zEO*h|Ku@<k`napZ`4l6`Z1((ImNk(;e^g#CV}IOfEha#TCIowd5LdAe!`I!3!Mg~@
zazt@BG^&f<I@=6FFZ*nTEw$IY;%GU<CUvLMChsC67=jWkgQViM4D1l%;88cHwo}|C
zOEWd2E?`)ZH$`iW%U#D_MTOaAevN4zxtvWuRqmt|J81eq+AMo*q<j7mjV#3@*y!YY
zyQ=4xbRue#$t8c=o^f!?2{r!f2r?T_vZk!2>TOR#-Ac-FfJnV94EP{ij!*-Q(rj_O
z1vUTI@1q;m;w-)JPNi-WIZW4iy!eYjdXZ$cV(tt2xqhp$ROvLeKS(_4TGV6Gxm%&1
zrPH+e5MoZc1?J+AS29h;{oOdLU=H$6s@)N`=c|o9gXEeDTu*vsC*U~bI0IrvfKk+9
zrRBH*h1reBci0(C3hB(tLXGZQxl-8P>olaSh}u(8cOuREn3lS{G9T4hO794p_%RVw
zv9nq48}cU?`W_{*V^nn!{UW*r;O3n~(ACNaYxevl2x{ckOsaC8_UbNn6<aObRNY=>
zS<tAtVOZzsDp_+bPws2VP)Xtde)9?Ow#w3z&YeFvMNg&95Yf2hzdT%Wt$UJ!0ok>z
zg6Z*d#J>4Bsb&iLx+yX~m)htBG8Z@S_L`JcHx)ylRQ8${n@a}kjk`EV?mkmSf_;x^
zBaELEy{3hxXM67(B6-fgJPzPt;e`JLwTYrSP4VD#4l@z<J!~iJEnp%G=n0xsV<FAJ
znp&(ffytw<@;OAh)DiaE$?n%UWbX!gm0bJH9faxT+YG0RPl0ZJv-><XH3A@y@(v&C
z)dSduz<HKZY*bC5hkcJ2b(aXV-Ov7zR?TU45VRJuq-ev7KhFxNejb%2c5Sj|rA{X<
zdzGVY)pNuJ5M!)F31$ZB_er_x7Ode&Rg0fOs}_?_Zf3LUicQPA>!hnqS42=*S@tvr
zzSWVRdpenxA<)&^xGuPbj>(*qisd0og+a4M6bSGA+wEAB{$Uq`Al0I*GB_i5&3J5H
zuJW3TD7P|L**NXud#>*^?k1iW-c&){0+QLCS!d09R44He59<Xi7=TF9i<?;Q2#r)+
zPoSRa6nsvN_cce~E8y69UO>a)@uMN`%0=_otYj6SZe7|rMdm8Eqi&J)`GFvSHYfKy
zTh7q5BlY+v%?FtZkG(;$&q5Fry!|+P@=w+2BB!Ldug}?~Se}*py^5zcO<D>GWXLsF
z`1#fC@+1+S@Y<$9k7jbw@<Qp~d{*<JtP>=Ac6tpEPxRe3+v^Y*ucL#`4tHgJEvv#E
zqw>+?nhf>OA0u%|Zc&_5kF_WGLLbv^PyL#u7^m7;wk^<Zav?DpYdM>H&ztL*iriA#
zZ|OfH0PCEEC0<pb89aZ#qV@*1$+TYkna+hWf;oG;jSwd>LU{OY+vRqcqra5R1PaSu
zTZ)R9h-KDX1n}znkR;)4WU*&SyE)a{RdSDCMvx06Sdq8V1lmJRNt=~5D>Si7yjnK3
z-A%5ccCmY#(PuS9f5fD%p6WM_1Vzk~-)GO@J$`o3txIt+)UDi9^SE8F#a08U)V*XF
z!vCGrhHbrR*acI)p{_gsq0Fo7XYm_SiEhEF8KyZ{`Af}=bJv@V{+OhYqQA|h0@}@C
zHKssewTPtl=ELb~>~I~!YAom`sOzC=1&(&m-wbeDFKBwT)>%4_-P?XON=;Oy>Jfy?
z-B&#hNogd2Cx()5UmL>xy{UB&uL@V?-`+{8z92ho(%vgA&csAaON{+^bh$$guI#@+
zWV=Rd{|zgdnf^})12YFB+y97_lJPj*$bED5X9A3@?FS%mAsP?kf_8ox+6lwK$UKld
z<KA3Hx8>t?M$VR!wWe^t1H`P@yERm6j*V5d<@PAJeEl~2`FuICc|4H?J^6ZlvjMva
zZ|ah`E{vX~AuLv2pIqP9qEgp)&13PYi<n;?UwC&oZgXPbIfb;o%QqX^p(NxtIi7{)
zK>di1_PKrp=br7}ZA#S{{~~FjmlfYPxAy+NNxyC8G9T6UKSG=3m4u5(QO(C=C0hjl
zh##Jk#A}qmH7rH4L_g9wf%yHv*G)Ht3YU1=<)v2&bymC2WhxydqWb*Of;Mdne{YxV
zk6)Qo>zj6GS+n~?Gxsu;_3V|o^i^U;b$0&5bB^`$Yh{7$D#f2jZ}Hi%+-9?&#grXv
z)2y4qkS&p9^r~p`sexL3nlmy>lqbQ<r`S)Hg>S}WvQW`h$*s$HG1mqr7@CvLWZnkk
zE-@urjlW6gMD^vd>@s~4w+ST*&XeA;U|X7lz)f)uTgu0WQ#d0i;4D|u8QZ)Dh#-Rj
zH4R=;N?cKq<XkH=zCQX&Q<~joQHeb>GyM_aDh1VQA(6kKht~JxH>&T?YO9)A^X+q_
z_mEcb#*yZ$rC8ei{F&R-EW>4LZoVr;`nSoCux#hnShCz~5`&6|slJVI?Bzz;rb2^7
z6t2ff{0@sg{2zlnXC_>Lwzn_NG(I8B<;v6Z=@J3z=-8YxLXx~v2ByKsZAy43^P7Fr
zP&X@<eGshO2P%c%+v}^n|NBQb%G<jCq6(L0q0k89cZqZ`dg-9Nz}Y4=fQA^6*r=f`
z8D5xZs*hI#qEAj<&07agnC;_V-_O>dl`fx63gcjTWRD=*Z(h=<DazW{VtiViR5IqJ
z&{&t$iuvYql>F23MbnSvHD0v*kJG0oTdxmcE%4KvF)AVp(6?)!1~=JiDUD+$XSJ+j
zqH8cQBU4S9g<`TC%-Un!GMH^CYlR^~-ZE{L)Rm}dUZsf{5s7H3I|aj99D%}cIW=4=
zF&IpYl@R245-e6l(al`8p0A&e<h;)~dEdG|bM84Q9K%7`Iiy5r71+he0C3}5t)Aj@
z-fUxr&&onL=K7(6qL;cjAo^F+s`bocFFml&CQYn}LZB-M(<hB`rJ%8M>?jmDFyFL5
zvo$G7$^tNdso9kXI^anK!o!lA@H_kL1^N;AKMn}>;DxNf9Q1R?PeKL5Mpi2MNJ$pm
z_%F72N3fz|KCt$x#0Dy=Aoj6MK85$NsCoLb9-(3)jnd>W27{!hv|f$5Y|(l(&TVJ@
zve&GW+;gg%8x9UY(I&Ekg4{Q7Jt}ve>>Yw(gAinr&?%6oHMeUVL~WUw&hEj=Tevsb
zrwn*t-021}Eot9e=_Dy;#qSzKRlbQs3u4kve3F-J<eFwnv?n#}n|V{C#}K+J8N4|3
z#xe5E*AZt)?<MtfJ7}9r1fU>7>Az)C=%{xxP#Z_L)Ht!q1pYGldVee?lYx#nq#Nsf
zfHIrT+8uFlXNX0<kmU~hsBfhYZAYKLP)Ah2Z>5v1Wys?-w{$(nAdVHNQE5#dWK8iy
zZJ<Z59HEZX&}6?I31oho9kpLnTp)S#>?p4~eH%nWYK8x-%&>zJa0YbKC`vs`@#?Cb
zllg8&aFFpT<l~LNi|<!I<HRWNztFgTh}scff)pG4+X<GlB&PxuG1MT83d?ROIt8|}
zU4XVbxaGMlMS*w;m5pLWsee8a!q8|1Qe?5GoHF*Q*^0f8jJ-VZ$GK9%=3q&aC>2XJ
z%S!GvmTsl6iDg8p?mEm|1t5m#M5*Cw6YPokwU2bJlFje5ht4qo3rVRu@EG7aLZm+-
zi?G%W4W1T9jQ#6&@tl*TG}#!W{8uZMWzHO|^_Ad?j?=R$o+bRyWuO*}O9MI(GA3tc
zb#%S@W%0UfcUn^}eO)0Ym>)z`zFqlQYk-f69aWun#*3WCE#$SX{RVfyz#NGyY7x+<
z#L6!cpB8OZRM6R|V4jI^FxKRza_qKBggy$!-83L`8-UbBsx8@JJ1!$M@+Nlt<j6+d
zJUu@ZU;vZ0>-)hB0rp+~-*aLG1UuL{aQr`vy<?0nQP;KGHh0^$ZSS^i+qQeRZQHhO
z8+Y%vZFle1>F3Qk`F?!QN#3NgQmLf=txC-`=NQ+3wnk+US5Q7oVCgbYhEVzQK4F`&
zM$NEDWJdN#E5xx!w|In>SoYiM3VdS9MkbaXAyo@t%19w?VJ9|QfGJ0X9}HJwI!Hko
zVA~$MBq3fRFXMhYVNs=R$PZ>OqKO{Y4I0-*3O#B~S3+<1gjt=F?Jz^fl<vIOTH>tg
zprCbLtGPxD3vrzOos94&Ffb!kAY;g^c=prB0tjpulSbW;iKsypciqH+3*S*$XjcoD
z<AADkouEHw**P&^zeHIr?C4rUM|g=XQ^+Je2#>o-&1$X71iqV{s4mE$oZ}z2yWElE
zS8L<+&COtT|Lv>0+o%J^`bkZ<z8vpA>-m^$2QfKL8Va3cl-WsWoFknoX&q{d49?eQ
zCF{c=>8(-fCxH%wzm@lAmBMmVM2Z6^{ZHcGO6s;E+C?=NnXPn}{SjKni?<uwky`ek
z>?K_yD|L30_I}PRO?G#x5tg<sDZ+a-A+$wonU4BnPU(vaP_4-gv&S9qrnG-Jcm9=x
z;DwsKmoD4jGS_-p>GrfwfOq#QgDi_g=}k*Zzo{%tX(1a4yt&LoA-_KxQ^UeAr*+)$
zlvW0hUkP5_GqGV^!&3$$E4FRpjWLB7(rU3NbRY1h%E2ulqRMBFK^3?*asAeBx{mJn
z7o$oOmU7Xc&JfKN)#K%Cz_QX095~fDGoG1&%}X~{a(pah$Kyr^W3aRcnVZBEr_1fe
zD&?5vFqa5ySNp+OhmGhq$~Re}Wp#6-bmcX<b)$qg0@<!2zL4i$;r5J2zhZm&N2e0y
zI7r&{*g-@37N;O9R#pplV{fm-KDMqoVdkD5x9oJXS^oL++LEZQkI!W%OXZEOZ&PXY
zz0w+cUCd|2bL+NuW${ZNwCve<MY_@TTOCoibSb-y?yd!c)IKbbs*noPA^-l@Lcu?>
zPnj!Xzjw&z9qvfPOLMnmhDVmJwZ8JaiW-cCVe#{(5~%l!C-I8fA{Qxtm6nTD-Gv=G
z%XqIcyeAY<a=T|Z?smEv1uveuFl$0+^$V{^4)&sI6efh_?J_J;_%uI)YrFRBoxj@H
z^n4=^@M+t?rc3RdZWWNn$KdKvfmOs`|F&uI$($Ga40O~LWGwsR?(0A?lLm7U{@mfh
zg7}A(!VA-4EQ@B;S?X%IEC@lR`+k`~f_|DsYq+DPw+KZVHgJMS7F5blbf#+KIzIW^
zZmKEwR$;>lOa8<ywyCceCRL@ouNXvd`$?BUp4SV1ehpVLVpG14V}y7`ruMg1R2J#Y
z5}C+Dcy#w7{idqczM6eO<F>l_*DKrx0NqwUrP}`P6_&sIc!&9%{cgk!OoeouKjq^J
z(t7Iopl||9f81e$xbR%XO{fIJq@$nk5sl4n^=1H982b2~n6I-U;Og=EQ|pm6`yD1!
z7QIqZ%8gaV{-dIdwHT)6xK$FlyFjUNVipfpviv77(|i_gYT{ESQzawVx0P&RPgvW-
z=m-oQHr7~+TPnb(5EgT_@b&(5b>{yE8@=zpmbd@KZo$g@zneGzXw~dVI*@-5V~+%k
zVX8i$p~%|C@bMe8KCn(2aLluY-G8@m?_ZRCRFY==BvVJdBY-<rP`fl}0S8s`i9;8J
zVGLnqX+vd6i+@nI1BUjF-hmzii@ymi=Q9U@p{fi#JaN7p8OXG}_!gw4D^PoIdtO}F
z5Zz&TXMl9|ZXHfo=fhzT)wa^i;dJBP_(osK?z?^4mwx?39FU?|;BoSQJRIFXY9@cW
zF#ACXlSp+HSGa1$^H*E06zNEcc+nIIlj<xLHuo+UN4wqOFCVbP2;s9>Mzfb3l*;7u
z{4$d94BEbU{Zb)aD?ZO<su4<PW*BvOv$TyWZXN^`D=KLj7;(?s!OAq>$L(Y1xRNlx
z(o}yK7B`L;1KHU-p^cDBh;>S??^8}iUlr_?cK&$Hn5+_w5x2&Va@n~q%LT=o^W}@8
z(BX`-LCmy^y^@mD->62<^QNWPH$3~zs#VZ)E>GIw;^NrmnX5$uO#Ucu&UfA+Dyy|Y
zQB_ao;lI#ZrrRZJF1}^fTT&tjilRF$0yt#Kb7^dvODr^b%?6bL!!f4S>veITV+a#4
zlg2U*DY$5B6ZsINL7F5ff!U)}{&=g~d*8EWXaQ?LO~jx0j>i4+!Rt)Wmr{dCgkz{-
zGy5e$c*go@FrcVb8Ol(bo!!hyC!EHYNHR^9xK_>FGi*_bN{?GuW<-tmbm%YnIRRB6
zAl0fCFrvpi8KHMzGpPp<640J`%m<$AOLT*t6D{*ix?WoXBf))zy_8gQCdFR{ax7>*
zH#to6&2q0QT$dK0Nqr#RmC<TW)NI5D&q`rh!OSekW#Cu;jkF%>$2XtjOP?T-=->u}
zn|qUp`P;#de1TRADFwaU!E9j47oQD=`d?dfOe?VhOh<t&TzE%Ub;c+rOo3;Bgd1tL
zN(^TDh=myZlfIG#^CCG6f^c0(Q6;k#sVNjzyGhuHvht1(JnAx;ApTT46|l;#{PGO0
zG^Yc+1(NVE!7&#|%hh_FOaj3=qf%7*fm1~BZ?Boy=U>v;z=^_3R7|<52oG=|M^ySP
zq|Jd!biOu=W9n4~iXQM)Qt75-(?40SPcDd#*n?bvn0(Gb*H4w%a^j9LCvJ*S;@bSl
zWb%T1nml9*Ib;fH8T|T)j{Yz(9PS*bJUgQkgOG<_|H^BF-=k6ZdJ3j@PM(>CO&Zu*
zyc?Injx}~y%;F2ti<N9i0tD8q<nJNWRA>O`M7@D~1;-uCyErxgb*|iuS~F2W%QkQ`
z^dJ{r!#V+I$4na08$w4C%r@LN-eNrvEBEj(gk2jKW>NozCB!^G|Ie>B`#vRE0=;jP
zvR}JVf-J9lbUWDRT~NqFw?&)Ui80OW;es6$<m`b&><J&coVT`|7B0b9(e`5g!;0sR
zyiZ;;L(pdA>bu~XL;xZwTyf7Hf-6F6)ieoj?3HkLMJZ4A?Zt-=mF}3j=iIVI<!^k1
zov@)*BULF7dbvAhScKHC)M}}45FO+EHtu3v)%$!pZ8FaLl#~UxzJs6BRE=7*2m>xz
ze)!y^)j_Uj@frtW0&Qx)h|krS`sMspikKKh$v*ZemsAGo+F_wWbrzk@j=wj}T+gDi
zex`M*h@}d&?WKcwix${<kc65QNwoenl_~e!@#Q%3iCFyvjbpDZNi~YFv>zUJ%Ki5B
z`EhTqEfSDnG}jOzpH4;S{BjHxToI9)l2e9UO}${<ob1G*w%;C=W@PXTG^!BWuwcj1
zhh5aPU_VFzv2hXCTFX_Hu5I+{vlNz`ht(TTO+!Blgv@b!hx=nPE$HmCUZgMpS0HV1
zZJy>W$<kjraw*zKvS$nLT+2Z0eNC%#6r{o0nmki%>aFnx8KxeO*>L`E(9@`njQ0ez
zR%K_I9``tv=D?R)x3ki;x@&&6s{o)qM#9F&kLs7T%8|r}ZMqx{+Cwc>Mz#pMg?LSQ
za&2O)n=rTCpIe-y7_EQHWx0~<#DDKhKJ9$g;Tw@D7@?OepG}N2oFUf2Nv5MsNaV)F
zHi7+D8s9p`p^J~lotI?-0465QXhR0%nA*PbxFpwbi%sjU5k9#nxfL0-jTw)_))Mm1
zaH8?+{pnq^x6`uzED{>y&q@N(a5+JzdBPsl`DO@P`Vt<Me-|M9MFyLaJpQwEi+H#-
z354IOgaUPkA(HcUsn(J5@>mMw++G5@fcMs<yK`?Z$?Q;ad%$<(J+|lBLvFhZY?=2w
zOD$pc)*!=erj0=%aUx~{=VNVx3AE0SuP)PPs|CF)WF;#FM`vQ6K9j2#QP;!2%S`Hy
zMbHL*h;sVUlhjZyD*}1RTXtCUTm{*hw%9th*y%}+;6aKSbkMEK@kuZ_yPy9SSy85_
z%etux(x&GHIw~aRKeAIv$uM-aZ%9}wp!r#var{LY8ygJoH$c{!`*GJTy5oN98Ev4x
z57{&WPCH`h7mck-4@}v-QrPp+SI#nA>`qvQvI}Z?uRerDR?KMEXqsb&gVE}?m)7X>
zaHiTb;L%p-yElW1Fs?Nf!9@~|-_eezd$JVBP-^vRHR|uLjf=D&U1EC2B9WDy`1m<p
zzrRL<>KO`e-eDY+DBqc4zy}CHlJ-JlL@Dn6vv`&@oSOM77zYZ51M_z_oT~YXkFPMC
zB?Do2Mc~Yc8gnT}tI~~;Xs07(3#MNQWgXMh1=r6$2p<HO(29D&EVddqziZynNMcb_
za~((4#ihMdj#NWzNdJDb`#bBmDcN1vvOC(4_WA_~>M2RHQ1h9Bmd)tNnn^!TbTuZL
zZ+Z~Xn1P`*_RPzXf2@woN=n;MQb=ELOXCRX*jslC;pQsV_qCxel_mR7QeZ{Zup^*X
zQ$i*$JzaCvD08Bg`c534kf2FmjCE}C6$e+95o{;rvNY40;B8D`PHq&Y640JlneP$&
zUU5^h-fPLaF~7r2iR}TOH6utvK@1@-k1P{tD~mD9yswq0)gkQXM($fPzqvoCBbb%0
zr2f*$PjeDAh7XM=ZkP1*%xtd5Xs%!~HAkF+M}2YGs%!QLt+3V2Rx}6B`+Xy_Is_6P
zt=vvFeNJj7qdgIxKByM;jCLZ%R7>5A2!Eld@el1s^$O^rg_HQ}F^P$$M~0o$R}*pD
zhg)1#GlP~4ED^4I*M41?mBQ8o>ei~6((7|1SW*9rd}_O^zP`iD5aSmpXDK;)4bBU8
z)xjMce{r7@&7%N0FInq<sp2>pS^rOy^8Yt2o`duMq{TC`{O=l=TC(>4G1R`R-=^zu
zRufl6g44oX))vVIfHB$-up|2@bpJ6y7GXYQ+WmO%-Y+l7xXMW3wZdxfMe53uMZY*$
zsXDhjPL;CiQ7Ghw$d<6SI_7gsXg9849}Kwd*4!MhJ!85hjTPTaY#f7^wR(he_Kw8R
zHkXoP+a47(Mlgh=Zos%kS-q<_^p1HWjVO-3VkL!!ZYOQ@Nb@F^z`Vc_qtmNNKSqaD
z4;%LF23oTmhpR}h8Cgnfhfkm;OGO7!k*j7`tB4ru@EYmTgZDPDcZ>3#Q6`I34OTuX
zC|_q>oEJu+w7@O5tPB!m3R+ZQ+s;=J$$)>czT6^oyh1Ot;9IG2LycPEbh8Ky%MPmH
zvRMF!RVAp}3xrb*16DC+xMNfwuRt2)+8zNY{mx^{R*noi)sW0LiH1X5+gCl6tsEgr
zPGpTN(@-uRK@as_ALHG;IG`erLSzUq8}=Sx>ZB;STgHm*KZuD`BzV}xjWI%`Eg>!_
zk1bn2u)vdVqZx8Kx1C-xF};L6VJ077!iEUKU>xA|sdbM&zs0v$)p`xXkV!IRwy_Ng
zai}A0y44Nyn;Epk)w5&Q0spekF|-T2$-MBuo~1E9#2m3h*a_$x<>PcZX~3a{q`)c)
zMMj8Il)RlnCy`}Tzr-2H*UC-c$+hDAGc;F_VaTuvn`AIt(G`oV=0-r{Jy;<O!r77}
zRD?h=s5%ak_oy{0kvP<p>-(~GKI1P?t-hnLr)R5HC-CWvqVE6sbT44}G_>->aB=Yv
z6wY9;?yvW0`gFi$%b-`YV>{bmMz2>8Y7@R#Nv0j1EV{9=@&aw$r}yddbZqrR{8`XQ
zi2;^K$rcy>D8N6~7r#5gg}c~rU$rvl|1I8VIO6hl_W9lhNukiNiDp=w2#?;Im7yk}
z*VEVO_qn&f#l7M|RHLV7!+vowHM}t;;Jf|t=;hp?NwL1o+10se5XYhsD8*o$#$tir
z?&Q2c6=2@=jRg_`bv`uma8i}Fs(zza$MJK+Fso({$m8Jng;>(t{kn<`qH708%ZdYb
z_ocnF`n)?|dg^)_qXQYO>8tU7d)g1HniE-rG0l9??)H!2K(6t{HMK5|)3(6d5a1y=
zjlA0!(^TTJh9^zXaae}|R|e9xWvrQw(ZpkCCCmbzoer?+8ik|mYf3o|Nb3-NF}J0o
zZaE3KwjU`xWQj`5F;Xn$iCTQdBzf9`J7=|65-bg*H~89SMh7!6i+iL?-BTSWTN^eX
zjhiotT-MwLLQ@eu1iVs2A?+&0i;`o-e>vs9GZAI-82lxkVns9IIFG|9G<4op1j}au
zZH$Ex8b2V1V-Q4lmpI5lR|t_%BV=PbNr;RxhtE}pre5ekh02A2SZEKjC!9+mgIf%5
z>ylF{Hv`wgO9_h73KWPBB}UDdz=$sd<%?G#rvmE|As8#6lxT^D14TI2>Idq$B0_oE
zk!R^bAy+WE!Lqr8X0${MeFi#5PFJxFl~k1*D6TQkmf83F!*R*+N2_!IOx-o8N(-Z(
z)DFuCc09zOM~acx?bkeP`&%GkIap*xs8PPd2-*fnfdqWlHp~KyX~7N(2yYDI5^~(`
zX(*u<Aw+s(;P{r*0vuDmcnQ#uc(_Y3GBrLLbjuj*z+gsUXhUuwQ@LqB>L7iOd1e#e
zqFz5D)VHj4$ha*q5w*L`WSek)<Rt%j@sKc*H*A~DBzZY>B%Lul(gP1x<&l-RDbN-I
zAqc7Hgkia}31VF3_E2&!2q0f4zm<%z-NJF83@9!c>$ces<`!xPSUv|+J_}QR$<iP$
z49kj4yrNo`xqy*uA-Yf#JW&2a^q7~g727;^7JL$mdO=eGjMs|O8tZ4W*I&)e+*={!
z3WBYI5nlk)5575v$4u#@H-(EAvPd)pTyEV$xJN?_D=MtB05Tvd)L)1*^csbY0d6c-
zliAOrm^QuCLz$1xW(rDO<Pxv-98|TXtb8q(p2*iOGMe{J0ix45uKfF4?)qEyFLk;C
zt86RAFN}?Ot|kiyF1&ENp~dZ|<N&n{@MeHAh}IQ-zG<X!#=eFPwhXBVd18SnFnQoS
zda~sHcvh#Y?S-VF`0HQbIki91Sj_5%kjVD75wBL1L^_2DsWG-He{p>P#FxjynASqJ
zA%nPn@_1S_Qz01lO!M%=?7@Q2+JSDn+7Gvt8;zk+K<5TV`uXF9=kq|~ZJY-V;jRkr
z;jSP>oCXNPoUP_X>nj76E@i8?P~4zBH2ZKs%2d<9UVkz%F3i6h{l-YDiLl-pjLnBB
zFG1MTa3^;=(03CcE`ZNq<}u}j-dTk=M6l_-o9q;(34^CNmK+*99gQi_I9ioAPgMqf
zuD>p0uXjMfS|%<_%N7pNkxH2^tc}i=tcW~eG|W<TVt<`gl{8BeV`E=Q!K}c5=2y{)
zv$GG95)LiuIK(ag93tsiC?SIN0ej-6%voS@0L_D#f_XMOCj!G29e{<Xw5!$$0&&kD
z2wjl@<-yEkxX0&y#s-}^Cf<E-QQmUt-%#AS?(AK#w@}*>U;DyT^7Eay#gWs`W1JcZ
z3ptQ8wug3dE(cYM-9b+z93rqHiE~q-d8RIdYGY6;1omuSsRgh19~O0T{xmu_C;3pr
z*SWc=q&+yO)3N<xf>`Sl3NtBfT^i!0`(lZSJ1%nIc9kWn#!U0&f~PCTb#-F$qsg?=
z7g&EcK%I)3gb6s{??fdy8Yb5jhliqsv1~&D^QU+d?jjTAN}y<oU4-wZEwn^LvhaH*
z_VZr}hG+u6Qd0DP?`P?LtPOQvBNpI%3c7~O_xYnE51v>CJbKu+{HfH3D}JaG{EDv=
z)O!wO?G@hvvDDA+8W5sL*of~@Tn34&JbvW@s?01GHgofTxrsuYxBP<ddVJ43=C{v;
zt*l-rU`82@I6s(JI?g?lZ);7%0)HY#Y<sSu?>+@BwD1SSj4CXkTkXIs++Mj)^TpWq
z8TeO>w(5(;@$CNuDQL=4if`=WR5)ZGUyD+7VDOFiSX}EZ?8g!?`_f={S?b<cPRlP~
zrg?8#n-AKGi3Cx!Qp~B2vA&E<W*$|(g8zl?sNDWEBJxvb_2_;85fCF{K~j^xo1oGW
z!2YdRF&d{*{8YD;4SX|+Tm2>mbqP$r76&7gZ*}*;4C-l9i0#N*{7SGfDJ1h>ia9RU
z|AxZ*UuOCIz=h)ueuRww$s0|VM&A(}`rF#JC1}TT`ZIAhiqGJi23wmU@)sC!Pe?TF
zxVjR>R?sdSH)_BqsEXWQYKxGs>pIxDE^AUrGL?gcUD{Zc+=Y$(1|JhEqM!j|QIfG2
zDD4w)M1ZIx(TjAZx@+{fxm724R^q%QQt2Hwv`^_rKSLDbKVri6mp{WF{N#$`2<zD=
z^vO4PAVd82c;ARMHxr+YR)srZ$Oe}~bt=`t9uee$L@|e9p!m^J3UWUH=ISUxDbP)U
zB0Y;nGk{iW_-2Ex=i4D%rCz8v^FSg0+Hr3)NMTWFLm$9NP%5$-hJp%O{2?@aPu2?3
zIpA%#x-!!CqS)W;C{q%pY00DI1#kq&_b%Eg(caY6B~K%dg>JXD@}KynTukbo-URas
z@AEvu*jjK2xiQ!~r@+QVzTX?ayayNh<`nuUt}OEKeY!`^$ONe2w^#?)L-_X$_vL?V
z$$x)6Kb<i2!GsQdp>*C$76?JzloGrlW4h12pLzSK$w)>Gfj?LjSt*4|{$L`pG;n%l
zc)(MkBOm};ePL=Q>r_C@8ZhKJ@OV>T#TB<ml&t{X9rtyOj!NZXQB?Lp<@_H~lsK-w
zSjd>e!pY(f#Dx8-SFTH5vvNHkhJOgf(Eiq5vNDLu5ScODL>dJ%G?qanA+pT^4MXe9
zOIzJ6Tal91VVZantmkm;JX0E0)7?^7D@53`vKsyh<B+i$O}jF}s<bD)*d)M85Cu)M
zQ*!X?YbuWS8Yk9^tS@3#FTg}9%hH~eImJRW8(@c3r1Y}bm0uu{i|5NsWRy{1@aw5A
z5K8XNl9;xn<!WP}-xeLpjsTm@U6+t#8bwHgwe7dD?GUx~cn_>bA8s;(fv1!MD)^s(
zGh!&VdK<>dTo#>sUL_N#ntAN)Vf@e_U9eG@9OT}DnAmyGLW9-7!w5l(OVyppNnKDs
zp#knYML(8+TB^<n7s&-#OUZ@r%zt)Ru`g(GQGxp;KpfJ6VAmuvgpB4r&5T6fh0C|8
zWgbFRW$<0GjYaXYYDv9>HmYvjlsc*@#m)pNAzqa|L4<WbOn%PL%u4by=DZUt+MFP(
zi;|od!@DRUizBPiot0Np<x3C?GM&zZf0$k2*j?$}Cu6QzrjXWl$fiaXn?=-fDrTaq
zqTfR2`w^#&BRCZM<ANs&vD|C>8cFJ|au^W}KH$l+$S5TecwPg~k<<6*;W}R08EZwO
zf<xnl37ma1>UZHM9cbLlY5a;3Q6HR>qEAJgR@7VVx#s&}ClL0*_YiZk`PxLbtx6Gr
z0aj}f0uoz}X3E1EYFb9hat8)JoaPLDhoVfB=`$@vu)AZtILQ_qJwsS8q2hC#p<PHj
zpta~8p1&F&BX7c4UoFv+`guw_a667pW?`+S@#*x*u0}a>Evo;Srit!q4KrXBbwd*(
z`{{*IY`Xa6RY)6x8M<sT4Mw4k4<0t4xrYjTartifVpo`<4oVP(!tUiAL&J8Z&T-{D
z;f@T(KobEe77DfxXS6WV4ttKU`AtX*aDWG4?lNSKWET#iikv6x>uhPNP~ZO0u%Cb4
zBOMU^iw+?+X@MJc##kCV8rjgM!`x-eNT9MKeSFp_Q8%INDPthCLnxn|DSA&lyv{qf
z-s00s|6mBf;b<o38snWVw{=`U52d8wYVmhu(}i8JPRK}_&P=@|fPCW>q=C|`9nnIp
zzYjFpV*TSoXdfY%^ts8u!Q1E)hvsH~2L}{vbRcdAqA|+p)L{VjtBDKxpuBPba$MUe
zbUo4}F5X?<k^ClPrggk$@}(JUaTP`hyrRP>mjjWf#g@gkmaeG6CtEo*r4^flhJ0<s
zpb*jN>oTA)%T+JE5I6~^W7wbq+|UkA+@2~|wUUgH$roR<b+?MhSf3Nel`6+{cb(q}
z*JlM~LuiZY!s=8P#KlH6vpJP8D;9q9M2}UzWI&YOl-!FPgA(V+Wba5fTl#n<gH|Bf
z*T?-_WKHJJxImlBy?q4miqN^{ucXZ0Lpa`tMCH)wmTSPRj4;J&j2HQvr+Ag4CA)@X
z?um+tO|H{|%Id$N!tENfUsgt~176S6=i>YX%~!m_Cy!-s>)n4hyag9duS^eN`8}mG
z89G_}42P^bUDy176gsGEP9`gW)=c(*(MJ#5>=~<;N<#PtJ6w+f;m3BRKog|%w9o1|
zYBYzj;%vnmlshF>g*Ch113*ndb<&sxQfDR4oKI}+@h+xS>0~!qm(|jR_PnH?>8eJ>
z{W+aHPd$_y#93PRKt>^3=d1CicZ`!9F)7K3$?}~+!+c}%sn^9P<cfuDR)N=ZR0C89
zGTxXiWbg?<&3`q_RMcCqss!I{9ym<Nl(*xWdO2(%OEOl?WH#^^ADA?GDt{3z@G|z|
zQ}Q!f6Hw|;tcPtAB)9zGcwO{&0Ps(HnFIJ0HsV`IeBbV`vz^oB%Z2#c6qct4Wnf2m
zc0`%7<2$iLn=%pO)cPxY&uo-tdZnf!eFK9uTyUw9G>#(ug8-ZhGVC{FQ^x}CDmML=
z<iAm9^tv<BGjohllaiuv@bnnowkAi(r$$3Q9bPwi?KB}8=N5vIU4rdILr#&-US;pz
zD~zG2es=T4n8Ip+7>)Vz49L?9DlbqVW_!;sIe5s<r)tWJ7p1czh{9BK?og7GujLD;
zaj#4=>87S~CX*-%%A3uP5Fm2Fj}SKRy^0s#d9xLHuM@qW&8240Yg~0+=G`yHhScp}
zgbh`Dqx|RVO;>?lCoJ&0O@rPhu5~QThzp_-Zylc?@`%zOy_X!Pg2C|ZbD&l*D{0+_
z)$g$h0%%PwR}bQsc}8MWiw(DaFHboQKhOAGXg+;d2xv$jXF1AX2%<O*2SX9Y404AD
z5MS0!4TqM4*MVRPe_~SMI8&WMsX1mW*Zw3QRal2hdyA}A*!U%QAUJwd|20f%w#g#M
z=QdSpv&@?o*(m;%2WiUX$gOS+lXXKZk?G3{bJdHCHKvADo@z3;pW1I8Q;pAXkQ_|L
z&a()>c+`EU{3nK9O9=zH?mhwUKL~PO(C>}c<V_aq(SdWak=m*}-cq(=rH!IU-P4Ud
zlFbEPd}R?abx|aI*Ix*93X6~L-<4!MtD2*e{}xKI`evh$o2S4JAKf^ICtTq9X0c(~
zDu&xtnp@74iSG?l(M^;&k<s7f!6(Y>_s^8bT<|rT83ex}BPZc)|63*ZUs4&F8M*#<
zmXxcX*4tq>{Ev3<mLOcMP{J2zAU~^n0;I<_;RkEna9SF+Rd}-@^6f_z(P+AcG8dNA
z!x~^d$}FlFM|nqt$JdSE_3Gr|!R6!C%ISeN?t!V-ix^5^@x_{ng9pEjc~p(1=ZE0m
zvx<$YkLXhX0NfwFx$oL@Z)>7x4Cd<`dz>sdM8$}k-;^NUy%@j2pS?6ax{vWJ8yGF_
z9t|Gs=zaOj8&#_<l;oXN^?7@7efXWqmYt@6WA}dY7DHOZGy;bLO^)ov`NQoX>U)<Q
zC}eskBQ6dR)QPO#V+_z+jmq(evkuK#K%hdN6gHkI4j}k1WqQ<(JBy71p+K0ikIBys
zR9pR?sS@lFgHS-JCl`ZlIxdtoY;5u~c`&E5`6^%IGR1w&6q>ulF5#m-k!7&OaB~3l
zjyTrLq-e0jS(N2nplSA#=RLtb+jIJ0r2YfD!TsUEI;R_hCzI&_;Iy`6OH_}&6ejs2
z(=eTAk;#-uhy+llO^j@xPgM&8s9n8zoVaYa$YU4-qoZk@j%i&UNV7tO3^1^Mx%G&F
zjf^b{7io(k&R2|!i`;m2ibm=T#C;rvdzN*-;B2!<K`Qw?Evk)*s{@i`G?iiZsTzCF
zIGULa*Gv8>+L(`J@i$&Ir2^zHYMbQ>5K$6+&L)cP$DQ)TOp`0aMn>%z+!`qPCYq9R
z_4XyMr<^u@&ydrm|0FLsItWI};I3jW;!`%jJzcHCQ#-*PHOD#GrYzOZU#m6Ojqo7k
zMeSPF5bLQT<n|%bzM76NR`-j!4SXYXXkrmYpE0tGb`=Be^_l<Op6%4_d_P{@_;kY8
zeM=VN2BI<c8@^Jt@!}zna(={nD=Z*^@gtM)upxv~r}u-j9Q^$m@)3IQ^HH(Z2>hsY
zKtX{sXSuH5Ni`rXGArRF50<G9J}q#>t44_;7gnNO(wZf9T&6y)>s+#GpL-*n>iiMH
zZie=%@}&@benEFq#>RdF6xd*5%-sNl+F1dgF^Gv8iV?XSOA!FK-lTG;0mnl@sL2O*
zi3BMzoD&`GOwUzvfGXQ^=|XeuCm6z(PKPN>LS!M9how&RglJ5@Hd3yi|2x&EUdL+M
zao1@FivikwYd6W5I1RyYH4O?6OjWVQaFq*RKn$*)AKZ{@aSI-RYLBQ7Ax6^F&VA{y
zHy0i%Z`|iR4_Jag6PqwT?ib6x12I>*GUXsz=vYyM1A|6zGSch2Q};>H<~TSRRI*C}
z839?i8C;2duXj1!o*({UIRru?&>d$8W`bBkqTU)hG=+*de#u$W>3ESG_vnfJQXMSq
z<><^})0ayyDshue%ah}x^)ajaOxLKZj&tpVVx;q@b$XN}QoD7b;1+ryn}_+Tljx$H
zrM)<MQtE>cd3<|mPHM)%6CQUrLyf0aE%3Kuzg)hN1y?Hi>BBzO`<o<?uZ<358wQQ#
zG?Lv>-!y+lh5R(!loD`!yZWLy$q+Ys8{1angpO3*pzC_q`Tk-eu7Qi`!ee}H=li+v
zARqv74>ioGzCMMXR!e&{Lj&P)sgYvWH-N0$=+N`_88U{gf9T!s<98{rtv5dBLdca}
zdH})IJj(J)DhT?-?4U*gs3fLx=Z|cSMUz%X4=Qu7VQ+M4F?VpyA_6xD)Jq9p%&yJh
z*(oU4?2z>C%?L6<#j^uhDQGIYwuuD8Ux{h_<%f|IkoQ*)7`0#6y5ue`)yoEqe_lub
z{7#I_?+&>IpL9yaXKwv>nd6w2e3@qvU`f(np@xEvXz2y^mbYxQG2+c<+1i?(zZ?qI
zc7rVqm9DO$3y8%M<NeOAqDyc+HA=O;<T~*ovVLCTA_bv>^a)Du`k`gJiG*DVWgQuZ
z+3JkGET0Hr*Q-OwrRzz;ZKL_gY6?di5tn)<t7LT(DK@64Dz%9lzRY-p9-I;^*1|1F
z_5zc|r?G0cgJiogOXx3-U#eSXIQPXYxjBAm7L1w*3$tQhz1qsMx_ZQv>q=b?wLv+F
zd9hh#s5~0uC$HKdr~0zn)g6BeEzTdKRF6d{EsmgVMkBgQ`y=MmS0bKO(~jxVvCHfn
zjacoUsC%9X`$ZX9KUr^g$=X3#_#-;*bsg}r^%t#LV&m=fD#DVAm!@X(d|H5RMsa}p
zv`3fW5?~}T%;u01yfe8dKa|4>l}X7d*=hBJ(p<h+v(R$DDt3_$V7UumMuHW(v%_&e
z<T=L1<rsef&2>h|<Cb;SrwQn2T0)t<P2N-ZtLbFx1ri<^_8M{)u`-P<{FjOEjP*J_
z*Cmr<F-h*<pG6cXb}|fttek1c9wx2~8t<1)6aGzv3dyqD=jmeQS2v@@0(W1%Mgsc8
z)XOmG38~pXqKS!i+LvibE|CoKlE*O5&{X2e`u3Me;{oP*jK|*@(=(D_oCq~wee6lV
zi<9E-RV-MHxcaoZxf8of4XB+Wo@}XKRR`U0V>fAg=}@@urIDKuPpRs$OEg_5%i4Mj
zFi-T2d`~}-z4`^)#8f{|4i>5f^>3S9;B!X+FGX8atD?tSc!0N(`6ZJZ+b}bJ+PeJx
z)1^L{;O%zniuMH=e<(j9G>%+vplgW{o){jbJGlo4w3}jq@JJjupUqCh<Nd92j>J2;
zjjpCUyKcNwkGeZI{%vcYtoRowy}5DsVKK_L*uSW9e&npelkf121T;5n`f^h|@bAId
z2kj#0D|xyeoeg?ek7b_!w4*tlpEPPGNA3j>=j|?6i_ce7elb-<V@7)mlxFpMZp4C~
z%eGFj+p=tYWr6<tF)uh$-u|}&gq`L8%mx2Hng}P;|8Dkor7azY^P}$iF)wU33Wgq6
zy#hf9n;j+uZrV)KjT15c=u#m6r<Ww!_-gU2tW@_>kCne-6-k@;^ilbDD69^YfS<R!
zo3Fczl27AUP=l-67YppG|4!M#hLus{CmPzw<DLKGDZ<(E)@wl8xfWFqy^p7R=u8p%
zhbdAjosD0E{=?IqD5t;%^$;k#VfN{p_i*-F)`20Q)acxOY-{V|b<n!J*bwRCBniYK
z37gv?@WZb^2qSv-+m-Qz9P$tQI~@#m1vwjy1lZl8`M~iTh`ssxe45G=_||2_ea@Rt
zLClmy+J@g*BC#0F8Q9x7U1O|iR{mpvuU-S`AVU>~ci+Ew>*+DS$Xw9RR<k2qWxj4n
zl{Fy6cw}>$PGOOGj}0=in0fLJU_Vw}q1w-#Wd7;g2pbQP#LgA{PuAUk@UzYyugttY
zq_sf_>@JSH$j=bp$L-YV=;*5Ddp34Tu?6#v<ol6rxCR9+TlSX8(}d>nnu|<PUKoeM
zO(}dQsb)JJ(3Yk!9aKB-&*1zn-Xp1n#r^#UoCf6pAyIHg4##%C?~j~O-w{QEOz*-6
z%3>qz>01k0r;P|f8ZG*E&8d>>4i3%QU)g_#Vd<p1EHph!xnO^Kw#fuxW|Efnajwn;
z?kr$<F!_hAScz$DLr|uUdfJ!`yMy=m&E|I$PKKv;{gC9lcsa?e1Gb{A-AGZl7@xSg
zBw5?K@xacHHD<Gwx%`KUuvJbVY^pCstlXlEo@`oS_Rmz@1Dm1Ozh>VmT8&wr<#~Vz
zrP@*a2JqRRe>hlo7t+!N&7I7DWOWwE>%y*`hN<Kq7IgN!HMH}$=g*%-mLK3+k9fva
z`K@EWBnE=#cx$TXyQ}YYjjMrZyr}ol)c0MAfL#%RD6l{@mQnYeHq$Jw=QNcp;bra9
zyoT|xncL-9B;&O4kS@5(P*DmA+Dn|UL5an>{kU^lDf1RfQ^Gf(v<1naapcbxy_Dh#
zEi&0Lmxq%w`z6!msH$MdQV6MqT=<JZQ{NuD&H#5hs?lE6)JDGvrC$pUzJuJWa;ZJr
z{NWGmf`##MFdIP?v92*4lsA!C#mbT%BD~!-8%857y+@i^A-L?dGel8Xq7aN!XNpmY
zYG16k=2;?y{+9b8c?PYVDIieR`4wy3+T3OWnJ9ERh{oZstDc*`o-?Dt3_7pQrGA*2
z7^?jpyd}mJQim%R#U)AT;H#V%6OO)G(-9kt3Kh*Itu~s?w?HFdz<dAu{Nnl@Z0(gd
z4IZddpYHHJ&sPBb$qT#}KnM_rLs~~cO#vR_njd2L#_Z+wxr|5@HL8vS7VrT=oNusx
z`%M*jLMd(shiz=F5H!ZsPqgUVdAi{pxtliW&E*DO9Htz3N(nEP8EceCcc-U>3%}`C
z>Xz?T{z^0n`~HwtoOPbkUNy9RFTQzSexy9YbDA;FBdbR8Xu(i3VRi7n{o(Z4jjY=>
z(ak7OCu~xE_#5V<wWCGIlsTzG)!}51zYpSedeG(O1l1L%WZ&_twj`tPG?`6C%7V#Q
zLid1d3GeD?!6D%@8OK;cHo|cpe%wn@Y@+N^d~%64hIfcE8yf}^9)t+n{k*~>ua<2G
zyW2!d`RwXK#xrKH3^kWj;m|dT6z&+abTn9f5v-GyV21I=qE(nVf9ZZ5%`%Uso(ew$
z$>y)XKpeRgDUx#0?+qp1mO3S<hCg1>|HLE8T}pZ@Tz(yd;Eg50Mdrs4fEU~5FF?3~
z<4s#rUEyH%X?g>@V1bPCV}RHd`@Rfg#1w#aUx7vsfo*-KzXrLJTqWF@$}mY1tsSE~
zN!%S;^-=TYhgSI=n+!ehr&te<7ZNm_<IHlrdQVj6&JJRHbQ?Gg_O@rUV4-K&7$o_T
zwUntmivg(UGxn7M5o=`Bc8r$w&cyVibC!H3p{Dtv|BzQ%tTG&<*S)8?H=m@E@jMO}
zoL%9`eg`9yL#DU_X@M*|hh+aFixVwx#NH+cHi{du%cyEaALtJRaFcuVG;BkBahikl
zY<t|3U5dAIU;Ds*{9<mOw_60?6~Aph;OLXDh%p32WjZJgBbqQ6Qn4U^1EnlfB)aF%
zMpJF)2t_Aflji#I^{^o9rCxB<7kj8N$2Z>iMjdkn2Ng_j&vfN`BM22_)pJqp$Uc}e
z?9BD$dQbujL6W}<%n9Wba_@#hSw?4q7GaZw*9CAuu2I^{%`>JUGiJ<_qcR*#arUC>
z``8hd#?H$nL}obFfyqlVED4S|(>KCAeXJ)7Tv-f<GSHIcyHeKTHkR%Ov=Zww(XYtD
z+*h}|_Aaw&J0+VO5qT?H5!~wz%+3=ZP$N)J(nXMdc9Q*SAWy{LB1z1YO;;>UUseio
z-ZNpu->LWs?y1S+r(Hc<c(X3>cC@u3%E8Ve)v^>DLo$Z!QW$-SG~BV<*LH?I8v;&!
zzjA^#Wef*QKLm^~De7>Nbx^zar_)AXQZX1jI#8QcpL$nKa4V(5Tx%UNy84r<W2aZv
z(Om9QIL#1y7AVmX&vIM6HUBv|&@5RG-hIga$rgrg!JIy6SEpvWA61Y?8?x)AU)C2#
zPESFha=|*LYZ=eo_o8b&**!Y$eiYnRFnti*F1GQYum+kBE`=D>Uizbzexao%Bac0U
zL?BDKIy6nUN=6stXH8+|qRu*TO%`+{nd#|kR;bXX;%pa9bI+eMl!T_2DREKxG!{8g
zQo4?PerFkw>fMSZ-;J*P)KV$4>4%%!v1B+t&z9^UJ<-~Z)&d-&!r$?@H(ryj$!9Wx
z83m`jSFc*G8|I~M%VZm4B`;tW^>$wUyhWB=Kd-16{9IHXj658U4Eyz3miG$*;N=%?
zGM;r8FYw_~^Gq*BE29A>o)@=dQBI*VuhK!qJjMS4AnWo+{ndAk2kDB7ma8Z2c(&QC
zdDm_3gXbD1ih{^FDuAedXV8c-V80Vp?j3=<r3EGvzUlvneD7-r8?V;NrNeAnAkf#V
z4Et4xYZM*Z=t}*4<1Jp4alp$BRdK_Mo6;x&e#chmH^z&179x_@bFD!~Zc{c5!Io9S
zHgc{w$_}qr**>Oqh1Cb%M5DbJi^cevYcH2}Yu9nPM0SOBT!8bo*x;!wsEiO>egBM_
zhJfb8+NUs5%eTjJkXF=~Jy5D(sy5)07=D+50d=Lsyx4l#Z@49vTt}9yBL?<2Y%1Ga
zSuae&8;KQ){H!}8bN`9UDs?}7e~>5!>WF*Cy?5@uX)RC89&X$8twau0lAq&g%gd=5
zoNt})IvBr*2=9>-EGQmYI_B$dE9Cp?Ks@@mE=IR|TyROkD&$AHbqQC_^20Hqij13i
zyi<6{04wok%M1<7Oh+Mk(K318ce?ApWDc(`tru?H<qx-(YZvZ4Phb>-riy*|jps@a
zo?`LUHq<Nd()!Iv{Td5@8Asg!)6(3PPr1sACF)n+aUGCYN|4f$EC+jSuPXZjFPYov
zV5?$5gd?@|P(eY-&keMe5N4Ty1O^i~R76@U_hgR9b=?~Rm1T5cmYQiqdMo#)ZO!HD
za$Ag+z1~7MjvZjptK9w1|14gLT_?L&qF11vBL7*EBG-8|f9|#gUxlDmz0e9df_?q`
zr&^9MM!Xuf^()YENu-l5zx$MWSn!?_10DKKgk6UMb0j+P9DOp@L%#d(t<kjF8>Haf
z)hhPHRV(PYtE(^_Oe3&kDHVtK$j-lF)1UprA-?YZTPe-)e`m2XvvU0JpuRKzuTkP%
zowAJ<$ysBB!vLzD4*=9*x~m<p96w`;xc2QYkwDAeMdm84MG>&^^Kl#98TV8j?d?E%
zzTQ1Pzdc?RKMp|K_wVgJtb)!XCVu@X4g=ug2GqkK<n`fweG!vsb#9+7Dp8B370u;d
zAHP=jFG0=w4BV|*=2KAP)_Co4g<jvhv5d}9<l4v0X%PoY4BVN<eZPKOB$=3*&sHx5
z0T{AH<yaj{BRU&GvYux~?3)t{QesZ@LAK7i_(yhP=MuC*T{@VDkW~MqWKOX*A8{q5
z8fvl1_oXJ3IviXiF~-XO__Ex_7!G}5S+@^_U}V$O#qj(nNK@tgzw(tI>)j$xa~yu$
zxVJG4`A=sIb?4anjto-w=BrnCbh)ToXb<*r@dkJZ=h+feaUeap8G$$5TLvV79`8)P
z@2@`p=Jr2gV2M5aUr+4i?;vXg=V2hVhC_#LzPUgCJ+@^~vQUN;W4+}iV6E&;$g9N=
zS`sX38AGf9fo#w7!@)rJUW)je4{);)B9ogeGSN|4NlKZSMc!G!xf!_|6}ucZeMny-
z%KN!Z8r1+rv6v=bREkm<<4qT>1R;6}77|*f-45{ThCNc}O2PaMBsV+`=PU;uHeJjd
zO3K4Qx>^10sw%sx_+wu9O|x$daNxnqtC0yMr(PPnMiB)okQ@wK5tJXlEHDoqg8nQ_
z;#eSQaq%lOS4$8n0jJK6-(WZ@PngE#ML@9?B}}?wlz|J^6r2?@f~%OShHF%39`agA
zuIBB?H?f#FO*D#j;jwof`CZU7$i+c5E%G7pOvIi<?IzkQZ;@1F{k;eXqv)BTc?Jst
z+5)CGg@X0~2=cntZgBLB{rc5??>tmpR&Tb^Is+MV@D}3NLY2A-^%e=N5KKB~C=)~?
z+qULI%2F5!Jt@LuM|#Lvk%8*E;Bj<HV;eq-d{=0N#z{IS#lIoN5TIje{<SLS@QwZe
zKqYx0)23Z5?%9ze#SDc@VexwuJ>;T=L;84Q=o_wJ6|IveGPm61Jd*FbOxbSAg^32j
zLLlc1EVr^T$ec#^&*qat=sTj?IS^kgF&F&T?wV`)z<hpB;x}l-tU*}lr<1p1wpFkY
z9oQBGw?+?#?hWkuu9tsBM&uO(-9*SNT#ugP$CKdWDRg4Hq9tb5j5680d&y>w9g3;K
zZHlcJNNbMv&+B>sO9{IoN?5p5(N#~rbWwHP@{zGmq&R&<809YA^~?!i881w&O@|a(
z#B55(yo2q6fp5-#pv-caI}<%~EK;R_Foy=S?ietN$p~)H8DnM-tXrV#CmSBh-FgX7
z=kI#hWzuyUXLZ}1WyO54>*XTu&K~M@GR_d+MdqHeo?KlD&y;+RFe|4B7@uQq^z%)l
z%}BwROH4mQ$Sup}i3d++W79clIX=c)Wb%f18vXenOl!W7l4P5{ScJ{~>1msD<O@)x
zcZCwBxh%^6edW4sJQ`+=c+=5~L}r@v4Ll^S3e@i{)oGsUiVVL5gAHxKo9C)Gajl5M
z&COG{45YDUr&#sk#b`zu>H@Ml7XDAh(~Qx_e3G#XPDif@vz`4rvhc3^A+nWP1s_VK
zi_%<l*3`q2I$6Qv0t|qDqAYg2MCKY$@~cb}oMesBoK$M~!jr=Di0&HnP|t<M_ELR4
z*zQ=%y>j~&3%Tj%@Z5fiYrK)h%%LPw%Ra22USar^`Ib<?Ko?tO&CXgi9;hSWEzQ+r
zW4VUiYgm^x&O&WvSe2z9)c2^A!eG4Sm`BBGwI1mDAonXbUAmXaSqc?TzujEG(MOJ;
z7WSY7O2qS?F&S4bnp`0D^T4`_WOdypy@y?(gV3NBN+Zab$iVh1Ghk(C-4RSRJsL%5
zqek3;A~(4JTHJD?`8+>&MP;=dgdP?})E`yudTRpRka(q=5-)$;bItR}nOs}CU~?V^
z3viOC&k6~}DXHM6q!dr|mST+BSXwND=*$}kB$^q&3a2YRw2a(hC%+1bBcuz1=c;m2
zo*})J6~kTIzES(W@Qk(PbU?$C;<uV%j|-sb-aoes)%Q*VxVv=zeT?jlkB1fI8g+^M
z_9vf;pIGFcjuuuVo&={l6#gOp%-{k+F;=RIN_)rNxWGL!aoaaaTj+DExl%;c&Xuzt
zIZQ$p)T<_RD#<tI>*ey9x_R>FI5<`oOdC^8s~y!W<@|zkj9snS?YJzK64Vsy$27Z7
zDx^qL+}&mWyJlSsRHKyVLgCy(F52R9I{9T<wSYeBMUb3FJM`4}z(smD@0lXj1hExO
z3m)4^#HCTF3<FrTDR7Eyt}wNWZg(^OTKyOpsb_(_v2HY+(=s*<awyO|SUYI_I!E)~
z=p5TP%->N=n><~qz4JC_2^%&t_V-Z#w|^t4Ke}8wu~o#ZTw}zfI(Jf9Yt&1H7juS+
z)t=#11J`ot=YD_{gFai;afh2+|Ihk|7QnTK%U}P%tLS5J;#ufZ^1vJORXE!Ol60-V
z;hn42)AY`(xB>9@V~myHMoRuP!Yxz18Fnme1r|?<`uz79s^@o|GRG6&=AM~?vEG%j
zuiJCGshvmV*Vom>ugtrOB;WI(*^EDr9G5!~41UDN=ZG3gycrR+j>IMri}tSe0#9f6
zh>4Tv8--(nvS{kU(yZ7uXde?P@Z6?>QCQ#CfRS*i@1s-$U3O!SQy3$=1@71+?9)T!
z&b1)a41(j!i(&MJO?h5B0hSFkMGxJ}1f(lHu0GV(6oLUcI5iThHSfZ7CqiJ0INS6j
zZnsSabHnxb#HA{9tgROmDpw{VkP9{Hh6(I}tN>GrE<6e{)QDw7QbTFR&N#YzjB=@R
zy}$e4AeyfGY6`KzZy$1xEO&Z61Jc$3uDHT3NMtfa%8FXTMd~n8#DIrW+g<|!dogHJ
z^j*MUtkwxr=phn1<@U*+fNlb#Cm}`tBu`gF<~oCJg<ACbWrBGUQIgaAvXIul+r0s~
zlc;_V8}mp$=Sgx^ZXpk0Jz#Qn^K@>jWP=+rle%51i%EaNz%g3J<3H4?4$<eH`B9ms
z=TCG|(`x&?cJ4M8X5il%OqvGzAv`*If(=>}%D>&~YX3Bv2t0!J0Dj-;Id%>>%i;o6
z^ZfetTYO-GRm%zQ4WCCs?YH9c-Mj{GFIYzX0Qvb*$vEW5HW2qyH~o8j7Ns=VV{%Tl
zVsGE=t`Mfw5kAE%$-o%a)Zvx(ia}rf1I`19zqx+h+(-a4p63S7GZ#n}dMgHK?ET^!
zXT|y7%D}8^6aH`5V_=ZF|5o#H{@+MJCQi2hojJBgizn`o9q#%?{S~oV?tg+DHm9$_
zB+`c5genQ$SI{QD8H6VaX{U0-+<mk?Y^Ip0{CvWT#SsM{_&%OpeEk18k+XQ9(R$!+
z_aXKpvVI303z<VA=Hm|BemTEBh+4b&;Jz_xq1^4gao?`t=R$+b0VyMJ@_K{v3g7LX
zmUJ7x`C!*E>E=!Spe$`zg=J8?al`ty{rwkB8?;2z=$9(`yuJPZy;SITlGG?1+`i0}
zJ@z9}!f>XSgUsLV2n@ZH33yeeq#dh|5>$nK`@wCEw}I#_!{xfou48%RV=1z5VxX0X
zlXx7fT2BpPqpveUkp}{?5cOq)Rx7b)t{8YnrixJmQk#xhoa=kDmpJdaV&_m!x`P;<
zRLLhu1T?4DprWW~1Z|TW=IeltFY?SFk6JFwFAmzxe5pU^;Fsi&1#m21)n#lIh9`f0
z(GsCnu+$a^0ss0HW?m-FLff<%9yfwa@II;)4vs_1N&XO3gmv!%UzbB=r`e{Of3O>^
z9_pj(?uN{N>Nt6rd{?04hCx|}^)IanUPOW2OWgX}zDRjnZC!(wgT#m0P3?B_MMJ?o
z<E3WTi0@*tK!JiA^TBDOFykK!5g-Fn=AUU+>O$k-C0-S~BJDsKhs0^S*9nF6M6(V}
z&h)RmNgBFs8=E1=LUW3D)Zpn=VHSP$%r*KA9%q|wG}~RE(cB4f!iKn!(>82z^zoz@
zT$F=fZ`$SVidMRH5Q>Hx3UjY$R+id~$%8)Y5HdOms$x2KW6lF!I^yrF(cH6P9!>TM
z-{4Sc8?KkGoudV)%TVGKi7!{dlpf>!+m6=AD<?xN*c5Ja>@n%=`_oL0{1X!Ps|UV0
z`Zjf&s&!T$<H_z1JYS~k%F`Y9oRVM`g`nf!MRqWdxK9F(7@s7gUC*umu2to$X`1NJ
zo&5?%14gemf7irdnzmXOZRr4vP9*zas~QgIsqa+&AI9E+NtCGF(k$Dyt4`UrZ5yX-
z+qP}nHcr{LZQC`~J<}aMaleVV8Ik)J?8tnv)_S}Eyl#Lw(mPIp3En5=NY|ro_%+gV
z6*QibIRmf5I|{2&w@cSPNHfA|k~m6M(|U>aa5ZzvWE8O$3hY=k2Y6T+Qw_CG*rA-H
zlVZT<>85((kH%PoL3PNn>{FX(xSEZ{Qq$(v^^5i!({reDvcxvNs-hheyQfU?a`uYv
z)|LL5)XS;W?1$T<G06>S8Fv1G>M;MwAWu`;J3~{J?&PJ$w2MhuoVGrQu5v005!t|)
z>Q*r5S9H;d?;H~T<WR)MEhzuI9s`A{y;yFxfHq5ds$T|M>iC-MsHRXv(W`LqoU?ig
zwDW6%Ud=MZ&x7Bu(d%k!OvOaf&vR8{xXWt7SrPI}i=-~mcsJMeYbtj~RGiP*Zk~}z
zK(Xjpfw-*fD{OKT(WFpmS&m1^0&4Qcg&duZx>kacs1z82;6h8l6@lbl+MLOR7T!ex
zWk0(hFVv+$_=kPO{n%(;A5(9(NKx%UW!V9HM#G)nG@rw|J67gS2G7j@H>rJEv}bF$
z>mb5OZC9NLU3n0+SzEI<-!%nj^&1GM9V1i*Hb%!T%h4NUgrk|ut_Y1!iy-qhF`nv`
zFo2Pk_I9HFnwC$L5em{1%1Ml<G<GACZ=2%Adj|H><Ka&o86?_h&T}iDY?+-!#U%nd
zSA@IXS`50nB8re)b!70%Yd?w(hpa_!p14S@GkjOmkB;?%vohOj9~gE+o*iJFkB(ki
z^!P@b{K#GYcD_KR-FecH?BPYxF0ALX%)p`*fQ?*d7UCpi;$foz<7;{EJAo`pd@`H$
zxDH$dQy_BU><HNqmkeh00tZE;V&^D?PqyN}^-9|2ap#!d(5l3I-6My>K(YDONK2d%
zm`+QSQXxOzvO7t%Wavbw9?5d`dq<{;%n>e%nSiK6>Ge$wlbwaS;-(9bkNgmt*XEF`
zw12L~w&hU$i>JSK&w1@z#fsLjDmLkLw2Lktpw{!qW{6c~_R<tvrMmW-cgP2@G%v~3
zPnHRC!3A`}QLU<D=3S-;1j$^W!i}WME!p^Y{uTRrnVEd@`<gn9i1?WGt{?hH<kFH}
z!8Mo{Nkz921;G87OaU?bk^Snz-=yIS6Ftgp+P|HI?&Oe;BX{#FV%sM?v~i{5y&n-z
z>m(x|WMyVm5h2fuUs-=YNc;2`a=B)DU3H|-WIx8BlJI;S7*_4Q>o6|MwzQOe*%t((
zBZxnfcCK^61@+m5K8>oKwDA`u#jcd39`SmwNIk^%&{I!wTcLlL(j=eCQbZY5&+%C;
z?BYRlE2_S%*WE;J*RDEas8VogHf}FIdr6y*y#LhzQ{2{<x^2`*$feJ+N8E4U6@t3A
zLd>c`-`hN2&TwO^cjmpSlQ*cXqdkZjBcji1pUQ_fIzhSh9gk5olvpdf5kvizKB`FR
zm3a_=I&FQ%?CtK{@KZ2ZsSy^N+ZL~DdP@=}@oR9I@$!_ldo$&<{})O<%2b>A1E)oE
z)4`M)`vcr$Xx;Xo{>2=u|7DfQ%EbPEyBG=RM9r<7j2-Aitqh!ug^dkujf{DDp&XqY
zj18=z+}5jAI%78JVJ^Q=-~FHBktoLc{`I-jVH4;CTm>i#Tvt#z{OAQGQM}c|wX6xI
zP~S1&#2DM%z-_MBbhyZ+^3y)0Gq}5YJCcJF$bqayT_XL(2ky8C{dj}fEsT-?0}LV{
z9^;&tJfP4=gC@ad&q2z+3^It?Tk}#se&~Ac=t=X%kpsa4kB+lCyjywF0tZ8TCp=dy
z{n|*id}-YfIvL~<aOmaZ<jM2%9@UL0kB!?p!GBry-U+oH<M%oS5B%_Q?aam>-UjH$
z8yxL2fEOSqMdx@C0K)csgz8hIkys+3j%)DCq|BwogeOK4$9KbzFUBW9)OaPLb4)8_
z+<>Wku==oIn>>jqB1iySD7%_RuuB(!^tEwEL1X3W7X&>Ox~3@Q;h&@g#WYYddH8$z
zC8J{3^7xWz*9x5DIlLE-yDDmA*iG-nI_&JjqAHP|{j!u&g&eW#?|i@eA-?DDd`*N$
zCAz#&0S9N}p2W$@Mb*x{sd1eu?Rx0&8Nc1xnZ}w8wbp}v6?2$}-rlH~{MlU_QmE;9
z#F~Ms>Z{nhxgA9@)ht}LZEV^Rju075cVSg}20^7D6piLtejsWH`%pJX7XixS?-4kt
zay3Sog|&}95PxnHo)F#Ibia^QPN<VBFVt_MrctG1{^`C@vn@o}WOgzStTH5%D|E`i
z3X7WDV9vMllTe^vYddD8JQrNun4fc{szd(tRf~;Q`j)>0Bo)2+9YH!XyS+RPQSzIy
zVru4MFnnMc<gqaVOp4Aeqa(E?mXYSTB#QwrbyVF^qV%wHXz7N2%^d=1i#angT0q#j
zdd0}ZN)EQUt>T>W^T|i1-0$&U?=lPh|7q9#Uu$_L_W#~2v0IbF@vsAN`iTm|0(cMb
z)CQ-eic*5gI+>|9d5!MY(Om~_S;eJ^ZwGdtpk~ib^d5|fIN=m{OAp^nlJ-aY@cu}&
zWT&(>Q%`q{1_}hza_ausTH#apzBNIWT_`eM)9+AhC$)D|sysUBbr3%(!H+NH$K_KB
z-qkNgjohc;bZ@1VkL*~~EY!?&GQ5gElErcUQ{j^#o<jj8WaiAL%@oPAFBmhv!DAS=
z8-2N~|BFefp`xl8MBh2PL(WU4En$yR0(yR+W%k%vq1M4T5p%O593>-b;a6&o4-PWn
zyEb^RRjX$ck%AY@+MOWKuHsJu@MvPHlTNUn<?GEjQ(?Cq;w8DU7%$0dI>Zt(722tA
z7Zg;5B91T*uSt>2VJR8HAR(6;t%42prC)Z^Cfjq5Y%AAcc#yg^pU%}e^+%%iuCe39
zJ`j=@Z5^aYW#g9QM+Tnq6j6toD(4My?9&J|TbG8_Z}RR24PUxa?MwR}Tkpav`=$9k
z&RvY6PNABspsBJ>QIztD1Mt@}6wJ1+5A0SO)ZioF1mq?$Z``cDn^q1>E$RjxgJp37
zsxh)C*(UDwaiCN{uLKox9gz?r_y<bL5sHppy|Umn!+1BKp9@q5!7@Cc=r0rk9AUyI
zSK?_pCbKO_KFh#WYgKCIS+V<{OE5Bq{<E+bt#pY?<p|e)&IQRgs!j_vwq0h6D2sBR
zODLc*DM)aE^99{|Vv+tnzu-8;zb0^VI6o;fGzB~mB;9!}|KvgE$P|MxcDd^=0O&y9
zQiSaLaJ<%$fdiM8La6*BNq#tw{WQnz!55Xmac3dN;pi9iL59qpvpwDyvRj^?XF6MU
zyCz?q9knK1gS9)do2u))AM8aD8veWqEy`U%15+tO>}TA&=3XZ;+>IPp-i#2e16li#
zM<d8cptENKSSN3(h}=>z6Inv>uFBX+tZy_oWD&8qFEh^gWjHdV`b>}*ygY}gB`stH
z)h9ckEsirN&NeH5<f-aH2YVni!Jo(Vq6T*4|CJBCH|4KhTbuN@iR9pI+HYfOdAVQb
z%&YcfU2M)ZrYz~W*r_v@B7#LrdR=rX8yqA1Zr_o*aD0$IrRSqE<{=y9dPJc=!}q(E
z5{fU^wxDr<&UTsx?cFd8)}F-RYc5=jI|?bY2y!vZc&N3{HvDPP`qD7n6UyF=I+wZh
zvcY$;m&9rH>PtItiecY(1Xh4oivDnLVwZEOf8X$71Uf9T^r)-%EDaCs<hd>0pz_X;
za8^|9wg`yYr&Vl+g#^1k=uf6^o?UZH2V~21#oBA2b{QFo{#XR-l$L8(@P%`|?NORV
z<Cqyp9zy}VKyAe`U%d*4o>z)NO}*wZ6~bXy<$C#2jwvjMh5vMN4HEBMkNxz{qTpo^
z0uFUlC=?3C%qs^e4Hr@X_M$!cE{wlFScosAQ@5$K6LREj?zRyf8p#r#a_KLx8F#nr
zb{w=1Zz&oG-J*@zK$wx?uQpe@HIohQr?@Ov&;3dSbZ}m(*0Ax#_e|b-&XiFUypWyJ
zdznE0w+`UK3sLNE1VG8nTjdwMT>Np3;@uCNL`K)1SN6fMGPuug`Y#)p$zEb5`Q*5)
zY`W!QGk?<COg(Z$jDu&W>o@GMeh)cvV|u%tpv2KTZtzU}iNb*0ADkygiE!F6G@*f5
z%(brsTjwBlwA=z|Kir~+_|T!sIUS5;A#opE+{Md^C`k)CQ8RL0adDXy6dO|{_Rr^M
zn%9$dzSThfjCt>RH@1U>?aXUY6ZJD^6!%JcNhSeB0e3}2Y$#k5(yA1cmvtb)i@N=j
zF}-Xzw!qP)t2DR-B_gFbe3-z8zyn29TLOmR1>wOtRA+K<jou}AAXIc*exM~s_O)WA
z0U*b8)hMJdr#@J?pdklnixAW^>_x_L%5V7dOu8ZPH5}L3?5O$!1B?no>u@Mds+D6u
z!P?nDLM6p9mT?UZ!M?OhT2Kw_#BPy+DCab1Exy#x{rO^32FG=5F&;HA`Wnx`S<)9^
z+sE^&6k$8=Wh>^{K!$=)9yX|Riqnt_nK6ONluz!l55}L>wS7|G;&S^ev5W3hz~-O?
zz}7K0X_hQNQ-KN^&!CoB&WLJtJxMZJl`H$qj#{DkHrku+#qzK#R_$y8+xFgU5r_c&
z7hA*z<kSa{C*Ls5YbcqIn2Ix%x(Wk^969rn&0=Q?O=q<D;?~#GutQ}SRoKq&-LONY
z@4=RbFlt-!wo*2K``5R5o$ju!Pg8(y0YI$OUf*8G7ev9O1#N4b)@iFuU+Nv3??uNc
zTL#QfsdpmGtJTf7jBPuQG3=yad9EU*E&*O0^Kwv2rG8q{1DN<ajcU%s>KIIp0u->a
z6LyKiIc1E;NxjPVsqA;1d<Mb0xsrpbqb^k~7YzXMP7l?mfk~Bn+9j#^vo495H6;#;
zqzClX#A=iD?f?-`fy%|>>6Hi8o1zf?Q|u+wmZPI}$O(?Ur#>3>aFgYO1gj!R8MZF%
zOSS5y_vgaI(z?bOQXC0&9yh|!6BV2{b*rtuIcb_Q|4xosV5q~9gQ%=o)gP;@X2n|R
zj*ZW4s2~HnlRGq{n%diCuHs+Lc9hzC>HYH|kuCP}V4E)S6Q@yTWE*D-U70`CG{d3z
zt!bU)Hf-w1m1XkAX8D83Fq9@B)cG=H7EQqsJ8J{%3VBS5NV+wRUdTyO?!}v<5|3+3
z70&)UaDm?eF89|ZxmmG~GD?#+yeJ@xf}JtmaSCPBB`2)x9@8WUIB{N=HnqFK7(BLb
zJMXN+s|$Ai1&1avHU6h3F5gY4P3OC2s~f^Hd>^}tM;EeJ%R4}?AIwco2#@7c4GBcV
zOq*u*os34cKpZahTMkLter63P<ouC@T_H#OwQ}=);X?LGK+?z9S`F&Ez*vp32mMNV
zcSP)pvL{Lc%^ZazQH!-(ctPh7CP?L%@z$2Y{nhcQm+8nH$03)P?N7SvApsa5O~p#n
z>WRmd1H?<z1oVK^eZRz&0Q8v$a@T=0fIf)d1*Wbq2>{H`$<n~FuqK%B-t03*Cr%YV
zfuA&f%)Pe+M^;rbUmGeFs5*CZ@`Ms&>yU8I*yynk<9#@o^8xSF*1x!o@#Ufg$jNrh
z1V$KJ^*-q0R>)o%{GBj2_s%!`qY#&*|5TYV{D&<s)31v1e?I>|UB>_P-u|LnPktwU
z!c^gl&=3K#CS=f$Ekm0Tco8D59;*}`|HH6RUTVFb&V30x3&>q>9Dza+{wb{KQk#9r
zYXe7@2`|<+{fiM<*qTU{or9M6NiGt#<ov~m<aMOBUaw>)t<O8>K^KGMP0V$d>NV~J
z+zaQQMl(8<e=X-zdh;(7pEk6^&k3y>^S;-2MP8-6x=|VO9gRQL{ahWLL{kT*5Usjl
z`@SwbOdM~7RO&~gNxJ6lu-4-WK;CvNJTj(Tf8cMWPW_o~3Qt%h{b6eO_+GD1(Tji$
zB*B)njK*A(mtdrJSq1~VQ)waOl`tPQaVv^+8ZiSNa%zepAwN}4N(PD56u%8ERBKd>
zH~VK-o?!w#_8wx=AJ;6?_2&XBQ4B#WVSUStB)ZZS$e<vysF@!INsYut{V`$Wi*@F2
z5~gXqCX$~2oXeBmyw&2p6rEBa`YCQ|IZopeD$ZO4ztSYfd{}c=b|qwgI{lu7&52oe
zp5@=G(zW)vX{Zx2m*1WItylL_H=<MaTuCY1Uz6~Oiy<&Nl91$5uzj?mshRe74P*ZF
zwp=n3U+Q&?(|2Eq>oQ}4R%`DA(Aj`W7l}+l8(3ViyTxni?uo<{9dZo@TcX<f9=RLs
zE&cM^n=76hBB_he!n<sAGInUknr%wCn1Pjg(bTN0;4v+3j`014NhO~cI1-*er3u4!
zw0Uu$L84omiuNIg+2nbbt(+uO%xj*;u73_cP6sVwC$u-p#1=0Yc>mS7{v~^Se3jo}
zT^kX_$JI4Ei2ijUF*LsnVT!_PAMx>Nvf&*mdtf1>c}w<%70MA?sFo>4^YT=z2a4v1
zp$=41E2|%NEbpMX+YA~Cvo*)yX@CzOgi!_6LL7{(z0Kk6F=-ep+7YGgrFmcV^?YrZ
zg~CXLxRBE<CBVN;QE;Ky5OW!s1!B2kHWYpHDh=sqkZ{8sk#h8mQpkR4Vi=M@O(;Yu
z^&@h2f=4hcBY-KW<)t}71lworhgAlKHK>wTa+sp=V>J&550x3f&EWjS$Q!G?d;I+t
zuOs`*18-uYGL|Ev_T1>os!vDe{n?x@6nFr>EK9!Czp);vUre55Iq<;j=z-KIoXu<R
zQo^X+wG*|5TbZDF&AfQ)0`SDv{DV9}q3zeLAg&3Sb<ocw`0q5UlFPrs95+Ev@83RV
zZmrpEzyh5TND&&rpJ8pkGMP_OA=W1Pz>*^jg;cH^qry6>UE`wNO67qGhU9ow^Fi3O
z0^Vh=z#$6MjB_+fpBT*~-c+u5M{(?EG}#w%uBVzbunjZl9fkyaf>Jy&LP~aJr$xg|
z`P*2^LVWxEwy=Y*aSpDQBKz14t38;bic&<i`kY>*H+1e7`q{VHA{|CCo}4WGj|gbu
z{UrnVHcQhq3rD?Es_`U^+K6#mk&vc;5d6R+AyDLHZe7sq!g;3@TM@f)_S%RzjMJul
zgoQoG?80Zk_j}WMArd+eaG0$VeFl@Zy9ZAOve@^|$elEhtDIrG+$U`y>L{~A9K1BU
zY5tfeZa~9?L%_<3&7|l<h7d7L=OU?rP54EB#vOV3PR2sy%PJ;nroY;|Jpb$gmXt6Q
zMa>!i?YCT{Z|W8T--fz+WOvP|#7sO~huzFZ%+v4RnzYlq=I-<Jv%B#CPVAU<r?)|8
zhj$HJkO#6(i_)fX#4*lfs)5?((X#hOw-R4JW@l$=ds;~)8^?Ryw)BoJu5sEo(WX;3
zy|Z3aG#y?2vu|77nuJV^4GGQMwLEyT8xkPJapE9rVZRh%`4=rW<Bggpm!0Kg9JR2=
zSr-Tb{WgmAQ@zVe_dMRc+a1t6n89Yqy*!w4`pykWR;a2~E~(?6Ij|WaPmMWV9Z;Vh
z`yBuEXizwr6iy2aGmnAOyg6J+M>cOKLkO8-hbnfFHG`H2&U{%4aQJjXGrS|Qoq$cy
zbp9tw1<MdmguWBh*D~VX2(1D#Fhmo@&y^yRyVpCh%QR7T&1rG{wy2Ob@8{S1QRz&t
z68EX@Pk51266$){ZfEPKc+R|(*zA*~I~jpxRSW}(c-^Xo1&cp~#mMq)`vw7x0O&Bh
zX7BJag4~!UJZC9z6LCM8{;cu{0!Pa-<;_K1eM|*<ypzR1m|VQn88&{o?-eTk0gG2n
z(c;C7%|r2q7o{r8a*5k{cZob^V5L}ABPAgoDwQbk#4?$k(=Mc0rAk1dzDI#`nXTnw
zKnc>A`Z~1O1=F-t{IXLG4RE(5@?8avh0@p2)t}2gr{a_raf)HEbJO7Ht!>r{bvCWG
z-k(p^_in}7*ub4h(@JIt=9Uc$F)gB^BB}9M48F`rs>9?C6GV#`7)cRksq_t<D8TTK
zMc9zUr}gIu?~YXr$@q<b&{6+%cKaasJN*lES^C3+jFte;NC~qecBOrRgfAn@7@EV{
z%d~EFBw<4`oF>Y-Cn=iO)s|9JWqr2;*I?V52X!&flB7Nc#BC__;Mws+vvR<5$+?>z
zCRtUCHiI&;QBriUIaqB{^H<kz?8|=Mii}0e(s~A^B$OLh%(IR@wUN>%MQLcV_=GW%
zuP&@o+G`Ci3LeWlISn$AYI^AU;YCTtpyLK20eL!j%K%;H*K=vfER}%cB44yKvE|T9
zUHFAdEG!zc&|JAps)hKdE`h<Ua@VTzx%ltL9Xm#1eIM;ruo0jT7=Y-_Enqf}we@a{
z{>L%gaPd)zorvnzfV-P`UPnP~Npq!rCk%?B`x09Ju3%Y1J%pk*Zg2&Aef*$4Fq7mU
zaXQ}YD-j-=A#$qZe%q!d-`mUJ)3#TU$gN_Fxe{bkSuIT^`mmIfex<z=1Rb3&=8AKB
z%d#}+T(z@ZbkjY)`a$RRw|IE7TG0?!KxhD(#c8&wspm$aG+%_eZ!f|bYuJ8>6ZXKX
zM0TL;M!8f0#JMoG!<I5iLMP3_c=KAZbPtQVc7Z;IYSEm1<j5Io`tFMpR!?i<oLQ2?
z|KHaQGz%qJW8Y7npbUZhUVrvEbIfOt&O_ihgHZHJr!I^jA^sZ0jAd4x(J;duY`Utk
zVax!7D6~sm!2n~nDZ9{v4S>NQTiM<jxj<wnLqlsNVp3gD8s-2pD6D1&;G0#`=L2<#
zelLi|H!!N^lJS2k#2EkU>Xng|k@3G(h+Sz-*ig5_ZhcU3c0wXtL@Z*{W7)M{0jwGC
z>cpwUr7uyNeETAvf0dbbPE2d>(vENGkjA5k;Ng`U`Mf_3%K7@Xb@6$5BHtUOsp#K6
zeosK20+#mnot3Em0q$$^Tllao_2sdT$*6(WNpK`QWqjP4FXiNes#^p)g(UCT{WK#f
z;srVb>TtRgx(fD;aJbFLdHcvXRG;$Zqd)7&`uX~f1<#2(o;3ei@{Q`%{q?Q*VuTVz
zn+&Hd`1*L>!WRLMQRl=#hxF%szuugshWf_*jbtR1eVW1V%?aV+`$MNmK<7}MeBK3-
zrA;NL$wl^Acq(#(fAV?q#t@UNJx2N_IOm4SlYEFQ;-gwS#|O2~WH<NcP!x_XP_IR%
z+JypaGrOrlde*N|=l&C7)R_LxXK$XKLK?BA(vAt@WRdAxfg9@PI_}o?H$U711666a
zf$>f7yPz$}b}n_h6(W0ih$AtJNLPl0ER_B%dfYdldMkSwah@R#c_Yr%C24t_UM2@e
zGje6)$wEISQvqq!y<3xMmc8ZJw)KJ~!5kR5+eWvxesXVdW{issR67DNPWp^>tgloB
z4)|U9%vHKH!Hc$eo&u18Y|r!$K49E&8$Azxd>IZS%}~HWtT>YcFjJY0IFlrq(7r%c
z9~G2lCl{Y>>3$xnNCuU(V<Gl#5+wHl%nVr0QyViTbnoEnt>SNUEEFEMU$jK=bt1Sq
zh=(!0F(na{(j^G$+H`H~LT!edDI3|bS+lG$T%6gKLARfLO9;pg33=mp5$dQ;F%m^9
z(KU1A9zaf!cravHmHWNGgd~{!7}R@gTmQbVw4m5blkd}bRrD^%`lvOppTfdl`;P^Q
zI#>ozYykR@^q!n_{{}4ssLXwhw%9M`rSG94^_}}wYD7QOq~w<wD$pdgUK<jod!c|Z
zC-ns&e9DZ@=+2J^i5Td7fiTC`15V^2!8@pGAo!+{88oDQ>kU9F&t$pvkHuGQ_f<qX
zVec;=13s2d)Q}Ac-K?ZkyqY$}qL`68^ZVsUi?J(O`qd<Kv^!J5sMf6s;qjLyU6ZST
zQOl^S5V`=-r))~-i94FEAvPVYn|e$Xb(ppTYw9E>`{KrTdQ?rla~YA?d`b6RMPL14
zlig2(Pox?DHILn_t!)Cmzo<phm4EuKpb)kr-`7=b&!4Bb2cEe2Zk^cPL%a)4EVUbz
zd6#fa8-bvsV%NjLd5K{_bjYEqozkFvT}FS{^tQ|sDXLND`)vkXA)}DF_U5S!r1FbQ
zS!JAcYsh#cmGy^0{>)u*9`HUX*F)J!xWhmF!Vy{Js4?Wh{c{l|!!&ELS<8B=F+71k
zz}q$FA|yGaWFDdK9c5Z(CYx0j&cgt0|GbYWKu1}UhcA-SAppyT!R1+pt~mc?OG2FO
zU1^pO__2mT8(mI%WRfyx;TI<`B}#rdKrUBL)7M8A1#Ie8&{i8+8dA{Aq=Rjuky)pq
z-d3aOG@78Zh!F*kYd#i*(^5H5%3Z6DYQp&%q$lZg-?NObl*sF$mq}IEY7I3n@9ak8
zM^F!<6FC~a=xZxGr9sOBG~%%7&}+<JVh7GCLz4-uLgMne1UCy^66wcqa`D%X$+iB)
z091teV-8)p`G^7>$~k_d$ey9zeReOS`huyTrWoMX)AXKL0M{2pGb%7~8X*3DjIq;G
zk3q!URD?U?AawNmDYAKnD<##<Za+cEUST#w@7JcvnHMA-MhfEW%WQ7PF>k#IZeRTy
z9}i(}mN*tdBj>0b8SybcNFGa5VKql&6`z;DC1`EWGUctKSoZFa@xlGVo_&2Bzgm-B
zMXqOUHoWo$#Z$93ZCv@XoO}+fCv+s0d2&%iBGYV#QbDz7i^s7XCggTQFi75h(XOA^
z0`piR@nGeYJb^HiJQxHv2Jn!{LXJW1iWH)5KUN8?_D~JVD0Ke5OgI9mj34Z1cxuNh
zR`i#P3n^xuxAp83QIWA#1KK&AWk|StF)1M(PXu~hJ}N^dQodM+21{yzB20iXuFw<?
zg28~rkv%>OneI)(xiGG>MY3*k+<tn+)p_{^e)%$WnW6BpjAj}cpRjyd5K~Lsms)~1
zTm%+ou&*2|B90^y16qS|wbZAUqKSoqH-C2jmS)7JZl7a9(wVu{oPuMsp#my0WZtlb
zTbj8_tl;j+^Kl|ZB`z+8F!!;m`qr8_TG-@eUwWY>4~CNW_e~Q>U$g-;cKG<T#*-Rf
zFd$6yK<+rzaFS?JzLOQ(?w^5fBEjxtpt<BBfo!dw$5*KIdB&Pxq6LHHj^3?)!X`rs
znz`#7RgBz<4kdF`kn}B(UrOZtNWgtqW6L}Ua4V~-JrypV{g<lR+|%SSI^t$ZH&r&@
z4`H~|6<q$%Pdw$|q3J<GotG#B9AmXnZ8M1I#6IobWam81=IYFS!HPEQep+elHPrY*
zD!T<vb;FoqCjq#mS}Jzs+_hwCY|Io+p!YCd83P2>LIVTrRq4~;4<L;jwSJsE`n6n*
zopiI6C@2epNS1Ai&TvXXysVVR*v$qLNx%0}w0=uftAkXwK`KAfDV*Orxd9qJRodAp
zQ2uCk^S8eGsh3`^x&~@gqQ8isxd9k4nsLE`QybGtnl0y|%ZEn`WjdA+?VdOH@PJpZ
z&a)+qZOBwXkF%X@!9zN|{qCGutsG-&I6#CfQwabTgJfSZTn4<zd1*sLkR!WInM|qY
zFuNN|j(zsKY%8G6h`n3Zh22!F$Xr4R-dSfRX-Tg6kffc1E+47QsB*e!=CQrEgO@3S
zaO>Q8V8diPAbBzmkJH9F(=!3a<##I2Otd1+?wlzK;+wDlla1iAmyk@qq*`FP6Vgq>
zIco5m<Z5*bt}}UB6E<$m9R#uGorR7Rg6lnEvLnKBA~Dh9fqnu$`z@Vl%M)Koirj>Q
z19$lTLYq=q4cxx*N4ejB^GS%3ynNN{Whhh`tQs6w5wpk!U-4};9@$A4_)>VKKOVU^
z`4YL&_A8$0k0O&8AicN+f^osd32OD+FlNt<Iu_L2b=`ChZ6M#=8RKWig7GX!9mcC*
z<H^l=#Q`#W?ps10j*qkMi)QTjh3@&u@l&29aqY+Q%b_!!xjr=Vmu3hX`2jV)V@lK~
z7HfZRYok$)n>n#B4%tV-fYJ938cdl`H?FJtIBSm0HvL35>kmF$QnP>SL!H`+@lJfh
z@}6gNlTHkuBG`J+3LxnB^<SW69Cv1zBt9^s4c`vUzsPpbnP%18-#e20;wHWCqi(g2
zc?Db0UyB($Wtwzr-UnQi&$}$EB>MxK+am*C*chRFVXtV2GaDf@#+g@<;p}z+e`$L-
zi#@Bwj-2S891}feo8y<fFFfZvQkkv((jea1r^FKO$CPM|-^7G!Ng`DG%-+)F@2xZN
zdzUrl9$bXCGi97}863($To;X7$9j8p-F?d9TF1sm&s^WZ<<{Q7{{Y`nJ5K(mT9oNO
z3<X)~{~zta|Ibix2z%_e7QMRuHBloG$Pp`Y=`-tptN}Vq^tB4e3Yf?`j(+|SCJ9%o
zEU9#J&*X1fClAVy42jUl^=2QQ;mg(L%JF%Vdpq=F+qu1eK7l+6&g;c7n@x@V@-cQF
z-#*S#O)2K)A0%T_6e&D#yk8#8r*}}?G5{aJV0i4lkC|_=f*%2RSZnj&D7j79Ux#n@
z+eq2B$J^j=_-s#neZL<h80Qp0uzQDN-+F%)fImD_)arpiw~x6yx%+LZ1W6;v^Aq~A
zzMn6T7Yn^6d8o+>rOzj^JF*L-Gt@%xBtpMf;Xs&w5OI4()(Dh6<BZHCd`~%~#Yb;i
z$ErOewis`$M_SqD<I-49aOcu|5+=Eupw`3CBGOFO`si(~Wlo{xFsaY6V!rS6--p4F
z(-kZ{P2`|{R8AFktQov4XE|iqj75J!oF;{_#B(QIeZ-V0uZWU`yOy~dCq1HV3zx|-
z?0L@e>$j+<nEXeq_KjyO#MS!^8Z$r$HN{H~KWWAp_oUj!GsQ&;=1k5HN<Ya{jJMsQ
z>4=?qZv#O!Gj5=ZI0bP|R}4U~b`VI7h;~e=A4$Kr#5Dr?8RzJdsU}bwismm@#x=-4
zl+1k%@cj8Ao3MMSxJtOjgy9PrEC53q8E1Gxf);ZBHTaORgUpnL5f1p>XwcO@{TT}c
z;p0M~7cdL=#1om%_#7FvDi{#IBNwRgM3Imzm1}ASa#EoC;<0u4dr89-NsqBr5;>o+
z*_jSNCP8@P1vHs3A3uKo2azx8@q0)U5baMNIe5jnQE9L<y^Ja}rJ3KeB2wh-G|E+^
zqgu)#AmSaALBc4GKe%&17`Xwis1Gl}Eg}_CBUxQeMhM5c;qnUXKX*hsKpM%hDv6uI
zSNmkI=cUdF2*C~d17M!96RfZh?xVlx5!Txv3}L&UKm{7V9%vf*J;8t#OKuAklv{2^
z*7XSf>jVD=27;9H22w1;(V!GCqC&_9w;D+H)y*bkjG5N}t#Jg&SzxQ@iNK7U7q?42
z#&`_XhK7YMHy2uAXKz2FBvi@eT^VPC_)oJv#5`roT1ROoHbf&*$qi7%V@AYdAVhe(
zEJeuZ?@HX0Fc!{f!+#qcz)!Mf&AUfoz}^4c8w@kp{f64z5;rjO12kLD_7%x|Mo^$R
zbw)p!qpMb%;nHG4`cbtEJ?MQCSQ;MXPxzCO!)PI<{4q=8cP3z}Yu3&JKSt^_zDYF{
zIQqgcuzilWvUo$|&-RiA+EQ$cUbW7^9tI!*AMlrv(=p;w^9U4kt1Jf<`wdbDemy_H
zPFp0KilJ>Y&NqLNq%otO31A-Q^aKl`<8QSgDgin-1H?Ku1JaqPvjW{g2G&N28Bxu(
z@~kvJeNoK8*VwAm&xVO4R~ll1^jsf<O6cPgN|-4;$=fBzD3!;dQUhq6k`F2*BpLB0
z7;%i^+7`Q+@4;$qvxrT%Tx#|`_HYwuMi^DhD*}D+T`nwTE(<eK##Qpmn#{z+<O{y6
z;#A18j##4IaJJgyUe0;tuT2FbkM+2IiS;p^F3h4H&+dGs1yXlnj<V=mytow8uXkQ#
zw$}Nr=e^%ww(VB=Pw!AD!kcybbjCx*IN3nIx0a40tQh%5f%+>U>L_Euy0d7cH2WDP
zdXZdp?YJr-(lZxjHik6Rb6Vqw+{8lS-qR)QNMq3bVyI=}h)QP8|G|V>uex*DjP+E=
zl|L_*!Hk0F%t3Sh=XEoJty@db_iS+Twjwd=f;l(1gQL+xleg2bq`-0H^p`dnsy1Am
z(~0~XZ0<%Se>@Pg=Mt)uqqG^O+or=iqL;B?YiC2=YJ4QO-oDCr)ik?%@&(LG%`v{Y
z?tLXudeVG(%gO*oiX!4|iD7_VNa!6|qaWUQs_gQVN=_fv&#qnw!23**cR7u@x$Dsa
zjmrOO@o327Sk=;@AmLEqH|!Y&$jzghylKm@j@_NVIr<J+Zy_K+;fz-Y_n<1F17E*9
zK`aR{W=i4*Ul8mj{spsGB&hQ3>B_GKEpM3Y&W!lp9|ICs*(Q-ud!qV4D_fi<>Uuqr
zmrdU&??i3~mN-S+9oazRE@J~&A))Ih50i7eh%2Y^xZ*N=0XCIV=anbuawlQ0n~A~{
zl=^9cjc>ERV9NtnO_#}r>!^Fs)%eO#p&ia8<)mZrK?)pp<KO8kac6Wwq7z}S>Hi}i
zD5BBe5h*m@u)I0;&f5Y?+XxfFE#QYaP)$9Puir8}=nx4)*ATS~0puKYG*9PH`Bl(y
zd^K~1xOI?YbF)HVc5h%aSDVi->G?(ag;fk$6Iv;03>sKf++a||N7}~BYEbUU1qR<q
zLwLO@JzK-(7wT+=0^;#D>aqfI=l@j$c`rL)neiopO9vy%UrFYxu;<e6VLD<N@I1h!
zi`)b5)MXO(;`)1}a_KgL*xtDD$g^ta_~cnRj9lcIwH>fvhg}W|b@UC3>NgTT5~EMT
zSED~p&aMO&hzK{GtS2=kNwp8PY;Ttf$<a=ALT7jAjMv_*JA-}n@$LqC_@74gEwj8i
zdia}-`Dhgy15cp802NZngY~?NW7hMQp|cBo<#+d`1Gi%`Dw}dux3N%n1}yJ_YRpGr
z+Q6$PlZY$QxVdW`b$PUy>-|dS&%n^1Sr%!ujw{QmA@dT=&txVn<JWmuW11j!+}#I;
zlR-CI?oii1pfl7nrO!Q63-61CZTL3zq`){uqHF-`FQ8~EDz}E6lrlH~lU;bY>1?Ud
z1?wv7Qj(Zv)Pq|zsl->?nLMrk*;HomUqboOOBqnyoki`CgMyG|X^b)#e>#Y7w~r1W
z8IO_p${VXKxTujTt;^wO#lRz7P8e_5I0ySFw@IlxufVUahX&?&xUNvI%t?XkT;x~P
ztCmHbrQSV3;7Z{Q7%w}hxbo0euDEnUcYmOVS>jqy`!OEtj>m8SZc}VMq0Aqtv1rrk
zY8zdfkuy85grf)N;JduFT4?U=yr>0y+0nQ?acdo27bMxzp@Sj_&{T-P1u5N<4i%{T
zC*4=xsK}_}qe#if`<hX+>}7Q5YO&>;^!R#ls}$sdcDoKLaR_F(uZ5Ks<N(q>A791P
z1+%0QHDOR}JGCOiue;=K6l>e*<Sy5k17J47csEyXf4EPrEJ~_DdYY=*d<K=L!v~*k
z{#(uO^rLoK)^<sbay$#cEhb?BwN_KRL1jC{TUw+ToXw_lDb9Y*IVo}jz)nWEhi|N!
zTv1{T!u73R@ZoqyFfaWUK$BS!ls}r@r0sq?>zWFgO~R#f69Hm9hwyY&<ILuAb1v4i
zzx0u&J1-P}Cq4K|BTngkNo4a7hSH@|lYC`8y}PW~o;<fP-paOC@QJL;w4yOnk<y%k
zt9}QBOkNVq*JWM%c-DApSBj*PN<lgJIRqb8XU6HU1l)$Oyb|s|&@5a!!?lhEqw2~%
zTl?BtRfaf}g97`3pH|rC{!i5>^Z&v2_}^kdCbs|PB(yz-gbilugvx@6Zg=jS*G~Z0
zu^G@*SK5S~@iLnA_S;wgoI1SXyvEYQdO+YQKSex(h~i#CCX@~s*QdkNqw~WF&4GT9
z(&Y8_`J`_b9I7@-uuwJ7Aj<4tNNeptEqsy#><lt@f+FEA!|hssC+8NlNj`X>URpqC
z_I0$8dDQMLy_*vb@K>;BnA5A{`&6rCo*~QA0fv_;e`ffCtu5cq90@b2>__Z3tgokc
z=f?${oB$={gX6%B-pwbxB0Ui`KD$~F1&`N<;p@XSx^xSDAw5|SIj7eLV_wRPL}W|=
zTKTdhi>48Z`h<X`SkMoJg+h;fg}%u>4E&_72Ms;*#3Z2?7Ul_>zIK~9lw}b;`i#KY
zsDQuEk>SAJ!dinwdNY}{qjYpGl_cwUm9q9f`q+`hn!lY%@f$#(x=T6C!cdKZo{f$Z
z85R@)ZOZ<^l8MK5vJFynMm~@YY&1_;M%B$p9M*@F-O(iL^qX2v4*bAw*ufeND+iB|
zDYa=@#?k8l?qp!9RrTg?UHm6BYH*1vHcf?-Ya4QLgVScyNF!Vpx>6RBo+k_5QDi_K
zVJk});~c-8!ep#OQ7#J#DnpkR=yOA00rB)|OhLvX0eXVnKQ7psR(Dzse+1tmn<6qe
zpj-bAg{a44#!@WMZnGPLCYSC=Lig#XwiFQ7h8PEkvPeO6whRnIPBxE-;cJ~yF8(|-
zQP=xDPOC+q16F&cXW4`SHyFA<H-xv#78>COUXTmcn&+pf#L$T#EsF)llG)Jfva)T$
znom$)^Ck(!)PRDeup>hwrFi3=J>txIm6-FZc!3g^Bi+ojC1@&UZ5Sb54GBO{kHGs+
z;u)eD_D*ppTPJqon2Prs4M;c%c;ZLQsKQFRE-A@MlkoRla;A2!p#O%+H?FDcpv{);
zzV=39eaK<r7lhaWyPJi0y|gX;8$00wFJn3cIAohpj7!xFpA%eH+L#6>lXDI_e;~f?
zX4t*7Ic1{{3fyPae!@+~<c9Z|iZ2dE6u7z=&-w)6Y>y~$og59EKD1ZaQ}gC51<PVa
zg)O?MKo!Rk3!kEu2nLr*`5Piecs1zag=5ZI7#GUpp2iH>Qc25aLzu}+Fx<eLSciT)
zFTDEZD0RZhMi3y(<H;6u^#BK=veV;xZVpEQ76pX0+m51b>=qSLntCT{WZe~T#>>ID
zqAf0(LDv{-L!HJyJ&>ypMPF<qtzDE$??GKmK8@PPE_aH?^4!BwNdvfe!~)&2o1DU|
z(y<Y7uyZf4aj;bKM@JZ#6S&N0FYxiCd4%S!!lNh;>*<pq;<oxx2gYl9KKxf;k$uM(
zv5{}@L|m3i#(c8Une)nuwAr;l?yxUL=7&lk5<;l@MC2J9Dd&WwciY5XT6Hx9&ShXV
zkU0`20uO_wLt!Lp;`>ySw>+8V)-CzV9LenI)W7y!%2aU=YS0NrZJOY}m`Ri!%9%12
z+fZe!hxr*Vy<=r1l(U{FT%(-@C{s(SsQxBNnCrRou88H<RZA9z+Uk`oLcE8BsqM-U
zdgH}-4qhVUuyytpmLh3>QDpJ=p1dF1Du3TqCCFS%RTY_5yPo)n8#B^&daNon2BF2I
zhw}S$apSX;c$}5VI7qYB5y7Uw4Q|a4P}TYJtcRNvI!Ii;QKa7&n(FG?K4oY(3-rC5
znkGy=fuKkY#?K>AV3_V@+Zs|nHUaUd8m6483um6P+Jbno9M{v9r=~~P<+%s6BZ}S6
z&j$lT9cGQyL;Ds=>gW=vD^B4(?m}Ix#)hOFOHxjc-h#}s_dgGFRLxSGqg8v{tPj>I
zjJ`jWxpS23U_9-If&=5mIhGN;yBMJg{H`9cOjpOQ8opBAR66lTQ$H%9f}1pei+>z1
zFwe`QsAlilu;jV=Yv1$rh|SlT3ADPFc6&($ls9O6n(PGHnA7eXjOKIGvAdpir2S^^
z<oElHFagi#Lai}oWM?!C_AVN>P8L42(WBp~y&XA~$4+PnR}UtdTAE4aJ6Ji>?~CKb
zMjd4L9G6w=AhJDCat&OnXNoS(uC4%&=SXVqt^iB)|Kxp=wSt2rg^O+g&-5Y={g2~_
z>SU+vC3?Cycqe%=+CMtwY<l4)W!0Hq=!)ypW)0WwIU3@c?_Ct+X(-XQ>HE<QxG;;|
z-trX2o(F<UNV#?+<K*nY=B;ac`yMDiefH{^kUD-@KccgSe9#eA`p)>J7+l(ZIP4%8
z%<3;F&+gDxV*T9Hgg&Z&De@^dyio};I`z_ni4l_0-^``Vd<kxBXT;1wlV1F+!x(|G
zgoFF{aW|*DIIX$6alaY3$_q#pJT=PX4ZoC4L4d#KzOxRnW@m~HlVTKNb#&V;>2odq
z*(P0rj$ti^y#E}+iz^lNXp`T#0Fs*bXoBD}2`2!?HTR_27lmTL$vKkH7Nb{mo*?3e
zkOk0Quu(2IkcF6Ht*b);+neITR%MBtMTl64<Jaio>4Pv=B#LUqi%dy4zaNtVi?kbo
zBUz;tY{!3TZb4(RJ?U3*6aqnwxV9Z{w7{iEtZv^nEzHrIx?qCoz2~~g`5r->qNjvn
zoe;UJAjTrM{k+Qcj~uPHHdD}(f0614Eubo5#!H%X{|BsP29m6T4S4NFR?7eZwh}BD
zTTmd20BgQS%a1@BiztnKHWy(CSO7uz4GJQ_V+R7nKt2N6)e;57ehV^?z~Uw*kiZ8V
zfG};}hXh2a3no~fiX9KyKpDY&vihp($O8c_pllli`XAF;Y%oED*~v~ouztixJ%aQd
z5n4cM@h-v8QgqlcvTgu?j1&$3dm&g!d4-UP?}g7XJP9Pn8L1`>xK;J=7zKxe9gK~U
zsWBo<0@gN7jH|uGM_wBy4bnKm)<z$<I>^u;#FAJT>^`F2K!**5bwIR#(|@bm_KAyI
zQR}R-#c%=8IAl>8pF~>xP}9%hf^mg3vUi69x;B*<%1WbGL-Zc#DZem9D4h0NAhjYe
z1~}ayDP4*;e@^FRW=U_me(;)-=w5<cS$R8>63^@3E7z19vCB0+6`fc?cP(}+7|(*k
z6nXw)TB)cy5Qf(CfmM-NvOdC*?(HP@9Kik8a*LUR>HmmK{QpL&vHowA8WZDxFVA`D
z|BX_ESB>uz?7QY%2N4GtcqR<b3*5&g5rv5+j0HT|IYIRloo^$xq;Bqp)R!`96jxPK
z>-G?l>`YQlP?ep(ok-A9qzZ@Fn$Cg1y_!dMjdD6F3O0n`xnd|PXAKvXk$r}Nm69>+
za66Cwi)J9u&#c7UG^`%U!wu$1w5OiGEIeLc2Bpc%x+qU5$`Gyn7r0q?N)S3YpnJq4
z83}pB!wsDbdxnG|^-g*Y>s*lG7*EmwaG7C%+derTP}CCjQsHiR`7i3}q;(o_q(h})
zP5zQ$h`~;szsm}VF~*n{5}s<nozeP=2NI@>19#<sysSi3h&S%@3#)29iTCAkjB50Z
z|B(lNEPbz;w3!>P_1vA1WYq9^e_?R)B8FDeYsCORtmZ1lm3p67RRCbATRwt0*6=B>
zH%C(lFUu0p#|jXvMScQhDn=NsT0dTQDdjAhgBw`7nZUH$q{xbaOw%l`=)lIZ9RYWL
zEIqwu4t2-fQuQxUf>X1bTPQN|pV(ht88Vt>(kP4A?pq*L^CkjQPSN^^k68?8wumWx
z6)y$#yksLDW@-{M3Pb3!17m4FV9w^*a0(7bt4MrW7POGUjuygF2(%a#eU%r^Lj+*n
z#XzaouxKDT+gGzyAqbjCOgHql#W<llh#`+U_e89Se$-4DM`AclT&CTJDQeNXugvDT
zevjz>CbyH2I=ANm_{tMgi0P0#pZ52Qo$s77^<m*Ih~kOZo9*n}tWAoexQe=1Ji_8;
zqfkP;l*8+_J$$wV^QVX7t>g2hk4Ux)5{6&gnYLByMdiex(hGcUOK;b@uGLC-33zGd
z1kE#hFh_Y4qYFnhAVJ;BDS&Yppkg8L5811<ldP;NT@9<PZgs0^CHjn<TJ1`13s<z>
z&&d}S*N-D?Fm3C_#8oJNsQF#$1$<_#n(Z<^Szd&gug1^&^M2`6X=5Bcdxk6T$S2ig
zkTPa%!~5O60~5{Q!t=~DnDVeCOxY&Od5=BM;T(IfsugvU$I~7D#_8AF)lu)ldFj+r
z?L}oiKZo9z{Zr}Ej489TZq1I>jK2RSAlh*fYCc5MB#d0(hw1D1+xbY)kzTdWTGw_}
zPU#L{!jQ<If34ql7}?L)*ci<Nw<L5WM?w33E1*-`qqhG(fIsyl6OFLPE<1FATxDk+
zO+JVzSP7wtBVd7@FrYMH$=cb+X<jg6A8g%?M!er5i{<0e9fws(5Tar_3mq%~G(Gq_
zpH4VE03kgV5t@DgcLHqOw$!lc4<(my&<zna(dHSgcsr>bEQK;7@rn;H1bf&!Ft-~J
z$RY>GX(2)0GB$+#+EDv7A^!_Po+rdCC;<S;A2kv(>)YR;Xp&R-=Omf;%Rt}0qSnj_
z>7;OU2rlbeR=Zz=p+HF~7>HoN{BR#owl(}Sh%8W+peRelWjBuwr0NvxaBN#8j&c>C
zTS*)7cL5|C1bae)rMy7RD=r91X9|HYgm9%JGK!Lo<Pvs6F9I?R;*%c^y^>%Sc(kDS
z3EaB~;`al$ANviAph%wSuE*}r5y53MzR({hypQBM{0GDYAlULz#Dsw%c(d`pZVu2a
z*WjtOL&%AkIT$=(a+9{VMJ>2yi{DPO@(opadAm#Qc0QmkW+?dW<J|hYTd06GXxl`v
zIi38gb!K=fWa@z-NKiB3l5kc2>)eLeVRUJnRhBl(3h&LpsVqJON?S;6CCyQ@#U4Q<
z&9@gjQA;(lg?XwaI%0(bvG%gHOeaq6QH~y+_Oeu5Crv$;uixPdJ^!FdI!y>zT&7$}
z8(^(W2)M{4PuTx}r82}W_`^}2?>htb*~aJ)TUqTQRUVP6Q^hG7$FxNE8(^QCI-ABd
zbO4+!L*%m41KcWQMX`&sSr`kOKT|p?_3cSRNjfPe4NNJcjdp>U#q1~Zb5e<QFo3YU
zLs;1N65sh97Zi#4kW{P0Lrnn53;xP3YMT+f!l1x=)Sy*S=7Lsjz+5YKVV(e_v|}ZF
z!4*3nP3?r)j|URQd4R`yI3#$XgmaRvGiDHqmkbe#7YGX;wNRZl7EVg!LSa#O35vO)
z6zOvTKi~=!v{1@+XjDV~7^o#){)#b@j8Kdl`jFRgHU0Q40f~v?-ea=7N9Qtl;%;pq
zH)*Hj&v6Mc*&Xb5A9|<SJEpSgF79OA3?p>$A--&FLTf#ptMYQY-`RP-&R?<rbg$g4
zyf`(=VR~mhUSk>W<blf8aWb|=8=9gRf}Dy9o+;5jaS1x=bprCBmu1|-X<)ep6=(`R
zMFb^rgxMX9UN&y+d|vGH@H&<B;NXPG#~>T>!;X*47W<h$+jg)8vJt)nSiW3Fi|tWw
z9fSZwDqSBcoCkGuP`I(ro4k;YKa0;O?9Amc61)Hint?_xb^_c<X>#*nr=sZz4g&jo
z&h)|;$>R)CNGj^HuJ}Y5wIRJBas|d7m!aaP976Fbr2^##1{?syn=Nv5!Z8=(nu~$!
z1(#%XaYoCbrMfJL-|DT+?5XRsf%i<>qcAvgi@Ho2R+Jiz1$yPFW&a}>E7Y&MF|gBS
zmQfdzQl{(AI|up6nI0|3$mhH!aU_11RrvGwz{{;x1P4J*WpYcCzvM|%dp|EpJ{z42
zf|E0V%9pG;vWs5qftaM(oRv#70WsKa<Og|O5&C1&yq$O0{4HhSe%t(y1D}q;H5PQ;
znfLy@I2O{%B@7-Fe^17pua8}!m7+qBCyI9fPNy(fVuF@7-c52?Ct0^1WUX-Y!7ipE
z6q`+#Fzhxzc8xPXPvpDa%@QG8XZ9WZGUa-fSGXr#!&z4@X=~X{SCk`wh}%V$NSUQE
zX)v40AE7sIjH%Dyf_*k>`Cir<qa2t+g)#nh8SOFN8Y6!rekd?(Ul5@2U=)H33L9IG
z38DOt8ft)aDSEnrWA3eG0<wQy%*29<b@sa-p8##Hx4sp)HPCMMEWEc^9F*XrIvo7G
zLeDpzDs5RKXI)SWVk!m;SG$My9dyk}DQ|}UWZqf#o}ZY>2iSwT=ztu>-hZA2#L-~n
zA~QqTKX-vbM>Zh2<ubhH?w<Ni9(fyn^yrE6h5l;^#zOy}uo(<Y|6T3h|B78<*8w_o
z@`&>Qhy;B$aRn}ts?&+19MwtwzZiR`7*V3NTf1%B-K%YTwQbwBZQHhO+qP}ncK7=G
z`;&8Wl5bz^q*7OvR5CBB#vF6J&mesNCtOzgTxWWPxr;K6nu9rsB<xclcA-g^+au-m
zalXI!IXKD&ny8N!9KZN^Dn17+lN+E|I5G~=--NXIsk>+!n<Xi|O4i}$Sp1v)+gxDd
z@s=A<fajM5Ss=;b!9XyNN^pYgaKaJ%PWFs)IHKTwy#Al?-#0t|>wrhNZ0~Q<wl8rZ
zi{sk&qxR?K`FsC1bO&Uq58&!)R2TjSki0)BIT2w@F7rH>w{#Z1_sL+=sI27uFqcko
z^}$kL9!enIGI}kKTpScg#d0*sH%X3o(yxj%jXf?{Y^6}fdSRE>TMPq%=WBX*d^G41
zTDwE=s39Rx#4Mdq@zpY@S>z%)lt0WKC(Vub_0Dy#mhHpbuy$@l6T7Bf*v>&KjTr-M
z0!5SH0-p5psu^fEqsR0bz``o~ad()NCl>0(oqFu`p5?=7h~z-&gYSaOAakOvjeTbF
z=`cWqwLoCQjW`5VA;Z6Zy+Q-~4aF6u!i}yHbMQhxi@~hW5Zgbw$>$+XOMxqLD#Bzi
zbvcaC)$HJquPro;A0gcSTEG`w$RzrU{=5J<__OH!D#u)1%Ym80l0plHBm+F&(o8Y}
zYydZ+rFBxo%&AKW3A|4E)J-__Ibp>Q`dYmqLgHxtpm^N^f0s{c?DvJ4dcyfRmh<d8
zue?8dp$iCVc!Zb~TX65R@Ep97pl>Vc7k}(ixb+QL#1sh;wxl%xNP?~a@r>}d4Z19e
zaE5DeV09gu?y!kpV`s~kv2+fgg^B)fP>0<N5={lZg!OWg7=5-I5L*H8-GcDdnz>;C
z&Zz>_(e{A)B|6^&yMyGJL&EL9H3$xM1*BNuLxaYp7VT@oaQRe<&s$fcq65n)qrf2n
z!^r~y*Q)(Me6Z3y2lyKTfqj!ZE_WpDl#wrP_P9m_W?|kU5_$)@XQp`Q(4-xj#4#WJ
zGv=-`7-$0wR07wEf2B}hB)X@%Sz~fGRlk3HW}+;sR;J($YPz)4*2nv8kFBgy6HaZa
zJn@)XWxA&*A!}!ptnLBLH4@9}39UpbvOcLy{4}4@;1NDHs&Ac8$=I5z$e$1q-JzmW
zUQcDSENh^<Qk_69cnAf4v7|x_y|Dhbvz!4*<2jx{cZ(^wV<@!iSsy`cdty{Pc7OEp
zCMpb@xOLT+D96NL1;Hzy)@=U%%y_b0r9dJZ(hdbbEIWazq)$U+nw}WQ9E(J|x<|n?
zR95GyU5o<Lba{$BQ{zcYDBP-#+F@XmD<nFcj7xkW3_>DITYj_$`sK#Em?D08b8UKE
zwKJ=?SZ+ho`Ob8vH)-rUsSYjk=}lWJXA0T7Bba2YCrx*cMh7+ylMUh#&gRBoc0%s`
zgrwSq0h%~ypbFAA>9fW{FB9W}_tMVxyx?N@K$Ws77NUR$Sv)_5(TFNRxz&c1g7k*L
zvp90xX|3dhmXG__?5QQ9_5s|j8Ab5zJcLL-Ldq;n!8lk^@8NtkF_)=W0TF1f0gcBi
zsDUybFFy~uyb1x>f6EU-8>3>8duAS7R@1FCeZ4)}18BowV$61FQM-&xZj7(&AI>v#
z>=nBqQ?8x?Xlhe=)TG8#VM>k&uH>1!OeGREO)0yoKzBAtl0?JbXRD&I3)qJEE{hPW
z!ZC}{HBL$HBe94n2F^Wc0)N1n()HC3I+XWY)F0~XFg#3O18>eyH_lmZRoUY@2g*?F
z0*?j(axl1AHUG@&KC0@z&FVcRICAe1sP1(r?_$&s6RIR+m}v8=zHCc>>{X!N4GnTh
zQz-pewR|?s{Mx1izvj6ZbEjxjk5W{R!vDEe?QA5<Jy-xNex@FE)zs<-TAg9C<Z_KB
zqDhuhdRitlyC$hRo;X>yv}SsYMR3yk2CmL4KHPNT5Rm!9mVm(`6mRm7wL<wht84$1
zwizp}y9NrJ+gVs@_WG?V1P%N_h69U`v*27k>y!(cC{=L+R|v`;0;U*{c>s;5yYwcD
zh>OvZ^6O%9UYzzI5?&Mkyg<?ag~+1l)7`s7)TO#{N+Q1K*l`JyH|r}5bbnW6KNCdM
zC3>Jo)U_V{K9J}JAiO5ol}&g>k~1S8<bArT(5MtSh%7qz8tU9DE+VUQxRGTr65Z3V
z3ygw`jG|fyzI6neUV2$AKZ%n~M}J&5Fv=MEGrPrZ;S|F0<gU`jU%B{aDPL&1XyhNR
zD>>l!?i1V19qm?r>@C!kZ~EUL%l5%^+@#1WJ+Ap*@m#k9gzesGGREpsr}I(W_-ja+
zf)_v4`mN>gVqgaT;MNJI1$=vj607mGc!hSB)pq99YR<22*U)8O=NC`Npf7W4%e^+;
zJ9iiphI~yShAed6hz-4^`c_rRaK7sTFI>ho@#9w-J#wAak&3MkZS$E}v~;=udWVPd
zN=hLT!w>&pFE|>h!@K!cABFQt%=~2;MPrbhHt{l_qVA>5*^(fQe=e$PSdxH3gJ=|w
zELRI=%e^b{v8=IuC|jb0M0^^C<ee236mjPA7KXW6*79q1O)_^SGL1Jgq7Lr$V#9ME
z|Ct8t8Bd%}l_jg^OJ}%G7n^vQ%+M8TRzFVtiD<g$(tj4QnyGglP&bi-!J<jh)Nf31
z?Xz64piNx$(jqJkOuj7EmE1L#z;n>M^=oPKwSR&;%HE4O$mNLZx+KOE+i^~yi*wb!
zb$+<Ha5c_0VrS}X_}Y}jw^L|u0UClm{X>PJVpLLjU=kzNe&LZx#ln?*m@i>3vofR;
zWhaTc1LR+)Yc6P`fObqhs0KBhWxgEpweGvv@`+APPgVIJ70Q1&lcT5q@6P15HO6Cz
z*kHO}RhQtSM3Bm1>0zEu0NP1@8(J0P(id?U|M^0Y$Xg~~zbqsk%h75U7qlgMG-&nE
zI=r9npI`4k0@~RRU>`p{yco+r@hy}l{Kml!l914renB4Z)?~#kPTcds3P5wmXPu>;
z>%Kt8<{<@S6LZBcxj2+Q;LO?)w3)m4?+7mGF+L}z_FG7}w#Htv^R~xC&EL1Syu4y2
z8WDPjt8SUT-96l2&-9UgdBS|nU-#F4c=+Uix^$R1|KL*Z{?qE=`eID1j>J^k7=5js
zWPYHsoTpkSXj)0*kCaHTu&-x@2#ak%6SrT86Nw`hN<e`BlvOuj;+Cho4>ytaH>Ek3
zB8qOW;<qNF$u{*GF1`m6`<hu1qtfe2oQ_vvFe3As!oe`;XU9wR@*VfWRwQZQf_~=_
zOTNN8UZzH64o?WW|55P1GJzaFfvDN&Z)kycf~T56b);!+W9|52yj>8^Qg@Pjl4rSW
z9HuZ4fh4KV<&By`f+Hf{KNXB4<CewNS0YCux;OGlVSm7ApL?uWL39U=jDRjy4vlZR
zK=c*OfHu)l2EJsQu-%w2OGD<hlPl+-3(%2Il)YpRS2Btu%=nYbG;P@>luyqBGqt!K
zZ4iKlwhxYNtigf>$^lSNnaU>(Hh^|BVvRjIDyK320CzZ4I?Hi%`0+%H4OIv<DVku1
zaSl`bJJcF(q2Wo1;@l!HP&-6t(ru)U9K@vGOdVc&Y*YhX#z_;kLRHa56ZMPxZB_Vu
z-m&3Qc8gV}3v`c|*SJw+$&3ekWV<!;j(CMnDRZ`@gGg!gr&rMGynOBuqU)-97B;Nj
znR2c&2N15bGdjkm%mTFVm{Wyot0<nbR5v66ZUj@D<6XTA$^t^tW#JX2QSRhzc{O+@
z+-*y@v<`k{85|tfknfVWe&0m3q#_cS9FE(hgn5sM7NsT=rnzrVp!r*J8z-%<nN_+#
zaGfa9o;yC?^aRo+bPiOY^rVGuz3By*C|Pv)SsQM|c<CQKR9O+HX)=ieD?tf6cKrTB
znJj_QU+(MP8Tkv7|M1<qDFFWi`(wJXO+{REKq0uPCe4_p(U*|=X*E4mogKh%uf!qS
zQj3&I@Z@Y%IB5Q{ab{;>=<)g><}OcudZUI%7NOZNa5+4k_&+nLrWcC$X4SmiwO6Ec
z*(UNjGopD;veAn|>XQ!Et=bZzdV9owRb62r_|23pUBoBGqkd3c<Y6tokU%VGuF*n(
z{WYFdw6TBjz{%gGY6I+tC11K1gN}{e`)WW>w>5<4(PLj6VUSsfD@ynM*U7M*Sq^>c
zC|#b}0}05Y3sKX6<4he0pr<a{09`lzR~eTj^<-##r+KDW_o`*LsX@veB?;6@sdI47
zF7>O1D&jEi5EiSHt~#Yt!-Rln$1h#8+Xi^i8giYB=G2(u>5U?UmA7Dm^>7vMVC9)I
zlysskZq$w0ZNoKMMF|-=hQ|<SL5cK3{Y0Ps4y$8HJ6NB~0g7;ViK14zTo+{<SkoP~
zz4WQjKgp4K4%w%qHSSeKU6S@v!kN&GAq*rCus=9a?mctF7>`cXfJy=$zM~pU@?0IL
zon13hNwG)H$8u^HN~P;{$2*WpgpbJ^uAbLmsAXASdKhL}Pa<YE#4r0}IAXHN3f=_F
zrQN8UUuI@D8qYFwTSCuW&Nab@02H9-?%jrVIBpjrTU?)5Trv86?%UY+>>}?b6{2_S
z*lh823(wCyHRZTKWttgL{pY@Dfs0^exc}zUEi|5W<QWvc-^wsn|DvNFUTGE>XuQhE
zGstrzqbAlkDwWOrYp|6f(v<ht(Ay+7@(ZHNzJ<M8V7J<ZPIzDVusIAV&Hu^h=myG!
z?en_^i~%W0TRA=OE6EAr^YF_DR?`CKTThG=OlANQ%|;;#?iWZs(@{JW_hgl>CJk{3
zdD{Y}VkgS1IGj&r-vGi7)mEM~>)}4`9w3ETyseh8R+k0NE%^YhEsH!=qQ7<DdMt)3
z86Rr1-ijUW<l3sSmR85~<l-cAPwreZxNG#@Y+p)=>sAc`*1`Cj=zSmw0H|C4D2x#U
zkmK=m1)|b|si3%4y~~N!G=3Lj`u54*1~c%sqJoy{5Ln3Ha}lx<Ww!6J=mo`1J)Vct
zETwkFqA&zxYHdj<>2R?R$YMPr_!x0^rd_z>k3>y!SH>Azl0#Q9f2g33`)wf>bnbW5
z4uCDV;De6eYRQb(B5?&6P62p-K3Dg`X>X05JXf`gThk41SwjI;QET&o1E;Y`OC7sy
zzn6d7KJ2#>eM#1?T^k;Ao7rp=ZKsC4cP$n7vhNM)P<l*3b9p3Mf@9Orv@dmp<_mo|
z&fid{olMF$G(#G*LkKw8z73o1mq}tQ$G-b>^2B-KkaP=_EL+$j;jZ6&6Wc)+C&NfU
zrl}6v48-`Z=aCg}KqP3wT~@+*;d99M>$^Z6VefYw0&GL=RLa#IS~ib>nz_k#_Xn%>
z;$l%-XOLAf)0ulN?R_m>Az3RWlpot-avJOCycc0*zIf84h`{P!&O%ASm$(it4LEaU
z*-pBhbUKgvl24PNj_w)rI^4;BEv=KbepKOaX9uUG^a2+bj^0f%sLs4R>#+CrV6OM&
z*0<YnI^J#PlTlg!=z{I)?&AG`<wY%~|0j{a`X74S>Dm4}k#PNg>9M~yhLLjt@dY3?
z*xIO-pA&}_dJ&X^s&@S6uZ@vvEuVwwIno+RK*0sS{vjrhlAAJk$`9AO#ofK_?Y3-U
zH$cSB(&dfiXCw9jA|8ka4o*-NvVT2!_ZgTpvG6RE8n0m4f%E<H6dmdbo09|30W5}&
zbCd1QC`#4gN1J&%_sGd*g!8O*Gi)#Un;U0`KjTGUT^q!=?cEt4XhP{9$-R&K_qS*J
z+j}%suNZwIbAOi3^*&^tph>lfao<c|_a7F|_V<Xqs76A`5q#3%>MSt!yIK$iF{Hdk
z<QfDbvG@wi3RwPOHMP(qA<4qP6qR<98*(tfG$C)7hnf&`LJKel;kYJY_*&eUl)#MH
zfDq56s6j90pBGUhOHj4p_++a$tXvlnHSjO<?R$RUBH9EM$kBmC<9MB<Qu_h3R{k7f
z7SX#r@US4HWs{>rsPtAPR2nqFk`Q)zRyFo}GU8KVIb4k-BwF#?XS>{3U>{%(N)!lv
z*>0mSmrlG4f<C7%25zm!Tp~hFeR<a~&WBWz7$YD09~Waf=@dRJ*T~PZB1AC%!kqQd
z>IAtB=tJC6fVQbv`G6}LU##4gMPY=BO$j&cwIJBM6$r>}Qa^Q_D3U~LZ?dh^28@mA
zbE0``k&{eaW!xyfzlndcDU**Aecj0}F$^u%S+pFooZ^C9q7Q|*%G?qXWzwpwjY!7q
zJJ>BMpskx0nl~^`30FA8CkNPPlQ^7j9*ZLHwF<S1>=bS@F){8(d~GzO2j<_g+P!c%
zqP}>Sok!Zwd&aY*3*Ar1Bm3?iIkdE%qXf2^4e~p@88o18nRn5>sT)&V7w0gWyx_V}
zShg_2(%gi)PyHdtspQsO!-Wvcxu3dohfTx&iV(uQ3lI<F_>_u)G#8UE^f(d^l<i=I
z(>FQChw<3FQb7gP`!8NrU84!1$T)Y(&7LZU@q?WH73#byX+o5!lz~3_@esM04BUO(
zVjQ|8VhL#yEE++44PU~zFy{Vc(|cQgIf+wi0tDJ&upiKXmkJr;+=m1z60*j)9qh3S
z)5<zakXODfo0&ueOipj+k+ds2;fmIhX@`SJm?)Q$$NNNciH=*v^2sXz1SlR!O29ZM
zbfT;!|6|#ru)vs5RK8YL@#adYa}Xvdt=9|UlFgNo7wJJ2ra^&C&l5`PTxM|co^Z}a
z`D)i1#S~Q0eIvJ_c)%MK!w5bF=F&3h6i{I;lu&x{WMw=8lQCeS{AM$Tj{uHVE`Rb8
zi+fzt3`UY1Z2EKVvmIpULpneLh)(<GC?F*n@*$~;J4{bc)%+BIAbjLRay4wd>X<{9
zY0^!=&b*88a!M$BI%+ahRf8;{(fLd>R#c~dK<e7(>%A1GKvcJpnIF=?m1jScbDt|X
zy>4|OO?F(VB`qkC2UE@u2DrQ(A_+07-ILX?zR1ET{=1#aIZ~byg(moqge%HDi16<w
zR~U97Bc_0xfBv^z(7ZE%UP8PxDH2<}-9O`@wO$C<`O6M}`fgA|Q^LITd~q#_GFOtA
zifTi#?5vbfN!(Lej>KQKqy{T_q;075N$?rad4{W>_VqgHqqIr8m^q5q_QB#Iiw4v`
ze6>7}p<q@LC)T(oTeCb+;n}Ak2WIims#|t_u56F(F<C`n745_e)3Jx0FC63;(aP~D
zYfovNU6YOvg$E>N$P&-A^}wC>M5{F&T59z}3<_9-O(8kT!JxL3sbrHS%3@i`qJoXy
zrmBi-=ET|YOA7UncTUz~ip^5l^ngj##srS021+m8cLI1S>5Ug{JA4v+SWQ_@Tk})O
zk@f*%H;79kLLAlot9bznC8<NYIzvnXgbULtLi+rn18dyI$pJ1~f&zRXk+NA}|Afqh
z=`MjzxnU54XNYtg#fgTd6w=<UYV2S~(V{l67-f<P&vdgXl^%w}03QZ@r0w&m+6Hk3
zg{&H)$Q?noc{Pt7ue69#cXGw3Ic!`XGJ=cq?ouM24;wA&Mj}3A=wiAviLiNLK)uya
zJn+wrT7=OUx<vfdxpMK?voI1PyFD+v<mxhjWOabQcdAKIiqYv^HdQ&zJT|oSo}pxF
z?NI~$De>uXtR<B!&387IE6kJEZjAWeM+<IovC*x5g@R1KoxAvvu_D>63rn;ljLdKC
z8OA}vqXqL3n0Y~IH_o;R@jSud&%%PrnBHrWN@Ux}bK_%j@u^b_$~jSWHD-D)fH+aV
z{Y1PF=jv?wL)HW#8QI565<}h<G0V*ZoF{;Y-)5t0A7ZJZd$WlIO_BxID`mZoIarPP
zSs<)HOeu5n!9d#5$yLe6c1i87Yo92c+_&>OtkdI<S5~6eh$GEew7Tl%znhP98;(Eg
zD|Dq(SzYA@ldQwYF%;&#M5haexY)*X{&?_MlJW$ow0Nae^%IvHe5qZjD%+jD{Lq#I
zok8t-S#x!zy?SAPL)w2oO=*)kKdkz+rspN}slTg6kB$O``{IS3MC4%E#Sbs>FZs{T
z!hChcFoNyX8QQiPf3LnETj!ai!TQd;aoArcCys+_T<w@J)4v^Bn8(!=u;~`)oa?!l
z2Pr*tr$_EH;}X0r<-1nDqpIjlM&5hn+FsXHnxE}`k3w<g2UG|Ri><gF8+`eB>D}ak
z{}O6g<@(ICK6`7wX}z!cM)W=M4|=y&axS+gN*tMtwJkkKZ<B>im|kqHJP8tUXp3F7
zYdtMBTD!5O(N<~<pxNqzsT!nzq{k<D`8|d}|C7}C?IEHkur;uN;^O+h!4#JNazgD|
zW7>we9k%CHQOQP~U?uDWT@T~$O+EOSHCPWQ2Q}E=8GX_*8#wpxxP&qVapjvDm)d&h
zDt`ypa*~7s@&_Uk4sKb854Sd7&Mwa-lLtbzI}48wPQV|6dtNasCzl6+0rRw+9@&3w
z$qF1^H}?z*7SXYzw}8CY+tDF@+yK6Sk?>pFZotM-DsKQhw6Xm^oI56*7ivC*E}>2V
ziOg_!xIo@J-;d|F;Oap)Bu+aJIFm?7B#~RxF|7W@61L>63{Tk_+|2^!@k{8uxUaI+
zoV+K9ys<PSl@`5mpID+#=+&{oq>ErDS!H=aq;Lz$?%_Haph>#nqxE7^OJP@rxVooW
zrY&&~NFSF8v?J{UqqL)?tY5G^N6yIe^9%E%qc_+4=XaBJc`BOw-%rP%{u|L?`GFaH
z9Qw(bKHu&i9&byEG%bIEhZx>-lNW1ee~)R}F}#z^o)$PP(M8~qTm#axND5o66i$F;
z!cw2o$SU>plWy_;2b0eze?k5dLUx;4u?#Ql9CCExYt9*&GCS>Z$1lWU2h`Z(9)PxY
zNU!({r+~JtEPdBK=r%KVZ`?6!*<oPGz?+^+m{H;#yrM)B?Tow666R_TKnQBx&!B26
z5(CR&jhJ=(q)}nkK#acOu&pA0Kt%YacTC3-KDjnO&-;?>LZ^YX1&we~5sfktO}s#>
zf>UU-U(WlokgL^I!On&Nt7)s^9_}#87QfkrVO2w%v%hbUmr9}=4Z(SJqV8Vnt++D0
z7K|xAebcpuj!qjjSLbTaB_TJFKUO*OK?G5YDwRLSk|hzH#t|e>C>=X<JgPw~5+XTJ
zBO(;G)hoZ<duiH}N&szseFPC<U>wIyK=={6EvX6^_l68D`eBH-63mZSA6xK@=lY(R
zyo>uJPRlc)_+tWbXhhgvY+Y(!m|kh|S6x*ldNS5#e^#CGG?YMuVFGrz_~jFmJ|2N9
z1CVb!qHT7r`27fbV;Z)NS6|=F_j?nb86`CB`Hw_i-{c?P{XcY$<|sb2sV-tC4WkTJ
z9RzhiMp;$MXTYZzS7N6Ztmm~}jjzX=cMU%lI{O7srh&-JEIPdymE&u~nuCzzA8tKh
z%LpIOeX>MLgEanqx2cey-dPH!IJJg#i~mZNqWq!$kvnl5#s|!;fQmkHcYV%2uHyH8
z1zmD+|4Z#4^gx*qfLVO~f&%O>g5#0eZR3YV=OXAqL5yD4i$rCW0LkNWv3Y02g}Kq3
zV{Mw=E#rVfj#9XO3)1qwn9$l_MGG!lF){h5wl-Qfp*6QrM)4q`G^Xx8aKjU>HLsce
z>pCp|F9S}ubg;BQ`!efs4U`C7L^Qw=kZm-=g&*TC4TBZygtYHFz$Nf7(&SfR)Gu$%
zw6Rhyr3zjdq6k1TJ98`hJOU-9n4D|{HKhrlR(QQ!V`rS1MA9s4n0rIoQspZVSOypD
zRUN&5rrsLwF~NPVd6@{X<u%c-NVBOF1TuuHOvuc^Ru>-?|9w4)>%SLE1xReKx^9dC
zX!vAiIXDX6W@h?y>fz*5(`sGZIO9L80<~_E+fkvhLLrrKg<*zZG^CLl`fF{ANiE-T
z!Tv)5roqCY11Pw}M7Go;_{y@Q_~pAxy5=Y`ZAe*x`APQ>xS^3~*{UIRXIi>teT#F9
zo}jQBT9rS=ea}?wG!@8qVSr~3wra8xXXAGcgc$9(IhSe+#ghJlM3C(4!VjVjsBo<O
zWY)J;f~qRVY$h!Z71OR=Z^He&B5h9`%i?mA8y8)xgq<!q8F5@BpjuG*jk^ZY=>S^1
zWJ0M4`oVLG3W@|XQcaINpEl&0xN=#NOi^u%9BspR!Cq!te6U=$uSQlO*$YGfUUT_?
z{Y0(5w_Z9?*{CjAh-p&>MMJscZPY<$(GQfIaFv0W4|r6iq&~4g|6Wp7ipdFE%@nuk
zBA((fzD{!9$aafFYRk@J8y|VzepS<es^7c{MC3rzD4x|gbp8iqu<R{UWjzUbWUs(1
zs(V9=C|xv}{MaMn&}LgVkU#09`hud6zeZ)5kP^``s^-<=!CN*MQL86(6<t&<H?1bm
za5rBzz_tAec!=CyyUB}3@KmzK=!tdaT|XKpWfBiNgDDI)lp`+X%HzDC4b0P<`y!1o
zhHb@g^jYAPK^2&xRH>xYLnD`Fc1xq7tNX8LyO!D|vEQiv>s7%;p`Ak-x)lmB!z~m<
zi6_m}V+R-W^)mQ~o&FQSWF9oDk~)p~{6oOE1&qV|Tffh$;0Z;pgc8aSzbkqc++EK=
zGgAjxO(Yd~DI@R<5!PMO71tHqn|1w~$gZMFl<@Z6;oF^+s7zennuq!}WE}9T(nj^V
zOy=s8S(^#ylqX)|V`0nI+7h$=`L2q?7^%AE^dC#kG{rGr*-4w?14pJ*FZO%bBJ&{<
zD${ulto-Aww+&M9LpfxLxBT+VKJgexPo(m#&b|GwmWFrP?5s*XVGtHcPt&Pu$nZyl
zbnnvYX4fp_oUEr96jcJme@H6vKP-3k0&<XW9?@BzDOnToF0|4$90$lX{EBhHQ(NM8
z=a9{eVl$aPVHpx1^F8B`%>^16QL0=JmlQh2_89g@gJI8)rP`wjXmbx9>aw_!oJ7P6
z=c3v^Gu2fBuVA$`Yqt?LR-StaFgSrRo60G58m>?{@LM0x@qNqLEhPsVffF50nb>#;
z)7Yz%7MqnEucoPtNhC~iz|F7H(<Lj#3NSgQyEFZ>!Rc7uf~@0`xW7#MzRvxc2xR%*
zX=PN2tSDtwClK!8-A1A_Bk34F<IRq=`hDU8G#<E*Dw%_XYN(VO{v{(7{M)%G4))OJ
z2pZjX34_&0({lU`c1t?*yZoDvxQc(*J159bEk2r8zBPw8opaA%eEgDIyigv)-6V1l
zX_YWQa^b7w)5&Vfd@u9K^#<<Xv=cXxn{LaHpUSQ+`x(2l!LE;X?Uw_2Wy=8i0Z6nx
z(T58%W|lSQZ7C?V^52eW?i|aCI~70fL|LIduj<Rai_e8OCMt9b#>Q+eg^~ugn$Cuv
zd{cT&oh95(D4mtso>zSy`*Y6DD4`qE@0Dsam4$B;6}t7&B=d`t;^<sc_P*h>1}${c
z;>}n27>oHfqV@nYo)dAHU#}><T>%CNZ!joqcEO`c{IJm0N0yV7QL3C2cx>B=DR7Bs
zykZVkB>GQ7e*Y+gWyj3Uf|9e~f0|MSKJ{<twAps}yhR8>Io7ZXdsMf6K+1#ff8bp|
znF`tk-X0%LGqrwTlt`j|cfNyGu4cND)??UjPfQE1ncSO*YJ6{zyVu-R>84{UQ%{r9
zK4hohf8a&9)@lBuGUWK*%8;3b^}m}Q!xoRFVtM50CEVgO`07exGe<E`VB<dk%h)87
zfnc}y`{$cCq3L<fk{H@UW`9<6=#o6ytlB;(>znm<9$Gs2xmcPRkZ3FeO(lEVn+w9T
z_eQ}EhvnTMKwmg?{yn?+Plr7ftQK0QuU)}Sz038jdUjTVJ?qb1_?%n#V<Tw1Ueb`{
zo-rZJjOq``LHz1d)ClZ>eF$<^-|oP?^Uv}3C|OR}$rN@+aqZ9Q=I5fW>^B+=1Zfoe
z_s1zRHvmbR!$T3y4VH}WKbMd1^HMxZb;-)br~6m@Qk$dGAtAE3@GXpLY*MIckXJH8
zF2yHuM#T0(78|TM3O|_nx9V^}+j!@`+1YT=+;bVEDUxlD$;Y%gTtjwRY+mNrAyy)>
ztJ2KV$E1foSl|@@hIECuD5^q;y%2PHsRN|xibzlJn218~sAV6P?ZUX`m^AZXdK+Zh
zry(+|QjVd&d=YQKH?u>;0^qE0$oXGrkVYv%#0#Q&qD%o-V?z=&X*g!Z!zo6?nmg!1
zGE7}yQmaA>SqawZd58&^w@kFgvjcB+w~tI}-$<^w5JZH#W=mZ{{pQKOfJvQA(F9AI
z@9AzBZmC0vLkKPE^z&A`gSO#uIr-E%%@A^|ia7KUL4&MISOvuO5*DE<6PSC)N-aJ5
zPp1U|z`&@CGq7KtQD82t^mP2kG>;R{C~=lfIa$SBElwZ|d`2rnoYk`p5a%rN>im0;
zMG*Q&2_hWP8yU^!EN(PXbdQ5RDH!U^@r96p%sy32J$r*ZiCCnRHO;(<!>;DJE*_G2
zq>A*j;#$Fwmf%oQTKQv9M~;v_jw3rQf}jwyZ7kkpePTl>B5~1b_18XvRcfz&8=NNB
zMBH|v_DLECs0i8bL~*-ON;e7^fr!d4Y4RP8Jo0LKBN09=8S{7#7dWmIJ4n(aBD%j_
zY5pZ9MK1ehhUBroE?~I^{lky|<5iym+wF)U*IC?f8WXl%g3o2nPpqg>HLTC4=!IaU
z{Z7Q~YB-u_%Md>qf=-&gYqBkvPZn70c!eYPiVPjRIqEyEEqIiLEl#>B9sSg&{_Ez4
z@D6HC=sRJyX(1)76^Gv(mC3iCd>=SHX9&kg$Ole;S({NrCE%ggK}nmw27G{sI_XO(
z@fZ1J^3WI~lD+0r3RJbyCtv`x#R9OH2O4t8@z)VVz7A7wIpL?G#hxD(494wPIUp5_
zAOE_8jCY<Awinf0PNc!!Fj-|Jhtfw(OIGz<E2Lvy?#nY;<Tm$u@d9LMtX%{h8e_19
z0*O>)6VM6&l0EYn_pyVLg(a%*Bqy6gctVyt$u_?UafKmEGryZiI%IKNDjR(QqC8$+
z5EjV3>jbRhUWo{Q+AZ)z!USR36U8v9Ba0MLWNt-~RPkN?sGb34`;c{c+9=Ll{8mhO
z=oqBcMR<@9&dUN$EhRHGJ#Ccf6eUu8{@>e=^AL?-H2rE9Ysvt9#Km2W`=0Ng&a~P@
z{TY9;k@4&mf?cDyXJyKeTkT@7@JbweXw7SE8kE7f-b!y^@C~**K=2guL5xvYFVJWb
zrN5+m@y61j_x<Kpq25H%*RE`dBR&LD{`XRP8lTOelyfoS=l`IU{z>Ek)k$sqJNVs)
z*_eWu^Sun0N9=l{mlNRO7xQ731ikS_DG;zIYWUofv+Qb#s=I+HI54d)D8bd;AS;4q
z%GTa_aPubSN=s;dNi7R2@9B&*hHt&e$U4R*l=?JO4dkqc5@lkqF)An;SXMx|G^c8`
z*bp}gQ}$E*j!aDj7M<_CU4li8K%8?D6y2#~sfJ4M1lw`CiOG7yX}A-ADtv2*Y$<_<
z?y1;aIvo)V+C#cCW_mcS_a~+39Vgr439Ru%eb&V8%&j|}*(RS>I~A9X<8n#lD+2Ec
zr7D>ErH(r*^u?l_24<L7FKx?mK_CK{C%f0wD253vH=iB^4R;%PBVQ0sybG>-4-b=y
z8D}qMsVm9|{2@WtkF5>&rFCG%BU`PkB=2pRou3UHi)EJx)j7A(ZKa)+!4Yd~oz(>r
z9dy(@bJG6FC##^eym4)Jfh&yC!w}^prDZQ=sZ`P_kbL4UiEnOUstZ5bU|u9uk_^1r
zN<es|%%7_qJBJOFtW~=F5yk?Wc~NWWN^qL$t8+nVw?-j@*pP;+z}Rk{5))x7t^1rN
z!F(6U_6Mz*<8h+MXC|VBhGoj1TRjNbT<>K1$$*Z=c_=%V0;(94{A(s%86V5OTD*v9
zyj9oxneh<GI)GYZ@TnZ6mK^7lb|_y_FDY<+<w<PR@Rwqv1iN*u7&Un^-jTMv%BOmu
zo(j;+`7{Ks^VOy+2WRI8vph?{&P#M3K-GcZuJO;n^jlq7qu25`()^tDli$iU#~z?!
zE?ZIqi~GvbgU-uEn(bsx`ybo52(f2V(CD-4iU^zunGYn`#T$-M=nVDJ0mw88lK#u-
zD@O5rf6QG#S(ft=Nf?#A1)`_>Gm7@1Gks4f=y36HI2Rpo-h{NHv3X)7GrCG~kP)70
zDOv7#+$P!DON>0gNc~#@9Aw^ZHTf8an227$6C_5*<2fa2t92hxz4^0xKTP!CXCz69
z@h&;wrRLY|zXxYt2Aj#g#YdWP93%AQTm9PPRj++vc$BI1l33Zpd^F_L%5a>Cg&D*!
z%|bUzMPO|%#E;D*6`uD_e$<{>``Tr&#`|V*UVc`|$30lYIf%1Lg0`l`j(+Va>#N5T
zlheH%Nx4Q5KvPg5#jQOj?oe7xJYnX6*?ruONBNxmRi<<Uz!!;|={-%2g0rsNo)(Yk
z-R)7I0qs$X`zXbmkSs+8Hw@hsuX`}3YBIn;j+nf2?a51h{NY#A^zdYy1C-1+e~GlA
zy_o0E^KWIHIHh#e3l`<^pY^!!C=@#I?(~e1oQeppvJ-6m+B<UQpN(jKE5nCm_z7Fm
zPKD<~n!jFL-5cg7+x`wuG@f{hOQ&VXXD6Gy%9rc?T##&0B!jUhRVx!RTcXCE5&N8z
z_u%3bLtHmx_t&R)Z(0I|SW|hNW!rYpK*0@$pcAcxI5PN-_cOIzrC?L*=g>V!tg9<{
zbRra~3f8wu6{<XYbrx1f;>pG*w5FkLbZA;nej>f_yPmN~9ll=3FnLbjIt%P8Pq{9i
zKUA|ubNd#h*+wXxxcVksMb2u$YN}atRC=m>Mf$b?DB#n1K%srSs<8PYbk%zA60Waz
z+xmC3JTD<QpT5y$OcF2H{kUeff_<Km`uzC(mNkNx7#2SNCzZ$kzss>0SQ!4h!L;k&
zJkq8JLiZ)r8?cS5bvHmDAXg;N5t^aRFQDow5&r%kzCyQxm3(1*2ykpPjz&?25lwA{
zrTK!4F4^DjmQK$N?=5869T33S#l=*}r+{VMI?i*W2Y?YoDcRi;UfU|+(nA?=P06Cc
zOZU0LEgu315AZWo1YEn%ex@Z;2xmVy)tvcHiX9YAH{IG|_fn^T#&eFje|b%Rc7C3p
zK4(<&tHWutO9tO=@6BItZIl4nF`yTBcU{bLAYnkWg67ny7GHf!=sI~tDza0TO($a4
z&OcYp)84iCqg42tX2PlYKq|)`j<lGQXi*W~`7tXEx>oZi9DxAQBz(5O)r!YdrvDHL
z4{t^YOhI9qP6TJ{6PtdHAJCa7eaWnI0K<t+n4vp7cXDH&>UU+v9L+&GqgQGS7$%={
z6`9@}*`|Ggx=~u_i~~!@VHnp$rq9y#d5xngtA__?j!mT(OCZS*pi2!ASwNDg3)=dA
z=2L@@v#Y=am%Ni%Hl}ejSq?^J>Em~dZo$4;>7nWo*+`P<A}!7JtKBOH0HKz<{G2=F
z!W;FC`p@q@(Aj6p&|5UY=o5<xe;92Nmlrwec88e44%K?VDA414@jYRNwxfVOnSt2r
z0W{D?yN-OyXPn5S+QAw)MQOcKfwlRWEvd~G3j3=n|A`Etj_x}_4P{$iY+Ium)VxkT
zz3LjEi;E3n*6DD>J1u`x{TT;AX9@+uJtAg1BN#rs>VTyXw#C_gV(A}Fke@m!-XI7H
z^&!2M@-EH-wf$7y;ldFYvy5kp5(OSk6n}O2V#dOVbzC?OfmSJ3G$64k_mi=^In(~*
zFLbzl6<@DlbK}bkz7wJcr~k^$Ee~lH5K2LxzjXa$?h|v{VB4kfo4^UU7Ej8N9-YS4
zpWdz6lh)92kS&uc&Nx;RsfbXVuxTs0)H~J7oh@HuoYayRPi~mA4p~Y)XgdmlDpf+$
zVlq&WtThbWOy*D4BF)b6og!}3eqEMGrW#VP?4+CrDn^5%lhsw#kC+KUkWEe8-VC8~
z%o@)YFWUw|!*Z_)2)}t&2uKqPrIBwP+1BS?h*Gb1)om2v<YBgS38pODMn#i2eu0n#
zn~<2Vl2f6_#rL3ku>hsXJo8)>U4C1)KZrkyjAECdSxfyFSvNML!eHj+WF*MI-dD+T
zG`4~K5+oc5chKRY@ZPd_9MYU-e34+f=^DL>1><c}te1K35txzH+|+14IpbKQmV%5U
zrV??aZ!!(0(E%_`LH22bs}|!(E+N^W`jtP_H)t2Tj7I|+jzc)v3}kwtCfGg81{a$2
ziAB4zy+w%vE3mde2|S$oBB5G^0|Q3VQ>hSJ%`{X;k|P;m7$rHq>K>)5te}|GCHw3p
z+zaY)G+wvKC<24BT7bu`(=Es%hd(jqo-3RS?pmbXys_S9LuYG-n52lqk+a4x7?oi-
zNb$a-cbu@uy4&H}>>$o9GLRB)m)EABC}P17dJoQO4)VvE`&)|}OWl^M5B_|fp7Rj&
z-|al>)%-$FcJ5}b>qcR|DRP+aJ}E42`ra|R&4S#V(A{mNxTiUw{o8@bV^g@1bd>#I
z6NlVZn-O+dF2%aYH7(70%^kDS+q<^12tpv>5Uezp+}7F<@G}!BUbE+Ch1XtKOSI{3
zdv@F-Y=t9)j{TK+Gs&&hV_ml^2g8`=ix<)G<!dnCn%>*{OYjO0!FyXOz3?EG-qyw`
z3&2LAc$%(KORTArS(Z5H*))H?8qObV*XS}85RC^IX5F6`j?$$8J#7nOR@#9v{wpmY
zlq{;`Q(l^g^X%2FY2wz%b1K4Iu^m-^!Oxm9^ojbT3&u?~Q_u|)Xp4WL6bNS!OK5xZ
zS6}0!^<g|1w&p1J9j|Ytc6@||_r`hy)e76I8Z*fn1qVnQmuE433v4ZOcq}N8ot4eE
zK2C6e2G+A%VSNkPHuX!zxPZGq!I|wMG`L$d0+srmsYj+4$tyICyk~!7JB?#UA**ol
zb6x%cid*dN{HLD>2mAlF2v`~aJG^1HhNLY{J7Ujl?HMmYhA^70AAu2b1I*D$unj+)
zKjL3-&-zY|qubIq@`UM=E-phzGikQWs?kvc?zN=JR$<y)ms#6R_*yJ<vM#{Vl-644
zx!=wceuYAoNMHn&!G{;a_tC=t393k;OO&%{$J}s#XOx1&=6C>f0}7&ZakB?8iv0gU
zl~FlgTs$URXYyNvP7==FpbFd>W*F;L?$7(nYc{ojB%)PUH1FZ!XzO-vr#vW8U>#U0
z9k0ths0G4CW$+lZwf_6%?&RwBX}py&skzjT>CwY#tFhU8bsh$BK6y2ZNid;U02;bw
z;M~2k2%>Hga}j3Gg7Ai+5TJ^NkJ~eSaah91U=#|ZojHUd@f5xM%o2zQ$Ga|)@3O#w
z>ck3EokdaAkwneLW_}&q$c_gQI7?i&*a8v{TKbHj2b!!CP;)1ctiUv)`y8~8Ncd&N
zaVs>vuA_2M6=ZrQLo`zcVK{F<7COImnQ+n+O!sPOXcyiTU7tLrK0y~-jAad?Xp_35
zpikeBrz-t0wYWvBwM&dS0p0Pp*gHrGZa|6O3c^teG6dzp(GnWJV%wGa4Pf`<<vqE_
zNIdppBq+)VqaeK_A9WbP_S`y`?k(z=B;i{8<>d&2mhh4#AZ?V8eGSwlO>>=sa?!Pi
z^;eLIF>cK{Wt%ih%MgolTW|p)AlnDL*|x!C7p=7101}zV+NpEsO~aT`(_DY7(oy~t
zjI{n|>XByi6Bhcdk=GE?=d#p^3{%ZR)al<_;@07<!gY|1jUPr#(C2DuB6ubt+lSf_
zko~G|!!3MrtZf7M&#^+C+mLiVx-P4DI+TjEZs&zyBE5MU_e;e2+`TG_a6#QiXiWrW
z%xPlb2=rxSp`Ht{9uYs7NO={v-LDVj7N@_iAFv&d_hivON*-F%7fu<mHpVIpsH3MT
z0}#es#ZyNaXAc?6YfUY8&BsM?LId%Ou>JqaWbi=U{yIcR63R#6OeIQ;6d0@<;A=`c
zc+g^FHv4-TwHwU71&tNUT$fr7jM^S7KywW+Gf#d}G2gh6V5LQOH{C=IDSuEJYrN&<
zLZ>##X;DrRhT!W&>jv-@7iDh*=C8xf1;C;^JK^Tw6{L@jQvndyw_g5K%~R0TsSDW0
z)CbLFJlSDhQ@aXCx|X%UHq^`{Q`*8sIe2HD9&2RQh~t|%d|4oL<xU*|R=T<2;NKv6
z7)FU;?XD?^ACJ#aUcdn?r;FwMNQF;-iXG=uBmw~kz5!fuD!VO=1z?~RLBVA&vZ2a%
zs}PX(SdCLaBr{8c?=>HK$$Mc5=zOkg0k6#(WGW6BSUL6@qPY9}71lRX-&Q3D6pOPp
zuJiObiL;}sidocuDMcf5vcGO~;Y$dgU%FLss9Bf0o)S!Ti$F9I!~9v~0?j8SCMZJ;
zuYq6>$q(T}L-)Z9>DbzAX~7CvZ;~gIbtZ@7SUlSnIk68Ea*w%C)QWXmptGTtQowW~
zVK-Jb*2U+3AQ}xm`Ia=QF#n>=sRL|cf0(+V+F2DFj&r$3ydLUB6ZgPdBPmISGfJeh
zKRet6gGDv#x4{2hO|7v|Z;b>ZqEPAOVD+EYA(OD3^^K;-Jj5~pveWL*yV2HK-_Hj+
z1m?Y7`yj1298Njq2O?etXEgHSp<_>Xk^J$nZ=^|M>2vWqfp~`wxGGrP6kJLThzuYh
z%WphQiZ~Cb;_bbr(K7>4lR1`pn0NzgW1rnmk(7#hF-~QM1H~eRkr7d>*}?<Et`E~2
ziNc@8s~+0US=8CQb2{r&IT?C?WRQTxIl&2OxN?K}etMk#_<kJoYt-s!CB_dmW)Zq;
z^6jup@nB8m6iv?H_HZ)>P@99@`fmt;z#7a1g;glQ^f*<e&@{7TIlbw*mQOm~i8fp2
z3&q#Wb1uIHplEG8=6OaPQ4RCkG=+NefviqPTbb!s*$u81ZP6LdhXKkT633IbUAhK;
z^6)$NbD^mY9%f@+T^IyEI~iSoxwI6gJ_J}`Ie!kMG<9&w;AR!khtb4*TvT;`Y>Kh(
zbosQFfe`ueT~L<Ec6emobdQyfyqLckq$hZxsJT0(GF@D|rEc{5tx&$#-z7zONC{g^
zrKub{0KBQxNY=W44e0n+2t;WfRvjtW!`E~BF0>YygqxINu%QLW$iu7RTj!WbfC{ZQ
znX@cLg`|$Dopz`1m|91ks<yP#6v$m^4JwsaDzotCF3GhJ%9STBFqNCj3u}^h;LR(r
zF6vNqW~&EkUQgOkGA{~doM@9HszMK$s-&qMM!5kL9kQQre04*Jy^@!gJji*+Z9pAP
zoD*g`E^^$^f>r!B#WYosYJIeh)y~>+`dI(vz0zsRuXzG(U2z5VjOKB)JTy^BGesMs
z>U&ITCM;UW)dVP$y_1|X&YFKUTQ8r?Qp1yL-;Y$M&9o%#Doq=2;!Pi3co_7)XyvFs
zmW_{g{X%hP41*ZS0qm0MQ}AisnMc+gC7)BTEVFBt8!E!sYCCHNmqDG~D$u$xB`1bV
zsbbe#eM+7+uUpDOG<47vmny(a4Gq7kThnh7Q17sAih~Pv3O*4s56Iinb*%Aam&tSs
zx4K_=4o3xU^$WwTnRH*<3W6-U>v`7~I4pd5YzQjj71>T-7s4yBBC@fQr8X^?bA;&@
zYdeI~Ua#on#Wls=kINSHakJdSoiejZA9hpBXMS&!Ctt6x#@p3N74X{C9{uWA$NX*_
zn?}BJP-LDgMr^UyQT&X?v)|J!kdh@#q}1tTDi*0BUt67R+fv4;9ZYQP8Rj;)35h&>
zs1j9@^Dc=!M&;I<t`rYxi0MnA?rSL(E@5vrw`D;^(LApj^h&+Zgj^c^^rxv+5JZRT
z*uD8E5hC=GUl$c?YQ?Za1z5-5-S}~@i5`23L^(7u`Z@Kv-z$e@bxy$7C$k(kjy35w
zyer=nF%x^Gy5^*z6>dh!JuP~@hAkU*(@isJEWVl-DSv%`VN(O?!_+%=)b9AgTw<oT
z3vgu%#*VvI3F|y=tHjc+#Mn9!m|E^9Gg$iNN_eqbpcS3kZmC4B5f*=oZ{3e<^4vy=
zT!D|7L~lZNMq{?h6*$Vgdsbg8-jjeP3<k&SA!-lf%jA1*2A4}tC|^8I4Xx#o8g4yj
zzB33zkrWp)I*{)>YRYEthT8&b5*UA(8hyVJxI>09)M`n1{+}oQlrw(>eJ%voQ6gRj
zLd@b_cM-6f)ZB`++2~kNMYfKt-0U1v8)aTsc3I3k97B}75yQ|41u!&z+_ITjT*^yO
zZEOS^dLHDU`nK=6(L3o`G+f>@M{vVs8MUTc>nh6{3Ed_@Znh<_IHx?f9XaP#M3P$g
zEeTKmooFvx_TzFnNYNPcB6i*#Y&IBRV`;%#&FsR40gu%t%$S_*?T@k;``4%i;QmIE
z5jqYk%%;0-R$(<J9GG`Jdd>=<z#gI~%YYN9AJTgujMDI*JIUaoe}MYW$p`;Ikuos+
zKMbD#gFK|?VEL~Ot)_qDdo-mEx3y>Z^!XJSB$ElV;y4tYLOa6h+}%dW!;}B$6!`cT
z=S(|pEW3qs0W&Bq@c8I_X*}NF&!6s}2LGlHSL*d2KHQ%Iapk`wGsAOsx&a(Rq{Q~7
z?&>y)jLwGR7qTGo=J<mDwA9Xm`0oPb08YYp@pu9^gv8<RjV7~v@eHQ|fzL9gj^GYL
z%9J|y{qxr!Jb#Xj)y-+L(x^OJn0;dT`8+>7_mw4rL~C~7oAmwhkw6aNlkWqyQW80Q
znDtHN&Go_UXEn-LO@gmX_5INZiQf&1lxIS3c3iP(gUC>~L^BN}1f$ks=C!a+XMCK2
zSF}Y!*AqO3zXf#pgGCy01#+*Hr*=}SuY*m8A=RHS!qPh!Ft2hFeZ`jxZDMS$&dXSy
ziAd^PFzIGIw)z`SUWPiM7rSO9vUNbLZa|1VIsmZW(>O@84>m;O8Kt0~V6`Vv8y0)Z
zAH*~!qBK1N6k+(w6=+K3A4G1r9rfjsnGUQ%EYG6p61O~!N+RA_XRnX|#t;XDFB*o9
z`u6?Y-UcJ0KDC!58d)c0p!GeLWG%v}{|D4K0RKK<dz%DH>1uzZmFdP-6$N{3A8c{o
zWP^T?TSs;c)Wfw%6N|uPFkj|GqemFOseO<JmygVzn0{Z36Fh8p)?UqtGRQ6UbS+Y^
z6{ss*!@UWW=Uvr~Au@)j9*(8O$fcWyjHofm<l!$rp$tNAZUNgF^P|d&F}ZP3#z}k4
zOw6C9cWM~_)&T~2mGFZn!bX?CYibJ@i@U#oM=VrJ{94Y{+~}qPagS*7iAPFdShN#q
zh!q~9{x`qY_L3Q}Ddv^I8>G@OcXUmRGZLb(LRX-?Q2HU#EfWbjPT7bFRd-@wcM;4;
z-8vW+Lv>f#gO#0ZH2y%OROemtr~_>pd#<tik$%R`(1hoF8H;Y((jxdV8>EP&GgqJ>
zRlp3<oJfU5vajiS#H~{DNwq17Za&t!*Uu`wZT<x`6rzx^;Bb~odQ0ll?6p8Qk%3K_
z#ifTv))iDG1S>{Y+OY(Ql|E25XQR26V>Uec{&ZU-hd2nWLW1(z;W&>k(}KN$DlCpC
zKRq_GT{a+-^O^Ip?$tkI&Id2$<{ZYCsRx_m(yA+BBC1Mt<<*l&Hk%I!VNav(%k6=#
zziD=7{l@C8q)X{DGLe8Wj*LU=(EGHZvCfPr{*9H0^0r+AAf5W^=!Q_9`RG@k;pE+I
z*jca1=A=~&g?w$P%LlV$!A&6)R{HJ{e0!0jl9qJ!cw^)OyaPZxY!>S7f84MQd|?kI
z#NFu*n8s5V+MxDXd1_egcg(}#dmMIjMsM6_N2J%Dt*`ni(Myx10}g^By@cUv9O+5W
z!Vgve<gyd5xK3h#uO>6HBGv*BR74R)&~-%KRQJctr?nYRF3JR)yTlhi<Lo7U;w!BU
zN4~exy>t$2C=KIC%UTaQ)()G$2}f-f*addP_33VCM2xF;H?ddz9#^P4*RpaQ#|Q3D
zyy(WRE$SPPCybIW2fDswy6J$Q6g-TlokM6j>V`e&74SW1ucw^H0b1l5D=W>Ep6LHS
z#@?w(lt69LZQHhO+qP}nwr!iMZQJhEwr$%y>zflZbF(L6SJWS<y2^Mn^GVOuEFhHk
zo*GTx1_XXGV(jdWbmB&<1BqP9G*4Y~n#N+wVq@B?y36QCsAf#{HnslKAyO|DuJ=E*
zVYi5M%K1oNA?Iej(;u%zbDsufDQojUgKLzcd9g3jSS4abMBXOiMgeab1#mV0hIxe@
zU4!FQ?{&bzOpDx`)kl(LZ9iDc<*xPF1d+MODSi*wY*q?XgjGGOLMyVkKWAd<+|eT>
z_BXIMUC?2UThl6@^HI!BS%ic6sfafvR(HCjL<sd(vaygu5Nj{VTtEv5o01n*F<$p=
zZ*LjsJ^c>vM4@Q4ndiVXqxxgktM3iB27Ebff;ZAFY2j)TS#D>Zz1w`()6I52aHR}e
z&_f{v4IzB76H;#|@L~Gb%bAJMpo|7})55v8$hAEk((%FYQ^DDdxEV?X+`5p#eDIyZ
zjsJziQl1^d^USrZsl9cEUB-iv9T(W7B_E1F>hVPjCPAK-IXY}v=_8{S>4m}fIULFK
zV^BuFVP`IDQ38@-U*5b=nw1(PH*sLj9rKCaY%hLf*mhiq77D6(Y0B`ONJvpz+484t
zxOu`c<XqXgsEBlwi=~PjwrYN1-B8fWywi~#{#rN;r<AwhjwnYOI9m8lMr^*oYx^N{
z^7SsUv-AFGN?|hno^!57X$>OpQ{~QhxYaSCK%gs5?6uy6<A%uImgiK-DoNjVX{)UQ
zUF+m7uVSjs)hp}f?&!fZgN<eK<K?&or9FDcTG4a9RBWw+eO{_O{n`LRd|x(#y4V+Q
z7s{QwQ6OjbrQY9KH|Lt+clYfFNB)+8n(`qg`Xdwvqn~(I#6H2z!!VCeFe4rXj}J~>
z1RhRoa;_#DzVem!$j~e6EuLMrfGLr3p#3g;bAM<B2R%B1NZkM9$f|Yhc)+vG@n<?W
zs@RGj*9|Az>dRm%v2j<(1{cEdtp@XV2mi0n-+AVwou`Z56RObwR{fFoYXtLHI2h9)
zw@)mwp2|Dg&hC-r=Rx%!9;d<~2dXe+l_q;ar*Grp*G*}^kntU7*{!&~K|n?d+dhq+
zRtfm_vnEiFQ(k;ZDSpRU^Jp}_LUSin3jmMvodo4X<BOMLc|Gi){)edg>jwX^D&f6$
zvuBqAczlANd>vdVa$Ae*L>D*&cR72!TN4O+xw<>0!Dn-oNDkN-eOGlx-(>y=rBi_W
z2CI6e2W1?YGPSp9_SRARnY~}oOGj?=|0F*C)AAe(3&VdUK1Q^p<FPs6dOy{d@au`B
zitz#rtFMcJuV@8#(Y!H0lCKCcrw$qX{VUp}RQ+7JdBYp4RFEWni72CJIDNB^o*bQh
zcDlV6I5hVms<ZO?<ofAJeuIw0gp)xL#D<=}j6Y`Oq)gwq9zvC@rqhqT$9|i4tsJ(1
z-(V5+>vy$6c*LlC1L%`S9sIA|BjdP^-3zxFaLr9Hz`tN*u;o5~`F_2Lo91rw#%1*i
zloqNbZD^oMRY~j~j=XzzA8L-&Y5GXeU@{RK5zgM<hk*<-AdGsLEAa0*pk534Pt3Kw
zM*dY0Gcf;BJh1-bo*|kiiogQwG5!?-gs5<jG%kn^eXNi`_A*GAy&ee5<4N;SdP_Nq
z<}EBkEgx$LXJS(zQT3!nSPxz|RZ)s3Dym|sD^z|bacQh>jN_j^2}(Z$voU9GlAK5I
ziPt|UGGI?eUe!8`(H*S*qs`!=`XP+Rub&x(YxW^{(io%&#b3mvzgs{E9V<YIqUt;(
ztOqZHG!SlNruavI^A9FVf6j8iLe-r;#S;tBiq}#@2(NL@Q>N@J<ipCDsxKHRjoRX3
z6-soEh_TPCwYkG8VLB`L<X4t0t~qUrBHs}wsB8jotVMS3+8pI&OV?kvzeRJlq|MZ6
z899h`NzPVDK<`djPfz4sWu}*>d-XWgPD&IKpNYy1uN<e1V3lvFrO=4;ub!yV`v1D)
z1E#gJs-ma0=tw3#Y9i)7LDPJMuV-rHZCmYJQBzK?Ot7BaC)(|^6$={o^qOd>rZe&b
zeZ`2V!s8fZzjLut@`+q#^nsES2jNj>MHNrC_geGOUog{DTvsV`y*Kpyipl+0f)LPg
zBf_PuloV;Y4=O#H_PY-0zGXYXV6H0WT#CGzC^RMkslR-1^7wh|{xmqujTLuFvig2D
zgV0L^EOp=r;zUZa{y$v4+df!-&E|>|c*wpJWF|SS-*bc+=`pn}Vdi19_fW}mB0M$K
z+rnOL52KXf6H#EIQJ0M+6WB3}ED1Q89wj)4Lj@-ukVErZ>hDtP8G&N3AgDQ$D=T22
zj1AAK(UnN#3FIALrIB>QAF0x~XsUXFWx}SIJm-)>#1b#7j~`(JW;>{Ii9xQOc-Kgl
z-;8lN$h)-^peg3DyyQP^p~}z6_aI1k7PJpJPO%)ir*CTy&#MJ`GA3Plyz?IXtQ9f-
zkYc<8)ksy$LCo5jp2?1Vz9ZdV(49cbLo4@0S2_mOfvW1{$EpUUXc^?id*PtVUXwrM
zwmk8bS^N>%*WLEq*hpE1YA8*lBFwqP=y5r%uC{mV-+NXIl6#yK@4N{-Fg6-Ra=tcv
zMHs;g0dZDr2;Wgf#RKRA1bmWsIt9tD>{0PtKZL%!12|@lSJxW}v%OEeo2KlQ)8v-*
z?RXTZ7ivQVO@0*h5^$#NJalk`L2Fq7l7&J0&1gRyqkJ&~+ThOm$rEA@5!lISf3<|8
z_eF4hym__P|9w72-+?T%_YXzHu?~v~a9@mnlJHS;*2_Nk_G25gvi5v@Qo?o%k;4G-
z(;r<9qhAC|J&sL_N2+=Dwm7C(F}xddb%1fOJ<B+ErxgtnsBcAc+x;9RQaMC0S1Eyz
z50t{g#`E;SM8z9g=1;HXx$s{+xomz{nSYAuirWuiu{H?Wb6uGO*2-!|JmXyt3Tvyj
zGS~WyToEIUVJ<JS)7}vlVo2^-{+pI&H&_yBi3(Lx@fCk<4uElzMIfp5ca!%Mjihjm
zPLrH)D67EsRp0N(ao1yM?1q1SJ;<nqjM<h|ya3I&0Ckr3Xyp;R!uwp8=xa!Cu~xL_
zo<x0SKRgFjI%ieJhG2roCyNVlQK!weG2ad>!Z*<5y8}+)5xfn5SzSP68V`di;xtLb
zvlpj1^w*zE(nT{M1JRWu?@)QN=JajJn`f-6sl&LeYikC}J<FDs7PuZxw$MZVXS6({
zkC8Rm%l@ZSAZ8T$+;LcAmx!0{p{mL}=Z~1?`GnbxcB!;FUPp=PgmT=v8&&#zgyv%!
zey>TL?B?oK%er3U>TaoLTX%z9S8j_(9`w5f?YoQN-KY=C^SgzWX2%=e)4nj5pt`T5
z=UMdWtCh)W<8;()!BUXE>jbB?cf1@J_Y`rM)|h~}n`)5qOpkF>Q~knzN>vV)OGr^`
z;SS2!^OFvL+DFvXIp6s~kEKVM9&2_sC2A%S7~Lj>s(-`3Y9hxmjl+;rSm5O)_od4D
z+_In9cV(S+-8+T!punW4VN1JpYKv&vq6suy(yTR~rJbqF64vIXPo>@R>(Z&B@(KPE
zBk|mYX%7tLtoB98vb^hecHDHO1}N@lgiU=9CAQ53U6KQxYEKv}&HbhG^tV(&4HoV>
z8z<K!=Ho0hE&(;y<13{|Wv%AB>3e;1R%1?$2xHpaAFi{4ajN_sp)A$QIVyMEH`9l~
z&YA7(1#2+TBi-EXip>-owPG$ZDzH~`P^I&wtX*CA-0dmiR1(@1w_TK2DMsdbH(Vyl
zdK^U`oet1}XpD+kG{$N_6aQ|?DIq?aIVeacTia3BSWyxuN#klNYEPZ)SEQB#&3f)^
zH@$L-u_d+n4giEew_LF1w6cf4g3c4~vgYj-q%*3g(<VFVOsH)b$*x7bsJNo#BO~3C
zC2CsBIbS1EKH7XF6j~@(;okYQ%Jc{UY-3*TBbn|G@N!Uq++`c{rRy<Kt&^<bo5W!i
zPpzG!IC`3!Zhx-8t_l?Xb{RfLnBMX%rA?~F9$1ER={X`0x4p+5qq&ZB@(;XSi}$H-
zFi++kuF_TlG--YMnb@ZHulsqC5*@TZh^mU?$2IsVx~N=HrFbU)G3NLN<Yn^YQt&ZY
zcP#KJx#4*7EBe>#rdRpox{if1171}}!vnYamDop=8DINo8+Dv{ZsC}z*j9I__Kv{f
z=k-w~!x@L@Q4;MMzS}1SpQAun)~3J7xeTtH0A6OVqJsCSQ%McqsVk?7_sO5(I#6Q7
zQ~!j^?GomLMM?8(z{A(mxnGfjZiMdPRK8z{YIEr^_=S>&&15ap*I+(6bnfUvylbOK
z35Mw~d+fI`E<W4DJih`Dv&MS{4Wkey9gjg8p4Cfl8R3!&6!^#ESCwB8>bcC%Y_h0-
zo7j4O&%*B6T4w^aH<j@+U{x8Noj7;lc!D#wZnymu@#98lzX={(<Y~?a|Je@QUqC^5
zJZ%+>NkF}cA@gi9GH`L!n)o;C7PMFI*pl-83n&}&8TFr39rJ(6K{GS{SE_FLpUJt!
zhR}OkH;1ngCJd(^&`{pf0Q;2I-}Wb*$YMcQ>B;eI_cj4X35E2!c69=cmB3mwaVXOs
zFQ<?qoIV&Q&mEq=%UvD|{C`}|u=H8r|Lt-H|BuUg_h9>W{!iDPdlf*JEM?J&_wjB|
z8`cdd!xyj{oCOW-5Z+<boB?&(;0b@bwoke&CGL*6NxO8XI^xVRgLFRapQhuxCuREi
ztUt?Wm4vo1_<Xc}Iv2#v3Saaj3xV5*`G>8^1eZ6p#NA*44&I@=dcKaO&+_|xfP%oj
zamOV9Lzxo1?Q8q7s7c)BVD~;Wp_kLnT@fMlacKN019qhVwjL@o44;6U8)dK2H?~GN
zQ>-wn2P^C;igOJrt++#=Tv=h-U2Sb}V3kWWsk9j`lL#IA=XdCfjf{D;%$88ce%P-0
z;<@2d<q0n0Z`UNdLCPUO#8sCCHnUjL6L6g5dBDnHH7#sFX<?_<@BAT~?{+Y8^D1vr
zHI$YncCaMU3Kg=XdhPWE4Uk)bk!+PFmML&d(}10)3N)Oai|ga}J#^H!SaNorb$Jm8
ziyejY2GU62icopbJVf_R^L93O)AaS93%il1H5+Gy?ZCo&_siF*n(8_-QqCIg4=>!O
zhwj8D{dQpFLea=sL@!Wo*AfuLc1lG|OI1)45fJA?vZ8<nFT25r4iAna1gA_XEJ8>L
zN`@r1^5D4&R2+W$O7YIds8>+)fu~xmV1Y0d9+(xWq{NByJ!><QO03MC_MhESBp427
z;(L)sAvT?~Pt4byRK)u2XBIqeib>ZbLK4H`-w-_j&b+Z84=tr;v5v+wjdCGB8+Xf5
zBvwO(Jt2#e@c0y=Ed~wn7)AKBUX@tb15dhcTV&Y;wa&6IiL@*}NPIZ2S1;sQ&a@Oe
zwk0;lLT^c`muc$peyF!FAAk^L?9BI-)Q`H&^HpLxHicG;7!3W_Y_Gl^XFLN>z1C3U
zrc)woUh^x$HG|xohcy>p*-^d@e@1uCp*hMJEABGdAofC@&GeQH=F_I@ge1QpS-=p6
zsCiu_?3q0e^}+WS#hV(Ee44jLJOJvA<C9jer`9{PR-t(5P8)X#>q0UcOUN9$x*cl9
zE))JF!>5`IC5M{1C_$M8p-!ot3^hupyQ!19lUT<cfC9nMqA3(^FvMA(k1M(Tm+m--
zAYEEp$I*nGoT4hCu)g$2eKj8?pp~w~PB|>%$Ce~F(m5$ZRTO`nS8{<4S)yebGr>*h
zsp{2L4jdl65tE{<Ep><7YK}9JE*c*B%m5sfnXxW1PHLU|xc6LMMfsZ|kTa3OvvFSn
zh5!C)$?BXR^$mD~wr{U%3hPiqAx^yPHdPV|JV^9p!Ksm}<a48$wrl2irW(MDVhT8W
zQ9d!>!+(@m_FbqTnGn~M3lsZ914DY)(@Dmh%VJbSd)&iJe3QOmbg=?a$xSARXmz_>
zf@D;5DUK}{sPS24C#J}RhV#7zeUpPmfb$+@JI~cEoqEoA7_;K~A$XZROh_Lg^X->~
zcHF=Mh7=3lO0rJk-o2OE(IT^W$J+cSRr&397}Obwq;e)SjTsa*SqEud*C1)DD#B7;
z;UOitKZ8+ZY@!}5_$1MmZwgw<sbjLA&^X?7gL<AZU@*87BUf~hJq4ZRQtB@N$E92;
z8Fn?t<ovn<hdY&EFxoOs1LbryVV2$OHu5pF|4hc59%PxU&c$JLH+(_)=${My%KbY1
z@YybSUd^~l72^^i>K|q%df4Q&_-;`Y5<`lbl1lG`;bt!ktz=6%o2@%u^C@}bo3aAv
zQ^|DaC<#$A$VJ7RV$4V;<Iu#}2l@tn%Xf~I7=s#sDz6DCxwpz5T&T-}+ZuBp1z=0{
zq?EgHp4{*!nmi^2w@)aJ!yjnH@OEQD&jrcJF=&E(naZaQ96wNYdAxb#XoO7Cvpgv(
zyH;?-9%QuTSj8)|SO&dTD%v5*B_q~3Mc#OG?lPaACW_*Wpt;#e!j@B|k!k!WYF&G}
zN<G%PLN4){qtLR<qTZ;==b&0lRh_g^O2d>aLq8KsnXMa|--YhvG`1rYX%{t(x>HT5
zd9)>=6a+l!g|A6&&0VAL0Wduw)TAO+ldidK<*jNNiN9c5)305Bxc_Z^L`TYG>NCF?
z5iRa4i?As~s|^-yD$(MKa#MEJF<DDV^}J$dw0zN()(C$o0AL%fDOPiLk%b?cT$sL|
zez${%Z;>B|xDDvl!P9Pl6B@!;Ch_+TTotW50eRTr3oD0JjV{Mjk;)ltTA@w2Nzp-V
ztZ4O|@X6w;v8xIJ)|I$OlBVxf7o~A>X`aUOtV@~J;N07O<Qbvclc!I+_#I}$gQl_H
zJYqB3k)ImveKhMUej}M=lzdMpSv0asei{d!<Jg@WC`VzJe&Zf9Tl%klidpuh<=wQ+
ze0FK>X1uvC{UlH6drpYEwll_US7V$0;4WvI{%m(QjU2IZ@<Ol%pH5TqE_9ZetEfM)
z&fIkFtfEk)OpGaG^l0JLr4M5ZuF4>F=NZinf^K0r(pnla-Uto5x`<9ld*#&5x}-1@
z>Ezl_(b#@C$|N_gQt&l}I6b1U`1vQP*|rP4p!}*YtB&!kB`>S4dJkPlDbj+pn;tAE
z`*^#Sly?Hzv$1%vGkRpgpHIj5bx1MzLHz4kx>ZO4pohT9UQvMqytCgUonA6mtiK-k
zPz>ku^3D18%Hm@kdC&lUo7nQczGKIrR{njt()gQF#5tPtVbZOWsq<8b(QM7nZ{ZG*
zE%#II!6!uLXLJfLJGN0o>HStM4oa3r1Oi7FI8=uI5OE(OV>$3){VoFBhYQVlBAv$G
zWMVq_{Y1RsDY;TiaVjB|x$bPA(334}s;I_$X(TUAw3&?OS$ZOZ9^Z^(#j0<r-p@+z
z7<VX%!$+yiq5IJaGSR%7M#tLaQY&_O>x=|YjZ8J=fz7}1K=c&lo2b~g4z}pZgR)HT
z@z_xVy^x`R<VAMKu7CYK^{>J~Y${B20i|M}SOFSc8=cBxK2a$;Xxgg%^qebaSHLEu
ze0BN!Z1&yUcvQUa^o@ZVLD^K0U-cqbjxY!HX^Q}p;&+jK6%?=pw4J8if;$51F32@u
z?k3o|!@dD}FW@>_6a9K#Y`+DZiPU8SEyCZ+R~09oZ*x}aExj*+cdk+w*eK|UI?%3v
z;~hjT+9@Vz=~Cyvj=A=S+}#}t<rk=%i|;?-7a3Xq1AdW_^}hr}|2xrbvBCUHtli;b
zh@w=WGhncAuY;V?4(vqXMTs`?l87$ieY{Hdd=>C=)0+pN>GpHC(j+dBC;j2mbf?d~
z;j_WZW3%5OcGUxg-f@x7=iWol`VBr2GUtXbz!7?VYWcJ*YjO19y%W_UemH*TyQ|x_
zM+900@B>bxU$f^Gv<(iU-<wQj)4@Ni9cK*xpiAXWL(Y-A4#+1Yi~P*b_uD$#u&ntb
z-F7a{$IEBiYkP!HO)QGid0LnK6oh~Rwg{+UntS{6)$-Z>PBHQ_!O}+gTYdB6b1w3?
zEXrsp=8xOV24qwLrq)GHG}K~M5K+(K29woO9z7l}DT+W$QdcRe0l1{P8V{x6i2V+@
z#*&l{dd-rRM{mS^8xMNged~E1V{7XZ!81wvr=+}B#Q^o-x(16jF!Z=mWQihfx?kUJ
znkUZX(E^RO;$YqPWKnjC57rYFQxvS^1!&irw6L6Lo{-|E6aD?Y({`Gz$m}Q$jZ-;p
zYMUe5=7&=c9YwF120z{9vz+JcDwcD7RGsrZoFXm88>(=epk-DdxQ&z;^eTrw6&W~)
z^9cx1ILJ#W=2bWsR`m8A{bg=S%+pJpeXtPQSeW!J)HLUwZYMM7o|igVZHSUsW7@;f
z<g42g-&^kqjX_?VAvrJxmnfeOLcbhXS;a6)jMA9|V5C+66Ddc@T@tX$vf1u3gYqSt
zinFH7D&JVHuxF~jaI`I{QDlY+`Jq~TCfs?*7=*~P>HuPkV!2}}R{FIzw`ooKzi_GE
zfLwrwqVA$hDG>N?#|AMgm0L0@&%>uz4s4DGs1~9W;XNCC@=Yr5&8Me^F8l~WoM#66
z1Qv6!!Y;~$r@=(FA>=^H;f#(gXY5h8ido6<c}g26_4kV72)u6XG!TOm&3l*h$W{b(
zSI!$D;PBG&bT+>p-|yWQbXnl`q<#e`dCP=GMB%3v9<5^(_Pne91v5FsI3`f}IHemC
z-7&<Quk|WAk2^L~UXu*ASc$%pr3wng&<572%tl$1>Q8y`GXdn4bap5G;owSzpfH}|
zZuC=1@lxw~;%1xLIiv33samS`_lhl_dv-i;7UDbo>B0tWAY@ekqn|m51+|nQvlv`H
zYV6s%dt6E!_y#&wtC;w>F!Y4;ltiY_@(O%wY7E#`l!jXJT+~J#4L?u-hIL@E2I*->
z;$}%GO(uzo3UuSp_v7luRVcPE#y1+HHbZ&D;qwX+Rjj7k4t_%k{WX*Ivs%{&TJLtz
zc4|V(5Hd^`vRC$ZjiufL17S+nT?4jj1B~lt{PW@Zodd~U?iJ=6%zXb{F<G!&7-s!Q
zmgadGDGa4kL<DaLXD5ER!sAKs)%GK00|;#sWEjUkH%N<|_Vz2*wQJ)zyt{lUsZclo
zzIiQrW##AwFO{XgKK#oEC2DE#W@eiay*HTU25KzPG^BGy#}1gM&?ZOkAhc-b)3~RM
z>t={W70J37bch;S0@kv#D(t{inYapVy;S}+RhBIwL84%(K@8+5PNhJJ965mDdM@Y?
z&aLmLEf^VX?7Jsht?Kg9t+~Gy0!1Upy?xmEo`o7q%{|6jx{E8`c4bnbP2vZr%b-hc
zh<Bt35wk=Vp4O{%HUqZPkc&_gZiL$Ty|9jv=~ZYn*DdG3qr}ZQZzIjC!x0f8JqGt_
zr(LZq%V<Wq3R%^g1GzedLM^v2G(0-@OruyN?yX;qpg@maDUAX%{+6eTHYX}=_?>#Q
z@67x$Tl!o+SWvK;`GXnDRAZ9&Og^X_sJbH^5$P1<Xha*vK`xM4sc|pwV<V3v6hsp{
zJz_4r`Fii&TgNs%7KcFC#l18+bVY=my>7Nwhc&r|7g9U{nkp-^d6A{jx83qHY{2p1
z7Tn3@8JC-@ie=b)c0%P+ili^Wi}A0%XQFlMM`}MhFGuTgTfr1e2wMj_<9BFRI<S9t
zmX}>lC{BcTYEwq+Q73cqgoM=g5387ypzG`*JNsNuuAG$KecY&m)Gf!9(xP$Nj)o^8
z;mC$dwFO*yQ)%}$Toc`ZQ)$fjq(jEQ;^~3HYsp{Iov5!aGP%Ey&&O<$S1}m-GU|I-
z6Z0^o%}-lMY_yDTr}ZV7_^`&KE)l>s9z_p<f~-qzE5v7b6H{6=vt}uTW*oM%O<H2u
zP^j+jDJje;V6$1)fY`D`9xIa`A;a8x=`7`O%$<h;{<$S>1dO;x4aS{bF@#q7&OQr5
z(Q}t*9lDr&As@j78S3>!Kdhs*mZCxj*Qb<ZVtOYt@d!7+5Q+|uI9ry}$f|_*|G0O<
z1o9#fil1Q+gwGq!L?J5v7(-655F%6w`R-;N$S9~5O<)i!)Zol5`>rtEYG;;^A*n+q
zA5!MaEou-tgc1+KiQ!Gb#k@!)Lw!^dbc|A&;jvLMlA6ak;H+4vpIdvo#QH@+GEj7u
zW!xL#FOGDK`AbddC~oN2Ve*O12m*>{u)wRQ6axtlf>eKw3-g9ZfwGyzDC_InKWiXw
z+TUK%KDr(gY+4Q-m=d(lpGIGhzU?xV$H6HBcER;Q$~;gL-|#Alm`NL*-#BhVplPAw
zJb)}#AwIl**ZMe@oU1ko+JIwfR3El{TG6(O`ThQw2MSz&DjmO_M=Cf&qNlP7V?8Qp
zx95%9&kw(XL)WAiiu7xBT7;B*X<ck43JE8@$Blm;eIo1%biWK_Nhc??6W!C2<P-FY
zUUMVv8Yr=Jy}B<6UB(W#CAiVu_t3?4Pb9s*jyhJll2&R-^lZMK3Ef8C&Wm1GVO%`W
zb}o5uf-l6Dq$!=NrfC`Z@%CY85`;+ijdo9>XLS2(K1#1dwyPt@MaRC?H}kNO4Aax6
z`~vuS*o*zA6V!i#G_n6%cl}=vy_ltqi>VX6n2n)}sfekuy@@FwAC$9;lc}LCl*dL7
z_Fp?}&X?M{a(sp%@drmhm?BzID26tWUf2@`2(okGg|i#ZsoUAZURBEm6Yzu*qlTT{
z2i3O+<Ne9g?H|+EeTrqH3{eIn&-Y${?Y$>-4q7W67=T5jv~>MRetb^!EI#iKikXt)
z_`=<^-urXGpb}gGcYt$Xd3Z5FFEOYO0AClMuovl9jQt4V&eK)afrhNl(}3CmPYd~9
zulKPy@;{u;8^2Hbhs$N@+xlIJpwqO-YP7q4*I%%tfs)h_%w`b{eZNgUyS|u(XgU<>
zAE&qYll2d*kn=#0|FdTSiO3<TMyth5_FY3ELNb~l5<<dZ3Vy=Di}gzo57C7x^tONY
zCTTOlKq@<0%10(6{*h3giq7ex!G5lulM;_Ca(FI>V0$#?@5}^a6kT-odUv=etWvRX
zPnXpBPPpIz9|y_=;w~Qt9svRx>2_@Cj%XBUr#JoXd6>lkI--Eyh@}MN`I1tGvLf%z
zgEiKyJaV|2;u{fDZvLQl-J@6WP-VEZm=eo7mzgK5?3PW!0<U?xTrb}I+08XX#TQ8+
zCqM73D#IrD<S3DSWMYMHB1Fx%65<<PP7w-euVyr|ZR7x2iC{^4SHrB6+?<dPrSj;C
z1@weUe4mH}%Qp&v@jB}QG7gKo(_DP4d2nslQcAjLNm=bpGdo~JhCLS+WX$?2?TW=D
zus8c&B?)%lE64I+hpMKY92cAyXUaNoRs55~2u6|cN~S5MmGqTFGlb<MGvi^ZMo2T|
z#B7EN0EM%WBTap3tS3QTHiSoI0=9%|7$1~P2t$B=*A8@GcYOIi$6<i_VDxE0NvTjH
zPm4^i4z^^xQ26jL=?~<-ErqNQiWD=EjCiYHy|kUhskC2@{knQ6+6d2g4?Uc{7hGE#
zP+2iN?-o?K(!m&1h>$#GtQ9nhQWOhAO>GHTVpNL8(Lf}FkJ$HjHt9s<L4LZ4{=?#`
zhTTGQOk-io_1G!`(XmBKNRxBeCaKmbY#0k)c!v{3TMChR7&Z04J8GzM0m(_#8}-Gd
zz4gGm`;vM)EE`X#tisxbfPz4gN}MqB-x7@`RQL#VHcgr(%X;0gRP`*W+)QnxoA&Jq
z4zq(P_qh=8v%+ym>*A6a&VUg*r13mkvnOsPnB(rfNe`OQ2!L#NyXXc%hOLwl<{!aU
z2R#-#>j8`_`jKOU`!VSWGk8%yDc|L>HOl!}G9<6Khj8z=#<PQt8gi}t^Oojo?cm>O
zd0Q{~spaiyrc;64R*~IODH|HuQTd9H`e>+R)%?BKRf~WMJ&^WHXwm1E#g?L(`M)Z>
zsJbTJ?#oo&CF<Mg{MGXct;-J;QtwGLm$bF^x?G<ZL`LOl@R7kYaCAkdZhw`sWr1ur
zuLzOWF)+EdoI$i_nQFBF9~RY`@&I{28`rKXuPf+YYs@nWiE3HY^%1gVz2+rKKFunM
zOihqQFRBm&P3|ftMNe4xJJLpau_RuU8^gH|2RL3Fo#9v20CM4L&KQUH2Z()WUOw$3
zVJf1Oq9*LLFB$@BBCn%%3%k@)+CbK@eDg@aF8wNVRd0%n*-Nqy<BU(~8(vT?DsBs-
zQux$Pf@3UO+|Zb9EizP)ZL!g|@DFVqtScIi-9j+DPEc!aMkDvKEUm@5>mvNMb0C*w
zJhd?yiX~{I^R*AQ_o&M{ACk+69k$QMFS)HFkEj-6b6cFIl(IkYVq0965V<|a9)|%B
zaFL#tW63d?G9r_ziyK5y`YCedl032>wa+)zgKvEj7!7(B1xEp%x8az;y3OI3!nWCQ
zOkzB4of8;^czqKXC%&FfFBi&)=9+p|gE-J36|(nzKTKsQVZftShfqWN7~dd^c@ctk
z^`ULQx<j@>1h}Fnf79xx1PYvXP5@(G3`oBpX>Rzw-VZ<RcYkE%>c8IVcBK$goG>`L
zzM|5@LP%>?CTBftB{5Q(0yybfw5knI*3aGis%d<*-J+uS5>gN^FXa0!G8&^;6!GB{
zH{W6reRL__c)J_38oq7NnPpK0_sS6V>v&S0s{4vhAC-R={5LMUQ}aH2*M0Tf^pivs
zp^=q%;sbI^{jHHo`XT%r)+AWzEeA<L>ov({6rR>JY~N|4gDSygTbhy5F3=VDGDvMg
zPrDcHa?>Hy!V=OSUfX*K!voc|Dnzk_#c8-2dkpdtim^#`{QW;u^NNz#->iP(9Pk7r
zDh-|<2>xxYuM9zPVqRNOrnjDMb84M3vBZciAP^gAJ=Ow36n-_bkcjT!Q90|=p+Nu$
ztDr*^hXqYuPH9>QC5K~bhO{awmR%2Ol4Wzwz@oBB^|gaF<nAIO3$s(r9sf+6aJA&s
z9KrcqG2J=zW$6%}<A_h(Y9@1wen)4U1h)XweU;TR|2^l_?;x7=X(1ABa&=0aDB&sI
zNhhzr;5?X6^@=XZb2Y*3Ol6J~i!a%k^)a~#cwHQ}2~|_(QTMHLN`&QsAjhN6`bdKt
z&R>NFkNkuX{s>5nlB#@YPXTpuA6=C}cBPpQS)7J+3!!a0I$4=<I<QK^HtBF(3$DG^
zcG`XMhO}xJUArddY8a}<{$y0X<v(Ai5bLNu>+t0sU}N@EB}aPY7A~N<99k3@95qZq
zTFyNg5aO+!H9XS+Q1~(M8oY`BR$=KC#*(aJN??i=hF=?E3}w;=?83biSFC)vToCLD
z8mJ~VUF!`1xbBVHfxViR2322Nguzymr7r17E!uV|a2$~aeHUQWm92vgLB+)_^n?|3
zr9SQ-fjewriwQ}v%SIPF*jL80<TdA-f+!Lg6=tTOre8PhF#E)S!&j%uFSXOC^07ii
zrq1V2^V7bN!^?UhS+!BggdTX_Ff8I1ovZy2_GD%FCfn)h;2(Vjn!JK*Qe3;HI9TU(
zUxA#Ncc;XpyqL8QZ1+PWA;@kHzyb8`^g@^DBEx#b?Y-TdgL=mu-srH3+ab^jp0uyM
zbJcn)oYUDwoWM|tz8o};rOj0G9U@tuCs6+J`JjAK)Qpu6wZSZhTkK>(e^yRKvWrti
zm%Rz)SubC2TudqF!U(n+?qKZdRYy}=+RQ?A%RZxJi;=3%ninncWOKJ3XTQ|cMcvv?
zsur*kI*zbBs#BlKzN6`9w0)A0j0#RRfhim8hEr2PmU0>>+R`ASbe9B(=vn~+c$lY6
zowURB-@Zu`lBTLLJi3d~hLSbR$4cF=*dt2JpP|K8SGIGJJYLaZy7Sv0g|2OL7WenM
zo;4xoyS0jzQv5V0x9ko4rKEBKyb3wL{)R0v0~?9(oZL;oGCbU9$~p;EIJ?Mh8HRTo
zDpPt+-mKCZGThx_qt7DiI$WGAl1&}YervI(YE#NyCUlij99ve*rgYUhVDa1!bA=k{
z|Fom0wb4ghllN=LwMv9!7Mx7oU;ecWveA<HH6MhkJl!z`0XK0#TE-Q303P<FaDIjW
zX*Meuf$+rePuX6=v8pTZYtzAoyFScjyF0fdW&^O2E%7@=+zfdB+H}B9YMP3K+<2eY
z@_@d39wOLV;>Ki#`+M`I?^<o&wZ%<cBJNezU=IIuQDL52tjiw`GEqpZ>~M&p#3+Y!
z9T=9Tnuj9-r{vLdu9($h0kIx~R~Uy!&c!bDd?y>Z_v_RrUhyAPFarzw|81N8zxQ=n
z{tw2KnTh$oQpnp{vhg?_usxsZXBkHYXrfU>fg=E2mmbIjfM~7ol3;usJ7NaI8G13B
z0gtzj<yjYw_OhFgig^9d7<g)_Dn^P&hs4W9Nm5p#R)hRs&nnRZ7Acj|NKL~$c>-sL
zkY%G-Leg5Sk*meihhnllbI1^RG2Ay4<mhqYS>0pSTJj<#BTZT5PE*QK!@2ilvLqsm
zQv@SYqP&SpGW{1z*4_HTSkhL~CV0#$g`BaX-jz!SRa|_3w7Nm9)l28JV3G9Vy%V%z
z9m5>&Dce^GB8_6>UZ3%-H3BIqc94XkN4U});<Z<hl%byol~fNaR3+F*<Y<g5pNDNT
zgb1EA087nF=7#U<*)h`VL9)Z-;Qoe>zVSrH28QSkhKGY#J7T`pOESI8XeY6?j`K>=
z)2;+CsD)aV2(OE&7*A3Zdw5?r(2V&z(zD^eKvs<Kc}rldmX={uFDjpHFXssiBH+tc
zkkB4dN-71F5QZkp$U`cf3Dq|{Wu#%enplMT&Y{JLDafjh@p7oWY7{vs)iD-9`PgJs
zYF(HV@^00xFb)t}(hQKg(jDprlQ70G3JzjJXg!$;19roC2ap!H-!^oBoZhgdBqdcV
z#2^PP3Gs&RHSpQkF`^-J731kc$Fr7tK=HFfo)SSya;FJE7}}5Vf=J*~4aqmGWJHs0
z9a3)nJxnGfw4v~UvVY)mgOu_}&4}u8gku+0A(*P>f4}LPUUc@rATG_f|GA&_C%w{O
z!%!|Zd9&+$ySv%0YGRI&$;NlTbxWuB^Zho<9t~-Je}DGt>*<zo?fiNi`<ADFse9cN
zS&TLBisoJM0IO0A5v_iJp;8VRZ4clgE*}Gp0i&K@4jmTFr6cM#Hg|AzYkhxrdOaC&
z^Y8svEoqvbydL{5-ak41_1hLI`?V>&81eIN@W&r#sCn}F^n2ZZn11;Nk#)Sh-PGUd
z{V{yK=Ml5;2whdr@Q=)N@GP_J2u}S%rlzb8u?=rrMbnDQvXKPSK6ut2NNZ!+P5g^*
zB(1sDxcs@+sOycx@CwG<imj961*h@7VWu;w(dp9d20pgw<L}sztv9Sb%8l>I$Y*>s
zA~|<oQU_-hW0{SP|I-s+)z{PWX20=O-}Pnc&6!4B$L7EO<@$KYDT_YaX8&^9l5Qh`
zIL^^QXj$L3RR~sPLOFN+8M4~9_SNj&vB@LrA3*|XAV~Clb0e$!>*4kiNqbWSg2b6p
zAB;pxAaPWTB%n<_HIBSY1WK6v6mHEJ+DZ(&Y6F^5gs7%&q1q(TmyI@*Y!8ZU3yG~n
zdpK|l5cJ7}^n{ie5b1*hsas1dXbKm=2JLKP;qM_4AY%du9>h%aQ*Mn$qN6w?*8Bvt
zLX_KokkC2J#bW`hZ!!x-Hyi;X3RQ<XkrII*0!Ht9&r8;`t@}l(tZVoT02-8vc*$7}
zJ2oRoQc29jMXzR_Ad<QvNGGV(ug1-*14&29+zp3A?(Nh0j2KreK=BVMgy>lM6FOWp
zDPTf^2fXtG2Q-f!xc2D_G=Ij0@QLHEi;w4qBMHDF9DPWyvrEe@?i!$=;Du9G6+sUs
zeiCE>j=WEl83-g3gG1mR{o_z+*Y+1syr|X*K+Uc+e$uly>N1l~XZA?1)_k?r{CVOz
zMqS`*$60*9Dy^tr>^&dVW!PKoQC#TDDob%%x1xW+X?c;#LpYLb8j<28NZ^RLuIUA|
z`qQMpP;EbJ`^>T72GH)f)q#4|yB$?FO4jc9c%p`kjE_~R79=k;)6YXY*v7#vc1L=4
zo`|4o&9NT7UAWGzEVr-kRuR`n*=n-jb^JB02$OJ&31JUqQ^0G++JLTImPN50BOnj_
zat<lOTGP$|TSX8&LwX5M3Btd!R*;=SeAQ_MNHEX%QQmSMI`9)(Ik1EUOWhhpNU%Ms
zY=bx0V{mK_7bF^^{d57rs$9A{V+cw?zy($qq4i&@ta76edV)Zd5l8_9Ud&66qHsPR
znDu(&uVMX(xWv&n+xxA>lIoB0Y~aH#1P<c8{XWG;#%loa0x*tfldFL2lU7{_8#EWR
zFPaV7msCL`5H0X&d}~*LA2vl%0HJ7;wwTpLJZf%Y5@5VyD5p;ZE)T2;Vj`sNQ(|*@
zE)z}Q2ZL9wj-U$U9PD-uK6DubDWs6|UQ$>e)=RJu_T;ktS!#lDSB4Its+|OWH;`Gf
zBpDxtVhJ$uh9xP8!-YZ%3Sz&AVJd2?a9PxT0jN&P;QPOyAQFSsLD@tQ`o~FQfVqeW
z+&(XL?+)>O%FhRPF3(&2joM&Uk)(}7f+!Ttsq6|=ATBw77HM1_94E@OT5|7f7s_b0
zSK#Vg)Lqo3moD<q!D?_vs!74z5bTla@@~{o;tb>>%tn^fI`#sSrL<)vY-snMu(U_b
z1cS)BHE)%5E|m<8yoPfoJloUk6+3I}uJex=90!vFU#0b3>U?c9ovq6_0Kgn6)|%s|
z<5d6?L;8GQo{PnN4YWb0X1H*ZoJH&iAWyD#CG2Rt7^(EjUIG-U38cWxh~OGP9xhH5
z_6c67-?BDC>!G6Z7x0=Lb0rwjIjtCx`SFXZoG+hX?BaDo*8wEGmW|_<H#|?`rE)%p
zYo0)``N0_V(*_mW%Hb*?Zf}rPGI7|!6P9JqClap~=)Y});7evQO>1h9+Awkt4E}h1
zWq=6*wNf#Wo<L1D23|Uwf}(!7n~e#m{h|5&Dsb9^t4QqJm7oD(&hZ?80j{3~#W8w6
z?z{J;zmsPAPdvdY{+=eZ6w>z3azcQn&zq3>$B@k1Cn27ge2edAD11!l;{X3^JF^t4
z#?5R6m@zL1f{)jS1N41y5Pv)|4zLU_)of87u0)=q!44a!c>4AoSsx6#ppq#{EX}h4
zDJ|hR2XIT;`Kt3LyA!N<NrT1j0hDRR(G8@;!4gcZJH{07g|pHGg`8kwM?taH0K%%s
zpXm}Id;?5^zeL2DErFDg05S%HABt-*xoB57XKIZ7+X{#qf#CNz5!Ec1wiolHmq)eK
zJC|y#t`JUP-ppVQ6T_wSj!W4im-5R8MD`JpWm<gB#TDe+zaNru^*@L&Mh>R`E581B
zv67MVzcWs!|8b(QIg$PsC+bfJD0%^bORA2kft}rwEkk%1V=doWtlg}kpFhz!LGoXj
zismB>TLNjM<d|bdwFRkeNK(G*JUw@Oy=r7ZvAKbW)jFX==urQbK|y#T3)-ILgXgZd
z(_p31!^e<#<q0hxAiwxG$nc(JL~{$DB0zlKHSjXeSk^ZSA0N%&M_Vrnm#M_v`F3%x
zd1|e)`!>0^x1m10-v(ux=A^%i>K_>XynQ!68?jUYiVN1%y*|71(B>`8^ANC-T&BbE
z{?T<^v&8G8GW%H%;3;qSWMPT=m?OmqqQ|`bFrqaGTI4(;mWd2SXM9wk5n0{?;4SVU
zL3x(%`=l4DQE;gZ3`!uws6vGx{kw?H5~eX6rg3i+Tm0gW=e!5CaiNc(mAzGE&kLdL
zC3((K#pLI6PhH}-=qsCvLC^9l-fTf$>J>4POoc4LD(fe->pZ?QPuA4A>5#A18XcrB
z8<J~F!)74t4Gi^0AIU8_u`UwqH7TihJx{*HZT$#DVqIr!AdbZt8AQZx`LOs;tz;k=
zNWr@cVPH{3pF7~<ivGZLV6@XhYOt4aFaZ{H^&3v;G$*uerU<x{KlrhCUhv?DrZJX5
z-y!aGC8w{I=;<ji?JY8!rNR*KqL5=y=A?Ujxc6${!ndm_Zh=%e-JMOXP}qWL>NuV`
zqdql}3BB`E=s1f!X^zQb*tV-H#j2c-G~bEyy^Tmr0Rt*s4M<F2#iJ6x%v6M@MN3}7
zn|a8?w&N0sk>*de@n9R?6+M^`*ezJFE)EOeyfUrOlu?TmDvJE}XcRarC|sG;{UYN)
zNn=JD$VF4GY!8>9{|W(E-0#vaKqprLq><zS(M*}3tU{10bKaa+(&_*lLwvgm9hT=l
zSnrnXd(N<;{cDYDp%>ZgiZsi|67zNBM3==4W`>{_wH8-21ghNkrR5y|nDqu<UBexn
z^lq$r`;Sw%avR{T!QE)5`wygHt7m}?GW&ekBIM1TRT{Dt__>|=ovjO28VcdI<<!Ks
z>YnNsexV=Vio;^f!r1;!D46X^C>=C)wc^Pb2^&J`uSI_qI1kGIM!Yg&Va!2pb2)+O
z)>4W)G-071vmfRT6Im1+_@0{4a8#JBZ&zkLZ9h35qy*hSz^{<o=xaBus0PANbkYS&
zka4?raSr!PR`=c?!)EdH89q_@jtGXb_!)uNyxVp%RsY$yQXGS@@`an$S=Ixhh|~n!
zUZB?PKiRa=pom5JUCAvUIl9<Ri+vNgKXv)T%D$se=a9`Akvq#jlfz_)88zE}k(S5d
zfhnYM7ykHVoHV)J73v#>#OpS&lW%JO*Q;4mBYtCrEp=qjeo{Fp@nE--pZ)K6z`5^Y
z)Eo)gx{M%C{&V;=l>l-}3{~$TC%YnCM2z?E@B!FFf{8oIK^+IwK$ySHD{C5pMzK-b
zZ+gFHzDQoorEL_!5A_2Bn(Z5W1?X*kqN<9gnEb*29MN8$)9L#RCCPVwgq2hK|6sX9
z9eBS-)-*<<LD$)TaW9x6f)UUnlc|#fSwe9X(%As(56@%<Y!SK@?OLEn$DGBSqn!pv
z1t4Pw+m%lUpln(kzt-SaE6LI{%E`ifh^lzdeh)vpqJ9^b{pyw-RPK)4RIT}%sSJGT
zAdOkf4OH;OVI-(2jI^qNS9zOkDx|>8&0wUU%=yZ}xNWow5<3#-ihlorvG`TVVcsuw
z#654W6_%({Di?ls#^DbfETD{IfgU+ZYX(3eHUVCau^n*#_&MgT^Bz*o@0=AfuZtDs
z=w4Iyk2lP>I98^Js5%Ye>6xh&b=^B?<j^{9JkbIMG<nr@IY8-oNjA!-2Lm3eX?y`<
z7_DXx6ev^V+5j7j98%1yr5q#U4s@8yyD=HSc^NIyBEjtlZ`6&F=1ASaH)rn131WD*
zzCkn((Cgv1vR6~W$j!x4fG@E*ZIaL1T{xJO*k5`VDpXrdBkG&)J77t-bif|G)J;AR
zDYJZ7cM>hn3oo2BuqhY}{}WP!X~f9&ybv8r!Vbs5VspgnUc47RXu2H&0F73LMkeC^
zQ+woRH(@vDW>`9JT%Uzypgq%nsTW+SMX&%k$_wYGsIaW)QJujcs!Em3+P$}giG!B4
z1_#ds^mtj7XO<r8KeW>{n>sJPLhaf*`IsG6xVdjr7Vg_Rm404plySbk5NM89X2$7%
zN|u+%xIRW5OYZl(wiBK?JXSbgE9-f6&gxrG)3x;|rGj@A6ukXq5qhg=(Aejd-`I^=
zZm)s;v)o-4XToVrPW-p}<K0;2(#`?xC$*=ztt9noVdTE$ipex8N%sfg(NgE<y6mH_
zRD{MmMzBw}gK9C&j_Ipfr#6(<SG;AM6Yr#qz<lQXWo5)MeaJ_x1!HIb)$XyGO~bvd
z+SZEAi%-L&op}YI@tHW>P2c0g&e(pzWOAg9Zq5Z!JS7dR>pGC%6h8Z~VxKG-u7?4L
zXw}Bhty9~E&+V$G_wVTj7OSi2y0BF<&x2hhmvfq@@+e?w*JSl?#`$FDm|*325}T!#
zkL3LZOpY0QeG8qVa(ZUZwa46hU*2&@iZm)T_$gkgH5i<z&-Vn?n<kP@M!u4+Gn+^(
zn!Oc><@_NG7ETpLEZv5VYu6NKjC}_7^3P1iNm17M(}7%<=<K9#+CJ;t@n@bjIrRN$
zjPiAoW0&CTWpVNPIg3>9CgNtU;t~zwhURbuRdtql4AVL#$Fhq8^1;1N#1I~83SLBl
zi;ss}QrD0kb7!LW8Jx*lWs`_}a*;kIVJR$Tw{Cj<_id<}XKwPk26YRb<h-Vj>xq>C
z_PA|o@<ernZp<pu=@@XxyZ=zc6@|R+;o&IJ(8!J0zU6lDEJ#@9<TP~JT<Gq?#S&6A
ze0#LgA8N`U@HUB=$N!{mIRC>64kzn>m+`;W{uc{t{~wT<f*oW+p`_Y9C=1s8(-g+|
z^yn5ECy3Y;&J{s6&cmy+4?l^k^!DKZodNJeGex7umxgL}_;6Vraz5V{Z}(P@H&_l#
zlN1JLk548D@7`12-ctFpNr(Z~@Y93u$8mqhk_+!#Xb$k4=~-Jz--QSKpc1^Ge8?YZ
zoLkr^7BCXsAUfQCG&6^A3GuuI=H1%JxaF;QWEc9JU+=HYN&Q<7qmT1rI}@c~LyD`X
z_s{J*WY;f-&>i=jz$UAYU6ia$fDd*@GZrkO0RhXZvR{QU_urq7WBUAGIt<v)y4wkI
z6D3IVvfFbtCat=p#Wt6C8EokxQ0HM*x$?}$B(9^tlKn@Q@P)an3AUPDpDJSu#R)S8
z2(Xp6IM1L`%E!HbX@8d8wts1V)zXOQaho@8dD7VDHWt44hqfQ9TW58w=h^d{Ntai~
z!pi2AuuctX!Fl}yEJsLHL2na@6f6_xWRcSLBf6pk7<H?X7G3-62@7gOi_kR8B*Bac
z8`J5I5>0nIBCSjzJF05fYibS{>yk90;wVaTG5|75YzYxjA$FpF?!-^tsBeiPepX-N
zOix0;!uh!#wh$(7MM3`r9hJjHE*JUtGgnx#Jyej~F=v`f6~273Zw{8;!JnJUN<KR=
zp*bT%Xr#<4P>=)^JyZ}i84W$bs~kiQTu_OR!f_JH%jvX2Zjzx8lbDq8Fqw*?r|-E@
zatkN@MmtBAXM{rW!h_EWnQXe`3eS&1)wh+UN0i%dn2Gp>lGlpE;a7i%M-X8NB#8+z
z3D4=~f`02wq02I;T$ld#KtSk@*ou`8vmk&p$8)eCDe`U?B9BVsxxjj}(a%^d#1<+x
z#ibwC=j{behg}>nO<o#B2CTFhc3G1aS5QQmDmA=VVu!wRWx@Rg%ZeNm5Z6T3%my0m
zcfQE=3JBz9L3K7h2(8Wsx$Zjo$+}n7uc-O~sA;$K#nIvCWlo`NyHvuMyCAp<^PPp$
zDPKgD9)j}s4(TnPvP33Bg%z=&`;6fYOM2{|+3@|ikmJ?$ig+nT+J5_He*}1b$pw+{
zf5g#m{KfRHc;a;uUlQqorMF*{_>aXFAE}+h!{%KbYcAuBZ?cL8D00&?4F<MbcD~VE
zh$>#uEfe~e+q@c$;s}y(7e6?ZrV7Xh1Pb!%xyQ_3=qq>H*bC^I%sHoGn)g3?(1+h}
zj4xrB_)7xZ)0U|Rl=m)}3(+2bi2aSgb9S~fy3J2#<5?_MA1r~7wu_U}Ds>i=RLNh=
zco~e;9h#>8CT4_7&<ei#bjQopegG0AjUDo)v5Ph2FAud(-_Z<N{*4&2p-sQHISnD_
zpS253Jc{LRL_3Un3s{GXGO0kxuMyjWMZsmBn5N*<lUF4E_gTav$A|h>lV@3fv2I@u
zpsr{uW_>pC(=#28g7;uRp6k7}_FT*;n6<wiD|12iF(ep)4StgXDb_G=zRoXlLFQ1%
zH=q~Pypn&Y*z<o`rs{l-a(<y7-D$4_ldluzj<gI@Njg22*Y^xcKTRN$wgN7D#?P!5
zpIO?g&2umnBEzB&4c4uO2`IV>n-PyzDQhzlD|tT#4HI&*_m%{^)2CzB4aL-P$#%sY
zNNVE;_Vf+@KgQmvJ+o+Q!>qVsI~Ciu?Nn?hZ*1GPZM$OIwr$()?1S!u{`SE>nb%tL
z2dw$5F~_)X?u3cRHxNQPy(v-R)&S%j-_csxl_19nw_&yi7csk&f?uccJ`~A+-v7~5
zzqzE1r?&5E-SgXn{Zs-XmukYDxr;@h#FfT4z@|}}?{I3&CzgDnFVh%>N$*a2ak@Q3
z88Id>g%~7UZ089oM95ryKJIf2di+|rg&D74Rk#k|-9}c`HG0^l=3v9>FBo7dszHLB
z{o5gF=*MMh^ZF65n8HWyc@c^qq(r&!eAN3LGXAoqRt-eOP2_7Iiqbd;G3D$_ZBM13
zO)QB!4W;PpO644m!+6qwNu}iL<OI(*0Z`-nPn9tRrOV;1>WEjna(jj`*urr{%JazD
zD$iIj6E6S0Xp02}<WDS#<XcIJb#iZ_6P?qF6u^@>p(Ig=zr)bf8BMS=2~~`<2d=zn
z6$5oRP}&yeYrF#(Ne`kn#HiF>EE5z7$r{~1CJ4Rl(L_;<=)|NfQfEm2Mlxv1Y(~|q
zxz-C$#1wcP0+HZgRrs9I_mI2kXYrgdjiLolSd=)hNSGpJ%QO+;35#%MB&M2qC|_8~
zW*BRwyf5R7m&s@!Wh!jV^A9?%7AtN|vPxhcv>MqZKQFPWo#Ik~gq+{A_K@(f)OWy&
z6}NgNbilThiqxbuV^TOTx%o_q#xcnb_wYLn2BEK^n@>cBptNf3wgr&b@R#A)6K-ek
z{K6~mvjA+0hSxNB@GN2?dDy#}g+PSIu>Flx?~TjX8MU5u%`;)Y%?J?w+eN=iDcYTY
zkNB67z-K7b28jsx4`LxOpP}&V*yhkq)t_1#LCf7OP6@7qM<U^CE-6|c2L?>rVWWi4
z_sapJxzr_vaq4+>JUJOmu3>q!CzNuj>petGZlzUK&Ozi_-KT_O2WDs4#-jZT5V3N~
z*9yB1%<`Prhw`da4UQoOeM{HqSEbPhFMWb#$|TDB6%-Qg1qS_WM@ABdMy$=;qI#J~
zk6M<t*?0-0c70~P$5dy}<z+-Pm6+8feL<}C7I%iPI1L8t3QsmW^Oe^UqEgo4<z+Fj
zK44^PG*pb<4^1{3L=^FXzw@p<&Gv;WHXkgshx~Rb32j=#*nfKB_N<(^C_F9MkUF)R
z+MAS8#`IG3Ru#b`f~x%XXo@(A(&CwNZ73EoCnFh{!p+^JsI9zQhpW+tG^Lz)BO!Q^
zN0T)ExRM|<p;VVG7LBd9yfHO$;*$+dp?-w3lo+hG9ASq$os1;Jq90T}6q`aWXmwlD
zElJ>yeLSOwNJuEW)T=zR+NBo$0@|2nj|9#Vy<!a(zYMjqQ-gk9l2{|C!3#lrgArg|
zdKXjL33P&iWj$WYCcjEI9c@=tXs`80b^3OdvmVN`1C3a2e|Oqw6c_L<7~;Tt-xAT(
zE+QFqur9oM6v68HTN`mcxft+r1rW#+#jd67`9yB70zc9X-0PcKYOMh$t@1Hhj#RqS
ztO=X`gW3t`gnx~sa+J-2VZOCCXJGk~NE*9~cWs>bBO0XP-OM_)(3-==?93%n=15MX
z^qSrUe`#ttU8JZ=qaA(+N(nBr`71;%{#<Bqt~t;WT+TjJC@LaDD8|<Ht}9oF$5K-U
zX;Lw(!X>|IYt>DLeYjobqK(oacn@!(UYT}*dB9KunNK%>BnL7Qn<DxxK6VD=`L(B^
zXgGL{VQ{wrc0ba?Vt9!$#J;c(q=t6*xe3&z!VUI{$JPl_ej}Rvtr@T=DlCON-Y`d4
zE1q0k;t~5{$<oeKX5U_kyoIOIc(?8i^3R|@h?I&8o86ZXA_>wvyAlqE7@ae4g0KtQ
z6Nr11;6OY$ErMel+!s#H>@i-ynG~~twJf-p^uXV&7F*X5UpI=?E_gsk9~D6|79{E4
zZeK}vM+oMG*N(Sy;ObXJwu<H0d{y!K8OgrIFggj+qSJnc2MjpttWBZbU<)l&ISPt`
z0a>X<uxcgt)`xu$H6oTS$iLHnx^jU>DbPz@T3FrnbDJg%w<o+dLp1Y%uC4PZxY)k@
z1R{|dJ13s|l(dU@{c^mu%CxtKvq3keYvuxV(da@m;A>)7fsC;Vn)XI$7RLU|@`nbA
zmj{XX5e=WO0Ll>kjB0?_Gn=X?lREUsjHH&U3`ynT0y?uNJ`gy3K;>{U&0tv541(Bh
z1sI5e#!xFJnsWexRyceiQzWQ<>OW6mfZjE<Jz|`qM~RjqFK3d<W=m@cs~TN<imp_|
zr_DDK62*c-!<c?;g&=SZ0kKpoH<);B11?6Z9>JJNggEX>3^RiNiUN6X*jY%>(O!hl
zVAgft{oCX5H&B-JXXAef^^ElYg$2jK`oEPd{}bwoe$<xls!Bhhe(6M@(>FNII>y6y
z12Eex0@tfo;*bhimm&1!T?#>xR6eG>_~h<42=~`8g<AfAEWA*h1#b(S*OQaG8<)3(
z9D~|HLye=u`w!?>&#g~y?d(Y}RBt3w|M~Oou-H^l?=dFcjuwR{wYSSlSq!`+`vt)b
zdfqMKFa$JVj64#aUyOv|zwW!3uxReXZNYW{&Z!%E*`4;AmzI4#op0}*G_$D=NZo$1
zZ)b0w56=^knGL1_{g=B>1n>W-jO_XEq%j#j?&fbEZ(7MTN%z@FKTBS3#@A;WKVsz{
zaO4k$OnC(5Z3EH3e_TepC4q#V2U<ktH~#RF@IYV;@pt_Z$n_|G`|c;%guqiQ1~8$R
z!ee1r0X<}&=&N{*`0qh=4AK~@ipib>s;5?dm{f0D=6w*Xxm~u23YkoD=dsm$kUy3q
zt{dS6>UeZ>1pR#W^U}i**wh%4B=X89GuEo}iCiW!Yaou{j?5oo$0^5yPBgq3--#}e
zSXmm{*A{!U2`sz6M7%`<2&ZQ=rlNS}fHPN22{gN`k@|3bmSR>vcl&cBXg@brqz>Vh
zep5icsH7)hlrI&zJD@S=@5hUGI5@`e;}yl(>SNpxgD2@Z^0I(v)p|_(Jj_qi*cEC$
zu8Pr`*enBF^MJP9C`!(rHNZ}FgTCt0DCQpV>D@_3!m$FG^kvwMG*Mqt22GdmItR^{
zBQ=yr2TeP1E;-L9XQky+5n`I0lxlSPT6c7Zr$}@4u(8?D8vIU?ty@LtlWvO7+{}w-
z)dxsAqx|AWF-I#m>Iq&iA<EHY9DoVyz)&>j%%cr;=|3J`Anm=pSb09<ExWthb_!2m
z$h!B1(YfnJuqGZS7sqT$72Fg|6->E3L=hK8f5$OcqEY0sDvcEcg>1qHug(ri&y=02
zzMCbEQ~-pIq&Fr?*cy<4pgim{(#0k82S4T$*=Rss>ralBm*;e5^EtNa5ScyV-h?sC
z$nE@%xIK{7B?(imb+~hQDIqf{4`o6`A=_<cHWn~75c+ypouqVOoLS0%oIIXzfPX_t
zq~SqT`l|QmDf$|Xyu-u*kEbgTgXvM#qBTyB{y)@6&4MMLuXEkLzc|46Ld5&Mcnv~d
z!a}rDbF{}p>Y_q()E+R42EQesj&9V>a@oI>BSg-Tqe~V)CF(>j0r5^Lps>*ULiCu+
zpHK->nftZ#lAmx>PUCJ3i+E&H-&iKOLb;~lY_a^Dn!tc{%YDS?Uo^T~!#xzu4H-Gn
z5=P%o>tCC}x*(i#*5C9W&kP?>e7&DnFg~u%=|?KJO<0r8Bck&gzq*`~okw70>Sv^F
z`Ap5|HL7l%UB+a=*)gHWt&wX<Ld2_Qf6VLQ9K5-b?NSFYaS7)K7pQ&?Xh?jW8j13{
z$+cL+BxXSewqZB^4OD@bMZ)It77IUpz|EqQB!RtHM^3jL`uzh@g0r8+C*P$nhP*(;
z-;K$^;8E14=BgvVTJ}M38Ah2K(7=$uPnS>FfTP<Qs9m#Dzsx5M=o93Om3VU*%CZu6
z)k0zv3t1AD+O$dIB{oP`3eYdKZNX=`)|uhS6aX9P-b<!TpIFSDHuJ@i?2R@IWa+S+
zcZ<}2AsuC_7HN14{w;PLl59b9NJJc9N+%twswA}{cc}F6%h*nzH53wxy!Pr1R9hDk
zMQFkboyC@%1E|^vF80m4-u*fnM%4E^8?i_{r6V62v6!3JR@xWL`GKt`G?&ad!`Uhq
z9TQc}S+uXY7R`P4CH`)-tISWj8&?$<JY_7HCkl;zbXu@%R<e~ZIx1ajRGgBub*aqD
zcL-D!m$-WUXK7XWao^IIv#<^DESp=0dQ>lNz`2w!TJ~p|vv^(KD?4EgHe~JKLh(z1
zC1_Qq)iN)uMn#Ilb?pQ4EK#liE2B`|C6<X6WIT1~W1>3rAb{|`t5!s2(g&rZ=jEsy
zn2AVvs&rlVeZTF9Q|qoS>q3~dvUIP>YI0DNU4_4saZv#mtgs?}RK~UD{F;nB%~XVo
zN}Eot!a=+3PJ1z0d)oG4P&bNW+xm!RyhV5su(Z!!qyCx0eTe8XkcLOg1N(>ZN;0TP
zbwUZbF;?&6@|WPNsZltvtI7A4sD|g%m$8~h+&qfCxCO_K#o05{R(WA6oWvLQI&#G|
zZ_cl?&zcX^e!>qexDvTyNId{ZatX-mG4C^2iLv{0Vh4N{@Sw5aiQZ`t=MUzAEpdLg
zNZxO@KmcN$@X&xztiNpuNH2%y%k=tkx7RBf<?_wbHoZT5Pa)bC?Dl<ypjfzH?+r~i
zHcSEuE)1A|F$}K#pX^~4=iHxa&A9C^W>etD4zo`*4IU7P(nKd)X_u9mI8xK}&=CIG
zrlZ%c#;BkQMiHa7e~4SaN!+X*d#+fhI{ouqimEt|HM=)X?4Go)?m^#sqwyIW>|h}J
zNu~rvq-Gdx=s=($&8EI+8kzzYssOE0#sFa=e<<{+Mz<zH>PBW@MdUd`s0Agp^ELEe
zKVHaK;?9BkJou(?jt!O4B*GugPt|s}=k)%T?Lx^B(`>JHF@`A<xA<`?IHu_!Gnz(Q
zE3?j(Y^RwJ-yh|TN)-MuiMe<{_PWs@d86I?nSZH9UZs8BTQCS~j#;!OHNO_bnqld|
z<x=$xJ-b~CvW!MLDO`D??lMq1+07|%H;(Ov&W1BWcYlOvgbjvguUg)lwpEmXS;gz(
z4_TT+&B?5i=O4wu3p5#rByTV~-m$~6{V0dZZ#B^O9FMaC6b_BhxM(jQ9p1`T#7{}`
zB;^Lk9eYyVFJplSH>k7l0qqAoqEuFtz_b;|ucg9d;HcX`HioInNvo*xKnKl&a<?_C
zZ``5HI&=@y4GgP0=)so1!oZ4WkaKu`d`RY)MNOe#&RXid$z7$KK5EcBpN5MODAU%I
zn-bjh_Ujmg={(xj0zxu4j#x`6JN9xIQR=%`e}oF|Ppv$>s80Dzc{mP`;88S(|GX*y
za?fcC12Us2jkT(sl%s)3<ESJ`wwvgxRFlbwKrqJ{WAZpQ0G=0)%IoSQZ>wmJ4ba0O
zO&x1}A?<a}N*49Tc*l<-R(tu&1CD5`wXC~zk^BSQgoGE_bw)B^j2iz?O6GcVh;32L
z29?L*7nzXdCNZO@>;_ei&*fJ8NGTK6a31HgAWU7k8d30xl9w7H(bOf^r?98X<YZU}
zP<ZP)1DRGA-8Pg1V9ow&!_JMD+8mPebU(ovD3?n?qo&WY7P*U*qeQbyb!UeIHuFcK
z#SkM=^!rx3%xCh0?!(P?-Oucldmm;&LvOG0(6%q(=Jyzy7Ciwqq?Cgd0V|f%`|Qgv
zAA2MN4PRY>wj(sbYMuRhAP^~@5IbhfN;&T3$E1SnB%L-yf&Xmsm2Vi``d8YO?vcHv
z7QX|vu$?A%>1W-j(hd&i4hBNf+@qL=DUE`%*IjFc)rp{mZ&hoD+}Qf}*|#kA1@8z}
z=whgsBwE%tDq=Yc1uul3lG=gFE9DQSsfYB!-_=MiPE0E5(U2j3aVNAr(W$U)1Hxwi
z=Fp8-rCik%l9xw)VOzA`h9~?+_doNq(ggkt4Z%~4c3SK<VAiM-T&%=#IMmgitcQS=
zP9$d>bxhf-V;s(27jy$xms2FXOd}LIRjeLtF7OW;=RvQXLi5dlwexG(OuNpPQ6O+3
zeRJBAqbeX~>cSJ=-4kyJGCqUPb<$tUIFF*tG;JU-_O|AC&t@~+h4<6i9Gf?^Ezz47
zA_x+7$n!ASC}(JCh{W4=2^#sqpx-TTjScoK^3Wo<VHWtWZ{lGgckDjqG3+L-3++$2
zv~e3D`f+W$nw50it^JiUU(X_N+VbfZgaS7;SR1Wwh4RZwNisaBY4t&JXtb8L-n|P0
zCQZVr;kZ^Zciwi+F-7}=_wKW8`)~Q4f#v_UBrvo7Z-UZE4H=sUR=BQXRaYBrJ>Aay
zE-+A_cnN@Cr&XM0ZHdUbBK+kyU*cPX$SM=ACak;dUahf$C$WMj`R(a)nB2Fw`@83x
zcanMCgxSw+`-d^GXYVbYt*7aO5m0Izz1J7(=fgxRy=&)g5qk;}UBolo_j!s95>C@E
zPY^6R*H-}QflXl#^bO-Fwk6%CLp%}ftG284-)JUoPly+6&la7p=hqyvY_7>Pwq4eb
zf^J^k9B-d|Qd65W6!*7p8@b(}_?|-K1muaA)0mtR-Rv)^AQfetZgC&xoUa!rLOJgV
z4EJaZd&356deG>e@=?=aVyG#Ma+ja@UR>>(cS))DYe3Xc9%Ka4zq+)=%|E<b_fMs{
z@`{Rix$=&wSx$o)kA({7^8gD^>0G6uj^iBr6wpVk1&*q5|FQ&BcbBJBK}mn9H*uOY
zSjSk7sm%SsuzC(c4@xxjEJXOueW>tc<%uZJmYn#*gyyouLd9>Wveel846xs5Vv)Ie
zjy~m^StfJzL7#7XZcFB{_tI(1Ow8_$aL=sM97*m-IUyzhuPL`kjw$(3wxbPTfh>Ux
zL~1!}e6{1>0)I_tYKpdHeAwb<lQC(?0u)$Y2xYuMDd0UoG5ULcQ_9@*-=l+2De9FD
zAoGH&>~{Fc?!7~Xaa<vUo%4drGXBeMlARS7)^){ksY(0s_g%5uV7hW4xb&hbx)Wdd
za{*dc!`go#s!Uj|#8CfQMIG+d>3d{fkrKy)Xm^z6Xh(M@!8lTBaMDr$T;Rd8^pOwc
zW!!yub)u`#({Q3ZXu>&N=ZVEU_euf0*uHEpkN95&?fV<9{*61-K#%37vKAQF6Ilw2
zTZ2Z2j#Tvk0kxvKlHsyN4F~Sa$1a<4L)~FTiV5f#-r|l(k#7lp2ovK5yFw21BhcBo
zV(gK?Tnlyx?-t*-;BQ19WMOPD%Wj-|{<fB1LG;rk=9N7@Ofs|`MVO&nC4tMknDW#j
zyth}9?C)AS`R-f`Wlg3qJJ&#BAJqgv6)y-K={|9E$H07mO*k~*cZYe3C)V6_`dXS9
z$uw$zV5O5S8@Cr-sT3<bUZuaLRndTy+&y~j+T;Oj-F^n)OSa{GjvNInO7F=D3~5Rv
z$EmHF@-<dh7nKPl$CoS5>!OK!$BT4|uEQF;o`yy&B}DKw)eVLU8cBknvj$V;%%#v$
zlWs@`NvcpB1i6LZuDOVv_{)uh9}oqli*O`5eBG?e3YpSWWe-CK6w6e{4oe4~=SOjp
zN^O!a#Lp8}RVhbd26&Dg_HF4PH`zC&M4REC<EyRqy*qF?;hy6cVF!3hoIvcR-8QW@
zo>A!9e-o1oS#kVRQy1$@`cOrqa@dk!#-&k7ULlO-vvne6HJKA5wlb<_$#1BdR)I%?
zCKR+OtX}*R`daAvmmgTmI}sO1f$9`j-0Hmk)eF1~wGD#U;L8`Z`-No4`g&p(w7asU
z3dI{!l%IZAUu%Wyb3Gq%O(vUyfL^+C=64YevbTH+HFoJMuHss3-O=V8LiJt}(^l%^
z5q*&k!?CgqxDQ`x5}l_xV)x!-8KsLCeMIs0zPKcYR?k^<xDp4Cllr>xs4^j!y-cvr
zXk|Z3N3k&#@9}r+Wz*`ArO(7O-F+wR$!hfVg@DBo;fzBRY9hH%+AyVzY~CM6@XQ`p
zj<7BSGSg?rt+4Zd$!mIjCKO9Sz)DS|Q1++ZHH)`g8L2xCOgL3G)pDT1{t_^*g$`nn
zDYj_Z8w!13<l4~pB#}~fS}u*AzIL1If+R?6T9>~o2;$lHJk+<uEn1ViJ7Bk)lHpWQ
z%er0H(@lBYxc$Y~?YQ?YZ;pi2qsqXJq3ebBe8;GhT)UX1NVcjc=UScc)#rNNfYAPB
zQ6Ik1D0`_Be~+BhMdY3^d!@)(lYB4HTX10rM=iGl$Jk~+X0xr;z_LQTFM4m9Z%*H^
zvepTPHP;*0_~6}KUW8Tg?uBKgAQr+;-0jf~OvN~Lu9B?{iODQwt6y>@JX&3I+SqS#
z=DF7#3zp5p)O3QRe#YkVuD~<c-=*oojHs3l<b@qcaqzH{rL*BwQ|F{MPWTW?PbL_f
zE9oI7_i7Q+r;)d~35ndNor2ZHa%;YAl&cXkCcTPl@-xh)KvNvP_7!r-&am|^+5U{v
z5P_S6<NSf5hj9ECsUFAw<``ty-fO&18e{!7;DDP*kw&d4;M_VhK`eXNQ8qKzqczxA
zRO;kFZKDgO6qv(i5Z}fAbURTaL!Wrd0|Fuzqm^j5eTQI8>b~n}k88k(lf3*fMPqdM
z#3Vcl#ww2m7kASVjq%9l7upc<II+=H|2Wb_qi}Bgfb=DDcHEZpC_||TjdxqUpy(9_
zz8j^T4C^{n{MSPy0zpKDMyv4J295S8o0`>g-$e#Us47XaFI1dnJ!)3}F!(Q}a{L~~
zaHX(}T}yQri~Af=fAfqP&3X@NMHhdQa9MKdoGbNv<v4f6<@&7qM7H%3MdXk<Eqz6k
zKf9jRFE;}5)?y0x%SSp)^c%Q>@H~y<OE(V-EfqM@iL?RtRvX;G4V4<)^cyGUAT<@s
zpSgXF0AT64#&gzBKPo`WdpuSm0#d?q9LJwH*Nu$j_{NspNl*#y9Cf|Ed_r8hj6@;H
zfxjo@C;Ha3{X4$7SWV~++91Ssz6~i#XJHd;7vHAstrhOsc0Z$2k!T<$WJxO%L;WS<
zp$n5|rfb^yUhFuk-%YrLutMfyuzyYo+-Mbd<lMzK3{HwS%LdrxjLv6SIoij5$CEqD
zCLX?sBDH#D>d6Ax2K7B+MjEP=8xwpxYT$bCFePCBp(B(MBO3o$ITJ;C<(Ip94Qg4?
zq4l_Nq8_}t^qWmq*QBbhI}j#GHmcvd?@P8(1@F@Bmq$2~FSjsc$SvV#LiK)G=9|x^
zIWnur`5g@2nDbHKX|FlnSKX<RoI^|#({gqc9>+nW$9vW@<3OS>rhUtZywKI-5U3ci
zfFY<=qKh}@6L|ba>MZd9MFBMp3p3Jx&;3rg{a{~PZI;0F^3VWUB~qH8^H&9ReaTl9
z?pfzNgQn|nka60E<!UwY*`Sp|*F)f~QNC97*Xu_jZ6#-8H}x!%h`;7d!h&K-w0MfS
z#-ziTbN6JZj=5aD&oa?S@#*pTgl)2=iJy{zss;Cao~pz$iDdQa(Mn+~5|dD@@W_|&
z!BZSU;#eG>)r0NwYJ|eqOtmk2!bq^f*^;$!hVx|tDp@5@%al+Hd)OC$xfpcoD<9H0
z9}ubm$<2Sud;blu&BXq{xjp_<&&B?+&7G@0;9;Z>?jd9R#vEk=x36bt#t#Xqb%t|}
z-Q>C0mW!k2O{X!_P*@l6Wtl0kW>q)xQm#Z_w}0atJ?+1`Y4dpYcen4uwl#PEWCZGr
zeLyE>(O}~ErH@#=d_B*HTcng+4Wg|fDcpv$>G8($4go61<GbaX2mJ&u3lbwBf<9w>
zk8u{W%P=o5=8e8iGkdEtWXC^^aN@r9egAmPH5!pjxZ;ZKGgNZCylL|@3lA+o9yNb_
z-Top-^!vfQPJw;z^;x%s?&lhPku#-`;58<0{#v&r?+`fvgV^ssx;bz{aX}TiC2+r6
zWpMfKlsN>`w;);r9WeyGp@VjGcrprei9LEE5s);(U0N1_Do8*<3(ppE;?P0bUS!$k
zq03$D@ZeYaTEn*3!~Wim+qYhk`^E1!)}LCdAZdCi6V)2MMK*dra`6`{@^jR{TO%Ma
zFnT40PlC7}4M{=(ng@!XEx+g>_peC$X*FbPgXf#AN*{))hiQ8(#C>exg{p*#6kX+D
zTs$}9GI+@oS(V+=oSk;#l&fFilr-cAMKxXdQ-*t#is^Cl!a2ua_rqJA&h+dXqC1ax
zs(u)~aQ4y}D>-6M%)-^Im2sfoUnfmndqgLv6zwcXa)h&~h%0eCq?n5}q~Y{NhQ_vG
z4qVZst}?u%iYthmLTz`@CHQ`F^8HboxXRNzwOF+V?{aoJcKL(KqIx1S<Dk<v0jBL9
zIX8sfeRir2ldfvV<ux@%Fn7o0+)DbfFBBiEOeD@g*vmEXTR>Q4qs1Vrx{9nWr>?~|
zqn+~8`*f|1r5E@DYlBULC=h@kMe~D|9g%H-{}gT=K>hoz%sPX1R6+j~!+Q!rEKQ6+
ze@90IBRtJ*is1!IKTjV^nCeVpbpZ9wVub<NDAg28-*XZlri_mFm|I}H&RLzMxr^<P
zr_o*YD!KS87*!uZr7VK#4z4LvesD*Y4!~>zV1Aw!H$N?GjdZ7bC4MAZs(?L8QDL{D
zv+wZ!dKXf@Y`ifa|Lauvk|LpNI0v!d?JI$IXm2=D&To1sYVMNibQ!ysL2LBM*L=P<
zl=~Hu_NYnd8db;UUi+_5C~SVRw$Pm=8v>l<vTy^+_3tGtCSgPjb!iDngH@piPLmwc
zfQ|^tT$6X%xj4@mRyO-M``r}p!iU}`fVyFW8l+W^?m+6#{&`dBtF)I(Up~s>0WKv$
z<WaOMo(w{CeBZaWo9sCLNAm1frXCXRwY8gg;MuSJz@p2b-iVp<%Jn23b7ew%<f827
zBB(=(t4_7Vs^Cj!5S)AdFnUI&qgESzHT+vyP}P8j5fx1q%VzUjH|De?ecq=;3zf?J
zeMoBozsHi_)EruaaHmjt#dhkkqk;Mq#F^pPYBSIM{-$90d|=R4FD(CZ;SR_a#07%R
z4MR$i{>}XE{Ec_^iwvwSNcuMFcNxd%E9s$jRCEjI20jYEMbL#p2aP8lVGw$i>*4Y2
z6!;*?Uk%XHkg)1GgF)BsYmjs&7Sg>ra~#qdI68<C5)a!(k@||mU+&+(V(yF3Qi>y>
zCwjy8dh$x@P@~D?SR=62<SQefVP4&|bDN>0*1+ZIAjsv$i+?GY=eaL&LW?_8Dq3XT
z+o7&Gy(xI0Bi>$DF$CxL#I9f)M*0>i?UoE+{3do<zy`-88icx7d}?gl;22~YLLG7S
zQ$9ZKUPLhRDZZDGl!MBt0Z6E4Sw8+7>78rZ(wS-0lh`sjEJB)NNw8d<NE8I}$Zw?3
zl73EySWkHybt7YiPBv<egL~!X{Ug_sN&2Xq%Is!(MAn2T7vWu{k>D_f5;ZZ`gV*iv
zd<M7!KByhF?7xe<t8Lj~-RL43^k|}k3U3t)|5egVXY*wDH3KKm*r|S;RFqp&Q8HZt
zoFOH`m8W6m23-b&!WKK0xu>a+_G}H-woZZs@~U#N<&53*aA&p=Tp_5jUxMGDFqaKw
zMo~s6$5fM=))D9^hp0h4u`!7%g8eoYE@gC86^fH7L7`+<Qq0lbuvC4Q?^wx5v9vt`
zYhX6XVdS;<h%y<vSn^T4kaRcSX&)V&k-wH_J!J8Zrr`)}vNru(a}H<NdtuET%|Qu@
z%rjaCI;Z+rn~Bn@*rU<Ugzk>QsiDM8MXk`1VK81>a;^)@7J7V)CmJ?FN%KM>;br58
zI>XXH4S9l;BJ?#OOO=Iut`xFw>M1{&u}l;UrSlxruNms1OQ*F{{yvpgsAm22C+J#I
zrHXaU7*V~XTDJXd|Cf-nO)PwLSdImva^LvmLGc<k(K)4E>1BFXpwC>_Ie%^L-M;w>
z_TQ<x7gPC!MOi$74_3rRctt`9@mM*A`5Z-av|N}&x`Vb~nNdec@oj;VSDQXv-mmM&
zy8JW$_zYCX+<(*I$2{=xy`7ltEfa3PaNvf6YXBL)nm(fVa);+X8*5p;a9=hM&PRg!
zSbU;sj;4Fm9dGU=nsc0~ow(2#gT?7t(ad|DV~^*r!GsmGg}<Pb-K@Nb?~j+hVAOoI
zIu^-)r#tHwZ;A~qs2#=&j!Ltuy}3XsB4FVN##ory<D+00Bnv=BqZ~<FOxsu~^}VTG
z&x}EfZNZgrY96*WIC`d#<|~sY7T<9-<@D<BBxz`r5)#5sA^SbZ)6lnJ?mB<Ja>cw$
z7VtWDr%i#7qpYd5%}TO6#;J@IQc$w1`;2W$UDxILHJ4Q^Rx>KD27W$ah0Y}6l~_xp
zHHQps3@`e|o}g6r7&61cLWtFkh=fbw0np)4X03Y>c`p;1Wxa=2Uq@x5Ri+XjBlAWU
zX--Bj*cNZ?EB6!W-5EH=$c;%FVUFLf56{Blt84_1I~AmkW$11Z3ION}6u4KRAhfqp
z;rZs>v4XmhY@5Bb{4xRBT-O;yD008w3%fi#GV>BuxcN$?b*Q>0<7P-2m=Ccq&=|Q8
zuM_9<-L%{x(TN)W2|MJ%{&nr$$Nh)q`oUdDSx!R{2D%xL{)?mQcJvT9bydDJ121w(
z?_k9X$a1J(^$V*2-WUzfHo1luDXFBRyaGCDf|KM>vhll|pCK~7_CPe-YqKN<9o~a>
zkwPMm1ryK4WSat4y~Lz;XUu_Hs&d6|r!$Ykhi$-qXMeo4j1y$qO?PJEibD18r27--
zx2%s?`8U|=KgsF;7SEXeuSuSf<9~z{UE+*eYqeYP>Dskf_-P<MqUm8Yk~#ZcA-SqL
z%7e<ZiHC2ki}-XWMwb%Za&yP*w`O4LaWiouNyvMKXO8T?<A=zFlq~m$pj&~{`GeI2
zFWZ3ZfQ0!7?B&)6Kp@EW-@G}ztam1k?VdyGib_`99^N|ixV+nz5bVJ8km=om!Mkkn
zs~4bl{nEA83b=Rbz}qb~bW0xk(+4SA?Bu&muy=mCdtKwhGa|@bGUG;MYgI(gYZ`Hl
z#;_)xG;Bl4ec%$&fIScfDjD&Rk@5i3-7zB_p*E?gBkydq_nT(giy{{9mTAr^&iY=Q
zX@2ctpuBA<>6LxDjj=zPdFGA#=nN|1RK-?~Ir3rO)E#N8O-&Ew4HC{1$BiUk>O#^c
zXQlagxu1@SR0BK6QUB%bJ{A`PRuCd6Cd!r(ru)m`#rb8THyI>Oa4-EfCaJ$>8-n$&
zNL(Y4+-#1+f=dD<Y|jMTq%n5q1ZQ%%Zau|r*sK|fQ{v?YNe?kUj=VluXu>3WYBq+O
zi7rT$bXLNL@pW?~Mf4~Wc?Tt=@}Mp---r|5vnhMl4}{zf$pvycsJL1Ejg;>e#u&!r
z2D$YJk^2M)DL=9i`_&$J+M`lZt}gwsOOG|#$R)@@K$s;Uxy2}@C1j0jEwXU??e)Sz
z87j~&ZXcxo#Q{#3sv|uM`1Du=p&*oZ0>4cJj?yY{&YTp?-iISWp5Iq<Ts?+mbk(fM
zS$)eNU3B`Xj=%<8N-YNH;_<G>^s9vsULbYn^+M`#kR5Rr@@6yDcmpjk{ZCrlae)Zk
z;bmqR9TT!Ano>=m^4Y?qD_KVlg97fjbtP^wAEpi49A`p4KPz}1!Z$IJ>4XwbCPSg1
zbrdPBtHiX)(W$#}QwtZ(WK>d`{-#74=7P&#K(Ic9@e#8P6;p;+5##-)B_$I&%IX`k
zW}!8b0M;skfPys~2i}u}9UcW^K<r*sqEG4mYbB)0j>Qm`8bBu0pjs$g(2=c+nFIrn
zXz=RDs;;5GP<;%MOx~ZOQz%<dx5}=Wl<#K|nNlx^-5zkOAxX0}CR?S_7lgHWvJYY^
zGP<3}&9Ud4VRWc8C101U*ljN<bEY~qf@=|eJZARBNw$tOBR8mF+|J0;czH$~jtPkK
zBTlA(mDyHKfCIodOY<;Qft?qfX4&CWXQN47X9}wpTgPv%I>zj5RQ-{PaiSZZ6XGGE
zYV8?B37OdkZoNc!TSeypA?fhL+(I?suMU{(`c>0NF3q&Xn8JCFy1eg%;n~WOqgqFK
zM=zc3a^nqTtTFxK9sL}k@xs2pgSsUr6{2EDs!X*dM<x@3X)`yev&hC66BC75xL#@C
zNoN;wssg?jW)k|PYRb2~NA;O2tVu~G_t3EilmM;@FV32OpOOry@NUx1yi@@+?+ajJ
zb*};SmXW>vdcMqyCPoZ!nNW*_Cx2plqT-p$+K_~QfT^($TBOhkX1IkZP}~rK-$@T`
zfAPcJ(dg0M_{9}Xe<qfEPD)-;=amV}fB44X2=5U@LQe~J{LdQl=sBtI@f_>Rd(9K8
zS=36$cm*<#d#H$;a&ZePC16e$JoDLXB*CsTI@RIRW~3iDk%i%u+}8!Ne8}gN91id6
z{&QWo>kA~w`Q>Yy8I?$LLWUrC_xJ_%PXJ?!^^1HLguYONxPPbuk@W)@skyC_{+}04
z9GGJDTq{#ipTgPD+T>nxJ&nwLvlwJ5e?X|IIKnyk?)xkpk}`65M~S59TXJq<<0(~j
ziSnt1d*iklW%b*#?F%aoFM72`iDx>}!Eg4q18DRXs6JmYJ8O2$kQrCg1)e?02XbSJ
zf=TalJ;|-i*%qM%-Gyi)U+9xI6D+eQ8KK`1_%jHd_qV9L|2p@?ljNC&5c_L|gCz-}
zg4&5fN~U#vlAX5g7E6x20v#YEp$#`}P3To+R?+&#UgKdS(yY>C>qm>_s~DCE>)Q~E
zNx%&G5e|>@2+TItHyc`~$(TUR317j5TB@=!{OV}&)5k)SP;#DpF$9*NgO&KFz9i7*
z9~7jDQw_N&`<60!(*|qxq=?D`f3HfMa@r6?;6>C`l4q*vcI-w1jN^dD9MUu1Pa8KW
zrDqBvJV`eBFuJ3b#qyMPDo{J3*!IFPdpUf~K0`)lMfGjJRMZ2y@>&691b;I-#Cb{I
zF4oGMib`)+7~Q$K#T-wN-VXwNTE?Tvlg)YE%z}~$F~==uncdD-Pf%7TG4y9FIdJy&
zREW~{XxQlP)kYd7?5P5NpKLe9xX19sL7ifNRoa}^f<oj$l1ld2_CDcrA_8H|&A;PI
z(goXhpbMqFH!<a>`U1R`V;H8K6D}4e3Kil2s}vhWEP?kcQ=<x|lG%!7c`e-gb}~Wh
zbJS9wL{T4fiU#hO_#GDMEA!^&_6%I_29~Edrp^~_QYSC{7x4>ht0BjK7?zr{=qfFa
zs4Wo2ifGJPb?HjDP4lfNm^4m?2~2jSR7{gw_@wC2xXt$Klete1P{riluJdwKfVbKy
z_KFaGf$$U+YPM}G4duC?Ub$~4Q++*PHm8k29H5*ktOdRYqg!pqn!OxEEVAfpk2T?l
zktnD;y8T=7D*7eMP6KvE9ftYyjZT-?`fcLCW)Zy(83C7gEt+H9Dm%6;wWWf?Wb8i7
zdYv(4pBz=46$C_t?<&iq{x6J*1*5J=Q~If>v~YJnj0LM}kS)M3MEKMTugmEv@&MMc
zEhtqo`@X$~czK_!?gHI`^^u_HSGvATXOt+(X$2hPLs!{uj>5CY6NUmq#neX>7X*U^
zvc{SEajwWXsg0thf;=S0BEHKBjuC>(1&&ed_Rmwj%>qZMXPeV4TbH7-TFuY)%vPsE
z?CyG=L#$9o*U!?Bkrki?^>mG+MEV=dqudH&Cfny7CKb$oH?fMM?6FVZb4jj?KNz3n
zdya`eSE>@dNk+K&#2-NVTf;ZTVg*l?f?v5WH8mmYz(@{LUN#rCqmkyU_DX3-|L=CA
zjqmj@mBCa=c1y4b$(quWL@2I+Zk0I-RA<(AH06WZvhajfOX5b+VVol|ckNSgecrTi
z>|>O;iX`@m)D5#|8+FR5prWq>$tsU;1qNa1qu#BRWco_3RV467P)+-epBF9Y%cYIv
zf2a~TE{!Ca`Bs{JfbFSdTgGN9NWQsqOGwZpn`-Z7l50aLErX9LAML?^SH@EArPo*~
z^jbC#Om(Y^!P>L83Y=)KuUWHq3hA;Kffkims}w`nD<42m!#*GtqQT*_t$-yQaHe|X
z?xjloU_gd!!S>`@63xS}QUr5`8QJCjCqtjsRi!0#7`e+7-4BWSGg?v_{X45xQuHy$
z`Di0DG*+-uLsghP5-v45b1m=L)5sa~CBYb=n7J>~<;>Y|a+^=Caokf`xDtrB>aXBg
z_~lQ9e}T<BGi>hrtAO)j=idRCf7az~`Cgp`=!xvPzyTASQ;|5#=}cqH*y=Z!y`iQ`
z>$v4sZe_QBtbnt^(=}wp%irVeXlPGI4GyKuC{+>^&N$~~gG#g(%p@;~mO!P3yVUbE
zx{vQH9|=7-K$5cN^rKq*8d0Pjo2Mwg(XqP<npbA`$g#P!vonGOl`)dg&$5~8$H1_X
z?luL-fybaZDQZMu9+8YLpiDnb2AIpW&W=WMg`QTFj&rM;=imp5-3HbWryCaEPMPb)
z8se0_uX0Z>DQ_G>{~S8Q8NRj@gV@ASLdRusR;80M^`4ZS#7w0Q?%QjK8#iHjt(E4+
z#T+d=8_&zAJI-ZFcbTurpDcS;Cyh3Wrb%?CD=<)}M7)DY0O1>ow3pD!gL2!6cjpLR
z?^~UGvZqr*RC#{?z-q9@e;C>|{1v8N&3o_-2FPL#<YarMs*&tX`;uSL4D;VWvBv?v
z|1Hh`1b2FT8v}DFZtnjVsKmhfKVGgcs{enWbOw$}LWO+X69c~q02{vz&C}A85;G9+
z!3;jw+x`1!GG08-Q-|{!ZU+;QRN+e;VVRB7ee3Mx@PELgSF0;S75E3OZm$anI;I~a
zQ37O9WC~M%@9!F44|Ob@U7uVlmxgw3UvO_I;9Y`&0SH^MJ9!~oHWW1sks&uN*04tz
zuG^!#!ymL!TqFPSDF<$&+;a1NJl*Zf9?MvuYQLrPp6SZbwez_&LdvQkvUU4x7TH;d
zr}e<@8z&yE<4u-q$~DeZh9?fi?@l*Q%dp%z4ACRdFHD7@V8ny}YqF!IOwe({AcX;K
zQeUyY-Q;!Z79ln4N#sRw#;Z?D13-@T6zbe7-i^Pgbp~7FZiXWaS8kL7W2Y?n^5Ro(
zP%T6?0YWT`@npDv)E+R@O~)@va{LKsL*|9gW(yU(iyF0IovTkEky61tF5qM!X_Ssx
zA>fQz+Zuy|Bhz^BV)27oM2#w}+6})`5a8n$gs#WM&x*inD$`c;?YanZlgvFrdTBpD
zY?rg7S52B%1J%Dv8cQ+Kgoz!LW<xYovBR8Dvhf$|1mn{r8jqz=6u2R<)(@bYPAlhZ
z9Xe&|PtcqNcGFsmf^z0D=d}e7hvzpD^3%oWX75Dv<$0$7C0abfIEOE;+ag*XJ&Fs1
z*4!e6ygV9_QQ^$fPZSaF%m%T#n=t=|AB;8)(r@J|+A-8evVcq3r4>sq9H1Jl&L%?q
z9IsQDdEL!YX+ABpCvk^B*2`QV$e*)H1|7TI*k>@eh`CgZ1E%ccp=mc`Lai1+A!U@3
z)!U2G@zHVs+|gcAvEA2rS2xPX8Y|Y;@q=r9vG;NoJ9^ZT;Z{;(j?QXWJeiR5?Mr=C
zZKTwcLkwUCnf@0Df6WnUUHwlB%vd=A_pRVahf!rLIDLH+wkPi8kid#De;S}~r2&9t
z`O>25$d6BqL_@a}9Pi6?t=Gx}Neog`I>TN8r5&|vO4UlCroxFp?P<$e9`g!LJI=(s
z)~U~O*<Zd3qxV^)A1XJaI!U#gZqJc7f*g*e&_Uo<C~+3l^tEfvSX-X2cdRwLhb#gd
zdG?drHmZxU%0pv7EM=kv3K{r^>7{}_NCkxMr_PWy76MGHOh>9X1x^l;Vg>z&<=1>2
zJV!kmc4Kb=wf`z(cVzn1GpW7ry%y;O>&*<wReESoZ|@@_W=ip&Q6nDOl+mpD=V+=V
zn<>YScvFD+m{+TtT`}VFA*W%VXQp3IGp;<FyA7`1v)fB1VXoWBd7IVQUW^`;l<Q?}
z8)2VFH#yt14kD1#AFD(=P}r}N<*r;uI7EKWD%dVMz$J7JF8Y;^0?JIbUAz!@4*yv7
zw}z4Ia$3C%vE@iww7piUW`p#Ap}LV?hGNHbm!>qNqqr$5H!J#kj#CX4Xri{(s!~>=
zyBEbfPw+4|u9$3y#Pw=dK5E-r2&3L|-wO`zWIE_plhAJaVubl3PCKETMC30OMx2b`
zHqvz9e$F_vyt)7|>vW|v9qYU=4h|{0HBZr_4jNNatVx+*<Z3}0wfm>{P_ZIvr>Uy#
z(51xkz^|B2>+nCJjx%WtQmnnEq^eZ+vM9^ek2j5qQ5G%jN3Y~@^%erM21lUU(jLot
zttQ#a=}s`SCHr1g+iD$k?R`gk+-cci_~ddw<=H26ItUf*7zJsh7{J0dlGLD|Hnu3?
zXVGGatO(qiz5V&ORn|`%uld9+|Cv7QKZFm67VXk&l)QZBee2GJbwDsRFUOcGeluPJ
z5wu7QyQ^Y%gy<J!Iz3W;c2q}z)V1WCLY=oo$^&p|DMTDoamJfM$wRhEm%4hJ*j0L`
zo;`V*f(poi60tA^_Tr47p#X{Yli;iQS$71ue;q<C^}*!WGUuA<R+3jyHfuqL5g1WS
zo@8HAjko$R5@mI(_F$pJVnw`Kjj|LNOqu*69=DjQsOtP%D9}HhP!xrMtYi+D%JNUo
z24psOuXxFMdob1eFhdXLtmeo+_t%owHw~g@&XC^M`o4**0;8ce7b02aR5<=v_hH5D
zeC@x6H^`IRhIh!5xCOA61GYms2m>;Zcx^mf${KV=<dx=hvN$9qhK1ckjzSycm)<o!
ziWc0<R3**CTuHPJl13!Sm7*3{viAJuqIejU-SW7K^L4TYGbr1EgH%J?GrOq<suEm+
zrb{L^4Z<pT<psZiu0x?{|J!-?<M{ahpJ$8=^#9v1$l_147&aTi4<PX3=eHsl$NnV_
z3fN$fc19C=+dng$Gv#0g_~$s|`p}A(sv?$3p6xd;(zYLQ!-3mI9xbs!&hc(_bM1P)
zAzm~}717so_f7-t+<Bm2=UHIKBuGs*aQ|ZPTD65suZCztIaZi6K4~xFvc4|*rzALo
z?dKSF*(p!B7{{NID>cliq|+wT0Mh<2#kXuUY<_hP`9{M#OVX^H<I}lYHlN9H+5M*Z
zeSZ6TRFkpw1JVYmDCYCtctMwhBBhBnnTR#?`R3@{_*l}UvC7Llis|l_*-Y3*4pl)2
z)v_1>7w&~rXED1>eW)P}1;U+bCEQt1>@a+bdH|k$;uXZ#rY!nYZ2K)*nE#TxRraEJ
zx#tp8FA`Qpn^bV7<$0yYeX(`{!PvW5ILj#Ui-a#0Sb%qkK=s4({0F@~*~bZ2=}?Zv
z7b|u?iCp)!J9@jw0Axh2Wi8`XjCSuz%q-woWiE6Slo@D4-mu0ES#=c#Dq>9XROK&i
z2yJ$7&u)k{tsNHHMmAXf7)s;03(t_<p=^A_zAr~ojcdrYH_PWI^U(SBNN=dae>ix4
zGzztpi&}8n-oxp-=-L3!MxTyp=nl=>1OgzP^t&huqbQz83$$>}K9Us$wachiqyP}P
zdLGvtd{Ovi3QCa&8z&o8nmq6<-J(+iW^Z7J*q8}#qrj5J#dQd7VmlIV*%D7iafdAt
z2&wXj2g*#r7lwSJXCK_S3pl2BNq*1BMpOK%7PCOa8V#iBnWB<u8^7$7tH@|wCaAbl
zXu6#Fn<I&{A$6Eiz2gy@`~fQFeD-Qi+0M%I?4nJLGcu<%2jXCgDT?w4ML2z$!d|Ca
zW<=cxhG#YQRsu-SB<f*1Poo-=uSQ+|Hy46ZKWiS)EKzEj>uZ~iBoPILTKq7dZ~uO+
zUGCpcnR}XNUD312gW>QVZ?2Gyxv117s+v&4R?M6**QDGfZ}_WOcL{r*|6rhz#UY7D
z&s2l={mZMn9sTOb<Za>TDd&+i%DBAd`_Y<)vIrH1|4E+0Kvs94IuS-jh`@_Gh}qyg
zQqe$)W^XrYuVbJ{#_F*2lR0#@1?QIrVN51PTjhjQ-`?p27}^OUx^YAGW6OiXn>gRP
zvg5QX0e#3!Gy8N?+MjI6YDlR)z8g+^uhc;>`%c+|SQIE*H{duf;W^F-Daxc=g|3`e
zqMt~hzywm2mV3Ic4$fQbB#Kyz;$_B>?-*8v;8Zd4Lz?($@3M!ik&bLv;bGvVcN!j1
zKho2`q$e0+>ygJMZER(YwyHo!!pK@Qyy4d6{zaoQL~F|}z)h|v8n)GUj#TyhcHl<W
z3J5K}7SOBCZj3HagBm8Lew&n==0YNu8W*{D?{tGAag5L~2`9|QSm5BQc(2r(^VLz9
z4!jm9$>bdRib1tyNEdP}ss4TJBu#%xL@<)h80xe$y@E^QzLeGtJQdkkG2tV#+7W*l
z0krM|Up#Fr&5sHdlFD0^zSP+BXY@gW=-26Z@exq5oF{~7G_x|)JbZJC)06B1Or`}a
z^Tp>oP_M5rsg>q;$x5A?%_@==n5jiD>*M9{%ER!HY6<C(?+i+baHTu{(n26oynNte
zT$h*(dH8RrtcbEOrej$O7A1tED3m}__kMIWYP6=$u0KUs{EYL>c94@h%WQ@3?`Gh5
z?;2;&=D|~YU@Sr=ufBj2NnHsh@rDVl&*_KiMW&z4OKs+W#B_Dq>L|nYvmj<6lgIM(
zyy08N$r*Ka-o`FTGT#EFbF@^{aEf7Jp0WtsQ`HktzaLA|yL;znDec?Q^gJX|tUB0r
z5wgmVlQ#B3AAV|N4{W1f$4qeOH<`RsUELOf{A6%w2${`c;>JrN8>MJFN6_3b(&C)1
zm8~nW7KIQN#a*7ZHGOyf7Eb*<5LWsjZB!dGr9<nD)nc9+@E&f?Dzk^T86FV_z-V8s
zTJy(JGD0!1RrXgo5iaNhtpc*M%wO)Ht_V4Q&6?`ZsBBOI5@)>P%HkiWL`QYmB8uZb
z%_4o1)rFbZ^{$3;Msf>)PkKShyn8hD+IUu&#eJ2%;vr+)(!Hus-rZw2^1g`xf=P(N
z#?dJs<>S-$L>1w7ReWF2IiF(0-vLQ5<F(<^yCEldG(O{%YH=Vojd{KedkVl;+%^RV
z<YRJ@@d~fN{Zg^sy>Vu0mMAa%c;XZnSe1Au(L1ywZ(I@RMxR-SAeG%zUL;F&mBt!c
z8CA|z*`se<7Q?mPudbZ^AOcd2=j-Q3hh}JTm}-%Id#TG5Xha@YG-m|<_^1WkwV}Y?
zXrIVK^iebPGvE929DFg22{w*O(I>x|kYYF~pP|kr84Qz$KOQo029WtJVAm{6;+htr
z&9<h_lv!+I_D|EB>;LJc9p^+M?{>Q_<3wTO`XC4F+XMvIwQ*rQixSp~>X!i*_~QKT
zy;$OhH26m0^inICt$&J%1n2>9<2W1ziLbrDvfg7+2P*%LC7jVYE0s-I<Cj|Qj8`jA
z%O|v)UB6N<+q23t@k}%q8lfzQxqwIJ%wgpuwa_|S$ROqHba0YW9SH4<ov6#>dONe0
zm#OvUxn8_@-ZYAfoWDiUdOR#?lAj6zd^4jue~Z~GkCE9|<IVOJ)v4hGs7m{vfh}q@
zqnvr47;_znU=$XyFa7qt{Ovv~+1W&Fy#}dDx0WYXg#PGlU=DPQ0K+4DY>|YV^h=h7
z3n;Ur9MqP&xrVNVs!F9d`An~g{tT7YBS|)8iuX`Y0xf!|;4&inea<M-kxbwTuUQ0%
z@V{jcR)+tZL715Sw+!+#g%W+<1h+M<_yQIkO&nea3JN6a84tYKn5sRyIB2D+z0viZ
zzfkjd%$^Nn|92yZEQ}@&MM8d?%<bjjc%UR_i}u<^*Cw|(4HKn4bL%_xKZF@o+qN>h
z^q~I`X1<mm*25giPTOXR*dcN)im!utZ$H2S^U(aWe-}z+c27tS@Zz5$IiAjhtOf9Z
zaoC4VZ+OYrM;=w@<Zka}B)EQkejc;oMp@G=)Eg|YsfQLq-ckeJs>S^oznT+EZnP&0
zfm+^0+LGW@mm@3%>m<CSs_dMrHt0+=uL}B+y8;oZV9D6x>;E6d-Z8khsO$EOZQHhO
z{A1g8a$?)YiG5<*ws~UP$%*Zo_jY$x*X^gPo~m7Yf86`Ss<p<RbIkGk`;^S%VK(_1
zeQ_GfqeMZ&S<Fq97-X7I6Q+a%khZU3f9Yxe6Fay#Zpv7iP+|M^NZ+Ct#OE66>%u*?
z&Ja3o&Ilaj+&Ii3ot4yRL%Z<e{LO7iPqx0w5Vr(5?%OL)Sk+Rl#`)+@JzpG_8iVUe
z#Xhhgp)$BLUA{-z$5XP(p%AN_8EsR(L5VVSVq7GpfegE*3W%MzKLssLSL9GI^oHO=
z_e35s8~ACTeNF_kk6#Fc<sYFm=40FZWQE8h5Xv;ijq1Gf%UYW^XRTZvlpe1<eqQ>2
zI<y(jub&C}ceg{rfPqHU6_sL?_I~qz-ag6DvST33aJfSJCn)^rGj?~sSoArkivN*h
z6(lh7X3BEfFkz}}g`!K(Tp9K@1JX)rbTbdKIAF2^XLnJRyhSe;(-HaV;&#9~)hGn6
zk8N(58PumE1B}cM4e0}|NMDLcw<v?DJ|p)L6}*08CW6vd898z%qY0YD%{l_VrCGGK
zwSQ)uvGLW3IdR$m-YKkfKVPNp?x*KV+HB;~JRh1w!=XWHMV0t>O)QtUYDaUqG{ZnW
z#|y6lk^$S{1?+KmUUy7h5b)_$WdJhRE~A18-pY1Iz8=U_?casJlrs>=w{^N8B`@3)
zMsp;PqtF#b!he8D_h;-b5<vk?K|xs@wl`oD28%XJhRPuA1V~l+bK;_&UL6+IXOV(L
zUvCRuWn6*YRX+Mp9q@@|?p^a#lar+lNb;ceU}E-`YOWIT=Ae&cgmh%g>p6*_Mzo7=
z*a&QMBiYQJep-pxUdw`OWH$$0Q9F0At#^5o?b?Ji74QH~LDxexT>q68O+QnmSFLKf
z+lqDM6-5)u{^91L%<k~!T;=nlvRr2T1$$=2MT>rOB_<b(oJ#9&B;bh^+{|$na`KhI
zf24iJ!WyB7E1Q@MCzeR{yWYu(L0R*E^U;s~M*QhsB{~gx5Q$Dl!+w72#HI}A06PKQ
z)Jhz9ph`X1!WKVLy3rjn1yQm@@sMcbB+`X!0yU~9(CkQp47E~>xXg1v2#$*Nk)o-@
zMjV+WUv{-mm$URy8<z1(zMcW@_o~7R`xO<`zhae97+7LPyeDr=EL@`*n9zvSTN@`e
zHKbD^Hz^X-RO})fxdZEm?7L7G=g~^2Ww(HoL8@f)+u7K($$1B+oAzgEq(OdU=+V<a
z!V7y+L@?ieqF7cEfP$cvooZ#rIpq0Rq|GkF6`GBs^9~*c@c9PC5&xN17@b;;TA*-O
z#z;qVZhvJ4*-S5{&^zExy7W&VeZ@wyjR&sjUh?1}qKgcBLCWXpfFny@+R~#QHEeIN
z#puBIWv8$WczXgNE3zavQgRlY07VA@B~cQLgH-lo0Wo4c4vbw^*bj_dOuOq_jvw_j
zB+Zv!Ed6=#7dp31g6hQkr9-xyFX=r-LJHIh-1eypZm$q}8saXVc;*#U6rq3EcoO+#
zLk=YT?~C$INW5Rwd7CH1PbnPan7k{K+b|45)B~e+kkk$SFL8Z~ZhM_B81L!P#{DFM
z_X3<!iwSq1IT?wHb`y%*8m=7G#36k*VqNRnJj}|4-h|vaoN4IVbLfE?%5aI?yDUbL
zF#<N86S!hNru*VmXHXl4+V?gpwt!2945@grOwSRpUvS|ckT`1EG?=n+Nz>HT7JjJP
zYSST@UofDmi4c9Q8Ndahi}wOWE%3Fw&!AyZz+=nNJ$F8gP&#~J(q9i{dx<u)>P+Y>
zWZkG=!gx?P86~5F2#|?9<}0Z4n0tMI3(B9{xs0RX#pnT7-EXFWK!qTrbnrEr?>|Bj
zI6+8wK}es(NKXE|NCB&r%2^wMJob<i#R5Fj8vwIu^<ykxe^YnFHY4dVFL(P<1pC|k
zSRe)fFSktn{a_{O*I#`zy`u>v#MMs<m>oekgJSduQc1Pfwdw@><@><;t*p_3EBvUs
zQ01C=NYlgRw9{xE-5(K$-HZNiN#dQi!uhlcWiMF?eOZl*=e@1l`fD@EhAYuv<qbYL
z(fpadhKctCWAn1my=C}*Qf_fP(%fq8(aG^QlPNIbsCZGe(REgFKl8>cs=uZ?XXUCo
zRZYxA7$K*H$k7oY4sUXzRe2{>-_|N|KEc4R;8l2MHx9TXuR87SMzkH6NcjeCNtRx8
z(~;n<ogLks;3z{=-oV%$ep@*5F=P#5ziHU>#1$f^x~a)C{f5q9ZWA8(TFIh!plpm}
z0kQKhOFpzK2~_9_J>RQ51lJAcQEwqIwI53>IwXeq*2O1P59R>5aQHPivq<qSIna(B
zkmfH$pb!3=ZOMPcPkUQD`RvHfYtpGcaL?4+UKn;ZSi+xBnZwOsZ|-^>7Yab;mzA6|
zxYcn?{7j<Wv)JZll$Xc_4`{auC|Hn|`bk`gr=3cBL6MJKcWO%I%>uY7+UuZ7o^c2$
zM#dob1P}~=2+*GgqQUO>%h<I4#7L9c3DQWXR4HRr{$s5#Xx6XE3YYfvs{;BAyNotb
zv-vEF%>$g1GV}WNp>z7JBj3y@TmlppKx>Bqi@={vZ^YN5UicTwQI42tfM+M&YdUDg
zadl;}8dpC&ZU!-c2BqtTtD1k?D%<i$J2eF~z=^6vGjo8r2)fF^U8pMq7U;U_Eu@i7
zMWZPPH1HtV?FMFsl+5*dSkx+?**6?aQe2AOWy7lc^lbTE>Ec^pYWuizS?d`j%ogny
ze?kbh3B`GV+G(;AP8{kR#e7T2;MxXNOtHzz_xCn1on|$k9qYP9#&%SW3dpF!KJPcs
z8uX&xe^OeU|6irW%Eb2HjRtn>=-3~2!2A6V<~2e+xeS*HDszeg_Q4MEsrpp3{<H+~
zC;RYB^xLnz(zTxdX(}7c!5ok(`O~|IPo6G)j#Mxp(C^{*@k-*tHb=)|_I+mt{=udZ
za<N;pV*{0t%kKEe`SUt5u<0?n5dYuR9)-sTs(xj}Jl?$1pi+t8gb_($%)iA4Sm%se
z)bq$EF0-vR1t#pz2LXPAjipZy&-VVlyH@l6yV?`~wQ|uKzC-AR+=K!*^zgY7Y%c*V
ztIZ7Eakls{Dp*R0Xy9ECie5^5`)7ZF@LFZ%PG>&<ph&c)SveLVFv-|)ex3QuUL1el
zVg~&TZs(L_ZSDey2fLXVfr#>^3*?|A6wBYMv2F~#yjGQ8L%1@u$XA>;Uvgu@pAw!H
z-D1}QWnGRnNn{Bp%qbx>i5@9X2t5*qqWwjuXyC66%{QzQ8-4FIUI<XwIy|0(qkssN
z`B0e--;4$csxsW${$NEia#;qc=oz3)qeDOLULy{hed4oYN=FB<()aF@T{}00wV)Ej
z$>?nn3t6T;3P=i15qC+%=6s--JmiBdI%NBXt;$M`a&_c1%P2achCrLzS8`@%yfA@d
zwi=)3SN41ZoX6}`?X^!sa!j|1C|Q3#l>9ryg6V%$mRvHlgr2U-!ed$}_4F#L{Xi=$
zex|xMB31@N1bzL$@nwIpwJa=)6nMM3ckpL~a-x2tEgidvrYZIWxg2ak5!lSqgS#Ku
zT_K8t-Zxx&#?v5vf~_uyQI_iRP@wqLfwJx2QL+-Xu0IsCJ^)2HGT?UH%6U-D7*Z!Y
zYo0ky5!%ocoN%^aX4W#9?}{)V97HY+n7~5QofcNjoX(C8rDx0{O9KcfhR#}TLvqR&
zX_7{&v^s-Om5lib8egaaSfcLU-DhMvjb$7*I}pN^{$>VZ;O{X6rOrTD5r)?^>aGhJ
z6rhW&L%5aQ7cih#|3M;z6XmHP2m`YpUV9d;QnaFQ6L(DIW>(w?Afz6Uxn8T)2>u;s
z&>D|lhy^U8aC`p~X6HPDfVOm}pJgat+fN&S0|OZQ%I%N}IFIWfmM(0sDhDVD9TWy{
zLWrVq2P|^kA8)oZh1+IEZy17+J1vO&d?m+hiMO$posapTgUXl4ABS`W)SYr6=iaow
zM1qA0Ikdazgl=#ufY6C94w5ZCku8pp{X)z>Wm6gObJ+dYwoUDSQFaBLrvWyLmF@+-
z+cHHGniEGeHjP)E=ALSUZatlK4Zj_~989JG_xdOupNZTDuOCX_)+Z;A-Z_*lnQVLB
zfZ~2A6T<I(NDZwjrAXlUCGZpPu5u~|R`h?RUXEv-#*bmgf>oTB%&U8+@cF#WzQ_0a
zd_%sjKYSVNr0mk#LGAu~p#5S?cD0co^c&ex_fS*%06lJt-@w3z6vn^no4@h(5;BV5
z1^#M(EVVC2vZlVRgn;~O>oYoK_>0E*22IB1__^`dRAZ(VSjGg~2jl+)UkY0J5FGix
zRet4k&X@ViAqtj+*Gg4NUm#UdW#uy+f~JY<&eI5!F`-ybA5+L>Thcab(atF5X<Jnv
z(Z>eODHo7I)3cfu2>hb>C4RiAJf4#kp<a@Zb!V~DQxGn%ZcPo4&b_2^Nis+CKn*S0
zD%*Q~e>Sk%<c)n(_66`9Iwa=LFOVKtAEK-;z;OR=h?ep^&-G3#;sgK2f?`c0zdnaT
z>a0~^&!Jj5pj%zpngBACF2|oZJ$R@BvywDdaj8>IA0r7c1<V*8Ww_iVHuyb~`2;+#
z@xv^s!8;x0L8$nt2!*y=U=!}C6-YHb0J=$mK=jnee9(NJ3S}#Nhi+|}<5wAN2d^Ab
z@(;gvqu0&<&IEc%W*&jRWK2kW8k4xQl^VH9e-_O}%Qxq?{c^9Y`pY&LD9#xk?0J&;
zDbAG-0liQG0W0}*u)<=d5&Y<6SYntMvukcf7R*|rjyz(lZ3{mo+hN^|lQTV?zV+BQ
z&r7<#LK#_Fkj$<ID_WmytVTQIm)C*ncRXPi65ozP>&)l~H&tCjhqL}FMj0bG>|{i;
zuX|x&M#0xEw{R}5NbWN9gszg1stZs(!JNFH3ANfHZ~mvZ!?tZU*qhY}9T=m{!(1b)
zzPt=TQbcLue3;+~mrPbuGFDkQKhVAVaW__1&1!g%3!0tWS=XXx=`m2(u4974n^r1y
zzo{;D4i`ip&)Q~ToyWF6X8^re<?6tIICpQ(Bum%TWjGVA1u)|3Y#Q6>*aUkqoUfb*
z1fRgj?e#S7yC5l2k1Xr1&-<HP)|(Nl3T1gR_T^w}(&gu1w~IC1UzcB|vLueukgqw~
zGR)00gunx23ZVEmq?P$7YzDZ&Al0%t2N^Ag|3(I`hz!B~cH#XBRGHcnqGvQopLJI5
zjCEsZ;+b~;7!Y3z3RGoTilB?t5)v-MwXOGZ65(akZYO$7bx~H-BsE=>;lzETP>`Yy
zZ!L;DX#sp<&w@;%H6@bR<_dM{h$)OM>&HOcYj!VC(q3<%BuFY1>T{A);a-F1&!Gsh
z)RZRk^GfF+J|?!?5}vaM(sg2*F|8+`Pke#)a_i&q!=WZe72+ZcW}D({g=dFK^%=fC
z^TJ(FO&lO11Mxs`P1#+($Y6uSIPJcB4VPyznx*H?!&@K;Suq9?P*qF^clA7@Cu)xO
z@lH|fh@#;$*<FHfH97igCg}ot3m&N3i}21Sf)SMhp&f3C>v-`UpeUjCR%VM{5-Ay|
z-(mJslV@C;eL}gmKZOPaSq?Q`Oc2R<YTS7ePRqW^AyK-)>Tt^Mj6^v_a%-n}YHvw}
zb;^)C<B;Y-la?B#hd<O2&cl59`Mg}-m4b=NOs9q$!;A*3v+FvTj6P6j$wc6nmBS31
zbO)Yyp|%6cXl>84>C!YJwA85lcujz0pwL}J9qW^Y>c;o1W`sd~CXl3K(Z^PGCejT6
z(}&BXR(ku|7QNgCp3w?Z(4{03@md!$*_f#lmifQ?4YNfdf*l{6;lGq~k>8E51^*FJ
zyZe=MCl{>I*@bNfCRHp?%@W&m9+CvRtj3;e@WjKphTZ_}Ec1;dDz=2&L|8Zy8wAJM
zTr_^(Jm}T9on-VJ@&J-pMkfHRI0BfcuIQpGYLQb$HKuLr(u=643n1`Xo@<J2Fp-@^
zOjm#G5AD!F|6raITWzO2mRGnHW@XByXi;pjk+Y;Ffo5w=radUOR}`hWP$s%ol5d&8
z^cPe*gHjKZXb7rbFME%BLFy=Gj$^AA-3RKHV7s7w8cJZmOG;)pe8%;}$2%5M5o_yJ
zWxrW8hGkq$Rq?;EtuxGsvfuYb-y;PjiWp&7C^s483SU-w$`x6gC`JKg6*~-U&OBDF
zJZ~F4Z!nMnu}<ig8%rUD!8Y@>PV(NH_sl{Uwe=@ph%V~tRBdszK%|Tk%@ORYodMBx
zMif=Id8Ida0=PXaBCaZE;Q3CVsKe7EJrM^S3Nqhp)%sJ*#z>1jBWa7LmGNR9&I@c0
zUpKE9BIlWsj!Ys&TBB8)G0VELq`E!3l(>#^UUg@AFubGP>X$d--cIUvw`WfzY!D8K
zDwQOWvt2=tK(BOq^_bEo)b(WX9GUasQC%vt^Z-g@uF)vQMGI$!Q5Udfp-pg9^IB;w
z+qh=$eec$$b7)=A4Q{BY_I94Ugt8G`U2B1hU5Y?5L~qdY{2l8EfRQ15LlqT%fof(0
z1~iJPa6XN;XoVU>4k2Z6m8<M{hU3EiCc$SvD05)iU-&^)z5GfuLlap@X@Ad~?IA5q
zmEj&0VMYYpKO?-#7*G1@NhHntWRQ!jtYEV0Pv@AqXQ>=!TA!A5RebBoJvKG2o_GTM
zr%vLU4>Pkrd|Zj!DV+18HXVGr-;}<)nAqE!wXG`Nmi1~q0tC{0o*iV>8qDiE;Ez4H
zV@O58y<cDvuiJXmjQ0#YgO??j2}Ts`^_x(_*B4`zOB1&V$`=L)TcvZhX;(>Iy`Ud@
zF0!|8HV1b=4(`k+<?a2ijMDW#|E0BO^V0po+oHUb`G(|bFDm;F{|wBW|F5{f|9`w3
z*Z<>&Vqs(d?*Pm{csYB#cI2P#@lE<+pOJ6EFbD>O+sYh;5D1nHAqTRHVps4`1Tpp<
z*6PQ5m+`D7>jtUJa?_<PoR+bBW7g7PrABR(Ubz=dxiW2}Sb=b)PBaANEdhKsMqF4)
z*tg}PslZ#)A<M5;F>;<nB@!oZWi8ZOm-<4p+mN_u?fraw4cg(ss{G;ub-toKOzpP8
z&B8xqP9vHzDguCOa!2Y*Xzko|HiCcb%r<<h4sqUCD~|HiaDBupbG5kJGsaA#-kGC@
zGm-C1i{RDce{)x*ZZW_Y6zSjswnThmmJ^x6Jv{=TO5o^qmzWv;jZnw&N~0{`(96wN
zi&YwHo4~Z~QtQja1eUhN^HNzMcad+?TET^+lBb|(+_KP!%!vrv6%tZb-xA2umk+Da
zrLMauB|fFojG$GtVsZRqz|i3SM-b-pG4IHL!MhB!kU_&B0o%&n&mG7;z3+?F8&i;#
zw^V5eCCmw#{deGC6Ck5Zsu`GC%n|lCXao!RdM%^?H?`L)zP_ng-CH--uQ~>!*c=Bn
zwB459b~HFM{7m!yn4NC)je`(XomvYBj$UqrV4)_}mRDUtGl@BsvXVchGI0v(UKJZY
zyS6!hZ>}=clwUV^kF-0c7<8U@4lBP9y|Mo$Z3g`bzzIP_n+n7^f@`SWk$wOl5U5Z6
zcEIixO>WqxVsEm)tFQ%uw5zxk-mS^T$UXy{L7-XV7&R&`guTs7u)h~|cbH|b&iH-r
z*8TQ$e(u%ib^Aq_#q~CMQR!d0<i5fwSZhB6%h#h@_0{xtfBx>AiQuQ(yPmCLr*-qB
z&8I;}2QjG*fyBz>RR6}Ycgv_QsN376&o9%fa;*tYGarSNDqQjPMg5gy3dTp)s)Jh>
z#*Oj9fZIo%2Wyu{Xb7jdpsm0el+KzLVK!9_A?$di$sTA|a2HA;!j4%rG6Ur{4r0(T
z88OR-T74L$v__@TM)g7Go|nXHcAJ6Be~L7Qj$|5Wus3_gQKrSw#&^mAiM*VMO9cv-
zDZIjp3-?R19-<G42wqta3^@~eCZ;Y?CfWg%bocjRALW`4I~xWlMF)EQmGQ7bt4N1M
z3Nod$7_QnlPCha#*}R3iYAH%GBm!isk<~hgN`3$s-eE8@6qaUgXDpFRNO*uBx(h8H
z(5K;Wqcc><Un0R8zqgGY;Y~wZX^rchu6C~Vz$jpU(k2T~<n1Tf%In&nDo~e)_p5ad
z*)Pe;leDbqw_}=HmzPlk4#BQf-CxU*C8!oF8n`izVTxkKY0MK^WoW5ZW2|C`6=vSx
zwhH}ETpn*Oj~~ZxOVN+VK93)3Yv)IAY16~oLjt`YCpZv#cFP)B?)g+<(CK~dE7$h=
z)jM(`6!M?0pZ6D&*ERmws00JR$h#oVJG(PbO9yHtp<bNSrOY2m_xU;d>|rkJ9A0b;
z@;U|(m*I4vw_;<};;gDtFy;(?VCP715OJsTdC+)4gD%W4VIWh@0pQ}J1`$oNtf<?2
zaVsktYb=Zh=lHi+A-UL9h$NH9R{1v0jBY2pVlUi4Zl(O2K~+dQ@p<ogD<Qusn3uk|
z|M~)*f|mUW_{He#@(h`Wo9qzrtJeGUiN4Hxi`~uR+Kp#wjT@6oenF8RMMOZO>;Pgo
zpAC_=_RIAZV<87Kl471R!ia8c2&i4ClVljCXgUnT(UrYOC4zcMwv9fI5o-<&WCx5w
zmW6blJ{{y&4y0QMKA8l&pV;qLPc5QELmZlPM;j29v=|}BG-w^cchE6O#I1ilFLNii
zpyQGVhdVo{HA$qnZ*V<rCys>9vawAWR^EAzeepNS40kq+u_5;h-E7Kxi2CQee;)MU
z?s{x&@zYCNdFJ%mNr;6+V69ycy4)m*@i?>pc-yE+&Xzc28DHJlT*&OA-_a%}Rl@^o
z4jyhA4H6ipCFns(uqP`VY_3ag^o!$9;O&)9i~<Y-=q35RRW0-03kQ%oIAORBBfP5p
zh$lpXv<n7IBc_xb`;!COk*N(uzgOW_1&qX$PQKcFR&Dk{T&sXLPrq`j!2ekzo<_2V
z#I!KtUR)O>tEdWD78~5dj+?xM{sZk5o3kN7F3<~But0@I#E_nV;#Xo6<*;F&u+fSy
zN*+4~nD6YqHm7JK!C;OZr?dF{U@#*q&~+)Esd$^lm`{L_BF>OXyB{_40kA1NB}1vd
zrJXDjBpb+SN|4sQ$g(PzsvER~e+D=4y{B-HC_g>ewy$scYqe(~!2IaDJ9T5svR?b#
zA`~8gpL7U!@gFF&N_Z|?@?k^N|5`4x3;*Jg=FJwasZOSQhy>yS4ok#mU=V2!ATMk~
zO;G}l2f1>l&D1e5#g<nxP?Br~$7{HWVB{q&*c3^*<|vTiG7?RU13v<^i0vtC=2MX1
zQmxUTAt-NyAc-%cF7e(gVEHEa0p`Aejjr-mI6RP*iT8J;-^S<7<tg8=N|SVqa67S*
zq9f@PRpxTDnu{bU*jf~s7EmUB^OW`A=X-Imq$pR50whJX+cf3C%PD|3&Uuv!7q?mP
zg<FaL)~VRqG)Ij=CJF?kEqX`z)ea(P%_rC<eG?*epTjv%#P?^n=uDl-|D(!@<0hHL
z@OD$0b?S0!-Q360iYEMx8Jkew{HxeG(yN*gq`ewBfnL*tz}5L60W-CnP)S?8|6em?
zhRAPm{(Ag8`6J>Yx4$+)`k^CG`Of~cWpGlTUd=~KT1DVnX5M{{btxyyCePmQ0l(w9
z`5}5-)#KxGW+eJr&k20&Aa{CfUO?{1&V>^wmGe@OvuVv@;g!J5ggkq&H6+2$dWC37
z?<IWng(_J42Fcq{n6VsCnW*FCqy<N|mq;GPYv8ch`vr^58Vu+IC@DxdBDkM)V05MU
zq9X56piF;jhbsolR~%1Y1X8MXg!f)_G3n)tH}MyhqH*=yt+*8U_Hy5pp?;LiZQ6e@
z`z;(U+n<AUa(F;RxK=0*ozNb8`sf4o1&PH&!A0reGLtoSi|n=pXIlIGy8+7NKW#45
z3dDa3$VGW?oAzU+SvZ#2A!vSRk;t&?Y(?Jjv&H2_v4hPMNrp5~3uz+Q!470KW~?P%
zK>kDN{udl(6-&&(UeXuI4^z@ILEgh>9gWwNT+G21PvRC_dv{C%>>bxy#z!E}0eo&_
zPtVz>)NSSFk5w!Arm+u1>gb8}t(>TjG6+3FoIpeQdK3=yGs4zZ{_tmAnAH)%Q%n0O
zd;icZ2Y=k7>Aox^-?!ji);D-tAkNr-(0?qPT>qgv&&>H>A%u^#$A5vfBL5KX-XO)2
z)O3N3fTlH3i8ygPum!;D6$FO9{Usgk8oU<hFz^nUGLf_N=z7+plNYX&vcGPfzI<Ny
zqW&jV<jC;t!%q1@Xx@O_>Tl*aH91|y*?s(38YgxC{D~|fJvoUl-sjVUoqTN>-?ZX`
z81`G)Lghk5SFjMFIAPXK{wvnc!f#8qaNI<}w>R@qnEO7@5qNsII1w~3HPVhUJeYkM
z{(gV)d7b?+E<(!_rJou2cMU<}qzEA<K^hG{ob;CN==<%9Z74&OzRv#XRj@rg3Ch<9
z$!}Ut2NorV*5ud11sjpT9*mHOx_y7hbe9&AmXpL<ug@>dL}CtUc+IoO=c7HiZH?=#
zeUKVEv6s7t(k=HhwZykKI$nVyA2f}{loUeq`_0uKPp9QMy`!D0k1na}Z99Jup7B@S
z(j`ka-IZbr0X0#5%s*yI7LIgi2yunTog5l~g-)aLCXpD;*g(|)JgnXvH%6PSs!LVP
z6H}zd`5>{N%jTm-<T^)YEy`gw_ABO+O#<^W^5EBQk<ugb>af1)B4<XqwT8<%&Rmr?
z%u0rvmv$4&x`W7WcTCoVjK0Asl*4Bd5*C1GFSv1ld(=|}R#r3xcg=OAiE!;VnTJTu
z2euCJ8@V$z)L2*iHqPT|a0^*j6dnE|PK#s*2_q-z|0GBloC1vlwfXG`sX?-^Xzs>T
z;sisC`>zexB%1f>M1nE1wEerRGzXv^Y#$YGywlg@Y39riOkJHG$ft?z5XN_sM+R|x
zCg@D4ptuk`tO#$y2Fo!&54P7AYN(mznBK65lzz{!lMO4a#J70Poi1z)-^89#KLFES
z9&|gWcKpfNR{K}YWxf%^N4Th?C+L)X+@0p0{|8x5GYg$Mv_rZOYX?3Wf0R2dSnXgl
zdZ5v|^H(KCs}8b^vhDXP-~66?c_S*~wJX`({PF65grlStU@ivd++IE+86AU9A>uEk
z4ewY)T(IRt#M4395Bod7KCaMqiE>y8@Mw=Nfo<<EjAus6#~oK=7=04HPqh+wGNcrD
z?}vBP`BuYajWqssdyOm!9{{ENT??UnIX>`^m;PSASSlVpz&Nhl91N^Jc8tH$P42-E
zx(E?IlXOV~sOZQyO-Pie9HL%~XhXj2J*IhIHFtl8%Z=G|q;j7auf|6m9r$g>dQ)S^
zD96(Ik>HIqe?jjzWk}rlp-J217K5a<N5}IM8Nn9NR5@Xm4aCa{r!7<LlO#)0-zwi*
z*gH&*9FHGD)|P&FS{vPrdK{>iC#6*XrqLi>T>wrQr)+Q#(jYT`r}Eh+&@^w)DSjsx
z%EpX#HcTKAdA{a;Nag32W}Vbj$ZtyIkx%4Mk-cNf1b-2w%{<X2A<bP$>8uBF+f%zJ
z4_l09@EKqs4H+796>;k`PVKT+?TjoAk<)Xy*jBVn-MDt7Vnfb%<YYT+=_!RN`$sq<
z5g&gw-!1S?K%ONrAEoU0q~~z;qyt+2la2#7O}C6cFYgFPF+bvJqM}I4vN@A$CBt-~
zxFf=+o<m<Jep!h^7K<wC?VNaKzN%R5rv;tYgtWSG0V8jvBpqwL+aH`K8f-1xuBz4f
zdlnZ)!|T=@#p_61kJqFUZ`wo(9}5K6NBsk@+WI=Q*54DB-xbjdbWf)b*8r~$_ec}&
zdy)%V)argq2;SbFsJP~th+Vk{yyAAcN`aPZn#UFG2hd}C+FnF8Y-@W?nyU}&a$;lb
zAVWxXRo~_fgh58^BFrM_OD1W~@C)iG<yukZ?vtsI%Do^l%;;7vj+9Y}6#$Q|`}N9n
zOa3e;PU=J;ZDGmzA_SKh%}EINv0hLoK*vNa+_cL(lA`J6jadJ;D1KC1jvnB@-6C*F
z?Y$W7dTr~DXDztoJU{!S8rLu{UCe5}n;Z{ea-P73owhXWcdlrIrlSM%9+?dWtdvV?
zk4TQ1CyaGhA7ftbY+VdoY<B>VJ&7x-T@xOpFmBOv&Dc8V1<avI`^h7nTv?^Q^nfE7
zOOnH;9S4RbAxKs@tm+g`53z#>6{gP@FHy2J`nLSKq$X8jWjf~WsF+#0s;Je|+9@8h
zLHEG#NB;ieS{n16o)=DFJ@$@N(@M*jnyca<|3!-_vI|>xeDj-DU7L|ATZ4zZzGip^
zi|<X8GdgwN%8dI{i%IHr_v^`}NuP93@8nV8OdfPe7UC0r>#T5wwXp3U;?Cx*>N1R3
zpo)3-6t~^{0MD?+f%cz{knQz-?}2h|EEN1Q`>Ur$HIo-!hmwo2-`oR1>yBx?6zi=n
zm%`9;JTu(>RTA)J;6F<lm)}~5kt$I0-}6@utD>1beO<!cln!=Ykb|~=t6yxTz!_Vl
z($-vKXS0p2aE2>C5GfwrOS?p6<yXBc;a^MX;cRu-C@)`1z;I)g;~i2wp^E)kXx2MM
zyIR+Lb%H3Yf0>OsB6A0=5wmwB=Be`b{#__{F6{-=$(AGCQ`FwbIc;;>A4e{c!R~v@
zwJ)<VX&DTN3oXs*3u}T^C))}!+R)5PK*v!`*W}&5&0w9}Tttfx2|FXF$V2?q8uPWM
zt>_=579@8r$_O3RufyqNQziT}OiP4U+scf(TJbl^%b-0+{;DpN8LAeiri?AWh*CwV
zC7^@WEki}QAoDUkzLn!89lSJs>m_n?$_*9kYR8pF&CHo`XLrIMvmI0&6!D`}*RKqQ
zE@OeQw6uJPPB>0O25}Gro25~)m344Gk0X>B{zi3^OpNUo+&wbvrtG`ao_zAN*{wY#
zx1T%|{v%m|sqA%boqMe`uZt4@M%UiAbD3Fu4|VVh;STYF`)MM#7LDK6*9WS#4jI1F
zQ*)TJ=1R&`WG90APNwi>0ag7@Y;x|x1?rb?9bKes9utgr*zvp|C--3ln>`3hSfylU
z!u&3Nt2Lf$6yCfKg6i@TZ_e>L(IbZ93DLaULWseysRc^Z11Fu%e;wy2i}Oh54x~?7
zD6a9{GmH=5v2=8-Y}#Mg<a;P>Vn@MqNk1g#aTMG7=C&rZf-y}gbkVACzZADC$P)Q0
zzbm^^{F5kO8S8CT`Tvu&;`$H6gv=cOowVxK?5sK9K)U=ye*<>rK7=C-Al{u=(I?Ww
zxC&AgcB`as_%<+oNhDYHlXO#7Y#_gzdUv2&`#d#$cIDJ{Gnm2QrNs+W+dA@w{1CZA
zClh2_)B-YMb4mNNMF>fqvd4TQMBRLB^xAsB`{)fHX93g;ybt`jO>)SteG5cDQN8!N
zPKXl6bL_s?ZiHuHstNf)C@|3deSfzhZ(eKzA1mH(BLR;ctJ$KS`AuviLYIXF8-9Z}
zsIm;A#S+>n<L-%If2MdG#%U%HT=dEVAH(wUh0Ol$<W_em3jTrQQuqT?8lfb(OGi#E
z%xl7&V1rWXq2_?N9G7e2EyeC<X)2u7T;2_e%?sZppRja$Fwhj|PrrF5o&3FKli-ke
z0>EZU;Ylkj?P=mh7@YcJmc!W4NS|(uZRhOX9k&dfSo4Gp@6I0^7t;I`n~N}?smm@c
zSqjK0*#`Ee!d|;3wy>JoI-S#0icXIJ`HLxg>5ZFEAWtzKLA98bhd-vj78xE4-(NBp
zRyIndBTuIn@yP~Iny<g<Q}w3tQcU&iA`!bTL8o3(ZSs8>D>E6Qb-s{P9M{*OGm;TR
zwTH3?aShp6Io~;JcrSINa+>qZyQ=-uc>}@@HO&lK8`9A8{^<Vf`|J&$H7*4+ZZjWh
z`g|dw!T*}7wIVYkVg)xB<0|ZAZES%CwJ%1m^^#%Y(xV<6giR}LL?>g|5=cY~lgn?^
z-)^j3KIY86q@3bcfq<)j{B~a<934ZcQdB2)hPOs44jmjmL4GdieRoMC90;Rd3+7-v
zi{Ak87Y00bLTCdSeIe2s1lqPPG5VLvI<!lRzOa!qU$r|j&K{;u$Ob(LlAWqdtQ`m$
z1C-D1!-Xv)-D?TUq1Ou5sCR`o30tKiz6SQQl!EqPCwzPfAZ`t3e(MqCXeL+oFfOLw
zy}=~~7iM4EjNbkEebLOSw@KrM>*&YAB<SMkNk0oqubKpJ^N$F4-u<wf$?*8>Plt4J
zZ6PluTsuS##$Q6zS5gd0lMp@8{dg!ZgtC<rfoJHSsnN_WLo%Ytnfd0$M^i9}khB&M
zufXURQ=){VB9J27cjP;HsK-Nib1wP-+yIy{Th77zwTX6|AK?3F(v7Bs$zktfjPO+N
za(^;&SG+#m6w}$_D1p7?R2~IB-iwR6Hzufq;6-v~(Ktc!Q1Pb+XxsAU!I{(zz~TMt
z?WD0$!iy4k3K_bUwGZzM8GOp3cL3O#tZh!zDKS{-@gIEmzUF6&u79h0HR-cRij)8`
z*ZsEdt6-JD_C%=a6TpXqSbG-w(%$0>-!&||Sm;LZPby9MUem*7hCBCtHhLMZ&Q*3M
zAXR?{m{wp{du8mc%bTg(e8KTPzbMtaWA3)}UYH&;_~@Jn(tE1cnDANTI;INq&sFWe
z<%gWLl`5;$bT_G`r-aut(-PV_WzWI#<-J(fCo5<_8l$+*m$3P)Ifl|G1q*M#7Z&uM
zjgf0ww;7(_wN$QO%$@gXs%};v+wW7Yg)d94lK<2SIG0cMM^n!II{tlugDxe$|EJCE
z2ZH{8PL?rqFmwG^|Bc;R<8g-^aNW-uFOUFaYHLs!aF=#WqE@yyj6<}kL$>g#Z$gpk
zpB$Yu8!1BFqeL@?_(js<#09~mY<)u0FM)2skEenyFlkGZ>WpoF***m_|DdJ;Y<}!W
z5h=f)Nt=5javbWjAxx=qbc57ueE*9BPVt{TP7v7n<Kpv~ZWGEy%uS|kSv{ech7XFz
zte*##<E$W~_ctR{Z=;7eD(t+WqYRHHV@x>BXXK7)B%LnJ@Noe@+e|tOiXM-68WO@d
zjMHBExWZ$$t?bj%^D57sS`DuVUiB9qGUA`tp*nv2US#ncuye69DG$+}I$iQ$K52ch
zta0;fDxK-OvBR}f#8;^Vqso-J)1ow6(Pb2ii$X*UpgMXUOI7Un0AUXldiRxgtpvDK
ziuVe-EvZ~gQ{^c7vz8YX3vfE7<6#=aY;^S{Gz2f5Mb}_sN_x$6CIE+b&E!~KiiBDc
z_gpwz&{}1wZL1bZu`dp+B<>CUyz^WQ!a;4i`(oa269VUbGf#}GN8KGK*to;Vf9j%B
zn}<4)QeGps?Cd?GaH4ux7f9|%^hhlbt250NQPnlSR1;L)CS8F{CC1sJ9c1=V5~hdh
zv^6!LNzO-`s;D;l`^#mI!HFm2Qps58WhfBf?8Cj~2~zQ}sH&{BmUW0_J6JD>nM==&
zCmOsq`oK8?MfGl{E#Y!zwPEWkp^i;*4L~aE%(}Tw3Y&{ZHxZ%umRqGq$S?7qm$)%+
zWmDAVi6>5Jh4wc_THDEFAFCAzsTFy{lhB6JRJ7%K+Y)LdcJ#TWvo&y)NT3Juu?s`2
z?Fq<aoPfWnFePz?t&7RYJhV8-&xNfk7`#x!2aKDy*&iw5E-s6~so2zq+y3g2Q|d;o
zCZeoiIp~Z8=s_Nl=JnG;WI~80D7%GMUULZO=|prr86`ARuAqtxQ}e1H(sN#?Siz)k
zI+(c#^Xax>VP8F?{4Vm9qFNhM@}De~x3rI4Ct*HnGK8omTPFY2VoHclV-nKOj;sE#
z!B8uvdfdD(U6nbIl33?tl$BbK8lk*_uFgxV^RS+JKmYOje*UyL8?uGIYh3Wp1@F@D
z_r^>XWh@hjx@$S>?7vLst0he)E^H^6Xxm_Odl#-Rj~`z**HN8{p2Xc*Uth0C1PLJh
za#LVce_Ev$YOfdK!vz8LW;Wzxy-5O1vXXA3_MRI@<nL6ma~+e~hIQ<6Q4u?3Lvjjo
z!JL%wj$-s-=Jdb0*rel8R=VX!qJnacNHce0>t5AZAVIiLf=!ci^jurWjp$+;mjJWm
zqVYre1I{gHd_@+wVrcVgus5cQbkUIP!is3kKX2k?x^518L`pNUnrk$lZywzO%s|J}
zS^*bV$FV2lkP<J1WwVBL1*FBF%5lz*G#f$?0qISrF9gh9DhZ!}4`4zTYj~N+dKvMG
zu<W3rm^PQC!B6^pCpesp$?f1rS|p#Qs&<6fHlajudnRxsOy_=Rp2CKtxsl&?6sYlv
zpM@O7?h-vFO|_vxlcA+%%?X2^+If4JoPSQ_(vk#ZPX5s~i}2ZAwn;z6epI^lY1hYn
zlhA$T1$==-#yJaDBNy2zw4oo_96iHVrZ*{0mQzZ^tpTg3-r*N&j~qLV5~T6s`0SjE
z&9IM9>6~lh^KZ=uWU6^rcBXps!Ad(HtB2`xo?FDvuV>t@lC&;Q$CWN{HT_#;zve}N
zIzXmYCW7~u-8iZ6<im;e7<=NQB@onI`V+Lj)vaRUKf&L2r06g14Ks7IE$8oeM9?%c
z?<+vfG1;mDQbkCfTf5^(nyIpjbEO_fY9Y}L>TaM8kJ1~T%V+1Tb$rA8!_zWl+l<9(
zy*q|5ZwYhuh@F*~%J%Si`1CS|sN`NqHPL;?X65Ya9-ceC>ef09q-*1SG7lc&|5hIg
zF8Dwc;j6YhNd_%wi;Fw*sMnF^1fygVSpR11Z!ZZ5`ge9}Rc_lsfEFsKmG?1xmt8UB
z1oiMbQ&>4vX)^ffZ_+#(uLMCqsuSCF6`z~qf9R`vP5tLKQ`zXU0`}!4g3e5~qQ<(h
zqoZ-@v17tI-7jKfAnGIG-~&B0<sz~71myZ~>vpO%$@_SyG^_KX%K~+mwNg3I2y3yZ
z3rN@NEkQEg=g}7k@>_7`z7q03oKLJjL(c!_`NYD?_1}46hjAu<;c;JW8z>1HiZh-A
zV~uodxCOME?QR$&1{)smlSSub41Rnips0DNK6WnP=FKypi!1v?7Lk?4OUeJ?+<2ZY
z<usvW1fi-eeYQ^;Kt%c0K!L=U)u-Nz$^1F6zI~FFd3{=5G^u0zM-O&@{rArL%L0&K
zdZ|K`&(HdG9>;cC;SfKr?h4&sDsPn@;<$oVa_flnA7d1J-d5fjamV3gknumI1P?YY
z-jAQB#{@i()Zfl7;^BSk0MMvxek}vzOrN*cXB)?p5pEO|j`ClKhYzEpY%w1bC^0DH
zR5_rhWMyPgOY$?~MyD#NdA`3<6hWTg33a>0J+0WQ<~@t@rivoeNj77N_7v=etCgFf
zYz&Jkl%9}S*#o54K%7x{5X@}_BFZpV$*fT+rG=)o6dGAEUY{Oz$<FX3cWT8b55&o3
zAA6e$ejTnzt6WeOAV#B)2?Hz)*@RbjC1NJ@<L0abN4oo7oWKE;B$+!AbJgeK!6-7#
zc>={^!xb7BZ;tQ$)?Lf6HXq%wt)<c<G86y+co=;8ww_-EDvP+{1pp6f!<c?F)}zR?
zT)43zOkB@s7AZ8G%eZSKHwPUBoz1vwAXg9Rg^uZ}-l7)VvYUxhYcZdY5!IJ%V6^z$
zK)szw8I}}NS<)P%p`NUey8beh%T^D5wX>o!<u_=J`QKAwhO+fB06c!}yY6TJrab-U
zqt{b5sx*BJ{cAQ)Ec>RIh$h30k87qOQ_Kr)CiOAyfS7AR21${_GX0d5`+HlZFp%`;
zLz_2S2?AE<>WEKp#<_pw^{=Ae8&jnXAiC(d0u5oxlln=1x1cX5+H$5;ck~8y_<vTi
zbrK&J3d50Nc4b`u401q`&!rdle<;in%n-=6(|WPpQyIt{X7q}_)~xh);vm=TA`}2v
zSsPW21PS^7&?a#1Q|0YJZnTY3yDTSKPt9X;^Q2<UaZSwm|3SLBPtHRQj07&}En$m^
zi~mN-k+>E`OMhwGjrg1{{r2^FA6?mD>P#2-B2$!oa2!=*mKA(K2kjIl&FZ=5E%_GM
zbt#H*4bNeE56VN%lg)<gk;kp%R}TLh<5xNtp;bJrS1B)l6iV4Al7~&hhcv=7@zB#6
z%&Nf}*%dJW(T$9eD^ttWJ5`|Jwf+ne)%e9~*CUt3NT=q&U}iyV35Swg0dBKWR1u;^
z0NW2}Fq1Ta>+_+&C>M+G>9A<hvj-PHAPH{efr{6c%pg(8TG0h$x~35isWBzzXEYo6
z#N&lE?~mCQVX-^<GXX4BK{LyY&&;7$dV9ZiJqfukobR)4`Katdh816dFnnP^_N13c
zzuMLPSv>;HXA8Y}?Yb#mweuT}^*}+$p3SUR$1l9fVBeh@n3$eSnEiv8Mm`&ng8&yU
zju->E6k=MSqL{@0a+9Uz!rfwpav;yK|LP~6>!^7lx7AxsXX#bA`3BNQ3JM>p`i~5#
zIgaE!SH;j@s&)J`g=?R@whAxQ`W3eAnvL|)vb<&)cIucMmvEhd5{_9Y%0SzV9h$Fj
zV9lP>-=TTVWh_gs*2uYXN8{?MEo<|n&faNmzvR}?{po(*@1fvf$q~9w4$SOfvD~4G
zMW&>G?<bE>&gA+zPIF8L>S`=npSvp5?Z|A|PDH2dBX1+HOx>7HrNeq6;(M%IWm|1&
zRyoJCUgtJUu;Y=ND-meN(c7L2(6WEqz@+jpudt2I3AKP4f57p9j(Ydx7?(?HpY19D
zgsE0z=M+_9O|)ju_!;VSG}fA?8_zXN{p;FL<K|d|+1C4AqLXG$*@_0fI)B-sD2=uq
z#tt&NA$(%!7@Ui5p}UEfT2Ziul#dd&y(s;13WD8GwQ0g!M)~M5RdURQnaRis?614R
zEuFJwC;kd8?KSd?mRryq`}wp*yRd$rlanqRB4#pmZq2N!AaVT+`Tj3*1;6G5`4W`f
zvbM%X7G^=J-)DE{3g{B;&#h&lW5vncQOTBsqGe*~+$mx9QJZ-?8f!48nMJW%nHIDK
zwW*;~o)#g^+&Z<fvs;~NY56>?-eL%=f;|V-zE6>w-?*~?v>FdR&2Om8Q11Zm8sbz&
zeW2c<kIdki?JHJMfZ;TE$3u1T1r`>%g4XK=MURnz+_Ls`sSX_{tEhXPU;9k=@_2i0
zb4B~6u0@Y<a?BNM@7?*pMC$J*s0-!M2F^h^vuk`^`kDvpanMb25|g06?{P%4&7O++
z<z>HaCXq{4rkO|O)N-m<<!({PwBQqZfIxEfkrui4w7p_m(<DqH>8J<~gfo5Vv@*}T
zR|TY*T>#EDtky<S^DCNEoGPzSM6rK8ePTmARj1ZPHrq+_Mpb8Wqu|PYT1(Q@db$~h
zWVxaC-qJeuq_%o3zTx(|*<*et%1yqi9VXe*dx~<emFQG?2cGnUxUqif)iXnP6E@Ci
zex##^&M^f)>`0hxU#eUKs^jqv8{eQSc@T^63()s0im(o;?jq-A31kc>9Wum}gmeh-
zn>=~Yvb2Z&b#@GbYOO2)(pIor?C(S6E@dhd4H)u%xBV|T)!N00fXmV}s3_?~io2a2
zF~!}GE~`i;<E|Ppth&o+&g`%FzmN#Oj==ehV6e~rm6SRK7kLj!YG9PN+buJeQpvsl
zWh&p6X&2f&irWy`q?Pyls%~=hbnx&D(5Y>LTz2p8#_gGPK6nI2ddaTr!2-)XH1^d9
zfv#FR#HXQyBZs7+BhPVYp3_30Kb>O-Z+%<rbWTReRrN4ZTpn?wOw7JAZnL*%NLir#
z(|NI8mj5+zQ~b5RmcCw*(%7jS%#5~Th<&0N%siFBjy1dQ6~rt$de)HTcV)G`-66+o
z^jF$@Z(Ff)scN~|CPmsk(+i@^iFK|8y{igbScCH2)mOxV@Sb`waDh8_C1-}8TSfGH
zKdu9HeI-loJ^q#eZfLY5(;8s_ibc(pp$4Dw$9k!>KzKHtJb|nn_U~la%VkM5>4<jm
zU>Qs;hR$WxXB50Meap<EzNgUrDiU?5WyTCV%}sWEOPQ8L)cW|%tZis`G$bJodiqFW
z`#pOIf{~3!s`du97LpLwc?PD?PG^BWeiCEUHlLtn$FLiw)g1SsZvsCU1U;EJ?02AX
z{l-X~&lQcT`-%H#Dt;y{^e@NXdF_ffoXg9U|0y?_mH9suM>*M;|Ev1J^iSWJ^e36-
zfBMctM!$#QnL<_xejvqsH}qphhC?$(c%~HLAASU_`m6LZE!rY*P1<^L%lg9-6Aj*H
zrq5>1_ie8(we8LWxK0(`K37nTWCBt_mU1T%Axn$=J#M{vQss0txVm)3%WZf)-ka0E
zD#uO0_TY_vVzMvNYbM^aShkZbvKU>XI|syXvH7}G7@U9&I^tHYo<|Kh8>r~S>uznj
znibsRvk0@)eG&9m4LeF#<b3)wL!K%~<2%p&zUKJt3AAgv=_09L2gj~XWL6Y=Fd!kI
zA*F3%Gq@7Sx+Zd(g->5-coPI=tagEVM!^rTuaw!Ce}daf#pCFZVHTW1B#HxJu^A4A
zxSdh)9;*nYKB2+SpR=pcP1Po~#v}DVEk_4<tR*fo2#AROp{UR6&7IZBAigE#7+!Ag
zDR@i=@}I>O<dM0Gyn2PqmMWviuoKI$nkP(^SnfZ2%tYszZx*CYB|cxd>(H7&XB>|1
zixzh7k!rE##~f(;mQ)*_yvMX#eKd-wMugU<Axwy!mNREC@sg0WmWGO18x^06I(2$U
z!g*ysb{QX7ala9#Y%kcxVDP-8ZuFR&NP(T17n&G~_9V}O8~+%UJ_6+XBl<d0L>eL|
zZt;F|1`WtwH}M!(lXhEHlb~bfo6kdAnP;QPXY}Wl4ZM|)=zE+k%QlCRrm@6e>cTcU
zSyg0)iM}i}PHQ9+jAmGJoiqRlPj$}U)1-^L4X?Gq$Gh+~L`9~nCZ@%5ws&G7){O+I
zSZ}I#5+)uuu=f?H2)oT+O5wi!=l-0!mzBxfP&xzZ1?`BL6U7}K)c+lA5nV+1Xo3s^
zt*~0k3D-FxYk!|I-N>W?f`L0a&%Q26u#=cpOH%Y$l_CIk0-v9EACn85Rrgzyf{qXg
z+`3|#==%4C^&(y;y3gHwgV*aQMJDWjIcolhTm%5S8Y_65b11@c4ZI*fdKZZp%}U2o
zlzs->`ynxrsNs{0=4v1L5~5bB2S=Ey)B=2>bSE}nB3@n9ePUK#+7X7U1UsjlTRl%d
zvDv;*bQ9gibI8*1T5L(v)KTU!qm}0fIv~Oe)<p4f*HO`z$)iIPQ!o+AQTw%jFXQS^
zlLpg5chQ7PS3IZoi~`42uaSdmK%p|tiYy^DCPwKr+Js*#)S7L>n)_mJvHqc_rpHor
zm|VqR_JPy4Dzy06mv4uwExd3$^(viS0+KL6m4>I@zOQMK)~9I*qgM4e1<;u5v<BBB
zO7KGAdYwEVBUrz$jxkI6EOnaA0*f>UH0RAR%>7G;^_cSYZ6!ZbQHZ<bZ#|Wi{+vkt
z-O_c%>vfg}1C1F?j}g1lY%x=YYKS4^T&;iZhDO4<&Yvtk<&zsyOOsyMS?FQ@JC$$-
zerxrpm<VXqh;$h$o{A&Gj3V`CpXMUtAsMf8Eyp!4Qgg(zN~O%qDLDUD#5@;(a2-q*
zllc0z>#sMM^s$rlX<3i&^v<78ZE5HaQ|zLH4##F`${m{WFu<g6w175Fw`5x#R~&_3
z!Km@iV9P~QOSiRfsriY33N?F4rql(;6Fr$As%ll=t<Fm4r{Xt#Tf|V@#JDB(sLEL|
zA&`WagKAwGs>I*4#7_GWwPqNySe=&+Q9@D0K+VidN9$xdHV%qI6c(GTt<!<_rYMSv
z`R*rnJ%qnjx&8aduA*~*bj9O4%-Qa!P}O{%Q+HtbJ`QSGKNi)ubaAQ2^qRec;ISo3
zSOog2VY1m-$gbuQ(BJ;<Q^pU=nl)6MUq1_QRAOySs?w~ke$Kt48Ah&{miMfu2f^ta
zV`h|=2P;5-7J?r|kN3o#fsFIan#1HdUII!dz6n)8!!7?`jGa?(Xj`|fqZQk>ZQHhO
z+g`D4+qP}nE6Iv&J2(HiI1gvvs{J&p=Gzz#qgC&{_4bkBQm}5$_)bM;h=wwoaUZDw
zVhE@5dfOF^lwr3n(20$X4nC*Tgcd59#KhWc{|<mB;KG!r`n=jFiw2z&oViH1Ise-~
zUyx16OEY@No1Z0^t9E&U<fP#N9CRTjHG?&ZX~w>ws0!75wCgc=W4@*sTa6i8qfapU
zmw!2flofll7tR|Wm^m+P+3ViJf`Kn6k%C{zVx-sfIZc?>D1qpfZVX6k@_zrX=o~<0
z2%c~P1d~$CU@vW6T6^*x=$DHLlsa2KbAKNjO8Z+O{*IqisB4bt8UY1W%aQK`_7ACM
ztM|7TL@lFT>Ds#u@f8%Jn;6k;?4VX(OIj052aKjVEE-_z&{#yBDVsHW4Y=*k9rf8T
zy$SB5Bn3W5eQR%v^B`>g+z_m7c$K}8d;MDio_ZTREp}4*9Fh^EiR;t1!Hp&~pW^IS
zpg}zzi(6CYH-LFw=k!0dE*SqO;gf-b<=+vXmNlhf_gSHT2%j%-QL07aJbw6n(yk>x
z8Kg(556aHZO`%smfKN~L785cD!&zh=@)nfzpH&}T&5rl&)2r0M$@ksK%={UO*_}GC
z?-l=#{6n-RXU!{YzXs5bXS^*uvFdqAuXs^SETZZ0?o_JW9B6k8z&p4xa=*nyBKDMT
zy$N^PSyndgNt9v%Z&b|f{7B^>T5c&Xj?dSxa{)B}W=PiU5vliwH{X}zxro_qN+X-+
z)mWjAKxp#)dMHs1$gSDlPP@>*nia{ZDQb6c=U2ziHK(uL<n-nWc`Yd(1^jabDrZ$N
z6GOcXK&O=})R1{>7K|Q<E(FYpO|tBMr}%f0QDi&=FO!*}A~nQDFs<4O%5nqA!_d^v
zQ-|wEs3Zmqz;Yzz%H2xFrcU$m88dbXlgRsB4}VU_&&=O)N*{<jA_}hFa+w<?X0L(e
zn04dbXykT8AmZi2Tbm1C4}E)}A<ZC{YNNxi2+))E=`+D;<c1O!3{O(N+ogkK*glAR
z5_1s#p7~uQBh2C=laKUQWjT^HQn^JdTN&<{0-#rpy>bEVlsK^=+c@-#vn09Be4b)v
z%ilqDUNjJuQrl+au$rT5Fa!w<GQ3MLy`L*e?JcyG(qcvYaq-wLl7b~KUsu-2fa=hm
zSkv2Ha}z))Xpyo?H|9a{3jx1StH4HaoL{ghBTsfzm1<A8OKPMvPAB}WovE<zg!r|h
zIEkM0<Z`x#_4sJa?hxrM(?vv2mt=9geaF|fuU>$(cblQSZrwnPH+kV1IFTvOu~0wt
zdm-}xJXq7WH0tn}@hi(zw@6ABg_4IWIWRIOar-jwoO$qAAVSh6Ir_x5dZ~40pN!go
zw6;_R+Pm8EgfI~2A(33hB{JP@<r3fGZ1#M83ayW#FJo_EjK1JTm09zM@1){JMysAh
zbA~aBSazdeyHgv&*qH7Qsm4Z;$0FE$mD^hSRF`%OKH^<XwnnyqHT#%w8*dLY3xyrW
z40K(B@|Glke1TFzs2X+_i`2XX@7Ny}&gr4M?q#)M4vuw$yviiYgUb>t{)8!Uno=+~
zf3)L$;?;?flHG+7;SEduN9sRbYLAMv93(sr@nkgoHBTKmqL1ep(W3F!UbK~WLM1M4
zD^*P4s(fmkJ|G*2**(!yoT<wV^_8#elVy$E!$#MjLJ7^ui%j$&h-lOR`%H82@oK0~
zn|;j4$L<HVAOI_rb!&>E-BHol5S)RoO&S9t7MT=JRQiUYFa-H3taD_AlOa}9R2D7;
zd(**8$;lP7)=bkLDE6Y4tQDJWI%KzQX4a7iZkr{U`bl8l<Ej>IwoQ^kHo+<_HXkHN
z<IMr&yw`RpR2<RzqnnG04{$2je$fEVe-*9nrZM3#7}3k}wxaf0DB&%0Q({BTGTgg+
zA%#?f>GwtI@}U#YIDbanSr6hmr7EK#lJSp{8##=em06h_0bd=~`m>UnTyF^X5_k6g
zs&_1y$@8Wo&)#f$NJ2WOyXUMJqF^TQPdkyK_4_zm%Xt^;Q$IGiD3y_y<_ZGEoK4so
zeM48H&Xii?nkJPkTg1|22PQO-&o)pKlLshP4cY@tJ7ShuA-h3H34vdH7tSQtoSNjj
zF_2vh2aJ;A>F($#YUjnRsSyLOK2sZ=%H%!jah(C>Ms0*E*}3hyWE3i`kIiu^=GcMm
zoANMi6u`8T_MT@}d{B-_t)y#cLVBO>R8;gU+qek(goLt2l!P;O+CoEZm@bB~VVBu1
zLofBQX{F0ix(mK5o(o(n4rOoE?5(PK7sjZw&8X>O0|_}w1|6+_)eL-%T~HI97N=g^
zqniHq`w=^SQ+gvVks2~jhAx0LDtPxNIqz+V8><~55x^)JrcI%be<(rv8W5R2Qc-}6
z$=;b;_i4NH+l+$q^1TZM=c2xfAOSH|j~bysluEGTPm<O`Q^AY{5KBYhsNwQ1+{QCz
zTgH>lRSTudecU|&jMGhJ@YHbuu#mkvjrJ@kY^mUwmuN{cfj%?mV{d3AYO;s-(?E@-
zWc9Tou=r)JqhW?&(7X7~8VaNQe*vVWN1@TTk%~Zbt-%5A19d0R1Oz#y)pz5$5kN{h
ze1%ACDDiUL(2)5(Ij#2mAQ-9eog@vZT$EyfE_WbdQHW{|&rKv4>Oy6@<kLnM_Thgr
z#emREk+taG`rZa~*XPBizWLffZyaIv6En}@fU5##1SR6Jc*AS`(vtEHKj9mDI0EL1
zOa14R`DR4XMIQ_A8M$ZRWBo??y+FwC-?=}9f1fsX3LtlHDniLp-l(D2RK*t)H6gF&
zp0odAPbqAJ$8^RmB~5d2{tDbuFGnyceLpV1<l+5N5z8SvDlc(xlI{xX?`FR_5n9j2
z+uy%Rke7q2;*MW|5vJoYy4k{0WT!($5M-B~gr1}%4+R`43<XxqN9FLOEfiu+f`jDI
zV|PRu2>5_1!93#NAfEeHl1a=xskn~7OtuEX44jhX9X8tN28v)m5KYc<JnNSx^<5s-
zn1t0o0w;2sS*kw}o1ttGvT)Ys??)&n$nc?9i*912d|0Cj&hZpv<a%Bh7dL0cIehyO
zJhbienEQVcd?vR4F{@zUU|{-JweXMN+Yzxr_q?l}zzI`HE<u4uk;*pk+i4uAKFEW}
zc&;IDUFQGpU9x&ebFdjrG{(}uz|bfti$#|3LF&&**Ux2Jh0Bz4gG@h^4NSh+4F~+h
zKS2Uv!JMWBV1ry6gnB<z1h+WpD1i!=kD!%$j_G-P67(;`<?jV}1X|No51RaVce$#)
zZ<I*goNq0H8ekf?SKYu5J{UmJ>Ab9k<FgZ1RkybruzA7EwZ&uo%Ubwm4%52QFq$g!
z0*GRNNSr9>P>%in#pSiJ8EaTN4iYVar&)JBYi<6Tfbb9ik33?&MF1Y4FJc;XO!<)n
zKZHfVOyK9{JbxB{lQ36(x0oF$5>7sb>X{@S2PuP2Xb_^{m@vX{qQP76rUO;ZC&3yL
z7++{i!P#B1joMyM|D?*qQ|5m`Yo2VM>B0$pD!@JOsQd)N4Z$L{C%^!XG_28?LMNub
zJxgg(zfrp0RKrv_lhMXA#&LBht;sJzm+#ivO;311fa5TQZaI7=rY*C&u+gkv0xfn<
z8rzP0qtQ?4r=9uCD#>=*&UzYzW00`rL7ES59F!&9ptVK(lV<PQX1Iev0&4HmV?034
zUI802R|=MI9cGZ?BQO?B!>m<o?VOO!+h0O<PE2CY?3?v#NWWT%*wXe;`Z|r30wO?D
zn}&ocW+tXEfo;N6Dz_kf<v_v-WOzWAN1(zXt43*6th*fiCHC=BRxj5r>qGzWvAnL6
zKc|hC1ZCCr=#SYDoXxM-cw0eFQOQl#v_r1DRmaftQ4Hrzh)&b2QxeN5M268l6vJpG
z>?%C74Kyq7x(qC@N`YFQThoLB8a89;>M~XlH*Mr-FMWjw=1mU1Q$$gL2{ne-!z`3*
z9Z%R_?pZc!50k;ej_D2qvO?qKD|jxPYzav(OHw?@5|SUMVwXY)%+Ho4@NUQcYvoO-
zA=&bCqXKI%8=-mpS2S5LyS6gmUfC-Hb*Y|Jcu$3*$WB8-vD$vx1a$Z!tPyD}bl?UK
zIavIw6aE`^k}ZA&fz+|SE-f0IW}_c_ZBHv@B#=Pyu;;pNw+k!M{NeN}r%?4wyqV4-
zc8TL$A`9p6ul|Vb_bfTD)i9M23HVlysWlh9<CcnCWPy0|hsyPvRn0l}N0{<4bq|Yq
z{Sj8RCv-;m6utqiOfb^C`U3dg7bYSEGbB(j$Z$F}+^1Di5dcb+m!i^EnZ9^;R@x{f
z4)qQ6)kVlAEefUXj#9+R4P)U=2=b@q4zgJD2n_mnxJ?L<<A)Wctpa-yP4oD#05;4I
zzGNF=h)8Nj-53!DD@+s-0vnl&q7&{7Z9YO9;=c&1R4uMel6_a4su!fZoQ}jPgnIk1
z7{wy?vH+V6FEB_t)EH3nP}H`g7oLW(fHxDLVY}gSH*`l!z$GyTq8h{h9G<~br_v7|
zvQUd1C4x#E(<?)^D?5_j@^?puc=73dWBB5tFSI0l(uWY$DXFHTJP(pPBzMrh(u!14
zJmlsqK+Ut0yN|tKmUZEPHD4TDoId&z0N<Nj<|ECBy-DqaH|Z_vfrj(t^gcl18lpG!
zJ_J6!O?td4E@({DI^%xv=Rz7sX)%$-BL3#iJwxGU1bqwyrPG`)4LTD1%2h|@Jf4wl
z=RfRz+)a9qRv4Y7ej)`_Hnw<zgPMk4gETS>uZ{}=@ng;X^7MT>y=9|pTfhHn0D_YM
z;ls(+nEFr#gGL*4TKk%AT>(9Jp`1j4r0eHc%<V!BTpA+QB4AOTw_0QIiK!`J5UR);
zxIe!M>ewSa$Bcp#FjU?=_1tVjt+`CGmqE%hpcI!d8X|_NCmK9z-#lAP75x@yle2c@
zJFrOg7I}<HiH^1d5~E)}V24j5)5OyEr;6wFjK*e+r}|Q`RCGL=sm$CWU-uVuX$kkA
zrnQ9y>Ne*1D#Co`<m-ALIDkP-B;xk2<BdgjO5#K{!4MV!t!q5JyEZe-$M#oQM_PW$
zn2ot;2Iu{ND)IKYOT_h*&eZM0Jo{#qw+37&qC>T^-b(5HlJr#8b^IjdHnZpmi)u8N
z`e!j<5M>h$7Z84<Fl#>r0`&=qG#+kq4IzE~%x7KC$TT?z3M*?mes*$UZApN93el>8
z_w<R)Vs8gVEzMcYoQ;dJ)J+!AhG`DoV{xBFj5_kPi_GvrL52`pQ;yl7RR5Le#L!i@
z(_*pKyES4?ep1NgBw<Z4W6`FErAgpNRl)whshs)^0{OIiO9DOBceN27h=E!eI*AjN
zZnc8OXlN6#&w#?LI8!WpWq=RE3kx&VUd9&3D>`YeSGiT!DTx;%i6&CeZ?UO32in^!
zzssp6XT24T%Q4|p(bumt-(qO)PS>3Fg5mhivO{6M%jSAjS(AaZhKb6Y6$iP4w__uf
z*U36eNO)qJI6w-Lq}SvW|FRWT>$)j~)@FJ#+o4Q7s2X@)?^Z@NO`c=LFK+7(G0@H`
ze3uGLO#005S-39?cb>nIl4yCZ11R6ft8i)Jt1Z97tKhb06GU(f<O91za|m8adP{i1
zU{$Z%e*+z7+*19Mj5Gg(jI+}JI~niRoVML(|Eb$4K7dq6;+NpyLChCj{kt#mTZeN{
znVd2PXWsEr5=hofv+!1JVP!2f2#RA5$P#L<d|f-<jh;O>czy%8>i%Nfb}hxr-UZ9x
zi#`$%%Vrbhpx-)Gdw$lnI)CvzglY!enY{n&-+OukLV5!|C<i^ucyx0xe=&K#%9uOR
zeNN|fU^kg6@k~wRex8V&V)eoX3j3Y)wRzg0UtJ*eyM)f;*>~f!HF1)xDI)Lvd6hPr
z2>n;TD`?k%-`4f@R_vCZqUB=oE88d7>DBYBY?L<@fao-U<#8A(0OoX=;jy-I9PT5)
zUgpXWzz&#Rmxpm%Rl4=vD^Z$8c-cBKF$rTq2D+f}7`@b0F@^Mb6`sdyB7&}O;zbtY
zjO6$a;Cb29OL+#118ZibAJ|%Ox|8|ZM5dLXJDjZHvX8DWQQ@9?o8K&<pvr~u?h<yY
zRFlRo#jrNod@f2^6_FMnO}WWz0a21UbbEAn{1jO*$$=WZfe#MPgDo+>UWdwb2l;L8
zr?s`y=&u4z=@`p8A#BkgAxLW=B7c!^rSrQum^f&B>JKS#=2-S7tBlHe>>5HB6wn-_
z41<OwI*G9q2q#MVObHN-k1jq!=Ku!xb)&jwXhgY|^cOU3&gq|xb4@e|=nHLpf7UiZ
zePy#RyQA>Mkv!&+$2z|jNM`F}+PrMSyVb2?8CeFqnW%<<9A<LZ>=M)v@r4beZlgvx
z1Xos9)4|6g(F^LB#h}X=TZ;K-=41<sd9^oRsw~CP^bY1H>7}-#wsi+d<YYJ7Ngl{>
zgM+<EpFsMuyiam<TP<JOOs}uJxnCmlTxpX|5~J9eh1_gTzVs9>%j&RkU05?J6dE%o
z-K~t~QBhEKH6;><;uRDD4an>yhhYa){KJRV&g^irEE<N2i4k=+{<Z2g`<tpk9(2~D
zO!FE^nvC$QB;OP^WSncR<Zs)_?p41{H>#ZDOx<ckfMa$?)>&54*|Sbw(QNINE5MvH
zW#|q!XE7^%hn}z{w_T^ss|qD-t=-h_yDmUoY(_gK>6ZgY%Iv{kTehv^pEBj^EIFq7
zH$IEGNzpXAk!YwRsnG;uqAAV@Cq2raZsIz?aqvPS?_RWqh=^^~(ICdz&IlRa1zZpu
zN6an)nQqD#l2zuF*HY0$z=uXy^wg2_0ZkBaZ4%2@e%L~ior9F4E~>jrE3g=dIBczu
zczuwYg;py0E`^3F;yTLJv^B%{<j}fj+Db|(6>3*|Ha(t{B-Lr3=@kjk$7;_NG{LpX
zUfMM+%EOA+5tc1UqYK2g!NR3o&(DTR6{k#zRRsX{@oI-+u&k6|k6I%=ZTDs1_6?s6
zBql|^I$>Bxl!|*>?H%MGrJ=iX0h(e105Q2~{Xi6pSr#L9epS0!<9#V(kl|WHoeQl>
zm4Wu=F5Vao6s$bf8f#MF=kYwYVO5WuOslK!RWdXX8woH)LpI*7zfyKouw@gW#C5G3
z>3y^+x3rmtglBF)=gtP>y-n@E7T&*X^l2QZs$Q?<@6~D37%r}-^RKR!ix6S!5dQL<
z?<jpD1$$Xys)H4l2+)n9nn{Wf@n;6^uSnu#Gi#YT$7~g~h(vIo!Ob_gZdM`v;-mz(
z<%s#z+C}huZYuIy%q=TiN)ZZu>zl&Vjur4Eiq3fY%sl=u_JFH76Gcj=-EqPN9yZzp
z#Z|ihus&RvsHle3U{R)XW7R5OkfGKrV>w|C-Rp@?acO>9PapAh-Jok<N$QaJ)wjZA
zngZ!v2&4tb9rI?w4)kmTawk<7yr=9_dI`{x#rnM9x7*dyO`FUB!CNTrnte`#-VFR0
zOnV!=T^$?YfJ%8N1)%eFSdas%I;Xq)CiT7tq??~A$tX^GKhTV#Il7#{;27*`f00}X
zJQ)hErY<O8Lv7aNB&?hhc$QwwaB;lR<RNM^dxWt*8)-sGHU>_`6kC1Q69h?Eu3zY!
zEOnsD^HdxK*W&?$fm%o{N%A^|H)?$yOT<fYjaRHSS{79t{u+E=0&RHGc<enmEw32e
z47KIm0II8melCd0VyY+Vn~0Ep-NvR&%TYA*&BL%2(_M2?0c`nW?D$zVluGP*RkryP
z@g@K2%Xj$&)<+sL{7-_!@;}+M3~coOZpQlGb>$RZ+5c%IZ(yOFR7Y+`<V0NMJk5Q6
zhbqpd&vR$3$!=he7|n`%C+>|3!~F&0`S4`t#pBI}fsyq94IR911w05&k(;}lyW0Us
zVxi~$M*ni3!>ra8YZtdNI&i$(n#{)g&9GzgYYZz+1t7MUJs|Y%OMS)Z)#g5!ti4bh
z{x=lN^_&Q$AX-fP#(wT>FOD;UdZqsUs{8qG9v-|ej)y%haL94!-=FM5XvX{)v4P7I
z;XQocxIH+&*X69bnk@%#D}%GwLeB?Q4El41{5CZAQ+|2F70l{bN#V5(fTxiw)S!8d
z+6CO<?eM0OzMNj_KO9-@ejX9j2x4Rn=}ALLolPMEz4=Y**|h_fx1@Fo*1?zfrA(e5
zW^YHGnrM_Vd*Er7)4D!dwMiA>@wG)Y$FMm;h;@By6LNzg__SiD!6x&%oX8VX?<lkM
zhBxgOoDud(V^KnXSuHUlPU9mxY1;gJ4eOtTd|m_Jp*CKHy29o4aY$&mU_mXpCD)Z@
z?KTm;*Oti1KeIMw4@PCzugXm~-PYkV?d|~UQX7^cjvh+gb?$G?!-E46qF%+_m(^+P
z?T!X3_!CIIN@#G?tS#z=LkT6cgsY$A>4n7bT!4rOwnt@GcPtNGth!4~)*|i@r+M{3
zI~$o2<cI}#F`m8!QjX$#!s`O-CL9<E1*HlJ=rYQyxeZ>5_#__Z9Db$g5EamdK*hi{
zXV5glj`$;=k#TOaSz0=WViK9k@`B9SJQA!B#$}b|V>rE~Dm5X-DQ;BieO$pSv#-y=
zs#S3oRh?M<W7{0T7w0@PVWPhtsNVBHcX-thYq7Zg!s$seci%k_JKx&4xd$LD<ZK^4
z7?X~td*$Tdi14B=$|qI&5<8g5m9XRl0BMeec(VNR%WhE2@a&1y_w4s@_HC{T%Nq;H
z;kv1)A}>uD3^rxU3X~T0z~%y-2YT?O)WTU@O<JTbZmKLNT3zgkF2JyAYnAto-mKz}
z?2ammcgf72K|vly5_o63$R-Ct8Ng90CUQ?rWE5*R)^WUBL;<8>9H-z+tv;{aFcgA>
zd5SnWR}8O$uZNmJIJ1`O6X1Z|D#1dGI3Qjo!u_NFJJr5H7BJ0gWHHN-s#{InzZ$3n
zPE9rJfEQ3+Yr8YR5z@si{*I+@5X2&Tzj8r9u>MaGku=QjtA0%C@=sMz@<*QsDM~C0
zwlt!aFY8vi1xI3yKa+?)#yJf=MbBJBr&U!CgE+w;$s;{;a_{5q2PjEJ10jn!OfDD7
zo=vlK3!?y(WTX8lyxt0UM3-s7!6K088ZUEJze-f_7yAV*d!#(myq7=9KbZuAomoN`
z&J-`8e%1-`r67uQ6R5EA6K3fL9%Oo@hMSB%dz{iZPAgkKNYcU3ihu8l8!VQ_{92;z
zT7LUZuvph>_f@Ga%G!gdp|bCOiQ4EObr_LI5BNK*w!SC)Kp%dO3P)tRHRLej`75WS
z<?nz6gw~OF6>!9DYI#akgPdlJQt0yx#NjfgGFCjD60>@v=83-#xwHAETE(JA6V$Yo
z(BOFVY}nuPjtmk{RVi0%7%bKU_M5(H=<NyxTe8|wA1NE}>M$g&R@AyP<54C2)t(WD
zbRDIrn8FsLswZb}944A)8Bv{k)W*|5i?OE6`2-{J{8lTADK@4hsF!51tr|3ZFC-A@
zP!hh4Lov-62<J#0HA2rgQq@2${x?5J9d|TwJ9&WmdbbNIt_U{%w-`mR)BULGipVUA
znVbukWfSN*2lZ<)y^?M=$RV@tC%O}3B#=rouj<rYPURu$8a;Jet3LT1`PMg3ZsR^l
zwPRjvMEV7o_FFiRjT(;wQ(APx)elP2R$wvbGVWa*q^?4Dp#ZleLSaqfSE{lN_C*J%
zidN~!2$<l8=cB=3mI(%wxGQ!xA*2KJJ+`b&0&qV<w^{c~@~>J_vut}bQL}E!*6A74
zE<I-MbJs0<&AcXojd!=S9RJiC7yHq?GxBYE9$>PbUmy0vzpCPCC%gI|pw-|0^4l@;
z0H4Z_<d<py4N<_?;ET=b@JI3)lJ<!*g`L0EFLD0({?93(U8(5X`pY5QWe_t49+`V6
z_&GWFR4U5ht%7X=p5>q(_zFhkgUZ|+KK6x*Txn1ns*3tt#K&He2(JNy0wj=%-&Mjm
z`ms8SIP%%}1)M<L00gqnzF|N!pd0{!rqTY|mv|^oee6cG0}uCiLuOPR8M>uPf1}M<
za9aP9<oyF|f`Ojx-@zu9)yHH01HAf=ttiCQM%&NdFJ6=}P77!&AS+;v3~%qJ)7{oA
z{Pg3)CkOP_uN~IJi>&=4t4GG=vwb0(DJQl!o{7KX!0D6DucyuvupBB&3a5|V@BGI4
z`D#tWiNl3sO%o4WyDBi-;6n&b3HS!845|NqfRS_lu9KYVc|FnBoudF&<h6{z{Z=IM
z7%&64{^jy!V^le~BAiOQXwk)$#rkb4`W)OBCi?W;D&yJjDFsNjqn~E+_BH!pa{pfD
z%1X+d4fnTKb}*685){}RwAzVK2UGxhjhzDxrdW+6P_+<s+RuEbN})m53A>=Fq5yjc
zb#M!b<>M6n@{ejYY9iU_HvOw<vR<SL@7Y{dqbkHMO{F4Cc6AGxrAy0zssaHNl|r>S
zOL6UnWyla#sRwrR#pFfqjTsBn42FH%Qr6v<Q_rhDAuSO!iCdM*fzG0PP0<NcXXP>C
zD50ifh>tmDA~BAoigk3BQrKl2(RGftYQgKemB=3v4yjON7`meNC0dv&1@L0d5_8)p
z2n6U(c1=J@9fIt?4=~{{h%t>9z;knR5r!apCNxr)wz@u%%ckS(k!)2hDw`~NVA0wP
zDWaMNJylK4(?Y{{-Bd*TNK#plwGq=cmXpqHg?017G}AlSH;)0Z`DJKN?wvNSq?W&R
z*J|=KZv_f&5l1zc%6?*kjLJz096DUITff6lYgkxu5B>8nnJ2EksEfKd*~|nCQV~`@
zq~M#5Nl1O)53EIXsuZuF_dql`wC#i_ZFgP9ZfSf|r`4VJ;Z}OJy1AGW8DV?OU01^<
zwQJ80q^c5X`*bpSQ@RePCQuPICS;<{@0)h^+YuXPD`mztQmU%`g$k|OIhzHFAWJGt
zDELmKS#^&z)G3v^(1{ycG#FRgwoKY35g9{#jxm-&TbnW+HRij=P>W~!xJ@f<s>(-P
zjC3a1+V#b)*v!qJoOH_wJ8iypqeitPC?GJEL&X#;zzv(fU@3fh>x~sXyIbpG-xzu8
zTA?uyT{i%*VQvOmStylGm&~6Zqb8z+)mXbLf99BM!#}xNxm0Er^F5tm7r>qHREB`Q
z?N5`M%<$?q32x>~!*B>{EWX}YrplQ6ukljFtLm7`pOeL_ErLP(WAf0E=2tI!_WA|3
z`GKTf>eS1BdW_cC;U>*->jYLg155!9e&gFBH|L9e^AAB|^6_tyn;%hpIQ}q-mud%I
z@Zi}J^g(v>hoa$O#-(G+<Ej5D(CK{>*pE9A2%>$<V_H-+5zX_YUg>3|P~%ZTWI6SS
zc<X*cBC{#c{=S79oYw7DP9m)wT?aemc1zYaOoq$u?b+j@me4n2(_R=btW^|pOI9@e
zGIHf;^})Cv<&bepy^pxfqUpk~gP4AwiAY-hv)l9W=Yn2wR`SUgV4G3g<{w|5{~4yk
zz{vXVIzZDJ(=q>XZu|#z^ONUSL<}4VaGwh3d9H6OA}7*X%H#a^TYe(t+g$E~cv~7S
z21nyf+_y+v)c-6OvZ9rX{JAu?(@snR#gmL+MBc}{R{pP_FQIU1fHL5i$DA5p1s^T|
zIlxV#E;PPwZP~2EwPV|JmNV!XR#Ye-pD)!;zk*?m)DGdi!vXAOVnDbq9SwOX;f}_&
z3j#7^)3Kxg2oVi|lTN2__XaO&b1qhpSIom4W*p4!XjHaXcZ{Q%A~~jbXU#md%!-i4
z-ziS`BL>V_bMt|y^}<S=go%?iUdvI@lV9})d+K!&wlum-R7cI-T}A?fjImm)$m>Wp
z8OxiNi|PoG)!gR8`C3$!J?(u1j0w%C*?^;i-TH%unA^@h<T<$>oTog7$ETo;_pX*S
z38ORj)j7CMc~9JrEt)YbaSATQ)+ozhvpBx6u2FEqIq@h`o2?5iH);=>hqWdZ=J1fZ
zmTQ1w35Y5?>a_>MYJ4t@vtob$dXn-4Rt8fi`Zn5hL~Fk9k!yCY$5!i-96B>%6aq)6
zmfUTT#cVJjP?M6R9JD=sR*)VZEWf_m*RqXSqF55UwuiU>ew)JIKB<zC7z0BCdJH1P
z3J}W6x$LqV4T@}7PLDD-N&IWAa$F%-UD<CStvcYycro}BG7!xZ*ll)GWahdpO0s0i
zJ6PnVd!uoFE2mE8uwxdP&vfYAM&AKGUz4NATGCuf;tDzn5e!5GnJe~JuVlTu*pRDP
zazhGYiSvMbyp`?#@^Sy_FGl5dq*2~kEuP!!FE3|ZA<l|j+{)bFYJabNTg!AB_xMwr
z@PIBu(xSRjMN2?Q&;;Sr<03oEf~fSsY<?=6_~;~siisxpj2Am^9<e1H)Gl@5bdr8S
z6f>UASkce%F>(YWa99^H4C8=d<cN5kxg`rzIufHfY+y^09@Oj=3Y|lxVkzI|bU70m
zO71tv#vks2@qrKUu~NB=YGo%jgShR$3;U!wu+#W5v(O~<IHH%PH8At*1=<aHQOdB^
zJ$tgJfXfL&;;N;*){qPBC%Hubt5<MW3jl>?J9(rgVq#>JYYTTao!Z=d6u&zQEhku=
ztAZ(rC0?VsxIWmKIn{SNdck!6&HPw)R8*o2nxCdmK+3&8zYBcxK@wSW4L3%03e%IO
zRuLZ=i}a!i$qFs4j@a#iW;jb_Ep`2-Y0$S=I0+#cc2C_;xfI~%Fp6je??zn@>QOwD
zvCc+WkL+x=@)#`@81HEk&SI{oXgF>aBI9m_TdL?rUJvSE6iy>?7J+DWP?Mz2^nQnE
zHQn7aq$wnus!sKNhbUFVcM;}FQ(p=vPA@V_9tlR=aO!6~Ilfa9N`gKlu}8;2&Cu|>
z-yl4`Zq;C%ly2*22ZZCy6(ZHe#gxG)*OZrJ1x$^sbB@`03lBx<9rQ_dwb@oR?T*_5
zVgC>&vjju#BluJEXnVE6sMqv|2EPh@C)sOi`nOC^=q6^r`u7A@x!RKz=%v(KbCB%8
zip{n13GCJe^3BD4T)q%{rw?UxT0R1uU9P%PHEAA2o(1si`Dhb*fXf35jIPDHtR=g&
ztSlYVD`J6>y~6W^uPL#Erv?+Rg_6>xY35-BDa#oz`WQf1*IcIV-}V|7Y1}|tMLmoA
zXy%W>)blFJU;{$I^RwUB6FHXu^mwxUL$`wEU$VESGyd!G9KH5{?CmobmRtg(2d9_#
z!B?yd|KE0n&$keTR*TVu1h&}u(~bn;<nIY$l277j+a9)#*9I={E8W=<R^4AfvD3$`
zpf4duy?O0(3j2UbOmp5}=3nlMm^gZ1JwN2_o5NSH9l4EObZ~9p7g&3WzOP}e9mGKG
z&=*b@m&B0U_f%)N*bg34m{H96M2|_x(1-T#`}@}H8JLu6r)|oWGcB_iFZwaPFd0R0
znIETcgTUl8;R#l{?>bXzo3dM(+t_2904uah`{~o#vRp36scj8M7wc<Hl2l*uU(Ry4
ziKJ^s8qe9;&`H#0R;^*{>7&w_cG6a4*=#0d=W)E#{m7Tc+xGX;8$^xTIJ@shqf}dU
z;0X|G0P*<y>Dk^)yUxGG)=zD!8M&ulakjp`cEaPUS)vL~m=-@>dJB*Yr%Bcr`TP-H
z`KP2krZ?tpngQaa+GBvm2vsM+1ZC^bX@yn}*rzNX6D6*9KAU?o5<R+v365W|#f)dK
z$OlVKxW+b0O@UDCwamRF-UqB(Zl^lGF{|n4fWM4uJyO6H`5Biqvy;4acJ9knn4j;S
zB8~$h)XPZd=Cht>-qA`I^#!W&8sWy7_j8|YNJ&y1E`2~1Fr0-(u!DZ*#6pV9a-gur
zMAW?*D+g@^lws#aK>jMNiWX9ic`V$c{u5GN$k@Cg`pXGfc0HDM7HiaCCOul39@0B0
z;+ZIMq<|F))`vp(;1igLNVAks4yxf->G~7lZC3$Q?{mD*5G|GYQ+v}pgWYN&;n2WU
z#4`-GiK@Ceo3IGe&k|rOe&JZ`q88+Cxr-qzEt@H{`IoB#Y3Kar=9g#5oUjQ(bEXj=
zkTl!@X_WpbS|ZHvlO*FBF)UGxKm97-H2I2=hW5a@9T}4yBRrbT2*^A1X=%{1L{rTf
zq4kv%MSbK$5PfR&KF$#f^@FJ`?SonX7G5Huy>G7ZVz_ukzb3XiRf>690&f8tX)I%U
zX$wJwlJp&t^bL|cAJ<?Ss%K29t<7jA?xu?1qlGo#CY)CP1Q=IGLNB%ujzbPd5K|>x
z$5oYeBaCb6W}#oR=41yilch_P(GaYU9utEMBP6u!Xb(LSs2HJOQY&A6NSg&T^rKyC
zaS+8N6<1Ses#o_}2|Cc!HoqgEraxE3zwaQ~8}w6Rg@zk%3;yz>-!6mypcZ(E$G3*H
zKjG9f%#!}KZ{3}R<O$i%0ZdZ>F(~Dg!(nD#fXQVaVNL|I^;@FHex7Pu2+DZKNk0-*
zdY52uxGpgHxqZXS9s*majuYIzfc9L!7MLgMr9&U^ybmZ^=>)0{VA_eOHl!6FD)C7O
zo-!d=zDgWb@D)}gYy&Gk*g%zev$G{zENGc*T(FF0i>UHU3_9k0V8ttXz%uw%ot!o+
z_TEsOZ(Ge%6VcWGXnmsK4%PUQrAWLlQU10A@0=jM;7*zS29x+m2Fv0%Pfea#oZ#=>
zJ}fT00u>cb{*B1_`n^7C<a6m4AIa-o8`8>Ldp<MLdCBYAe$l180*J;nbWm$)v5?QP
z4Mgi+ogcQ|%Q?K`+7%^RuVKSttoo?GYYIOZ2rg;!_v0J|G)j$j8{2Ro3_JM}DS*Xw
zuqAs^H#>ZFP!Q5%`&KQZ^e4{h)@608507aQkM2ccHypawOC28w?nZdxXA?<i50!t4
zm_guE?}0{MpSaHblp49}fkwV}sF}!Rv^%PMxZ#-Xl)=DRA~)4i%5BxFdbbwo1<K+g
zQA80ps)FAlQiOsC9^^2UcbeOoG3v>iM~Z9c)CyO_faR}kw1~{xo;t74XOYY1*KLMg
zW*a-)xe5Mr6w}URf8~J3Y{FVxTdrGzVQ*^c#oY}9Tek--3hJ%-n}IsvLhmXB)WZ5L
zmbT9cIEs>bKuEFxj5_aT#*)A*LYLYPJAZ)^EF}s~Qkded8TAzd7&v8Q8H}7#RJFY5
z247d3%<XcmPvV~x?(7e}m2{2h=-qg^xiobYwTVrSFiYB`V0~}w8*62KlgXM}9FN1F
zjNXYy)#I%RZ8e>2FW04SLN8M_#Jxl{%5R{?kem8{(i--E7}YZW%QnHVrcCS~cIfW^
zvRG3_6!7{9_@z%3^=-S-|7TWfBgbR?{cS*@)nGIcjx&0Wo5bF7C(fgAkchbTVeWkY
zc=P7*;wbvZ9R<A0g148g501Gv;3zW&6D|*1@cy;&^<2YhzpdM`QIp8=-W%s7J4O!N
z-|Hv5qzB60`wsO57V}DIbGxGl0{HWmf`^H$CkA7XLG&SL7Q_FNpZ)9ebyUDCv^j!S
zr-beG^z;7aTA<?^pVFxQ_4t^8rw$x4&>~AUXY~DC>;ChmRQS@y;!E;bu;cZy5W$U?
z(O-ede}kg{!9Oou<qta~BnCT@c{<2m{{|BqZkI_1bQI|8A+|o^yk!n#LfFtEe~A#$
z&Jh>0Nn-9{vCD7g(5>tEY^9wjVOGt!YeS5ld{_tP<Nb3N)-IJ9A(MVH)Xkk|Y^Zo(
zMv(9#4)rjsp1(Yvh1Ctpub{36vvXYJwK^=dZ@!&7$$%9awS;@L%19~?D17`D5{`qX
z3L1s(&8k&7v2wDR$B&ze$~lIY(@khIlu!0hw%rQGQ!AUGikoNR?b9~Ul7AwArybye
z=U<hXGR@{FChvO;axjP#FXRyoBeCx)ygK>*CRx+B06bni`~}z5s1nsU?m;ddFGXe{
z@m0#NXJE+TA+ZigU0IyZmXd?&3_%vao@PJN=+!9jtf635&@f_mkxtTpEb|UR=$U<q
zmm=ZHr7XuOBh#I8-ixqMT9yHBtv!|aO*Sv6gJ1Ne6EUqktRR<)kgweeoS_7Izz;JY
zW$<vQ3<rd!nd4y#2sOLB70U~2B{bD8YGCH8zB#ol0CVqPV&A1ogK4ouHYl<C@(><V
z0#hEB+tlcAkt6W;+)-51WU0e*J=J0AJ~*@I`zoP5@U-13S&i+Nn<a@=_<drkCPW)I
zCypzERFFb`GZra>|5%ah#oN?{N3%NOm>~7IGMT5*p;)Rk8!aUEc9zozjyI!K&Tc!n
z5aUvK*CZDn6g!?43p;XbQgg&1>%i>t0(Ke{Wd>w54TsXjV|h-?@w}+)H|Nz<VV$MD
zf@vCUZdspLIQy(g#G<11<EPaye(B;)@s`ycU8MYcrmotUtT+}T$L90n1l_qh2fUZk
zsO0<kV<?PjG5+c6xrs;nUjHH0?h@(kHMCgKIz1iK(4&bDWkjrv2QS|UduZIfdwSiU
zLC6zqfmApe{kD+SnquHs)$|m=B$0=d#Jp>4ZWmsq5O_zDTK*a;6I}=3d<zykI2swq
z=~Mx8-35=crF(qVKMZ|1f==OCL-5Fb^7!dQXt`n?m5!9>WlE>>T%ZO9(#6tjWsMEp
zrz-vXgse7q97sRW1mcIQFHF>8K$WZp{<@5cR6@0p3&{31JKfZ!)uq&?0#F5NuxHJJ
zOYS?yV7*FV1w<zMDV-MnZTavhV@!g|58D!91tKDVuBh?@XMuvsd7r2R5h0ZcgWVq3
zT3h^KkV9KRWp!Lc!XnqUylPaCDUcIZToijPsMaBs{&2R(><FrCVMzXM^s?4Dm5&aD
z7XiAxPGj}N2(=<PO7Tz`C!hKU7_-4HYl5XD&O~n|CWkhUrWfo&Nlu1(uvC+`>I!{=
z8>Ky^grgfd--8|nv5;dDdhayBvszKXTJf5n;3~b+J;nTHy}?RW#YpBMMxsT|jI!|#
zo#JG`3Jelhl4L(eaYh_tT9EiT3a5az8~nTJUHv>*@FMy}>j~*i5p-I&#?MjmaDR^C
zoLI+Im??Y>_Ja<c^uTqSm+2$oFM>%-<;Q;dBUL#a@y%o(CrU@N`hAoZZTFfB79Qq7
zUHjS<d1sbddFD)kJ)^vMYCilLoz{~X4zI#f9vjx4=37#YV8KFYDY~E@f5R%Fu^|Ye
ztIk-4tF13V;@&EcYg;$wtemo!5@{XPmR%&}qp6`b(}b#A+o|?4d?0hicsnz2&6m0U
zY5EUpz{pPj-~QYGKLdm5|1mJ=SvmftH&@3Fs}<4bxpv3sU7j|Si8xdQ0pBL0rH=-F
zh+=N#CU74-T5FWix_#{PnMdeSQdKQxwZgCu&@kFH^x)zq>ve~6SW(-af|3IgLr@~7
zEKOcV`am(9*0LFDogwkGA<>#LLt9cNj8-MHB!OxO3_d}rK|wPsA|tQH5=*xpDBPm4
zkX#)_Nl96IVJFO}$qR^rjYb=(F&a2Z8W9d3MXX&$i-EaPL)Sqjuv%kTA=wlTmv9Gy
zK0qhx1y};#A(Kvk(o)eiep(k6P?5As0G%KWIiVvh)kp@B5rQezx>24UY@J#d_I%AU
z348I5whOhMxClsMqZwQR#Y)(Ldj87MK1zc?WGX>1F)wx644Wa+Z~=tO;hs!Py0p2x
ze19V)8bz`Gl<<jMegx57G2htDpmoUDXuZg?L_dw68G=aBvIP7|?T?O2!IG%5_;^i!
zZG*SAtn7|ym4O^e#1>1Z#2?F+NaK-657~sz{)Yu@+t|2FEs#jFkj!OTK$mh+A)LoR
zNzj}fcyi^RoZmR1y7e1*jv><eNn>&2k8zmdD6#P@8(~=5NSTQrfoP;vNp0jv@;}gC
zV=d&Dn--o)lZ#A;0Xf>>k4qHz^ARl}6(kt|7U_@p#P1h*5l){<6FW_ZXah-KQKW1G
z2tBoqO0y$T(}LWm!Mp*aUb|!s<(CM+XbmuyRQHUbuH1(WWGiAR2~>Vb)gp;u|8|sl
zHT_LzwpE$PXS(HfA31zs=wLIXBLCs??CHwW+s>J-L)UJd>YLNe#r5j?bxXcu1xhUy
zJ&OK2IDSnG-CT2)j`qlgkQ=>+&&$<*X%FOJf4M-WoU6;Dv(E0DtMg^j<?l@G_H@n0
z!EW1o{d@WQRalCi?yM}DSi!p31&;a7uSzlqUPE9%emr6vw^^JLz6;;9-gGy*O*C+D
zfjWpCSbz!<CmKE8LB7VlU<{Rv{@$;r&YyCq1p;7Uc>4AA7Dw}!W}n0OHbC2o^Ucod
z$Jx!@f%DNCC_B4peVUF=TieIm=jbq8Xk_B>)coCm)C;y%7l=Q(xp`$*(-@|I^wY4R
z2IMCWemZ$cv0SH`+Wq3-`d*pW<DSX0>662}R+o#Eil|nF!$6AY^e5_5kG$^d0Db#z
z)$CP3D6dO&VQG!Qn^LX6_9Euk#$27We7F2Fz$JVaVT~j)TpeK(WDl3Y-GE7yHRM=_
z0Dx6^`$3prtKid&D1A5`Kw8il6A%nzz^eXhp65J!1CRnFpf*BuygmGyVjK^8d>y+W
zX+VV-0Ggq<i#kwHVJOD{fR+X1Fk4d!@@NoJ6R}yY0{CX_xQfrfvG0o`6{<JxojrPd
z=pgMu{f3oDPKMXalD~+HzQPu8WS=`;gi4yelBmuQ_+ou>a?c$&cj>hp?XHd7`;M&i
z_8H`GD5eQv7*!3a919<;Q^*5&g>hxV+eKLvoM#byfG1Y|0401Zc6dM<fTw&6xBV?X
zo1dsE)mFOAS9P5svAZeB@@GcbxWO`r2vAnp(ro-0E8+*Ji`(oI_&(8@;%0D15!IO{
z)nb@s_LO2CJ$Mt?8K}sfkP||mI)K<OS~0Hp9ZyWMKC38>Og(qwP#*eF!ZpYRLTU9>
zc)%#*Ex5|k4{;B;>}ojkiaAyMSHt_rH~r5XP<qMH50T>79!h^3%48N4iTB0;Ms0vG
zI8}hJwOASj5FLa7a7-y$bH2HTrRWbUo!ie>H>JvB2XSx$hXB$IlOTH!#)pf=CmT=E
z?cf~jm4~RMM1~QbI0jmXIC=OD3&8v_i&AO;F&}EqybmF;$cKYOoVlbust$WVe2Mcw
zC;=|cS7+MxVtRZk0y~COiVXjLxSYK3?kAH3QcSg`_BUhuX_Fu;+0Gr!m19s>*2W>0
zQGp$SxbGNe!C-adc~A^|;)cW?)WWJ;!C$ZJ>eo(q8dpLGc#c1x$wvoAyo7KxuCog#
zJc4JDgwbgqai3pUyu>)hJx$e)pUOu}v<^85POMFDS&4Zxi~@;q+3>Kr&}=4CGV9@)
zr>ZmY{t1loicbRLxX|RDem^!N0ZV~O8aOHe*X!eeW&~A$csf`4SUC}P;1_N?E~_GC
zbro{5^0fxU=lOYrgxv1VlW26!j?O4`k&rVsEQ;}GVm=HZZUTDsqi#K>0$y<znYp4M
zQ|7zGe3=QN4dV51)>h7P<~h>a-)2yQjZcVB-Ms|Cm4~nmZZJ8m4>F4b&rc@1`eMM0
zmstzjQxE5)g528Go^`kITldQcFgf(csjGbJ-*nq_sHv-ZZd-U}k<V;)pQj%}&&SiD
zlRr<%*OhONMphS3E>G`3H(cZJ`f2m`J@yE06++4akN;O=;d#E_1T(1e)-L|Gt|Cu*
zp!-ZAo8(iE)lcBQ);bxS1lc71#m`vVNhCM^^0F8wlLM+hB^YE#F~%*IV1*}1KwS&m
zB&>T@uv==(?<qtnX#RKfEZyZ6@uA@>nn~tb=fNk20LK7A17iT>LVrruJp~b_d})`_
zpQVm-0N>1%JUZfehzx|>GNha-KyQrW97d97`|mGTnT!rn8H|l`M&kT|_mcw89Uv;j
zbIdlBgYpOQJ{;J&qBnECF8sTvP6V%fWOHvnQcR~jYK)ep0?a$!*;xQ7ek7jk?rBC!
z;1VrOj4E9zig{m<cqh>x`@x#h0#C4(#Qs2ZzqTY2b9=#`WOe~2&7NAhQu{@R-a-qs
zJ9D|5-vEMYJpbq$=J<zVl$GP(VYa3<#%=dS|M$##Y4EHYP-tw8i8U@0tmhX8i0B3O
z1wIqU!Sj6u8m$av5={&#5PA6U*+!?80a=a+*_G=mTqj)X1~^?HN$t>JB0l%YjSP<8
z45S=!HtFpRrEmKg=b^pMKBG=43vKK-VDD4iD)LV}rJx@o&eJwLz9!)4-dmDGddO2K
z51$g-3my~<80Nf|j|#aT-0<Mf^3Qb#;Z3L(mT;12`3#N3c~^X%PBD(Lf#uGc3l3Be
z9#lRwE(W+hzF#+>(;y@|b%XLXB5n{H{5HQh@UD%98PVXION~C*LLB!N*_r?8ZAH;U
zMt`;nqeZI9T-OE(?>M~TI(*VeQ9nx4@*b8nQ>d?j(J#$|BI>9m#tG>N$1~(%rtLge
z)}nAQSL+4ub)Z0|RGR##<Q^s_Kpv(&`g?D-`(6J{m0P2MlTUH7cZOQ0bh(89ABz-b
z3BVj4L>#IiKDX(u<It^DopXIDP0z!j{qq<+bQ#4#x1y{jfZy*r+?Q8c4a3IS7j>k+
z*J(R_fnF6^T`u(3XE?0La>#^3gH2SDSf4U8Yz)^)TScVItIBRwB8oXg7I)q2Xokbx
z5DyKh1|3&wiN8RAZoMXEBpfS>t2)B9d01I{?w1VD)3E_=j)GDi+Q*t~7gW3knrF`0
z){s>1@#Qq(nIyHz;^@fMyeCvn5y!EgU3anyzFgS_OjDp3D==7E5_B52FAc6ax3OHk
zxHkN<XJctCM?<U{_TMb(%_s9nUe^@0TS-PO3S=POR=NwE<49jp=+;Fjk`)ls+lObq
zsJr3?LTZP+fH1BX&$ur5@jk=YoRz|D_v>NyK+*~NX8vUB!3<v}5bPfWwZ#d_ZW>>M
zhlODynS~IxV#70thiw=>_t*IkGLtpXAjt~d8iNf_rSs7DE12Aq#1e;2XE}~+%Ujd1
zI1bi#*C#zFh%~VTF$hpLXbX9x@OL3WV1YP8t9}q6suwH-Y8lWIUZ+)!go<qYyqv$D
zmX7|l7mhgY#Xt!Lj}VU#=t|KdQP|NU(VgHSTZ6}rSq@X&0uS^gA;(_|=GjrWh`M7B
zP)LHELklTzg)7?@i&T2u$EasS_pgAKcJM>>r8Cl}Q<){CQW)OhDp6dyqgh#1@t$0L
z=_xQ=#xr~-c6}bMSfR*0yP!pifz2hUOkh5k=ZdD*5+Hw?6N)ca4aenp2Y0Tz_M_CX
z7Zn=xGfamr?Hs6HZ5I^vRD(BHfkat~T<MEl%b!Py;}y4|>~BhuL<l8EIX8uZ`^M9o
z>@ch9!1^@h^itHsJRia}X=CP0{s~(WL26rDHGYWDx%U{av^cM9`b$|_jI#m{IO)ZJ
zNO9hRm(<C(j(<`_0~_cP*mQh92G1&aZD<!pRBgPRufJ-T2G@9=;C~@-2wH0AwI*F-
zS6lXDuJw3RpP0zR-6NYt%V6NM7_px1tVwU0jHFp9a*6!pTpW}cCnR-}6{*FTao!(U
zK&^g1ia$mYy{!^>9{2SJOE$y%yCNP^nV;8a8<#gBxoF*r>=#&>3S<vng6;P^f<jXQ
z`CXx;OzW8Mh%CfDBW?J0d^&tTF}(CIyWr_0Z3Z;YiYO^3XV#zF_gHC6(d%l&*s^v`
zGy*!7vnSOEwi%Hlt=osV9I3A_j<5h3n&T%^*nl>+vSeBYu=x`WWoE3D;-4MYV|<vy
zWDfB_pL~umQQ-XUPW$g|dmbm)1^IUZCwS!UF~Hpg*DS{^{hvQwYuH7*os3`>b+cK4
z<^!)N`yxm9l%BSQtlYhr5!bWkQrk05arc@QC|S;uv2ZO9u;C5(+LOg@63ic^a<6;O
z@~W}OW&Ka(E76~sk({t|_tH9qD93iURHv7pXV4j(oyB^k$gP#yYqK8O4ITB1T!ZKP
zz8f#_GJm*UXrkGRU-&k$%kl#t6Q}D9AFjJOrXYZ8n%{B?l4oFwW5+`>TwIthzdiBj
zuVi9Ic2ROEl;0Ap18Mu6o{`P&dU)b-Z2*<y{F<@vDrIAedd60Nomr~krF7oNTf6Cy
z0T8vP?TzN_o%&o)L_K!H1jZ)(FUH;}I+Ji)!%YVrC*84a+qUiev2EKO+qP}nwr$%w
zS?7$SvDd|}Q8zW}sxIoA^PPBJj<iQ;Y>F)!K!FZk^?Z(JJ7HmNi(dDPoc1Xy#s%~v
zv>_xIg;vXyRZG??Ez^RxST`SqgmAL=_lE6ej*cmrc7=}VOVtB3>{IF|usmNJJ?bfH
zUc81*8+SItCZ0qQs4NV^e79uf(dI+s?%a7_eIJ2Z=_k3(JnFpGeYoQrEgK)G;eV0Z
zOmzRV$;ZINPWQi*)1y&@>~Nj$lzZR`#OA0mbR4Xxbj{`mR&*i!dKB++izip?6W8#>
z@{)9Y?c;o;f6PJzdZmekxiw^P<1=_{vbI^Xo!6waL7}y!ueV8hz!AO!2KY@#;qfwp
zPd|qrrbWpU`_Iv72@988-9B(`=B;s=Kr?>5fZ8JIzVE>9{8M2Y-h0$IBEeX*!JT7#
z^xr0&wG|_DFUJi9zVj=-?B4pWjU~D%&V*G&5+OhR`J1*N3bYC@>t<Xc0fh<TWJrB@
zisAmI@t~|l-u1Oyj5OTC(8x6<eD2J=f(phV+aQw`Omk8JjFaaX{%)m8^&G<yR}~2A
zf(-6?fB1Mw+3^(d4NOvdNvGltR-K~o<TqW@)P(5D%R-9dcyz+othmDGwj^v;D*a=Z
zqhP4Y|CudxMO+((RRL1X6|T&LTNlC5jaclQUF4Sg;I+d|LQmox3Zthlt}eaqkAr`Z
z77}bA^UJvD4?U)TI7%R;{tTU*R)0%vv*@s?+v|`Mbt!(SvwuF)*T(9QljI9M$4sen
zI>5)AFdExy=tHcW;fT+-J&2DRDuYp316L?8A>^;yhpqmYx^S=BiZ(HbPrcEG^J4Pf
zVj3NTvJH|`vd5QC4z)PpyjXU|fPm9cjvQ>*5Y5}g<1`k-u+3-d$u^{k-sO;rMKCO{
zj#U2yx+&$;EPz5?X@&>YWi8;3Jaq{%<tp>}WfQ#25!)d8+vSCmGE|^l>^(^Tg#(;0
zWrs$EK;(2dnaE!+HMxdz6s0wY!o@!5gUQc{-=voRT%H(*nBP~;)))Gg{ZV!Zi(NOC
z8!lcuBg%%yJEqC}gu#+`Fs|$I7W%~QtTvS`k+x`HY|Y48@oqTsq|z@FYb2;M#Ej#!
zCOOsomi?PMH)7<WaZzPxbW|H>6(Fe}BdJOief>D{q_cloq7~|Opxn33?q}Z&->V}n
ztBXL-D1ue>Kf@D|AqudwqzQBJdj_(-X4?{lR{E_9i5XH#XVm0&Zw#_ogiXkJvzc0}
zUMAsk6vl+*5oicV&Ecl)zWKr?y<q3i!VQ=<>2A}n8+afFjYs>`l%z8BSvusYWIa@d
z>#F4Hv{)F4lRSlI%w?{z#NX+Yx%ok!TYdXkj^4rmvX6zUYg9Cbu$@nAkbWVmW4sJ>
zGz$}0t`4dwnz3U>C>*V?4BNe9@@WxZjy>(0ZgO}7D?cBxDb4`#U$`2m00f`=)97fM
zZqH9v6`c8DpLr#&@0}~iA8_c2%DzA4dGWR5^$i{+;TRr4R_kQoFcF{Z<K-aBjnf*@
zFaCA92yfYZ*FAXGBV6sNVw!9cmEZkxiTLst`2*MQY%2xkRCP1b5st|vGwr7++toIO
z>yDAnP83~_>kY3gR$Vd5l@6`;(ENI9Vn%?T9GzN<rR5#wh@?uvF-S)bx<!iF{N`(P
z?8i$*FdAZaUwMAlhh-&Ugytd#Z1jgJL9ud$5<zJumabtqDOIe~P-!)Q2-T36ojl$-
z3iR`}j?at-?#t5gZvBiKP+?H5wL6^*B;T~K#;crFKgE&2{P~6ZLq(|kv06D`DKGpl
zpR-O8e@5Q2pz6i==n&8!l=)*jgBvY<I{%?S1TV%uoSo7$zIstYuIGE8D)Kg)wca?w
z^Rxw&d;|UY6oK205%TMRav7mnoCKNH<&IeuW3p#X4yNL1Ze}j1Iyt6mYQ3)E29OsT
zyLd`E)uhA1>$s)GRCOFv7d_}GdslA)$gj0n8VWc%m07BO{V*7$;~TV7>omZ!z7Nrn
z<fiwysqvjq*{YX_7-L<l7=6mJtcBm4mOZsx+k7(Lt5=$A{^W(;%A`zBEo>4&UH=L#
zsLpRrCC=3&$Z=S+#)xwuBQV(c-9<64k)uJsQTJiG<-RY#1ao+^Z46&wtOaA%2)D~%
zs9D0&RABGrWzz^ZE2N^0<0>+GUtL;8`b10c28t3Yv3&Gqi^?-=q{%<q*fG>~gy%3r
z7;nu68x=5#EeD+*bJ2oLTL|wl`#@HUPUf~iw2D+IW-(4@=CnB?jv#3%n<C?X^6n>`
zfLMh@P$fcJ=7(BhN|Eh$!$G6*x=Vh%hMtA~S7S^jzAgArLsx~n?2Toio9K?h7!;yt
zuEyiZyH#cB>+ThGgWLn{v$2F$SwR&aOwoEIx7YO(dCFi!Y}_X{ot`yW<`_PZPGNCX
zVa~P!a&+-Q%kDA!hMy+6F^&8BBj}dsz4u>>m8Ttyr`)@T1XzQedFCoE#gX$3Bh53t
za-Wvfjv!;|cVI2^2BCa((NTv0@keB0a`2nmga?U+-?V2<Q~A@Rv+7zNA&nu>$5?Bl
z@pK+32h_5f>uZja>0XI071DiZVBUamREc<E4QDA#pOV>SP1mj}`AEe^)BQvllVDkt
z^}650Os3Yca7efp@#cvX5MOs(H=hPpNpbPf616;ptt;bh+yg8gwp;j(DR4hTV=g~z
z`(j%+9|U+UbeX+R@o-W=>D65Jf-4#PCz%sSLZp0$o===S&s?@){q;ij!q^T+5a)9w
z+A`&Pe8igUwJd`->xry=*zz+e`(Xd4_!U_5wXAt*ZB2m3`)Y>89b^B_hO>ZeIJNU?
z{(xex_KCfJH!M5vua0D<MLJF)sgj6i<xO!er13%)j~rN&n~3t)3rWIIjHr;t)B4Pt
zEn3f$w&1idiIEbT-#gXOkcYFWovjX1bXBZjkRp#0DK)Btp!;6Z`KkK;7WeLSBr6w`
z2L2WfY$Po|AB3njN?rr5A#_yyAd~ncFzE3u5oXufTdj&VIPTMFpUxfN4cb=#>~<rH
zXUP&hZ^13ap6m3o-e^Hj5%_c)b*VudN%v_2Qx{B#1wGt2zf2E-Nf$ZN@+{hG5a8P@
z9JrkgaV<xf-;eh?rAXM%8)Y&42QaKx1IMF`)z;^pNpqHS+~B}d3*s)_BgByl{|AAo
z^>>XeuF}%4!oR=M2~QHY*4RxTZzriR);E1g;BC`O>E}h@zl_@s<AM-ZV%{1ihf(p4
zAnsJ6-VT~H%($632yduEid50$lZbcLV!TUc46pw(^TIOza}JI)ByZoK>zDpb*X$pa
zIRwF?aysS^mT*kncwo#nI~M(BfqM@406T}}Dz$qdCi4rK^YEq3p?A}C5WESJ@4$cQ
zBg6frvvR@d_frJjYixH|?QhYB?4mR0#rD+uum<X~4O<SN&Dhhs3G(Q@>_z1|)o5WI
zVl{EkH#o{T(TYuGKgR^3l}he!8TY(P6rBhtyFnyz$T}bUj`8-jb3^WIYp#{uVE=|^
z(NydIZ+Rd+{r}r?kutJ2aWuuFr(^kF=7DOH*6Vct*_D0(Ru&FTpvr-mX&Ck0R5@U|
zV6x=Atm*vLTd1a9wc0CyZVJFkk9)lj#hnxj#T5vJPv5k2@S!8iz=Fs?8NLQ~6}Ux4
zf~JN9$E*Bh_d(9XOIAo7jSa6fhK3V$4gOVlnd#@J3iMO``PJkC=SM7ZkFj@~vGf~z
z$m+T;!+Xyo+evL{7dthiLJxK5Y5aCy#3&^3fn_5&`^(H{<74gNim-ipDy*89RSy_e
zjL|gr60Z~d_C5Wj>vf^LatNtAo~B^n8aFS#!w|GnA2hL12-ZIbO$BMm-K<YZ;14?3
zp6P8*K)&dZq#$acljW}nUtLNJYa+{+FS?#Q<%7mx_wux7s8tI3qJd|j*VUrGT=-k0
z=G>(k;H0w?`ttng5;l2~7=nPan$CKP3!98wBxbpc&^O}WFyq7`D`MIr>loD6iI-L&
zPAQc$Vgu`}(MN9d&Q|giB3+u;F--4|vur<Q8+E1@%XRhBjP%1loaI?{Rt)+bLhF%+
z`t8O<uyb^IgmE33Z9@rc5oP_)Oz1KUy=tM>EzF1-<=2COD(4a;D-8M7BL#7Tm88{1
z2yUk%(@%Vr&Et*n!!UlI>*yO*T}(%QMA7OK4VE4^_V|}oMA618z0H0~V+WJQbnX(f
z+qnr0aj=_d3YCj|rDpEK2k?&esWLnx+PU$uJD%0IOqnZ&bFhjOU<H_YdV&qF!=`3K
z&g^;QR>Lilk%n=EA`Z)X#=R?Nju1wyOBr;0nPA#>OsW-cL-?xIJdI1FEN;})X=CK+
zo|H4Bl&`4%KFk!`?%3rfH+I=?LJwvqRcFBj5YwtJ(g+EtOtv8$JFS5iXZyk^nQ=dZ
zC)dYd&H|H7Pq^gFhHR~No6GM^PBMvUQ|5aLxt0?Jeenu^7(xg_bSMdB=M@l$<Tcfl
zXJKv&u%ee391s!rpEr(J;UGiKLaxFUGI}ExyeWvavM&rbc|6)MsgbS;t1ahMm8Olk
z$J6GNZ<kcy0<!nDZpXP5+^e(`MusIGQ|JS6hH1|H^%Rf)oMVjCa%ahdgtJB?N|lsm
zPx7$D6?ZIAdm(LAjmhg*4`;buPYRUHt9cWS6z(h`0DN&WMLCP~#j8rdhp&%H47!+C
zdV&W*lJWobOexU0%R^HK3I3RY;rymbARI$Br4kjc1C0_m)#Cp5&+gRvT|g@OTKgr&
z14-C*AE&(j4USu-_?CrxsW3C4E^cJfvnm6kQYpu+#>E>N%%?d>;LD-<qk#TJR)ege
zr_{`5MS%V$BmU}0^iRp_H07(NLu(^OG~QInbOx>sHov3BG7dk$_D^x4;K?QM`pntc
z&jDre>E2;amJEmQ4ny;;6TMT=TC*rK*IRnK{=8J{6zkmmzJCPI-_^GB-%cEc|9Rpt
zF#azaxs5-NCCX;Ft^bVt2&pCraRc@9?Gb^uw0qb(3#$sRE2Cb1>i{J3wipfGiL}_F
zOhk}gZ}E9Vvc9-R=bK4Kx2t)xzV7|LKs9olrn0<0?10+{Pie(SI|kzuB}~Pezc@ba
zlN8j+Cn>!m5=u_@PkdyzUEP+3<`MZ^dUU)$Z*@qlgYJe}sQ3QKpqaSum7IV`rHG|k
zd}XrcMjOqoW&DHiS4>YRM3M-Sc#Tv+0NYbD_i|&wn4$vGR(j*T!o2e?f-pLdXW_-J
z#u%3szT!#gXYD%PgL0aZ-Ncs^<=sq_dJn)K5}=*qOEgUO7n=Pd3KBV|_FBhVJjGQw
zd&<1thM<MlbS?b@`djTL%%ht@Hp$1}i>Rwg8F4V>i779ZY`;#huK+EMl8wmFcUEAi
zR~5!^Nxvg)2@w@fT?dT2NbavP_F!~bW(0N1J%Qrv2z0fq=Ti9$nZMNCbdpI2YT2se
zteYLu{Tw^Y%T*|Wa2;(nG0@7^&*+QRpaLqDo3c=Z*|u5GhoguvaUcS$s%cF!5u_Ap
zWVvi=MY?5i1@!zwLIDMipu3o<_PIG#_s3bK=vNhgtTOe#kz~yBh#Bqaat4H}jh1U8
zSp}m^F3dv0>@IF(D<X&P&L#;B%dR|a?p$+N^m5B-WzInv$&cF4dspm6tWMM|Hf>PM
zOeSxybnnh@*EbJl+WbILSMT?8@-4i1(gvl71OD}S-tO!#*SBq{Ml9++j<(-#paNr9
z@xheauPfraVjAIWfi`$SIrwn?zRo8n=!>(}=*2u3%KjxmaNY!DI_xPWC>Qy95)4w5
z;rXm{hQZelpHni3{ZV?Ru$PM8M0O#R+nAJwMSZ2hEUlxg!V-TsCZ7LLa8a&4c-rvr
z3-ow+3LrBsP0|koiQ_T{M`8euj}Lv5&tTa@ShuF_<^G|Ce8etO_jlOED4i1>!x}0|
z-n4VY=W_jdcrlv~VQnkxYlWWQOk=;fb%)b>RXv_y2fD;+cO)~9(~CQ7S`-d=?ZECy
zUhCB>VP1m3ogg^X9?9D&3)tOBL#WHEwFt3}y)t@tP0$dNTeOfEqr-_@#^0fZ)aJ_L
z>n=ku)spOc0%@jsV#4txHX|qB+~tqGy2Sce%dgqNkxI#Kggtf!C0w8!*(lbje-w@s
z4bNNgnetA53j5$p?Ig~>TU#Oq<+*d$@HheanaI*LPPKmR$@<)`5v(MiB^D{#O!HK3
ziVzkjydEYZd}Eh_0WQ2SKe`L2<GX{#M1jeh&J72gCquWzeW?3r&oxP>I|?ldEN;`-
z3ktMaG-hDvUuUkGw<|9@d+*zoNd?8oHD%?<3+NCjwMdAmZeFa9q3G<QC9sXFRvWa&
zHp*W`DVb;ND!K8vrDRu4NOW}SFeS~IZmc6gqq#l=Xvr*Gb7eO?DLs~>b8~RRc}>J<
zk#kqC_1M>J?$IMNuS<?Yx|~s)CRgrjo43eOx})jqg=ih>O<y}W&)jS_unQi7@p1a_
zAS3pu>B&zkDN3~@&1cMsO*Jcg2SJh9vhM$Km7oboCnsfR95!dGe*3suOHM%^RVis;
zyj3BZwCL+tCzYQc3eD+5F^9)1z=i-Yx1m}fDT~!$Tl82Bre$*F4W_k<*9@TBwi6#%
zXP6QH)EFnQOoC51QQk=aIYTXLGq&&pXaeJ4-1*LwL#BeQnVG7Q5DkjKvhe{Qjg-gN
zBhJglqyNq-E8M5;iHu0g){@&$e9BDQxwG6U&8>p5!)-fRhFR2>sy0oNaE!_WGHn?f
zGSvWw<PX||?kPOlie!>bclM7Z(x>FZH<U7I7NAqD7Me|saM$MY6tan}X{9dC-^%s_
zhzzdn#KSSjiD~G=7Z96Xcc5Wjlp0z9A2X+gm5#N2;8vPG&ia<t#uN_f+$-Kw%i_c`
zF*r-NffDtq-N_8d+r;*n3$q96saA{%9Mm@RyR^TBC3Z&<Oh)<K;D>+%Yz>d6@ecx1
zJ-%tNO;yw((*`|{{`()~^p=@>umm(Gqmuu8t-U+py6as@V*IZx_oqxdJklsHglMf=
z?@g8-C9>WPF)_~P`rj^U{sRK4mjz(_E}qF<t`e{ch@kIL-dzfsc%BWb?k3gOfaOLV
zFMH+qvZ4pKYNKa89J&qZsGRhGyUe|>U$~Iigs5EP`WW}O>9K)!XLKBv>fsx4;^2)m
zp~p;%BW!xz)?8>7K3>!8e;a&Hp!2_$)z|E1tH`AVm=$*)o+<ss8t0XodS$`>vPy4#
zEWz9=P`=AQXz~loSBXwf!!+Gr=&+S0h9cK|!H})VVE2NhNTz;Xtqe?>N*;#Bc^k*=
z%VBhb>m|gc=X(`qh=IB(V{Di#^~yQo2ckN39?(2LR^fNVku~`B-t4w4DJ!s{v#4%e
zGt}xo8Iv%ucwVQQFA0rKmp&=gNK->-YW$hOz`&|ro-?N=aPK;5_%tfd>_{?uzE#&G
zZiM7X@)-`2kvVa8HV5ODRf@*UW+`X}CCN-w=uDfOit!r*$$zZF@V5SfRJueIy#JU7
z-6dacqyJ8@yG#DUg9i<vH2Ok6nEgQ1@62BV-LF*SSZ0GN`cum&&#a&p$v6nE-RNdG
zgzoYS^=MP~;4ur|F8@TV5t?9F{oHtrBU*(t%#7qi<J}xy)+JWp7OFl2)F)6k$H6CP
z?t@DhQ<Ky^_vHyPEFpdcOk(BN7CD9JBJhW(nO0i!u->!MURre^x9b>l7lL!<R?fS`
zN@D}?EH%<4$pyy{lA12=@pBwvqc*yo%=Y9NxsJKd-MH~*;+Bad5@C>LdAhn7y1-t%
z%i^Y_Kp|?3MBFpiGOf5fb=#Jt4OO*YOnXb@d1X5i$(#xQ$KOynphIF^zdOYF;Rw5{
zEbqZNDe-!)9JIiFaLfJ=#tZ@@h2P*Zw<yJ~;ffujr_&v?{K5#U^;_=w0rTOltcp6o
z6{FKB8>wvw8W{TT1g%SO=Iol$#Y?(@0d)5ty1`|8p49FmVuj~3tGm(xRqYL&PUu_G
zEhb~W+-b)5s}{+QYqd2uzyaY?ooDg-_FQ%099G~}TH0GJMp#OF1F9O}8O=HySSD2M
z_a&*=+h;?>_6nQg1ZQ}BjdUo2l8(W%59IM5yw^yKApb2LWBmWj-2LAIkm+Xy=>NQE
zMa(Q6jqGVfEcG0Xgp3Sq42^hrpd1|Sjr6RbT-P>#{3Cyump_ewGjI)Rg#wfwx|ypc
zkb|Tg8-AJIiHxI|*Kfg+%CZ8A>dPxZxWl@}7^Kl*;idj)e)w*yr|b0N{M_Cj#I>#K
zH;=EB=_@F8KoU53PKDp@=kd+5ibccS#UXxWWbE*$3$4u-H~LK-u;Fh*P~GPPh%uai
zJ3c_SgLC@xz6H+s`(dpBA)Y}2fg}G>@5Pt*w^x36j|LcK_mJxM`AyfW%Y68B5>hPg
z*K2Di;(yR2AV8E&)-R7%KMF9Pl#^k;j?||H$NLklEtWU@cK`~kxkV}_tUpTQoUWA|
ziHs5KT>26fX;D6xy+Bx+?@_|ED114RcsO-S@rXXW!w3?R@amZ;LM!^;SW<KTU(^Qy
z%_2#5%$P?0*Q)3Hm!8x=<X?Z<%KJtYzri3Br`>Vm6dr_aW5cQDwSdlyfhf<x`vFL1
z<@AyCahoJa=XTQaeB5+7l~p#1=ohAO1s+2zq4K&^d3HnGX{A8P5pQY^N~C4uAXj$|
z&JAY;m7IRvQE~?BhDDxyhrMj?PVbE9+OQ0pmm6W8<X)%0>U{otKQpuVFY*B90pIEV
zGWCzYN{xW+sem%&9dNqB$T|$srYE0pa<~$zR7|f=I3Df0s$xdZ7kbYO6}VRz-1a^b
zYAqhzRd{VX+ocL+1bvZJI(JN*@WxuTmX%0tglb7ANRXeDFFmwMFF!nC67qO_azrBC
z+Qy!Dj7ID|6|bsg<Ew)&kFQrtlCj;MDs+Q|^V)GigRD?z@Y9VddvA&;sG6{tKVce>
z`L;ACW}gG9*-RG61ex%eoFH-#0L7Xpf*A5(r4JBI<IBi*pC(?QbI`KZEw0clF19C2
zBAaOusES{+&9eiCh8u;ij-MSC>Rvb&o}TiRz4nY6EX`Eaqevo<GjJ_ZkCEPjMI?4e
zu^=r<B=_|cl3Yw%f)2)fIAYwyf|50_cP~ESb1|AU7%%P%A;R1zHBRUo5*ui?J!1Pn
z;}L^m_SQe5@a!SGV4&BU0Jz-DmfTp9cq;`nsn|ovpx+g?G%HC4qm1H5RuGl9EiG-{
zzALjY{voG(K3s6NiQwp<1qSHSY~dQHO6ExC&;t;Y@8{7lWQ!+6C^>7^^HM*kanB%|
z;KlOdsAq$CLh~kmf5l;$|3L!bu7N%yXK4Z$MtI;E9ZYT@WsU?X+G+ZFnk7t}NHY2|
zzI3t(e^a+xt$sp`=u>6)r$pvqm3sB=l~fsS6KL8mu0v@VCaAz`z;Om%E&`14%+IA&
zD>((owZ}&R!QzXj=VG}x*!>bk3H%U{rzWLsNoQKV;X$GCS&bkW$yHRIkXwHO?7w20
z;jjk{dz15s6>pt~gZQp)InQpI=6;8tg<%vRN-3qsQkpMEjn_GK%Mki0$0TgY{{ZSA
zi$z1n2cTQ!s<%RNm~8eax!1k^djBpbf0^u;wIoN_c^Mdc7bNyq#o26gQ2G*|OID}9
zC}zmC1OzqSmQo(+0W+%ImkE6~8ZUA{F-@J50}WK9`^UT|ymQCEF`wM<l{g`BT5ZA9
z&Jj5&{gMH+9b&dwKZ4AWyZXUS(>4fpywvRzr_+c!)?ZoqToYq@e-Ud&P~i6rPMt%7
z9*3u4$x_u&Li13S5|?`4>MJ^lD0C5#^OmUit)#H0=|8~DBpYIvpU-|QJ3;5Dzm{@e
ztp%m9br2c%MsRH@1<{BUi+m_q%+iPXkTOhN;hL&kzK}x<W7I6^|2^2(;E9c2iVojN
zz&|NGv3@yX2uj7WE)-{9md-U0o}!+T3ewi(&Wu|%m1A)?ku>tQV2sZ+-?`Nnq7T=j
z6sKGI8SXSes+gRZ(Uo&&8w!&A7ox8HSYw%mHhbEFK^?hxijvDKpxi)%A6~?@g#MHe
zCa!5_^C7Up_$|~!k1BJ!eKt`0)x5g|QnF|=*{_ddg-WvsK6RJG@WNZetN-5|8<~F0
zfBuJKBi#>E@qZH$E;UAfmgoQcI0~PDr}gXCwgAC`ShZr~HN*VJQrPh5|FjfV01(U9
z{nY1siAMOf$;#BA)gN)lSG_I`pU3BKKTqe;olJrVYsx&n=YDO+d_cxw!aAYwp#@%j
zYCgY*Cyd=V9Y{)sjwBy$NB!^<BX(NAZ?JbzeEaJcW3u@NQ3JogEwtzE%Ve!i-MV1L
zZ^D#f`s{Pz=iq!gzTb%I`xIOey@RWtw%6b1w`D8e4kGvr+IYLUgD^4pa}+a0^%}0b
zz8AgR-YOF{hT}fO-tbnvzSo$kv+F@R4Z$T0{*l0ej%qJv(rfFsBx04wDj;B1OncV)
zaj*fCh<I~wswI>VhxluR8pb0@tCva61;V)}MVEXP1T9v_MgG-QqWq63P-F9ApIXY}
z5glSks&0W|_LoIdiW)D2%Xv``6c}UqFBB(5s6ZC)Bf_z^PMTLqiBZM07yst|`dGcW
zGlMonk2U??&9!_bVB)P4br^@=#&$i;@55kFh^9x4iuD;|fT-?Za%~L~1ogNy(N3Kv
zp<F!9@HDP%8%PTLOzm33lgV@JO{wvD0|N_MzB0{Wk^^nV?D*;WlMGew7B%ctZc>OJ
zkD)V0GxX~DIK>gj14HG9-M>FQ%TYHG#jk7?gVe<n2io{0egLALRY#!`)&rGnusZIX
z$4sg%(JOT&DqA7al=tME!FXvy$wr-F_r-80O=R91e!39mrE9~u%X8vrRT`k4*-`f6
z-!Y;Isa24Ors^PlReoMT#n!w*_(JJd+l;~0=_?|=eSGm=w|7dk=EdG(aiwNsIL`>p
z8(BLNByhK2hiQoGthcitilH+E>gAs*?8!KoYc?wz_~cc5fH7<P2xV|(i(tnD+-;lT
zGD4EJ0lH9du2rMdtlUqv4Qh0qXcA%kU(Z~y&lB+`<tUGKVZ|y0`#s~|7i>|WR9w9h
zJi%G#;O@v{@pyt-M-vzcw_vk!&Cb+K-`r&OANvVlqKE3~@eyS7V+u>*GkIYOGL2;4
zRDksK;$JToZ(Aihd?i+0YkGrY7&z)G0_E$|3PBxBPiHn5HG$;XT=V-h>@5@SD1*>i
zV$DARERab{@io$~lWNtaU@=CmfPRO{u4ER57X1|&YAevoz$361LRZ3E<Bjv{oyg~$
z*FNHP`M1FDqVN+IXLEG!05jq;tWE^s!+R_o{K*y$s9|-Nt+Hc2$eJ3$eQTmE<M~gj
z4gBc6^;sG^E!$s$XQGbbmM)=Syi;Og!1a#I#zZ_jCW_<pW!mZEgrvl|reZ^h_a@j}
zuNo@?B#geTJa^O3K{tAtDUJRNnN8Yc)%7xnuo^{nuD@w0lWk6oR16l~0*)0P8uWNV
zl^PrF$M@~u-<*7t7ir)zbI?xx&Y9Hxobd+SvFs0s6EGce`ll~uB|S~sCDK~|C`Exc
z6d-6Rh4Kgx6Xll9)wi+Hx`kG`g+9rP(}R>kts>*jGUHwr{K>)NCemr%BXpP)HO5UW
z(C6bAtEALsSne>i7P;V(@0gJguMb?Ay#Xl*K~{DCbVzYciF+KUQ5tgBJI`jIfU&ye
zO*FUVL$|Y3vWsNFbDydMu^I^h1At+@ho4jGz=da<RAyToW=Hoh53;jj?iO)h9RxeA
z!{5xM=}7As$UsHb&?(YEdy6!ne;{HjZOl0517kHmOH@P6YyG}nsVO9g3^$HJii}i%
zgFM15F91Hi6oTen%6q~-H~Q@&a#~F|lJtov9tG#X#+L+oQ(hI*&#>~dsG;6eh>!wd
zTMx@1+j4lyrw~4~;L17mrueYdcq-9zn7AzQ&h)OqIbq~AAZ0QSQ2#a5m|%+nq88B$
zhJog%Q>d`0ql*fuldxp6v-?}lowimXB7DR^vLZu2>{HMb*`8?9w9}FlJW?oXV6%r8
zi}57YIyq_^?Hp}N+v*yg!#iHVxB1d^MnTr#RlWy~;xVH^ipX39pKc!40jI^h|85CV
zka{}4Fcz&0w`;J&sd+J2oN7lx=qk?CZI6=qK7R`@meMy<^yr~QXt=gd-EhZ$85^?8
zSi4$HFdK)5rqhZzVSldZIe35Wyvs6LK$H5h2;|3WXbL>zawP!{dB0bmdnyjyJi>rs
zv(d|!s_f=25IqL>uwRtZ5?5MRmbc<msESk**u}3O5U_>Ay(Jh=JQ6{m<h~|Uja-o-
zZ`l<=B`cb-fXXD#%~3j3ig4`pC{DcPY&!zcLti2XAge4GHJ(<^Ud~@VRR@RF?BU3F
zRuaq_^kf3O3U+HbK=+y7`yL^FA>6rf882PZJ81k$Zx(2YfI8oBl`cM6LKRGlte9P!
zGoZm}?k+i_4k_LPT+MTRr?hs=(>Lk^_7)CS98yFddw4IU0eZdcY@ArW<c?gaByTg+
zZ!^+wfP&sA{Bq`&0lwb%+b$CbiJ&;+1T1Zvz50oq)*NZnQE5nqcwf6XCgh1~{%@6I
zAvj$BXP9)k)74t>66LwiKc4UhfpzAhp4(yt`+^Ypx*sBIK@{-2$^_SITktv1ba)*N
z%~9z(pqB0-;wtv%AneEu$PkK4A3FRnZ=|9ye{S|_WS?#LV2qZ?_L%wBDR1A(?aVJ1
zTe2GkK<JW192A4LxKuokcbm@IZK*m>#gaqGW4Gx`Wx=c`P8IcF3vCqN8t#K5wj$iQ
zb6Md$xso-Xw!cF?9SC%U_$FBiD|~?T8egmbo4ovQEM{h=|6PBvQ~m$vp;UtU2Q+13
z3hAWs|6xh|2YrcPq#4HTjKc{&9BO$>e78^N{=##z+uZ;9d2={9v!h(IW9RzK0>nG>
z0hxxHiOuC}|2y~@w_oSSLFv}%D7=9sn09&_Wc%`Tly8UCHyfxLC?ku@?}B~=wcuHH
z3wzw#lib3HS0~C$s3E0<PHX2_SJtQd+k1wfMrF3J`gG~r-HZ3b<z&!BgmQuH<9<H|
zC+3$CFdULH5aY*(%Zu|%elf>g-H+h&_Qq{X<Q??=wc@+RZh;9xAlbv>L_-me6Z%V?
zUVl(cN4lcL2t<w4BE>s9E8{nPfwyFL9RCDmwI_9Oy2?FNJi%tFndXDM%#7z76B}lN
z9*7cq>4-?zn4tJw1I4pARGvIe5||L9J0k3H+M9efnN6zz{rOb&+VY+f%OxEnMfIQ_
z6!H;5csvq$oF|EVSyE$={@UdDtADl2$_T}bo$UcKx{g)B^u410{zahn)v-|q0WOQ{
zDwww9)W_juFg_V(C!?g^6np{+W7b!C2E-Wozr$1Z`!!(L^8kHKWpk2$(cr6@gh%$Q
z8imr`?3|FuqPh1{NVHLQUus0viBH_Kdv=g@AGNNd1y*2nACeslQ8!7Q-x*G^iB3AS
z8misxC~agh0x;Z;U`jOB_{m4uhy7=J){_##RVApgl7D(Nif5v5#B?N;jJ0&ea`7Cy
zRM`(~5ck^{o*m{g)Gol1)oOEcJHqOgD2IP7D7qy&#H_5~FPAZMkst5+3p)3EF$4!<
zBtLC9ULZWaYCp8hq*e5xiB_|K37Q4RuQu!DCESe|W+t$cfVTcL^_XDERh)|Qh+-{v
zZ5LA(*=KFpdT4|7cfm>^YIlu%kZEDVGFpE<RUZ%54w*<|x^q(u#(?om&8pw3e*{PZ
z3!=`XsiM?$b)BDGJ$R`q5cH3nPFjhVu4kO*)c>pL%>qPyMG2()*;ncpuSv`paUU}U
zv-4Aw*4vFW=Pe;(T~7;;V9>zO)X7m;=YJ+J33(3dR*4Cgui)b?Dq6zb13TJ;QX#gU
zh;l|(0E*@hT7e2F$A-}gP6pMHx|jwpMP9#i(3q1j^XUdl67#SIsXGfOXZzpEB?Sc|
zY-Ss<WFUq&0mk_ANhaDFu?Pew?Y}9-6crUK<F{4%inD$a8~y_fyCPc({M6E|;l*Zv
z4U{aX!XGc(tkTe$DIW<>jDH(7Ga&o*`6*sI>KBIQo?)G2qZ+(*P?KX^EI-@lHN{!_
zCwa^VOvJ`JcvDL0iP(qw!-sW+n<lZQ$_Gsq`78wU-k0+6doUY{ea^OYU6aW0?-J8b
zRDILH4#4B=fI<qM_I53|dpOfY{$Hy`$0Z;21=IDwOj3#Tt#;Ob59M^z1RESu9#o3H
z6zXpvI`=0GE?$WwP0;S4eLJwNG;@RfI3XV3VH%_KO(x3)rQYJAD0=eZ{XmdgS+BRe
z`QC)5y<#yrEA1#6**)Sxnci-_*|G&+hZQRt*%y(xB^|1il24j7a+2?IwPltw$Gh5;
zM8dCGj2^rg8B$&H_Tl7)J&rS%lO>-ety>E_jN~kMo#es;$Amh3<S-4gs#RMb8!WI0
zGJFy#c}?-c0oO5R?T158`382%;29vtw9Lc5qf^%i7)adL{)}AxJ(2ff1-)vQPI;HC
zs@hz*J&<TRZK=$TJHkRatQ_ihUi<{Ok&C)&K5trS3vAF>RD|tUFgWtprh~Ve-D$1)
zMYrFKswpf?!>W)q{Vsp!wd6fo2)ER;WEP<bJ--hjVZtKtvuKgV^fqNRsLX>VXof@P
z3-Dk6iuVv3k1N9!c`qkfU#I5R(R2c+P<5%7iYG3}d+;TKq9Kc<sfsH~_+l#dz4<u4
zuz%0~0yT=*WJa_8<4Ui?i$_dU%(``%hQkWHss&7499YTwn3yfWw2$t;SIghiNL1?^
z=nV$Bi7d9VRO#R4#I_Nsrr}7#f<W;e2W77q`7aYW&Hx0ExDJD11mYCfW?j664)G0M
zUC?Rv-y#Y#`+t$AZ2!B0Wv7OC^f?<`*K+j@E*ZBxK_gHQ5Use8Z>w~^bpShT+xtSu
z>9-G>c_R_09m+B%TR4^ChLCpwRq@Zv@8|yI%j462o?TtvUo~d0Px}8$oPxtj|A<rA
z`DLB?Zr%c=;A%o0vce@#^{>m9bhv8_tZASw5H4U_*Jg-U^vV~1x^LY{QFrakIW_kz
z+J2@E0_6Y1ly_aY@7K5E^a&WrUFjF}t|6an9h+|t%vDwmDB{J3?e48U*af@>MMU>q
zK3Sh{p3mE->R59v$veW+fP;tU?0LF7O_iMF3O+L!`C`793gwK5TY(TwaX;6jI5pjl
zn?YcGIZ#J|uNF47!hdnn{Di_BMa5vhyT(Kfn=$pAO(e)Na3T4phqzXuy>;>bNlz$#
zOi{&T9Yv1_u+G)VEeOh(E_v+wefGTHc)7ANxWaR)YlV=JN3MXJGP>!=k(^e@B@YFd
z{4Fr)A+?E1=A(Z}GWIE;Ba}OWS2#7pi9*c)&-M_aK)_S4Hw81-EQlL!phbY^r(i4Q
z#*+<kkwZX@S4tJKn;VBQBb;fA!fy+~Zwt?D{Jh^6u(tQ_V0Ndji(&0+J(I;k=L-Tt
zh`^pufK3zVQOO}V2dz^?wZ9$Z!*4Gl7{<(Y0oIX0?W7j*DSx+VjT6XC#8rjbrEZb4
zYZ>uz?it}FmI1oErh7JFY#!qf@^f-S)?C;h5!%2$yQzwU(oFK~4~Av*%!(dvJ7JDZ
z1nYg${AbB1O(r2Th4^HIw0ZEI@ND_w)n+DA*?ChilvuMm8tKWr#m2&vE)#F}AE~I8
z%`&i-X>4Gc&wfSqfYvcHQ-(?c9sClRyvR4ICxv$w1Tl!l0A7Z`Bn!Q+!w8bM=UB7D
ztjR*7dry!+fCrefhJ9tGA-e~(qcFqc1Eaak^9}qB%<mZyf=r4gF@7N9YN4mE?^YMV
z>?WW`ycIE@cyQOpA=6&pi6$~P07YX~P_gG$)I=KU(|^+Xb*8t4q}JPO3b=#{lY;`q
z=$@IK^iBEcLsQ(SBPfSyxEY5EKF~dDP~;YNDb{nmz))S}@Z#zu_p@@&2v`svk#%AD
zOwwqJaD+vSFU*}4JBcTt7^14gCl3?PVkKZ%i!n^I2sjAgv~uw_JNFbQY__wU#(<-A
z9oM63-A%zKXUt9_nkL0mmQ!;6@=of}y!&UVP+W=kEJD^z8(7@>w=NSA2&ZP#4b<LZ
zZ(I>P0XKNr7DU?zbrX2qCGfgHUJROA<LENsb?1+K!%D7`Q&qPicuKpvWF>8On*yY6
z)Aacwkli03I1|mgS=KSpZE{?)Uj;68xR-{!YqC6$z-S#;2E1#d-pdD?v~V`64yIyc
zs?=zv$uDRIyMWEZCPh#tmXCVeQ|(Z?TZ-ESk&^gq^spS&F-T(O<z=?JPYL#1rhUpn
z(Kst%r5Q|={6h^YG!n|agrIC>WhDq@*g`V;N_*p=wxX_yXS#daGD03otRtdeBdd~)
zj|nJVBc3SgLAE&Je<J4*gUmV>O*pE}o{tE8MThnpZgqg!G22%CVS6ywl!RB}M+q`u
zt9|hG$1y^(x&^v@isDw=Ox9U1<yfPI-^K+cZC&`^&(hI+EbaNa=T?VS<}#zE1D_O~
zetTXF+}fY)GV@E(<yz<SZ8Hu7Ca&SiOB#=lTnHkkLCHlyeYMpV=4opO=Z8$CorpLV
zs4B}EjQ2aE%Lga~slPJsi70v=4PTgK0n~O4DNLH9?Q{Pw+0c4zEC<fYN1C6^RCgMb
zlTjOzsY)s{EzqYescEX0hnuTGJU6%>&DhZnFl^&lioouDD<7G!i?~u6OKZ};QQ&+s
zscA$oVYK17IGRo<G#|7q@tbB^ua+&=@~Ew5c`VH~3AAf4|MPI^6qd`ysAh0^jKBB8
z#FA|-Ppl{TO%>GCOWf$;WI5!$<+PGZWgrmhP)&+g<6kE+&?gxZ9u>0kI^9~5D(NV9
zmcLW-Z@w)F#DY2dnfo7QGT!HXj+rvAx(#yAGlztW2RC>0p>IY_!+<?PQ0!H*F1v6<
zsPX%yuh{P}e$_WdIS(wS8027=T|T>$2%h;t95P_7Q+?76wu!7~+N^9o1xU<GvUvDm
zTWD&nQ@*-rhnKqW$9|baSp%vdJgPaZkTy!EUda><KUA4kZtj-imJsk~r01KRtUY84
zRro8D>;2yMQ6;Vyn$$9I`15%S94hRPgI1sDRz2(Ev3gDuo=H-5@^Podfqv4t=`P{A
z#i*oY;2?$MVt`0eu>t9Ul7`$OEy_qv)1idUz|q-D5J~eM(@2efB{`R)(VleTa?YHK
zw3#(duB5&>GcL`UXPSJ#{OMsIXvW{H9N0W@v&Vn7WYp=AAyzhy@0KD~nUkL5;zMwx
zXgxr)9*wuab3E55evl#B?u4-1pe>zzg#ypIkY0kqzJeno-l$=Ty7PH^bIS!D27UbF
z(UYvo6od>N=<K2b8`r{wVyni0T|d;wro-rK&ka|3lY4HW-?&@nv7-fQ!M7@jE;=n8
zzR|BUT})MkE7A4lnk$EQy=PsVIwi2ee&RN~cUo$JBLam;Y&WwCVH6W(FgWwKrpAv+
z3i$%*LX+hS!5*Y04j{jyD=%Vwn425SyeIR<;>OB|swt?SvRfy!TxoBU;YdD05==!!
z9ko-63^bGCTl2TW^wYl1+!9m3c6**{A2sAv9<#%nTb*?B$exa=H}&}DXO<_+GfT0s
zG+sW##vxL>TQ}Z$9dRK_{#(w$@?Ute>`ebVp6uaIN&J8XrgK^K1zdzME)OMVmQU5i
z$#(<cfqH~|WWo$C^_w?vtz*PUbyazj-x#h+I7uvQ?@mn9MTc|q)1Ivho2NVOG~*5i
zuDzKSdjys`JI83o{un5M_++cg$%o4{Ql%;%AYwt#n(^^;HyND`(J=wk3c>(<^D>G~
zW&v^N&-OjL8q$H2eXpiXn~aR{3!M--7kc{k{dT%p2fhuWLb>MF<iq?sa|`YpOXHRr
zG$^>Oy9;{{RxoG8z@*W>Zu{EDMF+0CQOQGFBk9M(Ve`3Ki}B7wUUw$%KOPQ3emTPB
zMfEIXVa;{GlRq8~m@L)yqrat<K&4n-Ul;q;!$ijAZ6)c;l7%~|$_$gqMaFKs|H@F6
z8Ktw6AsSWVnn}k0T~(N#jQ_zfSp!Y_sgX+har|=^M<OLe43j}mGXourm>A@iYEcRM
z#}9U_8Kwb~AQRhnq?KPw@ejR$;mmj$lZ+v}cETC<=;w!K&&r+{_YPYkqb6Na0L-*K
zN>rsT>EQCJM~)6PpKsR3o%?USulglGJ8ZAnZS_)GU6o4Cm5Aw-OU|j#x39`PsAbP)
z<+C3AC6gB#j%_n2vVIY}lNK4$R@+DL5Y|6co+qXWWi0A}v$ZlDbs(QA<Pz;*hC;U)
zG^1Z~SQWsM{ASlNbI0O61Rw~D8EQj^D^4N8q#8;=;}?hUW4>^Ws9mpx{`MRGUP90O
z*~|e{RN3Dst{N^Z)*f4Vbl-pkHM|}(7*DHhBq1WN*XrAm45_NhGKuu0M1XS-`iK;O
zEs$`kevA(G9XXUcZ4w}=8xaZ6jt^iO?!Au$BE;fL8ScR{+zO{o^@d9Fg?|zO%hlzY
z;)Xi+g5EEh21ce<@<ehiB1_g2T5vwb8XoLBkt|<xRE5JFyZHW#J1tGbzuP@Fd)P4u
z486+z9RP)~3cpJrA>XM-xnVFVAz%jmj73I@_C&aYqgkHOO1PGefCOg!+n$7suF?=X
z6?ui6(t2LX8}}bM5;EvAUHdP+ux^U}83`Yd9x7W~i5nD8^_ZnIJxhWUIePu_e$1Id
zL;-7OGFD$n+F_x~Twm_$8-EbYplx7kO1$>5FY-3}hO>fs=c=CB;^jp)sicum#`eik
z8%C2j+l$G>3}Rly*Ca+0h_rdotAHg#{NV#~vz6$IOqGL)wo>@`$hsBl3lLjo$<AZ0
z4~UjD^WVw*{pv?zJ6$Vp2*nZAcIZ0o@O{gc5qAjYsCI;Dur0TRY%iCtNudL?rouE^
zWXoQ*O$Efr4XGnWtS94L$z4E7bR7h!HVNn2%)WN>D|Q#XkVPj4V6^L9b5h@No4Dcq
zA-81k$U>Q{RS+M9D<S?PmuwY8*R|!!;1Y|yan%@Z#NtY@Pdw>sy|f0EiA>t%XWqsf
zA;~<Ki;uLQZWW7ofwhDSP)MT*l_B^g1C&AA#TvE`_4+4n+M68g<aND`89NO$S)#o=
zo+><|@P{wDRO7au<Pxfo(%<J{eO5WyPwrUA{=->Ow7~{p#-6c5i#w^#ER4#O{NV3T
z@(dUNTIF8K1df9Z2^!DOG88}mx}j+@-RNypxB3O>sV2%v*X<_;AJXz)8WhBhRCmt&
zQ=Ub&vGqMWROdlWBu2JFIAa7)nW|U-DKHa#Ng?-@bnS^2FJ*Jl+Fdg(utW?t<e2Il
z6^f7CEtf`CIXQv08Z^K5PH;ZU#DrQac)l`S+Ms382OzQ<44l->TyQ$#k?D31AVmRG
zcOcm1{VEO!P@jOo;1vO8f*d*jax0SS62G2T!>naHYS}i7Ll|)rS#-WFvc2q57&?z}
z-}1DR4}LtIxxzd;kXHvb(=ALMNq3dtlAuz`xD8;b<#=$&gl213D%HP8Nr+oXTw<6(
z)}}r0B=0nhZR%#`sOi9}3?*jt*4uPiGN%X|T9OSpscU6f7fm+m24EYbgy$z{j>iMX
zqBuF2?~U*CSQ6R>L*F-(!7D5E&XUz0J2~&#SB8dhb@o>lt$~m`?J4|{G!~^9WBdzB
z?&?*-Rs$mO0q7xE%GGhT9jLPkEvBohmu@Gjlg`vjmk&2W!ND=Bq?yYK6i5=K)teT@
zjuTfoh3;v$n_vdR9#lrtOSY{gljfO2&==18S)ZJNXMN!Vq_eWFojP>EFH`=Q`$$#l
zt4Q}X#E<Lp(ym^fQOMz$&H)zCdlEFrN7MQR<;<30k%~$VkaG#a!tfUEG+rJBLW>jB
z!v<){*zfG^<i}cn*CoIDR_n!SQ<*)*I-=zHqxPem#0BZm`xNC75>xCqv>2Ec6j4jT
zT=UxUe)GS-uI{)Vgtctirh+V=GZ8cj(k&>=b7uWXGD(3gEM(0E%CBs&j23>0bvf!M
zi)IZ5$H`scpq<$G#siJO+IO2T*@(}O&1c#XP4RZMB1s_hn|n-iY^CACqWkpX7W$VW
zqe$dnB5-G<(|%%Visaf+oG(qzvbJ%d3x>m_ln#C!!08VaXOp%H`><k!&j%Iz1<q_w
z<HtB!1hw^mu@)^)7$6d~WLXkR`O*4Z4&b*hIf5m=1N`wIrm*NH9{h$z8wg_fZ$X~*
zf8swF7+C*zrRu?-Ab<a#!om+UUV>Vp043+o&a?vTLnWpY=7bp?cQKDt_qSfjV%-Is
z%*EUB4pmsc)`6&o{mV37h|czu&d=JWIaY-&s0yr(4_ptgm^avHc<^s%yoli4^Xgq(
zTm_2$<y6`P1&b~!ALkF5Vi(w`YoISMSFoK2>%VtC4SUdipQIK%bKi@^m4)B;Mk&Fn
z1;EHf(8+wK57X7#TZSkv$+GqVhzFIkl5-g0{%%NSshA~rNDQ1mZ$?q;2@ywm6t@0A
z1}M8>f1{HT<pvTQFW`aQ9WtHn+Mqaey@~$W=p`Tx0`53x2aErk9<jn@SpSjyz&$1_
zj^O*jzVV*l+I+T_YV$K<KC!u`9mW^AaH_>`OW4HlD)!D@A^O#{flSP%I`#aZf&w}`
zKGEkFBEx}%9nggjILWZmy>pRJ?Rc&`la171W8GJw;y5eSuQ0rhGl;H-XGS-9*>o+9
z3qkeKqgC}Df-*Z=vtz^1fEhlJ=m>5M9j6e#4sxlFsgQ|{+_hAwg!cES;v@27+_uOT
z3Wu>YMxgRGkWVE8%pXO<7<DZd1@h5A$1gJ4FLDMdh|P3@+D>}kkgO1Ne6D%5dv$(3
zS8g{LCdy60f4`kmfB2>mMm`gQXB=`pJ$HxjZvPlr85`7N{d~Fr!>fDUo#<4mHe7E%
z)vX~Th>dGsuPJpyDNP$mxR3=PZV<`wouSQGCWhL8$qDq@%HEHE><%F)yMpN#)DQ$+
z8DYhN{>d*=WrN&qH^ne_4F)h3TaL4oUsMuBNDxe2&C})r2Ws5KhQQ79?XmQYnGQe(
zP<uq*^K}-bfl2P^%1#g{Ct5i+(+-CRld<!f7aY4B!L}bvBF^fme;x$NQ@d%$&B>@o
zp1$I;r^|B_W8jt-(QJP7I@Fa=N7oQ~Cr~MiJqh-lH?BPo__WE^`L4d+LE!=dcLZ%|
za0BJ#5|y|Rd6L6K@UK6KT$*DdUs+AMkVlOZoy-Oy9{ntk>O4J6P3A>7s#Qk9hl#|W
zhlQQzIH-#Yx;Ns}v6=%Q$y7F5OS{|K!pX^`%a*9z`=Wa#dm95l81jpmc2Sz-<Gro9
zYOxxX3yt1!m6;${=+|kQUSHKG!e^hzr+}fCjr{C`&w~g4x7nXt?=AM;ABksS+B5%q
zg7qz$>RAycJ!&=@ZtgnkJ4tRmpC$_JGoT9N_Zl(p5P;3usWm8)ZYbF1iSSc}L8?Q7
zr}6L3!MCH=bk-HW_@pnEN9(gAyaZ#JLL2ES3m;g$o}Gqur@uiRK8^b8tlVIr+@zfQ
zpLq-urV@*_u@6jRbQ-5YlYoA;fts6Oipglp8_A^o4TT6w^?;Lx(E64^i%pinxk`W{
zEZ)3*YemOOPL7cEAz(Ylmm$M+LiSW{CE?b~Bm54=m&!_V`jFwNbf08x2|}HQLDxNO
zj0M{!W|fuj*Xl)-e;F?3b?o5`=Pta}6-U^*qkGrBL&tsPb?c^CMUT24hE6QCv#O-l
z^0^$EI);rEbdZ;?V^LCRi2Wr^yy<GtjyXdWw{cQ$eDxr;Ap;1<vbd7(_~4CEYK!uo
zB|V<YtrY7xQ)~`+cd^S(ojvO7P_`YqjOJr&zx!&yfJ%k^+Uw!t-P`^-C~mrqmpZJq
zJuD~gUoa*_3d_uDBP?HFWdn+eEuDbf_rX1FB!CsYKWMrBEMe^lT5U^!GNFCdx)VWC
zs|EGF*3w?ERHQHJ1xW2ltXczlaTRaJ3<?@*#NefK7#}Igha`^4VO{7isaz{xacs@Z
z^hVtDEc?L!gRyrC@+AJYJ<IB{ZQHhO+wQV$+g-NRW!tX$m8;9Pt*H}p=iZ4p|A{#-
znfa8FFW=ZJ*WUZH5+S%a;JrY=Xc()CX+N@oiteRe<F#Wly(8tR<>&Rlt&{(Vc$o)0
zSs8LkoVF&HioKh5s(}eKE4{f|m#y??6Emc+@ntwi8GR}QI&nLsNPr~88P})ufLqT7
z!#51v&|grxq0DPQtl){|^CX{x;dFM6enzig1gHjyL!KQ-d1Pyy&HyEm=mxa1d>Q3U
zq!Xo?8%{%)<bcGu0xu)q_(|4anildi29$vzL~K*_gATY$-^7*4JjmPglX}FfP>x=a
zhf1Hg{j<_IsLW3aeZmriA{eQ=4G_kR;k{?|eFJ*~Bben4nGtzfN_?o`CG)qV{sE)-
zxPIl7u7~R2=Ic%zSZ4#4Z=9w9juTmbSr@&J_Vg_w1a(=#1F;EwN@4~R6%$upZsqyc
zBUPeF!@v@AC5_d|n6gOQqk+2(Epe5faif{^kE;2FU%fKOIUJ|R$MzL4w`g!#QRdPM
zj98Km&{ha`HUB~F${U!wq~ogFp~R>vH{&RsqwI~EVz;@$0H>g-XXQ?yN4_gHx18U3
zA&yKR;FTr0!Ek8A{{^yOZeZD5AybUf4Ej^KYJ3`bZ&-BCTbu~bLCYN)If6}&2G<IO
znN>u><=qt(0w+Q5?)^PlL4J*T75o_}Sy=QUXMsOqTe}bO3AU{HMRNjt2Ikh#pm^;4
z`dG@S&2POM+FDyiX#lAK6MKUT(H&D4j3kylb|`Dphlko-&-y&|=o-}}H(*qjyN6eW
zL-CTS<%<BJ8V6<|vA$_z^jY9$5kh$@+4)0{+_;#DMH2zxnWRJ^d=j3rsdcg=59zAG
zchZuyiABU|y=M^)1Bs01D7_^Up_w=>*V8)vU}ObhMd&|N$%#*53sW()KB~q;!0q!W
z7_kexyAO+E6Lwb0<v_K>ojAq5;oqTzMzjFs`Ezfs3S%zd*d!cmSJoWb#@BTSDK5AU
zNKEyTzU`!WOxZw6?MVI(kmf7#5G}bf<;_=icl)mkDnj`p-Y{fpN8=k&>KkWgy0mZh
zrc4R{)K#;xeS>$+qMkTcrLy~#k7C8K)#b7386>ciV96Ed>g~?-2njBfdI*18>q#6n
z+*k#GcCyI~)!iQ6y)HzcqbS!~C%1Ynk5wz!!_Ss`a&&x%TCVC$t4q`NoU3>9qZ^qu
zH$~Ghp60@e!y7Ui@na%6hVAmw?kSC91;*KD;32vw>FV?8@kzisnZ5EqX{Dcj%YT~@
z<YM}t86wj<lX1A6aGTE>iwH5aN{hcRVHW@_CWJgVTi{x{n137Vga2iT`3WpGlB&Bt
z0^FGe72K=HqKYNY5|qbJ-${S+b&eT25$v%kD7uv3OBuGw#RMeZ>PXqNS-46<maO0I
zoZmKzQVdTYW&cYtd2zj6oNn^+Ssqq|7aE{8Q$9ZID5f%DpOhGXQ&<q(dvG4K@aF2L
zS%NVlRv_w$^;X?HJTv%jjt?*%&v2uPHae9dC`ga)hLy7LHyoVcAcIPxV5|L!*}jR@
zw!P4K$I^QnR?i~;x;a{^^wLFKr<?&o+&I~n{UiN>E5=sL%zM!Cuz8@7d8ufwZ2%?6
z5J8_uC1uI(3AP`~Bty0)uk6gh&juEV{$gti;hA!a;tG`DHa$gK$f(|`G##<`4gl>Q
zAGP(Ay5Xb4Dh-W+5uoh$r;6g(1Zi8IOmnPgru`=JpDnVUeW~I?1%EY#NQF>8<5R1~
zUE{tkP0DNo8nCi77hy3b>=;4CC*EXD{>%5>;NUiPwBcdjo%yFMZ8hBwQVot=wX#(D
zKz??<yuFaE-O~4HQQfrsO`iTP1@EeTT=(%fYUUm@`!(-1R|jULc0%s@{p2O|zQOn|
zR$SZX!mHgxcWo-w?m-WbCPhETX>+M-ui0+hVc9^op}88AG|=AvH<|8W&h?~v+2XY_
zhl0<H?~ZM(rrS>TTjsPCi<cBCA=kOgfk#t{CP;rwlqR9DEkEJFC<)~Buz4OpfVz@i
zlP7?G9Q{-T)MpWpjGz061PXIf6LF)O-Xprw@rWDiE&b~p1NM*av9SGbDQ<vVY3{L<
zxAIm_<{E;x+C}nct7f(0@y%NRqQI#&X!sWMh<&gI_4X*Nj?Q`c8iWdO$OvuYRHy*Q
z2F{0POXphNN>@J^121b=BC+~f0<uPFv~;DIEhoKEl=>2d4lOPx_ur*(LR{_{t@T2E
zd2t`YdniQoB7-OiFq})89$(XK{oV^dgG$Ki1mLnd0ry+TMd=eupZO=lw|xRY>n4SK
zGiE(%QrLmNxcMxOHa~_SCMh&=4cW<DRp7-2PIgQq@$Wq?M4wZTD<T?H)cfq>RWh*@
zFgmz0k6apzN27!8FB#$=Cy;0Cq?vGW#6pFnZYZ&+Kw5$Nj7$kv8zM?RU*FP{r>lA-
z&73x*Am~23kTiI?E^%O385DU=uz5#`xq_iUk|-B8`W&gQ%o6oC@UI{S4n$Vd+2~Lu
zH@(xaj}2N_l0pa?qCYQ(*6xI`m&)#{JW5d;lXy#4<N3`UWn0}y588?k=U&f3)k7#+
zYI*FDlXUu{oF(X$>2+1ImSUc`<;~^f=l{SYkNmB!3QFRk*u=G!!DjL!3x%M#lE<B_
zw1hO*00>;q%ynE?<&Gku*)S-9hXc9;=}6K^kvPTRcd-oXQ)rS?J@fn>uSIzRVH*|_
zbWK-Xw+Ato`(IDvY*3d+pU*K=8>1HN@9S4y)cb>sn3_`C-VtT!8F|O-ZhdTA3aj+W
znt?<&?fnlK3?{~K_eaGnh3leh`Bd1BKeQoiJj>+NJ|Q_m!`NK~V78McMxjllMaC#v
z$S$ZhObS{owxY48q~#6m0$N%x1RFm#4`Wp<kG;N#NBwop5cc%z8kj;j2lY7Hs|jeu
z=)*RmmLhO%!?vkE%wf^*f;q1s?|fEWrcUy^e6jAuFVRR`S(oq;+4ufD#1^R60b^Y8
z=U2M(?B|<hD|p%+_~mxa&Txu{RbD)bhp6_DmDzjbiidV?4EYmsHaO+AobPn~u*O#=
zUq5&|Y6e}0YcN%&J><)6=n88PwcLR{Zz8qdql${wfjfaqD^8OYjnFrlIdZ7d5;Npu
z8yeK>e<mg_Bq|fk;;>vFKYzb}MkRwqK|Hb==4bN)L)b0Ja^;bs#W_5cxIlSYvhK9i
zCZmf?;x-gJp?nt=WCV_BJWNfHaf{%f-OS%d^+l=_$w#?aFV29VtB3Qe8J$6_m=Z9P
zbZu}2Erfk2+<G!_OK|>}L$io2vq}o>BpDexbZCuUwE<xmk6DOr@LQ!E_g-zO5;rLT
zjQPz+Exb8=P_P_L^|C(Xy?TFX?J_I+&}Y$+m_57&Zt+`)12q+a|GNEfU<A!DPz~(z
zBSN;8f(pBhHfU||S{3X)5_A-M__l-c!8P3_s*F+n9eak&J4JeiUwiqsjZv=?>>cr2
z@9CD`uU?ngIet2VK5F)}t(?Of_jK}(EZci_ZgoD<S<7xc&3hLd2RdMpA!&NRx%v%T
z#+gjYBB<eOk$0)h)ZeZc({abY@93P9x0XH{8fG*DCB)4yZEgFuvAom22H^rW8iWZ&
zg-)Pwg`i*u%{t|3)b#hv8S{?uHnD4WBI^HMRaPH6n|j63Qxm_X67Qj<MFa{z>yBL5
ziK8JnJ0Hadf9-9S6^<9^16^w{09}G}f670C5ZW-pY**+5O_bW?_Ezc(?&OMilXXMA
z_8mP(a{FZ~U+cL*z}$c5C^sq^j_E-2Qy$uqR-H1yG{4_mUvGciJ#wsDlwN;F>QM>g
zKgfX#d`=nUU{DY3=MlViQrsf>4iO4jN27=-zscjB^)3G}p@?dr%shH*K9O|BkN6gz
zco5{ZBBSxn!JJk0iIi<?Tzd$7O%APdiN_aLBescCnc4kHZe3MBunj(+*Hgtmo!5P{
z-14H-W5Lszb^ky^%J|!8tA%XpRcFm)*2t+t6BJiUL(7kN57@+ZH`O48Jfcgf3AR%W
z1KOy~YRMgcpBC53iT*&L<Q{Wc3FbkPQ?!qosB`XD35(t51=Z$%FJFJWp3gF0@x;EZ
zc3W^C5NqfEedy;x@`-3tp0N6`<|}Zi>+U?$imEEesaXEA{1J^*KqVlDC!^I8uGOYs
zK_f~9gX|F|%!zo03e4<V9>u9FUd@at1628&(MtgR0de7QZN0FQX2rTz-lh9PvlcWW
z9wVc{wS+nIyw+r;s$n)UtmbBzBdSiX-_AfsYu>S{sT56lj6+vi9f8x!LKT$gA~deW
zsT3t0wfT{0$|(x=4EHd6ui#go#3&sB95224@1hfprs>)!me04#2C{x$n*rh04{LtY
z)~I~TvDub970ahcN*BnBbFnpD%i+stG|WLpC+rSw`@r$_n8Sl1B2Z!V+%CmS2PT!S
zC!1sj%hoFby$Z~ozp45JE)j<|nv1lB2b}0<9Ir4@%-qATN$)5{lF!^XN~gwN^nVDv
z4UK5O_Zdrj@3Kr8e|5JSplj@nstKGvLp4bu!|$K}^`+y#$|irq0I16?n`=>@FpPpE
z+LYuO$u}^!u1@l4+I5y+uK)C@=7Y~bQNtCd@zoL77Nys|6fZ|_AgQ{1E)Emoow+e<
z7#|;<W8}L!<_tDxOj>;DmmOW^vX2dJhS11P2HJ*Y$B#`heQIxK>gLPT8x8KK$v^$_
z=iuw4L#lGEI-5GR#c%{(qImGM;D{gx!3RYqMGg~jU(1akCxj?PokiAHc2HW^DSjW7
zS*cIg14U27pUzB2-NwvP==w~4Pvw_Fxk-(_CLRsxaR#YR+$$|%ZY1Gi)x-uwmx)gB
z{4-ws^KQed6-6O!`%od_T}>_v6FrQL&$aFoQO#F7KZ>UAs}DNJa-sWksh>7{F#xRA
z6jvkY&3KKN)Da!2u|(O=;-Oo>I}`DPixou^kYU6t%;e+cW=GQ5fqSI+`Irobrta_k
z$rABz(4{hht7!ke>uP&C@7{$fCOt6<-mh$9OtYpkUJyY3<JwYK@I@s~F9u)0hMupr
zvY@5l-2q-~Tg9Z#%>i2J4#LQIJ7bie;&^yw)K<i}x!KnydnlbOp@=PjyernSw*3Wp
z^(%n>Kgo42=Kn6&8M)a0M=L$kSu*xN{mX#sSF|tC`6#bnny?8|<ltWC#I1#8g`Gcn
znBU(8DUbD{Yq{asatOH|tn5-ssE?6uFAu(Z0DFX4dxSjwl1y4>?COlI9=QMkhYzs0
zSaAjP04!+0M*QjQZ1%6<Q5O6tXR6zJkE@$Hem-f3A7hgtfDWikqKa^n1RNi+RzT~X
zYl~s#FWpvrxDb}UB~o#|Zr}Fq>BGv#H3Bcz)E|yr){la1{Vn`2-~1yRiwsutxA#$z
z7a=Lbg~)Lk^OQWlFBV_#mtbkS4$duP0oFV}zdXswl>@=q`M`geH8_wU5_Og1n6suY
zRNRWbySgUfOEmbaLTI?aaufpH+?uK7obaKg$ZQ=N1){%;P)mb+UrpFMt%Mq*MSuU9
zPm)H>=KO?+9ARV(k_L0_6^*7d<c>6eAf+~^bwP$pjEK|XH;<H)p<{|~c9S(mOx1_v
z(I7Q|uZ~tCgHu_?!&0jmVzgqD*M=i%C>$c>^2sZb3m@y`+sigJ<XeIe<Ta(;f^iDk
zNzpO}EJ!xVj?a*QBqw*QhKt*zQHEIHlOaXo0&;zzk?pRU6@!fEFKiFa-%^e6iz0Np
zr<|J(A0I3^r%3!e+0+^7hQMv?k}WfuTX2@%ExH{zXBzHT`>HYIS$Mf4?R;UgEocBS
z9JRfCh^84mbM3jk)!0P9C@BHW<N8{TW(nKIW|9d5da#K5&I`!gatj@16tUw85T+Q6
zfiOoyg(d<Iua&5!Z3Fl&QlZDU?Iel6rdeOitakprb7>~6Fm;Ol`rW|pfr}F27pa>p
z0qSFlg*pd$P3L9>Q4pS=p}&j~LXW6BJY1E%ez-_i=BLXgVZB$V&A?DeTc$9$-i$G3
z+hEx%Y{uog6+DKl`%YM$n0lt@OAnBl0x%vW%29!c0JNw>1@zh`jq-RlxdWJ2XQ{x5
z$O0H&F{+T`iu4czc)*T**r=D76N4t59#mj<wZwZgju!{gQ!SPg0Fx^1+r{eqWy7Ou
zBz00n{L?E3H@#@PB}-F+;WvT5N&*cAZnUk#@c0s1%0*pyq^M>EU0Y)7z1QIoBR8r2
zyt$WziLC6M`#BM8=rOtX`+%*%dRX9EC*T8&j6XSlwSpiB8X52MLnqI7`!+zq1xiNb
zK^iKP5ICG!G|cMIKTTP4PbBUPVGdk--QovP;9*0)?Ca*f<Usu!c1PxrnCHP7^Yyr9
z0RW{_BSaw80}<jIjW;jdoiC8i$~a{{+=no9W2EVxBKsa6ZdT8~uy5b3SWbUnqz&Fw
zClZTCeP#zW@Xa2+S$E>hNStU9N)c)maAvcC6Ag>indt}<S^TNXGRxtiWyUNM-!w|B
zr9<`%Ntz8hF40W#E?)?H7ukYu*7C#GUq}^aJ?vO(pWfFzO4oCfth^;Sy1p1UHz4D;
zGblkVjDIVSO&##9a(oJASBH7-P;V1qSKlWrb4YF{Z37joMSGbcT%qOxI*54OCtQKy
z{doulg-SAve{ld3B7Q*fG+@W11Hpq9a78S`J1!ow8s(yLT(3>P+GQnnNC!l$#P6{p
zpTB1>Q{)26C+mk;%YPD{_$vvdaIVipi!8T$M_D2tdo(XY`8U9USEig9ko*p#O!!w!
zlY+1vL?=^0(^_7s%9h?b(>|0MWpr4n%uHj}NUTPG*!7B$$2O*2W_@)^XBEMo)&sIc
zPBl5W{^q4}km_W7maIpsuCQdBvD$T~xP|uXG8@Vn1cAcY)TPE?a$zkt%_UCjxMRM<
z`cn46YJs!7q<hmOoB6<i_N^{F1!CC4!eq;Ah?1IrxkPY7tXMTeaWW>uS3h2eToXfr
zmF_eLwH`KKGSRhQF~zA;63KFplYF^croP?h*E!j$@?wq2MRswAw#J&mxETe~s?T53
zc_a_s=E;`ld)P5=gcVJlOZqFi3mBjeHNTxstbw0wU(N*f$RYO;xx!1Jt>UM6P>Mtk
zAV+1K(Zu98#|*BDzFTRM)HVa_VtIieBLo-kpjFV#TuQg?YgDDv2Z5Cfs*%;?xITWL
zTGB$`H`EOLE`u{&sv7Vk*z1eX$Jejs4^@og5$Uk=;p%qH!?UuKm9f>JBh3KREbuYK
zRmtE>bZD&c{X*`DEzY5~31}K7QJD&cgT9jLjOdssj4kwARS8*wWjnBab&mDN)b;5$
zx2Y%~bBWHaMZ3}F{dL%N_VwvhUjad$+TfkdxTnjWPRGlgj08W4&b39qp)iI7X*74|
ziFvk>nXHb8l*{*DuQgF0r6R!#zZ#ZZl<|XCTTFW(x~E5K!q2?+Te^1Z?>P6=PjZ}D
zF;KnW^4y7fLA)G?{)Cg{rJ_V$!*+{q0ME%>5`^m$J3@Jp6e{DqGAbzx#0lCL%TqfB
zc){6S4MreE;;4xROU=0Gho>J=nPAU*dTnr$7v$u{mve?4d*-ymT)73LwCQ*HB`goh
zOAi)b3|{$#z}jQHo(X__JwsJoiH(6Hx8^B_@0~CNiYD0QUV<UNgKNk3WTZYohbyiZ
z%!l>dxAFIx4Ye>Hus@!nfptI%J2;;RxeasazJ}gvFb_GI(T=92#R}_uG<SQ`%a@h3
zjMZ%o%OZ0`dovA%K86@r>$Q96Yo9yiO25ti?FMuC?C7`ctcsu+Cpx^ISboxJ4}Pie
zLSReBiuMEa9wGkLTv>mhQgEH3zQ<GCF|O*0rDoLeLPqx#pM~$ZVYZK=L2)W3vF$9w
zBK+`AAmyO$<$aCNax~~1C=v1?4k6Ebz`%MC(A8%v`VZoR<%cT$zrDi$?^Gw&|C8#(
z_CNCxm$ZM|<F+Dw)f)82pJLE+*^>tM1t70jMWHj&VptNN3~F(aTC3#{<p3XVKV#*}
zcTE0ZzFlBw2h+^mEw~0fJ<Y>x4%dZYCfFdYzp{kc-GUqkv&)6bA&PK^V0+mtE5Cq3
z-Z~8%&#G%e7fyiUNmXz5;loS!LPD~w9V^3RQ|c8#3vgqwM9(j<nfH~+0+DUx0}Z0N
z!8e;ufhoI`bOVZ!ozTih<=fDc2rH37q~!yX$jT@c-OR|?6qm#VA@>gK1O8-#`>Zx=
zhDOAAkJJh-3zj^TL1d9?kfrxUR;jQ#(DyJa8UBMM#{%=qpp4kFL4-`Jb+F5iAp(B4
z`2Cea=Wu$Sq_RjQS$>9$8K!W6%@U}*g2_H0|E2;xw{+MTR9Q>WJ}F5EicsGnfOgaV
z-KQbwDq{p{e5MVpPJfdr^-Fmi11X~X$UFqC*i{7DkMvZ5h=-*e>Vhz<>f(uv6AX2k
z)SnRM=t?J$$>ngPr|L>4zA4on3DUY<AaramNc1-U5Eg5-v1IRg3bgFpn2|B^+yV>+
zrvYDf5sflmomkMSm5v^67Zm1-xoZ~{oyi6f7*c2ahji!Z=v(Yzs9^wpYE;wvD>`Md
z=$AKrgzWJk-<3EG8*0TTzLQWVg$;<=VPBsP?W@zOlNJ`fl&Z8pu&27U^WPgIt4;I>
zKxPMBpE8R&hb>OUlEEYCxz|7sTBMJ|zkAM5_bSmI^3}-XVEZ1dt?wsCeq4QB+O)y>
zG6b34x6!p>|G}Hfy}RSrLq`w`Zmci%4@MvNYz>U-qq50;S>)t*d7UZAlz+7By)D={
zhsnE)hOb3GU9b1y8*ZOeDw%y%?FudExB%eqI~f1mVsNu)tlPM*h#qg|Sa^zn*o>&S
zCdSZ`f2oq$i(u_i3K}5t5pkfpCa$LC$bU8R#bX0T65$=JD%gk`62jWqH4nwZSFRYG
zuVKHxZYg$(hNE_z4Le_4{QR6CHGtEJ2}lPfP&ohIj;V8S4&H3;`8aw$IJ$8o@bM1Z
zxV7lcFfa^I+lKY}>HGONWz!uSp(ze#4G$wTZ#h8TEtYgyK}TfQG9Y^WX{(`6*oQ)W
zjMn(cZcUE{6usVgc)Y(&-}rrbdN%yDUK$OK6i<w~dNuVuktls5-S4Z0%O(M@OO^C5
zHU2aH{8_%8FaNqt)&L>$H(z91U>FO7?Hw9*>WAgg*PcWxr*j^1V_$b5^;lRs5cp!k
zpG>nTRYjKhA=#G2<yir_x@_ruq6f{$t4gu~kf<G+!_g3<+9TsqVpS_ji&5sH>LJgG
z=w#fMM<a?5?NVgqs#4NTEDT?i9DH#bzi~l}+>1pw5|P!&9SXAhIqq;z>yaLMWx``c
zJq@`f8@9`W-8m5yAnl8nPj_lmXF;xub92GtLPNU&CxRj|nm~vJ2cxCG58i3ITw2;9
zU=IF?s7a0wc4+`#=anC5dd(K&2f9l?8#jawZ#i$p4{zvpf?RtJoYI&3T5=xp!k4l0
zW!qgQ05-ASF2IipI!kYBoxNp|6aSGb_0W^?)99jfklV_ouqnBdu3w#u<cjD^xL3_*
zsauh~)%OJ@+(Td&r0ei=txlxv-k87QA3(kPcfa?4`#zpKz22Uk8@(LbzMhvM+`8QM
z?q283%$kR%r#Eibh7HbIXwWc!9ph>_-YiyE_xU!>VAcl!Q_7b^hG?1>=UkOT#+)t9
zg%nv;pKQ&`ZvYAcmEtzS;Oosq5I0MLTP-OkWt)v5ssrj<!tZ2b3&Vj6cscSO6p3ou
z)IsP*_=Lq(49^C!$VV2Y+O7(C-FZc_DiPrwE)Lrj+HPQCUCJ?$D+(3*3}hQJ5*F}b
z1Ls1$?Og((SY^D*kom8ZjCxS*Ga1FwI>1*_O8XwsSM+=b=;@K`D7XVzibk!xX@GUg
zd%?rr&nP)#`x4Gf`VubDVXb04RG(HfESl;ws_nG3G0B&?;nC>JzcPH22y9tRYp<+n
zuOxKHUvw|U>Al`GKUPn*uaf7V9&NdBwr2UWXImXxB-G#c)7+Dd@VsC`R8X_tx!N-j
z#!)zZi_QKtn2Q%_L>f$(QZZkFyeR1f*t!6B)!uxJ!5}=`IFkta$;ORW|9t^QF+@uJ
zel@<jUHe0}d{mZ$9yp`7Z|XM?iCFcPZ@rgDmK>lv(l^N_VV=_`CAnk#dMJ8!0y{LT
zKrml2$>;n)J}Boq1)V5?!Z}Q;U5mmQI5*F7;5QS)FLmCCtFd{QL!qYfn?%FP6Sf?4
zJ1DAF;+I}LdIvt;;**jA*Hhz5P)B9W$_lKDs({w-Sxtl3x5!KF+S!XEi<cFrZ?Nht
z_%RSOIeTEOE`wACl!RUWd~=&}LO=AJfKk4iBEGhoWp41;);2q%K>>-)*>V1^Gq4fb
z7zh!4SX&)Dqrf#V+Ju3P`rn86t6PR`z$R9I)1YJYTO&<DC=OwL>={PB%yikK<y*h$
z(Che%YU)>)5p8Xt=o<*2N>e_I%w(%(0eqg*N)iuPM61U@9f^}ba5InMR~X>^Fr)y}
zR=&R=YucV$T-CC#q^;A<{2G)ex)WgX-$4~qD4d_sNp?BHK5GUvXD>ElsS;*8BP&sl
zt+s2Ut+v8{tPj2m@(Yit0-3N)B^z~(cLmOQ<95}*Q|P{=co5A6z7Ytk@sD5oQB;?_
zC&V^y=RxktXnHgFJLf4Iw7q0ao6mPYzo9=hEdPVtVCDMnv)UZY|0~3a@c##BRiplY
z6OG@0h{k48N{a}JV3wJteReobIBwYMxn7#v-;5JUcsgz!rQ?BVvQl|t@hOz7br937
zcHS3P2p6&el&IBQJ(vr?iwp(S-5G*lx#PIoQ+1G{=V~p~@QGSvmIBr^Sv;lXERqtE
zY@%76JLR=TRl=TM<S*e3Tl{ExT6KG9BwBy5#2lqYtvX+QKemk6Fo|COP0PKv3b^@q
zF8fFUuSUJzFCPMrg5lOgqWdsk&mNV$&j-cn6x31@2;F~KV^7ac7ziICbYuLvl_)Rf
z-uDHW5eXG&6+wMm!BbrnNcYVD6aO48Srq}#hTI^i>L#{^X0{L-L@lE8*hcs5vHH)m
z*cy^Q)wka)K1|{IM7Agno?SBrajunJGuxAKrW7bG4c%h0a;bax=U*1X7J1%*He)2k
zY?Gcq=rCJPexIbG1=L%mCuI5*u)z+HwZ&-^B(5)ADKlXos@Q(NHI><xp2o|wI0Wn|
zk5~GI351iyC+D;bYCDN282$1S@Qa$=c?O5v?=1`BYeTg6q&|vp8*sReuj*0Y{IT_G
zl}z&!);A-t!+hpShQgSs`1WL3L|}J3`XdfK)2iR=b2r(Al#3ylERe5R1`a8vJS8)`
zYhqVDK0{eHZcb_iWf4uW{%o!wb%QQOG)LUs<}B3vVa}dhIc*gMOwf5Mz!GhyJfjL-
z=i%JVSOz+ddutx%%m+(qPHn|^87L7n4{O0P2=KrFkw#<g`7Pm;JuZ3ccM{Z9@T)7}
zlrZ_9{jSoG{SG5}yYSUIa>|`Fi`x?NwW=^<5jZ#XrY`yHT}(=S*Xm_Z=ry8|K5Ta<
z&NG$v@-u!Kj<pGl$m`d{Q43Zbacb3vH-mNKC5j>AP{&?>=pm(iI~gM)hh*-7x|_~6
zm~lqV#+E_X<9cYP@f<Sf<zg;RC%C|$xxyFlN<>&MU21(5Xax34{tqXFjgjU52y)r~
z#r)=A`JZl3{^Xo-B6c6sJb)TCqT2}?61MC^9I?m#gK)-nd9FYJeEVk<iYgORakZdL
z3H_A=G~X_bMB)3vd%e6B1I~Z=K9j>Qz>-+!*?3$2u!P%|ufNu*?57Rcs2A>*>UQfe
zYUZt{B|J54MD92}uMd}Sb0fPifL<X6cCViWkR7Y$SUTDEMmUG#SR9e}AVPP<^nHo_
zGM8b;Oi;A;grG1T+Qd^u6xjs}&XUH;5(ik`C`<n)@7U__n@OaQrYQW0il-r}Ju!vI
zD2}~;O?q}_kCEe@8VZ)vXW)sCygyG$@1L?L&MG&JiZ-}cE;8;;Xmb8+4Y%;nYP6s2
zE>PHKA~D^_h+LseUIhsO0ToLS&8%G=pIz2p?QJ+7qnf&l=Z;~QLXL{{Y>AiWNsbjO
zi-?AY+@g{I@7ULZ_6JE95c*)fdxES_bP8+9eVBgh_dLi~tDkq@n_M{p{4j}nbD%7R
zux1AZdUN+MWb?=N1NvZsaQyrU{CGdtlBV@5n?B6nOdf9DkLRj=WbkpDJc3)6GbV5Z
zqoZ&7mMQ!8;C$pQyGePYTA-5M*i4c=rep4-80mF2X=F<GlN4tTKd9ji=<(If(mV{|
z;B1HZ(OH^f@0X*B1Yc(<Et(rr0ROdxF;Py_K5+#!;kA8LDi5!PIAj<_|H9o(;oa&Z
zsSmBM69-gvM^xklo8!|4HBp9Iuy@V{X{E)v)RROvQ>>`+*W+nL5*ly}7}z#?qqnwS
zu6B|}ua{Q^$Id6L(ALUTObbyYO3w~RDa2whrer9Z9f8J1Yb$D@suLBf8Xa@<`MAIy
zFBn|WQ-z~srBljx6~rXj8qewAi31Mp=E!q}WoyQX6FLFt_*@CZUEs3K;tud=G*C<a
z`|T#N`S583A*VPKqT{Kf?o^0(Z3#H(c=P+xpGo%GWO&cVlx*+AQ`29e_8=SY;<?Au
zOi>lLq@g_mIp|Uz36<lWs2dsDQ2fA*Pl@i{4{nqE(z*}j+a)5ZDZs3<ZwzFwJFxP<
zmF#m?=Q`$B!Lb*wZbF@Q8=B0pZQ|lXWwE?);x7LEQu81pt)1@njEV3_U!2@TPnv};
zNdUNHO5dJD&!^s6nV*S*Yf`n3=<#YWmHnjqmYB=CSr|*~7?M9j9n9Q}-&uqtLvyJ5
z+K3b63~f<Q+=q}QS8R^rg__r#1$FD<AcP@LTe-ZN4Jc2ymUxFL{K<JJgI^%EE7PUZ
z54V(aMGz$jpr8(3wQ6bflll2{ZV1Z`<YnXmZ0`$BhU1}uJYQY3`FI7I2C%JEZkx25
z4kbtht-o6zk|&bm4N5K&Cu2<Y>ig0iGcOftp0pkZwrdUo2BU_}o-lav;WgM{?>^B9
zDURk#8CkVrrf5p;#pj=lT2&rBBCG^1%O;B5Tff|KiEFKT_Ve5ScC>eXz2utTyHRd#
zAU0|?JbbpdTC8NpLh*iMajDU4*Zw;#_NF9OM)&YcXGUA(5)YvNBBXgq>K&|Tn!5f0
z#qcO+wSgVuF}pJpS1NrB`!*Tauu7w6UP%)i+n`$BvI3a}>!FJ0fRjF${=(&G$}_!#
zpO|^ny-x^P;N;3Yk6_A8Pv*ghtJ5H>Q)o@b_dcC2lO*^hbo@a~@bHg6F!~<(Oc<}S
z6YwW->YZ!dV;UAu!kJ{W+2+{M5(SmyY$*`2Fx3gPDj#Q0?l2_cWVkifX`yS3lmZ7d
zC)Ykh7T(FSYcvDo;5J1LP%}eh>CA{)g|SfIb<jm_((}?quE(@KA}O8Z0fl$$n4$uK
zSgV_H>^S6PiY&otC7B38x?lZUdmPY)n(<PDpH6BWtt6wpwoH93IatRAW5+^NI={fB
zWIG~rnodA$((}~g{?9lTS&?-R9kCDlbm2SR=+_G;XT3YD4m-eHdJb;bH{-WV6lhnq
zD%$h0oI?_x!&SWue}3wpVkY0M)y({hvBN&(Sl88r@oB`I?B9^Rt41>4V)H*^IQ3l>
zQV_Y5VGDt&N)x+(klJQSx(vtzd+f<~t>9IuH8tIrRky1*sAV`eRQ<$Q9U}nSnwv&o
z?`FO5N#`JHENy2nF@ThLjQy};qs8Kb9y2HznyS@{j>2%$GzWsCN^WePu%nnerLBF^
z{&iVLyaLZ2_pwlm$j)rs)eARl1!>Ml{$RH+14#L5PoiDh&vW3)#gNe>=0>MPM~NOL
z&P#r{aOrU5Y|Ayv)7NeGTF_kD#R!QX!VUD{Z5Q#`yy%d^c{U$NvV`D69L`x?yfh35
z%^L9*HFX~dQ)C~u{B4D%WZDP9ECEN>f2xX1|1X>B|EY@X%*_9z9@(b0O#B8rV$ZwA
z2?Fh)=PfW8#PUpspgr7^`hzMd*Wuj9w@5f~1ti(pxpgwRcfw@}<@D6(E(b#GWsghC
z`=x^ye^)1>`xY2*{O0{J{6pB1LOu6M@jcK8lFO_?r@*#Ow0vE%zAas%zP>-?XWBlf
z&>U92JK-7lDCBHDN%1V=S=VHu$s13~@r-9S;q?5m$%7UDiEq9O-nZA+r(tx%vS_+r
zY28y@+3}!2xVkGT6TRT`^NL7EKUf%u+mx+j?AKGpyU$CJ?<JvZIgUV&?`dXYJUE|k
zFUK-hB9<2-9M@Sk;M%1szf3(^do&8kAP!-M+*|AxDNmwg3UN3X^aX4*oD~VM&lJ2F
ze9as{3BAfYqlpgws|{LnJHTV3HWUdNmWq2g;=xV5;zOv8w7IelnXL40f4!z0o%uQ5
z65mveG9I1eU3$Y=!An99vM&11_$}X>J*@ZO!8Hf@G%{^fQAfiza~Syu>&g`7x;-q{
zAV2JgB-0)<Hz0~A{*b;Rn&{@=@XT;Wn*kRp38MYWSM<#&oCS6Mw}(N4bjCj}8F6rp
zUzhoieCP&JX?$>fo$Zj1ZehGK&+qzOWQC4Dd2gszeERplwQieLP%o)il=Q&8`S~gR
z?BE?}clgl}AK4rx{L3IAzqe<(ky&rRax+y@(W&`!`*`4Pa*q>c+<?%>HlqL+6liEB
zym71sB{K5|y3Ok~=YBJYkt3an(*<pjd5%K^YVimIi0PfNIHA?jlLW?%!i`93C!1vJ
zj~Rd5Qxn4M@g?b)f>o7yEXonhl{tzSb*A--eT!t>#Wg|l1}p_AP~97o4|&sbjA?8~
z@#_i+d8dsH^fs}%-!4<tII<Bd2>28m6UQeC+(*2(GXve=yr?YW=T;JF=1dTaRzN5X
zV&IFmI0z}C!)TlBm6^xYw-p+KW<n4(B;AA@0C?=sR%KmXFB$&H8*(4UYnHVg1pwyM
zjZO=$hQ9&7v+Ly7Np|WkH^s6IE;hdNRS8YZNKdR-8_gv8v70w1ymb~xvKO(bYMQyg
zm%WTz;qYT)K^>EGfX=0RZaGOtG-+kiR*5*I)U(akCMunu)GwFU0)eri{ylMxoE&xc
zkfkN(BzZ5JZ;G)@WFpfeha@?zeV!t#utFKN+tdr@0M3j@M2Dr$oKib1DyEL`Pc7QA
zG1c|g8*wpVC;Ak|m29R6ZaWX}4uJf_l;itx-kPh_T4oRjH_Fwiq_-PAP?o5{*<&7@
z)vFH1>ytUOYS4e-S<|0f5msi2c2Z;(kVoM2ATth#2#}{c!^!J1!hqvYiCNe(?tQLX
z4ig&52B6UKq1j$AH{-It$7PjShs|@w{F*-a22{VHf$e|%3K~Ihq&SZa5BzM8zt`CD
z%(mqk|BTTHrm*0vXt_)DOCm?|Vh?<&VU3(ir0@JdmLo=GMfzZvCl!W3A@ivtpI?uh
z?2Mj>&LlL9Qko2=L^E|D?YqsWFvt!kEwiocc~crCl|P-)R9sB7YE~C#*M?5+>{+eU
z4_(c2K0ko<;bzBTf|Ub>Vw*FxemX%bubnb}*q^i2k`|6VsT3uV($|et@X}k!Lf2_2
zOv3$%yeJGb<VKPu>(2<5udpR<JScnpP%Nc^ihc~LItRBoCnGx|dsXBPf6gTX3FViw
znn6}sjC4AuymeD(5|HlNLI<DcL_R^=D4fPI82xSS{lIxJnY=X^$4*!eL4ROH*0hBp
zeUMzo!>xSd*U1!Q7yB`1*Rrs@b_t=SDOdQfI^E67tk!TTLPyB0x)=B=(6(^}`Whpl
zJ3>lxk;<fqU}X;3sgn+6LXfl;tN@!oJa?Ae^#QJ{!=8T7a!qWlEWGS5109aG)u)LF
zvBEYhTxvJ1z$!6|aWcvBeTYyPE=%k{(1<oe)G6kK!nX#=Gu7yQREn(u#?p3hU?+y_
z#o}r|-#^olwj-UMh>2XwcBN<WEO&q-E>6tl<i%kJ1U)uvE%=49TQJUeF#%7y-O^1w
z?VDbw&R3<IsPSt1dy|YBWwd=W4PV2#Z}a&&SWO&jc!xq(1b3&}$Z<^Rz|addOvBMV
z{HJn%ht7!0ZstTan}#Os_0(T!(1w+YxSVi^gET*3Q^LcFMvOt;t7UsR(l79a;9%hY
zRHJ@;jQ`zA`Cm1PnS<?rhGt#T(Xsy-yY_po%MdUW<2VDs6u+GQ8`wiOrZXQNoG(R?
zRCxSQ;NQC#nj&7bhPQE^%iibZ-i|D$jQX0GId=F?oPL}&%D_R88>80+mEWb`<D-N6
zMR;46B<WJ`Q4+#p{d(v8I+>HQzI%>jpe}84b9n3A=ib)KM!f?S+FKsM;Oc)~BOKMw
zCBENQ=8Xe*Ry@vu*(_&;I37I#c!c)SBy(<lKThV^o1#6XGjJ3+$YTLB(;E11eYEWK
z9h1fPBn}J#P4=eCqDu$IMx)A8UElCwf3$wYfeL)w|F|Q*pK8$APj#%S^y-$1Otb=%
z8pk0moc6Cuy^z2}AV0B((y=Yc1L8%v3Z?@qha=VuDS>lJ{-oBZ?3y*5<KoaLyzmF$
z+{ohzoXn}q@AD{9l`W#nk;9VP$m!f1zQeDqG`@0mUp#QbTK)!hJ#-v-m1_M~d0}8$
zm=fXN=IUw#)baF-T4_>99D{Fy>XG^&7fL+7f67CD&0sGwLdD~opqo!~qm%%$B*&sT
z5hJ%G-s12e^b!Z;VL%GC2_p-Vqbsd9jd!1Z2HUu=d1kz>wtnU5l*HEWr0IlT1$72h
z-}Q$w^0m3T;K?4N&tmUzHF_`$9XpGiPtB}kLejrjogPp{ZL&}E<F!DDqX2pLVUsQm
z!C;5DWKP$k^h$!`%&D<%^c!c7D?#t*m$^sJLoi|sjWyv83YI7_`#K$J*mNpUscYr<
zEsYnhWfJ5DWlkUgBQV4`9kf|=a8k<14r6MyO*b!i%*{_q*bzqKYe|bCo6{v<#Gh)E
zwoyN$^o&*t%+(DGROXHw@Sai=;Cu^D|C4Pr&DG%2*44-zi%P4pBRHFo0-?kOJj?Ph
zY3OlLoGPgG@+Q%3+aPr$c1UJ4#K%67@#xPC?eL8hb@Od_j^Z?h1#bZfwUtZsyj7S>
zO6V(El_xIwLJU7^GS@yG^NI2Bc4VcR=PH+6^MX_4-KH@69Z{L-uU~S9xvl?p@2GV&
zFQZVlb<0$j-hxI@3s%v)P4rZ)HQS`sN-cG;7!Q01pX}j&Ovla|Tkl-o4L_KiSW$~M
z4=K1?eVS&IIevUi?+Wf9;1P&#LS9qOy@Os8(FkVd+3gLzL;m{y`!HxKtsyMVi7=xI
zf*66r+p|xlc1uVL+GnMN6J+L5+jKVvlps@&z_uGHtJ%+7T5G74HPsh|aUw}N_>D-b
z0&CCY?ER|D8`e+!X1R%NI$iNy<}t;+o6`Y<A8Z+NCjpEVOAdD9U{8q*yof1HdAUm|
zd@h=iGhce~iwL_5x)4}+@MJ(J0s_7pGF}N`zL*PSzL+0!9tHvS3Vc;Jz)GEPHkP(0
zr79&wg?YvSjzda?xP?rJ;wRp(Yw-0TreHpUu{KRmMV{RT?u{^yii%0dvUYI%-^u~L
z2R)hV#;if$fG|@k0IH?9#AWfIAOwK~$a0QhkRdhNS9&7w$(BCwQI6uBRj&3x&6F;w
zl61bd&nRG@8Z#i1Q{*-@%}c|{1=__pGA#oyzS}a=58sP$`aVZ)byO?a!NAnJ__5M}
zLyzvhJO2_X2VU^>mtX<URi_~EQ@{Wno3kBoQL5UJ@?!ZQFiQ;FzYOLA?VnNQ$Tgyn
z6m#e30l}+h+48)+XA|K80dw65gnuH?kUYC(9BZ}ITQCHl=HLbc>YEpkFacesZXK?0
z1)VPW4#yMMr5%!P2xj;!_y$D6+wBtL{7uToQGaGtTyp{B21vbL|D5SKIC)ge%Q?IK
zZPhQ3{K~4X&9mtL`BZ1&>oI$JdE@sKY<#tx608RtYPz+FGC0UFK&gW4EpLCf_;kLq
zGS}!tSuq|DJV2V>)}S3aU)G98@-XBKg3euc2|tV6z%m<2juKs(gx@I)V{60pq~y3x
z&J*vjFgo!b)%@&2E@e^r@Ev0JXrdz*77J)?{=6a>xELpGJ3^!VPWIxGp~O4el|;d@
z<ejP3){T_S^ZUBJohePNo_WRRnORkYw3dzdK+5)Sf+Ctl=JLV_27Zo9)0OnMIkx&`
zV!gJBC_0&8{=-QGixQ%?w_(Ih8k<JZ5z@Kx+%SwB=FK&MOL*uJ*2N_-XED3>YIO@t
zpEu<q3QP^xpDf80#$a}#;xjAOw(8h?m0NMp5Okd#N#-j&{L8msw=waN07c4Y!pIs;
z3g<Li@a==~-_czbXg|sGMz(-K{M_RMCK2V8VQvOF(h-u8e$Bg4Gvwgfi<%x|axBiC
z52zj)b*8Of1loaaEl^dPDgb!79-FC%g05vqN{dvC+yWg604*JW!nd`(<=>>>F=)+S
zX->4a^PmFjHny#&X`waQoFe6~M;~y#*;I~yVLRFfx;53QBCT;aSZayRfch=nXZ`Y6
zP$vYEXWd4$*bTBFEvljs(T2o3<43>B8E0x`m0$d=M<R|Hg>r6bQ%8&3&xiaU)j{Zd
zS~LPMr5*rTg2e=7M(-=6c5TR%>732C6oZile>phWKPw1Tds(>1juFm0H8AqNop!&K
z;mXD7?{)`hNxEY8tE&@d<Yl;(({*$z<D2+`!kaDu-Ru|uUsq8+PUF!?x+!GEM=`09
zh7SPVA^<P@#(=|4I!Ujji+L{ruI(?A<{bAOOr{i?sYVr>6(v`C`+1st@4{TUlGBwl
zj7c}_=OqpYr#xNjq28WX<3APVE={u<7*R~*7YsJXwlwY<J4NvdSJ9!Xz(HaN7?Jeh
z1p4%tu5;M(3}$^PEZ3EW88)Jpl?>fYO3KcxDk(lUX7DR%{QekiF{Jtj9Cb+viF9OB
zH?{N{htn6wf(DdFK68Z5`B-gSvH1H}NQ<D2F?qZ$>O*<Rx+_yYSu@Cc*K#F^+8-|S
z`Fx$O`d8uG<ix;+Tc;>w2>ssL6{K^S7y>k=g9Oe{bGm+gajXz#^I8;KO2w6f_558k
zqprwV*&qSbAlrb7>$KlY-_0DBFUC#{K*dryhf4|4i{1ho?=%qHMlRl?G2H(GnKB1M
z`k5^4VzferQfHEIj=#(N8%p%R`e$MIWLt$wotM~maz1}5BcSdBQ7l{UptaUd%9Q(!
ziQ2Rl7ft<61#fUXeZ_dK(OaKx6!Sb}fG9YWVZ(X04&e~5YC#GcgDzdkucN=E?t7K6
zLScNx4#89(yVvt9RBp4SK*X9QCchAc8Q()5EazW&aGWK~R0e0^S}|gyG*KAv1eZlc
zlJ2&a3nK{1^Axp2wpjy3IJS;qC`?xO*w&sc9U4m}TNcG>4F5B!p3GChrdog<B{#`s
z11MVWQRP6NPtGQuSywm7Zo7=y-B5r!<5k=Ads)(I)(cnZBi*c)z4C5O7Zn+DH8#Zb
za!a|<k?$bV5o00&nt{xnwoC9F`ow(;Ut6R&C(bv`jyPBKUULU=$In^w6EIH*@AR}6
z<b$~jzUp!vgOxh$r3mpNNrBNcIE_#Ja){(7ES-FJc64?TlnAh~o5|ciAH4lLE%Yzf
z^*6buQpfbX`Y-z>%6yl_pb{SOPepzliQn2+{flLalO7Z8KBIWr_F?$dz(;<h;%Q6f
zmIbwv3bP<ecPKk$UW3OF&mN)s#HddFHMlFP;cOc2Jx~Mg3P)X=&k$iNv72XCdeo#b
zcS0lQVNsx=OD}Ql`vNb?D`@{o68}(`{+lFb;{2c4HpAFbcDNij{B<V?5$0&oID}A%
zBWZ#Ea8J+=5TX7$*9Wh19lTfUE!(=ylE<0=*C?$d46k!YE98;$b#L?WUT*PXXLiA$
z^U5m_Q0y?UenU(P{)E`)X9u31PClI1ET6o1?u7pUh6!f<A-|g3IUqplfW9Dz1svS>
zp)rj;;)V2|JtUO2MAlko9Y7nxo<~p2_->y2+WNOXpC0bzF^so)*BQahirmekm5-&E
zwxlZ*EozpB%5K8ROQZ2$QIV<X0H#WWALGYb2I|aH!a8W@JzY_zB{HfxNm-5RT$zU>
zZB$V)AHTUSg>KX;=uRRJi^%~^X$xy;!oQD+_;+df@2MFQp3JWm9|X<1r<)G0H8tM;
z-Xh&WPf0AK2;+drYa<EtYqI#7Y}OR?MzDk0jw2Ll#0dp|EN5ccn&{v_#KNps*R%P3
z{_a1YP~kiy?Vdf~{k_Y?&85n}g<yZ3>t^6iBLY<A6LR`GDc?f~I#JJ>&>QRa_B8t7
z_-r(!-L@copY}0%Mmq_9kiIvd#B#sRl>SQv4865v9II&6k>>uPCE|vOh2nw)c29Of
z%Y<~gzb_86^`B_txVGOzPekkz|F?8CE#cHel$|Jf0oTuf7c%xLm=Nl~GCTsfn?a8N
z_%_4KKlFF1m6ra>F>y7$jT!o6T{+U-Xf7ot3V$|8;HZgs)~-3eJ{^7pja3I&;c+=#
zH*88H!cb&tco0|Dqvw(tV>W6q0q$tXLoUcqf*Io+CCmY|9J4+TNwuhunKzom-hc=A
zy?~`diK+Tx8~saEPFxKN$j#mfp&BtWMtWRTjQzSbN>md^->pB5DF9>nlkij`4NPRg
z1$vFG-{Y<+R;|~V_|de{=OvnWjg#l2vTAQ3Z~w*<<tKM3VcPS8dHGb6GSxH|@YlUo
zN=9auv+(o@CC5Q7lG0XZbn*w2@2W)2x2TPdauAbC7)4M-O-l{r&6iq@WJ20Y3#bdp
zSq07wr9u~JLh*H~i@nq^IUi&5kb{(WZbRL@wa>dg)TDPq;JD?{3yGtDouj`q)s0pM
z7h1|2Z^1ijND`5)CmFn4%Opc#fa<o#uVRz0tZ47qs>MIl)-|S~QRTs1u}cK#?rsrn
zJw2{5=XQ~wCJ2e~fAqn`l9qw$F`dF<BB)z(Ln;qoA0CD6p+yGDD(XX>!~RQwq_1AZ
z{(wovar898K-b>aFxZ<yPc4koOXD(G?Oj}t5ji}C#Ll`^60@(fv;`dz86jVg!#Y0<
zGt*VY{x?u>YoU28s<=E+j$*}4RNQcXKJ;r_!HSK9k!kgGn<@}3zhvwLH^4n;`6eE*
z$NT=dEM9=`8$PP_lSD?V5jTk_>{Q0{aOY%RA#aj~Dn8#TNqUCU4XX?{$$sPP{l1BW
z9i9<)8|S7Gh>HhYINP0E<*uQc_*(GLLx=-(QY+8fkfa<hVZ4rYEX$C}`!Mx+MemMU
z;aflbu2!IIbJXg&U<|<hKxJ?m!^!=%mPFyOMg%8F0e}&va2xkW{UbVy?+vtm&%>Zb
z&&fPRG)mh-$<hqKo<&=@)f0-TGA_>-Uk4H(3=h-@W4&*}tufsDh?<FT6i&2DyZLy&
zPj9$Jx%S%jHRt6B2$R9$*u<>Zsm_-U90;4X)KX6$-UOo&o6NXtf6?Nzm~2ereCb0_
zkx($dFh9B&Tw4HATwcIhz<ke_htc)-3?dH$l`3jbRCh<2!``%>l67kD)sLR_MGFV)
z?qEE>2#wE^w5Z#?4s+gip?X-Wo}9<@{f_7T9Ht;^q@ff34YeFRB|1u3ww@4@c_t$F
z)k6?~k2O{cQC4N{J^5fkv(fOm7T0+B%|@4=y66HpEIn()%|L42sC1a~tbI?Dlc{ED
z?8LF^k`C7rbH0-8%JLz|8zd<sfQ`0$T4+SC>d0Nl74Vm6(p6$}=2*#7b!N+N!a^tn
z?`&`29~iN^j-4QmEu)289qq0saMo$t%Fn$o`Q=DvsYmbqw00JAnc1Rj4ijsxLeAAi
zK6rxx`%1UMV^k8!ttbf^Mh-3jNZyld=O*feIvm{}&`#Gw$*S{gaR;2hVhm@>=)ghg
zLVS*?gR)V0_)lkrBkiwkEkAa0>I7bRit&IXPFiP(=uM>;oH{L<5hQE?`%F0$(CcNQ
ze@kNI@yDMWXubVkjGa@ADB;$o+qT`iZQI)2yKURHZQHhO+qP}n#`MMHznD3bIagJw
zR8py=l3L&QuJt_PSPI&wv`*RU!5RE2rWt*lpbx~7qd8I?nuxCF3hmj#SwOPy&x>99
z_Cz-3rREh9oqPzvy>%QTs8O+Jl7u!dGa4>4{~91UX|oI0b@XM}a?=jnf^8GUSy8N%
z6_ttDP1!&6xj!VEC<@XYq4Dec46Wnqh?zg`S(YZ=3Sf{l>NQ(o0tFP(&kQhxdbxK!
zXA&q&D>pgki?q`g{|Ts9;@{$R0?kAX>&9y8^FKzA`)1wXs8C{1bm<P}-^g4tZsHHy
zsq8~Y`H0~|tB5w$i*S}43Y;@}^t+XJO9zKC!!_8F?1>0qmm2hJ<IMlDHn#=tSsS-T
zzs;&~Tb8#-xo^InL|e61@sbgk=y^L7`oUtzQN>yxXFIj>U@DDGcUjzXsKQjoCjY##
zP?jO0JHLpIn$+RK(l|s+ZjoMeQh#Vl*=|YK6{d(~Y>#4x1qTag+A?6s8`zL`{e#fJ
zP&G4qbNhq43(J**37PdXPy|yDGU#pGM!m-0-~hKeBt;LN_F=!=ExMO?_v=i<jk>i@
zlGfvHV{OtJ@Q!^RO@px`NxRPx3K}=suEMN8Z(^5QGTU(K2L<08nu1Oasv;h#dzx&F
zX#a^SheNCfGil37BRhf{J4-^Lprv^XB&Gl(acC7lt*9^R_@I$HE=p=9vwk~(uWTeD
zN3U`7WB$W^mc3Kd4X5sA5c4`8)-k`Ap!9;@E6(lxpK2NwmjAn^Vf`=Z3RhSY(b#Qw
z)itM^_58ud1+PG00g;Dsew&@C8-yXfp;-aAhHN39-sPfbI$Ky`O!wD>qj^QvtZJ1n
z6%;9S5t_W0&odn^|9~UV`zUKtPG%zd(UE+T1_eyeYZ7fmXl8bjbFPaP;c9ra{}eH}
zx^=?8+TdSDLa__&VWZy3!@J<2L}O?gFi&&w@)w@SF5}*}$}xkljGUTv_rE~tQN7%*
zob+583OsMk1(QJh3I#M$IpKl!lqK!JBVqyW{D=g>`=<$Ml@~<4E$q;6?o*YVln<Ri
z!KNlxGY?W4l-W&x?N$h3Y~S6N&19~VRe(9F&gGqw4g^<IM6x%=?~B5ZM{14AD`X$J
zVK=hJ8<9fZA@y-1;Xb2Wf4;>S1vd~L2|;&JBUvWs$Sa;sV~D$Qn_r8C{>DvPW`trd
z>1qI!3KNQ?60pdBdHJ1;D}|n=dW=3SeF$nhHd1_frDkVmyYSs8*Q-eXB|ctg!R-Pl
z2ow-8^~AoqpV&#6>4;%k6ih-vN;$-*V!{rI!;feZ2^*9MQG{C$83d0Q#72Nh>$UXM
zS#_NpM0``1x|Fn669NfywoyR<@5{BmWMUdSKgz3xCE4~yq8NaPukY8u&Wk$Giia1N
zQiwUq4%EL}N58S1d*wf2-I!5p5Wc!h9OFNMx7G1Dr72fy-q}@DXfVvF<3-gGx>Xg?
zjhHOY$&E7{AlNYU-~Lmtg*U!2vawdO(m#kw_E@+3u-cl(X3Nal^Us}*QQzRNFZ~T^
z3|qlSs;|Soczs~Q*b3~KfEpC;6BtjH(0Hwl2!Sb0r4m)xP(J?0MOm#WE0l3)jZ=cJ
zpL#&fR$=^~OqibVoJQKsRhel7w$H;S)`XR{FW?{(H`j!Sre%2AWfXJ^QZ~2j9+NbS
zO8y>4iDPL5wt8T|z&}64ogs1AM5B-)lIgkSh$WmG6o^GRCJDOQ7tGbYN){x&L(q0h
z>rgd%&?+9T9bje%U<(<%q7#yO_SE&Z0#1FR=(LH%OJYj$4|>L<prCEVjpIos8vTTZ
zpH~WY21Bv)t#n7hWG(7QaRzf&jwA<m@vJK^L+@}nb(lq19GFdBSvT#It1LAtu*Xm$
z@}!~FOO!;pH456s_y?(C2Az}0d+}+rA`1|8lHGU`Xf2CC?XPo$s1K8Dh>@zYN%FUK
z!eXm{fnoAGXB12DMhP5dbWpY;r|e%+jB?84nCPLVrWir0ZAK@A15}K1{nz{lHoeqP
z0Y$Wo!m3u&oYz&hVRtUk-C$p7(82jPGtknXwS{XCy@R?bKh(^X<TJL@Zl7$vsR-pA
zE<f-0Z!<aH#UQk=Y*oFxQDtG_dT+`2aba^<@DnTdk$6JpFz1ItQ2`{Jd2*h`w3Wa)
z$+)>*<N;;U*k9rWh$I1OQZ4L7FxjN{U|R?W{0JB1H!n^2hL!7RVa4HKJcx<)$#`}=
zHTnPA-%YZ^E57S4dFeW0l5!9eG~(+Y=B?_GpJdz$+>`GSoLF&;_JVNd3d}Mwt{|%d
zu!OGxFiQe%OKXkGI+XO?3W-uf5YU8n3BGeiWaxGCSsDue8dUx)Au*GFCpXHYHXAQD
zf5_~Nx6`$r3%%s*ZBkSc^tB(s0nF+G^@*3|!e)?Yr%gAsA-H<dKX)e?jvYlAq26>F
z86wO6-8-1iOEv4V81(K;6%dEKnF57LPw*ecICy|%I{te#!K{0p=XFt%#0<9;!+d<C
zjbXY7#3vLRW1<XrK5M?CcqB4)M@rtXlBIh5t(>k#b=h7$GvuaKzWoq6Q2RTSnu=D_
z8&hXt(UXThyS<SK^};}Vg%Xn1*5O=oJ@OdZUIw@6b)C7~or_N~%7F3p#EM2z;X8uv
zIwXhD{^rG|pHNW5);|9y3{i3=b)4QJX1>Y47p+^~*{^?Rbm5-d2<>RP>2cnES~gxd
zIk#$7dMv(B`<V)>@N0AC><`Yc_1T}1_H8Fkn%b3F+F{;EEui4i@%K2LkK~GF4-c#e
z#ZDCW^5|vis(p8M3O!_=WAZ6wqdupdqsGUd?u7f<sOrp?CoyL<SMuar!c29zEc+*{
zOj++Bt0v-R@7pdq-*uZ1QXeC{ofdtYTM##Ix|P+L)E=f+eXC?>5j8y=F1T8n^>?fO
zMwYSoHvV|z8$6Fmnb#hvG_VNoR<$bv1DJVqo5}U~2Ul+e3OZN%qCXWpbvhd<CdHZL
zZj?(&?hGQP(1H}Hc9lA6p7*AcC%r^VI`uVA!A2Q7_*_~UZP|L$$&7ZqPLKXsmMk|7
z!(?lej_nC2Pk0*_!%ddx8JZO+nNd+q5$*9VV^FKX9+pdDbaPH+4(OXUb4^c8XI!2_
z^U5HeEDHPShkY(xY+IG&Lpg96Oj)wHrRs|h(mQz(5lT_?-oi*O@vDVe9$;ITdQcQ8
zO`COig`f7MXY^oXtdQ$w=e$;jJzB+WNk7e+tS3D`l09@R@fzrA`}Pp21m#M=Yp&({
z+uQusugU3e)l}(pr?l9yUKLMZu(Oc+PU&6^HI85xVv?fW7b@*7Ejdwz<USI|@BlST
z0@u4eVK<zdfbv$)#nwCBD|i>Ry4It;koy*yt1MSvqRm*xS~5*91*nLV+#OAmI&>^1
zflIeP&TkqUN9*Q#bI?EA*vDA1cZDB0HmN7gjjX1dyDC;^JLPKG8?RV(4g}hVlr8Xq
zx{#Qlrc(Zz_z-e<d)Pb}PKM)kzP(@99q`@?Qp_%Kxyw<8#?<_-{ge&fRNbeGm|pWR
zG=iy591U$2c5oHxGcT%FQbXZ0bc5aAJKS^2D+LRftkuXD(<7`U75Vp@OKTZ4t0vN`
zK?*#f>R)Z*9TydwaQEqW(UCInVAmf^{1%p}I%ZNnc)5aJ19waaU<Xr)sNgi3pGt5_
z<OhLuD>&t*T!kIY3^9*r+VNp-=&ViSgDW?()1c(kWM>V(+}bL*@k~+IPDz?eNT3xN
zF{=->DG@5@_Qkjzhq26fq&L|M+nxTuQwr7W>IjLkk0d64(EtycygL72?Ga5<&0}gW
zWSu2(yq=Zq_+lnM6w;=eIlVcm)GSJ`)j3i>2T{u4=_KX+U80{?AUT6h$#5a)&S1*U
zm9Wy3GHJ~}Pk`fHFp#mOfKW_D1*6V;fO4CkRTUsnUb%vXyZ|WCULm*ajmH$40J(F9
z`p9M7`n*6%#^fUW@b8~b+s%FK)R$e170Vq^bBg?~q2q=L`!|&##yN!AGAHCBfM0In
zcHMD9_r7T6=nP(>#-+kJ^i62IaAi5`!q{V%prVmOg|d-CT}L{nDHGh@?T80$c{-v2
zDYt@OQiJ9#I!ANRCA|(#T76r;BFpPKDL<B9(0sM2WM!BlBMfPU!MB}hF9jAjeCY(D
z`grsUuJIqLeMtp^7l5sWdHBlb`#uY)2WnPP)zk|{^?BV}ul1aU&cAE5LU$F&=KC=a
zeA1zs62kjWB`E8EAO`%>Q2(oLs41+r=u1|{oSIo<ZQT%KR2$vdUdNe6zo}<EMuUQ&
zt5LtNpFXr3A;k(05pUaE<p`7rCmj?WZ;jXcg@w0+vyqC}2t8%LzN3Q`P*eW}I$Nwo
za|QrOXjNWs9dGYF92PB;cH$GkS)<dQq7F_T=-NlHSs3>A-A7B}+0a}TptG(EG0q>$
zg?0%X^xd7jr)kU|-dS!0OZt)@@6V4gGD8lB^`EDfyR)x@vzQV@r=UzTgoz(-pL=9^
z0Aw_D*=VuoiJTtl9^Nk#kPZ7`AiH3TD#Z~@qBPc8)2q~5g*c}8T4hQtjGIp!3SE&|
zuvqW!#||f=9dF9f1#z(H$l11V?gW#bg)=pW8H%2u9f|bTp3Zm7N2Z#`zwxr|k^KD=
ziHKRsI%n5Nfs9m6MFTo|%Q0z<`};<hh^K{#X9(j64=RS5NJ8vgLy13mK7IsS5DbSS
z6$c8>9=+n!>T0B-#xx%U2($SO6oW9rSfoK?vWcRD+DHS8b+;6H{zM0L&YMB1&oHA$
z68k+nw^qBiBol?B>%w7Rj+WVX@q*buR--^w`i@nzrR(((pjzw$eGNwCAqa(Dm4iXX
z^Q(k*{<K?O^Q@U-vqXW?>z(mvqEpndch_T21`hVvoJC~hKOmePam$Nih?qwrl@PV)
zXNs|4X@Z(nwsVl+C6xsR!biB)I4_SNCQ)N&?jhJ$hDVFC44~I^_27renAIa%>K>`#
zg{%A*b(RGM7;^qC`Ew^7RYJb^SPg>Z>Lba^IWa1Bv65GBjUKG(fXDubEIM^rfP(d5
zH`YwOvQeX(jF>c-M#vSYBWaQLqm3v&bz?9c%%lqv`*sR<(|Cu#$Ng!bngoiIMvE|X
zU#%m1$N7~pHTR+(afwV&<G5lMu;p5#;}X<AAfsEOE!fUSZ>5L%o+z_J$3s$WMrXl@
zTfsNE&c;XV_(RS5|J*W2;Up%gYu@Ja&ca(DWGU8K7_wlF0>lnZn7`ZEzMVKhmtgG4
znAk~SKqHx&;1$w`>2yjkSQ?h6zTWJ#iFIM&Wm$21lpvB{i!KF|a0%Te#0aP;Ai=bt
z*3^sNxi}Q!?<F2zs1)vB3UBij<?7IA_iofL$Ox`PctlArAlE9|iqIrcX2Ofr5teG`
zD4O$qki_iof%akC7LfHX8XVti9^Nt1#}_yVUegJtp#}oIwG+fvnuxP1vrH#Lh24qv
zVL4IDX*9g@&I%%Ijj7!X(2!NMDF0%TLwd`jb&?j@_X~-T9~TW66*5D&T_YA?P+H8L
zP&lf^RoSqp6ZMJzt}(zRE{Cd}nd2r1yfQ8l*&K)!!FFrHJXVlMl*L8UlIcmB-K+d7
zmDCD4>kGy`C)|DABX;zPlRRekU%i76qztF24#VqX5IOV}#uhsTpe@@KQrPjYZKRx2
zc9`xs9GeYGHH45-kp!68q?7@IoXM0ou^9Et41^@XR($1bn2y#sFVoS-q=|Oym9v}@
zO+d|7Sa6b>G$EHgD44i7MCi!Q>VYL#`?1@xCXTV)iAO0`>M1J&GZ>3Ill>eI#(^~-
zXv7FnAu<I=LTjwZWwKf94%7zcwv*+tnGMBQ`HaGhslvZmhh$S9VEx22?)xj57#kHe
zvf>y$>dXt?x_a>yoN^N~Dw0xHq99DFSh{qiK5k59`_BkqDcL9DEle^=d3QQ3_l~{3
z4YjH{#02MegS`~Bjm=0!qkJTY0{9fu5=e*H{qF=B+CaY9kzLR&uyf1-ToOByB8@Vj
zg^@MraGs<eInf*M)xSo9uOY!DIaDH!5@e~mO0@Q?Ply{1V;DaX8V)kn4lc^0{6>m*
z!~C*qp^p+7>UmeMUqF+T*_%J)G>&mLtzSy~=~aT7{!S9@%*XYL@T8k_?7h^JjK5P-
zTnNWv%thMjtdK0_0S}5PdNl?BsVm?`5*<c8i&@*l*hHimWgX1LmZ0by_~hZ=+`9IZ
zY_$^gc6J4XpFg6Wyyy$FXWotVdu&9sQhaNzP0+g)UV4PGT$FD$Be_^kl{t4<fzOI-
zhN#o*2vErx>f1nz!xv+}IfL`PYsoFr3RN!!s8H})3l)ed*n(OmSTv#W`e>(sPwE-j
zzzd+p=k+2|hPDMkNVo;6&3V*5D?A5=aCXD(dNVv~J>t;}q4WZe;rwZCQ_n*@etY&F
z`BLovj>L^=j6`}1X408pm3CG#8dK%$)94)B0?E2B0b+8OZ3#4tu<Vd*h`w1dq#3gg
z<dk3L8BF495gQRss42;mI&ur;CLfVmk@+YD%<u~^@WkEboiCPU0(}<ceL)cNUdPq;
z;WWYeb8{$3gvS<x*fTpL<q0yBIL^_rK)4a?TE)P4DpQtf|HG=P{bC2X{n`3Adn87}
z@SA_|SRvePEBX3JC_z&KEz>ci52_>=mGPsi`DJt2(y8etxX@+LOYU!YBs$+p_z4sJ
za;E<`Q)czOAf-Nx83x%FMRV;w<s@hPmI>Fne3y9WoiJ8gfb7l$ZeXCrbiw&3@K(Uz
zE$J8C`@Oj<rs`^s+AR)Ka^GldWLV9E#11eG_nvDj5;H$!$qDs58BQq7cS$aA4~31b
zz`r{6Idui}f|hOUC%v%Lq>BTmK4FGKnGEC4cv*~Yuu*V7w>3V)$r*_cpra5$&V79M
z{ux}2o;?mR)?%ASqY9d}qm4uRUKedTcc7_+8m7wVbOm1E^e(^&$rCT^%Qc}@SGM==
zT{gU@q2eYEtv4O|!*>2>dy?{eYVe8eSKH}C9eV?Qq$Ht5JM@f@#=5@q(d(HII;q9b
z2VIa_I@U1%O!z?dWcI2!nADv^-ir6uJs8bcZ11lwxFklX=^)cxW&UNH#ed#{aIgzl
zz)5a`h5Fp+L+~?&TpKeuuEZ_4Zlf#`c!Gb)el@Ca3pcQ0>*$AeNUz%9Yivlr7)kbU
z)}DQT-#z+F!fKoRrvm5~{QiIDiVSS59RF1Tbfhj}gUyEUyCk@U58b<!>P0Xglk?ry
zd{usw#c1*fjUeTDym1MxIkHztKB5l`08_oXzO-ehgX72{S|vih95D82adS~j=eUMt
zAG^lG*ai3N`4ejj>j1Dxsp}$hxk5!yIAjr%f+k4OiNovBnX#qp_Yba@4&l~+YuSuH
zyq`gwaL29$xa))aI`$MQ6nrFPHx&Pti@g2DyW_)+sJ`AGTq4HtVrP5Dir310scFts
zMfXJp`#l%c_J|~5-j+Vu>pOtQ(%PfWiP6tsJ_hmff#`y~$M6Q!ZyR|Ti|GLmB2%5o
z0$Amd!cAreLT{GSYY@6><&+ANiE0Ozr&4?o=s?#!Ok_8RckxsPmxRnrGa>tktemeP
z_g$&J38(UsM1o<V!k%lYDj-{f=8?aj_XtXjvSxv4_)Y$+5Tx4b)C5FzDJK8ou>T>^
z1T2Y}_Dsk-;w7W7h5)F0;G=Tq?KUT>%UP3(jq%wbY}j~fj5KWr-VH2Y=v~B)-r|lU
zvH`Wr`j4hAmHMG9H#r!yT&9!A+Y#t@>I@}CH^G$19W|4Mf~P%U0S~-A!Byc(bVn7c
zV#(z9Q`@E)T*Q-eoMKU%VtopfxmOVROp^<=2e?U(>?RZk<HrfoR2nMHqeB|sqVj=V
z%%gj5Ro35z87gj*7j(rp>H9RwoeQ~>8MODy11|-sNc{zwMKcjIC(3UJAd6?kHM5%#
z;cvLG#OKY3IF_~q)6K@5#mGR-1o7f=$q5#&`3INzU{eYQ7uzSvHSeGT)a=Mc8@3_$
zP_9C$OgLvG6Tb|{GIZ=lDw8f2kJC+dOnFLjV|?B*S92VE3wrL-xV&sg&C71y6(San
z$kWD;)O*}`@bh$zZg=oBs)e>1KA5RA@aqpwQ(9faq#NRA?dYVCPB*W(l#A%CmASfY
z+xQ6j`^FtH&y6KvjoyU}ujcmIFVd#S_C{|;70k4a2~EHKAkTR@lJIeD<=APZ0&;sw
zQG4y!MTaR9*qw)|hU9bu84WfKasS#4aT%*xitznQMu_<%FW$xx5V}Rv-fj%JvJlce
zsid;v0~@>GW+|-m$t=Ufe9uXVe!6Z}VVddWBP@-!M@N)U9e63xY88|Fh^&HDo#7zH
z4LD6}dzI%5O9#zS0*ChD@WZ(0R~LY$lrbWUGo+N^xJHJHJBowyVUj#?i@u^eZTY$3
ziHdrAw8gSEjx4U+T7>1L!iZ|p8l??O9x{Gk2Q~A$dtIX0DU<^$$Z3L@bDV{AHCACh
zKOPpggs&3Ic=`KU`Y<TL!63M^&X;$8M3wtaV?(cH8kJT0zWLi~^;oL9+rnP$P#(Vl
zT$R+DFrG4g6L5@HY$qgeeVMF7CM_|WHDL(@99km)bmy8$3e|Dl1{yR0H$>_t8Q6>6
zVkI?yK#}Yit}q*jZ25J2gDJM)l6JKogzWP|FD2LV6;BxMB{3dgnWtyCX>abdwK8P9
zxG#rate8wxgM#RFFztO>Z_8HfD8MGit4}ZPJXmK8j1$=%7cgEHs|i9(w8w6E{Q6z4
z2<uzcpr5V{+1q4Y`O3kCZ<oB<W!M!PfCmXiv~pzRqUJt4sI-7A&f}<aZo+Hx3(#%+
z!4uIXs09xJYnoRCCN-9E*}&&u><w_lB)c}K&1z5ND8Fw&)7@b?GyfkGtL?9^*V~1!
z8`o%~n4eh)OWC?4zuh$j_aW<{jX$U#sXII=tMl&MA7IRh-<=}@K#YlgNwzbr>YUtU
z*gLX(NG!X3(h_Sh(N=Xr!<+&_l$^1H;xrFlmSMsbX5pp^N&(xV;&>7@?A?+1D(&kp
z{ow`Qh8f_^ZB^eeP(MS?{HrW3`_8{_a4|0^ud1sx+uTaTE?JGdxI8PKxanRoXE*N<
zk|s1Inq7sZK_1l?6j&b~C`kE|?cEQURW(LJK~vWU%RX(7o`6)Zl3-GD@=?Nli3sjx
zdM!4I_TX<Zc|oH-Li1M*@v8Sw6}XPkV~_UY!@Xou2{^bPCb<w)#aLTIKTB3|FPVjs
zRH0>Qc>}!z+!wo2<m79?W<E*>gjG~_%N}Th3Qbh=Rf^0+9(Eo<oqOTZaIoKNSQ|Mt
zK$CZXs?M9G)EY}?u1Bc+7XaZYw(@5yt!AK{=_yFn>6q6~$wVSNnl~@OO|ox(OLCmD
zSc_wNca3Vja(7Lw0{ZUq@+Q(`gCjdjzB$5LJcxz+8fH;yBpT)xFa8a4mZlt+@(9Ed
zSt_SB9i==&hXLK#K>*I-0s)7nM*Xom(sg@|iJWun!uKp+<t;&cSqQ!AQ_1O8GJ%?A
zl%%jlne04XlL~nuUB+p@=2he6!}A9|8AatkMpwU`H~;5vlackm>`?E*8nY&9d#bJ}
z#i!4R;R5UnCD8!0zX-MwfC~_5cR0nHl!SZx;=YcdPHu9t%%w&Z?vDEXsPDb@CN8dT
zn?D@m&Fd!3YW3YdngPCh@AzzSnBD6DCPh`Zy=i||9cW>A!97t<<o6BkItFCEUjOv%
z!Sw3%9stL{D+5N0Y7&7?bMa8(Z)<70NFE^d<7nP=!9D=`fMe!ze?MH^aAk*mInG<(
zNkaBbMamnX3bh8!i-7aW#K6Dyy@c5s-3Np=ip2=(Gf&I0E0y<#_NT2M|0}_S;%kyH
zFmob|gS^UAWmkA*FFU$NN9w}+x_4LJ6@f<MkhILjp(Htv1t~UiM^8Hzm&%<lc?;{2
z_C-x_nsjB*mq&ye4H%6V{p5i_EX;$(1>A{?HkVu8ccARiB<}(ub4XEE!YJOYB;vGJ
zQZ56jx%Zw_TrKpTuc-9)8y@xRi0XDo{QEs%efkekkzHPou}~o{bt|C@^NDUA_r;<J
zk<wJ2rBC#TAyp&rC2>!+9DE2Pfc0;*@2tb*y>B3gsCXE8Sv_?1tm5A9P{e4Ag$LY>
z0}XP#-*bcj_`Vav1n|n%<puRf-=ck?E?S7Yq*y)S!bL+MXdIx9ny3Q{Hfy@uymoE{
zB7Uzz424;_1h-O+?d4kp7NU1T%bhVO@HMZ03!K73Ket=!sKM-@(D4PkD<{js$%d<l
z41}~d$P8}rp|Ik9h<okp*Uu!AJn`)Vu<JEFNxIo2!WvuHaYbVn<)cM(YG!uuw-D@C
zeXHi5aS>SZo_{Cn@&KdKT<v!C^NVV{$Tbt8>_M00^|rK|L;|Oc<$|Fl+qnk44CXTZ
zF>|m&-!!Bk9FKqJ%=LLYLl6cgQHn{+gq>n3!Jh&{+82??<<vpx@|G3T2I)~tkv)*^
z4Hb5=xE<b;?1Hiof>?_v5B0&E@M9<W^P;%9k7_fO=W4TZ=M(i%5iZ(=@XQu(J8Aj3
znaep>h3{vCfMkgeEqc|={Ol$_&Y>b6^*NQ1f{Es|HwA{Sy3exqt!3Ed|JIjZ`IldW
zP8Waeq3YWS$*92*r0o;)1OSa7mSjb?3HF43*Jk6-*YqhS`7{dr>v!3lyFk8Z!;Z%~
zDU}->f)<{c$O!vq<ZA6!IPQ>7KDnE8TwK-5NGlTfDO#k!OeRKPh8lN+oK);h*nQ5A
zW3&Of)99LnMO=|?o+Tov^u#k3QEA0j<hwLrBsKm#JM{~yBzP_h=gk6x*T~!NPH9Q^
zRSpekqEa}K_L6&`OkI|COU@RUDl(e$uk4l&6Jv@0cd^MR??ej0KNgkAWS&}PA<BXT
zdd*tb0pv7s1S?QJw**RDXM;LylqZu7>xJSYQ?r^VSXA9Q@Iti$M8&mzxQ-PmmOaV^
zp5T8YVl1C1pgI}ET-9Vyu%DV4OYc>rXd%2LNwvI}@D&%Pom*sVfnnJOM*yQ<hy^*(
z{(0XlZ(>$=%wYWX?(lNrpMB;Ob`6PJ@|cS7<q|dkL;Rgr(<q__(pShj4WiQi@i_8g
z<MZ*3O|SW!L!U%PshDxv>V2#d<_J|;;eF^bAWMp8ri-`sS0A_d?o{HupYzsfO<0<^
z{etzKC{myGgg%ETf`tCeGF25JHcwE<cU6R`8{e^3UP<n6J{9I34gZ+M097?zMZbmX
zsYlvxsMdAaP17->R=$p5AOvDBX#UB(ytge?2nA!{4??-dh{CE#E-OsAogZ!$Det|Z
z7_1Q{tqD0oUEFP7wn{l;6nLu@jG*i%8%$|7@EVK*rfpY9S!RsYB1dT~M_Za7Ta-b(
z+3oX{w&FFSRCCDG=LW$NF>l;wYl1;Z?;l<zQ3=ILnMuzjLp!XsnKfxgut(mPxvt!U
z7w4r7K@(OMGQ4yWK?2yzY`x|t7CDtj3V24mkRPVoF$%pfaY+dLRl3+cbc(?!|Gkp=
z+T^A(WEP7^q|9Ce$)l^1M+P2IcMsGtjwcuu+mj`FeP9l!$rw_n?oK$wlEd0SjB(tG
z?REiVHC>>v1f(^kz(JH}ljS!v?lErW?ywT<V2d93fTOov2bS;3<X~E!HM#dvl#_h0
z1PrN)p2;hU>;7K9c1wiOBSWY>o$_r$lWMp4PziOm(OIE*(nkFme+A1SJZj&FQdm2A
zA=*PY-30sacXq~+W-0|D3f<a7I}X(cd7rDubD%AYS7B6c8QV1+sf4-}Wy*^q$2eH9
z#(0*1=Vaj#XMt<D>SBgiGzFWqi#{d1qVTkcVLqchl~qLj21|+2)ce&HcY7A)TQd$4
zi7(*=p0Dsyh<X<kd6~}V%i^v1gWJEViUrTh>$Sw~zG7=M#gqPoBOQ#5^w4jR89=K>
z`81xshoLIrRkHCewU9{n^b8f|PD4mjq?Ab|by`=RNq^3eZyygH)o!w7h1np_j9y*z
z8X>J5^yK}K{3;j<B*`ak+B$GPOdC>mAJ~w=<zN@k0gG)qnTj*MJ2+sv*T{4{=w@D1
zDkm1ML6%D&XVWSxYTX94@-#z@=}ZB{f*<<yl0rh$b)s=zE2c}e<B!o(*s}DRx`{l8
zG2~z;x}dcxa~XCXmm;}cu$hVscF<(ER9tdGK3aQSK)PF#f3q0fJJ$e#hwGdD!#C_J
zKWwrS=wRmv(clT!AFSiKp-6-$%@Jl@EH-J8F`M$(rAdgEBEhVUl7Hy!HfI(2$}QF)
z$@ub-i<S&(%~d(LZJpa=C~>f*Z&zH!PK`H+`I9DS4qNEM44%VE8XLwaOv5(n-4ndR
z-jmHwi<ukE-p^SZthu~ohLh}iB;6i3Imv}Eu--CSiz?{}BdEp)^Z-0J-}*6HNr-kK
zLe%$X5E!3(;szaK;W{x(`V`hzqt6pK6$>yqdiPbI?jZY1qc_3d?j!VM-8njAb6gj3
zI_y=985O?zv}B-~%}z)YyjK4Rwhh1tuFrI=!Fd*;VFiA~g-$8hO(4H6JtXydCw5wb
zyq~*<>34YzU|k+Gr}TP1A};sxaVx<XOgC}S{FDlw@+U&sVq!ZE4w_j05Ze3;J*mpu
zrse(=k2o!3H$nWmd_j^!Opksz7?-$6en7o7-pZoV@8fZ3UN&x`())erUT4u~L~45+
z-HqZzTR(UsKD@gV^5g6ZGes?xH_JCFuSwcUFsrq}KfR(DRG)LWt!NddU2j0_b94zh
zzT9a;9948Db+xvsf>hwh9*g^q2B+h4;|@RhUmf5U3+$s8x;)e;5ycQxfM%5jL7*=J
zmRW||oz!TO{-0f#=izI#W3`t{%@KA_d6HsW1ITyD5M_TgEg^apwoNG)X1x=H8O=5E
z#^h=Tb*2l>7M<~_t=BK3>4+<9n#UlI01=CRciXBlp1>u<ptXc->R9ipv!l2hSYIvN
z>{RHEJF049EyEb&alK9{r5XhOj)KHrV)zZk+(Pkn{?M>jl7x*+oCA)%2W8HZrme?$
z{{~YE*=j~h33V`+9vR`3w)w=LqL9=QIA|rZxu}BxSHdUmuI2ap<`4XIH)farw9CWs
ze|v107})<eE$aURi(1l@u(4=C-2PB0?EpJ9V)sO*hY=0e%I!egM%6-<QdVRB@tGsw
zAP(V9^dz1*n-WYO9I?&bff*y37@z}pd%ajV`#Kw8vx3tGf!6+o;qU@Di+(~QVQ2BU
z0_-v4Z2RzhKB$Qw-8%KB#K~KDx_#o@ZgqPuAlQQF!vFQ`n_#m6&|sdPgBt1TN=@ws
zPj{ic!kWj(#4nIJ0Q>+n8@M~tx$*IcAFM=~D~dU@`hI@@e%vWZbGqvEMS%PHexh)~
zfE5IHn$7(y`+d9N1INer<wd%ImcU-}S$BWCK^IbXgkRW2F6S}NVF7;S_ctj$z(MYG
zQK`|yJAOXj@fcQzyh}if0;BNrj;1$0Kh6}KOQErpFQUD^TedNN+l~Zz9^Erb+yl$P
z(@X|u)IrTHq!fSLtdoA7PkuTsKPpH(Ka~P}e~Xy(B4_er*1wZ&9PV$TcQFi+j#ne{
z?jkqm#r>e8$y46X*abYzud0;lD)26^DyA54ttx(-#;YZk`s03T4Ch~*NSpqRr3a3=
zj$B8aV?lVfyic^C=g1zCce?%=TKC*RCNH~DP|)hDv>hCUJfFZCRP5buFT@3wp4eUw
z;kTPglN_YUBe0fJeyU`AR7qOc=hgLj9q6W{)>u==`=9*v+juRYH~KL6k<4*KNyXLB
zuMVDtSAl4B*Xn#tAr$7divd~C6eMjPBp$MGiueS4GCPaM(H1Nlp&N=Sw+_VLgnx>m
z%A+q-97~VdvQ?84Wxvm2Pp)TsGKEoG*FIYXsNr~m9g%@&`g!#1$Qq}S73v1(-!jfS
ztV%15d{r#Y%3yRCDheJX9F_z)GTxaT(U7H(L?0N6ugy^dcO0#kJl1Ru^~fxvZy@6I
zK&Lp=w4$C-4eV4-0{#Xe>I@Mqz`f{`Bu}`AuzQw-hxg85F*oP(+Yl9n`=Ml$*<uUc
z3_GFCjM^luCP`0@r4Dlf&!)ZBQZg0mzjvU^VZr&E`Wmsi@HUHVuXbO`lME{hM}%QZ
z&-y?73t~|b5$ORV7H}x`fg7L|nGgiqr5v;3zV;bwC8+z`MQ!S5QdGo>anFM9g@#;r
zip$@3Gy+AT+Z*)xV7adJlkX2Y-nkUUJi?GL8Fu42-mOjk?WG&x2C8a9%AlX#vboOu
zfZT9UryHrP?D#HVP*ZziChMf!glnm$fgn%4hD>skdxka1Y|jKoQ;pZ)M^lqWD@kk%
z;<~@r8hChBwo`K%=BK*vjZMGW?}z`;`#Z-xw#MzF|K3BnBjslB18_7|*|9O6;r4t6
zRKBURJ&E`DmB=Ql?EB)9%=U1(0A<Igv@J<>Np54D{hY++xjcPuoO=IyKYU7#5_nVI
zUFXyJ9&D0TQ)<&2@h`g}>T$EEz4&QUZ6VNcm&?xQY_V#Y&-1d$o%Bv5>2}L0$Wajc
z)l=&Z0Q$mLO^|1P7<q`UoUB-B`6;mZsGLems*12@^`2|$8;i|%wKdLB2X5hPRLAPc
zQGIah>P;epoWDG(3Ol=xmir?-9dbcsu+u(g840RTW6?srgYN=+bD<T(<Xokm_qr)C
zUc>CVr+r=<`rz`Y^TQ`BElcg*xMn?pM8##~R}^6}$O%z9pWobUj3AVW=O%Nsv|K*q
zRM@0s(mlsfuaT`#GWUKa-$ZEHt=sg3RUajTqM6|U+_LnXV!RB!6;6`2BL^hamS|sE
zG&9VYHp7hVTyv4G<2!-WS!(&hpI()QJ&Ihhf`bg&P}42xU0>G;C0GZYsO;g-o2E3S
zhF8Fkmpx&Cltv{}Z#q!KR3x&bKErro=T;?a_}nl_KY7HVqD<AaJ<g;xXhSXADtLZ7
zN!tKBu~$8(&eKU>&FX4fR2TQfExYw<(4F{M{4hXn#{I&AT@hkc8Z3UN>;u@=%*&ql
z&V)2sR$ELo+mKh0+vr_rizPhcW#}g}lqsTYVguJN*r;%lyXt`?Ia@;6UE#Pzdtya2
z#@)Ez6bo0#pEwc3KLhu4qORQysAbN!>Q05eTz1GpDS2FgdPQ>f(oZp$www59u{@kC
zqKn8H;@;&s6Td-0KQG=(u!7=pfYc`(CJry4lh=^t25TSI9~Il-uqDG^4frq_ulQi=
zZXNvLoRY6TU(@;uVe1tv>V!v((`RhtfSROBrdH9ZTAM;7>1`Ta(9jV$?RE`=Wb_?U
zswv_XI3CmyQL5P=`C98#B;rN1q~u9QM0=F3QK-(ZvvQ91u8!KmkkcY4u=9`;Awk8|
z!Q6COJtq&}$?P}|RSj&xywDO(HDbMuWrO1qU!D{-FuxfXmR*Ly+NdrsPE(Pw?nrlK
zx`+yl*piGs8h}H6TzWJ}Y(K!M>0f6OSNVZOwIDA;6$~al(3^!#4S+kt8pB&P!E)3n
z+sp^sPZE9HW>K`pwG2e~P^=Ek{5vhLBD`yPk*dY|Oe2vN(V4`t6i3o>NWap^VRPUv
z)sR-Medw6(f`~71TqUb$`Y0*T#Z?bFm6$kYvXQK9z0t4Igzo{>nAdSeVz=T0y=2Pk
z+o-OT_-_Xbo+O8+3#esD{rO@75My7Oo{osF5C4hvl?aXvL3AUa{waCNfpe5fdpEwO
z)m<~QZ)*IU<81+L(fIE~wzAW4`eaLPIumysOIK1WMP}~LlKJlBWyfBfbvM*-bUS9z
zlset~Z<t(F^sCU)-f|zIQ7)<&G1YPpj~2|(zPkxCRIsoq>c?G5mA4C!FeGC+W%t>@
zRWWirFdv07k@i`Y2h?e@>MT#m1f%sVxcUf^j0VV+D1XWekN#5_(@=eM7oZk;$vN4;
zi5_p6$id=uAa*Y)Cy$?#MWwRjOiY8H_@?q8kkK+3^_X}Sge2mYS8)DMB9z}vKvSKn
z3nJT^x;%5y*+j}I$36p@87=S<#VQ|h6y<Hv(EzF_F?RHR(A|j-sMlgxQ3TpzKe+!Q
zqFHI&L9EdUS{|p$p0h-}!xf}wBCwtMq(l`9BndA04%LxIXshx0W3md{oTAoh;7x_Q
zi3eA&M#)SOh1U{eTy6N&8Dfd)xPSBMkNDaWq7-Ha<(g8(D-7HA`b-~O58R#x{87zf
zk~?x7^Edl7`UJan)Es;2=0N?$esc@ySLOwEH)2`GO)^Q$0-%A~8+SG;vzq3JpCxL=
zDq=k>wPus0>H48@5~`gs*Lh3Vi^fsc*|_t%7wrls_rvYW%525M@bfo!47>@-Nz+#m
z8tEfe<b+axsz!C-huLErDM{eg(?o5ZRpQGVh|W}Sl~YK7OKTrE6!ShS2~<I%V+B=+
z<DD4~OwU%BV^y3v$6==TrV13~>_`jkUuQrU)PM~^Rvx#N`<b{KC<Pu+hoQ_bmk*8j
zWd(~2)@C!|7hs6xgJU~VX~=VK<))(ahRSVPXG@%0AqZuOhNNJ3`MNAyz<Y`MTCA2a
z`g1ULr*4<#Pjjnuw+yst>hBA<(R+#7b~JnF0v}}ks(mJ{laE;$zJ;U&=m8G$`GtZA
zBSVPVagyqxfP({0EeVh6NvEh7)&?zzwkd(6hz=POihxLF5jYiLIbpaHmABcUk4g{`
zIsIndBE=(*Z5sgA9W554!v!XEws8m*9B;jBCKX8NfFg-f-zNmRn|q+>dt?yjU?Gg_
zX3T}h0iFSPzy&?6Z=SjentH}j^GoBD_Lt3F0ga55HuK+&#hsbxkcI~Ib+^XJl^YHX
z?b6)^As1Rpc!B_f#|z#hK5fOO0{@da>|^Y}d7r6KN1*w;W4R+*Yi|(UJl-)=jB95_
zvam%{VB|I0rv2bGrH$epFOQ4krH$xTGxJunT{4qCIw!ZdpQUf^+YO|28}knq$H^t_
zZpb-vP6Jr$pr4?PH|~%nICN_ULTf-pjA{ys(qbXZ5u~gW7vSZ^0$lr<hyCbXIcd<=
zByb$Y`Yh>z`yU+k6v}^IyF%+xtW?`Q6E%vW@K>*DHc1FB<P%vLJ75eDhw4slM1$jf
znS;4_&-EPzMGMST_#b}kQ(<59#arq%fE46<9|M%IG2+|mv11VX;KR&%bEJETXwh3<
zlF?!-(^;BaY><v3^6JV-24|29SI>8L?7>TpkM8nQ<kxto>);w^JD{a&a9R`3qtC3-
z(@WuCu3c;AnaXADQ(N*jTiniT?&n*7-<%B(xLeNF{XzNHaTia8qRxwfpwMMhyHKKe
zWE8S`>kWqWJi5OAK|Fr{7VUf|vaf9aVfDgF{~zW==vn@|+1DlYspw5snC_WhW4SV-
z@E1@Y09QvEKo`rZ#+<_3VWQB(cZhtwp++OmH!cc~YnQ<U%3D|=uU!08j0md^T}#YF
z3D@~Bq)l(x@qq{Mqu{(w?Bi_uFd#(4tmmiY_bYW#z2nyDygcFj(W#G&i-VWI?IO@A
zEJ*wA`#$q5a-of^T*rq}$m_<*&Gz~UL>;UbzJJ7(-!|}A@W#_j%nmIh6_s%9-Hz))
zPq(M$cbn~!Qmi1?#^=)ml8z~epEk>B06j*!jt|=oU5`VZkRdk$<Gi{7r6E6(W_ukg
zL}ZP+;bTW)F!PE_`{*x`;VP(jrk$_HoB49ArKW4Y6oh<&gi4ov^tCL^$1quyhafEH
zT0r`|nk1@tkLPb!l+_a}GuC_4xg8p$1*j^^JR6aKkSK3Faa8Ssw5j7vZ*}q~%w_5|
zB2)chY+G>W+ARFR^4`fP6voA8`oF<Oq^7Vg$%E2JdW3901Ddow_WnZatlQ9Zg~W@(
z8E2kw8wQ@kMm*`?GO%Sh4)1aI8DWoxlk$A3Y>vE_Sy>zuF%$o+wEm`bJ7xZ>Yl-8W
z+(ZPwk-j-VL`SDm@c^K(gdTFZQ8vtf&{1B<y6OXEu79aRO>wI;$vd7wNcUz9LG8EO
z@XO`B8P{2sL5)dx&ug8jUX7GokHxt(#8RF?n91nMIc7<~7w>8Cah-UMT@N;zc;7&R
zXKGb@W`v(jb}8fP84Cr2`^Lr^hdT5Ru1%Kf-(%n~=+RH=2Vf(ztVHe*!WkGfF|0n<
z9@vvQmb;}rJqN0+n+*mzd1x!&3$_3;XG3dA^ft`o(`?Ft0N``thp(;z-2~aRKbcqc
ze~AsuX2<5FFa!YBJlhsX>Ias+rrZB!?5kbCF60R4yRY-<y&J@%7xOK19*N~-iNf9+
z|B6MsBNXD&(({cgL5YBP2Uf*zLmAcn=>W0L$`zlFwLo&4`=x4GvF=Pyo@s*&1x@KM
zmo(urg#+bTVE{7hV`l{Bqd}n}b0&16y61Hr3{SHI8DXkZI$b+wSQha>nLQ$Z28r?Z
z>9zi~ba2~Wm^C>*<Y}Mz4c1)by$%<MK$5k{o-tQyu8^A+(i%|-By$}nGxXz`{Ukt{
zuB4Tr?VAvR7QPF!kQg7S)&l#TZ)68YdL7Ih#(G}mx7@+Ev*4BDI|IHtshBvd!@@Sb
z<9y`Qk$1D}nvLRTLb*uq3R^3Rsw24^lsA{U2EH;|4dkB=S`8p32XVpjdR6R9o=3E3
ztSGL}m;dco#~)^K<GD;Cb<tk;hMCkBh+hLCR<5T9C50?zQ2YT)Fi$<v7$c@z9eX0@
zC|K-+dV8V)#5+d81$#gKlqa&2qR<1yfD3ElDFB!t=F1Bmb&7cpb`l=wygWwIzF_&M
zHt2lDwPF=4)R39Fz{#9^?Se8ANx&~sz^sBZVU|qPv&%l3d|4EWOOs%ebz(V(JqI<2
z*eyPdhR;w;Q}&7uUMmD`25x}sjIlCDAEvdY9j3(XL`Ra;9E{LK%qwH%X?=1oI&tCF
z7(P6&y#1cM(qwLDui`=FS-FAvI<X%`Wd@G<YGs6{o)!g74Hh`JMXdvs!y{o%Qte;C
zJ_cD9bUpl1Nr;Z~IL;Fmuc?Hf9E5AMJEjQBYd@z4J8lB##%LLmt1-~QjJTK&Ba8{#
z;Mfm{qg~0qm$g4{EQWjp(jU>ZEu7wXyBR8Ek;I5gFO<AYNs~&cD&BEU4{X(EIFHtA
zSoIR12F>}GpDA@S)(nWV%b~OJm#{6RwqkPTeC#8K(K2k1FAh2GBJ*b?M20{!d%S+l
zRf%Gx`1zYjqr0tJSmS}FN<Kp1iA}<(&2#%{JIFbDdQ%$KNxMZ?QiHM+pC>WhThGI(
ztaza+>5+dy8oC#|z7g+=_5DG2W$cCXSpKKC`3E44asgF_!A>VPLTc;i4jlxYYk?yw
zS^rIgBY;7+C5l2A*+2YRI^|z_yF5l@w)e$xP~C^RkhMjcJkN+3(ch{@*Yp)F2^!t(
z^*Rnq&P7owJy|Kh*4|Kw@pbn_Vq9N-?|I_u-x!F?^Y@6uR}w!#Ojr5dil+o&XkBpc
zTdk`xpv77B-lv1lkilM5ka@j1dWb217L{wU=w9S)Y7ZCj`7WWkBRx+j+-NGR%dN&a
zOZ}WFNSyE)_{6Zn<M^=E&SEbyo0jPKNThh7;Q{wF9VAoj#AQ-7muiWOuqoOfjNoy|
zaK@D)NG%6LU4nzQ!X2VZF@O+YCVR5Up8949pE953tvQ1*L;NZSuMu3c^>Tvbh}Zwo
zyqAuj!W`(8*7;)Wl+T9VdsZ#Es!%|A*Gqg^DF>^^Rl<5CAc~{Lu97kC0BGSAc_E3I
zWzq7;qdh!7M`RW>l9WwnFtqMOEPojROn>3J4|^q^I;KYs8ja#@L8VZp{#2W<G~-87
zNCO9T&(41Celaoq5h^*PY(d+=Fkj(SJ3S`@V^{!Uu38FFb!L{h#-jKEKqg8H0x*`5
z$g4U8bXJkR8-`Mv+}WGihWA5ttl0}=t$Jlh2Hy;6Ss^32J2|JF+ms}{+*+IXtb2B=
zs>C`Iyl*&u*mb*?5K>u01GI#zkFiX_hLrmJM}cP6bM4=2)n*-2bzs4(iE9rZHLSaa
zQd~@unuQ*8P+$Aqce^ghNhB6kgT7YSQkov_mu^rKTC=agmMdR$QYO=Q#bh|ml9Kn&
zz1<r+M_Ibqf2b&!f1QB;6W{;8T|G?y<LY7lFR|fE>T)qCZHV3{Dz6M~cdJTA<@Jt&
z?OgpN@`MDt1Y_ag(b&S|rYDUsNN+E1<2+rHvkfEl8^0|rjMvsJT37Gq6*Ze%g%Xry
zVK0hQ-drV=I8pKCF-*y=qRj65yjbah)J@9d-w{2@jb_Z^@=f5TqWAK4v+)iiHZs-+
z+3D_AbG?y8ERB?U)>U@WbN6%gifzNM^cCTz^3V&FBfHt=1*anHB{ah%nM_Bm0=gV<
zbLM5vxB;g5^8@|nVWIuL=V7jr&>rMyKn00^#qR(Xk<#|<bXnA5_8<aU2Osv&#(2O;
z#TO~8QTsW~G*K&6mr`L^8SNzd;!|0&lQr}!Ns4u)Dsf1euIZyjbVmqN?EcBfyGBxr
z72g%+f%1_ri;CuF1Tt}h_2tnIp}?|*Dof6Bi(vx~Sj%C3*-j6qLl)oS6$GMBIMTsg
zw65QJGHVTOGU74P>SACZcts4NaI5)gpIVS8Q{pXNv;tCH8SLIljfg^(HsSp=scthS
zgWt}s8xRldE{-TY9;wp@0g4jVZJwLDQP+dFnoC{)k<GQ5o5IooQ>~A1Bp4!3R+3a0
zXfGIY-vXbcH=;SX)*tJ2J{gJ_z)2qicHzNhs(=JMVO0PVRgNs|<pDh5F#(WB54eTm
zk>ROO0rg@qIf~;!{EJAqa9kp!rj-247Kv1Xz=5YNJBX0mJI4qj^*1W|b`Ob0D$7yR
zsIi^56+=&RegPf6`@{O-)2GLpZAVA@`Z>eo#(R5YZbwJwo6X95RqLz1tDXE*P22Lj
z{cYoTB!Ux{d#ih0r&fa%V}h^^<d(|lH3Zw_^?K;|BXBOq%6pw>#|p2AkHt57H=B0I
z9bA9%=M&+lKlmSg*~qF4+t24p58_Qat~E>L*HH(XjdrBA*~7v0)b;sV4V{h^^1qWt
zw|o1Z9c?Yrbhw{;@1KQ+uZb_#rL)0<ts}exJ`vBweWS@+c5*5VQyc(%7H;2w5O_A3
zYUi44s~7YPvKTy@Y@3~%lClzE_xpQ3jaA#W9qsEKu%0)J{j=w*>#e1VtBtVx{YAd!
zo=)$#f$J|-ZcrY*EbZNIs?9L%;RVN)80(c7Jl=0DJ}0WrI|Xo+o%b@b(ypxdzD6kY
zE8Uf^TlPFK6fJYKS}CWoJ%n-=2Qz4XLKXxfCf_w*kwcQ%3kE_Wt1ni<lrEUA*aLo+
zoNM8H%)yrUn!ep8`HamEVVqxW-JLi3hzh6uCqybm+zvjVqJ!bEIV=_QA6I3=n{%25
z+RqvgKFEL#>~z#is+H@W5XelfdKq>OMj}5Pst{9iwX$r$JceD9Sp-=b(KE{<l)x`z
z>(|x+mikh|+&78ZQABTE!DU1MK6Eu=XhXy;P_;w+SfwN1%Jt*n$R=}#ffErvW>Z)~
zLgK+$3y_?H7A%v(4iK>v0-0nC1fo(f<8#eceoJvR;IIT#e{u{$tTU{7jQ&|fSd-XL
zH%tLq;lwP=aMvK$Kh2RR9nJn6i-23E!hZ|szyIvChTvUOQ3v^<g8*qdz>7uS+J=&j
z-dZ**Qj+Gick&nLW9nLI=n<%@SJR9`e?E`Fl0*x@6w#mq9RUFH2Ty|HJ*-z%?ZTEs
z@+orh6OV&h(}NL61?hp%MTLfoqoV#s9*JA7phO60NKKh47y+DMm0N&u>r}b*gGB=L
zoD{Rb3wdL5E?M-%cUJR{Vl`o;0{9$(_cZQMLV~k2?*7;gH;*KxFT?{og9AAdUG<6%
z_GBz%&wlP@c&mvfaK~u@ZpXUW9I3yjsOIqyc1^2B0J(Zh;&AM6!rA$(aZFgUvgH`#
z?1ex;^Km{ax!QM@b?88#|BJDAjIkwPn?=XAZQHiJ#`f&7ZQI^s+qP%+*tTtZ&&=8H
z&AB%xId5{mmCj1{zfL;ou6nATs;C1*Xd#w0{A0Rf{VQuN$GIUWy(LpAAUSRscrswk
zA%n>%b=EVV1l!nh&Ph0!po{UBXFeyXFvY*vEU@YUzQnvx)(2lP=tv<#W}u1bRRu8#
zYU#61SIAT9Ow}Y0xHWGg;|WezOR6d2g?rytRJ#pi`9`&}n&pkI^mbM)!NEF$Mg|3d
zGZk(PLSekF?Lx{2(4(fEQuCo(>DgV*`_@*Xoe_899k*i761;X*Gv!pRwbpm#tnrR{
z7$4Z5KOAazHN?H`E)H^C8TLg-EQ~v_mAOHNlzPOq`L|pzAc0i_A9UD{0rrXT6WpaH
z7MT|%XvhJbcR0#v#IF1=e-PEBGM=Vm)rA&?pGD1`>GAIo$KCyD$*lI)Mzd>xenw8N
z;3e3QA;sOP^f4^)pT}|d!k=G%D^40K-(`i*f$($-y}x6(UK6(vhYvIXg}tyV4rux>
zJ*}$upAzwO&co26I~nZ3q0s?3RCo=#$lDdM0=&K`wsX!S2Z@}M?Inz}Mp5jQD89l#
zBSAZ1d#LHp<R|IRYAUi9vIwqGTG7!8h5Wg&WCI{lj^@Ilw1CCR^Qt0suF9HVFXH1C
zYE~CFL@y~D^HxHWacTNb9;V&QSDYK^-NBY`4}S4(Agk)x(`s6cv%4KO$3mg0ew%gx
zAEMtQK-?g>Bn;>7Riu5k`LR3JU9%gE6z)=E=*4ZZt=cCOWbEYQ<2W~%|2tN5%=L_Y
z`5CL8@KlzNHxQT?hAv?_`-bJ~+|weSx<v|WVg?}qP~?f1YX((F&zj=1Ut2Gk>`@wU
zIl>$(36DKB=3>rRz^_76NB3bkk;U2Tj|c<Hp`m-SP)u>0qqN$AGOYT3CeVC4Kt*$4
zW|$I8a#V$iRTM|%G;xF1-yBk^ZP&rbz+uiZ&0HZgB>XQr@(Cy(ncqMoWW7@9LcncB
zbG(o+vE)EMKy7nWOY19ZR#&3mQH4i#iwY*Ood(X_Lw-zf8?rraQcCge6SPwOIL5Gi
z*v7O1_z_NdCO@1&7cM3?t#W0zf`3)8S1xUL`f|>f3-@gOv!m|s?B3Hyy0$_`K9z>o
zP8;iIO^wPr255aq1Gplt`^%w5Ua8@hkF)#v%GYM@lLLt(VTCqR6D+H0l;XKS0*8dx
z54&*2f`&N7un&BU#N%Tr2L)BbrCcJW(K-CB%J2xMTyPC!k;W1a^h~=$_E;H16=Sj3
zRV%~5ZsM97Buve+y?s^|Ocv0{tt|w9A(DkS<3+BwG*`-5c|An26;`%wrifAYoQQ-w
zjRn`y6xb#8ydrC8V|;RHzac1991lTsDiEq}so+-tZAF4z#qX6CKv`TRV~=;N`YVP4
z^<s+KG1!zb79P?TG5#b@m^&T+05cP!-u@S~$HDx6S0%A<{#$(be}4Zz>UR(Bc-&U!
zy?{Xhp+)9!PJe#lO|Bhi$9lR>!cZ_WFC_1n>_4+6S{ezp)w%p3Fq@UqFWc(W;<r1K
zy`j^z!_UdGXw_g-g<$ld&sX3(@X??ChWWDr5GGP1k018G*Qw4%cRU9rTxlqb(S7(k
z&wi8xMu`2Kd-p+KGc^4ektW6cerY`q`|HP!WlIdiPXqTmn{$MB{6scf*FWFitC;_B
zYtGx<{l~3|S298q<`B6cBl%7L%KLX56xd@iLQL;2E`*9qZ1SlZ52X-<6lZ*e<kd+Y
zV6%v{Bl%DZJ!dgMG8{>lQDCLe?h`*5&grp=EV(&E&T~lags9_n)&q5*lCiR;b9-0}
z=0$qF??;8FpBmgJjbPd6`iBhQop8Y3$1te6CSAc4&_D9qgoN{3AdPdB??X{)j}518
z@j3)Oq7ofYAZ=<?#m3Nx?{^sq6~ra(sjOwgCM^x1M6KgO?bK8#o{K3Zw&HwoRDjE-
z?#^CvdPBVUY<{+`2ZC&>qzJU+6)tCuEX^`a1k0tMik@&QZOvbtVcwUIhc@?D2NGdA
zRddAmt6R1n#<on)U;%c_<U9AUKyET7ymYxSm9CMQg#VaiT+^A*>V|^032>gG8PcJR
zG-=C}3Yn<)X6&%I<x$CzCpN~+;LV~Yt>WaIE@caH1m|z1e`gMpHt~j|S?oY2guydr
zCLDtt!87YlpNg*IBb)Ay)eluhejz<y&c?oKq<P)6V9&KAHM1c19=Fq4zb=2pJ1(MF
z=rtY{kJfEfg27o9H+Z%`>Ih(E^i9M)KMzg9TTq45R#s7r)Gq|%&PPHmu8mm9HCq;`
znM-nC3_plo@a7#1sySi*5b&t6scMilKxyocL>C@2Q9lhHu_{E(#&CB6IZwd9k2Llr
z-jVxLEFDrCn-Zt#W(yaVls!)24W;ShwnRC|1_-`PE@7@?bBno^aY_~4d-p{FPn0ne
zLfG}|`OBRT;YU!2Q&=K(`*hbw%qGeNYci*1+yrtrV5Kkc^*Q?5=kp1Qh45yNH2{fz
zQ0qQa_hf<R<uLwr$I8D6gr0&I;3>!M*vs_(gu^@6zi?;kp)|;nK)jc-l2?G9Y!l8<
zhaCeQ=EC7Z2G&wBheY&Seeku(maLK{YD|Tgln<QB)5hb^JsS;QSe0*SE!Rcf(B$j>
zv20>l+FQ803G-(zG9?o@25pu^@lwbQDjUMg!X3t`S8s!96Pt0p1I3sWmzif~TufEH
z3_X0;Qb;`%2>X1Gy@6rhAsaNAWr^7aX)r?s<r&$p0*Zl^-Ce(9M?r_B)05o2D<P@U
zW^A_&qYy=k{-|Rz3`^%+cK5O!Re1cyQBpzpWUvgT<Xb!WYS1V$A8Z^H)Vi*TtF$ce
z5Vg9NHmHcPOyn4p@)tEGSmJPxepTw?l@z+JQY10KNY_{Ed#9{rcpQ0XuO(&)rMT&=
zcSTy}*M=#1srX)#L7zW{D?@3G>5u1rzjPH6rXN56vgikmID?8CT+pZ%%#)Ept9r`}
za7QSZWx=@{6Qr@)P&mJ79`N=?xK<|Cw~|@1Q1?rLGN}s!N*^uOg{&{a6<cTtt^g@X
zyega3!bjaMh4tE@ipw)vDWTb8B7+AokUjKjX|(HPJxbJ6r)9d%!|I~;6HcHpObku}
zvWqc)4*#Zs18G;!)mLVq<n%r?rp<~Nx3Kall6bt4A2cp|HJZ|^N1*P7n}HuLL@^X{
zCv^TImzk<FV<WvEawEA;gOSQ;tRrJ(z}|D~RWPC;@)Ap2Vq&#tPo*O>r;io5X^;T~
z80(HfuOw;z*4<f3;rt>{!x>fbs_wktUPY(Ac8xXA4w+NnuY)yARvSF3u+#amK4!)&
zvS!PUo@d`F#rmpyBTVK(6$;xty*Zx+2|?sV$7_2u^0AP`F6-yfEE1HoeILF1I+jIb
zK#d;YlX$uVJQiye{jRfhnbrELmA*7_c}@3WwS$vLRKKjJaMCuqiQtZ?1-_}_3AAV<
z7K;_voF6qvrLnRskLsj%jW@L%ZlJPlPNANrl7%Y$SINOLO<XLKasAMFR4XmP?X#Y;
zK}ecGFj>KH$PRnXDGf<W1YR1>)mB1e@4yfg^QK6xROX8-N?H&`BL0AZMt|@tqz1iZ
zI|OT5s9zEw6>b^~jV+4SafKQFq?36~R!oMfyF*+cz9Efc5}|}bQis#gsXnSol=LmW
zAxU48-GuC(z&vYlUdf`WNg=wcqj1(AX40$RzDOnnHoHj-?Ws;#%OYf97Me+FUHs&p
zK*Y@L>?Se<;uoW(E<enY1|IsUlgvh_p?c@$Mb0UAsdJtxEG*0?c1x}2urnhEW+=C!
zlR1uMES+wl5)GrM(<W;SePpFH8Si|Os{D0im^aU`kj9)7y<wV;_`-C<Pl!CSPlUZE
zp(;Bypm@ZoOh3R%O~D7VS=(qfVN0JWdaYD0U|VXk@Yz1loD=7Y<c(kC69qHhywF>y
zf%_0ulDFVKeB7Q3XVj5!@KAny34ee^jNbTLWO#6d_vCROoK<=rw978aXcV1*X$X&l
zXTu2Pn;;U;B%eqFRC1K*JP2WO;fc{0zopqI`-EVq<Dq589?V%1-C1oY#du?4x5n6{
zdiG)sYa=R&KfTuliU+<O00^W$q(owj_RLI(f~NjJR>FS*-KPQ1zkIt5s^o9-nn=w!
z9is|k$(GwhEhQnxII&XT(8-kH!Z@A2<w#{#C$`rb$mGPN#Mayu?WX511P@)pxtx0(
z8dsJaNXkv9jq?`7A(oQly_tpOxxmtIYJWDV{d1nq_uS2>>ASVF5#c%xo~(&mUDV6y
zFQ+DnsKj^_Pra#&RC+LrEqFS`VJ_ra$tWGW{1gLL(L$I7&<f!?joB)ry_v+nOssZr
zgWKU*-xh5%>-Wtzw$?KNTd#khGpA7!-gNuo(~rqS@X?8mXBOE`pQ~!mbFUC7igdou
z(z6O@Z(nlaLb*6Gn-a@3eXw(^>{OF`I|@moZ%W8!bHVz(Wu_3?wG-btoelch{$4qm
zJ5)pHw$6V;nT$@Gj$sV4JnQ@mUi&=chw^W#Vx}kg>Xm*>HIpo(R$0?B&)3Tf)!@nU
z?@La6F<(<X8i?Tdx+-u9u|1tyVi%-5-uAUgLjbS4Bjk^+)$wt#HzM|*^iXrTsRCLA
z$+Ququ7{S{n9dzQI+llFzKq4OJ-N9aS7HL_?xssT`BBGo@8twY&Yq4v^zeGfs+%m8
z@21n@gpQtahY>v&lAvA}!pWOhZz3gWHcu+7bAw;lnShHCGGa1NY(^J-oOW&QK>9m{
zaT!>7B1=DnSZI-~qQ<GcfSuZjgTAJi5&Ff4QF)n1LYQ>T27ZQ@3vt#MF+TU$;B&@M
zc`A7(yCka9<Y2fW;CL;=Y$~+4uFBDC23#(s;Bzhb1sN#9nAzE-;%ov>!x=I*?wvKo
zu<FB;_ArUsy3CA1&f8G%Ggcy*=I0U(wOi`7P-&Qe!q;~c0GQ1i9RnMVpIo;$uFZ&G
zV`@cCwC+6Hl5a#s_SqwEp=~5iw59T?vdB9DHz{1{-R86GoopWrcDmP8FgZ;Cjkf4(
z$^M*qu<WF#;JR)-NKK_Y_~qqg6%OKV!Z7n*1sHq1iNfCR5O|#Q{i2+=exXxz7wqf%
zOL&lQQ2yWWG0Xq$4rSqF`XAzB9qD)-ZiIi8P~HFPHeg$T1DVwnxcj$ImTZ~Ag^iN9
z$zpGEhyH#Mw=mv%Qg@;Dn+ozrk_al3iBtQ$_ud^p-n_fM?iU;W+h{U6e)zCOehWG@
z82E>ghhYX{P$n;Dj|mC6(;x(A-+*UQ2jho#5&E8u>OzJHKz5<x2_4sTkk$$vVv^Qw
zU4Q_y46(c;Zsg^Naj!ed{`)BCu5YK8yY76<lNp>($-N_GX{+0WU5TzjD~a>R*Qas%
z0Vs0Z9a4wsug33>yOV9(X^A!&a#bbY?~k7>%|7yH+<|9~Ce1f<sJYZd7$$dF`r=Pg
z+H-I3A%rO?NZ7GbKqLkHE%+~Cs1QrCCibyCYzLq(^^#Md{dwidWKN7=D8f0t;fV4B
z1{bV?`<#A}htDPfh%)UllVCw^a>Sk3U^yX@c0e4!9y0dJMu^Hnlw4X0Lc_bWc$yFs
z7@w`k&Y3t08Q8w5I}uHHU`p*mW!)89yQM+AWJm4QM$F2r^gh26G#))=I3$vK(Dvl*
zJ%tr8y)B&ucR<o>TN5Z?9b7*c?;bG3zUAap$?p!0;%^3@-b4m5cwzAva~IqPIrac+
z0zx>oSPC0a$r&p&8C*uMx^w(55pJ+T6o00pru(<Mtr#<a)#&(VYf&jjr<S+7D?dG5
zv0U-cbj4XvH#;K6l@Yz+2WO*g!Qu~X?&D7q3u?i!KO7!>wi`ypOIplV(0D0E^P2#?
z66M93S!!jG7u6jlp}<@adi>J|k3rW|_7KFj8Nxo2G^1P}Fn%0LOkvQZ<`cyc!d661
z?|}3Y>vyj(|L8z%1uD6XZa&e1Gc?o{0GgmHnzcO)d~&U8l)+JW<IGTF!sQI>8{H0$
z1Kv;(0flX#NnBchm>n78>_Hshp3`3;v$;7*>;o0)Oj{d_kDiWmMg-2ScP&=dx1Vng
z5$e)yWXoy9tr&zNHJht`ndMV!W}?NE7lmx-%K5jm+9C>EX^7cHPXTf0J9&Gbw&)k#
zm}i^|b7lx%L|T74ar)ELX2#mNTrUcCub+~%Ho-=`GgQM&B5)Y8@xd8)Y10Syh`qjZ
z9ole(Qa4(!2Il^BH?2$a24G97ccKeq7EI!^fRj3xux(l#=Op{5YPuvZ^fE|wT#vPI
zRcBi0k!^8nBD{z<yT;ljJRE5(8AM!Kgs?<EN+taRA;_j;6@=9Z#ar=K&yjW1wIG8r
z18a$Yr^a*fpF~$eji;Dw%)Xtg=L7PStE*xxK(_^wtL`}-ZT+odYroa4erct1yWwZ0
zTjK`|_o17n`>^Hp2F@qm6ObC??OUY(^yS42@)hn`raxu!TsWVm<=vt`MFZf|$Ra4i
zJ<dS8k=RcEbS1?f()3c|JkdCl&@8Q8fh+!c?sSVY=hC6=>FHKT<pZZlx!Uo&`PQsI
zJSBgU0GZ(J`D+h7ZL$6qZ(IWW<oTwIYc9379lqP=|GBx@2R3B%Q&Q_QK&J|O`#bMk
zso<fcbu#|R1}iGo8RldjVvY`v1ayrUkq+K|5}|l_y~SGFU$M%3PiVr-fl5UCBAwDA
zlLb+%p|$kguQ@ifibdS8vtAe=Mjl%tiI*oiPwKmIx`i(JWxnpK<7R;JpXH>6XlE2r
zObDh+pcO@QxM_(<jP<P$<WyOt1f_cn(Z3SbwD#9g(g^5`7Kxh8TrDZ<QBxfyiM*V|
zHFCt9=qq6_@OThf4J9X@=_1Nb5x5BA6f`KE$on6k*XB5{_>#VMy`(w(+#qvi>$F>M
z+UPK=1tZnlR-po}vpr3H6W`J$Hf%GEt(_1cT-#|SMwAmRbwR!zIB>-2D-PgTCI~A5
z`H5&}AcG^!;=hTX?Swt-T3J!xLsp>HJ5|=<6>3D`C5nz?Q{L+@DPKn{?cGdsxN?*F
zWhdT=uA$^6Qsm2rYweXCuM3>9uohZJN|dMhS72IETCaEzSTveIIY+e<@_F<x#$^+w
zM~?XV1`Gy^@wN86tDV$wSe8zbt8uJe2rFYpJ3j-<?DR1`Z(#aO!xo%bNy-vQUa;Nr
zt$!QX+Dt!Gp-WEw*q(QloHg4Is_569rIs9-U5m*pf$OpzS5pKBCivb`+m#NrS`4o<
zLG1OHR1y`F!sfutCB*KjPPTufOQJphDQkjOp#Cu6p-NHJcbkS0Gr1=$QfHzo^Ri_`
z=$Y4Aos}LfIHo8-YZe*4SKQeA;;>&4&wdZO<9mX0YWI0^Xp)ugYKG=H$JYjlnX7Tm
zhKbQQOmf@fsoj+0V7KH|O;~k>9b^i|T@D<8lBv%rD{ePYw=YwMl8J*EM<AgBCt6?P
zVG2NoKJh%IxzL(i(q#$!e3!o|N(8&HNiA@u+-C}Wb<MH4V^B3^u`lFSATENFT-Wji
z_GYivrxjE;#dxyO`F)w)VTWgQWxDtfm6AWzY5L%}6bdgtE9{+sy{iaZW<m70F7_w?
zI8cz8;bH8)0*3{Er}{2!dRu?gw{j_=nxQk$%V(!hwZnQa&F*70wAZv~7!J-1(5k5+
zKo*@c>#Z<lSAq|%3vM41jJ(q%r;Qq?7{-0^{mmDwzt=h}hesf5lhfbN&wnGKzEIy2
zW$xD7v1q6s!<hQxbTfa%`GM3F>2gsn5&?G+lTt>W-3Z*#$aP3(JrfO>TuE8c%5*|B
z8luhX+$n{xly3ax(t^_+CR_@zZLGn{9BF{yn@Boz$~lZ_t6=D4e%q4eToWWG&@C$v
z7|6X6;@)E1ctnmxLFzrWI><^Yzf;Pbb7C)SaKD5zmdm3Gd(}#=K{vxhA1zVrWEo{H
zaEUa@aXMEoDurGsoL57@a<o%%6?Lzk<DgOtVLA2Ok(-o04Uj*yqSY@=yDF@y(+0JB
z>PfN{=D$I>ics3Z)oYwzqm@O?%MC*Zz1|MBWRTL1A}~E5BIHe}_pN;W=b$m&!UqP^
zf_Zh4aH&NklEL0GYflu}dEN3dn87Md#i^Tw?$?ET{lToA)azA7Zk0H-SCc~cM@#nA
zhfjc|cr3B0?0zM1ZWBwn6GBYjn#%;uq*P;hSG9Zs&E8v)OVi{X*qhq@6Q6?u?+;^i
zMjMn_X(^A+3yqn+UobFy{VPie76Q^<<gRLQ*$1UdyZ!TTAfsvy;dDqBZL(_n>h6)W
z;H{=3v+NK<ajNrqOa)x8w0rdfB?YTgF<-GSG4Sn=;g&d*tPvh|Z6`@jjkWDLmE<`U
zAu5L&%&%j=5mhpjcCNS+dWyZl$S+E4x!qAIwt28isscZ8tY=plCB6x|d62<vm+D&A
z{8qf_Tg%)+hY1MdN!3)X?q4+a*Jfq7^fP$!x-dBvdtxZ3wHRN=+&@CPt-{thqfS)!
zk5TJBJU+Y64;F-^vu_^7+lAX!6jUG4%^G`MtA6V<Uz*eNt?2U`dfknMR8zJ$YI)iV
ztf1zqKKOCwsI(ge)owZ9gPm8Z1E}b--048&)`S4M_vk;4ph{awQp_|JIEU+-#=)Ad
zv4p!lBZRcNFe$+f*FSmpmmsC=JWy|Ip=RV=)WqG*Gdy1L{7O{rFCVC(RU{Y9J6YXc
zt|r_*ha=BFE1Ey3{|(u){vSI3|63%-!u3Bxw%a(9c3WHzJO3uXOhd-#1_AS$o6?YL
zIv~3c&RAfmx5QdkzYsq^b%xWZWVWwcaTS;#Hj=0sHE%IX%F3hxgzm5LUG9*@<5h+O
zS?BkcV1Gp({pa}3u8x3HMAHfS0Daq5ws=1`?~>Z253`THdsW-JL}0nVe<4VI4E+zG
zJ-|HSadO?)R8sz>OY<K&f@KU{nmn-(iUf*$?h^iVEP`v*6ejC*!~h5h-2Ja5mTaG3
zVsm<ZcE@2w6tSkzVw@a5?oXdwUoF8lS{u2hfBg95hsI6^yMU&6ud`cVcLL4A7LIgP
zsp?>E)6I-CBj1hQcl^eDn&tYA9evk!x#;njrUZHJsGP&8W)=93s}@-%@94bK_Q!QP
z>xv97@k2p$UK&m9u@ArYzxp>H!Dg43)etDngGJ1A3T8OqP!voJ*5hv`dSfz5ipk{C
zIOY^3l+xe>`+%q)qUJP;LzdxMZXSCZYKf}^n;XKNFnvrMl|=`xgqIp4&2hcA@`YHv
z-t(3-f`*|~?rxBtrO8>;wIX5*Ut=uCE)v0Hpws+Ti>#zG<=5ko=w~hl(!iL7F^{9g
zLsFJ0t~J%pnX2;2>!!?+MVI-eK4|kzqk3oek`*+jxt_EQG@d>{<@uHo7GgvxG!yR{
zzGbiUVq9Q?EK(3?-EjD!u0SHD6mT6w+7%(Y0(R83>+oV=pGAo;P-+-?K#HrqO$0?p
z|HfQ7!v7R1-9x0krzJypvE;HHa^oY`=x||08ljQDyl+kr50pv_wJ5A|TK;J%GYNAW
zJ0xjhVU=#=4>eZPgm37<ZYH`mXTg9D)0@&a9Li>sZxrCgW4;z6MU(rYJB}u8x}Sg$
zlrSTJpU;JzKGm~=nJ(+8Bv&3tr}bI}KA3&2)MClLK-+PkNNI1u;zkaT0~+pFms(rs
z;)@oifEX5wPe8eOP4jj_CQe7!_abkM&c{(*sr$2r0<`sVbcSZEspNK+6cv5mG^SP5
z#o;?c{ZdW)amVH>P>$pvzB^Pkm@58@Ghj7NL#v2JzwY7TPf2hzJzLRkaPq=%=E;RD
z$#%y_=;N!TuTa*l6^lwfeYOiB)*t*bFzKB$m>dgKWjs&`eYKtmh}OYc<4)QzCh&<S
zR4lAU<Q0|_+txzIDRowne@3Kmbm)UtuqJ^556cMR=BVT*;z7k<tAX|?Mq@e(T=!AP
zW&!C(Q8_!R*nNyRLX?A|>^rNglz(nuvb!j9WFrK*lTKj3Y83-5iWCo}cC2X<P@o|r
zPJO&0U4|+(-keE_VqW?h1y|!(krv$>GI8+DI3?UeAx9uyFsAIk!K-~VWBtB8mK=kA
zK~Ti-_C>NFuCz|$m~a!{(f-9~(FewH+G7Gr14B@BO#YA^4AU?RaTKsnm9v&N>8As6
zgRUfoJbZYmRZb1MZD|hpvvNj4U*Do%RQuvIi$V=vwxzuZOQ;|nP2yMsPs<zG(UsNN
z!=TYMB16d;58GP-Ya_#nsa)$KQ@Tu&=)N?-7?i;uNgD8k>!?fZ&4CANr%89@^fy^N
zhj;+UoY^i^YNp+XcAh+@*{15m)w$J!b5_U9lzY^sq&I7sW7qyPYj%=KMO{Yx-O`2D
zTk6Ye>|@q)S;iua6=`OdtLnqr*!3?LdCo7D0j#U1KIt&wt3mG5_b4)U$gJs`ZU5Zf
z@54z#U1bHpPbzwLN&~ep+~VtB3vMIO{Bs>oxW7D3*~Wr{gGJ+(2pV*Woz(*V8E${#
zKoYJS;r)^a%Bhev=xt3WK==XH{apo(JptCiu66cbhT;}GL`sVe36z1#D-)!3byY`7
zeqYWx6II{r*S!i|fY@Z@cR;0E-8KfPFmf@NubCOXNNp#4`$){$`k%O%mylNBB3*|+
z*wN4v&eSm~n>ENfk~~3|NJEuK7m7V`hmu%KbD3I6><cuwp7^Wi8xaB`M%HH6K0n9a
zTCabq1l(5PP9fGrQ8hL>&%{ZH>XqH>he}at7B8<ZLT+9x?7K#As*|ERom~%lJmH0~
zLjfw`#--0@6fcWS*kJi|<`$>eY-W-8CS1vm7`P^JZ_hHI;BjP<H0Cz?@Q;k1mEo7<
zdN!-ds&G=ejOQ}@#DQK$3&-oGf6&u1H%Hf=+W@a2msd5J;$~$IVr&q3p7QWiC_YlV
zGdAN!#nZzDz{s!^y~fB^tRyD4A3$r<B^wO}Li*;xw6QcFtM2feI0Vu#&VpZdGf<bC
z0*3}K<6W%?Xf$7i!TwR--7{0(((ZB!edBX(x8)6?GB7%v{PTc|LvPDb%9OEKi=)Gc
zJ3R4?ImezeK#AJE@u!U?c@_7*Oi1YhaKI}UTC`o0MM?iKYg#8LnWh^?tIx$%&wu_1
zn*d+0&DrWGd07mPo+FZ|W2?2z5?hp_*^eEIkqnR;jJi4^(hqNn#i1pd<@|(g5uDdp
z)hz{=^SWm+Z6@nA$AjqVU5wMe0%aFDIdMhkO)~dg!`<`{bS7&Hft|YBfhS*?Ha56M
zH_ha?6W1k5<)T5kE~=?)c!74S&s^TcbfM>>Ow-i5{*5~BXdC*PfoY!kH<qz+%lsD_
zJ);rIq5q74Yh1x0ApHoK+khI?ahc!}^*qGb{<-A&cdc5Vb^3USy(zUW`r|sreagi(
zCi^a&=|NWe-tgQJPhZNiDUZ)x01Xy)Ig6f(8~2$_QT!1*G#X{Vcyo(0>lA%r|Kc0z
zs2fW1ZM>U?b%@r{;pepmnz_jx$T>itWP^Q=JMK+z4WNFU5U8S~m1Ski7)gv5S1xkN
z)%#mQTM`1b^+yf2CAwzQ+xbOf+lE@qOnau)gO$)KnsW(t{&a4so%IMDTy}f89R9_*
zF&5j35?Iig%*b#v>w0lt99lJ1r1uuYsuG9SGyGXHC#@95F!te>|A7&;m?Ci)Z7B@s
zAA_U%6k1{QA$FBQ-~*jfvHJriD;w3F5oAlvGIZymKUCIJQB&}+V_d26Dkt$W0O+R&
zMJTyIjuT1(*Eo&Vvmek<ubj&NhML&^i(byb_CL!6T*euX!|S>hFi;Yl^V-17XO?o`
zG$G=I>;>lrgSf)GBAO)bBeWtGSF5P9=6ekI6_9(Q9Es+hm~!;sn>blUx!zl6kn88|
z=%^&YMHpo8pYq-A#5;p=qC)rRG;J*CIzI+~0ojV4louF6+7i|~=sVuQHg5rJT%)ic
z&Wne?X^j3hV?A^88%&}%x=0+FVs=0sivf4wy{IG*E+~Gxdz^pk+w(1VedJ7o5)woz
zYt|-4TTc1Dk<C6NiW^`P4prw=a7xf*C<~Q1P9C+}PLop+B=D6zd^sL;E5=nFCzMoQ
z7R1#Hj$4!>_(obKBtAi!T(~qkn_MPhbbM>=S=*$N;l+}TA8(Cr@zO;S)lW-<Wpk9w
zWJZie9Ei(#aX{A(;A*DS9OJ$^#z~y<9rfp9th0$*fK6kV6r3Z|QbmKMXK7?3&wn+I
z{N6AIwvOGz3cXP8NQ|-Hdizj<xuZ>{XNcF&n36jrSC2ePaZEt-Eg~xfKgDje%*>W!
zk2Y$w_rP0BFjgFz3PPlxZEJhmt~Qa!9oW_qmbSCMUT622qHJ=#I8y2=u(lqKw09tA
z$vz$}QQ8SBU4-{1GvB1&QoBvR;1rTlzR&fGHr;A%*3NUYXen@TvaRA26tX1U2H!Ze
zX|(VFVVs`|7TxD#wzuV>6?j;v?lj(4$Ipn7>QX~hpg5XC?rDoGks*-RV1XhdUw$IM
zO3`!P<BMMTt%WKxvz-jthCLY8*)#QgQ8TxbukYDSX1h<&Z|tPMTkb(nWEw|WDv<)a
zSXI~&7`x-p{Ge$|ZC!38rAPX!>?}FZ5P$Sfk%o8t?^EsKVAm(lmd9nzE{wppqn6uv
z^dR+*F}HZw@QUA3fU(JNVTkbK+}Au5!Lp5JeK3H&6$5`~!IyQ(EZxcbeC^z>2SP$c
zwIbQBPP9L|hB5M#y>FyL!?rZ-gAebXll;)M-Nu>hE6<-Gw>j#`P7A;y{>%F57saQ6
zr@e<yUE&|FLYq{7v~{xIu+J8r3#C^kgt>rWLX&Ct7-8M61{8Ph5`?+0C?7j3{E?{g
zu~qz!FiXN1`Xf~FCi<6Hf|W1VyY)gniB%+WqD{k<-aJ!~K0+A8bT?GS1RP2!o|+h>
zHmV^y=;lmOEc!tV<IC{<0zwY>SFCo)f=KDHTx9Ckgh6z$n{-^-J?~FY93K>B7wY#>
zwiu-NSF?q?A^es!PHD}W<7O8Y^{BbW7;P-m%18SI!MnViX#w2utPoS+nYgT;1|z|J
z0p%$mf0B`U1Bb7q<XpK%#BJhPr}(WFiWqvj)pu{<CL<f+CSjc*xfGVWmNJy;if(tU
z8M79CHXC=9SG!&i2zlLNrWr}ZT><Hyzp!#H-zR`=hk+!px%w4lhF|bf(tS%PL%L;V
zd7M+fo^d+K2kZ+cE%#f~pGy{0Ek3!*s4W8O`yl0sVY-I1Ev=-L8P}qk0O!sNFqnRj
z5B)MB^SEl6*1MwA?vL&!d8`Sb9o{tCTy0>WOqmCIvm?6`u?q@RCM&c@cZsnoyE1W1
zvBT=o^p^Av{(l?yg-_iQn<iKD06)4wn`I^WU>EXHcJH%E;lNN7-h?dU>J=#+9v|0)
zyM15#mj<ao0siJt9YXwjs!@}(A6O_?u+$^RlD=`;RQN6WDN_A0dWGmHzwn_GAbJ@9
z>X1l5(Kb(#1oG<$z|xQ$GU-?kV!HW=$ne;tu?Jp)Z$!$=X;gg3*zeRPmi6z*d&o3`
zCy4$d6+q8z$LE_5PZ>uUaDPpgA=joT4pav#4bd>E5j@F_lVC~`LDFA+5}Krg4T<J;
zKx4#_4b9nLYs9M$piD5sfiR8xiT~=y?;80JTcS&~$WZDk%Z?e3Q&eQiRq1}-)(PUS
zxdVOt>7~s~UXai{O-qj3gX~P_<S;Pra?OfsssO9Z0Kr&TgR&X+gok!DbrHQWqAeC0
zJPl3M)Z$03%R`Koq~5Qb6iArm;bQ6tC@qR1OXJmoYutlJvV4^hwHBfNH9}GvM?h^{
zJ-}Bu&}~w>#H61|Vb8Vz6a68kwOF~$)odcir5ThW(NG7kY_sIJBSXBj;SMfYhKdVp
zTFeQoi9y-%Z*#qM^8o(&6pd^oCEYnHW|0ba=>#^`OT9QXQP|qbrb#XGC?gE>Pzu|6
zJSC~VZc>Neb?LdZf8x{u!hXd?+cr(UXt1f&HLvR3s^yxCZMuT5M0c|SG^3h!6oRMa
z!*~x(z0UL|$IZ1Bb8TmxCN{Z30N%!G2VwXvM-&+KI)nY6ZivI&#2|?fMRhITY!R21
zU>jPn72n-kxPX|ooI|q%RUhNLBGyu3=A0a{s6Jn{VrcNAQ9P)fA6wTRRN5hVq;i3q
zvy1g=Lw~rubH9X<rqm>=!je7eCdi(qUkP}m5WPolL6v^3)PNr+Mrt6D&?f*QFPMg?
zd~s|wYD(&)*0`LS1Nd!Ki`v~jYgW+J08ovTAcLM~bj+d6LmQa)Mg)d)c6M!imEEA<
z7VRW8q_{J*Ts6(k#dbT)&&6iptpx60LjlQ?)DnpY!9E5j2SvZikW7OLMUD+u2t#Fd
zC?-A%b{aS`bDyn&XPE*U{qV4w{P6Ib{6KEB1R44p{fhVfsV*UDbl}m8IBJi+=Qma<
ze)8c63(T;Ze$LN+EI)2h#qwUqe+HoX-G&DK5}KRjMiY<VO+T+Yv7TPV^pLDW`VFA(
zCSB2@h9sUCHX>G{M&*Fk;5w5ljXn;0wSh>UgRR!h*z#YVYrMVMgsim0XTu(4ffj73
zb-XOLK4TatGgq12d)KiE^XRMR*x6*^%-ZQGVxxs~{{7`ZmGh)<9&)!BBH~H@bg^Xy
zxRE@%9!eZ(B77P8Xx4#3Ik6;mgCqqcOhDyQDO0PQb}U~bWCF@$1$ytF>Xc@0kwO}{
zqqb6v|BchF2xek%j@4#p5z0PV5?njpg!;N9(K&>6FjmqDetCPgN+&-D-V9bbws_VG
zBk(N!4wzA1_0nTlLf(Af`fAr5WS4TsnJuXT973%}sf!NI>?S{3eCdA|EC+h18Y@ww
zd574Z^Nu^H{MNwfU|%<5<=i=>@X$YVj+CRXxJLO4cCUDYSK3DD(${i#$LHg>&^7$t
zI+bA?yf~G?v7vS1GZ@BgQ?<t^D$(FfmO+H5wg@UM37Is6uz8@;5a?9pG*@s;xi_qH
z^}%r9HM6E@Wk1hoGf3$t5q(?~%WL{nw~gF#>nzD;@b7s}n^pgK6AnrEHELzFJXzny
zG)<-3;V3wX);!+OFxjG|bHK>eDUL-2gA$LRk1}6ItEaLMAW16l)F3jPIgGQARx2mv
z<Qi>K9-bKG4t~b!Wy6U*UIESM7(J~)z&35_2W~g(4*%X%+^uWc5LG&V-j4ch^1|z2
z7n7P?t_?slDvRw9n07T|n^{`?Gfl-S&DKE+`ut5*398ZJDN3rq;I(Xbdj+)+LWCza
z<~YvlE*bM6;?<N+Mx?NKv{qQ61Nr;|(pDA8^50<Uzqrl+Z;9uB5BFGEx&MdV9XgYC
zcwNZbm$V;{VQIIJ|Ea`H3cRxM!2E@b)zH+A;AD^S&(=9k(>QjMYdLN*1Tj)%lss(A
zj~H#!;iGr-@O^W>xNXlk!XPYG7NjrWtA)mwaMzbKceag^g0fS7yx4YZ|Eg?-m`C}h
zNgW#q2<RjA=?YI)LiVEaC?wb~B2XBhv0!Ezu{e8u$p19~Ujg69f|z*s3D>ueC=Ukj
z_sEe-IfV*tJZ#qZ=J@^^{P1+e4ft|cQ1Z+%j~vVPfBP)!+mTKg=r$y*4$!g~fRNzM
zHwzSmeB6}Gw_sUaS;25G#ej@<>8}C@Ugw)>p@jJC+6sPcgZ4Y8OBG1Uy>v$`qZ3M7
zWoVKsXN)vx>o6tANC@X(oTGVg0C4q*>Fd*kKY)8_$5%A~3*uVa`g>-Js&)LTeXN}D
zz#CP6i>NWRRl5=C!g?kN4F#~x(D?^Ztke+h20Kc$m}XOW8U^RMmP{Lcw}W4G1i1FA
zWm95<>w;BwtqJ7}b8UN~m?-IB7RhcCO|(ztk`AN2sE3`mBoL7tk-jo?i)EAZhN%lG
zoK0vMnK49Ji)G7de8XzQAAzj(p1efBG`opRGR@M=Hm@`)<oTa_EBsTZr%-W3PKqZa
zi;iSqxy40zIv=uf1}%nkv&UbiR`xf#EnG~WD|iJ3Ql-wfwC9!!0AG4!D=VsIwp;4j
z%atBKBNZc*7rmE*_e!C^byX-npQbCV>-)8$Vtky_Qu%|tsqov$*{~DDv%6R^^ku_{
z5Ls;HW1S5~F>vVTGz2VNUWki>Tg?fr&uFV~B_s*d>Fvq=Nu@C|JiRJ+U1g4P5@tC{
z44WM~wMy-}!*Kin#q)I1cM{s{Vc(IXTx|~LOJzp8JKY1nHrR0QG6w8-N+Gblj~>7U
z;~c+q&41zGCF41*S*+KrknE36y++AIC+k9PFXDjTNVXta;*rAh{d0dk+wl%tlcvZ<
zrzvnn(|9M@o%tjyLQ5d$Tc)2?#?Y70Z?~v!FvY6DO58p`9-5=5E6@_)`QX!4qBOtq
z8*b5rp5=$cni-b~;p+KHR>15qjE~E)umt-MKrx#(Sk874plA(X&=O>g9@($|8Tz`~
zV!tv&I;M_8Z$BM^d!=FDRbw9Ahp3X7y_yZd8>79N#LAlk(_<MF_okD1jfn1FjAzFB
zRGBKF(;%f8DQNTqo5G)Oe(N|*Mo|3y`<s&Mm}`-_POb+K%Q@N&2<q76n)Uee<u!%(
z2-22VZ$irDwke@`(`^Ikm|CKF(i5TY?p)F&Y+>tJOZT1dgm&8ynt)r#X#>w5zE|f8
z(N8(_LU-PCt%Z}In3<4}{Wrr1JBJUJjpN00(`U8uDe$T~xznoXkw&{W!dEBF{jk+8
z_Ao+{;b9_=r0jS|aS`cM3M6ZTGvL^FBSy?VWd`|bp*oFekERrVkw?-FE>?i~X~$7v
zlxs@B;g4s30cfk<XP2XzNy!|tRqc}FreKFHW+y#|bt~!Aa@GO!CRBFw<!40c^<!{j
zNmAqC4h__;#*OVpSo=_i)TZiWD)#EZ=C=NA_!!RHMN&2asEWdN8YlYd0x;;nCHi=+
z14hIu#hkgfFn0H0{fj1i(fR)Sq73>9#T|Q>&$rv_D2V69mvR#Nky3J@&Uf3tc#-7N
zSAqVv2APs<m0Q2he>#x=`(H$I!+!tAf57&|iDvj%=j$8~PIiN?@9>tY8HRyDh78CA
zBOlcUj<}&Mux9_!$=nrbLTKlvrtef8sB+$F8UmMG4sNbNzCSY}T&|&I!Kd$0;-C_s
zw6e4|*2jX5m*qd`fs^Wgg7_n^XSCG!yNhvhJDg(n*B+tIC(yf}Xf_a(DpOQ+N~RnV
zHQA*!9(BNiyY{<?P7#KH7i`d?5?IAv9fyt%T}wT<3hEpQ%#zry^Jg4TX^VQI;k8sd
z*kL!60B(s$h5GnQpP>Jck%7^i_lw?mHj8%i3?n5h6@HL$san*UA)ZbGVwZ~jh6A4f
zmisJCqm`5id+Txr^noK`Rs;wvA;yTY-bF>t^N`U!6*=D7eW#BZ()1C?4eCI$Gz0Ij
zueQioLi;2*NVlaFbt_NeX)3ArZm`Puwt406XKkYACe)Sw%8eEu{+WH;QoeO3-TDo}
zxXMj>9N9I-H8#dfM?J|~8`xAxP91oJo-Hft5e*~7OW47i-g5YQ6&TXERNF&cnreF8
z@|gw@HCnDrQ10p|<$EW!623g(YNOkP*YT*i6AMvS`dfpsaKXNMSL!_Xq1#aGFxB>s
zCfATWC_9bKiyE$F(Lwd>3*j<|N`FQZPBAn}2x_VW5rJh>gDwG%5It5fNlA`*MDWS?
zP3p<LUV=^dyU|{VZ$tR24QL7?f&%66lE-Y8ZaW1WW)1FRs22wV<oU15MakCBDLYDy
z^1cOy>hCQ{=ip?}Z~sdij~vODZCH`Jjnk5vma90cV@kniyxK`7e9l%A37A+kMJjVC
z!5rb{3m9-A?yH0&n46|klYXm~I0pG?D+;K^)UB<N^s`V7{_Gx`RUjuM*NUrovl<A%
zK>lCTGmlNh$s13lyG6fha{v$Vc+Zwl-UXQ{VrOZeJ4W&m74E_3K_c_CKn`>5%GO&f
zU4v&S@vaBH`6$jzT()M7QSSXkD<j(&=GvJ&?ASApIxGIg^t(u)shgHR{PNbB1$kT!
zaUn+P(P`A1_aR-|BK~YZ`3IR6eZdTuD8f-PTPl}3X`fq^ZplMa8^f#>b>$GjNS$4i
z91k(Rn{=keNKK@->}A`lOXnXFbAOTNS#z*&M-$(y-+L`gJono7Ef-ZU=R&?}knC#t
zwM#RARx#e33z!Z15Tn+{8L6cTkHIF@>pr!27)%-ip0W2Sv_~X?3@4gaf)&~gmz2M%
zoG3Z_*+{cJU`&P&X4_nAnJ4h>Nu9aSHUjT!oU#~hW)l^uc6xn1idGqc3^!ccvOMi9
zqpQ%4+bBMzKJfruU*$_%dNZ4xU@p*4@iOyTHyPUzZDdXI9^lzqYASr%z?uO^Ye!LR
zJMj4J5w*&iPRiMJv`K9+&+Jjo_Gv)5&9s3F0$tMV!HmIenE_Dxk0#-T&YW;yXGhQP
zq^Ck(G-#s7ydT}(E#LHt)?^y3){A#R$|`XLdcyk+y(A#*lGlJ9ZVNS_l^i|8N6HCO
zwwYL1Xl;4PUu>fP%!qRIQ{X^xu=ZP4znuZEJ%0r8T~nu@GjEAP)%;4+VP~-&(z>i>
z%VRD<I%)9NCc^#5F9TfM-YKMG(84>NZNjVSN2bWWG&~5k+hh1cu<OhxwCxCkL5}W(
zr+cQO;=LB_Rcc2)7ipks)T?&53zxN_FBhXYV}GNp5g&Gh5%9z%zf1kNd94y7q=u2|
zD@|<DtTK$#hW(CTwKd6P-(}%*@U^coiXiY3r?+0TwZsm}O(VbJ@6qooZ?I%|Zfxw+
zP+|v)j@4J1_V#YF<yMb2T(+7%lJ6(EAOur!9nxXY3~`gG4y|})@UWxR6wQWQ(YVTi
zHaU_J_|j+qx8I*pfVgl5v1}`mcC8zO*U;<dNvwWfXJDGCgx@sjIphQn?-?n@6(d!Z
zVjaG;GyFAs*`jvH@a+8rPLZ>wa0_`RoALJ!lJddx_n(&}ztGcoau+xa(&NG+)}3>H
z2WPXqWG{g1{<LfSkn~q8D7UJ;AT4wNQ@>>iM!;&=B^Ynok$!TE3DU}1#)tlj!t%c~
z@V1UKPTgN!G{w0Lt=(r+DtF?O8!hH&T{1FMxMjva)<g$l`w~&X=&-4Sa(a_`$Cf3m
zE_o17lM(h~owYU-6j5E}G<lmN=tEidEl-@<jSrLFOr3q0DjR&gF(#L7k%&(>@_yB*
zJLJO3C6^sZ)c@kHsW-&|jnGkEqLZWSHfqSlgx(O(zhYOqpToARY`!Z*<gKX~=e&W_
z$=CMby?dSW8eXWdp{ul{Oe-4=tqpmX+m#-pp@Nq$Bo3=eu26yMTM{niadR;|wJ>!K
zj@tqpm=`bMx!h=I3@--?L2w31dCw1wVOYx$3}!7fK4+?HaDVHSG;&6=?b1Vg4LP{O
zgUPmH%DUOo^QJM)+aRW#oA)@JtGt8jY@FNLTl!;D49!hGh~t*lyhol>*ri{&7b7OM
z!4Fa`)(zU-Ke=&<Nl$nfjZAKekMtVE@n$w7A<CiIr#3qf3MloZG*el}koHnq?-SS-
zwx_jJV^W(>M;>PVs>y61nbbD-I!Cn^7HGwk+|8~alFmBvTb6r<Ju1O96f+&L{M-7u
zZD%iolPa;s)C(3R+>vKqV+DfGc{1tbSL#Fy$*Hu4rOv1CaHqYN4&f8;Z<1(I*nYZ8
z@K6v>8|Dv3!~PWX>ilp1xESlz<WEVT9n}}-t)jMcKn@HcjPJf0w1y?fig7k3yzBhO
z1-)u(r^lq1Ipu`2=xQJsgbXBWKws4()AT3I*PEW!v9T-5joc?jbahi(V!PuOYjh<N
zjoSRf@7v|o{~?=)<iAlej{n<9%*w{W^*`FQ+s2!WBkj1}`KKHh*6bkW&qr)7>JznT
zP0^1TF^|b1@Cv&*kKKkRNkuBBZZ6vbII@7Z+#1Mad?Jd)td1UB6DQ*lXY&aBxNy6`
zkaJ+X-k(gB?nDL+$btV4PR1DjA5Lcezd6}1a_{!yKsgjWTDOwqTX~siJq!`V2NwMG
z3_zw7PO2t3wIe#@6qRoXQ6383&pCYp^dt~Oe%P$}A0>O2_Vjv93isnVEAO5u63f>0
z__!<+6qGs&L>oWNd$;!ccLR5qLsQ6@Jxn`{d42pxm@8C9h5AXzJEg~Mee?zd0yCPM
znkb!6X+QmSyS)skN0cejNkpjUJ<h>$x=1@2!=Gewerl%1fvz0zVqXq%R;Z#0gEWFB
z4n#_@C{9FJuyl>5QSxhij2|kcyh4VZ@Xk&o>&sisndv=*MZeWG%gkaX{+%so5ul2^
zKlp|G=gnvpl@Z!-(SifM2yPULD3l#N*q<$xlHbMvE7BthRIg}~&1}W3bPDRCW~L+$
z3OPmu1XNZ(2t~U%sO^lbhPaXouNL_6cR$%5bc<WH5_NEhLegb!(2_r%BCpA-Hv{)y
z4{RQG7LUEd5xk}fTK$r`<~D`6gq2aTmXZrijYg(na6N@0t)UZpW?7rH$W;1Oj8cWI
zJXsI64(7RdO0(EGcuIJ+GJ|rw8tBpc`Cdt<?a77?u`ufEwC*R>9&6i7FVSVzCs3Zs
zGx(}&qw?=(P8D4jZaz~dm`cUI<4DwrR6Uf2RrtSZmv-It#2m>;mr!nQI;)aa>GX6?
zpwYs-`>LH$*c9_WXj7m^Wg35@laOQrE~sgB7vBmQ%j}3NqX!VBNrnvS<tsZiD8f-_
zlPZ5BuNxe>{`w`BVofm4Nvt=VY69whg=3coU1{UZi0tpV+miQ304DTkE>t98^-p}Y
zv$QCYQD%a`9uFU>soSwCBMj)UfYFq7{(NCqP@Ajc#mn|S%ujFCjumj?0pF}_)s9zU
zf2KmsZr8<ge!S~2fF>i7P~+PdJyR>bCLcLyP7R8tNQxWi%QetUkfDU$g%etLr!Nfz
zLx$HuEgwV%JTDRFEsgyCZs=d|*ZbmW-oP%S!B1+I2pTP$W>~vVRs4+EE%x;Ttbq3%
z8EmO&B$zO}g(a9QQa-S&O~CAA@INyNG$!^_8l3!~1S~_ZJ41u>7`!DAL?QRD7<ie(
zufagJDgm*A3VBjNQ@S-yK@tCpRKmw_K^2v==hx%?$%i}Sj*D)D4(Swh$A>%n=qe=a
zN+4Xy5iO7mB4w!+R1IaI<!+$kCk`SDzo|}pS2W-$)wpekW(ihdtnGi11tEb#Py9oS
zcSeQiMBCBOY594mG$GZ6FdaJFTmJsvVPCx>b%g8g!_JBuLT|Te!Jy&^=o;&|9qk<k
z!zW9bQy~Ao(;(~+`mV(_&+Np3#DWYqGO4YilsAWVVf`F9^*4rNHn&)$Y??Dshb^(Z
z^M-6=^`XgU?KUCsj!f~xYxvqeDlul>O^gxYD=+t%Uq$M}f!a93NvW}z4S!J5=5%;t
z(${=As;O@bEsx+_YZzI-OyOXNw4)z>cW|HL8Yr;;>URT`j3mj#)cV*Gk%oYMyK*q&
zhcNLl;?On<QRhgq+G0tFP$E`mIl($>FP_?2R4GZCv0=fKNidRx!AT=&u~97&qp+}c
zm#3{)%Tn<XhrGLHGA>|#&S$(tpBa8)3TtMt23?5l%(-s^I#tyGIiE%Iuvri9MQOVd
zHMm?B_-nF1edIK^1HJapbJ4Hn#<HjLaDvB2+CJi2i`y;m*1wl^yRxTK<XF)q4UfRp
zAJ^=Ci-5i54vF}k^=!}39zSTrmij}CS*uNz0uVUT@ofv+&X6Sz8z;37{>k6b2@Yo+
z*d6xWgqqjqXV~8Ftzge+K1_Cu-)2Hrk!JX6U(IQvwxQRMxc2i$7+kubaIeilmZ`|O
z*$NtGs>El;d2lsv_GKo@irR(2U+~M3G>D(e=_SxvO*Zp!qC6SW4d(d#H+$UN^H_&?
zrRQ#y!QVKU@nLB_$UTKeTmBou>FJ!83{hg<571!hKW~%Na%2_gJ%4j1H<*Jsd`}dp
zZ|<aXIL_3~t%GBzO$-;PP?&*;kE4qo4-42&B0s(P{9lZnQ<r98vaPeyMx~8P+y2sK
zrES}`ZQHhO+qP}{)IPV}7yTF38YAM3m@((`-JP|y-zG{5^3kAMwZHpIoC}j%WvSGQ
zoW1`!QZF<C2G^v5(%8Ij1X8+6y|sTxU$6n%YVvuzC0F(>%qPht!yz1d!~U;Y2{#Op
zU7^ECDO~0;j=VItVo*#5QHyR#kw(r}PjxFGw1V)lu$D&E{2b@5@_bW(IELB6`JUw&
z;mb-b9=*S_LNYDLPzgFZT1qUz6QR}FWIxt#orKW9JdK?qD4Au<sM8&Qsv>nA<iK<l
z)U?})`g+2QqYpo#W5b*3f|$d%$wriYk5p9`a&2dB(LB9t&-tt>+bgmBcyvW(I}b@8
zu%Md)ziWtyr7q31Kz`CIaHGHyi&EHPQ2oVGN*dvLKQupEOh&blRy2evr+fmItq!TF
zvRa*E-Ax7}Y>73}&H@6)g<$psZ82?;VP(gkc$TR?Cv|>-oF!`!QokFkB2(Ek)|+J7
zyx^IFmeDLOWTm{|(No?#EHsEC#Go`+iiJ<xK5gX`M&##}Mt{u(OIQ9T^*2iG+2b%Q
zbox|C9w2w?o^PPi50o+zxyBH3e~o<O-Z<TLLAk?9wh^n88#IbBUIrUWI0qKpcQFJ?
zCC`fymo1bFg8idv@M{t|??pr|u7pbof#j;$f7W$7G;tG9|GIQ83uz;^&xV^#?_EO7
z6UZ#G=GOQ^dUuE6-0xfg^;DZ)rLO0MKK|RcVAj0Y6#7gIi6A@$5ZOZ#cGEe#K7&C=
zawlgsJlrD-zjK<2v&LcfNT_h|QFxnUs`c|C|7w_Cbu1q0QFsg0ASY0V@{Jqy9rzZi
z`bZPp6RO)(wQlihg9utB3Yn7VBb$dq(}UCp5B`Y7sp*->bF%$|J7JbON@L_?m=6(%
zUWvbknl?Ut4*yL9s<!}rKm$+f?i+|Yd~>*KPm?nDV0z<Zd#EX5*}~asD&$Cor(<u+
z(BIBR6;2AFf=Y<qYgTcyx8wgy%6QmytS4_>s>%l+B7`g1i>U;|*xaj!sL6hKWAf7F
z#nx2SDR9e*d8cpSY&U4`8T#15gpFgsxU<7&N<q(QXJIQ}w3Q*J*qz~Txwo#lS57_I
zy`gL&!-kRUOBmF)>i46Bc8Jm6=6g;WVFyAO?0BGH?-b%n1YIKPQqp5xkk;y4k2l59
zyF3%9esGNY8aU2E9gmY_0Ldrryz%W%f;(78hE#PVu0rL;EU)oC3)%9L*L}5mK%cG0
zL-pc&wVMiSP}S2~n3n-&QU}FNxj!;bTb7$jQv5AIyg)<WTq@&~xugrRUcy6<s2EEz
zV`B*BN<vzGsK26QJ{I>0<q5GkyunESNwI-tdI5g+4+#`Dtt@R_)7&^Wm3w{kj2Tv1
z$%88Gik!64`gj46qYd*SN1AGpN|wjqBH<hbgOj_p{HnzG(FF2VS2HZOI|n*N^61O>
zBlW#0o?7)%UAfh7ld}dQ&hD~H+H*D4qMu&dNDZ+h2{$#%lS~b0O}o@6&6~nx4m<d2
z#F%%7Y|IaMlj104{hqTli+uVBSu&2q-^}lZC(cFg#c|{3rMr3aBZqYq+M1ulI~-9p
z+o}S#MSmmL(j$WOujNDSbM0>gc?#r`Gwg2{)dL{5TI<RpQM*%SA3su$&JPG1Xf$3&
z^Y=A7^+v#S3{%TGocR?gvS51BPf~Bs4V7KR@8eQTlAWhA>xR^)`cbK+Y%;cU$gO9p
z9GU(G*JMX?RUI>8JOHcB8LMsQJaOvjoiS0V$)(&aPg;4*MDWf4;ufPvRlU3BmyTK3
zi5vP-vhl09r^o465?7H^UA1mz33jyI)hzcmIbLv9(9Hq@ecf6+6oqUD^Zn$T>&gk?
zDzGxvuan~A*7MVhhDcW#&hu!N(wAN0mY`n>88RfxqOSl-bcJQ?UKK*d=42|&&gQJs
zf9ffmh7QgU3fcre4|%ldaV}PiVI(U7Y$j2XAJCGs1<DDYhD(FxQMzX-d6FB(3P}2a
zTPe8k=Kt`=$Xk_n&FC^O@^l*#Yny*MuAB~Ir#F`mguslb;!sAbz&3DiG;6cgOpPX7
zpPNOUvD~a5*rSy+Q;xYdav~hvntT8GzHQ{L-<g+=CzBft484E7MWi+l;&Wf_hpeX&
zsTPl{m#kr+iD8>pd`1c6uDdM!<w*m#pqVIJtJVNHqTdIS`;Oix&Ac(hRW@=qY5zm;
zhU?Gr^aRkRwakmRv8pRYI0+ukg#TKnZiOj*XmGUP*Y5cFJJKePqSDWsmQ^jIls;v(
z{9Vs~3f`aW?@C6%Y!*#a`@pfW0@;D1Tf(yXjUO5{0r9IVHJwgS>O!wHpK9In(js-K
zX6iy_sETE-hN5(Dzn^kA@}~T`z-z>{${@g<ONeP8JC`nB{LO$@50x)*Y=nFG?NkyF
zyXXq%dI85Z-^dU1ulPE3ZPq4Z+;361cUCdYcup~}1M)KRt<k)GAmFX0ME=`k$M&DV
zJ0|-7(IdKqJrRxF_Af(ajyIQe2SJxWC>{WA|Ipt;5EAtNBUG&aFGA&?)+SabN;5O5
z&*Up|ey(4qTFKTibZ><;F%t4{65=AePVEh+?B3-0-7eNjxZ8sQ<mh}SUoA6n^bm7o
z6C*$lk;CK}iKsp>GH?U5^Xxjy2oxO1bq|W~#_js&SH524KiYv~A`vNHWk95uAgHF!
zE|xbrQR!XAfMGhl^kEok!sz};@kN27<BeeYyX#+m&Lm@~QBC&muLm7`&~aVO+6Utx
zwR3;CDHFX;uy#_fje>>IHapiE`(y%eW5$1%D%`|`7!kFOHv^$&Bw=(1VD04(Z_tFa
zv}P!tIp?^s@f;6x;D?r|cXHkv3yzlRAOcKTgUGY-HxZS~r}3@xD<ga711BLn)q$p#
zss-T_DfkFy!E^RY{4^4$N@um7-!VrQ10`|P)}g9kkrY3bs@^}%u(3LS?FxW~P^Z;4
zD5_C%r;P2Cf$g_MAG$|rC-&u)OASk)tb#u)q@`kYQkyY;8SB*;i!*{|)nXM{J#62p
zZ3DnKjy58e#MtB&OY*nY@OXBuH_>LL&q`a@5pQCrNq30Dvh|Y%<^ej)7?-)X*1r=T
z=^nQETO+eH`W!{aB_^-XFPGgo^e?D*fNZy8FTA%CA1mTjDw6HTGvw`AXtnQ(|0;Mc
zWPW~Scftbgs(0^q2l%xgp54?!pZIkyMpP=#@q4f3VSiOpR&nH~)L9*0eq`4}WZc-=
zQ>rY^Td%Lt2Ant%y=i-N`}mlRytf@LBz<oQD`DUIoQIo-OK_Ua7(2@EqLNCI6#&b~
z!2y+7B86}lV2lqhZBzu6W5GIe{5o1PUs~!P1bb_6H~V4q!41R)P?kqoKjrEl%1kq<
zd=V+s?$UcU`qWauHc-N%s6EYHItcL75?W^n&Q?6y6}oQV#7h<e>Exax)o9o1$kgz6
zv~1G-Ss@?5hIq6~4y8!Jc@{n@w9?U~v$n|1d?CN98V<CRLPIi1NHBP|tGD#R)3<{(
zvVV=*(7<fat*p&J^j|jBIOIv0JIwPTwNEgAPHZ7^3H<#h5XmFvFIG$`tezyauoy8U
z8dmdAKB!61%_|_kw=qzz1Dc?vYdlF8lSL~zJ;jG#?yXoQ$VTc80Is<Z?%Oxe(yH4h
z1)D?Rk(>irL|b{gmq@DDs1KpqaDAA>+>=a~s)(*#okUP1sqsOv5B<u`FsXEQME2ZR
zcn&fYby#E07h_dtIq4x9ma}uw;}h>-w25_~J<V{20k*Q#SF}v~bc2U5h~B@{5S+b1
z>%sIl7lw_Yl4x<UFLvQEU~LQDK7$rlFdoV~X}q&rk`C98sEB$2SwmD;t)Zg6Q=AIo
zO+kU#fwYwQsXNzO@Qft(_xyFIK0DiVwM45*CEq_+l#Bno2%u8tb{?nXiN?7z@2+eK
zo8!LC$kVhO<k?}l#QHRoJqGRdAF!a(b$Rve_N`pDkdRfZOg`td1q_jw784ut6yx%K
zhr~}0k)Q(yxham3l!3SN<E5p>=k1<KDB@_tXCAVS&(6pQb|S|9Oc&mWPo+D>;&y$$
z0+OEv`f3uGCE`{*#Z(i_Qw=W?cww29{`3Vjssaj}GK-b{rP)(J=p9a;LkyHEvG`-!
zZ%{FVu%^^LuKM>s`HNPMJ4zzuY@h}TRpiUT?X{iTvwS`h>)Ss4S2zn2>0<Z1AfhpF
zRW4*f6asP7sqjDe<~W+VAGAHdgqfPA5>HcB!amkfi<*Dc9}2WF6xi?7BhUs?SJeUu
z${4yL3bE9W!7M>W0>uK!>f`GOBKiZofT#XO;DjJaJIZOIj~oJP7rAP)zEN-8Le#OJ
zRDl<_$;;9z-qNwkNy9#q@nLchJioWUB&bex+1)hat5Jhbz8a1D)#_IiPZgBNN2|ZM
zN}nZ;z${jvcGbT%qANEa^ylcu?HMhf5-|cg`4`D%<|`%qnD({QmTPz`YR=O#Qa|Sd
z<k)xIF-yw)OAYQ*Z+>f)D$-5r@{)-bJuCXtoke5&W0@DH^&CT)Bqv2(F)AoGpcscp
z!IiR`R~W>Ilnh71-s-+TA)@)4x=pG@;7U$>yq~_lzIeEGfpf`U!s&=DLhxX`jt5|h
zd>xyR1g@cJ)?-V(SG0hPJa=gizY8^z`W}(<mU{g-<TyLEgsyf@{o)R+XsXlPRBZeF
z$r&ayjSZ4fvt2e1GvknB)*e1!89aCPzT&z(6j4|5##8X6Q2kH1c(aQV)(Ho!h4Wig
zN&<%V=by>3J*Zdfv)i6dtk9FE$b%6Y7ysCdN^@kGuEowsYce``Go@^9TG);KsI#Z5
z!p9<Ey?9Q8mW9@nKKW)PZZD<w&^grG;#$%ta;QN-HOJ97`IqW7L4abZ{?a;E%+oqw
z1Yfx<k?H08mx)O3;=evjk@MG|Jg`6#+f4_$2v8`1<Xk%uaXkC8fHJNs7W7q7_;;6x
z&?k&bfxI6tnqgf0bvrRknxkQ+xs~hRy+VDWc)vAd50lY+IZn3x_#jRtg`eBV0c<hM
z(?dVK-dG<lIwWAcmY*bXHoP_1IMF50Yk|Z;v9=^O$dPtv=}$-&Y^Y#*h(BL{tbZT3
zwuS_vIak>6i`S`m&gY|~_hq4Ywt-TaX6R0Hyj$<!R>|OLAXN65%HM=wG_jG{?x)((
zC@tf?d8_C?z);6lq!y*V+v;}KA?mb~#ZK&WT?ZY)^wcD4K*JXFG=#Y&P9u<xIjV~W
zVy5Etp(d|UQKxni^OoOj86G5l{5K#OqNe^;6=Igr53_|Ad0M~q=~l%Y=wex1X~wX4
z{G**ic<kYmtT@}BrTy;g&(Rk1iXf_;P)jM~#I+RH&DjNJ{@6piJLh3tIVH1=M1li-
zI+8${4>{G+H%|>VjNyLeE&Ngp_8*<|O02E0hI38$e>M~>>}o+sESrjW`j?UOK@+m%
znnIq6;a}0GjgVBqToyRxl8mGkPIGEW&Em2#H#wDmqoOAob^SEE|E`PuEM6xDw527Q
zI!U3O7#l@uTsLB;Y)uwruPpOESG(O`?l{g6`Z0F{ocW#1;YiX>&YLgR;ImN;X=3#R
zXYQt}zVwx_5XinRmn2N)*P-o@f0RIG6u2<;^3DP?Ppzns7(~JAf0zFO3v%-csbN^J
zH8>|tlQ_Xdh%|6nmF_$IEpnd64ak639%HN}9ad-8E0<}mTrkRsqEIJudy~`NA0`JW
z3Pm^|->)PgXawgZJZP7z8ccptvQm0dw6*Xi4^W1Aw#8o&!R+nEa<q;)Ooqb_(jF+Z
zep6&LIWV*@!fKKK0}fNzT9nK$J=I`rPE`;R-@&4r?S)fQs&qVPYJYChdTw7`e$RNF
zFLuBs+&R5ZRFbZ43oEL5lA~6Ssd8X&^q5-G%FR@8o|Hb1u%HQrRY)ip`<*Qzd<|!=
zfpdQ_u`aWy%3=q>v2~@JS6`Dps?bPl`h+S-5OS?XGi2}4cBY!M%6}bJVz$JQW8Op_
z<-Yk?J^!{e@;peLiw+@FwSGy^p`81i$um`si$M5RZgFca3#WbM%yG$yTx`WY5G1~2
zpP30gD)GaJuf#!U<GUW?8c;;&6!HzEk+ig){c5=gbQ55l6?643y&qiw#!%|EO*kT2
zc4yw89+qpdx7EYT9H0bZv7$TKsAYLz4E4(ex2AqVX=SuIaZovG?-O<1^)?HciVn|%
zE*izz)$VG)b;X)ng5P|=K_lx~S^`gVh1iPE-Gl0qKN(}SW&Vbg>mFZvy`%G&27jKD
zzEM_%(RmUpMi=<=eZK<)DdQy#>7lA=bKvtjl{FXjkuo7^4Z;&1n8H_@N}Js%-$?jF
zN)+n{tB*f=)F*bQ>EUE~66OvQqaB?54`B_acpf3Y*TqWYu$uOF`W)X~lr|G|nZ|Rv
z4dxT5&K*fX_HC3)r28S8_fToh<uJAgh{;R0E;CEw9`@y}4^d0A*?*k^{!`y!WaMD@
zpY)xpf7t8I@qbf5AAYK{%@Pa&EKw8-xP3#OMyw#g;B+7Kz4pH`K+#F1M$nvm&3lv(
zoE$!4tI0;IQL~%I{>{<VmD95oqr(-6&#jBwOADnt_Ekk3lvA}+z8|xRyNlYVEkT~E
z6YsWO9xEVaWarn7EhiSLZBRekIUb!8r+WpWzHQ80>$SZ+<8OjxY2|seuo)B2Vf=Oi
zX3rbFaLC_;I;wSn$#l6V3HC=@f8uy>o&L%oC-qJaE$HNYd&X9RP~;cnIA-MF<-qsg
z8b_^S1V+?y_u!lcu}GxmqNCncr|kT6djdSO>y*`hw;b{kr0n$2ARoeP5KA<MIl}AK
zhQ*tbcd(-e<*BTZJ{KhV_NyuS3$v)pho=WR;!%({kmn~sA|Mif?)_e*REn^Gn^a`M
zg48btpC#AOmFe&0A36>QeTU8v%IS~#Ju6#TvV|GZnHen;kKQLZOh$cH{wW?08%r0J
zyey%LZlWHZDJ=8u2BK^bBqBMZI@SdH254I^zr;#ZMf&RZCoH*yY9-Go4IxIZ;tCEX
z=hlRhU;rD1fA!0v5wxfQJEiJ7OnU=qXygs+`88iw%1H^TM-Pmlz_5p%zh_L2K~{<2
z<47K<s6cM&{1tQA8U2{uCZ<=91m6dFbP##x%uX<d<BkAUe)D<;aaqLg8{RXoJr8O4
zB>hZzYSgp>baq7koK`w06{N3M8$r5Sy;^E!F_>v7<arUvLDCDO(>nv|bI6jBnJFKw
zu=Z#MbBi_;e9V>6!U(R--bgA$lgTpW=;ksu`+R?Vsix`B&uui_&Z(b+*x%_MBaanx
z`R@S@qb4qlq-D0MEJ@I8`R{vvoMl_?ZRcUpQ9nx~KT9w9hCYX*hwm<;W$kFM8X{`9
zV;0C?7>Syh3oa`n)d!D7C1AVHf0l~<W0;hK=|I)!*?y`dEO-bjXq$pDb<xbV_RCh3
z36fODyS1_T(Qb}5lu}9mu3d=|&wJ@Tbh^l6hx{7VBm0Py`WyZ(B=P{JG{r0b1Q(R)
zq|)D0b*^DH2dYAeV=<0XOx%`qfh5`$>Y(sq+cXZ2R9JIu$SFv%d|QO!&vGPPJ%1)U
zzX<eJA0H_~dDwIsFI*i3<hwuLt%`<>a8t*IYtj3i?xBU{Tn*8E{S-9sd&Oy!)bjEP
zu2L2QMybbQSQ$-(<x-T9EUjYZ<TZ5V-nj;5>jfSlO0m)2{FjAZ`FjI*(zxr`8>A7L
za4<)wM&YCvOR{XgJE7nq`V#5f8ll>J!r{&jSZN6)p(ObD->MD$;&V_Nj-E-VW_z|`
z>pX)j@=!?$ZObk5<`F{cGq5Y`&Lbjbsb7QTp-suN5;+d~qSamYZjQ|p!9m-;@*y_V
zhrdD1v5~c^l?AKbdX%7eLOs>+om)hxE^>Y`xIS|!?+`HhfEAuSSZ*l?>SS~ek;QI=
z(6J9SG!$t&wBR0XB6TspH>YjNjVO~^j~Zd9h@?8qp8aC)Rb!KYIi7F~DS>AW4x1r}
zGTn6wiuS2ULrSC4MK*A8>_?J<vi)-YWir~iyy-CFbtLQChUMtvs5?34a9jPPwE&f>
zEL~;O70^Y8EbxqtCq)`98On`vN^|9KY4fIaw6*(Io0X{nKMs=T@=E9=Q5XgFkQ*eQ
z6%nl1oKp7}5R|Q{V$jeuk$%p_x{<YFuT28ivaV^{(V^V%de(gr>G^_PeHz$aaucK9
zv%TfyY#)`-O=o~pW1w}`Zb}X}FS{20g63ndnKnH@-%u&=Re}0NrXzph9R(0`bbh`l
zH&Xdb`qQvKqHX=eU3RAnkM!}wmrxP;tY`khii@_8Av#<8Hm<V}n!~7+KL6mjeb@Rk
z`6T&L;H90hU2-p=i?zB?&s=;`uM+%mG~t&WLxH{bH}Ki@vFTR3EM%)k{W6f~urufS
z`b!@~`A^k0)pPAfN`w~A2KPM~w*S_`&EVs~f>NL<KG-hO_1n?yhQGX*dc5gsN^VB~
zeNIqsD-?PMa|MgCp2VQ`_3fU{PWSiOn~lV7u+IW~n<_V31x9kpC#3I2KI?>wH@pB9
zf}R{vz8G@?ibO&F#W1w1HoUS2x2C{p?hX`I5L|Nbk(0wa{NdT68l+=xKZsuCtiX*2
z0<|W~Z}H-BZ~zk8wq#6e&t<-7`};{fip$r(GzcyDoC=<TJ|8n6+O2fXq}p-prHV%1
zTS#=T#_G~P?vq;Bv)Z%Ji>^W?;)Ksw{RdDX!#7e@&>93;y<mZ@(80izFPPnVzT(T3
zzpJ}|iLM>`x$GtWEmhO4dUAAJH`q2G%Q{S$MZ>i(&65lIe&>4G-#&UFWEuQfHFHzb
z>VNt+H0Ot)9jRIyBOz0uh02SmgQBRvbvsj?%*U&!%yFp@rT7KJ^~i)?Ta1ot*9N0R
zw;@g+Mi3R-epbXB%7|`iwQp<?H$h!Po+C32R=;ii&*`$zb+bM=K{=Jw{0GjsR4d%i
z;b@f<Zu1E)_^A1Alg+!4(N=K!b7XnDx{0>49KSlq+VI!MVXxAegBL)1(F;nl=AwKp
z>*~q!Y#)6~EFl_YlQgUCZPKdCCIoi^E&t$m%S*8t$~fYKW_r!CtCD4wEJ=q>do!zI
zWy{87H+PMY!O5$|+^PUo$N_JPcbaB&G|yt$3Jj?dX)UvqvxJ^)iepcM<-UWjuEm;T
zzusC@Q=@Jg$YqJ;mg+CTwaVE%%jT&n5j0@j(%V~JMb;vq5&#Vp<Dcl&#LpggZ53I>
z@sAvNp0gje*^*JOqlU`t?vTUArWQ?SVwDLa)a#h4lt(|j&uRCvxg5T;re9fOEXE@z
zumw-i<FP7TJV}+U?e7zwe0%1ok{pjGEk=rQIYbWDE5YYgGA5j5?^B{TktdrI46tbT
z9;S}t+)2vhYcYqfnC*8Nl$|NED_&kF(5{VTJxm9ufhtuiqR#<l>}Km2rBPk3gDMrZ
zS?#{cOUHV%kZ@j6o=fADu-ri_mgW6hQg&4tyPRDs>snVQsQ2erF)O!dkslVWUwi{E
zKftrlN}m7i0maVrAD#sUR;K?AWp<^;VME1+c>P87jF1|`grn_C&^zA%TNtrND~5Vl
z<Su;w69H>O>(W#o-6~PX={k95CsUpIoqqWEO#JY&cW|~L-T{?3JFL#+^_l8xLhR$;
z*pCUo3=<H~e&_gVPWT;_=(rnFxEPK<(Sz%A@zp>yLgj0+Tl@2MkRr7S=_uqJ)w--4
zTa?`@;%!JIQzG^GH`HQxqzJW8bb0+yEh&dj5`wUSJ)G3IvTwhfSst2CrBIHw!W}j-
zEo7;K8ofe{&d&pbrQ;RGG8Nz$udoec#Cyf48?v^oRD}(Q^@!~8Dt41l2j3b~Zo9Xf
zWwVW<atg^S`y?M*d{%}cFfe7x(gxuY>~l8Og)#cfCN>}`ly|yXQxJCm`Y^c)2tIwG
zyzccZM`Fkq(CKO1dQ4v?MFEy@Sj;zc5DI{cG?AHxsx8lSOZ*4JI}36nOMlwD54ZJ3
ztc><d_OF*2BSmtyYT{Pn+>#nxq2&+~FXFqxk&#E_l`&)a8nEub@$#uqC*)j)7Og@t
zipaaBE<jh>ep`%y&vI^d<!er`)R_Z;kuX}RTa}w92uUNt<tm%K)cJEPjJ1(<UAM-`
z2Scijn6C+^SlAWo8)ObGsOR$9<pcj*hp8VAmE}~AqD#xQW&#XHLDkzweF<kg2j7Q=
zE;eFkQw!uF7t{YQ*}9r#(uPdg0St;+B$l;8ERzFwDQJm2Dg$$kV-l6}*!K_FYqrxJ
zDI%nf%|OpeOi2&=oj*t%Q2O`l8_KY_-(SC0ba;$l-V%J-o@gP-dDQrpLIfIQ!-lw*
z%-AkNKp}(rj3oqyk~Fd(+Ns5OA&&hY#y4UlaKn8}YKnLik+V4cLA-_o(Q!-)X#3<C
z@^8yR7{T5W7(&=7iULhI7+rA;m;>{s*Z_!Ccw!5v8nxnySnYlVH<CmamEEyYHPV?A
zI6t{VPHeZH$hTh9CE*-R(P^j{81r8wUBP%{gntM{!tr>s2rXWd;^00J`D5>3KBFUU
zec)jfiT~~aM^xzZH9)yda@rl5;meh*iP@A4f#W~}YXJCM&Nwu(qBWtt?^!<kwNjuR
z#MyN=lwr{t(Fkab;DlKvG(JHXWjGWF){XR%>*w5I!RPxOu<%aL2b|#-B9j~rs^y7I
zD9SDM28&>yOE>%EG*x>B4+N~0j}*U3?!hUCG))it<4!P}ww{W-F8t;BQ@G*A5|!YB
z!s9D~T>By+7xRvq>y1sR1+sj#^wb`ho*M=7GT9+4RvUqj8i6pDHthwzuqdB2)pe;<
zd`EnHqBfZcs&xa4NNP1(Q=|Gx0uqmiQhM_%y%XWuv-g)DX|rmfhd-maC(vF{-?Rul
zis}$1Z$O41=NRxSCT!o;4%<1R?ZH#(S!ZA8;YnG})gX8X3OCvB0;*eXEOlW(HdnmQ
zV=rq~H6$%^5PmI5+`x=cw-L-M?FszMqE0m@FL<5kbX1%60mPa#94&C2%z-6**&3R4
zn}RU7CKm-&(uS=cAOGdPE*e0J8KJ1>@L1$53!*Zdo0<s$EO~S42R@y-_XCS{`I|aO
z))L2D^W4mZiyw;mgwvRFfizcFs^s8cRaPT%!Q(~OG<t~2?Yz;|c^S*2P#*e5Ab8EP
zPtFfF{t(#9FLaCW%^vBOA5ecdvcPE3a@o5e{np=e(yK}n2lQJ!?wa$tS9?d&KZg0g
zT8m_6N?A;>jh}Ipd|>k?YVKk_i6vwzWQw+2WSCk7>CJ+d*Lt^rg9oNKYtkUZI7X%V
z-&{8J7TCt)T@rp3IdxUJr?Le`4`^lCUhkV?(wJ0Bk-eZY_43rzYGKjQwuiu+KX{R3
zFt(t8baCi!07$VL0)m8cq{ei+Vr6KYXKptRXOf2i)mndV-(*nIQJc3b5)GW$IzV|n
zK+5?M=Yfl^uDJ{Wf-qz8yp>)0r`lqL9ID$CJKCVp01H}?noBj6LhZGs7T1XtCM9Pk
zDtfxWb`2QH>Jk)MYd=AQiIFGNClnwk`D4E^i`>>V66>LEE=G_N+WMw^t{uW<jaHsE
zTn)*@n=Ma|xKxqrb5S;0Zgr6Mp}lm<YwnR${{<RJ_wRu5(1xt#Z@*?1vQ{JY4DoB{
zg?&r&@0N;>OLLZVE%!I@;3@&>aR;sq3Dp#xneItd+rlsO(Vp+U|JtoM{<B+gF#d1n
zoJ&{}(HrcSJUV;$7*XVMbh?PTjMC7TRuOG9&yB0kiv!O;2$JH7mSfJDAsV?M#L-f6
z#6u<I9G=*Rw{K^!caQeV?Cg7R>@1$1xW3wA@8IGA%dqf#s&E4%FMBmA6iG!haXjQj
z%a3-CmuD+D*(|R9K<*%*zieHdA-P4V-bcAbv)rcZCg=F@a|jd97|bVZSbP-SFNmzK
zh!UY`t1l<cLwb0>UE?PWXX2d5O||zz>3!iIbrpqBkqBu_C;5lKVq=L2S!j7wLfL_@
zdP4ntcFz91@lLTd*XjccE?9WZ%wXiqxRJP3$C_!YF+~YMN1GAXzt}(5HrkD2^8dbE
zn`|XQ%kJvDqMUW^n4KA3;KxoD*}slzV+s+%ZWlpgHa~d4ejs!sHKg5D*d{=-+$?x~
zX!~$^cYU|%cE}B^*LywrQ1JA_3Vb&2CEz~Y8|fKXDdF{~tOSiWWV}s@8;G$HWV;Wl
zA{6@_(YTyCjKvpUX<G{pN~tt-Pr;Ph9Mht}U*Od#?GQ;mRxK4%R|?0e+{F|D{U_^)
zDsCszGwrmk3Apl{fvU|CwQ+8lAEC;`aTbUohnRM6kl8B2Q6ZZWk@%Wo2CaZ&o0fN;
ztQXTN>(i7)R6+cPfH9Mv^^Zd2FqFk3Y{9lY<OCO5BqaL)n#&0F6;>MxmEJY5bzGPL
z6)@(aQJSKu!EZ$bt2Hwr%BR!ENrsQ7#~dv)``}8O`>Zn(q-dS6ZDT{qyL_G9Aby1>
zDv@Y~=-=cKnuioTWoI4=+a+^{(;Ztn_Kb8U2O)@4V(Ue!=9wh2N?GR^6bIoUChu|B
z`r!n^M*FUeXfxmFEq|o^!vC;v@+)J6%cl?t8ujNwZZAu-NckN(k=aS=MHPqP>)$zz
zFAxexPOk8`tvGEn7FV5O9#oEh)=`p_%Q&G>k&ZC8)oc4v$k7bZu@=;8$1iHIQmhxT
zlBTfC<IIXqgUxe~^iZ0>9Y|oPuUD_}C!bqFP!<1eOWhk5-fDNp&-b2u{KaAf&BDK#
zpDk6HnppJ{>k!CTatLR>Sdu7nZW?_Kn4p<q5w1pc*@t<kJj1%9IMYd46CMw}F$tu<
z<{jW6uc7cEol&0-qr9R$$>Y8eSU%FOSV9}8slj#nUahpEGmbo!W4BM<k|EVr)NWGu
zQ(x?q^d2ykTvdwE%BGuFY>KewJ}U>2wJ_g~_%H(&q$n7|^Juh?;#3}p?T{*3={Mso
zn2*jTkx$KDici*@9%>@Iw9b+kI+LfZG*u)8xEfG#n)D=#BGb4p>OW}jk@3Lt)2fRF
zK2>Q#Ygaq{-rN|JMcdw?Qr?mvHnT1wF#CcA6tFa`G#1z}SnFH4eb;;KT(0=e^EqW{
zRM4JR*9SLe*#~!*&!tz!%ctnbE$h~TS~^P&a7A(@v)~>_)fWeR3md`BpAUCnAG%3%
z)KNS*FL_MBoCyDAvb<kahe}~KZK@rwxp8=UY)BIkhX3HK5>r!VVE&U5sDo0-pZPWC
z)ULs9gSe^HR9PEEfU}8hXj8sch#n0cXFa5b>j5HSE8us=P|QiXOz+-?RY*zCzB8kq
z5R*)y)k?R!Lz3e>1UOvKk6b$R_q6Vp>f<w8GDt4@V93wE1!hveg09+dUQ_wOJr1I5
zpROf09l~^>!V>UX1MAAR+`>~Qj{-q=pF}VP9yCzA@rc~LPB!OwI-}Q)FBkFyf5+@n
zl)NRdxOL~2ZvUAShLErz9ji4{sB@G_@;vN+lT49nuB|X(h=H)Y5^Ep6N(FMlDDg|3
zsN0CgN91qEyD+~OeCV5Q?UzSsJQ^*^bbeC2`}IX&fzZdcEf<|?m&ZQso<HEasEA;j
z&n^3T4NTW=O4ujdECG{$K1<_?@rx05eGpB&RbU2Rn<9C-o8?XT&)OC>%3q117v>1~
zgBeMmUM@fPVUXVOJa^|+T*r7TgZ7v*p0DydK2Q856}?1*vQWZ7-G+?v3dBIHm^qc-
z-5vBEbc?R(u<!Ekcd@ioln9MaT<7*7cVxxo!YLi7Y>sUK+iEEW=^>1t-tCbLH0zU7
zH++>%y+A8Q^D$+NzVy~J)n*}g!yuC#ifvkUl1Un;T}o>4D7~M~<*+<|g&VK9Z+b5~
zYxOP_+fgT99QpY*(s$82a-B<PP5lq84W&KWkkMreoi^YC<c<<nhVO7TG<Pti`>}5(
zA4?vv6n!{x*S%%-mt><8H&<5{ZOibDYA!_?l4;I-yQMiNHJHkwvmbA&B1>V&9-TAp
zE6gOol{vkO&@ZT@k)h=$i+|3Ys6C2Xl7^>kih8=jSQbFaz+UelE%Gu1Fuw#uiMPHP
zwSK~MrDs6)|8O6!CD2N0@1@JYFU-&3dNhdaTXe%?)FoG29h;(0Nh{iHG_@%?FllqO
zEu`3`mRXvgoK2!szeJJiQ!FiRzDek|5wPX&Um)fcMxf?f!X2v)to&GJ=r5}tq{9R+
zSWW$DX&ES9Az&L~*WzF)<M=uX<>`<u4}v~O%8ljPEV`J6k>$KRvhHODf6Pnpk;Rw&
z0l7iX1^;hdo#VgEplq!F+Y<l(aM}m|aM~vnpTHHU;&Qxv486wQ(2vVQo8j5vTY#&)
zkAFPv=1if)1Sv~tdH}hd<y(P}Jj$D}^DD#Z-pkgv)5}}XCZd!fBzvbfJ|CXy8z>bU
z0O41!tN-59$?b~Ga(+s4%5f!0JlY}WI2*4gQRqC#A(R+`Z-{Ikzgx)P^sMc^l>2FK
z?+`xqOm_kcq5C6>q~L>v*gA?s?NoteoQTJfkRF|%)ip+Oi};z8H_0D&@2#(n1qI`b
zNE5ct2Zt1Vc3?py2*~BwkiC_by^m{II1fbQqpY)ltJ{l_+-px7by?*`>n@Y(MtEhZ
zJ@A`^yie2lc8~gGBWD@r6TG4^YNT>CZ||_$(9_D*7?ynUwffV!u&>Qf?mWzt5e}co
zftu^lTZGYV?Ay|~IpqV+NBHSipmsKvgIMg3v5|YJy9r0>@8df{e>)ztFrbJ7K||!)
zInQU7WPbb9;e;nvZ(iwxBTqr^0SC+-rw5&;E|_7tQ?n1OI4?|(Y~<vf7-4~KgoOr6
z2<`eAgPykVWz2avha^leKL{)W7_1^?m+o$U{3gZ8XX%7S%Ne!$12r-eZqD>TxOKIv
zE^g0z))DII@6XQd91FMIL~m<;YMvUz999C-1OZD#H@B%cCb3mlzpI{NYD}BEZEt1C
zDevRYL(BuLhXqb-uj!;##_BF3p9nWpHix!r8|R+>m2i{A`sUYlOAD6wlc!MlC}NB$
z3^l@OYkdUg%4@~YZQ`}*rj|%MpS1ur?v-X%w@$5tZA_}}qPXnjBXH)(DApB`WO6)d
z+oAQ5ZOcLy_4<|xX+r^fTrYoWtokUlz8cX<B+%O^XTsRaUwH{ykR*gfRPf%XQl^%U
zxRV0EyV+Un=UB(Kq)VmZRaqin`rU;P47#`HyzImA*K5&LLQuxd=!UTcpoQZBq?vFO
zQUiAeiCSA6(z6rCBT!K;j_rD*^T|`c$gGm6b`8R*=F1I@C@Tk3)>}S3oleI}0C%s`
z-i^)?h*WJb>D!G@6Yx|TYBTZ@g!f6m^(IZWd6SIMr5gbg;6dyqqnLI=(Nq#5^x}x9
z_i6E#wDd|haw#8HRQ|-Z?XHof^Otj!21tgo5qWMbe<{0;7vV}CqlRu6P`7j`NDes)
zFCUgAURbNTjW#WNi~Z&@Fces<RYPT|Vvo3R1wdypcS*(%umTPR82dgbAK%h8FqJOD
z1p&)4&gI_zlebcobjCe#OJh)I5c8Gr5~;%BE_zNl5FoI)DWw9Jqvzy@RF%=^UJ%Nn
z;+_5Hr*~U9OLoL+OmeatE1~$IOLhchc(+*<RwPb;Dhg(|vzFxUIcqtw?6@ilkmwpX
z*J0ipI1`D3OKeSe738_<IH8~Xd#ax3N%Rszq`z}VJcBE8O*LWLU)G=G+0!Q)IPRHr
zYZ=zC%Kb^g_C&%oPO0#t8jXoA%AjGl0};2)lLZbO871@VA1$W?Ls6of$^x`T-3K-&
z-KbRLOmP4iM<h~W@oKE;+?J*LzCqZjAX-i*inMNmi?o#nQBx!+krelxPdNQCG-lFJ
z!emg7@9(*R>*SpIsvct$R6tZ}e!)LywKTTOk+wHhdCBV*8l`eO1{$R#%#Uy3R{T+K
z<IR6xw%ch+*EF}$5(_L>(Gu@kf^yQfRuaNYzG2AqbjrS-Zvz<8oWF1iX7Uurzdw$$
z=SIK=z7v0~%oT+TlFgAk$I!-Wdz^5h*+bt2+=dw%iHh$0LrCe1^2ibD>>(@Nz<Uyr
zGWUi2;O3<MyhgBggXDzrGat-%AcnW~KHj+sk$wDh6QY^SU==J#5k$Kemf>=9DBm1$
znSBp2IdgBeiBX&)(ev(N{5jAArBp|w)X!&o;3a<k^|gPVs{C&Gq;WPwAnf)ymBvbz
zQCrYan*(Xu{>>`~;(W+su}igRvh@Z(X+m=geOD&p;<K$ll>BQr+F0e^&wdHioKNU*
z*vl)lJ_MkyYq-uRd})0ud}e+d&n&-VOBS-d4Nr2IYf*!5#IsKKY%ypxeT@Bh)cm55
zf&A?H%1>S|BTy>6aetQ@JGXt8it+w9V*(6cmCQ7@X!3d$n>2ktrirwBzDqqFf1M|o
z;9fB<)vi-xt1k5`bxX4_;kRkA5T`X(UKwNB7_yf9F;WL$!|S1r*XMwZaZuo5p;~hU
z9>t@OxDJdH!4D6t1vz;HK@N)WRWPaUO_|hEF1M7cVh19vW!n>zo$?p7R8IeXpU0$a
zJ;9^}V$`BY!MY8gr-d3WUAIYQ*_9Vw^ZN+B@5PseZ<9a$03%Bi_f@Q3VQy))o<XTT
z*ML6w0YVA}7W}X6@;_r{7})9mw=~2qbxE58R>bZTmC_d4S^7*&czixx0_S}&cZ*n!
zIfc2c1-$m3pV>UIbN;b9!!n-HcoO^Ihq%vU!n0$Kf{Yr3k(iJYrnF&-ncdpA*JWSs
z*?SmuYjtZJ-+Iugk(lF|e|Yh6yTk!gY2uIs3px`X?%%FvzwSW`!Fu=RFh{*0k7mmd
zvK?{^&i>s|$kN1Ggc&|4j6*cbY9&h3Huvx58=Ha5W1rrxjNBk_VykHwX35Rhs1DMp
zV`J8EjOiUXC1M;6aam4~M~a!$MR|GuKBWGc;^0NbUQJr2Zly1_E7d*tyMJc!baQpE
zw=)I==W!He_WgZ_h>rm*jEB8g80-1)kaN5-fZw!gS*fi4UWmUuHFG2FbOM-hpJlvQ
zS3jPs3`j20#hS>4)$beaH13$eB&);L3%p-g8>kd2F#58kkt-IzMG~jYGnvMuLLO1`
zUf+wk?nF3VPzOV^7%M$7(J|iQEY(;r#IZj}U3P6m(V1siTRIR#k!D@Q`ei&4GCIBo
z`qsLfHt%k(k~Mv@(s{lQ_l98O=DAYxbG~U1n`}JOhA(NDF@k5+|M{bMQ^<_K&Bj~u
zWtl`XvRzfHL~T!mSL#<<y2RQ3&<zw*r(BpVz3+)O6%0d)uH2~di)@}Y-<gM79hV~^
z2GeC)R=Uq=#c41*De1{a$C9cFyRNE`W~6zeLT+W^dsSUNdT)C*x=F;cU?kGC(PP=T
zXJqYK;RMx$;aLVVebxTrU2A5hzmc5*+7WC6%a5NpC`_nZEB{F%8cL|<fp$>sD6>Li
zhHh-hHm*6`jYz9!%M$n3NE2({#5OIGc-=veD^lM4wr0P@!95X-6|;&B763|4Sv*P)
zRz?|g(uUzT)<ZU_D+<K;;v;X>_FR@;tx{-+T$?XQS3SiYqq*C}c=h;|A)6wg+$k+b
ze`i3+^olQs*u>;oI>(^OW9~%tyo&bR_`Pzm$EsweZV+(VfOUZSj?=Oo0T1nU9jqdn
zgR0FndqUL7`JOpqQPDy1sAiR(b=uHB=_8YaiidrQ><`@=(9{*mW0dr0*qY4qb^U1O
zuR!>U%QPrKq%-E;)F$a{8i@2*Z5#~>PbF5Rn^V=Eno~t)MuSJ$Dq$zg>19xp*-RzM
zG3B@%!}#FHGQQFlG~ovS0Y&4oU81Nht;;nparzFhq^9cSc?^G&F?Dlk*P!5qd3#SD
z(8PGte7iZD?1rlIC4OL6m}2R_Wro~oSGUqf<2XCf{2U)Dvc72ZUSh(*8*405lQ>dy
zJc5TON>>)K7hM^j6baDZ5FK-?DT=pKQEWQ1mV`@uvW-r+RB4W?diA~Cw8@DUBZxCJ
zS4*}+cS`zQi;w2r0bGYae$|-P)&cN`sGUjqdl}(pAvuVl3Km9M{@aegNV-)ri2Z$7
z_cHoWReZBfKvTS1p0HY7ryTJ{bB-`M<n$&pUQwM*d4kexMIo-a{@dBXsB0eL_*sfj
zWyd0{2Af!-@Z?h{Kg*@d5aqWFVrO2t5ivhoEg9>ugr)JzAZbxFOrD;#4&EZR5JJm#
ziCkY%Rmd{AT^;gRRgS8V7RhvJ%bd~=$77~KPTJ`aP=Ka^V5b11PkW7C06B~vM$N9<
zE{<2vM_uQZ=h#adoy=gUURc^^@^)U*3ulp+T~k$bZ1n3gn@CcX#mVw#vj@LuiRhL^
z{%bnz;qEw`SrAWA?dv@|Hu05^>zXmHExC+xKME1lqwXohxo#frl{rj;QKDZ*M9@(w
z>(zo<ssItx39%&(2Z<d4S}^42`$h`N2qx1GND?2@`Q1cHljKW&2XI_@(Cr*KhNAfe
z=$Wdf1hmx?N61Y3mRO9h{aYn&_QKOYkOTTZQ+LAUiMh|_ALKDxAaQ5WirwAy<ZkAU
z(PZ%-%ZWcwWSZEZHlUCCKV3je<kqW|z>C(Pi`KsuWdf<kka%T>W}`pB(7s0eQZT$u
z;^PZw%jyG&%O!!4-HLb~#<Zx0BD-7L5=Y+{{gv@yyHwb!%XNfgNWr&)!bRw6A{4FO
zfXQ<h$hyV9QGuhopx3IUuk(NO@Xrt7XDLNIQZC`)C<Z*dLF6T=>xMX1+R=91n)X|G
zSmRSQ(q>l-3}i|?#;S|~-`y6wz7Le*JMK@{o>dOnLh~14RK`M5&mtqOY8|;_(_5z@
z&E_G9iC{fqTsJja9!6JDcm@sjCy5KS<Bzya5e2s3=w>Y+7Ac*xZdm9p=n!-nP1pt8
z=}PM^jaj8nD3Xm2KZrzubM;hh+r1&_nvYgq{fbhCS_<N8Npc&O%tvam-_3o&p@R$e
zhe;AqZGk}7vtXufjfN1~`^>kJdWcDx@2y2LPZVJ|Sjpy#1^ydFZ}gTV%)ahCXkLo+
zPh4-*3}Zgkd;F3ArM<Gz{})69Gd<J)bW*mYE*^v3fz*AW;%M#07i^5`(u*-TpTQ8D
z5!LoDZ#cl!{ub1YF68q|(awT(O<g@6&d?XHjG{ig$w%|>k#+3s!TP8E<F@kgFpzE6
z*8Pq3-%90-SIoT8*dC~#GF5xKq{ln)Hx0nUy)jW@?%=eeME2n|w^sonSI5^Kh?|EU
z&Mi{F-4H8a#%;1YQjRZ17BKsWk~Gy*{;B)9#9(#QiVi-KS%!d--FL8~{av4+CPv|Z
zRP#C41L(T+-gqL?2mrQq_u8@RP$ozmfjS$}E#CQI@Y?V?3sY8(4=(f(k?HA?)|yzU
zd>{}%nO8SJ<Ae|oo=s@tC;z56)BK?<B3xrWlg2+Kg!@Bn_<4J&mz1Au2v|^HjwO$;
zFC?`xpReuPoZcBLA10QWWN7Yq%-Afk+tg9Wc#g4z%+B<bN%+l*k~T+7L#^KJ%~KAn
zsBL=l4pq?`c`u}uUD+lEX4({+3y#WdRL&WZj#?qD?C97xne<+FNW2q}G<RiaFUV~e
z#uB0YEgXhULu{88LzIw8KM9cxp5<Y&_mqnO&!3Gs!;N(p`zT+i8(>!|1Xw$|A-#wn
zpq!5#^9|9TA-1jaz&%$WM&}J9-jsK%`$DFok1M3DHytfC`$lJJG&N&Z9O!k;Bkv9m
zmI-!x@?SWcB!X5{{R#(*XrF8lI-pn{Gk$bcGUjabuKH%R7og!Q^HX@jXntx!t4zj<
za*Y5pfl$S;Tb{tYq6}9zzc6AovQI2Tx-|&j84S=MTEQ@QazQY!ap)1m%Su3)8No4^
zU1-goU^4Ebs!1ffHde9X2TBTJa_-_jUxr14v-4WivZOCR1@<}(8iVqL#DW%cY-XXN
zNItKQuzq+&G`TXDG&of3i-=pu7`qn=2wt;v&|!7(p17Z)u_%xpf-VMaq-82MXw4Z%
zIPQ&LshZ@{hZ$yj&}3a4%Ix*nO7r!Pca`4ci801i`Yia;bj&wUY^nKOJ4k5Hp6;57
zHh%7Rz!Sq6j{$%b1Z%-YEviUAt1MH9CH2WPa*gzw;<;Gm)~5iLb0Hz9oel5eW$Etb
zrN9RV)?bIeN5vi+Y&o~f3Fi*Jq*Cqlw-mP8Z7t1NjH*`#-C+WK3w<_l=E(~CSdP*<
z;4EQOcg)an2wsPjhXd8?dnR9!*b@uQ@so$+{~fXyP2N<1q2xHK{6qr&8~EtHESoie
z&Q==~huvvbKN^mtH>less15ry{zm<V?g%>)QXLK}nXQjP(005Hi#5kshnk%@eX4j!
zt5TGZq>`i1K@5&iWe?BgD3l6mneSLyw<Ey)1oyeJg;FxdRn|sJ{xfd7{SfrHv4z%S
zXTJ(}hiQA=>{Y}@Sn$H?>F2Eaf(rSdgn&7}(Si#3%4|#bgi@hJ&eb4*0PSDN4Z=qy
zxbk2wXvs@k2+fmPl~+?dhoCD0E-O56Je#Kw-#IvmxMh$aF~ne}i?iYvK>|{g>hyQ{
z>O?WOz0Mm_JpNuE;JYK&o6T-tM5fhv$5a=OEpXAy*J8ZXBM!wJ(1vmO%k|5BHQ^6a
z-aRkawW;tB#wRX}n5bP_8QQ|)@u0LXD)n=sTs4`sI=w1{mkrD0A5?}hs!jPpYr0Xa
z(R%DgaZU#zoJG({(ksel8XdY>*NFs>zRbyc(Z=NPd;jHD3UwDD2LQtgKaddOdN4_g
z9+@Q!Krps`!5eD4;uH)i^X9Y)1Vxe4nw6Ty2U}`3K0Y%zjj3ivhORW@o#&EVr-hFD
z)FuK)#1}S!2yv*a>(Kv{2I^lC=W^;t`}M@PRr$9EP@hz^PjOYeN_%J&btNoGr4W6T
zJ)sBg`tUq$^tKwOvLN4H5IQr3R-OX@b{MPSiZ&vX-=Sfnl*XLeX;^kdc_^MOUe5<n
z+SRijF^{n8k7s)Bs#!V<jszWj>M?*7WN7;i4WCEv50{uBdWeXwh$JM#T%>jw)z%Qf
zfev6dHS!V;nT9^u+zqzp(y;*rn7mp;`8@Jh3W-I2JL&nOCzgd!SgPdbs~b$_qQ$DH
z$PV4XyNW)MkG~!&pr4v7aKGJZWFdL-zzUqRq@GI5(2NuTY#i=e9d9>DW?Rb>j~6Q;
z?O3^7QO*em-wbu@07F{d1Q;$%)$35=5R$SbEfY6&$c%VIbvLu=a=(@kR%jW=tvv}f
zT`t+gQ;E6Swj6ArCKmR(5oSPz;QI7TKI_X9nAA;KZbx{(;+s`VbE~6s5%;AQB3}pS
zG_+pGNy$t9x>Fk3f;#X3RaM6nMPDJJfGABF)5$51BkUGUN%LZ)<ueITJS3=O;jU1z
z=^k5YlQm))a`XUKrQUe5a6o(Ao7b!y?Af~($g#r($XKt>XS|SEtBT(P2wCv$XLK*b
zgeDR9lS}Etlkuz-6bsvgGTawG;ckliV4mx8w1cNOYWZukbaq^Nq!;fVP&|--p+bIs
zxr9$CvPKz4TKy865mLSjCOk8H+1s(%`Tl%&fi69};@wn|Nv{V3<?Sj_5gaCLvwhFr
zp`oK2_2UD@l<o2wG&WN8_R4ZjKSv-xu-I_1h7s_P_HBiYzz(O)o;B*1R@j?bG`HA$
z?6RV6NW&CTJ4bnqDm$$ElsK3m_dY7|qSG>4<Sp51F?%7gSWtDe!&W8j@0;IAXY~U}
z@mI{v7nTB<ic9WPChU&-L^n3yFN>foYgCB|(BTY0kqu2&jccl7wMin+QKCpFyj^a@
zoaQAYsUM?~?}=cUkZW){OOLKgV`caBbf4JSFrPa-q@Dg<REWgl0uUHnAkqa7f9aR6
zRfM3Gd7Nq8Lmj$0IGNPhsz+2tL)LvVvM5N@119^u+EfWaMKTr8$qVVY8;D+~A=Nr5
z_eVM8uHpIR94oVs0e?f8O~53nU5~3_VtM}`WA79sI<&S8ws+gMZQHiJ+qP|Mw{6?D
zZQHiZ>F=AGnX38E#mPld$xW)VlC_?^?}OCr=e7=7TJV!PyrsmNg3o}$`lfzb(;bIY
z#}{p{$JJmt-pgFSFo5x3EHz~9=z{wUUW)R00@^GOjhp652|w>m3=UFC-OdI-%lv9L
znY*%U=4{)7P&dfnlA#)%4DZg@urc|&*db>v8`+#hIFg-SOpant>$7BK$AIMP_9oYo
zeXs18OrB#is+Ky!V7W-7Nuf>+)@&hE+Q94w?;dX{f-MxOmVxsYGeh~0_Gn9+AOTOq
z+f<qp4(Dbo{c*{4E5t?3^-Lu_$&8<fAF9i-oFCu|iS5+?1lSn<17KrdWBsoHTbIU|
z4K@d2*NWO9T-cPJ9VleKc>P50rqwHJs?dO$C#-5giqOxua4dBE*rY=esuVvZRhkjV
zn}Mg%`O!vx3f!N+H!nGGJ(=3t07(x7uSX9npr`+ee{usOwrO;*g42K9Nq#sitUH&V
zvuP`ugxZ0=qMck@*vyEYe#rrFEk9qMH<X)HNL4VN{aH}sU{w7G;<0&=*U!Q$LSHx)
z{l!7dC)v&Ax=-B`<2$~*ljqcakhNa!+@Edm;CRVVg9232b2LBHo6?4*H81|x;=uFw
zXwGj(F}AutPe*F<%)7R@d1w_;fygN%_o?E{oDE(zF{)=_)3mwqqP8iA$`O5y`uCs4
z9O_8OxshFWTx_Jq1uVre!t&_Au1I@AN-8t^ngyrtysYCv?E#(XtbcqAD4ePJ=G`!c
zV!Zy%%qNL{+5Qj56%Q*O;w<KxHx}v&0AgH_AH6TJ*pE`vRH0Yg|I)YES=-E#d0b46
zrz)H>R%?I0v1v$hqE^qGP=in){gAgM2T)Jb(!Y?sP>Hq=Q6kp<GDcpXX7+hU7x$v|
z_mdg0qjRw7&8?rNV@vh|Vwi|8CCzc=_hB0Tdqvwta(!(xheI-Kfm~G}!7O!-USNqR
zsr1?ktwe3vbr$*M#=?(W6TSJ#$j?}Xqjws5*9;i=9NG(lh|5KvXTk+GpmZN@N9`Vy
zSMQqXRj8sEcL03`70W%%X?~JTI<QH+ba;F{jB09QyT0LQUVB<;>w#MAS>Et)aM}82
zf`*Cy=9D6tIl>Jpx(cqhhNZ`(PJlCkGa+fH{c)j`AGNq59$L*xif74u=ZQf!LMVtw
zNnIFng*l^NuMF;3S?mTYepWS3RHlFqSVTP2Y?<Pg^SXWQVdDd3w_9}ZkdszH(Hv5U
zM!p+g<~?FxA2FH0h2~ZewSppLa8x2Kfy&tZu(NI9z08wKxW<ktxT4%#)i*DFS?9gG
zl|%!e-+rcgv1HW+RxF`(`ITKJ7H2vS?|?dj(Qu7v4W(^J7eN;|y_${)cs6-<pIypG
zVhy!9ngb?!H{-7!iI(86y^@%FDUqB^H9$}L9J;8+!G||ny>f=2qxPJ<VB}$Nlt{S_
z%6ZpVGHzP%95;Oogc5p7Z6k^t!~r*JhhV(@iKEE1)ivlLCFv756@oE{g_2A>*tmKz
zsX)LJV-sxfve3#HuecP0gC}Cl60mk1yG+i|c&2gL)SU@@i{_W=#TI{%x8+n^gRQpH
zIDAneF@Z_Y`Fvh<a%t1RxL{gn_@uA(P=YD>k(u!>K9A|2@+zTTmGF<W^zCA*2eOYK
z(2RjMR4>lt2t$Ka3c`Z$3lOW1BS>m)6xM8sYjUD>AW`?`=NQBVt+Xiyucudnt|gLU
z0#a6c5~jWbc@Q-O<pwqwA(N>@$b8e|j7AC!ua~LRp&<!M!#HTY^5IMJLnD$mP4|eA
zr*TNc4s&$D%p%}QkL$U1+1zKdza?hD%U|7<9ZOX!GCQ7MVIfP~sbxlHO}<$HHfD16
z*v)y^?q+mHahpV9+RlZ9Lw}V>9@M@0264o3shVcDS~UJ~6i-?#+PQdDXEBm7Syhiz
zK38{`&LGLaxr9!&xsaQ4$h~3~2;3N0i~y}I&lfTC2$D>Ib0rco9Zu5Ul8qUb41cqt
zmRte;ojNONGOEN40=cE|<Efsr<d9<Q2EGHwc3~BasN#B_XVVO3@gprx@R+7RXyH7w
zt4M~_PM=H&8P`OHbdvdMCU+6yxO#cUUxtBj|21++s{G}Jkn<Tu(~9K)b^>}gdujP%
zNB9NhXr;uKm;D*81_s%0CXIrnM@`>i=`Vk6L?YWyA3oh$pl;-UYpJVnP*q7ql&{Lg
z8UceOfppkTY_WPv>3EMSAKLBcn^3L38_1H*U?e`0<|B$!>TC1qu(KHf9j5r#Kp3b-
zCs01JRs>QZ2-y&V3IiXaH<>F0j8hGUQw`j;!kxq|muZVvOjaJ~YoooGtL7hbi3|6c
zk6o5Vayy4b?fcdID~QT08V|3Sf9nkR;rJ`vXz`{jXP7_=(DZUy0PM%FT1aV$I5qez
zV=}b|1R<oa$0*VKrkYOz6kwBl%F5LTnRMhn-oyQ+G9hl*skS>sy>`l!I7o@7ciq)O
zwA3aV|F*rt@i*j=FuSkJXcOu7us|9S&g-?LVp(OPgo<9*a>_s1;~(Ob&=HT&@vlf0
zv*7S~wjlkKWzi!1CdW$6(C_)yd|Axg@!yORF(=C8Fy!quDB{~Kn+<XMml8zdc<!2;
zzmGScV{&_F%X351vMKSSPj3*UU^<D;Gs_~6Mv5v?M)QZ_VMCOuM`6htJnljzY0PC8
z(Usx;Flu{_CXa|KG1YB}pTvt#asf{9db@y$ilFpBWbW>Ggb7!c_lVf~eB17G9{R4`
z=b7kt63^u-BdNE&Q)I;C2Jtm!`)xI(moswl<`BnpLH~sYRwd|!&QzChWmGXpL|x7k
zTqx>ln90Lcs292FI^H|$=ZSn`XQ01Z=8Ig-@LLq{QQr)HC*hgU)FZel%+)~n75ZO^
zZ3^JKlC{70*O=}n2?tiSXmKw~qFq3<&n#E;my|$ow2I|+d^x>x5v$2^>chr}x}zV3
z#xna?;NJ}rpUw?Fym2M%{=?JU*?jWFY<YKnpIyV<g#U7g|97XGd#7~6mVXDJs_|$m
zT|-($S9q7um?L{Y=3T>oU>Z(Wa@R13xA3|KrIRyER)9Uo=F-qKZ%X<kdkC!0Tyt38
z(1mK)qC+xXYWZ{m&ZU6NRP)p3)MbN0I#ePf=ovrTdm^wmzO)E45cK&_=5h!}cSa(u
z)r|nWoQ(d?)fu3#Tj;ktR%72+*!wv1T?hY0ky)nrgJ;M~RrVjy3@ao3|3`6DcDFO8
zlQXbTa<ZnA#b=<WrxP`IaCE}wU}K^Cf6_ylS^wM8)Q&AS8?4WZZr;eY5Hle~_<SCC
zEU8SjD}^302fnL69=2&*>&Yc%#|3Lw&o{3Hw~p=d1>S`OEyZjqQ$Za0{ehTb14Tmn
zuA+$pGGkD0bi@|K5`+jivUb!Eh(lQ|#ATuteK^$^n4unxwVi`-p=WG4QS_2|FAc2)
zTm#Xb(Zrt4q>-Y1Vm*&>W#&3Ns#ruD0YzgtMuJ{*{*sM|SQ4;m>?u3(g3uy^vC_W^
zL)hYYkP$-G@YsXZWO$d%kCL@IRu%dnmW|dh4ap{!S~3WNHB3-NC1pK0lMTZ`C5mPn
zVzM#}{CXto<B0vj(y6E^sUffCB=o_kVt8iu;5H%zLqIE9#d@^5gwZyMC6PH+>!7PP
znj~%MGKx@FN0wlu4)yhgAQX#cHmmYkzTlbK$gXA?a;p|;i|wgc#RvWBt}6lwTUc>=
z)xHpv;~=SZp-HU`uq16*>jg^K>=8(VfOW`O%!H=UF=O*>1?N@q1j52&kzH~ub$tEI
zHDEcYLZAI}2L(^yaSL#i5TH&yJ?;KAi%5#f7_<%Y1goC5-EL*21_QIk6g>`^+D=S#
zB|b<s9$3oG{CEaFo%!ull!m0^Fv!CturW<-h}E$E`QXy7bo@S02_dJVp$IT%lIE*{
z8<KScc{~9{Ky{cyJ?T2nOG$<U@&MS>z-o!VFq=B?;=jT=W#c-MTwiP6MS1)_voV8w
zQYg(}UX+l{kGqlUrJRxko1fdSLQn1tTC#<sexMrsQC%lG(#W;7PqX?{!#SHTPIp(Y
z*M-Z448IA~)g-vvZ){s2g4%$89?V(O>%zB4)p)#{o(ri!4l~GI{QB@_@F5J~{*DB&
zf{c5+f+x+QFZY*M^VQkEq@aWL9Fqsv1os6RY$x9fpEKN^yzl*?L$kLK;#Z?w#HTQN
zdEqDM6d<E)CC*b^4(XqIPgYNsi|kfB@x6bEg<!(c#AYgwbZ~b&U5&e%-t{URuK<}+
zsCa&gv2k=wRu+Beapc0#$%KL|BOjIo9PV|i+$)5NPw{V<G`m0m5CkBNMC`rVFDrzM
z?vYVYG`kRhmNrX-$BHTlT|v9R7@!bDW(XBC5^9>{b^&^u<)Eu177#^X!cKmQnPRg*
zX2?3`fQU~le+?rCsl$tu#aXZ_V*F$!`(6fVw(r@AR4rJ(;Ri>HQCB&m2qaMUV2|N(
z05gg0!bs#kK~H0=*2oM6*(J|IV5tYg^Ys1LB=gG81VhPkvsD{APj!ASc2t{Q;~k~S
z78()~0%;163jm~ugO<quM02N2>1N~N%4o~U=XM`CFy^#px3=f-(d2Y#<I<8<mo#2e
zsG~gaiQEa~G523&18I&Pm-X;73I?rw15LvI3Cd>a_<FlMJu-SL-tfI|^kjK!<Zd<k
zjWW-805f;o3~!u>XmrOnN@h4k+ZL7qnguc_Ewz&Y=bDX{6JXnH@0A*9D3i<HM{lga
zt<-19_#-Lcb}m3VeTweOrYbfsf<hE592bMM?~S8;^3Bvr{x`Tp3K)z#Tv)dUys$aX
z493beII$2$Jp2-fVooX+bG`@qOgH3_jvH(=c7-0{zJqKP(;a8GpC!ERJZsY{rV&iI
zBnh#OIJVSAWq+d3Z7{jznBU+&qBsbJ+$gf=aIc}p2-3)@ktF1^*977&AYD11KeUnM
z7^2OVW|MfkcW=jqd}u13^IBQ4q!yBdLTDG70VOmwLP&bZzwbZ=+NbHK)>JHol3AJE
z?7HFTM>TJ0d6QJvfM-=}K<MSv;Qd(4?Ky|er|!r1wN^Wu_Ius=CME@Mdi%X#ZREP8
zA*KXwTgK-G_lEE1t?B0q1b62H+qOlXAx=(J9F8(ySpbE;I$|dK=Eu+#2EiSn>o`{*
z5}yR$O&ezBxS?f)>moHEDaX8s2$Iif@ZkLQ-HBI&+GRP8-J>=^DDMwxgAhTj_yae$
z&Ow)xHa1Sqw5)(_jPKe`{_bC`N_p;qi!~2y;n-x=LCQUw?&@_7)9z9>WX!wbe&-GH
zu~WzLO#@<uyS(GLQ<tjmuE8!>lBGhd!t{CBLQS<o%uuAz)ahVmJ9ErdW|$EqitQeQ
zxb)5}2V(#tFvHad%U1Y*>RXjk5IT{PKt_2Ov}VH><kmSy#>^RI9?c1gU{E0TW_8}J
zKvib`d7$r%^$YR>s$m6>-iIm;6aHOUO$=ys!p8#iomQ&UrRlxj<W2erKqpi&v*fuS
zF(Pz{^gpW*>#QW?`GZtuR=09xfXkFb>4bnN27rOAMeA)IfZnqxVk&(eyDr+G$`1Kd
zaH2MG<?__)fDHiPE|l(ZVD(u^TUq+14+m#^PS4LVYoCW%zMhPpJ?vRDSO<*C*Meu$
zs0nX@1r100eK!ej`&M_mhqpbxyzJ>RXcKRH41!y}>OP&qqr(amJH<u1SW8SAQGE+)
z<k&0h*SOZ++g2A3*6Z(?6PzPQQrIc3m7p+@2+~6|_>~6qX1g{^pHuuqJBz6&Q^T3F
zh{1|IAe76TVb<_%RFsKq@_BpDrRr;fXRP*(tmQ4}YIo0XR~>b$CfMeiV+Jd|7aObs
zaWF{8>%o$enhN2tJf*NaW<8-FOTv>UR)6D<w8>66uPoJsX-UJiB2AxuJ=p;;P&8Z=
z^5w?f#hq|0f-%06sd$k{xWj{!^jX76%u0!n0ZM-wpG(-D2{4!rFx@%c9fml(mf#_>
z%MCyck_Ihhg!eWE-_}d3vIbEKD*HsX?t(u8?_ejSwxB_`f?*`v2-dVQ>CvfFp0I5h
zW?^qZP)=VQMpf7#isxxnt3I1xKbjG)vRb;=B9<E$mLFjJ%GH({_o0q{v<T0$MY+Sj
zR(!ODJ9cp<&G^3GCZnDoLOM#wKa7V&BFazFB!tjGFoPJn*=zC1^301Ow1Z_Ka1xp#
zt+F0XPNMirdYeoHWPl2hQV?X<nA$Nl&tijlbPIxp{d)J~iv*oNmBiT3*@TqY^VNoZ
ztVMVUAPp>W*d9#g!@B><#}rs612rcw@IH{}3bxAweL>EfY?lu<jZlQKeKbu^r8#;p
za!mb6gwuPv;iEKXqBJilEmTonhfA0?v;&{nd*SE&#1%J>x)p{9KaHaJFydPfGB!OO
z<z1YT(-XskmD_`z>A=Z`ccjARaQkqJ$e&Xi$OcKYbI&At?8*7EI9b@bGq7jXa9S9{
z+nNhqvtq#58(g1W2^9YTlP|aChi8s}WdCp*qQECifxw9q-jlDBIjfld0faUkY5osD
zknLC2^8ZhhVr2NukN@B87YJH889UI4TIo9(3mY5S8X5EQLOMD*80%X@x~=~z9;&g~
zDtBs9ogm;R@`=v@qCh{Z0nQ9Mf0dbo)!KtUx<T#7xh5{I(Jg-ekQ+ip2}L4W$UEq=
zE`7LoI<vQX%oaHI5Hino_+0tvh`j>}1tPP6;Z=oR-kUxi0u#sQANXV96fHflzn&le
zow#ChPyLp&Vjg?nvY7aZFi#cUw7O;=(0y}uhFtBf1Co|I*tL$=%K264jE@cdZ{tib
z&4mU<ZUE2fPI`2F`Njm3k^6>o?jnCYeR#e+_k}gjg71qzJ}xM_>3_v05oT!7;y=%F
zFK>^3Le~%C8;N~$FT1_ZR!2+vkMYB&<~0W)G6WEBWK6B$IlITLSmhK65lqAWhJdpT
z5`spFUw{Lj!V*>oxiJcc?ew^%4q2F0i$<#;+`!<_BYwoG|GQ(u0XEnc*p$F@O~`tb
z??Mt~YZOOB1B%VX3^0NG8xOlY?$&6S`OdLFb{ibdU5OXoKQ>wd*utm^!ig>$XT+@q
zh!7WDlLs~l8z#UG4KrAB1z&{_Kb;{oxb~osT)6ozx@^Q5xrF^fY+Q(wVqX&&Rv7$G
z88fwH_|F_AEUz*`L{Jr}q?j|vQWyer{$h>1X09ch%<B=6A?UfytN2H1RMhia2I!+;
zg8Iv3kr*r`a!=hSU4(|RWGG`rdF{AG)XZ9luG*bh0#sD&%9%S!H1qR^s*(IyoH~+>
zf5k^M711O361UMuH0O^RQB*kWB>IAl6)tdT2d~(EwRLAeCDuPd1qid2IDzWb@aCf)
ziW0s1&C@JKU8GTqjs@9IdYNFWjRm(n!YL}T`awnOVEb?RWS^*;dy_A4ZH;`WQ)XBZ
z7eP|$F@ZF=U&gmg>;e<|vxw{(f>9R3^3o)Bw=ctN9{6-#4yP<5A1Z15^FvL~U{T}X
zB%MM4EBD`frN0gs6Jc*H1{ic>>dWc@A}cWD{P9Q%huQd#&mnEI=)rMcR44r0MY`qe
z2Sy@hxTl<wq|Kk@&rCIPtd*UBYd?%uU9Mjqm>V_^kj8o->mvPjWcqIY^x6We+VpNT
zsiJ5g&^kj5QrX`t?CYtR2t?4-^dy)zFj86G9DWlO6*c6y-q~vdpX#5*f^rJ<Z{i-E
z)h8=au=DHi)lnt6$mn|}*|%luGmeTJ^AH{^pz7Ac>#aGjgyOMon{dcd7M>}scysa2
zu2lQqmJB?h>=PGH3UBYP@R{<nm!U&5Usj;~tFxCnu=<Mu!*Nf)(G4V5mBVogb)3+v
zmJaFh!>)5%$(5*ao_(pN&*`Z?GA#B2P!H*tE~kJF^P*fX&nK7Sv{*1mjkgYVpDJ2k
z`kIi3QiaE1i8>;~Dp=x}k>hsM9PC->blerJ#A$3HQX%wseHfRaci%IT-s**U{W0%;
z8c1Y-jaxP*;2aPvtR^9h<W<)OCQ-om&2G7A`tVvx_3L_mzw)d7ShRauJ5qO4R!>$j
z_b}0*=o-wr3+k?MIiw&^*AoYr0ZbI+0eBU)NeM4q%oI+$>BwA6vlSdc*t}(~0E2am
zZk}M|dNo;=l3y>cd?1ngZvR$!@i|fQd2L$IdpC}P!<*4vJxE)FO}102F#IiK6D7M2
z+NVHmM6g=uZC0|uz*vY$EWsRvVP?;QLoTXJLu^e*KS>U255a?5Q&1g;I$HX(xS5A1
zs<NQyX<u$O`xf=2(@&RMeoM(vo{fnB9dpUFre{ZP^gDfTlz)Yi?7)r=eDcGNI-}eb
zX9t90q!dTW&o37}u%d27aE$=a2J?a7693M*SvnOztcCdzqK?&F)geg!(SySbX1M_Q
z6Pmz2w&$&SXd!gQl4=`IK(ZvL;&Z~<hbB}N*}>kK{}GOISmW2G2ZJZ)w1HihhKxF>
zA5tfAeDiJDo6UAWv5cnaxv0R9?taCNyqDH@k}{X!5~0N!G4YA^Bob=HGK5Bq>Wa97
zks-(6-9O0^at?7WaW7}9fIf#I_*m^^`qZ*kYjZqAzHT)qDHASYALM}$gy>`+#Q8nU
zunz`<<ZT)`tX){Av<F`0CyK7$9<0*d0s&l`cN8M|Su|X&I|$_(%#6X8Q#xwL#gsZ|
zIznH_Jsw-+JoVNsP4wn%Cr6-s6^Z$xAk=kq$G@<v4-?UnFki{AWQrdp2A}J$3$%VE
z_iOD5Ii`E33;{BOz;ADGV6r;hQqSeV4DtCyM|*e_Z#TlJtc;g=j5pbBe{i+fbvsvM
z+B7|=JnCx~6rha1X;kB=ory}8jRD(Jmjys?P>v4r?j>4Nj-rpsZIuL7LhseTL>Hhh
z!*FsA%4^1|?8DoR-W`r!zifM&8fbS+u&e32x?%*jmf%%9M%qX7+(b{j;?kuIvVUJO
z#LesAN5$yyb{g}TZJN+G5<**pJ=BVQp@@+QL00zWURMn&RsLy8V=T2*tE5=+ksXQ-
z8Yw+QNB@12a^}8A@UFz%ZG^*sw1-?v=?r)2$6RWTbN^^E+R6uA-yQOd6Ee$fSUsgj
zisS3{_<db*GT`^U%inqoA6lcQ(CyiKg*4J@oGs6ocy~lnj@8?3E^{)e+p44-CvUat
zqEfJTiX_qiL5*ER=JHutb<Ud;|F~!kYiXlxMi7d12x-?1d@(@FzhD1tb%Y+M#Z{g)
zmbWE8ohF^%FwBT&wmUz(F7y}JRGFMQ(*J$gTrv9z_J^2*Cbo_uOEUtpwnTK~>rzfa
z%kF^K5iT8up32|23$L*vX+hkcXm9&zn->ENO*iaXZC746hLpI*6+d@vY2AU2pbn+j
zYb9JwB~!oP_jV0+A1*Guzr;KKDR|EYnP+tBZeg6;Tti!vba?UgWPbzuG90S>Cy>JQ
z9|XB<jQ^c|#L)k6kwu}qKh#d)j9f<dI{onZ_y0jXE{oWR$cebhd7gg%2yw_sB;sf^
zr(+M-9}75$!@eoR<N82*yk6bjySzFuGP|KryJ2s4!}S8E`wBRgGz9*UpYFYVc6@zM
zuyoqgX^mSRwtM*GyItbtK?mmsb_0pt+&$|<vWo$<QLlE}cai?d!JSvv%rHNC;+<MH
zW>p<`D`$#H>SvlEzB1S!I?!u-`^~lV;X7@osmhn~T@cys>c;uz8I@T-O%mYx{CbE5
z8}(xZ4D}ly+FRipcp1QpH#04A7kR7mczrU2=h-PwE-<07JgLkw$6&6=20O`TG%qzv
z>$b3Vw;fz3+-9o_fj9B|bbz!=CNytONkXRC_?Mb*T!onY@tH%R#dQ`_L}x#ZXt!Wx
zJ2BOX3V~eI)s)kDqycB!CK0%c(l3aayQcGQ1tL;uUj~CJkkkCYS~$BK?|Ls|bKD7i
zsBDB3sq(J=R+Hhmw0wwy(Nj`&f=st=7diE|$dNUj6@1QN?VP@6`v3+7iMFrf0A{I?
zC*^z=&QOUloHOaM%%s1SWSdEd$7!pSwZ{P(hbeCWMY*K0&a#Y3N>jxZvD*+B2|#i2
z{1}VThYcM))Hua3{4r$$DmU&9FReO_5oDy=rsCZ>)4on#a_8H0-9~+TWHbPeLSW%a
z8?I@J3^=XqSXu8bYukC2PDZng<ccX1FQ?7hK=Dw7oiSs`m`Op_mNR2ADmm2NO}KFG
z>JtxB>Y)-|Yw6aDNRGvc+bU^7&d7K4$vgfP^x8`3sWWFh<5Zg+5``pPqT@z@EbDx<
z>X>cX-OpU@oK3-tD|Ws_G6yyc>IKjWy0Gam_feHI8jK5*d%p+9gUZZ2xBdB<(~^EC
z>uyNgGDvYnFOoc&H|*kf4tzN+z(DD6TnSgmB&8(Ij#&jier7Dl9<yP!Xqg$6uf+eD
zCwoj;b#100khjy;#&y&N)l#RAD|6OXHRGtuIMk&LfzhNXw$bFALB-)hKp&4~UC-OG
zy3+ph4?e5cZk7yUj;GdlAXLvu(p>f?d=6A(S)L<ps;b#g$Ic2nG@kMNfuohwhGB#H
zker@?#MEsI=1;=WJ*8D&UnR@T;Ti-1!??C7&rbF1TP9UH!q2g^CdbKuFJ{bzpre3@
zw)}+NdFHby^B3RgoY_&E#rhl_=nzw%L4sUB#JW^Dns@w4U_wL7pQemlY6IoC3E;ji
zP#TT=MM$N|?7lOX5cgKiq^)AN>&6ott`<S@^B9X+9vx#XCjon{)2F7>%jx<E{BnrS
z*r%ZhaW7D@o){=;y9k9Z<l(3`KD>j=TOKS!`;g0Cm&rAQzf}GkmC9g-?x%@D{yE1F
z8%(Kt`$4VjZ=KL4*!FjpE>Xw<`_UNE@;YOB2G4(t>4i!!i%VQQjp+NHrlCnoy(Phz
zQe9A=AEgBBbAHdOjp!LZx^mrw23a`kPX=)WuRoe_-yfx$@u<}-v6Zl7aY>pb!-fMo
zg(o`&XC|(ws012Asy^3}b{mJ&xEa$XbF*cfa6HqF(IR-r`h@XZCOIXtVNsvmCgC~o
zX&+PgfrsX8$^bF0*Nar16E_@WEVufvm4yVSMp{O3s~9s0lk3?!)c94(-=d*Cws;|M
zUTzY?26)=8zP%T=t3*W@ByCy<G5AZcLBfpy&`TXyKfuWYCl?5>QW-&paPp#ZJAsS`
zFW~|hQ|H57@Su%BR=9pIK2yk(b}ifK&lqScULtsP@HtpHV<&?QNqwt#sZBPw^xz@>
z`dZ(Xz#s=I^zMqucixiLby>tWxHh&ojC-ywoI6IiB>&ie)drkH`Xw=L>bvLRwQk)%
zRYFDR-P{(P<2Xk2bA4EeBE(H9vVFWGVRT<>luAW>APWbar}IHzV3xa>LgvH7Tr!tr
zw#JI+x*3%O-g^*C<a;5yJ)pLYkH9Kcu0Y{&2_IQ&mx_aZTsae>D-+B!Q|QQh$wXAL
ze}|%3NXl6!y3QYWJ)*rt*45jr<>>g|DCP_!>t8;<KYp18s;!i4Q(#xvF?An$;?u_<
z!RP1(H*KpTE_YA`*<6epj-m8zSp5}?=`sbX9^TdoA*Gy2`t9{z2@6i*x8BA)^t1S0
zlq{gRziPi=l`x<F0t+#i!>;DgiYlvbPDNS%e^EKbJ^02gXv{iCopKjrfQ5ZH6ij7^
z%Uz{Q@e^zVo=fq&1Sf-SW?x*tY2*;o+;ve!tMxuc_bchQtvYGx7t@OtMY6J>Q3ibb
zIMN*s{_l7P1li92X&q$wKi%jVm|6e3pY<t?sb6D?|69UZ5Roq|(M#A%%tSG}f%^YC
zQ+&ZmeXS*D@~LI#{uv1;Nmvw%cpn|z`@j#K=@@W-IScZr?Z>FKVE4uH`}1p~bQo#`
z289o<&;En{{VpUa;)U}%uSj@r{MK>MvDSmno(;$iB-X9>W59BY1k$bCUV)@La_jwr
zINaZ%pz5^~vc@>2s!`8*lr}=Wmt<-o!F#%IPrujw^YLj+_E2AVi`*Te^>Ojl`C+?4
zi<pImZ~l5RLJMsM;x7v`{vb6E$M=KNef4cM6g^JdlfWB=?c;-&PNvH=t;0ASKbi_@
z8VlWx8`jxNpJ=yitB9$SWzxx9qs=4=nXZBgbGK30%B??CfJ(C#X_6OA!sInie;UHT
z_BpUBIZ=`j80n%#!7NY3;AWl}lF;~vzdbUxS+L;CEC`bZ3tw3Z`VPm61$*Z`)_{i~
zX)ciaIR2Ny?<KRh_8n}r3ZnoNtZUJ28oiNM6-B?Rmg|iY8em0ZuipSWoD^f`_+)fV
z6$sR2Ng`-H!@L2>*s4BfIh<OT9Eug!(NAj|wDOw!OAz&Ifzmd)lh3Ey-YHH6f0pzi
zl_SCBoK-#q(T&lxPNJ7=YEX%W5*oYLlM2%=L}s%WS;Cd9m}515dc~IXx*CYzmrD`a
zlJ1SdBC^d_AfEE0ps(F2i2)b-y?rO*Yi@8JLa-38u@UQWBFt#df(=gB?gO=jwC9XC
zvTr&wY9VOFUe%bE<s_?@)0;mrq%2s$7B<a&msa_mB)r(8H07vISu{uRf@2xOfC$;M
z)JdmhQkqnZLK&UOjqbc&XRUD3SK`%<c%VzageOK`ulqX-z7pOQCCt&|-Pb${#EBQl
zSEAu}lP>2Dj}B@!1o<;noTeL;T2rN~Kn%-z-Y$G~Ksz?{Y?tbk#9xnwSls5xJY7zI
zfXe2`$G!6ut5@)*C02$$u@axL)8GILPk!>h?+H1oUYqF_mcs7@3o5eJxG;JmO<c%U
zhHb?Dm=@CQMT7~(@>VL(+mt7w?d~0RB1-P`3YN4uD&&j@{<-4}yu4F?JT_H)3~+;Q
zN#_wwLIv51y|Ro%DEY%<^B{}N^In+F*xn|{pHjoTJ;cQR`Tb!z5OTrJy7K88E-;Lp
zjV0<%FlCM9wlhDx5MM<DC!|lRrSXgF*&VcnBqyM7$m8}GW4cbCiZ(?pq!i3aFW;-j
zVtK&0LbZYZ3u<jJW%vT00U!#>UHTeBIVnarDMmOAm&*NMf{{i%A)(8Ai-s^7oRe=)
zR)4*^OCJhtL{Y?;Iu{Vz6<vhlQ1IG$%vDRCIi&isJgv2%XP^17I%Dk{{4de!lIfkQ
zl$lB6rEe=xRm%4zH}2K{3nfufeY1AA29z|?w{5v5n5ATmJuNN7bX{BS`ixm!G)F>@
z=2o+yk8@9@;~MUiiJW&AC3U1y5i*E$$Kd{vHt@j!!DvPLnY_}L?8k)$tog^gVXFjt
zN1$q4?q#m0MFgu#2G1b93KY$;D*OSAL?)>p*!+PKo#U)2Jy~UXDcC*>#6{mS$!qx~
z2-&?`u59wy&hAdR$Ik3A7zi&Ccd9`sZ`p$k857BRo2rJ%0tPEcXuP|snPm^&;|5CH
za{y4nO=<@64{c)M9p0q2Ce7DmQQWGS0NX@EuPKzY$R!waNgTR@Qjt@J!ZAu^pg2I8
z>OISRf~y1*7MhE+=}O26sye*nUis<az9M8V;P7c>Hg{|#W)&nI+GU<@I_EPRgccD^
zt`Th>+daPH>3bxgxPv__>jNrHGfZcIY;LisXVG#wWWqS8?V~hBntAsCohRAJwdlF`
zja9t4D$~aBB-diI^K5|BQSkFm4(PZ;oi*M<qc^@JQKPpDwyI<c<bdR<Emm++(n~<C
z%d`k$iw3p_4Rr6pIIbCB%gf8dweM!3f$V=;2H5@+6_JtozpB|DX?8_hvcX>ekCTB~
z9QipMIyC(=&YxAAz_S2lfq2TH<FCjf;`5#7v^uzBREp>r8YyBC#X~a5``Z-(Tpu6K
zZoV(4WjDKitfSrA?-PMJym;QELQzXtSql9o&dwk2*C!3h+DGg&0g9&LkvUg6+lPA@
zDmfnHuL9Y}(iB@*lqNF~xI?zgGmfDfx}AMrd#n7U;f__U<L%zWy-6N9>AfV6{6>QN
z{GX&evd+)v?;NtRh4(Y7&ps%m%%zlLBOJ*1&gpO?!F<`E`wVuXd#-O%bLiipv0s}{
zWf4`c#9T<G5I*M}hn<m*IWyS9>+w9sdGdef_$T4PQ0QE+zy~nXR#NwtR%nE4mX>_}
z+9tX3ct6-3SMPCcWcRtHQYLg9tX@i-o?gUYt!=^PjD5R0f04OiVC%GXsoSul7nTjO
zCvxB@CsoHweL-ZOK;6f&kaDuh6g}ghjA1lV((?Rrshee;FFZ0I?=*zk--2cx5z;C%
zc8C&*mW#T)u@4%jlm4fcw?XEhMqX96G3-tQYAm9SWlY-#quFEmF!J9##n?0OvdNs_
z@!dGn*kZ)8;0~von9W+f`dy@S)5Dz+)S&{|^V~LDgL3i-=7~e4Q>~V<&R<PqPIUnW
z^W_@O9qkPACh*4<D-<0Z<^(NnghT*{2eYs_@$KV^Ci&k9Wd+ysl*)1ejZ5sS_oXri
z8OGT*iTRl;rcOF=``k0ypKui&?BL?$Y-QtP;RjWW1wANW;KrA4$YoK|4DJtHf5-rz
zFZQ?B`;Bx$K_zm$=ph^t+|Q?Ts;18#2oiCbsAfc@P^)tMqX;R%5v3B#z(X|^Y(aL(
zX?A5JqOe5TxHL$DlnWH_OeJZQVsZ@e=d57ATK_Vnu|DBsGM(bmbfQ+Jl9FkcxByHe
zvN1HiE%jV8XN!}$Abjo~(l~Jv2Vu>lfQV@L#q9k=b3a*iE{Bto0DdrxQotk1kUN52
z9%_>jt$7SkW>N%F-?oR7tXVJ@=aKLnHs8(f#IsL`^{3lpMoyktR=of%f_=SYLFYNP
ztf)l&QH)A71C<<!|M1U!MnB31Ty-xp71gA^t)zq*09jK?bb8p$x&zUmNjJjaVK>&c
zCV8`wf)*v{P??5HrVJ1h+(rCuAuLMed5d@x4>{R~F65F{>{2Oz_uQp3<ykwhf99*0
z-mgGjN`c31T%rruYLSj+4c5FGy_w!@bmU*q!#v;tg+iRt!U4;hy~A@+6KJjRJABF_
z=Le7})owTCDLNqSg*wt>ZL6I`b^weUTB?^VrkHgQl>YRAEQDA`OO5!U_~>K@j^VMm
z-H0$R;<osq;R&7Oq4;Fr+qp2Wy*GWF+@g;-p;&==NslRE4h5|LK5UFKp*Wio=$rWl
z>lG55F<Qh6XcHzy5>jsRn(tt7&N4(QekKG{%^yElb-Ggx^8KSWPIxfAw2N?4L5?DC
z;%=ddVB~0ULCy_-z7Uvm?VdeY^UV3k*zPNN!D`jTcZrDt=aa2-f+EeJll1WukJ;J;
zyj+0Cw>G<njHkTrX@wEPGNINJO{hT4ys=N*Dz!w;*em{`3WO_*t3Iv0*(nd!xc;km
zp-N3Z<p**1?!T_FjiJyU7%2OW9oPKXrfL!yL%f=Av#|#Y;$&}9W*0nag*H;caj;gY
z6exb58m$F@Ae=wId(qFD=+Ion$>ICaex~`I{%*zFx*+b2IIx)A;P=wnctPAqO?W`=
zjm3`=n5%PiWDy_y+2QA~v8KhXaEEE7HXDx)c877Uy##oJ7RzF@^0xsX+T`OM6c!Kx
z&23O3ojC?hCZJ41v<uV6^ZWu6G{@~id+83@kCDS27C(QP1d(yKLZUTq@DsI<f8eX5
z`zVU0%{eMQPLLNOHF?Igvyo_&u5SA?d3lf^TefSyi`%G+=^o&aGwmUe7(RLZu2|gG
zt9Gr81|c5QYkM&3dL~PbiA5;LO!!W8t+AQ;DnG!o1+H)gvIe?~&WffU^Sr@8;HR6r
zjCq?rzywr{Fnr5Ss;w^0r*jEk_o8?j9i_lc*w3gNiy?QH%5gPB7ryP29oQ8nFv~iS
zTEqYL#s0x$Q;KqtRV;TwWXm3edvG^Cle=0>a9&JmSuoTzhu?|NnJ|COkG%f@(TYdz
zRu}btO>Az_rf_XAp2$9T-eUNr@OE^4KxK|&<_S!N=~56u69m@yeexqf_uhb4kw$O{
zeuv+K9p-rWA{_pC+@C=B8GRJ!<wN`+aKr0`#L<&0$t1rHd?71W!(lXhD|3ZfY*D;)
zLkN{)gBXBXd`B!oOg^-?pj)AohH|3V6}2Ht-v6x6)az3oQY3NR33fn#mAmHySRgRl
z>zBNf#7Jy`P`UwQU9?JBX&yPRi79G^DnFx%9Ukdkm32$_BP4|nSIUU0XAO>*$&rtu
z|C9-hO!YG~Lk#Hzmvx(x#Hi>2kh@|;uh<DRd$Eq?xs!L3k0gR?x=xb{<EQf~S774p
zn5c*w#9$M8-{~Nky8EL7uVYSBX*H65QO4u)LMz*eS>lPHEx8@k2RwBx@1M83?EII6
zhvd~oM~gNunfip+izS7y+>*_)&z1e7b)yOhQ~<v<JZ-!4HUl)B=_;AWyvod!%!Flf
z{d<w@@2Z~_bWsd_MXl<=3fJ!7s?Kea)&@POIBodCu9NZ6sxC#FHM>x4G<F9WV<fLO
z+L_l?v(t9`PpG>EE7_~VzY+eSUA32xaR_2)PQ-SvR2FT|HoGJ&ca$TlEHv6~uT%M|
z9@)%BWjeK&73WYGGCbd;DvfaJ&*H1qA3%|-mazYXOxT(K&u%g!{eS1N^joNn*<eTf
z-$E^Os0gYqzn=8Q7A$?s<g+z>sEA2C2XUapbFHs$WK2G}WHa@~D44ud@u&e$4`ni8
z=u9B?_6@9UM~*D!AF@usl8_cT;8`dMUxE2tmcZYh^1$u0^XpaCZ%^4d7piD6omOln
z``YBEux|l6m(8va>RArfUzB;a=>BikDd+IXmy#FS3aO@fS7d{6b`_+tGbK|@a6i+O
zu({A?<aS`*h5OIP1|65i;Qq(O#`Q)tsG=l<@s?f}s%ro3-Nkjw&e83ku219Geo(eg
zHkmI@^0FX%|Lh*<pLOrnLeip6c{8w!#Msh`be%%xZtwyPW@z`tSak6SW@5s4b<rUP
zL?tk|DGTnSm$1rB6i5Y#2|+D7vpv+jBKcuYk%m~>nt!<637RJbJ*a&5b_xQ+kOSuu
zm+?08D1Z{w)IeBRIb_d9n(Ommnlk2pHG2etFnYrJcNs_2p?7H=Ig0RwI<HTp64Lq+
zwa(VGP0ZWbEiRn+#3T^~sV&br>XoemhyHqFC>_1z&6%MP@!l9x7oMr_14>CJ7K#2%
zb{S8N*d@3@Q}fXO%5N;`ucpyyrPQa=Np>j+LeklNga?$V?WfRAymc{jQZ}Cxa(g1r
z{0kcj6eF=P0!CY~g17R?q-AK*J3%O$reN3OmZ-wLJGGzXJbG?rNhpKixA2B+Xa><+
z+^PA{1kTDkto38L8AgQ&OdjB5U+@aV0xpxT{K%TR_mMMDlGsH`d?6a2e8?wi!X+)r
zo&qY-Caq~XSx5@RnAFz?AJw0FX`fszc98N<$|X~25L$s}$gEL_l~t&{ca-QyqEb#2
z^9ArAh$QB`w7<TpPjKFJW<Wm*YT79vEZP?lE+H=Y&vFtFpwDA&%6`O1+6_&HRTZy@
zZBP)VQ~Cgl8o)_8qWqx{>c<iQ38i!&m{w`#UY}p=;8hUhX&`0=!aO#b?}%^=6i4S|
zPK$D3NSc%OiQR3Aw1JKz3J#Mrv=0^CeO0U#p|c#1f=X}%RWe_4=g^C3@<>#AHKXxi
z<4weKKH9|oRItf1Kut-eeX=GfS2bE~62}b?-4NZ(!3ao2`GOfTqzA|xulr>TKYU|^
z0V-Y?KF`nyzPOk(sDKCVahwxW0KN!Dw6hbZ?l@_`iB7X`{=-(W83s>Tx;s=(NPf6M
z61D&KvXOTy<kqP?DhMUX6=M5ud@7R<;RnM3@~fL(A|vGa+iC)+^tnV4^y$e8Vr#b_
zLiL{&&RYi0>4;qIDmTVbKJ$XsWalH!X}J#Nhemq&<wbk00C`@ke*g))i%TqIU@KOi
z+?cL?@?~#e!-5CtBh9y#9CUxUb?8O_svNWn3u+b^9W9?D<hY;*D6CRXQicQ4Mq0ns
zS9DXJXI~TdxjL4TrR=L#P4<N4o1kRUtY|$#^CIMx9Ijj>p6}8eM<FbZ<&x*?E2fGI
zos+C>)9q5oWDzI4QMW1lb8H^3){(796&hOR46F~amC-oDpWn78+ncET6cx`!)SP$~
z!x+fCvB2@R;UI{?*}#FJlV<JF_ci)7%bU&7{T0~=7OqQtd41pC4tmN^F?fG|QD>s)
z2o!j>#r5L?x+ZF3aXZ&f@d2o?c6cj}jBybFD6>D>vVY6~Y*{0i%t(^`qZBsD=6#U%
z)PABkj!Y#Mv(Thap}LrR58V~MBa2O(QMuKiZk(hBmhZfW4vOz^#n+UQ`I7Fy^+N53
z{1XQAS3!p}NFUUB;UTcBP=zAyd~_#EuCX(SSO_oEI;7?wl&o!<A)<NrnKAS<=N3xp
z7-pac_&Ebe?!gzI&nO^HjRr|cP<-iJXD5r6If^_p7E2ee@7E^k59!ElwioH+@K3?q
zI$#o984*Q?TV?z>M{3k#cg`7;AdhX!g+%1P66tlsH=E49GYBN8ywEU_8jB^A7oinP
z!1z&pRF4ZO(*Jb80km&b=u8=DBLVy&13bhRuo$2LYbK>qA+4r8Gqn(;Q>vu<Or|Fo
z8czNaGNMwlnl`O7j&lU<lneHuG}6wT0Sl(*_3Nm0PXT?M6qh|~GW!wbIEN~V4SxD_
zc49fZiVdKjtzJfb=qINg$~;Hx3uUt9WJFxZH|!g5aF6LgsqXf7(Y*odtAl?{ZvpC4
zfQh;4M)Wb{I>3MPNS7p8y)70>i6Y4%XD&%Om)8f?7~a&O7Hh+%$gZjk-KhlEF)|3m
z<OIe>E=!03+)#<AzFwN^a`xFee>doQN{j#9*75O9QeE-j)f@)9h-#<U%f{}2c+=_t
zzSBxee?se<D@OGZ_vBZ2b{)4hgNx!SEU`DlHHWJ#GM>Gv-z8UL!LLNoQ<1Bxm~0?e
zPHHH|02g<C>K&F4V=!(%M*8E68ymVQA2)I}g<J#LI(svi!)G!Isoe2~p&NSj0Y)TP
z3CNZ8l<mkP0!azj_k`THImb=mLW1*_OQ+?J#Wja|=QK%^14(hlC7xMS{kooQ`qajF
z1E9V$Q8yi~H+G`%90=pzbG}jc?|72PyQ2UdlaIHqvJTWGQjM#S3F*-@^T8t8|Lk&d
z>MEP32<#qqMwC*{WFCN4vcQPDW-X0*qe=YaI90$yXbT~K_w<y9>5E#JO<tMw=uu~N
z#dv9Oxm`6fq+5W;i4@B5Ytzt7D5-c5<xBb;0|{5(EK)GQr?99@HqUZtp3Tqxvl`ST
z2nKAxc{iBLPwi{RPYskH(3kjDmZJIJ<mZ+F*I^Q4MS&Nr4Mt{3!&6j@l@bP4PYoQ^
z{QjDBU-3Ch6SJY);<+>9{YCqDNrnx#eU!%oVQIEHPz}k_KaR$q91JOxLze=kPw6#|
zltjpWr_LM;{;Y!ww}%oATX6s3<dJjJz|XWj#XzDmPKNaF$ngN!HbKue#fuRt-T|7=
zk+~nhR_9{0{{VB?e|<XsPtqLwf2Rdq`t3(!vBCeEcih6I?%m4t!dr~F_-!M;sy@o2
z|MLusB;|R$aSf><Y3tCDQiUe&)kI9ehV8+aW(we8=;OVHh3kNOkz^X_0wh}-Km&Uf
z+|`9=H-m8m*re2ho4IVUBq$s@@rywfr0~M_d3iN|R)gf82JiwF^V0J@V%`k{@-li)
ze0jw_`}mI7o?92Es&-}8DqLe!N`7%aW>9MFOVeEC_uTWOf5m?8;rnbydQ%VhBc*$0
z`04yo`_YUAXWn@<uKE1hS%9VqsHRh&+q&fM{gdFm=8ajl+yz1_;Txs5le0>H&oO`v
z(obE<QXGf@Vo;OT<WJ?1q_^B6h|ZlU+RBWpO(cvMBBUA0<!8X}IrOgsVe@sMD|F~D
z%e0EgrV8ORSRUJJnZ2vw{PggV{er!G8OKaurUFtasW^W1BCSH<12}Ri3h{h9+7$Sm
zOyLZRzS{R-f-_N0NmtRmwkO*)%tWRn4afOm)WT?7=d?C96Pjq<ncQJMmU(NbB6BvC
za*TA!B1<*`sl@ON?}0^pdCftV{N$?9;;su|+d(eYhGbh=+_=Nlr>CV+kLBVheV&TT
zGO6Rf?(%5%Iu$dbX=%qolRVH0Yt)B7?kz`f{+{e~AwY|5F1j1PB+*b(Mm}?j<a8r{
zZ1rAwz#QnKbZ!#Leb-avPLq0fsalH$lT~~8-ExTY--n1mD54?jzeHkc)cr>OCTt&e
z-Kv^1cNiR4jrG$m@mLh5_wbgR82w_V=1CuCJA-wC)=C!BclFT*$h|W1k=S?>j?FY|
z0{Jg$!RPJTila<>AY=!L0jc3PDFJER?I_I{hHb$S5W(i}@XqX^=HUt6&~1-0`_Ydq
z?*yZ~qsh6{?uAiB%(#I>i8#DEDBPw63pAOIXP{~h!$;wFBbZ=DTua!2UIr5P1m}S?
zS%<cBvXUmmqNu7eEu<lhkyOZ{C=Vz+#|R>Ef@|BP(CDUbsTrN`snl`2i)>69HhQYW
zKGcXb_c7eayqoWDkc9n?Tw7z--0N?ON~^L09hn+#UWS9t-4r90p`-^+x;wP%C+{v0
zHLZ9GG^4~^+4f-<gsM+!q1KrG++v4w!b+qP=kj95RNfsg+mwPb$b$knMo2L2;rAib
zNGo%;QDMW25=$IDaIC}3AEi#^YCqu3Q-zFFj~Vx@NGgStc8b`DPvizTw-INU;xoPD
z7JU8)pvj5NAS2GikSjg8$1a7}T7b_LPAw^u`Q>QV^dSU};DH-6tybF5SJbZE$D?{|
zHN&+AeaifX@iHp;Ywnb^7#VJ5kC*F0$vro>ylbG;_@0ExTb%nZHZ{@(aa!U95^%ee
z^EOy#J|fIQcifRC<aes{qoz{F9vk!{3*(hHC|AR~nZfEjds_pQHhJ8`kBstL`&dHY
zy!Bzg?`9~S`acYqU=E|p!{8d`e*E7aFMBy9bUobP{&$R??<uO_AqYMGyK9c$W$^4M
ze%@0(G@=QoS`5Rmf*MneCwefl<WLvAL>uO3^tdCbdYE651%M0A|MZU=*-^EyWvG9V
zNbsM0_J7A;-OYhX740HM?ezjZVUjvheD>ZNB7A4&zA0s`=XA{<{Vq9_*#dh#sV?9h
zAY{dXyU2~P$pq#c76hk2Vs>s2YN8u~vPx}F$f*|4Y}msjiI0a{=lT=kj=PLOjgY~N
z${JY1Y5?fd>G%>J=Ir6fj$(As<|IbyNR10GBPX5L5Nh65yhvG8u@MJT#%h`t*YuM@
znspH-p9b7%B`Iq|mD@%K@dejzeUdQ8q+E@SjtqELBv@gDP&w)Y2Amg5POEY?VBmgO
z514+U70Uobev6$z4e_vvOowZrsY*f#hk&e-Dl&E&+$>Q(+f$!j8Q2%Y2DBiBQvGRq
z>WI@IPKeVxl{v{*rz(cxx(Y~Ysg#DfDs$Y|a~sHh<xn~Nr<={@nY!Wo^PgjYEc)rN
zD$<HNa?>1&FH~&<X&8e4)JwP+OT>XRCN7ZM&BH?s0nEx^5gT+O4sRIM4rN7(yP^3+
zB{s(}K3OdNAT-7+q<ma0m;f+_K6ov``~-Rk4?CQI=_^mxgA*(ajQ=QD8f=Mc?Txb3
z>ZgF7y{tV;icD!f9y%1V9ulXNqsE71snTqggTNxUl(RJDKfaf%mchevNz`Gtg7LFm
zntI{L?y#Zb&Su4=_Pwzfbz;JAirbgYWGPuI^*b#!i%=w<YhH$LGB?k%Z2@jPJLpIX
zJBU1x#<U}kpWsGT-=HP<<IUU0m693l(3Wff54GX&{w=w;#v`vet6xAT9AMX?SsH>v
zyP*~j%;D^GG}T{y`p2{6zi!U6q@v{_nuF4n!=!&`T9u;5f)=(m`n4#gesKBe(WUJF
zVRyD{X~2v!drV&KWYi+-Dv<$5iz-TM#^e!E!-3XC#k))VlgPbQdk{fG(QSHY%h1eS
z5SZ~^1^kqRD{xkn9hk(fJHWtSr1hj!_UiBex!@LqkzQak#@z#SQL4zYbSYU80J+`d
z0CBWp;!yB4Iro1t_D;c~MN6CIwr$(C&bDpawr$(CZ5wCXwrv}y_g_^J-4S&=;=av@
z^}aIa%pBuOGp0Q)>0J4ChC_4wBwX!jV-lP*uPx(hzgR4ZYMzo8Ffu9IJr|^9w)fu}
zOYn>nmy|Th#I)X&hq5JwwHl(sHB0q$S<1>PRmHXV8(*{k0n{!L2ME2~{E_5xO~0;8
zsoO5?utT)+dEuc4-Uekp`80;?qFbF3^OKw$eN&M8N*CC5!&!zk(>Zl;HODe^C7%Ib
zdrb?%F85{TSw{n}1jgB%-kc-J=_506{F6NS@3gepg{b+}#)O!1{e7KVlt}V|F{6!$
z4Sa_<4SY_U`ZLJs-5`zq<N>$^KY7xNY9<6al#b^&VsiWy+lJ%;Snc%m9$42-?$9d`
zee5779qWhB0IJU%G*pg-=^rr28<N$3>T4MPKUC=dAwL;8nc4r#!ID(m)~_Oaw(br;
zoG8*0XaK0IX&tDCb?twQpGV0;kN@CHbX%p@4;xl>@}~z5F*QoxklrJ3IR7X5V)FR>
z`QaY)dGFUqwsiO50^mt-j!Xip$;<&@fEc=ZJ9L{@tejd)-r1ZS?#%gdb8d^rOW>Xf
z_ykhxJTL@a79fgnrubf~d)7H}3y*)$^k!R_vg78^I@P37P^~*LWmRw|&fL;WSSvC+
z@b?+-$Nl@6C`w%s1iNnp```2X^TTThbcKYtq0!spCn@h71W*rF@-q<#hVK`vchA?z
zcx;q}C%GR6=hquseFDA6tUzLJNz2I86HH=q9k^9DKauuRbzTVzNe-EhMWXg^*4R|>
z*V}<X!mpQ%aZv%LSiLd3Kq^tTCR&qKSPShM5|^_UUm|`T0+OkAzquan4NG!{1INq@
zWe+GXr(?a(m4V3oXo;T*zdf%oQPX@;0@zali?qjn!FAzkwGar^CjuKP78*#F;!GO;
zghRw6SSiPAVs^4pfEJ}vL<nC#uC@95DkGFF)U|kLZkFW_gxMfwXsf1pKQpaNK}4;L
z6gJMUe#3!n!L|x3ycrp58->lVTz)w6BY_S|3T$FK`OesiKf+*S2{V}{dXhOh(kLa>
zgQn>ITnV=XRxNK*kdJ59&WioqSL|vVru%3gY%n9hO_p&&@qIPo;V0g#2B3=LxW!&8
z=+T?a)!R60BE5RjmscQ<jKYADJFiSml+gnTvgnY<Go6n5>>$p!&sQqYkh<oa9f9Pb
zH%S?NyK@j*xh}7`?$h*0h4o?S8zrv;yu2_Ii9c<sZ8rrWkRLSrJw-bV>UW$2$>b}J
z?J^HAUO+YI{~(OV1nVgV2*)Pf6lM4qKIdyT>H&EdXbKSBdg1R6g0;6Z1CQ<29~?;Z
zoNy2M1C~E%X09N?cinQ`$ixj1?<={G0hq%jw844sVn(7q!y*0ipZb4vZ{`|O!3j)6
z{jFP4#8v0FZFge3T3-gGUKcxB8wWp`1;C2dCPd<G`lNy&jL}40Q1wyLOclS$k@BH*
z<0Cm9sm%bACG&-Oy(Ik+Z%>TNWW`-Iq3hbF>Z1w;0!o>eC94GwgF`lZpIfKElD+P)
zw)4_8w@+v%$!#rg4>DwP+DkBplbTSAdJ}3m=%w?(nUrWZ878bUDw2(D9~{<TS@(UQ
zzW>4V70<!8gR>^IaT39E@M}}laAChYhSklesgM*w%Ba&*TPbneQ}3LJ$m6zP&-da{
z<-%QqmhhWnzj<EWITiDcQ1bQ#7$OY3#};;II2`H{$nmXO9fvUO_k3ITT0StVd|h<7
zi)xubUG}QZyhR_h)(v+`#6V0Bq}^VX9qMXWdPFz1JTxf{QY^0v)jG66ImJ0en}Zt8
z-a~~FX<<!#Na5&Rx^Sm1Mu4joOi&ovO3P>lZfj1|ltx@?(UvhDV)MBp?PzFQ))qwN
zurr?&#NZfNTl>6+d{|)8E-P*Ty8NBFGsgwx#e(wAQx)&i|2l3$TgbfjjY`HF)o8Qo
zQ_$187A;OHw|(iVwHXkmUr6Y{Re!ZUV4M}#^aA*#++J>npRxvHX`Zk<9BD&M_SlHA
z(}~i}Uu29{tU`j+gE}5n<x}Ea+f0v!gbqB6ks$DwCHU6CN(#RxyovmD0p%R|UKj6*
zlFBx#BAHyvh}#Y+hF&pQQr_omNxg554tGumw2|coWQZhpg~K?GnE}lDmdyc1m8UqC
zT_P`AYwfQKiz11>5KRc4=3Tg>Q{7*{xy4i6nx@E|<5AWnrFm>}hEGWWn*I}X@Lq_D
zTt^(1Wmd=8V8fve*YO#r`<~CDXA66n%8YPmqv}jO5<Ft8k!Ni@F5!ZJQHmt|@lUZs
zxmRhh_HI|szk+i{RZU6~?;VRe|N0RMy%cHx+E55wE-glL^{=+2xCv%=j|f`30f0k8
zv<7V+;K7u>SaY1Nd3K9TSxpZp%Pl1Gd7qixi6UpgwlY!V9e%0QWY^w5n|f^Qak}V#
zHA9v;n0wXrD>Z`~&1yFD!FI7w?>@}Gpq*`g3cWw~aAvF+DF1^Sw=H-UE~7Eb7PXg`
zj9+-<p2S|ajOV0&9}dbz!3|eR`2$|<p5KyD!ajvMcoa~-oBnpBKlX1ge3z21hl6Ad
zqsgu~?W>KrEM5j7nn6mF>&<}=o+*RkE8vDyvtPcJL{LAl%2TtRj4gAsYr*D0e0DHZ
zSHgoYY<FIwyNKymO>imIXA<JGU8Q}xKUn*eDOp$bSO%7b4^<gH$ol)rEtuPddBa~Y
zx81exIt^qYMsc-61!$@&L$PC7E8{*P5JuWX=6|k%vHG0#yBVw%u_hapGh(O&SBl^_
z-Hr=5i687#Qyg2Ivk0=#V0dDwBJAdZ)D66u_H5c;QnA>cAlvo0iS;Y8?c1E0=K9-O
zn!>pyX+-ZrR)x<X>$Gi#<HE_R)?aK>6*Q!Ieo5#i!Pr}8X~P|(nf)WS=@BH1Sej{E
zV;-Z)_p7&U=@0SSToj4~zExO!AeicEV>*fIZ7^aZchwuRR;|@oqU%b!{{d82L1g@o
z<^4afehk0SjQ<Vmx2-vCd(@8D`$_o(G-Q%6%-3(;zp@5>lp0_sAS>W1=QY;*k5Hsh
zJR!N0w8IO>7~auM$+$uNd6}NJY?qw#`{U{3{%NvnWE|42zvBF`4e~6cxR1bME^{0Z
zqRQao&FEva>3@JfNqET$R-V{iuMd_qyZ;URG4t5JpCFhE1M^V%Rl0eeynR2=_C{GD
z*EH`8uQSc6N^A5yPGwMb&q*h<{p~s9K@iKQ|M&H_nRH|%foT1O#W$j7_v--gz=F4k
zLCP=n`@TmRRS25Xm5I~^WB&ftBDa<G#R*@k{;w&5tltkWEHR(@gg|2QZ-sYO1d}Kn
z9p4~u_D(5@-YsEqW+Sm6%)yorgz9AedAKo1B9dH4(@=mRQC-L?5Sz(JWboAdAzA4{
zH1ruRyBACXNVhCLxl6sGZfjF8O+k{Z5>)GgCIQ8Ec`pTc0|?bOm=Y=3Ep~l6t}zTC
zy*p#F)Go-MsB7FXE5J%iH#4YRe-z+A7JXs3tE3KtQtiKxw)mr!E>JD~!I7&$^imz$
zUTTydWt6XXSP-D$>Y{?PY@)&`jjFI6vf4=nvd3AAE|uE_!CI61s3n0`-fZd_QD@8A
z+gcUtpeo1197Z24VZxn}iy|Jk3Yu0mkKd=Of`vuUbf@(&+r3zXCDCtarZO|uU1fFJ
zp-h@SK;z~w8(N%w%|8G&abs8Wa-dFZ{LnU7>QO<QD5j2^&?L>=<6}TNU45`>i~Fem
zwkLt{l-_Y;-9{7U)|}$xoesrn{(x_poXVTuh{05-6-o7{9Q!(&zu$rVZRV5{Q#xAM
zV4Ervgq9|hK7wrIj{z+=LuNDzY7`i;-m#ov8@2^81Q=lZfy{Vp9M|R?&EQ!3?U40b
z<KHmu<iH}D)FG0cq{lyA%EvMJX2y8-kL8huQ<`S;r<xTUcLQ2qpu<lM7+)4E&8mVQ
zJha(`;t-4mB^8F;Lk6sccW0`eux=)#BO3d~T!&;&(-bND@$H`UiD9=qdF0ZrtGcjI
zr}!D)F(_6qZacOyrO~r-C`Bx_XBafUbXhx*>ASh(624RHTbJDaSp!bBq`2X<N_IAb
zcr#Ia5iaM(3(o|`!loqkX^>p<y9YEV@^WYDYby_+$vYI=vkp!gENINcQBPL__FMIi
zVa@oF$Ei?yH5=d9Bkg2l__oPsd5Bu8#Pqry8MLnfblAv(5|VN?R|={L7%`&-n?>iG
zB;=CzLKPEKY45IdAJd9oXrKK2;1{Hy0GFFE-V=yJQFNf|(I{Zv!fj>btMy%M_VVf!
zuav{0lspK357%T2gEq^{jGZ19y$<!}gEyYs6DEZmb>sH-wfwWy*Ab;OS5;dikTU@R
z&HP)*eQfk-;K326(L{_d4gNy7r;b$t+1rcjx@J40WP81O*S;cu)~P6Yz<HyuWKOw5
z70#?x6zVY$aqLFz4=r(sZOmvRCFh!_hbdNpVa3L)Vlj$^3M5SGG`B;P|M|MfXz;n=
zVsDP#g^sS1lGb)cWa+|~i_CJ5@{3tfcI-1G8kSV9k}4@q^;2vE1GJ9ChOw~2?b>1y
zPdj#+`;hua2GMFabllCmYD}v$rgh2E6L~rm+fPc<Zn_u=%6n(+86rF}#I_EEVK;UU
z3y|tUdDog~o68Z^LcI`7s^b^cKN(|YgFRLnio<CA61U@!G73FUPukatdUhr^edO(q
z*o%clDj5JiRr5zp(|49@>qISE{O)>`qtLb>(a|_dCW;9q+?4O3cO&i1a|2vUD)CB*
zsYuUZ%Bw;Gc5_WXA!6%O89bSlElK3g5%|xjPI%*K8s9?<M*t2-d7Lg(&SIp2<)#b6
z23i;9wWPX+tI{?OoStSegRR7p;K>7Uqu2QZr>0prvx!WhY=^j|^@a86OIM;H#0&P#
z#R;XGP>$^NvZ*WAC{0ii>OqQ{v+t$H#K}g!u!{_vggwWJ1ukh<t4efQJ80hq9;Z{2
z=lVLGO+KrdYgXhO8!i9(a%@j@jSN1BZ;TuHQmKk}pibAe?^66_S;#8q6q5)}X0gbr
zsVRlM__pL_CEdqXYCgs`@Vc%Wk%JyKhpIX-^xgpx8z6R>3tDY!`7Hr{C$Tx<z!mD9
z2j^gjS$2uL?ieUoGr@DS&q4kj#ubG_+M;_w%V*LU1Dk_Lrx2rcftA4q>~qcT4tX5c
z4%g+T)rm$`e-+}Eg66Tz_P$~=2zlDCvD9~QhfAOrVeHr}OW<XzYbUYGf!HR##UH18
zCScXF06F&P0T09_NnU-h$Wy7;bjK2oS3YyleO>A^^5y^(P&<>mfb?rhM<NGu+a^@_
z?j47Hc?3>*oIB=1n{*x8eeQ|mMZx74af#NS194`Qs%$Tk1+GYmNVr=u?9ZNn-LAQj
zw{uF~el5!A?vEA_D9M+G;vf13kX%z<rUf{Z)$7AqubSM4W%EhAioBY(jkPVd(6%rW
zx;@ww+_&L0z@b|5EV3dO*rgYGzzbc=_A4$2%JLrm+mW8S?S&#FID=seHyJ%Y+Jq+i
zEwA<twgyQuo#x$`=GDODO^l`kRaL(HcPKImvih5_1B;{LhZ$P+7f_l_>Y6UEZ{0>T
zE8icJZz$rpR}+Lf$vK~v>*p@A%}sMdR+)#a)>WlH`d+W!(3Auar{C!jJc`r^!ESfU
zn&)n`c#IbNUW?ghosep^qO8k|2F2}l85j<c-!!=vCg9ksX?gQSBwzeBt+ls~dx6eh
z-k<It^k1e%N&l(gVEVroRhgOD|I3C0XFT?*{dz}Ffo~4$8v0k{QPt_?zx6QJMo>n;
zT*hhq^AACi%p^j$>WZD;(Vn(dIP#B(RASmL{lmlM<n`xq8od#{fN)u0ulGM!g**Pi
zvPApEE-86Q_aeGoKkc7;vXU43?Y%}tEMPVNUBuq)=jCz;S`=;txAvuHR&`Lf=s+p>
z-4EzmuOA<u=~>5JpsFYMjusqEtKjOs`eR(eZXqdZpNaOpjt9)s=e%7%uLSWGdOf<@
z`J*6I2FF55BWHl|!8IKH7XNY5w@?R8_AmOr$ENfKGcFq}Pfm=aXBj!3EvMdSi>SUG
z&i(7LU1!!C*aG9+wJGG+JBY<vL?HFOT8AIlyd5s6T|_Sa*0oh4ZuA!e0X$*WQDFUU
z@z_TK%cO!@$5{&E<d$r&zvQPD!3h+TEZ%V4HWba9Mo$7HJcvqQrw|H5WM6g*{Zt=u
z2inBrSoDP%6vCK=x8fGT#ZxlyYhsDW&yK5x@*>d=`m*%BbAmJ%N}?PJ3Rb~!y|HJp
z#1FxIRVE!w!v^kw*8ph~f{bfG9?4p!;QNAT;T9ECp*40Ks(jii11cfC1zY)RR))(v
z(ELN@)RpL0r5^!$NK~3L8HUmdGn9q}ma2$8YZc*W(IPU%4HXfN&ag*RTQ)k#CP<JB
z_8prFH`2)8)np%Q=7oAoe`D=ns5-l0-f=(MtJv3{Cp~mnC4(8YTa@9s@x%l^&MeO+
z-1#3c!l6d6U#7X-i(Hq1mI^`3b<pk@gJ~1IW3}He)n0BNKacs4i3FrVLf@Y2_;eqD
zp;@xZ>B-3Ur8~dgZzt{O{Z926*(BvZp?x_#-XC9~yB>iT1Al`gioxBMLTE_^4EP@f
z#FDXG4vY3FA2h4zDB>_C>JS;X0so>xpLNdm){hFJiFmi%(p;fWX2fDz>v&8TSm6AI
zF~dk2W^rm|I}8NbF92t2;IKOe2!YI!+bu_y)O*-JDU57wTfARHGf+B)OU^IHL_Nkm
zC%%r}3mqH>9mF?hk75&!g|Q5Lf^0KFSdldvSR4u=Bq5JyLjLuTY;GRxgxrQJHs{4J
z6{XIx@ytnYa6r(MNfGBI<*&(s@T@$#un@I`9UzldS>S=67d9|TFP}brM5sZxh&{B=
z@aBPEACeZ9SrYf1nq)HHe_BGYo2^oG2C9Yq&7CLgmky!F7t1OH;A>s<jD8yIuupu(
zIBu-kw%`>dHE8AvFfEGlcYYOKWW-tkY*KO^F()kI8Hm76FULThB>@4-kuqHb=WxR0
zgZg_Tdznj#@malVbkyDZ8)a{N^_d<5mqB{$*^#o5yFpZ)NWMeqR??L(_mthJGi4){
ziYdCn7oTo(f*0R_9lpfJPJO7}3Js8zAsfyxeiGk+z`x+heW4V6Gg`$vGS%fcKu&}c
zs*vyMrs}R|<P~?!AEF<n#cKY9v3<&rG^3KkU4-TQ7D1D+^+mQxcGG=pFnHK2VrlTp
ztX@J|x}5vaGt{pL+r?yzPUKEB`*NKJ-t8&kxXV!6s64-x3ChakdznKMih-WNFQUY1
z@g~?7X*|cPcYZ4qO;bRQG7~Zt;XNpl1ykDZzO9ppQyUCyy+)Lk_J<8CDd~Zo^E27d
zfXr1Cz5`d(p$*9XEzDITyiQKPIKsp_D)P6g0~4r46y%T!p@`cl6Fk$8vqIq9(IHvo
z{y<5rS~MZ)dglIMm`;9XO3cgKUiNlPi64E?pe}^<8cvL2x}h;M?WSTK^vN`LG*&!^
zX*gPljV0%TN}A*&yHc*lYG`6^yQvP}7SbA&LCo6cMQHF6Hxss&9rr-&3hPE}&^QDx
zfsP#N5>__4Hb7i{KpN_je=k~4*!|GR%n6EEPd*7@I820~%MoCn=D;9v;a0y*!b9y%
zf7#ls{Jy|(bvTi>1vZ^?^<HI_MzS{zO#*ZCKCwmy@zp`cTe=W(Z=Y-6NE(s!k5S;j
zU#EWlcY6^AT6VoHN9W5xa=OgR)ju6|CZPQhVbPr|Skf=*?fFa`Nx&oIxL5MzdT(Tp
z!!!73i9a45f5o=wK3sTvr+jWSHiLOmGcg*k`kEt4;l0)kMH+LdV;i<W0cKNlH=uba
z8Edvsp5uPcLu~ae{gIk{$0*zGrr#<pIz`Dz9x^lme)=|Q2{)|Cmo=%6*CHpX9L7OM
zJTb;W;?i6=0Lgg=hK$S(r+_Eh#_<=ql1}}7twQifS5Bee!b^>l&!Ij{C0r6e`vjO3
z!ke?bL@3`hJx4xqY5Scq2VTocguzH=OUWEV`(saXBt9bt%075wNzVgPsfadN+;@q8
zk$>_Vbqda@js8j#@R+ol^bb@EwQ2s)M#a6>GL90t^!@}@71lszi=j~0xhA8wSBn!W
zMl84*0-E{ju?pi~ePAlwDUF22uvP_o{hByo$xt_9#h3p%jPxKAkroN#(b?OF?{~LP
z9<?F5eXfWd-q1y?w*R5NkDxWDEo?Mf7lgN&MEh6~vo^rjZ4BvsM%2>Z`ZIvWL0a6w
zQNaK&7uI$`p-Pz^_3w!bY7Tu-PeE+l$}%Z}`ErO6%4SvZ>7OC(F#jOMb3~N7m%oWm
zx9!3A>Y{ETcz1I$3VQBK%(O8|5VuL}M_ySC1ltIt$+S=iB|bnm6S|5`Vb`M(_>fn%
zhEr5)hN#%}!3OdTGG?=4Ti)SLdJbpsCHAKM;ddk0&m$$D%(6LKX~_notPcmMUX9I)
zL!4FYnTK9vQRL%~a5mbG!`CwVk&SbTR`PkMZHgCj>-^`LDY;-$N9MV1x}xVNS9ki1
zxt42B{Bt?4TJ2%X;Rs7prNbGI5<B3#JHMN6;BF9Q$;*CS>oT5I!*a&8PX_o9U6PB|
zb-{K0veBg}MlW#H#d!{1*qej92WMFO9<}{9%xa)(!`Go}5n&%QldDEL&auAP%L%Mw
zh8}Tv%cjcr7`A1D9aiN5G9IrWxI10OL@m3~`o|^LF(1q}xj||FHq70jX{`+c+gab*
z6s9WFxQI1=Px&R>ZadezK0=#JPq#<MDs3VP#|zgMaNT9Y!>DLC%k|$78Ex4IkT#K9
z={Iq-%24%;EbQ^`bS(y54wH?F1tTVc#jJ<mD^oJznQO?cjbFebDP}+GQzawQ?(b!l
zii>L^&jQabqsu6T?(*dPVyiFvmd57ur2SIljy(te_IzjkCc0@Epe>`|(+98mwN8ak
zy>7A-EODg<{Pd>1u+U7%E8j>?fjD^9IiVQu1)f^Xj;PDnLCZ)V>++<$$Cc=)Lhb_I
zc>4rmRli&YIHn;Xpfn{Nb1}kxSfj<g&~&g%HcA^vhgX%TsUlphTUX)L)ruY}F$lt@
z&G+)ET9R~|gj>5%V3FVU$f58dL3K$uH=?X<g<2a0Up-(4zhi-*{PPxgT*+Ny!>;rw
z=Rdlhf~vbUXCG<1Y?sXk7&H6L*G{JKMhJfieW@&EkbmeN^pRRF8~LGLCPOY*E(RKo
z%P<rsmK?$+nAQ?*3(M=rGRIWz1~S&@KFyhPwkEf-2j)=RqG2m+-o#c<y9}_!Xc|q<
z*2bH-7!2HmX_}+*4$WV4P1uhvTVBG#A#2eQ1G^m60;9Oc^rA@vQ6n-7tAA#%>UD)&
z{|>((%S!qyq++5Ek=Pw=pDvGU85CJq8KAJo;QM&~m{mziSNYla=%2B7t68aeD`Q+!
z;$KcpGLt2t;2(3lg?5O0@P*07Z<MoiQw}_nMuB(mQv?{t86~;<$3Iq9EV_V+EhAOa
zpddq~hM*itjT&?%W)QS}=nKi0l9$j>6HG-P3NL8EA)4|douFv*txOx*;K+IwXlDnl
zosbRzHD%J+mIO^4FQ+&BK{u}MrH>J@pf5hiv1;8!>mFl1AzXhZ^av-6mtrnIzM@1A
zdrTHgD_y^@kFBO4MRwS-)67vNs-n&9H09x14iXi8#*s4`q!BGUfk6F2$Fhs^#u!1)
z5@#5!E5arc(NdLK=C|CDA3pe#x$~)Z&Ve*ah;Hr<hWa9`7qeuS=R3?`So!A?;qhnp
z<d_JzGcamZvLSd+YNW{`BMG-xv0MOF>FA@MjNd`=&Qx_E1yHDZ!`QH>tQJ#F`Q*E4
zTzwR|jBO%!IzZN%$lvXhJ#3*ug*o|}+lJ<NmB2H%p3#Ezz_ycIR;w?G!zN93ZazxP
zk?^S`Gp(%GSQt2#lKCR(=BS{`GmFpL97JT;{D3Vm0vJN9wvT6u+!l|;qy$;lYUKSM
z23=FTQ~P@JZh~{2k&OYw1JZmnG?SDlE6ap~A~A(wk!C!jX}4TN-st&*H`DAoT5K$Z
zf~URQAvY4Nid>}KB?V4Z*aFk>y?1^*3Pg0*4Qs}+64KL2Q7Hk~3q8n8M1iVK)W-EW
zxIB{yzIKcr6nqf6S#h&QS!230hRd@OSX!=a!wCsphOLIpDGZT{DZJhmVaZVec`ZT<
z6uM6S@b&a{(hraX{MYAbVr%?=q-?(@J3~t-Ru1O>u!+yk@!xIYFKdn4692{(Tv9)P
zPaPm_0SI7dq(Y@@p?YIa6)2oN;8cPwZioGRL(&b<GPkHZ8jbL4q9sKW1U2*!OqaDq
z$mOxm)@8|&gC?2CLQ|Km-X`^<<M=_x#)!(PLYO9}>-x$5TXUp?<(+*4IF;Nty0;6<
z^L#B19>N3K4F~2md^Hx(k3xDId!m$k!x`M|D!ZeLj%hw{$M$?&(lVWRNcO^G3gLMH
zvGv{yxdojOKilTh``5egp^@!@`~!o3*mih%(l*^LZiI5UBD3pjI|wZwNJ&m^kgjI>
z@x6Nfc`_;9f?9w~A!r-Dk^=!M)&ZpWDjlLG5xnORLLvB_O*ReCm8SO_K_xV~I_uJ_
z$Vj{;uFxH9h7fSikqwTeNM5M}t@!Qy;O(#@$wW9kCbHiR2nt7GACpz&kmOANrXpcJ
z8Ldhz;@hqRG{RtTBrfq_i&xSHYFdrqby*Lj9dXi(+}=rC3!?(_Htm$S#W-p90Mx)i
zPH8-s-@qI}bQ5=CQAV^`dpy-DL5Z=4M_;}Pt3{e(w09*o1wWUasv7Hr9Q2@z_H*&o
zVx~G<Ot5DP7g7i~5|W9%{qTY{dY4i2kQ^;m1RmtWMBrX~F$y2UN%Ni|<-0+_?^}8Y
z{gfj=`<o!>GQ#d9R_k!MSJX|s5!4*#9E&SpoRnt*UjRcV3KO(*m=2Ks2%N*(R*b6T
zn|&VClW3nBkU8!q0)to7jcxV>QFNLOX@~Cmj+RG4lO*WZe^qR%I%E8#A1930vO4((
z2T43MV_DpO7(Ze4I_+)#MqF!)No!G5y!)WkOgF(#0-Nwj{(b%rWm|=tIXosfw^ATR
z+vr?-as&z$bKB~Rl(tlq#8x2<fl?#qXhTThtSY7D39=air*jj#NhpJtm=%0pnDOC#
zm({9c?&>ElhoTwUtA+g@lDt+8*X=+PV}gjvV_|CD`efZv;`*djO@=)yPH%CF=sJF4
z^pUw^+I1(F-jr<$UI)o}q5!x|ADz}T=Lx^e38$EUq2{iWDhE0#F{#qlE<wh!M-qx>
zWU&>0r;n%G3XgnmybFBKvNhiLE{Gm+EPN6<N1TJ)PO-%Zk%#oX@O*Ra@uf)yytc`m
zA@62yZMHzttxiI=k?u^nNL@DxPw+@6xP7)`$Yz36C=76)m!VQ@Iq`7X9}3SC#H=`!
zG*Svi*wi(vC7_+Z%pir&YDe2G&7N{xA)i2Jl+bqG?%UvF8rUMrHWqrRjLf{FU$!cI
zf)XTg!@eBebVFJz0l6@=h|wmR{P!@NLolS+d#_5Xq32*G$S&U)7x@`4rlQ;2p<EX+
z3a+BtyrG`{WGl1fU7#26cmCQ8R)2wsC&cgJ;!_J|`S0gZX^7KLU5^~)K&gu=0uH?V
z3&mJqEsL9xq)5TRG^2a)bp|F7Otwx_pj1cjjc&pB7-<J*=*l-~%^B77dG$S4^znGm
z7lZDJLTfuvcaf8N3|D3;YXL-WF(SQN$FvX*uMyE?k~nn_Ti_9uPqxVKsJnHL8DWRl
zr6_3U7e2Qvk#nVMC^S6f7`f|W9%waJfqj|<NnPN64rpBVu$q)J6d1-)RecWAOW>Aj
z#}p&drWlW&Y~@(*bSJ@@%CO8lEn6;e6`~`j4ryjqu}V^480W3o+C~x1?MIibbqL8*
zsJolsF`ntf9?Y4UE_W;95_>{+&5k!jrwiP-ltsiJ$44lcRn4B6iO1|g&fE$&H{uni
zP;P)>G%&Ah|5*y%bR-grz1=O?>7i8Wg=Nb!22`B$-#0RXbI4(<mhdLM{eh*)>wszM
zG}ou(jkD@dvOruo_>j72X&*!rpAO&*$d^^b_1Ew=Qrm6ETcFGHO5>L38YgsZ6s1kE
zEMdO>FsxW%6&<OXSC~_gkE{VNc%hH+K!ebJ_JbHP7)hHLT?hs?>QMCbQf%Oc@d)J?
zSyE=9GdY<f+Az+Tm;@QLDJZtW<Uq@c!`!=6*H-18=(Qiil3x)?y=!dVY4-Cd9v{B=
zi#oQS`l?(nUE)yGsHDd=9n{%7BT~AFRFozCfT#>N<y{F<ex-Z%-tEVhHg$;o0FvyV
zf~0n8wy{0d;n3zQ*<46A<*I8H>>))i($I`28k$8KULK_Tmb0+wDwf59km>kCnX-K-
zE6l|lQz-@%mAQ~|Ue|yyuQ}!<p`8?5SD(l|TfW;nb*uVK!7>vv$Is}sba-hk6g(6?
z5ZV;9OZG&WS9Yd8w@MJbx|pYAZ7OI9v9|?bAalnL<rnC;?fo&k`P%362}->A^dfdP
zLCzmf-ks%JT8Gv5Cq`mC?rU2=8hF19;2<SHqUco-?ay5Q!kx{Ae=`PWi<PEduxgwK
zk$$TR_^1&)HWyM*X~YQvMXKt-YmrT{h6<X4j0x7hR(|4WXdOqcXh_|@MW2!y&U2l`
zqtAx-xZ<+t#@z6|@EZ1)Mq;Ht0S#GjjB$m8!9++#9$|%hR95f8Mr|95q-i!JC^Tkm
z2LyzM<caEvH0^X}1*8QcsAb`7gO3Q@kY+=PY7#^BK1U3tw6a1%NfwFrzn>T8IIkgv
zysTN_Q_3w|0GrY>c|Ora$mW9zw9uInhLkQPJ1C{(KGCw2ZLI_0fp*sV0Eou_73O5+
zRT=f^*uozi9<nwGgvnekih$$9R1rm>dcayEAl@9Cr3eA}$~A@oG+|+1ybA8O#zvQ5
z-`ADO?o1xWdF8Lo3#pA9F9-sazWoYx@ChB{kI5<7Qv2WtV3`z!@G&t)1~3Y68p?c2
zl~Fb1ouQC`e=XC5ty4~3GET*|#Bxed3exyf?bps2ub7ro*xXfYS~b8r#qv^mRM_F>
z@AnBo>{Z@i+_$WldeydJ*tqK_=hKE`nrE5abD&yVUm#zp_Glz;)T$5hkE2$QLIKOi
zj;6vX!(nbAtJXO#7Wi-UYkJUi6jw5iI9n;&oz~XQqU8Yg(Dqn~wU4Go*7Ow9-#IDW
ztTGW#w+_O1n_RZEFIEq%F{{UB;+GXut#dIaDeX;ax_rEP_E&Rgi^>=2mNulK;>^P+
z*^x<e2h}EqT+VZ_4Hfta*6bfm`P2*xoGi_*xy93MWtf%EV=FO%+K^HkdQ4WQcK3v_
zR@pmw(gkfSbBpoy=L+}{l3~o+Qjz9Us_nnlKW6zLxI3D$n`~$6EY!R=#p`SWU^FSW
z=LhPzd=tI8F9UeFF90WR6>a0+4$Ts6J(b+O{uwX~(NU+2Ne~`9rH=&<VfrBRV=3km
z%=E&bQgC`+RnYki5F^o*>Jhze@)1~nij_rGm4*Lil}`QmL6AoTvI`NW?83z394ca9
z04x|@{-fq+y0^bZ1d!0@MgHMD;_?^nso3k2yAO~2BxS5id#nYJo<>0fRMNPL@rS$t
zl2$ePIH~<B1-Dy))Jc+K2iA6I>=ne~<Kz=d)f)5ZX1IyNj??6?<GOT@VZaMZpN{I7
zF3J^4nQ}QiG<Q>;OcXrs+i~Ark@ke?KenOuro0a4^5}d`{G4@4VdI-<sOv5mOXQS8
zACT>Q=fnYB;qT>aTeLM~C<4JEc?;0`d;~cmn_L`eUz%j|V4^zF)7$f&RbIhsp(QA+
z5tf<VrmGQ)Io)dteqABW!!hY!+81ekJ9{E;Fv#cigzW8n5m(QSv-u=V2V8IIQS}Bb
z`lS*p{c!3BNTBTi{hwN87RLXJ=J>TL{dYQz7LD=PUz+2$Rek~%Qga~Jhlpn(2JRS6
z*NRs{_&gH&`WJ?hC6y#1QB&e_W)UUROSSSLKJ;>m*1F3UHane`w4g=aEkNFu$JZ*3
z@Bc$Im(QvdMo~R@eOh;|rw3SEa>^eECLf6>_V4A1t&2S**Ob6#fS9M<%L&Uun2?95
z=QtwQ=<D@{);G$Ey!wA|Rl@01OpV-!DI?LfMQK<zVT~tT8*f)Xzn_MHZrWb(%Vq+Z
zNxZ_D%$NM{8=1~<Bk@wgjL`b%A_aJsN*lZoQ?QjflFjt(V&<NbuS~P_z{)tphADQD
z${D7_BO~48#|f5Ia-Bw)ssmjKU|dG-%AsP3k6N>8j>?HjqKbK{bDd^MD+a$jx1L#u
zH3Hls)=JmUXr}9Oq6FCbwc(je^UZ6w1+)5yI^uLUd^f8~(2eC$Ol(<)9PAqGB7<-s
zYEXBLLV|TdgW|Ep+7FJ}zI04UNPLvhpzlE{YKWY=$kg!#m)0yBnmsAi`!j<gmz`#6
zX|zoF&W`y9bhXfKpr~0U&Df7>lgQj=rMOd9$wbhP0p7&BXUhGQF<~qrw`JSNYT9l8
z&~Y(>z!(X9z1_9CxY=nxBLsQQDEoh38-L+73*>_<xg}OMcmLV?S^pfB+cadi!mMB|
za-th(+NmIHtOz2@Kp7i5jZIi)>i+xs=`bPQ%~~K$sXyZ+>K!ekb|tXRKSp3ImMGtb
zDDdS3<)~^)1S&aJMAKFw(Q6JBLF{)>=Gd_V6(i=j0G|SSCGbQXqu8oFaS*-@Rt>t{
z|Mr%G5(TMwyTT$Y#)1GopC$R-d^nQ48ywKb&S$a_Xtkk3_l90WxjobKp>o<}ix+z-
zahU6l<J^Dql{}J?`*>$1?i2G?IeJx8w&i5J80TUa>4F1#)(<c%>=Dx>-pHU;3wLeZ
z3^Bho<3qldz7cGm?K3WAQ<9!sx=J&|%3<J&X8g`}^IBr<-XyNw-^Ozo@A|$n>z+K?
zIn~MON|g~lCCkgRdzK0#IwZDXSPtu~$q!K{J;*#)<t0eoFr*<@eH6&A4d8xNr__|1
zJ(6A}%jn@Yt@}mfjR#J~7UsK7(%?{?=Ij@W3cf?ANfb3CY*N%;7AM6=yDae8Ca)OY
z5YJURFQnT3Bm+xx)gC=}E`L)MEDu1z|Kh{M&7xpjyfEf*NSdfX&m^;uHq~qs2QH#n
ze~y#m<CU7R-*Yu?L+~;$iy4!9aL;=V(!DgPfM@QRMZ&yejgBVPq4U3wP6a!J40_C>
z9Ldwyp*uowY*~65bhFaIKR|IYP}jo<8+;kFq?MFm4R=qOy*qZp$+E`u?&p2!(4D-j
zG`Da~w22hv1)_$TbfXM90>-gEIRihK@R7UUO*Re`A*-c_sb`f|k=T5|r!|KhUA3iS
zkid?F?Tpi~EB1C$!N)6LiaAaopQDke-Vi;+=e3Al9ysTqnu8GK`3BsP!VbdG-y#+V
zxqipr7+*f3X5`|`+aF88rRG0lF`zecH-?#b&x(#|F$}nHv*+>`8gZ5PRO^RLY7ird
z)a4@ixf=0NQCr&@oCHC?iv+4x6Fr60#Ro4l#@Y3vu;J4=+bE(o-`;~v9RnBnAo6Q<
z<{5!k^89!7XWKX#^LJ+0?P@?edx|sKNO$ID)X821c8X;7>#OTRoCozP_SW>;CqY6)
z{qGYehlD&O18Wr>nPn=s3dN9v83Ty$TSek#5vrWA#``pVg=Z+`gHT_QggU2c+E<{y
zs2*_HRDB+Zn~Iv^tm%fPT8;^7DzNwm%HRJ+nx(A{JSt;xyl^e`+tcjnaH-=*PdU)R
zHR7CL$LRO9r6-`Wqlk4+wY}5e9q4Mh;}vVCIxsGPP7k=MhyM4WY4uP$b7!Fv-JT8#
zvw**Dtb^kqiW429pjo4NYTXO<<>zhd3tn2T+xS0J0d~g!&vM-VZz2*0!~dIz#Q5Lx
zep+@~9f-cKb$f@c{2pT&`}XQIK*sFTu2FOde<|ly?t)h5$#%F@2_|JeZa-m<iM0}}
zCtGm{01an3j|jt@-cJP6GM&T=mZS-tTm^wGAXS|5P2Y5YnhxBcwWrDL5gz2qLlsd#
z%JL<7f3J);1{9#wycv$>(_s*oW>m5c0Q2CWUTnK4z#TNAT4S1)<*19z&uUnX#~0Wd
zV=Gu}@kkV;_@Il)OXvnr$ypr30qGy;NaAQkZ-vgP_E--+4U$mCpkXzr!JKjjqouu|
zsaDJIR5%>RKNg@WUyF&rrzb`3#+G#80pT3)t)P10@93nUbUNCok0??BjtY^t0X^%a
zQEn*hCA^gaNS}Cnl8GOl2;E0Zrve`xZvj)!`n*shAz6M^NWz^E`E2vG(8HcBQ2<gA
zw4x46!oIq@+-}G33Kh`ZhK)}OQnd%Y*l^`PMOb*WpAj%gS!23>%O4H{aiw%x+?Ino
z!$=f^@bXy6N%Sz2iZ@&~wB>KHsaw(7WvL|Q$0y2wDi-j{aF!ZhxV`4_)a_#hWalUW
zj*l!f@J2Vc`XWbQmJGb++noOQE4Gfk%GYFO>oq<6$|o;cpA=~QklyU-3~;-r1JIdO
ztKj{_e@XzYTq^@FPW_MWRQyKsBmJT%Lwy>+t=Ed5Rj{&yv-vpUEiRSeOlu{BdRmt}
z0(x4vJOcVPZ<s=A*c~OXVt3WOQNT8XO>RGL!`T0Qa$x4><;a{4{S4PR`F(tjulo(&
z4LQHKGW`+1dw1sea(VN3yC<N>)0K-o1y@#74hN(|#SAQ6DI8d!NLzyoI|Yooh3Avm
zl@3GS|KoB`QLU4X^dc+m=N{@8X9y@31NksfBAn(?i?&_dC}~|&%b1{n3=_WJUuY_?
zS|vrNivy4XQq_m|^IqpS#1tsBus(c#zCQxVxqP;vnnY{rVA0SIeUqDk4F?FTHw`ls
zx43w2{qJ!8o}}InAyiH5@!i-J&HVl?ZT2zm<y7oZVk`=szRb+&>K^HAePGGW1G0Q<
zeBl5vHytu}5g$OVqnP*MYW>cVjb|rs3O-bgUKh@eEblaI73KNjY#Rh7XPVsX(P>os
zQmQ*QTaMmzH0co(=CewV%uX)WH{JIJO8K=*H0;eclpYYV^R1rx#~Ts*<HZY{hx?7u
z)4iSKL3~d9PJB%K$qA2lrn*FQt>RFJrr(8~(v2ymV=PldA_(?4TVMT(F+A$(_xmwK
zl5Qp|3QEA>rD=}R3_tr?bG#}koe8+MaRd*5WfbP!xwf&EIEa7?Fgl)#3_{WX2jkL%
zJ6Sj_5gTm=KS~r>c~?XPc>ty*WEPzVKrqmm?a97svIhe)#7ctqsuZ2%LHO1YnhSS?
zf_7%9rgeBCrnCnyIG*I93tqj4kgXtSXi?>u<^t4?V3XQ!ZXVw@NGEj@8UdtOROpr+
zJ3+_qGkN|UbiD9jnn{3=JAz;S@DOWq552c8oi8qLhb>^9iZLp0iS___oc=hm`U6BN
zcOYq6!rW*P*(nasuMl|^Z0$&s$itNC#Nd`XKm^#BxKjL3J-%`P-ss^9sy1=B7Y;d_
z7K(p!k+@hL;!A&on#4?$ol$^78R5=}k>sa3+~?0;VLL@_y1#Mz&KU4QYqWJ1(p2#G
zD_vna1RJ#!nl@x$=DsN>cK{QBKO!ksH#t=$rl_@8YRncLP|m740jNA%@i;{m$hm}U
zmHDu%n5(<ZAW}I@>uPI;c<zf0Bch49wAM68sR+?~C&4=e1x+(p2q_N;RFt+rK{M=G
z>$;<TX?&Wkv^cOrRu{3yayAJR+(czOCGJg)`oW!d$JhOru_sv3$LM9=_dmDW)5o||
zKfF*k^2b;7zLCT8uzfr9GIFs#oatToPki!Ga<MXc`k4>2)B9H?^{t&0RjFQLe%=nz
z+p52sE6M8bgp$?Qk`X4+Dk-iB`&p+CJ5%919DsX--0k2%B<w_si^huK$ht$TBCavo
zyHpcyJ#N1y3dLJ0?aR`^mciyrG-kxYLO18!{wvjsMV*vlGAlOoCj@Mku0{b|WOCdX
z1s9SF@ip=2XDi&xRjyKFw%&$&r^!DImX(O1xXnjvOoBeo!7!4l#rPII_BJQq^u$L^
zR6BsAI|4v#OanD(;L7q+USV+3Da{4=!1wlwG9!wbR5v<*js}S>11X>eTBzD95$An+
zLd%j-pCP|uwI)(E*9nmUGL9e_55fmp<flZ#&4Pm16ABZVTImBfTCNmP!gVVEFF+J(
z5ZkHj&2>+S3Y^Il4S-JsSyN5iw3AQxFGaj#Z%!hU>utw|KxP6P6<|oY&HC|LaG4A!
zP1|=UB7ycqf;Ei-trnE`$Y5_`g3~YPNCQ4H0Di^<6)M~pyG1516DcwjZnM~rp{Xp|
z$Kf_G!?_R>LYax;+*0Hhtokze&&DqcPQ4eLlK4d3$LI#}L=^h4A~bqDqCs#~0KJ+d
z^4#sHPF~Quhfj0VUeF@VmVV8q%{Q=gP6a0-%-Vqa1J<M$SayrO>KQE%<%!tQ_E{Om
zDB~MS!8=)n*G7)+QeP%#&zI?+VY+<)nxQ*vq06qAsx|}X`@~U0p-lCl06I_9ltedF
zG<Zx^TeCQL%L<Q^BQzU?TwZ>&b!<`*G8<|KgHaPSF5{qHml@BdbpDe7%mVH5e{Jvd
zJ-1oE$Zl7HpFV<3GA%0D-TD$dm+jbAt+!j-o~;_qwf4p?swcHb%OO+F+lU)lx`_H(
zy#Tmqu<CGezY1Lep)^Dd!C(0m6{VMzNb^CUA_GBY{<-s45ISFea~~jX3k#CPu;4+d
zy^*WM1O^KrcArzT&UYKB6>y^T&DWLlxFREg-Q#ti^$eBnw(iEldi3nmhk^*SW`s)*
zYWeO)O3i8Z{SC(lA|l!Jhp@Q-1~|mFV`LrzLP@=gc>;`H{Qww&KL^Ipdj|ZnV??7E
zvLJZ~(v0@?Nd3Z+_>NUdfOZPz{}7=*{&+(>j7r+|4Qe*ai#4^&(LJZziCVOk%T40{
zXRmRZ{+Xu-i1-9JJtvpPJG`4S+%~h~{Bi@$7WS`DTq+?n>T%qfgXi5I2>1ea;q?(7
zo<~@yg0O(-6deWcK*t$UyiO)}51?uM{OHX4ft+ttr_A2Ys23R7L&#<hOL4Swcq)1k
zD7G(yO7Nlcur<uTuy~XI{@)jY$i{!@>^NB2|F8Q!6B8%He}z3PV@=xPj9>HW&GY4N
zc*Z{i1Pc^5kPxtA_kwcLM7KEA2VLD5`0-V<SCxt`Hdb#7^H(ifqm2`3c()wAm?Uj3
z3Z#C1j}FPjow1vMXr5lZ+&`9{#T4&Ka_!UFs#+S5^nA2_AF4{)lmvTq@=LP(vW=UI
zxN+=Y+y)fIlQ{q0?{7oeQ8$3wQ5}TI`vpjCQ4klL=W+4Qxl;Ok4A#;SPJ<yszWsgF
z8b}~3@u5d_N&%YzNu)g_iFm!arvwAQ3|cmMeg1rPCe-G(7S{)%khyRDtP!Q7AvU(e
z-hd@TZQ2DtEc*3&|7@1!rR0P5jiNdN7bt-C`Ao*mEH*uwerD$Npr~pdvyld}pt03~
zPVhsx7yx^3z~$%2gYg9U#>^4)2d!8)C=9xzBT@&!SuzPn!fL+z9QRYHm?gX{{SBA<
zAEynzg^vy@{^4*QHC4^Ws_R=f?hd;Tm=K?HTkMxjfh;LMXHmN(^;jZmb(x6>?m$A5
zw*Ny6+pjKDi(uG-K3n-o?Qn%ni_m%qRk)cd1gi2C(^mj+QMXC5QLnI{UkgSsAWpX;
zQnShOb2qOy#YNA28M(`hi(L{M(7X*BLX;ycK@}G8=%YLv56(iMZtXdK2N!l{1z00i
zo6qxJUS)=L=1zzWkrZ*qPWo^6YF6+1YZq#<IJoGoygaH^0fvwmzK~LMfEnsxG5<E|
zp`>~AL_JDSb=R~2({K)kpLfVgdq}G=$<!qPb$3mj^5=$$u+^KFVR_7aL+?3Z_D#X!
z6%H}j#Yge(e9|oAT6)po8LZ^+DqC-?u?}+U3DoA=Ac3VZ&H6|~Ps~0~Gx=D%#98=f
ze@ZgPUztt;i88@YWj4rT-wP@ttaT&u#X|J6N|P*D0z24dnp9GU7OFu}Rc=zjQDe5L
zc+-D#k7|Vmkz>6ZWVQ0rxF$R-?hUj%4r8qF6w0-p{A|_CnV)TXkGe|SJHk`zbr`KR
zOZ_Q2nkQ(!#j7rw)`&6Wl!c!7N=nj<JNXgoMu6XU&bH?DgO%ti1OO+VpovGB2QUWE
z${df<Sa1f20!(F6r?>#Fxd+nVx02BjPk7kP?hMK?LM51t&J8*X5}RIw7RvbtZ%CAl
z+b)`yQiNyAREdfLq$E9=EJ$)8f#sMB<30uTvBO6&K-GLj{jSST;G{brl{(Y2DjVy=
zYpDrg#PMvoY~E>uhfJ}_!HUBh$k{2o9xO^<N|YPxc)-~3tL6rfC-37g_MjZ&dkeN5
zkF}0&vx!w(+mH4$+y8b%n-xuW>$yHVjJ2v9CBWdO`;!W6o0pYZ=h-^kV983gJC$i`
zYYc93R*#gY2eDPt>a38KkpLQNiEl@o!cv%VJ@f{POfV@+f{hxfci1U4LaCYT>Cjip
zM9XUDXaw%J-fUB_H*HH<(!5U^z@y8foCQf}t_{`mZ@tcyvGCn|BO|Y60QRN%M4S#_
zN$WcfPB}>!ahvNUFcPcH<yEM$RL55e!S7nbplUarc}do3w%E!Pgm}VSIPZeIs01BJ
z!Ck|dhcr!CUnI8UV9>id6*g3+Gd$G7po@}OduK!Gsx%QW&&Hsk>>_oguy!}J(CRM)
z{#ICm;%s}J11zm^!EXV&%WK~#z_LyyR)%2M&0M(6K8is%)CaCi*p9J(Unk3H)c2Yx
z?%dnobdVpFF@LwRcP);u6ZT%Fj)k>LtQtSh9u28@ATI_LHtR>Du0@`DMOax2k(J2v
z8A=^)a(Y`>xK)hF^UL%$Fq7Ljc9h|3XnT#$eA&x~@n}z|Xew?>PCxyvICh+b(xBLm
zH0+s)+h`c!5LT$d?Xxdib=|VArTC+?uxs%M_$gz<>v+J1@A#{~g_f=P*f4W534Y~p
z`h>oUbr9{?cy_yTt-E%qAC%}@d?9|(+VDMKS=%}7BnTVkjg&}y`um2oDYx~*v`O#!
zjoYOYJ=_8%DE0lNt_@X!D2Q`Fb!?4&af2O8%;QzetrvQTv5lc@4t&Is!q|XN4vnpk
z5zBtYfO8wQ>E6YtpHDUPZ_ZvDs#uiLBVZ~xs3Z@4W-JDI05rXH-|Fn&c{GkZ#i{RD
zw6>Cix%L_-@$Y|S2^REiCxgSN<D%;GM3(_sO&MUX%Jcw5AXuo)iL5qYuzNwx>VkLZ
zPY@BlJV|-xsWNv!PAO3-yeoCj#&L0ka%NzIhHZgvFH~y+^wCLIca3uvi%?y%1Gg0c
z^NfZ2u9(=Y>|dN=;IJmVFvI6w-C4PjVlw@7%9h-}vOGFAhmX~5hCIn}Jic>F3z;=m
zoK0#{-j6I_EsL8MS<9Pr8Vay_q8n{#E(HhW)Xrwjb#cRM0vfTy8w%dOG$Rkzu-wVI
z&8d^Kj}zeg&Qh=MhhDkaR2)bulu?EYsXE^*xxIt^H)KpkH<!4zG7rfmUiEewfR-o=
zQ_(y5`%1hfl^II6ySUgS4FNRXV&Hx%2RB8^c)|mR*A|{cec#X3&Ctmj))RjabD?o?
z`*KGEzuWX$XTx5Z=QoHUG8V+YLYTBH{k<vvnIMsz%V|oH`(Mf>Fd~*4i&b`Rkbq4r
z6<Cs33eFz8GktT4<S#7lNh8kouAs;WBscv?9>62uIaQO=o@X?sq8Eu<50GCis}W~}
zHXHYZ5iPo;=q{6o6ss%UteW9qAb6l!F=({I#l!z`=5>@g44oUTN@2$Su4O$HnKN4|
zOtmlR89j_isf6*I9*VlIWL3#I`bOr9%q{ZO6-1;Y%hM3(@d6!I5Bg+|yj0b~j+9*b
znbjrm>&}x{z+L0poE5FD3%`tUlwS&X%Xo0fV{RP1f>~}e8lt`J+hxxAe;7Nb?oa|{
zTgSF-TPwD0+qSi0+qP|U#kOtRPHy%X_wDS5{Zjo4y1T08oL`}d>l~3R9j4=EV)>rK
zn`Vwa_uA6$%eV}+`!8{>p1oqGc6>u}B}Y7MM0h1r!Td?v&I)gQ_BCpX!&}bKUsoq>
zvKorS9cbrL-0#9|;CV?^!n!klE%%jh&%Dg#wQiQ{y2Sp9#b%tNRB)Z|!%k8m3qe01
z6TRv!TOGacFAZY{NM5!kgMbQ{(lHE2`PFk$?|(qaKT8__TmQuNfAtZjKfkDm|0Onh
zg*{=r$$r1Bmygdd@Y@vuh#FnC;Og5#dTU)8AX+np8rL&BrS0TWMI=!uo~yPRgR5Hu
zXjn#6P%URxMO^l{G<>+bd-Hg^&y_g!6SFUM`&{|yioZiAfYG|b@vFmbUBq7FzbH~k
ztq|*oN>-maK3<;1jBvr`m;%g;*a5`2O#?)m=Se-eJqfOa=f=8y#`H$l?Z3pBP587(
z#w|%@ct|4P-AlizcO9j}!(3qOP<llzn!iit#JEw3Bd2j@fT!Gy`1X8z#)Tu|>(06f
zU=$Z+9|X9F&T2BF5C6GmR^S)E{s_hxL60}b=w^I_w1BJlwhe<kPRik#DKNw7sXvfD
zG#jz|(0HSEj{lVEZ!8nmR*-9X<`GUrqet<q<)nns(Jj3GCU9*k%uN@-Gi9VC8D0!a
z!Ylqe<S1%ELo?xpm{nUZ5=LbeOc@Lih@jt$px=#9kn&OmlM;TiF>=20aIyCyM2i{N
zgu+xw&j<I}Z;p_W2A6YP@o})TwK7nOWw{@NlE8cRlXoe?iuslW!zFH5(1hmDFHBHg
za8?BcG@!qF9FBG<)^aO|s%#j}IV;GLpbfH&5Uv#;tbl>&Xqark`H(CAZ5%X;PoPEp
zuL7$f0l6b&&A4b;fJckX`LRZ5o@$R{&N;mp{Y@gFSyr!4Za;(gHjhhR81cgR@Bu1E
zuD)T<V04_@kTs3s4myk2Mj4jn7-`xeHmfH2(ly@n^rZ5r0tLcIC>maRDA$Uhz`i0L
zh}*^w!_!PjU{C{&$|i_tqPm&?y2k9(qWJS%=c%f{W3CT=7C1uR8TPAT`wrp)@0{gB
zlt-APhK;-W(LM{;T-BTe!Ni-#wiJYFY1hZ%b3;WQ5=+NQ?Ql$qbpjtvDBt%Jc}`YT
zQY<?;QM(RWD2L@Y7h<Nbg+w9lqxkaGnWlEevcL5zs}{A^mqn3!OV`!PAya>CBdc_p
zIu&kWXe)0}2vYnyWyxAb4^*7{=O3*Dp_V(SgOK*9ZeBne0y>QiP9CC!@^@<0@)}J+
z|JcP40t=>2<v4?!!fn8_;MC>CP;D`4SPdc{jj1^nP;u+*{`RpmjQlgxTIQI-L?OIn
zZ{n44HN1st3)7Mo7VpclyHDxtP4LPkB1_4QV8E0X!+P%V)c6DhQ;(-~ugec6>`>+0
zniYF65XqSKgRbG^Ecvy2SascL#7w7!y7K~?e8p;HWr(779?Fb-gqnN@PaZ}Kz1=oP
zp0c_d90lql>YIlD!}4wdw_A4uY#-^{KGc57sbP0srGFb?M^RN>NE?Dq7V31n=*gjR
zY)PV))I1e3<MK4&zA<TPAeyY`^l8sB)qd?6#9Da`r`Rl6+*RdrKuxT)cqhZEf(r}2
zF#qqPDN&HiNx)@zO;mk}Vx7cZ*MEa6e*nf84VcX$31G1CwHAIe<P;$ICw4RSAm70o
z$y5rSCilHDIFn74=JJcErM~GQ>Dt?tF!wymE$S4#;F-hnBt{}vFH=bE>h}$On|*qT
zA@jN1w+6}szDoZ&-)E>KIT}Gx1I#k1%lKwQhQt5<xVoU%`xX&s|9B>sOU};!f&$t3
zI`u8c2U$r({>bssf;v*wf;#kf4tiM+SOEOv%?{&z1}7RXVE#CC>V^l~?~XhJbw6^~
zwR$%gm4`9T!Z2vSa`2%GXIdeO;vfTz`5BpCWcNn?H@9~`*E_lQLHYf>_4>A%%Ue=g
zq*A+zVieE5?$Iu$l7!g)7yfcgD6=gFXdtN#8Dni{(4>9CP!{x*f{b+(OuTgnLT!M(
zoFx|FazYr8On8{@T|0vQDqo!}3Pw8QZf-LgK`|Q_A-Hclvv>D45&O8s_I_^UaI?4R
zKi{8sbsa^t#SmdI<O%T0UX%UhwoLaCbWULm6n2|PJE5rp5+>z#1Qhw$xE<{CR+~Zq
z*f32&g)A~!83)1ra;v~ZX_Qm%v5E*2@iFq{#AfG=9RA&oa1q;#+ZWb?<2k%6<OP!B
z;X+BcE59u=`5-$T!&t~DRQNR$S?A-Ldy3m(92iw#POIIsPzbD<b${-2JBB+n6{5@j
ziAtnDxr^^JFz=y~g-5%yki>HU!?o~kn=uSW{nrSh?kJkyM^ArnvN)9%qSRFhQ&Gkq
z?TFuWY#?p6&m(oiT^qGr^bfW=RVlmiVmAwj5@-1$j+i9w<m}#`eGjX{T661->W}1l
zciHiU;so;dx^SbwiS|JNZB+1{BGK2W{$HVXcEJaWoubG?rW@zThMNX*8t>^Wa-43>
z8_1spnbP8@qZ)AyRTe9AsnL-@cD18<I(qM_h|Xcar!SO;WCyo7$S1uMYo5Lt8*t3y
z0oyC^@mKZxY`V=!dT*IXf5$_u(<Kyr+1=Xedn+69&7Sgmt$DW(Iu#$A(`)+?DqmQK
zl$DiSL>*sVkeGz<kep&kwF>r?#nk3b4n}QD7)3$|2G<tG(U@of)-~~D0Bdt3Dk>{X
z$EIOxGsN)<GD1}k*w<kh!DYc{3rF#GA0G71ARq-ZOJku#`L?9wv)$#0g3zN_{Tr$K
zSjjDVI>G_9a_WO}&ha|R#5(t#)|DY8(A?kuiime)6Zic_+*S)N1y^(n+azCfJ2lq_
z$P%o^Hy#jMSW^BQhyM+Qc~TbmI+m0)2%!|&{reqz5M3g1XQnCI8U_s%&lwFioXfDg
zDio-#qJKP2qF4sVPV(5KO+6=<5!TsRVqm})$_x(v!@iug<^}b%S4bn&evFJ2>LIj@
z)zYP#rDkCrS(oFDc2z~4t+i%lUbn}BX0TmoNi){|^MA@)NwX&Ovz=xje?x+SSZ6h%
zZ2vec>x5ZaO~|@(6y3#lS#}VxHL!!$w33x9=+_t4iDITk@Q-(<25rN}`*D-9pwt%4
zmq-3V5DJ1a4_Cul)W<t))h;e*NJtK7;=(Y5-6l6ZE9$~OE=2nV{#@@<DxOTVxZSg^
z;mPC-agg@5#?fu<+ux?l=mJzsR|ALSYS*L$q0okr<WCM0+*;cDb+BTmNU%#Fy5c-U
z;4J^<7<92}8`OK|6ZUaPqts_R5D@`%Z%A8lO|IsqWmxNob9EE{%ISVWph8jw8<nz4
z5K5>+qHWv)Z%&w-@APL-!?&bZtC9@fXezL^Tf2JaCg~drog4F&CNzI%x@%{~Gn?}V
zK+^EL+xFn^hug7S2qH~TufCj8NGbSd8Ln$FQnl83;zyR?+HuYs%3%VHio5`I5LT-!
zsY(?9-BrH)sYA8T!s`XQQ;RZ~3VZHHY>RdUYsob9$S0uDMfbGe;$3zA%T$lci%D%O
zR}bD7xR%2CzmiaY9p(Rrgv!SHzx9SgzeDq3#}$v>9)4I1MSKT<z^pVY<ZP?XVfE=i
zUqR>JRa#G3{P&OIniLbSVZ50VK%7{U@O>E)k#WaY-qF3Q>qnoLjmqIZC^7w|+ix2H
zAHp|2aa$G^bpQhvm>a&(ULNss7(MbHdCBUxhR=<=Y`mL4?AL&U=)HjMU9AwC1^)EO
z*XPN#C$}GWyj{>n#8vFuA=^w(s+bL)Q#`rML8bFbIT@|_CrW-#zn$R&M*SSOBro4K
zHT@j;Wg9*N>@NP#_xB1{I_8{`4GmODwLI=I>2UaKj5Hwc$qo!GrkCL*YFX%P!ZG=9
zw6k}iBG(V=Swtam69bigEZKlRR47r^r?7`}*32?5Vx(}W{9zVW5r0n>8PDff(v((j
zSN$L^jf8eSvf^2eYq8iEXce((E)$(UQ6}RZ;z3HKRYx%?RAf)CM9+2<8f|x6%JdwP
z!wgfgG`K>7P(1yb&7(S{csr=)b@DR2NQ;Q8{3D?Q&G}6kv`8(KY$OxCXTD>u5}pv-
z0f8oI5#Nl!%K2MGjNKlA1ucXb4oqC0c^ey`cw4uN9Q}=(?_>&t1$C42z)G&$9;hWk
zMme~Nx~l2x1O7m3=M2wkR0&GLNRclI@&g&IUc=R6LV5f!f{ZmegD)hwkfz^Ka7LYF
zza$+RiFt6wy%>~{2iPodQ2YL1Vv2&{gGT&^<+&M8oe#C~4DY{oJjc3T7#acEjQ5Pk
zZSm2TuBzfiO`AP){6uF)`<Bm0XF1KDs8xyq&i>ES_6;XTcwzF_b3z~;_&F?t>THM0
zBhRoH5Itrb2L;1%4@Nb9*P!sV)aH)@%Lhlry`_<gXXoX>V{c_UBs>@Rzk!~;Agdp!
z7^t0mX$r)(YH(0-o_^W3m%!px5X_7NrN9W?b%$mM78<@2ST~E(Ev_U`<P@119|a^C
z42G@lO@?~LC!5~c;bk$iqUKkotGsMdG2K_M^wWG67RR+x()A;zt;!rjsW?jvidngh
zj_A#gN*8#kpl-#fwDHyvcl7kd2#qHWyaRAPxd+!cx5BrW#@;UDGn}voCGJy6mhhG0
ze#XW7_Vf^Wkk|!(3aSMp6*1}*44AM5&_s+L1`Nbkmk-g*M%>;Ro)5EUm+^<UtM%o$
z>iDiP!*uYmUiD{wFQ_m`U-1Zme>Iq?!|m6;U~urCl_!M)C&BF_rtYb$VwocF;lg=%
zhB?z6o-N7A4jN{-KQ=7|f>8U6CL}^`D}UkByoo+I5Hm{$(E$Bjje8q(88(-(4Tbrn
zm>_N<5`%<;UwD6<Fn6JDqO)<QHvxoXZ>U!ZbeR&{qS*n6SoMBH9+I($Atgd?Ax-l;
zFwiHS6^fK9D}slFp6n{ik%DrPE?-RHz&ny*!0q8CTz-b79N;!p8@Jak^Fk(qAe>Ai
z70yB$_7&$0IuLHTwG}hAAG>fvkcwa#m;1BLmN~G_n`o+QSz71(>}6Ib*awr@T-^@j
zD(^E4yGFwAHd(DeQGM*Rl4gep&eGo))~UcY2~QwKb81I<4*5?7@;d%C#N4KrdtTN%
z;wppeP{wtH+Jj=5zv0sqCdFAbsPyp5@N3`%W6)HD(k+*MMcZKWVTuj1Oaf!JbupH5
zkd&U2_Xdqx{g4*&u~A#ybhnD1r|?;yB|O4`@Us|6HwJMORyR*^o@NLlW5C6lB^GOw
zy#O|926PaqV^GDik!|Rl&!O>0n&+u)`#^_PSDz+A8_J<+u7FG{$#+p1ipGE}R?Ff3
zv961YL}`bWq5qj!ohay5^e7l1GOuI0l0P?3F%e1>Rr63IzzV4=AupUg8!9QSM1eMi
z@SZ{DP$5-MgXWp$ghdk&uQV(1q`f*kS>1J$mqOqQbwj%o9{wiiCZtM53`i?6tbifM
z(Ow<H1q(DjpjCON(pqn{B-v9xKzHP1T^H>bBM3~;^n5&KSbyrzG^#Dyq~W?o$!=>k
z`w2tRdL0^WB~DBQ!?jH0y6P((^lS?>=Ms(9X!?5T=W?@?(y>C&rI=Zkl`r!;$7#Lj
zP5HFd0UM=eN@%jweG~&#s7&TQO*hk4$t><0O8yd-Pv3QUPrBqg>yzEkwfaj|#v)KF
zZAWCYUE*}8GqZ@fhVOU?nQb=`e5R-SvA{L{Tk<=~U+D%4?Z&H+GldCg?OGH?Mhn;>
zr8kyL@MV9IZl#Fm6^sUSDoK4zmRHIP!}w>b-g+FiJZ2-lHA|%EdilW~<A}+&HS0F_
zCDr>)tVEKoiWY^LmO&Mw86#K@hyBlzwQ1HV=#8t14h_%o_!Z++8;v~>l$8tS7Gx<C
z{9z@IN_W(BV|)b9LFw)C)DRm(`KAlVK-;U@oMCz0f7(<q6(s&E%j855erux<8%nZN
z+8*PAph8A2xyrYfs&_^WP=r!6(H(oL1%Yk5nr?RZQ3iMAc=wVV_13H%+n(k7_L3Wr
z?(J*SIu7C>Vffm;XQr?*=Us%_dpKe)mS8qQ-N)#+@#r~N=$m494xAT&p*#5BX9RY;
z<-Fd41*ONgkl}E<u(INh2(h_wr-V6tgmaZ^-w$x(R05Jy>D`ujk_I)oAL_b+E-Dm8
zd_9sSY?n~B%7jY02l<0j*gM)2eJxG7q|AI-_844eLAD81TJ?p@t8-AUE6Vx*GR8EN
z7p}V2ofX%Q_E{{D6Nww5XIsUxid%IyCa%Iyl7iE~&L}h~-Ky_IJgB=^c~HD1#wx>J
zoYtLA2IV4RuAW3lW(~$Q5GC%k<nWPh0I|3KhFa3UP2w4rvef(3&ADAOb|QfyJvPpG
zZyu%TR|%E>yTjYPL&$s0fV&St&@cA(8nWn;vR(&Nct{zVLa`_oSDJo)YSOzoVP`;v
zvTsfPqxR}o>@s@c`P>c-Rn0(i&Ifk=!a6w3t?T9b@PljE2q1EaqyFOuE9+PBMnCKQ
zveDb?`!TVyqd^#Y^8=}~#$?|Hx(&;>Kx`0(k#@m%-?&#Cx~mc1y$A;GI7v5wJN|F|
z)}>9Tg+6cWPc@%lCHiTcP^yD{lNhTr4Kj-iXedB^OUj#1Q#5F$Lxi}J2r)M(g@?8K
z;hl-HJB-?gcE^30FPz&>Nasv?rEyX;B~Lnl*MP5-HmHA<->hQ`3uvJ=PewngoraC7
zrk1Bw;XcAno+(szU6pKK+u<73RB>^BCkta0P(Dkt*G^A=XpE{Y=8Pl6JvSW<AGR_I
zlN)mA$c~=xvBQqXqL!1K{k1<WWWpZ)d}F6S0QEbz;{>>(;>&y|M$Z1d(I3*%>3L}>
z=qliTB_BMjBqW;`VuxHdW)OaaJ+P2RG#ld}(!F{AI469d<L#K8v$=1>*OqYB<8#Fs
z4BB+gnl^Q<omBeWCdzj9E$>{#9@4i(Vb7~nXUOy5OUl9b2+Dverx%aGAfYd)*bhJ6
za*t^Z3+wxi;26mPK`@q?mqlY)xf3T}eGvP+Sx9}MbGdj19AjB<Sq=oIc+A573CbAu
z>lV{Rn@BQGk#WtHtT9arIi3}g&vSo9Lpc;)`D{Nc*|S=zCWB0nuYWiMx!H3eRSDN%
zRBhaa2e>V*Tl{b9Xa{sQ#x2R<?}wj<!q6?n0l~DBQxS4|i3XQzP4tDWx_qL5HQ;ay
ze!hJGu_d-+-!UyLYdgL}KYyaeUw#02LL>nH+s2grzbMYktp6KF_dkmBVcXq*Hm2rq
zku*JiBE+^{-2b*s_csjTBgdu>8Q>Z+g?#vi#1u$K+m~#x`>@QVsN#n_`NYT0K1fRC
zLLV0P1`7W_Xzy$T<WX?mHx8u9)DfWHGA$>kRL>_tQR6-PEL71_hEMDzr{>ssan}L_
z<9!a>Z()uQK^XW;xu-(xeQV@nfw3pj3Au)4XK0QoRu#0-lZN;PI)n?4a%O6S{hn^0
zv)FAs2kB3S>Db4D0%$m)qS%3-yI<!*tsEb|G2-B8vkOHS+p*>XjZow4iH0EN%@PUd
z{ESs$DIeqv()$te*vdeKcXB@JaIkcqh&Yb8A(M}8U0i!P4KNT*)G85~5RuS}tFacn
zd@$&i1iLe7aehZSJg8=#jv;xKX#F}YlWa#qeYzJb53*WK2*wBoZaaN~eK?#CcJ4H;
zv4m3~IxruhY8zdeO#8SfRCtF{*sw%R+!4Mt3V)Y3Rt_&O?zg&nz5<2x?ZV25khecp
z5Sfg>D-h3r7=VAdyz%DH5Tyt$WWUXfRh_}#6gT<A=YyJ-(`0A(TlDQ4j5W%Xo6ACp
zuoDos&1z6#I%HBB2g4G8!Z8_bH6s3rlmAE^y2a5)FK8kNbRaVS>xk5n>M0nWV@wnN
zt5`RBk%&L}GyjfTq?2nRf>W^c8_t7>6e00=8x|zc96K{Nfr`tIYOuZ-YT!$77K)F~
zsK$T*b1vRP!sDc?E!=>JJG7yS6VNZlF1$2E@Z~YR?aw<SX$X01DA#RJLawSC@6FGK
z!=M*yTMKy{xv*iE&v^3DMxE?{1t$hN2VD^k@b2=o`43W<Zn>p$y`mANNS$ZWGx!I?
zE{HUWzV?S~*|6CjY<phQGs+HQ32YL*V4dL|wjkq|@Rs-Os4zt!|EF+=up*v6*v)-T
zc4hv)*by5TN^J@KVvB?|Et+}Di;nv_m54EDX|5qjjFp-OUDdp!U}Q!BUMF}Pm8#8K
z5@o6N<1Bedz&e0x_kL=bskTpu6fZBM`e&f<4!JmG*RVq;zQ9FrEW_;)cf~VU{n17x
zoTn2;!UgZT>VZiG-8oX%9dlc&-NY8jLbVio36p($skJcZzN5wc)F|9sdT+}HO0I1a
zhYKr%i`kOOdb-1A3VzLcpJyc8irVU^JQ`44dAW$NsH>y2ftB=R5;BbNgUU)7rCunS
zNQJL4%Iyf>CfpIDzC2-0k-1t)MZUn%y0DOkwEFDu+aVj&1?3{C2Cr#x`{^x;@`{nV
zJ!b|X82&)N+fdj7T|c{A3#!gwsM`7W5#Q6{qfkV%61t`!MWqX&T_8Vgo02)iVmTiC
zNk{tnr`*23HAWqjN1%BLt|8&pWj3IzmP&l)NE(X2m|t>j-(hZj1_~x1iQRCcs1?gR
zY00YWMbaEu^W0;{!VE(%MYSCPo8e~n&|8Qy2B_Q{yFCtCtR0G4o)d)GaEl`R7>F-y
z_cP!|?Q~dSqOJl{<?<|_9y9SK$OhYJs>z-HLS%wt(zI;f1AfISs(5i8;6gygLXzGm
zZzT&mUALZ26On@6r47Nh6%kl4-9OqTRDQVVD?s=1!azn#6XK-jRM1bF_L3`z1xoZ(
z1%2MS#vMrnJF}0}2)2>!cjC2=r;rl($72a@Vc{(PNUs{-VtU!~-Pi6oZpy`VN2H-6
z0xBiRVL<b&T-Ku5eId9xjAxaza9xi!I*w=Mq4caivCuw*SowI*qFUtQQ-wL9Q(vuq
zWSEA0(;y_1VxDok@Ytvoo_hqMy4pUl+Rf$MV-XtC`C+i+oK}zf@0S5rAotSP!mWa3
zF!!CyqoVL*)!jL#6uNeESUy12$+M(aF9+ZpeB7};B!7M9IX84FPWHkOrLQ|(y2QOz
zt6vmVA;A-Lm#TO42=DpE47=PX_ah^edqL&+&iKW{DuIsRWvK~Ke+ML9M^BKFh~#hi
z9px%&;EZrMx>C3*fysGK`@y8yFs(1zRX(<1K0#orB*T(OjorOKuoiX5)P`=9jd=!B
z5OHd%OE((!DW#DA#LGmqUvCjqn;hS(D9ZNWC?T2=Vf{k|-HSFPTT0f1)K4qRBN~hN
zhd^bQ1L%u!u1rn|%Ts(6HeYV{v}F)a#23KQL82w&+pvzu`}O6;&iCUTyEyRd!*^dE
z%E%a%zlrBW44X|Dn4*A&PfW=%sm$=Iwp_VB=?G^QAV(cguM}932YCGz?44N9s2Wx7
z=Xgv4hz{-$6boSifl-yB_N5S%N#y}4yi`)uJ3@VY^Hu!8L)AN~_c=XuzRFiN)NWzV
zDZZk9yqMW59T8m0*GjRN2^YtrJl+~~=r6+&ixeq)MOaH7Y%gjV1?qYma=|_!lD_m6
zsLN40A?ywek#>~7fH&?Qwj+i+>FB3Q5QBeL<avhiZ!9MhlzYebFWtg>(u8?T9t)_0
z6q|p9k#w7)HnY#yL=1P9m=omywih94Q#;5-cjW#O8j|My!(v>L2AAsg?~Dc>%v$~g
z=MlxbGwSa-#J@b|xVrjmik4*&EA{AD_)zxedP{nm_{BD?bk3p?V*<=6<W@q=XMc-!
z;|lanb93+!!@6^9n}rWoErYF*+86R=^3GAKIlr6>JwA$lY?`&JvwyGM@M_we%DGvE
z+KkI%>KXP~2Iykm5;5-@G(Y{t3L#Vo<q92p4n}5i?S1iv49%P;r`$-IV!p$yVxdhG
zI=olLrmZ5{)44A+wPeZjWh}M^oX}RcFh3ifuBjIGNMb1@Wkayj^dWL4u4$(Uaz~PD
zQ7-#fC}(o3{p-}qrrR3*qJvgf2b+J^tPEWE3ArjdwiGS;?f#(@y<ihZe)+dFIHrB9
zEW%R{jn<covh#+%M#nZ;jW{GdSvaiVT3;wiA?Z0dF@#i(u6}jx<qUe&kO$&4tg>I$
zB>v+sj;p9{u`N#TQDn)C49PuqTCmjdyfr1(nwZ3R8C>R8Jf^4GSkPF?nKs3`-XL#T
zn<je5T;_Tc0^k*?1Kv7a*|*sr)}}6;^@dA3x+7YdDou#$wK?2EutR=IZkcMlwET#y
zwfws-QFgLGe(<rdK3yY9K5<sV#g#k-m}UV2<o!=?ATeVoF3I5I;0MQ^3firbZnN$F
z2Ox-j_w&CrV;l_s_ox9A>+cEe|3$v;){=_lE{E-%DLvh6Q2eIR&zUbPKFc{vc9ddi
zr6{H7Eafro{UHQ{0WGQ6*$B-52ldf6`c>~YzgiMI^|s-5?D%s3U3NjJdUSt$yOiDr
z59Uf57B+I#CfVn??)2V#KhYJtzT(_6Dgx~u+}-%~UVVd?`SBzTf$vZ>|LjHgqshOX
z-fQLFa+LJG@XdF!ur7=r{5_|YvXe=A3ired?-#qGgx}g&d?21%KXF^XYGm;%JaqUG
z60<N&fcv0a6u^D{T~U5y7KOjQ84u7wV0Q;LYuIl?EH+OA5r8J*C2eMn+hwvagqMn(
z(N-7;S{ApKMRUw^E%tZhw^BJ@GD^9OkxgsvO&TT+p2$B=D+c%PcpefGHjC~j8Yewx
zA957+ln)i5i7Doyg=a2G^u)5VJ{2u*4j6w6k1R5>h_TK>Buu_8FDOog7YTigDS2id
z7f)T4Kf#NLc@cbF>R!0bw>s`RX*(xo<3#z^s^HOfdSIBEE$y3Js-V$Z^_Qi6Io9&8
z<VVYx#Di0KUNkC6%j)yD?N%TwKgl2N3kp<s#IXu)jEa#Lx^%JPA_Opmz)!PMr^IL9
z6nU_xETz+w9O`yHx{*|JovVE{$P;u@Wn|E(?`){}Tr`F42+qWJ$GF(Y!>>$aUmZ)q
zHz~xx5Otv7%-8Cz`s(f2{bXJm5KULzb^gZuQgkuofdYmpANoNWKMs03^>)7d#uQ)U
z1^3hL2gbMXd#}_d_$XDy!cF7hfdA>Ek<O`_qgMw#($b~LWpq~XX^K*#y)<M_qC^jt
z*>MbYv5r!i9L;N|lbC(Q58-Rde~PYg<YyfxHRBvwb~9dcUJiK<kch3pl314ah?h(#
zd0Oz=L<wrN8SpwRb0?_PDTON>b2fcG+-EV@V2#em=qRBCTSqEfFry1EA~&;0x!OG2
z)3R%<P<d}zt3{(R(K-5z*S)m|7oh6OsmkZV%SREtGbl7wGM53Ovz(4WA3l6ddL0_;
zU6~QDh<#ZXxd!Lr=JLh5)Ab9TA|K$Fp4!H`fpw2aH`okuT?7m18Fk;N$Bt7X7M~=o
zdLAmIKc%=0x<#yJ)$-i%Smzu$eF5>Osnyoxm}zREejTD3MB;p{k>;WvJaa~d4G3bK
z>FSzTlva_y&R5(#7}+GxO4GhM_)@eexO0cbJo2sp*Fyv4*4OfE)&#iqd(F7}Y}E?Y
z0p-T$I9_lDQM8e#K(M>scz3$vS)HkAKG{~^$@At3)IFZwW4}zE!~4%_tJ^Q`1_#ZK
z%9+!fryZ;B{8Nh|ebnp!6m3yh+d21z9WFDOQh}t3plWzx(&&u7P%ea(E6rvg9;0oa
z_K0~5^Fr5I;!M%=L+=L>R#M&LfOHUe@`u&-s9c1$wcx_;2pT$wX*GxnKUI*0b^Ls?
z<3n9ceJ*b%)XeKYi0l1yuH3N|oje~wI%p3unUUh8jg7a?uK_HVN=H730)XL44&9+n
z^e?)yZ^cd|;DD6>cSz>kO?jn6U=}=V>0g5ybq(Fd;~X9ar5B4Usweiko)1}gT(O=J
z5sylhE%MH3nSXmbRj!?R`r3so94Laq7-&vT)RwAm9#zzzsb<7G{1koyD*CpK4D6Om
zBBGh^A$xeG#b}eQ-4PAU53+Y)oEe~;4zsl0ux4n^;iL&-qI7TwD0TgabqL_s$295V
z+~pAi@OMBo&0A>uufBhfNYOD0{=ZXwz<~i4GkASxaR_l4kd-}Pg3&R9!1g)Z#H(Te
zzVs{q1ww)*NCT#w;Wq>z{jS@@NH?*;4_0oA;HQZYwrh2%>9%Vg4|o!vb+D-JGIU-*
zTAjyi&D2{~DFelxRVPXoZLcN+DO;o3{=h)S#G03D0e${olLcOo6}i5!?+{+lW!!Zo
zBK3i1X*|}_Zm_>Z>cj{9$s5e#c(6N8t7F7TJoSYQ{8q&<YSZUvS7-|AKXANO42w6`
zGHFnQ3gH^C>2)$)ul3}E#{&yA;f%JYH?_{I(Wh`Ra?!VI<B8qb)=v8;Lb54f59skg
zhKS5`=q3<~?T?7;v(IEsED67uujKdOixaDw;Xmti{Xj&iR<@cp0)qSsce6;^_bp*3
z`3>jG+mDrUT7U`;aB4c4UVp*3Z1CyKXpK;IiBb*x$4(ftp#%wI%nBz={w5~S>1}vQ
zY<~-M6y@Y@{pnHD=_rk2D^ZAutf35NT4=55rOHW^XX&R!Z@OUSN@7bwQ>Bux?@k}C
z<m@3E#u00ZG))&8KK2lWfJ>jvckeiC?VH+Wd#j*dQCNawOJ;jS->m#rJ7LIuOz8+%
zR*b}7lth%|f^B(eJ`_rcU!#^)p3ReSBkf#mRVcS;{I!HS;{07cINpyrn&fo2svxg`
zQ}cTM5&;;uG(dpA`vet~ip1WahKt1lL}C9LlwuIQ3l{;1x(_A-SpE<Z0pKhynyjQZ
z!(0LXEzNx05Po-Ff`|a6ukOF)bQ%rNp5EOD(NC42%IE1U-Qh}ZX23<PS1Hp3zF7Ow
zJD`OKYj<(xb?YieRMi-VcSg%WB!}x7F&b@~LN+?PpA3`c8ql%f0k@s%Dk{!0L&wFF
zf<s*au+&nVn%%9oy?F6zS5+!n%cKp!mXD2qwz;Ou?aQnAv$;%7GeB><Z-KA@s_jy@
zJ2gr=w`)*r-?tyi3}%}CRwKh}#a#j?`Z39e-EI)VRyk)5$ZQ?wW*V2F1|0|$@Oz{^
zb@g8wA6hZM0I)dkpChlpDFPKc*X2K+d5?na6xv<&j|Do#fIJxt_sTwVCnx2k)s?H?
z7ZI<WRvpl5VKvGI%-D128s-s-Zv!xCbRcR?6e1|+mL88#Kv!*5yVDVk(C9q6b9V{Q
zCV{tg1VrN_CLUnYoJINm<I5jNAQ+O`oBGMMgePVet|^LG=JjQ^ih5sUNpgw6^mH=o
z!wteVSv-p6n;lWKCvy{V2zEXHqDf{Pj<dh%ot4!<7Ckfumf2FXA6wcQE{!pmQnItg
zxJ3<`Qfz6(%1~@m9LrZ5bO8FAlgaYv)LLoc7PWsDjre*x+3~e{_or<$I~X>l2@MO{
ziNk|aDsM;Qsnz_r{#6W%dV|+a8J}Z`?S|QkwR8zI2QVHCcZI7$H$tf#8UiKFjwO%R
zf8)!>jxL$a^@rueL^bUhwb;yEgy<5@H#prGrpMbOumj=FN!T<$VxrMjP@64JxQSKX
z{F$R9nrp`3Ix<XP2%7G##xh+WVUn6AcXZA&rl*1x-bhY62Px&EQ|`+(|1Ew`ET|$+
z1kaaKdsgfiB!3^OFWwY*oK4_D-&7M+#Ne{Cy4}o&>hv4yJa3!urqHjQA8YsAH{M)s
zv~)O(i<*Vud4b;;fpD1P32Y`eC#=jM1s}gf%5o#qX*6kp(sHa&|N5c6crrcUE*JPw
zJKW=6zGFL})EXy7N>yo2!%iL%R!Mm5%iEfB`Zx{lqWG9B+0Ptg6m6oCpQFg!@_K1)
z>}yfyLNqe-Ye=+SHQq6eob<U0cOE3AtDGTSW4l)8O25C~eLnay;xz{UTLb%Fu5vN6
zFme1(4eXWH#P0=o*sU40Mfj+y7hndgGm9CF(P{K;Pz`O2`o_Az%}fy=z9eDEq>A?~
zxC8>Qb)MoBitwUQCz53C9$Lq@*T?&}v#rc}DTKN5ywn*zbWlF=S3L=eM_WiqNV~=R
zw~_nO7e!0>T=Z9U%4mN%zg}YR&aFlTL`w>A{8y3kEvGst$*6$Xnw@u8-q2rE1;<ay
z2AW3k-_d51>S~1spMebUq`K0K`{1t(tLw%poi~{e9n5Hd)%!ZHWr_79G4LmpNda6}
z-|x>6^6-c>v#Vdcb}Q=zyy7psg1OU-+9?<Ay^&z2=SYrjN#4?cm2NRMm^BaXOxmXe
z>qw1fF;-dgBJtM*wQ{l!orJIXV_W**c0IJOo@K_vRbheroL|A6p!2TOu7W+~AG?@9
z@>$Z->=W-@_U={@9dX!#n8Z=Jt(86o_!Gmvp}H_^Dew~D8K_~IWs;tqI|2tau}<F{
zA)}jith)|T_E-Ci*{gpXxC@QR^4`x76^q(Z+UqN4P0D6GL^E|Q;>=jCx~f=LR;zA_
z6NZ>L8(ddW6rt1Rp_OXc$)h_c`yLK8JsSlaNc3)q!bJmhN!0POnR)K^ZBWaCRmM`t
z@_JtKSvvpksef!JPT^;^7I_kONDeo%mzJ)KM&r5CmD-q1sJ$Y^mY`s!%Xv-uRvlR1
zyb`npi_s@-UZf}$hnaR&@URHwWZP@!m6N#v<<Hga4ZFDF#sQ_&%mKUelCxd9l<d@D
zf=P@~n`3%lq-kZBk1po02x)=xL@K^qFD=3xA025GpM-&;^8{b2HcWJ0!wX<fv<ek%
zuTZj^@6YoSyq=E=5Y{JgYg{DidO1oe{?i<Xc}kN;&w&$)CMY^OX#&eO9j{k#0Mud~
zs<%c?6`UX@y2jayc1BP%Lan+{VNf_ARSTRtVl<(Yb@Aeko341_xlR&Cv%z604d}+P
zn7Lf_z0vorSp`(}cL(=hogCgjf60J<b~nJ!S{KWMQ~>jUl_*iu2^7Vb6FC(Dz}XeY
z$=9KpZ|Ah+vto$EL4BLI*Nw}AWYQw|(IoPS-g72mDfBa}R3x6H$h8ApbwP5o74uaq
zO+R+@eDoXbW>1MfY7BGw=!0B*Jf0cEP9g)*|A9@{gRa(bE)BDrJaZjr(hM)HBFjWO
zH=6+oc}L3#8bIWtu?})Bov^ORp;v<I7Xx$Utjxg7b?9m{^|S9hI)aSOqZA>Fx^(Mj
z3FIBhfQ2SB1TfKOh}2{STX!Ldp8>HzUs|<;OJ-;3^N)!)yG3PJqi>vlKmr)T4DqrA
z#+3Y2V___mYv$f7*r*+<L3+KzO6n;9yN*}0?8r0Y92?2W7XVqPlQJILCk}|ROdp2T
zDZERy7J;hOJqx6qwL_U1X-vBXnn3s@U;a7vd_=vc8Go_}=oNCEDHHDAX#sickYFY<
zALwYupL8H!CUknMc3DMN(Fya_8n#Sh%+(k#3In>}0<~+Q$ou8wjTQ-qOe3MqN<SSa
z7EFEvH%gHhhc;h9-5BF$6&6ps=Nod0&PGA5`U=n>#w$_c#w8MV9Y9QB9Wz&p4#=ae
z5LctBUD_EB4a;vZ>sdb}ynY#2^@L&5R##lgtyW$gB8hn(t@CrB3)3d?iO|yU1h!mM
z#qOQ`QM5fE!db88|3+bBY+Hzn?qi%iH_>&WrNa6f^H(=S<<jU*wcbI5$kII6xV4M2
zC{cHQ`K^Alsv3>4YQ(osGJwG2L$ZH4Xx-+>J`g_oyU`c-3*3wEhOcy)2I4o#bF$|4
z_rzP-#TxkliieLCEIfsuE==xkqrrk%5*lZdxPhJq$SN|kAnYYL+9N~A(&}Y?GJ_ty
z#>8_~bBd~^h3MMO@DxzhQM}BD#lu*xq1>|IHNzf@*Eq}2T2g(FqTeILhu&!ZE=Pws
zcIbUSFt}8-8Ni7@AwQ_)Dij#pk|a3X61H2vFm4?Y)_^ZUXaYdF3&p6oY9A!V`A_*`
zh9pNm-e57ozp%W?VC@J?<ng>eYWLH{o8d0)W7+G6H6(J+boa(O2u$VBV~I}^{IeZU
zK$a6oE0LCY#mITgV-w~Bt=$n_gcL%Mf!rWx2gQH2W(FwcOMnK*0PLyS0d0|&Lp6jP
zeM^#2b!#8jPdp~d0a~&YBM&}a3}&Z_e=*cyrTd1srm?g{BD%}Rp<ajWEAUy$%|%k>
z-Q=>&z_Zy3Z4Rsdo>E>Gt5?DfqX>v>nTf2W--*_T&?XetO=cG8Qmp9NJtpqdC4H0j
zg*gH_NHY1C{e(}VWx?;*$gfCTiVre-86<0Pk}!MEy+Y<cj^QK1a!e~_D4T9J$i9~N
zUwIDw%TrB1IGJUSF=y!!R28eG2%5srYs5hwJf`H;qc`v<+*Uzi&!bxULDxG5d=Ps6
zp-xq;i3WxDk6LVRka)Lb0(bVAtIZ>CAPH2$XvT`ctN6u3oJNm?GKoe=9<LgR^Mz4h
z{Uzj|i6!&Iq1@269Y>64S+C&;vOv6?{nHBgvv3Y^T9mttq2k@_F*?;#A*-k`-vy`)
zgb|l}(ON~gRz-^#w>QW%^X^jm<&{Eey?avYV9s}uo0!@{8tzh?S9%;Evt~3nVxHW0
zCC9}oDXHOOd8lYVP>v{{v^Pt^S34c$xOYBU)`TM)6>+;aSmyl0=d2HcC`N5Zp3<oC
zzT>5jLYTo9aD)N>7z-G^Koe@o_c#~^yEht+DEquK?in?<EQ#Kb6EJVHQ8(|!**3OT
zy|txq^(&w%-9-ZY!QLN+9h|P%53A>_m~tQiom7}PV5m4|)w|gccNIGcW{tS79k@#3
ze=!(7)v`JZq7i&qhZ<KdYpq%geV?;{Ua}Vs)N(H`z`}jaPSe4w;m+EoPJ-OukbLlH
z0e!=4ZG8T_L((*b1{;c~7d60`%-3+0^9rVuY`Q$6MLomMx1TByLXDCxha|hY*+Q(l
z4<3YYEq1%hJXR(8Y7PSb9M3e_on@wI4Y)l}q<-JyCxK}Zl=}CJ!RoAN9gJE71l~F^
zV~u4%g!2ydW@R#AeIdGv#U51Z?`br)le*^>qq$=0CbE<#50#Hx)pw6sCy90HzjxM-
zJA`CO4%w&dfi;2BD=M}5g>{1i!j%1YonhS_+>T2}lI(uXPAzX>PQNxJY$uM}NUhJ&
z+fBwIMj|ri39iPALTOVjPunFozVT5WF3W@g#(#%AVzr3|QD>ceV(N3S=yQTEIZEl!
zo@i7xk51e(S=+S0HWzgXqI04oxxljD@9aDRAn!I;AZP_qG>jBK@_lvF`b>hCZ(sBi
zE2VOc%031i(i)!dzae(Dv3cDO5P$`H$1J5XufD!@7?smrr1?-Jym3NPPgJw;KbY|T
ztfl@ie#Y$%3dL}w1W#L937H2WVO34JUK~bL=Wk`K-p3#YDG_TwRJ0%z3Kw@{62L3s
zQAGZVrrjMIq{9SN8f&^ft36-wemQ^!X_jp8{`mqqsZV@eA_I~>d)OA~lU2^dJb+xx
z{zSbV4qFp%9SkXxt${7A!5JsO$gs=2^ABR-d=n2Nx=}{A_g74It$vzUlE5$S9~5VZ
z+q`G@fZM!N&rz+h%{O<?<vYyV+qn?yD7nj5=N}TeEdoJ4FXbPPf0uEKdHXvG2gTIg
zlH8{|jIO+@xSqVSUGBuAt`&5QAFh;rW0cKJf}~y9k!%o?Yyw)aZ$3l3fIWfh_HP6q
zGx3XUdpY{|Fgvry^xo5Vw{Expjmbd4-L&lJs7Ke?t~91<(d-z)-;5N-x-vVR2kH`?
zzkFX27?p^t*{s=hAq~c$f5abC@>?lw*J)~fvGQLy1HIQJxKigmsbbieaaT%@YIT$X
z(vpog;E~qM<{zSibc-ocgp74v<YClu_7eMANM@5Rke1UZ?0_>dILq=-h~yFj!n2ES
zQ{!okozXJErT}K1MD|=}RCE%W0GP2TOeK^g%{80$)vInbXax>|^iV}#_soGpL}2g?
z)YG_$kgQ0m_hJpLiC>STzC&rav_%i=a$m-8m+y_;z0Nbx;Gr1uIC(ZQ|Il3};ZE1a
z6;1?h0^GK<VxJVJRtH%wC=A2q97^-9XMk)4^28lP10{D0<)M$yoGXgOfg6@f(3|Oa
zYL}L6c!!|9PvpF*9H>oV+<S^Fu!xqAq*&MWgEwOhO#C&<WB4Q4d`=RPX_c7BqYnyD
z!Td$0!>rKeSzS$X0&vDzYyYtEC$$s@J;=?JG~Q+ym2F?)4sJ^Rd=W0C*sI7?yn&@Z
z<d!+%p2VYn<%R6nJP69-+9dF+w(GC8K|kyTm|Fkhi-kP@rimpfyf=9<b3lZ(1O4$+
ziGyXS<9BxG+w&|e(p3V4tw|U*tGM$mkzbHRLTY9irNy4IH-I@7X!fwQQmGh0VqfYm
zLc=EsfXRI%QX?Lx&%9Z@BKEEfg$1|a-fB4k+kzzA0mgg&rJR5QBR*Wohx{VRfh)=%
zC{KV0EKptm5g8!4!ZHbZw*0csjF~_}6m$<TKG<$PFm8u<v$zHMr#g{?v4jOSgm+)S
zV=`UN|2DEQ{a;%&Gvlu-<bTl|{zKo!=|=3CQF??9GmRgDMF5Kz#R6_$o24BkWH3C@
z2Y>s)xqqj;ma1h|j2rx?&TQSFj$O%%w&wlLcz5=2_ICAnM&-aXNyT93`N0dwv-`+r
zZ@Fm61RyPyw(TS1>v^bc{?5Gs&XJnH8`7ioUO5eqMDqk7g6|1r@52DGNaW8;^o^?a
zuKoG)5a@e60;Xxub=U22NsDCSAj3l+-mi2$mYZ4HbCOu<0yh)H3vNpiy`TCJXpfV9
z67%kVhm-U3`8<Lw7qL5O@w4=U>D%3t?ZXvqIR^p5BE932{Ry1(uLw1<sYoKr=jY<_
z>(Qi0^LU>ni1G*j?=LKYrO8_+#32i2pjowoIf?*oR8^CU?5ofmgQO3o9DIsk_1l<v
ze4I&whJ0L-^t_1G<iHMeizc@Q4b4L(#Ddfe#id}d5K)CiG0)v^M`a2CetpFpL$a|l
zo%F!FgNjR|Bx}ymg2UVvlRsXVY>S{BPdCAsZ0bx0>-UUCiXENA_1oCrFLnjzj#h$X
zmq|Vem^5bdn^nnI4u;hWnq7xjdX?At*l=>49ILZbs29!$8-iSs{WUP2<GYRdB3-J7
z-CW)i%zFF89L^azp@HI)j;Psirn_R}jI~Mk`)v<Oqq^q_qYVvtw@+<xMG8#{&Byy>
zMD}qB5=saO&}_U&)4;`i()U*R6d+JDaeQhE2N3++W*^#0%VWq0;5RmyoV*sDdc7Jh
zI#U>CENBy5U_bQVEGP;dk&S%V^c>$L%$pjqx6k$btJNn6s%cF3d9I`!HgUVBtP@FN
zXgqI3FG=t=M*EmRnjn{`Ex=F`>qb6T&{48{YCx=FM1DWrA}KkHA_;YH-OFYW&*L5K
z54v`NmBm0)0*dq%d*W_(w3Hp}qyzK8#oMz-b4e1ps3e4tdbfW;J{nXBu!#Ibf-x*H
zQ`Oj@m+~ALt?<-<VAp1;R)NhH`%s2uBjNjYeEP@Ih7#~M+!N6Pm??^b29{AevL)FR
zu4Ge!f4apTPmJMT(cE+&`=m??&LN!f^rqU#4#X@^O7*74t9L~GG=L}n6f{8vvegB-
ztE)8TXJ8sy5+=eSa7|5w08uMANkzc`1nGETnf?d~s^Y<mD_)vE#h@6JwITs1tnbr~
z!=ldVMW^WqS<Kb3#Se#GQplTxR=RJvz#P3X1;;~MPMs;oH7~U>IoR3`CINz;hh!yy
z^?m4F$HI(6!jtSX2Xb!AFA30O?Kl`npT{z5>tte{c7TlMJ4JuftHqtn?8v@s?Tnp=
z{>%}}!2O8q;IOYcOvmft0Rl#VpkLoguD@W`n0Z$Tt0UmC!(~@$d6E@CN5r}b7*Ie#
zGERT{^HR_-$YRGSOto_5+o!}wG0Pk&3;Sg;$04KOb*9ulDEnnR(;cp8Sgu*qs%-VY
z*%)q0RDn2;P?W_kvf^+Z$et7G*||A^OKWjWvk7;50ZfbE9epoK#{*@hD=25iS)4|H
z#sndtFX*eQ+n;S9E>K8QXumV<HONC|N7>%Nxv6Am_*ZuxGg;U(-IUOe%{{8)u5I4J
zCcWsa3?z5K+A*DCaOK{(I{IgD*p1~hHZRurz5Uy96ZIkK?hPEv`X@$~+O6$4mBbXO
zt$S})F6&T-6c}V@T-`HKk`s_tNuPg68lg0cypc3fP%q`ng91@6vDKkCwKFJHRqF*C
zI^H!YhsN)k;X(6S{Ed5)zta<F<`a`7lNxD~01(ttNK_f<{sV15Sq-$i*6!WOMj*)G
zYk5R!<Hn-Q;<?kBU(Bhg<)RCUl1^a_gY$8K(*L&`*;$>o6UI$CqXQj!f+%I4TX(|c
zrKJXQ!fj1l(p$Y4smWOIG`<v`LIsL0J;7FFeSh1E9gzYF(^nf9q(miJSYRfSxJciW
zSI`x7x{YQ&05e6DDy8JUlxoDppq~50(zzne4pmMqRX{D|>QUZ(gAWsdh>z`}V|0BX
z06?HoY-dH6Pf-6@x~qu5X9}U(TwvrA*<1B!-@&5;LlEv%b!AJ^ZC<{)bma=+G1ddu
zrK`=A6t>Wo#|Y5rFv=wVGN(Bm__-?&Z(0t^3Y_+0F7ufO;BQ}{r{DIApU+_byFZy|
z`SGCTtV41K4=mQ_Pe@tlTUy^{=XZw#Z)sOfvE^fDsgwYNTC$Q9!$-&O83C6yvKt?h
z^eLn4tL8=*xYlM%IFdV+E%~O@29)J~!tWT{cSMf)S*pY_PC44ry~m~(N*-(T&e4Xh
zelW>OPff1<3MZ6al`{!C$jQ$YJkBF&<@@}NBNz(2NF==-Qwl){7{BL@+66mgi?;_9
zWtlR5&dL&Bnf!ED*2h>szd7?TkVHwXD~JBG1wJ3|pMloi*&eg^JGq@?Iq_A%tZbj5
zT#)qk*(Q;%bRVtpL-WbQ7kb5mB1mhDi<deJeT!GtY<Y}~>^gCquQM%Nn_2lVQ+xJZ
z=EZvHSr!vK?xFdNlaDo4)0%S>fs+1^pQG}k`fu{+nZi%G9}C};8kTdrg|iRm_k^a3
zCb$n#AWg+!nIaMGc`62w187E-#`bX6iY5&?!Jq&>js;08YC8)Qkj!pV`PXsuyq$rP
zwTQ{~;&?7L#A<0mwxeF7j|e<4jq+l=F2*Bo3x5s>@b<VS>_6JS*5Tygue@J^cW?YO
zFvOVo1uG8(@MPX?Q$J>HL!>>c%S`vizl(toX+Q4i1R<r)4lp2=t<-V@^6@8m&<{s-
zOn#JT?z|{QLsrSSbv<uNN%gf1Q_R`<Wi>ie{F>$Xjb&3<aeAOZactdT&cSCHOtUn8
zX2m{j-6?mokDh3RCbd*%-PeU_im|pp2j#GH0cudQwkql^)0nb$R-cp2iKjFbqaDP^
zVINHLA>J*DuW8t3zR2uylM#}6#M|7=K|4#t&nMEm%}vmHSPHc>Pgwe~HS6i1Q$WA&
zIqNX-;hZ~kMQcWh^|mD$6>lbGrP)cU$wg#*Cz;-~ym2NiO$)*rcH7yExOU^zJhEr4
zZto%69>_`48rrA%`INnGE3WZpU^g}zX@4re1_fRk^v87>KXdEu{&>coEY^e#gMKP7
zS2EEpf$Ga`_rSRXMc_y$<6e|oD{~#F*FWwfem^heSHq>?M!z^M8<yj{RGCo$B0sLM
zfu=k0SSFu-gAn~}#2S87Vhpwpf5_teK!f7uKm!+hK*e+)8%mr{_<V+1nbg_pXL8@x
zK6(-i(zz&ptM;cKWFN#V);wM?9|vD03A|vvayL;tIg8xrH(N)y>{|+wN*bE#S*lB!
zIBCi!I`LzAdFnWDLQM6B^0J8b=3V2#IikY&SQjl!O{>9aXJx3|d^@$Iqt+B6=qXku
z;SLMy&GnS!80>VXtg5wk>p|izDVn69OB~G(8cfRDJLhj0KuewRG**j64TH8zZ<)7-
z?z{iT*gFPy5<YCYv2EM7J+W;&6Fd3Ewr$&<*tTtJVkeW$fA^fTTea`0daL?NS3lL?
zy6eGp-xm&m5luL44%iaLXQTCnu<h!2K2SY<Cm#xhhh6$?LhRZ`K}DOD4ZfEI&z9~k
z`N;KH#1qf8>V}(7+0gl-ff_b564S(}c&hf2{!n`-MiRxyO-dlmqmdBZmnVv+#pxm?
z_o(q7kzY}km~|=oj$!w;^%+6@8TcMz96rm{4lI_Deg$sIPb8E5vc-PAsIb)yUA0vw
zL7J)8U!&Q5JELyYw;7SO$juC6vgtmV{e05l-NC3U`SfJ?v?P!W1(L`-`|8&S#bce&
zNV0A!(c!P!Lw+D5yE2<BZBBiA;QTQ4uP}Xk>aQ#+0v!&SoG|?;U_ajB4&_@{WV9i*
z^&t*d1_VTJ-Pu#fg1bFyQjse6ZjOcNlF7Qs*V11!`-owf+;*i(=Ot-li>X%>cXT*C
zUm<A@{%%ibraHytPR4Jo>MQyl%oGfF>tm`Xv0ZUtj1Bgh&bIO4VX!igJ1Z>2m&;Ms
zN!pv%#H+<5ZIAMa=^-CW_Q!0}$?-R+zdu`Ca8h#9f3F!29WQc7E4(xJpkOB(EN0Y<
zuS!Vo>(Y}n<DJF2&kGpD$01TgxOfN3du<z0x4sWnE7=$ZI{ZGpB7Nd==$N<i9DfvN
zHJZrmMRz7cC+P3Yc7MVZpsl>CDUw0ZlZvgw*Bhafh-0A5`LTx)#&PhbN?|dH9_WXP
zfNSg4O_JLINGx#dEM`^-8dZu{kd;!kOd66Hg~I$-8RSF})GIgSw<^WL=Eh9BLsMWQ
z<@CyO#yCbopFf@6O2w&VO5=;KkRIPxSKYq6AV6@KZtX+O%^UyA1q}E9o$KV_{_jCk
z)3}qdTmOq-$I2Y$4Cp6Xd|-;tifR8(WtxZ6Pi5Lk{I0?sbt75JlJYDM#1VnguyMC$
zwQ+*C2jS~a*QqbB4<O}t{ghR`vB!r1kni5Bpo0w?pdM&S5^Kj7!LRdF7t4F*m1?GB
zX!Kwg!S87>2P_OJ|E`}A2+>^~xK*-GqS)Xra#rQ_W1qnr)`zrqZ5P5aH(zXoPM`Hi
zW#m?q=&k(?hlmigHtba{JxnrYNdNPEU*4nEJ0D>mh&c%G`mt%Czz2J<Sml;hY*=kP
zW}OfyN{3^m6t3s>DTdH2e{*M#C(K=}0*-?wa<(vdR98?YX}ck+y~1YHY%U&?d5fsL
zYeA<50MMG%$FL;tSO>%!%<6rQM&mPAZE2W!Z(w9H#3g#VGmUoM_BAsFyHlF>0hE3x
z%9PSwW0R2wc8k*3mda4-jCW9}R!wl4a-oGTwAbZIilOJ~5lAFEl3}klEB_w7BDc8_
z&J|0npK$_>ef}%aLaA&kqq-|1+s|>sAcyZEuM*&iYX0l>O?F3rN4__&oc}lMlr_Ki
zhbIhSG8jdUP*^|Rdz9cmW^X~C0!^J>RQ4Hon(`*My>eN0ENbkTgTgk$Uq(Hb&e^jO
zh9P<vetunu#mASY3Fg#TV7wO%Fgu<BFg0B7d~yX(Z|{yN#xn}(0d{MbNPtjDYVo;;
zXfz|3fW+3XM{X*`pbdz?3>cTVazL<5G&=L+oYpVYXGeI@zcgqHi)FGXPaSz*1P_J$
z<SPI}Zemg^GHa6ou+y<dLVT^iPw5sej@6zk65BH|^+8&f>6n?yik}c>D^d_E3TN=m
z!N<a;+G<Xft^rP+mZ$@)_S?Z6>@r#Lk<ipsE}92$n8VY_Z9EEl<jtZlITf*XI&4w7
zb+8NUOOyFJ>2s_ZVtg*k`+~X5ZuADffDYfXc~kZ>(3s{!rzaz$E`<nTrg_b=W%cFB
z^n4&^L_Uci%h`TLmg!6JW)JzoSsr<hp`>lCLgy@@BiUeH30^wI@s2tSad8&lA$SV(
zqpRRhsmDKXOyM_e$-eUELuT(9sj~s73INIa@!!s=`58k<O+3t4I1Vz*b(dtXL*cj^
zTJ7WROD)w2!@5@u80OZ?b(RWp<=lkDAeN;}LuC9K-eM<M*QE~JIZ08?JlRnb^T;aY
z1dDm6!ITs#jSwZCHQ`NGaF^DS+JgWQPdM+aGo=g|Y#k=q5S)NmBL+79C9*4f5q)|Z
z0oGWoL30U})C{sCzR=}_I2M8-GPdzMCqP+0(_E`gMw6}E%0G)@&j}s#X{XFGy$s}4
z#quxq^V~btg^F+Ojy9v@C{`(_zG69KY#ZqpA(kq-6#WF1$XJ+=gu=c-nMB)zxs>7?
z82uzfyf?g-qe(uN?eZ8tE2q^bnv89eB!FAdY-w)Aw@OP(ge~_p-mKOy9KpB1DKR;N
zD8x`cMABesZ-}vO?jlu;!33$9@Cq&2hd7Fui7_|1E2nJ7ZNH7QyJDhTxCXR4>+xcD
zG#+H<uQ`bABxE#PJjvGTfqEbctkc1hL?A^wHz;OGNL3|B6l05wMAYS^uDcE*1=hNw
z;LWrU2Rd+j=gb~`0T4RWeEV`MVzwq@IRUmDp4#)o-`nzwB0QzpddrXhOe;E&={M?K
z2LoGE`BKPAMz?H-%16I^3(7EaDaW}ffP0`oD4nEXH&SJkse1|u#g3M$ZTOakS$Zem
zN*tq_E3Y3(FVa-)p_C>H*yJr|tDkX?)V;7|O{0tWLD;VcocPKA5FI+hlM{gA*O^m$
zR#ps|pkiF2Iho>A!F>P;@4ZJOqn9$vYHHCF-i@Vgcbb50E?gwK>zuE7kD}yQXvLk6
zl@_;7S#s|m^<5~R*u*6;+LWfWOXRzcAe+^YZQUMPA5+Lh)w>>D2WORe>Lh1NDYi8Z
z6Ko#zIJJtgJOk@oF>eoGFgi6(Vt7Scw@T)}i193AKyz}|56flNwo3k?&c+KTiTnz!
zS-q{;ttYICO*%80EKkx5;hka|Sx}BpxoZ7GW|A;tc<_acjuQCh6cP+%phcXZ&I;e#
zC?Z0oSrM-Y5Ns6@v0B8WB?AN0h&UO(XyFVuw<{s%OSG0i)Y~hKyZZZbHC|lb>4EJ(
z1Z0Xh*@G4Iqz^v#@nY};lU%JMr4Urd9)YDgkT9Oq4Iv<CmZieUeS!Wa1$)R-!d*8y
zC%+?JVgUD0XQBrSQSU19pd{zMD@dkGZ+3^{U!)_qMlVcj8}BJ$N@oUOKGXXVBHcb9
z6(9#ZruTfgLrt$i2gtEeR{(WWvbelI-n`uSy*=aNtlj?0n+rlbGoEkj*tY;I!}Qbt
z*_ZI;OwY2$k6v(YJ`(0F?UwLgXyEY1m4_LGkKg#85kQLx>ZRc|ZoAeEAWaC%Oc;1z
z6I=SjFH2llwL_$YABNroR?}(sHu6iR?iMHax!d2k&JUsNX7F4lr7Z0yN3LRbG6XFv
z%w@L}-ppnMRM@PhIJ8eMA0D!jI=?|RH=2c{7^&z8j*gZnQ8ksC8oIv_{0Am<LDkpl
zBM=CxPKt~hQ>Y}A9$UA%NKFL88s+-+r7f?0m-qSoM~TemPDC~_2<MKXR44*c`)9sK
zk!f-dyaViaTzh<>v@UN)Wmlb=cy5*{7<J;UnWdy`*G|lo?>ZgP;{je#hAXhrH#3a5
z=4^JglNQyE*_ng>JkK^nee+y#bjNIM2W|Qaqg)16v+Shk#967VDMNCEc(k^zfJH=m
zi3sS``L41DDh4Gb<b>z3!b#a}{46iUgmT8Sla!(mTn+R;(9XCha?w+)+i^*w{e{E*
zyks48{G$U`K$J3O&5;UleF6j{U>Q~Jvh<20wWpKu<eoX$IB49^`in~=VptP7)S+}-
zzgGMZrvP0H!l#FaHEeX9g)?OIJiYX`H<W;`uGM|oNhe!LbGSOe8Qr-6Z15~!z9Y)h
zpk}42#`QxJS>tjvhj(m5)oJp8Iz+Av1Z-mU?Jjs=5mZts7JP^lN`NqLO!B6ffdhBh
z{+=0jdYt-9j9F|<73uU+Ek3Z_FsfCr{7h!|)0Yoa57Ao)<^28aM09H8VH&%GPON8C
zX7*=Vxmvp|={GUJAIG(Ley+NEBZi*AT6Eh~#Xxg+%?;f1HTO?f%mAoEqNPD~nc%E#
zNxJ_jY)5W_*4Vc{tA<u?7N|wJ`kl*N15VW4@ZUxgZB`wP4c|4Yyuw|n%RMcFvwy=B
z8*s9>;#c6j!LNE93L-jfo9;v=_osdaU7z|(4MOecL|p*WqKY9mUUv7@9s+j=Ed&3R
zU{~NcJ%#e)RU~RbBH-}|#?WOFeCxFXw7<h?J3#2B_peX}x9@g<p>F$MAmxJMVMtx~
zGv?jB4p8Ag{~5rj>kid`<*?ppzwC&NMtfdQh$?(iIr^tm7pY<OS_U2L;aC|QxTjGS
z7!Ql<3swd_I50&SyyV4120giTGL0MR@pna`P5Cpb$u~yBscg6oIupQ@5D*bC=aa<?
z$G?z53m3fpMblTT!Y1)--zK$7`mgs-uaF<J>Kh>hlGW3HHK}F&KNJtLW_A`XmPD*v
z%>TVf?J(Z|jAj0<n?p>W<P<YB9NKP$9J9r6hIP_}V@Vp6kKe|-|DaA(L(u|gCnf*`
zHslh-R)U_0s>nJCbFKsgw|WFV@6By6MQ>E+wpVX*0wGcUX`tYAUj`}<;4b@ot$$r7
zC2tI#LJ-oHFyBJox+CuXJ)bWp0Q_K79)(MbCiO7dKNuCme=w?-$CpJDK<bFOrl+@L
zx7m3O@()Ip>mGb7%YR%0$OH&bYAU>CJS%vzLx}%GGmByuMw)rf%0#q}7k7LAJqeQ#
z`+7Qm;h=lhgE)}W6tnaAn*!rP&e2UBka&Yq%%tqT-8@!@J#92&;7(FOYHI#?TT7<j
zx=QI|Jx+t$NF@%E9P^N*3k#*3+uiFF|7sKa`^a=(Q+fleM@|h&4Bp|I%A0HdaDyJ~
zR&js2!VQQ{-`=n@k*<MxrU1mX3Ic+n0itDH2SQa-Rnklbu(qtB8MWcl8$2;LV&PxG
z3edSl8S$`g#UhE+2T+b<+v<*u6n${nB;b+u4#W(?Q7oK#=9Z2OiFd``?;qF|Bnf$y
zebil?^h?bH5z*SD-Zx?B-Na#e$hltk(2<GsJs};3pH_IaurJ<$Alc%2N&DLlHi>Px
zPm@azB?kRmg8<rxO(7&F!4O*J!)R>n1$0I+h48baN=bvf^}5wML_b+5LsvA8*T0*y
zVUK+{<g#_V^zf2KyI*f{e?JB>5n5L!v|5WHS2%ctq>w;2J!?XO?F;ScHjs|RU|_ey
zzW;sBH>4T}#=)-iQ7=1(OV5|hCk;{^0uv5P1eE%`e6B<aI0<RbC|tso`;QwR12^N_
z;9F2wZkq473qR^p)lnj0G64r23<monq+9Cm#>v?S!#ILwHMk?K$c9J~m>9Z{Y(SIO
z@W@K7Rf`NW^^fHEL(yoxM{IQwgp~)*O~_Hl&ctyetdsp>Pau6`nUt|@0b;c7AVru=
zgfKZM-p>EG#v8eJpqjcP$e1qodbK6U<AE&l$MXAZ2l_vE9dX9ukc2|qIJan5DLRzT
zAl3aputh~k!U+hyT`}fN2rM)6_yToz1(+$$#eyFSF$aAw#je1Fnhp$dY;<jF2`<3}
z7CW7$6)yN1I|R2^rD;V{SW*`?<8qma8}B2@?@^6jvo=yuXP;RK;!%k_plmD$b|2*-
z^T&hSIBd#$bZtwSk#Xs8&TUgW&%P!hhUWNmYsWDw!}Q=qX+H-@?J-zsKc*Mko1z)e
z@gaeK@$G`sdHqwYk`{Q7wAESCnSw?quJ_a2eTO8)EhsCah)Bdxw87Vd^2iG-QC!1U
z)wA_7!4P@-T~G<L^@tYt*tPu`-VmWpI#1X$MY0&V1>RdJ_%<&zBsWdl4A{dtHq1x?
z9eWJMQOtXck6vQ)WSge%u&>6-c-Dg<l>^0wt8Y0V;vAa+`4vbEjq7k53tN^=Z(F0b
z96>Gt)47ID8Pn$RoHiUcKD<V4QRF+Lq>Zhoo}&G1dG5rDgF+ck0+lW;sKN?ICgXEw
zRK{m@eb_-x2vRv5tGw-DMgnOBkxTNT!o{3KxpS3-uZOqA9ENO16!WvWTAtO%e3GE)
zzJ-LtTKjyakT^1xh~Wp?W6M6EJ3`wt;!efb+Hf4C+*@DQ<sp(YT9nbtK?2=5O)&J$
zF_N)Q5Zd^yFdrwwzl>Kt!rAYoMX@QCc#UE}nyCbbk8q-b6Y$DjCp*L}y$F}5Owrx$
zpbje^=y0ZBgNo%PEb$Uh%VteOu|z$0Jea~UMH8^buhEF@!Kd6(pup*~S8$V0=eNLF
zeqJ*AHgpX{aAh|h?`FD$Ct{=Ro8x%XkIdYSy>gbcKOZE`SZT^8Ayok+Rz7>H(ebys
z7+fb{@-?k0zvNDO(&eyj$Kec!%FNU>3n#Py=0@{@|ME5GUf2N}+1}<M=E{(}%Ijjm
z9$@|1gKMITc*0<6`C{fJ4Hjy-^xG??#HPS@Si$y&H8pX%(};R=$AXT$Xa_ts(THSr
z)EdtC&&*$zf3|DetuI=gol%FT?iSbr>o9Zc0QY5e%K}^5kk*%sntFTv4Va&I=maR;
z1lII16Za1b_3OH`i=z;sySxzMp^1=@XT#$P@2g8kKXu>oa75^Jun60Sk*5jk|9p)R
zbk$86bVZ^ueDb7Z3s!1-kMbimWNb<U`ktwXaVHcWyxuu$3_>+|W^x)?F*S7_XCu$q
z%1;1GSC`KL^|n2%Lgy_v27!c9Q)(yP5g9GO4#cZVABg_~wM`jBg`M_E@kXFu<;bHT
zYo31wmFmI*MscMjGlXr(!{ARZ^rR5SZblasU-=%Af&3|?W4H`t-m!B0lU-vDg2O5o
za&|?#%=vT%WkeO@80M_91149DTafTiFQq)*Fx!aIg!|3N=D9~qjmP*?pGVBoj5il;
ziwpILMq}<XOIxBBIl6OuMx6m=MJ@yw-K#(e;NQIP=&asfc%Z%lhi$*kCAgscrdeaN
zYGr2P$J@zXKkx6!wq2pI4#ICKQd7vt8EOYn-_$&5oPAP<>W>_puK3a$pR}Q2kE~VX
znotHdH=DN=e$D#hi9bDd|M=M@<`Oa1G-778Dv8}ov@STbr2~1`7K?jVW13qXQ=>QQ
zlOJh?`y0+Fqsx@eEk;`2<@Ah}1UtiSwg=M!YUlfUyDa+EGKi}{Zm6-;l2S8??!&r~
z@s!9Q)`(zh*b9FZ$y?1xb)FxE(q^$i%y2B=>(B&OKpP^1{VufM=@9}7wI+t{XdVz-
z7Nf=cDMQPF8P}HN{r7cE8T$cTB%rYiK4T({+&bePM<I^1lWus7cN~%leI?&LjtUsj
z1<1r6mbYH90)CmDv`>%t51{n8ETh2I#*sD}xS>i0#KC57Q6R07q8p)kL~g(ZYf<eH
z;R&HmaG{+1E1q&tPH22|6@}zjyvSvFT#jFj%OmC$<WEWB^9upqbYc&!K25LU)+`^m
zrEcJNAbSLqh^YUY%ZkBie0=x~Vf7~tg0)l9e(hP=?8X9a+j;qe!|;hk72ER3iY=%e
z(4VfbT%m+@;#Sl!qu|AoQ?=wzgaa1`pk!~%eKgO($*^{3`_~B5wWWEZ$TZQy^&;&%
z@p^(63;s$Kz7tv8I2K?pKaoq|O4ROz<NLOzsyz2F(9*#vMci7B>?F{g<xti~V9lFK
zq^Ben$ZE8#lDnG{*&u6SFhgB4m13!)d-5V9SHZiI5N#sc=`0m^S{|0dTZMAkH3t~z
zq>p$Im$Wo48Y`fqE#+4#b~qZRbGJAiw<>t=srW@wu}uFI30Bs@Cv*8H$fZJ^<6@(z
z)c9JzhVB=Zn+!h5<iLKjj7#(q?YNE$I_D)#9gKEU?7kE|Dv+7BoSLBuyP9ToX(h3`
zwejt03)f7G@2XrpG-VK>Hs_s}?hC+Yv(J{6n2M87b2l|`9>_dpn#{iVDJhx9oVMj9
zNTAFh<few8q=WfUu)M6+Q;279HG5qNxV-SJf7s$Ow+zTz|4u$QP~xh{m^<g$sk@O`
zI)5RU@Ls-SstAp}gH?T0QY+2NTny=w+6=QRKGL@cK4TeWkLHreJafqf{xZ>`WzG%6
z=+{dLX|jIE4|nuo<ci;o$^Yko70FuAYz(2j?n7KVYR3|x=3Ytti+QgDJj}z6-lqgs
zwd8s+^XfelKyS>3_dAjDRHT9jq2qb|O4BH#n97TyUE(m!+^gGEEM~O@^T^7@7Jekt
z>~UuTOD7KdR+SWg0w8M!>Oenz@mMy2!LwA|@q`S~^J>N$z7KSF(B6GB1f#a3s5vf;
zBB{1jg3ivhQfapux9Zx^VxT*sQbMObP0>4!0yl1{-cR3kjNarqOl5Q>@5*3@OIqCi
z$2PiOdh5Ba_BvF8K5khHRxF=i2x?DuR(0YM>E4Lqe79!<vv#0Y<K(U7K7Ejv=DFcC
z_@IQnRl*<C{ThYby$6)2WbU&3lo_|%c;IiT&XvpG21wrPYmppwfsBgeQOc%2tBO{u
z2X)ZwnZ8v$s_>LvkcU0wSp0tt8`)gD-34`}=>Xdc0jvB4ADZqt2+fnDbbH-K;5Iz}
za%AHT^psGITqL62D;F5!uV~F{q=(aKH&!{IRM4fjC0v=JTfqbImNe0+iUQ{t)z(sS
zmR%DN)E^h_*N-1Ys2EAE034O&eL^nfFz>9iBA53Cjm2}qF$Y%=B~pL6X#{&sV<B4X
zUMeol)a0tlio&1Cp}uA`*x?GQIh+}xsjatn(4H>Q%*_fo)Lvm?-FxYTgL-XTXAM$;
zC`du9XM1RgPopPkcT@U^Qgdc&o4kv&>=kj3r7P^TI=wlO#DWB8%`;YMKWfJsAOeN?
zl&|6#xeenV-(?UqZ+M)r>hct|@aiP?yw$C-nYtd{L;(}HImm79^^pv9f`&etcboxy
zNrMe?pTkckwuHMRqc?HXqb`l-7%a*%Vnn0&H9UI&f^A=v-f6F1`Y<{|HOXy)>Xz8m
zd5v)@gG4Doao*A+5^M(5So*%4n*=LJe7_i3Sc^k?7TeRP7L;z&ka_oRx;)n&!$p(I
zq2$S%d8u?0zpsn>%qY+vb<zTHGbc|GVk3GC&_DVX(~6%Cu3#V1@67i#v%FvOkgiW)
zQ-gtlrw=w=L2?03>df-uZx{Mbe{G8G?d30=!{^CQel}_1^05tN7lExUI!dITTA6rQ
zr~f_*gCjspokjeEfPo;Iwfd@|JH!44wu*zh2W*gDb|pM}kR8&L<gHbbz3K|!2R-k<
zx!Cl;-k);~J-oT#8TV!?i~<NNw<l`3EUI}@==j&rLZh{6kWLPm@^Wxn94dR~G=2mB
zeN5R99`z}~m<g?0a{EVQ5Hat*tzn!~_mEPmE3ika*UOV^jqv)*o17x;)Q23x;pZXM
z-<OnE-|NL1kxxPM#WvWru~K_p6Z!Y~G?Drrc1F6m)j5m0PCv|mE+?oqR=IBf=!QR~
z^td=b$Gh-N;R=fn#!%TmE%q(AW4EoWK_EHd{0}W?ga5bx)E&;%eDfWuc#MrzcQkyG
z_(g%%oBGvYZ~y5(!oKDb$^UB6%>IAJ=UJJ#x&B)y+kfKocx}i%H+5%%^=wi0AW_2R
z<{BVQNd-FbzlcO<`eBSu<9qW+w8)}tp8rNSw+1Xv)sGuD5E$^UrEW52Z3zhI3J88a
z3TI768TMCQKdi#v1`PQLALKV>K_bZyKEG`~w^3534EDnQ!=`Q-5scdf_1#!33iW60
zJ0KJQ=Hdr{auou9tG=lc-ri*P<Pj`3HUC8DFMTrUCvC)%mmxDf$ze`;RP@&t*&ebj
zbM9G><uh<&s*au>%M00<9m5`gsDpUFz8(w+$)FF89|pZ0%4(qIZ4G+(KU?>Q;&gJq
zh(cof?Y=kf>w#jvZIwMN_!5yeI`&$39S;<~-;OU0dOktH7(VPXX*fxfL$F%yA61IB
zQl2*J9yg)+JVXgZlQ%M^!GIU`f;;Nrz1In%@Behd+dQta;e<eoQ@3gc&GnZ>xPo%1
zdP<^wP+h;!658mt(6cPY#5~d}6||gF*49-m%s!6#tycWGFxVViz1~NaTIM<BlZO_z
zipraTG*MRuH|i_$WEo!}J_gGKO{YmYD`|l&`t0(l6jH;9p@bWd>9e<Ob35S~n9&Q1
zS1(Zv!?9j`4ZtgHI*?>lrlqS?!rfP5uq+f>phx+;+v{EA+Pha7chxlK304x)VCUsR
zP?9ZWrdV$T4a1g2fl5IJ$8=_)Zjo2%Rqbu98<*v?fyH>N826=`XO{|AnB3APS?JUf
z?|=7NBDDvEaS1(fL<0<l9YkfWFTZhMk4hu1H6#<2ZgMTv{|Z)Ot|m79FPq9!U{=?S
zEvYLY_rTjrOKF02QJ~dwZ0@F-UJOmQD2kK<Lsv9NboW<3xdi;i#gTW^d|oG++M`H@
z7K@bxTHKqEH(&(#uiXqNf`+y>tr4X%mSQZCm+C}-mMbc?`O}SUw)mg#G5_8Ukb@lM
zg_?t+BzccB+19f#Kdg$|U!@Lc3LMmwRjo8)=)<PqSb2GY0)k1%e{%92)XOh4#ERE{
zRCTRduBzsXQi)!)OPhvB4ias(p_tyxJH}LLNp>R_2!Eh_XT|XOH6{jieQUXLx~}Xu
zbk88N<*H`LUm9Acm$u1So><$1sCCU|+<33mwKgan=@#i;RA+L}n^})M_4J<}T{1s;
z<DJv6r(#&K*eb}e8<6#8#S|PZA}Zn%+(N@swqWTLHlC$L(W_c*U6&NOMJ^`T2`(9X
zFhhkC6;d&GS$zcD(@s-`RpF%`xA=)?N#3O<LVTqkM)bBPv_%RM61y8~ElVZ6hC6lE
zD09v<F#_CJx4%PK6VZ`#zI-{?=&#Yn>-Ml(-A{Ed)}^K_GBt2Dj^%oy)oVTe%ta4k
z`{ir78sbDH!5No9XzhkByol>yW~BNrD}{u>z^G4?Bv#jfLTZ$j;po<~XNL&Co0&uB
z8$C0w6`1idm#KIAbNxweEP5qD%twcPz&`i7S1m-zdc-WVPcP=(=wk&pjh3VaIUF55
z#}psfeg3UM6gaidz)z*swZY^FKX2Fj?fmh6YgIZ?H2{P~Q|z1$`Jf1l^}-ykjUTd<
z5Eki8XP|fZct)n!oqjnKFTSLr298}=RixH+tWo>KrWVsaQLVs&m2WJB{xpA~kyCPi
zU;4nYO2ZV`Vtg-}a=RnE0Xc)Vt^#hQ+)SufD-8(?fxQ=(C|^_5()0CsQTuQRlcp4q
zx4p2g;^usg{<sjlBh4;56b+9<WdulmIJv-qIRG|Irc@(HPo@tvnREdSg3#GgJbxVm
zVo=55OK##n5<sWARkv^qFGHT*Yv|-PeNBU_)zd?#a+Cy;*wXDJcdN<}8EK#907t<^
z;=CY#MK-WYOvat@LyDHfpwm1pYo$<XLb&f?1YfwnafUQZ{-zPi@HeZ+M@i_Cc>OC4
zmW&JpD6Q5{x*$=a2FmG>UQg5pX{*)*qQ|+in?(U)U;pEIaJwr4fv8H)gh$TB{&d+S
zsoe=k2&w4V)y2OMYe(kh|8=&E;xmp~v`W^7==Sms!K*Y3vI^6mgU?N0*jjZjhV2C^
zO><}`KF;GMO9X9_?S1fYf|jGsz@~tw>!Cbqg(K~NV8zfpS)|>%zXAI?w^$X;H}Xu-
zcZZO8vZQP-UhEWzcm;OKFc=9jPU|v-^{tzGQd?(tX`M(jDq5LsUE>2w5}JCvlU+ms
z)C3A-K?$#Ch;MUEcQ8)>Fn>f>EX{gDsCmsOkt=&j9_Qa9#2V8>S&vTJME9BJ$-aDc
zzoCJsz_oV!+lJ4}^-H^G9y<zc`siMiJKn45N}#UuZxzLX4|;h=0Yi#?aeQUFFS{}R
zKrPQq{7hxw2}VsUqC$~fCLP1DgPy$FLW$VS=P#R_h(0;Y4!I*0I{NyW#%h3}oxL;!
z56`%DH(pG^tf-)jXzs4&vFOC31$PHU9O&M&M4z6#YsPik6F)%5#UNxr^&tkQ-6^A?
z+no1M*!HabQ6ILbvAHsNq%$`?dlX?<6&Z)%`Pruj1Ek6}b+WOZU{=Gp5e@QUu&P7F
z&~a{dtEa$miU+*(wv&}7x6yLBSyh?&v*1_fQR0ADg)~pA_dwR*NH|dEH5+JA-$Qb0
zI~#DBywk*9q-AA(Ge7p={;TwE+{eVkP(~bvqZc(})<J2H4=g=<P_gC&OHvzPga-my
zMDsbCvBsjBTAvN9y|$zOMU+PyIU=`E-yEC$X%jE8BCiYflk+LL<pso(o8~cEoo?Ay
zkIUAM$~yX|6DW*lL7JO0nIJBmP`-pv$#6`B^t8w1?^nuDjzf@J#r-iRr=2JR&$fJA
zw(%vRTP5f3#bzgxeTW*)r_JhN-xaASwGC&-A%?=AmR^Dlt932lv)d+KdO-NZOpcP-
zAUwXicA2vJ(n`E|at(%TlP-2xS-UrnSh-lnQWu$xQ<(>tiD$ZZUGe^8@p5Q|(Hu9S
ziS|tLRdULf&WW<K9pNy)akX+2#|A9=DhM;QP)rU3q*m9(b;-T97-OjXE!)bkqY*r-
zI3Qpp2`X7xbCq>!L%94FZK{g}f*^0Jz3S^ZV`RI}c*qUpO6OR-CsgYfQC9Zli81RC
z!%4@Y{b!!7?G~SdC85NeAZa5DrI5dcYZK~NgW=vc=ECXx5M}1v1Uf_RSwRRy3m>S9
z@=viZ{4*$BeRJ$LbM@soTgMy^;4vqitW5gpZ@CMhWjj*zmH$Auw%3xyUUM2L+*cdY
zlw0QZ_bPi7h^CL>pu#WoYSRMo7o1mNXgj54?wvnI&b7r`QwvxsZZi<?<VpCNedr&K
z$LTt(dpB)sQZ(ZXbOd$jT7b>!_(NvCASe7~tC;c4*t^~d6aN~HUVO0^kPkNPOn*LT
z!Z7P4cF};N(v==Csf=UtnH(>{l`lD@?v8-Bx!02RWQ|u+C7Y)=2{m)>q|XU^n2AfL
zXHb=L8&FNUj*5UnA-+>FmMR}2cHp{*6D{1-LbRX)*>AxY9-pjMEC{M&%;AXpmVOLU
z8FUJ&62shhb^guRUWpH)PIoxl6JIcFsJNDGzy?SkFT*^KFI{}dQ`{MRkB{Przg~kt
zU`%D4;_tW9n!0luim?Io=Z#M62$E&%$U5*5>$eO7q_wl#dU2OC*>{I4-TzeOxobF0
z{}u_U*Es$9Esge^>PA*j9ujBXd;ODSom-^seac$^+9%!8t{nD?UO}>Nn*~T@S2u4L
z_G=R0xbMQhZecK3_cw2z20_@Xec%(_Qh;bul2n2200WChLc&LN@@*5781wKc2{7sP
z-A{U^TU@H4<mo|^e~a*S2l2HaH}rP72k|ZHEb9OQ_`4}&IuS?(cZnVxHA;3T#t_Ib
zMOAr!W9K1C6E#{ONbT+?Y5u)+dpZ89(WR+&`^u|Haa<)_o{+qTsL}P4q(x#TX2TEK
zCG_emX}(rapZH_7AVpubdL6W%X_=$u?{MxXj0M$OLomJLNLzG0T*L!-+qV>Ce!Bl%
zPe^&g)s9hbv(*&@jK#=y<_;%-I_z^W$549%BSaw4H@$=is_3|uk!G>trM`sN>n$;1
z72Lq20@<lfPhP%wMHm21r_jtkSQ26Sv{Jjdam62Nx4hHoC*_LSnZVKImckNYC4F{t
z$ji=i2r&VJ|M5+v$Ow3hQ-N3tce7Xi=jaz9t?4oQ+m<Z-Qpa^Gt8>ys8im8o;fP5X
z_SW&_WMsk7k=8Fw5SzR!gPXrn0q!JOb3iVVi{>{Ys`1W^Z3X}2jxlSJ6ul1U;_2jR
z*f%Kt#2EU2v6kUtWBuP$+yB4go#p>?yt6a^_gWT~we{?E#^HTl>ag+|H$gIakGvb9
z;QpuM9Rn(n?mv!q<H-nREvKgh({CR{iH^wTMjNpnw}ddRu7xe)uwMiOGgKIiBN>{;
zEJFk$PN7;v!G?TEF!Nx1LXs3G)gl}jk2Q+2(A9xh<5H)Np<~CWlEK)&a<bsxidjYN
zsU(#uu%ve~bV-7bh{Fr6RE>rtxXYQbn^LM*OuMMDqU?i=MhS3kF;rA*(zDykt*CJj
zs<l~hW3jTq2Md{kZ8TktlZ7TKs|?$R&$GY>gDYWCipeU8aTRKP)%W1*I#@mrR9dKU
z2rt_-I~00m21YF47&IdoUAgOb(>k&Pjg-?!XTUT_RQn17vKi<wQP(cpz!PVm-~(<V
zm~!kystGH|F68>tT=@?Iq1&~?TtjtT@KW)ERYMR;i!U+G`bpFV_WG)C4wx^d4sOF!
zDuYLdG(-M6O+2YxX_vkS(sd&-^Zr9CqiK|iZSozVP^vyGl&ZcXVaPxe$-z2Id>5%N
zS4~X}Q<W?+OmtxX>CEKee@1B#r;INgzp9J{lS-ovmaNVqR+WvF&k0aF9(t!QouIyG
zi{>itCnt-q)URe--9w2P?ASk8URDdUkSy@MB(7K-g7g?#b!RQEel}W-DtoUX!mdOR
z<xedofJtYxG=`cIlSVu1D~3|}!mtxLE1wp?36A|KCLiEf*6IrAT}m1gsLuDrh*=s#
z`lC`<Wp7H#{`$ynT9D1>WuBQwQ8fNIE%$9~$j9>8Vx|9d`q1xU<MH;I#`uu8JN5Ap
zP;d;%NO}0yWHBv+C6C}4fS8pg*7V01i~y;nSYTxE-0I;sPrC<edLDb7ewWP-LeJE!
zf<i%u=f~zW(8La+;h={A)~rQlOS}9XfGLZv6f<a@W%{Ye4~`-1i;&V_#$`t|v&=j8
zrOZP`g5Fh-Fpko-uvyv9Nu5}lz7DAcMVa<UovMb2mbgPbUwu!#yq3qHuJE_*4&Bc5
z)A`NdouBN(p(%hnUEO-y^I`QQ?WBxVTcC?TcWO3#eIOdCORdjXf`ddLzKi3;WrvS5
z{Y&oQ;pn>9)5!PF>s)Vtt7f8LZmu2IE!A>?rjkeCZ&9kmVG5v{Flwl7{N*EM$z(2y
z(@2OU#J~IE4oF$|q39q*pT4w)#3~btf7gl6a!&v93HL)Ru=KB(umbI4KsT9y7Pa;P
zuS0&4L&O&&;2{_2PWDYe-9$e1o%)F9i+9A2I}2lBCI^wVvntesT`C%_<T?;^Hd})_
z=`{#M96=p1MszrfS`Tg@d)WYO1wrQ<`s<JhwZ}Dp3eqj2u=N`U2{RSep!Hkex(7Lh
zYJ?^=m|5`7$AlEwDMn(@f|MC$r4Sw)%6K;pMbA(q_YMhxKlNzFH72Y(-R_zXMm0@f
zp{IYp)%g_(nP9LS4qLkTUIH<RR4u{r0{W;ZEU4{fUgDifWs^L#tO?3m%{D4*rD7{=
z*Zpd28)|IF+Xmi~*2Ge@Lrp=nW#;3jIQ-!P7__91Bam3~g{Y(C5peCfFxSBTUD^Br
zCkz@u=>z^Ls*Ml`wNqxb`tVnktq*{?1%^wLsVAya0F4I*I!CFA2?vr4L)*3<(u1my
zO4$}yCa=Puriboe|7{<*!!nWNKk1U3r$gPIQmuDvu|7TPt?HW^^C?Te_`j|x+rpF7
z)RV{C=w;J+3|Jr9$_n*)OXoRyXLZrs@^H&*w@B0)^jE>c&JBy<qCE)BMA!ReD7xv9
z`rU8|sSONacCD5Y^5bskYaDd4YleAB)2gFvY*4O}V2aM9FMry(dcbUIyfncpJi6g#
z@%N23umQ0t^$`>RTWC}SvgfvZ?d*0wDo6$@Iy^PrHJ|(Jx4X?P0l$L~WZj0zOzt8E
z0Now2iK;Um4X->IHZDNw!_X48rUyI8?PK3bRgez%%fYUIDh&Y}Br*EY!47@5g6IH*
zU&*_Owb>GGRh=mTy}qhj-4r7X!vHUSwxGbc;CfFwEP14HQW~@m{LHaWve_a0<Axl~
zjD6UgN}8Jl?Fr2ZKXQSfLQ=+{q$YAfUF<xc24t+7K#{k#k<SJ?KvpOX*FGwCP<wK7
zd0;z~BXQ7LM{iD)n?EZQn|Nsuq5ARD)EayG;pwrZCRnVu=bza3*R7F2Z^l=<6$2$E
z>xUyV&a=eVofmILk3GcU2`c;Q*Il9D@7nio+kb{N*VA_5gKyQG*oc6y^N(Tn)<RCa
zSxSDm&l-2`t`3~D=jz_<{mq?aV_ozYhWDcR>N0G{d&R%8DI|FD;T*u^-lBx?_2Sou
zx2n2ewxBHC?8NmFaKSztd<dw}hG()h&`JthXiCq2_7MmQHA$Ql_BXL^QyMOT#RUwD
zb9t@I{`ir3in4xxL-{_SaLWc*Js~{?#PxG*^mBA;bAfD9*%@wDPF?7YDmyfw|NLFT
zOmyag2Fo#N&tS$yZPk!>nJ~Z>Z+M&DlL(ky-Sk6K{3U)H$hI!gLd_$+sUJfURW3yN
zlf|wS=^3Beu2|D6dPyB`{;0Sd{u$lP=*!k-^-@>&0I`semF0ab%A{P6&f8LX^cIve
zauj~p(#O;<De>$q%;pI)CK&X0l6<#jOjBC)*R^X+K`ffHbrFzEGA&Q}rW+8pLk+a9
zWPw;9H{y|GLir34hysGiWg&}%{W#9frT^+c3Q65y-B?&{7NyRi2Miq@0+Xh&yrT4^
zsle^$sK9kk@_|ge(k7qFUBFDNq^}5H`lD|Ez9Tk{9&<T}yq_q+64A!pWSl2rO&;kE
zjDV6Zi=BaF1JQ}R6aRq$ZPs{*i~|gEN5Jw5SV>t(G8aK~3iSof_<{WS!#eHoO_Q+B
z$X|1Y44wNjv6up)hCsy^F0S^FS&XX6?ac;2aWGPIWcR#1C~RH}@d8<Nihf`gdXQuP
z1f`oS!?7<?WFK`uS?6y_TaIFH&eJ-;+sDV({dxM={aY^E8kWPX`Zh~q2oTn8oX#Ob
z8@z<$=9mF&pnZmoWxMTd4U)DIe`yUt0ch(a6@oSyiZ;!j{nwY!+%Ee0HS?YC2(~ok
zix7cl=F_N3tEC_z-Fb$BFsZ1QU-N!kVF%t^_<Z7JC^__`un<xIz$fz6C7{`Z6K@m!
z5MMqsYqIL(@Zi0a6IUKGkB)%X^3^AFxmwWWmr1<RRHXk^)I15~<JKKA&UJRk95H4Z
z-!KEioi!|M^7H(;y;L+~^a<w@tY~B#60D1t@T_<CSpykh1Hp#M5DDPd(@1<4VY+TD
zGC>@mlP?1_IFp^x9tHbLD79M7@@-iFS0unGJ~kci(Th^K^-SzflI=HQnRWetrMu$h
z;P~GY$p42RW9DZ2@9D0VwPoy&IN*9tHC_=#(W#XYiLi)RrC=Uc#O%Zr#M~9U&i;M>
zqSk7pU!Unp7W2C0GyQ4dQt*~s7;F<p>=nS<CG_=VW#c-=#zVL<fQ-k~77&&xI_?Li
zvNrVi7Wlj?X4m$Ny^miR89O{`OS2c?!?|4m)g<r)b?|TjTQC+P!0Ww}%T@XG@rb@T
zSQDj<ac^iN@G>clwr@YBR`Foqe{#Izc=nEMMmWdc_?0*^M&-l9XFP{i=(2#Q>y6;o
z^ZlIPMA8hy@h{rox_9*@%#T33g@|{U7|zefFPw4ApM?NYL>0{K_3q{NY5yhlYDji6
z{VwR~_Hc8q*?T+{TAQw@#;leORC`J+s6{yA^>;H|x8*boFA99pRHkSZ)H0o)=R>_z
zLi1X>P9$E%o@9*J0~vsHzi%|5vXILDO>(hZ+Y|>CQeuL~k~_clC=1|K)EAP%$2Sw1
z_i{9mK9(sd$)YhC87%F;=>)l!j6PmTbYff6K#EJrM<E;Q)PQ{wN2amdxMbA8>tz;f
zk`RSUmIt}2Wr|%j_Le(VN$rGHAsQLsv=r}w@xb%@_iRUTU0dn%gw>|>9MKI;<R03<
zZ~5^X;YT0%V_drEYODdR34FSmm_{5g?T*sdf)a&1lXLArg49~-z-x0WgmSu!rsSJ8
zHr^V%=N~_pn7fZ^e!jnY<Fyz#bi4Y}+9QLO*p#?2M!kq`G*%$p2y-AAWA*4axMRZ&
zK0yXImBqDk;Sjv}SLv=;W>@a4jF(3<k2xpxLCy%LOLWa^@<-}j*%a-+lh4hxJ2Z@~
z?J;Vx#V(xvV+1mpttMmi`w43emwbcB<XaYVb8<jZXI@$YS@b7DUE0*>&Maijy+HOM
zDF);FbswrGc+=oLS5`D{{X%hIkc@|G4hgt2jzgF|0+5aP7+nGC=Goddd*ZK>Uka;h
zirLbci%$%RF=SpnC4Z8lF)-I>%DM6nMC6+mrED3CY>TX;8I*d_M~(FB5C@?mZ8W7m
z8*>C?Di0mZf;NnXTnzW~c{Q_J(6DODhOHy{>3&S~7I)_2a+VQR>M$OBGj6)$tT<d(
z#D<Ikous;n(245P5v(c2lxe3FX=e>z+~2KlAL)d>RL3ehesx6SdFK)F<lW`{jWkd2
z$1V-;i6`<IOlf;OwsNW_kMOT9`rtMvMF#FLo@HZ-tL;Ui^v0Gac7oYnQb%p4zMz3E
zYN30Av4)3YDr=~N>|$}u1YueHx}>}M8}`EBQgm56<mSUv35jN2o6-|dw&k2zG%7bh
z6YnWU?nL8Ur1ObY?A51*WYbLFahD<6Xc@n*hh;nml{HPq{^6V^6Apy8l$Ulc_^9U(
z8q&eQoXSxx7Ld&(W1}VU%Crx@_vO=t=3W++t`WmpQss_=Y+!GvgBQzXbeINZi_Zvc
zNn$kqqN4YZdY|i$-k^q@TZ~2@wo*TJr)+a^ypvyjngI<`i#-*&;Npq18c#+-yLWd4
z9mR_sAOYiNe0Sg>jG)Ma*FqLBBPkmm^t`1;MQHbn2(dF)kKT2pFCsl2j(Y(WmYLtd
zw77?{E!R(PVlP?%9x%<|B~MC#3Dl8tV<;N%C~rZza--8J>N@-uOl_0@oWXA-8L7E!
z2k6=mIPrA_A4$*J(Wd62k`UeO{t00eoJPpD#ZKT5TRb_4F8SjV4x|u8(`A#^Ayo=6
zabgo_e8F<0h~oqYs#;-O#L+zm?>5Ce)9DsX({A-*%ra>@Xsa<CE8i<PR#(TyQQ^=j
zn8N3{W+o7;vrW{l`osIY71>=RhoT?XzFPXc|In-uKtAb8X!Y%l;h1oeap>5tWi1e+
zUxQLUkaH_3@kJe;3j;H8?2in3tSr*+w&ZQs01KCK0bg24v8WY&6>2vvJ*ufVd)g5c
z9rYNc%*&0bEzccg^8`ywg<tx!XPd`O6k$?~UiOa_vSO}pGZod0#;n8)H!bA~2I35q
z+PqlS3Xd^fVbi)}SRJSLaQL(*%8yM(7}kLE`c$5;_U>O+HbXhmLcu&jBwzWP=tGha
z*a%znGebr8AT&5}Abdl`k$&*d`gLki_z{P6pqB3YRW|BUX}oM6`Uqu*x6s%e{i#m_
zeBMcAOl5tK?HN7bA!FZ2pGhU3hS*Y`^0x^Zr{4K21KT3KgX0u&MaZ6NNF4U=@Msh5
zQyH6+HTAyjUeYqzUnCsQmz5Os;u&a`(4hU_*Bu-b$^vH`5BkdMZ|%<=S>$H%8=H3h
zR5JTzx^AFO6GF4yQ=td%(tS_txh*hst;sLX><=k_O54^uG9B=mlsv9mtw6>mbX`xV
zwVg%u$wJ1c9BZ_Dbn_oFt*5BU(#NBm$KOB(TTez%!$UXz4#&RHU7|jg$5czdG(BlL
zcQ-L)y0>Y8-~EDrdH~pZjvRP3fNSQwpwa5~D(nYo)8g&??btVCqbbr*f!I)upI2w@
zObL7!kq!iQejN|5%y5HHmHYfDWC<z5eg0Feao-%+wd2b=CnZb{RSq54o0D%ru3!!6
zv->r?mvR(Xlib`{r2({aW)*QWjkW7719Qh18`ReFUu+R^bNwG%ME@@%&cyWJW5kzn
z<?L}g?`jPix}SjA2lqLFf0dW40Rv?Abmo5{vz-%Dzw0A@edG{x>n6&q(@J;>fl@6v
zaSy&7cqgPzu8J`bB9xsDc2e46Fb0CxrJgdv?!iWU2Mq~ZU?LJ_haSIu+^>55oOaJb
z3;wvy?+)+W`|Q2C7LgnvG(Y!n{SR|E{Yi0di|*S11G5>vgg0mS{S>;;JBv0M7FB^-
z{3#iyf%^d#6Dkj><^o8yCV;XnW~Cxz>D)MX8fnxBo@~h6o;W|Rf0M!(!jj1NU&VEI
z=XY<nmqut&Kix;?X|ubCk3Vj49x89!;3T8BYr@I2!BJ#eJy{ajf3p+EZFQ4O_Z5~D
z(o4}UGaK@(L`r(*jtnqzb&cjzUexqX7K#H9?Hp8U;P`1bxgfI4I1xOpq~cLdjp5^k
z(s0JD<>;tQTQQTf80^M(oP|W(FDxwpt#26{H`1|_P)gijaU*mX$=xoPhZNtisl1kT
z5*=xjWscs{H?~j=@z6w}1n7_OUkt2?r2;ini?0=Ym?t@WJKWI2S2Uu?&90B?+n8IT
zvBwEtV9!el+>{NNp|(bQ*%75aiEIb3YbwA|$89vl2`!^GPf!%cDhtXfT_QmvH8;_i
z;vY1g?IKB3mJ*3nXIku{76UBv5yWYuK$>0=XK84XDlZdMj5)`BRe2~1V%)10E+$Pk
zxo~vqqBpi@FL-;c0rJv*pNouOkwoSQg<uO?L7)3+zF}Q|q4b|XCSYLiQuPh{#%QdM
zzak+$CTAc?t2?{jYMz}~A4s)S4F(R<dHM)7>SCn1itNjGkPr2rHVBpL;M2vy9{j2e
zc;|2EI(kzTwM>S|2FT5s6_oCM(Pc1@<5~+!rc`KqiCFE&>9{CVH9vlnm1IPm=QI|&
zAGxQLc1fcT%C2YCR&^NHy7diu=p9d^O`XC_(lBD{51?wN7k(<4SE1P^jQsLVA}O#T
zgUt!%Szn%q9NDN<Gb4*y4{a2Xc`Ar&s0P?N+|*&~jJ4BbXbu?1NjI#wrxe=ZlE6c7
z5z3>1JMXT|RZzgbien-~E*CG$R&^r1Xe*7hj7FmHf7FCKV!ql0k$va?UFFJsG?W(5
zHqx%{Qf`uIbU6-%&AUg@Ppcum@5spH@v1{VUhhVis}WK%R!q7|R&|y#+0}R=7gl^1
zJp^)A)wC<8urdLq5?QE42g;I=$2qsPT%A=YUAN?j>SVrPnQYE|>#|$!06*wZ!wq+~
zKez%*1kB3XDjOW<me<+Cnc(Kg75F3}X^c5L601?<c#cra3e@%5rYK!{Opeisv&Y)f
zHktNWtcWbj9X2yKF)w!YQ^oQY#~obZNGd}t`Sd>7kUHaTOs8J=3%Poj6(vM5Z6Kgf
zS?51E4SSZn4Wm=EAz`gd8F43eMl{hXPd*x$SYdN+IuG{hUr%QX=UI;M2-a4BSg4}a
zrX#v_=qKW-@=<u&nC-ue7cQb&mmQXCE4Hq;q-Sgj$BR^50MnhSU4fg<v;QSIHcO(&
z7&R#^pQea%H$+q-5|L4B3bh|ZotEcg51`w(GNay{Tcl)y`=P|@^qMp1r#X$*hR+#v
z={uD1c=TuRYiuWePQ$IN6V}Hy(nCLfr<6tJp|--l+9sX0Hj){rnWN1ncNt@qbkXmw
zBj_S+X!t2ErY_J)5<q&JC=4mtt&zin`{IxSc@3d4-$&PP*Bnnz*w>;a!=DuuGrdA*
zT8B)Zm1(!wG@6&R+=|D1I?^U+$1lC>Bn^x3TT|UE_6TI2p3;zrF+|g*jD7rtuKJ*Y
zJMptiUB<;1g=?wQ&dPr+$cc{ZWTYn`i4ipyQ87NUb*6r)YFgj3p2-OJ^4?hsa+4n`
zFBg2f4D9mL{E2!}@<4Qcx)&G`fpeYm`C7Y3B|JKNyxi0o{QG>nAwt>{1S7_rVxb$V
z_nwVgi6rs~pKQCw1Jf*f1R;^Wu1_CT!gT}zT?v?-?5AQBDuS;r+O<S0!By;mclcci
z$kxQP3{8g^(3J4(w!bD)X<<n4s5G5JHKMM648MgFi+i-n^P~~^a_08dLLjW-D;n{8
zz2?K25X~ln^MJ3B^px~MCiXXW6piM79<ZFW#2*?EP_(jFMbZx>4oP07Y2sU8n)($U
z#e9q87k7vW#&nx0mUz7YgNr-^jY+CSj@Dp4U)w9_$LL^yeDUic=k7nAcg;3Me@p@;
zlG-J@(J$_*j!Bo?>3@gBqmHqk*&nXMI{F!3;+^9V#gYS~CB)4CgRyrAvL)QMaLcaR
zW!tuG+qP}nwr$(CZQIyo?eeSpZp3T6dm5*+V#R8$jLgV6$3MsX#-nE&<A+0=KL82#
zGaMj;hSq@jZO91EY+%QHEcud4_#kGxx}g6-B`M4WjaMmm5o`~(;H;61R8z;Ql#98w
zCHWdyW^wjd>7lp!Qg|{lul+ou4uVPMgT7573L0-(a42<v0OG@e2|1`l7clZedco;|
z^wF&HYMgv!6cAmVl|5Nf&@Dr8FMnb)FP?zYi-}N~;OwnByKiZ4HMZ&sX(>@XDOm+g
zCy1l7LggZGc8`T;fGE|ach9s}UcU_|ehsL@Nsb165;MTD{i=d>?_YS@+|~)}Ec$9#
zWhf5MU|$crV4H}&RB!mxm&pnZ34`b(=3W}PGYfWsQ$jM1RzZ{}ZkcS2BUaN{YZvK_
zj4!K;vXt(lS_y8~Y+57DAQNjA(AilUX4%3QxKYm`!L8|rJ10i@yK>h%q6$wLLAO);
zu0;9h7fbD`;!ie-GT3l`hfhYP7E>=ihO2hT7Zg!=BiG6u!I|0~6tVUSd%%xGPuCX0
zSBH?5*06*s&EbyTr<#)5s*=9C(HWE4>(1<2Cexn7iV<^~GCFGjp60p3`4!nFTG4T6
zEq}KvS3#<e&~i}p*Lq=025rVxVhsNR2M9FGFv_+_YQBMP@tzG_!Z3`PajCP~<y>8l
zT((zSv39>CV1mNJjq=M-S`cUvXN7u&606Q>wR*peJ!~s4+q@|TR3=6&GtZ8OxdPOH
z=-(J5dKdXIOyIgl1?NS}7_5jc`#}GV!1{dJvLDA?*L}7b1F6deg*M&MRyom}ukIA9
zJc(AUWV>$rG2Ba%H1#uAiX`mLxsTY{^xFIMiv6wa7K_9x&E)ZjAd+!TE!(c59h2m;
zmbzN2HqIQ%>2}^Wj?d_;2C=qOg_Ny>{6y1O2?nh|=fTZQiP|)?QVN@E<*sl2*5dXf
zjj?r{;{*cxR3y3^W!W0qWGwX+*V?(KN+LH&kP^|^6?s?F+OIGC$C#7266>)sRn;a$
zAuo7%R%Dkmxo7#NXy(*y$!kI%BfhuoSM~)Q;oB$bPbxgnb590w6*sl5q)@+W4)Weg
zz;{E(fhUgX+0`tXPS5nMmEZbY4G<P$i}q%<>?d`hPpl(|6>;=buP*i*n!*{h!#PE-
zrlV~Ry(c|kVn?rQe%$!HdDW56Af{B`kt?-Q{n$74tX%>;x@SL>r3E|+O*UdaAx!p@
zHhwF*xuv0A9i&8LN^%AZZx{SXKbxx~O!6m(d&;QbV$5!k!rt&2^BSJ)c^`pGeoZq~
zq$@#=OL&(?QsrdoYy|BXRalv*5qa3Ou*ywiX{D)6uS01OYC58-Biq)O10is<#m<An
zVYt19zK+R`v-zHAb)dU$@Z14pI~-xRuiYSY<H4DSY$RAz+P9Z>76`BH3XZkK&!<@d
z(1A$eja^4ig41tPyrO1o4<9&WV2!Ft35+r)GDK^#pB$zd%LdcEiG?C)#B*CSDaR@J
z`Fa>+mhYIDJ_3F<Yb9Po__+=T_w5$l^Z6b-MMU2BM5~eSgp|2(^kH{zaRUZ@%vrhy
zgF@BBZnyD^GR@E`zYlTU@vTW|ipRlR_=V`Sw8);J88_e-{FVxYe#d<}?fRmZ=Y0p0
zY6GN>ZQWTyXz$#UO+3bTOxw0R_~T__ZdN)=iCQx5?T@)5;CAt%on%bT!0}9U2E!MM
zd~=_aq4XDp*A$=nFNTz!<_4iJ5#=?>jbucL;Ll$9&<~&`LC(N`k^;ZDlmC-#&%ng^
zUpZfQeh>S$S>b=3uk(2HLU4Qj4d#FUM+qGKDglG#8v@$Qr}@6UI7A5vp=+&;W)}Q9
zVOTTvKeJJZJ62>}aamnuGhJdfoOFRIxt(QcvIil<ek2U%)@Lb*vQ;-1H6L%Pl8)^+
z_WxsgJ#oOhvub101m!1>g0@TE+;uOaFGQim^u0h>^Xc5%Eq1f8jjPJLMQX;*r!XoV
zgYn2?=%$N*uzg?^=#^`4{o-=m@@2tnI{tbI{q?-oz}Z8kt89NiJROOe{^|goP8q&m
zIEuxi$<WAQGcs|7m+w|FnkZEPunTEqzVqvI+l|R%gi%d1aVgAC_21rJSL@@l>U8n~
z^o#5krV(;Tn~d2Ga%WE#@=@>j#Xt^VKrDJ2d}YOGi8BGfE2RikId@i|!CTxQ&0<S{
zDw;CM`P3L-7AvvI(bJUR4Jb-hJmF$6Ek$xtN$6*+Jkvkw#L0h+6`OjbE`#53!Mer5
zAM&wdew%cpV|RH`7$fvm8<$jslRaUQtO9zAffK#CvETm#x-nA`&#keBVkDl9DM7rG
zO0#trpSxpELO~Lv(Nb>3NW2fu{$k|p(`Zeu9#B>AT;JXJ>ji<`OCd{9^LXcMZk*u8
zl-mEeM~+^nSQb~<5SeARFz}%Y8^D0uR^N&F@!V`msPIY7Da@hgl|EU%MfR%p>CeTj
zKRw2)hkrIGwAfzAukdcm5BUUBoObgmR}gBa@TTE<;)YV2#p!Sn72e@`pu+*+bLC-%
zKZlTd@45;LA+$2w6q58vo45QVX|UP-dV7BudaPSl-nr6)L|)*JS7C6rdrJt0AsD@J
z@@18v;ItVgZ(+)D5X6G?XKaAucEb|F%~}neI6wOzGjXUnV@I3-FCZ#VQbA*AumkqY
z6R#ZjRV8IaKtyBoQ*!-r#|6UURP|GO?S&KINcCRVj8xlOa)~9*8wldDKRf3qjf4th
z5qD#Jgz*$SM?iU<bb{nleLqf-lCPNjK_C-6D(Hbhjs#tX=1E7i^gmjmeii!Uf=*`E
zULm+>wP<CcmO)W#Gn#NuBo&R$py08Ovo`KqlqOzTUvO?VD7XRjwVieJ;6YumTHdjG
z1BSO>4gu%wKV;bVoWEAabjhi%=t4k=d7Tr9V7s=Wr@S~c@W;idnOxNHH|S)6clM?O
zY65@91EK$5bV3~ZRyp66Br*$1h<~`Na_7-OI@}wwigg_0dGSebOwp?pg35VzQV~Ip
zePdH8R4;p%nzt1@z8GkaLYjN&l}oRgR47`SAK9v_Jl|>+ZuM~R&&Fdq6l(gLSq^i=
z;bIIm3Oe&+Gp8EKKaO`AhZOauRoge04?z8#pXzgHZc7h}Z14f1cP~keB&~$^DHD~t
zW~>=wl_tE<3u-jsrQh5%p(u+ZHMLNHv8)VzAHwa5=dV{@O<~e#aBZtVOnklaYNl})
zVBlPZhQJ0(i#rLnLJ^h?SxO^F9`5XgMQki3+P7Y|Cd_fG{ceB|q!9=^_ttl!dF!0Y
z>%)C??6H;Jqy$aX#)G$TZr)Xtlq+<yF{(xqcDU%@VV0Cf(2BI9NRfv&&>+jwWo08x
zB8*K>mNtwtlBXeutxADTy^~Y`x&&!ocb4!Jc^X*E)%&C^?y4iQma|+Z8dw9bS+Hg_
z8qYPbmV8po07jFKF%&LuTAetVd$_qO1sXj}IR~Lmj-}c*1TI1QK`O<?5NjOxVU8-p
zwi9s>icB=sj>8|8maPWQ-QvsZ-35;_r*c6QD^%mtI>yf4;-VJ?6Jg3YYOYmzT=x~N
z+F5?0XoDQ~rJ2g0F^skARxX|4U%KWpW4}~Un5OJ{2xceF_ZroWegI~oW~QhdNL>sS
zJSiIG$yJUeUhU3Zro10RdnCEo^MU1K+lNX^%QQ7)H{_&nTo{9d$5T5wD!S7<$4o;)
zoqwHro{rJirgJb%gENwW`IJmcSQS4m>Tx(3D$`A8qqp8Bav0b_V$|jYFjg)BhZ9X6
zie!j_<iN9^9(9}A@bSq4ElmV$YmRp(oYxXTngs_Bqd##t>%g~T%IL=PyB7mN?86+@
zf(&xGE9~9%(B)D$pG~M2O!ywXbf@A<Os(gJ`GE1+%xPiK$ZS0h0ocv`DeC6QBa_>D
z1P;*f@>L)DY2c9yrH^SlLlgb!y7n+qy_iK3^mlt1zwr*}E$8AR=qTc{o1U2QxP2Cx
z@o2QiWWe6Yrkg)wKD+Kd>1c|9q@X~<QUrIp>akG=0w%_){_4Em{T<RC{;p@qy=R`@
zbUzbo-NCMU%I}W~Mf2uiB-vNyG5)pwN<v0PC1Ky++k07q+x;Z~O^hV%$!(Pyu2L_u
zvUJ`Q$B1WoQd++p5QRmc+_F24e;EHH3=xCiIvn$dw-u;v7&`_k$zi|WJlmg*N;a3b
z%BjsfS)Z3XfuJBnjeCDMfrySPzsOS6YORJj1)JRb$tA(}iCJg0c6e+8VOj}>GRq~x
zE^B(6Lj*b~d#jX@Fs5eIE+G{C4)rFgRYzcorcxb-c?^h57tpI9RK4>dsH9SJjJ^y(
zVHkSRp)@Q;UO%>Ch2j`>nMFijMfv?01vZ0(c@Jy?iM9CA5EObn{QY+S&=h>cK4}I?
z#h5&Y*gl}+o7yj9^rHcEcK+D}B0|i6t>>ERzaHgnH)CWU>(Js{=EiW@Duey<Z#8L#
z*3BOsiR%*_bU0Nxd#>Q*4G+*sn#)xyyl^F_d&>#g@m;M;8zc^m!?b2;EPEz1MdKS4
zy(DS1;NY&&E`WfgWN4d%#(UW7U(rstK}z+lj7OeWAl}{u5%0rEpSbf6;N1t)`2X!t
zSn25*|8IxF%*Om*btoslyE#{_aJzGL<9O_xo(g~V@%yD`i}~GN^<B|)beeTs7QTN(
zD!^?r8j-dfO3}^-lZF!|A_*xH^Tcv}`6ia+=<M+5^y)@)yCL$sd3`=z71{Cc>k{3U
z1}3OTvK7kW_+)>+NK0JZdk#!Wn8E1&XE$f?ElL5<jJzLOESb~qUiAQ{=x*^uN$<Y1
z`1Xv>f|@7G;^gfA-ObUVxU$Oes6f`WDD}1yU7B7KzRDhJ<}$N@;U%FvqvMpsN^)kG
zBQSaP{60B4L(@PRP`mC%X=l}R(JT?MMXd-ZN{Q`56pKm*R0ArA75*IFd&wM)O!j{W
zDkB4^2&8okKTjmTI3jzw`Z_w9ljRSWEQgo<`8XnhRWAyL#B`H#->&ZS<@tJfz>{sM
zRe_3#OXkCu^ZW8Obn{gn3P*$$;smE>fFW7X{5uhVib+^8?pr6-mwHlGV#T8Hux2IS
zgj^$x1T2ei4Y3lG?5$ASqpGq&O!hT#D0lN|-n#U^L66yc-rLS+D0w!V9FZ3i+9jTK
zj5}%<IzE^b(JEL^5t@(Ux*185(Tr;oIe1jgyeE}2=IsDPX3#+db5^wB<y@8$QuJj!
z&YEORLfiszj1W~bW>*fAEEfrbIW8i=8*e%;Uge%WFaidubqK=Ep-O#qYsO8&UE3=f
zTUk61U;OS*ptJ)Y)!mz*0cd$WVRdUj%{Q(|x7(<ZvhKGWBA_1>VDRoP$aOlw8^SDL
z0cI*<E3=l{;xy5JX^s_H;7*ypC%nnrHq|x0zzfk<HzI0!5<m}n<64#b4sw_6-p<ud
z{>g>eHi(dfIY=En%ibNO8(?N}b@u8KWC!vi<YY*x0?@-r3=z4#uyO~+%#5+V8DX->
z7~8YlB#6BoM934W;M(LtaR7=t?nxEVGxh&tAw{-dm0-&tBxTmzm_Do4j){x_-=~)O
z)ahtcJ2Owr4YZSiVIWy70TD9eOwLc%LuQbKDJ&~koTZ9IrV2CDwK_x!FRF@=>fo}6
zBDvNi<B$L?23!<Hzv)Gsdx!kmRo0mP=N%hXH)E|)doW>|_V|4OnubWw0$2z?cSI>v
zcVRFiqg|#%>gPt()~Pfh$Qd<nKAs<j{-&8X-ih;ztZ=hWnKkw+EU%l+!gVnY8%}(2
z(9IN7pb-zb=v_48)whbQqPb4N^kwnlJ;RWw5(TcInXfB(q}S4IkvSCGGj$ZeJn*E#
z#)7VSRBqB*Td-nmF$@}ZS@GC$_nn2|oMWRPYs?W{+D%U9j7T_{Mn62BNjGdkWEGap
zxcIQC$9$;KTGXT$l<Jb*{0$*_wPY7bBL=ew9$v~;YIFJXDYIeE6wGiRYwfgZZ8ZG3
z$Kt^^ah>^jZZj>URxJXkZ+OmO^X#KXP!@#JZml|Ip`6ZopajKn#q;o#g=l(+9~SK4
zSSCYd893cB+y(e(yP+NhCz=7|+tu?5eE0&gqCXg-jp6p}!yMAk9bSx&+?I@Xor5J3
z(<&>x?L0a%jM+}keL`%OvP#}?s#yf|7f{10aEAP9%x#lcN+oAnyxPh8ZqPv2s^PBf
zeO^_NUhr+4g02Oz5Rn#y*Q+Cif3r_3+?8s?-uSEN*{g=BY<M;G*5<;OiIiHDP&29p
zDCm<|w7SjOa*v*}@#R{y;&I@<ec4?b3L2*`leqRl<OI|-6({79X2a^(8>pMHvcj$>
zDn2vX;&D!RTG1sweEga~p|C`Bj>4_w4wwL}N}CSnOdl*Gs>Cd)r*;d-=stfXyRbMm
zgl**OA&>g%cy>FWpB(Ta)q-GYLNhp(ai^mMR9-?ZJ3+5a-`SyGxK7LdIdSSGWUW-L
z?d}Z~A&1ZUB{0D<8)r-uahY6LqR(oQSskgtlWDHVYfDMGH-vP?o>A6Ju0&DDYDAOr
zI4ecF@~=Zv^(eb^9`M<2$H}+2IlV9FR*jwFdWZ{Ij}nQ4S*B)vY_!ZqYxa2OgG=0P
zs<+8*^ZpcrQ^jeQfqqt7NMhy%q51>+Y-lkJuU7B%K32#=&cm@c_@=qD(jT5ZP_@p#
zNfaaiRU(8~7b1m>w&8|Y!s2f_lvnGn+^f?p6b+Z=vtp20v2y2YE*z|5i4<iZxmN2g
zW#9@71W<LMhi%LUs-2S#RuVL*e{%ESHaxB(0OHhYM?DUV+KU=kYF<QYRwcLnhuAsw
z#upV5c<gGqGQF~f5FESC3uz&)a$tAJFw=?zmG~lB6I<<t@D6p62K8}KTX^AROp(tv
zflhtUSVhME9L0w-VvHRwKwEXH+#KjYOV!qWwWF{2db2=ZXwBSj{2iko)dZ_qD_d0%
z-m@_1t!e{)q^5Xq;3P$MDT&QNQFP%lJP~Tw<(w#Dve7aX53by$9!a=DN84}Yn0Cod
z8Iac}ThT1v?rn>yE~uIBMsJFugCH>)6-|JyvqysBSPIw+Q?yN!ECQd}=d9T0uQ+X3
z+JYz<9U9}>A>ad5Xq4f+COZe}fmzW{i!zrVh=PDsG@!v$Hpw?LUC3u#EHDo!5@73?
z?M`Y4ztvB^Q-7lGVF`!GS-+)H0N<REQI8){582K@_F905<`&h86)pGAdspfYUyh3~
zf_oyS4|L*yi`Gx(!-&@3&I1vprv>*%(6$;2K!|<_&<82<29BP`oO6Pe10X^_;nD}m
z`U0UBEKBMg(83cULT{0=7be3v1+w)13Ee%fu#2;<1E_41tH52qD2)toN>BL6jZ;Yv
zRM=c7DY)S(YMyUclq^FEq{{8_Z7O?}9Ego-_|zl>r7aCS;zXDjNrwGHs1`lBlt4xM
z?0Ouj^bh%15)?*|fETdCFged(n*k0SUoH9&cpRndiTOB^X){#Y1=<(A;}X>uspGH-
zEqYm`;KwhZs>0pPzUtywF0`g}I?rFTA!qZCCIbu2I%xik2hT<Jl2k5*PsL0GI8rhJ
zuUtn(iEAbzDiKB&Pq9s%NH5Q+s+=zEnBb)R(2(;(u5^eXXwfm+dg_h#*sFZBlo*k>
zHjUC3oj*o~X#ADEy|@NlG~=H@AxaNgY?!H2-{|)jXLZ;0(hI^=>eN;XPOm>8nm-9a
zN}9hR=STp(XH!IDywu5pbV~z4>g6%%K3B1qgM6+20~)nj2PO@}N|?bBcYll$B@JwC
zwv!knNMh=D{eVa@81&>1^dclYm!AXtjiqiN&5fmeL6JEUO!?qA78Z{1kBt%95XN0E
z3GR+0K@nfr74{qYj|3!3x=mw#U6Jeu+lMhtG=48{-}et}^x6+W|FQ4;zaTS4*8lQk
zY6NQ{mWu6WSMLDNTx407f5`vZ_7`M^`_7uaq@Sn+m-NFIoLJtYy7hP}Ppd_efb3Rj
z?)!c;EZ5ho)3@i_A?4Ee3+6q1`uc#p100jvXE%SE0HP;4c=|fK$ivHq4q8K@8|g@P
z*Z91-tDU<=#GVC+`Azv{@6!ThCj_KJxIK(roAKl8`!Msh@kd$J;>#M_2*i)#n(0Oe
zmdc!{TT$YBJT%A|VltQwz>KK+cl}9AT1zU2GYcFHv-a56>*sYe3^7Z3)c3CSb9MD`
z-Dk(66%7OiCj0aKIKwB8nC^T?jN_=7^Zh<|@!ID1B+o=kBX@FC$0xfp{s%D{7#W65
z^h%Vz9~jPvl=O=TaeBC8aE8LTRFr^XyM58QZ>(!H=1QyW-o_AYdXphSi*OpT!9r}E
zHv)>!bnZ6P>j0$t5VAWHFfaZrmBIo^m(KUZEoKsk@_TiS{3blj@6GP8D6{O!!=g41
zklRm5*P{n<DZJ?<Lt6<8^XsH`XM%x8gXm0pa^EL0!zWYSm4F~)oSv_AiOIM`IhOS>
zqlSPXLp6fER3@{6PUQSDbFJ#ro6tl;vQRu&S3^4&G&pW)i9f~fvSz|aMJYgLo^}PZ
zkfq!oa(e55eN(0!fbwxHL6B^kaUTjB$fLx^jCNv28;S{`@U9XupGxsq3;R#x5eI?I
zA$Bc6_FB_19?!0b?MPXKR>COOOj>iPJl;R_uXRScqA(R%AK^Ic7uNy}r3F0S*M|hW
zVvWMUl=ehlv|3`_4sOQ^?TH=?m554i%o(1q$p`2~ykZ-gC0EHsIBhttIua&3tQDs$
z)fy?@Il{2=-dRNin19iwKFo;bA-sFxyomk+SZrDe{zr-b3Ot+xXT&2xko2NJG>iO_
zLqhV3M_&!sfiU}8(6x}q^y|RGl^!yO2~RGh@B;pt7QgIh8dlr$4Dag=t<0yWVFWFx
zJpDc??7mg$4f8^X-ney1F!C5(8=J^PFbZQM5Qj2sB1mr{+oxd`$iLX3upP>C!x@C|
zSPC=n+N%v6XcBZ&%cyPHP4r$A5*Z>sU!7(7f`AAO7pRwVH;NMa*aH@R4?vGqdaG*)
zj(?Fw(|W~wV@(5+0z9-k5{z6i@O)<_TpXd)c&r|DfFvLp@uW{12!dakPEi`hSOwVh
zVlxqk@d?6u*xbybD(GLHu~ZC%)1qpTTJyr{?C4@<L`dM%LXx?nFb^03_AcDd)yuZg
zEQ1l{Wd(xw%A*O$c#xB>*;T7C$iX@fp3X#IquA8vm-y_$!4)s{ZpRmzr>88~IS1{O
zTONm?DCD`6b%4!L!dEqu1CzLDj+CB~h-Eo*$I4H}WwigSA=@4W<alcIKx;dh8||81
z1Uj!a;vRYjIy3&71&|e=%}?e$^&HX8TsL;dvqZ+T$eLB~MMjU~?qE!dIRin4otZhJ
z$DK+98s!e#<dJ#Y^2*#PA(EI-ck;gu3ReV?FW-ZR$iOV~Zz%>A(lD7@>{#b{kJlZw
zQ_NiEyA)IuE4JfK_lw}xkfM&q$sc%Y)V(M&05Z)!HzfoPS)iT4kaF$oh$priN)AoA
z(SgvLl=dh?%v8(Qt@DuN@aQtdr6EqD?W?Gx%LJhY<cwh<ji)i?C{FrNl2z+#hO1fc
z0v;j0Fz#4{u2J17j<zRVBs1{FsW}cc<H}6W9H#8b1f?gn_BoPv;y|+k2MEqas*MJK
z=|f}?;wXjyCxz)a0lQOB$|r2erUIidblEk&qunX$hs?#%BWai-aE~J$C)QxRH3S{d
zs1mUKVLrnMn+QYC55Y8#^?LO#^zoV9X#1C20jk7*j_=D0=wOIcTFfK-JNYM%cGm3I
zt25gsi+)cL-4ggaMf%vj(i+dzB=T2jlZ#Ss`Nz{Z25Vwj3x*LVe#-}2oI(v+gv^3R
zKu>Z`^!=#@*`dLD*X<KH6%xJJ<loP5mIPNp$fnS)^#?%ZS~u=g18{>f$dml?61bs+
zpV#<Nz&fGLu>u!_GCC|*KezNF;zgf$6dvE%HGc2n!?|_)&|S2tk%)E21;K;!q^(JG
z!~xO#X7n;IRxn+<7e8m-HZm9pd_X`aSf=QztQ3X&v-(R%Hgb9BOf8;>XHXGZOIU`W
z#YcZi2r*wS=xb9RpULcGtXoZ{(y0JvTpnZDW%tnFisk6~T<<LU0TNdpbJp_TS+~w4
zYeJl00%bD^Sgv7Dr8G8<M$W`8MFR&Nfb^!G!9){g#Wfp7&!duW*U*jq>6RN>B?|gd
zc&sH&mDmcVU94wO=`kxV-~;Wtt+#y#JT<%sN}3u>cdLu_Jy?&gL`;ckw_|4Ui<D8C
zZrtQgwuy1vsB?c8&SYR`5yi#^FA{9!vd}Ou7wYTENwIjaGu5zo6hTEvm7K&a6%p^J
z_=2tBQL74D);?bQas2ux^z^Z_KWvK$!gH6K+qn6FQ+PJdx|qQxw#4neN!7TXNz`8k
zjH(7b5UYEwW>Poj9XwB?I-YUl+x$a5qiopznbhq;oOrv%b%MznI}1pxd5!MM+*K=~
zGq)9z@07yq-BKTenr%Vohxxk~P>5r(ZctE63~IV|U&FW}1r2Zk7iF*HP$n7{F))DY
zEv2xOIB~?we_C;Q`Ll&#59Z~|(8{l3I}fJ?+x}HJLkBeoGCOC*+V+UIUFe}7lfC@3
z5p_GQbyI+MX(4WLVdk~>c!hvt={D@50VzFau)^0TV%Sc0)hrXpwj<E_K3^vvR!ut6
zE;piKDVVGvX3a@WTD>QCq^+)GyOC|Vb!o{90HX$3);<U6Rz5yd^0WZ_OZz^6<t&y3
zRpC@K3Oh91BV+bSI!MDW!+SqC@%#x9CacqyW9_r9Mmz}dZliQvI`++4aduiyvI$A<
zLXeCTyi^Y`D0Y@S{?3`s&^Q>gjnx1R&w&X7T7F99Em*wnOs@&FI}?5ZJoQkz&tGsn
zRl|3?ZYLyfseR9fYO$2+5T_JEQ<<v5uc+NB(bxDZ4`(HW;ZY-Lu=Toes>wBA#aUaK
zTSARxv&u1v0!)LsS3j`eDzccjoq^3jmZTCBQsc;Ku8RBIbmoyOM=!c-=ZeR+s)z$M
ztZ4VwXe-~!Sz3A9D+<#0eAhKt3=W0zp$_`FA>y!Wa$7&Gl)sd>up^e#DxLg#JLXK8
zkY(;lJMEC1I-?SC85ccV686%7xr<b54%*2$DWdF60-(~axN3jqdj>GGk0bn>4!G3#
zRAYUaT-1sxiGGCm(npA?t%QoB?XwuGYXvOcL!rffM4Im_qrkr7n@jjJieH{Z2b4qw
z4|BPB<X+<*You8ISwZGTo!PS{Yc9Pri|#w~l@x@cp*C|%A5mrjIiggJIBF-3mr$?|
z@&Hq|HdL7gfqyPoQ_r+mp$&i4amJ=tbf&4)(avFcY6myo%N+FBv=k*?U3(zQ3KV80
z(KYvL%-LRry0=DoBA`Y`>02AgcT)4>`}ymI4Fkvr{y%l4za-uN|Gboep7FmsjVw{?
zs@-OXxxQ8V2R=&hoNv89*dL7qF5Z!D+hC(uGVy~}fLnY%Ml9Yszn&@{$<0m3Lhp^G
z|L^m~z1L^m8-0;oMB3Kf7bhOg-w(k(@h~1<?17!T@6S&!F#F;CJ_11ZS;9Y`Lo;n#
zT(Z>u@gOz$G%cucCR~s9cW10PmOi?+?o^kkn!>zBW@U>t>aH4z+L=Ke;sXQKremHh
z&B@H7689qw=+%=`QwXP**pqmgqt4IQ&j&Er@IPC@eS!gs=p3p71k(TsKGA<!0;ZU(
z{Ej@P^)~2^*YAvcz}Dg|j{d=}l&N7j(8tZrk?EDc4Xz8vQiI$FfIA!~JHSWKU%)7L
z)VDPJvONiztKzD?(zUK`OvYqd6qBM^kTia0{l*c=W<YWy)CsPfCSjrdF04TR7-Yo&
z950Ee#PnFdP=W_Se&fT;16s5tc8XQ8C)rxc0@MO;?MWoMfpB?S@h8AA14E^GWDa%}
z<%jqZ;<h1@l<X=rLLJ9xSp?A;rV9_ay4lz)6tnMZrzunV#q)J&7O<2h70u{CnuH@d
z)22gnTB3)Q={JP6ShRT!xqMpDsV|j=G)WFMoy-`ncP(s5Cz2cLSY1wYzw9(84p)0n
zbJ_3JoE?+xF}%h!<fU<LuY^|SHT1kx4?B<j-H2udriAaYA3Q9ksY6z$tY0ulPzAh`
zRt|BV>pnv>vs~x#5Q(~r{MZkir-VeVckklK@j8EKq1W9D%E^l+zdMf(3v!nDah5)F
zw&eP|FKsTWQe(zg^kTTO#0AczEs(Tb`{jV6c^8F0i!rwEg0jY*C;9SN@OT%*#mBsj
zpIS<6S~4UtW<e1fR^ev@Y?L)8>~RfvHGE=j_Z(y$tEWd*nYt}-PpoyF<^?P?4so#G
zR5VoR;1=1foY-gzYSvzM-K#W95H<FMOeC8&uv&h>s!=X@HYuQJJ6HCinJlNUIqps5
z8d?V4D#-&exZ+!>*c5~|Fe4+YI4RN1R1wAl39RvuM(<ahNiE<}&7bB`NPlEYM4&4)
z8EfvfAFP>yN;huhD%4nRd)`(N^`!zOH-L`+TxUyHxxA<fhk0$|ZrFY)(@r2(B;y20
z^j*y3eo%=HbELsG$MhJlV1O0x1e9k`5i46SjwDowOpQd;!P0}_CdKulRxedksHT@9
z>`XAxg`a*sf7`?j-8%5$u-CQqR{@=oM<*{XB>6V6pascGNm9Jgwt2D9@piv%dv5D^
z!o0rn_Rykvft|pxWt~>4fBoBoC8Fmdl)`qhZ;wf9J-upU5|N&p*v8l-qDfq)ODN${
z`p9cc{Si6A^nabI7DgM}$oscInJYf^xb%2UHY#(S_@CY;jQ^KA#_+p-^<T|4RN{Kz
z2I*mfUwK3H{FfvAhzj8FBZsqL8wQ6z>e%BYo!hgMl6sxsb^Yyg?Eh8BJW|4CB$j6g
zdI`pY98s?zRsm?qW?*k(-G~VnmYB0M6^f6i0ym6EafoTAiFLu2mmW^H21Sy&fzf6{
zxCEiz?8lzd-?9Xl{#nVJYL|jr<MoSbjhm)oP<NbA{r(_N<tNF{bS$W${eZv5Ph<4~
zAb93%{g2oDAJ)s*7?}R6YaY>%k|Sn=@13i=!$YsD()Mi*mPcO~n2bSbBM*ZBbyP3>
zjos<Xdnnqrm74=bB@vfdT5;{r{l0WET8?78^fkKld2N4dc2fvgnGM@tIcfs(HTuVv
zo?qlPiM70XSEcqmmBMfyQ-Gsq9Npv%`pxyQqhm)Sf(sZ+*ZKW?e^$9|6)kKa)8Qnb
zlwX>i!eJcJ9sVfNVdUaG<+lGJlQ>sNCKGPDVx1Z23MTBt5J@*aYV&5g80c0VUt=Z~
z-O_YezJuJYqk=7-ML%XF6GSo+gae|zvY-8;fV?gb?3v}Wr712TT_8gsi>oBP*sr)b
z)sC6v(QO{l6Kag!;ep$7FFv8&yVrV8aiipdKbQ$@UISJi1Z)=$z_|?OAP*2v=Ht5^
zozTc|Y2M@Mx#O**-SSGAb~9?uB~tdSpivEoBQt6Kr$2(hv+=X2_px8juZR2R=?Tj`
z_r(p<64}C;UgTL57l%&_O~R*UPS;C{o?<?!AvgI*vpxWPe3;Z-5o<7ipmSZKkCXz<
z^my?vsn$GeD6>53xv-(o;(Br8?wT&S>Wq|eUM9u`g++za646|m#s!1UR};*MhF|$i
z(<b+FIcFRJY(yz}F85Eh2nw@6{k)i}FzY#m^f3vg(*BpX@aD%>@VLuigcMv6T(?t=
zWfe=F$l@Cf;VKEHF!8H`FELzD9kqqXB|5Qt?o!2NyKsSMo3Q)$$3@Gm*MY|CPS5!U
zkfzLR6^%wHYJag*ya%l4r9B4CmOQ<qq!-66xrW4k@18*rqV9u$3n|K8Ym?I-G;k1p
zE-^F^;b~3CU<t17kpqbANa%jp$ZZ8~lD^O~jZW+&EzD+&hT*1LAq=^b4rZR<szm^b
z;*1-BlC0jqx`*W>msA*f%Eizn3^EZ;9W6DN)KRTp0Gf=K+BMSF?v2bUlSERT-`=WO
zZ0W|MxLIB3Tl2ibzRg49aMxzS=<SDkL#yJDP^*+hv$0+f@Pul7#){xVhF6&~|BJx}
z7YUQewk2O_ZpJX8K_V2Sic@)0{kWALSIdWHPMuA!2T{t*Yndiom5k$-%Y@RFh^Hk$
z%qZ*nx*l3vgOnx#ma>$GA9{J4Vr)I<cYH9!0#yZfucA~?)vD9Qfl6LtYh_(gLu$YE
zwTw=kWms1tt*tYK(S<7?w*yokP9m}(4s>uST8WxbVGk17$;a)b$0NNZ3T~~8pm3VO
zLh&xAc&~CSFNSQzeJ+<_p5p$^G>W0cA}K3x3h<HFYIph2<PBVl#$4!7xlPD!4VP2v
zoxc_=Yxbz7?harfF-+)=asydhPmTN1`dy=SiFi{i_qT9^YIJd=Z-+i#H+%XTN#`pc
zD!A<Q%O(f;hw-JVh``X)Wl*|-7906z{R2PgC_@fnRY42Kv^#TIed`5=ljad%mz3;v
ztdaU)phl-Ta8z<0F9bo^!hrcBBWevGHV-TUo^KIO%k(-L$VKR=X6v*CbjJ<7u0DYt
zHYhYoj7IDyuAG)_#(+_}==IE=cET9bOcublw6=|o-yXp;f4AePrXBl*hgOJ%f&F*#
z*{$xIIEh?r*L(flOfL)n0k@AN$HH5VeMr<j=n@$gZwXkP0~D-YI@iBs^kW)HvpbXS
zM`>K(3vpuirg(^5>;Uj5X|;dnC<<Nre)L^O8vumI)o`0Lwvfjq3d{0}gA#8kI~wA>
z)0+uaqq6M3w+Q1OxK}(d56(YNCMNiTsSx0lW4apZnrW<!4O=qvQ9IkcpkhU&XewJU
zb4rv{w43prK*=0A$*~{`*Jep==bes%>iyX#J@uE%QB{y*Z7n8M)B3D~-UOaKhn(tb
zVs`aLm~CH(1;+QXgt;P7xu8I2Mj~l5{HrQ(CKZ(2qpgl@b7zK9IkvYCsJvWB4_xuS
zZM^5;=#^ZJvv})|S{>Kvt<M~|{qnX8Y{@T<X5-!|wOQj6tPq2qAgPtkmtvYl2p8{c
zi99>n0>!~xQ%aIVCS}QdSezkYqk88xhB41w;r3`Lhoffi2P|e^fMk|@ZA;N6xGE$(
ze&CoQwE@dC{fMU5j!kFYVzZ+O7HNWRg50#H@+Jno+RRpIUZeTDll~T_0ncM`FYY^o
z9)f52TUVU!N8CI=6FbyI-Z~ajBsc8zIL|HuZob^)euS?$rbW6p0dGLSX0kqc9|7A=
zB`4gSvrAZKLVAR!c6IKCzXnLX5+>1@4i|D?WLp7O+Jw#1M-rS4wUFw()tGiEIm|>3
zTHLEqqjB0lK=9qIVE?K7G5v?~$IkxW35XHRj#y$gn3)f%$qckRbK!_(Kp4z(j!l0i
zI#MQVjNKU4yKg`J%Sxfj%bUu}9N5Y~IK<Dw337@PH@}9<y+3oeVPj3BvruCCDzD!q
zz_sz8kcrT+E~xzIusa`ZH+8?1(Sx%LyktcyU+N#%Pnj`r*jO!q`=J(qG4Lq(cYCZF
zMLiouyqTJ1Xg^@TJVB1v*Tr%Cfs?RUzrG)b_ziW6Zb-f%wXfUjue-ehHXRFrQMcoo
z`N&ScUK~MCAV?#hv*i8B+aFgwIKNdvFO}l1gQJhNTFSTV2B-dY38Qe*bqb<zZL?@J
zPGyq6;bXDfriTntqw)U|=*GZ_fF<#V=$@75kHpRp&B8c!3<xz$-tjYFa+2!{;>_si
zG$T~=TSp|lO{UO&xXG*%Y82!xSW-uVZt%$?HOWX6u7FsLgI9Z3E1h3L@>3iLQ+_?5
z2p`Z2jm7(7u7(N;wSjGg6(wTLCW(>Mu#f+aJM!!H0{;>#=>ybx4=>0_^2R+SAcZeS
zi>kWU#G6Oo{WVAt?!QK)fgs892wlTLl*m_z5oY2eTZM*hy=USW?bF4=It?89^P$W2
zTGym{vHvwct*x6><zhwWo()TQrqi1*0U2=_+uMJTw#LnjxWX-p-ABw)1ogOV<jf<_
zdM}C<unMb-WVlyPI>dJJ<57~va_Llsh@>0|{gi4b7J)(HII>4O5;HHmgRG`#8~Np(
zY7;BqBafAp3!XrbbKvw{r^H4ZDc97m%3o4q-;Y!mp0g}BsNN!9Isue8bY{#zjPAlQ
z00M4)iTZwtGGQ?k4#?rAhfTlZ(f{NGPTit!fF@>r!)@p^hCnw@M%Hp<C^MST56At7
z;ebJn=`JKS43*KW)S<K5ByG*Ue_F<Rs@p7K12GW4B`6@cv)p0fr7&U)fCw%JL)Bl;
ztlG^Y&+Lv`z`rHsWr>5{re03MFB!vmUSWj0I<^6h@knc`V<EvF)pE8b`OaTW-0jN;
zc&a-fxJV7?4tGhrOw3;N4yqPe`z({hW8VV4C)CGc>ao(Q$6;iU-iZZTh_0J8riISw
zal0-*q{kq`n(!d!reNfl%V8O^^{)j&T(l4Xb6}NJFCCXyXXcR96bh1>u~;vCgBc@V
zkFj(#(JSR@`m!wCbzy45lQ?i_gK=Ygy7b=E*`gqpZe(nPFB1{l4B{<YHWcna+!Dzc
za|zlZMImY~+sO0w^_Cxv2UG}zdBdux%CUmYv4yt`Q`c?Vq9+Xd!D2ETOD#-EGGWTD
zvK;<ekT31A<0e&yf#ai6Di=Z3hD0=NP0+{sX6i^td~>zKZLip(swi`G-MUVyiLhCH
zdHWvb$KfG2Z~sy~Ke*O8vpUv0s&2HauOWx5l-$vbh$b9@QFpdJ!(yeT@WE}FfHtoY
zhN1B(lgVMD!&C?IQLc&X5(k%VdSIK$Pl~MnFFtd7tT@A3+tNv*c@1?p37er~0;17j
z+Q})w^-$mSQvK_-Vr@sb&E~#sj~w>?IazEu?)OC0dAS=#3a*}~GtO-GFB&6~9ZHDe
z*C<7{{Nf{gf>^k<A&0)i-EN%$$Mi4ue2dB~YAx<BVNHj(Wm(3bM^lTDik%ju4#`7l
zew``heu3k@l;ca`V!IiqQVuocnrG~^$S&NM&^GI(ZzcH{?yz0&I|}m>7Gr}4qR@1f
zG<KzHd$419^p)rXPNWh%e<C^ruIw9c<XEK|FNbW&F2J$0k0bJ-&49R3he{W^SS1U#
zee>H`BX*ycaz6Ai$tziGs^6dfVcIgY-cnQfc-`Q_gHUdXZ#%rg8)gTI66$O}uj=C8
z-kJ@6Ac?q5kF^ME=lQ`xjd^`ZE-HSvaAr@g)Uf?SSdSklNCj~{KqqMku58)vXJZU1
zf7aE|L79E%^(}}2cw1<zN?GN?^`toMjQNB{um~nHIS0d*ngGTAK+!%}zqz;7)EXH>
zf$ZlUN?VGC-EK`R-UsTN`BuGvbBccyoc6n)@qcd6Y6y}y`ZM?<UgbVV<wMipL1iT4
z3pv&g3-n+^;Ci;LaL{s)+It-RVW8CgK*&0z1w@_?$cWd&S_`?o{zmX=+X*vgwv<wj
z<F?TADob*+54N9UFUyALW)F;E%V|yIV4zxzvdM<7a!b;kN&&%w#ZC8NZ?fRQAc?G3
z5ltB;UH0h-5<7#5b7c3Y+2=Am487;@;%$B9pCz}Lg$g9as!a#!j0JKx;-qY?gi(I;
z^_3DdU<>v?0k3;4Np^S?47#*UvIqC06^iJqM57QtzMGJce~p6P3<e|6&;uES@*8_Y
zUksC|+nWdjtohP`n3e*mhONu=_0L$hnPXHAhBzo!%O+Rbr4(7qF==YIUl2G;_*NTC
zRI_HVvZ5bXc6o~dB!}K;Q#$&aCuwqgXR$c-CZr!X>ww<e2suXuZ_=!~2=X1m=sjQg
zo~mxwzBwchyka}veZ8WOgRtdf0WF6qYF8^RuUj7nerKwKIsd6rGymU8%}D=Wa!_V4
zCv2}+?{a^Ww1P>@Lb2e{>vP)q?PwgSJSpEjJ;Qo`_?+7%2}!0}xAVb8;6!G79mF|4
zdBzXU(th!Na%jZ^P-Q702aj9%|ALPDBp>n@f>b2Ym50t=4_?_Q2JV~=!-|EE#~$p)
z?0nosPTWHmKkw8^;ObJ9)rXwm7d4qJU>~U2AGPj3&__Lb(nP@^X!%a>#?I*I7?G$#
z*bdCxTYh}Jdp{o)CBTt4{|%T^B`Xi*?dc7ID{7)&fdd?%!18@#_V#>k3evQ1veg}P
zK!KAKWBUe!HYMye(&rc&NE$WgLLx=aZ^Dk|^IKKXGeGYXk}wQqf>wD*Uw!sT{AtK%
zhL*3&;HPvhE3CvkPb@$%F&_h<s6ic`y38A&v_5V_U=%lovw_yAc%WVcrZ(SGaGG6D
z8}6|T7D>_vtuK$-&K^W^64hW;P?e4|9U@FE;ca3-xRW;AZJBBcOt$$FJW<nzyudlj
zeza3NR$FTiVzd*H3)A2FL|ERc>~uHC9y$DVu@OzPtHhHMDWqU$opoN2izJo*m>?3d
zrYU2ubznrDVQ8~=SEBa%@<W(rV{o)?44Vt4#hj{&<QQz^j+rCpbZ+?j?~qh+=|_F`
znRzm<hR>8Wz}rCF(TJq8Po;z-Q*;?y7DOM44d>{zGWF=c$JR6!khK5BsqHjf&u0-D
zL*+!`JOKFwq<F+Tsk3Gk%!;9xR3()p2S<4BJvQH3NQ^%sbXic;pqMX@vBwb9|1NTp
z&WK6AJ3nc;E9Zd_TA2yzEqXNdg4rJAKCTASwgD-$BM{X9wW92G?BF6!)!*U=>%h+B
zk(yKF!s6moY)_poc^_xCiE>uTt>%J$hx6D?JYzNC9mC`%^|X)>!Ekyymx=-GITlAh
zcZX11lv?b<){Rl#ptEdTV2}1AWc%Im(2bgIMq5p{WGEBmm79lh<i~$q(abQ~sY>B!
z#%gxlEkp6;I;&Ehom-0}KB#%OteI}$t-ueKZ%VQ#*vlu$U>2D_l4~-2VKpj^qc6jC
z^CA|x0DaSas7b!{FPB$`r4cT{3meY@-T%8L`Cz1Q(@lV=ecuX`5;%eEutP@)8<_GS
zt<V4%$c9x|*#z5_a5GXrR}m2up)8-XX`HsyhUrIsxM<3RDto&Uod(P7XHsDz6sxsT
zA2+}_7&+62+N|13U5$5gHykgV+H%v!>BFXOMsQKO_+u$<nSTo@@p@}YQ|4;Ig_)Xz
zk&Og^GxGk;KjuqHeG9BPQCVy4P3+Aozv6gEThMUg6ZVQZVrpWVxqUJqa#Zx^w6MTE
ztGZA7sBJcuH)8Hp=m}_94b2*MwF1x9E<BA!Q!@Zpb9{>zeNZc=%m~B2Mtyzhnqb#c
zN+#uTF|vN<v@IHMYL$I6Ij@=ZP_&O#LQ<)a71UCR(5BjA7YwqSq~mxrg6hg<4O%h=
z-(9QiaY8b)HKw%4C#ld~dDy=Mh2G|cyY~Rx4)nsc?>6D6`6uz3mowJl--_>wzK`R?
z4*#B-V^{+`mQZxD*aIBI<#0FANwV-Nzj8BCM1w_#EX7DMUSK&Wyes^Sqvo_kGm8j^
zl7f;fZ$1<;RdrL2M>m~PO4OriEAKUJ#{zxdYMdgZg$whPxxeG4d>uvqoRSQzFy4eu
z?Q853#sbYRbqjj4FGy}TfMZuN-&HVZv!`WVODoEGQAH9V#aR>a*up^T4uH;@YxD0O
zX@-pMliwUPrUCKI>4@!v;mxs$Uq*Oju)}-A&K209X=S^gRB<=btU*-86vCE88x5mY
zxSV7YSnz$0aqALwO5+>4=^kF+<4eUr7un;<tm4$f%B-Rld<kbV2ZIcVmWZR4O@ZVp
zmz*4)vunTt-Y+554<A3Elql7kXQFciJUp3ScSq0z$mclZab4)9`Ucgfb`g(SWs6J`
z27{>klG;?PMmV+Yt~BXh5-GkKb-9C|Qv3ylq<gPP-<1q>jWncViH`JY(@Om*FUMqP
z-6{JCB`;(M8QbZNPA?ThsWnMkUUGn>l!+kP3sOT$N>3D6Z>J{cVhU;8LE;P=uGZo0
zJiSUS)?a7u94CzMxps>38Wiafz*Ab0!vuUKcuHnK)zv_c<MT`=?$Dff%(r_PPT+6b
z@dr(bQa*tT?_^Gvm)pmr#X-+Eo=&J@HxDSZoV=4C!6d|QIwuGlG9;IsRDns4VU*id
zD{>i8TxpT395@Yfy!DYvbFwv=qfJ>)Ie8p)*rq^rnuI0VW$KhbMHTDC+|C+`RMuo<
zLti<=_0|SwoqURyw_~;%EnkVpQAknXT=be5sanWnLOTYX^-7gCk#7?Kn+0^U6FhYg
zUKk0mYmwR^rGk0mxop@6)HzF(!GQy78ETep1ix)Gf6h#S(N&#}R<<OzflS#^+F%e(
z(eoshV3oiyTS(brV0xY8OZwk0_SLZ^uD}IJ`8J(>44Nw3fIIV7G!?G{yMQ@z7p{av
znNQijcetEbb~4WVy@~3xH<E4~M_S9Y;5PET6scCX@1CtC>TfOhjR#;PDE1UEazLle
z=87rp&%=^|)mp>Kk^}hPaV|uuvi?j_qT0Zxfkw%7OI!J^$O0V1tYH94@QRwBkM&$!
zwk?lFt06;i7&W$Q2qGaYe_F>pn-_6)g^gxtD@etVPEd)<KBz7zwG*?lU)x)zVjiK`
z`Z_k>&`xPBc>hxu!SWvlhV0D$9ddU@lPhjZ6#nL!DxDVW@Ok#;FEl#@d(I6|ag>8j
zJY~2}NBECl!Uov{0YyR~b$dQ}!jy+M6BnJNs~*?JmoIw{9&eAu-|RcE?91IgXMQ$f
zpTOd%RcAPU_5V-rmmCR$OIEPp#P)h|W*v#a=BoM^1l<5(^Kl0MgQ;1d<g2pr6<wly
z7BtI$bn2F!1Gg|K-k&=7U%@Y;P7xTRcNFpC>BINwW?o1|6Ag<4MrTdG!?y=_4;F~9
z8Un`+QPcUqir=v^Chhl~6qc~_T8Gp1(BY3p96(mXVhWPmIEcQdffXFv#^~;zx)^Wx
zPZXRY4kRfsAu(^4ds>lM)f|UiL{I@U5EW8UI+=+jKw<VhVHi(`;2z5)kO(tVLDi91
zO$(=Z9`4POCn1P7K1+NNQ8!9@ji4uzj5A=<re7&>B7*xIj<9Ih#rfnBYI;l?g~Cek
z)Kq#7LndLE4nJ(LG6D{Pgeh!y>%J((;TOH-U&J`b@6H@+RO{t{t)vxjH1*|>tf^js
z8*`L?rHehbthz6*dAT;@AZkI@Z!!BP{-7y<Fp&R*6nTz|Trm>=jzfegDt{0^uoa0e
z-tCo1h+)&S+X(jn*}-eKo}8Og)`G&0UFL1@VdYm=u#-^(lnqCM=I*RpdI;rGXT&}B
zba%1;kEk#Rv|r_A|E8~H_%3F=p%IHpuZyN_ApdTC3r8F&H^l$W-BVC5lsVT67YD|c
z38)=w(CzRy+8MweBjn&kt5CbCxjZ@#NwZZh60s_$;0-FrRG>uIZZCYkI3=Gl=3mD#
z!4N@3YEz&tkE57e^T?yu#m|JSq}ys}h<p#Qm$R-$891j}Z(e~5PV_BwTQxouF_wt)
zjX@i+YB*V}ThMDeDvYlMnN1LRu+c3S2AHF`nRs5Q!x(w=?K4GAbM&c&S-(uaTEbP8
z<De`df6jiTW6fHw)NisGA0+%2VWZDGWr+de(u1_e@#kb6heZoFpZeHmPvjr=)<R8G
z>aqZkicGZRNV!IGtJ+V(F}4nd>3JrK5$ZzGlYx{W^Oce^2(@ixa^sxZpmT>*L2<ia
zW*|Ju{4-QY+e3_?noD}dbBFui9LmyFN+p@U7hShfzAXY}&+6DogF@^uVL@Kp?FE=J
z92~Ls<slBp3nY>I{$3WKIVYo9(`I20v)A*?rkE4Z8oQR2kLr&Hw6@<d$f)87VzpzI
z%w`V7Cv|;3;c<&Z>_6c}V)~!sba=Ed1MfPAB`?kyjA3mJrO?eS54IUyOKKiV19~~R
z3*XJPaRjaUSI{@EiyEo!4eXM_jb8J<wYJi%3>|?ou{Mt{lBlpTONG-ZLfMe|dl+e+
z?3iqDR7E9rhn@6<C9(Y!Gwr;6BMPMLj3`n$?!{KIl(%I@o!}*x>zvEWyGoWhFf1z4
z?d~S!6LVCa0nEqR5c?$Cm--+e>jtY9P3c<Pq8=mWgD)24zIerFDFsC(7aZ_t{|{sD
z*qvDvW{bvl#ZJXG-mqfZwr$(CZQD-8wrx~wCpUd=k8{o#-FMvHU-lo^YtQGIYa%;;
zJyZjlz=M3m>WZYcGKT-sB|Rmxa<03&E5=y+EFn{Sok3z`x2U}d_TZ2-+mQb<FkP<P
zTCvEXf|H!~@sEl6^hrTEjz<0&Q~jxMHw!B)0vQM`a=NOdCn<}DO!394@NyHQ{y!ib
zcjCNR6cc#Gt!7tZM>M1&9b&1orBNl$NnvvuVK(+9ang~l(G~TIs|%W)o<f~N$#`Y=
zUNhhb&eRq{s(+1dG;%H8^QtD~xYNR|BD)+#anmw7)N^ZcFF)2YQusxFPwx?T!a}!^
zXNjo0<Yx)1$G~ms^4+r%_Z!@3`z^qT>Ed#aDix;9bwiS?d8leEfYIwp4=sXs_p`8?
z>DVLGK}6oUB5*&cR<kU2o?W`?yeslnJHswX9bMCV0&X75-=*UE{IU_dR#9+}b&XLL
zMk25G_iuuJJ2NZY+9SZ9vn|~#jI=Ui1@B0&&_6R2EgEg7b0*l1G3~@^W>n>WrjiJn
zk8-9ZH5PJIqfIGZ3CLjYlhr#f@Tn=m7loe7%{h`V!GqBC-Z?*RFS*>PDJZt;y&)wP
zqI7p!`^q5AE0fgAPieOui`A@fk%Z(toZov=&X1nFCm$hOuL+~&{(cEvaskOj*Se-T
zmwTHgsVb5D@?r1&-q#o#8gJ~0HjrJEppl02#v;AIf|C<0(!%_Tt$9;Af7HI(Pab0q
zV3R4O-lMFwb%axWc6b_DWn4K~e?w~oH~v_-`ovji6xE}mQ**38Pw`JEFczYs6HjQs
zxK(nplm5arhwQz}2pOE$GR#j#jQJgS@MeV*L%7DH;vEH*8Eq7vX{_@Wz{I7yV7?zC
z-O8!AYo;|n>xT}6dx1rk%EDqFX4Do!%H;6eO40Q1`?=od!g+~xp%?n!KMl36IxO+-
zYtOzPg9l$U>6q-3$Qo}Mv5mWAi~ND!#`G6Pfh;-Apd!pG0r=AN-G2weZ2uh$GyLaO
zJH3dxm6Ndpy@-{*ld+Jop{<cIFE5OvlY_CoHH_Q(l!j#N!GDIDMyVu;@%ZukNxPW%
zZ+E0=&MD4qE#P;2f6M(l3N_+-8ccuM<%lnqB@veRI32wbCGV^)o!B^AsQgdb{iz$&
zhw$|m2|M;b8b1c?j@H~Qf*A#>ft6tTAKJZ##@E%W^ga<{gvak?zZB?%JQ8Wxj-Yy8
z&3Ybls(Uf{y6<@=KYPT}+~|+p5$E^o(}uW-K3{i2X0p<MknO_-dKWqhB7`ZZ>gYLA
zD3X!2UdBp$(m)8UwM(_BtqXjm*g09KeX-wf?l6c*VwmS#rC@v?kFL+351#ND;Hd(4
zVGsNb?~jF*8a+qJHwll4!${JP%mHf7f)W(Q5k4=$&JOCOmh+1I#*C1SVgEs=i$tmH
zI}6}cp~f{9GmNV97Q7^)qxXusAa$1rs5tyq-Pw@Lr(N!LCWd6ck&swM(P@xsmG)T3
z$uLc|3FRTJ<rw%O=%r9ivzxp&^zP59Bu+5hrtq=5k|Ap0-xOOw=c>@qL@S;uL=eym
z0g^pAxlwTT;;CEbm9wz5(uKiR%WfDI(EN+5WLLjEbWC9d#!TRkSS^J*aCrAO6I1F(
z&~1U}S>AZj>HbPs%+hYqjG?-;inI`WC3K;2T(m$Ax&@2z=7SA}T{M(xT{YkXh#9OV
z4w5zN=0DKeUD-wSnzg!d0>iJ`JMc!TUH+O0<2Chz^w5j32kZRYXp1{E@47WImj{z}
zP2*W~N7CTl4yEJ;LvZo#yI0EdXHWppFQ;WA0*vax18nkz`DEWidy-lz(~yv)Mv@u9
zN!2K~YXg_|-4+UCJ(A_Zb|Re?pEp~Aj~{Jm3Yw9}&(?p_T8N~#R<aRx6=4^zGLEAx
z$6iLIRt@Qm%?d{Q7V~;fd=PaPA&0<cN)^tbjCYBciN`s^kDqj*%a8kLPy9<F|0#p)
zMyP%cUBl^3B5XQiM`JRk6|69a)T};3JOwMTsbd>iuF!S^D-?{zSI|}BSKM;DdWFu>
zyq<d7eA;GNe;_uxa;H6GSGY@gpdOmrHEee#nKL_N6*j$UBs#w1aFuD$N`TO(mouhk
zaQ`}|n3ismYdH6ih?Rjw{3lPz^9zCK(l>RE_=HC^=kZq;XyYsNRkW7R{Bt0aJ6keP
zX-v~0>MdS66DBHxb@ZaqSSyxrteN}$QFWIHw;w-+(HTzlPV(@D2`bSb?Pk*gf>13O
zn@qg`*{WH8r^@GZ=|CCxFO6=GbtOQdk2q5SF4JFowims9e}5r|X$L(hX=DN+?xvHq
zSufn%*+Eer0`XWP{ERBX69O@Abd=n<sA2iqBe5xK7Y{d4_TW7Yhn4oDnEN*S0y}d1
zIWNfC79@;IBVjFt2^9PK^tL4XJ<W^OF_~i5i?erKO$91Q9CkPv5Y=+Hf*TkMoUPWc
za}uH|m{0@O1KRb(6{3Zls_(F;!xZnPPyscEq~-PNLZ0qiV5nkMGsMjz9t;166LMmU
zL`@q-oFOAAi!HHM%DZKA_!GMb^6I#WD+MUg<<6I)>E|AXyqr+CQE)#4Vy+T+aPvBW
zH~hxIKgIho1)@r1$q-6{ah{q_jLgF&+@80CERIPNrsC#N3M~CndK?LfgDdf5xtWhM
z9qB~2RiEPS1eY5~8;s47C>dQ`R1R5bb|pVt^pTyFM(CU4`NW{3W2(iKT|a7UxTnCa
zMutMACKnUIz<ejRH}RCBziHGqBc9Z~pIF~y{FLjy(K~S(K7q7A03Snl1rnl>C)yo+
zkLu$W;c#iwFT%aRoh_wC{Fkc|f$g{vZAH3}mCvY10r4no>nnHPfi>Gz&5|&`YthMT
zSBpV<SHgrp3u@St{qLmoG^@iihg>h!k=JZEmOI9`0_v95rwBsyv+U@py2q5v95Lx4
zl}Mhhae(3~AY2Mh3e@qx?}oTzL=#YJ3K0;OrC64n6R#fW)PI-NUEjzDs=c^KNTR<l
zXXWp)rb6QkN}#S*=YeREwq3=@!@JR3B}J*{*9M9}V*{Quw&v5(&OQbtsIo4gqObax
zdj+&;$_E?uzh84@A1u)Gg6|1}<*QTfYVd&Ipm-Mf(7jLM6cFqvEczjn9sZP$5tJg(
zAAQtHFzzKSv=yTSjw>gpEjm=9FymwRyhhi?-M97h#Ovf&FWEG3Q$0U;PlUWlCs-PO
znp6!*1d|vZ!;IZj-fV9++z-{6J{?#!+i}5?4D!l6?yDZf13;{zci}9vSZS1HnsnXt
zCuQq=C@&s|Q~S4}LrYK&!up|^K9gHOYR<{Tr;9tpmkSn=3sWSW6E|J+Z4AVw7o{7v
zhS6Gpjg{E3F1S?L$2g<@k%5jk+02X1SzQK=&Gf8?^XH2~kBg?rlqeEt0n!5mg>~+;
zW9xl^Edi5o{}o$V0j%u*M-%)1**L=v_}`5)%uI~`BemCmG|p(X|3~8tQXTz9=xGFM
zd8~mxVW+ySB8hce(I3qL#9!~jx63(<Cd2%1*Pdm!k1MOUZrn1Bi05`SZln>v5fPnR
z<Jxv%(8v#)B<zE4Z52k`Ao4p#LHHvhNz>%A6`JyQ6O4Y19%)JIsfM=@3W@SxG$-N7
z+i2{EjhXbppcYEYGf>UxOdX01i7<;OvJppcQcY#y;)Pm3nAlI^){BIM%O~<k#uTSD
ztA+h)Wsf+hvQTi=TBn$d^}kR@#t^4^8b^vqS4CkPH<2I`TWgWeC!3muVK-w&RS{Yu
z+uK@<s&H#%p<jX%x9+j{s*tAdP^)i71pz?Up(_OSDy|2R$YA;sf{V?IH2#t@Q*i#h
zM&c&JLLnuCwCf6&p;Q5HkT7Y1qC*>Rr9@QhM+(IjGDM*zXi*3@GGx}csy@6aqF;@g
zh-qu)X)-MZERSr@CT@*^j!~W=ou(+MN26beX5nUX^_w}u+QqobqO&rr0SbJ-#8{za
zZTD`_L{;y@h*i6d$(~hKXq9kI<&PU#GjzbEOrkWtz7RAUECQwS`>(oEsmQr~CVCNS
z62?e-xODqB4WvqdgQg&1v4LgT1okDo6-%jEp&P9sAhBSgc%7n4*QzjXH$84LA6*)C
zo!Ot{nA@tvu6}OOCAS`;OBV_3q(+>Uj=`I9%_U~`FUs0QpbpwSF;lJDs=5DOvo)((
z&>{kqxpF%m37a-@jV9mcV)v@DCwp7#7T@=y?V-UFIAu16d(r&M#fzJ*=R0+Tz}mqJ
z9h&y@#GxCLw=3uRwKj|QN;lTj$*0qW+wJ`d{r%4Nl<58T-2UHbBkmCtt{Zyr!yMX%
zz$qjmn{~vt5DvPO9e-1CLB$^(*R~Y-`}Nza&GYe)%yZ8-6CeA_r?2c!4e07X-fb7b
zj41?dx%6#IuJm-t`=^yZ+taDDuN}*q`<?O1vG7H^mOJ9L>}`p8a>=6?cMc6%#ZM-y
z3(oEu8}=@35GsT}gg<2dtI>6TKbzt6)y2T3Pd6v__lPr}YzgFNlRC;%N3Roor>=|n
z?SXYK-;QvRo3I$O3?)!+L6~hA`ckRVbH@q7LhrdxW>>Bg{bsCGvl2W2<_l$(u87at
zR@Z)kQ$!5SHjfQ6AG;eU_=+h2$uB&v6$rOcq%dei6Zp?V3ar#VG7-c-UYuRW2la&p
zYCj$Upbk0pa;TG$4L|2@aHyq2a7gK3Z-L}ppc$o46E-BhFo#?4rNE*WtKcS7difM9
zF?EIUKeb^U34lT+8$yyKZ+;3m5uk1Fp}kFWA>`tSJ9JnDJUh1`sDaEP47b3FTzShJ
zAk&(>G1e@1il(GCFxuc{fAqL0ShF!upF?222n5-E!cjskxMKd%K!!g`_=FbqpQ8J6
zLJj_*VhSIy-2Lv7Od<LBLi*i(z254$357(Hzx!b(1{1tk$NT^MR+<g_00pdWbhJcY
zy$(IR-xrQyK7AZqO1@u@$lkx6UCd~^GOXpk?##r*ii{NA=I2TwcKU33ms}aK@Z#XP
zl7D(Rac1Mnh?0|w9t>QC>h@;IysyDEPUS=oL(aU-6dp9jPsPAn^LU94!i@l4CUNMA
zcirJy_-_gAcD6$BBc#>6`0qscjTz{4xC@a2aXryB=upM}#4RQC#iKpIgPDRl3k1;X
z83w{Pd&DhvTDv6}pi#O*!1J;HK88skb;Z6VG0vo2tU3y#SOJ2O^VE&++zsS)P=5+?
z&{tA(GQ#3P7VT;R{xPc7<-`88%DgO-ZRvwo?)@%fx#OXZ@LvBjHAqr6^Vkn1ytQ20
z2GWiHMoEbUEIlgA@Bp7B`r9^J1C9_`Z{7@*AHQlX3V+CcEKGeTVcYQrmF+K1U+yeL
zUBHa0<TQ+~nqwC5*{lAZ9&6`KGConhs2DrUbHlPbIWX*OYjxh((u?wIic#JM@_~}n
z3E!a4jY8p_+_ao?IP+o~f@zuRdOW9H$apOMt)Lf09xNKpEz#~0MSETgqjx4Zo)v>n
zwM#JbAjQ5Uep)HIg7Q^KGCNvGnj&afmg=lgU1D0VwOtKf0EMj=G1$iA{5TOuTY<Lv
zR^|(P)3kP`e0<V32R`)#zNDo)jV~)gEv1&uq~A}n#7?dF5=;blu;%sMA#sgKf_Ce+
z6LqzJ@TI{$F6_oqiFM(YRSs;c|NbZ3RvAF!eGqorY(>W<7l!1wG$Q$FsOY_{DZ$6r
zVG*bxAH+!xp7T%0>g7_Vswlz7^$3ps*4hSI1-g>MB9+XXqK)eg{^7pmU_s4F3ekBq
zFFdS9bXhzdx>r=94<8r6+hx}GN+<R>_sfUVH@q(0M+kR7b!*PbWqF2Epx5q<eF7%I
zlkkj^X+Ph7MjvoB!D%P(h|iI)Sz`-~;!@rfb8ML93eQsGcK3H$QgYf*9g@YM4Gey}
z53L%^PD!FV8B!Tdnisf89Q?*xvKQ-86FC2z0>mORr}daRtJwRGdVYt_0w+DDu*$4L
zCojA+xb!R)Q2x)+Gb<#lEr3=kepsw;F0X}xN;xWvv=q*sHC`Ob7ZAn|YEvn1m>`+T
zyP$M8oEqN+E}pAGF=uA@SZQ3bh4bLUmF0CJzFMGQxaI$%Fju;;Tm&X4OL8E5ke{QJ
z;EQ<Qvg+8$j(RFh#M=JcQ-({<?T!$#2R)J*Y*CLlH`|^Qp=uy8u}`?sGMbu5Uh1|E
zw>bf{;nW*1-EcXulRB~+r<^M@{o`>a-VVTcsxQ5v%E%KFR!L8s<?wH+DS&<w7Lb*(
zu5CtK(L1Wy@C*GNPD31C6-?q$>@x;Lc*g;x4+MpM78z$kD5m_GgA|ki!p{*DO&InT
zRY)FbO+238LX8MwkI?Ep`!UYb3|>>E)NqvI00pWbR%!OTUFq>V1qu<~Rr~7JDHtD7
z<I$t-hb5^I+*9&VY72C8e^+zIzkEWbKD;j10qIBtZYz20WolMX26LZGEgt9HF#o7Y
z^EWwk9%aYoO!kyW9;baCk5J*MDWE*0eYjc~6nC3$h$u0!-Ld*UbzKr`KiZT&Ke|eO
zPiHy2+W(GSdndT&RoMTpqZb>)PY%+5u_Lkl567!+^{JQ*j-Rd;$`6pjh@>(CJtV?J
z8JG*}$adPNCNs}}`tM(%#YPe>##}LpR){hN>LsJG-%(dSP95(KZtomk9he*LXi(fP
z1Up>2kQsdZ_XUO1pcJNg?=Negj}@$(J=-r08aM-~Cp-SVXD<eXHwc71FO+?^Lz-gO
zH~4{{ox_jUnA^doUl3os#m+u_7Nx&}301kjK5s|(VH<PlP|*Rxg`-2$?v-gSM@f`O
zjuz<{yi&g$E$O|2b1EX}kY#jrGdiDR2YN$OxYbLcw4P^X1U>ZplkV$5zsOEHb$2?h
zL-{lt`a4uIcFy?3Tzh?-rp8t0gcoV@U1H3rhzO?oxj>AOBZHopCDDW`9O)?q0@eM5
zBMy@o>x<Lu4W!j`?Ie=GsWV?_2+}{tU>bEPAxB*n<V1M(Dz9QRuxvNBOok!p>*!b|
zp|ATF*ordU3=XrO(%obz=iT{yrYrQkCCmTHHeTLV>T!yJm_{TfGYZ=#5wRDV4ZhRu
z<BZ9g-J}L_L4OHFt_wCzeoFl-qSTn*J$l)9ihw%iA?m`xdB~Fsm9~14iy9w*v#xtq
zgcA`)WRA!s$iqb4m-5|BaYm6}Yd1gN(rXki!`#Qc2nAx-OCryL7JOC|QaUy9JD-WX
zv71)#DY-tp>{umEw|N;Al~ozeQQn6F6HbAcx+#Y2oWWP(hDlIW<_l|>c<f2xCVM69
ze(p)X)(;bVb;a~rvU;mNPU_I}<}xzs1F7q8!_Sjzq~5Y~xVuw=uDcEIGT`<KEmqJA
z@8h1t6E4LK+1i{#yZX3qgVmz8gRW2Iat-cm>)$$1kJoaLcQa72qENUfCy7XRh|7xq
z_xJe|OVJ|i7&R|BV!%Ywym%2&80fbW@A8}Rl1oaYAjDrVonsn2t?@(o!mQ4RLrY<=
zX7pnL=l!4WF<ebGg1s1_hI3M-ejQRx1YkVgSrIzO&ug80?GPbqY<J2w_@uy-W}w~w
zaFwX~mL)dc0+~x!jt5L%Y{sUNIzcE^Bjz?a19W9cT1Di7Zl(z>V}yJmLO$oSBL{Un
zZ7?KADxeN~QNG82IqgNu!AjnvjlX;#1A4{-o_HG(x&5?x5yxOwO`Bz<G}ngJdP0h)
zTB^=iQD7#Wo+y8>WmQ&k8`XCUo-dJ;&_z#u@YybOI&o{|3+9}PAK+Btmc!E31*=>9
z(x#!XnB7XL?ScWgQ@llcCv);HPKV9%Xx+4^0f!o(L<KythSeyjFGnu%gD+bPlB(ll
z-)kUYIc`qK+q)QYS9~PQmW_JTamXT3r9`2EkX0U0rzIGeRI2uK6V!58OkSLSrPxd)
zNVC8YMcL~{MCgY_FvH}qVp_yoiL5$A*Hum2CK8%YrYcL`wJb%~%sVpJ7N_O(ENuYV
z9^RcMqh|iyZ<9ndM@!1AsXMaipNe3hv_$CtZE(Y~;sY_=g&485zFS_f$>O`8bM(fx
zw>sSV>Jlz?Kr^0fi<YQT+X8jvIdaT`$CbuGAvwv0aPU^Z47M%zkL(F*W&I^G#ftq_
z97tzi5B=dvgjW?sbAM|CMw}!<D=x;RMAIo*g!VXAAvE+=L*Xs-%@k3-z-}MK_DH$~
zOLKb!tK3=O1=y0gs$eU>rFX{%@1F1@$rfpnogfp;()Er8m8zXFjx1b%G&6{5K*uN)
zm$d~5B$ui-R&WUQ7*ePFb=hKfDg7AiAgW!)Rg=S<r7)FsgZC5=J{mx8^3vv6YNJa~
zhGs$w4;_N=7Iv|$vYGeO9Qn4aLSQU2J27;gvd%#0HLl7LF(Kl0gxi$!+E?q&?_yN3
zb@?u-=Y*6+>)!A`q1B!wJ6=%CUF_F;dykxI%=ak{#mW~0+4M6|Zx8sb@#F`@jCvA&
zOBR{5$JdH45&N&T^_|vE;_R=x9o7ho(mg?#$y!UK&}OZ91Ss#bBunuFO?Zh?qtHDc
zD?xpd#Vc8vejF=p1l#3+F!-jbv>7)SzMmpe&TFf&puKb6%hRkJVtc0mi(=*`>*_`(
zEG1?BR1jS(DI3aOBShptm02?(5Zorr{x8!=E5C6w%+M>;s|KnW0}9Gf14WBuUuiQ9
zhxlIe<+QJ|g(g;uuH5*Z*3DDtz6O~7q}1$%5tHAeDUSp#Q&Wfk9Yk>e{tHAg0T};h
z48WA8xUEq;e9yb8k_{GNH9Dm{tR5zX>IUdxioY$t3_n>Z!079nXz{J7tGZ$<5Wuaf
zE{h@@F(E2JO8?zGIx!<#OCYPeliZ`W2Sua9>(dR?8~1_nL%w&Y0qLV^@%nUr+^b8l
ze01Fv7Ar*%OzhzK&QD-~$9VpT-fe+(deq?c;hyEwhN%*VDXxOqzren_LR|2#i{ten
zQghEIdU?OTNbU`aZ^*ph2@ZU131q)LD<nlpM_(d|sQ>gI?ey{lV_^T#?DS#%fju8z
zxf8Xr-<YPfO~bPIW8QzX|A@K2^1z*H_>A>C_4OwWBRTgA5r=IKVUi{_E2q)?nF=5@
zQE6m&;Nqjfo$8;bq1cx-0G5B5z8v&!mhm?1yaZD!cPGzMBxZf}a9QpG(LuY#R49+_
zBDJUl*-$EMz%T=EsKaK}&=gG)dd*c9u>8bT!Jsrfn`Ws(TVIP8k!xm9J-p=chQM~y
zY0%Un!qQKh|HC*f(P=4spk{j9{6=y;&gHYNk!&JiqEji!oTC2FEXH-^JBx;)aSC{b
zoln%RoIGb}_e~!=>h4O;v9B6i4?M(ecr)|K`kfqQZt;0^B8}A$4h3s7Px%<PWHlDO
zY9cR_p?SeN!*mAV!a6RNc+EI7Ngr-5p1Yv4PLXbM-?$cbD7n-rN=d|JJAgh{)7~)7
zAXRW4<6YISwHMlk33f7SFNB0%j`YLH2mumKc?gKJE*&p8%@pq{+K#W#c%XhVvog-d
ztmb)HXvIML4BcoIA7K;caFD2_*NuX^2d>3-lS#@atpb1OnF_`vp5idP`$HdQE^XYr
z#wTo3?_k+*sN@5aYCK}a#XH{If?6Hv;8y!N`TLS1kgjp*Q|uh{epBp4g+$YJ>{>#R
zi$LI@?+wvuSXmy=zO8crBXC~8gA4A%zb0)fye_KPBVg^Sr>Aja)o{ixNmE2WW_pNT
zbi%E+!-rjM9zFPxBY-qCK=;b9*w6QlV0^XAl5%x)g2zfo12$C6`bbj*bSNcVHt4UT
zZ60mRw8X3Z@Y;#7_*vcb9XVaE<!dp&3c_y+d=r!2x|jhH+eUoPO0*rEx^yGDJ=)qc
zwW83wC|<E0q8VnwXSrjZC8WeL`ePjm$eA6Ba}oe8oD4S@WwkKftIh}0ClEkl+12#R
z4Dh7nD_7!o8pP;PM5cS)0;fb&nNbAZB%^9?VzRWFiJw=wCR<M->i*X#*>y@nhQTSj
zZSH(CZI`5!hx%2ac&5zd_1BpzZFTNgN9kIi-P;F%KPk60sm&2Ac5j<rzlh9zDJjW{
zu-|fg*N{Gw={bW6o#QyusgHCPmFfMZCJ?7w8qF+oPf{yuMh+_VD-Ip@BiS@Y_6D>{
zy?Uzmc~#j_u5plmsxka^`QC?4(%wAQzPS!{Z4S$%g2VVmKk&=Td!_2QUlV6Oa$8ot
zWiMr=_v^A0jIv3Gz`z~qLGbHtL8jh#PM<vaz(hENHTyFoHO<-ma!LkW)j#ygi#wf6
zvf%eSKgyh;ZAujj=3ciC1O29!ZH~4(`REE9q9Zp{2O5llyy#VN8?Y4?nFpUjS5IaF
zR|OfU?p;~x3ti!Or+pq%AMogBc!WuWe=YE5w)vL=jRL>$i2AQ4H%ugi#=osM7Xkh8
zm6lIL(!it;9*w}5mAT86_}pY(wg9w22tVyig-LhbWD<W3X)ghH6?f%bw@g!YUs)n~
zN+geRk<+5f@_>d#CH%@1RNzEPwq$nylq@dA(Z}j3s%Xh`4K|ap&I3*f9+DdCm1~Of
zAVb?_=I!e@ea8NJ`YUp`RX(vO6~IgAu0!(LGoq=LHs{_VXiE<X<k*u)q=f%lll&6Y
zjJ`&9u{?!l@HO1!#<Hf{m1gtBFppjbX!129>0#2&>OzUrWu$XC2ikT<F@sa;T5<n8
zPss^B3y;_A1EvVoB?SMLp@#^pB!?7|Bx?s{{TO$-_DhK^j@Na{HU~5)Ld}4OZKvhp
zI8<pUBbKZ!$~^AV;%Mqui>w$V&NFjym4zTfdQ-AmigbF-hFd24w%7D9<e{*hNQUbL
zw}eX=)B;>_1N=nA`>z1hj}7Uaig*f?7=N&LvHbc&j#>wv1BmIw0W9^M&Wvg@Oj)0@
zqOkUaj3|k#i788+UolGiG(%{aFj_X^2xABFlp=O^cf-sBQhPxIXEGg6FV)mjJTjcH
z+vX-me$U#r`T(YR7D2$k7<Cb;4qA+7OiB;4Z(<Vr#YP=*Jwx^3CVp@e6Jd(|gEN3!
z$}}Xo)N45ITeCHcw|8cJ5=e2Qc(|PJG74D?I^3iPsgYZ1DIo3aNaMK3@zr-p@`8ei
zcz}=@c?#l1*!+Fds>4##;9^NpDC+thwfTv8D&7d7Minaaq=;yR77EK`fl^{FqXDD(
zS|f&Pq^N2xsig?e6NCp&Ul&vA@!bZntfFo~wouS7JqgoWU+V)vmLc277P~N+Gevkv
zK3INHij3yR=utM551O<+-uJ!z<>YtCVC@Gp&Pf%}>MN|fPKb9zyJ47q6jQ0srV4r>
zRJumL!6y{xI~F{rZH1t6;DO?`L7qTP;P&$rad=}Ldrsm8(NN`h!z|9Zf5)PT>OcV!
zxB=t&SPe!Ul#9~C#VX>Rz%l_`2eP0<H4e8kpHcx`b_d5PZp;7G>Uu9pEi8p^Lx0#<
z9Oxr%TYlEr#eQMz8c9K~Nq4k^pN&=$m|pQ~(479v`*#8SpiJ0RV!vD`_!=<R413Ye
z;9zGR;VN-J66@pLS?Qn>xBROX+=C>Xtv*@t+7Y{g;i3x56agG~bA@k+Gk-B6?<rA1
zulR9*!gOyJJq_9F@Ol0>D((!~@@^U{XjR{rj>le1pz|+V$R-+`$md3ES0dZQ62#Y|
zBFxvGnHp+ij0!!4@g>J~w&zub?XMUm9)g(&x0Zx-JmA>Cixm9<v`jx%>(o-cL{>!h
zmPIFahUP`*vHUuF%V~#zKq*&<3GXHrwt*Rv;`M);qmt)Je66E~SR_$*9Bq5TgmKSa
z*v>L1aF$W73#Wk=3CdZra~xCx#0k&*#76_xzj>w~p;-Re>Y7oQHtO>oYQp-yxvuf2
zaJjr8btp~2v!eBJ<xlu@hw_(Zm`pz-v#lJ6gZ1*!oMgce27xU|lROLiVKQ(^75-zD
z-O%x^Vx6-H!eL#|$ZK;RAsPykh+>1q<`e*oLtF0AjGuYXzgA^M>1K#VDBwmk5tJ(>
z{py#b>|Bt@1pMK`1?$o|*gK`-2sH{dG1abXpyTMIJLa6^oWzfQ^7$x^<A;P##2Yqt
zqRfbe%ejx=Gno;)lTuR|6qB8&Iyqp>;&e%pfj;&z_yn?XZ?ADIRa=>7%1e8?0=s}A
zg*GS8PG~id)&}NO|JB^0OmCUm7`kZMfxF-qbld_98vNv9-+m-2sll1f+koun-V~FG
zyB#5@IF<G^-MEE+AD#x%ePRSkeuc2#3CHekhTD9+`K!Y+LCGQ}?(VNC=LdHm5y|Y|
zrfJX%!+3b9nZoI{=KHFQMk@r1gRBaXsGV3&7(v3<#OaXynrD=dD%?lXmdUXv#_5~A
z3(e@H9z<u_GtNGZ@fdYJgjwabD&Quvpg@k2Sy(Kf2ZEVJpWOZY7%GG5;d2Ax6Y3E(
z_XBpHmWFsUcXBUXGfUr%hO_9a&i01KjTlYXuZxdJ!j-R(N=sT}`gW~CGauDOyyx1N
zooE92qIcNJlTH;_&^qs~jr!E!#3zr-L2k~WxBJ`==9F`Ty0dBicjBr5vV*K&2gOgO
zsTr3xKy@LCtUL_Pc0`sNO!8uwC-Zo2noGnRZsiY}c{F~Oy5^WQ70ENLv8CUjFu|Nh
z|J|?-_^-n{GXvm%=1yJu30*X7zts5&U8IP90%w4!+d2YvkSzQ^gF2E$z7Vag{o^E2
zs4CsLIO<R~>J<HV`^`kH?K6RGFL!be$8ogX-Q7WJq#p+E7Yst+vQ`$-B$fT^?fVgd
z>U44N2}10#7Nr~R$MxQVR~Pl3eh9-&pC%FSb~d^_{G>cntukA!JcA_fu43N_ZOl*i
zC~Yi~m-`jL_s4B3yH4EDOhi7@B$s!+46~m)Lzm-1_1n?e(4E&h!{t_8hGm~5=o7@v
zqF*3HS2TqocMR!irqVrrUN~~a3q<LCJZD88E8W%YEuAxN^v(khe<-qR_S2yIAA3Vy
zqoRr@fh}T}=a=Q-)|}l-3M*pdnx5oO;n%@?&W_j?d+8iBA~H<l5}zdd)?wD#u+7kX
z%i{+7Uq<&t^BNq{T`kUML1On3#O&FHrK_+UqF2&PdZY(BzgtDtE8{3IsE&<E=GIg2
za*U<DJrQz0V)*<u((w~=Xlof-8wzexCdriCj8{AY!X43KN)pgTFPruM5E|7OWvf|c
zK=6QX5tMsXndi~w$|tNBzScZp*`&khubb~{t;=FuZLKirbNp&p3l>cj&l4^A@_ydz
zhxqBhCThONRYREwP(1kOhJaU4o&AzqWm~ZQDR^6<*vTPLjM<kZlbHvHwCpB9tX6kV
z#6`@VZdIJlwpU5ksks%|buP?PISWYK5{xF(lvI{-VUZL!vkHNyK@umz=c?iuj;vyh
zZJ9uR3^Ki%7%Yqo7r+27A_KMWu1;3;nuui?1A|JEsgp?&&7=c-g%TEqa^B?4|IKRw
z%?ZcZrY(yQfoWCY)QTMIN^IPMo`NU8D)K4}sqUfwkUBJ;cVnQ;GI#4@`gaG3<c0s)
zGX)Z6n--w@Ip=o*;Pf!+B$R!FXB8!thFES|nr5aHSuc1m^;!v52hfaOZH9@d7xOmt
zuq|X~@#tpCF^v3491Ge;m^5i3@zDBo7LOL7;IxlI0M2#^@gz(LpybgHs@Z{BJP3lk
z9%vdI_jCEHotZ*DnQ1xx6hu4;vhzz+|06+1fzfba$ZbrZ%-;hw7F%kt+c8c^TOa>+
zZIQUNGSPCAikS=cSs_i$DS^Gz5)K-N?Sd(10-Ma(`MAM#>sld9T8ejj$Qik*1PxQE
zxbG56loW!r$ET46s=(g?x{jFyu5)IrwFzCThF~VS&gm%rWI)tsCh15v$#HSZql5mB
z)IH+|i?m2>8*)Ybt+W~rjFS2*G4H#5CfFQ!AY)t+^FFKWs&lkRU`0pmalq$4+TRy2
zv~KiRCStL;nwDraQ)WrFKGPv`!p(QEoH$Cd;(H_+I}@kXs=G{q6b!pJ0U#Js<~8?K
zg|SyGN>K!vZRjVReU{?C80UhHGazm|oEhREfZ;VoeLJ1hMFb|dsY&;d_ekk(Oehk)
z_?}4>D2}7PDui4rA;niyg&3IeH~&KAi|w1t{);5;U}s-GI%zpM0TRUs%++`H%x0Y%
zYP*_Kf`nS@UduZZ8}M;)ti2+Hix0OMut1@>+5nCtCN1qoEZYgrj}9s55siE?Lx`Ky
zl>4g66iHQi`n${FIGyx7Gg+;X+rG4G<g~tSPUc*R9fD|F69~SQ*BP+7_=YS1ueewb
z9|AP*z%Edh-l#}?sb91>52p||ZCS348~vA_o*T&kYK~2dSFE7Hr6A&sl6_|b<Y7!$
zM`C;Fbx*s35@TQ_+;Pm4&s7$Z5}sMV;2E|>!V!F(NsyzoTmyz|G3ZrX8r&5`JWK0p
z%K#w>Vu_qr+44?vzp#=OwfeZivwXcaLs=Sxi=?|)xv>_~_1zpHH7axpFHvTdGQ6%3
zlAhQh-fgA5Y1RKx7OC*O?%3b5#KLHJAU-Kh1kOu;1^bvKXZOW~m%s8Z`!;B3TVrQ}
zHV11$q;)tyNr8(?Q#43HhJkE_0waXExU9}FI{Y{WvVyuG@EqtfOIm_;Agqxec95@<
z)+6d#4>lobgsTt~2Rxp2VH%O5u#tKVlOG+iZnB{hw}yK3n%aaVX;W70gP_P(8JA2Q
z1gA_Z(w>tp0-kPO%C16xsM{GJI)<4v;U@HP_esbNb3b6XEm~`H5q7TfL1YLgXPxHI
z_hOY|;4K^TZ)=HnKgz7=sMX8Qi6gkM+NP|g2*rO|L?2#WSG@RH<s`QUwc=fWHG!xD
z9jFDM)H4S}mS{l=4}ZUDW6T=wB2${M%Y8aM(2iHEV)|zxLbdj-BSLW>dHK4`Q0&3Z
z5hOI~1b?sy9GD!}y6Qc!437F+LX?^VloZ=l3V|p{`N#VdPC4daLyGKvCAT$Rc&3T_
z`5}jGsWH=UY+7bK*ioWc;)<0)PWk2CX1NA~#=0uSJolC?H^+qhF?ez?C;&SWl>6ev
zDFrrJoTXBGX2sc)Ke+4MP<d8I&J?sKcEk#wzy{X}ld>i^rRAC~3qsE@LX07dJJeBP
zHn!P0{UvvOE)}NQVX?`#m6VL##`Y*<qKdIg0@KFyCt_LB!xxXLvPH@C-reTKT771!
zlo-VpBAi@~r5wT2gG(u|%>`Ij5lGZx);+q&=@J|DACxrdMupi<s|p_4_vm(s0FiOf
z2d{qBC{GE#t<3Tr_}f1*g$Ffs?+@zW5RR6;CL&>zzhlt_z&SzaWnnuy!3VJIvZ(Qk
z_xiZ2ZY1{8E}y*y_W^wL@7rZdRquBm(;0`TU&Nv4ygmE(t4TdNE4}iWj<ZcfX{Zi|
zW7H|@qnHG-Kr+V<Yz(ZnzRPN!Lr!T=HX_CG!uZQoi~-t_RltH28NkDF!&d?xD=d}L
z`&~73Bk4YE7GN~tkz+fPiv@A#>J*#nA&!6J9{iy%M;x<fV4f|dG?)=<MJnK7H0to+
zJOGr=)&&X;rNbgW)vr?#4fNgt(2VkX^l;G84U<@at2SebW`oTZ&i(8n`XbTEB0|Ki
zA_Rk^BDu~WIW}p4dNyB3sxc}akGXZYcbIW(0!NZ<%mY=#o(@%5wvv_cJoEIB`Bmvz
z1ON&ZQsc9I&DqA|?AL{{dkWcvw!F(gmER*8t+({{o;;h#4BkQWOvi?){sRLsM#j2h
zl2IgP^qKuPCPqb8I@~^?BK_#<b=+MeU1b#W(qc<_Xvxyy(fz4Hp8~-w-2$s=%scXT
zEo$=(E+%8bZw#H#Amu;<a?q_}VRBF$+^qyiV3YJiNneE4U}m7L=-K-BC}7K^Ub~X0
zV5n?CuW92e4~?F+Gp~x1U68{{$CGgKRd_>;YAOHg=F$l_2!fvev%uxijzIq|#wHMR
zRWz!jn%awwzllI7aOuju=a=gA@X9N~ft;!EW=FUr3h~l6x92yP_ahs=<hKbZYS6jg
zZ=bY1v;7R``wXxGx?eBPN3E^Bv2vj$A8*jq`Z01Ce>rC6=cB#odpX5~PL{D3U?Yr&
z-Bz*Z`K^5esu!&594BEI1$b4Sb^gNmv-UReoyBd9KDX2+<)~MVc!$#MB5Rg|FU(7?
z@~Xh`p1ujvMG~5+HZ2y{{%w}=T;jw#rG-Xk+n91*3I0%!Z*-<?-7o=+&`oSyKY9Z$
z7cp|0p7lkgc4~JuAxE=Dzs#$1ZWI2rhacAt&4qqu$!jKIx-#OB_1lUjffSIEDpjV&
z4sEK7cChDcW}fWRr}RasqlRL%UW(*ix4NsUL_b^M(UlJdwx?P!UdpBMDhI6?R+@6k
zbLHYQi|yzKIXuJol$$A!ZwGuZ4d`*}!g@Cq9V}-BYl&dB((U0UzY=DF00{W?J{$`#
zAqv<nR%3QUoqW&#sh37}FUidOc4Hgi(FB3h=RG}9@ayy#%R_xhQF4zO2rDWgEntrM
zSzT^HL1VeT&uB+vm*+D0wb}L3H7#?hR8OW(7m6lN>Jz(%1+b5W^LW~ix1|{KM^pEa
zaV#%A;6$DLc$WA%Pz0myMo?!qqxD|m3&y8z1P4P<=X+kO%dI7F_p%hKuwPBWhM9`{
zq3rc#bm3`&FerKP2qk+EeD`e;v+zwTl}{aKFdE{tO|oLdJ%nafiG+OW9Hpk2^$(H9
z-lYrAS@UbX5J_WMw^$k-xc`Qe+_E9vO5MPX-|(Sg;^F@{I|d`e|0Ngv&-QUfz)!B`
z|9bv^r1)DJQ!)P`#h+1p0T-bTFY@sl0%A@V|G%a9*S-{wmXZ<7;n~u2__$mRLJ_%Q
z5k)kmwuxCj?r!eBFHT@tRdz*HIorK55xQbu$_Y5|(khhtf26N>N_Y4q%i(qDz4DWQ
zywrVMyh=+&<zhT3_9J(w9v{r)gBdp=ov>laqJgi95xrEsd=W0Ww<U=|!BcAo76`6;
zc}Ka`e==}l_m;msJ$+x^^JTOH2@l9gjM_gQPM$+=K+vEKKZ{tQg-zaWR~{bsT63(7
z$=s#hhyb52Y|ga0OddK+9`l$55Iojk@tIM#ApMm=_&pNhG+6kr250g1^A(BnYR7qc
z3jV4_n@tP{vZAFj4y43DUNg$~obj@oVJ6dJ?qoIT{Kd!em^`C|?vL(yBs%g<`c4JE
zV`XDSLLSMkh|)tnB%SE}?gVk?V8A?WaW#J&ZdT>meg}I`91T~AchAH1M;#a=DeK@I
z4M}nggFq^;;qxrCx+Xr@RdVPaJ}_KfHF;qYKa=>-3LffODdP)|up<h8w2m7?Ao}RW
zc{-SWQBP}5vsPeEzIX<I*NmK7i~0$}6iG{H3)-L?23W1<xT%wkKcVrg37F&r7FcDB
z6O_X7o!0Tzq$?U0>_z94kPd91T-n^j(sEu_#O)tZ@z76Ao8=vY1-zx{VF+_Kx?%m~
z<2vn(l!OH898s7$jvMcuoLbj@0Okw6YkrwvR0t?t#B`xqab3sN`5dR#yyfB=m6Ojj
zIw}j&b^wIVwKS_k<aPU1yaty(+!N~ai|K4XHM`vAUhX~qyrS0q%-B=s_8gyUr}|Uj
z-J5eA$?R!Ps23LxF=W3c>*;vNqbaN?JKZ-O8j!dEu~!+$#+Ln>Mp#$f@FJ=G8sUYp
zO`4$D&~#U!!+~_F+8VB**K5i9T*$gSNg8NNgME!nu&^*x$t(HzQWH(KVSM2A^weI#
zj1nWDoI(;Br_pLo{qbxcVmj;hTL&!vQ&{|(WsXo3Q3b6CL3=i>*dA(S@SvG74!8Yi
z(;DpIYhCF+c{U9TU%hc^%Zg8w3_b@RyoqBs$>@8#l1Cy*6F<tEdqUkwVo)XtBMu28
zpDapc@IzDBCyKS{5^X!?P<}@`aXBVI^c_Ccv$_bU>Ph;^Aqj`g@#;`|9G=@#nYm5r
z_T+27+6EQ)?-;_qmi3BzPQ`{tI*chgw`PqH+-B;_47kDK2r=gzI>+%E8gtph2F@KR
zJY1O0vLR#0nIJ=2nR0ks)e7Gjy!>kbFM(jP%-T%2>0~~Lw#WP&&cD(ZnIq}rbq~Ct
z2UPp26??f5f~51eckneMnw{zkUw1yeU=)qY=!tf8H&>+r1Pw`|8AqxPqUe7VkilXr
zYlkhlGkc47%AjiUX$>J>q8XAVGJhpi65Dh;0S-`}u6B@w(lT}jCYmmfNlEvbFt+ti
zT+_Fr(J*CQ(Wf)t3=4x&^yFXhp1X!r%Bf1OI%`Gq)Rmeh$tSoV0h<af;HoABC;4|`
zny*J#`qpFlLa*#XO3sHwy~(f9*@avhgSt^w0@+o!$jRSX|1Kvuq3H7B<M~gQ!fYbh
zy&{@FSvV_3>Ta?E=e26X9&~@#tr>!aNki82LfUJ@ok*R;C7%{&F%SBxy>*0~BnVVw
zEy}~fF}%M<1Y3J5cXSj=PXM%3v@1kj&PIKXqJYiWj2}$c@`BWol*kTKSm!OhX-tV2
zLC?Ib%0m5mCgQ_P=rI#EH;gIqhJLj)Hghk^FP}^ZYTI%oqiube2`Mskd8R&a)o@9^
z;Wbahq`4$LaQU!PoIRsHr#^T_Tv>_tt^{nPJ~TfCbanj48Gin!E_ta72!NGOt+)H%
z{|Xl{S3_LE2!m^%rfvJJbmY@Gx`4}a9lz4CJ*a7JNzBw=m%@L4n_y~6w1gz`;hH>q
zD5NMvUV4zrZY5W?t{4(k^%UqodPjfKhvNbPumkjK&<gSHiD^mx#oy(DrdAF9`+$cb
zheUO`Fx2bPi)E18?1|+cby>>ep?=3yo47sGTe_>OOsJxP;euQc#@Hx0)5=m%sy;bV
z^0%K`fa95xkmY)4qUPC}U6^8X6h`{WKQY2{e%eMh9C^GOc)ibQEbdrx)VaoqM$SpD
zIq7wXN<rM}*w&1hyYttIJ?=4l`FEySVc9|jV0IdaxT;r-Dd_%+%{*(@Xw``7X$PFo
zes1rqJ89V(Ptz{FXF862gF4!HE}XkK+E8A8;uobkin4ko!=41T%6#FG(E{?+eC!|i
z>d#hwov0GmY($(aL-D*Trb(8WB8K+ChW@yIVZJ^X5#upmQVj8ENEOO*M|muYF@zV5
z5{eykwwLGsg){CEI34W`QP4PFu3rOM@u~2tNKaXX{IbR6j#a8)6}MT-fB+26EyUQk
z@emaE61qXCKb5C*q#4CZ@IIm7^UNBN`!4u`@ESJ50Q(>aLmoXi)A1!GJ}XS7WgTI_
zEf7y`#Ce%PvC7Wk$JP?-w-))eLX4O14~jwBl1)7QcR<egUjuR&0sq4+?4=(d-~3M?
zF9<^^Cm>YQb9HI?V+;SJ@}R^5IGXMH&QGFQL${uZxg0A2xbaX=kO+JK=95_7T9I+>
z&Cb!0CD5gD(E$gs?O375y8)f@<##OnUjnpz&(G!0S8>aO7mt0|=9U!tnO@MHv%SFJ
zzW>)%he*IJ;1T++w$N*-{T*|pb3FMI-fC!Z@?7Y`DX!;^^xA;v+x_J;KcrS&Aj!8N
z{pIS;^XZ8##jX+`nf<dYu3uf-mOcRe=*%{&l27V=zDwU7-UKD^oXmZSzXdzqznXv1
z=0@HZn4FlOgkhgxm4*qNXeqJr{L2oV(fEakp!>4_m$-Ax53}gL3fKx#TvcVWsct`(
z9vHnjL2+56eCdXT$dWXNZGT7T$ah6|3}yc;e>s6_$!uc0JE<a(Wve&5CpB%jSQnp{
zBfCWVUsGyP<w`!oYyH<)p_K-&>;qTn@3pnErkf3j@ElZ5Vp0(NILzeG^w?eZ-a$R(
z2ZqGcloJXt<EYF%Y4-d2t!EX{VEYim&n`3YT3?aV$%v@N*+SC7<8r+{fV!Gh#kmAx
znXLxXztois<Hbx40#BN<lHuj%X2Wh}bh{pd*bR(7;LHQII?i=i1eH&sN~oK7jMj#7
zz}Habhkpu)w|-t9hW6-_4*+dr1rDkx--fFa;{s2pQPd>kI=GJ4dBdB2kOi+3$iwsf
zz0JjE3B4yY4KM9Ct+d?~I^qC)*W7ojz%7+ZAl!;s>O(@UR3&u*%9%S9Dz<-;Z^k7v
z<-F_@Gh1z7v1+E-0t67X^HMVr*|0pSG&-m5<Qwpm0UZ1X)h(Tw)~@VpJ%fTOYwGzN
zsOv>x;H@Q_GlmJ%L>%=rH>?!7)Z?C10<y^iN4tk1-*vipuXu`@p>;MB%Bn)7pguxh
zD#s$bNz8lXOzZ~Bd-kCAo{>q={?ciyS*d?LkAQ+^u}D2KB@jRY6}vElx&jC1u+jKC
z?yR?0Zw5czTOFA8UfxTze?<!b28o538VGgn1a`lQCKgHz;%+p@+Eo<k(t;)+(9K;M
zMw?3fNRU*-ib#j-rJTOza2x9Q7;mCcnJje9Owc)FG*_7G31Ez*jC+RPi2y?v46sF_
zr-gf?tda7T7jt<HGIU>ohw))ah&L1Qq5_97gCDYE#Ojp+jlCr9Bm<N(2j5BD4NLqR
zeN$S*j@F}Rd=`o#Ee}Rfh?POV254whZNz!{7{koc1K0n3{qF7GjoJW+nv{tzRE#Nu
zI>fYq_DStI&~RVtfm|$h^iKj{-RPb*t7GXw?`1rg3ta$>X<%?BORrdcyggA_ZGUH6
zeiD!XmD(p0cRrw&lOOKES1D@gchK|x>dDC63fn3-@SSmLi##y}tw>D2k{l@($tE9G
znC^wIth$;WyI9)U(LI)#Sz!|eY9XcY8*O2@cv7g^QMsP!u9+9*=7YaJ#o-Mme<)Oi
zqxTZ1tXCfmmJ&F%H#C5^-|2~|f<^x<{g``2%CZ3K#C-H;eN)H0FO6`?+g8A)4+31U
z-n`X4npO03<1#feqqQ6sLa>9$6PmQ;eJnzfE5Aj4G9k(^A(J4Id(0CkY@MR4W+Pkb
zK_VxFu4a5jD~|iquhr6VhxOJw51LbNT=mb<3>Z$DrEO?-o%M!@YaQ5w4eO!SsoR2m
z_>__K?kjSgyH7XrmkomxNp0Yzmc7rB06g4QwTPafHDLud4|%*K)*v??d4pqi1tDOY
zj3TNU#PY;b)yRGZsfFfEGnQlTeo|?AaIdk?RA(fd+S$3^VWBszbqxO&<`XU-lUtMI
zR)YX8Ya~vK-gL-9TVQ{T5v|iEJ$$nurstOHFp8?H*hYC%Ttnp%3di(lB~i%U9p#QA
znV*^u3yDFEpzxZd%Jtp%)3q<a-4x*CjOcgMb2!~nFL?)=*aCVFd~)O3khIusf7E&;
z!#!~{E}tdbiw+(r-Iplp9{(f_+oW2lYC0aIeL0mFhh2GJ&@~jPN;vZUgdL;V3ly+;
zLf!Ax#XK`D<`Hj*HHt@QL!a*qD<8`UG|kyxZoPc@u;mzRn7UF4G^<5FS0y_C#?F`z
z=y{GAdee~w_Yilk7vm?OAbR2dh3f}T6E8~kps6t1f&+iM;s{LR1r%@{BU2)`p1{W#
z0(U1hrLfPdZ~)dkA!96t=t%r^mo*zQPpe%I#}ZzPw$Cdi-_&<7FfL#%`?*F4@*fd8
z8OXeok)xBmNYU|lhpFoYJ?wO{H=8P`I`BM4heEQ8-5DGarSIRL)W)^gB&?y{4|F)e
z_K2Xve_vg$?V_MO8Xrlu&cM(6a`B#5-h1%vBxTucu``n(Q3h%H#jVf#d~IK2!ejJu
z!UT)*8q%DcSN9Q{cgWaK-@MxWl$ER`IHmWLG;(<1Z}I`2w!YF!>O^%vOB8?(-TwxB
z&RK^oj_wkg@gib9D{e<w&{55bNC4~>p_i$mO2V(b+_gC9xJ#eTXtcTJCn4259uD08
zki~R=C>h*CRO6NhkDvO?&#76VEW(Q#-ue7s8RVa+`2VhRjgjfUB03WT>;Ej9eF|IL
zhM4{4O7#@<xGrJ{o&ij)h4qgO)(iR}s*yv6;PH1TYO4)hw^+me;jBi+R`Ycc_lI*_
zXf0ytb11#3FeG&rL}cjmHRLY%ST5mf$s8msVN}4td+dO%tU|%w`5=aPIjU}oFW1-d
zh$whWF3=Q4EeI|FEx%i6u6)|IGFf-w3#Hed?e6u7UGC0LFfEY(o<E{oOmFY^|HIfh
z1$PoPZaNd&b|#qE{>7Qtb~3ST+qP}nwr$(ClTFoL?AHFjt@^6^ru(Mr^v!wuJ?A{n
z9U)V#&Xl&5oEbG#5Vzdeg6WWWM5y+wUdg};i~gUX3eOe{1`iIvN<c*;J@dFv?rBfG
zBqLA@@_N_+9wc*B|C0xN+^_e98zjS)w-Ej&>}cD_LOes#4yZk79i4V>>uUsWUda(z
zD9evtAC$1x@BLvxRl<K8O9Y645S$+$_cx!r%lKwX!#9)GcT73(gg+yPrDJyPGV~Ac
zOA30mSP>y^FnWp&*u6eqRE10>#HQoZLy)U`Wvl^RGzvMU=Gjvzjvf)PKh^--WKzUE
z30$_Jne&&grM~Fdwa?=h{8|dF?<vGSIlwvI&ySURHS((ekS-r7KZs8iZt0-ZNs=_M
z`-#{SPI;Lq2THkWXcI}1^PnIRZ|p)^H+atuu}fp|AO!Pakz`$m1r~$?v8NQ8KCOoy
zKS;VJ2;>f$Fl(cF-Ln_Wg_?;JE!o}&)F5{i`qBnt1ng7u9c#vt9{?==0BT-C(;0=(
z;X4(5M|=VM=ImtK_zl?wLzCsGfW6HeI(U2n7P7?{KNWGs(3hp_))eR<95n>RA$Epu
z`@EBZN|Nmp-E}SrYLX3<p=1#q#cjYhX)H3?bOaRE*tKPQGD(k)1de!A#gyh>4RTcJ
z=2;`b4qTJ)sd~%bCkgbzm$Sh<vnZy_r#we}pwntqxvH+Gr@VdhLe%8Lp^9+g0nKsq
z;mrf=aq}sBtkSSh!F0t9;-Mxj@0aIgKp~4|h)!q?=etTJ)*z?adxXbeaG(c_hBj1}
z0kLjlA#@!7Ot{YhVdV$>mBovwT;tuGbf)vig-@xRs)fc%TB^5W_ESR%R6%=JPE#|r
zgn(j+?t+hUSnk3<gvSpznaaPxrj=^^FMDJvToyV5swS+)Fss;Wf{_<3!Zbt1=scvr
zOYH}dTr-vPb!C-`Wm)s=iN`1O7qUzbu8m8DVCMle+?7WYZk_cea;WT90^kO+^oO=O
zZb{%u_Qfe6t)EuDG};`?xn23=mBV7h9ZKBO{weh~$1MsA3JVr6&k9k}bvE@a{vZMj
zYE`<;cBGuTx>v)%2?~djW{jzm8)`Yes7sT2jUu$y*>%@{l8S%{y_3JICBiBtbw}}g
zwe(2VvuaXvsckfwxitWRe`lD3I+nW7h{sV0;{R$==uAFL8%K#Z^c@uPHx}x*HhKQ2
zJl_YZ+z$$d&DSUOO(5@sr!tl8i<!e~k9923L(AEl{3&uepZLQs;oCeK=e+DdauTkI
zw({i;x_Bg!I7i5n)0L+$BMQGx2Ze7?)7n&}NYb8!+NfltQAa}e>|C|y9<Rqb2C+51
zJeRRald<pM*X-K28)l2kIF8!PzCt#boy|JYQ%z}V%(!GSJ+5-l3RlP;jLi!N-(vYA
zeww01!HwmJGJ!X)m6wvP7Vm?W#<FUSA>uI>*5;&4^5PzuQR%-q(ux>7rMuFJ4mtS?
z?JtUB{uw97NP|X}V_JlLJsy}q^|JMKeA3+}?ZSs-YWDu~{O{Lx>3eoz2kh8}YpuFp
z5f??U&9!45b=%IW*E`R612CUFfbb5U{XF;|+yc$XVJY?V-@k3%Z2^JiGj@BoZvQY0
ziu0%l)oDg{fMSz4YI)E%I$Nv}62<LFjkQR7DL|{=o*_UeFcPpAlnE;}K2R}Dp#2iK
z1MFI6Jo~pmr#n}Va7}W4&Za;stNVxV(nS%^jdczoX}vULBlE<fK9d(AWCsB-ls}CF
zWN63O725%<5O@QL6caR){1C3i94A>_HsBr-pgrt856SwJfPbcQNKz}bid?9FN-n)d
z_I8MkK{NUz+T1-OgY1d>{v5BupX1$GK-~IG8QX*0f>6{XlY0<?`SEd_lt>&c^Gll6
z1;6+*Pj+(Vt>)Ugt=WGw&^!3oto{pd_9q%`PRD{m<9?M}8Gj4{M*E>9e>5-Sl&o6A
zO)<vK$ulgK&8e0MrktCfAr|$B9lkWp1=XXrS0|^GYO1fUSfn!I!>;B=sOYb0uW_yN
z%mGi;1qIh5+B%%A-5vR9)k2ls3b=P1u#eLgC7RJ*ydgTId>!Uf=fv!veV>xUl+<k6
zRlW`BQ<LDX;;(_tdZMV<D;9{)Uj9OURNl1p^*?M+nu8v7@I0x35{-&sX89$$cto(g
zaiXYVpC67REPMHoQ9cQzViY<b-wZ&aL=4mS6xoN!AT?k&v_z=o7_a%+5kL4qUY(fw
zc(x;iZG|C1KK(cvamkNp27f|tL$9{T>IJ-O8e_7gI`?4@pI=v|VgX&-3qLzvr6A!}
z)jw_k*0d5a*12!Njr4T?>9p%uG7dgjrUl3hKZe;G$YQgd=Bwi1W=Kj=&{%1RIaC!u
zJvB1gbs@NW3OQ5G215BWNW({Y4BSI2!zDG^BogvI_-KJ*#-m?ru8yoSMW%Y7s(lG~
z0$Iys(q7k_2X0+3k&D^pAR8Mp$nhSfi3s~+ZtdiSVT^|3rk?FT;%eQJ7O41zo@)>|
zz@nP6nlg0}x9wgc?n}#U|Mm@0*$n^npGt!N$h&9z?;^#ex`g!xD`MAm^$Y>MAbc=g
zk7`f*6ZlD`?+@rz=*Qza`Ge6Fg{O#}>oOR$^uupBPaN5uV7rbzym5@_AKu(`eBLdv
zwdun-{O9t?NbW{>svy>8I=Pd8tX#hPxPCL7pYZRLb0$#UOe#6Qo7*$}(_cADnOqQq
zTgmKuM|P0AL0E9B+;S`56b<)OWA<60L+0JTfl;2&jq<PfJFkxS4}GV}4qP1*2mD+X
zv9eFCAZ!5Z?Gpqv2M{#HJdhtM6ZMfkCAbGKV~92<Hgu1&UM-F5!S31^fFw!>{1%0R
zz$t&2Ier2xs_TDg*fSzRj%(Dlw;fviGNHW6X`0CM*GozhC0qG)OuAn_5Nhx)J!?E5
z*!`dw3C8)VB4qN5$m}UNFPsxHRAPs-1DYP2ID&EgSB7t-c^sxJf0*Wk<h&oD@_8v6
zCBGJUfxHMRbbn?@j0`m2`eo99{LvvM9y^xLsJyYWG$pIM?9VyRqt+4BNs{aGlnm&`
z7T!v@<V{f=-Awhk^E8}X6!f=DECr22ingXz8|y^#JcNC^Gmr`(7Vtd1D+Q_gBM}W^
zo|$^3RQ;{ktO2y(@igjV`IleOEX+D0>ZU^Msa}830SWIAR0<<Ww_)9Rhg2OC(S`IT
zxy9Im&>F&zg$EBpASej?fVkzG_>yD`9k9}X-FZxty)jg6)%d|+k#Lf~VWf4{BHx38
zo$efbXyS1ywl;^IL7z{q9CEQ*?!o@I>Y&D0$^I7eYie4#NGDp&fLjUXlL?Q(hCk8s
z+d{!drhc674ok(Zqwfo$&nZxx4%1^%{Y-9jcd=iELydPV?W^bC?u^)BC;ZikwX58r
zE7xSljhIF~P{&e{x)#?IR&CB9hQi=(kW_kV&Gtk|rRn>a&|^Z}gWvm@??7&4r8bq>
z&F$rv=~G{c*wx;v$FtzNhw9nZ&TO_cW?ds_I209z0EvKnc*hjQOj;#+C{U&_37NQ9
zMy_VyvK98Q*Sh^Zg+O~GlSdFAR(09Y4)hbFXX=)|UHF>A0X!*S&R)A0sGcr#T_nF5
z;bq9}jOO5lib2RCDT4XAOw>wfO<awcWL2H#$qE*^Zp}{#Uud^4$kgz}?jjGCWV5vn
zqo|w_lRvQ-9ArjWIE5IIItOZ0RDBu7Eg9#`vR<~~)XvZ&n;wc*Z6As`dYrR;dMNSA
z2@WyoBM=T^4>i-kMx--dnu?dMA?&ZZ8DlANk{_roWT_4r&A4<aOY-LtV^DY=mNJVg
zs$`+rBJoPmg}D-zTcIjEi)hNcceKliG+}US(J!O{jHJ9t^o}Z0BVJo)l#+`=p!j>S
zYf-P=N4if=Q!Oish<r7*K8wvUY8OM48N=MpHEuAv^|802B2Hdf=RvmOvSl%tzc2Sq
zx-3k5XIl!d7Kk7l8m}Vq2QU@p)f7Ochkln%N2n2bd0u#Rc3gdBsl^NGcH)1%HHAOR
z4f2EFmGQaU4P0jj)3tBAmt{B=#5DZz#^mt$=5%o8x&2@kMV3`Jm!!%;3o=-q3ahOe
zIu6BG<;_W&+sw$R6*!+kk|gXku&~w3p|?}xLB5~Nl!?EaDrY{mg!0#{0`o^rlXq%Y
zoo*RvWh?B#I6C3XP^o1m_X`5_BotWxu71eb<3C+EnkvQtvX+T_9F(U@2s;Tao$?`8
z2uOu31=SX)reHXTWhiK1@FFoy11MCYPtZppn5|BRUg8sXm!ahxEv?W<b>ypR1io9Q
zhuM(k7N?BYZOIG7;=pQ*Jxr03WS}UQexX9@D^r3AyH)R634j<(1Tbxt(`-Zm%WkCS
zx()UEhM^|ZFH+S&+$bjROQv@QZ}H24Bx*5*pKqa5-XROk5=K-*2~dnnv@6T+IYdM^
zy__^2xy$bnNPsb8i6<O~>-)>?<GfM3+`(Y5KCC9`&0kgBY800WH*sfcywezWT_rTX
zS$gP)F>Nis(8%U7+J7CmA?Rga*s@sPh>0XTNUj~qA<Ir%K{Z)9HpaYjV`N|eZ32j$
z#?Pl<cnj}^ljCa&?^#iei4&CIGZgfO&yV0Oz#4ZqiL9TtQ)mEb<sEFsb6J4bqmN~O
zGnbQs?F@)~Z~4KN8?#poXNofnqziyE`etcqy^pB^FP3Dae}jFdkjQ1wR9*(y#o6<7
z6yIh<8dTr4no;kzi?4}ML{^cZ;uqLFa$c*#{@hy<WXY!ppsQNtlCN0W{A1<dG~7(L
z8PfEh=BiY$RnYBI6OM+R|8AybGoW$B>dY^Do#^I;j@I~M8Rpti=C<v#$n?`TL&0w0
z1c^5R|4m9*a50Qtzt`?5=5o=TJid*l-c!VucKpm@TaG6_<xNe#`J6^vfGVRb-|=vY
zWC*T7V01H(=GYBykaT{Tt~}Ts@6q1c&7OO)ehv$A>=9aC-Q{yUU@NBtf4wf;3d+g?
zgywdfY=(ic-p6%BLl=%Ga^o7VG`#=q>iy918Nm<=^*>I}%nbkIbRGl8&#ChNW-<N{
zd)#XOe?#u6llxP5K_FoVq&xVmwGMtNd<XY6!RK!Tye^X1B<gX7H8AME$`ivjTXgCO
z-f(U&cMcxjZyv(f8F#Q#u-Dt*F;FzwIfgTLJ3t9Sm|NWl-Q4Du6<i{&qUL{HGd`{l
zCxc}mwUvR!iR*!aWugeLVp&Y?v>RLz#whLM*hbz-qsG5-Dx(lV@j+cLcAuYhvK;uI
z?o&U17lEU#)2oVz|5K49Hb_!X+qx@0<dgElfC8Vem6k3DVHj6C7%!QaE4`9DuC^=5
z8dvr`y4B2$iUJcafF_o@JcZEDf&n*m(li&q9qs_ng25-MCQQC1QBe^Nna4U^VE3M)
z7k#pQ<IzKs;Wf;vp?Z(rw;1G#YQw*0HZN8%YFPB_;ut=x|IJ#Dg>u_gS3i>}L+)Zw
zRdFqOfAn}mVM}e6AGOADb>YH$w`fE|D*Ya9_)wU^GY{=4on}Mi`Wp<j?4b2bY5-~@
z!@1ZD4=UO|cak~3N|ZH{gq>YimNaIl)y7->bi5+de5^%ps+_BCBo@h>w;p5`)os+N
z1lHXeyiycDI-(z;c4B<CdRCd6dKLCJrLop@Yp9oN*rf8h)SCjsZ6J!~5{=y{J#-2H
zqRwI@KsqK5zq9}wJl`w;WDc~xMr<W@RY^)1!(a-F$c4rNhV?h1j!v7k^|g$*jtQEA
zY=!B|{<}g3;Xgwy%nmum*Zt|;_4#V<8G#&Sb*BS_B{j~;k>$NRe{>!b^HhysBed}7
zMsQS^@-iFkavX{=(gy4k_VfWyC^=1Hy%Qob>C{v~0EV>OK)n~qLpl<_cfbWvU@?5%
zES&cKSq>WKnaVyGJ|p4Wf&gg;T1Jyv+n@|Qzgi8#fiVJuK?lcfS9bCg6!S@L)$N=_
z3)81w=<oJu8iCVLQ?*)z-xTJwnD86?s01vYBq5U{xO!;0&tqfm1~0quiwr>`O$w2i
z^!I<|YFMlMjwz6f(MVaR>LP%tgt8PKmmN_}GM2%-ZjY55%SW7De5Dc_m0P*-`NE@-
zPkcZI?B{ZS#);F=#<8%YjH=1-m<Iw^V}K{I<W=<>JDPDY>*<}OADP#U@vn!3iTI4-
z-;I3NEb|GIG3OJidOQ3iyWK{YFdQtAxrZ1%b73Zm<8$C+H%CoqB@Hna!E;nIKh59O
z9{G9k#c^e`31(!j0%G{~5z+x07}29K(C}aFE|UCYEL`+n4u}Mp(SQ-0+ZK&ax<P%r
ziiTm4yL$hZ6qgPDPW<ThL4uHBZU%zrVY~r|0sXo?Gq`W-J%(60;dYTS3z%BRz2|ZZ
z7)L=v04jO~!S<8L?n=$?b;0UIRf6EbbZdG+{7wQGV4<PS5kz8AogSDB0u}V*xR&rE
zDyUg)c{zjWCjk)^WP%9azJJ|r;uOF82UnLHV2k$D1>Utc^e%+kNrSWyXUVUvl&V+M
zfi~u<5lGwua$sMx*=A7&I)&-_$;I{b)esvf252TVPw^KcU<Gl`FDhV$3MEbf)+rz7
zX302~vO6`U+<g-zVbP@c@e><Z+P(DxhLsuV4Ofm^nGtYBcz-Gj#-$g0asa1?Tt#Co
z8>bTXK+mKDkaytAe+Gt1fEr?vp09Dnh7yRXLF@Iv$QS)!+y|YJl1NH#4%|%E8};i7
z^es2orYiJq;?9!Y^F@Q@6f-_57LrQgFGo{i+^AY3U)lY7$?pKSf|GmoH^9B~ahzMK
zblA~xW>K9Qf@*`qMVz{E!XMFDY>oL<c8_Tx@2qZFI#uaQ8-D!d70=vN<+4c>tmuiX
zO~LZc#Y3H$azphJ`2mi&Vi?H8ku*y8T~3{0jQ3Qjg;aHkPTk?V(X=*}7l)LMm=$w*
zRCoa;6MOebg?{soQz>29Mz*nIiK=mCOFvS2I?|AnQlF7!N|jOr%eeOI(hlZI7D}}_
zR68dmV3JiDEnciYsaFeJX4S69+}5vKzh~{ngmu)7U)M}8qPcuvgey`>0-RJW+0&J1
zjER}G`TWbQm#h8Y^6i)%iAp9jQRp71O=})0U50MmI5tRfCMT2#>PD-PR16LqyZc#v
z^z@<KAw%n4A5D`e*)V(-(bHlbI@|0FQ^bv!OvN|{tH`<#ZIN3C%_@>kMLOr|NRtTn
z(i%A=7URM0U0$^+comdjcI4)(@3~xrW-=QxX*5TFNiR3Df06j+a$EQDtPa+){;`YC
zwOMac+1&^0uzt%)SVx;Dy(>!UZ#vRW{3CiciEEW$(#|P-DDG<C$rav?#S75Q&P$}g
z2h6w191V#E(m4u6kBb!WY+~cmjpZnrSK>O<)?Zp5{;PKz;;&;8OU|~NDCfrajzQUP
z?c~m)Q$^qIF}!Wj&ajKV?r({(c5S8}<E3+p<V6hYO#MWTfgOt)oGSuNTO>_tKv#8X
zHnZkUt&qJ#qj5EnSrs&J^v`I@WBVs-@&|7ZONJ=BJ(fEnCSpN`ebXe{Z8dVUF_23x
zyZUc`ATCzZp}{fFI`_1f6&m(QI4T?F>Ku?&KKLeTyg$jIk<_P(pfN+R(S83#dr}d&
zynPjGrcJY@yX?+oW&=w)fizAAuV{LtA=9>)M*Zup-8@di;*aQNp4ST^8SX#*qO&mm
zj|{-b{$s!SZ!*B7ri3+BJKR>b&Mq&aR_Fo>{Z9sv&l&&Sp11Dg$oBaiE<Zi1FU*uh
z(rnByTq~R)dWIq-|FE0<t@(0j-F=na{q@4<$}mBxyL0z?9iD@c@r9Ow-Pndnlp<{V
z`f-0ZO?8}e+B#3nAyP0rvz^4VcnKC*h%YcJXdT$niAli)_sAd8P(lnRPv<gIoHMPm
zE&UK>ekGD~8<0P|@%8xnti#2S&I=P<lEQMA7iqw3w0AS5B=|TSj8=XpHdw%d#{v;2
zHoFl=B0G@pG0<K3{qB#y7kRYvkVq9LmbP-VeH*<Vex$AE`$c!-EeE*yklI{c@9cUZ
zzUgem_mMPpo(*HFG&o|rq2MCW{yxIyc+8+k-<~O+DkpX<qM#824m_=oINA=171ud@
z-}3Ah_*6B5YMV1f)ruz6I<hhZ%`_IXHe@LyC?yOx34m9R*qFodG0&po@nwt_-X7HO
z5Bdf8hnW#1E%l)_vYTZG!`&dWwLiklw_#}g;L_@fm79C(2~(Gl`8%ArzG=kYm6X^k
z>YswzWDmNAZ*)4}2w@#q4l$^yu2f?UzZRBcG$@RjLNqXOClKrN@0da7Xv><?-->Xi
zV2A7Gc_a1P&AGff7r)chlQ=*Wy6}OE=Xxpt;ZsR(={v2MH7!E-Y}j0BFTrddxAhh#
zIK~t8!ARFzqC;I@^=m+NW}djL;m~*rc_~6RY>HT}Wjnsf=!nZ5;zg|bRS)E-{Q5<#
zG=?3H`!6OF*&=88EY|qP+ItK6Ht`SfQQY8T>Mb^8r?n+FcRK2?wT}cac0HVAnU{^e
zS4I<#T~D$4^Ed{XP9fNE_e(@bCxBGW&&phFU}NHQ0o*z*Xi)nNbYBK3C7hTugAjU>
zpu)<<yEJL3u8e?bRbE=GyTLaW@csvw%mmyc)X?^3^;+M(5?sAl;;dK{HSCc;qf1gy
zt9E3aL}sMs4=l@Yc8mf;Rnajr*P+%iP5MBltU1+ausY6><WSQYHGd*aBqsSC-PZX6
z>?F<C6;O)l<XhmOCebB^rPU~|z&mdQMG&l`Shu3?DqL3wg5@I0wI{x)dz#N+eO_qS
zVyq}w2|*3!O}?zqm0JJICWWbsrcS$L<yHxHV{Ocu41vKibyjD}rO6Nxa_>y$;1HmW
z);#n>p(V9F1^B^s+=X0Jfyxul3&7WWuB~1h<FW?*=u4?N+m_$dnnu4~EvXfZFzL>q
zsu}oTImNjGr9gka#NwV>Q?Ha;&h3MR@Z`n`RxfN>R{r&EcGe6mGC>A{$KQ%{<*Zqe
z%)9^rcR{Jmk-9w^iP9r3@rDTS69~8H8Tj>kBpN5DGQEM-cCA%#T3T6-SGLQxyW%Py
z^JSc~F)ePRWeqa~G0w~En)4ZKs&kDUA1#^AI4zwn$9zS8X)`efmk%<)@kVyuAfE`+
zD{u(aQA3{f8(1(mw-QzKYEG39D@{uG==qIw<uw0Vlha@T*;SW`K&2k=p-&6}eXe{=
zEjkWagL-~GSp(j+Gkq?N$hLt%w0Tj=Oj7hUbMEEHI>6jYn3PemANq^M36<F&*_zu@
z1m`S<2|DAB^3KF!2X}OdQNWzn>>=G#Q9*M!19p#51jPzU79`0nn*D+L6fYfEIm-Q)
zc?hGpjGFKrGvjZQ5F>$YPRE$T=hOJV+4Xs~)~jgE7`B_xGbnbg(%gTc2kI9itvRud
zukwDQBD}#TNJKe?X_{j`<MXizNc;Xh4Xdb&Hp){j|IrZp$h6X~pv;L-4SFdj6}Her
z`SX~k)fe_YzL|87w=yB@Y@XQ)mps(#Ldh`_9oBa|VQo#+d*G})yND2CSS~E-f#_@N
zvboL|g&f3et!>@59xr*FSV280lvB8JHk+SN*Q;kVnA13RoZhkN41e@Tt*z5RCgB!*
zOAtJu`Uh#=UCfEel8#9wuFhJzF+NFq?V)C4rv&kDRHxb&^v7m2do46>xFMhO4b`?x
z@q60I9fT20wQd#!!R?T~U0@G)0pMM1)*n%zlcbsY582FHdKS!%TCz03@n<WvyJ14v
zFb+4BDd#G;s&7a{fJ%D+il$Q+LHH7vc||g+dpFN5V8X@ZqcoGxg$}U~7XjLxN;Qv@
zeG^JdgG%$$e@qY8EC=_Riu&hSO?N~*MPd~ObOWMy;iXl;9kWD*fQk^Ou@{0k^~%$9
z1iTUP?NHn0I&+nvhsPzLG2dmeEN!ILIP9sG)MZW{MH10EpQLg}-TzD|;(E)i!YL6H
z2NriKUO{Ic0(s2p4iXXU+FEV4DtxkVC}TkSd!b_~H?SN#(u@?Y#QV(;-kv_<LGw^T
z&^Lm9o*j141BOp8wGmHktp*yQ-LI9x_8rA^9XicJA9Me@U$9g}c@r-^Roi{iG#o+5
zAQSEJSg^Ep$H%>E6q-0`3Y+@o+V)HKQzp_gZ6}y-B?q7eAI3-^IEb5-)b)@aGBtu9
z46qlwpKbQV+n~4$YCjEaDD6er=D(@kd0H{`DP!uU)U%6oG`9bX{gol*(6k`oM+iYc
zmKgKW(?^qRu_~=(sG2VS3e3Pgff_H#Ub;9FuMTHlc@7r$*)XZ0X|t`mzZRdm^)zFq
zO1s3IAOjC|jtz6E#3Hz7;YC1#dE>>h&aL#3W+a1G%96S#f&$o?+AGfGk*{z(*tin@
z#8!^Nh&5c>$<@6TP+(+!hy6n<3R4uIAGB90-9;-0_nKgZEU7~qZBDvgQcLKa^m3-M
z=vgT*Y^^+Yp}__w`8z|%Sg^CzInhi4d}(S#53HZ&kp_TpN1V*A<mh5L0<w+|%ObwX
zgbj2(Fm`7Yc-$&?%Tv7E!qg?Q-$pSDz=ZZXdLqC>SR~4{=zsi<st?Lv?^Z$HxyGq%
z`kWvgsRlW1kXD*`&D(PtPN!boZ*ya|8WFq-d%&jG3D#*{S3P4r-xs>xmwaXJ?~&6m
zQ){BL7Q_W}#MFUbKA^R>a;Q~&FOu2{o+gCduVUwfq&lJ5t><!@Co0c*vv|pXS+Q^&
zZjeJvJ5NijTB6cFZxjcl62aY-;6lu+kg<Ukw;p%eN(n7~%cUMPwY`ZJIpD9r1W#m0
zC9n=yMk)^>R4f-vZ%+O5dFGM<2E;mZQBVC<G01RMpb(RCLEpq~?IN?Bjzt||4iuxo
zsg-B7Fq^fo=%%-)gCeyw+B|P+YrKfZ0f3DqhPP!|@@lMcRXAmG;x4&gLB42M%pIs<
zvM$(aUtFwA@7EaaeM1aLV?94+gM*6V{?!BQT|*4o_I3jcWy>HM6JI5{<(HK`vxWDw
z4Prrn1HWN;ff_3<eFyjlY{cQ@xk)1=O}|xCDYq4G32jb@eWm4^Q@*jC9h|AZ7!H*^
z7$G>$(UVRMozgru44k9hG25TKsd<SU#KIML&e^-w^h_TI%siRjz7txvcbtr$2>t_}
z8rr3C%rA^++QkqM+uBuT^o(9j!3-wg^3ovlLd|&Mkj)E5gEd=wVNgQrBnS-}M~3FC
z;Lp_R3vy$tim6GnSz{r*q*mISt}3)O)K}l7d4iGcTO>fRTKY?^VpK|YhZE-+60x&3
z4(fCul*Yls97zIxd9f^kKb|zBE4lxW{|zSA-C6D#W-S}0ZL2?fmD}!?FEDv}R9(Hc
zja*0-xcmI~eu7_hTGQgrWOI!aWEK~O@e--?>IT1z`@+2HiJ>->+3sX^O<L{H$!cFw
z5pzr)koV+fd3opM4<Ru{#53-%AbIW^OysVY6(KSHt?nb3MdID%{q;rwlPHGtpRzXV
ze*iYJ{XG0{vi73-XcP_`V&}BV140U!`7SUfNCdO8&xYOv%?L$^=N-}4w_un>9fmx&
zJe_i!U^yx4JDQr3hmw%E^ET(^=beL_`<n~@S;n0Ou7kA>fEb~{1Dhr?31A#iXVTVu
zy?LG808X_mMmvWsNBd{HbK~_D1e=Z<cAQiu()DX+WWc!qLExKq?F%nK+a{?!_Fcjv
z`PsjT^EYG`)a}>D@OrfjCxUYMzx`#Ga&J#B_qTUbOwJrV9L{ic-=M4MAL{WhEaj!Z
zD8D=Kb^>oloougiIS4g%?NeS=XWlMHrSas7jvOJ^HT9#^yB<oTA0rg%kVxm<iIREJ
z=&TTJDpzP)&ji`)?cSu)-zI2=g^)a<$vk!BM0tU$S=2$>q-f?3wQyK(z9%xQHUV&M
zoaGsSNnq63)=GpR_JCBw#GJ>Jep}^S`Ws0x<R-MI8AEGNc&8L<Lcx@J*GBGrt76Qk
z-AN%sbX@4Y=!f5tsG7+UV&lMagR|?`KSxa-4|PigzT${}ql2Q_b4o=uj$f9MW;bUZ
zzGA7uu5THC`d1<e#4aif&vFJ|Owc%ZTfR=YomR%Ms9B>Op091eA(W@)eKX{U@JFzI
z82M8rZoDB(E4dMiJSj|+r2vk6f%aggF}jxp-f+*?!s~Qlpw5Q{{B_9z^=al`s4O5G
z{ET_zXdd}JU~&AD(2zJ_md033Iet#O$%Z5O$lfn(YT$gnBb+gYM<F706S%MF;}h<@
zVTJ>iM7XJ^N+CFPK>A{?b}aJs?AG<0KIG%DtE?gGD{y+I{BUcQP7*D0Ns}@}_ZW}X
z=&t526YgODtk1@Dt5Y9h5shh>j2iAD(pGJajI{|Ha>$v#m{Z@U_w5ogc<Sb8@GAHm
zq+hC1Cy`T!AK_OfJOQck$&_H1l#DvmAu!JM6~4Y&c|-ZpSz#=CSQk8`1k9`=dwa<d
zhTIQb=GylY%0UPra?@<PASC>(K?KOWqGskIB1WZXd%m~%0$oVFWm^bg1O=ThU-|V5
z;%8ITjRNg-5`)unQ*K>k`H-4S{@ja|6~E<F0!>VuJ2i>BtVAx%&#-4`QFpYFZ1>gO
zI_oNx_%WN3+|@KCcl_*%7-SsCwb1Jj)wF;2ta=vHTvMlMIxIBtKF3CHYIP`x2d|k2
zmr*b}>^*g0y9oXr%;^cl(dK+0`}1g<-@qKD3;Wq79>&E>PM6b?9sohh5++euNMNr;
zgF2q)qvC1mZstRDWgVTa%7A+wTM!3*e|@2-Hj8OjeSt65DmR*aUoUib;oLdjJ}0AX
zUHWg3i$u~DB#wD>Z6RK(HcNgVp6m4UBuXh~Cg+~vL_Q|~eM9@#Hi2$z!I9I;4y1<I
zCYCN$1K{_Ct8+l5aSd;uatOrV6>Qs|yI<SUHI3${SPbhENJg-K5(JfKv6O?+{KptJ
z=qG0YRv|0Gmpfz^Nie4_Wy2gD1FYn_#i$$gKPa9xL`*S2AnH{HKedfVdw{-K))d+@
zfXz1<Y_Tn`m6m|tb-4wrD=L>&9aFc#i~{NDFL|sPY~(nJ2zl~lKGl1s2(JVW>xyZ4
z;j%FEEo?kQ8J%3}zTx8zGo6}ASz0jOSsme+tjs)hg>#+jWXF%%{dHy34lIA-py0kp
zV0`60H8GaI5eDs;!|~U|D8EA>=6o2H1(vHV6mjpL2ntL%4lSbexay<#16ZH2=689#
zm+L}KM3Qv;Xk-Zs0QsUAz$box@IJQ`o}Gvb1^Xd8{v4<tG0WAaN<nNh^<2@!`GB*!
z-9FO|{M|ki`qoFYSa>7ze8K1XnDgg@on!HDCD0A@qJ(*K81#1~EHz6iD@WpQxGmjH
z=}{iKaPQ0RnWtWRuEYHlNlxdNufGsOg>JlcdV%xuMhWl<752v&2B=o)K-f--^=BgG
zxL)n~`FVBtySs4cIwG{wVk?_X5hMkZJTFH@`@<nnh0UCXR<Gn%zjwDzBFBrsYF;q-
zh$r+-2TgSn(nzS`8wvB}=&k7Zq7*1<hk5eVNiBB|hQd+LsZ8lGG^Wh*Is!};IKV^4
zb~w1HNPE|qutW9M^zHwGP>PCwe?dukb`vufqofC(WSj~(P*kC^s@8_i2p9@`)CzR$
zZ#0d#Vbv}VP|Z+SS06+<4n6h+${kOBK<sVWft!dCML>uzw2P-XP%4D!lR6r(*ibM9
z3HIDYuL34y-N>5oy&6Rf4sT-Jx~4zH8zyJfbjIR&$Yf^U{U>q<r}P|LsU)GZ=R0J_
zxJ&RIhc1<9TLXekv|G8*t1Q*GcR6AA7J8S0RFiFK)l$j74qL&G!QyT%yZBpS7(PJC
zR55%0(~SAESLFXi95XVr{CDlt;Li#baqDeW^$7ue@HGTI{qON5pw`KrX2Re>!hPJK
z@Xbl0?rcgYv{9q_%B0FoSmC%qvhzW=<bwys(bMtW-PfB_iwo4h4wNUSufqj9z5&jp
zy0h&R<fI*P^Y`Zu(bMu~2-OfSY7~*aDn8xB9&PV{{NER{@Y*+urSI+k(6W5Fg{O@p
zv!^jv3T19eckV;ZZFt1!kbrv>pmd(UJ-&WGbG3ntmI3#scfHU)g9${6FbS)WF_M00
zd_;IVjW3}8*6#P9D%f*|z%+%%Dm5~$b|~(FYBJ1ac6rQgAGKXS-d&<$jc4qH&lX_C
zf7_YA`Uupy*<Ga0`2EXB?Pdcd6=uEe>1G-=?mK`H%c@OC>{BO!u9H=zPek6#S`$(E
za{8Tx?$AOkXMSw|dI7xcL926r=GKYMXx61|eqSYXToQC<RuIx#EzQyALfvxJF#wPc
z(`m`$U0w%1Gm^v!imeD)gJZ}P+m-#2!Fq~eG7qdXEoq^Q*|^7htx)GYPoR}CC7s6+
zUFZ9C&m?p=z0GH2kYK3e?*1w^lv|!<)Zvl-;Y!_t(J<7hu4Xb~Op`j@;a1=oST5kX
zB+UpIOj9U$JH6Pu@lqXIgQ*O2=lgSr*{&)28-Nwd^@jEFGUw^;#`DqD_BST=W=wKh
z2jQiM@{fn80%KqQ;@55U>F};OLLzP|Dst3>wY0Q%hveoM+uej1u>YL<aT2WCu8%X_
zujC5o7uCo_I)U;)2-1BuK?iLg?k7ZuFMISvA<kEt_-@tLrI)i3JJeG`bB?DAF=!Nv
z?)8cjAFvj2D9y4zd8c)vbffdXDtXqZ4xq3Rfw)Y31^!~4>m9BeraG%Ot{Tegp9&%6
zpAzL2Y>4_DE89f{D3K0RS&K^J(8Q@rSY0|eo9fc{vGe+7BmTmrHs9@^3%#RF{(IAP
z(Wkiwb*ojKU*L#0AT3huFUnRMP;${uWL5(S)+&u~dkcc-PcK$Y*p_Q$Istqsex|HV
zV+rS_&^!1%r%Vv%GCL53!*nzR`pa$MdI;508oqAw1sl<R38U(mg<ZF^D@V=HnfW(Z
zW85ti_vQ*X^+_r<(G)i$Ldukh3dqYPhdvZ7#mfksI%LicCTEBSn@PN=5vlK3=V8Nq
zJl!5ipPiW(%k)mFP}u43X~HDjUw0ZoEvjmKl6^%jd%%#)F3I{D%<3?gDsgee2=DjE
z@jFiWkwPHJp|?W3SMjolodEJco?yS7X31^6!pD5v)Lo{bh7c2kO67DIq<Nda;`RwU
zCcci8AJz{=T(<Y|Z_N9y6&?xi_=PSpj89Pcj?$RvC_~$SR$hMK&%b=cW~II}@9JdN
z_`gD03;pcCd$3=E^*BA=E)R#dyS{-#2OsVTHlWptXW;!NAH;QPe}_^?@t(G$ErI%v
ziuysXOQT;ci?8%RjQ2gC{_7pqkgYOWm(gVT{dXiFtv}^lxLitThMEBmOHYl|gLi#U
zK#4y#l-6K}EkMBdmMY^|;RQ|S(s>KI_=_9QOES2cMCsVEhbDydXxz`6xD?px#=I<%
zn$8zl+%hpU)dvpW%CV(oCy5Psek-qlP`(hTP=H?v^C&UMR2x_p);AeiYx6A#kyBW=
zFs;+Zlu_8EQjQx<%qhw0^ZlI)&)rv0Cv6&QB9rmo3()D*5G1lQB-JtlP9V1B@n*&)
z6{V=)?OcSZffgPj->ajU=vc+3(pJ|M8iZz;y;U%~+w(?9um5^qxQl1~>=5pZ6+D&9
zGi&WVw;Y|&bfQ<)x?ts<`DZl6bIH4+J8=0@vU)=o?K5bLw?g@|g2K8&bYNDRfYzmI
zNmIBih(jTL`|}Yfr{x(UvsOqe0dg=m)QkBRTh%j8E75kULSL-Jm|-iUe3M%(h+%me
z6V-b4E1rjJ0=OrFffW*Wy9^@9@e7;VU8NF57?HCS;wWX~EbM*^2oh$8#3;h6I<b(r
z&V3H@yuDG)Dx9#+p_xRLDX|$cnaB>`qmD<rrHGA`-<$9fWr~?~#YgRU#X{l=t0df+
z;-!A@r3-CY@q8!KMvsyK<i-n$|G&qUCS8M`TM4Z9Afys`8;m!hP2<kjN$sFpY{t-Y
zT4)+;H6zs=+z=*ds6*>D^gDF<{ux2#=<>KGI@dOox0Qn8+TF1qGw+q@RkNEKHX)T&
zC913y25d@%jt(J3uxrd^E*$$``(s=!fZ6XxDr?dU{0tFBZuEm0+iIRGi!M+}4tLS3
z(@mQ0to!??O?i-_V}ZGqleZ-JW(ha!g_e{8<&m8u5#_$iX6J-)z-evF*eTj$+CCZw
zX;wu|{8}D{W9C``mDaMDi3W}{Dxvs-)MDD5F-|vy(D~y9cB!mFY5v|2Yad)O{5Zg(
zU}Fzi{)qzr(4!MH1C8<#sQFEBXWmnsSlCTvNRCe^H{YDnrphi_3Rt*wTXVP0KjAp4
zRPNteL_O-A!;h(_>DjRgWlI{S2YAN@p=Y)c?a!A}!*k5%Nmf`Ipem?Cy=h)@D*E26
zDNOOn!X12Td%Eq&{!+7vudGEY9$|f^PhCDDwZ@EW^i{I#u{Yy58Kw4W59^=TEj!==
z6XVtU+EQ9ei6PrW1*Q_OI%IuL_E2#c*RUd2oXW0H^p{0TQi!>Bx?JR?LX$j8jRS_l
z<V)*a8dJc04RjkxAl4K<)HLgkx}<+;85!h6Xu<TbJPhfht9@Su(fdph{ETXGQsZ=M
z66l3>Z?bR=;6`DEnqebHqjITtspo=u0DW<Orvz+xl~_d+L;h8wO4yvLfR(!`lS$9A
zg5G;6K(&+*n1o2H4<Ru4Mi$}&jd%}N*Bq#&H{{3r(2ryb@!A*Cvi^w?8KN{WE7_9E
z<sQ;b)i{|;+g!#Q!#z)4lyAwTktxEkd8J`A{fn@?CK(?)|MB7C?r0eyP?jlST`ZkJ
zJ^+pQmCg3yy@kVvRZDg;c50Gd7O1~BAwjtPb=jEMlVfuVoGZ$4j>a;_|7FSw%n=gN
zEEBEsc072IU-n(m`}E}(i2SVn(CV*^8K@_dpLeYxDDl0Y^oE!js?XjZuS7p-lZcW#
z&7tSe`mC(#B$~BtRVA?q_S;IKkAN#1WWtr8IqR_3SMlw|_<eb!O<m+9D^`%r5|F1I
z_ce(0+-Vw!9RA_NI5&QpkSHk}e)wfajN0brNW;m}JzCI4_+8_&O^7pYT*ByHqUcoB
zfoYUCRxz}>|8VuD+f!)dt;s%%*ZK8@61VdG>(E)y{6C~TdRF%Tlk)z*4l(xs%OUo&
zap}M7Os=hQ*pNT3y8ZaNAK)RA@BMM_WI;vKC8{AX=;KgNGIeorH7w`tsr!=<TED-x
zWaSN%#D8my2~o%zH;j`S2jAmw$7sms#t4W7f`qon0b@7{`jG_bA}dlxqzNo}l?PgE
zt;m;@w(#ecKI?NJrsDTd`m(3$FsZI>l+|FjE_!rq1xH(KSq^SWZb|ev3T)YusvcP>
zYNo?0liJJ$FqDB8oo?c)$~(c(<?Q7*qOzw)2N7X;gRu0~lfE~a)ug%4nC+z9Qtx`U
zCVqjsSxpX`a6oQ6o0EndoL$NzK0fFC<!8J)QpBm)`(lCd`}(BKqDR2mZ7euCSVmPo
zonFd0`2tdgIomaUG&k5x#zmz&p6-i-l|h=$NlVuuR$Uw>gg2Y8(#<oYwMqpnTrb6h
zLg^!tMTR{K*d%i+4{#XBJqx{UyqQSWOt|wGw-#85jV$VmM`WcBC#)#wVJl?&T<F^%
z0I)Ugs|uobtF7;|1+IK|ii@mvtHPppPvbplM?sV)I-h@7(;6=dgTu1$M{3X#ey;4h
zzk21n!T&_QrR&E@z3rr^E8MP>YiwdLH_dVPNTf?~m-Nyz@qoLTPIii{ln2A3+hx|h
z2NNn^*(*{j{)o8!+SA&Ea+AJ}NQ2FyNd4Lyf44S3^~(=o4py%D_vey|Koea=x&?>_
z7J4?Ir#-v^t*J|2!R*XQzM){DTaj<kH=HT(!w;oB?xb4>C4YBf!^zOn4ae^eUY3V`
zb9#8bUekL3x_Ah*2>$r+X2N>ocwu${icX`C^Xo?$wwRA#Lzu%eaX)l?1mr`3#z${R
za^5lepZAmZcTHc^3rq#^q3NrzejVNw55HLdkw%Sz4{k*emTTD>4=f%QH4s{!4$l6*
zod)_h6eh+%@_u+ZI4EX~7IZy?v44)CQf!=0xWx`DQXD^e^2T`caICW}-Q~@|kr@Fj
zMaPAsBR%uYbilyD{{}hTgLJTR#@iz>lDUdk%p@GOH@rXGKl}J{F=%4Ph0E>9+JW1d
znXv)2GJ9H0%<jN}r6p~Z@u&ETmtZr(x_VleP2uc8f;VdFr;!7At%E@I5IMhsPV4N8
zi5L-BasFacz`pJ$z;z~jUOtiJVR{SCq@G@XEA0yhOOrv<FRLLl7(*hwO*Q`)5-}cn
zdhpNmDEQB{8*h3CzK$NUx7(Nh_m7n=espnCCunzhIeyzpXR!V@efN2tO(l9t<QwbD
z>wRDD<5SPYk->{4<E|tl&zz@Ory*l0Y?Sx-9#)=9<&88xG)nr=qNSAwXDcI1E86<z
zLM1pS;O#^<0B+Q_#?a5qTyE49*BW7MtW_8|nVd*}Qdu6=A+OP=j5^uqEL7b5U!7}9
z?{KZc4)xZ*i4d&qd>8-sJK_&ZQzUDV{80e&@53bDmtpV(SerrhK>Bvf7Z(W^v~gii
zJ(dVrD4d1?O}3D5x<EQJja?YgEQlMSeJ~xxwAsIhtwi>M$HG}ShH!}+7`lanRaUCR
zB#@9od9TnEDXi-L^yG9xAVYKNW`k~97E}SH@(wIaBAALQ2g=V-x@Eutrn`9bY29Q7
zNd7}O1q8$waAi>(A|^A5Hpc<Fr7R&Mnl(49rR#L>c@##X<*g_&Lvo$8uF)t}mKvo&
z1(Ity3Q|qZ0;PZb?J@wnRH^*DUDHr!30YKv-l4j6J!k+~zA_}eg!sP)f_&B76j(^U
zf>hGX7)=+U@<awq{h_2#pt>KrPz`HlL&ciJbGGqGA;~#mub|AX!LYmosU^la8KzHy
zpAftOeCw3HO~HsIDv8&S4U=khKi8!XX<<{y-CQ}UWIRSdA#o9*TIDG=o?pao19@@M
z=76>kL#m{|BC(daf*q3dRPO_EdI!|2P-YrQe|2V+FciTBARG*0$nxeWX~#d&Nr&ai
z36KOW(M&We&8Ohhv6m!Gv*F^NLGd@fh;KUH+<|UP<W5^ew~DIdAT75s-2Z?OHdJ0D
z&I|!x2mo&5W}di9qV2!z%E98>JJxeZXRd#JfkHj;uyiC?K1<P4Qb7s7#enVupoE!B
zFL3Ob1e+vU)AVD7Xe!}^3P*SfahWmXt7d80l{}3{sY7x{#$`RhB*c5=<vHrA<lYiO
zs3|a&T2*yNUCUW2bU%a|YR&R*ayrWj3PCzX7<(wZU2a>}ckuL(4E+DC<QP_|avg;C
zvR3P*69ibc`den{rys8vt@2yKDA9F+ML)JO?*N1z#q$L=stW+e^0EgD;91jHy7QN)
zICpe!L=)7Tas#;)hn>HZ4^H82uKhF_>snvYr61=JP*W}v`OV~(<-*S$MjLJF0?be^
z;;zpcTSs0;(#Fb7D0dtb$I}}E1~e;MmKsU6Q|6@JNFt&@-}Gvz%T<%y|G6T;yaEYb
zHKQx5!|}UAW|q>PNE*32d6hMJUQ!Su^CBq<?8<m4<9@bKDim~7p$p&z$j{cH${G7M
zzLPwYbEgj+1y9<if-;mDWv7qQ(p7!2xVbxSwpJ*CB$AqsQ#BV!nHZEVqK?QTX;!5-
zP@^}l{n>8R7XPBT<GvT~I&R?aQdm0ln0Af5=3F<wRoG6|wa=f5Y~n_h-5$<hCv40o
zxTI8&nXCY#P*bk4VLEI5p))zJc~v9kghA=bmLp|>>TXBg)FXn>E3?56@y@_9L1N1L
zPFq5(7YRgpeFu?<P)kDk3O~Xbup3`~BFz{8FgsX-z5ZD;xPOd4I7Bd2NvQP;rP>FT
zhdnaAE)zESyV?jm%$voj?nW%Ltab+?EuavD@X?rx=^MO7m@j$iYAC}Of4&Dn^favi
zHyJvSNg1{?nKDlEt0g$G?CNY1SB@u*<qKwIfm?TStlB+H3&|<~0mrm`wNCbTrQXPB
zv8v{drRggXc*i@P9Xxn%a<*ApO-K1%A3Z$~hrS;*B?4Nxb)1iM{>?nVL<C=>>rTd!
zI8RY<EETvnddDt2(-~lhUA-ns0T#o(!A~@9Hd3>e@LWW1!$PnYGtQ+K9M>(;7m+6s
z(u`(<zzshyH0<mDbaPVwjfM|jsoTi#q<#2`%mcuoLwF;z980O^r4$v?6n2(Wd_hnk
zg7N-`IL5&KKT9_Twx5jl-$brUwb3817-HwM$_YeY&Bz`&Cit54M$bmkY;$i0xR^AD
z<kvTS^qyX5oW*G6gDHI72D7=Y((U40-TUn5)#2&G)5B|Ct9>7iU77p)7Lr%Y3rVPu
zIj0CoMbXuz@$)jjxZ!QaO-+LOzU|A^E$gsM8kbW{ufI-cR5q?cK`(wKa%EGfuzoV<
zzLX2^u;UDbso?>T7wQcc!S~NhBW7gCtK-*ESn5dK({3_0<E!OGm`Sv{V@WtWaIMNl
zb`3aoT$cFoxOGs7W9YkEB-7!;vLuX`3fZFY+#=3FKhPxpd_|@ap*=&25+;bFjU_PV
z;Qpq@(LS}L^oul#tWN$}UbL3^=2;}BqKSaZB&B1nxSO6XegQ#ou*p>r2i?0`ZoYb-
z&!$<vXnxv+(6at#YMT`6n}{~_GDV4we(l_fyv3KJ*Sftv|E3=y(q(`{U$7Ld!`o(b
z_*j_6Kd8~uq!iILV2oOrDA23Ht|f@1pBe7$q5Pf^RUlF(%w224;Z;|++=lPg{=^UU
zgp&|V_WOlij%RIoVukCwa_OxQf9#+fxfAr-f*VtU#(*A2dT2H_P6L{Ifz^C$di`-$
zIwmVcDoQc7&s0|a>jW!GIS4k@=8_CTQjXi3556xr6#kCO`I@SZ_Q1#EHS*c7w?3%1
zb~xRE6y}XS8MX#u4xEBoj;gg4WL)$#;0~qJoqIv5H7uiMsWD+bT@A$LZCq+H1KFVX
zcn{ZZF>Th0o9Az)W#dJw7M4!VuP$OJ69f@+{m#L4@KX7T$*;S&x{GjRF!+Wh?E02G
zpOnt6okXa08`xRWnCpC!Zervm^V9p#$xefq9xn|uwh_Luo1}#<sabvBPLXmTC%<1u
z`a--V(EkD~AwT$)fFOty)_Hm2V3ZeOqS0W2zygIoH46QdS^{W#GtCIy*viWH<ygy0
z41vC0-Z_O`fEpvH6T~D%sHz{cWIiAR`dHT?<(Se`dQuP|D<k=qOffsD3TvVK3bfMa
zt{~Qj4pc}Cj-PC-QL_EiYv;~hb)1LS&NUsw8jA59@5TAc^L7bI-yOPqcM~}^4WI9(
zrR*|A#3N0>=9l~^2(1N@h5d^z0_=p;m$KgP+6y<5Jk;D0z{(EedTo%2$U8^0Bz*KP
z?ZvyvwEAseO!jFVG3QErm#`wn%DTxu`&g2&)TAGbO57o`rjeZ<=uB0E*Y2xt>1@(t
z@zlp^2am7`r;NvDvt&}IXY(7y6C9nZe5nkra^EzZ#G*O=^e|tfAyeJcJ>4j=xo@gi
ze?$Z`UXd%>QVLMvS4+_z{wb#tr&8u8Z)1h!L=dJ;4?AsNIVO%MQ^rno;T!8LB<B2c
zXx>^>?kS#7IsZ%c3U6On14*;@ZeD;%&|;>g&8T_dnAp^@rHyWR;_j^A*FY9ZK@y3M
zUljsfVt*m{zeAQ^Dk$KNgB*Y!12^8XIT*-;oy-0<pZ-F=7C=Ne0^iQ>nrqk2ed-=*
zNlpbB>OOK{e!#q)DwZ&Vr|S8-S&p`u7WJEIU>Po@;zoCofMWo9`juiyY0)XKWL35>
zRduUbX0?;}Vi^|dQJ7bJMugyeSzT}hntE{3K0F$w>sfndkUdigChI=u*6QumD)1xf
zIU{KQa>cx7{8r<iGSN9Fe{Q+lHfFJ0a`MHB{={W8RsxC05Dy$6bn4(XA*m_)ZR3a|
zz4vuFFu>nWyuG@^7772P|Harj1y>e!-8voHb~?6g+qP}nPRHulw(WH69ox2@oc!mU
zyZ^iRZq}|^m#cQwJKj0wct!?uYq8)t-1Cl16X&2(c$hA{#mYr$DFk*ci>57SdnhLj
zHvW=wJE>61gm`TYjVy<st=%m6PlVefFa=^fdm|RYh&(HI)ZFVsKnx1>h=gTwAD5P{
zJRLKej=6CcPA6N8n5~}lw(BrkI4BF$2Q=Nv6$N7Uka!9@oO%#cNdSzrn91;B1lS0Y
z%IEyNN%<AL_rU)18)~0M>KQv(RkPRemABTz{Zp1*Epz7heMSHcf(YcA15|ZKCTAP2
zpzzIB152+o@A=PQMY4$w)wgy@UU^*6U{(!t)K*SVEZS!tC55IAsmWtpdQd?Uy53h9
zkR)H)itja}Md5c8r-kq>QSj`$))ga^ZBdso>Y9VsGEH(5N~-@Lg-dUZswHXHa!dtI
z#u8>Uv;d=7-4TBfI%!uk?e4JuyC*#h(`4G%z)k`MB>5&_Cm{*$8WvykbTD3Q1{VMR
zY+#3RVr)K%!$*dC@KhwVVc$ilK+(v`wyuA!9(Pfhnr%s&F=TRS)e<t&WD%B#`MV#H
zN!!^Fa@EWFnaD)uYr1Te=)lO8@%Cbk$b{JT@L?2Um-2!oxJeANAcc5z?&zC5o?C`+
z1l`vFh|^3!cI5AN+0GXTLZleie>ydPiW2^pRmJ#UR-OH1o#Xz8Y<jHzf)K5iEW#Tg
z7$9lZ@RO#RtUf6}xw1sq{Pw3HRcs(x_p~j!%<uC^TtOH0^O;YbZ<UnGbq{~uEKOod
zLc#AZKm9xhbrA&k$GZUb4uC=y8M^&*{k&6^G`ZkiFe#f$2It)6bKCp@JE!mhTqjh0
zUt?Gi)+hMYcUrdh*s|{B#J<43J^${T_#~)_!3rP(V+X%}5>;-A>7On_z2oHoVB54R
z<{%4oxkzZ_(lC$wDH7z?Ba!eK716o`hmRMY4=?ZS&t{x4;|TT5hmmL9E(4%Y*T5U3
z*qztg;|JI0bBPqrze0e-4}`;q{dJbC*MSPiM9}1=-%r3=fh3)xqQGyz^E^rOd0<>8
zgQzeG@9U(Hs%N>PQBh=yDE5ELBrpk^{?0wrH|pgUzEB^>Ph5pfPPurga7QUwOTV@2
z-U*hAry`YvvY(tcm7w1%F9vc2$x39ATM(2-e3r~~Gw;#)<=I;PTXC$eAz`npzJE&t
zqYcY&kwP#8aygItGPf!^CxQT|;6X_8QKAr+0y~uqp9D)NmBBAn7vR3#S9W(zVgKO$
z&li<HyWv%0B5e|D$FA8QEtvsEWye)DhDrk(->gd;-)<|_1ht>OF3HYSpH`8g@b5c1
zOuMgtaY#d5-V58y3>@lMQ+B5sVN8$`fEbKPj-z*FyYCZt2-AFFfW<-gw-Y#1T&T#1
z81};a2mR${wq+p`H#PWQRQ<hfQ}dxY-4<Wb-opnu|G6q_gUD?aHBSG`86>L@S~Lg|
zu!n$suQie*3TlN|fpc5fmcYEx#2Vf7!E3XGfYcA=XUMW>p~N*hBRYt_DPX8(+0?tF
zU8>DOEcXLzx*Yj@n8tLQDD|{o^e$<Ytg*w4`NLMULf_7?(~eF1^@a}JSt@Ihh?9Lg
z6O^gGJHIOp<KEI)e`%s5xLJDxJ*U`(?PVvJ(QFFkkyz0u+YOpi&9mKMaHroYiadG*
zkA^AS)<hOC8A#}aCppI2Z16NxG>Uobi;drOq_6dzo@DMrcW0YoC}~i``l4h0p&eas
zKOQj3Ya8&5CNBO^90k+!KGG3XVK{SFtkQE?MiUJA3x-*4V;Dk#Xfs5y5es+7St~9S
z7gj=%w=Ej#eC3CyvmpSR+`7XQ#?usLC1uKozVqCtMsh~!mn`|%$pgwx@3Y`7uQ}{N
zvRps^fb|iP&HTJ;P8kZptK`e!CkV6%Hb{riCod^0n~OtPT5w{E$6paot2r61zm;2x
z;!5Qxu5d-kk3EPy3h_m3Cw_#-S(a-@7fZjle$5q)<R?alcC2-T-_GV)f*d}JO*u^V
z50q|G{i9L^E`fl5`PDi_&eXuzWaKc`O>Gjayr82$1@ug+U33w2=3MGBkp6J3L1>k~
z|6Zebp|g{Vbej`3>~I?6sd02{*M_&UOa|f@LNN^6pmE6sbwh%Ea4#`})!RQ<GzW3z
z<ukNaDiu;AQLD^J(J%IWr3;MHsaG-FRx-c4#Lp3Rpbp#X=yGEWlN&(~dA5gwOIUw#
zi@GpNws%dHTF^?`COJWsH>z9AK-WY?qB<EH(qXZ;&Iwj%BIC|=#8c?VDQ(-P6c<|b
z^RNzryhHZ;enDGHqw4NHn*PToQR6POggdc!3mvNe7@+d0ZwB<wm^;>5eIH3JcbfCo
zhJa5z{mW`B1m!410~we?2~(q{DibTHZ=LW??(`FKm-0f2Cg8Ay>7HuKH+ripgk3&~
zGsP(a^WBlFJ_#&#juDHS0$pM4^~K+)*a3&Dy-o>If}badteGPcEhcS~C2a$9r~%wM
z0yRc@p$n_VtH~;d+9@~}Bku^6Q7i~T6;H-Q1mtstR>b?c#-=3vHFlM{7zoPA=9I+a
zIQhG<*KslF$)ikNp6ArEE5-#cX&_r8Z6`uothzrQU2Z$Cwl`)r*sWG0MsS=9UhxuP
zG7Uep7I!|QZB@cTdbR5)_jMXO=qZ|NTkik2zj^Z;EKe{PD0qLoWN8<RdKU?1he?w4
z__ZDttQxy~ID+XY1Noi(t~cvmv@4N1jZPT1-6n{P|CMNcWr+|f!6@S5jB;Xu>Ojbc
zDp}20C*9P3s(7@jn>k5{wu|mQCHP-6vj32-A|^D^=b6qHfz)|Qqn_8Kn*TEAt8_3|
zCyV^V_8k|;Ew;^UzOx!UuH8xZXlb_q)>TTa_LSWBc2a^5d{>Y9u$?PP-8UmAeJBH@
zASl@8MY4H`ZsJ>+2Ku>X%vY=iTa`|vse>ynvq~gA#TUWI3PgYx9M(lP-)wD7OI+}0
zDY=`^#{?h2h~@MqwOwc1A1lRcQL3S9eM0N12i+g0JPWD$dx#x+Y7MNtt`3-adMD~W
zt$dgsRazy(AVgnP7sK$zALmvtB(23=@H*4|=nwRfgj|*H<^CQ_eNrXx3Iu4^)@{9F
zIoe_R7-JD>bET?*qhI^=c$$l@FxK5K4CxP~uw@DSLPr1;z1HN4Oj(lH1;bZ2!Jf*y
z4a#ndkB?EfDoE*^#aEe9G*gV$@w>c$jqy^V(el&!hTkhLdA<{MB*7bkd|2VHMDFNy
z(e_Q6s+VjcS#L5v^2pLux3mtrR9GC>oO+Da%{`nJR{I@}dcODst=`$R=sD8Nf3><$
zwqKR5>u#a^wGP8Fz>emunJuZ3*92+UQtdi9S2vD=6lbD40X&Pf!t~4B-rPc&V@Oss
z2ClS&H1Nb|a(g+Q>YtE?uZz7r(+R=a%LRrxh%awSVmU?_=ce2+XRl{HtBULPQ_b{q
z@b=<e-Pd2>9(dpF|3Qo}G5qf<JQD-we|Lr7{2wvG0pGJ+dqY6cOcCh?Bp6VkvJB+B
zF1U&Ajcqey3BUgRUALfJLQqa}wUy4x5o?-WFd8SXE$1)e^wX*IxPLm)`Ty0({yM0O
zFo0eHyNQDn$dFZrz+dj0N3xs-ulE0|5r27e9)9Z^7La)O4j2bNH}E}X7V*lDs+LzQ
z*~W#?X@TRv9C9;1*Z3lV2f|rL&hPX6Wpd55l;0~a>^U!s%RhxyJgiG8=oW(}P00Oy
z5AV71w_(3fmqI16$cyIZpZ(>xbH4h=T82n?ll&Frw!hO>M(stt1QQ>Mnj@G`DX9@Z
zO>7alw#EU>_Da7mr@Q6_JK0Id!5Ve_JfHuc_Hc+I5sM_z7G`3NbSfc}IcWSMc`_&d
z=T~?^WSOC;<}|FHiB~i`eY5J-4*`cLM{<eWgPv3K=b?f}(p{_*=YUf3SOot8)G@#5
z%;um~e$=O@qF~Bsw36s41rA;ABkUl$&Z<6~f>bAI>-}|g+=4&tK*J=$>x|!$y)i+t
zR@H0`df!KQk~(STm)0*~aCef?V^HA)!1?70S#pp{G-20J@F2;8EHw=SmiW`Zovb<_
z8zi&_$V5G2C7oNVE9C-(1Ao_@EOtyEJ<zBP{J$q*w$!F4hJ3H^_0)%<&0*&W+DCAp
zh13wSECOOT@cZ#v8o(04S}a_ulYlN0XAz2HtBaNlvk<Z+EH-M+=?E$DAG~aF^l)db
zE2>N<J<O*|Zzr%P=Y~o`R`=tPfOfX>!yG4{QJtK!Mt#;yA#F<tPs?~A8ICe(<IX33
zrYdT#!&q+Dc0iP<Q~MqW*(s={b^Us)>#`BW>+kydRXCznZ8}PEj!@?f)y^)v?lXJK
z?qL}E&4eZH!nC53$zb%XLZbag2bdOMRQ&@(kl6ghxvmKxXWYY$vG9EY46CYU$gVsB
z_z;Psix_?2-S5q!g6YQubXMZDCp_IGsfMogG&7M_sYWA|d%ScdwYEtS71&hgO;c5K
zYKw(F52br7xy=Fq?MAhZ-DW8`Ew0lR6r%O}!j9x%s_gWm(vU&t8I)vo>5<&kR|j-4
zQh)Pw#8-&HqfFU3GHG(34fmXLY!yDF-+ptvg`X4Wl$z{e6V>6pHqy!o*MOX)?A5Hb
z+^v@+-a4A9pf14KxEhIpoe>`>%28@GV8ZEX<OEd`>uMA?P*F$f)UI`y!u|)|;bkPx
zTY!ZZ610T<s})IZnLC6;?b0&s49#BoN+{Gs&;3%aHQKbX*^Oqbm}8yP_D@q~XB~_*
zg~f`^wiN~_E+X86gY7U-HnA`~azgzSty&mO@n!z82w2V>p)o&a)Hfp;im)^vR?@kw
zy(#J<YG0a<XWW#f17$x?A>U;&NBM2`t%q@^mh#fcX!@<2>*U8|iXAI=>GVUGBF2fe
z?A026+@Dy8<^7Eg;U3;vwm{$>)2><-^vhol1DGZVC!coKrxzt$1Mb0{q?rkVX*;#|
zuM%J?jf5@i7g80-J$pyZVniv1AsOnrgX>4qg}|n0Rdp&rs>I}l(Mw54YPOWSU}x5a
zK8uRsf8o<jY^T7Nedk^|DQnj~v$Fw}9a6Ludml^<z+q<nBi<fppY@CoULK+8cd4X*
zADBd~a~oqR1Ze~$aaY<Ql8%&XbdVg3AWqQL{ZQCfQ{rH4+iVGYfnpgSk~Y{t0Tf}m
z2Q4WghE2R`>y+Wve(%h2F7})ruO?DYZnbh`y9G~2+R~9Lc^Fq~!N_Jrap7I)#s0rv
zzNF?c^!L5;Q^6YS@lc(U7qixmMLH3j8@R>o!jyYV_$#uiw{E)$iM^A&$h`-6x6p=R
zajp)!*`j)4uN4lo%X9UX`<N=)E$ywp*DM9KcQ=L*;G$tmKMV!$uKwcECp(wkR!HtD
z9k@_3QSI`Den@9hN&_1>Cs|?29=>UsCPlh4e1d#x75kmWFpVdKH>3EQ%|7~d5zhQB
z>QP53SA8%}&PNPM?8A)iAY4w-tr=EMT5N=;lZLVJEY>x47p}6eE~G90$CjWEdFX!$
zLJT7O3{^Wm?eRCN;g#<$2uqcdc<i=TsE$l9kd(=>7|aI1PRCJ#Wn59@!V!Z=$4zr8
z4_sv@xr*j~9GI1OM9mkpKr*ue*`ZXWDe4HgNhu^l-HTVkZoPisX-CtTeYNg8gN}zS
zzSyE)(2XyW<83`;phCpF+hrQaH#*=`QQO5pVPuuM!;Wgo*SLHMMhHBSq7mKAHiSvl
zF|NmR7=YjAf(vzo#sIUu)TgNG`G?eb&qqvIVC)J@>A8?nR}gl$jux%GJfKMFXQjq{
zOzlFhBSy!{@c7h2o9<SQV_8!E+F)CgzVzVSNrmcNndi{Z_y+4F$;Aq~4>`m0Ts%!n
zDVIS|h+(p?F`HFw>z$yfHjwT@KLG^lg!c;+183l6Q5WG_m0kp+;%lcV1s3-0*HAIk
zqbyFeqRF5=WsqHTpYv9%1WR^)n+Zm;_3spS&EiT7{oa&pL%(Cy+KUfeacj!%CM;s6
zyU~*MvGTAdKhi}Dtx>pucO^JsI<#3@69)|yi77$~Um{gK@gZSGrcg!rvAj^`3y|~=
z+RHZ_tvl_`ZJMR6Ijk4zCXmSIWo>mO-@{NSpCr71GZvRW9lO;X6J{1G+dpWLOy~2G
zHUy|do;)*|ODI$A>UTq4g+LVC?QXmdlRw=vQ<Y$etOx(1r`+|~tRveOMVd+xR@U4X
zJG_l)z0UI&-M|prw_Xfm*sMd;S|3(lUe8D7{eP_|gcScL{lWOZ^#>yd`+uiD{wL|O
z&xY9jPW=EfWmX@BGs`f0X@?N64ekr$1kL39KQvhcN;@qH%?8t<U|gG3)SsM7vvT6*
zvz(hgynSx?F|(3+=)8gI^J_Io578?Q2?F+}CXgXw#|*(H0lgB{U@rStd~(D`{m1&P
zDnTwH?hMcrK@W_BPXnkS+>Bt#Qh}s(ym>i_eqe2_Vb-#@rAc1jL%;}7oWI|fTfFAV
zapPa<L~Q9GOAN_KQ^HLCN!@$Z%7;vzZ!@}#_9V&YljN%r#*~qJf4pN{qNEsMHt1&`
zw>paCrGZ(XZEvH=QPCHmD>N3U#ZfNeldc>zO~C^yLH!v)0)ybuud-WHP1q&Zh2$GD
z3TmAmm2W;igl}9g7gT6lAM-va6QJ)mM@V)^4e)dB;t+^4zpqVS?hifXsU-#dLP~L6
zi9zr56Ctd8R(r(+2A|DMfdY7~4b2hz`;IxTMt)fr60rJzpUX4Xx~zAP(g^ttTJegH
zB|E%6WPbCt4isIXUZEXQJqPs%$Sik0K1@tkYd?h`pc-jJ^3F@M-nXG1NwPxR<fF#)
zczA_qWTLd4xPc183#eL97@xnQ(N1iINwLLmBg>9kC8))_O@mL(Y2Ph#(8NJ#v!AB>
zFc7R~8xm_pM-gS3sX$}pL#$|MNXbGJK-Cf?)6EkDz^XWUd!^MgB-=11gJD^2Yi^cm
zEoBd!t~f!ZYEK7!RFeURjM&!||8}iDq1%XAs2rM4NtFw;%fS18DQCOiWL$+{b=mY`
z6|~1D@R}0iTm0i-9yC^)TZ1BYCFrbyp^_heN6rYk9go+0e>W2#9vWaOlt}#^<4Kgw
zxDJ5YFAYTRe(D;oNqDLL2QZ+eaCEa3nKnI^AWpe7!S`yg3K3hiw4XB_aqCjcMHhyt
zM=t_2qX*rvhhYHvux413IYbmBTP5g&?a5Tqts0h$^aGiXblVxjl5ZoO+HrjVe3n}c
zqYhkcb9LzZb-OflwxS%#p>MQBF$R}?PSl!-kZFhU7g*H#C<nCklKCXi`x%$KbjJvs
z$Tqns^tMRt(Dz%a0JfGKz+g1L*g8kb18Q+j5z5n-=sWl{z~?-1+@OFdMV7!~bzux@
z{eqIb;h0&SP}($LZ)g_dK7S~!_0x2Ma>2uY>5&PPV4j#@Lh@7St23+P@7*><%5$jD
zOz5}heLYG{dgO3Tq`qXs<4Y(2GPGjYQ20ZLB){8^qTAjYXqGB4@eG_$?P9#j1xP5Y
zfy3`qhgM#!=ctCTB?hAtcuxpHBN>^<X$bk2kJJWLCFUfQPZ%fg#5xMaKPK6p?x|n!
zf{c3nBGx&S&x3MSi5CZlQ>iYoOZsJ$sam)rs?<<>{tTRl^gL?}#0?Qsv?eK`R^5Q_
zu)?A4qDQSzhaA?iP&8JXc=3WLujqIdXD#m=Nc{H}7BBV)<#ND2Eu77!lM^y{6YHWY
z&j`=q<yS#l$8P}Z_L;ybS(FWup0V)lhR3e0Y*(CSO-Tg?&WjaW=}G48t(1bXQJpWb
z4VjJr%TW?_h%GAb^R}INpCV;GtPT_hxO&KN^xHjLOZYDJU3PRdh=-$<2>cZDqMN>H
z_cK~5UGH1NDkYWiOM;wc(!{^le2KDzRU+75kN~%t2MxuIViNg74jp&cR?-60?M!F{
zj6HmFy`e{G)=u`U_r;*;#>h*}<^@!U`3<WJi)>81)-CgMCe4@8Ez`Sz;iJ|puHTLV
zpp5{+gVYX<JqPz%@ZApPffmEGm)Rmt%xT)qFz1QOP+ktw74D3I93i9OaMUMAJq{@E
zj?v`3*ic_l&*HjTGybwV^&Gs>-C!6%)Oq&L+snVGPrG9=1bC9XkUyW#+c8V26L@6)
zi^B?e@^X$tuKj9(P*ug=E?z9%Nt)Y0kYbgPivP%TT2`8s+sfmNy<6v2F%Z{U&)gIt
zgPs^B<2ZSUS*GxHUZJE)#7TtGhL|fu(TKj-)X$QLs=?ok#a9Pge#<$at&mb7M5v<@
zy0~iRBMyxP*aP$#_ovK|RGd8OhpdwJRjJso<R*7m#88N~rQ3$kPJGZ)9MEdKU4RN%
z=WX_#u}6QZ-5|O%5(ZE9NV3RQf!@tUR7PfRku^RhF!dxfLfdTMz7kBISEr>r(JGQE
zHR7~f0@sBhwI9o-sod3T8Ein7a|I)Ix$;nDWT)z}#saT(&-`T{Xw%bcSY5bpxG<cq
z5aLnZ;B|BbYdSUrj@|^VDl;}DI?!M|ui|2DKM(c;T%i~1d|Y$9Tx|1D^_Z`8Tz2Sw
zKul-mcg}S3o%Oc6rN#4^wYFP=497uJmzsf#sY1%5+Jd1kOc68cAnbyYvb(9$lW@+T
z!0Qfgc_r#!Wcq6klro9cS8_M=?Fe^ch`_+JL$~E6J<M!lie0mffK89tX7wxfnXzlj
zxsj#HV2Y>;xRS4`F>S5(1r170j0`l}L1j)_7jSVEfrnSVwYk?B`IzGSs!!33ZTz3k
z=%2Ed|7BP(Gco;lHb2go-38m-Rjq;@b01HJDve#hT);a0XF&VV4Q<*y;TUE3*f&8*
zTU0@6stso!jFZB|6uPy3c>}H=&cV&)?%UIihcb)CE{g{4mLFb#fBXkjB1Q~8O#lOQ
zSnGCpZf=pt;BpFmq7r2f&G+R?4ZS`phaWJD{s(p6;U7{+?-#64%UUMOxV_Sw(CfbE
z<GShhZWbm11CJZrJe{xmr)!Ax$m-;L<h7T64P@9#B{7UTA99=;0of0+P`b-O_8w+4
zFG+vpm-penf70;n;n7WZ6*LO)3!glfqd>DUwpLc@O62#+_)=38Oy06-?6%T5T>|eW
zFo=Ugnm39-Q_yo<SNi_|0@SqDNJbOHlu0m(Sal;i2kg>w%yq?Op)QM@K$L4^n0?sw
z_HH1~?O*>c+n+PhYb}WZol@~8ndIp%6_u}1yWM%*J53bCBm&1+6(iab>ZB2h!N%9G
zMn^V~<VhLNn?bWZ5txfne`Z)9bijUrH~@oDYyBnZDvAv{&ZJxu^Pl?}3?eh>WjdOC
zs5Wka-iEEWUaiO3BYbCamYOLy3Jm)D`$O_0S{PRS(CCjB`Tel`d|%4LGai?>jpL()
zQim;gS&@mG+(4WOyT04RG%1%&7RP4j`r8{SNf{A|YY67*$f;|M(}c`c-IM!&!1H3P
zB1S-;VTjn}Wh1Qp6EvY9dSyg`A1y|~2EXSGWl16Tuh<Z2_+Fhf@W-);s=MhE5OF+O
zVF?w;^O91Q&-R&v6>>NLY4DMEnnI$lfyA3&F{UuaM_@p1K#y_ZtK>_yOw3RAmEeNV
zp^d!MZS$$Zelax6=VElP4yHb&-e}@vvA=gqX{3CKysR2TSjMHNp>Z{5ldJwPes3}9
zCM-2zPc1E*SM-LZq&Yr`_W=(^H;KT$1--Cv$+H}2Gwez2OXsf~Rmm);Nso^Uwa6}<
z9^csvBn62Ca)&ovMm?5WPnVYzTe@^i*$5Y{^42Xc-4fQND>gj{?cqvmi$4HhZ_L<=
zlj9B6vvMxV)oiP25^XR4f;pTKRW8@#9^w1~^SU4kgZ|54qOf>|3$N|4CYa4Cc$;iA
zKZHC|N|sVHy#%*Nuy$<Gf0cK3W~w$Z<uZH<8*ha>3c#hL)ytx4!gW;O4ktptJ4{xi
z6q{CSJrkH>J64VJ$5w9;W7Pae$l})$ntUYfJp`QT#fwbXi8KHF4>+WuZG43!myM%z
z+NwJDQ}%hEk`;p{((tWY;zw}Q^>EUIyU@z4k|A{qdSZQrQ6E>l@oE^5suh45p_RF>
z9B<?z)2$+3@}2i`s=AIEoZH7KOvxRK80{`doNV=Nm>r_(Dkp)ioRaT&|ApeZG+hBf
z)FYe{7V2D8^SWE4zLjD~PgzrdtYS;iy66|VAC|Yzu)PKFv@<oq;x7RGQmjB*Pg%xe
z$~6*gKCd#M<4nbz_DFj|kq+(X!MCSPMl)E9c1k;aL{zyOpq>{7Yu>1}KxRPEGonV-
zD;tS-Zb^3lxgPyoHDemWo;1;K(?!o!iQafpV4pnP6k{3!@=C`C7UU>A<{j3YEF1KG
zPjkdRAaf`QO(7Oy5&Ow-e%f#+$8z#Azsdi~af(ieWk4s6TZ?v@$9HnM@NhztbKZ6Y
z4bTzCIe7uqSjdM~I@&XSbKpRn>!`H6@0X|^aQ;;7d_c$|*O(X*tF?M)Ul>Y&oX*<7
zzTj4l*J)R%y~=BLs#3xWvGw?nI4--C$~<mV{Aok!3Y`;tNi2LCi%Ic=m#=1>gqI9L
z-F81id@6rbP&*}~J8sifjwJ<KQNva3T{dydH-g=+uil=>*Jv?iD*kg?F8BBey$fby
zTVF#Aa&AHr&rQ*OY)Sm)wTDTefs(=+RL-HQg3?-1p!HluOD@rZd+8a?N=8g0NGnZT
zgE89vO4ZpJTt9--pSRjZo9T%WTb&;X_PMYBEkg@EcZWWqZP0NT-e8I&P5jZVu|^9u
zfrOwxg-z7XcsuTZa#k+6N~y#RLu%&Ex`8z&U(XoCN^aJRzo*ak=^DK?=D#_Ac7Bp7
zhzmW^3SwS_F!U4c*4dVQ6WCF?`j#V@e^5+z4&rEmr5||{9ST6Ta<Y0E)68QG>Ar2>
zY+!~_sd?JNWY5a|HS0}BVVL?WF`<j)`g|VicC{U-U5bpS5#@}?&y5appJwjl(?bCZ
zPb3%Zy%=4h!rX@TNt%9i2J{apaHfl|*r2dY?esK*f7L#vtqR0rQ7XC1yU9uCgk~8l
zciPvVqpB2zZ7(f9Fl{IFEl=(C(sK<I)`VFTIkd-rSi~sXqL~X?o6`rZs}1ymEx_A<
ziZ}0|B$etZIo7j+kIJ@KWE$9H7DS5=TP2R;Is=a-e$u?1cz*WaQZOU+ZQk}`ImJaS
z`k9w$|AKDBP<ONbb)SqIM9l4{*N$Qq5U&ewVJ-nz2ARF%iI+CtX+A^?PvPa_{>lBz
zNHQ$3hfU#eZK|63F3o&nI84{ybKQNL5bPZnq%$+WYOJ3~yZDL^bEHN({@GHi^lJoC
z7NK08dT0%5eRsiKW+>%kL6A$6FRx8d;H2KifUf3k5_Qj;x)dR%ERZEAY4i%I*^a@7
ztspgrhS<uLqcpax+qiEe7Q{R=cw1|D@IAIL1+$eeMQA``U04(A!ci`-V#vA7Z`6K9
zy?%^J4sAZ>ShO59+kvm%Rpg+(I_b99Sn}c3^YTj0r(A(%MUNw2ZlB4BX~#{yRLXg#
zFb|~p`^GUz4miTRH=7O~5;}kK{R|iC1Y7O4Ja)$x9-nJ4;W?>oe!QOhK?|teVq^)j
zsfS%W-ra`N=IZGQ4~TcUy>sgOnPH;|XB|c8l7VUj`0~p0k8ELRhRp>f9V`cUQA2z7
zxi+jr2WVYHflRm6wtQ*HPl+N{NvM{P{;EI4#Sq>d*OQE3;x0a}WICD`lRMIcbKt>R
z<R+Nh)4bX<JFh{#vac=+?X+tJR!3#JA!(ezy;sMVwhEUlwr1gR0tfB9jS#9d_PEN>
z`<0hLS}=1B`CXQSPUmOifFp0(!+hWz_b$hzqX(Tp36ER2`Qfzg%=fRw8kB$kNnbJl
zZ+-PM1NrZgEB^z(|CxdGysMuegiH{H(-Shpb7X>fo&RnvC@ttL<#qY~&KGu)P%JW8
zt!Ebq#5dtaWu39d#}$>^Cg=3!<mtuZ>$b|ux&hD1*y@oLU?=enDiXCOgCML5JG~rv
z_<`S#10H2Cq{>-!V}EEJY$FhGIc@>@f3AL7b#R4z!k#Qrau?b73oZY0&9TouE9;aV
z|3CQs4Jswxx6k_}q2`Kw?QpGBMH2&?%u+t#nyONPk%SV7W<)v*32hJHkXaSUh6J~_
z$K?kH!u{cVGWOF`ZeP;thr0&{N?Zk#1dNKf@o;~Ay1x6IL-U6&w8;D}JYjz3TDkQu
z#TSXhlA@Z`tUMnrEV&gNluDK59|mB&QLVJ9JZRbsUyTx*x(S|l$b|e3wZo!RBHcDr
zar!RdpMF*ROG-m{NM>w9B24?)6=A5`%Jot(Zj(SyauKPIS^AS~hqXMK1-YJ~yd)p_
zT^@NvH1hJ$_%e)5&q0NRCJJV$*-^Av<Z{0gMHRM7gc_6ek~c4(23Fx5kN%}{aW+sX
z#F*jDxvC0dM4qiY+CxRp_&*Mu1zMnDAk$^)K|4$J=>S1ZQE&0qcpO+tiw)6V#aT!e
zi-g<RzoaEsJ)Y+Rbe&=XL^Fv}X;A&S!RgH;)o9ddNm&`)c^8@ceA#vW;&wPO4KkGE
z2<AM8Nd9sed46nRpnSO~+v{LkiZTG-yp9&$FMm-C@TGz0u0aYf_x0$8i>vP!Tvz^P
zjZ8)YtRrm2#b%*%2jMMjgS6)C#Sk7ntSD~+)x!(~;^;V;BWlxT1#-J7^NN~S-75VJ
zV%g5zb&uAtMv@>i08I`5z3V|9sqa?N-4YZ^#iC+{yIhF44CUiWLU>3X6FdeUf#yG+
z!HQ08zJZJ5!r#$;5^r-z??i7!+SSNQ>S<I>xj9WBIXon_j*D@EMXFPFNf$3;GJdvu
z0P>Z|jNz>Vv8CFTuaP<P>fsvMx;IBXN2rBlc>`Gefc9-d#iyt$5A&+dd=S2?$RfU{
zzc$f45Bk#m*lQguQNy59Xg^tR2Qsy00QnornnaqrqJHF-15o~Kq2@84+WUo*7#0m}
zVY9w<eo?b*bY6lQm1B42FoL?K{J|f3RWdtm(n}cnBnQ=K+|YwjqKMJi9lHXjiJFZD
zn2|i@&k>o!{sqlP6KQ0FSN#^R(#VFoo_S2ffh5E`@Xyj%u3`0T(K?a1Nu-{i&muZ$
zl<!IDqazI3DJy5qSMqME+#*3%>0(6-dykPs(P2^_$Bwm5hEP|h`SOg?aeuKWA-7p$
zT-BX6M{<Sd)@0uN0RtPXbCaZemeCoM<kTm|Bo$A7B-~}xQ5DBoh0@r`jk$d%(7P&?
zdS*|^EY%>Q<hSAE@6si+YB`Q|TUM($qWr|Rd}p4PeuZ4kA=rtoFq|dd-E>?_Uf`x@
z;G-HPbC6;O)d^%ItRdw>@C`82RaHDCsR<NwNz*R%deqrNSExrrS3X$;M71d7pz{NI
zw9Wysdi>oErm*&<Cq0_QYOyabmxTTwHET*O7o`A{t*cbkj=|=s4)wJ6rZm|&6-<RF
z^);_EfI`^7E;agA*{$#=7YexfRAv4x4XHIqu{U(iB&edytZMp5p&1E}K^hEYlIQx#
zCib{J#ny23@Sh>DEpK^|cSSLXmU}q$FXZv}g5kEVAl%K!0fnLEJ2zjarCx=<uAz<U
z2?gNUnT9%C*a3iKk8<92Wt?+cY%1xYMc2#e^twtu2EDSQ7-^`gpnzsSDp&JN`x6#C
zc~%=aJlM&cDns~-LcnSc+$}Lr4O?Yq1$w_g>oNIIY4D2^Ir{wy&jy*zH0xj2C2dae
zIYnF@`iuEWS`Nisi--Momie8-ardYwA#%5|Pi|4sf}LsVp)2lgz6}m-$KEQQxy7?D
zfib@o-muLFy@bR{0hVs$8>WY?Pl_&YA>OIgqN7b^HS5Rjeaop<?vyGn%xVX>E5j)y
z$Kr6QbSiYhF+7ORyLDcKP1}2mE$+_7t!-C}RnOg*E{R2#AWSPP*&!5P?P@1mqiQ+P
z5x(=rCStKD7jJo|p{BSAHLCmU=VQL4+;nlaW;zCT6Dqqaf!Fkdz^B?0N=%+w#_R^R
z?W%78Rbe0W^#ER_jE^$G#gT!Ty_}QbSe~hs+@mV#zS1;K0(Nv?&Mic-|7;jl>sQ_b
z0%S-j4v$B#4nDcM;3IPO#Gy&X>~$|&(#Wga$97?v#MGI(#AUg#gG1l~`zN=Mf(!Ur
zdR1H%9d@T8fy__;10!7LG?oJd{3Q?#4x8(*xt5mcK&4b)yNLI<yQa6jurugHBlYx&
z3TsaCcCUbk!T#19Y;^MEmWI2}&n#t*RZaMJj-|07P|!82R#&dIG7=Y$n2z%yKlag#
zQZDm@q2@+_SZH;|L1FuTq1#fi+|HWaFZP)b)07#nETw0&kS(aJ?ZV5?3H`s!l;qzq
zrE<by|LJi4>9zY`!w*J={|?2~ttDx1%>F;ynqB;eG3ZJ#h?ovMj)7ZR{}+Alb(Vks
zjZjQt<@+aFk_kfFZ>=rjL*h)6?wiKx^>Fv~ad%>|Z!bjr@Zs*wR*{A9R9o__sg0vJ
zgw@LP)BSZ{Qp)!9UWOnZfceJpc3D0&U#OU4Mp1+iL*?q_lDCJQVeF!3ZKG#wk=mnB
zS;jl!r$tbd;s~)OubA6V`cq5e8$Vfv+CQ9IpPs9yNATrcR%J~*8wN%T3Do=XcrA<z
z6gtpRM`1PBarnMH(Hg#eWJ}E`X?=$=e)!Hmv2f3AO)mk+X+n3L&1nIiqhA6y+*4S>
z_envP<Z1(i7eN4$)Pj@sYxHcdBGfQZj_@R7kR@9MxI&wpw`2tl8huEg!tsrIuP#)%
zPA$!8A$ra(2tyqiX4t6t%7QOM5|htDJ^wF<3zWRCjr~L!kUsx3PD6z(r4pXAeLeds
zp_){O$;A3k3@e`VmOBH)@KGwPjC2FA_4r_6`atjj%4>D^3(c<uU%cXaaldUH&#TL)
z3%?ji?JVx$+ERRHb5DF5fbqCaE1OmDB8i28jyDD9NjwpjM?j#V0;i=(8_)>yTwTbD
zxMv0cbriUvc%;U3r&o{i_1vrkaxlsiHM7u5mGsDoy@GQ_qw1Z4L0wgaL+8z0Z_i2Y
zx&$9gD6Uw!bg|@gz-ieRGJ|nF%Gg^MwJ&q2{Pi#oSQ`&BRrOY@)%8Txezv<1KMft)
zLY*J8vZJ4xjF(05pqWZZ*`Yg@iHKFVs4_2DjWJS)lnhh4F6znNo3+W$$}G#RWkge?
zyF-$*$no%Bq~XxA+7+(%?q%itweTA>EeS=o8`|!GOgZx|9#6&a7z|dZ^EulDG_{o+
zBa1E?wyYMXMn=l;f>eL}w)mP;Xn!a4+3YuRHF<Gb{iAzeU5PT}JR7Iq2|U#nVb9Wm
z;1m4qIq?1{nYkj%;BRzD(J(2$9t1IGR7hRam4=$lpgc^vW+}C_$T1$C(#utAGK7N$
zLSlbkDmAqcVtr1gg&dRO{rQBcrEdFDRBsOx<lj{$2f^F((L|^qK_{Mz?G`1Y2~Eb>
z=A3dZ+YVPZ2mAXV5;851j=u&t8kDr%L#AnViw}W!uA+cdaFovB3;SrHxm&EqR=YDP
zK1hn#xy)Ry$ceh=hzxh9h*u}l{M-DpC<p>{prGfYJP;IK@3TSHv)}G`8A^hYYE%)x
z<$>g7meEQ%Oa&sR&(VT4z_@EU%jLk8JVugY!jhHt6Vbyfw71Bmk_e-o|FjZAVd}tZ
zqG3j%3%N)2I1$UomE+CWAE7xh<-E$~c^hKcUS3uO6<a7=dFxGF6*}jj?kV@&8UBK~
zoe&VSW6WI8fXti|xnH9pfmw&&->@JVGD+6MV@WTy#tiK!49d1Z3#L|zBSjVeD|v&Q
zP3H?b?wwM4V3ta1=2ILp?2V@z$+wj(WxVJeaE6`bFkEwh(~dvqLypjrO2wV#E>+yH
z=4KE)kOX2~G^$P?m_uqe1CA<i!aXe)1|QgzoLHHnSA6=omb2hUlTWo+hDP~R>2iMZ
zN_NL#V!=f4o9()ZDe4~&>F^lG?ttO=G22nB*j_ZQWSXq`D6u))1X-IbU-{zvK^$Kw
z-%$aX>QplS_)0YUfCaZTNYwt0#$M^3A$h|FgBWtmLyfuA9A;b-=Wu&cxke2*Ts9V6
zjp%kqmF+R$34f;uld+j2SP~3z>~cA(kd%`3P==eq0nN(E>BDd~Pmx1P1VWF0x2kCw
zI2!+SBM3I?HDXt96w{1-bX0QvQ^5y@DNCKXT7^x>PuS4stGbq*ad7%~YQ<3mn5}?9
zYzHz1ljR`20+lqZ`5JyBN~3f8nk(PMf!ZRhHtq4XV`WoKIS+5YW6U81-Fsc})k9(g
z(Zb7-+96jnAM@;xC(<}|_{?J;ic-$Vq}7OUoJvG;RWJ*4V}%AP>4rEb8Jf<ep2Kvc
zz*2{M>sdsjkmP!fc6Tbv-kC*vWqZlG_!Vv4u=58bkF8uQFvmG`^;XzohF_^m+`0+n
zSly9P;Y7Iy?KZH>gH2&Ve|@Z5`J*0Jw)s|)9Cdp-9u*aci$#E6vP-c4*%ied%JOXF
z^6ZkB-+Y>-0St!ev?6s&6(8UG)m<;i#b0HGsSmXV4%CuE`;v`R)b!Z7f-7W@+H@og
z7CS<_t5BHIqn3w-KN%J*Yg>N?Q8y#*{`JKb*=bSK&5LQ*L6niNOl+8${@t%sa>ZKd
z5WCnQu?9P@4A7-uNgj|+!~E>$hNmh616zeFPqNmb(r>ARc?Fb9!mo_E(}}_eE9&Xf
z?a`S|wWz|ro;6Qs+am0&hv63XYJ-&Ky4k9SY4N-E!?p;_6}|bjd4#Car5|I0YhG*}
ztTJhRz2=?g+?p-kdapI9$#deT!CT#DJa7JBs>(X<N0S4XyImGDicftKM%)elimNZk
z4>sG4XIhpp&2(_<s>?qhISZ;gHBX8L{w(wE4L@8vLC6O;RV1t1o;;K_b0d<+rA8Ds
z=fZP8t2i!_)Yw2^sg87R%8=j9S=7nrDm~&r{gz8dQ`_oDwJG~2;fo?~NBVl;3p^EN
z)Nu|~(v{T~3vT@W3?!<apAMrJmE+(t9tS`PAZ5|PS|Udklpr@^mn&M!=BC+c-4vv@
z%I~h_rt@U0MZTi*@4}caPd|$Lr;7$!OfWCT?ChOJ@cnFX_Kh#&oUNr+cj_>%dX>uG
zr;)q}TnH_O&o@`T=Pd~BfYLAT^Sqx>62{*3mcIUXIsE%q>m5nPe^McA|EY49f&IU-
zV7j%&?Dl_>mjAORVVE?u2M!8uL=Ml;3gv_|Y00oz$|v*vZ5UN4VO5`8=ImgiZ1j_|
z)PpAQgZ1&?)!mZcwnK&(tU?A4d%h0N!^!%BmiwaxCk9zkjJf3x@6YQ5pTURmMtH0Q
z7@2jKR(JJ_9{`{T&4ph<JNhwDeBgVD32jijZV;bRwgFUUAIECv{YF@yQ6qBlSPiHB
zd%qp`Bv<U&N0vzz!LjD16Mx*xMB(xm!M!jx-|!dhbSt=d`(C+cG|3W|_?`8sDAj2x
z#u#F;nWlN;0b(&BAGM6e@OmT*JxfB|{T7Uq)1!X0;Xg;yQhZ{!JiH6@!by9;7n7E8
zUA<jyy?R^n-`UX6gy{gg*#Ra=fzSK<1N%HetY{pdG!%}HH}@C6x5m;8m394KUGT~F
zl17%aU^A&zK@cX{W#ML7=uVL;Po3c=FncHz63X#hef}`J!^`GAV1x9f6&R|V&r-&V
z24QTn=nHp`=e{a>Z9I8#6zjvH(DSUXC=}W@Z>55U)chorEwUB8h613Cq3tm_f{RhI
zjYM3ye>Hww0ogh5Q81-c?w_7I`8ufHv0XTaW|3NtIRsts7`R;543c3Y^m*?pCTUp=
z+9A;R%4&;P!1Oqbqh3++X)f5L)RG<@d^6H%UG@bSgQppbdxaJOr>Xg}sr(fY;)nxC
zuCxm112AX{$!LsGJ?Qs9pw42?z_+M-wTv76QQQN%>d|zVUG-`AiMvvi=?>jI{d)NU
z2?AwJ>5tx1d0^VkIK;%fB(&|RDr;69RwG1W!K>W%wG)&i=jy-a6@2ta!7yNA0N0Y?
zQNYDWvRpJ7H^)*rctM4B!3F~KmSpc@Q9#;&{Jmtz9Xqa%HkLWQ$s?B)=m*hL#F)H`
zN7bJnL!~w2=1*>bA^Ay|%gW^oz$Jnr2$pk__pk|ZvM#o)05rjdU)QQkNr;@Zcd8gg
z6<I~f$B36sTGp{0V@13vZ&^Dspc1~2(#UaBS#wlULcNdFMqqK}pFYvdjs}p(+{b3t
zi|?HrH)Hm85qFm?TA6hPZ_a{q)ehF5_=I_?1``!n7k91Pogfv3E<;9P#bMNz@X7^T
zGnQ>KloBrzyn66um2DUu*CD0;b`?(+Z)bFwWTld5!4^gF_*NeCN=zdZ`t(JPBM+AQ
zCE*xS1&3)5n;Ag_0BU4ZQ0Ag)BNp#(kD<luCTqrdleOZCh#7CXNbPdRda4R1JPW1@
zcZG2mcjElFI@Hsh1NEb91#0v}`D{mMi?fkb_f&Ef$vM?#`JAAcGj%R;!%kEdpO)E5
z;uDQ9L$D)P8!{p4_`c7WV$8=n87^l7O0sf1@%2=r7Ou?Iq%`&vgG48C_h;$48L9es
zu{IdyLbWLRiq3#OH@QdGza>^0czSM<HCTpz(PqN!i*3k&1PLpLC@9=N++|#Rs#%Wq
zqGV0PtNEpT$_bv^`Ij5>KxpM`)b>LVB@#btL7^w@eO8r=p37bEeez_A&&jX0_`-jP
zV2-W8p7%pd&Kk&@l=`5HDS<ouAT}ys?ML{qoUTSa$&U*^3(zgNG=~hWz?>?_#s)Ng
z$6<J=x}_ldW*VU>W+ST8|LYq%lXFf8gYOaFZACa?(hSYdu*tHkr9vtD`_E52E%2OU
zKWuzL8E=hj7?W851cZPgsax;UGf#e>s$|y#;PDja?J=Rakt(0w-MsmW@?FkI{cZCY
zBThouf)s^Zq+st!n+2*Ud)*;fUX%6lqVIBRf1;v|^8VihnKpS-8YAzZ%{iCG3;COR
z02#letC-GdxWeiCXZ$RxPDs|Qn`<pMdYgj)Z(!E%+AGc@LB?3y2~!%EpV$diD|(k8
zTuxNt90^$^bqW!t_*(i+FTqu+46&y~!c&%4+L96anxvB=2@#iJOERXbeq>S&(mJ}Y
z)r9m-o??&S)m*=7c7ZHX+q|V!eE~xm_S^wS5IxBE?csn93UH89L{LX6{r%O(ktPwn
z!Y`8y_`*_(EHlF4WOgi)%+BBgz%h&N3`ddNl!50blGegk^FPBD+Jq?7)m)f^8o8aE
zxuE9Bpwo5}W0<V75|+FIXZjWRlYFZ!A~RWfT{#f%cd$F#YL{&~$XFR=G7Wo7BRMF4
z6)A(Mac!&L=u-PCh>%p2ECl4jGmG3!3n{|-?WXX%>!TdCU9y$4Ins~V7FnV;<C+7B
z6GHMggQl0hu-Cw46vUd{AbmrHrgd6T5R<*3X!*?76d<{^gSh4ZgYSn6AMP$<;&*>e
zczK6Oi`EIkW}-V4v<Rd4squC2N$naw;6q9U-aPD50hm!kkAx+~dx~<-hhe*cZ}EwV
zuK||s;*%Gsi<?zJBrstDYVN3++&l@vsVFE>*6%GaJZBZ#M@YY&?6?M4MZ8E^D-ig9
zV%A5=8&R@Wh+YW0BkzO=mJgM_b%$^LR0dffQbu-O70<r6)M6y?m~sYuPC`P(vH4bI
z`wp|wB2q(;v78pZgw*uB*UM1SchdYf0%@3guJ-|QRwk*HP{7v2Ny(E@9|!6Cxg1Ni
z>Sv&txZwt^H7Xx>q~cF6`jH>#wI|>FVHfv~M=eqC+1c$Y1Mr$<m8~kjHANF#e7t!k
zEU3b*4sYT@ekFiS78bl?x;5CRyDNKs@Sg*GPydrh;P~G@0uuw<f9E4?{_IlXvcvvR
z!xs=VQYjQ+3|JWEbJ*gwquaqa!RqDt{(bvHk}5VUEJ{BzX~HiqRwqhCp|+}5hhGQd
z_waOg@%r*;u!Ew_fnsj;z=h#kys>0}^TPLI9#&&*`N8)30<d{>>0LOhtBa{;pC#Nk
zeCx;nu)!{K%EoR4LCl_TQ}qhVTKdZx8DQQWI34`Qj#{}*%jv*#%rmb&@1K`{$)^Tl
z69ZLG=68aXMJgpz%efLKhc+Q1V#Ra-*pc*m)1YrLUOqy1oCl?r^<gGLx&DTFNcXi~
z^w#i?1HPHtW^Fvg(dRjtG!L86&fZWH_ivz#&fC5Wwi|KhgxAF<d;3Hgtk#+iRR82i
zV=<~TSF{R)lK$NBNg690jxwz<SzY|81S~0juvoQ@otkA>aXWR{11j(5d)$YQ(Td=j
zQV62o&0?<7d&+ws1Jz1>_@WN)ChO|f%6yO0Yc*Hi`<e{1!bKCS3a)HC6}W`vBbyqk
zk0mwQbH3YC&L!**#|C(We_*_&Pa(@U<+P;^^*st<32V&!2qUH$ZL0i~_s41sRU?ua
zUeM~4smE0W$PQd1o@zx&gKkxB8MRJ+mZn+RnM|3Fsnz_J^JyIuKBeTOQe_9ud84iL
zd&}qR^ZMD5b6gz={vRIW`_p-ML~vnOCxbWl|F?lJkxKk`HA{6lL0Ne2HU#5cim)}{
zz<tg8pQr^`E#jEAt<to7>~s3&F+)1e<ZsZ1_h-~OYU`w;G+D(-^p#mzRl12)zNnW!
z3AK#|II@x;K};u)vsq0IoJBO|>vHhed4oUx&{2vhS1q9NjYTkD8t0*W>_}v5rQbv>
z^xWv+9E?lf3Cw%iG0tL|5@8d+A=N?F`v6`q6M6sCYtu%(PB7holz%_Ok4?cvj1v-X
zXog*ZB$wn+I;yQg&_R;)oEy%L|2{yGGyHi0t{5d3!0Wo-EoDK!D9Zrta!uq_LI53k
zVH<D(b)SIpC5{#(mEWjL_vkQltlpZ;au4mVXwwJ;?IW&xt!bw;@Su307d>gL`3EKA
ziaEtHTS*bG6BV~)V}I{pe3nXJy0#d8S0{ye?iFtSSv|v@tX4bg?rw7KQAM^nrPyyx
zg7L4?>lyquxTscjLEtZxY42d{D67X>gvYD0BeuNDHDj7Md-UH*2e1w*kUAK^@{o?+
zR6<TK@dVx0RLDBcR2@}4S_s%{$oGWhdWnfglz5u1N%-rd<?#dg`e*eLK8h<Af;tKm
zyp=z!rb(A1_c^IV93hzB%_=I?Y<9KkUPCGvn|oNXSNC=|&@*MqT_rt)LHqU?7K)I)
zdZJ0vD-wg~z7M3k7CY6|yy-l7;?!D-wHW<CnKlXX-Ph1{Q@+84HJo$($T7joREYaY
zzQU(av26?lgc#+w|H{dC)LKDfK)7*jEHd>{FRKfAL2+>xLAZtzV4A?=YuA?Suh4vy
z!Y)Y4GH0mAj^3;WL-M*96Jbc$nAvOKT*|aMXiQAXCOjq6sf*;o`9_XN>QZIjoLmyC
zNnGq*9qkZx@`w!XZf>q0t@edj3(j74(P*I&!7fm3;j{fp&=ftdTdvSJxQ}^TN0%6i
z_FQx{J-eBI`WE-o^YKHu&QMBnCR2Y=kgI3GcI{=gL>LtU2h`Z<!6=CL@1Vaw{N~c!
zXh#}`rJcN@5=j2Vm<`?Lirm4HhFUZ`XQohrRX$vFSTlplGUArxt$$v9zyYiydIoIX
zU3Yt#%r?4d6k@YWWWJpGr*jTv3_3_P7Lc(~uP;<KD*9eOu#KN^6HQB5Oa-av)Jswr
z{d}w_4TGt`eB?snfIAQIT5`9N8!&&;ER4;PjV?~nl*6kt>efSim|i2eX^4hjGN{R<
zyKkxK?g#F8RjLl-9MA+@h(^K-7N3eiQP~ATG2XOk9p?)YBJBO>$b-X1;o1w7z#Y%s
z0&UU@{Ju4XQ&k9Rde~EAqZj+;6HO%)ud`-Jn}vy7E6`zIY5;ut>I!&4hCHJVpwaxZ
zr~OAV&j`GK%!tXX?L?QsLi!bKZWAdO9rZdw5N6z*!GiHUo52D`C0pxzI;MBWix`}?
zj2AZ;{o^V?P&%sBcs6m*5?0IV#tZf>w!`rMKZG_Jz-f6!wznZzD-5>YdE<Ay$Zup7
zeARDc&m6g5$S8O)eL$l0clCq158Z$NAI8q9$+kdQvt`@1ZQHhO+qPY`%eJx0wr$(C
zcd6@kbVNVg^Kj<Fiunsx=FH4*48_+uj$E{(-k9z{ZH^b>zwF4_Qe3<c-oOH)iTnL)
z*cjqOeZha(1-}H>vi6+BuVfYYg1S@A%CE22vIty|5-JS!s@J}29lR;$RESMj-I=XL
zua0K)%)piDv9-V5mrX1BC>Ig!>1mz+Hhu~{fay$=TwG*t22@3&p0azRzNB)aBY(ko
z9}@JuU1Aucu==Y{>s<c?9YX$XdT-Ehr2m`DqO_}J&wr~^p1V6CYo0qfAG_HLh)!;+
zn3)8VMZUE9z;4huLff#aX~WhBx<+HFMnFk&EpF79=;b@^5MmbW%fcd^m|zC|jUF1_
z$YARjK`UW95;-qv$VS|(YUy0u5~mZmq?BC0;&QqVql8^9A5+h4C2SSVLae0iCktio
zzs;8!z{UTU?PG^OX!^uI9_i-?myP=6nZQQPL37kBOKk@{@44v*FL#<T>hpp3PLoP`
zZb(*P<b6bcr71|#TiB><Y)A7!f*iHG@Wfs6_5$hMjf6sh*FoIETka;V?IrsPQ^A|>
znMlO@*&>QN=d%~ktizlAKeQH%9Gw4u3-VuqH<|xC^!l=v?f-x`zi;#f^Yy*J(<eRm
zHw!2d*wXai=@5n}=Q{f+P_!+#ooR$qqqn~Pd50}%C6$m^Pe$SrB$~H39yxG4PQi?e
zgz7p(S~;K~MP%bvU@2y!Fco%402P9yrNk<zJFyUTaQUa`#MY2@YCY;z*9cdvs0b94
zX~YuQeS%P}gRA05{&lR3#c5FVs8WfEic4x~0f9mq0z#XZ^0qKmfE;iFP!(pV<bbm!
zTc)ryRvFEX>V}O&1rAgi&8SjDRI#!hcZ95=4URI5FTn~8oD++yTC*6E6KbL-C{{r7
z))qW|QOlu7R76Bc#*{58R6yED`ZGobZK}X|6-Y!LX#+zc-msxPvWcYa52*4dpi<e2
z$=PATR)?%ism44`?$0#3Gib{2@JJe^5saCFtQlfiN;P^Y7jP>*RQVQ3pH&u7Uz!d|
zJsV)2*dM9CF%{?Vz{3-<^ubZ6#t<4oHKdyZ;KKsLG88D-gB?UI_C>B6#jSFrteSJB
z*PwjT6-{h1y$v;$VrjxEF$LqMR+%B1CXCdE6g1XNj?Eo7(wCScnd5Kyzoa<ou+S4N
zo0xPR$gK^ga*$BdYJ<SvOEec%ij6jg`2kxN8r#GcD9X0Ng`FV!W>)ztDOjdFG>8u%
zGS!M{*i!Xr5n~#$ifLf_i^9{rN|xv`5Gw6iji}0LB7+^(EL;OER<p%CPi?Elsr&}Y
ztVT|~#siygF~Td~N7>hhUi8D8_p?^V*RLZVHoiRF%tuI_Z_!sHAI|nnym<Q4Zwp!e
zSa~_|UD=&(C|DGXivwaG{^2x_FpQDsqfeg&uF8>@gH8KaJ=QG%-qBc4MRXBM8}ZeL
zrd!?DZ4)o1+?Ij?#n<?o(Hn1`zUO#?jr%K9#t#d%?8;S$^oXk}OslvaG2V?!xH`Sr
z>cQZ}#~&B@+D`u59J!g!!JAWb{*9kH|Ga5Bx}O!--kce~Ua#wTa{EsH+8w>CyqKS>
zZ_npN=bur{*MBw-9oY9Au|QiNB)#)2R3%ot`+p#)`4qj7DWl`Wowju4;yVw=-;Zs$
zxI*RT`_;iK7Wn)Q^z-3AwX?1=i#N^X9>be628Vq?J^FWi14^oX69eVm2Xh~WuY*U&
z#oiY^V06h)uc*WHJTze-gmx1QI{SAt`hZ8>1N(w_R2~+gd3m}lqE!>$fu$cRETLzD
z`qY^czGT~TEJ5nAC8;48=*API^UbAs1CUw_zVlxXb^h8c&!BBdfODtn=H}wJ`xARW
zTxe@7Li+$SGI;a$0OTM~ko~p*Wew<<1CVR4AYGV!<ADC^oVW~UHcY%-7`XAe!35wI
z8Pg++&ZCc`y}emq4jkCiBXK9NvjJKdhnNF&`R)4>uTx6V7-;aRk{WH6@b#n^j&V}k
zkLO2k4m{q=y#D{Hy*d5ZxwRP6vhp=tfl;V;AyFQYP^hmXe2zdD+r@f7ww=6a;wTDk
z@FV6H5U0^9bvA(x*3lmI4K=U~?lnyz%F4y{-YKW7hCO!djayl$N%GmJCmzN&vr>+m
zr5i7VZFEC4;5tJ!U=IRSyK^W(<s<O|bH2Fz+!<|0!iIPG%d1A*%H~p|Td%ITnoyq=
zAB^uQj{beT&2N)s3~6V~&mx{HmxAXVoK0sP8x`^hL7cXQ;iMA!^LW)dPYViCc?uyi
z1Td4=33rH`+!8~D0A@43#n_xPgLP~MA=m~XY(s`vKnkhd-2ic-jDt(Jqp3+T3hzIx
zHT;|SyOe3CL3JN)yhSw_uH03Vt*zW;n@uvjSvm+Hho&FwM|qC%D8^;Rjp5FqsaI;;
z@`njNZD~~YM%$xRhITljt$`*3iK&$WUXSUEGcz|%FE94AMFs-+<IS`qeu0uGiD)%;
zXs3mcBRW1{1tJ!~9<3szScqb|7%@~bB2aoqsboh93hV*eCO=Im$!oCXRSMCGa3v`U
z(~JnBQMp0}^`3-Q5e9Hz_32DuN(YI0wxaeBV=^IPGRx9KXxL)j0D#;aK>i)2uQ80(
znxZLnqUxG#_6r-{3f>1@B1u*viCZv<u65ex&=pXCmxPd87lD^;oXw`Arbr>Qzw(;W
zd$UHQ{WkboKpo|YByU_7HNQZqwj=_p>-wQbK5EkOF|GGs(%?<es~k0^vwDE5R1Z{c
zw84|h=6rZj2th01f^{;2$r(emcsNoLw1SZ5Dp7RdtS2bSdY>kqbYiGReSBUQT<Alh
zZM16GrIiq(Z9oPvM?29w_P{#PnAv`P=pU(j$_w;HR3!GQ1dlQ*l3EqQ3FrisbI9w+
zHDumGtD8tX*JknUwnNO&v45DMOQQ~A^|?O<a3q_05#0Y;!JhX=+!STtdj=gCL6fF=
zbrMfUJn${)Vn#Osu}?1``35LILSlmUeKHQswqyV$@b~H}lMEdt7j6=zqip9|gVRwO
zltr&$u+2<5u`a_9LY<_Vm$XngvQh6(@sw;^m=VF_fiD<P0HW=~g&Csq^Tfd*59SlM
zG49Hf5XTkG2%UpQHc$Fb#x*2LiU|r3)JRf45QI?2p&ZK4iG<3NcBlu*6+tw8GhiER
zA0f>m-;!uc7}cxYRfilP{@N;cmedOXs2Bn!eDwkReS4?^E>8OZ$Ff7=cAR@L+H&F<
zF`N|KP?LM=^h{`pU2LIgMN!z~U>@+H)SsG>mBKdUI}c!phjHYcGS>}zUs@H4^eTk^
zlyr;UnkO*9Y_ru4#{WDsc4FlA_5Rm;|D-<DbpHga=H)j|5aRoJv<=jTlC0~h8HX~|
zCKW&8Yg`Gj+rB0K7WPQs0*sGXb0wy&rav}Q4@x;@c60>g6DHYDEmSMPd&s+#!<N2Q
zpp2SNMr8gnuufOzN!64gBq8J~6@((Z|JGidscw(CR#FU~Mnw&SzwWy;((e1}32lx-
zpp3U*E^-r4YynV4(5r+ygjvut75~0fVsnFm+5?S}+7vRdrOx~nOMPRn<LG}FB@Lfu
zItmTJFrThQM?lamaPl)J;poZ4FNy;FsCffQzE*^mtktdCiA=Hq8cN~>fnRRd4vs<+
zE+JXIp~^S@jOQq%^jz-got>cOwFN}}4NqdOo&Fy>F(yvd|KEwpnA(}USP(F>urd5s
zo1o=iV8`bFI~Jv3hKIxN8|d5Uq?^+Y?ndQBm6hXU`T3bE5RpP7t)tbJ$_%EZC7n4u
zoO0l+EqLD$c-*|X>-lgo+VnsM;bpzm$65l<--X~hE545yVPw1A&E(Gm96LYF$A|8#
z#;<!7tZVQ$g82s)<@ZE+@I9_A<NSnMYDd>;2Y0FMvP6CHOd5On2?3V`k`m4p=J>Pv
zamzh65)?_dT`K*xy7_Ult833ykBE*9-mma?dUx0DJW!NzM*Mlq!}aWge>!a&zj;Xy
z{Wn1?o&WybTp`A1en&*%v6npt!Q&)G!Xg)Bh??EsSO-uzv;N0#$+L<9QB97k4l4m$
zrbBDE2vPO(9K}JcienJR`q%$XQol&XWYLmD3rP&^WjT)}j*KsWu1_9x7NnJD=B~8Y
z<nNS_nx$;f<;oN*>yc8ok~KeeDu`dte%8Jc-ZEewC=xxAlqcciD^M7g5Tq}itO+tM
zhP*C|z(YGKMyf$=ZFD%+hOs&{!igMeKP=XmV30`&1#jVF6q$Ev>A?!Vj(Is4Khs~o
zk{rC^s!6Q%K(hUj9pBPejQjd%0a)dvFWsFR?Sf^}D(?&{aoj1oM;b3(Z|Z3y#^eUX
zEJqM)x-P*~d`gT@Ln2xRx+zD%>rzVIhzwDa<)@*7sqJ9dpFT{1<2+X?1UC98kc{Ik
zOpuxge`QVxoJtnx(X}mRE(@R;C$Lkb+X}v@2~g@eKALRP9iQ1fsD629abz?9aYV!z
zRIf2)U-IeB3y|GW75as2g}^_{lS+3fJ@N|J*oOgztN<uh_2FkfSzTCx%hW>LeGrgK
zm@w^54fv(V%<O?15DeGYwzok7II%O!i%B)#6ga7y@MaP4L7HTw{X(u<QVFcFWwBnq
zS!gHk&Z1?D>{^g~f`Lq1ffr6I<OIlnf|QCDWJg<P$*5)SO{?a6w-8}>t>;r0I+Lo?
zV4|smGcOw4A(!Ifle2DhTbFRw?HJLF7Uh}2SoBA@MfkjQLb&3iwKg?dI~_92CN^F;
zG<x#B0c%_bS_`|jOI%%x;&tfH-nhT>l((s;9~zRnAqw+6x%<;tv227)z{ny<o}n1S
z%z8<W(oIrR`EyE54N8gqm~Tih@6PJN3u6&03CyvjSw*<%qOZ!i4l6O^7~idQ)?<<^
z4PsszUAs4>TcQkx5PP=|$LLAi(yHv{KdHNmj9E+J@?;0ef2QMl7`qzjUB?U(FyC!F
zCeomog8`588^_~k%O49&;WlHeKQ44-u<PEUB@mUu<u)n-WJ>-j3$3coA3MXwX$>V*
zSq;D-=MKYVZ9Fi#Jr3)I6Z6uWSti_zS-A9Pc_;I9jQ<qZkxwE<he-9pl!5L^ZBKaG
z3+NT3@Goisbfu%}aC&vj4cJ}aR<wY3>O=;lYE_l&IS@Fv5?Lt`tv(htF?}>nsxW<c
zT8*`q@L-i7q8ei`WhHYygu<G&W1${qp|Rj*5Edr{e`RV67F4nxlm^u^xn0xoi@GUh
zL~;%~0`8NP+O&q2Z+88q8Uo`GcH^aEo<+-k#B9S_lyBAZ>tV7LUn5Fe5wfZa>_*Bd
zsi^BJr0+uF6A5zOc;w2)oY;;P^-0qeOB&n2GL=B5T!~sIXYq%MD2tLfbbE@+F^KU8
zKDQ&`y6krpP->SjwgSae<JgQ9Qj@_dl}Rl~X_bj@5i+W`-&{6PE`lTvR5no@7TPks
zTzNzhTHa^mY@?K=U9r^cqbW<+nTLlt(~c#S%OBJ+m6`qFvY4n~HOfb;CMpzbsfBsy
zEEGE;f2qM`NnM4-R4qFO$}3ydO;l2jxG_{|_ka4z6C@xZ=?zpc%cA);qp;lWsRrBk
z$r;gQ>3d_L!VK0_EA`;-1&<#~gv_x^D2$Kk$<<p}qP!)4_PVB2OMh}^)obv703Ei=
z^8P~wVE$eE|F4Tg7EaFpsseQXRsfuK*q+zAa(;#ekwh<mU@@}|W-zBl#$Vf6UqnuD
z_u4kj!>e$&YGvikMeYk0#cQ%u;h4NnWKwpIywg7y4?k`nk7W+ceF@F<-@rf=zr?q`
zq#X;3W=Y7rmDdO7*P)^m!za&Uu#(ku{BbYt&%(VR^(ZgO{h)pd7oUcTA<XN1!id$M
z5sTld_^tZ+fwJM*l_mxTB_q!pdiplo+KX>urUqp{khgE{<L$})=~!MJjArs_LIOcM
z{q%fF=p!IHpuw+6UH!C}_mkU$%NMUQsWWH~GC>I@g(b`a3>8D*^oLQ)!Wvp(3<wE2
zh=!!+VNi%TSRmIQ6R+?SR(W?V{A2Qz!lfjRCc0!C0Yar#DSU}rAyPhH29hq2r|J9k
z^wrYh{h>mUU610I2=L6v1N;V4j>0qtAf;kXEAu>9l~x78cpz5ch;|1_XoM=D2HgOe
zND-Ym-4n_SaxB8Elv;O*QHZqFF5oet49>O(c2-PK2VO`$M*BAxb`E_B&`ifJx-<h@
zv!+SOgqp~?)5f_{D#ct^U=mEgjPb%W$M_hOP$f=3eX)$&;rGB_xp$~td43hPNUS=#
zK$?8)eB&~?UFVTFYlpRyqJeqXls@V$U3Sy#G2i&9$_DAm&{k*rz`DpRgl}EBfshOQ
z_YN|fEN4!u$*+Q}`Dlu(Kq59<a7B6|3~Lo4Yd?4Ous?m35s(RJ%jhKkIdNf>#5#kq
z?iff>D6n>2-GOR$eS>o(U?s5B1|z)A5l*`|a|KKH?edD+WUD<zu7uU;>|&F0TChXY
zDM$I(Q&*nVXtVtImQlk5?x3r~`9n1`0s>dibUO0VfE)f5#Y;$II9`&7Y;e9a&RyIz
z$YeIQajfJV*U1ZBQTHRiXrB3cSZ4az&~DX{S<_h$+fK&RH3t#oT(gN66*$t&_(Z`p
z7!vw=auXn&$gA4xc|S$=EBhFzAwGLu{8^j2O4nzdv^QUpBN`drT>CELUt@8zLmN))
zP=|f^{Ap6CEZ<rh_Y*NGoGm;U;Riv;S?o;s{T8(mfa{b^j`d~5xn?thj917Ut^R=z
zF7TZg0Ube`oX?C}Xp<RX4;Obc82Q`<ci*gLgMuBPKZ{KY{uqIwc~faPcL%R}|8xO~
zpQ$-ue99lD((>x@m0-yn>{yI^`B1Ntn03ST8|zcH1tEAz0VNd|Ig2m@R?^G2lwjpx
zO%9vx=oI&b)6@)c4zK6792gU15jf$EOKe#OJ}=t!Ef(Hjo+L4-`KQ7|S*^Y^<XTKv
zHsu&<aW6ZLR|%vwZRktGqx%jMH+`i&&I}c39$^e8ty+*b5__9|O=z?sH15AiHaTET
z2U!cVb3qInLdf{x`5zs49*Ozx{*iw6IhQ=si)BnRM|)NcU#8<$9_{5RSLzC#wBE{<
z(yY0es{`3VDSLRMYmHhkIApO}GZBMluwOzeg5jNSCAGZ75@nBFo(30Q#eU=X1mvkZ
zI~VX~MXN2G=wJQWLU-U0tc1&1e(c_TIeHy$(mB29yxo6T#?|vA1K7Tv)qtXcX^*dJ
zHD|j&F{KcGwHu6Q(W&jD;e;C+;aKHq<tt^kSHwz)*FkhnrGh4Q%fstu&f2%q=lcTZ
zRyjVL7aiNYq;yhS@w>L058tX-W`40@mcrTd6^QlTmgz=z27<ofVX>Q7$rJ!paTnRl
zB{G#}-gTU4f>mzYMNbAE(&o!!gDDu@U9=$!pO>_uLU;Jc?w>5fJ5x<Baiu@91)896
zv5cwytW?Lr0O9bx{y#6Rbw5f$;GgU<3CPUiQSN^Hi&ch_0FkBm=i5#fcKz~F9%1om
z&;x2gy=jik`<E_zo*W>6F57+~YI1ei#kKY{dfaTaQoYS8a76tJTBRTJ>}aJMFp7l}
zq8`vH4{LW5hm+-ZSh?TbF2{;L8F0I$gQroYtDp2sk+j^ss0^M;ZX}i9O>CZcv24!q
z{hS}MBYIO?U^F#?VFH17I~#v7GVD^Cnp_rA$;ebGl1m@BTkIwHAu{hIs3e0t^iM-n
z74cQWSo$}<zyC4Af6N5)x%Ng*1uy7Pls@I-h6wTei84pQ7LDHGcBdQpa#>m;C7y?5
zGf&?V&UBw{J7LKPoK+sxPkt<b{O3u})+Hp;2Asvn;g;@@PGXU7v#iwsa3nmO?}6P~
zhT!G{5PdqKcW+H~Sb{4;6&Oly%AQ1NlR~t(YB-vsHE5{C!{OY#JL5z9m{L@+9#a}R
z+hBiw#A&))*ZCDkxfR=)#bZbDYLa`}<j4nIRC?jpebd0?-Z-Y3n%GEVV!t!m16pf&
z8j<v%crx5jp|~A~4RfS3phuIIT8|1l)h31&opV^B>hQpEb{aE$qy$07bJ4;-)%{{6
zm~>c(_Vl^rMAM?{kh|>08XNnWDU5>2UYoe1LEKiWCAPm7cKA{bXr?RQ;C%LyTtQ^>
zsuI`nrMThlu0QSebw2U=ta71^y<A<!m#2Ld-!Y-dE1^;6S=0V_!ma-4={z%@_ro^r
zeH?OGFxfT}T9T9}No>OHPEE%V#KtYE$!*!`0}oIgE3P%j3t-BT!+?GR7?BQNiEwYq
z=bhc`v+l|ckc|e>YS!dS+UKapqwr@Ilg#20w_Kvdg;@1P<s5dRo%`tU=oLrLDQ@On
zt!6w$L$%9i#dPsg?tz${uFy$nGYide76=e3*uAZbG4&(<;{Ie4vybUtVl(&dG{ll3
zomu<nHq6Wyf@Sem(?bqALlCM*mD}Fp>X$?P3j0o9TbQaW!q;Y<?u3JdNE3j_rjE`$
z^>N5^$s$|mWgt@bR1Wy%@CQzWQ#H{wS?J39i!i>%hf+tisKfRR&=Tw6vr+>m4iz{7
z<N9V6R>2VUfa^s!#Uwk{$w44tIwDv(>h+^}$k;}V6;95wv$NC@B`(bg&$QF!wX<aI
zANKD_PJD!^?|(gRVJVoYPzh9W4!O`Pz=p-x;;YE{O%w>1(KTKkR6a`j!$Er87S}P%
zv?p;=rb@HT)mr1rT3HP^K7F)qG^+g_7aJ+_N`Utxo3Zms++xT-Ny*}fo#rLC$qz6~
zZe=`>UG5>HKK%~KW_EPw!iI1SMJSh)ntPute?-Yjw9zi=;4^n1Ce>Ich{GvR3J5a<
zqEO6fPYV-B01JfJ@(Q*viCg)h1UZrwp_k;E_jEXjG^3_<rTwY@o{@CQ$Tu$$%l7xm
z&($&ngZkElyt>fm<SBcrXYRUOY@;F8)(r%8TGIw8Q1l|yGyPDj_!+#Rmw)<%y_SCl
zOnN8L$Pa9{iDT3CI~Tw6C;In1{ilYI>3>I57&%$~yN2-p=10Z;C!%r&Z^A`#3P}S)
zvhESMgY|zSDz8o9H$VKzj|;&Kw3f8w{e|Geam9q3y9qcCk7qa9Wu5fy@pbOKxZ^H*
zpwM`ot97_{z|;PMkFw_6@cB4`um8F~9{;SIy?Gr?TF4Hio!$oWU+?D<++iEMACv)v
zd)pJl?z=-wYB-s~+-tuL<@M#kpYz4XZ=90^9s}+Mj}GeB_3oZqRjbXH)SU}|zkYhX
z-u{jSk{amTW)c_pT>SlerQu;Tj(x8BDkN$@?&V*fZ{R6iVPpWn#JJ6`K`>k<Hu@Hd
z3ZKdWj6H->VcB^oosXi)Ej<?49>@m`ZGFuqI^S{q^!Za?7ha#@n!`gfhCf+n^)Xkr
ze%|kEklpoqo|MINjgw3JPhB8sF!5_#k^)-BvbPK~dU1>RKmiI3YnW*<paD|Vm+q*v
z^~ti$fsn6H^F^_f8P%j#MhCkWA%9Jt<MJI2++2YwPk({({6s`d&8%gie=$eQk033X
zQfR0j!(hss4QurKM-2w;d+mVZBUV{7{wK5H^EPjtew8b^P%(|?g~K?BL2G5vH0*@?
zjZ+<WkMm+x0)A{-dk_p;lQ*bQ!=!GNVT&*K7A49{a;!(~tr5JXC@0o5S#wBQ3ntnw
z;n{0RrR!6iPiukR_<>QK_6<HMrg+qJe$|A#xGt%8;I9>CNlb%eq{q=5e>h`GsrpsB
z$O<&Rp4WEhto2=amk3r=%|$ml@UZot!|!!BXIokYTQ%XVhqa6zdZGwEtmjbHhQ~Vg
zoS+so4i1j~q|u-9me&Tw>LI@D<ROf?E&p00kPls!TL#c5g<XnlhrNn;e1>#Qt=41N
zQYw;gUz02L45c-Nm^|lZkL(GKY3YJj;00991YjhNl_nNB^;^@Qh^>`5S>u|#cKKH7
zRm_hb{+X|~k#%cBO=KDpH-Q`T#ZFNP>LCHF`a@_|wP%0UsPD|nyhMcxvXii0iXo)W
zpCCTG{DCR>XN6~WN$k8(GXp4yg@TnZ(h{RXmi=fq`M_e5-NX+f)3Z?WHKTN2u)YoL
z88`O8&0JmU@~408f{dt1&e$-*8b!`?ARVXtC184=45AFw^UP68=irI#C8bZ4y&(W9
zC{M7+`te@n0vxNxdts1^uBNpaCyqt?HWSfoKQYMzCygH9PvO3ePrV1Hj()90U;Y87
zbWVLhJnW}~Cd8;-Op{_GGt^K5tF$I`aO$kMBM%kHKO-{FCJMz$>F1KMe`omP`wX}&
zvXhmHgv;*OZiUy`+UnFZ$zf{X$6_Uhi;pan)pWP6wqWwiG}|_*ZZX>yYTl%TBt;R_
z1dQ~>r2HBwYIDxOhvo~~g6k`Q&KuYB%*Hxeg9;P>qV|P<$L+n^lYE1yQdW4P>!*aL
z;nAl0H31L!+R)+SIcZv0xwWxcmKH&iyoS)vv(0Qga0-VVg%H(;pfO5DJA4jj98)hH
z;6NrD5)xv*4QGL4s6FftwwddSR9auj-fKpSOvTY2r95m;P}FxvUMr0r`EW^vO8arQ
zUDbU)m9EBfFk_g<4s78eNxY$Y`nO6%H@U5?07s|}3Q)cVAbi*)SPEcYG6$#BBu)}c
zZ%T!{ve)V9*ldw|Zh7{!BXMgSM>CW+k+Fv_OAA^y990Fm;t_cJ&Z2edmK=n>?Gd^u
zn_9%AtIf^4FavjG4zP#ZhdL_{A*7%)u<4e##v`>OD0rg%ipH4MvFjX_U5u+hq^pJ0
z6K)?}TTsuew%H>8GJ^c8hL!-ztWSKg6G-T&+SFKe8Vg^(qE2;G{OIOJA1qnW^~=wf
zjI;3dZ9z)V=(1S{a@hv-7!ICFd-*#>=c%TzcVX9Q4@3YF?T73V#Z7hIBA!wiuEjo6
z=NgFnNlAJ!2b*KErlyQU$Ov+)XyI2AFhh?r=JdE7^h9*pEE`iFiB8oW^x|;AK4o6O
zkGt9)1(Qc!dWG5?jAx>$hsStBL9E$($NBkBd%lN2QY9GR_!k;ebcBOY<u!-1+y2IZ
zT>7Ri!e+D~LxNe#@1coHEDyq)VW-pCJ;U6RD6Ie;55*Emrxmd^=$6+sV6b0|_y8H1
z-Pa9!l-!3tZ-m|xO;^F(q#1V&b-9^!g*?5QbtSx7%il_Pl!!4}tR>zdA-R)ug&ctn
zNOygSg`lJBQ^8yhVk7DaU8aPjCAc{cNS|#6-X$T`k#`jjO@{nBtWP29aO*2VR^!%G
z2CPZhQtpxv3rFkCt=nMhS8sHzJ~>*w(`){fr)t~WO<=MR?$LIXW(L;g&&)l{1f1}=
zx<wXGJZaLXa~WseDtp?K3qetytQ4+`hkGMGNS0$K%10#Pn(ZMy69Ephb;GZNT)7)Q
z9nQgykvR+rdJ)$`@~nEC@e{3*$aEsEk~Ce7u|kg9kFe4nlC_gX_8(dhK0mpba990H
zJ*^1%fV#dTJFO_bg7?jp!uVJ@tqA#Ny?xa2s4|f5&L>a0ZISn=B6N}QsM4nhGk7bS
z1+Nl$Pc5$^S;OyMY*lpQQz{kn{-dxA8U@Khd@@Zh&KT)=EyD<T?yW3S<5iKT^YjH?
z2Kn)y?l72{|M%>Nk&~76znuMW+aIwZ+<Z|d(}551z;O<O|JCSwxdxL+3+^OPP3XLc
z+WMhq@{vwfqxM{-Yf7SY;&^o2X~5}$cY6E!pFHZL=Al6-@n0*CFE)VQgiq)sa9CJ0
z0Swr|%U`2saPq|6^VI)<BJ@-9aCgt&%KPtOey?k2`|jt7LRfjktD4lZn#3Myo(J6b
zK9n&|pY^aXDHwS77(F<@zy6H|Pfbbf%X+}He%CEt4*J8VxR!uo@+Y(kQDc3--y`S|
zBrvSN%R&0RKkn{t4(9RBccdJlo`)aZJ$}r^`b@8Rj;~1?N06;i5=U$q#GrF^Nv-Xp
zQ%0H&GV9ZDGf6>_G5PUvYZOa}SzLQZ!4#5+p<yb<BqOs12F^Ypujj=ue55o4d(7aP
zl9Kx@uj@hC*5}oj``iIoN19OXCwq3lovZT9LoK<1=O)Q#j|bxs*r2%Q#Q1Cnyt}9m
z4<yNVA~xqFGLhc^IZ?!9jw2`(l0MChiW3JvpBY$|ZJ$;)p&GdM;v0m|!bL#<rQ|4n
zIsuEeNPzX0Nkl_XJq=hh3p@1AP&)x7{{DAC$_?)#V;hH%ccWrM3yrsjj@4f)vbCCd
zo0<&2BtH)fTrCB!EVAvEbRlsV(}E=CU-gFs0g42*WFQMVQQxn#DHolhh)9?>GLIUN
z2CCe1G{j@W7=J`A$cv5a{Pn=H1<_hnNL%=b@N1EBrNc1En%*H88a@Nfec&)%F&5b(
zW6M#JiFwSQj(3*X87!k7+5vz(hY%+5mMUOtOFq+bag?#FgbE6ol-ruBl$HkMh8lWI
z_h9UV!f~2vwo!@#3B(5vD?#?jiZGNzdK4C+Q8ypXn}q;|w5O$(LNBTAhXm>Uexal}
zCT;5)dO%92dp};(yY9y5Z$g1v`jzMqR!Rc1rqBj*fo3GPjXQUQi?fx8D}1)O;x@ZN
zx_Cm#SGL7e!}s&l*4bC>*v)|gRe&OIh4#y!CD^BdX3`+eTg%MFj7iaL1vn~k<E@xv
zL$=46XCMnm+h^WPJ^mz#WJPY48o3S|Ow$~$=apHLLzsJVrDNQvBR{8y=qr3*O+YGg
zsLJzknWW+Nagys^Xi25Pkxy04aC;M>;=7v@&=lQj2Gc`h0w4YfW>sgR^v8{_7PSlQ
zy0l5#!>(x4T`ZJOHZc~NeODH5_uQ~H0HgI1-YFG?mbL=6a}8Azst>+a8=6=yo@g3L
z#0iTkp}p7~wrVeLa2ZgT#%`k$n_8EI38bTxEB}LOU7K#}k@YYWohp(0c!PNkix4kL
z%p5hqA*ZMGooxNYfRYyLvltYnGa%TQ)jaVuHF7b6?g>o&mTh@?!mp8+qrAWL?$put
zvVIn|df*jXL^IU~g=CtByWNsiO41Lvk~B?fS7!n8A;Y$SiPUy1G4WMFWX23%8z&Q&
zCp=*&30g|2wW9cxT^1u`xF?{hTihVMq?psrx+ueXvtyRgr9We?j0^owGxlh_nYNEf
zeg#rHIEL<^M}!c{`ZA81d2vN#MaLN4HV66hRXS&}uR9HkDptsPJF>aa*+H^VNVL>e
zzRQew-XmTLmuxi3ScNEw%HC@Rf{(N~ig+7)sF|EF=O&nAwM1<&++{N%24H~OZIdg?
z{i?+^xz1&?V;z6VR#?sGd4IdbZWX)J7Pr~`w#D^hyI^xgxvTs4N9VTX8KS3q(@osQ
zsr5Q;?Z|FZ?Dwsv6@hcUSTRo1dEzUF!sMrNrwF;3lpM-8l~?yq;uCdjUh9*gnR#a~
zXh41lUUMm$HeNl)YPj_B=e3Lk!iLUKV5(<6Djr`YVm>V7otZc>0eE`!FYmRX7AIEu
zZ~-b3y*UO<zFjJO#<MC_Ox@MG;#M>R3Oq`FST#j~a|rYp_5|92T75JrZ5E?5VY6Zm
z>io@%YZXW1o%noE3r|P+Y8x}#Dbu+0mYJhApLZ_>rnb}GVv=I?UbJU$zpN!@92fhK
zWQyrpY5WFiE4kZ5l0cDHy_yy>;Eh718}&-ZL3HFU?KPCBPA##Gt$O~0XwJS}Gm^s@
zD6_HX$q%N-_^xd2S@?~Mu>2FzxyUN2q;hMD&OS2Q3s0(Xq0R?bs8B<=B2UY2waZD?
z*)B@;bUTYH?j33N1KVmUVhC5;S^l!k$lyzKxjuLHwN{eE@Lk|syox%TNBmexiHMMI
zycy_}|6I~5@5(bmAw8B^<7%8#eq$qFq1lYXd`st)&b{Wa74AY=YD7c}E^+kKWv<`i
z#5`qR7To2Rpzh%gd-fB=>O#eP#+~a$i|O}@^L;rn@UeV_E~RbfX>Qs43KpUxpw4c@
zM{xq=Y-eD{J66)kG2zs;@bwd>#U0A4??jVwuGy;5fGzQ^%+wa538@_X@YM$?{nBea
zSl0$`VR=gJ^2$~1yrc7V)wm*^<^3wJha!$QBe$>T(^aGnaJ5fxd>ngUS~{mcajY7_
z<;wxU78bWTm_5}Xr%|eg-RT~Bb?maW*8bxfxIl&W$L%W0J@FLMoWahN-Xdz+wL3RQ
z&l=!#E#)M}DWyF`88-8lZG}glxF9}na+iKa^TonTKK=6cD((l|^rhnDKYiP?{-;a_
z!+*c--};q>-TwMHG(?MkWnserosAg)-K=Zc^UCu_$-@7`pU7=7)i#!QYdPybycb5v
z!<QU)dgB}0pB$Xr94w1o2u6+TJGgtX1@I$yM<>~3VbK6EM1isA5981MMQ@Hm<7-Kk
zy#4*$z0Z(~hr;Ov90KPD<>1wVHH7IcVu)2Gi&gwY%YBFO@q>H)e|sKy4>*M~{#`%s
zgbYo}E=d2uitc}(FL$%Trno5thv3G^2`jMvUmg(fEdNvS4MNn--riq-QJ}A>7b=W!
z8jK4Vg&9V35JaO(-3t4PN-<{rEYUUe@A*EsH~uHVr~-aI;TjUt;XdQH3SQ!y%5$0?
zu{K<jljrke9%<kR@B}M7iVJabQ|e#u@4gl{K3__x!I`XmO03m`)*M*6jo-xyfnZ97
zT%Cg|{JLE_$_Pnldz+Ff205`c>t`q~fQ^u?O|hvVY->tIv*Wfd1z0o5<>0npuS<Nm
z9LgIiIlNF(=GdT8Zjzz;i4WPbe@}$60#&<|6BDY-3h}utb!BW8w`>kDl&Y)5P0uMi
zMeZ=rZ_>*W^UBpVKdE_|_bVhf=<v`-<IyNs2C#c{L|23Fp_ZcQxH97iDjQikh%z}G
zmdr~|<PgMT<YF|6+0N4+(Tn5U4ScXpB)Znd3`=%m`~RWsmN<8N<(_TDIiH_WT`HYM
zbJTXdRc7sBLT5c{V(Bst%D#C+0$7nRRk2^tjFZly_!9LLEGV9-@Kb!2vCnFzji;3^
z6rYCMAO-S{r4qYs_L2t2X7(G0wo&$&8e#Vf=>dfM3ez|kSN5bLouhCM_sl;YV_ce)
z8AEa>*LtdWO`dWR@Mef>3S~o@BPjYU;w6?j*I!LVmFlh2#AXtt>!u)ltmF}f&~tb^
z0OB{6lo;$M8wV(xWJgvA7)e1@$@loRU*I0$kV7ip^EeBFhRPl}<`NoptkbBB<uho_
z;@psqj*x(hLwC^7r)rGr#SIfb_-Fi(cST~pEx3OV#H;`{SE(*qBgNLBR*?nhn=&vq
z?l}(Zc?lFW40vxdy!lkjhkT1%RNT%k)5D>ekmMNDSwl04><hj^H@tKe2@uZ(ed6K}
zo$tgA8o{y;24G7uBVgH1K1o@!#og7%*@_v=pe0|6{UY5~e=lQ$e8QYMYLeJ^t|DAw
z;DEY0338aUI3&eOl%e(f>!pl|bZ+8QwMJ$Whzs&BolE0ktT0SY2Pu^#$)h`YT}UVc
zfQxgQwGoTSq8h3y$L2K_ok}x`&cNvMqdn%PaLga6-|xFg#DH$*^K;U$C?h&7+-13j
zkAWF!>ZX_y;mB3(CSsr8#csNc(irvii<asLIw-}RZ7j94>>|arH|E($XmuLDiYF@h
zR*u6+AZ}AL`BF<DIGJfMI#{nQglbuDn&J97&^H#uOE8(?^aBhm|JVespfM1Si;AsD
zHl!i8ko{^+#b^b5NrasHH|72@0@%K*F*}1=1bG9pfWy-)eck}rh8z-TWtb9eqvT)a
zsGd*e@B+xT1uyipT>>}P3sr;y#v|+qNE-|st?Cd%LVVnJilP2p&*?$z71R#Y>1I{A
zS(B%#8-^ITos*+FG+wZnLJ6IPjyA1e$la#HFJHnG!;y9?;pbsF45P$x99*4w6NJq>
zxY%}dePx8r6GF(hXV8?Y0i+SYzp?GZ1YDEbv9t})UET}YzA<`g{uC-oH?-s-5hONl
zk=Mj##u&Jl1$Gm!J^KU%OnDOe;D$_>N~bcW!IRQ_;7Qd>OsFc1=W6es6JTos-fB14
z6%9;Kt|#RvZ9R$)+}FXh1!osea-}WNL2=Q%vpFS&rUaRIiGUf6`f$mTFx9TKd|#YN
z<M*B-G)^(wY})FMYk{5Es3;Z??QnJ7Y#$YY=Q=IiQr*(_5L%0AsHNqU9uVc%{VRI=
zNGAGy`XN1WsBQX(bYA=6!;ck@sgi4X!j%D7x5hwM<G|^f#?`O~<)(cKgH;pl{Eyj%
zTgcSpZra6Ng(9yybo%;(?Yzwn-1t(i_-2Z2`lpE!u$dQ~E#=jA*TmMTY&)br%b_?u
z$z+Y;&YkA7Ktr`OYwRYT&rbcF(6O=Cn!YD)EfIo%$CLZTE}~Ags(m%q6D=z65YgX^
zc5SA=hShfBTOGi7?z6%-UBy$KN>n)deQr_8h&TvPyinb9{(I*3QuZcoyv@MOw^QHU
zZ!>m=4>?WdJwVBez5>kazRbpZNj1>Grrm#wNwr@h>+1O2kgJpAG-|Zl$K8=D4*DKP
zT8c)#j%&8pPN0D_f-LpYrKH)gJvOdu0uf^lJ56hPJ8=>~IHlQ&ZX!|*V&<Z5B2)8m
zj6t^{X$74hf7f^Urp%9UO*Ud%h_v=rIP;8m1)KyJUc}W}5NL#tX`4J@hN;&8>{M+a
zb5!i76Vp@ygPp(}qrE_n{J?kXpDD!beqBNh@P|4Wl6Fnl=cl^aKD=}Nl(r%Ja`^=w
zum|qG0ATD(bL#9p7}DB$bp}0vfMA=7*g1G<6a}$VcC4Bs_AxI?AUE#Y`_U@#u<M5e
zT$?fX=iZxWfVHo~Tj7aCsTtI}YKq@#1kJXO4rkodUSJqM-wlgM(RCBGDNz>e<|lEH
z{5~UGbfMSdAJZAmU%jp!B8va1J78z{U#B39><s@MnS1(|9J%%Xjseq;rySk@Vn7dM
z*#zuieW<;tGP9j5zJCe^%Se={tq)`8gR4=82~Y2*gNZNCH;c<=;b-aLdv(WT<UGN`
z7bjOCk3v&>^G)X-jsTKL)Af4f{rN0o)oRV#V^&0*INz_2XUfEZ=$QkKlUf1F#LWp7
z#&8B?q}dh|rUn|f1aI&U+UCG*U3COkf=}Bf&g;^4@m&nnstgWiJudV3x_Lf7<}}2W
zYz9qeDX%}DUSGrC1^}op2SC(78y4gJOY7kMpQ`9k4+T+*YE1Pg!D@_N8oU%x^i78`
z5Mx9oh^Mj?YMt@q-7(D(Q$;;!2QrAVnHp?99uGH{Vtw9uCXWd-iZ?O&MyyLJ#Soi_
z%)O$@?<V9s9;c`y&p9Rt+?8V{2k4b^NAns8pG8l?l$-2*7Sa}rmMP?M?9gDk)gxZO
z`WZBiEf8c7il5sZk%oQrBF+x8#_qKuC<OOCp`q1ziw9yP1TFB0at)@G*OyIAnevkC
z8ARwE<=<sbYW_{b5agRTf@ZTZXO5*pqch8DvS-#!%V}|};%-_{JbIS0-oKLGqckhx
z7$jshsn4YA-59-EZxG%080gBUaBMC{jc<%8T)UFOm>V1bO;PKTeue0I!W1^_z-4Fr
zzHpItJ*yx+GbIgO9f3*>Ioj6P%%yi5JWi+0s=r%vAj1}d<W}BR21nU#(y$Y=pLMUz
z=yi>@@rAI~`~%H(xW|+7=$tYn4rOhuI&G;!20!F1@|ofu52Gvs=t}v6lKj;3u8<5_
zaCkg8tVUwQ!N~iou1RIwZ@${H@io>n3XDM=bMcfk!05X=E&28%kY@zd9Ne89<MgkW
zvy%H(9j9*<_hVr~@*;aB;umFwHM%Zk*?!Du7@8ILSt%ZhOk<sIkph)1ccwwq2`%|2
zcz{)-pnaOFJVYg56_G9uW|^97qb|Q3;WvY2`Gt1woMT3zv}bk7g1=F#VN%7HKlx?6
z>|W$u>&>b~@zfZj<MB|%Ycri`d$bi#a<(p|YeS(pZ#<N$lbg<T5;mWMVO^~#__5Q#
zEWXYOXJLrz-|9GYkZDq|E1TzxyUgHBe@<m5n*g-Hl2>4a-Vv{|*Cj%D;Z8|?^f^O5
z6h&5B-?GnUeiih_(JGd%F-SvnLy$%s3F(a5_LQLmxpHQU#+qSkU%$u454+UKw`s=W
zKTf<=r?49G@AF*bv#ET=$#g^AnpeIv1v3Z<;q_wf_|0`@992K^uBoPa+eku6i0x}6
zE0OB5i0&FKF>`gCVz?wEVm&wG{D^qNZX>FWnLmF-2<8K|)G)x57j)fpziTSqwUHz!
zL-?jpMIVSb{;nb6*lR<pw3_(%gHd3}I0>hgOYs%JQMVY6gYs(^m>UEN=J|vpEJ0t@
z8rwKo%zac#JHHxWL<^g1%IpoNqKFdY_)Y`2yLGbJf=*FqV*Z)#R)z47D5-}>8B|6!
zj>UPtrPXVxo2T9F0Y@2)K<}j<`lvSO8uFEDm&FO$(leOcb=VQySRq+YfuB1~*Z_8j
zAU)>*>_L&QFt3~wnSuE0MW4=ns;jsm%n+-sjXTXJq|%60>R5`x!Te0h`yS_Oye){B
zrn9g`=eqBFqZ8R@;RrQY^4iqxRqb|dU8l9ULtgHsu8tZ|2T9=}(DI87{k*mgS_Wxk
z|IeUl`=URspA{GLi~re(gdAv>IX8mf6|~=Meapna66=AJM%{4%<*0;zd2*n@1ksMx
z1kNehYh@~ot~E3mr{FV8xAHQuSW*Q>FLaj3gtSwq`!cYSv1UhOvvHekho7_k95r=h
zrLjvZ`x#0^B^MOUE*n_=MS*D*zlUZHrec5;Ai04ATZJ6E(3g2f-gYH;M;<Qd`m-eE
z{qz3mg7A_q$AB`7s!*-#h1jvGI<dai7^XmkQ-=WWgkpE4r!sn93n4D@A563=%}0pJ
z$W!{(Xqru|G*V+4m5);SFoN1?12}Vj+%Pk7k4$Zo>#W9GCZr}F1xl?dYYoIF+U-~w
z{wK1hGZfRCOYnbRE2&G+jPh=on^Uand;(lmM_6ioD|=lXttc*jTOV9pdhZSqNtfZv
zTATdVY&F|1t=1`9N;=wR%OU!k9jw~VE#F|(R_)Gh^^A`e31~H)1sT?-hYOY5Y6<r4
zS;OtA&XZui=~-2>%xPlI8vfKXm2iWXXC%ZeM|$e5`O0^=2HKl$)&9}AoBZ1_eC$6(
zef?xw>H-ph5BP)ivc3PQ=5hS*Z8sC+ug3EKKJ;RiHZG=4^kO!KE~X-;#`Y$re0)&O
zE>5O~woo1$)4!AhY<7fS%E1$Sy5X-eiU1-#(H!Uq&Dc&*PVnjv{QVEUl1`Qhy^?mr
ze((}fnwFlej+UH08Rysc)5G`ECia5KuP|!)>Dv?NCgBq{2|G6TS49D___2Ck_vd#2
zej3V8LQ3RM^W*lpbP5j<CmwJD{|c1NuMxx$KAJD(JYUQ?R<snF*Ef4QH}m*wqEImK
zx<jJK`SJTXlr%Fexg-08iND{y{8;ptS9J*$!sf?34{>7odVHpI`5!rhd2EPjXwMh7
zuiroC;tim~fWX8Ni*3N7n7Iwv5s$zEYCz%YbTd*`>f%9rHKoR+A%!{8R%h&ci~~1c
z$zEY^j6vBX4aavUxb)E8K%Dg@8T**g(I?0?aU|lbPkeM|V##;3eHt$cHU=+N%<DJH
zD7+$`P|B962EL~**_3L6re#|n;c>ESRx-GMDVvR6XH3V^!6(*=oP>=9%?IZjq}N5}
znM>%SU8W^ZN(NP+m}nTW1eZ~LSJ9B21F|LSHGfyH7^G1^cafvA*g%!FEbdepGJ?;M
zgls59$OGhoL`0<hgR3QjY8dIwKn@ftyQX8+wo({19iKG+jIYtmmKZi4qp%*W;1)%8
zB-j;l`l>t!C(aWkK!WJ#dG(sv7(Hmp!e>b`o8y7$S}4)lCAP@Su0Nc@$4pkVX}p1#
ziQ%b#9K*<;N$%^8DEG6UCE>DLNXE$}Lahk|Ge=KmP%R8t#EJ!t{gJXR_R+*1bgH5-
z-+RSLpeH`CDgV*Ifj&t-DQ7CXz@5K?#?(3XS5~vFAk`!=Z!p;w>@Zl(-bP0wu?{`0
z?GKtb)x2~gFAGVr%La8yMLOcspH+;rI_K>4X%;bDXAalIwYrRG?$_&H;V9VFfy|9&
zHM^`geQDa;z?Z%fIEV7;LyZ{W98NtW@K$;lmLt1S4c;xj_hqE08Gc=;=9FIkrYlpu
zpn5;!DLoZcZ?K6Fd8+-VAG5`aGM^7{Dkf6<21IVMZ-_rqa>ol{g$i;epaHJK{?1+f
z+2)X~lvp0QytIxzPBWM>Qp;7&cCFP^T%c8^s>hD{Z>~Bpy{Dl^lC#T-*Hm_pM_ucz
zSrT{FRj25FYvG`nE4#()np+x|m0PMn<^EcWj=3ZZe6yjWvEmTB$w*#qrk3U(S90pc
z_+pofWRjRw$7#0ZJLmhRauy#K-fW}$P?CmX3`L27?lUlv9J6zwRX5}P<^zL6)p5hq
z7!9lmPZ0T^LZZu(vT|M7677QOF-U2<<ToiH6)k9c!)#aMe+p=!1E90i=fa@_=2INh
z)q_Sr(rs=Mon37a$6qD7v&;9Le5*>=OmY|WnB-sDPCbs62_7o5O*K=^t5at!lVMYm
ztv$^rhi3bS^X_VB39Ifk5(E%eZPuW5AmP*BEviuC&I@U&i>>sFa}SjEw%uEs8&+tc
zUnT!^=}H;J=piXVw|9J;8>cX_=i^b*IyYlbK{+SYQF^Tn@2O3}?|qE&t4+w&aj#1R
zqz^50)=k545Zv+WG)pfN^oEs#035*YU?wTZ;5h^hP}<#S^X&S%PAeJAxii`CNs&*k
zI^%Foft%+D^PAkn80~9ah8T^@9W#3i*>t6?S7R{4(*o8F^}sxC*rD!)Bw4Lb@{E%M
ztr;d_BW<QRqE>$b+={)nIcR5>DzTE`LkgIrFnZ^r#g5_>SOyy9VNlS+1Hmay{v;!M
zw@aXhmvfbtP6^3rwW5^SYjJGz-WSv;z@=&T7#Y2BWT{F?ALCbL`@<IUk(4Vyv-}H8
zF)n^fFkW%N>%b))r$AG-8Lfj5Uu`&P?EV|*vn7`-v3S{|$YYuA>!|@qBp_&lZ<!i+
zP-mN}w`PIn+^HG1?mkovBiZ+5#2CGQcb3@0i11mWTa@;kZw~Xv+BNoco6mRPScAcg
z$9CDa*)~ouPED7Dy1FqK0HA82P%Q`%7figJVv}oE%%J^@LRx<(&O}{_tImBR757xg
zJfx{8WM7DgAquS6M+>Dk>`E36?M29Lzy7>q&#)b-8CmR*_|P!_%^F;U{%Sq~!)#Wa
zwx((w9~EtQ(VLR(Z~)>Mo$2SS4}x1!XUxOigrbF<rrfQ-u}d$M#h8b;Nv{T9n_=y&
zEyMOeMnt)i=1^<4Yks3ENVs-mWMw#+Z8{^Tmb{d=W!@;uUI+vq{8g7)F-Mq_<j=dB
z=MPnh18c6bZ}&H@eLLVy3dwfP#@C=A0b2leh^HU(v4YHd<H+JS@7OC|1SJV!#<>GQ
zY4V+b`pjV1DQumq!}W=1n`d^)V};X5`_cU;mte;qzMjjG{7<1Z7sH<1{K1OChC>+F
zBCqYQ_v5%MN78n(nP|%;^TI5+h>t;iUY*-Bp9oZI)9zt@KNTGDO-@kcWn=x_E&Lr{
zlIxL6YW|(Z7Ig08hk@LQD1#q5&7eCxv1)Mt&8|r>VxdKbPO;lVWmX5H=N*0q)nLPk
z1%Gt}_FOqD&wFHb(==z#{(P>zN!fKA3Fn-VaMQ=ozo+jtv1&W8hOM=EDBgV(SMn1t
z1l<JrdCAv=2|s{-NmGBJzE&!~JP;x3WE1dQ)~WzKBX{LFzS#`j6S{7+b{nmFHQ`yZ
ztyy4@Z`Rp0P{wvxCH~gi{!PVob;;3&uF!f6_lBwjj)Uh=laH+J8a_PKDj@gusrZ{`
zqTzQ$vdBE1*3-GV84={5;zOL<qMVJMY5po=^;y@|f}lEL25LT<5|*=jxar#|&RLLx
zrHL;+PiPkKs4jm5zG5+m4Im0G_4IaR^25IbVe}s}7{76g|CPaDX8iA%2i;oJNr!B3
zJ+I1dpkcxYApVsdI>#?DWmK)Q&$iV6hp~4I&MfS{K4aVNbc{~Nwv#))W81c!j?Ip3
z+qP}n>DZY3XWp52rs|ohr|SB2KAoy_s`kG2UhB7t7buDe*#7y#NW@`40_rv{Cwh}*
z!azd1k%<fCe6#LPhsW37#+UKe5>)+r(&yiPQ10NTbqn6*<R?JOBkkUQE}wQ+Yv2!m
zM^~y<qQ30ExbDzqb`wMRfQ9mVVeh;TYDzdh<Fd?aIn5Is6i*~o_bf}{9lt1WXMXSE
z^@w(+x;wva^CD_=Z594Wo^>cow*xFv(HT_@kP>~ERSl|=jq8`0rVN)t-wv}njjnWU
zL0wwWs7*~jy!KZme|XFx_n4A%Qev{{FGU&m&=e2{wQj@%dns%8Qm>uSqRJ$?YN+yR
zd+XYioh0{9Z-w9AK0j~2Qi?zW!?k!6zGYuoK40hLn>9U5ZbV13FvZ<puCM!t{P7ly
zr5z@|6R%6r%XKr&_Th7AQOf>dJd<NqCCwezL!-BCJepYYgR9wYz#}8{MtLP;cfi~0
z+jIiX%fqnrqPZw4x$<)|A3XnZcv4xCEDM5+5;kdflNwIrK^~YjS;|2a=h_@=Xvx})
zs~sGa&}&l_jkVH!ZYd4*U*5@~;YAVIpJI~sqB3{!t3NAJm?s2>dnjlfpS#+)!o|!(
zClAZc|Bh0l4PF@7-mRDtkZT-D$7*F?P>JP(HI?rel$g<t`l4@KRqKzZB0yIz+V2!W
zhWz3<HHS^@H!!m09a#zyv>Yv4-D@6&zhFFOH^M~NJ3H41uqt4)YcP>?v4SZQCCjYz
z&eWoBH))&WB&9ljk8i51ENSzXfB2n=sAR$M!T!)qyPO7FDgBY8$>w3jQZ5xjA%j$U
zhwavht<pXc)}vwk4+9TX(oiaUghQ_(rJj>fnVMcSUTwLvE>HW#3oc)wi?M>;7(szE
zDqaX&jg1Y;g_sOm3rRzEb#+tq--LhTr8bVkT1DN`0d}~~vX*am8c9>Uafb(rX??SI
zRbE{Ni1Yn)Ny$<nrz&)prm|>ZAjS!$vT`y9J>zp|*;oOKX||5CddP+2wU*`a7+mm7
zYQW-YBR8faQO|gM`5>GN9MnF-@FD}LGfT3{lIG_=u|){X@Stc@j=i10)^rayboi|f
zQ*S)ld3;^iDGG^$5HPq?TE5_)qExeqS1~_m%95Mt#mSGLaiff1QpuH|ey2Oeq_B`j
z$fRGLPEAR?0+6zw0cedE3WU_qap~sk%AwOPv{z=^YbUt^nCH5X&oy{aUo>X+q9dG$
zSuh$SOYj7^Oi_jLCtQVPI)<%Wzp|JH2Y&i}d(zU}U~-5KJ5Vv>H~~Zau%R;$vihHt
zpr|CS{Gj|cy%1c}y_ijGVx;72;8-d$iPOcCl^Q<lGor2S;M6}x(<q>>{%p=xm<v&<
z&j)dsrHkiZ)eYTCXS24c|EcrEoD?7B58wmo+U<KgN3YGDqv%Flw@JQ(;xH?)U<<QS
zx=+sR<&Y$Q0*=<)fmtzT=O(_dmOi7$+Br1oJ5kEynfBsZKu&9ioUt*lb^fOO7Jk-k
zPyx?h$Iy4sUj@H$;VO7;bCUSEe@@$coR(IV!hbdElnFx@3uXN5OT?ReLP<F}=~Z@c
zS6}-o$_MTrCeCJQXrL1LE%p_jcbQJo2j=*xCpG-Rs2JC__B1EW?*^zZ_f041Rh`b!
zOGT*LmR47<nRO4zdeF~Q+`+x;)eVD1JroDH#D{lnqz;-LT=c9zP%zO_1S~TOz{w+3
zowVRLSo@-u&sIgGReivZdo;L~tlV@&{|1^2@N)dH;Ak(ov(wS`;lq-x+)Kn{J~mWz
z8GUD}45hiRQu0Y1%uYrCSDhV_psiSh)?D$WpJlXp3ylX?Dn8ybLUxBuoOSF55v_Vl
z7mWgQFg?uKuKNK-q@u8PGIdvET5?nURGP=)LhWzjd3?Y=+r5S8Y#3G7E0sDgsLN-P
zb1Kj9<M~pxX$W7Gh`g|o5Ij(s%MYtmm%)U-`#aAT1k^yrc>Zj;p*63$azIXdd82_<
zp(=K%SUGgxWuG~Z*x%kx(4xH1S>y33I}|FKb?_&@-Pt|32d`X_hg7?$YIELT>lplI
zD%=BC6J~tB4(98xO+1G8$fVo-;d@gCDt;jMo~|hKYIQ=0W{+$%q(v%{Oo@o(z-Q#x
zbO;!tX$J>NOA-&0`O<x7FG-cxQ~FVW>y6i>)FP>KtIzo{>sXV46OW(9QuSMUbgT50
z)NN2u8rvRGBRAfu_FD!r&V7w-y!peS2Cn*{l=J2?Fkos<2_YulhGt@{yS?qQQ0r)P
zFw>wmUx}Zr^&fe)j6kmA_SD1Z<q$}oDqHlW1b2nTijkpfxiaf=reR+3L2|>qs7vf?
z^i?BovSnxsn?t;Z^$^8(h@4RmrTTiVa-V$}+Y))g29hf4j^Fu<dv%4y8m@UG50G}4
z<~urAR4MPbX^Z??rt)fdw`T?G)xf{nOz{3XPX`A?WEYl<tHfO>4SDm-zJ@F4v*^0y
zGECjaz)ZDspY!(hk<*^0e*I0`H(neLNQr)~_9Krbk9fZ0kj=%IVVO6F`Y%I2L9i4Z
ziS?vRpjR;A?9)p8&DcQASn}g)o*^`|!;u&`92*mC`>LNsH!M$=+(H$odLZ_;b@y=%
zJU$4_Yh&>qG{^m14;PQg4o&5yROaAW+vEgaXCYeAgmvle#+Ig8xx+YbRB2CmC!g@V
zTy@+^|8Cx5wjk^5r;nJ&rN0$y2)uP-;$j2g(q;4om%WX<Vko|+DK{o!l;obe9TaE_
zr=MWmq%+}03qOo20bR7FSTC<8ldUWLFLR4gg#`V(`qQZoQo--Esj!fpr)wk!i_HxI
zQ3;iYxXR!?>r6Pmga_J0ym*Db;#Z!-D2Az5c@rW2cvI;rE=3E&HL0i7Z}mSh>9!X4
zga5^;Kv0z$rB-d?!^WsXlh7pV06$JrQ|rT%sADF8pC+TOF%w7xasFb)AMaU?oIX~w
zMn8NHXX*g*Xl&OF&39*-nF0M{TcfqAia)da0q)qxHGK~hi#H;9&t8*sr=RfQ=mEOz
zWb2PNMDCTUFkY#0o9Czzo7^FZsz?F8Ab_rk|1{oV;^6q-oz?#zX^N47k(vH~!4H^N
z|4ZxZq~<qjY8dsiO1D;^yWmGdGc$KRy{H~>SUJldFq~xg2FxJ8_!E(^Uu77WEguiw
zCUWAyA_7Gc>)D#Z*7jEK)Xdc5)J?}dt#8(pl-<NCgaczE_gO(W>;(%GL_3f*w=H8N
zXvcx|5D*Q9nXrN?wHBC2AlZav;Zo46Q4JZdlt4n}$qsd598RGu{-;b56bPWmJ=9p|
zb&+O^#UX)8H$sl*h;BT^93GRA`$>L}g=>4DN5_1Xq$daa1ZY&5A+Ng7MuU!eRML+8
zn1|rD6Zny+Jz-I)7f{>tJGKnn!tiJ(sPKV8)}R)bku&a*g@O{V61lgNY>8>&SLy{3
zHnY9R4;&5zBf8<8#7Py#c*U?wA{BvzD0;DCPBmJEjHOTN3E4@NSp@h5GYeX(N2Tux
zG4IBHm(<B>pcQGOxp{qy48Of6xJRlLL_qPP3T=1`K{%_Q7<2Qilmq9ZT>UuEiU_DR
zbp-uLj7@II*@70{=|64&6<APJqT!iF%$jcDLY|5&Mi&cYZ0>JQUBgF^#P#k$DP++!
z?$C^Q@NZp2x!lCiKHRtpD@T*2FD+l=m#I4XM0_@aSx93%K4eRnph)f*u5e*ziy508
zSfYAQ73yxn0TY=J0a_ba^Y-)yt^o!Nq8?1>UDzp^rELu6MK4hC*u;akQ6BbxpMdRt
zzl=6<9^;`I+}Z|w6B%G*ZIZ&E2!&B+?2<KtiH6Oob|-I34$fmA<!ozN*VXpZllQ8t
zwJmKz>*><SgD;B+`oQ0A?xXT8_YDc#?6TlH@++MXy<YV%h{7dZpTw$alU)SgKtku$
z$)l6WC83Vi$C5X*?T3{MW9wOTn{UTw@2B^}xRs0dT}sv4i^1dgWhsAp$GK0Zw`->_
z-)HCB+ht7kdv(g@>ES`c{|Sjg&pvLB7c1<6o_^6POmxg`2H3UVP-Qu@E536Nd-`$(
zy<PsW2=(>3F+leHCv&p}@*|7>0o%GOV^Foeqo!HQ@6T`tqjle<+no3Q%&WpTJ!`VS
z^R$aZaIDWvQ6^kMo<bqgU|w+GgwC^vx<ezGc2)DJW)MbMRouTIsv_X|cRUIa`6ADc
z!EwU^ztAOJ*RS{G?n&xSa@B#__O{Lr9&?SVHya1eR3hz|si7+B_NzrRHhj5if~OQ;
zZ~nJWGq$uS!UuLF3RUI?UlckTfEh2rUs4k4C+v~y>%+%`!HBCxCkGdnEJD|69U>>&
zp?{{7QYId(ukMkIT9ZSn_4vXt-x=IQ+tN10ru*qcEEh{L`*#tu%?VGFUa+{ls(ig~
ziXE_(8<wo~+xK>hZQI;7p~j@Pq0W>IqnSLcG%TjNab+PTGbB2DRY9U?L85?Hd9@0D
zaWcnjp^Rx%$#v3E-Zz5b`gFc$$OMG+$+*8fnt<+f()ikG&dUh`#})ox=8WPI<|uVP
zi~I?m`d@@0^^gL#?8@Af)rITfuv)#M5o!}Lt_mZ>O+sPP?gfJxZl9hJ^amVAP#8Ci
zVGOfe!H|UrAl!*fd6@dq(V+c#{Z?KOtV}T9q2|XI;qL&!(<eGTsOSPPr@YW(>wdV8
zt!!}n%w`aCd8#$|y~GBS<$)5TFrESw6mbYs$!tG@AmDlPC%XK?BTDGiG(dMOa*`PR
zjsjRs8S=CsPT6m<hCrH5Mgc@($9Bkqe~sjSq1aWQjz{{VQ6L0RgW}jiz;nomsF=IW
z%!(uK9oo}YWZ~YngR|0%y(XR<wV-WE%Y|L#dPdmb8KefQQQ$%l2C=}nql^(RaQpY0
zROC`jD;=2yhTf*0Ip$Bc&#pSn?ISK`JRx%qSj$@|m3OHPhMSc!REylI3{wkf7go8k
z?Z7~0Cq;u60)^5L*7FyrDqpXKz?$S$f!TA4cN3&)N>N>TMubaRLf7$k(F~$F8=pT3
zyz$HFdQ~fIf_f7UnP)_4Q}gveh140H%7Q7B;;tcm9kxtBAr~Mjie4;8uR*J`y=!3w
z(&)sIoa(Tq8862S#2JoC^doX*V9XbI?`DPqUSsD{ey0eM$0{9_BR=0Jj_2nlyzEbS
zEW;h*sG8e(p&?>vfRH1v41+)3_<7swjr`9MSmemDNSwT8B(FHmE15!|%-0UfN<`uP
z(~QOo`8UT=+QKP;4^A>06PrBHW{k(FfC^6U<-G>xh;S?bZjhiOxsCQM9Mriv)<(aQ
zx|Fi~8*v;<D7@f@5VlS2Ca?vbiWaM;bSYl?Gf~M1UUaA?-7*Us@vp|eXh_Iu8g$Zy
z?^RlWg>?8?sKJU^D{goFFEtd$hhNiE0aIcrDaHBq+gBs>mp_1OR>IV@ZbJrHf37Wu
zi*R?>8MoIJ>qM@HMfb~8y0tNOmwug74FzR12a|-HNzqxF0Y$qG=-X5m2|?&2uW07L
zQQLQldz#1KylD9$yPVCjrJzKTLF0@|9RJtp@xZOofvQ*rJx^i-zmJbQ#10V;H;_2D
z`4f2dslO)HDh1?SJwnY<3caYo<E#b))*m-$c$rD2I+o;8IaET-WYYDtf$0d)1q%)k
zi;eJe!OW#W6l%r>0T8Zr_OaM-CNW&{sGB-<CNZwT*XM;&1<<+^sY`TY+64zd$cmn4
z1uM+Mhm$^Cp1ZYPHHW7!KD%Rzu=D4~ZwT=i<dHLc@Wo5)p5>WJWd1#CO<-G(LpJy%
z0nu7)kboC3{He8;i<1&iBQn(Rp7m56cFPx;2tJBi*BfvZa~2IRV!{-<&#TPmCztEl
zL;tFat&4VC-$G#e5Mh&^3lTBhl;TXQVZI!>xpXHI3jEm1HcXZ9zMd0p)b)v3!bdGC
z&&_BgfLg+g0siq-g{g3Jg%`KU;z+a^)Jes5MHyHb)QJDza~Sn3Ptt4g>+y^8xuE@M
zN~02C4T>}#!+X#`uhqLzSZ)U7ZAS`r(bOa75s^eRj{?kgk}9CtvB$;Ki#W;@3>T7s
zdP$9DXU~5VD;C8KPd_Jo5EOeg5;AnffL_9HMetvJ^~-^W%gxoE@BXPAHca)YW{_L=
zU0mG6^-yqjG>Y5joUBezKD4}_c0lpp$Y-n9IE4I)PH|2WyHr=&yj7C46w)<f;zeh`
zr>6$5e1X8$H`=+ug!YT1I+uF3WKSb4v2%>8lX-n~T_Y{`)yr0l8`|6WL9K)5RIMWs
z^8PUQTg$nt%f`JE{kUnMv9?$kes=c)|2euiX~iW}bk4~^NyfS0itk|O=Bhq+p(Lnm
zxzRXU<}D0m(z<HofAIM6l+jA8NA+(_Dc}1E=P<ByvM*b2PYTYFX#?=UiqTrq{Kax(
zE>l;z+mLJmMu*^>GOzGuf7$7zbca3{7er2>6a0(Cs|r74+S!m$!sGurGU=DfLl^EZ
z2|5DqKA&<P)x~!IJ~d~?-k$AK8b6V@;*y*H!@`k;@jomanOOci&rRe1VAWeuJC{`t
zkXZ(a1M%glTFP1Ide~vxaQSco$bS^{PCV~yyFx6JG1ouZ<ZyAT8!_w+$~I;ANISeX
zaQkd}yUlZO?jmxqdVlbNdjC9Sm%wW<aRBMF_}lfF_F0<(RDUHn{f4c`{1Er`^hoQE
z_%{9Z%!G77a&&M&=m<MD04y72ELkRjlpbKby6zcE(B|Zd#2x~ke%$YO=i9DK%(S94
z@09#Q&D&j9pBu_cNmxU{VSiELMV=omhjoc@^tJyc!jt?xt@f4McFh(qbJXKOVVKh}
z<kCyT9C{vIBWCeU^6QNO;u#QZp9ly1q&R|{t9w1K{xQ#Rd#JdN59`2-rZST&Pl^AX
z#JX=4pp4gt2WuF4u`v}KQt}dq+&U#5E5uD;dJg%6#ZJ1-<ZtBYHh%EA2@z(1z~PX#
zJ|XR&0{6M25z<TBP=)>YF`V}2_rGo?MC6_QwsB{DAQIXoo7^}8a?kOg_w`Ye6ZP^m
z-Puf<JV>Y9D#7ZCbLQG=#zFqqNPUk%(E8-=zbYYk{bN)0*l&#?WBo}iIT#`&^TRf+
z+))z|+r{)3g|1~*wvgym92%I1*DKYzL2V!-{w!}EkzyP-5S9k$JfQ}HVeMF}gA<Fc
zSU3zXN>WhiEAVd{LG{Toq{fk?BCHi0oY1qqNVEKy5F~k><w^w3o7fb1tU9^L$!401
zY?@0ayC6!X60uO2LTIo8A?u?-3e}*cAeJge<@;)h%5z09Y}k=xdbN8c5c#UwUD*nK
zZVByF(a)&dp<RU$vHplRqZkpM!R7rNsq}`nJ%{D*Uxx~pyKTxkOWHEYI<)?{x_%_E
z#V^@&x5Y<Uxn{fXlUjC9Z{1$^u)yx9%zh7Tc24X_i1Az}hRJ$j^YHjWI-C)B7P@4g
zV8e#f2xDLN7R!S{+7g_}x+WKdW&^aus_y`=UzP5s`eB1{+c_J^h>>c#Vd3vJBm$?0
z4Tw8nYU5S$K9z;6*o7eYJ0M%vDoozRmQpuoF6&mWfYT5_I;xIGYq5%|&0luez(2gp
z6K$uWbPYHYQHDIfTJou#h!eF=%*&}K7+oN~0!~lfE5P0oYt=YxunsT1Eb}4!TXsZj
z*f!GwlUeq|p=;)+6hwXLT+(7-5_S5cpJGC|f-*g%LU?F?yw>DiyJhU(@hrck%`fDw
z!xHMFNt@ObZ%md&(ET0F&t<(Wd=g#n6N!we5cdwLEX?4({$^kucFJ2&CuI&^%qQhq
z?uyL%FZHST&w*h|RBFyo3Ll4nstRf$oWr7y%151O-~t!+t0Dm8WQw?XJ*v<3efdD^
z2%!g%gTzc1(jHo1Lg^_>TiY~-1DuRSN$^mElAM&(u73t@#!O5TOUWPmqT)tn1ACNQ
zo7{9yxj)8|QG%{cMCT?BpC?8uC&q9t@m3JKvuKSu%Zlwmy8Lgb5Zr*6#S+|pK{Os$
zJ{|GVpvGeTaL6qOPpnX-4cmuw`o;5PSFjhoOoNvVR~cA!@`bg*QjKV+^kxJ+g%b6r
zHY}Xf*IH%tLgh$4Z7`ECRke7h0|UjZx^||%=DQ`W<w2`5PH4U{`S2N1x$~#r;S44}
zPTV}a_s>!77zYR=w@hnt5C6PWky~zi==YhA^?e(A!a&?UyNtYa?0xgV78$lW1IK<g
ze|^}g5(`i~w<VolQNaK?IYS|Tu#-mT?S*-!>HirsosTe#1^Q*tdlRLIC^p0;LRO~~
zV8l$v9yFM=v4&f1C}X{`>TncVe6m3gc%Lf*y|@N60pKCn3rw%%Uxio{Bl)rE7H3Ig
z_;((WH*_&}pe9c#2#r@R-cCTRZb@!HBS4=t8RAwCmh_Jln)S*IR(2W<r)5gJm4f<l
zlJy9Xo(Cw2xukx-seICp)=tY((Pm~(DYVEp3umgi)$tq7TBEZOOdv+d@mlvnrm=Fl
zQn^1~-T?g|eIqQ;A^nFvPt!Y5`N^wwMWnZP|Agmm_auC3K<adq2Uk)Y-zC;n)kD=U
zsW6tZfA-gnSKMj>TYp;oyv0k^d)3#G1#xhCt=-Qnz)&UpF@Co%=ERi!5jJ6`O=Nm3
zQ52vru{jL-I2KX3oyQO+7nLLINd08LVW7(+86FaLry=@jcfalZyg&TZ?fr%gqj9aP
zBlwc#@dL)^HA#ogz9QD)Y5JbXl20@}Tp$w_R^*{>e?w<!(cgAH`VsYj$AayWGR9(;
z$K1W?Zv;XNK=swSzySp|dA4qI`bAH_>=K2%OzRq7an$l9X(>hdeJuOMGZj$&$$Z%*
z*FKFonR80H<tXo@1De2AxP^b@xAe0sUNogOhShcjCnF+t7(;S=*&IeG9`aZX_kXxE
zkrM@U!XjnzRKkax(f=2BX5<q6fd_rqCiKe=9Cr@F^6hQCy5j<x?v=%pf{EU%8bst&
zRH1r*)GSQp7saZXih*91j!DYIm&c6B_SL%~f>|x`;BbRuDc#W(D+)OlpByt*ehIzI
z()4nlFn|fF+RyY?4c@OFR~{o5$mFPn*kbUczgTNBnvz9k4zFSUb4hU?Fn4KzFP5@_
zW>1hkZEJ;GnGuC7!w3C^9JV@VnKY1!TN7QI@&l}g-HJwEx-@qnTaSZ=oMkkL&GBkm
z=E*)wE&r%WJy^Oh?-lTKOO<eEI&&F>_o<{`;)IeAk2M&{CzTT!N$1HjPV4w1$6R)Z
zlg%7uy~kJzVdAuqMr-0jOY2+PW&jd2A*O&3k!ra`=%k3&Iu}uX_f5Z9hO>k}##z`7
zOSbWJ*)o>}_%f^RJvNtW7(1^Sa@O8EJ?n~xP|>fw-gg<YB=t1B3$x+6x2xf7$&DQI
z=XSb{HAw{eN04sCVS|-DfRl_rrAafZdwOu0-NV~Y8A_~Utz;gliz1gO2K6Mz=;Di?
zo2Rgp>cnwvO?Fpm`%70p6(5Z_b8LAF+Zs!+1sa3@vA5qnn{%L(Fu(8z+ulO`r<RE6
zKde7lzZvoWjRvbzvo)HO6Yb&rJC1JBgux9=2$aUT-2b6TVauR$a7xYa;=Q}Jm)ZTW
za49ku7@mq7$xRga%je~zkDh;vk8it+PhJP19xk%Q>(d3at@o7P4sYSi0OYUO_XGQv
z<s`@4(>uW*qnsq(2-=#jD*tp*e({2a^EqMfd_=_UZ0DNF54b7}yT)drxxPZZcD>kp
zY}-%rLgqpL_<Z}=&i-=S4^yAYcue^$|9beiyY#jF<LobO?m9chuw(Z!{loS%7-F!&
z9{cBCaQBzH?v(AV&sD`qN9S!p1uO{#Ez3eYQP|1EI_%20Tt;<;cnD>(1`={=G^JNd
zqG2*>f%w>4LGT~-H&h%6ETS)W$Ny4n4vu3)O(;{bz_G6<eeqb2*dfHB%wE|$7<rK4
z=GYl2|0xh<TqAF9cbi>J{?hh{7fUML!b)NbEt|XXX`$Tb*58OZosTEwq8O=oi+2>0
z?Ztg!2?|$AN6mYE*nOTDqe5hMnZYuV*{t`PjI`tnTw|<)%36HFLGh1Y-Wkp#lyuo$
zrO!$ceziMtsNq_bHCB)JmK`G_|AURhWbU(;2jns#MB=%fN)yx?%-``hU<N8G#rqTF
z30IzZ;4#jTI2y#pZYKoQi`1p*I7Xud!}jdSooSC*h7%@O<!uQYkaK&X;(chJRv*L~
z-<pLg9Q&49l4sIA9lubhyX+_vQzroghR<n|S_g#8DGme%P5p8l6^-I0vxU@YJ6vam
z^ZLy8_3w{koTs7;ql4Qx*&Q<gb1i2&n7E&7$Gtu5E^&-01c>s~gbyi~d(^3Gc?y)q
zBCJjIjDb)@<C?vyB8I+6CRDghK23MfDMu3?vF1zSA4<m5^bb3uN;;!#!XN7s&m#C<
zm&8A??JilnBk9)YZH<1!bX%kfXwj~3jg6dCpUUngk;qWmh-9f3pP;F~QphRYL4SyQ
z)M;(b&8t(aM6xU-<3@G<mJ`F{>YZ85g6qEQAcr0Pt8de3N@`C?IuOX;z-Jk@EXv$U
zCd_RnPF4*xzC!hOPsBtO466Fc<x!2$z1<ngoRcvBK06?L4d<Tl%Mth#jaTY9VO6D!
zH-jq*9ZVD^sUYU)5MGri{%=&2A;gOH$1lAJDPfd6pK+G%u?xYZFpFR(M!`hNynb6Q
zW>CY}K#FM4eQQ<LaN-VJs`5l4?9MciWwyDbI$g0dC1&dOIE_rBp{FL{ycw!OQi#!C
zaWv&HIvD@DhIvA&$`3Wmpkq`DPbE50a;mZ1$;Rd5{!v6%q02Z(ommsGjH_-ieB#xa
zI_YVQYL`d8hps?pUPWPu$%ZB$I~Niel{G{DgbMUQz@pALm)0kgJ9bswm@=YVOjn^W
zh*aUvtgCk$Q?aSbEKd{^<D$c`QC{E&{3>0$JEuaLE_&ZX@5Uhdtylf`c`M7{yC?uf
zq>>4qSN-v+;pBRv$h=3ViX6;3&t}Nmy7@|g)~7nu6r~VHkOG}#oMyZ%(h6CV7)1=j
zlUVQps%_EJ3$2V>$ZHL!l?Xvf+RAVrZ@+s`fw2um$E&WPLhKTQHoy~%)JthY2-5CL
z>W8F(VtD^lZa=szjuLS07v*>ntW(pqj1%k0FCmOGS<Se*$}=m_1RLpFdVkplqsFzz
zH5xs({qTuX51FlgegMjIkbMz=zvtLlj;bg_%kNk#f|C(mlS#DjuBTMm14%CNPryfn
zlaxeuRvhowbH*pc9GYH%%RqO$Nq}w$4}#jw>;b!iv0_r*fPUfhg!9P1Cj9ZgOqg2*
ztBJh{OXn?2i;ZK`DWW#Dz>(iYqFHB}A3C`ib@pkv*r(w}Pk_5S$Z{|Q#B#iXXSm))
zzy|5EIO#_h4>}5aGcOBl!<ncBUg}i2i87}{f+|u~F}@1o&2p8E<?nT#)eT83ZomWe
zl+2`iFG1W?SLR*9_?b}+7j#i2K^g<RP_;Fh!si49ysW>6%>CI$AK^Rz>04ZsHE!4%
zToZ!UR-C0NKdTcG7ik-V*LZ>9PH2M9UOq2eB4#o8YsCW>gN07Zwvi$Jr{@u;*~ya1
z@_m5}i07E*t!q%^?>PQhhQ$p3*iHmf{2~?RMDV15s6q9OIH#0GgE=NKqk)7gTXyi~
zgX9b>5<KWuezwCzCH_W*zzk)>mg5`Zwe5pNj-rJg%N#9@o^>zo@d7N~Kl1KLqkQdl
z+bfl&VRw9fN<{TCZxrL<<rNZVe=2a6#Z7ttk~->e(;-XQc$L|b78p-ODS5_doi|0M
zzKBTGX?C7MdWT>VEa363?K;GR8XxOq#S~=z*_PqYGv*n5)%(}9*>yHdWEzuf@$pQ+
zPtkzR6;KM3u@!_3Wk#s?D&xkG0ffA`=o?1t>?m)xx;c!?HaGR@T26I6A_pT~rY(C(
zp4&H+wA))DqNF0^`b{{fda<nuGfc-dwwZNYw=TEFT<43YI)#1hxl$a%b6M`BnhgO!
zkt<0^7!BjvfgDq@X1=}i8%vjjB%)Eu-`b-NOI~@05fcHsF$;VxX>zAtN#++!HYs4U
z0O9}(1;Dj=dG-%IDFpLzAuqPA5^Pj#6Va30F9iVN&f8qxRS~Wtg{SELCwECZfnw&j
zHz0kmWuGU0{#i67trQi;0e0@yVY=4>eN=Vl6u+n_Pf@#+J)~609NW&^yu+3OWvGjB
zrMhwxgoEB4b0RPPm0!|@Tc7;<&}lVQ_|ICBcZ1m6ydQJ3btx&}7!4juS4k>X)KxpB
zrel`k?!_ICUKx+bG^5+-U+*s~kW&-7YTRB;jN?cH@E6O98^d|i1kM*G>Jhn0%5BBj
zjfceJcP^O0J<UBSM+cnY6pnd?`bud_xSut|*kQw~nNdDUf)903kr{Dy6b~T1@z_;a
z%vcQ!444wtx5}qK|5<rZ0(KBa>z_q$O_8Hf6>S|OUIDt{OzdaT&zA~VNqwEP9X8fB
zeBHrdGN6|O+fK2#ETXk8>&b4Ly~g2Sh%?h|_I9`xCd9AUQ(2l<JBG5724-Iw;4w4}
zLDBFD{`&Njz0B|r=7YLiA#fwG79$?QE@;Gt`P|J)NVeVURiOEEy9rtrX~E<xwY&|h
zO-vYQR37HhSs<xuz7H_*a(){K%&j&N0n_X1M_y904x<PmK0^s~9Mopi;o#e};&C7l
z=q8tTeb|XMJi~3Nd91U}EA!qIgxuALjQI-$&+#9*P~m@Rq}UM$E==3_h*%}e;Str?
zP?Bd$p(VJBYvNPnS1q0I>F1J8Td&^DyEMl&(h6?2W>+G43NLT(%a1tcmZKIFt9K!&
z1lS{LG~{OALWIQ{qwx^8(uOV_o`<6;UGA`2MxxyOU%FZDE!mF>^z;siEpN(Yvbh>K
zkNr3y`w<FzZKqyuR5Uwl<v9J$;$5I==0geNrnZuE$)y`FxUVj79_52`=Dt8(AaN=F
zQ>)4HA1rPh>@5FPt2y}pvbY89#xi?>1cUw?`sUq(Z--}vf9vehzI;!vWfbPFEGk=|
zD;qFaGoHMJC1tlMIX@olJUpM@%k1m=P}NvFz0>_{B|afX1A{5xi6etgFNg1s^S|}}
z_Iu!k3u(yX%hA*K_mYA8xB_0++J41ySU-9CzejJ=771CiJW`qer~iHb{r}<|BfBA9
z%lUSGd=fL)1KrSn_nW^z-9BFb6E@a17aWo^li;wM-{vRkLxeyZdX)r)yLNefnZCc=
zjfu0;mcC{FM&8`ty|F}kO|IR=H&RCZOf#3m6+ECY?3meAnIrOvi&x|FIt(lnUrGo_
zP0ZFMtN@acHawDws^pU`(bdT^0%aE<p!mk+jC|SpE>y>m@(Q8H08`Ip^_>`NdmNhd
z&pV(^x!=qi@;fK(Le+;bm7FH-d+OCAX`*`w{oIx_tAleyuI%bEHMLuF*1RSr;L(7!
z8PXpGc-(+wGa1inx$H=U8{uw&m~RZsml^MrDx!x-0vY8LI7XHA5;cYFUW@i(aeNwf
zGwI^_D}v?Nhkf1*eD!1t(F~U>M%@HQs_a_|>$hw<2ckLxWm9zV8*723WUq~L{^w6&
z*|a*|oZ?&9`IOkJks?8C0z>MVFU5)&BvjV|EsbhLgVSL}g{wWv#|NWe7G~6Ij}|ZR
zqJuGm`m1}^a5_YkpZ-?`R~-!^26!V*Yht1M)T~w7!tNI2sB1`=N?*Sa+hehE$dZcR
zBD@Sig{f1Yq-R0aF@|I_Tq4sUWaie7ih(tU6N|0KR2qs1F?9gpMBN~d50SxrbkfU(
zmJA0duNWe5&g6jPRjD#bU)$NsCYGw>4=EjIaf=D#hXZ>ueu=FJC<mzkT@XBlOcErQ
zXki|IWW`?u+#`u1jf_F^njgqXQ`DowOFQJ0fsSdo8Jt8knVOtET=hxeU3MZ*s)Lw&
z=zK8;MnT|_E#hp&f5<Xg^C$HN$%srrhRCwSAGYi3l!acQnstQ&7m*TEGzzi414)Pb
z(;}Hcd)!MMh^scuV{KN6l8L17{hNxQm@=nI&2}=awN6BET@kGb{)P_ycCoJ&UJvch
zTbV&*bKO+Ue;O`s>;s7`hf?h#$nOWwH6GequCf%K5Tj)aM1!=;KLa?0V%2wFm>9Wf
z=j<pB4!7<VLFN5ujSWt66ryIX%JbF*11z;rBU;Y%%?VXE@E=;({pt&7_M2vxE(<&d
zrL*!-{2qmv)vEQ-Hr7;&_8dLRBK@|`6wvi=(Qy7MmV=Z2UFCKrel3=@TK^2|VU?3>
z1jkYtl&rTet*zo^V+W@AuOTTe@o)nHKFsF&iH^`*9zkj$ATLIz$Y2o`!#K{mG&?G>
zmORh|H*s!1JY9+H`4J(_EH|v?XK^OIvlKaPf-NqYdg(PGokLYJpDCZyF=3tYPYX6?
z_vCcVW}$xi#W7UFNFs9%t{Uu4uD#OIpS5t_sl^O=O%84@VqU-N^BCM!G8O)mH@I-R
z*~8vHY)_CYsmLi+ZfH>I)CY>5nh2_)UbU8T6u1;Rv!NaYpiEuLBIGMFE{S4`90tX4
z{LN*tm0~+KcIe$GP0ic%q=_F~#|CG)^g6_7KD-_f9l4pwF!gi(A~9;1k4?N(N|*s`
zja(TD1bR<c4`2AWbXb*fP5bldAAe{^9+-}o;UEIVJoC@1M$6n8E$i%!+EzR=+h1yc
z`|J02Y5m3R>^37wtN0Ij^E-fR8};DYf`l^h<v5pl%jAZIXrL}GFGsPE@lpTjoYrqK
z`d#(@)-2B)|B!VyrXe<0_AwnZFs};ZZWQJ6mYA<{`NlmoBMD!IG9s1E=&}&#EAhn?
zUFgU#YyL44KDy(sb;yPIAfZmFK3b;xmOJgginv#<56Tm*$=w&q+RGD8q*D{+7d@%L
zEC0+B_mGfR25jN}?7cHo)C6<aizRdVP+mEy*csr7u0eC7KAWpinUJp5J30hZMVFpG
z%AkRhM;p)SV1OdP9T>69Y*NXm9DvC6y(t3}@PFraPUtsZ2F06@=VRVQ)iDB2J<52+
zq!S~+1&%>1F;48?yur%gW|a#g+_(G^(5(jm^<+`UTn^+I%&xeo$EL|0Q>Zs(&1NWL
z&DKSN{G13`?Yt+;jQ$fWPaDGFP(5|pK>Ya-JYM%CO)4c&g>xwEm;hq9B#;nz^(Wz2
zx@i@nAWF50*vE+oEZ66MCa>x%#RW!>flWya`E{#}#td+Uw)$@Nr7a-Ypl_F^T&y;2
zNVbK3+95I-#vHu#@(E2~t4=>|Bw=e%b2}qBca#SNZhGVZP-Cim)VL<TGnC|d9rCY%
z-9dlpU#E}(M*}t=SqE3m{sOX8<wJ6KNM7qu8$qWwce^0iKd81oD13_A!O0-SzR5V;
z0&|>$UpF%EJ)=qNH>BIcXRD5`$l^y6oV~~%$kXacBVWkXyL6^4u3dq;m7)Eqv{$|}
zJ}Hj>sXqS)Kn)|?_sZ<Q83he$Oxo=Kj|H9j9!!5EbvG6Q47`^8yFO#Le&^9_@?Pe?
zeTE<n{*X*G(u}U-9SNs$;OPFr51ie6lKrO6%Ki&Eh6f%eHV+9teO5Bx3XNCK&W$uV
z4pE8CcYd9Gdd=dgRV8Q@u;yvWem*)h4^y(}c;L{cbg&*hAB+qz&O!f-5|uTS_vlr>
z61%Fd)gFSL<N~0$iybjk4|i|*@}E>xsQB+^-z@QUcK-O}Rayj1I-IG6g@bAD_9fVZ
zXHuz6FsaN~dA-ZNTHlRPvB;DfkG(f^e!BBqZt$9%g4f6>POZ%3Kvq*f=Wf!;g_)Yl
zr`OPAjQ;Hqc0*<Z#>{y8vNxG0G7iVsfHoMh7H%o21yE<|A`c7Ct2!)Z*PkPyP&(bZ
z@3kpNr$^rm-l{;DP>KtUyMq23SM9{_JkONX))JmU;>0@bH5T_7;Dqg*rS^pTkUY_N
zyfDA!CHQ?EL($gpOvXYFw4;$*I#8&JhG`uvZayfIbsJza5}m?cH$YTN>66QFikw!s
zTRS?w@-;H48B3El^@(ui)Z)~k#Z=K8#5^g)rSnYYhDM~$#oG3&H0Kdd(n?!`#x=CO
zx*j=ha@#Rnq-Vmssdz;!_w(v}eH+s$?Y~ePuOL|LBTCZEwti91P6jx6x^oq;{56^T
zJF-mcw8r1}E2-djG#{nuK3MdL6g4|O&|nFbQSwsn_hH@nSM*_jx(o)S()Dgr7~?Qt
zDo{p<lFs{)uAk{FhpM5sKbCdcI91@_w~|-`Ah}O0XA`Lh>=JHc;(gpANMUEry_tM~
zJs>n61mIdZy&7H!r*4c`_$*)3yZ9PB_{zZt`gDq!Et`;x-v$Uc%<`ruLMafB=E+ma
zC3cH@OWP{K^?X~3g&U{^)v_E7EUY=A&y}w`YfQKYzeIh7q=8#d)oMhbw=~fQF@E2u
zV5J_n5NCK`nCAxa%_${wNZ8csOAKLfj4(e&^DN6})e~{RH4ve#wbf3tMv*8?G~~+&
z?PChFGNvz=?@-WoR?;Q#@bW3AWIM`qCTDf#wVGb~_e<Eh*%DE@lh-67UJe)6QlrKz
z;Xu-MF($AdAZD2<K^iVk9y2OW(@<;fuEUKb59zgbQJPx*2G1nq39>FLYIWbRO|MQS
zRf;oWd{0XPdE^)zmQ=j`4g#7z=yZTFoLZVb-{%|dL|DQ;pcbIBjD12u)%hKRg?>A&
zQC&+0MNK<@CXZu2m=a*ZBT}Z`LbPkcEXT3R@aH;JD{7=9=Y!Iom4yE9Q~9U{^SO1?
z5<-jYs2TSl*vwdH$dxLB07#;Fk@_lgYSH6|d~c6%c6_$8uVYmSh72xx400S26S36=
z*0Dq}hM^PS!Ai9fa4s;k2mE+ibxqO_RV7r&XVa!c(5vJaYz#aK+C7_)l^|Uc8xcc7
z!Q~c<O@kRVd&Tt>Lnvr)#kZsS2d*;R6iZ&6*$#=I6pCO3rEU_`AVimeRj~#%DI&i2
z=1?Ao7Op+oFV*N~|B7YvHkQHA7^OrXD6JH##{M0Kv}<PH@SLa(%syxZv6x}+;{%+j
zG0<Irm*|}gfoeN;8{)?Hr8rw1n++S2zHnJVAGkFDHfUThW6^`yNgvl9y<xn20}Tem
zcU`Y;=Uj6K#?s#eS|)J8an97rF4=h+gi!n?TRl&AeHLZ_olqH5Nbsj{@TFV>ZT_Q}
zR<Ty~@!1Kj_r~m+#6+@VK!!M7cFK2@>#YXr8j`wf6Lbqw{MR8<?sa0C5rS;9z=!ah
z+TPoe@4Y9ypnRx|&heOM7mSClFcBJ8e}WuzlaYdFR!>Z&l9(rCpUN5{Kug2VceI%d
z1nnY(u$!C}?P=xh;Gaa4(aiW;778v|V<$E{UvUeUq*!A9;`;B-Q#nMqShi=yk}~ug
zsQHi-r+Xfo7mtwX;(D+xo@cd30_xyY#BuJGQOQqpZYe=ZGF;QNbagFuNz5+Xr^l#;
z`(3r)T1Wi92uLS7S{zMq=-tg9-!0YwUT#t2jC@foX?ic$32bWn$$w%3g7=4;L06eA
z>wI>JO`&avklP!InTx@mRvS`NAeXQ7yb3sKD9xJDzK)fc$j(Wlr+*F&JiWH=Q2OY)
z+~}gS5vsGf>VtlP_B~+k{HM~y&h)>f#Q(QinDM_fX$@+O+I+)OJD*ffprsQA_n=r1
zT3BtsTL-V{MyN+7%#l*R_zPO-<27^+4v5E#Mhq^i)mXB9q`f@&v~}d3RKpVVRbavA
z&z~e`pk?_Xiv=qwNFnBd($)OM{8Jf$MaM$O4!_BHgMFIsRghhQYH2?)c<+@%L%0L`
zs&hHf_B2ZO@g(c@ikfqt@_!URL|a}_qww0_O!sG3cW1g8Jbs=%)5$gpiZw77OSBU!
z^jUh9zq8xty>Bz|?WTwCOV}oL!^@G{+4(l^9Zdlkqm;$BNgzpxoX7@DTivfME^sv1
z@nD8q2ys0+JPzUZ)g6%mS@@vN2F9|QVR|h&E=LW^9lpHCzPz1TpPZO<enx<HN%(Ox
z>a*%>$}dpJ8@&yPo{XN;<?Z>W`G2H~Mr${O?@T8dmhpx2oMuIdnJmQf_k0U$Mm16k
zjZ$0!YmM*E4~BSRI$3hcPZ(4s;zI-MUry0lDj=o-1^6n@c0Ch76W8dbJe1wgNgOf8
zx)r{PfEAXQX-k)1y$z_3qkqRVf@QI?R2I{P`QDvJNsl^;Zh2hE2X3*oz2(Wf5gZI|
z<@j|>-cnD-G=A-K;{mohW0&nbu<c356)+w^V%hFRlwpemwi*#c6EKUU4d#!O+5Xu7
z)r4#|FP1`Mf$bhF11rjHiI^_w{ocz69$4mqkHY%>g4;4btp$#rOhnA49A8=$Q?<rQ
zMZQ&wU{{E!I8Xj8w=H0TLlWI(xfrSr83oVz8}G95lY>d@(i?=<<<KGRXBmB>ltZul
zZx+`Yf&qxj+d>B_fM89cFgx{$+Z!<l#bMvL;;$`TYPoGddOWaIQJZM=a{xVNIUbb{
zqCTDg21XwGFJg<tT3b|oq~{JCzfxbl%>lW-ZAk-LttxB|Q$B<=jcb4&^?m?OVkhVk
zAfawkfWZt-L)w=e2cA~QcnKY<E4;tYo&pgm;V=MRm(Wd(+|~_1a9QpsmlCufZRQrj
z#JQS20?!HCt+3LF@4l+~Tthp;yR#ur5R8Lc;!an}da3>6iUs*_@$lm?iA5TR^NPF*
zzP>AhhU<t*HCl;L<x;2$+a}E+P}A(jY<eacbkAR(4fxJ5M~<(XMv~y*{(*XMDYmvU
zaZBCbldP3G&TEYQ=@QCTXaGznZRC{aIqIOEq8;NE@DjAEj85#LeCBUQgXXGC1m|e$
zg3)Q2=VC)sR6;_-qOxw3D~lFh^SvCe1acZSQW*Q&NTrZ;hErzwH+NEmhye73TxHoQ
zEt)u|J+-$v#G&vj`!96rh3EXU`faNB`=tl(p73-vp&_N9|72BumF$Ctmm$y&1Er}=
zMsQ=U-26zLQgroRb8ghO?xAJL7^Bik!i_f)MT*x>>SW^YV7;aozx?B6m_heVXfM%!
z_;9}X10LkK9cT%Q!i`r%t7^IupdH&ym8KL6#KNacbN_HP8VW7kayD~xSZCq2r#_X>
zNMz>?5!y85IArovz|3B7H{HFtkKhrxq2OZ{Afh(ZeW~#Z4V4w-E1A>f1<6LS@}m{W
zkK(Se$Fzzk2^V4H_Zt1fUZ|JGE#974&KG|zXOxC#v;t~St09Yf)?`l3q0YUd(+Pp}
zYVn=}9{k*@8e;&w4J&tLouy64SVw$EaYRyM9puu7^lZJJ7D#*h_MJaPkL*nn?desX
zJouAA`aC`5Sk2%8=Pb>9KOcz(TP-VGDwoz@VIkF*4JIKaBsQzgm8aM}qkH1K%JTS%
zqA8N6%xE6_{<V^tk466aIi?fb?~#xw&clY}H5Mh?XY&trHGW0KSA%sYJ$>M_gtoTH
zv;L&lz7E`&8H+sF<hCU5-FtE(?q~zoaM7E~@BnUM;aaf1Z(LzWP$aE;;rKWJmiHRa
zn6t*^omiv>a{CczQQ&j2W}m+g51n`@BhIT0!_pfVC8p?(IV7#<QW#Dhw6)Q6HqJ>6
zo^dI(M6fmhT?#9IAyg$t!PX}vyZ}{ZYn28U7@39Ea2QZfKS4zB5=c$lK&~A|d_L8f
ztBg02m(laS=!hR^?n9&-wbt@&g>_JlXls3d9cc55mU2()Pj4@5wK|<M`(0t|yzYRC
zwW_rE;<YBIbLXC#peRznyKLjMy1Fn^dJfo%K2NJ7)jCnO$g9{vBuXbapTnslQGkJ!
zAlHaFgeQGdGALi464vd6ZMunF-snjx7fzV>upDx+J|Hvv2T)#<*H$6je>mBgsy=Qu
zvCaGkx%{|Zwp(bB+$Rn*oH5FJ|EK(6wFD;~>TOl4>9CdqDFu<-9Bxd%dAbO{>TeI^
z>~0%oB3_jyY!H*sTCS18Wh4%aF&~0=1?KUsKlB>gf??GOl}eTy0_|)ID5nw2>qfi0
z)=g9UjZIAG*W{-@B{Oi@wuSkL8}6hP#fxKuqIpj#)c=A@hutlSnAug`Xa0Ry;JJ-T
z1OBUY3T$yjT8J~KLT4w1_c%F_AdeHrkqfYy%biE$DT&4nu%o3Dk!_lj5_6B$90l(x
zQ4^C_V%#+=N5&d2f!M9WDeY|6KvE&dFZK}N%XzNG?>c4)Gj0c3V}d(@pWYai!M9;5
zF<hc@n;LuD$d8d>?Y~EE3DJjNEB`v3viESxS091D&`&j=WgqjSE29Ho<zg2Lc~?Da
z#t$Ic2#2s$TL$9VNF+~(6p^^N`O^BgnaD~h=a!@z!OExS>~k|OMMJo>?6llfJM)=A
zrD4u;XX;v>iPUet0a_~Q;61;;QTu4A(iKfA`xp4)nV^|M`#tM8n^W{~G9GeM6oo3X
z6k{3I+UURH<@s*Nmn{GP)RA!f2YfpR`+s*?Ui`L=um2CtRq+He`KLetKQr!hMW(Zf
zhvoO5k;Nq=$I0@~mr5co5kj)b$TR&wZ@h@=9d!U&U}2L#^NNq3a65}HBkHmP4uYt?
z`8%iuadJDyeMaFJG?96x)BF7MxrCMLli<a$k?kAp>I>a9x4SB^gDdd9+YS<9YWtJ7
z-x1+)8gH53&<!i=8P3aFa{t|HL6#Cu8u>4D29j^r$74Z6&BGcQOF=+$=A8=TB+0JA
z{6Q((dVF1P#Hu>hc!!n5ttZjX<U}=QfUBAiR$_I?aRID1fcXVr?ecp3s&HP8XoaVb
zkCG^t$GK5BRz1m!baxq1Acl^67kXu`p?j2t-$+)AKR?xxo5G|(a#q-M=sQhYy_~mM
zmPAcOxT!)bXF{*;t__}Y88N!`5=}U+p1p0k+g3Kp2Q(Z!V#u8zi-nJ_{%vv5Rc|ux
z+U0emjjm_EoTG4Ha2%<?7gWZt3@c>&^~?LON3($=d9qjBm(F(9r*~TIj0$Ae$8vMA
zQ$HB>u_3gq=ziNjUsv61pA@Cxbw1jhOwu%(9SnsN0$b1PA{xDOlN^JtGae2OE0?XZ
zol>eWhf-vB?P(sDUrWsst=`7va$;QwTYoAH6Z3kD?@^nK-pmToAq%pOjL53>7NnpA
z8YY~F$lVz-!;xiFZ6*`V1l8~3(K;U&R=70km$@_P&^&#sf+x4)UCzwAlSpofHuQ>K
zLCOmZ{+3rrt!&W;fV%khr6P>~40ffij!eJNMHtv8o^aF|Q-ftt3lK$gT1A%X*}s-I
z1gyohl{Gd9t0>f3@!bxPgfP<l9<JzT79wi=gAauHi!hX}H*z1G5v3E(c#sq$IClM)
zZcTas)cx#W97@oDV^IZeUhhUpkWNuk5w_DX_J%p>w{19UYGiar8ncCYeJ*3aL4{<0
zba6*IuG<RoXu`jTX{(dg2$ipn5@NYyJ7MO0L%}XNLu4}|H%Qg-3j2xHZPl>mrWmOJ
zoq8=7DgDliv3f!uGlyw9w@L3z{gw=RXIP~lQ;|U_c_9BpE;W-AZuQDiGk&Yt6I~n7
z_oWMr3IF}cBZ}1QhPwy9bVyG#>%nweSO7Xsr8W=36d5ejd!WYY%!A!*EQ$(ouLAY5
z+|wK#DGxS6ZJ1}|fLF21*0CMbf|3D-aSv>aH!Gqc{a_npX2Urta|XdwukYt6k5^Tc
zrq<b}YR|<rkpGD@3j099LaORdvFFp;WVJ*4v!*aq#>o~)vK0p<J8`)Pxn+n^P5>q8
zUP)F)kukLu0Ns{Tw#On)OQoycG&-Wzl*bzQb6f&(vcxXGR|gkl<%R0eFz?%;D#8Q>
z_ExephH%8~)Bf8`UNqWFs!gx=jU&M`!%`rF*GtC<?L6OMikCQ*PC*n!>jCjMR;yO>
zh1BnbenDz=@7rqyk~ULkFvV<pb*;?v8@EOb8py~UCUX+2Uw^2mEG2>QQmO$t&KWEx
zOL?=@+aBI(GfhkO=0rbm8XyP%EDiaj1a5etC#nWG-ErowaxY!DE-YVw5xq+{stGN_
z{8XbyT#w~1d}o`&axJNDCRSMwO(aqMIC~S=kGm04E(U1Vfl6B&{2z?HV~p%!*Y4Tf
zyKURHZQHhO+qP}nwryLxZCiW#J!d8}$$2JuK2%bv5C5bpskPR9-PihEQRiy+wb;JW
zD5KTx;#*B@;Qpyp<mu^6r0=%;>!?AH<W7}(Fv|7CC<rWy=ao^wj~YwIpcMo3xcX6v
zCA(I;<{{LyPBe%?9eI7G+=N`95F5$Vs|aqM6!pw)PDEKBK=^DO-#dvJ;FO}6hhf66
z%fYmZ1M<$`=c<lH$EAm(Ct(E*O>j77dR<hl!OKbrtF4@89g$KsNm@{!7qVVe#LL(r
zMAMbThbhNLSO3ZrqZuFrb~~gKwQp*ZfCt>DlcD_w&+x)K0uTN$sLA&Rb|^OGVLFvq
z#tmo-xb$y>A)9tDN0y{<uVPN53I#GazLXALt&?TC9RXCt*h;?zNC1S6%VPL75Cia_
zfr0Z;L-3D@|GguV28s{Dv3R`sh2a3=#;72l1pw;GSKqE{(f7O+mNoOmUJAJQ;epbS
zp#m1`vppGLj=&=qW+`+Fw2qs=k=aN!>fBZ>zuG;VFAd${0e0N@5&!Pn#<|5cy4{>%
zt(1nryMT1f-&eCMT{->+I8(Q^c|`vvv`tHQ_RSc<qj1J2)Am~U#4VxilQ}pIoktmF
zttiR!Q_Bp~SnNttc2YVW>&p+=8=7V^J<(y?9ds&AI(%F#fY;KUs*V|O$5+rf8Vp}D
zXOml=nz^c4FPsoj7U~ZzWA>`nMO=mkQRj1@0XuC?K7Uf1;T^Ly`_z2a!=e6!3Q^~0
zyC~K}oibU{>2f>BnN#|>Qw&L(A1HF>OsVhF&4jQNqA~5bj1MLl62{=*RDZJvzx%S`
zeLULFf55foBvhHKO5r4UytOnrQ-z`9-s@s7VSSy2L{7!Bx^J{Jp%a*pi^5-jQq^oR
za!4x3{fYKs$dbd{B-PI0lJX%fLp(KuzDy|^_B)GB;`vYQ`9D!z7&-nsoa)N&rpFO0
z^siv>1}2(7Zutik##v1cWSl0b7lZ>OBl{%&>mR{#uR3j4>a1h=Oe*<JTyYegAGP=U
z{rU6vd05l%2z^y>!14DM;TdR|+rW9DY7|1C1x4!X?G#-e2bfHqnkT`j;^Nf2n^fM#
zdlGW0D!2$;C)MyHS<v8Z0Sn@RZp9(4MAg1<ULddDm&@jD`$;a)0+O@b_^IBWpXWiu
z#vq1xX|q=IuV?^i=3M%q41bnZ_?RqB_b+Ewq%f+qn*JyDAITD(_p{YO26S>uy?y9Y
zqvFId-GQOc^JQGI097fbbL!g?QQolpKDf-m*|tQ72^@g2m0@|3J7FYxFyRVi)=h?U
zUhRDo+G-YgHErRiOZYicW&Up^iimvpZXwS&sbCTvn_zM;LyrbbIIWi)`%bv%pDKzD
z3UBtUky|-CkUCnliel$6B-WH&UXG+hrPRnY1QajxghHv%Tb$s`?sW8qiB2k}aapv?
zH+Dh~t!iE1l3SDJB-8rF*1I@z%j{+fk!|;~tVS^}oA*ej;kPUe(V1|`qv7W=ZO5ok
z)O_gjW2fSR*{n8tfCH%nt`XN{&M{$4GAjkj<gkHCVGXR7>V#cgi}B{HvG0x<Xy(<+
z6Sdi15J`UxG)rJxM@|h;N1Jz`lCyG`!5UyVNTAh`i3FET+-7BOrm(KO@aDb7Z{wg?
zX5kPK-Yj#jw3cOMalYtHY&+yl0ge%vS41)atK$YGi*2lpNjRb+`1|f|@zOgzq(CmJ
zuCm>lP#he06I54vg{ZhO%>ewYKk90CL|mg?wjWzrbKZQ@RP$8k%5%cB+2peL^Td76
zI%Zyx&eBd!mQDe9H@V`?LK(UWUDe?|feN|mjZTXtxd?t}@%q0G11Jm0K2G}qiX@+6
zEr{SI!OcPwg_|f0@RJ|PyTF~6H`=^+{9B=((<nG2lSpv}B$#aYVbP*|&$P&H)+@LI
z5hNteX2M@~vv6b?37+xu$N93|#nKpplpQVpxljY205i5G*i9__umAgG@RR+l(p%*$
zgA%Yb=vUdq&%&DIp9d8p&hXb|Rh;v2*)d*-8O%slOhB<va@l9&GMB_Pd0$a8Jlnb?
zE&Ad9OXI;GjxrU2Ji-#&#>J>1T9TfhL9wgz|2ZH4H@mw{y*8W8xwdQ~13L$u511XG
zq*!r2b}R^1ZW(IAE<7l9KyZ&fWAujTs{B==w;*7+v3T}Fd_Z~WpKiXxGU-QMRb`2q
zHwwY7BL<+=H~?zPoolR%F>S}!u!WZixw?B-C3LZrD@i|ONvkHXL~<?1MOm#Szq%W^
zH!OkZeVZp*h`h+zoMZ9vOey`fadagxMeY}^@6^bGIH}ED35rw*nqJRW2A)%;*#O{V
zrC}ns!ceGbm7;5YeiS5wh~louwRPHo6h_J75CAzICVRBtm|`km?kG>oBq)&BVGd&l
zs~1@9oT%Vls}Ausj3x+sEkkt}$>O{W+CEE9Z6RI-7@|puv1miKENL1>qS4X>6e(|7
zek{id1<#)6NU_*lh|F7>hANW9ss5@7S}=~5^51Iz7=Au}IhLoDI|3#cg*v{UKfhUf
zB=|PiCBN<+-JcJi_m`(f@fIAmUlhfv<TH?Y>9~D)0!DwUZ!C<&R`fjCv=#P`rcj62
z)BGQ$ZLt|5y$h{*cei7!XX?90xn&}w1DmE-z<ud>=OI|V$zX~z7;bTW+3Msb1P*){
zq-2ZHM>jw^7wt04_jtqnqe~@h66R~4+7S;_E7P|g4hWzS8>Yz}Xpsxp;Bc0ap-lmk
zbO+a?znpVitAvWt;F5|%5=oq&ia3Ca9>aJWIh;QCZ!)RS@pZAof~VvrE03gfsn|*p
z4t_2!x5clWy3yO(aMjwkh2G@;m;o)51jiR`zN3IZu=oOw0xL&JV9flfhNEKFnM*t?
z1ko`3-JxPduko*xV$YYGt6--UyYP+sjeds3j@JvblULLcS4N1>p2@{)>BZP4itW&n
z*$#OQLs^m$?OhKnfN6(m{!*NRR%(OyHK~ano`8k`yWae7s&g}Hm;msSOcKjsLZwVg
z1L;`yQZ9l10l38;`F^;keN#v1V93o?I5HwaHNJ4=W=W6Lk+}y0t~Z#=&vUocDTfmM
zr{z;H6^ON`m*QVH#r%4&Pdk+fz0TUi23d&hR{K`Tfzm-Z6m}^08X?~oMDG1arHqRt
zD@x}ycdI%U9#e`%Zk-QTqe*melvl>Cr(M=8!zU;s8kDc0O6UGA!+2>VY2Hj4_yav!
z_d2dr#)CDid8e|1IjM!o?tVF$O_Fl<OKjw?@<D94d3is%m*BkIqMkOE(^l0+8LL%F
zrK=lrV_Fuohj4)ya_5`6UgSLjXSK`X7R)-bf&2^W%Qn5c9}n%tGXa?ruj9o#O52r1
zb&ebeKc$TFyW)!@r=SzIXN+v4EZ@zg4jRWs+<haseLahj$;)uR*&g3sQ;%dbJ|2)0
zumtiYeUvX`$7~GzBo)Wcb$qoHVJ;(uh&3LVzLVU?)moF+PC2Wh8lvl1nipB^tFao@
zL9Ik{3A>EHDWb8qoO&M9Lr`+`jjpS4icKLhA5O`UC5~^0$sKu2%+oCYEYkP$@#sl!
zglXo$ZdfbB-}%q?fdPVvftj)XSkGzT&gNb;4J<QyPkyn+RVYDaf1huL3KA7<cV3Vc
z8;F;FlYKx4EYQvS+6yt{I=YGdk`4P)2l?az#9Wbi2EYuJ<lMnBK^tuOOTm=i^I4pd
zg*{DMPhv#r_lpEzSKjIxOgnz>=1sKd@8}AP^Q$Q@$G`mo69p5OsJ_<Z75xicHNL?V
z6v2(jpTmldzA@$jSyo;GFzU(Ff0E2wM<XsZgv+D3qDB%9-%CSr<L%SbZ{9l#^6DNf
zCsFgo0@>_On0JOS%szKIkW;6-4-TH_&6%^$!8%2xZti6h4vurjj4I4*-?s-TVWhis
zi>HvnMtN-W;KXau<>eMF9zOp11MBCDh@BibwcRIK?(_jVRG^mqrwNMw|08hx|I=IO
zng3sU3+sQIWust+%?9f`SBIWY*8^rnte8A5gn(~D+k>m!AMVt@1LC{BzQxM3X0@J0
z`2G5-(DksX{@M9ri%k|`Gz%?ONc?apW&x>%eIUQS$e*p9A0Xu#5N3S<R9oJOFNES$
z03A6-GR9skexMkryvc*982JE=C^WfpYa(t*I}zVaMDatb@^Eqy&;H&8aeH#*IT1>{
zsOehKnS2*IN_reXi1Z;2RMG5Zl8@Zerc6+PLI@p<7#iw@j@R%i5Sf5_8)T8cty#(*
ze{_cUBsoL-w_-GsHyn)dg^e5zd*>utfQi$!0y=v`>o0pfvI~U);yv>xvjOKfZYlhZ
zC!Tp^%GXZxL@O(-=#=V~5{7<UT+jl5SzGR9e2FJj`!q`jNOgSUE0sEh${NyH#z$8M
zSVT=vHrU@*T=~@P(F7>e{20V3BGeupHb%!LNXm+t_;#!2vc>5;Hg5&VdtTs4l*@#b
zvRczTssKkDDxLdKmc4S9@`NlO&gJ169xTch7b=JH8=!c^uc)sbK_Qw&V%&?twfa=d
ziO+tk%*eQ!egupAX4nfQXbG^Cm*$835}pxxl%0+e2t_R-YHre+^8yr?AtqAk6~>Qw
zMs;JqNRp=I6A&ds2`w*9b=xX=2f4<%D8V}AHUKGuRaJQu1%!>Rp=^F(wTXIG3PRj&
zEQbaSt6C}W8bX?M$1FDzv?{l{AI%6}sz~=pmaf2#T2>INgf--8Avzjl?w9t(Cc9pY
zy)Oq|9KGzEnUkTn>#GA}&->LOKHL~TJU{flArMV2Og*`{Q#&LDcxi)JOyH!v;6MU*
zct=14w<l}vuRGftn43Pn&)c_y%3q4&=f;LR`_JPu-0zF-@%{Z+Z|wf`^-+(v8)~d+
zZ&Z)(%eB{y@!RKXEDR5Bu&(T|YiXm)L2!vNBKfTNCwXdEZBX`1m_Fv$SxQ{tm@E6{
z-rdcPJ4d&hH*<fX_Q@Z_^Ht{O@$-f8vx9yDgJp;3q^aG^;1n}}gFOWgP7fpo4U8uI
z1)NfJKjOa%H+E}gXL>mG_cFM@Iow^qpG1j4nz?#?6`V!@o?TfV7@j{640#2ya6S_z
z1ggngau7HVF@NTT^sOC%9$*q3c$_*JXfVMMT6^_WK#68#SsLL|*}!pU7c`obTrey#
z6!ITw?ZBU0aUt1DR)A{4uO^Ln@d#ru*8W#0ekca4<xf=D3GQR#Hwhg#a7R3PP-MVf
z@&ijSrhsBJ^a$Dkl)_IC5a~9vNq{q5ronyokUOvmLF{Ok<7B!$!X3&@U!I@Ky?+ke
z7KfGMUS@c*a&b5FmGeNVRlt85t@AaM{IzWJDS|VEFE;p@M&MnLWZ+SPr%59-f#@Ae
z*rWGmWX;gRPLmGL`)m24!IIv+Jsnu*4<79BPWSL;$sW(+QQ357W>3xzP`F5~taG{r
z7sc<!MnE0_+2d$vx3l(1eeb{C-R&8;c(Z5sV!`P2V(d!sPQ})|kM$oOkwJ52%FG_6
zPU&X|--eH`hDPq690jbtR^yhEUX$$1@Uq}gh)XT+qJkzscAaZR=Cf{fXvbA9e>S#6
z>%7rpbOhKv%`*Yo1T^cbm<Lqs`Fw7e9k{vv6`4TL_w@Mezu9OQJ}Qtl#%Gt%ik%Z^
zAo67>r1pGYM!@^tGM9YvkZ*Vv>#4OvLJM3FYzi}#={HJbQ54ABg~KotZ8m>V96X4U
z?}9JV93)!P&<+UW!KX=AwHLQXOG)bF)7qmsq7_3Or{ee8EmmPpPZ9Dr`8#t`RLXjB
za8Y;?qXxeKc~K<AU&C7{E_8LO$WE*_;RX@m?+i@Is|-*LM7fRy0SndXA4}l^58?^G
zIe5hD?}wh9af5rqwVoX!7FhFpOsENqfBB|{Ar{O8tW}f&m~q|#&0^vBQubRxA02lF
zVFvq`LvW8-7RejyCd&U26F&Kn4*F0NxyFX#{!<Y4am=~%_#Mkv=Ec^6N-wSa_sIiE
zCONpmm%~5j5L3U27r6eBrT+wWbbBCO3hhF5c~Xj>`nnP*EDRupx9>JjXO@nXe27j9
z+}<J7q%{a#9E2Sz%y6Gpf0*qB6V&lUUaD6Qo;!S|6ca$7iuKY&($i51_NT86F0@`~
z{r;iz?dJl~Et~~l`ZkVhD!(VijrW>U>H(7*oCg@UxOShwZ84Qg_9zyq<50L(XP)-y
zLYfmNxfvxj+>m2I&dr*EKFd09tHCc%yF0I&D~C2=*!5%;N1`zm{t;55Usb3<j6X5D
zvwgkNaJt2x<9;dAo|sPBz4SWLh!t8vEghmZ{kiEDAUh|QZnRunZo~CJxsX5;2YmJz
z4D84RupQv7m8o=#vKW!ZBAPrYail#-M)To=x-11nKiZeqISr*;mYN~myLz~qVFDKN
zPpml7`y?ppkwouSDDY%>VSb@-P-o?64mtrkhaN@(2dLSs+lD-|=@HRajs80SPIepQ
zJ*25ERc-!0xu_gfzg0_*PmHsZmAHG-;(n;=9(&Bd+&$|&b>N=fEUr~a%eLO3kfUEu
z>Ssxr)A@P#le6x}jqG_BR|!(cVuawiaDj87Jf{MA=eMt%7}Lbb)(OtEFm?E@;8)^6
z7Q4n@WEq+8rJMe0tcKynuIO=7$(BgcJlqkx9gI{OnKOCUQ*L=`79eq@00PzJ$ZP1y
zR95PQtD6%1KWf4bisg{$Cfl}2BS9&P&ph7vXGlVfN$qWRDz1tT>f#LVWP}0V&T<Mi
zmbN7EJFcB*2^7`JRle4PcVDCFtt3eH&)`Re>GcodT1R7EzBd_y)W;4lukyLTlxnO;
zl>s^P+Xh2r84wRc$VYj___44+z4upnuo_&Np4HEHLDb2{<$vhGR%bnv@Sf!%+EAeL
zhQp!RqeuF{6Xtt|Cv`>I6%Ywmyf{M4PBMRv;LkD#|MA`Qs5Jcrw#lF1N8Ov;<WFh=
zIZYG9ZmvZDO9_JydlA@8WKW(&vowQm*;|;!ckY4-81&ClSx`k!FVDkyOtbTC$y0uP
zslXw((Q64f_#luiiAUXYq)!t_7L;!DRu3vEpC(U%-&7bmfR$~bmqZ5Z*phE=Wy<NY
z=AQsJJVH^TGz7wxx2b~8t2~UTo~=V6NlGDW+D^Bjb}PX#eH9$wb3dEt`H<d&8^I}P
zwpOIvnb*WME%MX&>_1kWsRqn`iQ*7xcs8>08gFRxyKxV3-<-m~Hi>^>9QUj+)*0)J
z)j1!By(l0dc&tt>(nV+tRF8;zKHf--cfsN$=$Z<ZlN$Hu8HoCn9hWQ;e@}qGpv;eV
z7a1RBR`xX(ByC|uKM$5h){npVlXHN(HI8reBl5n&+DH&E{9){$h2)>5;hJ8?L;(9e
z^_y%B7-;4Sl*{^aap0%v9xC@AMopIgfMxm3KKk$O(tdLwY_WfHATFssz=WGnZTs~B
zEQ%50wV-zbvjf@5@DBdR+~dmTwGdsB;AGx5Olqa4tMEHK1mWu5;OgD>W=DXPbpwVC
zUA6%p1*Q54O#Q0(f8<m3-M@Li9Of}=xZ@t`SA`$g-#4bVux}x{uK<YRyg}G^xkBAV
zU*J!gj2|%)b=(u##y<&ar{42xVv&P!vk6lUf6YBR=Y-X&qgD?=s>aoA53K@+cS60j
ze(xWIn;z;buo4(ppqJ_q+$vl|ix>&Y;}%4SjbF3Aq$6og^Ik8FS1DrE$Qve-5tZvv
zho2ORIn_{TIW9ZNZHi<y=r)g|N~tQ~WOe{wz+*7~yr?|eV-JToJ|MJ@aAHGYM^xi&
zOoG|+AmPLr;;qX9V@{&#(r(&L*Sh>}J=@atlZ|kHK77&m=+RAd(BV}8aDBgcyt_Vo
z$~23Ly~^O_L|0ZZnc#q`qVI!N%Ae%irE|v`f6`wum7Yxg&i_ataB3xtTMGiKqIwbR
zQ}8zRSAoeV0YPlM2c!#W^C%R(CifsZr3m@jlh&BqYd*?HJ;$t<+cpX`TeJvR0z4Mn
zt&J!Wn8-(DAmd4N8vXFxSR$klf?487c;Q{1w+});jjN2cUO?3>fQ73-X5}%EW)Z`g
zusCeY@Ib4A*m&9mxITt4Q^EJSmbELRzi6P*E~@N_coYuPIPvzLU<5@@Xe&I{&gn5j
z5^r>VfiMtne1G-e>%jz&1N5;Qa8b~Xuip&oP9mCx(VKnqGTU&7M*)3RH?!o>O32}6
zfo#c1nE+K}tiGbUQ>74)L(k`Z;1xI4tL9#V4WhBC?(b5!R9?=a5g?P+$zQ#p3Z7Tk
z23Idr3WVUS$O#fqdr*x$B*}wc;`>uoXZ0%6a#5yKW=|}P{{5gl$!-MTS<9hB1uOve
za|0kF%rN}S-0z@2VzDfQ3_3mRS#e-2hb&IwrLx>!napR3q|wcnmZssTC4ynjxUfF9
zl1d0EnookXHDI*Xbj{vP^`jytmQ%hn7?9ALF(@V{C#DXp&Z-{`6r0Ql;n`bp76*b|
zD=4f&xlI2IQFaZe)}foA(m9V`K^8TbPI_v&mV4${_VHGzzX=jRDwb3#o5hOgFu_75
z7YOEI!0pmW?0Z4M<znb6{oDPi5+!~bJY?y(LliPpPS>^VW^>Wi)!D?ac_h1;gS@Ml
z6+wbtK(?&i;>?3hI$z8V`=mXyeU*t6l{piRNG<<4T+K(UfVN;6qa1`lBsbFmzP3OD
z0xvS}u?Z+FQwrIP*aFYdAIeSrJ=Z2(Y<Rc0&E~~ubsYY&B$0_|%%E9!J_f`-yvQ?R
zjAZ?>Rq62l-T~OH^x}|nnknKs*fO8ZKXlZU+3RZgylIeVhBeb#bn(Rv@h^;Y#=Tau
zg`A3szwIue-)k^5r+oM@+pkdNOV2AwIP(?Vs+F5pHBEU6t;)-PzL0S-ixXotsma|3
z5;g7Hs>zgu-r7s{<IA8EUozv~w;&3Kh)sk#^&sp}U3shw<ZNf2M3hN>?mW~#ho7?7
zp)I>-0jkjcMzo0yRPx&ntwc9Y6y(Hej^~+&*>VIT-~;KNvxDW}k$5I5dSN_GXw~SF
z=5NlKn7bs&iPYaY<Ez{~f0w(mWr}Rom3Im?&gD)8N4e?H4AtA$Kw^hTxYD&w;89l0
zhe^VhGTZc+@j$1VfBp9u17!&gL_CT$gNs$Uqp%zB5(0mstMXqG-?d~Z>0ch+KQ$|2
z?bqS0hib@4|H{_NW$p{;V|UGcZ~?K$H$d?2zSqExJZL)5uPN+%e|JmE1{CS;_fScr
zPI;WkUFjOo9(4~K(Gk=Xul3Z7dx82925>ioxK8fr-?6a<ROK#lf%;8NS~b<rkM7$A
zC9Yq(^;JCwsKbHN<$g8T`Iwm4jE-4+YH@A!-kxliB%`*$dEmtii5|MtzMi|RjVSq~
z7Ux|smGxBM?o2A*(xzXOprbqY<oTM`w#iOjCN66IblBGGVf6qQ321Ext*M`b{y;{t
zhI+ng2=zP}hS2XupkywoM>^<V`5v-~T83)LFCFvgWIHaA@dY^d3e&74xM2pp!;MZ5
zqI~2yf-xdFjlMo6!nn%Bu_oQF+ujgzYhbk0?Eb5YW}Ec5c&=50wzfVcW3}CqGx5Wr
zb~<WfLgDWcbz6#rnk}O!$-K^2BNBwads>~4;j}dSjEPSSaP6-9CSikTg0OY}M}O!v
zgdy-jMxe)3?`8*B`VTPPJ2CP<H9f}vGY*)6gX6zrk&S50#BTj=TipKU4h$aO<@p=v
zU&o-G)eP+Z-^-A%e9pi25%y%`iNu_sLv_T{xIO68qV})5!+tnCy3@I|wJK$>sccXx
z*t|VBe>V0XC~P`RpL&34adz(??k|@kZS=3~`y%!<L_XAiw|823z2f$7z~*23I{r&&
zRrt{ZX2oZtl1niOZ^`ZZZ2PZTzx{`Yja!`G!yY10*3a+felV$_PHpa#nnjf%Ifq82
zT%?FNq_v#NcOEEabETuJbiwnuFm$adNaNRaUi5+9?Uty(9}U<qX@i8|)*tL3ytTz5
z3~D`pDW{3c#lv|f{Eqod8J&W<!xHTojqyktOylhnucybS_v1w&8DhNo76cfE8ZpGj
zr#A=#@@uUIBS;5xd3(P7`#F`TY?+Dnfg^zOXA<T^v4b3X2~!6-Gr<0`6tc)C7n;%x
zQYG2X_nSSQYLC}5ug6f7S+(EmO*wvw2sjNIUej1u9H@jVYf#uc$6LJbI2|r+<pq)`
z{oXgG2en{$$JJ}BUwP31XcR(W)5A2M7zP!*BvuF7In%C2uE7|1{{r9<c9&J-?_dhO
zsqyS%Q#kVRl+(;OriPG*g6xA&e#LlJ>(T1?1C{V=&m+5-`~EMZxpOdBj)4F==6*7;
z=odVQpe#wgtvNb~Q}2i|%J^}R1iEHobr%)3j)~Z$0mn^mhXwaSUWDPd%2*^5ArtFH
zT_SmWY>BLOFH2oj`eCy4IN_9RCK8Db`k2l*4pX9xenH%!)GT0D!)?)9dPNS>cV-Fc
zlThMib&%8MI95hwVi4P@pMgRm8-x6v`qfbk&CZtF>Oj?JSPJIR`nEngMjVy+Y9Zm(
zb1=xjwztWoN12oLWR2jOte<~We?9?a-F&fVCKMj6xPX))k9u;zc$ZR!Pyy;9M@86F
zIhO7eUzAcQipi25SL|ZUn_M_9RV2Y8twyIIggpYL+GRb&SS3cITA|SpH_soBd(xUA
zHG$FI2pW6n1{C4^j4t#vsBzoSGu@@Fvh2_uhi<H%QpAFY6ia%IshXAOR@@0)RDeD1
z)X2ArqotBITAJ=u!b;M&duye@fI6)qrlm~Rp!M=da}k)TJf^z`o3_Z!(V4uzgf=9^
zf7~chw_iRFRps!PY5KAES;n6nstS+)kx@Dxeb-A%r_iO?<!IgTz#i8ad`F_D)|(2t
zI>Nv<E&}yj;&F-0Vox}BE)qT#qo`c?n?*!FsD@EKlaz4ql|M#a)C;gKK*!Y1av?%_
zXpCW>`LU8(wFF<b?Ciorj20uEYN7eD+sHq?Inr{2@iu+AH)3ur)@p8qT!b^D#^aR#
zskD>vRHhPmlJX?#A=}>)13nTSiIr$L_t-czuVz0{T^Qvzwz^%`Nx}j{T|#Vof+S0M
zKGA_HF9EWpRIEg>#>9&<{*9XhMGz0#(u!7GAKo^`Ca)%ME;?J&a`%n!3?yW#vlz}Q
zC#$@*Au`#?d2<nT*~4eiMM(mB*8^rYLdR+Yh(9D)5|ffa-e+jUA+3hpfTu|xZMZbr
zp4Qe!p6`~nz|Cf&06mzMdH$q`Rq7DK@Ip6>Yx#J5h&Mrvy=!(V#G<%aP;k5W;o6ZW
z{Dvnhe9<V0?$g$TpH$8Is5uM02)UV%-YtpX6Pb~C(31#PB(yjMnnz>3M;gq@dL(U~
zQtI0pNVu6d8~}la3m%`)K@3Qk_*U3Id>I0Wn1%j4PryIC6wJj(c<Q7!ZuGQ6Cr=rD
zrbZ?YIoBXyTT}3bfafnxE8}u0$6`_^1*0L^V$c-Ok|6h=pWk0-pl`Ji=ZBbAE|jmW
zeRwkF4Q-cxXl1rIOgx8-=jiXAplA6&908B;6zuOu-r(mQI9QUN<#|<;CqBYp1;#o?
zm&gkXp?{OCuDqc;1hmDC&`P5qds6*L^>Afug8@f%&KpfU{vlh8T@4VU#DY^$8pO4o
zink;46KTXOq}{u1yGG;cl<w-k-X4$brOGXJY7uR<xwYc7++6F3@SX0O?B2KD>JpOs
zTR_YV&}3`Xu^|Z%;53bXodjtPM}mO0jE9rm!a_tz7z4K$7MeqnWwgz&cel3RB0Y9D
zi5S5macM<z)SfR<Nty+<`)k|!u^m`arQ+A(-UpX(DA^WzK$@J3<F`0TnJP2z4LER}
ziLcxF*ZO)sdd9G_)8gok@{VS#hdC~2@OqwZjXgo;OftC#;3$j)itv7^$HLcg97BUX
z2JcmmJ8&pR*lh^}<pupk&GQ94M&EXM^GHEB{0`se#pSDuc3;de2hdmSPns3}gMa<x
znip}M*mTJ!`%a#L$OAEqYW`hyXe*V7&Z&mARAs~{NNpWeeA5=f#<D3}KTwgp5j1WE
z7u2Gpc+<=;*ZAJSm5QQ7{-rZq;%TQ3hR{*`U21$86OA&<L|os}CsjlPzHfZFp<O6R
z+=9KJaDI{&(dzXye`-T@8i#B(P;a_|3Vj^A@AqKdyaWwzlv1rMmdTgHUrXUxt0miB
zx>R-hZf*bOlqLXv;Pe?tCCSNFm)AH|d!SP)2k6@aNQI?0xDv}-e<oGbkW~zMI58u2
zQZNGNgJ?*cbA3-5rP|)F(d=K2A-mrpmgkuz7JV=cArYJaOX_#%_{f(WF)i_*1_S2*
zwCh6u-vtu=ViMxE|HoNccZU~Fs8EdCukW8D@tXv!b5(OzVeWNN@cu7M(N4oRnv5$8
zw_DaY>L3g6JYi`sDZ7_PXE&F(&pf-@zL;9t&QB&1PyDBfm<}_uT5-^vrRSUJ+cgmp
zS`V#PUb5!B@yl+^jni*&@*XSl?VwCDhfhP<0Isiy;ChAhdf^LI`@Qz<2Yc0{FHIB_
zQd;hw$z-;#@B85wjhezV9~hk%Ei&+HT?t_9v<H!;uETc<pi@NTBM@%}j(4J$2=L~R
z@%)%wPOGZhQ)iN`u?+2w*Nr<>;>}a%G`2{%J;FpYpm8)@!c_1|^B%KFEJ*X766Mdd
zF%p$kf0R%qEy0m0o?stxJ0v0YP|LfT&~}y;C70-#297wzhs#9S?b&{$D=Lx4g&N7y
zoSLI}A0_Pz;q<C|uJyc4YtF)9^Q=?kvp*2+W=GtOBjfE$KZx0%E0)VhmF$c(({6%Q
zG2^fc{2096y6da6llw<27_bf)4)Za_C-)bS@)z^~!S5cj^vYM(m+OmDVfq>UnvPhg
zSi&IeL`EMtE<+t)A9=%uVU?o<_9=nPeia4ynHVC<(ZaVNV@6e{19>OQUM42HoMvA`
z3Zox3k2{J(@T|`u?r>nm7^{0XvD}x-hk=o(>CGf1@=meQMFI#1v;!z082w;NVbm%q
z{QC@+Rn$-;iC{Nhf-e>b-St|s9aFz4k#=BuAybFM?$2OJ26h}JOqSp()02GE)mXZz
zOHK71@miKv_>hX?_E~~l%~H9HTWl)43RMqWO`CWQ?DhDLVqGP~rK;WZ$OBF6z;mBZ
z{RRf08Vq!b0fef&b0G3PBYhq;Yj};0WXx;Xtc=vW4<*xu+pAfE%<^R&XV4Uz;8GBG
z&mmgH)o3(x75Cg6xJTuRY`ziUzfzVIcEQBq8}@%C+NY9%(Pqm6in>>}TsMclq__sa
zpuu8(K!jq~wm=?A@f=85^+%DYNVwbzEy2+a+k;#c7DP4%rq_z^1SvB6GP84=7>Xd>
zrq?*;;<u(YMrxF<h%eG~*b8YOTI3&OAvP3oC10RGh)SZdP<M;u)&GbgI974Q-BJAs
ztj6@nj}Z>p2uG2jy<996qM-9v`mkRFE7X4yWxs<H9=MW%UZETT5;*|LpghXItLY1`
z6{lHR@fD<9{G=(49)1DpiAvrZS<%R0Mc<;QVi#b)drrD@2AJ1>k{etCk6ic!L;t6Z
zxDlk!W=@s7LNw%6kEP*6{at^kyFRnRFm#jeZ+hQFGH)$Nf8h^&JGccA0vk4*=^+!!
zUx{y!`}@sVjOyYKKqGGlTQad!Yo%6*F~&Ew|NLF<j+55=ME|-XMLt=vWLn&_@4I%z
zWT)9kkbLrBsv9nGY)elc>Qn6qyIQ&~nuR!xPbu(zPdoSAhQurqIzcMxHRiDipgEcC
z0;T`8HZep?dlgZ;s-2Q^NIVoNXie#F0Ha=MoT2JA08%$M05t$Yt&Qrn_e+38f!s3Y
zcuT|)kghUpn(s9bPk?S0gsm~OQqH{Be#B07Yjgf6rxsG-&#V#JrLl`&Wr3G-mv>LC
zrkq-fv1sMoPSS@v_eu_LG>n`t85~SrivnB;^WRr2`g)_wFP?!+hyjt{kbCWUxhKCl
zf*+YrxaiyjWhWen`U;Mf%)=o2A#wx4^2PB%<@$>NT8bQFsJ(rTxs7A45dFnLN3e`M
z>14ZF4?*aS**L^JDHxXWY?GF5&kWHQ2md3x)%50Np9ry&!H2)^B@e1TQrlK9_g09Q
zLpv0m?HSh(9g42q=`irRR+yzOuaD8qM?d$ezGE&UI4Fon9`6a!v&-^4eP_*WlkF)w
zDnW{H0MV4f;{gpxX2F8m^-FG%W_xoH6$dT<)8aNDF)u4_X}_39lWj?9<Fuf->N?aE
zCSRjY>(+)YvxkLH-;a<3rHun>6XTuS%RgyKwLkYHpN}PL0B84Mm`xh=vi#z5SN8`8
zV%#*c)IX!M?jikezth&j0@I)MoQcgR9RZX47w$hYmY{&BtRNJ~{Qr9J$HCYy5a#jf
zEj+6>ZT7hbE367|DU1R+-@A>LEqH|@m1X=~HFxmV7_<T6ioT|<&<wqqAlnE<mD1G)
z#ChBS&|*c37#^o|YI%g*ZM#+5c>&DhO-Q+HX=Bwqb34w>eCC{3a7d*yReCqna(S{H
z*CjW$t=i@sa=VaSmgz~fyMA}yc`aytnoW~e`s1U2qHU0}7(sHuUbqX^g)HW-+3&lh
zFhArmy*3wV@-I*S`s^)-{K`wsaMH11?XwmDT((M@Yar|W>9e;K6kOiP&WBZH(X49F
zvHUIL@=dz-ONJ`6ezr;h9~|Sj>GVg0Z{-8r_GPzUin)XD+&bIXy+ja;%!*s&<l77o
z^*?@}6SMZb&Nb3&aM%f298~es;tIXov8Q)a0M3UHD4_k_C{s;oLaa;E5N<<wV_fuK
zptYn#w>*8fZyQO+A{D%eeqe)+5xBB%1LzBj;K3mzPSM`Je{sFQ)TxL<M47UdbT@<P
zfp!v)XX0j%pdqF{uty<inH$(jAu0kWbfm2;CMT_O>8ssXOa}9tk9(VbM&bLK?Se=W
zf;g8qk_86%?-BSo94TYtK_`Xc)&kRFOdY(vp0`BPi?of>a@;#>7PCH#p*Tx!=gHhw
zQzssARlJrqCNBYV=Hx(pnuzPHU_$~zpfC06Kj?&>+<81_QWm-AX}|)zD!8Y-&p#Yx
zHLv9`r(dXg+DswXLw!CC&SD$}8X^eH;{O4tAHKx?r#Xb>|AeJ8a4`ON7TW*8H4&5j
z;vm$X@YGX*BEy(5Cz!0nCTXbl3bPBZa$n-V|AisT&oevEO<m6v?6AT-e_a5o|8HEA
zv!~a`SLMu6u$aB2=O-(mPy7=SF^pz5wtqclX&c^B9vDSR$*FK$f})it_V>+s%$OV|
z_Y6RBST7JZAIBdDn3)C2x)s*C1!k$7bC@UJ*u!RCaVnx;F6a}?_I+L3ULG;jFDd<g
z?d{^9r-#?aYei&jGd~t-Rf)W!jNd=^NIdBLdRl2xa5&HZg=;#G%e6cvv5@!)aQt}p
zxe)DbybmNkl(0fT8&y)p|M=ymJAWm|nDvTB6lZ&sgdsB<lY&5z@b!7BmxQaD<A6pI
zC<!`7aNmEjC?S(hYWfnRpb`}emfX|;tRT(l7K_mdS4Rb+?Oyt(a5H^gqIO&I9tJPi
z%00Ru7oCGAh|<Vd5TdQ6H=hN2X8IB5pc2(cM8R%C9(w-`7e+~VzDv9NCclV*8r0Y-
z;+-hEq6^>wwRReX^}Qrd1SF}vDmtnX!-u2jb4gCw?(an2%<yf+JsM9>VDRd>LIpc(
zw%5O@FC)%F@{n|b#~`VVN~mL7_nuHmXhdgh3*R5yz?V8;qc#gx34*0)T&;HNBTIQ0
z)Le2pYh2YqZb90A;=gx6!i$WQoNv_&DEn*?po<>Ba^DmA1#AosK<TfxIFLuH1kJgM
zvIN>E`?^je;*7-2-Iy2O$?*Jr3+E)B^$Fb@u-q5%uP<BSjO=Pf#<|DRiu7HXx=%`p
z42hy3BFP`^cLQl7+zcWJ+o$ABU<sfVrEQnQfEVq^9ST|OJ1)$GPEB^TZjjE8bDUG$
zQTlqrX9i{|{1=NFRca|Wgjv(ppbWanYP@TucXGgCn<a^c>V_e+Wroq_J8aux^{{**
zIt_oU|E-yxsVE#(L*sP6fjO~yVfh|PY>sHD)I%6898jC7&0uZFhwK)@FuX^mk7%0Q
zfMD;lg(>9*UW#=0&SFb9AfhmY@EpJcNG*vVbIhPW{0nKfx(oq5)3QTBitSkRaJyYL
zu>-~&%KcCXJOu+F-0I!r8Y}lbFJ+rj>UfMZF|rcLufItSpu0cU+LU+961270N|3_H
zz{-q`f=siwcWY`nriWP;{X6c{ld=qqW>S#Hv&ori;yI;`WsoP6;2P_HWp^j4gWy!>
z6l6OWjl1s#n7*^loU=w)YdRW3REg-CSy{YTmDR{}mUO{5CTcl^AJ3T@yBVs&W<uzN
z&IcbsB7wt)VFC-yBFi#SV~y|Z0?L({z5tHP4wAc(j~(c4ds;?UJA2x}vKb{*U)`Y%
zo!F7S3UaP9G{HtP%iDjPVIh)P*3HWm=c>S&3w4rOFc|>DHx6t0R_c1#I0WTQo`=Z=
zyy=>Vm6>WoFK-CB?zI2#@Sc+RM$%DS7Fu>=>YUW~3$wD%;8QB(Cacc=0~|Ms?P)pY
z?3u-l5mj;zp?2RUk9m8oJkiv6yY5?C+X7PHLVUjHHy57wJ<W{PZtl&yWcD=XZZ0ZL
z<TXAn_jsJ0INzd)>dnN*P02nzLgzPgXaRMw<sz|FDY288&2f+!9@8-0q>8A|O*6Kf
zf0jx)X<0McSkS;j3Ji_wrmN0m#3}ZtjbG)4NsGYar!wrZdL)zeCHmAnWHiV{=Hy1o
z)=9>1kwtGs)uYwsDU>vBB{bG&hPR4yGBl~~0ab2jpD#7Wnfzgk)Js*zcET*CL_bSy
zz(~Zd;ZNLNI!SWMpa73g+$+UIqOHPHZK`5Svkt>DuTp6u$!Rumt=$q#i_)%3(^8K7
za=PMffoj$jOe~J5b;f|Y;s7fuWv{eT9+b$^e)y@@3Be;;eA`;-v|p-e(ryUlBxf_W
zRBL{ne7vu&DYov4LF(~>(MtxCYJ*@UPpZqpS9wQ}z76_61~NuD>TYYi5RAW3+=|@Q
z(e&WgxJBwUQzR<8TK9U7Gui=_C+Ecos|{-Lm{mmEkjt))gBItt%?@FthO$k59YD1o
zZaZxmtMPFU<Z8gd9hv<FH7=<>T2%Vu&rp|ZmJ8ZG%yTTq>;;=Klo6Ix6CMgnb55~!
zJTX>{YN>APJgn``P%}4e@*LWp9F|Ho$Sl%rLy>ktF}Ivs8;LC^#q2sFw88CCxpb0J
za0i~id$j*%I5I**7@w8dcjJsI3>xpV{jSmjTBldfev%2yF=6_SeBq?wRjY8|V7MZ=
zc5hvh>$Y!Qn*0{Namio#e&1^EdGxI*blLZ*#Zt*0*QHk}H}&O+NpY{@)OG#ypYSS3
z#%cB$)4?=B6|+$mb2y}psiq9XtL|Y7B;NH%YlR9-)!vKBCuxCeC%9pE#`QW1Z7*~G
z4sF*8I1RH4uY2@83U#kvI5EkI{AUhid#80j#}#WOHSPLVx$;)JsYqlOZ`XHKSb<gI
z>lpv!z-KNbq8m4^X=h(Sf;bz;yUy0-zO!rcU3k=JT<M;c<T#OL8|F)UT8>7Sk_KFd
zMoyz9wfN`Xq5+CP-<o)@wXjO;(6KSMn{3Z9Y((bRKY*=M=eYmW&skaj4~7N<^RL+O
z-{75kG$w2h+2MOWe@Wj4>LZl2{QCHk2Y@H60=7XMFrA;vVdww&q#aU(HFOc%;22xt
zTN+x$F8_|@hw-@{j@{hro7+z^v19Z0!r|jJe+8zdrGkOwR|ot(JiU8OS~Wk$JSMD(
zNtl}GNlSQz1MQv$NR{dZy28U2&?a(%K4}>}VG%X`8x@Uz=;Tdz8Pv+D2=94^IQGql
z=Xc|bpjxGWXoLMx))u$p*G0;SIl>Bkl;b!G`|oFa1P-{m8tM<QGWlVh-O%ff9?!<8
z^i<~8kmoPM;%v7Y?Gl7$b!nB6t~yMW;d5X!U-~P#MyyV=$y^MqLs$rri-5Y>o7<gT
zGNFc%^a+`W1r;lCJ$Yo57%j@C!Vv+M+Y~mk@ambYqS>8@QpL7B2I$b!wj8aobz?zn
zytWxeLkhP^TVeExnwQ@-F-a$ZZ4x(fnp-3M;M)LChe9T;#B<t6Q?s<}6C<MBm$6m*
zqu#HcQ!zebL*AA#F?RMvHQL#CZ_bvJL&UHf%l0FYFl&O};j!psf>6wb%CiBZG_QE=
zV8;3}3wM;#vDXo$G{mUs82wiM=1HsPFpbunZvah$A$W8R_0yr%WRd#XCy9vT-|73W
ztzln3fY^?ii0ydzatddDEkoQGqG9y-7#!`U-Ft^*Ln_#_vp~ufnQlsGLncdVm!A%H
zQo~ECCl&b?*-~29<N5~I(z<vusn4st;qH~Uy0D+C-JrTb`Aa{RJ_Dp#`{Xj3xnf@*
zXW6MuV$_?KCAx1q35WuwT~>|RX-ySNPqsxd30W75t`0`a&ZJp@Q6=?cJ?#}$_VT6S
zMZep=Ub(5E7;1J_q!(%)`5_C20wZ4LSxaWL^MmXn%Tcuw|3R^dG*uY!4?}AQfSUcL
zd3jDSw!ql=J3N-2eS+C~Q%2e1=3X~)mVjlYQ)EXY3hN;D2#^j{bUDy`hHN{Q(pp!7
zkd!I|?vndGYDzX{?Tykv##thdn0w6Hf-Xw2?Fh`Rsy>W#IiMz`_%Kx#k43JPWe!vm
zru}qsR;_|tL@{WHsuGA20*?%jBW$LRphg3uahn7!Bk3y@_*J;kDjkQ{TG?r1<C%of
zx8QPl-g3{3oQ{t#5R;*#m?+wYQ4Do#*CKikAy>gvp~1nlm$VA43!+Lcv5+~Y{!Rm_
z#5La+v4X@v`lWitDO8%nAhEZgY&>#!6sN>YqeQKx9oLFj$Q6%hUdmKf=DM1lT#3TD
zXmfIXEZ;1l<wGxRjO|;dzhNXoQ!zQfxhW%5FR8>kPFUzy{S^3w5N#q($&*3!&K@Zj
zX6m~wiC5AqBL@Yz_G6!tEa~5QCi9d-Nr#*??OsMJh>)=$;|U}QD3dOZJHp_xL}q8(
ze5=A_|1afd!E2TIf;wH;WG4GGC+ubc5I>>x2I6*6q&2lc#kJI6Hdt$XakA6$EbCZ<
z%=?ouBJqWtJ4hqRuJd5-bU?zQVVC=pZ=fK$jzSKttyrws<dGEEtnfSfAv?9a0SMNl
zA5)SJU7Jf4WfC?JBd<!F&CF8_b&T6pSOTJX8aN5B$TZgc!!&+wNnId2vLw^(`oXl<
zLEZgg5B_$77a&$h094nSFfLocjs{SRPH?vXf@A-=gIXu;R(iD3pmUNmE`wTlESg@V
zBOn_^hHi7S#x|vj<@fN*C6Ka1Z%EIS%B321h#vjbBRtbnOkc^J^(S#Zgkd#3L4(za
zebky41u5B@e~_lul3ZbUFBS&>fL$;CVO1#D^0I^RO%R>>ptOkBHa~jE=@5~ClkPJM
zj8bkn1$ojR-lE35t0H3&9`m@7F4>U<Ez(oY_i80y{W;|r?Mu(T=IG5mL66fF>}zsq
zvm7wp+VX%%GyAkYeYqU4-znJG*ut*O%t%{uZ9P}?H%|Wt*pm2j_@C-M+kZ%QVEA?Q
z|5x>1b0+qG>Bzqq%!T2|6?pm#{Wru&`E8)TUL(rPY)A9mfAW&4)LM-^OkA?Ck474$
z<8pcAiH*3uaGq}uPaiK&_lEzw=I{2x;IlP*2c?Qm{Ti!Q`|n<i-Yzw)njbt;;4EyS
zy|aIjxl?7oMgIvPHEtV(-NzpI9^>4|-g|ZZ9)qaoj=B%OkFG%{7jE-{T;M2RJIMay
zH(HY{C*|ASps$kRf2#dz0^){<l7ehg^duu@1T*Y|RMVw9+#+}uI`V`yiuv!q%0rV_
zr?L5M?-T(T%5eelfz1A4DiA%(8SLKUYQptPS?8dI+d}BdU~Kn}JL&SlPtCI-Zz3GC
zJ2@dqI7#f_!`t#yq>ld51!K$5&e6)TNLK<?nkM^-T1`YOZ!8XRZpod#$;Wo)jrtf@
zNvaX<0X=zvw#VAT_bBuA9s-(7Xf7$o!L@6HSdqVA`oudn(OQQk?%|olX?TcIjGFSW
zBZYI;RaA+2IRUV)3{JHS1IC54S~i+}&5x+_P(@QB?+D2x+ifEf8#&?fYRe$;&Ss7k
zS`(yWMSM9gk_RHkT4^sbtLFSdo3~39!!8+cOT&eh%5;<%vxBIYl*wPPJ94vxw~k23
zDJu5_--wCF0&{rK5eEUy^{Q}8vpclCI%|GPt;(CF9Djbe`lrJOI3Qc_UN;7ty!SRz
z4Mfr0;o+aJRY`QOCJ5G!%h&s&wvVsZM;%k_`Z%jmsGl$Q2-g3!*}^6?+N{5)eM4<!
zd|C}2Li3VSOowQfe@|ZvEZ#HS+Ve*QZD<}B1*_DfPw~01HzG9hMEjJuy_%5=HX;_I
zUlC36=t-Lqp_WMZ+$Yo(mYcjup3!Iue~34PHe;|R%F#S!dQ5gO#cb@B&KVdPY6ai2
zT{>aURk&^BVTW+sEi;@D;7tA5<uP3r9-cv`(~1f+uHUr9Y*%gLuqmQ)U@UrS3))&?
zLCBbv>x^d#O?c}-m|iA3I)|~`VbhU~Okxfbi1j3o9>0rQi@UAWV~S}Mxi-MpT@P#R
znwc1#LMsncHC^q6Y=V7qY@1>}QLEuYh8CA4IrwVvJWXhx6U*{j8&5x?tuN&A!MTwA
zZ*U8Kt(|m1Er@(P_zdOxoVxy6CJ2qzqc|RT8Xe=4U`(&bCT?!Y#B%owj2{&UiZ*<{
z8&EMpheT;UL=4k~j%-oXB&<ux4eWfu^dLpd>mhWzBCl=uQnIe4`*6aoDeh2>P1|yP
zwH!l8+<1p@O=Ri{54e8g4AqCSlYl?-nkd11IY&m{1xfV)oXEKMVZ5|tXT-ewa6M|_
ztx4^F;r>uwtZyHg-fI4GLhM-ZY|}q_RkT?${WavEjVUncW)SJ!v5F@087_NV%Vgh8
zafRFOyeMvmwF~wYW6V(n3NK7eW`OCSwcHmgb`7>TQ@+MBuT#TW*Gye6kxZBiYFk)%
zXMZS9fE!}L?EOKpTp(m9%`KVTJBB0mD!cXB8$5RLxXh~DPSZre2*_LCNtzr1Ym3un
z>?BFpwB1R25}H>&{**Fz+=wvv591c(F~}k&#VkMIAhc71L+D~wRO~N%;sY{u869O8
z&S<@~N=uaTt7<Sw7VA<1j_LVrsh-ya%F1i8QOHjevtO=~9e&Hxl>AVZrxv9bzBbzW
zpww!EnHS%mx_<A)fDWo_m@oK0%ZRG#n1341j-o;NUkF<8D06d=ubfiQ9$38K3E%{o
zn{sAL*M6eJ+<-en7Yai0Jh&0zkkC6@hf0lgokI*s&R^g8T93jMA%?jeMo-;T0<Y+?
zoGudQA#b^gEL}x}IO=}V4rn<&t;PBoE}L0UW@GXqqzGZF3HmE>enp0SIw%os?xh+v
z=pLM_zI<<4X&;dnNE2B&EH5KBZ>%Z@W*Oxwe2R2<H=DW9bvZ`+PhQNfqJ(vOfCCUy
z;5)O{V*8nV&mao99=ki5jBPJ+;qr*Duf=qd8tB+9l2_ogB9`i(ywR&<#~PE{>!(MO
z))q?`M$14xLZqFaE1Ms4AGk!9DfCAS{oT2H>LIeL!BSm@{mLp-+)pp+V5KBc)LnOU
zR>mwM9RxGAuLAJ(TgPv*0cypp0alorFF22*Tk(?tb2@d1oVLyIwYdELh}(F$zKCL_
ze4F(bzI+oM{uXiiBg%0@u)Q88;3#dw1yFZ2WQBh~D<f-SH^F5yCO}qIgG<z#6;$^M
zcEbT~%e<b!96!A6wZz*n)wxqtdhe3vnCy*?Mdr9;tuY-+;oNi&j^mWSbx|)chCSUn
zu3L17;+(dOUC}O@4`ZpRSUpp;JbkONEjM>R+^}FcM{-hBaRSl&_!^FhKgY8ENuo?`
zGR;*zUhDkcPb9ru9WPx5bX6A&I&1|oDT7Z2dO*|dK}-YqbiSSMB4^7JE;nFT-l5D{
zQ|>mo#dt26n6{V;BP%C;JCJQ19c@U#w!_9EC#xEY{C_caj=`1n-L{Wy+qT`YZQHhO
z+ugB~j&0jk$H@*lwsG^^d+XG#d!AGER;^ke*N0tu%{kY9j`16F3Q=<+z$CBvZLIEL
zHCgQ;j|Z}h<fi(r`N9CTY9(~2*@H>1vCsC-=x{}d`!29{bFGpmP07sM9Lm-VhU1tp
z*ehmdJ^t;~eV^5i6zqIpaxe$@EDUX2UC!}-ak*7(aDaBjd$OEl-LxD7-;}wE-xqWJ
zr1~g-0Hfpi4;L6QW5VbeauNQI=<0*4u*Rp0hEQ6a30rRq2l5rE&f&!oAGHXGEcRQe
zO(9*ER({A{_?DtfQ06c4jq+ud<I^MG8rI}QzwVz#;+`9ZEVTXj^&Hu0{Eou|Ivy(?
z+SWtL8~wfqXI}a;f`<-%ytU)_<eVRbPR{bs|2Bwl|BKR%ll?ziXn)JQ<8lA5y!(H{
z8?ivAN`cykGj$SGklp|y?_UwBN$1sT)k-q=M35Caw(j-m8o1Oyd<pNbH~#*;xPP&A
zf}-I;HS~Pr!1B)jyJ3lQZx1y><uLuv9C8=4vitW=lyK>f_A6LNf~y;1UR4y^AlK`h
ziU|QfINLkIbiLwsb5ECv6$anp^4vqx=--dxZgUvX**Vs`=j+vtD@whe%QRj+hcWLD
zukNo;Z1o80+<6ePN-u!$-hk|9V4N~v$&e^<EiC^}R$tx^;tF{iEAJ9t?jP^({E<nu
zYF<()w@pto)v1vB+?bG8DGl%CREP#m?qqnme?>q@DxjyN$7YW$IjDJRtK^bNtdw|_
z>64|IRT!4gSn>`u%mdhk&lOUo@NF%Hu<xU5t=tzTAW|Fc1|i{*WZANAns4uWq<5An
zA;LKD=E^8iP`ykb2VoD@G)N6<=U%+kh>Y}TD!DmpV{4pIJlG;8L`Xqx#_g=G-mk}P
z+zB%tG~G_lnF7qL)JbJJ<9QnOjUrKvb3^@kG>3^^HY9M(nyEIx{N@Xb68E3^Ro@Ja
z$PyytHUe9k+>u0gIN=#95<2k=mB+p7vz>&cAfY<Sk#YcEW;&WzY9v)vlnHHI4c3k@
zKO~uF<gU3y*~G!;P9R#vr1BL(t_ci?@jSnM=09+8JU=r2&ct~X*E?f|<mMWCTsfoV
z`3=Q$iz@rl&6x)Xl{?YXy&ty^ad6H=>D1gLei+>%luHpGe;gmo{SL&a!f6b#%op8(
z8fg~85TBKG5$*sbBI|`=D(Nl^`J@;vFGTnu$3V4heOqRB-#OZ2i6)%rd$EpL;Y<oV
zZ0o}e{e|<U^E;ZE;UdrN7M4kH*}CBeaUF#xDTwvRS!5u(&K0mBJ9drqb1wuWjv9=4
zR7e%JM}^5xaj6Q9g*rEW?;1ap!|aC3m6UcI>&~Y7<r~^r$0Bp%<w%E#g@Y|L_E3Ox
z!*`9PP+oVh&zi_tK;WqqM($>Tz6`Oza4d`+@^A#zF6;Y~i=QF^10tl%!PqX36lRaJ
zmD?ar?V3TF0U}ff2ihJT2fl{Ai~YVT7WBtY(ylB05m(~3Q-MM6;%d&7%e%I}^u>jz
zKNF1WAZP7`ZS#kVn_W-O6wBHu&xp%ux>!tsZY!P9P|skE?%6AZd?3t#E?;%zHW2x(
zSF~lYwF8w>qdiq(q?&!W@;V68(&pOHPX95~0=rqcS9jFSL+gW0k@-)8+GKD%<+UW(
z?U-xi#({DlC`P>{rGnb)K<pdvL*5`4*b5U)4z$o%8|}(!s@~2NP1HzuA#BboGI`=$
z4}`&tbtP(*<?bP;tgu0OhJYhVywrtfSSOs?pNzdp;%a=JlMr(c?bZEL8h*tKr|1mZ
z3hY{tKG@jCBnjyQ_MuTf+_dDX6{3@c#T1HEGte?Ch?<LFu%R0z7i+8NFu*E>^H!w^
z?gW{otHhm{y3a6OW~)1sLPnv^k?*72>B@f1!;-^`dBHH&C)wUgtPD2@BgOYrtL1}G
zd!?#w(MX(wGH=W>?j9cV__~s2yS7B)7h!PTp5%&EFoknn4Vnk)hi2D{V_T;WXF)Pq
z;88_P{{a!>Dv8vS2f5126c673eXjXorwo_;AiVjLjz|Dol&i17LDHqo7GA4=2lh{T
zq1C;<Gkt+wr8idg*1)@(elP|aJv}yC(K0ouF3fcw`kCtBT|E-#sL5H{;H7cEyCtYG
zoEkHr<CcpO_;C83_?Vk42nKSNa0LCnX}KFm3*%cVl~Ktz`#M?YhO6XIrW&)P7DNWc
zF60Pt<a5@EKPQ;ap9=Hae*+`9FzbEKvsw?AM#$+Zzmg#4qNevMVY7~B-q)71-fVHH
z86L_(>HODU?<DSBuv<yuA$f9X<^<a<s*_(xZ1}-oI`K}o8b@~Q4S={#%PlG{hl>8H
zq^?_MN+LBbqYB{ri!usbCmlj<!A-!hPfZX<9RNiF9kls(_XugC54x>y)g3Zz;!eph
zvLqy;zRJ0$DLqxMiz&VN6}}eZkFHjG`rk_rR1(VP{;4iYm7_V@{cAb;Kc2)>S4~JC
z@p0P)nFVJdc`!E{V1d}E79R{Hl-~*S4x!a|xF=B?9crUm!5Beid`2Q}F{}v6+tA?L
zg}H8bs@XV}u6S#h#sNz=%v8r0;wN6N<WSquH)e*t#0^daWf3;!JU_cXlZ57a26q82
zNV`RF_T%z37Wye@Z<$@A6=pBT@w*O;>1kVL%r60G6HMhuA*OX2REz3=0XU}~;Vh05
zrDrIQ5;;91C9C=_k2E#44rO0JrQpNl|I#V{f$cL9IT%~R@bUeZt)7*I`9J(*AGgQl
z#_2huJ_Vn0GyoE@l!=0yq7QNi;f4#AdQD{O@IknHz5CdVQ6tYVCJ#JxRs!^1vE}v1
ze}j&*^%1TZq!r@5NVBu`dt^Zj9N#hUVpe43q0G}V^?MY)d;z-JzVgqOGid{(dkzU@
z-27B~^&(il*VRa)kY(UG{RHHTT4sybVS2Z*Wqo&1D^C25lh7c$(7u%NqXFIT2Ysyd
zR}4W0MN2%>K%I)efz?2hfjQvM!})1?L3yAgXP><Cq~4n3efCF-f8Zz_Bn-*&U6LAI
zG3e`(k71d+p>FwrfT(++<RCeaH=wYUxez()^(xC86=OyEb|eynk~$C0myg0Fd;?Fk
zSgM6al|<+?l~I1=MzkU_M)V@6u_!Z}44QTsg`fSsdjj<<0Jn71<d_0y{`koIJGplN
z#jex;97}=s&`0n2^RJM$d2C+Q($v0}$6(}b-Xxwy%Cb5Mo2EMxL_sEe0g{j;RQ*iZ
zJ-LkXf#%1<)2fx`%l#&i&UBYMsp(BFKkMw?Gr=f#e<FoYPPOmv?Pjz>6<KUvW2Ir5
z-|TQb9R<liYLdN)BN1reu0f}~SuNvcM+*Z3<eeyD>37o$DJu<zOTykR{HOMmnE=?1
zWrRf~1+TbAz*u`M<9He`D~lnnfT7D)HrJ+xMoMn~JFv|D6c{^DcPsIFA6N!xGGPnp
z3|jtF&`-pl51fixaIO%Ut{5Mi?%s>^6g0f67Jw5gOv(jIih;nZ>K45!x0+e{MGjdV
zFWOOFx?U~shRbELX8S{K=PaVkJNA&<JME4NkZmec>6rpiN?Ue5m!;QwvtUW&%|vM|
zEF<~gVXGur7cUcGj3M%zZwWMI(+g=zIZ1EF&2POhR#;d!w{S*f#e6-MOVP5DB|6Vf
zeX|qfBE_=EZb8j)9mL5Yx8_163=R|k)nHm(V+r^nd#q|KQiB|WYYARs4>qY3vu<rW
z8c49>!8k6J=+;CRCoEJk>6kK%(r(mQD%TKJYd!#;*t9WJCEcpA56r+jQTkzp?U?wi
zSBIGPJHK((1(#ALwo}X%;7g)RQ@auCC|hd@ut+~t^{ZZ1m!e8voWamgfv#yF86$Rx
zvECr2S0-lB&0U>g5459RDnv+kCT=(<D=(H1m|tO~<XfH(kc$kRRe_6$Or+^c{o))p
zqX0XUp20)RCD@jc!v<Vm&ej_Ezr8#3JD&D7p*C}&6!*1U{YAe#?Uo3*rtkHEXmoK$
zq3biJ24!{i>ByzM$}0;IKPyj5wOi!WRwW9|YGGCX(ns9^_iR<Ymhd!dCVE?8#E8ZT
zA5$O?-iDIqm#O74ke{dZdwps^FZ0hsFen`yrBgJi@>tuCEA1G7qzC*G48nP6JPFfI
zPnG!N5e8{nH+OuxAG_W~(cid?6q-l}7O`DX`1?IQE&jLxOc7jIi)PlzPXq$ANYly*
zt(lB2CCoIF1eOE3_2c8khyM{hBDd25<c<jKR7}vDso2O2THW?Be{C5ou_a?`tBWKY
zD|gY<I$LEy3QGof{%sWqu3SJK-zsT1V&q`7+c?M>&w2377`w8G0$~UR@ckmxCr#hF
zo7$DS;YXJ<lVZKC39F7{QhP46)>Nt*bF?j~xumjtTxT=RRJcnkMsRiqB2DB&x#q0w
zmcB32R4!qv{WhzDx!vH>+e24?opC$V-KKd8-P_E|?`oZ~Vk?_FZFjS$t(sF7sbOYq
ziWUGAV8g4b>iLNwm>jC!=TsM+y#2L&Kd8z*-+P?t?|j@?-fBn1&%yx%CquU8qk5-C
zBd#abs`B^Vu0kvp<qRcbSL9oPd|h_sZPLPRM&W^88wSxis#G--n^5I}`&f|nJkfO`
z=VLQa*xO66>Y+Wpj$nu~a5+i7g(tsW88LOSgsGX9Qn$C-yUJ%VB4?G4Wm8>!#|CZo
z`%@CmnVH1o-=5dZGf77D#kRs3F4(!I!Ov8-DVa?bHz7VkhfzY-qsnfAvaM37XR^}8
zrnbp*riL?Pyv21EHd5Zz#(yS<n)N&h_&I8OsFM3-KjnOjCIow;-TdM(gO*gWnJ2?M
z)eaN>tRbSS{<`<GUbdpH&)I%IM%~DI@8;BTQYYNo8A$R&8oU&z4xhzaPH@7f6aas_
zUd%UXjBriRXwSHxHi1_Ms6?%1!OM2g&mqpM8G8u#!hK~&n*<%G_R12sqTiC-itiJ`
zFMf6usMdkyzx@3OysxMIk&4>#>pZ9WbDCyY3FAGu1U`o*egt1`u}){b#`)1KJelfh
z1XwZi*PmL+RofDe&_X`85$)Cd>OsC6L@yWGF5&EG9-}b-l@V(04D7EzoU76yC>LUe
z3t{{M;?0;3f6}u3KeRO<hEuDN9x5tB8js?x+IQSIX>-i`D=Ix2eVFagY|F7?F+fWV
zT>GVw$B850q%@>mt6ncLtC&7c=sMM!4!c^9!|+ZaVbp7C{y@)mX&gl9I1nO7xFkmL
zx@4*DA145i5WTb4`p0!@TFK%-F<?@Td3JJ0y>s5_7Zv&P*|m-LIZZ6@Oc4<-gay5S
zkPI%B=tbViaY`l|M=bF=B-V%%i<oe53e#-~_Gokv=$!=MV6gplRT2J@X5FSJNqg8R
zez$Yk>?k}lV5Q6*oP3hXU=V0M+jRci?435JaT)az<Xm(+8vh$_h2!LXm@YkdwDgTe
zL+Dsn1P>8xPp!GU6qnYYed0O%KF-MnB=1aYNh8o5W&cK-_RBS4mpjBibf25ae^frB
zo4KfKvmFG<eqqR}S%uG1FT+pSj8){YzX^4rE8i-Lev6})HN<mg+9jm(R4}2(8^+@n
z?l{mD^2FIaLl#*Yu9z+?jN?0G(Vkvh+{TWAp@JTOVf^^v^y_oW&e|9}`ImuX>uFNE
z<_WEM{Q|3lCtX9x0W9~*>)+2bhL1(Kv2K$R+SGXWPBn2o9^R7;a@LGY_f8r^A$B82
zPPM*P<vOD%Jap#2*5dt)zhn32wc8C;v3%e&79BJ!Ycd8Amqsi8<e)m2@Ri1v1>+&;
zR>!-*RM<3zW?rrO(a8+ipsziLLVO{9&4c*;%Xr1k^#3U9|0kpoH#gURjb7}`|Jk-=
zS=-J5cNF=v#-LuXx0tYL(`!9vi-c*H%<Y#ja4Z-29~@yp8CN<DWVL9P?9W&JMu3ts
z)pC=h*l*w%7Y~=#yB5~jDZTl05g76^G)XozJB)Iy86-%leNp&dzYJ)EaCsZSOq_*9
zCfB3zBCTD1?+VfR<X@9b56`OcHCCw5D;|6<p>aL<NP)Vz@lOPWv6~h^{JwV~uiIB@
z3lglT6@vIZuRrsRSwNiGv*s`WO*mfA1s_p46^1M;H<(tG;Z+mVI&LHq{WPt~hcRLY
zC8S&pF@pNG#ZpJAYUuWtHSc5v1BR{JtoJp*kF_%hBI>^7{#;b4F4Dmat|9%+NJjAE
zLN={wZAqQv!)ZtU67Y$g26jyHy}wbMFEt8eXUN-)oFt?^bRm|Tmq31FR28&YD=Jp*
z7#-QKqB)cIlDdq+TL%TRP``C!(^a$``BRO%xnw|tc^m|?WY8Q0D-Yw)9@%nfMAs(W
zC>G5?_UpI@>;{*d`Z94CkJn-#Q2e1D%5a}0LYeN^kp@rmQlNcnCy*-vuiW<##I8JA
zz=U5}TOBr%+p=blkREPkLrD9P*(eo=uZsaDNB_fl_f9+I47k}E^|+F{$<&__Z-=Tq
zI8De+sxt7@3CIk`C7>b5`N3sSEkREa95>9jS=(>2ZFX;@Wv<8w-h+S&CRLdemxH2Y
z7-6Usw8CD=xnZM}$dFk60&dOFQ8gequlcB}r%dIb>*+9~h%n$sUTkh&j@)sPMux+Y
zzpJD3V=dFGH4ks@ls=_tkSo{g<VolJYOR&i^&r6VQ@eUN{2`T|BB3jlgjgPFpq?5(
zJ{OYjbNOWBZ^Gg{=N|0uC@s0)$!9=$I&TN8?=I9mUU78FH}{0%0bip6A#OVnjgyOu
zgEL6t1_}ygjT*-XMN&w|utgaMFTO0iuw7j5ccU+@wMPdpWOMsNF@M+b68wWQi&a$u
zo?)M(u|Cfrp#({0VnEk^@ontwt~bt4uHT<5n|^xV-)3PJ5mFK!!*XLBd2@eyfdtJQ
zr8K^TQ6@1UPdPJpXXDE;U}%jh_z7_HV+XK*tWBTr3quj?e_~&r>lC@T-5JoaSR43X
zxhedv5m)#vBMeD`I21h!QHUU_gR-Q?)AjdkV+$4*OuGXUn&MTTY6N*t8e8BTw?zz{
zXR{NTzUa8JjBr|^)WY#?$NzceE8xbSU!eDS(enOo{C(>%%aiTH?EdpM<L={NJYwE}
zu{QvFCog_GTYe{Ygpc}QXlV3NeX?z(^dd7MN$IS9L5EC#eq1wN9J8@#Z$<!*=;D$v
z5}U;8>Q@B6iqv^0)tvntN(F<ueM0r&^TkFU1>vlPp300DVZa~vU)FN-NM9kP<!VU}
z_2?f@htJQ_Dl}$afplA?e=R7eQ_oMl@Mhwo>1MJ?NNq@LcO7*Ul*|2r2p5tujW+4O
z=g`w_Wu{5vWufQ^=J_pQM;SFR^55ukp@a;rQ1$JLVRiocFL*>W+Wz-HeA#|<X3maK
zjupVRQj$OaIBe33Y;+by)~$dQ#Zm$g>TESG-qld-PVs-H;)}NuSEXw7>(Tn(_tbx2
zmjEk)2e#0_N76Kma@^fd`;2+s8Z|JrTo9((3<GlwzGVPg0O1b94EdqjR=E`5a0Swj
z<p}cOI$FmAb>^8bDVvDLGy+@9;G6sXN+B161jBVyj<-cbXlxNwud*47e&HXB-hk2a
znr($ADHKl%ypL!rGp}Egs4c*}8aYleqp%xedN5yV+>EOUn5=_2_+<k6yb-vV7}aFk
zCzV?@=WN%fg{H<LAnK_fGuEjSJR7HW*sQtiny0y3$<S8@1&cM?L}toGTFJ_XBvMB7
z|6Q$i5wq2z?IVCeguCu}5R16s2%I$Ij%sYxp&)0P>2Fh+wTxY-b|`pOLV37x=t5-t
zZG{mY=s-nN%mQLaWkLXuJCQ7>r;a;I-SDseMlJRJuR^ssznGQ_C;F)2bU(U`z&5tN
zyrl1Jf#HQe^rV5M?P`UklED0Fi}+CWU_zGdCTFCB;4lO6QR)+egG6@RI0nlh*<2tu
z0z61Od3o>7w|V2|l%hMPj~Lu8Q3Bx0{g%9m^wG~U8qzVN2Hp+QF7n%atN7=935=G}
z4SA&67)_Yu-V8@%Nw$8+R$|ze=t@_tt>Gg!m_of%t)}av%t3)K?1T8_Yr1BBZ<XpG
z*A@JQ!vvz8b)dvRp|6E9#F|tMocbP6LWScj8k+KkP<N_zAxGM1mHN=YzDD*xJ3}SU
zKiI66SEQBi_j#`&QwM%)C_^#Df^=<@4}ZbHM2JXd-BWS4h0lbLT5W`;+}LC}TAMkZ
zQe<0N7%swdiOmro#&Kc4Nl#p@m)L@X-n}w$Ik3RZv~Pw6t=AYDZrex3Y$eP*Sf6Bc
zU-{fejYZs?o4y~pHsE}j5`Hc1urd4;m~R8<{4x{G=#|?HOu=Uy|B_2ln$UJd?^bZd
z){ChrKpz{F31MNwHnU`zm@`FNXFzj_D%DuzexOd~uh%qX;J3w0#usv{UDijH)=RS{
zPK|tgQ$H<%5-I2r35Bd)41wY<P(0NmVUzlvT#Ko_hpdx5h=eezsA-X_hFx3HIR@?W
zm^oO^KE|zWRoQ6{Dbqv8J3$L4I8z+Qvv!kkbtm6{qny1${pFAsUmVMFR33NKT@OWV
znL!M}Cm`$8(V4t7e;U7JH!$<SWod;!AjTaL;fRf>vdvh!(V1L+Q+X;$eFW2Q*X>qY
zQhh{hI!?0FXGmK@zC1ltW_Arq=!u!NIPs|M)wk|EaU~yK3fV4vNGO}TK`|slDNXIr
z=Y*+B=F+LASr)ji(}Vjt=VyrkZ}WRHf4+A!c-o0$C%KuNoJJ6PJhMA`2_Fr?)eSvd
zbAsra`?ZSKjL1G!`S3`jW{Ksree2}1${%Dw-$b)SBAC?^RQ$$CbvrN#-G15(RbFVX
z3NcyS!8f?EvQ^yfqgv_Ut$2QzlPiWz8!}NBGI0|kanrK%Jj%0UB^%pOxv-2!sxkQy
zv1(ra7FMD{uw<Rmvi(?DjFYrP^tF2-CNZOCMPTGwfeaXl^tN-BvtGy^S(p~c`xco+
z#}3ZL9vpTap6Ezp+W`itSYi2vef?8Ntpge%%NGQX-)c^6rFWawH7`i6z~(A*Ba!|O
zi%x1PL@-1!q%g#HlW_QwN077qG8UYkM3=U(s}Io9t)IT7Pnkcbk&?6IhW+Cl>iuWy
z3|p+(YF`giVreuVTk5I>UYrT3Na`^oL>Rss^Qtu1MF*OAx;J={H*8F8?!_EAB~;gT
zj88B83q&57o$6o4M-JA1*<<~u0?K7g8Hde(@-a0|5L1aGzVoX`K%!FM=XBuz$*(T`
zKl#-X^zF29#c7AjJQqT=S;<E?(>}IO9u}Roh<W-1{CGb+=E<CrIx-w`20>#CES#{4
zV`hHt>OI|dzkSSW!?f}58QXM^qThH7=ppoaW?YA(>Uy4T?q!VC_=b4!<@7x4-N%jU
z+=q4!ewc8NKCMIOPzC1t(5&tyWEc?IqCe7Mdyx`*5H|o<Q(%)Bj^x}l2Zp3};(7lZ
zdPwEROUYVR8{=94^C5928LB|h1fmrtGnNY~06RXbN@2%lve|y&=}m>Jffm3D2ku5=
z(xao?oEnhG9)!BB7k1^qlJQfPH$sS~JYYuWveJVLe$)a$yi=iO+Xd(S;P?Skt>1Z%
z;WwioJUj}&(>;e~zmQ9Ib&(N{=54Pamz&~)lbl-K816*Q*z*P)k(0|GpeLI6Q*H(o
ztc+)0z7Yoj)C}5e23WW2XkWgc8JZC~+?gjrqg=B%#z75QB*~nUF5A>3Ce`GVEd>0V
z36nu%;g3!sx&j3`4WHYZ!7pk&L^*9_v7sEh>#Xqug?Lo%KG9LCz5D&`6ubb=r@csD
zKiTLPk%n)@Q_+E7uPY}0Ph=<i!Xg_nnFZVICa#C^NZtM7YkM~!!6?ELBpPdOw4{lP
z%GayO$w7e>RU(oRVzRIIPep<{R8(Ve8$%MF5W)br9RmMF)ykO_O<(6%WS@qa6F|E=
z@w%`SF39Pa<oFydCSJZ!kmbmWF6u))g%y>Ie9%^wk2EroPagX)0S(TtZsB3Ena9N#
z0Uj@cd-CxI{FFjj_nw&XmFKZnJAounNTxFSkGTw*MV?|CPigZ5Zu`E)0RYjJ?=ehK
z?%4^z3*Tip&DYBHcjZKwi$Q9_aj8)>UX~y#K8!R=1i`OJLm5JvJdTZ2UpC82isa5H
zJx--VTbAVJLhsXGPA^*B)QQku!TZ~}$k2b8uCT0r?YDo6>SZhA-Ym;sNtzA(z5OI8
z2TyEY!7ZM(6TF#EeEA0}%a5S+p{Y}e#?ood+9_9ga?<sQabo$a$}NUV+^eO8wsD=D
zEC)5hGt-H#xQAqRoBwPD*K>~7*%&1DIv*rI&xthemNTd*YhfeD$n4Pd3$L2YbLyz0
z`10i*6HY+hB!E&{`{vP%T<ThIYKWPuis_+7$+~wbqb_&X2Ea0_cBc;)Xb)|`Ynz;8
zyCGgWao^ZWS*6MfVygsq%pU$yuwZZg%~(D$uX;)v$e4bo0V{s?rC@z1RGP^;wC}k^
z2UleN$DyC8^eOyfAX89}kuH9Wwz)a3uc$-_haSGvzr+d=#O-<-y6_GfzN9ks4-vlj
z&P34W2GJvQwf`KJ3&Cs!Ef?!8bP*^asuotFAA?M09tUB<`WdsWNw6M1W49XItd`4{
zx2EiPmkIZMa6B`eJJAm_1BsFcC;MtV?}T{%SGhySO0d<*J<VVbnut_ZXI0;_;FIc1
z(--%Ymf$wOO%bkQKr`7Y4FTQ|X>~bcIeU-19l@YYrJe!A=0V(Yd?Ee}dTR!+*fsov
zywX~o!fKa6tHSDg&G}*(LCXQZo?_TL{4=t+Xp6i`TZ_U@;p~@TUTz`}1w%LUbPc%t
z&j#B}&uQUKL8xfM1BEk8wl8ct_sud|dylm;S_IyWmW#!!a*3@nn;)*Et+o1gdHw?3
zfB5yB{&|+xXp4ddN|$q<eD3_nmRsH>({1#}&o@_R$DX!hem$j4r@R`f+VjQbjjJXF
z#?6pWkL@zr8qUi_^ybd*8>7djSzb*=k9}Uv!}dC>u&`<wKB=5am)m{2UEU?xugp)C
zH`j=ti;KB$_qwibF_E*jS~gai8|b1_-z<Dwb@0_pdO6M+B$_{w>$t5T*2)rIZ<s+O
zef0lU@7bCE+ct)ogY7?s9LpqZaKZH)Ydj;y(%ck4B7tT8<N$wciD{>QZ2C6h8@+uY
z%8VxyCg-fB6pJAeP{xZU5soLM@^!qe{Jp%}JG{8wi(TSAT<1F3?r|6QpSxYP^s;RK
zPDG9=^?BR=@LR&E(q(i`T8VV#emOs!Df>=D7RO4L@n=7}|4u~aW`QWki)qOVVS&*<
zv<H0R<xReS!|70hb+Kv}1snqKT-E=P&5wFk_WF7AzI$M)LaN$|SgcbQ$vs^?iM$8F
z(VhuI<_8yCj%Hk*?agXiYgKko=650IH#JYO=J(N<?`Xxq<j@Kx1&V@z#`KeoxSaGZ
zYVpGw0^W9-Vo~YXCb679TzH<sX_Lvr1q6EBmQ#B96F4B1F>vUoMm?tu=~xAeSWZr{
z=dq<YQ?3f}qKc?^YUWjn19*OD;csP&n^$suP1v`R?{<UK;nK}ZXIG`gvjqB7G*Xg0
z404~4*9z55NC8$?dLPp<yCc$S&U!HjEfu-L&B?oc!h{M%nKT!PL}UC%u=2vs=3m);
z%&FfsqzX(ixWFM|=%oA$O&)vBx@Uc*UHd4iQ|7J68`g|*qNHFlWXDm|U7dp&1?h>0
zZIP6Pat=#k?{B1{)w4`qvRxWixf2K*TVpA4O!qR;nO)wh^i#8LBuR;lzPD-V4t_`3
zeR!CS$I!%Z2Qo)Fv2l(13?wltzx&pRI&({GpiCSGGU5B!h?`gS6<?=b8uNl@nXqI~
zL1G!w+j;sUM&TBBtN8Z0K@BYsvV&V#Po%MA#DEGejptJfFCO|x`$b&nhwp7xZR!If
z2c*c&Y)j`X6P!k0CW6)TvPo7&$bj1UqB&8S*%TyRwIu8EUHiO0TJWaFOqR_VTcT5?
zwWGt(ay=Gt*{ERVPP(yr{Gv*-z0WbFe5WapLi?V*OU@;#Me)4nvDUq_q+`RkbB7HW
zUJ8eTVPS?_j2BfME-x(KlP1=zGe%;g`DDi#UqGeF5~xa+H$w;Z+m0t+l+}gQwdv@|
z&rNDwY+;QVEvMjVYln0U#U2J`Ep(D02BJ{yQsakLVtJhaGmlAR1YK+-0%d-tatl@4
zyZt5V&H;C6Sp(8pKAZ+l3PLUV<ReOdP=|c*za~$BJ<aHslZ}7W<HWn1+gp2$Nh6E)
z)*oVXeMsM-q6OKQ_EIW<xkx2SD_12cHVw7R=Lb<+(3)LX$f<mZ9&_`ozHG9kjZntL
zJ3G>lE`JjI{3$AMH+yIv-psNPvh=vzsDXoJ73O;;!a&uOet`F}6#4Ml^3q(&T7aI%
z7Q+Cosh*XYb`$ppkxq7BW!#|0#E68*N}#Qvxhte=VGu#N$<YBS3r2z7i`;09I!06K
zQz{L-T7wMXl{*a_pW1?<Ex#6TTyQ(!+K_y+xVK8Z-SOJcR#I1Q8i*H!x6+YqdhioK
z<u+q9aKcNYTZ6Un0PZTy@$(@T&G$z244+3aF|;KP!Q9I){SU@AEGFIY@LM!dVMZd$
zTeP9f>x2EuL+3gR!R?G+iRm20WY!CtAJ5os*K3pFA6+2BrDQv`IR*?`A!zm$OF0zV
zW<@94oIgKnGj6g&jq7W7k+I(vp<lSTmt7F6i3=keMy|?<J6}+Pm1-`g;^eg4k_0%f
z$-x-KkZNzpBR+J;)rrHs!(w!$CQ>Rbboq*>dEfTd&t;F?KgobMvhX=fKYvqf3g^9;
zR;n9?@#iSz<F@LD{H^JNUZLvsxN~&a2E<@|Dfgfu%IvFC9fIi`N>fj!<Jtz$)_@74
zyFYBawJVP)$!NR~)qqLWIHu9ue5i8k7{aQ|Pz<fci0I~0;=D3r?jsnF!Tz#!{svdH
zA|0y3!YItZw}HydptlVq^X_d_qko5sKig@id03>W<Knq%yvCPpqulgn_SC)=#&$)|
z<}3LN{pf0HFtMU4va&6i^vNvVNXt+Vwk5Czupn3EEqWfhhCG2nuW7DtfcugGXEot6
zA(9x16tBHhX1!U#obPOERNAg>YOuCkGnx8D3~~xU90TBgbN3{~`?zn}Lu*P>iZp*?
z;TlZDTKdh?l0wd(w*1MSR2rgk>bJ@@Uamr{?>VlvvND@B$fl<qgr|zmZ@Bzd#5PU?
zz{4>I5lO79e)6qQ*>nAfvrt+HSY;)YToMXi$@;;kw-uK41v$XE3;Azdi{rn`{+U_7
zg$4hK?0;HI#^K=qb}dzOYGuTHs5w`P@2(~05c7lH{D>_g=PN){IsHP@f5>X<5)W@U
zdZ<;>KibZpt;OqGM1c5YAtA?<L~MBX+70~DKS8Pa^L`Ty=m4LsRv@dFph}E!R*Ilu
zh1)CGH-Y8$W=1F#z3+Cc7HG6nFyF^CZqU8<?t*}_@>)tf@kZ&`ETCbVCX6gWg=VzZ
z;EQXG<ottson6Y0h_s1UN~*lkk#tX1cRLI>%9ifxJ+)#dkF-kJj3XD~qjyYxd9~#g
zbhFQ@f1w{21HnW-!J-)o?C@v&Ie~W1&zkFf#b-la%9WWGmhsAVSI?U5o*shFANR8J
z*pC%ANO7pJ?{|0k0!Wc?#O#q2!0f#}>^q17E2=R9r_7Kl9HkG1lagX+=-eCx<DS@@
zosv~_{Ua4r62diK&kscv=K8siAHl5OK2YSI6IPf?e~HkQy#sul>?Lge7UzoWQ!wZ?
z{S`h5u>_Y+^$=Z&WVc7F86iJQ#!&Galu=8nc6kbJTTw)@Tt$$OgqLULstEQX;3{Zy
zl1-FCPjYQBvf^LD>H^h`@xR#>Jj>kt8?<w|{WtTFk{y0obu9Mz&r$OpS~(Jh*=y(r
z>1|KM#}u{x<%SrD_^gxbLaU1HPVwEt^r5bRkE6p+cNA2DWA{Cbir2$kNm<W0M$e!y
z8D=qgY;(*F&#X#5>w1~Deoa7k^4XU2Zrb*&iW~Khyri{y_ziI`+i)F(<H*KrR4s%j
z>_jP-O<_^)2fP)Ok=>jw4A}L-Tlf|2N<$LKy#{Y3Gqap!4gF$?B4;fHgjyV=SxY`*
zjWA27tNdnpRI-$@2|~AmIX@tIHb)tyT-9hyE<mc<qF=6g9_La~nkiSJ4|j-`<4`MG
zGNYP0Kci7%8V9Rd>yFC(N7_nE=2_dc!+xRiu}Mi8cE?bGmI3pYgqo^o%8l)=q5tSR
z7qv;UO6?E$q!Pk&wDu+CQI!Y(wpUinScT6ZO~R$EUi7vg6tB3AoRdv}TgQn4**ZL-
zR_*+h8JUMS7zdP7$xp08lh5l(<qBqX-j)+~<K*E1pUaG9<F!K)O0lHB4Y`1${IlJ6
z(H5KQ;67|3PMZzYZ3VxWdT?#7s5wC+)+WvUhNW_;+#~}3slyb*umeM?aD`y>8D7+P
z@zpMT<!4jD!rJ{72x{TPVEBjl9S~(MxNwy}h{~#9J3!PlFr0$Xzn(ZGnFp^xloesY
z`*QB-vg2p?Wr2pqwCKSq@Jm;TLcM3MJNT(`^rZjdw+wxkvYeCB2V0;Yt~mKK7#}Q_
zQKX)Lm?I{{yvz=K%!19u8Crsv;YPkr_5(l{F(T@YekpMFvaJlHywLCHoS_X60#^>~
zt1kgT&6`nJ+|5W3Vrd@CF(su<WKkhu5i}bF-WPcM@%914RKk)S$)xndok+Fuz@x^5
z6cuAot|h!zf?aVh(qBx?XX294Qc1*qWMtsLUsM<suZf#n`Cn+V(SJDLwr&vg{KcFk
zWVMxgf~6{#^FpPPf`wr7Vx?$nu7--yo8vNFEqDv_TiTBGq~x4ra6kD0QGz4HaSaz@
zayj|4PdfCg)Mp+|+HyAo%H3hx&V}6^hv+wt0o)%RcEo%|QTu(?8dv%FQ<U%w!#`4t
z<Zcc^<Tq$X?sfD$zCrj`o?A&x2M1b>={jVWENYMzUW?TcEpyo%2Hv_RElYL5LWHAJ
z+n!HKTN09aH-tHyYLov~Zn^%&Rm;NlpSfy>wWi|_zG+>@)bAig(f_!tkPXe<Sp^;N
zr@m=jCzn=;TTcS1$%g_w?%A;^1bDpBqS1uZ$$2JOgmHG;2#kagJM!_r3`s=#jt+c4
zZ$r`)>R3+Np@BwtE&2U90{rVSy8rn^Xr38I2XIg1Bq$&h{^P9m2j}7k5AG5_;<8=k
z*{N}2fi*Zpa1NL`XyDg!XbdIGm`sij@O$^Kb8B27%Okv#3wV2UdVZR7z|S$t!<p?5
z3_H5Yi+BeG#wAn^u|!e4JIgzI+e4&pUM#r;FNXq2eaQ6!LGvEO!@bIJB6bx_!WuQE
zGxEa_(SoA{j>+BNg@sb@11#ZKONu@c_WH9~13v8Pn86N!o7A6w)k52TM4!Ot>j!SB
z8?GsDGnffP5k`ue+{(VZEX_>w(;XF?xHtM$t7n6{GEeas7%@TBY@~;x&S!Lx3&l;Y
z#FO`7MdzL-w2O`yAFCQ74^$IHtMv_}2irnR$xuNRyDc;8OtWrT=^4a|*SNfL<-tI`
zWbv6LV%t!?^q1~utoBEsb*Py^xKlPhGd%8KgxqE}P*@t}a!*i%)IX)v#+dXfS%Vha
zBH4%%jYe{ZA4Y$d&W*<Tr&jb@3X>F-Q>R12xI!0Sy!5WX72FIzP9;NyPOZRXNqnD@
zzz-zAS%q;H`itw`kLv4ru3uWn5#^inI1%Ha#U=<HcZ^aV<VB7sT~?B=S{E{MmLSlb
z#7X1KyeaJrm{`fGsD5+crK(`}m^88V+?5%KHRJv)xB8Xx4K`{K!-V}td2Br^XdXDq
z9Kj$y1@ZNGKp9O#!D{Yi#{-KG%{0PR<f*1j$7iYy%K4JB-@jo~^I~+`b@k<6mSW5s
zGKWG}$c>w(eHUcLDBOm^LN}J%9^ktQb=P*wD)b$2z?f(Dx*|`k%SleW!ghM0GbCP;
ztgO*&n<`4}-&%d~ttde2UeOu{Huk8U1uR#JZe9SDTxIs{2*HVxu*nnOl1k9V;yw1e
zdmY2r-r6c3`HfS(&_V`Kb@NZy`x3&m3KSiVCy4pu&7`43^`~}&74PC2I<s(6H%q|9
z0*Cv>2cLQ6D27zN4EDPx!j7_pm8B?am7gtqiNDiqPB;>|4MKCli0I7gPjisXa<_uR
zbR5(<>gw6Do~;pAQ%-B!@|F!V#AfhXzp-V#5K<;GxI^Y9Rc)dJla6V(E#&qB0A@o)
zYHj_ZcIQ&G-^t&TPzO$6P~2M`+cC<<$-uE@P^^S1fM~3;4#?E`5sn3yt8S$i0HxP`
zfS;veE6#v6i1XTpd=JLI+O5}if7RlCE6qUVsfk#tn%Q2~S>d@?$-X*3WTh?LM+PRA
zsnJ#8Pc#dppv_#|TzahLQJhYnXK6OWwN~N~(a^QOdxx$p(ad>QqaHJ={)#py%0ss0
zk1`(=HzZ9kkLg8foaF5bWIp_LY-X#`Tn(wsq%zuZOwW#AA4IwF2*ofjkaw;n4=H?l
zQX74XTUY`$iSW;Eq&6Snu;MV+aJ{QNmj=-S#bmZ?HqiDetX_VY6hmp|iWLFj;p5w2
zQc5^$s|re8y1ac6#2J_JPsC!Rb(fiVO+z{z{!i3PL=YiXsEpgdg0smn*Zu2)Rv)Gu
ziVH10&7E69vce*6sNIzs+1XowCpa7O8w!>;?j=RZP}r;;aQ@RAg>10da1H)f2EWCC
zeS0tMf2pUOO#gD|!p89**VAuG!T(cFiJ<06<$}5|@>_><F?d~m4bHp~rYDoHr<+j9
zr~C{#6K{DYO(ql=?|9q(d+GOdTkpezeVTm>htQeefIkLG{~ZUAe>ek7Wt0E*Jpa+W
zic|f+k6cE^4gmp-9{j${yNB?G?Evb&y%^F+7XW_9hpx#RxA^uz(WcPltds5$q@7zT
z5&&i05HTR&y?$|htxRRkCS6-X+7weps<_dTbbg8?X^4X7;Q){fC`*Y@i<+8P*nO*%
zjP5zz_6VeHrmLz0Jq*I3pDk-c*)8mGG;F+bb6z6%UNg&COBkmqV^{7UF^g_}I$GTr
zz3kOfix#hk&jJ6&Y<|rq==ll~akR}5|0bgNcyE1cd23AjbuhQBf>IK;r0)B(E%F<A
zFhfF3`&21RT1g8{@sgAs+dDpbysm^Y|1djyrN?^v;CH>O1HMKRvFa(uQYjC0OEEJQ
z_aa3OaWk{7x^X<PB#X_3-=m4oDj-HDLBl=yQeC%>eJ>6UiQCm57=1JoE4!bqI8Ha=
z6y!gps&e?WVeay6j?1t(%Eg#QH%-+2$&D2h{v2EV^3)DAMp~WleK0LNE-ZC)YIubX
zu40SiNc!h^w^KLKSJH8`4uQM@Avq+jZ(kBk^b<rnQcAKqG&R6TuQ@^g&h{t`+L6K_
z!&Vdqx?h{|C9sZgp?nEJBUGlN{+)#82LLYlh;miSV258HF0nq?mr17t{8>vI_#i|$
zw^~rl0=nrYIV`-o)wxy3#6Zv@&)3f&-dakD(#4Lwl};vzQU|qi;9&)$pDf4VQZ5Fl
zGPwSXem^?B51>l)`GdYd$zM#y(tL9pCECSJe54gn_2rty^b0+CeYXI(2=9Obk@1=7
z5<hRyh&XR%5Z}p@<{ASpS9&dYFz#|$Om|cVENo_$j33B>QL`P`f*EQryLLc4+pLI9
znD#J;i`KG+??nQHUQ1c1?wwgsreRvp#2;~vW1<%5kNF6>e|L^VQAUk8E}h6hs_NGi
z3zpN{Vjq8UTDvhr$_428vG1Eorc^jN_P_^IBeEx`IR^@dp*inj%^B(=`RdfZ#P0s^
zi9mdPa+37V>Qi;HX88iUFg_GXp@}hDqpA!9U)jXqA?~nP6lzLKXaaNirO->R*i~>H
z0W=nbv{A?9f;qW#GRZciA7IHq&m)-zuI{F6(o6Lh8dDSCZEx#INp_KxKD|_<1Q~d>
zzl!tMq)g+~w-F&DQ}>K{VIf;4GUudqZ2|^>`P$t_a9~HR4FbR>>s_+o>>T{==F4)c
z%TSXRy=sZe13~-yv#A)x4&1L|7*)n?wq=Tt_^d28XWuH0$s3z8&a2Ys^fWBNN7y|H
z-gZ#8x_o#UUs~kj>OI+aDpY&~QrariM)?D)Lo2nLlcpnoPG_snk|bJH<Au6-oA_YZ
zuIE)gJP#qn`AqYZRYqFm(gR)!bEriyyZBU5lE~3Jw*>i<`4<h+==d@?Uz%s|x8*+<
z{regv1-yI5;U&2WIA_!R?^Q=~meoIMNY`d78hms>F7G0;$><;3nC+r8Cq{8#1&CaD
z*dNK2t>}f9S*nM1VbvWID9BRY1v4QC2_}>h_n#0wB;tbyyum1}drVsr$9|3D(^fVA
zt>l6RTVJXQDC<zW0nr6ZFZ^}-A$AT*C#f4SA3=+hZacq4(9>1m{{kpQfTKe3B;Xt|
z+0G-|8r&{<zby);2X}yhYX3a5`EkYwO%R?yJNQ@rkDB9^AI&A;kmtl%k;=5LKFeT5
zNIfzHVrM4$?W0QpS7@tEtmDHIe&l+^hp9TUV-=@0&7^VTZtxT!n==fyf9bwviFtB&
zph*@v#iTrzApA%`YL`4YT9Yhj!_fhui4oQ%WFxW#N>E#KErT72L~5lYhX`gAgn^*L
z)Yc7a`;Hg{`upFa2RXhWbN^+7_@8uo%uGzI|0UGp;QUX8df(B5t;qhXwM26n_d>eP
zTp|cDeSxT}Rxsd9beNVlQ2kmwl+tE7jy&M@0e-!EQ9RTgjf@@a#eI)IJ9)_>2KP93
zrtD{kJ9eEB0FlEu<0XYFhLp_OpshdWBF|MX<k2*og2dg`8rY*N)#iZ@hWd;UThZvr
zN2S{^E5;?78r&quhtX_R5g42;S<NxY*B2Far%L{YDWUNcsOeQ*Si!{$*JOqylfsmu
zgj3opNa0D<tW3nSHTeuHnc*fZl_<(${$^0h;snn?*`T8C7`m`R$s}IG!(u~caIej@
zvIT@Hp$%x_PL&Mb%V5<Gu#j*_r>%s{{pN@O8)aS}tHV@T1Xpp6j^#GgH|OA%j~C0*
z!y1&xf`20s(oQ%JT@nuxCYdZ_^0D#@%U&|hqEe)4YK9J-$8>=r2(>0zi5_<~heO)%
zW(}?QHRC=$z25>>W5=iW3xWQgYTdP-u)5m8mAo{Po26ZN-5UUqwmvX(7gIhXuE~Vb
zUQbiDL$GCOS<Ow4#8@&6d&Dv6l(3Y#BYxJeGN;ADRgXYMACNAYR>Er;)TWVNPtT(j
z$56q1<(anYMp{YUpqbaO@zK_UA8%UNrD5+4&KT7D^CLKdvv&b-x<I*{y(KTEE=K!P
zZx>_j<S}*c378+zsLegnyvx`hk1DIZ?lKbEPT}{sv<vzgcet|MOtcG-k&C1zl7xgd
zXDm`mOaHl`8O)083L0w`Jcgb8rk$KUPb7QVQ_rWj=iN@s-|P2x-xkL~&c>H<*H4b!
z`-q~wZ=}#*Yg^w7ACA640;U6L66P9@*aNaph~4$FJYxMQ0p10CgvJ#H$8P`Mmw@nN
z&Ot&(d5h!0z3YV4osHgFe|<UzhA$R4fgA?^9`EA>3&L=N8h+jLJlO#JPydHO+ONs=
z>r2~5`^-5ug*sav>ML%Ax*xg<C!(4Up$9%u<TtL#@ez|RH%y(H+5TJ~JFcGxe_xDw
z0tC2ojq?TYfh-;o3x@;*d|D!UJuv(v4jwQfCu2&$Mg8jFsrEt75oo=6r<B$;_;GT1
zWBs{&Iel1h_1kJm+;d|fs@I$YQ9uLohijxQhl***1Mzx*%s8lp*e#`jfg(llZg#Mw
z6%l&&$y~{mt9ooz3;G~|gV}+?xj?L%s+5d;5S#oIc~Mshpgf)e(?=is4%S}|jroJ#
z^ca3HDu$hfrE2s%?bG)q)%C3o3Vb~{<df|(cidGrHWoF52m;lH=h3V^G{u=l^#IxO
zGeVbLb9L!RD#!7UaU?rB6C#PWB+AaA=65E$4z@NnuD@S{NUV~d?S^ZlfFs<s7ablN
z2r%^U>dJoqfigJtXJ#-n56oCLm*&j9pz{+@Y8V4IlaQh}!YoA;Iz1n4J=yc{3FR>z
zA~YEF5xx@yF(q1kZ<hgO)My~-h641?F1Y&-_s;I@0=fDOA?;hQM&8amFXoT++%@#*
z`F^3FF88jH4ee6wguS{8SsWkR8{&?(*l|0$oK2jkZlvL75GC8Y+lkvJ2a1FduUt!`
zQhwP>wjNEWKul#2en)o51@Sh(L`2>oIUu8NSb=s5Fce}^LYk_;P^5-??B5}3d|b#%
z_Y09$xB1CWX+VT<=;zE}N1CAW6bXT0kpaf0H`>r(;8J;t`!SlqhEE)4%28zBoOw=K
zVCz9af_G{nVJH>8VDe<NsXuN&0S}YDMk&O7WYGW;V9B43Vx3{&{o+e?FtGVsxr0R!
zVFr|Ui{qV*e_5z@`K}22UP%{_i!u0(uBgLC;IP`_6@6!S!@xv=Yx=psKw7ZOD1FM5
znH{u|g@&l)-kZ+xljw+{j+LN@W}-&_x~1-&jG^zg^9K%)9tXy(vIA(bh!vtYn<d5M
z6VE)$CEymqd`kxjq#h*8SMi#x+tsQBKY<c4@LFyRuf$_)HaE(LG3&)fyKBTCd!}0i
z|IEVt2~mtn0rJIyf9gQMf99_z9Frg;CyO4_3hGXBKNxfbYCy}sGoq1%&c(V0#(qRh
z9N#${pxz=0#~xy;B^~l8PQO<o@p^v1NEVmNU%^PGw#c)hmD0qCn!-pnEYU|w<!98w
zo}%`H=I@FE*)f{v%|Mjt8jWb@4@G3G&G7xs5SK7Jjg$9oPtm_D&fxufDDSQ1VbOxG
z2maPSoct%$cS71M*47Gsgf+hkEmE|^R;?6GypLTJLp*w;xh^!TYM?_ZrryOCI6I|=
zNi1;#b6SC}EyJ+Co9%tm=h+w;xl5b)*t?!iU9mi&doAYW?@J*)8uSOh|AhTy6HVME
z!zQtYgs-=;SxRWhHp5Ro(~=>;WHo&AoNPjGfk$_S%$5KU=DJ-4vJmY_wXtTupWpp%
zfWP}Z2DEcW0^zz{QnYMQ6$?BWpZ0SF9r8FmAIE#?1mzy9T8|D0XE6nAr0Kf2G6~f8
zYnHuB*bqlV&UiokGK!E&y6Tfknh~KGKR`P{oWmzsn<-VBd8p4knUfBXmvJs5Z(mH_
zw33=3O-{6xCHN~x<bW<IR7maGn@&<JTEIf$N-mExL4_yUg}jVQo~Tnn-7FOq8w(*E
z3y?r5|HR@UYHm;tSS1!yyl-1aN>UNbwF*;&KparpjDy_cdYz#4YzD-?L>;O#N8RMB
zTE>z{fI#(EUv%pu`;_4OLd|bSzTqwSr9d`hNGGM)fD44t-g=6~n`#yKap#|E)!^rE
zKK)SSB3o9|VMnd{X$6LFtvlmY4cgDr+<l>4Iher=Mm}6jD}vNkQId<BT28$NFm<?b
zq1Rr(S~jy4pQ1&lBVXu{0$1My;vL2kDS;u}$DhOnNZqS@iB!+r4#gWnO=kGHW<Z#3
zmG(JFoTfOuuSnGd9JGojOF5Ri;*Mi=EpQ~Snapm`K^4>A5+1qb9r@^ryb$XaZF_HN
zcn=r4E-3N*rOI)&q{ROt72&>6V8v&iv=?g|rgAP1l5Lu8W-50!(ZV$Xj6C9qj#ms|
z1_Hk{+nIAEXL%#bN=r-)Bo*tE479BCSZ3}q>YF$ek4Obk7GJlG6pr+VF(oh7fs!CG
z;xrdR)|#Thm*}b^JC5qd5!-5YupIF<5<@RFcSoboU5kL~mq}x<;vjt`4vG>&;|eih
z4L?(po#2!%#`2^1Qx3LLLj)qoEVdnZC&x1YEQxFi<FqRd%g&F>X`}QzcoHPt##G36
zx-096+UJeoilQhS=cPjJDKy%zgS;2If+lag$^m+Fh*h-0vA3yLoJFmuJ0O$C7y9Y!
z`hnQ(x#9iR6uaD7!NKPLTnR?aqjeoMr=Pqu)7W6b>cjZdMEnB^)pSu!mV^RsqtY(q
z*-tw81~xGq3YARMeZDiD7|TJILH6j&7ypZ~cVG@}?Y4D;8QZp#8QZpP+qP}nPG)S|
zwmoCpzWHj`K4;&%*R5Jr<1dW<cC^+Wo~2TSru~H4Q3V%jCZe#7&=b~RlV7)@v}Ll}
zJ>T|kw+7p&5qfp8csN9U$6?AFrGtG~yd;o0lz>YC$L&Wy_8n!p^0J5!HG~I&bqb5M
z4LWUgtieTIALZ|uroQFNcib+ddI@k%BY)Roo@9{fWUWk)>cm0Pmp>bmIZEDL2p_zK
z$3P7KwBw*>`@j9Il(CJelNmleD>Ku-JkgEW9I*UX65~-^lN}%?kX7skpAE(f>H)IR
z&xLOC8&5%tMBc)=WyAace}X!GXwPt|LR12d7nHk`o2!S{i=7Zg#+?N=<oP-<9!=v5
zlNyd1ZW`Tq4*HrmlowVbF!@}aW*$?XhV0w3?fxZfSQ2W$VNBC-SX7ksDV_zqjdpz#
ztQe(pbZhjJf+zKhUlWJ=SKDB-{eJFs5BJI5MXKSfXU@-#!%9{5v|{;p&KO_D5N0?S
z@;D<#@*onbtXQWW`3yeqe(L-P8C*AZ*BVZ`x_Sm3B~he*P&|E$2crl&8mu)fW?Ek3
zHI4^|h+TAKb#VM`EpO#-tpIHhtoYcLmaCQX_W@ow7!bby<bSyZ11;+E;_X2(>;*yr
z`$2|;==#F#;Q3rvr4(OYQ_vtCt&eM=rUfcb`qc7`kAEtvYI5+VjwM2`e0{vpJIQu;
zHN6H30fL@zGwAL&h3QrOBCT?21k9?U$?>w3@6KQa&&tOGUa8|So?!9uv*%WMxwUL2
zr8ECz;L*qykYPy7VPu#zhjs^!eA4Wj_hRd#CNWs`6b97X=1P8qFcwlANcr;>0gZG|
z$!)*ZF-nc#|HA1Yp|$Z)?SbR)jt*=Ox+W0?w1>@JXB-pv^yGjFfQB4!ZW?IAtK3zK
zqPQc8cb=6l8fF=1vMWJ+pSHtnL<zdknR20+OG=fr!xTkCIuhaBteoK6q#!fVB|U{Y
z`r+kYK+%k`iur=LxGgA!(Z<Dal&8!6`Q(!EW2|>6aU?ioKWN+Ry*MUB5vvEu$Lgdn
ziVD*?^0Vra9;+mnvAdtz{!KLtJ>H*`0wP<TZ^f#$LmT-!=yXyEHMZd(TqV~OKmeD-
z|Gc9aW-?NO(2@eQRu|FsjyP4wu>?Eex8HSEn<6{0X^{7|GVGW*g21<X1-7KVRib^2
zb|eLdjxIs*avW?|A{I)0je>UU9}Xf@NA=PqM5ZdNv+Wh7DIT5tuqXzby@%5H7rN+)
z)?8@fvSSq$HCsZDd9={+m*nI|r8?LmaMFnHxps%-=XA048V%+@i=B&%>P}My*-Kj=
zIcTW%g(4_Tn2r-C>m_d-!9>OKc`7a)k%S3o{z6p9ruo^FXjZ|WH1+6%47xy88MPE`
zyA04Jd<i*2eY%>th7?jYM0fhI8cAfyP?}Vf#ykD_CdUTi8MO99$s+Y@0dtu{8TjOM
zKDajbh`Gjd=##rfZDeZ?{ue}~6oyu82jD=49?k|OYBANRsEo!ype*}5ve@uwDH`v|
zAoYr}9L+AutR+wH8dDyJ!WFJaA6)H~jMzWoKu8XZ5L{%$M;Yna#k4ekKA51cQ;|q$
z6Xvb6c4{X1(uagndv<L13^W3U(EZ=U%<%MK%{rzgiQVDz^}W{J$-Rw^-t8%4rSxi6
z@cmwbwT-p>`;NL1cw=N+P7WjjnhrbE$`3z&^MZqWls{Ix#+>Io^8SPkggr`mAfZ86
zC=<1t%_Lrz!WKyKMe=MlEHyZfLUy24*m|sI*+wK{B5s|jl|t6ZejjzkR7L{EV7S%G
z(WG+g(9Sv+ry|rE5ZXIL*}q@LuFJ*u`uXl@bB##3<((wC4#P@Q;v??E!9!d`?(CHz
zSpM)OmJillzGVIacMK3Ccnu4>Z;1OiFXlYYlZX-)fuEv&duycZ6w=%N>&T7gemV{O
z04{zTvSxucm+<hn4jz3eB0B)rlnlyr5l2jz^org*Uh(P0ed>M5EZBURuzEKAOh5o@
zxXou$MF-roUzXfP)Ymc=NkCkf^*|AU-8hlqY0X3vgQBu|@;Yb~*oE>%{9qTE7vr&P
zbTi*f3^6?6h$>?EagYpiSnVQwMh#{5>c^EW1U5_&ug7|$#i?Cb^j&?-F$Nr)P1K>3
z8k?Pp^OOdQhPxeo83hh$=f%81KeHCIMPfDL5_!4m9hK=9O<4O0jmXM;l~pb}<UW+Y
zvDq~7NA(s1G~R0X!AB-jbi+~B!y#dYL5Dk+*$OGwdR3~c9X+!~VYoE<N)kKkh;7Y@
zGUyv(2m=E2A-60NCXB|#(q4JVfLj&3cy!G?rvkY46>wtNbRe?WsngOoJ?Sg1@bGW*
zv-7~5qUc@xkMj4VtJ^-}CB!q_*3{VZ7B6VT3|8|`?s~1R@8(32)XC;fg%K(-)H`fP
zvek9gvxBH{O;u5E_VYsuTnoDDf*mV($gApzN*~Q@KqRz;N*&{9jqw|++qeq<$!@Rm
zR_CPTF|px?@84o%cG`|>7}}b5L|j>4ZfBQ=TN5zxr#z0EOIQgND#N$n829$~rLG>R
z_)n<gEs?r^dM6qGx0Ho}j``mSFsFV91lX+5KZ1KHTbNn$0<j-}B%mh_*h!VYt+1?c
zq9W|nH*b;$uZhcwVNEoyiK(kd!DuY9u11@e?!n#R(bwtG0@1ud@Pgjb-Geot55X%E
zF)x!x9Y8<+Z&}_HUfW`&kQ&c!F$tmW-@dM1E1|LAF`|A)kvl&Q28{r^(2+a|$GIZ*
z(Za>}>)z*%o2+9`bECY!uLkRXVSTwjeTe=;LJAhZ>Oq$bFh`kA119W7ax8QQ!MS>a
zcb}hx(m>f(cnR6vAkv-mXVLd^FHe`f+f~g|KGnm;>&|>sn>0M-i7KOdVIsBD9{|TV
ze0@04gR(E~vP&^(Gvdpn7#kT7B?8U?6m*|#Y;G#$0;e-6DZYk`ji*9Eus3ZpD24=K
zBsiR*iIr%$U3FIp=;&8e&c{WZjAFoA0?^e<k`9sreMut-#&z+Yq_mh-%rTvT6g3MD
zMmeyGIxZrb2cm^sVZ;D|;{2<&#xkTZ1BF<%Kmu^WQ{S`>CIWG)v}9i=7_m4L4mZfC
zPYO>eVhEQyzFHAfI_)vb5zBdW%+dulJYF-uTnje(P9JCSv{9+fy^v8A-ffnWM2#+U
z^->w#dS6i1bgZ7nCQCYq^k%$xf>l5(`C;A?ZaDWV9^lS7@z<L$QvEaGSPg`2fF)E5
z?Ly6rEbDhQa#b;E$hGfxD+De0OX=gsI(XIO-F@!i#uOgo+67l}92c3lS=+})x0G!6
zk=b7oN%H79kEFO2z1?V_e0A|5rMqEjHxO|YXdTXt?Ti8zpt%|3%v<D%pG?ZX4PHV|
zp=7!pFMsjXC&bjpF3t8xY)vvm_akDBHWP}2X3|S1;&l2IJ!zrSaNH!8*%QIE@P8w#
z;^cUBMZGqbLq1(MWS=!NoX+W}j>gkjkPMsvH9z%q?^G!c#bExfW(!T$ag`^b<;qB>
z{9Do~aet-bzk;1UKZU2WAytVo{Qk_6XuqUKIui-9aU*^(bSd>1)z-NxHjKh&doPGD
zR`uc@ZjZ{CvFz=YZSPPEwP;e_NF7bzf1Z4?(kxgN>})6e%&V9MWIM{MkkLhkChi`Z
z;aFaUT+G;N7!odTg@;hGD~@Ll`5XH=yWAb(hO+OcZzGw(SH<|9-Z@hkK~5Q7$XiNj
zuI!m~oGqoTP1gm{9|DC-O0-EpAhF9wjTvu93Cj#&-1L%iIw&Op$2FD^UQ)?q)JL+T
zpnaD{nIIMK(0wj&G_;SK`Q%iMXBh$oksPN8npB{2@A0>gz)zYE8Kegkv^Xeralp?M
z*p$+#FyeDbzir}LOOz@Wk063BOXRd29uQpi?Hdq*r}3HJ<bvTlxl+^ozBnrP;%_&+
zsLz>Cinh3ggSTkid>XdKU}<W^Q9tW4^Kp6cc;->+fyE1tjN4-pu&!%p?60_Y%=Hx-
ze)si9R`5j;u;M9L?f8dt1%zCRCWkJ$)3g2Nuak*$%#phaC%0zPU%6UtWq>A=j-J16
zI#CULCC@gIII9LE7Mfa!w!;mbF{-N0c$tW@YI4#cTZ)gBp0xIcLY0Azi|AIi?T_0u
z@K~u!&Hgm*M`R`<bz`H_)_QO(>Z+JEcZ4$z0-tSEea2HgRdFX6k(29{oo=maQ^{(j
zymmJ{$dE03r7M#E+Uv41>5e3^Y0jJXD28JrQhOPX&vu|SodjvFJa3F5aWwCWUEi;*
zLQ<H|KRK7a)UBOyLBu(SZqgTTBfe2E?&*FkL4hdkimLB3cF_}lib$9$_nKrs8Vg|*
zCn>r@>KD7IZop2y=%jIKn^O%d6x4S*hk%YAL1uKRaqEZTr0&dB?ewhVrZ|Mdw4J$p
zma*mJ6c)?U_RQcECh!DopC!+@s-tis44pZpwLr0$*utD>YxT7cHPu4LRVlmLJis9P
z)6on^RXS1>uFXDSFDbdSIU{QR3sgSCG}nMBmhCMK<?L~4-6oQ@%*XO9UMG%w2#yj*
z=4g7e`9A2V#wuni+qUvm^}2}=ulo72Nz&4D$EMzP=VWsF59M`GGxZa+{j`hp)j_Sx
zNPD7>h$jKNi`02VG<QDBlVC&nm~qZYB-PV~(GMu={NVA9yqJB)^&YaRk<Q)r`aT`)
z@J_@I*+;NVE&yMPV5NSVH6E%8QS2qfL7fU-b@b6YhBDU^31f`BZ|rcgk5N``H6L6s
z6;4hAv4xc>04};-*{l)&cYjA#ek~8k^%oV(9BG6qmtZ=%E}LMwNY4UBga8p+kLU5x
zOXC^Uj2x+jtw#uhqfa(tQn<w%*yw3P(&lE)VRyZ;oiNz6u4Ab*^7B#qbIloJC=qfn
z8a~fqN44&of4G=--4ifSpB&tcfz4rNbU=5KWcMd544y(2BW5OOG0bx(AwsK6{gFZR
zi^1^O4k3#>XFMW7MUMAk?*^xGVhY<t5W7JL65tbQN5mM8aPhI9)lM&>0<Bgl!G_Nh
z2$<g~x&6Vb6t~7tzp-n@-Hre=ZZ5~Q|EfhOcuG1n_in6DONLzWyjG8|uD|u0YyFd|
zA@*?67xYkQoZl)_g_b6ZqfxD-<L5Xr;{nYnSiLP$0I-3Lju>;n)T)*;o$kD(ItCMQ
zRsRU`8T?1JuphQUQ3hQmh0JC~O<M+F7^2mXbGJkhCdG{$EKVflTQJI(MQX@ZMyGt&
zBzI(67|NFV`h5&#RlZl(_dU*Z_6fRK7<eE`mxqgpL&nu5@EfpMt)cLr`Uxi1|220r
zFw(RCtGRpYhpf2Ig4q51Lsmo&soX;{gPy&zfs5OKeQO>yuN$KXJ@^^Ci6fEm60D}o
z5{7WC){xb&-_(S~b$mKTCsIq#M;2sR7tzQ%<zZ>?dUTMt;vTAgs78+V5)hGRcen3{
zJ2MIa#WTq~8WKWpY;TvR7CgQJhvp#+5BVJu-d_ie5te4K0y>c_HT)SMHBQP--r@hn
zH45g#i&~lK@%s26R$ryW$fIM?s$2*#QyvS8AdLu7u}m0+_l=J4S-wRZ)Md>nyu+4u
zz*M}#l={Z6rYElP3IBFak*fc2y0@6p08q$VDU@ntQ*@ZT9#&emhFyoKA|({A6FX=x
zKu%wLA#Qh=Q4eOiYA6^^W|mJXDl&{<pb|Ae(Gb?2U_7DR-#SP$$P!wE(MoJHPWjy?
z?RSeyqhR*t#^m;zh|M1B^I31MLd~C-p@kFp$HElr9Qpa)_dWxddP;bChd8p)zv&S6
zvMSET0jxHCd$p8>(Q*<Q$7RZ3u;ZX{)Fwg~N+5wELruJ7q%z4vGuDTxpEH8Vseg{S
zrdy(cM-wZ*3(KT5s>&S^=1yXbOc6FH0p)j(iM8R;3r7XSFYVI@{ZuAh6(Lu?SF}Jh
z5*}9lpyUyo9CzftG<iJtL+n5)k~i|VfpKEvwjW|`&0PCyAd>IfXU@nWwQkkslm>KC
z)4UABR(W%d;=%gO91LRexuo`w+N+8PLfk95bhuocO82tu2j#}_RHwEFOkPetr}orH
zsc;8blhM}w6bICN2lc2+B#TRtL@2e2kY$ya67jjX3Uon)&~_u<poBeVe4ru1n8A84
z9TIv%VrSo7*A)yF?KG!RbEck0BK|Ue(s$Etce&eCZJV$BbUQb>{_Cc3=V~t)+UYTN
zPngvchqlw|b3!ibZVmdbgW(<Crge5NBo%~}p*XM=p+Kil<D1;d>{K|f-y-Q$C7Bxt
zYfKHt%7`(#Rz6x)KM-dj=!$tp&H;rIx=-7$-P)M~6(`%6#&0)Gk`e|=?WRX1f@wY=
zm=hc|{d!<m?G<I#Ohkn*@R}vBj{`c7-mC<QwWazR5}PiV)2<i!dZ16u2|cvGycnKN
zD%etT(S*4zt)h9R$;{;SEQ6)VR`|J_Sk{H{W;1DNiL?!V2Aku(EXDSYZ!OE_HwoHG
z-7ej{qjwq~k$8zJW}rZwLuZe53NxHlG|W^jl%$3x;C!sk*rlRXz!Sn?f!uslMTgZS
z>TH!UTrX)n!nQ0np$%f-W|1&^gRScR)Q`&8LVvExZEH!!9DcnmUSPWG(aU2r5@-Ix
z!ycH(;-Y#_@G@FQEd<-%Q$+*-At2YPNeJ*-ENJurii7hiwPacGoI&2i+r{{9aTalW
zCt183FG)PHd3!zluA7TFhY1p_d0`NRs})IB1HG6OvKX3C|M$fmI8BFPXwBv_)(*0}
z`w%Q8ZMOMDU7k8cjfK-12y0&%L5q8r_0wlS*M#=kUn>f$7N#yL@d|y)w-)jy@UA5H
zLz?R&3Qnz=u8)fMRwGT7EL~9yrA3b=!E#)T&EdA;W!N$~x5`H~uDo*2B3bKKGMKV7
zGQ1K+l%`GTA$v8#Ju&6^tM3Ke%o8E?J$}sgx-T)7Ga0rB@61UCAKC_{<Wc`0$KV02
zXKj`XG(|vkmtdzC##fgw*C&0tGKHicBF*>fC&l*cIQTgbK@t+`;^EA<j*s`-idaeq
z5(RP9n6Ai?VUf)N6`nDuIHrVU^eJ;=f81S$3%wfOO_ezM6;=LqJ}T(xN`Fg~mo*2O
z;F~^rdM68gr}JA9ylu+5X1-n?*k<J5!1N0iwda1f|CPoFxj!synyKQQK}62pu>K!p
z4oLOxgo`aPOQV|k{0Baa0}n<lGq}QK*R<$F?6!in#>@RXK#qLu7;IJz$5{z8yF@R`
zchSynYV<^px#xQ?IDRw3=?8Dwy1!k!IqgrOEzL*cK6I8}?%cAfr$Op5j(7op=V6`w
zk>zx=(O%xYxZJ9}9blU02vrc&WH>BUE?ffuowX{l*h$#G9f-;Pv@0i0O89aPC9nby
zZIB}R%?U`$(5>tr)!|!#!`1Q#En(zSODV0)ob-b1E*AMnBLt=%Qr%vPXs{h7_42J8
zf|suz_Fj3?o`Q!^RDK<0ou>owcyaKQW|y*{g9UJ^k8~l77k5t-_8z<M?EgM{%U0r@
zS5MN7S>(25MJLbT<jh+1osjXPERuN;;`WR%IbPe(FDFwD0KT0nAhSg%bYP{~&0mYc
zbMq?KSA=Ow1U~D^DVV`|L)aQC2hwUS{2+)}m;c=!Z@{%C_NmRa;-St=twGJ@EH*Hn
zXh|P-e7>HGo=hY>lD&1L80*2RJUGHS_A(yUOte$gg#o@Oexv{7F^f>r-QabXiGWO=
z8vwUpZo<7`aS3>cA*fb+EK~_%8KfmcGkxPCTgTZ$Nru9@ohSi#KzChvN_9FH-N;hp
zf*d`Cl3dO$`3ulWM!WtJmpyjMbI$>H0uEI0{%z2XW}<^Um*0MoP!yyYd7_Ovwp9dj
zp%cJ}Q~Yj!YI|l3lVV=lGMAx#-&E>4u60buh8k|A^~%GT3U%)^B6e=c@zB(=^7ID$
zMJ(ie^?0vnaYG-Zq;_6R&(B(IDby>|M)A0#s*%5t%LWTy)()$M0(c=zTr>7ua#}G~
z80#CN3cwH1ugUhiXZ-XoOCV;3UT-{FkblH-JzJn`!4(oEfX~@Pn@-_(jj_%vOvlul
zUdCgzGZ5Jt{)kF;uc_t0MHdI4&7pVL;Xx(O?bC$kFEgGN+v1{vrwpi$YjVb<=OA;%
z^f?sdrOI)Zp^;bPka`s}Z<BO-xetD0P?QjXcTVGwoXt%VpW_P*i@!{NUIE|X=T-FJ
z$v9iPoVcxd``tESk)A;)mjvZryM{tkfV$GHZa$YpJcyyao9wH<$BDw?6FcCFOc#;z
z#7F5IeDm_;&~w+b<>kl$^bckDZ+X+Yr1NX@9$Qc$sX)X69~JX!zdjPB{BlAlEV`ma
z^Z9``d@rY`sU8pAE3bDqSV~wP-F?Z{?YzI;-k`#kN&g8K$;$i>M-3*%f0tc)sooWX
z#SVS-iF^-KfkJTxr-rV^c@9$D5!^AL94c9kocc@~GcSe<XZpda(n+B(x^6}s@%9p4
zgudl{?)2QfexW+UDVVL=7cgS;-30uV_lm{@$FboKa0u=6e!Znjt5LY$yep!CH86F*
z<L}ww!vOCFs%G<nqW5_KF^`#JSgUKP)iZqad(#1Kv!}#f2d`;CT;O21e%JZ)`71V=
zW}N4`fTy(VzDUhwDT#G<&_gXnsFU+{n&q*QdT74+)9uQAvz{fTse$JDqZ7)`Mbt96
z5HORrlR-b<(+U97oZgmTkzUiJa_F3f0D|WfX=S<dfrBL7-FD|X(MinEsyM1Hf~GI4
z-Sr1g`FYl}w0mNf<Xq<npQP{f_1Z0g2d)N&$0>y3-;Vd~qO0mFQ8pA)KSMmpXE|bu
z0T?Z#j>QpLg#9cnH4({bF2l}ts}h=vx5GhyciW8#zzUhBCxG>mupU;fK`*3nN(^jG
z>&$d#pLc&SnMQ-n-Uc=77=nyuj4Msdg6NBw@M!aSz|>7FWjitmDr{8Kdr;fYj#8DL
zb?scMcy*~aa9~xgo!C&9x&34wvptRQ;x`C!tfFQGY>79;AKG6wS%LGsac|=nAP}e^
z@q^0gdTmF<5~&>D9Q1H*a7L08bglp%>9aa18u-&thJSka#?^8cdbZ9%6>IXa(uMn$
zhj5e_eSmh54h)FI*a|XGG8L5`<<prfG=IYlmZn!9h15{V7@rhNSBH|_gkr9Q9Z^yd
zGdFdz%3LCs+VpT(q5_>iy|+a<{U}reCM1~7odsMc;pClZzlkK`IpwX)B)oBfJdh`l
z?~*UrRRj|S4x9XluhhDtdd$2LWdPKwBVDeFPyE{dS*ObGdP+?h8n{n!FeY(Uh|^y`
zDn=m@gik~S1}nU)V0Hm4^i?->A0xse(V7K^!>sj_ZemTPEvb_r+jJ=#u@VejD}#Yl
zJQ$_OAZ3r%*uiec=a(f2G4{u_hPE_l+%2JDN+_16)sly#6D6BQ&;vmhCn@F$l#6}S
z6Lls-VEvCQK#V9)b+1<^6=^9vl<Z|K;~Kp#*{HlpM_h_rz2nKD5kLfyMHCT;%l$U%
zL2;6sClGLq5<V?S_Fg{gc&%e%Ln<k4<^~eY6H#UGVxF&kvFn1@ys*45<D941=LL|o
z1R*QG%@NBc{*vfn-;PIj*o8LPNJSfgCHu$0r6|#e(;uZr6#{757-<lRo~0hYK>{o-
z_=yJ#ckY-&W_RYc!yxWukSk*G;(M_YZJL@+1ChIj*XM~<6`PNk_E^uLacE-@o)tUs
z(%+4j>S0lEEQk4VViR?9^6?F<TNAHRk+DWVa1ESrG(9uSs5nR6S(09oN2VXQ))m)P
zn6R6~LOWR+n1rNPOifO&3|UjZYsl--Cq!YM&?xG`t$h^Yg)io7M+u2pDeAR`g*YhO
zyBXah@*SqiX25?~3D9x%;I60|K}l3l)CNn}0#H(P>eG=(Ln(iLlgZBOq()H7xWjrT
zDqBZVcOz?Veh#5VU9bzjELud)?ygX^<iSzCkLD`c07hl~*_ip=f^W*6fmtGH=noS~
zZP<rqLA^f)N7<A|01^gA(-fwC{ksX=JXgVjpD);2oPwB!wh3H3AY^4#F~zV*ND+HO
z{*B>!{Wy`#r^@aQzgWj`WT-4#=sBi^9t;$X?5X|O`NE$>nyq-}XpRa5o8A+G#eadv
z3GJy%oDfWMW-W9u<Y)nX>XuL{u5;fwHY2{MaHV27hte|y?IN&Tx;KSkz&D8L20>tt
zOf@TrCE7wp;p|hHjNn#aOI)MM_z*%%E+nx$eSCpS5l<%*$l>EpU`iGWV(Lms?@b`1
zLE^^epA2Pe-Bqdxm4f4<4`1e`CZvCwpjrQSKs7znPcF{CF_&JdOaI4Q+V!7EgdcP1
z85H_2Glm~^zV+;efox<p`;`9i_m4Win5_hdtZoY#7+P+mijZ4ARu*~a?8vh)vt{ba
z``(Lg(GBIdo7Hv?bFZ%1*RQGQX%<{QcHpV^*rS^NlzG#UAF32IZ;mhANAm_5u)n9@
zVQ3G?Dvx^rQ^-5I)d=QF4E-k5^9QI;Pmv4mPo0M#xEoll-T3<a8WBM=%!QMeRX^jR
z7_Lsa`<1Z@J0UjMfHY&ZkXco%mgBvVv$JWlhB>jUt~z`nsoBqFwP1i~go}vi!*cTq
zuQyNkQ}$!}YoOD}7pQ{)B%`fiuvR_$R_%QYnI$Y~qrBOPd;DZgMyi-~H5L}I*MxMB
z2AiY3bGS<iBxNLPSZ{xh+DAQ8;RH<M?W=6y<8<NQAFD;Q;H#i`SU<%-BlnwShs9T7
zu>kUQI$`>2HfCl19*o1-+6+6v;X_B8d<|}Y^S(h+E*Q@sQU;H=hqd7F#ialo0y#fb
zWMJjt;VD#QvdDz&4s0E@X~lu-G9QpxOkQ~Z;5T~d2er#F+bMcf1TJwsk5z8gaA*G2
zF)_!X`tS;S8BtU-<QVkw^L!g)kM;2x@^1ONVo*cJa!Xn4cI@fL;9zi=#-NR$9P-py
z2w_(_R!1%@B{DK*61lKC+VPRL^aVOSr6l?}?98?_B@A<lt0v%N3{6PxdM0F|PRxvV
zh0mXusxAjh6W@*Xd|6a|9d)=Ky4X%+sqw#L=$0Bo&>FF}Xz5JYumGT0lM92bXM_}Z
z`hP1<#iTG%a(xHnk1yUZ!xJqiE50^evstMM(+3;`Zf4AiC=tB!)N5X~rRj?uIn!%O
z_wv~=I;>j*GJEsuvH9Lee{<2j$43J9FZwk7!FHi9eBqw)nKnps{Si4p3dEV>q1TXW
zael0?=I~lUWOSt%vcr!Q`f$U$L#W#;o0IU5RjjvA4Q&)~tCdEDBhudo3728KUZueS
zIxvvL$;V!C$q*#p5_@L)<6O2?x+mXgShdTXD)TtYajbK`93-wYa%n=E7d&0?*s^c;
zy8Eg~*;AlquiKc^dw0R66;`_Vr>NB%UK~1d9dn=Gi}@2ZK2x135N_vs`VpawN)}0S
zJcwY#Xw*M%b{e|+Ru9?-RmGt;vNk%+U^aCe1sW)&HmzZ#u$2}t3m%EM{A9%OP{Ms}
zKMEh5aH#1>pRbNz5y|dYvZo3rmmmK^YUdoPMl9Anx8*a4r?P_YZmhy+_H;Ynj9lkL
z&}u_#dtVw1N*w4OtCf??vM1p1H$UD__LP8=mEDnhqjCPsoM#9}q+CcQaMRw>!Hrm|
zrd2WWu~Tk?6v~*1j;SQ&Ow7Dpj6yR7Vs-q0quCi^MztL7PM7ePEI0djeX4Y=@bGEv
zOZRc5u?SP^AJAGn9MEGx>G}!2_)W{as#|eI81<Fc9+_r5)E>#FQ+(x+jfAmu@J4!v
zCrCmgmuX(D;YrUoPWZdB3p@$SlGb4HdL0UyExjfZ>M`}F+8fX7uAI?UI`EKz74p=5
zCAi=&yqAHX>kj}^(VYa2?_L^G()q<Zy1u8QpdUnuEJT#a<Jk?pN#eaoycX}g@-Lw=
z6#?A#Euqn$_n$@by(gVP0a0qHCIQ*|JPPAOU~KQns^`atl?YWfOcH%xO5CyML}xRQ
zbf@36(aPK2?tiMM!$g?Tp;5PjIS#to|E(cm)rX5>noaV-F|Ap261ryJOv3tn=yY_@
z8-vjv^?mIG2PQ?^+>>WaT+}X5lg}ylXy-f&gp}*eVc7Z>Q3*o;x?Ld`fGAX@6ck@P
z3?}z?D@G>C?l7T;{HgPfHSqxxt$BDl)j$qj)`85$)bPit6}`1rgg)q=v6N;*UphVc
z1ti;9Gv9Cz98>>zd1e0x7Zw{M+rO&wT|b-B|3u%+{A^19h!lu%`{DOWtwYzgP#-m~
zE*NYm!fpJ72%PD!mEhScxWbk(MhGb&zj`BUZ+TxkJwH9XeLZ+Jt-7O(x}9wIIEVPo
ze$>oyo&32UCO4IS{wjPwFJfl<>bwJMYEI$J=>hYZe@$feH$~rr`?i>TO%zRBzQKkv
zu5vbxK2*2cE!{nq#5{hK=gcM+`l_{ZWdHuD^NH2NJi3M3^^5LGzdKrzP|OUfM%U?#
z#AjCYSV-V0LI-qs8-f~HwVx~|BNkEM5XraW<K7&kIhlFJVomEy3j)<x>TSu#x{z_q
zqq@?(^9m$w#A1J{cj|NLcTR>vZG_fE76hfi^Gs|hQ%s2TO&xIasNVDa&eg_3u7a6!
zgVz!In9o8?3&R=#n=mp5SW$1Q@ytD5%{_hRvNu7b#rMa@c6j^`uy$EXNhEO5gr{_d
zSYaLVI+;<VTI@EYYi1!aVkYtXa#9Aj^>;RZ<aqKz4H!Q`IY5_%4AP*Y^7;AP0y`OO
zIuD%#d}lMj0wj4Yv<tXz0Es3xGuG!@Gw$9r)x1bXL#-xf?2V=;2ylcP<VsDbIvj-$
zGb9LvDnp1|UtScgPwtS}D|%Psi$~Vb=&R`63!_CygrA&O_Q;t`?_~oICXW<nOLr{Z
zsJ!!{&hLRjP_?jZs}2fJiJjf{2_H-GHNN?;iB+Gd%sA^Y@~S}ix&Vd>7dk_>B>g4)
z)WikQr&tm+Y+h}nhp}zV4`Zr#54xK?Cv5WZrG_*9QR<>imQw@{;}x+QVkUv&6bTAs
zuvMa-NOTCUhB2H$x6mV$m#88+$}jvYq3j&~1hVS{xExqh*(m6#^9%Qlco>UJBN5iS
zMSU^MkLZim*g$!Z2ZcLQu#F^A2VnuIq70z!5MJ2h*3p5LLv0TQWWfy*lsjVA?<;Z=
z#Uw;DmsEwERLkuAMzQRl)m0fM@AG!3B6V}@0csx*xJp6Tr-j~85UrvrJM5Mjr_oXj
zYwJC{9=3zSbaBi(g6&s?VtKGpG8BwCV%q`o%FsOGgDblUrQEPw4JyC8-&PC)K$;CR
zMrYj-RSL7^dZqmK8hzr?``qDI*kywy<y_j7huUO!dZZ1x+KK~wYt&z$Dv1M-hIH<A
z^CJzBC;tb$nkpn;qT$;ur(g>n`*05GN_XNdX{hoX-oxv@)s^;ZFFbI;Fvg4rs@fy-
z#0V(!`5DaRe!10$^F0<{nDZkJb0^Y5_f&`Iv`d$Uq&xpI)M8eUnCfLYgyih1$DxWL
zL&cg{?&ho8TU_$X@$F48S{2gC0zy%NsfmNjNm6{070q&Vs$&w{T2n_%U06-{nzk@|
zb@ghpAbn=>+?p&DP5$YkgZyML2%9$#`t<a`p)DJchj}X_wh;00S&a1<JLC|GzdEgd
zA^C&yVCrIwJaVRJdzx)K)QYCMll)Q??#fd$D$RRO<eN!CS6P-|GYL;7&Nw!br;{44
zl&-YBjY^uZqvf>EhefE*)A7HW^u#JSd_LEMwcZ_o#)u}wa4fpDD&F;5GM_hR!~}=U
z^4YcN<U5A+;MEwRIJco@aBGdh+pH|5OD=A2QRG;JYIBi>^A*;~n!m0<od#K22<P=z
zJWU!F<q_XAiU#XUfe@0S8KY5zOukG~nr2_voUIygypJshW7@0%zL=2JytNe5_$COf
z$wLH2P~vK@fekP_6$vJh-*dt&$N{s<)&AtGh)@>qr#2KTf<*16;t)sj$2%*01H(Yg
z1UGJi#8lCavA;%6vq#?;D@line@-`K=bbS)WR@SdTka*t&mmxTp;z{z>7DNtQPs-{
zi0gsRZv#PQ3M*@>au|K;R;5`GwO`#8qw)iv$kj~cdR!UKG^*TzuTK7&9fmh)HUvR;
z3Cw&woE0cRrvIuI0FB2O%V)9z3G5X;v?JGAmk&=MYTgI<W0-C}8v#5#`L4O?e`#m=
zeuG|lj8&Pu@d`c~Q5MRzqGevwYFI(NsEDu9UP%&RPQ;eNKx1590XDA{tqdECZW`)s
z#|n?5J2Uq?yUzi<NzM)5tK1h&aNV6*Vst5D08iJxs2V{Hk@RCaV4z+fG`N3f6&oz-
zHG9&2${ye;y?YHim(8Me^ujzUkQ8N7F#1-y>L;0F>cKX68SX6Nc%|kjaGM+}cN@NK
zM8@Vc`u;PU8FUA}vhm=r4xK2O>LY&BfvaWe8+6BdHpQ~MC<O(5*@B(aGah(yIH%S;
zDxFOSo93<D{kJM+a0qQZ{q{3r&X_$CAUS1szT{ji33IJ&Nx<;ZY-_H+zEX9Dd_EVI
zqieSq$VjPG@zZ-XN!<1&8&S}~`YXfUYt#2mtqvIHkRz-0MmG(CLzdXemk1;ybf_J9
zu<+9bP@p#DmaVSK6I~iy)Y&`u1#a=iFBbVXd|wC!{pmM6ara&Bf0(h@=>LcN?Z4H*
z3~c}IN;#!A6|*7oqk5uv07{4b(O-h9q2G8k0D4%~wiJ{So>GJ!eZn1nh%b+(Tt2Q(
z@uw8O67eSBiLdSE_QC41%-Ck`=-|L?{{oF2*mwqX>OG@X^QPc(10aG8X~7%Fm0d14
zd8|ZJKhM(9`3CE{e@YC@WAuBRm-o9QlZ{(G!45U9N~|9~{k?gP^Uy7M?d;uXumD?k
z^|#r#>$AH~S<SMbsONN<*V|{~qXR=&)KqN`{{lu0{p+hv9Qc>j;0WNHP~Prw%kKM8
ztj~&zit7v?h+JMJ_&qc^@&Gm{f@#(>4EfSbJ*9Lxad8Lb5F~O%e-VByt)aP$hbfBE
z^2Al-yyy^zd1VQ`2NHy+;pG{I2g_kYVti)|JhS7jWW1zx8K|{w!$CPvS5mU@-f#6G
zeF9(&29^iH8%J#?ExX3PU+3@2PfZZxokItL>R!482i6!^OJjOGrRviDMX+;B2_LcA
z?lL~9_%nv?Jq9^Uyc8NPV|1rIP8^Et?!OqPysYz5vKMy}OP~_{n9U3qCdEOVI0~<Z
z6!bz<&j3aV@;m@eQLsC9P47a7HfnTbGJgGn@jNlkv$(q`xI%o+8rl`tEZ_iW|7whg
zWNz<*qHHpRC~yr&jg%1;U4AHEbr2HF5!I~#cW5xLJJfnITUm_}j%vAv-D4v2JnbgL
zQV}%aN)E86v%mOFj1<r!QBO_Kh#SFY;WrD>;F)~%2fgJYXE_!p+M7hV!30C4#|K)u
z;;h{Fg^h~rtvVtRrzH1ctAX`w3F(;~c=~z1cVZCG@JKPNvW{iMcg^|f==@|2Dx3b$
zHsK4o<zb#eB&IMFJqnz9_Fc<GF;C&{jhKb;sG8m(dMHlBj^#T<NR*#Qx9vzLygFJ3
zcOv`~pfz>!^)XLT3b@T@sV4Dz*+F+I`usIEC@Sci_1l?2EB9N5;e9PjY#W?buGLV6
z7!m1}$dQl$9h__}q1gRcxGm;4Cf;p^RtwwiFV?UGm76UHI0-!&C@`v63BBn;dzu_p
z@lz@>Qjj*v>}>z?JGy}PoL|f0j>P_tV}CT`GscAaq1MhAb%?*CyAmBS*-}Ib_NeC&
z_F$6Tl#^rJ5N^zmDler=Qg6-c_8`x*q<Dlvn1ncSs@GqC`3skZhof>{ugJ5BmX_^W
zBNly2F5bba(WPGOUp8hf8)b=BjKcpZ#i_fJKCPQox?R`O$aT<i#c-_4nAz|5x;`Sb
zF2|9d5zfoxR6j0|Y!!X~)hAg-#*HSxhIWJ@h6XoLdjuHDm$LWg=h)skj*ERK@8!p1
zvA#d?lIM1yL7&3wcRA93*SiA;Imd=743;qkNv+;^NXG83$dRNIYkvIL`aJDM{q(-X
zAZ$z}TbEDB9HKFHP>No<MZggi3i5<qfn@w<EF7ZP3<<SRX1=WCEdP#q^I85Zf>)1f
zxt_R1_7;aEyh;60q@Y!Cdn)#_g{-6@neZYCvQ%?b9U<`mr8NCo@f3B6rQC8B!!1nt
zGqd2WD*8R2u_*=#JTiOUvNHZnGuAvOin;XY=T(PtrrN58JHz4Y2;zgYzyU};<clHd
zu|fmic+SkqskTN)$Og&dbuJe6!~c<w+nIy?6_K|krj#Z|)O+$aH%fVREruhnPn@0$
zxsvPPY#=A4a*c#Vcf9(PUe|btQMj^H09Wku6eYd2B7CI^E9$WX_ymjKn34L2vdT(N
z_kYD7|9_@pc6O%!<0)ohW&T&=&X$Je5Bvkg=eb4=pVl0Z`gqf18rdI^ta{-Wq~0pc
z>@ppYMgdZB{U3#--H&%4(@~USxd`+4=A>Ra<_=EBtCjRIl0zEjwxW3oSN4_lw!L(T
z5w}#RR2i{780Q2-O8e5LuJ~%Z7=M?gb;|RwO*}m%PK6_YI68(%1brnexvSA4mkegZ
z1LW(X!b!kXl)_CF0G&7!shIfXUVoHgL_6?pWyns-yp%w;YIOI?<VAW#KGZWc>TXC^
zt7zW72t7oi%T+9+x|j;EsSFPN@!tHu8GV@?W(d`NG`~oI4g_wMJNZM>e&9Vs>VE;`
z#+tl`It1tyE(e>uA7v83@GP<w_6G!@qUWIrB0*TV$`$xn&K4WH23E8JOpwS6Pr-yV
zG`yOo|F}h2m;%6H&t{qLGa!jD9{}B$AX-AXf^Lm~UrRu|@&g_r-ZN}3mBd2(7h}dT
z`UnM-tCg(z75dp(B2wxNxAhG$%vj^XFClB80#rpSfR`4L+OKN+9}cbnRamg~0x@AB
z`zgy+^aAPw8=k<3D55?9`jwhw&}1vZV9WVQGRmL_p6^*+ZDt&LoO<cE^PMgQ%#Ul6
z)b#pOB<nw?-{w0cLiwru{1Y%lP~>_|5-@FkdF6;o{x{{)j1ZRkE&fCz1i)Z^8B`ww
z`ZG&{n74}DC-=WbnVEoXI3y+sIMsMb(3PO|ug5F^fb1ls`(Y1}isPyWLg=h8^(ILW
zeQ%aj3rA@Ar2{f~u|CgtzWe&;ITyz+yn~suvheVFJYH^|O%k5kbaK|ByVG%HO=o0B
zz!GH)_*p$Ta%RXH5@XO}eSLYqpI*JVa-AC8ZjC$`z+%Yk*k-$X?|z~3H4DA|q9hSf
z>m3*fEi7#~?fYS19G;x+lNF*RKHrnX)B12{1FXp+&Dt^KW=+lzQn~@(S~;+A=i*3B
zqxuu-eA2OVVfu1=e>21xu^}>jLY|wdVFYMQdN?^SSo`HYhkW%-pxy=Cq28qcV}No%
z{|2q00k`xMnVHd@**E__Iq2CsIX}IyW3c>Hq*~}3)s5wSpMVNCLkf1EnpDR65_LEa
zMx$~{#RM*U5zT*D2>UlTwmZ8!Ee6+Bx8J*sw2~a64gz)PD>W33R=gC!o?B<MqY#jm
z@dmO0;6_yJ2t?zahrY|hE;xFw5b6Z%zVH;{<n_4Wdha`xcjma%)E6t-ha=<L>#LnN
ztsCdl;p@ri+V0Ng;qw(v_FXqdv?jK+M$|CpfS&JO-@VZNLZ=D$F2C$Oe;!MS`iJ@f
zD`MIZIQKG@iAQ!VYpV7+#4~9N#|z@c)5={Kw#1!fqNl|5EJorwb#i)fVzk{_RNpe=
z&WWr3cUglbP3Y_#;q^M1rTAsLyVYuh0!S3k-e!y(qw&<Wdg?nn7EU$GVOq{SAEgw*
zuuN-Ejd*}O$_FWkRvJCP&gGm+VF8p8>QUan+pVNeMK0TlA+c$t5kiU|UUAE;L|oQ5
zF`d|t0&B{schW0YamipQS&8J3rA)#&xxAI;KE6nwj+hoBAr*Q|GUR}yLre#}02PEH
z0ksgCzXhOXl3cmXwV~<2nZ?}aWo27(ozjxR6h98?TQy!s>U#Yfu7u3=ehr=A(+XjM
zguofWBlNsrYXp*0fw5e=dRf#p2)4B4L^Oe))L448=wwZ9xg=>;XfcWSoLS&oj1s*M
zz<Ven0+U9TRbh9o7UHcwhz5>5ZL*n6KmlkT3bTms-J=1&U{)k=h}Vg*AVS6pj45do
z>cbshpy8pX6b8|OiQ(i_ZhJtUZ&HPZ4}-l=1q7zRyymr<`|p;U3zLsGm-;I2dP<jf
zkJggV32F~=`|ijIeT{gastGkvX91e{NfHm2qQxaqzqcEyFkyjoZDHY=Hm^9D6GdTE
zj#{Iw)e13gv@~+bVIYl*x=nm=W2G@clcsTg`?x?tYPhiT`(Q!1hdGu&&^++%7RZo9
zI8TeL#L8fM0Ypx4v-T!i8HDwrHhHU%%4@O0wSnuU)<`LDj<n2l%#rAI%u&xB;m~5&
z=bkE#)cq96(Q(sSY?G9x?ct?LmDf?S@wy!9+W-7u#|iZXrOLAVBDjd?m6L>%IRwjR
zEyoEZlnSzEl!v&CkCmEb<t#a&ZnAIgn;YVsxfMscQQEMeao5^awx)xBL4==;ht9q{
z{=M;GE*zVUx+>zrN}>r;12mQtZWF=H9P4^}6M>a$uvAfvu3CYvx|TVHj*R4$OqNX8
zru2144m8%eB#~E85_ztoAX7T{?ZRxJu~tM}7tQIoQN;LtuyvZ;`caqPI;ufV<CH0<
zbw`P8su{Azr3BRm4U7jW$f-G>%YrO<$vzjI*~zSK;$&;mRB(kKb}ow-(H3N=^4<`#
zS3Iss-Yi>v%gaX`Qpmr1%DPb)%v;OtVcriGaO2GQt;gRwuL1w$k%_NG>um8eh5y&b
zlQjPS`hNWgbK2nIwp{HQeloM{5u@`Us^i-VpCj*2w%xmYg!v>V!MCzd^m)QB^h(Qc
zd<y`(@CBM`)NP0hJhFfz|6*|%7>b*N37G*Z=pe|qshT~I5<;^x5ga*U*ACE2<ZEBF
z9rE;5f)WZj(E2oRfagyl8e+D=Cs3@<U&>HQ2Bg-=P{wmUUcud@Q1z^eLauGF#^W5+
ztZtdqlNwmKau!{3g$saC)I_d_mw7mg)Lq{IoEV~UcA>HhsmXILri>H~Kc=n*Rh*~b
zp@r#LC@(Q6UFW#Q30a=*kBg$`7Q`JGSnG;`508Ac5%V|6+N=H<Os^9}ZW9HZh~d@X
zMM6lDZ-b084W<`@2{#A<7a|bA*ipd4vv019L2l>JU@;@yX?yR&!J90Z7(m_Y4AtGI
zj!ls#nU9b+NYm>UMFSZ$nCU-TYA-91=2*>Oq0Xt!RM+!w9-C|x4j^tClM*&7l-6_j
zHPAUHMNaol$|&FUn8BKl)K{Ye=w5R!qN}CRuEokxthpH7-Kx2CHr!gZT;9=I*^GUl
zFBjqu7uep>A&()!S9wZEL;|{0T>6N_vSgIeg~^~SvMkYlIQ2)Ox4m7;kXfqldS!_`
znfv47z|Eb)YG{X?g>c)nL1vD@Sd-VG3?x8a$*r2SjpLweE4~}^rn>rJ5C^Eu)HM|z
zkl;$8rL#D>b;iESU97n5X_Vpl@-85~Aa6=jr-MqqN~`wFj-vQ|m?CmG=VnQfNp};V
z%bjfRE-2Co70c36{A@j4eIsSl;piK1&{CD=ADTio#{bb2{`aiS%=+(U?XLeBmP@2i
zfXjz~FHQFU4a+?<9=j{FU&#*K7iuQIBYKTjIJhy48<_YHoBpumK`)r7maWqp&ks-R
z>(^Ld2q+YORKV`#@y$wMlG%&Lo}^?MvTjNbXZOM@zyCg~-^*Spz%f__!Q`$ha!K2G
z5qrc=Kh=)^bt>KTeHRmh*o}Gs&-drkh`5Py@eQdLOv~5GO3Ok|sl2mNKonk-ynqJl
zPqZyEt{J|5C0;Lh&E(Db#?9x0DQy$TFd#5dG(iSHpa?vB%2hCAEF#@th?z>pWkLpE
zJT&5~k>4zr^T*4h;faWjZuwP^0LPYO=&_yjUvyJ+iqQ%m5dh{BghxnM6)DEQ9BjR&
z+7sx$E4%O1U07L2R~b8rD_bmip!suGwSY~rJf|i>dIXw}QO#QQJd!q=Fh__85)e0Z
z1EV2-n_>aO5r-SY<rnpxna_h4XKP4lHyW^SHW+cnkhe+@R34&;Tp|=_6SXj*p6n?}
zC*9=OL8asEL#~G<rMk7e?%ovY^>E~6C=aaAn+~Dw&yotL3m<oxNKJznjuPk7xF8Fc
zOfV;mlBVs>Hoh8<dO<}RHLa<ts!%}la{bw%%cNnl{a|~Rq*h3Sc@r3fM9erj(%f~<
zM-o_u=8lsX=t3;iKFiv9g2D#;Aa!NpCCsKoT7ah2_z_LsMzu7n2__Y}%@c}*U8WSE
zlVP1`Uq1&it6<#Tvo<HziOluwN<Jlwsw^yU3{5;KuHm6lP5Z<dH8-%2u$;2=2--P+
zWv(vjW;z&M-$8XKi}5B>+EQN$C4-OHHmB5_3}u0ZAMInUoYdV-7|Z^-xID*T;_;!h
zO>hWcy8Jsu%SzjV$qYma$$TnGnjf4_JWdG@xCXgSsP0KE_ZlCNlNgldL0P<2y|jFe
zhq+Sa(iZ&)_nE$En5;@p>jIiKEE0QJNuOaj=_r3{0bOu237yGY(z7G)$i!MA#;NVB
zMSbTVX`*4(WSXFfX#9?mNqscu9EK%f5*WNp-6Upars3)~cn}{}IHl!;mCGlhYEmMR
zT*)P!m2)VvJ!7RJKwAl1kJ0eTD$PA&wNkyRm+-(6?>`+rmxBl;+$?bsl7>M!mOp8G
zU}9}=XdlY|Bz%T3Jg3rJmQ!S(%_hn$PI~{@rZSS@RY<)uMyuUj#_)bxY1Bj^G&s4E
z<<r}xef!vS#Fyn<wk@%kw}_#UNjRQ5Z0b-H;zNB{_4g&mik;50ccs;4-Ok;|Pe9P5
z-o161L!nurIJPs;2Em_wbdpKq#BQCuAEjh;t$7{A4RL(>LiUwmJ;S<kBbvUw8YEGT
z|1-PM%ZtDm>(@1NY(?Sbc^IZAiD{v63+awl1Gm*HCaT9hi_a}|>=xBR_NC{N8pYIP
z*>%}4o+1Y&+sZpmv1-9&&_bci@s83$JkIIl#mfk~{^?CZb*qqvtJh+Y*_sod*fT|w
zIjEntP)(4Kn`uLxEmlSFLUU?oTeLDk@jFI`X6QzJFzgyhtJN-8r6ieo_9DnR5cjPU
z#l8rIcy?503>RZ;niukH6=G!!Si$Z#<FdpBJCPo#84M*<=f&v#`bDUyV?c$@Y|3?8
znabwav!o%lfhI&i=2$#p;bO=1QC8A#e!eZE&zdi!I_aDD82*ffLF2REvTfG>gwoQQ
zj2Vq46VWRN=moTk)!e+5X{DqlG4FrgS>E7KzNX5H!Gm0y18GHf&Zcv-YG-HiCBO<X
zQgLN8J=3V0nCzS!bLq7Wj_sq4%0a`j<2-oiEmYsc)a<59QkEfrGbL=OM!RtQ&wqIt
zFow*7t6cQjKxaj1&U6Jc9~-UTNmrP<*I18uc{9s6x;7c~ZbdKIiBUEGv8(z*%4S8@
zF(2r^1?{jazp*_V0AN1G{3vNakxO{*kOfaOCu14-uHDCA8ht(T6l{($ZS5||O80p|
zh+3)Remmg>G_El_>uBZ-B|nC9zv3E&^0i6gU|nO}`DR_yFWr?W;7jXks{LUo5keMe
zERXB6y8xztLE$``ghO9(h^{as>8+gJhb#AD>QCnQ{`y&1jQU3QriuR3!}H@}{XY?B
z|NZc=vHr`LOO3IZ%^xAnv+4_2ff#ZVU@%ZS<tgY93;zzNBPKEXA>HfufLsN&2gS-f
zcW6Y17V2wjm{$@}Xl@|3POgltY_2X7tc+VBHnc4-s4%bCo2q!b`PAPeM3su&%axlI
zeu>%#_BpWP#SEU9pD~JgL{`JyLjdD#wnhSnmtDyohP#lUW{IU{?g2ULwEW}^`+DIG
zzbXn*AYL!Vx99Vg7;1Hfmfwb;2^M6mS`0><C9dX0AlLCVbCyt?H3)#fm(?hKZl7C_
zJeq3fy!|-n=sr?4lROFmRn*rze`$`F;a#%KGajc<Qnl~@VeFlPYYW>g%-GnmZQHhO
z+qRP(+qRP(+qP}nwv*23?&|95^LN!hFV?)BRjbze-uaGaj6nmA)QGB&1N`!dm|d+M
zgb^c($GAvb!;U{jCr^<_flauI9`Ej9veVto^YyXF4Wx;V1$F?Vb9FmYQ~C-_*}vPb
zUuAFl`sMR__eh4ONjP0Yp_BMgOrpg35{NSV+9#NFjh73ub~UzyJg(#@q7-9HfnkIK
zRVt`N+u-OdvM1$$KVSbmbCV0)@o?pmWUb)ttX=<($hUQEPL+kauHH>}&G1-w$Kqj_
z_XI<><*v~zZ@E{g*?bFjQA97`R0h2)ArMX+9L3mcT(*^MSH(N2Gys-QY)><HuI#_R
z)F9cAV7v{*g4DQAiNx=&sFfq=)#Vk}Tz?$UY-m?G%Xy1Z_B{CR)lDF{=7t@(KFL%4
zvrpQIi8fdr<Vr@Hm%#J@bxU(Ik<#L`cCbeiJnSyQrPM;qPKPp*B1gs3tSS!G_*mF8
ziEt<Xr%f?wior##&~i+)634ArYe{bkK{>hP!HqrYdAU4poaeq?uc=rEH)Ha3lH%&^
z-3#sDptQ&V_atHqFPTUBYscu^sCLb*NdU0V0WamM`PjoInVw$Dco=9k%HvqprRk@i
zlwNt!i2^n)Cb^^*47mwKU?xscQB_UVJ3qvcKs=M3A|9dHz*ewGEI331mF;&0Q^8{c
zSq`&Z(k_!pe|0^i*WFd^RN1bF4^b#q*&I~y%);ebn7uhCFY$t0No6r$N#DXj=xHb%
zs~@IDiX@IKU|9MSD2`8QDTB)aNz8j5;K>dxoM2Ur2Rey}pjqM>Ccg5L=P7*;S3aVu
zE%mgxVghYpT5DthGZbv?w`HhL_Bv-=^IoI`RFW6Uz$~+70<F40gHLg%oo)<;Xid}r
zC?VixHh`H|U(~1-CL+DDSdwD7#i$8D>;;6>ENJ(Ke-(11%b(Mdg1gRy^la{K**_z8
zU(<Q&Cr4PD_|*j8V19DtL%h+czZ_$j1raCQ#2Rt3zzu&6KbX;!tU84kW@Z1Dy-!PH
z)Z<V|bz2~%$<46R<wRJDEwfrxF#l{tXN^7C0VwOrgt&qpXN)>wmj&n&zPM>QrjV?H
z)R^1$YVyI9*_RQSofp}7Cg4tZh<yL?Zv8Y1vY`XZI8q*gZgBh90nNCez|DF2aZ}Q-
z^N$k_1XCcr;tp+>WFb>B+y-a%rTsQ3Pjk|>Y{R35%(V?8Uf8udX@zx*Mn=bIWX+@#
zhQBMc+WWI-L7{zu)eLQFsKkZ`?Hy0*XR5)#l#PAzcO3_mawInJ#~ejDKe`DUPs(E3
zISF^F0_R8N)#ovu+TYjd7Mi0BY(2zR<`VDkB{zRN#1bK|UjxCd*r-aCLSk6+fIk9-
ztHV>GC#nrQ9a;08VCZp~#gW%YyaJI4QU-WTRE8*e+(UwXJ=(dfMEsi~_vyGr0}fHv
zs5Ux=XZtY+FW%!#204K`p_b>0v<FhJ5zp)ygDgOGmq(r=S{&eKUOp0Gwo^lXVuYyt
z>uo*~b-FojF8qk9A~Zu>DWt!=iF1{h+MXw-IEx-An4i&Q&?*A39sm;mmOpU}$x0HM
z-_o?YRi(5AE@|RxTP5R=(MoIvC)oDCFIEdI5dU3P8=3jcu=MN~Y`8!$0-seMc59rP
zCsyndj*>@6v=;`RweofxD`T(vrJwn!giX=bgyAVe5txjEfXxD&Iz7AY!HIBrhH#FZ
z=Gf>y$f{#1H<LS7(v0UgQ&3tACk&gpyJEId_b1SfA!|+p^&ORDLW`#QMid>=*9+^c
zBf~88vCHY1E9TB={1eWeTQp=MG|LBhPuRHG<0b=Jz0X(>h&r*;`*uU(=LMrUN8F!3
zy5RyKX=@w0sY&fVUePlwJl4&Nk&_F67y;OaTnXfHgDyP0al#aokbriC3K9nt#cQ}B
zMPaU(5A-3-3*D~!vULtI7C-p!g<+?Im@f>EVu0!HX2hwjG9`U(vzd<LO9vbKY>#<O
zqo|vMcGqVg0?_as8@!apL%0uE%j?Ryk+EmtuZkT_DU2D4j@ZL5)kjr&$qdaEib<Kg
zt?tx-%SA>5+gZ?$E=`9LB>$QiHK?*#gbuS=GMTR`*B+~>LdN{(gYJ;Yrm{L43RPt)
ztxz!%t2G`kZ=R&OIB|)%vT}jHYNAy<XaM5;J4$Qn1-&iY{GXJO`~lrZax4_gTT4vt
zsgj?603s)Bum98O%=*8Z9vSJ`|64uae~hC47yg7EO@0mu0+?Xk_5aTF|M4S#s;3!q
z-49Ph<snY5Bk6t&FA~4I_8RG^p*x99ZNW7$OH<Joez~gx_Ut>uV&_>bX9AEWa>nCp
z-+|A9mdR!EAwCIvWO!^T)^qU|sbvO2_Bo~Ly;~B$bux#EU`KPw3S^elK1DlyIm}(~
z9$|Sw)-K44_4EFG?@z3?>pHMXuq48mn@*rCs1#>393*i_9n=Va-29PIB=@8UPwXr@
z`PMjmU)EATDnVXwY@A3Pjh05d1gZ@Ld7vU{HIA8(3{bkEff%0HkAZ>@XCv+=y%Ybm
za0#$UU)b*aKCyb*p~Iu|>w&R^HkRy=4KztPe18~;^ds;$rmsc{;E3Cr>1ns4>seA>
zcwg!P^zIwgPNWTv4>pxtlZ74v*!?CchCE@PiA_B|b)X1Ox=h3&cEIiS^}Q)4yU_zz
zf|y@hD(FM|p}6IYLQ5cLX5v0K(Iv<P=@Blr|HCXL0snqbaV%il(%rnxEMW5kLlo@L
z%qF^PhAOJdss&PIx|xP$-H%}q#&m4~fl`Jlw;k%U{TC?Pc>H`gCNUT;XFwVIq_*|+
z32`#nWsmI+|AzF)pi!%~n-k2hbaW<vHq>~o0q-a`4YQWAroj)(^c}_3;}YzKK0BTm
za&u83Y#^*0U+zZDEzM!<KEuNUYZKN^*2DS`>1Bqo*d+?Ca9W{9p%gK?A}4K>vCOzj
zb;E6>yhX+L1#`<jK3r2+D-h2pjWje3XJwJsp(;ct?ZDI^GOTjL29P*grG2lXePyLE
z3SSb;DuTk6`JHmcZ#6OPHRc@*`2})~&;vcc5@kDccBG*H6h5Q0kBs_M#=C7Gqvr&9
zSzFE-4FO9FXvDnX4}E>4vWsE8zF;Bx(3xg4B8Qi{0DE8D28V1P;#cW?tkrqGsAyg7
zamJtfWu;HuhgwB?+pFE7E{zAc-Lb{5zw=WbP>w@EwhPEmIE^yP4@u}z{lC$BHSYI0
zj;U|-$xdjXF&O*FXhbAo7#X?>fXVr2^#sKj1r-$2>=#x3Skytr%>~r2)`UWH$@%e5
zr)$PXmq?I619TE47o=ULa%T`S=4`8piqJxbfLymxhV=|S_r6=lC~5-Zl&SjVqSFMe
za=Yv5vWzsCIOAE)v`b1SS7#KMV27<%WBD2@GRnn9a4s?^_Pj?b2?9o)|59e;gD`of
zx9JH=F9sVF6zRLrh(YVH=pEN~!Ww-W>vC;|5-QjZYzVAy1F^%~%F4D4XUcF?7`Mzv
zK8=TB1Qk08Osu$uA3soY<@RzV$}FMsJ9nihHsCG#0gyU1wElWrAV3tOPGw{zeAxuF
zV8GXylS)+XtZw+Ad438Mx=@qRcw+U#A(iyO`}KD7=QjQ`yNOJ=(ETVVYB$ic4#%{>
zOX|$)p+i0DQc%zl8~ghsd6SI*)Y!PrjYUAR()CACQAc33)F~+e_@9RdF+ByQ2X7+Q
z5p~JF7a4Ao{yIwqfTJ!}Rk~92HP^wPDwCqCn9|sYecq;3_0Vh=w$==RAvnRbqmuL^
z96-e2JR+jS2wIR)?DFzCEuf<D2_G?;MRd+<)(INz^5!swi1hW-`2;Cj^l{U%uW{HV
z1Hs!@6l2;PY4~)E1mX&NXfGeMNh*;OrA?%V-=1iOV=0jugKq3-J<%vHr#$oy%$Hpe
zNm8<8u{J(^aj&yy>N#a(JUH~75Lgizwv<`#vq_{aaQyBV(mGXZNc_#Yf5Wu}GcF6S
z9zjG&AsAMe(=QzTeO)QaH!eW{kHgfS{ci=tdk{1{qKM+)6ZY}1Q(YQV8KszId<Pt*
zB~t+a)0<I?GG&HN7^<`N_%u=~p`bYPwM&ebuvZ58aTd`(L(ldk>#$dQmcz@>^hYsm
zza3g57DNY<_K>6<(iz4o59ZrhbIQ;-L9nKY#$ZLxukCx+4oKg(X3@4$*@SBJmVR0>
z!eWpYwI$50!86X;pz2_Po0~2nB`+D<G|t)ir<^0z2PxG~mxnqrHTLLG8j(B{oX6tp
zZ^^RIGLb{58HGX#y&2<zXbGnYNX`qHfJJAPJIMY7Cd^BdU(K_-63cFLoqbACCwtBr
zh90;zmrJO$@3v^!nN&yKx_<%HiJYrx$xLz{YZ>mYATP>`66#@&QaubXYyrvIu?MUp
zbj)hV_+V_FmL@=!$ui#s-YJTvX8U1YsiXX3bP^`&ExDs~L+W#N<Z`xSAOmv1{Ax+*
z%_o@4MUgeOzyC-|gTRR1EC3n%l5T?}F(4F=mJe`-8#p#BePVG-(^?l%Q9FDtZWTK~
zU)vV#Z|S_tsd0(R=~&sOpUi39qM6+`W_shyXRiOee~E6}>pmM#_8uisgpa9YlT@)S
zzF>M|2p$wA42*2DA82QX-YSMbb9^qzp>D&?jtYQDZ5>)9$=)ev-hkt#y__!jP@bMz
zNEZ%tbI5gGI)n~PPiwZ{A5Wl!UO7P><*8l|W1-}k&uD)A!$QfpXPVRgZ0vq~OZq_O
z=S5E5ysOn!*u3Kv5qTT;iLiW3yjB8`9~AmJloc6znpX7?xypZ`upE17ll%RR^~fny
zZ0zW;j#Wi!w!MMP88&dTL|VN)gH|F*_ch%7BuTqjRwfJM&8kshXj@by2i}~Pt7yMi
zl&c7@cv=2R#zVOTkC*5^C5O1mWZx~8f*r^~vl^x*1|?v!xjp%bTG~UDjbaxSDtXFH
z83`IRfJ;3|mS4coEaMRWY2;x0-y;VDGuwYjTx*Q|kA1T1gX$byGoD-#S|8oKqXRr`
z1H=bw(jx9KY2fe&UPwZjhH_=!b|lnsEom@YE~&BuFZ=qNJ!=PEZ=2dx7nBjV{pBWm
z4{$11lxa@kBwArQ<NC|&<we2D{xj=MpsMO<<l&{)?zf&3$vuR?*L^W<;r~lpx9q(*
z6+FEE%a+A1u)i(!+y3D;S5U1uonLvh^y})z`{sHq4^%lcJgHR8%l~*eA%M*zIP$)v
zFW0wzxzKv~x>cKKN)>6w;`RMH#T-|2(51iRzQUdx^w>pKIL(S2FUjBMKEMoQmZQTf
zij0E{4)*ql^i%=z3|>})d@W#cV84yVbET`t>XR;_J(vEqMQX(uLy5(tioy{(0zd}k
zbUV?9M-&k=Ecjq7oYcqcCfx(1M-^mw@bGYCmn9pZqml_ALoZBfOA8<_1dtN}J$S2~
zx_tSHM0z2U)~v*0M?nS$D(ILP$pdQ1`FEr0B>aV&t4c~fB}B^|(x8^tPSBR0+(OX^
z%EL}e$?Oi>+bH9m;JiS7pFLrzMmd*P6xirBmZD|<4Fnm|IMl&<Ti0R+FV)=3SV6yZ
za><fZR@6WKZ2=T%C5pwi?nwH2Yw5I#KDseTs;n4Oqcq6zi!4zcYUOUiO^|`gCSX6P
zWyn=wrD|?aPho|{mmh8!=j|b}XX7oyku#1Jsw*5FXB1W;a(AKHOoSXSwA;hw7m<y3
zX|>zt%FF(Xd=u~@mxLH9BaAa}HU$8Nj7-=hnwI(~eO|1$mlXj6HWkt6VuUVz(YKzw
zn03z4pJk<>ZY@IGi;nI}p}Mx|h>r_COQ+&DE+g(>Hr1YGa1sSNUnJcxqVnR)Y=J}O
zq$Z?cwq1J^X6n&}6-Pz1eDr91uKtvLW8;CUyH=t*($~*IU&l+y-2HSYC`vOFc&Bjp
znIfgv@-pI~DlKyBHs~?!JM^_`_-4zD$0dM;L5;l`wB`j$mn2CX9K8wNR|C6w(GH$p
zW5U%2`R4w54PIQP5;<$x2v&TVu{!-jAGe)X#^#L;#l<WWtkI#8*Y6^zVWHzsFw8XA
z_#C}UmwWRWlj2|{R6A&Yya$ZcPjlO&6OZ!U5M9v=uo<A0X5mW&AQT`Jk{U8$2cjPt
z5({(}17MBB-`k5YSV6ZB$i_MLmXE60SRS80LCYU<ilW{TF`7GH<ZPGvbc=Ba5G0xo
z-p6b|Eo_Guv`}UP(rJ`LWsAn=huBV#{bX2ii0Qm>Sdy**e-+Tj2nW6n?paC(V#&EH
zxTc}Zy%D`~lGAun9&88=qnmg{sjUsxI`!>-aiSLPSue`3-Z+%g4LHJiN79;)g{OMB
zrc6>tYtq0T8SD(i4hJjT?SIzA+MENw5ZQg&M7YPhZ#fP-PrdJ5%i<c)<%&$r$C<$W
z!nG+_3~4<nY+faOiDlhU%L$5ck5iS0leSNu50_LFd3qFLG`YI!{L2?$w!vh@%R{r(
zHCp~fGn#peQ$C?3NwKH=a5!Rx3a(6foLUY)CT0sid(N2Q0F9!FNhmzcN*IuETs)j=
zhqGvH#>)pY$+Gx@a}^8FyV8=vOvPF%O7uzQdQ5dB(i7cdFTP)a64k0TdaVU#cc#va
z*X8~DLEskwUrdg)8CCz10%u)u0%XqwJ&mQN(aEv9X*;$w6d8Ph(dJ3c781L_BtapH
z#p#$%xv$fq<Iyz?>lUpFze1++GC`~e)Ng96GuG%Eg7xo954QWp-{LYlA-CA_gDTO$
z7Q0-MoV?aM{go+<uIqI<hDpNFWN9F-e5wf-V61TM7{BMrMyCx{m4oTw!R~S#*1vvx
ztBy*}2aB_1twyJa^5Tb{`3oNWtJ-$+G>^5qk*}gx(7{Ss;Xv6RoMkr)=Q4SuaVKY_
zAE#jLnzI!5op09V<RZ4r)QYBs`D_8y?W%QyE%eII)ldBKKb;Be|G`qvLC^YMVQE8{
z6SfDew>dSl@JSOGaC-du_^$h)o>p;Mb4qhpzs+(#e9o`P5@QM3?hW9YtRmAN5}L9<
zvX1ZX_TFBvj;hQWd(7%t+ubs~bi{rEhZ2Qk|D#NvU41;B{3?^~*dB37l6SSATX#x$
zzDm39pys%rsJpkBBoJQi(fy5T>5XD{l-zenuW!i59zCg|U=X}Q_Mtu7zJ77(tCAR$
z`O&#iBz?6|CenZieBottQ39c*yuumGAm^7xo<bwF<+D<pHw=3=**=gR?(|#lW&Gq9
z2bEG(qpGW`)mnxTH!8jZolN)=lVsT&DDgQb!bs_+O*i^)^=u0Bw@9X62FI*p1PkKI
zE9n@5LIQEvqHTU0T<Q%lKy-`bUn<-EOiqLtiQFY`qb4QP`l+Q+c3X&JiMOAYy`#@4
zt~8Z`Sb&OZUQgh~&_2|lipLCR?FYI0T7^K`9{e4Ydqo#825L%>k;Y>Q3{^ZqD5SmN
z@3+w;P4x9N#VXQR{URH*x6<=BDc4Dg{e2NhjJath1`TNIGE?M6L8;K@a}U{_)IK{C
z)IDDo^RbS+Bsbj++!ad*kobB=bCjXny3NKk9G^DhSgS~*83-FBa{NSp%6o#$v1p(i
z!elK<UWFc`IkCzf_fp5_h-UnRLlvHXiIrN~^SI3na1u${)A{ChY4lX?+zjy~B8QN2
z#^6$zAyi3k=<s?K74xO@Q5yPkmDcPcmu;bk2o1|*Lp?u~_Q2_w3snH(S$(Y1nn`-O
zA>8)QW`qNa>73x_1V3fpW+#pEIZj#duwp@06AJan@!wT}P1auN9bze-HbiBQ$5;DB
z!8%rcEvJn13rX|^b$|EntK+j{1+(EfsDFoUBey-<p<PI3fJR}QdUC`JB%u@sstEvI
z@i9Owu)qZZWE*OnE%E7@woM4r-xycF_&w3HZIH3hk(IrbiH!k2=mk>neun#=9=;$n
z5KZ)nh~<9IG~Z7qAFo%VpbQl*ULyi*6WZvx_7*&%M%P5q-TQ$ZwT*o8^R1LLwh186
zv4E;kR28KZ^t<ZcD3roX51VpRq5gOrfB`4;3zKG*zL5~veASv9ZS9jJbtS8&EstcL
zdtHF#YQeRQHv(2M79h<%GGT~q8eoV;NOQIaO&*^zFIsEl3vU5i2n~`-*wtr^R&$GG
zC}4U+p|)zg5fQYQ$Q$G!L&2dec{wM59_@ZQmF%yOEF^leJnhSxj-#ReQuj`<y{?X`
zqeuyuX06KWs+){_-I-KECP(MpduxF`SHS2~_P|vKHMRMFj~zuH$~i1&4MppvPpB)l
zm?D&Q$AdjP3EI?b5)NLb$0-Fg1B_qAu%qs>g_WsS_b$bf6c9_=y+(88Q^hoFP2p<h
znjkzvQEAiJi;7g<9W><A^P4VFh--DAplQ_qc0fr08>nnN!0egjS0#Fo|1m20<P<Jd
zm7yYFwBRVw*n^oCZIrQsuD|9x316D(I=y%O32;JJ$e~E@z!)cyGd@p37eOCyfi_?T
ze*)a+akM5ES9bwElmgszLa?&X?#d8}r5Nm&DzVTj>bP&jZ+u}JO<X83xy>Vm2va1q
zG1IMy^N2-R+9X+)4B#}=PQ``{gvO;MCA%JpikkMbMvPNk_>9+#bS+u3icdF)DZ`v}
z_%nl(RFGBcr1PU8c~)1@PWfC-HS8c#EgJR_q2M@eeD69UB!kT*)|jUjE<7cX`q<k3
z?*X-O?^B9y&8Mx6xbjcc@uEMpta&!lMexxm!~W0T;6jw;f1&}E_|o%ZJ82!CHhJf?
zaQG~46vsIV-J$Ly@9sP0@rZf!lA^Lto1sZT*xy+VJ<<~9@k*UNg$ZC`x#h$n>E4o5
z*Jb2^g8elMNOz90FN;d>4wrB)%M!@%OPtbXGMQq;cq@A&k71ie6__Y3btoB!b!~c_
zQ)XonTXW8Jl_On=$vEH7*C-irjlm`jh*ECN3eUcSY17cCsTdZ>eQFcFawqWJk$=w5
z@{CT<+F7(}9>+!gH5~zbD05EGcbqaEw0i=o%K}+nedjoWo1f4!W)@JNKuHkPeKfir
zS>hW+7trdEDTfsUH@ha+S8(ov_?=)n46`Xof}D#tP=F9A;NoQ(`EBTCN7HLlC0eyx
zZn!L^hb%P&&5An+S2~!||Cr}3>NBk5=!`nKnd*;B+<s^Da75<4K~=Pi`5g&dVF<UI
zwx})mCc~L=&l>7-a9GU5PGUc0BE^3X8mA6Nmh$Q+#o6*!a8LWCpIYmpt!)%vvdP-n
zHrGAN!DQ2}^T}vK+23@90r2;cOuBRStF$dVQj`+)hNn;RTpR=LpQISG5f^0X<i%^A
z2Zp96H^OBtH#J_RYy8CqFCpVz8C=sINl9$#$48A^d{FUQ59fNuGR`+Xuo(&J&~7kv
z`t9kHOOlmrD?z{Rp0)2x<Y^|9`&r%3SCv%a>5fE;{za&;UKk9L&*~liA#NAT*$Xas
zKYu5TO$Q#LxQx!UdbpA&6=KomjF5q6{{RNIgS`Hy^PA&;=NB_D{g+sI*Y7*s!T+7}
zKSf#*fi?SQHWO=<Cb}Dx12mmuis0cVlw0vvk&L8ixeBD#BCLpL>QpOod1#oM8o55Q
zdTBMY=!9B!x9s+12BTx<jUE;<!?KHVw9cr}o6?ihOdNO$kZLp?KDy)CwcP+`1tf|Z
z3T_#0gx8hM2xO4n(std#9wzr3+S>ij$}#d?lqG|#6A!!E(et}>W_WhMd&(%Djtscm
zR)mfQqPY)&MulS5GDkIUs^gY2y4~9EU|B+^oSEn0vs*B%q^Vh`II-S6;7wB*s8o@K
zqxHT(?)O>?9`;^Yovsc4O#+@C$)u(@urkVhp`aZmYcR5Uv)ED#^N2}{parGb;w{&^
z(d+ne4GaxScByk9C7<hlJ6WOa`U8yRZ5^h}JpcLK@>ciJkfK>%(0!6LeK!kdg)tPE
zRnN*CBiz=zB9}x9*U&9islejCSoV;u-erfEHQi9oR!^6MyvpnS`Mstr$IAz1rl^n7
zRhDYDJ<n2JWFSgRg5*~;FZ+#7BMfJC9A4&r+rFkP+TUk}H3{m9@0$pE?^r)mI!Sn#
zY~4t}9}9DDLv_Dc#d0CxwqaV^P(lYE??PRx0QnYvc@+%htc<KpQx+>4p&6qbPi{0e
zw4T*LXCVr|QCg9Q*QN_j<l*qjvc#4s0_@Gt4IUKzn<?AmsGyW8(_qKTm-4kevUp+i
zo24q_L*3!h+f3tZB$8CRF)K437I4mu?iyiAy&L^j&762u8Ks?vM&Lf7EV>=8ODtS`
z&~++OpmJ#^N=KACvoKC_NT49%+H_}iXwbK2c!5!tp6H9Q2NAmYlr~U%pTfF24#C2~
zRQCk3utYKls`ZfO;|7kTJ<SZHioZ6>%%sj)J;3y5wT9I6BaVp%H<mX!hi%Xnp)jpF
zd<znA(0cfKvT6u?IU9Bo+aJ{#MPIj3AJC=g-@oGeUD<yV;<k9#!R?pWy7w<annISM
zTT%)rRxwQ$R%sE_s*lFQIP}Tf0l^Wg`c&_7glaRR8riHLXuIvAm_cPdliVd%(3gkd
zzUTJ#5Y4nNEdeVkt`FWKIw2?0;UYb;CuT4EPG??1#HN8fnQf$qMr4JzK2B@Hz|dq@
zW`qnJvD7_qh;$mYF@r|X*Fl4W<SaM;bA!9W@Z&355!)YcWQuFes6z>#`mpacg8y*R
z+Sc>~6cr_5$f`?FUP`M@gsIP7oJ*X^=OoJqY1L25tnAJrdUS=)klpA>VxC0J4wq&=
zo5BPO4=6IraTh<<5calJET01o)a?fv1w1;nM$O|#ftKH8(#V_4HdKg@$C$y1<0N{#
zKtKx7R__`<A>`R$6mOw<N>($m!(hmwGYPM9uvng;BxDMtm>N{0QmH?5dERB3Rii1p
z^hNE`DsS5Pz~Ip>!)Cguo$resT6kJhi-+Z4M!K7PK3=`>NSZ3;QIc9Hzu06vZO6Vw
z0XFk8jFqrq4(+xe1pK~VZFxyX)oT7~i?P8A+FIB{u3JBAZ{#f0x|e(So2{xsoCWwH
zA;k%Vb`qnOBic$llIyU(8jX2$S~TqGas9|a!&ISf+Nr!O5xPF+r<6S^^H`!FS9H`f
z9H9H`i~UjA*_Hm{qc`w<{*7Hf4R#dTc(ew(@2Tg+f-aHEj+wf75|vD|<Zu2M^!-Y#
z+kZM(_+ndDm!sF0O(4IIqwOz%nI20jxAJZwHtwmkK_BaT<bK<_lH+)5xMNX3SgTfi
z7UBhxxIR9RzKrq=advAMbn%uI36(e`)Y7ykJq#=gFG_0zGOgGc7m2%7n^so<{qa=}
zE*AU4H_vTl{vZCVIT-$Tv<)LY^MAR3)10XO)%><js6K%UC5v+S|Ll@FDdly*U00u$
zpE|QdPXEwJcA1RTWsBVq*3n3LV=Ev_q|5zCd%r)~f3tr&if~|<{?Zga-nk%MyN`72
zES4-71ZkoSKA&md?2d35zGm+d(+Fo~r`z)?+1+`yi$dh@b6NzldQr%}Fi!$uMhYVg
z;$6FfY62^4JF*8jE1OF8;r$zDGqF2x+>jIL3*XCn@1-xX13jUHocK!XGMX5qt!dxy
z4!3kg^3qsrfb?XCj}JFzAA2u3md8sv!1Ahp<h8AeP5J$QML7v8V{>UR$@u;VfCT5{
zqJPB2#DF9<i`54_WcVZ$<29vp8D9SuIr#kj3LHKW1=Nurjt@;hdKns<rE?w=i((Zp
zMG(mY^b(lIg((6T`(GGG7Bh^MNn5~Vj2DS5mA3yq4qv=%D=PeH6i0`<2a+UR!oPvG
z?+E_mI3$0OQ}_$7LPDfRrF%-5uc8*_Sp|y5HR+=aYcnR8L`}NgvN#e5<sT`Fk_077
zqK^jI6zCzqKNLghctup5n0q~f?WT<of0{N_B8@{k*#i3pia-t!8xBhCViY9#dp5=4
ze^4m;NwMEYX&)-YKPyI17fW<mG`f-Ix7gZ8S^*|C@ka{G&lL2ZmFi&%U9{-gX<>T+
zUQKG}o<`Is?PZwpi>aZ+OEEc=ecfAosG-mCh5ysM=(EKnf&@JEG#CvOMIybaZyx0^
z+f+eGfDj<g$(n7FKHbt)rd$Uge$-XQAG2)PBtKv#{ieV*vsk{T`2`$Tv?~}>By_YZ
zq$5e?0#=ZPLmy$ys+ffY+Y8a{?_<`D%@S)9bf-Vj26LO@UW4mV40<lffpf=HJh@~u
zgD<@f@Wq^6vl@|aBuZ?=+3tvulbqdl$1JA9J9{PEMg1=-k2c^YPqYDBu(x0ji5KWk
zcNG++PN!PJ&<?~mRFSZ{`e{#)UqBJ6cqz-|pT=mc-q|NAhx|Jvrnlf49Hij!Z|KeY
zZuDdTkpJBJD@IYBg+?|59wr9$soka?9*Ga-^k|;cn6pT++|CeS+ekdo-R3qmZWC*v
za~NE0eV2dJ)tS(yDcmPFbC<TV;mhW4ZZRD{AeVVnYoR7yS@WJFe4j2v0FKZ^j=3)1
z`E43PsG6E1Z6gS-<Y`RCSqhy|`X!$StCF#J;2P<fyWkATJ?HVfi1C<<l}gy)YU7MC
zxjM5|Eii<7#vMam6eVqj*IQLQXol!%Lh?2)MoDU25y6iT+-Z3>AU>z&rBt&R!>vId
zP7)nCe>n$-_h^s`py~M8xuzFPdf0o0yklRxO#7)m9QIEo2(acaQb$H9y*GkPltI~#
zVW(z;*sUZK`g2imSF4~&8alaA1}G{nerkZ?C>zVobMqCK=>vLLOTtP$`})}nJFtyy
zLXvmC>)|zbTn_OU*LK-xI9LFrSkQRU*6MPsAMH2Jbm_>wA+(r+0Nei95l0z`rsY7V
zaIW%J;mG%Mt1o#OQ50;OLEjfw%5zE_@4wJk`c_ki2{Kh#Yb{JkDYaSd^<{kZi>61a
z?%5`__%KOoZx0LZ+IU_}!<*mGm!tXC>@X&vKJA~@a$G3qQU$QO%&_Fki;L=d{6$_k
z{|luA2Np1;ogukudd^w1oFczV+dyG}Yr8D!LqA@1<pRa1=5(87B6y>y<k5$A5hDH(
zvS%_vlPo>A4C|&G5sWAup0jDzn>|s4^J*$qmcEaydB+1{OTIHsd}Ok97)SUCD$X^6
zovqJZ7H<id__zrU{~Z+4h&obIO*6skA|)ZA#l%myFbzejJ|KTnrElxOMt``Jf2at7
zlDJ!zf!HYn>TQDM?d_Md!+K5fGBbhC3koe6Y(JOsW5XF1=5I1VGyM?NZ=wrEhjqcC
zF66XGf00ZWr(1Cp7+i31{2Lt#ABra8r^L4E703=Hn_gmHRYN9)$A3_|nCS1G?`6W%
zR=8?dG45^p0~Cpxs;bKi;t_jNl))$8fhZktW`ejvKwi#8Wn&nK*ysFNcXi~8&}bUN
z*sBXtH_0pFpmRerf`X%RQ>=-8!WA!&OB$lAZWs-au`tOh(^LAGD>;wz8`sllsxS(Y
z;R}*+0*$bSKekw0M)KXskk%o^KCn-KR>R!scsEHQC|mt#)DVB-A^Q|^1!**j6)r%f
zgVb%G;axox;k7J~b0V99G9hj<iJOQ+8^~8HtJK$i((G-nqWT0}&l7?NRTPKkI!$sV
zXWR<xdH`it>)9PdL2c4-pB(%UvkwhV#x;%=ny7Ol$C-tk(XTa|>&?FA;B#+ZU)cO&
zf4^5>9v7ns<kV349-sjj(R#zlxGmgS_z!75r`)UeLX2-);fmTjws?`HF#WIp?zQy8
zFr*1L$@#zQU!+P1m_9;+$2{qNU5A9)iehX5n6!iS??-mejFZfa^#MOTi3vs+aBu3#
za_a14b17;al>QgEOC4@4Tgl7OLM?*8)K33P4C&I-bm<=bb^Sq{F%<==+O_*v#6uSG
zWG)wr3ZI=5qwlS02K=_<_F2yYvxOFv6rT?&+P$b6FEI6&WV2Rot#1x##<=H&xs>k3
zW1Vj+DlaKjW5^6rwuR8I1>SZv*{s6&vTnv?m|)q=(AEYD5%x!ZCQ7bLly;1Jf!O#I
zv`4r{K=nnV-74T^qSCxs*_6o`RdHs1w&W4zFK2AiopZ7C7vFN=k^Fy{CfI*d<^LbP
z_y5;z%+C7%xQ!WqcfJ4TMJH-u?QG&mCu(irY$9x8WM^!`%M0b?>}X<O1LeLxrMY8=
z)rR0ZqgThTR}Fvc7{Y{I&ki!GlW~E<fgh%nXQPi?U$x2MLM)ULsQvTqD`Y}*aX5T2
z%5%|oABLI0l%=<0&xSTn>)IxN9rV@XBEnfGiR19>zp6ls-|8AQr=8$ZmJusV{|=#K
zrjWf`UGxqjd|pn1hfs+$L~$iv#F<=I#4^rFdH6wk&Q6La1v6f$M6(({`XURN)T~!*
z6ca>~5P-NtWFlXDrN=3IS#Lmu>N16FR9+o17~^!$&}U21aVno-_0*<LiSS`ds3S?a
zD%2>YxbE2|HC!Fa|5d;%RF<f|fx8J=H&FnoY}>Ps&hg%@kBX|N@s-3N1bHDL8O5{6
zpRNz12}SyW3!3yE6~#b9YXPQbTfd8Me?6UQ;6L9O#HbiK5jvHeNzD;B{wSPoSxV8&
z07gQKH(5BZN;?L>p%#+Q((sePP-v_-Q|n!6OVr|GX{N8Q@bm;@_!q&m*O;T_fyvuA
zS9NoA`BLk4=~G;C2<9Pno>gvDaiLccM)~M@X;)cB(LFo9eJVds!HFYnqv#E#bI{T8
z%%L_!CJ#S_k^a2~L-`>~latE!3{y(Y@KGgCHGGPZ#pPDiZHp3z`hqI3Ev+%xNiCz5
zWw<)wN3)Kf<flm#43vSt?}~$k7w`x{9s!*tUJ8`QQqe1*4yJrLh7zu_|8&|cZxcDL
zzGA)dFMS0R({Y4QW!n{CQ+d@RK<)_LFusZl559s}yrlR2Rm;~y3xuAad(ZXHcdVb`
zW`Gy?C+(k3A6FM&-^a(ZRQ9qUn3c{<r1Q`SX8M2yYW^#%ES;aPPw#H_&a^D)jzgsO
z?<M25ELu~Xn(ptnh5OKfBV&#$bCz2}UM<5|w6Ue&CZANaAtzKv@ed#~{OGIHQn?Sc
z$hWaQ=eK6Rr%#a|(|3zEjZwB9UFo{D8KqHrEm^Y(DYvmD&zjQIhl-)GN5^D!@6*z=
zv$eyHx6+pBcmt#g@1Qs6$sjeNI*yfwHS2E8QPmATkBx+4sqKU#vjA{H*dUl7IAPqd
z=Eb?DNt@W}AKPyBY}yR=yMI4YKSn=1-MiQ^_}^pqGz{P_Y{Zba$ck{+j_F3Dl45f2
z0~H6wH81zp&o|#|Pv>M44R4ivmp{+%Vm}|psYmzw_i{cz4{P_S6$kc=KeIhO@AeM%
zoL{RyPv@m-dtV-Y-n@L*WZYAUql2Z-viBf!6i-QV@fPD{vs@%njM+tr-;NIRep%Gh
zq$4K_Qqn&IQblFJ-lupD`fQa8%v*5}u1ywCCdIDP0toJ-)cMH)X+nbL=CT4%($AFx
z_n}|#4{Q(L9xt#Y>~vzvfru4^gOyN&s|rv_X@3yT1Chfi1{J?skmuy_<FRQ9tBDT*
z_Oy!7?ObF~K_Fe}J0zjhU+SiO>)11CGRMZo`=lt(SQBH_m^2`R1UdetM+9XZ)h%V|
zQP1#ZeVMhtqAh0AZWDB|6vwl{-72^#VdhchDZ_FClY=hCe=*oly52G_eFDs75rT>~
zSK_k%lZ@Z+TWrf)Y90t9C@81|b{+~EhqG@}LQe++=NeEzgHyoU>M%V)tEklAwo1X$
zSb(Nm!3t&r4-IJ6$zf|%J;LAs59JJrHx%a%U@(D=@e{SDyf2K}1NjS47qhmt7oxV#
z6N1$a%XV0?8<N7Z83$-P3czOk7mKkM7GpOA#$E`lm7t>!_6RX(Wet)B{QKl<Og7;U
z3*jALJ<__^>A8v+aN^7GsH(YyQt-4w^8(U6YCAr&0srNxbK$v=KCC`;qf{6x^l3g~
zah`n|$cq@&N0Z=YAy_FUjQdMFnA9*ulFwKI2K@kT|El8OJoKYwpk174v{|^Jw7Et8
zjVdEmA-f1dWTt!*nG}|5=gFuHvSwb>G+Je^M<d?!UqCt%vcL-T2)Y>9dolFyJ$Vbe
zZc+E2_*?dsxK4k%{MBdwNb-AQ0b3CKapBQ$u57kAP9@zJ8GFb2E-ssgh30xlKw}-4
zXxobKdL0@CA;BsOm{J9C#sCG0z@6CW-5@)qDkAp8P%YdBFG5?7TU*(|($WitEK>jE
zZ@jpDu!z-Ow3lFHjb?n2;~IbDw_(_j<(XZ4_*KFM8uGpFyDCum=^8KCDMm(Y#&V!W
za?RI*_H4yq!vgJwf0i+rZL7dZa>zEX<4wpE1H+G?6v)}3^%_MG6!>vhDdy<1D&y@c
zCqVv^TP?~${M+QkllAG=j+hH}*7iOO2b;s;QHA3@W~)zh0A=B-Ee<cy0vj;`SV)1z
zx27;DjWK=NA+uNkb0PZ7>#+5}!|w+X7m8>oe;scuaRANx0TFULbaEjiIRM6++6^(T
zF1veG%H$Ofs$UEd0|6?x2M<|_zuKQ%jB(u-5|MG#bT<(Hajh15U7w`?HQT?H_Df`-
z*#exg<^r|!Pqf{c-W;~rtj@SO2GJJ<XE1d`5EVrqID3_jjM@{<Jx6<Rb1LW(%;Yh3
zGxS>)`0zeb8dtR;f3va_Dt56UA8?VK!=dsoI4HnM{tNV8Ce|a%=L=6U!&^-|hry8S
zS_$}9-TLCU+{bFww@e}#E|1FQIL6Rub{T9Xj-S2g3cS6=;TRF1(QgKS9nxm09YtZO
zo3P;X=<h{X`9QJ-PPO#KA0wAm#ww2%ePB2RT|RC3-MG<uu^ESYz7d9X1Y5ue`l=;l
zegvYm2K|QjKXODAV15XPm@e_4;a(->XMIv$aTje-)i4-C6HXvGrXl$VY{a8Pq8rgn
zn#cb#EGIKS8WyUqK6rZZly2F2tNZX$<8XiWJsw)S%;MYA2d&zG0k-v93$?nVk+$mR
z&}BgZCWA-pfdC}28%h8GNU{Fsb65e>AvpAi!OBshZ#?a{@5MgH1H>$Ow`TL|CR?xg
zaRW-7ugBMA21@*bi_dSUGMU3~!1|=3dx?qCl8y+a$GC7lC&>iBRbm7F9+2CqDt;;V
z#K>`P>v|RB=S$<(1z|rT=V$I?4?Zz*$a@J!3gPC%hDhnP!%y$8*AHljl<U=hn5;M$
z|F<a4$o30t|5wwK)>sTt8$#EL+9`O%6iF)}C}4u7WiK1sx#_F%<fH{$&JQ1ZYhs@^
zakA&GPzo_=BFykRAz>umH~Z~CY|F#RMoVTupouIr6?|?F9?16YtC~GF^SgeqzG(Qs
zNp7g0k3B7{u36__r{dfC*R8!$-fpTr`;b}pJspCW8&qL;uoLq192HWYLi?1?-O}C}
zCG?RK9yadE-3O?XgC4%0yJ4ofEDxma5cv0t_qXSr7Rh#HX)M3lX?l~%JwCtgU@UxQ
zre(ZNL$J@injN^Fp0C7e&6oXG@%O^Fy&66^<3lskoRd**+o)yWhav;bjzET-GJ7fp
z_-?1_cn$jjXn1&caf%Qj-wuwt)`H;%ayhjm50NC02%@S%{8{yH;rUPm;;w2-lxow{
za>Jssspd%t0W&M)Dp4SSvsun;$qjB?--%}~Up5cgn7fAoW+{G#_Hd{>9&KR*dC2<F
zBH0Q!B0j%LNqI|sL(>+y%jJJf`tt|@_y-C%NJY*u5*&3|JGf>+Y?B3d-<qQ2)#)H2
zZUS1-0U;YYrJ^f;Gi6GbS3>2gij3!}!~7IXROp5C&#aMgPEto+RC6S_tU&d)v8WrS
zWgBFc2^P{J<#O(zP5prczzN8@|LITfImL_aMmX@@mEN1~Kwr1>9|DnoNTU4)-{|B{
z&JFV10F9`;Xz#LXp{195o2;N6S#h?wO_x5exs6_m8Q)W^V+-|0wcx|^oRL(Ssji->
z(>*5)r6a@#8R<zXx6TM4vk9Sp1O@Kq_~#<aZ7ydEJIEd&y`Y~r#QX+PUq~c}6Ir;T
z_OSdy@>;f7C@PuZ(0ACJnR#04HVdm{2Pv3a6@JAIu~gmTt-sF^<(1`=yGN^G6YGN8
zvKz$@akitj^jkz9!o}mD*4c)$0~=diI{Z{XtQVOSl#(s}26Cil<QVPeMggHOTcppq
zG}$n6CZE8;MM2p^y8swKQTlS`ZbK}li$SYracL})h!j#=?$^KC@D@`cE4!rbt^6<W
zpR5b|nc5(T;;4l{ll2G5OE{#z`MLNMvdGoY;Pk&yZZip1rajiQV>J2R*3`cy<5`B=
zvM`mDMN9c#;U92B(02k}&JyTLkpVda8i(L+i1#fWQQCS+=O`#z%`i-;P?E^JVy8)z
z@Q^R4{izq=Dq?Eqc*S9P!0i8?V&!XQIr2rg{KQyj|6XZBWQuNdDq)6)wjOc)vBa&-
z);K;n2$Sdu;zUcb2+?`F(@ARFG!wxzpLYxt%&uya$6TH4SP1XbEaLRsmQq^HtAx^5
zZ!7g%wbR~fbe*1Sjp!5c9eThEeiNr$kUi)1WV)1tp}*WL7$$b3hapi=c}Y@>B7<n6
z#S-mFtshY!%}M3j*v(rU2aDMkPbo9^<u3m$p2T}iRc*D)?Vo9c(wmKO&grj6-r~$j
zj*qAFPh7i@LoM#qUlt!<L)a&|@PTtY+Dg5`$kJceZ1)O(go&2paE60G%MPdb<Jo(7
zm~#F3S?4)jE-=qP1CEYiL1)6)5QjYk(Wx1^#2p0pA-{K(G~!YvUp1fY1_&<RqE5?T
z`o`Vy5K;4#0l4+*%p^in%xC#jcZAO+z^hI?d<YR(QYvCt{3o_7n~>ox3Um);FcFv#
z{~udkZt5<kygN1~2=xU?u{l)_|FSzL-y<MD-QT=VEgFFg=)2Z*w!gIl=|5!KQOH*h
zrdfFJnDWZYE_F-R_eUIA0i033GA|@)`vQR{p%Bf_@?gSwkf~Al(l`uF>sU9$mkESo
znjYHuvDk0Ca`7a*Tf|D&Ii8Jid_K`tCMhyZ1DtQ%e-Ro6<<-huFp#pGSl|ksvUK!I
zFm%M+WX<b3wcr`2Tsd7Vtg<lGt4?MI&S<Y}qk(L2RuoXC9oL^1HIBy%D;%mm4Lg{g
zgL8D}NscG(w6R^enU5c|=WvL&rQAdD)8U{?^+b&F1}Y#0MoD{E1KxHxq0S!hZ@z;D
z9}+=VrceJZ1ct<5!c~9CqE%dcVW(b!*Ok#wWndPNQWz$b*htZRDV)Vh0&jWo`vzws
z-LN)+4^$@bzo<<3I>`_P#0rCppq?S`*OLzdhFz2LK<_rS70ki%P;h4}4QbgX|0A4~
zd@3eEld<u15LSpZg#cy_p%p<0^go~$%CgZlps?mwsM>3+zk}PP18o==AMP05TIs7)
z3=vSn8r}2VeJd_jUlOnhTv49@_kwrd+Oilz(F6g9fE~&HCT*k)HUI(<shDOkx}2D^
zG{1qMx<*s5QCUj<gJ4kskC;v>;(|oiYu*pwH0oP_4<<@A1#mxDNG_wDDrXLp*k0~b
zVc{IcGP1?PJG`oGFZlZHj#P<^0jWau<OUhw-{|~YG0R+i?x0ftE*Y((=`U_n@M*V;
z@eXc!HCgY-vzb|V1?ySVo#dX=yR%3<hQmWyB<15_tSC{H{wS|fAI5;D((49wqghvp
z{7^O%G2?cIQ?Ksr+-#KVn;iP5J=v_d1l=PqIW9*{R<J+DG2u&Oq*#kgLJgs}LFe-0
zR8f$jf+V{uH7oREJ=*0(_ue3<5;L*rD{h{98yxyR_-vKRR|>t6Ew&$6xKXl)AICBZ
zU2W3>ec?;52Cl5z9#2Klk(549sa(3cl>Y+ruauo(%$uuQtozR&gjW*E{)`!gQGP8c
ze0!WoW(ML6vVbl#Ab&~?9#+r6cK58Xv^Z`N3Oao;NwBKc_hQ2+`_sFP7)vJis#15U
zuij*0#nsztS*z@o%L<Ycq2*cxNf)1FHv*1ohKgy5-eFj;$>;!RoHMPLCKK8wC&p1c
zx>Ctp2h6Qe^&)72A;p+SyX<Y?{)qT%aqdA^K<5VRBrVC;@_}Gx=|Y0IuZR(a8V+Rr
zls$#LF<Ot{qRO^@MpX*&+_J?EOE>Bjo56NWt1)Vbg5k_?yaLu<qc}3Xs2(S1Uf{mX
zc3t;$%O=eL#|Ga5##4v@e8W%0bsZ+yhKTTI=Mu9>?E9>6v|{#z2uMXt#l8t-7-kx-
zq9YtB4}4cTpB^O_u@XPV*7DE8!O$b5<vhcbCx|~2Ze2msHb*C1ZhyX&1m0$wQdWpw
z3(is%O{6vPI5=_$HUtl=F2;q6FRWQv)F`Kr3(y2FL`St(Xs4LNEJzfmWF4!~d!4V{
z;ynIU$j+}`UK;*=dia(*wyvmN2KNQrBKZxhYQ++iUw((R6LDF#^{lWYXj5)EhzC%}
ze|krADzt2OMrQ%*X(@CQlo$+7;cZKdsh&5Weo&NghWYwCdUc`D7h~gMZ?e-oyw<o`
z5k~y=S^~9-8Q8t`=gY=H9il{dLPYz5>G6qB(%>wW!d_=vxg*U7&g43waRhrNv#jsL
z99rL_qMJZ5UHQIiLSE-a?}=mXSTGds$-V}f@`5gnxVWIE$NA%fmdGVCpx9!3k<R>z
zi0dabY4)}*g10i9t_l^d!pJ+hjAO#4p&v$Eznz8lzP(_gj@7V6Y0iX&-aOPJjD%SQ
z&yL6c+{5_soeuD5n}L9w;H!d<Aa+A8YoW8QI{w>aU#d8oq^<F1aFyKV62Rhfsxrsf
zkh~!H<pQ;7u836!z8gGLGjrHT&fGn?qMLrGLCSfnB695tk?nZQ!smtsb-m$ZD26ey
z_3(6E%FfC<eqMLz>EX(6tZSp`KTXEW|Jy>q$i(vB;iA8F*|7($2*34~vv}xmkwg2y
z=D!|-5#UQ3a35$#sCteof{!0~3SJULcqv<#gg_cC!pI>{UzH^6Z<f)$lZmH;qYs>?
z|4Fm4e0?<n@Wp?_BHm$URtGTn6B5f8+S4Uj2IE)#Ij2Z?KXPw3qSo#$WX}y0#`lT5
zdp~6f;p7fe&Wt9L5y&mM^$O+f1O4LB{R@pk@hY|Ie1G3<L{3hsP9sAx`qHaHm#A<U
zS+47KQ$ScD{t)RV{+62c<Fx;~6=rHF8bJE?k$O?-)H@0Bm`v>k=Zp`%aRZN3yw@b3
z=b^jJDH5cn=&u!BH)*xZzCN=m>7@267KkgzsqbV;4Xa6a({e@oNpA=CcvE$ozoh#D
z>SRP)FY-_zq%Rg_ovd1<ce2^X)Vk-F2}r+6Tb>MQ_F%~6qm3-2&B?vEb5jvc!xM1}
z$At0zK`*sD-WWBvVElyQ(oad8+6o{c<;px~NeC4tqLr}NG^!$LFZ8(;^p@LZZL;8p
zIx}UR;6<d-u4O2|D&l9HzFl%J3+Y+CPR;#%Z?lab;u?iGsjWnIDNR@!`ARcKQ2O2M
zzBp1L+9Ur7Dpyq2kQC8i)Y~B&S3#24@cwDPQZQlR2Q)8t>ml1N)MYq3U%;+(&spr3
z+H$ycg-3w0-W}2-cgIKhU~)>V%6K9TW7GC^{qgng{PaRy97W&bf%|!VAmN_n?^ox;
zhca_`y~rKFqw9H5Zuow;g1J^0okuTN7LUSr1m=L#8~4*K9)qmcFP`m`{)#Z&+<aq!
zJl+;zx?+a}MBYXl=rIHqsCmceQ?TME@7zvsoQ;*$HI$AbdtpX(!rp&GC=h=>BHECj
zW?(uVW$i>rfvwGC7_J)N&U+O6NLP#`alIhYYAH@f#|xLF`Iw)qrG_&3!MiLsFpkw=
zK}y2V$*9jvb(U`>;cCMF{yAuU1F0#+XnAq5rGd0D-L!rCE6qhw4-0j)F@w%U2nqK0
z!z3~z5bJs4BdW~F(FpvUL%q&<KqF~aVyYVb(-A@!bm8}_fk%n}I4y7yC~7`DmJ*%=
zA=wZvZ6p0m`Qv&)n#z36iB*GCFySOE5@?|;672E8-Y3RtFcyz%)K`@?1)#@%+VVm^
zrAfAW(IH`Bcj4uSEqOi)8aRW>s1&K#Kc+zEv82c_$nt+M_Es^G1Z=nN;0^;c?t{C#
zySux)+u$&`ySuv%?l3rwySux)9roF0@0{fSF21Cay6C#;RHu5?df)Y|Evm=lsON0u
ze-tl!aUX2hXXlsbQpE6?ml%_IkOcPi5O&rAA&Ch^zG%9zq~>}J*_E&Mq3&}dS`I}X
zW-NOe1<x7<$H;j{j%Ni2$SOxX6il6UX3g};#Cq>#@LSI4Gr|1l`Ku+-Rt}&%JaO9t
z5>p9AK60Ei%P;WSO&GcK*lekIo#jBcn(ANr#XmaDp6c|qVk}~L7IQRWEcU>yoB_t*
z*3!k!n;ljMO3q}WG!iRi%315GZ5C|u^5y`#u}-pXv-6tkb`)$0m^!@LuVNJOK(i-S
zpYVRVL&03VT?+`ti7Y#^a<sn`uTff01)C8f31;L5B#&jlFWgxnciD}YO&e=pF2o`(
zN?L=q?Nc6IVA=`F!r^4?X+5)U-V82Zxjuv=XiJ&)#SQ)<ODtlLGRGQf(iD@_eMaD#
z3<?@PKiJc>X3dQQ>{R9e2C9;pr)BfOyE?X`7n7vbY4!F>8;t5|bhx~>A8RHluvphD
z13jPWLhYd)T@WdFYMM(0mybKY8uKONIKH7z<~koOdQ$eM+OB1@Ij?m&9gD_-5-Mlv
zM6ryXO{v|5o;H2f0`E}++hI&vfbOx9r8=uMPbwq4IlB=w3&sV?wE;BkNQ`c93$g{t
z9-Ys@5P|n$72$7f@kL)3^MXuOn0|fAqt5ZufD4&$3J19Z7`Mu=Jl1gLQ<#IUAH6wa
zlr*uL7g>G3)aJAPbu6e&k$s~C7twpHb)?vv|BhPnZIsPMHG5~)sAoH_p82i)7DS@C
zOD~Bny`hbiu#@OOL<mbh#iI-Jm^m1MDO_z$x329@qa$rTlOaL-uxmT1qwh<7!Gdxv
z;2)e1ujGgM@byym8^7XHHs|<J^tQ0rmjjtUfoa7gFB$RZ$}*<(1&DJMlm181jB@Pp
z{Hu+kIM~F)rP6ebM|27Z2Qy!rx7BUNpVP)vqmLlv_lG*1Hg9-VAg?=?S@VlXW&a))
z@)ccWfFh^8-e6btfYL;AL{j6yEMji96>Fo*mlsPHX%=UbVa@a2v)39ve9qMvG?pis
zB<>Xn0WJBXFz2T}j|up2@b20E^}+()A)T22JmdZv@EbjH2VFnL#CVov^E(0;wg#Jv
z71#omNCEk1w<xlXC3-yWu1X&ziyIf29|i->+FaFMqQ15kEy!JNBZF01<Zt6`yVV|(
z*&QE0;Z@o{c*s~@^5BN2ONhPa<MLTZ3+s^K9CfJ;kPA+5>ky7V-FX&mbFjU92kJ&1
zNVU1p0*dGUq1<nhmt@Ny09??VHHkj4G3eY%LXBncW~S8IqGzc$Le@2K`;Vkr2C+aY
z$T9T*up+`cIKIZ;9oi`~!XWwRh~pq2)%AKoGfL}l$+;Yf=W0q$cLKKz`d-)?XdyN1
zb*p7YjE&x$l*HV{K4sg)LNoAHI`SUH(#oRCF`LTh>J~=v>3`nQJghHeVoh1!x&nQh
z1iCv3RE*VA>Gp4Y7okOyrkVX2KHA2vwqT{?ps+Do3Wi$hxthfR<9fr-CAmJiJ$A00
z5arc%lb0+%NYK`tyFilES1xS>vQ#uf-Hlkn%6^}0f8%u{O|=T*e90C{99%k#FuB=N
zzpMoLqft{OprknXjUK19UmdkfAXBp49E?mZDQ2C(jufV0XpMw<50WKt#qt+Ej*Ho;
z9v?q&()s7KAnV~J^CUMO?#(*5(=*8x^pG(9w~O$3q0``(_)q=9Qh%(d99P~b^#Dt6
zl*F&$ZczT9V^0l8;|(%*)fc~yP_NBc>v@2xD2|w7lP}v1Cd;7dZ7%{RpSiK|_-eg|
zP{A)n7Fxh#-O`-0Y*><Gl&^C8ENS)SdmBGpCu6ozvkqE#FgVuy+X-<E##>vn%`nv>
zn2Gz6)?dVKJc`OkV5-Y^oF_bFC3G)Rni&rVcyh#pAd1-&co3l!Df8i3OxTd>->8qC
zzkX;`NBpM_OitGSSuoDX%*pUyd6R>hqjvk>3%^g_-eiaw?fMVlKON?)0(Okv-};N`
ze=qz3D3#!vTsv4N1!*%;A_vuL@gtM$J}q7DJh^&ub7}^nl{f$NAAWd&-3pA+TUa)h
zIQ{^luw3zH@aHs%Q7>uIEoMm2k^6bIWj@0Br?U5skjjIs(#H@z#=AjxbenYWk=XBu
zUJ^c~F4KJkHgjn~ufR6>b$|VyUD2$LTH42{9?`HrunQbIEalPly?vx=xU1vfCNZ_K
zs4^z~Oj^OZnG7p6lEb<AbKv)gN}p3BXuo5+Lf!k%sFCcB`6sPu<zHF1sb$v-Z`s1C
zSM4fe1v@}=ZFVh67y#i9QdBF7pS=q7Lm2%MCb@&-`82lYH&ib)IlP&E1xf-}V!7Oa
z*!$*D>*ML-<Nm4~F%qV1S@3QtR@uAPAjvgfPz9v*<;Csp<KD}*sjjG<Jc3-IscA??
z(jb{pPwWnO10_U;a3qnzBZ=PdAzyG%Bzny82aYmpVdR=s54HV)Hz01!af7&oC{so!
z)#rc+lU2~@l~l?3$ggvl@3kIKmUBxDpq#0vfRPv^j|V+Gipr)QohNf-P&iI3bCNqr
z4mS4BI~e8omFn$xk~>LDJL$yyZ(7S#l)0j9f4PR}4<cX(-aqGY7%s(7O(<*c9w<Pf
zVjl-7SJZ*O%P|_vkcR^`6?i$CdapLnNvR*}S%*S7L1j_H^Gw2NEG^~r`F+9xtPB%p
zEL?c^w1<m%jE0mG<LrhAA;D_6hM9uANrlqunovZUxsK*!W?ghs(<g)kv8fRvjpCqE
zLO&XMWTO5ivEMx|$Z8w>*{r|ahPW5gX?Xy#YDLG{S}AE9O(@`6n+3_76|`vKg**@?
zo#AhZ92r-7P%%ymL2nZo`d7(+dG#~qijWh-h)u?;*rl0wu~^Tki+`PeO*V_ImkS#Z
zG}oVD!A$VMpim-|a4jRxvS~bCN)|GvkZRtrSnb*eUJSO-zeNaIn3P^aK)wD`hj;#N
zeyQIqfa2<h3Yg2ySV9v)ex})O7E@oXK07M?0?4b?4`cDo-F~>OO`6es8gEucCXd{j
z1WFtf3ArH7R!OT|DE7oNO^-LL=Sfx+(_G}cCM{zBETf(l2`W7G{CnfuT^37^F*tLo
zU^NW62q9ucT9vT{$qY?B<DXP@>M2HlMF3Wj)Y^5i=eSg3W$bPU3u}$_i8D;n&nZ5u
zp)MHmMmNb<y~?Ks<cI#)VPM6I<h1te27GvEmJ;a)R)!Fq<qeK(p(@C<`Q%<`l|-~S
z#_uvAHoditHrt_n9u6%1^cUWCy9D(l%1Mpo!P+SxK7L)HjxBtFu;WhE7`s-J%m(y@
zRI>Xn&nL~&dnt}wfQs-4aNhIC)5Joskt#umt5`s2`CurVt0%M|{5J$C`J!;ic6i&J
zM6tsA2glD*;4UMzwYJnb{`%VjHiGny&fAi42MzBH#V?7t3Oyeiq_7{6z2IC-pvsY7
z?5U$vB%A-ds#7>Vr^U=F?&Cwpw}aHlX#VMr5bT(8XX%d1LEmv^*!b9Swt*NY;n-gO
zmb2qKA6z|=l^`H0(es`?LHLu}Ss>5V1!6+E5EE{bDs%E7H<^H?{J8Tpmi|1%rNT#t
zdp<Bvkq|_-<hYkIaY_!V4d!o1-P(dCs;OQY(g8#mbCg2Qjr}vHD%b<Sxsj0y;rh|K
z{fvlClC`+HsKMhTbJ$fst!}`=Js9El?@nN1pkLbUdl3sY9e65;A~iM5%$7TG5B2sz
z9-xi-E80YhCerK6BM)$AcQQ{~toOCEb7<R;ID)7TQAZO-pAPio*pj(irIf$bWjW!W
zZ~8F(PjAoc|C7ka#KiI6&6S$4$G(YtmwNg;1Qw!2n*xP@RToZvI7?5r4C15zBlD6#
z=VlIl`N+qgANyTgF>Bd-QBR}fomfDVS{tduhbvbb>P9a`FXs-1n7Tv{Hw`EGB{-YM
z8VUAO?ceRI&&$zw;_`%Rrnp!of=;}LcVl8M5LBMIcUsgQBnLk$1X|b`V5VWp995v{
zUcx>7fjJDrCqYdd(yNj4`S7|Xh;H<vN2wy_UW#$nNSGDem&gL9l3g7=y(!y(N}KUK
z!zb9~byG))5ZUL>Eo>bXr-ZFc7xffiDPAK8Q~dyo`rWQTmJmvYl`gT=W)MW7{iGpA
zBfT2;-V{Up5|LyvY>%p&Ur67FpS~tSE|f96yq5v#b<RVsh<y9H`oORXu&J5OVtisr
zD~-x_*fCB<uB9BKIrM*?5yxL!Bstq~-}!fb?4zD&0|n-yccs$ST_$K9j8xZW0B4_E
zNEOWBNbeP{Z|GAdKv-iTwonZ#8bW8yJ$j=vv?1y{jkDu0D8DP5idKQgV@g0CIbR0(
zwNhVZlD+o(x6TUoZ#4(H6tpY);KW^I8`^80leK4ATP)&V&e)<1%?xhdf0wW#7eISu
zr`e%%Ox%G5QO>T8aU%)d9^X%2wGwgrNp{8%L5j0WQ=31&#zRUpa}&h#Ud>_Wdy;dv
zu!`rNzuuTA42x=YvO?2F9XU0v9wgU{$h848jx%tiu%#27s|fL6@kQNY3W2Wxlx0p6
z`u7|VMD_)rytu^doSnh|As?-s9<`nQ`}8(EFOgkfNbqgYv<6&A+O@KgGG2BGX=GBJ
zOTl;wiRz;CDW@*m552i~KcO<n>#U`sgEEQoklrK(p?X~*E0i%vn@Lp+42!GTYpsHO
ziH~A~fdo%Dw^#R-=`fsV4rkZMe8u>knMq6(S;dAa(Z{L_SB~t&>KQ@z;!BxkVwaRL
z$}t9RzG-OSb5cT4`PH!M@2WqRDq0$?sRQ`#fkG%7I111JyRM$p0T9nfF?Tp@QV2@x
zt76ALWZ3B$C5FcVPBOSNTZf$CzJ911t6s?WUFcjw!%c8#vL$aWu}(|$Ei)G?R6UnP
zUVrq~21c{}X7O2)gSpnN8;I{iCY;pXII){hjxU>5kUeD!#XlPE^D$<Ry-slQialz6
zV;G~%JDq(zw5<czj-iU4eOjUsr^^q=czS!mHi!FTt-ub1sn#?)Jj<G6svV;LR_xpJ
z+HBpQc44&)S%%<50`t2^t%Gacc~iHhO}EcZsYN;Qw2OplV0F|-m%BX_svklVm@-I$
zajN5?z`f5;Q>R&Gtw-)zgRiKoBJvyIQ1pfl>C}ZqHZodQ-xzeSz_v{IWzt0RSh%Z4
zOu2+sCJui&p>)v&TN~Yz#tv`P-_`*7!)Fs#!LhF#^8?c9YNND12~)LVeOKL>4#I#2
zTWWl!c4~ZeNUfDdA;14-*C}CYa~perExHO+u#VFkW@{_br+-i@b8#n>WEZpu+uI*|
z-uR}saIPO6_x%Cv3jz}{ll}%tC!k7<3$1!P3o0bjv4HxVUDpXng)%VmV(1F$ogx}T
z_j!I;i|8M8k6|e9_&?vSt@TjM<i=oJyZg6h^QIM4^tFe#F6U8-{n;5U8<w`+n#3IG
zFFZVXH1zcR@_hbau*SAPt&yPr9(M1-A&6;UpdSS#8ouoDkEQkxW1Y>LrG8)uCiFBe
zrp?q;F(g$_|Ct`@>_P(^QAZcA!&xjq{5_R;$tV!5Q*Ye2%+Xox*_Y}cDYe_O4*YRj
z0mIEx`x`j(0j&cBbTKU&?kl?%HvNz!LO0NZpeKN|nb)Hp$BeaY%w|kQUy=&J3_3w1
zo)mRlj`?Y&H>~zA?MQ2Hon~u0_{mD|CHf!|K&aB|fo9Ut8$I1kEI&UWn8t|au{ppw
zhRd_@QCm+FD5UfwO-(WXu*teL#@48IBZ^e*MADZ^ZRbsmw9qv)AGEA`i<>6_MPGqg
z?MHX>n|dRE2@9w21Rzxa@?XSa=eF3pS=zfI!cTelVrOCb_C~8Vx+-!}<BmnwgTK1j
ziODT(Ff;v)b4f%gk4H@FXKJ0Ba{V4LTh+F0D@u#6Mkyr#8qi*vc#CWHB4yJNzVeaG
zx`EeuUc~5`#rdW?HH+oP>?39nHcL4INk}`cg$+v4a4u4Httm|U+u~}p$%<r`dQN4h
zT!cs<jC{9z`o~jn)lGkbOhT0k1l7f-)-J=q>55hbpM?h43UAxu&c;`d@5}I!Um)Qh
zeqwMmNPrjxfPU+)kQ9{B?AHEeMH1w13KX)%mC(I7wsDj}E7!FL=QNiPHEfdYy*93@
z3Y_%ZgzQ4tQW@1Jn-Zl2v`13v+aeN^qHXYiV(~9NQ5py~J`^6|!D07@DW2rK&ko#f
zn$TF)48fW20fyWvt>J6|z$y>J0uL6vS~ow3Ok;NU(O}99OL%v2_BTJ9qL2d}RpEAZ
z)~&=aEfO--P_M9e?yy^HF6P871!oB8tx>0gK_RaYvL(9p+sBL7x4so1X!jDk`w?M9
z+jsv%r3w=PS-pbG4V%c6YwB?Rx3^M>G&7}9G!$JY)A03X2U(Fw-*7Q8<KNe5_Ye1K
zcL~X%H0l9n^BS@g0m^ap{Bze+*pF@S5a|h`i;}QQgGo@tT0Zps6>fbvz&U5aVw2ii
zU_`KGA{c2F3=Un{mDE`^+fv>E=SJCD__?szaotm$B-Q#)8x=K8!LN^i%ox<Eru9<;
zNjl-1e-(*h6s94+yhnE166;=Pg@DyjWRAcXo%u}7uQmj3g7m|b+rxO9mrt-h<4t{M
z-nq#+tnSr}ix_>|cvHJ)yIEcz3}<=zqKDxmXtCO&vmyGTQ^xAkZo`BoD=yasfiGE=
z&kXa*>mppTD3(gsHjq_Mp(ar+lNu@F1(&Vv<#JQc^~OA0ijoj*?n$#)JADU)hmy-m
zg^zwu*Ei_ShCh!teHC_i+`4ZwtzzBL8tb6eF@s;)Lz=}po)!<Y#aT*bbH3t+w&sRC
z#w+N(i%<~*BU_Or`d$A?Iem+y|GVq_|FMjj*uPus{+o@QChbu>oet!UGwMsEFthkU
zn4duswqpXneW&S0elZxC8$x-|N51<;2s^e5Bh}jG`MlMS8X<(T#wF==(>uMqU%!4{
z@5t`$2C;9?-+tPHyb8{K5th$n41z#c8hU)Sy<bR5nV#H95hN*nhge=OPb_l?Slv%S
zLip~fd+!Q`!;H&f)jX=2J)%nVPDSMNj-+i$Zv?flD1(FsdA>eAYLV(0#-@PzY<=81
zRl<TSaG|wMxDPuM{UzQO;GXQ!DH7hlMvY}&%AREaZmb85*O+Ef*EcW7q+%J*?`1k+
zTM|sD<%#nW84N`*Jjt4Wyowi>IOu$1t3N4w1{qIKEsxjdmZk9Ea@SB$u?&|`&icfY
z*ltifyf~9I#t4)2B(t8Q3#q4QU@R3xNk};L(~i2KUa3CV_*I7#jTW_M73$ZRmrHpH
zcd^0~ybJX<>+!cVj%2uRZk5brco5zPy6cZnbkl1!ikzg?K++pqgfdixJO@`Rnvv3L
zAfM(Xaj72bc@l0Iev!pS*kigwJaW;}*6);BmP}%i!Hr5i6)SL|t#>gER+UP96>FO9
z8=87Z<mBt;S3X3k{phPc^T5o0Y4p%Fmu*QNUY79TOW9f{x=n8UC9)seP}Ef8qIo7N
ze9o&fq+JkW!@8*09B@T@#pTm5=(80`<I_0V#M1Yo8Jmvb(^>NmrgVwe_HQw8lUj0*
z%5%Io7t1zB;fTwUW^ZU=LGuBmco1(-SUTTMRGj@0CHjvbvIJ^>tJ#G|vA-{Enw3+%
zxe>exzfzGcwS_$dfv3wZVsu7pV8Mw+>bW%%@<lO&@3d74sCr6wxOlb1-S1iNam;b+
zaKqslRaEq@Oqyvmxhv-AN=CvR`>3ShG*AFANL!ILDyK1*7Ifn3;s{v}v!$RpyqP#!
z7<#|ZK=epThl09jfjGJ2SKzFWmbTg99jpxnK^gK=LW<&${i(R6p^XN7G-8P-r!B0O
z&ghJc;KJV=Dpv6sJn!N|t^IMu8Qr+ACJWp5b786qDE!=pPokZD=pD$>mN9K)z8JY?
zLz?t@pwgj^6X~~kW&EjQtPBAP7-t?*H7GuH;!Bb&?6$~LwAZQ+?vU%JxKQ3}@I*M@
zwNG_U;gr&|mRs**Qs+K?l@7O{P-pWxsYhU+nsW{&-NlGdW>wJ2x7axPL^iJWyN^x>
zi<uNfS-aiJNNk<3S@>y(TTT0DY3l0YVF~;-_42DC8c_Wg!HP1G<1abZP2!TSI+VfH
z0r~!AmGlsAT<i}Y>dVj>^OtU@%V8|y%cj8YMgM{)9SoONb*R><LbYVm64v|z&(7>H
zIb0=LziPV}TPeD$2L*l#L4_WV#eBZ4LkvNb$M>$6=$m`osedqF*0OCWsSadecB($L
zb(n4@Q3NERdkf|xp-S+%V&9MTUnhVxj}(;PBaZN#oRTf?5xc4o3A(9BERL#poEsG$
zY5)HPK|Vmz>N{K9LPTdp+YV70jSqJIdyb2!w3T&n%``ZslT=SgafiPWZ||0<t|^t>
zltI>7com%-IG>|2Gvp~Pk5dGUvirPUJ7UGJLPpoow|2@-H)~-XYSavj*55PseY^j>
zezStbLzPvUSjt7ZfdbE%2_B?2vU=Ie9cu}tXP)u(W1&>n!ckp5@<~ZL#Soz#nb|^<
z9VIYlK}*%am34+8NoR($lHyX-uk(-$i<f?*Pq(l*ku?tdfnr&Z6S!&uJDW%Hg383a
zkWZS=JR%FhB_0y$f$p01%O3m+%@nPrYGyKN45)Py2U>g2%G%aPq*ttLPuV$@xS?GP
z60GW*BZ|PSxK!>?q>?>{7SNDE2^+}NBjr?xk~>Gk<si@|VO1xAxH6$KPQFQ<kf87v
zibU3xT|*mVPkV*6;f=TxLyB$7{81cWUuZc)QN^g1(@Isr-E_|G^A4f5j`3#8+2M><
zGlZ<V!`VBhBA<4dwTl|;lKTkMe{1l!BOjl~&tHRsPdG!E3|-F6c~(y;m5SUmn&_Y_
z0!i|roqT)9Ig<5qEbSynV7Ur;@$r$)O4{gJ8fz=t+XC|B<}`8Yvlgo!slhHSk<6Lm
z<Woy2;d&Tru(+#1H(g98pP&h92g~N=#I%Sjz_{3z{Cwe92hEA$bvXBvJMox%;3}pT
z$*0uJ<>;ziOn+N;cKW$)Jbj`0acVI8croxGcUwJLEnG$?mHjGQ<Y}-I5}0GRmxH0u
ztjGI&lc}w8_j@>QCzp~jxmgG$7qy4jY)^)e+H)exoM!gncykz`;xVV&O`B&_E3<8(
z#5R^T1PFc2FxG6$h}Ni}W<M|ORv8foG)^OL*VRo;=KB-ye!_G16N*VghZ&5Xoet>*
z_a=nOL>;T@w|UT_I)IdvaivQvp8;Cssg^S|z*(KjOh{Qkkv7kJ{FmtXPzSD~_}-Lb
ziggO_;qSaWV(|5W^)c-YmQs=aqP^ISioC$S<YJ<qpIL)#EGkmqy?d(;zIqDOC=Duc
z@3YaN#OAK9mo;(oJBGV&ljUAK13?TEj<Hk>WJ^G~Ap0@{mAVs^_f#t){8j1-)nUn+
zDt!vrIh}@)zdtaSjA@NIFmwpu^a=@W1M}KF)TZ4}Idi2gk@_k%w^6w5sTF#yb~e&B
zM(QVHkha*7&aCv8YWVZGyE`2Zp))9J>5s>|^SE(W+X)%;NLIMOxw+tv>7KIGUN#R5
zNmD5Qy=h*lKl`5k;#qMoP#_e>_hS3Iv`_8gveZK#6gedg3bi8&WuNjz`xnb-)|fUJ
zU?oPlC3Nibe4Tr>^?Cb105DFu`idqVSsN}a-%ibaSwIQgzZt(J?%}2k(+x<5PI5#3
ziioIh-F4ToeQxKO<^gR|(rqF5o7@+fT5^^t6eTXiH)l_q94Hc}?a@x)L(IHjCV`?#
zx2s*GG&H1+Jyp1C{eFM7o-FoZ?eW~q<6XWApYbVbemK%*Ly<@ZQJSxAVH@FwJdJi)
z9vJP_of+MgRxBIen9}(R6k>l{hea%3oAe+>PRs6>l8k@mGiS?{1V2p1gW~V3pl~;6
z3b$tDSOt}gh1G7q56uvtQo-B~bFw?=3|RJpduYsR<X_cEQs+8)6dP^qP!Xw0dqy3t
zbzE{l<_7i)^(B!cjpuq78Zt$K%qo34RV1uY5sqBdDO|h^El}9dZY$HPx)x{wkm+XP
zGSBzZ-$<*#ozw03Y|nO%9onuVxRhQDnzfngO6Tg^ZmD38L@;vP<PhWSTkY^8G8b)|
z=1V2!jcg;kFc`cCUz?522C!6BbsEJD&TCx~9@h?bSDF*kGMAO#39~fip~66-g2P*f
zu$p)<$`+$^6=j`$;i;F?U^%R54OmC(Q$Ot5a1kv>Yh)rjLMvX~OB1idPM~XOb#uqO
zaeBXPcPX5T-(i|8H*LVb<lmZ=`4>vj6y;3IpTJ(1e0F{GO{pu22pHlCtIZE<k}7B2
z7;@RQe|vCcUQqiGf`h6uEi-d9jIGgctjgc<r09u;(Y;06Bg4?PjZQt%_`T1;ZehDg
z`+frUql}N6p47kUlIi%`lP&;tDSK(tFk8Qi84SoZ?B!QDgoOPs?;6JcsB$^k{-?@i
zV*Ky?!7eTCSYmeg%O`aXT~HG=#R9!QhP|8G1aWks7Xd1QbgB{iUl{qDwKwC-N8qh(
zH$W6oC2#MdaU?(d=bgK;gS~^&%7H*bHGkyL*jlJOdX5jIEXXD<%pWDBxO|^n@1FZQ
z7M^tXghxuV#>WlCEnHnvbu3^2-h0~KlLf+ORu=JyR<+Ysv00iMG5frymDoAY{|yB%
zkniz%fA2$Wc+urZiqH&%vW-KY7}yD}v|Er@hl{+;H?o+h%OJi9Cy7-$LH@*}*ql#_
zA1}b6{8u_h?R3v_b59j@mn*?^@c(LDg(4Vl=?;M$G%<SL?5}h23&zibrJ&)tILDIN
zU@kJgR8le~nJJX9fzVX8>s=4-S}B?~V<9;dDBq9MKIfDm7K%<%aL%dpqu-FLzdo4w
zMThiut6K8%9qh=bhg%CrxKT)bi#65EdG#75nJ;?BI4>~WPoHm`J;q2g8uhv2`)IRa
zuk*S_XB!RHBdV&M9_eP24Pv4PYL@kqLe0I7X~}AC{qo1~mrq%uudVo1nN8){vwqUM
zOoh*C*LO&lU_PiE=bnw;^T?1R5ivF5bG<NI9gei)-yhkz&I%`NR=xcdBg@971m{Pu
zUpmXk^;aDqH`%5$0}s}fzUQVQq^IQZ026EQc(2?Mur#|byZkx6KpLF-(0c?1Yr(Aq
zJ*gdic5#ml{{zLmIm3VexEL)2$}Gok1=u*XsMs`|n(+s@Z9qPzRW>X5-Q&+8{u&UD
z7?o9Rt#2{ZbY9N0Bm{4cxG-|Wi*AI6uuEg9ReWCH8)PD-?75SILY*Q~oY0h(B%)eh
z;mT`j1q7)SqmqrbVYEuedubC>HgztcDx{o!pvftCB}yC3@Ta!Ll6uSYf2vGW`M%}v
zN(ZA@jN>m!-OM>X&tXxXoufKSv8nE<I#C_$h>kUawb;7TV-q@}xkf{kWaxpSi>-R&
zf2tob)bgu|$~f(5Ix6s)2YzM?0n}7bnp7A02do`ra}~GbmGIa%;y8?TWVF<hB3Usg
z9e&IJ)??jie;_V8_p{btP>zC|c0<|LYUAC<^=mA83oO8ho9?TeU&J{4`s4q{QQY2M
z5fuyR*Ho!-$ee%joM26C@>D`youDZKzN{C?c;)7;#NTFIFq0RuUCvkQ2#lj4@`+aY
zG1Ym!d=3>@OF+V6qPdq>?VXQR3~rv4rKvR1q6ckCcaRC15(TXzma>i=p|w-M1C4}f
zqC%a(xHawbFzYCFh_WSh(4h+?^k2(|B!PdtTI`RN>@U6{3s2mhm8p=cRL}!pjFdGQ
zR&qb#*c6p6>g`B4B2-@WHmY(dN!P|IeB;0hJ@#sJ54y+cLCK5kawlqIhqZaQ`vKz_
zOss(iQNd;vW8~W4CZPF=xQp5HSfF1{Iv)IX%M?bqsCh%Wzx((}ddW)KNg6mpB*Tko
z4kZ04!v_1hkdArl)}o1KT0Bie`*I8Nr_FSL37f7Eezw#Fdpp?5V#@dS4tS?YuF#Ud
zrCEvjFsEM^NEIZP8f7;3mNC&;^x47aM&Jip1MNp`nz3|PJ7-NKJO}ZC6Zuriv6A%1
z0ZcJXyS-Sm#>EMteiGbw$Ej<tC@Ni8b%iY2b>`F-`)#TrgNH|IO2VF*yH;*%0xh0%
zkWSiF$z9Lp2_9o?B&8Wz(cD+{(b8*9Haz0BEPXBrHEAf{2wMf)AASvVi&K@Ou@$D%
zi6l8r0Vfazky$1v4A=35AWwEN^=FbIRzh-fF_fKhf8oHH7n^Lipi!(DlX%IUQVJ(f
zy&)U(ltbb}k6y-MVj}IF=gwIeZa4f1e0}tirZW!197D<bv~O|7%5nE)luXM`s__<f
z{pSYlyS;O4gt(}-O=&lnFv>en-kbx4@bo+A5;6)3o3+VxGOjMR>h-$zadid6AxUnE
zobzB#6;~dXkmb!36x$33oM2;@FV@od{yBzUA(Yr>Qsy`29Nje)vF*rF-D}!CR#~-V
zb6!kM_wXV4b@_F^^R7XPhumJj>C^k-3NqSSB>jjZ@_XL9emF>tKaQ&$a%;7PFYRwq
ztANtTpbQ91u)=%2d^tpwH+>-CvC!}OK{9m|`l^53XjgQ@U<loLE18$32*pok>OABT
z#t?7Cep~FmQN(7*gU+}|XVewl_6M2_h(eMqW>hpT&4xo?;Xr3`HG?5a1^VV0+6{z_
z9(Q7`xjrIipVWl!d>V95L5UGQs|yctlL$$I!R__4fw6%8Cp)a_zg8nGZjGhE>9OM#
z1b1JnDhmgz+!Ee{@K6rFE%(ZmnOa|d23v}+Qbg&DC1xSUXzuOECj#SVx-k3pIZQB%
zqNBG9r+b|0{UlubSWMWg3Jbf3(7ai2RMADtz|NE)R3gU3zG1b;4joe!r1T+C-Mn8C
z^CYJAT4Lv&v^#l51|z}n8z4Ah56n|i3~D5vx#7bF`n1aBg66o`!f>uD$voYiWs?Cj
z8SthSkx9hIo&tDcBaW%sq|$I5Rf(lwBLImY7-<gt6`hvbE2>F7!-?`D=>hYUQ7b_(
zC8Gxu7ZsjsLS}z+mP08ZR;qg{{J<42jT~@pH!G?UoDrcQzsR4D>Bnf(m;Mk0=b3~(
z?F{3!&U_t$!t-qr77(5Hg-MUzpo78yLKO7XLp4)N(75yC5W`s;&McNaFc2`!N=ZI7
zw7^9@ngt7%e_I2aZr_A3i;-KFm9uB#ut-L~`!nHozIO=RaN8l9+q}>G{Od6<HxdM!
zvX7dDlU{VcD04(aZ{WmjFRm)9k9fB7i+<VWb-Unsawi|*XKgG6cFZ_C3s<nLl`Fv<
ze0cA6YrGh_!F|sx(#W@%K7C@^x6;kTSD#M+!6ufF&0qbv@Zcww6R6I0%3fL?9eHb>
zzN{gV<OEW+ThkspStt=F?0yWch$m6A2R(L#f=~a`*<)w;|97VUA7_t+nf<?V_6EOO
z%KmrOZXbyy6M5ZV2qdAM<&VAH{KWO>_5aD*kzut-P?lc=N>#`(VU8Kw6lHy;?ccAh
zy_{XVsh*6&*!S+-KfHn83(e}s-A`wXe%pEmudmLp!yVrvQRkA_6eY@UTkqmz&D&df
z+Y%h}bz1vRti7CYmasA4unl;~86mb4-0j~jf5SO}Gz*j<0#IG<<?8!+B?fC!21N7k
zVSHWPon7$TN4u)@Ei-8DflH))J{)O#=l_E~>kFa)JwB=RlmE`yWh+-(|Gg@rKq{iJ
ztY!=>n1(feH>pMewpCCF180gLg#$2Av=S3oDE|e$v~}-qaea0lGTvql4{ll3yo7PF
zr1_HhK+96M@cVr*{w3%4v#t<G;B;8r7Jp~hNa!w#iwDh+adsDC>Cq3=QT|a-qKZ*T
z2xC?7)5K5~mVpi~i^SjXfp($bM##FKa_r2JSA#l0(aoY@!a>X;(p!ury4|oZ?2N>9
zWOnGoa1sl}5%K100%)-yb^s7`(x5sBV-H|Tg*`P_Wjd&iWBJA}C@t!*A!X_n*s;2R
z3bL-99V6%@UF*=`6Mbm<h%7`|>IuIg)m@3o(+TQfTo-N3{YeC<l8H)S*+`=`k>?d}
z-1&#gt@ucO)WpaYPFn6rMPDsj4JqRw2SZ06yz<R1Vx7alF5S{Oxmgz4(hGW61EXT@
zG~3@BQ=6_cCqy#)jf22_I)B-am(k>W+N@a{@<a>9O|)Lj(~=K9UUD%t&Vk~R<Im)V
zt*r2w{KMsd)q$W?^4QhKI<s{WL1}QQwEs<FbDboZXD_3o48X_)=_``zwqvHkCU}y&
zI;*V62wK`57QXrL6icqL!i!Vd4IItPPoRyJi3!dOQ0uNj|94_GMxZJ6E2+Ts9(IZ`
z6HEKL<KH<K%QT^2AhTEdYCAilH|9e(1|L@=t_*Br<PQIJQ7`8XMMxJHC4*G>PrcYg
zn-JB7VZg8QlA8j_+cLY^KUvIeBLYqqgSw}CifAK$^tk0b$SZ`RG=^@m7{RJ8yUF+@
z9WwuN-%5V@9nvJ|E!h|l(GrSs;;O`URr=_x#^3VJ#xU2p&xvAEd=+LE%U|(7u_Wa=
zP?IV`P*GkZ0xNrB$q9haD5gYNI`Uivnn}BbsJFGur*m|}AK6M9ZlD3RIfj!7MC3DO
z5bU@0(HJnGhXTo%POkY-bwdBBDCAs`s8a$MMVZ-zdH*Qinxg+vUS)t+V1te{ortXg
z{Z};|mWT&|(g=>Z0j2q&NmCYaVDhp>$5hoX1W8L1?9zN%DUDN3A?)Oh#-gj_Wm>tB
zV%V@w84S>utw%%I6Ej1eh3!w|^-zKLg{wYcxMFBXfwxUIPuitjKDefcSK#hr4_4N*
z#bszKNJ_~AQgkiv=LIN0dneD3Zx!P+s8OJCklty2<EDpqbva_&2b#xx-jk09X#0ix
zh#&G6$DVyDU4!XOqa{VPs{pc8ZW{AviCt<bx{nRp+rfpP{?*;cZ|rl*BESpQQk2Ty
zrphPoRK*%QYPps0tSZH5wqBmZ9B#zcg4XY5?IY!j3~7|Ad~;!1KT-+P&$4xFcn$sA
zKg%mGZ7GYGv!{?uPuo+ea;-mNfT}(~sWpfcfC6y(+M_5=%$mA1wO!69!`=T;f2pC3
z9>JbjzOcmlP(kY_3-F!^vGP$i>~;X1yBvjWEF--B>}UZ{W*X>8ml8qN@)&*_^P7*3
zrDNQKt<uX9LTejfG?l8?WscNGqP^W@Io)iha0wF6cH<-V{UbLRvalT=UyXh%nZ9%8
ztDZVb1HJWFetZmD4wEiRlUNqrWB|&)eyBdnSU<gy&tqL<0tJi}tNu~o^8J_}zod|p
zixkXuq@()P3)JqS_hc4Ea|PGOGvzu`hD}@fu%~dHJK2E@17@yf?zyN>&}aYfnlAp?
z>FuIk-dcwD@evJ~!^CC#;#rOtxRUahG|;V`rKu)!!{D$J+L!2Wr;269L4tP<ckw$o
zSC!*5dskJ^3X^_nZ~QRPOz`Z6q<cBtq-&+X-WB955H^*|TsL5Iu+^B(41u2tlhG>{
zHl)?YQrf3S3D}WHKv^*}!h-NqA;pcBxV3Z`Rwcj-r~-YLl(WD#r<w5}C6ANeWn72~
zZMk>ex2!;B5<kVGv75|{L}h-gzEMP3ddxlx8O54Vl;a%dl)e7Wzo~a+-s-q7b@r8m
zYiBcm=DEJT#FPu)Z|1gmzts?;rt<39P&f3KCUv0H%E~_^gPOF>>>}eckEDy=b^ZGd
zd5y=-%0g8os?qLUwh^%qDc?2gg+_}#V*i4yMJ9AADspoJedBZijACi$d^J`BA?D@6
z%D5`K^%1&(c+m5wvco2~G5a|aufnZ#C|B-jn<~o-M8>HvRx3xe>?bV=k!kMcMr4tp
zi|A|}DpYly!ei&-Vk~ZMv$+A!$iqXzM~K>)fj`h+pVOgAK7}0eVWM0bE44F}u1jaG
z;0mSU&6om`IxBNegVR*rd!)wmHWm+%F9OoovLYBiyj-%vPY&eGT20F2c6k_`a{Apk
zvq6c+97SIC+io_wp)-XpJ_|aOOEGGfTJ8@|k+Zm%N97W($!S&Hj5RqA<7mu;S=mPY
z8dX=he1uEc1&EJSJx4qgtDtogqP*{B{+;7OVCh-s#zdKIjx*0UlB2861hpA7N!j>1
zbZy=FQjB}JKj79CmHc57DbsKgY3r@6>yfB}4qChowX|ZX_(lr6CGKnCSgo<8oo5eX
z)2Qe<qyNmi0qO~{)?VN>d@c)<VWg1xAHKx}wL%}(eiF*nuwX?2%s-mv1Zu>Skr$gh
z6lcSs=KQ6^VhzMNAM^y>c4C}W`MX?i`cl8y&`?-8Xdze9qg4<$?#^TSkwNJo%HZBM
z-t?N#L$C-fiF!zTU#YdfC70dQKV<cc9Zg&WD%@{n9e<FvQ66_Ho2_zN(2Q?3@KA~!
zTwQF+E~O8q>PHm}2Zna}FrPZ$AMpCn)~x-w5LuQlylmA2<GN2x{OWnK7ElELAZ3@s
zJW$VdPjMD#22x*MNs+4v$a(C6g5(vJRrvn<AQ5|s!Qr+UVtz`Liqe;W14OkGKQFDb
z8>(Ej+V`39IBo?guJ+jXLGn9nffpgurG4G_pf;b<x7+ury$n9dK3mm(f2!tw*$qS7
z{C)BH!Ggo)%mM6?9nM>Z^?2s*qg}AClDH@RpGUf-|4Gs@{fA*K11IBuCFv%xJHDC3
zbz9$^0yMF>i28pF4KB2yBvSi-3sw-OQi)!EMZ%Zo;|ohRcRmsYtGfX>uh@$_dbl?}
zZk=A;U!49fL!f2DO#66W!FNsHFIi~a*T4-?*vmfM$UnUaT3_FD&mc85r}+AR;+>qt
z)Gmsw8sdH&S2GST{UOWE=S6>QSbQ?4xd>q9Md01K<;s>{J%cA=OmctU+grkz7(h(&
zSDh@Kgi-;w&V!GwJ!a-Bx{JBybYt=VymIkPBsx-+fjCW-W{LFnD)mn4&7RmkzUB-p
zBxel6comFAH10X{?t${KUrC($Id-{zHD!YLN(@1BpQmfkp7H>~NI{Z7YDmJ(!^q+W
zP7wzsIc}Qp6xS(<9-JqnF=Q8_d$ji*Wzkn>3Qqppr(?zvb#3k?TufTq`)^tKpxPiS
z+h}7OXRbIMih0fBFj<N*(xd1PE7xtN>9R%yR57tiI?>%GFn=~BIr)N9#{In73rJk6
zQvM8+q_CxT)Z{Kz>w$qmTkqsw#xwM9lQ{_0c)F@wI={+2T8fWb&iDP-=hNrStvxXb
zlp-l=JO7u*DE)LCL6C^!pSsiKkLSv_&)ZD!v6}Ma66Ozr(hdSYH=_fb;f$hi;fpS3
zXT&b2fleq2#?sRhYx>$#k#uE-hO}P2bs-l$%}kPkU*raG6mgnrX(v^FurCeC_KC-+
zvi=l=fqF5jlGHc~m(2)xjRNiS1H~TZ88wmj6+EmbF#U)gR<oS`?9I$rivPH@Uw#LU
zA88N`x3YMu4eAK(59r~H#%~Dr3#TM#Cv5u4f?{<(Q7~d=^45>?br2d)5|#LSYc`=e
zC=zjb^kqLfA9P7gZvOGYn2%G*+a}E4Wn!PH0APPgir4D(8=`!VN8TP4M0S$TXV@2i
z^T)DzMLW{Q+93JHtJk}W&RV+rRFQKh^n~2^j!+QtmGqv#;mEz(`Zw%!AC;HkN}B^z
zbyt3(k<@}N$SJ9$Gqg&wsF4SkBR5OumP~+GR^4UanKTx<sS`oe!0D6^+7Y2O7RTX|
z^bS>P;8?0yFlLTqNFc?(8z|9H7>B6kr*L2fNh`Tve*t6~lB)sA8z5awjZIu#6vyjc
zTX{*KXF6c=mZB)qAR1EsYY3j6yg{WcGi5g_>XgejJtc$M`ZTj^wH<A^t@SY)q^7fp
zvzsK(Pb$NC+Sp<K=eMpKbdcpuR?$(tVp?RGQGBJ(MgEpLp3GJzO&p|@(_<_RBA64{
zK8eRG8EQE4@N$|y%1~?#Z>KW89TSxrBfsY?8hZU%F)UI{J=qVFhqyR{QEt`-mwdry
zXIA>vTG(9RJFLBbX_BIZmEDwc$U@d8J@0A)Uu@|`p(9{3-&k8y=pQjei)=FmluWwW
ze7Um-XGq%kk*Gq-(8&+JClM#*Kek($gFN}kF1;dhbW8?Zzz9~Ht)ZcYgWp|DiF1~C
zD84Vhc-cg;X5DidC1vy}2pjtgsEhw;hi-556o;CMYq(ImGCVS=5<tYR=_bN6BI0*0
z4h84|nd4|3$p}YgQooOEHj6QHrT(Q%ZK!K;Jk1lt{iHZSr)9#Qr+q7Oz0mF9NgD71
zoCgvGYf4H##4E|Xh7+_o?g!edvdzWAarkINw?ELv2C2z6lrH82YTJqD|Kh&BFhmg1
z4%K4GGAwF1^$sym*?y#(PEsYYL_Iv)#)OQ#j@jQP;%p7U8ndl2KE|Zso0+pI6;dLz
znl+=`UwH0zS}BM8l1vJN5-_xr52Q1?IXlsql;6J?7PG=+3CTz><!$AH&03BNyLU~~
zsP5R_zPYz-Q2NV=v|TrI>?x*+ii>?KRKW~ELPv%TcYEI`%xZO(#w-U_2thqoXO<_v
zoC0+a5YN(@oEFT0)iX3m(_|-;1VYgbCnN0t@GP}*xOw33%H)yh)T(2e9`3D>l+aWT
z^_Y0doKK*#cfs1YAh3bCt|NR^QZdCL9aK$$_A6Xu`XN#3oW~DL%9X!#AmoQiMe#88
zmj$9Rgxyx9%_V+4tPp|D-v9(4IQ8EoI5BsPCMilC(lZ>IHDcZJf{GA^n|B_pvFtef
z*Wjrq9#;WR-P?NX)c1>XPd-&%FjcgnSj}Olm@KS^H2)B9z2Ug-?$|^tYf62Lvdkrv
zkM*z}dAg?x$T7++?zMezV5PwDWJ+jDfAG#YQE0>2jmB+&vhPcMymg$Ql<zrLpP{EL
z4-87`NwRtM%P<uMV$+z|L&H1>KYc9dL6^ak$?V48b%nLvH5^6#f?hL<9wL`afVn#F
z*V;5Kw2#s;x<(G5oiycQnpX9`LHNmabZnaTWeunODv#B(SlGQjc{vC`fgP^#Sykfw
z_S3$d^WQsfqFq(H4pxyj%ydh5)!51st54k?b%5a)c>*3O*8u_#@rrF}9AT#Fe#j}0
zJd!aS&r69_%J!v%oqD&3mh^_uRON*e!~C=+rM5&SULb<W;5S}|(F<$bke-j(3z;5W
z!Hgb7eljC>Qg1s+_Og7%nlD9Qmr5asxh(FN$3yztopx6<;3olYTfL-c1T*MZ;9Vw+
zKl~DlhFZ5&L7J=@9h<(SeI#&N%@8{Fy|W_18qtIJVWGp-ad&c_oU;gaI-R0mgf3)c
zrVz~m-`3b>)S1MMWSDA<GY@S>L^z%>`tF~XD0Ao?$qcm76~+W^oY&SAYe$8_TfN<O
zR|_m+2<MwGXxVLe_4<Sp<v2mcI6D|SP?q|9Sh~n_N0o}oLvn^ZUcj^SWKx!oa_a+X
zVa;yue+mya&i{&t$;tA6iI}X6|6Ou@K}*+On;pexs+O>hek1fc;ZD>^5nN(Twj~mP
z!=h{87MQ1CQ>=QX;a>%?^!v!kcOm7j$60YgJ^Hu!zkrzU9Oh=1O)0EJ2DI97pXtJ+
z-bye@)ZmA`B%2AKspp`dq`&=zOqYldZ66YYvQW1I=Y`0A)5^t->i~(IPXkd%29dg;
zTKEcsY0jBpLaPR=k)w<T-x5M}2969Qs<96Z0;AhJaH>p|Q>_RsyQ(3W9`PNbBHCyR
z{H>_z@@tVGV0^SsY*`)p+}p~fQy4=Pw%5B<`-SRaqw>>=pC<fCUvM2lWY>{IwP`Oy
zQ5IQVMu37&EqbKT)xOn&EN{50gfOeiOhu=kTChgo$!arZKCJIc2|kGZBcg8<u;mZ<
zn1dLL*da^v>_a+K-D(=R#q-rIdF!-IC@LMa%RIE_g@AxYtGztdj#M?G@;Akq#f)09
z!KL}7l@{_!BSd|X3O@UP{P}ipDplHAJ3{3?N#AMaH#OEI9Ut9om;Ci3QJnI%fIyA1
z7k30(4xA?`DU|KjT1YagRo+@!07bhVY;nfa8iV**&#r@N0UvBARi(akS@UaCh}LU)
zAaj#JHG?>8)9$r!pKaMPtORE94gpEjY|+zCiA$lrzUHVv@D`Hos`?hnqE#=AUda<r
zQUiJH!(QIN4Id&3mL5nE4&jLwC{Nb?{V(ah4Irkf&b``_5V=mkfRD`jdX|bVJ4V=I
zd%&66u2-%fpZ5d&Q-;Gt56<V^)>nSlK*Sv(rzlLRy_{SQzxy?m=Q<0O-Le`py#A#8
zVPt77x%gb~CK0?m+1P=LBeuoE!<BuiKEJM9%3Pc}LhNCw^!N?5HA25q?g4T$-&^*n
z{<hF$Q6;^@FY8k)Bz$;z^t-~+xr2tCQT5o6t5W+^>&Tw3iw6Mn7e}yQ>DbB1?Rb8{
zBT;u|&cqf?Oo=(Q%}t4Y>2Sj+#$C&lw@zSX@r?7mex7-qmU=fo4{!F9f13K-me=CU
z)W-2~3%Uis9WW4lgwy{jVEAm*&5=X*nsxGW9lMjcBkk>fG-bFqbE`=faoQ3s1%S^@
zF5u1|e=a(D3MsYMqw(n-ZkW>#3C-MzO=);JxpsMea@%^|GUumFt)2R0gXf=2;qUT5
zjy<dm$=&Mo^!ajsJ-eP_-{K4B&xNO#JERYRk+n^rm~-@?WGm;Gx=4#rwc<^AS7Nps
zaOo58U)i;AXXVrypPGK6k8Qa57*5)+)kt*4aggO-QF58b`!EUIQsQsUy<3_aG~v{*
z(kZ^Q+8NX7GB`T$Jz7Y;WLFJu)tEF<tY;RB#xls14qx8fG3-5-53N1gnhLG$Gv_p9
z3|e4{OsW8y)a?oo3(zWph!i6C7E=8AM;(=?;&^~6(OyP*U7XZ!`(r|^#|0Onvufp6
zmIyR*`E3jY5GD%z8}aDx-0S6G><@+pnuJ`|tG6BF7Jr`A7iN+{s>ZO6pFueYFX!1)
zazz#>`Z#_6HvTfzJL7EO!CIz=AujscLFJf?gX#h|<`kR8gLs)DNwKVOdNH6VFm6GG
zxXdSz!wgWK6dTJiwo<urH8+UFO!K-+DknIMPj~TYEct+5SL$UmNy{-6HbtR%kEv2V
zX!H>B<olJJ8X`Ynz6T^`i7nq)2o}EvpC9&_XaRXcV2BMHKdIhcbpsc>9~IE4CV+jC
zw5EX;ZPN$wHZuOn_DeGWxxM4-hgM0D1vZ(4QQ_}agcV+Kxqmu^AZ5rWVnMSpS%d*d
z#$Zf;jyZ>iLQ)C|dE0?uQG2yRXqiG}BgC?^5TD3=P}h)uTE--f0~RA7WqSRZ&I={&
zwAX}{ts*3JpJoPz5@&Pq$Mp@0z$7Hb5MjzUgM!X=80d}RYmHTEGgBBVmcEbv0#9lI
zzD*^0%a6_tMR1m%v3Tg4)RQg1MdJm4V<cbm@L3Mxtq|gE>YInHFGLk#)FD~xJOEm-
z8bg*u(j;Z0EF}t)!ai>?L%BgaywbV9sGhYzh2-3&7#Ryzvw;n%I|$A)mK9OX*(ukx
zF#!1pjG%^1$jT0@Ri*kza#968z{=#h8e?e4SzHothyStd&9y#FRxe9t39FGUmVu__
zV#m2i%}bumS<X-xX{1oO*oQ^bQQ&b#WHDGr_4`{&VyudzdH_K4yI7=SRQD5!N)vvM
zjGoGll#irj$t(^S8;6*D9|tH@?RG>zmg-r+g(8~lc3t<=cxXi9LcH9uG`7VOErl-$
z=}P+l3jOpkUBmYO=Ow#MMNAcM5DOhzm?n8B1!V}4NQhUlhLS|^HhZa19$l&ISi`<&
z^q42?)z&KhRdAiUg5)XU?nBJ)c(i+t*2$eN`}yG6cJuTrPIHZ%raP~xCT1U3_EDiI
z1dwt@t{hX-VVdMw$RzdsUq(SplJwU?X{bOFE0W9=-p-m2mu(1lk~unHAHh+kN+)i*
zw2X=Xl8phEtSK7#ygZ2?t^)`ihu;TU<%azKF!m12l|}6qZfx7OJIRjSvC*+@o1Jvj
zv2EM7?d;gL?cDxu)v3Bw=c~GB)%*=>jycA7pSN(sD(RrsDMV_)r8*O|Uz&THbb|)+
zR!r`a!${k%j--;JvRbhF&yP9QUbdpY7=GqWjPC;s6za(s5myGjvqr{1mKlyXl5N*`
z-o}+hwy<*Oz{bIYP;2mcf~f}uh|}@wFphJ(U4IY|9T#FbTGZ1iYVg@9H<0$RN6>b`
zj4cos!rj(rhYfKqJL9UnsHTwNQ1HeSe0WfJNL*T$c<qR3{`{3BEqfg0%|uExs+3Ms
zSd-GC+t43I18COd91M=Zs4dfI*`VmQ#17I<JLT>$twAgB{4N?rp$qiu>mzg!ZWK)J
zqY4*b{T5)(1V@%sirF^z!LlKIhfIg?6jHlo2QH{*8CU2r`m<Ui`<P`oEAGg0KZW#6
zkmb;ZK=XXr<6@;Y&E38k>Sr!Se`vpZG_v!7VVP=n3I1W;(?59&3MM(<tpEAj@T4o~
zFy@g7O>GNB?YHh2;e|bYRj_s`2h3{+5*h+*i;*DYkC@krKM19xrqB_%E33JKWxrU|
z+dgnt0Oi-P_bJYY&kU*hR~VwW+@hM?DEYDB{(>zis?r}Xi!&*}vDJ&A3zO!L%g}<H
zVu>CS0o^Vsm+!ixd&Ncl{P%*=hd=$raI5Th9jc=ffB<u;NAxmAvW}!I8F`lknQkva
zJJ+#$kN+HJ++ZF!TgFO#&%jt@FR-JTkBDsV8brRJ)az=ZDY9qcmUYkP{6V6><Zc#3
zjE<GIgsDKL2hhu3v4&v*FX?m>wtQa4^V`om%KK-&mcOsIw-uqxue*Z`iZ~$iA}FLL
z{~iMw1-@G_s`_DKiqi#2rIkE60iZ4c(HQXN8juemO`GB0;CLy7`u|%!a<Z`fe|YNu
zlSBg8|8Gg;f7nI@oJjxJMx}zY^?D27EO_huY_=qv(9YTyq%2EYfn6&Up&y@$Xl3!g
zVy@h%R&gwr6Qv7j;>iT1<AEts^2Gn<G6y;;>JqR)6PJPK$5Ok{Vc+y&QAkuZi4Nj&
zHQU9%yHk{TKXh-*i`b!l6MZAS1Oewt{rNF=kbCG{dAusn5TAau&*G_DAUw@0y!?9M
zD{(7~nI9V@MN9E`TRumImg-j<Yee1IJLtq5ji6*ZD&!LYrDb%2IzSMRO46z_eZ5HP
z=Kadn0zGV~dkRR45wm2Jw)mNE=m$cbb(E~_*o^X*zS-KU<^H5D5=2!diZ8mGHaTp#
zd7m%Y>8QIbMvC0bwyOiG$xOjmNCUiJ$%Ufam6=5i?(jh}m`rswaf2AU@oyhv<>Au^
zBy%Tggrn*$Ou!k3VSK;;&n@Lvn#TwDx3RYrVnLg6^)A><MKlizg9?cHkP1Ar--%@H
z59VHx_RKY*IoL?>eK7l<52x#-zO`^c_gwWDM5b?t+rZ<2kvQuP#e0ZCXOK^bOc0m^
zHt`jdbBM`Sj2nA7J2AWD9x*4iuEQs4vA5V1zHgq|AxqB(nkvXi)c#HsW4Y$~m9DQ8
z+l8i&l}ek&fFF(rrP)CHS^hH>xHJz78l8@0n_C)n%%e(}rI>;Pi^5PPyNPZ~tm|ME
zs~_PyF_xfR3NcyS1Qgo8%m;f+;S8NNad4_hJ-HlXF<ALDU*B4lG<5loF$*SDI4xhJ
zFwz4{^QpbHq+#K?feg#$G;<3$cyB^?vylwLplqyO*38EGM1#8GHbl1T36fd-)E=Hb
zdlv}<lQrwb6g>n~q_la0{7fPUWR*1fRY)~OsBi<{v%CQ@1Qo+BQrE+Ap}X6(c2U`)
zsOTaXNDR(Nd>Kt3ov;a9%|Eeyzo!i-tYv3}f?D4w*~i-YI0_QB8Wwg@LD*zau+R=*
z+%369nzor5g}zo=7d{0VYtc|BUgFLag0TE&;d=!^7->pv{y4>+v@LB;KH_I5^kkXS
z9~!u<E&^g~3Zw&Ku=kwP#F1dAXm4ZSRC>kfXWB8F!bL)av+tXjU<aEFwi4nOK|Dt_
zE5?Wi9kKYShf#+O&!We#23l|1pwAI_I79R>bS#dymlj|rC<~LDKC=wutRom5F>a!O
zr5sXKh6JXNmA16X+LV<~<k0<!64Z<9I~137y-sgPzQ5oqq4U)v+W5m0z9hHmEyO4v
zt2#|p3CLV6rR>=<$8bk*nNU(tL(0}7cSB|SW}Gi=Y4>&N#3nL4X^-CvnjG96H<dK-
zwwLEDJXg)<U-P}J9)XqvszjCC>t{fYG?FQ%n&DS^lxz9$;O>4&a0T!+QW<(il-K$q
zb#U>5cU*dew%3p==R6Nz8)swT#g$xP(GCs+J`A<SO@}JUvguibm$9je8etpFCBmx8
zN*QV9R1t9%90t?IU$lWAe7nk@EQgQbGAp8!mOXN1xM3{&JdyA#Hc8>M9|T*U*AvIT
zecq&CDP90d24MXt;k90GYI=27v3=fxzD3U@l~SepVSiUfi$9?lWNX3EmqHYPpwRtS
zIP1qtHD!x1#1!nl&TH^$8J_;dVWy)5ZAl~VYaI2z?omADRU;pgTTyJi#Jq~7YfhKD
zoz)B9W?WwQ2nE!9#RI>UpKj-?%Fv;QtYn(#4yn+C!n2D;keG0ts_HVRnk+M`8mO52
z|DHvt5SC9TYN!8u0oNNLZ4A!T`5HeBh(-vMDq_aga~0=DT4Iznmb!ku13Agw|A-Q&
zn;Sa+qJj&~oad<Sjf|>f?G%ovJ|sSq%i{jbR49fr)Sw7LTN-c8+*vzZycY!(kdN~5
z%pUQ~C%#@p)Ef164z(MG*gk<^K(ek-=7?ui=*^4eY4(QV-r!3jDfzjnvL)J5rtYyl
zzrZQqE_;4VA<~t;X@wUNK8W{$nZn3O*tnZ8e(wV3U*S)pO=kr!VIcLhNMNSZ@JOnN
zOesnEbd)qxi?**YH)5<N?|?+^evP&;3>eGx*G`h45PgJlqYP_T$3~q{Kjk#lR+gx*
zwR<k>F%!Y}IoKdO1jXe~4y`Z_3uH2r)-#+6F)CU|FoJ!vX{PI1;C&S9G%v&yi?lMO
zLqknHGLp8K_4#i-8+uR!y(vnN5BD=~rDy>TDKF+|N9kq%xNfL=7YZnBkw433b1R1;
zfadl4GW*W;jH&snkPC9Jz@+vcv(o)gvM`yM(oG&kWQE~QOV6f6V(&p&I*8O0!Fv{d
zUP$6o-AlD9uFO!r1iz;-K7VZTul&BGv2&JC!9@Y}Wn1)9X2a`|)VQ{-D@O1;VU*+Z
z{Qg?n+L)ha6xWj&0Ia^X9aRdfU+w~uhSgH%i0F|Z;ZOCI7qj?<wlFEmrwNju`!*{Y
zs@(;Ay023oziV3jSNcyJq3-&<_QSvRx&r#gk*^lDmwB)6xAj1OgX`aefry{(j`IYb
z>&~fIhn@nNbJYbW6zRH^^V~<?<o3#G9iww1GzdIefJ=n$pKKt-ex$jR-2JXvmw+1a
zQ^|7H9ByicYAgi6m3}On)rQWB7!<=Toe&EJd@X}`)#9fwz4h?P^_Z!6^V%hM=s4ic
zPJVTc_7kgwFF5@%i`E*;CF;5Hw)3A6ZY&t_(=W-Is~}*1`bas#l={7DQu5OOsFNq+
zV4_Q<_!4jxf?j4<Rg*c7IZ^R<C^qXFklC?#a`c(3h11FuUz5Y6t6?tpV<D^W@)f#9
z<0y**c#pKC;=UV7rbi^fba1+Sx`y@Hl+?=2oDlVyV<)k__FWc56(s<sf1sI6ieK>q
z6Ggf02u&$k9FtdU^+gQ(^{36g;(nXmi1YnbrHuj<TTLo+if+X6@zTAlAkcyaTBV3`
zKDcr6eIs+&l|NEo)d#Z*UpFUI=g910@xhB|<ot=P!7hFw+QmL2-ufus?G0$f_?ZxT
zNbhLoH(p|uGu0qp8AQN=_(TR#D~B0`<W|U9|NIt&;Iwd8)OyW<m8;QmR1vEn9DTwE
zIacGP+Rt0P<yY5@JJU>`T|mO_5hQ(!YRtg2)yD#lpX@gxIryX+V9A%j+{<sj-~wo9
zQOaGXP63cz>-Pd*e_75>A;KG>()PxPQq6Wv9CeSl6Syx6>xgo~!Z+xSlIXPa@IU|$
zN8o8X1ffM3EWj8!kzfe0(n6ShOawYHJh*AzczvHAc#%V6?SI$S*kDM9a`LSxWIlEh
zYcV_DP$F(J*X@bA6K>F3ip_dGAAH5laTu^CGg;yn5??a&q(}%tlzd${d$#i(ATGfo
z<AukHEI;$wDdrUzsV>9Ux)a7!>pVhqluQoTtZi2EWuz!8Owi}#7$?(lRy|GUk5J>?
zy35a+jQlqN31IzS_5B}Q|A?9Y@Av<+bf-&e!tU^Y*88E-o<Txjx85Eg&FVt0V(MaM
zRh!Mdev8qD#oN9*5>J`za@A4B?d=nUM-zSFKHNN>9$uc_56$kV*6uiZybFNz%zmJ=
z3%a=7fQ;kfZ22Ml=Ch89uEFk>v^IXZ`|#fB;P+)k@qp0ie)-qlPoS1`_K56bU2gq1
z(@aC~isU0$a?B@aRRV-lQsd6~@_E0Mil~=kV;^tl;!+4-E>FjG2AgrWR``25jmB6b
z$`GtUR+pSy6kGR^tnm~6l=)d_c!g(pT$|Z?g{Qm<=M2ce93`W$g6CrC6n~V%4BBiH
zk(}7ny&YiI5l`<lPls`cH*xQyQ1LSkPJcY<qMW%zL7v*$98u~iU=JR53`+pdvLmkG
z!iZM{Ch*TAQ$1sbX@e@TTf|#rNKeY?ZenB0w8XYB?V-W%3Njr3oLX1?$DZP;ziNxN
zNd|#fC(*&9emD`JJluUE$vu3CnBbCI4U^fvKR2n4iL1M#)bV8uwe$06eR-3{(wjs%
zL}Wcs^lw$&+1izLu}`<CZz2#0%o^y_uDPqTM<R36z|cxGXKPx839&xACzCb>HE99r
zP|`?f1D87+EWh?T-x+#qGSyHC8L#R$VmwFLP_({$RkPeu@>`wMG)u|E^5uL)#Y@;A
z>E2Q37pld}>+{X$>9e2et<3M4n+(MAtOmDJpAi&P4Xs~W8>i)9lzfd85g_=wCIf?H
zYcT}r5!Bmeiuh98I+jX8M(CMJ_>c%tn~@N=nC*C@aFuCE*@#}vQ?2+|i$6m!hOvTV
z<wOCl6;T(o396uUTA-M)asHDNPgP-@Q`{-sZk3Job<&V7A)8R`Q9^5fr8&e6n8zNX
zDu>&C;h9qE!$#3@XPtA=ufYajN@_>Wn6M=$%FfU8D}lG!T2n9sN_}X9vYQ@*N9kz+
z*Klnxe5zSXKe(w)nAM|@=(p|=_o@Sm4*3pFEUI(nhb{`Gef5HmQ22@WkxL*wmtleX
z%tfFT+}KUt5_lxUzgP?d7}=aC81yC|B8J=_HJ_#r+tte{Mn*+ElCX%9x-Y3@A9j4H
zKLaiLUrND7Bqh+NY3;YOC}9f1cYk}a8ur)!CEJ3tmO;T>FUR!QthC{dR~`5y*za-z
z=}H8T4%opk5H#eC@r+x(lsnCB%(*TzD7qjI9Q^b)TmqLrKHQVgF3ZrD9rcVulHCW7
zoEqILTLx(^g70F6Ha^zhXq2O$znVjW#mMWC<E4;K13|WQ?=eHgtYXOH#58_(XdZfY
zP6r(*^>0ZFR#=s0L-t7pMv`MTb}8{rgH$_mG|Dtf`D9)sc0hy4zCM%ag6ecD1P5o@
zwK%Pa)_`cmN!2LFFS1>}=6x<vT&Y2qE+f$h7q{D~)wF>$OQD!K(h+`1v`vVs5`9!1
zR%hCmJy|E4=;YX8THO`XN2sLHSr4Du=B4d3!9WfRoIUmg=()iPLSW+Ry<v{LP9GQ9
zuFgSxIUVWuk7j>`C~|j{eIt=vp4#%!AkueH&Z~WK_$(k4A1={+qI8|L+@P~Cw_3{(
z?dF)<pMeT@j9G;eLCmy@b%~QqTK5omPxyxqKQ`QxLi{`*Z_3KD)=At~5a7X(yKLO9
zPE9Z>adXJM-dQkSVh1wHgWf^i+Z{yFcF9$**Jy0(H+A#Yb`#u1^Y<p;+HDp?+sxr!
zIASw9O>?ym!Taq7BO=>~(>vK1N1T#h|8q+z>^VEhVUjgWuLwJGc-LqmZBt%R)_szj
zxcH`7jBq=&4*lKT>M|6dQ(3#U!;T!+-0aK{|2FA5o#*D=*Aut@`^zlJ@GYzwcbGP2
z^kuIyR9@~D?<Zub9=6?|C1<(LX?eAGz+4v2{Yij?%YS~k%t?8iX6B1EuAMx*B}6{3
zD7ZsaAK?;Qt^Uf2vA3%q#jLhMYX#w)+jK$ckvP-K33^44qkarT+DUz*>^zJ)1(Mz1
zJ>;Gj>c2bvF4L0l&$=nNFV1{v#|^z)eO1p_vJ&8}ORobi)4gk*xs@oeFk4@9g*w89
zo?NY<iI?z==yMLdJfy*!>#jA|%?kL-oeAjq1)AfezB#odZCn4<PF$=gwWNGveWe4L
z9_`z^Fea7b*R(3255^`;7Z0@)U<x{PRyKzXJKoGsxioh<F~3CN)L;U9V$>Qw{jkq$
z%t_I5iy=Gb3xnph!^FfVLPN76Ms^<Q>H<tOrcGHNG<Vru#&hmOz~9;X#nroCX5itw
zi3_njdt5%wH1?)n$VHGG&>G`AD&&Zk>9<7kZjIzKci+)DzVf){%0)f1q@=lNl6M&S
zi{OJ%J?zlqSi_Z9W)sb|{xs!(;IZ-6>~5~|?JRuGVufv8F_t~0`0SsDm=A@OyVgi&
z8P(cXdvMc(nqUODeD=iW`rDQNj5fJYUCsro5lFc;Se#P@)3N-3zAh*C{%<!h$N#>8
zSy|csZz12N_CGu|2SWEu%?&Y5)jtad2ST!*4btn}z)nm-%w56z<m>w<t!5(STJ*g`
zi4gLttW6km+=t{Eedu){fuARjFQ2a`o7)YU|DD&%3kOXn>8(EHZGK=MJ!PiSz(~yQ
zGd6A0zznRQnzZrF{_VB@wu1);-3}b&OJ9dHVRo0QC35u-uMO*s){z}L-(maU?GYEk
zXNgmjd=&U3;BHKeevS#MTKLI)&S}KA*QdwZds%pO6bLr-Uum54P{H4PMfhp~4gyG|
zFvlM^_cgogK>JL(Tchn&Md7@>e}+g_UL!h)p~E!HK0hv;B4|WNU1_G67$_#Rpq|Y0
zL8(rWk&+p5AUI<sToN1;1bY@E@8!A$nxUii=kf7#fUf7aY5qkej0L2)rmtMibgyfu
zgk0|}p&1pto`?2QKb@hjDi~)#o34WibD-%U)<H5DvNl|rqy<ixn|ka?N(&9P!i*H?
z91GQ$%r0O0wusVum2vwSQXgy)ZWn+J*M!O1#&iH}`GCsGCBuhtRdLg9AlC_lP2L;-
zhqR0DL|fs@G|vkaM*X|KsZ>>wB1?krR5Bf0ZTlMQmK}Wr?u>(ocLrQ03zZWSo(fSS
zc}pVgN}WA75g}N@x9h47D~yosm`?M}vs`3O0`4_i(Q`lTP}_y2gwsWKAX8sEO6~G!
z&sq39jOj;Yh6zFa4v7RNYv|QDq_4{#O<9}|ojnf@+U28PDAn+U1!9AfQ>C}Fw4;!y
z3Nq{MXFAi~usG@yko`K#he});?@q8VDvLY+q>PGMY;{oiUBkG)mqS=`JcRIgrVDJE
z=r&NkgHU$JfepYuh>+ZcM+vs%vWyhte_aMwG<uhne#)nh9D)FTl`knS9wfX&rW=ZB
zO@?3lSD7uJBg2z@uMVfqbf=-w<mkSboAE8jJv_s#tyD8%5s$$DB_G?wb>b>vTu~$R
zM%A(aE0?>V*h@13Wll<ThF^#l%a&3^8EUCXo;~V7D|BRoQ`HXVDsm{~z1+VS)JH|2
zP-s`W{!st4oR*khtpNx(u4ZA~Or7(Gxx}da8c4SZ(~g#@F~cj$WhY$Tcv(!3TZBq(
z-X6~^hi4<vx3ZIJAsq<bZj~RdTQ6r+w5_4gd_2Vl*}?M)nWR$OXSICmAE?HRNx5_Z
z^Ve+J(k}Q{ph@N&g^cCUhv_k=yho~%%0U?nOUjK1_|i6}w``Mc3O?uDh(F928w-kS
zE)1WP#+5w-3_hQB)EhB^Vz0SD_$7CyO!wT~KU>}@WgeG|p(^(;;$=2R{&0ndnlTh_
zQmF@h5;mo?zWkvMrPH7{IA`PyYHKon(pE!)Ri(`hv+lzTDiL0fRv43ga7wy!ri~G;
zvM#wqz>5(%KwSA16VDd)OD^D>0QBTMU7j9`&_^3iG+dviEr7ksQAfQlvVx)_P2zVQ
zsnO*Y=tq*}DlRFi1`n8U#^pjJR#diarcPaPsSl-mIc?L$3#F`$)y>aNR@t}3HqPRo
z5am5S+9{X`39`*~sy3wQXj&XS+3|KrOuP*a+HS3I_Y}g2ak6P93BK)_nSn3%LeX^`
z;7|>Hn2YL&g&e7gT{@>U7nlaaw&(19@YVz4mep?*S#~Y&=~fa&gl5e^Wf53i!zF*i
zIv!!&V|Yc9#!65>d%X=)yu0xD7>l(_Wqj>p7`NipLlVFmTqwk9b>@?J4CGz!^!>a3
zV!?T$D4_0FMETN0S+t!&PNiNAd)W+DM|Tsw41!MVxfM+dRK`L0fh7$>$!RKGb8yj>
zlXq$bxfU)N@zD{HT1S-S)k=s8vjE`8<*-ZEL){FqNMU+6Ob)$mX!%cg&`N1px09Mm
zu~6h55patY+c0sp@Xt)-%Gr?AHLV=f8Q&>hg54^xP^$BRb46t@joi}|N|l;nGwInj
zgRL4im|nJTunjiJbRv)UEDdQ)x|WZc*&n3DtRKuQ^=oX+MSabgGNWirE1_CJ2EBAd
z9#Y7op&)_nPo2i8%Sb{b4aQOme_sSL>+lbg9#Yp04=@)Z2{3FgGZp&n<;Ev7T0nn5
zl8)_T^Z82xl#gtKCv_(`Z6l25@DCCBt!fSkQ9*+swSO&pq=2Q`A@|mxrxmNwG+=oK
zbq~5QL&V19h(Bs4xkFw2gB}1N=GcIN5RsabA>;&%%0p?psI$aVkXw{`US3<e|FRUd
zX2872_AD%E{q%PQU~Uu6cR|o=N}NCVuTC?dlt`dIj|v6+OQ0uM3tz8fcTj^}P@&>7
z{Vv}hMrabKMIzs$6=cv%FBc{$j~c=mk3~k%)@B%8w`#`YScx~N)w%hXe;ppEGg@>*
z7f>|CUE@E9rNrx{mU16nhEuC1ESx<PX$<U&y3P#)OENm3f=8?hGh7il0+L=6tA%+t
z<|2`Z6oMl6qIJOwI_Chji#ZDb3-ll>{|D#2Fh}27YlDUAbG*s(%UnagyC0dv;8E)a
zlCTw@k$cO~G$KSf4&f6&rTz$CLEhz6OItHHrf;SBvrbd=r1OK8w=q6(HwqwoGd`<*
z7L+X+)3svTcmxuOhx^}IAJ#frz?~fc_Rl@k1mt!9%mN$Ir3tZ01&NO7T&{tav7FpP
z`vX}t1s-Db<8C<HVy~}mQqtmWdT3rLrc&|F%aW7rMvp@3I(x|zZO+;{>5A3sguafa
z;J8KJc*cb&pBD$qJo}}=S%k-gOz9JiEY4nS7Z|6M==jGz*9!lL^~^DA``?}0Lku)e
z1c>erUx9ip`#p1~Pwm9l7xp9l`Nth<5RsHWTbeByqrc6ZGlWpIL!+Owt86W^-|u@$
z{^1|koUc|E>9V#M7IJig*{XTnnj-HubuS&y%&yCpeKGVrnWolaLt}%pf^8_<RpQT%
zWYJm?h~eXHo*z!Qs1Qym*|BxRy&0;+V>R3F{dV)*EZ>d4b7)R%K}6DDBN{*6ANNm*
z(Q1FC{8uIPzcJCQoa})AlNYi0512*4jnwr{e+MmH6}$#A;LkQ;mus(Ex^5sJcqQ;?
z`1XzbleRP}oU(<^0LhY<oFY7bv_wKWxui$_?%{mm^?ial2TEQKqM}pa-=5RLo8ZA5
z>Aj!k3VO7=kvzWN?+-QV|ItFYf+R&hUT$C9d$*r1Ba}e3X(EVyDi+qUk?9`$IK-KK
zqqN^}Y>Jp)w0Ezc9|+$6iJ^EDsqw?h?d{**afOu_b47wc=C8~B_le*7<!5U!=X+yN
z)RE35-yfgxdj9qOuO8vzTz^Y`e{p@j-tnfa;ZxDDN4iX}^|7qMS`Uo1(a3r=;SA^Z
z`-gyJBJ2z@$S=GdYI;%6c(SU(<wkF1XCn6R%!<KrG+NN*X5X}>3W2y)2!6eu4`%zk
zH%YmXfO#<eM-BDmHidF>p^tc{N@4rlv<bl9fJN3cIJ8(tq5awYqn7xO38e;?CTh_E
zLztaAt1nQXqstS4WJpF#`C8b;Fbpk|@I@fL=@OFbl*kgMZN@6^k6fh6VS~%r7N)g9
zB?vryzcgcbizMe?k5+=)4~U81<DThr&x_ssrqi}!46aYO(r#W*)4;4?+}~v4gnLtB
z;*8Em-D6<n#wawH@H_A_MHL^@K2r8im5lns@vJ#%MG79-pKbSpIyfo>@eU&}z=6l%
zQgG>#(lI5ry&Ze@rh8yehIP+3<Ux<C=!Isc(FpX9BZS;dGnnJ@Ne^y2SGdT_et{H+
zh}y=uDV%4}?%1@COFrpNv1{1GT45LMPxDqJt4J&*NT#R}q{{~mD}wa}HBlM0umP$p
zx=e`M-Z8u?$qW#^=hT69Y{c87?96+F9*YbM)h7w{b?yW{$>dM1w1#2_se0cy@R0yf
zwTz?ODr}-Mk__<I@1;q=9jcJIY1f&LvI*s7C4&Oc%dZnx#*LJtt|%!^Ej5?As(M|g
zrI7<gObBC7nRH$2b=kD5p7I^lXDP)jVPvg^HW&I&PUH4YfcDy-Xpud=+TtpF8(yzj
zj@#7*PK`Vkr`B#e{T~0NpfPCWa1dqHGsR0N_WSyfz?kYm+952GUhPtzv`#gx9_1Oy
z)JtsqR8!163MNl}Th6y4{pGc@fETG!bZ}t9b|j}UOe2W(3qq1s3;14gOVrlWInlx{
z0TQ|VD@o`60D-J9q|3ujiJp+vi!V@W<B!L21w#q@m25&icehC2no-IJW9%_beJCU2
zgY~*Ki}Z=qX~O=93SU~dj#&K_`(|(5;QoaPiOQWcmnbMWj;`P4plZS&13T<K8jhsG
zhs`q{eK9v&oA|07FH6@Wf@5sjK!vmLB$MY^Qp=jA>EEHg*9=Pj?947Y6lgNDV%T)-
z|G=kme@&?^%SCzS0|^O9cu-BNP*bfs49$5DkcPwkPom!;?t0>JvWj$Pj{1&-@U_i@
z#E>Z#&M&{Ys&|Q~4%zn%O~^vUc&5E|lU=F#BgC=$?MWiL8&z$bV~-D`$d(G$^y<oZ
z?@TT?3Wt7pzw9Mv$(UtFY{z5MbNy9h1}UC9I8;<Hc2qSs;ja-E+MiDao;g5JcrW7=
zz#v3j7jkDna-L<UU83Oy(qtwTb|s|fV&(jCzFdw0UIJ&G*P<2V^mvbK4sKJq-}-@J
z-)qAqnAP*Ftyhwn$I{OV3-r$X+rBjze|o!QFU#_R2Wk27iromb=14rFI#ZO-{grkg
z$HW}Yv^;RN)IC5UC!r=1LU)GXHppN8hTa&l)E?_WFc(ro3UA>^lDnX1W%}UTMGFr}
za9pYWqNLkBG|?m~HJPa|8txlcG#58WYVlZXEb(W;r)}n77thME7S8bQzNUtd)b@Ee
zzo(S+V;0Ll+yH6MtqjWO&X9B|>0GuCFF$`W-mQNzt9FW0p$cK#Os$GT)lngnlJfrf
zY9Ov)p@&MJ815(h#M=HwfeX$oqjkw#&6&z6R-N}z$!CevmP^|G7=BM#L+zrSt(TC4
z6-J37-NS>%{mX7&YbYYTHN!5KtQtg?{>Cr1g~yPE*`$kVx;6hcmQ)|hdqFz%8rM{7
zXL3|9qXLsRep4peQ}3CMQLNEwzWOk>BW?f34Oy=gb`9iKo-63qzkIrhB{|0V7qIf9
zSR5qS8fvok`Vt0?#@Ka`KAvB(e~jhvOE^>3Yj=QMD_XYpXBJ}4Nv%;Er~un?8DoTd
zIo4{@b7?iStTvgnLk%WG1mCjw1zS<^x<AHR4~Ky2d%0N6Wau6Tu@*ceI(lSZEk)jC
z+&dpE6MUr!<Q^9jr=b#iN#2^y@d+OK&accpt*w(2e6d40==oNA?I{U4q=gs5Z0C4)
zO^sbB{wU5_AbCP461cZ>hnzEV=jS{fA$$EL<@2IB*D%+9*X*#3!3LERW(&R-u_`|L
zmgf+!u;wRAA%-q@pzLs}s_`Q&LRr+|wfJxjB6t<Xe0FuBc7!Y!lPkSR|C$^duOx`0
zb#yYb@NtfTj1Phxo1e39T4S%e>!22vHPKe{Nw)p8rP(HX7_J#Nq1i**V!EC>ws9mt
z<jgHU)kUzywH$-AlJJ^fn?s7)P7m5|T~w1oT=`m`Gtar!31>-ucs#-0h0=7tHU_CY
za!hr84@+O$N#zO0_fUy8ODhH}$B{cYno^?gA~QC}Atp+_rYJLo436+f&v^Q$_PBq~
zBjRvk>lTe(T0;3q#z2C-I*tzkz*NcF(*m3NJVaZ*J+0A%vcs>eq)_nq$^)3L@*Zc-
z#&@coWK&##&Cv{=dcSsi{j(z9H=OQS6#E#B)||0@+vS!HE;w$Q;!3Yms~WVHh<2(|
zfU-+dp$ardmQ5dRy+l`tM&iQ9?%r{D*N}hb%akBR6!uw;twk85I74<G>cisLVVk}4
zkuW>pwPM!kNze5jGLi~4EH1-$iZ#rAuoj#`;j@g$Q@iXy)?1L0FQSEL;Whj?TFPB!
zVcdIFdUos?-e_#%9u+c{VfaN>6qf=isn%ODn8}=#|5-8kNn@!XFMLZB^AV;TVwXVH
zO8{TBQz$~#xdU940cHsMoJrB~RL3-)ebcucoGF#|5**DK$%;y;nl~1Sv;?=()pd(k
zXed|a4j%cyexx=3o7qupue#`w&CxW`!)Mng;c{~dM;wvkri2W%LC8iO`Fbv|7Ni)w
z@?2<+pF{7jony&cl?Zz=vMIjzmG4YqpU2Fi=iZ~_19F^vR%^^8<vO|jK$rLl*J_&?
zWQ;_Jh=;JNpOe{;xSUM(_j92RY+VG^=c<ajy!PXd)8Mdl8G-3~WhiYp=K`6f7p@eY
zi=Fn<DOETVcku!KZ?yYDJh66H`$SJNCe|>8N@bQYXj8bQH|j;&#O9OWCvyoP*IA$|
z9MxqzPw;E2FEZ)V=&-T(sBK3)TJ9&s?KwEw4~oEaifUrGWZgLL0H4z60*tHY;t9G_
zBhTB4q@SiD&8H9DyzSeMG0_7fhkt+9?;_s!UUJZg9{IGjFETeQGkj;k+!>xK!B_NC
zY*fSi=9}anK0*KCjPm^DFjj%>F4k3HpADSHT1i=VB|&LZ4!x_POF$B20Th!Evimsc
z$tG19s<~4P{Zcy{+DJfyCeJ5c_N*8UQG(dt+uRLp{rrm~-m)cWhj^#;*Je;Re8`Wy
z>W%oM&f>qyV9x(y2(tb7zjeAcwWVU{|EJUCl`m>5R1nat@9p0vH`^jCFC4FebpQ2j
zSlUFfv{z*od$7~YC`0k+h^fk0vQ5qXxqtZZcz2oa=+uwzVD<K93-%8is~^j5F?9eE
zw8-N7`Q!6l?jJVRH6L5L4A4d3`}AP9r4R5(2My8thmC!&`1#k`B3jL}s@XHDSo2Kw
zTIibCF>?vp(zFoeMC@Lu=lkPv23s>IiEQQNNB6CLjz06Zr)54pV<cjNFlA)6-|IeQ
zG&zo;ZmJM(08t5Xk9<$pM}&%Pr^W8a@yIng{%e{4aX@dv5+0j>NhQm$!V*;8z5*`a
zhsW<G<vD84@a+?^SzMZH2v{@<i~UL%0yLE3X%eBx3AV^^<D}$Qnu6P96+%dxo(U1b
z&un#a-(+*B8j9U}V)FasP}H@%7Ma+Ncd>r8j^^!aO~1~>tT;^?jeRS)f;4MF((0&)
zI+63*o%$H%8Cw<oe<VpAzZ#b%<t4RZ3KlrEZwk-QIC(uxU1-J>?Cuflp8=CiHacmt
zt~jfGdbG6uAvh@~xsv6akZ~ruT;!tD$(8wj3KnS@v&u{`(9zE9Ijc=!xrr%O14k~L
zc3OCOirfC$+Mrro1ux+=a>*30eXo8}jTqCL<bRR=>}X;hY>>Q;Mz3UUaiXFJJx82Y
z#nL<JApzLs<c&W1Gn;5bFrv<#5<m@I5W+I!lPq1GrAuf%&;)HEc?6-A_Fg?X#dk;#
z|8>aw57LPOP#~KO7_ZH>$tNVG##Ec7t%P60W|)#YK*VtkNeCb(B@ri?SRx<Mdm|I8
zLD{D5`VPI&R%LZDWl7GPqiWNfi)rnwkVbVH5$_(Lw5PP9mAdtObSn5{6%)N5Qpc=7
z+Vy-(dyF218G=-fqR))chw_=l>$D)3_|<>G`Jpbqs6GD;?&#dm=cOr(3x>eAl%Qp<
z9D)QRj=2>_Dc$(<>?Nuw=?dr@HJkBlk&Nq^0xY(<ZyrBH>v?$UFIp1DLGugWZyFwy
z;M*7<;5{cHv8CkyPM|!st~6mQyIO;Bl0V7iIwY*1`>Hg}LxkVhUjzR1%v1A}O^*sH
z>%{8o6Tx;W?IlVmTnf~QclBOig+cIBjdwP8#~%@3ZYnX9hW&Lmxc9f;HM-ZjXVR^n
zPiR?%<{^%21OZ^4WPv|PbR}>?hKq#x@KK0Qvudy+c(Ud2w}_AA1mU+g3PA)W1#VVa
z?4;Bhw=*-s+ZSAa{>=&3Vb)pakyo>bedxsyQzbS?1JgBjy0l&mZdRZ|5I`hK*64UB
zQ$~ctbfjhRyh)S*)=t=jAWLGMq0&)d`ca%6*H7eku90bGw0PX49`Aw9uds&218qXr
z{nfV2CEP_uhPd)emgR_DLvMW8&l&`~gZyZVNes%!#Fm};M4AVU2LX^YB@r)^1*DNl
zWog)SXQdH91I{7iWa;d$AcxSj0cb~hH5Nraq87AA38v5;amb$!NpY+rRc2zjI$+Li
z9qMW}Pb)HCOLj0&Yn|lXm!_-F>cP9V0>6fk;#tdD4rDy0?kl75VgW4y0%fnG5+l<h
z$01l}YH0;DvD}D0$U}b3Zq<?zlHq0*Ykn+#(cbZ{>}v<d9UPA<Htt2?lm$;#q>Hv=
z7BW^*!X9^;@|fsLJJNGmu@~H}6!xofObhNs5@nF>!4LQVU~*$=={PB`UcwU6N;FoZ
zF15}0IpQLuA8E~I$_#pXYi)psebA{!f{N_`;>5J>vV(<}_APKvJJ|%@*;s;Fx6Fk=
z>5$8<I(3-0vQVtDMN=xLc;D~pam|c>S5PZVXe3E#TEN}*4J99DRNd_JsCeFqv+3N%
zykq7bc_nNK1DBSy=Av+dCZpTb=-r?Wf8}_!QPwoN{tfBDzrLWLp|`bO!ak*DuF)#$
z5>Xpc0$|`W8RzPBU7lIW=VHIAu^Z)1?p!NnHPMg@&ub<SDJHR;iebG@i2SSltcneq
zzeCyi$EfarEa^7d<F?h#UvkPO7fyK+{Oc`>hR)Z8cepN-U}v2an>uXljt%v{?RAc6
ze7K$kMWG~O^|eg_B{n^)q?$>|U&9PAw_ai8TFq8+fOFeIA2q<SyU-#p{35lt$e)7z
z_9Uh{>!jwBkOi+M57In-%{n-qCFS5XDZ35)T+|c74929imgK}^^YzlcipV8iY3_3F
zrwW-9X>ya$s)9{VBo}&z)d{A~=^>r^SAfo%q`Hn_n>ItwFRgAVW9&T4<6hnrm?#nZ
z=mvI0KH!Y0s6^Fetpxoee+`8VT@L}IW_6Szg7g$kQRt5iKQd(Y6YK?$CC0opIobCr
zeV;|UkkwNWK;9zz5yv$iUVl*{(+kT6eX>*pap2<=-owO;=AU0c%`i(t!Bbx5#z!A@
z&cu?U{fgWU%T>9G^l*kDTwRqYNKhiqUk$k673Cp|%jLZuYi_;RhPO{xy?s?)CcF#~
zS2DCJhumiCGd@#dtv4f&tR6RqpH46*FLD5|?bRe!m_INf6}h5#>TU@I!U>0+FJ|v|
zu8tGSMxxS$u08dutP!-n1h%FQp6wKKxMXV5wHZ#EnFSb_30qTfW5?@(<t2$|%+ALj
zThVP(%cCE2Ea(u!e&1gOhRgPq%Idz~^U|Qox6sgfYOLWBE-NwxNlII-1vOnf9NtYG
ze&Z=WZah~gNgya6e5USxw?mld?1Sbm=9wLBw2kvPwHc3;>h`>_l~mumeC7o}nicAA
zik$&>)$hg%j^cMf=#Mdeb{+Wae9^#=*qdX-T&XJ`HGyPKBq^Vt0Yuv9%zhod283cn
zY2>y)V#s^a1WfsXyubB=EmJ+u^3n_oz%ejH>!JxBP@asfb&SY;A0gisKRXmoJVD6V
zA34@Ay0;c2O7PlB`H&}=`!@3-@uB$+nobasCCCe0@h;*^$WEb8OO7CxysGtF3kb2x
z+ZEsX4-7k1e@=+lLVb1R_!bnjfcxeGksvyzkI<tt(06H0ccU|i6bL=}E?a*5jHV#I
zTg&82v$;Es907^?_5-K?=5y{L_T)*+Xv)|XY@n|qT3z8Kg#CsuXlM!iZw@2pf2qr}
zvi#qW{9RhyG2|_1)5q#ACx{^vWD(d5FgTX#$R24?V+ax<m!p06-^hBHHB_r9?DVq#
z-fSFW?cF>Yn@(@u@xb4DOxMi%_IT9PouDPf&plMR(D7e?3Wcqyph+r2PCs)_TM`n-
z=UwyQix$$EB)+}dGwv&({@-Or3ZT<9NCC^U1Xc4|wm>|&hJ~anp=VB_l1E{!F<yvv
zhHIy<&(9uaRyKOE+7IUKVg2oG{jX<bO?53Ce5weN-u{OXLFNyT|A4y+`K2O&J-_9D
zcYUy4X<15X#ed;&f4)DnM}DLyVN=j5Xj#ja@R<hx@T=#5nTm<OyU9olNet#}m-L_j
z(4GW+c-_~w7yQ#fELRqiQKP)wMTCfDQ?2a7GZbmz_8|)~o|$D6+WB;`hqS-*Q#WD4
zG`SSvEKT}Aea>HU%)Jd6nb=X#n0<VMpb^L{I(9&pJy~cB`dH$8o0Lt>&X{Xlc)>ge
z4T-wU%wilJVc=ZAn!G_U;vHnx1GD{M#mYBFYurIfO2+HOwmVE6QU|3VBtcQZ>26lq
zy=X;Fa)-0rNm3r@!d%wy#jSr?+z??ec@@063Q{gwDh#Y?*lRKimlx(OQa2PQqt*l=
zDvJE6VKQUWUoFz$4HT~*;xsP3Rii1umiLxfx^k6E8AqC$aVcyZbIvn(-L=__q+Mc{
zu=Gb(-W){~hAVjxdb}L|K|laRHZM@nG-n}^F8bjQ=pd4CEubw&&kZfDPK0s5SPb=!
z+`+d?$Xc`j<4lc7vzx#rit$!Xn6x{9>b4!xmtVIS!OXO(!zluJOGhYN&UBl3<=9nX
zpcIDCinCM+^Y<8c(RD0!X23(hgkNKE$ropy8?qS85z=Qh&?qVFV4cat%zUa;GA}oY
zc>9&yb!UHcm|p72q$q9#$My4&={Cz<qVnFR;kqhvcyqBD-CAz9?@dX+!snhp)G~~>
z@2UN~SMp@Xf=lrreu{*bDaD1rzkqMfvx3hsnYEZv|B$(T4slq+OPet+rx-<X*>qJ%
zJX-=oNe>0h8Whj{-@t%VF-B96aVjln7{d+T9NPHb<f#uCI>m_Q#irE;52@6qK4Su9
zZBctM2!JTKCnJHwcvLI6;GRI#N$31zhq!j*g4L?-{I5nC3&r&8H+FKTvvCW#w);;0
z$z6Gr-+p-FB|%UgMtI_dsnK?9$@)$oM3pnBq|>^?k>1K|%9IFDMAB2pj!tF<8NQ+F
z{w5%TEca5l+R(XRO1{d&OMS2s$B8AHZKA7%$HcM+uP6YPwv<sog(c}y36+y=l4AG#
z)l*L~1fgf`JCfmv<jG{~y0~ZKSc1fEk@m%W@oH81BFqi^+dM(#U4JkoKb_zL&Kond
zz)7&$neZFx4=<gCMO;7zhU{{sQiBXZ0MpYp@!>ceZ|I+S^HTcsO{4Cd2G|9C!=p+9
z7xHDwjP+juxqF%8QTfVJgBtq&ORAcyg7y{F;A=3~v1jxOae2SEQz~zUY}moN@XchJ
z^3)F>0(q2>6ad!8Ye^D|W$ug0#3ede(WB$%ec@qW-SYHuz=P>xx(QYWuKAMUJr5aI
zsj6UuYO=dG>+*^_9t)czl|@W|@$y5;X7kM{zTDf=7*91*l?^oB3EjAn;W&ocQe@_v
z*<#F9t%bfACT{Z?mrfbgF&z(YkF!c{ZY6`;w7oD54B?32Csj`&K&`6gmi^=%AozQg
za=&pl5w)H2<~?Pz_23jInu`wIU>h?)Spnv`-H6j$Dfv3I(%9I{!Od`@ic6Ge(`C-;
zttBn=gmvif2v_&WCOT;uO2__j)@f%0SFd8)*8)~-RWZ~%&`6Ni?VGz6cvD6M=mmJz
z5zya7UKr7>42y+F?{Bo!lQ)x6PPI%qUXMc$_0?B9vTj=pV7;+%X1rG!R1oCVBUEUN
zkBm1t<fujU0d5Q4p~Mei(6r*Qj8-i(jGei8GFT_FkKDu&%EO9j8*O%{057f;!f~vT
zy|1a8#E9?1vb!zF4!oc&7=oz-UU8Ycof49VOa<MXW)BVI%GP;7O@s^M{b+jwkVw_y
z;q2)^v&@<hHOvvP#?^le1qTAjY^7|qtbF?nA~|p}>#IeO_Y5YqB$E)^voLGvf#F4~
zuAB3t6bU70Mpj+w)qOlVdq8&%K#g%#XU7pJHnY(zX0>P%W;v#^TxShK+x@gG6PAkR
z9I^ZUT2o3eBjQ739ew{UH#Lp5K{Gh}HgLw>F<Fi?^e19=T<sXs{yHb+&SO}IYh2ON
zw^?{H@BXjUOJu2@W#^Rgk`c`uzLAb`oYu<Jxz&QCwQ9oa)*Bh0r!!X-r(lA5>k_?D
zg0h+|ARRI4YN~xG3B$$BX1(4Se=gE*S)!8>e#Q~##@PPHb-S_a)5At);S*=QW+GZ{
zQI)?}wJmSV5jsi1HnpSrAL#ChYz$-of2doJhYw{d!w`;$-fR7-sHXel(d<Jn=cg3q
z?6j&p5F&;PxLLOczD+*+N<6}$aiLB7PP<L+Pff$A8$|p~WVmt*d)@Ajgtfs*=%(hx
ze+f;?jtM((v52MNmxr*O^F)`UtrcGPP*!sXUH(D$9j$Vs>`YUsHR|y3CeWQ)Q**ig
zHQLZrA*C0XQ?;3W7Gu@2eOb1>U&w3&1s4%hOthlJ3o6g-)@kdd|4L-8VOSj_^OHuD
z-F?>2LLFK}s%HUJ*GrcSI|#g`R=W;4_%_nel(l;;j_?<~f?qLKdwK?*`;e|k0gPY3
zyty}#18~)^`S|fy($`Ii)JEemcwg<lgrMAb)Msd{IVPzwX~IW$Kzk(6&+WoiaJgO>
zmhJcYLl9W*cCEEY*+a4IE7aA%cF)6L^B?7}2h*Jk%WF{MRrr$I&oz~)gHi8ni$b@K
z1@ALR!I#K`Zuf>?_g}<<?JxQNP3Lg`7YhmS4{`N>X>~9D1C-)(qjjDCOH#v+&^G)<
z41YO|^=@qZOHz>#3of}=qV$kH-_eCxC8^j{q}vXl+Ch>i!c7<=55-sxA57zcp#%5l
zyPK=M2V%A3=ckYVvebWLfk}hH=G6Z)7WmC4T}IfQeG93K7swow1Kzz}0$l($t%#-n
zQlYGXOoUO@$1Q%`G<*nvGCxJ{ruPw?=c<mGms%s;o0yYV`1bvL7RE9w`Wr1Uzy5N0
z_;~Z4S6>}f%OsBy=rAXr(@l5+t*FLDOpTey;`O$la(chJ-)dcCI<@fid25O%?ngHc
z50^y!ueDsvcNx@V=)?r~Q_%D<ubU!w7E5)XqHI_INF9}P0ROwbMEK&olHp{H1OaZ)
zMJWvn8CLQ1hDiS*ZP@1!JZA9hy2QE=Mp^->6J$TlE{VxwXrA`ojcyqBCQq5WN->~C
zW$s_bscsVYT$m4H>nfHQJsLhwaTWiqnFc#09y{CNoR}gD6lWyBDV8;hlRe0k!-y?c
zV=K=Xw##CnisqfS!|KA1x7_9Ju4Ei`pT=MiN0X{Cx_geDklAhOD$%(otfF<wtoJ$K
zp#VoR<@Qae8+Yx1*TxTazp_Ur#QFm>|9zP^fH%o`Q;@ZmpGZ2G@p;KhW7r&KCTwBG
zA(b}rVRyMMC5<vyAvQQ3s(wP%(m-Ea9|sl44jLNe+PB*gp^R;}<o8<E4n$LQP;+Ag
zZ(!<>gTv;)#}4WfPMEM4!2n?tvEi+TNW<`n7K1PJBw`bL7h`Lr!hHz)V05gbMOInH
z=`#hFq}+)eKoe3vt0Lf6Wj;^SlCZ@~a|8ib8DYz#KMd=7F_9zA(k1;(H<F9yG@*1Z
z-f7wk6ARyYn0fOAw}H3iE<}$WSNet;Du0$dBL$g6BW<PxgYH5$Es9OrLxV)4f7^4w
z*buknPH3=lzdwD(-({w!sP6GofU|mbM80%Z8H2D^Q%dfM65R1abNs%&Q4bd1x|W@J
zy<8^IVq>#9U<U%KCSWBTjAoDptJfuyBP&Wr1RC{IBs4DkNlq$t?kN{c?h&Qcky^RL
zZ7S5}w!#I01^GS177%~7{$q!|Om8WsKd*8&N0fg;10j26aoZ1U3%+DQ=wORlG1`*}
zCPO(&;C&k8XHD3?*;u~n>fg<*hOIiSSM8u#e^SQC^zvg^MhKmRUQ#vj^ECT|gDLZp
zIfN#J0dcMBMfhf+(~t^JlV;LA`N)buJI=>;yn@I`u!?I#^m^A==l3HEq#v}&=4!2p
zd(Y0;^CCl80RJk}it_KME2t6%zy#TvGmiC^^*}78F}A3}#$##`SUAls?cqA#fS_AB
ze1)~s8%<BGR}uqhi%An$w1=<wmg)Ckf$w}Fe)BK_>p4Oxy?Vb#{us>ma@uFKuU#-I
z=Pz#8`0jm2Ccs5x{DZbI*<X&~K%hSebRCmX)HXHW6|xGCGdnnXVAfZ<1r9c;sPi<s
zEM)KyU;5YXroFzv_c?GzdgqUzC8P-EgLm-L8{_Yxeq?&E>yny7ZxjO>5jxa5vOlm>
zc=SIrd*RP}OHTcR9C*m;INp-Pqi%%P?CiiKK^rY0+DVT{FGmYxudM|~bpU8m#p6)E
z1-j!}Z{g>jn7&2o-&rMl<==9lGr(QrMRWES2~wa;OR+pfIE>gM-683JrUBbVV59jQ
zY#pe~a0mQxOS?d%BmOfENU2JB7$>hq)L-QRO|<vy==1YUc(yWTo>qv)KNwL>KY&3Q
zGs&!}c_ybj$yDEAF7%4u{ytY01J2!R7PSvjbd6rvkZ-<;M6LMVMqXls+L*XQT7hv7
zpmM~r%DYlaTYuwtrXiazz^*BHrHV`x|7;fOHEc3p7R<w{v01CdVZ<xrA6$I3W1Z7p
zOVnR}TlZ}y9mJ61W)8L$MsCVz=3DZ_5z*$q@sqZIRl;c~QDL0#9IkAnDWFYff+R}^
zn-xY$n($5XhvsqQkPBCFs2)C{!1_ENPK@wsWO~X7eFchREdI2*>l!zfFD(&3_vswK
zHPcglpbr9A-a-hdZNgYLI%}uTi@tx7ngCYt_c%`l4f(ZZ1sGC7$e~2v#qTyx8$F1i
zi_wQVe~j8rl~LvBl{;~U1L;R$jH?{bu-;zmDt{bTMY{nMmjKsGz|xh2&Y(>9*<x5w
zXzcv0mftRNyE}nGk*rR6M@6$*L3VQvlp-w-8RQQ&Kp(X}C1k?b_CzZjH&Q)!Dh9fo
z6+Ws}-5P<O((CFxCkBh4l;&$;YP#opa!9m_Oh{5qCOU;Sd-HY{JvoP^OPc!)gKFCQ
zV)6Jk1CrsZ{zdOdm@MhiV7@HpIu%E)dqfB7WF16H^_}<0e<FdAz>dhb>GZZ$HE~R>
z4JHD&hviu7(U9u&HBJ2S%4pAtm{Z)akho3KMLs2?s?fG%I4KdN@kihjYtQ=apF>%;
zT*-YcZnY)%6vk&sYE<Y&kw3qyiT$AsKl%&d#yd=)P#Rgrnbe3f1`4&iD-dT*d6T9H
zz<n*cLTn6Hb(|!OJBrlp(KyM?djn(K3WZ^i8?R8ZmCSaREGgXXAif}<PUV04iJRp9
zKa9OobY@)_Zd<W!+qUhbV%xTD+o;&6m~U*SV%xTzoa}vdTHD{nf4eSMYjcjV#u`uW
zekGPor)nNydJJw8t<YW<JDIuM+0B10>yu)DmZo=!CBp3B_U}Mn&3GJsv9VNmg1(Mx
zG^r7MSc{ycbHxW>`8&4rWZkAaaO_4XQtx=bPkafVko;xKv1uxaPCfUUf6I9?gwv%w
zMSItH<Q{D7JH;Pa@+(}sn!Nl?>)TUPqR>#D&TO^Dm1u{W_h6b`LKsoLAJK@65jf#h
z3}LwG{+v(S&K`HU=28{k3>1c#<sQm5>zbc`tHGTg^-P2IG$a{R#}<U=nYBINdDucm
zM&ji`hh^~(p?3B7NkhVt!afW*pB^a&HqO*nF`AZ)FXb}}TbwmHwQK%l{6g8gw=^^Z
z-&>(C?RY9-G#i}&1;4=WA``##HVIa}rXxKHn5ICp@KV<U#96V(KxaKS@dV$((|<K3
zN1VeEhCcs9>m7UMN}U^&Mn%V=$5#1$+z#}Ho7Sxa)UJ>UikB-JUbghEI&JwPyMc{*
zJjQ7amss2+myw1PW-lM_%$c}T!d}FS<i-UTX#-PzNp#X_a3Y@gK>31!XPmobYi{c=
zb12P5ZYDlE;Zc8o2?ld(H&aSrivI2XdJU%=d9a|Tm*=|1Ec0>Tf^(qP#$g2=J+V?t
zkY*$VU~}YRoZ5w;LRE$uGkwo^n@6ihG4nb0gA7Y;d_nsv<-pSoMop%=eK{3Y>=N4^
zf@>Rc2$p)f(oq?+LGK!K(^yB<84d8rNvcB^0EZq(ox!#zfIb>HLi>vm5g%J$7v$ob
zb;7~-KRQJw#{XCE&BVpb@;`No{|6wk&W_l9U8`VEH^Z2{3riqqV9;d^aZ=s46_p!B
zUWS|b?SGO$Y?_#4GoiGG`%yULIemkZ)c!tqdbK})dwKe~s&JfQ?y%D3dnNcaa|f>l
zPUj9Us0p(9lRt|`&?-t@R@5tQX1a5_+nk^v*uuKG1Fh>k&<_~kkw82iWdVP08-8+*
z;SFWDgSvag9WnKrQImic(qr$2_wW9EDsHU1p1v?A6DwK?o8ndy_D)bKHcCtpMK>(f
zEG}v6^pRHu!+{!!kDsiaU!NC{r}KnV3v}Pq*U#G{0IjNm!PBfvsDFQQe||lEPP0MR
zrJ-|v<5=1RFzB+}QN0se%qnP4=vZ5zMc<_faqa$LD^{N!H?P+7l0~A<(&a3+#~3yh
z#YE6qp*HRhkvVCr^N1Rxx$v@!ajd#c2-kcCp^8R%MH!OR{E*M+BF#H^yDd*gM>tC#
zG|fMXVYcCs(elKNkR><gP+n;2Bm+?)F$Mvx?c{(nTu6;OG)(zm1j-7ZbVX|liKzpd
z@7sy9GNF}KL<uI{BLfNr3&&mZapZNx+p2D5inSD`8)vtEy5G%NWb*VZde~8&(U6MQ
zTwDHUkE^m91zsgAbKUYWOERW5!rIPGJ`*T>somXs&&;i6Qtz7?fBx!K6cilUT}Ixe
zt^iwZ@fUKEhInXdtbe9q8xb;zA@Cn3V$yznEl74SVJPdNLBzhD9zUv?k_q--zyQ%*
zOyzu<Lb>M_bm-eDUBS%FVwOk&kt6~Ns`=lYfsj2<C4O~9up-?3xF{Im#G5Af_8CIz
zCX@ktN#L|5=v+6sps=$Xsot!l{M}W-W_GB|G7eH9YoQfX!;*i^YZ<Q|ql#N=9fJe5
z<;{>c(mo!Qtak<!Q@_ScA&1>X=oiF+XJskr3yR?y%vVPDtY4i2V!@c(hk0z;C$#re
zVtv|yNmyXbWS~jGufr0_u8hrQ?j_j;`W4vuLjgecePYv|tY6-=&Lv?-_KHz^JluMu
z)UyqREUk;^j0Hh#x_yO#(KIwc&dsE2ZB-bRh%IWviUS5fh8Ub%lFsh)gvhzWvt5(J
zO4hTxx#BfcrB6+wc#kQ{TRl1tzd4txYIE9DEjj7g*rCh;bGM_<Sh5rqe`!aJ)mdj%
zmdZq=3s>u;<7i}5Uja?L-H1B2;3(0wGjJ?bg(pH@0`p40c?jz|V<b6fP1|ysoqLl^
z;tkiB3~-{JA*xsn%SxD&`U%@NvX@5K0*PI!4Q{vzmIue9!m16OL<Ffg8({&<<&kZ^
z|LP5MV!<P%M{H^5I@Em{f-1Au=iAtZ2u9nQ4FK8enB*Xv)l;TpE3?qr*%R9%ZQ!Y>
zJ$yOZ$}w^@b!GMpX<1=Pi4Y0z0gsG!e5%?epU6s9olW{+thxay_}y^d>~1~BQapZ2
zdE$(=bd)(3U~^Zm1ynKbYhyCWvDH9~!w-EUuh`YR`BnHrTvZdx`2Z7%SF<wqt`mXg
zV#yOc*v3gWt}gW{CtWl1gU++}3WDTpl5P8Y)5Jxt!)D#l>uV!C>r@IrozBKlYu7j?
zF`C14O+i7chV093OTn%~k-f#?t3NhGu(;()qoW1o70BeglCRLQObyN@)~W(*V77bD
zFp2u_U^!{bRy8D-5|q3P54tO-gU_t6IXXQ20FvRlInLdtUlN0|z?T#??Y)VQA49K&
zmxE?rN2B&&syHb}k=ds=kJ2E?>$yv33@jZu=NM`rl+1r#%M+F>1SGq=%-0B*(~lD6
zA%1Eh&+>7!2`bN340s8Mtb@!HMe;q>*D%t&p)!?wmEsL2I-C}aAErTH%tfdP5^JgJ
z!e=`s-xv>RqzBn&sp8?zmyQ4*K@?}CNp_g}^H77iFg>+LA~W9cMA>gm=H}nZ1|I6l
z0`bx$aHaIhzbm3#dTulWo$6W*`dIxsn<NsA`J8R`on#6OoOlefv~gle;?jRbRck88
zZRJ66tMze>Rhrp4rTvBceqGkzQ~4u&ubllm&}&@YbX=9*_;_oqBy-Mr)f9va-SK=5
z7<5XGiPffmX2q!GgAU{L)zf>Q+k|4Gb{ol_6;<6mzO%wxGzv!Fqt&}<xno4vTR7!~
z`5_Mw5>Xs8LeNbn${01nJA-_+eODkZmWFTldSvez*%6y=C}h1wE?UJj!SXy^qC%%<
z1w!+Oi<^t?S7DWpo0C2{M#K<|uV6lG`xut7-1Zl^oP8o@+rVoEjk$SvcM!v<ZY$wm
z#&-6M35DO=9=41%GYHOlt#z0hiW4udxYuYw(ZnGjSqRHSu)W6aAl0P#tPq!rf=XK(
z?N}>5rP=|@@d6;!;NPv^Ms2xbhQQ~4268dZ2I(Esi^ySXqkWaRAZjUg_7X37bLAOV
z^}w>_m>mf7c6PiDF_q2Yq$bN-N)*%he>*aw!6duydRusI3twe7{Q3#DZUc+hyxZ7z
zcnp}2rF+hoBI_Mz^)evf8a-iPQ(rkeTA7_6w5b?$=ppZe@@00Aj)h~UlD;+^6+2G$
zq?5+Z@?OYnDoYnDeZq{+a%lWlLH{o|4I>lh|5VTywI+W+=09;l<rgqD8l^(N0HJ_C
z7iNK5?HSs$%Cnah;aBej(x0o$wiI$p^I`~bC}O3OQ6G|tO9T4B_}#o+-F&{Cm^>b6
zki78r+u%C_to}hqLK49c3RAr|PtP~&jY}slUWXG_wy^rSJrLdVr@?`}ECDZj{y=gB
zouLn~(@Yw<Y&E>b9_YD0k$!)v{STi7ISS}RhFplRhxc^``F7??RhLui_A&ju?c6U<
zY)y%x(V;#h>fYV>p1iqMFr}#xAk7-dfoFkmd-)#e!Y)1FI=v=mQI;+Z@+-?9vwHFu
zwO9Q=hWeN^S3~2jDO28U&uk4fqgA>zo?Fg85BiD9ntDbbFJH@q;@y5&VuMXIOH%YZ
z%IzUb8Yo()q97FtIkAk3@#*EuCSnWF7DSbryInOcC2XvEfgDN5zVr*(-TELVXrfYx
zk1;up4I=1XOgvac-3Jup>yA;Bb(f%~E97UteffPIuu&##s-BwdC{4j`1ZG0UP(u~0
z(OXX4Cf$(GuyMIclAEVpRds*I?v;Q2eIjFv>YV<ui~Y^wRR1QI@1p{hBGu%ag=$nn
z=6QSQQ!-*+f3a4KC6jh$7=oenjI=LW)!*X`|0|pfAcba7)!~@3cd?Zx^VYx)%$;ju
z4{fk}`n^3*e8Dwky?$0wGajxf?l8^r+RsYk&+*tVOm@$uU0pEnq5{xck`W6dw%)Q}
zZ?#b&^sH`}ifzG$tCM)2wP<{@*=4FlK95Srk4Lt9KW6hhPfjnrl?IEhnNTirt3-V&
z844t~_sX5JeZZGpKy$Nbn=ADz*@EVY`%Mv9^DIA^BhPe-yVNX$lv>{c5#Q;D$)@2d
zqdFWm7}G#xAk5euM5LTx%k=g*(L%x!kO(m*qBpw?1Mra~d}vQDilf#@p;6KG>0~6=
z3#7SI)68fHDV{v4Xcv!?!8pB&Ie=Bd<?n<&?Cm7pt{uL^KD~oj%xA`V)!P5`g0%q(
zgiOf;4*Fd19BGq`s-r;%<}&@Af{IhMZnm7>VN47<%9WMhiG<2EsWhz3wNCr86S#~D
zyo$}Qo@-b6P|BZpIT=p!NJV*d7fXy&J?3K>$Y)7<TzmZ9h&qurZUeLRYiS-8N0n7>
zYWQ;0gF}Fy8o`f2ps)M{V`HBlwkL}tnHI!?!rtda!H-CiUvHsE94hNDVs<5`9={TM
z_1ABh<VA~w;GO%huI%`GAbw*!7%G+^GQL!({jRM9G$t<4uY3^)JzF{`Ilx72>E->H
zL^}usB90p~O3ocF4;ah!0Sx)kXiGFpa$NCn#Mn23dY#X(!2s_J2I<h|u0igH=KzH`
zhlac~L?d2RI>p#i@qRG15`iAULyikPND3%NOre-N)4Swx#bTiE2cc6~`F>E|^DTjk
zoS0mN``wXH0ep0Ikk{SJ<!?<7?MK|RDWS;$aa|19%LQGOiqPr=#hs05YU)K3F$bc#
z5dvz;)WTSu{8MK)B}t2>+7^{YRH0O@nm3S(uzW;}^`QFPdk1c;<IIxnAqE0*&s(rX
z6|N-2x%_{3ToUg!n68)?4G6{&68%3mWn-?<cIO+YMc$xZa_}%=PWV8uJ{(9|vmOxO
z>-{pKx`Sh1Bo7LEw7^5P&tpiNAa(Q`3X-~=JrB?{7~J6AptC?o+E+gCK(Iv`NSjXO
zm}MAHJqFJDvCQ598-+eoOE&a2_3t*dXKOG<9*M;k{c5OPwKE^hqAXuNkTgEq$oZPe
z`YfnQZLlulKJuPaZgC|?>Gb;HRM;-^f?g0BJ2ncB5=wQSwwOob7z$?H#Vu~liCl%#
zt#%XcNiUriOhvIej`5^Dzv@GcU^XW&?0pbzwfIN62RDd*cwB9?PP~MWuj`=twM!J(
zS{Dkkn@9?HN>)ky8g=Z!i-p6MCS;6r`}s<bDhzs#gWdaYKRE33RD>@gsE}J7F@Scg
z%wT>{ePyi7HsZO;8xy0e+%W#6iR+Kf){bRF-`L;c;;hN%tmItu(C+uPnAU1C9j?gK
zBwZPZaHcFZ4j&o;)@@fYv(qYbx)eh7ImdN?Ew|nM+FeT5+(ZX8-3XBQ<<Qoq9)PB7
zu!hCGE9c))d8ht~n|T>jAe_?vl-xX>wJcuo3Aw%IDi$a$x|aq+yfS07P)!bRqLn)H
zaYw=ZN~RMzug&XJD@yHLv`kG$<K;-!?h8ZTZ|g~t-cll!qAV;C?>#+`g9s@Lbm~_f
zJ+jPvwk{pk#e#VmVKsn09HmKYdofOFbKE8<@uyn6LQO}JW(M;YEEU`Sbv(iAs$Z6k
zb%^;(;BiGw?jHgDW*H}OuBqRl#OMqaD(7(d8bbS}u6Wn4>?ZN8DFDd)uE9Cii*p-}
z%6Ih%`D>3}=%Du}w;-HJlNNqx)>;zh0WE$<kjYApD}hG4%<Nd|Lrvju?@q?c!%_;f
z<jcuagv?1DOS{%jYq8|pqqrif>|&1N;`&s*3s)NE>@+Q^wHNOG94UGYg`IyS*RyyF
z-r*n2ONpw~LbqEgZxVNNT8&7!8|AAJs}g}&vUuH}8SFH3o2d&<R@e_pE?V*yx+=5@
z**SV5C)|nQV3UeS2#{vJcL`lVC2C1Ss%@EEj=)U5Dg8nMbLym7(GG9hU0F>C0_<B=
zz9Zb=zi!&XuF8!^c|;a^%*r7C3joB#{QvrUnK)S(|EGT8{{Vo_Ibk=CX+A(y#Ud9#
zAc2c5l?~jKT{top4X$)Ihc>?TVaa3~Xa)9o*Fv)f(?p|qi4ji6ugobro;N3suU@O_
z8~c$o7Ja`?0th6(V3O>y&u9Y}vD>HhH}$=#Qg_co8B&z2yJ`L1KcvDb;Nm2H-44kE
z!ONo){o7^IdUS3+J+{Cvk^6x9@PvK#>`oUCL*v!wh2Pk{o)&Oprh4~za*|axR=io$
zR5pSd&HUN>sA6i6D}MJLT}zPPe&wncN4-)^dGT?pqmGos#I?B_XZeBA+oJ^VluUbv
zNI=%LB>M3}Oa#Jty?~th*247FvFFacA$B3#Qg=d`!4X_f)&izC{W}^=OTe(Uu^faO
zRli=FVGs~CAkKM8LLe^uzSlojK38ipKbHK&XT(5c2E4JbG_ZsYxQi4pzd9#UFJjZ5
zR*80)1mfI0W#esoP~UdB#!P5TF30n>1v-wVM!cEQiBWg7KX=)5NEl`M#8ZXC$WR}n
z<F0&Ad*!Q&WA^G%sJyV029PoV9ZqtTrfX=2uQw(ptgC}}>RDk$Aof!>b+UhuW_~S9
zK^8z%EK^-2dv)D)Kf+2MSi(OIQNs{18s@EVpSaeu;2}OkC9>mja1ZWCBp?{4z;jW8
z>x1Upx`qrtumU5A_w=`y=H>uJkMQzezGaiqzgSRZII5tFx4F4TSJHx#*@u)w#zv(<
zC8z3^c4%{U?eA+=DZ`%t{p-e^E77%@b~rXzy>d$aH<VCk0>`T6k-fqSwt37F>;Y$f
zfy4}U)MvAb#LTttGfQO3q%_;**^CvQPz<@?!is8L-Cut=VLA8dn7i@mp+6OWzy)GV
z!-FfyU7|aBwqqih#qU0xqxL*-H4@GhW^ZHJ2h2PMExbvMB40S2wH?)E@)z3M4dkf_
ziQAx-X+&E-A?yU3Qu$lv()%p<oO6B&>bP8&tEvz1M0MFCs?F9e0hu$;f2QF?OCpPH
zdH&jI4#uL$um<#Y{~OudbiT@Yxk4?B$tk7b4VyBwst2;n;(hJvM(sk*7$K#qK86ud
zTDa19%4}JQ)}+iUl5R+t0rvz3wXzKE7NLZfxBWiRVwtyKv?*FRs<c0!>2?kBkFvc2
zSQaE0qKN*x+mLrewqHkXER&-k*DSC@njPg~2-8a3k%()vj_sd{^GKb{^k_nbRf$p4
zDi2$7r^(KZ+{zZYbXQt0j}!Rr&RlNh%nS>gCW;AD)`7v?vM6vTe4H=Q)gqqM3n%v0
zr)0nU>Q2RL=a~rAt2XFTdfO~yT|Nrw17ug(6gHJ8?ccxceVcJXs%6m{s+C8RP}0o8
zII(J^G@U|CSy_M;aPwn!upKyc<nxV)Ek2y**Wy4q%X#(CIVSh+9TWG{-F&i$4t`hF
z7fTl&=T;?V;g-H8)IVf0j4&Q1?M}jqXxi%#*22#H0~^EjqjH|r(sDU14Fi)#b#v}9
zYMcQBKP#z|jt})y&Jtz;l6Whn4{gch@G9C?au;6Gp;2>IhwvRcnQd7#E2&W{C<!mk
zDTr%hmCTckAvpImm#yX!OHy23TwrVb-Aqzk&%C(&+1Y;fbf`XQfJ2aowRp+DxKxJP
zt~^4Ns_)q)sSstCq4qniNV!6T$!b=;vSq10UA_B{&g}8u?xeeNWI9~L%DQI%!P1(k
z0^DUmWR680$f{V!%d<mP0e)=j8y%om2yx@XSCxRgY%BfOC6$y#*s`#HLp3sqMp70f
z9H9r+WDKFE&;^dPFIoK!bH9f$5w=m%(Tr&M?VfgINzriU2^XRL9nLX$hl6?&3w<rY
z?W~@eA)E6~uv-h|HWGrp?LP6HXC_Pk4fQd8#fjz@?`uIVzo|IGEhg}?ti%7~kDDaI
z49{c>n?{bR*jq~suQX~LVy}1-?ug9w$^ikJ0A6uR!_922yi9R)8jYln<EsH=n@C=I
z6<%TzMBA3!BuNPIc(-}lKC7Zoksov0vQ==GVdivaS>?V%_xM@vLD8xgcp2;gp~#WM
zVBgjhCbYFzT8l^{UmgvDNcLN6$?^iE^<;UC`?ga&bUb(rDd^Zta*OeE-G3=~C;@nG
z0@;KJc5q_Vtbw28c)re}@|#G8y}R`MYeCy-7R+r!+>_x%Rl>h{IX7?GlPn2kWg+kH
z<|ihX0@C$Vr!W<(`%Ns0Qt@>w)TTthlB^n`gxk~}COf*d5xqco_8cm>N?5V0T2XyG
zMhe!kDl;BL*6BE^!rQCA%aSe3z6Gn$Pi(jTr0M17dF<oYF(<2mlXLf;c|R^Sr1yID
zwHxm|3Fd&z+{lBi2TrhQdenPR*zL=mfL}Zndshu=G0M{Yxdk;1^I@-dTrB)VPN5#@
zZgCjGe?#QO_eT&VW740i2CnlG#s<;*b;1rJVepqOVikD7_8_or>>DNKN%TLy(HVcr
zD<h%3krgyAFN2DwgDHc&k(IKGErT54&piGIy65EVLdf<1bAGU}{%-^MMJ-!<+%~wc
zTz#fq-!{KGGfBll4P4*@`pHLV0>nY88Ft2?<}RtHvy1j~*Mz6#=Yprra~<}MDNhhm
z>$tEa$w(hb=^{MMdA2pj<vLP|a-`~zZxvz=>UtvQ!Uc3&=r0MHg%q2u{#c~Ih>`-i
zheS8bSV}SbqOjI`O=1qxdL>Loifil=qUJRWKH699KvoNvC~PyMGE5cZGGfXJ{%}@Y
zCbBo`@;zS^FVb<j)G!wB?F=DRA+QuzS5!(trFH3Gp-u>VV`BT@9mVA7K&qllRw|~g
zGz|&4{3g~Un3tIIPArupaY$rI^<znBXk~JJp3^89Xt2<g<I$b{!3+`^5~f7}_+RNv
zW0EnSHWn|FCjeHF#Zw{+IfE;k6R>GbiBfX=co|~4k*{?=sUdu<Y}A?U8k-ci>t8{-
zTfQKCs^!Jrqq0*iDo}<aAB~mYuG@9{e;V8ef=o--hV%2oP43w%oh_1$b!lwqz}+hH
ze^0sUMg@#FGptE3HrNTKb{j~K+aYxu*r@VDl3u*;##cN&Mkci}k(TEb#T$7fE}F7G
zP=cDB-59&7>tZ?cl?NXtSxey!WbkZLUT7OK<rM`uWff0l659>c9hMfWctcVWMC}lW
zmF*vvYSg5#tA&zshKSXmZjkf~3W_PH;Rn{TlJD{DSf(v(8%@v}bM>X;C<Bm*W7M5&
zowv{aYRrHKbwu|!{vBBiL@j=UM&V&lyO>1~Xkz&GUbN4H4KI?K|Kh~fe;BFd=Ja)6
z`)d9y`Gi7nn3;bP`}Vnu$`A>v5KT0g2||V#FmQ+_Brl6!OU`PHMDXCn{Oa&+!2I%I
z#)Ml>e!6DxyZzqo$@*PW^S#kx50UK7VZQ70(EWLN^tm`I_AT-b$kNXX-=<%0oS4;I
z{|Q(CxxEN)OJ1cK8DaP&7N@`za7407_h3w~pWk}&v}eL@A$%<p^x(r%eBNelc@lv5
zx@Rl7hveOrjl>h|ZpTTpXpVNOWO705L8yBz-FFzr-a-m^TS+Aa8xQHod}pN~NXSUd
zDA`mT&Z`TCL3ZgieZL&{7_4xi3KkHn!P(O@$P4t9>^M0r1TfaYl`-4!+Vx?Pq*bj)
zC9?`npiPy2TDovy=jrshWvvxh<~@_w9T^VHkM#ItfA}z;%D%01t=-tLW*fwr<I~{9
z{DjGd2CSM+L4qK-5*8o=2uo@j`205b3F4>Z-;&iY@5`pw(m&&q=h4<Zx4hTG1szIv
z57tNyT{<y7&u1rJI{o?Js<$4jzl~<z9UtnkHJ%b|e4@ULA3!l=U29uTx=3Zq{2%_3
z|Edv>{bnJag8GA!`t71#G&OcBjUk@XxUb5ETO@}io-^-PlT`>Rl#**5$%s#j7nk%`
zp<$!qATz?<`uBWz#jqAIax$bCR2vFS%kn1Fx)oDii#&FW-`p3k7nfBY`?hgM{d;68
z8^&TG8>-K%@Q4mD$)CcQ(6H4g&GLHGdy*}`SLx?r0tkYBi;tGe99_U@5LfS!*mLL&
zu%jBHj_}@qki$V>9Y~rhs6;NmgoJ1wPp@~VA$(pW)qx$2+()b<j$9O#;FT>W)>Sk3
zy*@;z`i&RdSPyC^ihY3lnH$ixidrwvG@=%I$W$aVg^j>X8m2}X7Svg2z{X?2<*NWk
zSM>5{y!yV?g3bY?Knq2X`|SQ4avzV#MnmvqC00Z(@%0~djC@P2Jv3^Ekm{Ei=^iYd
zC;c9**zITXWmWB>BI5<6XQXb#GUO=RC~SpFrhwuwwBoJqb-)0Ip_p`xF0hld#LIwZ
z*-JAdVy~<)TXzIv6;T%dLzMqXEJRo1#FtAO78|z?bbc#7C9-YuS*@e&E%yjL>CPg0
zb_h1*IeIp*_TpsW4!ztYQG-q2m~=0|*6cWBPiz7`TSM&E1bd|ka9y(ibGff2?nIU&
z?KSF11MR-X8iP1iXlsl@SP7|@g*&#Mp6q>f$V@R{lXAZ7zItHFxWVqqf6eb(J&n*G
zTngSB&t)FLg+gEkO<?r>Rft%WU7#Y5f-yu16m2!I75FcFYtSY&-$Rxw*d}lMY|fOi
zv`%I~speLVXHREN!bo-TWw8)E1R9LH#0c57rWK44n|p(Tk_t%BSTMzfd5_R|U$|wa
zT+HAj`iJ!@dGs6)wI$6=tIdF>k%B?rF&9~C0{2#!@<(xr4;+N<{TQ2lsElGgl_amo
zlaOCQOYQj@x!1bfi7kypg{hyO))BwXk*VjO^g~m<wDiw>J?7uOLpO;WZY~C7peR=C
zM;n_Wv%@LEw=l;CZ0-73WjpFV*~Sq{><D<qaZG4Ehhdmtso@)nO}+()^}Ps5v>}Mi
z35H2|$iTb~)<}&RJf})Arv5~fS_(mkIr?aa<b^uI+@~<0&Kn6tN~cC4_VpF18%Wzf
z0x6&-$VhRaARrlZuoc8A)HZ<CmvKlwad{COW&+Y$>lUI@C_L?N3=SAi)U6@ziIk|n
zgH<LlRJgY>{LYY8?x<-}2FRUoW%<A8KdMbQSaFQth$r9?Pt`wQB3L(J9wGJY+nrqY
z06}j%98$+cyN@OtII!mCs-XMAGZ6D$D;(?O7qyO)@QA-Sbv$sUG=Oz3tTk54YMAU-
z^YYMOruUj71a9b`BygUN^GtoVz9LLEzbvo{KkpN;{rFt|MetHwF;C*fqbt{2rU)bJ
z5koUIA|_^Rg^?<&wiHXA4xavVtSuUfSP?Y+^C%%jl<vWqF0T|9=e+yQIrNBmQO3X2
z$5yvs)VHG$0T~b>`(U1&fmm+XM7e<oq6FBcqHkFeXwee<0r7wAq=PMAd#N{m&?M`>
z)DR{3Ge33^*`*rZ{sG;yjA?+QaJ+HuvL_pK+zP-<(<LOQpLX~-ucOLO_`k<i+`%0|
zBl*KYBI_UwVbHwC;z=amWR&Hik|!Vn+(dZ-8}uFqFX0nor6<)pATz^R;k%Swy6!|j
zQOO{q%|AT-B~COs*|63aNrh?{FV{stqik;*k=8H=<?P0z4kMXV;$`)o9lg!VVMOa!
z6;av{gusU9HV;Ak5m?O&9u}@Wo{2fv+;y4fxlUtNZPHnI4$IUx$T1o+&VMvyOsxO0
zFU!dMzx_1X)cPS0alrnNht43xNg@}(LBTItyaP6jUeFIvO`I}C?!Lnls27R<{_Uu&
z{teFcP?|(ZACF4ag8zm4aC5l(`nh|L<p4>|jdF4G<;wUVINveLda{Waqlo|S=F9fw
zk)pb~Y?-A?!TOgM*c<-oJf9yhie;GNLt)u4uOGC92iznOPYswr4N#`3{DbAmedLIx
z@0=b>s3^m9UVKm2_eVZVBa1nD%Dv?c5ZY`er}4OCc582TNZt!hEbslvU>3=^WasXm
z?vXR7+i=UbP|o=Gy9*t)d-$tako$@WXUvgAE>0%AFMh|qcP5J0!*N)St30k4&GG4D
zi<S!mV--vsOY6U({Vild6lf^Y(4Hk-tsGdgn%GG++H`sw1VUtZWN7^mwN_PUoy=>K
zYB@F@M!Eui4fN|}UqhLL{Sy^O3<4}9)WfS{SC-ZKlYHrqtOR-D7gU+(m-%ROG(o|;
z_QJ+IZ%?M%K?agn;*7f4dxL*5166i9z&~@9swu26NmZX!L`SW(qj|Z6k#JyKC_Ln6
zhM-a&GF4k{xRn=7H(8%`6xsJtCW``fGj)TGQB)|Pha?+ic23Y6XMJQGZJT~qpV+$5
z>A+@;d}KltWXawTC8^D?iKYEQwp-R_ILUy7o|Wk)b=!%ezqG8ETDrWS!~_wqnTg5~
zk;8a~%stAmW?sXw^((6qY<N?fCzJH$?`XBmgW_1Cu|AqHyYC(%Z^uPE=2Eaq<!u}w
z&>yT_eh%-!)+ltR&gA^~x(p=9cN`CA#VjrYk~;K<)hF;vG>PujveaVf=HkmN!}g@H
z6r3Y1c-m^y60XICx@M!Sz4ZC$OmXnmLo|<_acOavNhC%WOhV-#9xtac;V%RH^JBYf
zk%lcCql(VAd33{0Yzhi+VrTW4`L(LV0<RAvzpUydn-}Nth=g#WoIDxX8Q6Ct?zwDr
zaM?aJRD)hKw2L=caMzL;*nH5^B!RSIEeEQ*WIRhG7J5i0qa0gx$$d&|fp+1v#0T@~
zQU~p+vasu|U9y#0?X(UBu*8$`U{5iF`ny~myl?5ba865}+B@+XgW|<4IP0k691F^@
zey4OG2m0eX7amFU4OTv}B^hgrBnmHA0=BoUB-ltJ*OY$7nHAV!6+A{)%qF~C-AKdo
z_|wyNibG}lfgYTjXXMXO1VSq`(pXI1T0Hk)X<yOXva!$fK6pXYG8K$qY~AlNRKdiP
z&@xUzg`PgHEY4K_uWat35oowId7GJIaWpgQ;1nwgF5^xuq8B-$K`zgupL|*8nCq9+
z!@}i1QEvC9(?7wGQSKneo_G=hLCCDowh2C$eV|kbq(DO#){BA?xPd@u*{QfDgnR_x
zam8~;Tz%9zI`Rh#sMlh0T+TeiWk=`OPG_rtB@I3#lj?@<-4LncwKs84F0I&>0;jdV
zfTQ4;7@)+kzSk2jdl?n*nDDcuWEs~tmGVH|@CI&U#Ugv9vw;#q4~3N7F{PR0u*jc1
zJ%gG%HwC^|s8^N@gu?c2O~Y+?J3(8`EWS#<_>>h;snyPd1@VIu%P0&L{ur(|uLC7e
zsDz4W-FzJyv)WP2E;_H2aVe{P*G|wpdK#vhhjVwUx%-hlv9ZjOC$~K@#*p?f`>m{N
z_J&J&GW?t431ScxfSvJv{*Z!P%Mzo}iK7iIkyt-el41h=W*9**avSe!b^aQTG(VMy
zqv(YbN6PRyjL@X4Ix4|VV}Oc63t6lMeApe1kL|)E-5Z)_m56E5#!V3<Lq@9jrpSxN
zJ098_ID!VEK%ozXgPyQA=`P{bv;JBtSw4s%yh!`0;Kn7WoCJbO_`na@j=Z?5Xq%i8
z_XIGO`uw;KnwsN>GpT@hsn=6|;_~HMysbv7^=4~&`FsLM+#q64@@?8hAJbJ(N`(XP
z$+tG|vQO!8Y($}RC<Wj}uc4ySbNRX@MAxwxq!ge%_Jc%lr1<fq@W4Rs`ZQhMA-#}q
zNdjF7xVzaxNg)ZI`L@Dp7x=b>_I|aCO3NhJ+qT|4<Ch;P@pv(0*l`Dnalua5na|69
z1@@9*LVwv^<gF&8yg$F04=Wr!Za+;|75Tg?eA}!FP3MGKndJ1QX+yP5Sk7**G;cz>
z^PHMKjB7~Eq$LMNqkX}5ucO6+shlc^T&$tR=AW3kbLtVK5GyEbSktDlA6$LCUuD$t
zi1sz<dc*_fYv?jx=*^aY1EP61FQACr8qA|SOwPrH0HXQcS23-3%dtWC)kV6jN$z^v
zdd0vL{Jd)&#UyJ&(a=5?K#u7c*KOd!24f}WRgFSDNO^!R;5MKFoM1OQYwaL7zKhc#
zwxKwG6vUL;4_U!~Tkl!DYZE-@OQdr1MiQIQQJ?Q-;qOX^dRmN6LeqNCEZpnHBv1d2
zKXb>)2uhbkpWCQ^Z6C4QJXwd|$fjJb`d34Jjeag?&eMD@d)<AuJ*KS`c}#@97Jwuy
z*ip$;zn&kEjrzYVQAXyqbB?<>K=-1;Rnv3S;G0{N!@9$zfBJw17m}8jCI1|;r=&+H
zXNicz;vjtJpZ8z(;VXuGjbtcufiBG(mrbRf%Ig*q=L+5KmIXE>H*-4civRsi_OcFB
zxzE;<@c|O5Dl8fy#JoaNV>Q}kDb=NHz$P~Ycd@BS8<Q$AaOQ4TCh<pA=e#s}`sb$h
z|0Q@^_4Z$#1Ka=Wjc4NcAN>iFT0Hg#oUm8N>a81eg%QLR5cHUYELoa6?b^jWI<Bpo
zR?+L<`naYtPZ@syNaP$sM^Iu#m3(}QQP)2#Z0=p%JvRAr6EY2mMI0v2pTQ3xXz~n9
zXPi-f#YkA_`Lp};Yo2%VLU@j9iFyY4#NQo#f(WNz_1=xv{(27=OjsVpb=lN<*u+lJ
zyrQ&Ecy`85e;o75q9XM)!RNjHp*8!ON7u!dnJkk&2FA?PBsC}yCdP!~K)DxQdA~!t
zBVUo)*Erv;18@oCk<_@DaO8vKx?NO6@v7y1ZH@YF5ep9qLPovcQ(40cr;%J@<w46P
z2Kp4xb*{J?8>On1k-t_geZ2ks^77pL>^KKp%+$%l6|n(j_V{1jaRx>pDv|K8;kLa#
z%HR7whnH$ylx|ngH$%7RJuxfQ-G_0pZzbN|`6W{gVTZdl^0_8;fie-~-%_cbxwI6j
zW5DWwKcX!rCY%fSeOfGx0%}1<OyI!haCc?=#HVZz-8Jo4;70HkXGa12ze=@aW(OlO
zVbnMj?w)s9?Lm$QCp`+u{IJgFq7RdAnx;m<tx0(?RE0g&_h0C_8!%NjU6KOlX%*!R
zgfI3hV=J=-;s7{=`9e>b#4)N&|KNu~e!rL>?*3s}x%*<iK+c=c$W^}EUl4n~Vcjo*
zk&ajrJ+4u*rQXsSDUS}XGWy{T+p6ZKwY`#7z)jEl$YQb4Yn`zULWH}HldX~F$~JUG
zAHXq878@WtUORdW$G<;#B){_gkrYzi;{N)Z9NV@d={IzFhmY=TGDhVkuaKv??1m1v
zBZ{(yO!m#}m0Wmjj(03rG))2nCMpKc^SED?sPgIgEyzBAq4<?16R~6;zj%LUk;Oa5
zEhxc?&N?ySQ15@!QmtZA3xB77RS_YX!|7S`PkZS27dYE>1qwjgpYUI*9Q@XJ!C`}*
zKjgn?nm4F;o}F2I)|=)wORRV9dLB02p`E6x;jN-Hl`(p3BMv4L>bbhRwW)bZX!y78
zUpDv1Xk-LdiLC6S{Hs9l<PSlu+)saCqB9#iT$<i=Q*AVzqcEw&(@8XQBVu$z<i*1&
z^Sm0|lTd^v96Bn{ju={N7POZNj?@|k(BB>DrA!)*1gEbX3C0o+ijLGSK0K080ZH`{
zWYH@HoztBNV(i+yC#<7R)>L$^e;ku~Rq5yiaYb_<mD+=go`PZQV#K>s@4SG@rM=P*
zj2$GeRo>Dw-R{NlD0~0nuGxd^b34JeU%(k<I&i_Cod>l}axbS8BX$9mK^;BSQ-p!k
z)3~@O3bD6*Id=u2i`9-}Dj!HO`FPu+6xfQRnVKKVtShmVuQrOk8j1myCyp5<JEAi8
zm3)N_MnfZg)Kj+9zQWQcZ5Z&)GB{X2Gj=|^F4~q|4EU}JN&WLLkjEuDUS~(!94V!Y
z9;DeP#NgE-xAmIghY+Lja*PztDmx;fdwI%Arov@4(b_}<8P9bsRyl>cHrF4+=1?Kr
zoV^-{(!!0F_y2sp$hHq|6fxOU%KVdsnTeq%A~4+T#5Zk#6SDuNy9{(mKuI}~Z_+wc
zjO^WSP5Ec4u`VaFOqy~>U#@f^?%?*1eC8mZrhNm>50t$KMhT*q$GGw6&df#)bQ>wl
znO0($4&@<ZFKp$<?AB|>_6|~~>DU=Csns1nVn*@l+fUc0sn`if#ee@DKJ`r8Xp#-7
zRR=S(QTwz-%9N<*+X18gOFgCLmv(w~L1eX8!b&epF$emQwpU-q-uEcXaL}fr#mSnK
za_0a`5=NVpa=QxmiV3A`3_y$LV1g`KF)Xob0rFLHsa>vpReuRyg}S%Hv7r`X^HFkh
z8kf0hn@2cFE7M`tQv<Kk6bHYy-TWV8KdN23xz2uzLL?Lh1>JJSI>aUoaH1pAiPr>I
zaxv%Tn;uf}iiv+wg|tl>4}D(yIq8t=(UTCGvQ;0ux_~XWmvRNXAb2>A(9ed7B!aEP
zT3+OmKWKjCRl?bYlZdC<@UeB%RTW!q4)L_T3tKiW3}GUA-u#;-KgIqKxR)wg%v47l
zJ;Zb28ZwO^D|XZjgZDw%z41XC$XyLY)Mw&H#JC1i7fSl4<X%F^K~CB1rjQW`Rm{`<
z>HD!=WRnYvO!;ScFsGi;1TqzzSvn+!fZpvo7`c>FTM`U^t!=2u-=`6I0sdOeWWvKR
zRV?S-4gH(g&Yq_g{|*xWTh1(ndUqJEQ`)*(ZY03#wnrgx_aRWv<1kNkr&qK)P$s>!
ze8r{umb@4Ft+pC0_0Hz6hIF?e!hTYPn3~QE$XUj{L+{Al8Xcix!5jY{jO)Mus*ddc
zFK>g1h2?)knN4c6#o<UGPQIeugRHNsE-pgU5!Q7egHbm2{&O&qj6&7;CTMLhJEooz
z$|M=@b8vevfA>&sdEHt$n%}+j@!(Ii4vK^Y%DM-Dn+%LIXlOocfd1Nt%Dj(1n#Qo6
zowz5DZ%mWSCG?l@yuXbk9EURSG{M+;LB&1baSE&JlIiXeaf04HguC;bJa-b%v}Xw=
zjn57S?CT#-LuiLKlfQnL`<;IC=OTRLdd()nzArE)J2J8Pe{&wdLD=|6HeA9L+57p~
zcH@1`n0TA1T{Q{VWaVw+G~bnIF=*E3KIVEFb^QxA-uGor0cf4KAuKg$OynFOUFNJn
z85SPiQ}xq|J%v$PMj44u7->_myM{W+Ot#XRKPwOBRYkn;9YPMxM=`HO>U^G>z=4&?
z&xhhR5IP!($-HQouCbq%^64_h^}{_i?Ku-6EC`yZw92V<>lofssiT^5HE(1<OvP1_
zI`%nDd34SYX-6JcxpVV-X5EMT2jV5diL9~Eqlp#tCk?<?>fU`buEcCt9*F_|VkonG
zpmved>}Om){GzDXVyTuJvqC*m4DJ66+YcG$S(|l;GNJ)=4#W<x3jqvdd-<YMOJ6eb
z4dlpQ_^BJnU>-S9knWrH<CIcNjR<_{F~^J!`2Z{_2@~vVkHk{(0d^NZc~8w!C1Lua
zW-se{rU|;>4D@|+l~`;3gXVs=m<!$MJ5GCW8oDzTGKV%t1%ft>eKJ50EZncF)U%Ar
z?J9%&3@|2aH|d&X<uG)_hLF?I2c>HG>O~m9^BEjvT}f;QPG6Yw)-Mk@c0^fVKp@*f
z$4*q};9sIEOgOT&?bqrt<E-#01UD*)ww5v%mQg{9C&ZkTof{&7#!KW?tW%0J;5gAU
zR%)ZRf^3JO^y>0tIWMG>B4L)}Ofi=YPrSmEAbikAJ5GKn-^%8ZuGHRY8s+7OWBTm*
z>mRq-<`_P8TUx=RxBRwhQ3w)e#bT|4#<XpWiovEzy>)dQ_0@?|b@NAwAb8N8t@k3d
zury6+m-M*mV`u_Vt=P(FFqx~)138((V3bNLy;xKdzgE-`HQ#hj$%958nlkZgC9P2l
zdML@u!yJq=(wdr=2X^Z{Qpeg%-y*T%#otQjPkrj)b%cR7%vv56CjO06=D)gBakzGN
zI4EFnq*R1j1m)Z=h6{#?%a7>Eqbm`aT9?>sVW`P3zg&DV*KGtVRS{0la6^Mt6g-Jt
z%a4s*;anr8)Lhg)+<QT_Ly^XyrhY(T+p5nJYN&$TGAl@$igt%k7+L4#J81UZOTOLE
zY)IfJNj*?(wvIX$N{M85Y$Fi6L@CqQI0T^KjEC6A_RkuhTN)o3*Sj-aHxdaI&vG#B
z;Qo}745YpwJhO%h1rgcDR+TJ~Oyty7n_xaOJR&px)A>A4soSBBuErTEc88(ee0lBT
zxjn$|ILmAg<JEjzSFm`z8E$`46DO6fr1hlvz^d8oF+aeo2<h1s`3X?NL%V~N3c`@@
zZd#a4<7rmKs7+hiC9eJ|k(9w+SaQH2=9|o?^qJ7t>sfQK?P!Gg^vHq84+p~s)&y;@
zaJ4iMnHZZ|j#q&n-b+zmYen{&bmH}swgl%SF=7eMB~V^`YbTHK9Ra4E6*IloT)S&2
zuQ<2WQ#g*^&mMx0-U;)5KkeT84n0o2b{9Tr|2(h-br`8r0j;IZ=zw(&9194Ch<gd_
z-Kin*6T3C@Xur-NuwiaG&(K%)(Qz?x_z=i?F#qK~pReCjGXDwVRc0diU)M0_f9!a%
zvHow0naQ8=<~}Fl&v^3z#83l<qnB}~c>kAk79pT78+_CEY~bOWpj1+kS!_(ph%6(x
zij{0huQsY=X>wYeqc*(kd0;=Z)o~ohVWsoiLJ;5L6^t%=iS;Mq6?F8ee!r|}-Q>hK
zHl&3zc6eaV;O^{2cykAi*>*tHd)mh|$ae>C*pY725q?1Dlg2apiMwp-|9=Bs^qvU0
z^8P*Fjx(AoBQi72Nm?V)7Vb&d1--<7knCtTUWxrDHnZ>B$Ii)XN7xd>bobsoZW$*g
zF->2s((UQMn4}tDdf}0OdxNu8=C>v7_(Y<EVwgfvh>Qf~PBQbF^CxsB+f)X{9>oHd
z&-59t!~GGu)+EPQW;AZ1A~XH{@b*B!6Z=BJ6~T6no}Z6KcBeC~p7ed?`_f~os!x81
zQd*`l_OuO7+#A<Qw^5+gP@0zL?4Q%ud9uxSPhN{W9*2C3Nxc9Mxc>Vy{4z>|E)X89
zf`rF5*ErB_Gy0|0MWVBjZrL=CV~42MVumt^%3Ds;fJtcN1SfEH%JS3~eHuK=e9H=W
zwXHW?JgyW{!5QI6GBbatq$y|&?gvc)JmNt)u4}8PH&5ysZ!L~GdI-R+o@QN_i*Kvz
zpH$#$sv0*7j@`hkG0w{S<W}mMi>-Y>Um{rILgu@Av8l8m9zEywgmF{)M8M^#h#w$b
z)U63Mmzmq7I09*DPHx3e4srxHD7$*+oujolk2CkWUT!ag?d(7beV3M`$r`IqoZm%j
z56GGz?ar(XIS-ePvSuU$$H|j<LHya`@$%74mr>|ckj+0kfHFIvjCd$~*1rfmUoFgb
zQud=Ioj*bx>y9lP^)tJ`!&_E+fktD<hF_6kju6boq%G$Wktk%M?98wjabE)=eg8x$
zMf#hn+bAt2I5Cckfq+4&=eTYuf9JWG_bo<vLk&bE6#mJ$3mSw5smn@L$(|@I>socK
zFwO6GO@44HV58@EdPj{v5*_MUcfNxia<UQD4w70b52G4VXr>V}TAJPV)~<&noUItv
zqlr67R~zxj<2q|n(Pe9s1r&p3scAC(nzDE3mVI4W6Q$fJ2pb|+v#d<pT}%IwNR6aS
z$y?Z=h9>T5!rL6xc?mg!lH=BeGHF@wtfUj>t1F5sJhkWvu97W+x6`bPfQ{nXU?Q4(
z9PHB`4Zyfi6%>%mnj59267cbs+=Nq=(zar`A7c~%26Jr$&?qndk)t_?M7ig!=0-KN
z@qtiV&N&#7t2u|LcX-KP8Q24xZsejBl)ellQpc8q0Svaa#neI)(UqyA>bFYwiyx5a
zwks%{FWqZi4>bS26TW#>%X+t#-*{S9q2ka*bBsujrL|d(rKF2X3@Y(h&YxH$O~ejV
zC40f^IKCF%B3de8-$oCM6bJVsF13}yJ>R+f*{Yb*%@P-}6{;=Hz=KPCz^mB`q>=Aj
zX4%W<t6T3=6b6079=rbYnh!JHr?I|49m_iTUFyVVJ$>xTLWC?wzkqw42~|K+pW4QU
z(mo}yBujskyo_@=#Db>otG99}N~L|x>^rW0Gb^9l5EI&*Y#!xoi~#yB!iG(v*AJ)A
zg%zg|5M&7|VW6Cv7g~gN6CC0^e@gZ8Itsc54rVkPwGR6)z*#A|V5@!_^iA)i-pJdM
zEUmc>!K>K6g(S<D`T>*5N5{Dn#*0wncAxr0(-FUGlL69$abW|UNCFp0ocU{-2qK$7
zLY2oQ+(6;i`aeU#Z9AQ)^1INUyRq$;a|iBA+1ZTl7nj;66A`A7r%D`o!Za6HZJg;x
z=!brI*OmO<*!sGIMhsCL>KoNOUQML9;RvNpoq$bktpxPMoOOX*IfsqCIE|XHZS}pT
z<v-;4%sbE70#Uwv9Z@$no<$iii{P2L1+o=8hrKr7rrs9G2$hVaozAl73;4_h^gHi@
z`W(EY=IP_846n(w=y}}l+#3H3ugd?S#s{(OFZjJ!uaN^S`9ci8;Dt!Y`L^lL@d9v>
zhaK?hU502s{m}q&T2NXw%@`%~s^QjQpcKrtXB1~0`K$gDKzxS(RadzFL)gK>@jqf_
zw0Z1FxR9qlX&xZLCXCSZ1BiN;G(c=xK+MCV;Q<)74)5Q=H2hidc2acDpcqYL^<zf2
zl@wzQyr0LU@&bBW0{VU(#5q+C*j3!VU!K5zM6V2zcC2qEKp{e$zF)3yBb6!RH}_Kr
zQjw$ahu3iekFPjE`}p~<yMSNE_GyGMyZq&I8myMEP7JQQ@;lRjnW_CQR%V5p4+nzY
zdjSD||HMD5<9~z;Vg;~df-KRdfK83$n&7fmmTzLUw?wRWWm!y(wob;kb)UmtN%Z|?
zNjxho;nL@jPeS}|cNY<B!(0k+Mmh$&6xM!789^_O<+l$UwdQu8ufe9f2D0cI>Xdq?
zt$X9jf;`5avQCQJwaisCg!hRDi0U~=kP%{Xr?`Wt=zTlQ3HDSFRy0klBqFqTj9$no
z?7+n}Hh>*&G_#r*75Z_Uyb$R&35~d=Lt7hJI;owvDx;1GR;C_v1IifoRJ-EEuQM)a
zM)8cT1_vr-&xxS89vT9Wx=CKKYW5>2R+{i-Q64{cuvp9Lx*b(cnj#~dp5$eN--=QU
zq4GL;Jwov|;`CEktc6uAk_DQM33G>xV6~aTqyav&DM98w3nUgLek^LLedewBN*-c_
zw5y4IC;Pn3(y~&%kt%@#Fd;UeA4!_d1@Z^dW_gBdi9)U<bjb@;kESwo*RqWt0N4>>
z=b`+&R=D8t2Y2j+exV#L9JO5lWZhIAoY83xn$x0VKJrdQ{%1mcx$iKtxPdjLC2y*|
zHs`S-{{%d`XoV!EmYtSrSp1DR!de1V=FsZG<60cfa&r@T4yDUfIUZNr%8EZ5`QIte
z62BhcPt^(<H+&7-s;RO28wp1~MFFAQp2Zto9;O>muL#z02?rgu3!}fxS!>M?)LV4)
zG!P1(YTKv;RDkqy83JW<UxHIq7OCs#SAq=Y8|&R1tUlhUB65n`w<ZZPIlR$wsWr5W
zzfu>$JI#h;ZiR)=75}c8&W8S)Nlg`P?8(<X8&Fx(_Qz*;9(u~$tdGJIG{mN`@tfeJ
z+I6_;58Ur1H*Qm%rKu|<p?L@Iv-#H@j5>$h)aC5lA_lf!At6qS*6Myh7qJHm6cA!Z
z(oQinc0*Dlt*^9bFnAO(Dbl>^HTVG_4OW?3_g+w7R|W_SE3Vo@y~7%SAJ<hov|r;|
zq!C-&+nlm;sb0bCgH%J~2?lE8r_;aXoP@-HxwL=so(#KSJYt4b0#pY+xDTh-{y!$&
zS7XwwOEu^*<2eGlWRNp0zoW~d^iE;Rn59qX*qKfbSv}6k#c!W!p`-SLS!nQZq(zg`
z(GExx2_^;F+*S`Wcuy!AXrDBJ#Lt>T^Tq-fL_j-;DcLe0PpFYLw##uZ7)K-R4jII6
z@07pT{Z^$)PVTx_x#8a*-W{m_kFj@PvMp@3HOsbb+qP}nwr%XPZQHhO@3L)kS9P5}
zeQ)20{yO5Um}|{HFz3pVZ;pJ%porsY?eQsa`fC%F&SF^SbIBDKwr@^6vEg!aSU{ql
za9k2PF`>Bi*I!!o9VX~{j#+S)y{2Fydtiz6h=W&p1(Ph3JMl;_RKXFh(*+qVg19y2
zigo;U;_KEj4I$HnQ9SRK@!sud6mV}}ky~<j28H2ch2w$#9;%kYaNcMAvR52n+P{vy
zg16T8V;O~^k_S#d-MK~U#as^CE+yl<sIvVc9$5xhfnIP>A~WXzvOG0e=v5U4cV-lc
zp13wfM-mdg`JvAsRe|Sau+`)q0lu&xajbQ@Z}e@oex2N-tm7Y+n^*uLU8Srt*~ruJ
z8h#U%#r4QQ+H5H_4kCMN>B)-Keiss17nHR&S-CT8;FC?R?~QIHO|C7pZS&)p7aYQT
z*;3eD;R*_iP)}CHD5|ILZ#Butqtjm#DO;}oc5Diu@5fs<zrK)8(?c&RomB+c2ivdN
z$Jma}B;$lfF^%G!OObQX_lc9(RtqCraIVEd??~&eR}ZFEKbS^sqy3NMX^zXVHMPhg
z@+RJ=(KuH=VQ*{>YaL2}DXfgHRP1(YXrT#A_g!L#IeQNB<V>}myAM-7YaeX0cADke
zyFYs#u^D5+@)DQF7cQc0DY-gWj8;6f{zN>w(%34I9Tc+$uE<G`8J>(8J}$bma6<#?
zPEzfaH(w#%#=Bdjr^tVu#PL&^Z2jHuW->y6;TULO+=Qc+c3k*B%bjlqI4MIJ!fs-X
zxd(L%mT*Ggp44~;j{BX;oL=086bV;wch__wd6VUDABqK6;G>W%;=S=Az2)yY^Tm_1
zn4-u7dU>u<6{rg!e{Q>fEnj76pP9iuLVrgvN^ZSEpO+lb(7lr#CJeG8COEC7Gu^Le
zfJfjb!?799kdmC%^QU;{@`o2_*sR-<9gQHt=D6Z0405U~Rlzz*d|y)PW>C@mTqNvM
zNhq=`tAF_q-)reIxwZ$+22NGO!jM%41`pR-x1P!T);>M1`}()OZoA6?_ubsNc2+Rt
zC$Be_VRIhjP5YaT+zNZ38XS7`vn{1Xlvs(k|8fZLBRv(Vwnl=u9<avZ9ZbY0?|dIb
z4>{KufYye#HmLfwXWJZtkiww<{j0*D;IEG~6j5$y|M@cr5oT-xXtxs{vsC31uC*+9
zGIu+}>SxT!A=n1X)l`<+nE#@18G(z1;kJ$^UP&F`e=ZXSi^Bm2(^QVSb8AYW$-!HI
zSZQ-%z*@#URd8lH!@FIvT+Ro@-FC0gb3NpWxG7l5b=#}-s@4rzs}qRFxpT*u1sGUF
z{;=v>yqJb0V+}_tcG7}utuJieiPfNexyKuIhZ6|~r-t&aR#3~#q4Wo|Fycq)0CUZ;
zNY5mVvrUzZ`rE;}r-1Cwr`#WV#r1L;BXN^#+zOlJ+5n(})E>gtW3qx9rimN59DG{V
zkkYS8QlpYu<>wMT=oJmzjFXP+z7dqsv{W@D&(zY8>#C0tM|cP^n3ps3(|L{}j{MhE
z73z)-g;+F~%)%^-3fuP<8>{ZjqN|sFvSsB_Eo~=ZT$R#YwcVZSHk=Yyj|yf9`sAUO
zu6l5vePUF%y>#Zfu?##K*$0Bu_CQ<4RPXjt`m1J~GO(~A4c=jov`chR0j>BZU-O9W
ziGHn$o61eB$WrHhZMHn8d9~!4WZI<`dcy8kd!ox0+LD*49yZZ=+X6jAb(QhZwi5sa
zTkZ|w8SX92OvDA;RoQ;2en@qBgK^*~@>}a~@RYNd`TwBFm>5|9UvIgLshzot1py-m
z<NvhR|8Jnnr}`N@%@84{4*)dqiycmW8<;n>7gc7qlg0N>0W*rEWU<~$WforZ!{Tx`
z$K%XU;`XbYo4(vy+ix5nTMjbbK*i0qHuzI$aUX%@g3;-3c*K0qzq4<TR3**MS{F>p
zmeQ#Ob9r|je#^UN;CQ+Zw7st*m^<cX@v3i4+fQ?rYbAOY@J@Y?oLu-tsZxY9Cz<1~
z(@lHwjF6dEq5`q~obzhtAPX|cIwW%Wlx?uk``QiK=PAtQLU;+FLUU4_AM$tpP6-_s
zQ`Q;`d`+KD&va&7MWjkY67bNN3s^*0@OS4zPJf)pVfB{Up80k<m_mounaCk?4_^r!
zFCC9{1=eB$AvSF6NV)?fFk~cpWHr)NF-TtPzTB@)gELWx8iP{wXb?@)>uQCgC+F^r
zLDwAj_(2S9eppSTr>4NtW&vz2>mUcJWC3ISF0$UM7?O|<)ZHW^jTeVr--_?gFSn0V
zl_c>~5XI~7561&~nEd+KcBy@d+&q8xUAewB73aN$VBL@4#?SYjDz#oR639f9<W(X$
z)Mg+__o{3FhGPl6&wxm&o~KBv^kF-$(o`a?0@rd%G3Ew;)448EWz1COpNIo_wcc2T
zL1B@e$i~FSbRqCs%)rtvS7LD5c}1y3BtG(tR@qKV`3Ru1{FhDT#RO!}d0d8~ao5HN
z*YKG#9aPFF!VXHfY*@MNmWFIZHCC?r6o*-_w#THDZ0S1>?!q-8PXBy;HI&cf1O!FW
z5SiC$Qok2s?n)ETD^eov#t$;!OKGOAUE}8KEoApyY?@3Y2GRsAU5^}ICnlI{t7zi7
z*GE<mOD*=~Egt>DkfPz;yO~)Ret&C<cSGTOXH*+^1s5#~25(j(3h~Rh<v)SZPIkCZ
zckN;*duO(X*pDL_aC&cQ|I!ke9wQ=vRK2xUP2i_2&?rVk2h92~?`#=ii!x47vZRMl
z(pm?aL+GPWS*Q@CE@WBy`J#!v%v3yN_p@IzNn~}P&%G$e5%wp^6E9Qb5alX0#!Wfo
z_s0GV9P{>jgT@ZFFx5eHD*zz3J|N10_HNnK(x`G2&5ClVC}a%FmHgcvNot@NXS_I&
zq(mH7U5zTcz=JLceM&$|bW{nUaPA&@ify#`!4rBQPEQnl^3tz-@fHx^eEAY1^w5<-
zN+t*aR_8AeV>;d;nlu&T6~CZr%8RiI!+oC6HP{}y^{Z)a6Y5l}nlkV`6n}7x>!x~4
zE$yaWW$*(QMDO~!dP_c_O)V8cSEu=h!@a?}!T9_WJQdiRxy~Wa?fw)LX((k#I~I$q
zYgf#h($!m<t~DiFv&!qx(?PPV004XHO+n9+s}c{gmaSG?%!Z{`40E1(#r?J*b~q*5
zR#`#PZ0iG6PM|F8oU=du{Gv7+E8qWwm!$UWaANtwxFO%UD*wngRdwUZm0}=;!#<fs
zc8yAvcBI?(SGsh?B2C&K(I1s@rHpmyj4<4;(4?p!1%C<~agQfDuW(C)=jaQCn1^ut
zFZblX#WQ|~`XLI#=FUJSo*`y}&#TJ&j#e&r(i=}p1;xzBQQ-1R=h|tM8PkpVVGbC=
zjm$S`Pnj=3Z%WgxNbWi6`$>q(4<bHapscm^8;2<6fG=FTWhxYK5|I%~O=HAM8i+%z
zcR$0Ax{Vuc#od(VMiU_6Z9FHUy5tHWD%_?TAxjjof3Xe1_@@XHxUTbq!k9w!Zt!?K
z@BkP9){*e3Ig=QID`L8CL2iQ{kA%Y9xY8UP4jf8~JPsD>iaQt^s$KjR7G<&slb7&#
z8>%|HoXu6`*SH&`*o8M?{JvFudYF0Etr;&uaUTWNrg^|}pfvBvpJ&P$PRGg}z$=-n
zl2%4HR+XD&%7owBvUpyGFsnZ3WW1A9hQdYMdiZv9*1q_8gU2ULtHZO_2dt>u4k2WM
zh`H=G_OKz_D<x9j0}8i1Xtco$1*|9BL8))phnsX|B&%zQD!$(;7Yw%-$=)XIj|&BS
z3}v#Ac*+g1D;wTatzF^R6Sk#Tf!fcR=O+g&;5aBL!)VJ1<f@7|TMvrj>JsDp52dW)
z?(Uo);xXN*7=w8Zd*j&M=<SE#^+n=44c0r%FQr)e0qQ?d4X+Ysp>B!3YU)h>^GBQD
zgcBJTZD7^;V$WwC><V)bgkM3UX__VYMB%TY;{e4Z6!c*mkNx``clLf3)L_)gM_VT{
z_#O@{d{tB&m6wLQ@va$>B*m(v&65tFjcC89Ge>pWk-92OZD~ehwb8Nc7?e^dzwXAw
zyyQZKVKMa_Fbz%s@j7$cV{*T~C~>itXQ$J5IL!)e=*|X+zL4-<-#QEFSrykGa3?M*
zjG9Wyss#8~da#=$W&{z$Y_;X1nzii1oH{dsbz((<0$c6*s{u}-1%E*EcCF9<RAm|e
zgYAd&m%96Ze||T(Y+Ov8=*4UdT}(wxjqOcL`S_rmU7Sn}ZJ|6iR(`Q*|0}dzqedda
z3t#|1(yURy9(TIttm5p(5`O!~pW?aEbV4pWno$%!W4KX72_L_3JYgtK5N<bD?p9u(
zUJ71~11M_f^A+$d^r(-(bAH|cFhr4|`v>QnLn0!KpZwc`GUXudcsI81{5=Y2?3c1K
zWDl6zrvYmSJExewN|m%qksE6M4ff3!=BRgDk`Nel8V8%4`{VO-Eoo*bKA`}~<j0@}
zU8+g~oy;mNluBvo4bkUugd#Te=Y%uF5M)`h%-nbOlAwlwxfY(;H+lbZWB^GDf+E|g
zhGv+1UM7!X3BG&qRyvtXQD+oa5^`i3D>4vj;5uhSH@zEqGJDR{UH!EX0mR~hLp_<*
ztg=?f%o9H{`VNfSBuK|Y7tR=T-Hy<N6*mbZi7J@l-C0ldoK>+!4XDJ9a}qjO`Xk;m
zmr2%Z)L<FX%Xu6L#h5EpQC*DQGp70&Q#Dn^3pR3L85ZjN4!K@n;^u}sR&jL?1A%mH
zWSw|KKsRPocL!y97M1BdW_q>&N_H*N1#3b={E!=MXFcE{n<OH~kg0cb?<)T8_HH#(
zcOb3+v=OECOm_w8vT%i?lB_Mxx;z)3G_Q_syi`%SokVH@tOdS{Qex?7{zGHc*5$|R
z#bV49JY`zJc{FX9^H0$-#W`aUX0snACAi<%Q-&jlnhsG=I&ENMUi8;*YY{J%o5n&5
zZS?nMYu0`1P(^dSa0eCY&g+ResU&rT&K7+uudlxV_O3`*5-xE62}*xAvuy3P%%2Oe
zS=cZltj~`x6dg+f<MJ1JY*7}j?+?G5=Vw20@~8+kVZDf-I8=u(%>0lF??b=-Si(4M
zj%FQAa8M-U1=m+}eVHaHeoP^pjyuii^^Eckp)lEoQuNKB=@OZUNsW6%<6rpJl+Z>6
z>VS}x6vi}p95NtXP;F;kB?pIRuo>U|8}q}o9rlEldsNK;cJte}&t~wH0|Nc32(|||
z?8h+bN?A$CxNJ1AE?$si3rOBrV6oA3T*qmQM66PNNA{)!tvXO<>yVTsPj{=kbwTRl
zP}iq*tqf;atqx*G>y}_8M(0hXo|>FYb1c{dm7bmJ@NrF{aWpJQH6(YIY^27)taZ_@
z*Si)H7V@;~S$*u7ZNUY~eOec8#wl(`If6_-a>O}r_aQZc?n_0T)y_l`g(}h=tyRUS
zMroOEMX}m9@G?1`qD{?|s-4IWA8M3V=E-!P5bnKqKMxLG@)cVQ157+yPqjIGD)!*e
z;-hTfVre^yTXw0L8wy@q7qFRE>)ARkWIlfy2PU;x*ru*7)V1)miS74^lro#n_XycK
zEc2#cI`nh9mfGK(L_6aP>!lt)nK7x!iz;a)nRvVf$x1pDpBEmux0}n7O~!Leibai_
z@wT%@_e?sGQ0yKP8XNB(@7RUU6Ign|6>-lq6x`@NKxO=*VV|V@G+iV`0(1D8D{&&z
zSXx+ZsTU9{Qlg``g?j$AdLE!4!%L#G63x^>i+z%7z#vtjz?-+uO9~ERlQ><0Kw0}p
zPW)HW>z+IB9ek`HWf^Z4DXY`aY^zN6B9a;r$=ANf#3^^N)fgw9>TwQaC9OG6cOl2~
zndGBdv?>-t{aG8__rfpJxY8~=lPIkkL2H<ZqIKd_F-SR8OdC`B&?XlnBc8etl4l>8
zbd-7dwkfB6v`Z|{*t7<s@b3&GU*bIAjVII`aciW>AO%~|AmPq%ZP17lZZv!~9bWI!
z#)22?*A<uO8;7oB-!)%BO+d94pdo1(yeN0DnQ=}F^}W(heXI234?EY)Psk7qQ(e@_
zG?&JW8_LO50+u}^WzZ|B95(S6o%<M9Rm~MK;0?EIT_!Z{qJ@<=F+c1~hU^D$Z!43&
zD$>RSw;iaMze+oc!w`TXLT!ypYtLkn2P>1!Dr%Lhk$o+?HCjc9(5ja_oeC{tpsFjK
znmJXgo-PGeF%8g*tFg>8rEb~fHZ{1!AHv5UX+o}&yCdI~IkhTDkDHZ=-SyV&HwHyB
z@^8y*{P#vvwGO!+<1;T^cCY|J4%McQkOgN7mTO8EMx-cm$|c^F{pPZr^T11EWGevz
zLq~Wr-_Q4CBvyQ99I|%B@Q|=MqbC1cWd=r7py?XEwR6up^(Pb>lxfqlqw79dAgk(i
zTXqRm5u(Fax;fB)ITH1Ct&HzZM_y-K=2<E&%Ro*eEM{OGX<m>$^?G&X?i0}oNk4bG
zX0cXb6(X%(<(5YQ7TQ=1Ev#&?%740ZVX<Z{G&j^zy<;9m4S{`xkA_N*9}G6Y!fdxX
zYEw{{Xj$?WrHq>PyLE3pP;Dq~{z_R<>{^hrHt|=agoyaqv*v9zTeSPtrz|UMZb)sQ
zi<f`|eA?ig>bP5dv#8wKeQ){R6OuQVrrugjk+c5ET!(M*5y-!Cc}+--J8pvD0?8se
z@s%xJMzf;S>-2b#@V3A{DB)c`c=C`3e6NK+YEzlIkkYIO)6~p~@?N{vgxjIL@f@6{
zyiu_Z8ZwgoHayLO?pUP6$+7429#f*%AV<0J>~BNi*Si8W4rOW%e&6^h&xfK4;rrW=
z4@D^Priwl0lZo!lW?SN^e7%ghU;0zBp~AgA=|Q^vO5!r~I)G)V+;If*``jrJiv1^5
z@PF=g|L@+L>3?Q-`o-?sZT-&KU)9gRryVK|fG}X&aoqztV)p+lz+6&}Q|9}JVMLXP
zvV}yuH6hrmKw|YpBWpMu9~i$khmO9CUL7>p5G^!eXzEpP9#$?dXeobm0AipSX=rM0
zzs`HMr1Vk<-JupmH_T7_WArfxDb)<3IO+z<v6q?P3oEn8M5oF{r^q7Jr`TR_7v3*>
zmAkfMcVI!(IfU=$-@{l0xf!SP#?O=fbzfQ9l7HD?rwU99HVYP6JI43B=L5R{!O|$K
z5)-+>+w<++zbEr9>(Uh;a8XUc$OA26waCOw!VN*|kz__qMTsCFN-#+Tg8c+Kf`WZ(
z9Ppf*1f=;9M!ComO2=Bs$67^5m_Iu2Z_maSV!3^&@1z-Woa;^`y&qLdmYRZ<j@Lf+
zMstj0atqGz#HW>q$zZjo+%qF8ehL1!slgBo=G$<bDDE&i=FoGN0DFgico4KoQL*5R
z^1H@@2I9M$`m$FuORsC?3rxS69x7#B6=s3+LFct?pmRzyfS&grr_4mKNkXV{yyR&H
zv|O`hVOmqx0>iBM-~Ft^sG#cz4T>a8o^Aqh)IVYyig*#LQuc_Q)k3JWD=~cjQ|3)c
zN<^G>dQugw*<+J~(q1Y$3<{kk2sGJqP`RcYhz~fpCc|>(kb||&BM8W^fk)US2rZEo
zB0Kg-#B*2hCY{IxXR?+rw_ezIsMjDqGPOk0)5*44hD$pBKa4?oRIF3Rf92Esu>;fI
z7zc1XYjUT1N}qFAhYSTkfKfzUjgQub|1j7OwAuacby4t1OzX&~cqg}`{8+|XQue6r
z;F*{^>hIe`y&>^1+a*w@s01RHWjB0ypb?BUeI82z{-lP*KqGQ|V)y+cwVgt)xEx-(
zma*w4WY%emh(VlID0Sm*Fx&#~xC^9wQ6-+ETzC9ZYYZ;Wzy!`_H{e^_q0&=W*D@va
zEzC)Fx{d`e795O>A7JB_=ELr!JTH!iy}nu}IXwub3*4u?Z?0Wb^`#<U>A?LsGI83-
z6^gwJ0a~#-@(iA}t|X-CoS>9Q9zyJKvol-9d1qUnV91+RqdYK8?2Z@4<PTz58DFl|
zG<U&*V(n(x;Or+u*~n29H|pNGPe&r`94yqx5Xg$6?rV@I6G|W(-Gw$wf5jweYh<Zb
z(7p~lX_pbp>~f}NGC?m*&{Py1Oaw51ny^)YKGC945#&61%QB`WksI_+qqDcuHJT;K
z`!ir5xICSXqJk1VK(J4NLOL0q+{T2BdHJSm;6}(%XfoJ5&8Op93tY@yBtLAo0@OjZ
zfjyTDDn)+;j(QQv!#n8?FbNSa_d2%nVu1<JDzsI3-C`EYW6?CouFK?aKn|y0r}ly^
zG@J{ijjT`MCQ;mMJUl`dVfTQuskk1WB;~z;KAl^`t5kGr8?Zyr@>;>^pty<!{Lke`
zj6!%9egKYER11T#9ceHg4n7@D)c%5(0;H7XiPZjuQ^S?`%2zpe-DasWuU@IO)Hl*J
zw6nij^if7)fwFF%I7ljj!HoqEsipsY<}ROvVZL)|@Z((TO4K5Mxn;?k%FX`4C?1mK
zdNSut{{XtUV4GVK{ewUXad9-PX--qfN3yeY$q`&hb|C=^a6*q7Hey4r;mDp_!FX(R
zjwXj8{RV+|n)4v}<g`^%q-Y|cKZb)!Ww_Ye4JDTvS4LPXVq)ynL4?LdJ8o*fIB7VB
zw^u)&5kZJ0Qr0tzr9Z?wubY-3Jw0ROd`3qW^8|+l#!sP^j8oxmgiT%8WHe&PnH4fp
z(r&jW&_%}&tQekJYt;VO{H*3LorkZN+Egb$u5xzJINWzdz3HM0Q1@{L^a_zlYixL^
zIh~4*^q^){e@dCU({YE`qz#t<)dWfVE<!x(AvS;}{hNz`-ANxnBccH;LX1EB`PtKi
z0L;vRA11a6_tY*T7V0i`yyDD{i$S&HPbyHQK?Tu6<VS@^g;fzL0i@LED9RhVkkZT>
z7{v-SAFg!O7d}vFW~5UKrZQ>+q^iojW-P8juHHgu1aFEO@R}*RH4JG8u11_=Q}!o1
z_Rg{{0w4FP*?AyW`pfUPCa(8zx<!gvQR0@Pc$?bG=v#2jq}E@erx^+Ff*9X&686^S
zO}(jlBti|&=M;z|5qU`M2mh-Y8|RmpJSMte`5c>=vGAZ{FJvJ_IjZ(eR!x!Z`%ctD
zU~cogI`d{FrjwQkhsphA`jqL7+zg%9eY8i0q2vXO#KU?o(uCWL0evaHVa!fJkSdS!
zTnwFB>?@JYpPEqp+J32W>sVYEontrWUV_hEwY&8r(Y2!6o;zxI%W0yMuW!Dy|BuKJ
z#Z!G->F=H2g13tgNXfF)U?49304p%y?Gfp!7SAdZA72Ph<4{uiIBsraS!gZX`QJqQ
z3J)(+<uc1=o}TP)a0jAovj0$U8JYi&!|wlIkO~JI{r^*pnOHdgryvzYd+aepzt=jA
zeERd?7dLZrGxI2c0AzJbC{P9)bPH;TP@07Zr3@4bDu>U@&zS^rsRE28BIPCc)Yn+&
zF6JiYT_0y}R%K|)WgR60sO4r5<GxwKq>lh0wMRk7<X@(e76IWF;V3ZOG^fM8W_mD+
zZ`kipAXyhIXv7o(*_yu~EL*A1Yy^vjuQ8y?FC`4Mk$-~pi0a>yMvK;nghQR;01yx<
zlFL|KWWqa*4zQvdXRCndho;W@Dt{Kx$~WYf0NFnY13_nP5n;(5SS7~FBWn!`PUi3C
zApV1mAqSn#QU@u@7K0G}(5ksIN)*F2P3Hi?+Q5T|R9*hJROh-Brf%CHRuv%Uo71KK
z#m^Qa`jZzmj#UcK2~_#?mp-Mv)UU;3tS&VzMy5Q_ScKhlD1Oj_#Cot?RI6ISk-DwV
zd@F>EIrZSPFLh|6qo#fg*^5?9D6Nm}AR(%e<NA20XfxsQa<jIfI}kge`;*tqx`jI%
z)3!E3oXU4A2)dRIKdj~f!xW#7<uS3_`mH+XeNP(<I%!)C;4THjZ2l-nG-hGlD!NnY
zRuJUOcL^PH_Nra|pdp$r+mJifhj<c@@`NX1p=IoY$V9Lq{XTU@Aqa^uEVGI}bBqCt
z`@alL-<lw9k1B)|JBgpkPS7=TSaazoa6=T~qr)1&1M_=qePtK}822pJEl*7<(N;ep
zp=LdF44uzi@_)Nmc{UJtbH!)M0%hdzq%er=6VP9-F6+9`$tAc2z4|V`zvKK5{9?ce
zYdD1%qV4Ny<>2A{g%hKOK}6Z(i+__{MH)#W{zSr(R2+|6B1?)mlMJ7NB}yUYc9LUu
z+LNm%onKxyJt|Us3!LfY&X^UQbIgA|%;B5u_T=Wwh!M`&<H(LOS38DKOhRHFRN=JX
zVWXp;+-Yun0W~C|PvYRY(U&DXt`)6IzCMg7zM4H5w$8uJ_}7YSypJKGV9kjguEz|i
z*9pB^xiEL+<i^>ZSe&)_Vd(1Q`ZB*e9Hd=nE}Wsv$=?@~BfculjU7(-5arVUTor?x
zvloG%j^xvT{plachXck2>tQW(uY6H<ue0Bs{knK_`nsL>WzNux{xvP1H-AdM`o3OE
z=IVE2<;=+ih%*jY10;@*Ybixw0@i6ESa@yl!l!?`zHT4Q6ni7~3IUxA8o`XjR|>2=
zj13KWe|tqD^{4(xRvXRKdrU4Oj8JtF3DjHwY>!7^qKT{o0FL+DycAF^>>x&yLa9;2
zm0hC<WQs9iPu3P~WoJyT5O4PIF$6Ii555;^7`p+gYPhDbXaGc43^5|)^L^=$zj@z&
zI&GSoz`*)<z4H2fTa{<lbeYBPx*eC+lxMd)`Z)1<bGrGu2992&-`ThGeaH0ev(nR(
z*?m20QEU#0;HJ3Skek73hwegX<2LY)E2X`^pvie_S9-dA<mLAM{m@r*-ba@JyyPw+
zCG!++tjxy&Ha&P^zS-HUKyzZ6Z3J3C(4MOp56z>T{=?6B7L~<zD&OFBOpgZ+ARe3j
zc=1AQRl{iuBwT(6yw$Ptk~QLE|5^}wb9=%r1wJPG;Pi;Q&AD6PeQkYy3wUwq@CCT#
zPgM?G7tlGcETFrAwCpKhhZmm(est1`A%O+v=}jX}M3CGsykf)@?*OC;z+O|{Y>PTs
z;Ykk&VY}Cw9jUYl7>%?Fw%)lkKpC8L$to2v6#R@PJssdnJ6_eSiib$(zyCp)P<|w#
z^U9V4lL1mm!QHQ@u$Mls*@A>BgNnmtmB&qSZ9-j(&Qz!K*T;7FZ&0RN{f`{kLOL9m
zM5F=orAob4e?v`>pPF!)vOB;L2w4)~5j5;Py;##TReqJMWZ;{MV&Fe2`pfpB%w+Fu
zRpVWzC8#<-f);~kvW5cT4S@&8@1H*;9js^dd~_MuZ_w-Jhb!9&)P}aT>xf=jI!7b)
zN9a!|#jCAB%UXFs%PlksTHq+Ff#H|^L#})KlS2YwfqX#gS->d(<t+4pQUEFjJ9*gU
zew<l)xjj8t+fR2PX2JpT**4*CDl$cVTmnPU_26SxP?e-r6%jSZW~A)#JxcQ(mu@&!
z%P71e=y>$yk`oW7FK_QAKFnFU1#~n%x&777v)oS{jKr<;nt(V^L0gQC3&ho_N?>I(
zqWxWpsU4M9013{6mC~feoNc;wAl9UYPyMWl!6c7#lR<@8K?d+zC_?iUHEV^5WGJFM
z;CM+hBCW(^7Uk!9oz_vVw~e^70%rxNeiBzJtH_`F&@j*9^YZUqBXQD9RH~GXfhK|+
z+z~qhNV+^NFC*QmOuCfTqw<v2!m7`WYT=2h|8=Dn!_i9OY9(?tlQ?tBy9aRx$?eSR
z6mZ8L*jom+vz{B1$S4mjLi!5WLC!$j7NQ$gXQpWp%>S)hvoLfBau+ceREhu4bOZ1+
zNLzq6MB_r)A2;oR3(!X}s+B44HZij#mx7?7cb5WCEW$spRxdQEicd{O#wrJr_*;92
zHU7$bN4;va$^NvY1|JJ(7q_x4f5zW1klW?{udMWsoWZUHSOgx9aNBh>Xsn2(w6P(_
zgWij^{(vdm-5A~~OOoscSCZ@|m$q|t5?ewxvhs$A5=y8ovZM=5!v$H;*}_0;Z@Sj{
zg;b5rLx`5-P?u0$3Y&}He_z*j{oEY(fJnigLHlC-gJL--{O9%2aA%Bsu^eefW1TQe
z8v#Y0;A6DF#LviKA$9?jYI^=ViOR<DkO)Q-usj+z?5vP<7Jtd=`%;0`{cV$i6S>bO
zq<nxB+K{4XcH4vLyd1x_{<8LV<`$_*gO6!V(slL&)&x3Vpb>9C(24uQsy`Cfs;yR2
zAW~{3fdI(D>qW5sli+ud-@BPf0X#^wBdSoi6y&d83kOd=TJt#4C*T2%OAQDACKZ9_
z`X(G!v(Lb=wd;IMh5V7O{6sqL2naKw)pq)|0>1R%Dsd90#Jt2f19k~8bnp@oedA7#
zMT`qqh=%J1q%X21OoRFV1xqZCKs@gaB+*)x(CSU6nV}Z<`G%cCZe-b=TI}|4yvq9}
z8kIy(YM0Q?eGfoOywdr#5c+9vH=5tk%WX8?00DJ}|7AC>nIE{%Hr|q6)$!|k*hbgz
zYcjQP`z>btNPDs;)9advHbfbFCm)~cOSDQ&#VN}q-cnX*nx$m06`VfWqC;(}EWfwm
z|9T<ef;*vW<5qV)Tl>Nr7n!HktmX$9*H#|kM4{7Lfc&HKrczZbyfd%6wWNQZErz=V
zRfw3A&nEC|{7@DCDVK&N?%(c6ULG8O{s52<JGcG^naK3JkNv-Sd;d!cGBR@hkES)p
zzj26%Z2!IgFcV*W0EPj!f!7tdwRol-uN<GTgv0dXFX4zPYUR8MiF8CDT{>)3Q&&lG
zzrWcZ-Zu1pc((-Wz%Wf^cyxE~ir^BwRJX@czG4t!fHiz~@B8*V-m&bGeJ(Hyc*fvj
zdtNO^Pe8*<h=S{){%0vlaQ}3kncz@o-VttI`9a8B@Kl{&?h$uwOWDB%!#~^a^X<JK
zxhb3Tn*LwKKlgteU!D}z(gsppg49DG=tH-=@ZVriBh8M%$<vTHeqW3}ykCn-GuEu9
z`4rR=iaORMWWUEUl&0V{fq}orGLaynO3_4u@geFi1duVm@c0Xtxa52ttfUp5ggSY>
z@N5^kCSuGaygt!?3?y=sb8+^+Glg$872*NtT}JBOH{-%<CSXuu%MHvOvy2%BHANJ9
zG#8)6{_Cg04vtwUk2(|qvSh6-LeLH><{X6Fg}wB&PfdXb#(-E*#y%b_Gn>BNHg*-i
z{-yO#2(?3|N@%W?g^=mbY;1Zk_gDnWbQFcR3LmxifaZ_%b`Egig+H7(_s0(=_E(u<
z+L}e~FFs?=<d^<S8E;%MRHTzs%pZ>oqHk)F`Z!8d)+KE`xM(OMac5$?yG(2@US!(L
z#41&%b4I|6Et7)kxsY!Nx+Vuc(xc*0(+1{WiZ#fp8K$D<t)TFiZnL5u(vy=<YuTU!
zxM{=K+x#Xdmw6$BQ)g-@3}Yy47bHD$M5cL3JV2mSb-9~w{L`*Q&L!{IJZ~$*&rd=|
z>|7C#V=YTQ89Ev29vQkDCf_{zfsfJ!N!1D8s;hrGGc`TM707t?jXcUcIdy<WFhAy?
z!Vb{01>({BlIo*gd2?xD6v4J@5nSRTtQz6Z+0>{h%`h-=X(R!qv}Mh&+oPFgch(8Z
z8YBCAW8e2+I5jUY8*ucgN0Oajp0WPWkR`zooIYju^@fe0myAa#eV<}Qwe*T03SEHI
z^@(#r-l?W%uveo<T(rS5*Rkgng@hBG%R$9v-7zLnwoOTh?ugLc!KWk)Qy#u}ip`c7
zPawsq#35M@v+~rU##L-(%f0zrOMnh-^Kj3=c&?Dm@zq<hodT{BEliAiN6gQ9^i!@7
zvP21_-Z#eQo$>;9lPfIDdknHhd;_F4K-fh$WZu0nRwn=zo2Ks?xuZiI+c-%=CxFju
z#i7<oSlmj;>)vMy8Y(q_G>_RDqTkPJop`iaJ&cR$O?-lk+o6>Lc?ZrtGte)}EOe(T
zhRL*%#kq`!EYg}o^PVL7kN_~(Cdu2JF=;DMp3q#al(_3xZD;}#%^5qvw&>mm=Brn4
zByvQGzH%APSwBQByN$5uUW}{CD&g9O-jrxwY%qUq1&1(7?-<=zbnuAT(1#XD5uK$c
zJ|+O!Ecm1<WyRB8;yEOF=~W4JQB?_^^-Aa{A%JZ$<^-9&>mQ@H7-+r3CPn&rR>x#_
zC^ug49^#G@a+qmk_k*R7&e3@zCOC4g24iPrAND4yBljyHuPM+2N}!Hl$tPR2V-q%q
zS!-7}OaN1u!NQG$f?Cii%<0Umf&ifkTW6M4rN(L@jJO9F_*8-pu($)FAn5K8riW~w
zi<0m|$?qspqgk#5dO$Lu{yR!MugHB3Wdftcf#y-~K4EJ}oGMkFP2p9+Le5wl6qGab
zT(c;4j^HM+wvM?rpE1RNyuN7ek$(Zc9Dfz<@l}W)ylp>=k9Mh9i2<ReY_%14XO>el
zc+r9ygE}aBslr|R#4Zz2eTW_bWL9}#?TYD|RqCNQb-`#gc&a55XzNWo5!^m09p$q3
zxC^;qvk(={a`(`fZ1$YOMw&R=d2n5<;f@d0U5LgOD8CW79*#L*Y8YS5;*_v>pd;N_
z_qX<EU4PcA2x-d}g(B;2VyQMakBFbQu^y%Ro66u<+_;q^2LKa#_ZHhP_6CJ>cVyOP
zC4W9ErlT+wGtYjooF`q6o#SB>KNu%PduVdR>w5ZFqkJ<PxJqKX?43#4qf=Pyq`xH6
z!Vc^}J3`)lh^Q8llNHu@^k{g@0gNBARA&Bp*YvsqP?b<-n__pa%#{xWayx%2!KV{b
zb&{ZkEID>f71!1s)JdLKjWs#omByOx;R6qbTM%HUuD2PYm`Ygb23AczeSGj$yJiBE
zuGo0k+D!3SH?;^S!A!k>WYU(#r6Cy_T<(B4%11>|C<&VLhfTvOE43V3XqRfeMoIpI
zQ6LiGoczq3Nl7~i8*G`My~;y8!AmP%O_K98+3b|BCG>7CjT<A|awipqa?;I9XR)>P
zL@M3i_1@hbMj>N}w6&4dqZ@n;js0cu_ixJ^nfiKbvIw+7#)=}vw{z}MgGS~}wFh;&
z72{Qh3+5Z;02uY}Eb4GulA2j@wU5L}-dk%Ko~QrjR^4kCe?2jvi8%~HbNjV&kO!~m
zF~b!=e;q;n)~TwiH=Zy_)?@631#^OP<!p(Mk>Rp}k6Z~})zkB>@~V?ZY_M&gCF&71
z*z#SUJBk%<I@Ycc-+&@k<lF@2A==+FC*u(Z|CLf!=OYZ*I{9h9MTK^Yb<4wV90J@H
z&71-}M_((Q6kOyyEO+b!bdGyfCki*Z2;I6DYoCguMnqMnza%;(l|klC;7(G3RzJp}
zr6-=PGTsFpjxnD@Qz$*$??67=mIu^~e(OWV)G4R5_?uNDX`pqQC|IwIDc|q*%w6r@
zA&RK*xC+m@XZ~ecIck-tVI$?%VkgDmW6f^Qa^XLPtT;@$71&c^a!vzbG`VAcjrnKu
znx~^@zFa@lAl^5^TZ)u<{v~0skCa}^8*My3AKsy-mqFJ3CsD%sAC_m>nEq#?WJOy#
z4!Z-f_fq`{A0d2%9ndfya~+F-9ore&Ndu$7xjy*nB)%_?DsjZD3EyhQ0u7?Rv6Qk=
z?@L3q`Xc4OY$WvIr)xt)qFDn`1_M?5&z}Hq!NdN72l-9G;0SU<cOTYozp=o_caPEd
z#3ie5_iyg~_MaZh2o7KxUwd@`apLv?t9v*#4_em`STRk@n0<hKc@{Wp;^!oaf8GW-
z5%7M#Kh}b22KjUfw-T2<iR0O%trC(1Cuo$I#<v&6)69>efsStFO><?p@aO>9Iy9)W
zF|PxwXwT`JcmZ|>JWa_9QA^FgTnK{$$q^5VR?29?L|;Z(t5xuZW+X~c77|#`FH5{X
zzP#T4DR@%RR!0SkW@(vxU#|uFF$6HZad41#<9@%5-Tk{OA=h%N`jJzRE=_?`&7dgH
zu4jXh!`#37oWcGBJSRCPwVsbfQ2|V=6P?mHy&OG9JyW7Fw%7PEhtD5ocXT-=v5tut
zeM>n+{vl(at?ohF?D`K1sBb@q`u7EFj`AWShto-AO_uUdM%^-`VyYNOGMk3TXomA|
z;y~x4XKF+Yfk*wZY@x?5w7^ESh+c7S%i@qKP%Mi*42mGdLP?V@EU+o%!hF~28N=~m
z{Da^!C%SF~%Q{%()Kxo#wVaX&bTu6*<%W~Er)GwV6h1K?#wP{3zcp=T>2B+WRQg&)
z-H63=^`<Od0dBzAEujHYa#8)5I8=e7&V%cX%j^h(!agrf6-j!IksfpvT~*{%O0WRL
z+rdPS`{)F9<+<4<PE1r_Fqr{oaQu>8grQt~CK}!<WmI+=F-n1@P|Wgz7%6@jV-c6*
zrynrC3XNZYzEdB713!&{j{syr;Y|hpqBC<6X_EFp$uhlE`W~&e&M8CM9*&(jhbm+J
zp&gAq+LiHLU3UwhEfB`!U6EP{8z_Z6!S?h7BC~Lh92{hmVSWqo)XiquW2s!yJuf(O
zq!X)a-KvNNLZ-xvDm5#kQkET`Fdc!mz6bN6Ig=3oa?nWGkSZtQ^fzbtR&0RngHc98
zgGJfkZtc3cTelv>P1#ImL-QrVP+(KH&1*|3g}wQK!IGuK4CSBAy$H?`bUzwl#KHh}
zAHZJXMT;K80wgj0gRdu=_h5G1?inQNNrzULUg{6B{27oQpx3(e19uZ!sSKditcaEl
zXJS4h$YG@2aSZ3Y)e|(>V-7=+ADyDyrzVcP_(OR$rNxPhzuC?sh<B)36nzPB7pc~x
zr3-{BwGp!8qm5etSzbe!)re&ZcBG4|ULs@i_3DhH`IuTsC85=%YjMxeHJVgZW68oi
zM-JNMDXDt<AMy|$J1g0%HHCP&Cb#ai$h^d2(L^iz6D?oxnH}pMA@+y55qXrkaXkC|
zN)Hyk#f-JMQZu#Mka1r~K&Xr$b^D}$<i}?uN7EQTN@b%(Ivg7IxnxC!5MI61wfAxB
ze5a%QYWA}pWtz!kj(<#cN6}o%!02V2Qq(gR%BJTJg8wWwzkAA(<=@m(eV}RXsU$+|
zF|)1@aOfeQcbok5g(fFY1LQ?V-*(cKQGBr^#2reFobdGCSBIq8R#PCX_D>VFk#b?#
zQ}6l8n~uC5vKdw?-LiCwQ1)W(=P(j@-Lr+~<=4!**RP*>EM<8+2Qxx=iTTO~x7Bf7
zI&-4%l=W5Cn0O{K?Ph4fDqt(D91Y0FIzy*HUM+fwqmsoRzmE7cJj`ENOtsm#(1}}s
zjI*X+JKP|S#`v}Gff^iISw@e-7*_Wpgq|wSl6;l`4#qq)*#^xoJUut&m*X|)(oeDb
z^q$1I`{BK7Sy@n<MmxcZD?k>xi1wF_uIDx7LfO%=snD%L%F4AIE_Tje=((X{5^w-k
zl1reP5lM|OerP*CpLNF6VoHn!FmJ<xX`%LD6cu^I9T)|uSl>{%s}}akt<LDt;FV6f
z)Ld_m<l^-Sx0msjq3o)G-MkWRb4mScCP9%k-A|nvARvft=t)1q<`%t#f)|>|s8AF1
z3#L(oWG058HWDG{8I7}R-LaG(lx1DRbueZ$dG{VP6S^BxaE{Y;mJ^xAD@7VYy^|rk
zU1#ArB8~X&C+cxiG{~Kga3Aw(eZeEXgrT2G3$)fCM8}j7w{B92VFIfFhO<r{EEx?_
z(F`5YZ&ebKLo&c(csrO~iz4-H)y_Ktw`slw{BdLZ(16Z0R8Ld!B+j>V+AfjafYJor
z7|&F1GA#;GpND)WcU2*@0#y3p#8nMluLW}V*D)F6V0^cFRJ9CcC;tJQS&Mz>;SMe_
z+;m)kiG&LCHKZz>Q#3S-nX|V9OiL2ERmtAP@$<U09n1R-2g@%aPkGe&$>~b-N!<J-
ztc>}oaqUbI3Os0@QjX{alsX$Lx~bT~mck$f3>F?co>7>D^cSCTQu@g?^8h?gZlsDy
zSgGvEva?bq=-x1CJqq>?yA|ayZrxmF4be8t?$kWTB{vZ95qmeP-Voy+Pu_mOYxA7@
z3#4H+Zft_gwN-)-Z-*r20!!R-#7w{0-4)&3#L~Jk##GMtQ6={R>fVq^=0VAGP*~>B
zQ&<5+gk;nAuo_u99K*s4E460OR0wspzmoj}!=p}Bd4`4iq@qSuX;7^oxcgUjDgQHt
z{R%;YQmCF8YKsvVqpKh*aZiq#Kj{e9&K^~H1Jc+$TM1qM@`Xg!l_PEAhof<wma0j5
zjH2#uD0W<Xj^t~>T11JpyZ33BGB(u1ilDRU`?(r>_f2pT3PX*?)MJo4t;1S6PG$Xh
z+MIenG?CScGSwLLEgY}3IBcv0diXB5%Qn66!|vJQy_$9S51ge2fPz&(uC!QZ)(1jz
zfrY2X+iTpd|N9Fzgw8X5W<(mi3^+j9zsb1O14}5du)JeVh~X-CV5mp~cHk2!gW`(I
zxA#p427p0}l2J*1ONAnU#aQMgS8|1tHYEG7bpOJ!qUkGh?s23ToJ2lV2rnWkQ(E2w
zFFnz?C(FEFHo?cK!y6w?FG{dh_qMmyn})Ea${(=hM8O2s9gKWdD){pN&RBC?hoB;=
z{>43!b*54UbMtwGe3588LIq7#U(e!&jPQ6A<vqghn=y5pdCEv);xN$*t_X<S0LYR0
z_4dQvUB47leuo9=$D9cqL1Vw@<k$c%scG4Gwa%Sd7UxNq-0fcS@=USV3R{SiaZGvd
zZZ4zokE!`TsVY3NTz>1;Xj?e9&u~pokmnX~^6adtH4O(6lUD*UbN4~3;7HoqD~2%4
z>!*u#y`C!(*aN1`zULCH^vMh>HxttoBH#Kz3bm)Com&zo>aZ74WSSR?#MAP=jFM`6
zMe=pT=6;zh7jxH<PekK|rH9r4HYKFC)HQ#*#gIDDWto07zF#3K@z1%uM&Nu0Ra?6Z
zg4;gN7s8f~{spGmcr#_{`97LyYWVaW;^jDaB`6w&WI}x@=G^@Om${5xzf%3nfHu;l
zy&q(bA-^AFj3Mvo2J<8J^U~7!1Ne|@g7cpw>3_gAF)^^Q{*TVd{>#5~D2C8;N&N<V
zaqbHUL&(O<^gDm(1?2=08|91+mT?-tJyo?Q$tosYZwBX7K+&Gr=+bj-@pxUnzp-<F
ztp9XT>2#jP-Z%O6b0WV59x6M)Xf1IPrmGfud))YS9O!82u(z|EO6@zn-8iy`_aNah
zhYW=3hjLuypgqBzP-N_=vh@@(NWZj0Z`E1i+=AP2Dnhj5Y{U2QeRUsd{-w%Lnf^Pk
zS%ztWI`OCR#U+6Rju=bnjQA9+591EKEdnoT)O+2r)X~#a<&~XdTD)&-fJ=|&mn>qI
zZDGow2?l)LBBK*CIX90cRrYC{Yre>nn0zuEHTnUJDR)>!d30t+4yNgx<C5w~slET^
z$oyP_9u2+XG{2@@q3;rBCUKEh()kd`39$?jrcvC+j~k=^WEMeToU!-Y97lL0jT}m%
zMwSslL#GZGZ0bHfy&Jyz;?-|4V9LC{br<0`GfZ~^WDq9Ye1G?Rc6QlUnE7%ZdXV~J
z<l?Ylz9*tP?<vG<TTiP9PQkdeNikjvKd;$F)|IQ|p`;Wtx<m(Xil$3d%qvBk5yhSS
zSs^`Pnb71bd`n88FFKC-TR6O=km|#U2{N0!bZx_$1Ptsz8lHOKoO<3YDZha{v7kU>
zYd%LgVv8gtMfQ)UAYO7hu_89S=a>ka>E!0yNc=z*?MsG~5DmA&!qH-7HdeET{t_-R
zlEl~1Ox9||`hIh7n&}~kImK<-lbzXDPHdFkB+rzbswzYL+k)cWk5Yd*1r}`5SOfEt
z_J++NgGJ-(m{eslVw3||Uq*ovOb!yED_^mB#tPNI#el0Dc$SvrBhtCb3aubnpc79j
zsr$odoC>PB4N5CCbF!1*3<`m`vH2w?e@$GQ3I!8ROxHA%$g?z(Y)M0tKCKYeVqbl`
zFotm&)Dg%w(_17pr8|Y05N0UktzGC3lR&dHs)(?yIadGxGJ=J$K%r`luC(SB9IC1A
zRK2$x&mu`>QbHQR2&L93ExIaEk=U@O0}@Nq;#~DTr|KqYlG@TM<qMkO;~N=D{Ay!>
zRb$Hz6O;sb`Vn)-3g&)M?IXs_1b4D$t)z4(Bxz(R3`vuNQ*H#qGCP$$^#=^x%kw@<
z$(P<f%W)yu1N)ZTWj&r<1(dZb(y^8so4QPS2C0_(5lm5yaG(QPY3+@rw3P_o$*kHQ
zMFR5Vq<~*4<E_O9SzTBnolymK-&hNh9vC(J62U7w&sswLjk^;l7@CU6LB!D?rc~c+
zg^X$<glM!7yN<abjujtGb?!Tf&r@_(6(OdQ>*%pKI$im!_Gvvf8`#FAeUkWUlNw={
z@yX1MKX(~y8AqioQl)YnAdJPL0A5of8O_50r@P!DAG+v6xYPom_MQG+9567YlG&Vi
z<xVbtm#np>y5#+NegB=`v$vOfe}tPAk_f;droUw4DPnruww(1Io6i+DI+Pm#x^lc|
z5Y%<l-lm+6*_A2YTZzmuoMXdMLG;fsFryLbNwXs0!=ipC=wZ@M_r`!ni2#R}%cBP6
z{*m%a!@KwMjKVv|;!msaxQw%VZQ0y555`*FiPa*a2s}q?j+X`lwLa2eWV6ClO5V{~
z07R~aoeD-aeM}FTMZ+e(GbQ9mc1TfOX^W0_p%W%hrARJtvj4B*LJ^r@xHJiankm5;
z3#4o^M0N<bi=X#sE$?w|=OS<NB80dRZdV87#8as}AIW+u)kP*x-y~i$(SGw*-x^<u
zoJtZ4N-%m^MF_U~x-HRMe&9EM|E<&1&pBJcJiR!0Fpi^UwB2qMo5G^P;(*l^ZFrh}
znIEDR-owJSHpEK3oBSl%r&P$aR<!-n5SuA1B%HTPFDZ>uIg^yD2vI@~(_xo-mio`v
zC9|fhipX_v{cG9o$>W&0_?2rht*nyQ@odZ<VU^ArYlu=Ebfd;n+cFeTXHl<9q3TQc
zCQezg_~Hzb+e9J#om*V|{`Gm_t;+GoTgbhe<IMtm@8U`ahWxrbqf56-qtFaQf)%L^
zb&xZ_*AAM;Aj?{Gs^QG+N`O(UDhstjn$k!k7W((+mQwU<UDn(bXE!e3t3*`SKwq-7
z<*gtbCkE!WvyJ~TKzLZpi}-IUIL<87x$TtUCUC47z1Im~xS`KUBJ&(OiWSta4Isx}
zB8GHNbkkAj{*+)&Jlh_iU$LM9`s4nTC3X|fcd7(`I_=_H%RSp>$s;O9SUP@Il$wzh
zVRNrvXrq770{!C!I4bNQ3|s*A7#Un(=BfoaUv*P&enBa=K+le|mX))f;uod(K0A(s
zjjtwojfwANrTyk7b)Vz@RdL?cc`0ueq+h=_jlUq1l+b_;taLU_wPnI-LgwR{WOEi3
zz|=+?zBT8`sw#TkOe!_vN`uPz&(-KXfEM4m1xnwJ6ZCb;09{rc>MHAM!|#ayk<8hu
z76bgu-dYZbl3yRkv8m7~q1Qy3mbXc1n&vLGv>3%%rQr}Ralv!oWyub8C8!}0p@TJJ
z(sOM&7(J*1VcG7PHr-p2_AIX7X&*RgwvRcrO=l<i7<RyY*vSxASzG8mc`r<YkJJZf
zyex?yV5|&AzTbm*d#tE^9xSDAWAR!!lE2p8pE1aLuDH1dh@8OZZ3yyTiJ_T-pJzIH
zzCA@w@B~yB=fy4?)6-@sbcwTND@+>Y$d6&W4629d`r@(??SQ&SJkfmm`ilr^VYCT?
z`{+;6kKGY@z2nE6*fs9+m_H}#llOm;7aafJF&rZs!~e{YHKHkPzs2zzO<i|~|A#2C
z5Kn+XfN2KgtVYl-ikEgH*X!^1&yW)QR4dO*G%~yQ9Y+Fb;>R0GNmpLZjUJr5efTGH
z#w_LBUcL9{LZHw51GE+xt^0pr$ZoFHU)O?HEsi}C$1G$?;*$D$y5Bv-_e{a+c^**r
zJWgSTS((Ba>PWNJ;-?3bx&?RQ9CcQ~ZC#5IjuJTUJib1!F9p#JZ|4N-B&~83g}Ty`
zLz56EG$=6lXNhVMJJP4L<Nhzk-nqH6u>JOoZQHhO+vvDsCmnTc+qP|Yl5cF=wr!o9
znVP9n^LwW1soF1K|FLVY>%OnG)(7y;LU0`Xv6*#fPuNXPeKk#>(auV14Uq*nSnOll
z1-nrWk3k+wS4vF|l^O%(5aG`u9oZ>x+Z}m`v#bv@voWNlF`O_i=mG17;jkhC=x2QA
zY$JRV(4h1s*`Cn3iGwT&b|c&d0>}4)$4i4VZC+n*#`d?SSSdTaD?FwDS=M-B>BX-$
zdde0rtm=GT_88EKEWcOn?S(piyMH%dj_L%cQ7Jv)Y0vJ+dQnjDpNvBVuST3;mHaFu
zE~u>|`z(x0N%)T8v2yY00MQfk$NSlf4ubc#W6Sg8{F<dYP>ABI7hNp=9^s>yBYN=+
zy&F1uZ`b97;k|qRJl@hrMIucd>z91g8`ik4Xg1jzzc`OOBDN~}$2F4BqcXBFX|?(y
z#TtRl+=}&fcm$1tdnYv*g3H0bF4T#v%FzgOEf+1_HCZPjMraO5jS!h6)gQ*>nL-|a
zjcRZ=z`xRt_>-d#P?kg{jM=JFqJ`@w57{u)1JNom(L^qimk*@&mQ|wIs26KcaT&wG
z-nki#?HX~sy)^`=D`}%G-gOh_z}IXaOb43~_m1`3l%lXbt-%&d09$aI;BVoS+OJ1C
zMHB{qws*Oy(Iw#`5LPkNT*?NigtQ%M+dO^Iebr+WNzcjbgbND7J5sF5t|8@mXwFeG
zD$B4qCqDRRv3zMSY4ViJJoJ+{DoE(dlE<BO56bufClf?fqW!isChsF<`^}hyfl?AC
zBFaP@Ox{f1mpKk{6K6F@ZMcAG4N@L}&OI+zl8)9cHQONW#(<(1dXJN$!cJA!1}sPa
zhH)o^FD4OwZgTL*vjeYLl!@|)$Cc<+l%FEgE_(t$h#Y;p53=fKAlMK<7||72i0xA@
z)IRz$117!DQv64DWBHm&4K%{HWx|h9vz)lZo)z?ot%8``4M8yc(Zu8euB*H~PYKZl
zp53fDkhxLy#C>>VM#;<fhS5x#P;@3t%>wfBaG&A>m~5QeRFbR|?U@^JSa^rEaaoVE
zN|iRMS>J`{F{53sUaqI^(h&&J#AV~h&(2i*BDl^19>?egeqGEX0!GZ(Kd`i(<`*#3
z_dJ){eAiKHorb<DtxAp0=0SE5jAGX%FOHQo5Gx+pDxP?}d%-xIp`)z?mSAz&g0X<p
z8hoFI#%$r;f0^hZdm`ojsntbJos|3+SG`%y(xw(e7Jpu<eo11E>@dYzv#KSYL0yc{
z5-LA_p9c<}*cEL2llk>k>M5Uq?F!B*oXcB>IDCe%epBk=!u&Hr;42A1nLvJ)recRJ
zUDH#rvE6P#oaeA1;d7>uhIQDRGL<}7gqQk@?;&@>_z+J*$thoHf?S!qSp`>f83Vzx
zi+c8vZIG(aJY!2k%O3Ky0dJVs4eJ`<oQ?BwUZcCK#mw0}=vXfv09GzK-vN!lMX`2x
z9Si{*pp5gn1Tp^m!;?*`3hxNNiRB^QTC}Ni%Osvg6R&(%ZGY!09kxh&8Q66~cc<v;
zxeWth6@F0^-*-D0hKcI<a>=GeIsjbHvM_)n8Z0+yR|qbsIPE(N4zixL)K$rH%0skn
zg51p#B{nl&N9IOFoKvKRTquHRvL%>v6W?nsJ3O^hp)tnsOmF_;Uw|(Btnv!Vm<ywd
z9_zf?CkyA-UGWp_-;;=h30A;n1_Ney=|wH)xqnC%-x>{CQ-=4}f*`xT74`B6rS=8z
z5M0$SOF<Upz7X89<`uJn=yGHuxu+km=H*D8X5T;Jzy00`_4<x79d<fFY{2LsoML;?
z-M1yRZz8?TJKQx^4{aeGztFkVmX>K)wvXia^KU_ZJ^VZY|4QBr09JhR8v3DV;mM5)
zQ2e?w=(0E!1B(}=Z?_vLfhFtOTL~gxppQ@sqyIsWSbwzVfA~UvJi`CgdG}IhG-l(+
z7c%|7^Ez>Z_*iT^c8a(o^g|l~Spi&>gnQqDMQzeX$_sOw`{1L(8nG!xmyZ=x_q*5r
zh=jRQwTj)k#W-|M+P|>$1rb5)|9oiv>9lzG9sFA&V)t?U=AHHU_YK#6K36IVVq`ad
z&f{q|8bU4-4q_+mz-hKRNoU}+l+?9M;#T|%n(htV&4+;Q@H0OrpVE!n4(p!R+y3J+
zWllOiH|S{7ZD4%_adn?oUX?Bv2qi`>)q#i}_p+O$Xu+TmFC?fr3?gMS9)RTW>8v!i
z44V>zy^w!jf9`RghNHT|<ttjnT5zB2mNT`Y)K{p>3DFsS!72e%S~_$?#x|?B@fI(E
zF4A@BfckAA1iFUSo#+<tK)GZ$o7KK=`cHvxSEs<|d<GmHC~>G@H(xNWjD=YhVUVGI
z#_qX6|8eOy)na?41aKR(_A&@oNi%L#k{q`nCK1<wMc5|9URG}D`Hzhcvr^Vau2eFT
zA#Jb!L9t25<#7I>9ViC7kg9M`lncWfzsR&t=1n3jjVfcrJo^=P_K*@;Cr9;Qkn2{N
z&WFE?6))#K#zRP`d}d1J3W=$W95Ce$2M}xihl0O!PwpuT4@WPl881$pdeIgDw$|`0
zMVpQmt^6RH@`WH!EnaK?;m9Vm@F+mg)mO)ngwkCChEoi`<ux@2NuL*)$N9%+)R7Kc
zo%8kBZE=eQI-}&5t!xCorVIIJt1`!y<tObn^lV5o8G5*EI}Y`V7n=DJH}UGMpr<7~
zQ#Q6-kXlIfP24;`Uo_shMorUO14J>4oFP7O<-PL1->h)G8nM<a;IS(I;-zcHn#M{|
zQnYAsBL32)^XsEgu;zQkVkU_HJGbtwQcaR!aEcbxo+iT>G?H~!sc@62i`6>(_4}_%
z1j_6tkhZBV5xN)~d~;Eo(Nb1|GdkDRRnIBy2vo^LS#G<9-9)eVXP$UI@trLQg&GfG
zl`4JBB;JG}dC0E_$uQ0UG=a*;U}P_u1-6L+eM1A=fqFE!8Aj`twk<bl;)o`-cs^16
z^dclL?8GSj4dR%`F-M0u(g&3v!~%m@b`9VJS&D>%w<ROhLYAi6M{u{?qGSC)bB_7u
z2}@E2`iR@*QXdH%4RpJ<p6I+=*;rV^y?;m~&}^*|G+JCjZL>tpeL6}tR_INKy|C(p
z36YBv<OVTZM2w!r;#yeO*=ESwrq7yT%V2l{;;ZPeYKMw-JUFf<8be^ENjehP9Nl(J
z_r&zVV_ssKH-n3gUhWPuokM7xXCQE1hplv=qMBl+kCkXVxS{6`v^th4?VkA<d5pFz
z8!9@K;~{!+t7aFqyqiuo>_(2K5s}0U|1ZsJ!&a<oH%c02jN{YJ5falpt(>(<GH#|Q
zlOOH7L?+{TG-mhKHlJe@`EnJU5B<n9d!h=dqk2-qE@$g|k|f~dy4d#G&5Yfro#CC0
zgVu^hYfNsm`o;XgfL+UY0<pu3S3%6q<$BMJJrU8>p3vOXBw1&3-H2cKmsFWI>hqJ_
ztJ$Gso2$cE#vT2Y9$w1Lii3S1lysy-D&$5cE?DUm#xnxAa7U}`D!aLTa>Kd$mHaAu
zvweEQc`{*ogZ0>Rkxk#e#27`6bZ7Pc^c?BcF#x+M?4mFg`XFlN0N$G$aT{7Q_UVwX
zj<Ry&Y<!KmSX&;@#TdvrF(#pdcxyUURUBmRU^+1A7Y$`vzdts{a(=PSW7?L(G!Rpj
zi%W2-R)sqs^z!81rN(};$y#dwCb0;oKR~V3u0mZP*jWWpP8&GEmhCis`1+U*xR-m8
zm{NFO!dEnejHHf38O1lAd}^3WGO%x^%un!`?@>$@cw{kN{5z~ftkj1ga(h1R)t`qY
zp}0=YNLilSsKRKukpk&;3NI^Be){}Q2c7$GYE-X4?i>^)jzs|ys+Ch=u@{R%UNqW3
z)5eg3d(vr+gHI9F&%z&N_wOtKvbD-ABfY)bV^q_Wr=^P4&k|{TNe^m%u73h@WA`Ip
z4!Qye97i{izaq{BwOcy$A)LkY6zm~JFG@U`v2<8q@oLEskH%n3%6GxPgwOU@Des?&
zrGPfza{@v6eEegL5J>A^bY<NA)BUi~@E{!Xs%|nfUwuoKH1e(EiB5q<=pO2%+nav^
zR#lm>n-3k6$b((oYY)x(@eGS^xD%WOLbd<gRRxbX$tKkXajN}olD+iW&q+v|r^@t6
zH~u%SRtk{3Kr5?5oM3+pSr0Al_=_Mx_!FNJO-uK7w4ys9`r}kWywNyI_hFbbg+QJ!
zd5Y=0a;el@G>-{mx#hb(WbsxK*2Ibn5huYk46U=CEA4EJL<ROb)WSZJ3WI3{eB<6F
z>sx-i41p41mQ5A)Kw?6Rz7}Jh)5s@qeEo3INg!9*=SAxP%Hc?nRdOfrm%@i8oydpx
zFGKcWJYs{y!}hm!AIp50p?L(Zx>i~=h+IfveL9*P7n;|qUSCU#U;940LHYxL4(R9x
zB|mTaoE@<+&U3R<-Pgzg6m0d%6T!<7xG&EJP(FgH)|BY|)8M$wJ(JOU5qd$YL&z)J
zw_5!^sB1J`Ox2v-BkeZ)(X{{77XOK5$J@<_jqMoVkpBg#<Yf1hVTv!Mj-v~%>;we9
z4KBnTtw*b5{~`e-gEbvDqzu)(EF*}O<hU{lOT!ATTBq7#s)U{>d>q&)uN1th;%67-
z$XkV?O(ZUAh|a|yD7-VlXEA4B_0bHf;U5}85UlPl4^ny8c+bXE?E$1Qc%A7#_Flp*
z(EH;vaRxWKe9B?NM4En|NpZRDha({z)R7)Wvp0=i^6xyMNBNwNFkaJ<)RAEAzeF*0
zS$!!(r?1H5^ko_#7h;OLf)vvYYUD-6%x}ST*j;ygv*(Ll7Z>2E=tH5U_p;9uf2Q@p
z@H?dO0~f^a>60TVLsW8W_Eo*K8!;tU?fWW4(xktu_Ja+JzDn13H={c|2bwA{NCf=o
zGOV@X1=oKXi&*}{Sj5Wo-_^QKYD?RbwITrCf1D{;(PGNDzuA8Co&r01f^HKydPq5r
z5x;(iD!psy{CxbbKLtVL%0+G7Cg|NC4!f!LinTx0CiW<dAzI{cx?8^8P#!&}|Ce8O
z@F&QOmijEn`0f3Z0sPcI4b76378L!~`B!m{NliD26yfur7Iuyq*jJm+jk2d*zK<i_
zxD#CyP{a6tCz(rkP9&WM)*qLj4RYL!TK+&JP+u~v5s_*KvlfKJlwpawO1X}`@Mc~+
zxA(I|{#6<1B4JQ^N4I1cVD`D!9StX7CR{$XOVw1&lnA3)>y*rU&5*R&TD;EtN$Bnf
zPt|hLrmz8HSpGrdtjk<c3wJ3d0rl64Zz<unAnDj$rH?secgE6}^eOK?tc`D?B=(6A
z<Z4S^re32W^p$0y*TKY@#KSC9I;>*Kt&}dsmoO<Eq>DR#V!z<f3WFG;-*7d_{D-8?
zJUNMCx{>k=tN*8H5rUxSeaeKkH?tz6B8H{*CBgg=hQV4--QCS9P?I4C%>$QKcQghu
z@$bk#_|6rvWi*0<SnycWJi5^l?73+faz@`o#OcEA?&{no2WS1i2R3m{(TJ6l*L0$`
z5;CV`fGs=&-DE`wxOt^V@(Orx&9Q%V>XPpanBL#*%&q84tf&F3rvpVfXiG|~o4=3D
z(`TUrKy``#fJo=_Oxr{=Uoxd>;cc?cnqjCJXSL&s80&L7)Yk@xT>jpvFo&X^D1H2V
z^{8n)?S8xpZV8GFvA<qPPGosr%AU4FNVE}mcS`+-Ri&^|tb5Z2EE%Y$!a+BfAqs#-
zQ!Gv<ou^fn(dzX(XWl0r&Z|04Jy6*XV<0{GR)u606p19;`^S7wX(f83=%h1W-LaMU
z3ST&ne<;8YG1x>X7D%zMjRSR~V#}wizOh7@eX#_d;B!g<QYUb~Pqyv{RIue3Hx`1&
z7B{T~hoj$wcR7+S0~GC1kugn7dB}sW$GKO~1nU5l4w_vjwbYxGD4)7|X=Ik`)S~rJ
zD*}uH&?|K1TB&|dC9%2th63>znri&U>n<SV!{{S}48Y{NNmEE;Eg+*0r%7(fs9BpW
z`hygy7z5@OKGDVj2<;!yFURCE@aHNYP~A{Q8d{8+#ZhIx&yep^zB|J)<P*HNnknw1
z!Z}O)rr1-EOPFLfw+;Uk6p#1*F_)Ji5t{Nr)Pz5SM%5N`$XBp~?lW>`cR_Eczs@<*
z!lSb~uF=jL0N`B(y9Gbl-C1Z1O2fv#zr5}cm1gKy;eeb#7=6G0zJ8txk}jIi@3TKZ
z1=zJM$vk3|2YMD@OyGC|fYGyU&JF2(iyK9K%6$XQT}ZdBt%0KyB@%_0VHr3`{Ya9K
z&2%033`$I-q5U43k%O<|M66&Kk}jT#Z-hfroCx+Di`JsO7-x-D07XeM8wv@PMd-h%
zW>O}lv?trrr<j1lOldbO(~#&ubqb2;D5SSAiN!NFNx0U>>>MC8gm&P$+!(HBeG76D
zAECfFcXRgSl<)}Ld^fT=_7jT|``kG{UTM@JcyiukgR78ROtAO2XjyXNKg*eZ53M^)
z6OWvc57u@Lm+400kFDoal(g}U#bPgHQ$Q9J2)C0B<I|V(yIY`xswF<&tWj87f6hql
zH-Icol4W(tC@Cba^qz2QRKHQ+MATRn?cok$abdsX<V`qvmgZXDqVp{taAVt)qL<vJ
zwW#gT|Dr=PXgF-qz=I7CKTC5?UMnI-u?fVwwq8<57#COEdeRPtYsz9{?oX%N9Pqpl
zMqB0JFP|5Ej1qg#`I0Fdzid<yVIh7X+jabF@CZ>({kJEenjT$Xj+>(G!f_?j{_}jY
z6Y1`$WE(5u#N-zA)AMu45@#S@M%B9746O!xL>f?43hD_;MKKeD);Vo;n6p2g_Zt-Z
zwk^U<1XcVbzm_svnm&Lz*Q_@37_blvr96K68M9VFi8DrRmvFq1LD{zyN@UGkJ}OoX
z9WgRyFHgnXJ@0@V@slA{o?C4ml?FHI-zy)UEs0v#1VNTC0ktg!XrMAU>w`}>*P>o)
zBc<dV#}#zALn{<_l;Emad2K%bMv2Au=6-PKdrhY}=wdH1F@pNFVek~f%+ZM}?bH@^
zzJ}3eogkOxE2{__47LYI0$6jrrdqk>u9)8dZ9{t+6$+EEkpEshm22+GBeZzvt<-M_
z975R2zw$4VS)@$%{0Hgk;6x8m-nw6&S1}p`_nB7%#5D!iM!0|uB#B7<nq6T_$Z5=b
z8f3Mt34|jl(OTgxoj*U~Fu1??MR)VL2&fx~6d)Q5t?;f@%cmICLv2$@h|IOetNcLU
zMdiHx${)*}VHP#Sp#5%6y~w6c(SNGSd<g31ShRPSH_)eIs|Eky^5tFYGmu77s3O%A
ziXhc!%{@La$?<rqSnt(n(pMa8aanw|r|^gE(?zXiWsHRg32Ke$a&S(;;y`8H#LWXT
zyXUCoEzTw%P7Bg;xTtQ3nOhiHD7d^@IGgKc1<tZKOr@}m?lQ1104f<2vz^2_Wy+c>
zyk@tL!x=eiA4NWkauvr(V=+3|HE3WF4;LAU$M)f!E|=0Z(2m&?h{^wJ;2b<O*R)Sb
zq6Bgqr1=mIl_Zxo>H=hi+Ga0{8^+vZqw5*ynWt*ekGG?jcdw72F%e7Zga?V>gEWSg
zfEQ145i6W7qQZxHw@)LF-rj-h_j)~El}t*h!;}Cg#)GA4(>IdMq4)X35>y_>wCjEf
zL?D#L@-X)V{(;p82B96*@q!pM?uoHDUKh$R%Vaz$bt`(3Ai@Bu<eERO1&VDrrQa2X
z)(<(pNttbJhU9VTcX>P$Pg|yUeRPjn#=j-jofpJ1U6P`3I-WXZ{@yOPo-jID!*I^|
z>(&3BYt2%>mpr8}ufMSFfKRHRY{k`a!-trxSA@$-_YLU)NT#VQbaF}zFORBb-7xU3
zKE;PB@TazR5^3>Rn!KFr)_~v%H;WGZRw&b}1Z|8ZNI9bAcm{y@4#rPBHbZ!VO||_?
zwO|XXB80V&R-u7<?YFHcr+wi+uf&=l(Jx@D0)>MAFio+r{vYs_|G$hnGZPcb{~J>0
z`Y%VVMQv+)onhq98UqJ^A8A3i_E>ieVL!ezwqQ`K7T6~CD3O#jv2s!z6#SykNB+ix
zVzNXls#_in5H6gpj@b4eCiK)r&60K~pGCDTRdtY%C@Qo+<me}~4?B!7Bx<l<Nb!Sf
zwTP7f(;?jG8!KG!ZA=^Ymd37|1eSnxljybtK|eNGEI|-~uCYlcRRHjeD6v-A(#({M
zZ%C^`KHgzI#v`eesh_}Yid&hSH=ln|KFg_6@dLpRf}3ne=h2P!<edBHchoU^(;iC`
z9eWWE+esAnOttd4H-3UlnSZP9Ub>0YORiLKQv)_DS%fl}w|ZfAL8#WmxUI#OZ4aAL
zO9VF)`C?6ZlYdx0PC)AhJs2s9b|@LB);Gc=XlhLthEyF68<eysFBUFEfQjwS5WLgR
zUwODQ!dzDrGc7#WSl`*W`AY-h31_VmK}Dl+m551MN&u==fJImApJW<bL3->+wMfAi
zoVPI_$_Lu{mMPXClY4YB24lYE!_CW93bWS}1_~Mt!M`BbrVVm9A+#1pKGbM)Lp<7i
zbhX9}+X&#QU%nEcCDGhjkfagUf3SZ8W(IeswEH0JKrtw6{HP}vW0=U{5h!Cvu~BrJ
z7#lFX*m9D4NeP;I<+ixidULb;65u2EI^PiRbv-+xS}|H;V;!CYxS700cejy{G1?7@
ziAk(jj?*A~^c)cZ5aO~eVG}O}37f`hTynxdLl3{KrPex#jbM(qfQSRb8VVH!v%6lV
z*IzI0@9zg4pFTFvO+P#w|9)K?J`L{eIs4qQfI{C@t{RwaPXN5MGz4=d6;l|IOQ2)M
z6~cFr!<xWK6Xc*HKdhc!{d#k!0s5W4CIC1Bvvql<!5mN!PWG>{xx3w&G#5GO5ZGf}
z*45#=v^Tj$hJ-bvh9+|fR^_!Zc3y)Q_mel2<BfydXm9I{_Iuluf^@0&jNfdT-o7m#
zyPN0E4KJ2Y`yIJk0zBE@eDCKS_PHg5+Y|cTTAexg%k6mPQQj5X@pzfUQi|u)iL2Df
z*GO8?N5hvUf88BebUMF~wR1lCB%cNOsO4gFcK87gCttY{;&Q;@`=u>w39q2<YC%_p
zltOKKxgL_kByC8fR4^%Qy2-jKpm{*os^x$~?H(%HiyplJW|Tyg^<8b($E)UBSWA=v
zD=Oo7pI^J?EyZ9BP=4Cj)cIzmDLVWk-SMufMBi;EP)08u{8i$v{+{OT+V{2kB!Rqu
z%hikP?-xoSX%$IIle~LZ`>oD@#mNQej@bPJBTk-8LChdBzFnrU1_j>QB)&x>e%)Ir
zQ4n7z?@a?2JqOnGp{uEpoRN{o_OXgbqLB#w7pJ$5+}&J+XJ^N`zFIz?UhbV)h^CJ_
zw|idU+yt!%PwmrpSfdqSkEeoMdH7;^-$Ddn6VGP3-W;AD^<3@(27oO7&$HW$!H?9#
zN^61FhnIJ5G#-Pg>C{x=&@#kZ8=rPJR|l4EUFIUUhs%$Li<ihj#!UxuM_M4LPsSNm
zZ7usn?+?epmA3~;MqcyC>60W|`WNjyKA2Bc1AMf4d2nP95M}6bP|ErYG-wSR^1574
zvm(#Fkb^9;J~wz=l^s11LOhO@IG<h|$(7%3%|=_0;J6`GC#C^%^~9h-a=-uxS&BF!
zB54J?^>h$skhh~rsEt+YAx8d@4Skg)pu%RoM9%M(A^m4CJ}3YDNxMDPW_Shg#*L68
zYj_DF<2Gi0!>d`8*Lo!1KO-F>j+{5wHF3lLz}+XF{rRXh<AzCV9I%5uxO%{MW9EWZ
zdIFUg4pVwOg@XGp*Ekp*>4lQZ+!m-f)vO@7-y=fTJoz39JKpy$?=DIkzo!_#Ttd{l
zUJ&e-y1Lct5{Gzzr~K4P@Wh!|pk;)9QkG%SEG+pDor#;IR!Z<!!f|#@+Jh-rl-niz
z9oi)Ny;4sWtI#Odvvv2R???{bq)GPs?5@$)x7BSJU5dKaT^P%@aOT>uDw2j?<e8d$
zFf76sveNbL=&-OgENY8jb0DcgCHI$d!%NyOcJxn*j>^<tT<zs)aK^St7Qzg5Nv+h%
zQBq7!M^z@lKp3IupLIf3ni~-F+6;*3z9g>|a$y3|{Db;vf5^rE3B#k<n6F>RO%-MM
zi{w;Bge&yeEE9w=dtL_jNY&3T#U%qdu4|*!S*eX_G;)o-Tr?}}1dOG*6pFE?QJi1O
z$X)uUF)IBQeNi%E;#WgdlXTZ#lYgUpqJ6m54umi7UqfB?)<!wWUEG!*v|dNA-K^Ye
zwnndwX%`yS%`;|D!Va#dS*oYe$R(2Z!!e6RMOPBXRnn%Q6&~Z4u3Xmwd4(NR5_K7a
zphC_DqOx+3lflJb5K1|#=0Rhlf+K)ULb4Lhv5inu*J!LsvXG-%7`v1y(-{x`4T{Dn
zr**=V6$X92c%_#})i3B#@%H+%ck%L=`FiN#nt=`kT)`wx4NyRA_QM@;5CYeGi0CFi
z#LsLYSD0-QaVm0<K;A?UI|`^e8Ny8reEcc<DqHE~!HHvm7QBg>`r{s*DAne6se>xA
zp2Hy3uP^PN^Fj;q_fH{LeT$-lk?r=LiRSp`;Al)rw9&d@Z3=q}{*z_+MUu?Cx3<j3
z;<oJQyrZXJ>R|Z6rju4yqn%@Vv<^0ho$1TV0H@**$;j})vGLwL^X+Tei<g|20EIv<
z)2;1Up-OLhp#aE-((05~Oj0Of?P3}&O5Tm|32Dzemlyu5WB~K4q+kuVR+a$+V+=4q
z>48q(6q@=d`eIo26>TJ|xycryy+ahZS9b$f>&ksIWQtL7=`ZBWJzX_$aRsJ|kbkb?
zcY6@Pb+pj-r7J5NbZ`_RzuD=e!c;GSf2$0!1ag;9bqd++F^Iv-O|N?zf*&;GqtPYN
zj_8r*6k)(CP!nMWqe_7}@)v_(M!`$3D}yJbRSx~FB0da1rjGHTiqRIj5XG*#u{!4U
z$sAmouy4XkLkmK0hXEi5*AF?chf#Wwk8jR>_l_&~yvpEe7OKg*K!UysRr(_`I5b0^
zC@FhU(KxMuf_td0=7TzytfbFuEOc5vonnXzKhaAG;)(7<{1n>ZA_=pyv>t_p{TL0-
z*#HVZqBj3sy->@!vap@EJ?9mha`3iA{RAmk)=ryWz8+Mv6*;4}j+BwI{6KDJSPhOF
z{MhcS+p7(0igrfTq@1U`qa3G<CCRWqVv5&(1l=~pv!O&!NshJE`%TUzB|XHOr57bd
zp9$&+#0%z^c2$&_y$mgkWF;VVB(@$7)SM4hKMXIl;6e2=?%ac#&O`!F3?}<PLwIT?
z3#pTwDKe6>2KQ*s8x_Dtrr*&*3{RtW(Q5}uB4lj?^;8`#Iw}>pj^4vlQNIE1+iVZ1
z8VSF4iMO$5M+<z);+KjZ&*YDYg3PC1oAgIEWT#s`yzNxjfeRCp7d{5+<A*g0iDbWQ
zcSK$KbjwI=#{kqrOvrDTW0FWbYzSM1Aymp9<XME6G@Fcjcf~Qy1oFw$BM;8k=y-S<
zL$oJU1a=!8c+++|IeOTt3@rMHOHl~pfP`FpOV2ZqtoI_N?81nkv#5gL*ru&yn_S8I
zqAkQT9>^aEDFo^uM_I)gts4GwXaWeigy{*k;+HDsr8!&#oso8M(@a6%$yYmM*4a1<
z!GT?D6oBp;F1BUiU{hYDD)uz@8ZaG~^|sY|XQKD~5&jc@2Ij%#)>>t8^$paQy!Ib=
zxc}#P{r_|>ax!uKS3BInpKL!a=l|Z}J|L!<)mO9ufdaEOpcA!X{p9FiY;wKAzQ5yW
zI96djm1Q+3LQ~i-%r3`ipH{9s%{#Vp^tW>a^yB{hZ~Xe^dMm;~NU~yG!$~_V&=9G!
zz2HB=+-ev78qM|}{Ca!9N6g)1DI%0|4p55h6-br<7|15+bl&-M!S!<<-{I69{M~24
zdBYE7Hd!P=)&eBw%j5lX6vMDuTHp6f^L_cawbVBUViq6gsoYpXl-b4i+ieaP?1wVz
zxQ_c5@c!ET{`4Rz`}en*sp5BU%MUfcS_6P~PENbjxWzhMRm#WpcZ)L6qq3F2yCFrK
z_1_5jTH?xRuiSWSDpV!fD(=973L2|6#(Kg^63byu261wukIAh$Ey<S<x5$&T=;~^^
zZvXr!4%~u3G%UA?@ZLZ~%1!%-^JbCerF1V+2=0aD;IRln9jIwi`;-O%ukJOg64jLZ
zaU)G~YCt@p&5)aAi{4p@_BP{@p%ih;>g(owQoscZ4$Q{A*K~I|vrL5r`ZK_4>PZ}c
z(9y{%KwtH8GV;d;YmSuFRzbO(`6=~{%5&>cUvvBLede9%olVkrHntvDu)(J>*^)!U
z+pj20kNK7qZdiIl^hKzj9c>jeFnH^?EB2kED-7G*!BZ1z75PA>ULYyq@9W?4e`0N*
z;Rjt2)Q(I7o;5Ez7EQ&mq8g$u3YX`ys%X8Iuc8J$pnq8{r#3~CER(%KR4=0VDCb2E
za}dW}u%vPHw>yKl{QhgA=%qR#l2mkFA||RxZJ6h+=8&9m!&C|nhB+V%249*QNkR`b
z#+r?lO9{LSTkFJ(ZGyH8ITXDTDP_d^C$cMx75e0d*#lsmSR3&}2^>hMzKH246BF1%
zNCB}DYr)%1)IYcF?x{2j1=+Sb$asHs#q$q_=mTOY3xy8UC(of|BHq*|1;sMo2exJV
zw#rUaneQ(<9aA`T5~cwQQimNVK-Y{$P#0*j$x=HG*;5)^>a${%w;B*(^A8CcNPt_F
z^m4r7rqq?rzVKiEhmvd02s(>9bCBgkFR+osF8$zc<@#9PNV>lG3pb||$Du<zKl#_P
zn^jYvUiFL&25D=D%;~FJJlOVq<_#HV!2ODA6OH$h1<05^y$M4c5XdjM_`69K$<AS<
z@B~5ReNbdk(OZ8}6VcZ(q*cC81yM3brqQl@IBxR_QWJr<HoU2Ja20eSe-Q_TmmW5E
z&-u!{w%OjVH72d0KJM8`?72*`mqz6$pM{73#e`7lV3hkd<w_&Eo^P8gk>2vBd?!uO
zTP~P?_JFxS;X8WGUxFWhR2VT)z9PO&_*&vQ+64Lp6)1$Tjx!)%O7q*!W8dlb3yl8V
z{-OAqzE2}9S4GSk%1TrSbgD9V<5%x|QXwIrgmnJ8u{ff`5%G0l>qtJwD49RS$E~?$
zEAht+2g~T5PJ;+2w&{D<U8aP5g-&ugOm~~pmOs71S%PX|o#dv}Gib8@+ZO&ymojv2
zns)8zZjH3{;30AzCT;{d6iZwHnLqaN$IYlJ{JowCEUz0U|Hjh6JAv*{&pS_BDx;0-
zf}WFkA~#>EZk}C~cVWW2kx!kQL%U<`a_=4x?De|7w8kx^Nu}A7XsI!2w#t$yP;ISg
z>oibG{)|D8Ko;S{#nC5Y;3{L+aZ#IGUcZ5zyI)RcTCfW(P{eehrQH_GG|bEs?bA-O
zq|oqdv-}8QR1C6OL(QoA#Ylg6nRKOo0V@Jlh8;xG_5%ysG`KQ&fkMuB8&76%f?z&9
zb-Mpd_R+bdi+37=&4=0wp4I?C!(3v-Zk_gp=P1B0FXha=!ATF#5r(X6Qw#LIC!I)|
z6#{WNT2MZ8HqE%9meRvP4m_tH`^Y1}Z-K7|_W=^EJb$yyyB#2-;d(ME??Vp%_?&D!
zXED>BV#WQuD%dY=&D+TL=JD~!;O<yL+^WXPQgzH_sL3##whhUKrh9fc((t-vR?``>
z%+0p0<hk(5&`!<-9{(P^ar|Wf%rrnSP87S8yr}!#I#tw1@AB*VG%XT(-Gg9`VxvrV
z2`F@Fn?w}B`Ba)|;@}Yh{>uWCPdr@<GnB3iM)W+hb!Bd=30@Ir&ul-7svQ?;4+XLw
z*`oNb{czC^S@?rQVe*lBQ^h6Du#Ax5-a)FcaVtFYUT+0!{btpfP0LYP8Da@#7aOK8
z)d5XUzX#Qyu{4LN99puftDF7p_Kp{=4LqxQkbTl74+1{LM>6qfj9O|4APz#(@m4s@
zZxD|~4gCLbvBb*8^*`K8%$!_I|CRD}X-_7sQy~G~sUJXy*_`iH4;_;)!&RGG<tJ^e
z6?H~)9DjJdhPYr*aR>3NID}{tP#ItQ?-qF6AG{;G=eu_vFJALpzjmRjaP)oQdI-tB
zVH0@a;W31;;QP;i>V^w{(2GvHki`obgyX)vcBc>0f&0ipA2*tQ^OTWDM>km2L+xu(
zPHeF%ns)x{PAA8&`Z(BRZa%tkj&JXev4L`9M=cg*(p-^&^;Eyi=59#J)72$Np`BZB
z@m+*lAn|J%<2~avDfP2|v6|RYNBx$U4hAqqxS^#l(7qZ9?nJL!ryw68`HJ8j4$QVb
z5FCKNmZSzB^YQqG|C^BpXB>JoZMwXl>{XdqNf)Wu5^E|HHgxDD2Obw*#tF)SmD0$g
z7icIH$i|OrqT3%iHcbp}c@j=v9w1t?ZRnoQvL+1eWGHb7q`0kbuP#ocNtjM#U1_A4
z2%W8Llw+2WD)nGIcoDa4zJ4V1l)HmhOHX)wp7)$c<05@hn-hHs>KvPm^=^JXi&|hf
zg!J$cZLgUpA0!=&ojR;k!1l!NB3y#y#mYW+9WARZC-EHT;ao~1>1gRKYF0bPFOrVw
z1gE`8YsM<XSUZbUWGS0HNiahF`a&Vg6%d|^yLU{IhEXD4KIM~6_exs=Uiai6At<(q
z#@ac*tr<T&qfJYSCfA94NUOfJooH`}kpVQKJ{8GlOR6bn0Ycz%dxh9ZpLDIWBh4jY
z>yiz2m8aU3pYM(;qJ~VInu=qdT>65WFo71COEyXT%-fDJW^IDs&72{h<=s%?=DVUP
z^s+bhan(!Ecc>NU8nxrbp=GJDZPqBxegR7+so~)+xu?%{_$QJ$LGaZqQ9zgrXReUW
zr@`5`f&nqF_*Y<>wdS;$C@ui#yJm_zoY|vALb8oUi+1-~)V1>%hZ96<C|+$JPWpzD
zRL`b49TC+=5t;er<2`B;TBRay9U2cfO<5<6vI<``0J4;LJxgBp5~0Rgz;k<r<2cd3
zvnkh16o13x@w=xJ?8$k4l!x0E1qt=vv4A)^HH8_#>?~m;Yg$rIZWOa==d*F2>Qas2
zIV}YO%Fei$Pn(*CCE+8k<>U$62W4kXEe$ZBFRqL6DD7XuH0mlRH9N?8oItdr6%yOc
z0=~$b(j7H!=GNd<ELf)gSo0AI`8wD!{NyTK9&cbe3rAo4Uv5U_Ie2Jur^Kkq&**Pn
zj}E(FLT2vsHI`9esZL33)vM{mWp_)R0z9x?jDYJAK-x&%i~C5%lasH=a_zw_k_6)g
zwkjnmgxpX!kTYsG*Hf4KeM7g#Po%Lz`vnCNcR@kAwFRlj`F3v3-*5JB%o&?2`%U!3
zB#xzx!`9MpgbyGRkW8v1h=4@u7*_-MIzCVn%THf5dI_+5hr8{7-Jqvg6mZZGY&nV5
z8VwVtTO>U+ae-rY50zojLWFw$!lLB^edhstaI>d^c4pLZyGBJ`f(*xHPqD=-?|#*{
z&zJM~C_s*aZ>)z{JhwEn5cQmGr6dJd6DWlYOne~@h>+C`cWoiV#k-P~I(4H<9vyTq
z4*V?+b5tj;DM?i3V;<9$ksMkDDWU?su<f1QRy_k7M(@%vs^H34sdywqTGD`?gucZ?
zT3UiOe)!5K<oA&$BKW-Hf}=Sg*rsjw3p9MbamF^G!YSt9-yUN~YZp1)Sxcwg6E8Y@
znr&c34-ld_R}d+qiBgMnS=a~E5hXd}dNEJl^!3Tny3~lAtj*Wymp#LHe<ILYP_USe
zyQD9*@0}{9w{)rel1z-YsdJ-8Sxa?5!#h9E)QT|3JkM7L{K=^<+!{-^sNJd+84)}w
zCi`%Ugl8j7erQU+C!B}JV`eS$Nv^o61}0rjwB1zF3o(wcbjk#dY=w-wCOngEt$KxQ
zux?{`t}`#sUkMdEV;HcjNlqntR*N{g462MmXG)IG#CG*mCQM#)Wcg+Iy9s30B|;!P
zF!{)dt&Gy4Zcf4P{+&{a`?1N#mynXO<~(UDGqIiTdF`Q+uW$5#%eO3*$&uw^U~~$)
zcml#83404_9C~~Gk;$q%F*RAtu#2|SOMrN>(<XClJVj0`7`>t1@$5i@yQL!jHlR6U
za^MK?)H)5q@C}xp9{QGR^pB8=V8w!PL}aJ!MyZYz)?c~v_=aGl^D|Neq;Qb1U-9sn
zaHW_GY~f>;e@tmT8VeoJPjGF>I9z``(kj6h-X(PBPk17I0q)o5z9LvZ^umXh_{^+R
zQO6VbWY(gV?DF&4HsLNj&>bwefyB7R94`eEP^Uf8%j=IeW==#FuG*Es0zH<KP5jWd
z<BHv+*<Fv{DHMvjrJm1-WE_Dav|fJ8Zz;&vzHg?p$Nxz|IavOuf^z-03aZs^N5+XX
z`H21rTJX!{79<GROoA6AOH0X&i{&rpTF?8pf8_J*aF)i56@2ETPu0D8yrLkbK$jrC
zK(4+53T1y42x-YQA?ggMg0KI4o<<_@Z|%N^SL?U)WT*U7o|)hui%V+$H~r?~l~zD8
zcHT)rDNu)b;LdhFx5|KPam=+;26@a6Chz)_tLwV$$Ol|9^^x}S>+54boU)FQa=10l
zDllqkN($W@`V{toxCV+k=cGU#Jzp|DC!I|q5oz9DP+=PVg&iayH8FUSk#>+x1G{TB
zQ=fuw3~EQ8jhAr<VWwdYjBDr;%EI@q4}3qrJv*TmImZC;(QCe_jp*-DQ)*{-7h*TS
zTj29ut%!(BbxoefOIPvx!{ZJS2`bPb5orQgp}SjAFy{+TQ@#TKx|bnvqP?WcnFaHw
zcaD-CEm2!QWBuWdcr0VsP_A}|dK3&={Pq6*m@Uey6SE%5Fnx&&B$(6JB}8v4F2p~6
zA9stN>72G-P-1fwTLFiQ;k1@(Hll-)oMsr?g7>`zx+S%W3Tg^=$*dIduSwztN{qrl
zO*Vc|wprm=2aV#VkM3)8*^1tP!$=r|^HJ1N(KV$s#OT#Hrg_KsQeHwU5TaXfYZ8%*
z9($2od_q_lEJDte{rF45_9UATnCNkXgclGz3Xf^|kT=`fZio4ZK#h)ABtjqWh0d-5
z)bFViY?7S5cjgeD_T7lRs+@v%a7DV3PO~Nz83;QFDy&MXJt-rLg>-bUi>N+5EDXzp
zrUOkC%ZR^Omc!z3;E{>~Tg&&78ucP()-meRgM(tEZ`O|aC^5hj8gLTE$DaK*8>_+Y
zD$$%(`{LYjDTWXW?RajS5y=C?dcUdpBN%18qy|6%G}rOROi6Fd8PWO}{W^a-JMQpJ
z5&HOmHpX13MG@?EGp~2nue9!wI2;*f*<p**{!S{8Ntq?J{ifU+JaxHRMxv?tacH1v
zW>c#r_(Q=rZW4C<((QI8_|7<$I!z;r#OofY{SwhahG$)xAPb80NReS;8F1hbbngmZ
z7Kn8hh;y<<@K`1y{USO1uHiLL&&IV2VW<G*7xF&AhaJ#*1mB1l#t&k2CdFMBz78`o
zix41v`-Qw=K@C~GBR*V}$m+5!DvOMc+T2%Qg?ihZUlK1w;iz?j&_c#Cp-Rn!roC_r
zY$GYUG-$`aSRd9$>%~P=*if20>A|0Ih}SCTe;nk@dgkhkIK~Tas0S5i{fZBqUJ}J)
z@tbs^<`gBKrJ#OahPDYa*l=uwgDydgBR#BW<NA-66Mvf<A<jjg-4zgBH5{vrKpZqK
z$SPnK5RQW?0pReON{W}Lv@B}I&+(oX9E5cyScan07;vPCUddPDO1W;g@+9sxYPeOE
zK=*=ErP&h7fvt%6wH1cNiClP9`B>F)v$^LWOd82?SdN{v*FsHNIZRn)F*H|4@0AIb
z+-1FO@u}pswZua8o4<wpPaIvXPZHnmp%|O}Pr0#qTEEI=*v4UxR(z-BxI25qT;I~j
zzv+LMB;hsbM5#_IVl9If+9s$T-W*H7duX4gtPQq9L-|(yRyxA8os2Y+2hk8_@jTa+
zD#6yYz=T9dlfPj6z$LOwI?%#ac}g!by6;CMX%JIX-53yy8lq$xjP~s11u5Fw{37SL
zsO06wqG<*4SdRvGWrqw8{b!O^b7rye%o+azM8=8Kz>w$N<8zWr7vQQqi}>Er2~~Rr
zYeuz6l}5|I!m}g(7OBf`wdCRHp>F7yosDmn>iP<2L)L>QzI<v>R#g${SDuHYr)@mX
za8bY&J%>~~jFQcs9(JWe1k!&#WdbPCvHjD7P~~W}%Fjt0WKafL9urSUVR{tWh^`?I
zNl_e%;VLXg(GPo!JQ-5g#{0&gDwV{Y6r7{%aIfr{Dc$a$hER^uUwE^a4JXid@8l*o
zy3O~`kryH23c|9;pO`VpP20$jkuxhs(=xM?B{Z}f8UyYjCdQcAnnP5Mv&v5(=j2p%
zt(imL*{}an`~JyK%Irv*q4$*|ZmiFMxW=x*T88bloA+;vDUIcR+9dT{(~Ixa-91|0
zB4Q8I(tRi)n9wf;aN7#44T%THZXgoAiH`ag*n>*KzK#1+T7yugYsm?`>rx}d*q;+O
z@}iy2Zp?WCkr2ovRS2)u;5KMSlc*P}#Mm=FKyqmPTk@xgu5_lvlLsen5F;tn7pT%P
z0o)tc$$B8I%WbO5afh^IhkLYzmhj3A*qQ6#zvH_9li1tmnscW00==RA#e)Qy1VcDN
z6N*kaG8dXWUH4%4gjw$lbb*t4eCRIS;Y_i=OPe|5p8kZ#LVvM0Rvxd5Xn`^6tu-p;
zd%*=<uft}>zfpv4U%~VDHLh5Y8^@nLdH9n5w3FCqx}wEK=FVQ>X7MwFT-E+XmfP!J
zj#CbA7b(oYyh9Ovg!BN0Y#(%zR$JHfBn0?!v59)VWA%jgV)T=p+khnQE$WdkNPfk(
zsQ;wWod45~%KYCJrTxT4W7j!;&bdEC+n`;z?Vdir<qn1!oxm=1e)J!x4%_$l47MYc
z@<RXL&P@UnX*BNZVM%SoZAku)zivLfUH~j!Fm!w{ye$B{5Fy)VK#A~m2vA<9|L*1R
z&2i$=$*D)lxRouOo_`MM=H_jpzZ7fFthfPiOl|<UzNiyiO7l=Edx+U?`jyZ#YV7=5
zQI{D^I70#S*UR$yH6(qXRWhI8UA3>LH}8iNmbheLmSW(%w7Yp|w=dpHAb6>l6ew7A
zROs&U^6vU>Y|h1q+<eB}{=?1v`%JVK@;SNu()<=LXCr~fp)Ou$<QS4+eUA6+TD;kl
z2&Sd|r~(u#y`TX}m5Q@Ib|MDas0PA#@hGFo(ud9HkOtSO({5fCc1dp4oK29L_K1C8
z1BII1+u_TowHJ-ggjX0qHaOMkBbgszo^QwfkzQ6td0%8Uw|Me7H3ekaWc;RzFENMy
zwsE>Zb!3)3HrikYKHD*l!x=UjjxPP%#b+p<w|}Qs<(J0aEFdjY!Qdd+ML&&+OfAt@
zBjk&Tmfr-a(zxOS68R#JqWidMbn)9-bY%hmTy)f-DKm{!8M&^THw<Jx>wX!LPs1Ma
zFXK$bMe4>|kM{F)1dZvio>k#Wx`nr(#zs#`86hv}#@6FmiN%msa>rNS)J?=!{An_-
z`e)tF_6BKUhQ(TW<`yr@BGVDdI@Vqk@H`$RvYT!!eb6kwz;@Td5oN7B{NPx;E>Cnu
z)(*S4_8fL;tEL>w9#_b9U&Aro(wTy|@aK;*7WazNj;Cc@Q9!tS!a<RnF0g3z#!!85
z<gZp}JJJ_K%+4WHmkSdOm<MT)fX$nKdaRj7EwJPi$fV`M{X?%Or&p+pLnLQwsf@-Q
z`hAqWL>v<;KV(@R>7f;1KR9X*;B=P|0S2lT@nIuP446=PEn3qEOkNa;UK(9O?jztU
zlNNv4@`U_YC$m!6QYbL9%$O-YdHp5;BZ%sKHzraB8hX(JsEAl{kA839Cckkvd=bU0
zb+bCh#%UQn8YPU5MD8Oc<dIF$WWGuAa4_0U!-=E|I0^<(pT{N)pEOQyl0v)BI|^>5
zh0|nIrm`~s0rX~H)B10L@Q65QbW{pzUGr1@kBbt8lJY#Q>35{QgtT!H8<-5<BYqhM
zM8KRaZ+p~BfcE{{tU=?eYSa8lKyEsW8oek-l*i49d$#As0P=8BJo|Np5ktV0#btz#
z;ha4qA~z-<9UTMJZ|mg6vKhJ!6UrsH>;@_i7+lKCf!Pm}3I|@IbH~8jmnH(WUz~Ea
zlB!8&#_svbB4iQFo9$dX+KZLM!?=|YsrLR{mY$j|jGjZHECk4B+zmPD#=$UCu08S+
zU?(QfX23(NRsjbwd%RLQh98zii(q=#mYllA5giaggJXjy5N-U^)d9z**UjbebP?p5
zRW^OIH&5O4ArebB+6+-Dx*Yw9fjl8weVPGlmVMHe^<y%(?Pkm~>2D-8bjiy6M<X?Q
zvPW5i->TsiU(=+E4BRcn8;@E`_OaifV#Z}%%E8`}Vxn&cA?@>2tfn>`g{_4+QDr^*
zU3F%u`xs5cZ5bXhgh94(ebM|c%qtN1R8ZYJ4Xy*6O|+3EsV11)V5TTfv;t_`MD4*(
z#r%MHkUyjV=DJY(j4wv#%(>=NCMbWe$gU_I_}Psdtdr5Zx)}k^XbhM%vwjR12^9Fc
zHR`A0wy1$4oBX#MmAsy+*W}gZb+<un>tK8oa*3c*B76MML4r`*Bx)-jD1PP3wL(?5
z<-&Q|&A4n+c}AuBJ)aLSeZ%+U*IW>8+>v7t?IAFR9rP?2bhtml>UsM)d_#G6XwzNY
zG`>`h$*4o#*nI9jyM=%-c4*fOx~WhNTgh%xsC9tgUH~rQ5T-D0Lcl59$!=b;LTG+z
zM&VP5BXdE!$6qJVw<OV}L;s`6Hj4ua+`7rYjY<G@+wPiKILM4r`+1d`1G&UWoACgy
zEio%O%b1RJ`jMx3I(!pM(j!xL^Cw~(ijB%k4y38vN*i{s2SBU!aD`2m`wswFJ%FO^
z@Iqb99dX!Ojr;gTxGbTdkMdnOj|D5a@AHeM*@W*CJe|@PQK@7TmY!BAYqSxP?lsLr
z02kM?f<lXg7Z?Sz#245-wiA}_o!9-`Ky=XZ>_17>4*~Z78z^PwWMTfVu4I3fk#?j%
z%gA@yH^`8F?KWUg;1&v)I30+e1|{(HtfRQsZ_x-H&P=sdjPa>_`E+><YvZnpgolUe
zqEdrgef?a4+%U@0PT<h_^JnOT;FRt>)47aMU@BRruCKJOmjZTecl`sy>hJ@HyVY?E
zgj}f3Dj?F>PH-+SM>vPrD&h&x(IbAc${Q-ukr%vUQ{Nd4Nf=>*eCVs2<8yn(e@whi
zAw^N}i7r(^8i#URzf<B(OXE@H@Ft0GO^6(i0SBBPD4Oa%@PjTaq#Gl7T5iYlA4@6x
zS(@0K-Ow(!0@nFu1}xh!(a&6#gLiH$Y@cY-L#y`P5$JqbjUX&Eu}mWskrJ`W(^`!X
zg;}t}(i9Ai!a0S)OQLK_X#NtyI*0b{;Wnzoa6>c0TDyH1p<0%KZ?7~<sEj%?MhT7X
zN-DDH^-7Q7nR1t%Vm8?zIA!6y!+L?JIF)b=A^wgf<L14~{ge`K3G$04#MPdHm~uQZ
z=a+n{#m|pCYHN>uC-XHa%vj{C**&CA;;W?Ao-(O_@{ru}zOs^0%GqF;AqQAFHxGYC
zQZR6}ols-#$}T#e;pz9B014>$n3a@5B_)(uGs*fzyvKrxH5t42Iw%l{DforlZ5)Bs
za}Hj87+9%)g7qijw8WzNT$ILL$+HmAS~ZRKPbQWB#FZr8lNF5~Bk`bk0OS^{FU$y^
zzTH0w9p1EQFcmWYms&!_Tkt^TrC{hP7?aXmU&SB3>veexQ6|{T5>6wAknXTsK0y++
zt8N7phXbH2<+15QnnrKZ%UWnbG~oHl<R!ysNoxh(2y6H=25O7DB&mL&zRVdW63#M4
z7LVt!nN9K_2umx(iDpAirNV`-a_4liyhrc%cHb?YZF|-Q&Z+$F0Q>GA(xtV$528hK
z+Z^$Df>qR=X}^Ey(SZ%WJs@fzT^h&F1ky-pIsX@9?-V3Tv_%P)ZQHhO+qP}nwr$(E
zW!t`G+jiC5?wIbGn0^!S@-rjyGtb&<p9PI&oqijX{18E7?83hd%2yWP&94|BFHjC8
z4+xh!EU-z+fgSw3hd!Z+5DAp^+5cJ{B0C{JQ;{`6t+Cwp>mTOOb;1M)v-0R|>8^|b
z8Q9O1(hmZfa0=$eMD{(>JBg23wkDI4yT}a|tvO5+KDmuUP+0!l^9ZK(2g4~{49OZf
z=fh=m5AN9=MPjPV2gloQlXv(IEdHi#xSaY6SS{tg#Z|@vs@R-`EuTD6p7jLf^{D2h
z>kQHU$t@C`@$HC2vPE67YBs9gj?Z{DxfD$+O}%5il{=H{NbtpG$K3oHkq#t?JM<3A
zlXnIfc3@Dk6r?2WIy5id&5&d&L`u10mJ~rP0oa%!`Mi&sX4>&bersHnqlK#wq&xu%
zrOCIG6G7~0O7j+TM_PagBoM6JY|M^;^XX;WH=z~W)X%mjB4IfJ7vn3>8P1Q3t)muU
z$H)c~HV57;?HIoKz&Ouo+Hm<&y)Fl^U5j*7c0zY01&A4$VVpHcH&4r)+^7tBta*E>
zrDjs{lcG{qJ5#z6zR%>iNbV-p%73L8V#E%l`GOeIHLbXqYosqC^{r0p)aV`le5~sy
zFD)y1ItlC-Px?n8q#L9Ho&QKr+%zhBB@30Xt^wv2WZuQ02j1)~vD4+QBBATLe}ETa
zfvG(y7LakNtp07+Y-*8H8&~MuZx7b^N(!-^`->M*h#_5a>E=TbQP?V#@^DfosV<!c
zH*OHnW}HN<wHvDEdZjzTa*{W>=;C1Puya-7A+@eXekpd)q*-iP9VGG+Hz~#&f@A@s
zTe&v1-rNXPNbY4mg#|=GF1M(7cqrBC_s&(u)I(a8@{M)Bv1s1`7cUV+Lv>hiYDn>W
z6W@ZqghLr$i`h~jAOHC?SCGRn=`kwwI*~=Z<9d)30j*^3KHY7M6d*vGz()WHa7zNI
zAt=7tIHg{3?zNw!p$GY6U`+pj^s?)7w(7e&NmvX2O)w6gG*9(DI%dG9kdQ6#ifK3S
z!P#8PIyQA}Kp&Ro({XwR^H*Fv{aUMBJHN8dtak>Edpx#ic0ON_8)+JSuNyB>kg+rG
zAR7r{UTH?*4_6?;4Kf7N3TO5Qjhe&pOkL_B)g~F3WqiuI>#L{jimPptVfMkK`-oa4
z##G(O8p=8J*U$9m>B+_tO_y{vt`;P>?cU;3F0~o6wY}Z-lsR4O)ZqvBeVBIg<>}Wh
z7aDE%T1fS5iyoRMurJytFy-~>%fAY|&10@rmu;T9ZM7~bc%e1T-H6&9YV!tur(JG0
zYf=~|Q5o@u(4{xbi+D(qpDP_;a}s#X4~0a5@AptTJzuQ8K)T|P)n9=9k-VJ$0D~-y
zjQ=MLa<KiEF!(=$e*SO9ma*?(_686G+8x&}U=Qm<?M0Q9<7DykTR2osBDJwH8s$Ad
zSvJ0Zm;I^$zYE6u_2S^?{pKQwmGb~I3uCVjo&c}qCpcAbJOD@$&Cu<O|Kqc`q2W3E
zkgyu@!12C0l`S6+p=Sm#MXV1f7avDZm&j@Hpw;3DJ4y3MWS{WrjFbK>sEtbr;_2ff
z=il@FQvltp1&aL}Q~%Sq)*kMUk97%#y2;RBEV=>f@9!Uh522xneq>To^D-3gA65_V
z|75ahrnP#6zdH1<Erb1`v;e)OpuE;Juh6{4=q0#S(C{l|p?FTKIVXnhVf<Xufr)|p
zxI8*sHOdCn=!_C7CJR&>NrqB!cKK*;=q>qgek{aiY*RSNd0f*1bMMAg0V>yp`C$+|
zCIat*lavQ7&%Y*avb-ad!Y*KWS|;>|0N~-+#k%H1dTxgNTdSi6oCmw<!`b0b<i5b1
zvs}#_!lGg8GhMlOHStfu%)om0sg(ldy@p-*ilI}uD3GWaZH2EdxG1A!AV2wqIwA{-
ziU7C3y>Bl2ztlv0NZ8gXu=^4=HL8J*^lWe`3)0XF$Uej5Xx7X+G=^MGBR?_Xs9QpE
z3)zp-3;MaP>D$9pS9H5!qZG#I%keL`VW`4y7Q2kQ&po}ya@TFQ5sUNmpu&n}_vE=|
zY)r9`sLTYArL=$lz~#L{hX!ZB?t?>@Zn8p!+Jq&N5*l(V8c*hwIJVn)VVOFwDR1xp
z)qmtGI$|=)<2o$trsh>&Nf<+UzRx`T1D?hp?PW!!1Hvd(8mE&8B~}SQUy?N2H<!Xl
zafk*n_C!z`K=0mVWm0^!#VZ=1s#zfMSAuNqc22aD%-Ck@tY%u|!4bxOTp<6J4_I<f
zq_5v~C{-CIN}4v7$uFF;xx{l-S-9j(FW>CwNLT-%D<r-sHyX2>GuVE%g5}ButY)Dh
zo8`(UOhX`VAGWolQ-V?IE2lM~QKEO<bs=or)^T>B^8iDjp0`|U>8l$EbydG^h22)=
zi_i-bZRLe+#Ktyie62U;#$I!2WiX;BkwtZH(m%AH?m)p2BU`nI2E7%#8cAOcH|lYm
zta(Dcv?!Kikx?zpAR!0bOwmSZr^0AGlG0H8{T-=V*1<8}!6?T`N5{WN*F;NqrstVl
zvyy+r+4Z@cgd6Zo?0vmQxv~RsDatpQBmnNK4MIu;>J>eL)XI&!|5Vg#Q%yrDPp6Lv
z?c+#X7;{Mk@ouGrV$&eX%kThntzc6nlnOOvDI>95(xeWIc2YkJbH(}y@fE~_4Rb0B
zdEPm<i#fO=#aIntL(A%r-GO5A4+pA;>9a-=nTx#S$g!==dSVBaUlDHEmPsW`oX$R@
zQj;wn3%$nadz17?Nsdpnd~ND04=n-GNAJDpd(Ao}DyMYEiR`kpC(mj`J4VmR2L}ai
z!|tw3jq!5B?6d7za=MYI(_x$ubsOBC<^g}=t$PJUa;CQX!S?lmE|MPQpEvXey9Co-
z?K{5XwU~`Ok{k8GS+HL@FxD@rpB_cJ4BMaAd(4s$_hn_uw)=A;yU(a{A<E8ygXyxl
z5$>~`5te#We>#zopJ<AoB2g^ks7CxnB2@H(!9dtW%pTy)V&+cvtm5Vl{ADcBC#ncP
ze;d;6>yOzNF>zaQ@58Sp+l4&mo`ehE*y*jTs|q*Rk;ZQcD6~D?S_A}sJkQs8<!-Eq
zsJGc$DlJ8AP_M;+Nvl>QsyKmZ55(wKaH3VkEt$!33S5VOA<mBCc*b&fYg5|{6d)O)
z+mc2SKmJuYu^{6Hd;OD~VK5D4YY24{XOBF!v!+XRR#-;&rq_8^UTb#RmJAY9=Uv-;
z6@y?COs(DgWrJaVM~J=iE82?wetSwpCgP&2SKV!-->B)-4%Ulld(l$g+#Y0kpwnCV
zWq%UaRd(u@t*NMZd8z`nf2?y+KtV6>t^@+m%9(Vl4wQ+LCoAl*4U5~fgVa(diW}5G
zM*(lt*pQTX={YP+xHT6gG9$c=TrfQ3$7t!F22!lsikl8ewN(o!Ld8Q<bvVx(i<Vj!
zHoj;t>d!#sZH~vn$+Un^NV~cFlu>KJlQ}b<=a%HToG%|m)rEAPkX=S9B@}WcojI(4
z*67H(MvCDQtvOU;rQK;>uS5(XUgn(WEvac3+N*SKAb9Z-^0Jx&4VKT?B{ZKmRbT<d
zptD5g9kT*zapU?@URUyH!KI40S{5&rsTN5#m+hW6?xfURU0|as2mJ8p^`$}&T$2G&
z?RRmXXo(`*GmAmnxO-}6(cf$Ti_?G@vle+OG_$zjqGF|;j@mqtLn&v<qQ_ZF+_!@b
zJ*g2>4NJpR{W>ReBQbZzi5rWHtzdnbkECskm5RLqVdH5%F|Kb)L&<V;6>@itZ%?W4
zi0B@jQ{E!>(nL^Z-WvO;YeU176fGc^XjQ(dT6|YLH}Evq-c8#02B&^un)^?X{EvkG
zf0-#01LJ=ulbykm{Kri1Zq%Q^r}vdcKp8+Jnpg$w7=8YUWSKZ;2p|6uQxnH)JLxbU
zNjN8Io2HlCv?={)onF429lTy07#cYb6*+qQT{Ka8l5f={G+H`2O2b&KonHDrUdc*Y
zU%4kh%GkpA;@iAD4-b~fCAiS1!geWKJlqP8uosNnIjrnCjE&N{CzU6jQCBYfl2k;Y
zg82Bkp<DYt9tpBGrLtH7Y_v-Ogc>2u5Ewo`BGNdufQo?V6bbu;yXsRId5b4KdL;ng
zZo%d+0+`E{wR#fVdbU#IR)KOetbRt=^snJ3?>!^+9l<DrX&6O4zW9>>B8wbGkqH1x
z;Ji$?DHrRL)2-UvM?V$KPKnbyA7uXx&BC7C#^luDC4IUxc1yZWG3u0It5C%@^0L5?
zkh1ZdQdAofa=u=<KD`~hAHB>ZMkx&2H2z-xVe|q84e=={==$x4!@1+lp?YQwGYu$7
z^9TGCP1>PJ>QBvt;ldU)$D?^Dl9O6SAfJjfk&S*M$<ygdwSN}pyfp+#wb2kvqom{;
z+Z0pK#MuY(ME;s?x4cu;C`e=z+V<}hnhlN&_iR+$7J<%`;ZiTdth(r@DjrHZLzc4a
zV@k<W5WP%MQ4z??h6k_UY4Mt@jFEYtQUkdkH_eeSGEM}wdXEUDLE3X&xOlZuiD=0<
z8sR9gYxx5+jq_(M6S7F-gEoz6%cQw)Ey2@J*f)A>(?PFtH8KnQ#i$FCw2n+Qww9n(
zNOK$fPoozZ{M`JMbqw|Csfq*o9_-EV_QD?}+PqPm$?Eb4+7Ul(|J1E07)++u6xfU;
zkSWKVfEaAK?ydy{_p=}M4d%U#q`E4eQL3UtqY#F(d*UW=DHu3(5@@Ho<5xwnQyNy2
z&Lj}%*t&Lr23q@3D-ul|&d^l1-5oNGF;w;5bpIf=g1qb;rX1ca$%Apf_Q^}sV$+N&
zONMpFLrOJ^$jL97)Jfh&FciIaD%+(^eACQ_>#E<KgVdAI%jHv@&_#8>>U<{gV`OIJ
zjEk8uv$6G$*VdLpWv?ZpZ(F_WZW-nu`347I6}MKZ!oMllG&2sd$I;8xs5$L5AGkXv
z(O?r$uC;)-2SHS=0JYMT3lpros=6rN@H13JrSA8gZ-$R-L)BH$9%D%GVmxcK^jbK1
zUR_$^LJ%;x&|!p9I;Dqxky4~V_O1O~K&e>WW4t1Ct(Sm_T}8erx7t&FN9-$VQ<y8l
zfr3_>nW3iXMWdy!8x!DdpPj7RuTr=`*9#7{?8RIiul%ok{H2pwR}t6cX0Z+<&e1)g
zd@qe*w)KIE4-@WePZ<aF-a;`wj=Q{x0o9}pfn1`S@}OZ*N2R9yohO?%dwA1lp+Ri%
z9+=G#<MQ=o0}ArB{LSV$1Z46C*nqygmtxS`V|~I*QB%6LkQOMNvaoxPrj)S^75vI+
z#KZFA`1sdrt%WreYTYw!dhXnXIhqUGuPLun>mH>2Uz@mlA<wgL4nZ$VpH7Do&9gtl
zy;}kDb92o<kO#lyNf=DdFs#FrIUyaqhr}DP<)*=_n-&3rR-^6DCq0ZdnVgdY5t9j6
z(pIh5#fK%s+loDI6b;mPKI&P*(pw5qQFNqjFR5dV+%4!b5`{wQ$5w3V92?AJ<y~D3
zw=oH%?>wxB#4m><X@Pxk;4HYO+QyekMm{bgHL3)w-#kRCjDv%zkoQkAOMAGthQupR
zDgm#R54i9j9#oNQ0tV)TMu6?<ax9k7f3oF<r}CtVZh-=}F=6xDpIon*LQ~6p3dqXT
zOgsQQtWbU2T7^4XqmtuoM0}64?OUwkwI%tEa!Z{~I<w}iZ_=apCH?r?J>iEy`KIdN
z+eU<3Y?ad7(>M5*)Ya9O<7PEpZRLMoake?A38sOzK$~hoBYvb1?LeFkW8rC`*HxF4
zf6H=deUIvz$8_U_t$DNuttT1l`YT>F(|LNP%>&4}VG89yK9U25BAEbTZMndQGBK5I
zuBls~mx|WoyY#Z=9zp_M(evABjm<3T!qLi|o2IH=FBYW9_q$u#6PrA!dYEfpLz<b!
zB((FmYhBDL&md~U;BLbYn1t!nWSd~9w8x1t_SWs5Xl#u&>UDc+sS*0l%gwO+mpVjv
z+n})ZCNr6Q51mir_-42U9B)QXebD{mO@A;cV>cMaf(OXkFty1m>av~hYFailH&r_W
zjmsy(+9jszDJ=hVQYer(UxL_ZC#&>S-o$yNPtr*R*844O?h-C2_xwopvng3X{+{Rl
zhzz9~9e?!jp+dQVKCe<#u(bB96i=JsHYEApC(plQ?~wRgIT+pdBD-jUst4Jnv)Ib=
zdxxC%6e6vcwqkS|{jr9m$>n*~I18amM`X=)OUI=-$(+g9-iV2q6jL__WSffn&Q4@7
zUuPRwR;E>Og0_Jul^sR-l+Jxd$M}v_*wHsaz3tgxR=oh@x!ne0O?=P_Rs`wE0?%@U
z<E>`N$LlT|9?h&OTE>9ya_&{ScADjEovg0nR@NbOv}4gmUSIJ~4gjc02W5|3nbNHJ
z88Ujl9z5DUkj9pdHY0m99!W3L!TG&30!a!xm9l?#JDEqLqwXnxMpbxilTNi%PtRf+
zwi<NxPi@9dmUL=2$nVr}0ET15HXrTOTwea}!DJl1`U2OjXB(|<eJI(v!<T(N-~SKu
z0`tEjs{cJN{QtSuIXPJWw`rZ3<-eo%D_RS7I_z-1bLtR(({F@bM5Br-L=g#W*|zDF
z2qc>NZ@>f@8_8q4;>IxIqg(X*!XLI8aW|h`M{9Ed?j3GQNs@}-;!7$sl_d(z(gZSf
zC<0qR(?a-{^DXw0f-98v)V?HPIoAltIFB~IAVCe)XGv3(@Rw;QE6n_ZZj}8+3r$@E
zffK#NIl&K3a0|GR)MwTS3QZfu0;MjzV3RBg$bP0_fK4=NdO{0qRc|{AYE@s%=dF);
zp@-nR15$ubSG6S!u3w}eZA#lM2+1GKNke6p8N>)BHu4<>Ds<or1zLtl15JFlf&v=i
z6;YBt{(uT5;R#^GIr5;LLCy7mPS03f85`)>e$#r{t$QmzZBY@XgfWm`*h7=R4BtU3
zRsYa{7G4z(Sngjm4*E#k_)%;?k;W8;og=RWTN=&O)+0@J(w@XvJueVA#$(0MZ1KLL
zb9R8P5D*#|r4eLm?}<Tq*JNdH60aAO^B$mjwR-sP3}*F(T4w#FWrdr(cr@g+xM5_t
zZ#kcnny<2>6slR}sq~RZ+JWxmzh9g0A!#4so=1kBmstR0#fOOcnI?7bD=&S{w9J5d
zgj5A-)1tls2eX$Li<$?OX6;f5QoIJ<*Ua~wmSVWqK`sfcJgp><QFC5PHgJ1r3rSP5
zFDz`F2YwpUfcmB_YkL>KEE5tM+|LvUV$Pswq3w9xB2aCIf<~*{S2LVpD{lRqiRzO@
z(`tU%yQ~ja@57r79|H%EUb%aCyPoXh-Q<4L$LQ6haX+0}>%Cjb-p3z$9cr&XKbneu
zpS@sJVlk@Kh}4LXAdCcEutHjR%-(kt*rgA@MXN^}O1?uygF>ZRXI7UP$^E<}{&Ggn
z>jq<C==5X!`sDwGO<~ERSE1<%KK%Y>SGQ2Lc$SIjeIU0e(w}>|WNqXtmRUlH3VPG8
zmz*goKLC-PVkFj<V;#D?Hu7)u@#NUd48MXMyz^<(nn?!_l<)odKD_8kuU@ypJ3AU!
z8X15x=5Ym(N&BFXZ>)mYb9(@#gxS;SbCe;W0rv5Hvhe8Dx$4fT*MpzS=fTyR;jddo
zdo}Uo`?Q1$hEsbCUP&CPd!f^tjz=4t%JuXo`Cfp3-0oRtkB;Y!L^;>N?i2hZ)iUxL
z=~R>FvnM?_c>}Na=5=8A;?;oWf#L<SpC!LNeushwu6<0+FdpEdrD;T?N5H}6)=)ee
zw8zVt`rW@T`^;V0IRNPRsz=}b{QW0jRSOsQ#}3PXQM--b9nBs2r$IP`4~KU@?SuK{
z`PujU^7QxU>hs{9_NZ$f9`1V1J(~`wuuXS|PH<E|!<}cD@=0?KD?XmRo3iO(_wM57
z7kcdDVQvkd3kcC5hMx`j39D<eAQ2coquPquII7d-{mXwP=OhWA-`F;l(@lu+Z!MCO
z;t<6EoM)6IDA|uq1n@$HR~e}yy^0i+9AHp`&GTVHEkIyI-+`uxqYw;eqK86bHHmHm
zuq)^$O$GyR1VsIWLvqp-2(~;YJ%1N72zD4zjJ|uxS_;_f8dMbCrj=VHiH11G5A<i&
zaD@|P6X?(Cn={%U5q|JLQMo|rF&?DEdXjjcn4wL$K`S`XxpbFC5?)=8rajFZ99p#I
z-Vk>BI~9l5gg<uCMoLO1<(Ag&per3&JFQ*fx8~uG3(8L7Q*M4IAT>oxgCV@4(=Ml>
z=MIsWyaU#zW7eCqvP_`LDY4lU3Wx+g@mncC?4`D+WhW3{yB?LaYE`e}lA{|SE%{vM
z1#!F&AZUJa*1GN}Bv9OakXd^qO5_w$b%@yMlvxm-Lu{X3vSJI=mXl{TIHpnjDx^+U
z$}G~+doTWE^?jD2D!mGbtiE!hD!jE3rMBwVP;cP*2cn88z`f4h8aQ<?d{^IW`acn*
ztopq}C@-O^Fw;FPPJV`zx<kja6fy5bf3@1v$}QBcgra5VY5vxetx#scWX$TrWmdln
zU8sHJEZ1JMxVF||W^Yi~q)-2K7U6%|aKPl}>+q&({>%Ro=*PqLb7$E0)k^LphR431
zwCE6-r-LmO+|x$6yV<IZYQ6T51|m+Q#94Y8(o=!fQ7`B0rbh0A*Kkqt_{Z!1Nvg#K
zja#r;0R&HW9t0Z9bQ+muU4Gppip;dcj7pgWGb3{<KQe?Jw9l>4uAS1AyWI=#Ya=t)
z1~-1shB(Gs8~e45?Z(!2Yf}po6gL!6#1=%c3r57=>mXRltN>K+bNtIKM1i))%BU@b
zE~m``_}H|vg8ogV))(Jd8k{eS_2si6#{i@^_{d0CkLO`u2@qD4gq6zgNSTiD04y6u
z?mh*RQp!!*Wp@@Z0|a2m33|-OxEp#+LWS*~F2aB5BYdn2L6o0(!$A>9M50IEe5_<W
zo>*nERz$~m#jgkByC|8OC|6JiA~9VWP7RSvq+E!e91sdntWDqCuPY_Xu7(2Wj!s>>
zFp86&Omb!haiVjjJ6Kn&k7e^NW20el!<Y7ak;+5nU^BzN->MQRL(u9=Ij8?)a!AUg
z2gK}Khyu|`tc6VS9nBzZ0HwcxP0Ykx<Jt<_=is{+kG?UxBD7umIV1KAy(QZ^#c#Iu
zHEo`Bns14md){puu4RGL9R%%qOZF%=i!6~YFG!~-@~FRDX5neGnrlSP3N}BP<<oDf
zH>FGU*L0kGc=q<qUhe9v9g?^)x#h(9=P;zTzc@i$P^>T}$j&_mc$|z|nafvpGdF#<
zcsn^ub3n1;So?@#Z}BgiN=c3=Wz0q>N{bs{twL@L6zfuss1)+PHbGhJM?o>-Rr;2c
zI+h@&Jrs6pQpbu`)&lLG8(NH)l}I0Y?LS{aHyG>KkEYfViA^qxoizv>NNa9m(aIaQ
zSWAwNLq;IbD%IOi+lX)jJ|F4%RiApt=VhEc1@Su}0ETN%Y?ns@({CYeMlhQ>%-;gW
zJ4nCB5q+@2Q43<J3TRAi3i3;H`+;wD`89lzoHZ?c12dItHqk__nq(G1iiYb3$Sjke
z3H{0Q%3rF-B`)E7RdbEm<xH-$xaU!Q3GqTZe;@{>rF)Qm#`<GwcmO~8g|>dpfhs#}
z>_4mHZmo!5PB6TanTawZ{GEuu;Olq-8vkJ&V`2V3g#}E^od2EIYXz&rj#`{)O??`^
zVq4W0I0zuo%ctM$vtZi6q=wz8;pf+&`qny3SZWBIyCCf9ysqvOMgOPk?)Y=*{O~+I
zIw9InEpXuYFdp;_w7jq1Qn7jjfPvQF{muFPIn;UiiE}~Ri4MdU#<%@RrG5^G<{Cg0
z-wXQg!-nLIxkcQ7huVdQ*aAK89P-KM@zQ0-afTN}96t!szP``*>oStu1lo)uW6^Fw
zj43FUQ5i%P$!yU}VzeFsM>2z=Q#$X^{k%{Ai%x6Jl>9^YC%s(YqUGOxL<o+-I0Si3
z_KS}W?Wt+trF^tUb~x1>f!tXiJbUoVsm!2s50_6Q&N7#SLqPsZmW#6=S<d_Mvu4mc
zi2MU=D*lO?AIjCDRLKeJf#ZWcZ!8EZuV|ngJozHFyWl~}L>I5w1x{-~VG|nqEQxQ8
zg3){Qlab!{eXh&KV3nNt%)&TFu2j~NF;0W+v)z&q%C_AKh^(329tmJy&~2iX5_A(m
zAC(?fu1h3%I#3{a#%Ys!>s=9Af}Y?)-sC<Jro<nLX$7-sv59jcP6@7mO@O*ZD@io6
zgTTyq6L3vvs58&IyHL}*+rx6~j>_!%`TM4_{Ef{kfpGMXx-*2Cz-sx%moxiqG`uny
zQzoywRXh@1hqBmOrnt;bofDrxk$H1P23wOG-n$2pE-xd1>@<By^*sEkxZh&yrk%|1
zZO%OS_X&9MBx@e+xUN7^<((z$#idhVz-x;LcTYM#()ik@+F!yU^-ZlLR6CQ|&h^NK
z>JQl>!Rh@Z+JQ2V>6~;f&zrS7Ubj>!ox<)h1}J4<ckl|hDH3py<=I-`M7Ea#2lUvl
zAi&Rq{O%%V&e7iVv}T^H)M|e!`(YUx9ThTWCpb+&Q)$bk0;(IDlqMw?1}BeH%wdwU
zJV4b5u!WT*dm>I$TLorQQb&5sQ=LQECuVy1iYLy6)xPd3<8C^$G|Wh2UU^02VT9|-
zvzvIJNcU%z^<<Yb-q<FWx-Kk?E|>AqIzg$zIErwFV;T8@B={&R)eL=qY@5NWsW5g-
z!SeKisYgWWEm~5_PrV(D+esjlxWgDM6;_Xj>&1Gl9L_B|N#xA<y4K%iw5Dj6&bVr8
z-KdwP%V{qMgvoOQrVFBU9V1}Xsc@cG8-7&O%I_F7X>}SmDkSB+%Ra|m_t)Cv%YS8U
z5`vD1o*zE*<$pp!Mp_(GIyuQW>HNRV-w)4siT{$Q&2r>M`lG~CZPtECfTkRh_|~;5
z`A!yOD!D19<*lMgcC)voB;ka}<zK+=Yxi)}e_2EkmSz&@^Y1v>`FOP{Gh5g;v|!o?
zfp7o~rvSfc&G?sl(L$Nzi+8~+CWQdGGgu%$PsiXL)F?o@pC?ZH1Kn&(O_@O#V&M@G
z>(?vbmPgQ!Z5xR;zMn(c+mFFZ&QZ=`*n?<{z>_jab{t3S)@>Gt>cl2?)%gS2{rj?u
z#+)=GP>GySbt0=?RU~VOsvV5@9Rw_e0Mjd#(Sem7rtl^Q=NN(4=K#eKYD*EazAd?*
znYQPW7_!?R*BWs857whw7uNo22xCP3Zag86y|v^1>>ZgAA_<ML4O%7gkzoicvN8Kb
zz3~KYvsk-t71sJD$rTzeJ<53&am7FG+#K2DEQ1deTBL+t@jUmsNc~CGx~O^5-_j~o
zd9@`XI_vQ!G7xs!K@4=`>2As5Jd1m_4IaS?-o0C`90ZmE27|!fZB}uZl6SK^!+HC9
z(tAqv9P2?y%8IPM%Dkn$w26`K<sG+vUkV8bP7$u8dwB(Lj6rb12Jn>mNiOsY;7R@j
z-^H&!vP~8y?*UIqIf6``=<x!M-OKCct0VbxNEPuV8>dUIC;ihRyeeeElUrQdNBW%a
z1>c(1VgIt~q$<z|ME%WK<eIB&!)s|1Q4(~G^I1@_t-&&uT6c~3GWu|;<Hi!Ipo7U=
zfGWYM-(7<e;OQUSSEE?qX6Q_aF88<c=(Rar$)Xd^iB6OKZkL+=kS?zUI*Ab|rEE%V
z{dPtIic!jDsD>Tdk;#z2SSQpew2G;x;+|Q?nJ(M3-60IoTAQ9mwO;m12kM}W5nR_8
zZW4t|6n&10(}buLThptyH3z>`Tl%9lOfKlIU={YYH4@L$S?89lb9kw-2V0ahNwpft
z_fi}h``W}1$GYPB<81hzVGUY#c$BF;&Sy(?)?ib4t&$Y$r3O3$4#)&TmzXTij#vNi
zS<0V(+H`8U?3vli?&RYuDKEBO=C$pk!f;Hz441iXjk4z=GN~uy3XxD1WCHn^y=ufD
ztL-FH)vJrn$2MbN<6+8gGXKzhMx~?j=hf=-uOwM3j(zWvGLgpUc8LETEPhYP-iUP$
z*6-u0%-Z|WDbaIPsDjn4G<U0vmfl<jS16Tnh@lgn59V4e%#fx}Jn}%@`nYs8ll)S_
zvAg^+LSg+V%@AgUBoel6g)I0H%Ser@r&~_07)=`t09kE`g_R2~%<+O;Zl~e&)m~|)
z$Xet&J-QTL{gJAW#$borSlRjR$<ghF2#fHm?b8PMzzn2*A%$E`y;iImj9`2NKBEg_
ze@M1n)P0vn$ZB(%gudJOHJYcVaWyZiaJjJ<gTMx{7?hO?1`&{*yPkUxeJRlN`yTfe
zNqOLiF%ueOQsPAHnm+oMofsi*GY7KkZ-JJBps1UzYX`=Omve7kw_B9^N`eh*iG#GA
z#SV8QAf*r-pPW&NsZvxvQ>0+Mnij7I67$Z(vE($sU0^bXN?hV!c$+a81o)g`m<0C?
z#-W>hhDc1it<2$=w{4b4!iwFv<eo8Y9Lf*4g2iY5A0d;!<L@V7Yx}a|yY!;{U0VTc
z(ef8)6m|vL?~cISAKHpnKYg2>e9~o(MTW`Nu58X9-DL0?LngNFf~5>+A$Wyw&Q(p?
zl3>V>z4IR;a)mu{0O9`6JVHGb<B9z%foUuPJO+D=0$Z3KF|Yh~dyd$8j(dy`@A!;&
z0J0@RiHL3?8AKT}uGi!tiNsg`e#Jq!u3tT1c!F}`RNu3H;h$J=JOR+ys;4#o?Qb8x
zdQxN+V%YQUhLDgi%JpvHvb}|0HCCO6W1Zgc*M5~9g)parbjtzUjNx_z_b+Mjd0pD@
zLnZIpOk(h*q9)WL7>1)BjNw?taq<^kr#>r@p=6t<-_?x$r~fpgvHZV+9{*Po4KwF|
zS8p|=ITLrt2Ge^=`2|#rD5=OVfMAf5iE>sO-0{z-(B-`${Q3=FN;45>$}Pu~*8)D$
zMsu%!BreX*kG1{PX3g(5N6r+hoHt<jc;5r)S$K@zBBN!86VL=QGd6cJ*RyJ2+QJmG
zD#Dbr<=yeHAK5<zWq|pBws&^~GtAo*#&AuV{gO~KSga#hNAIZj@;}?HOdw<};Qf8S
zdY3dcYJ+0i@0Iv_e6N1ipNoho+X@ZsQ8hQezMawaF$_#~Er2Mkg&r@pAAMiNWm`{6
zO{G5!x4%9XRG4qqJp`lgTUw`y4+~YNdmFD%A@^-~9;;ZRjZR2}Q`n8jLK2ha$l;q&
zG+hc6lua1R5u<3xSTpkDGlg@)pX8#ah7Yn#V;#?Rb&BI7KT$90E~L~jPwnSbqq#$P
z6o3;I$6Q$+{X%(1brgk?gbzk}9N2Vhn34+0(PEl777Q1x%7>Cy7+W~QE0`7(j?0%!
zSRVQAgq)yo{&jtU3#z|#1Xe*(CDc{|&0Iv=pvc6mfdeZuf4SI*Gn?CZpAu7}MUm26
zCU*&XSaP;uwev`!j<6|vNxs_HyU}`_Ihl!eI9(#RXeBZub8>{+@J!kdF|w`U7$2F;
z56l*4J_aO;7)^_)*-BfCO;2OwJLHqzdH(TR&8<HtVOu9J_e*<lbm5dm-QLJ)Iw<(M
z$#}Y~tJidwV=0isYZs>kPg$1l(U(a?Y`!%nje1?yHnh6ddrhY?_gPBQWH?<tbIVgI
zaZ4$Pu5KDcnS1G~ET3IxqNl8sZ^^IM8jR?cLDsf6=7K%gFv;#<7Y@o@{Z$e<ZfW4I
zXV29bVRTDUeiYOLK{>FVcmN(zDy@<-b9%XfNonRRXV6FKZ~#y_10PRP4joni*95pZ
z>I6tR!r(56x}2f}!g5IT=U~XS+yE7Wbdzx&>1h|jIn=PcwTcL?0Lnm(P}b$<Xii{!
zd}mDB9nbpINpD`WapW5=R%4J(N8#8F=f1F3@gt(?9%a;rg{26t+d)hg%=h`u97i0~
z8^(a#A$y-oddxvbE3V@B+b}pk*3@rf{Gyqclf{+%fn-aGn~K?|A9LQ_TT*#2PA8j=
z<U#|{G%rQ{Sq@oG-LSTJPHKy?_VEyvYi?vZQDh=WD93S>%-+b&CE81mY9jr9d7TEW
zI~n2o{)}fVGI^&mxoGmDv~gVEQjo<@heKI4J0Yb~7&ntsUkRPJE#<eYmhd#}mwV;L
zI#xGHPr+_<N}s|{WL!yz8jgK2oCQb1#}Kqk4aW(WFP~&ie+-NZ(&UVfu2{x4Re+pu
z;mz-@FDpnsgD$&qQU_7sE3qrH*by!};3w(`wz>`dLOn@s(5EHqY|tCaDok`>+q|Pw
zl?AoK^M}3uk6#meeTD8)b~<?t)7k1OD~i4yW-aQ-%2np|f?ri6xu9H7;U{xaBEV;)
zJzFD|1jBnn>B=a|ZBY4j+3xVXH|e>HFs^k}d?bvEQnf+FiD8UC*}{Fr2@8=s3pQIh
z#Yo(=m$MfRU#(kYc?*<}YA}>w6u|CUL7k(Xm6B#=)_t(oGK3{Af0a0>HE2)9C-0x-
z@-n=CCDzJI?t4;u(Y{q0axL|4U-Ta;{RG&2Yvr?lN^V8uhsN5hz1N12ujW40Uq`;F
zK{9J#8%=>P57Ifboi!{B02L&;8LUv=5@PznhmBz5^hufp<8x7p&HRsu(gD-yyL=cK
z^ZdyYa3htuQz|U3Ks7=41NVPx*6^CgpfmSR+_--LCs2cFyHo=-n5n^t6&yPBYOMNu
zfp_)G14&H_;R3O%cdVvxrQw#fbk-fGDw9=wrH6D;X(^+Aff#f(-2c<@&GsK`X3YOy
z{?CVA%+kii)QMiq#?Zx7#MIc{#FUQ@%Gt%q)X)~nV`D^f%I=U2W&4x*33%wP_YD9D
zIFgxeUKe~FRTni^bEfF=H(E~#g;X)khFq#M!BlzCH|LW>z2(}^rS0?j?#s)A1KE8C
z1TcR4@)hnWXhpuE^Q`0%U=-1%p1+G9zgDz-tQpUSHW5#+51e=E9$IjoTi|KF8eoh*
zkC3<c72#zr=`x?N2U^|>jF-RE)t67pJUOHyd@^L|zJ6^Nz74}#*5%xqtF^DU7w?Bx
zWhpV{&)>*|eT1$)f4oIdFveN&o~xLGOTX_WU;eL!sM^UIz9s*NYrpTatW<l@9t%|7
z%banTo)Tf9QyrERy63a|=+>gVdrkP~JXG?K%yjru0CdtQW|79!3I;7j8%@SznvQfc
zj&)8SlSgxEE5B460`bno7&kNZ9&Rn+yv;64!*Du_JdJ0hUbO5wVJ}qqm#~T5!19%h
zYL5cAq@o(vwufu5=~FGm+|<9;$e66nDAI#aPBu7nJ`xlYq))SB^s{9jiKk!HV+xO*
z1Yt_unRPY>5ins3^xcY=qSXptvYruxU}y<fxMnECRi<N2mLz25t*DG?x0nGyKg#}c
zg);*amO4Jj7o+i4n>MCeYQgmV8?jbBA9@aM<SdyVetRTEjxyXpYZ5YxOd53oN4o$A
zaDHCaJU8p=+L8I`1VjrS>dYb1=90Zn6!wh<x~i{~XH1dMbnc+hqQ9lo#%tP4zXJK=
zWFHE|w;y<<BhGpoYDH~a(he%K7}=v*bjQ2ph3T`MOI0($VFmqn9JMLk5?3i@r~&L&
zx|xYalbBX5e7Q?Bz<lTN;)Rjrkj*u>Dh(k#<j<MTr!GS^E=Uczb_lW{y-Zsyr##~|
zDdLh!$85-?-Lx*&B<`;jL1`)NzSgkaC5J@kMp*jmt@3+I7P|_Q>vru-lR0#tm1W;j
zUNDJ3<^^WFuvA0iV8T78-5wLJs%7UeSC0)zPE@|Z;svKf<#Uz$R%|?a9$wX62?D$i
zv4z|y5EmBb&x<+^7Y>)Gir+_0;rQK^QQxL6gC)|(H<;mjYLGe%ugM8h7|}lM)$U!5
zQlI2;5Wu6SQE!W39FkO&Oi&)0v*!>o@vTprK7|-D!BQP62Mm&dkh~=O!Z()y7%jvW
zD`$un1O+1(YLD&~!@|}>xE&6)RP$oU7}~I{us8B1C7KrG^)k2QOuqMw8xHkuLxhdC
zkpyy+TXB$zh%?7N$P1LQSlqMC?ugs6YPiZ}`Ybl2H5f4t#%tt?g<Lk>!}3K!iwG8q
zH~3$8PVM;GRy6&})^}*yO+@88?D9TT+C}2QI+<B$cXlj0m`BVT8K*A}zgU`f$&w+@
zJGpxCx?lR*uUjFw-is@<vpXD?j1}DIx337IyP~H1`(vIZmIY$1Q2Ys1P7_Gi5~`K}
zClSqdqTh<rnB>k<70PS0EFg(7u{Tt(+r^HhydaE>ZCLz~28(2Z^RIDpgcuGVX~j;r
zaGHj3Zzr^SVvk_TFq{1|$k{TD9i|(<qzTr{y~%<C;)q!7P7-m`@y6&^(dU2NJ^mUb
zgWjYnrd4(sr1Y-O=|Qsj=o=!};va6mt!mNCy<AK|pvCr>_S$I6V!(z=s7(1bI=&Po
zzoV8<XM`c@x8;S`HiXPa<!U8wnA@55d}GKLy#QIQ=vSb3D^MHvZjahk*yidM+sL~~
zYcEe?IfHVNdv3*aH|Od-Q*jN^;Gd#u9^+dXYRiy*dWmugBQ1e@9rqAg8D`J+^Q<eq
z0l;g9^Ny|#DdoYNn<mDBou~GOnCY61GilTCtl=?k`p}t^PW%F7J>U}krvsYfKS<aZ
z+5fv*nrp3&I1+Z4nN#Wua5ZYF61ake{^_e2Kyy8re{mh%*tUD$zx~Uar2jTcnReWc
z^XXMl2Q%VT$4|aVOZ4#a_VD@ota54`K-FdK_Q-dj6SoAO3-cy`AqxyW-x+@Vl9e!h
zd2gXgRx|WdySV;~@5x|e?g01E^*}lNnjv>EyTz(rYS~XQCtb_T{y_eGVXl5V(!{`^
z`T8GUW1IExOw81w?EBVlXa5TGc)z|CG}W~d?voScD8iTW`SAupL6Anh(j+v{&QA50
z+QWCfs$A(O;!D_*mga7i+aFk|=cLv0Si&k4^O%ChWk)><he}I$zKBR0o9<HeX?U8l
zP*{Zg{9GHP5@Y5!UQslolQde#sYWC?6_GLCXzUZe%|h4fQ*cFu-BVI+-m2WaDO*N<
z?0D}6ps&c1N<-48ml~zupCi=@+SJdKn<g@O&L9WH9M{Z_FG0EN-BoF5-dnI2x|x+u
zhHWk2LKTBy1f^Tbde$DviG|<hmsZ$RH=r6iccJybyYxg+d+s)1RE!(o_qFnwq%im`
z?pN~)B@o`U#k8au6^L{72E-N~s5E*1(&>ep3|TCleQWY9E`6aLTdSolXdD3^;kXqX
z)NYM*TZ~O*#V`--_o#Sowo*Go#HbRKD>cbk=)V&-BXCg+!r0}*wrQ$IqzJGpPXnUD
zpuwhB*gmLMSfDg*NSIcK+n_`#E3A|4$2N(lP377_^7n>afUw!@&y~nSJA=#qJu*=w
zMxfSb`b>R>>Oquy@71IHm!>4xbHQh9yK#4WWk}Kr&|w^KG-hEo5Y#8`9Dez;om#eu
z*&DKE*Lt3(0G&sHP4E!9j*zhGa?E;?l_dXY<->Fe(X@7|aTyBg9mqmBEybpsVm<F<
zO5R5~h9Q<I2&_6;-Ta=WZy_mEi09_z-!oqjO3G>hp>8{KmWawEq2+v*wP`Qr7)fCq
zBCe_km5V69{?S%a?;KJ)?2u~Df9&8Fl^LWiG(XmzPwiW=LMeB#!Z>u7Dxo~WG4Z-D
zs@IkMf>OR;kS$jijIq6#bb@slgI8gALa!V<!MmnF1u>#_j?et6XziLh3nf?WnAWO8
zBA0QBlWZ-<rM*~u^0i-a>Q0{5P9k2+Sq|rAbKQCXj#-X+8E8W)YueCww3iq%*j1Q-
zwreRZNy!z<hVF-5KVXyS;x0lmSBk4Gm8hQP{Sjm~%XgI^z1GiqC@HCNSTR3rqG;>d
zSRpyC%AKxW1#gGq)E06mvV46f7Y}wD+6>$m**s`oj2NBk=eQc7RKA{vd4PhouH1B_
zL9tCR%DQIWYo9PcFF59>LmbAkjtDWUr2)_}2NX;&73(Y>80CEno_+$hnLSW4L2JVp
zuB-f$kpd6Xp5%oMd|if&jFl8mft8CMwONxFq`hAwoj2mxh78({zt~mvNfy&SDF=6V
zkfRN0(C#^d74I$vpn>EA2}7aoTqJp{ndnR<0N}@|yOCq)TE3<b7Lv{((F-VpJTJ9W
zj8$SssG9N5@Rql~4AnIp0cwp?0DZkKvNhorGaBbc?C;bCRA%OTEyb`c<GDo$HRUVa
zt~|3ycDjISn$#h25in9&jt{K1GHD!DeSAy5^3SD4hxusgG=iIxV5w%qpq}xkT649_
zXduIzdd5q9p??a<okz{?<g^pOvbDuh7k3fhvR8w>e8TNlf?7R%5r1p&MhvoP{;?r0
z@?m(e1v$Dd{BKhq&`#i#RJ-!(#HqR+rF()01ypkwfwMvU=luhZW#J4@@T0kGdFXc0
zNO9~;%`M8uepTlmHhP8_tSC7h&0=KtpsRMv_8`uK`Y4?5B9Ibl(x8j_?oB?iJD>{n
z$<k-cNW<TaA!htmk@^Xx+BJt*#+_nStIUl2Fug^OKF#ORlMd9Gl+@Td8hzdTPZ+;Y
z0)#NQ0K^*rjy))RsBrwFr>t_Z$>*I}e<|?9ek53)@yVfg<Q%>*=2GO*KYGB1oxYVR
zW6hTi7{j`_zff$6^vBp4%f&F48oi7Y4Bix%b!w-!3}Ayv*T<QS{*4xiB@*r9zXnb<
zCm6cV7;(HLoAdcS-tB9KvjR%;hO3A*=N9bTS}1$sQR%|I4lufg2~=_ls#(mTlvW%E
z#5sjF4NYy5_J}d-!k-jGhGKfwJAhr#YU&0z++7D|`KP2?wuQf1>^E}X%*10|cm4%#
zlEnuJjDb5Z*sis^clL#jB^qFB?fe4s06{2vjsXtOZ#tZ);;-9pI5Ipf8@T!U40Q0p
z8vOXIHm;<y_N8o7^Qy*=JRr9mf9?w~yOC$l{m@(Sk<xK0C~@H!z+lUW3gy^&jTB&*
zQ(e%{s_+kAo%#+V_JxbBM5Q!x;&<K6AqYMD6MWnR6+8V|Gpo$Z8suKlhOFarT)-rE
zxw}u6k8xg>D_T~av)lR4p=!VSmmE>%7W_b@11N_GSkkbX3`bRk2j*+jAWcP=<a;XJ
ze+nB)Q4Ru9b7;Xzuhvj(4=m+hm;Z6hMT#w;uHE7?2_(Jk>XHWQfc7!c<Bz1o-=X>I
z2KPlPjdY7o@;dA~Tlv)+w~vj-&p`k7qvTaR^ULzo$?!h<CwBh|pa9k&!Pk-SR1O@a
zeKdnv^~rm*&UjJ?)UenbZ|*pFB$2mIgtgY6%IJ7($LBwY@uLyv9Y22-aPZl!go7sF
zmGu0#DtL#k6(82|u+zG4{xr^aG|*0ZP1q{#G~)8Z;!R2lX?1ydc&-Z)6(9z@(ZdXO
z<=XK<+@AgNL%^@9<_MhzedWUpa^u}_K-{`@63+t~=xX@O7aR&$nBRi9<e-l3`Tyl=
zhuNxhwqx%wJlsOc`#f<65DvA`*tu39d|alJ!SG(U3xRUqSNff>`-E3S+yneivkWK8
z|0A#aU%Ns!*8gsnS^1Y*efY0=_<xYi8A8nwSPKjb?@{2RG_l=()x-bGDNa;bS~1fW
zA_>>D)ud7LtjgPJ^5{;SU0*&mcD@AL;5dkV?DB6}sQ46IB2Ri=#2TwCrJ*o07Jm}2
zPT5*|C<HDADG>fol0SR$1vI&YJ2@}dN9F40SaE>W3GSi=%~lK8DXo8AbM}Q+zw|*+
z8-paoclUKVW>21(dd;##5Ca?QN+F?Ej%jIVfx5WkQGCni%ie<hQIYFC<F7lJT4CQ>
z_8aOlI{(euwU|KbF{31tB;qc*&A;89^-#wq7_htrhWt=<o?&4fx)4@x<=Qay$w?j*
z=G7Mn>NSb3G%4LlS0CL8SkRy$8roD4TS(xzhfUe9VSjD|ocWtf#Qi<>xu&Vz`G!r+
zk~NE#O@#JrcipIqA_XRiX!C_7|5Z<qU+#Z_vJ|)MJfqJqcMq;2rh$%SDhjaq^|k85
z|2<m@Ctb}dpk+P%7XYgGJ{Gh92^cSi@{}H&vI|4f=XIge&E?ovK`3zRI>EYtZ!EYY
zW{V&+hKb@)tFoBdwjd~{h1p}D_fFSkTu_Np!q!mN`EJ?bl@s1-o~U9n_`%JuQK^q?
zH*;#+ZvkUtyzaA<-h@6ODPi+1%~j!DhhUKGY&anft;HyYa638LhO5-9kjuQ9(BHb0
z9q0f_heG<@kTsp-?xKcH4q9Pll5K}=jPv8j*7MgORHZJd317$3I3#S+98(J40*x`t
z7R)^sp}9|+pz~^;rNZdEsIsRCW||HmOcjw6A;U6Pz?Xy$yEvryiw7+I9AV)>Ri_u<
zz0y>0y7L|m`mKm&1P53Do}#yc9+B`+VYO~tbzKX!*7=p#Y_!icI`5ps@eUG<ir$2U
z(#xH@_L6^7lP6+Nr8T+RJ6z6a9;mt`E7O)0Fu$0~%$<J?jYY8OmW(ND84)_8ODX*n
zM2@A|8Xx0HaVRga*T+6D`O)`#awIavYSnY7)tjb~TUPDlw7|k>*=G*qgIB5Wx{#Aq
znoFmBOxw<wZosQ;Bd66Xm!`72&f2%~y9~@(^9fN(mQ?B8qR{1(MjT5Xpdd+wh#lQI
zvynUAQYwyCR$~X%Fj4s{gW1%mzuHMfaa~}SIc%rKsU=>*YN^pT>k>#F`vO@?FE{vN
zlVA<J(t;MywKCJje2%zFYP<>hFJgkHRhtPcE*?a)mP17yU8`bsC?SLqmy9?stPsal
zBP+q|YA{p~LinxAo<#s%YcsGpun2VuJopTl1;kA4duUFqr2~Swao5%OddAT<$7ZMg
z+w`EVjs<T3ufI?`v%T=7Y`uf{&M3*wz~<jca8K;?CVu-vv>$vA?c-G0?$}x`gd9Q;
z!so(aSREM!9*CR)O~ZSE9kv|<U|Vc@4=}?Qb2~P;bZi^1s6U)pe@Nq}3A<;YYiW1S
zFgMaNFr4RDY>P1>_d4kFt)NoiYsYJAV5=edbTf@?#o)~Sozbm@qs&XU4<m7(pR_Zu
zA<P4ByD~fWNt~Ihr(yBU><EGyb#}uuusyflrIaG|aNQpfoQJ?Ou`R5y0)8=1z^EL*
zDmw$W>KqOFbTKO=vViDNcfru);$NCGX;K*wD>8-`Fb+B)^7K{ig-=`UC@#Y>HrIqE
z>M?(ZrG&m(r_;_I2%1f~aXEM_#fxE7zXFGn$;_~W_$@pZ!5}PFGiiBvcpGuqV&rW^
zeKZLK73q#H0aPAr@bsn5PF!S6Ris*5%F3~t)#Nl{wXq*@hy^!yFPR+NXf$FceSbhh
zVQw%vyc});U$j~>0RO7SySdkWxMl*6fNwAs8Q&h;WE~5DV9XNgr%t^s#V)!LDn-4;
z;CFW>fyWp5uEBk&u+^j&Cz1VRN2?acD?d9io=kLnGjzC#sqZ)s%<>dXfL7-qe_z7C
zb%4(?0iD~*n@Rb`fu=zMP2sq9_1*PY?Gj=^n+bFJ_p<h4Qqdy8Ojt`tCW&3%=scG)
z4nEE0vvVYK1ry)1U|0N6G!h-6zQxAj+ffzv!@p?j*s2oOUWA0mlC}%<o=k3!QyCnB
z-}!_3z`2uZNtwLSs|dElW*#?TOGhBjEcu)0^>sJNpfAlvu$Th`oG=+r(QLbJE?EvC
zo@&>1R6-`7p$qLp<JddGf7<glYS3^vKI37Dh4-328<9TssfTK8egW_cgeI!~pD>J-
zf&Kr)Fb;<QZejXAFzkp8rsq}t4P1*jstBz>LBL&p4WNVcOglk2A#(|b>DM2k)ibeM
zHI+IUXKN*mHp-_&f^W0wbM;?h<I?5WQaSX+ER?u`>g#<ga4*3-y#yZ1vL=8bLTCK#
z_ANPO>i%IWJS`~^e%kN9o4r$WkpDr*2h;&)=hTEVhF%M#oEJ(s$K5O0e?WhHArE<W
zrHO+<^5MCD@9(FQrzfTMW#8Dpmd$VR<^B1etEf>-e9U2>EOlQ`U&|c?fgz23!NojS
zw&VXV#@>NR)Tmq1EZeqiTc>Q>b;`DF+qP|-vTfV8T~oKGd%9!hn}{3n{)Kn%wKH=)
z+0#)g*R!lR(#xT*U&JVkte=D^98>J+zo!TkqHZu{asmA_<7dWqX!%~eViYz@Q}FX}
zprsc5`D|R*C;kWTLS@&HgRBmQxf7uL91~Fg?=574%{%SSO+vTunEzl0QprPEql8!|
z|1>|iUjN72nz(a?jnkT)XLJ5|{w<md1{LUsGSn~UR#K6inC59=Xo9npvNp)ZtAq(n
z<jN=C`YE*lxEOuCm3pjF?UXdY5LyN}E+-mDS(0L*mvX1U@HoJm5j4p=Xp~6>U%_el
zJt^BZ@gaQL{=qSchd88iFCyj+Uv%;k4pQwjX<zE&+bRvyX<D3pow6=~hHV`e>4tGR
zk<lr`YM-<Y=r)Vl@)xY2;l#pBk2qqjI`Lf6V+2YB@<ox2nq}guHCKy)Nt)A_Bl<o)
zP^tjE^Xg)(M2}em3wcu(-*mXZMAVoo-2vrbVFofT^K6Qo3I)P+XgmaDA5+_Q9M{#x
zvxYUP@HlMA;2@4OS4Gy9JkM576dkqls6mzOEuG9z+bry#wRYlye4MsK{4rI5-RCFo
z*R1@-)KItOYWb~<1VXtW{)j1Y9v_9r!!OjAFBa_p0?Q#{>4YiM1N|CT7*_$!<1UC~
zEJv)w^jIuu2Wlr9{5TdW)Ssl9IY;O|<5$j6EKuD4nX|jonN3V((~>~*7e<$2x^E$v
zi9q``TX^K12-Ea$@<9h)b$VBWi2~q~T?5OMszje&z*V<R<SZmL{*rb4vD49`YOBg{
zp@IeDbM5&c&w#aabzdJWlpA^0fCuVP=<0mQnE8TwrkH#_5*`-FQ89}HyGxnm$LX52
z_B-Zb(6aIjN7LV(wevV5_E-O_wseGZ>KD<!$8~!S>JpoC61U|7W+u-P&Akr|-QDbh
z0oFxUHgN{T#w6&881eF}`YH2XdQj^LoGV}@M2wV-*DOH-lXA7qL1seN#e(@UeT~<s
z+oL?<v{V;e*-u&S*s(keddw)35u3yh&BuSickz>Io9#Dwtr`s-z4^$95xl$lshCYl
zEwyG=lU*n6<j#}}fgkPok47-_@1_;Jc%)=Moi~m>Ef7vrrzc<=V)e(;7nZ5F`VxdT
z0a+nv>x`26*#Pa;j%Kw@6t2m}Bv*n2<F!X)w9UnWsB*7J$29Fo&6>Db9%pz>FP0We
z0%<a5qgE`gNGC7@B4~b}%MDoqwPOuyLMfbK`TlogEmoYK%?T-4TB6sc6WUjx=Req5
z%L&N%=ed&L)VO$~;FZaEtNM8hllOb^uwV|pW-vTueFDHVBVWRT>gKlug6h={1cX&2
zu@L#~{xA)W?(NpnV9y9tO1ygp!DrLwF%fOSMZoR`r1pjSf^I$sWjW5Q<RG%l(Q(9-
zgr+2Z!)-roq)9+Y)(UHRLE+-6jd<zx#Yfo%HKVJX%{1NXDzcE<q?VK{!$PZiwr^*T
zqIT$GV8I44sx1>IB=MkwslxzjH7#|NIH8U9OBVF0eDg{C$T`bM$D@X7s_a9~;)Y6F
z%A?gDT^&n3+!Z>bYc7u!hUjhv8@hU1G$FH9YCIhs^%sBhq}BSC-?Wyh7RRviUoI@u
zgaOU4C+J7I!{K(6PQmnO2tBlDZL8$;%mza`>EhpTn2TqF)pZ2n8QAHQ=k6Alth91M
ziY}*=_U+Z)VBwizPqz-&S8VehMu9TTPoe|rqR2vAP+H}!Fzr@VQ)rt@wLbrN!s(P1
zM^~D*{2g%2mz29aY8?B}%1(FcFeMA>H|6JYIDg$7MT7j}6$wOi#HzM9X#7;6gWH?H
zP}#u3I-rn)tycdGGYgD$JUH*NiSwjhzPKOkm7SK3l|EB1T{$|nu8}$c=BMv`%l`se
zz2$?0yF~X(vUBLR%?d9b4W+yy)$#DqsastV1F(>t|MwpDuSwkl`TC&LR*Ckz_`BcC
zUvlLNRLSCE@Aes}?a|ya_Rea@<+0L?ZA|e5H8Ie?o_*{fB=&GV@T+{cnH{jbJzch~
zT5&pK)EZU+zm~b)WrE#LDW97UFiW<#UV~S`N@lda*J+eRZm?D!tNqnKBaE}dU3kV5
zm`rMOl^->1q`-w$?jLvFDo=2Cg>RFB<k={yXbp?l6Mbo<FMycn`r3bZ9I`V0had^(
zF9q|z<3t{5O2uye21#ye?QCFYDTqkV0cpTi)3;#%C-YL~e=slQ&yA%TOnKmMT>SBP
za^giI-d_<UVsX0R9bO$;ecL}is~+`(iR(Bzy|Df8#lJzu0z<%{38Dh_jz(@rx{}OZ
z*B^?Dl_UO3?A|edk?=pj6!5xJ^4lV>B%Hm)2@Y3H7cO)|!+VAE@`X5%zA29P51dI|
z`+B?I$e!$vNytYs`qZyNo2(QcMlzdnotj6e71)#gM}d4tg`&pra#p?_nr<+-J>njg
zuKF<<IFfg3=wS^73JL|oEcVVh(klW^=fz)i!{y~`W5Qch8t#(>&s1G?GK`p6+n;>T
zXPt#P`~G>UFOeHjHEZUv*wKs>Y8)xQ3|(#722O%enP}-rIXbfp*Acr2+9i~7vR8!D
z3BBi`iwLVnEbf+0R{-KMCkto1NJq@pC><!sg=4I>q<xPTea`1Drrx}&CwxjM1yD#T
zaMx;8sZYf{%|-bD&31>LE`6|YeSnpu;Erd?npUhqYZEOD%H3lVTE$za6FW!hQkK1o
zOIN@n$>MIbyn~TzvYjl?BPS+7=Mah5Xr^Zgcu7wB7Yy#hTt*Q#fUzIa1rE+(@45Kn
z>Cxe3X0q2WkclD8_p1#Wo+XZ989s=BLl&0*2d{hUYrApEKpAR`<t*^<|3-sOodl$I
zN}0nb{F5RM17zq#s}F#BkA<vkYrHgskolZOV>}TTr8x48L99eDW`4KIgrvqjq`_DC
zG&!_%R^$jXRUg4Tx^M6Qty$liQDx?I^LwInPrXet)(mxmmAlu)U_jrc)Hs>MTLdu3
zZ?j;&cgB6NQBkg(bHL=Jg_(XvkS`#8<QT`mJdhn^G!(Z;<J-5%h}WzIWj+bo^Q=$v
z4xIthWE5ScjqWsd!?NsV5X%>O-yW_BxUIoK#vO%JtxS*@jD-+nR_u#7eDy1p9j;Sf
zM5aB*qlHM~Rcxh57FP$aEmtn)pebv0X?9uC(y2abxwgm{N90Uj?$7s>a1&8&Hzj@M
zPQ-OVSJ!7Ov2mp)J59PpGf`$=B`CO6>?whj<V(q3V`^x^*)caW59VQSaZ;A3m-o-f
zYQWZzr$KiOxR~Q&hfX%@WZ2rPZ<FSEd&n__b1Q3XWq_Um;5sZaH%TQ;Wv;8hWgeju
znkxKY1je=R00XSqFT!kGkdGCEbgHq-t-xuEY7ZM<qNb+A+0BbaW*ByUN?Hfej6obP
zHa-YG!>dBTTTl_E6)06?%n$<PAiFz4zjf?2?%vb5ZO<zj-I-j6Ww%a_w9a#>UA%lU
zuux|4Olns_(j=xEYfth+aum+(IZ*B%Y=PWy=T^DIZ{Ik)nG4g|M`}Yq*AFpxy5;#C
zj&h*%Bf9Tqjw0BLfZin>LQG>rta=NJJE>?d?_@<6O#!Qp2H;-~Quz9)dd67=_$Tlx
z-xr!_{<z06zqVyo&Ih{{g->QB?PJBH#XgipH1xVr{N^_i$|nZdc5|6IzLZUw5i7&T
zV?A21p)`)Q)%!M4i?8%l`NR}%jASOk=heAf7?a~NG=KuBTye+HcxfP3FmPOAU#)XF
z3^zH+Qqvqc*uX*ue>6JAIGbc5*L>*MUB(Ml!qJZH&5{-m)#teetN@(_<)4=}7nxgm
zLnlN=&paCO(GI4O*jex=012v$an#(GIf(>HMCsjipQDYgEy|v1_{1>^Ofk@B@w*$g
zhYCBn!htaTCN|lg*V^zsPM7i2s@7Xlu*H}=JW*X0Y3xKR(23~_+D^GL2VG-7;sKRr
zMhb!~h1?;);RvOhUFzr$1%JuN#tG~g4ZS38%NX{q({}GHRQ2W$x$SIZO*b_;`Cdom
zjKe8_uvtXKgrEm1)C}wy>sY9ugG2Iq!;E$@A*W;Vc+|8`F`0YUfdnRf+oNR+x!Q@8
z)+2NT^}f*4`l1TYRJ_Z3i-;c)Im8}pHs-@3M)i`zam)NA^)$}|)i|P;Z@k(PMwVqb
ztAu5Erg~mW01O-2DK>PI(k7q=tNiCHP+60p^fXXimc#@nUo!fnd<I72X_E-O93f<3
ztC!TQeU|jey_&x2ko=l`)|1)F$HtdbbIyVDmzA(?kjdtErcaAl{uY%ox)U!icj0{F
z)DeS6(M35)s0DxGL5NEW5N#B;B{G-QH4q?|<z)!nti4z6)urqx>QIV{i=1=UU&hx0
zr`mG$`~ElDr*#ILpQyibWz~vbtQ|mn4;{vZ1$eR4P}vM#VJ^a<Q@YQE4PJ1Sv<B9k
z-Ab;%zu<}vR|V`6fT(x@?b57+SE+?ru(o4JK}gJbNExYwIDc2U*CvmJzY<WD=K=fY
zRt0<a-c4KfooR-=ACFGXQ=O-M%4%)u#y?~}bh||YmUYaL5o`_zszd$ZyRab(`p#dI
zHI7|_RviVxN0VEPvZg}$Q2C{VilkeV?kTd|B5R&X2NGc~*y&Bl!OU}tp;I5f9S(Tl
zn};*kQg>*3GOI`)$z;>q{{zfs`lQK03r!by<OQ>bn~?NLp!Qp-<isveP2g3FGrRcq
z$hVLw+<|9OrGlC8+#}rms+e2}hwmL+dj-So47sfwcR^(P`}_A;r>8{PxA{*a0Mq|>
z1Ylxd{_m|=cl|G1%PH`2-6$2EA3<*lEpYqLE$t}f=%hJZ`VYS1Q1!pHNUfIPP*o{y
ziyieOzE6(Rfr+hyjjh(mp+GhN-pqlG7vS5VW4T$L)6-qRq((cp5BJxLvS#~Fp8G&e
zt%=x^8~+`rj#dOWPy~J+XuFOVh=$M=^~4v=)E9MK<LOT@AO2B}o7i=8e}stxIp3Zy
z`%>uo9XbVDg>xRN5FJX=5I<>fMlzn08;TwJPlh40@hq>{BX%Xx&XlSs^2;4MK(=}$
z%FsZs9Jkg{A=5ViuatzuusX!XDjMK2n&6nU&`1c>&Nx&#qu6wozX`L(D1vb5Ji|~@
z`>J!{gGW#EUmiaHK5l9xdeuef@UrG4EnQzvPO_y57-0GEq7)EsZ!dNRUjFfWmWdR9
zajzLBw7N`A2yYKlBB8xm2efCWXeh)6+Z!og9jwIu!rR{G-PqTu^tCR^{K?(PR4Y*<
z#X$z~LzouFMlBmJF=s@*;B{h$R7u&6oP8Q-{uqi7J3{hBy4Ru5i9qG%p=&%vb(Xqb
ztK7_=!x%<eHZ?qzh~>(?$VLVtj39VTzE8fVU^1^Y8f{0%BvYtcWfd5JTRLIzY-JL}
zsMtqzsdy44XoLs<wZEWKQ!7TT>AizxdPw02Y@ipCz0TS-sCVe|viWEBIl+U9K}*!^
zte*EO8cgkUM$^T^6>Vzm%Hf~L8%w;NXU7Odd9QU35CDQdjTko4u#VoHUMS!0gz6LU
z#!+IpF|$!oUL5W23><n2hLP+s*=|zul9wDUB<oJ`N&0SKhpUw7m^aE`P%%qyJ#6$b
z+9Mp(Jw8{039KSmlX`-&Gfgu@sb&vwwz_Lomo9@t3DLc?bPpW6(Tis^2AbL*$i^v8
zGu_!;MLtJmVw(KCI;bsE&(){}fwM6Lpi#@gBt0*Tkrg+oFZtNSy@o7o%+i&4`J%50
z$&s^wV|q(jLO*_p@p{!m5pNhqRcC#@97l4FloN5#@^*pDA*?(-S=QzZL~kq@I+V04
zVlxO1*SP_U=SnsWXDD^8mc)U#3iKjwwu+b>irK1oUwRu>m~inKZWN_`Bx-j+EFr~V
zPXRdEn#T=tDntJwYS6^{2^*;uomb^5H7g{^oxkq-WHu;zQUKp%q(B3rlPMvA@dbd1
zPbSfCb#K|M6pwjYmUUs7s0QbS#KJ|;!>Ex58=)@KDut$<L&0!Bvh#Q;HWF5V_?8oc
zbfW5WIfk1;9K+t0&dnF$KBPxPd5P$&Kvj>G9KL#jbd$VtcicYcFw{RY!nz9Q56dJ;
zWweLrpq(`sP*eAh>w)9&^we_{+a9%37<uI;u0?{jbEEzwYst*isRf7OAnupwIDT=Y
z7>jkJZpYeMWe+}UxVF#6VLxmW=&S=E+ZW@{L(e>9ssm#%sZ2`|V)qTWBm#<4#C4ik
zL47IrTgsFgv-$=c3@s*2m+Xn9kLgS4*s)Zi)S!_h#}%)m!$v^C;zRD;Y<NrfUQ<d7
zm4eCCC609=K7?Rf#x-F+{M*ZpCHr1p{p?)Ph?LH9RUDh`Pn|;gF3dN-i>gs9T0g_c
z?hRiIXjSqs50fHkLo10Xzv#fnm)hud1;OJLws$;(ybJK85iS?^@MfeW=OQCbFr>g8
z;!|<S>cxWNoN+Y(dA3U45uDI!@==r6xL&?`NzTAOO6nsvL6wwzzdBCnIy@JXutbnr
zWfI~mHm_Ms>NYf_0Drs5u3Z9~Lztf=y2?UPt-49UrAOjR+F7l-W!16#>=!XMO%v+I
z2y=xWugtSm$c*cFdWm>hsF7@>$(S+bUlNm}MeniYLj>LXgz{(X`0+7jD-Wm!vSaJg
z^994OxR-qzPgK?3TzNhyQDMcg8CdPT0n`HH>1aHjdL}J&zM?E!EAwBjf4}g#3z>5x
z)e}s4&3n+{9C(dzf59Q;0&Ekdca(0tJ$w9jL2|(fQ8(^D&78FT0_arID*4!H`Q)mD
zSo=P;SJo0+3x(xyhorI<-ZO(;IbUEyeC;x}zXE1>6D|pkRDr!ssm`@5UG<io2y>9g
zaOem_;z)A5?s~cb9Xxv%aal5NH5)(DbQ@xv)B@3!!;)M|fAftdS5Li?K`#Zo(|{~!
zk(Kca8`4xn)<VW|)cR640Q!8rbD{Sgr-kf*@d~%=mbPn1PM(6+z&mo>CUf+S8~^$U
zfaMzO9vbi6JR@1Q=g3$C8F6j#O8ur(4Jqj=2j{vCRM&-3kDS0x@oC-MlUn-`&kkTm
z{*g~r2AB+uEPo045Ol#yu%_A^g0m6{QL7!e#`3U&u+&A!$3}|@{!FCXy3U2mfLaB8
z<KV|s!Kh?ARycel?_WxNrVAF>cOj37#sfS4Fvo+x-tvZBhjfCK<p%P9w>YWju?D5k
z?k??8WhgX{9%tEV4z(q>9=sZya9F2cg7pt!#^vg~oTTA6u(!skwRMD_Ybf`<&&SB`
zKiSi!ZCCcIa8iY4iZ|zwqmmQgsg0Xzf14`G-q@T3YksswXNQ^X)p^U^EAO=fwWu-!
zJWC1geHtxXj`R>Ag?J%Uud%wI*+1>PR)E*0I0T<d9zfy5Q&8$orkdJ~R{wmvus8*g
zx2iy<rf=3He?Q1Og=oxa!0QaisAjwrYE_T866hBSoY-^7er)vA#~m1%O|kmQo$hFB
zeZmVhnxOoL;emyT^?wWV|9|WjMuuOR=YKlSOl)lb%}ruWOIxiG#II{Tdw(AZez%)x
zOMDT1f><t@J%EsGAZu8{JW?)&iZCh>P#a(G85(zqilV=gMYLeIzrG5eiHYfP)>@Ya
zrE~p+WV_)Vm?ODF7h(3|rn%~sa>Yw|HkzPNnG}W}7*_$l50%cewcEKx61mx7OaY0n
z3Rl4)vfY(9(TBj9_(rXIQXQD41up5m1k4RnvmUbM(hej@)TU;!A(I}kB4a8Wn@g2-
zkmAUgHrY{(qxntI!8uXIv24K`jpi&BP_#5RA)Bw-oaAh~fcBnxJy;Vu+X`17lcM%j
zP`#!%>?eV$SJmE1|MtwL-ST#y15TzW7`7yKJlE1;V;XiUhNXI6!n$bhAZsh@UsDt<
zDei+7z_Ho>S~nFgWQ!Xl&G1{e`*gO8zaEZ#&HCIh>-w-xjE1d$`X<Y*8Kq4htOL8{
z2aW~|BZfYVFPR!=Qro5e@)1deHgD9f7jIUjR7<f^E{`&}RM9n7_U<oJ@UUZOrfm~B
zEAN4?HB}#1oA9%n+B{dL729Qk4p>+<s;h}ZE%%0?jqy>o6sSLb1q+b=K3vz6*E=C)
zl_=jkk~^r*i7T9!GRRf0EgUZ1K`?d;5VujEctfF@*?MYjw=kxH%k8#DujN8$i_MIn
zz5gnz0hf#f&^3y*vCnz=h#+^L$-slKvu314G3-O8L~mzF32X<|oE{D!wJuF&UNd!*
zOda%%l2hAI6YEvCEFmNh7F7p&G=>LS6QM=R@%?oEeDkXvd`C`9j8$FqWJ%8qFR&wG
z668Gkp7mVy^aBTRoasf&nc9YNVzg-gh1mzR#-Z}uEs}4`;yKg%)!N^J=%?v_mVE28
zzMWk7+AsB9d_AA4f4(kD*DlZB<-C6m-Y!csTlbvqrG6f!=+)$AhI^iUyWXyiU*8IR
zJsI6MV%@St<HKj}xn&9!E0&r00^w7ci(O~soOluzzCAp+qCN*A|Gu3?VjbY;@_GjF
zw>EW>?z<$Le>13YA}YHqEbZ1vvjzu!k&gjp+RgjU8N9E2u_xY;f30kpt{N3ZCt1p3
zQSav@6Ao)f)b0?RyeZbc_xv2>;sTBHe%>GN?;BZyOQS&rWJ$r{VR1iBBE^+1d$@e5
zB}R@tJ}sR(ZHB|2zNZ$hS@CpTx_!T%yfoVC&WJYJi#?#vkf$<N-Qk?!;PL1kaWe6`
zubaakZ-cR0z~A!UWAqL*ntn~39X>v}J(u{tJw1GX+<mN#t)q3~d~t*bq&D6iU9LO2
zA_JCm=WI&LjjT8V+^X$F&rFNy$!@n+ON0$+zkwW-Vyqd95#21cJbJ|NSih0mV!gIu
z!Fr)Z^RT4wvx>+*KSYagkW}_H@|gSI#gig?i8v3iCzzhVwEg)c2sI&DD7NHLhcJMh
zJO{vG+C^+<CD<aF?vq^G&jsK!qM%leXX8~X28Qql@#6u>TJW2UfPg?vS0FZ;cq|Xv
z#bBB*Xd`8DPN=69&)F|2a~Fyb#N<|?O5lkDqcd{%B}_u_6W<O**hLjjK9eQnOQR23
z4!|)-d-dTq_sRm$i#*es^VtDdX+qqsfQPh;tOQ6UtT7I=#N79w`^f^N!(w4^GaF4=
z-NF5uq~UnOo}B?eWF4E!%SGb%_9t<Kcx;@Pmy0)&p2Tz&-~ZGLEVFiFh9Cv@5m;0d
zK}kwG_$C~ht1GJtstYr6EQOi=WO8$N=r$vWaR5M+4oIa5grLw2SfO!UCzKFL_Co^u
zlo2RuO*e3jH4GkZxaU|tM02-_{cvXC#qa9E&|_~U)<0MJv~&^}sYGe>C{7JTXGNhT
zFtjmI=%sVx42QIWgP$6MtGfWn$V_1ZI6}VIVk%qV*X=tRMdgh*g(DCW9L#@Ihwyh`
zA?eV9K>owd;I&9WN?f??04=k~*UPuU1(P+u7?|UFHU_|bcnJSu)SIvN&HV_HH=M=~
zw!78e_V4t{8<2hsY?ssnVz;;fhzFZt5)6hJ6N`(*$_(bz447-(FNQ78FyiA6ffPmM
zJNz3hgxj|L5itniJ)-&^;Di7Ii=Fh#I0lq>C>Prtn`UURV-%!`L|#YpU;{b`M|!oU
zDygdk;NV@o3}1Uj_K(B&&fR0Eqj2vduAqa>=V8l?gTO+g0)<ZqxM=q7Wz<W#O?iXg
zcLEV7pg?o6!Z;xW09s=*f*|W_`l60toSOTpThXp0>Z5=7b0I*2pfGNI0f5`go7u~b
zA-vUHoN{j;Mu>h};QO2=jCAH=B*d#)J--CcrMx^OuKrK<YkmSoP#LG2fqg1ML_q@J
zK(PE8mO4eU5Xu^2Bpi`=%IxeV9kZRt6=hyT*4HgNUS3;rmfwvt0-!p&k0<%4fbXB+
z$dn??;4xaJgje9=E0bE#83I$GP9eFc1Fk9V$0QKZ1|>0JeK7b>J2^1P_pW|8UPH(u
zG+Zv51SPY;!^3xk1f}Jet9b$0d-;wN|L4R1*Vl>PI>u>NPVJyXa_FJGS%L=CS+L6~
z;B!=zFDj3DdgHlk>b7j5B^uthah?@`5~&#qRhndx1dQ@V!fG^_i;o$xW1IxFE|;AU
zLi;&><6ZRrCWyEw3>v%wXoMk-*M=J5GH7dME*dAS-V)ad6JkonO17AdK}&igcHc|&
zb?hv0FMp-t@#vYnFn080#Rb$)-$|cZN^eoUk#)lt1!jNbsv#u`Y&qg&HBL0%FoJ~}
zG(ppurM#<Qes*+tBBjY!ok{!x-m{LsnKkHlaCY#I(+%^lhjW%l5D^vrhCNzUw6~ly
z?Fg4gD*V<?@{|^*P=s?wn~>Zhdh)5O5J-g$g8w)?ArNbIm;hc3^|;@aaqlw|j(L)C
zu0QLdKdbUO^pAoTSb_?YtnR)9C)<rxD7O0t?u+-REa+y<mN5!+Fdi?8*v<AAgO4w(
zbw`AtO1%)?L9b=+nOXX~d~LpTTx{;UfPSEFI@2mPCqW%va=X&6@kJu}!HEWh6E3Y5
z5wJoTF(|hDxWiu9m|T$#a3#JM%2O<8owYyS_)DzI?QP`vfE*fcY?BWgF|h@-7xvki
zY0fiCUXq}ARCa%Szr5eCC?ADE)p%kvFT3~b9&Anf!9V&UZu%(f^i^yXe&Z5w$Ia`c
z%kv_<|FrB72@DN^f#vN5=ZVZ~X6o6)3RtF++$+H6H=`}yO=XgLoqFY;tLJER>m2XC
zJJ;7@s?GyAYBJ)(BJIjZ?g#&EJm~AyNPPqXjg@=@0f!LrYiMwhQpOFA9a=orgGjrS
zqFrAp>2Q%G`m-7%DuGJT&TDq<ebMXxnbRy;=0i|ihBOu%NlHQm&ztd2tT^YZvN^fD
z+$?M<=2)o|=DP*2y;t+c?}Hus@d<9?!Ij-^tC;w_7{@g|LH^;SNlFY@XfihedK3NI
z`~aKpC+(B)w&%U5k6&@8R8nk9xJA-wsAqinpkhu2HC+EIZIRdSQlV^D2jA*t<_k+H
z9S!~*{u6<IPlkR^r*rT)*QT~p7!Bvi_i67R=YJSyS^v{sHS2H1z<<XM>-udI_{~cG
zKmBh4`n@j8fF2h_tp%k8|4MnCetz<WB7_uBidXBi1BoJI-5)YBMdA3tIXkanZvLG#
z&IclI`wi@UjR3F^p3#eAHJR7}^#9^Rd|`dN#KTi^h<P<7i%;zzHx@Q7b5Yzh0YmUI
zKv(%WVeR3o6*#U{Ts8~MP;*XUANdCD*MHH)@q(kVJ3n9FBeEt2xif!kBrLMgg#U?4
zGMVlF4J#y!3Gun$Na?vt-d1o(XRSZb)^x<%876_unZaY@#3ruFv09V|Yg5VaW;u|~
z>s~+zi0ei@wTl@BeL<4x0WUlYCEUQ?FZ8}wZfJyH%i!wX&ZxD$`S;@@Vy)c2%AkHL
z7;pUj)*Ok96;=)nTmX?5y7L3a@8a8{+(^IS3N>hCWG?ZV9zO?V_^Gb|Ep_wpPYp_@
zYCLy`Kh)&#Lu(Fq?1)OA8WussEj!q9&0i6qeDN~OLFBEVit8fb@V?MSB3CJ=6j&_R
zoX$J9A?MlSEf6&(pfnl3qaiVl=+&<JYQ^~iRFiX?S|F)#%8lMhK6R}3uI9l^TV#`2
zxcKP^;I@oxlrgO!K#W6m+SrsoqQfcD7cGNB;syy~txp2MoGH<)yGU1tsT)%2F~4e%
z;+ahp6ai0X;87I5W<Qn{NO@oa$6>I6H1Raix^XE|k@+Jh5%uJJ*+h^JGgDN4I08Z1
zGRWP|in1gN9Yb@aIP7kX(uFHif=YjkRSayHenNq>edCgG&(n!hDUb1xu_Ws_6-CSv
z&4N`Hr*TSKBUP{rKY19MS|(afT$!;fHC&S^HL_44gb7IF-ZA{6vw+v!U0Fu65q!bZ
zq^Dmi#WDSAkQs-P8QLT_>6G)%!u`+*J$6<-&#bQ5B{zvdEBR%;>UW`IIERONEdq5I
zf8|xQGJ^_CbYj|ln11PKEZxZV04vVB$5Lul7mhkNTnTOmq>yqHGx>3|4U<X-ZS|oW
zB&^qjni$-Ltoo}3uIu5k3$BjEyJF7+W(erVho9>;?Zv2WZ&1@*z3_Il65Jogu(Rj3
z8z7ujWxkHVO>m!quRhL1fxqmL1<_uVeUJ*;O7d^HqNo)cwGQX14rz3Jl96(yW;u0u
z#eLEZWJp(WD=g9oQ!)3$mo;yu{cHg`oMg40qW?@hE8%<bR$x3ur;{4#sFwipWfuK8
zO4CZf{%(q_LGpv-8%oqH?SYsxh$eT1b58~FtSDr3YsaPIQW+Te>~zIk*f6FGUzN;C
z<t*ERs`H6G)IErxamymIv30B0BeR*d09|0`bCwX4_r{^Rj<Fz}IPwyol^BK#%)r{a
zq4(ml6g(&_YJmKCOj^D+OmE~t?FRSZU6L7H0D?#t?_CF^ARDPfDy&Fab0nC|@k=yA
zYyc_|W!91&2z5k;PWjE{(o8;#M~2h$H^su4^Flj_^Q5W0PfL566c#SsWysZHxO6bu
zO7#|#Eez%<JB3*wgh#-pQq7$5lG`YhHXn!Va$+mGa{H`YGTwA@7%WI<%#5*t!u8dA
ze0;>pRErW)ub&3QSl5o3f)=4!+3g0c!?+kb0XZA)LhiW5w`5L4YI|u;?uxdl{HWqn
z9iZrN4HM$;bG8_gW!tcGuwaC`a2TWxV++VPvY*3DSl!hz5nwrz%2KNUURtLNB5{-V
zLMJJw1JM<A(q5y0uhY(eZa2^l&PG<1cRKUwI_s_w2|4Zu)=Q-eIwS+nw{n6nHQdjs
z!R_s;GNM&2O7!R%sFFK&WoU}%Msqa##=>eG<BhGJ0Qa`r3A|(x`iMbgX$CtX>cVu*
zu1d)`3-d<*Sq^ewV6gZ&oq@#SR&&9$<EYy~r1pv|n+6U=B>HH|dE}2JahC3@+_0zm
z(zdq@SAFVZeg%kAcTcv{g+<KMlr}&9Y&-cU&5vB#JGz<z9Xfq!`E%L}&#_Yv%O0$A
z&8lR|<?aV#`n@QB4c!=_nO{dd=m<Djyk9@1qX?Y&PLTf(upJAo#DBUp*f{>T2J*iG
z_WxaAT9?L{&7tUjg%eM~wdoonngK+CTX?JqT0#DgDz?$}rsl;4v8%wU6lE$to7keN
zf4<yXUD>+Za<S-!+_gadBiCP^;GY5ydb3^&3MYU`%(GoTr@s|6Sh)D~Uh37+djk(Q
zy{26J7!aL6Vemdsc0RXghq$_i3_X_GJ|nkOmv1m$J}DocK8vywaP@d6J=?zAXF$()
zNKa^rC)N91*owpY8c_AZf5QiRsyC@uXlQjxKfK+^dg#4^NZ#&{^p%L>9bD?BsBg$x
zX4lvN!1LR#2^SgNO(OQ+h*Cg)NL=`UHNO9JGGeYVw3A}0h>*J^|K4Kf8y{yqN(-pb
zMCjb$e>ARscf_*}{Xx;r4MWR_R<BR^w_X-*9&{oPj!#gW#~RidU&D>(C<|seo5n#M
z^XO}Zw4%QYSc7}rtcDEs*CLJCjd(F+{yK0K4tA&U5W=15jw1wk?RaQh<wV$c_l47-
zV3r*bV=!&W`FFK_jBHqV6}U(ePtK83=I5a`MNK&9&cj`7)uDSRm!TsC1JewaW?d{k
zW~+}JT2`&pWaX<*vr?};`*a=!Hc38uQc{C!t#DiKkU)}Kmbp>vVH~1ozP105^fi=@
zG(e8nl2Lkl)-+DGtP~d?6M^xb2uIPPSCA^jNP!)0QU91+UY%nd;{>A`Xsr0~nr4z=
zn|`&nAuv)oImSsnwUkFuP6}&ZD&~`j;ickfQ;gCv$Vod(YhnOH@lKMm!W~<&_#<LF
z^xdtAL~G<$%CR|F^a=XAudzaRAg$-0?$zk;&{Z1AySqDqpq}*S5)oZf)q|*}|H<V}
zQs2u5ElXxFCI;`tEb~;+zUTs~xzJOxw`Ld@Y)j1KSM#oZf)e@^e-iP7SKjG-zdfH&
zD7eD;!aWVfwh(EnOH|fD7Q81)L)UUf^DZ7gPfZWjhFjj_Z75cvNUMB7MttFfa@whi
z7OH#bhQDRe$<7g~X*-IhtVh>&A}+cAP$zLPTB7EcZIjWg_vZ+#6-Kk9M}2*wJ}U}t
z-#a&*<d9DoAQRy$$7$e3j(kyDFT&X!;wWV5nc<+I?bg{fP~0FB*(239+bj9A3`E3I
zeaVIeXM>Dx_DfRs*kEu7q^^m4i_1^aweTDpY2|S{mvYHp$kfE#uK;)4!%lB3j<_Rq
zy!26-qyhF-z`ckD>RK%Av1I<+S82e*mxs=Xh;kMi{c6xAJ5T8wL~-?=gEts<x4Sa(
zsX`LYaw?i3G1JG$0_s@1Jz$pwa&6aL^1^_I8ppX>oqNG4osmthes8719qVjFouqYA
zp&sAD-7W7?9$QW{{?>k_oNmx|AgUu%`1xrf!U;Qzgi1X?dfy9|IH!od+kAR%pS?DX
zPR_!<(I^w(ZSE-w1(7W|wx#3ZSW@+&^LqQa%c6XN7Mv01Ly^|eqX3Y0l6F-x-jwdH
zow5*Pzk6#=Qgj2mCu$XC7j1ffr=8nLsi*K^SzmXoTZ{TCnzp2PK$NAUGlr=Pe-YT1
zPXzA>a{*Xic@jLTe2DgMyFP-j6^O%Yfh#YH@F6E3a!cn*(?x~+&=F_1TD6oGlDfoY
za;<NMWV7-RfII2PS7r$CMmfr+6q-*>qSqm&&Aib8(*(@I1V<%ACKUM}Va8B*&n$u=
z&06dgHRYmjZ&b#Fci{I`P6n*ul^|gLrB&FkX&BNwt|Uj7x{NETG@Ka0hhw;`gJM#x
z5`*TFH#c%KcPVz%Xx}hW(6r#)4o~uP8NZ3nNe$jmps=D8Al-Hth!5%B>EMzH_yVOV
zVgBaxo2C|?4TE6J8B<2@pUscz@lmFq8jWbb^|gBVO8-x1H^={>U;i(On(4n;AJ(PW
z^<T?T*DsWhKobWroS%R&z?M7jeomZL=tWQt<%);fJ)?0%l;J4G=2t~=)^kLhKRc81
z4zHGR1FfkOBNM5a^S?7|@BZzbA;7oLJ^62}oJAeL00jo$7rrOENO)i+n4Y#+(M#j|
z>NRt^77B+Sa1i|m+RmpLs~^WlbdN(V-9Bc&^mcrwcgNNs>%`OCNWHK0;_CBbgr1%W
z@xS)GO<yZ+w)u6x-6@L-|0@YM3KhrjdG{jqD)gHx|MH{YIo&(S9f(cUyHI)1U2;W9
z0!B+>2{9w2U<nw79j3^yq7C>9#>y-{nP*~;u(RN7q}|xd!c3tNcPK^{e0(%MGSd@c
zarn_1&OgB&U+x-IyQdjeHED;Ctg-OQHjiZPL>%pJf}3<m-f@8Ov%L3W?d8%p!g^On
zHeTazOq-9YnkkKpov>Ug437;Z(lI^P1Q8Iwc3V4rjAllP<4%xxEMLGrmTTpcp(e;7
zlN76+i}<^2kao*`gR(Nflza0?`r~Wm>-lo_K%IpIyXGK@F5a>-8rKz03uu8!mU%9r
zi%C6tZlb!!q9_SXr&j$2GuIPXYN=exhy-H1wm`BxYU8$<R2<0pMx;0irq0%QV5S^K
zIrGP(g(A^3idzTCULsFgPQQnd<t&(74`zyl*)CBb`WSbm&uCtJ+Q9BlcNP(AhY9oh
zmzJ7@fUbD$`4{sL*1!Gie-3i#6hdAuPv`f~Ebj_f26gfH>0=W3NB&8-IAuRVc%zUf
z^?Q9XHLoPP$^w{XkMUOb6VdU1i7^ty09^#NQmeRUdw~d9LG_=blJh+AQ6;zx=9bFL
z{-}q%Su3M3<tcpTfhb1C-#}7!Z3;@32&(gQ_OB=)#@7e2KqS|3uS)%f?tVK+%IsiB
z=7wxX7}*F-nwiF(y`FGFISW+fDnnurA4nY`Z^BDHh`OldrqGUd{dh<#vr@rX1y(7p
zJ22k5g^nk56hJrE#xc1P7I3t&@cx=Kk#%jnEs2Z&@$DR-+VE&`ieH6ZWUh8@-@lf=
zW148<*p$JKuN@lQOkl>pJndA!i<e5-iqT(z2B(1t$%zZZFb#lp4-e01_H7Yh5+S$d
z^-RbJlm_|A%FY;0NcipmjVh_TC!(i+#-yP>>UH*s{J8I^zO0hZGecZV1Pwk#kfiVv
zE9V+07g7heor9c5)v;|&$fNWGgOYG#7zpF6Ay_r&R458Iv8=rv#xOWL6G=<SP}1_~
zZYT^JsOzLjl&anj8K(~UN|3tvT6Tm<jsmD+-=O+bbgMQ2@lkSFV%V3r+tbE4lR!XL
z8WN0QP84Wdw@<{L-p{Hq1)(kV+p!xkkG`u>m7@IlRK6X)$%0=mwF}F4?FUsA8G;HW
z6fjX2l=S1lUjv{L9=5h{Q=x9!0i;k)0S>58D|oSIC61Ny%f|QW(oYuQ9qcqvy1jzq
zOLj<Om?}w0*dm9J9cZpZujVBi%90P#VpXACfvI{-$pjq>)t1oR4ur>h!H}i*w{KcU
zwDFFuPx-<-$4uMTL(Z80;=7GR3Ww$6vcb@>b=7bm<x=;jz((~RX3atlE71+#j&;nj
zvTTJ*7vIw%u1MnxsyTmRfP_K`w6zqg>*nW3aUSr1*E(y{p_cuwZuC4OM8V01-*5`u
z1_-`a6=|DcVL~l^BG}5p3pjC(=g@%eE7>}<7u#Mb`g3?LvV=m@QEIVfKx>^0Q|>-e
z-}pBE%5hVE{tuJYgP0{f1I(L5?WW0^1@b6=&qhM4fTIMu(8C?XbZh?1hSj_wlDy{4
zW?N-jGoL5c5ZF0GN2D?5Q=BU1Z#3KsxjV6wb*X|W$JRtqMCsF_0n159G#gj6&X99T
ze|h)3p%#b3U-YiWvq$|#8)$#fGL4s>G8|7>1-LNnd=(UB+;<1ChWVRu+WU*nV7%l4
zrJsGSCre2wdw11@^;^u87&ITVG{_p&^8@zdm7VfW7iEvrrdtmt170PLXPI{&!?Q4E
zyQO&3ub$*!MXQ1qG{I8t@+5Q&620(C!UQS$3WX>YMx8TC6I+Wl3X3S}R`(6tm%kg}
zNgKTG%CZy=FP`_AbULaB7SSo>b!*5vTziXTo~<Pdmrl>LKj`*LlV0pDS7<*Shhqn$
zK0yM;LGc}A3q@#I>HUJcIx<?;A&raDdUTTMFJV6E8}YUaBfV1(vjSaPQoJ_Svl7T0
zfPf6F<SHp~9O=f8xyRy(=p+}g_8)#~*tBs<%|}JESTnMK4xVas*y)KUnt;I|NZQS(
zo59E`(G%h$A~duf{kUSK_4Ua!*C(JHk2AOCIhFJ8jE^Vw2yGqlI0)1@i@F^0@CW*4
zkVI-iay;))Pss~U<6VXHvMHmL?`Pk0Y6U0XfXV~qr2ewJ9=n6xdrj498}7T008gE@
zg1n`=`1I4~YJL?eFLrm~bYze_myf!v=5SS^8Dbfzf7GatnYlEKIx*HPKFV$*)QBGz
z$99&k_Z((9%$i)-fn>ouvmYVCSWQb(Z5WcL+U3Xoyjw~(Jcg*^yI(>1H98i1>Q_O^
zA$yrCa+|^CGxKTy?sI7NtR&`ScdaCfHYBNZ++z#7W&lq^4-8okt>IP)&tR2!%mY@z
zuwvOwE7*Qus8g{aH^J(7B5y|puBSuYB~ZC^H>aQ^L=lH9Y3bRwLo;nTcMw=oz$dg7
znd?ra>$;}*2(sC;JSzbG4eUr#n?kD%pJG1RXn|yeO+a3XNI5}G9Z{Uy-}j~loL-iZ
zu`QL`ue$Z=f3TE{h@!MTo5CXJL}fDBc4HtD18p@Tz;8$bYJlLOEStmu<<#R$KQ8$@
z3K^5f3N0sn8JL@2$)>)Cef(L)rmIS#iZn-)Eq{%{EwVagl=WZ|W6gTwDhXhu9H^-;
zCJ8X~R=!PnoBc<Id%E2sLo%|WAr6N|QP6g4-e(zR6N^wqD_Mr#?Y*dG!--4Xuht2|
zQLMTX!7V*WJ6ERIcKYK7wD>yd?|&MRIR8^Vkd^JfrHMBG7L8l7!)%>UU&9M0Q&<26
z0JI>ZpVq+t7p%<Nc1h&pCsGlZb*s;1o(_k+18FEdv@?DpR)mh9Kbt?tr^7-!>sFLg
zov)id0#E$CoP<`hk)t?h$>QzR^6g4QvhI;{0lauIgD=MC<<VL<xB0$F2=jebsaQ7`
zx8k?Z&1}@PdCHXKDgfG>`kim+e-ygK3sU4hT-@OI@Qn3Wq4W*w9jtt=e2gsY$R#_^
z_OI4q*ao+_ydBh*z5*m8E(epE88FiPe71Y}zB`v_-Xu5)r@aRHe4LChReMp7LnkDW
zSF@VP<Ea}&=~clBj)-Zviwj~ZrV!@GqZo)Iic2yetbiH#c6c;PtQFjznvCD3P$r6u
z(xfN!j#B@K6TwiE2o|5KFg**aMDsLnsZyvznleEv5OmIyLxGFPj|yfIHRX4OiMS=N
z7`3=`fsBR4s59xBHNfs5QNBwPdJ`ExwDk`DOJGn^v%O<>$8}d^<ojios=qxybYpXn
zS1CTw-a51UVI-TGj{jbzLQX)<8!iKKr!oBmr|IV**Q)Dya)UStohHu__b;c5oUVhW
z<}<9mBg~$PmejNE-L%^YJR`hx_4&dD1JZZ$`$#1R;c0-0#~S-)_8KG)q;O>B9Lau&
zqQMqt6D^Zalus%`JE$J9mBixtxU#AKo6Ewt{hqwm+sve!*)*%sdWOM!;zV1#TIFDt
zYC2nq3$DcJPt@gZapaA1=`N9bRZgZvt{Dxb*_47~mXwjhop$5w`r*k%G`_V32@VN0
z*GrZvyE&EDjIVkW?QI>90%J02^vB9aF}ngP?1-xY%u$e&2yuu$Bx406d~-XX!qR^{
z4eV;~`3hCTPt=YI0<+Rx7QCgN?rizubFwCoXGmH2f`L6m$cM&Mc>)!ZSRNT?>Zg5+
zy-pyULdO@tf*mYOf*?nrr;4fQ8xYhVNv_+&HlLFQ<JAKyN{X<GXQWqllo7E6gFUR5
zGk^Ub-J+^eo@r%6w|q$LKxUWp!y;7qp{q!#8f^wJNi&HuZojGollm?s&_C+7=)`~g
zY9Y%6UZ#-EKs?xU!x(ZF|J?4N9*ERVPx;M7_lzp#_7`v=_ke!qU6~7mQ?M9ChCwB>
zEbQsT3Tt<hRtHH7*)#Q^@iB_&t!e1g-78zDCFIvkx98o>hy)PxaGmF*MCvfO(5qwD
zE~?^g7;^!N<G2mrXh)c?rrh1JT&mxx=jL>D+ErJwx(=g$%+Fa*XFri2z-6(fPus<*
z$lFqTZ%2BCu8D)YJ9jgx77t_RMurz$=3ep4)qIRrUdq%~%#WFDWei|fb8@m$N6FWy
zFXR)TE>YNa;ubeTOm{s!oYk_LhL*!o3<cM9+%@XT!NRr#m0w!_mdy8dr{vE4rNaH(
z$XQ7(EuCJlh>4bal2X&9OO(s-=zb`1c(6YhIZ;-_f*LV?QhX7hjQ3>ULxJv2WZP)s
zw&W;$4-ogjO5EfoO_|XKlMv0-(SU=Dd{Z59vi!O2x{ZlqXuX)2719n3kc(T8x`HG?
znk1Wvu;$*o3=(J0b;5<CAYT52&8TSpgP!w#{M2w(um5LPI@vms8x=)oVvAMN+jXaW
zEia3Ljbh1E<zJg5ske>;PgBLPqU_Ss;!$h@I42!{{>j#*3IH4sf|sKA1>dN3r5x*K
zu1CZcV+ow9ANFt>%o$e7sJre(4Y^7yGu_Va=?3oKOS!F%g=l&BYXP~|oEp?EGmCF&
zsD8S($}zmGA`6lY$}IA7qhf)|k;KI_ZW$R@&Hecw^9Tg|?!T2?+qZ3vg(h)M=P@_N
zNt}Z}qkv+_2V1R3`6>1dJA&*&d=qyUl1JBOE@qJ6nvfgHSd}@wu$(BNU9HA2)`E%5
zqOpgz$Jthh3^^WP5}0ua5o}zhVN2(KTnKsvy+d`@Bz6w%CQ-O)Gg$=(BDh>BtCkik
z{w&KwmTEDBLl9{9j@#bpT;p?f?J)ZS^lNO_6c!2<XXQt))d1B9O=@!@M=va5FHT1=
z-e#OE>+xkMqX4oK>~fHzk}9a5-FWa1Bo?!o>?eP0&-mrbqG3zws))|tWFK@g5#>-k
z<Nk>lx+vY5l1_3Rsdi`MzFMZ{yVsUaNILJz-NdbUEa79_g8K&aR@AW<);$wha^I`F
zzk<t-ZRJaY&o~~v(I*vJOjw)O#b(Xz;dukO3Ucr$?wR<a4`K+*yOKnQUz$EfcFvH(
zZ2uZftI<3AO7a!}&9_byJB*r7<XfPv;(x>Q1x`_8)8@Xm<|c!G@7Sr>1;Kj4KlV@m
zyHnUp`!kvDb>(h8pON(H{xF!@ilL7|J$Ag@Hb-H?d^S6PVWV*NpqqmmR=5Ufrk7;(
z6!QegIoj71Ai^mmcMHe<aW8|m@GqB-qBZ+5X9am=NXH-DI-vBo<4}3m6!y!qx)X@^
z94PI=tB(qlwSy`83DkDE)O^0fJ|nw&I_aVukafH~6qj<(E^`8d*kGxXX>XB5XEI?s
zDo*mCDQ6Dl+BM!TTZFfejs8tfQMAk2W$dkcquCg~FC&aLJ9APSYBa|d<0yF1f-9(G
zz=H0~@Iu935#{<SmqoL5SNC0(*d3F~Ti2O!07Uu3l{Z^5G@c12dZPGef<-)IY|I^H
z2ln&BJ_C0>ciaEkLdIdX3hYg2VJMIsul&*QPl@HW3Eyx&luv`kr``xfA2jf7FN*Ou
z+2%JrM}6Dof0!!R7#aS9XwJg)-<m4A)PE7E90;BNMWr!VojQhwfEk?T=-p-;wXQE6
zn$9||d-w^ZRZOB$fS-ur1rLTJO&B`AJ|}a2D4d)6xV~-o<ij-Ygi7II&)vc9hmP(k
znme;P2v1^O^!m2?_L$4S%>l-nTgTwR>jBpM=gWk03#!56Ce`G7FK7<a&A1uET7zqN
zi;6#ow7#+!d#AviZ4!3GdHwVKX;=!qdZM&wDV~iU+#V`Ut2!jO4pg3X3s4T~QB?$r
zL907n|3%yKL00?j1=ZA%(L6qup2|L%;x<c5FYGG7V2in2P>=%EcR|T0{!M!KkDU6|
zAl7&<ZYq9Ed5)A%=|T7F<;wA8r)Gy7+gQg62h><#h@ao}EEoapsop>}zUb}A_7{w8
z%4^A}zGY%NQDp&sORTy^X(@D|o+R{Pj9u`fbz%L!$aJsh6Egp|Z6O#LFyecr7GGxR
z;ck1OpUm0|D2ai%D|)jTpHLs>Urnp(?s}7mUQ7J@_zyOYQO?tlA01dTAw~FcRH0)g
zVZz}!Z311DlH5Cy0`-Y*Pz?^$-r|!<RX5KVI3mpbdQUd%dH0ar>6|uG?S;2TgHY|9
z3pNYx>8wn4DOWP=!<t=0A1ry|0pxJTGAdaY{>{YsuDXUCYLE*1^>=E&Qkk&r9}o(v
zkgXBur$2C(=>VO>9lNs~W$jy2mzmAD&PAH4F`h41Z4)*IT7|NCwHM7Ilb#vptWzK&
z_60Z?%H7l&r0x1r+Afo{>(;C+(XHk$GG-0yQ2?}B3ZYoDg+<yVR7grv?GT$8H*`n)
zT}&pI49w^IEL&5r)u)aBbeAGU)zVg!w**ComWkPJ2GAy3jR^xECv2d%QFbwEoY5Lw
zmKA2i@XL^;unbY}2aIAx+KV789i%oWy|+C{fuVhfZYANtvX#JW)J;;OVW3jKC-<Mw
zRzwjahAf;tv|8M#f`*oeJ!~g}G`AEpSMZ*yZa2kXpt$yO!sN9#)WEt(&!i;L&6g68
zy;EC0vN(RR2;sUG=PRgeb@%H_7v6nkk@Mz|&Xg<EM*$?o^SXj3(5Ws<j(zj!EFB^l
zQmHKai@u3ES(Evru&V}GtL3!G_Yvl<o#V(A^0!9hr_>S?+P64VPB?UT8I7|%%@{t~
z#q6d<DENkc@H7m<19?{0>iQV5ZbaQWne3>@!$d`6Tg7Rs$L}chwH|w7sIXvgy9x&1
zIS;vipx3+28a7xGgo6Iem|x5Py;hclcID%Lv|ya|Ls8DAyOr<Rk~fWHb2<>?^7`Xj
zGy@)4=*e4m#KAhaFIjPHlk;wu;w_f?FXO<uW@4nT>dW9dKcFv1ndLbqpbzNgz(+L3
zI5xCW#DDFe2}MoB;QylRoq|M(x-H$ZZQHhOdzWn+yS&S`ZQHhO+qQk{bjP{T5&hqY
zb2C<CKILQPoNJ9azG3qWDnE2K-)ea;xa;&ArBTGgwL_Xo-LT>EkBBDl`Q;QqG{o@2
zpL89^yoUaJwC^InOihFGdvxy<<kzFn1-H)wLr1-N^g+oiJKS=nx*iqy&0=+_TNvTK
zGc?fs2PWNewR!#0A-u)g`^W!0Vq(w8v5KXj$iaRES<LR$kffklgybV#rvopK?C;;?
z#R4?CatU<}O#Di!qvj%OY~viGVZK$#Ha4uqGI!rRY}9t9*pOtLa%a;9`<_5n9liK1
zj|$fwY*y?fxjHW09$^~o&)U*~-T5QKPl9CPCFF~Rtz2fR^~4FwQ{h&gi{cDh0MW!<
zo_B#yq$C1X`u&~Y1<c6kn87n3(`@zn8Gp6?s7iG&UM<Lo6Xg1ZbPOw!P<=8tQYNR5
zi(&AOd7_VR_-2=+@Bj2K{yzga|N9V}ne~78FV^IYp=m*!eo=F#=u_uYMU~$*__wu&
z3mk22z&zkbsO}F8s(P4nL)0-!wl;#(;W)d%d`VWJKB5LLF5e^P@&52%fRz^Lw6O~P
zzJtfR36NhIliVE7+4UY^98m_pN7nbJnw6_t_l-#t`R?e=bGL(+7a7C{m`3jtW%uzC
zwvh7^ZjUErlPBbrmgfQY^*esp{a3~;K_$}T0dsyI8PUsbWT^P_%cO3P>e=Sc{(ex=
z5La3r8avBnNIg5*$({!WR^`NZ7axBJ_W4<PoLJZy{zFYWrF;xuQL^Kk6&#l%@>j9p
zw$W+XwO_cT5;nm}QPG11dZyFDMnnC$AKh-dHWi^=)~5qxReO$06lDZeBYeHy95!O{
zFzI9|QPL|pX3~ef|8gY>*jN;vy5Rb?+)lh|G~>=j^R5rrnk?N7@}_^mB~CY~<Ze8+
zM>sqE&-4VaY%GLH`PQgB`)FrLc&7%ondLHFVic-efAnFbW9zzbeEe4Y_R-^|vm^e8
zM~hrsCy+e#^pjMLTa{&g_~G!MKI?1rD#co)Z8dBVO_Q{7bT0KCqG5N^4=y8BRe6*q
z6X29C_4MSJ&J^21m)3r6jNr2Ku^P!84$o{aY?2P}Kb=NUJTl9mDVN-qC}_>uG-y3w
z-hy%$0ZfyIds9=-X(Ac4+f|j?C{5U6l|!Ztkar0rw4+_EaT&zY@%HUX#*#|SV@K9x
ztrH%EBG#_cDpH1skhw~oK(8cESLS2=1|bi?Pbt#jmb=IT@5t&2#_rkZGJ%3J58J(_
znTqL{7(f|K<0DpnHmxTF%?F5vC8`NRm%?)Kk$Qc)fHds<vW`>RslI9tJ`(R<T{D;>
zU{ZJeiIR;#OOxQNvzdi_qnl?PbBR}Cq+8F90L7qo8METSPa-VS0&%L;IW&TAyz7?Z
zN%a+&akXWx>Mn^C?%ZjiP&!Za$~b6fmhNoEvYS~%e`|1hB|x`bNdfvOdmZ}g-kaHn
z*C5ZZ_$u`x$*}V~e3>tK`d%H5e#eWXAh@O_C~Z^?29&qca+J3X7d4h{3JjuyGhm<6
zz10qA36)j~i?Q1X9hp+XWrAMM<GzDq8?=~FybMu94m3Mj6k=mhVXr+yDno;S+4VU0
z4Y_}jHbhNg>Ch(E!<r^)s*7=MuS%R{UO}JGr%Nc>?<6gfA$5*vr9I7F<Xgv>&n!JP
z6?fhptqcy?M+`x6qRl8YMvxQB_HRH9+di)sF-)NBB_eWNjaQ3z1N;*x({O(tJFU5g
z%nud%qoqqbhup_|;L8|5QH4*-K3`I$$|(onP{r11sdiBvr?(Z~78s3E<MmhM-`++F
z7;GUDA$?f%!xnbvF4@IR;VFY~m`CDg$<-<BU8Z72%X>x97`;?G-IrXc<8<WN*lzMN
zG9R+B^ZV(<2`Q?VMnreyX22{q2JHx&yMU8B_cz@LOgA0COgWe5Td&s6v=A^`J14qQ
zB5fiOL=CbBvc1)fwk|8{v;MYJkVUKX4K7JZ%IBhwa?2DF1Ixf1V6}*M7_mBNzc@6l
zs9YINrvF6@Gt()WtBiR;UdvB$N(Scl1GR*|3UioUN1gim>(vTuI2+!mS+&XrKHIOM
zt<;hU7ShkYJ<JX&*MY$aF2!!F*zr0XdrDmO_eFNpSysySqn7f+vQy%DKd@cG39jqO
z0!3GvE^sBHZ67hSB5zo9cx2g*bp(QWy8O@b%F37sjvI`ZL6%H!QC6N7l*#(_mkxC-
zqhpi#X94Iv-4bq=NOe{$Sh=r?L1dZxz}xjt3B$3Musbf^Lee4#AL&zBR-?vOVeYZ4
zjo4hrDmp7=Rq}_HFltdGTfxedA(-ySsZDsWa{a2mrs>*5e)F0tmS<V5s~aX+)M56n
z7;187-6FfT^bGE%>tX3?m|&0qFk&MSoHo|6TchyHB#<I=X*7}clNLVx7D9ldEsDm9
z1tj|uSKQtbC<fYm$~IAWl9L<`L<*R##o?jE*SaZi6JR3khu=64mHsX+g)NIFp9%8(
zBu44takU2K5FMO^xUYH{+F4BEbT7vlqibC^rK3~ela8OIrK^_xNs(#G$<?RUz7nN;
zhoOrueI=2ww<quzNVS@$fh4%P;o6(!Fz{e*teok4TFyx4K_*J<MKgmEE9p$2^+KaL
zQ4KaWMj{V~Sslx=Fq2+RwpDlgpq#tq^!*|&ael=^y<OeZ$@^#3Y_-bpC8qrl>^8~V
zyXl7E9cc@v3tG!8dsSnLi#5)PtG>8}8TG)L6u`4hOOj|=^u2*hOF`S!hk*IwNbx=f
zih&K4C!Br()ZESyKrY7{uG#k<T(2ZK@QpwdM=ub)JFi~eu2@R!TN63?Z73Uv%<uV)
zz|o>Lt4D0G5ee#-)O6DWxyA(Su#dD_aHU$@o=;)&k+z$L#f{sKtpgw6!SSHtIGMvR
z)X&vhZxQuMx72uWQRX+GV~MP#Q-TU@o%mbQ(2aW{AS&?onWDKtrn{MG#}0ZyqAg5>
zSpj(isxTz@{#qYI^7*qgPJmMJsaUsI<j|Jv#DnAMBU`}1z0dH9E0lL!3@F|rHLn^z
zvLG#a>>88{GTgLPj{YNZpIe#ALIC_!n<=jawFO#~EIcOWEWHN8sZGUEW3Y0Yr6+n^
zm@cU$ZZb`o`b5Qtt|?k)-O*zJ*5He@dGb=F#J+>d_*D*e#$K-^Dge#zoLTk#M=-_$
z8+Hlz!#j5rlKh*3k@k<&k7jH589(JjPx7fTl(J4s3l8#7c`!x`T;8#w^$_K)&dyB1
zN=J1lj4v8E*aK$2=%utZ(+~HxRKgNx5ie}SREJDw6^#Bdn{BC?)lLYnbh!3%%tvcJ
zE}OqsGwZFV?ORodOIm?{?vt;2Q#oRS9*eMS_~Jj2g*ji-(oen@L%c)@%q8h`V(;5I
z0AMPXD?LgodE4pLkOZiOVx$8L(TzV_POd$kTt-sM+VE2N|A07u`tBM_9$Wd6z>*%B
zC2GfP6#(SnM>?ptZ2y@xBwPISJEGZc+3xQ}=8HPSyIq<NtEFAlA20~aG8vKbJp4&J
zQruJz)K#HUe4M?*WjuaWREPn@5S4@4YF)K|RlWWtyZr!H_hE+rhwh7&`Tsay|NpYk
zadNW#zYaPU_W!AaPSFmV6V7L*1|yf=6MjyDgfu)v)bIMay(NT3frWSXI$VE`OfI8q
z5e}!s{O8LD$8*2Xm3?8gWeikuB5h9sMb!JVG|qHd)|!RfO5~1zD441(HXWw|x)sU6
zFQa@2)CS2%yF3;R9#-k81Sdm8<%NJ$;=GEPH4K?mi9~q(8;qHR8g@M>+kPcQp!_@d
zB55?%(IQ^#fqgb3`4Y*4dE!)9pFZwF(lWF`i(pZqx|)15mJA;qOq^`13#!c3PJJNm
zQ#U#`{NiZQ{@h~%gp&L*4l~}RSz_RHun0!7BiR8yhA<Bh&EaGZ&{KRiKxNfC$!b8;
z>k#ux^`V564QM1cCLSx#)*94ye9l_|1J39My+)FD8%hO)*llc=>)vN?U*<3Dm%#q1
zH$`#GP*)yE*}E47`)z1TC6FYyv{L}%fl^+k(+R%|<a5%#e1voJu5OvFV36b0dueU2
z(u!_S;?hTL@!+}901Mjrr~e9+*$r&mpHEhaB1$P8Pbf;amsyjjE#6=fK1Z0Q)OY0p
z{<`-}>~@*0OzgzA2fw*~35mZ7TsK(7BdfGe(gTnB5JhEm8G6VPskXGOm5`?9l1*jC
zogtoeWtWJG`U+EZ3&b=4*VhJ=sUJ1Hh+=geDcH`-XGy~#)d7W3sCuHH6GM%j;U@a!
z!F)*Us%0G^s+JA|$eMMhH)M^5Tj?Sxrq@mW3UZ7@nU4Nugml)u?V;)+2;PE3g?j|>
znJ9Gx&)d!w+-x;y6P}cXy1Bag=+V#JKi%G6Z-$$E88XEW#a{+o+?X@luby|MU!EUM
zAH;OEhB51E2M2H1l~1#xzW_Q0KuTqXR6sRwN{_}BJl(D=&j4_<grDY_>J0wu99i-!
zgO3?C{*N8Eo0Er#iJs=Gj)u37ik7{<?|ENW=L19EW3NUZF%^2c!8_9%Ehh^tSE0iu
zD3o(y*;1HGmWJfc{!gm%Bw5FWGzW<IEh!BL&j;qNY#yF`-GdkMa0oFU;3D@aCGsz~
zo-=t+RtUS&{3?iWsof!X=SPz4!->!EO53)9iAtyamYhOTtN^I^9)TrhCkJ}+B@OTe
zZ-#uBMoCddT_&f;ekDUghP(NF077U1v?KKD>*@x80Dl+xtY+U0J{BII8ojr@#I*rZ
z)tNbodPE|cdP6@xI_z^BHMxg?>X(H*KmKyHWzLA1B@{J?mV15#cjWXqpmSI^{=C2V
z0*prqzaO4|ewr5ug=0}Msh4qL*n9jdE{>f?kTrhTJ+t~pmO>yf-@zK>7z%I?m`)jN
zEbIYp_d86JuOS^8I+kR*@Xf)SGYv;CHgZrLa8c~dnLabpAldNm?%6Kqm>+jq{K26C
z;PrXPnMoX<ezl;@^>(LQ08Dn8u|49PC8`7F?$P4Ik&TxRXJ)67KBvcAFNS|A#y0S2
zVy=VHI(2dc0_udX$ZCJ-)zO2M8&8k!G<N6R#LJQG-Qc|}!h6Qu(mC+O?C*l|x=@7{
zSqgOgip$Ly6pLM4T<eVHycz9TL-M1Bbn~QUAz<e@mqi`H=DcuC=TU|{+$7GVQRGG+
zrxvVSd=0W{VlGLHX)4wWyoxK}0Rf?$Eu?;IZ)hC}aP0a7;4W9y9!F^*>g`_?Tun?E
zwmVf0IT-K@Dw}%HIdJ%ri94jeJ+#@J%8py-A*s9h$-eOmTByxG0qXL`R?%`Qy|SQA
z29a2BeJxo9u%Cgd(PCcY9`+uNZ$8d-B<Z-2)Gf?Kn}M%Dq4XK^MKE$H87xVH?IGs*
z$Iuk$!RBI6gTm{^t78{ujtRv|rA#6T5MfEdtCF`Ci;t-bCS_J<^@$&a(o28ZC);I$
z+Ga9-SJ8j3AWuOLLc7tzolqzg!BhbgCL%#aFywtkKm^6%fXpRUb%4iwl@Ke%7j!Br
zA#D9Uttf$#?iS!oi1QCl02BWUCleIb&S=>r2}SykZWY7@l>!@3xzzYwr!J}>(m}(R
z3KTiaGJ-1tWQ5y1F;P?+K`-QKCx#@kdXRelzO8tv-M;}_vDhRrHYbGbJri`2a&x;G
zdMhO2tR&4pykldfW5$1~{1pO;p5|(CX1Y+^cFk2GRlP5;9@Ef2cql*6Sef9=w=|7M
z3ZY)9$ua@Wlk#nk&O|6xbltfWVJG3Q&*sj~Y7XSl8>Tn@y?+4*uG&;9KSt~LVMDpi
zX3D0^Io%`3CKi-BJBO0hi<G<kD;6QwU=c|0<_-&94Z8&jP9jVPQ5TK#ZXtVx{^Tr_
z70!g_H9b~&z;Y`0!ar8A8Om%Vmsv+H{@+V#qai==oY@}~wRNK>!+n^Bp_0I1%|lT$
zNwzIt3aN@gz#TjjVsNqtLObK%q4rkc6xlyKEMaj+Da^*~12`Fh^iqhCZ4j}Q@4W+R
zG7^JXA<?zc{(UtIZiXP{0AGWSV#p=x-1|}WG-J^tiuMFqo8;Y`i_}a2U`_=VZYsdD
zeA``<$ZbVt3L{Lvp}g^S*H&QhB*Lc)+A30gwA!C1EZYRKQqY}pR<~dg_TCcjIPVg}
zwj&SZf^O7gYww?h>!u$IcKtCAgnGQd%9`~{pN_1<A)mzPUKJ1_K~4V}-mq;Wr*mVE
zNc}D@e(bCmvYUSA=W1N}^_a3RhIH+y>5YlElZU(Sj*5l^2TwLmOkb#*6a48Mo(~RQ
z?zNZbXZmT=f`ge4i-{Tg&=Ux}&iyZn4i^6Q8L~Ei0@C$9|72so@*GWd5){f*`ATa!
zEa$Fy!{Ds^K}BkN(X{3dDHAM`DgYKga19S@083CWmbH|?#p`z&6KuY7ksUUAV3xhp
zSe0)XFc;{-Zp0R;aJ~Ix->Waz&3j;EUfZPBL0W#sI+RlMR`f!2NBTtFSzLxH6APGl
z>`3JnPYD5eke}h3*3#o!;BZk4GU=%Hr{x$spGsGs%r`|3Yw5vOEuv=9Z!wM+Fgyqe
zKRzBjudp8L4{)Vs9{4;&0nY=TKvymL7@JZ({&$C&@F8{rg%B3VSv&RIFhRnF;2Cb3
zIeB0*VN9EI@alQGaI{;LU12*=Nw{1Fy%r)RMg%nm&&4jq()i0>sIor!Cmk(ClGnY;
z8Vv)SE;sU7B^kx3BU}u{Pi_}oOaq)#cx^&f=|)XObJ2mwL+6NzkGPhOwojWX{`JDO
z(lW|a7z<=nQwpxaDH+mS6r9|!ch1+hF2M9~rPK=m*&2PD*nDtHxgU1tf7VfOvD+bV
z3^k_$zQ?tXW^mH~`!R7pxp!mE5nK8n`aw39|JzQ$#KQFdkJtZSc7pVQs#bttzzt?R
zf)<!}v_lkQ$8_QQpAeL>eX8+f&rLu_g$`21q5X<3iSEy)+oOw#x2vOtKW9Ts;QixA
zS06;Tpd&s02g1M+sQ#Fw$u0O3cHq-ReHW1_$J&&gG~YM3=+QM&95{pk^j_41cT1Lt
z-F=Kw#db-OI}|vi^sm&nFDTguFWK056n{JA{KoHITJ&@|7_~SO!u{?d>e30Tl9{72
z{R*i!4`BW=pDx_}3Ga<q-ji%i_zO<g1#fil6JCAG>j*JgHhi+4Oz)MSS91n^HI9t~
zo$~1Y@R-;Naq!fHi|rm%0HZzhCWd#go(6Jq*q=5$L8&moA3FSg)>1<uNp2di($4^I
zItaZcqtphXvG>zaC!s!a;_#_>cm(VxTpMr&(pqGxiD}w}jVXidi$H-z^MI8N6;6i2
zgd}(w+w0(dN+%E=j6KFd7nTZ*Oi0j=7k@k|Y!X;tOId}r1;tgGI3w>b@z)OV5slJZ
z?^^EBMW0Vo<uCGOdd>pVfu_Ec1urxCJ7%$!PXRY<d9G*%6YLDrn$;0&DahRF;jn0s
z#mP9E7H25)_z7j=BFAtvQv^;$7z`h|Nb9y4(J)gGGW_D((QA1Yr%GVL3n7hl<1|`_
zzwupS3CEAr!#`%i3d-hq*>~~#2%p&!_I-i*FyehG*=VRGQe$2?O>sxYrHSG>672Sb
zoP^8*hw=gdj=@_bj2~xEw}rq&<M~!3lKw@rd$P>kcGUsnZ4k-8oSK2Hl$MkpH4egp
zNr|@4y93hBS)kN9pSf&KKdHx)`>&&wII6I8=IES4{@#Rs7x`{QuV)<e(I~)EbaRZ|
zb=a3lO$t8$+VF_>!XPQeQ}xANR#66+B%?x(x;>C2IIm;=xUcK67BWL_M5i#n`oo9U
zYffd=8nGO#Dcedu^BB#$&=<W>c@$(nD#E)p-Y_=a5p_!xv)XpQ1RiEb65c5?$+Anj
z$|BM5{^&#p%DI_)1RtiKm!-q(=zq{?``M+$z(ag5V~nI#pk(!eJxR;Kd`;hBRu|cg
zT!|}&uuj`OdeqH_ZZsOfu>PmJtM&RKikx1Ry&TA>5z?B&ms3%<ia6nD6dF4JrVN<q
zr)1(oDL$t_@Mg&&ZCnA#EET55<3>b{b3JA;7XNwO4yT6#UorMnEHAh}B$|HVC@VTD
zcOUnuNR0@*SZ7s}fzz2V4xau>UqYSq0_In#NwzQuQjBIdprL#0U;(fVBnPu#vB4hQ
zmC9()V?MQV)Y()fpeQ3(1960&t9}l5oZ119-+}nxQ-S#8kVO=Si7<{_53wJ36u~5$
z_U@$+VCVsovjw#ja59!peYF&MvjB%s=<-BU7Hu?WF7W6uurlkY^YRcw>Kpn3>V!yE
z)wyFS-k%1q@#e<BNu`r1gotDjK;~a}Kco2`!~CQ9?~>UDio8Q7T8kn5N;H<U>21l^
zd%nsE0X|UbRpN(M>6+*tI11g9{u|R|L1@{Kx=2h*N#VhcsnsCzf*XORLA&A>B{ogk
zaKW1J*ChZfY7$&d%69W&tw~i90T&Q>#d`8~(Z=~%bj_vcG%X4Pk>F0k7-nHt8G&d`
zAWn=q*t~b!DmM$wWXMu+74buZ&*FW)Hc4xCsE0VL$<q(+(%h^qh)6aYPB>avD@5#;
zKE=<1AFvjql~1{hR)e1vyZmf+NGbJ{Yy?esj@R3R6;qRoKu(UVOY0A|R6Q*+`Sci+
zr<?9)7(*~k7jrDL%W<43Q9n5nQSS7&Vx1`+P*ttQ@fDOYUBp`|s+6e+)R}`<?hOP<
z!e$3~#i%Waq5}*B!!=UqVGniif3h2O+&t1f(r3=EAWQ1bU5M@as>B1Bgg-K_lk8@T
zxws?cVM`jpkz9Ij&H<$=z5HtIr-rGD<@R5xx@z9*!w!s&<;A9#N*}^h=^KJexQd-$
z=h07!h18BLT(^S{rNMOH=N%SjaWkt%R7F5?xNuEqm35Z3=-A%=D(ysBaw7xHG$P+P
zP^S5#)fKfym#|046}(}v`|yg@B;IUJE;j3An@v}}MqFi8k}RbOP6KE+oQX|sO}o{6
zy5cn%Nk^9oPgbTcJ52F!8f^a~vZ$^d(X%?_NvVs1o~gRn6bl%iYEsnPsOGe-=V@(@
zpU}39UTjBRS6L6bEXUQmk(+|4PC_-XIx<KiR~w0{TRaE-g0a*ezMeD~b;;Qcu4|l|
zhGVXJxpZALs5=wUTFC>&Xlc-~B+-$)WwQAh!abM1KX$;l(jNH<6TzWY=9i8)lfNMz
zdxhF3dm-NNMQT7P{c}#zrhfMBTb<LHbwK>ISpAsDmML=K^n6YQQ4C#=UtX~#^Lyht
zv+{XvohFNCIqc%1W{hqAN;7)OXf6XwKTWUya5wc;W}n2+SXLyT(lYN@s*5!%3hh&^
zYOHvw=(w|mZ*L{2+!&JZCJvXpW>mavXKKg!gBY+Q0j}h@wTZ8c=?#jjOQ@%Kb3GOz
zcFkv;K+Lsub|ML>Y*uL`_D`%JlLUqdeNxXgXJ3J!E}!wh+QmT;rJaq(kq{fH;c2WW
z&)R1^4Ky!pMBME*HVdoJP0(P`+i&=^WniT2L-R2d9qNeFd_{D6X)N<uefs)SSACK<
z(^s5PhXTXTKdm@*8s8F$&7#K+EMvC)-j?7C!VVAG+ejSf3KN$b=ZYz4#}zc2Ju7nc
zfny0+p}{ox!G`hlEpfN{TQB|H^oJf-iscR<vQuGY)Q<kpwrIb@aphqP(~~Nfm7X)2
zR{Qa3ak3}EzO#_?zK*h(t~i{It$jTh;|PgiUP*adQ;|@b=Z~2}hxZnjlqtrFr^Ue&
zTTaHBF>hW&z)P``US2WDLH-UnzE$?|u6hQzRVqt*qNh&*WSKz}t!n4@NE8{qE6?R9
zXwUX_j^|`m7Ishj!m}_-8X`j)we-M^qo34YvY-s9H}?|Ud}`_s7k2bw6ApbUJDN$>
zlAY<Tg)&F<By4QDGamd%OGz+(R?sm__I%!g)bO6D{CV~auoW-c+9}0rG;O}imhdPK
ztID+YzP!Lu(Cx!XbGmTs5oXRjURp=bfbOz&P3A)*eHrbt6tF0UnvZ)F{~-sm#!X0R
zc8*?0W>SD56c}b>qs@uzIT^{Wgf3fq(;Lbn7Fv_k-Cz@O-?N1!Erv9^)sYTY(@vMu
z<B5%lp5@bL%m38H{fCHOCT52JnWnm=DH%iEg4jJ%b9w;dpQk~ChAx80kbxJg{K?QV
z7}7fmZQToasr~t&tes&@A8(o%0H2Q6ly1DFhHqswaki6?*}cizv*Yb1#H+CfQJKEo
zBiqX_@eLhshm}tAj{#!ggMMKLUWsZT$L!mj67kLG&38nNi;sxI2Pl;Ple+hA&Jx1t
z6;s86CY=TBjq<5qZTCXCb>^Upg=up4?E%G<>-+Y5g*z>Xc7=*C-d<IH*s{;b6f(zz
zznrg1(%dYhL?h<0QQ<H%y>Rx^L4xCJ_#nlt$H!n0r;<pFBKFY`bXkO12;^-%-D+9;
zmy%7s@0ni1MJyRfXW$XNjih-$sOAf!kJSb>UW^GXaTxe`tQ6Z`@<UGmOMU4#_#$z%
zBX{QGP6vRVo#hun@aOUqs8%)?1&YasDYTzU4Wxiz)d}&ubdm5A-)vJ=5z78Rq%X2}
ze!z;zH%>R_-Wc1E6Dhf6k3nmZTVLNzR1qmHyfyLY)!bMi$lyr+O{{w~(E^hX&^<Xz
z7m0R&_gFqqwfl0sIukipUpos|7)#W41)+D6w4BMzHCD2Ih}J(1GtoHmzTUPpfH=HI
ze}FuTm6CXp?_LA$UiEJgl`=53x4r9gtXk9~#dS?3mM>)Ln3s1I$+-MEjdZ8g`D?$c
z){V7gr=2HJ-fM~Ax6tvS<MY)>G|VOtpgYw%IokFlI|aND^Ge8*`Vk7BE|6QbfzSna
zzTT=Meyem+#Ak@<O4%Y%rS-i?@+JR$Wss2k+EP|NGsR(Uv`U((0T=}{4=L}xtLE&o
z$Da3vw<bGS2v1jua4-p&zA~nI_gM+2$RG1u!L=oF7bI?|6Q+QL$SUL@kCsstwU!fr
zh<Q2y>4VZF6ZG1ag90&~$Lyj+jgFM^xgaOq(v!%<83@5}`z5tfS8XfQ-jnmD^J%@*
zO?w#=S2jkXOTaCU;q#{aAZ|xi=Y%!pq_~p$3mtL26+?A#RTz3Q$MCh;t5D2gtv2+;
zY=e1XnG>AmMf59QW4DSqnn;DnRGi{8VC!5u98)zH1NFjKol#Dlqe1dtTo^hxE9ZFH
zZm`rpi_O8f9$YByaXcEs;>Jc2!LSHYE8-43<SX}Xa>1zdRU^zHG)p7@dH>vmhwlwx
zEQ70a2xG2PpFw_%jF2r81!$3T@vrWl=w$yrJsJ@BnDO<Gp355XWZ?^fBVXV5neTA4
zK&ETWB^t=tze$#Kmt-CJns`HNRkvukPX&#;mT}?^<LC2mt!nTenECK!_r3jOeWQ#R
zx((>N;e^(n-Cmg8U0pk0?YE)YHc;S}?{_I3w*kkA^XfC0?da(*vvxn}-CN%La#~c>
zgzhr$!c1)$*V8N7{GMawl18ELxTYX?s>Qo5=Fi|YoSDZmB$ryC0U=@ER-D9`$EEmY
z+FTcrNpOzI{EDByuN7Z9Ay|DGn3AqUh|q&NN^<f<SqUC+qCheplBG=xd%(%l$p;BK
zfo8|sOfLY-A)tj<xVm*`k9ev7+0&+3QerLPk9yG|DhX1A#s!knB3v28z|OI%8{*hw
z`3YNhcGK)Yo2)d5vJ&^t2KTbnN8Rs|{|=_XH+x)ZSU2Bm#5G$tY=Vt~Wah&a)~?3L
z&(f6ECX8_F=hQZ5WqP{D6t;kGowdD8^A4oN3gRyCJtm#V(E_i#K##ZT<ZVF_xwIXv
zG8mS(ugE#WX$dZ|TYi)xT`;iqu;DH){y@nv>Z@F-&_e-&OD_-pH@ksF!5f@v5XkeC
zB$JKqJt}n?1COG{EAtd2M4E#sAXLQB7jSoG3lGPbC-jQ=Kv<$D*nZBf`f_V6M9h(b
z1r~)isRctzU2Qe`CUPq@z`eFLO;~l}aj7_aXA>N>;3$XSH40zgWz}=((kgu&hPz<2
zrOQQ}C@sj|T6GG+9%HBwukX$0TU)!+JzYV`*5dbuLEN#m8Z{7jRYJO23RQ@vdoJi-
zqD)&LoF<j$x4>2e$frt7!64L8C$ej~?ZKf;2(zLNYb}y3b3B{Kx_WX=hd${2t?ueh
zmtu4=l62qJe8cbFTsXPu52p(^BKA}E6hG$>9g}#FWR*#T9vbg93?RmZtKk;8?kH(7
zf>`d1gB_dS{`N7tX8*d{r0;yR?*%kmSv%=7Rn}fqopQ{dOl2IyF^OS#6x>YOh9P!|
zsrBcrn$bBtu%daJ<q@dVz7?1iC$ZUHTYXtLIp$DxVAAv7(O<U6-c6>ZD4C*-;!7Ca
z!f7q1+FC9;T!PiX!qJeF?_3A$RyOOL4M!LLBp8k5{H6{k!Tm~(tXkuag3z}~dw!y9
zE1Y2Cf1Q?Y`<uP6`f0)^HyzmqQG4Z9AFdLM(RV{%BV;e0SaPM~+5ytmF&;>md}V%A
zW&ySx@agR*3UWA1MLpb@Ez99XXCO?|c`gL9w&h<oEqf(`bz$0O`5$@{2-;R#7W!Cs
z!;mFY^Cj9B-zEG%ojq_1FkiLPHy;<d`)^CN)8AA$G>wmHRr!Md27Qup+_<yTfnflj
zDn8E)R+x-~j@PQ|BJcWRuhrg8y6KXg@-6@lCbjaqpcq2>i7RXB0Ws77*Gpl0mq8E8
z#VQ0p@ZCNJNY?TSFuSIL#7Xrt8JuROPh@pHz+I!|(7Q!|abn~24@lFe0=arr7Gk0y
zA7fRmt7i}qd1+1m^72^*weW|6m!y_hYTp<^j*3E11Nn_dQM=lrhv?fcztwQ)!@g?k
zW~B*FzTOL*tZreBTsyYbqK8DZsjmGkfK5bxA7HTp7f-f1G|sR-upYPImwUvL%LlKw
z8Q)*aJb+0lG5S~{B+ot_9uxL2nx8vT=1hcB-S0r5{c#ncLK`!y51+K&*iy6TL`29J
zu9>lI|7%nIm!@+Xa`)JiObHUAyFH=E9iN7}^O;UfX<fIt@wi>ovGWvtU0&~n^ktIz
zH;ENWk8Zx)Ezj2Og;k>1i(>CYZ7BJJ?^-a^K)$=+dV8LYWvc>af_JAoeg3)tb~zVW
zH7&^bou-Sm<HSJa%679`@hDBfspQ(9Z|SeMXU_+I0U?OUf9j>!{||-qzX2QUzs~Ld
zb<vAiSUa0I(u-LeIGc!=7}*({@bN)8IXjvd*g*YV@6zP9`%UA%{-P$+feSTJBzgjf
z1ifr}`E_ZxLMwr`si+-(`yxu^7rI{=UFx78q%HmGo^d<kmX#sC;JjX4o<1F(9O^io
zP$1hdc01s^=;=TH$713E;0n|Ix6eke4~kX?uO7$HP0h*lGc$fSjU6ort>6fHpJ@7S
zH;D5X*G96KRyvqQZ7FO2HJp~X?(VfHM-CZBZwuV=adGv+O*_tqlh0RF_E4(sx{ydp
zU?oQ2mW}8x=4|r380_#^{p@m-hF%i4BFJWSXAfcT=t^`{yKxpI9-e4S{x8O|QMy>0
zqN&?`SRJ)@GSP#Jgg8Ln@e>Z>&F@)KJuSkGW<yIfMJC%;&#Xsx=lZ+DO8P`&bf^zg
zA=LKuV?@WpuQ%a+RH_t$<MY$()#qhIyZrL(GaHu)Ta#PW$I$!$MR9kn^L}_+KO@2K
ztm(C6!K&Fj`!9*<AOVV``;N+`a%j~{em}AuIM=n`AI9u7V?oimMPeHa(KyAF5T^53
z?}Bl`xXX*yMj57}q|8YHCf!E@y%FRqnfm@kb8aI%^lf$V)WwFW;3}O7Zu)t+t{I_`
zuG6D+uJYwE!w+12B?3UGGxg3FzKS;<#UtSH6FncAxtYaLF?fxuLyQPoc*=dDt^1}U
zV$e7ZnOV)!Ne!PPMUwSMHp$~-UB%&AgYv-j5%k<qxCb5=77~6fqd1YRp9(N}Z|U2%
zCONU-sAs$?TD+9iOIhwx?)5@W2XJQ13rFt^=X>O=eU*|Wo#$Tt9iWa=s_r|MdGMh<
zW<ok|1CXXkL^g(Q^<VzRmZ(i|4y1$Xj@O4rW@veI<Xie)%engRY~M`Nl8#iRaejCl
zozP|i1FP6!ghUd@TQ0_qY`Iy}nu{5yY$-w#oKujkXz6K}kPUyBLR{WN(8aYa(R_Su
zWE+b6{xdoETVYUE2RVZ#U<LOQtkPkpj8_Y6?HEVIP^drjal#Jv)|2-d08(i-OyMDC
zL0=vL5G(rlA*5DZQ{+?xv(0DfShT)^H*?+raluIICHsDTU`R5OzFz+FaF*RzFXBN4
zaiy#9rAP!-eh-mrezX_+!B&l8Qd-OXW%rGUG`e?-tfm`P)_u3WXlew*qS*9w7I#=4
z{=DUiQeB(jNO(dngqg<{loYMI36B?2qUK10>rdOZEwsEhKv1RZ<Wj7|P};Q(Ljeei
z9tc&Y3Pc|tR3kb?DXn_WyCbQ(TFU0Jq+=b2k)I!wN7YEV#kqBSYZeqBG|v0(iIqIF
znKu<iI->O%TwRl%UvmPV4Q~i1^RpKj;5f~mWG`Qh4N%8WbTzUt7|vFcL{Urth1Ys-
z+bUu7i$!|kyfi(I@ivEa-IHa$73Yq_s=gI9ptBOPA(f961xP|`c8&fRTCHwlHqiOM
z@B^#j?MxlI@Q(6&>>heE07FG;agaYE4=Eh^tz0aL4^^f?3eVvrLpW2J)IkGK2{;Z_
zr>HcphruM~O?xRTGz2gw4ciAz`iuvtax9$U{=xb6e4D8tYN*_j{>ojpqW;6N{<j)W
zfpuZ4H9c!pstcb|H~hUp2nG4BDpsI_9&A$_u`?-58h(1&(AuEmIFvI*<0_a&Gucl%
zKv=<P^rvmNQl*&FLAKIM3{`}5+#HvvgRSH0q}CWY-=j6mMQVAmwWT)a7HOaD6i4Q9
z5BB;$JVzLP6m}lpXoonkLD5&v$j!qB5+@Wsc)>j6J>0lKQYBcNKZKN7<g_~^Ow4fj
zQ8M;oXWcv~D3j7iMwqE@D{<=NC6!T5Vbmta%j5V!#)H|CY#vRQA)}welS~&jYSEie
zmbKKK;c_Rk9?RLKsHd%c*^~X*Ljkx1vWaF{jb0tS#|s+lL}D#CRAy*a5~`ZqwTs}>
z0xo=Wl6v2|Ejg#^az>q5GvFKzGhkt(JOiG5DX;f{VK{NtpKT(V{Ek9!W*2B^o7WFA
z*idoR7T75Qq`C`i(`6FRN*jL_#OR4dVvJ%)Ha#wvl9B7oTIcU;S6%W7rE4cd3heXN
zr|uK8D_ANTFh6qQD`K>U&DG<EML-v$s$C-2UXc=y89tOq^kd-#;FX*}IAKuHdRWe6
zb428`piVdq+o+6|YsyFlte6-H=jNA0-u7?EDH`@l(}Gq?tz3mowPz4P{UXTKy~46=
zc0ZC=%R6Q?47*GPKkWrDBt&o4<V{JAaOGrfFYNTBumW580xwv=7YKX2*3LV{At%p2
z^&rbAn45h?3$7cIN}<OReM#m*9OYxUOpb%`>2bxqwd>k_WaW<?oHqkV`KDuT3byfQ
z4FgUS&rt6^l7Zg8r2V=Rt!JI2W`#x<fi5&Ip5V!GsI2S<)&@VP!`^-hZ(U{oQyJy>
zzl9~nU*q2YsEi(Ia>x8H)`=R04&12YTk0M#1I%j21@Qb)txhb}Xi+=S%TEN%4hrd7
ze3dk^zKI2;qv@+dHp<)WOK-@G{7FnnN`!wOLqz}R-oXXIDQJmaUzy^8exSZ)*xtd$
z?T4RJ*%{XYXePC2N}idt8xJ3~L0O1E`)L)zh#7EyOdg-qaGlf$W4OndT03-mju&^7
zo3@jD-<+sTKhDqV>r)uHfgWA7UV;j?{(mtJm4GCJ2gANO5ZGDn;7IVpS%7g#ALsKT
zuk?XWQX){*IEk2GQ8U=FuIhH<>Jhjl8GD4{J8GGXDCN*&3K_s-9ux^?fjx?;poAun
z1We>SwZHz$cJpf!=9f3AR$^SVo$Y55wf}s(8tVZNl!?MFG?weV{i@l!Sg_+=YPA32
zHP;bHwnFj=tXzbdgVO|BaNe<_JrUJf{52UgCi^KXb4CO1xobf9R>2|O69nPC2%#6|
z>;9Fnvnc}YPC-4-JenaMfau9&$kwrZq7@L?rJaZFl5xMF%uEO@E>baho(F+SDz3Wk
z3q>1`WBHr7Jul3kItu+}0sk32ldU|2&}?w(Vo*OG+$e`?^RS4pe;FXc!tE~cS-xqJ
ztxF!Ea<WNmkFrLN_Sirtr9>k+4&kg3Wchmi$7%lk=pi3I9w$v1d96rlRUny$FT|xw
z1=9AkM}!H5IWOA=cGBlzHd&Iv>B0Por&3Z3o2gqcWo1%4E+=1~MdPe7x(d!LN;>-f
zNXYgyJw#xFWrV0@rQh0&r)_IOR!_GSki4aQ-F$;c!G(L2d#Sq8o9-<@x)Uf$Srk^z
zohAzLy_SSI@77hXH5NyL*94yqrHf5<Cloz$;n#W1Q@&VwBppsjh_4<=D1jJJOWB^0
z@06-h4@Jp!r4cGsns(pJR*X>_?;^SzUkgNU_6@loqbczzv=$yK>6}Q4KlaCidpgmt
zyCr>85VkQp)0A>ML8gVUD~Oz9iej-QWTApd5#16YTV9{b^34(Zo^c^+yJ}x-E%4H8
zm$0Anly1aqN)$@%-1{M4$|RM57_R|&)~~^j00M0O5o8CCK;l@U*A3)}#sdi_t^Rs$
zqxupB4FljhG$L7syM>relR3nalunFcjNG+PD`3FbXld7R0yQ9En34GXNDgxd%a=6E
zcozCu9ZrWA%J*uQwDVj(^|Z9;$#pG3-T3g0symY#-F`aQ&B8FL>iN22TK}YEF`AWF
zUCU<AvywxYU-2V^kL%9<ZhQ`BiLnqYHQ3C&UH_~@Dg{b?!2`)<$zsTnLnibc@y=QY
zlq0v)a@4VOFiEL7+BY+p%;6^8P-X)DX9?wDDeYonsKg@C3rp`isvvfb!GHHek8h|D
zO5Z0mkI<0kutZ*~$&FEja@;luM}V5lWo!+hV)|GD2Uv0weJ|zjOU&j9Y$DYiF#A1H
z&Px1jl883I_jQi3D=KFbZxyDV_?F!!CS0P7!zu{)vlF7%f;QQ~f_64)TacE}Zot3&
z7_h|r9-vK-QXd3`FNj<ZA!C?6xUy1vLc(?WrW;Bj+H8SsDJ~-iwkh%ecGz1Ii0iPV
z%y=>#D%CwF58Z^N`U?YT-p%)edicancPM==?<bO?_3EK<OigXKk`%Ahjwkc&vjjX-
z+t|E>iq?bgO;%KXsjV|UyyNI_OTJZ>7u>dhLK%;(_KUAJ>OPorMW7zOc-~1JN7Kx;
zLGO_^rxEK0hgRBu(XBTa?1ikASo?4W1;neOQ7+i2-jAV*MxBI+1tG5P?MMr*QX7s{
zg(CX-=x%&Ml@$!iVO(T?<;^upoeT4%v?H45A#)J)TI}kMim;SX41Fpj66iRWUUap1
z=zXulf?Qx$MluzX{<z9^X?6dRv9dQA?j7#htbIu3?;*Kp`y<x$prs)e73^%Xh<BvY
ztXL&k@MjBeop>*fnzn9fI(O>fBcK5*Vl;GH+vb}6r0Y2tYhw_8C(WpAO6Lm&v1j2~
zX=L3?@K&bY3zuUBk|qLCvO3lnQcHgNrAAfH#gk4{_aEFphlma#cBWOL#V@uEY->{Y
zmZ{+v7kZM|%~srb%PmcA+mO#ki=-BFG;`V%*HQReG|HJ+DaQEv(&>-o`TVW^byf6A
zyfA6vqhl{GX8EqPl5AW~t7avpyeaQD8}0;~wzKm%W(<LCusWpU6O#Sn@1J>8_s9PC
z9@_`OD7?xixU1MCzEX{}Chc{?ui+cs5x+ppk=F&p1@XsU;ZQJf)qhG5A*}sF1%D^x
zsgAO|{ByIr2z2^KRxBhbAalW(Zq|1D;dxbTl-R>^))A%V$lI+nBz}@-BBR5=q`SmX
zop5fPB{>FGl=JIO5Mo<OmNDb&+Tm#w52KHUQoY6h-s?#n3`1JifXPy2Ws1}o)CeGN
znt>~nZH}^vLSb2It++t(DMN<KD^ZoRBP1Q24xjTpT63veqvjo$L?d9rV0vZVY$)Ar
z@m{fEwLJZ9_VD%!tGlUlpswz=@qKl*f3Gssl=Dr3te7UPTgb+RBEL;$ZDccpIwWNr
z3y4Y4%|n6RHHs6<nro6O(5T+_{iO~}j<t9(-DD~6Prgv{ZZ)3fR=-W3k)~tUy6a~8
zK-oO5;t_FhzA9RktsARHcaJa67c^`rnD>8Z4A>d|!~UA}mumSx*<UYdO51I6!2D_z
zxA>t{N`-j6`u}8pBb!@b9knhk>aNN{u73DwNGIxPqE6yCC_Qdh<4nRoUrj4QZUXSS
zucPTZ<jI8tl(YNxZtraX-1Cmm8)-E0aQ@YU1*hN-=y_Jnjk_d9G_r}&k<aK}d;MkD
zH36^vO~~B6-G$g^uOHcct*Sd<7jxZ9*c7~uanR|8Uq6>193{2dh4=0GtXoLasPd1B
zxm*5zIx_k7&DKVX)xidsln4ym&cx_=3HFY+GZ2&muzY^Hy*7U|6mexOZ6Yh+fl%1c
zJ^)cPhOfadW0_=93xwIz^w&X110nE6yB#aOHj?>U0wxSlloAnY`1ttXrzf4%G(=8Q
zLeFMH<s3BqtMb~b0=ityT=h@mMOvmjms}+-Up`FRf3Vo(j&KOIT$a}ecluYIzE(&L
zGp>=oisVGJZk7xpC|8La@5d;j>~u#fD~7&iG8ww@32T-dK~}sa1Zlv3p^x(UueGMP
z_RgJ)a&F+>d-2S{!*}QH`uutT_>vj^P1-|d)D-%X0VPK%IB$qSYYFRqm}@Ayy(YXc
zOj%c=&2k|G!5u9nhTOFsR3)-u%tWXCE7C(KfEerOu*6s$d7mTAd~YbK;>=6yS#Vp+
znN}~79GCeA^9EIyv$kJha@JJv%i6Kvi<~JH7H`cnl8!2c-CAtrN;Rjax*UOn0$0ZA
zbC7;d^wP>sZ-LFSA~Nvd2=Kiz4TQqIzt#p$pZnt$O=)m<=!P=rX<UIkD977H*h1#9
zb9P;uZBm|mj1nD)jJu{0LYaGs@z?wnIE7O(#5)ageDRhLNH4Z=EDt^|DQM5>ONHyC
z`o5Prjz&GhSpw!wS;92In>Z?%Bd$Zz=O`p(UaLDcFEp0=JK3eMhanpeCjVk)2w$dP
zNm0O;6BICL8P+~!yKu*-3Z_{YNS{`65vx12prEvO_1AKZ{;&@Q0bZ(vPBq)E*UOzM
zJvD<a35|Lb67|wp;%R~NQ7#0$f>8V$iD<Br*>F@N-Qx%-SRz-VUY&X>bIYzy*jQ;+
zu)ZmsbdL*Kt9>edUuWp5jtZ7ke~MqSc}VI?HGxKRQ`*R3)kuZ<+77PU>P`}X(I+%0
z;fXEH4~NOo`Fg5olQ&3u40jj3j6R7-`Hv0v!nB9>RaYp8%L(*(Njb@51286g0lp5Q
zMTw^L$EJ$0&hG~{CNci%&t(c7Zb0Sg4PbXf<^ls9tjfeEVQ{Dwj3-|t$08JNRM9KX
zdGQu6d^7!b*b}vo^F1|sZm?6Y=vH=-dL;$K_VUs&CHnh1F__{`=dUeB;DV)7*lAJ{
zSw8BSkQ=7E^psj^MM`N;>&Q8KWBjmx`Uhqr9^%+lN|q{Ai0dWt{YF_yfd}hZ0d2`6
zjY_`R!F4fx=b^EAB?#&cgA}zT?Vu_SNwU$Z6fcm*ld5t{VnF;NXH%ep0tPSArO$Nn
zeM@K4A3Jq#_v}4L_{p?x1<J0*8>z)`U?}uEVq!t8GDObfbfWAW8>{*j=U=D^DumiG
z4yUJ16F1bjRB8DW<P*Ri*a83&|NP>-M;5*gWMyIIjT}{+)5oVd3>8t2fe(vg%B?#1
ztVmtCSZsx|KFhhKt&23Bh$K|<t|pz=(ik-wsbb=KKt*QAUT^WaG^Kd6v7@`v{W=Q8
zoR;Lj_gs`voW#ZCnw8xP@X|I!D~?tc0a#skJFaZ@wA^k7ukg|FklT%DdG4+h_N-?+
z?0i=qJ$Y&+GfJpULwB3V^TK_>1)nA+T@l<1l22N(JOsmmi5_dT^-R6+1`7fYGS$MT
zZ9`iuTCA<?)+24j_b-In1Pdd~&Lrh+Wtco1FLE(<^et5yBSoMa&vDcr+Ck6LMdhi4
zaIbG2W0GW$4eqU|0e$<z-j5d?FlfEye)>$%0D?mpTaHCb3R@0nTME52Q1%Xfb2ONF
zQTJT<>|ak{w*%UyLWc>O*AK%W2k5kOrf|thN2$~XYT=}9Y4neS5S^3HBvIUKFx*fT
zt#CLBJzZyB2`y!J`=sv%1t*8!rb~kHjea^M@K^p?9jQs|wRzM=$Kd4;Ajt$@=zpp<
zOpO0G`GNC)sEl|0Mou66Moyooy~0<DkvxHb0me_7`fazTYRxInZ7txf|M*gnUYM-O
zUCI=L$28Q7D0z7il_d?#2*T~aoxBW8BrWO(E$BP^{b2;~jr~L?fyd(gMQ%a_zvvC%
z*C^%holM5W$A{ui^x*b9-9i7RP3OMt)c}mps|u<QI6;gw84t4#Hr)2x?Y>Y(K6uba
z!JzQ)m)!DwzkP2Ak{Re#>u$#{s*|R=CMO0Zq_<HW7pn`oL@nx8!`Uj$X=40&`50E&
zl%Tyt-vv8Dm9?C}DvZ8A!*YdTldz1*Mg5trM35y`RMp$;yd}Hj@9cK@d4KnLJyB>S
zfVBD4?}gl_{eHSY;G@tlUWO0t55W5V@_co8?J7w}+3YEB6S0vT=$C&E6f?g93y{L_
zsX5OCU)4Ij;J{-m8EC0hOxw0Q(4GROsx-Nvc7cg5nB?H4TPpDRCrRo4X=t38hs2zL
zb|G>aKP(s4D!Hyg<NNk-Gn0q?_FYARlJ(0G+#%`=GTTp8s@*P%^#ePPJ2yV`bl6TK
zn!`3&v*V&GU8t^WRXcL^g|{+2fH283^%Poj2=%w(v?X**joe&?2OR;}atc*lG@_tf
z33>{z49DdPoJx%Eq!|*;YeMp;<Y3uQakuP~xgZvkcPwBReR1YwSwmc)haF3h!#vG@
z5W1~ulq%97Q-)~B2ldt{xUZqCO<aQpOpMiUMxOv<G@(uzJtpgqLjhT&-9%58W^_mX
zL{LCzi?%TCeGNfS!Cxb17OC%s)MxKLN-{$xRZ}`q9J1gY5Y*k!cYH6WJF6%o9#cqZ
zziMr|;Aq(Dt`C|h({O+*Sd~&H>6ewFWyXGqu*B_=)2<%R)eNH7N<$?WH7dD1HQ{K7
z>}|hkKA)9L0_?pAjusW}z%2r~NpG;vCj2K~R(5IiU89KM+yV4&jC!l)^ATdVQN(bD
zHAFx(MPdexWEBI&q6^z5tu9F(_k<x6{F7szCI8jBF&ovXjTh3+%mc5D#Ac67Ehqo0
zJ~=djWJcS=R&mB8CCF_~2^b93?j}gL-!O~itbx{v0_~9<6JtKX@8FSz_Q3*W;UbHV
za*3?Uk_qT?@?eE1{I&4O!%QSE*ym`v()wx96{Wr;`meKxlC(I5!3Yl+MpW4T15iLF
z!|<cy6GB8qhTKjKr0I^Bb>)M2s7>;0g5l&{ZsuCsAjVOoW(U}wu@jc|F=O9p3JU*N
zp{2}HC)p8m0uA+jrU)Z#5lCNv0G%go3Y#kGN(WieZ>?nWi%$=2%1>YEWEzlF0WBg2
z?yfp<%r8zL!UHFTXJyc<&b--{GAc3@PL^U{+<DJLp;m@;w}A`DA|-LilNx>|C5c5H
z->M)Sks?fiZgL$Z-ci<R)8;#DSK>tJ)sSgE*E-;%<P}`iX<Q87v{yPe)2935yR@7U
zj%p9nxQaX|EgVTz2g}u7K`@ZJ)cWNOS{0t1#8_0Cl7Zt%Wp0n4MTL7*y-w4>OJ2=W
zk<qoRq`q_&4V8n7Me~FMh;}3lDwLNs)<Wo+SZ8;~v+)2ZBFp#$vLR0?wmT4b<C6|L
z@<|RTTLzbGop|BQCv%b0x-%wmLGSKs-WEuCOZF7sH1LW9;3d<b#60NDb}p0V_5U#T
zjlrR`>z1)??AW$#+s=+{+qP}nHg;?$JGO02&bf1ErfR;aIbB`dzx&U-dOgp=FDKFm
zK}XhE7@4J;%*ln17;SbuUhE`KTD)aBvA7InVsM&b6rsb4ZDzo1AOOdC#|U@fNol7I
zByop~7B-GA=#2+?o*7)QAcYxjV9p4+pU1fCw4cq>IwnHygwNlRxW1V-`nt!rvwB~F
zns6D{(L7XBC?-VJRZFs@3<~>&ImaGA$c<qV*@n43QFlk=7ex1w!n^y8cS?4^0i2N!
zaVqWLhX50WGt19K7pp7bGALU7`cl?0>Xdu?5Rjkd;p0xGTa$pXsN<XYRP12ea`-s-
zU$hV%RszS=zs=QJYp-VC#a~6c<VIL^4han9Mz39T4y)}nPvzBJoIr6(XFAq}L<Sng
zKKax<%ZdS7S4d+wCx&n-mZz02L{s<Hi-CL;mscFNx>d^!gww%@ABtdd(T}XZH)^&w
zCe;ecH}#!uJOeY|SfE|Sn>)E|6)$>d-lE#TZG)<oE|K?et#0ZwGKyQ&!=_gk&FUN4
zbEPgbXfOb<f}D>?!MF#UhjTf6>&0<az~E=`SPp%7oWe~iYH)+Wv2C-t?C3?+Kb@=q
zk_N`WFVPYb%*p8fJ~SlZciTneq<doxqgczi@51#Gat=$jY1OCs>!Q2vX4}=^KD*D6
z;CxzhT&mJ;VwsL=$oXf~ikZ<X@XS8SUtn5om1nUjCnzqReym?^(E+>{9gp`S>ytb-
z<I8vCa4}<9(3vzeo_=D|Vr#p%%1{FB)eYD<R3SL?C^zNkV(f)1bE4^CUUp6BJdySH
zH&tvuXxf{+oZK+C*(sIg?m%!Xmw>1#!!1Jr&pP_{5aAP*Zj&N2!T=E&ktIr$=SXfw
zI9W=pg+$9tw8H5r8bjAQrb!@7rkAq0EPO|tZhLurfXlKXDgOg@vobONZ|wemc~yVv
z9_jy=S(T0cKQpUtYFgX=YDN62(G$qin}^>^xEF+r?DIojF^2-Br$#rkgc#P~B$GwW
zrpyL9-abz6-tg?tzDRUx$pU&ZcS)6k3%$SN)?QA~63<7I^l$5k3tYSa3QkxAg-gZ;
zk?p7Kj1!U~scXkSawTs{u_Kcj4L=2_RS{RYk;hjc5YG@g+M`2Qm$)Q>Z9niCLUKUG
zOR%4LE7b-%T$Ex#2yZoX)B!c*r<p&nfPQ$!6X%f6j+XLYzn}n@Z7ij+tLdhdN7o6R
zZxB%IB}+qL@kKvtup7VteMYg^DF>k)WARJSUfU?o%x|041a3Xbr~n=N8#@<ZXP=mi
zBAjuq2m)53t0+!_lXUh>{fZ5G#CDrg2`1Cc5*1+bX#bnD*hjP&6!oc;1@uu$oa3Y_
z^~`dFSFcKKESMoy6PWt=dZ4Z@dXKDK`OTB2kB~m5A26b#*OESl)k!3zv^ckRm*N37
z4f<;7Z!c9`({C6jJ*@T;*hJY+>M0WEcTCRFBQ!?x>4S`dV@_<?D<#xu5?C95&(ESc
zbGoI)SpS-er&P0f#*SJ$JgGHnzv0Heo=OY)iyO7c`r10^)PCkyVoWa$56_x)A>{?V
zD4NrCD#L_qccp1UW;Yh8?99}GQY=kpHhRB%RXv0CJ4_CqXT6#9l3;o*XcGBQ#UoDD
zhBRc$`|B$@xssk*=g7$td?+sfmESvYrFzH_5E05Jv0f~_f2pb?I6B8xK`4f)7yPw@
zt-tOsq*iJPtnIzh{dD2D%F)%9F3Ep4ad2ks@yJGaJ1u(GdiH*%p}pCXcy{pO^v(`1
zFo39LX#lTQv`WKSCd-%~6L9sqH#x(>?e2WBeRi7rJbl<aDB!yKy1n!K{$7$>I6ZvS
z>$n<mUC6cB8d?~m+nizR)yB<<_PzY_xgU&txp#j)^13U*o@<LjGUlwbEsmPZXky}2
zV&F>rP-Pmh=H6SldOK@+cXoQTdH5N8MX!82fpBf?DwG)c`rI7=d|z>Bm_#0-Wyz(8
z5H<2X8Hw3$f0=%K-C3}%6s1|%Y?ySE96gyIuZi$kTcnZh(O5>0Efl1Xr@7FATK<*o
zC<WZabKREFmFY0ut;_;RhpgT)9{pP(4QjzH5ABcU-H~gVuWcww*r@`oI7BX}IO?c0
zpyJ(zMou>8Xm*6g6Baj5M6lDSZZ)9Q#J{R0m{nNEQbcUj6O1;aOXP{^;k)*EulEh7
zD;<wSXEQw`*TTx59HlGRB>n)>C?qALAYaM5h2%`*p|GBBd_+zjFp-k~?&84P&e`)T
zI|le2I^UBaLv}<B-QOhD*PHM2%8fH4Mkr^~jki6`AnF)bbWD+N+|lT%%G}3aJ0pDK
z0_ZVHvEdc8A)O^dukWRLnl7&MmUxC%UMGHGkuHb(R99O@Hg;rTWi!%rrc-VIc_9&1
z&5hyBK1xDcdIWvJl5V3hUN@$ej7_a{|6F7ES&J`fW-GcUz3W=1$ZgujDd4&GhIITV
z!L-_Ar6XrI&F3X>g(u;*{2A4fHmSXD5J&fKd7ysa>`_3Gz-S<U;H=j7fdK6^W=@{>
zRE+Mkztqkdpa?Ep1Ics{ak-+zRMb>aaZ{+7Sd2i87I}U$VJo~@f7t`%k<g=X&p_=F
zkC{N3!QLtuE_|N?H77Se5I<DHdr20TvWgKVFp7q<3PoK2k4t`fZ#dx~s7T2;ZiF<L
zzWz%+nX=ftG7I6by^zml%T^vF{}bd1CdZiH$~^8eSEMjI&Tsi+GTjyAJmgZ{QW+9}
zx%h3&7idh}{?w=4w?L#a@?o01wxET0O%R}M<pGTp4_XQSS;~Ndwopzrk^#W}!W-}~
zvQ}}wl_8|`7Pp=OncpDn)%>uaC(S(MztI>7jC`uV+kgQ`{UD?hr9jd*7J*#iqJ)El
z`@;CC-swRxG_3%}0@2oyzUHen`<Mh0q!WyWsN80CvkC~1{a^s)go7ZonZ-(FV}y{r
zYi{_l4C+?)3=K7Q${+xDQo<Bk`uZ-a#OIRwSp{SlI3MCW0%-KLt}T$itC}+kv49?=
znxrPAKi9p1sv5xLRP+8|MYL00k$3rXN%p{BksF6|8TH{X=E2~|hrp2wfhXkykCUJ7
z${Xy`-&0<3O!3MOb5NwCY_E_nk^OEGt5taf!78(RxGYhW|IMmQP5~^`XRQF3+!tn!
z7Y;)q<Rw)?QGw|fWCub4K>~U9W^$q~v6EOMMuI>th~?gflocnV138HuHuYSAAQFaH
z^tL=+D#jG`*SLp*Vb%mdYWlH$ZgTkbFo6|#X0ZjGJ@8jJ{X|0Hbx)lB)pWYT0Ds<9
z3iY+2)naJS1@mf$it_wO%fq*gsHv0^6CUo^xNKDEjUe9dK`Pi3FBIkcRazS4sQB~a
zYA{P9=jhz!oP`#eMc3?0TF5J%2?-kP!=iET@$k8ZtfYM9pL@|iZu!4%>!1fX5OEyO
zxdZ@I;`#B|%hA%riER_!Ecn<qUMn^Q3}QzTWf7#b+0<r7D0+7{;e==5hoMUUP?}k=
z7pqBC1swhcqed81-~cdCkXF~!z_+J5%A~k>=Tk*pGzq-T;+h!`p~~@H#4azq5ZWCZ
zT&~g2Jm^vKJUDT-X+vgP2X>PavSfA@<Z9>ln#-#Yu*rgC3q8%MSiK8yCeQ5B$8{xV
zbk2rl3su!VSHL{C)D)KCb(d2mf9#B_`M8}P=TzEY#vQfI;^J%9g9ag~hNBG-E@qI?
zpo(1A1-EKA=$t-~d+6^n>F0{_mUb4aPAk|{4lwn_#Xl1n7TmR|tadC5p$l{l9t2<Z
za4L&Lc=*aD+egYLIgmRwoG3<kQ_wgVx+i$Ywn~Id;tG%6&<FyJXP)Trl0rS%DMM)P
z5Iq6@8>tCljPf^DwH_p>8lMm_EghC<>&r+|1jeZ;>GkU7sdJ7s43?k!Bd#*oX={Ar
z+dMs=W&MOb8wktRs{!VNXwuj%`s;p`8^6?6lE4Vyfnb5I4PE~j4BX&9D||q}EGJ08
zfigo9bRJ5<+jSGNLwYm`5)OEZBVgLKQYQmg+y;WQQshrAQUYktTcA?STJcFlIOhX5
zrkZzeA|y*1NGX$RjG0uN3UEM$M31GE{;0nS*Hc|QfF5@cdJEurc;ffZb|FQ<;y*-i
za&`=+&)x9Iy>N}zdW~O9e?3*pX$L+*Sljpm<+G?`t-%Ll@TRCd+gtd1N7|MLEv;U-
zkTA}3W-1e~=CL+5q0?ZM*htY}z~q0I*k_^*1Ax}~h_eTOQiqS?wg`N7LFg~%1;3yP
zz51b%2dJXvvHFKH)|7pH-xfT|*Jkv`o^ev%DmmeiiJw;=c?p{8A-bADt{0BEn1ze@
zOD)jSB#PQ4@Qt(OagNYe0uZ;p10ZPn0N*(VGVR2<BN!n#A&ldak9<^NAh=Yuuv3<-
z%g^#YE7Z3m4NBB&CBGDr4Wk^wl{?uhPNsX$d9)@PJiq@1B5rPQJuCZz>wE2F?-oRz
zKYH{)&m51<UV|S0N-dz)XFq|?BsEN|vW!eo48PF11cq1Qs-<O9?7E@pAlgELsrdyb
zzQj-e55K)^%>NKp%E8R^A1z3_H6?8KSyBF73z;T#$792<&+LMh*9zl6<v|``;*o}u
zeVXsv8_Ojk8I!mQnMYnQ+DeC4esj7pRQf*kWq*GNXU8~6`TOAEg&WeN|Fm_7xopWO
zNFQtH;f3+jWw?FO>GD)?8t{bip|!Yk-C96Bg8)V6K{af1Iq$~O3~say%}E=WSqk@r
z_QdCM#Jqby!xJi|#?trF$M?f|FtJXA5mMTm%{(y5%y>*3NVtPYqA)D*&b1kukz%))
z>5C?$ls{F%xH^knsDqk3%?GK}{^6WQV=O|1%G_x5cU&-8H2fCaoTfMyRko3m0+RGh
zKa}fS!B)n2@w2OLZ1!ltdog3#5|+tSuyDBlnK3js4<bYA$OhHKzI9m1>Lh5&d?cdk
zU7ybs>47HcW?!m$&{EpzHYjClY@uPWVTSh2L&GaWe9E(ueDW`_>59|dF?i@}oa*L$
zt~`WkF5COVT-%vm?oR>?nwOQ@9pnU|;Kl_h50E2qYc_AWnG8TobkhD`A_TH#=?5tr
z4RtWwGS~(~2t`&U-p_@7Q6$i;k->yG@`A>*Fvgfb!;Bd(-nQV8%E2NC9bvn<+{sPL
zeDiNz<9u(G-qT9vOL_}O?(5tro}77j9hQTS2^`_&^vcj!iAMzGm@7T9AgAU5q++yi
z(~FbH5?VOY#OaQI+RE!FGmhd#H@Jn17=yW$56=FmHC>}bKZOaDr#T?wH=20~M;OhU
zEx$c4NNGMrqiwQ1i<?mOF$g5JXc8#RyKCluBuJ$~NoC}i9uGKl11DGdM@U44I>jC`
zI!-U;%hbh?hu_e%5Q3?+s;q=6O$o^=Ldns~=}lla`spJ%F-%CRjbjz9wNsTVD``{2
z<!k0ou~0h!Z(>D`&*`{(il19u8RToOX%pO=X?Tk}-^rglVRVjh1H<weZpZA%Z-fry
ztiavnRGR{-5R)0%4)&O4mz`5YkD}9EfyD?;s?vd4z(xe~S1X_cKVF;jQZ*-9FUZ9h
zYhUqlFGeA0ZXq5BgN}Jqmd`9SiXZjQ&#MNq*~_O?c*>k2k&NxLXlh4TWD#IdI1-;C
z^>R;fCN*;^3}ao(7)=|%E?wx1Wv@3TW7GE&6jFMsv@?I`B(%nNPKc|)Y7*+SC<ANS
zx%YI1CkbW^$dXqPFsrL;HPE9W3k8IvA!*`)(g(aT8OVV!xo{BAFyGAp?aK40I!m4d
zUq!l2S9)osCYZ+#$0*16vvkYX+o2?$VU?COmT~3RHdszNXh0P}twvTUA_IvnQSQzH
zXm-i)R9Gyd?;=UM%;xtg;X02AEJ|#c6r2+~vZOnN{}u-0<_0=+%*2n`XH1zMW*fox
zjbRYMCK5ziL)N&WPS+)mjltWnpcs%Un^2vwh2j>`O9~{BR&tjFr0w)qac6S2fs8I!
zTroLCk^0Kys9EHwu$_MVReLSj4wN|0_OOXkH>@J->G`_h{Va^LHxhSl<_Ey~a(1N$
zav}|&h?IcB{`ukPvBmwMoGA3>c>>oT0iu<h;i;q4NT6-|OJs8Lbnwf(bxFHx(|@s>
zM|Hec&~93^NA)l9SUEqAGi^-XGPR8f!}0zaFg-56ATcNQtLC1=tcIklj=Hn7gV0dY
zDASbDrVctF5ZN%%41S+g(UPRnxpR%-s>-OO+gNe!qb~7{5Zy7y(Z1qwi`!oXa}$gu
z>&>YVxUbbfR-48@K3Jcds@hX-Cl1>2VvAM&g2BMW5_lP~Efii;(i4;`YxpEff(edw
zS9dAjEPDP7(rNzd6yWHnt0$FT6u<g@t@A4FTx^Q?v0@DuVH?y$qrpQhh9B;tR#MDP
zlCL>->VpfPo)mK$f(yq;Loc}Y?W8V?wOsifN-F<ob0~e{5$ht*_%Gxvvh-OJ6|?5q
zx5>;<EaIF6W>KxeLT)N9LAsTDAKPQz64z<S2c;Wld=neNfqQC={gf9=CFLcv0Jkwq
zL7Jaw7Gr|ht`f#yEV_nB5e!ZIhQyk2WpV7-+h-#^yLPL*R?mkA+*?clmC=eXFR_{!
zP7}y1z2{$nOXCiKN0O(A+*T<qdh}d3pl-*H<`@m@O9M*Qie74o*lsn=<15Rj{Ur>+
z&sMDoQ;qNaWEHNSDxH&$12J3H=Q)xgF{<XX`j!@A3%AjTQK4Tm>~~q6@qDz&;n~{H
z%AGyWcva?W8h;&D|9YN6QY#*y=jxKs(PIti0syNI+AS}967cRhP&`1Y6B5@&a1|Fb
zQ_4bi6|YWXctUCeCNzO3IFKE{?#Tw`c)7m;OWMER9|niHJg-N-QDhf_8kc>Q_<2oK
zbpqVLv|L?u5<Y@5iXaf8tY<MO1wG|LZX1;z0elV>Qu-%Q^ue3<AswSu9it59+!hmV
zRGLqk2Dc!LlHc*3G&~EKQeljuv75%+68k*CgWWye$oG8li#>L55Ln0tdt3Dit=NGG
z-2iH5EB1Jm3;$`Mm}>U{HG$BN!8(w>0DFZQx!q^z@$=>%Iv(7~2oTyMk;asPAj{Rt
zd#WcqE{M{Dn!ErRq4h>kOby1tlHm6eqPe_QhWP?l-H1m2Cy@VXeWS;>HL!r<=Kk;T
z6$3j5$A1L!n?HN0jUN@nzxGm5l#&Iw{D}Gq=3uKEa7Qhx3kCmE1@YJ<vtgg*24w{w
zgRDVFaknoObJ644;K`G<1)od1?j-95Qrhusi!%!>k&ACWYj6w@wDj*H&o}Dg2B+*3
zMnwym)WYkWD>twC9dj^zod=Yimjjp~7G|-^F4c`L(FJPf60X@#j+liv{OTwaART6~
z(*5mg-(N<EPYWcYmI<1sAagYdqT}M)@mEKJEA3un=CEU-X(VFnuxB4!yh53^%zV{y
zie4RVZR0o!>I>(m!K$L&qg7;nVkDfRDzZTtY&nXWZt~J)$5P6SZDqe)L6{#vXVTRU
zfYZxR`bCp}aQ*%9@bl(6-#WW34>nr`KkEGJ)jn(=EUXloI1oa=BUi7}Hk6NJ&Op&e
z&Q01QCB+EuR71o6Bs?8?dKJE*0VQO$IW~8ji}mhly`{FmG>%xB5T9blIOy*db%+^A
z58+!vKVc#j<DdGwOhbw{d9Rcs?K9qEs}*k3{*g#PhP)ffhdbpo7?KDiqwtE-QnHHA
zfkH`yN0kM!V&3O<S=X|2B4@+(brq`R^gnowMy2kwhMBC=axNDEab_V3*5tl>r%0`&
zmILCSB8sb-#b)adfn@fsGs7lO-D-$>PVUj1`!<YkPaPP+Nu;*W^lPcKh);?}opaEt
zc7;H@t7g;H!|>~6PsGil=QNwvag^8QPh-j6ES?%9NwrZ4s}<xZkxVZ#V?fQwk^?F9
zNJ{Fk$x9-Q(9F+TQb-WGZsD<{;l>h*31$P->hLh9+f2WCDLF7~%Y5j2$+^<_nO&MF
zF+;Tf1XZmd^+#>yP$i-86UH6}rb<+V(F#-+O9SE~<3x9{Sji>jC7~~4{Sgk%KOm{0
zd42D4L3PiC^+!xe-A_!WxXAX%#%k5!*n%#;0H-2T-(^;ll`e)nY$(!4jdw^?WEN1@
z*pcMmiwlMIh}iSU2o+oFG5R34dArg=>&~cViYrw&<L9%TIVjX-9XmVXyaQ|J<H)AQ
zcTf|eV1Y9-A8}>@?kBOf#0HV4H$g$81T`F;oK=kVxpId+1GLHQitQLYZn$pyUa;Z7
z6HJ=5hkULgT{!~~0|ci@#dnD5&sXNDvAAH(ikUsONT5@}E0PT}S&wS!b}FwfeluuD
zTuoPPY6Y>MMzTHU4heUcsTR&2x1R~t04fut{4*H?kWDczfSWrjE19A|Mo*6B0rCO~
z^j8)lZA4WQ`E6>iE+h93ZgRr=k{hyM1d{X=cT4e!W|d9OJsjX$<w3A<q7t+W4#9qd
zkpm{O_xVZ_US1s^U7Kc{K}9KMKCA#r7Fz!~WUwOxya9`&e-MSWD2w!>gov>$<w&WK
zlbTv7K3bw8(AMapdy%ruR8zxWJr7eE{HBFRcJ*HQ88zAl+rLP{3uWa;a6;i&C+bw@
z+M5!aY95L_O?v0!D&38j+|6rkb?_Q~+G@Hc0)>WvAh4Qvt5E^@TjQA1_Dw?0QVV)V
zG8ajKr)`g#Qz~V_jV=WtfWbU(c9tN{BUw@xgZACt+8co>Da1EfoQs3}7JfC#sTm!Q
zvysY*N;KUGW~n`~TkFT}gt8*2w)cc`+UEBTuQ|MRO;CmLt@CeF=ugd$3mBuy)vc@~
z3Mxj!-W40`IQ&#ia9Mi+xz?vneL|ez`bo)IJhUKZ)J_kDw^lsyM5TBVok4UWCC<q;
zXG&c20ULR87Jq)lgKr8LC2chBj&8tm;yKF1AcNiRaGU!av6u#_B=QK;>}EYFl0hHx
z*EZW7oPCP%_F%=pTxvRtx?a8sz6?Dy2C5VxevaMZ;^QI1h)9)QuEtm=D@W~2>u|G^
zQ3Dtv#j-1Isd%f)(8d3?9>GqKM`}BO1=HGs$lJlvxw(9!)aQPMYhDq;GZDK=$A<V*
zXtW~^Yy{HE6&mqL30Ndhjf+>j?q0u8QPb{{5odm@N*WtrK4ZMCHpEyF5@2zOqocxs
za55Fs!&!1-Xy-NbyF92gBVnSwbtENeq>cNm;cyeX&?9yxB~gPmw`<!!Ld>Ci51958
zM8mRSBe*px#DWgq%SC7uC&zMZ3u>gT(x?5Z`^@*WdmWl=GiQsdAg$?SG7vYbks#68
z6}|-s#q_QYctm;3Z0oRxBi7;cVDL4T3U%Ulj1M<r$;!76RM7TjF|p$NTO@YGk2*P;
z{hI90kc}A`C6&sDm*+&;EX2+R02AKp9JFCvq&sGc{AIO{9&ePWrg%p$bj14IRpHja
zQ!S2RR;HPo?`lAypP%QleHLNE_s{5K7h|{>I9M^Ve(fFegf~bjyWl{0M(R7?LEJig
zuQAom4&I9<4B@8Lzc&B$N-cMvzJ}$Bw5<2R7DRIzN^RPq>jq^CWB75W(%81D-xsxx
zO^dekZfuRnPeGt>=aH=S)ZDA4snog{*<z-Txqxkc@iX6C+%Z4Gp!mpa3^)W*wd{+o
zeFu%Q?N#XX^Lxl7dUo9E1sS{7*PZVJ9MF%g?w^30?f+Ox{kJ`k<v)zCztotrIrtHj
z%qYEp7KEWZf`b4Xk){K5vy4`sl%HIg!*4$GB_)xprCh$sL7{ls9n_N~5SGX0Vs3h0
zJ3Y6od0XMahj7*gfW~f)w8FgvANA(F<~L^lxC8m0Zwue|B`lr1E?)?L8n`F!y?5qp
z@L)i408{9EpzFMEKpH}hF)miHR%#izp`W>dVrTBMR>Urw#KXL9nO1GTKVGa08|!03
z%VIUoxG7R}s8PpKnl*7x`&M|j8=HroG&ZjD{+22{Usul*%Aj%M0{vRDs&-(0ISE#a
zkaMa9+)=e`+4~bNFd-lmCK<Y%cO1x$-o+j)qOpff^(I4?M3nqdRe-EEakKp4@wxiZ
zSiT`zSQVKqR875lRf~u1F3>0GeSx3?a^~~3>Amr>B{@}ZHIfuIHnv`Wi%eFiv<$X<
zloWddxw{mhuJU<*@43S5?px_82G<_GstIfb*d;yRT%%71Xj`s903NoAf9~NK5uN1Q
zDnSe`wpGj#+E82Ex3u5jQ0Mqym1J)uHx^$nVjy6vz_E1-yR7FD$-*Y{jXhzTDn@=T
z!l|I+$r#{#bjmJTPe^uWRk}qnKp#?9zbRpmz}!5STiU*=?xO8-n`Ict3X;UVM}=Ye
z_L0Hk?W+}xDO?LMX@y5Rc)Ux^RVv>RfY)Eg!W66Gw<D9$hlYI~us<gf{po|Y{cBAr
zIyxGP!?v*`$uq6n-0@Fn<hUP;C@jzjy+Y!-JBkye8a1RR+B(xg{ize}dTzDzVVh_*
zcZTS>sQ_oK%8>sDmUwAj9E6ij)#~P?(8CuP1hFMbut4w0>Lb4&?WBQL4(Vr7fBkBb
zzWcj;KYgHXub_)a{TJ55UwFa-!q7&X)5)$Uwc|qyJuF_-vhpSE{@h}VC+N3>c957Q
z_?b2McduF3J5pI5WiQ>T1)^`XwqZgRG@P0mD!X4)2mziMXq1xnl6`og^KnII!D*;t
zpcU!`z8R@K%5YNd`%8J~k;sVTk@YXu%Dj>c5twkl@xcdc&J}#QZ~Qq^&c&s*DQTfh
zTRZdBiFeqe9PR{n_yZ#@qM!u<jBh~f$Z5E#KI_wcWOBMp=_T5F;Yio*lAD&|Ur*3X
z2(U%W6g@z5s+2-1^@$kTAss16$Iu(a>_d~{L>bRSj>u<ixtklOCS!ntvtVl1N)~8&
z4&?@01Ru~}5-WR5Zzq-cQAny2$*xPix6Dvuly+Ed{i`(6?%8!rbBGS@TEK4;bnbA0
zlNL^A?_useUe)}6haAaN!k697fFfa{L3~Q@O~4|(iZ|SAg}b~~1{Xy9nQGvVm84BA
z+LfCD8j<_suIDiBsKEqWo22TGyBC+IBnNGVhn~3G37`+(<<))UTKPKHiZ!p9Lss?s
zXb}48cN+?@i>P{KZQiD-%=Qz$cc&8+P)4;~C;kWANS4_A%hHHfg-VQjb!pv=LRT$h
zeOeVz+@Ds(8ICK%wx%`bK)i`WBA=>}id(}Or)dhd$$OM|YZC2|5L>Ps?dEXl=ACAl
zOAh6fF6s6$+cZE>bCRM?<%){#&YVb6-2+eA^^qp}#>x3IuP2>BjbZq5EmD)y&&^n9
z$rX@I91T$-#8U^=U2)JmCJCyl0qj>vTT?xS$3D+P08I%tWu6^)L9jm6JRZr|^))ps
zVldQzbP0zQ^o_ryRwOk$sSm5>Xh<9?uEb6!8j*0%(wm|M-Mr4F39`jkKh31&T*)9*
zPK*mrCM`XCz<CE&g>1OR_0_r8S?JP_+8+i$^eP^uZ+xg89UprB&BcG&DpWPUQ_{HE
z9I1}m_{)XR=pC`Da8Vvun*%du*E3fEkzEf%zd@$g1net)XdZLpxwcuTTe|SB1R-Ib
zMmK5-Z#&|_d|N_Z^z4#`M3zt13B@#Q;{W#E$E-8*_u+6|{}KY7rbSwvhoPeBJlDM5
z;eVZjWMA^&Q#-$1epfy>ZQeRe;~?GETkrrUa@cpS?+`lYVWba!;XCVM!{%Ie<D;XR
zvF?&X{XXl=U2}!ciNy5w*w|#khsXD<IN!=L{1fgAs2@UG`cLS>@ek<2`XAO`PHFtm
zLRN%-&wi-~2~1xAfPgiwJNRv3pHv@|i8&5uKfZ-R<-(JUCR$uJ0;$srOh#>SKiHE7
zV+G)PadNida;LGtsebvZ3~oFFz62lW#r-uis{<G?!`SkH^Yuwl#O%p=6{KiBnz(D+
z=fcTlX21HW2=4`D=hX<hfqkB%^sbb76Lq9>-s8IYlrlTM{)zArzwxSvn;q%h^5Glh
zR;TEKjNM)S_VniZa4Zv2(~P?X_8^NY-M!8Yy#ewMs`mi^7p&Y}Y2AI_jn%bSB6&@G
z5OR2VG(6GlF}dk7x+!81hIJc<AY{YX4~DBna(f{pVNiEFiX@ALNenPa_z5URBS}Ei
zeAgu`BoEIQ2|+15AePa`FntXlwAFNbQDGcJgxQo(ZNOB!yeytYd*AY4g|pE|mdqjP
zI%3cXeW1)S1GVr3&PkA&-V=aFKoO~ynOK2MYt=-ROYxc)WwsC<C$gs+&8N3u4fR8!
z$#>=D#l##4Z4343UaOJR0#to|1K{$qi3`J*UrQfK1t)e8XME`v))bb{N7v0vM<3lW
z#l)K*d_i_*qfzRm<1LQnKNsEf8k>XLm3`&Ag>D$8T|d?(Z6bvR^UYiNuvy|$e2d(6
z;QR85*@eBIRb-bBR3L`%nu3|LRHPJrZFAzEE@#?p!vDCXC#ww-`0W3^We)orSEjOI
zH!__f^IjRpU{0gh<gg{RgXC}NX#%4YQB3h6Pz*_x$FKb`HX62wEdC;0LG@}XTu#1B
z>a`<JxX4-E6hH8RL^pRvxN`_y&y}3g@x9bJw%6!4>g<6Jl++50jNe@|;ps!|t3%=m
zGWWm8gUy;+C9DCSckF_)w0{0uLZID$N3X>jJ0YQNO;f`Bh_y-0gH_b0tOg#suhk#>
zG=BqlNWfQl>_qK=tU?z>K9pU1j|iuAah94UOnyevFnL@aGii+J`9JSZ5))Lo5IYSK
z59e<zICF(N^`5sR)U?}dyd$nylk`JnPd0U*(_F6tpyXaYyhSp%jJTWj?sR8-+X2mz
zZ(4rZ2I=a3U5_?9x7(_8Q^!Wi^Yg8YbJ)mRq+!13B4nuni)bL3Dcfp#MB#{|Yz=U7
zk=Ko{G<-Q8P0Qg#(p-;_slgN~9$oehp1LB#x-^1){~Q{l|MSqGjkgoG7@5lY%7VN-
zDjqk{@?a9(=~`8$WKC%d!mPw5`50%OvU1Hv9<z9bLR_;K15XFQG^M-$GzNMx2a(Mj
zo3>r-;fyUv-b~~xD<?j#)nArpjy!TVBc2O#rk_CN9@b)=Vwi_Jn&41g?*pU1K!gCE
zE@e(R;QV3e)Gg-6pbJ4Y%Oy!Serguhn>@=)s*KKXo|hD=Ku^tQcQ>lgMaFv}E2;KN
zeQhJ-5Nag#Ve*tk1e!|qc}+>|$NotvNOQ+s57D**^SkIh(n>qonMwp&_lHhng^kP5
z{bfzD1Pij%`oE|!=&Y9vGlZ7yx&w+*WW_DHf`@;KS9pp6g{iIgt8%9=kW;0DGF%&-
zNxv|<dxF_Smx}Ew22g@iRWf)j-ndIv8ep@Ofg717Ww4ylO2d{43Ii@!>$<!GO)jOE
zo0J3cCQmgKQ|s|uf%R4Wr(H4(;HZK<mB)^50RYDDNizctGc8Yw@2j9S8{_)u)?SH1
znje`(7?yU=_&Da9G<5CI1FkD-+48e^dQOW?ic_`%#x?peeEmSRWX1w^Uyj_-p<U*;
z`vcO#7Qbf#(1Mzeh^SZ162{3QMW`}RVkz;;${k~7>_%NYMh7U3IJfqk*XcCecDXBd
zvZxNMmkR|ItuLRmXtp%VX)AwZJ~(bwf@~UgE_wA4;fq5c=#a6QcD|XiIui{_F?h2!
z1@#fGYU`U*2DNa+_3EAB!>oqxbV$&o2h1P%XwWI8AhT@}q+H6~9XYI*hIbpX2!maw
z`!Jp#YwoClsnFaJ4v)s`4ZC#(3rKe=lW)OFXOZph6ZR5Iqg1JiB`fM_RO_&v+m_Ub
zgyAI;pxp<PkRy?3GpKn<fS06uT8)_$>GlkXiizp=1DrpNQ0Ow>CY}YP8<RRMXt?g%
zAv~N;!>uGX8)Epe{297$qjVP_jVRb7r3R68y#4Rh5PAIUtBno2(SE0_#B`S&9}}vs
zhW!JYzfXTSW=24=8~v}lNqsC(tp(RqMGoU{XzNjv`hP$%cKUzvelz}O8_b{KrY#X0
z%nyo9{>$QJi%*Zhbqmz(9RDxP{pAlu`qA7!`b<WRxR|uY2J+O4{){$>L=<-ya*=a+
zab&MUb7=)-Lo`6aN}uVFXQ5@c`^@Dl#{lSQ^gX^gKi?)fEZ(wCh&a##ctgHg9+XGO
z08>o-3gEkcPz)2&2X+b(V+9Il1tKHZ)B}h|?}NSizntI42j#zfJ-=W366>-!F25gQ
zZ&x<ncWZ?ZaX^*UWi&CtewA$Z;(kCwMwsuD3Lo*yZhxiq{PMvpH_;trA|D5$h_I|y
zE+t=-rr<KCNuwvFq9hj^nR<_=QuHgBlEh|f9(7s5a2GSf2q@!@5#%9T9}-FsOqP&C
zcjJMq%w5o8L^kK?O7=+swI5eJO#U65kJ^yR5&ecJA$(w2gl>|4x2%LeD6tmv#_=0u
zJnYPDoig{rjN+o0gV&wiJv>R4Nx2B;p!UHyfA4r~lxQ?B=ER&iV^?EKcuI-0T6}JT
zriIy-l^LE}|F?-T5G7JULbBMq_-;1wwK|j;OYWS=ISo=5v3Hfc!P`+kEB5CH#lA-j
zhc%`QZjiv7kl`9a|3~$84Is7xquo6XSs{q55hKhjSAd#X7!~IbYOQQ4>U<eKmGN+s
zgg<6khe`pb*dVD$4TYHah}jl-tLqEDKiO5wF|Bh!M-}WYu_KD0sC81daU$I{+lGj4
zg3_&3VZ)zho6M>-mV3=+R3AfVdY=q`D=;ltz_`?%rpd??i7k)>{>FY8I3Rn~G<(|K
z^|&Hzb89Ti;^$ty4@nH?rMZmrh%a_m#f}D2j^DGUMV3k!2oez;y~}5e6Nx}`#UO>;
zS_knM?ASiRD>S+*mUQ5?6;kPK=rlbrj;R+Fzvv03enqSCfCL$a&~sOnME}v=x!~cz
zO*|K1_E_Zh0;}wEY`k?07W7^J<77C7(hu|0ArzobZjwpPkWGFcsg^%oYIk0-l}5J~
z1?dGr6lb^T{J9aj+2W94-wr7Vfn7%*0~xA8W_g1&n_WmC(uFGchtYUx^fcFRmO6Sh
zBnHHU0t;@Sbv8edS`nYEJ>h`@R{G2V;C=Y@WRJqhil`kS@j-Z@<er95^OL+;zt%|Z
zlu|cxSnq8Xdu1LAVym8<iDK880&h$9N`c7VFl9q&e(#qR@pIHd#!Io=WK>7|48^YN
zhvtBsXNh0|XP+hVrc#*zMnhgQlAt+h5H<yv%5UIRDaGUmhof42A2z`~^l>ynR5p3h
ztuYl7E?JVOU#5}izZmd#wg9&iq&Fxx=S!WoT<+k!9U3^|O9LRd3;UX}U%RUB{VQi|
zg?8dT^AQKMoEzJZh!@y5ghn$`n#4jwj^h3X+(nlONs^zdt;(>?ud|#_YKk^H2~3*L
z&cLb8WdNq(DDWyKJr`<9nVhoYl2_-i%j;G=uxoX>W9t_-dlHN1FwTNF<KTQ=h`nG9
zGWQWD8^}`yvGk0+hpzR20Ua=k8TZs;LyxB)<hY$eR*dY<Bd!Zdv1xSi*i^*KR+T!-
zQIkFFh|BTF(m{uf&JN(BBnANsI6l^4np{VV%4$E<*!?0u%YqAuy8E5<pbT!e<uDs9
ziUGm9+)Ck%wVPS;vscZvg%@^Zj=0-=h0|X$jh7V==3$DqAknZLUA8&`CabTDdCIrw
z3$tgv1IwAgy89+e`YR$<2`=xWLsoGH4hrfTM#A0*0xZCiRJAmSHglw}?siF&<*Sk8
z?xoe(X-a~UUmjJ@)$s^5tXQK@>0x!r_ekRU*Yes>%uUy^QQV<DcPoxE{%{{4nG!{o
zgwk_CHI)XnfyN)7Yg+PIpnKR7pDy$JV4RtG=%o9)=<w`{2m&pMTOhM@(vE^}I;cXj
zddGzAm7b5k#8PeN{eHhMDo(B0%;i~;Ii~F#r`st{eLBa{p|pj;wc~XN8hk1CQFObk
zuEH`hg?5YLsWkvEef5}<f#?ynxDuT>KT`2cpW)!|CRFD#IM<O@>!YgrVmx1?N&i>8
zr);hb^A&+^%=nQOSCb5ObxaaAzmwqzWsE-57ODAi3D>wd9n0H_YHCWijT)-k?NI~d
zJdKGQ#da9&Y^xeq*Xgul@q}mQL#1n$ietycLkzC{A-YFr{JJJyTXyH7oag~c!BwwG
z-DJ7_?LN&e2_MAEc0P$TI5AUF4%|Qq>&+&PJTS~=6)@BQz4;;hsb4^&vc--LPUob5
zs$dDWxsY+hD{$)FHBj@C*8*gM2LUik7fhXO2~7ADi4Mw|6LNWm2bf9<CsiG``8g(}
zbO9!ZW44^G)1wP>v3Bbw(Y3p3;9QjAH_psi{2p)c9slEg@SV6_uAqCKN3ncN1ir8A
z5g~?1{YE?$m~b@(P(=KGHi9No$xcK*iqyM%UK6^$xs*o{*o1NS%`6BDuwtU|Uj*6Z
z2T=rLvm^Rw_aT`Mg;8?bEc35-RxEzc$d7X@*gRsP_g+h8fY-({m3!RZ_UU`yrE>t{
z$2v5{k3UVQ{3Jih15Y-rEW<nS<<I4LXnVCa--+HqMcy|ykM-k!qNVH%9RK&Wl9ie5
zKiXDyYmEJ?Ke_u~{mD~x5!AH&`uHvdU`MP1Hh~<lCGU%&r{8?i_NnA_4b=zu3ninJ
z%Z8UJ|Mt+ihQ@9#ls)JN6V<VMePH<MiGP8NC5C`O<3|PT?yuaeG$xt7aNHCYD@Wi>
zY~#c>eV6pzLgD+o(DZ#}E+IVKU<I4kGMdNkskj{=x_Lp~cWp}%0fWxW5Xi58f8LJB
z9t}$F%DiBOe64JZT=nYka8e5Vfe(=ffcySf4|azoj(P-c(9Q|j{)Xdo@qH4eq8ihh
z2Orygp2^R2pILPuTm8w*G+E6@6t<!N1C6Rna`8}^G^BnSkC&puBniQUnJWrTMv;Ih
z`I<u*NDj%wf*U@!H0h7=dkn%NaSOS~^>+@rJ)GxW*Jh|x#0D^hBsi>5)2X;Z&M|bt
z$W758w4%^^nAwyGMl~q&L6E^@BYu^{t-OCGFR?R0vX3E;8CuLdWC4PPzF$#HkV0D6
zAcHMrq1=RWznICS6>`~*!gG63eK4^R66k=Jzx@}9UWJr8C^GqKmg<(rrfF2I8GY<d
z9hTnCNXUQ?yTZ=dKrLlM<}|VY*?W|g2i~)x;z$`vg;(u)AnnUSC($Frgzb0@e<Nkq
zBtCjj09tex)GcwdgMi3BVRRH@x2)5?I*JG7@IjWw+TiJl+t-7FYI07IQu0>%dv!?V
z#F@BQa=OYYXterY1}=4BE|yb5Ymps<tjcKDj_(wk&gm-p+psDcIL=)ypHqW{eUR?d
z1xqf8hDJ>@I!&6e$F4-`wPUnD>AzH)XTt46wmJ1uXyK^vo|?+d=>>&+RsDXV2$f@|
zRO9VK#P&%{;9bwA-k%%GlHN&Yh`5FwwoMGemAN(1Q4G>JunWd+!h9keQFTqZHy3?%
zySgP+Jo8m}$c}bUxt(|9xvwZ?O}nxV)%#XwIJ%=$T2Z%|PXXrZ(=EC)up0q^jM^Uv
zark!y^d!JL7(%|7{5eP1fauaL&Zs2(10~h58{s-7;=0NF%u0jA7J$V4S+REWIo9ov
z;?}Yc_~USb=qjh%04d^rfm2X0^T{To=reeZv3H?vwSRRftkD+(u#x{|ahmnJ8Gra&
zvZl1$P#S-Iht%_@E^A7B;7W)~YjMp!U@@9=LW9w?E)oT;2u-wuZtQpSvfw(b(ke%P
zMn9S<pGlDuyYXvSXzrA>k|g-t?<vT_tn3O3EKjS-&S4At2%}hnvEWv(f^$3F;e>2h
zge?HlP}GS70ayeKk($QzXH>tL_5uqZlZ=xgVSR0oT(Do~zi5rZbk9*dvM$n-zJ8Uw
zD1=n)s+rL4gpNbiS@`Qo+O3e|IfR9y@s-w1(`{G9pD3JwD(i4;p#3;D<sSQZRrSoL
zP5WA@_h-wO_Q$KT(9K$@ZMY&B_)5+5Jp?wd#3VIpg!6+(6~j+b%Fhf)4JWkAYp<_%
zCfzd&j)ELcbX0eik`!5e*z|_|G2#lL_AfVu+0uAdjA@I&Z#PUH-D*+{dVK!$oX`a3
zK*tTRy7;>pMKqVe83|NV*YFX!tBsl*K7@>Mc_6L--HTJc9H<)Kp94!=h&{X|Ys-<~
z=IGH>##xb&jrEbMP~MrP{8C-p&fP0Ck-IpJRP<WBXrtM3bzV!xxf*qmHnVV<+-2cL
z2X<`**1N0MHy9pq!aFnYm<_)bcMteg43q>axNiCRS6|AcSxy9eF($O8c&y{eghI<K
zw%9(;n+C%`Oj53i!71D!-l^u@p6UBMJLB()zeMf+VS%?}oBlH>Pl!hGGy$Fgoc-6v
zbw$GZI&y`gYS*d9>tVjL+}p;HmvGw=H^X*~x6QNs#N6^WXNj&RuYmQ_;=-ZL?*rR2
z&QVJ(%yY(4A<@-IsP1768gv(W*s0=n%_Mp;>IK0&G-rBu0uJlUM!Msrwlg@@<h~1R
zoCk+-dnO44p$0&lK=bf!*^DlmISdVwVj#Fn?H%8YB+4KRyY8f1YSd5w!LziQ3v~bv
zS?J5h7!l^u&Vjl)!QFznqqxQ@I&%%by^oZKkYoul`@s?su~*5J0<BLGYiIzPCPux#
zEhlZft1TyG)n-~{Q|o9vKCP`ZoKU$NEhlL@=2{PZyvr>B<v}ggTxG1bT5WO9FR<Q&
zSvxu^ejX4^6D=%&7s!^c4`S@r8WUm(=i{$72ebElde8tM|HMX_{y}QP{L@1H@9X~;
z=&3Ol^TS4e!0QZLlSr`$E$?^VQxot(iokY2R={M|QQYgdP+7AiWv#U<+)|4Q32p2F
zkxF$bUSf{V#>F<9UQ7sSDR<!D>E}C$gV2QDY=fD*1Hd>|+Lo`ZZ;wT^YE8FxQ8R-B
z$LrN$Vk;e37dIev>R-^+9`=AXu^fi9XROR;40kB_JIK5C$nyq1v&zCC{L~YG&d>AD
z=V;nL7N}MqSbTr<Y_BtY*`}PBBdpWd*^T2$dwp&8V1ZPZgU1RJGJC($du+cjf>~8G
zT@LuFO7-|kejjS})0*|mYEAt89jeOkKGvWn3W9!dQn^g6KBqu?n+-w|)Gu}}44ej8
zP*c>Vg*dDigDY!cTp0yfsG3A>Kf0b~IyE$mw5>5)1geOp#%mTHX|+gwON{PnO7Yg_
zUC$|+c?e0vxkEG24LDxOj#WAO%4OFTp+LuA;xHK65IV<1i~Hv>t$u7nUS8aXw7Tww
zT+CYiwWY85EvO1ZFfw5VYd-iDxkpuc>hNVuDJ`ycdymO#gY&?x-txJ8v$Nmvr2aFm
zm0vLd)<z3OnC*b8`*3?Htmz0D0G?`DG`v*iUdzd}M_jKUt%Y*Uq>CE6W$RA@3?XCj
z-58hSPr_fbL!xs;RHaKj4FP4GDM(0?tcXKOSvo%hYj}ucygcn<nm&WEVPAUUK_jbb
zxB(S)S#MG{gg7YjZRby4CD=Z{d!E3=6$Un7+ll`MkP^tT{N3VAyPGW?{>(TRh9_VF
zl)>Pk3B4ET9-j~$$vJr*PzM=WsY>o}jmL0R02_^<0=)7^LJO!!Lp2SjqWX9XmHSSU
z;IXQvOE8vxNm$XK?!7W?IUmd#?H?KxCd1p(nrj2H_CKyX=vv+kQ$vIzR-$AEz2o#G
zKUprgQ#VBh)UPQ}w*V%1?>gyXC=Ji#^@QzRh`ZrXs+H(wec}Nicsbyg{HUe%_*9#8
zePt~1viaH$G~NbnaHxqrFU9n)&I(a8$6}1S=01cxNJ9zSYVjs>R|e2gg{KXm%f#Kq
z77>ETQ4Ye-b;qGq>^K&S1ksn>s6U^Bq$83b{|(>yqwifi4r-8B<o1^@Afh%eCWk|P
zlp8;^F=nvT=O&EDxd_<z@GHlb2lKeC-oAmTMML^fM`u7>>2pM1>%ni=bTOvg7Ywlv
zqjsgTpLbx1&gm#p&Oc{W7K}IDU+5v*mSPEJtl8KRlRA|)%WPt5CP)k&Ox_L_XNVqV
zCxb-2%QrkGV*cg865yR3D}1MQwZ1}Gyy3pS7Xgv7ubHS=FMH4`yhKZrXNM@jI4~?;
zICG1?@IeL3q7%CQ_BKir<F2OJ#GM`^YFINS(8sFSMR{Q79@L~8@~&VOJzdeQXJoxA
zLB+im=~0K5E5lD!N)RHjM&_afqss9)xd1k9buz%cOAFQjrVLsWg;sxGte=_rM#cy=
zqs!~<C48Q)47CTJA#iT`D3t^T`QV!^iXed4mYMLzD~ycTb(r%m4-LeTL(m%Tq^eNI
zi-PQe6eC<Ybngj83ui2}k9<Yh%FLlm+}Tk(|GR*jt1FFNnX>BLa8SH74X)|0u*c4N
zjP$Y!cJt$sR`Y9ma&%~WJQ`oc5~<83e;Vdtu=D8Jd&1;j*Aj01E}U3d9ygOK9D}EF
zS%7JBm`l{Z6wl$_@e+qdTB_{llfCn_x?KvKlkGSJL~xP&((&@SW8G}75#G6GL;JV%
zQhOZVm7tv>rrWJcHl@AD*TJ^CT|*(R>3p2#hZtIk9IqWu*R+}1rb_D1D=JAe<U;of
z$mpQ$TB|}NyZ3+}L(oTZ7cu;zFSp@Sh<y~eT80T)S>bs!dI2rB(?{P4lhfOaltS`(
zrzx_pafqpM8QF-ftke1F5{)15m&p3^|8T^!F|z${e)9iv3UhEU{I}AQk@Y`pwcXUz
zveg<v^m(pj=hvGDzg)}Sm`3yW*FIwoPQ-G7YGR4<ODGdAr>;OjZTo!uJdzJh5?s%h
zK~sIxcc0$XH8C~meKmPbJN|oEqdG5Jwoo$PzSG`R6DP2<9FGWc8#A?Ex-NkOa!uWO
z5=9oyy4ayQLG&=LL%{*I7vRsH!Q?EGCA3$A5#mp*Q;r2y=aO`PQ|T;x0#@!agqDyL
z(==S078*4{v>%~Y4R_sEsL4$1U&PR$-?c8t)$c|-NPe<A%df$pTX+JvRHpGz_tzei
znf^S@VURNke@Q*mVI*OYkRB0ROm8XXTpw8^HpLJpO`8#(rz?^11L{2-rWZ?S?>C&k
zGO8nnbZuP-s=Bzckm(2FL_QN5`R_1IoZb!?Z-&s;-`v;C3pX{+lAI5t>N`6_nw-rB
zX@wT0^<Ls&V?v&*a@irS7{fG2)fmOe7{iXjLQxkO#m&FD0|q&)0Mz>Y?b?MT3eVDj
zl?JF0!7#R*X8Woei%ow)KL`Metu|pba(pEr@fa|02Wbo*%!hC|ow|wNs)IT*!|WKp
zpoTcZo`)oYZ-`N^gaGWSlV}EkMRU&Vn$I)~tsns(OjCGWX%L9ea{*xr1gtu<uZd_j
z-NPvf>!o3UVZEhHgG+IuJfqd`26w97-N<(Q;TB(KDK{>+Y2=PdsOQ13KUBu=_nFr}
zM%%oJ!6%5D)D#~CLPVqYuA*^Kp01}&MTiW2gnTDM4sK8<a>;_d#}KyuI$4l=r_(dY
zS^a59?EU=2bJI&KL}ZbJMeFHmZ*zNmx#iL{C{V5|!qL=}O1W;ic%w5xZ+wlc92;xe
zKbd`;A&ne%@>!d^yN$I))9Y53Mq8+%*5gjsp`{;5^b5I|(8$5(jX2r>$HOnC!9xIo
zAczhfiG4S{UmJ@&-825am>~OfU0{pUitc@XUP58|rVkcOjXXWv<2113*t$P`yfE=$
z?A9$1=z25p@Zf#*dO9+0stcPHt)2Ca$<cxR7OtG3r8AxI_u}|0Oa<or!wBZ24b_CS
zUft)8GEP2AenUP=enP%V4x`kiS#3F6=xguT6;@h$ocew7He-6faOyS7*L(K8=)38A
z+dz$%BPWMO7io|5GEP~Hc5^!Bk9$O7Oi}wb!x{Be@NMz5S@qrU4stfMjDlzdPh0W!
zJ`#JK_9UdNKet#X6XgYgO_AW>2kc3>CZDgr0py>FmRAc<JvEL-(lQfyb2{GUR8Zun
zvXlkj0zsVyAgfqSF*p-@RbBXP#qu$>=j8$!{q^>7{_;E)8#%o6i1+ptdOsY?Hgan8
z)bV|K^td#4`f&U;!?w*W#i#Qyg5J3ABdC;VMCqhCGemkNvrds+(UvR80sPV^P3ASP
zHzW5p5lQ9rv8nXz@bEtvd#4~#!fwm9Y}>Z0cG<RV+qP}nwr$(yF6^>x-a38HjqZs4
zZ^Vhn*L=&!Z)L7I$3SThW#QdiTU56YZ4CTSJXEjEgz=D1{^zpI&;tF8B@L8ET!$2Y
z25_B974fGoS-zIib#^NmAYG(W2*`*;vRS{e=aV+PiT`?G7X#bj7=LTGscPiy;^9a7
zWH%2u<Jio9a^Q#T4Z`mc;oD4NSil6}NPw2f;2~(#USejm^sGWr;NDJD{FI+*)AIcV
z&Oa5PB%6gXF>Mf6K*^tsGL80o42>LI!YKr84|(I|fe%cI_G9uV-=|)XEBdD|gU0uB
zYGwa|i%xrV5EALm0T2<0IOjYAH_wA-$$~dGGbeXmSZ<czd-4ncdUuwNewMy0$pimg
z8{R=aJuVk*BNDiazL&k)x~e5j70Go~Xn+&s5b=AQ&~TgCqF$3Eg<v7rB&eKlHJr^5
z24zbVm?t<@GMN9RM-WCLWpZ<JPiPMqzJGjIZdZFIZ9&)<g|Ma7Ile=SsfEM|j0rpw
zGV&dI(yeYK@xEYQoWjx7LV^INexj~z*gc5R%{+rcU>8?YQKa#o&L|16s^)e9>3QB^
zZC5i$6Mw~RAHk^I!QczU!v~TPHTJOjRv3a&6AR;N!v49qgG+Hcr=m74Wo;~trjK8(
zRKl@*za{9N(lT$RaN(iIego}B95w*plWveH*y&D^nWlsSb)9?Zig5}))N^8x*?hd#
z%gD~6;Rxt-C8;_C+;8|E$G2HnIExz@M^{8aqA9Q_pNvOXBLKu-(rQ^i|EoM^w{*ys
za?b~rFcp9UmkqB!dG8${1%PJ0VkZ6o$QI18Qei8&?-L+6iU^lQ&td>jj6bf;vcMrX
zU|Zp!cnbCM^M?o569FE*D-X`brxlT$Ju73L6`oI@-XG>>vGG{hCoMHHUF3Sbgw)CM
zev8AiqA|gM%11oZLDLKaJ1I1VoQ^fOefe&<Bo+7&xlvhCgm-|~l2$gYp>7Z4WLT?`
z(gg5}etkRiE2V0x7aTn|@*^jhXs;8VG>`ue!?Y^|og4)sXg_JD5)Sj9c8gIw<SFsJ
ztPHJqQ_NnlfQ4vd#mQV$BQ*BpHB_SjJQ?2B5<NND<0cb_@ZpF<9)x5e%&wK-L>H5b
za&YS=PBB{q`uuYSw#(>vTWTe6QI(yN!uGZE9QMu{($}^PI@+h_Vu!HQ{K$^$LNj~Z
z{3SS81<&|5JFv_xGUhh;faCRgYV+rW^AGVSpW=7k!q&L*D`Lww#Fj1AW#70YzK3OJ
z*mVAVl1;otUe*C#+c5Ua5{ii?*Jv8TA=Xv>LJCgJt>V67)0uoT1tk#%Nc_NWNu<LO
zu=ltaz3vj}dBQdl7v$Cml<}C10CIvo)OjCv-1scWl9zH|7K!x?twd)9S`DnqrqE6h
zi4S{D;a8D&0K40Np_@uaHYxySF;l1(;RNEOZD03k%fsEZ+?27K8TgA4<ob5LzTQn3
z@Ld1A>cOi&5x4+d<czGnT2a{caJYbf#9T&OYi{GZ>>d%AY=TXaJJY_P1I|vKVuT2b
z1OMV(#t3CIW3PRT$RFrYhHje9B8P;mxLQy;fC_Lmlcb`cmvS`+Z)CV-;8r+q0q&=!
z${=rZ@p;cn{B8PSiWoD4#g8AB(j~onRo@4AnXM<6n9jEDA5Qdc2Pthnat704lcJT5
zCECW#Up&{z#l&~%&NAaYO;V+|8IqfVhuB$@8@A$0*{XxuhfdK75EBp;5F2<dRPCRj
z7+GyG_f{jq|B63(cjj-dbl6aUcWk!UlPf_kL(wgw5a@7Slt+3`)y{>j;1wEU-?E}e
zzSpe?4zTE`Rf1Hy?TqvF@Zr&+L1qIFFewvC#~Nk<H>t|r&tjlr<rltVmrb^5Tfuv*
z(Fj7*mUsYsB~D2-?26X(j{&Sd#nsHm1M?XNz5`4v4Z<-+AE-8|30W(?`(&l%*X0Nj
zJ5dD@X39GBa_{)0I=kBVv<*^G(AJ8CvJFc$)#|0%T4kCoEV*1{!suGCJ2qePIJloa
z6QY~bzAo0qa>x?>ONJf73Dt|AIONvNd$HKQeHiy0(m$8<_ICJcNCQ6%ywHMkC>4OV
z7Rc=FKa5odlE|G4i*`GMggAm(DS^D#IkHj;Yu+~YB2FQ-Z?&>g5zL@?CdQ}pxi)j?
z>?2SpDE8`<CHMo)V|m^8ALd$imj8Q^&cwp}KeO}pXmQ1nv?0#Es7bqH)5IC0>dhPa
zZ**V-&eB+uaWT<=h5|vG@$>DDycBhD@O-4_*SW2eRH;8T=pNn;-GKYPzWqCT`gZ|(
ztDmkiI6ib?2QGlm{zc8sV(G#l$N;;4`#$z>&1|-$^dtoD5XT=+$WP0&>Hrx+iYK@*
ztQX3`n~CBR;~bbN78G|3;3?6}6Wa5?LF#-#MPcR;PM;?iyLj$~sej78^1fNV-@o5)
zcf$rde;-FK^Ir$v{%su}-X}Q9??N3;k%|2F_M{KKkLS}Tn`H@WM?hXRKEHic(j_o4
zlhOoU6ByX2G5-@N0i?)$AacjOpe+LE)^7M!wIU8ZmWZ(6$Mj%OO=C-0UAKi<8j1qy
z!8uqh)+4r3kx1c4!z7@Vx{sh0WcExz5xKG*XrJcA(E-*#hqaEyPbBS_q8`G^*J70{
zr6mz9>``6b{&k7oW#N5lPUAibRf;yFg`swRNU>s^tp_UYBJqZxl~jvh8lqNI(tzxr
z%d>$MkbN*w$GqYG3-PKUF7lr0n&Dzm<?7IMsbNNIOIag7uAB5`VPNGYNH?i=tlSQ(
z@c1j9fDVxjfSk5kDb?j?HK2MKn!04_jYaBUqolv8*x6YKt5W;<D@;?%S7*&u&9LY<
z(G|nwn}P(p9=$O<nU*rg06<lRy;#Ux)fVL}>+bC&0P&XUKR=KsZyV^W3oQ&Ui*P&K
zZYcEpbQX7c@%nvEhr7WQ{jS|U77(h?RhZ;J@VgyQlY6?IrB_<RBsb16R)ntBF#FK*
z@&u@w?*qmq!`8`d(I!;+i;vtmAQRBFQXqiz`Exv|QwoQ(LOn>rfLNe3Dyo4t@DYAR
zX~#ExS4yGxJt2CE9R<vFPiAQ}_kGl)8|#^QmLAB)A*7pvwpc}&p2nTT-aVWrmn_-1
z{WsPYBBr9p@p;^pPN|H7mQWE&erQ6Gt0XOv^Mq~kR)SL>l~N@yu)eYrwD#(z1mX0t
z7f%jBi!DG7x^aym9)dN6O2i-XOc=`ac*A0ExugW(YgW8B<_Ce^NJB^uaGmGm+Kc<_
zM-N#{Jyu#yv*-;z@CU?Fj<UIQ<S8GjvgxSC#sq;n$H*wKl(N)_Nu@a0l(9OEBMRrD
z`{OO?qm$xiCx=@Pyls#gLKw7LqA_{)l2wCXX)?=(Sq%1CYLpsd$w>uWyYVd6pmG>Y
zD4Bs&el`i4(Jc8$38S%u7vL?jaqnaqUMu}*J@MCd@^(lVAu9_FRHrviV6s_j#6qip
z2o7%vzpFsie?+Jti|cf6G&OBQ+WJW(wDpCcLZ{PG_PwjuVy*MBl-jc`d|o@T!a5XT
zuzzm&o0F5p)mvJ7Ux+yi-#*ENmB1$XhsM*z9w{CV7erNxQxWkicBiX0E7V$Y)LKXp
z+FX8yJBA=O084&ade>C{iM`=rZBst0I^T>h#xr4pzKkz=N<NtXkkO2q`z`hN&ZTh4
zWCUPkz%OKlfWGfh%@T-QkBovUlWwIIUCmPy!H^-c(j&IsP&z%<Q9(&c`^{<yWc1L6
zYhIpSwahqZi?B1Zqk7}6pukosrheEnIE^s3d--=1p#WMhO$5r2L+jL-(Q^`~+X0wM
z>VSs=@#kRU@}8^#oL^3tgx!rGguY2tYxIK7=x~F)Z~6pS;<T}~eqQ5tyW1X*qk0+Q
z&``oAw`T5aI$)}Fz8|GQVPwWbT4EKe8n<W{RBo}@yH|L_=nGe2`eU42vubCv*~u#{
zJ>-X}ulO!*uCyE3{0S)H9@-E5uwsNXi+#ls_0w;rCAFbS?B_QyvmQi;{^;Ir@#!_B
zM8<rBNC;oHN(k3BzHcxgZk5$D!)9FEh!iijU^Dy-;aN!uFFwKjQ{u7{Mq!ESugXAt
z@htHxT7s&qpm`hO0i|8Li~d(R5S}FvU>;--`T<Cf%8==F_zWCqzsKC<bTG~34=2bJ
z01fGjv*<Q}11hp<I3UvhM({8AA-@TZTA|En824l5w&oCK7Ze*-(H=J_*$^{G3!_?W
z!W8ynuvTra_4!KE#|A&p%mErQfhf*vna@jdiQto!Ms(_3K%lKabeXv6=C>&h-5K_f
za$$1iDBn$fy3-y~Pcy3<lN;Pq!^8%P7hk;16WPL3;*N1X@X~{qK3EY*6-1AdzX>9}
zgj1OrT?x@W|5WPlp?|7$ad(vwjX7($nqPKNA>DCZ50J#k67Qt}o-Da$b^Dt{J81U=
z1fM(G85+d}eftYu!9%|WDEgM$CFW8<jPlQTai;!|+%9OvCcSV4s05h@*=}y&FN`9k
zm!xIZ0F?=S$T6yVi7s(9Yh(9Cn&v8il4{z(`cY89$Powjhn81`Jghn_NV`$3B%f3%
zNBN~)3T@lls7(8SP%w*_752%^B@-Fml;lo!l`RpkPE$ha{gmb_{n)_9$Eir@`_umG
zrDcgSZ}Pb<j419CFX&RuK%lG!y<~v&GPd#f;z_z}Y}uKGX{e%eq?A<iH!WR^Jer~k
z;m*`v13}9{gMqzO6ZRk`9850V(SG`i1mic)n9Ty`zSd2&WUTZ!mkwr1p#CW@Ab{9i
zcl^zkZ4y+F^y(0xyx6qJIBC{54${`7!k9Jhk#s6cu6t`e1$a3Rd!_;5(AO%~`Mt|H
zb0+Z0x4o*1(<bkFmD3w`h)e7nY#b>3Q~JisSY}mA8;4??_hKtk&}?FZ4C1A>%XVFL
z`YVa19w3HAc}h1OwWbtV72~FT*h7&PB3)|ipPuQs$P|-E0M^*aHqHSSO{qS*1x$3a
z+J`ha(86X@wb+ip-E)$|3=`&s=)XwRhr7+o>v}8%K?Gj5edA;7!c!}Rq24fIzT<H}
zJhC>qd<A+6lNFTk$`mrt_=hC|v?W}1O%O9MgY1^=*WErL8PerC@!hT%l<7>4DaKx)
zP}a~c3&S8~zYHA<kYjZk?4U%QO3jZEdu<3K9;~i+uEW>bj6Zr}vmOPr4Z3GHq5#)5
z*e5X+;}YYR0UkYQPrPL~v~?a?N}@CI_b~w-IWKvj;*LixOs8ijI$4XirF(uWB9O-A
z)6d%v)q&~fYyqDY949<3M<G^h?~W}AKxHkoH*(t)k>(s9_jzL#1_ki|l#^$npFzWn
zZM6LRYpwWCYX`|}_00p1;^7Y*UgAai1SH4Zl=_bG19EY4f)`dei7_fD4BO)c1qz0H
z@`)fL+36x+U~I}J@R=OIs!u}Q5>KBRqcPS#{f*G1<!Agwoe?ggI3vNbu~n(;IxO}-
zdfB5Ya!>nG|NcI29Z$ve;hmiP>AwWXWC^PmSjhYSM{t1ORSz&3{NuwaDKb-1XjA^|
zR`lu}0!DG&b^sOj^BBLI%=wAm)WKu(pQcB4rvLp{^uMcWObq|Ss@gC6*zUioYX8+S
z7p?-Y0B67;$e2Mnt%>>@nHzbX`#k^k6Q+nnD%Hs2nrH!s>}4E5#K${4hsp7ke)zaG
z^l&khRuwP-88>-$_ViJF8c^IPxtKpPK}lMzuyiqZHkXvVxp<PnOHji2=-_;F7I`98
z#62S~j_+By{I(DfX=NU};8D}y5mTV?E~qB5R>m%S5!B2i3GrCKlg#(?^!^cKY*GTj
zx-&ZUdi6V}?uJ!%vI5zp&j$kBF75K+?*q$IR6{2fQfsn(y_>y%-;PnV+LBqwd<Qwc
zKN(?(_A+`3AbBoo4#V--fEMOe!wwFMVe-6Eks@i@m%wYl2PdiKr2GPwYb1rKe|KKO
zfeJ(h(7}`>6B?QR<z{b5%QvFpUsM}ahi3wfzv^BW*Z4tp4702dJZ660WaAP~gPdo=
z?GfZ`fP)?XGLpO`cM~#HHpwnm;$1lXotw%Dbr=t}VGlF@)nPjV@WJyCSVU57p&L)%
zW~O6!G8{}H=A}<^jis1oU)pp`>i#z@#yMbKx7$dd$?`c@$+j~wpG8QnJ){@LG;!kU
z<rUKFYdQf*%Zq<WK1`QbIY6uj;mW5AriY0^1XhWo+m%kt1;rE!U=<%bjP#cU%!N`=
zEW10}BrQsUQ=qkya2UNJL{RmfgovvR_5$oJ!}~(4ptO6nWT4bIEyzW}+AI-)LLhw*
zSn*dD0XQ#CNXZG<68gpn2&oeBgIDtwR?TRel?sBFnu-iuwalcS+Sk^LKFGKIIdV|$
zx!5-?JY`ghrEhQbpZqF>Ho;2|34cujG^c=iiPO@!)^h?TMBcL2<VTHw<q17Zn`FLg
zt#aoO0MM=U*ONQJyD?R#dyXOMEF+&i<ut@9RI%iSz-rnW5Wp%|Hz2L#ju+c6i=_To
zZX$z4d4j@ru@|zpvtc_W>#lEXunHm8X$!Kl7y=#w7C6wYdJ7ASjn49N7|Wc4>G9K+
z`{5V!>KN(OX)n*`pb9qghz3dIPR5@2`x&{xI}8CBxyeOfqg*n-46(4>W0PZ_py?xQ
zZu5Kjl*u$lB08%q+K&4ymoEjl`6_=hmnusHZ60VJNIOwY*vq=lJsnp+RgfqqXe?6V
zj<r_H?b4Z~F=a*$mdq~vRq?SwT5Lfwd@p-cF=4jYiqr2Bk_6s5S8rLUiQbA4&Vf`B
zG<P|Ca{t5Ck5GsCH5hi0$_7|Xkc?L4`BlZ&RI*W~&J!pXjq^csGS=NGl(B_>S$`s~
zJ&eKV9+EO}<2(_!j2evxU2-Jjq&Eu|X?rHoNHeF<6t|y{FxvafN-BJR=?J3T!SVEV
zyD*S*OitOEB>QCUo*S2d(xm}ytf%L`tNxZ}ii=r6B9Zaw8yGoC*%vp?WG5QhK!Ac<
zN1Q8XvfHkgI;`PBZ7OXqSQXN|@lA(kN>5|y&ot-BONRZaiGxFI?bU}mfvQu9`@H-@
zRU$!7vsO42OrV@@#CwiZI6bclyqT2zOH8!R^NeV+S(smKX>AD9!X)mhL?+eVe4D`T
zB;>0-Ri#3hd$KlZMw8@@)oe@1WR_t({J@~P*NfB@ktC$DXralSahUki9(z{{ouOXk
zs$*a_S5jV)VicRwp)+}vV6%<9<j{c&ohPI2(o=>Iy5+&IrLq08ljRekGHC83erf9|
zVhC@ElKM&7!0{uw^b{Lg*N7)lO&)DuoqVsNYi8&Nc(5<uqj?*}vYN@tM(fpgNWr#s
zrt4&?wS8=?$@hYxvQHQ}8!)uF9Bznd16InS@_-dk8}nM=)nJ4!-z4RV`mWf_P&UF=
z&eiG>*Nwdf=etr5VM27Ud4ToZ;Wc7pCVUyla;mVSGrn-i=CWxsRsl+$cqt}QZ}DMy
z;N#b5G{1o5Bi3pZjRfG?qA)7tJVrrmelp|?4!@A2d;TNtwtqn+Ci{xhS=3V+*0$%1
zzq{Z+2;Mj026PYIo25-Y!&LP(+3x~W?U2*#JKz#{w1ypVamb;iT_$<jE}J;-HoNYa
z(`|XkX`yli<uQU-n`9=Katl`r%s(;^qq||mC4SP(7^Tt!x{_rxH<)S5m~&Xup)^Li
zq*1lKiGkS9Y=?~>6DHJSnz8X7y~b({uEBRO#WCc>+1fPge4j~M3fwWX-!99pUG)gt
zI)r89w92$Z`R?GCevDdb<l<dQUlU`YH<mZG&f|=w>j=(hgzS@kRmE$?J)w#}R@SnF
zhAIDfWA~v)jp!XSw7JIBlPJL=<ZWbks)`c*S5+_R<m#a$n%mRsw^>IeC&9G&s2eNs
zcd;ClKg~sP{5Xx7-P}04B9qK1xJHsw{PQ?jeBk;K?e4)f=JjZMEn?U5mV)w>LIWgj
ztxddOjjMRx5q>^DKB<W`ME52^a2K>#kKp1-9I6LTynyFvY(EhtwXslBw7p?X@rQO(
z&6%E_&=s!m6FGPQ`v%6#o&bohn8mjP|1Ah%@!y}(1;%GFS%lQXxGJspe;MC3B6U!@
z;%q-S`?Sjp_#;m)G1wIbz7daTmg;(2DwY_RXe`rp?XmJh+)goA&m692&8K}cGK)5r
z>c?%5T-^cRBWYs;ZwvH4;9pw><fQ!FsFD|QrGAs11>Wt%UrPMln<1%e$+@(sa_Zk}
zb|+(CUe1i0`#l>)QTVHdLA_J)zAkH44Knn6qqLOb<$RwcmYB>9p}edZ_|_O`Kk)|#
z5co@3`fu>b<?_S-X`=Y`_x<10n*V7I6tgQ9o2|aDZhetXI_b|4VFU(v_AEK@nW1)C
zG1NXvKokb${Tn8)+k{R{u@~y1We3;Ii@2PhjOV+flY94@yKrXq{VIDpUZ0F8p7|TA
zCftVXEXIg=>DbLuU0yR->YM;?F-xI$_UE%>OB?>YyT_3F+qjaEEgw4*qUesmc>Opg
z*b!}eT2;b@4mbL1zJ?f#0B@Lk&)3)83ARqUO*SCal)-N*%U)%cNEM0VK5#;`6Gbi<
zIji;#WxqndZ7`<gs+gx~5Z58J&?ul479U+aX*b%vhSqL3%_)tvt)E9uAu&$OuGJY9
z4h10wiB#5SR7#t{S_>MeQXdQk3?9J(UE^;lj_qFFnr6UU20UZidiNRCe(a(j9dn{6
z73=~QiDrV|^o#sWhc@+>VAeI3vze6onIT}>VY&yMUHj;f37R?YNPcseWx!~~)=Dp7
zQL_WjqvnV4E6~+RTPoH#xx@!!<n<O;Oc#<3<#&DdZ2U@A(%c|e2#N_*HTq5q(VKfS
z`&WFsR-SJTF=C6#Y>Z@-L0WcD)WoB{t!9QD^>ji{BbFK&CD@!GOl^l5NkQdsm#Fvq
zH4F*_I(F>BvU^NX9Vt3ZE11%{b3%^rAEyR!ONGo7Ev*`u=3yjS#T~ltvBN9g6#vph
z5PU=NStQbMtiX+<#)EOfcS_4No~e|e&`a(tx#57>P{KecU(Ar0cG~H;Sqomgpq(m1
z8Z|WNHO+>Mlc~3YzXxyC3iVp}DEbF&Zj<HZ<nr7J7Tqd@(I_{u+OXl*tLl-GU3C!X
zsZcmisPuZFU14KceyX-cHOj~advP~xcLD8{hUjGe%2vLg96nudCc~OAOu9;C3C-mh
zHCHjSlbU=H8K{cxg)yxg>=k2Fgu<cLVohT@kD9>DFYM$W!H<vkw>~QbiOr<0o)Z*A
za+93wRs{?{vG=7Xt=&!<!iY^wv3V#8Y4UsvA`%e{iu0c5b*MVQ4OZ?&dqmDq-Cnt8
zho6K)=SH@+Pj{?n2#-hN{Tk>l@wvx!hZEZqg>eA4kr|?7G5XATY+1nZu~EhOun6+&
zIwCq9ef?~?sbb<%rOBoxcMJEUfyiB#iB5PL%#h4^J1eu*kF6Y4@)_o%u^dC$fBt4$
ziS8^kL5${Dz1C8mzK?B?sX0fJALwjST}+P)kyB?U*^V2K8*Yh@pk%=%>oWcjfge3e
z9qW-Oj@LIX7w*Vsv>=13S=&8aZtrx^fD?<?4V*aLI<d2-gh_zP0E5@1gb{}4^=c4>
z62Rwp5n}VwY5^8G0={^t&%P~g9WoX<ERkL3mNj(&YRl#}bJrYE$UUfsC19SUQ)Ab%
zFU_()bJd~3KZFj<VR9a&d}!?C)&=*NlODNmPvBy)EfWkX!f%(mQbA1w7{{g5TpIXk
zKPCFH)a7TU<}X?8J1P*~Kbx5pU3CJkiu=B7>)UG0gkw~gGq1#{R#4HGRJJ%Z)#p&Y
zcVdSg%U>-1>d?C^ng%~IpT3@AQJ6^;W8@u<1pTL2c6}5K&}#aMVy}c$)Ddg1B{h-f
zYbdk!>C{am0^|=b$FI71Jy(`x(2A!sVRcL>Cj7l3ASEU*x~*d$tB}tV&*s|hV6z-1
zk!<L~o08CQ$+P$KOFQqj*9|+Q#yUGoQ{E|;(1V;?q~w5#ssm!1-2D3H2ozCR1=JPt
z@ebJ(!~zms^`W6k^bB@|zFvIO*dC#bPcb1Zdgk!nn!``?r+Mv=f*4Ns%3knc9ys7O
zBq3YY98W<eVQ>P?=P!|yG|+K#7*g`oLBn>Q2?R2S$MLCk&@gZG^CaVI3Hm9J=4-(e
zIGKyU6TW@=(2>oB&o-kO8`RTpkWkzT2@A0XY$gy&=0;O?1Gibw5|L4#`91X*B-q2L
z2R5S}P5xj=53)W6iq@uXXPC;9^;dXrFx|P5>q!NrXMQHZ79ysp(;!zgZ`Y6_Mg?Gw
zh9c9JnIMuyV_ZD9v9OI*fx8~aU?GA{=ePaTCxLN=F5a@%zQzMj%yFughYc=yq~xxY
zCGm$+FH<7tB42`$BcG8JZcnU1EAJqJc3TGLJdv#+h(Z=dUZ$QG`!~pv(oi<cwy@9W
zdF1Nk(!^+@lJyWbVsg$}b$?b%>_3Oz*4Qj6`P3?STBECv@XS#$f_h<TuA;X@K2uKg
zG^lu1E>NXi&9%J*VuiG0H{`a#@uEPNAgjNAv3N;iqJ&?HZ=VWSsp%)1<_<|pg|?QM
zok<S+h@Rz<CP7bPz9cevB=1aAGi=&qbTg;-Q{OC40U1P9&H3udxqk9K>&j+EB0hYb
z)1N8vMsteW^+slw#i=?0dW)7q$li6kUgC14g`LA04!=i%v1#-fBrQJV#6e$s<5jt1
zFTjh4PruZ^HdR4M8{EM+`CZdwqMEv^symHep<SVVSH+HG)UMyhs+|e_fMa)LF8+r@
zn1kUzlrjFW|If-8R~r9EyOJof81J`?F*^$Ac|p`xSXQ{TgxC4=BY)0PlyK!RDw|DL
zAJ$+UBdnPB8qa(2FHZoTK5P5p*f)-u{RlG~Q@0P6fWFx~FjaOc2v{D?!2OH&^De)o
z;r{ZFum<tS;eK0cYKsemp4k5=p$%Y*OAO&;9D~_|Zj&3_B-tYot=K2}i|PMOw8owK
zk6q=}_1QQxKKtT7CYG!wAZ@HHmPZUTDHKaECw!VV1d0%M{Y~rrSj5-LV&1vop|Lbn
zO(m1=J=_2i8wA;tyc%;87NxYWUHWd18xM%0cGc2Pq{j&{V&!io3@8tbVIMs%`FQ&5
zd~^F#yO3{)4F;`S(0{x3BI7MUAgO~9tpI5LdiQc`^Wa6_Vy(POMlTg^Nj*A;!19}p
z9l%ml-K=_+1&dmn?|tME#oRje$2q!eb5qR<W5OU0Ufe1(qKJugBNB#EzU(RVFXEEl
zjyFfTisZ<a0oO@IX6eFWMmaCvOiJyLk{VCNR4^K5P8=gXxFyE&oc|@*HGhDW1-|t3
zOxVc4jZy2o;24fJ8J?Zf>{Twre0%ygEG@>8soCcyo^uU7$Jn1FY;*Hv#Iq8$emKyi
zu{nUW6qzlOZMZ~Jr=BjJkg*@!oR&?f<%x1JaQI8mX_=O;qPnLxB^4`|j5wUg&kPY#
zs!KARs&g>)AWUk>+HM5uqhOzt;g4<=f|5jlYts`4Z~(E4h;}_G_dj2?v!ajU=vU{_
zm)mMx92rc)NQClSCtaqj{M}5iiF94YVGxzl7!Yt?gV~FcS{{FJ(d9DJoLRRzfoPk}
zGnVBDZEDJC)Ylyybr^x1kYP>5=-HJT)t!tgE-{t%VASl$dZw%~+BI%ZL5(an?H33D
zqD>M0_=75Gk&(owej?3z@upd+%IT0@tw~c$#UHD1UveCabk-;jtYgTn0KP#rJTJu!
z(#J2dF#Db`o%a;}R%8r+Hof?34>w&?CTbQG`!eU=2WeP6S7*wNk~tyd2_pxsI<&uY
zWiRrJaP9{z%q@0aeR$$NbSeRpdR2les$^FQ7}bEcN3Q$x1w|x0wojfXqL{hcvga^h
z1yuJ;<8L-E(L0RSo$R`E=oK0WWzGedhCiU%MFB)LR{96vlwuKsN9kqt{3#${;mqHw
z##@`(l2q(Q_DspiKDlV^F#%=Hb2Gephc1jTVwQ2$Eyt=)h+9K>bf4?EOxBRMVLIxs
zL|GpYB`eeuN>K@S8Agi<EsN^Q6L)IlHqHVxGcpB<Z4eRRIOdEeU9?Cih%?;-`UNWA
z13>`+t>@OOXm9tcd&l@1hV@)Zmv^w-$G>|8Uzlab_=3ch@zp2;?8CF%X0q8gk6JRF
z&5{^Of=>TndhX^mI`di$Vy3aErc-|U?QvCx_+-3h%643@Cg&qcGtlHu>u3vloCCvl
z!^TJ(e(Dz^|D?0VEx_}*-6<I0(QKIRNCWn=yJ1q*MtO*y(Ktg?gy_hMBi3rcYE;0P
zcOK4Bl4fQgZ;%-MZiU&-yin28OCBj=nvWBW^L`03UVt;$>3{|jsY1Qw0qpW1xGLz5
zAl*sM`}`~)Os{$9H@-!1QCSw<#9jLLJz-v9#elzJ3*hk(w(rv4{o*XcXMHJd?0IfT
zy9s=!BzgbfI~UmY;R{_mIRfS_dJ+Xri?@>nPw%wm2%V8ubJ|z+{<RF3@T~*ZkKNp6
zrW6FJ=nCSZ3>L0kpkqeq-nd)>w&#bx0dD~7lXvbWZC{e?g!2F-uD;3X!vmC*^0#w%
z0o?5_%Mrq(?SIf1=x<eneIsuKKHe1T2%tUI;IMxKb3Jzdr<5uO(|>@wurM<H4~qy=
zb_f5>{8UC2QH0Njm@k$I=();o8_7lc<CQry`)fXZYPKGeU6k;$x0YghJ3A0cE)4Sn
z=jpzJyXhK3UffGIvp4;9(fK=@zk$-mq=JLz(+u8zJH1{=SQ#C0j*Dm#PaYj^Oy#`5
zL+FYEOp*2ge1MA~^d&Hw-?y1Ruu&9ED_j%2oUGIBg4?(xBOG!LTm8Iz;B)Ep^X3Nb
z2wLJ&Cg>!H%8-n!GuwNk*<4QYMBjK%n=h(oAVMo1uctNh<#JnFny99V`syA|jfc&C
z0({aX`V`j(aM;lR7sr3WOg~U2LeVAud+vWsb`joC2veuHs|t)k8Qa>ryn6Eab*jp2
z{P)Yd{(f#w=aK-Sgbrr3{MV`a&ilFk#gV3I&E7{2E)8Nuy+;m$UBx~-5#Q+8SEn&I
zqy-zw<n*ko(9eN(FH)KSC>orojfWYgr!Z3Zw6)c#m7ofD9C@l-Z8}+8N8&$OyvB3B
zI1LMi`iUd=faBou$V6*=QS;Yko=ooDdTw=?dPKIjMMWW7ALwdI_;UO?#Ug<6eQwye
zE+paPGuKe?gu11x@G>2BRGlod^`T&#lY~`?jnz6W5iH6MyEIDTV2Hmd0$qI$K!=fw
zL*(!`ia}iv@2&>%C;}M*M1635gAnn6n#-4J8y*B?LoQ#|h#$Y7i7gaN6rYwDZ|w+l
z2D`u^6!p0&_(W+Ezs7T~RFXf@EiOzSMA>COIO>%YXgKVaTmSAOjL!-Tj+m0f&1_VB
z%;#vbQ_79cHSLodP19d`lEj+mzb%+FJ%LiaAhW*OQQ{%6ytPg%L2Y8Ahe1POs4`3%
z7|G&uK=`<domL&(t;t$;6;dYN4b(<e+~mGE;`zhhI+l9d9?d%u#$Co8$z?rwqBLfF
z1K8`GB#J^^E4kMRmSRk}PBYpf?1q&0ptj&P@m`kEJNf8z*P`^$Rtan+mEeHeh<Y>K
z^BzyTv$`SFM$!R0`Zt`DkgmGSO`tYzhHP&8{3<Jq&N{l9fH{u)#)Jr2kwy$G_R7Yo
znm76_pkLK$ExdLd>Usw`k^5|o{0#}6JGya|oBYh~dBRQ#vW)TiI9CaA;wAp8B+V)1
z<Z!`hZlar(R%t8^C(R+p%!iXqv->hq(GLB`tcKQYi7ME}tTM_m(DFjGx-=S5R2D2x
zy*t4)WOKK{QwSXe>_y`U+lYFX(>`sItvNq<jGk_{GD_CpHqA~cKToB%;IB#9kH}FL
zLpAfOT=j0L@X05|#`}RwHY7#PLF>Yr&R>l7k1%7$a$4>gIJXT=P1V_V4y+&}-<t=M
zN^&_fAlEAhL4?zg$IjMLRC8(JZJM!Px{e{&Lo-%MX{&PIps5LDua?H-)-ervMbp1~
z(ke5pw<c7l3WHjE?(&wh*Spj{cGUr9;J0jGBxjw!56!JD?Krgv(M;{@HXQPB{{-75
zSurQ!FD?U<Qte2Cl@#@04TdPDgtwBGq}r&^yD-u{wj~wys}P2?4GSlnKLV8sB8g_-
zllJ5^L41=g36I`0jEl(IH%|$oN{ZghwfD+5guVZ@=yyt=iSQz#c)9`SH5PLr!m|&9
z5?SRmdv^ssRIFk@jF00`TA?|Kvz&1}7q2&Pnw+}>eR($Nz)s3e{L{vtkIus;L84%?
zCA>RgBasyMR2M&Ng295K%GY6V@0y#kBr*ZsO~SS+c87U69Mwd{s5=sE%v_j)Sn#=Y
zfNW!Ywww~Z#H5ry;26-pZ@Ai}JE9}@NJpFRt<F8O#42$5oWS<)Et!jCIlyR$f+YUX
z42!RpFLEk5Jzo3VRp2fNSJU~sIN=|S94>X!4w#1$hBvM6EYHjdKAAdQ)?+-2^@dO7
zGdV8t)5;5>%h-%7Pz|oi%;Ue1&M1?T``f?j69a~nDE7*&2i6o}=P1p_12=8k;uY<C
zgTd{m_PK-jI^cgkf_TXM*(OuxwaPkbar};A)zE!^-9hVt%bEW28@{Lg`A_3M^M9zX
zvHyMp{7>p@E5Asv|1y~VmoFP5h9c4vK)^p^+Bjf`^M5H!|392cv-Ol~APMy29(yWj
z<o*LOZI~YzhqssKho>h;5ni=Dh$^h!F8CgLqMyKhe!&0`L=c1b7sfZccm<4}owo%g
z;#=eA?p;po97L=cz}<))VC<Z#(67N%zNE!`5$jOneAsR8!wTc%Nf!%~!oAlWo*nn+
z^ZiP~)Rfu`3KWwMqY8A1DjRe%Q_W&0EB((^dZF@XsJq?F$Jq<RB>EhQ){bSKDZ&aO
z?ze$DrYV>w-BL+W<R8&ZroM)Mv_B_*7PgGd@a66tUFVaUm^97Hj>{mqz%@5bb5*|m
zCcW-42H5Fy0q$jeVu*-t_@i=OrQ<zH*7zCi;lsS?AU;$l(!EQa5+(ME?GYbq2JjPo
zi^xPN-&F5Kb}}9cQ*uT7g)!A)r8z7Lk=dh}^KWt8uwUTs)z@Pb1RIAR-l|BjVMiPI
z<=8Z7<gU(EMOUkddROW^hSRo(u(ea{bjISGE9J7Bj<k$cEPAc1J1%RN-c`X8@cPnD
z>KAEe!`u=y{*r2GYfFeBQv~5~>_6$lZ>SlOWL&CMKd|JTY5O!+xg-KDn?I?Wn!P)e
z)tVYAXL(BG=rJ#541sfq)P*xjGen(&QN~CA@n+Ml{6{X<`kggrSA|%V3y}g)w-t`)
z=BgoD>d7CGPj5!wX4sP;N>R2IF3GX8oW5j9FnnKI;#PF0ib8`Ef6(Q`vejMo5UeBE
z3fP`fWduXaGx{!|L`g0__v*=$We8{<$NwI$F*q!%(#a-p|L@u6`SI@Q@n%poL&YE&
z`}_S4Mc16bpxO*C7@X(x)9dy7@gv((_39Jvk5_UFbEvr`pC|&y&I|-Se*P7S(OQxK
z%P{J^0!Tq|z2~2%q}|EOyDO5WGDID<>)cc`QRc=7K%|-$bH4lb7(j&eB+}{FF9?Xe
z{5A!6!1yd;f06-Q5Pn7@X$1bn!4e4l+KxZ~PsW^L<W+^4L$A|*TE~Q~8B+(*8PhWh
zLBd7n2_`83^Cpdw*GyHBkLqmg_?gMthiI0AYQI*MEot(bBd)dx;lCgem(wF0NSH_4
z0z~s`jnAU&M^&k9x00jKF(`(nJInTy+aj@h?<ln+)DyNgoNh4TUYfF2=<UN_yY>;S
zvg~#@;Zn>FV8(dY$rX{&*s{;|5u&u?d>E<DZ=1oS$gOxGlJi;tuTlIFsT9%8FKRAe
zq?C9QOR#$&F{qkr?}$3G1GS#Jlpfyqie1&>CGv;uKq-O`NAV0Ee+`|dwqPKw{$^u`
z7;)P;NeN4w#Gcfwqceaz3iKE!Vcc0MTgY+n$}~1uEGsoR#@Kd>R<>BLL@i{MWe-13
zrcE_CL|-@u1PB1CSm$ZE7W5EsfI+5BGu%@&QRc4s-}PcsskSRY^EDID494O^O-L=m
zPZ(PzCCpuBF~ZcnlihE}fR_2%C913X2WEKGxO1i9h5O1fMoO@RzhsM)0OuW%H}YY=
zS!V`C+Y9yQDyTN@N<L=qU_i{79t6fAv94qAGRb`qZKsqi4{AAuaC1Ovk-@X`uTxxe
zV!n0rA1~?D!C({}I@73_EfcL@NW=|fM9^}>#2my`Jyp&_)wA!yKT7ke$uAQ$H$q3=
zu*n;z8ULH3R8$L$910O@ir;%;owEoEV%}C_6=nSU@RA(EIEbuEg&k`tNyj7H+AFrP
zExJohjNMySjgE$<IR|-`xL(qY<T<c8<Sz-Vs_uOt7kGLljt%M*tZ-R^^q0bhvL{Y)
zZl`?KD582B160`p=ERX44TZx)FVgia=;VsTqF9kr#&9??zqVvtdf3XKO-&pL6CwaB
zS%`29ia?~{u|O(!DW~QEQ~A@j*<GgSZf7u91xQsRpBK7$KoFniZy$-WIL^gsEU5uN
zxuQy>xs*okRl~Y-YheRls?8cJS}I^Slg&<NISmMWHfY<$U>3WS%Rek9>UCTN3Q<s{
zV>}gHCei@!AC@CR3|7uBfR?gyE4xdrASisNvA847G!E62Kmf@onrK-pcL!N(qRLCc
zk{>RjhQ`u2XLLj(O1J$jiLEVH-tbz#$YZFXgHMz;Xpb)P8TN3cW$Pu$d>N`_$W6<}
zIc>(auVW>|;LdPS)Qt*MeF^gtE%Xn2mH384=_>8(mmZcC9j)z6=w4A^3W3RKuRE6?
zO_4RKuEzwqi?7*sAnd1ci?SSZ#h*=6P}ZK;zg)Cs+=blCI^JvZaXd2(KMji?=9Qn8
zhpsy%G>>DsHd_kidew7H=ij<MsI-_n66&WWVugx&%W4am3dCb|m`~7LFucwm_3tHo
zqZB(4AX?w5;Tpu#5PEmv&g&685;;~4;OF2)q6toiDL2^?-@wR|`o;g@cxC@TkJtbI
z>frx&jIy%)&yL=iU&ko+1nl><9wUFBHlML;sL-sG14Mim^BOghHbNQyRzLgrGw0n(
zg0=-N$xko5nC3)e^24U$9MJusjspit7(Kihn~CITMlvNf6reD{&;m6gOqwBy1JmFW
z9QH(UaJqQ4Ff*elX`%!jb%5V6ZU6-ts!6Hg<3uuPMqnSM7}Ji3PPD*VVIv0yex<bO
zfrc_g)Lmkf;=-$N3O#0CNCSY`ft0F;HjF?t-9MOw)5?7Vx!S2j!hfFh7?Bk_MnW5B
zO+=wkRp7<xd%t&O@T0^EI!lTy=oLX0@Fw&?YCr1Eg3(l}K;^4ng}ufrl?>M#`4EW=
zB~zuq{He0R+=-zagrk_47-|7Iv9C-u<Ia*C?FrX&TA>$`ti29f(5|d&1Xf|o=l56A
zLlqPCUZJl5xKk%Kz!zEpph4>4g;o;Lftkdnl=upRjbi>m<(gWtA`OsCq0gewEg_&Q
z6xW#{1j1RRuosSAA0-gEIeHgnU?d#8rOu_rdx;zy<=URwPtKGYNW`1Y?;ljAOee*n
z9~tt#1X#U2xBi4`QueOC7tcuA8A1-3`eZ_lsRBB_tO6Uz&=%@`g@CFBPLVckX?wLM
zh;@b?w%Q&Lob9iq5`iJ9aF5I{424pqHu$Cn;v#4_oYxC_?<L5C0^g>9^^#@5hC=fH
zR4A^l$j6iqmyWRP1v8N7T-sw55yI>9lwYAAAU5-M(VBmR#IgZBb|3555!q~4hFz<6
ztiL$@xXEhMtKB-@aOXagei$<L+>zboAnE3^_+I_|`xy=m!E3S8Y_`oMXAD{(Ekc7N
zJ6?%7kT9}HE;e&FV^OxRs~h{X#H|IzcR@#faNPbjmX-H(<A&dP?fINm>fyKH)6UUi
zmJ~N*$A({*m)(|^*|*#6Y2(*rDO*QYHWR+dE=V!&(YePiXzj>R<|L4w*VM;f4r(j*
zuR23!F7sz6de+Z|PfzadHBVP@pZvA#^X+Vj_|qf5Aa`Q0yQCtwY35fswA?iDlxAsi
zE2kUU+<5oKwXD)|F7q(w_oWq8K9l(c5p_4SvNFNep12~72$`zD2WB5?m@w}4hpE8@
z;sr2!M?+wxJu8W6gG6Y#INiokNcsgN$T&QAN33e2cb|DCE~#qb@zv9LH8USR{rIW;
zV)JZ${`0;ruPZOMOJR*@_&f$Z$svqIcR73dK}9#8%F`hVy11IWSzMdjXR|w3b~bkF
zTIb>YE$Z0jql2phb*WD5M(dQ}deebzNjAAEobAfm@ehxc*~4*hn{LZVS`j{3p4Hx%
zCi)r3oZMZp>Qv=6x0vYzt9BZnry6#HrS8XV;9hQgZa7bZC((<r2ExAQ;EOQ5hUdhh
zR4Ub*F^|`pqMaAj$D4?ZN#u^D5C~pFU@6x4MmQM~u-5%qv4<(acl(>gz3U=3+pY*R
zDWtjb=?onRevgo^kB;DVg)qbhnDbi+5{Cx!wcXiX!j~cw0dzgUt)$_O4)~mP1MJ{J
zqO!Sh3uCzOh2hTfz)10Hfv|tmp}mNKpCpJGuqX9-PK>$J;A|nJoh_KA3Q$o#J+O>P
z-rsNMTkV%%b$K}yruruiTh2kQv1|lzh3vQihiNaRKo@lXR7lvdAwqw1G$<P?fnc~B
z{}@L{f22cmzg+%>F+cw9ga&Ngh9cPRGe~D)Ah_g0f4KBui?~>RVc}xRl;CZDxNr!%
zEk$4|VX<35uzXQ-l+8ROex`o6sd#-@IfAUZlwVpK&R!re0=PNoE*kE2D5fRJpk)W0
zfA@+-ncm9|gAwzW_uGVQ=}p$cS@m7CxnT1oE5aiKJg7q}K!*|#f+7t*F7pX2yusdp
zt*CO$IWCOT2Q?mHxX3N78|5Br=2$9JsyMTu!Ytr`52Y7PQ94M2OF9nlB;|1~|Ayh>
zx-yYPd1Nbcv23hBqn~^=O#_>uFazyld~#$P-3_EqAJ(tzsN>ofEHzyhV2<YB8D=xz
z(>z~8vm>G$y<=zmYECO-Zr+0UEFBQo&;yhlpl@kf8JiAQmXHHEraBJ08E(0xW?|J~
zo58XcjFfm<w+JfMAmvD*<Ts!HCav{q_q^)4L@R%LF*%}1Y=!P=La6NA1N1^vz5Wm$
z=L<n`ng#F~t#dc|;*>|oqays!|GI7!;cpc;fu=*p>OrKj#{cq$y5v&apBk-HUmYV}
zkn(t(*^OQJRFlwSAfX--Dj{YBdyZg|7S)R$epLr2sI4_ktdy<Jvb_PV^t2qbS`AA&
zeNHue&HMbdnH2F*oNvH7CEi4Q{~53r8>P94;+~caQTH?yq<|^d@3mft_2*%nK;`Ye
zx@9aV+~sr@7YGXt+JGBoo8<6uxs3mcHwt5^9WfX*B*5;6WIvD(x;}=%v8J$qAh$P*
z^#CCEk0yN#0L+diJA!+U&L0;&QGYD@aDn5I)^&;X-IV)k&hsjKH1aze`}Uy!YtuSL
zV~Wf~CRy<XV$`*Q{1+pNXc(4()yk?;|FO9n86qk^eu+xNDCY)#PurgRqO@r0j!22L
zAWq_V`#hmECscE{dSNHf@9|(~BW&RbNS}yj9{(Js^&uUZ$NG4)D@SOF-+?IDk^=CL
z=mu-7q9jBZ0S;pw5OZ=45wuZKtZ$WWw(v>x*z<GQ!lR?J=L%Oh3g}~_miCDLLXaAV
zvi1F$J%3nkjThvFPVFu3pDEhVUR>em$oEu7fU?!5nWA56pLQXtq!4>kr$@MSNPs5=
zn?Lf^$`@eMie{S@cJnHExgDG)d|u7uXT=t-;xC9lU+BI5GX7n#$(7G1uY;hNeE<4u
z9zCRaM;vT<rXd%Vy_u4-u#T-Yn`-nBT11TqJQlEX(sl*fWYJ4krB2cM(UE<=Wj-lw
z;DMKOWN<ARcfJ!Z7cJ|8<s8H23}u#lw)*BK<llX;@b1K#OWeca%4WVJ2ZkKDpd4JI
z?_m-B+YKJlY6n^hBXl=zsJ)@4AO-v$%sK~2KZ5K?3d*-Qf1m2jwN_U+n<e^!kmcGZ
z=xMj>+sMB-U*rY?M>!h(USEU;FGOf4kVh$^hJ+CWE-i+q6{6Ow7y$BQ+(8=uPJBql
zqb1o0(uV3hS-?Qt{<Gzkr@AIr?ygHZJ{?!Ra#IZuVTY~)2C!pwSf+O@UY_qZoTi77
zi3YW7jc1@s7v*A4ZP{^n@fJWHUiRsqMobk#em+?UYM+4^!lZp!fs&w*(&u}5a<Y$C
z;Df<Afw5ti;;LPdFS^B^8GEsmozaYLT5jRe-n<=dyv@l8`<7b(SpeLsT_e~%>C;Br
z@Z5I`ghRN0FxG3c32<E*^}FU|&=oJ%WDj+M$1}lbWBl(Mn}PYcr;Lt!Il4@zm@!jl
ztC%u_7^(0j%?&~p2hB!zqV9%*S}Aq_zEt3bg=)xOuE_Z6J;2{{80D!eTn$IGdQ_KP
zp2&)-eg3+=&L$}FLk`V?S3c9sm))Rvznd%#GAz>yl}#^#!fsekiF-m0)BZ1hID4BX
zN3AOSpX;{pAK)2CX!HLt%yRq(ISCU7<Npx6t1V$i(gpkbujrJIA@Ug%1Ebz+6n>oX
zmUf(lqetrdhftFOtwNFZ>P(Hn(Ok_fg+y}yIl)k4=i4c%jjy-I$NOWetD%jcP)T5y
z_uEJ5EnwO&Wty$03$!Gd?fm)f<SLJ^c$wF4=YCAmLi#)Sljpd~>&;eqnl^bV?^C%V
z?Pj2Ik-Uwu^NFSRiLoQ4_oQlaZOxJIMQ(A5NpLdnQ##|<?EU$b6+H1wJv+~i7}WEo
zFM@|``bvg-L)sjruv=XJh8**!?dO>MqU-0C_(nu|si$6c9Rp0#MW!9xgchlmZ!1@Q
zdsS^K<hkowZwL1aE>HbAY{or`n6qw8oM%?>i-rq-(tB{~rE!<_VQkcQ&D4@rsnH`%
zCpTc~SN=((V@{iMnO*heXm#>|Tb3`^&xPl?$*TCJyNX3McE+jd?WjOvRJ(^<#C)Q3
zG*IEAMhJzjTD0g>=7op9^#DR^xMhQ#%6PkFg<aF!6dSF3y`AvK1O=EGigrVbX~*Ie
zT|#Qx9RF?(3(=)B&@Xz?kR0hW%Eq<$?^^A}0*5XcZD15}2>IP2y_ZJsi%)Ag@puG(
zGp)T%QKEv^#%PCNljBd$tZiNyyQ+6Zm6P2qN^`q*TzhR1#HP-Gi)e*MOuPD&B~KnX
zq^5_#VtX$c<cbxeM7f)+MBs`UX+dm9Vz6tQlwc8dzT4Wd=#UJJm{s3|jrBrJtkw*$
znwCw3*TOk?1NV*1_ed+sk8Kj*7Y*?wC)*<jAn!{=NkWPkHRL!qBwkD6RtIbE3Z!*W
z%D>>$X4U(eI<w}Ox2c{s&PCFEcRMbN68b)0MEW<$trgt0b?S#<IrqiYVQZjRd{IAW
zNB-Q<);}gTE=RssxZ~BN&E8?FRSo4NtxpOa{;pk)0UwL1j7N>lL{wWH(f3`6C!iY2
z6P8S0k}kH*fhlzf8XFK@_h3bhI+|N!t~Ip#dvXG<a_iFQNf75iNfQt%7iMG_OMd~Q
zJl%D#OUCNWB7i={>oKHk#M&yZG}a+$BO09ZXEpjOJC#p?!IZS*mMfLcc11-BXoTuL
z>1&M84Pk4R3nMqx9@YoY5K;0Qfg8uUUHMgu2D62ObD-G40=AE0ne4D)KBY^TxM%9Y
zaxgNtt4mXmC<Zi>)fkoy*7;$;RNM~ZgQ+)o<P~th4SunYxBpxi13}CYNk#wscE<x~
zZ-xsB0QuPfnrtlj*l>oIgMHFv@;ckc6?>?dy4+`(-Dy8j%_9*xnPKNZR7H>NCt#5&
zN(VW4(j#3@qAUDU$z?)ZB0cqs6?>dMKqLLa@v}2Bqj+V}UGFdW*SEsWf^(*stv#2d
z1z&?dxTykt)e*}Z2tOV<Bc$>W3Y>=!<*H5_kZi>?{#A);(CO)Pf-o^ilB43^2HARa
zz4Hk2f&V{@y<?MT(XzE!yKLLGZQC|?*|u%lwr$(CciFbOZbx^=8*xuZoG<GS%oRB@
z=N$Qrmxm4CzUo`_COcIlHwM|NzmfEqQ7$LQEoC(EAltTB0c3)3h4921uyq*CTyFoK
zBLgP5;Le=rk?I4}t5ofQi=5?&x9}Y~ShbXd@}qA@3iI`MzM;X@dQ~HwP0v>l5>Flq
z$gQ$l2C@maFzO`WJh!H>4HlO7xW{^2YNIBgK3s}o7$Hp`QO?K6^KEG>89gLY3*WtY
zqr>5Ils<i>3>7hSrnrR$Zt(4G=|sQt)*K)<!$}d@0uA=-1zN|!D{I-4_L5vkPSeTd
zPMkN`6>ArgKFMXy9tDK2JSn|XOe7*#@N~R56A1%&#T1hV6fCNvnf%gB>TfCPr7|7T
zwv=J@yzXeR%9*f|>jV~;)(a6Yy6Jg!3F@O3dWLwDfNS6n6$Dx+<Sev(0`X>|ZwKER
zZwsG!h{mj^P|WO<;|!Vz;Gi4R5B>AML}%&;AWTE+eAB>cS~``wL6j(~Vd`@uEI_?f
z_c=D6O4f-D4oL!L$giN@h1xqwPvP{fEJf^mvgDgr=A%|Rb~ucy3%4OA2H}%3%gXg^
zysb-JS$)B(+3sE2_eH;9->>o`ISZBcEhM1=aw6Jb3&zA-M-rv+U_?*)*@k<0GD~3_
z-3^kbBjHRRS`<M}C*hT*z0%(&k=4xOr=L(FCdpZ11A}Sz>^UJr31e%9s^E#aosc=?
ztv4|G?UuY_nLi*rp1bj63b>1EMoSsIq`~SLF{rxn<vyVID4bokRTMa`?6sl}{0$4~
zI-wEVqrF4i_;-qz72b959xh8dmqd34WErI-G+~Qy`ED8JA)JHt83nl!qk2gKyXgav
z++^+AsMIag>Xh{f4imES{m{wLGt$gMk}~rdP9l!CuR-EZNn;j4-4lbN)Qny@NU9hw
z%3Dpr0ZZi{UKCUfmwHtWIWq2BM-aNu_FjOCkJFB<o_kxi5(|w*wZ^&gK*zbw<lr4<
z2LcbY>CMe?>dy0~jYgH5<6EUs9&Jl3OroM91N;p8z<BiWdB1xQm1}Z($FUPc@_oL8
z5Y>qoqq+ql&`Dt94-$EtCbMCXTO)!Wg@y$?kL!~!F2h*lJ74Yzx|5?DJ#>$Sn2PD2
z&o9%^NT<2$mY0C&E;a-f_SAYehXyWllyp~E_+A-BA=;M+0j#eDktc0G&Nvz}j1si)
zX78vb4I2z8YP`E3^Y)n#(PT!AwEMi+riA8sl_v=_?Q}xOzkN6ISF&;XhKDocAqvGW
zpmh~xiG9PP^^@Q!48X6)FkZ>51E<P#&f*syCK}}WN_jr$_y7GvQ(wuRbXq-nmZG))
zn`<32qN1`uWMb-u^Mw*hk)`MeN}?U;;dK?RB;FgH<gvyoUmXSK;sn{BZR}AeCT(kY
z#n2t&FKv5eDodb`Skl753{l-p6cSPJv^OGMk}(%YjEz0m#k%Mk(6NbW0?R*2o&arl
z1jG;+gF{`7zW^^o>`Xn$iH>lDazk$?Iq6AUGk%5y|INSOpnfx_F2i07pUNgiZeG9s
zx1>!Wvlw9GX1cxVT#RaY_^H>>7)Ht#S3+jwkyj_){y6D&y6}BORl9<XmxhBt$(|_L
zL$yH7l;r(UBm7yc7}L5cM|ohr9v#_izqJ|UjK|Ep*d#O>N_deucwKeHTH+(vyvCRm
zKXkH7A8(?^#(r9pMp5y=XNE)hdeF@AdQILYW}!tvpeHD?j#p}+&rt-C6HEwp(Zsi0
znMv}h4sJ$jiYhA>y)GJahs+88?kRgPk$)!AXhFPsL!_N^A*ECxi#x#v-Ex+emgQ-!
zhxFuQ(<54xIG7dry72`;pW4Gr2(d}0H&2ca_l>2<t~eKkzWctezQ#RWM|~$jpVg_e
zMIrP4+Wc?$dNlool5G<T@W~X)8>ei!`-AuZf*V#!&bJB`z-)&R)(bo%PMHr5+>e?3
zI1MA6h89jFgBI_6$kARju2=h&s}T4#{>h#xLZR}2E+9Gu0d+!|eT>d$`1P<+f@#z&
zYP>Vnc$jy!x+@?a$!r1QaRa0%b3pjio6q@?2}<LHPQue+QVa%<D$I)2sgTC3|35Vr
zqA7wJN`ZOmtx+@wASmpC8f+_7iX2CFO7i?ClyZ1k^Y{dM_T;gj=s+K8c=R%_eE55E
z>ceZ0D6S|Ybr|8<+@5>NV<K5Dx}*-m!-J$2p`KX*h(qR5MWG&#^BKXOw3!gEP!urw
z6<udut|vV+G}*yg>(y+nsGmK+ckJLmh6OwL6Tf}@tv`HeafsWc#IwLA*1Rf`5<2o*
zUQqHHR7s{Rd0GV7V?)wYiY8gH->SL<2_55mkK;Mm=h(DwNww>!z$r08^d}zKv-pk@
z*IyUuhr;TTkupW@_zuM9DQd_O6u)#FL-Up6rjMUih=Dj?+lR*KwgDog=L<G)TXW^(
z75<l}Z6x4g3v1Gz)$W*~kEhkq)}KF)su4^7yIp6eXZR266l@F(|JBp(FIGEd^Z&tW
z8>kzhp!Lzux%>lg%oN=Pb;Oi*m@v5i38knhC>w6r%nqc^MAZ1WpKy3(7~9|6|8#lv
zT4q<@2d~21>4EEmBlZnA6cGvpffpUTeg3rrkry+1W&b}(Pp+e`oZbt2uQ2#MuM~Y}
zi6Y>huQ=5Uy4DN0)3)VMyZ$$H2H7v3=0@!O56%#!S>HY%2lz~lif%|fAzD8d7n}RN
zLdH7g|HIX;0)QCg<AdA(0|;UGmG_%lGg#{zdK1d`p#)WZjf-;Rhe8rfkU=hqA#ez4
z*fh6-Dq^k*S#LKKfCLpu&D`<1UA!ycKUOgyrI~ti%LPFb7h8+h>jS+#anY)vOP9|0
zjNO-NC!@@MZ8yNdbCB+yu)0V_6u2IDeJorb@rC0>9_$sMcX*IcPM2=FY#w=(i~{N8
ztO-41{)W|U1XZ^pQA(vz+(yS>-H}A*%+h5TaWp}x=*IR?KE{E<JdnCrlvrX?S-TrS
zHr=MS7hKBmLYFRP#vR5n1TUH97O9&g6hEWEnsMqSV>;1z9erL7DkIVh?c*u?4&<$>
z0k$d}D_vAi8owe+Z~L_lusplkwqO?gvUC&W5@;4E(-9JsF^PP(z^W5D5RiT|rD|eq
zIoh&hl(pY^r${12$GcwV@}zPW?4WW*S{n8D^wcSvpnGd#?NE?j=6-Hyab2p60FVPM
zx$BUPMJpT~Imc1<Zl3OpnG~^6bMA34N$u}uVqebf<Z{xS>bLOw8n$)AgD#_l*_4|p
z3BfbyB+J2zGU7sp_m4eLiJ;*3z96Hqz8Frr&6>12FPOc?V3MezpnW#-4aJ|+w|qvG
zNGCyV6CbPaS|!PAc`xw#5*H>97-KogU?LA?_1a)RIYm2cj#(0sa>GIhCCCUGW6@?2
zu-V1{CMx+B<m~V(yhDv`Xcz0`J-cgXiEmLzmVpc(CD`ut^eVKO*k;<yVO%kg6xaxh
zGO^LSYW(r#3PLN?*<_vZskdzf16n0cQxN0g>N^YkSN*%P1=zNE&jh?u?C+))fHA0X
z%8Z2A>wtDD1u59X6|^m5HMKlru{f;3QWnk($2K(_e^aSF=`jgBvY?Z|ZE0|(<1{#W
zC^m~MJ3e_zYqxkj=NXgCmRj)>USUCE+KNW~Nrbnzi6eWeGa|^4%tR~2f5IlHwJdbp
zOOSSJP|-~C<^t-X<D1ofarwt1a1+ID4fh$Ffor?fEnsE*_+=KETOK0)4k9XZXCB4-
z_(V<5U*QTW70vHo*+GG8q%LGs?F*tsJ%yfHv#;XW(a??lwdtjsl-<R0SKa)vuBVBa
zo|B_Y+<yGzirH0NJ1g#DD#73zm(i-sp#x><(DgrB3%VdLPCa;AmQ;P`IjGvEF5MT(
zMst<Zm&(%kfRFckeh+zfOf<ZkOjRh%F(f(g=G)<LFV80>T*pDYKE&i3HBAZ{hm4}=
z=nN+uZXm6iGZUXzBbU`Rym+uo9!6=UxnZHr!UF1YUBUcIF|-}9xBooJ#scR#HwSS&
zt_PO#4gWM{roHP-iHV+k-Wd55Pyd@B!*Q8nYt!CDlK0a+P`<Lqub~H!{mArdBSh5a
z?Q<+3xR(Fcfk>A3yK#rp@URA-T<>!na^_M_u_2ZCu69+f42|mq9F6{9|G6d8np*~~
zc78-gE$3r-J~J7zdUA5Egk=+&WjPgwhBiUA?{%e~0goHu)Frr?G_tC;fi_f7*VSAQ
zzC5#qHqyCIIJ-tEt0E4nvjDw?AlUTM(fqdN`qM6r|Irf5;uy<}ZevQkO$}1w+pi2&
zwm`DHSa*~cgGf1Qh`pWJg}q8p{;>t3ofGBMnFn+PYI12ZF~NHZuwrjt?Ry;T^4D9t
zqqd%b9J}=E`q2E!1D)5Y6p)&&KmV~T-EQr5Go~!%q@0WhYnl-XI&7W~1q&rvMs5eQ
zQ4{CA%xCmC%ennSFDcGbX~A%`nXGn~fX1~8G_SOQ#G3WQY0xpkmb&YME7v)j;I8{$
zjGG%SqyBeg-L8gUTw~oIOoYk%w=72Jbdk)m5W}A~o)YAhIQN#}$D<D+O9s0=b@+0R
zo?UYZLHpI+`kn;ACTRAB>BsLF+Wh{FY}cxFY`_v{OUWI!?cKmH=kHL!O-%6*!7{}(
zHzPn)8gZOp-v$jKWBCrXl?;t_Z<4{Q>nj&=EKyJt$z9rpN|Y9Kg%E+gHEo5T)Wz=!
z%>*CUf7Qyjuow;z_lj%tFsY-hAM~iL0zW{-P3pS;N&Nm#X6yeOq|5pXy!vl^P5+ZA
zamWtceM0pKq#zu=1c(e-#;yZ1WsTVmbIg32ql8!T^FzO6p`}Zed>&YsL~W1sCO*;d
z(VDouTDtmDnp!zB11YAj^735><U{aAC!WK+qyb>S>~zD|*(<9^IXFwjM_#=0ss6ot
zCw^iI6Qc%jH)#83W)_v;a-T)BxO=0BCqu0~M{Ms7@z}jLRTLC5Czn^A^XKb(G;wM`
zbvy~x$jg)(vP6Ygs6?C^V*P<h63av#vgwQf$t2JV*+d=4$r$IYDMdXLpYGY6PHD(e
zUE_IlIkky!J{++}8;5#b-khjvwYtcv#V9t9Ncu=PN2~?mP8|M~R94tsC-*o|+H6)q
zn?=<?!)qM!VOE^A!Tv0}D<?NPXpt9WY(UtM9%S8gIdvP1rZYODTz81hLWHB?z26{e
zwe*nqE?No}|5dyBqG;|{U*_G)JM=D5CvAIRW&|g|t(B=p#~=Vh(=9Pbij*FJxcI1w
zlw8Uklxf#lH9;AzrwsLC9Ojdgl>wv7-|}ZW%Wko_ss$5C`fwjDT|jj!3)qY(=8~3^
z0t>UWo%GdIRXq&JsAXL!Y)?L4K1qP3Q@-5#@|XuWT5Wzjfk><PZaP1HV`j(*DEMJ&
zqSNDH#Ld{GGwdAb&FykIJ%Kv>5XWe@$LdY1xR}*)fM(Wa7rUmgIrtl#ef<e|7J$Y4
z1~TM*L6u9xDDj5lSO(R==DYwFRHP{++o%&T;HgKyCYL9xjV#xGU|@WDkqj%+SpN-|
zG$O+uX2}~_aQaH!i<Zuh8}JS}lPIqlP@DQwtWGW?i_K>ZfCt{ad*#?T8SMaf2Jc%w
zTugo}`Z6t%0^-WN%+8kn-+7fiN|J7U*Q+nrql>vT%U<+X=um}b66VnJkB`y#4pY%6
zBXBt=e0qvGiL1)+!}G&Y(CJz6QZ8jRP0_ML7KQUqdGRu`&&<)ac8?6tR_<7pWj&u6
z9b!6*v4}d7Ja_cvdDG1!_&4+*%etXSiDzl%8yUpY0q90>B$>**@>Oik{i(dg`O-or
zm@1uwmXKTZAcTwxLzrxPFgqqBsnz$K4ayV+?+iOCRO{i`mMa>x8fi)R9LVg&)NM)f
zkxOC<sBfg)FceFxdZSj|#6TV-zm_xN5JsCUv>G7g0j+1dVetI5;ds1K2{QehX&ZmW
zRA?gu1yH8@$*T07*#OV@k|d-Q%@DXLHZzGR+1GEUz|&w*Z{z{NT2zi>A6nt3GBl0^
zA9<!I|8BA`?1aBZLq)OOm&y@h#YEO9TCLNfVV4u<i;W0a;e&W&GxY%X<*S{uLmaMe
zRwzl^w`A`xMkY&<6gZ4`(yDE+{BzG@;DGIORxV}KH&x10y{=D_+wg(9^kUPEqbdfb
zC@Z*g#KLShDU1Uegq-Pbs)J50Etu=^O*!T_ii-UT7+v$0H6F5gGq1(jXhF(6Cjj(2
z`s8)0=zb|sUDY`p?1R%Ho%tPFO<3bb{bL@CjRrG_lpCBOdzyf&A?vJQFYy&r*y#z$
zEjdn{50a$2Y&6dLRO}W&UQ@4Ii;E1je_F9=hNVi>>+Mw`if5{4QFPven<LV6SeIG2
z(9n#?P$pfp3Nd6E*zJkDfi&l7Esh{BomYJXy1GATQTzT$$APH=X~y%vq;Hqr>#prr
zynrI}B)lgB<+E%jziY)%^`DBn?K*fVXQ*8UgvforIAuqqg;eO2c0>V8MN^yF)SHZ*
zo-A;9<6}{hU<j3=*ihA<DCVi0JRp9Ia-H~vF*85!TZ#=uUEGr7J29bLjd*9^*~!po
zkIB|qKZ#la@sYJ!bG?`sGvQ7Cbr{cs%zloUq1_^M_--x^-0c2Ww|M>Hj}M-sNBI?w
zKemYN{Y(xH8gm|UVVYf=OJG|VTp61mvN@=XKdle+qJQ23Xv$7=W)}!Zr2^L45HmSY
z4j3C?20+vmxC=b*Us(X=CeInQ3A;y%PDU2Tq*2(18l|;oVOnzUP45Hd#N}hUU1lnt
zb_BX|)QGtXa*#>qeVe#qBJc2a2#-l%b_94nrnWg@M-!F>kQcHKqZb>wLxMJ}ccf&{
zRjO3jMX%ebk?1%Xx<b89;GMJkyAlB(!eidSv2|78e1?m4ej>*eTpAvi2(<?^s<q5U
zGAs+Tus~}L`mpS>Qu!o4(yrjmbX9gPc39E(`Hmk8ZTSN<PcmQmpLC4*KMZjktn~k#
zj%kkD5V8GFHqtxTl!1yp2=t!>Y_KsJe_H`rfrLeD#vfmCd+UVmGLlV>CK6}I%B$r{
zi^;QG30Xcqot@mC?)Z!<J4|ZX8(+EKDeebkoGm(s+K(Q1@#E#}1&w0<)_Gq_LSiWO
z=q85m=`6kP4t&n*PRVbvQ3d|^7JYy@RXRh68?yEj;?o`W#l0t0Gz5iP?-P3JEcQh&
z$B4HIc_*lL-|+Kk@nf-9?&(xq5R1>eFi6qn{ozB&Esr0{@C*2}JUZLwljZB{@uJOY
zTlyY5e-Q8c`5@Q4?LliLC$);lJZdHTc@DgK+b9YR8JFaWB1O8fg;}av9Vvq)UtKYJ
zUzhBvl?|+ts!?8c-QL(shN@9wjn<qbe?Vh0mt4>_RHUquGi(J{HTYZl!Hft1ZJ>TW
zd0GvAtkhGJjNstm_A8o>x1wuk_@W@9+T;y<1LCjNp^CepvYhsE@cIjsn(|65bnU(-
z+8m;k$p}Wjf(k>fB{7r4-51B($Yw{-1f)odWLstFM{)CSDX>rFF>mDffN&c0+9(Q>
z?aA~3^Nl%BdA4K+0SJ2^!-o$UG3VS>Geoz#h>nt9QRx=Qp*aDHW|BfD4CQ`hRMK5t
z1sm?0O9xqPao-}|1|@f`uIG-*6d<yo#1idIaoboLZ;Zcv%L%~+8MO3WfEnYr4LXu$
z3L73(O7$Y2_-lnOmOziF76^*vdLOG%wx`!SnG)WGY_!b&&I{0`w+$I2$aUROY&X0f
zqcld($vMuc>>Zs5A6JJ7my<yOOs3XY*CJ7aH+#>j{tfW^Q}2uBRe9!^u(3t7R8`sh
z)$efCdT2s8beo0M(H4HZ5;~xgnI0pl#zAQ@K8F_$Iyg16NCoLoEyUh8%rvS(PYB#%
zS8-i<kmWReC?uwS2&_g?)35G0kMK_ofy#9}|8y8ULdO|uM<^`3-BkZ3I5<L{c}GMw
zZO|}F+pa{Y!1x`j6v^r9@zugb=8T-cCt1Q9gr+Q&u0=N-!ZA;{u8+V`m)lRUJME*%
zg<M*ODdO9r`iCiq7Dtsu{&0s#{l>}{u>%(4N#mc@c_*3PtwtBIuP~Pd$Jcb51xtl6
z1urN8^%OZU@lpu%yO4~2_s?HI`(n;&ka8ZLCz6UYa}i=-&(5<8%yKtc5@5xenKvym
z2}0jS0nKaINQh%BrGC2cY){IUEVk(d#RQSM_pX>ToRe$bK{}PIOhe?uFer%%!Nx>n
zoD^)k)imf{gbmNn9GyA*5D6rT;lg2(B=d(v%6$1Un;~-I{ln|L*1-X(qp|rH0kN9{
zaT%kjv}&wsX_xQu0-FT!;rPUsb&xW^OU~m#^ZL$xHkoZ$RW(CrCmhXO@dq_UR8KCi
zP9hRce>~+r5{X<D%yz3}iCp-MB@JWT!Xnb#vC+FU3T?trZHZ5s!4;?Mx=diK#B_nB
zBmv2Lf`S!W@_nu5QqJCoX*fQNNUbt&>taf&LgMC<J|u;(Zb~Q=LBVZC{R3uXpNSQT
zZOx>FAF>@ldCpyIC9Y=UGwbhP862>2WDY8R)dEo*iHf%p;8aLggE3|IEXv&#cSju8
zI&R1+e)q*(b2P+sck#JGVQ-`t&@KvEL!DkgQ_U;`5<%?+{-lFPXgcrb+;mwS4Hl**
z@5mndz+218*7v6An<SqDDW&$N%fety5U7R7*36CN445J+ZN^fqHSrWf8M?`#uS-c?
z`uAqX)N?UQQ8rBlZ_MTSRJYPXrxjQdD<{r4SN80Ee6Uth8#V2#51bDyLhpjDaW4#1
zd9Kj>h{MU>cH|M&)fEw!@3_bel!75KR0R2Cx2YyrYpkdbG9Vai95iHNfJSG#&J{?R
zAs@IUIaiwrMwV&zA;bJDnPpWHl##e^*3L0*{C5PaoJe`FFzvAcU{TuO7ScTyl*wK+
zCBK>bP{h%@@Bv}goE}!u{A!_-q%4*zl(Zk{T@!-(sW4r)NH6ly38!03^$@w=qFKUF
ze~9RQzPUR-m#1v1nHo^cRhlpl2h`W@$}?B|n7HVnxmg{u8lu~h%(0z<tpAHvN`FB+
zX$3Vt>s`mi7^OLmPEi*?=}iLwYw~t`35CIl073Rd#^{s~lO1it;IuIT@L`3WeG>YN
zz$*K4gl(7cWwmuznZ{-rh1B_$o5q%C;(xDz#WV7oK%g}@<(Wg+u71={A*cz;llmhB
zsY#V~1v?s&`b}e~l)x4~*Y$hvidIP?0(*zYklz(VL9@)fc!ZJ4`)~r-m9<RDzf1Ye
zb`iXUNiKuHM>LN{2<qTPH6SMpZAYHLdwcOY3vO@V2#!9$CsNk#*$WA+MP<ug(^3zz
zUi+y60C39AeC7)XfKP(`Zk<iGqP*+Mjd)vSpU+*5GBeSZ{8tV`FF?|m*EO5Vp&2Ok
z%|Nq|c$xVHNZahgS_FU7gkXA3Z0ojKt3z<Kqae7ZV-;dbx!LN?EyW828clP|&@#2&
z*sg?(LgQl<SkGP8pueYO0g0Ivpsav~($Ay<SuOdgY7!~^cl`=T6-%~9bORd<zTNOA
zL-pWs6H9C`21yGZ-&EP2DQPrSbg8!x8!#T|i_flx8UH`2KI{Kx>oKtY<{tky)9bS4
zcnp>pboa%t>X&F6#Ww>0{b>~c;}FTvieHNVJRbV`17}cOU88DgJ$aX3TC6FK+B#Yw
z3a9tGQ_<r7wdS>fg_(5^4I3Iu2Q(Ix(wk>IcfSXa#6s`oh4t;Vf?3U7dmpzlGIns-
zn963G3+<-yhdOo&&?*;OV2j8S`nY@Km`7A~KeJ0<ht6L6`FEGI2>vjh!)~*e*Jt;X
zuu8RmB=&CEchlSIdz*QFp1j2%Z!Z~!`}4~hi40s_0}b>x4;ps&2cyU8`?Q=@SF`0e
zR4VjFFVyDHQ9*aEfZLkpLD=04r4g?RW@4zkirZr~Lqww~RSZ_VfHY{FfbWV2yLe0$
z1IZ(zKxDc=6ERt_Agi1r!eNC^;&VJKY?nNevml>qqKwR;d0Bw!piIx&#9_>@Ej-3#
z3nhJ~^0A6{oPzNh(%XWeJr0bC-zcd~R+LvS?}0)sY0zP8D@|->NCdGrk3$jF)HF=<
zpW())ZEjV}{$RtvWlakeW@%q~4xW5zE)_W<xgcBNs}Gn6goJ{RTzsqrJVlxQ<rI|D
zV`Bjx>)<QdT}~ACyY~}?AqofIeTrDDOPDR~t_JCvN@tq9>N<M{4f>bAtFg^IpToP9
zr*XDb?!|3{BYHLWo_;)vkXbr`yNv8`(M>5VF^98gRxErk-xNn)gAp$FG0rOb@pzLu
zwBGd8aF*>fK@4Xl>%i8NG9ARrpwqaBb0O}<7zEEs-T6}*aSymnTsPsroB($^|2`2q
z7wHGrAbcob4Y}ya8rNNS&Z?f&PdSrY4bP%`ihc#i-m?8wZST7A-THNFZm49t=;s(i
zfJxJ6uM3UUM|ao>jn2R2a|*&8J<N4Q-0c{}!3%vD6cARu@1UoECrKcu6x)Av_No)8
zSC9@9S!8?PLjuJ?&b!kU8d`};l8#8H<Qn3F{@%$lB@WeT)KeOkGjqSO=(HA&wRT=7
zEpV)6*Yu1!mnP}wlTtUkwrcLm53A)~y@n(<o7J8qe+klR@9&fI8g!tB$%W;N-eNge
z4_;DKz8p9(P@K+gK!#GKY5Kp=D`n_HBRf@%{k7HXf%|4D-_ytVuGD<00qiomzAvW|
zx%hk?R~yuy>H6a3+;Z*f*O?;tj3PA3M80>a5MoUE(8Zz~O!(Io&49v;x`>^LQE0V;
zQc;aU)mO5K5@i{-T)4K*(ncIP%!`CTi8mtUG6nllHB{EN!A3~^Ma)+tVc49C?7&C)
zRBDAgx@NgS7HoiOap$Ijp0pW*o2g6UOkyD2GUW|uVqn6E3+~oMg+}*8AWyb8$?G!P
zm;9}rdzpZQY8seeUwc)Iaf7V5D=DP>=Tw@i+OxD}T@aQIq951*lCKY~VWYI7hf=RV
zKYnZr0{o^C>$7c;%Cdad0QMiRGnyo?+1o#B=0XnULMj21%JRi{Mo64StH_>UT3Z%6
zDUETrGW&K4)LaYv^*5#P`-K-7GxUwtgvShJhr&m66v|jO6ik!CQU|%P;^ZsiFaHYe
zLT+dQjQD1xGUo`%ld_<#_9V*TNJ6iuS-KqX?6JRG8bi&w(Dutumi3(LCxGe-A*uC$
zQkZpB8*HfH7?tj`yvxBXL5?U>kGYaIt=)YUDOfFrJ<@%*CA|EE<jqT-5*&2?JP_;A
z09$WjokN&7G%b?89@7Xux$;^ews_Mi;Ux7)6%%~aw?Q0IS{_6G1#LK(1^p<9T&<p;
z1>2Y)qpTh!1$#lrklSSZd0+OM7m(!AuE*rJJq!Ru6B*Pv>_vZ`ZxB4pH29goG$H`?
z9|Gt>G6B8YBy-cnJ2nIaPr(a%FcXh})Lx%xWj5oKkp1If-si-%B9Lz_Mg0Qk!JX!E
zoeJ<MyOIlcn$6D!*s;Zp0$-amh-$P-x--UA0@0B;O*WTJ?SbHQlAOx_vLr?6bTSXQ
zov!({9o0~+$Yl0;)pT7?R7El@N4;y79q`262t1N5_2=Q)QqN!Gm2ZGrT^ZbRfEv9r
zst4NS<$%wA$=UppX1281S;yJX72y&P{Pc9w;HAol0*)H8)oGCxvic&(vrk3~aBZOw
zeoYGneznz?!YL!Aw!>(2=P%j%I2J2}zt5ltenJBO3>evMn{TA~4-jucNX7qFB{mM`
z{{s#4|G&(_@&DyuvHv1s|2xIf`JWst@6}p$e7dv16fQijPAy>2tRh(@_A-M~5iKEd
z@`#9J&H_mV|BoBbV_jWENb}4V8_T!;m7U42wCwD3N0aN65=2FMS`PBi1!bY-JQ^*0
zR^NCajk^#KSriy^e@DSW@wjtu$zgCj8V=+dF*Eft(eIL=l_4w^c|M{MMAKXfC46&U
zm1F+&&J_LbOQHr)N)}HjM&@(THu!U@L&P-62UnIlc{bDp^HvzCxcahIayyBh8v8v;
z-bjf(6W$eOeKR&fFoQei>3}ffDu3_D78@&=zgyAP0uj!ea%(=VSAqgat~6?fchP82
zl;b4iAT(ExLuLYoEC^0nN?a8XnW}|RuBTP1iLLZfBzGynSxaI?hz&Pr)tC*By)ALg
zA1Ir%?O0Oi3Y~A@L>4K9(mIvSH}osBz%)_T+g@oCh)A03xMBZ#|KUeZui`_Ju5tKS
z_cK=Z9_R;qvp_Rp2j2i~%2j)UkwAV}oN7BfvuHch6Xar4RBF>xn9weNs@VrSL#=NB
z-_9&Uq&M$v-JV*8{mhM1r$u<A*TXv}beuVq>g4ikvO&jMe9P9;FrAk`Ha~Y(<rnK#
zpkg|XU}gYLt-a(U>uc@6kuFodb;!F|e$}FHR$Pc^<sE(u42)+EtGR53+<Q5&T??<r
zDt6H|pF{P}<rCnU?4YS0bv~4=*nqC_jl4Hp1C%>uIE_j-zlqj?1<1YaIG*UNs|C{5
zS(SpRvniCgJT9vx+Ug1ZK=gcj-M9As43<Wsu-YmU2~|vHuGY%o@_4At7RBeqyu2tc
zuH`iXc^o5O5JWO5Sv|$Yl#2vkxAF`gW)5Er-w!W4vBKeA=7f$no>{Qy>4IM4{oI;7
zZAUIUvA`|$x;3<E8$^@+6O;c?$x+6RuB`C%D^`-d&|BGf2^EVaolS{ck9<4-<yy4L
zk@C&#R;?iZyyjMNsD7Kn`qQdCOfENv0$-K5@%HL&QJX$o@JhZ`;vJi!E!Lvh>b?_-
zUc-6?+(|__jh+c-bN6&<KO!yrycMSEpg&tX8Mtke3l-L2alpV}VsbMZnVrs|lTHl=
zxu?<Fl3HDaz%XsnH~xo@BZdj>_7oXbwTot2R~v`U+u~y6q;K%}YUI(n*Sa}fw)8jq
zo6r5U)Hz<xtZ8Lx5$1k}D!2*E9(*J;k`QoAb#vnQf#15GbdT0&TbdVE4{;4gDj~{U
zH{p!b_>pTY?l$Mf2Y5UUr#gP$D6~4u<ZWaZk#GwjaA5)9PX#Et&B1t{ys(g$1tOH}
z&R%BNL}vv577M>EilBI2hF#hLME{!GSsu^8va~~!MJJ=U2y<15-HUb}vfG&%<DmsO
z2|m=&yncPV2JqGuWc+#)LaL(Q3UC4b!FZe&a0^GhzkvMM3gO?r7-&^|;qh|zd*gB<
z3jy|l8U8{CBQm@R2GjNB02Fn)0XZ8Q6&k-@?i~2>S-n;<0gPJLU_T`b85KV=NUI>v
zC|BeS#2A1-0S2gh=wa^&UAAfnw^9yeptLH&$)IJpa^Q(!uh5~f__&(H6FG%>10h0n
ziu@W00(s0iz+tgDfM%jjaR3__?UxaG>-;C+ON{-96$w#ph%r~;A$xgCgo`$QfFDQa
z<R-r<KiSUQfk_hz{MI4(5Gfe_buLEaV+Ii?NJcV42TyKf1giY&6RIJKgr`?Mjb&l6
z&?eR6{Gist7|bI4eIpF<r2e_fI2eB0!Tr?mtNWSYOW`Al{=|rgn6d%j!KTM+Zfc;G
zQN)1te}aO5)djdSp!K2W6Wg05A{ymaa1i1o2^<6ZJ$B+S7GhkH$7~F930{u-@ac!0
z;7nL!q4-e`qx8A=5QxtIkSpkfM7#pSf%w+*)Ky#Ai5u~Yry67fqjKed(~mEpP)Iuc
zyQpV8(IMc^>cbn4&Zh@tWoFEQrkAM1dLN*6>!b?6SD+qy>7&3(L7VHD7rrI(+qGSe
zqZD2@h6SuijmJV{<eymvI&x!FzWh*ivRgdlng0VU3%Lh41kwl7mvpxk0PSJc#ULR}
zb^EE)%I`L#LeofZHE;oYkPgHO8n+bz054|@=V+JJG^A@)8J5x%D|e%ZtYMaCpfJQI
zhZ}~FIjwf{FMsu}Pfsr9`qJRujp@V6xxOuk-PirQ`{(0+*}~cR>tp<|G;+BzbCF%^
zlkx5CIJNSi@vSxyucT7Vw@1@E_S4xtCW!O<)5GE2!=5v}cuTe(0cd09p*=fi)gkBP
z%*n=u^}swQ8>%7XcOT7xRp7b1i@jT8x|ZQPvUDi|o9}3egh_1kv1g35`y~N$(e~j-
z3n%1r_lIx<d#=%Z=;vpYvp+N$)Gz6ct=PXt`KAB&JJd!C<0K-Q+EvL=gT=}`F;^y3
zSdEo*ooh?9B1#8<P6e2xrVnxAXR06{-bE+Qh^^G$E?_ooSMlNiC4^7x8r!Ue-VYEF
zbAaaC@wUG{f>y%aYA$%oMRj7a7=gb#^Ga=5Z7O+!p?K;=**Jst?ajpVaq>C6+~67v
zDYjRQ<>duDflpcuT!$M+EJxg3rO-7HrL+bUxS=UR3MCCHH?>^<_H6+Fi59t`8U`s{
zVT%SBKj{aiPs(mT*9`g29)K*qr@1Xj06y4!+$trorCeF}Zi=8HLxk)x6vy+Pn%oQz
zX=e6JOMrPFK|N+fwID$w`NHJ8kNQP5r2p;5Q`C5*l=)M;cr^U@_=E|8`Ea*b;)Mm)
z5lgbUch!us2@<&bcrL_UX7fmg0QOh~T$BRjL0AaXv?}C5%J=U%J*nUik};U=dpIw$
zEwSxgh#3q)S3e@|j+M_wSA2fH=k4)VOAf{f%mund2L{ETg&KYK?V`v1mTOSl-4Rb8
z5%qA!bX^_llM~0&zpS}lcdPw5vF3WnKLBlldHUF#Wa=e$^p*@NA6?J*$RzWq4#jRN
z4+g{7#Cm7weNxPzml>;;UqGDxA%qB`OEi!}mam<qcIy%TuFZfyD&yjr#amPpp2l;0
zt<adO8HDUhK#S~X<l2uuBOqQeKIQs>rVs#9aF(Uo$X+!&i_vkZBZ69|*O0o8eR-K6
zp%ap>`I%SQ6kFM_t$Y$%l%bE(?NS6gyGr}1znYe#X}EI)?dbv89=rL6r(5zA&3b1y
zu|jlo5Xb}7l=_K)tga)o?_7+e?hEWJM}91><~*)uqJ-hFn$a%$JAmIe)o^>ylK~3;
zKFwK(x<OJ+`y-2=g9*RlT_$|V7#{<_T9LPs$1Jg?jYC-YhqP1?4}FS2!TztpB=`{E
zA$|&e@E21aAsB&Ms&8}wk6s@+8^P1;fE^m~yygfweWe>{_#?;grAy+ckKb;HRQ&VO
zdhJQ}LSg3W7o*y~*%j*oTW{HIIh<!pTbC|PDO#BGwmsJTozj65-{f;ccWD(?6<K)l
zm6GlA`|_g&y`@9tPVX#S4IBn5RtAyg&;G>C5yyb}4h?vUv@*t!n6QWdp)Cq94)IeS
zqyIJOdBmf9Gxoxwg|Y?MBQmI9k%zck??D{gnS_jCvX{8rW!X-?d|`Yb*K!D#^geEJ
z=y$*xPRZOZp(mSCAL^n`$($baMY<ra<hDDx^NBy#p+Dne>#8g-W@r^YrENlW0sDho
zPQWK)HVvAk;?Ch~xjYt8<CB&=HZfLOtdgd>=F?xw2KtR^y^LgKD390dw8Iy?YmFZ1
zTkaJa@VI}r-%B03v5#eA;!k13(WRC@ixZ;DQT%@_KKRs{SLw;ERb=QLj5OgimvO;X
zH3l~wr?qfe(BcpjA4qrK>DYhMp8j{Q&QAXy(is?7eoyqj>c@XEj$5qoJs)bPa6w0g
zCgOl2fSnC&{I)P&R34NWISyvuKly{B^2p7awpq>r04dD{;xnGdSdwx-vL5fRb{_7p
zSC{(tlzDe99&g+jKKXCG^W5U>?uN)5Ox<4T-(E`_h<ao_{FW!L_n%z1IB<Ab5Zr_K
zdY>z4gJ5u^)%zSXDa|nxn4^uCQ?7$AhL|T$3W>5@!ac9q5TU)gKE89w>m%vTkQB0q
z0(ZL6OeNu@_WN%T$B#p3i$Wz}@;Uyw1mLw}+5%Q`<~tO<vrfsw@Rvm>w<y>hXH<?2
zIA4$P&FE7iG@`X*b7PhZ0Cg*I&3(Snn^VQLvn^5Wx_~`~<ccKL1w6zcfqH2*6B|sG
z(Fz>L&idp`hUdqeSaeq{R4Hiy5*X3l5SQs1E5mz#t?Hb+IQ^$44ISNcPbRw3h9!NB
zy^W2;ur1d?u|%hkHjxX>-j%I$#1e4*!~5#b(JOl{q?fpvJ_zRif<*nT7rLu8bO+6S
zM!7NY9MgpNU|*|kwZ@g=rEguY@hE<!tDQ6Txp=5Kc~23JX_;AR78Lm1Xckj19`_8Q
zp-YEs?zSromVZ*EZm<UdFvgDnVZ1R<rb95_k}!Hkw8h8X>(4k3;@6sq(^MXf>1+^7
zW|`D|8D|>O?QpcA``EL$ekiy~0rT*gS(Ypfv^1RL^_edpdf;3{?q0Q9h)Djux`7Er
z4!c2arqMY1muZ<U@@ONWu?UF4Woc}4Drgb*t~qjoGGX39QuffO?B6~`@labDD+v*(
zU{66K7kni>MpEoec9)&h{}1E$K05Z~*svw`J>R2fcRD3LL2&Wv;P&zJ{&ts?KMp9!
z`1@m02oo$Y!s`%tM7r<sDfjGQ5bvO5S^t1X?W#&cCaCQhQZdZn?_#Cd7mn4Kv#i`?
z?`36Wyqu0PY*z4<_!jq3kXL<@*h=_zooDZTTTL`tb0?JFkPT=O*@skLl>cfj9202B
z4mcBj0wn7W3DFfgm++B+5XC{5)hS90dGZM=wt~qBEm<uUev{s4O^0FYM|(lwv3RFS
z&s3Y{ktf=e`q65o>s}z_UCtA1oTHukG8dXf>R&WC#MlNWC2Jrs$OFsfD^fu&+6#{5
zKn)}hAsnoT$;$VWu$HD%JvuiOk6jrTH(f&9PpSc6S`(XY&;F>oJz0s4(iy1ky+LPg
zjq_J_=kgo%r0(znCtnNnUa?df3tFtpJrX|ULW!zVL_?`#uhllYtafaw7a#k6LW(+T
zF)9ocOVgCbYnl2z_5r6*QCg6}=51H6kSYgetMBhWdszy)gIvG>_T73{J_G=SS0Mb4
zv4pwYB%v~&VUKHfA>SqZVdLJ53b4s88%qNUWx)&32j#oH_7Xb%8OAb$g6^=-#w_o;
zxw^!*OUJqkq)O&s)z-rm_nDb$ts|PAEx~)q!LE(fZZQ=2g}Xdq#5C6OIL?y>{3Eut
zL6{$KG02RyJ-ROJbLid$IPBUtIem^V!dQ}kEIX~QT-M6;PNT%8JQZu%`fmyUK>i*l
zQgM_-v9JjEXK`#B@{B{WkqAcW=Eu~M4BY(Fue|_CK>~&K;or2Zp2>eQeAmsqPPsOk
zu|4-D>TC>+53HP~BU;}ZLAhCL>MsA84v2pvdEfm#_}w2^=<fQTBLmUDcQz6hqk9O*
zb{2SLN4V$%Ye0k3Ja*~p#i^Ir&3x9zyeRx(-Jbk3k?uT@w*VX*GnK_JG=*;Yv$VG8
zW%3}DAfn1drH^2liN@eW%oId}7>*%8Cr4Qv{jkip>P<N=Q>N$Qh7>t;z7Xo~Ur!T}
zCs!ft%w3i_Dzlx9D`kQyBfI-Rr@(QK5=H2<aId4(jk;Vy<c+{F%cUpRQ$K67iH)Fh
zHhxjvBne&mR|j^aa#T62`|Sk`XzLA!O*{hSkDWIlZ&%c%s9(C>KQM(dsjJ-t(~MX&
zEVEl~=YYCTZw=l4e{vvT-m7}_tNIF)YjG=suu^o{m9j!>#Z7R?z=#x}41Ql0<8EW*
zXWN=kZ@{0Rh4Am8mNqAo9H&zpuO2($lHR0R<9FEvdiW>aFT!p*(aZHDYXkX!s($K@
zjRU_X6o%PcCZpcI!ERb2hW~>MG0-#pZ=^pH%YSEISpGje=+kwmHhu<qSGX*3o50>!
zQ|9#J6rtlkcqQZsn&eV8oyK6^aTHR-7xm}yu_9;3V!y4$_MP=hF^NsqJt#J-?Jiav
znx>Dyq>v@%FP2i^;koh6vX(`&J;tGaMc9F({Y5cj5XnCkQ~xlb6(E#s6n;}2XJ9GC
z%0fsHN=`8#Y4)zY32x(jO(^aHd2(>qR?n3)aG7$=4J-0}7tgly$A!m<*x1Uf${v!L
z`R8+M7!6uf0}WFc#V=FuhjrK0ccmLuOJ$v7B9L8D%W`T+!7Q1fr#=OlW0i}Nc+@GT
z!MrXFEZQ2^%PA-+sy#*yR^r9{Hey)%=ix$3iRbN^Gm6XYh-X=Rk3<(qEd`Zapv2EP
z5}lf;pCup`B)3nt_;HBa1NXwX>vFOC?3(G>u||hjI~<cUn7<pZkSFZW;lqkW>rCG6
zeRkZUM1N>F{|D^?N33klEq311!JRRwi_`P{s(HvOhK1d@e?`6ZeiK9vLI#<uqWXpB
z%}9e(|I*qXdU%(U{1I!>1)+l!rL2MCaB3eXxB^)2#IA#x-)1uNwa61G)bis!Zu{ch
zWAyTq{yNoD^C<{H=`*2vO`)#jO3;N~xw#|GBtQb>UXX}!MT^3EE=}C!c^Y%xAx;`V
zWJ_)?<~uDuE$3V59=N-u2Fnt8@#v^RX0}sZNq6+2L&hdse;9GHVmvUQi+F26W0j~6
zteTENhANK4+TyFdIoQ%HG+dpI><)Kti_k^GIm&MHO{_fTPbKg*s6m(HFWuj-`3@!D
zzi;9nw9OVs(OV{azx6#&^ua#k>=k@t6Q&ox(Hh9n;0b&><OM2kCo%mD`6$yQ+wVhT
zSpWLpd=2o5RE*K)d=(W2u%IsvzGv$JWr&XCg53ib==#8vsO5aiVU0OC4YiGM=LrOk
zYz^y{Rag_%2C!XuF7z%ybkyvI*j|SlKjRyH2VDCJ?<`n(MHpI&k<Ox!0Qx_x#m7SW
zT>6e(Q2J;lsR9Zmn3vM&uyS>~y4hm|3GVm!&V(1qg<S9?2uABKt6><5*q_jsnj1rT
z>L$v||60p2#=)aLuq{*#yz??tQf!r(0N2NF4=}pfIrj?^r}QJJBJcoJ|Ahje4biQt
z!}l&3BJ_UJC0akiss`OA4c{ht@;=c>(OO9=f;bI*b{)>};!wrIt-r)KG}_k$$EPQ;
zsJBx^1TqvBWg|^JP-{nr&aWL|DFLmUT8{nQfzRs1>i>%uZnTOZ-EswK<evL)!Ftt<
zmCNXHCDW&N9;`SH*d%GFfhgmx*SC|wptevu7I4(*U@C-S0qYbF8;ex+Lra4ONl-$p
z`NJOYo|Kcxy_krql&3dvb}Hfg&oup)iy+n{B;ZmXDmo4OGCn3tekx*A2JVXt*TSQO
zw=0bYgR7&tUYdpR?_~j(?XRKP0o|!gdyO$vAiwnGW;dhdMpjTBRT>sgN^m2R8Waw)
zy0(gkJTmUAFb+3aWT?7uzBTB)f{BL8vaGJYp~a=_=wAxzIqIQkLxEU9-7Sc|d*)`f
z4c-iF-$>)ZY&?##`pxNtltdu$$zeXIli>nD!v%J>pDq1I_G?n5Pbd<dgZU0z!Qt~D
zt6y$p9dx8ymAMlgO=06*41_gm<vq&CbX(tnE$tcARC4KGlE%rF3OcG{I$!Yu7ix#J
zLhG9fK0nd=O-0`zy%F8^lA}e1DRW2ahtgK{(2*aV4UD(zXPycY2S6L3dCh^e+bMlj
z^i&7!OW6rhblcY?C5c$^%^z3m8vp-Plm8PanBl)9yf3Ry#%yvzZ~u=qx-cS1T!J3~
zKhXvBv_`~5L=CM~$JzelCsJ{C%^{03UD^W-J=R(SF&35rm(t^v;dI~q!ELR}lhMU)
z0*M_Pb{8xTPV+l3MYN><4>>D6whye{u4Od!YN58OrQxB&{pFGDCK!;e2f#Rq-k%n@
z89{6!Z|LLtkOy?4ZaX^HxDVQv$={rnQH%iXdl;8Z+fX?yojwZ`cHR-KpXa-}H`~R$
zLWLm%eFsQp=--+=1cxeK4+T^JsUNEM1G~$X`=L=;eVr3_k1uR)QUIYa7BL1lC5mx^
zso-o+fRqG&uoYu}C)d9Hmy`@KA;%j-ta#g&D73uXE}HhCpbFpY$L9;ZBZY1@8c~>v
zCLas}?;ndndo+D~$$CS<O_e^rKbh~wgoAjUh6Zl?eM~zIEn9Q4E%K;+OyJsRM-%hK
zq=PKSUk2{V*m(-zMfKPc{Ez}qUtlpqd<1H;N)mmfq`95Hy`u<}#5W9xl&Ne+jPs2v
zbh9ji$LfJf&|8MxILYavuE1Yl_w>NVz?NX-{{D>ZvRv#lEp1ijfH=tSAH!e&Dv19w
zq`}4pwGyM&c=4gmh9ctyOhe*!Qb{WrbUza7c&>m;c1!mV>L3m{(^ArCcPVo-H>8-@
z*xD@l@_D$T*hUI-;fgh?5OX()l+{ZctMt_PRXXSBhu5P<sA$byj~-+O56KYry7Ed`
zVHHiVCc{TK7AZV+SIctW*I&=Jq(oe`;kgeH%e@FbWN0_l7Gf$9-5hrBdMtSh6+YdZ
zt6?M<t}<g2=~~~J44_`2?o6Q0OOmqcI`|VjQ@g@qI+n1%9cj0eO>wRl{lRMF{+px;
zn`!fsH!Hyruds5U!nKSHL1!4nm5W&+A|yFkQ8^5U;d{tA18zrz3<V*8C_t@KLhd4`
zeNYQIPZ1q9JKc$fimS>REY_p6FR&q*?kLr&tEj<W7hxarPun;@8t~v9Aw+a@D%JXU
zMHzvPnt+-MCCV3xL^$*bHS4Qf)vhQjTcf8z6<$Z4aj%5JtiMz7K)mP)Rfz+DiD262
zBGKEhDH=j--sXs(>rL8ts)hYN75rMk!w%W|TBUeI@~CK=rg+3cwXZSDiJK;6Qcf<u
zO;$C^b>favL&MshAcDIb-XjtTVx4zs;9`Z5&<*9xb36<<+TmL>Rfia3W(fX@CMO>H
zHV%gm=gyZaO41{N`inKGXz=|7t`H}W3S`7g`SR?1sBuppLE!J|tPe60LSYhUOhCiC
zS2~XOwO2aG;nv%|@M%^Go`ubhi9jJ^V)~diO3C))&{G6r&<>?oYZCrC^2DY)FO<7s
z$->`UUv2TDVA>;*=&({1MPeli%T<Z_;*?ZKEX5W!uTUWw%FVCtCGwnY>fR+!v~tDy
zw3&Mj6-dOBnP(zhhGgAhNd+5*E*O>bx^Hila{!%kN&X_}#Yo3_hxiHvPbx#_^H@$-
z)si;_saUuIgvkMdiV#zD1KQq{o&rpNdr4RkHf16J`JjsjjF|I8Z}(7@2xMZJgGu1C
zth?gejV8d8l95{YVp($GFdrajxv{$cRPdNt|92D0%=}-vX*8y7{%<lm#V6o$1mY7I
z@}C43MSUBlQ|tQD;hC)Cx`*%3V&d~r$*hv{MH~N)7GsG|VZ~2!QCU7R4j$|0*-SoI
zOfs^rKcyunct~?#5<dcpg)E6c@T!B4pO)_rx{}6c8;ik;7BZ=YrgE;0-HUtWVCWsE
zRDI7g1Vao<VwK#go7|#GbWi-QdA}uZ@fE*1iV3*m6zj6pm&bQo9F$6=?v7z+q?K+q
z)0YjUV_{&8UgH2*Jnj3tIUEl@s*YA#G8};Y>)Z6b{*84;(+K9Jh#ann+`O8BNrZwb
z>MpTa)9j?8z|Fp?tccoq%vu4XtU<5>5bqV|J^*B4>1p>ScX96itGs$y>yHI8Z0ucz
zc`YnBc@<{>mI40ye4Ngy9H9UsFdjsQ{b!AXR+E9nfh}oF<9bg$TuHkdMQKy#spGI6
zCAf9wtOy7E*W@;vQHdBr{04DyH~l{N6V3CcYgy8ru88r%QS-dUxLEaWZp3<QJt!IV
z8p4H#$JR*oV&y%^Q8_<9_yqv@?@XgzVvS*{rLD^Fhg@LO4a8QLxdm=uKc3h@i?zu{
z)R0M3gg{~hhibk~u)Q>Xd9P|?N$P~k0#GM=(k{{%sP3N2-29mHAm(Fz1M!4K3*^#<
z^pN4yL*u%q?KB=J)J6nuv(xHkp^^|U3sf=+zi%V4(8930xD!Uvn2q=CD2u9bS_T?s
z1x*TlXC`Djoo#IkEs$Fbm$?$)#8g)qB2dEUm|;CL!gc=MeabzTW+4hWhSe_qIjw$F
zqreuh%1__mMgm#jpP#;K-Eam8a}K2TUAyK}GO}sdV$rB;0ehi7YU*yul-j&DzN-s6
zA^W7Gu3POW0$bFqngWgo8FPl5fHTlaY-z`zxn$)~<RU?JI&m8GtN<!bE=fzLRq5Ta
zQVO?{|BJDIjFGkL+Wq0SZQHhO+qP|YueNRXYTMRo+qSLMcK5%pduKoIlkDqF?xa4<
zs-&t?XC=ou=Qxh>8;cIiN`r?|837d!=drp7!fZtAaOyTYISzyKM4q_b*Yr`<R&mo_
zA^9<!SEuHziy`aSNMu-~RliUnkx_9qZ3^pzxjC1bs9SjJ;8T%RM!zzBmYl52+IoA#
zc@(|&s(7Vy?S+!qUSVeYk3$8L*?=+~EvHlL`byU=fAfFK`BdNzF%DvDyt8gr6rtsB
zg&)JTR%jp#aK~yq!(>SA(QZopa8~c~0=2=b@Mo%l^Qqfam3g4Q*lukoQJ+zgwkSAn
ztC9}tnVNM5^Wnh2CTX%?D7hDFj2Bq!kJt62oO(ahV8$B{3rX<_U=MCZ?ToCKIBk;(
z(g`=Tb@4lAVn^2ehC7hPJpcR>Z*xQyMLyfjnp=fq#t&F%8GGy?#s{Dt-A^HP4|0Px
zzMDgT=$T-Qy@windq}+2S}t^|ydBvP$1@ZE5H-oQwK*R@NluD?(fpOK0km=6)~n?V
z>q<A1vC*T+qkUbX8UH(Jxk43KlmupPK0SeIJUi|JiJgubgCeLO)JJ;SS?x_JI1wlt
zyQo*Vp1@S?;ng=0yEt4RTdyx#eQ^MM4}EbU4Dq|tivuKI_l!@Jh}q!tMZ^ri#gqsb
zE+kmhX#UmdMMptKd-jFURuGv^M!W4K#8Ns{Wj>XtxCQ?BOjaFYjwji1gi^Xh3&vwc
zq(tix7v6PK<2ltgP8u?&*hMFq<){LMVob^IEGafg5(alXgr#~>#YE|?@OpzDas)Iy
z_td&!nXN|W1-Af2@I%LY%O1F~VKhoof7n5Co}IeJWo*)EUR^;qnNM{h%WP$ZM4esM
z6d)q;D6<^~;e^WKgljA;6XV(k5)Yk}nnlRN^=T!hv5`I)_s-fjY^-)vH1uPaRMb0v
zu@H6IBnox)U2O3=tp)mT2*No}F|ik$=l1F<D8RPQJDP~Y@#&RMfepCvPM)Zf3qO?d
z?H7j{?S0u@i^1<2p>?Lvtd0&*uI&SA9^WqI$X9$G$IGXGQ+>97QGLe$GLf>SITia4
z+5F^3^(zr6mH7V<{E;zZ{<GaNT76o6dTj~6^~|4~NV=wGbK*X42Oo)~Nkn;v%s)4`
z^(^OpleIE%Y-wNSJkHE<3Ae))53a>0IFWxe4M<^~&;4=g=l_$UY*M!@YG$0z_IaMa
ze@X(RXAUSHQx3Qy7e`nb#R>Y<GJa<jGuxB%ivJ>wop{TyjztP=(zV&(htp@{!hUap
z_nK0=kchZ6D#uvSrd&pP%1kO&=pjFo_lB=j%19|T{I=I>yeS^}1lGy7>|2<oLW5P7
zrGGy~bXYr~and|={Gj_H9`Dt?bGb``V(XxCyP2w;37gu<sz8V_!sdb&By@T{iIepa
zAb8U~MNkDe_xoD%-S}9RnyR;W2Pry~ur<X`x+<`V^EFREweDiz(Mkv}$LH^LpBi0x
zHq=i|CSS1%Y+WT&z|yf#HuN^pc4--lMI(Sxd;g;4?ukCh7x5m02PEQ0kB2${#W*Z;
zi1!Uq089hY&5v)I{uAI;{R$!0NwKK!xP387WpcylC7;P|<5i0n<L9f-hto$1-O~K1
z=f}NA`746RT<c0T@rL!pxV|lH-#(1$`eVMS9zoO^HNdhM>dOg4d~Wo%APt9cPU@Wt
z>(gX}(WI)^<>}-iI`30vbEGYP>|xnE>XL%Y&`5dIFW5{iQfD|phCy$Zh<sH0%r>4l
zxsK)C1Xi9DEfHvLeV>&sf?lg1^(ZYK$QZ$TlB9#w*H~WIY(*UFIU~{KdP@_ePbhwO
zOER!c<pQV&<Q`78@Db2Nu)jvm#uG*}9-OqGhtLKcF)Zo~MyuL-aPKJnAR&X!oDeTZ
zQe)bem5t7r=NP>+#7Si<Mn%YEi06y6tf>BxuSQ0axx@K#T=^0ykjco}csAuy=rpA%
z>uyFUJ_MA}q0-O*%XyL|!Vqj&lGBwvIK>sVi8eb4bIWAdK@|-iClTj{IxG&eYFfU7
z3IMukp2{7DO<jY|o7)NBgtBV`Xz!jPWh*Zn0;D~_orLH>lSwQcfN+JMU4>k%275X^
z>Cid0<ni33;tVjD_=)=0P!Plj#u>0KMy_>5lGp;~lBD%UuVxMgtP%7FQu6%j+wOVM
zoFA!wK06gh4~K1)I(f_@e3=}S?sr}Va=DC2OfNCz=xsug+C|O-)8xrp(mS?bH*WA8
z^F|@SeKtq#tifGWOLROlHAZqamkxm$=ONj^Jgm)JONg>bv`V0z)dsKzcyLK~F?c;3
zh>E~7y#I}3TLp&Dnv(^MHX)@565th3l6@E@OHZ%sa9x%D(v89$05|A>^zJtcKWoeO
zfO%=#SiO@=(W{gl5Hbp}504LeEAhyD7P0X1X|!k5U!IfG4l<duQu^fMh`4c`arI!`
zk~&{)h!_A>rnKD2j4|_KlZrZ5<$hYP;UQJP`#~wJo6c+J!w3@cZiUTEu4BHYejpH3
zmaPPtAXT97##xOo4=tf1JC&Hu2D$>IiIAv!fBjAxv$Q<LSNThtbJzKT?o2QF>%!Jo
z3u^UC<Ff{A{Z+Pij^AfGoq74^p?o2^-+)~lqdotkbqwtPD=+^4KbiMGm}y2%4z~Zw
z3Eay3N#?O5h2DHpLtM&wE8<N%gyM@D64++8M4>?J)6D7)pvP9XSn5*N9T9y#zr&g0
zjn%jE$yNzfLTeKnNx+2-><KH8lDPJz#3v%R%v?sUnmZ1T1)oFc6M&4ERD|e6BMO<4
z*9*;_Cs&b(r-{H3<i}L0{f-qHB_nOZs4mXNm2B8{svpk6>!398BISZJn(9Y@D^8(&
zJ6202Fy)dUtuwak77Ec^Du^anQY++$ooZ7e>@!=9LWI!nMfFP5(oMDuXg0Yb71I?e
z%r+impemZLs;w)Igzc?}s{2y}A|8Akars;9UYnM##3Z(IX~?L>;6pA@Q~p!}BR04w
zp6V?hU}mIDEtT--=i?ErSgDFc%uuxyFJi(8B9;gy15(w1O(XVnUsfg6DglFvEX43Z
z%7>RDxnRqi6SJLct4&Icl4Mn~$(q$>+mDUV92_X?mWFyo#>x8D=Jp4TNcu`ofor4_
zwKN?ebJA4^Bbw3YBGs@tCnYGgs9(|&MWG0@l}diTqg|7Lt$@wmL|Hp32H0+#uvIN_
z0!NE_&9vTw1Vlabl}t6w`JT6gX5|>Yz9jR8cD7whSCI9HOpsIiG%+J<gGK`(l?^)5
zs(31pv!z9oL45J&S#9%FU}PoRLiixW*zaHvQ#7Sh9o?h!S@8WHG3Fqab}HM$OkLeq
zsDWCBouYFps#=&49BJ!N0;d)(4UzC7gA|5R2+}Bpek?H%Q@)q|sJ$~rH9EtHUL3!d
zXMcrU-(R>yeb$ETy*+IloID@Mvcci%ZN`CSg!qhC@Yek!xvo-K;V}9WW?6PGR*v;;
zKXaD7*qrC6+V+OdoH#qPx%GTM-?r|u3V*tyjxALG+A!+>jx*OlFrnV|@(eW??ArDb
z%r@Lpd3W;Q=xWo+#+^zX8Hg&K2~}@&pTH*8|Gaezo%YM@3w?EB)7R4kv+RbWH&>6k
zm}p#EqAJ&zdv3gEFV&~p#@$#6ZjO}T5&V>1Y)WpO_hQPU*Ah-q-@?afjkZc{jrJmz
zKyQLBhi->1hOT}avv9sJ=IHS2ZnB@*ne=xZNQ;`dUo&l_&4$^5+kw@`T{xeZSX;YG
zx4IisBleaKLPZWNAZqZfel+pepM9D8JGH-L<Z;f|CnPp1NURJAxFNJF4-e12xh&Jx
zt1c5<xJ}L64|^7{q<uHQbl>8qOp<qcxVuacN^X2%KEWQvdHK3BadUZ-l$SQqs-6cw
zAa9?_uiyJ*9<>vyU*qrnzV>?gyqCAK)BCqq2d0)Dtu{<Kw^;d*oo|L;ii=ib;W2xr
z9%-U|HcpCj{E{UXRz1zU`<OqM_<r2{1s_<<^Y0e}|K!4|kgpQ1ZnX?wqs`V*Vq>ey
zjxMbz&&oHlx;(y_2$8Xk`T;{wd+X&S2dn?#i_l+)Xh$DJ%4B!*ArjB76rp0H?|8>z
z1U}-C>ROb)(P>?7j2?q4T^rIvY|9Pp^<*|xVDb-NPukaZy{k<;`5u8KMo<A%$Si%~
zJ>$u2pmu1PzcHcT!)5IQ%_@3EQDo_!I88-3XgqS}Nzylh7RV^TqRpun8v(gm5l-oq
z;6q;c9#;_$Yn9^?ZY6k|2js5ml({b#;kf_L(q1D%jo1_C+z=3dgR{;c1X%pH$G8{y
zz^;vwkxU{nc{J<(tQdZZk%QPBjd7@AdXi~xJ2)olhv5XoCLsqbdQn;C%}a+wXqcl*
zmEbOVcF_5VV_bUN3J-zAb=OW!Y`%r0wfcz0l|!qmx-syRrEGdjMV-Ra`VvRIcdT`0
zsfq(vYVR*a5bZs%-yr%-S6shWue5hSn74$<v!IBbcXyGf<>zGPHkRfGGJ*U<cqEe+
zH%Jc$GOIu*86}ky<5)zK!n@?K5clI!+%rfuuIoVV!|Ftm6G$(|okc;5bR~30bt!aD
zA^HlSAHr!8y0)HU)hWJ%Nb$IlDH>>u^>H<UPTq9p{TfApxCdFK4&9bl`J@P}7n|Ig
z2cv!6dUbIjF9nQM#HSsGp6J2aMe<CptYiwv*X-4m&Hm0~kwVQ1wcc1ouEuy=SphMi
z%W=;j;gIXNZ!Vvw=i^DcN0ubEx~GP*hWzPZkz_7w9q9^V<pj5%(MD!r1w<~f_&q6|
zTZCD067QZtL9#^%l>ko&wZy782!~W3ByvfB$Sv`+oEH=-gWe2z3Aa$R(+w_ZF>3Wk
ztYG!h)h{juu-Lv|!R$;iknwPDR>^J@+WxwLVTu`l6O(XA2WYxv*Mj&P!@de+e9{X@
zTR{k8-<1zB9q8CQ_j`-ou1=Je<5%icV+g>(vPnTe+aK@e<MiL=>!V3a=Hs3BBqOcu
zk`KSQ(_uppgP{Z{hAA^SGH-SNw233nx#BZzm5xcESVrb>KzLK<<wLgZE0J3$4X!rI
z%z=P_(RN15oY65A!wZ!x#VM$b7Ud*2CH+*0&U+o|_1U%{a*~zEiyu6Simq5|W8t};
z^dgiGo_3s7x?)_rf+hQIUWWn<o@=s-isr)NTjZrU(hln!A%!J4410a2R%@uy<lrwb
zBGr$sMpzHZCZ(|Dxegoe{2Gzq7M?Dq{W`-zB6veu{5mDNxUS=x1~EBw+`jP0&hoYf
zKG4%7p@#ouufWL3@c&lBm{|YIWKjBluwbD#eo);4%vDCH==ndUY&7<$EztjP!M=XD
zV5t=JksEHM@C`JP6EhV?zC{mzeP;(QtQz%w*D?Q3ciCqv;9*Fr{;ZOPrvtz^7RD2Q
zXm4(_81;%6-TXR6d-lh()5+t{xrjC3=*N=1_&FE-Sh7Lw)%EQ$30iF_+@qhQsnYL1
z=OW%w=5&wh?`?n0Ga|K0gM%H9i@dMZ+AnR^GYgar1`W6Wl>aq`&_SxI{we<p49xh4
z*;n$v>s@GU@nVh#VMu9D@Gzoi4PAm;$1qK+3KvJb8kKRZ<!qYv2Ny<!DdJ{L)MJn>
z%Rg>%t=F+OFKrJ!ZOl)#bQPW_*>H>gZutE8;Bz6>*Uws;{oO$Lfc$IbHq+yGvqQeZ
zOex*E1WUuIVSAEHh;FV?7xEEz8kZvb#3n^gWk;Onkx7+WSfMgR$&eKexRDN*=Wkmd
zs)`CIPqHjA+PN&3;3t#|Td|7x?hy;l_AbLDS}y9(H?0BFnty!SCoOx!Mrbvmk?}d3
zLl&y!Tvx5u`qrek%4%1IQ~z2EeqLv!-Zr^LkVj-7*)1AU9BGf3zdn<s)iJyrq#>!G
zjeYCCpM+hiMpl)rIyJM@mc3#D$8@}0i$~s~Ig6xk3X&bHH-;w+bF|H3M0Yfv4SPs0
zr8gQ>jJ_qYQ(nbdQAjP}L<#Lz1YS0c#Hs7kZX2**wC#`Vf`+@)VtY7S^`yES0Zhw|
zFz9Y)mN=e9xgg$f)>a#c8Yv7*&GjXYp~s|n#2SWL?mc7+d>BpSWjZ5C1mST1zH`cJ
zo~HkW{;Repc}TmG*}Sj40_nfDmKkP!%#|8-@Qo(iE8(3wboEhaF=LQ2+U@)I@WUOv
zBWYFmcoC9c+?@>6kZs{Pkh?wK=LN&zril;XdTUg@KZi!-q|4eNFPpLvYSlWLVWxl~
z)YI_+Gd+zB-5~cp!Xf9`1L~NR0sT+c17DyguKit)kUAY>g_-jC#%;PE<J*mtHIWh$
zGJ6s0FLi*V;H>H7qmK-uDmx<A4H-JYDo7x>uJF~N!W;=>y=kds_rj^RHor(l7i%^w
z&h;@bPXz3k$MFnnokwZ`VC+MexK9i705Cc;^Yk-p+KF`HjcfO8gz?ZLpW$im^47p<
z43$ehuv4yTca|g4udxQE-_6J_q~9>{4KAuRoO{~jY{fM@lU4CPI^RPFs^js-y>?E+
z_NFyF_Y{>vqyiL&5>+1MGMI=K$g+yo%N$b8GCgoC<&jE~#bLFR30P&;GI;0EN65<T
z#jUyrSo`<Usw&=lTm!Oait4Sb)x_dzsu`(<NcYut8$zm)VJ(4)t!8#B6y7Fwi<y@s
zo3dTXD+eavj{0bY<VP6xqef#WNK>n|aZU9L_D1z2+@dCFbslpTCz?FCSAsNRwkYAi
z_yb1C%8YB)I<Bx$opWe)Z7a`N^%&WT7$0E<j-`*EnwyXvi9-Y@yvDB;CeCQXx+7u8
zMJ)2vy%V2N>F4xJOZrAu_~gwXSmMn3uH4;nL^gg_+1cFUd4BR1bKS8lR4s3hUD_@p
z-kTW)2H8Z;(t&^hM%l$f526F9TibO0bEkZ1bOjpmFmX`?EgpZA35-#N*mr}CLkyvN
zTHS0A-9F(@J5etGMKYOwpot6w_C{7vJUsumxry<oANRjHLM^5L$3Pgl@r3FXVErwk
zgropsu^{;~Tb1BHvsImcW~;vAr@l4OvfE`kGH(PGuBHz5N<{K+c)jrbSf<Y&>%LrC
zIN5g~(=es>z@xCVz4=E(3%dX)+8A;_nt9#KMl(`~Q`X$<#PNDlT>O~>O3xc`7~cu-
z&K-s@B7tvxr!nCfouYS4Kr8W%Q2T!@(?cAaY}4=Ov$z~Oqh=_a@8J^ucc1moPV4y?
z)&H?fSBJ16idIAVFi-))^7O;(JNf^yOlxFg?+9Y&Cq<aWG1Wx;3|hvy)NxVK(a>W`
zXt=m`DXQt^xDTz-?ubywU9L_}o(nqfjy!epkfjH^MLi-%CQWUF!%!ym3PAux;UoAG
z(M0jvtJqt@+l%MP!G30t?(^!q^d9y!N~#hYXEPaSD~WO=Adik=;_YiHs;Z81IyGp`
z)0{zE(L8x)moXi7N8~0AXg91CMpf$PTI-}~{bP<=Z}pqkdW>->!J(%nsZuasH1+M<
zm_MrPU*t#kGSSX}DUl#`+BuE&APq?JSOr_y?|zZ)G&2(J4DFAdvRavvnu4}A?gKXz
z-x$J|FAHqRG4(zonXOUe79Zuy$_1j|RFMifl7K3pZ5$UMt)jqLopHiD%vv|6q@%2R
zj)niE;Y*G^iZY`(<rwWA^|)jzx+^0TstZ);aj0LFm+`8Hq~Pj)1u{ccD}u)41cni>
z%&y4Z&9}sTYpd)?03<%Re5rvD7?3)#L!D%WBar0aK0z&qBXr$u)3J0mv{KyEYDHPA
zN}rG_T%g+LvAx#@0_iI7`M@Z+Xzk3U)ZP+<n2K7HpGx@=2uH7}v7XCx0aLkbVq$w5
zq4Px4M1K4UlZxD*pHFgueoZE4!;fIDnGz-TRDy_nCtrq3QSsZC%qGfDuE4oVaOJqG
ze#+dgW)5c5oF6Z%nr@v-8g0%EQ)I?7KexyekDztl(ZJfqOb9PKbo3V%A@TraST0T1
zO5Q_Np8R`QPkLyrfoD0-s;lLWNK}1^W4?^|S_!p`$f@oCETY9G<+c@rV7W$Z-U@ZX
z8h(kQ6ELcG{|%F{j3s()Owh3K_Kvs)*@2XRrod>YR`~Na_EZE*v~U@9Iw6zu_Q+A5
z6)d8dEFz}75Kb&|=Do=Z%GzH&@SgUZXV`$+CuNnxYw2^tEY)yF#dGp2xjZiE4=&2S
zYj-M9T~@;kv`zxpL%CRQX7Dty-<uEMpFI?8cqzn<gpsBL)Ny8+$5+q>vIh?!9*9=O
zl|qLr3Y}JalVXN|@$NYx*!hbVl>&h-*vVL)t)e|FIQBxM^$lzfXv+#?6y<wpnQWCe
z*bYp7tgL<h5OQu>{`}<Jg6d94;mM{QZGqX4Wkuu*MlH5_X!?h08HJzT6q8ILa;{It
zzXwP(&rimN)7GNaxxTkk;0$tvI!SQ>1QImq#p`{s@B_|UPf_TZzeD3=U;etRMQO*6
zTW`Z9MS&LkHqN@3&4kDCrni@1`aQWj_+tig(LlI;0WX23&;FaLG5@ck(f>u{G5zp{
z{=0~{Y}^Lh4_fFy5%I~t$Ql@{Q_Ii~%LewM@}h~#a8uv>E)sB-k$PjXM%=mv&B+vr
z_b+A~alaoQT;~@iuFqVbH-1eXs@Cl~dVMkh_RfCbv#4`%djcFml&sh<=)-Fj9$ZAw
z7qTMi=iUIPxqbg4aDcA!xz!H%1h*%6++zzarf;7QYeV%rL~8SqdEDf;DqVp1fJqix
z|NeX%k%QK#{w>{Fj*;i*zx5+)vkA$rj!w(evkag2wrP48b>pubgcT5GZ(p8o4e!o4
z&Ffg>GAaOMQrzZO(G)J@>kORYkzKVS#gOcQ^CHd3Ga(E?rg!Z*^fMVcC2;UgAtHK&
zf|zV8U%ua5u&jA{>mLca;)voRf3s^ai>twvk>$<e6&QggYWDeUF}h6$LlL#;XEe?e
zwViXi8q4#jnWb!H@TAS|(DtGXHGCuV2o<0QOu&^!U&Dx*HVP#qbX%x?WjGE#C9m_k
zmd%C+L`e4lX;R_jE&6$tCpbj1dL8|V(Wo6c;Hxgl9cMl?jE=<k@kWCHF?CK-GR3sB
z)m(qYLJh?1dpkn?VZ@v?$R5zpk@x#1Y`jy^%>fRK!C*7M3(=xxiCU<xF^qpFZP0~=
ziRbHCdXTGk!jB{m=B`ecfx;m%(RI-2D60>i&9Yzamw0JGM9FBusiq;NE+FaSjSDJU
zacXx}!^|wDpnv5_&d%N<d4BEB-cUa%jJ4z7&MZ{oL9e1@P|63escVZz0V=KJ(DLDF
zJNI`Xr8;lGLJqp(eGaNH;zaUoORLauh}^StS&J(J^N=Fs($Z0cg;2oZH|VOcOU`~~
zo;UPSA?$%a!>If6I>P|4O@=!lc(t+K*8E%B%F~7qCO;owLgh&WEwN}K2x2UXC3S&?
z)Q*}`S@dbur%~=G)Ty9HlS$`L)E%75V>)u}jNufaft+d!!9|(KrwU?p5%+iYfT+)u
z{|UosmL|V}Id}4Bl5CzjqEugEA>?U!4FAC}L@eAD?9e<$p$EOAXgv!8%AeOs@1JAK
z7g0Tlo<3e{#8KjqYkjtMSCU4e@X*D*oRK6Qg}k<Iizp3v6ngkk_^m%#sz_OtKfV#C
z!RMn9C%`&IDu@_oBj-(3)RE8M3t#c-h+-d0$SK|Y!>6lmOLy>krJU2W-l=KG-;=^1
z^*oP(9)hEEuM-Fqay7uk_U~j%wQnOFrCni$b<RSYo@ahaQ3(a-)>5=y)jm8~t@jy*
z@yvq?7B%9~k2;Kb(r<UUo1#k5sjO1=iTe`>M=nl4R)i@GV}K3-uH?;fBLN22#mnLO
zFq`8S+Hf-9BIfWiBbcilOr#6Ak=%BE9GNX17CT`U@EWy!Yn-lTXZOd0cVez!BlUso
z(yAAw`OXKL-7$E4dqpJ+eVnbEf%^lBlrpuOXX@D>S+WbN9B+FVl1oy{mqxX3M+DK+
zY&z9T#=ZU)EiyYLnrE%fER&<X&DGY-7w(Xzlg^#N^A*=fP|JR4vsNB4B;q+DKeRgy
z=?bzpM{-xav4L@N*Db;rE`R12@Vjvw;wvaDcz|)Dr;8q(?+GBsEqTbh1;O{R{9NcU
z?`Wlt1EPM1gk^Ch8vaV9Z{<G?i}Xg?8k-_ep4NN=Y|0-Ki+&0QYA@dakDB10|0b2J
zEdTEYkd>YNzm&=;r+>V}cghDKLr5fmv5aVQ9Q*8y?T#ev(o#8hdGFJn?}FW05-BI6
zkuf%KQAB?LgpY9e{KtRJ%nR2)TfmL6{oQ182Q!Z^wg4B4S7<G)3^4Hg)&9R9#xM7J
z)&qAuqd{8Q)Nx7Q-CdUc=6w?o1fE6c2L2}q`#6}!vbolrGxbd4I7hH|e1mQo_$(I`
zf<1yJz&GDsU%sVL^;)2@-^1$Pk1t*C?kh6vYKURK#wceQK0b}nxe<Ocb3u$$0C4<2
zM3$?s(_%H<E$;v1n3<8Ma7;B2Uw^GpUFf)~XywyKJ)7Rv(5xSZ7&}V2?XKDjd5-XH
zisPH{oBLCsOkK^)+z7ASJ(jokk7f$nWBEkPOt#u~2__&cqXij}L<o}D^NejAZAIG;
zev~kgVaYXcMEfqCv+~>juC*(091pNPjq)NOni@mLJ5aB$&_diAm$dFPZXj-|<+wI9
zN+G<FeaPb3PMrGEY4Y=}0MJ;+$~M~FUe<oO1f|CmAKw#lo`(``kRxZ5+Gy&1(*igB
zct3D=W_4odzPdul-zJ>*xv+x1ifPw(cE(0@qDLzg2cM{#&jJ{{stmBcjB3~?mR>TM
zu%ScBf}^1o85$;D1y^35ARF=X*xX5wBdB7cf*5kR0JTb8ZS3N-Ncz>b&dj+>AxFHz
zUEOZg{>RLO@o$rWHu98S0%MK^3^aY}qpE`X%#f-W(qFFFNRRLlhx)YIUu0v5jiFd<
zwt5m;xpZ3c+Ts^9r+>6bF>Y8<Msk{IF-S`d3^zvdvH-@P%ubU;c(TR2khteu!^)$x
zI9>h0pO@-wV{zD5IqXD-haD%nQDcsvvvjz}<(RDobG0LxWuf2eZC)%I^%G)L81vLH
zKRW=j$oG2SAl)0}qsroQnxbJapOcN0hq_7I0G<iKo~DtQwRcr5*eRScR?<AC353gZ
z>b3^sMmkWrS)ScU>PkN~1G=BREHwj2<G8{9()dH~YM$dzuO8X~`q(*xzF;Y#5xA;J
z3T*&=S&85v7YC~3mdMo(-ZsNL8p)6=QAn!k3QGEyF<46OA6qKSo=BlG3AWWqMxP91
zLm|OqMwFWrD8=H4fs{5JhIr^VO9+C*Zj^r{+M@MIa{a+(v3E!o0lq}Vm$PNIK5k2e
zUT{ICugX(kKm#$xSLTO2Om;ZOcE3s7hu|r~y<*9uj3a=UCS`bT>5v2~8|g1H^&fEz
znlAnJZtE3X3J#|%e}z$1HoxHRxGOfI0x=z0p3bL^qC`=-5U`IG)z+krh$&e?vL?Nt
zN?{vzHl`w<DQHl0G{eMH=DV2ugotJtKQ2|{5k-rkNzcBawpPK^bi$PnV~W*|!f$nz
z8OsEf^j=_twYLK@;LlRlym17|6kb><{S}sS?$}gQ)3Yx7?&G3^TdZ$p0iuFL#_c72
z7OZddiW{*ErY6>jjZ|HwpYcc8fgB;v39Bs=C|4?r&i<LIAw@?bB(3bpY76_9fV6|V
zUCBG}>N1ob27kEC5HHcnmH{|Dv5uO7*OB?FnvYTH@pd$coj}X5w57<ga=fkiMJ@Tp
z&(9oNW}vpWLyB#AVUyt9rj~NCwv}%!KT>tc63Z>#pt-==Iz~*fyH`&a@kp2Y9g@?S
z*g7gVPj=<$N3=G!tx~D<nq{hr5qH5g;PL2%fSj>@J?zv#cYP?_i8?O&h4ni84mCsk
z{`NX__295<ypnJ5{HM{Vy_nc@xj@go24{7hN!L65r95Wl&Zign(I6$5SbKz;>kieK
ztBt1_^a2z#6&1e|#M*Ml$Ud{`K&Am-cm+k51(xWeCL_AODzN(~<yS(c7Ir$=AR5|7
z7E_uB@8p5+rPZ}lL`(G_kny^kCfbG+r#|O{XwdNFqqq)y0I;#RNmOlD#mIiaU-|V5
zZgk>~3EGBJJ>LsC02oTAd?PsmKNm+{nJcJ9M%l=MV6ZZ9YWP@+@VFj5`!2DmmWlT}
zb|HPBn~fdQ(}<dVtSW#yTc)st%(DN_{7k~k!b}BE!gvzC{E3V%vF|70@{GIZm~ORd
z8ovHcTnY)hNn3F8TPw)QHvo|sqwc@mcG&-=RLjEf-%W71`d<;2LPSbM{y#(li4owZ
z)j{p3yr^5eR|W51VM@I%sN9ZjJd{wEeU3yEB##Ln64RsC+T@)-?K8g5o77vHdk{2k
zH~QSW;EDW#cfSj!fglR=y}!3VZ$+h8d~jdqm8gdP0f62#`77;ufFtOCpcs6cum-z(
z#PvVaFrDEJb^aZ6-+d>nx_{Th!X)`?Z?I<Y?BDa<lC-vZvOQs)id6|ZL6=IER-CuM
ztC0_P);VnRYXa>S_uWL(C$&1dv`I5HVwvX|u0~6cBpUCpS3+X{2GbtsU&V<hz<JSw
z<D3X7NR}4H>sj#GKC@P&C~~%6{m=PwCp&Ch%pF1=$X$qbwNthC@7H;{qQEHx2i(AN
z7R4RCJvaTpuvI&6aEczYTXDUewVgZK<!f74yu((5dRTlTAl@v2qpo1hb@TT4fROiq
z=9DbfHTe8Ju2bV1)*(QzlRd1kdi|`BHywHIR-U(F3~HU1xIGtqo?<cGrVEXAr5-9_
z=62KGtkq}`1fB8O$Y3(Crx>WDb`O?(W=#=V%%mdRKA)&hxZ-Mrm45=T_2m+zJoKu1
zy6MRCTIYs`muVR~e<5ihsp(t9Y7t7OvihT4(Q9ar@r}&D?+zU}m~pupLWEhly^MMe
z<ur-EFoC=Hj+B;i4d6wTEJO2_<X$EdR-M!7KBAI8V(Mi*8$(XEUBs&wb_XSo<zABb
zPOmd`vm<Y_<;4M#FJ*h8NCe>$kDIWLJ0egsP5<5U0AmY%xlFP))awsc&`>WgAP`#X
z{|PHC${B&V_PlXRDO1K<No}Y;#*r}hu&<o&sGdQam?%&zK==!M&ZB{FJ?Fm_SDRM9
zb)aKNO!MzcKfRL^3*#V-<nZ_g95#dXnd3_MeFq~DZ^qpX>X=~z#cB$ccF9-nZKY57
zPOB;BDfl4~J@TXqd9aQnbQ364ce~I?NZ|K~T9#hOh)V^yRTl+Nx=XJO57IC72=Dbc
zxs86X@|hY8x7kg^rbZ<;S(>TfVm79al+t8G9x{f8h%*1S%WHe$U_go0T$5R&{J@)D
zztc2I^fi*C&?3D3t0t=LJx$~$hwU9!;Da_DBayE1G7d>ZRBEU%NQ*Lv2xT3~!8w7p
zx}2!h#K~HFB-f^#!yWl$5jTmsNzZE7N-4ou&(zzD0-^*>yafw~#Kcl`8|0M7rRh1*
zg;F`dL@$|YO-+B6RxqB1A!-q?XTZfI<mf*q5zVnQOfAvVIRyq@|HYE1IL)@U+kSpm
zW7gVHM_gZhsNB!F;etJ{?Ll<`da{V?;bd>1$-|5fBC2>{a&*OZ9HXYvx@rc<8Oz<p
zAzUTtd<x5Yj_g1(0q_EQ(2&rk^J~OB-{#37w87~LW*XuTbO;-JhFwWW7CSB}I+Llk
zWW#q|1jg6HK%*@5eOXnKrhO@t>_W(b)kgo2(Tl{@t%qs$`Os%3p$qlYI?1*NiFlCq
z@P^7h@P4)8%P)^CDXEB9);a&}%KYF&tG@pAV_B_k<Ad{8t+r2s(9OESO(z6lpgxF6
zFp>BQW~Ac1K~-Dl6g=nI8ocRc3mps^9LO8(n4}2W=CM4+=3nH7M};M9$pl@6@48Qv
zNjMn$l3(A<W2{ebNjd>xovqCaIP_Xp#h;nScpOujhHW@ax2<a|=-<KZZutjuNX2$B
znw+!0T6j>}X0T49(R8@Aof-(L0>h%614xofVK>sn_CpVJh+l^}RxbT!Ilwm&L-_G|
z+_J5k{2a2ai$2K@a5}RsCM@l!{`U0ye1E^TjVR$^7Yp2fA5Ii5xioh1%Oh}FygXbq
z9sAR=Y5Di4x8k%=i$%Crqa_{n&I=^j{)>z=vi~>SEDqNHDJ^p_{dd{)D=kHP>~=)I
z=URf<v<u<3Lx)gdG;rW{Y3FVta1ckRCe{yvjgIWvSd`H_!q4Z=(0M7RE_8b*#B&V4
zI+mi|J(#e=$&sbh^eE_>K^<J=9#q(tn^PiF<WWp8Atw~ric1MBHJl}|X}V%CwX`N2
z0}eMF*-9019Nhv09kb@#-Q?epFYzWoGQnoeBpNYnNSO)h9*&AlnaQ}xM4|NOMz&we
z5Veag)SSN!uuyZ6vcS5CE6l2X=EhEHO~vnWk+=@lTcSwJvrv(tXTm-}3J;jcat=@j
z?i7VXlmM64J%X-QYspB_DVC8j0ySi|w2@H5f=J%R8OfDo-T@ZvbC|GV2fwH<#Edyg
zRl^I)tiKI1YR}SeOtyO#$AS(B_fz!C0#{NDu~2aQ`MoBfueKmDSFd)DS#W?gNzEj=
zv6+ZW6C`;_Qp^@<@VIzsL?2$mvM4FZ22Vv7Hd!aQbhIia7S5a$10->5(;=yoI?FQ}
zzv?p#(ui`;X_zcy2218_w~Ub`T4+fSt|@9jtf|imYm=$N*e8^MlWoN;C1)JP%lxDs
z#S3F;#W*BH`P(p&Ysuz*cfV<wY^^_ON|?a{30ofE_&D#6Zj5X{|6(SIMU8KtigZ+p
zZJv>LCX-mX$@a#BwH?+E6$uKP!;+Qk9`7Gu=WH55N=aXh6ANtFlABT8GY*?Z7X6bi
z#9Vd@FqTnNG433~2HGIkNWc*P%hSg78sl`qu1oLv*K4hRp}wymH~KBAxQUyur>B$0
z3uBR3+!ZHK1o1VfBF3+xOX@&Jb|%$Fz%YN1hIn(Fz0L#;F#C4{5GGXox|D0N{YejL
zA9rHwy9R9H6<_||Z2H)3+36BG%uQE%@`)oCM&~B0ugA~3u)-q~PFKgJTOU5{{(k3d
zIt;j};d`ujqJa)Y8fRnLSq`2{J-TCa@r5JD=exE0>&J&rr>p&o@1Csw)htS%blTF!
zqZHEKC4|}%_4|vVwZ#r8OI?218G5pvsU3W5B4c2yV&qf${4=T9l>LP@o1ad&adZtX
zyVq?V`&U0&1^ObrUA=g{(e8Y*Y#SdtwvNJ!h2myCoyo6*j+^7C`?KG#XmTCBuKaF0
z_#>(ZOAj`l(Tj(U$IMT9BEux6_LPL=KH|J*xm?kHi=Pbsyf?Y+_E1WA)26?%6i}Q*
ztlsbU_T)A*;Rnp>k?5YxBD!`(q+sj9NQvT<W6hq%Zx45eX*x=?=4MjS?=e4KM=J5J
zZnE@KC*0Nj&pO!AC-iINKi58NpD!La+~ucBdtLo|zs`!k9~-#4v-8}LGmj+&XgOE%
zIOV0vb!N_V+LE1P`Jt^i<jK#fI<>MNGyS~~6W()onL#J-b_Xo|^vX1&RvKVg(=MgA
z{L2(c5f5a`GAhNld}a&eAT`&S(C$nrmup6coTwKmstYXY#}$QkgWtZxgkrpom?y}V
zHH7dv`A;>1V7(ox_!c3G#J6_5%Z$hQ=GYbyJ8?{Q^}^Pz49c^xAUceyzJrL`fmabg
zRP(4}1dk693lJsb`oM8RqW7gSf=FPuDRv=>Oz2s_B2-Q_#R-KU-Qre(GgkND+0%!}
z6CopSg_#W16J`jIQP@rTYjk7KR;#UP|L5Q8R=^t@RF@%Jwrov+T=Ng{CP9uf*-wz!
zL}oIS-)m7tfFw8p6mvv=WE5iz$1y<>9yR>mi{$+)&DpL0yw>#{O*BU<me7B~098U;
zn4c@5AWu?8j<kpjc^(1sC=w_|zeqnru?*RVA-ec^vMnxlJ4%5^P*}EqPc~1PzZ@Ed
zd!`Xj@8L}xks?6x<WKyn1t}Vd@6S~a-4-MbLm1XR2yo}#Hi+PqE)7Fa9<PoWGCAao
z8T5fm#~9r8tt0e;mWJ?#Lzm#*2J(y&=nd7!8?u`#<RDMjR*Imr5*Vx(u770=mCQ^}
zx&}XxE4a#GjvtT1Ky<lKN**${|2bsu*W<AuD_7ew3b?_FjwzLuC*T?EJC8e^6h!}4
zSDR&hU0=Y1<U{-3Ye%-Yr*M&aAAE}Mmj7;!>04ATR-5oV<AGE$0GBZ0*x2z=#|ghC
zxXmXl8C=!)Evp-uRu|J`drR<~`&R<%9^?qj>f|w2yVn4QNHWGbR$yHk+a;?2hOm)H
zIp!8Q&C{NqIY!`j^l6N`_BF2XoHo6&0^U6nm}SqHf=Vk4>GS`3__wpu6;Z%Rt<}NT
zOKVoWe<z42nAi}$z0nPg2*ER>GM!7ymQ|I%^dky_i(0{BUwY?X2d7gr<^NG96w(>T
zujIBuaeh}BxW?lw`gX6|k()-luSBzpxH}9X`6B8Y9&VwAt8pLxEZ*lHd}8QTc`Y)~
zwi|c<5b1$_dBOJi;mJNiU>|-1Q1i(A`Ioyf(@*sE|1oo!ewqdT8{6aRpUFT!w#T#j
zDY#4{nw`LGz*arbA=<z{Xl3S7PLr?icp_0HTB&a9Ly%FyYCF2-`}*AK&f1)7zqM|^
z%UtjMX-fM)hR?Ufz!|f5kXm5XXE*^3AX_7?*Ux!PV$?I0JXNd-JaRqWo>`wiKmt^Q
z!vxKM&m`k;Zxh(Py0pu>FcVa~L)bgNh=T@xi^@WfS6CwKkFU76A0h#ErE1-+wNG>J
zHlLgSO~skS`0A`C7${l?fA7^kNRZ-kNNktK#MX0f2Yu&`R=H|gD(yhnPsc*7F7sV(
z5PB1c{12Oe#N!_}LE|trQXi+sN8_?s;7~WuK96)d0+}4&Gaec#R%N&l^_+x8OmO7U
z6w#W)@*hp<Z1({!T1)1;Q@`0VjyEN~#Hb`h$9fC#MX14iG4Aw;s{^alPiE~$!VA{L
z7<|balS2vn7-H_B!-|#Kck)Wc)YMq{4^U@sxneaS5xvkc4f`Y8pkbSl`AP4o@sIdq
zsU+cj%t~K2b22j#s2tl#xug>nPV%^%TQ$p)A)a27a?TmXULBSrpL^@vAfC{J4+Zjf
zjJhKw$<SPvb_|Q0#3mJ;(PCOQG93Vq#nZk&8C&SldSJ{=h=W*J`fuT!yG#|Ns_{uD
zV$%EWqhSwfAV@bUE6gB5Du8r^3iomr&Oq3Al#C=v)3y4{l&I`~r1@fx9b8V+e0V&c
zlmA*OTwV|gZ1WS6XbbV_FyPdfd=(rr)I{zwIF6^xwY9GAjQFfzCYkR$`f{f_?a2^;
znFeTigQrWe6QHC^A;F(SEsf?gS2tm(lZ?BO-BijSlX{;=eN#%gotF#DDr32uimtm;
zBv8$ln75ygI!co!(AT2W^IcRs;y{&(E!`|&Ml~Z&D4HQi^RDe-I<tS^`NmXQKehvv
zzb{(CR93cTQ^UqQW7INu@}y9kv3Y?lb3({zWH0SVNBo`J<AiY}!{DV@@zzxTjrLwd
z2nF1daV(}m2ac>zaJ-xjBRSKJ^T8}G;YLv^>q?jvFEfvc|90!vfx0V!ka7Y5o|&@b
zmxLDux)CuD$*V@pI7De=F)Y*5?GFThHg7`$O+xc1k5B~cxLH$@feU<Ub&Is!Ae$!a
z#h8myv=WHDPwAV8d<ztSfJ$1Ol<-V##J3db0_h+pB6kjF&R0}vN$GW<`Ut!aAx8(Q
z*GN~2w9>v4*7-^G=D=<1AXdL?Vf;e5z8mPYyT0^1i+Y3lyIZXmfM(|I5kT7zcL&v3
z8Q?x;iniFdYkE|ADsIyvgv*-5PS5;#VKMu8_}9IsJsN+!yAw&<3AzT=2ek%Og8Ew-
zuWxoSl7#(!8e+M<R##=lnWN=mMf<M|BZ8yu-pq{cA54(TsZS98cP3#9JF5#Y7gk0H
zMmr*|sB|Pjl|GK}M{7ZTxE$x<Hw>ulB`zh!gS*1)k1vVh;taLEyi=+)JDg5UNpYSJ
zw1F=w2r6=gv(RFazgz6AlN2BaS~JRD8c07obG9F|O1_%TznSjKriE{k7t<tfnD?hB
zTQkX*?dKv)&wQZo^9@@nuJ2{kmtq7oQ$fJ7SKb)nWt{1ivomk0-Fz+rL!>}!E6Q0G
z6!ZSIMq5$YCmV9*pQ0Jd#qy<GTUdf8;U4w%T2xoh<ZSWCd!JEIb6NbQXO?lLUGwS>
z@NAZd|8M%q$?-4x$@<?Gr6%mC*<rTcsgJ>bhoh8(I~WqUF#~p3@9NH}{$yWuAHKeS
zCyQZe-%Y+GL<9lpMw?obMDi3#<chfayLWhRT*76YY%823n>(!H_P7(!H~9pnN>1be
zD=Z7Rzs!Ap$yztvT^<rv$0Ur5xziK>Up##m=qev)V4L_d2F?=}_a$Qk)H73HG`*wV
zKY03Nf?h!r;ngq47uUkZh8f>k=}awgDHC*AO*xYUq4Zwo#HUSfjpY6+8r-W-Q8%&o
z2Lt*)frpqRYmk8@2_N&~QGKQ|N!b+Ke&{~oclK%VG8|<F#3+s))_&c1Y~MD+z(0`t
zg2vqxMC<R5`UrZ6LmzamA2duyp1<o3dvCK6^YT?X5Gi2k4<;oeOpWn<nuwD6faW7G
z9|Ufq@ck-jiIGJ^NZ_V)5<7<v*s3L=TODQcly04mP-F}RY(x~HJfMSKU_|wWcIKLC
zvu1L*+cETQ&yc%k??a4a{-PXZc<8SXOk`iv+nt>Vkzr!d75V_=8U?|QRAl;ZM$HkE
zR}1)F6KdajUP2RV>M{Gg`7Ube`5s@va*4{?K@YKEnm5jC$;5UEz99zvjKH#;5^013
zqT#5nK0`izFz@r$z2~{*Hmq^0w5wxVDUpHmIjwQUP#_$s^aP&mx--AweC{k?|Exz+
z(00N7Y_=yosD4ixiB-ZWRA?lk_k)f=Etx`cip6q!F0@r{(dp*Q!{FP^30IvCZn+7L
zio7`a^HU$su(~$t_xM3eh@L6Fd{S8pf5J}+TJKQ!<)E;NjOXk`CDC+QNf|lNeIvr7
zKeTY+sYEOSGbEfX<LR?fG8~3oOlv8=!;!ypkBH0lr^4R~P2kI~f2u9&UNMzHYQu`1
z>)B_FB-*rU{npE8OhxFo$%+-PupU;o?cr&I;J5iMR;>4apvRBp7IdUpn1>|Se<wp#
z5=&;eQ_JN4C|(2n$u!i2%%<xuU3XS{Df7+HobM5_@=eEvlR$g9XoA!8w8s+c_^u13
z;LV0y-S<|SQf6(JlNPD1Dps4Auab!1@FStE%viwA?WTj&0@MfI#9|bKOI{pB4XR3b
z8B<)&b@NYWO?;@ITlbyJQ(Ffc8(P4Mwbfl(zrx~|N9RX0?2^H5*&dF@WHHMr*`{ft
zGG-h;pGNCf>_{%xu9s<v4uB?<0Sn>{RD8=@6nqly#b23~_S1jEcb{iAJl9oTP<umc
zi|oWiPX=&db??-iD_U%?L5`e$xK?VB^m^8s16L3fE=!*bD$6Hq_iXdCfnjt6do4wf
z1};nV{6;idJM|_>O+2zEPK_!0jyM{KFva%P7pK9NtszOJ=x!1IpfQ7#vw{z+HAy{H
zkX+KN@3bPR8KuD<Lr*ttd=s2R(xfPWhc`=|D#tfTjcKAcOigIYA7DLbNNtm2#DAm6
zr9Ng4<I<8;-jmVHmkcXtrjj7DQ@hxx#o$sixxLiM<x}tA1HZo9e5Zs6M9^bUXWC)~
zf&qda(z}<piWv+8SzV>wzW6J41`TJg=Kp7JS2CgAH7!gvZRPZiBbKt;IE9i=j!Ez(
zqTa#X;z72RU3@uJlhlHE_qL*{VcsTmnTGvhm?GIro%%W4pGxb9g+E+EiXQ2M{Ka43
z!|!f2HgK_*p<BuBUj&nx@n0BN>_77<|C<eGx7Pm;#h6YQ#_R(C4R|4oli!Nf^Har_
zl;>pm{+``aCLyivy5$1bq>jvHKglYfDfK7k^!4@d@_l)za%|{NXkh5}%R}jn|4^62
zV`0%O4xY2}`egicO+<pxm-i`1*19)+Z9U+giqqoEEeNOgt6coBPz+(^5f|2~l+`Nq
zMALbxdGUig=-QSh1_t$0<Jvsw)&KU0*IcE-v=AJYAO3BirOK3zF9_;)T1=7}k<P}%
zZ{#=H1<|jO4qD1Ho=Fh8W6DwmOZ<{e$Mgt$#4dx_+oS^VJ3B8oXT5<W0KMpef}VFO
zP7KU*$@vAfT$icXHV?qEU^XV1q!zroF<SAy0zRiPwdNF$*2If;ZYs`kGZoWut5~C4
zA!sIok2|be-_s-uv6wZ133a?0wh+8bYjJ8^O!{x}0k*_}p=<|QZ}Exf3WIja8ry}e
zb95g|r&SbXEJ(aFgXAwCPFb(Fv4mzk-`9={=?4NHZu97F=X!@8bm?@IJ<%YGgbAiv
z4cg26VJO?hn;{@h>e+Qgq-|Hqz){=l^9J`(2-5QLyjO8glSJYX(S~ex!EA?jE*@V`
z-xvF4R!QekVwvv`4+wgN1cqg2IYGmsEI+@j9{vx`#3?B0pD7-1!c6UD*q&pa6Vl5H
zSi+0e79jO^V%EQ=(P0_I?QR8TQ5Fr(PVZSoTT+(&iZHG~3};OA`OLKq;~PIB_T7o;
zWyzeWj#qy%jb3$~6wjiKta*`w!`IQ1T9@~jWW>sPS;!9qLiIhEc}qQUCY?-{0u_!n
zV|#iAO?lVk$2CmF9&ub76bsiO*=pXjo0xU<;*NP5Lfl}a0S~2H_kV?Ci`wSJ8HvG<
zx=A{x(&RAHiF|}dG!}uNM!PzD0DH-$nxQD{l_Uiyl%=rUJ?Vmi1>aIjMYgPA5%Z9L
zCftzTAE{5|jkaQ2x;x+2xlZ;TFB^TInzCQUa}Oem|7w29y<JrPsPh=#xWI+G2IX~f
z;D6_Z6GM_~?}-npiN`is6XtqKcH^lU<3OG-Op7;N^>Qv|C?BPMH!)}cmR2-U6HEuT
zX$2H04H4B<cAyI?wAQ8IXy|laLl_7W7Bsr6f6K1<M4}8GCX3XqSzb~V5Ed%z*8Nl2
z+`^<|?Tt~EmnWQuRJlyxSk5bKfPUr->gGHZyc_J|R+SL|Ef)e#y5JCZG&fB)!x)Vf
z!Uhead%MdEg=-BObTi62G|XngO{ZX}<Ip_49hbi$APouP9aIK-yiIzcHi_c`K<d!s
zj#J)KdBm+g9T(nn<)MqaR_-@z>pH4qn&b3s)ZuAMo=Q1PMSJI1+Ok5U66-yj3Y^&k
zU4;=W&5t@FGgtKoMy;K_zb(fa+@vz+VSeCvO*SrvG7WA<t+U_gn<8)Wd4?3Nf|c1m
z#p)}?-V+;OUqjkPI(=FP;ej$|Mnr|r951a%n5%H@BFC7$Mk>_$fHa0AZfZPu=cG%H
z&QVVKOWV>Yd$uB%c&Ua%dd8btx(8pj8f<Irmo14#4yMZ5m67crnfwlZFl&h1)DB`W
zYreoijnSs5r(W@$FhTA_{VaEYhBUFsLe&<@At1&^cnohCJc#kGg)5XFGDZ-onmuPF
z$KOL-c|5S`6#&jWJ{W30%mjbXGXd_s0yE(&b8C58{o=g*F%+d(^8A)pT)8Ucw4LL2
z7r4$}sFrl8JXA}1YbYoCV18ulw0_5)>{1%%M;;}9<u32{ma?V8%q<!R5?^m%8!6`s
zQme^#v$5I|fL+vGXYK=VWbN6y6YZw_Qntgj_DWLcxi-EXG_E3ejJ!muUQ^G(VAzjJ
zODrm%PA#}AqZLzTIq}o`8(IZ>20JRCJ}3=<t!atYTknKf`%j7YyQ;GcF~wpm6Rrmp
zRFgAWVkYmRBFbPPOZI73T;Yg?VJy6J@PiDXd(eIMfegh1l!C4YsT0?L+QZO830w}H
zVA}2XIBGayQ#l0_+S|^}Eb}bH*M6P*0TOH21N4nd=9n=+Z*Plc!z>N9f46RuZHQ_o
z3C`&fy5*C*G7Sij-jT4If~2RTC2Ir4B=r{DrR&iLvjHn(U#%~AdyW|cpv1BSR~g+7
z0FUJ^@jUc5{x%1Rfi=5lP!A7E%nq?Nsq)gGY3JIYO}M(*yv^}AE#5rhbV5+Iq19LW
zbn@5}wH}}VPbjjm!Z>GW@Cn}pdO|n%0PWssYOo$N(MI2iaeP3;1qHK@xJeW+{C|wS
zQ;=xEwzav-wq3Q$wr$(SF59+k+qP}nwr#WP-hSwg{!d4oh!q*@F`shIoH@q$%GXpR
zAWB*)LXCdPUu_U<zLX{)CJ&r<Ht`rvI0O2AzKqwA{!=kz_&=vF3=Awx|79_3i~gOy
zOr20Z0!cQgQQ`LH`lYR%fn>VTr|pf_;;`5L@Y>$Ri-jvlc5%Zlh6tfl=xh-kJ+2Uy
z*vQu4$?|H;%c`&eRusqZlK$WC+0uduAn>62@2?*3PwEmzukJ^Zl3}AMM?29yXZIj|
zci{NlujIW)5lisfd-SFA9j5b~qXmmtK0rP_fghj0L`zV}PrOu?P95)$*EexvW0PrQ
zNCs~P6{uomwgWVxJF4_Di(@feT3uvU?|f-@0j|L9*J{WH)9frM48hou)%(nTO`4vd
zH}W8awC=1kFbR0Kud{Q#z~XrcQxlJDAO-{w2{MsD%VMVGh<b_Gp=?B_`3Zb()ayU<
zTe(P&{G@~E7b}h_ZnW{I>hB2YK%2&v1cb(n5;VLOa}tT#m}zncH!131&9{k<G|22F
zB;-uGun;9aYlzD(Jw6+Q+E|8^&fFUE<t>9{)$GcGV-}Q~^HAi(s{ksD%CEkWi7L*P
ziLq4jbBCbeW8dhMC)sF^rNe3_nFpa{0dPM&5Sk$2Bd6Iuop8KCGONr@gwIKFVa_;I
zT0knEw@4`2LXvPlUf<Gdl3XgTsMP@@q&ub8{KI$Y2Vw#$MsO0PE8a#GJ4(Y~)XE2v
zS`?_{4=*e4U1{~x-6_Vy!tLtyEs$=U&8+c&iw1cd!rY0CdoN^D^0Gdf!=tao7FM(i
z`Peb1M%|2=DdC$o5KrM5E0b+EXYuJjTZ7HV^-&2^36*sPN;mtP%b6AwGpDK96Qe>#
zrB2T^r}pgQiyQzO7evPQc`|xoJP)n1q15wGg6>04FqcI<s1~8X<O_;$D%k%{nYHEY
zIokAtlwy^su-@5Az@Q_djiQNGFtVoHE94R~MGMo)PzsZ7now^{bDR?0qQScw*O+4*
zQwmQAY}U3B>`{7~wjPXWHW$+&;)GuqEEU#fqZt98!X?T$>zPQ;qa(HXi4E8m<{lh!
z?D8(PIX?8OI-)29<M{$8+jL!y7BhLHijp$h4YQbIG!;a3B!%@}+?sUYt6ZZJC7BOj
z-kxPl3)~^;AL6N5I&61NtH#3UB9$t^{eUYV|1{`?s}wNPDxAEg?I&!9>35(~$-9b_
z9O#EdbujN2|GES^brr9u8Qx5g(ze(bRGFi)FVcSV`;?aq?rj+V6*Vj!o@@3{H6rC#
zPZ4O6<*fj+alG7j8atwTyIheCN}H&>>s<P*M9+zE;;KU_k7n+U)T3==>&XBrxFXd{
zyuL4gdHlG(n-~pga9jhKzJJ}6-0}g{HGDu&DYDXieT+PwEYNbVWwsQ?e#*S#r`4M{
zoOoDobs6#((hFHp&_hRMjH=O;p`rbx4;(l_*h01kV)W@K62XpT?sJ4|R>jZy&D<lb
z*<5?pq`QIHrR5b=9=1Ngwb!-dby>5&f?^dw6K{DOpRp9O-YpVgx44enC006+z_IbC
zl1;B~zIp$7a8?NlQCPxcr~ZW@!pr3&xNAY9L#&k@X%5GJS-pkMGr5>|p-##*l)TXp
zXVX7sXzVeaGk@-An3=$6u*YS3Tm?1)5j;04*;tgoxtxWS$MWn&jQh*k2a(Bce;kmP
zM=O>GWmC+eL!1Qe&v%OMgHg%(-@>v&W<?TuR@qMUs2zV+96TYYOe{PIy6ah9li^aS
z%M6**d}0%>O*JA^BsBtdX^5{=4PURm{`>a}r+jMkBHE@NirPQP)>@VCj)U^x0ae@o
zcvI_r8kZ1xj?XJ@iv`Lj_<L{Nv$PIrdyT{xqC&JND4kF2+JVe{;zb_Nl3>Z9D?xU(
zE4mhJ-;ch9u)_d*$On6k%EGHhnhOHR#nc@WVc!hvsCWts-9oL!=vu}QQ$Wf&p;;u(
zx8W`o7UgEg&}5qhA(FAZSuLm_po}He*yv@X&uAE^z@BUr5C2OvZ{H72EU$=s8<Skp
zX)pg~I>5|4a1kYl4A=B?t5be64_hqyc@W8<idYEyVLWi?stxnuhmvnCpJxwT{${{c
z*N5jEr(J1-fCFF500L?7_>d~%S7T|fo@v~~7xyWcy6pZZ*@m!J-;k$7^$_fq1!tCS
zRJ*7Mf=<dM1Fq%e7%;NDNkesHmU!}cuHBGZI?X7pA0Y9v7$5ZF6-<jEE`QB<&gsEE
zZmV>+Iuq~tpr*i}^Cu)P+P^cpZjWM$HexlV6cxtB8+(Ie94Pacl`0r~y~j>;5kq3s
zRcRa#G@3JT5}$R@#pv+$&umw1;1K7yzY?f3cJDO>ALpwpJdZAgt>b*=Gn`!s(+CZi
zi=9h0Bc6=mqeD~$xe6TIeT*U|Af)6uIu&>{Ves2eVsM`WG0tH$nD{-=H6Y8Lo@6gM
zF4{O4mGS1MGt*y^^5pdEuJ!NYPFt&Mm*r&l(fc#Z52@?%_7?YP2Y+bEuldrY_93#L
z?R7|{Ss`oBw$s+J4Hd!*=~XSW(#=_+&bw_G1U=-w=g!QnKl~!1HAEb>OSz2*I#tIt
z_Ak#QHAvIZJAqeI$;9+px!6*r0+YwdVylx0?#vs`wd?y>PIL(<J9OJrKrvkII@<hA
z9M1KcU=2-wt4c3lH;Sh;XpuL>WyM!Alo`)gtwfkL-LF)1Tw3)OWJ0eJzkZDhg;^TS
ziY>zjuLOntPCTAXW=6(k-%x-Pe&Oq&R}NS+Cp>$1TV%D5mi%y+A_Tb13Kx-Se`(r}
z$VAnh(|$UNE-1e~BbC_8OK5IAcOz#m(ok9AEmi-*J;nQOhL#!2mY-6YVpbGTmajKv
zr;U1Ww4ZZpdLsTSAy^+E?GydD6XVz;cD(m85S?HyXQ>EQ2{I!3x(!awLsEgGM!vBz
z<t+bzv0{DYk)}!RE<B#_6x#KCDV~MA9#M(6bYUrSzu9Kn==zcy8VsUTSItq@op#s~
z3!G#ED^KRLfXTTs1+cf^Kpj7upAr&KUUlust(Ns2OiZ7q(&-lZcgq@%9y$?+19f6g
zmXDe|)lNL$#`kvo=N8Yzb7Ay9bxDl>f%?t%+dBO>BZH&gkpUJfeD{UwEnMo}t#mKE
zIc1#RHukH^qa3=lduRj+_v6jWfkWHkwTS)YAw1B<lKt4;<ZVU?(W)rP>eltM{>yPG
zlfx#KeLT0PULYFakGS3}=5c7N@*r;Zs?CakSi}Wt3aS9P7HkgpwXsjp=s!>lEmz6L
z9{cot%t?8M8fBIm`BY-KR-|iIJM4>q+P@0DjD9RQ&po`K?MOkt>;|RH)!xbl>_+!#
zre|g)noCoy!|}bUVefTST&J{(m?MAS^`@2(u{`cJ27_k?iG<_zY!<PSt$C>&B3Vy|
zgP?mUTcgUB)(w234?wfzoz9*CGgl3+D2j%<%)l}REX~8mxSqc$Ic8T=^r2M`O%lI_
zLm_beK^-R_`xToU4TeY>r`;rnJ@g^}GA>{cCe&1&_oiU}JifHt(ML0R1fVx;`<POI
zAZp=}$I6J&LVfDP8Fp99FtO6SdK?0~W7mTnN#pOo7WLy%7tiDnfBY4C1V{r*?gM$&
zw6ch$hN6tWtMz^qdF<@m;LZ+>D9PeNykTkt>$@11gJPXqFtRte&N;G|Q@z{Jav!ga
z5wuJg*4las_k8JIZwX$jTQN#d;3FRQoM)YZO~aFM)@}6J0P~v(>itC&;3t@p>ph05
z4RJ47GqbUMX>sPtL#O&ok5$8CGn@mFH*FE4a_l-(K}zx~r0rwF4P;@9gH4VH;SZCS
zgTu*iF3kQ)o2FA$7+a3cT^q#}ATA29z{g^)%n$Jr1ns5z4?>CQ2eipcAgBv%VJ;p;
zHJrjFzWPuoI4nl0{rkJaTe8u{8-l$}13)lBkjJtz5Y8v3C*j2+uGgQEB8peUFV`Cd
zW?@oi+ChrI>63GhQZ)`bN>)T@RNofz%9>M6ANul6%F>z~=4%5hfK=qxMR<muLw}$|
zr-;33{#5MkI=Y_y;gs8gtsOW#!8uXLof74GiAMujM`RH-YX-F1<zKTKa)D{2V2ciY
z`sCvztwlH0UD_{9*79Jzy}P1c$-;2Zx5Y)rZSG+@1+_NKtwVH3g2G1fDNn7SlA6##
zb=a!{yp)IRKn_U!iz}e$WWhT)^Fg_3Aj{u9<v}D@K=O3XF%y>ra4zSSs##fgG};$>
zLc&&oiowsJ-!Anc+`M@`J?eYD#j>Z&94(kHH_dA_Xcv<ao*P(_?z3tzzwn8KBx!sF
z1sbyywA9jsMg(W|y%I%zydq?ts8`*HwxCLt>^Gy|0O$aGfYb$8C6JCD&dZ2@7fsC8
z|5u5x{}oSiAhU3ez&~eHm=GM&XO73&=KJk+FJ*)`^xs;K9mPI-O0Zlh?~!PLiA|?*
z&UnL-iWwxdDeAcSq9s`i4#{GW)o$FCK{em_tM#KIc2Jfpjzw`Z`f^@p1kwfQmfR76
zf@OhIkL;J;k@@=HvBlL;**FPxEWEnz>hPGm^PCH%%J+yXp4u(1x$?UCA#xqV{=R)1
z@rqlbL*%FQD#7u)UH@if%02)R%7U&=D>?gD<|u68d=Xz!`O2$h!)n5;O(*b1iXaZ9
zn}gSITu^oTc|P;qfZAzWdm4enaOZ~Ll9?D;qb<?O_KWMZY30W;OpyJg#%#$|Fw>$*
z1aIr1ppi2tXXZ8YaDpNo_SvJXa;O_>OMHLo8}$)|<STQ;>u<{-c!Ihm6EZ`oZ4P0w
zC?*yr;Yw01anL9wjWgw&io^I9=>p7dh9zw)o#;<1WpTZ97nUKQ6%*b;hn3UGPxpcp
zZiei|)4hccdNm9O_1bbn4YE<)KZjR_tCWh2ZLh+95|$Uk(WW=yd`{Cx`=lIs2h@&E
zgwcH12a}6IXDb&NQR3YH){~>Vd=07G%WhNW>v?XNt|kzUay80NF3vCd_dWIdq!A5p
zb1u}U3!bQsF9bNrCsPbJXJYDze`gmV23BLHqz`Hg>tlZ!=Dw$n7bI-E&hsxYcKF2g
zTfq}L*sn_hbzaE5Rj({9&LcaG+ZNfT935Add>=Ho%CZ8;H-S9w+j$e3E2tYRXdU*S
zi!FjSWr|MMyn^%P4k@&nX#~S><Kv{CE}T>RB2Go6?Td>7v!gm`vqBb$IIp<i4LEN3
zagDMqjAF=fvW}IlgCBj0V)CfaOXAjdM`^DAp`$2v#+_^64X{~9aqR7V20rym^km}$
z#^jF?LEe@uYXD%t1mhOV1h6CPu7>-y@xNV&2XpxhM7H10y=ugEvssSiffwRIXuVQ%
zE?qQ@#dU^OX1{aYnj^!EM<VxNbb>@Xlw0gJ(PO!ga&@w}b_vcG8cX-Ml@#w)*ccEI
zw}eWcv(&qGT)&TiilgzExNg!Y5Yri(>b<=&n{!4yaiE}(l<&U<E85yICqlPLHh*4S
zC*PB`>Z+h>jof3wW~F<^Nqw@1ZlWq$jETScr2v#@<KFW^QhyTur+WMAdi_5%A?!^5
zox%S~Lo()nvm#Yr;UZOtgvS6u0Z6jJW;MYaerZYV9_xZHKX?*dpRqWkbqiOyr@un`
z4I*K${R0QbhpM@$jlD;^*L#U#yK*DjRZmYVJ{r?EVCsNmtpDXy+;zQO$y+qqXCBgN
zAdT%EdQh71@W40PgVkc&qwK!!LI;{0;`TpMtvy1IX|Il99DX7XAGy_3!6IhY19N<T
z-%aqF{Ojp-?=miP5(&H9R+uvwWI%<nTzJuAj(94T{CCG%G7xy9z4&_66r?OYIiB7}
zI&*J$>nFnlAm(5+xvm|L+a%B&o&a*UomMn*+H4V}t;1)iv}^t^M<<}5k2~0i4KQN_
zUkp})Wwcy4hteisW8kdn{psHB^<uUISiJ5j3WQd}^I_yguvL#Qkz)s+0<QS&&g9|g
z)`FV3#=<K)Hxeqh3Y}N|Ph)tS6;=Jc=?isJrAliy8Y#6xb|-?<*GP<wH1qxU@(dmb
z;$dAgc=iCjUyaH>NtI))-!CIpWZNTj<~V9!hBxDh0XO)n{!);Lv^7@)v!^hsOg9Ej
z1?)4ZiadIP@HMq5v%co)s@q&MD8KF*CGnqukDG*VnI~sP*LAou)n0#QuodwJ>5Lu+
z>E%fy@ZGkBNx%LuyN6t*8#e5Rl{HfMetIZ1$#?QrG5I@<Cu4&xy%wA-?!h#6m1#HT
z$*NIe_Pe~Qv5uj`Hpp<)=Uw{jc1TW%&I$5|a4=3ke`>&)uS}{B+{2zhPMStND&JUB
zxmDZqem^c(3#@%g?#eV#r917nf8CSf%#cTTh*ub~ku}7VC}bHq0WUEt(t@^_x1DTi
z{r=EFMYbM8IKsIbeC2evXx>A3NkR^~nn0)PU;cbP`gF^ClCc*#N1S1pfHi=vKLiP^
za_t`b<78I0M$US=5WuX7wb+}Q*HX_-CRb1e+}M`DpA@A;xpl7p&J$C?{Ml#5LftB3
z=nbbf79}i2E4#LgY>957{1T0|1^j#Gcu02k!8zu3*w%{=tkOKSM4XbA+s$hI08z(6
zE%fS28%uZ$qyAuh$r=?k8+i*>jw)om@lRWB1uf>p>wfNLa+46=v|+2sL=<7`5kfT5
zeo95x*ePhIll3dNA*@pT)rI0pFtri$niZ(TF7cPAHRih2h+}-3QdLfDAvJ;|*0>sA
zY-?SFO)~nuAiB4$8O&GCE{gb;l~-W_1Co!ZK^-3118eY?RM~9PMg7q!B>-^#U@{~O
z&#~~ccEdb8D#n<9D*UraN4~~)YO?L7)7&?XYofMlWiT+6jzbRUP-IL>J}?ak_JZ50
z-jU+Cz3v3QMIuL01l7<coIOV*z6-b>oI!k-hKb!C$X#sw8md7`cu=bf$mFh$jsE#J
zq?=}Lh;BvD-l7yD>yNYGZ6Z7+Cn1$nJxR7cTcIc!w~(@1$sErus)vjhYkXCBman#2
z16HRA(SQWl45Gc*{sP-(wR@Esgz@}GXJPlEl*oY?@V7VTKUyiQ%fm3?%ny@}t^7#N
zgDM7v%_z=et<GboO3Hru8qVgCkJgL`Qc)0dA@vk^S6HLGi0uHN`En3Tsflqm&T&^b
z-#XzmkaU>-EFn5{(O{j{ia_1!H5wGlVEMWim&#T1FerLGK@LhIP2w*p6~E>3-LY}x
z?L>KeBHv`=@)SZA30E+C#w~Y;Q5MidW|Q(zhow%czfBBOM!rM?y*6f(=gHNbJy|6h
zk}d@FAbl|%?dY74E&3tDMr+LzFwo*2q!TJHv?^}(7grM~Qp97~#62BKG*T*u2LgrS
z$aicY7`)Y6&k9b%BWJ3|y1D0a9MY7h71`~<<%BH(iG;~u!M(01<g5ma?%R2`C|jqO
z!LdsHih0POu!rwP)S11Z`4U_L(znDkc{p`s2TbP*04OHQ4KEi|VK6lSmRA**FAd0{
zf)KupaIzCQQXN^*z%l=3TImZGqto(0mJP$aB~_EpS}aoG{<6dM9c)jAFozfJl<T_H
zGKB{62EFFq+D~tA9?xm)f0f2a+=E;fbbsU!^HNx$t;Vc3PtkL4bbNKmxv?r}9x{SO
z31+36&sEg&DIjvFLg6UPXl$?bV;oL*=>$G4djXVYygU14ST0jKozV0wcV)5C@yBP+
zcyXqo^viL=!FZGcEkf#D7Z`a@<_ZaO3dUTE5UqWK0&3F2Tiu|)nui@xk6Li-1Lh=S
zMZjD{_i@HYQC6M#AT{{#Jj#aIL}jnF2c-7h3RsWE{vgsjuxL>uFxB1Suwi~;W5cZ9
zuE4)OV#7dg3SKm&6ws#?>}_GgV5)4-${v}j{Ba?z1|1VKW`V-$C{a#+c_%SKL}R)~
zHF5zT!bCas==Q20-cFe)E39!~hbq3@wKret96wrYe|n~3ObT>&p!Ur}-EBkK_)?p;
z1%!i?@{qsx-t~WiJd19v0^UYzR@}QN&BCe&YE0U9np8<|S%|^cX5#q7pu<<^W_cOp
zC%)5y>*}LK4`u%D(b~`i1rQ^aeqKni!2cL$dkGt7jtS5Eicz*e{FnBn@wK>JrRJcq
zMxC_I?33mF#`Xzw$DX<JpDO%sz}f$|!ZR@cm#U>9O>rAaR>c41A*SlbQPKDk^d=X9
zP5$zlppKcu?8gZne&7^|BDS>L<tHHRXA_2u6Pmo6Vr8?x*~S)PQetXSbgkDjGjG9V
zvbw!_0k~ryX~i%bUG4w$o3nX+xxGJ0|3UyB`0)sfmLKgNTS{!WddzQ`fMR(bk#|4O
zp$A<|qgCFttsXejjmx0CfV_G}?A&?9C=VcSI{n#yPA+t5<HM2wyB5ERf<!K;y*BWd
z&U_vn-^OO}2D3thuM<&C2gW!EL$mvF5S_#l=j{>01O$7cY)W5`P!E$kEAIdo8x0Q8
z#ASvLQRu5uk3xt<9mf8ZbBShqh<8-a>pCm)w@Ps4q&ctCnStR;a^DI06iif<XcH(Z
zl!iNqn$O7lg@Eg@kOrv;Wi<l#?BuWa3G=rx=7w>!(!ubo$;EepvxM_{f1P@3_vrlQ
zH!MkbLJE1k-{jEV`2ip=2PM$=ThyoS$n^AlcIY;#slQ}t5s~w2M66=Q6={D$1fWef
ziX5tO@L(adKU5^$QU~PDFmuLq8jXbQoR9xy5L$ecwGTHA%+#L`BCvy(#M$Dj=rcL7
zCz2vJ)DqO!X~KA8XCK7J2@>HuWeqbFMU4*;rr$*6vQ*z3k2~|l8uX&>*W>pLIMf?s
zlkO2z0Fff;7A;BUPo+x*IZNaTNh(T09RJ=LpT`8bYa49qTP{TMzP&+#Urmz3FC-Dw
znm07srm7m-f@W_SKyoL|d^w^XNKLN0s`czcR07_rfy|LjqFu%J-L2Um78yRhg+_Q8
zh3zbQou@?>E~%#CP{q*5hcPN<h}(>`jMJK(Qr}Zf9lT<9Y?Wb?)N4oco{I%1kCfi6
z#Sr9k<EL5gLGjwD#h0H(CNb=q?4!lFj@plk5EmKVaR}#5db?NjJaIW_$6Mg+{-e^8
zAek3$EF2i~kw#Z_cplq~v8aqs+GrjdzL6N796mYwWJ3(gBN56=nRJ>{O2N>!q?Pc2
zj7v3AO(8IZ&eF4GpXyS6#54x7Hh5;H)3)I2FOUW^l*N4t$6P{YM^&6~st9cL9I*5?
zuLI+YTtUcbKMP*2&@>m;8c(2%t3lmnzM3eTU$=#7<|noxItG{+=_8ya7X3!QR|G#?
zU<Q2g3;Rlv<)j7a$-0c-3qV`|sfTx#nuoYu?Qu2Ix2=wMHPv%X`D<k=2M6VI`?vWN
zK+sN$Wz-67o+tDVPyvqByT(Oe_Yp~z2&I0$E=xr5`!L`M(^#Gsi$w9J@@Fd>qz#Y@
z1(4u->Dp3cg(`oOmBUb_Up?Aj2Sp4|-QlMX`}PT@Z^*T-^TuE8$sui7$3{ydsSpZ@
zHC7QV66gNs)BGEwVNlpajACKIqT8q)d3EU}F3tlzeh<bey`jXnR*Upuvb?QalF!O<
zx+E`hV*ZVLM%Xcuf&5Pf+3Nlu$r;|p&kc6><igA+R`fKYF6qV<`+&gfCZj@m<sqnV
zWeb+5%0UCV@GmJ3U7(*J<@A<B_FAH+L@6)tb+<iChz7axT^9?r2`S5b*-AzP4g&FQ
zZAQ6Zz+K|m;S%<pPaU^b69%Pk(D7!?bESQ+1>3UN29cH<VHyRpcUMrj69~+gHL_$=
zf5h$7qS-i3NNJsT4aK0@%c-YRzUv}G>#54Ta@wkViU_k8mJfnuKe_4WPeTot{C3pC
z<p3hCkg34np!~ofk7G%@Xy`&++S+v0ICI{0O$viPXi8m7FJ;qyhM&xbl}mpN!d$V$
z@dRq9JanW|6^fx`lLtoX`K0H8Ig*6r$Eac(1^|UO+Lv~zHfYMGRjt&dxSUP5_3A0;
zawP;5H}C1*q(|vYiq@J+AxUC4bQLW!xftY?X$Yv<n)1U3$O|pl7cA_W3KTOWGJ-Re
za1k2b=$2K3mCSQr>!#`gx)~Ar@NFgTDj-h9c=IX@=i#F)@}{Ua`<bUBzv92(1)HVk
zeZX*4=^3HijKLNJU6YATe{*qD+ozaU?4DcSTuTHX(gFfZ$`Uu?zqXz2Yy_AQJYD@I
zBncndD5znk`uJF#G%yoq7R@U8vwBoZ(v8I#`J(zl9LuSR6%lWi%3%Br-WeSwLL;Bg
zIF5p^d#8$JVEC({TgOq_Za-ytjw?jZ*%bE){9>0Wua?^SHgS1wTaOgf>>(=eAJ~;u
zQ`*`CQ@`6c7i)}xX)qTzZBk88C9kk#Hh1b2WlfsGwN1tGV&xz4C6}v>$^|R_pd~o}
zGndIf7K4*Q&=k#cTvNo$2Lab#<b&Ln4)?+(pI79E;WX%j(U^p!{z%)_NiQKE<1g>c
zPC0&}tj;{Z7JuR@L^`uRqFKr?7VWh)PW4(*f4ciKt}fGp)hi)$2+X!M^Q~f}gISnT
zobn7)xBv^{{s6r{zcv@B<v_0Er+NaQIO_#1wFH`KoEqqZOY+aR+5Ea#C#AM`a62JB
znDhPfgaVO^55#=@_zD!oIVt?4uaQN9?F^IRR)*qoCnQ`VoIRQw2zdhpO-oo8Z)C{$
zE(trKVl6s$gh2}Ymv{0Jm4Y;>@a4=oF>)7X4#ul1rNJC<bUgY%>rTGHHjhN(UZ%#K
z%Z=Lpnkwu7y_)H>0zy~hKl`ly(j(GeY+Ea~A)%{aQGTl>sp+*OsnQ~rtmzSwVF#b+
z+l^9Uu*y1e?GY!=o(}X5ypR-Te0ZYm(x3lg#!7kY9R-)#t18-#qWHp`^^XTyoI9$X
z$4EzfgQ57baq?s?2d-s2T2WC$n~-?~X)yENiSp{o#vj~0caw=ZZ?D>cV}Gc!K|;r6
zeK1ZVlan}r3$Nz-V2F}i!lX#HUpvX$XTbD=4JT3gyTjAP^y3q3KF#0!wuNIbWBS9Z
zT}{inC6Kd7hA|~x&ikisf&h}hsheYbYI{l@x$l>p(0sqtMI|?09Nfjz@=~9Cepg$&
zQW2N;6sl@*YsF(9MS4Khus05xb8onY61GgLbk?OrW~uA3ysRj03|ftex4{#Z-;b1Z
zrFt3@mWHIh1vSCMu`SKYx|iBwskpH5^!Cr2N5`1lVIr!07Dtvgt54CG7P<Wyb0(hI
zn=q3icQ$HZEm_*P1w-xzB6=L*aG-8?G$eNP8t2%8%b`?kO~OU+ydj2IdQTy0y0t3X
zZ(~5ZW>v<W0}>Q|nE7WTmwVUQ=Bcigb_{#va1Ek_R4p_ERxgC`u_Kpprt`658O`M{
z^yJFN`PY6?`V!D5-HK7ssHCdR?=iJ6^@hg8uojrY7eMr_Qra+PrCaldn5b*1dlGiF
zBB~rAbEYo?D+=jiscKGm7`C)uh82NAxrWk7W?a3oMMK_t_<T-^Wrz}@E8r}(rvqLC
z;Bl2;ZXkc9VYuZydH=(4iP+&XxU$+43HKY%|Fk3LKQ)mo|7Q*)0|O)bfAuVLr8!}}
z$%eYM@@pNMj7>?Q?WJ3<UIjHy>FwyF!t5jJ3dPFG9{T*oRv;plS7<9IGI5^+RicnT
z^t$7LRd2lPcB=Q_;p)WZ;WE3<ycJ<T*XeeKq$T1>5-MQMQj}1uYEN0Y&S_DIAA0hQ
zC|bl!8E}c|XzQ38F9>01y;Ujez#5^5=Esx%QnUJ?$Lu-{?dbpFY5wOKt0DsVlIGv`
zu`<C6%K-K9YG2wPhdLZQ9giSm|8n$nTOc@cLi5BUmKVUf<<qf9g^U54g{(quVh4xm
zvz2rDx))=k>+^PBfr$Q`v#RV&HCnpcvGP4v8#_tY`|@eYpP`3N3|M$P?aEJ}Q3_oN
zquUNN;9eOfveiAJLY<jnyqJ=Nu*`Vo*H>CL8cUYE+<)`YS|`R_E5@Nrzpq#rKQb&D
zNn5c0uXA{&h*IQh8K>JhM0TxwI6gxAZ;BYX+9$O-N@o%l#;KF(ybLhf>a#>{z)&R(
zi%jE`6>t&t0Ajch_;gTpeK}-Ga@>3oL?L9*ywbJRM>qne@SFnoA~;Z9l-#uNi_%(s
zms7}%uFgZP4f-Qj`-`D<BCLSle3VmdqiUQ=xW?qA!s79DkxTSCx)CUXDZJaIq7sUX
z<lGWGz=Z6USYhG%RPIKJ!<3&m{cxs!^xs8)j9u(N+97Dlu?o{({@8tUP$E?CP!eDA
z*lfCymNl|6=4Q6WTl(&iF4mNAtK+=qmNfs2$CjtE=-wdYF<CFT=C+cHjmbK>!;epr
z#(H|val3vVE`17A$EDFAhEkQ4L_S^;jjvN9NweSz!ugCj?D~P!Frsp?u0qmx#4TeO
z%jO081$j~@y?XyJ4f5?+91MhpP1tzZQqJiVGxfAWHi+TK#d@S7k0HYv=#IJybdE9!
zr>@J=B5PEgHMhib3j=$`ot42+O`b=qdshXgU)JuKCXHqNU;v4mgn^Uu0dA^@Q^woQ
zN()r<s_IkRF(P&s>7d2<_j;j1`|SIh+mgg#z^yq=5A}8YK%i*fX#&qbq1yAs^uXpt
z1yyqtS*C>M%)8*nC6msb4zn6)32TwkL#1*WbWpyDA=Nc`PdxcC^+U3NSIbYmg3r6>
zO)maCJjT#Me}vS2GAYlvSU^bPQYaDrkB6PDm4T9E%Vu~6dx?p=fl=>dQtBMWKxjlv
z2@CzKuwYcs2z;=j{#K4LddfF>oj|I>^e|~_#uD_tI9O_7_XR6+)SzYg=leLN^hQtI
za#%>7^Bn6oW@})zGdd;VT71SK@rh6aNX;wb`-=$e9)-iuPbEqEVuY{M{!fLfvHZV*
zZBa%APaw!yh(spdxvoIMgy;a<3ZJch2JkXd#v*G*6&^6594#CYF-hcm-kq!0w*UyA
z+0H8rh3Gpo*BsRpliALq<^qwcj90KiaEH7NoC3rfI$JtN59x7KPgL_q2ldFowYUXE
zog`2;i9LISciD}R4N_zA(ZX5=N>E}L0`)@6k`r1ekdExemrtv-MUuyIcF5QGbUA5I
zhx23ykQy4)VsOI9p>lDD-tp(EEm?4t5`<49jTfptSum7&Wz~_zgBL5aw{9J<-Ri(7
z{_^Kg;k$e~%l!;yZ=ND~nf65oOMc~4P&3vdckhpZK~1y6&B&~Uq?MuvxHFOaPiNw8
zNzQ#5;m_{8RNwerdfC(~#)&N+xE`Bh*Wf*v;WyAdnAV)VH7eg%w-tG*Dt4OqsVIGu
zOgTgA8*h0aQq)5S^(q+9O&V5ytjLwNYSmXI3CSYx7mQ96dZkgpQ?me0LsQaC{@AtC
zL}<wZ3jJ$JxJ{z7gxZ=I&(QDW!Vk90^ZErK_U0hSt%lo{oUk1<PO8UjLdN0a9GZve
z+1)?df2ng~yfoce3c9inIP~IWyO_?ZgD1xO=Qd-SZ*f6uFl^9{tNM{i2UPTbyr8fn
z;Fr>(4}!JPy<^-K%!S9*^n<~5s#sOU3Q_)MRL729Qp?-Xh10X^A=5l4noPpTrIR}A
zSKmwh!uEdtuuZh84&T7`I#sp|cX9%*Y(wB(%8Kr;V^XbuMU&U%y-u^0?ll9!n*WLo
z12DrmEd`qxtapOVuS{8N1e=gs{%4W>T~v1dVVU!GpdBuqH+ucgUy4=f!=+Ah-g;#H
z3X2kFPDx81>iv4K6`56J>?12%2PEWg^hw0k?&KxtbYnCxvNfnl%BLrWUjmmks2%-i
z?vMz)$@gD;8A{^BlNT^mu^pk~k^oA39n-cG8&8-a7f@B07MxyX${%OjCAm!it>0~d
zsaD~%{0y%sZ2qS3gqQRwTRd_1E_plC?*7ltY@2enpjtU*u0gb#y6`HtYSs!bUbZyS
zWf?bGcFhTDtU}zTA^w>-G=n~LqRh7ui?j4{)ytSx=hEX&NRvuK{_Vi5+akV|<KJM&
z6tnG%fosVG^hOD+OgzB_oTmU5YNM<^0D~-gO*t^n2<M_>JO>z>Ke|9QlbQyqrKF27
zsv2PO%&HQo*^OqP-z@?;Hco2Wn$~LA-9uggbm<u1vO2+q<18z?K9<^Y*{uerA0ln7
zxwvj|NRdd@BiX!4+tkEv^#q5GD?b3Ri7G^inc46K5vJXk&p;aaKF?#~bXHc>lMl8#
z%dNEOHSo`=_$h*iyTxdSrhKkQWu#v#7ZLsg*j;iClz3;OWi{xN5SrO6?!>A-84aSz
zP9UMvOIl*T#k*#|n9!d!*RpmRRFF}93I*ByI7twj(6WikE~E_?zQbm>y18)HpX4M^
zZ3sY?Im64heqaonQiFVN7X;DMC}9H?0%LR)7^K#tg5Hv;A$@cJiiC?SGc~VnLD7#K
zNWO#+jdK&5=y2XfHf)JV``GOo6{iY(V^-sTO3H=U%4Tf4H`5^dZRajP3V@igJ2V7N
z`|tFOYdpc#Wj*B~^TaF3!G1*P2DKy^r@j5*pF_-nLS#)KVg8rTXYQKVe`<#QLpT)!
zE8TzjW~Rv*L&%CaeWH5GfW_u8W^e~W1&VGa0!q^{WEc_w`z(UB`afFUEus(NCgYVp
z5QP>E!#Xuu@85gE!U_`a*Sn*)&%0xiE5js({_f-VhxjaBW;aMlh(=Z<yp*J!`;YVc
zjL9_r!if~#p_arudV7nb@=TAIep!&*W1dft>-SLZtygJO;Hv8Rs;B~G>#XwZ(-3dT
zceKSdc_$I%`&-X%IO2ZT@5olHxRKK=AktKtUGW^_GgJ(lKwfTm2gxYtoBW;g=*ClX
zUo31fczvJv``IgLoZ5@h@d&L+xjPHmk=rGwT4iz!j>tr0(l*U~N)o5g2;XQ&vWw`O
ztLB6uV$v?Nh_3sHeTL=pz^Hd2^O2$_{&8}P?H)17?Yl_Gxc6D<HK}!e^}#bNeRUvf
z587-r`kJK4jV8`Q<N>EtzgXq*D&9)#Y#Vi5{kyM!yrz;cWj#!tyGe`Y<>?a+wZ_j5
zDoFRJ%1+r@fTB1H)4g5!E%t#Pc<ludN{fJjj#)q}f5bN~r(sBhR!X*>xX8ypB<<eD
z;;h2KfkeZS^E(O^>dz$RWf?H8<DNvJtNhy{J0y{pPI*yacg<PPI#9R{T2-dx{?K#h
zD9QQk6SWBE6sxB^V+ok~HX5O*{PfYv;Br7$%9;l!)X%G1w={D_ZsVNQjllzL1*3@E
zBpaic4<pY4*L0i-FW%HpdCrkatm(r<`<KZ?GM}A*YK^#7P}4MYy!diNUeIg~A!*;l
zzY8wPY{xX!QptSBv{9N<u9`BzPn)wUG4g0@PAnN&$4K++lv}^W5vF&^Dm2v$S5{Z;
zFtnWN@oEWmd_`=@&)cq<B#qd`rplH~%XJ_?pSPxur|0j-5%}ys%pmk0?>jA|KVXpI
zp~G2)^HLUW?++~tH?lw1V<e;U)sc!juSKtCGc_gGV<~|}*d=N9RY~S1<!OHj$VBfJ
zl_`CsE5@6K;oxD8#l0C7h^~tqxrJ9v5agpz-DY(4C8iQ+kp_>b8dRPOAm}4zeDhOt
zm78*|WO87=z05w}n_MrcwF2a!nW42F_{WPvh#x`RcR!=+vW$cMB}Kz}rlKj^41DdU
zdCWK8@wX81NjH!=oh0$2)vMo^F<O9eGBCy0lVW*d|3-~pc;EXj&RpP<IUCy+ky`io
zNEA92C#|s)FE&br^Z!9CH#JH}<lUKuk{4!g{T46}Fl1SWA(<%PWZH!yZ^uvwL^-qs
z$@k+RV&9AM!&eq{;<xX|`Zc9I1rU*9o5cI7&rAp?oW`0-aJ4V<+oz2A#}t|#N9LJs
zf|wR;)Oo+3%#=Y*yU;HRH)?<1JtVTm7jZ4s&}|%#r6m2yNr9H|42*Mn&g4WZjP)-#
zL=i(y;UtZ9E!z`#)fo;DY47IQ88?w8V{7oIQep?qr&{;NkNXoX&wsRxjR!_f<S~bQ
zB^#Qe>EVfOh7NMYAS2kh*>0U_rBn+60Uj;pPY4$XOg$76418s(8JnAN1s4E}v%-nb
zKk^13Hmk7i4_x5o$tzYX*#d&tScL}$VXf;BhyQ&F0mymN6_o#PIcoXaomV(WYIpZX
zvXxK$oOji#AGk%oAkZ0?6iW~)S0yGf>P2Cd67zs82W#t`0Ro<t3XGk~;gKqFiF!D+
zvraBvD=v1q=v{vkc+3t`YGr=7%E4(`Q|hcmV%yKG!bJb+cx6<&HT>6t5wmK8GHDsL
zn%X*o9^VGZQy!xq5GXz;9$wE(zLltPJa6;9a*C=0&<Pvf6F~^oh&zCTT_);~2tDZ$
z)$f0ySO1jBWK4o36s7fP5jiLu0J=Al$Nxwf{N>7YdQcl2BN~ucM_<*$iz&p9?xzwo
zlGhi#4_44;-3Y_9I`b@=*Oz@^#kA_+&5L<SW|iBA@$qVHYMW!th<T~;12d~P)58%7
z>JdINrzdNN1M}RhNWHg81FtcL-I|3ddJjSW{vF`#ml!)4n&mIwPgf#Z(vCeGF#6?D
z{&UdRF`MFujC5$nRf-!7R`f)x0e8^1Jd=NTa>ZlNWa$fLTfoHkwjsY+`c2X*Z{C;L
z4>#hVV0WC^4l~0IzanlMiS@vzwIctir?nE*?050L_D!a>;?Na!;XZvc;)GKKqs96U
zZwF@OYQbvQN&BKdU^?W=0C@bpQ_y1+CgIfC^oFFoHx0f(8hiBx&`<$x*&7%H&?t+}
z+z<dn!pGLEc1^sqNAMY9?fND2v-+v^<S=UtY@G2*>N?S6(@6{UWDo}GmJvp*6=QS$
z-q#qO_4Z9Q6QdIxa5F!DgT%7`OJ5OvtZ<(D3*`u^BJHKDS#Y?H-Dl$N5ABPv!#(3+
zHJ<Prq4~<2DUGzNNqb7oO2!V7EcuP6?53mJr-y;Kykcp5o)|Z~L2>jU<{q7hH0!Zn
zEz=W`I#)$ga(!Zu6@c3CAqZ*$TFhE}(S`)Uz_R`OB{-Kdkr@)I!|s#V=mrz<7zE_K
zKYrde>;{xeq%}92Z_?b-{^|ScDJX_o8$-Je_7*mtwjvdvrHCr0Xv>JGDxYMBxyTBq
z=nZR-Nel{uJ*wFm{g7tGSvgL~;Kn8|^B=10eBiva;eH2LM+yL;OiPd?WI`h9h$Ou4
zKhYqq)I*`-w%gqg-<s>?$Ji(Pp7*DTr+*WvUB$=~WR<+O!tjpJMeh>TXBOuP>@e;}
zlgm;p*w>-1)q@1*$nlT9#?NHrM3!4{^`2D1mY=}X`fPxArynX7+;g6SgmQDA%Gc-0
zNe%$dzX+9)Hv?GP5e86Y@jAqz$@f4}%2;HV6>KT<xtj`nH5HeqVjaZG`Nr)BRv%tr
zJ;U=sWd_oX*PE)d;bSy?)LK41RTebWQ+Dj1gy%EKpQKq!A~xwAp97kX<PiaUy^+~J
z-t|_3!yZf3$NQ=k1-t`Qi40AqdMjk;!b>1=z_EHeT%U0HdA8)GoN!sqCN#N6VJ0}4
zJbbtm`+-bufHH6h<!beDd5OJJp`10Bu>CDWP!}z;#vzQm7R#v?iz~DRtQ<f_-51p=
z!43m|L}^$k>ORY5_;PQOENuSfG+h_pnvt?)y0EFAsI)cHf0-LwkSy7NsJB$O;#zbh
z+czDD$yM|6Zp_?v>TDR|aAOyqA*Ao;+1n^Ul-oP!XG;=KHg+3{Z@|E4w^(|Cz?f^p
zmZ6F7x>qGQ0OUDw)eJTltx=M%M#}dNW-j;T3l3XUuO?l;3L8kBE<+VV^9x&!9hX-q
z;XNv-=2NV|#LN-cvnr{oG#sUC3={@LPZyO<_A4MU*F<RnH5p$B4lOR`w|`SC$;J`s
znd{pd$CS-ZcIzO$npv#~JSo?>3zA<QI^G&>7>};PORd-q@CG$b#{J$fAr}BBN9;GA
z>b`K9*P|nmFaBAKXGEQIY`u50q(sUlDZhvr@G~+=SVMD8(IeL;eccHtTVeLlf&}Sr
ze+vt&HR?o00f;`QXEVl)Q8{YKDV(`xBgwQ`PqI%RXzcjNtTPj$Sd}0Qo_rnaS`pK_
zI%JMM$#Lt}b8Ts_BC!=2uCE-i;t0N-v2z&ulBa+cQ(=`FgSs1RMWlEZCEn7iYI4W^
zht8IP_5YP{@&8NH{D1QL80eY*J2Ur^hPI6sE3DT`4g74%wV<PLcz&TULa!Ce7L68u
zXhZKcrU-rANL9ylBt0(L+4np5qNf{!jOSI|-dKQ@j<&=vVmS2BX^)DgHPOPhzuH!&
zI+QztPzBo5E-<H`xD6dRuug$)`<%$g10^%r9C#1y#!pX_0K+9Lx|FH<Qmj;7O;MrV
zs+Y7WN?DX#1#2L3QN^FA5Zg+2atcCo(6>bx>ACc<H266gC&YB#fW`p{4YC$kvq)tG
zvtSA)SIG*_K2xxczCeMIA26thF8E0zkvo#SbW;UXdCVAo0KQoBaQAOn8lb~nqE&S(
zIx%`sUE+rOxdBAfxQqairm&uHlt(V26yAEOKVky1F=7K-={`|?l@y)J;<-|8BzO}B
zFvS}5FvS=wfk<YIN(K}NG1N36=p^-;;qX||ERm5t1egJG2eLy!y(EW+CSMqJ(JdG?
z3HQbw9q<fWVxqoy<en7-h=-7zvL&3<LhK~y?tr~Cg}wP??yA{5OYtq1`9BOL15NeG
zlyx;bjXwW6V2Xcmt){S+Py>qga!8QFqsZ&c`*d5|5_at2VPIeOOr%XCl~<$00sRrq
z+3FE#a4IvaVCH<`T8oi>X12a8X#bj+#y7bzjd1a=LU+evaZvSB+lZ|Gt!_f60r#Ud
znbGUZR{?mW&LBh8m`*aM;~LkbV=8OGSQ(M2nS|6JBuYoHMN>?r!Nb+In88}8WCE5#
z^H8_HY`JVz5L(J3Y}s`9e0}hSj$9(Jv7PNhNzt|cf;7F~?j<v+c=%I0*b#KM`>Vl9
zkahRqDP>l^=(^EmVQIlg1<|bhFSenJ7xoMfU7ufDVixe=N&?Pi=PY-hT$zS&YxeAs
zb6c+NF<XUfuhxKW1K6gYePD1xZDCzq-|9UWF><hBOGLLaRBJk124WW425-JQu?<di
zL*PxBaH5M<LHXB{T{$~3b!BCORT-7)^KwE;P6BR5@M0`Qi67v3hCd;UG1;a`u46<F
zn&n^`u3n-sZPPJf*iL|{!FAxLkVho@)&;6Ryt&wMWY)9yT0cBq6kmXFb1&%)xzcbE
z8f0w&i+1;3Yr#Ll)l|l_3vh3oLRM}{ns@u&PJ6b#?cWm4-dq=PoPRt&623e?HXPmE
zo>r<qo;=?-o;-XQzGuI@zxM8j&TeC_@m6?y?s{|`Z)k^aIso9D;8S0ja>8IC{ado{
z&Egegf~FA(5JLi7=e4C}+Tj8c#$F*Ede~=$XknQ`dU{D%VRO&gNDi+g7B2J1sg=9>
z+Y0-e{DRB+ar{cjk>OT3ub**}c`v=Od$3$;Hd`>oEMTBJ69=VgSqfj3LcfPWWBbaS
z%q(-mQZik}R<-mk$zx0{j}HZtit5<IXSA|6S3#K@>zh<^%u1wi7^J4!8JhU@#aVOZ
z0D@YhYyo@#Q!TU{;H%eGG(K#TSD_tQNw4*7c2@%!m26laR{&R<EPI`nz<QgC(j;4}
zp=|%>v*RD7CiA-S{i_>ySs33Lo<%4}i4{&9v-vs(B*FIX8)$$a-@jA>??P56ht|o)
zM{-Yg%FM`euZI`wM;#e;xo%M|Yi${S2|>Quzgp@BE&aQ3Vt^k1%7fqY0E8u1t`(%+
znvSX$2!lJ9@_BDEup_|3Hvj%Hu@KBe=rwUZB*E4}{c0&7VUIaq7YbV>Jf}ZTITFrS
zusg16nF)yZSMxMzFN7xK&85Y35-XRS7aorg!Y7Mcp3F_6ZiE>Ydpk-~gegvFTX$>K
z>Kb#YTy6sV#NZ)4R)Fel&3(ey!%jxf*;~_@l^h9Q{Y!je2uo?@lwhbIMm!+YvU5(N
z*R=VAlG|Dej&)H6qVyolIzVC`Sq5+Ymqg&LupSz#H@B(ddRl9x?N|=x&vMk{lZuQ`
zqje*Z^+BTq)SCpJOs2==W<UbI37HquUqM{&_~W4yc%LhSR@%=8k%xW`TR7jJR(~H0
zifMZ9?V;M_{YX6x_)qZLnoB6y+1(ror?9b-clAfwL}iPzrri5uD*7YM*`I6%e0<M3
zN-|{|E0yRqD8N6QeMT(*%D`nHPv?d3X-ebp_`x>?AyZ~kZSm^<6-M5whiLQ>I<3J2
zW-Er-IzS3q!~P~DG_l%06IACiYUG6xx*A_{Vd#FZ)g}N{^P{d``{z>4S})Nz=!Jmf
zXihXdMC{R;9E^qZId1Fv*1K_Y#$<M9X6+cN0Tx1ThJM%<?>ZIK*<z6r7z<8&vB`t&
z$^4k-G8My+Mcom|QLyGFo>&2YurZLjc2wK=!)>(G3p+^d4K#7Jz{Sv(oZwCWux5VR
z!v@Gq+x9o~HYp@rB>QFTDx^O|ap6Q~xu~3(o>5U0Q71UVZ1c!b)2qj&7i}r0`)Y_u
zt>{xq<Sg1pS!~n7KNB^<ksjDw50H<hK)vi!^_47!Fz!2ETG+$MWKwcV7dXMdt*k?9
zI-(jP!5!(ls;Rh=HKM#jc!IxIVx|r1oUK`KHLUM?GG%Hx)dUw*?0t_sRD6vY(n;!+
zLej;~yBmgX+T}$&TDEMBo8criL3y=osF^iHCE+d(OS}a)Ol7=1i~Ce?vJPzQpN~(G
zq5QpSUa&J{%POYUq?p~bwOv#2JLh*1b7*}yR9r%Nk<)=|UaT01WBC9Gmt;f!-S_#`
zkN5Sl(A_cP(|#lKgQ~+nGAmF!FEMxv4G=qxMqGup?t|VUd>}pK^?j7IKMbBW4CSFF
z+m`Hq1SBWJ<nB$B!qe57KUf08C(ZqhK+=H@<8UBLmp!tH9WB)@b$-UJesUj=EG|F?
z>Jg=!F^YIH>t5fPbaj`<2WQ%TytsKnU}9G8;bUehV&+C4G75h1;&``9|3h2D$o8Kk
za18$)_;cu2<i5!U+r3hAi)SiaxXm}a>J6Hz>EBKLOb5=gCg@(r=#bjOTUePwyrFzu
zJ;}0LO`)FuIxi}bnAt6B|L)lC!|vs(;{H2Fux0D=%<$C_dxwk1tV+Z7t%WR3SvkwW
zD-x2N;pWzmC_1%!+*&AU>yqCs14DN|rS83oTIl^(8q@DqS^t18+;!{6<HvK;nl%h&
zMy%K>iM)B$qvOp!HfXsAokYhXwQR=U2x%fD9m2Lh&=8`Wh73Mij8`U(E6re<uGsO`
zK_50B>`Vn|;GQeQg`NzI{o~%<rT<2FWs9{DNonJTDju`1;%-{$IjwP+JelIhRor~+
zou-Rp{0-ydGyU*tb#}jzcEDW~cP2zPJ5ioJDiFi;?%j;~U`;%ssUdAGY#sE@R8>}k
zskxSIzXGXYmPh=Y8FzA5T9PtnRA@x8u(GmXgT%gELhimI4vEs4_MK_a@P~;*N<r<9
z$r&Trdxm#OG~V~i*9LF*H>7{?SD*G#AqXb8w|kZkCN(lOc&2yZj*6N|QWhWZy*)Pu
z?W&j>XmW7Pg7|{@0%YCUbAV=&ZMNkGG$>)^O6fVUwBj*=!eo_OH=fnM<8m1UBMQq~
z+T!@(ds)mJ(K}i1^_@1#sP`Gica|(3(kje|=#E<|DRQ-U%>{}nm}vZxIZJB=9Wmf^
za!7t0Ds`tOI|3_nLBw_QdEi{|zBam~vhd!vA_NXrgfb}t)}xtjFF}Z`+(iYhMQ}kW
zEI;^FZUQssW$Ul+CuW!rnE)Q<9L6*(IsQNL?RCNM<Md_V_lB#=e3>34(=hRvsq{m7
zT9R(rkwaJO8uPgjys|L7^3@p>t0$69{HXLIy15ZNJWL#aWMB2d5m{=vpmM=MEu$RW
zd6P{gwFG!2f&!aa@;I_%h^Ha?$FmV*xX@;}s=TVeaPfx#(_HNwCXEhvO$YA${*L0x
zmL`Jhfo1;&vkhUh8cNnok8~p$#dIw(SPz9x;}@d?CFVH|i2YxTy<?MQQMav|cBRcq
zo0Yb0+qP}nw!PA}ZC2X0D}D355hu=v{YIR<e!`4cWA@p48_$CS1%QHK-S)0nQAy0l
zg`7$51h?WXWiZ1oy~K5VG=dwmf(}tcMSIkj{FE2T2r~WEXsv5|5Fa92;&D*uAYzV_
zCa0*5)3NB~6ve?4?vpha!`f4yVT+GK`VeA@&4JVYI3HB-y9A3@F%QG}<in!khjrty
zm0d&fIwd55Cs+uzymermu9PZ5RD>ZBG5uM2IqYM7R`Qawq_Ei^;O60Gib@KL>ffOt
zhHD<xlqgE!S}X)diA90E)(@s3;*oMzxj&&d4(DXI!>;NWlZ<d>;08W{c3UIP9rWT?
zW>TB1y~Vt?tUD#KMG5^o{0`jqGaJfu_uM~n5&SrhkJqJFJbx!tSq|k-s#5+cN^dw8
z0;D%v?cL5dj}Mp6Pb&IL!9D^h^c#!gV4?6xYMm2Y0$I{3Mnh7(ARvjw3cO#1@zhA?
zU_P9Ll1?Vc-?vy?B3H4X4D%^?n(X%;&YEN8H_>}kDuism(B`pAvO<krH;6@CVp~ls
z_P>0Hh7uCWJ<2<0-4e5>3d-82V}D*sZ+0St?DTQwi^8Kxa8A9M$mk-EMw9cY=|t_-
zY;<%;EOjYU884GRXl!C~+_MH7g{97@TfYM?ot=Y;ZgBF;px*FMp+C$Y7L>fbxkT1K
z`}A2Es!Rpe<t=O=8d0AvMh`5@>Vg;Y6uho%f?L?xUvQ?b0_}HeDkSAJ@`I#YB>ZF9
z)0wXJ5T<de?NMpGehqYHryW$o2kGzw$V0jaS=+9$(ur%T_~VhQgm5jyvTWT7%H&09
z#Q&(>qMATEPV1|^Hu2UjcbVx0?O1_w69K^8rv2NC!)?|xtkX~qv-isu>x2ZD-S@iB
z%`Wb1_P3gXfREm`9c+)UYW4u*LPq77-myJKzr-Ehca0DOOB%Pugi7?dAalbu!yCjv
zas<Y4uCsD{*0gx=^3vq6nc>fUH}zO^oTf(sqiM2;#n)0Nu96gUR<WO{?G~N%Ii1e7
zGs{c)B`O$+UC9qC+A>c&U8Htenys^wsccY8sJ`xlBvy2P@P8fdcKE6b&MuaOVLIvQ
zVtX6^1n9Uz<T9<J6V&&UVJ)jGG=H-p<%wwen;UI<O)@bT_J=wWCcDE6?&131{8^Gw
z@yE6k`@3ZrC<TpKjTJ(!8?x)eI8d%ekSxmJDmxarz~*^}P+DH33K66XYp5e01{M^#
z$s=B&Ssl0%yZ8*PK-9#5+`7ZJ6h6k0I<u~%hps*KQ0s4YzyVEc1vE<^-vv+mRC<-0
zpWfmsomsc-1KHF?YN6^iT%3}W>oVR;a(mveUWGj~H5X{sW1~QSx;=z(T6TzkQSDhh
zJ;E%U4lbNtNPtpwJzz)})mY;=jj&#r#f(}JW<BLQKUZubm!X>tzG}A#9fdz+wep)Y
z&g0+yrYduKn8ty5*f!xkiG5G8aeOzan<_4pR$D<iLXBp>zf(aoJ@$zfBNN-E*Y_?$
z?H?^4af#mHNy8z*w2gUdQ!}k#yf_8sfFS1Y{PNFsg$>X)`}Xd7k2c%GG2j7bB7LQ4
zAcMi+81BhY#7@BzJ6R70xMU%FtQP;R9yJBmAfChT&=r<KQ)a8H<GFP^4oxtysX;ne
z$JYnMNd$~|9g>9As+Gf~2wmVixBaG&xhhqfOD5NuMmIFJoh<Y$cg<UL<d$Up46e_u
z3_UNW4gs{6{$wcr7nVRgiiwcj5Y|dQf0{QZUUm%I&kbw`KX|PU_K0W^<pLRfIVkmV
zlyXUrtem2%^i5^i9-1m3x=&Ix3++#94^lKa3M5LT3=@*xv5gI3jw53nr#qh3WNdj#
z)(5fICrpb_<7CjQELCfg77rQI^9^@wW&=U6q?h^ss(xCWA*ek*inuuXtnAF{E(n3$
zyDmSp?xk)-Ehz902uxtTCkDUfEAU}JyY2l}dvxhuO7x?3Mr$ssBoifW!BzIT($jSg
zy?OrSIj)_G?4fA|eY+nD{*nfw_@R6b5~tos<E*v%a**Z<M(Ucp#IiAX5vZntf*rh2
z(B0+?-zd4@_<4wNrWDj|cn{kTO?;)5sMm<iPd{PGK?JHzoq>myk)+WY2bnMN?K?iX
zlbi?Qk;3Mu<O#D5YYwt5x=nmT57K#1rtom&9~<Rq<awYR*~huda9)x9oAoQ<z#GFG
zUd>dgyDQ77szv5Ki@$vNDfMM~e0nR3In`=<lv(sb6pR4EcUxXSruop<x7`t(l%Q<F
z5!__H!D*s$Y-4_4H;2w6Q!k*wg>SZ`gLZtUfe!=-+aFflT2QCG_V@yA;KD`??kta`
z@o>-VT8&~;i)SjG`)bpKO>-A4HErAF;x=mcj^5tSEa|@uivNx9%E9)(6N3(G{$E0N
zrs;p~O3a71;{Rqt@P>7SivG(H9VG2Ic5{+IL>igsS?`q%ywpI^P^tc2B<fn5bJ=5U
zo5!l=Wy49%oust2C?)1jcqfG%l*%PWQeAp|26#Wxlr(YJ+GkR<jG5fCi{rX!$>cUg
zJQ&a_+*-kkv|<pe@&&YfV#jEl!@CW9lEoVN{HNc>mJzt~`S$oAYHCmd#rlP1ceOaU
zc>HdPsAJL3>O7^1Nzif9D&uVs9^wLA3(z;|?Df0RWsRd<3&<GC{0cL9d9hlJ_MTpc
zMP5%@VUlaY@Tj|~{v|l7*8M!yoeXD_!9#bst)CW2*4ef0rCB_tYK~`$DiBg22U<DB
zGyHFgiuWZsB!Vd|uPJ35s)flOT2!4SNozbJ+N32>Gv2jvjrF07CTrYrfr5Oa(L@R2
zImi*)R(P@@<9A@o)ujLg*ZK4=ogGSw0D`O$nrb5kc5H&0lzUGXmtNI2j3O<g_yYDf
zx01(Rcdi1<O(}bNV}biOa3p;4G1rMn(y+48vE&YY)(E@7ZyS@$Q)9<(UKF&^JVl+l
zd272xv^T;BS8QpDZ<6h#gj(2lTh}3<%oOT%MF=mbMwR!6S1VGGSCOx-ncm<7XBE_R
z{8ZXf0K*KzpDJb~A$xHXH<OPK)BschO<t?<s#7~oug9XT70C&ScRc)(zfhS8e24NY
zk$s+}rJdOTW!0prU~WAcS`CUn2Ltc!v=bo>v6XCWJD;yzMk*w^IOe9sPMa7i3KVQ?
zhXyap)xpe`EtL|iCFfM3hX&~`Z0xU;F$42-z!9sGF5`3s1?nGWdGvXsRC<~>L{xkS
z<zM$n&zx8lD3;qekRgYqkl(F7aF~3O1F!8j7?OVraI(iVjQ7HNf}*hvbrd{Kxh57|
z5Lc3_Ks7KVwV#MjSs1fxGzz$kS~X$fkZ18LHnTW0hGbon4W!SDRQU#G3F%M?_D0hV
zMg`cp&cY0P{fK&xL~%lw@CPZ7w)bs8aa+?&&!$)V-%rv@n7^nzzx_2NpAYx9SUshG
zc0Yn@3h~hJn`d3{Y|o`_wf`J4{q#8JbSjxE-Oif0%e?SY$pX|`b-`T3Lk+Blov484
z!Bz|T2~c>UIn%y{_tNT19Z5(EO{qSnc?|^}9dHar-#E46MIA;|4TgYE3Wbg2p_p4K
z6ozxhaDgMHd4LB}hO!q67*B01oTrnr57He`ybCT4JU&3ntB|tGwI{pbG~JEbbXb-W
zNNs9zg;9kPkc~G<AUw(zdOq7$apl&Wq(g;3kUh}c!X_^gZ%m8A8lqS}b4wjhQDn*y
z(&c0>SF#*|#|RXX?<txmnm|lZbguS%A(yHG{84|EP~)H9+N27tlsx+wltRDa0#!l(
z5$;)hq%>p1?qUyyb=9C9CU8PuwC8X{U!3`>H)H(ON^HhBeF;m}+n50LhQ+3|0y!hw
zWEwE?Q#ktqMfFl&RMZam17T3=!qsUvwe&tiI_Fw|J3Hlis4ckhTYnu~V&?b;S20ZQ
z&=2S@ICHe-bfA(hZi_{m3tS*9S#s$9H{?dsGZ4DvBPXSh__kVKa`nHI{B9q5ZlyfX
z*(OJS`Zr_u7E;n_>%tE1ng!#KG;c690MrOM?6U?C+%zcvH~5_zx3tq6&+Vjr>_q2f
zR7t*lY?*p6Q)RtJJC5Z(V_S(c+)qEgjRnJXC<MhrEm-Fa7V9(LD&&*IT?F{3FGDU+
zO_C|wWh{?;t!LR}22jZ)x0Bf07<YkCQkFRXDaaet;`phsnUo!kTP}V?<24~1B1Fbh
z!GMO#Myj0?W0owx-iQFqEqKE3%dQ=nmodp`P)WkWPStmPdwf57$}E-Q;t=e##f0W9
z{U_9X>;yMN5U)oiq%xK@)szn6?RU_*&*CQAuR6SaxO!%#I1(ge|F>;{GQEys^;1pA
z*{xn3+|qC(Kr<bgjSy-fflbYqr_bxzD;u)<t(<lMd$1T%dG}MCc<nm-{Z{LT^`2L>
zCU`2vRA~@Fe#Ss>gFUAEHLE^JbSO$}(awH3I=slwjf=t^oDwmJW%&d`1)8FeomS5}
zah3*gFYKZJfG2eS1D>e<0Z&xsP5yu<iZ;5OvqlB-1P=7zM9HrbIs21|LIKK}q5G}i
zuA!c1jQye{UCiujrud+lFL$Th`>%-FAGNR{%XN!PidQXg0m=pPp%dMBdTTz7^ST5-
z*ftR^fBuprF-292+LgPTlh&?Fw5T5Cm)xmTAV;X=7UOLBWb+c{UPsl0vb41$sAJQz
z_UsJ%*tUY8DFSED8B&bpyJf@>s5l)`!}W95&5MJIb3${VTqkWF+Te;MkVUD5AWc<d
z^EF^5k-8#HLu4uj!qV0yCVZLf#M%p%&C>bY$Z>XV<7TC^i~BPz&CwpJ8twx3Reb8^
zod7bjDebb)W&X*Vw;>;`{(|h93v{xwRX{=fo@nBV-evF{1^-Mt-V`*Rz9rjvZP!;Z
zF&`#GeugQo*|TJa>QltoBbV7-{-s?=EDlZ)kqgzRtLsW5s(jG{9oP~ko)7=$nvB^|
zWh>s~cEM1*nHRCKi*WXXfN(<BuKv+()!Co=ly4w&-j)A`L||t8pGX8I4wjz~^FKd&
zQ44El6GwVcYXfH!VG|=eV-sFp7$;{(69XF<_w{D2>DV7JRnG<G8+a&L)jwFc;OIjv
z(Dt=-t#~2*p_zW&OIvu4Pt<XuEMqq*S`90HVwNl#RiA1;bPEojyu-&cdrv-ZcOjmi
z{>hJ?)h`RRJN{8k!m`QKQ4%`J(({|+^FTwg?vZ0&Oe#X0K&+R?C+h%(N{%PF0HIsu
z{NG_yq)|D7Abu2UoWMD_*_-O)C-R2H|6g^dmnOvhe*Jn0GB%xbJ3@@Pm6iJmjF>Y9
zlbE;45$tpU6J;@T+rNrR$aq=Cjo|!#`;tdyiF2Uph9xAvxqB0VB$Se^MMw04C`2tu
zSjvcF3JT_Vprd)Mnc4CTJKVXgO47bPg}-IAH>EU~;;)u~?<-UwRW)HJ*d{aIA-#IG
zQ_xD%_Emt}W{PzVC?o{V{jg#sM$&RmRL`T|6i78J;F1}Kvb>9QStV`aweaHP|3pIP
zw0<z2NTzW}5u`g*|24}HyY#`E^;wwe(Mqs+MAtMG5Jvz9gbQIOm=b$()>26MS)QK1
zwM-#as#0NiO$y`|N?9oH&zspy{!ux!@qrnn8s~fqnP-nfov4<Gy7(KbruN<w{m1rP
z^vG*m>6t=wr1iAXIQTCqpiCm9k-gBb$j9me%wrHdC0wFh9PQ9YWEo93aY*GIy`1Ei
z?kYBMNZ`mwVH~X$e#KeSF^t@0bK(;zeV2{Q&&4TrGMWR+o<ml_j;8|HxF}5uyPK!B
zt<$N&h^vdpvSv+v2I7!Y3zt+*uVg9C0&b`x`(ohtb|8r_d~|J^u9j+E5DypYUd?g*
z!-0$9RAaTde{5UsKs)Xt8md|qUA}i&+u8#_7I>iDn!Dn}a#)bwF%Nio)zV1XBYFy9
zirS^V6UVPkR+o5MNu&1~^e3%%F|3zqJTLJ!2$2>aR=JcTWGWkf=<m|Jbz&9mEDkGw
z0YjxfV4%5Q{~<R1s7wWdF&Z=uL;<dxCZ0^|a=Ig;0vOu?<GjHNs1Iz&Gq<3|z{+_V
zIILQm=gYGUPCujm!9r0dkU4TlY6TiV!X&mc)<WNkg%k2LfQvm~oa)GvKM1$gb%M{B
zBfbW7Sv#d#+eT>L2OI{mm_OG`XmV_%g<|jheRtRy*>95zBOF~(y4v}z;q+_eY+s~^
z*9x?CqbQaW8+`$-D)PZZE22f++j)g6mYYedhMWG9ZvD>0`j^Su>9paI(pPSw&2VK@
z2W(DLlM+Kgjm)^z4C#LI2aUro7DcRQFWlBlT&TEn4_VL-eHYtJVxifjt~(HgOzYg*
zQYeLb)R3lt*L|2H3*c|D&{yxDE7%}QfwmG>`7+Jy&MMH>SraQSG1_k(+U6F%L$vLY
zQ9FrtuALl|Svrm<+-%5XGBv+x7`Mc!B9%y394A({k6p|DDxJc0kkoy35IYZbYspJV
zK<Ibcj9)m~vJ#uh4xboLrZvN}&{6|z{^nHx?zNpa5w5LSaYjv*wDl#q{_@m=CW1tl
zp{l~j5X2qUev??I0rXntN7_1gqyP2A29ru6Z4p+EWNu|M^>N^Qux592YDaA42d5fL
z-KO4ZtNw_t8hEcda-0k|_~6vh+^8+H&E<&FBhg#dkpJnNL5=ap?$DXS9jFq%lg(zO
zA`ff`3U9mr?GLE6Fs!#&AY30oqG?Ja!!NEA5QMOKnkL<Kk7!~WIOX7=7(+|GgkciU
zrSi29>VbaZ;q0;7O6;kn*(xm$VFZkI&yKt>MoF-jRcagWuBJ~%UoAGud@}LZyBI$o
zO`~@Q+B44=*`^I)`xJaoP(k|mVCU-YAo$^mn(*V9+XXGI=7s$eq$aM0O>!DYko)(r
zv)DR#l5j3>o@|rC3<zslN!NX%G3m}76VqDlHEm+VMaV!K7X&hdN+!tUYVf(|cFg=H
zc{tbXVGj7KRiK|uWAn`W9HQPJj!kRgXrE!edS9Zsf?Ss$0?s$n&?BQB^#vOVq)5>S
zG~UGo^zs0cuLOieF}soK^HD%d-IoxV!a_x#y5SXI?sm+`M$1_8?@(nF3%`3rmz11k
zszK4PJv0X<x5bBDaeW{r8YeBG5qoun!YU;9Azg8R!IMrk+C$Q+V}s21;au70g2VIn
z94HV<*e4x43L#f&tg~suya1hsZ{S}vh9-R~WENiS@f*N*W3*BXDI1$|W$7P=JLwZx
z9q|ztDt-x4zC&@-F%A2De!=u?*Z4kQ%~QWK^`P)P$8qcC1PnpDlN2m0>jbF%Iq@B*
zCg|iVXN~!Uyp;2MU1=37>=)x~_8BkLJH(`I$Ebd7hMoW><+AHU%#u-zNEwVRpgtSc
zb&S%QlYEbvm5_mWC<W!Z)2La~;&F7+EJ3=Nufh{`{ZM55bpYh55c-6O+KjcN2KV0@
ztIJ=XX0;$yc2H{Sx-$4^Md+rv5+JhGedN#vzL-rdgyRR){612vXZH&aNj^_ASM`C3
z5*V9x&`rhS5_KTS7`Pb7y=#w}5V%)e4)vo3IzB+M%byh6GS8!bEwj0?a^nhXF<4ww
zws|;Ek-ygS_2@Iv&ut8rs?kONoR~H)!K^@G9akDEeeUUl-lnNYqs1@0C4}kKqz5`s
z(qoxu0n(pUB1V;^(Wc{UrD}Sc3?q|I1l0{aUop}5(N{OU?n(%J>hIOHhh@TwQ|I6g
zXtGq6YA*d`f-aDr-48hle8NU5y`2OK|HX0kaD=xt)?|H*;5}_xRYi$!KW+;>`##dD
zWI2xU%8!LzHuOUWcCEc6D}alX-nGQKNj?Xzi{oTXqJ+p|QnpejQqnRHw)!X)W?E#}
zDif5Qdnijbd7GmEVm#c0<C*^GrR+I$BwZ^Gs3yc<x9=LU#QG!aU+n$Ev}o?sj&^R+
zASA^u=iX(&1`19vE%u%^G=B)TukCLNBG3W<o1DZh20eH<zSMNEVXJ}nwuw*JbPX$(
zf}=U}U{mkmtP5kSGyPL2)+oVu^*0W>Et*1DH?xqGojXUGYnx``Rm)bQlkaB#mtrGk
z{5y9YBWuR~S>zf0P&vMDuNeMfrP+FScmJ*(M!Y)f*-{8S*PLUAey~NPuud_qN%3;!
zp@eOQmwwcaG*HLh<%kGWoIF}1XZxuYeg9@iaqD4wrtm>9D@)^1vzj=#g^~S%ft4|r
zkfVFx7zM%`zvj1XdUT_)p@EZ*j~e{i4Ac1SfTM?EOXV7;v=v^>w)7bG-(<|w?wCz-
zE{Uit`WCbxnb?tbt(pdK+?-UMR@B;H)fXGz4xHrWiW@1s4Q4x8ocCfo^z_#Bqg30$
zlp3cFG{^yg>n9(F;*w<YWf?I|Kj}jlKACZXM5WTtpo}lJVN*1ox@O~!rRo%zc>x06
zkFU>-ot`f#Xzbe^c0t)DHMYoST)tNY*pU3`$Dlk~16sUks=#PE6|VqZRpZIN^;;Lt
zIblI!{>3XXYmvP!W|<vl_!A3Q2Dv$AG<@XY^@)d1a~x_lm&hS!6JlNwDsEQZg?~&`
zyd!e&+8qn!-_l`ri+fJPe|~>PJ?WEQVP`Ag;Z>q9EPxreVOLLYDnUxYwy*z<(BK@&
zVr<%HtPa^-u)I}PSy)&Q*UAk4`4GsY(i3_QeGJOSimOf*jTb8X2;r7r5+9B?RX^F$
z^SiIRVLnxQ$QkFRjK&$_=YxZhu#{?A2)Gaxi0w?<E>Zp*Xols{o6y!CpZ$}DTF6ot
zKv93|VW}=%$UEKOyA6v#o9NUAn#)*XqYv!3-t>#Q&RPXo1LRugJo#k%q+Tv?<)&F}
zX6oFf=-SrHR-;x*FS-+ZvHzyefHqEp4iFvYs^$oE6VfbIDbw9%z1HvoP>;@VdFK<z
z1nJv-px0}63KU!nOnaxlwKi?X1pQ}NYT(sWy4%PEp!in}(CGKedeg-X&CO(Y9#@oF
zN>rpMan&8fzR`_G{D#!^#HRl*(H`@E!>=>^FLkrs*b=tHY`52HkC0P@8pFRZq3VpO
z`E9ZOp&z0eA2WwdeIroPiMTdmv}-xS(`s1%C)P9M_`p28-B^76Jex$bw;#Ycdc6Je
zk$)6iC`)WIpWaPIMqj%9a{BjBmAt;-R0>%FmOnbbliM?WhLv1~H!&w<m%_=zrF;i>
z1)MV*IBn+ddIsx5{pK5S;mj*eO&pvCviHMxf4`AyK%^F_yPN&D{O#%G_x4heSR;Um
z1L3U7eiGX2$2*Ks9$iNVrx`f68>aUSx0CmE+OKMyICF?V9Hz$?YI~%$ryD=SV*y}9
zxwfE=z6)v=;P*~x535zg-bdFxE7&03Qmnd8jVAnDMtp!FtWl3f11cC5td@ySaJM-P
zn@54utI?%hxKv4Wmvah7yeFQ;B6{nJb=W%I89C}bdU~`rN%k6$<C;6dyH7kV|A&i+
z&!v6<j`)B@a5r4;I*|>V8*i6l%RKyt*J0d!Q6lSxmZ2e*4_S<ynmE?75PbfT*@9zF
zf6E2{1!3Ab*-b2yKrMkTU37wTS^CQu5z{+Prb1+$Kfqpew#_OMLN^k<V_-{5ktfca
z?a22SbuACPpw|~@tVH{*C9)P9o%$DNz;5M477`hml8^XR5tlqzMbJM9!6k9oaHz97
z0F;6eBCu)eFv1-AE8m~Khb5pO)Lr{JeQ-{{J|K(&3*u;)s5!Ecc6b58*Bh98YtfZ~
z47CdXXQ)CV;=810Re>%#7(sAU)puXF=)llr*;)F->^jL`Wab7)|CWfS^4)!fcsD$=
zZ>bnNqbLuH*Tn{yHC|hg&?HHUsA0;il;;>?;VvVTywk-9ToLF+@CpxSx+?2>(x$`j
zHjAe|){!#phE0NW-OpTyO)j9i0tHkoWK>(+#}+aTuiV?p-G;YZ$Wijn0$Rmy47P60
z61HMq_@|!d1xVKL!^Gpcr+)X~hmOVegqf%6&>!_8K1M&#Ubt`Fe6Uo&R*bOjY}gh*
z`~d{C2CfRM^gJfq%BqsNch~)x`ONL-gT~R|^O4K9Kp<TY=nbr)do@Pw#;rovK8gG9
z+U1b(h3zdNyh!1^=$mrq33x=Fj#_yxryzk<)K9bU7l^g;tGDG-{<&EQ@V%~j`u)tK
zJ;J3<943=DiJc|C@mF^oc~T40utLt$#<rvt-c&z|y=Ny}!fMbT!a8~<A<ag*f+ECO
za*36QOg;4)fB7VNNSn|wXdPE((1r}S>&(#>fvm#h52a3iCM$7Z?*i?aO1nn`O~;Pf
zWx9xXamdnHRSg+49!!y&VbeS(hC2$bQ}ko+5nmCK<V`+Kx>la=dT6m35xK{m`=jcf
z0aKbvlwfr2HQq1OaFq*<jqQmC@iaun2QEVG5iK0Cf|t<lT4XeU*%`?qQ}9fa&~{sB
zOmrO#+MrOe8lyGsX$D<H8WZZ6q$C2C(UC1llSh0~R9gCZ`*nG<HW&W6Gp#})Q`J|s
zni7?iezbTR=`anH=he@qgttfCN*~4?L<Q5cof<v>M_)kswU3BqI?gw3N??6(Li~l9
z?(Sx;Q(jgeT7Ow+f`YMYTIjf0V%?x)?W5c?Xr6it|BcSxF1>9DnrsIz+J<fbBpa@$
z)kMMLuks`)!{)$~47p{l!d4EECh(;Gr)2=vrfZdvv2Yahy?*UHba2k!n6t5S>sr=%
z+vo_M4l2IQG0ZGSbFxBT%KH$F4nIPu>x7l3sF4`{!DCVAV87}34^_`{ms_%!ev#Be
zj|o+pLm4u=^`=ed6UU;-+p|1R{aq0WG5BiOS!m(!!iu3w21nWOO(`moR)@vd<?fE6
zoH~8a%jH)dUh}}{HC$c>OH29?>!beMUC_$rX*j3NqX5`%y6KHfud|6wtpr6a((3uR
z$Ydu&xolb4GaW)#5zuJ#iKii7RcGaZI`Pw@BEne_-GB4*3QDqPnn6p^3x~#XtnY)_
zseaNJp4+Kln<bWZ`x25*g(~@-GLm~T=I>lb+_C2j6{|61HnvtS<b5i&BhxCjII^nf
znR0j}7FKHmP7u48Lm#GbI+A(v&y=Z?P4Q+uk8B2XQiY%&ipBt&K<J?Z?Oa36UB}0e
zO=HrYBZyOXld7wU#0<mvtW6GHbXwO!5Do~xMmAp(N@Z5Ea+`fOt<Op4&D-B4B=~Ds
zF`!;+Y)W;|sM~RCOAnU#Guh6Lc|8N6EfaVyt|-}shdCCk?iO=YEg~3-av|sOuZPVt
z(Z7^yIS&XaJ=C$<0BucyNIwn}Gz9SoBgpS`;!Sua@yZiZw%4I1Nv(Wich^=lXdem{
z7*vY+C&Z|!9_{himP>zWnONho1Hq~3@&qq#R5e<dxxZm5a08>M58t=nIFf<<t!A>$
zRuD@<AHjVZfofqEHWTv4xLenMPo-y%`jUxfX*t8|TY0i7uKZzQkvrA^*Xay*-bG>V
zqrnw$FJ9rMPHj6*JudcA-@MPeabp26=4G3ZhYZi)>`HIMpk?}Jcr!IEJofv3_8EVS
z>3t)Gj89#WzTf(w4Rv`A)G}ETP?t8`E&Wy>GaTuR20(KpA#YI@hW3?8!)@D=&yGu{
zZEW47Yoo~119ucJ49)S<=m64k+K$B72=XeX!Ac}#rGl$JE&2E(vj#7o4?lx$YjlV=
zE*fzwfo3QAK`jVlFf4cajMkWIt8W>LKEHU$c*6fSIs9S!9{@E*rXREY|0WNdRv(Yu
z{6E^<q-v_4rC^^m-tSgSCs;=q!#wZs@9%iAh+-1SE!yd^AlxPeV?bQ9yxR`f)~Bnd
zH;+#bqMI%_<F;$<4rdSnwp$V7Y4O7dg;=TA=ltiZh}DHX&%Plb!rr3`lKbc*Lb?x*
zxb1c|{egUz4AB3>M9_2`#2%n|4=(F}AdE5cS&}7(LC6?5c&%OCk>iA|QgJ?G;aRUu
zTV3oeqv{|frE^8jCOYZp^~+vBU@gp!(;+!xcjfy&>&f{tuTa&+mI}=}y*_!V!FtV%
z!C;9l%R$M&n&l)#<~%7Ze6(V~=d+?^K-aSu@CZ#@nM(ffhBi$p<j07l6&WW}%s!-Q
z)!kf!8UiHaR`vCi7(YqJa&DlQwhvpkDeNi+ybmE{lEf^q2?+&w4PT{B4~W3x#b~5V
zX#(I(SsE+EcN>hyw|#-MUS$_ZBNp9i!F!&JR`O>!$S+$Kku`ic{2V*Jxfq2E6`{Ff
z4zO>INwfBOE3XA3e>T}%zgZG*c0_ATjA3Fy8bhUM(EmcLfMjx{816@W<|g@+wzN|o
zv23WzO!&)DyC96oAe-$@ubxad&qZb%5}eaa_<Lqrhe<tu9=HGy=DXe|Bau<6jtle~
z1OnyRwmoGV6KdJKHE+!@`(4f#P%Fibn&8F>d7=Wj{?h;LZ<gKQ`Z#3pMW$^4o&Gm&
z>T6YTXc<VWA!%xVZjhW<aHq$Vj_fZeQA-K*Bp@Lt_MU)j=T<7wA|RqsED==3{!Qqi
z7%p!;M-OjFo7M{6oH_xV#h%`?q$$6R+aqm!WUXFePkcg3Ge7x7Bct1oP0KcT;gK0*
zc!AMyhKr9-r*{|KOl9d?<yWM7=pk_Q`guQ`FPIO0#B&BM!rrL@)7#fO0MlM<U~0V8
zjf~nzBfOY2<>nt1<)j}z#>4_NYC|@~A}D1b1@2a5i*@EDGVjXJ8QK5IR~cl@atRhf
zcWLWVM_hbNJIWhq!=2~|zj7JKbPNiEjV}C$kwA5xf$Op+uF5=JSc9{n5V$O#QXdl$
zXsOu|iFEz<^m^VrtXkFYNN8~*1)Ua?2<w_!*JGS+$j{{<SP<cuU>`iIb5^83M%Or7
z+_}#Wc~Wv~FJZ|qZkdVuVbZBvwj5q2HJK$80rHnR;8D{dNa@enNp&|$)oO+n8^DnO
zweI2sEyTp*NFHs=VX3+*d*pMA@Z1Oq4v@cE>+9jztoB;#OjZ2PvL+kc?^MsOB!u=)
zzOe>*0Cj+x>lSNt_b5V}Zs;Iu*+~q|ASeDVlWIu4MKV?yTQ;@9i<o6a=7}>PkG8!W
zZa}DUzed%LU2S5#J>_a*T(PDla}CmSUq_-lU#6^P&o|UtO3o6GRkB(n`e;3lSNgHY
zAxM(v>#Dg%tM$3eLF=y&`t2pg2isD+k9(H;X^UHET`Qw;NjXwGT4l>IxAjFoe{-la
zhsB+J^67oAgw+k%#mNZOychF&onTyL+>^!86dKSlnqx~T|6rQ2C;nGP1<e%O)_}Rb
zxF%?|SY1{rs{SgDCFPM%8$Y4fA(ZCyuA1QmnM<^oSfpx6fuFW8uUgfOR4l7zNdo^5
zl83-dO`H4hci=YlU(`!`a4PUU!$;BAl)9_=%obd>%s`tgR@3YBG^-2FK~5_7MM?QT
zUw0*?%}B0{T$0nz5t;{lwQJs%|Lrhk{-0G`#{YAc{=b$hDslDj0}OD&FTNpu{!6iV
zT&eP;<rK(HtV1v;-@W^(%7J@^XWpJYzq^3;IQ8>RSfiyX__HH3DVxaUtr@m2bGtah
z6^Lr)w(a?dnB+8!B7dhMH=OkuUW}B9b-Lu2YKvSVh$mN~h_4Klb1|gROEWt|=Ww!z
zS%wd6#4fS$xQ%)l${5xfF`is6s+xYEw$=HUXpY_6k;fL6|4i*nHz^cU@<0|}LE7uL
zz)S4{ME~sp{E#>Q2ac2R=c)W}-X;JIDLdSs?b2-RX^}JH^b4p$u(;8rxGlW*ge`n0
z$R$`^5I6kztXx7h6?&(aCz(B1v}fj^A&<sKWqDd{lrC>BGp`<R-gb9r7iO8DG$mbL
zFAaq6{O!(Zu_K(+k}Ah9J|8__&!u%yz$dJC*GhjsdH)OC?Cs;OcPJLw((Ga$-#0Ku
z+=S`P;`Tlv$x+hDuz|uiOx{UPIfBdH%v9bUN(40AxxN9|8XHwQl5|O_hRxC<1^1tt
zkc+TzL(X1*{(h29@)CNHVMFza7a!6&T`tJ3NDtxbgdJ?PS67u<)`v=6pKhUUchxg1
zmQ~cKDz=Vb*{P~}GT8A+CXK!GKZ;8y?->o#Mmd9j&VP0l?DA8_^^AEYmb~*Y$&McL
z6Cvru+(z@RujF9YU-W(-?nJ!-9x99&pQ5Z!U!q>D9`=)naX|G2F1im4`ErjnjvK4j
zQwpsS4cW?$UGtU0Ce`QX2uX7d)03sc?!HX*u|2=1FT$%v{O0CbWSA!u*3~*lC8uRN
zXAiXso$X^Vi8aK5^X!i`rfsKtvehH+SI)PJkAzfNHJ>H>5X24pE^g--E>}eFG;EFY
zN+Ld{$=>+LD6H(H<msqx*k`JKhw#X8IjLtW(=8y-3Q}Jn$kA~}c1I=*EGdNJ%_|Rz
ziH?Zlcz8HsuU3sPcwz=Rf}9-meH6FPQGBjOdx37#n2>*SZ_dHhmeiMZH`a@8V3|d>
zQ?o`=$mG8K^25Puby_4kxHOrNP_yuOz{f+ds4hpgd0e<-wK6Hx(jJfh#W9n$w2UwO
zOG`fv+fmgYj;i8j`#OPyP9f#cF`h)0E5%2HJ3m0(GqJ;~gKvKy&e)QDgPiWWoiEZ1
zmU6D7sqIo3444vaq{pS|l^s}HL8jXNMOxe7Oh*9JAMwxP$Fn*cj`Db8zFY71WYbLI
z5FGkL<D0{36?}^mr3u-q#)iF!cO8B3xNymhU4x=MRY;1t?asG=hFQP>70%qZAJ57N
z=8<%Ey86P<jh0H4@u2Ws8MUv!e}qNhzv+5q_!%ocr9&fpUW%k?VF!^1pv+o|%g-R0
zv+I|a)<Y*V!n}u?<3_Hy>ag;|E$X^8tuYT1A6?qR+CH7M{yVGdcrFAEk2%J7w1lV8
zZ0}zhjYp6A5A4nt9u8Anz!1z<+q}3BvWQHcM%a97=!g2=hb<>Ee&nc>GmJwtI#I5g
zh_be2dY>iS3l&86aaR}1jZL$N=)bW87Bk!qIYUj?e4mT(lHzBJLZtpS-ujYwiNB9?
zT2xaLUdE`9RqI?}eH1lFz~q|!ldFt(w5Y<xnV*d<cdy1PYgAYWet7O<3r@J7*MouN
zE~g(%g}A@7)~vOW3bvkXu}5O@zzg7?aoZ*;mP1}}Dqq4fMF5iMW&yCEY517P?Fix;
z%!39rUw1Io(w6V(f!W^~Q}!xdBfc;=TXZpbcqFpkeXOeVXB=8G8IFj9y4wl^eFY{2
zU5Ezi55gl4C&jtBdr<Bui{^PO!Ns4x6iu5V%$X;9(t_J34}Y2BhXqehSfJP-Y_BLT
z0szy#`^Q0GPYUQE4Z=H=dtJdj8|69=Fo#<hD`ats1r`C=r}cWg-rJaC!t5n6I@>*Y
zU-kS~pAHt_|Df-Efc5WhN`6rrf$A&%Qj)2R@!#LAeTI^~`TBaAyxI0;gc<u%DLV(3
z9nzQ7?RilH6-WPz)#_aYEDlTv#L7_FNB6{loB|78ig<<Tp(Tso7v@0tQ;>7Pgjs|u
zsVH_sLn>xXX&YliE~%Lq9D2;cC0T<=E5(UY+&QNb*U}5fEkCzfROwn&&VQS9d~3+&
zEpQQPs*oM;pw}k)xYAgquuJ^2Fu<G)8gB)JW}cqd_l_Z2(!&jPBpNHf;hCipxW7d^
zWs^=I7QD?2W;GAPEIj-451gyoIRpuQCBLs-svc-v>6c2-$`v~P&Q1^UmRkOg)Y&Ho
z(g<%zSR3COtiZX#O2Nt%p?}voQ|4qsW()3GPMbZX?tq_3aYO2Q9?c8!d66jjOw|L^
z^wqhMvcS+E;1EQ)y47YN$bT-6CC0i3jwsIQ4s`+DqA1vzeoa=SE13u$HxYb_w-=nV
z3+DFEP-+ek#r86~usuJhlkOU90S4*;p~9tk=<+Ra&Jdgb!S0KZgTVEq#~fEpxKPSP
zVlr`0v3+xB?LURCnsBpQA@*-$A&xz939|7#*3oV37ry$|hj{B+F-ZHG?O{LY8t=;u
z-Su;VP1WmkW+u@TWT4IGMoHUaKY;T~+@l~awCYpNhq-VeKgI{(MsWMdZH$WsTf@ug
z-@+Y>4aMU$2l}CKR;k|%8;N`^ig1XH*Yz{h!}Ej-foVT;Ux_3VMoP}a+%yO#YZx^p
zdK`OPx`o~mH%jkDUL0B={Q1qQ1pXsQM8g2>QYZ+?rXBEWtQ@Al*LY8KIjhI4nvTX-
z#d?S{X={6taRB<F+a5EBiDEq%2>DYNpGx&CDxi9tnb#FD475DMy|e4h6PbnGD}Oic
z*bFw$_WN2iZVT7g3=+0bCWo%M!4R_AV)o!m*1TEEUQsN6jBe3GSYtJ4l4XzyjM7N!
zVN(olM^1(*kM!xdbQUDMioGR3219$cGfN@Ei4y*{DXruQ{CoU8Y!ZEKdcL%HrAE`?
z00$+3nn4#qKM9D!tImQyA3p1=WoogG-xCpS0cobdlL`}Ukv)qY$3R+fd;Hxi+abKm
zO4|L?WImO^-4v;0`{8+0Z;SkMPG)v*dIrA*@$0-b>TJa;OPYgWYu$g3Yvq7?Fv>00
z>na6$b&qzQQ#U6<vUP({JkguhJogTH2_W>?+xU3#Uqf?NhW|ONaQrVxM*vNYI9xV(
z-#k5qI3pU(6#<>`wt60;xPUH?y)gYB*%M|Ro4?KFUHr#K34CR_<?_4(#OWDb3WXxc
zkLDq9u}nhN?&jw&)9ua5;DJ<q@9FvNpy)2N??>t|w>C~qnys+-PQNIJplBJEC-&i=
zJhD4_aO1!8?9D{A3ukg6mLrG1B$QE=cjrXu)I9wlQ|u*Yu_b)t0EUtBBNou%P}+wp
zT7Vf6;mea-Ujm<?{g7SU%T6OSg}_+Z*2)p?RG#1ch<oAN9Ip&Xo1KtV5H7>BM>9VG
zqhsK(2f0ar0cx$4|E;I$@p@?=8NLm+p&&-w6+FI1VLUeUqNk@E$#94PLRXXk@;Xv@
z|4yK1%>=FB9wH4F9uJo;?h7E{jXqi!z+O!1QXCSEBl)ucQ{hq;@s5y{h=mm*u<EEE
z!UQ|6+D1W}t-_PG)7oDC8}2V-TD_4~oKk`y7BE@70ze8sYKPbBnsCsP8YJT^-=rBG
ziRT=BQj*&b*tUU&oWoM1H1>VR)wxC?Kq5#|*xx9%8RR2yGS;LSd7>SSZt%gf38SO+
z@^cmwu|?#M?!T!ph0jC@QzdtU<V+e(#`Z_xv!uBEKZFq|{X@hH7wP-Lm+krgQmwB?
zpQ8p$M586I%Oy^7TuBVHB+4~5jpXhxTJ$k>ml6+`qfn!6+_aJHrFzIezJ{T5_P#nF
zk0#YR-+H`#@HVD!1m5_}nQCv<H?pKh(#;oOqNmZohUOGLd{#9X&tnZ|)dkeAxFUVD
zJmA{0_tPt?Q8k6c+=rBRj#PexXIiQ_s+<M$bOb${pJoGKpRN-cZ?gOM>>nG+VId3X
z`y}8HkT<tn*dJM6Q7C#}?~9|ozW>UQ#h-hbRb>4I1PW$)m!x8BNddXMt5*DWu>CSr
zA*P!IrUW?5vcCsdZw)#?xuGZW?PiMd7R^x+feuckt&<u0%OZySe<i?K40|-)zfVDC
zDaVMcfkh<LEKsGe>JafLluSPEtumuke6ei#WM6Nq05r>dGLeln*f~FNEL5QTiU4h<
z7UD){g}@K0%aV`CO2h8SH*UF+6_Ww|co0(!Fn%nyHnYqF&oV-7=m35pnWo_%i4?lm
zaKdr1>fdx(*taf8y2?t|sL+!gf*$1#Iq|1dpEMU)Nu!#2TTTrW@D6WH>a=}UIH^0%
z98)FASeEkPL~tVC3#SO*5EPY`k6>DKfz%r^GR3tGi|x(ML709x8(Lg>s0bGM;Cl|a
zL?TsB<aMDOnaEoW;AX!%3EIDWfBCxuo6U0G5c^z#wA*3RD!Z89T{>1zeJJl9@WiZ8
zNl%)Sr5cU*{SW@;04Y=LXRjkALn2tQF1{AucbrT&EE#l13)s}Nd90NV7@O30s>^G*
zHD{Ina#wQ8EoLpbm+4J1-IZC)$aqS9GP%$dTJr)VwBp*Zo{i0$mQlt&yPTIT2iqHB
z?eVw5RG8OgKE`7YboD;kROxo<T++Xb6l+*OC{LTzQa#X{gjgHtJ$W=Qh%1GdX)-QS
z6;5*c(dDL-)}??ZXUrLQKX<nj=5%vzG+wpNn|gFsX+Smrw4UZxTpG!>7NmJ-A3I$o
z$#IgTIdfxt%rE9S3nV|(kCH_6iU}fgx}Xcr)6SCDew}VcUl1FI2X!C4RQX!fL5wUE
z6up@=+BJ<p=OKID7`B1yy0MBq!b_p=Fa-6Xt-)*XWRq&EsJUzNbx4qUpjL96y1BrK
zhtl<m(kZC|Ol8RG<hCXXI2#Ey9bGin*jCF4Wu}=NR!Lv<4Jh!U+cw!?C8c%s4!5eg
zlw1fnNq6*ocd8u(3?nL~<sK{-C4jUB@{Qygu1N9&<?kmxSTuDxB8F}I1fo>a8h;^h
zV>CO7XN<6g#K@Kd%(^Ms7c{)M@87rFS4a*_S1v-g95ue@I94hrkH`LMvq)WDgh}R0
zWx>WYRx;_e(InSmtkESZvy1@myp&Q4GW_)|pGgHQ;vO&~*7Yxsf1H`_G}N}KfDs;~
z43@8|G}(MGz1A=IA@+d9KyPGe$<tMP)j>D3=4=11-*LD`Pp%^72zO+Z`QgG6_V283
zf#^Asjm}LHCdoW5xTE%++B_=8#VGqD^$f)JA~P9!rxP-@W7(iE?bj(?w0)R}Vfi!S
zkp(@o9P(_yr}fikEeK?TKE`orBPZ*V9Wcmy`3>zcSV6@=T!}y9!pp^^b?kh@S4Wac
z@E5shZr`!fNF?jjAHiVA6ZzvgX$XY%MTJ|`9o$<&Vhk)7xGkc|FRS;F;MfYBMh`37
zPSHZ?j2<zRuPfkzs&FM5<y(q5!YN;gQz??i`B&oY5kRQKL=XjM`P&;YEQMtG-^NZd
zsy7pJgjTtdjQTIf9%44ECv1WNT$4xZm%Bwdz*&?OuEXxd*)a6dL~0GKl2oIp;UnbV
zp_Yib-Ta{ehiMO;u8s6g9$x>3*4I};;+YN5$|bDi>oX>gH%)I+r(*}%Y+9}ain?4e
z&-ClcRB!Rqanw(O+@3bYJC{VMIns_YU+fK<o9E^YOa8)u{W8&uCm16Lzz2GajPaP)
zF=auNR?&}n)XjS1tXhvSHs|)<N}x;46bsM?-4Pv^%_*aJC^ws=IR$k%)S+^zG5S{{
zVLcvpk|col5gyee(HfMSMgrKnr*S)CY+`2pt=gjMWL|$w5=MRVf!@?#dwHn7V+S4U
zb9{1#yX?7Gtz#*!iBJq`b&B6f=76AG)ey_=J;J7L)qZf4@^GM%UA;ZHZaiL(4BxJZ
zqB)nevPT=>RHA*|)O07<>V}D`z1I)}6eZ;+l3#E&-3aAIn^CoZv?NY$8~Z%!jDp4A
zc8r8)ym4l~vI7-Oi~Es18u{@>Av~!cv(aT9tp~^ov@WYlZ@99M^)azrbmswY+@``*
zb}qsfPrGVjvZ-72jG6Qaf@moqf>`>1Mx1B-U~<+^>?I{y?!0$MUh9wLFFF2(LramX
za*#1Vq1N%5pGp7I2O0;}Rx|FvC}9K__ArU&(R5Hmd%?5($z7j0rfs6R47l2$YEaar
zGQ7fRHu6wF*PHYhY3qhW4Vh)~ZB+?VrC{bBNkuYGzElVS;9{|*`xvX6++){WdSh|X
zRY%4*nrhZCc<wzoV-q&y#5%R#>z2Pl*BaAnmX>wjTk!a`DjkTKQF|e(H1q%uPI5kS
z@-~VYcY%Q8m!K``q%l5dhHj-?23&|Ah`mK#baaS1M-s}!%H~#WwPC5+t+^f<ucvp~
zMt0W1n)oU#!#%8=-dV+7f|O-E+JcYV#cyjnD4ah=4o~Ft*3;}Fhhn8JV2(S#s_W*E
z=3MB|)6@}m&D`lHdog1IQ_;060u+gLFWis4ZwP9LeVZ5WBNB?}>Q^zNP-$CFCWM;Z
z8MKt9xeV95i+pz?pL5s_Tbb(a;<hss$N|yFGsCZhvOBGqP~3_qRd&r8D<{M}J(7J#
zc}J8IMhzpy&ylWC5|KBIvsP=xI)DvdEO69XiadAtR0JpOeBHY4iTV4kb8k8LLifqw
z+#4Ih2ZNlcE>yAKZRC@@`wBCZW5HcKs_+NOo<xNmYMG-wF*zknHlm%HuA3&#w=g2@
zpgl#ZDy*|&r$YfJ+L$W(ECgI$K0JR-CwWcHe419F-*j}A;b#yN8|J6|7Z7}A013N?
zjVY$HcAEB$J4;9#5{tbPea>kmdOewkXXD-#g}>DlSjgk7b924hIA|7O!8eyXF<`aT
zJW;$wO|}$O0L9VCk?$?-$zE2wb1tvFzR{W}ugNHp0W}O7#o8uUoV>d!{v5}zwnoI!
z*Otx|!cbdVRq$&FUM~(OZ_{fXqo98;)V?7rffKI&Tb%gcSfuRi|GTpKf9yU+KXKx_
z+9^W%G@uPA6j-7KVvIJX2b=>uBl|f1^*anTZU5`B>2Fey=8Bhb{Ysn)4J+TrNgqOQ
zceZc$M;nDt-Bdu`{q4m|P#*^FHySPhHoH1uy2AeLMcg3YHhUTbANhy*Ov*sqa4(+k
z;CmG^ycpCN!QGc-v%8eh5jfu{v#KC&Eq6B;WQRe|Fu8a#b%Y$#5j%<)1ro&^)&2b!
zl?qwB?w%WF%?>e^rZk~}1?^bhdct3??~~fz?%QhcH@ZUDL2k=I^=Wy2WSvtAWv<34
z`NTC3i2Hjg8GQ_S8-GCyR0wWgVl$u{GI9SYrx*~BuL2w2vU|68xNit4)N!DD+*~jf
zLCmny`OJ}!vc0s>seoTq$7YWkh$1n5;7;JY2OXvN3vtm++V~EZYLZ419jt!h9Y$=d
z6RN7SiV(ksRni^`ubor(wM(cQr#wUc4JogY*<?%fp5gThbICokBs!3f_vlmA68C0S
z1w?%uwy}|~>#A-0vtc`&c^u!VNU7I=aB78*<fT#fcfz5|WlsK~Q-LK6j)^fi-$9GF
z<4-5pr?gMaGv*YL)SOI1Vx%gf=}6#SFPfw7b-tJlskAi9Lza$rn-~csX?|W_O_*3H
z?8}Ik8B$ax{Uj5&d`kZ4=&Ilr)z(MGeS<W^=93NJ{D=@|<ojIH4+AxA#^->(LqsCp
zC~*fKXSZ?)kyZpv5{zDIwU3UMb1=>K_YHlh8e5%>=hRTNkol6}r8lF88cPy&#rLRo
z;1!Krh_pY|F5;NkH&Kbq!!4P&v_M;PIAOY0n98P}LDe4rn`K7Rm1$_!A|+Td<xp2%
z>j~}GkM|dq2-&bY8YzQROW0WRbtYVUB+`D$U1=_^>95!T$0SbcS=LpCIn9+s!f6i|
zvIDW9Vw8fcgzBgGX2@g7C0`4XXOyS{>-s9+gl});Z}9RhjrPMPQ;10#hIU0oY0%W4
z^AiWO?R-Wh<Eka$WDK`FLnV|iSxAMFVP?1WPMb`GZijhdX@#MGiv+{QwvwVGfm}k(
z9oB=WJz0+mA=wbWamgSB&mHsz3DpP7#&ypoy&w~e6vF3VAvu&{(5u{CWf!tjI14vS
zUi?AUL)T|PlbRPAIK?`bB`wz!${eLKlz_({)VU&|l4;1rA6Y6s<clc<01f)p9au*6
z4ki~71kRm74J}l<`b~wH49r6=p<Yy7U<B^COOqsu;rAK%OP7o6cAlR@S&@{3qnY=X
zAz3`4D$1hdW-t+#g^z9*KG#Bfr`l2+XtGO5hZ9LDQa!Re6ojHZdR)VTQfS{4pTl0J
z<JS?TM`@3uF)4hlT6;bsAfM#fa&yZ6=8Q2$etC8uXU7^P7<PSe1%arH(uB)l4c>%D
z@-q_LkV#K;%+}kgRkDDR5=SG{knzE>@>E+&J~N5CNhd2YoQR<5vB>0{-rRc(ZalG*
zXfW7p=!#|#B4pdGqn!L*YC4$y>@K;ojkV$|Qbd9gl~%LRWWOzrX5XE9dnH%eZZs&M
zw5zsLSl|%3AU2>?6!4+vHJDBPUUvQwx=)QNvR@>;vHZB%{8eZn7RNe!rg-PIg$`V^
z@rM{>6S;Stvvl&AhA%O3Ujexx4ty@AKjZ&n?45!%jka#hM5WD2v(mP0+qP}nwrzZA
z+qP}nnU(B15vTif#J@Xw$GUniSH!v)bG&0dLyQ$XL9hqPB@vLhxY(m;mwS>g-RoQv
zd{N6oVoVuj)5V1#@kARl9tX#cPxUGTB>kATw&Savt)3xlz*7~2DO30YZ`XlqGr6P5
zIWo-8g}>qr_K4L8RFXP1IdwH^h;r8oqtC;%HG|wG$%2F7lOjFFtTaMr+yS`glBhur
zAEB3LdET40Fv9ji%Or(Pr{&O+Lmm-k3;;|nUsmKM44Y3qd-8NrKsm~A)Pg{7h=nIW
z`F8P+Lrt}AvBOAD3Xh+F6_^FRF8Vfhvz^UqauxG*j}!Avb^%{?2r25makjFkA!H$4
zCLtIapk>4N`eOU~e0%=AH5RC_XG%G_tL4)jfC2ncIs+t9k=C1jKJJF?>V5MjmNdGL
zILN+3xI4bwjL&X0-{H{o#{i#E6Vl`H6U$<ldN>Gu&Ew1&A0_w<tEq+iEr*;-q|F<3
zSLth;05|%a4ZUT1jKm{}#x99k{Yqoyeo=BX#nQqlTRRYU-B8=E*=aLvIRV}+X*EFG
zJz_q&s6Q!%N80Slrnt3jeM|6S2<hM(P87yw44M+qia{2Cbc!=qf58|sJw|hSzDf}`
z<TMZej)=U>ZHk<vWt30!_AxhLVS*GIJo{IxcLI&6`9>kt(%+qgv0NkaMA+V~g;j=e
za3$^`g+-U{V2cAl|Emyh%*1YciXaeaVK$03yaLr}tQr;3yN&Ftb*M@|`&BtvY!Z_G
zZOtjsJAw5`iQ$PZf_6b5neQs634=rizK@AFwO<Jw;K75udDJ4TytcV~dbo0#isH!+
z;DSsH@CwAo-qBAV#RcHw2*vKwLA#sw$C)B3Rd0zbNrEfJ9+Wq;qrd*)qHO%Lm9Api
zBz7f(x6Iq#urFSXe8&{olJd`h(etbZe=5&2`_%Dj%5^8YrK>hrz=cNzkNnWp^UN-@
zt+j{^gsfdQgYSpHfpG)Bd>-1~C_@`9442!GxaYX5%Y(J{2u_R)2ysX`;eSI<lTSDF
z&vRcs?Q}sUhomLTbLW0gzGGxS_6EsFMq=2<qp(e9_=7YcX9_!7dwMz`>Jf^)lT=~?
zhda$2o^Sz&J%ad$P{4~+9P$I4m(@QH2!Q{!EHF^t6&g|I6%;IJlmEaUUv5t*Du;I`
z94`;_ikcUJ49HkanrSHv9`LOzGAB9;fW%$tu`cbT{Bye?)6$+u8_L8`)!FiyGK8{i
z&!^(OB?O~3K(w(03z=yp=WOa@WOGWrC@F`UMTK=<x%h=NgR8n_THKX#27@_68QQOq
zQdX;=LmMX)4^u8ueHr&l3Nk<SlCOedm-nxSNJ;xvT6!j%O<XNS8dR)NuA}vvo6Net
z<hFPBo1NvgBfD#186@;mPdBk!Wvq;9`=PVNZMd-JaPhMnPJ*>2;igydn@JdY+_D^U
zdP-6@DA4MBy0ki5>`7YKWs8GOo2E-#$Dz&2XZTb55mvi=yH0n)^=ak>Ow%6yxZ=Ui
zvlDOc%E)$>sw^WZLVoHODSongv?qb}mRL2#2J8x6LnlXHSkymjqA+ced^gRiB<~GV
z6WU`nz8RF}E~D?(9-h@0{Ix#De`4N;pML9u?6KmM;!?jxmzZjzYVf)HB<3i6P_o+v
zMi6An=P$$5f*lH666fgwz+#&@ZNZ3dYV97&Ez@gedPOk8TI0q{djS^;Y7vfTkvSNe
zO{eqs-U{n=(1w4DNLT&$1-O8()71!WSW9L}Ye)CA8CQG0g<ofu9VDbx#&64?P<}tD
zy_Xt4zqC_NO@3#)4r~I%h5Kz&p@cXZzZBZhsGezzTdSqP^b=JXYgWf}Iy~*;87|V{
zp>5eR1{^Aisu7L)MRk+{lfh>3jjbRU^^PkZM~-vhowr@Or6+v+g{{q>50MOpoyPIx
z$&XF%;1e1mmMyiVidPT4EcUIm{goXH9$f31y~yX-!7qZIwql0Atu3(8TOFT?og<|p
z$qPRuN5#$!I}U};FONF&Bpw~7sP$F49XEFOHMIc<NRQ@b>*mtnPiyKeff~{o-;q5H
zeMjocMJIL_aqv~#>uPxAUwZEofmEe{FWYXoDsLS<k9-`33ZV?K$>s&M(n!Edy2z?g
z>MBA2H#)(9k3dvAS<|Ay4Funs{zxQ5pP!IlJ6YGF?hV8b*!Hn+f6mE6-}Yg*NmRna
zam+Ev9{g_j=?~~@Dl_?i(Bn*O%>TzR|NkVYa?sOBo7kE;n-egyaQt_@0n3^;cK;Sj
zzi;#e_Ik8o(xql#^dg1)wH;+|_VxnV!8X7}2{$DVmg-?O;2po7eL@qh&;N`z6c@wm
z3pI8&9@(=$PQkR=OVYLtaBw0s1md85S|N@ayD2)1s9?0FK!;;#AH0xvDpz%v)cj#3
zYR3q=@}HnZHtST=K3ZC<jKUflfu=oSH9-0;)nX9Up>D(sU8LN64=q}u5<?j)h+WIS
z5>ItfzK$fqjV3N7{bCz(SIw$rB1XXsE!Vil$$%P03Vd`jcD)Tm*{FbIg%>B5&wT4|
zC~uR%!eQEvgIs{!4<aDIp}EnCX?b9YdUJg5o?a9>@*8KTp-eqLJyam!U$nq!=t#I^
ztsAy-4F~uYd7V5#fFvn$NdoF6nPI~C2wHMD7rYEfFfdYfy+snzE7X7<B6I;~WAUmf
zos|HheCw@pQPMSzK~hFWXubq-V+=%68bp?Mm=qseTv-}~G+jswqETw&Pl!jcDb@fO
zYhvcqa`o^C%!Q+VQb>#?GGx27jSYR2bhCmx^l^|Rf#VSBoF8ip>ik{<D_Fmxqowso
zS`%EF4y{f5oK=l!U!l5+|2=w_W?{FbXOs;4CidEI(>cx^2x&{4L(IN1(+9&Tjy2fQ
zB5E6L?URKY6}2o633N51LnW}~DS@OadXg;SPD{;U>^*F3#HW9TOB9%%MJPB#6-(=V
zVHBaD16X5#AvTsAKl&!CG$Lq1J*iRtA1=PHrS9*;!d)I-KKy9y*GGN#1^gZ;`#1YF
zo?GAR+&z#(JlERy(IU2lF{+UO2>0PPg>kiXNYUqj%Y9ildl%*g;8(fXpGBS?yl;1B
zPWD^)j^5t)<!?W4J}e#WZ@}eFPV9~z?%%`y(VM~L(WvdD@3NcLs^4^9VU0Iu*7i%!
z<$0a~L^(&TiJm3O6=cpJYUE_dyOc|$UwHU#X@iG1<9psqTs*f3be)G35edF)WZT<v
z&vbEk*Lw(pJU&XWwt-&DUx#@=fwYNq@)n_L*?KuSxH|(W{U!Y9512JRY-ZK{Vv<bR
z@Qk@|p%A~j4||Nf`d#R}iWL0ZZ_K^_PM*#;1}WGrh3#y{@KA#}UG=x=(-urOKtKd}
z#FlL}e#PE>x&mK<1<J1J6U}@3coVwLmwG&vdvW04;6I7kb)?C?cC3EQ7L99egmzzb
zU$VKp%!;Lfm&>SP^dIk8x^MO37*%h3&hv3xe{sKUMUw*8IhF`@@eO5-u%OqkuO740
zH=VofG_C2wSl1f}7E?HUl}kr(5{_W8ch#OVy26pilYhXP8>?Cq<wrK27GrDN*r}P7
zb(eK3)Ou}mabD^T9N9|S;JyY5dV@I2ZNoS$8(reZd&QK(d>Qa9-1H=yz~m;m{=GPS
zGwgcT?ou~Q7z0(4NR+cQ$lw-yGk9GycnZ2%-+RvYy6}yLtWBQA8zTbiB19}+lZT|7
z7jxebhExTv+olaBY62E3A!;LrD#)BXVFYUO8KW6|fa)ng?kdq0SK%&cBKT!!ESjzI
zA@mYZj+F}OfZ#kz7seTMGo1FO-_LV>Beov;MR7GN&joI0fR7D>LT^qZAG9?KL_>ty
z5epeUEfqKkJIHKIe??2&s~(^<rK7rEEZ=QH;^sue_JlUR@k6D*%|^Z4fS78jNldNb
zV2_(hDMKW4OhKq5xEjf$G^+RsLRo_g)0_xl$xSfJd8-%uPWHn){QV*tRPSf8wLR;A
z4WGKEg&qSks>PzZMl2-i5xA%NdkY4trGT<QyU$W_nKG+Dvql`~4SZRW(cF)J;^&tG
z#NgtOAt-HVP*T<5d>$*2jRmnhglEVFG2xm4grtD%uiI0gY3dtc;Re+BXVqj*;Afgh
z!5q+zOf`v;)3y=!BL31p7GScdbvp}#|4_t)K9n8MWYoYb&n$Dcv{1cp?>7;yq!)Q}
zGZR}ZbF`^eO^v0(gRGYl(%}-NHn<T`NM*lh`g^^w@TFcf0q1?Z`F}*gE}0Z*hHZ<r
zdlhnjx4+|np@-u{RW?U?Q;bst_S(1SAyk8kDGq7X8bgZ(YSuppP2BUn#D)bqMC!#K
zD#2>`<2}xVU&I<ctOOXGZN3B~SS#otgE4CPU<j3K+%+Ux6ZXqOf&nE}AVw0=OSP+n
zmq-4QSE8t-C?gPgJW;e4)RftJhgvBMU@Ag&StFMVxha7H%67r&O}Ck0)Go4&)&1=_
zR+1gb$+2=v!W0~cUYBT$TYcmnl;RK>%`76CNklxKyceS=jP%U1WlUknKsgo^42EnR
zs@w_!Gu2~kHCY#T2W$nV-l<)U6}bzJR@nCeml=<%5)4A+rDTOxfTxX){O^?p5eZ8g
zC;tdch!O?_0U?@h@Qx*AOa#$%8I8Owss@6zU{>&>Qe$k&qNx&`{F?LDL624{D3|W5
zMJ_a<bb_jo20EIAI_F;fb)xS_xT?Yqpz_zev6Zbn0@d~elT9Wk+Lhs*@BlW_SnXGD
zw}9SMn8y&$C=0d@E+6~s&Y{&IrtFeN=D91xF3CMZ*N!UcYfi-7evbqZ6b!EDA4YwZ
z_^neFT#3U{552I6aWFD1jGD%*<l5+2cw2F0D$D`p6egD1=Et#~Y0zf7hW5<xK$(r%
zLA8qtd2{*EU{7FtT0`~mJ^c`(8emyKf(y*wT-waerf{kq%2~AeVJUxD2!6?VR^c6#
z%gDjsSZOMl+&qWU07Q2+7;q&d*o+4qnC@bIr5t*D+k<}<e^u@FJ%bQ>yTXkpq&h~!
z0P`lYUA$w6h$K9nl_U50aIt>dDF;sWb@TVU@f2dy+*Pyo_c4#LsK+|zC7~fr$(wi-
z#k5JU#QFNHnhVwqP$p(*MRyqymA%Bs45g1r?b!>WvtdONj1_09LlEw7)hk)q)YmJc
zd_^*{<%(`5k3>Z7E0{0_?9i4YZ>3L3gZ_`E>O&ukkq_r-X_9q9V&<aps+>ncIM&^$
z6{>w2T!gN(n4KSAhIce9jGkfeSzJXn6pdrAaF3et0whV*tHafB)yEblD~Rz=zpaK~
zg{G*EsUcH1i<zMlSc{-P>~vU)nv?MjQy2?JL69GK=7Q4q+~djV55*L-Z(fnp9gXCr
znpD&h7Fm{McfpC&6mx}3S`IK6IIZ^Yxea8%fg<@NE5ld`)E&Z+`AK$JpS8JGI#d!V
z))bgJvt`r()pS{RyV*-PzC7Os$kpdqS}u$dpWh++T`~Sd>tZ(eCmzP9{xcVt=f$c4
zu?uO*jUjf(Ya7c@e}S06Ymu(p>TVGd7Kyqw)3Y(hTM2GkLAj?hC~OLBCg(x%N{N#9
zl;X8S#5KuFqiIZ_RWKks!H_7EZEZU0VJlY}x^+Uw;_1ao4a?=_!<V^#@0jsoY2odp
zd8?FS8W4?9W?$0_sqICxXexZN(0kJ|rCafAZSQ>H3l4?B{0AM%%=SN*78qIo%filO
zEvckKR`{ORf5=-5c)h(M%gU;)T-QoRsY<&F9vho~weSC&VSiNegn07tR8QB}{1%x(
z0!V-Sd;Q5t{NEmb&yTMUk1w~^qJ9G>ucpQQo0)s#-_#^<obB%=q0E+!4`(Osc8izb
z^Vqzr|K&Cg?|u5z+`kN!`Jus{%2(-jD(yf4J%>)X<}AjZD3fljLu_9=hmef_avP8i
zGrhvlMaW@{Q9ZrB^Pw29diz#Ht+g348B+vC*F0x`Hd2J0YJc9}{)!}`7l+e!%g}u;
zevD4)(P@*B3KtZ2etyiRJ%zAHrsvaf9>f`UeSYOl-qQ8jw=_=#38qM$0%4^D{f3_!
zZYP(1x0L7Tf?I%&GCs`f{b{R9(DTdu<swU7#z97t<r25F1AU5C#(MiEOhIFbA#e4<
z+SPi_gEeg0h)|1u{-*=XH`#xrF?Zy)b{zC5CY2zmI{sT+Q;;c8>diCGQGx@j@D8y(
z-r|6nkiW4hH?rAnz>DwD(Mx@p)4$J5Xc)mETJ+7FL(Bxh9u|a!`S8tW9xi%mV66RP
z|D@re?o0GElmPZ7{Pc>ggYxj9?Ms<n@b4H0Ze3nvcru}VEW}kAQMkDo=j9qgGr96p
zHkNW}uEm_EK@5JyR-FN%jW&<<opl;j4CucD@c=7_{X+vRx^sm|X0N=1zpG3O>D^{O
zK~1-21+3|036ixQ7=yABb`xkf7?@9>)_i9XuzbPUXpWd~XqQ^e$hBvTrFW$LbOJ{y
zKA%_JYFfOK1vLOZYbLiCZ=m^m%Ddsa;6z(g)j%Bzt})`X31U-1b@;Z>x&PfYwHim=
zHROO@x@Oa(t)$u?%3kz!7eXr}f_XrK#}mSa{Nu~vjq-cN@hayI><Xh97l@sQLRGQ{
zr1b6cuv5*W0PRjAS)yOx@AoA+(T&tg6^aam0!$da=hfK0b}pYIMj1*h&d7Z*ahQ|#
zzb}{5v4HsE<|1&K;RrBno@+>%kE(}}&vm5SoiSg^fHOqBp5-hGp|js)j|NAkdy%_m
zb~^lg!~5KZIiW%Nz))a0mv|)0o{s^?XPw2l>bP_{@!rb9l5LLY&#d&Uv5aIa-Fc`R
z<rz5vBKFwA@rNGSP9cB7y|9ikM&XhVm-2!fty6QzM}1vN7cr$}ZP+6nqbGrrtvGT5
z9BD?G<YOJS_}f!E!W=gu5D&--wwf|BWjwy@^0Ng=GK&bvXl|nFyOIT@<2rnxsWfA?
z)ca4u2%E<*I^<Ndj)-798q#f;!mNoVupWvUj~SPuSv6Yr)n7fmCow}XJ9BAtDvr7J
zuz~SRBRg8*(<!|qCbDbDH8nLy0WhDenuIAg?cc<b)^Y&C+AcQ9>h$L!0c@!{m!=#y
zO^!b#-3!2(ucHS69v3b3>CF@DDQfzhD+TVc8?9l26JFQR#LHKX(D)7Sc)W+kh`O9J
z?T7N3=gNo1LnO^SLN2P%W_yM?15*xEP?Ho_Ph_)8(bdMe?txMMv_Y}4cYBvu7|Q2f
zJN?Q8b#tB$j#F3Hw<;ZWkV_~r*j<0S-5qz-S1maTDqA}k;$u|iOW>cuo2c#mO6I#f
zCdqS?kG(X&66})lq;k8YG}g@+RWUV~R22$Fl5t9%?zlV-eJcN4=J@JP?K*y<7Qvp7
zgC&Wzz^3S@$5V@ARshg<C}qE-$F0f7?ku=Wkb#<ZwpeAx?xqeZkVi|%864D8LnY71
z$O9x#<EHJ3QGj}iV$2ZsGeF~E=h6n*4W!f03xG~pv!`G1PRNyRa!pegHBw)Dv+uw^
zF#?BJQ*a(2yzRBl*wI2BZ~AZ1Vkx|39AKkz@RP8D6YmY@@kT0RqAxIKzz@Hc3aVHd
zhCC_^T9*k}(o3z_mX~%)#fzLrM7N~gJ5xl-P@mvSDl!tzGM!ifXa0=_#kHV}dcI7t
zvDytbG#Ux)1~{5+LzIFp*O-EeTaGol4xkSa&V`6F1RUXmef6mp=?s3AYC?pBBFK8T
zh_nLyPczgPE9D)(45zAGcS~c9R|2QEcQ&8$$qvkIkry+PgshZA#m*ndh{^n9J(*RK
z%QfT(^2i+)Xhl@B@Cu{TH&8&8vURO&dt8rFRmvU$7X=}NIwWzJ`da2)E$TDF#$(QT
zFge)fPZg`fu#fv_dTS%${uNCq=qf)U_EcA|EyZrIrVFI9WR}p-k9lG{RrKR<M5kvA
z8ul0knVhh*MUFXaXUh2xB|*0#E<#^&lXFqD$|aqlLL^D5v#{L6wKD^YwC@XOS+Yms
zb&>B=<3+f-$_Wxv25I$@)sy|2cB5D+lD`guyhtE-mMa_}R^o**Nz*v@9IEtjVUMOt
zKE#R3Om>9ivVO0)*+j*O9sR`04p%}rZETu0<q+7U5tF&gx<%|h<`OQm7*j{8!W$#A
z2%m1q<+@o^MpT$(2YsB>h%s5Fvtsta#dQWRSMoChcrz`pN$!?a2I^)lAUR8kYHAU(
z)FIz8snueX#Z;+9dvbrFf&~}vXj9tzoZrhQsEnHL{Pk!_pJjhl6$<eb^~rCjSj>Qn
zbh<(-EVmvm7Zandv7;L8Qh~>5;l6Gjs;@f`R5Y5|ms9@?m!%n<IS^B`thj73XzlP2
zMdBn3V)ZO*D$yQjKCayGGM5@?zw_qvO`s<yoYt08j;^69)k=dp_ryJ1x{qq&P+sQo
z{8lA9Yyzf<eh=qU?9Onw`b!)DT$KAf$ZDH@(&xKqS4G>0$>NDtEU$ua+pbG_>KN$#
zYm{lG(aAebr}II{EqU7Uh=|ffo?}$E*9FK&bo;!O@;8#(jPq^ieySVW(5Xuo`i%wk
z77BNPi}-Y(P82KFBIls_vx}wSGMNQd)+-|ZZ+a5JMhP}st#o7D%eA;v&a~kO0?6y<
z4BaKiEytqN_n);m)wGy(32OVWdE1k<=xTjEhaOju?e3ONm`WLt&y2sL=0%6pD=<Z}
zA=#;isf|rbe6$1Dnlc~3Gyu|?YJeZPMjL@lfF8$1oL(9$>(@{0NrTUTi^ZWbZSZ}@
zOgCK&Jymqxx#w}uQ+TCx4T@0t(o!a^9=8XNxCJ8VK*TueVOxSU!v%h>^mS6pA0=%q
zF57S{H4@F5)>C*~SWxR*&2R?IG5xfu9uJ}BBo~Tkk}@h{vaL2V9JZS@h6`eN?ZSst
zqgGTwB(|(w1qFaJCuHO$R|84DcuoE)RoJoW0n$(Y59elU^+E$*8@1AKIYC~y#5|hw
z8kJN3MyhA^F$<ZN($N}Sf{a?-l6K)U{rUl`lflVCH{ef9-S^Z0`e9qhFwJw^=z-2c
ztol?_GicPa=q!uQAP}E6dAAH3>_g$<0H45CXci_6;nEB4cR5(OeIY<A6uZ<mB5yVn
zeeg9OQP8kgoY$8BTSz;Ci{Y$x-A4Fxdg{jsQP)=QT3wzZ15aa4;Z9aX8SPbE>);}j
z`L^Lwqb`S1Cvm(XV0|LWxhy}SBG;!rV@7wEPxl*mk4?nsKZ#TJ|CdSn{}`c6?Elpc
z)H3#X>?P~Nj@~}LySH{_8xS05ODC?rEz~>ulm$Iq84l@>UpkRSF}Y=<aj*Uf!6uEu
z@CKXe<RU(q$NTa1`}50}?cPza*zwc(<*39ws6?LRf=m=yS#q6dPR}>@*EelRi?hx-
z2t`c6<h-5SuKTxXav2)*U&T{9_$4phs#|BB2hA9ZZm8pRnt_;Gt_~sol6mJa!ZJ3C
z5CtT`b&=iK-6jK|zeql#$M&-N6Q-d+R;Th}OSI9MTtAPO_;!iy1pjy)W}g>7@1LF@
zvs$w#!o+=Hi@o38Q3!qH23@=qHei-s@0Yt}_;$T7`dX=8BGhRsM6v2O{OV=0E8qqP
zMN<zG@meFAKs2@ZeqRn3H=i$){&j93;yt*+$TCuXI=JFXJFitmK>q9zta=JZ{FFdL
z9)D&g-?G{$CBz?c3Dm;Rf<>Gf&O!)T>4iPA^uUm`UZFd1w-s8#+JxQqKipXscErW?
zi_^@b{*>OvO(8&Kytwn#fTi6Ib}N#HM2{JK#z#M%!S8>5;T=`8qSjzg_5F=RYhX5K
zd-`jKv)Ej&C-UzfQ{Em^H;J-6&kz`|6%>g4bJe~g`9^QAqac;HvmfBs+FY$}GGRLg
z)p8O}gYMijd8lj>#5@WVSPSbE7KWc#6llxTS@>jQaY-Mlc9jYviMvmXZD4UezxOJ+
zd<dmP9uLUVcgm9cn>bg@&1XwtqokrWQH}~h9cR4Yp9ZST^vZ)>Yl1%5Uy}zhh2O@;
z6#89{mn|AE8t$o|n<WJ^Y>9OaD$z6!Fw|7mL6NdA<5hP9ZYT%GHG%Ki@GClyQJfma
zsvMLi#5Dr0$1|B%=|vb#b^4W8yo}!xW$L3(LB1i)1wn7EkQ7m5rhkx(q2E_gpWLu1
zvvkFV3JWy%voF=MV&aj;t+yslZK-2QVTYvSgTu*nOwl!j9{TA$qKuwvt_dMY%A^OL
zzW@^!N@WfY8m8=YpRjDOMu(^M5gPDozwJQ{pM<vB`N(MkA7HlDVn99grA*t4k=mjm
z=CBE+-{cDi|DmzmQSl$eFf2lF4lEdfINlT@UqxO5)arMobBRt^jCVMXT9e3(?Ft2_
z_DhP76@9K^SwMHhH>7d$t>-PW$7Z<^FoWz@Y}dxU3+hzT3s?i!)O6&=CwwRpklCzc
zl_si=Vy$(8;7(l*>k4N9!>Z%4h_jmAt#kCu-`E~0xCX#1WaMt@nPcBR9r)3dvGdzn
zkyGHK>eA-7MsmwM7}-q8{EoF;!IEa7R9Weei|ISp@ET>9L6eYqiAeV7Y6p1HtGnRl
zIPN{W^LgCuc4pJr==^npp#`Q5GSV!+uodfZCQt3b-Q`{c-J}Y(AIjl$J(+laudVA|
z%033TOQrR5?1SsXviKx-s!}utzDadcIW3c5k10;hItV+(Qf#Ex>(<FVF1Rx?2d!wm
zw*&-gv>dgQ6-6?J6O^UPo=v0UHBz+Amcdi9B3L6(815Y9uB#Rp=VpM$<W$yQlTawf
zHowH*a#LBQ4n_MI+g@UF?YY@%2;ywAhh>|Cka!6yzArl+n{XIU;G>jV6!w-DnwR@u
z&R=Zeh{Vu9Lo9E%LXx5wgsJ0<RzHZVnlrJRwwWx4T|i5{(nQw$>g$%FCh3mfd6uCD
z6%Go1CS4(Z7PE8)q!9p#mKU_X@k2d`&u^rWY}1Yflt6JIpmD4~o46p@)IYfQ@Ph(=
zaYW@!=h&t_GrkbYDrn+8p)xvAax_s*>MI}6th(uo$wW*2mFi7fIkXl+3Qxka64bog
zT@<HQ#1~p{Y))7M)SO38Ed4f9rN|oH2_|svTS1G}Qij;*FYcs$icc%wceU^Jl!$dT
zmsL%>mO`=26}$~$aD`#qD4-~;m%zJaLTjy=>2aP`>^X-pM)&l|$e<;CjO0UR5_8)K
z6`&sgJBx@1>t85h8N3}Tr9VnakdJ_2@v3R^I(seiawBEbd-g9dNySwxh0~%VagRH>
z+}I5zqSGP%Nmhy<^;T#Z!<&G)qiR_a%6%IbqjpC6@h9pFhJkjyool`F3dgUa`LJbC
zV&7-<kdpUZR06`zn>>vePCbobT6a9WIfO!0i-GVW-6+bh5_2oK!N1CRAc6eS;-YTA
zifICwmswLmPNcs@QbrW58vb%)Ln1qxwX<R_iMWGqXop)*n$|=@aFF4+?WSW!rogI7
z3n#Fwev$iE_T<My+v+jNkE)LkRd79S$>CxiV_d%CD00YA*4OWs#tO27$FUW-2@J*v
zg6pbG;ju=67K1Yw<Pw5I!@D$P>f>#$2DIP*+}T4vGY`(LoDL&q5#><G&+rFxP7IC4
zRH=gGjlp&-F}A#P)69BtxEaAzmFpLAGrgnW6w|v3w=IN&z)C%kQioo5@{lFXTo#YO
zY%&hRiQlS67}BUMGd0;y7`s+M(jyU?=FwFwV`5Y{{DCNSRmt!hpa(FpW{Dr1Bvcs*
zcg>1=1Q`*0em0Z2k9H;;0Llgp92Dm>_LiOruSj{TD%tzzYx8qUx__E;|C63#ekcH>
zJ)Qh=o(X4ufsEI!6K^kkg&fDi4x-|qjaY3_;Mb{_xD-bwa6wG&m6yKlpgZu6%*(Z0
zR_-3F`2zYgwC$)8DXfw$zNJTUL`v?%;Z7?K;1$u}+1!U{&_zmss;ef6e2XaaaTQD6
zy^OWXbE!C=71ZLQ`@;a<)&$+U0hB}mILspAa{ojb!lH;{86X?-V{=EbEb2A{M61X(
z7MkU0YJnBqJ^cnCe#VgtaEV8)kqvnMpcSXcU4dAOzDX+(F~OqK;o@uE-y*$~kS#}V
zd|)nb?yad9SuE%-8aX(s3bHRAIbgnk(vC(Oc~jdPIR8LtW2Y#$%-(CrXB`red}Li7
zVS(6=-Hy<$+Hzz&axoE`(kX7F%ci|1)YW6PDd$evn)rfS44nX{y!mnx`(2xhul%F+
z7z90bIRG;X!R0T9^BXsvglgVAdT2yn2a4n{EmoL*#<6pIb@W0KWc^;IWqW}*d(XOI
z(8Q-4Q0zYIG?4U4Z+Uv-jmk6`@wFYcMWsF$d>qiRfOhF_@0}~F@J8(WxjVf`QTZK?
z$!Cx&OMy5tAeT{Abf6QW$)7%n`Q^m}q2;YHc_ZSn?k-R^#^_7IwlcHp+~<aFYKbt+
zmHH!6SR6Olfy(IqhYsO(nT-C^&}1JdIrK@gHXT%c$B1;5{Ou`-2eWJ*Ll1L&lGvu}
z6)<kU^trcPCzKs;AiZ&461kb{z#Gabm0%P4=)OH+uiMyzITHn++T<C`@#teq8-BqI
zN00t;F?A$LTgT~kmIeA~pnrB(KF>+(6b5ew=E=#S*TdJl^9{oQZs_x$bR)<A+?{7-
z_}?Iw|37#@iOOW`COvG|Db*+FePlBEJimS)>j9YO4B)Jp=lQ3Wgk(a&(2R{AArdSy
z6SD_(_3K>w)g#lcEN(fB_E|c<uJh}LA3pmt5XGpN7Wipi$n9hEtBXBMVTGEC{j|IK
zk5?+)KU;trkH1Bqy9_ApAgZdm;*H3}k8i0_#NpW&Im(|i&XD1sQO?axE@;y4!d=AX
z@Aya01;Hq^!dTazfkZHcHG>(?Z`(IJ+hm{PC|!^)U^$*;h=33Ry;z_=NE7{zseGBE
z5Q$~3OHIrBj+84wi#f&T{$Iy}YG@Hi{$MUBZxs}>U3_wIr0|5@%WxZs&%UMRkZht-
z<Vn$P6$5@69>nS)F9N0n(qdXixAA;g_s#l~jEUS&?IPJl#-f`vhZara3gODdTu;Z6
za$?Cuxvu8~z^{De>0e@+*TQaKL1@p>yYoA0jcbGkuV*x#MJ6)k!^WnPXliK0oKiCR
z#xSFqR^?x(f?8z)S5!=0A&M4D+fbnw?gfysQD(x%PVJcM;CVd!MNWGtoJVE#EH|M|
zeTdELbW;kY@z=a#C*FB@r{&-zm(fxySayz=(r+24l}Jh}3~KPK&#8KK*8CU_ElU=*
z<&xc4t^jR>^bJ{ZBy6Fih-$xt)A_PiN7H8G3A^&0#hcWK$Ab6md0pw?SvmKaojSr`
zWET!EGunx6SK3(4ox7f^H<>|2&)p{;nd<vx==)Gyt!HhTM`EKr+?$b)!nDdC<B2QX
zYjDZY=GU!Lh}xyxz3*A11tnVomPM^^0(?qdjTI`cR<{<6Jr!Iayl=Y8S5K}lP=fs0
zivRd<|0hzM;a_t5zZ!T{;=2DA^&<F{H$)HUuy-vYm=d^9d}<w$O1kSO5(Xd~HmT8m
z#uc^)_G>i&o$uLzF9o_<oqM2wFMz3ly96xrcb&d5U~bo*53wTl;M`QWq}Xt)VmO<b
zh$Ak}S1UhOaoP}+CRzoBGFGxjkOWyBWyRN4X(n;uxMzyQ#e;+F<ougPi?1<niJ%T0
z{_pLQcCfYpcEg$G_^9&DtPN=U(>^Fv5Le`XxHk(u$Nx#g{NIBfJ==c?yZB$ld#tcM
zx3%efh$@9Uefs_yWGMNghIS&g*A;2sKSCx6$>bVLmjoQ_D4WO<_xJbv%V~bJ?+;%x
zi?<)6n|26!v?G-GhOlXIy>z8r^T&EcMUwq`f#4=C9?!SOvsEjIO|G}cj+$OtHR|7a
z6SciQ^|UEyBWYAy-|ufZ;SgZs?QKLiKFYM55?ey;u*feV$(s`@zCJ{z*gbBqGelGn
z`Dq88IB4`fr~qw~UWZn?W3gN7-H7X)S@{gm-F7C!l4sqwNUO-orsh8lE@?zLY)Gj@
zVzt&|)J@&H-YJFh4hV!XBWCXk?iu3QRW`0@Azv2Tza4zG)82WxWwcUk#Sd|>H96XA
zY>U;@U_U*2dz0i~yDqnTS^%3R<(JXkj*^|%(JGy$1v-b@wkC@_#1wf(by6%`pE=&M
zF(X@fN2bz*mMT_6_lF#Vf)2_CYwOy;HK@wp;Gyk)B#RJ?YC0j5TP16G>W&C`92g-N
za5-f)m8C9CfrS;B#wfQV<VLaBm~ibj)H0n`WcF?cTgB}e_9hbQCtu#nLt05O&k2>R
z&6F3-=Y=WHgdq>u4WW~MU>z%V-*4X|SagBpbEYeAN3?%#`ya^Rw0d@Y3_ZY=Pj0nU
z#|Nf*-A{h&1z~QC@0X}Ri|XWQgbMO)FwD7nz~`D3K9ufLKM3Mq7P$EPyNVjLiXg5K
zG@Ex;Hbd0PNXbVMo#D>&Vm7i`l{}lvA@-avoQW~{TXjHw&qa+Rw=phCg5&#8HrOyz
zVr(8ww5}PBpcuPeC-fk;QLX;PA^qBLr=LgPKcqgT=uZFbnel~}trRja6;|yvm)T1B
z$Ek17kj-p5)5-%B#NlkF88@28pEvwAEEqi<6jH8q8{X!)r};J3kf$Ckg6*kkj1e&x
zxC|hwlKh)VP2UglVQY3l)1!Np%C!@7OQIs>Eqsa6*wcQ(cV-qLGb}ALk9p)Bvak<Q
zg-$AptB9z{u;GMr50~d!O^f8*3(XG_nCAr5AGn({7XTow6FZST;xDV-K{;BSXgxO+
zozkpiF96P9NeHznndlx>p<#EM=oudpm&@co{|oRfd()NzmRed%HS?&WLwDSK9Q##e
zc;kK&?87c4I_gVQnX)Z+CKFX6LHB%p>LBu#{t$g8&I1LXRoqy6CnT`pK4I8YgBht%
z?A3xHf|m_(xg+Efz`X`fF0{E{>rL7aiOVd;SzfmH0_~B@|L0)XcitFn{Yc-zNOCw)
z)0BPa=Nv~Tlg9+`wU2{4iq+y#nS(7&rX}BiWb{4m4cA{~ZIt9NsEN?Px1uO#k+M;f
z?21C+<JUco&eAzZ3uQ(weA(&gx4myeIYGdskiX~~=<UR_EsMbAm7k<I=5TN8>;k*o
z6AK5^p7vGPWQ|<Az<7<>Hhr8hCm+w{t5AOwnIw$R7&F6ezPfjo4??hSaet&JUm?zX
z5Af(^mA`9KDBv|`kl?wp*Sd}Y8>L(-NY>$KE)Wk%C$*8H9G%HoJtE|YBdJ2iLkmNG
zjdw=cSaAv<j|{?tqAP`-g*NjY$dxlgb3%v(M$)qdt}^RK2y|qS(65NtgifD$Y&wGs
zo;_tzA`VCNgMM&U?9lQ_Kg5jXlf=h^mmykTFk5i7iMj0P$MPXdoo^wz)H&R-X8`?m
zhL=m$WgEAh#JZy=Mfv05u%jB|rno!yC5>{lFQHhi31x&t^S4<bBFzRSuTui8!;jY3
zC2@pKxZ{9|M5S|`%@l#zGlG~D!X*E+&dHavDED5Z#@y6$no<H5)tSy@a~shB#LVe6
zCo*52t4s)qC$4g=kW;TLkr$^;RCl$yQp}kMf3z$w*(N;MGnTBFFV4zY8xGpXPSDDB
zV?*-%QL_b<I%2-cdP;LvE#r9T*V@0ZWRI?=1O<eQwamAb*RyN*k!C9y=`S$WT5t$u
zK`8`yz*g>axXDBJX_#j$xgq<+1a!p}@6ce!#@5h|VgwK4Zuj1f=4?zgdY1P2obvaR
zI_lQ9?Fznks>57TWBWl_5@(_>b<z4rHPFLU0X(l&vtzo>-lyZ$(C6}+TRNKi6^IXp
zAlwG(pK-kM(n{07ivxu9IG4=tGfqAp52LhaS%HhB)>n>cKtEE~1Mh5SUPuU>2t1P;
z3;HK?>YjY%f^z!>NcJz*V#G0)V)V-_o|08E=HWGz-d<Np<LOrmWFSEO+K7fcm=L6z
z0{NciU!Eeqdr$zE%bLtd5S0`=#QH-M9U1^v>vxdKh)ulGJJNzTOM(AjFi5Beh%;i+
zZD>3ETTuw(y`$C7+92PnT>E9j(&fuC74Ek5`C=&>I(^+E?Y%_8^*bWoLe$5fiKjK{
z6c#qIWz|ju-n@D^xT#q&mbUW1h{uRhj~PcnJVT0R_?9!KE_Rt{FR8`ewzm6^$vsTS
z`4VkiPk-8LiMr8tF>Qihygs3bOSUbfT{S*U%xZ^Y+ouhZQRyT#3aFj}1V%*@A{?b+
z_kf4~L}5V;Zz6i(tcrZ1$k$x^?INDz+a#N6<S>O_lZy9p1X~6`_Z27VYT0I-g+y3$
z&<cXi8pcrqa>LyCc~9)=vh@!USIMgt5F}wSXD3GZIc#N(g{*#0ydTeqc!SG|ZJITV
z?Mc9SlD`X6sMH#E4C(Q5C-}FxF#OLE?}Ypm=@GhAky_uFVuj9)B4KA_3UMHAV{FjR
zt7PPSy@&XIeJYgaJ43l=NjWWFhH-t-O!=%mT_0Zi3ACs`{3n^g^dB%4?Ehuh;aE#5
z4yPUNA5*KGkG|PcJ`Myt=-O5edu%$uP5>ST^abyVAWrDpmpmewJjJuL!3ev)k^kDc
zo;;$E{DGLH>Ugm*wbkwP)vZ^XhT0%T7NW81r_KB`|L#YVaK4koyxX%X?Nu+emz&&#
zo7c>npTg=M>W%9_@73MTMHv42PJkAGe+z1EX#~gHai?`epEmmO3Fo&d{-q5a$6=tS
zkP+|6r4V;I+S~K9m_yW`x3e1)=y|!)_V;)&t<#Ts?9Fw2%gwS`%XzXlO<K2k_y@fi
z909mf9q>3$_ai)__x*i-Sc~twNX7c0W~c&FJ#tP%t9Q0sm>Uf<Y3BJ#C2J*wxgw*1
z05haQ#Sd<)4uzk}MGq|IrXbHpcF`jE%ShO$N)21c@Cf)<go!v0D`m{926C0MOE|SN
z8gX}tZn=%;@XHoucd1wp6IIK*Py0vPRyQ$#^VWwy6(h#wZ8Z#%ez8(l0D_sUK!X&s
zw@|}EA7CxO;TA-Bx~Z<FEm_At*alv5uZ9bl$@u5JNe}e2L6EA<gTKS^z&eRYvM9E4
zYKy{H@ptD(ST5Mk+kbrR!!&;c19l>NOY*#IuApb=UU~CVozy%~I@z~}hGeFi!WMNw
zx2$GPc!rIpUQ~DSAiP>3H71I0Hmi4VSAKb-oyM96mSma(fqa)^we63%gNI~vWBg;M
z9MvbrIY|6&o7FbPKbeCOC?=yv%lCFjljJ~|WwB7AZ~^FXKUW}&mGn?7Rr~0O!E7nr
zJ5)uTp;XxkF568MPdgt@=4H7?yH!P+rvHmqmTm^Om4;`Z%SPsi1Q4P7GMogoJvwu;
zO$}-~D>K+0))m(l_E(=LEs6SeqB`ttC<K(trF7^KFK8kwG=^I^R(!R`t~pO$Tkn#}
zaZvQ9K^(3C#1}UgVz%d~JFh}}X?U<hs=SbD-66!T2;I4026aWEjY_fZNR`t{!H4-T
z=|(5?Bb}IKq$(wG2<4thAoqwo)svG)Pi*vyJxkE;kyB#^CikXnoy=PpGvOq1ysptZ
znpy0i<u+_qR`8NM-ZWrrW=k*5mp7?~mZ|tGix*up=rGrG4>e}a@fhnYaYuT_aonts
zrz#UG;-%D@EOPMxUIf<_T~0jbO_IvJ+L5UbqGQTj;<;Y1DSEagQ(9Z<jwQu*0y1{=
zcE^mZD(9E&FTNMl5@VL5*Lvh-rfkc)6KUAQ=zQrm4R@H(AgxDT%w_Rvt(SC{W``-;
z=c%h(O0(4!O28xetxc9`Vm4J?Hrrk=IVpnQ^loC2w4pNJE|x@i^EHvv*o_V=*JdWv
zRTpPQitxX-S*>8cH?E6|G+5qpioKZ9Z7{lRteQ60GFhko2>HX^W2f6?5VAj=Q|kXI
zhV3UTb=z5sEL2d19USX}qbit0q->r<)I5y5zXS>#V*GT*41^1RVq;E+f0?b|GY|FW
zpC~icBNVbnw>dNq<)89n7cvi;rdP;XcAG}bX5>i5lS0Z4b#d~{BR=B=;(^KN1Nirt
zpGoEDnxonK1=phm-GYQjrjyPw!(4<2h0f!$FLM2{!3)mYHI^gMJ;VIdri!;p?>6pp
zMf<j_3wz3aa(TkNLWt_fl`{!T%~#9?4omWItYHn3Kb$22PryGIKNlyYu+}V}H|VvJ
zcJ0^)?onXrn+L4Iqvv|LI_%z}(QrrTnCktc>MO;VNA~v=fHTezOxs9dg^Ne|Y2e+*
zz3;jME=T0UK5a6_Q#BuH32*y~oa(kfI)+L#=iOl4_jiZ8{VdUNd!)LTGap|M=$+0i
zaX_G(GzGQqgr5>>rzQ)oq0oyQ6EvekKn$Q6ED`DYXZY7s7JTOIHeOJId(tKFs-eQb
z_sa{(yi<SYQR7n({~Ql8Z*us}$+x3tWsVnmV&TI)qON~2RIq5E9D36Ze4xV8vt{)?
ztf-6sku;kcQWiY#KJV(Dd6BbwQvw%1<p!IyhFAh~;*lR@A}PT3*RzlOckfW)1~6@C
zM5$Oh+ZZ1?*76x;6N4UDEXcS0J~^I=EqF0mr4>-S8jaIM5rIlLy-TZ2%Pv}Jrt+kM
zm;t4we)GtUzU7*zTRKJGO@()@OQ-h`E2v+IiZ}JmGj-L_5e}xf$9iCaG}MtaOJHWs
z%L%w8r}PxzG6O_($>G8xc8#(Ed7Hd`O=FkJe$Aa!oTGC-=m%!Mb`G|h1N+zqXALZ?
zNM6mNQPLe{zS$XN;vw`VhT5f(S~mpnx7qWbB!9P+l6^+6#ZadQb3^|Wq5w#ju<P<w
z)mt=|6`k?$AksmCP@@v}<$$)B9_?SuPKAZrKpE&8MM0PjbB;(W?j#>)%*<TS`s4(r
zYnC1xW$9n0y3PaKY;D7(iqbA}X2+Z-YRf|eG(EVYZS^Ma3U=Yp&`gilT9F$u1j5(W
zl9NSEYSN?S%NVaTWs5NcWlq9fM>#3qsRu5^%6;p9hf<&H2OgEHIeYUsJNd3SPCo*?
z(MRik(m+=XdUmgaU2tB>cns0v&sZYr#cScUT^RLMgU8XFGVxB+vnM5JFd>q{C6puW
z)_Lh1{m`xQcfgmF%NK&sBehQx%;E~7@mEV?REZx^3UN0l=$U1s0;M?;iK5guTMv{o
zM#!ionJ{D>CG4CNrD(Vnl;co?!8>_holY>99|Djx&3-+HMxwGgBDWwUF>ZFeT8BZm
zzZDR^x*^~N>j7ns%CmC7g;@(Rm}~+xjwyzhh)d9lJ}-3iHcq#(u3;ZV*9>$^SDaqV
zON!qBT(QQs<%&HMr<UrVC+CCGuEQPfhfA&}a!F?PrmJO~QHiEaiwgb@KJzFd4cZgd
zG#F*NmQgY<U0e>d$!3!^5tGrr9Zg&jYnaAAU7Xr)g*^NUlqT>!$(1Iu`BA_7;q&+|
zb8$H@;T{waY7^2nHi@0&-vx|eOBKuY^$JUls&)ueDRHTj*)nVdIe4L_Dg6;-z3YuG
z4A#Pg(L76gAmrq{3)`~Q_;+)St}FaH$1ue^5qHC~AdPBEO__}*g9n34Lxl(rxb}{%
zEH-D=cr#i2uj3=8h-1=tiIfhiZS_5=1-VAVwGE8tcf&o%HrWh>H8L6OMDSnZjbb>O
zgBloO75K7-Dw?pdN=e(_^rtvz6;|mQc;!9+d~D<D+kXuQYgDvP<y8WqbuJfC3y%Di
zl?j82nQFzvWaEk9iX6do1!U>b&kF!5j3>Q(x)K^E`}$k4#pdwmrfRd|ms3)b+JyiH
zXg!NJYZ59jbi{eaHAuqSdn(5ta435CJ;x!)_;nd9nO1M*jsO-S;~IOi{OTMFmL176
z=(m)O<y);oO}|=DX2q2KvQBxB%?=GnA)RmV=TYkY&h-1~)$+YEXklqn`|`lqEx$L(
z2n^d)(3S|FRl{yoPdGuS+25DV4%@xCz?4-Gt}Dy;40SCJ@A=M(!~KKqC)GoGWRY;H
z13?d}Qj4ll4~U(X`&Suy;EB35RbB28c#I)}om52Xp^ndt_UCFHms@tBi^rQP>~2*i
z3F{Z*fLP6`<wACw&JR-6S)h;{P^v%(4v(Z`XFzF*CxC`CS2Z>RZ964tlpd^n95I!`
z#byE<anH3-e8x>mGJoUTU`_y249dRH?YD=3q&#R3QhZBNdGAT3+i?lu(Q5rhm)on-
z5^aIP38eT2u!$wxK<Vkp=I+@BfcpAT%l_r<E8tvUh+*tL+s)edj0LSBR6p10{L>D2
zk%G5N{>NVD%P32Qj`X+iNCW508=tF(cEf)Z+5f@IW8?Vmyu4#Ai8$?lBKx)0-7lX-
zvmDMGRZ++Fe{1XjBrkZ^_zawfS0S2Y^6rw#-w;e0z&p-7JpUv%Ek1<?t={i;?>4Vb
z=hwxgfZ-_pl-0>Hx~KsJ<e%C<tel<hCBaOVf1iB*UIry^>_3Jhh+Dw&;`u>*TFJeN
zf?X9*^R7u%EOtYvjAF#ZQGUU_%%v^*e24bRlsnUw#PJdX2%6=VW>OB0pIYVodXiG^
zGo(%h2x3M^VjDOej_#)4KT}6xu<d(9ppA6w-&RE`()1E8UebE&+F>OHJ3MIJtC@-+
zNDLP@cK3@WLC0XUCtkOHtHD@^v?<z0FnJqaf7E!(-A16hCd7}Fw5Y$IZ9woH!e=Kp
zBcx#X`j^QtSlIm%ZZyPYn=ehs6;8njEE}-@{^IqyN67Jk^a$l?X4rrNbI6t0peic7
zlLfRe<VtJmpeIak;V1%=RlCjRY+CMrYO=Q2g4x&{2(WPIykRH7C8^>murWCKRZPUA
zg-Uf*-zu9|bZoz%J_|7r&z%&ci|3L|UvF%wbU)Pa%ztK7@|+tRMriI>$l+{sy=qWW
z2{~$x<!7O4>xFeUEt9Ck7~drcuB=dV4KKBFKzda$hpA3a39~U3)JFVN$Hm5|l(0#b
zU)92i#ZCc-Bf<iv*b~Jyac4l+JxmToBDl;uwWT><!AQ;4w7qO)d>W#0@RLZ~iz8x`
zu0xEnRsWjFqFo30s^OTYqG(Vzec@AcNo~7U>_bN}MbmObKAjFey?ifj(^(zID~g+>
z&hFkWK=6l|%c~`$%(oq{`cdoM@XjcE6hlF$(d0C`s<$MP!|-3oo~<#<<=*5-qtE@T
z5mm6Nu&XYa@swFuu|iT#doniAJcom!0`Q~z)+Cnlgb}#_mX0td+BA$TtrovosTy~z
zeLBtv*CGE78K?NvV4YN!t6i?}Lf)R5aH$lfTp*QKDaS~UDx9tLiw4v?<+5<J47f*r
zjOQ$cu)W$#v-ktG7E;ctiE`{eag&d8-db#uJiLSmIMsMLobv*%d2S-4H4T_|Z?%}m
z8q7Po^$>|X59uGPsD|j}o}pCO1ec*!dn?MI$re649uP`so>woD)5N#koc(*8c=B7!
z&H|<Qe{F@*7Au@NtD;j1v>n;|YRpygQLi`6Z^*#9{+&(CbbxeqnIXFZmmL0edgJ@P
zx=QEw{^>?y-`Bec`J^VOg*N!m-K7~dSjalr`i>W@iWd})j5H|ZC@#3Kn!#Gf;;nlF
zl+(n>TDM|X!2%7L&w`!ZT`borj#5W@K&W61VBTfh5!ZgEc*w-UKBrNrFj}OFEmQbc
zDe#%oaZsL@)=4+HJxV%je%Biuzyq=a=RgTOQ6$5ZpqO+|^d9b|J^YVXcqB`uXKp3T
zj*gSR$IMApai^9Aph?wS3#@>$KmxNY>~;Ex1uAVuPe27U4T$2LA_7QDlvOH48opjn
zZ(Y{ZaGnCbXpC<Z=zgiUZfQkrW=R|awFl{2-fm<BF1Pz1T2HlLv2v6<fw`IJLWgbj
zh-QxUjfUK-t$6IBzu%j#>vxztpwz;c%_yYHBg&Tx77d2_RODxpHv?m<+ot5`gb9Zc
zOo1Ua<%}h^{vkTsM2Hw7D1Q2f?2HIcP1h@YJqSwnSlTW-t?n0%eEvNUGRlBIhoaG(
zx=2X3H~%t44W_qvF$GEwyMx%1<iwk-_0$PhtodGw)-_A{e=+vX&zS{GxAtUW+qP}n
zc5=tIZQC{`wr$(CZF7>Gc|O&tI?q@2?&?2a|FUbZ?q0nvX{#huN!1R@4lMqH^2p#;
z<NC=#|9^C9+#4Ighn|CTr1!YsAOLwHc82S#w+X^U$`W-1n?-Df+RU{IEwFL56HQQ~
zK~6&LRRLf=c|73M!_v1?rk!Y7g$WI`L;Q?bexcfRGR3)9!F!xPtzIkU^_6xHpL1vf
zGZ39%IVE`po-*f>N(CBcgSW1T$p#`Bh3YGkhm^k+WOgMv*ITNjBg~-BTt<3;F;ddL
zUr2^go$?j{(*rx{k0$21SqF2{k!}LIs~9AFFJ8IXx5A*GvBIUF!Ph2?lIhQi2;WOc
zLmV<}&E*2wfDjvwk_T<ULRON$4Wiqk<yRU%LZbam77vy!TnNK%h;<=LR#idid9qp)
zh{1N8dTC+hg%)|g(*-sBD>EE?`KT@k#T>Gk-NYST_Y(6mY8`0!Vmi<BIG|zb_b;oh
zs^+}<)Oc}4QQ9fiLZwq^?&sQ*8!Y-}EcbB1Kt!l%d)?!9v@fe^`s4a@0fOW5bALj7
zwz+-5eBfZ}VX%;xM-kVue)jsFd<2UymCj~4$>(SD@g^g3fZ7*#Y${FDmVGL)#7<`&
zm0O4b?!HRV;>6OsO>DJ}pf<MbU0q6kM2DyOd<*&$aH*bD+P*;*gMj{pO@}x1l{GP2
zAtQZ`n-kQaHcUsVs+6fvAN0jTLccI~^14>6;wopdu1`ptabR^TW>F=lpZcUMI+N{>
zScs0Xm_7*r)fE{{xc6gYq9%eM+D<{mGq+IC%d3#<$kMfT{^;+>Rn-{)YEm;wNL~=a
ze)U`7)DA`Wtnb3w&cpGINv=g6CG2tNqch4<c?c~aHKH}#9kx&0O<|-tcpFx%gOlBm
zs`OcQ(<)tw*G4%B4iT$sTj~L^=J*nes{Ricxlg)^wbqVU#h;kTQMXm47{JaWT$iQz
z4tog}7-;c)F2%9{293#p49Epzf6DkCph6JVs6w#XsA3-M6(*2OUv_BFj%Jxc^bDb?
zzTF)Q!Hti5S-?xRPKuT7pG+$k6XLT=vNDfbSZQ66FJ}+qZ?|=}%Iz=NoMPS>_%1zO
zP^okp*-bZBunXLihcOEVns%78yS&xaBYR>bx-29iKoy<;Z3b7=tAr{#MfJsFNi&A0
z%m87l22z7+<iMOnoKtF}Z<Uj!XvknHZwJH$FSX8(U|3<Ls(`>ag4>R#skpkjiUOrs
z*7--JOto$QMSQPVD3K)v;&EmmenVikrSg8uv6mSA1Cw6mk%JxfoH>m+w-CoCFKwT0
zA>YSn;WU&*gWVeWgr#&f_Vhv^^lWf-*{9y7)ZbkN1k)_z?;bTIL5ze+pM{z(?5Z+r
z_I4`<)xqOQg)==N%P$MpR;H+&C?4x&Qi{T?Tiya%N^HhRkv~{(KH1T=yV^?wg&-q4
zgU=oH@l8(W@`gj=_CW@e>MAj;$0tV>DZ80ZYUcPS?c={_b)wEDTx&MBRV^Q-3dLkX
zZG8$l_pv%Beb2F2Iyw+mA2t=H?k5f`7|bS~HJF4eH@5WzVpY0a{K9s^G33AEN8!CD
zV*5OY^d@5mez*D3yKuY!heC(`5(bsBGP`K)sJUZM#+WhKp!w^0Xf8iZXh$Li7d(GQ
z{8;4NEc}76{)}Xj!}iw`@`LWGuO=-y^xad^BBwTOssFF1WQKm1ZlcFq?_|<9Mym|Y
zL1pV$5=tOd`E-m2(zCBa?oi2QyQ&iU%M=f!ZVI^pfskx3qb$_hv@>FMX%u*S#gr|T
zA@r3e%LkW0t<XM%=P@4fQ-zD3ja@H!Iqyc}(qdqZw%Q!f@bwZ)nT`)#m%3dx{_JXc
zzp2M&o8$cB_>%Ok4jkpfk)pA9XqcjL6nSsJ7*uioR3Vmu9-r{oPo#;<mo6|r5G9KK
zBbC@P&jjIR4BWtMRq`!|5`UY|_l4gC5(0z11_mbc3&=}APLKA7=4!K_exPY&B-28t
zFVu>xBRg_1CvrC7q92g+SNh|Blg<C)68{h00~0&r|7i7p{ci=1%?|S~plu4jt^m9o
z908*_$30+2>8kqd4@Ac~p%%;eAHUwEAgmZu>p1o-7Dyz0BQ0S?6jC3O-UOT-SjV<a
zoEvs}IWRDC9;(u?&oIG%hUXt@$rv>7{}`TOU*iUA>5G?Azu}(1m9T{UVtc$khPGEI
z<dLAx<BF0uY;emKM$JEzHFx5J`EFXyF@4y3&NxjmyUBHQEzz5=qu3fhufPA{Q9^q;
z{`49*yAVi_YuQKuFF*(o{~^4Nr-5A(dQ_-&)vCK|NRWJ>2{T4d{KCj1XbqgMj{g{1
zQ2^}6KNC%ZzEm-;`%Q16RVyG>Zrhext3+{itkPz1v^><srWu;HJ*3GV`F<|Mp@Hj4
zQ~uLOXf_oscVoF)Pn9&nok1;QA~hSba%soh&HV{^>ez=Iqg<WlQIub*3G%70Mgx@m
zpk<lvskVG!x>_4h7FiMchjLN)Zrio1uUYU}>m8CPTzhvf4#@caiUaPv81mAvgQNVR
zVbfW<R$gG;1v-S)qS;qaq{@C)(ZBHB)r&w$+{T3&S1^?-Cj9UZA(B${Zp6Hr5!K#x
z3rW&690L~aw-R46xF4DKL}A|DT9>HGGjUY#G^7#yE!plEC1=U1n)?u)1#Vi@E*diX
z!rs$i$G!NjbFbhCOrCp>VkM<7;hYBw&PR#Q0GYzgu>^>R@7L9rGbaObR9F`9&)g6e
zwdUmqa32iud)x6vk^zu-VSZt2GC2}!0P#GqWGDyEEOftKuebB-TPcMwKSTHpf1yGs
zUw0b*7>EfF)-L+q>NyrOoMZ@)SO$Oe2_i`Q(TzLTHGkKpm^t?9TO+Id6p-qy8H>$M
ziZNgoS!>}Xj+WH>oz|G;3N8_*j!Z;9QQBg=zmi|=>VGkRHg;*Gy1#MQ;0I4+pUoJa
z;`pkQT*q^y{2j%b>5j>4KtY8-uVVKgv}c-Hka0yrp3z`rr=iJFeJTr=#~r1a4Qz>Y
zssgeqEOb6G!JVC^=_tj8he%lox@6i)BDET=MsMx#rF-3$E^FSWJX;VQe_hv!nt}Ao
zETu5q%7D-x&`c-T_cD*!$2Ky=3zEcjIgfD_8xLgtVPu?j;g)-uwgeZYA&_ai$;Xq$
z8Wli*tLHkAiR^QtSh&L25|q+GBL+l!xH3w;-nT`3H~4?^y;?j#sw2F)Xu`}Qc0{z6
zBGr=Lw;wcH+r)<<xf;4NI{Gynci!4aGCwL;9y^3CNlWu_xmk>l9#2Ytug*n|QwX%V
zV5S(wPh%X=w1rcGoj+9%nUbDYaV6*WXFTx#MgZ<0N>5Y+xjuqOGnxd+8$E#(VE+=2
zt}bk{v7S&5o>$V3aNn^Cf-L9li&3~fHwQjeSN^MG`z7{=EB&neG{6!y38@vWBFdQ-
zFp<GpY2blldmbH&+=-ZapQMiovygo)r4jl|g<M@DwwWdv(6=OX<%kL)(cxVyeN+2X
zK%&;(gO2JA=*<@J1)8yNx0glO-{O;8`<rzc0CClhaAmrbnlhIfN~w}96+YjZ9Df?L
zsDaU9RKj)4G6^B2!}^Fp7^}eOkmFhP(lTtTSNSRZ&E5}LZMtDgN9sYWi2|3Ony6IL
zA%GZ+{rRMn*r4=lfUOkrrFQnsV!62Z_oSQA5C$Lhw~Kn{>`LdRn^~k3ve`zQTbF{n
z4)5^sRI7sP-lIZN&b!<wq({G~#UxkEle8#UuX`-f`#{k!%d@2f0oP_l_NqvcHmCB9
z<-q8&uk4Zxr=l-s$RTS-sxK>nC)?M6)A?#@)FG`o0-5?~+HWO9@CmnQSn4%?Z>G;v
zs84l@stZ)slJRAuj;bjqhTst|Ak(cs$d=KofZmBK%h`RFH=Wl}<5d>3hjNwlKf)Hx
zUotA`gw9U@USiZ03vV&1=-~^#`qKm%<g?55r<pk;FiN}TkEnU0ZFluEM&t}U3qJK|
zpa#;XpdtbzXd@eG{{-!7IufQ+4Jj<Twv3lU8*RK4!0(vv{^PH`haP3RSqT)7Cu(Ss
zR9`i^TE>0qgNLS4?WI~;EX~X)F`V77V)w(;$YziBry}DczvGYGxLqrpZxY{Gr3S(D
zdrU~aqIf#o;i}(j5(rUhSx}ZLtcpcsKsOrLWd1r@<o>*_HSt>D=9z1ux3O!5*350c
zT%krLt9L1Em4CGm@<7`n4}=>JP;Hr>lOuag+e?QzWv`_Tn=6SXX++I8K^MU=8_Emm
zi7glWbtDXvYSX=Bjf96~4ee2<+E2oTOfnfd*MMmtr6eOje?YXU-2OaS88+a{s<4S*
z!N_Z3mum=3osn?F$x_fEn=eA&*t$S3B$7LvqF>HU|14cYqFqzZg7c9sRxe9q3VXX!
zm@qlFOBW>=`sH&;BP4M&vaT8Z>frirqe~_TS?S2y65<Uxl3Jn1{-+nR#zWJ9X`AIe
z7hke^q7iuO32VBD=vY`O7A0#&GQh6(j-_gKLR#2+G}0_-qSMGAfOoc#!`RWMyhNci
zgKnK%qIs8v70|_#5}S7G(&~-#+pDv?NL6p(qe~1r`q{&@d%Fe~`|Q#+=e?zUr75=3
z%KrXE#q#K)iNpnSMcv+I!pO7JDWh^z-n|2`cH4Bkgkou#snN2|1@x$8wqg$ZT0)`O
zM6%#$H)Z_>D)ze3iHGT2>onYnCq?w38*z#`i}h*dS$sot%hA5qI^$_W17z1W?`gwh
zL&)Fg`%qGG+Uh|z?IOVLk!PIAt`+tO(Ri?NC{k-HWZZS1`2vmBwQRb7F|3TKTq=6k
zLoQQZQ@@||zR)zSBD-Hysq$yqjp(t~OrIw$d?K7isv=~*(#a^wbKQE4O8J*ni89rn
zF3sY`u)iNH=V_e86TJogp%s<mUVX6v<Tfp-+Rx87dXE5*_vWKqQWfOiS)GAlYJ<bT
zkx@PNg62_H+VgpdBcRLvZHJ4vR}F`U-B%Kl{xh%Li`#ZRlh^NjZkPVycnU^-5ksVq
zvUI^<jPU3Z%il>e(|%t(YU=+@EV2Hdi+o0Qmj79rZHmUUEeSb7&#dwjG>o+1J-KFK
zwAX5?qtlUAvas=R)*#nENS4PJiUKIhVy*+tK3;y=Zd`!?y&ifu58h24d>$SRouDXq
zKh10{^jvVsPrk)mC+RN=I|C12E#L0N4R_1eB?T+1a?lUl^Gi8#=md&BMPe%dMMi=<
zEUPlX;#LuZmBsq0c{}WkS%Ae<hP$YZ(21+5t^3FI`)QEruo<n7fGCg1<P@x=G|3E{
z2s4T9%#rTNmC!kA&4lQJ;rs*_KCh3@g6X?p!;6_OCX2n8hm9F*!xIr9`$N(sn|Fz+
z<z&Ym;1rtHB68iynE#^+lXbc&@6!9mOjo3UvCgG@0i)!PMdT!8cBhdkO8|W53^QcP
zqWvE((^3<}11cF+LhrI_3OBUYj+3jCvu4LJF?4Vaw)jtz9f)brJR1Hc#mT0Ym?2qu
z>p;tkNXommeYD|3whKA3p!zDIl5icrK*yKeQgHNI8`Rsb-`7|W!+?48$(OcS0S?^x
zL!D=tB;!u?$Xf>^z4Ms4vFeR+mE7j8o<H3&6*BjsBgl4Pi}QPAa8@yJaSg7tEh5F^
zK_i>%L7s1sK6H<mEx6AMIaAWLB5>Sd9WDjK;nO`O22R&_Ve}%1G9?I2!=_QDa#o9n
zQwvzT(U_@};kQhUT}vmmqsAIZh~wQQNpGg`IWk_@IpSD7l@aYX)44bmuhLaaTy%5#
zT}*SYJ^9mU&O)tSu@YTvVcFO?Kf2Um1&Vd3Izpce4~{|T9^awpn#8$?I7^tVX}nb3
zN599R;TylVJ3X58Gn5;RxdnN0#WlC39p_M>ZIXEMW6$5ll2~#{u?5Ajo;Qq_V%#h@
zdh-Wk^yo?LuD6DP@|k)}LeOB(FuUG@nLQ>C4$|%=j;lN!NcMH4|3z}9bv2MZKYoJr
z++b~hGEQ~slWQ@h7^iXM$+#SGO>h*F;zWY#l0%SV{bhQ<FOq%li50?WmpUQDBUMvv
zWH4??Ov<dG?O-@6i^-KOIE8T;e(<r4F+7GrUHn`riixwI{UK`A$ITQ=OTGO~SkM6E
zMY~W33cMQH7Uf=JbWMPEPsyJ*+N{m%)Wft+)S27YB0E0b9!a%qTj|&;S%8k^;}s6b
z0-kO}E{X@Vio)92<HXj-rQpqYg+Q=N#ktV%M$*A=I(MuuaG+@6Q_|pwQIQ)h3bK{r
zOAp%AyPvC5a-{=WAAtVa%MhdE<AOAjwu{|_7_M<Ta!Tfy%~>PGkYMgcMJ>1?b#k(%
zoP|25MV`_<Ht3t$tJ}U7$iDGMq+4Ln14#2Kzy5k;d+O!F@6x~Ef|UvdA0J=Glg|Y<
z%+u&r3e11sk3M+1XST<tNdT~tUv}DISp<-E?HxwS)2jr8lRR{>A$2Z2_MnlTq~o&t
z*6a0cb-md*MT@K|mF&3}!uGsijZ&i%FTv-UTx_)BHm9y9M$A#>y+vMBzOI84aOHrq
z2hEVLc~70*dgw*o+E6SF1(<E$-6#j4q6d@d>wgjInIfz5u<bg*=Y4;Fd@N4h=9{nc
zXCyP^M+c##wQjgLONtB(Kg4`G8<Gn*2$D*19G#TOOvr;a)WLSqfU^h&xHw>ZrHUsF
zgpa^PK|ctQp#TlWpT%>wC?>qx;Gk4@gc~X=6TdQ{>t=N}>8c&u+8XIeSS;sF?z?dE
z)T`&P6DEo1V3N!GdFg_NAb^2)KNDg@6hL;nlY$B@aTgYL+sskm^8y+eD)Iw-SuLhb
zF)^=F2@gndj3s&=-?)9#t&0=BMGCYFTw-r>2!&DBI+q8(UMI0@Kcy=1!t9QnK#Wk4
zCgOsDP$AlfDvk5p;Ka$$EjiOSs!hv;DvG*s*%%2E<R|l$ny=SlygurLf0@sAz;hV;
zwqfE3T7dn^B;MitGYn?r7;CHklR~eLzG|#RXIHJAJSmAuD@pJ8o7=9o03Dv&B}5yV
zYSqOHH1Hitr8xtS(EU2Yr1pz|!bO0V?^*@?EoO(i*eyjha!QJsA0G6xAG(cNw5A;L
z3(y?ALXi(_5E;U^t08K_s4Y_GIT9^TZmVUZW@Wif^H}{*wE}<a**)MmjWi4%S}3Nm
zj^hwWHIjeE!q@u7!9$hqS)7%Z=4~kt**t+At8!w0uPtQNw<AqTrjDQYf~WZRT0FU=
z96Q#x{1{Vos#MF3R(EAi64do&I)*~)kTR;T%hF1cQZk=^0PcGez@rP~v(+#0`u#V)
zkH_cLbAQ`x>nusT-LqWV1pdH2E4;z=slF@LsR;RLHNeWEjl8RyWlPyocF-^7XF$H^
z2dFUu`oEDgS=s*E4FTi-Y>?25Ic}@Pepgew2OqxAq>V>_k&?_IU`Odq?MVSc!aF3I
zChXTM>==r}$Gfif3)DRUkCof3RdF|2G0!cOPhP$gK8}$3(j}Zy^!Z6u^Y8H9&k9M7
zk-|L3HSG6OfiEw{VF9E40C+ScUS9Anvu|ivcUUw6+g=i$uh&OLxcD`bzy=YX&_aAZ
zGJUXIVA*gf1J7aU;j5tM{9g7t3U+D{?xGFRe$X2l-piYTt;wk%;jw<wrg2)}i^F-O
zw}?zpI2hATp^0_9OtZSX{f)p=+6Z*IA5HwZLe-`NT7wJ0+SU>5DO^CZ{<)oaAU|`r
zCx8K`SJ1aid!<9z@ca2mg%Fb*L8k#8W+T*?K<Z8cjRqwS*Utj2yflFWf*XAUxAENU
z55~lXCs1Clt5p_vr@2wLr|<e#^;{``<;5Cz;a6hNhy1dU4ez(ISSbXgq&dD}>;*o^
z-@Db;T7KRiTh@XvErFQd*MT|>o))!m3``Ta6}R39&Y@RH4W5jAzjH})6rgt<M#mdB
z0;DMf#~$9iIK_6S&$^LzwXU5i@~1>|>nTDk`Ati9j|nH)+ElhXD2}TY2)7<0-=bA@
zCy944QR=@HI%_RoL#nV1qag(yUHSIRWH2tZR+_P_AJhQpaoNz}X2jU0l4m#Sa>YHs
zDu*e*0dpof3<KR4TOBjrL{BHgMZ+j>Wl=3$y@p{aHSP;JtY>uuxbW`m(*Hu}Pi)BW
zcJuTg?y(De(DGvDcDQN@MViC>^`fbpI0EcU$bOrdPsrpR&aBoM(rF_ar&mJ|bka+O
zFQRNhPxo|f;;dZwzrM7vYSc(mbLeH$yHRV8&sKp5aZjNxK#@0yj2nZ&(FQHcpi!!0
zDxxDD_fp!SPTB8yk2(*4KL1u`Hrtu-&G`Z48C;%vTieec#EB3apUGp;%uMi_bgSV=
zy&Q5-$yamO2$FP#XO;KLc+avG>?^4$BrngmVA=DkwrJo>3=a7$m2)y^FTUvx^luXO
zIlNfXmoHp!=GWbBoyI87xb0LMM&328bsSYx|BRVZwOTTeUi)gj#`3FfneiDedpz_7
zRLhDGBgLE`YdsXpVy?W^zOp$#&P3UkO&AZsnypJZV-l_N$9Xm4U+PpDt$0pN2pS~L
zNgpBbWH103F_;}X)y`^2qRDV+NYwm&h|>RsnML1wps$V4pH!Yg)T)U&6KZdMP3T>f
zz!u>uniuRbLI1jBcA%hjizOCUkP?_R47Y})*;cxS41QQKs^NDacgxeY(8(yeRR%`p
z?4^Dv4R%IMyuEB+C2I%$j1AhV5n<R>dW0Ny@DElQe3{L@#lKS(rlo*X=eA^Y;$KZM
zk+bm_L>N_89M@Sog&aoZgrFmumu+GbCz}Ug9dd@y&f&lx%~~E72p-;6-i{DfPQxO<
z1!UBk355Kn+l<3jEvUJTrIN+5oR8O>z#Z<&9@hSDp~CKNHaBebyqqU%UHOn_4VowD
zLQi6l2f?H)Qi7&;Lv<BwI4P)l%B1hmjwf}NIk{aA-)jq2)py>igqte<D+YeD&@jK>
zwD{_xdA+T_d*yf?_Ler~{Cn7#o0Jz>TAkSSVr~Sl(B;Smo;ja%*P;=bOMG50iLP}s
zk1>v^O00AWZGR@lI!EMNvUamoj!%L61;8y<m&)1yRKfSU<NY{qN$qLJPm;`8t5vVr
zma;&NOA;Cd3rz+ThEwxQN36<`Ge>D=ekd@d2;F5wKD-0G66H~KL3q-JX_={p*1M1H
zlh_nJ`|w10{Li}~J;AIR)F{IEylWvwjW=&LFPKzlH5UX~a{D7DSBZ!7naoC(T#>}7
z2sh?R4gQN`O5DK*cu#j4=a*t>gH=ZTlt<!+P-<od^ZK5m>0y`;jBTcc2UM?+!-MGy
zk=c`B2m530gX3%_N1niARXul3Yb{}%j=(*3F3T~TYvd}AjaRL4m0<Cwdlo3?0eQsG
zbc>`zH}ah!b0??AP7n2Y<yPgq&D_HK_B^kk{QR|{h7C%G$|+kI3_bmv19RX|(*SC@
zPz`b99c2O5trk&*K^LMTo>S$$Rq-waMckJ#ltql{??@RCt%9avM9TK?$|;p!G95@N
z11Q^nxKtCG#^NjW|0qMDd1`@1NiUi>1oXrk;PQ|J?dNa9JV)y!U-~E-5FG|y5PtF_
zRckeSq#2o1>Aq9PUy<P*k4T>cKW)?;8_Kgj!)kNx498S_s1<o&H&Z?y6)=i??k>!I
z*$8G1B4=&>0NK2>H2t^6kp2IhqBAox{?E)O%h;2##O>GrNesb=A-PcXH|Tq5LLZ|I
z`VZ!0J^Lm8>qnR(B9TJvefb)fAnd|YT@pp|o;G#o%dFI`w>yV#r$?{vx<G12r1SI5
z$Lua-q*v^4aL;WX1GVpp?|bdbsJ0o_J?G950KA8K{}b%y_Xb9`4{>-uP&$IsBWm{Y
z^zF=GFgt?X#Z<+-WAhoqfX2Wm%z1$N;PQ4ly1T<dxOIao4L$K@f?l8ukyXcI!9RUb
z*k2KDJX6R@z=yZ$hTiq?L#P*K$7Uj<&4m)pqjVi0Mz}(^)EW1j!QYBG<_|SGx5+cH
zfc3UD49DHe=%b70MMd56T0b*o^}a-a-(ynadS=@l*+UQv`toek|E6Ss#i2QHAauuF
zY*I}6BV}Uhhj#(BqfP_G78Kgm!>`e3mwxIqlRa?=tp+{Fq&jXtlE9)8c}W%hfV#A1
z-s1htM7`K1a|pE?2~+jLr%{|_<8}KvfxOCdAD;gS*~@UfY<Y_7mq-Z91+n@HX0r3c
zwAi%%-u<PxDrQ>H@~q0yHR;d2Xu$1Va46J?zW!(w25BAlaJE=jdF-`4KgD^8B^uyN
z9)FX}nR!09XOh<0cO)leIeIFQ&F?0|{<>~uV!e7V<j_KIDs5SDOjK^syOz634lI`0
z%v$^Nl{l4(+icFC<A#(s04d!m$b!`XKfE9k*kO}X4L?CkU6xcGIRyo3i;sn#Nzztl
zP>OT@**<|h-*xguX4i6!)~IO=3K>@cI&4UNiELS_23HG$=jvAgtH=MvS=A9#CiAa1
z)L%)i`+#oyryF=UOa?98qBBTSc&zKtZfm?W8X)%~!SKOzk8i?c=|2xXdWzW0)-EgB
zaW4<=RQ59E*gqA-?omzCV?OSSETq4Nn~VAC2P;=ltM441K*s1iZnUsNX=m!p;MldI
zKBP4y4s}eNvwTm2%!sB!v;H`AK(ZjiF{L%p)(C5)OGd|p3>vS<UXZ;o;k+>vY#JIg
z0h}+u29bhUAj_==Y~qM<#(@AP=T*7R#W$g?E6>mr&~wB0B3A5z+Mxud8(ns>WRL5Y
z1aI1T97t%j-3Vxz{X>Dd+OPy05Oi3mHGIoo6jq--Y|5X_$eCsTR;8pSHlTj$U|h>G
zR9)Dgdz9Hk8nBadtHy%xF&6HLGzdATtTtkHw+>7bBSSg!sUEJ&-r480J@ZF!*-=ao
z6NLUL05OE8z2Z0<?NcUr2DF1o6AU=B<|904iLi*9N=D5Bav4)IBfWPS_auXIC*w0x
zeQvpkWGF0YLe)QyU;2RXRTwMWWgPM^%TSG|_^T-x<BLnCoY|Um!LEW$14aE#TYijz
zBa%sMPD}|w=uKak;rn_qCO$YNhR-(C>2%}e!1<}B-@|@s3C!-7MS8D1HMUxlDNx>K
zD8t7ta<96m15cHW#ltMyL(6(PMKi|!-I~nG)X88H7`6WJlzY518!P6fkFZhZb9#H}
zsH4nJM3xMd|L_o@L1pdd;rW$&b|X|XA|}^Ft5|<qCV~F?VyNDI*c!{*&eBh@BBTnb
zo|P-4FjP*IP<;NwdLpk7$>vWYEhmhA@nDm+QZ8j5Y;l7HWmn*b*0+!D=VId=R`6k%
zzw+fhbYnr*jx>IT_h}j1_{JyG655&fe3AKf`*nS}O|xY&R6un4ech+(aWFtJG_VQM
zBDnf}d%b>c?vkqL(omrZh3$P6^<h+!$T;%2!wh>l(q>>g*H>}<q^VL#PosHVXf2M~
zGz{4>`YUv+FF8&IBJ1b;`{E%n29fdb0^Roe6g}5bh`tmKvQJC#hR=Dm4%hT|kwioP
z(XXhB1i%KF;;^t)(7|+yoLOost#ouMtU`Q78|Sg6_0=S3Vz8*n^6or9QmAD^WGrtF
zP)KbtVbc;9Xi9RGhIDHzp5!Xl&tbPwGV9Wglku7{z=;n(eeW&OlJyKz+hHAG&VZG~
zC((Yx;Ou4zH1X7y;7USC`78`wxE|-%o>w(Q;Zpe@q1YB+J7Zy)lyfD7hF0E}TKzb%
zWV<qB*|0>7U8Za5Q500&4M>1RY1-V}$R&{KL%}+TCf5ufXM>t{CNP7YivB$lTl;Mw
zQY?BwF&&l(tU~XX)QQm5FfJ(){#T1cU<wuY(*!A%6UoXK0WS#-eV(_EsD4jMz15z2
zJ$*5Jlh7ndsE@*2<)AuGwrIZ@^6WXavf<js!S**5x<JKGwWYAIb>|A-wHB?Zs{Y|b
ztB?wu$w7(JvU!^8)957@nVh_DBCzfgE%=qQ(pZSxJG|-N*_7558HYX9w3|7yfHZ{K
z98i^2xNV%N#Ls*c%VVbIYk7l#r{Li+G$cn>slQVuLt;b=&3H*WsDi-Lr-~4A23-w8
z{nqsOi38HHz6nC`sNq!ZYyC(mnqn6SVH3KPUVPHsnq7b`OVUT9W+2Xfi^OPbh5Uan
z^6jcxSM54s@Ia|ne>y$|LHkg|Asn@V9Lo>MQIt>9rYLk<Vhj8r`;|NaL4mJ70G*;J
zkUf3;$h~H466z`HE6ZT6eHd9W&a;Y@&CeSrEkq1SoogepDD{1Una3{J-ti2Ht<Vg(
ziQZ`eIy*AbzOQ?{?1|oX`sE*so%$q|kwsVoUz59giI5DbCe$;)jGO~CKQ>?I;Gb$j
zuTC-jdHH66_-VElArl3GyB&B=X72g$2$U)x1<*&e=SMSbpJ&@;NNz;ZP@JV*(aD6L
zMa6aZ(?zUOgscAO%616koc;Ag1!xDRUVVTOkoho5{MD)@YRxwQO#Q0YtKD}M7_m<w
z`j-e2xzZHzGIYxtgn08t4a*Mb8Bz1lyX(BA77U`PGBNH-<~{aLqq#-_GtxjB&iEr7
zLIV#5b$uIAuD&;!@$fyfli}I=Xdb}rM4s(8TGS_oGknNiw`LH)k$QptO62-=!cLod
zmGly@e`Cc{RrjnZ@nZ(%i^eQ0aWgM08{S|afR}}OXxat;XUX#^TlC}yylOc^D5TYe
z#Jy(rIyVJ!5=wamaN2l78fwBl4PpDJosaTniRaw|_Zg7<C#EXQz1iP&>*8&%l3_|?
zJ*CW)!;O=I2ga*jSxo3WWahAgyi5`Zz;Bj@H-S^SpjXzG(M^%wfS2O1ba`6%tVpIU
zvK!K)>vZOVOoiKG#8V-$bU2n}LXVDWJV|=SwSA7;V9uSGUA)aUO9K$+w;p)c&u?{;
znR2q_`k3opug1Y=6yxR#RWWzgS23Lq+(xZxNW*=^L|R%gX4>dnXrk5a*{x!V<%=`R
zN~lSFmYLBDGCquI`1U%XH=Fi12+;81bB|=7^e9+=dd6{_r@%*#FoMT<klR%C_{2Nb
zHCjPXQ{dOyV6LNRt&;Kz_XK&Art%v5{q$5Ta|JR)PUYn%$Ql>u6R=oqiMRYS1ml46
z%7V9Zm|At;rnK+~4Yk@ZurE6C6;g&lK;hq%n7NAD>gb8t<?)9HDqL@CTx|<U;TENJ
z=<HJLOqtP_T2nH+2E85dt~})~hwZV%SI}|l5)`TaTK}We&wh;D)XZXSmap!_H5J96
z5wF0fxCe}z{7PL>i&6>Db+zi{O8~#u=vusc-UzyHE4-_;f;U`7U&aGj(gS$Kdwci1
z5D@TG@0QY^evkYCkNwGO|GB<_3c{L_(y{mg^0x|GZkyKf&!)!Lf)I#HF}YPGQ#vS>
zpFV$n5nbhnfog?cn%@rtLCVf6y8meBBv@9HG1VNHPGN*tiAtc(6UFaqo|z?4hjn@n
zdi6nO=b0293B$-%r)_nxn1zK9<Y_?Zd12h$5>uPF_kBj~r}}hJY!Zi(97Hsz)OiBU
zNUopIN1$~*Xg>_9%pq*wbxe5JX#CqZiI+X{L49+|xXC@)5z!)rh~c-^c6U-ekGBMv
zesha%Zof&GrroVgVRQ1j7+{n_;*z_zE_7?`J=}z?@))!B$;&VU)lHr|@1I2}+mB%<
zt+~5Wt%iC=TV(aIb8Dox=$}Gw=VR_I;;7d}5GU_|c7x3uRN=z%GGoVVE;#<i)6qtf
zDwlkv7g~HmoH-F^mkV&|vZ;hnqTe&&8YK^8&kVDJ+EU>%CY1jXV2LEszZJ~>7ZJTm
zXl2?F6d)0}@%8oZu#`GEj{m>49UT89MVs}1NLQT3p0qvWc*w22gJ%h&LDLh^CvYW)
zowC7hr+sN!TVKR+{r;&NE*(oW)@)fw4?0LEihA}<NG<)zJvtwqJbpZGg^wX8?^1-5
z%k#@d>5hL=m&9s1IZ8%St6V(IEA{J<f=tbY=A|fJdE)rkT`cAFTbx*gF**CASUj_f
zQ?dH+?Wp$BoK<@kQCIT>`OMu90ceLfj={qw%=vkK|BRx6aJs4z7G%J7-M{kZOUTC>
z7(@UO=AMXqWKTYq#bGoHcPHErNH@zb?e-Z$sQ)%o$u-*9kO)ZA|NLD&XGw_iai_#k
z2U|kRUCw5Bk$cnJQ+NufD@Py#vjg@X<HmNvE^UDqWNHHAKXU+<OXsBSbB8CVjK`~u
z;rkNp%8mQ}E7)<coLFOC8|)A(EgUA4F}^hPgS?HOQL8`+gE%TUV+DvP3tnMvCx;?F
zx)rSehQ|_jz^L{PQY2;ylSxP@{~POWs*`=54@wxJ4vLObCRLXthFmAPS#z8-(UIZO
zV+qt@DSc9{f5S~x@$?wYwjpNg*ga$Uk>SRMyJJ$_&K5p*t8r>lIsTz(50wyBRH<1T
z7&8!!&%f6CI-B-0y)F?uk|go-bt>&0kVzt|y`0<hOWyD6^l5aGpJS<jR!IVo_=A!0
z?1u5KsjU-Vq_O}o0#h$F-WVjRPS@5^DKS>SM<y0zWR(SD43mhrR#`+m$SenBY>}8E
z$H+SqOTMih@cb?F9D$T(ljIYWXf5y*g~9Mp_#s(=A?D{Ul@c?_IZ#-BeWU3#s10!p
zsP!M<jD$xaaXl2gAXBkt=2;?wBT&rGF!&08wD?H;&RQyqB|&d15LP{yfG&1U8av}6
z3(-+JtX8llYV|Qk*_c@u+$QHnsq44^hQ;8W&)JHvCG*yRP}qTp$~_EMU944x5L$Cq
zdJX=~a73@)awy{Cl)D^C_cStg!EzXTYi2{%itnZa>riC!j~h(8DOSH~weKAK(ql5K
z1kiNVE};io1}!JANjL<9sZ~}=IG@0oj!<)+G$p&8z}Md|xhsYbsv*VV^e!Lvd1Qw|
zt(eYu%gU#BudSGsgCsckB*3$k5~L*m4bt-fH`{KhH3AWqDjW@oF0@pTo=$e@C>(!G
zDT-HZ-~-j@bC4#^VC8Xu+?3)Y1-SkITn)7{Y+qioF0Pkia5~uMkhNiFA~3qMXb{)K
zj&0ZLSC2_mJ@t|RuwjhRJ;j@PMthz7X51ikxXeA6-1Kbsc+T%&9+DW9L5Y7`!nv@T
zo<K~NdyQ!!52OuW4Jq7!%~?|kO+*+jOp7-ngSilL6OM;a0Sui3xuUls4@nCRI{O|S
z!5?(;VZvm--4omd8pr2ir>`)UxS6<H4)UjB4;{E0jT&D~Tw?}1-&Bk9tL7GbQ))D+
zlR<kX)1|jgl`ERWBwkGjMHW9B;G~FIT8OZBCod*GTT@xD4{U~Ag5_+xc_u$J{xX6U
zNgQT9YojxYZvba$MMblOs@C8afApaI&L)#leNTeG5Cc=M3?@Yxku`J=q1!mkj1~9F
z%2XInTFY>rOpP!r5if|{nk!C+O=@Ax-m)o=`P?D#U9MEzk%!-OzfeA!rO>cI1FbaT
zQ14^x<<m=0)v+g-lVxq&f4}}p*yI*M<Q$@fPD=AoK7;tE5Y-@!4x}&VK&)a&YKEPj
zPIXFsq3San)}#ueui(<jXF=h}U6VX2vs700HR0K%RPIr;IZ1t#YF_W9h3#4>=-H_%
zy6CvJlq@CLcF4{vOLeYG9z4VB6|T{z(iN*wPRnsnHw&dp!s0Y4B{R}yz3-^ip4>{A
zSe8$@R055Rd?2b0`^ItJ=vaspPtp1^tH9Ii)c47O>L~)?zw2DwNsU5_Sm<$|l)gJG
zb-BW6r93djbto}Se9Qc);+ZJwv)C^>se1pNw15>1&b?`I;lQ3?TJq^VP*{H}DJ|n9
zl~j;=RCT^Asn5mVZo&m2HrA);HQ+d>)I#hac11BnbheT`yDOE#Z(*RETY_MUFyBxk
zosX5jUggm2r{D!f)7pE<vENP?2o(^xRdrnZ8zLX(Ufmh7sgwfhC8`s`JMFLo0_H(b
z2L#NeFsiS7^O#Dl6MO^A;M#eJJPLKms1&@SW1<o9VVa7=%-ec+QPsZ2iWARD`{gK}
zHm}<3uf{d0IWnOP?hzKgOX;aA2b7<N-AInDeapSRc729)twOk5+X{hsY5D`Bp;V(W
zh^Kh1v37pvU=Lrw9*GM2pK5zS8<dvJgTz(khPcmiYFThF$Rj3;6m;vCMq4uVIhYH;
zxn4@t;&*mdd9tFwi(GAIM*Gl(RJo{vJBz$nnHCSGSt)TpUopksxvZ}9?E%hWol$o%
z2wjD)&LkRHPN?al#?xh|)#c6kHp)QY3Hm@r<<pZoI++*-Z}8U2s8MyjHtEX`@K<HM
zLx`Z<j)%65Tp)$t*Er;+NNN!R5jgH2Fs_3U0m@&;xGeD8j9?an5%=q(7&Jj026fpl
zM%fq|6Fi+;-@fEv_aG7>g&s^o+Xv$upSPDoH^h}(jBKCg<R!cr{vT50Bd%-bLu(g7
zIcpHOlV82G<RkU~2gCa|$pQ4H0zpln?e?L^YI88f)Nk=Cw`5=lC-X*3W9!$$G?uKC
zBq8}y{pX$G&1-DJJ!Nor?Y~nY2vm(t4KVtCalop60qlO#%_ymGhTXmU8(OP58+daC
z!b&PkjAD&^CG)bGsF(4G?giyfFIt^UyNnr&Z5wf9nwZmq!j*Bht+@$-A*bg3rsVQT
z*FbvUtNRp&F7oJKVB#Fle-vG1$0C5wJ6tTZSJaF;2)y<iGqGw|2*y=G`3v2MK@8i=
zhGOGZ&zPBBJm1M$Z86^$_$mI_nGUE-7iQZ+YgfR!b$=h6l~~Q1HavYt)??eswBLMq
zb_TFM)8AgI%+xTrk;8bTeNQll?*DGGtUep9VK1EX?^-@5+(COojgc#^o&$sL+L|K3
zo6o1V;0@|ZSZQ33im8{LX`K)a^U~-Tmk&V+cx8>Po`j6Cy#Mzv^Z=h7)p7gq&fRA;
zrxe$s!<Aw~oK=7F>bPs?4g({`nR**B?u0kYXE@^&35gS4N>73WyZ=N+7$@{(4{A#+
zC4vaEEYWJ1_=Oh-<W;vaA%eOUzl^fd%wSuSwOPHDS=iIcVuzkuG`w@^RbSL47RG5;
ztK)0^Wbu*of+yxFuiE`fRKdCC$7O2<TEK+FM393t?tUbiPSWJZs>Qsp9#zL(O?~pn
zW`aE=q9n5aiVkk^<Vyt{LF~O3bN(bMBDU9oS@{~@@0fxdzx{48GB6VE%WlDvDuw>k
z!WTu^m#c31VlJd=O*;L`e70h2+4=eQ6FHhuXxzTw$N`3sR6>L`I&+OHx{-lt>I`Z*
z%Y@{o$oleUkz>2?zi3Prj{nW`^#4l?<Y4{(h=FXJEdL|(_%cA#4tpHY_qEoXfIjQj
z#d%`~_nNi7Ru9^x1cM1d1R(|rYw8FYr?w--=I7<3s<N=8wB{%b>Mfr}4d#oY&Z+t`
zvJSx?5of~4<R5vVtg&EpuDdMr5Y-_O6gh0NCi@Yl{gAkKa-<?;WHE$u@p*&#_}@Du
zQ{gg0B#HNbM<A=MOC<iqpP&nPTuFm{Qtz;9D!?^%M+#}pCXaxe64%DL?O5P}#|Aa?
z$U)*7<^Bm#T)Bs#k3LEAFbS@YO>Y=*&oSdLpg6|FKu<cwggRxtP$c=?&jae4<X*Be
zV2Pc<luuL}6;j2R!3`eiu8tZ!E%ZnRdC8)zj~F)QWni#ygJ>#Z0DPRnNSAPe31tYE
z5tv=Q&&x2~J_U4iGsc)V0*Q`V3l1p+)m2LfME76|*vlKmTk&_yIQ>O6`5Ws+BXcz1
zL30A3q?BhyA7{8LGAK|-AF4twz{K>ARf>DIav+S=;N~x?NAQ)gdP;X0;##I-f$A3t
z26)4pKXOp-$uofyfJcCWqc>cgAZ2e5N06E~TqF_3G;<ed%IxKRCGp9G7vdO0X|4!F
z#OpH<Te`^vV5w5?J}6=es6OGiohP{JnAbiS;xQbkQaqbwfZoveA^4oqDv_L2cVSR6
zF|rHzFPTM)6FjsW*(Tn8&M_=64aIvGaBQ|35(EZ%1CB5PWd;wNZh9S}h0LOq$vx-`
z`mqN(g=pn6+gza<8RL%vG|oi5VSkhwUAj2`w+4=PPRj82T@g+0%3?`28?@`?FE>Bs
zE=u%gM?NZaYWY+3+Q~!9x1#R{x%~wFUBV3EsXU{SaI~8!MgFfL>f4^E%C`HGw^zUK
zC|t>gRsmka4xBjnUPdbY<S=e`$jL4iCI5a0Ui_)hLM4q{f;zj!<i~P`%26}DO!Oer
zwPc}MU7pc|x>2Xefe5PGhirK}YXEPZKL^PXxH+|WrgyNRpKctO$z#phu8%HGvR<ZB
zY;B5$Gd5Dx2^^A>1L?)+qD41=B3JrO2kuWk{A$*TV1c{RY}}-Wt~@RpeU$pJGS8k$
z*qYx`*COErNTGedq&u!qRsy0Rh>R4G*IO9|O&pf)U7F=bxqFwc?~=-M69$UiZr6by
ze>U5+_5EuBM|tMraRSbA(!rr4(<Ud~Y-HMUR2uNR!(a6}&p+eltC_*DdW9F#zyQ_3
zFH%CBFHM!4r;>`rZB-w_s3Kd3*s%D)Zuv`gxh(6_#SSf*!v+5pmbw&`T&|ca;x6eG
z2hqe=6g={lUC6kk@2|q&!kI2A5}ZO+46<Z1F`gmjYQZnp*o#Olq}tMNa`T}1B3M%{
zU}VtF`zazdAYg>&RXcL_?;h`-`1QpcSgYd0#%=x!k)>XpvpcuEem~B)dYyK8cw^)D
zqG#=T&LS)#Os&Bb5x^M6{{S_Hq7f=SKfnyDiS6GCvuR1TGMsOI{_j`~3si8+3|1rv
zgHt6-LFM!3bs4KJ;ywS!lD}I8_ZNU-fTksZ?@U;rr(gCeRT&p?Q8SFl5D>Ka)4=Eg
zTF26LD=>Pg?N$WhqG$5u_w)^jkHEHKutYL-`5qn}wPD>Z=L`DsjzRVdvh^efb41Rn
zmjn#3_@)Uo8{@cpb<Au}OxC#`8R6z)Ob#Uy_q|I<(M3bMrovY(C3v>-ju#zylG5Jr
zX0j|<=+HrloRwweUHMepH_A)6g&EudGj6HSe>JbBCN;auGCjYa0mSVCqtB-cm0`>6
zYm54XlG1eK3IYTDJis336u4+{plbLi^cO)LYpDD*6W+PKKzpE9Ju@-dV7swxwXpyb
z1tFR)9~3C@a)Y_nIOjUT(%XiTph3rRmrsc#HG_LNV#G2O%XWS=ygEx@`1CbUc#X~x
z;4KOc`Hy;1fWH3!Q#|^B|GTF2%a9ouju6}wU%8WU*BH|51xCs)3?NQAj~S2yMS5N(
z%2PX8N2FmppCXJekQi(u9hMF2Mb6Sw^M2Nolh!d>p5h9R$Gu`vA4|ARBLKe_Jn%_g
z_o9EFG?4IsmQ(QJoN&cO4OjN>T{pgh&+6X(JiP5`)`uW#*7kEX1#T=k9zsDX6Gsdx
znJm0~<}+N+$PNKxA(>ef2@5fpa%x==oSL|3)72QG!|)301|$&NM_=a5K6`s>KOL=-
zkwE8*g78Ca@jx=%!)bJ7_}OGsU-ZbFgOL52lo-!%`+_VnWcS<Dj3(3c7I*n7rzjoH
zI3L9!M-sxc>;jKOKnrAEA{1eaO~A)Yj0syr)|{1Ra_i7w89lrxTl;ZV`fJk4Lp}N&
zT@R8tOMVnQ9D9Q?Ib&4yhr|Zr62q$PSor>PSfei`{u1OaEZT5SQ2XM+YvG$P@G4~l
z*=^45icTA^PK*b~YD-6qs|cZ|$K8KL=2a}YHESDmv%GWf2{e;o9|v>|4gEi5+a{nj
z{EmJ@h8vHS)+EKaqry~tO;$_0XOElB;-9{ebYw+ol<Zta#fr!?yYp+PP#5_AM)emI
zV~@V2TPs#X-c3V$yXj*m5cFE;Uc(Y$^brwBxVf=&H_M_*(xT^I)Xr?xsE1b{!ybhX
z{${>ff!7-!g<JE%_3X<z@Z-)}JFyD>t@1C^{sS&lnjP(6471K*L;*3O`bIdCtq0YQ
z0ANWq&5uS+Y8&T&-s*;+owEb3NIMnvh`!eH{{IwfncZTTU6kT3TH(^ax(8H(rfQD~
zT55ydqyJ8Y=3GQJS{l?#TRtjsSBk-`-j0{Q>8F?gD@~l6Cki0}>Kg&ujG8x4o<-0K
zv$mf+uj7x4O@Kv$`{l6brJkHlr9h8Yc^+m7U?t_C)y-M-YAoTeEhAWScXm#EwpCPk
z^W9$B4K)+E={rB%H&l8(jfZCFBf&*hVzR`G=@_ssAFYe??m+EivCscCd-}jup{hPz
z;rlx9?C{l&adl%G!0j$8fWd@T3=*knySs+q<-opQa`RM1H9DMMu6RL{X&LPA$%HS5
z_QmO{O?t=Mne=W%XuIg%idzkzn$!&+q?ZOUr?rQImux+?%8Anjt3CFr{1=nQOCfG{
z!h=jNdC^ffhE0iL+ws%OqYoDM+)B*jEnW5nj`)4)Z~4mLOR(4RbzKBg&}Cw)aL3r@
z(u;Sw?C_s|bqx*fBrDfrc0Zur_YV~2ASbDq@0bkc4ttYd-SfwtaJSsOKDOwz7FrQk
z(Q)8B+&t<uzB2tY>-lPNoE0{QS*LH#Y_No@VnnquWfQMwM^MzjIWdC&Nm^GdX3PMI
zh}kBbilQ`#6@VGl`&U!*LeHurpMbB?bYPJ8t+3IKaX(24q}>uj3uN>kk%*xm)b{%X
zZfytpGy|NR0N&^j40?quO`Bn?asK^R!d%K^?cS~9yNij+=AJ{>KCYN$VG-PUf3I2x
zk}jQ3hx2YY5<B0qQCd51!P}!<<_aQdc~m&kh7>`|J#h;oS&g6L7crDxmN0_N@Tqko
z87T=E?LK0bt&oTsqTgLv4s>6`(ykDu4U*r-FWsuc)cS;kLYY9qqUl^1o?s-uq@a%E
zU#8hw&8FYry$>|X|3&YzG5l|O_y5fVX8A{i`M(}|F$-&F6GwV6YXfH!5fdXjV-r3;
zC?{t}69XG4_l@EI97MJw_FSqz;bUZqqUre)3?+(#IX)WNM)87(UEy32bb77z=27)z
z8oOOG7D|GhC1es+M(UwnF4LCYCS~*8#?p7fo5ds{@AxY(PPRgxflK}fF66T&0VAjl
zK0P>nxX>3jIOCoRQ8t%M&b!Iyo!k}mFTyZ5&r|g6j}u-!-i`vV`7q<383M>ZRE~4a
zL11Icy^ErJ!TsK@k8im%KvRh{z9;Yxw;v~OUbiWF>M0Puo}N!*{RCizyjX!S463re
z-%g)wUQI}}GBA0kf4<-NV2OK;L3j}Y|IH7%KqA1$7&NhiPx6og6X%$n&rgF<fhZFY
z1il^lcr?e;t7r0WA6WJXBQ!tEjFIZG2@(0TgS6A+6E_?h&Oj{Hj|Sz3a#(`RI^-nj
zKyg>6aSKVb7>PxJBMQbp83+~sO}6^HqHsKNrzz&c_I)~8W_r`UcJnib`I=GByc3|d
z4ii8~Dm_W!dHdrb7;U2y9I9DOwBv_qe-7w66421Y2DmWNf4L&fBy~er5+-<Oi4~!#
zRalY_pVafB!mlRYAI<adDaaBJJGI=3pM{I{#p3i?Q&3tsC!g8?3TPopr-Y<)%sP&W
zsUt_^b<_8%bfZ{8y&Hjah-2itWca?WL>Y3{%l28NBxd;b0+EI(o5s^b^pc;EzTi)o
zhYv$Oj>9FAK?R;|apv#Fm`sCH?okXR?3_Jx6m+A%m98zvB{3U<Kuft_84r8v@##Uo
zm>44Du$cd)DMgN=fSQJVcFmh_m<%0Vr?t;3etuNipp>nyh+4*qKl3@n5G!zIrtWkN
z%Otf4X{x&?+cPQ;Tk?g(%Ab!d^i2Sm?&QPB3#a)MX@Mr!Z2jmaNfx=__BK9g+?=UP
zuMoCit?HN`L-r*R^ea1~0gf4tGMQo%zt;25_jDM{oyoW|4Cg-buYBtdwPL}r9uh72
zFhLwCUXV$Rx3;Nnw(;=1v#Zczw*=|8gXEKb(<eIvC*89M#);K`H9fLYM^Qdk3=io|
zm}_e-kaHepmWg9|WWlPsZv8oQu=YE3CZ6FerB-_501m>x+!**@jGaSpXyLZ4W81cE
zJK5omZQDDxZQHhO+qP{xdAU`$cTd$joz?$aYs@+37+>A1e`ezy-6={{Fx4NOkw8dg
z1T<(JF;1{}a5G)%J+so2Mt^5GWA-$WY>&x{TZF9HSb3d?9#Ek02ByFtP!Qab5rJ4-
zwrpgO6B$>kF;Vn9LI|;a1;OVJ5oCT@d$ub@aNFV@GCZ*D#3rR&`%oZ=&XcW>AJ3oX
zmYN>`LFn&8RHYe2IE2qf<}Ho=yx90g{IhH(bbq411VG9ug`lK`2PD;%6~MYBxK!i<
z$dwPL@OZF5yi`jx(SHgL!qUEZ6iD^_of>OD=HSrsOyKOrl%cu}soZFH=HKhXZQm?z
zCX_UuYRq4(*S*Tjz@&X1&9fRU^??2Wh#7kWo8dw=rbCxRirU2ss^MF+X@H}W7NCv&
zP|@QUzqpqdUYJaXH|#NqG(rV2De;Zm!2$$#k%`AEV~rDW>K6m(uaXfCq~~V!_<W_U
zc$FyUWXy=1guyc_=!-CJLpRp#A$R~|=dsJuH%JYBx`slE@g~i-Z6u<QOk0Xs5l9kw
zrePs2=;RzvdADH_<OZfo`dt7-d3OT_!{z`P0h(-34)bCVQxxjZfd#9fJYtLhUE9Ll
zyFHr>!nH@OnqO-S!cE}RNd6ngv}>n!v+0UpWM*;8wPsoqQ8lji-WEP85_M;7fQs1U
z-KA1t^ikoR4)S(>m|72%%18;gWM0XaUKE;2yU6T1DHhEbYoH-B*&<yDRY;y=bC7jq
zgFulAWmcxUG?&U;m@lP8mB1cknA>+Qn)A+CPbW>GGWLw8G8wWt#<nJ5uOuoD8z|q5
z=uAelcs_3&)qI}xn;%w+t@)ZqIrd~jm<qKh)Bc;fVcgZV$h+G7Y;nb-`;XOBSQ;E{
zz`thf^ItcIsy`}XmL(x`i~8R<{k#@~w&O+{skIoC>t<^tx^f6Bhuw>{f6E84i1hb&
zQg3utCISOsY?}|476~5TE}uTKSR4tHoGXe6j#ID~Q&cii<BklSvc;X^3GC2R687Qo
zOY&IUhB_P0h5Degj?gHj5x>ppecd-ClPuw;mS$R497^A7B_~bRDS|J&@rTUrGh3H{
z*p=FOD>K+U<H;(Nm{HmRfMUMNRaWK!#qu(Ya)4e+4K?Nnf0K?#%V5k^X`3p7yBZQ%
zpUFY!B_xi;@FeM9+Sg^}{U-9($&?96b=N|w9BJ~60(g_&pSQsk(SlSs_)-*nf%=zD
z6jm~J?^1QI1E{N&X*TD;b2tZWnGcxeonstqJ$DuQk7Ac}t{OG5qg->@0$&S5Tlfic
zfn-1yZ=xo9jrU}3go{gtLoxVEIIbR8mv}rX2rFUDOK76HKHP$YsDUsq2%<D(DNx$T
zh|yhZES}&nOXb)t{6-N{Pu5g8PSDon!uSnAXJQ&?f`q1L6!2~@&2w8lP27K&OAO?B
zk6>|=O=SkGx}GneM`!Y|cnp2OTRrJbi?Pj}pu)X&DdM}I8Qcageea9oR*#WSuG2ip
ze}FIjfKbI>wzExkk?L(nNsr|OwvD{r>~96$qR7#OxB`HesuHM;Ba;R$UCb}*-5xmu
z?fVr(VF5vMZMQ_rtPYASrvITM=s?@)7sBPBeK}UcMS#Zg%@bOP(P8TEjtqJq=5O2J
zNn-9%bYN0UsB#qeGau#b4tNm8X>a%b30Z$?J&}V^ntZr;;=gv^0`>6T0vlOdGq~aQ
zT_Z|1vkkiSgb>>{qW?sbU49kHV1p>?cp;5`>&l4&4IT$&XOEqH4!?W=$eMlN2DopO
zQPs?YZSF2G5go$MTzG*hS)ddh@<E)oF@l}Wh7Q5Z${^dMIsaEzj2oXX1D12W=xwuG
z#7(+DIyJOua@gYsimQS9_hVTl4q-{?H!Pxa*W5<y;Dc8dD2!peK1pmSfKT?5qo7w+
zDx#j-hB(kzEVOh0^g(-ZODTKiSeGYVP9ne2EV0rPeT%^>v2UPv$aZ>07f!i=4PoQT
z2P#9saB#_}a)N?dh<3VJ6{{tMqO+$f<;EgTwDqR}j|{(MKJw667tu(0wjZ=pxpLRg
z!T3*g1mD5pdsHeRmG{rfH}Ei8_w4a9U<1Mi?ZD-#Vt=L2Uw?}fU(X;NR`P20{tz6-
zsd9~uT4$=2^E)`DgXx%&^9`>R%YZ2q3gqWaZ#ZbN!+Ltw2%Ot$e-cwFUSyaCr1X@i
zYO=IVf@Eic<iCHnaMCpjIwvH$0nan@Io?Y#9&I^vyc7O?dj4i7&|{@CZvK-3{6)6>
zpGpjl-+0S^r2w%vY!2G*wsiCGAyQtT5ip{<#(=yw0y<H>DA2a925jaW-e-4iU^o*;
z<BTSv<B<jG{v<FDw868+55x$<>B5z)43wZ7h7jZ{3c}ayfP?ajzt)hzxU+*41+rVX
zezAUD$xB}CarGG&v4B+r>=1jl9|XuFnoxkFM^ZU?La6Z0-;@;G$4q&GqW5zRlvKX}
z5M!buj3aB^c1Vu&^n5=Ngy~VnkJP;O-DK;+_uzB9ZHwm>;dOs;?ZPip=;z~EA<3uh
z#rVMO!0U1@BBlhA9_1tq;PZpA-D5K{Na*E0Zns7INdl=}?P{Zd0U!fknNs31^mg}x
zX0-H*SFMtH0l*042x(#3vjc_lcZoEGF7<02aHzoM*Z0bRS337vgzN~SOrt`@#`EBD
zNs5*-LRgx3#N|Q&5C}H`OwD&6PTciLp>%Phy7m*zocDAV*c`S%rYuVoBAB*jqf;)B
zL16>uc){yj`~iZ1i3f5sA3JmaR!R}EGY4DB_v<xIc0p~l)0lC!E7s;@{`<r9ZBy9C
zZ;|&0G8k&Cm5BC5gv*ZZUGAxuPZ@IldryWRP20YE+i&%xvn_z_AP6)}b-NV`zmUVA
zd?&H%5MfyUiYS#+9JtzyinL*FEM5hdKrNhMJUZPt6uMeXe%Y;SslB+(4acijswIcj
zrP5E8uNadR;!Gs9LKSiZ{Gz_GKJjjv2MIOa`BvyY<x#yX;_AH<e5=a#nA2sklrXu~
z*KR8i+pTc8!*$YTE6WV8lJ;PbR`%AZgX$wm>CUvXq`jb%YYs})KZ3fwl}!GkFTHY~
zXV8viPPr4y!i&CWy5fvU6K%~t3Xl{T%&Obna~|zO={fBxo38$JiNFe1e9JBrL!(MG
z`_z%;+5WI!7&4o-h1r?R=_&1VG^$St9GdLXm4Q{EZ{|)ozwAK<#L_0b7&QQ=z?^;F
zog_RPCL$s#80}_t<5H65S%;#AlSP`P00{H`2RqdviFrA+^vlB`SU=6JYlCA!SUB29
z2fUt#{j9Q?Mj;F9=1&~8;=xVyXCxNdT5;&LuOPEGz2p-Z7R5eusbtk5xK1E~$^0xR
z|GgydO^Lcj7|8)_x)Uw8F~SMw;nXKKYk5c~kuYfVB>0eNMv>NYpicYV$<wB1_!^cM
z0XgKrUyJ_!H0qD=bxLtvj#^qwj)I@7cRggI)3YpGSmF3lW{Udz*(2x-M|zPs7VuBe
z2Ia9Zkm@e*c9W)9mr|6~LoeiU!|Bob?dP^I=zigTOGBi8R3{Jn;-g!sVw;n}W`R&)
z!Xq^0$5@Ki8Rd_H7(?FcaZ>LoC`Q6@hI6(nuvLUYmJ@^zpVjvWED{F!qC7kayDXT_
znWq$=g`7DfpYO+~ay*d2oy;Drt!U5*m}uHdLi8xq)R%?%nS${ytxx0s@B<Z&yHFzh
z!3PZH&6L?WQ}Mz*x_2^PtLgEk?K&~grlZ0%?-!a1=MuvGo2#ihfVW%g)W)f0|Lm=(
z2+ZZD7htq>=fArf24%yMmr|tnEs36<rRFAEX3=|6?illaLly@ff6_m6VxRGFa;D%x
z*V}*Sk2z<@h~aE$bIJ)4M8va==dDkfY3-qNz!ns96clS4y}L^+<7$P8l8&_!4=u|9
zQOYb>gl9S6vc1Wc!$k(~y9`eyA=pNuF|A|Xtguidqf{phg|p<1vMCndh}TU&O`75}
zN>EoS!mM3}-ng^1U>Q=xK4oy87-nYxJ;C&Hi|uMK3jFoeU1Ux9o98<<o!?E=W_gy~
zd0>=5r6MD(#@PMOpiAV#pjquag4@3%PLE{Xs&GP_J`-BXQQOh_WV!2zMfbq4WTR||
zL~!c0{usg^g>2Hb8r|QYkhV*>9_>oGI0~nc3F>Bw2psod)olqeEjF;p{rT^{cL{^g
zb1IvRsfUtMGK1p-a_SRH6b)uAmvxKAKmL!aWX}-PliblU6?FJA6Qdo!<W0bU#(kbn
zR3xMdQs=8Ho>5$@{E6MUpm`}rugHheYrTX=nDWwS&*$sufXzNKI0MlHC9T8V{D;*Q
zR<+n>mnzHAf6y;E%*vy~kn~?PiqMHx?N{J`l|t-8Gnmd*W^5#{w|h0qyCcD@8|<-d
z>G5#x1B761`#kMVgku~_hC}d9gB4e>!*5`&;6`wi`6>(ws~RN+Nlbvo#0G4|H@Ngv
zxv(tG#!Gp7%E|<Vr||sSyUO%)xW1;M;CE*ol-6tR_THSHXVGFxqhwxZpI-pUyfN>s
z0Q3|b2Y!NiV9>yPT>18C?NGcO`P?^$_ZED-J%1)ze0-Z>`M)C+kMszU<tgEM#;YEL
zgbvvUWZPpfE`=~|)r0PwZ;@uGZ!<H|qJejU1A>4wG0`*TayX#OO&V80nCYfxGJyZ$
zAhHN2fdz6zt9bHn^hquS65W`S0N)@$RSjvX>8kG0>RkO^LsWdT;&j7@)C?-0cs=Vf
zTG&aSPUbgBX({)Y11xXaHN&Na%EFo<w{b!>dKfou-EfqGh9bR{6hP#Pok~G&w4QLy
zUv|I&Rzl^$Ey?nA_k%a9Vx;BQ+Cg$N6SP2-=7+axa`1TlJVz{l`6}zC&j6hWniE$(
z3;<`dm|q~O*pS=V*VXwKudFQvDB0?lK2oYBS4IE=IO>K3XV%w*m>cS{GYjV_GGnLq
zd07pB)Osnm<Iy*{F7uX>^!7#lbuIT|D?$e!A1Y!gwl60;Y8+Z_!v?0Sa(Gggwpv;F
zCQMiKh&msWu|lRwMMb`l6Ec;yH0LW;c4h0#;!>ANS#(@Y?9w@_?fVOGf6;AuW?&m1
z#HFZC2mSK*pV@mtXXh$2$U+cchEeSJG~Gh%tbPzCY)(gQ$uPka&oM1?6wz0O-&GO3
zk?Kn>4T-CS14k^?A{=?*S6egAgADC<NrM=>p8vhrwDWUwNXrJ-kLTcGR70a|W(v>3
zA#n2V$2i>d{bg3R;Bqfd0-x?}A>GOqqKwLU2ChT$xPFxy1$<lPVvHIe@S1q-v2wur
zQjF#g^pu`+O9Y%NF^~!oubWFU#zj`PyzeDP@+Zpx64`w)EI-X#<rtTHb3#$%S#9Pg
z%G&JG^{jm{ZeEladB@!<2x;&8+w}{FWk<h@CGug~glEi@B5C>CD1d3XC5&q|T@J?*
zI7)BxF1EZ0IGRv^5sbT=qvC)Kb};iJO*Wi`>^npktc7P0c_oU?m3o?@iA#%n*IE9x
zW1wyxn!6}Pc98Q%3t`5t#u1&I37yNQu|!)!`kefEf3#Eq%cmArguyZt@(Eq*Hc{Dd
zESUr@{^Egu<9K9=L_)oY3~^$*KLt}n_>QdWX-cn!0Ta%-eW|GRDSD5wbDW6%a9f|7
z{z=i98$Y|2F@dyq)5(EWU^oJ_^<ipOrbeKAP41#0<%MIqE6IquDJxeqxNT#T+`NTM
zUh&9L!`DLBc8pjoHN+?`CZJik99A_&JzKW%*I?!<|1TpB?%dP8V^15yylEJ(i$5kt
zM(C~DG_um%<^$+YldX%+qq2lx5Y{@osFs)Jz~2~sjh2c6%YRn5UzPIlBpS!8eua};
za=6;%!i|(^ma-h(w5z1=D8M$mZ{^ce8M>T?nd$C?stu@D0u?5S)r|y~0+f>|jPAh<
z6GUetDsa(2FcO0;bm7NiUzJ1Zd0GZWUb#(?wiQBfDZ{{fmzJVGj7Q_vTdQ6CgFBK;
z(yP!ns;W%&9j;w2jAgyt9@8aFB8(4G+3D-g&lVO;Or_Z*=yA=7+<|MzC1axobX^}o
zZ6#T}yy@Y4Kfnr1d^G<_els)wU-`|(`QMeSEMZO9khEQ0*OcPdsuXVf&Z%{bu*Gh`
zctZbEgJEh2Nh9dsJbV$BNR(7i?%7<DG05H2wg}tLNEGMv$~k;=e|dC%nTVMGpY`j&
z+0_QfZ#~ivf%$Cu7$8K2p8E&q*OQ=R<0H>pL6K58p4bkrSNc{LkSufWtZ)H9heyet
zpO4SFpO@J&FfZbAHeX#Qa%67rABvdA)z`L<jSbyv65`KKXLcg9Btnp>->vHwKC-lD
z*8C+mrP8g*`%7XlD8+k5p`8jh5Mg57wv5mGpzinE*G7%553u0GH)_ea1T@R<&MHB-
z2|(_&1pm_}s96sZ91+5j*PK3))$F35>(;4#;nd9KAY%^So(1BH)u2JmPi}iC%s9ZL
zbQghex}b9>#<|u&xkX%Dm1SQuu;224x{AJH-_mpag(*qpi*L&-<C>3IHL7WO4&^3J
zgdoW2w@j(pe|A?q3Trx<un~xNPrVf>ljXSAOzMV5IhLY{2&&5t=~Q#8YF<ket9=5Y
zj|yR2?vFJFe{YDN3Ran<MIa<YT)1SK5=v@E+Q7q04Zqu#i^kKsrwPZ4ucVn_iN{6g
z1ed?$u;pj@-Mv-zWHYw4Y(f`il2!R-or&g``}*a>;%<CcdrsNACmZ=_21m*Cu1KFn
z5b!!At(S2h;E%HS?iwxpsE>J1hZr&bKD+y924+SZ^3H+F!8pvrDD$&ZU@k@`;br^`
zlk8gxh~H<L^O{TCe5@CO{ak7%6BX9z_-Tr2{_p@6tN8IQ+8_a^9<KaQr@m+gUIL;_
zf3S$;g$R6I<Y&Dd!~q>-cvT^|Nka9Z*s{P!F)qJ{$*A`*<JY%B9_9B?HZ%8?5mN&z
z<dBHyMB1K&5!WQpdia+{!gIc+-LMf362DHW$)*w+nBbrQ!{y(;Cpk5o$8X=OY3Eey
z7kowjsDKMEWbK*X{6EKiZ~6HN^2G2ZLVe!m$*qoK_>Nk@KOn@Zv3kxTU8CiM!96I#
zd)1J3lg0K@?if^WVEg1hEg$k#7WP|e4T*)V(ZmOQ0(FVd{g}MhFgdA>dc?l3cU+xu
z^<Q<P{VcKTYlSv<A(@f8Ab_Cl3}0_AXB%55h%^P_I{a<luQQ{Q12%`yrsaZ2r1Ki#
zZ&mkzb!ilge*1AQNg+9@A@oD~aYFsQ86=6MBD4wf@CXS+L{h|pCOT)Krf(^;H=^B>
z@-UU84#S8{<8U=x8|LGsdYM&duCoAT%_LNNY4(!ln1hRdm}IuZ1CW74-6AO{pP$l|
zBh@jK6KL&WjmKlef5h|<xf_C6+nlNWq#Fi)8YMFq!An4_%HeQ(m%owprZJg#B}6|w
zbu=1QQ$tdDx!HE+tx_k`j+y=inurMJc?dqlhHB~ALYt7a?SpRegk*w-glGeK7*bqX
zHv~KwopeI9`bI|jpd=`yWH6Sw<T~^LIE9WYbHrcLwSqdSa=w`<We%gSAjw)aLf;_`
z;7HaE>UlH#>4>0wPLz=&!OR^f;WSce+mI8Hj-073!R0h*532d1bM3;@<Dz{>toKwu
zA)Hc?oN<5k7TXcl*jwped!epgR?20r)ZMMCE=&3WhKq@!zmw*veIue+H}~8hk{Pd4
z%616t*bVwu>{JL@-jUhf-+;HzNcV9{Fwjb5kAC2^_~F&(I{>LWJ0p|1a0&1D3{EMs
zt52tJT-4{>-92MV-(r6(4e#cFZ5nmyq7EHeVx=+sfyXVxG?NmnOL<lj?;h#7+&AxM
z5nXmOr;v7wO#mygV~pUSgiYYjRsJw=DPO|@W@#G;_f$yva-l^y!Azx;jXramN<7WV
zHucPMX-Du!l#AaS_~4TI67x(TP){^8<T7pgBwfj8@VTf>ynd=FVY*Wi^6tYe5H2W8
zb>w94lm_=Xg$cT$M#6hUH!>uiB2jBplCNez&wvgRy3%l)_(ip(SDCpAX?Vr|&wWrf
zZO@^g>2dUB8Ca;Q=;3}7SZRn+3cTsS;f3M|8{=l5a0YhsNIk~V12$gMy)}s%PpqIm
zqOD!{oFvlC)1Q*m1@JS)lVSZ8m+kfFv{!(qNMg2;L3w$sH$Ls^DHZ5&ZfL9{<-02P
zO1yKw*Whdt#0q9$F)Hm86X2+C-Rq1b`~LoT!W)y|a4|>-Xp>YYa;dT4e^O5r$L0oz
zjh|8<&cv3Yz=9ms?V7bgSg67B`5e@W-^W=rhQQMoD}L0hIzbw_@Z^@AW{#6Ww7w09
zR^@Nmss~MpPJw#LWp^6mGfmJAtV=_KXl6p{YT)fywzyXooe%pOMOD}=!k{5Pby)Q6
zj!Cr4yLx9g{a7`{3+(dTC^qns-#*K((jvg@Wp`Bn%&_(*iEPuf6?FA799fIF3+6Yh
z+EnP$!#YYAEhD?j7b_AGbt;&py2c0NMY}fSi-KR{rCKx&gsVReZJUlE+O8GPr9Lmx
zlE#3Z9E}x9Oj1voq`D$j(w%n#$Fdt7cBZNSoPgq5bQjU9k&S3EWkhPKyO@NhuDkeq
znAtd`&pRJU1}z$>%W_p8G@YEd<(6{mqN!VuZ+b#iv1nBP(Nq;eR8&7&P=cC!hd%ef
zG?9!F?jk>$J767iPv%qWj(LNjDebR^zdr|ddJ`}O&|s~kGnJT@hf~D}xP?RW$QVyE
z^cWeHdX1_$>K^YkaU<snA8s4}$hVGv>RipJxv&%#%~W1tO;5$5oi}UhYHuH1#=C&1
zZ9ZN1>={G=H?OfE`p9+1O3UEg&V$sGClJKWolduFHX@MACq{oQB3d82_8qRnJB8C*
z@#^Hx9h-;-rOJ*b+;*GBQnP%{P+qu@iM|Jm!B>-sDPcOhYJJ#zo81<;RG1NZ${3#J
zMaW=<p+w0=e<e`0Ltnh-wprEl4J7UNCmzk-MHU0*9>K0LZigiEkdfw5OLO>uqefK-
zSsJae=C0hI3Z9qxqN%I50NMbJQo?V`;nn2HIroJ>Duh2TyNd*syqO9&o~;u$&>eiP
zL=Fg10zBr+n4(#v09=l8^~}}DzYo0gM-D4t;+f2}>3T;{WNT2s--mP_JR)BAF%yfm
z$b-kRReu)iQq5d?{wb@B`@ri-FIq*dqOz4%B#T+h7<Qg^6)O;=d&mtM#-=VDNbXO!
z?cI!N?x=AwYDwY7Wq~YV*k5+4v!=FH@?+4t98K%@H#H)jRbsZ89eMdyVU~(sKgeH*
zg!9qZ4{=P_pG=UmUDaR_*sNW!^7YJLuuApHkZC|Hs^FRe)B8!1VTPb8+TS=E0J!Va
zVII0x;eQszr7s-A_Y>1^x{z#cZ*6sH(SebhG3&xmDP69cs_M{0P3RpW(rFcBgmgrh
z?n_p<A=FvX=g>JVY=*$r#HFLZm*be-j(Y8z5L9olhl6IDyhj_CrWjtoA0FQl<`rxE
zKov{$5Cj~qt*SofoIt$=oS(~X*i2)p<k(zjisn??8^Gzj{9wEx%nESp5n8*#31K0k
zZ~z|JGaOsoYE5!pHR_OA$Ga7cYUG_o?s5P&Z!s^3!fd}|KhMAN+bIoe-Z2`$TgR*Q
zRF^t{#vLFhef$9&r6YFzPm1<Gbu2S-{x4to|8r8%q%jt~!GX~ALH!6k6+5&Kh5_5S
zwFP+0Cb$FUfc5v^Zvyrwgz{R!yzWQ^uAU~GYI#^An<NB30H-^f4!?x{1`GUuLR!Hw
zHv#$B@!m;=f|g*v-O3uU`1~O~(+d7Y2SVVIWvsZ--yU2UC*)N;Q*yy@Sjy*bV|iRD
zA1S8BGX-zq(c%}9rll()#AZiCN$6as-R(R-Uw0!CXu(AgY~NAj-cQf=?>B>MVv9sD
zKc759`iqig@W5u&iA)1EJ|Q<H{G4-jHl0LU;+9cqr^8f;h1iv_LL-{$>=;ZnK;{}B
zZ%>AL#5>*2Y_#?Ueu@e;cd6pr*>)6_il1R-@Gx(o@2p0|!q1FGuy5HEQ?2YeW=L&v
z429N_vh80AO4-FIErqul6zH$WgLd~|t=A`n7>IH+u5u9<3s^!yX>jpq(#t{i>K+&)
z<4(UyoIk4T#38dF<9fL%N+#ENO0_&{o$7<Mnk!Rbd)_^_>yp=p$YM+DVf$}<*JP1^
zgnTbdvNIf)ZcgHo3{$=7ehiJ4tZ3*3{f(f_H7P@X>1^kmV-!|S&D3IYMMr}qXOUV_
zB;uq%VYA+eC1#9sCyoJWt<G#;oBD$kct5~oV_ta8*d((0$(W?m^arnQpAkI1VTOfM
zQ2%P!g=jDi^prxh*-Oy*=&DmvVPrr6Bms@}y5t)xHn(+^B^u`9JXOZN-SZn&UTfkx
zHa8~TwF~3PA74ykiq8^;VthFL$nz+K;i3qe%W=Rz{T-WLj8>TV?8t0liD?@7*)&gf
zs|JZT5So{eya&dtukpV9g0hRw7FXa3xCO3a<%Y7CaAMT)X#|-~6Mp-EXW_cCWj2!N
zijeo@k81HmL!ugtr$H&#?1iWWWpE-sdK&sDG3&O<4|+i_`d&R#PHH(yD7=rW08&Zg
z)t4#9d=(GE+@Lg2Wh@7lw5drnoLQR*8Sk6bLAgQDuH^K6&>HE|(c|$e#a`<Z>B;g(
z<W+o_>g&o#f9$G~;)9KwA3gWCBPo~TXl-j4gR6G0A1>QN))rD`&N55$<TKAL*dF>N
zCQN6baxQGHfI;$4IgP)ZQ=bJ30AS%Q@Eg%jx8k2;V!N;VDV>PMzhY8yGjonU0|@6s
zh>2asAjSD;VZVdRk;7byuByuy+AB4d4$~q%H&(Tkv(L(NZ>CT)+*0JFbs936B~e(5
zu6NrcuWJr_<r0WYc}h`x!cqMpJ)7HjHWluiBxGVIZMaJnoG6>p<%f*TI}sfxA9pnB
zhKk?oJ$@`u1A{SD=OPWIR7%O$xKVD}DzEKv^~qTq{*2o?fn)PN<_sPf8JX*1?r*nA
zT>T5+|7yODqo|lsxKeMF>{iy?lEO~7!cX?ueq!tbC1_r?`rHkIoIWQ-K#>qqsydcp
zlfLbJND!Ub&BQ8GVBY>1=W1}wc$jWLHz(WJMPsQbdq|iyu=W2EQs-B6oA-U_Zj7XR
z{O3558DSNNu6aUq>?5h0E0%;5A$6E>2k9%6;KIn4lzvX_f<8#P5a~wkHP~b%p!h0%
zN8|5Q#4@E=xv7-Cd|0SN#Ap(U9d?o4TJ%rGWP-RjmAhivr4n?@RYt{KA>T(v{JB~4
z@FFycxPP^&8u_V(dO#JP)h+78z$|k)DPHlfqN@|J?Am;E?3l+}bZ8GXS0Nu^2rR?4
zFif?htG4inRU`Rin};TqNm-+w=TKnUCboxW7-)_3^r?DV(<OM_tP)2-cGcEI1n?mM
z0}9JzJPtg`x3U{fQVY|q^4~S1DdCZ1DqD;7@mNpb#kN=1^o{lBS(8F`4s+^#7`6)8
zDD2CWY2oa>nM!Ie+wY#!G&Bm{J%d%Js#HQZrG(;bbF5_ESyDVLNj&;L!Is!%=@P2a
z&z30-cQ_+D)aY^nhI=Kq{R0-=A58hKX;rNA4IA?g$v0eaC6*@C1eW)xXQu91V24u_
zlAoahM|=f#rV7{(HYuz*&_S0GVM3H>^=#OnRu9xS^wK#6T}6lWx<x{yk(wXvz=||k
z*Abt5^}V_tJN)tSBDc=ZU<wu%m|xG>qaN-VkIDav;;4+%z{n$qc3|osl%s*yFppH!
z=$klGRC?K{egm6+8r0O4EIojuCW-3M{d>Y}9yKWQ5E0Mfl;z1KiK=?(WDkzdt(g_Q
z<8-AIJ>$DPoz((=Fx3niT0Vh}yt5Gk2iB-^%o(DPpu>l^G;N&*(8>LL4=t`CCOdLr
zGWXLryzrg?|1p5?k)1nqaBC-(1rCKeaJp=GSRkkG&ZZzrc`JpiuiI>gdNjEiCkdE4
z)NQ3~88OzVMq4;^6ANyR!|Cl~Iva&(3ZNIbibPWBjxmf|#x-~58#dfbV#fgB=LXz<
zDFd=_-w$(dsMnYog*eP`0EF*%;ROHzACh&e)xV#6ymYB0psLjZu!Pkt%O8b1kC~dz
zy%AO?A@@CLwXSU?$?P><p?$qQm5C>eSD)-dnsK!+0G2)CX3lg;yhlqXGp6)d1Amy)
zDcAZp@_hr<cqc5z`ro@NtCm)Y;&NWADzIvjme)agr0f()>K+INAI2-SOC^#-o~%3{
z?Uax{ijMA*)5k#1E&x^yb9zuR9dS7$4e8#@gJyQXmGh;Wcr~{wgoteg+PgaGEgGZ1
zhqb?<!WVR^U?yWsk!Cz6!MlY30urY>p#N3t0<-f+sXpF=jNGiNS;y7k$9(ahRY{mC
zMQU{abQ`*Q2R6D7n?B&JeP;8mJT%%7{N+V|&&Cb}Z<1s)fn-wWI**`RouYwcGv?cG
z_E{An9+NzDUtyYAg&fRsUIZ?rU*zV2yFV)Q?}!Q`wraWLxg56oAOEA{k~y)cd=4Vl
zwht3-@>5VW>+@*TZ>~sDJ`62qtTN>rF|_3D@yh%Lr&lmk`%l`2_5YgX%q+ja<NwYq
zAG5*fxT~%?#YdR(0t6vU|5fMg*uTA3JQ#HjwE>+e!ruIU>c!DgvWK!sz=jJ<5{A~<
z|Ghq2A8dq#ES!Y6*h<<)yAfw@^7(9mbnf2v>@t{59S7>Egk3$vT+z#BN<rr!d$OEJ
z_Ko!I;B`+u6d>W5Kmg-opzgoTncoH8$j<ch#k~RXAuhx1)bjvG<_1Q{>FfgE41IsD
z%=mckhTe9tAF$8-wNt68ZIWSLEii!YFsEOfpgsX}p5^xL>6PI!$jAo3R}?$+a1F<}
z7uU}Mmj%0Bq?=i>y|qHNN#QVy(pO=J1A^lcvAHXOxZs?#^eq9?C(qpUwa)H>w-x@5
zo<TINA~tbqrX@zzbwwsca$#a}NR~M0+T>Wcy&EsRg5Q1Od8hoh6M<T!;10$jok%r^
zV8;N)81;%VI0j#@{UFZ;%7p7jo*RbxLQLH2*+fEI&lJvoZAQa1OTbBia!^eqW<Dh#
zP450oL=bI}_Uzz9at&4PIZv_LMJY8HOkOtuG=?e<Z-?67^p=pzVxOq4&PxH3y>jgK
z)^??XDx>&G+*xQvA4Z>`ZZ=fe67;>+3y{n#D>CMUIp(uAp7d9^-(8HyOPe;KZA)6L
z)o3|-RBG{Dg7sE?YEZlAO@*28uvPQ1l8nT*ogbPsPBOu4dGI!h!jsB|fRKV3e~vfX
zC;{W2b@C2xEBJN}<&}-ejygF~0DL{G){qLo1jt~FJg{}qwv$%4TB>Q96=0T?^Aaw;
z%dR0LnGxF^OadrHpB8jg*Ed>)lXM7UuM8knR5~<Icqiykm}{qZwj84<Lsc_9Kc4rF
zuMd9nph#>R)V9^n_I_<BQ1fy-EtGsH!qLy|b<R%r_t}K4f1kh45US@wE;7;Um6&&9
zZP4Pd)y@O+DF6hCl|LxN8v)^1E3mEF%90=bMnG~jBq-IgXwS%8Z{vzW{7~=TBkj!4
z+YRv{0-(wQ<Bl80T7o5KaTH~meZ@e3%LC%d-({=e=j;g+s>+v_$7?&UXXPdg)OTNt
z(`I!lBom^%AcXLRobS^5TYuE9t5LsuB7z3Ev62@;Tkz^tt8Wk0siYMqAxH|LuQEYm
z-aT(LK7(E^YXb>&kspYx6&M0C*?FNrv&nLEPtRws(~jN}_<hzM4^&wyo3lqFkfRW}
z8s$+C$-ku9y+<bOD}^o|>W9TjOL3A0<CCn2lRPvw08kV@8<3&D7ly>o2TL&ilbaY`
zpaD=C%!ZLkG_Km14zXkVjt8_4f~qZ7HDj7X^Ttaee-cYTjzo~r{LBR<p8h)2p!iqG
zruR$1%_5FaFfhC)vY@@jb7%=?erzLkW7zf~8!0n<mlaCxhXq$&-cMkC`2)Jtm?zqD
z9NI|(ZAtm1anvRF>bMCb)e~aUqW1P?AWvh>*APjSCIXRkkiO3Hx?v=QW-jVuoFNa9
zG_WZk39m4SE7UH)Fgmo4lNr{!bj)~^5iWuORe{WcjG?q`xXJpfA2In@K#I?ewF@7`
z=6R&qXiKDBskNbcQHBNDKz3jLmVdBf<4zMuHy#j4Yb+h37$oVG4Emucs1_kPEhko^
zNX@O1pF$NYFmt!#9i=TurRE!mD3g&HBcKAOa1Lp8|DNfm!GQTUVYbOx6UC-?HN5m^
zQJHqt9>4Bs8?4Q%j~GQN4Ro{O4lMyltJLg;u8e98vv}NRnjcH~0T$FGxzwu#=kS<H
zKTLxV??0>dm4YVYM1I#jEijmSTB#qPMv)Yt1fW=21qggIw&w8EmrE0PU?yU#(kn48
z4+xxta1$sd81xtm3s4LJ<{|gDouVLAP+RPE(k+*^9#?i`PgTg9Lpd&>pxGzkzQASC
z9RagC4$kdq<vT;np_QY5YAi8{3QbchI|(e(TC)6fv4fsT&IoV`PA##LYq;XrC(9%g
zkhsG+A2*5%F@5$kdlK!D#6(I!88juxB^k$Jr9rZll_y*rz-eY;-pjGeToXuv&zF+&
zg*|u4u`3eN2X2U(CX(bG`+vu$HD4B3!#b?dB4Saw{A+Jmgo85<scM!rPGx8oS@D{&
zy($;R&O0YC43mFw>=&$Jty(tkfoV%KC7D|0lpcsPxPjO!<;6BF4O-si64uX#JT?xl
z{!qM07DLO+*p=kcfq*7$DFk<_Z4M;eD*L5rUrED@U7}7m8W(4?(yRxcj16ODITuA)
za(YZRT4yPQuQXzV{8=cPJUa`uF-D;!j_h<W=!cD5i}TkpT@F$@l7)N{g=?2n_{EKF
z#}mx}d=7SX1&AeBG|Zba#i+?<Ya(}aO^HuQoPl|W@MR!|cb4+nqdj>*NGMWKfSi|j
zos>*m`t=tLtIxo+P)Bi77w}BX8PylBq4U~OqH8F4uK=g7hMkSQ+d5VC(E|TNMt*TO
zc%~*RYSoj|{@X~>gr`NbU&V03PMO%d`Vp4w%z;c$kVNbVBM=f*kk6(e6uz;K$aa!3
z9`jpCc-twj>j<fu!!IN_+_+G|P_|f~SNS=9sH~RmnYL5ZX2XHiVC==!KK8I<@s(g%
ziw<=~@io*kgjvd<ui^Y4t34j%u+pewx=S+mVB+?kzq!gW29X~B;5f<4+Q__z9@>{_
zpw}|a*()EAOW_wg87eS?m5xMWz=1d1af0wwwMz?3Yuzx8Bs;Bb%YH;mPl`i9^+H^I
zj|%!7E*UffUhH|AECLV-CSw#LPb_`?kPJ~aIY`-|MCd@l&a8w$Ni6SdgJ&$v$HBVL
z%YFL<tQ(ja`IyNqw_MLaDU^Z{7f*#tUN98@Ehm~qWoxW-)*^$un;pe%EG^XGYWWB1
z)uv{W#Q}SveBVJ%q#`CQ*swT>w1I-$nESEKj~U$9u7z5e$@dWt7W}^7jP&+N-_`x=
z^Y+82?+z=_iOyMbKf=?9lpQgHZQ0uA)B5B6)*d8+t&b2n5-d#y)U|oq<`8)s6bh$w
zpAD~6z8s_wL)3O$P>DoaRF=ocfp-ca!N|mH9O&2}Fu6YJt0ewla%O@=r*MYLr09$Z
zO@>Z7b%-<u7^`56idl#!#T^YN(s=n<U6HfRbzn6jbC<kub4Xak?}v1}12LW#vMLS!
zfGlo&LWgZylT4Zj;#St%zkNrRVVz;b(u8$qxJ~-b-An?5Nf%AASLH2Ob&+VoSSEG<
z<KD%sL=t2%F7{qGzXtp4Trkwmq+g(VV**KwHT90ejk0t@UOwAgTm}48_@;F_cyjp+
z^3Gws>`1;X{Fyq!h2Dohf>8aqWDWP$%bHCR*#CR_Z#mY2Fu$Gx>$^7_EUiZM0Cf7l
zi-mxhkkR}?1G>-l{9@PcO#r3I7HW!~%K%$FYkMqSFtJM5V9=`QTjhw2G%*z6FN~g_
z7?c+bt$f$0n<v$Ly^s}B@_nZf0wq<XS;;08bP$F7Z%6JUgfn1~KK+rtt1Xm~&Hyg$
z7=$+U1$${$+!I_74s##}vcDZ&wAyiwfJoN7L4C>w&S!iLi~he%_k6W@{EIyKME_H$
z<mi>SPl$=wYVk40`4^8#I>!sQ%G}{&ZX;XZMfVE=&Yg;PWCbCsJ->M~BlGB?14O6A
zea#mUyCG#UIXAK=JZ_T<XD4)$>Rf3jE!h~Gdyl0YLKpNq*CL4$I#Ek3gO~$^a@!-9
z?2;Vuin-&sJa7}us${siPH(R6X{%vwq$7Z<lGDM1s$QvweteyvL_O~5OkTm6dLG>e
z$^om!1-$j8bnn~}ywB!Z<mjUKBm?`EY#eRgbES^)*qXQ{??e=_C<4k;K*+WY{vGW;
zNo_jK7tCn_;`a=|P80W*i>_UA*4E3opJW_`@riqJJdbB$^qQw~0E~0{mq~nexDT1a
zBB%0s58{<9+$2jEddX?<{?_8xM4>XSMMp%QLdOcx&8|WOdN=W_;;iZhe<>|Awp!O0
z7hk^fpT0Qj#-dQoF@R#j`-fx->vDB4iqE5ds+r2{XBn}=CsXM5?@<GL3P{@_RPdPg
zj*+4b>)Jo6m0i-{)mz$wO@Zg>+f3u@d$SU{pH6eUSKFLqomm{7_)!b<3}s|xD*;!|
z+zhDhuIei@^VT=B)fZb-Jbs~`zdk}3I`4klpeN^E`L<yrDyGp1_*hXg+`Qkl5Fk|%
zKH--Y{G3H`A)aQapr2P@1y^X1{X#p3@nJh${%mwvvf#x_A-%71B#N-IU7HsR+U@`y
z|3*n;G9<N_nW8Yze#BG?Hbe7h%4bV`BcOA#6Z#t%TdsebpKS_P0<egK-Li*NlIrV{
zFrS&t&;Z`WAcu$gSkCnAb8vJD%HIu!YuCj)e{~s|OjUGG4KD6A%PBAPH3v}jq>eF@
z`4~r$Pgv2qG@HW()c3g?_XAtbk#QeiG*XLYUhKxAoKq&HH@MV@KP7v~d}YC>#A2nr
z#fw<z;3>9^#(+4d;x2Vm0?1APsR>8IV$)_pEMSi&^H6S&#EVUdMNcq%$lfnYTKG4J
zzvj^XbB5#}s)d&bUBE!Yt*3BEbTHo(Wz{=0i~G_+luE;!U(0m|Veki7TXXaDKh11x
z|956%V&LHXuV%I@jkf5+-)P0j?>>6KI8K?zR}grsYXL0NjWKO+^p@+N-jC1jzO=jo
z{)WZ?7)~j?BuzIJ-?E^jOpolp_r0UXyZf_z`}%%FHC8XrEMHskZ;-L5U@~Ze$l$Ys
zn6ux1T+-gfAo}l0S~uB;`@49c0&GMU;BKf2AVf9dHu)w8PH(&Y#IJ_nVsfYa(-c!D
z09GMdJ)H0J{_Y@R(zocA>=Q=!pywqfg<n_nRj>1QT)OA`4#_4LEH5h^qA+$^#^;O8
z%j?Y?MH4(>liVvd^ZVTkOQYKe+I<WyaTsBWX4H_Qe~bYH*M<~&-;^?p))&27&BI8F
z3Zs#Rcu?1RcfKEZMPBaybJhSQM^yh0X?@mM@tk;2^O{)>9h}OY=18%oI9Y-{_@`SS
zMHcA+WekFMj}+yAPukDMqvt93O^7P7zM@7yIYX%?Up}e{Z5C=Gao<WPDqo0YZPndS
zW~Ejlm*w02z;Ps1P0&*Wrt+e>SKCB0mmTmTg+zA4am$HMf|F4YAur}XA?7S-eE+Dg
ztcU4kP-e+iCdDdpC_{ZFGf)fei!&^#W{k$6WZtU$D1nP|<$DOoy{zRiE-ECD?sK(A
zSIHIN%fCPZg2gLr@4(!Z5GGz|nw!5RKoh8ziGGD>qUGa;$kp}3oPR~=(RqM0(=?S$
z{xo7uX+*@N^QO>fXyo(8tR<I=k3PetTz_Uw?&NGRJJ2i3QK^eHN|-#(dyYwUA&XV~
zM8ADA&fFa7cSJjr)C3B4*^HE%nzr@R-{WS?EGcyPI5S_et>ra*Do>!oK*Ptti&bkD
zl8#`847AsmR+W}NWg$+2@&`hPv2zgJWsf#7L*qxeVs!!N*az55?spvvhrIa)-I+uj
zbnBsOk)7xT6!Ly*$#`QV)`r_~aha%ogc46_RCBgxP9nPLN4fD}QZ+&j_8W%Owe}JW
zEe3h;dQrperD`4+(0Kt~#3ySraIrF>Pw>Z{Y)k%=aOjS@_9Jmwm^YR=K|=R?+9O@2
zKgTdL<s3e87i}l3m3w8YLjzM3<v95vpEpAf>(^XO5O)asaQ*2zM^o~TI7B>%*3m)?
ztTbflKhV*_B!9&so$7j3RE&{<mS{a#{6)kkDe~)W`kae3yk|!7tGrNDfKakZdh`Pm
zNhNlJUDaYIQk`DbsqiXKsxuPsce<AycnVy{BDt!6D!Wp>F`Z7pzS+st0@w)nvVA41
z(S7MQfuuPMoZ3sNMB`eQ7R<DrVV0pd_rid5=EB#gP+IWcSsQa-V$UX^=vmrVPd^~#
z1*4oJZ;>a;L&I`)WvhsK<mHk<wqFk7{9r5Qtro@i_;;i6gWA*iuLI%1e&AZcs4RxF
zT)sW$o`*kWmBlP6)46bDixNl8stw4>v7S&YxkIW?oUSvk60><3?~GAC;WYNj8E2H-
zS<2djzw<}z<Ow+8<7;ht(RK(}_MyA%1+w8}6skv&Ty)ZeI#q0q@%arvTeGN65<pzR
zgiO(uT<SD}erUtQ``LX9Pt%f_^GV*_1mRR!RC{M+hvf61|16D^r2%(<0cvwWtOEjZ
z=xg+HA7joj2E^Lm)ykiXs!ak|7s%08)+WjS`aCbDbj3HZ!tF`(r`UE<)HJ<@kVw|C
zaG5N&s{Omhagcl<X<g}Jyp*<qMaNsEH7dlel3`w3BaIJV6Q+|Y3>-+y7i}R`s1HyJ
z^*fJKnT&fOUbAW_G6!JkbZ<bf+v1UE)AL+N=l_1~HlhTLI3&FE7#yj(%Mweng({k^
zW+CX6kAQsB$Up*~Hj_g8Pm5Y?KU;Vj4XB&1EHx#~503*%#gf+j%D>ADGgk&i4&yu-
zj>mO{=<`jmaZ;fMUx0Gl^Jd-EZUdP;QfE8gg!{fOK#A*W+Ql3H<h|JPB_9j6mfbj>
z018^-5+}*itp_|kc56i`+K%32FvVIdCA4`$>-s8p)jZ@67J^hOkq@oqJl^1-0~K{`
zO+mSx>}RLFVnkn-ugaE_<Goh?VRm@88BKdx^NlS!S%c8pbXfvtmik&1&S2{iazh@T
zR90_ngqoekvau_Xbv_SucW0>-d*f^!U~?!r%NY&3l$B(flZ)9Om-aiO>dK_7=`=PE
z#0T*~xgJ;PoE9{2_uG_Uu|LGnS~4Q_we3-4r75sXE%!xV>^-(_PC3*?XF=OK|KpJC
zF}$vssue}1{*pic5LD!><vLlXpVIJ5fwHbqG=rtI=W%k}s^&5J+P@_KQTbrS6DcOi
zP@;1x>8&baY-uikeHM>%%b`WPJRZlLebk(FVyDk{8ER4`6KgC%+YE0eR24Fv=wLs2
z_gpijprIz;XImP_vbWYAin{>I!@KCW`Lem)vY>1EEZyV%Ht>8AIBJsrx_{%U5qD^3
zg3C;~CmR}jSa3#fZ7$Hu&MoaPYkLG&7AgTbaFruB36Oi;M5&BhE-<5i8-%Ka&g3?S
zsbs{+j8U$T_Q-NqfU3HjebEd%DgS5<R#k_)1v@qzUg4GY9D48s(B7fY$M<h{$NO7E
zWsyW2KZO`}dAZ8>vZj;pNA%A5E$8AB1UC|{*XRy9SGF8858D{C3cCMN`@ZLa$zs~;
zmL_JEIvtIQNh-1UZc*POI)M`4+If{7xJC)62-_E+k`5<ea>BX%NHMp9QUb~d7m!UM
z{{FCxNLqh#iJMff&^PR9g?9|AwMigR<3hYD_<Y3^Yav>Y<w;q)w)xAo8NT>o&hjR_
zUMff`iL-k$y=RAC=NqhmjFb95$&UX3Nnm1R`0tYSOBzzP#Q%#L4pl)ZLr1`n&iq9L
zsUNC5)<UE|)>5`_2>*CfMzO$~594hd(6aS%G>g8AcqfXV6K1@KvdLzAq)e6YDi{VW
z=qW$kF9Y<%enQ4WN7>N$GC)jCh&^mnC63=Z%)m>8jm8|^ME_h|!TKLy<-YAH_+CYk
zwo2dJwCLG@Kl$bOHXZy>{g}YS=mC#w3*}wi(D&FFA*n&x_OIO5d_KL_zco`1iOB)+
zvew{d-1;Mca4qH5^;4L2|IBq-XG^VCGmlxX!_#Km+(^scb26TiGE+p&xepSE7bjsE
zcJ|(u8^_`mE=~emz3E?tr71DY-!E+Ukq}8V2hkz2K9r_WS0VrV?CN4<s=s+lfs)-2
zBlr4TrcRRL<DarTtR$#p-REQ7h_TYIICxanl0WUs?p;FiqWTWk@!XJFB(%qLkEo<=
zjCAU+fa07?MkCEr5DFgDs7;k{&Lgn<4t0-|1aqeV^^c_Y_{HjZS<_aQw`mh(bi!X|
z(&Z=%Q7ezVAY}KuSBf%WETGGak%EgB{3BSkRm#-0kp{tqk}_(8caYqIbCq1p%Nb1z
zYmzyFh7~b&((Yb;f@)mifWT7+_hLA$jl_QjKBYc>I+)!M#~h9AnCOZ&(XFauGy_pz
zc`8OeVO!Ir&#79AnId@KyR_;FN%M`VstIi{aCxs3ARk?;S9SI}OUv%U^4E?KrJgNs
zL|6GhN~;HaBBDxS?O&Iq)DYw=0p(}CH(;i|0l8p~V&HW%FCxqUhV9US{l@ZB$k)?t
zG*Ms_h1uqh3>Z}q;dAQ^U<8V!ic8AGIn|TEfHt<SJTxQEHKN_KH!BPH#Fd(70Y^SK
zEgT$B3kJ8!#l|m)pBdq#cUMl$geNskbSKLqv1gua8WZzFQ=3=3ymGK8El}j<{-SuV
zP<T>A*gz=~@oWW{gPw8d7GYVCL6e|RTB3+8$n{~-=gN{e`NR8aLMHBPYJ4R)c%_9N
zGVx6u&Onv+djo-(A@Cw{NBkj(jGEX5eX)X`nr}Saw7xrM0_<wX5Ta~oaj{{Q$Mj0+
zZUz{?2B{zA&~iDfuo-;1NiRlDp4o`nt^<ziHYZsvR08>ks)hG*Za)OU1je(_2H*y$
zB&qtE%*QGsBcfdv&f|k7ZqNc4Av$w}Mi3%&j*sgYu0s#Q1-kB^+t5LAit<&3;3zpb
z5g-%h%5j<qlbvlwsnCy9IR=Zl*QK72;f$*Ybe{GH{$#2?n_wk+1cQuBi7X9r6hbGr
z+)S5}AVFE^{kNi#k0}|r=5NCCGcr3z<qzd;Jm9mC*Y4I+wc>RLf*Y{p{-UJ@segSm
zP-Ru%Y)H`T1(cr-g<TSYqZ$Mk!5~x}Sfv{CYRv8Q;~D32Ol@@FyS%pBbUsSp35<<=
zjyWLr@%87We<7l{ol$tDwVNY}_#XT;)9*tjnV=d$)rW=GYq$+_=#D^(!0z`ua|Jp{
z_)0eOL^=uhO18$3?8#gtFvZkA>>Gt`z|iX5b{e<Z&xAR(C=(wz7CXr}tcJSTJc!iq
z+7u|nA~HLk^R#{nEiGBXy=3tB@l~`niA_V(t%AI>Ttj)tqwc9i2Z8F6elAnha;Hef
zIjNG@Gw}-$v~HiIA{m@azF!q^`>??h5oC6ISha4-!jjr}p{p}|GUO($TdTtagNaFA
z2`-Npc5ZBYI=%N|7dn5CHAaKCvcXhd3%<%w9N_hby{ACKK^4~72qy~&>GA-kl>cpX
zA2zMU*5fQUVx#%M4rOl<QnFg%)5EW-vA!_TOyq+sR&%(#JE^4_UwDN5k}Lyt0ZE%+
z`P;l~P`!gqU%6F}RD9SnT1*}`@6O|X3_woGxCN<eD*A|E%BKEEmLouo8A)~;`!K8W
zC^gkkM=}o{Qj*ie6vPS^eseiFj0qfR4XR-hqC54Y9x<b!tTz`HjtwCC+xvb+4>AZ4
zGH%pw`1pK@!|t@LE$sS050>$8yIQAXA8|;;-?dA0V*WqI-YG`2sO{P<+g-M8+qP}n
zwr$(C^^~j2wyU~qo4tGQFZuH)-@B7HnVAPG>tr3RG4DCXxQcg<cCuCZ)@CK}Jam>j
zdNmRj7$gP(obS$aDob@hX2|(DL6LRM9H4du7%IO<RT~2f7x|ulJK|&5gPH1^-{6{|
zwMO+(O6=?PggJ_Ameq2N3T4P>_q9`-qckm#F5Z0a{5hk`yBxLbx@FW0Sj4&I3+2G4
z@QHc8aj}ZVZ1;gZpQ$7rHR7noDDk2w!m^%cxU6VBZ|%fs<Im-T_qm?^*y7QScRk>2
zpW`4;MrCriT*rJiRW5BU!L4g1Yx;azxpy|TE118jcEZ15=ET@Gc62~+d8(F{py7FO
z8a#~k`JH6fL4JnZ(m=`arx$xY^11{HeK&r=-2YL!OWNfEA)pJ<?e~Qg%%od{IzcPH
z@$>TP^QtuNDMz`(7G6|G3)Sw><-4MqKkLwX+%eRaqLFRUD^ra9w|2LBqMOTg+<+oS
zdy<uI`4`)1O^m}=8SofZ&rSWO6h1REsaj9xzM|90rT?TO0rv^SWL@*hki8<63UR#?
z^R1v}o%B&xiq{-1Hl>Zhrz!vA1iNTS!19aOiSPDh*A_-TB#A1w?yV_)h_smpMjoi+
zB#bJNe&d|r6<%Bo(&-CUk7KTKMu=d32fl&`3~t5FH`UFY$7<aEYC!_Pr1h@n4bIiX
z%jsZ^|8n+>cY#&^vycr3H)_q?TC6~q@L~rGXMvPfy?JIgUbtw;xZ><}JWb3mU_*32
z)0*|?>le6Z)0x$^3=x)|aO;s9ISizM%}g;Df7Z%$1L`N+#<Fme&STru@BRaes1JfF
z7KF+Fq(V5E{#U$&iIeTW;}rhCD&&YAu5Yg4jv!L~7LpN@m1P5ZUMG4x8b7*F3Nfb`
z!1r{Ukt(T`&et$A4FbPvB9kt8K_V7?GP!S;IDBaR+4IrL;SNR2jrMf+#ZCFI!dye@
z?8zQV8ros`?u-AcJ6+kQv)xgPDlUj-z$Fy0wT*>l6zc3K)S#RLFkH#rb>Nt6$hn81
zFKr6}oUqSJii4LrE2wV&n+g<YB!m65dN@T6mHnAXZspem=GD|rod@B7WZnSt(97R1
zXl(?6)U6w18rUd)Lh#g1oEVfn?efUg@kzW$9~~2(@XD^~$}<xPYxI!!)Ik;xy-GJ#
zPV>Pnt7AqXAghv_(?Y-{NE8-^p(T^Dymew-lHg{)*Bo9?#I4MW6{p;sq=8k@b@Cl)
za(ogcV}E#?L_XmW_R2l<w3D1kD)rIU$yY+&Sg<tUj9vcKueQC2$Qn+mwWW-;c8$5f
znY75X_nXg~Sd5ZJkY~<vy~#mzjcl}L(o(9}uRSSCPLss<$@W)C_>lLqtO-$HTes4~
zblJj2We{@7c6H{T+3{`3VM>x4)0*pMJ8^9m9d%P^>JcC8rkp?3Q`V^)#oJpidMAxJ
zQkSQXE;ZWgzHDP2@^~ZOE;+#sHme8hQVlE+G2V;vU>4OLvN(sUaSh{7g3ZDnGQOl*
zKYe$JpWMzID?}wWA{`nm7|)8nC6f3%B)j;4)zj0JwXFGB$|6%-{auZ>A>k!aK2bVf
z?7nZNp89<6;v@gEL`D(Lub>JC2a&Z$*pXDy!9yv;db@-BR4@)pNQO8?a%O9aR@Udg
zN_5KYQj|Oux*k}%tpb{$D+Jcvw^1bPHQV&pq44(3T{+mn5VC^}fYLCSD*JUT66G|p
z!Z(UTj&tBIYBP&asiunBV=$yBZW1y}4N7@$e=puzga*;%qEo}sJ16ufBxgTG7BGGc
z=@qmjE0eRJ8#pE-uXr^)UC={<w0Uo1B{Rf)3qJMV%S+?<*M}h8guI@S@ozW2O9+No
zldoXhQ~UxTv|1kRwxlaM9KfK&mIhv@OF*~yswHIEumD)UoD6FtnLltcN+LJ)HgNRN
zBsCf*k6G|Lt6Vcy;v3>d19OZh@f|Qk2eTS9TVfLTiHC1M)JJZ4(0@*Mq*T*!|5mZo
zrqnsLNA%soD~nTf=I~2WDb}TZM1)e|6yGiG5XwHMKA(&D<EtQ2e-oAM1bVR$tcSo>
z@O1TavYwxT>3%aW4_DsMgi_?iLRmqiXPqFN4%jOeNx!Btik^$^pE;w(E(nYJn4rdj
zx5%AfQx=XX$q*Xm&Jtc#7i~WZ73AUeEx=nk&pp<kwAVob*ZtK*x{cE1h<;sWKs@9N
z!x_2fyR12jevZc0iN9a<8s{ATOFpUto0TEY5qZC=i97^O^g~RJtJ4acQCd&MuM)0^
zo6;q=d_h@|rF#!a%ziqRLX$!`fa4Q#R9)ixZ#%rpn@~#_NRX?Sl_Um#Wg6Q?7D7h+
zC`W{zNpur7_)~3y^asow^<^evOptx<#V_cxy*qcF9Qi9zLLfaIs~}YMr@LZ2PP^dg
zVX9(WYhXb`m0G7b1_a#=w84S4QSLrEOHbk6Ju8n1ANR?F%stWPIV&UlS49HX;|KS^
zz!*^i1jrwzox9saK-l7JF~gHIc!DYXSJ79ELO_4-&(i|{h4GpI7v>ztQwChKC!^eh
zxD0G#2gBL~0rwBzhwLHIis);#A&^2X9>eft@Fc3H@X{x^4-rc&dh@)KvphB^TrFHS
zeL67Js{YAq*|lkvIJ%dd8k{W=T6T<F=3Gl-l*ZOc5#!~Y(KB~#K9t(2+W?Ug`q0Gi
z9!fQge9B-Y`o?A&qk8CgWhu~!Pi2Re!}YSJPh(&~Jh3FWLYk`{%PZ^U34P%RYG=%?
z-(O(d?l$4M&Mrb6<q#1j*6d1RSJ<J$$0?o}pAfBtncn#>6-S^DmRrY)Uzxqmk~9Io
z9G&h(AmL9JaA(opR~45SPg+}txrqlJtSm*TZ2&kJudeKb($J^dSb>e2u!5X)tx;%^
zpa#C4@W6iZ&-Pu+nCu>_S>^gmt(7NAG>t~J1L<DvRaf&p<@U;#WB{g;an_fzW>xbi
zw~?ldYV|VwobMM~Gwy*8qLMR>^muhAw@^9Q?kn?^3un$*7BS>&I60RdXIs;zM6!I<
z;YqaJb<rGhhu_}%%TcZ2Zel}KRe|V!j$za)@>7|_3Dy1SDlYAVDIodKws3lEe)Z>|
zXw4>+>N!v9l`Sd++%vkbhg)fLvXFE@#;(cj22S;)+<ui8d3#3+JOFLKliiDP>ja0V
zu@yb}cX!o$|H5vYtOfOdPP>T@8?57s3Df$gbdDuG)+G+F+LN4-uK=G!GrI4DgopC6
zl4UI&-cf3&T-DNzimo_yy|TS<Z5DoXDDeycDwKGIKFFZKdk6m6*$=@-3XA`L`Vn*e
zhnyz|=YPpe)RuD41HkWoYUBzS@+mz6!<D?m;S~Z<H_-G@fs^Tzi-(H?`cp-vQ)mCA
z>;qqe-T?YwMx;=mlalrOXuZ7xD$YOO4|g0I_~{JO5ShY{4^P4G3|M~%Tve&w$Sg~0
zM*}|Ezutvxezm&Ky*X+7#1upZP<*{4ysna!0wokYy=5c>ke&!-9qDeH-1w=|b5Cqr
zm=`g>KnP@G@t9uObApiO`(xY}!Rn^5h~Hu#s(-!E)>}stm_ps$<M+F+9%PXUG#ypT
ze2QUrFSEK_H*nAj5+Ul)e8shP*iXO+vjR593R!Dw`Mvv{1Bdo^YxjdlOH^N@_h=a8
z?~#;Ls}jmBlQ($@?M!gIF8Jx&@03N#BWDoGIc}@3wVbsuQcDaL+~HfjM_A$ne1iPa
zyiU37i!B%R0lha-rd=N#BS3Q|@cQ_51huSj<?9z&9`U^2$^wBXc#<*A6?J%Z=|u|O
zciE8VLNEC~C-2V~m%O!u_nXIW>%T=Dr6byXG8r1|ni!J;6uarzudj=7`Ou|9o8{8i
zv}1+Znr^!eYV>Aw6`9)XdRNUX=_87Far+E2#i8bWl6MYLpmg1JkKN8Xv4cmo_2Yjh
zM_-Y4P{oNcGtD>mq3o3(k~`w4wG`9c6V5D3&?llr#g=gZ;?ZMDd7}h_*}Y91+K_j>
z1+B;j^|PG@sZUQvTD+L_tx<vR5b6lNYh(7Vf<9vu=3H_-)X8L5OXf=&o8`$46AmVY
zYjb*iefY~*#g=W<;12hvVUYy6B3o){?El?KqS#c+&}TX{qspGAu*&1&s<m=a&04o|
z_$U-N&8949N{Gwh{Qb5Bg`|)HQS!3L&e2@1I}uVIn3Sc%pq>|Q19x656sJ-+S}5HT
z^f-%lvavW#gNNd6$#n~FKaf$t-MMG`)t*!*ZoGvo_a`e(oEXw_yy(I|6qlZ1W@>Ea
zLKE%;$m;#+`lbMOsP9!~(BDIQ1I=kZAf34uh&Gt9u2^5?M|z_3*g*1iOQ6s9XSkw2
zcu4E-Y^t}@q^DrB!_!hV)a>Um&9`_Y%5-<2Hd8aMF?w5g;NRKD&4G&J*{sTe4^r{|
z6D3hcXr|muTcHh97NUC$l3F42bqT3$4X*W6Yn13oxtK2+m!htBirG`GcZ;9#J=y4%
zpZ0A({Tu#7nO;OHVyOrF8G{<omea*#TUyDEheRQjIM6YEEfV_A!AjPSl<LIb7E8mS
zGVre&DI4;~NXDFqL=2ZdMdh2U{(r(J)N!qVFVYN*25T|$ruLzawlDttyjD1{-onm$
zGxX6E_%M1}zBg`fu~Fj4o{X5y=(|%mm#2qCAX|t@p{L$ei`Hiy(dT3qogv03)@^$3
z#eF$$Faco-{REq6*5O5n);Kz4iy&CTwg*+7^7hSWs>GW)4xX(!;o~xUHgMcIfar&z
zjLn4bqD}Km&WMl={)LI3Y->ASc7Lmgpjy)C?2JbBYLDE}^NCr<D5XEx%Sehvx#=D6
z#0`h$`fd11%^lOU&m)f>24)g>pg-F84ArenO}wg+=xvk-cB&gm+7XjI@!Y2*s&swY
zO#9y3Tp8=lKY)3>=(H(aEIb0)K5s}72~k>5NlMESO`eyN<CaV$;;nVQ5)E&*p*h9Z
zqEVq5<!~rPyJsl=y|dY+svE|i=bU#8V^8MFq}VzWkvvjD_QylhjY}Ef=U9P{suL|_
zalN+^iBi!%5n_(%M-S?8`rF$EgHSszBE(8AhK}(FRk7B<>e<-9BUnNt>jTyg6dLt{
z`;yD#J@|bOO||!wg0>YYqn0HDw#g%KI~IO2(C=55))#|F`yXOfng$_`+op$TE3>nc
z6p(`9P+<81rG8>y1zS*1cbgJW_sbNVFKdwBNSAm0;Jd<5;5}p}e(_nExu5b+=E#82
zl`!Pw>x7?zVvgSjMsodNy;-KqFpkK*Ol@E^Tn+0rk#dn+JXZjty{PqV1tRv4XZha)
zfbjKL12xu&v<@1*wXO&EO*pqRL7MmmX%GLWHm`R=&G$8_1JnBaWFI8CPC~BLe5(=y
zx?+SQbqo^j=ie6L&mI1@rzQfweV?>ho=tmG+gzsA5|>#IN&svyRSbgI(0)0-9eng&
zYE%$!op2y>4ZU>TVw+&5v1EjY{-sQ#mVQf&-5{nP3%WTfkaZP;`pt6MXLwl~OWSMb
z{1sd-IM2k6XNzX0`c;U(Pk3t)x3`{yIM?Q0I7@&et*;%2qyw?8aR7D7xgv?OX7q0@
zf;a2f8w1-6z}wPNfWF_5g9WRd_1X4yu-6P4I*li*l~k8k??n@+egJM`vy{JCFZ_BY
z*&KF7eC2EV>4}wQdpte@SV<F-1Q(Z4I8!sxmw~4NkedhU^{>JJJ-)wq*BzI93q>kY
zyL<=UsaSK}X=l9|@DgL~Wp+-Cl639;BvpEg-RD4^?Q2P!JtuSJh<*rTr))v3#Q2tu
z!$*tEXHvn?gi~7|Hig@%6~OgZbIbuMk!xKF-!Dy~PKF;z{j_C|VVfIWnlIAFPRlOr
zh&{--9(3^y$2m0RG8X;OGDhlxnd!e5wCH^U`e4|he!k7QuH$mQMR8K`CN=Ut(Fi^i
zv*3y)?BD4vE01evM%2b8@|Nc`PY0<84#_ncmfzUdUU2&7XI9KvM0p|Yd=Q!S<KQyb
z9<&3t5AJ-H@0t2PC0qP`5>TdvyqgsgX8B}wChhC@Iobk%p7y$vXzBk>86|N#*<lyA
z`_%HaGd@`J--x2WKfWf=<|I$L-P!Je@#UU7Rtr}z#Y<5clT7PAUvt7A#=Go}r+xQ5
z(bL0_)y6+;I$4rVMKBCAO+8}ayPesQ0Bcc`S5C~-?9$_Z<^&Iy*tuXHT(#ubz>1Rv
zZYaQ^jAo!!=$T&D3Rm*5TzI4<-S8Aa>iID<0UAg>1?*XRd*|8dCR)0drn2twUo!eN
zfLSqzx-qn$LJx#3?F!o9PT!swOCuookhuIV?)}`^t@?kE+8U{43w`i&?q;k4wHmgq
z&FMeR<Fu6QFlqpO|FCSyN+55+^dCM&Tb|-xy#*Dlm7Q#yx1O@GJ{~mB^!q#_l^b`o
zH=Kf~x)wZpHYL0tw)&}q1T74nBC-sH`V2b^LEy9|RkFaJwx;2@yW~6F7POjJlR>Qa
zd}>d$y<U5yjV0hLtk`f?^G=aoNORB<=Mi&z)8xVJ3nIMXDsQ6d`@!8p9YIx>6HB&D
zDnhGd%Pt}^HDg0k=~-&_GGj*qLZ=%cFLjHIQU=b!{L%%Ps7Ui2om+Y>Z;S$K?`daq
zCPz9g3rAz_f(s7u7vIj06iSIvnNZi<mwrbU(_%JHKG%db*INpuNt1educHlD2z(rt
z)D^%X=QP*k#W2P4e@sjgH_euP%yi*7@LYOAp3D`jcforRthP-rw_--AQp*uc1||m`
zF4dFKMdLDmyD;H}j~RW|6-l%4V%Bssxvn?#sK-$_@I2ECRGz^n9a5Oz5B+L^aOCJM
z9KSu?99{KE^zNMZIM8C4mJ)hOK-8ZuBk*IQNGiRG?%CSO?|O&rqzh45byT_0`t&ol
zQT=`BOt^}!XQ_iA_b_M_(&A)MC%+;PIChG_W&MH60UTnvAr4JqX6CG$p%1XNr1N3O
zuMvn_ti)jQ;q8~6AEb9z>r|%mI-%$uM;+Q;`pe*5Ed%i%#4;P#|24Jx|D~{SvHl+l
z3kS=8*H^frZR4ObhWuS`&_s{|vIAh@`J@H`^)I)_5pBe6ts(?V{XO1FrA*iWa=Q0A
zot%*@yO8`QOgC_+yL)`o^~lST!DY%mdz9EqbgPCm%v6;a#~>DF_?rbxHw;NOUS2k~
zw4hO{GzPYhy1O%yDhrViOhi;jAUa&LS$ZNdaS3l+5vc)1WU>^6P*fd`OeESHMQBp~
zmU$H|!xhh+z6@6|I$WwqY}+utSegB*3Aogla~Z&yZ1V4-X8#HmxM~038Rsa@qXl5h
zi^dm{cb^Y!!6*ddrWdU0ipT!klIO?B@689FaPf;qIePT;1Af$>Z0);H#F#Im1UY|4
zZr(=TTbv!`!lMf_N^NWxORB4qR2=5gN-;jQ65z=^$n`B746TYRI1(i0@q(;e|CS6b
z3~XU21!Wy}W;B}VhB3G!Yptz2^x#d;9(a!CG7hn-gd@~6pJ9w_5m+p3+6XX4XJ_K%
z1BHlVu^YfJuJ!=!Q(A5@Cgg-&%iCwyEg<ns`K=)(8(qqCc&`Ez%`ckbtghO<t~82u
z;Xz-CMx5SheI9u&bwR(Le~}`weOK!7i#|+^;X+xNt@H#zs)^K>uB#h?he9c*=5Ix9
z=C3{_>grTEr&`VAh$A#$cFNK~<KV3n@fd3KgqlmBl7VGKOjkA1WtM0vM%-D!vs6uw
zAel7MXe^0z*$=GZ##B|0pbnnu_-tg!SpYS0vMwPXwv#k{`9~cH6C}7k?$8_)!PlM0
z<Ih};|D!J}cE`-ZvsI18U+h@^tdkK>=iLHa9#bHX#jjqktsdU;jS;|)mX9oz&dWvP
z?g2E>CJ0Z|D|%7DA}`1KcjZ4OiT`#F^PbRsE{gB3dk{p)$nA%3mM=T7;9pyzZ|KB1
zh48l%)!JbbL!hr=!nt(Wq$bTmE)odn{L)lf1ZL1+C}F5lS9-#lWk+1}j6K}qJz`hi
z_3v{Hf5l`mXK+d`_wP#S8E2;M9eKIBFY=z_*4)auuG;JZn3;J2zTaKn_m$Ueej+W1
zyRi~<?f#$wL6Du@?>MGvi-vU=7f5%!y`C?%T)DAx@gLVtE^ildk9~i;_`1pphIhpX
zFvaZxmgU)!W?y{P3gR0mEWXcaMJ^=-p64KYF%IX{9I};Dx<`y0f*q{Yv-0$w`SMoc
zyuSvW=<uR@(&M`b^N_fS+=Z3*O<63=n}V~E#<dpSf?P<}npzqcnlZcz<tXX7V1;Rk
ziWg7xLqE*h45g7^hZ)6cx+IDtTo(ES2?LP`p$K#JS!;XU#y5lSM!lveIR|sNhD;Ht
z@Qb&Ip)zPom}vzSn<brP2$7a43oEUOp`W495^jN={&wB@iafd`q1G01d;NA1)mn@Y
zCl=%LK$pnH-S&z<c{l!Sac9H_o@cn-z39K?Z@pf5a*%UjnjdcsnK~g@8y31ft!aPv
z^z(XNd;B2)@@%Zi)@^O%u9R+q!~adMW5&CSb{Uy(N;AV#M*A>pd-*2zprxPI2!ksU
z>{i){*dS`FMF!y`YdM31suAo{o3c~so3)dAUYE3!iLoKaWj&)St#m4U5xNa+M*okV
zaRGJ%wg=rD2FK|^UaCUV=zA(VU@nxl8rxM7Olca!-K<$Qh#Ld1@6|KVpPqua&Pb*K
zRR;x05lV03W0eU8qiD=BH*6_dd)QgQdzI1>y1Y~rn4W~v9}J?JCoS%iCP<Ejn{Ni&
zoyh~vWzjYl6kcaJ70w)m+f0E<i!a7StLS-a8)+c2FC`p!>J0`<VmyqlNLfZJhE5r5
zzJdA@N9-xHxoBFGMOmn9rK8b|KFMcPsdUjwCerp4sca^-X$&;%>_pF$e(-BN6f710
zDZvSLUr0B^8DY=`m{=d&XtM$QgW=AsOYhz_xG~u(psSF^=*}sXaYGCmpwf1;+yXD1
z`6@#;y~Is69g%|~t_NLGAKK(9>=A>}(S0$h=XO}%?bYFRY>a?Tmls7W2^+4`<mmD%
zZ5sKO${+?Gp?6jY%)Ior2m7}?N|}cwLhU1F6IUV5EJF1>7HF{99vi4%`^ti-1`Fn(
z*2Atfgf>(%4r(D|`nQJQ6|vGsaXu9+(ixp$LL>Hwy>>T{fv7j={n%eIio)?^I90=I
ze*-blxln{03`$>7hp^CpZ^`GfG`VeDUTN@fm!#mR$B14t9!rhLT!x%Vw90z?18YP=
z+T8h5s2!=-t?y?3D{e<a)~bk>?5iT~t3mpo#WGHT!{Q~p1@^fS>oT}ln}53y+AXS(
zJ!Pku*9ww~gT$>>oQCP10)jc1-V+~&bz{$)4fS+&L5@y97X(G0p$ua7qJ)Vxy?{<l
zI_u8lfk_%DPQ<8g6zJmlhZH71ig|MYuq-zOx_rGFJn|IOOBZ-oCrdE&@I+M0<`>D)
zqUI;V0?Tv`0ka5?T8cv{IW{NU!RZ<!^8amPU&8;^Hbjj`;2*eLlf2pGUs`7*#N|et
zd?WVaWvMuD_;Hjz{V2cVgV&Pc()LhTlg-hV(fk_BO|&3_1n%mD5%6j%aUaktWogg6
zt$>Az-_$hWgGoxHoU3CnRS{py6C)&=brbnr*b7NG4Gc<xmC;7oxcQ@a2qySJia-JE
z#pgShHYTlwZw8Qgk{mYspN@XkbE6rsx`H1|Mc<sZ@)3tiC(RrqpsaQQo2xkE3f~K;
zIYuMs;gi@N`8|DB?OcOf(Jx5gh?rM?zrx47wn)rxXM=+1;#ZOYcL(co%bb|s{Fv~Y
zKbYnIJQ+qWXcyj?RAdP-^5VKpT$eG~U>eIEzed9Ia=r=<6?zp8uh+G`0xRQ`2x4N+
z6mWAhFMg3}-zX4!IwDm$DB?x>l-o_~WL?VO3m+xUVyD%Zf|t!K@nw;r79Tu*i9yu2
zGuw~sm-W3YWhF4?<_~7fw>Hu!fEFE*9F0AJ8+mo3CUJn||LmuIWFz@4kf<}OTc`b~
zd&hOk`W$GtQtWu2qwqMRbO^~)YTg<1`UnCM8G<MSCzGoNE%s2jFT!gxQ#Y|MB3MZU
z{8johG1c#zc*GJ-!Bx>u5ybsQUl5NNCw$0PqDX2LF?*+j6;^&s&R)0jyIe=k_Ce=N
zbzcCxi*`JWn`bF<<sK}%)aOJr#;d*4cPx^MH_`p4|2oBLga7#crE6gr_t@H^ug%cy
zWJfMtx2&szx`uY{c}m;mdjrwRd^Q(QPVXV7ck^I=^RLzlXZ6TlGX>z@w9PRakf{XC
zS>8tmbldttvM6Y*)v(a8lug=iqWV3x<g_L5IsY^_y2J);;NHTO%x+95F$*yg-Kt~B
zV|)8NRgkcHQX;JmgUNj^Nfn#{Q3uqALhV|QLsZMRXmop}a%;@kZRy|J$yxAJ9>SIc
z(bmxYg==*;p<JFv+&f%opdrUmu@<ub?qKKOyorNQfx`+?ILAFCZD<SyVFvtck&d;2
zL*{un=Ea2hOdjy#vk2(EP*`=<$75}tgH0U5V&%DpF;%;BAcz=lYpZs!#^n|eW-0Ec
zaI<9#qFEC>eRl29D4?GmvuJD%#(NRRaOdr^{3$Hwu!8J=o!zu*DSv<Njb5X&uKgyw
zoU-gaiM(o)8FA<5!(^KSL^NK%IoIF+1;!!~m-r9DnBza;S~%G`|0^&1$p5qj?iwfw
z7{*EV1__Z|xBCQclV|EA4jU(C;f@NqyNuudtE);Y^={pAr+}E$D=U5@7n3HdC`ife
z!^Q8%<JU{g&%J}dmEZH16=Wd&hLj3fl|>NL1YhdNIBzH@6O~yb(bJYLJ9BouJzv8m
zK=lj;62XWA!X?Os$S)y|ow<WKPX@<_#-C#aiA^i`r7okW4*u&t|IGQG`{#PA05MsN
zvOmp16+||(b}6`$X)j^XwS~G>cwh1kU%UD-X?H%3=--KkJT^w*cQ83Re3P{<)(ZEo
zm=bGK{oK1>aEXNry~EPgTYBhx!qi6ia*5yIRaQ3w+Ty*>NdJ87h)>Rag@uk2sr9tN
zNNyL8F>6BpF=%>cQe3|9k6b5%ufyAJM-Mio0`*Y-o3qw0XpQ}QWlG6K(F0;e1Qn}K
zlJ&?gM+U7InRF#_+|?KK#@I0Z5Lt4A%+2CN{Fo{d(?!h%@qI;>j6dt_6WAE55f<_(
zuoUqJxu+$}VkK7i`Ne;y*MG$yk)g4)XDU}F#(M4p5pEoFfvZr!LRP3rbKcf$q3ite
zbFE&mKDZlNmXFKGCoZ+`VX&HK0&7-Sk8Ws^Y73)(BUX*g9v3NBZIU1u_)z6@7%+yt
zUH7hMcw9ZGd48KV`Nvh|*|_8#i*{GTj#n71pW@_i0;f+iS#f;x4MoA3=&xxXgF0O(
z!%goULZ1;8sJ*Lw)~fzJ*kxX#rXYx<Uv&V@b^yBc%qc37e8wxGvMevVflW+mU4E^$
zAUia#CC|P}={0`TDM}E{PyI~ngfHd&a&@1n!so&@Z$3A9>%02xEnlf9M{l%=yMp)S
z75`!y#`1`7dJjK48e7t4a!|M4@YZy0X8CZ59X^>suT?qOb40MUaXwRxF;3W1^KQsz
zbL=!rSwR0kR2prapE!vk-RBz1CiWMWjNJQH0@G;{J65LYE-kG#105}z`b@S|1Up&`
zDp{O0+8KW)liGh4+csaUM}|s?T(;D=%$9&QmzfU>eT$+)wtkzh*0mMnCLN;b0g)86
zbx;{``?Vx5|AlP#IykawoAOjSJs>?hVE?q3S!=8lX#`U*`8?VZJq_2Dt+Z$JkKEp-
zA1piY{f1qvc&~4+4sNYA=cY({<n=l)8M-<nELXnERSqnQhUN_(zFIE|Yicd3Q3$KT
zI^rxIJae2<CR4mT8~OF<e@3K6OAOU~XCvC+xHhM+cV=F`0G_Wt#oia~!j`4i3=!WV
zG$=8(v~n;o>yN|RrG()%WScEFQni1@Lgyf8q)zn|B5FOOiWw~w2w}<XY%FzF+TX8Z
zb5@w|G6v@&sQqKH^B%v1N2|zxQeywHiDC=rh3X&b$=syPOB?s$!Vb#GYtamy1|2#~
zn?ZTgKpGH}i)EmPfVM?4_)df}V0+=_*T}k}C@%igh9tq*-W~m`5d3ovzfTS>R*yRH
z7bRgwa*9MJbRNd!q8Ph{#qcJ{=9gx=)lsTxj_SxAl@M(m)>_2dM>ZokpQA%Mm`m4#
z)QtNy0D06rH5_&%vEhwdKE(6hB{#hsz?%-^zO)S~UWGEL>Vke3DbAE>rrPr{VkH9F
zpRYx!Lid2mDFn@aQng0`^2JPMDO?!Lu@q3wDjQzhZ^>OV5p`D9v@33d&;NaSnjql!
z7amOR(*vx`lww}+ZV6U^x|tfA<NLsZMzSu*6DdeH=js|f(Xd{=7Id#7L|kGy&Otn2
zWPy-!gqS@=B|7d}pKKlu4WD?_EA{-_8j?mMP1Lx%8M%NMJs*d_%0CN@;BbW>Loxd~
z;rz~o+t;iP+op~L={5-ks0{ej>R4-m8mc)ZSiQvz%Bo8kzT1+F^Wh?HL46lXHIZrp
zl+}BGd~R8_N-b+jlEedxSQlIbtGX8c5XT$Sv@GeOG@RN&JG;%X6vJ7`@Y)=H?sd<=
zb5|kzx$J$$6aSz-gWcVGJ{;~)?{`|dR`t!=g%Je9yd9sfJe$3~By(EZ88#A9SA)v1
zeEOvk`qgDk8gOdTg!Nh))5()K^upo-0t0|2R}3N?KPOmw7f}9Im*dejpTg^1Z&r35
zJzc6u{mW?G^ubQ*ioYQ(PXxfBC~I}!R7LH(wtAgqt3I1OVg789^6hdl$XUnF>h+8U
z7^enXuq76Vzo^`iZFpL0^x%5cdrYx=snP}DjKFxA_{AWEJ{r|J5M4w$Z~DOM9@jPw
zZWmGFFd9%zLRhMxzqjLaKK9KXMqt=cnn>&aOZ2u987@CocIQ~PcjkEmtYa2xi#BJ~
z*&|XSa~QF;SB61KbUDx!pETKoZT46Tny09>gmI)lh0E*@DsVO~&0?dU7bKO!{8KVJ
zSoK%t97p&JS?4yACd0?J$)v1~uMtM*nJb^I7?<hDcyK1?mDz7`C(_lX0MP!s4VtPo
z4uzO9+;?~*$#~OaiiNTJ*OG*GUb)T>Ur@xDG-8%jm|LZ8E!B-o4~eS@e^&4oj|1PO
zyKxH}%BFr(u)`{Ah{aN|J&O_~sZ|L;84+!&tCyi5U?lJ4sxfeQ3{!8SJC^#6EL_^6
zCETOuBO^}jChadotZS&IoqJfkAs_mO6BPc1#PTnE1IF?n@GO86sDA#&!jJN^nn}k|
zFEgZSeK-B|C1Yqni(H#DyRh9Vty<j#HuF&Mq!!&OE={_(O_X^4!Dp+|li_n_FBreM
z;Ui%?jGWqdRL3rej*VZt?^QYPINgBLQ!tE&(cbRMvfJG8;s5JO&lV)n_&#4n-OSZj
zOI^a#;!-%}s8bMT+H)L9#1{wd{xME8oTt56c1i=e$(xO+z8WuG3>|C_^Swjz#-{-!
z))gHN5gX`77(3ShTS#Ls3X9X$SVDUk!R8{d6Hwk{Mz;R$3Asoi+rUdFCZ$viNGMyK
z+nR#5a3<=V%H1`(VYA5p)0WuM+|4qFinWoJ{V1)m7kiq|l`Nf*Z#1cuAH2f=M+~-F
z7_?Iu#OLzI0ztLr?Xsy`u6B%aBrO3>J&cphCdDQ+908wt*53ToN3ASQ+5W?^=68=3
zA})`81cg9=_6U8GBd|o<qA#JEg29xA*-Y1S7Y~~K-1+0D({<p*$k+Fxv4Q!qPKxv7
z=V~w3DdpsJC)*(ty1?yDCNN&8yNkV6Zg1-3)Uh<k1(y*=^P&}8rqcZof9_SH*`BV;
zW@k@#(xUpz1Vt<;Yg1aWw9<eMbitN-=%s)EJMiqxO70@aN-*U4nhsjrkQHXpv(+xD
zX}sOzgqXa;Q@MZkDDrH<)%IjdvN~qV^20BJihb{X6dK>{{XDNnH?*E8in_2hnfz59
zelbNt4DAAu7x#}NFRCTDCk|I+KE$D5|8r;A7EpcX&%+tsBo7$IhU2Jco$UU}CEd2|
zS<vA5H*J6=>(B?^UwxiENHH|%X&=`)hckZ{SXRSmH0?h2{ycfFx&k%pu9sh<xM^lw
zm>OEw_NSacR+D?<OyW2=nvS$<&(ZWF?|ACL_>X)V!TF1AX@wC#e`~?`5ZC@a<F=Yx
z(e01j35yGIe)oNOO=7=c+FKgsE3;w0Dble0>WkGdsjX-+8|%qo$u7cm=FcbtxRM;U
zH|<JBC8Jcq5G0}a$TrADeb;tFf_B|oh&7N)f31sZ$*<d|fF3rRVJte3P8L#O>=A`>
z_jte+#?PkRog3bw^Wvp_&!49;a9N(jwGD2xYOxh*3=wS)hVvcpU^h{?jzsg_?M$zy
zY;>3)=0w(XC)(MU?965}4UI4S$&nBd;`}APo1Wu-Dc`;ulF+}r{}*G~NZ;%~D)9f1
zCgo)3_^%Xr9rpj#0CXTPh-qqG%aoWy;0o&u6}!&0E|mXIfLBPWC)vI*$B-7TEZH_|
zevYOuB}mB|<jXrK6wq(V>Vv`R%Q@^(3}B%A4f!o(L4in^8-9Mh_SAtvl{S1CnvtYz
z)5H6N;Moyj0S{~uI07pHgz%4$TDM>oPxII|B=g`K_PBy2@y|%?5`N*BzO+EzDbeE_
z?DzJ$Wl%txbocSXT6~p?P9(eM&-cx>-@~Ai|3w-WKPdk<XOXdfuxr+<#|9LsVvqCq
z|F9DV!2in89GQkTBMm>*@M_0Q6)!F+cq0f0br^vlBkQqnQ3(Re{|R&Zv-M<}icx5H
zCLU0}Kq<>8?H0oZg~uyji{mO1D-x_J;niR8O=z;xKF+Gfx~LK?Y_mjCf2wxQr1#CD
zFrxN1JHJms!GK1fWK^JshhYbsM+oz-QW{UjOVS6~k2QV+xwqBbq?Mu-fqFlfWt<Z&
zG$XM8!ahwC{~YRsD0BqLGxGNE4FVKG8&)P-Tib}F-em|-nkuMEmuuJG|6Ob*;JA=+
zqd&k~q7P-V^A!7qMhOp=c8%d?GUIQ>E=T^jG8Q5~YC1npi=8c2!eY@)&|-AoX9K|=
z5e4O&$NdG9H<vcw!0`y-7u$9Mal)ufBfFrExzpLazmf@ZxFBqNyF%+tb`;s#-S}JA
zyjG~$)L1t=Du1FJ@b4S#9$eG^l(w-S@XfS-t*#>wLlfSn?yqQB_$Sj+R^HxCcm~@+
zCDS0qHfT~76VKz9|H&q*X{Po%HB>9!!HAZvykyaf71H;**vm=M$@oI|&Gswy`;yEd
zotE7WZ8`E*8W(aWwMy1I4!fo-)pHxHl0LqJW9T)V<ngLv;nAkz_M@Li;*0g*<sQ~u
z^q;5fIOTqv&Md9qo=arc{piiSZG8}Rmg`paXippBPT<Nj#E1Rj9cHrqyJ!`|a4A@@
zn7c$R)=P`t%~BCtKaf0|-G){G32gtCZ~_+%`u@sVAg=V&750v^h)GcwIjRxDaRhl-
z)ZB+%v)W7D$Yt%qk|Q}2sL|9i&U04mIC`uXZ>njAA9JJm%7c8hKF3o26Fsw^DK3IC
zP$3h@7H@gf0LkGDhDt@Kl%79h5u8geHjh=3b-(mg3UQ<8mq;SIDrPcZC|)ihN`1&8
zgAogvZIoS5`C0ob)_0=Me?5XPuIW=VJT)|Q<icFR$4BQJZ2kDc<rKhRDxl>H=xk)(
zsuYW@7B(4aflT|hCiHjm#Xf+0?>0HDQs_mr#!u{}fX($&yT_;fIN5vX@v}0FW<V`h
z0EMU8nsW=Lb_>p`+N0M<MFcS%M0?ZyN{poMDt7&3<}Gsga8csZECB~g`;zuna<*bS
z59G-?e-)^>RWOjK^y5mKNS`%Q-i$#o)py_v9)5}xlh`k#E+HzFllFPryCqm+e2|a{
z5~QKU-iFee*>TK^yOQ|JGdoL-q8+@T3RC7!%&TVM-2LcP!5VS*QUwI`k#V0N=jv&a
zm(s~WNNiWpk2+GSM<!MzInRxOME_b6Ol7k@Q05(F$BUb8Nl&X@u;kXysioA2&Wcpc
zp0lz0D%1*Msw<d!2?L^JIKel2B1^gjDD`<65iN?LQcwWU=-g$+5%L$%E7fO*uT(J;
zcq}{3q|7+(va&xxA#>mnZmu2ZFt_q!NGO~g?Iat6<F9{sAuV59Yl53KYIZouz)L&=
z6&C;Lsj0uDr^!eF{>6~MYyiu~9F5(c?2=D5hpYM3qq<10n6B)9jI_uEDDO4Xk8q*W
zu$3=%fDR`w-k4C|{{ZcLQ(j~=BR(ar!v=k2F2ui>M*n^>#^hw6jpcEevJ~)@BD+eC
zx2sS6QsP=VPMDOQB#^T_CX7?oDzcSF&_DGepW&KKuHX4mGSZ8GKMcI(0pXO#;2#Bt
zq`dzF$6AI3t|PCbjBI0COg%#BM$hF@#DWueunnj`G97ITE^}%b$h$w%r84}%8V&EF
z?rKHUP!^)gRUs&4Q|H;g2EhjJ4~z43y(srityL+SmzKh}3B6V6l08l0bhqCPAKEm=
z$nZ9;*D|It*0JWMk;Ev`(n!#SSYvIw=~saOtkRx~%g5D7JsOeFfXD(R4r+cF(XS{=
zw)BFu(D&;Wq5BkvTXK-Dnvu!`)zsIgi9S}THMeoIkuQ_!eeQy%lwy8IR`A|5FbA5&
z;Aoh^(gHk2M7R;gxoVF)y21#gc?8-rlqYJ=Bm-B+lmk>0$CPxQ3OP^Xuv^OAJM-js
ztnKt1xmFUZf3L1^=*6zgLbry8ZP)uWD1jmgEW2Y8sWE7Kr5zATIq@Ofpf^BFDPW<$
z-H1lG(wazhz;V3lp0085!jz)86)lutl&?4A;BT-qDjQb%R#bpy#Pguf*K0L{70I@k
z`s^g1(kSB_(E+gi_$VZ`YIOrQG-{)l$?ut9?#x5EZ%^tx=<D)5z@!`O&3^3LZp@%r
zKwb3DkLFSzc}QM&q+E$Z$<P{>Lt(d>u0;gZ(K?n0*l>SW9;YJ`9IN|e^EkE~H?bvm
z#&j&Em$`qe3(qaw?UI91<*a5N_^9N-F?*GzszfDLhf3tYZNl6xT4^K6zbDu-Yi`S5
zB0s370L!fw=Z<Y~aBlcG3U4*;CID=+x$YrcJWhpgB_FL?=;B!l1WyayJx6Wn@DisZ
zps!`tfFwv<mD^YwZQiJ*E!M>@e@r1#G{CuTdEK$AUuyVDMsH-4{j3b?8ET56U`FY`
zXgf`rdQ<PYGgdc>@Ju^O&ba3iBqlx09PM<+!sK~<%4|K>6AktZyT&!YojFJ}xQy=X
zFJBQ$;=xLU1M(*$m|_hop!9OuVffvr@A14m0zt^0pU=A0z8ymG+*Jw}erq#sO`n{;
zcN)x$b)97MJmtac`}StV3Oh?K1fG_pb;1ro@(A0Eg&Y!TYnah<p>Ri=oCY$q91gBK
z|G9~rU(xkwfq50TY;dUXRoUl2<Y!%d*Gxm?N%X<95kRX|<e%min+NHPHlI&sZax!F
z_3~@`^=?`#5L&2IpVY1nFELz_VR`9HMsO|NGTy)zQ(T@uJee=+Rf!Y<N>|Lw+A`NB
zT#E;_p5Y4z(#7ZerGIYBj)xYjNa~ScoOyl+xTH7Hl_wPETbSM#fd)6FyFyl-qgo?i
zcFEL)e^D<N$#H5SKl7%n0Vsd##)ri4=x8`6aP=+H!QO6BNd4)$cdacMCYmUrP(~_t
z+>E18#uxp-x8l$$h3H8uLn{zt!+8%unT%WI#S=H&9&;_kTL^DT@2#V2z#~Q=VXfCV
zZ8k{x)TNBYDPP=WB?5TSn>mE%>rRPKXca2TTk2E$`LNq;RrFBtRMkhWdJuvghnC@2
zPJtL-5=t9Y7f(ly9XkaB-Tt$EST(G*;8X@>bg^*?`$F1lGKbMiZux}lO#)BwwLwB*
z3`K?|$7G+?yGUAvHZ_WGjrrv#`4I-xLWr~hYg8xD-6zutw-N{2FB~Opr}XCMZM5O{
zjTPBioh@{1ySy-kp4S8z=)Hc4gl|O1HD0m4YH5GN&C?%Z;Ke-1bNJ>iWD`u;>p)X$
zaNA?ko><!-ZH#pm?7r&07mgjv?*xuKh@?D+uN?d;nD-Z0wlNY5rWA%Vd6V@g-Sh~U
z2j=$Eul<}|XCsi>X$B`fc5b-6SN)+!JaOzGt$nYT=J)rFBa3=u6#l}zb<5m{|3JEZ
z&NPV5<Y;pj)Bv#FCRr!9b#wOAwwn4GJUnu@S#CJtb^Cv`J%v0*`@HK3fAT*5O#{QN
z{7(X$iSvIYnwZ)D>s&RUo`efw=2qhkQo97Q3ewSNU~LEAumx}1wrRoGycpX3M?mvJ
zPD;%T0Kl_VbWf8OQ$i((NIg2AOX2<gc)5RnnVK1xfKHe^UpghA5c@X+r<RICSCFSN
zisSbCar;;Uq;4)g3t>oF!Rf`{@O>Y?$0+4{Q<lbPr*a246;5)Hzch9AFSt%@8mI*L
zm7%Lurk{$gC^OiAh<9UjpV^?z{{8tH%b`_CBwqK#`Fm2~P~h;_FNeZ34psYNDd5LD
zhGBrxsE1z&Nk408ufO2O^Yc_TEoO=aC8HJH2`M9)RD|;%uRN-ux${p=s^qUA96eU5
z!08^KkJ|6A?^7%5_v^LUjs;;W9+g#Vl#tj0<FBN!gLz@pfG+T^2_8vp<6MbI-;x<$
z5;_FkaAXfMqdG-(hOdueC~g{%Guk}TES=(>ag<!3Lw%yDqzBR|F?yBTPZqCe3dkXD
zwuEALPiK9w-n(T};ZaZ}@!uF1o&I427M6x;{wM+>u%{K$dz!$rYsjXomy!!*I#J{T
zuUQyUqIK_)HA|mm9(_*=+j%7h<8bNb>_P8dxr(J;N5I+M_@8XWB7bKZusqb+Nz02(
zoR!Km_MsnhiAnD2xEn~}ck<O^jJ^q_N7m3C&5Vo2t1a2PS3{zJrq1pOs^oUH8yfs{
zU+l{CNef0_-RNU;D94Q_Vk%?F*Py)7)B29fB$ESq#o@psZN>WpD5Tb%2{Lrpy2$7f
zXKY_q;aZVf?MX-`eJ#tis}%(U9za4`%xzW{uz=Awt0m%$ugKFI#C&~`iJ;l&1FaRz
ztTjK4k)3X1zU}nP$ERFbw?=7*z1GYF%e4bdWxUPinAEvtO~!aW^;3%_tL9pI)57L#
zY^0_<S7*3EQz8h?#{jGW29*240`a&POZnUt?W6s6&haJf$b4EBrEgB`7kOTx)R2^i
zd)?V&zwV8S){cz1^Lm6&ZE@GEt{-K)fH}0tCeP&#`A{c}PV>0p6QtNh)@v>iO0SPn
zY%uY4C0xWd;{2^>M%oH-x-7bD0VR>^aI8HRrD?9-t}1KNROil2uC(IYd5HQD8Rl7W
zu?S7i3tK|IbLg>8Ln=twro;?m2FT`QK3b$!2|IbtL)I(-Zf7LC9u|o2%!g@$Ft0c@
z+aYpogRJK8g=!Jq_;Aho-%4dnPkYZ&GZFM3l}ea+?Z-;C=5r3-!|&4y6231#pM5Fb
zLP<&WPu4)a^35^aS8h!gw!qh))c33><-Z$bEAUPo8z*Gz#u@Lg0kR8oEXyqt(t8}f
zCEdrXpS-Yt@ZBO}EcKXpFLf?WAl8ek1Ikvg!geDE36Eq%pi|k3Dnz61X!Lds2a{xI
z#*)>-Qjtg78gX{Klyeg3uMSwZlF1iX+vJ2wsCL<z+x0b%Pnz6Qu|ZIzEfdSVg-)e)
z|3$E)hU~lsaT~sBb-KjTq?V(OR-5~*y4Tn@OTE;ftm^dE%5lB&106O)3Qf(jtyvUw
zFz@Ml2mq=QH`e6j4jw?I=<J<YArN5k5~W|mfX4W@k<f{71h2-_KZ@pz_kn6CW4xRa
zQ0vnKfR!aypEaFI^7`yAOHLE5v1&wr=6Ku)7X_~wr>#lL`V;SR)0aG&{2UQS_3r3J
zkvS;KM!B3j@?(IkiZViJx0`xFQLnL#;@LG_kT8rFn5o4FTS8~!d=F@po+@c36>2@g
z*o!upW`FDS*1q2L8kmT#>X8+M0+n!(p7e_AuNs$6cXm#gc6t&DTx#d&-X|U%OpX2(
zTJf{1RfZCs1i#T~D(g!7q%4vZ+1n&3@oT!?Q%;DT<D;aLw~o0}Yoe>m`q-T6tJJfq
zUvsLLtwscQK-l{eX3Kw&9v)aV3uaV0{^%E7D`H$U#%4W=%aMjybp)Re-4eP}676?G
zqP~VeCREz69T>eWa@9>0mz)0aX#xR`r#lCg@k!aP_l2jC(0bv*qkh;Mf&yX{hkC^H
z@fJjmqyPy^X*G6d9Ds6LxMQ1EV1W}H4zlpEPRrJ@+mTE;gReDs1f%J2?3JaXQ#ui+
z+1fJCoi=pfI~L66uik0Mc4z0OQ5^xVHirQ`H(XIDjV<r!&Yo4@6y=lXmatBwR5W%e
zI4PGU)W&;M*)wn#3Q=C=mzL??_z9MBA@4~?M6WnjOxfY?<h_xFJ+a`ZA5&XR^Q|>|
zRu%3}yf_~}*Wyb$jaBcMSkZ69yT*KJXW3h{t6eNQ!~F&MNo@fUnjvXJGJW+kqq|q7
znZP%!x$sF{u4))j;qUS@t3{io+tSG1h2Y)oi~i}jfhwxS&uaVnN4eB+T~>aH1IA&h
z>I29&Wn@#Ez5^uKRX0)d*EOrYxJfLR%Y!@f_^DB)!@E4-yUvH2VX<oX+jV=+@E!oB
zLHE>i_c+q-2S<+(<`V*mI}yUOsHVYX45{qXo2im3nfUp;HjDV+dXJ=SxuIWg_DKlg
ziTA`u7&qwOsXGKF2QOM%i<m$&k+dFqX+$)P_T{B_78kOr!L=L$8W!%o?c^kPkGa;A
zE9wdeOdW6t3(tB$FX@AeP<$_x@KV1#ss?aBUc3XLi95qyP4<+D&+;UpyUm48QH}{V
z`l6b_EF_K`J9LYTn^w-=FAFRbES~$ia=LRqlnFg9ZK=D1!FS5gGhd)EP@%(3L^Ts>
z9KUk!*ans?+f28Kh@^NeV0_4)t;pdtL};Y9_9tv@GqK7CdTimmcX1~V^&-3+h!jg=
z1?xPIK^jON;%RzQhR*_z0C)_p<WU`U_38qXT-%Dgx6HFA3(a+PSEUwpZX|JCAmI_7
zaTyJoBJw;~7p>QZ1Opqx>MPu9uv6qd==MDM3hIw}oN_9@^G7lUF4K|#%_DUNPHYw>
z|3Rh_t55=7bJoYtFD8;WX`pP4apzLs)69kKPqwkxa-|zwmTa>}$5PvyE1!6Lt7Is!
zw*cT2?clrmcg^7%;hiP8=jOKEjwZD}J%yeJ7ycT=t@1W?8jl*`ce?mLHK<-{FeZrk
zZA$TpngV^zZtT=}X;CiInr-c^w%z!3loHmPJNUPb1D_<Baz;pt93x92WQ7u4-QeSo
zK8`&HHU+hB-0J`C7Za>&rw}}e6P}tZH($J3wbRNV4yAhGwH?m)AgWrV?EC%=rOqcv
zSBQzG9^lmEoR`Xws?pRVvWA*hQ)&Z}^YQ7rnNm>^kxh3pPb>i9otLM69-4o|H7>f1
zPB`mY4Uy0jTYz#SR5g=W9>&=ja{X^9mN)H<RfhaU?=tz1H3a2;ee+N_q5fVve3Y0j
zw|$LWnwUIQaYo5!l%^v#<nBfmef>p%%HNr#Er5M|b4-FJ_9NIf7tDxlj%4KvlOg~^
z9D-i?1?xB``38>UjQa-?`RMm`Q~NZt>+gA3((#Gipg55=laqqsnB`(|dG|r%@bnPc
zK1eEi^ty@w=9Is{<uuYm|4Df=v;QyU$;9<vnlJuk=@Yitk@{}`W$77*iMV`$!h_mc
z=;w7pdeQjM<P^AA|NfodH=#^UUVYh=XNS~67n_Y4b}*NmnCZvqdcQuu`*=Dpof(Br
z__KWSd<k(DR%(#AZ0U3e6uQ#r_1*sNLQ&T2ig(7WY%P~j@;CpcbI<y288ngJ8hG6y
zFNj%ujGz5=Et-Df8|_&`!a7K}jhCW2-CGNfanY}QYUJ~$VeS86?45!$i?%l0*tTuk
z>ab&eF*~+xJL!&X+qP}nww;rGu1?jztIl4vYE`Y<c{Ao5bG*;9oo%G}D?)#N+YL8c
zS2z3169;`x0P3&H{d05iOP?9iT6r`K+wYLcXI~Dz>@TcQRTDX7+=T4wT9M_u!j%Qk
zRTa(1l~pt{>@}<Z_-@ozi&!*ejB$xqM+SJ0zH9LT=CR46*ME5b#UHWqEBiX;onviZ
z=a24NUPV-Oz_ilGg<H!A)JcaFng_oSIA*_Hphzkou{#et;7hwIw}r4gtH0fECi+x$
z_)aOYsHu&dmeNYctqHR!abB{<<4{JqSj~ag7tHp})V#$XRa?H!DJ)6()sY&5k2_#V
zj)M(v&3?`5k6u-24{tCax|`Q@MvpmArpl|0mlq&I7$g8C9RlWHDNWZh?^DSP%lmIV
z+V7&wX`}7|a)NZ{mQ*B_CrtTNYJ4FW*^7p@{abtU9uux%nwap`L_;hNYAJ`On!mH;
zi-|Ck`7Lq-3AxV{TJAie#yg(x07Uk>V1IZN29n|G!f}!`B8Yy?m>)h8fYbDCn=MWE
zi)Mxntcp^jLFEn#&WG{RJ-XThk<Pfltzd9WI|@T4o9+<66=JBC#gG~zZUuuvkXOu-
zhMut7`;ew>2Syg#UCWM>2WRyM!NX$UE_3`rnp%U*xfOba!$Itv&b5zImhB2?GrOz<
z_ZXN`ycM6}kWz?agK2|U8;U?O5pTh1(3l$E6p^My+8lYE7^}=jFVywmcyUwy5ZVt;
zR?cAM9FG0Bw|;AQ@(H>(+;fX?Ynbzr^1__<`1InR_A&Z01)DB<hR21Rm_z6<i@)nQ
zv*7&IGqN!1j*#Lp|0Se-hLs)OMN0Q*Xe&xS*7u|MN&Exu2+D$9fn8EX6rYC&_ObGw
zJ8}>-nCjI`6u%6jjYYkMpPHfI$J)MyAWMPu3-Fjm+4QJOx9xvjq|(-tPU&vG^KW<=
z=L119i<nfJZ5d`2)=AJj1BGR$aE?z32KHe{mqg4#6fk$qbh%6?Z?<)M6N^r88twRu
z;uFW>xrq<H=#P=r4VY=gj?Z~UQgH9<sG-O?s%+VBwB&|<2`i<fnr{w!r%Emusetd0
zsDqR&04>VgYHEl!Uyw#1Eqc*{q0P1P(cS52VE#A_`l{4%imrO}9PBXq=im7FUdbVF
zDEF8>2~E?~tFqK=&2->JTg-Uu^3+(=p$+gH_;lSv{E4~{-{(bHNmE49PmA$Tx&~{l
z!cr_BEs^PX1?U(~ljAFY`)oaH&DB{R`Z#anTEMbo4p1Nd%|SG&Qx%Q|A3YF@D@=G_
zQ7XYDidQN6qFrZ#?+(+-?P6S)30ITsvcR>K#2>Gxr#7qWpl&OA%{GE&FZ|=<Fjg^t
z8X-%-lx}Z)Dv;<|e-9xLzgtJry;^Q`XqI_^Kg!Z1fK7ldQKJ)Ys?W>wh7@bArQz9x
zKxwd?b*}J%D&Bgl4Oybfw$<TX<=C4VU8JJ88$=L#dJt&++qi#8YK1;~g=#ZLcRnk|
zz=#id+|)o*g2rN5)wWRe&fU1^-$nbNEvw$?xq3##nA!8(*deK!e$i6shM}f+CL}s-
zNltKD1o>y*zLKbaM@&6k)K(Pwtf9m98R>8x%VrRn^g^7rg9br$(9#iRHhNRSWcf#3
zs@=hMmC593#@~EVZo-Q7ewVb<l`1P`w5&JQ!3pPSpsW~dd-7^U@0Zk&YFO)16wvqj
zV4r;|eFc*2!n*{~XQ|<tTpo9`e#rjkx+@B(z$?@9tpk<)d!*R-E6n9$$q2Dd>Pe>F
zZxWAlzT@O%*X4B)Zhsyxeb0wh_f5AJpXooF5c=RgY{}}35JfaEsk=Z}2sL|;=n6!N
z>fnS@-*Tlh!Dj;49lAPD;_BtTb#2naDg0~G6G)zT+xa!ieUkV7KUdHw5QUrIFx}Nq
z5P|e}<SFJiiO<fCR7W7m{$Z}3p)9Sp1_R>DnsPcSFrWr3GN$Kim%Sl6rPz*xN|uGf
zOaNHN0GwbICeo60(k$`ADiAunm+~qHlX{A_2tvHSuaI($bCXJHsMpPZ&#RPt;ftm!
zUCR0#0OIV8G>Kf=W;Rts?%jS_Ea+2Z%J{Gy^}RiDU)?H_GzTgkOvWpqlvRzavVwrV
zBRe@8z9oXu!L1CfeMPP^Ajrt<*C0#lFQkqe6Y<N=#p*#7c(_5v!VCeM*iH3!kj1W&
z56~DH$T8C?nc=PHcQ*uW1{nwh+TlzqOq!5cRX*k7yIq<i@N5B-=<MM*TnHMn_cR-6
zg^aEk*_@+v)Y&1m^_3vn_g(Sm;ok73EuE&_HYP9X0>w=0-HxJ880BBE2FsT9WbD|`
zF7v<pw`=ZdZ|z~iA<{|KX&T#?Y$BzUB(0G(XKap<3YMETr(ju|*CEO!`kUWVi4r;5
zpC|3%4?8^F%4W?D4G(waT5biaWVw$1TlHsU`M=elo%4SaH4kEs+mLczTvg8?s?$mp
z5C|bHjM{)aB#GIGA&Siugv!ot61=^WKjLU$^Sa>Y01q{Bp=~Gp9xA~!UidsZvcEr_
z%7x)U0V^lSM!$jW&(_B}kpu_2#4j0EdmK+>Ub=VeN5t936ZH3$_$oTD3>?7{`d|Xv
z@xVnoBcSB!+Tk6KHGfIEy`&e(QK~$bS!7Le4G+4CA>4xJT<#Y5>R<+d=f@{!-?3MB
z@!53Qngn;SuHlqt>2ltI;G1Dq_7#wFJ?^_ccDSdM#vDcT9I<SnUZ%prts^BqOY*t{
zCdetqq=Nr&7%kq$5UT!N!9!d5J?kMpsM2c%35TB8i~_Cco@|d5($!}*nCQI+jQHnf
z2)GnWugh3kyt~v07q>ThS<AK}wP~-09abxt@*fGtXb@Sel+p<Lbq6!7*yT#nBAo^j
z?};bxiD7j!a%B+26;1w=#hme<j>oMv*9Ac_RX7}-le(-^P+OwPQJ*#it*9MJelpVm
z$U5Do>hbH2A*A*X6>#i>LY+qfj3W&{x4Mt>zKZibxRmnwW<sdxjN5YV*R9t<r$*2V
zXlxopMff#r(e^kn1mQ=wBmB4buJT`GFq#4f86~Qh$ep;)ag#U;{rk}sOL-JOq57t|
z$)?^u2?Vq<f<b0P!^3k-TSp<bD+d}TjQEm(3fxJJVi7M=C@)>do_7iQ)g0=O*OEtZ
z$Rp2FMQLKQY^CRm<V3T#4mYSti>ANNJWjH=pZk4gMlhfI!;xu3de@i~UgbbjG8L&d
zH9KyU3$ZeGh%6{%OQ{>83fizF*3$QFLy15rEtS9KWCN|`5fWyh0<vtKV#xK`>>gb!
z=*+_1Q)0|SJUzN<`H&&Cs_CZq8mAosAv0xCxOMkJd-!^$(FCWEU!6kBS@D=t5D~UG
zQ;oKcmL)rnSw-<Hb?#TOllxY`U`xFHQs2qHBIQJWvkt~SX0PN>E|7n%b6)zWYwt)O
zza+R=c6H?3@s3w3(o}NBY40tI^5mvd`$XwU->tABj^w+y9!tCS%$dHpeX<;myLYAi
ziyeR>Q@eK^411<AWCczpDfLx>u&fCRAGOy*?)dBcLCLV<^>%*vT3oy>bl=nso_OtQ
z4Xvb7?QG6YzH{)Vx;(Ou>P{Y7dq2W*5*CBJHiY1&gnO_Cj5i4UCYvKk>I&(yl2Hy*
zs>u@wykLYB0o!o*&H>V{k}0N`7fwu@tMoh;_uQ;gmtd;Rsj~GLy2e*NDbh&ejtfSf
zCd}pjrBzRJLyp?=N<g<Rv}B2^F=dj7#-)lNB;k#za+!auva2c$8ADsl(_G~k2<O3x
zG;e{1Q}UxwKwL4HOC**~B-K^U=*M!*M_6nd>>Ix>UT(V%cHKnW4xYkj$r3{dqrYp_
zSFRFuG?uh)H2n!_wr5II7Rh(U*=8)USdsM65iSrZ&04opJ=@j&0&1~yp~Ny0xPue{
zD;}($M1o8!E)p5EA?H{Sr;Mspu#KsiV6{L!On$0se&&)96hKrPwk~XEGcs}4YFVVv
z9vf?qIo!6b<TDSyl#EAgyFmEKjBMLgBpl)(zu&lGt2WS%vg0Nk#e3BsK*N)zjT|H3
z%{$PIPIa5bSVV@)BSEpc`gN9@(~BkPYfJwl-;j>@D1Dke12o}Z)ftX%+6WAj2WU_@
zxoktI;Nn;@$J|<&S9><hZLQ~%=H)k?D*-o;nV6gOv%6vU*X%cQAU<kdl33BNnA0c{
z)s6i!viKAxpUEk$v0TucUMd!*XTRowNFu7}(0mllOeQ=loq2g`7SUUobX+gJd@3L~
zqK9PXHQ1`V@TSz!mPrfr<Y=qruL$apL7+mQ4r=uBG{s3dW+;pnAkerEJ1TgEBQ`hO
zqkHkqoaRc+QOG!E_4*tZ9X@6axpoTqM@`sDwH@fiZyvIb@~mfGI`2&&n*BFvs|N|)
z*v7m%M%FLpNjT=c_aKQ`!Rp(R3ete_X`N>wKtNTyOuk*IX0Oo8P584JYVJly5~_5A
z_5va)H_j1O7nQVsZnZKgT07d#K#<zwSGjTgJaH%U<~JJ%cPV08H#hYcxD=bo<A19-
zZ2!07u(17HfBe7C|F1|@($<g@rR!a-Lr|Y1c@F}rPhDz3$R_iFeuVfW^@_UzGJb;C
zS8%X!goH4AErycGNW;sBUJu<e{_x2-e*GO%W1EqXaBI_QB>{QK{kkbOn=@FR>`Qen
za_8dchnK#We|&iypRgtgjp^l|u(UJ2zYV;7`8q4qgFm)Mo8|lZB0_!@gaG6twxP+`
z1B%R9h18%%<{=#vUgFaSFakZ5G+D(K1=budyg~ZcMRF&`;zjM%`-;4Mpi@<IEF~ZJ
zMpM~@4$`;yd0LTIj7_7Jb>Qf%rpd@RKU$@X#NlC@#N$r-r)GNyE{2J<hM2m%LwBxO
zbp;96TMyo65=G$bL=G^n@G!#gH|s3VewBG=x9v$ky{TnB`|MTvUq_a6FpgWuit>8*
zOpBCp`4Hz+zdgaIO%@e9WtD1;TGzJ4TR}IilYKf{QRb17%WA6W#=dp7_~^7|CRP!&
zw1%!my!dR9PVH5Vs5PTO5lw3lV0h1%w>%Jesv({0rP-n+IkM{aAMY{;HfRaY#CS5T
zCaF5?0!o8W{kD{2iJ6{G2DUFi@{#pFB-!j+mUz0dmClU4-+-*>_p#Wsti?`7`Ipbm
zMIEwFS8NTKDtH6ypgMa{xF8x8KgPK#PF#Avt#Jd#XDGlU0^kgP>GOv;U7n`&;D0+5
z@MDEIYwo9gHILNheGgf6&D{eAB;hvio&a(+*{>^ADhUOzPcJ@?7|tkob_=|TFZQn#
zS@kL}Dd3_~|D-SrwT<2S7)4@f@uQPmTDM5$>5mG^c7GqTu?aG!=~pCCSSewh2lTsQ
z7H;`Sd@jlRTxg!_2lvacn8Tx$b_<$v*V&Haj56X!pEgz7v10**dLu+48qrH<uJ`n>
zY4^E_5Xi`Dyv}iK%JRFphP`KGW*fwCBItNyr<abN6-Sw~;wZt9s(6(SOb$-#zz;+q
zT~vQx>z?)WU!ml&k*Ak}N&eWMcx`pfcTmU<P*hIAQv$`zW{UZvpOD%<)FBQmcf-Zu
zMWTIO{`y03<LPFQEaHKHRy0Dh8|$<wU&3ikb&)BUq}|oc3EJNcgo6Ol-X6fCW~cz3
zP;gq32Du?=gBm?G9uPKZ)-lX9dn|~XcdI9Y3)e6daN_6{Rs)&FQ9u;NI&RTKmTQYq
z354$PTUfrS?sP1P=-#<$M1*NJM$*^6yi;!?3Sx3}k_Z3JVqxiAyH_pJh@(JG1~win
zIr8U_TUO5nYNJ_2J<)mRcGF0Xq3R;lg1~2`OrQN$SB*r2KY@<Fu9O@!LxRw1!|n$l
zzj$OOOHyzl8>Mg`{h2ea@OQU7mK^<Xu<CE}eR{3iMwqLkn!ox;M2G9*#71`>lv>MB
zS>6h}Y*jmJVGE70j<tNOS{AKbh+}v{S@|8JlfP^;F(skr-S080153483_&L~QUwq<
z)l(&C677KIvH~*6R{Q$CxGyz^5F||t#b_I<F2=58kNF}}!dHRTXF|+jLOr&)y%xNR
zuIl*9Y?GZu=(GNJ<DGt9^idXv=A-t|c!D0|&zuT%Hd_qVunfrl?xt_I*Vk~dUSf~o
zx3tzFgs>KR|C6==4<*}9>!sx<-W*?V9vq3(zW%+E5ZhQ|p!M$j6VMv0v6X(g-cMOY
z60&qzp)v><j*bS_BEd!(Y?Auux)X~ls~_QEXPzq-EzFW7yyeT3BAk!(Evj@o+g|TK
zzT6cNTjaT%Pa=ZdJyGh=%qgp}coMfaT2~S^^g|4idN|pZD3{r8Dycnk)mP{z>QH%6
zBi(c=LNDT^C#KI6l70}VCfs2<ji8;Byu}e~TqO&UV<@9(1K<4t`5b#(!De|(Z^32`
ztMCOnuaoZ4J1HHs!t<)-SGQ%lAi8H;{KXN0rEClR2$2>2dl(uWuHxi*G3M>Y{6kr@
z+y(cW*!2hYUvgPuBSJ9oD+pXXsi={Ge`7K~?-wovyQ4)XjZ^BYEK)5g%(nE*wf8uK
z2XQILNV4#pz&faha~4=^lcpKLc}^0AjZTVchojp)H=b2}nlyIUt)bIyjd6)Rv+>aw
z$BJiTmhb#FX&c96mk<FKQiXuKVVNS=oUO_;o;0^I@bFj3N>b|AUDK@sXlT4jxubZl
zP07b7xv$Ey#(R3{^5!;p828FBHoDzkH1Xw9u#g@+m<v-PC|w?GIPSCNm`GR*iqk9*
zB`wa9CUI0fXW<Oy_U;{On?JiVr>&6+#4n1JFI@m6z;4)@Gj+B~vCYYMyL1YHL26np
zk`6dH3kNxiNJj7WXsT1HR(Evv!RA@<4*AXtuR&5aI_$b9I_5W-n<l0#h+P<U9(p>Y
zF!>*noo=(OH&n;lBvh9-rJbDHBh){8>}^9VZ{h9+Ojz#~DoQ@$7!&L?ME#so%qgHm
zwjY-wv<-e8)Q|}7@SCrR*Nvw3pDL`E$Q`ctFW4Jsssj#qyi+5dGPsg!xDomwlVLVz
zRV)zZ2Oa_Oqe}}7N&7A<Z`4-mPK1Q1#r5#}{BU$*jN<gWY#fg7N3?VyzWqjmaP$|C
zvZ$A?%{i(JCDEt?eHFGWyKCrOhc|~-A{tqaFF5?WCju!84p_~YCe|eTwX+7B+!T#`
zY%`J9O%&HDp}y7|&gGac(MfkO?G%FbQ?(o0-Dkk-GcCNUq{Vt3GdySwc>IV!FC{Nh
z+`ZGcT93fO)9(+Pcum@)Qm2=n57379w!E}lNaDvg=}Xe{C7hwZRtGPieXX+z1tvsa
zzg3Mg3LRHm>>JRU{IC`hL6BEmt5V!E7Ob@6u0qH6_qvZN2Wx{S_=P`%pUpH!4Ol-5
zBwkHhS@uxzDgG8jjKw2@K=BGJz~<J#XYm&%3oV==BAF{+kI8LdQgKu3GxHgyEa9~4
zr=z{RVzz#w5`NVj8jL(iY`t}8_%k#VR((^dm<|Z}Phq75^ddv2YH}HV#z3SYf+l#Q
zMi-4Wn@5%X#V_y+Lirr~=c4|4TU8l{z<jl#(Tjv$PHojZ(-7D1mf*@5HIO#?j-k8K
zy-wE^#P>ii14B`gx~5gN@^jj|O7-^i6%y68khz0ER(Mo{F!>?A;7qs=$-$uj(yBlj
z+aR}5c^)*1O+l6maflbQZf#Wzjfku-#jO<g6@u9;@9%dTVF<84qhajs;@IUERJ{9P
z`ftx^L}b5fT-7zHLVavFP{E(jFCqTS7tVcr^KcUTf^2x9#r?Pa8#~kg=@_swvHWix
zgH7Dg80{Y)sooE#kbSKL+>EWLe8c)j7wZk{fb}0;EM)2^aW|nwuIMQ}z}gj#45F+&
zi$Lrjn(L^;Yu(7<=jn&T+u`cUU?kAk`QhPx-c9g8x71!v;2<q|hT`~m^I`R~l0|4X
zt4~G3(9ZDwwcqxQovDft6!=T{iz1@y77ZSARZl8xp8%y>z+x;{M7^`e{&{-I9bND>
zT9@+Ct}UZ4_o&p}{J1&AEH+~W%y_YFu^<MRSGUngJQ@Xr`d8<HK`z@$%kSNUJB051
zUf=O-sF!iUpAzsmZ!5La;7hL1s-F96u*R!AR+FeXC%O7@73NsPE)oU#wLa1rBtRS}
zZ3RMrp0d1C0RjZ~Gfh0drq^PU0+(Fo2Kq0CzKEzu--^V!hu*7xei^-D3K7eka-Y(1
zkwk`k{hZ<3jVZ{W#<#j$cJh2E4Zv=V@-GqW-pS2&k_!6OKMAovafQT}U^EGJWQix_
zJyTT2&4DnO8SEnlNV+s1i>j7<^HBTgAxM}8Il1C%Y+a<G;c9s1QZOKSno-C&&A<Q^
zz$pgw^LBREqPyOYN9#%FW`o)~2)ucNJC?@^walsuiDR-^SWUl9w}DkbNwYzIX0^&J
zll~pYekXtsLtNhAqRR1U*Tg(W9=L|{cVNld-vIcQ*S2tj;Yegy^t1#0c*wh#@iww>
z5_BO%qw*l4h1-YA5xvG{Xq0XB2SvZHDNPPuLrYeqZhp0shxhO;A-a6`FR|46s%h)S
zs9Tv8Tyd|u=~EHXRm6TKWz`lL%Hhy<xvpoOT?9(f5DG+1Mg&$_eCNaX)>~R0T(8;c
zfpw7~sRtt%zS~cv`fRoKFYmPueDI4g6||dYW3Y0?y_n!|+`TO2QD+eIsMW*T3$erJ
z+Ve1PvQ&=nRt@P8ep2gubjsaR+7QS?l!I0&7E1y>@>t!o4}t4ieL1Xl@t3yhRLG*`
z<`crJOx0&l3IKPBomlqnhyDToGNwEh$mxCweyaFzBcJ`k62zA%j?A2tqmvhtop1Wn
zX_a@oO@jrfC~1$#Ghrgeuxo$|)@l9Z7Q>>)NS$?jZW&9h|2y%jEBX1P@u3Dx=kQu&
z1I(e}yzX*o4SIPrfG1hf^pW9`C#Fe9qRvZXIk`fyt9QK%wIJLYVwroBuVOz%62%D9
z{d-(X$C!OpWj3?vtMG0AqoI(MTDft@l$NnRSJ^^j(PElfh?_+Yv2pLR06_BGaLgdK
zhjA_$r<xA0#gH%7zNV~y2YN>mG&r|~x)8%z-jWjH3S$yoR#WSl@CVd7=1QO@CWS#c
zs&F!!OoO=u=aj|=Yb%&J1br&)c*+5X>OK4eC23F6^T=-|ofkpMY`Nki&8n<8rr?!N
zizwX;FAac}%NcVB+GWbM7++;J@G*n6%y*nBHpXP~>n>Fy%v(AwzmLv5Db;Fi?eDDh
zF>jZ`N_}CHqE5&~%xn@jNEc|Z&!f$VZJJlYxFq~<q`9UEl~sWFf(aJ=herqlhZh#Q
z3=&Lvoq3TLUzTyL-;b7OtnU8KCrTNuylvY|A;H;0OaoY~>#pmNAKZlWn##_|75<(E
zmD(;AEqjaruD0@*omO=@2czEpo_ka({joe+YYQeRD}!{6`2w4Vq^mdsub|D?6)eMY
z;ja!OY3Pt^e?mjBL5?I5_$3YJ(LcJRBK2~2FO@Fp$8{p`vo6q5CTT}xcThC)lAuZP
z4<i}cE(2nl;8%W<^kAUsj>Y1#2|UgyE?Zt+v1PamNoN+8Vl+z<R}(RuGmGB`u{K*S
z(+o~deIs0k)r#&-n%XKUEdg8gw`b@ToMyF~v^{`orb#9~y%ewfr)5K}z&2*jd0kd`
zlY7YMJjFT0OZsVVYCVbcrRV%Z^l>shk#plq(|=AQng>D}qi(BjMxMbCS?JoUzU6jE
zda@|pd$ROP@yu!Ns<|^vyu}_9JbBYKGWC5*ohEJ9Mse2fn>&_;V!CC0PifZkoLmzO
z1=k63KX^Jzs4K!Ep43P9-=0l%Px)8DdW$~1=Q2z1J>};arG(CO2rQkM7ejd!c`O|a
zbY2N`CL=X${Wz2w9E7~fuh>#|URERVZqH-4FxhU;p`WSyTZ6I=rK{0l4BSNzk3JqA
z9^9X`C>gT?ytDG#a;IYEI_Kb-1@-o?_t)dc^P$+5k6ub68>BM6LP<c(QC=00xQcni
zQpZ3c4dRjwa%_4ZOSPrVn-&yN49(SleIJ9(4@e!yQ~0<+%K<T&40@GAP3)fnVI^uQ
z^>Jivk7gEk+Gvt;&_HgxJ#<)UO#F!O<ApZGvUIcA6Q@jl+|CTV4bbsP@fW)Ov&H?0
z?1n$mjNLP#lO&zhMesObs65629=vAj-{h{M!f_wKX91(K8Q8Z{02|~hhDT)4hI*)k
z&#O(5Ug%&CZsLuqfQE*?J}H;!3hdhz{FN+*c?dVg3RS_pak5+FcrVS<dnvE^wcmxM
zgQ1CnRnP^+W|mEqh_d2e@1JM}5IxKEe}NxVu&MaFW(C0$z>b7ihd66O&y?X!p7k#Q
z)I?gJ@};CNBmWc^uneMl*Fa;uFZwg%sextJ%n9Aw|01}?Rm&^9g*3DaO_Ez^09VFs
zT_><Cvb4c9ft)lRF^i*DjaO28VB=7#y{vHS8j8{NFQSqUO2O@fPw<3~ySZ67olah!
zw(J2;O)Uj(hX%ygPB`Z-{il==tkg<W|Mca5(php5Vj|B$hkI~Ou30jhKbRn4UgjyJ
zNln*pz**wT<|{!&)mJN)$j~_Ivqu)#s*i)b_{SF?C0|&xEfw*{6^dIv*5oy;lE~Ll
za|;{^2MQP0?ET?tqJHB)Cpz<M+qPc4w)YzeLMT7v=-+8@e?Q)GKeryhk-N6~v?zx>
zIh~l_c71Y--dx?d8s<=LwL#O=KxRsjbJAX?tBi-#n{1Oj)XY5PoL&pdpT#G%EyP%w
zip!2)=GTpK!P+|-H8kw7+&PIyfZC$>^n7oCt-u5A__JOC4bo$6r4_Au7w)r31#ynP
z<WC#b2({e;QZyIj0KL=g3=T)fb3wF@&eW56H}ycPkEE2X^hB#~t=|H&Bpynu@9oe&
zu=t-{K>zn9I+F@9UGt{Xs-@?H?*mK<crL&zo>rd<6}hAJ76I$m_Y65x)4>vX#?*#<
zCwLMJ2Yr8ZbB|JO(TX+K*%G&eq}i9VV{_F6@5K^1Rf?@_0-%M}rQB2aBcr0nc4M)(
zM<IeH`Ye-<FFiNc!Ia&@&ghRFX9Gul6x=Ps1yHE-s;umsWmrp6ash~{xL~#LU0L%b
z!oH=DZ2Wm=RA;=TXQC}JVRqCX6#BUt2!1D+*{P`0D;ccDBFrJ%8rUU63J4Qv<#>of
zd4WnEP8{2R9z4CW?u$X}o1Xq&ai#o@;PI0xu%x4K&mA9f{_I%=_4OP1nS+-oG*zlH
zlgTQY<*tYEM*mRqG*RZM8unZ{<z-yU_9Jq&n$iw+%gb}B)S6Uat7&>aKP+Vu5z)7K
zRp8m@t1s<^!)K5;pC;G^k2W+IHGv&Ew{CaU&sHtyjEgju8Qodrzetn$G)PiF`(nxz
z*Pz7RRK0DVU8{5|Z`P`qWRsC7=pEA4c%B(>8>H&Z=AQmdIr+KU*}Ha4d1=RqIjzV|
zex!2YPb=9!Ivqq5It|as>=L&zRX=rx%1B{PXe@Y*1ik8q-6$iZqA~nfs|ShaZKUfs
z1Sz8=h6RpJAK`D-q_S11o8wd^7#+u5(-!OE)hhJXPw(ICS_IzUCtf#kE2@=A#VZRc
zHg4ZqVW7)5lRAk#MOZxwd2#l0ts%V*zg4X9e7)r(ImLo!XaY7a;Y1V&Jxlr@{m+yC
z5;xjEG>kN^o9q-rS3okmQbWjd&+zF)taK5j*o_=Ve^>N$)Co$RE~_*;#)Z1z`*^#L
zt?h?dl9b7(4BTX<R!Sz}O?J%NVv8Yo3ZH&Aq-sDS^h1<eRs})hiqxedKQDlpmJay>
z)fan8|JSp@f16_G{@;4L|M7B=wZeBjSC{;dT&NTodzj{~oDu)7(|P>_Ng+~44Jbxz
z5MJMN$<8zut)C4!P(h8`RZzr~4kRMtw7VD_URykUHodtlw0~GG8JfD?bV7t=pAeGK
zOLRXh7x?k>=R?L=B^v2zR(`6&r3bFJ?RmZXhn!wzM5YHYkDb$AVui@2)@%WVZNm$_
zw=(|5IWzQbt}hg#XplFE2l$)Q>T2IDNfV8qS)%US>5r$ows&hzSp8^}oTTY(*IImb
zsLRE?6GNu)YhS&t3xP?|bCtM|13htuZeN|EBt4D5UvUtHjYDuaV3HPtgKI=i9|`PX
z-M`I$Gmyy=jYbzpf(J$s#t?8&=M^+Ev5G0Ru>Fh6>*i37wT@eo>^2$$A)P){LoqfZ
zRy?Ca2zF+gW~mf8zZ6rT%{sWfbMHhgQF{Za=eH$lfA~VUk6Qu*Yff9XnJOoPVWYj&
zUwHOE=I}7@p4$+SW^}AY-VdhOr*)+K&Xvfn|Fob#=3k*9Pf$%%?#I~6!)zP23Vx-3
zD<BpWt-~|SQh5+Tt+XJdFMpa&>QQaWTIO+J=AD+-C!TeE2xnmN!0L~cx;HsnSE2?d
zVR$D~Vf$9VqEprfPa~@_hM-Zw&fH|=mVLFDhA9rfK<Ja=N;i!bCUc(K+x@B_N~8KM
zoEMJr(I@UiJb?`fvIukBApi6zv6BJ)OTET*pf!BYJi{=t^+dm?I}S@^+|ND;MASWv
zI4;slobVU!_Vk}=KAv|$r?Gg+Oy%*AZ`^Be1EJ_T>b$%BJ3W>tNZMTzd7AtMLk`LO
zvvrDS7Lm7zkHtDx_IeYXJMv+y7ji+GgGIRpHz#41WV=*}#jRil6FFE93?03EvEv~4
zT%JSNd@4bzYri50e<r+hysCLQ(ch;Y`pgymB1?vkkQa7WXc3wQD(aK2jA-Pwj}dQF
zVp^HSQSZ#+zsG@at;642CzQkKTBgq{joq3VRvZu#O-dZlGC`!fs_8*Xj=$GCMnn16
z8J{{05x0jsWh%`uaI(We-B_t|>id*qxXqQjGIT&6YM(9OxdOSllbb;*vm`Q~I_vqc
zp!%>HdqrkR>W((uE$O#%!?IvMZF1}L95ejt`nWTw$>e$fA2cosA}uujf}NM?oF*GM
zT`x(CMlhpuD*|!!1gj?6^50&4Gj9~{Kfz;YX!B&y3UrWfM~_5Jg&~bcO+ExIFZz{=
z+A&Wm?HfF5A6pXH38P|e2Jtj&Xfy>AX~}ERPLBom1~FTzkacCV$P&W##G3l!Ap&(I
znx>Gl&yJ_}v<K$OAYShy+I>v+DoWGWtfwk8fri?IOQj){f;}tWbLsB=nUuOuNuI{B
z6c1?BBZ7!%)zK%uBLzD*&t(*<ZlDy?sI6dx$n16DMmOoFa2`wqO|_qC`eZvnEvI35
z9oWuMdqKy<&qK-d;dV3JXw<O^>(RPv!n(@x{i358@HN|tC73e4wtL4Ue?zrj33Y*K
z91yfxt__VhGkgVN^CrdVwWt1ePqleggPg{FhUtqf7~2*DF7|Xk=mF&jE>5>!SacN%
zLo7U=@Xx2mTrSo{cc<Az#>mrAB|e4?_$Fs{z(<%AQ&Qh-{YjM8(n%Hn>c9R}I|}6D
zp<N=p(Ueaqp_exoW?$@~o|o!3Ud+70m)A6G1FR9*L=)pIRbg?PnsqMZ=yS3kOFj$I
z^Ii@2OC$f3ZL1f%?W>=!RI3m|HaBeV9h0vepR~+DTkHm2)kmRZwwYOKZIo>^qzxCv
zPO&42XU7i$VY`xW-8)v}ZRQBAP0B3PBk;Bxz)5EfyuxB4>*}hN&2K^9l*-(!3+q4I
zONL1Rl?QT;`z9^W`B5MuO9N{UEaaS0_FZ1}cX#=`&L4NrnpnNQU9SKbe&4|KGH$;n
zuCR}_9O8Cs{py3*`voSfqm+y(WO97gvv%XYno9z~ei^ED-n@bX_NKmV7N{KDKNw>i
zP*r+^MBSvCb*;6n-dd(iRz!<?ORuzBdz(0vta!^V@acIfmH$<H(M8uTUjZ;ojMV9v
z2nsCBrmqYTvcvJ1OrnsAFBI;{OWYa3k!yImYpM$9BeRH`fu;2SDz68mN=s9^nv-Y-
z3!9f!$X-%DFpWylFaC~ZYKTg}=Oz)MfT`H5*nRpwzUI54jecACoOB`(GWQ>1KrA5w
z6o;#p>1U4AwR)cNo2GXeEn>R4{5VS<RK;89evJxvOlPK$DhkkNmS~tF%9>|yRhs3<
zX7{>MdB{bi1wgYaX{w-kSvV%QCvSnb)P|&9zFkN{J$tWRYxQ`7R<ZXsX1BKBYV-a|
z&q(#ytRr}@;Yr}EW20;tF@7ed+Er~8h~`(3&DdO6x>S(GXrEL*LRGSXHSBchr)huO
zt{XEkSf%-HjoRd*;de4P#W#&xgUGG943Qm|siM!ee2X;VP=3XuevNHBa6G&pxCs~L
zz@%Ugrdm&2T@cr_S&<tnqi%|0AkI?$R`+6`P9|&sCj~JW8~`v($dN1IN{uGwP%7eU
z`h2?GspsG&$kvxg+E|&Nd&WTQEpcZh$*7#qu7eNzlt3{re~ezjH~3RsFTBDW&WDI|
zp{jy($V7De<vj*uEjmmc9}W2MOqz%CiHwW$>sXRjSelkME!@11j8H<L`rT8D1fRee
zwX$D9rc8m~CJcuF#&t4E$7-*4(9ZB!Gh#fO(w!=(>|9NM1Ei@Pcgj3AC0Cz8MFhsn
z9IOcx?vl>`o@0Bpu`RBsgZ67}dbh8k@Jte(l^m-|NMXyH`Flms({}n%2-^2o9Ya_T
zY+j)gDayT3*vR>VZIh*VY-3;L4!2vz)FkA~6s`+gyTdqZK>^SnI{OIGg!gT@xW40q
zXTr;Weu_1&(1a<=ID=+VyX%Ra&b&mRGQnH;gQmvaJ)r<$!on}DA}5bKZ-TJtQuDf3
z-*gs9sO_SSLnBJ=$IYd}dNjG=YF67@VOdhtz$jOk{M9X!#o#SWjt-_lbhqwlhpIs;
zL^D-x-4MjetJag-&_N4X9Jw}E-refS4xz;8Z4|yh^1bryx~rSWy>_WJx>TdKo53jW
zE6Mxk!Er7H_D3(J_~N1!25Vij(Lcm9H$m+is?cY3I72QbdsBoP{P_jSD#8l!C@8Lm
zI)<^nv3(OU$R_jlXMEPFWejawp$b|o)hPrCeu<?doCEJx0f`5tfxsYYF*r?iIL_iQ
z{l*y>pE@-Ay)($duz5OHDdeBEC<tTqHB}xcy@#jXqwBe(-s{M4j8MOtNYaW_9&c4w
zomipDwQgVY5&-tx_Ff|bT2Am?>1qIkk6u!T0VxS3R(`eGD?X-4cO+<eY8uK~1{Zbr
zmrqFn9Gd7kfK;L1)L&;R1)rLOW66sH85R<vDsctTZ(AG9=Cfroy<4->ZMM?NFV%S&
zJC<oHC;vLVvIy}B;#Vu3Ybncf&`=nMUW&JGa-26=%wDuF;$JZ&ud5t>Np=u=?KL&W
zh*zlH;WyLkco-yT5g%RFAZ0KuAu=pn8Fz&P$qZ;27{5qHSdg;XE-Va9`8r8je@Tx{
z`T{dX6Y>AIeJuyuf9ZKM|Bsp*-2V?A#TyumIKP&Rh*R(c;{fSD9q+)7WDy@hvAPJl
zx^~mrA%s8rd=?xU0J^0qoK`QLgFBDs2bbsF!4>kb6P20q^<=SLSm<wB=)f)dU(~zN
znp<6rzV3gOJYV>Fho#Knyg)wich~v1aM2Ay9Bw3U6l0%zX|R~PrdTCX-s6tvwY~im
z)hqAQ{cPiC%F#NMnfpCI&eyZA$>KhC2M3iO-t@usEgwKsU=`?QEP<QBW#oFFftgli
z0(p1LvR*~=G0a?W1gd;s;PgxrA8#)1>KPX@_w!tJ2TB-n$7ni#0q`$X_{(k-x}34M
zi(^y?LW_$d(s$GgD6dtzdg%7=sMcLmqZO>lxZVu$3;EU4x3OY!cPE^<`7ao|1)Gpp
zl4Ya&7F^HnwMJVWh0p6U61;K5PiuQhKe(ql0Ku5C&k<CAc&E8w5#;f#8o-<KXUrNB
zC1SX)-#k(h5C4!uE=3gZkN(ndzIeclN_vw1V8F7@QBL6c24DXDF|=va87U>8Qc08F
zbzMJ$d6OVg5FY_Zf^j98^+V|s$Dj4;si#v(bYuzJ51Y1FnVb-mE>=Eri6S#Eo7`nl
zGr-51^xgt~^)+&uDOKCjz=MWgw-<r1S}tIZlBDI7Uorb9Jn!UrQao>GRQg~9Dp@QE
zxhT;<WqOg}Ktdz!HgSagmtDE3r?YuWjxVz!v#(q6p4K$2aO1D4;rxJB(C@ibaURu%
zvaZQVQuzx!h~LqUkbO&J1cebT$+>|KdVgYdBKklBM+pdNo1LPxHD^uEQsjQSFOg+x
zFaEP|&PoX(Xa~^H&bkm673svIb92xJu%=hw92of;wkJ-^9dTF*Z;IuxE;JNBwz7KM
zT=2>t)k}2AwDBt}T0!nxWzZ>^Eih3FQ-N_JCk%Wl{Xm+X3TCr0yzA24+H6<n^f5c3
z=Gfr7){a$it(>0%dH6B(B9MfRL8T%jB&pftIPk6g&_<7ID1AHS<fm<2yQ_aUb3hk=
z+`Yc*_!vZ^P@IXGGvc{r3VwBYwZ5|j4)#dFKwF;q$}#wQ=@k9jX3~{GVUzZ_-pJ9A
z3TRs$bXT)Oc~(URxe<j&))`j#Itw!gL*k@U6=pkARYw`?OsjBh3^4qFGFZX*Bn~ju
z{l!2Tf4bz>R-ogtO>*G6x;JJ8nRP>Ey#@EF_QVxc!*5YMkcRYF8!2X(AwNK?uHc72
z+TD3YknAD}mTW>tAseu{cz_4$>;xh}3n4%Xy7kKvtW;MrNzP2xdHFuhH7Ig~od!C8
z5f1%{{?NVwmcg%@Dbi8ZPpLT?zgj@!uv0;I`fo!<%p*+L-%GbJ4vMfbBiw}U>ZJyq
zPYMQgQlcm<wrdAs^9MPFu;l9;Mx0z>O%YuUKk1itX<rADRdmdR&s|ioE$GBPn7xuS
zZ;p(5t^4a>F}cvI_;PSfX|_y|fik(a=ihAh-a%y9YD%4ZJ3+W5Y||&C2k(Xc1przm
zeERJ_vxkVDL;-MLQ-i9eJIM9<P|w2CsDU*|RZ~c|IjAS})>3`vZb`5;m0x<jY*86=
zun%7x!rJgIpPO6vw_D?Qye7~Bi6ZW$)9=1P?8_==YZ21fBeV*hRnQ4jKmH!RUv_5^
zROZR|PD7*v&qMS_uWv14r+A8TK?w_vml10um!dPYAwo*d>>h@}93H{jnM1anzb(sr
z!MYvoUdl76d~v#Ce^%K4#9)Nnq7{WPhrECQp&-*vpB9lTnMbQK7=F?YSL)o|ppk-`
zVBAVMptKmQKIs=nBrN;x9o;`#>Q)x{v}v5Ur!a88Jce=|vvm9=3sl~;YQt)vdMeXc
zF&=+gplZHfjBQ6)kzqQVBQJ1VI|X|gx8`o?8;p`K6)M8qp3=%3!d}_bB26I`bTHqv
zTKM82esscb*QPWD=iu93esh;A^<+w6<%?K;%n)j(0?}3{ECohLL>vZX$@DL?kG#oG
zWs&XLf6dVuC$k*_C&Q(RqO;J)6rOpbo?EihE|<zlb<y&HTo2=X2jWt_s8Kb0*wL5)
zw9TQaC5Ow1IMP#_UuHIeB`fu|t0c5T#9o{$B714Qh&tl64j5Jad*pL6xeBnNp2!HM
zEt&xXi{QGACC1Nz_<HF2=Od*c7l=44mqdg&1{_ACP`2}o%tk|gpsrt(YE>q5&E?v%
zf4^N5%ftU>tF)%on2j@7TFYlosaE!Y;5Po*a_p>2vU?G4tV5C~%Ti}SN#CSBuQJJ~
zRTafm&G2lcK8|7qvD%6vzpjQ3wOl&SBf$yNjWP*YX>HQ@&AJAY+y@V1j#|(e2b#od
z(m}XtoG{c7TIk71C5Q!nG(O;pF>XTDOq*0mq-Wz)6U8bLb^((q0g<bBpGhr+UAoJ=
znM1|?tL^8r|HUmvotXgjM!8TxG=|z+j=NM~64qQUX`U^b@P@fi&;UwmJoycf9m_12
zG#;}>0XqhXI>)C@b54OHE;jB&VN48sJ@S<6(lGrcyXAU%3~YuNU`fLgv^kG|^E%^>
zJ~zZ%Qhb#yZpes#qn{=B&@-Vf3A{$vAGOK4)w7PmGf&RbL@G4Ed#|$!L~Yhu_UY|E
z@1ZmHua`OvZSyX>5DvC5bMk7F7Trrg3-r5=V8i=md707o?Sm6aFRA)_N}1R!1W2f`
z#j84Y>lX>GvwGLABU+E^j0ri4k*W(Yx(ty?6}=T`b~0}{sL>EZ-K?!U8#Hd^l9fI5
zqDa<ns1>Byzr;<7eb3Q|C;)u1Rb^6AL1U^G4#7ST%|yccP{DZ>^~c!5t8s^C=+?6i
z1b3+fbi2D7+sJY$b#x*=qYP^ip+YcTNo`4BQ4`yQ%Vzk1@^FHSxW(3;oSkZjw#!uU
zl`eBAePl44ylZ!_P{es%g92VOsD!&sDz8GiJeXE3?u{QT1OBpiFrfOPy{}bm4GPwO
z*So7FfQq@FrgTNl5ml~uRm2lvP#%k2MaJl;B}7&T=&(5WpUSh#Hs|x{pAagy!bNbn
zRM4m?DDW++mKm&C!%X|O2)m@NV6_krMY5;<g_xF=$bv=2M$Q%GQZ2#la38T0^d}x~
zkM0aYeb#cwTHuiz5o;f2YHPM`BCI8f#>EdMoU#tQ{n<_r2K9a`eG~D{{?Tlj8%-qY
z>W!#M<qL?XSaeb+UdW{wHV%m=XZz~T`0Cxzpl5VXZHqlLq73EEb<zx*B+!1a+J}y^
z7H?KAgP&J9y>*T5pxc>slMQq5!Dv%gJU3qVQeV(@lYR~tv9~UhI5Da6gunt}Wc}Mt
z@RC}fMs=6847urX*XK93a<ZASQ~r(L8VT<P(+$jFos?YqFizP&2&)8Ry93?*-I}3T
zp4wFo3l7b9xy)Gb=n?DZWyzH29frm<wahZ_rjU2mxEPbCI&EJyP;C@_OcS#TC9k-o
zo#21wx|#PgxkJ&HE9}Hr;_}MHzXWBQyh**-i`;Ot{=MeVPp@=pKSq6P51C{2Zq*|_
zCCC^1EemZdI-aw_kK353b9^3L_%FT~SrQkOx&2v+D-d*!gyF#fS!KfI)Jm4;jJtD|
z6}dcYa|Wc>DVdm|<uAEmV}W;i<A0W=T}d<B*UjBO45Q-J?Y$ViW%gopBaaP!6cHwj
zw1WG>a!R){nV2~>5QX}D%$`uGyj2bz^U>Pu^E(cahTda8{~Mo5xwd;b4TL*iKU-{-
zhT8P>V*V1<Zwam?hbk9b4lh62cVQivUrd@K<lP=)SVatQS({F{``d8>4wrsarme--
zgo#DpXW@V3^{95g*o>w#JKs8kv269ZF*IDMe0NUaf)QzBSqa>az0jcM&Q(9u6FzLN
zNKRZ6<HOUEi-77$6%Mw)xYba)vSq9DBn#0}-~54Ni8!YbQtb$>sdZ+`%l|^fNT9^I
z=?TLyf>o1HkGUQjoL(<hT1z)fhtv*##975R*Bp}^eAcPhJ%2bec2T1@u-0tlbVEC6
zsQcZWFk4CYUylU;mzc=R&CUHkjWI9&m@l<Bk$%RQPY{E4?I?+$YK(Y=Y~j6VJ!s;x
z9L&DHb446N*5{wdxXlo!<ER|C&N(CLIDKsoj>hM|K2Pt{IdB;Rz^gf~w>kavet*EC
zU}kQ*0qwF`ZTVV#tw@#P33z%%C1~#N-+K3`wZ9h-?;zB?h<fQikK*>AUyccQ_jKBt
z3tj`JTOx`9Z$tB($amB|*FfE%&F}Z0&6JXK_m+pK8DS%B_u@PB6Q)NXA@Pqvs(p;|
zWCSuB-Yjyvl#z8B1qKRofgf52Q+Q+V9U%@QoE!N2NnK)<DSxNVtyc~4`KiF;XxPGI
z053vsS@H#$;o@Hu3Aj<FK2i83yj`(d-B%5pbH^?eo1mYmqU$ep)%Mx$-~vBlsshBl
z;j13oNHlj&Y2;K|vF7DWm*P5tar33J0-b`?icwSD_6q1twtsZ(H6_Zz5Q7->92YNP
zMk3p6g~CMd8jHSvEazZuWl-kIh-0~3XsRB7^UPp|Ym9j`l7&2RyDKF0_9py@5#^pJ
zG%M8fyCkLovvAt=L60AO!zuGCo_)Dj&HblBJ~fOitdy@YXWSFeTAeFI^QNB}H4TX%
zD$$$D1o^OtZOaz77ENTAMDLg?Ed9l0xg!6vRYzTh*tt&{sFTv<)FuoyR|PVUtPeRI
zW8YD`D=6<(-2sE=a&y;6H2g%oT4@;T#WVkIRitwZs@%>P_D|h}-|<^oF2Rk!-y$gt
zYw7|^>)hbu%KFM#tRPLMJC(!>0cUWkOuS&UwGf@S6R;DOm<@*;j<AnscRqIYBA7_B
zrH*UPD543u6vgU5U6NPDMB~~>J!p`^Ym9^XByC3Z4|m3I4{u)&(|31%uJ4n2f?0EI
zvqWrpWdI@t=Jf1sj_qup>{1604BdU4ne1%y{+heQvKSzB=ikQ__+<!x)gw}(9~Ft{
zYF@<Z9-<n1(31f_f_4$cMyiaVxX^$7u?QAuI!-yy_%7)To~m%ADNVS2fJ%Sgk3V`9
zP>hF?hl#bpWT8_k(NxMyW%HDA)rE6XXyhcU9*TK0r0}^7@!YgapUvlq^d=chJDLx%
zBjOoOW&kerc+zmwF}^{R#2EV4Fo9F8%9T_FMx<+sPBYplcuancJ!XS>La$_MMB3l4
z_gV?2hT_=%;P~4K{wxi2>=sTfpGItHR6TT5f(#~cO|I<p+i0vob`+=K-$sw{w3!Af
za8QQ!7-gbcMrfF;2xAy!{bShqW~V`oK3#<ak7+qwV+w1(Ag)#wT>E<Y<wosbHCKSG
zD|NyW6D5u4b=WGr_>c3&!5WHG`a;%<q=x4#rXC&sI{gbbbu-tWOWp=!>=bb<Uoh)t
z+bnW~MY8Y80avL+pydxk+m#t0&|d5kKV?=pGyZb(K4Q+y=b@xwO{dO?yG_qeh-gMr
z7h{v<)HbNr2=Y_n>1>GAY^tj`6oF)SvEYLZs~y6oXXB4RSHkV|q~G}wY;Rcl{~MD8
z7NLcT(ac}jaj{c^pmhOd6-=uiY@^G~Kt=&@Dvs+bi4>pbmT(Qnoa%D3L3AcXrpIH}
z^C~bpqleK1(b&e0?GpA^yQq7H1asnB<lCWVQ`1VbD@zjT`FnbMt@$w+EDxBnmz~xs
z#@wDy86JuJnRnX_)WS};lnTM=-j}61n>9n!^wYaZBz+@0N`eL<KXGhF$oC)&gaihZ
z8{|q_IAgB9IcP8?14Hl&Hn_`LSMyI(RFz9qV(^53vo9s$YneMB8rO2bj8D?b|E?lY
zOL#_bh_&hogYtIPQ%sQJ&pK*Apqfk?3F)sWV*GV=s(tV!%AFoO`ep5T>vo9eP)Go}
zKzEhs2hKr175XeO#sna<TxIn9(A>A<=F@RINgs?bEh4<GOU1ZcebhG)+ATEb`Z#=&
zi_?*{K)(E$XbC7m7q&tbAOOY|bXsG|fsA^TcwB34+<YylYK2bY_T|MW=E>X2JF9ps
zArOTOzA1tP^CL|)<*zW0?K#yd)n8ZNQvHb%z_LHB`~Y1%|9y`$IGbOD42Jui2<l%H
z-2zS{`eOZBSQy9m136gUw<pFM8idPnP~Te4GcwZI3m4IG?BxeF(*FjB!0fwuP|YKV
z7_uw)TaR1(43q5)aSfeNI=+X>E~71Jin~~oCs#$BC!d={y59R^SqI=3WxscUg^Z)N
zx_7;{Vad{2O9_VD*z%{&Ic}Riy0p90K)4>z$UGA5i=R?e-0M7*qXxB_13g8s0;*?S
zInX(O=LFbYLZ+0HtvqLB;YwvRo&pRJx&HNA!tY>sU(0F&F=m7(Hv&mnC<S_5O3$qE
z&G#BSs--6dppT<q`u2(v<cXQ=fKuF{9u61GKju4GRd#6F-R~&2%|S>*vTxu0%6CG~
zQ%}2^5;mv`A;Q}2{EGd8k}P}<Ft4tKv>LoJXzUhB^94_#rgry=QF;q_U7I&Nmw1HT
zK?|wI?^9j8?K?!Gw~7^YuyF&5#JI)MdQ2Vce}vjI{>pUe-LB1oqvM#H-86K-S(Qmp
zis3SLIR;~CZh8L&^OMS?&U+%9K@Er?d8^5z<m100F_cFm%mJ5EKnf0WQ-pQ>;FI7K
zi+vn;xYz#te;9kG;9lbP+cVCNv18k|vt!$~?c^78$F^<l*tTukHYaChPR&&P->UQW
zRp0eRSM{^{`L4A--8`x>N6+M6%w@56>gizjsS9ow-$up0EQia>&n3RV!t%9+|4AIM
z{|78K6C=xiB@PBPC!=Xf;J2og9$;_+7>~rbu*%Ge$rWAM(>GOA)g)9kNUpm>frWro
zbT*adM|G6H<@K>Y4EWk5bA0$u=VRpP@#XkC3V1zGe|2zrf4G8l<-Pu9VdvuV06IjJ
z#rMtpde^bqe*x@{TSkqf9B=z~pT1%X?%@6L66xA`&1mkSJ(1-1I_$bG^S!j1+88U4
zY1==P4dQ`*!MowTneO)bcqkk^SmiseGCW=EDjcr!P9}}70MM)p(M0(9*0UZI{gtpO
zOF7ncPDac2?|j%V(yQD5`wWzEjv5G!q4rd3ij7vK!uQWB;i@x*&1(F!!|~&?!F>HG
z+CjnN+|=eqbxBerC7Ig$ljF19i|4adSGZv;mh<yEro_3QhywP=_{FwxUavdb$MbEP
zM&%<f>4@+!@Zq#(TTW?w8l=9|gJ_L2-(dC6{l&#75(TzkR;V{MywrFCPEzu44-lBc
zfM)(8ta>`>UAmQId^~7ok3h=zQbfXl1{7JDTu5kx1q7<Dv4>5%*5`;4o~DnkzGV+J
zD7Dp!_!fcJE?G-d7}Xm?%-nU(a*RLeOOnsp@^mL0uOKg@EM@gVyZ7T31(PW_G@C$o
zI9yQ4A@MpxgEJdTC?mtjfT_?%g8}wobAI7`^1m(P<IRh=miXPruMb+0rU#uX@$StQ
z!@xT{Gq}=})VO_>fHSEuYVq+-(y^o_96LJFJ=4_pV`Jp^ebW&GA@}Bq5F9wpgUY+1
zFMJ_*aB)X`bO%xY7$}a@_xtv#n=S_jre7Ce57WO2FDfN+C6*oXPG<y|VTvwe2^_31
zHu8JP8#GIZq*B3jss{uNc_#1q_K?B?=e0j?2lp@g^Uizm;##<&T+{*@_IsQciCyTA
z^v2PuXggKWa?=cC-x+PdJF7T_8`rw+J-&__IFIlaMDuj3OnExuOS6dXVq6eFiguV`
z;4&hD5K*yS-a>j&;%~b?15Qv3U*3FromGdEgr<g<r!-M73UCEWV^!h*Rj98+a8Uf+
zR(po!DgoO2^n~D4us400K*vP&*lNWJ3ClSzAM=%7M}{^w4m3SNZS$$tezS3c9>&1D
zJw*nEjihlP*lt<2GUaFOPw0p{S^YElGQ}G~jjSn(>A2l@{l>^r+m!<7YF<juX~Jo-
z?};NutX|bhnOqYo5#J`ZP5q@+dmnysAEkh^m3AdlP+FyT#FWRdmf4C&62L}+wPQv!
za_k+j(k%$qOVOUlf!7Nr7KSj%>%A63fbTK(bN+S3sZlU(BKy{n;W<^X#b49y*aY|K
ztm8Y%i=W%&SPl1h)%G^#PbN<jhs+h3#KGHWEh0G?nF&{=(jZ^g;FWea`7TC88bYT>
z3~X3BTdy2<*3FWTzFgFy_{IZdf1_G`j#`<?Ssb7bZ{EbxUMJ6#WbLCqF%i<@!^H|a
z-c2#@Q8iCTsC6az>q;~LY>Ys(k)>I~Ud<gh&%USEx`n8gmbY8nL4fW(lz1{;g^dkY
zG#nP<t&vgQEKA9|8%O2py^hc3X>qqhH#5SUzIN0Kx!kzCC}g=<$}9UlZLnYtzi_b`
z&EJ|G;D4ce8~6;x9^!FiquQ8g6)ngo)It})n!^?KEGsvj7qhM!l0&`IPW9-cv&wbg
z<wAQZ#qG81+LucGl^`0{LoKfvoq}O^LvxaeI`HXZ`sZnGMIGjMUDaR7j;QxsgQNKP
z*3mj&`$1SGj0meJ5KoQ2l#1gUkZ4rysT>wf7mbeNwf}%AmX!uYx9B|uz%m^dvN13S
zxm#UN2GLjStZozGT;PPzHFc$~MUysE4;|bt%XKgUnX0pz)vc1RNpB`v%qsvjpEuy|
z=O?HI2pN%GOGZ@>(--P)!z8e5P*czlkR;BL-T!RK2{lvtS+60>s9``<smBFSHurAt
zky|d&-MN|MoFQUvpVm2Dk1o~Jn=)I)t)YUjlZ1d2bN4qgEENJ;iwjp+gE2JU>iJXz
zF**}SE~$tm$Q?uE;dEO397!Ar=#J!<X4@d>urT7Yw3vR$A4FS^J&GwmyZ1zUmBC<$
zLE}$?F5S9y?D<biM8Ya&lAbl&<fzI;cE-FYGc{@0-JgE^@wmlp<XfF)!6jU1e9f+R
z<?Fs0R`3}vt!CPdZXn5uxuJ^(l-hZEH1lvjc%OQ`Lv=1T2Ud>XdaRW8Or6*1oCYdm
zMH<X~wxL>H8-J6DHuAu*x+wV)24pCA6)X~Sko`j1>_+a5dWKFRc^3uC0pN=S6Idrd
zfom6*yZJ3zR}%=ybDnPx5Al^TORq+qSHO60#lI@;+ATYJxS@+Nhio1tgR$@tyST{x
zrdBhkRQ=MzRBgFU<YsbWtA}`acU1t`kWLjm_$WdRtgEj#R>wE3e-7{Ll(pF5Y~Hqg
zHU~xCHDGl<CnEwV%R5+NwS9%JJxp6Huz54Ql!MN)yu>9WiVrzmDceWZffrMbHNj%@
z`}YjyN!tF=ZKgit*P~IQ!H{K!Z%jU7SX*c~(^su-Ef)`aY{W=NdK7+>ko-W~h88Kz
z6GXM~uu`v?%~l}gHbb~dS302`$?c<5j2-Yg)$-N>@=Ro8tKP+Ko;g=d=yDr<3S731
z)a3$yilVOdcndX8XL-f1;q`YeZkYB_dT7?NZ3HCNo2*RBn$J14X|1%wC%T*!4xY;N
zFf;U|tlREO@0n_np~QKW04-M{MC+X&#67Y}Jm=Qe{+EyCY~rH^ak8coqEc2e8B(IN
zqH{BlC6ERy47B&`F1p90yHla=r^|KCbFMYk$&O<4Y$1MC4d^2j$D4>?4*%wzU7LbB
zVblZcw7WSZ4c;HisZZH+Rrq!a=Iv&rX-4sh?V9fkxJR)46d-QZ#}d{166>n<*8ne9
zl`ZibY-YxOw1eMdM(R)NZS2UU#D$yOCP2^Po~ZQO7w6r;*+`5i|4H+3{Qv02(k8ZM
z&gKM+Y#jgH8R?+rX!JfCV&^;c18BcN+#sAe#!2!pu;X~Hoq((WSqZ1{*LUn5)j#8L
zJc%rvpO&!jMxG@>NqAob$Cty0C!aU3MGlT#I1bhhUu-`J67PUQSt|%gg36HH&-a`6
zh{Q2P&w`Mj4zbBuJGt$n4~f7MO#M=DOZ%?_#8R<!9=RBnXRPDp*0UtQsRgQ-HYQYT
zHV76d{`zaovz}aCEsaR+t=Y%kcUMQw7pDVo<)R=L|Htbm3Z#7Cj4qtM$c*dD?efj#
z19r*Ul48hP<ODwdUQqS1l|mk*VqU{is%$WUM2$EOSl+(8YF@*!oGR5@@QP2`OH6tM
zuYwmz5vGXDQAi;)=)y)im*fwY_0~$XQ8Nsv3-eLXxd6$8g(?lGz+tXjor5thgALhp
zXiP5KWD8j@X2%#fSfRWMXKJkjMdEkRhPLa7)vo%Ia{^xN4M7d+Pi0a-iU%!DQOTKk
zBK@r84s+WCvsse$ju~=i(zojy*&i{eR$(%ZJOTDFU6^Ys&4lbWW6w_-YdHU*@~GOz
z=(kXE#JZlQa}W5YAxtFWz$zk_G<lWesCZ@Dy0w4XB~e%qz%99C+t3XW>G|=*N#;`>
zz|gscJuRzeV#vkkuUuNxu{C!-y_}dT5=}PQCFMV6#GbN%F3OsCT@DNMyXa>9nWpo%
zV#d-S_ynFwjSD~YOCiV0Z}JY5*+*OYj!3dacUc>#D&E>V*<4-64s_MRe>)s15EV9l
zy(9nbpF1k+T6)K+%v67Fil6H^u}ENPUX*tvp1^Z53(&M2H>lAPQ?OWbdd7KWECll&
z40?f8Eo#rCQi_Ys5euB-mWx_i$SRstF&rhWiy%AbjlFX-ST$zV`?8O42EgEoyhleh
z!rhI?y`#SMrm(EMqzm4PCTsxWn!QkB7S*ko^l>-^3beXtaLFV1MvN0DA~Gp)*u93~
zS7F$xW*}kj<?G7<(ke)6rfZnnpCgSt@1)mRVxdi<!eYaP8B86Rmdnhfhk^3I97Wu#
zkXUl7ka$_%`jLjY?(Q>utyhaTH$mrOG6o@r2Ze<>uA_e_8-GYU@NTf-B;-;I;XU)0
z!~g6EtMa9R+W)v&F=t>c4rhdYq1@mA%Q%%&zb-7%gjWH%x|Qzp=&3SOuz3pYsO|)T
z)W;sQ(c4!L1ZbT>+DR2CP*kDSEqmF6W>dJpY~e?hWS{MSc=cD?aOhj{_6YEMBESPA
z#diG^y!}AQ-m}~Oc;@=yMOVVGcmS?1AUS<(D9;`3{r1fg;q5;n_kg4IhUd^1da;uf
zK4S&jY?Zr~OvHYTWTYu<{8k<-K0=9EUC2CpD?S3OU2U*lvXooI3`xs98Zq`SX}Vom
zKPBfXI?O0>K|@BKf#NW%x^cRN_3b6Dra(g`Wm0*BE^0OES6+;pmz6OrD(c7_I)Vos
zlAsVK&L0N_Gv*{Nte!5YrD*;-AN`|Kv{zS+KO1;5Hfq^y*)nLrXJn(D>=K!i_rA6S
z-8msT#p<7@a5`1V4B6^fQ&;SZ%P(Kv$+9=L@pKURB%MD{m19M^T%KTh#wk&y1QWHz
z<hs_|C~DKzrNdj1tLUivJT6UNCE4YxT2)#s?Cbomv=FTqESp)L#|7;2dyhr|vv8#<
zn_o(rO@mv`7zICBO>xO8k4kj;p7rXPz-~-l7JW55OfAPjL6-gKV{1`-)D*=sTpS=a
z2dn%`DPWZHlIO8<894GRtN(#4O7%c3gPvvI)#B$>NX#pB{w7Ulfu1`xOFF94Nza~}
zN?k5AP=+gL0X=u=o@<_)2bSlxag{UFhJndeT1B<#)!0qsu1gF*f(X`<EDPKq!XRdO
z3ADYzXhi7j-Ot2#3LGxwYw6v;vz0rO@HFf@)n#@b$gb^c*@WiZhQ!Px2GHkKvz-KQ
z8h})V;d6{Z@HUc1+OOIM`BZhf$TWf!ST0TW9Bl4e2qE%GhxBMKQQ+T^O37YtPt89{
zOEfE`^DZg2@G8_)24t3e7s-!y)|0gIxeyD`qgb{jL8HjBC#@M3tXGGOv7VeKMv`CY
zXd#rKbn1wXekCQOHkDW{>Hn<Jj^}9f<jN`vTYaduiyy646<*k~Tus$Ft6<ahoR2j2
zwa@R>%yxDwn16Q0V}O1d(37j{Bh<-`Hd=Q7%b{A|h~uSqULCm7l5#dEu594P&k9zt
z$B1-6#O(EU2^jUdHINlfjn`tDT|GGUi=F>C;QnV(CKKa-$w(U18nr#(K>itUcaRgN
zv9NT020gMV*e6=(Uf2^B46G$w(x2Z30q<l*2Z`EF;@mO{NRfx&_(DwGAIE+4zI+`%
zJRi4oi-Qai`U8ji!{8?%B{@A7^8158479s<xA%83<Ws@L)70K5$BJ{qGj3Adx33BT
zMVNjM4iEmvVg-Jph0*s1Nk?7puO8Lzc?xY>FVti@4v1GEZ4gW2fUo0WJp>-wH0F{|
zwQs8*N6>J;FTN;f0|*2b-S^w;4m|~ww5CH`vbAc8-gnoB`-^9yrn7~R5dnP<NB76w
z@3y$>(O|TZU~Q{`so<VK_$u6lFGWYfR=81taP80o8keZ;q;J_Pg(l_}aiwVg^g;({
zh|Lj&nXI90+@hQ3W3kCj;#<}+qj0b!{N7aC1J8nSg_)`>RrrTpA&w_0RBgs!21)Q%
z^O~Jx%xZGaK%KsI5sTpBH>AmuOB6|lvDwinP&Oh4|N1aIg!5v{!IQPNGD22m7EZ9#
zfT-9ZcbPOf8fB2T$W4~bk<{y5vDlTU2IOF4bF%eM5TO9w`8K^z{0N^|zQNX1672++
zb@R;JzsVAP4-zxyMVEeCk3J`G)cjC@PG$M#Joc4$(i!Xv9{A2);SI>n-UU#~5?H0V
z%p}Lo23*Q#SKqP6ww=$+MC`)7%S7yp&-g@aB4?gMh~F{4ghRZbxl5x+nj-lU?j4co
z9WrRQO2VEobv<&z!_fK&n0#TvD(C+yBALJ`%Ghf@ug^&~wPH&O`nwXl%{M^Jmit1w
z37x0b*5XfYiHwJeN}vu!Bij*hwMD+u2z(P~%tm=V=Q(@LLY&VV)Dc!LT6u(>i=8aB
zpCJh_64n&tNm2?YyCSqFruY3SC=Arnu1*bHKpjamf2g<WgmU22b#k=9wc%A%qx_qd
zcG;mr^vB(+GW{JKmf5aKtN7Y1weMGmp=7TdgY+)V{+JB1VNs@{4mMdMO_@(8|Dvaa
zSrLHVC3rGP`P$4^ryQXvJN7IvtA%<I(n(D)GCi>zd9V3hXhVu?BwI2Rvdo=(;Pave
z#-Zjo5)r`_?>MVq01g^F1gd8=aCyJo-V_-+P4Q43J2im5*7W#_It_wbl2iCLmOI~_
zZah94p@U5Jw;Wm6(XQ0jxbfU`wx|kC`+&!tb_RP~ZJqIS2gDf3xx!ja-Y&6K1-dp<
z#Syp8fqW?Y;~Pg{Lxx+;$rBjLhCBPs0a6?v(_7Cz{p-cs2ZpR!add<3x>#XEjEq#D
z8$g?CJTpB%ySCl*9|@0N6Sae6Q&3V@6SXpz2M88-HRZm|jZ=0SXN|0?scF?**0QML
zB-3=Vjs<bI2z~5)OQH#UjSs0e$ej0<8qlY0EMlIZD%yio&EdM!D1Zn^lyB2FQ$aJt
zkR8YD4>S3r9p@CqWh&EM;n4trjU07)<{NF#?!6X*;@jA0z`oWrfrGxnvo&wQKr<+^
zk(O~{sU2gE0n_093SV9!iKMM<Gsul~r9DaU*LM)~i>9wCzFn*<r1F&+0qN{JD^2F-
zbKz0?w&UHSWOOd2LY=}MFbkAEShU0*bUGLlaq-Nh-|sao`*=|%m7LPCr^Kx0k(Bp8
z#53ezvZZ;6q7hoPt5dZ=8S5-+RC=K2X{{2D4H5Dh;Q2%wk~pMe(zNbDQ5|`^K}V~(
z-oZ`Rvn(FrW98+>#PTyVIxz=kCBugQFiG}z7)hTjxRo9Ex>(oT^PhwCDb5|rNi*B<
zY@2zK;jaT5ZiWIvTS?=Gh4BL4M1|a8`o3+hj+g_Fe&e;kBKPN^Lmy_bm0rk|JDNME
zFq5%DTTjJxaLQn4XKyN4+vnj5(ZNBkMs2{ZTZ!J|ZY}oUuotkoXFSFHcx*9qhq=k5
zWVmj|oXVx|x@w~>oHIC9P1&>>L@@+Sr7Y{J<>6lH?2skfJjJF%b!EpAZ)_0<n!R|^
zy}XJdMsBY`B-RO}4_y~auKEb4<FnR+{(3jx(crH3&oalY&G2st++Y3q*(v{y<rCis
za~2Q-7!A^aZCN<1{e_URI+vUrH)sOfza`lEW6GsEyAk8wdG`xO&b8P&jVGLUEymwB
zT7VV3j(-^L*suZksy*r+__8T_LeMm(QW~ymb>p;JC{4Sn3LS|lRiast8PZ%u#jQ4z
z0|Q-7^NyAjX7O3iblLdLy_9)aB8^4V1Tw<l3R#TO7K<tl6)MKtgSrjbHI%IhNim~J
z=0*kx&nj!%lRO!<P_Xnm2bfA0ZZ_^DZMbxpXc`I)ujHFJ&Ec09^BJ01Rt!za;D{EB
zn%4x+mpR=#E#`~jTzQ%FEU>Lt>jqsh)yHeu3YHC%vB0=25!%kCve)xg8WAp;PfD~r
zu)K?dr2kr^<!B|dJ<gzjDa+A{N9;k|G@@&HW?U*V5d>6hX_<KQ*QLJBxr%2jlf8Xa
zm3cp}WMt;3n*9AO!PrNOhih{d!`1#0$xmv;t3M~*96RRF@Io-Pnl+~JZCR@$VIhuv
z(|9sLDerPtOf$%7)1ZBduZP`(29t1Pr88;m{iGB+EhU?`<cjZm-=M(MFX^6bfZb*v
zrMtJhRUf=2gCv@-UW~U{yw_>OY9OjHX_zKMU7poG$P_Pd6i57ytSqVa8;NjE-TO-6
z?bG(`o!%NA6wmVXL{C_KD3fZNOjGS(gUt1Ez8VclNToK1{>qKjsliZD8dsh=)wR1Q
z;kP<}`EPE{K;g8l)t7R3*(T6XTB*&fDzX=^sc79jtdr(<8k>Ih5BA`wO3IIv0oErR
z*Jr$F&oQWw#}iLOAxlIW+mvD*A5R_cqJEJ6BZ|*<c#qr3P~BH3xe|cC68GnynS}OG
z#RX~G=6>y^r%aQfZRMFSNTsy^qyHpgIRC@R1SccIe|0j^s5NeD%nrZ#PJIShiAH(`
zjsc;)cnNHp($lW58c0$>L-yT`A(@}8_Pl#FmESh@EQul!_ZOk;@Nzn%<MaIV>-OPL
zXMYmPuBU8$7@s{CJ^l+^0->sv6UcydgYT>D^NFaa?xA%avS=ZlFZyS_&sH;y#S;Z2
zl=BDQ-P`rogYY_!EWi1MCyw4wIXghIUd3zwMaGW><^=cU<?}5k#+U#9$>&YKuzo*Y
z?w|U~)dLM_<yCclKG&4Ifc&q79<8|n<Hpz5=hN^!J@aB)av^2kfbYrjw(nLkd3POj
zAPK6ZRTQT3ulR>>cB{adTNPEpczKAEb(w%H`lh@{*<9Borm{_xvqEPfDiOshh1=EI
zwG=l5scgnyJ#Z!k%qE4CgU`~e)qZDd<><!i^Z6j;D5+7E{0b3czd-X8n}XfmNIa@7
zWD~MxE7H;p#TZv7T=b6(hQqE4lukL|)9VWvZ|!t-^igs@5)QDkogFFVXQRnQ6EjgB
zp#q@6))<t)usbvP2B<i`qVe8e9+ETL5OW!x%r0mC(AdZh5cHosj9}l%Sw{-{o0{&3
zv4yw+a#Nas+h`qN$6n@}2f;)r%H_Nr!K?MQqWGbChA?)$;CLJ&_H>P4kuU=PN~|b-
zms>I8hYX`@F@9_;VcJD1ISgg_3VCB&#66j7rlFf~!K8?C*Zm8BY?~pgW-N?-PKs;D
zrATHE`8Ut=Vv`kU$8)T3P~kLD0kY(kr0l9)FPgs~O+AL9?isulDAuLYR<dS&o^38l
ztda0kZ(7KiCUWahFx6o4Othv5H6m?<UU|~$#jg!diiz!Ubet+YBgmO)W>6|gCHY*l
zLsar?s?1nu+alG|s51;OXrdYrq_v%3UmDv_-Pae1)#8j<W{+p<FQ0si*^7UOvd&ky
z9&Td;>z^u>yOYO_G?W+0JX?%cxk{%uR+U`^?!(OOFeH)QnAtLLNQJO$|7a!fy4fDo
z8!wnsAHKf+>qWh{Zq7=JQE!_AB>h~Awz)mV=^O%qtTgu&NtQU9;D(BXJjpWlK#-Di
zqOjou-W1y>cTMeKi;0j`j`0CkLEAf9FvlSMy8criVdE?i9t24|Au?!76VZXGQ-9TT
zJ-ybR(qs_!8#U)mA)u0c9Xy?-W?3%V$m^K1n9qfG6LXE*Nq5dg<QkO4&A`t$(DJ0k
zTaPkn=gexx&;LE5m+&=iz@jAWOhtL4mq`duENP>`&Mei&!g=Pt5=Fzl)oSzzQYLl6
z4QkxG_E6c`NDKLG!b6sBs#e4^oS8Y!_=&^DkQ5N<0j3IV(673cEu;!^m^+qz8*rxd
zK_TODA^N9CmW&g`mVcUS)mL&mxS-EB9o1pJEfkN+7^SR~#6ZC@ab=lU8|z*_L#<pa
z(ohFHq6kY~-`$b-GGF7ah_qUom?g5(-Ks5-DQ2y>S!53W`ECUU{LA|K1XP8j#q4@?
zT)(Fy5^ONhg<L3#(`8C)5n?4Jey5msg4<VHJy%4)F%u0fO_j<o8Bb!`R1=vV6)KdB
zta)$(9o1jaRs3cz0Vm+BJ;I-7n%Z8s)V_4aD2V}lgcnxc+`?&-WGL)nO#iGsDq-xY
zURs@#hm8}DpHMpOouY1yQe3wvi)9E?YYFD0-LiOm!|`(EdP>8-G~}%^FnZ|u$&v#F
z8P2ISUyD1dn1xrmt!S|ee$hRvr~&AA>4QVxi(#Kz$n>PSpt%?~SG4S8+`wu<NZ00w
zK|`*#OR@bC6IxJmhRd;_Jn)1koTUljsl__y(ml79bT%<23uV>-y!Aa6jYXcZR$E`Y
zUa9mcs8dTF)w;j7-QOrRZg~;9=eF{!?c@U}tv2+)xH*7iPioUN&|9ZTS%G2vUA%>z
zCDc_*^{VqZRxLVa`Q?Q=uvQLyF@I!>X1W2QKUIf?D;(R^{42D(7%9yZN`W46&}fZ|
ztnVgULs5fbdL8@4>+1e^2jQrkoa!x+l7>rc>Q{c?fZc?;p|iwjoNOZ~QmK0mi~Zgt
zidAr;Ykf?lGArvyybvW(g%@B>F_pA0MF78INS?B_aO(l6LUyIvFl;sZ!GvT4$v?sy
zAgIqyV1Iw2`gV2^Gw9o&I)(C)cAJ}n`i)f?s1Pu`!ib$>A;XW}h~+tYrNT}-cro55
z+*)i!UZFQC?f%K-R)72|vAfS`Ni}@ZR(!pTW!~H!f>p$$15nE7j6xpwCbtVso|3iK
zTZYMc#dV7h#W)KqO{c0{*`5BHDWFKfwpZjBRP9JZYxz|SU52)J2I7Xng*{wsaQ}#+
z*9hD9mL~LC%K4?sQ~+la+L1mHS@15oLz5}*%{<K~WgPjNYlo*IwR9|t1|syp@A}Z<
zvXB+Mh`rYSAzs@MgjU6(e;mhs4at_(2Zths#Z0?IaonI_?g}7tLbXlWI?cSH%)0zg
z6kNQ~h`p@OWhcFfzyx=y>^Kbz8$-!kzDBz<e4pn*IWqo`jEMLST}+A|LDi82)VM{f
z)1(QrERS#~UMi<|KVJ_LBGx&rGi%h{lsG|XeNe0Pd|MdzyCGI~TA9u;0)p}Z6>sV{
zT={g?#trbfZ7>818ugU<N&E7F%$nWrH|8LDJhQfvqG*(zN1lqO>&6@qlCz#u_-<^_
zq%eO*O&6J(i#8J6=x=ArxObm!Oyt12ErBz2cTW`APXY;dOo!Zu#t7jFquVS+$%;DB
zq?amE0x#pac@Ek^YBP>;w*`<D%I-4hqTItLN##o^2MxsLGhMgSHZ|Pmry$QugI;eR
zfHNB6aY?k<Tny9a+K}yY2IbN{)Tg7VO46F)U!VNQcW&htFW5DFS!v6E*XtS5mj@q;
zVTgSl+jp17$$7AYHkAW(HWJ`{RU-bzCCF_NdE6_-nPk#7=*ZBb4dzTami}5p`6l_W
zp6KYQPvsXu)7siAXoB7!7tjG-=zr1XzeE)V)a_2!A<FI+lMgelsAR%<0R=T(8z=2|
zyr$KLuY(xga7C(|*49rFsHQ&jPd%t}&E=}4W0?M@(Un3et=4Wkk~vzGV~OqU1^rYt
zuR+F*iVg7_usT`T3r)6lswO4LSss`a53ZTnC{`6u>Xeb>G)aAp>!iW5dOg(~jhE0<
zBSn;P`VXgO13|O49CoYS!^lB--dQZZ0WuiZFY#*9^>wg~goc{JS6c3@{ocnRVRczG
zC~#>4>z_)@nJ$Z{YH*B4CQ(j1U~PE<Ja1R7{AX!CW(oG@2~Fb`cY1eIi<-V|S;L>|
zPOM9-bMG3}mbePx2{1&A1qimqJt_~(5<EFR-m=8ZM0PPnu2T|L8G8y#8=c{52%Rq}
z7G3MsRYZ1p^|y9v0H%Jt)^m15wk9$-k5!Z`E1VM~g5fsDKaTvI1PT^8re{y_$ANH(
zdMA??5wwnW;8ZJm?bF=RRng_>VvW<nrxR<0@~EAQ7yIc&R(oIVCTgO?Wb=ln712i&
zMG<-K$glr068)b+Sf0Mi^5y)RB>IYKUoqIv4>k}g^TGNq8pfV#@()~xIk;=m;!})>
zqpFxQm$T<vHXALHg?=7og*iEQaiiqypC}FNRYMoDnu6POVwnCZp>t8?y<2(7Argea
z7J!#Q^obhgC_vwckqq1!&DR=#C&XSYgvhiYt=e4`%|ykP;u+R-J~r)v({jjq8h)fE
z@>eRqU-Lod<c0vptI4P;G8T}@l{m%MtY_T@RA~Hxsvv4Nbnu5Nh8|PsCQ-E9q0`b^
zhbeacw9Jny$P!oh53`iw%csMX#wc2_DqOZVH1N*bde&Pl-C)t<=of}#DT<06v-e4v
zcf#(H@?^}NrI%rBDaFyG?Udx$b8}3|H)wCO(d>WF>KqLJ(_+m0UpFOVw?-SaGUohr
z^TVcy-GX4jH8d1}XRfAB*c-dEul2lt`$xF_8+k6|!!ZXu;Azk-c}wViwTz70ZCCSm
z{~%cHZ;8|o9}kbO(SzwpzA8xgoS*iRkf%z=r}^o>z2X$PKeTT_6|n;GW4;kymb-k|
zseYEGx4&--bh|y}Uj|z0rZ(&^d8_QK9<libH@P}*A7zw1Qu^GWx697YyM4YcGb$;*
z9b@ib^^OI0EO7M(kPyRFUL<@?LY~GqLugSz1v9|I2jA6Gu9aQ!cl3+wf7rXuzMfKM
z#B?t)i!aZ*gjq`48q_$>XXI3Uted4*@Rsdbfj6r6n|~WPdI&d%=;iNdW8Y(KP4ry$
zk``;WT|((!b+_jeyrSil``Hz`l%Y?Y9Rcynh3jA@)onXqpE6r%OYD(4zJWiD-gO|=
zxX6j$Na{9|eyesmEwQ_6(K>q)7gk1{`<)jC94vHy-oASbW(qQ?#=jHgJgXnso&&V<
zb%z<V9I?4Es!^wXLg=S^nsM(WcfQ<NZ|>v;Z`%4&M?Bu~PR)NlUM)LZLC)lhkKCt`
zbiRi%Z>>nYT(NpqL3cd{zQQp9$Y`u_0+JmpY4n*^j}`?-TD#84!_AV<rylJVb?_cZ
zHY^ksV-g0$tcRyxAJV?;FFXV!G8Nlz)kGSn?Y7&vyG7T{ZLC>Vc60Ul%pzkHU4aP}
zHD8UEg7eo}vu$^5Qu>u|o4f+?-fAZN+e*l|f7I`M=n5pE%V3YQBvZBJT5VNOcF?N%
zi9KY>8-HsK$7sq19&=6R$<vugiVP(f8+7Pg==q+NtAr19ZLO_K8>n0gdhD&ue-%0q
zDgGcj=~|sQ$&O#Bcoz{+au%jsg;me9vrvNwjn+UHV`uVGCCkD!ij!@S7f$OMmq<Z?
z(>mf$xg5s4e+f{W^No5(ASJ|5KSR20LkwES4U2rC3=wWQS3oDBZ3F@pi7l00>raeV
zM&+XH_xiKhK96Np%`R%XQ;~W%V$a1L`e_!9q0=u;q=X-)eW;wM)SUr@3J|N6qP{hl
z9{h?AoW~7|jP>ALy2<(8L<#=<3rwnplLDDe;IP=*v=#Vl#Qg%qftG<|&5!c4zbwSB
zTyY-cIyIx){dAxnuK>Nc4WtFpYn~$EXRXOL0QAx1@$J{(f%CAK#GA0vk%d)}$pQ<K
zW8=z)y*a9^X<+6roH?nsqB;7-Cp1^ZlnXArr^?VC=OO?q7%{k{SBvkRG7T<BWUGCU
z^A5J6O&dvAs~?4}t~`iJZC$desk9p%&FpsoTj2sEh`qi}aJ=lNf;J}&cSr!ny(v&j
zSRmvOyZ%WF^uxj1^lx#ipFC_A_Lf0Qw&hW_1OaE_Gij_f`$mW~?EJejfssHs;^wGL
zSG!?2ly93WX_>qbS^GN=bQA5|)(vz0(sf@LQMa(zqdjoMVCoa1x5mb7WlfXCr|#3`
z(^5$y>j-hH;uFk?nW4718@4LxU!^QTPWAc4YD`OVSU5&oxlO1{5%z-Sk{ZY7VC4Bi
zWk@6mEXx>V%5T99dzlOkz+!|HWJl3y6uHGxCkl}HU45Z;)>YsU=g5M50>ON4uu+F7
z)#7}q_73vPOb`s&5gRG3IaK~aO7~X_PRHx8-DVWFO^{pM{J*~htoBv@FtjCbWptG6
zfdqEE9$s#M;dpD-qPjYv{jz5KQ5h+u@F#podJKD^lis>$(!#SJ{nQpi7hb4Dd1J+~
zz)pfULX88pZEAiqgKQOS;fH`TE<J6zi3}l9vhXQxd#g)THCF;i`@F9?Ruv7DuibO1
z5OB@&E06t5CX2$02g?k9aos*yFhmHQ33y!3-O4S)bQm;y0IC$b#(v#Y%;Q2Yli`j)
zO%ekMHNzNwE;%7wRkdnkQL~|#-40|5L8V(=!aNet?lr+SL@7<sJ263-8==~X6$89z
zHmt`v)1Ouqez43T(xxKUChH133d&bGZ*9|>*I$_?hMk9-8r?y4s<UOKREhWmOtm;R
zHJ~XEUydqrikA1$4{*S~!7Gnx?N({z*79KVRni4kb{20{G)Ix}`dtg3n3fz>ZR)Qz
zb#Yh8Hb`HoR=mz`&u|Se3}`XJQm_=@IiTgI(f#xnA@1X))s7gU|AV{3aigvGd^|W%
z1zuT{Z7*L@m~J$fdK%_5s9KdTwNnbU&&>^Dszeiod%2P|a8&s?jfR1=MOHwTTs1&B
z<`QJv_$SrhFRcgq_ht2fi1a5fNqR<-pw*qG^eL2(!>xjo(|50Q9C?Occ+A~RbR*-n
z^X$Gnb<Z6LE4UU6DI6mU>SGf;kbCg?+=aWn7r0Ts2Gy*u$n7k7WwHt(&HBDqj8|yg
zD7;+)0!KJJ4SS)n=DnlME1CZU8yn?B8;bTj%-|%={=TrFl$~8g>jJExQ7DSEO0>20
zMSWLBW<f!L_>v|UQ;0S6A6Ddg<sNq`{^;1pI7JPcP>RtWs8dL5c}mC<CU!qyY5h$6
zIp8rcL@6HRkoWILlIbPP9y=}&>b5<N=LdX$R<WH<9Zt-tyG>Qqlr5=p%APoId(s`R
zj?7)3QefaLeTcbM<#!G$M5rVN<j2Ow#jkjb^enhC_70auJ!8ar69n&>0`|G|ez)X9
z#i3Rk?}DnL#oD?jmIbOAcu?rMKXfxvtQQM}T;$U7AEtT=n|(Yhmi04{YMPn6Wew^n
z8MDiAR>qpY-_tp)ybx9e|C+1AziXF*>IW}r6@KR!W{ffDFFTo;Xbm7qTu7|0D6r2<
z4_px3rLR|jd<vG&vuOv^&T`jRETzR}ZkWf+J>ACdS!d`^FMvg<rH*HqJd~|*{s!=^
zQXyB)ng-;ZB8~IB|BAJuCf8HR?MC7?shVmVj+#<>HWs)Z;(dG5L8!Q}&?#$6M(_u#
zmlDn%K-rou3PG*<Fj4O?QqCTumfGj~YyHRx*_$b`36EdQ7#0ZO6oovIIF-zj%E~1%
zY{rR@_zU1nv8`)k&MLmYvJ=Gkb$X_zCqp{=5~JANvnHYL|5H@TC1@y}YV&zaHO{EE
zfHCRZ&&rnhtYRWvV+`x%7S+!eoC2PQ-t!j4BD7t+G#sEQPt=BMxea&1n((Hqp%_q0
zXA6c<XH|FKvG>C*2wA#>NiB=}?U$n*zM^d3HUpuE+h#Y2J3=zxkjDfi`PL4gavpZJ
zbX;(F%NTTDY*JMM)bJGwvKs%Yb;8%RQwhz67;x*yuGojesO==Ia%}vhq`iw@y3w5d
zn3~8WXGVtw^eQ~RLEw7iS~)Se2zq#%3U+9+>hHGP5tWlh69bDP;l>vv(>(Dp#h4zW
z&&A7bOVC+}qIM0(UOB{2?wgBmY#*_1bZNJouFMMVX*Yd@7Uhil%=?vQi7oS)ghc)x
zqZn%WySx4?e%Vz5tL_F?|FO;M-<&7~Ha!+IQ{?N}*$>NDRTL)1f44)^lKatK!ViX|
zhnL|M7qq56&I>uv;*kYwF(i0c-1>Pb8h#V+TqM_0_lhN|Tn1c!3Ppqo;O1jw71Z>$
ziY-^c&d!HXgiqIhFL0JdHw^wbZ~i${#`I7-4>|ISB2UisoC@dZA3mIDG@}`K6*XT&
zr#G`2KvCD`^vt05Bx~|A(+wgBknZAqE2~QT&v;&r)!f?KPzhn3pxy*;SGuUgli^0z
zibRKct(4fpAxbrvza$+5sefM8R5Ob^+|yzw*^DLaL&LQMWb(>%9aZjecn&jVwD~6q
z#luzbHylmL=ADH1Xc427sd`ycrKYN5vY?@853G<%gIdLVc&Q+Jq21iO@_oI|r@UH2
z@tx+{1%p|50qaoTy0=S0tBmLs-BjX=CY0)U)7-|cEbw`3eMDU%;h(pglJ_~={Hqs^
zhjNEyl23$M49y141Zy-T@t!z30Qupj1lWQ!i;4D~2B8^~CEYlwJ&OF(mxbbMmJq-3
z2;nS^@Yi39EoXk4Gg=T#OTJRgv4t1c`!Ak*RGmZRVh(f3+uO;>=d*dhV+r)bU!7A8
z3I5MyCMo*!QpS^@W>_PO>1T>aEY76zLY!-0;L*B{V;ou}_VJ>qhvX9mP#O!lJd))B
z3AD%GYoagg)8RI({nSXnOPp!*_^c90I!x84({>l@F*(~`?RcP~V6`c}-><Zt;@UwZ
zZnDDrbiSX*ckj*JPNffO9U31+PNSCdvYn_~Nxg<5R0!f=;@w-`4AwR@l*uJhW5`Re
z-~bz?+dDR|8MN_R(5=8WyL)9SxLKa_*5k#GzkK+k$<ghBjP(IosS<>z6GZQ=cM1h7
zgv}s-mqNF#ZExX*nNwfQ7TZW1MCl!p_dJGB{MR;X;pM*ZKK&f<?|CZkGFU8*Ihu>S
z7|c;)&s)+j>y$c#&p_i|JfG(OiSd_RJf%Ebj#t*KPNY`rywo(n!P`+2v%`@Q`0vaY
z69RS*8SG5G;Ds|lf1TLn4u{V@>ar(H&Vk#()xAD56`(1SYa2*^l6@(M-ZM-_`4rsg
zZhyDA0#S`w=*E1Qwr;;d-7kI-6?{QX)JqWmhyOG-PWJz=9Y)#1-h^Jx&{E0ShF+F{
zk%8fdXyfSQOu)&`LI3}{BAJ+2|Jz1HExUCNG@t7lg4mQ*aOpUn4mt>MK-$Dq?jVE=
zdVRZqJV(=BxeG~ua>CK}wVOG2n3dZF#VC+Rs1ph9-G*IazMLurQUMam?R3PB$ipa9
z6)~u$c?yPj9}p!(5qK~xhCz~%Ur?jJLw+q|S5Fp2$3OSNs9%K`!(ls2J0e8WF(~5Z
zT}1*y>mvxDxrx`nME&j|K}`c-tU|g*zBauouwDp}cc6bHPB2nSd#5CBl>)O8WMDD>
zAVRq7Vu*C~aTqjA%LOtFnNWa80<*w^k&Qjm<c?TL!0fYJN5Vx;v;30lG|46`y~Ugc
zp{F_;+_M-{L@p&wQF9kaSBHlV5NF0qluK5@{}bO~GJ<R!P({Qd$moZ<5{_I!CipKg
z)yQlWmP$#*k46e2t_mIs1;^TmNNbeK0KK8M)rCswS0A9s(_V80(H;HoXL3h9Cv6S*
zI-AL82b!$OBsX=YY9-}y5OF^mh|rYDRwoFG2l{dF{c3!65KQ(hkE(DAAx~5qKac(B
z%@p2N0na#6;>{Y8`3n%mIzrbmx?|6ZyjKVg!Cs?$UUOwRVr^eqhal+KyU4|GHtetV
zpj1L+KCEQWaj%IlppbyEE;xliNz8DNtBJNgrWf*c6p+4Q_2_}<Y;TNpq2&qG2?(8g
z%@w#hj+q`+LSt;wAkC3@@|jQ#M=*JHdmx2d$SYz)sH2|wpAJl$8(u_qoSZl~JESnb
z8_>tDm)GOHODHzRU*`yezF+VBc}Xcq;YaqR$NRUZ<0)GIKx%Wve>GymC=gK2;-$`n
zFSsFz+!kZ9&+kk4%0nA^Dd0S=H)SsM(a`z6E`0$up4Jsl2BZ3mZ3XsbL`Z%O(EjEW
zKCl2!ID&C6VGG68?(ovUk;&+FSj|QevqOl8Ic#<v-5)-efHEvV=2@`OXl9<`ilF67
ztT*MuO{^Co+t<1tH(2!D^x`&vogSr0Q>x4{B&BbFab%w>!9Lw}T3^UwLru$+{*}F%
zDhkTX8&9<C++4iK9Zmzk+qXvh;AB@-eBC&9>TOqrR&&-iifvK9a3?`WOe0E13?yq;
zYG$p_VO!kqn9i<RGRPmX9nRk3CtRX$Jd+*kHb?YCOKAaOHcx_kEM8F<{`4T?z{79l
z(eI}6+34<4|Cw8r-o`=q^|{Ob^?vHKE^GQt5w5W4lbtgv;H2=uhc8rFekE7)o%%$c
z;K)Vo<8fcJFQ+$T^lIY8kjB>knw#|rm%lN)o}H<6UMLtYg!JPDWe8Um4=2Nfvf|CW
zE%@58@>O0WR&V8@H-K4zwcd7oe<DGD+SD+?wL1N={r;OC_H@~}(Y<VUugTN8UTk>K
zwNdk1-y5<j-nha9IK=T98H5fjsylZ3CyN4N9YrOGd^J0eSH4<a8edmX>;~oPncfGs
zNmBz|y4?mn6-&f(qh)(7EOufqN|02h0Hg&GzM@%8^4r8gy9#U$peLlu8Z7dP1sd~P
zQYM>$M7uwhgE{Qvl);Sp-jM6_o8#3_y1uCOVS=P$UWVBXag#Z|)R|fVgjcY<hk_Fg
zn7a3n_*?8eW9z?nSjmiC#IKbGv|5P5LGTTb#jY^N0J=8P`aCT~9cL-PU#D#WbxIf^
z{4f62YP?*$_A<PRHe&}XyLyf)0OB1rL8V^{h+@gpfG{=h5RCH@lR+Y31?luJ)i`Iq
z4gTh&#*XqX)!)ggZ1W1eYJ=@xizP=NO59a+6<;luHSLWgLFux<fz)44RsonQcL(Ce
zh3vOq8}J#xn($YmL!W@eSK9ME<S>Quy?njt_kF`9-I(z$b6vRceB$+~{Nw&0c+%fw
zs#2;{EFXoD(g+p!1;v<A<KQuAgSI%9k<yuwAkZ9|;L@pzd2<H&DH6Pq{RU2q*~_0I
z({3x$^p5V;i*25qG@PSK6|SOO6P4BgB0)w`Ela#IjBh?)UoiNHdIu*6_Ei5}K_Buz
zFs)1Djf?NfFcr@je$57)Hbtry2+|R?_TpIWjBtG$B+qJ7V+fW%z7px0e^LM!w{+un
z$r{%rU7M3+Y1+e^e-e2L!T?20#k+{vQ9eU}W>IlO-RB0Z*<kPbT}L^i<l`3om32hb
zd(*1<RTPx@ZzOXDG7oCFbfLh)qEFL|tXw1K0meA88st_n^GBOMtd7k*9PrfdLTEc)
zTx3Z+_=^xyd?(KR!qI-X!=!FVwgF0<Gd^rah3h!PIwg#A=j*>5Do$rxZT$sEkkBwT
za#a!qA|p~)Xe-a=jZ|&At4#4b+N{eBt}QhiaR?_(IC1r!O%ojDv(y6deqR9u1N>tV
zZq))-L!lC<0d3H(C1|(^<<Hk>-NQ5!y``1olrOj}{dH=SfCMjdCfdK99=vpB^%;<Q
z=SSsNjWXAj9dVAG@~yRUd5&5kQkNv%rXB(M>KSr)71+<N#26i?OqK3yaU53!8>%jR
zyxwmUx5Sg1S{fO{C4A0`OVz~f<T>`&)PP*L*}yHa2gGF&1k=AcL6gg<OS#WWrU=sg
z<hpB@{$&YCl<8(p<%o~Pz>4UgG_4zPYfH7svzD~@h+j!&FbIyTEgnOtHsn#QxOCXf
zICg}MvugqqF${3Gd6hnF(Wz|A0&QYsEjBrdZ8kaaXHRG1q>)Vu7}AGcwmC6yk;mwt
z@yhNW4u(}$Z=j!wtK2!u9?crp(47)B{J4|!1orgxQ<n>sZKXq>l+V~6f37;OrpF##
zjcU;Q4c4jk@Qn$&A&cR3@5rIr8E0!o5hw9d*GMcFb@&oY$!)v7BHuf|KBim>y0W77
z14aj}zDe~}fvSMoK{lXIUaA)`Mi(bAV>mTo_}KdgPJcxuGRZOg+;VCj@99H5Jb)en
zS-o(hj_TE#!$!8uDXgs!O4n3Z9~@vgd1YxFftE>drrpZPQKCi!o*B+bw|6a*K46D*
zKje|Tf}%T0QhiMgi$+J<e!y3ZmyZUHueM<RJ95fCIVOeS`$jMU`Cj-B+KJ-_{Kr6G
zXJ`q{!}GuO5##?7RR8z&f1m+&Vvbg?v)z1k3G&}yz;g;I{?YGoy#VTDb#KluAv&e}
zd-RQft0)v+Xso5X25O^$Gu1T*c=4R+UvBOnUN5O1?5+5-GkJ4s1hh5t0hO-Dg~j9N
zh!zr)(~qCi$QHfqk8Wv0OFupB-)-@1OUM9*SQCeSvBeAWlL`x{7YpGPdx5oj`bd_`
zXxTS03vz~bM{q}oZs+Is@tp6wD`5_{`rwKQFGKagq{gg@^T=)4vpYgeBLeol-hYqo
z1CPR}@EhtyqWr5t3@I~t3Stv=7nvI@+eWkm9}<SZWvZjU;(u&scp0D7B8%t~Ac{};
zTR4$?>_)^jUK>*`p6Y;EqD>ljd|sxiay*e$U&7-?i*@D?`E;+7wnjEZ>+jJxx3suI
ze=yWKFes+mW}R@&w_pB{vsQ=~A|(<z2Ju3-O>lv#$=dcGm!dWydNiAFQ_UyLz)N03
zFAAqSp136AS=>-w$*Lq4osRWxAWB%1s_QH>gKQ1xQaDzZ3HE<UeSUA74jOiaNb=D?
zi6ahMM-8!os5}7u{#pe^KG2<dKk|x1U&D+&5$`DcMctvbi*IT}PXDLZB{9K%@!o=D
zipikBRm*%-B$ya6-09%0?-2!|?arjKF`cJrj^s6pYpobKFV~1Cn)COf$C+ls8E5sa
z<ArR83DgpirkqFj^z}gpQH)5*-~4*xPf7l$YcgH^3fLPFrrwi$i~Z&0tL4-E%N%g0
z<jA^AsIRwIH1-*gZ<fGWRm>^u?k)uWEuZVzBOtY{JMP$+MB8-s?LyrW=>lBzWU?;-
zC&eCr#Kn<ps<$-Zh>9ORPv~UVbf1V_lgK%xZcMxZ^)*SE<tRd(Ev1>^(z(U+m2=&E
z*%%Wm*n|{5{xPQ&DD3fe{I<0YsCjtjPi_^HLJ;$II<r6tR-&=D$ral9_Ex?SFNKj~
zSm!J)J1JFy!hTRuwa!2eigdod#B59c;9f+NCccGpyO72QG4i-95b4D!*q7f-f3~)&
zGRn}wL)P=Jn)&L7oTjS)rZef7VV7fC7j0Una?S8*90)IbsH}z(71W9JIEzYdlgZtI
zA-3k-!Z`wSh)dPfBEQz6_Gx(#g5^bk&ct+!(G-48bGSe-(7CFF&d4GBy~WxZ1&+<>
zjdi1~tEZgOah`<AE2kpO5{E1k_K2!K^730^&dcUJ57<=<FDib~F=Lwor@d4bUrbtn
z<AI&(Xro{j*-Wna!&J3Q?uKD0Lo&4W=MQ9b+Qw-cPO~}09!)cjZpx%W6bG18MNUov
zns_Q%hoCnu$i-UVP+8}I1dbg3zP0=`bGlh*Y~dykPGv0&>d#KazevG#_o3uTFWaJ2
zF`+n9P86u44ehZ}Y)W5CTGwknf%WRw%(8@rc1m-vdg?NwTX##CID6;pe$V(qaYj`M
zxjx3CA-2y4Rd)YQSgAR&^uN?=(?UMW%5F<>hRr>f9`4|6-BEP3&UBltsMe5|^HcHe
z>bxk)$RGNFmm8YCa0LoH7TaevjnTioK6XkH$<o<dyYU_yzt$!oS`jTTb>@O?aWy`Z
zEv=iFbvtyVqRHGCu~)iB+Z|lkwW;3Ts-{J@E{%iWGjz&+jO-dlo71`o4^eISs5>A|
zvI>)6@+GbPJaIh4Dbh*|chaS0aaQT|&7*$WMgS5Ml=(%IrQel~4gOknfQ@QP`~~w9
zt{jcdzM6Zb#gIO``(mo2SwHa^BX&mniG~-(s6V%7q&h4)C8G9gdYv7NV3wmynP+sW
z;R^!{qB2J@1NQgDrYqcsoq7zF?bZ7Z32A@Amj!oSpth3pXrAknwW8tYsfqr0v8Mup
zBw39&j#Lg>J?eQw=CpRb9J04qmkp9iZ0xWjN*Ym?=~d!)%Z5_7wV5;F$E=!jci1fk
z{K5vS`MuObH2-bm_O0hKykPgCo3o7VE2<p7pGp!u8K<0bir$mDf|$Xz8y~y7H336f
zE>uEHTZ8MFWi&c6y6_+%P|?H%I0#%8;gR!gDy*D1tm?JT*En)252a_!@yAirXEdQW
z#zegjpJmQJvdXlA-7ERaanens0LLuw+>G@E%+<3_=gq}wMzG4d-ajf>%`dYGb5^&|
zW_&l(7L;|bA9!blomzFF0KoS)`fl0dU+EbFE9etSvp??ehi^DgEvf{RoGcoqzN`3g
z0&K3#rMuV@{a2>GT}nE!0|EQW>)g~B0<DN#rf5-%xK-%~+h~N=t@vl%9<iV3t?JEo
z|7qMYvHp*7$M(NVqX_;F<E~2epDi{c?B?}PNK!#KtdN9&LI3|^?3{u`>4GiYr)`|J
zZQHhO+qP}nwrzKxwr$(CrX%KU?jJGtWk<bML{(<)%$485bn&-==|awqAW;lC{0UXy
zkhnX;>XhajbuB3py}yfZVVoIGr(L*a+QJ3OFhk^jJbpCdW6AOgKqO2I2S@=*TZnx7
z-pw*q=Ttw{0&ghjB6XiXrr}FK<ZEc4THO=|t_RMU3GU9~PI<Vh+KEuM`Dj4=6>4>Z
zXbwSS;+EDXF2rR|Y3WQDe!8$p*Dj<15V!cReR`#jgM!xKyC()3Paa>Tx@Yu6mu;mc
zQl^ZDc6D=S_TQHwnUErJm=u5n?*^$dO|FGBFe`>7Ac@Zd-_tC{5yc1qycuW`dHPD(
zA)q&cXyL0C1#HTJAQa1?IbVs|PP3DjY|Q9Ti;h7B_~%>?5S%@Ochylw=u5DGn@i*9
zM;Va)x}mR&KGoG_dFJ7bnm+=(x(k73YGWa6?LCCaQ}z8Y4yRc?H|;MaH}hTnV!6#U
z?vyNhGb#P&osgO+iab@7*-rOeFuhiRq)E}JLspX|Fr*jB%!<Ew?>f$#`cMqdv6?>j
zdw;tr`<Z<T?7{dTI%+tzO-_Usd!u*9G1;_H?d>dOQOOww6rIlhV6$~a7;%iOX)Q$>
z3oeZzLyoUd#&=H9j9q{+xB;|cz~)N0aB2$VN=D{Xa-hO>GIm;hHo{OG?uKx!xy*tu
zPPs;D6G~aMTp7>$M?xj(o3q!&Fqq#Aj#LJgW2(0XJjo}t!8v!RGJXxr4jp_wSq5CW
zA+1-@{cFX9XI}X2#tzb1b^dKPjBqSsqEv!q?08B#_Nt0b1=NFtIzO6Km5D8Z<(ewV
zC{sRqI#vHgCV$=_?ocPC1PQwxbsb{nO^!bB#BTo6<htSkY2-MJy#DRowVY9!Qd;1h
zHw~UL58gt-Qw3h_TaE%~!!RDrLyZ$<jQifeeFV2;ORNB8@?Sp!)Bki4WMKSVA^zXg
zeOnq6Hiqp8TQ91oV3W6B0Kk9dojLhZ0=Horq*c}qe^AL2l#TB2)t?{H7Sy#SQd6lC
zluL7pxV}D!WPN`~?p~e`Hg>kPxX1~!;QDiYJ}-)PBZGP+KoPn*-OGdHZ_jv3cV-kU
zXp6i14U3&Xa{zusU!7A?4GACRU~?{xa(Tb~|8>vKt=AD=Yr(Uq5!mRJEhM*r4BSQ_
zIy;D~RUf&Ovw-A>t7wCALE?D2`{Sms$X-Z4t@JFBPv{W&D7x!5;l6AaL~-|j%7Aoh
zE7=XaF1Db>`QF_dWY17~fN0QBQt@FVY7VwfJK1I@@7%f-%n>DoIppuzNBy%GEmk}Q
zIX^NJ1}g=>Up)76Jz=u#!|A!!KgOG<=wA22;%AdQ5z&<Xhur==w0llH|6-lHYOoRY
zJg)LR{Lw)<PyJBUsL(4;p*|mJ)QzCRd<(moG77h3>Np~-o*iGV_t#o3eH=bm8w+x5
z@78b7d8a`>z8-038zDtLSX$J$M^GP`pE-gEHhRC0{x(-2AK~#{WP&2TJ-B~_E>eID
zhm8aNbo{z?vPovOEF(i~vsqO6G3GgS9Wzkkk8%;MtE<L{b~;5@^D?OL>&||g;eL24
z*(s==azD!y-+Xx`KafCNWp(4ap)Tmrx`9|sn`U9>-je7DEILT0979jbSl1wYOOq*b
z)dwXKg8iC>gON&k!X=nbWww@BaP-TftVx9@G}^6TM|(!eZKS?Fmj1f@V};(Voe==2
zvqa{W0cA~h&r)|Qt6<@~mC<<yxkGg;jqSMO!<8JMLT<ZGD?OHTNMSVHR&=>cycyNI
zj4oJq9Cj+3rrzij85vFeso(+F)Hs_Fdvtu{r{$s8X_9TWE&5mTIODN0<z9fsB;_3s
zQ{=&@fp7@@KFM7<`@Bsc8MFnXC1H{zX=2nY{_Z5}S;#gq<QltIp|D?xy%Pi*cz)fu
zfOsE?6qmSuRxFN>cf$LlZeH)tJ*H-%<iLb`v;v^3k`EsSq)W|w(=jK%gaW~+@syF{
z<x8afVNRtD<Az+3-RyJ+iyH1U<xixf%PN$%5vDis`||P6r+Rv330~<wi)KLp{yb-_
z&6Z->!luW!_f<#YmB6EU3l!I5nd1z9(#<J>g|qWyXN85nRjVy#Z(vf#7xt}8Tet+;
z*1jqxM>mQN{uMFFE$A7mF5;N&I`KcCm@XZiEC*8&^h*^s+2>Fz>feU?1S9yVPeX0O
z8^T0E5Z#U5qv$c$+d5rH-N84gcPxl%@M5*BA3diK2BvczOXR>z0hx$|s%z!~1W>Hc
zL<p3RcI*@scrWiHq}FuFCd1%q#uVf*wVqeNWI^Y|Vzx4n7}Omkz>Wo*2uTBZR5z*x
zjs76KO+q4ufNF1HbJ}L72+}|w{M6Vgo*vcp)%$ViTrPc{t8`xDtZ7DF0r{?>nC0%=
zQ@tRDznLR44i^HMx2~r{k~y;E@O<Z{>3!&UJL6x^>Q{=OmSYf9<2(PMGB$?`^&Cv%
zMb(C^nHMNWF3sSL6SkTxxoVbMWGY;ePY`z*;p~l4ci6kqSq>lUnGuNItP@`*-_JO=
z2DO2|-aOP9;PPhBx?~yp13$KNYEtkqD%sgXjHUeRlJOdV&vQ&>SYHndd3bpIS*GUZ
zaXTIZopxdTd>aiWnt}TStyu<@Dv*UvTXO5M<Q)|Kjut}z7W_`D;Nj86#7sHcsX5IG
zqvu^cQ<bHU+jsQbVm@elfV{ygE^Fon@>B#0#z`Teb5-Huz|C@jkNdIXD@0%rX9PRF
zO@kSg8b#nN5CBF{3pee|fxY&+nc%zaj6$Lo!;R3;mU$Wz<weod;n>(%i|h9rloR7f
zGgFozlA7)18U)dpl|a$dRX8H(UVqe2*|Sv4JGV=dO0JqIrEQeEZJyZn8VUZikk<Gp
zU|WR=oCj;v@TL{HeeJE@S2W8x=R26kQUJY0+(g*x_3T*Q{<XXbQJ(ka)h$zJicXtG
z#7SdY79>oacOwesW@+ZFntd~=D5I_2T}g?qHzdVP^#8{lXA<4m9SfM)Ww$)dwUjik
z;~M4`1$>ZCKZ!I)&Lv_5PtxIDWdFg`jcw<hyF4)fMvI$D)VBrHae}^69&dph#PDKK
z1f>zW9ckFS-W}Mn9>}Ru`L)1Y!ZH|Yk-$3=FTTrn&j{u-H8}xO`q5ZN{c44NNpSh*
zrP(=lnooAMJu7bK^`%z_m7e@w7i1-1fOjUW`1;rQB6BCga}|a5u8m}Bx56Zf42HR0
z!jn9*;4NFS(3?1<ywS+`LU?MU486yO3vZBPBa-&6pj}6+NWj8sgg#}<HY_3YyM<9f
z@(>Xe9C*bf=HgVera(a|a4++y@_tT`_t8@&f9}v%DIXV;8953d_{(URG3J|*AxCeD
zYr^i<(ffOsRAhE8cs=26^=06X9f#wv;X?b4AK7$xetX_ovo4X@&h2dRyk@+QImJZ(
zH{xYupsfYMcCC3X%<;>JplMat-{I}%nG)PcPY=(EeOXb4a9z3jd-jQ?Vd)n~Dc+>#
z%W?7MlA|Kv`I>#+)6YUY8k!4tRL@DDceah0j)*m%7QrV2UdkN;oT9C?W+GX7oPk&@
zoU*1ti~VLFk6ukoC<q86yiWrV>__ffLOAMTnSpoH!QOgS;6_x-vv+_>^ZM)_wAzxQ
zSvPMSI9O`lHl<I6aXCi_FRxyKqL&r;Yjvrs!daKK$Sxin<f6KVG>_eK_kxqsFwpqZ
z2y8ka)XA;xy`#v`@|YDo-o_g6C-myEHkuxi2C>)$+^gVbMhTAni|zFEF4A1(?<6M{
zpmmzCv!j7&A;(H|3f1pNZcQ~xTdws%4;<p#r|Q*souX48+p_bsXBO?;u>;BQA$XDu
ztJsG7h}*zliBQJ11`ZYfKR=adn&&4E6MOxIBMhy|V+bA*Fr?0r3R|&IRJ%pT4vrPp
z#>0zMII|KMKbGFnd+e#np$um^)bpy0j>_Ywse`FDcF0p!+7`V)rQ!A(r>m4HH$+fD
zO&=gD)Av}>CoCVys(vfZnXio~tK4dirfM4d>!S9GMO(mq@ReG^+sEn|A-VDwD-7!P
zZLGZ;r;p;T%T2A$tv9m5uDI<islU6M8czoi$h@4;@81U-KLFV^3Htw8n*Gl`cX~GV
z{~f%p`?oZsWQE>3q5K4r7Y_G;hDWo*`T*QQ9<4d8IK4WjS^v>h!ifLNOvs+tbZTJW
zDN;}-cptYYl>5zfw$Mr{om+b&%e$q`<*z&)x+{CLBO8}<`iSZzH`kfCBBGv+^9zN0
zFKE8gII!%Hvou{mo_|T>*?L<ukRt3WzIgiF_?RKuq(np*6l-(j*YC@|C4zr=(d@WB
z>nH*7QlESy@sKx*!l{Usl$#bw(Mpm)ErOc)H5B%B;g2jO+Tf=f1PN5+2h*7aP06xT
z|AfkKCJ^3{Ru_yCbG0k)RU4>~?70w=7{;M>Fu@#T%r2kd67N|Ewrb`8ypWn%reavR
z@u|Ya8H^F7Z8B_KqNG-?`NWQZc1<o*j?REq%Qe^FrsE*%Mno%SXR)y*Uaoj;`jdj8
zO1rAQrHP!SU#VJ!t{e836@2YvJL0j4-uen1%j74h-t7@m3#td2x-PG)uL9?O_ySQD
zK92n<Te&<>6R@3(Ouy^^2eUP%3#E**jxk&cDW?e!wUWre7!_@>C3W<VGFIodap6&O
zoXEf9grLW9T={llH%@5NFbbNQkVcX$Tk8=w{c%ko@QL%FwQQbB<e2!dWyn-xOIxOt
z7fM%ArMjAMPoxPdbT=ok2i8r!=9RWrnEiW!Jw790$6)UF<TSHoWm;|><7<R-F&7V5
zaO`9W>`O_^39~~wR7$*qx^4UH8h&`0EyQOxS5Bc4QKS9x1`dM4w62slKz_v>qDj$s
z8$$<?CaX7ze}tGL5($1KabV>JFZ!;wB*^@od)oyzF!&5`54#kFQ^r2pZrS2)76@)V
z6B@As{TP`I9;<)_TVRa=^Db-C<DVF#bJ8%G5{V3?$k_|nlEwkuIngw?CAhW$>?9OI
zwU>a1L(LjqH68aOty8rd9}Bvip|@pK{R=m|IrrX8c~(rMP$EM?(>uOdJ-lBH$rKYE
z;Ns#}-HzCEeli&cV;;F-pcM%`x+kW+HmmC*M(uL%ev27arF^8Acoe7<0C|<0n7)G^
zwz~dDq`zJ=k+vy!^E0P0$l~A{!gx6adFYpw0Szc%Jm#jD{CMuPX(1CHKh<YirxA*H
z?molgD&bfI_17_hoK}FDWwG#`QLgI;F&gffWdAgd@=+GJ6>bL*Bt35#)ag)4RQNc%
zI4QX*8qI`4Px(;glb=2Au)f}dovX;G_=<!wGSAeiX2iRZ;02|k<t@PxO88P1GCA7a
z=amkajx?HKv4{Cw%w9PVoJZy9`S15~F1RR4Oyrx&B{;d+A&1jccU-8imYj~3nV5x8
z_}+$WDNlF^<uCiwOcm;ptX_NFKtU*La5m&LG+LD$j)0m-Ht*9)aqgflUx%>vQD3Qg
z)7r}5_j`|1w-I>{{V~L;IquU`<=^|QxZ$yL(<BetF;T(2{I8xZ+2reLINIrR@SD-(
zBE2)2=*>gKE7Rl%<{s70c@3g#M>w$}UHvzxN}v#9uRKQ$1ckh7vh_~#mon&T@L=Dr
zc!oQ^)_Nc`=h|K8?9IUsRlLU;Uw91K9emANL-ycNPnHU%TGEf2tIe(3;?;$+I+biJ
zKI|Wc@BUa~I$87wF6Co4s6G8dPrbhb{K^@kWi5Yu8L!PyX)v6kt{~p$pTc&xjKnmk
zdTJ=T(wu`Ai#*0%bZ;+6e`Lx#PrQhuf@)&ZQy)*;Rur4oTzjGbn$fCzBn1N;g<=&;
zN=oTzJh6|Co4I8;Ic!b^|FV~wvR0k8G=yU*Ych9MLz=hD26Ae{zkdS5<m0J42n0hP
zx@2xKbgy}(A%JUxaD=ghZ9MxG*<G(a*e{<^M$$7{vDt-DuWphpOH78-oGcG*<acq#
zdRfn(m|is7yy))T(&#!|UALrcsG94ZK8tjQfFS98(!SaxP_lvAJXAQUXfiEhUf?tb
zv|%lJtHMpoq>th(Jmdp;A1cR;r8HcN*)Ew6sl$dY_sRq77#C~A`Zh*C_8lKNfoLg_
zq>Y`8_t<=IzCo(hR&l0<HVmn@K+GTP87~a#gm53gUtQI6ZaF^?zAQW6a0RFnUDlc2
ztVFM^Zagqfq-nF*X8q$gH(Z51nu?0>-+J8)qwHrh-mvudhBFdEVP%f=RYHhF%%wk0
zL{yLG+z*L2qCP)j2RlKFYjOJe<5elE$#c}m?8-je9{XNE-{J>p?YY+6_%514wKY=H
zYDN~(0Y_?hURlmo#a6!oHLh4JPCM98hNI=`u;apssglZUy(uz&YdixlYmFR|Vf&E*
zbbLqc0%|yhkmNY=t?cT_O9;Eha$_oFAINDTs_szbTHn7cHOFaUDTIsVqP?M+S)j?I
z6GU>*?0H3b7^bU%l=ya~c4F~WU+&dgM7u%a{TBe=dlZ##%NL3bA^RyPv$9W=)bpe5
zE5KJgwp`@A^MGbtC%AM(+<Eec-$1yDeJt3-b-Ha?n7=KsjO)q!vB3BNmv$kQbypF_
zO4dbQPs1@o5U*A75jHevb36C@k1q=m;(z6A=Kp}NF|z*eX!JvB6E#Gn(7)}JML1}n
zRW1n(?efkOm!eC8E$t-bq%jMWu^(Ck98gqY<(+ng1dy$#fz=T(M?P9#8jrU#2hUC~
z9{?JDVbcKZ?H=k79kX}l3>=SEV8bM^#x8GNpRa=Er$w7ZIZFTw{~E%t&Qw`F0&jgR
z=J$cLRM|h#mxvcnl+9qxoKFu!pl;{A9E)QE$?FnRy`Rs|55lUZ%^KbvQoaXU7aK>l
z6Q>;umG`izJH5K|P{4(F>qEfReVIo$yeYP+x+0B#)6Kxqe?v8<ksKm{&7F^OoDQ-P
zmasn(@o=iFPH9`bfl$O8`g>_wwyf<Xs?WIVM$Q>TTLenJ6(R@6R_jgekU-0i8B2Ck
z+_jcPOH_2cs8mk7TGu$?PBHRFbC>ZBg5>o*<4RTSlEL}SnkV`mPIQ#O`*B8Nhv0%R
zND`n(qNo`CYed@CKfM#QW!uc!c2-S{s=uPFJ!sO1e^B9c^usPApUay!`LH)q<;kn8
z;Lw}Z<R}kdpR*Y}w7d9^G|EI+7iwX4zb92rKS3w4jqedLP--y~oX6ewNS1FFn7=Mk
zzNMhaWIt3%#jycr00<K+qhU<pk--r}&Ks~v(#*?4wIA0i!YYVNWtd<hIG78tOJhp!
z7IRq`#>L?76BzPYWfz*Xj)JBMH-}zA9L=BGy})viXz)c~h+5uNCPYQARX*+)#j<Nd
zPurI$V*5_@hlmaDHAsPE;Yrn(3G`hyH6QUDTy>;c)K&rpsJ3Gg-`PB;zjFp-@t|Bq
zmb2<qcGBBh7*e}32r$j_qsdYeuYc~kxXM6Ft6F5$rzEB7y%?q`(n5l*gPlJ*h%TT*
z1`r;cV?e0q`df;T{Wm@>TPt=oxrdHDC!+O`8JQWGY+fI)(V1TW3nMtNJmXJ<ttOrq
zLE{|y-3$a@K%i@HCjU7RT_{IgINECO4K0FFn^IyCnbTDA@M>ulI9T4&DQ2Q|jSDnY
zNLMA{MiA&#EwpGsSoh0%Q8&P~pSQL;J3MigFz-^Frca0<)n&?@s+2{-iqrEh%%288
z<|7~*<R%Go+H%q7Q&7hcju;Fb8ZkNSppRt&QcvK=Iv@)#H+_q$Wrj4b>08UxvtU*^
za&3SY%4LZk05;wEumMR=eckf(7bw#Vw8_>HAyNH{y+N$V5W-tJnbMNhAcd6Hko#1k
zFTmG*Z}6l+zlwqA1pJn2!ZLEApt4uxt4(k&(6dH4{gN?60*)S9jo3oMv`)T)I{Ar=
ze(<ix36|@(>WOP<x0a|<7lXLW^5GfyK=Ex%%uSS<v<pc-7S>JrF@<dihJaHfHh<bv
zX_b?^zS3TVKR%*(U8e7V^jE5{;gOmPKU3=eF+H_n;W!YX6>a3k_S4X}?#F{#?HBP#
zNk;;UT1Oj-Sq)plg2~LhB8d`K^KhvA=B79pAbWD$wBcQ;JOt>axFo1NDSGcy4y`P|
zXv_<nAz3;=^xAT{wvxqkQ>O7^@*2O@;MQxgYOxor4#KhI48AkK*)OTMgtsQeY^5=F
z?8PhO5||SX7BJz_w6rX7DJ4#RLql>Qf3s!$`LwFCHcjYNF&XobS`la0nptpQ?!awm
zZ45$>H7?3}7_rB;Y3~|8g=>if&}sLfx!A86{y80Q-nd`u=E%%W@#jB+!8*;1B|nQd
zWWnaFy3zK`NVp1xJ;?X(TgA@v(ead5NquXl8R%ovWgiPgc*?mDT990Q`qny8!5Wl=
zwaG!re4fR}!ZL^;md~IOoKcK>ZngO`gOm}%9;_J3XMi=uW|N)9{})<#*H5cialn&f
zr8VNaF7^pjUx|VLp8@-S4ua@e8UAnk`~NoxQjVR19l(PRy77+i{&V)-FJO?NW+=_&
z$f)yl_@S#!#r#YMJrQRMum=LZ@~3mg87(!<pXi@aN*<3t7r$=185yJnk{H7x>t<{i
z0V1iZ@;@|f7HpwvVvzV5>4m~=4pxxqik1V4-~^Z?MR+M*I+wEc8J@-)<7`Y8g_(Pe
zD&y|j@@#L)R^8}GGPTk8sGrQC+(2<Y3A~i*G#Q%E_$$0r4le*a3vRLhdh=gm+W)Kg
zU|{)QdMRH0aze4XVR~j17onnjs23Ib^y02@TKzY%UQrHj2mP`Gv5nXdp9IB=gp|uu
z)!`4}fex+};*o^I4@E1U?hWTAPQavdM{41Ev%NK<6nXxkrPS`A2HG-x+d3(czY?2O
zW(}+MrttEbTL&)N{o|co=lqO(<=!jx^>K=e8*uyN<o-Fjz!i6sWBKzh-R`C^09^VN
z0wkScu@KEx-t9XEkuPsd#(6W7XQ-l2gIgJQxv1)nOHdj1#|{N(CiXRdhvlxc7Ot~B
zN_DvZ{%U2A7lrBmq*(2ppR$#t_tSH1a5U&2zR;(W-ekuoqPWk;N9>L7>vO04=xyy}
zKC0YuCNp*KxnUnWm0Ry;)r>qP;#CJ)vjwVNxXKyYE%!sSzwHFjYFNe_k%;fHJd>8d
z0)&Zak|aH1!Y9u>(<tBune2-qY=eS#NJBx1u+O+c7%MLUujA3kCuwdvQu}uYmzOS1
z50hmL?9ZNQm>V$3FcYMdlL$r=O!X)XeLeGXStG)6p+Ua&Y(*R(o0f(Gf^3pz#%YTh
z4g%CMe$voiw=mw;161fbbxUS(?qr>}B`O@`!G~DW`8f4T^N$zX-mn`m2XB;I<>|tD
z8GpD5LZQ;XRq80Buv~v}V#jvfg{Bt|-LG#iKM_G-b+=$fuXg3MB215OiL|W<k(6H>
zz>zOBa6b%sH?8pulV0pX{ne_&^{x80B&O)lfNIc(Ewk8|2ZbWa*3j@p<sXFRMid^$
z?vxd?ipi8G*+htsxyDF&ldlDnS;DOqBvwGzTltdMpK~+1s?Itx8J}K;vGW~F$J9O9
z9yCjbJmO~7ax{&MYTQX?g#HF(9bltO*s7T<UrX(9d&@d;v3{H;jN?drs2!n7Z0;0k
z$5>7mqMocaS!5b61he{{mqH?fgKjq^`V?Ivjj^;zJOuc*eD_Q0I{V}zV%1G8OFZy!
zCQhbWBvRFmf=HCG*UxtiD^ps14xK3%w>4(=d*5z5pv1IM@=qY5y_b_{PbZ~|tA<6N
zug->1&x98vcdk7PrTEFtz@|;ta>=Hqjv9E`pq~LVj#4$5<{y=|Kt7``T#d4fEJ@EG
z3nOm3%W9<`@Np1x^pco#&TXH3In%_F@Wh5xDT-|O{tWg^4>-aFHQsItTEoovYKkxn
z$j1VT*%yeD;e)aUy3r6nmf*<30p^Q{6m#cz2N73!H|FY2!ggi&trD!ZWWYVFmCI%z
zC5dl{KgGi)Rwtffoj&P_Y{jN|jfXCU!WO9~YwuQLA(?eMQFrr%&kn`7<_%6>>%3+H
zEt9Tb?<l%abj4$7iOLuo!{(v|>p3)NuGz`%r<>QW(TQc&Bi=beR%sb%G6v5Xn(Yqy
z1BhOXekdK9aO=_(weCOX2*^QXTIx!#iSqaF6`4=NDw!*U!Vt3FzEfuBzZGKb<t+uq
z<F>oe7G&`gU89T=RCf{qH0c6im0IvitV*?Oa@xP1!~yx%{6I!u{O-|&&H_>`_F1rr
z2ctlf8s|leZ;=7kw-gi*gNBs^hj&WyLkBO9D$Jrm_cov~g|&xh)Incb^fPo+f>1ul
zL8U;17}TBL@+v_a%oY!D5MOwaLm>A3SPqCwqji16Wz4phIVC%bN{=o!Jd*o(+J2Fi
zbjtk%KR@@X{;UOg-t0}k&!AW718G(K<T({>D@?(gCJr`Bqt9^zAYuo<@ZhdM248O`
zCP2R_TMXwD<L3mSG{=#;Q{UFiP?sY;LNbeM3Tz3SV2YJ&geZcYIS_Y;Cu6ifC-7C3
z;Z1Kaqd~|{^58${oMWYGBo_7|RiN+AKDkVfB2kz)OxsYJyx=q2$7>?09*sJyBxmPJ
zH0zukBQCbN#cF^YF#(m1)ARjFfh3iv4m9anAjoEfwi9eEkl4OH(j#a5Baw;^Z@Iw6
zdRwRZ;)G*o<DGNO@WDD*BiJ0Ny|Eru_L$(;1616^bayW!vBB32cQ>ovT8T+DbHgc>
zidC$Mmv2X)9g`lx$x$>|FK8vt3nqG7wpJz_!m<EEyIyrSZcZ$*kE;CU*-5^e4GG@F
z!r4}yW}1v3BsW*GwY_1G?Rt^Sbxgl{M%)1ctZMnlwOw@xUr&xWsj`2Sp-oi#vUmye
zv@KsxXd&PKtc+{E1XeRM0&C#m=xVF`HnFd?ea%!`Jy)(ntzElV?Kr+pq&mkD&XB4W
z^V?kiw?p%W78Oio+N|tU0GmJ|>Xae5nz(ICAL4B}lEjvY<-JgnB}(?A5)*jZlmj+F
z_iQKaiohIxok+$A7dR>Y6T8h|QkS&`)mq)nfuRldvwg+tkrv%eNEx|D7MBuFY}43*
z?*#9{LykTF>P$~*pf$o;BYJhRI9(A2=2&!X#_)78w}$?#Hh<Q3;=|I-`gaYh!F*>u
z1$E0%n6X-)^8GzoaJ@-FG*AM}=0&t2)d58gN`tpp5(iK!X$wp&vx&lFFGmoLP~#b`
z1=jJG<%TiYf0;ly1)z=1t(b8t>cQ6MrQ|R#elMcGLP=}jcc+g`ai`C{5^96zTb|!o
z4X<z`$(n)fW5yJS`<o9O7bgC`cd_RtIjd^gHsItppUu5!1YWi8R%!O&1W4Gm*1|2i
zesENwwa&+AMrun{<GU|4BMN~2L&Fqr#aaP0{5THl$;o;Td3(cZD#<#t?p6sENjW2_
zmPjC#Q*=~C$PSkP2APkmXIZ6Q#xa$YV$8EmQ`TaNfW;628)8khsS@kWiB==r5&^D`
zU-HDVrEg`G@`~Xq3)+=2Sa-=(qGEw);0cddh1(`}<g&A58~5DH#@)GxLRGrG)9FPk
z(@g|2T(#G<w~mho#d^05@W_0x$W_B$?0By}v?PbmY9CAtI~Rcr(vXBpI9l6^0ng>4
zXWij|eI;hN0ox?v8J_pcbq8hh?q|$<f`(jo#`I$2zCgKPvkqPN9@W0XPDK#kmDy=f
zM1M-O#OoCdF|;{HIIG$8FNGH9SP=z1Upa)GwVBu|#1hF<BEZi@m9HF~7Rx*d7(uHl
zp0~-#OdoQzDY2F61+fSM4?CyoQ1)jCl_k(Y*rJ#om^KxKc`wHkkC!RA@G+mDg?3@L
ztZ*M>;J4s0$|JBI<?{hWFn5AOa6j(Oo%%EbIg+OJkr7PuMHJdhvF9Xh^HM}3(pnN-
z9|r|LK&nMtDjB7A-5`m9zi}qN9j0&J8>YwO(7buv_kU%U=?d6kcnJ6m`?o=`Fr`3y
z(tbE7a=RokGweyhjhSnLpPE682k9nA+TYVIuEhLTFH$5yOkFf5{|^RcYqHMDTZ?@9
zJ04_8A(`hTNLv#(?4P8@kIbTe;d#qN0?{<(;KWF8{|~OXriIm~P$>52Zw{%pqps-$
z{8ID>hpRO3twofPYf-pxpB1zD;+^pHt!359?nh0q+s<4H;PR#5on-Ba_i5nbK+Vkx
zI)|ob8k+UP+`VG!%opA-%M#iDDnVKQqXhlEYX0Bq`~RoH`kioGwjumN@E(Ddf5j)z
zKwBNxP)5HDJiJOhl1g=>Z(jtSiU{Edi?en0rcUa3rFxuF8D!#=A1Z^lH-{gux2ycA
z{T01p)rk|_NGVpKcS_-o)64y0t?ZPEliUg3ogg_H@9Q(epji%YuXpVER&U=Wh9ZRd
zl9?2q`>!R;^pm2WrYb|{j!NZ@6yMMnRS3%z=N1py()4xF?3*#o_kW-6KPx9YI>6xH
z8-pRlRV*H*rV5<4%S|My>qSe|p*KX)?Zu^ORQBs3bfDMUmC!0HKgdxuFY`Rs2|NEd
zwhxsB<1eHlr`MO#ON13&8`tBS>cW&YQMKM(<jam~n?~NUeLubOlZ&;xmk-9(uNzRK
z?@c;iokn(HwUWCoiLRzs`ALT&$|9S1+vcZc9&FrUx0ve!N-67OM5op%qQ_OLFY=63
zj^@~Ie)M1-|1KuGSRji;cTjYakdHIkS9kS-eYV_YGQ}TBbN<wTL?=|rd6r9UrS=2{
z*>{i{9#&|fnNCowO#d|Lc84AMVp?>a`UZ0M!xr~c&SmU0L0Z6mkuD;<`99#quC<9W
zm(NB{EMaO*&Af^UaKC217a2^!9WV80%-S%eeB}RO=VQ4Lr$qw-tqlop(TfeE*%JpA
zzI_54E#y1*e!MDd6C?8VTM<s(!uS2c`u=$Jdb^qaRH;;k?EL(gN%a~uYd_CRDNDTg
zcT~HR%B%CuQ!}n86bG>xB-yj<pNRx<H@W{8DZFf~-O14ac(b4Y%<8!CYZ@=WM)_}O
z$c<NZ+h(UxN7LlRWcdp6&1nHYu``jM%RrTa=77}9C0IZ)7&8CENym+<K50NvtyPyQ
zX~ahQrzC#a{=8Fj2fi5{fEHqqN@~Cx!l3^#>;THrt)&@^qF}8=lVX+M4epneACFfG
zC^~ZJ7zgJK{>^(IFw0xcV+~L@9ojkzV_A4Fc!G|)TTM{<nwfh@&`tGtLLeITEI`dQ
zG2&n7(E?w;zV{@qg0wB>E9PujmCj}~KDW^y#&lo5ZRo(f0(TmO=?U}mTXnd!slV)r
z8g))6(^r2ND~J$C9|8mExYibzMp^i7lz`t&E)<(FUpjP)A9(L6gWpki4fTs`W?ps`
zEu=o-2R#jig5DOxSZ7q643y*NOaJJO-OpNpswCcW)_Q;eLK(_ut`Uy2X#xa7@h>pn
zZYNe}w2>T-jpDiw071YG{_z~Lc=P`n1URXNCAWrO@{|8%eS&*vMJo(Uf!BxIIHx#>
zP+~-==rlqFKA?#59gL7;Y3uKjC<F7y-|ieU(oB$E=P@aphQr*zw>p8pK6#9A{jyVI
z_?ZRNkFi*u>3+1A%LNp?7?#p`oQ$#5b%L?JK;<mshNjk9<SoN|yae<ubfPB7&cI$v
ztix$Dd-=G{0;=f5h4XCn&bgcmbl^<<jw!hoC7Nz0JyMBP>Nbh$CFqSeNR}}B<sCdz
zeAI>Spw3WV%W*!*8{|iu!e8EFd{@3O_B(`KH(t6Y-Ug~bVUmF1@l+lcEtSA4=Pa_u
z=C@pp7YLLmImgs0Xy2iBUtS<Z&>N->`eu$W{@Dx#LQbMJ4udYO_e!Ehp1J?gc+JCw
zGNw|*77%yFfB;4Id-cv=L|#gQ0WR#|92qzXdMNn(%cbXjyu&8ta+Kq_t>ZU&<u0F`
z8yq~V_qY6SWs^NUgMZermp7^l4~Mcak8z|AQ=Pa+fn^3^AFTF5o~<Vln8Pk#5+Biq
zswL?qhDPI~al!&<_o<mm^{lrI9P1>Hn9p(%9U6&hX*`~cS1AskZm7pKOGBw#)w#&I
z#jO2{+KR}_`Mn%x;lt9M?x2u?SSs>4WBQz~k{N4zGgF;~_wA@L-sR&4A2CbGAw9_)
z(sb8TAJao6O~8bsMb)fgJC{E~RyF-QNvVD`Sp{XAyLBreY=I%5iBZM4iHIm(b(~=b
zkP(F=<JtK_BhF)>i`iWV>8zgE$zM?zZBckm5h_K(%Q~C$1!|zoJZnchCPDkb$m40c
zk|(nX3d|c;ZN(S+OZilYNJ6JbOiD~Zx_WX*C++Ie5X!Dsbghdd%wi?fy$)RNK4wr<
zr+Xg=gd{ERM_r*H8;1FERq3seQt4inae76)ZPWx&MyEfg9hTC<nBgE?l;KJhvDpfR
zllv|VlBCRI_9}_+u1VS<S?WzXG7BSW!1$4L`T??x?~>es19D}hpeyLq+pZ;Mi_!Vl
z+3_YbqPGB9YF4UbfS?eE)XtM?Bz_ZNZJ>o5B^Q|OWy<oB#HV}c-x0TJi;_4qD)ebP
z?^AH{ZYXdR=gdCT7_W;cP~H#?wjNMi@Otmd2YXQGO;!qw&B;6?i3enC+x<P!cZ%pH
zU_h7;fjKz#c+cDbhGQIti{H^t6!-=l1c#vFTqjw~=yijCUXZjLn>O@l11xzvupXXA
zm(plWM-hR~(LR)VujJsF1US16AqXvQ(XZ5GH}e;xJCUZ--v1b`-XQ-@D33TBsa8Z^
z6$F%6@S74<t(T*7HlmDSz9vnYYh~L-c<a^qGMAdcd^8}$J?}wi)I|+GnLL##!%N%o
z2oi<M!&~0I%QC3~@-%Wf4wCi>N~H*|FfFAthRe_&CFaJ0N0nT?;-{n4fval|&V@NA
z&%v=`$t)b4zW|r<Y2-zCdaq(%I>)iE1`1|hU38ciJGLr5xDg9J^1s&LNt!@&AwJVu
zVlU}l;Sp3;dawH`X+)}}eTVg1b$S}xlI!h?BJDS#bDot(Yw3Z^1_-q=23z|9ORx=7
zw9n{<OOAO)-znF7Vb+)84KracjT$`52j?8&mEF9ih~1#q5>8a4N8UbfyK5Y6!48ss
z+)~8|Me`cxzs9@1s$w-^;HgEa)v#T(7^lceAHn=HIQ=V}S+sGPU`%p5(x82!_bVcw
z3MVo(4!1vcKQ$ysBV_M!A4{VzEGP0=uQJ3T_z3(aL#|{UBCU(ON}F=YFrB~MIsZBD
z)(7G^u~C?OlHVoJoNa$prkso9s*g7~*7E~R26r-sKoCDyxvunz8oJRhsE@wTx@MjA
z1u4(c@07!K%#CoA&5o=eeQ|mlY*w#jcZK8Vc>1lHw@&kl-wg%A=6p%QG<0wh35#Go
z6?d3kZz~gEJKL+{V@t-i+jKO9lDcNZ-su`QW-DXST9&YpHEV*{;x(n+oH}rg2%M>$
z(R!+~z9T{L%Wi}@E`iErcNRXh+kJkoUU%0aHhIVJDN5NmS*xDf6*N<tb#jLcCekkY
zI#n{L%Bpwtzrm<sZVxA99ljPbe5Nx0a=1frc9O4}$Lwu$jzBHA3g+cLVbU(q-d2xA
zZ4<Y&C!D@2jW?NIzwX}ddHbRiIR2|<{B<(=f7_PyO#e%8@{*cl>?R9z&rR*E{Yzbi
zSiA?(89-z0@21+S=7!>|_F){LG!zqWZ_-npvFD{b733owA+L}Rk+64BRp*L~D{ie5
z&P`6OkDI}czr0m;<JN;D*stc6*Q~VbXxc(aWTLWmix#&@q>_1SEVjvnke75tM`U{o
z<DLP?2sCZWc81sZNF|@tL2^7jn?O{!-JWxnpJtVGDoVZ)Z<AEWBf*pN6gC3>sbC8!
zYoJfqxk17Qk9U1*JddzI8n1gl&QEQ7uBS~<sm8;VeddH{p|j>9%FQ&jDa;HqU2>0+
zEeP)65<iI{<kD{OxN$Iq<1V^?uV63+c@1Nt$k{a#ln1)4+n9xC==0I8tn#v_6@Nmm
zUcnSyjOT^J?XZ~k9-%e2-PW%DoL0UP4OPSv-MC|Zs*8suaEMDoaJ<r|7YGwkoHR?Y
zN0lAd5Se+Ht8gTqCGm?Ig;z!L9eP9G_J(Q#_ygY*u&&6b4lUJl6xgT|x-BM3HLu=C
z;1T8@g{b=aenFGtI05VYO0I1~*i{wew403(SD7AgoE&~pw@Y{DZwA&;hoT;T(J7Lt
z&m)4y#wVLTwEJ4=?A_VgjEG+Ou}b?VAtn)|u6`54K<?)NO}+g>(12#HPHK;)8L!Jc
zPZ9U08tYx<yyvN#DxhUNBb?l1O`k9lM<@U}tqW9FrCOA>t6Ul=fV_=XeYV}WHQK=+
zgTMQ_;d#7$IrrrPO(Dw80IP}?RE*+nxnjk+a+WOaQ@!CFAOyjocp0uYhF=3@2i!~D
zH7l!a4*W9>1!UQuih_I_(bQ(B_rlfQ&b-X^X5n^3y1GEe-cH_0M=9#kF0uy2Y}!aD
zyv4KO!2dEtfIUvLXWLZVQ;D8YrJ;|H+C-MqE*Od2u$YwaK1Ikaa6i5&>0ad55(<OH
z*7J+5g)mu|&Q<NS*CIoG)I|x&oyl-E!{AV#T%-X!p}P@HxVNy@trtVzP#nEM@k~pD
z$kf^j_Tbju;pNrq>14?&ADTR2CAU{+2-S!WJ+}S~YS?AR=SHW-C&z+XyKJ7IfbDy1
zsddHX{!sB+E)Yeusd*DV*U4cY`w4T3U9ob@TZfAg*3N+j=%jC&If8@|iDS;qTtFR$
z*x0IT@Xdf4%*-*-Ou;`ihL-R!sA%!$C<${-i(oMpE^zC;e0rO8j7LTl%cJ=b<;C7J
zg0%n}cSe?|smd-zl#I4G{D<})oLhoq1!z!!)>jD@{j8r$s<~NyJzCOd$aNk$&0}_5
zryk2RLMNqn`|UWKO*dXm1EgFWk!Ug6)VZZepTBeMvYnK4ZMO?@epjjqkZJ>O3Z{)N
z+d5{<WK6BQr*l~{x9Ko}D;b0i@$=hW=$47_)yc$(a5%3J5=UagFx0SR?WS$zaKUFf
zIC#Rn@Olbye0sjTI4SEJSh2m3)=R(G58G9SJsrrvTp(lV?*WpyuPZb}k~eQ0pSuGV
zgkKhjpBFU=T;X{x+%PyPa$>oN=X**%9)?K^_Ie8yDLBi*+Hq8M283RLIF_WY{3JsU
zp}D9&1@X|4(P!6N>1$UEt{Mn~tzWEyPPl2#qizE10b$!W6_@r)mxh4h*6qpxtL*T1
zYS!cY&D-qm#inB{i#D$SLFl4GTAb<HGa?*Ru>NSSfL$I%oGD%*w)I{|w@kgJOo7Jo
z6zCZp>-;g*&?Pd*rdU`b(*(K{Y#yehH}s0tUzmnP10Tr(bLRVRuiw)eL7Ma9@9)FX
zOtVY2nMS!}*fBtlE@?*O9p#+V?HOA^y3ss{5Rk!08R^Fkd^Af$*2ZY2BNLbsv=1Tv
zwESgp?or)XbY%AQ@Y(F>nx>?rr4eK7scY`wHjZHt%m=E4xG=$g#`FIP*Q94>Wcgq9
z0H!n~ZGY|RdOlQd;nMapk>Js!yTtvr6%JJ%<$$F<R*|-@@_l=EZi&MasEJYLVe(cn
zU_u{q_C^D`8TxTtT4g&cJ+0GD`Xy>t7oZZ|@NRV>0kVyCB(9r`ew7I;miS47&XE(~
z`OGx2H-B=U>sA;51OWy2!;?b0S?@XaPUCoy_Y9`!=j2}6TtmAN6v#BZ11kreJzuZR
zPO?X%Ah+4yC2+5{wid28YK%A3hl&kmY&ioksoP;A$hj`5KEA;`*SK4RE>r*|=Xoht
zUsRX*e8xO3fFDV+8eNK5$@8QJvo6|7M8$27BjHgv=;BuRTm%~?;-gew_@@<^08#3w
zg^%VfY{F{qX&X2L{I+rwMvBvKON0t9T;-Z_I4d<H=D>((HsnAperOeN_(PH+ajbVS
zKs5@uOvfudG1Yck9%N?TS`z<cz$nAjh2J^#r^3876mz<DLLko*Z>Vo<%&l6fuCAa|
zb}i^y&uT$&+N%w5n)3|lTc5TKrEIWU3C?Z%tm<P*L=U!1uj&z^4mB0V-0bWC5ok-w
zkH39&?@dpHlT>Q%DOhTOLl2%&?VVo+jQ|-d=nWA+ONtQped*r-b5{ey;~tm1=n$BM
zoziRH+f0k94o3xx2h1zC)7-ezm?)vpIZAgC<u@=4j2H~3^>sg6PXYrsKjb1ZO4!AA
zvqYuS4@r8h5EMG)7rTwSN?u0QTQk$P%WJW#<L|;3kj&Mu<T$OdOO!xyrP&O&S(DNH
zv+n#U*a&E)QRfQTylt1D^F%#(29!@?#QIv4kWYfbS`k7@U5zY1?vdrK-*<lFFrlB0
zNS>EVpbFpb_{%a*A}<oBiz6cvPtqy7DC3D3OK+Lh<a!DIRlzskKX^r|ME*4=+)j7g
zk4S@7Cb&5dBnq60W!LpVA|eSoaV<8lyC1o2fAhtWuN0?sDj8hBC@7Ap(v@i_4HG1H
z!M2=4Z*AGA!EKq{Y>4ZD$~(k4@_D+sc&^@)he5M<1f$FNYYZFapUqJnlGc1oyV6};
zU|J<C!hng=9kLH9T$|q}_lR+6Kc}fZ-{7kCS9jeGq_g@!=d5q<FXB}-nP#(PVs}&E
zq+2^thRnK0F`fn`O+dP1Ae%{>v^OI?wkS-)4^51z#_tY*^+cJy%o9ON)D9XvuG&A)
zKW@>+KMml=3J0<Bq(*TmAjZMFV*6W5JGHdD+&?jk!#Y3sN{`uOj(>SxQf%;r{NA}g
zGH!rI6oGKuz<TTpkdqSp<WN-l7&p+#WGpa2XKR74^ehS&_3STQxZy8Af#ml5$yChN
zZZ6=}+Rj_U7?fE~W6BPjog$x%k+<fVyJ*YQo=7*k>$VJ+E1VZo3Bz|k^~z;$|Cn~7
zS4$+x$QBo_SpDfgiPx@ET$VUN@FC3;bf=WFPn+*!MuvsEWtlp+Q9cGlAJyi|Qzk=_
zd$ZajeViSI6w8gVyBy#^QddKyp=fkT4Eea3v31{~RN;o^LCo$}%D_Zw;O2mjQT#g>
zQHqXrbu8kI-mRDVN+~XQLXhZ^9$6Ct(Bf$X^veYJ(5V|)8jnlRT+A{yZW1q64G<wL
z$f}YEYd$Ud>r~pFzz9~SR6u*Fd+1EySWzC9!NZ)O41NJSZl)Hs!9E+WoyD@7`L{SK
z`QF$O!LD8O5hM~#2F?A({mBeij?SeZr?Xlq>0K$a!crbCK9)P!Oqtokp@`IiO3HuH
z)ZL7!-t{>~p@2`$N8C|_a5_T}V~BYvDI|dFGJ2la&&Cs|`0e#oM!KW)!P8~5Nv72^
zA~RSyM3Q(zIULG2`yJ#|D!EYrTiuvZBbAg3fpgm!NTig(zKu0WcA1mWmyX2p8Ep>v
zvn#~v)DvmpL9byuCf%j%^`Eb6T+L*Gz}^U4iK3u6VRj~eU6Q*>5&dbvv%~to$)bM#
zfpp10ZwX2BTUr*8iRy8m*~+9tZxfo*y=aNkm&t}gZKWNMzam%?89aQ1li4pd1AqR(
z70AM{f?>KsEKP)PjX}tcJ-e=9N?o^W-v}PpC}<451%!@m^X@k-!YZgdQ+QTUs$4|e
zkBl6ro1wPdYIDvAO`PQ4&7_%H;Lk~p!^FD448g?w9>Im$l=6&`x|Yj`jj~JOZhD<w
z&*fet$*h<%IT#U8g{3Y(xGv94GF1#-9dcFCC9)T8GzX=ioHBSeEIGW*Sk;((%~{zw
zh4WY%lAXnzvE(%E4``Y&I@NQ^oQw+}B@U*h7LW<?&c`Me4>xnus*=Sb;o|9n6;?j0
zN1tiV5mExd<>iWOL5eVMQ5?aB%&ptqg)&c^|7^d8K^QzPvZS`J8F<L4Rr;_XVUROR
z)sNY*MstoXidY&NTfIs^y^v$Yz(g|1V;lLUS7qK|0z2~C;1^+@M=am!6h<(bxt6Y`
zRPxRTAj&V>R2+&S%UWBuWu)kKlZEChYPFqnI{gci{rF&9T42M@ZDBq^dxzrJ@8&93
z7NQr^buq%czA4}CzXQ=>wF{lUeLtCH4*yxbB^2kk@sn=LX`TMVye!sXmvK?)_6kSv
zdKTvUfXMs-qZ?%3{EzI;PWK<{rN8O@f0f;bezW`G?`>_S_7*oxWl^&a&&(tNVs-<i
z9hn=6KM{vT*wyoNhxRE?=lS2ZQ3tylR|<JV(P(Uv5f8VlSH})Ej!UkLf^-TyjEWST
zEaEOyOmAQP-02t~9ihI5Te{DdBMr0;Sl6#p-@fr(?_g(7Pr_~-U>x8rOy5(oa-A&G
z{81-()9Oi@TNolw<^|=Df8^{3C~fS@-Fs_jt}7SR`&rS}VcPTZC^31oGW1(wl0Fkm
z7A2N^@DE#oNV}7p$|EeX3co?LA;qFNu4JS%^>}j1ET*w4S?%+`W=~7R;FL2+BsP&*
zgQ;N=OB1E~dMgps=_`I5p-Mrs=IW%8syXQbQuR}D^!?PIon%%ssH56^y9o{B@1p1_
z!&|T3DV3sUKCoa@GkW2K;TGi<mClD2>us8Klp%Kf<Wzm7TyhEC!`w#>BXjp_Pl>P0
z9(6clPVCDD;9wC}6qrx#0;|Wra~(s=TV3(o+K+MRqGV%HjE4yF$1b2$!kA}SIU3SN
zi#~ZcOk>?Tb3}?xS`pRSh~jjSnG`#xg8dgVMEa~BHjflS;PG*O?udPOo7bVGouOMR
z?GWuoG8t!g#;7{KeuW(L5MCEgk5{p0<}dDt8FL|wG1~L7#|y0|=4Nlif~CI_%}u>e
z7kwBkLv3jhfP<Ex<_>Adc_UCpa5TN?ege@UgrKKUge#0c%0_iY6r-ZN@|QIrLo`^;
zM)RHVgpc=1786k^jL|j@bs=(GRRAKtT633>!H*(T1$z7;k+WbYDR$0AolGn@phAo5
zpayB2mJqSo4BXm4j;uvK&ZXQkk(GHFKz7LOKSBsnOzlr;(8BG1Z5EofU&%Jv-#|ce
zrTH^WL~d_sOAMFPo!GUO=QCRp+>az6?8p14?pIyO^^N)pPTM+DJY?kz$dlWg;Uyq^
zFag?O0Ft^bs)6Um>530OB;uHOBZWgVNjFPs$62JPJBEu(>hmeZdoX2?L$V%yf)sDk
z5spLq$U9JB&1u0)fkvSE&id>@WJIHRFl1WAW$idLVzZ_f-60YX|H<VgxO~3SOfF&#
zN+Zdo|5@sosD8_Ih%j$6ddFeY>DvK~q5NjCc%1ugk@mQ6o%3J-(gr)XEPvPLDW!`y
z*S>GMHNV1K86c@OcOGBJ*OtdYWsemc-yjZkML>LD0+|}Re*zic3xxr29-}*D*CBxR
zh_TeCGke_AuM>a&`|CS8x_w8u$vc_5WW0@Mj<dCCM45H2BJ1QHf10A*^Yn81ZR_CP
zUM}R_?KViC9-1+6Ptk^CQ+ov+-WAYOU3t3C(NlH15@AGDl7@o3H5^b^b1dCUVXMh9
zDv;FE0d4-ioZO-W__Ys&x~|}P+}?vH1cOBjAP+Qx9EQRCZWgvq!p*;E)~#GClmX(~
z1z@Bfwm;u^SvZt!+y-wVc8$*VLNcjVrOiU~(UJ#|LNfWTQQibK-H~>EJ2Z7`tj}u8
zB8h$f`jk`Td{}T{Llu+p7-7%8Aeko$VJXu6xTkChUI;IKIEi8VkNpoUf?zD4X&igZ
z2ce$~+DM;{0|b{{7r-nGma3o4;%TA3P39x8Q{j_J*5_tIF@W}*5bidJCt!8twnh`U
zr~_9ZEr^znjO4um|Bhf&NnmMb(`L9otwlV_ALN_-zMzdiO1+{sQ+^-fWP@~z7X{zY
zTD|^_i&EdXUcm+pI9u6AUM&4?B|*C;5dZ#23IH`^3a7taTvXB(ekhs5nreO>K9dHK
zj|450QIlpaDza{*MKt)WX>ADrh?AO-W_fdaChlFTReGzpMvx=sV0dp{yitJWj+d(Y
zSE9=f5P-2eK$w<Fi+KHhLmShWmn@hwCg7Bwuc5%Ez<vgdGtW#(MU)5eTQ(@HeU)Ht
zWdLDsb}l`~bRR=Tum6X!bKuQ1eBX3EwQWq@x3+EDwr$(CZQI7wwr#glw>#M^lHK2C
z|H+d)AK}7%9mnY+nAV$Ds8>{Fggm#{r(_Lpek7DRWL!}eKF}rS$H({e^vXqF{PS<M
zPc>5hYYLC<eBLv<)!y}ml!r@y2J4W#zflpJg(Kn(m%4VLftg#DD})*$euJe(Pmo$P
zb4cBaa}3NDln72fD-t%q$fD%Fw>jpw3Lf!=@-U=afYc2Regoe#blkpWUIePw7sqvO
z>Gr9$EL}D9?{>wdfF#{U3VLu=K`RwY#`5x0eu#T?a?g&2<!0tcdy1Oh7boc=^+`%t
zRpwHFn9-%+?r>0+c^_*&qW}RV>k4pi)eH5DU;s3vF4TJMc<b}eq14Swbk9we*MSqb
zAY%#KcS5d#HJEM#vrrHnkv19mg{f94?&DT+x}@2cg|H;3jWyhq{n@M%iRQw6CDofp
zlL1irif53)sAd^9n9yLAVJwrs9QIy)EiG^#n*^m5w)nIh&YqmcCcovyi>~-|5nH_E
zLdre1xXkSg>9Qz_P!9@0Q{A$PI8}z6uIjOat)&RMtjrL(>dc4FhHI>gR*GVO?$PSU
z{*^);dR&Xy3la3R(Q9O(v=K*ON`p8c=qOO(NWzwkwU(W<c|AiloAGMLLUWbVG9n6F
zU=Xb{bJI!TsvQ=R!sS&K0(8(RL8`Tc&8DIb%}IQ|itG=lcTJMTvw~#d?|cii3e)^Y
z>uHQtSj7#X-J$wzh59Ka2r@cg$@cOUeEv7?2q#NdkP2QZr();FXW7Eps6<UAz*@TI
z%yQ`IiL6U)+mSr}L970ud}bu`dwbRW#NK}7Wi;Wv5fw>Tfs#~4LicpW`GyP@6W?YW
zK_%(L08fa9e7JeRsD;okk7(!@7H%q&9c*k-p4P@GnElU=OqCCOAt?u~Ykq)R69e%H
zdgqla$#mC{lz|UvH9ncAkUPLbN|FPW8YX~z1r!1HL9z8Th5KWwjMTIt^R(G+HMr|m
zl@dpdrO75F1Umm{qvTW1$z$TjZ)CS|ZR+I*D8XY9{y(%g><rBRdl~)zHqNsBPoo+m
z1N(o?s5Y(nPwLZw^nIfzkgr!uFnBz5*TN3gqn*hb1_5_PJ-2fAYkg;9W<8Op2IX}3
z1xwOOi~NnIuG%W_H}=g;B0?C#sYwS17sJfaRd)de6&MMwa`-3)LO^ze2&@Tm68i-j
zBc&<Y+L;Xnc@pSlim~L37Y7rL<Kt-%k>ds_nvix~1Jmz91bP8U6*}yP;xDHD;0?Uc
zqO&S6Fa`^{KwZUo2JTukwGXLKJQYfuQf)8FP*N&RhtQz2t6*4i@GJ)QQ$eZ-q_iif
z`}D5W2m|r^DNG^#IQfhAdGW~Sb>m_nh^Ig9fjW-kh#(mUn1U-_fl@S_fe52g^3J?i
zaeVSYE<Tc8;3tm(&3>>~LUZ2*dqVTPAU9+yZ$r4c*D?+H2tB|(AQekpJ78*FQ1ihZ
zE%Pdb<Cq|9l$W3a(t8^^!5iMd$QgSg<ap4<mXPWI;W~NZn+t3KLlnu{SyPm8F(vVk
zs~70b=qO*9;&LUK{FwF)Eq~?d$eAJpDqvYy6@==jgbT=>;bc@ZLBA4OGmum5l@K5?
z<PT3k){)M29seLU>fP_RYHHZXVtFXoYRVM@CVZK6LtrFDknq41*$(*l0)!$1pm?%#
zPC<~qC^bSXnM!K4xom{G;{D=Lg+-WiP6I!G0#2xsHn+&%LL*)xg;H3I*99orQ60Qd
zI2DaO!yVg=i6Ea2R0guleWU3)ysC2+26}Y?{MGc(=$(*|+80-fYaibqC#Q9*HjhK@
z+~?Adri?r`Wnu7MTX-yfRy%Bbg#iG!rK>jb?E<g{*scgvsWSz}%|Mra$Mxhl^E!@p
zj+^G8KD}r^3*Yw3@8YiX=&YcVypP-GUAI#mH=Ylr#q8F+&B<{WuAKFz@uRE87V@Mj
z(-of{>ny%r-=46(*<eVc4IsfML?6^F=UMoW*W*cy4-?P%mTeC&PL7QA@uzBzt`2s*
zX!$_WXB1R@>4j~Jl$ASDbmf1LL{rqw=SjQfO<C4t)Vx-{_1*5Y3fIj>#g{V&EX9EF
z9=Sp@mTquj8*o+zyXI@c>EUPbUL_PpI33v}eb(#@Ba2#*1JeiRjcD66!MFFaYmwSm
zT2hjp?)(??4*npiG0?8v_w(%!CVL539aC>dH@9XLKkfR3y8dP0_Di#=xG@&*wS|lc
z?!fq#Rp!aq__494JrAcv48)YrqBpKf`qA25_gmWLP2I2Uaw~&d&;4!VErVz`dzPcd
zs-o^7jDkIMICeqe_R0+)yKp3>Irp#5H};%t+~#eA$NTH9lhePgT%PqC1Y%|Ypn=gt
z+tL6h-jFX>){cza<|YrCWhsJ<#i;qLahnoh;V1EsxVukOxxugEx1bcBchpDk*PbEl
z79s+kG{fvhm$edwVNQ0=ePdGn9B!@ZGg@}mOVP1AT1mz;m+R^*mdRm3MaRC;J>I}w
zY*&EA7g^VeIdQ7yt=nM6n0e!H>8YdX{I1YUz4c6nmlmO#m-FI3PrG%4Qnyw{%MlUR
z$mEyo{zRz9MF--9AmS>_63iCN0*o)wYnN65h9O<CKvsfaE7E!esu@d;kgg@(`vbhK
z#)N9X`Qr5QGqtdna5(~j3y!YQI?lj1HfC0Q#BgCoSlgfOOS`4ry&4h^jcSahbr~+J
zaK|-dfLZX2S3SsWo5`>ruJyQ9AZaTw&M7>A=>dHMN^>{>Fz*<F4*u(pRQPXxgjpH7
z;dO+fn@@t(Z7(>YpAaWcAnp$U)D%`0z&r#q0B5mZ_tE3DT|yEY8uy&w0C-3%|12bp
zKU_FxfMGpxkhVJ*cH2#{``9^-#l59=7j7kYA~wtuim2+C?8BUN7R44g1q<EnGbR8{
zOm#OWpNzlbJnF>CyeQOx^cI5p5j|N3dk*=prs4#d@zpyFr*L0Ve=5W~`29#^@1~~y
zHDwn{wHpnE2?w}takN}<zm%pA^cCA2bUvJvXTg>o<)MxwSljs5f#l8USA6tp82QA~
zFX%pl5f;q=d77G394lrY0NZDq^(}wZUs|Cnye+aNodr!ysPmgx=)@AntAMv$o8o0g
z8f(SS<2x&oH*)be{{0zVa2C(-AZ9m#wo`1nQnbU7`i}dXst)0oW$iPowzXB;hPrKH
zgsMSd8k5mnktK4Op`2cH=AB<SI+zcVM;9BssqnsJlQW$G4lOry%8=q^PtN|0Z4yrF
zs5S=c8pvFCJy!ZSNY0q^db8Eu$WepZFkyoCBA}Q2{`7gE>z@7sFtj#(zXh$sXRoi3
zMYl&7Y!6t{n(^UOteWwmRIHeJ9wd2Q+%=Po_|eIC)6*oXbjjcz7{Bf=ObmbilK-{y
zQ~dXPZc8LMv5k);jUX$;cL5@uU-ws}6um-c(?^IaZNtdBBpN?cSCSfAiFg53a;w{c
znp_xq%uRp@!|-Q=ML5+@!f({1dS8*K#y@$Bo&Ze#6=Sn^UfvIIG_w<J+MkIvYA(Nz
ziK{i;h^Ff47x_@PKKX=m=W-y-P2vhJpS7vI#t?oDZ<D%@cp|THavl%OW}Q<uE7ODg
zjHP?YyDVv}cCU~YLLQ&W8vd73B6c{6T_bidiY-Hl4R-*M87(Zyhq3O$P)yKx<_6S)
z#KJ|@Z7%$ECX60axx)5@lF}?M^^C^db0r~5TSc7*QS+%ZR_8H&PkkY;xh-VHEQCrO
z!8Vr+6o<<?A3QD(p0@iqY>{PI*S>sMmiVg}Ia7AK1i6F{Kozw}{1t6HHxD3g#1}+D
zW*xnNxj-u@Ca_zV?iR^$bmYA#MQOh;uwYgUE+*(*k|wHeLlUtZXAOJ$2@?5p!qZKH
zi%t6vkbfUREo>}G124Iv!|CaxHd&--6J-i+ys7t!DHV%6H`?M?ic3$lX5{2L3UXXM
z{Q&|koss<yO%Dgd|BXyB{V!(f|8uo@jWZdG+i}gSr@%+wh_5UG4jVnyyba;_IM_xQ
z782kB?;39n|Mn%1OsP<Lx&~h##xi$9i6)YOx~q73K3QCbfIydrzq5rWEg)hSn8W+U
zisfB!?^ont|5u4#bb3ZQuUE#;Yhh#N)nfp`{2JF6<QwmPv(Lwwj5t^TGRQ94;{&sw
z8x0fVe3MK)+liX}3NP=pEiON%)#5Lc&(>z2-}_|{nc(i0Pj=yhRQfi0iW5-CD|L$h
z5`W<+&gWhx{%WYY*kX8v-1&Bt_w3fXR_TGXXw!>jC>06$y~XNbQGzfq&1Y~OGk?)<
ziSe$7?B6LJe{lEO<d|T%r@EmD(Z%DL4uiS&pX_ABXJ-=&KOq_f%TWQadh{hNgvmZ&
zP23Emi9x|$zEt)htxqe7(mgfx;$?BeNrxR!m4eQUpralno#D5^<Z+}|T#Zv!>w|uL
zZHBs!_)u)VI5LhQtf{@vJObplJ_*JP(J5s9N>14tDBRc7-F}y!O8YaT(nBy3n9wN^
zSr4W2iK5<zbelngtF*G39C<7Amb%S3zGBqtzK@e(qDIcpaE`?c;~X^^9=>|7bPoV8
z@M6EaEsftsEwa=wli@AqF)1rRCVen+B5F~}*M_oxV@^<QT1%6Ro9?JK)teXk5PuEh
zX(1|$uPA1^QFEo9Zbg8|?20PVwWOG9iAP%O<S?*yk;_h6E7g=Gg%(FX8%@*kk333$
zZcUDz+2lBjhMWvU)nLCETIf{?<66<&zRH-}A4k|=b=Sv{DKv_Bdg8y}o5QU*>Z!@T
zf~A#tF?dqE{Pr%{vZ0U{S$smVa6RCa%tg*INmhleTZB<R2Zzngdx?A5;EBK^jH*ke
zXS37x4zhah?3grFAj0*;Pdns4wQw5VY325v%*>%Q>DUp51;`Jh)e!xbwo>L9?F%UO
zranv^bs%8lSQHhp&zf)9Gu7qbMZ+GByqd(B4&Z1?T)pthnl<{a^^(XPm|t_|fY2CB
zIO_741!x}VN^Zr(b(5ISn?7O5@Jak*nFPeNFc5HBs<s3SJx4Mz3@>M9yrR{mq<GQ_
zCu}A`I6nAic#H60m%BG%wMIEjJ{jMVJF!@^Fs|DQazqXrL7B;Fg=5HEqRj%N`L9ZZ
zunH>|xx-zZ1CBBodVGns!Ez|?77Q3rm4(b>8SC)dV6&0)OnQagQhU)Z3`RoZVvFpd
zIrNkL0gHbGsc|J*6jmzi2R;@Os~9v6x1}j3wV$bH)0}B@Hn|UaI3{*->gUx?=Q6f)
zkgr>!@Rc0my)+uKO7DY+t}VHIf4OOFa-AtAKUIvAyxJGu?;Ucwa8XrO=c1*v*UC2E
z%rI&egkhR_Bpey10YH7&M<Mh3J69L>8?;-gFXBj1SwVyoHa*sk-bPSR?f9K2(Qxj@
z-oJ;n5O)ZxK`yi<v!PCwikv<9T_w|16BDcuN3vF|=xbu8J#~^MN^duzvk5^PniqB!
zXiNE<w!kH&0M3*!DK&tS`i0DnXr%Fh+A*lQ=T<V_ul+9+n)Hh|`+2_SIj!&WdXK}P
z!6nCGN5frH)``sJjbDF#!U43BVn3Syi0^7G1>V%L+O@!?Dn`fTytNHn=!XawPO+`(
zMgG6rd#)|xojf*zp5j&-PSB-9E@5+R{*5Lb?3ycE`+KPyHL><I?QTV4s;U%x&AulL
zUWZ<kdl0-*hokS9*Trai<eayXi>55^QkJfTTALegD<AdcOPdDUk3q#W^H(CSl13Sg
zUPYCyYA?)y%8@$NDNX%B*K@+--oy0=+lG@2t8ZG{)Y|38t<2En6McYNR$2|CB|CzT
zhFYXcuM^!#i(vAo>G>^(tJ<}x4;80HKTi&qL|hgUE!Q|aCz`^*1c(r3TN!Q2Wg;h)
zZS1zX{Ai{_jS^cagNWl{^{sQsim>zPcM7vQjAe!zb3OGj>f``1*G1cag(tvEhYfm=
z?pbE-Yy>r0-e5+OEijE*mBr~Qvg->4di-BTc(S5B%M=@9b700S`UwO<F|<<|$!rao
zDmzxx0mz|<S&hq58#0%Fct!X1-@?vx{>DnY9FJX&YzcoQiW`5{t1zyHLsenoKOcsp
zx*M`(fS11Ui&@#uxg_eiV<d7OSup>uvvuf?RB}!;D5$zZfTr^{=p;4sc|@(V3{~Kn
z27IQVd~&oR+v5snZ3)Ir`{78}iRLf9nCoZzZ3BdVtH-VZNjx+>b90{F?0!g@atZ#C
zjTEz{GnU3%LOy<G|Lj}W*|4k1q$~K&X5p@S_0rHN?2JqlS6OiVrrVP_rmT*UXg9$^
zRN|6W{Ii5CV5MwLh)u%IZ8q7f*Cm!Xn$M>pv%)zSgGmWGEs(d1t9$+BZpwO@p<4B5
z(7KKL)wSZ%L-bvE9$~eJ4Eswe7RjLYn(4^xS!h=i?hhM_{n5XzWh`+VtN8EyATK8Z
zP<Ynq0QkXPP@avQBC;-a*R{2~oDDZX(j2ZBd>WUHrCE@sb(N5xvV%~3c8JCVIatDT
z?ZZsXj&MeUq|OCS#cKiUl?=;RBE=Sj1X?BSQ9nz2C}9<71z+c_dNZ$N4(abT@OC$@
zo>=>>6dJVO=s(WJ={d5Hm+{RVo1;dR=5dS5OK<loUcpkV<UCN)hC}VazPuI^`se0p
zXa)s{0=4AIiGMnh%W47YJChKK#T0*LqNu>`T;=gHy&N%>POuZ~pfi~{B7#tMw`c)O
z{E_XM3o@wALthErf16>C)sE10YJ!eDHfzSo*@SC$k2<a3wVA?XMi%m$p&DIsNM5=M
zl&@mR4oV<GXx9DKcy#C<ztUs+Mmfy1Q@8p@fhYfJxF(`Y#cn-+$Wbe3K~9`|ynkNr
z@1ms65R_{3@tGf|o|XN`dZ=TC{g72Fnn9NCRKMW??|xoR+WEbGRw=Mw^Fb-gQO+|+
zW%te(2}!l^O6#3_sEV%yGM9@OX)^yy=!u__CtnGeT?Ew^XP6OVcxQ>Bg(lY5-67*D
z(uReK$M`4s5gJ?*wdkpHtXEV?`y;HzxS%vB=(I>ObE<L9;B&+jbWrUljnF59Kp;`+
zn>#;NFdMn@<}61k><_5G1}D?eMoK&;;xdYWgixoF;;&i5O8I1+bJl>WziU$fx4En9
z)I81*1KZ^eZpQ>gP+17Q$K!zQJ`jfiQUAh{rcmMU$dANMv`w)8_;_T?d;vG}LEHtO
zHL}D6CflVvSKA<0ej7ZMosgXaqd{5s9t-eR)d=d6S!jKKYhN3am{UymwAS|Nrh)p0
z^nM5a9>q_O2ZC5(&!P&<5uczzWL5)Y6Gyvg!51;Efchum+vkipI!iN+`WvXOMs3`+
zZUeV)gS4l&+ik6`SCqO*Cb};`t;dk#W^CXUetFRj1?<Ibg*Y=1N9$ZWpJgp_R5WFo
zEYZYSeH4u?XslgL3}$0r_-ebIz1nK1DD457EXfq7F7lKRL>+Dwz<Np4`VAHg-PJXp
zd&9X)Rc5YhpSQ|R4<H)-*Sakr6pr){14L&S_R70M>?=guJU`H});v+_j_XR;^xGnx
z>vhZq#tV^N#j7MJ1uikxckAU2!i!uj&+f#ZaGq;wArdovD7%CxH*2!Yhv9a4tf?Pt
zs}T+-V$+B43XD2Q=Kz1AsbO`GGv1H(MDtc8dQBt093=XjL_CdtU^*vEsBc+eLXq!k
zY>{IEIvxwy9EMk8`@L^}F7#%J!)$UFTkvogFM83OFq(F6^%y`r4pVN4-iN4x7_pnH
z5O=pbvAWvJVkp+Hb+1M2CosDVO8C4#c=De5+d*Dfl~j(j{aAvYY)&_}3Ad-a?m4Nf
z)T~r5vsN(_T3bq8|DM)L9_;dDZ;|2|SufZ0UO&@N?_5`2zrTsETz7GoY2cOF>PD{f
zf$>(JMrWXzKbrBp{QJggPq*YMPNUf-lBq5P1;*xfb><_EU8Oi*@JR7K<t5A%SGPZ+
zNb!Vzq7!ZVT^fpUOuvc~Wp!EGAB<wZyM?1rnX7LEYmuj_Ml?UiM>G@jbyU4x{peQa
z()5?GR(a@Ff-P^*t?YHHBg$P1)U}!eMZTrgQWd^t^|Z!SP5w&5J$z6s-Vi>wvv`+)
z!Kj{t*KSdpNA~)?uAj>CRW$iH{^)lZHM;cU9N%^wynbLOTzZY)&zNlZj&A9_?p)XB
z{WMDL*p;tu4n*)(^~nGH08d7yv;C)u<v&P{ndq7Rhvnh_GqG%I{Zk!uAoaYc9U(A}
z_y9x0sii4{c;g7M4TS{&jsF_IbG*CdY6rtIW=zXn0K$f$P$5l{qQ{JOr~7j<xf()8
zkf+PX)6+%W!+sKD@A&p)f^-)+^b;slz!HTLs6m=4x0}|d%bteLTX#2~O$|=)tB=P!
zE6t3QdkIlqs2_(&H#Y}8O<U~k0brzK?lkH1hl_hN&1@Pxt>L-F`)P8Q_v<YJLtuA1
zW;bbpb2{HISy4wXILHo^!~%1q{BxeZ66wwd+`^?2cmX?JA_$kPziqZAI)>%Id#dFL
zWwY~BqD8nGiFS>8CGB>9{1pEXC&$dO*vyJ%QCpQ~ES2bjS;AvdG4nT`r6(;{uZ-fl
zrkfpljjxa9Y@m~GUC+o$or^G*n#RYrha^b`hGRB)AAa)!O?!hvuBu`*MWjH$8Hk!f
zOm)&#o~qIx)2fn4s`iCWA3ho27UAMDnq{kM1(W6uGaD%`ZRrHj$wfYTyOE5;lQMcc
zubJ!+!|t20@mzI{3hF`gYPZ{~EjA=2xV8rRRnV5JtL(79Q8iI-`-_X^a7B1d8#H-+
z|42jNvcC=zxz`a)Xx$p$UB!lE{aZtW8EAtVukg^{fS=IF+wZsM$voa4Doo}Nx#|gw
z>0+b-y4^<-YdW(Oz1|eDq2PEyx<GLIdQXN}2Vm9E>ASbbhvE_=L9AWqzLCNH=2gYk
zR7v_{o$An>q(VmTt6KB}^<8r!=p_j8rV7U4vNb6S9D)L#*2uVfp}YpF=J$_>ClinN
zlCwqib3?RQr?G*-z6@UQ#X*6163!7~@Z2$eT%jDWwJ_77<e4^fNz<NXXwx$T1ejM9
zu_~ugWJWuhz~1^_G=sMd-a-gN+6Bd#Wd>yYTqD&Iqm+@ZO`)5&v%6AVb#oUaZ!=%q
zFctg+t{$N`>9IVd#tr*RlE>O<!(Gi{LL1W#1p5@>z-PS|o&Nj28sYRFs<Sqlec7rZ
z{_BIOT&m#EEDFF^7^mJUgQ0oDSZn>b(c!r{qYX#VOD5iy40WmP2ff^)#Y<Glo)A)n
zNe2WEVZ$BJEXu!FEcxn2azbm}bhNAo5nay2RZhj9361{Iyu?dY!oRksz0SPIa;{n^
z!okQcDQ0BSFk?vI5yh%SrT0TF`U-%z3y&;YqHL0v|B3COsJycBv9_#nn3HBU_Xwo)
zYJ$tnkc(vJK*FmKK`(AtNU7M8grvoZz1oy@q+Ky0p$Lu5z%g1xXW|Vztb7f%OuEGL
zo7sgMRZ2Z9Mu5z3&@`S95lQpq>m_R+9+AM3{j)sOVORG~(2fltT52QS$m3H*tF3=K
z(s=NJZ#K6r@z1k}lS44gk~%m^V0dJSlJA{)Ics;wW;!IVBK|gyI&5ENJ+pMKaG>jJ
zNU8gJGZd=jz*`6_ca&;;FKlhaE1R&gZt(NPGNv)YI-n>XV@>9tTjsZ@?A@?L%u(L_
zx>Kx?)vNd|K@cMa(s<urPJ~YhibzO23UmeOKRue&*UVo6TmhVet){s>k-jRJXn(4%
z9Wn#Aywh!l54ka?1x>rP=)cmuAC~Lma+xzJ90g{2Ytr&DjOg5&)>!5WnX3`92*q@>
zxoJx00V{W9(KF_xr^RbS!VKY{fg-cTx2SON+O0l`=h9jbFgGfFI5EQ~!-KuPj<AU_
zjPtA9OTF!^u8<%OUqs|!^6<MuU+|3IV3xS)y8?<6l>J}#YMaCW;08Tymm3u>hU-Vy
z&aOl!&Mr0p;tY}(AVgS9m7p%MXdYuSK)IlWvl18^wzLo%(L5Y&um@5MX8aU3mR9D3
zAv=84(nxQPmV%3JDpOrt;G=+1;(S;l7#wU?kj8XL3^TY+)mAL^j69gnG~*hd3WsEp
z-yr*duq8`pcqg(=y#oGR=CBQIf%-GTtn66pl-4%`dsAv5VJ`OSg&TPqARXULOn&z$
zDmLzTw1<)rNla7{d(dT{P%zH<_-0-%zrskgEjbH$Dkufvia0Z;ud`lZ6#y4h@h_DC
z*p@kX`Tei=dcvoDB7e@l;c$Pcc6H_*F9r$G1sF!qIup67*e}JHj0^3?z>CNYMLqHW
zc-}b@DI3aG`?S&yMyhdz<V2FYx=+eGz5KTM#SW}pLe7Mf4OiLBft-m9)%N1#tQq2b
zDmig6k&FPcy^ED|IYjz$Eo?6<(%8KMXx1BfZa#PKF)Z7R15Xo%RnZ`04JjS&uyAb8
zJ_(>X2sA`sGt0p~^`WAN$aaK4Y%e)|MKmfj1a!T)<LKtGkcP7H++S-pvjesf9!%U)
z7X_|C$I0)j+<3enI{dy~N{Rd$BS2aPzh8!Pthh`10VzYOW$fI@f&(piCEZUKK(Gj*
z69GuMaFd%A*+C1T78a_M@&rQ3>+G77168wS`aaGkB`z5<9+sRQOYEepG(h{9M|^Yl
zQ|tplg4*g9g=yT+rwz&Nq+;3Udzlm2VQAK#=g%}%GaJyv%JhpV>-!JJ)<>63&vRcF
zu-;$pJZ)>V^BL(UX#SD(+n@cw>)(`r-iIKpOpYRlmXdlht@*@UF93^0Cv#KVhwkyz
zrZoh|)PE7t2>hyyI8s2mkHsWLLWzLbA1gZu0==8bBk*>nvWnxn7LivCVs0S4`mq*^
zE*lGgt;|4H-m{4jH);<$w0Nln%;<O6)+D`T)L|qnRzP1rxaU<yyW2L%Z(08I2Y7ub
zy!1V{2I_6Ap525Jy8U^Ez|eX2tl|1AeU~W9DAX=Bmi8KVEF}nTx@`GJ$vNW3wevTy
zHpS_fgV)(GgWWacdh?P;X_kH)MggKhDbEmS%dOV+L_WZfGq3a~%d`pnk+<F`qIxbM
z*?9PGkeX=FX`**Y;W}yMGcdXa7|UrDge%nCXu-=@QP96YWM3|sXa0i4)UidCdCh5W
zvSC=$Otb4AvB^W?&CJFui$2Xpes_sKx%M%0FNT0VzTQRaiIbQILX;%%wz8lJqgb#7
zCZp#ytudAB!+9NO2%j3`llgxQi5ieLr)T1b%1H7`!W|H{zFJ-I?H?JY3eozu1p_D2
zT<O%Fd)1YQ@izA33@G@|%c)6F(RxK^2g{_%n3uZ~=Zf8)?SkEmc9K9N#Y`vUnziE#
z&WcSQ`rD>}Ym1LYbDern0ulR`5{i(6m|c=b3Zs+xSk0}DQg!h(+C&v*3eDECF-<2q
z_^MyGqYlrbCc8Jem1H=AR~zB1`Zc35MY!56%Mj2b6O9`@s!IY<6<lJ~Sr~$>K7IDU
ztb{IFZ{75z15>$X;u4KaY8Hv^9dyv_j$UB61M+l<J1$vm6M+TfX&WeJZ&c=;5w7ny
zry8+*GUPcV7*V$)-imR!Nh@toPjIFzr6XH1^kTC4mYi{{w^m*`kZF0?{w_dvaHOdc
zNvZLOQ%v&oYbAkP?8i|zdJU<<52l>z4U%N1FG_ClJJFO+TSCjx42$2aEJ|ABUGygu
z3mG$_k5jETms&iAE=Du{<v6WsX)2qxN2MvyrpjDAS2twGnjc-pmRR!zZ#U3?s0%Rb
zmV{J*{~QwzdXk3xvg57diZ#V?gZI|oI^TW%Zi7fKRq1ipr+EU{bh+zuCk-TmUTDtg
z0go+w!@JvT*?J!W6+@c5{&a!gsM`x{?xVjpKgD^%Y|-y;A$0f=DkC2peSucTB;<Wd
z#PNCU`onVc!03{{jDA3z6nsF?qwT@%ANhb18B1!(!V4z543CrB7yANDf^RA~;<n{}
zr7rff_LgGiY|+k{vS-p~wY3ySz6fryo2J<;xhLgpy|m(~FQ|7;3}}kc0dL&XaP|?v
z2eht;Snv-5i9={~tlOFNFP|-2%C6Wp{lZCKzNtCn+++46wSil46&b5?<)pCcYMK@p
zBE>%86k={B*_yU;y)K{_XPUPCa~Mm+Q41aPP7@EPSO1IQG|w7$YZ*V?;h{E!ySr>&
z@U^73$XJ6te!f_MhsAe*d@TnPq4?NR%we+y!Sg6oKK)T0t394wjQx9|`(+y>KID~v
z1(b>UKTQd&jQ{(X<9|d1GqC>`Nt6+dN!v?S#Q$!i*guYE@&*bG7><B(UM21-tSW4)
z>N@-S5ebt^KsK3_h^9r=K<1#!<zO;CdYb)RrUy@V8)k<QlY|!AUwQuZ0rn~IpcntS
zFlP)LJWt>Alm5*vSplm@=UtGj`DpUicF2{J%gl}s%nkDT??!N(7<Uo#`avh8;UE5M
zTn@vRwTL1aPc`K`gOT&?{qs!NRG{dN;v2g8aDM%DcWYq5+PGe=|9&|}<joH!$i+f=
z82m#2_38EQ`}(hf#YBR}Mly$%M?s6jqsflC7Yvo5qR9{FQ6>Py{zk7U9Z?{8#f&FW
z2{Z)YS_;XUt8R)D=F05P+z+tza_gmi(t0!X{1)E-G@gTAYK0U`l7>8yrx$@B8j9F}
zY_64kYWm9#)kHuAW8yb=yldBmN`Va4O944a5FAQKIWckNB0#$3AK0lrBHqLTc6cTU
zh*n;1<K(Pcv{HT^PGt5=h2=-t$JoQW8q}|%whKxrzF2h9Ekq>Tj+G$FG+EZ;&{*;~
z!}B18y?}=%yQCL-`QxC-S)CAawc{jSdL;l9J9}H<<V7(rX@GVO<)EoeOjliDg}em5
zMpQz9;bcCP9bSpqk5&8X8F+}E4w+-W;47*3jqZ<5hxQEMXNZnH;p$}Nw08U5#BW@l
zTK*?i<XFfq8EIy^qCsn|wL*hF{tiu&bVTxIMf?@FN6Oc(^%y-~5_L!Lmm=3J85b<B
zCn|UAhEN2G(4UX>RU3+hZ1lvt%T(YR<b_2wS|NB0Y{Y!C2~G4cC(qabisakTL~-c)
z)|6shCksjmMCfw}pH+IZER-}I6Rjm(igNR~<BelS8>Qqyr)2cvwFcDV;6mto`7@(=
zwhjaTVHL060cdCo0b-PDahN~aiI$`qA8fS`$Gz)7lmP$h)v1<V{Tq?LJ`#A<DfQfa
zIN4M?QabBf%GE6Q(9|7s&1Ge4%-do80Ps{00tF?70h_D<BcWJQ7KW;jHTOG;j&er}
zU2!d%<%{gc;Z%+}7WCs7OKrs_DHHA)YHTsIDrrlq2xo####xi4Q~2_NQo~Cpe1}>)
zN0Zr|6z|j9W2?m7k6f3j%-Nd0rYYK5&Wr0i9sZ%|Vz-o#PRt;l86@#csUo~XNpf(L
zV4B0CIQg*-rNjqC8XH3#VAbd9NV8bb9B$zote`$)=m0{qzon9Fo^G}Ir;0$6>p{LI
zOdK?jCKi-LBh`*`rM{<N4t1<^d@t5+>qY70cx-J_UZLxnD}A+&x<A0Gv{<J|px`|B
zjmNr3$c-bb4Wy^qU%V^Km9coG3tLkIk+^pK#r(b5WISK6$=#kGAJM4B-w0a8BF3Wt
z-Ini;{leC7ftbyXX?T-R3Xw9fKP)r0#{MT=PF_s;I6iI=PfcQw|6;E@L**&eu-Psp
zA|Glz6IH7YJKOQAc1npY8rZDV<y|1v$FmRX1;;s&yV7G{idtK5*;YGnd|o(Tn%J;!
zZCaVST|Fs}saI}lpY63=%FMEDb923by}49osJpQ}ft4FD=tH)7r&u;LcYY2mYZ{OK
z{HDErVfv_Pr8B};pppBid7WjD!&yL@U2&gw>c_<hW0uKt2JM=_F<s7?yN1|Zp0(^r
z|MZWUBr_;L3*~!;xT?OJh?h4t{%S|r^>R(i0^!T!^8U=XnY^z*b!>0$$@XReQY>MS
zk*UTnX&S&UM=RZiZW+ftmgkU*A2L<rKrOdw+MJTBm|Sc*1Y0(4zmhcQR;Kg`)5}B+
zwMT@O0yp}T!;EJ(w&eihS%U5C^<~mXF|MHJ!Md2F39#JE%tB~AvEe$4m0AGUjxChQ
z@h^QSw402qv|uxL6QpEL(CpG4)qYL-yM1kuFX44fMBo6QI2P;Z>3G6->CQ)@Cr_0>
zYYXSe!8RD_iOaubg%rmoGPw?=BDGlGTR38FWfOS}f8C^HOpnrRxSvbeL<Jff8Rz_=
z5D{IsJ}EP`;`F1NmiKKcs@8mn0=PD~dqw4lF~g~JanxYzRnmDZZoSPjh-#eiTR!^b
z*Vg=j-S<*6Ze~$$CRK|HHT7+z)^fwou~}i5B-H|hunsGF9d2%ezOe*MarJu&Hb6&N
z&9YhYVa&2EEW&ooBC}7}8@59q$b7HOE}0ARe-Rc6?Eypndj6!L(f(~Yb-!{A?1?kp
zoYop-%oJsu<@UJ5f^lr~+D3)T#{XOCLjk3+w*1Q;xxET%lvA(nZ|WWmQ{RR;EuYGk
z)5n~ZcuC8Z`<PXqMU{g#-(5nhEB1keS-1`R7i{12KNrr7=4#Sl*xHtbIQi2_WzA$Y
zu*hKIH(!(Jy0@)I7K`EJd$@FC-qdfXDdoFADL2f*t`V%QW$rh&!JO!|_!2-lfx$8<
zcNwxCvJ`P{S~l|L&*D^Ps7oi^tR9|fXj(hFI|x6(?XP@C|Eco+6R!TBOh41V1>Jv%
z5?RI>kJawD-qE}F%{KuoL7-w+P`P3KnFier#Dx&O!o4D7;y8K{0wh!PlvIvEBw|vi
z@$f_tQ{E^~9$yltA;_i4$=%vQpwk7a=+fo&(m?Bpf9gr#I=kLWLYXZ+oPRu|B_(g{
zp9T`dFJOMUf8yP3_HyH(*@q5dLGMyLKU@jqOyHtiw8=U-Ot<vUv3a}F%@U?8Sv+_6
zZcXa(eZ5Jj3h!=PvkxwKAVTKVGlzrj#iF}4pKlVn#@9=cyv0=6P4d39?^UD0V`Ctb
z+DpWTTh`;+Vn1836@$e~dX;#1$Ze|<Vb~KLg!S2>1^!F_BmCEmM`Eu~ipp&8q-yN3
z>mK@ws*0oiIAdGfK2;NhX1KZ8gsF=HO+-K$w?jptj&|<dZHiQ+?cU?I?S4M&N3`2?
zquWkjzfCC@4ywV{p#{<3g{dyFMYN@w<GoFmMBf2ngs4$-%NH(QrC6sECI`Z;Q+&>`
zg2gO>_y#tV-WmRTv`l=+XDyKrs^6)0O*{r$9H&%qyJ2*vScL}@PA57^DG+x`mD1tB
z6X#`nD@j3JsLLm<7}qAR03qw3;M+hOx9zi4;k=ew!Dq*)Se&UV_FrL+u469hsYui_
z&`bZF&U&$y_4JWaH)E&2cCASX4XRBjVO8Yz{4YX8?L+1J_Xbv6cv1?}YfVzi@9xJ7
ztyU)Jr-1T{;@|?T$5+Fv3jD{Iit!Gt`j!GNw@?fgucmai>N)dYaMabbTI$XutT6u=
zz`r{#k(n0tn}Jc#`17bTm``8cA!^wYn*rL9&{^CY_Kd+M*1sg2vc%zvLa}yME8+R6
z=4W1>7casA`92~hKc8RcZQxMf^~RO={?f3nZ!d{qA&jpGv%_3>l^Mp|jb&~>pYO+V
zd44J&X<yVbDKL8sm6)lw$MYWArp=nYDQ0nd$&I!O#D$k4K}8|T=;U*QtalzBAlD8r
z!`H0ufEZC=;AYdP7-X_n<^<L=A-FKGhL{E4-B&_`<+eY}EX)f^fr_i8V%l<53r6pE
zW|@&y-@V*E8o9ixwVJ2da*6oO4dVlEK=mUkr2HO5>}H$tsPp~^!zpJ9w0}jhdTbDV
z%<c8#2MN2CU#kLanHM38vO*Ej1@ejBbI06X>E|O!WW;Og6#%S|W<(L?m0JMZn?63j
zy35{!N`$Mfeh<|RNd5r#Rt9KTNt9~bshi+`kK@xPweOiO9qN;fFd9&`Iw(L}H;03B
zl+HYEgS@;`u+eqbLR0B67)F62=f=`&_iw^MZRU|))z*4If67ZBC$iSdJ=T~kcM=th
z@W%yRkhV6rEm5`gbGmi`UV}O1F&`tz8f)edY7Di~2?gCmTZnSoVN9H&6WVcd3log1
z@J=ux7s8jYaX~PDPcK}tOK~M#$xU*Pz0BHF+QpY#rBRyKOwn<ZjlTQZ69osZ9?qL{
zBu>K3M44(f+9TCC3*J#Jiyb@^0E(T<PR}GM8>pXpxyu&hfz(X6QTMn`)v>Omk(Z01
z!zpYJjGYUyPP$?Y>!<+*!pZ6GTobGiJaKvsG)DI8dq4}v#)ayWPt-YK9w{E1SWa`#
zThzxK?v5tfsu}>qzp9Vs+0H?#LEO>L^J&n>OVVn`qmsIn0HTZC^a91&EEBQBSm(+t
z(`0(Fzub7VG`sdf<sz&{n#EVw1{fhq!4X^jf^Q@vs~1ZZ(~jt2^q8-ov>sMA7Ri?e
zBFE@Y`H8m{7nG~oJr{oyj4o*PCmCGGm2emNQXMqm@bfgD#BZ1h!C+fjD~VOH3aqUI
zxN>e&axj*dKy%0Hhxc`l!JCv5+UmTNsZEHdG?X~6%`YvSq_Ry*P=0|c!K)Rt)A|4w
zpTTTq?j|y=+cs@K3f}9-v=gW<XQPeM0vZMWF?UmcU1X{n4ZqF+lT=KT+66IpuGVa0
z4u5>9H#Bx9#0lN24V|fR@tu1d^hp3wmV7!P`6BTE08=D+C}6f%>72hRmIQ^uqqhl?
zNr~W8WhB43`&AOcx@)Z!OEsCzOtt$VE7ZYe%N$M_m@SK@Ta5xdijqH^YoyYxjiOCR
zEbLPqV7fds8vXR;E8gj*Feoi2kY1qqkaFwcbI71&Kz|rv+9XLmEs+5emG<Bu-qc6Z
zi`PszJAIp8V!OQ59d@EnVC`CXHE3nkyFs+rU^PEGTZwIXO1}%oq^7uxN7F>px(R$q
zz`LxQvCE4BO2LSDx~+5tPxY{9@mbhyG@aovV_Mwfpc*5{BXZklLbVNlytb^{M&$s9
z&-a4E_hG5Tv=up2$|GfI9LlJq<|vn2b+b^MD5B%#swirZtobI6c0y&H-Pma5r)EgZ
zHzsVvwqI)vtFW4@2e3n%bW=wt;c0;__Hwo&kG&mH4C9LXI7*o_Qn(A#0HK(=|B5^{
z4F?%D%*k;5)ABB6b1Exc*BxFz7<eFzq@y_d^laccxOU%`0)%xa0)@ZmPMLE)BC+qV
zx%<=V?eWeIT5~}cA}s=)MUB>oSz&OLwqY_AE5d1Zt>1yR!2l~QbH{PB^@sDlL(Y{v
zZx)+aw0>p{>mbgxSF%*iF+BCU<$JO`ML4MmGc&Fy#nKM&$O!&Wq^ib$K>d}r&J%t2
z;MsqM8<Y-(5Ym~yW0yLl(oCUZs>?9{hK}1T%_815C~*Ch;G9@%@i6PCe#R|9Tl3p(
zLX`cI`2rj|uAxH4RBLH1R#$1U2@3-mb{Q9S`Wn1g1@s!}x}%CyKlu8X+g#wSc`^0x
zP852K^I$*D0nbIFov89MTYn`|qL{N+TMlKj?w9DIP}xI!=i-yNO{_f36GU<(LMvB%
zpF+mk1|H^`d6|N<rV$b3v0|nS%x3@Gjd_t!R}tNf4(x_M91{GG&~@zatoo)nE#=$h
zJ1!<nfhBc%CPr4#Pnl0_vWVn|(mO7v|J?-N+G+lJar^1QsZB;X_Tc6(F9O(*8e}E%
zzcD=sA3{9{*LTU;e)TT`O{cK;oWEa|PpeQaYwM`phXCo<;UOv_Jb4*a@0Ps^^(kk3
zTRS_HfaY}5PA*|(^A~ZK1O|_`PV-wejEoH5y^{=9Y?(|GLG5>CC!}CE;_6}?QdPV=
zNiCl+H4~ou$|*Vg%0V6_mQH21ONzQZkgl`^Pa_+$$<GVO$rdWtB^b(EHp|Ryy3{Kf
zE-n+H6zCmF%KA5+)KG<9z`uB?q%D4xt(_5S`SROU#{+(H`h~JMI+Lq$;9){$&vOa*
zo8%irm*|hL?upI-R^3&)^#@hm8kW(f8s`Q}%^5c1*aguLSsuZ3{%?Ksa^lW&dQiL=
ztn9AsGIV8X83b?;&b@s_g>;ltUG=$J4M!cRF9cPhmL<yGy^!q=TqU7_0*A#wvAx^2
ze=$O&D;v|7idUNEw73GedvzAv)kzom!3&n9bt!9E*m#cPglUAe{=%vbZkO*eMqdpo
zDg%YS?Hy)DZvjFETCJ2P<%y-s6Pc7gr|J{16U8vW)(sBb%H{zo))&@|Ly~oM(s&g)
zf6p0Xvnsv?dYC4D`(*jMy26pQqu6hW@>t7!7(=TWo9`&rhL9$<sEp6<0E$5M;SE%s
z87$#pj97KXq2u<}GCPY=ocD!(47LF72B1@2x2Vz#l^FSKwN8)x7O>Nw6_@l`OdDlF
z2#EAyL)<<)A7jD_Mh_@o&=>AO2`wiGDI<mlSgKp=ExKHo;5iGWXvsp4Ef1yi!Bba;
zze0yEniW*mI%+-jBMP|@5v@eR=`|&0X0B8}mj-dD&I3ji?%oaen3dmkn)G_;6n3gd
zY)!Z=+YDV0iuB#Q|6S)eb2t<t5=RvmaCL_P3eh+01g}?hcrG(ivR1R@4zE+cKF#|;
z+x#-gAQzi>eQbPMbxc$Uv!fp+r2-~A`Xra5({&Ebq*+WmcvZnNJc68SBy{6iU>2?G
zEk&aZdoKuog;j4HH&U^lzLTT2y>;q5s=7J5IDR;b^j7n*=ic1835Nbt$ZEOq)nf{R
z%Ty>JaN^$?6Y0<C|J?11xxU8Xb$aUYj@IK3SPATo$~rE@3D4sZSvc0=Wg7M%BezY=
zKZy5z`w<eFobUi+xtaS35wSIiPRtzmx{6w@<I!q8M=QqN+ff~jPG|CTcFC)NXFOce
z@ay&f9>C5ORUCOiANhYh12X|tu>JCM8GM;3xpV&)AKpUSmk5^Qvsx;nGLGm1Dh#z?
z1z-tHk*j3`u6TJ`!7l+dV4mWURGwb~<G4~d(Oi^mt7laxL{|;@@vO2HJ`9lEZr>wk
zz955z|8PFs%3CVU(y}L7M*ld@HznMiIo!a+2S%UdYMIb3=FM&eKcjbCN)Dnx?Qf3A
zsPA1JbglR}Fk5i@8c5|u0xYswuJ{4fdEiI=r=f}C|Bf1BWaXg$FAYsS*b=cg91pc>
zN53LX<6&w2iJT>L80p#AE?Cot`Z2o@T#LH?mAbsWUK2^mv?~y0MaFUcE1dB(Iv71(
zT-mxXH;fAOA(}v>$;(slGw`yXehc|SL!k83`fgvX@3&m1f^(<BxC3p;cg%O2bDTIi
zWSVK9OyDPoy;oz@?l`AnncczJ5|&d)A?-5_f;`k;w4WEJPYh#zJ>Q4@D23S^*M9GH
z_vgpYcdG`%IHdHyBYAzeUNIu^OVD5_CX*8<cYic^33y}mCrz_g=%0ShR(JJM4UXD|
zImRNKw5^6?odV_aYg$~T-_;VFCLCgi!NJ~TxCIYFiD~oijPP1+Y0*`Qtfu+p;(6p0
zSP?zmzf*d@VtxIMM`sPvKx8OQw+rVPC7#Iju@DV~BPivZzcJkBV-V@6Jo|CkNQ4Kt
z4koOHE(PR1gak8_fn1E^h)D(1A{2-l5D(K8BiwOP7UIHO+@Ic0)51zVsR)t{Br^kl
z$<ZTgyL99cWDMdzknzztfsCztg=ihOO5`s|V{Kj-0f8k11V~x=lZz(9{l%O}O@XAB
z29j0xKpv*GLL6ieqy7c?$LpX;5!x8?#+Jn?CR64j#f7?DL$&Vi%M}7`Sb(AWB*7l4
zP}UuO$7r^+$kKkC8AAkeT_&RqM3g+8%&&{!LuS_ubrFIrT+8$wVwC(RCY3l(CC(Vy
z9=h3*=ph5i9rS^WoDntBBvq{~a2)9(Kf#TwO4<bhO@CWMVu1h5q1A0d@-%%#D=CeG
z&!-Iq@BV$+4?@B`EkkY!PQEL_icxOHZEeJi#yxh1<?$)i8JAJ)or8ZcP5ABQP@LOj
zZz}i@vEI%@29b9&V=k&92Sc5{%!!#28=AnEgVzE+E+T5s+%$TfXrXbCxM)`RJ$KE$
z7l%(`V*1)z|I$H=s?miF#ydwY2}z2P%Y?vYm}f~#bZd4-!m?kWM0kKzV-+-p3un^*
z)+k2Hegt)}ebNoH4Y429hRg8^cvR**0x5}i%W5le0>{eq)3?w?iZ=bo%UDy@UCUp>
z$}ica$iP!E6ma+#4Py$#4=Ag;pz)V$Ce?ItwHfPjgSr;A*{;r<Zq;3ilT9uOa#fH9
zbBfZ8;+%X*-PMQblGDC2M<C~{yu?1Tuf!X<G+0Ud${AA9CUkth?yWw>#9?qv;!Mj7
zX$^cSa~-;Io2}-TYIgIiu2_D6fqtXmx^iOrmZE}}FtVwsLOkE3l@7pGFo48xUP-d=
zw9!>5J1Z4NkdeQQqPTeQxcGC};cgqG25krI%!qWt9DDK1j-Yi!pkZQ8zE5dim?5}{
zyl}TCeIQhv5l%5t$b~N`)V7QQD(B!*cy{l_uAEbZ)98Ivem*=uo6GpRNMLF7YG;rp
z<v?_SPo)LL%81ZPEf{I1vc3*722<<Fo0(*ul^8m;aHK<@q${x^2gc-v@=(zgwhK2;
zFqzujT1z!kqk6JhYn5ZxVgrJ9zTi;Z@tm$sDM=Apq>|hx>7gzp_|pI?k<271wsHFh
zV2~bA%(p;Sgof7tgxs>M`vrgd{rT{gCiep&I{AHB_8^iV8Q3)0_F4`HNCHW%yU(bk
z1mf>`Jqi4+W=ydPUJT`7fc4spi+Y?yC^2$*rp9+cF)1&|93|wp{2JjA#Hnn5hd61z
zb8E$w-cYFpmI;mhJuW0+^+i%MTlHD=(>&`zRrBt{;jIf#fTE^gWz)T^Ni=U}IZD+s
zsv`N;FK*7E6^iS$@REzHmQ4h=hHY~O2ttLczBQ#V$k^xagFX{<!nZh^5dxzY`Xlt_
zg}b{q#7`ZAN-9ESMZwn~t&}531s&0qF!cOVW@oXDq4`|!7Hgl|9i&o>D7O$PPlQxG
z$Ga;oweZB<HNFzmGOjE(qSI(*U1J5rzv-im?M$pAmr@_sl>yuzU+kKNWt9JK4?{VB
z+DdQ0MFEZ-&q!R}5yd5GjZ5N@u>#03qCyf2I+USR3^~$Wo+6(eH$yHVP{VTx^dXTi
zr|+wk>Pm-d17UD=SE8)2TI?kIodGR=*|RAtH*(EZQdSFGxq$Q6SJtcQY#0T@`Y=~+
z^_<3vamKaK+b7P+{Q~-%MnU8_%oaJ_2o_L=z4mY}FAgIDRk%OiA|R%E9^y}J5bu@E
zcd*7^;bfd*nBZ__b;{lvA}c>yEet354J2P6K^o$CiF0~I-V)FTE10PCOzWdh;Mdy=
zJ+4m~4qJ6_!@B5dA)7!y`W2ePL0w+mm_x6dNblH+B*sOwMgz7Pail%nun(RWg2V*?
zS7fPB0hfX=<&y6(v6aviJOQb%<8jWK4&jy)Ym)9z+l1n5<4_-mF>zs)@io(=AoY7%
z>veuJ8t*68o8(fI&xtNOOTe<iNqUXxXd}zQMIPg%j2-TQ&h4c88@QyLCRc|~goigp
z`lVT1$zT9bTrNi;StgL;1DGVq$#8ZY>gzd92UGm|c<?TMFHUz-{o#MF!1f^AZtg8r
z3Ij3_>+r03K+nfwGZn74i}7rQU*j1&Y+Eyrw7i1J-`JY4yD_KHaiY;42sctV`@^-K
zr73^-hbYAloQ21XTuW39uBG3{R+95<B|LPoqqtI@+&j@ito(E|#pqTuaRO3Iqj>b1
zR{d0zLQUb8yvXuEC*HJ-+)Af<(~T!M(v*4YIiA&d&F@SlZfOJqBHi~)1_9bFU$ozj
z2Ty!e%rjpr4D@`fCa>GnKB4(TE?~i7#ecd;>!-ICyCS^S;M75*yL1jToL7YVj9uE|
zB)fMyyb*|!{!W48#xkw%%URo;4%~fsyw~oJ(i%2Vt$S6atpG#<`k@y|{5QB}X97#g
zrQ1Vw;<OK92maI}(tl&@oq{_H+jh^`ww-ir{bL&)+qP}ncG9tJ+ji2iZEN?gnW@<|
z?>AL%)l=(WovnlQ+}Ew&C25<Rsm@QvYkr>1N^)T!y+eZ2#I29}2%`1*Fov!nkzbVl
zXI5gi#|td60g;unnveI`d(exK?`(YK4MJ*XMROUL<@DOuL;Buep-f}@g7X!ZVnD8u
zP@e@LgCrl_<uiTkCFc-4i_+ZvgZ4`=*XjX&XJOhzoBHcHyU9O^wo8g2f<x^=t~(5c
z1GkjirIqdE0yY6(y%qy!EDJ<-7^TeO4B`j+9&A`w+OE(J^5l1yW@}oBGo`m0g4vYj
za!hwqVzZ1l`&h8EmQF#o#gk8Lw|A%K94MbE)hBcq*_WK8Sg~uH2a!-}N%Cgs0)EG|
zjgQX#uDX6=^iTytzEuU$=0I7)&e}mtVm{}T^wFI#`TUi#jYBWl+`sMnb_j)Hdd01;
zDdvWeGerY1xzW<pMd))IJ>+wY0~1;&meHmQ=|WSEp-}2AR{DJ?*H@0NGIyD$q-2hZ
z(!XuYmENa3R5ki;)1!sGA_s8<o`7xhKH}w39J4dbZPyGjYt)j1j_n5%9k8O?NSp4h
zrU8OKPZOaO6Lm>{+em2(CJZ`6nhcN;GX6A}tQk}yNI%O(34Fmh-LOaAO|O(_ER7hp
zcAl~&Iz*A~vRADfgq?B~xXemznPV!H7(W5&>MzQ4X)h}X27hF6^7=2hGRX#s>0B_R
zrcS*_e4qibdxl<Z+KkxkysVvEuGmU+brF0csulR-|EHb>{O@_(|8En3nT_SYv(B8x
z9k)GXf7aD2AY>dI%@HU->fB(D*D`y@I7E7saYMN3Cv!^Q$$O!^PNXfbmip_jNgnC_
zAvN#)%5}T<_IP{tIGCgyWFi-YK3K{C`i~U#13D^L2npL49lW<Vc{`cmWN_=*C+$E_
zu8ZNv-E}dU0~W>tWCqj&-1b)k{KO&F`HtS4zNS5vap)f^YBFcP54^f3F0J$P{j3x<
zE||}k>nRw!AcNKQ_0#d)s4rqxINfJo#<Mi+2mW6qoSQ(*#s;#wy(a(0{T_<o|2(-!
zBjo>qfJ^&IE4deBD(LUrGx1)El!SDB#UO-l981UX_aM`R-zYHuSF6+^c;(4Xu7fLb
zT|VW&d6EH;i!`f){HbSUA}$Lhfd*qlq;TFj?-x<-N8+4No<kLkusTCtEme7xQ2WX*
z*{|3MTTI(~WSUUMHL!?0Al0|K-N}{-psF+HNzq=6caGMgOOok2(ws$H`FkwbAI%~~
z;uoud@Je$qfmfOkdBGZ7hZ4-1eEszrh{h%7K!#gpx^ciY9h_pEi)6w*(faxQiL&Z5
ztTMao6YK=}kDM}XZ$`UOOY*zlpgGHBjgwnW?mN=A1lVTcKnE;O%Kg~8G9&<s9|4`0
z)W|KQF)piEP$9SzenCQcT0#&<xo5^uJzOgS{Ag}935)X698QKh6n&}O>5#ZTx)=f3
z!1VFNOyp1>BVm8abD3BXYq2Q?a|94_ujamLi>KtSh6UP8cv4Y%r$bQcQpBPW9)qH>
zuO7c@zln?j1PwP4iK;R5AuJ-?OSqUNpzNL`PElk+c@2_k%0mVsZl1vu#Zt^Ko?7dX
z^G7ckX6cJE(LYE8#AK#pm$EpYIEk+oGd`Jtb$p7viK_t+8%qoJM{LC1`0|cq8Ni&n
z3NG?avYSoLUtew>kP;qr3UZeG6KG!p&Og9C^*367Oj=$T$sB2(LJK%*Pw^BQjqof)
zsvrVUB0GeG#2);ZM@l-Io5ejcl2A=~{?~6w7`Af7F^bH*K-OA|SYG*0;+Z4Fj6_Zp
z8S7wF7wl*=SGH<pzS^$XhgUYNHO1xmj6NZAS_50Sm>+@-i+b1g%E&Z$HmpEnAPp^L
zJ8*7<mE5=R4M|(+72UJz%v|l8M-!14MOfVfoVR#{LRqk?3Qm*a*$)(Bm{3jWVUBw!
zkq^l|qY-8ha30G$s)0i)f3lsdd-O|jF-w<vOOuMze4KZ8X`gtrP`=~2k@I7=db(tn
z)Ms&aRA6%>Wnxd6Cb9U0&wCoe?+6Az9Lj$t*|rwvuyQHsQIOO}-wyAunRL)*8TSN}
zauMK4<%mnguj?e;VWLb_=&FYmrG~_jSN-WsSl)K|ql^?kMvR<6bfGzeNMsZT<dMq-
zaaC5Eo~HdbbI%0(K_*idU{)T-WgbjeCNkHL#`y9TuXMzkoIBbD66L~(>qZiiJR!)p
zp`T%l@h>!XJ{~+1VD&+wfGk0yR`NDolU7L`S<S!0o?%xv2KcgYEQstWOH0O@&dR4>
ziD=Ib8z4VgSSfh;;3~A>s6BJ{5R+F#vTY4S|3WsprRgVQh-ZW>Co?r?QU&`F(qLXD
zUVGFI#;%vas(DEr+p0REq*{}Kbc_I&g*+Ai=(${`a_M!FaM)Qll#tbN4q-BfObUhC
z(EF9JT9O+4sxo`$;XJW7!#)xS3s&SHRY_GQyfulb$>!ZpL9pAM<(6wX&fxU^Ig}^I
zF=#=Z2&;_IxWe%05&(=_it2ZCbFGX@eGT=Fg-auKGQveC6G_zgJlrXpj2f<aJ=KDb
z7dotm17QrUZD=V#(tS_Iy7EO+UV6L}HQ90P@_l6Uyrv?KhXYMnd>lA1lWsl*Wy+{y
zOW}{*&)gp{MhmH=8wSO^i_k?^Zu|1&@=e26q~%L^gOW|BmnHkucqI`a@FDDOtX}GH
z;3l)ksJ&S~cs-%PdH-o%v{HO+{1%5dbNOirOMV{l?{+3lj!)3cY}KI0TtF$itz`2t
zGMo#HhKI2w2=*|#P-dH$2Gt*yFU&DJv5FPMEKF9QP2tPTH|9<KPEv~Otu)!tx`g3H
zF=5C2EobOh53#}E8pL4|t_?7#1w)2Rc#T9qNDp%{)eqi7gVONTfC^Luth71>mjR2=
zRj?mx6IrjYgj^SSb!X7%3Ix(1X9OeJb)%JGb!SUqgNCpKd-L3V54QUU8u~v>sYY!#
zb&hbc6y9;N&A;D~^f`y{0ULt&Cur62JeVJGoZMG8Vz1I4%y#j{FqA6`Y2%l!ROy$a
ztO3!}JOiDgWtPN*$Q7F5Rs|I~I!#XAW#^}wh$=bews;jSNohjM1{IS9mPzN1W#_jH
zAr+;ELmM|GGBD#!%eX~@8F%AJj5T0o(_=IUw8i>=WwQlOPZj@QZAWTJI@0hlU62ko
zFGyI?^BU>@J-(+zF86}cpt%t27Wf1I*TGJwg&1$f<f0I!pE_*k{6S(Jr;_jblma>Y
zxV)&46$f`X5hpP0{2rOgizZ~^<3PxF0b=79#)r!o;@3t~vAfbj^!-R%aJ_}tE|pmp
zSxfS$)m`z8vI05PpOLtmvkzs34Lv$31WF_i<LBu35OD>s=E{Nfyh4jwt%>H#-m-l#
z@C#D#utOtO*`Ro?aOpSB@rs<UcMgh)`5lE{(HXs^R)D<#l_~Y~v12i4>`lXv3?ptI
z5<C<q$a6XU9XoXErJUcP9<=J7pH|(2-yj8pM6U)40-UoY$2)AGbPes*0mF@!7Mp_k
zS?iZYBw&K^FL7u30)5C4*+L2FTh&(uRRC!G#tCtae3(Th$dN*n0+EIU|J7DoFwClo
zu97gCd~R+P1Xz{Q(k5#Jo+-SfdVqqzHxjn3-{*dZGpljN=4VC+-vH2bLiBPR>@-uv
zuxr_#wI`m4DsK239tg@8a30`GEbS5<Oqmq6wWXEaE8QprcUKm*Q+gkM^=ul8BCXkH
zjlE_tKQy2>X_{%f@oNyp2Tp+Sy6kn^M+R{!@T!$5L1~J}Z<vAtD<WfHTi+}{$h&Vw
z0Mfz4EI-R9KcnyM113$KBtVSYH46sQ;*`;+cV8?-x<nq>X<(;JZHA{={m^POJgo(U
z{Od>rQIfxBdW|;LKSKgomyXWM)_b{F_ai8BnfN7Hk_lFk?GvL~vVzLCx(Q-j%XtNe
zOUk4gN|kk`L9rb{6o_g<a#o5d=@eSM9rUk$7DStoZj{zjwgYDO=)GopS8w4H#%APG
zsCN&79zee;9qot)a>N)lcKD>Ck&ml(b{cqdy#DH4-<pd_ZFv4@gJGX?bo9xmT;yqB
zPk-y%Wb}TXWBe4cpGz(YX*0XaJrqO5KST;LxbCID0PpdWpQwd${6B>wCxDUj{}qn^
ze-JJvMn>lUhr-3e$n@Xh#W!O&IZ=FO)u7VT*1<DLDkED(k%$&~+YJJt7c}%ugGTzU
zR$Q|=Q}?9VKVG~dBmd^CWpK76u7}zF#t<byuawInNl}zbsaAAV1R$6rsgNlbCmV@U
z9bjhB+$nslO`I&Ot0<`BDx1*9{03w_dSKF=X!iuE3ej+W(W%8&J(EVIy4wJn6tPp8
z*v@34efOC>FnBI}7JPv)CdLkF(z?uPl>Y&N5I`i-o;0M;9Vnd+BYDp7tZ860MvcPK
zB1{k})*$e~`!cdg^vlzYJO<P*HBLg9I5l&EjE1Ss#`dY}3w8^ovQ@PD5cZ2cDa>8I
zyvdc3Bm`(q@la?Cw2;V}y7*uZ!oi{dAi1}KJBs>Ozqj#p$SCmF3-wyKV4@FfXx-ZH
z$&NV6yPVy8j(nC9e9F9yn|W1d4)~Az^7})X8PozKY=DHpsH38A10u-_D?xUxXjJ=^
zGUy6bZud-tk{1Ttsm?z_wMDI-c<r(A(-J^6>!V3j$Vz+rhZ4H@a2?VbV&U2nL>2m*
z|A=&FIjp8}$lo^oUI9&G5*xr{beK}D&SEUPy53(%;ZRfC+qdetWo&A(rxN@y;n~wU
z3sZYyG;6UeiBD{q1DMVO3|rC6&){V;svlHT{RC<HuIeex_Y{FZ=^kM{2M0FqW?s$q
zuH1D$%3C*?14B%5ToP~cp=Vo@(SkeLQ|`Oo+n=_8a8wJX5ah*!UGS0c*xm4vm|okv
zARwF$n;0;Fra_R1{Y{U#-ELH$EVykjPafvm%CQuFx+?NwR4YxHD$6=Q-@eZG-YlCe
zsaL;l^K+`Nx2p+H!`U^y>V0>WuzNcDDgbu;e<g92!VR_mcB=7fH61*dJ~jBdKIi++
zcj@W;?S#a8iu#$so$}rSnzLS>Qk6Cdu<bgTZqQ*EnQkD3_&T$2V#EGg>hUX20{bmD
z<s5K=>5js4q;%%u$8I$#AhPx7ri+2FxJrq)!j^r7TVt2}jftyol3(#%uA|Ks#a!B5
zMz|bzp`^Y0Mepy><#~pZ5iWqZA1w-o%w0Hz(ZwdXFr4%5?bE&h?3XdMz4XE{7^jxx
zs+za@<NdAv<Llk5Ii2=Xc42Dz&Emn5aj}`{>Y#I>;?qLV`vy8FW4h`$$i2m^-~TpG
z)+yL|{Rb329+{Ye350cp)vQpPhAQ29P}HY);P&Kc@L5zVPlaC$qIAHH>Nvei6bbpa
zH_ZOz$Ua((xG~8qJRErFVAriY%r@DdT8@>Ar_J_u%kAX@wZi2;HajbwRgI(6D7fUz
z_F>4u$m{xslB_Fhl;S{M4ZiPFUAZfh%cqVTOG(%`i)Up#&x3ECe0G?X$LGfDayx8M
zM+T=FfzT4>P-25c<rEoks^FlwWcvowp3e<sU@^@*zCV~_IvU-X)uoG3Ir!Nz6vVY5
z4bKj7xgcxomy1cHpJP`n+)TK|=L|1ksJO-lTD-mR1_b>HhOL~5i^O<T7J~`YbzCHu
zK(6=a!iA;sQa*(z3fCUVUIZrRxPLj`6WR$gJ`w-~A|3lV0d)c+Y&CPL?C#j%r0ii}
z7f;4dmOd6Ew-3KA&Ws<eS$$bLAeL8p_4M@U#3sid>-x~L`Wt_HCqxIm`&Tgu)}p?0
z_}~^jS<bUQZO)^7Yvc{sP&6j#<q=BI?wXYH$j&3N=18!Ds8HY`(cG`zHMi<>nd{7q
zmr)1Zimrq}5`_;Y(C1^xqa6%B{prs3??RlSqQ{xhWCc4=uI#rcKmd55>JUKepEc_F
zeqZSamkmclaKN&{rKYUms>EK76R8m1<i-hVM7iqb$%!{!3>y4%SPABCWSd+2$71S>
zmYHUa$7UGKEMzFXG9d&^-Wf9M(Oujr%wdm{5sCWx&xGS5?h~SY7LE^!pX3en+M2$*
zwLLfn79mw<#bCIJteEQ{fSVuIMtcW{vK_tjiJ%OO6Czx)@OkC0gJ7t-$q9(ZCsb5c
zS0F0XfzB^@IJNkCa?+hiY5D~v0IicoAOzeM=KFd63^Tm=aq&1yI3dbVj#y0qA6Vtc
zVE}%RY%_b?dn`w#A2gX#^H=uIb1FxpARN72eV7@>n=}YCh5buFD*c!Kh;>LYp0cZl
z1`Hw-voMs^XemQA&RhRo;*Z`Gbz%-B(UzfKt$cRf%Wr5&N+C3^E;+B0F7jKd#5~(R
zWgjQtO<)mPD=c*DXOyaBcsV?cxh}t)M<vsL36xBZiP}-0@~iemY%1Of;uYzI^efyc
zPCM-R9}niomNBNd4Qm;Y*aoP$8ElPQv`NtRtt6G1RhGee^6yt}k%*o8-hZtOD+jEC
z&NV@AD}^bs3pxsOyPI$8d+)j9P+@d>ecHe_dDYeu!LnQ42|1Ttg(1)7o}3!9!dQL6
zrl{6oBgA9Xl<~uaY3QfMVwi}<GKNZ^=#02oU^D#<oSzlDtaT=Vj~Hsm>65+g8LzcA
zGVou_MX%f_dj2D;F|24ZyK@r(6MtR}z_$~S6govk+8_GWm-1%{c@j9|DFD0#L{fn9
zn)Er3Ej)(eErr+k<I-P+>yC;$c{9HnkIkby@WZ<foOcxERqED8Jq$j+chWnI?_D+X
zzCFJmR|^Fz5XR4=H2sdLg+hxM<MH#V%#z-I0aurEv@u*lZJ*bTPYZ-{9cYoAXk}T6
zo1C}hlphFYaE)ZiTz$ykzboUu`!vTRZXh0%eHeKD`{!V3QRpH`Dunzc8hMt8hoVf|
zx!N^lw%Ho_k)EcTa%g%H9f$%BFQ3wpp(XnAf_GO72Z`c?nY{TlF2?Wg6#GTlBN)4*
z#9R<PfZ?>2f`5<uaq;5wcN6$Ylr&Ty2cDCIg%e;OTot}(EpnSyH5Ddn-)B(1qj6BO
zuKn2*{}FGk{oa{h6y_4idOj@=<B?eLdNfXhGx_0d-R9?N`_yzDvyH-6+z!Hbd?suZ
zM&P)N_gv2XKhMA&j*k=8DIE;kH13kk`#T&;g3XN+0jw#subDynF3~+CR6hZ@mawM`
zg4Bb%JaUotcIw5v5TY1Lo&;q<1(Yi%lD%5;wK6N}%0z%_W6WkM6cTl+w~QpG2VyGW
z97-(M+op{GqX5|kNW$+a#wq^=QZk|DD|aBcbW{*zroj*-9WfgvzGcBZ{1)g0?*=Nz
zIshz~>3wo2iwMQRR*d*5a88nqDw-KD6X~(hqoJhGdpKI-SD;Ay@}B8~(2Mgu0k0$r
z)?Mpv&8&~6LvVv2+jn^b)lRrKn_EhpiEUdYCTlq=>yc3gV$2+e;m+FQI3bSlN9YCa
z5=Wbz#LlUBiLdeb|9(se&ykul{*pjmlI_roB%7X^)09Z*3L+qP&=;c$5|!33$qv;o
z$<IsUgd$@^Eeeb(+m&qHL(2c`E`z?&iEZ!}k^03uIu=TIcH-}8|JePThzS{G83H`g
z2JS}#Z6}!cj!Ud+(;sP}v2oV$mBo=un?r0`I-Rx5Qw}{SD7ygz4g~+rM?F)Sq7VF0
zBKH*nlrdpz293DTJr?<s6#8%g%^gaZOa=i_pbA1+1s4ejoUV6}A)^in1RN9^9E(yE
z(jJ;OQ?XZMIe&Q-0{nLXl-t;=3_^<BJOdwTQ6ro_Y>_k>^i7j%n+eRqRj!V6N>F%w
z2&s`Ii5Gvc5*XTV6dF4P^?&#>AnMdBPXt!rf?wD`eGI8j2b;v&N>Xf~^kTd&lNbWv
zCe77UEM4p(!4@+KHUq;zL}M;lI|}@Y?MxwoLjA3S9OROYD?+;iwF=wq#8NWhUcAC-
zKuqO3tS^+0Ypga|Bwhi}ZfHp42*I!;I0I<!IXEz|#&W#iH%vgNYvy5A11!AeAkIe`
z{2}Pvcu*4FDG2{Q7TvY94lEOD><}eznAFw35t^4Y{Jmadfq^c*Ke_LzPN^4l<5~J^
zms&_d1X!Nz#MWfKG3{vbh#~bO@A14h5>&13!;hxz2a=<CaPmJ>1^^@b|H1A3zgN9j
znA!h3Ht&`8bPaAN{B~YxIwoi2cQJ5u&_>TFaEElf&b%lx>lq1^ryZuxmm+^MWy}Z*
zpKB11EYN6+*dK|=KYWUhdyA_P@;7;MdAZv<$+>tBVw?b<cV=k!zCFMGg89sG=>95{
z#nPKaJ%ns&*jm&s))T4z$$m4$o8u!5XhdGHN2FVk@HDg1Vd5tRzTU|>%#mzVc#nXa
z=OYcLeiJO8soJ-r+Pv@2A(q1IX?AZoyRX(&dHG)cC+mNAG5Pf4+lP{Sev1CY<hZoy
z$yr>lyk0KfqsXK(o){zQcaXfFcTPuYoQ53>WVeG=jKw1bC#917%8X3NKMx|$<G=n0
z(f+gf#GoB>`Jkl2TBH}9QJBILo~A+!ltC4Rsz(3?1FqYMAJOM54E^5UO@Y&YpU^}n
z<*#fq>KblrtW#x<)Z)j;pa)s#8i$xA?HV2n;#M?Z3%0t!jP^=5hvVyx^}AQxOZ4d!
zqIU+YXd@f+5&jS&FJX!H4e!0-Q(EMyAE&iGl;ST==<LNoUQ8WBjNq2D(-H2};u1_O
z&7iR70DiSM?U<R11*m$<k9TdGStDo$>zEr1uVtIA&=qGS&5IlW%s^rC7%W1J*(t8>
zBLgZyygIFx1!_0mex53t8~-+1(vd%Sb{>u}SuF|OPa2VnyDdn@%H4~doCf8b7C*YJ
zR~=?rG^*#IXQljQ8MWEbGTn@hoh@lzF0~#dSv&UO$;<$-{G0N&gcyAY95vcj;(Iu%
z81-W8rM{$TPSPnRJMgWq0(QX4naMc?q4$^^YuXx5)ytk-U?epy8dC7T`q#sbpI3mI
zmH|lr&blj`%)GSrm@SIpU6)IooiBs@aosP&tw}#>(8u^2d5wcHS?xc7fh7}wm~PUi
z-zz_N_l4#K3Vf?%`aX-6mGS!)aSU+#Bmhzu^2E;g`FO3T*ZX<)umcuk@PnWOJQkpa
zynEDhQVWBXHX^%o?>5;7X1WXPAJgEukPFR8tc|UX!zV9~qbf&vAbHU`$Hj8hoNR0#
z97pdw1TBGjk3_k$!n4D%q25NVBvJy4yQh_(I2u~ZNFW|kyC>Rfq=xkrcXl@U_(*KE
ztZs$xe5Xr|`nTc1A0JW^-JY5o=vAa~jf@U7o}Sgx#8uyhrdoM>Bc)%_5*!6$mU34q
zqk1fwPQ=z4L{kDd3&u+4Ba(niqoCN9(o`Qi#T4NYgo3)^pIgtx$?VQUTP@=`uZO&I
zZiuqI*xK2;`k=lJNHrEA7Jqc5quZq{!l(OT`DOD0ZQ<53vi}*MOmGg_suTL<psk5O
znb{P9$)jy1r6BCFi^_nMeFXF2Hkx15;mkAuEcx2fKY+M&<!G@-0MrW-3~;~^6i(`S
zv5-k6mTSs`4hznMX>-Of2JJb*1mCQpfd>)Uka3V*<|6)Y1FPa*oI)XuBV&lav_;Ex
zb9TtCkk~7Kw+e?G3NDTImECfzl(sSUlT?*IH8q-kz*6*#cGwHqgwLuH(PetSV?Scz
zVjis-4;W|xLH@u~^b6L-d-~y&?-T-yxKONBk!ZJkZ-{#I4@$;>g9TvNR5W#N^Z@lk
zG+c3|c~R(H_nkl?F5^bY&Wo-Yzq4rmC<-kM-iJbx%P03RlMnV+%oEO%YV&T0FEPqo
zi2$<)P2R-zM5&N8mAl#@Z`I#`Q!TSEDtRrRvor-CbL22u7jH#V`bJS(ot98mPtsJ`
z&)qe+A*e4KesK%^8v%g-D<kUbTH8n<=#Ne5z+pXCK9RRWyb4`RIVuZxEaoPV|51lv
zo&JzNB*K-W)ll<E6^fc?LI{wt!|Bw%%vmtOJdd(XjHwWM&Yuu78=2Be-yUb$BfvPy
ziEj+_NTKLZ`)o)Q_{r*<(8LC^@Jg{t5<uGp@reoo4=5H0GW0hskr23pNHioyHKDYB
zw|ywiKE>$C@ViaXRNM8yOMT_C91E~}b73r3EKd7jc7sAcwB9(Etww(8)akyhZ@Ly<
zdzSNIUgO6uV9!ZoFMmR`f51{TV<1t2W$o}HY9S1Vj<1h-_peE#?s+q3S2AeN_4Ts`
z^vk$ZD$6|Qli&)$f+Q~URd80PaTa*cZcm=W)mSo)=O)u>Ls*AZxL#SGF_?UFf46>u
z7%*1KpJ|4_9<k3hfVF#TkCB<iqxKn(lfw*HS!p*xEAF^D<Kpxbr`j#ZH_jsu8TZ(-
ztkH2sCyWKuE3)C=DngZ-sYhET*+>uE`727@zU3j(Jofvckfmhsf_)vsO5GYMzXq>}
z$Qo5I4`?#llCiWXZc#Kp0~{mi=G{-21F<{_<}>VzQCq|b$Z0s{jX|C)&ugMd@=0dp
z>_BFuZ6NFuj?MFw`C!O(qpST3c@_qNOC-tmt%|a}tgBe}R8lh-G~BoXr~tvt?F_?J
zi3^n-^x;|CH&{yp&-Jqap!DYrN9@&E!>u!F#Dp3p*%@sTI4Jw{U)2->p?AW5Wm=aJ
zt|`6nZ~fnf2Y;FaAnISH>nI!?l+cFfXCg$4SM2$bYF~v7b=gp~QxXQygbiSzYP=;8
z!n7#plbrDQh{tX8<FPq3yN2+(dPB*>#(kqIu(pK#`O+JDN*e?S-0IHoohA@srV!;8
z9w=t#Qs^`gw$-#E(kM%@zTm6zMV<#Z{=`~q{3g_BBqUE8+<~jdTnSWg1<lp^=yby_
zLPzp%)Ad}JBy<$GPwWce>KSrE2%rnlcA~Xt?~#$BPM3sx`odaN)MUK&<`;%45CX`%
z@9fsE1t&~-EgD-I`i?I!ju=YcDYgi4fV_oa%d6eDUt2yevtcv3mCaf622=COPD`FC
z8A)0KY2O)>aR%!}E`^ebps2~R*Y&9;^7K24UGsxXka0?7*cSzT6iwr_T4XLrEdppd
z=;JdjuVow?>5n3ssvr+`YvgBAp-cU?p1wLRKH6{ImxQn1f!GC-k6gE(3~o4fo)XSf
z|Gh=*y-=!q$$8|yd9QovcJtw<)Ol&#f`7l0Z2z#0r(NN73mTfep(IG1VX-H?nnoCM
zq`3DbBd5qEo1q8rxqgP0{aWmr=PFwFb4#9&F+Xm2wZYQC<;hC0`pPtEbaGoJiYZ+O
z623_@&rphZll3!Dg!wb0(@E!Oh!L*zZ9b%ln$B6GPoB{h=iWn>U_~8Ya-_O%X0a#m
zMD9g~j9h>}C7VJw#ZWG`L|KkCnk~skG?qyst}_y?N+F)6c26+j>JedB$kHF&GZmNt
zyId;lG#Oc2_ArFExN#a|Y?%s4$2>pfx_X%iT$`YpQ(m7<K0B{^V69M_Ta+=2=Ec#F
zJ&w?1w8@xR{>~Bvo_mDjGPX&p=G(B@`-_?h2E%DaxoJg^HcGv4-!);BrcrIkuR0gV
zI?sa6Q{&CSNSU%H9TO*$3Uad^v0QZbo$iEEjpH0(QJt%(R=lLj#MFuFQdk|qFM8@?
zG&!Mtm)|a4E^lrP-MCO%-)uRaT<Kw-$DqcAsA<%A{-b6oXzl{E!peItMYJR)z1wo~
z-op({MSdzoUuNq$)%8(y&QOzzt}Po|cjj{KprSEYv89ORMQ4w`-B;4o3$m&gaPcFY
zP?4-Z`7qPQ=H~0Prupk5f?9cys*%AEo>%fv=zjCG8HZu%tw?YveMb7xw7RtdD?$~B
z9dK0gS-2ANlyJv3n{}W1J)*t;=Uq0MdLond<UooEZI_Urqw=oo52?L}zJKtoD^^bR
zNiI@mHs|UyQLenHso46{{l-NaotKoU8yLWA!Z-v&=9Ik92e5s*!q_~b^brmo0AizX
zvSln?S?Ke8__c?omAn%H`B&M*!qt!fv(R+eq5csjgG28Ewc@h;Q{H(avLJd_v>pVN
z*0c=H7t5a}5oD2VE}IZicwTPz5dC`|J$N8EFYS5g&1A1J%NtiU4^4077@LEn0&kK}
z2!OJdw(#<~6E)wz>|TJOe1fxTlmcOD;Or&sx_$FA|9FYMIBDfD{FwQ17?$z?EU_Ye
z;l2bwk*XoWIW4O-?gz_8T^J+KkT8QN10HCxYA<|MJ|W~w{ZzvdHY5=;A@9vL>t*Ul
z(|ys9G2>`ROf)6WIisnG6p@XIw`x$0SQT^;xKe8{vsHVZ!d7%n_n31o5u7J}-mHk0
z>h7qh{vIt~XFtx61aV5`r_s<RD=lJOZ|bD)?}YbZ`yxZzMck?%nI0j6%6=5ewdre<
zngeQ6d`OQPebpPK^=|+kMJ)R~WJ0jCsQual1<Le)qkt5YeWSXDCa2@?_vv$}64Hsc
z{<T5u@QRB6YKCO_-t?J%mgskz8`(Ao3ub4qNa()K#u&%ds_1qDH2VYVHp5f&KO{s<
zEdRqWIWyCLmk=#$j>qD1Aa!5Ryn_DGFm?ci5E^M1^|zBd#&|{ItaG{rb8sBHDQ#<9
zrz_V=9gPKUUnZw-UZrsa`xpZAYV>+*^SaBA*SKa?A1B~v&L;%>LJSNSV}b390eOBM
z^6f?y!g=p1LWC(m^P%-|dt3_Z1`#Lq7X$hNv0u{)?qn5f|4cTZod%S!9fEde4HAGu
zVVmxx&F+02Z2mK0jn&4E%3i6|p0_*&h0NS$?Y7tuU(DsYHvw5Dqh5&JSaf=}DndF5
z-%I;!%CMb~Xxd^>FGuunnIZ*26qx<s0$~|m;T>#CGx+j_%Y2mH<?G>Jx?{d8oqQ+Z
z*K@%`UoS8(v`1M@zBtS|ooY$g(-HB#)Hxuoxr-N%_>AUOjbfE7RB2zDk)+^U7^SLS
z+c7=?Je_bH<{_khD!~p4cFv>DIeWtB5?3B}^XuGu8j)cRNozDO8r)~?+D+>Qh<*2*
z)+-|C<zg4Nt8&%8DxQv?YMFy2AErty=FFG}%ZRcEsGSvxJnzyP)tQIS5OMd&HR*ei
z3dz`Ix@1pyjzfqV>9;9f16bMD&%Cq2J?x{75d?K$E#2<~w!KEcsl*+kmcdo>Xy;ty
zu;HuH25O~=lCVp+I2~W#9a8Pa^Zu!i=RJ#;kOU|@_2i(c{I11IJ&Z9VW!vcbM`ftp
zAW>I_@At>MQp8^1Qs|#x>_B1-=q@AqHx&oO3DA%19r!dcscC;NYJ{;l2gXc@<cO5p
zfUDn{2$5CWWqF{>PR^7j92&rCGob+{L9xp2P&Dk8+c!?~RocsRB#S9Aw-~D9PJ=Yl
z42=U;ua7q-l+*7)S9}dyofY!6r-ZyQ=%$i6eaKaTy2T-6@ELHxQCwZ*L3#Plfh`4p
zIfEiaoc_Y(+qtw-TNgNkV4_s3s8SY9YzXid_zSKSB_5_F?~kd&w4-vllN&cvEVtYc
z=H7S7vtG!)$Xr|sHwfCS>Eg=2357J1P;((?d$?ok1U`J{Kk8H#G120JKnbe2((mm+
zuGdl4-^RMoQfR!$uDMmBi|oo?u)*L=s+M9!Vq9e!%Kf^c6eie03ql4ut%^s+f!-Mm
zG6L2#)M=S`2=h=s!%ikAB5|~g6JWaX%JW+-R|g-9L>aNpQk}M^1t~QG4)M6^(8P1d
zmeBX4LIpzg(FpC5qsQCw<VB45$2Vkm<T4+4EH>9TFUcd3rnV2`HO?XDAaKL?Z;WHq
zhw)6!?r5J)RN{b?mHYaYg)sSKwu7FJE(<#k4dT-L07MZ<(rTc_A$(?GtEfWL14(s5
zkEU-=<8TP@&UwkXvoP}>+45%5h7b9u=O`!D`88jZBtZ{UJF~B)83;*vdo{n|T)N`0
z9&O{ZI*<2o(_wjwLl%kA=JE|N&bn+;Xi2xONtBn46nn$Fmvzk?=a7zUaVep5tw~a#
z8mO4v&r8awNxt*c0&%>Z-LFi_ajegJr%I(i<;qIH05mJk5=}anO8&w;n9$APPF@Dv
zoN^jtTtCa`SPRir;F@K-b81mT!z&Ua<iZ8AvMg<^F=8F720oJ>H7~<Q-Y(lvK~BUd
zN?Yky87MNUT9O=Hqgc(YwVg|VLFE5EO4-J~AXbWOVIcqYx-yc$WeV4l;Tj$_^kcSJ
z#>Sz!U1XzN)}GwfU_)EjX><YA?#-L8Jx}4PMNfQA@i&w`!C3Xaazo^dmjgttB_c8x
z$1+AaicZWjcg}ET47j^+8QUXmgkFK<ewev)YGF~MeL7M@p?QUYG*am;<sNMnMJ(53
zM7I$qcPB^&dC_P`dA>`u3je{1uT6f6mgYD1IQe-(1{mH<WWhfovy@sP*2~qSz2ja%
zCp{L!eTd@Xxvh3iW`+|_2%5#Cb*lI2yH#ThqSUtWRkgRe@R1)ojkdMLWvw9!13*>&
zE*~5nsds2lH;CGG#SBmpfP!PpLJ_?!0dJLK0wdH7BO{AZb)~<=2rk9-YvgwNB>)1R
zGdU>?OWph?mDr~JI+8fD=gnNB-Ae7D1~YTcp6g)0S@|w(L^A%+R84v!UTsFXnJyQ1
z)3GEwxVSDuaoJgHN6)21$IYTR<S<h4CRuOF?T=*pApGA`Q*y7gb~BTT6I9Hp?HrGE
z&rFc71nr`!E$S^&Js!<-shRzGMLNrqxKysLj(u}U&G8U?k_OA7o2OVOJ)|eHvMmL*
z$Z<KkO#>piX7^!_<rXI*-V5M!oOaeFDYvQR*(bNdMkSJljkF3!s)09G%CcDQI>Ei(
z9?VO(_dB|f;-}QiAjZcCNufvyN(!JoJ=dEVY%x#j*<8scw8I{i6JcI-?&*gsMF^0z
z<a0>++n&cbH8&qRDvcV}XznGiPu$C0sw$`~hReF>Oj9$o$rZ6-Vflvs4Xw7gTnCUg
zLGy4)A|Y&yDKH&8tqDcukTFBhu8AunA+^a=PUyq<4p*TRP^ZZF{ED#gWcL0nB)@Wz
z-|R0dt*ct?wQ$q8!$_}li>kWDd^%Iuo{G@SQoYJwB-#_Fs?!Jx^wehhryrDIntI|o
zC2;5hr<)dLS=6ChW)*eFnlq>7<QQCkW#c033h)X!L$KUonq@1>_dv)aY!pZ9_7MNT
z-*=@K^;0=!j1U}xP`6A1s+>UuF&kOcNBbvbH9WrvkptQ5t>_eAgPPq=YMDVa&ly&z
z59!y}pYty+3ZTnzv55?z_hu}C2X}-k*v3Slan2Wjt5<j6GdL{~9{Yyq0wP3XDZ}7x
z!==Z%f@i-5AXfVmjV;dFI6R|@?Lwr23j^h;P${#AOq3n>ChTH=D|)pf0c9b>6hp+5
z7VeZoSxtA7{zA%3M0(6nNmG?@`=v(H%;7#Essh*PN2o$6O^&JVfAM|sRmvK0h1Kv$
zJ|K1^^R`}-Rm|v;DuluzdiGvn9AA0O=zG*{O(k<%4WpFS@(s%4lJ-B!o_{6Z|Cl*u
z0{nN=A;X&Ej(8k!|76cQ<d8jZn{s)HrD>Hzik*YA(^2wJ`%u_x;tpaP`}lToy|;OD
zNti6iF6b=~0^v8o>+Rj%{r&BvY`=cGsKMdwy$jMKcxi9oqHs(Qa-aoyXqw$%+OL`7
zk_#bB`x-Qz(C=2~(ZxcPR52(V;!|8=xj2X+CA>oG4r%_>?e-DCVGQ5JBB_0UvEN5_
z&E(ljnmi&8bmq-01L6-K8gh<F(^7u5VP@1`oE@T`@5x}^<56h<%fZyG-oNd<cuK6+
z(C;6!*Q>o%wqf*DxHkNze@$cR$U#Z0jzURHD&QipIeZ&$2zHohGEEvuu>8q?d4?N`
z-}~c(c;Y~$2GwsSW_t$k>@RBH<`o`DRazRX(O}mvn6jG|5PM_e>2~zMVdW0vCeykK
zTh*K$TUoaz@s^M9M_`<XERuG^nFzr`Q+0obC;=!-^5CGTToh&012<sS_>rjBDK9~=
z<#Fywo|=6$OJ6&Tgy_EC?W)=?ywFFybaI}aXmVbyD!O=*v8U0v{M)cIzQRSRK@cd=
zVg@3LiQ*waGVXsv@vYh67bNeTWCrM10{5d6;kLTx4vlEvXQ}hciNuTdTJG%~@FZiZ
z_+h2SvdJ~bN{5173^$H!=>=o?gmEx?5N`+?#c2AM@Wjg6Px!~^(Po>61W{|6cr8;`
zLsL!bIn?#}7M#YjnZihXrmZO@s~B;Mnu^33AIvGC+>y>d@$P`<jEaJ@=+1+U$$}14
zU(2p}WMLfmW-*0yJu+KhDdF-bf$QH~0ZzvXF%$;VW7;(bse|ir$MP1+E6DeF9@)DN
z<P7ruGJ(*&Zwb_KRpLkX+oI`S_9XdC=ANB$Pydi8LG{Fr(iY`eOjAU-lOOF20edE}
z?GB{_wr_VQBLm2!4;}u@Q^@b0(E<~n-Id4AjD=Qhvm`F)#ivH?>8ICf7=$<%=-Sf5
zU?G0VL-8)hPMEQ^rjgPs1?K&5yq0Q_g47n7$Y*q{n1=&yfa8lMulN_!!*iX0`qM=X
zw+aOBT`$P3tQ2;amv4-@-=zDR9|ksI5^OuBFm;(00XO=QuH-d}`crZVgS*@VKurWr
zqlhZqx|AKgvcSJOxnldx!a}t)5_C;Wr%tbfZ(0A)5&iQEMY-}h2F^3khg^rkHT%TP
zr7q!2;R|^|mFC5oj(k;QhAnx{yk(cK$Ng)VI-MDe?4B6HFU<VT#uq1267B3&bs#<h
zQtBN@CY~x9cXN`-um<guN}oY~Sl|$=Stzn|&OyKaGs&sM+s)p97Bi`|wNa2l%kIv|
z2%BQo&Hj$T3gx0;q)cs1F12KZQC$0AlYOP>@GM2$4@ytLiF9Zrw}Mv?8_X1WoQm%H
z_l5+Mg}U3IVVD_KdLbB_S-itI*7*Kf`Cc5u02d=YFWE)3!+1mGon0z4gfQ8SXk3?<
zDzU>B(<Z?Zz5VI0;5D;b5V68!_w|R#VRI@k#O>MNBS}G5fxmxsX~a#0RD>{dZO*Y-
z*u}9Q?<-6&aeA!W?zUvVi&&8^yyJ$}Yn1nb={?AFNAKti4TB26r>z{zK}hm%30BH5
z-|t4!n~Gg=ML-&bdc#CBQ`pZZjZ3ynxtBu@8;;ngTT<?eN&nDFm!)AT;rh=U*_4dT
zHT=3+Lc&Gvf_Y9`bAg3mWXeu%N(dLqrGtEEQq5VosN3Zx&}R=9&o=Uay~hLmyAbkA
z-8imrl3L(H{1B4PwrcR}DeP=EwLO)&R3$Iqxu(HAEdmPY4{hUmRCW<Q(iF{wj1iWU
zBfaY0Hb<|OHq$2piI{EXI7&V-^TCu3y@gmxA8)^E_pCUb%}uLumHKG_qNdhBR{+=n
zqUL;W3d@BXU@9f~AYj4~{{EM&UwNC+g&|jFXokf639g4mTc;nQAxsiI)YS3#^kK3@
zF(tpvC@5a*(o2d(#a4}RE7Id71={?uzS8GhmH9*q(zVuz3132p%(BDu3C(PBgZ*z_
z@tV9Z;BqogmE~Yf&xbtXsae6wp~Z3!>TPRKh5GX0AG}854pafgHG0DLGkAh_hw~r?
zJ;&pb0DsVGPK|=#CoNOotw<+?Zk6BC_ku>U({iI+rn&s7=U`kA1T?$(i_W~215{S6
zx_2TeTzw7p*St?nk7jLhwDuPkPYqk}y2pQE(YGD1OmjDcqu(C2e+6JiICT!Osf;!u
zoNei?u*pgU)s&Cg0({sL*;@d-k?k8zS42BOH^Azq#TRZ)ptxFK(G65ihRhd2DHK%(
zj7~TtzdlHGMCn`J`SjDB=VoVSaqR#&)$gq!T!lO*9WLs*$C@e~FT8K@$z;j(-CcHn
z`O6ama;I>*L?gcO$khEAae4W@WJ~>?=woF!WV)DM_zs6gi?q1E5@SlwnRO4p>f{gP
z6Azm&=ahZ%ygUR=7jq&cITC2cHeNg|4ViZWP#~z#V|xEuco8vf;&RD%FkF-_>$H;O
ziB>0Ur@VIh{85RiS`9i+hRM~FuLOTFDZjR8J}yxAp{d58%xHd&ZR*H!2vT(`belid
zF(zI(*ZP>S<nsoP5gUc;h?t;Z)zqSR8j|QVf26@G6zF~UvUEF)j%3F#KfY*Xdg7j&
zxgEY*rXW%QvtGm#x%sM}0C}fchCcn}3;W7zc3hHW5B;IZQeaZMVz_<3hL^=&KIwy1
z8ORyXr#paNCS@!pXWdl0>$TMN$v*gz$wzNT&G|BU5x(EaeO>-DO|w74t3psAG*hKJ
z^II{A?k9cvBpAu9!MOHi07Zs82F>Uz;(!YHxMV{P@(#TEB=9QQ?hZ*<KKR;9+YuR5
z2A4@=Cw44I%SFu9CaWmyQ|P^@UA}5x<m$B9*K+Z7-T6xCTSxYhc>Q(Gr!LWm+ASK4
zl6tKqSGV@s%-8>$e>^)v{(ma$0RD5Uiiz#NcA*_l++_d13OizRur1VS6=DseDm}v)
zSu~|&Gvs_jX`jA0L_jD+qLQQ*iJ+p;(zS@Sg1<dqo#qBhxAsh47Fq6glR0+n-QU>#
zoz36C=wepj5Ck*=_l`bppSoAfkGRH$HBrZp4m;C2*Lz8CnL&)fBoO<)bBx^iV~O)_
zutq1c4=6$Nj#{^Rjo@((E=#@fUwV7Ix6iE2+x@)Pzq)BCND>UKATJ~};0Nt?-1PjG
z2X7YMJmx0AIX;s@eF^)lz;%1R**+USJSOd&HNgjiT*_FLp0Zwc4{f!3h9EHr)vrh5
zY%ayK&~a`QX!YB9uyK6DpC(rr@`hj#tyGaXO|(y|cT3C2Ev*WjLPjbABms5Z5Joqp
z{SXdttE@%g3O)AGrjZpkO*Lklp5!1kW6x|Ozl^I+z}9I9KlNi&O5BA7_lH?{Zo{E{
z2@WLUKc0f)F1M(_P{4}2uO9JU1RMg24#xt4A)8ia5Il;%bo)eLA~jMW8=0+-o8nQR
z0sSTsw^*r5*S4_Ym3IWd8l#w!nb@`ctJ$*P8M+y}wAJ<H%L^9PE`OlKn=kNwWf8sq
zA~v1P#+L?5*lm+Hp0W)zoREeKpT8Nkb|~xJ?-CBRFY*_@j5y?-)lEcc%oNNMduLw2
zV}bBa$UHuG$WWt+tkfv+O`Xrhd`P+$ex^X<<x;Xg&tNqwK}mC=Ae-^}CC^Y_{yHVE
zGE(FuGxvV3zef=~_sbmmT?spNjGBhX8oX$HGo;iV6Q+QI4R)8aUVfF$cqU#=WKw-;
zvFtQ7imo2%K7@wvxj)Rss>)vek>6>|VQP`mFE-SAaD6Hz!L+_gEc<9X<dCen3W?(d
za4uQ>3PWScIql@+y3loEJ**Vf>RNM%Tu^1BXQcF)qQF`6s8MBtYD7vmyjgm!ChSFy
zeVBdeWnI+0)(;k0MD1@KHa2j35VxmAT^zI4!9B+;p7Bb&qo`n^DP;h!DNtL@R#Lzg
z-$ANHFD$2VKz%Me17qCOT-jNTZ^27X?)1A4c094833<3{Tp90*mT>&D#H6Jh7?#?6
zRXTVj*P<!KBpP-$x8(CoctmN&VB+y%*|orwT)DaKVlX&-h8-+g<5GB0seI-I5%udr
zE!`~8kwRNP_g0VcR**%?MswlV7+(VGZ1p`vo}R%|bKSWL6tY{v@hNd?G^<Y-2i>9r
ziKocq#|+~)ck98B5Jj;*Y#+&wC`fW?5jd~zjcEKYp%yg(S!XK7DDy-c=?#ehL*<A|
z4^@C96=fAg@o+x+<7*~A3ms3={9gsh0reFF=XIi4%^vHyn22wMT9r*Kk0}XSa^e$1
zmMjxpy*Wc{0<Uz{e^I+U%TVJKJK7PAI^wN^yiHY!Ap?{u?iOhOhHNL`l{>;Ly!YkO
zVQ6R5CH3`VjwkhXHp>lh|Hj9Y_s%WxTI4e=pcKiE<9)&G@}Fy`&`Fwb>{0#dM#zud
zZ=kh8YrDG`Sb<WPN+9iLJe4y({;&)qI(WK~A9?T~ogx`ek$xmk#BIJm`b~*93XGZ&
zCop0s8cK{CrKf@RFj5CiqTSs^OX^LD94l_aKHLX_Y{j0jx+rIa%1ZuKJix-WN9b(e
za8!v``oGRbGEwrTurEnqe%&6aYqZ`Tr)HvDjm9%WWlR}mj2D}tSL~vlb_8|R!$FIr
zAa>8mCLi-OvB;?w%tQ?ED&`ar6r4qbPa<0a$a0=v*?yBtxzX66MF#@eX>W5?|63Y6
zDs!!D8kPvmH66V%ZiOZ6ve_e@g@K;M=xJ_jax(iFg5|0xYkU|SSquw)f@<Z~?H#N)
z;Y-f#b>wWuN-)L_dGs7@Z;<3PZ{*ATEoQ*c!OxsK`iCG})h_M<^eVX1v&`$7iYE>S
zaYb60Uc@z5YyW5@ji%Iai`3cGnwJ|Ucjbj|cEg)a@E*w&^kp8`h_TUd;)I;xSPFs(
z><FNp-tYPdk8NC&P*3OJrtwkhG^tffG4k0F^Vz}F8wJ%J;HPzgkz=hqubug1`sku@
z(!6p%L2;+BuDjsUZuh#2erIvK)6>(b#z#8Iu24ofmH_fE7$-pjr)Bl6|4@B{`!fll
z{0{+$nGx`xDK<MZ=YNH%+0v4Z`)7mRbD{P~h?F+W5fBicu)-o>%YKY;SjT91qR-&U
z5%&3|d?)MrQpUD0w$4WM5-y8zF&$?dA+Ps~kM|~)p))VHD3jI+rz%~yN5-Fl^BWNl
zHzt=Bhy|r2?PihxAAb^pU;LADrg$iBxEG)A$$+8X0I`q2-wp`Dt8$n4If0%R{Ark}
zR6>|%&&_F#20?r7RQwY!Vqh<i!Efho=rk;CYb+fl6PmbWcE_z(e(lZ!rrRxB^5*ez
zCnD1U$(4O%Ux^xs|0_Vp_x<W3t=H!p91GzUF=jW(P$<Z4@#R&;e7NaWPG-k3$Y6$x
z5tT6xkH*>o4i_)EzWL~(NojyNexrGMr)-_&2<?6=I<1ck#ci6_M$0s$EcKY@thI&d
zLDdYp0bTVJS3%O~j3#!r@RY`9<EEV^_Wh;lgZYXNquVp$=hlLjL)sF&q)*CRB`YhK
zbt?bKMKz*T`=a?+b!j6F+bmgR?{^H|Z7s7NGDj;b$q7YK3e{oBz~8x^KY@Z9Wyz-n
znHF)*$zm|EcxZ{Hg{~JndA+3pdJbz&57EKTL}W@EGG0qemZ1_M@yoVM#!5F!0MROM
z9*G!v(a7PkMSs7%g$E`Pja)aX4bRN6zR~hM2B{7^k*_ulesf7++@_@VOe;~z>9Kna
zI+_gHsC?REAgxD7F8`wBLMCzxw>8{@Gu#pem#OMk4hl+>yXs<PF)3I(Ay_|&io@50
z^4h4KK~$mH__4W&>_YdVHb+Hin2eAW`B_RRiV$$Y;71vC4j3USz%5vUmrY^8`0o1L
z5(P0paJNRf)~aHVc*+n1+f<ovy3CFu3k{;NARBvJZIoOiP^QCD(^$Mk@?g})(PR8k
zrXt3($gQa@!0`}eDDZw*P%cJzYf~_?)JR)`;d+;77ap~(0)d4cp}}q8h^UdNDe!PF
zruX;N=ULs$2)t*FGssK)2^^X=+|H2bUVgMxHpC<kd1jnWsP)JKF5Uf_p{SNpC``7i
zk<?G$Bvo<xOoH7F@MH2bJpqw@Gx|aFO*{$MsHj+(!#{&H<N<ezEA5)eM3Lkz+{|c#
z*+C9%3%|VTW+~txAnS5`XqNsvx|+U)zQ@tt5{Aw=(q%Gu`m83`^E&_gvV)o<C6KS&
z!nwa#B8oX5SS4ttV%dw(r_mV;DN;g}c>auKELe7O>y)NP(ags~b>v8;Mbx6+$3Gw+
zI+LX^pnaMpNU88>1t+DCu6<P)MmPFGm`P5r;b4i50|4EbG$9%sCW{k?3u0J3{zbD4
zj+pSx0nWfj6!!)fWa!yECm49>yXjAjFit(40X_bGoK12&ED!bioPOPzxbB_QrJBWz
z`Y9vtstY)UhFPXx&P5x|!C<W@J{q%Mv`sb$lR4TF6IS!7tSd?jgxg3aiO|7og@^JN
z+leGL$@?&NcWs~$yaxuuT0FK;S8E=XjsJ_WcM1+ISi5#(+qP{dE4FRhwr$(CZQIU@
zZ97@fpZ_~`>QtR?U+kM%T{Wxkdd_~k-!aBBtp=^Qsd5E<nIX`&J=$xJgig$^v^!F@
zm4%~|SQG_aO*>&0(pS(zdIEMhBI+XX;LOVIrZ_jfeBQ-^$^eE+wi5aRat`wcp|DJa
zz`{|Rtc4B3luT~2ss40rZv`bDIff}ZDv%54BZBt&fzMZqY8;S9emY%6ZF(wh?VS@~
z$&qa-GcU;0*_idoH!5$=h#SLpN|T_G$wlWw0`}7x+sRaH;2M#=cCRA3<f`5^JI_K?
zg92=;4_FbKVi%wS7G;j0rJ&80=n+d9EHR)L<jEO$2Rq4OOWLD7xW{*xmSMer2vd6X
z^&2o<+CyJ3655$9x;<c46i;_;ef~pC5IKY!j*$I9jvW!FV5Pi|TL2kGnAmK?3pNYl
zN#@wr+>R3U5diy{w8Tt*dU~gAh-@u5X6QP+$M}OiI-d69xY-(&nvzSdBH}4jUBY6l
zn2mi{l6d5GNODNptbTF+qb?!^C1=vh3yqo<&#y*LpVdu7ie3GfiE2mhi|e^(G3T+j
z!+93n0S;1~t0iMMrwC&_@0y4#ht-Nbu*u$5nkcs`A+0@cxs|TNty^y%2zzEcfDbsb
zeZxI58ryB(ba@H06C?}%su9kIXClr8{CeHja{pOcx1_4O1&18s*kfxJ$~1B$RvTW{
z4H#E)fQIdvd0RScU2I;?hF22BgVIA1uhrGdU*N*5Xf9}H?Q0dHlM63=kyL_-Qf?n%
zNr7o6)T&8k$gX2qf>*_{-jop9$_Z|6cnz}+@vKw!gS%bk53+6c60Hl9y>4=pc5QJA
zqgb!24xVu#1BeHFMrVegh5{h!r;jl+PC`aknDw@F!4u`(PdnKhR=sCdb|kpBA2?@M
zMqzgCPF*lXGwde@TA9=_Qh#z!)K_Q5poaI(X&MW}O6av??|>fq<OV@<NXFXK#(bGQ
zv*;QWSbwHsbW2UU3?x|&Rf$dHOTBolY8>X4Ibyx+(8FtOB`v!vtBr{<t<uzRRw>KZ
zj{+w)kg|{#_7gN7q*u@nU5zDkGq0;5g|e3QSJG;;#1?^uPf{Wq70w<Q+iLUi4((Y;
z_o5xGp;f8&2C(veD|&;13yqtaN<&DLPaT_T!|{%liZ!HC2@cA`O$7m|IRD>~!t~`f
z0Bg5fPqvxvKm`&m;c)CXn-DCJ;(?iR7E*C&t+2)~uB(DG>esDkD2HGqs>6EE(Oylr
zQp$w&1hwzCe-uHoq1=ZC$#iJhn=}&7e6C9V2F7N!>T+l7>$D`C+2+NLn}viTtVwXG
z?mBfJe@FXT+BGuYx9x+Ql0z)J=}ihU6;fk~-kRH#iEVe02{59iZXN1U5ds~Q$<HqX
zttTM-aUc2@=M91vHv)pW+G(Cl0!t|$@6v)Gq0+o0(i_1+rfU3T)AMc0A-f@^`vFcY
znWe>h9xotdjw?1{kS~&@05wCpAjr&XY6Bs2WxLu|JB{X<IkQ*X%1kW`mX!-|Dm5QR
zl}`kVk`ZLiTNE+ycPpC#Dznu+nqZC(GygI#xzi+X0%!j%<7@`p0YKENs4PouDv5E{
zE|sFsCAB_<?H3y&LZ5m6F3lh!m<PgJWq(Z}7k31+zbE_>9m?U%lUgm2RKKf9TiW-2
zlS!5Bt01Zz?y}p-c<QQJ4%?`Zd&5l&za_yrkZaWyArT}1wj33w>7z4yu4;#O^Vj|7
z;GF$UOOtnuUO>B~8q2K6*`Zwe$#H@pjgCXm%BD`Ae%>%J&-O}XTMUG@L>sbt<q#|b
zs0OuPP@;q*PmBL6#E;qP01kE-1AEU|+}@{Lh<fYmi;t2`evn|(@ltcPtUR|lY%=LW
z_^Z54Xsz%|s_@_u|E}-eHaFB{k6jWK=gk+{Ai^%?f4><Y@#l$jC3~G8{7Lx;2tD=G
zat~+jkG=&#zC_gk?~`c@kknF2i8YM^W^HpCct7z@&-Cj8!qan!0s|#wGnnZ#ujZd=
zrt0y%Jchu2caoC3ut7;WxcYXsVWbp=56pbID7#vj-GTax??gTSci(SD<wWv7RHO`S
z|D9i$<NxyubFlnZ<?4!-j-7TJqR(sHp3y&fy3-&WSws-~L@wA&w9u47{7^g`vjh$o
z5+Pmmo4#92T?bOJE*ly>dU@>B4zEA%H}Os<I1iIjNy>?7ClQiDLRJTp3+594-bNKk
z<3#0bIFBYq3nHZ|#DeTS;;{EY3{$|(C5T8PZdnvqV^n2~OGK+kFANtNg>6x4a?>Z<
z2az8+HY5!#7JQ(MNoTEKX%(IpU}&K!VvyVV1p}#JbP7u%ys+cMBqZZL07LKGajGPW
zN5AZW7x2ob6QjFOE+wjw{xOIVc;=4*TVjfFla?Sx%}RKH6iF%Ky19C*jVX+|+2FXy
zY@`*jR?v%SRhEo7aTrdcymgmRA+8<dRvwh%0V`?YlmtG~Fc5E4yd0O8sY3W5q%4aQ
zg9%qS#61QpDGfTsLug|BcAQZ!98eOh+d*hy_?ctxhvV3C)6x@D7U-Rru!?39qs98F
zC!9GolUBk}9i&!MtzeyuIX9Bt`=cpWq=qvX%DCbdfVSumVS$=!5v$J1+X6t-T(HIN
zi(Yr;#bO|l5{ene4`*T9G;`TYBljY(3_?4G;ioy8c|3=399P<%c)rs)gYWNFi{jF$
zMq~LH^t#*JC8w#~;vVl3B!}T-q|XK&!g+dA+QB`%tB}+@BFIeBnZcld!vP)0jzk*}
zQG@}7)EQrb#YQn+{KGIK3l81kyhW{e48&{`hp$Kh2!joI__H@hA@NQ(+JnYA@W|}}
z!D4BIk<TOS*TD09JKtY+>sqtM=6wI!_O^HbwV`*dBmRRzb7%Z${PFK$X65nvlC<EU
z>W@vI-~H9&RNJqhVm-uT&pY<A&ferA7Cf#APgh%;j{V{7bKML+7miB6t03_*bVf7}
zw$FB%37Z9*2b&3-3;S`_Wv647d(GpN>vn52XYO^5uikcc$Ga^`ZnwMr&1}Ql*W>x?
z#gfN}bNAEW@%bdGtl8VW(Wlp|vU<JWM3d4@KAde1=4yk&s_HZ=46^Odrnk%c_3_En
z6Qsq+xvDbg#9+8x$*y<kAN*j!<wmi(1I!3WCd-ZhpoA4nZzpgOb`pGvD$sR1j;PNe
z0bvT|f3OgzI7=TG5C#>I!Iq%NgBT>^)WU-EluR-iOc3D8Hk^ujA&xD+mG@`4Ikxh`
z=7H3)MF39rpCShQD-7KJ`}S?)$HGn^2=A7GmC8)#<qT(dDbV)PPoD$aaSa~reP55;
zQB%-+?^8*2hr8r|Zd~hhe{7rH_eXBV*msXl)Bf1`dOki@?fUk8t({ca;r%fFJRt$#
zZvh+FaY9|Q@+Y@$S=;#A)zbBOIoO47N!F&f2JV5bLpi0sMQ0eFT<UwjJyxapeV^2+
z(YtO|c@d-1avG1~SuD|%H&hYvejHpu^uJC+CO1V#&M}G|`oZ)BvjKC5|NT-t=vhKy
zfs~105Lh8N98jB&5NTxRLuAcAI>jr@#LmT5!43g-&gmZ&k8mYIL3BaeDB6on=f}Yn
zYxp*&h)flPl7=Z4TxZOY!-<Gg#1ynck)})z`#VVCz^l?hkPI4hB#Z}rmW3Mx3~SAs
z=12e++F3E9`3U7hbw|j)@*y=|wZD~X{r&Rl$DbvXtkmiL_;CC_fK0&6h)W%<4HNhk
z$!q}O=pLNKj0|!(HC5H5HkWI+L^P2P?3Uje)hymJbGv)f%eBe7X|qIhz!pcqxG+Y%
zFoG{iA18>j=J-`Gc7>wHw=D8RhvfM9cJc7O?)vKd_%<b$W!GlscG?EEhYp3-@Kq8B
zTk-u3#U_2IJJ4DoVnZdDHMvy!&|JauKf@+Dnaxw5f6L82mYR9HZ{Jz#X>+~aFUL*v
zc|Pwhdi#iV*lgH7Y$u$`ra^q#NNb8t?M0>uQu`{<B#hDMk$gBy6742=3WVSZb2vc2
zx?s4+ROF-lK)hGw-V{^f`J;&oV-X!9+`@swk&*)c2Fp1t4PKYjgLokrw9^ghFrWWW
zV%Q#S(GxFhp{^?eHm)xAw6qdIEc5YT$nzHu*;&H+6_(wTdm#X&R1or^K(zxYPbk8V
z6JY#3nO2SeZlOHutql((7&MH4tI+hq^#yf2ql2m>QmJM{@uCJO>coR`n83{jTO1cy
z(|ohvw+dNLz$Hx$!e|my%|<Kl&ba%LRyxn-&Y5Hwbk<)aCp0!QF(yATuVzu4YuUy{
z5P(V|1d@~{NF%ZuKP0d`gR3VvAt3^?AfXV_Ce^+QxDQ7q4=uR9`)c8m9>Xf&uWQ-K
z@8?ifzA~cdR7pPBKrLRu<O($gMS!zIi1URd>IQ1wS3rp5F&Q{3aLnIf47UZfkc`BH
zK(OXSvjEKo?1;pK+Tf|rRiv6kA77k-7QniIS|5NJ1v6nTC?5;X#vs^;xecthm1oxv
z#|r|q=^RboQ{SQD$;NuD;V0l<y#jC=GH_jvQ{+dDS;d5{S1M-?l-aL`LKcS1ZO4Yh
zv5G(4u4s*<Qgx2;(C{V!cf|LNc_k>UB*wliEgwE$3hc+NT!0)ZxTIkRf_&nHkqR!9
zh=qlG0(*H-LIj1t{`JLxvgjTb5%3|&@pA{8^^77=+Y4lj@q6Rs`ZGhPfrlk?Q<H&)
zf^P&G3Z%jrgDVz}w1PILH3&S_859|XFt_qRX&d$p07z@z%1R@%i@<o^robe2C!_L<
z3q@K6F2}5_@q+1d`2#FAj@lzUEc0U+x9K=vq3x*esyA7HS3h-`(Sl)8fwpY%g$n3S
zA+vWKMrifKk+W@h9sI@9mL^%UzOx>;KC@n4Pk%`+<2|;JJH0`e*(O~4Dmd5ve}<V|
zLM(1!=Igls7?h|{ujm^nEj2b9xuzSKp?p-2jibTg%^ZP=RGAR7jVSn_jwx}Z;(tI2
zDl^980^h`w$AnUwCB<Fr|2S(_k2IGHgRePy6$Lssfb{|U4C|~B@G9^A{gttEOU~he
z%C9)~PY0N&|J^Sl*g(}C-Sm9g<yeKb8xe%N?P|cRhUSF9t%lKh?Zx{kYCQUzza8Z7
z8iLXT3}h&*Z<&QUEhC<!QjL=f3Y8xez&e)r7Y}6EYAHyU&f9op4>45muv0XZ=qB9;
zvSCiZlp!iV^1f7*jVnJEq*N-bgFRg?EKzeooA|YH#8nEvOa1O%JYHy%cHHU?G*oEx
zPq48F0>TY%b0b|VRC99GWPwRL5=?+OM?3NYla3RlbR+b&Q2`u^t9cHf^Lc)hx(8sE
z^R-o>IoG0d-zNy>rZ#*E!YNRUA!zo%0A}NXB{-Q;cL?$-l!<0ElkvPx&0j_@BOLjE
zBm0KVGB8juMm~Tfu}om5u^}D6%@UIe@?KXS{ESbX2a-6A@JBH)0*o;}u*p>AXF?j9
z;?R@jK`i^$0vMN90WSLnL{?XS-IiG^?OS2;$jLRzOnK^q!_J}LGvw#{5x80<;A<7}
zMG!pOTYh?9oj-!Vc9fi98eaS6l;tZWzEI5?m8VJhN+A5qelC1a-y1g{V~JiJ@kMT)
zMHAX6V^3XCTPR~UQO0aw9ij{MK4^8QFiI`URL3ajMA-+GEx~)nEmj=f0?VDfTz(-r
zf<(E>xO#PI*DDU)aHB#vK?CseS|;HsV7SGK?VKEbu$t!@hckPWv2_ACYoj9NSm7-y
z;44Hlw+vBBb@zw2*)G>(m|WG%Kj>6@CTcHisCSJnAGyj1b}2U%`Eq^RWfeC|%C8oc
zA1o_QS{Wy;OfxnnnVXZ^lh*V9@5+WJ&q$Z6m%FcYY`v7IVVzDL#asQ5DgQWSFRCSQ
z_Cn#+E#5DpFdVA&H@2c4h0*8<T_>70*I{d;Rpl|dck7<TEfkPLiGU>1jWg;4Bm}ka
ztK1L2Rx;Tv(x11DLro;6;G4A{?&@11EW~!Ep%^gQwf;;M4&CLt7wub4l(K#Tjv1yh
zk1@Gd>YE(4<@Ci5;4+!_Pllr7Lx{Yo_sk(+Gic6-Slv>KJ@)@iOnat=$2k;mlY+Nw
zQnR3?u@|7VRaOO2AOz{}bt?K}O0f+aJfsJS(e3%DSbUag$1}{GCVKD1>e3C5^DqA~
zfWQAHQ((==lwulhy%FpV|BC)0^gkRsn0`r;|9`C+BRj``cM;x$DPc>@epjn@2Yg_!
z2Otn$+yHxY(%(i9*6&Xq<`91ze_JN5>To`?WMpNWzir8;@v#%=V-U!DH7nO8_qxF_
z7nGvgP=-#nAL9R`*k9N*Ej5vodUc-i<%uB3**<HJN!T1ZHFOK$=Cx&osUH@Mi?k!|
zvE@Y!ZxRRRu4X}>-GPE}ftS1OU>A%+mBTjxdRM~v{qpoEOHklkw@f)x5;N^F1W2do
zhUUsMk^(EhH~_0Hpf&Ry!2Y<RhsO^mKsPBsePduToczwvANCG|erM`$1^tP+H#OQ=
zx`gY_(qAOwHAo~^g++i@v#&U+b;qF()B~^$bxxieE*3Bs`{svlw_#6uMC*>tpt7y5
z^)}(z7_v5<W_!z2rx6G2;wINs(WHLa&X~a4ck^vpe;;tX-5(x1(D<XSd!jgmKYx>R
z4-jZ|87NhK&G+?ud|Y<AOB!bMQ(6-UxUM?duRBGok6`ht<_)tYj?S9I6UYSP1RxvG
zdC}2%H|;9w!4J1+Ri(FNc|?X$(mAB*h|6PL4be$TLI_wL6?|jWCoaqilTjoo;wdSA
zXYCpM=uE|1TIE+-V}AkP>SfR+mR#X7`)&*`U;*uR8pJ5Yl+qs;=D#|TlIpyNkW@@u
zAPmn|-?;Ipg^mi|ff1hqwIu{2Whr2tVy1ciR!ZtGf?ytTv-FJ0DeWGTVp*8F`CBx{
zv0;}BpQAyx?atM?-#}-3p51`X@^#cU#9rt4;}emnbH@FxKRLJHz9#!A;wWsALrf*z
z%>)2&j;kgvyKAwF=CwrwD`vZW1~ZUel0yb^r&H5kBME-CPiqxI+gn^CDV$dLCnSOM
zUZ`&@H8>7Qx=kqHX$6mwyBOP5#1@vSSc^yzbc$x}E_jFvbeyT29J5!-9F3V3+|`F=
zCttuXoH_9WCE*>zZiCupn8EI)TzLl9d;u3u@06_;C6gVei4#w!_nb#_n3y|mJKYiX
z>t+JhfK|+<ctu4gd$!hV>=!S=5>9caaLIH@nNcw!3=(}L4~QgMR^FD-h(`swf5p{?
zf;AO2^jHgasLsy5Pr<d2dypi^w;yj!i$Z<M08!k+CzO;(^g1;;2)wE>m&3Fr;Ftb8
zRi;Aa>~)?r&1^#Msa2gGc`zNqE4i3Z_|<wN*NbZ5SGNwgljTr&(3u<_5Km(G@Z2MC
z$p&%kd$XdCXjL*R{7Aju9<quQ`(u2_Mw9-cGtyju*$^LPqY_q0G<DJLJgU%z*d;&Y
zYUky4|7;yVE)lKq`Gz*MiS#rN3NLaRg~*1g1z4LZdqHMhSyILAix>Q-Cc^+5g@m0V
z1N=7w<?_8U)I4*pl$BNTf>$RXq5`f*M2*2PbN*Wh86>0GAyUY`3T3#cHEJN>DC?YK
zhQ?!I)(g72K)`$aU-asHTe;SmX@9c>x@9WCj+vk0Bf!WDtUC_F@dN}ZJ=7bpAG!A0
zu89#Bzp|?-r!AU$j*K>gnw;XGN{zVIoJzY00yblc#|YBbDN&lH<7Nby<&jsKh{bgW
zn9P>{E}+s21l6w9y8q{wxNk{Vvu{+#R=9*YXKH|@>Wy#OX*oiN^LHewSWR6RkqyP`
zU}etgsy_w7arD)c!jZ4$FL^{yAetxG9@r-~y;Tj@xQ~QA>``hrWtPwl1v%-v6cWIo
z7M70B7AV6xOs#+K65L7srsBK?E1wp<v~0slRVbk_qCOO)PrKN6T0J<R@PcfVN#FCd
z+R%`~6^*17j2AQCs#|1*i@?*&;9I8lksp4e<}63|0U9z#>k7=X65hzXpcJePTC>Hg
z>9)3BDUV!BVN2hziw%XoN#)@WR9t(fu)!lctE3goY{?}&N5HwK4fESZG92sEQyE;_
zm0Fd|Kif|SEF&1O_xU!^-(@<Og3EY4O6TjsBDI)GIoN_k>D-EfWn85>onN86P`_>-
zpix>nS(Dr5nl07>FP&9QT1w9?<Ue}=>p6iE$l3g_)x;1cfRPUL+1r%09GwS4cG13e
zvyXz0%Sd{=YAjxwwf)*FAasg2rt`D{ODmv9g4CG}c0M1~wxXoJ{25G<K>|8wSQ=gP
zvpbWeHtmy5NkXVIy2u+JLj?~7WOjqCh4b&;B~psZId)nLCid~7WNXC-!^a9?Le|MU
znpx?N{^xL$%Iybp!(p|3&|j_Emno+{fq92uDmqp+Y0yhWK~`oxEU?^@z=3j^F|zOF
z=g8iy@0~g@$v|t*&`kVZ&L~wVbJ;u=@WoyGzBWV2e`JU~xj<QQM!U^M$?zMFW%Vx>
zMux#jgbyuz-sa>U*%P7#S*2E1&-o@(`B2p1S2~p_D`lA|tte~$8D8Rd#5ze8PV7G4
z_(ojc#Q$k0WoP|gJ1H|G(|>1Ho6+WuBkn+)eN{Vyhlv|u2Nduxb#wM-#?Hcwjpi0Z
zfBW{^JLnoWt<vJ=1qauR(Ilx-pR0>pewgy3`+T@;d&S@N^$>Syn6{`tcsP88-b2Uv
z0gZ)^=~V|XCDHc%ba+2a`3;MHj-?;zNPcU4Ya4R&@k-jW2ib$@w*#=XQypN9NucYC
zVn92kv4FkrpF4Re<*-LI-MtUF4C?*59NX=^pFX39zs_~As5Wf(Jw~FWEG~}^6oi7R
zu;_o~{#ACS{mab@O}Krg97~gOI$$>lp0;Y|nk-9cktHTV<&A}sYZkv0*=WyYb5QeQ
zV%KCYs+>*;5T<{~G{$I$$`jeinB|V;G5JoG_@t<<%YaW3Bjw{j-r#iDm)@U}pr!XK
z>|T@mL1fv}A8p#}4}PG|&6_O0uYuaV4eC8bLK0;Ti-uVqlr5?zyk-<9R#cUF<(j}R
zDX4u<DsNn1N2FnOBw8|4;6u+CVZHe5?Az4J=4=RJakK@W_!GUfqAAdOiYsq5^LhQ(
z1n+r^F34LgwUmT;R+)3w290-9H!3(Pk6PtB5~-30`H#QQPNJ$7R0#(-6YFOAF=pK7
z46j#vOk(zp3<j8>O4RgkA)TYjIRnp6MRp~w408oW7#<NXvZITl2&pabCIu0vAfI67
zH7!NLz~E?1hO~0((F;_XqA}6&7XNvb3V*96IhKrt88LcdXE*p^1)V!^^niS^Md;(;
zYh#VFBxh9AP<9I+l8LIa!dtRG`cx}pPB*drkbO$1e*us`5A*4Cy5o$04a?je3Z@aL
zl;207Mlk*@**`5CvRj&;9#ljx$3;7gstdV&2m_PBbjdjfBcRerP?xT6ozparI%xgu
z5G6osX<G0Z(j0H5qQx;|K_Xt-tSRwowOAsoe$;_m0ch6<>btX(ZGu*Cz%R4qMP4Q^
zqWI%4<fVw=TAW6l(k!^VGP%03L4KvEF;%?kmjK2Z$~LyG8RV4b5{U`R0+d^A<bO9G
zn>-vO%>1b283J%#jiq)H2lU7A(-8#)A>DrCn}>|G=is1fuR1mQF@>dX{134^+&AZ$
zt)d{xd@0`80nFHRm8xP9Uk2gP+FP+OhSjxRF$>71G{ikRrA&a)_&z#EwqTf(>4|-X
zsE@9L)%Ak2R(#xpRXIKBktX!!RcGVMVFOpqd4BBl1n;U4Pn@PXsB3%VY9#YgbOFuh
z_^J*KZ$6M2<lth4%RZ41-G1=;NiZ{WblQnWU^C2sHR`2uCn#X%a^(2qN{30O%UJ<)
z0Sd2hkFj0E->p-%`@}uQtSrf6NE|UzLQUuHczp*=^?v=E8XWfP3N$H=;5n65AxdEu
z#(C>{E5I$u=e(*zA&DI5lV=Km8Az;{RDp(j)lqb~9JMu88j)GnUrO~K^^Uv^$G+=K
zK!1l}Q&NhztmT?Jnk|I5NF003iFknKgaTLwP}gFYPD*0pg0+|Xe~XD^ig>FMQoZCX
zMv6tq<c}ph0hgG@t>(dPqU5Y*OWb<7p^C6ruqx4wJM+B{i5Bb6fYP+Ga)4%n%yeeX
zkx_i&YHk*oNd6c+EM<kU_6QdcApMF1{lXR4%##JrKKHyiHN74sg&b_LJ!*fvGkK|U
z8X|I3iCJBBuZ@3hGaX-_3cX_I#7ou=``W*IB+3p1d>-ekt>FDBf^KVED6P>0+0BX|
zC5*<_KYN*>skNXFtHJCIVURH;=`{SLzFIF?Ov=q8qL&Hzi;~SMTb+cj!xdSRNbfDI
zaVCJsu#hJejU+-Z%SwyN>MLUL36xHR(olJ13>z~Pi|}XCPLEE*0V>{)Qk4!F9j<D9
zq9`80O52mE78^Df32Dh?1Ol|$&IO$$AbWtxnzqE5=8{!;uUi(55E`b3s+utzp2E|R
zK&4Cw)j}S?j?Q&DQY>w?+J+Mk+Pa%k0ioczF5Vhb>b6L-A4G?=?XHuUfwu1#CD?On
zVOq<go=jyWjhp73;w0Q@2QvT4aPn?QrtBeRMBZB$HdCU~Uyp4__&s6L7CVtD7vut}
z=AJgfTLaT%Q+FvbEZkmU0zmut5$kUS;q8sfrjSKPG1zLGSXQ{0O?B7++oo%sAFvR3
zbR(Vmrwhhr4<id&8BI7gopfFgQ2}+G))#5`;pw>0gICS2GQ15fAC$Jv;2z4)9l-a0
zu~+ykd?_yNQ-&i(fkqU8=MEZ(Mx*U|Wr+5?FFjEu8)jA|{98Y@*JR$G5f{PXQRg(s
zEBHd&@GClTYOP5|-x$6&(r5+9breW*V~JkSg~qyGdYvou!No7?!GkYce9aOkkBw7&
zWUpz>x?RK`o#DaTxwHvT{nc3#T7cObt%6+8AE}Dvtt*jWM1iQmb$v#Z^%`_zxJd!>
zs)64+LX>7})hJ)GY`l?kyE5Z=v82JDDUd>nV$yDVHeajpmN~KvG<UfdbnVN#S#V|7
zt3Bms=&w2##ffFjnA0f6nuN8t4c6?yDk{P6QuEG>V4z!dFBTXVXEiFXSG24+4V7ui
z#j{ynPyLy6ci9tF!jp)X^EE`O8K0v_VmXMHcD#x4kUz2l@nBz;tMNf&Kk%i-s(lDo
zT{g9y9MsvQ<!-RP?bzrb?V@nouJ$P@Z>t?sr>5rLJsp*1{#b03_S=Nh<vAKNR%Gzw
zrWh<UY|ma$LKJDVV6!P7zI~e2ke?Q))a|NwnJGDxJn>0Kv-!yK3mi><s%TEp@u;hs
zaZJtFG&&H7PSGr!zgETjd+pp(F&bh7Fn`QanSuoywM{msh5u{0M9v*0AkS!aJSu6m
zZZ);2P93oXZ>hZUt}>kr3>3%Heb#Uk-L3{FzwS=kwA4|SQT!&LIMXd!E~p-0T*B&d
zLqSUzr6LRH_6zR2gxn@ncB80*D8p#G);H}4sBR0_cJp)~P?@ktIv5jqodyxg6dLE6
zCZdg0<Vax~nV+hb#J_b`xKFno3E3K21exX$Z6d0w+uvtoonIyBG3>M*v|-`6Dx)Vj
zeffBu4I#n2PuM_h$?D!_Lmbr?IL^g>m>IupWh^7`@<QI|B@ii)YGT-NMZTuf?Y1kn
z_^61<S#&+jcHb$pO3(LkZjmkbFa{exO0rf<yL?t%FY{l#lxw=^Xqx??XNTGhcaj1p
z1#R*y5lKi|g2FhOODvtBacQV%)2g}I8h};>%RnMldPjIj#uH}Ltx2~P6Ybh0lU40J
zy8q(-1@lA{;EBAVdEZSk#F66wCy_@Ah|g3f&_@~EBFUi&EOvieAQJpcS7kt)eZ4;Z
zV_S5eo~)ZZwAT`2&G;)fA>x4z?*}g0Cdl$H>GkfM8UVCp`=bd%hs$j$UG?hd;YF@Q
zQqy4Iq%P2vI~=!i>%oYXdxJEBa2*RN4USgw;!X>8Hu{ukee(&sxx*D$lrex9tfu^0
zFIzsU91VeP6+C{q4rrIzZzmXX|9WkIX)Mm}M<*lv6(+dn^CQk~zaz0i1X^4e+UfnV
zZR&RSo3B>VuO4k1@^aq(I%G#$ByH(}h@UX&+Su2#U+q>KfOB;5NVie^S8c7JZD&F0
zYw;ue>?{|q84Td;^XCC7j%--~l<2b?WGcfzbUY?Ntf6~02x^jr8c$zV4;ZY?;SRmb
zb`*G%>vQPQwv(K(-IK2gtoj)9E~VS@vh2rT0#ULkCm>CovH9??+Qu0+8jibaRmz`&
zL;Fv}n;&>;+cOLVpa8@}^9}t-w`f&KpY4}+L&>4Yy4%eccp~9I1r~4yq#wwaW-;kA
zJ%c%p&^lDVvS4&ikE&c>yxo{(wk@WuU1@Rrk-LSp!@}((vu`wOa8)Ags-X?vl`tMb
z&ly##hZ4d1gMv)-H<4WVKM3g_2rK@`<0f6tMwUry@>#Wmv#jn|OP~2XoXlR`2obH$
zzdw?I0Nb$`PaY&)k>b2azfk788)~8|VKYZPMD?zax$vs(Kre%q-PNzlLd&@oU-|2i
zhY`kS1}D<um#rLBg)nZ_5tloi$W<_J-VGo_<8ys&iMs(>T?k!C#?vuSrK>=TE2OHw
zE7jLOysm(eaX5vyW3@76@D}h>D-I%WEUncJkygsR8!`S(f~K#1+x)WKIk|0u6>+CC
z8!K~}ACD)iCq3++I*X4G5D(m{4nM$j#k02m=@7#CpLDJNC9Y;$>o+lj6Y1AF`tZwN
zv?rHG5KXKB`e#~X8^p;H-EgW8U)lx#^?|$9rk$Mlj|uF5@QcGn)g&@yx5zm=aHp?B
zVbthJWTJtlOCN2Jcc3GE{0#~d62J%}2A)5hUymt@rFTxFarwzkpXQ%7cUpLQ92i%C
ze}V5{`|m}d-C=#&BX*~MRjq0KK|1#uxOyO<lUrxVzO%8D_j<Q|IOfL5NL5>MuPc9i
zJpX*Uo~^4UBI+-#<$<r!aUj_K9lk3`S-Snn?ZxGL6}FOwAs|zxq{No?Cz3Z*oJ6|O
z1(lY^p;Dj=cz{e+(gwa2-mCvSeZ5s$dpNGCpFw4U;58A&Msc(wl#eZ779@A@q$K8k
zqow=MSojN6Sr$9RITeW}d>-J?HjR|R&m~R=Nc8i?$ba;!>6~V{V+Wvt-EXIgz=B0t
z$6aKQMM=*+NMw-}n1}1!Q530l+4I!4w{FMNLkT&p0gffne~!s!rV5W|AOKH5$9gY+
zZv<4?#asJjF3eb8)h%jzP`XVSzz%ePaV}TmWHSv?-g+$s?`Dtqwo5W=5MdNq&W2FR
z3sdTt%EC=L-3RI|dsPt2e_-7qo@J5@r}QbO*U#Vhb=@L&@)qa;XF_U7=u|~yURreA
zpYAa+f0vWf%p9{$PqiP%mxvsm#@MhhE2oYTr75PRxOsPe6HSN9ZaUq1Rvrz8i1!uq
zFqJL&vcPBRii-6xS)@wOh~gM)vnmVN+OM}Oi|)5ICWF>w_$XP$3C8iam7IwEX*ygB
zN$~I<%JkCdo3+CCgf0^PlLc)Oc)y15uiLncqOIn0o%PU7!aN;ND-OkK$E;aou@UN7
zKLK*x=Q5!uqzyr+cqMY~B6%^`DRseT9z|HukA0MWq$QN92X*tpBS`>4=n5HD@ZNJF
zxtz#5V7JF$ADv?DP+Os9q0!}bSgD^yp-(OYxl9*z&VY@6h_g?^nZ5-_cU4Y|P7mC}
zC6TjjUUJ5Cnc6AmnQ#sTl_`3XWhp&VbRmxnRlF)HBY~Ck2ZkBU3BcEMA*ZtX#=M;9
zBpVNPsYcUY-mukWp&k3M#vz>LlOkz)t1gFf(c_6~3bE9oH7y)Fy{z47=`sUPQ|b?E
z(`>8|HtXFYNVr!s;q;JJbz_&}KJxZ*{fj=M3j^msi-1)TDOvHUb<-V;496s~sBRMQ
z?RK@8+8F<6@c<K$63!<Eod<-WkQHt}KT_1W9Q_on`)Y0r6sqS;(;(a&SMZ;JmRv;M
zI(9Hbg8*^DTohp}6WDzs&W^LYtw@tIYq}mqJFPA0q?N|u37G*AbJ=|U`*M#)1QuuQ
zZeyVON$8<$^#UlD?)jxrifM!Lb1_agg!Z$TMI(0!6`7kB=0p2fT1ky*x`oQs<Pd-R
z!^}pcL{&uD$H&Qr(2mo<>P|`{=PYEt;_TSPZaM>E<;gEC>B38lCrLn6H=a7lMd>Aw
z_(^5_@dCgZta$+!r^Yw~G(DZicTaUjqULKcFmFkQ8C%~{x?BfC_OOO6rIeikT(h#L
zbD!YVl)=&4uTQoDal526Xu_k_@$<O}cI3S`_=QDD^YJsn{4n^RkMC=IJwMREp>KPg
zbCUV7Al9Y1w{Vj{V@0_?2Tep2ts;2gCFuY%>4`vSt6cEYuzF=d`dDUT=?a>D?W!~e
zuxJe{`h7MfI%6yXE6MKmC{kA0w_@Ar9g|A#rOenj6ABSV+r+VQMO#H5%X1aKvh_LV
zH#S^8G364+6t+--p?2raa@f#dOyvNGc^pe)C1j;wCP1PBL)MFNK`UCj^GF40$XFz-
zLdv~59<5Z)7G=k>U-3Fcap|<?TCcwlId!WsX$l6vPBAIgM1G{8Fgt;^uiqEun9rDk
z-HKcO;{jW1u<}yWUMd3_E^VJngn%bq?#NAr8{Fc;r2CViLFy5><>dv$o^xJVWDdca
z6ynw6j6T2#tl}1c=xr_)0zeC}RX@m}pu8N@@Qw;yD5jlbcMSDxL|FMl^|O6U+!;-n
z2HRlErucEVAbTa%myC+aadxlMLBwJEidx7XPI)Dbnu<r4HC2dGN@<ST1-K3sE~XVD
z{;Ws~wD0eARJj$(biHm_BQE#J#AD)f`;1gO64IJ<A;oUC1j;iU+5K+CHQLas6lz;H
zoK|OQx?@a5JDrOIs-;QOgu7Da6e^goPApO?^{JH7ap}`GOM5AA*!*Cki%HD776QNQ
z*`;%e=)|YSO!~KEh{Ezq;&ucwGn$QOJ{bOKYe}12*%n|d>)}1had3f1b*@~&A7@zw
z&pC|Oovc)PyOeL5Tmy$%Ms+iGqskJBEJl+spM8EBJgt`Dlj<|DQg%+!AVnfO#aeXt
zwg!58l@W0d1+hVa@T?lH8^rU76)_Qc-?TPCHn}y^;`8(9=P!4CH27+ce{d%5oG%?8
zAdN-MzCi0Pvnkjh^@{!D`7_LOG~}5PW_V?iH;G@l4~YzuC9R#eGpDlevsqGyjyn@6
z=yX95-fXmbqmQDd7^{RMUk~CM&V(vt(db9~X=PxRo8+lC5q^=6^IzoNFiQp|A0GRi
z;T%IO3HJd?Z4r2Hf9HTz#nHZu<4CS+5wG$o0FyohDDD9Sq#?>ND15WS#5D_#0V=uY
zd(0vj$Wx-PHHe5IeM3o^S}J);pRmFUIu-uS9G-<pLF`WoiaBWAbJFJaK1^M}A=|f<
zVRCS6*U@h2vz3EKLT@B4_g&M@P06m|R;Sts+zw|)8LTkAP~v@~iy{6d$m&)@0UOnQ
zwUeP}+NL?P75u`ij2^yTmK0sbCMbSmR9N6UZ-3k;Wu#;<nWdy{xdxw5Al#sqcN5RV
zYKjTFL{8`k`;DL8iJZ`LN-|S0Q%zhKV6S$mn(W<dKXsRyOwjTnqfJkZU1fj2Nw;<-
zX;at};}q^0uf-KSORC;Ux~q$x+Sm)&wme}NgiAP+Ho(%o^@k=@p3R+C5LpulUoafB
zla)yGnvb1_ZoeHAUv7_6R);(2oGnK$89dmNEuBI4)(2K$x3eA3^iHi;EZ0`$y7SNQ
zlNJJN3f-l5#2zb?kiE>d<=PYt+q%&r&9L28>p)(qYt2gTr<;GTXF5}JsRu1R_zT@|
zVX!M!UVjR`a=Pm^%Z-mOk5V_S6fJi1H8AQZk7x-y9P-waG#8Vm4z=_ns4F2J5;_uB
zMBO{ry$)B?mW2sRVu^U$%!6(4>J#OuFT6W@Z;|czuKUdKrfZ6Nn<b1m$mEIXR|p%c
zD?2(qcy&QdJ?hYJA?m*VnOt5wFb?LBtx6_Vt~L8O&7G(WmX!2{Ofk_*p?85+y~C)x
zo8+n($4x<@-DF1C)=r1XG=4pdUAM;`Bz`u$+@9heZOnzav=vCpd7Ssbsk_ZZz4q{?
zPN}qa=@0b7tZm`%(t`K-m}LFy)fUru^J87NsLtaByc-2e+RoXW!gb3Q(opuX`i74y
z&>YD1#|Y&KT=9@<lY*^ep6UZ|{j;g}ABt0E#{UrK#`s?rr!!hJj)!b8y{D8<Kn4;(
zpIPho(%IY2rERW`mru**FUvw7-(i77gbDG+>#0dEdw)|a;C~1pDE`Ow8z*=AT86&^
z5SCxaJOFpQ3oi7>{GSzz&XX;~C<VN!`;XPzbJXhU(xo*0GM2w?0H2trTbUU{f+~il
z!5O-d>>S`sS|gaDPn!qr_M;}l6}Hc7nc1@FruQp!55#WplHKhe-t)<F)rr%%cNyQ0
zm+z<7BxCG>%3#yCue!2N;6Me0YkfuvT>ZDxxc!f%IV4x2(sR7y2mF6eb9q7S4^k`I
zi^ZJRJx)0%(PS<Mdp?TsepH!G#eia+0*PCgo_^;fpZm0#DlB)y9SP<yRstj<Nw`pp
zsMfa+(*Fb?q(POAxgVvvyX4>7-Zxx6jTFJq$B%$O&v~p!inx)+I|KnI(4RWwSMAEo
z>YCV1>L_;``|dQwtP;E-w^!>lV&B)IkHoo#JE@j@X`dQ2Ek`v3Vxy>g9lul)j09He
zigL+()E(uaSmj(t5H#@RTPhf^qkf{<u4?09(Q_AwHMZAHT7fImj}8<V*zTlQ?>ZDd
zz#dmA%@nPs6`thh@2*`9a45tTi0mK_V<fr8SL*kQ#2byUYP<<TVPH(5vTr1($gTSG
z9eoXvJ?$p(a{idXQ&R093{fAF{vE&1BuP4K<Q{C3l-#Yhlgflf1GRW;t#E`^Fn~D-
z#pw-++T^qs)gVW$5ZUMhjq?_uCm96xmjF#8*;{jYc2{n`ju_FuVV1q43bqe&FjSO{
zoTapUNgO?dw?@SQ#6mIYwXGVjSk>vEO?D>X$Lq%<M~dkF+fdsuPWFM*!J7iSV+y3h
z=o`>}V@pO*LP4x1*EJ6Kg*w!3oFZnv5_Zlt)tTQ798ur4jyGY#Rhi#YEI#W_k*8iX
zPtVeT`oo$hiKl()>G<RLi=_-tiuGO*oBh{ni?2=Bc%yp>=iuvYthSUy4#usL+5N@f
z)YdNetQhc?P}^^MMv3PV()kLhybq<!LbA$6OJySLbsTs>t0@vz0=lbY;+@mGUh^<}
z{8g&O`ut+PkXB0aPkn8$yb!L&DyEkf-s>ay$o9y8hw+B!R?RbO(nNB#3PVZlpF_28
z4U|osq(#?f6{Q}*&6;1n*)P2X#ZRoAB;|~SfYY55j|#|v*s4&Cl9nYN_m1D@CRzzn
z6S&}%*@_n?ShLIs8y4Md1J8L^bSmildKFL^t(eDIxlbJ`wNdj7;Fe-e>Y4gzr_#5m
zFirhx)=jOqEh3vgBslTk#=74ZhtwDtR8Ykj>kg<4gj4&<9bVEwDlkuN|7da?N^443
zQ+w6=zYRQT+6j<CzSj*3PAp=I>&{GLWe~PvyEzsdQ6Qg7Z@DUI^#`Prm`Mu-Y?qpf
z#)X>49jsdxyfW^O7=iB6H2oCWS{J0*ckTJDrC0r)r!z0l^%E1HpNfC!Z`Fi|Z8#pw
zQ=mHX1qS)l43d7wuJ2Y8L<nt1b=;-JeesvBe2TThMZ#>RIdQ@np_SeZnNejiEJ9>N
zmoh7J#EK#|n?xigGrrGFtJk%Evw#7XOQQs;xkeA@eTH0%+t?NT#mz|A*=go*R>-+S
zji!r&`@>x4c1y=&c`PwrSEuJLokwEBkiS>0AG@F9(wWqzY2Q6h|IU<uMprqfSWDMK
zPgFlno{^Ap=T~)?kzSG#;d9#Jh9L-mPC1CjlHZ*ujS}uzFm8ay2!Z_HhkvEU5>5y{
zI~?M88n1~B4A3^lk%LeIJcbKE5B4oS!v%tPni#I(eNGV^gRTZHpZ+poyqhSl;WK$G
z7eBYjY0>k5Mi~FcvO!qB&q%Y-mV@^eVTy<M4q*rv-(7ro8abZf^FYEFEEyFuyhVYO
zP~4`^?qb9QU-LRK!cUNC+|izUlNgb%?MwK|t21QfrBvtcO<FNm6F<#d!#wp)9L_{r
zzjjs8!7{wO!gU1C?BDzvAXqghppHPMD1g9CYz6&9Ag>yr$5MQQihd&v|Jd!jvrt3$
ze-*GJxlb;zA_YMo*pY%aQ?Mhx_W_{?HbhQ$Ps??76dU{v4}De?npC9s1+XFk?K|?#
zvC#*6L0FMMK4lJM9`E$nk!Xp7!T8JaDF3qnj+L+?S4EHUHsn#lxnV^lejI+MbuY=I
zRJr}W06FF>E--IV`@RV+i~{aRB8_i($t%#T!hD&TgN~E?;yYzZ6Wi_N!{`@t<85N<
zAb!2JzuMe?I`72IRO>onz;-9`RbgM!&sCT{&D&7j+bO{=5!hkI`13$d9LJOGXJs7e
z-S^JKPJQwb|B=}e7K&U$DKXM{=}eJah%wN{yGC}72rFG-*|al()6@K<#xmMR_fhr<
z?fO(#FRz0uH@c=+>pUjA>YIU&m4PP*iTrxpk!1rrD_5QJ-IyEHt8UkJ%l!_3Ve-@s
zdHo8>-UEeO0vjJ*x}N>Z1&dXLAmtALSP~*gJ6AeRTuLpT(@rWv_J+b!ea;W?&93rE
zuZota@_d(GIumljtVef9ko~M_lpAR5*uIkNa132ADb*Q&1^gR}T51=QwM^Y~@CKFT
zsQb7GgDDu5GE@V{tO{KEnFpe2t|XQ%_rVOWZ5REU-eoI{)X4}LJ|9#%3R%-QwkliF
zL*%wS1brQ=D>Fsvo*7x5){DJX9qpuxWuY~(9ZB1{9rwcwnmkhPJ#LY?|8(T-0uXC&
ztjNP|*;~?kWgrkGU^w+@CBo~b?6ErtHq@q6p|O&H`=oMZr<<H-V+xgKrrSeEFGs7&
zIV2mGJHxHxu#H0Zvfj3UFt;(gxzz$+7I6^1+t>5w3!2?7NA*7yD$ET3yOqzv!2Dm?
zZDw#fVu`yLck9ye)%g-tc(5>F7>Ztv`wJQ}_LpO`-+q7}DH2DDw3fAKulTDJ*yB3Y
z+0}kq>@vMy$D3<$c75A`yV-Z5oL?`8d3IFY@%wtDLqb+N*@}aKo*(zOZGC-G6*h0)
zA@Gur2z+#(*F$aZZ*Y@gh*Z(?`!S06ySSCFJYYONmmF-<9mlLsyq~T3y9!hlhgWlV
zkeAdy_i^}sUu2N;4>|=)1j(E`)k27?q-VXw=cX$Q`619}&qYJu2D`+A)<Dx3e*lTP
zPwk8jHi^4Oh(Dq3uv5e`{K3KAVq)he%Ri~-*xR^Xqg0!gIP}JnpN+jUcgezSZJdiw
z40L`Exys9Jw!Lb1Drz?`PMDGYaj{$Bc8@!`JDKh9(vQx=-QWKpp^n!tUJ2lLMi*jk
z#Iz^}T#ImFt`kf24~kX<J$2qF`k2;2*!Xx-fQH6`0@{8+IMegKujvIYdNd_@(>YAH
z0I9O6F`8J&w{R8itCEJ6m|Zyvdh+fdt0q3i++goyKfrFRN|x&S&8(4%Q#Rt~d0s=D
z7iXd99SWBeB_AZ>7)s}ABr5nqK8OoZ`F9D4IY<GnSuv3*rHPH+r}V)H1FphH{_+ca
z{2_J(5aZY)W8xV$7P*Uayn%?p4a9bXXqM_wv8)M*>J`Q9{i^M~5heGjL5$qFi-Aq)
z2<MKLw6dScJ|)FH-Mk3#cdw(iGr#D@BT=A$DSwpEJMx*rKC2ciF{+PnoPPy*kLV?*
zk$vhB^(nlyz&roQE3;pE-O~|UgU5aDOSGzi+jkeI*vjR|uW!PzlF7c4Al9=$g`*4<
zV-kdr!ZNWHff_JUcazP6vjo^;yxCdKI<E$ys~KMHaA(~HfOk;eD%pHgHPkX9EM!>J
zS*p)~s3rqylNkmJNmH!ro<H*~p<lYSr)u$=^d5FwSG{Ex5;GMxV#wEB)>TB}x1@=*
zgQH1Y93jHGw{Xf0YDX6tU)BX=u1okOKF`QirLoRtYi;ZoG*Tm!mbKPiq$fC_M#0q;
zMd$?9amBM{Q(0!Xa<KBR2beO@hRA3Z#97uu&z|3HHun;S7<_rzK^Tb(8JYi>3r60N
zREzHq_a!~p6$o&u+w@X_)z1hqA)oVivNg_L5+2jN#_S)Iep%b4!yw;4MNt>&ReWp5
z-wwqpuf_f5rdV+C`yU>t6m=;4CGtFC#yIg&?>k<*iqBbZR=ArE0Mb0liQPw^vA?oY
zg1>P7bXylMnpiEx{Qd;+JjuJmo#*9qjA1*xuS2kR5eSojb@hLs8Pz0gSRda@->k+B
z6cx^3(R4@2^G^71y?rJuHI{PxB2v#gmPoC7lY%d6<08UgRTDGRSz;h3D|}p+rQ)K1
zc{VX!gwclIbnszi2z)WXAS<}oqj4sPd23X<5K1=L8VB$RbaJ#&kqzcW8ymg}*KE#)
zgnj4<Afcec4y$e3j2i6+19=MnWo2T*FlN~~ib_*7Ud~VsO>G*6BWFE&{CO4-EIv!d
za#IOF(p;SW?P7<2#X1~F<#-~G9<~Sdvh$6>+f(<R?*c%alNBG=B)TcNVhNqW=p}He
zW9>gh$9ald#)~;d`r+0HSQOOc=zoxOf9VTP=M{3R!*uOYpXKr@Y<d-bAvW}!vR!mt
z;F=gEqU?;%K19hK)1{fbyg2z}8ij&XKHpw_DHz>WnjI+JC@-d4sJf6xDa{xKc{vpX
zDU~OV%K3UNZr{Bde0sHEc=$&ck_yjZU&56m(`Tz<3T?2Mjj_)?8ZnisQ)X(7ZV+lb
zXuOhv?*+=Yk~&%bGY~jHk!GWzR?4y|!lhFJ^S&0vG=(R>huE)moVnhBFmC#2N!qix
z$EuZ$cql{TJ;k}raZ@Wpb!6PNcGj?6kCU0U^xeKVa{ku9m(1}>=Y+T<px&RzfjAr8
z){?*I%GPNa>f4U-dmO}>9ju%+Ek;{qeLZ|DpN*x=W#qE;i9(O{!NCV+<(IQnM=Z6J
zz1y2Fq+cE_^R|9E)g3HRA-7%tSJ?(Z?hGnXb(kz5i#|}=#${EB%k(9?<6aJ#6SCT#
z+_&{|#(HpSp@yp*X>TkgS(jZj9__yJU!R(;R{<J|5BOK-<U$!{HogTbbpATm-Z=?5
zY$Qyeh0knNkRmHw26{GY>^KOHv`EO@8-b5;L>g2SJ>pgxc~h)ngQ+*;t=lA5k|dwp
zwsNukepQ+OfJ|oHd;;vHuno5)8ByHIBs%9_$g50Rm9n3%&kIZVK$RSJ$w=o4>9qB<
z?O!PWm*h`H{W%*4)axf*Xp#Q&N4!$K6K+R%U<y#RR+OM!99XJJTAW3Z;npPOTsbnZ
z@9F9w>cJKr6H<{ppIM)`YfAl_udn{mpy6{BD)&;gLFt2pG-jc1!2Yo82G4EQ+u}8n
zMTLe<q-(RZId0icJjk}}xa^qvJqh{LEzAnbH0893jiQDFNKG(QNF@bzwq#zN#Qy1%
zM&CO1v&+It)Umu>)_lzw4F`wV@;2|jp&EoZU4XI0!Zpga#qv<Gs)kZr1i72Z$2qN1
z){`*|=06y74AL^6Cxr7HWdX)`36E^;m(6;kMANPnA!dt4zP@*Dd2bi8{xOEW#?{Y7
z<^?-d?+O%G_>hQVHC^gyO)ErwJVv-d>tY@3eEq}LDhbny?#0iAoZ>x#DhZbi!od7e
z7lIpp&ceWZ_#U@<f>^|Jy0ki~yBr~Vd(I@1EYjM+cV*;Ck;|+cIv%W*4i4~JzH&L$
zSp&pMb*nT}vC2ux<%z?7O`;{GZ4tfVd8fQEO3j{H`5If!ZR3E$$&TKjvXACqVV}+d
zd^PX<^&_*4L=4J?_!W522ZK-yqZCgfZvk!_D|{%j(Ed@7BBt8G;Qjx@*gM9G5^dYM
z)3$Bfw$`+5+iTjkZQHhO+qUib*1qrkILX~FIZ0)tlKN9s8CB_{jo!Y-IbzR-;jOU&
z^+1;=5k+*ts%Hm-l@VHcNw3Nv(FhL~muqR*Iv(qp(HiT_>E_>YP<&OZa`J5(%Y%bQ
zf#Q3uJ?M)<Jd-Iok@N>v)M5$;!`hu^1LDb6Bu3)+63a%w&G(CoiH+Rh4$3Eh=PcSD
ziA*9AFP|&?pr<f6n3m(dSU}5P$BkV>6ZY1dqoJ<cV5yATe1}SZsZS;~WV)o8r-51T
zeViGwBSa-=N7(oBg*$SAnnjdPQU+unL%2%|oEJXvai$7r>#hQ-ZnWVkxz2dqZ@GgM
zM@ZQ~*{rQ{UN0z4tQHIYD>&a&@VKG&lti>I`fzXJ@|BU60R!}gJBJi%%>+oZP`s@-
z#2U&|L#6=qDjX`>Qq=XO-0tuCT3$D*<|;G^0Ximwa%T?XMZO&*f}42k<_+9=XApHv
zwxo6TqHHl!g(IxNF&t(}5GO}Z;yk!&pS`u};3tOZSO;1vOxJ@IfIB-+v<YQSMu~0k
zsE|>YZl`9G1dw*sB-V~|%Q~k5G&y=PNc8+KlkmTQY^WmLce=iM?LqX2xA4zJi_i6d
z(3Mb*srlvxe9I<69mB_tdLRXHi&5ukq+KfrVAraChhoQy%4rH8Cn}vLpfUG#k$WF8
zecbezR7Fi#bs1FE=N_u4vhpM5*$tUsxW{^G2w0EA2)01h!uqbsdzeJZ{C|+}jV8~e
zy`Gd%BC`qbcJp!}OK!<eMTEQo=4Syic`%JM)k&9pvbxTej?;u$32Y@TAxVhWXit=9
zRxOSqYr8<a-=pH)8MX|{>E+iXlgxLO>QbUEV{X7=nDLrcY2{QV=dUalmE%a96Ge1J
zt2Fwp==EHWYzBpQ*>iCXS>o${rJz5TVJnp}XCAx^&fp7@)nvld4<x2O4g1qA%a#XE
z#?5A5n+C8~6oJ%Kx}=u%xtlBa5mQf1+Mzi2>f?bAx<LXNmkWtUvyLr^LhkCQqy23r
z(+$+{&ya0esY+YPf`@KOVmr=HhtYc4>{PJO&&H2n+RVE8(^e=Sjd~G-^3~#;k5zyq
z+4jsDccMrO{NjV@N5@4sR}tF+y!vXbWRf&>6guCz(v@|5>LW7UDIkY|x7yt65ulh8
z8d}G7j9BHksyAFv{gXHYSQq-dPX;Luo1!qV@M_=t?G9Mi9SY{Mh1sPWA!QQx0oQ>K
zmaAz^bW?1zB{Mg->Dp$;f)tR%%TJFzrd{BX%-^$ZMa=-v-_=1L)G5ul^Vh4BJ6Dd;
zmJ{_AlQ>p8m3HJc>`oFMwcj}cWQgj_|E5Ib7|1X_{;@Ap9ejqxG$Ba^A_3`Uh@>vG
zh<;(Ki{YG!kt2u$G~bdK1}~y@r<&w_T&hT+PKho%yRLy%0LKQv=r1gAbpudWCT1I-
zs3cSp>uB4!Vex@5`^?U5=#FXYgQ|KHa@HrKTUgO{*|5R09wI=WY&zYnNowu6L^Qr3
z=BFSGVAnOge$Sehj$;Br)B>^09v7JqrZpqZA&rWIru-QVcp|0qWu5z)Gz$w7Adi_@
z#1Nit-h8Gwb#XAKe%jJOAD`rA`~BOOO3k&(Y+3b|*rb?Mthz|9w&t&N6Sq5BX9AY7
zjMXR3tYP4^fmlxTF!#sL;b`kG5HN7A^?zBIF){v+9_#<VP<2jDw*SFMVEtb$+Lkl4
z{xK3zgLl8FA)ZLRA>hZStq?)Rie-_9!VF3Uv<4YS+z%*JSVCAN7kqc^9>a(^nUO1B
z>is6_hA<!EB9D5iM03!z8D^UHhTP7B&4pq##l<x9_p&9rCp;4y%vBiEFVINRk$8j8
zq@y?1OcpbE@*LQ9LVvs%`Hpr3P2x(jKTW-$k`?VoPOH2@?7YsK2{Cq0G#Q=Nt@@8M
z;jj`Qk)Vu%tc9+4ndQZ4&C4_7h3r<P%OpGjz0JE-BM|YOV9Jvme`_%3EtDdXh?pUQ
z6OKkOSTVwj<yNrAn_CY%FPv?OXcQH_B&40X*+h|@Kavu{kx&IE$VMu}l^7X4ana4|
zhlVi9)72wiMvW65H)15n@tbi8l+Ga3zlI1jBgl$-9yseMMMtOCiz+Vh9U03d=v7=Q
zW{KGoog33LnKe9Y?#cd8QGg~sYG=YGUNv7i(srllt})ZjOI@<{8An`-$@3PMeJTVg
zE|Y7rin3Qg%0z_mZI&hTVP$G3JRvo=zpge=I{N@=IYz{;tCcDtpUOP*@#hV>>azcf
z(#0~DqEx()3hrLX%Ouc+WiqF=(n-7}%@L39Ps0nzH*&|Dwk$rtBGsYnx3yeOhkTL8
zR+Xc**S;EDSX=$k$>!cm4Y+Q$L@8f2ysVyT{x<<bDRNXRV*INcXJvYa)^z061DZX_
zJC@T3hgsOgQ5t9D7~F0rwg74!tJ(FPyIs4%X%Y>={VY6q<ng>7#-F}@b!$uh0c%$?
z&(D|d#rVBFJ$~$%y?c7x+B@7%!T@jHpIXl^&m+xze(sE&8{AKZHv=U0O3-#8`7~7#
z$r~B@a`~8_zMVcSzrQ^`7QBYUO42Ny(xTr`3Xk}GCuThf`zUQ9X$0w7>d*<@FIl~k
z3`vi15%`eRacotsk}?#ji@W(F&bpm3egzPP>n|Qc<(Cq=&NSVGz~UNHK?TH{Ew{B0
ziZw&&ae)x)35ZZ${KiGBX;x{&=yMS(sk8ahf9_{}F^L5G$!2S3W^!(A{MOectLB6C
zxmq<d&Lf}KKMu|>3lozOz5AqNW7#s7hua|^M|*!--Y?Ei%o~|}eh&6OyaqAgMXS3y
z_$GwAusmfR!KK%%jBoI<IEmVKt`_A=N6{}6ODm@v&+k*=OK;VBZE|)sS4aQUsseEB
zR?V=dV&(n4Js7`m^mcY==*TMU@_9CYet0^2e75uMW+h5R%X$0)l{<K`0b6)Icbu{J
zegmKNNLaLZB5~EBD3t$`uY(?d4ayH?fMSetKvATC!5cdN_di49Q%)}Z!Ct#tW%+Px
zNGIYgH5mSPjpsX#)?gL2!Yt{E#MY<Q3z{g>WtXY18F30MV_jN5Ttot7bdd=U(ScSv
zPcH!kAjr67C(69mTmOwBge~ZK*>FEu7V*_GsQi-k`7%ismuNBTJlcRq#c+b_4J2sz
z&IBw(%0|f-I00XoO5Lh)cHMT|5oC|&1H60p2L`m5;2UX87!5d3fGMG1QJ-aHLYvqe
zti`ON8-B6i-+X*ZHWy>T&8V;LZWKl?t$`awgbqe6@(yZ9Cd5cR1Q1RDGJP+ZmOj?>
zGLv!F)ByVwtI0Jpe^TN36cNDHI?myK36ehvM%$qCUerk5r$QtHCRtAUan{MB@Gy*q
z$ew6KjmYXIKp=2Jin0!Z2z**HMZx?~L?T4)n}aiVTR@x<oQc5R_BNa_;uw*V5LjFa
zsuTz`1=o?tam+qM7V$-xR_aJW$WpeX)tYJOnFq8%Zs;Ig$^r+;NRTc5KALbrwT3Fi
zFeq!eqzpF&_QRz8w2u8ab<pI{Ly+S$6$_+xrVr${b_&LG+*gjVb^>tFrsG5e=*ZHH
zAM6+fMOq*X!T^ta8E;~h>0}>5K;)7;`J^Wm?>E2*6YLyEv@Y+xj$;Wf9LDjk`*EyW
z_Tre<@?h9W{a}kN4CV${O$@M_X<)R{foLUy(2oRJ?lAGA2<}f=e`f#Zpe;6Q@}xcR
zsU!~sB*<btj!(g;>!t?x&zt~*UC9ByFTRf2HPzfAE<)~tf<bOuIfa~JLw%t+p0|6-
z1p*&Sm8auxA4awgYL0%M)s06W65-5>3+6u-OQMz{WJ!v<+SR8R2LEQeOiVI)6qRJ0
zi7brNxV}5<0zuP;KnoU<etIN~0fh_?J1i3<DG)AICA<_=A^n9VjKmu5R(mOsK_ofQ
z8Ayw`k=k$_>KXLf6za*vgRF3*SU!V3K2#3Mz`{Dw0h%8%WB?y3?-dtJV7OiZ<^U_C
zy|ibK00LESx>ZhQZa*K`4jRS8#1`+XA66u?{Qla3fOvmsyW&67>AzJe3SyieE1mVW
zG}dqjUP*o^F3MvRi$z(DhC(F~>s?8pxu-ABFu6`^hA>Wf8XriY-!l$+<*$F9sOv<t
zk6njuOlFC|Ff+)BJfzmY$J`*7YSM&Ys$e8dS4p9vPpW<4FChb*?XLqCW+$9#qDt}a
z;qx456f*mfed7TxXd);B4AdYAvK6d~>*cFJKbu8{R>nxQK+JJ|@I>Yk@CrJ$vqD)1
z2U9t#ZB-?(yf!2)>@m%{bI{_LxA3xKyhe112WV6TF1iPiPKWA#mbl^<pmClNMi}0D
zWcaQaCGq2E>wIbcWX#$C7G^yQU7kM6(BGn<v_q>(vaJ`!hsPqyXllUG0FT~Tq;Cc$
zJfa2TVX<&V5p~YJp)d0ct6UAx6HZh_Gu!|YwOBr9Tp@&scL=~f94baee|6fCzQkV2
zi}+v@?l)!BX@HDB76-3hW<umk?7s)HpD=p2T>lvV?cf8S+s*Z={^kF0Xms~}PqpxH
z?s&hySj(yM&GG5@cz+(HSL>D5)6uDlmAjTRTfATYaPwG~RUNHt<8n9;J9>Nw-SB$f
z2md0ddI(=CJWTyVUekX5aA555w7p92^N((l)uWF83v#fir|R@nYWF?5`!eloUx*{+
zHK{XP4JNF40_8wT+|_L&RKupcu&Fh&;Nj=%DMcuph;h4+01+Jb*4wURVe<#R^#RU9
z{Whf4qYc(#BZS39IHRTOk8#iT*R!-FDNg&&1vRt>gQ`Oyi~_+@PwKnxUuTzp_i#D=
zbY}iY9Qk_jSSZ!Jja}&d8dv;I)o+ydgXmITg-n?veqyWUtX;#UNo{`S3WWrr5$PBt
zu3Bm5YK}*ajETgQPEChQ($nto^w#&!*%Sr2kMtUbr*h@E8t~x=L>K7UJIF^x;|aau
z)XvHE??07UIsI|)bav(F(Y#?TJFC~7cVO7)@%}kW?_&}}|8#ynO|g8TGc&ny(`>0)
zynD02q+%u!>uLcB@7TF-Cnr_>uLQu=M3SnCeD%<L<;$1s`t>ebK<C}Ts&KOpDsiWM
z@=oBN;%~*?+DaDw(C!qL>Xw*q{N!$WqUyc~UGvkL>{?qdpK@KYv+nbX_ilx48%KSt
z_MSy$y4ig;$+{eP)+mm`cE@CYIA~Q_9Zh4hJ^Cl@O}53SIuO$zis_C-cm2DZZg;f%
zx}g)xoH2e>R(2cT^u8REg}ubEyO=Jiu&v%gP}HYx$N_N>cs4SoPbCWm5p5@#0`K34
z=;-v7LRG9PVGtguriTgpcTC<Ggi8_fTK%Lg%5+y$;b}*w46WNi%#qHqN7*3X8JcLs
zI1#T44<H`#t>E=f@DgG?+vL`nj#kO)Agy;l5(34B^bx#g4J!XweZ(2I%%dQ65(yYQ
z2?2s!{9EUY0!C+&&tZ0k<vS@Ga6#=YUS2+?hq!r5@K>uG`$s^8?QrNtqjAv_5_s+W
zg{Yhm`-Okk1xM*<ug31Q#jDdAxyPn2un?0=9(^euz`bD1!Lz1PEwYg)BQZk4Dmub=
z9jtZ3ECsa$k}yJdNx8bI6?E_K+!KuEM`gHWt`9uBU3o$SDSE9U`3eS_7ry8<^GBwq
zDPce`!pl`?1@nVwIBheOuensuuYmm=HBlJ{AwYG4e1qo;@rd@s7gA6TQ$4SH`r{i^
zKQ09WXQmc+6w#$sgMHQ&Repr>%Ea}hZTfDN=RbQ*<1*{=iazE5uI7b*o^|IJy~>|0
z{Qt7-U}pWl_k)<&IRAIn&S5PHJ8E{Mp3i^Q&NPGgVOTc=ed<|2N2$F{ec3>gQXI11
zUxwl0w&#Krnr_Y}TIa;!4UMk~QaRu3qx00%*V)uuK9-!UJFqM+?;m#{4}x=g@eYgW
zaX^U5{ipZWbG_Xn<&+xN4tdGybH}IkMc=<Z%X=0ea}ast0H<o9ceJtdJzr`LG}G<I
zSQmjkn~&6zaa1!tg^_2%9KWBBtLPdS$wX^!j6a#Z-oIY&x06#Xq6qr?Pr1E%2f#}N
zO={%V5{f+EkE7?ask^2X-HJSMibc&MFp_}jCtBD+F|iGbFyYqX9Kr#ZNGOZ1kCn+1
zULPSObP48Q;>L-8N<8Uy|74i-{6RIT2VDqwp5^z=F5&L|T_AxoCb%`r@k4>;my>Mr
zFPRh<PM3CW6IS;9fN+#?WXkm0J5gaV4wt|QFBvlkQE7E5PePQXb+R~~#*)VhUbyfB
zyMSpTikmOR%uK0oeLSkNx6P-Aics<4fmk>W$UXZ*<tcM(6qY{f^y?BkFA04(CAia6
z5Viol!hr47uln9oGtngAyd|EN5W}i#&Lox}HH(Oavkk~y3&wB91<64VBs$D;n5N(u
z5}BL$G#|br6}Cx3e86c6{xzmV9A2~}PnRRBq(Y$pCewe;*?7-Ot`?CkuIMwggxpx}
zLEZJ<0P$$}0OCde3QnYhP%@=&JcB6FJD8mvq@YB~<-ehF*??h7-@CeG@U^pJY=SUx
zZYh=MirdtNIn^YUZ$_WVs-%+P*4e_GA%N6DrLjz2o$sV+v}!FnIN`RDI>vtfuyIy(
zJiNVYMWpi}{8MiJG&JkN6sgITrM;(x7lVByzm3+g+;llnoFQa>Y-WIjs)cF=eI4ug
zek6Zf?rtR91i_SlBg129Q3D%FDfb7CuCR$UEF%0<-ztkVRGt?-3FJ&^4@>A2lVDV2
z94-|p;R2adSD=?18UZ(q5a0En4XiY~EH@1yX{9O?T1xNqsYBc~ZnS=ffY1L@cKWeC
zgtOKFnlv>~`I7-$CevbO4+_1x<4$I994<WbNhaA|2OeGql0v+DZpY}$eKJ-7RCAhM
zb$X$kq#WPrFQudI2XB6>eEcIKC2;x9on);;s?$Cm-YOWc4MrsDII|k>wxYM`WeT?^
zEos=e)g$gl($`2=ZvF{m?xE$&q7bnpW(&p18dLLB`6bSs0u_}mBI{`cTkBE$a6~5`
zO|!jYekum<$$-1jDn6^q=H8%CW9Nfiu5|<r<tPfR;G|1-k$t7Dm6(jnVdsPQRL-==
z$g67xI12T6L3V=_itpx@<oa(EVp8$S^H#D$sGPH~Mze-G$RBdfG_}-J%}l*xEhW;E
zFZc-u0nesL4P))*<5r>NR`c;n(Xzf}odEy9YF^eqjIkAXndi^Cpq+To|Bk@X0KNR~
zgZjFsHN^m`Z&s*E4a5KdYw-99n}r~*-V)sSLqe;xgZ`V*!4k68i+rJV7pt4kIfg>@
zW23oF%$4m{#1y0!oDEyh5Ehk5xNR61z-zgrn%DSz;?w*fyleb$26Y3hBA25ydr&FB
zsGRz|F9DuF_`yXe>@+|(+P+{={#niq<^-oGc2sKS3Dri_o-FD^I_FMZXo!Fvh{Gi~
zntRkIxF%BiQ1|5IkNt{LOj@Tq=HgG+p$>9#Oi2pNaaJ2wv*`n3^aZAq4(GSoI`Q(7
z_e{IeL|R`{YEBe1s=YcMt8H7~n63B+JQ|ZS$-D=jjC{2jZ$xc5tT#Soes&AZ0s-Jh
zON~WKXPuV*A``x3y@{Bg9j!pu9IsP-n}I?pQJo&fvcb1?>__G-TdyDA=^fR?ntGFq
z&Fji!024Dw&lk}ME1MHcST-*Z(=`2TK7?C`XrKbQC>KFDP%3p;F<YXpW`L80vyA$v
zf}k~4lN<IG@$<0y5zLE11DpGeLCxDa9uQ(6RO-bFna7R*+So-Vbb5#Zi%mH4NPt)A
zUa|l%xm_B{P1VUIS&bt6=8(e=KG5iLQb5f$8@S?F|7AngcqfDvT8nTCfm$;fliT9M
zmicSAmP|?qBj#>2WGCahz0~exZZ;|uzA2F8t+hhQWCmJa^qwTSU+lJ%>1HDG>Ud`L
zRAJ1p8qP?>o9p^51p-TkdMNI(x)KG=ex8pQ?J=1F_w8-6{Vp0KyN+^_0?Z_q8yOHT
z4n)-o1m;87u3>>tW~b)?xEXn&x;>ay&?EWJril)@c-)xra8ubyj1QKMiS4)<EqdeN
z-Fp~^&=Foyn2O?=*kVOD0V-UQfuC)tM>abS%gT24lC4SY0DGGa8vwPA8YK#^a*c@&
zm!|qpmmc-k3!~E}FUd3tJl;amuI5IC=(Sq*pyuMBk~>ORgK8Sys&~Tf2xaU6g%>_z
z%#j1G+p~;8@ELsfru2CO&ob|-VsvG$+LqYpulj4FCWDjFD|71X6#PwmqWEAT{=W)(
zIc1ra{qe_c8xpvJFn=*GJ`$3iEmJafF1}Xoz3J#(U7wXGjlNl&lrh<><unq%@K7g1
zsBu@ieAaVUyiE!NtB6iWH-C6uhp+~qqS`x5Ze5Za9?P}Uy3;##Uc_(Hh>{dgcMXo*
zYo@uSu&lIT&FOfE&SRdF5ktCj!V3anNJql&$y=qH3Xo0x+?PQHxzF|^9hdC(lvRd`
z)u2;Ry1r98<vDCFn1Q3C(J3~dKrhy;35*cRmT5)<6Gxg+Dysg;_1@&_=iKirt{d0+
z=V|U_85kJ_5tuHRjgn&w%p<`A)zwgWS>g#any5&@#RNZy)lj0Pd<kEW)e+$jhdifJ
z3|$MF;!ScP`=u&90E2Q#KXMFTL&JjxZ=Ey}>8c*D)>1K*Uc(8*#rHDvr2;*#Slb?g
zt5f#i<2u3K`_z6%Ix_$S0`F$rgKZOPSq4QY)h^IQUbQCukf5fz-NG7R>X>+t)TO)V
zYj=SLCU^cMA*lE@j(~Qu+S&A-%E#GIhul2p0+oe5w5FTqKGHGNZll7UCMOr+ZloVK
zb4;k=6IBG_!a>?8CKg^~TKw_PoMU_jp0+^ioWDtl14pHm=KFUqalTM9z$rwuTLF(a
z8qOm=Cn{j=H>J8nXfP(c-#9qneXr{PTYiUQH%_y-pfQ|P7;wJ=!Mlh!cNrfgN&_(V
zpQBW(F$9DVr(6i9($=Z1^(+SlAc22w^U<-U*;5c+cV;IJJJu*%rGLBVTMw?BKv?CU
zuzgEP>8y+vEqV@z82Ijbrtq_3{dTzqZ*X^Lfd71IWo{!plokJj0m8p>VADnct7gZi
zlIrUGsVQ`C-0sfA)_Qh!4P_oXdRi9Q?iL>fNLZvctX0_k8}iWfH^ilspW&6k9g6i_
zO+Sk%b-U%6_E9Rt5xCCrV>3Xk*L|W!|Ek4*9~{M9)m-~@9hw(+3BqppUF-JYxcydl
za<(?Us@<RY$!@G}`L>F4PPi<(9el4%wYZAE!RcB}z>0pjDz|yIIY0WaKCo2#61<qv
zNCrm7?qLHyP+Ej{i$oXQEq0;MMUujaUp025`uw!c5rzO2o8cbka7}%3t1`nY6t?5*
z_vihAzY9KC{=b&B|F>C?`G4Dwy^K8>gWZ15r#H_}Ko@BT6o4S{0uz@ByBmN5MC=0R
z61R=>@JX34Y{tJ-IB(4cY+huVIP6icOY;8YeXuyWcs#jCKdUH{!U?M?t=A*-p9l=_
zm}oIMG(Swp#njwIy_hWMz+7k_lvBln(F42C-4Ca@y*hAx_&#=k(m)k^Y44&!JH(HJ
zG^+r@ocr}GxCmf$gC|V>)I0Bw%T#`T2O+1S3H4zF-4V&lL<nZeEYU-njp3HCSTUZW
zgninXVtZYivQ$Vr<+DI_O1&H=5D6n~EMR;DIV+1C2uRNg=;t~2F2!Rt_QMBjKW%^>
zHY<PQ1rAu2N}{njuUsuCB;DfSQbx_vABV0P=3reeB_C3pj`w#HH%}Q3QI6Sj<emFd
zidMZGTpEjH#SBghvG;NK^w4lgi6kOZ$c=^=s?ZL~Eop3#q!BVaT?Pw%9Lj%HSYXtw
zXp5Gcd4Bo=v-#JCWS0IE%R0nJ_Lxz@jlfd<=_28UQzieuuROPD=Lz;ZyqoQrT;B5U
ztgP#Tx?G(9o;p_M{;~$jbH$u=Bb7{vpI#W7CjP~t#Dx*@tx35=K1N1(yx9Uwhqwpc
zl~f<E50>UI8E>gy+E3b4^8O|i1?b;_C<wq^Rlf4{c^@3onI<0Gc>T+QI1P&Ues;8A
z(KFkinc8_0`ieZE7An2FI|+=UCUD9LCAPmrnEgtMqfzL$$?!6O&sWM+(aX`44et5D
zttV-k1qo)TG1UsoAhbbCvri?xj@3^D!Aq;Q(hAy&(%lUN3tgEd%z_e49}bX3_v>k*
z$nDjbKqjXKNn?j7Je2HLs`j+7N(@Sfig)Qm`;Gv$5wyffwlv^=bLOM-pQ(ekmVKBd
z#cBKP!lVh`5sU&+QfcgjERcHigNvi|a=kO^g<ZyIOHa?IN^bci4B?T;AcjF;N3Feb
zyE%N(`jl_`QHK~MgZy95o39q;@6z#Er{b*aVwX9+I2!q6gS?9^WA>XBuJh;;ws0Em
zOA(mH^=)m3in3iwfVW@t35(5Mwg<eIY3LqVfcl(VP6>c7Sd=P3e_k&3wYo-7ihh<6
zvKox=Yy_{^pbH8HJ!#Js#YqfhKSC>lMERp;sr<Vl<U3Pl=_@A)P41AG5hr3obS4M_
znLtjrgd?OD5n5-$%PXN2Qye~m04DCt8)j_svT3@)AqVT8h`XDZSRC4{ERvzCtA(NS
zy9XH>I3yDVCr4~ZSyr}GFZerXh3Xu%mS$-*S{8guI|)b57%wF-;iro_$kno)6c-*$
z$CMzDBbd)Ec&RUv?_26Mo0T;P2!j~q8#_g@^g2N%Xt+Sc9kWDd&Nj_}$AzLfd`X<U
zd@W3^T7ph5Q^+O|zq4@(K}8eVyp(mh?Ba0VF>GKPyG)sDK!j_T<1vk^)0{tWyH)|-
z{|yX-u(zy=A7Hn|*9lURf_vV&68aO<q*LM^Rcc6Z-eb~CuN<T?_oyQ&z(u>;l)Qap
zRG3QBt}Wr;G<MD5z%%eo;m~e77Ra9B%Do<m500png7AF`07|A}HVvD#@U9G@e<5g=
zXritMv(-zgMO(@ADQa@m{_X@pG@$V-NsaLeb5@Y~miK7|hG8j&Psc&23X?+ds7_JD
zhj%@0p)%3OiWf?4>A~0;;aJMM=v`!3vK09>9>?0wt%7{{(`mWScJ-dury?zCd4f+k
zc~CE%6X;XNpv_T+OID<4u14}wZq&p@!btQG&XA6=8(l6uYD?+f+w>EEuGPQDshf6v
z6N#?6#2n2F>WV0h-X5m$@QEH|HqY-dW#hX4tZlt2E5>ZM-@3R<fZ}FRPHc;!#f;N<
zeCCe@IN&+HyO!5efJ*mqgOu~`xrQlxLfQtE0jK}2J}3L%y;9abC3fX^EC5e9SRXNQ
zp7E<Yh>p*l0xl=#NA+k(SzOXv;rW~ID%iM(LWKg&2BqHOOv5*JF=pU9_-M2A2Az7(
ze*sXcaTq?ec8uZ#)S!6rGF2VKNR;11RL-?W*T_&rySKvET(3fM_ej^X1I6JzZv0B;
zV}cu)1@`l@Q3K%1`+gkj?e>13U+yy4<@Wq?M+wY^1pwZb<enpvbm*NUUgEd+GuWdI
zh9ihb71j+i-T<OI;%RC1$jTWULKV5#JjTvlr8`Q7xp+NIu(?b(HrZe%12nUoO87QK
zW43Zh6;<aU<p7~zVdFV?;i2LYS5N=Fi75PG<n_+X;Y0252=^6irDd109+rl!3QB2i
zhZG21!}W%(AJSSHXg*0<TA)aC5ZjMHi_w64tDI8fl%s_mZ-+E6fdHlaz}r+)XOPcK
z6B-Gzz^BPHR22aQJ+MBo{r-L?hy98f)XlgBIRiM0Pi9*b#G0*g091`?Z_=7#`j>u{
zT|JsUgJi6=n;R!Fu9D$y(G4qr7%Zn0Xh`%jC;w@|AL*w-vd<PeQdj9HW*%5igECWf
zq?Zn1lGf0{yNxs(J$Cl)NO0jO6}~H-iGLbZ8D+)(NNR0jDuYK%8q1e8<NaoPwR7zT
z$^DlDrgP%w(Pf2XKB%w-KI_BCOlBhAijnz#gc-uo<x46eW%|STXaLMd6{B1|dbvZ%
zIpif;<le|0Cnm?D+mo61b>**GFBz48zLR=9Z_Gmo+Yz#kM#4wKxaPhB{u_tH7R%ph
z#;PDsPdw6j*?wl;W+nE~<B5Hy&DSkeWxF(RZi1J#ZvUb}?nW1w7M?P|RQnyJKa<p-
zmZJp)ko`^N&u_t&kqdQ2fi3Z@;ej4(lC@4#7}|WksvmPPWZk*Z>XQPNzxw?UX!jf^
z#697((=a^~ZYRZQuO`SvRKrXQ2=qd`QkdlvVcC&1v60HcoDrg^l`1{k9MqEX>_w4`
z_S>dgR9m18qe>PaB5p7;*1SX$k)YIY5|j5Ls~l?SrFe3@1n%t}d%!T<9D_@alb8!z
zt!gz4DH<@t;2uAxqKP)>fkFt{G2*dU*=h3WsZS9YZ4D`B*DkEC3WPYLX_Xs2%UwIy
zQAjYaql#RbyxAckt9hH4roE3t4Siud7nm$mU-(gh`w#{NWYkszZ7Qk-DM<wz`CF~c
zv`emawC#H7R;4x;d$ZQi%Sp<-!Fs2$<l|;26b;oP2Ck+fxTB3Hn@E!%h74*<lh8M)
zd`t&YBc(Jer)_HsF|+1Omwa=k<NHE}y*m)%LRIVw3D2Uih@9@p=_&G)2qR^fwMvur
zSmxX`_h~hom_DhVM2*{3o2KbtgHO^Qt+3AX*c0_T$6v$cV9e|=!J@`TVN>#{dg9ud
zFb<3kw4CM>TECpn+hto$<`**DbdaFd{_I-Y<l=-6*PBG(*7${SNex$PJO)DC=o5VT
zUWX(Pk(LX>m#3+)pEv39E+hutp`;Lm+kqRz-eGIGdmJMa!d^rxI41{yIXxsDG}z}1
zMtw^z1lBx7X=&S-fbC8365oyxG@^;ObBk@+5}dNcvK<Nnb&eux(^Jy5@-~9Pmk+DK
zLYI);mBKo$VssW%LB#eKA!&onxT{IWMwiZXXQ0Vl&5C_*@+f+pmhrYfWsbR~u-%SI
z<Pd>k=Qr7vu67qTIHHV-9#DkNb(I0CM9=Lfonp3>&$RX#pMv<<pj=AbrF%_0dWdRO
z1+|S*5~`7B?_;VJ94s})qVxreanV^1?Md*RHO{CcL58~JE}jL-YB>A}BbOl(EBKiR
zXBi)?K#_74o|5Or?tj2RLmUuqPO8x}1KXjT!;S8PW-=U$<6GTN{TF^vnqVz_BR!k7
zVV~4Nwz$=9+-+5JtxqXWErte~r$}O_H53gTj}!3x^mm-mp)Oe2xj`VGxwgKjbSU%F
zsWP+1lIkW~B1`BT*^^kxOxDnFHMd(RB_0N?>Ggb?vN6lhKQoZ@$0KSBh4onY_)?uG
zHvi(+uo>x_`dI6IoL@v|I<fK@B$bTLiDll*m~~Yx)AGzOYq%ZO5cEb5O!4&!%9{W}
znuB8Oy8@v!dUT8>7klN?H9r_7o<6kLSu<kSjU1D|g;bKI^>lx+hfcE7-}PT-X@=X2
z%)$h$xe~Z`S{r}@ORsqLojdB%s-ahOP)BFguw+zIN3V(oLv60%Y#5q)a&}28(cjHn
z3rSBX5L8eq)Phv4FvFaFaNh1E3miG@!~2DR6iTrCK+h6=nE$s0$$vp8GBGj#?^<|<
zHKk&4$Ps#OYt#7krz+jV3*%izx#Jxpn_DSLDPG1y-+u8G1q6kiU!`X!IiFr=U`4y%
zq2fAkx!XUlkFO7pAII0S(E*cIJ7ev*gYijzRFJT7ueuZmFq=Q$>OcR|7rVT5?Hd)d
zfZ@mX0{86SjhFfHqP)nziIxwN+b#DzG5YUt{ezc`3JK@l`MSOt7bkE<`wXAb@`m>E
z`S`{Um7&ZR=bjJ$dj7h-zXms63qTjU`u@5}T(3|V1ngLRVHmtV=5Dv9@^j3yB@Yn-
zE#QAWKYqqEdy#gFhI=H=0K#}EAthV5&WIW;loqf@Y62$VqVS!!;vao_L0QIUIE=L&
zEWs$&l?i@IqU=Vx9EY<=F9pl+tnzCl<%5b^FuPM{<xUsH6?`z=ZHR%;SEWltl*}e*
z9*q_(Uze+MK7Q(MWOZfm+f>boJ+3!tpU`Sx;cf^bfK5+6r8uQHrMgLdnLmjG9Lz7a
z$~!@lX+(}lc=&RXjLPaP)K&SICa@+fE3e^BJHH(n!y)QuNia&~xKk=;uV{#>uxiYq
zr1;1Yb&=v(=E2dWM?1^YJ$p^wXpE97jc|bhQQmSq5h|gN(%q6^N0&2~GS*D@l;yrS
zpFxXocP8S0l#yCorN$_&rpvt&?=?@vzg0G<OQFpnFc{@wA0We|9pp5w)Y7_{xiI31
zD2Hs477t9s^E&h3kkcBXr2~y8_oZ_iISgFIE@Nt_YPeuvGpt0PMlEynvaJ#g4sZ=7
zR!0-%J((ri$QDtT$0AV<QDj*X!UU?;$o(a&hK^Cvks?+tDDZpjvF^uSutQLoH60Lj
z;(W%+_sY6=o;UhunikeDcRvteM^EF)rL6xe)V?llh2)77n?QH3@~Cjd<hbBhJWMXz
z`w0BRZz0j6@~+9lD=va7u-B*{f?FW|Gee@SE}Q?+9f_KicWJB|l9H8~Vi2JahxKl)
zp+=rDj|NI%t+crpQLhSe$+_IMlAm%MpR;)Omvy20&yNu<Y3yY8oL!eqo25zmbno-`
zc>&is!?|V)0R*c{j!|=GbSV)X&5C`^O=I>7LD$JgbqR2a<`{>S0H}>GSeYLj&{q%!
zs^}Uu-M|YC2ZbIn?eU!EM1jml#%$(FXJ6@b-cpJ+#Hg0QUfr3lkeT^sR%7K{nhn5<
z$%P1?+~xwBBpPoZA4hy6rSDilV9F3yflXmbf>?l(LrL!zKWREi3115zo9`;<$%+rF
zGn_t+Y8u3a!<+;bqVbWW7U(iadJVIM>??Ye2cjh5TD|IDF{S&<tQiG&2A5sxt!wf`
zm7%P3vZf4ExwCb;!c%pv%)_7^h_L0*Dv2F4BLvM&%m_=k1flWgG45aXc7SOO79|d)
zY-IH^15d9s<gzl+B<3A@PC=L3x$4)CDW?2LWNYYqsTqAI6h_M#dq=Q1e-|N1L{+0Q
zN)KAYNCj=@_!Xe<vn)t2_Km+C!ZdOtn#xA};5gXD6OTzYc&29;Mg5T1Y=1rpo-5+n
zHRw9A0-#6wt5M@w{ZJ9<!=%RmPIdwyq-a6Zkbx@7MP%;SiF2d*yoKWjIMf4$oA`%3
z`16y>=7Oi2LK>BlJ{S%!IR&(UA>x>KZa}S+H5e+8=X8T+&{!U}0p^%_K%)9kBW9aU
z2xScdUK@8V)6Xq^3WSc>56!B_NVKOt^jh<(&YNjigpa-D?pmlWYN!xk;~r6&`?6K^
z6}d|I#hj-ptC4ov%rJ6Y?ABncM&z2Uj<Q>;8I_4`Hfe8MyxDItcXLB52lO|378jZ#
z`^rvOoazOJ3~4;f(ER<BIYo1y!&SfHD;h8Sr{rG)NV-+F(@Qk3z5Fc$5D`kUoff2h
zsPF~i#$LOQYW<&KhU6%(T9zV4{N?*$lP&r9I5ElyW`b4L#5->V^#a^=Zy@6~DjG#N
zZ^-S`Q`60lcB&=oo#^DvpX!>Ml-DxMa#YZ@jZh~#?z90cv(PuFYhFvJdtraTsbQwd
zmo{ejqx7ho<1`=BH^lnce5!@&BIkk>#*ZUjh1(ZsR3{<X*V<zRWE9p^nw4f<gqYpn
zH?7qx^b`n%GRe;D(~iqAg0b%A?xc7xFyZSBvlF(a7t)kQ)1V5^Q{bkxPg`2Frp*56
zEA2fcS=9yP;3l)SD;n&`x^LWXOzdybCq$ynrH9v@%`@H|{<+M<%CR=g-i#fMPRDo5
z`?o`bcf<`SaNbi`B<?vX7EM~ez}2pG*(x!!(z%G7+r3t74{(X|W~f)-nl;Hur|`}3
z5EM!ZOC!u!(w;qKc$CE35aji2I=hzSB%o!`eW=YtD8bgdq@JigjlM`V+3a}H@wJq(
zEOTa?3Dp!K6oP5Ask*hvPFjGj4mVvhL=idtle!fySUbwK{n%l7oaVFmje!jc^on`B
z3aLB?c&ZW%j2lpb%z0b#x0C+T)!-}#4%^|dI4<-3SrEh-wsX)A+%zByYj@P7U`cJ`
zI1Gbj33p}XnOgA1YXp(I4_&gK={4N~S&fz|@eu${Pf5%z+s_krfXox#njWE6wT_^G
zQJAsGYcKtVW%#@bR*H-UW;5b-#Rf7+lS%e5F;yFGYPqcH%TO%Easw5pXPZBAgtn(0
zt9~zdaA?|8C1u(^R9mHU0^(^n1$m%tK${V{G|MG92~uv%DC>bI*5R_80f#4h6S&s^
zKURT1RneS2Hu`NCQ@)*|sfP$S1{g0Cwxp7zrW3n04G|S2Kl?|Ixu8K8!y+dMv9Y0g
zI}R}!j0O5@7E^(ctYJeZ9+BRJTWr3IpHA^SDR{cBV^HU1P~*0GIoI1Srn0ygapgU7
z{k<fzE|3`;2CPsEcEi$)(}uF&t^7$SnC+T!X(NGxi;LX;S;*qXjElGCN<2$XFl3^;
zw^Kn=$QAjhdRGF24x&!u8@2gvVP`0Wx@&vJ-8Ff&_SUt^#+*s7#*W7E2e>D3i}}A5
zLY)7j&so~U*38MAfQf^V>3`MBy#4<wvo*D6_!$3DW`PN@*8h`E>;>(pj#2-r559gB
zw<~wA;N{%9wD2UI@Sk*I9U4{CONo@X|D(<BpHJ?mnw67HW`R+c(C?D<{eww^#{~Wl
zZ5EI7Up%qgn?jaUIDVurx7WpCAV?^JAHu(Q;uv_<exrbRI(|{VVb0SaVV*<Rzh3e<
zZ4ixeZ-QU;d_T{Vx$r#5ANP|6BvTOMS(!aQJYRf-TazgM@c$>C_?8mp5GRQB|BEL^
z7W8=^eYNxb_@HLUx~pXfl#76b-I3v6RI7H#KzICz@8u@zvt$Q+6tBsEF(Uu~G&jV2
z<HAj<gDG;EK8t0z_>ZJu+5ke9E!x*0z%0T75@nLK@;MKOq~aVtrbI~W1+D(Db%%Po
zSarwhwQ&m)(dd&2w_7q;N+d^rkIxr<Y$}n}hoY{WCRW5kLi@NvEqrj7V%5#G!K4WW
z$~;4Xl4U>T*w;H@8LWY2fl)#@B#L1kmUH+P0RX0(D4w8cS;9NRJe84`6jz)<qV@ap
z!7TR~lc<|vCiLj%(VxV&2t1LxZ%a^NUR;JPrJ9k?Y|p0X&LtTw7YtT1zs9n+d-}yA
zI#}SG3R5*24&s%8^?{TNA3QIYX%d(tV%03jA+i)8Ut}t&T3+==cOanAgrMh@%5|}q
z-!vL4PNab59SbqA@IevKs(4n;;z4zNp#EeL2;dMBen9KXd-_T?MO)Zo2zzOvyCU>;
zL0P4u@4QhVqNHBzja?3F$4~O1+#17oyv)4PmU2nK=#KK;z(@a(a7tc#jh{FQ_a)*E
zav`k|Yxc$4+Hv^jIQ}#ymOyP@5w$okECZ2kl7jN8u2lp1VFwM4aMtPhR-he3%JB)M
zd_!l9n3_vp0$h6`Ey!l;xoaeNTf9<aorm4&!spLQnbu-FOm7*n!=F{6bAky}san%a
z>F--5_>{wYVc0S31*k#TL`{^%r*R`u0|P2TfQI}8_(*A2$z+h(Ttv+YL46Ee{Kk)*
zOI0y#wpdXU!m+04Mt7ccX$?7i?>uUZa3{et!e~gS<kRfT>lM(~i5614LpD}RHFGBs
zM3#wCIg{$?1_fLb&_n&9XdJN`cPm+w?;Re_5vfz514JYU@@AO!{(-Yo^d(qusg8Ue
zhST9PY$T@QtSspR&4?qzV-*P}Fnby1X0<e-jRW;D&h_m#kbF!(Y&E28?)8~mf>lJ;
zPjN|S;7cJ4<88VI+s9C})uZW(4A8&wL}_^hzOEB?-rF^KQAi%MvWOu&)>`!fr{)Ru
zLb{tNuP9nP)rdM#)<flr%C_D<(XP1D>i}b?m9d<|6>iuy2W%HYe~>f%Ugx8dW*|$H
z72Fj83*rlek%grXE%o!d_8EqIU1E#Mvd;cu+YHo@x^qqFhFLLl$!>uES>lS5tFXlL
z8ap|Vk?*;4!|X92`S2!#pwUem=7KF0cZ;{gVf1&Tz;9~qmDz>M6>HMyW(<U}D2iz(
zk#kv(_L@PhJv9g2!l<IqYy@G^fpHpxo1yS@(0~i{77ffbQqA$tFyd3GK9tkYrJ9rc
z6-DcapQV?16N}F$v>K{WV8f*LDqFyRFpoP^j~_A!ncAnDLvBNf1cV;tE5YK!0Mi`V
zev%w*Qfi&soOSaSpm>^n!LwiyT9m}O9vu1`em0+z;WZu1&KQSQZXB+yHlPAa0_krs
zl6C<5lomh3VmHZGWmzhg!)ORw8Gxhe2C7G*Q&VqUdq4b~{~VGmHZ>t(`i-4-<>!VU
zsiLZ%0Kl%^$8eZUtBM&mi+y%&c9pmhG|iJD-)%b-=I<W&YeKBLz+AfwE9NyV03O8-
zNnaI~C3U29-A62;$#K%%V>u@^|DlOTC(H!J%FS2mA(K%eD{bN?;XR-w`a*AR%ws=J
zynQbQjjojU&r{FBw6ZU5<N#JIP-1BzoupdUc2JI0rBM?Uez&PgbZ>5CP^=CZk#vG)
z!^s*7cP~(Iw~kjlg)|>UTAIZu*kYRA+I$!xLWo=1mNk|sSRiG=R1zSaaRAn;T*;Yk
zgX5Ze<a$nbF3mNz-a9b7knVs$jDJutRM|94>tIv7^qBUMj~kP))jKh~0sbnzJy}_W
z)e}sA9#_zA%${23nnJzD)A;WVp@z!c(=9_yga>2iek<i7KJvU`3g--^PbBqW%P1*>
zSKo{Ygr(p;7Q>K$Pa1%Egrzx$Jzle@;{Y-j0;a6CMqolqUA{JyX~ch|jf^;j6k%21
z*QBU?z=czte1$y~2$-}f;+l1d<&7}>+Z$H}%^2PlzCC1K4UUnkwlr<TnY2C0GF~_8
zCGkeCF}db2vI4A2AOzp7%C=xS>Tb{tC(S{8$(C9*SoRugp_;_|P^lrr{;d1SATw&A
zd4kyHULkEwf2dL}k3yvFM*KW4s|(SvVt*iESa9G>^xH7c_B7XBvG9KmV7!5aLkB_I
zRqYiF!bQ-7uLELT^SB2Cbm{9N?1N)=fv6*fO$ELh_;3YNcgjhdp6~{~@^cdQ!!^A^
z(Eo!)2fhaQ@gm}_)3<P_)UCjPTW1GRG8m->ri2J5C;;E@L<U?pWwp3-*K|CM3$;Rw
z^3~w@Jm*QF7JHxZ@8s8t@gMp&H<|~uwR%_@=oa%)yu8g|XA^f813b#p?HTWowRT51
z^7I9VHhaE>R@fD1AoL$+gxqS~YXA+p=~`j#RjW2^I?A?CJxE!4Y+Y=$bMqxf=Qcyq
zpFM4<y)(u+mCG45QK(y{l+J$Kx{dQ!a5vsbx<SRl3|84K7Ov7Ma6w?TLQAYub(^c5
zZ8HURqh5L#zf^7YJP%6p1-5On@VB(PBVRwxGSyUPE)CecwO%VV&(&ttMqK^kpmJ(w
zsm`j+Q)RGbxh%-gsF#ikv?8MNyTd5gcrkpv<>5VpB}l9_4$n`P-Dm0Qz+N1K+i7M<
z3K&~2z2#c0iP!!9*{wVwo^<RPJ2~oYgB9MN^_8bd;P1S!{3+718uvL00{|zKlU6|{
z<3HHn16+1HE@LIxA%?PRZcKX>!CO;D3We3`PW`@xCnU0dGM~$;5iA#J3z;i+t-f^E
z8crG|dbs4#0^fn{=hQ*11zP9nK~`j{_JyPo^}cy67Bp$0Ac>^%=A-kCXN^&BI(zYM
zm44^eGAu&GM}}Fr|E@skVXW{jhmOd{4G~l86Peq^L<|nNI(Te!1s&^tf^{{1Ejupj
z+(IyMyZ@S8>RI9C(3Id+*QzjiZs5xS+x&X<pqdoUL3Prf6QB^KMG65fPfK^gYuSbK
z&(X<Z^LV4b3(X~m`HQ3aNbFbS{5t6-hx3PJpe@P65hrdm=T^YSUp2Z;uFnVcwX9Z;
zX3$zmQz`S$WS`*G*G5YDDAL)NcirhvuHQD~A4tPGk(G$(L@6u`or$TLAFv|P422|A
ze=6BOnb8}BT%9m2qfd-EA~&z(ZL+T61aQ9-j^5&ky>`A{Qqej4TKN&h>xM&*>Gnoq
zMFF!R6Rw<w{wdoGsamiNYDXpC`+oJ2PSMwt?oRm@hDSzE!=ML3PGbnopoXm}KlT=8
zUv)OOYqtU(iH~z+D=FK7?t_0;F>O9}@8xIGc3pU{&R2Q4TO#$#%MA#0JE+1RkV!VJ
zJrlo^e2Plh=qCj=#Gd_1;>q~<QK~nW+z54eKrSu1ehLdUF{-{!DAe$-?P4azg?)a&
z<)7Fd{!1mz!oc)@p<I}lSQ!6Tm2|hJlpQhqKjxg;8@xy*Iw>pyMx7Zu^wSDaYe8wj
z<|1Cl@2~uYVEBUhmBhuh!GSvQJ!!<3xJJ$o-}vHc>E-5PshW|TE{MEt?oLlON>ALI
zngouM{evW=+0yy-=lLr-W%J-^L~QZ`h8NEd{vGnntQcrkzJ165yJU85w+g*`4CJxW
zXnULKG|LmUkL#%!b@GzKGiWZWBFFFj%VZ)=fR9J8y_Do>wZI1nX+x@C7L8-m*m_Ta
z6PqOX*L20yF|<#?5SocUOUiEGz$hu^jm8)tgjr>96hC#Vh$>b(*1KXDeegY2`Y8K6
zOT4pTDAJ(j8nKmUKBh#Ol`?V6nF6RH-E8;Vmbi&5Vf4^z%~T*4lv%cML?PR3g9(|N
zEUlH;1NlbE(TP{p_E~wD?osM${EUM%nNFUyffHJ{>P*RanmMhpOagOf)Eq(ppM=3Z
zB-qqBIM=+(Q|p<k;B?E15CYg&TXXm09`dUbE{^jJ4I76)F>t?|SQYcVwjd(}rBk_N
zvQ4dS1=3(Nx+bx<y$C1fOEZSn;<|K6%6ihh%>lu<!30}r`m)6eQQ!mp^-`%!`6I^n
z0INfcmQ%T1MJMWp^I<0@w%P<1N$c2`w!;)l=Hs*?*oie8#ouO=8?MC%G`VIYGuVN`
zS9u;1+%D=rJ$MXx^BsSO!6Y(87Xj~mkWy<r=;@YG9Y%z9v5JK^%-UCoY!Y)rBPNPJ
zj)Ordw1DJBA7AcE)*m!uV#ugTvG%jPV9fc!_@`AjL8vnNS*ih@<Vu6HV;L+rttV7v
zc?B`|Q|0Bh06Q=9FjK#1Z?fP*##OFy%K<d<X6W(M%o(qlP_3AMje}b(DRs*uiiB$t
zt%@nM0>+rgEBeaszpFjzphfk{LbJ}+puB}I0`Zia9J<k$losxzm<G8{HFo6;ZJ$_h
z(J~N`PRqf?Ndv0u0qCZIIN6pHHH^@+3qKudhwv%U(Emz(y>WScKD<25icvsos$0nV
z`uIlBS1AnWA1IALV_%(*>Xq8b^~M-Wy(9=7U;*C4|D-;cvOj3`M@0^ikmHs#!XZ*h
zSL@5|q31huW3qeOiRV%h+EH$bQgr6?$c-@$Jbf0=v!r?VgXV8D>16vWx3ooXEaVyo
z*@8sZ#aEVA@r{-)R@DH|<DA4A45QVXT4yzVD1JLupekAec;77Ds=-7tU~G9Dv_3gX
z(k#G3OP@Y`GMLG+ULu52;u(^gKMr<0FU3-<){1X%f_X~bt+1jRr#1yCvr`hzPIG1d
zrQ2bL%BBOR=piFK;$qKk8NhnnNi?j6n1p0#*06hppHOXUkhcV!;N8IFjUia}?@)Z4
z9?+a~Z~oVH2+5={MJJhO#qV;u7&B|=D<mH4;Bq|=r5Ou#wI=IHqtu0Owkt-qY<e3%
zStg26&E_~>t}TWQh&Qqh-w+?JppDlYAHEN~H6m7FZI>kYm&a`v^{4|I$RWcoglW=(
zKcUREkww8J(!Rt*n1ac@LXU0z9b%PHb69bMl|q1_4?f7u{782Zym|`$ji*)qFm#ko
z@h<@3D;6&ra72Z}YSeTJ;PpDMgC^+?)k?ZOVQk|@sv46{u_t{AqpX$M@5Vr;!D-c~
zIw(sv#;Z>G=Z%2SPx!?j#_D>e<88pjgZg2^OJDycFn@CKPh^5?are<OMp7Ptx&zC6
zG=9f!|Nn=vcZ?Aw+`4t!?p|%%wr$(SYTLF~+qP}nwr$(i=`Sbu<ldZplbuxRPbKy9
ztx<D~IUesun0_g?YIX?qoH|htZ1`CL$gtCUoO$rfHimlxb<Po0usSaF*VW`Nm<Dp3
zEFbva*cugGV?Y$5*xuX$Vr-uxV8G89<S4$0>pP!Ou84Zo9e^}rKpX;gbA64j_ims8
z<Fm@RLK%cHP1I$tUjg2`Du+SvCiH~0?Q*y)UVG7O->z(sixk77tE1SVB5{TF)BEwb
zhu;bz*N1L{^p*M1PyFDdcd48hrz`$D%Zued(YdWg+&}hJ$>ZJ)QfWFX%hgU753#B>
zBz#9~l_%mW(C<gZRc%h$4;5(~QBEsp?x4SAiZLMSVb58l!vrqsOVAhlGmQnho{@}t
zU~eWmR98vz<ticFd#Qy~(sRK#2$pBDd|K9}(A)LhMjs3VutqxT8YAUU*WhWG9hLcw
zmRO|gx%8Mtt?e8-l$VK(up~><PQy+uahVqs6hH~+ob>&)+i7R+kuNNax1K7TM{Ixl
zvuLm9)N=uq{I;V&tF$nRZ$mEkHM$#|x<?JE;x}`bU39s8zGUWYLR%bqY@w%XE+)Wm
zlOnuAsttIy&gpAH6?@iP5{6#hLaLSmoU0daHcV5S(aO;V;d-oazC+(b`aK-uY!0vt
zF8|tI|DpnS*)WZAN;R=pd)%{_52o^^FB*5w)Zo><Y;p3b_vNU+LOF@|(?H>t2!d2G
zo$}%go&WVaXwp>?y{8F;<(Kkm`ifKk7!*RR{9)?@m(<+0^83@E(f?`YX_xC(!o#i^
zpVI%%KVhjGL$O^j$3L6s(uuVPfs49lfDs%|s|w{oX9H4B<Q6&gaI-~{tg!+Ja?st#
zC$Vp11XPps)Q{e^H4GCtg~Oy+Sk-ozE)%h>apeY{zY}H(H8%~rX-RXGodhp|Q?#KE
zEw-<sxv&uRI*-u(2;7|r<35OPG8J`NPFNpEE2j$Af?%~Is;aG2PkC3#UUMK3zoOy#
zAl26(@0&XiYdo#1rIipwf5Bl4o40`!4dy+Zi;o1RGd;|D$tu*H8jf*Urcz}IACg$X
z$L>c$mMr#FFYZ6|4-ZVir3T4rhpQO3VS6jWA<XHNO#+kNd#p@jKijV`6Ds`c9*lVB
zz=?OhrJ5C*dV`D;Rsy9bJ~N^dk?d`rsA^pxXnd&!N%0F@neL)%x2u_@38!|>Z5gqw
z#|?i+s$}Z1M>73Q_x6-5+cX2LFoA%;^x^b8$vs%<azyVT-h$j-g8@}t{!A*DqVYmD
zpk_lIfG@>#nI@oit$FkHs>AUy4F5t;Gb+vNKgl7+|4_lp&i3EQp(`z^7_2sg?hCa?
zyj0eTS1_=^$YTS1HoD)*B1GgW$Etvf<LLjHERJ$%;D@lpuNu{<SB|KONJ{UJGI(=i
z^WpJ!O0ul2nb)#=e{%xzz(1mauwe3N_^pYj^8TcLJJ5k(df=QD1;xheCi8N6to(6B
z#C`%e2YdwC-J7OtmGdr~xIurPPc0D!);sHPZ&d;U37Ly|%s0)&{eFL&0guB#X<*wk
zbi?ZD^5XsUKwo3UH0?c`h!HYVLo_+F6ZANJdfeD4sR@-Mnqy4LB{+*gU9jCL&O1t>
zW}_3g=|`{PYV*f8C?7-To46$(li`8RlK=}}t}yeati)+&@o{sml~jOS2kgo{T{zVM
z?S7RrYOy46p(5c6)VvMJKPY%yc45_w?lQMLpxs|MdeE`}WKslC&)F$oQ&*pqdoPlX
z*H0wvvdn6fc>5%Tw9IvlXMZPtd(9d)$I9jd8CpzD2c>BAa$lIA?V~2X)W+oGS-7Z@
zf(Ap`7Q%a%yJ$_d#b9+GYt}T@!W17~dO`h7zL9FykP3w7ZV@)?PDYaH6(qmst+d^F
zJVWccUC5JJR*7ylF*i`={fi+XQBoVq8HD!5bR$f6DEHm>(vh+Ry(DPNAZf$RzCkY?
zMw!K)okGo_@CnI60V64x=`M%<Kr3p%Y+Ix&hpm{bEAZ%FLA`O}ltmxJAVJM>m^^+;
zJ$X=33G5cH6{z>8@@kD3K5b2s;RhZ;W`(d;{8_}T;(T;2BAFz=kd6zVy6O{`J~P*n
z6i}Sw;xqpJ7tG;wain(*(4uKUU67%Y{7_JBIu9*J3OK23Llfn^iv5e#X;`|jo$!Z&
z4aS*h=t5|V|4h0R%j06=ivpb~urCfV&B69>3dj^P7<f3F9sfFqX4A(2JO1CNAmIEk
z(<5VOWg@!jc7H@JyfN}5tyF8n`m?g}^h(5?Dyc9_=FNvp`-B5XD((~Era=ui&);$(
z<m=Ls8d~Ur52$IGI>!o|#2gd0G?RbvG<hp=!1-IbDG8NP5zBhvc<Q6TW8v6<{yO)$
z&kgabfWYA3P2CFfij%0&!$85Mmopml)bt4jS*$AJ)Ql>u_Y4gmgB$Ysvf9FAhwWt}
zovkCEZ%qaI4!(23O}gHSATSlWr36q8*(Axfm=e)b_%Y!EXWM3IW+5;Nc~DqVSOjq`
z`ylZ80%8v32o%V1{WcqZ56yExu^_lsd@+@%2<{E;x(RHw{=kt8R!aIdHNDkA5Kzd$
z0BnLFbdc5`JD`N*wx&_fK>upVgtG$?1}Y=c`;YAY=8|X2^@ruXVC@c0yEFZIa4!!3
zax>GK&IkguLm>!a!SMm^PxwIyE?a$1R30jIG2>ThBA9{uF8nR!PBZhXX4?6{kH;tJ
zE6ne9JTN=!Sky|>NnH=~lvq50qRw;viT;3_d>W=BiJ)2@(OjA5x<v_4j0d3w?OxU8
zmb8kG($?|<H=wx`y2ZU2gHqDt?AXKx0pXPXU>$Z~9I}|x19l7t-b@C3#Ww3_kZ+Q=
zBq3|cqdE7dexD{ZT=2YdKT(#&QIsux&oKac?l)F6rW>EdPk5T+US-S|PIO6*hfb#W
zh-&~GLAvRA+UM0DOgL^=t#=YkEhiHqg~qCp)#taVIygJU!5>Br*=w#<|CO@lP+1Uk
zgwsWw&_1<txJwT?UyvY$G!wm<AxTShaT0SuE{q1_lz3W#k&c(Qtl}=Nuw~EYt8N}+
ztGg7ZHgFtU3nH@~4pr%=!tD07CVBAmv?j@X4MH5o_iY<Dbg=Uma#xjdL6J`+9<;b*
z_vH9ZdA=tSj`!c_>==)PprJ%GGB+%>)-4C~!0!_@&;bNffjpbU11>U&b4)k3J*cTH
z{S~<q6v#_lG!q-X-#ZhO|Cd;iC@jEggOJr~8=&K&GPI|8J*T=kE8vKmNRirya9e+Y
zT|T?ccrE^-4KTjw@H&nRO*f0IcfcBi(v|2#2Bf@Tc%V!<(^<frY2H>`p1M0QC9T@7
z5uklpX_<J{1UV;Xt-rqRzsaX+^pZx^jxighJxECcGA8nAvf><1-!T8Q`|suANZ|Yf
z6T5I0Ms85zlt^033E70j&oq}n`-V2v1sV+zls6*y2-2M51llpX#xa+2nl2g^1i>V!
z0ABTgoHivKG54sxHNAV~b^ujlWdrQMZ;z>GL^Zj+?#v8pWPXNqc6e!*%dNrwW{zw8
zB)nTVY}b5sWVuL`JzMnAbfIXqGq9&w`dZ6l0OR@n+ha~!qpfi?5?M#^o)bAODp&PW
zCw$WE>2i~%!kuVnq+-GyxQOXwR_F#xyA6{(yUl@R+IJa2&qoi6{<N338U8Y|b~XCJ
zr$_e+iuWQMOTrAYU!!AJNl%Aozd{dqC<=(qgF~xF**fP&okvG!O-ypcn}g_96p_=8
zuuyJ$m2D+jUTR&Jm+jV|Wc$HGV0ZV*QZXJoUa#^j>$itZLy3K4Sp+gV3bFH=TvP!c
zeE<O7qn=e)?A<v(8n0wE143L5>NAIeSqLDQ^w1yh7WlRIYHy&Y+xu%FrKe9C_!~!5
z9+?0e^4AWVSGzC`6^HSa+rgP8_F)OWFJrcFus~IvbHr7{?3M-(PM*gLC~A80G<#}r
zfU3j57D|D|VftEEWQPEeqLO>*vNkF{hp`t4j<}_fA2S?K!<;GTU%9pDaa6;6iaEP=
z=JcYRe~ka0>!?dyuYo`!hf;^E;%}os&VhX3Ob%%&jKiwB5|kpE9>yRsH@9gn%yOV?
zNE&jGYS6bK)Rgwmc=fj0f$yzdCXUCJH@%`5M1t^#kim~&32XNGz9)+L5G~+ky$PGf
z$&Cc~fg2@SumRIVh__NUbacI*@_%GF0bV;VaDI71p6N9C3#{;WZ(*@Eb=gA=?Zdpq
z55Po*%UzfM>TY<u(N^T2;uN9`!oF%3VNHrE$amMgAQ0|d@@B7y|E=~7vV-`ia?0yi
z{A*bPq^jDB%#N(iQ5nxN?df0%*F{9)VY&ke+fKj4fs72@0W^n((o0f>!`Lp^5dadU
zn@8OZTIht%hgPF>2S(9b+7nxFgDB6mQwAQhwz?g(L<|;(1ln7~v{BaL6l_m7_7FBm
zwO|?uK<owaUN3%9zp+vpqxTy<;Ul{_M}DnwrkRCp8)8wC)^@a1s2o-~2xZW4@Tl#v
z;C>A4eoUq=BR^JoN2Sg<{ADOCKu=jlVyH!M#9Ug^zD{K;p4EawT0^b$C!oL58Jen%
zR9P(DRmU-=m&bZ?c=VJ+0?L=jr&5N0c5Yt7frZWK%wD&NS1uHy@3l|P{kXdK3F#nk
zaIVk*X(X6}<dF!q_K3vJupfd}1nviBtc~LNj719Y6<0eZ%u4;z=vWz6gAj^7ETGC~
zZHOObx&*{&j?xZI38kw{b-(&F0;uKTI$(3j;MVH1QCBNzrF=CKo#pZV;FqZB`SRrS
z@{Wl~PX!_}!a=aLk0PAY=X@0Sl?p{ITTitD4?O*sLW>|zEAo;6aCcrMmGv(0vdyp+
zoqKa}f7#YL=djj%xgT{C@N1&3oas=N@M|14+&=32BmTVLFWO#_YmxNghtxo(HFN%U
zqIGW3Y^#*vTOg&2*B(gu$;23~*%!N3mFABz*{Oo+KmGd8cl_;fc;|L)VcY7_k?toK
z;Er|d1iwC^wy-$aFK5Rafp?@95JtKqQHiLkU?wH8`?yX12JKC>1L@DeCL}M{dv&{I
z=(e>^uq4oSgoO+*y)Tnh?9w=RiJqt%cr|7Y&vQQddR*0aNmW>?AJBN}TBQFZ!kO6r
z@0lbE`+w)|FKPbnU3DOAT~NJ)ha4BY0)Y?^O_}*!TDhSa*&7T__rp2ZqP~40j}dB2
z*+gj?PO&ihiCVB~P%l3qOK^B+96r`_`D}Z$W2JUNsJip~YH$iKg1esh)zizvB*f{`
zvxk^7ThHIQ<O~8Iei6g#-7D8_3lBFsN^NLAF4PXWv)gU%b5RWBW0jH~wj(9;BSy~F
z9$?}BWE}_pbSpjIzRdV;N&Iy>9CwEouW<+lA=xGGo%7q&@O7Kjy`DhY<!0S?6|V{-
zipW&Iv}CabJe*q(bP%2+)LS3&KFc^g(H&V!J&3F>SUap_QLG@Ly<;ofIbkf}Hk^rV
z<57w^3H0|ajw`&~!(L2}js!tS)D;KjaMDP>Z~NA?RN`M;q3QKQ8*Y#Dn5P=$_n^5b
zNeX2#_n@81M@k`%$>FDZcGgaoEFnBmu6Gf-qC@jIQ?@QSO@$@Tw=eH`ct>kKYK#?N
zGiRj9sK?Z_Jcu(p$tKgxvFwT?`?th(yr>8*yst4Qo*%ywV10h3f;GztbtPJy<ef0y
z^we+OkXyVA3W|?BxIV#+Yn2OXs^ps;)*fzXSEN9#)A=Zxpurt?FCi?_yBF|~(pJt9
zmx}|bxy%+Ur!^HfNrTl22IV5tRv2BrdpUSm{T%i_rMz-n<Xp79OwArjTg2VX)1v#F
z+dyHuU^={?rcuI@nI7p4Y0dmy)}~M#Xe*gZtC=LY;n$uFbc+<pvVe!~lv0(WON(jH
z0C5$hN~4_Go4G&v=yRT~O1x0%yk*=$`yhlm=$?#4i`f7asb^qRE&{?>Y*aj{g_lO)
zP+;{2o3?t`R<#mV(V0FaUB(}74V8(-kAhX;utaaleeO<N{Etw?Nh0X%!r0B@yYs_)
z?}nA#;5X6nxmtv#pGa>;h`A%m^>r6>c5!_mp5#OvJV<bqlQh(8LX!g#DI8XwIIY3e
zKo`tvX5cUn0b?YSBu5=4|98oPKH92IDjXkSc{Pr(-86*!U6?RnQ12^cExXC<@9ZF^
z!C!O=$|g7?#X^x1%#)sThL#7Es(<>d_raEEl<&cXY@TL&TDXq$9Z70e@D`<@HK{$S
z?oXQPRcM|0V-rdd)<fOxYg3x9n|)B9)W<F~5x8`R${8nGjW6HlN#88o@X{4)RmbuH
zT#INX-Yxyyd<nSK8`58T$zzql8*pz~QJU<H1f+YYquPOG-XTx$c0JsNRY<>z$^F=4
zbu&>tKE{pDAx^M;`(@F913<NbxB<k!j82}J>nT1j83Gh%6ZczcZNYw;k;mEX+!ri+
zJ6k$Ys1(D;fU)kE>3SSJW|Zp|YQ!kt881EjwnLX@bx=W?pfG=v%3mo7=#zS@_+6{u
zhJ|+7#o#N?ig{vlx4z22`q(P(FgR3x1&5$*i9o_!9UW#dC1erTa?DYV<q%eIJRTCl
zTL(o_^(d~KC-uCfeYY3c$^(Jrjd;lTEkv*G-rh_h;#hc2c|yG9Zn;Ee6=rgXxWYOF
zU?1LM?U9Rr^9CnCN1;{9f^4JFM&?;sZ54D8Y;D5l<Te{|s{1VP^f;n{lZX1HZ&{jR
zsWMdg_Y5Eetymi9p}+AuE7iT|fwM=YS!2##A&aA08q<=ZzNIP<wXQy|tJn+)KJk$z
z+O4)0kCm_-yVsVj)-!-$@DQkX6Douedlw61GH)WePe$Kt_b(Uv`LkM6TkIu(aybfh
zn=NyVrN5TL2D${RJC)uIXlx{Gzd<Yp0HQELt>8NoKz!I}7@RAJIPee`K%xIWiXZdz
zj}L%3mL3W|Y*$SODPZA93jvB{opIidR;;80Tsb=(JGg=eDYx_WlI;M7TvfFU^P!-m
zZ660I`(yU#OJCCEM-BXH9ng}pv51K?&Z|4E%lflx%#evA;|tEkBhepKypHyQOt`jX
z=XAI}`pPoO5T%*xcTHr^Y`L1hfX=jMioZGO*G`FiRdNingSq@A81{8FKcY|b-aA++
zfmPJ1cRs+rRFvV&T;P1S1jYUvq-W1G2UF~ryjj_wEQbd9bwObZS&AD(l|Ww!!Hz#a
z-_K9KD*{SDp<iUqL!AKzf(H0s3b-_=3~&I&H;<8sO6#E#25)}%8PxA74!Ik>%eThs
zfXO#dNNAQ{*_s7@0A;KefsU*XWrkUPEGoE&L(7^cAJbS=YTn_U#lxZ=lFDXQ9)owz
z${x~q-`XAfrM|KqH4E275=)#{Ow<X%c22J?q$(u*?ggzzmI?Mie;d0cNj#WypsT#P
zsuZBn=0&kj@wwCE)7Gu5r<#JRI#>y+B=cC8ZtA`QutS>uO3l=%*X*$qVVf{BD~Qa~
zAYS+MeCFEk*eybT&^qBJvI13R;6}=738$%wQlaH?I@fer+SFt~cMGg>E~E#-3fqmx
zAQ_fbUggnUADH;-|BWBEf*;jll}r&_KoAOIk}*TG2yNF*+q`*>TI@yKy!$<rl=;j^
zI>U*`w9dhAi*3S7h_wD(R#q1MO&y4zudwk)h4NZRC|#7E$P2)-{0C%N$@sZhdu@ob
zp&f1AHw;F7u`#rHfP=U{M^7&L0r8ZZ=@#;#nqow-l<lM7ToyU28>J1yZ$A{5hE>Vo
z;5xoP`ZNBsbL^g&f#>1iGj(S{Dko3u;%JMemRin})uO{PER(McXs@k3JqefJRuBwZ
z^85YX7aj8G>z*oV&X%m4xagm}-7ach{zupM)87CC%PHsHqv-v6M*PpMNEy%LCo`}z
zaiS$98FLA$eFVS5nFuSdY<a#a7!ZY!FuXZo^G3GV;`%lgeJ4dKzUmvM0gF$|r6Ngc
zr&s0e*}|oH2q%$)(x<=mw@rQN!vwvmK$$3XUKisiSj7X|7@SGGzd~!|s2^$kxpb|F
zPA&_TYrN(+NwuBY5oJykb=yas%8GN*SYcybu-}g(*gayi#@TO0_f<bv>P7R?U&RZm
zZuJP=a}`gvr37$rDytPp)ZJa?dX0I_yS$7sNkQW;R1nCdA#hf;)o=j3bGmFS?2-{p
z?J7$&qd4etc!MEs0FGqLqtbhHcJks0_lK7l!?Jx;qya_|eX6diNsxbq4kFO;BimT(
zkBeUX;dwjS;zO<K4@^2M?u6tBor2aDiH?oyqzyxT#A1Yu!<v%p)8M(bwVsS?6{9kI
zYF+kvM*F<aqA<HLw~?id=ee5KngR;-apq37_O_dT^gxrg**ldc`fXU&6w+$mBDbnw
z+oQFb>gbyVqF1L~p$S)i+u;soA1D_s<`))oaM9!f(VD7F7Ebls-n9iD7q-v*s>c{Z
zJz-gwHK;~L7v4BBo%s-06)u02J<v$J7|uxdyClC8GKbq5$1|N{DX6dEY<0c-<g%I0
z=8wS5O+!Pbt!U1QGfiNu0nWt(;XU)EdkL~ZGJ4dK)(`Ml5N^_c(wHnv|F_2EU}OES
zeDA8?=@WL_-FD6BALb!%Kqy$XBxN8kEFrdmP=A0i@UhwvaNivP#sC<`mU(d?#icP~
zj1S52qYJ{67`hfYx*ObKS^?pb09{^h4P>v_+n#v((=EhA#O>0fg_on7U(-AE1nhrI
zZ#SS%oV}H<E=&}IkN^y*Zt|yR%*ysf4$}RaX-9*puIoI!+v^?ODrLo!B{jEe_mmc2
zulI`xVc{JfZS8zJ3b%&fqCcv=->_1C`UbAht*d`-d#mAwvH4y7!j$Myp@Nv+yBzzq
zST13>R)ad)*s{zU5l&mu0(b-10hXSl>1(W2al7`=a>QhCS)671SQ-_w_L|r<G0cgD
z$Q`*A$*3M->L#0CJ;gE&w0ZB(6JHAEnsXjdfb1uDp;{m``*Do7>b$yKJE0R_uzTPD
zDIP|%M)+4yR#4xc%d*}wT#3pl$)1d3|LlF~XX4|GTHaWT#wpH)C!wknc|FQNZ!=9J
zuS<MDg9Lvn?)9^~7jaaNR&-N1PVCrxY8ACOsF(G@Vd}0!9L?`1A(5I7UXX4*TNQ`^
zJmH#}W;D6XD)+nZdKC4mSVUy&R1~9@bI%`Ofr_9;;ZeylNR+mxD9Wg!EH~}tWe6Ag
zSxajd($h)_Lgf_degjCh^n6w=)D(nrP`yLL=E|hI@uw5dA>Je`Z;C~d(qdr=EBb`=
zGw++8(QiYdTuDabHt8K0mgj62lEAgzoFz%gs^euZR@gr~dOS6B$^&)>w+4KjxX(3;
zknK%cobVkO%T=y`l4Y{YIJl?MOg&9)KnGsT*`ms3+!iCfF-sI+dMnf&e%OCXsgm;?
zkRLr6C82SKxF<dE*$ya1pkG>?^m>hl9#>Hh&+G?2hWxSMoa0V%NNqXP`^Zc>+~hAZ
zJz|^WjoYDXLMAh+XYv|nv50<Leof|KeE2}7QHohJQY!#u{_!QT)pe{P$e<U>wDK`X
zoZYBsx00ZMQ7msDmB=~@lNHNLJX*8rBYv&~aektvb7Bn3{hL4<v%yVv`QD02g7F7K
z=7&-N*HyMh$3Xi{{=jy3?eEcp!m^U-W*Z9n*zt}<x2_S6#q@S3&ap!xtma|1M~^k|
zcmzmfqXJ|F)=$#bOPm?Y0+clt8E#a~xSmNDjaz(*iwrHb*cmtF+XN0xzGW{htS;mV
zx!no!?b|zp`d0B|x=yeS2P@3m5(>!}lSMh{B*QamEhLVwZ}d+q5MhMFHDkT>nC_vx
zr243nVVu0imN0wicB+%8G-p{Md9<06Od-Gv>9$4dE)X2dis1cW7Z&n3T=717;k~h1
z+GVcvGk@CjW|)M4Sq>aD%z|5T)E2T?7)4)}=v5OLzli-LJo1I)jy?7GXo4f<p4*xQ
zoXUdTiSo0X9e35mkf#6YGV_a#DyM$Cryv<}SU_#x*`{4B1DXRTBT>f}{LqDR&TT6K
zXjabEMhprZ)tZ?2iKo0pM;=a+Vi$@*6gW8R0=D<yJ);#@x9Yg(_2P#1N(RYVHx^0!
zfThPTcgqvP0NF`iGK2e8AhmRP580@kx&+#u26x1<DnIRLii!4iILPKB567P?bsh9|
ztNF{t3AZQ2K#gsedkq-$qPtyaZ;!gx&LoNBmLJ74-eRP5BCNC#izXSRD0{<e1S1b#
zj_V3|6XU}<u93!FT&n7sI}h)ROtYCaLAP>YifF5MjjX*-k#MG}uqmG2-J4N4^!Mef
zXOe6o#c??F+p&-hKS?sU2Aw6QssHFPY&{|!F&YkCgrYMr^#f|hDm5#y%6?&auV;x*
z^qiKk6tUU{Zi*n5*8a3iHRlu_aew57PnrQIio;Ki!hFzO=6C>U=g6)_aB;;potKlV
z!whX$FKRzfp2ghtxvkPG8-+>sJ}LQg%g)VcKcld*zK1Yy)qgzBcZ}}sOTd{bKXgMw
zpUHwNEnD7U)805KB!!r#j$l;Kewl0L7Gne1^BARK-uFluvdl~5pn*)b4dDbXg^H=-
z1WhY|e27j5A+W*(!Jrc}y8Zkb{4cuwncYwCA+9*}4k~bx+G=fIFszG5DyBNytyM0!
z7sfR{KJ~n`qUeWUwoxgYO-pPxHpWKMaBt)aQAEgA+U=1~1edy7BKiU)la#6fndR8~
z7a5?fHishF#eLl8EX*F^Ol2L_rCPyEbQ#$&j#g#au+C`_me!uA5YE>PINxv+HWvP}
z^33gcmRU031X@dXhC5m}ZUQ&~5aoDI+31t+?4~Y^rm*GbJ8IuK#>AJWC*zBiO-w|#
zCjPUV00(3RZ1yj#N2G?h`R+~knM`Jg>=$Ha;JnHqTVOM)G5U0tjeN7a92ewfhCwP8
zf9R;(55~sR1@yh^Rjw8>LR<4(Fw(aQ|ElBu8)=028DB3B7nKqrT)_!u^}u_Mz0Bol
zo?}0mzSz>T>#2G}NZlxqT&*9FU9Z4<spnDq#9A)2@jo=Vlez<xZ>VUC^HLWuuL?5W
z(_$>GyKt;1w1{{3B`FA)PL1F5=-|&PXIN|-uT`pP{+LB=*t6oujP;RG7IA5yfUx_a
z*rtAGI~`l>%JD(0kP{KY{4U3ItFm7f!fxVvxES8VyaO1fR9_%mxWS4^BR@-om~=Pm
z14$`i9QJ<bH0I+phal8g%LEfIO^w{ly#B>X=li}jZ>p+S6pkG^xuN^f&^^4jr<l4#
zqmT}_dD~G831+;tjU#l!Qs$!n04DAJcv+$Ief#V{ti9}UFW0UF7|Gth@9O4E$?}q(
zVDpL*<fxg-7mgp^7#b**qW4E__Me%ESOU!L7!~a~nj^Mq)NB|Mg+v0DwgOm%--~s%
zDge2`!XvKa{3oa+MAUPnV`uFX{M1O<a~#^UpsfBhb96Lue~;y`<bDGz9j1Vl5`!ny
z=n_$(4^N3c#33alPhJvRNmflv@7BUJer6rWxiekYOO4Z~L7s6GFpD|G+<#h3uo~f#
zOYojG;`&NI2Sp?!k!1(D>`p$TsHzaWS%rP=>-!yj(cNDjbJ=qJkS>?iAmy&IyMb-S
zc&S2rdv!?0o@MgT-)4#c5)9|t>MVl=(7J;fB3pFPg!H_zFzHvn;J-MX#*ibyK*|o=
zo{fp*5>Jrdr)5yNb%o3BtkHFJ?Z1FLUz$7;RMfiQ5dn14(kpg-;oqnI!U^pft?s|I
zDm{T2f1!xK6SLW-0qp;|vW-TMPYgPr=Jwp!v3(xTjIC3cn2B?>G(EA5L3|h|ID&z)
zy>r;FNIkH-1cOB9yi%G^--Q+3pnhqk=FN?CGt2=;fl*F8!?I}SAFU&yQZL5psq-Wr
zb6IEkoy_C1QU3f|W1Jbn@Q?&LMOF<+RYidQ_<%uqgLLr?x@=@<?Po#CvSHWt#9zYI
z_h3odouN?%-XEKk($@BOQYF8;<tAPSyy%_qHHsVRMD93-y|G1=WbuyKct4fqA)8Hb
z;KD%0_7pp^2|uKv_T??S(MX0mNCv&cU#!{N)$pe^^T-JXvBZ%x;<!=K8*3@=FbSCh
zq5(|($fv8-{4blU&$#W)OqMUWxQ23O-lW4qw5E5oY`hGvGnex6n@YtkjIur3<?0o%
zR|I**h8=48>1R(<Xco^EfIar1^|KlivxR7jyq~q;6&r29l>~G=5`qIKr$mIaQi#j*
z$Jz8?vy&e;pjl;M!bQ)A-;wAExG}$ue^@`6ESM591%>YNHOlg$-N3Vm$8&Bl&iz&C
zxSiO!I6&+t6sY^`KjsIO&?Sp&0#H7hY!z_xziR{0=h%B`6Pl1<YylQu3BQ}r+%8aa
z*JU1CZlx;(s&b8)`bPw|Y~LVMT-r}JU1>ft)o1h7qF7qA=5niBu2|Iw^@?cjJ<jd>
zuBTg#q4!kM8JSyo!I;Av7>ryz+BvX)C}$I&<rS9IW}@5GC|$u89H+*nJcC(0FJ&8o
z|6O?oQh&YiAtqGD{Rg$r#QgtCx&HrhG1&hP7lV<F;lC@dUDDLC#c4tCeyQ0teD6Pk
z;b1b_E58-XER+BkaM0=qfFHdNu2h~wDXbX(e&K2GR7>`%$**Yw;9E`aIE-&NOk*-_
z-C$ga6`Uvn-G>n(rf2Jk2hM0$5R~{rfmB?Dg|*lZ2~MB9A^be?2DuuJ(Ldx53)iEp
zEyOG9w^*uC$(hzd`EsyFn+*=}X9lHgvMu88=PhBRz8Y!ZPe*n{7j51y2<Ni6i&Ft=
z+Jl459NbcdmR{Ogz+``!DWkDF)1m=Q?p}h3nY@m$-=!Qn=_Rjfe=#8-JO5BoKhM^L
zPK%E*#ZbMw(}YxC@)58TtM2Ai0ZOiUQ3lT5&H;u^oud~Ua#QirZH?2~Syv<%=Cv)5
zZ+#_~5xN^QIFb)<Ne)uIi<2tPD}Ahju{>Q7EEA5I4d=pqMn-*Z(D0>WWkz^6u#r%A
z)}F^teBXmXea1!A981a7w<iABM^zmWd|F1O_w7dxQboxLApiH|3J-4HZ5He>yT}El
zd`gciG0WWuq-bo`l=}sQ;#M3&n;I3kOqdGnS<>{59jSo%QiWb*5MwU}os3<fUabG+
zg1kGhYH)O|>C%b{uo)g$eTXbQIWmDfiP&6Vw5ia*r#0P|{?-NH3L5U+2~uH80|Gs2
z$?taLqfy`Qkoex9tjD3=ic)b=bxJnQ8ysZuweJDJavnP1TlTba3o@ZFzY9wJx#+`y
z-8jJLDQ|vJvsXs_QNRH(d0sFQBlKeBL(jL(y}p(e97*nr(w@Q<F?9J#e|!7z5fg#O
zipCY8k-OI)@wT^rcW^E8oH|sn_xA-zU9`4&FI0@yVCc@!u)x!uduJ<?rnx<3M$f6n
z6c23!3oEtV^XtqkWbE1GcF&H1OFPqT7w(`dJ!?`f#?_PznCP`g&-CWSHG7)R#{Cbd
zN#z~uHQ#X7smPbpR>Hu*8~v=qyO<V_Oxp>M3=J#>EDo%r;m2m3Pp-|w<5s`#SH17+
z&MZ$BUQZ6r3tZvV^~iO}^%^QHpYD_kniest0HlCL6y)pJX1;AOa|FPI%L}po>(q7V
z+lBiJx7W%nLI(&yOD2fb@Goo@3r+QeRyHyHuQ5omxwJlCBo@|R?Q&t^p>U*JP60E2
zS|t4bT%hhAq;ABaGg=8JVF&ZMH@FRCga|5Gb?ePa7JzOEU@VP{UdW<S{;s3cUEq6J
z=09}ND%L}wx2c4!fT)ASgzA)MU)}(oVy4MOe*#YA?9M3h63PAVdK;4bFsrAgz1_N(
zO7z9vGFpBBo&6igv9ocpt!l3LAxnkIl|rC*n=U*F5r&|C*OSPOka_y~UOqXc17Jdz
z#9vC1ngyx;y5^xw<(P{&0ganAjetA$>CN1cP~Niy$w^zn6Ye5d37Y)oF|*6ZN&LjS
zYSqHK=AFsg7A;p5WRc6pU_Ii+51+3tyT-=1wYn#^|LDSb)))fKG5D88iQ+^xGgC)_
zMS<<$amF^+wUOq7jq*?X@1k(>$2qfKKsn`h)uUFyTxD2*d6Dg|X7(DzS_9c7>UuhW
z1O}W^E9R+$08gm}YO?-?VBBq32a(<(7eo_ikaS|n7LbDt*NFAu_pZ=|_>{Naeanwk
zgO6eKNAvD&OP>~Z@N<|zuvfedeK=S2Q;V|$yd%8DVAL)iWh;;KcUu%x00@tiKF~z?
zH{eq_#SVWAL!w(}9gxW+ax5|g2v~Tc%U>(jJs)Zj@LQIj4bO!yR77C}j~^S4riKkz
zk3Sl)q{43iZ><FkSxx>S|46S9@D83{5(olCs1HCL5gf}2z+6%<c~ga80jmf?NlEqU
z6t-2Ro$yWPgnp(e86IytkJL{Dm3Z*P!?X{R`-Mrdi+CZYqSCm8JbRr8l<qAH=q&yJ
zxvZrDqpoK@M?dQVtXPiBN?jB(Md|D)#A@HmTSGR6V1`6NX`sC%knphrf&uuIujf8*
zg9N(B#|U%@`+b)nkr0$eV651HYt4bSO@U_cK}{@UpD#=38Q#Jqtr483(gP}tSx?JB
z$|dkOxj~Nzt?-X$vTl1YDDEi9;FnQ2G$ny&F;vh_&BDI`=BWBX)9C!oFrA1c<<ZVX
zkNhQPU|a&51K{mxsQmuDrwNBOHHNOiTtidec|`DB9fcrgc5Pc&5-WrtGR8_G)qDsS
zHA{JI{QkjvFs^s-+o(<KYR&6Pe3D!xku+>aKYy;?W<F!aNH3G>GuEXF80{a`NvQ*C
z-eI)^9yVigrj;cVhDS%T2hLL}VXMM4u!*Zucjc&v*ZGX}sZz<-3Ut+T`RxV}8F?Tu
z`>em0NWA|IeZa&aScr|UzFC8|>KIF1!$`AbBt6t4E@L~B@X0jTkp`i!5(a4M5C*t#
z2YEGj!Ta@Gx|q$F-I(vq66a#)X6I_>N?&4#RgDNrdH5WHHbq1bKqVm8c<0ldB4Jgm
z?CQ0d8hufZAqgRop){b^MIg%51l#uM6!Y*=sv}GbtddbcIovUljer#7q3F-wT+cv)
zEE*MSoCNFI>^5T~FbSabeaYg*+I(Ap8G66mWl>zAf>Mlh-T)coK{vrJ322hS?(JRN
zh2=>hD8PSvDW;1gmi>YD<t_d3M#oVi81onv{s9k3%7?)O#i!x1^TZ;F074bk%|$3=
zXdi(xN(Iypx)2+J6>DYbllV#7E06{}f=$8L5eFmv)1hsVWPaI>l2%X<#LpqT6SrG&
z@IgoG^@9%@y)O-`=#dg$19HL?DcRLB(UN6jzwO+#)Ih7qiCn~boUATN*vuuX*QkVO
zwJoH54s54eFkJ2^9O=(lAfk)V^9|%gcb`o7-xr6RT}U^#kpfukXBH1bD{2o@I6yH=
zT@+NG9@uhXD?XH2e*&(5YUOtkh@=7wvHn!uB1@}mxFRI+fd?za9C6LoFp@aD&#-K?
zEHbOU_ZAl&BeA})I;ur1d~Vkm`EX1~+p=T$*dbuI2}!sHBKb!#jRs3v^ieB9EdB{c
zDAu@ucdfJfHCRQCC+p99As>;cX}hyntU9FXF3=lbuw_*;ZBoH^Uw@7oDjH|z3baf?
zb?1u4hLU0<){~KBT&WM0tzY;dN;*<I^b~9j0(kmFLG4UIEnPu29T9MH9AM}SW_NDe
z6GT4~IHm~CxF8>Gi0NFero-H-FBq_!P^k@Hd(i0h*xPMK#Mc`^?$eAK-&@ZOFEy$d
zneSEmj@+MTS2!2Am;Fz-8c$dHoeOpb=dkvIgSG3=n1ZjsHQR`Vp-=QL&fJ^16al-g
z6^%UaiY!l^T^jZ43O8>D_8h#cpJyLduJ(eK!rk%n9Q-=pLWER;aX)@gZ-U^h?Tw*t
zJiKG@oc-<Z{eiQNR=yk7tufPq{o$QI?FmEE9f4~&Ur%UFOBV14|2#+@(eH$nzH}ls
zw|5#E8eYwP+S$F?BKY}Tn;TZYZeM=Bu3q)zUZQ`#o=^0AygWyenh?K~coE#r^{O<0
zK9+UeC9<aH8rW)lcs@oxZhU=}t|8P5!I5?P_|O^d<uHb*M3BMZha~XmCo}DWZS&3O
zo-qH!eICVc<(BYYdFuFh%c;RZ>|Jo>hg=7GuRZ*n@T3a<2N}%F`X7eg%xwQ%uwqJ!
z%hsw5_WFy8R0nnRcldS=iVnuzoegm2utzJ7V$`xu`0*1VU*kx8-g#_5Hsoe9Elw<=
zh$j)2T(Tp3|F}}}FtYFxF`R(3?Wa@{qbJADC;ko{kE`kE0MKv2?)~ZWKGKjlK7S!i
zm#Apz!SQ~1w$<si@V`+FdH}lJ96=}Cq96Cw`|9Zr$>ycd-Ys-H@&U^B&s1HfY<6@%
zpRXIg3ty55R-foQuPrYzB|qHub#s_|eL3GxA4;EqED%>q5`a(;*N-DTay`6XSqWLG
z66r+JpF%m`uUCWFpVaBl@yujZEX51h;)-P1m9T;%IrYb8bMv$3agF-oSmIP-J5~1|
zLj!~&iAKN!6zAf1G(BQn<t@V2u>)>-=Kd;h^)X?Lsf1$h-}h7FMincc(DBxh({zu<
zkDkqWNj!KWlw>D$ryv$;Z}<`>_&tolovQu{5l?c8HZMVmTs2$}{D+BqVqVVa87Amt
zm2+MpwA)U-wH>NUs<{sfl-?S7Tqa|ZA{Ykv(yR(18`Y;W$uD?mfy9>L>(0ksU2qkm
z9?@EyjjDbsUc<?zqb(Ijiin_?yV{A`i4y+do{txc#nSN8K?2sb)Y&2exg%mRTDb~K
zz}j)WldCv#iGpd8%gCZz2{~BX<-a}g7)0ho>ZQ&A?O<Xr%Vq(Hs87HGi3*v^wFgp2
zUOf6T*@ERG<}$p24o|3F#+sWi>(0QMP@cUU@%Wb6(soo{yGU@NnDbe7<JghRdbxHr
zpsl?}?FbDrK_Hf{P7WxnDPKe6HEG3|4R~K^bp_`!V=V|=f{$2yrxq*AEkp2EpdKS_
zb?6Q#B8`MWbA+#M`a*P_PI`XZkqbaxc$;A{+FkYiA9nCSu6tD_#nzA=7TY*=HWE3;
zW7^y8y9}e?qCrEYffVPPwM!ADID1J}B>{G9&j@Zxj2+pxn49osgzB_Is;F+=1ev-C
z@dA%uLj6=iVX22M+#o_H5F$v}_9MOUOgt>s8+bCfh&oWVa6pmunh`S`5e5Y%C___7
zHg+)Z{`dhCz5a`w>5}fW(pWv7_l}RavI;4Ue%3RE9WJ3FmZ!!GiqTh#`eEQ!F;L%;
z_vH~Ju7@G)m{H*3X+-#J?#?{n!}HP{Y3Y+ZTsO!fR^OFv$dFoR9F3bbt|^E|X~NWG
zt<{)`XWAi3izFILM<VIKnlQ8L_?y!*1@fA(n~ONej-(F4-{c(;Cq@%CN57Dd+{8~P
zi)*jJXG&&qnu!LJYz5bol|x;@jy@))Cjwjbi2Zf0JYbk_9iry8F)@a0x>0&{BX_2P
z$9~N&%7I^S7;+zc*&|G-ak!lkSlaS)W8KOGlDEu(f_o=&-=F!Bg-xgOih3&!KUF!F
z8j*4GCs-754lUu5AQ%%8yb}pjtSA*j<CI5dvvs<Y;pXZsxv-fRXvJgF0LS5%HR@oV
zjfam1Js%1sJ9j8pmg=1Q5<S-G_L8jzZ`qjk2`6oh#F~-&!)hptyx+|%5=z0qiV~G(
zAEqRrWrl9=bFJ`&%oWCM*#i}xuk>tW^72Fnbj;HiZLPJs(eLD(ihFDwD)&hc2MbN}
zU4S&$DKOT9CQ#DyL;)`;BEgn!capsGO7S2^+|&VY`)m-;z&%2Ae)&l|=&;T7qqun1
z*j|W5e6OqEFgVrGyY3(#V!E&J3q%ka=2Y;cph(Tma>)_-ayNe9!*-t-Y`v5QpR%BA
zIsTeLmkE2GgI3jorOqxqYfcPCVloaoZ%)6;)Cev*l9Dc@VVT)P;g-FG=qA(Dhr{|L
z<WdL4)P?W&ASjA=keS>W54qDYNjytV$ulLLZuc?2m5Et@$TIGGriDn!g7TtL@;)%e
zq%M;+F?Om<Ogmb+WPx9{>EG)k@&h9eoQd2_|LbKEQ{I;%(scXx>!#hjdcL?wCB!k{
zA$E!ukF9FowA6)}6Z0k7S>z1IQYM_(3tsG9(d86wV2{{RMB&pone1mE)<Sgg$biV|
z%6SHA?~BZWL3sDPQQ-P0CU@<yoLSNKFu`KD8Al2Y1tx1`vrQ(2O09f(^=`SdGutvf
z&@4s#W;b&e6Ago`<A(Toy^Kc7--sgiIF}=d9IOC35?D6(LF!|DYEhvcS>)xh%&ojN
zg^>N=bQ`B7ziAW{+nD<fGzdz`gnp7g`!P9c+=;tD6Sox-utR^xX(Lqd@NAz$rc7Im
zL*8B8Cv1mp7r30^IUC)`Upv_8UL{1PamOS3{qmkXWd6Ijnr6KWYlqf};>Pj#M!$;r
zZP;$->xsPV$N|Z830^QSSYr)9?iC@C1Y)P0o~repccLq6oaSrMi>Ez9Zs1NdNsehY
zH9$_$Ug$ozDx}Tklri*hni3K{R}=T~Ej#r$?J$<HIn|^@=SA=)3_0nH+SiQ>5UFHU
z_C-jLasWLga&o+&3@$s|S;Ss)+|Dc2Y3Lp10xwm)q<V#iTS#w}$b0{8O4b8^Z<TR%
zO5?>W*GDMG)C`w6;9gH9ITVzXK1Z7j9+ftELnhp9r6jpIjFhHqN+4XaE3~sp?G(Et
zYrUPixtvWolp4_f&ByyNwfo406c=TRe;6I!#9q(HKj1hjlJUAHsG)SHrGK$sxXM?9
z0vSlR8v);cI-)5_SE9^dvv0>k<$j+gx^kAUp3a10CencjLeXxBF5cEfxGyf4Ix8J%
zDGXm#+fb5V!@c9QmpubI-(q8lFN1zso$omS8c$dQAO{=p296y(KwdvaEj7HIIqj*z
zfht9LU2FbJk8@NTaOzuBV46y*a?i+K7Lr{sJx3x;*$JU~#!@yd-M24a!Xd}_gLD{U
zPSCF7q9YF0=J1KH(sFZ<d3NTxF*sK??+8j?BuAoKG2KR=6R8RH!U)H-#QaQyw>W7b
zBS#mUjwff4cJ{TkWzUuhk~KTq)*yJ6lkvc{mU(!3oH)8O!yS;@r{k;c@-c?VLMZdD
zb&I;N**6T@Y5Sp2l-69W)1u5%L$gJ#Lb;@OIVN>NH~T*3CC(C)!%<#0RuNZpH*keJ
z&=FY(qk?TRCM4<{3HnzgUuwshrNW3(@cl!l+M!a090za3%4|zJy8{kz3qAxd?hOz>
zc=Rm?pLR=PiY3!+4OS7_shSIe#}@qw3a-fO5c}QQi5gh0%~-zQ&)2BM;%xf@vc|b0
zxw9DW^Pf+r;`DRSisbP9kFa!jm&+0tr(nZ1PoY-u2;13nkZZcrM_(x7Y82R@=d<xL
zR2a+67{c?*hR*B9$H;;&V7t!Te}V!tv;7A;7dsREe-(9T(vpc`E`#l!sX5)OSJc(z
zolIf2DC;ogRJ75Uq8X<czc&A~_T&2x5ICg#{Dxz3Q5gl@kdL)F@96p^YxC<@_&z;9
zSm~dL)Q(?lt^7a5p?aJh5C5~|e|-_N)w5l+2v<X-`x~ZpbN}Sq-MaSu4bukqqeAtN
zH~s7hbfvJ8&gzsjubwEIoig!ueEo@)JYW3qL-SEG__?X^?UeYd9DX|AbH>W&?bG#j
zV{Nj*2@hcKlEb$>4^A(cn(He0dnoOEz3pt-+WusYHUIkA2&`XyKi{9;Y+Y|j#Aqav
zvYe(@_#9Exx%62P{e^NO4t|KWP&Yp5YfpWc#v}>~wT`S1fz-ah#<xI`J(3z5k|T&m
zd$5Y%?r;_=oTAW0Tio%&Fs+dQLD7vR0mu=5P!)h*BXcQU{_Ia|^K8gOEv;D!hlVGS
zWr*4{Tg3Mppxb+fz||F-SPY|3sxBCZvmE^xyUP;waZJ(0H15g)=C{<5Ntxb?;M>t&
zLW(n0`_|)!uxVc<E&10b4{&Q40wC-!Hk(VKMx8BuNwbqBE+82=;Z;bxT~=>S4iW4D
ziU0&*C@}<B;oOWY;%r)0AU@mv)zuOX%6xLmDr1Cnmt3Kww_5D?H#hlE=*`GW#0UL@
zUyynxtYK>gN=%FBVzXtE7)izp07H`h2#tGa^m;l|ByUoZzW2x0M<?I3*epD6<~YrG
zgofIDpt1rrYU?7HR97DtGxMK)-vj9&BetoCw!eD7%6t}cE6t*g0UA9k#O&az@|ORG
zD!ke^8x#0i&A#W5SPvd0gFT{2^x_Ce64xLtg1{Ud`SEDT^;v%#Wny;ynTHoMupm|!
zx(wfM&jG~D?GCv+S*M=-c&eiLk}BJ|uGn!ANtfXl_0)4%680WRq#k~Qyxt&;7F8=G
z?g3R(ud7bgl@K~;7(rPibW|l4<0yV6QwBFRVMwSd9l#^fok?~)RXucev(<Y0$D4#Y
zlN}WwAkj<hF7FW<$INVSOTC+F;X1H=S}X7NjX8Py$o1*1bxVb_oCM(VB4+vEsAPPw
zx*o^mvx8Cl(`NQPJKv0-u*l<f2t)i3$xPJ2*OP6JJevAVJ}4Zo%i9DUT78WPPYRQm
zrPVV(Hpxhca+WFPGhgHQPv81RGd`Z#@|tc6(ck*eiNjyUj!}YKP5>(3%Xxd*rqdUE
zO~w0JRe~Dv_eG}v3*I_Vumy8|HPP{Mx>+H7{$OKNXmHQbfm+UnCQ~M0?Ih`~Lb%I1
z*8S|0lt$LZ3b2C1lZ=Pt{I4-hoTZe}QGjt)R)lBjqOhY%IvN7ne;sk(VrM{u{{)&&
z9GT}T($?Q`$qmw-jnXmoP<7+~nh!yBM>WIliqwWGt5W}D-TZ@Oe%s(ZC;TMMxE$}a
zuv{i5OQG~H(`Lc_TH`KGAu1&v=4#eypp;=fjsypok4jzTyK4jpBdal_s5~w#o-mw6
z*OU)ItxF7lk#3fKl&A`ZfBkF`iZ`^R?|Xdng)cpnP`afp)8kylb0Ysl1w<kOe^F@O
z$tasRL_>{%F=IL&b{w1KEmb`(u1IY?eyC1T)S8WmHfGsmWjnH264CLC;|#W5RY@hr
z;`iY0u+G1mJ7ATaK*?a3uapWgTUwWi7HKC2A0y708|I7dE*LIOYKdGC_i!2bCw2@w
zfo1lSWe(@FuMZ0i9|h4>&t^Dd1i2H~^(kF1zf_Tb(OGK~2rbw)yK>bve#}e{l#TK1
z3@pwg=z9G)sA7uGnm_{8U0Ur0ODu_888<iuJKZS*MM%ovp-$;^gq>%}&OE-*Fbgn{
zMI5DHT(s^A=|9HE^N0!ZDhkS;vvkSyk|$M}&KkwySHk{)#a!et#jAD$+?itG8`fv0
zpS(vxCEN<X9kw*nRhKlzTSZ}OPNkL=z5}0nzTCiC<xIusrnVJ+EU}=O;4NN9e%vrl
zJB;0y)RNRvCS$g<vO9q+R5dN!BE(*veMD_iphPnfrII}Eo($xaqv;LI-i#@<g8Jlv
zbMh2nS|Pt=IrY;BD^Ma0RijVuUx*WUX1&9ro?15yq37NdJ5=6Lx?O!qYwW`ZHKTRa
zVSb|p;o;t4FtL+bN{=0sra<}*EQRY}wj?}?8uwTX(LfvSKIa@rC|gjs^mnC8%lLFc
z;rl^0En!VhcawY~Lmw*xUzw13E&{@XbqHofWZ$!=o%Y;9c1^~PFhrNcbqm(M+ap(s
z=(NhY%!_E&PVIyLr~oyVjC5&45@lgX6LG++3%T<x1Nx~>(==0*A?~Ejb!ohDD0pc&
zE&q!clDm&;o%{Hf=?GJAeZl%u;mIL?|BzF-mL>WYLio~>auo@@W|>c=sYIdi25#(q
z<1pO4M%AVG5vt3Q1OQ7eQDzE#vR2F^Np7}3@EwlV+|VRyhJp8YWl`i^Tq0XfUr1p0
z2?{q5#ce84k&<+#lGb}jao9WCn#&1QqxR2Qp6@zZg_->vBNG~3%P&N7iF=%>TU`vn
zJC&jv=)!p`)y;dB)Ew^<>$S3C4l2iWzpEOHs7P(qNFL~F7lnHmpZ+S7f+T~u!{dLk
z_m*K%b!`}^g&-hE3sMS*fOJYD(kVy_2uL^5DXpZm(nyIgpwg|hAYB3i4(Tv-4>6oQ
z0rI}zch33Fk8@q;$MKKMp1s$8)|2;huQhwO7Th%9+)W(qy(02O=4Jo=safOb%z2#y
zpAF}%MB4@3-B$g}+ZJ!t6z|swT%9C*xqFk$i$B^xJzvx8g|}M4z3;(&=^-ty*`elA
zeJ`xJ?Ls&`iH(Q(V-ld0DZCF%TPxsJX<0-3SM5U75*HwaP#<|hyF_kUQU1;k(lbs)
zp3Y)9+pJql*X$>Y9b9nX6(lRVOA|)YV_m9O{}8v4Uw0Onc$z#nTK?r-M5uhPeIjGX
zzy?Ik3eQANvOHfr7e-fN+h6OiV&I*r^jcG%d~jXE7IMQ;g0mSc&~4sMFKF8`ENLK;
z?^~fJGdmZ1ty2zU1)F)QcN-#{SejCwT4gsnIeMq+Pt?tr$W_VwEmaW;{}o9m>BTvp
z<xl0oP4drdFMny&kN3TCebp!3EzyCN-C}V@6;#Nz5_OTHH1;#DTld<~Z#xC!n@5+o
zO+0pMI@+jN41yH;yQxi?MkAP;3yZi3zX;<k2kDfyJ-rV~x#O}BatPx{yKKuK)}Yk1
zoa*3~m*`vGv1d%<pd+cfs9I~)p+DO9b2co5&r#IUGFQ_Vx+*<wtW((Y%{aB@ooJT%
zeW<UB<pcZUD@L4~9M)iMdGFMPMGNLPTwk&HZb`edj;P<<{KJW<uN9}S5ocR_5%PVQ
z2)j?}q8&#~7o`@3Z)R0>*Vz8J>$>f1&XuX3qgNtABVE4DZ;LMxzAW(t=YIXe`PyIS
znFqrQpV54d`Ax8gF5bA$Z~*Qj%84J&q0h&6Se*#3O?>XWBNBnB;hAbOmi3iohSI^h
zq&UU*6TX{itCkb*jOtr@OVUNdNE|*BG7ReZpG*>gRC-;9j#$)&a}UN=zpwq&4Wk+m
z5!Bzh2;*R%HsR54p!(ra2p1RoDkmnjPdW0r=X-q2ke6WJ`k$!4%SJc+2BQMCnoM+$
zKw{ll@tT$_iqR2P@DaYdE8V7jo|``J1>|-$3aZp-GuN2w1J+_@N@IAap3z^~q1nbY
zpUy~<9BTT=k-9pGOK49lLyS!tB>6gE$$y^O>+>{5swXjHbW+s#T8hMg!ToF2bYm%g
zOA-nFnL*_-WYVFt+{=HG<{J$@*i!E^$jXWh1z7$DO=CTtceb$+U85;K`gxD4pOP@0
zPN=4swB2rv;#;7DL8PR1O-1NiH?@$}3~TtOO}@!~ZaB@=lT69~#;Ef1|K~Ph0j~cI
zTMs|&vRmdOoIC`6s>aqVvw?Bt>Mx&_3u8*ZC!c8s(ih&pd(!y!{ga%Q*n0&G2Y6<;
zD<|2gNejbv?$pk&^rqAuZ#n@lq=htp?GAsIFk0CwW`})pdm}$~KM$ai;~k_YMNevu
zx6BdG(2ByBwZs*r_QT(|GHfI#dGqjgaMQ`XorCVc1CJa|e*>5It(?&Wl77FZS9fm7
zGPqD2hv36iT1WkS;0JeJKW2Kb(3nRCA3a>`@-q;azzgbkINUR2hxj1ywY{<q_<has
z@!?8zlAps67uoN(o3m?T_RDG0vJMhizjI!UXa<x@zrAUPcrt&Z^tbEb0rZLV)D;5n
z-q;@8wrpCWO8u0H$$o*ituIc1^^PQKg)nw7+mmga^eLI8%Q)$pi@~J}3pQUhY66~9
zuy|d1a5KyMDXolm6Z3(wi_Ki{ute;awQ529CXaIQt6VUf-4zYy$izsmA#%pAp%E9H
zG7VHMS&G7fvH5eQ9*k`-WhCJ|%DN!%F%g<-#xydt5G0yP_}DWl=e}GgV*zOEci^k>
z&6v$u59-Y*>vaeF%}fTYPaiMi>q6$T1#rgFf_6vx9zWtq#Jm5NmUjOx{w?EA0uLYJ
zQ}-3l%BOtfu%!+l9B_Ut^{VNiI&Sn`9BDx#(Aw94Z}&os7F&urq;beSKQvtk$*+16
zJjRO|S0{<DbaOH2VGu~a|Jm(eic(dnt_0>A6^}?>&B<+CjOUrtl1=BW7hN+FZpfxg
z`7r;q)THoL@4n^*0#(leEJl}@bjE82ST19h+;h*<;hq=tB_hdU@tjt1cUTucPh7*i
z<)V9y&kFqYiTo%1_^&a8@0U#(9EQTLO3G_XVr&+eaW`$oERy&;Hf06=5cenS!wY-X
zQK2CC{W1GTewKi&OAnk&OR)Gh>+bY1dfFBSPqH&#Y7QR6nS-*urfa&ho~m|9vV1PF
zLZ6Vv#o!_z_S#ch9Us?Ak|CmOmKf!M3dSFzB^~{v1xv1A-H|;Zk$=`9_Y_=1sHrMc
zMHk{}n9%g_n$3_0!_6O;Z+iz5YQ6!!yJ^M9P?WX0d^PNv&76mn1>?gHL_ULzcRA(`
zT^Kl8o~G#kG~o;$`NWJ-D{gSlKN1-GzNkr>N{y)C2Y;atnOB#IZzN{!O(v#EF+3`9
zT~d5iQ2Y4fHSt_s@81#_cdz<knS6U~uYzIWxu&RRZBzM0M`ze0hxd3;CcZGzflIrg
zxc$A-q*bo@tO)%(-MsA2m)O+CW24@e-o&Hbo^>{Vdz-nu*xhM8g5raMn5$8NdFG$J
z_M*_2%n5Ew+;1T#ax2ZyJ)>_B8;0)}`M3D_Q`3L)-;uq%I+XhJ3*FI#Xq%jDfQrkr
z*m<_+UpQ`)l@32M7QCRjX7g5cRcnLXn9F1Q5`Afw;#TO7&P>+<astVOF=n>6-cy~_
zuqXx`n#w;P+^|C3e_V|yVxIM+aEPoj|H`{VM(&{UUGE3HE<?drfK14Y9HaG7mWG^;
zXO*5H$@31-T|-hoB_F9{Vq$V*h2x>;b6cVBK8l$YFG$3?GAi>4_dLVj%q=H;*%PSH
zADNe=Tz$1NNkwv6+M`_Q&GK3Ug|M^u9<kSIJ%ywF+Gkj!`2G`Ss@cLvvm_0~8$^s*
z<3;iKPCvH~*MGiinvyYelxPs70Mk+G%ivz(enUfRXp+itmvWAUrl9Dj{)+edT55wK
z7RcGN$*h(_moBfJk|;sgLR<6?OncJq&o#x~Zoc-339I2>hE@G|>rX1iQeCU;xEmkR
zK}1xvEPtsLJxG+}d8m573;aaR5fY+^3CkuIGI?OEa5;MuwtcMe2AURV^SHYGL#VLe
z+@mP>0WS|_Jg%)r#vMMJuF~B`mrByv4#vlEEhNXchs+|Yvb&Sr*+9D3)_1B1F)g5!
zCYoKeb<Q|O)$b;QnHpKCmuH>6d<4l)y$k=Y{u?*6@XAVGf6dnzT}|+HGnHIsa#GSf
zkHr)l4hxq9`jqan7xCR6O0-$ZDQ_Bg>8gLa)LanS(xQ@om>8S%h25X<ZS;5Y``w<u
zez$&cnZo2Qd2`jjurRL-j0v9j%w=f6F5sqhiNr+s{*}wnce2c++Ycj5j#f24lIO>G
zdwgyZRj1{?Avbn1ryRTM$)U7EWPZT85+@S!Le>b&sDfyon2j`d8*Zn;qL+~qgx4Oh
z_Ew*>CG9Be@fMY)9b=C!j?UHG`956g*_wQP(nUZ0*LR)Gmw&3ce7Z!V44Td>GAbU}
zPXDeixb&WK3p3lqi0xY)S*@SzyO&0N)w$B6`j>N3UTJ&0i0teXaoAEwP{I=FE|LDl
z-I#oE^5Y1PwfwE263CpBS!7Ys&dxp?rq5mdS$+<u6nZsDp0QT@CU+IvhOA74=Y}z~
z@7^!v2EmWA;?bpxw!8*gZ96~pi(Z#3a|^~7UQC&M*m<y&>n4{CJWwcEa>d}guzBt*
z<SNlc8-pUy>rr82J|al4)<^STO}f=57Q7n|qTL%rg<7l%daagxZoT8vXV+RzTc4tt
z7ZO@cKB!jw{ir17{+kU8KCvgpu@Al)o8Ag3r0Z<jwISwgv28YO5vwbAh#0e}d8+M9
zi8ELCwcE+vD(x13;k93Lb933F@*j?UY#5}ijB<3;y_SS1cm*R`cGfw?d5@>KXn#C8
zat^*@h~eJmpr$8sFKQ?(^0@%7V<=_(?bW$VZ(n`}1Ia5n9dDCsn9{GSh_+#y7)~uj
zw1AR)$?zZlY36NDOb}8bOENW6?&}=Hyx;ELgLywfaPjNl;%xG_Ir*9{k6*4os+ni&
z_;%xEZt?dGGL60xQm(!W-%y6#C~alpHllLqoTz0_Yfl)ks8N;-k5Z3KUd|W9O)0wX
z%4%B1y>C@gK610;)2lTNw&I$K5rM(p$`OI4XbX^K3dJ3`yg>F$<n5KG*Nuj#ibFVB
z(h6^T!|gVf%iogLCd828?OS<Gy*7H~g=4;T^7B>xx7Rl(CDI|8uL)>C?^L`$e4<Z^
zs9&4-&M;dtWD}^8nAYsbNl*7Ue&d-qTbo+6Zev#_ub_wphVA<p0-Oppr|a$QFpCGR
zm>pq)_M<zvUasBy1<Uc{-1-50lVHKNw2`+b7Qcr^T_+1J)xPDM)k|+sH~5ipyUF#t
zat7Qi+l<Phd^lXI3OCoKUGUZI)gg(90|l>*zFIOrn}TYx-Vf?ZHvT<(-|gb`^B7Vs
z_dnc=FxV^`T+iYoxHVjzkWR=o|D#p2W>eW>tn{PNs|k(kbW^pnaa6lRqt5Y?k{2hO
zsx!xwtTQBAUcccG^x!jLY0vRaer))0s13&f*KrY>GP@@!tD&JdN3%o1u7IQ|x3&CT
zGr@hROzFq@r}uga@!v&}sU~(j9ZMS)YGf7j&AFc`mfWD-)c43v^W#K8*0j?v-cB+K
z7N3dh=6ZZ9yo}duT1ap*?Q4x*vEp1{75>nA<he^_84Vf8Gidx!Z~-b#<ylip;n+So
z+_DI~`g_FRsJ-eZTwIDJPLoORiAZSzwgbCPf%!>T`ifO#_(gArm&aC5L#otq)?p|7
zysMn)1#PA+=7in{>CB=lu7U=2Vkv&|LetQX<xhsmR-@IJnigzwbY!0Q)%8YmU&$ro
zs#e`rg?)MBl>IY{J_%GdM)ZePKdxa}c`0yU^6(89dvSiX6E-Ckap}G97njHQx_gzk
zv8)v}NQ(wM*o)R+-^M8%+Sx~24wF)q7Yic_t5&ev>T;>onLKx((qm8rjJ#NIm1+9F
zA!^_`4;-}i##Y$E!tCf*J8<$m_}{kUw(4|8&EC7wu*h_DNlBf!HF_v4K8G=Ta-!&<
zVtmlTTp6U%dq`99_2u=Sa;arJnC*CMVF$<9olg4?I>6A?{R%J9Z0aJ1E1JNg{jdHe
zi}jP(%*^_B2`iAJt{&g7QMLfrP^O_Q{*||mud8?dF;RdC_A~6?{1*-x3&Ey25=pDf
zeVeI+9g=CV#xXI=Yihw`m<`I{hLgpW-ZaI4g~OyueHL0zhIp;!y}ikH<r})WcUqTY
z3{9`n2r+Wv#kL)7!lK=-H$(ku?SohKbQYP|w|Vp+0pNHBa2j==m4DvP?ls#!l~TK8
z7y1g!euGv{G8=Q2iH+6~f31p$0rr<N)WgC}c_s%ffdks^Zbd5$b=RW7BXE;jzRsFg
zdIvlylo&80Y5NK;mJuai-+@kpWj-l)yHZQx{k%DF^F4DIajE$sE%U_@Q|k`(_=;>b
z+MlMcUzLY}-%A@a@u(+d7>b3O&6Si8Ndm9SD=_LHcu?8@s3>^rI%`3H@pG9BsoKy(
zsD<Bm*Q9nw;XbRBBPX`;&sd2Jmnp?N=pA^u$OQB*<fxi;RxG=48;nO9FxD2mz6j2>
ztrj+x#WOa^ny)gGg9d30sfB7s#!ley39V&B1hwaAbo4JRM-DK9u3Ua%&MH*;Vpot}
zY|cV;_GU6q;#6x9Uv}aqPk7LT3)Z_km%VvnXq(GFV211uXFy!&S$=?&W$<cVHAj9e
z62<3SkHi38sU7z+L;ANC+v~f5xWv9r@1M}!AKx15Qn-*^M|LlNJC|who)8hmtGWg&
zonV(2IhgaQ;vvnHKiPsqd{Rp}AHG<6-jk)c9|jd{xg*r&LfOL&!(&pE)TXwt@X+yp
zV=Y<vaNA;i?snq3src`{OM^Y}dl&k#SV6cT6@CY2O^fCG_*F%6QQUX3l$fB&YDH;;
zsZt*%iuGj$ni3D?n=UKh<|X9@jI9{Bb})>)aXsDR3XHg5Kk*J1l5+8>pi!>ngNv*s
zILtPWMiR>fM#lGvl0ec6T~3hEJawP@Pr4LX*oD{3$I16bixT5je>KX{&2;0kGPO9|
zD0zvK6%o8Rf6qXpVeS$c^~{QN+tbvlN2{)nveuu@Y`Btan&s&*{JAj`Ls%i@Eu<vS
zuD5`ZE!8~e-74&qRdKH%`0gM2PyAq~*Yjq(yFY$78gSPzaK4Bi;oj(?di*iCsb@Yj
zX|$VDUp<ud#rHQ*|8*udMGuDz%m-ewOH+aPCZI3U+SY511u{lH?1QdF^OJ;YZ$G_&
zIQ3-n1auS^dwChqsdDV78I=6KzKIuVv(`G8=U~PEbMm0eI-B5CvEkt=UHZcj=~@^;
ziQw}Z(;wYlpFCY7zReC%yaK%}+Zu$ua`_l-c)P}UBG_`6N;OkfJ6BI`1W$)Nz{;x5
zsSeNOxgcTjNIqs=W<0Iv&NL+0b9Zg~A?tTu_MxGSrQn{X&yGLe{TX_>keC-mmg&;$
zZrhkHqCt@K(&P8nclb73SKifl5<j(is}HRXD1h!*;(u^b|Hy+=E=Pb_y+a}WP@N!F
zJlB-7yMI=FS~cSEYeT_XfjAXk6%4-@UNPa3QqvV=W7&Ac&SdtEn#<w_yJQwOdQ`3l
z3>Cbqh^nNjkbL0K=M?UfS(eq4*4)Pw-OE%UE&f5Bp(~__JMkVX1WTRhnL>JyK_M6G
z=qk_Wi#+N=<@2u!xW_E0%hUGu0zg^=LviaAdOeMoRU?VK%VUc!el(@k_?X*3pi0w9
zdxc%~o7yi`K?+_B8`ku#-8}I2Yk_Mw<fGw5b%oVNf=+FAUosZS8@EbtKDfV_wjID<
z=jK$K|820SYD7jwW;!I9nS4N^g^`@y_m;fR9)(usq~x^vM*;kOeTg8GsI8|#cc2<~
zJ6m2DcR!*oR&;9m<7&#4+d@S|NF~1^59;+ud6wf=@Hn|&!EcOie#R1~eXy>BRJbzB
zBFZx)@K*I$QvJSuruT66wnlWPR(*nf%E{%o=4)yHtxWKrdpA4s7aiIv##W{#&g|+g
z#?BrNrtEUIM&_pMT9(g&cLND<aI?#rTAEup)AI6gvp=+SexYjWBx!H!U~gw?=S(ZW
zE@^LL@1*WvWMazx$kf%+#8ge@A-lSg(F<$Z2Rs6Zw+*SCRU>XL4%7od*`+OQoK2nB
zrGaO8n?5o%v43tVDvJHW*~!$%7TbM%WwJN=)|mfKN2-7Z&*gNQfcB{QcN+rMvKv3F
z%!Cb;M#kha2Un^0GW9M2$oLr%tvLL`?a*M-Z+Ec*u&Oa6NSm&PZ7~w3Ot#)tT^=wp
z)#IPs=Er&y(4FF&tfeBw*ShM<K*F2BPEBFs8gpMWrW-%3b;h`2Z;uYzG@0VZE*D#t
z#%dPlHdVbpxz=#FT03ox$J`f7FUO1ts&||6V7$i?5zML!9GJIS97LS9s^zox=SCU5
zTx{4d<&*Zl18NJoSo(yOKZTWe?4D57So#9@a9mv_k&CQyu{mn<hZg>c^m1z7qVTpV
z@gx+0uW8Hjt^w~p!$ypnYqj*c5|2SHHl#x@6!?6Zq%mL-szRli2^@R!aaM(@fqR@)
z7k^E-Jc?d!Adz%h(0ey`fERI3l0uQ?sBGRQ6{-N>4&zgsY2fI#JrZHWr4^O?B*GgY
z$wTCc!6Plm(Iw;8vQ5L0;xtUjbmb#%j6FIu{5BdFACSh<OGxw?6LLE2V$UoEt|KMv
zY=k1Gd-bK}(a>$=cu;5>m}*J_n2W=Tw^+T)Ps%4~dDFnt<SY4j?7!+QsetLTvWjsX
zVY!810=g)xpU)7qS;9uDPNFb9rxMS$AED`=^)Ps4!wzVTlE0~0=p}M4q@{8JBUoZC
zSO~FFvdZ_`ZrLwWo1<hOhXGGtg}bVbq}VrD(je3vz>BHzawMf4t!gu`X;_cjlwc2X
z))+Wd<CBv<hgEs5&)tEOjmx)?lf2&FZ&WGO<whkB=rVX@yPJR7d$Gz@6a#5E-W__0
zF4Xnm_Egdqk?Wx5P`7kl89*QfoGSS~E7VlvNyJ<%kaBGd<m0thN0S07%2@%4ELQ;6
zsE4Agk=<OsUC^Ap;e&yX{MwfD5L@>@?K5O16H+OKaFS=RiwZtV2yc3(r2ilaP|CoK
z2hR$C|HS7p?_bYHdeW2du^gc&#0}}MUk0`A<7s<>SCyc$%$thW2HhXZ4JZ^B6mrr`
zeS-#E(FF`pzID91j)zHNAqPmstl>+^s#iu2csogv$THp<M^hTOo=q0`YC^Ap+gn{L
zlL63Y!OgOz*Q~mPz)eE_OnFa(sSsCOf9wtApeCOJTnKS}uqM1f5<OO7no9c-iLe0~
za2}wC2Oq>0F0lgAQ+4ky<FW%ITD~GBFH0Ve0V6LhOmmX~{Nozn78ti+Zd9UmTSI~a
z4{A4lk3c;~2Z-i^_c#^F+pO#E(OVYvEs47P?8(n$=mUmSD@gX_TJR|oSIERb%+3R<
ziufbPAG@m4HKYV}7>K|O!7?f2DLDbHm3SFG1!X>ES9B2?(DM3e+rL>L>5ot$wyX0`
zDss4mx8b+Tx|diH!n68fsCYyBM3Em6J_3Xh;u~Q4DOBWBq^zh56KOsT<>O%>htyWr
zipoZcH*h2J=w&Y1>Qq^j^Mb3rn4Fkl^rrbEfwa|eoD3$%nNiEfvICm&G@p1<uc8K$
zFMq34GYjMkaU4J!&EnRamr-AY4FfTXzjgsF7AS4aE!0f_P%Lu3YV+;=UruLO9y$r8
z0`!bj&^I6@q|{tO4+6y8&-ZB2CR5gg>H?AXktCiLjF1W0L#rX|A<M~77AtznfH8eA
zqEY7!A)*{HF4C55b=*hILMkA-w5^ReYN?WM&5*W0J%~ZeV6<Ct`yl&MVwfp|9DBah
zRZt%FWDMY0Sb(K~#N;|fMNW-2z|W3GXdg5HF+LFG!#+!&(o`hSv*O)pK`Z`xOeIRY
z6dzs|+L-TSqp#LLIvvWC+TPIhu_XBk8PwHIfF4L$GX#kd*|*@COei2hv^RPt^I?YH
zh^7V0qEU*L%HrJ^$wJQ&6@W;+1^|IZ1x6<`NZHDiFQK$>j93_-w?HK9G1&s^uL6P4
zjh5}rgu|rbHZRYC0a>DwcZ#t*kCgNyVFG|VAe1#eCGGoFt&b2Ph^;*k)y<Zu4+W>u
zdkvu%9^5igZuA%0i0jw!POl$_`MSA>2A<6t8Mf$Id;O%E`qs0KbV`!@)8q{WEa*W{
zVT<zC{(4wLvk+4MJdZX&&F;_LPgQfTY*U3Q$}W5aWlOLT5zT{*mPpLX85pGHZW7gE
zDQ&86-3yVa7=c3YC<pb*)9}U+BA97&Jgg<<{zag$jl-OO{0$joyd%|jTQGSg(!<>S
zbn3{WSVPwwh}62Jm-QlG6ST1dB9Cf;;)uAetouKL+?WRC!H61(n$VN48Wr5$7$_%A
zU_R%t_ZwXHmlbkXP%)@hJ>F~=o~S~FjnMIvPub0@9qTd;0WwGlomxwNzqY4}a-ad>
zxY_H<2-?<GW&{yX3mdfzrii7^NE!4frt^r^Z*3Vi=dK#o2-bTuL^#|zqEHDXDB*zf
zHrmv=A3?jStz&a=b>lc2fog>i4@i35QMmTuDDT4P!q*l=)@Y=cds#18{b=t;#pJ*)
z-{$4S0JK7Y6DgEY+L6oE{c(5F5uMs{j55pLqO(G))lw&z3Tb_c_pS}oR8!zgbb!kz
z(8zS1=A;p7*HE`l+;vE|#^`bju?NZ(%4x2GJdM%9Njlp%h%Gb=2_uuPC6IJADZv{=
zetBC-G|%d}n;+y-yL64#!ShsVp6%~!^z1k{HbHm>N}^qhg9-e)?e9&d&F5~6i(MU2
zxj|Mye_bD06?}s-1g{`W=AW2j8-gkgz^wt}J}Gh)rP|V3e~Fl~xWd5h2s)<#T0{8D
z7`Xv(8jzYL%Q`ZLhq8sZge`#Z0Mdi4{Yu^6PX*8DPKqTFo&{v?aWe}S09h=mqdz0+
z0?K88kVEGV;NCo{7q!NE)=AI<7CeFF0{{dygxa?mz0lr<C!XGi7W!3QgMtH=6ZOaR
zQS1BH=u&0P7otD#NC>A<JDK|qu^M^^P6{z9r`nadjS1jeDmV24;U7YfmtoNG#W(2t
zrlFUcSk$OXedyto7jEi_pMq~%taAv$4^e;ya1kIaOic6}xjo-UA%fE>Zm_rma?t&f
z!9kb`khy|fm&hLByhv@`C?AUfqeTW(i^xwq$gGXf0y9u~Ohf@Na5@;B7=YzYKb++q
zq*2rW$h(S8PzYgwQUiXOixjLc_DOMGG3As5!VZ80As@~c8^n>nZic!oA_W-*2{nuT
z^&$7UpbW>LL0AW)f_mH5V2gRf<WW|Np$Da9z|oJRRPVOUcYZ)liCBkbr&(LJaF0ei
zms|SRwe=P__-)P8koG4F)q<TmO4MU}RMAKT5ECJ3Yo>;Z4a#VM25#5zPt$CHsgQ**
zlflvpKx%Dt&Cy8Mg}4HdO*)bI452~5jY!bw1esfbuj_dMG=x+V;EiaAg1*)NwF`<X
zsC|=$>jUua^gidH2|`i?vPa?!gg;xpl7^3fgKPPLatSsn^*WT>e;4hbOlkRL-#S@9
zY|*ANrMS*59AY*5&Ju46Cg&$l)>|?M&n+|xA{y+{Qsx~PUMP)W4@zn@2&sjCdNrW?
z3dgp)JGe1qeyiz(+mb>~Q185V1JV@$d3_vN?h(>p0Gbg!RD)2U94nKdKzjuMzt-m>
zb!FwxCn2t*{{4mk%v1L}>mtx9QhRKnROfQ<R6}UEigp$3GaQ7NG8(s{^?p6(a^k5Q
zp8{C)W#>^3!bAX)4L&<rtvb1e0<bXkl!p}+q-@gNfkoJ9lb}b=WS(bZ)Oo?fj=vEk
z;8{sX^Ru43%f8ZWtX*hy7OOFDFrTU(s@Mf)C`l^2(&M_ZLlji|4sO|EnozP~0Hm*0
z$RNh((|r20>hnny>Cq@$j0`};G**m&>PxT+bV%sPJq5>TE&NX|c{qds7)TF9W98HF
zwO70a7fOC#MCVHsU`DGZ<c~##xw$(a6+$qI_XsQn^f2hp%$h=LQKw=7amz7%W|=^V
z!owjWp^PZoT{c832EoNo;Q%n`cNt;sG#Zs0?xglkYdkWY7q#G$J$C);Ci5hD4_>Pm
z)4(wDlS)$kY}_$>2M>bdrlF-G-G=VR8+O1d!K#ZOV}?Lc9*4wJ9dL`>DEZy>pjhYh
zb08cMg12CLXF*Yq@zhkP8R#R>s>=OmEdHZbBfNCp3cr?BSM+Qe7(#-}QMi6Q4d?W8
z9+M*?{0Z^@qZEk(=ZKq0&Yh(^9#uqsgnaIvLZP5t&KI-aK<nHz;M6JEIMx%UsD$XR
zJNO+A1%P+atUT3A_n%5l-5%vpf?HShe4%R$#P9;U<av$O4uz;}xo;e$Iv5T>KK*Pi
zf!u8>vCJ_|H{?^MIY~BDsmpTkt%JKijP3bl%m<0eMnV|GG|uoJGC>>jy6^&J`EQx*
z*{f|9I<1w6?9v&H=;YBpmGJAm%NY<vD&>>6`F>t8DlY;#JM0YmBWDg&{k3&Gq&ow5
z?jQxtR^rXZfPsD!?Re*!0OW8$+1?=$KE2>kx5Qt$AIX%82sI;p8<9d;xh!~h{s2Y;
zuXfuSHUWUj{cZhRyCS90_eB%fk-K2p^eq+||BV1f5w7Npyrhz#&)(2_4{fncrdbj^
z&JgX*dKs^Vc+ak^$_74<k!?+-dz0Y~3&#pt%_U&-I=D0cV9lRy#kX8h0*=1f5S7cU
z_vEHOjC;5}cSkuHVuHv<5(%pSk@zH0u@HcaB2<K{t1JVmKaH}n_wyi0^wqIEvgkyL
ziZY<epgKs@3=t*d%x%v`z_#lb1loNVK#!cpX024=VR19E-#@LvsOr88U_aEo<dBWW
z*(DV3`GmOYSkY1;G7pjhxPdJGi2D{HdU+WrI4ysoA;@01YG?dck{0}#b}264l3qYP
zRU#$zvTxb)@{>1bi>{iP;equ!8_(B7l@j=}^@{suTf-@&t3r5ySK9J~xPQoZi6=EV
zP`95e<UOu|e~)GVS~f7a?Whx+vEe#+N$~{<4K*w|j7~Xm`d~AI9`Xax&7M}fdK3q7
zrjrCK&J<?^sy)*pvYk2oEC--gs~@3m7tWI0DX9260<&0D;<?Gsk%F25uEG_d2GH6%
z1+gff2Pzz=#W0Z`T002WL-w?nPc1<4JbYwBFLa}0T@=9O|MF|pM`RWBuKBLD{<O-A
z`{-Bw@m$#`4ZHh7k3%?2iQVu>Sj&B3%w2>3n8NvFVAxx(*h2s~d`+Qh`IR+mq`>8k
zmd<kW9^6s_>9;Ea8=F{zL5)#HL9qL^Uz$kr7~)8Ep9Vr2@)^h%Z!5oJOAymZ9Cc~9
zImnu4?)h}n5xgfY$_Vk)R8B1BV|_oPAV*;f@vwRvkLaLI)dCa=ULe5gmHz7CTAA}W
zf<N)Rbml32Y6brc0RKen3a}d$+T#<i=n9II)u>Qa0KC_!SBop40LrO00gW`<^ly%w
zx+CDU)^9Eqe@O$gplu3t8Hw@<eoK9%r&cDkD0l{RljerZwlwFC{RT{jY`RYGFRox(
zY2h;M@`uI!>~^{6=4%|lx~PL!beXY1-|yN}O#wwd0-)N`cMp193yx;;li`TYL-3c~
zT3+u1ZgZBY8J23uXr^~?v0(_Jl=B)m<b?VNPaZ-$J+QbbW$KnY3E&iYl;7{wei;yy
zSD<B!n~MEG_E3P4^|Sk!!$RkLVo8#>W&W)&I#7Uji=HBWPD1#dj7otp_B)n8a`RtG
zJ+1&^@hJJPKl&#s>-l~i;6*exD19L8kX$o92f%1l9stNP<%tl?`xxg=0J^)7LG<^C
zl=<&CA+)f(X~AlC?hKN`0v1G|pl$+K>-uA%={3AH2UWD?5CH$l$0GsVigUJPsI;I$
z_ItwUgTOiH5^zqv$j0w93nC2sYPEZhYU<*d0Wx{dYQd=Y(mKG98!b=yI*micCUQ5!
zOJLSj3sVWxjn3tJquv1yp2sA3kCP*l{02)tZWV1906^6fzbWo67Go0@KNa=mTxnfL
z$VP^8D8bzhbM|gYVZS4~srCS<h`zbDD%e?9-e|6i)zH&Yd1fXu8u%U9)PQntwzIGA
z*fumd#;|!aY_;2=ftnv?NMBEVQj=t;p>atU=-KwpMF+EnP=?(mKzaW)f!>SrngZq+
zW#*N#48SDRIfNFFGGvXqa+WzVbJfbX0AAp{6>OT<&Rg2^w8K2jXh0|o4-o;}K1!X4
zLL73wi~w!r2rU}@oR$f|{3xd4RNB7NwsTzc%!z<wthxYq1p=s5nkf$oX+a+Vifls{
zT5&=gE5`KvzIo%B08S68hLlbQzepsVd`H$)5udmsx3`UpBUaq!9?U{EJDUdGP%-M{
zXu1MURim3Sri6?PV)2vNi)6gxm`Szix>vq5ND{(uWAx!uJ4b)O3{h1@16LqlHq1qp
zbjj+&7n7EhfVDt}$J@o+ky<+?lAhml7e{v+4kBE`HTU#Gd;I_&``p}n)Q_;!_eYHE
z&+-70lSsoyV6(F{0O;rgk{LPYgN2AH=k+(N5Iy{eht<+j4nTySI}a)hdI0pZ|F|a*
zkPZ}o&4{6U%9{N3`AlA*5aDTKqxB1zk6BAy3lE@DQAPj`oTE|DnGO&^vtV=9V(k52
z)Acb|OF=&YoM{?Z0yR8>twPl#=9597wR5u<yC&8Qalex;s<EMFUiGYIKe;ia8k<hc
z+!NNBoDwyd;bGRJS=T?W<8v&Jvzx<E=X<T_2ns-(K@zr-zE3rSSHO=p2r@2NZu=_5
z&A+%M{kD*m;#P55+=$Cl;(BNV&h6r&oysXaM{ugU?S?Zve=4klUF~^}HB|q_(?$sF
za9k{WM~Uq?e4<z@)lrmlOGkGoF0SHKX-H}SD%=M&&0-LMUOYEB!O>F;_-|9dGXItG
z{WUo)%C{hL2+HrwkuQO^#0&pbx2}`{k@vT50hYUd5rpVq(qEplYawyjziJn$1xe6I
z^55`7dIT&JO%9(ktp+x*%JvypZ$UW)BB8xy)tx$vTf`kNfsl+bbL!T5&*1tNz)*u9
z-2^`pq&g|{fT-oyKl8l~9vMC1&G_BlQ_e91;aRMQ=`=KP_f9Ri-DVAvu!qhkvHx%$
zRj!GK0sBsh_lnKu!|>vQNho~VOsnc@W?_!W`!A{pX!N`2x|a^Vw>_BI#4W8~s2=em
zoQH}O$GB^lYdhB~fHHc$b7Wc-w130I;-(sjEup$w7Yuy+e<cbH98M>|SsT}X%;f5Q
zH)lT?GWvDqcL=-%y3-eS+Pt5YPYm7p3Lxd{mCfYiq5Tob837M+M#N|PzpMicfu_9A
zX{+#F+1-C?>mul%5EQt1Kwl@m_7bp+h0d_ZLjVT}aFc(-KvKNxqQbebxGcefT0cB@
zm0BN~(mZFu{|rt{5Q2O<Uy-hqq4{s-RDd`64g^|rufXP@qm6GkETP3vg|~SlZ`O45
zh*yaNLC*qQTpkc=b7ZgS(#i881~XjSdK}mX-rw4;Fz~H!oe=u`y#=PG3%7onZ#0MB
z%A+GFR@|N>*0j>5*W1qWE{;2z<76XhPGxwWBq_bAe$mPMVQ#MDfSTY*#|g&u`X5`+
z2*%A5K9-E8lCdq_66=kEvV~!*<F3^Yo#tGV1rC_A;fI}vaHj<;7>$%KC8$d3biT-D
zOe%92LCn4qY7RVa39K^oO3k?p^iOyTXi`!fI_2Hy=8lOz37KLO#@@RejTi}Y)d9=E
zs38%>=(K@5X8_|qU<G>$;9mcE<78lA;UA`iA~}b@f{fhg(|z7&WrbL!LmdSIizE<D
zKAgKjK;I_7hyC@FZ~td83t7qD0mAVXzFitmli}=@lwZcF0#WPGPL>GubZ4_Vko(3w
zi%KbRuSeMn)(D{G_To-u8xZ1FrfCatTeR6(9o@7A)w$_Cw%=eYS>@exo;OX(k6_`6
z*p`*mx}}`IHTc{zt>H(EonZRUVgk}<@FE>wp1k0?qi4#Cbq?RjGFBu?{9v74CnR3s
z_Y_O(dyRdv!+N*L8lK}HBx|o@Siqd!x$-i)_Kq2T40oDV9i5?KBnclT&hndig1q4r
z3m_7dCX)qno$=Jz0C3m10W9$T8yx#<qRab!!D+{dGUNFBVV_a(#($V+Mr*yL5^v?K
zTSZhbY!I+#;9R&#iWmMzS*0xlu;;;9Lkc9vLfW$A8YfD$4Ut*(MxPU~G=-W6qV9?#
zf*n|cIPW3EWx!gz8zsmDL5#oH+l}J^7>&6`JEw5iPD|H<*cKFW%$xph+Bx-Zw}HLr
z>^8tBr5Q93wr~De+#>Re9Bc=km6Od=n8Lp6t-M(+vPIJHF>%z(zQ+qcHert^L%@)t
z-f-}BSwPbRSaKwpoF6u$0r(YnqS#S|>72@Y2SK(U-U-1%<(uU3BM*f)tddUp=#C9L
zPoj4c2R5_yDU6_o%bx5pvwm{AzK*X|+SXrq93+^Nmh90uk?V$&pB6u4P)HOO6ixwj
z@p&s<0k&c^3e`0Ol*T#VRqI0%Eaw&R9~~(o41zQf{InOaLjpA@!tQvwJ1ysWB!3`)
zfRg(+ZV{SxKoZr5|F92oDQ^5M_aiKQ6@(;S&Ixuq6WDcv{){AHC@$$D85p$DkVy0;
z@Dh7Xb;NSXS!pc1iLA5x7eGMRiPUZJM9dRGMg{n)TX`IP8R~(^+l3Rer%XRR;#G#Z
zFM8gD8P2{#F7qh2XIM+ug>dKy6SRfw-RluguH-p^938Rc944F;xD}Mex1D^5;%@Zz
z8+=#at|-`8soZSbo1|zzXDzIlsHwisTRhD(@W3*dBA*HZd6WLAzC1P2!E<+qdeL!P
zv)h`_^TevD_fzI(zO?5nPpZ)ly%TYL-$*w1!PX4g`qAlaPlEQ%ZN@aiiYbMkrN9C{
zV&&&F_JH{%1ZwUTG$Xcc{PkoY9C=*;cD~}g`RmSj078pMD&2q4Jb?Y=B~VM>=hQue
zth|>00Q&#I_8@|5nH02Ii)Ze2-Zab@pwnGAC_A9i$3yFxGr9@k3pe_#O6i&>{$Y45
z5Wu$lB`y%YfQ2ZH=elK}Sp(LrXg+NNZ5lwcZA%MqM;;xKP0h%S#+}p%n+m^}R|?%~
zehrytG0K5f`$4P(xzq8)ypmQT1xzV7w(q(zCYs&tR`d0>w;Oz3FVx`uZ9`0Q@Z<MU
zEzzy6=(s9D&qA*I>&~_;G+Gc9$CVrsVwHS?l18^Y#}EQC`={xS586qR)2U#k_Ap0k
z_I--zIJkdKzI3Y1*2(RGLS|9TS-;=BC3f{EE4c;|d>Q*WDZuN8W_%z@8F172xcMY~
z5BLoQfk^j~t$%gNeSqRxZ3d{$CW|bcsQnP<Y|IJRIDzI2EFe$H%^W_ZN>Ae%p*sxF
zKO1Ji)(FZDZ|xbdK8^AvL=u;<m=|j1FaJk!oky-Xo%1&U#d>Y|{If^j`=2YY`d?%Q
zj0ps9w?u*@6u_m9z#1%1n!$snE9{hv`p@Uf=lLnj?x*#XaHO9g;t}!%>@2(mpwxZt
z2;MD9_E_muQ3KfBonq!j^?IQyZ+1^NSC4~1(gooo+&yF1kC|4OBF^nbGRpw(ZB2E1
zaN~MEiEkST#i;#FM<sH}=xqo2E-}%p5#euJj6+a<O}6((a!J$!2JJgmU4|W+qO`M@
ztM*QW8{OAjmTXEtZw9a!J|8U9Si8eme`hgnzW1BJ@ut}4x~W0eOor*hHG*ET<P|`B
z|E?YXVyZ5I)BqTQSojBWa2G&xAk_Z-e-CQJo)93!uA(-{AS2oo05o!$<r*1+sm~FM
z4ga#cIw}c}1IH1b04&3mvL{rW?-~lv6IQ3^A^BU$e;5nc5_m40Pft88#<+kV0Z|Tk
z2ZB1%`cre9^SjDn&rDQV@FeGZB4gCzuSRFy2p}3afDS|j@<TwgM^P|Fg<QFvoBwHg
z{ksR2a4(`%9&XqJb8X<o+XGZUH~9d(cmqGjU1kQtYt0xbps;ZyAki=~B79-M(5p}Y
z#AVV)km&lf%Mx7T{czq|ObHfA0V#-qFS=)q7P20P_nhc2<}$i_WADL}8->%HujeO6
zPARcxyj!^K_YnA4&mxyUSd=&VEunJQYdpQt`{III&Rz!h4+PhhclCTiMFrbGBN7NU
zp?~(|oeeHG#a#A>Ezv=RR0&DI2zsppZ5u>WgWS?^zP#Q<pc5pvy^PxS0HB(Gv&4w%
zimeNCEVDsS&0IjrMjUYp9|4x=Fp<7@Y85OHumDnmXK>p1BO58TzK@K^ihw=~h`zmr
zLnIl1a3VwlS%q8(NeNs6Sa!!>?her5s41HkLz=axDziZB@ibk5I3EojSVs-!6o#4G
zRoQI8Ad&N8t%fSezOaCN{eD=ztQ+$Y<qQ+0IWKDvnW*e0xt@V({R6`WGRL%ZXC1&|
zqfQO$vr7Vyi{1kty(8N<FALni!@liVo{MR;Htmv3TQA<?_q$7dfj$0~XigHN9!$4l
zvz)2kb&t(ypwP(Ix2D3GY*Q<d$Hel>QO1OWLhYX69kw4O6~@3Wh+mfxMnh{5kqq-;
zC`g%j5xE};5q`i1*T351MYid%9whsP)&kIz|7Dobp(ZbZ-l+F?cm`lEDje9hb1sq}
zoa4z(&2!eJqeWmvGC~Lw5L(y{=i@N~n=%kf&S&X>^9)L#CWwvGZF_*l(Q!tCSn@({
z-uo*Mie(m<sCC#|xY@S4!U%(GTMKa3dz*T+Y^@pA0AyM||0c^1_FHU(z0(e3yTvOx
zFh4L4OS`3(b6uBZju3B)y1=h&2O@}bT}nN<_5*I%`>E$ebse2McDOt<x=~$z$GR1+
zn=%Dtp7rAc;j-|6STU}_%e=>8h9?gzz>&KKO^JqGn`D2cy5zkHci0FGjWru^=SRm<
z1g7MV-kH&zglg`#ayInsEmZ6^a<VNV6H0Td4U)`9tGZLdIu@~)2++`9<(~rlQ#&I2
z^6?==M4eg<5kUZIMdm;Bg<<tB`$rd%t@)`g#{dt{#{)K@{q+GLrqCNnfdqCM8&`F4
zkX-hu8S!+1eW-Q7^2(_M01`b{YEOkAE;}X8PLG_&&8LIi&;>kBX$1*`(9Vjevzy4B
z&Q#4+z~)b2X9&BW%@>>II$Hr{ucT&ym9VJNW1td>^3~Dhvl@6b+-S1dv-7Olkf+9%
zzG=kT=#ZSW%W<<;U`y<VWmQut?{PTL3d)!r@TVI%CK`rpeg;P_&}BqUt*|u;x56S$
zBoliLVX>9of;}EJ3ER&V+10w(ji6-DeA@>TfYznN!-G7JZ2suP{3ehkV5qj?REta&
zz*bUV$utGkJ)@lxfTGAPiRauD=#w)R&}HPWHBnR<dN2S0BDX$KpM}7P_@RYqdw(*b
zRX<ZVdf^bQuKyl<dH`8MjA%tp1q#qqS2?Sahy<5eiC2nZ3Qh+$MsP!0JEW{gDNn6-
z6R|Jws`F%{^G=E|pNA^2`|%6-rp47wdHvqIh($2E^ah~19Sw6t@YMA|he_ew?PLWG
zchaLyYP{IVVJBNhyetOHH0JI~ilvSN3L0ys4C!ND2dU1``oma5xf*qd<^Yc48pD0o
z+ywT5u1Z3`?nU7#d0)TT0fkBc+`KyiHKR&UGjiDwU>pvQP-IsZ2NLT^+Vlf@Lm=x!
zBjifCyoB|}FE=VrL^p;$2P2F3`?TQEju-HpsI&8W@Ij4UYjNH*Aj9L8lbRE_IWmlH
zB8$Y9By^O<d$;jq!B5&kzs`N5@i?gw-W@82WIxcdyc`M5ILU83=&GFZWU=c%*gILB
zhVL>OA`ONT2Dr)Bk76f-hNw}ZPuB7Mj{B4L8o$+ABEsw{qTUo@mBCl*ReTROl16>k
z7VR64U4Omb*=VOZnU40`NU>j+Mx=A}P5m`53{PhG#!T@8Qq3@Ccsm^j(wyvOi|z05
zyQus<F^xtSv(ankn~WhGUMnV>QhKRO^l-U!Z^~<p-FGp2+5pzd|Dptedr<?Y5$PT_
zdX>YE7X5bNDx2Lp1iCcd>)-JG_M82-glI(e_q#B~U^8&9jC0fD3m8f8^InITZ?VI_
zH|}qUc|xJh<_s1-2atTRgZ}9E<!~6L7+j3z1a2>Su*nY$z)u4|TAMmqPWcoEG^S{q
zDPw>t;fE)On2iVc;BByZ#_(vvK8@Jn(bk%um~B^Q)=p{0@v=-ihIRc3{Lo*|Z(YT2
zgUk<ZezGHW(hk^fx@-&bKAi`1*tmkM?eV^jN)%pnFD|2*my&)fck{m03`e(-z>6<$
zeR#(?=9de<H^8gO@~#n4`nVX2JssJp^&MS1&~Jpg9T)rEDiPy=l$<QFpKK?cZ0?Dk
z93g%Za9H)I=nAM;?BqzsaPOPrTK%5E?~z=EB;WNwnxp2$L-<}JjbB^Cp7$Dcl5!EG
zcKX=vU`%LZWp>JaHgcakO|~(8NRBAgS)?A~)o8H4sls+t3ak|Lx*i)rK#lOM#pyF4
z#I|-$w)4fPR*v$m^<LaL9zTRG_?Q<=PEI=CA@SZBaeS>^->u^qJ`zgfdov-Sh{Wpz
zI@NVxr1|^N(HILBEBvV;hn}w>yhbRB?P#O1VO{2Asfp%z(}oN`+*u3Qn7TZuBIcvN
zrknW~IO||xd`fqCqIa+7<DD|NnDfiZ^#ArNBIGZ7&_6c)f8$p~GEPPwr@tHG;o(I+
za*_7`?sr4qI4@qfaG@y^Ua)9_n+g2+|Nr~{yas+>RlNqI|F1swpKr70K7E<}|D7xN
zsDHL=p~aotA~_OQyT#$478L(TDYs|rWAVz+iJ^7iZOCLnU~iOdIt@O}Tl>okw%XBP
zQ@5!WCIV^eDE7do7vNIApVJZ%<p1uY6k9O4`pMtcbn8OTXQ?P6ug>h|7OD=a4w4pO
z>F5_Lj&k)02BM`JxV2ku9od*ySn|pEiowZZ(R&x;|A-XcIBs!|OrZ5LoA<zeT){P*
zFRTlt-|Q_V^muLO>10W6oH#c`Z`RhrEnD-Zw(@|rAr)7v$n(8<4jHk6PJ9#}$x!w)
zKG#oHZJN;z^_Npz-|MAKeFA@ie|j;DgICV1e>2pRs4;{0wdLUpzhy)H$@JS&x>izx
z&kF5V?YKr`Ka!kC_zO1Pb5>SQE81M|EVk09m@&E)kVZRJaUFxb%qQN>_2wefhdV@P
zEMK!;tI{l^A+Q#Tr#q_qSlT>oSqNt`J6Y}{?pV*|&7W^l%%nnl3V!H16U8;ys=8Em
zUaSa%SJ50vkNRnDm9($pkm3#A`7gHQ72y2OAE*DXk5Zk~QLw+of1}}0Yu$pTW&POV
zNmhm0tv{><_O|%J7i%$QGNUE8whkR1%*ECY$y?yDQi$uYw|Y1%q+rSf)N_2JF1{kN
zMqM)XYvOCp!idM(&spxced*tROv3m*T#Uq{mkjH_(Cj|E#~4)I+>T|$3q0-#Q+UNF
zU+F8}YpMZCU(chcw?f4MoW3o;bbBNDgCFdhwh?B$IpN#mF6@X@GyahGI=!%ovG8Sd
z>XR|>Z9*@ZgIFwecm-tbedBBG#~#CBUj);J7wi?s{Ztl7bJF>6leHnksv`E)u8uCM
z8K8n;8zmR|%TKz@((anjLoxr*ANzi(j`bSP-Ku3?i%M_#UL&9T?2XTCJ-tfbmVWkV
z%rY%YZn|mlN?+$&3&xu7uFjoBurj9sqEN*(mi(p2K<C5zG7WSo3OKTl@>3x;)C@_Y
zT-mIZU&g%$Xf5?Wq?tv@lmxDPkenOAG<!1FWM*OD*KnJVpZ;}7_#aO~Df+1>xtH@n
z;fZsSF~)DfoU)S=a@U$cZv+zO?gXpffjyBO>LA26=d7u0*?IB!egv_)0M31U!c@5!
z<M(g&v*K@7;TI=u(g~Lrq(~b$D(^nY;K#5es0#U<7%<nZiKS~<J;);Ps@639>$qb;
z+cJH3q{fezf!htLch}{yS(xxs@LmN>{Cx2N@8+zE6wY&TeR@+eC;3|V4i{#b8k?KA
zf=kwA(txnbzc{GNuO!bt6mcn4pEC4ow8Q~ntlw&@xS7Fhd%bC|siHeH=WA>19JDCo
z7ipeI5>uF7XNjHbK=sl3{K^j3t6gzDiJ{`#@;g$P@r|@BkL(rNRtm3}_Nbk(kmXjF
z#aP~Y;}&)#WYk_;K0}T9$Bq4qclYFzzz99fgaQK_9j&YFfjnuc++!@czGQJ$e^}lx
zm5yFx>i-t{C25n-Yen%RF_W4;yUi^Q`Dc%7vp!LCZ13`k_2l42O4SPN@Xf?;Q2x|C
zcw@!kLQuY1`}B|~AP;Zutsqvx8ZoGQSkOh5JS%|Z){W9v&%6UFl8-M^g0#eh#TCTe
zGIJB1Iw^5j6K=}|6%<HvakmZKT@_T!_!2EWlW4_I?JBZnX07U0POe3vF;$>2`*A|#
zRpIyqj4!0E{>=-@W0$Tk3%`hN&&82a&{HP`I9&S}^QDTbeunOemP6xY5m^EMfJ9V+
zD8WT0V)c&}38F+7JIb%#)z&@~BfZG{EJNSYgYA;DREMdBOesIZ8;s40e)!7lM_i`p
zqj2Rp$j*}6)zA1gvDelAU_~Fd=9#kb!ok~9<oKk2m<S0XUhFJXCN0{h`$~s)8Ne0w
zFm#QAD_B2ic}n?Ic6#o{OgYvQ-+72h?P$lt=&$f$g~dqU-B?RH%b4y%%5bzLiI3CQ
z$uGC1lOo?I^O0WcR_nj*n5rc=)$aP)N<&IfUW<1?Z{}ecgk(FIxjs+j(4Zn<iOhB9
zNAXmg2Ixf(qasEwYkJS@xX71}kBZz`^|}%n=G8`{W$zR|!naA&w(S+NSAz3XFg{WV
z>yavD2BlXQ@hq|_{MHbN&;2-Vd>lkYYThc3!F1qPMWDU+?D8x5FLo8{M|*=T{q--G
zvOLU0YvV^EU2O7$X9YZGG^Clb$uHQl*LLA%K4DOz(oa+~E{`n6xcthim%?SU+I8#O
zl%WkSv37U}ypA)m)rv0wn_{eS{C5jwkn7+=Y}<^)girnNq~|P4*u2HZ+F3N%#9@)2
zgDLRnqD-G{^kNc9RaDVLKF<qMy!JMmravoB@QOJxb5bCfpT<l<@>~2pvb>ZhSZQP1
z!U@?8enX}?cOQr4GCrRsz5bQI)+4B%T!!2$L%T6@skxZxhMon}e2dlUW7?Dgo#c3?
zCx)P3>A#4bsM3Ftb`%U&#;KEhqkEEUR{l_FK`s%LYRfH|lA2-JJ8jX~gPl`mtfha;
zUpjbzd(KKfr&7dl!R`AKhp7ip0t-9W(LVWsg`Yj^AJav9vE+z`tt8zGM{H~QzKmbJ
zpfmm}mbEj8bLgwK$1mAclMj<`88V&+MK8Vz^Cl|{dAyv`kNd^xg1m@sJ3$*gww#T6
z?CV3EHPFPLg#JyA45^Z`X=!yYD=(Xwn!6yb>a+&Sg-<KL+*XyFt=JAt)&B^`z4dH9
z?3~73yw1|~yXnTpeLnpBjWMaj<;AVxhhL&+bniY)caGD{&-+>V;^e8*!+8h2!3J`Y
zoWxPjrK!nl^4{jZNFx^$Ojdq&b+US_HI#m01W|2J9vm8PD}*0P9XrbKTk;&68INaR
zDhq`8W;2P!MHJ_mH(ePu0RMn+d^UWauv~sXgE6c#nm3JOHrf`*d)O*c{3vF_RzDT`
zhhY`R(z{~5UVwPXS-yWJ(|gS1US#NeLd~9m0Hl8Wb{Gt^c#C@bBpp)=^sYq&r>BLL
zQeVrq`wiwCr%xj5_;gMkuI_)UxexyHfiS$BygdKY<IC$szB^I1$GG{n3I$||7bF%%
z<`G;J51{j=uy&`A3;j%DCQHX;c2`ktWK^isv#@AYH~H%pnBcdhxw+=*yYlZ{hv9Wc
zkm2T9EkjZADtCgS&4bO=HL_s8H~S^A`?;lLi03ToPaQkKc0uvu;Khq#z+=q!H}<{0
zX6&wxe0ZsH&-)?RFrlV)Cv*Llw~eO(m{GA2nqRA@dLfo}-KVo}arMy6&zHe`CNaYh
z=xhB-<i9-SXp(b*O|X|V92*$;eiehm?gU4}H<?5Zx-cRLub!~8iy6YVh7FSgHKea_
zkS&(K*?&Q9Q&;*@jFr$f#{FxS4kd^1jEd6q<YZeWu86#0XssSaiOW{DtMAbASBup4
zStkkF2JYF|hZ8q=L^1ZSwIz3*&@Jvx`*Pz5f4GZZMC1y7`C69GMlx&K-`DRwc;*w0
zdCxxNWY_=odKY6p`{PIXRC85jSB+FXZxyFZ8xYZ6qQ$u5Npv6cYvrTo(d7^2`TdW0
z0x7$m+`7R`Ivb;slLue9|0Yz<B5vj`=a^91qgTcOC3&1waz4JH>C!5G<(?k+Uk{}D
zB?&@T$%;X})v)V=U?ux284_P3`0wxEWn<amm{l5r?fBdyJ9HP9QV)S{DaI1<mhxDu
zl?`BtlXT_$Bz>JAanHJ1cIf$a^RlYKrVCA%v}P3F5H)Xj5jriiRJ#PLX%X8!Ub^I^
z(ZBuXeXyjI!c0t@^&LSifkAa{Esav10odf`ZSI)uE9%5aDTD>#3QJt-7Ut9ovvPMo
zJ>(#gGZDYZuvVr=yen?7xgj`iT=q${XfwIqjpEy29yxKjtlq$)BeytJ%Eg&@p&$3E
zVQa;bWj%DS6b#t>>%1npv}~8{>Ix@z;5V*BtXI%3hx3t!(_#yB7*3>*dlE?A$B2O~
z9kvGX;YZ4k($mrUy%#mJ4378{{8=@Y6LRRkKy3wC@FjB_rytbD-RsW)Cpx}-*A%6<
zt7^OUz9u$&SWN)rB`wJ`JySQZ_YPbXQqG+z1U!4+_8#5x!kX<0AJ6WkIU(mlD(<Gb
zhkAI!oODD|s*a(u<vRqCy;Apdj)hCtHUIpHs(y0&GPf;$4Slv1ss8Jdy93Utqy4&y
z;*_4<%Es$!4m;!w{Chq-j#)1u?ndG_F5nGpHAz2GQ|rk#_cUU;zl2L;+m55B$n5s~
z$*__g`43%tk<{I$n?~ginptPfW+~~n1hgM{P_5ZyOMd)67(2({Ormey$F^<Td}G_T
zHL)g~*w!1{wrv{|CllMo+&Q<-t^48pKb$YCTGhL{y1HvWYpwnK@-K2s6nl+jVS&T;
z3wA-g7dfnCgr)Q}8d0vQaQMgX_AbAw5e##iSWQd*;fQ{iYEKz2Yu5dA-W|0sLe-sP
z)%5fQyz`B{{jO4GMP>KED|mLMAq8j5rmg=hIe=#|0XgE8AH!}Zk>5kfIb1Aq3Fha-
ziZ?1OA$F-|O9{aHp%@Fn&Or>H5SJs%SPWkx0hp@74ZalFK5(&14&*F=Fi28y$2PZ)
zcMYYs`oo3{U3>L+5kyZ6MBX$=Xf^TxRakZP4t3eKy)3zRUA9_R*0RmQ0W+!dC5T`3
zjL<^A6*L6s9O|rd*CX!xJL=Bv)uFE8m~H!b1k=~Et4;-_)P%FZk}`#lAhBim*&?cl
zh;HStpq+L}ujKfq{#o62U_4d2_!K9^X}J-8C)j5jP?FaqxC9j(mnjx!S=&aP439n*
zPP0$i#%;XB!$v*9nGF_D^dWCmx*m1_HWiPp9sYolhIcg{%l?2e(yfrTA``;}XB5Rc
zoFB?zSQ$I^xEmwCVNF@~1H^p=dMas~#hoEFO6Jk~$}u-DLnm@gF@(LiZ@zn)Pd@Y=
z*6yn&ua$IR>~KbxliLZLQ(+3?YaU6(O!)&B&GP0MMVh`qT89V2cJ;#J8-t~oAL%)K
zt!=U+3t7R@q`o?J(RKXBs4oTDNQbdO6~<1b?8Yh<R-$c-s43^i3&kdT2={JlMwvw>
zi8+Z7`2$&FJtG4%L7F;(a?5%w-c|YUaS_dv&m@BIh~D8JlY~(`e>t2pe{j%x0QQ_L
zexrxGLR4whIcuA(6!DIaF*$;U4hrw|0f6=6Z2n7AV~nT<kF4>;RZOPEPO3d$1mOuq
z1k3VJjn-6nZ8x~wX&R*=zq}5h!W%c<(<#xH09VuUrhB;t^QKzQ_gR0vy*60sf?e&(
zKj_fAbE!Ch>U#;#t&!%;H^t#qk%uIWMT8N<_oUTCRi!9&PXf3Rw)lkwNp&QbaA6T|
zWg50Xm+_kC$+90n3`R;b@9(oP#Ae2pm=u+YBa+68-$~0Z2`_byEvr@B<)OCCmaqhk
z2g3cZXPD5u2f_@zabVwYNDwQFx>}h2V+8=cJ=}Mdz!Q>SusIRUz%t^TQ&><}$SI@#
zkXP3%%Q6VRk-ffj7SBc5n~=0Xv00m=ELivNp;TG-X_N#9CFG;_8I=H0nk!!QEZ=_f
zPD%f8ak^o71`Ev^b78&ZYc&h)o>nz&KNCr$NthMGcq_1$OcQGXqE2T9pTswwLVE!s
z`YV#mlQa+ZEA3&j-9g+9Bios6fFgQzgdWd?ZR|3%wrD7^qExf1DCz9w2=}g=&$qP*
z3@T(=C;B)t^D$s8kqqygS)DEYS=rCEd_9YsS__wcr3Oz#dTM25G^&>2ha5%Y&df$#
z$4SpriP*CFPy@R5dOhjY&+}Ax++jxrDY0aLB0f}ue6hwON(=0gD%8Tl(CH#Wj&u3|
zZfW>R0;x@B)c((<I1DzpqpQe7wM%zt1*E&2<yLfmn+|efU+Q;L*kT0Y$q(R~T*RP{
zPkFj!(@=2&b9p-Ms>;l?2$TjpH16^Yfw?j-Bi~f{ER9v1xibHlR9~ePQqViB*?Wc<
zBAqpo3nb#m<4~4?U+S7haiC?cY+yBL;e*xSiJhIKh-0n!!YV<rkk%hY+q?tizv&M>
zJHuWmJDr4s9yapucI5_h`5RcKG$tvKg{`4DC6I8Sm=Lb2IIg>NaQ)3GCTHmpP^hF>
z(-+oEXvO*Fc%8R;S6?%K-RPI{;ZQy07i~@Cb{Mteuq$!o&LWf)5PCucw?~5l$<ivF
zAbp0T1|Wgd<rAX-#|Y{yfPIyX>_f+KCMu6@E$?`H$FXvBhkrW#MBMr)iRdl=a86uW
zF{N9enESM5fqOcb%Gi$<$-cOKwJ{D2fN5(p7$IC-cSg00XEp3ptt!~x*LhBA3)vk+
z-MXREa(I8YYG5U)!KucWx1aE;&Fo+#>W5U!{A5hEIJ<~hKI2c$Duc0o337&So4H2I
z12BVL2E<Bxynwe!#cZ=!_}HDPoRWEK&>an$Y7XhrGh8SaNahOp+?HaQ>Trg#Tm$_b
zOhehvqUxh+`04ui&VB;(Ey!c#zL}u!lMB5jk7fUx*8v*Q82(EEq0{+ojG@ank?4U$
zG2f+D6ApPuuA?QpXH&jp4AuNZ7lxe65#@gq2Y5E8e-Bv%os%|!K_@A)q6QMy<zWUW
zUI{-r%yEI&nCx2WE`c>ey0a*=iMCNyAL}kBxodV<&R)p_Ipz8R+ma@Yc~2I$nAE*%
zv<tRn9#yw@Tzl1|TxUKp88^2`#ELEJ0+~mtu>+s)U@Lg`4IoZq>uPdfAh%wQJt8@9
zU<iL_k$W(^9u;(l^O=?&B=VVyH9;#66nomz?sw^TO2{Q{0;91>@wiZX{7lD+TMQ+b
z%2JdG4M{x!5|)9Tevn{nIerM*w$d0>o|oUaifEfD(ZP05PTTP;kJaimHq2KscTMCx
zeW&AMG6a^c&C(HivG$;^oPVfW97prj@BcHaYc`m!5ZKjFbg*!KUp6qy&tqAdE4^g7
zLgsedR9p+3Z$yzEJL`<#GJrWe%3*RRYKsKzkw>}MF#UT56YKt#;oZQbB{`m7H$xg`
zV2|WhJED<~f8lf_Y=&4%P<Wto@|`w%Zkv;<Y~&Vghf5az){>>MIGdGS@2`rB!4jbu
z3vW$aewaPh!;1(%C{@c#sAXHB@dIaA8}!10&KUlEH3X6AbR~$*w@4vh;&s?Ar`3$r
zaE(ojp&?nCa5Tef|6fI_Y;J+F;FMG#JZ*a;3+lC0IUArR$8rKBil`|JY_5WIPK?A6
zXl)iwx_d{Lwq6l)iXyjZBLih5$I84TWQ0^KH1vO1a=Z#fL3(N|oCt@KgeNYf*S@!F
zqcn!I-Sy-WUg-T$=DC^sZ4GiNTTp*xv{`QBn$rS#?v*8ij@h9yV}BFQWu)Ad!t~{9
z)e<dXpBSiD>HSuICZNYxpF;66GeP|I=qAWunJ#%?io`~Q<63UjtSWnHR)Yb>xh!jf
z@^6Dsd2=X35l)P4Rc7D!2eg^;AnnNMyCU6Eg)YEyJMiGm%8Cuz)n{nFTGe?Bga%Uw
zjcSu#CuZn(iI_V_YJU`;9BPK$%%c3?3bm#H5WVv2n;Z~1xhObV_j=>k)sjKXjG`({
zM!Noiy^V{fY8seJj-HBcT7wbW{;Toeu!2DyFq^TgK)A@U2$K9#?4(Xn)?}`}6+G#+
z$3^@<D%3e5F(0Q0r?~coU@zK08fZK_v?CNV0W#l2;UeB8^2x+%B)(-maW^m#ry?fA
z(=9>Z>HQSNdU&uK0VA1e)VZ~5l5NSsDQbd#ZmgEm+)c9^M#}cb(t(JsIFRFLnx!Ej
zYLUYJ*%=4(-DeSA-MZ2RYj;892sQO7yNDQ>=bQ}6n0rvg4cIJUwf7kNRKhFApO`>p
z6lus5N+iZA(|Y(iHxY+5zR`YAe!gt1?IayCIPeA{796FW2aU8qh@S9}{|ym-KJ2Ni
z&7-H$p51V($c}mXhI?6deeP-??B;~ip9%b^G^LdA=^|9FEWY#)xbn!C`Ze3J2%GPQ
zSN+G$kklWnX4JE5CEep3>e#2#u<`<ELJ*tHJO1-@x<HZZ8!jX`r;ot4{s|ZyQ)6E=
z)KVWI)YKYlf~&VVn!(-NT4TIni-Nfpo~q1pw#D2;5dBD;lU}ma%tNS;k>=Xye_6fR
za7IBXsZ0G*Mjj_$P?}RXs{pNJJx$0Wj)W@7&sorazl1rBzx0)0kI69YBOuHxWb0t~
zQ-K<!CuPD)Jtor^iOu1jcUdNx83pOz4yW}!?|A;i5*(SIEie5LoArH+ml@SuMr)?8
zJ?@10y*_otuL>d?&6`?#b2j_BWo<|tPRlPhSt1Jo-c-U1saQz8`Z}{OoH7m5T^<wv
z4dQ9ouAnD7%5K-L90R4X4}Z7TX47pn3)@^{KKN7iY!S)ja`thN(Hs$^r|gZ6enD-(
ziE{9YE7GDKD?f1Vbt$Fi$@`5UfA4SBwF+%BeY|TUd(8<-G=0XZ98|fjY*5^IWWx^-
za_f9A1oPaNl64KBhH||{v)_4|d!l|snhi7iP+_H94_OOs`yJy&+O(L@fb`?ir>??L
zRcfYZGG9J7jid*U25gdMsr;C0Xo>Ncvr(F39n_=9jMrF#q7$LD8_1PxAsW7gRoaRL
zG5;Z%5eR3<swCcCP`SH?%GSX0mDbv`qR*7y$;>$*Q*{2+otB2^Ot!{*1f&n>(S~0Y
zsumi3#t{>k2E9I{(!vPBkP~{Y7(1z>;@$t9KX3Q3$it86(m(kdqp(vHBXucIMK1o$
zM38XF{Dfuw4-zCz-S*GpGkCk(X=~iA(Q)C?R#rBY?F{e~QCycH*?z8^y|8CNHh`HA
zF()2p_QuC}Ld=yPz*2`&B5!tR?KE#ycx&Q&KB-V(<cNb;XR^^XTV{iPg4EvuIh4sY
z-}b<X29rYFUOcX3%U9*hT}s638MSs}Bd_Aps=fCH__tI!aNfY9R-d(yxNB-w0rq~Z
zV!NV~o&|gb6#CLG2lrmA4YbO1pWbkXu~K?0ZA}jzx(3`OBkO~|nvgD>arQd?pu5b5
z|HIwOCdM}tnbgpcob3X%?2Y?vf5ZkgHph|m-P`4~d7WVS@mJx>#8Xm3-L->ASyu;O
z$T`o*FS$1(M`*a~X@|OS31APLD;B&^^FXMY&SYuMON^`ctN!PTScsJ_ox>+Y(HAfS
zCBOCK(Yg#8yPlSz>%fB}w}S8|asf{rPgiHT=R_8mZuK6X7M4a_hNaf?ay;YAtbNAr
zNi<3?A+VFcC&$3Nc*m07?4c+X5WL}5HBVpE_Fx=`G!xuzpe?>j%V{(@j9g_1^rZZn
z;(o629md#}@i3fet@Qjw!qyl5V}I%T4^`@e_1jd+-r=V_`Rn`a@3yxblgOo)OaI34
z8v?m4kBjumF(a>FUZTs<*LpwMcr!qsdR04gjh}Jn7|ovSk1nyHc$*S`mUP!<KL5i*
zw>_)mQMWkNTgtS5&h|MD($xCS9%im6aPW-%bv%2iuQe>uZVMRWTtnN}pNp(mYbR&c
z_}tE?FzXrPOHF6IV}!S+^ADTHS+mC{S3$6yWe0O*WbC%=EfYI^3&@U5z52iTvx4sJ
zb52B6oZa(ty59C4oJBljc0Yx*&nohOzl!A(E$<oiV5544%g`HEDJvd!KuJh(43IV$
z#I##?n*^Yn`Z9KTe5yC}l3)}J6BxsFtsJu}3UOG(G@4hSJUBjsR(^P7^c0C?IM5z(
zy#Iriw!MsGaNBxq{z(MV05P17KIk@Po4hAd9@%;+E33ax5~b9!UUI39>S&;9EQ}6#
z`VJP6$9=li*!UZN0TC(ieizm%?0on}OB8~7IRYt;?+uaEPSCRY1QLFKrft!*T|loZ
zWPIX!%<NrzXfkj<@e&h-sFs5G?(-B|<Ku2O4kr-2IiPFMD&Op2_13Dx&Z(0L^!*A>
zZQ2!(RiGv)IW2XJ7w7&eF%?D^sDk02>m2-TdHOKgdRwbr4l7i2EL~vfD$u_9YUI_0
ze)u2qja=;i6Ozr&&Hca2H?HU_#Itjv{lg`W-$_^lzUh9|s}##T&dO$ctiJ?6WDDn~
zrVWwJe}6G{*?X>UUwI&-;~-_2^BFL8_2yPK@_#-qMdp2fJAFOgJ{CSsut5(_E|n20
zf%WaZYB_aV@0o`hOGYl0<tgoE^8%raw0v!5QwM<~b4a`U_vIl&{17{&w|b;Z@tW}6
zL2U{z@V@_6()xA)YzsJ%F?e#71JCcb%b1COX6pO@eLdUnL%wQ`QJZ#he*ZoRckDeV
zBfgztR?60KRQem;_U*eU%W?$p|DHQDLjHTnq)Bt!%!f;lJNDA79V)lLV^R;4GU>S&
z0pT)$-`$4?*>!xa9rQbE8U{v0>*RGi#{~1UdCgMrP~S1dGM5SG_>0^z)cL12pNo4I
z7~=pRL`fBN6{BgQo0CV&AQjevMjr}Qel0y&`S!@~K3q`J7#!GtYSKQQL9ShAa4%Z(
z(b_N!S*Rz!-yT>6KAw$iXrQTM;20(QP6blak*`%CEa|SP_kCYANtsNC*L@)LnD(0t
zD4T~>(Timni)LlYT1tB746_G>EofyxgbsIWAkm?jq(cT$5E-l}Q2coSu(!yPBW^@j
z+4!F}sbqUWbIK&V5z*IVI31(VU!|+s4#)x~YSU5@r-h65m^--&=CMQV7|zyuHESd>
z&sMRD71awJk7$q_^kK_Sx!|^^Ugg%lvyF8#qrCg~2w60iE;BkL2bk9*YCDTP<qN-T
zrdXJ$KYY&dA-UxPD*p{?CmVvI1Ai-a(0#ElfYSxJ86_BIo9s*vbz>J{vLA)0r4M&O
zCn1*q;xN|lf?cWc@tL>!rZzkiP}I6LvBYDp-)`tF8F7ah>`xh;SBMfvGQuREKZSo~
z1c|=?O8}OY36TxZjf6igFWC^76k`Tn!WV&vI3)5TUp@G@unbg#hz*r9S$KWmVoJc3
z5}t`LO-B^sB%6&yQ~Hy34a;q%qym3~uGylM9^q37Lbh`YOIVI}>}VIt<IifO*Zgii
z>wrm96wcPtbNpPFG}>A&M8_eDZsWm?Q3{X2mKXVP_h@aSxZ^0DrQheu&8vg#^zQVZ
zyt<n{;%f37x^tak08CIC`^Yl75&2}Qj0<R?v3VRLub@l}#}=g*SkV*H4L^Y1F9RMb
zc-sA+^!24q@-n^Gd^K*T!ZZU>#l(1G2(2`qQHNDI7!YQfx8StOfi5)6M*$A`c4+78
z_%Tk3=Q!k~4WD*6eoT_8#-X6QYo)FDnu!^17kC@v<sgW7zIO91@vFG0J}<bbQk%!I
zE8C<{PJ~Opg;=1vo#`a8Sk6CE%;#TdC-QI9l0<f;)|N$)OL*2_B+7@w*a+NAZSVuH
z+6#rJP+bg+^+sZ?=qPGi47M?uQXkqIGfxQhApuZSX4dbR<oypfSIUrG@wG1Vk7prJ
zRS8k%?Qj+4E|~XvzHL4-q*R7ba<6H3ikv|sEyJMnl{vN$4;%-0lKd;u<qY!({bEk_
z!xm+DBEk<U2_cqsej&KEESwY4WpFIdFg_8O#vrWT1Gd<u3jhWr4cMFLZ7`9EQI2=n
z5p7#ts^Zu6`hg#pNIe?GlwYei_9i0-BAOH@hxr9(zqZfpLx~b<anZEa_aZG20_?ov
zR;_VnQn|AL-0J;)436}bgr37cMDM~D)bgU$<sE2^gQMN1eHygm^5dAfA0!wO=qDR$
z@uV%gjcC4@lfY}WYIaD{qcm(jpV!F?-@@gzP*I|^te0oH4B^m3Fs6c|4v~y)S$w^`
zaO8Knyo#G}B$Q>@;__3@Q(-n(h{Tm;IV18@%uz%bt-?0VU}SAPlD1>=u_O{74hY4U
zw7EE^c0rq!#a*obK@1xA<$r6k7X?7|v+x2StV0el10-?dz;$OHX!7lEw?H{$Ibp2-
z-D`doqaNIP#svukuLJ;P9lN+&o}>#+P>mX^pLd0_XRXh)7F+WKDplm_3d=b`2qHy4
zY5<z;^iB>2FCP@upX3toGl2|VK`CMvOa&zSR^V3$0FNjhC;i}7?3e>iD8-z>pJt`O
zzBq0FH<zjeXKSo_VvId`AlE6`^ZdU94{}tjdQu2jKYAtfevff)vb_64SA#ur*lt`T
zd6%l2A^T)0_5YZH+=uf}q`;xFkSkU<647K~4q76OXs^!CLBnu0phm`PcOk{&cy4V|
z2qif<YBi~27xfe_tEd>|O8EJ^I*1E|@y}0gCOPAz=*$o1ulh=Jz{w(mE7y*@BwmG^
znda`xT6jj9qx$WJeVtQMiOvyWg0Kk}?G+VTgt}pRhZImQ_9Gu8V~z$+ImsC#UmaQ{
z--IUuEX5WDl4gPgj;31M71<X9l4geFzb^=;5td+y086zTQ_f#KjRC6-2b*NQwGn$K
z_5=N$@LGsC?o?0qk8#_BJ`LF<+M-6U-829*g-=Kaf~$Bkk+~37EKsRl8SGFD2&0C0
zjH<j5<XgSA-Sp!^q7SL*3tPZhHMS{=wGrE5tyVdW)Lq^3tcOw~OlUPYCRzmvD;f3h
z%pWyATD9IwRX!ka?bSIX%$Fp2-t55lolzM5_b0!pKgK$EMY?pW_wq!49*<k*Ft5vN
ztxX5Bc=BHPgHlvm+PEX0Q{skt+)Y`rxOg>L<)ppchILMd?X#M5u9NlF>=*E_pk*bR
zH{&qIG(S{AqZjFQEs@CP*IVLMEUBb(;9L0X0T+xqW6@y%mPwT&GTbY?oJ1fylJvLJ
z2#P@{1x(WAJ;U(u7o(qq&$M%Fd9DYGD=5*CDfN{4alJ)r5p(=n=2;&N;pR#q!r0u_
zWK0(xtW^xSBBy$Nry?JHLbTXG=ky&TM?OoncTY+px!}v<YyOmnu<z7t8h_3N;i57~
z{Jf`qJ5V}MrJrgl^(4df&L@#aGYde|R14Ll!GqZqUfiJrMx*f2<dD1Mo}MKp%Anc(
zo)X)dhFgmyV8Q;apR&qzjcK*7V&yS^=__N^s>1G<r4|;M8D?+YWu|W28Y4(c8V{Ah
zZeFBVU=QcGes!*lpS!*&nnMkluucIYZPsN@vd;SR{P5@jlbJMMV#%wvzzq*v72h3C
zfuv(kW~9yL@U}02?W1;!Ev4iyV7*j!9@FA339-Yjl<%-2y`5u&@#`Bg{2q(lvBwep
zq0+{K<R^iji3wafo9l-ucCL%G7}4VOWFc~v2$E)tD~*Zw5I0Lc2sWceGc+P@yFsoI
z(%=G}Ic#q+HG$Pta?AFh0hOFeCzk$_80<pY<>#JZA?|J2nL`%mT3d=!y?%OV$nu!Y
zWrt8f?U|Q!PMxebCYF|ow@q#>U&Yjm<8!v5QT#fXm@_-{1q9NU+tf(W4dM~LjKr>j
zW>wbK*P$p}n7dKad`bx;#6+M?RWwg7^WN+PX7?Pg``{)ItJ1@gFEGrDH~mGBY{838
zly2*ocj^%Xs}z+07DG5rb5A-wtBR_9chNvW5Y50D&+$0tZGm7l`F!f|ZIr~Rq=^pn
zTk{&#b-WA&O)BSx?a4W5QN7v1T=z|$<LQTuCf_fkf6lx;JXY4ttjRqvgl0U`xtQdv
zDiDm;Tz`D5oy^|2##Z^#qyl-3>H(30)Z%d%6!^UnwAgU9N9dnaeLF0c<LwnRx(*QF
zU1y{`3^H2<9lfJgtZxw7IcB0s)-X(bC#kc&w3tzbg{-rGL7g{GI^9;b&<>bKt~i(G
z*rk~mBEd5NRg=cly`>pd1-qq})+_&}JjTmpU)VJVnv^VztB&kxqXGAKg?sXr08Wy|
zl;b6%QdsyBN|D*1ny^HdbNVqCJS++UHBHPJCFVWFf_U2@7a$MztRIxpgyvYXM$fdN
zPGi&ng8W$}Y|>Q8neBG3v{v-z8-be*&h$WyxeF#Rdk$n<!Bfmor8poXCgVzk5q!U>
z(Zyavv&L(|t<KTom6{PBv*}YVwd}Z6$_5aEskrV!-@xMmh(h&}>bW5Jw7Qwpm3N(D
z_lqr_nrx#K_;3V)kNB?KD}3@|axi;LA0=yyWv98H8L=nW=5SYv8W+EY2ukMo*|BI{
z6vMKF<F_M869Y`^YN@@E6;EiE;Q{rX*LUx1;?O+K&iYwxTbOT<OO|7qm}OGchhjQ(
zof!Y2r|I#Flp;F=tS;Q2l)|C*_iK3<%fGXKOe(W~%^`;rcZ84KhPq1~e`^Y%B(43!
zVN-NYqadf~e9BIF;48CJDzGOk>y+5mEKkjH<<up)N_9f`E+%LCc&Njp+h6M(%RK0@
zmQ+3vy6{G%o-ry>@6tFrt~t$!B0{so8m7__B-;us8p&3_pfXY=q|GVW?n?jhAg_Va
zOKlkX;~=@_j+_-$j^wI`i0C>ilgPXDg^7p9?>grh$H}>)zsg4dKL1Evz?QwpWt>X2
zzNn0w<~4LQW6>*ZQHOm-Po0i0RbkNCB)aiq%KbjCs=^6t5WP#m^_L8byPRG+#dhAB
ztGpg3b>rp|<Hw)64Pm*O3BcLx)~r7FlpvzKPS4&w02>L<I=%T<;b`%&=+9Q~r00s1
zL&>af>7@a)%t~H@6?FxJk~tT}VeaGh3Q`6E0M_BvDL9^N?i;k7%wEsz5)4ntZSiG0
z;Y)${a#3bQh5IkXtW1<N6Yn0JD}nc@e*LX~bZ}C%M2kTMnRv|1p-dYd!ex0e16i7N
z0#A;QjYKf5Yb^3!xvTawfT0rqe6!>6pXUS*TSciBH8tR~)=I+fJKTxlr7P(hHF~U5
zWT9Hy5w4n^%`I)fIX7q9e+~Kc-R<#JihE~ZJ6?YN0=mZW2sja<%q-59UO=b+liq4O
zE7hW*c9H6J@z{M<-kOQJThdAbpwanIbJ$xtHuICbPSCVPl4?;^ySRI^Ke6R01V$(R
z3ipsk@`OgDor6&CYg5zo6)26gEDp{ruQ1__zjFPNj#=dq@N$N<$Tnd?S$I^DY!X-(
z;~~7q$@x1LokdTy4Z*mfkURO^c2~>zFj&lB_qeU`;o~o;AjoFMf6&4lTrB^;-}V3Y
z$Nz(DsJYoQE0VCXu>5OdcX4$i;pOG}e=Y9ptXy3GtG=8WJ-v9s5mf(|x^$`g1VICX
zwKxVjeN$6f)Nv}b$RlVo;{M@F-LA3mx0kodxr`?!_JB1aWt86nuHK!!wY5`PoMPA|
z6^aJc3d=muY@B%DnVH&1@52VHP8rRrI}I|MIIrfJ$vLHaxxrkaAA|@_S@&@x%Z$`r
z3T`R;peAAw^paw-^Ws0=gm1+Q%_wl(=Tvf}!vk(3tLZ7k^Tx>oAvKYL2xh!;IHXux
z_=KOM5|63)Tg4Mye6sM8j{q^lK5?a`5GKkJYM~G#%cZnNfyKD-I#H;@#^!1dt1gE)
zxT_ir%Y~)B<A3PPiqX-Rx755C*d*;u-2-DL(R~zfHK>p#QX4=#=TfT_M9l=`6S+{9
z(?iDUgz0%9$+w#B7~Gh0C)YHoF9lveIfWG`LG%HsFit3w?$bjJldqWb>JJF&EOY;y
z$)+F#4gmw*4v~<XjeL(mMNOlaY!IdWV$f*G6L)KZ8P2s}Pg(I+1u)UMY|RRWyf2{W
zE_~l~$!)pR6?`4dtaDUa<m0F5RtBnN=_ia9dv4q4O(qrPUPS$@kS+R<quA9QY^5RM
zI}d!JB^K7YHFmW?OMK9-6?FrJp^>}(AkMo1Ssyt$H6=9#a+Q~ZRu)<C@WB4byNi0U
zDO!QKEhm|l<DoDOSSnmkW7GPjNi(nn<<vVszZzkca0O21p)L<`C2B_llxIQavpr)^
zr1|udTQw+n)~G@yZwSX=!FGp_=Z2s6_x<Oeg{rc-XY`?6V#O}zJR+oz=%xxn=D+X9
zw{Jy5OGEyJ2eBU~R^eTqo-ba~bzZKXMy_4~IM`<mHi_-+o{ld^BX3PoA43P*z`X5U
zzm{LFKV5E!#BX}vZ>j`B{Mw+ynufkFoibQW0)7i@8Rho%{%q|7gk^O<#eY3uD+%%b
zdf=hN8xZ&Q2|8iEHJ@E9YS({j8YBO4q-ON@Vz-_o9ZA8e&Zrj<wew?5AXup;+TiAT
z$V3`1hlx$M<yr9kwsTpkyGzi6I4>uM7&|H1pD3Ot#|KXZFPK_uaqM!tXl<9^{``h{
z_M81mkhkw=Zy%28dGhwI2eL<=VClig&{xxtnRuV*EjZ3E)K~f!O%&Hm;Ksqk^#N?O
z(1xG?_sZLK(;2z#HaX#X1#b2b!>7{AwZT$<e?1e{My@s1W=;EGufZftF&ZT~awrCz
zkITs@mj|&C<KN5MKc5H@-*2}f-!BU<;X_qZeU~1)Jv_dD-?Q#7KgYI_>wft@ot%8+
zdHnt9{rVlP6%Nf47wu@ma1;*g!Uh8CUV!|jrXYQ;Gk@MwM+UUuTc6pR^k!YAgN_T+
z<bU-EG_4pW9DEDOzlAOBZq)jjtkvEeoH+an$lIMC<5bx8SX;pA5{zs=stw*5S*@M<
zQ(B^R+folUdk0N_cEQ~H2C5h>Mf6D%eD`JQ(vLuMwPdtn%FtqcQzG$q+x{dCtFVl`
ztCd^e{5<WS0kSOL=!c=wUGEaWnwaP{!}e^?u3FO;>N_IoFnGq;)*kg8F@$Q6kGtN>
z2{rpb_OQ?X`OYwjlm*|Erk{lzK2U7xHVP8P$)^7_Cv>^p`{ZZc<IuQA+#{@_Q4EPu
zQNW-=&05g;pXalavym)pq11=y8;`o&UtM-Bt*R<$KL?It0Y%imKHPWvcrVfglA^w!
z8Pq{T`TtJJQ!?t@=MU1z{N6F4vI_Ho-bZXE4cRrJ%1C)<>Y|!VUZ=C#+3VmQBo3mJ
z5Hb*^PjTP48;lz{Y8Otk)9E4jC2|L|oz0}n<m5ftBwbTxliOUH$5xNX`&l5b_d$kj
zQ+I=I(`hJ?S#2g2Fv0fBvFAndAbb)y1D*%%j5qHaUVHtxl)vJ|GG*IT&Gj!ZWSoox
zq>g!AIITv>*@-errWgz?&-molyex9GR{uM+AS1W2@V860YLvUl{;@0BM^TZ&rtw5Q
zqiEOs8L9geCfogjICzB#b4|HPx{SFW2Sr0;BlD8;n{cI%9jsSQV;2P?2-5a!APRSx
zPJ}d&Nti?`|FFL=|FXP{Fmn?Hg7jk`7GDw6FfTm;0yt6U4S5zCZGk+hOmr89LuYEj
zRcI+;{@Z?PDL`AS_4&9ZxnA~aGf-F<6vn$MauRZ7dy%3tOZ{_VfM~$UXdNY#bmK!0
z!&qrV21wIPDhZjrD;(}q)wu7~zbi-X2S1n)rHO-D;R8H(O1hq0r`p&PUWDt$Phu{Q
zeEy+?wY-InfoPPP3U1>BTTJ>?4sozG%w>hKSoD)GVn|1&ld5tRsCaG?_(5|pgP|hF
zzy%RlIfgov^{g77E{yZn7=h{aqlOF(y-u8-8_`bHBrD}J#a9dg3jAm}SQr$*(!h)f
z%8w9*2Sc?Ef@HqRs1m?EdYYf743f&wBSzN-lfZ~YsfK~5mq{sg#z+MfZ0P`=TssnJ
zso0})MO99Vvs!>(Xibw!kSTmA-Yu=%>4JAF3tOzjd5dF?w$>_AZz{1KHc46T#Lu;;
zENS(DG+_c8l(376i}%MeQwnhxCHQGEq1_UJrn2yDkrTtbmvfkA)WqD(M^2y|I4aY%
zp9d!$q1RxV3PH3VM<J=}1CnSTA<~S}{uKl2HS8sUTpvb43Ld8(sl1B<PG=2vXD^LC
ztRo}W(3{ka;2TxkzyZ0wSK2@tAxX2rCLUlEv^yjbJ)w&JTGOS2rJ~S(UHQWXb=_V;
z!Q6^}x4R|kK-(FyVsCBBC)93PH`&G^91I_@0bz9#7YG-2Mz4<l2^tpACwybu+xyr_
z8r3gSf6}$Uw&Q5|;Wr|%mxmX48lZ|J0*2u#M=bHlI~s{3YxX9Ic^9PV7AOr10fSis
z9mhEBE`}j<qugZrWU5mhyhGZ=tW2VVlY~{t3?aYoOWi^WQ?SV99$*=41%{!@KO%b<
zJS>vrTk8HAJ{HyhwgbJ7OJ6Onst_2?Q>QIq{4nk!j4Qt4D~IAV_oJ>@Qj#+5yCG&1
zA&gLy{`_$i(fuIo4Zaz#IBnM(8m?YDR{|OsC}ge@lR1R5ZpM#bOl!<`w@Px$t-aa^
zb9wy706iL}GS1Ha_m2Yi_oLF+Icoi<-wACG_Cq&cn){N`my+>U{}mrDCHtlu+=O}<
zUUkN<CF!rd$7V-u?+WU=W3cO5AK)W4xwr1g!>YovW(euB_|*kH%YCnv-?j?wJH(D$
z!zWLWQs;V@^L&i_b$R>vco2Pw^uOHDY+co6k4v*>6nL|;T)4Q8s?PxFS`H7B_J(i2
zNw;gwsNU{W^IsblSgPnRle$&KsxN!YWM_JD?brYHPv&~m=p4QMF~+S+98NTj;Z)}*
z5B~uh)Q+=7zpLAP3X=M0TLLu$g@|Of%>O2}#@MCt^qbJR-d8iAuu(TS4qJjz8Jv4~
zG76;0B$_IriQOh4rVV_<nL#eg__A9^8oJ{=_iU)`QlsEH^m3Z0ly5+2;2j7cs|;Kt
z^Xli2#GSSmLn-Gid7-~i#u(g_lS@7&Jc8jNBo$_@9y$*r*(2c7e{{nGk(gSBtlZt3
zUZ#%_@7J=QDkU_h_XWdrn3&r@Lw{=|H%3(-AVvvO@p4P(NBlTnyvR=pF-;-wtNa0}
zXBesRFGa-@>PwB<<ERqHj)3-zgux=8f|m~s8F&j+3D7Fs07->`P)8BoI26{8mP`7g
zd9x=gSO)Qum_LU(kgF1A0SBoJ;;ToS)JjAY!5Mrm=TRgfXxW2nArB1Kt#^8sVd#Hp
z`$u1dH=hcF!+;5pHKCen&tW9<7_Ru_XfCsV4I?Se(Y+wFaPo(XCX)wN8jEF$j0E4u
z7oc8cEg!PO%pQU&u;84O|05>N%1(>#hk^%}$<`QqWa(0wT1?gDCkm=4BgB$OGaqRD
zbQP`8_6etyTR~(sKWL;<nlZ$Ur=v+@np9}iS|c>!6!HRr-$nv=r+uq9UolaialF?o
z#B>n|iip>wv)0@#=435`!4Q1IyA2`B0{1K73}jL1u=FvoJ=8ij#Sa3>n1C1+=o6fd
zpA9teBA}ZW)&)aAZ?1&1hY?Afcp!IsrF2;eX#Rs6lAV~NHZPA+uoct>fEOV<p&yf$
z7N>_IjZo;1Dxaak6t8xS0QE=+ogrvMN3Y?G%m+J*`xPS<Qf@D|3lcm2*C=0g;;$E}
zNJ}aTDw=T`%*KkcGp9YZ!;28Rs`StQpkuf=dH=tT`ERJ~Kdr+4ztSozdNT0@U5Ni`
zdgVeajpYXY#wc^Y%}BU~Ua<~ujfZCkknWSjzWvKX(MnX^wKg9bad13=vhdt0#I&tj
zy$&uPPA`vV*E7*vS=a+XTAO9e_<Qh?e}ji5t+<g%vcs;HUapjoRq4WiDZN@1FXR0M
z_#?mC5Z`b@T!Q`vzh&${=;01v{9W@1{Rw~6<?YqYZ5$&{?j+c+XiWU~^DN%85$)x8
zB#aTRD8-F_GMgK0je7FE)2r0yGr`@ts+dE%d|+e-Ua+fso^v%vgRIT~OS(F<j03*$
z9I^E`U)xAYjZ)2X0U*Qb(<geD)DQYAg+}_Sw7FrYvvs3@^C|u@<>&nG!d#mp4j~tj
z9_UF`x&78^J?rwGT>6oEqlp-HOxOZAEHu#>czD~igzX7FT3(yLX-M973a_hRC&`;*
z%1VR@Ytmq4%DmMleKt+!8FR1j%Gi@>{@2Iasa2=9LyZ8pEza~YPdS`N4J+yF>mQTS
z1wS?wb+9wfBv<D=hp&p+sjFtRKgt6%TjS5{*sg{%Q<qh!3P$+0ROyopw5bv&>C*TN
zdy8{fpPDsTEV3&qp9~m|?aApp#AbsyEqse6@4W7{cXbIGenaM*u*!q>b|=q`wC#O_
zH!|MS-E`{8%}K8P09cHi2<+;-2HMX%`>M9a&P*>10T4=i)!X1iM=ZyG;Hx`koWM<o
zn$K{Y32H<!*qh)1TPEd6F1oGVw+1J+M3u8Jp#BE6dVfH_ulZdtPaHsL+sjcFfgrPU
z)2~$;RzB}~!et5>IeJ^fD%WS_Tv}Nn2N&OGo%Gi18*c_;Y@A8*s#m-%C^psDz4UVx
zULf_X`&>6H2kdklaBW`Vv^+OSq<Br#q$W;vG%74}TNv&L@ps>=zi1pYE;8iucP!p$
za;r;TZOyj>E1BY|Cvn10+BVLPvo#2Hk|P6F507=OD1{b|Y#R})8OBFRVi4>F@#;De
zlH4>FOj2tXzlTu^-LzlN`JO}SG&S2ato57Ff@oHffi0z+zh~KQh?@|XST}r97DhZK
z+u-aBd9Y)^Z}lq$y&RI_{iaA{TJ|p?{j$b1@qYJKRrS@cQnoZhOWInMYcNixran}M
z6{Ds3Q1`{kEI}KNLQ%)qbCp#}x-HE4de8dMz1V)V2|P7P5ql+lmii=l2kEhkjfsvS
zAc}zx*STo^jQtS5GIw6c!p1`uP)_{JXcE_>uJ(912meFklI^Rtdg}_bdxqB}eAhc~
zm|vefA+l}Ck7$v^U6HdJdL0`+PUKOvO*i3T9&$KoQA=4Ou-s(0K|f}r+r${SoVWaP
zo?f_=v@?m#>sZvH#lI}nD?kHr&aHujacKApcmKz_gS>fSYOOa{M^d7avik?opRcl>
z4Rhh+r{}kiZAW~xZ<DFzqIow2kB_yqEJzZDC>m<$JwkuU*uP0QH#?tS7eFE3Z%hcJ
z536%J?qpLq^9LU<ap4kH8m-N?_sft6q&CQi2sF-i0dcfaTX4OhldD-8X5per?~AtO
zsDhvU#@UFgByP7A<c$595n&u=b|p`}canvb%!yvM=EY?ka)en#&VX3i__KH+H7$at
zwX5qXkCz776~3dUrwKM^w^v6>4HT$uC?*}-MG<LnoPRGvU}IE-{-lkQZeKLY1rf}U
z8MLou1#fP7@U@gGJK8dz>OFfC`H<x<ig8A|rXF=nXIPmvgGT)f3-;wEuW!xS$axw1
zS?9O|I&>_+9Y4%Qa_?Kl_2|bEUxDy;#fwwWVr{iyeYH4$?oGW7n*RO4s5D*ow&$^6
zE*4BceR;<PSOy^~ddMJQTqdOm!C`I$bWI4C_pAV-lqL-?mD@o#tyy__=|gEN2k|QT
z?0YE{izYeen6olA*T~ugFdD~*m(rAjogW^<6?Ul#MO@t@H<YYI_8|p?t|r4$O5dD8
zdt>)mJ5TEOsKKFBR{#<3_3}e9D}1Jw3*k~V?xs|q6j(SaW^MtRS+35;N&O0``wsQm
zJKZ9&6GDdfHV(&4ybV+}rg!c|s`5#Jfh~P@u-A(Bv(dfsHg+{Y{e^2p?92XOo5EvW
zIyR^Zhs+F$wws-33|4$>AZHjQA4Nv0jw2AJe~Avy&X3<l<Uquh%)>@X-454D+>s~%
z0)MwbSTD2<mG-Zh_8@Owy|^&HwDb(LnZqzzfbo163uuV@#b4Sq;m@ed+d42R1R@Bk
z(PZNw?+12{GPpS~*w!uYEUsAxQ8dM{JW({he1~Nj5)*=Vi9KkIKAA!{VZlzCQ^pru
zMv)B%i9z!A%!o-QwL5-#g#xwHiW&`+40QBBGysIwIUL}Tj?p8Cp#i$C)j2_I<Bh7c
zY!H0=_%yBDV&ai)G$!A&8#^<p&w}XmtX;;%H9%muY9;L}hX`7Nk+MbFCfT$`GTdQP
z#Y#$B3*4ti=#l1XW#}c%#bLl#(JI9=<!1~-cRUlW)%z=pz&_Kvy;9{M&DK!@cTfL%
zNYd+uiF)+Of>A58C-I3+)zd+kbl2c<4XGTSiS=8_j;T~+moOiYV$d@XeKh}OxCjbd
z2+81ZHf{_;Ee!VqmUUTUQUTpI?MNQ%pk1`)o-NTrUj<PSOiWI(azPkNv@Z#oeB#17
z<jErg8ax+z$~v|&Na?ZFJ_6eocN*)l<rC-hE3F-eS=)SmP~Ouhb(@aBm@S?Lb1=?z
zas(?1N0%{!XvTpQb_5s$taP^S-GPiEI|l^gTi@jIzEBJ$nH|))naO$;-Hp1X*Qh9V
zcwhV^ev0de@qWoPV8pp5k@DVI^plz@ys9oE;Ghuoz+9*$$^B69)?_v)OwA2M$FmgT
zDPEJTTidbq2TSlmbNvX($y2Nxy0{YE^%`21p|Hi8um&=C+ySQuRMYS-Io6-*3y43V
zRm3E)s>=6s@8f@NkZc5ylB1*bDgrG&$Z%p_=hu)>H3Ufj(bEi5Ghs_doVq%o^R5D=
z#2Z?cQco><t;b#E?!fi6?VHvE$MiBx$NuZlmTi+OwnsA~_N^U$k3L<RmLXpV?w@XQ
z7W6qy$7P-z$|5Q>b0bP;1>%v8-35jHPx{{bWi=h;1E)u_6m=eY!jv3J;4|0xa3*Cw
z^I-9dX$9pvC6`i+4^y@@Ds)QN7xL$RQAIm5K51o#9jW}Y!ab9S+MN^qh`e{@K)@;-
z)}f41T^_GHcvzJrg_EwqVom;ilOL_mMQw);J;yBLnxN6wl4cg8G9p&_B`kbOf(T7h
z{~ZN#u+{C=7BcmZ`o|=tI!iT_ZsqdZ$a6ccBB@$Mje~S^A!kDwLE;_ns{BwF6C<jQ
zZg%#m)$#ka=knI@P#Oj1{Q~lvsWPa=zYbnn>55vyacYFBKhnQtHryR+LZlL7d$4i$
z0uKT^b$MF>YrZuQHdqcy<NGTBEzw<2%bi+F6|MCfv7D4CxL+G2?o?FM;uCRfI${0d
zNatTnnLUUEub~w+PV}q-r0qF@HuS{;c1GO~8RAb0S^IsCwzE;j#-<M^4^Lb7Q6-EP
zPP6x-{W!Y{9FvA!>-8tU25(`M!d$uex)smnNm$)L+&9R=ciw;?Yf4uI$1gEDd^~n(
zC44U}Gi;{CEhkS=29?P$R3kiEV%e^(xIGC2(;<_Z)qba#dMd75vlaqx5(kH-#BiNH
zHtvx7{ZZT6UK(@FaZW+g@`cW=M()`oVOk7?Xp*eg#U=4Qs1^e6aRx{zd0cJUUbR60
zi=&Niq7z^7F)q2I&X=Abt}*t$Pl|o$8h!&_dLi`mnl6~b@y@5)Dr7%b>!R>^$xVX_
zH~3;^J+2v!NQA@1;{o#DaV)@I;%yN1Jwz%@t@aM}+EA7|Dgep>T<*zza9X*qNn|va
zB@lx4;+{e}1^}D1Ya|~uygx?m4CH0Fn3dTQq;rxD>#D}7qfj&do>8Sw^))?P?`{0^
zIDalgawQ_!bLmxMk0^?#oi>jEXzmX_^i1(0Jo2ol$5mxBkq3LW3v_CEb<=9>nC;t%
zF~(<d?@7qQbTF*Vc>nTeDk{mO<f06#qjnB*+PSgBoeD4v)f{eJxa8Qn{6o>hMOpfy
zpVy8mSl=E|KiVZ$EhBXJX!r2^X}YOEt=l!-lb`}~>3<D1MBtypsg9aekmUUJ5#h(_
z@(Yh33D+N%>yzn6JJXt3%wo!<zXJC@mB-tyV@xI4;C~V-yd3{SsPM9}{qKB({>xxw
zlLevoMC%nf1(Q-0IiF>|cqV`_d@IbLZHejQZ@3Dw&@zux83f;D{$#4fuTMc0jI;9n
zwfr#V@PngUUoH-w+ZaV;U>^7+BIPfYiTtAh(jZPAeShR#qm)I>4HZ0Eix>{-HA8Qr
zJOuBIKsN9rra!NTFc#vaM{hJUCw6{Bc`%XczM?zfRD!(U?Y;u#rdtmgUweVUO_oz^
z7cdaN8W{@fHKT^f&@lL<FwT?~ccko%{<Le$Tw634<)Izt!lmId%+QIix-s0yRnuZT
zx;ort8nN@iJ9@{mlQ%kPUdkSr<~?3MXgB9veeUpNNyR2VP4SCfFFl;N%pB?N1UjZ>
zvXk3`@FHmXDCfcz75%Yo1)JHq5CN5W2%EW;y4H{PkG?o`R^u-km{5-8sDs^{UOTNe
ze5@|UV|qyvY%a$cb1T_R4_Qx9Hng8=shXHU@-SPVj!c1WtPK~(T>L^)P{eNGv5369
z)L~SGp}xi7lU5<$W|Y{fAo$B|^H-cU$ts<#=cnv>!@AmUsCmhkgfh`i+Br*>Ekr;x
z;KiPx6a>p>I1TTtmB1vbIP^JOyMs|7?Vy1*)I9*jdpHHJ!XT3_O+K*rr)(BPdH*XB
zTy>!8b0U@+W{_a=_gci_xGqWmIuU81<i=6Y2}q4EAsE`ro*pS>Z)Ch|+=e+ndHr>R
zj9qazmC8pOT5B{h)n9lEv9QM6*M0x{ZLVG@Qe)&!yL{++gS;CvnZzszVMM-kA~<#F
zyrll>2Xvni=Q??|T$0n00Q$`jy5XAH&~%0w&GN-Qg#z+4<?mfN8nc>qzQ0nGDXc`I
z!1sJf)eVc<i5=JHgkhh>5`iE12*%#xS8&!7W7P3J0CqyKb|zl$*kbv0lLcVg8^b{p
z)cS7LPjAQwfwX$bNdP=C6Q@+AWZCd}FP@wsIEMFOET?LmI;Bl!5h;_MHlY;qqZV3p
z*SrBZ?Tj2B&)Djq;0hQ$XDf!aMHOoEl{46y^jgCsJGG1UahcU-BT@|t!#YtmeX%}T
zbuSXYI=pZ#-whd9llUx`on$9(<h!`h0(7zFM0p+u+7-`HPV3iEa@JBtLK?9B!K*m=
zmg~7__-kAIR-ElJTgB<icu*P(Wb5BLFtd+Ux8mz^AKt;9@PXrVBvFq4=dx{3=3SDV
zftA*A6%GSd)b**}1T`6YV)I0K`n<sMbORhRf}m{Y`VZeZH|zg2HDF`m_}?WYwsof*
z4!AJ3Kj<I9k8kf8Nm%%1he121NV*E6$*dPgicjUSzrKi6ef97*Ba=k%g&V|^yO9lZ
zeSTU90uX)Mw=ZwFiIG$h(|8p$5nkfK_u=FI4lh_g8b_ula%>49_vt$`A=C;mjdY~6
zGh7o2J~_}&6hIe1@9PzAuhNOacPf0r0<*41OxiSxF<!`7hzroyx9eVn9Dk}a`B}z>
zm=})?XpfL^+cuwc5{`s;C?aIj>MPpv`VMiA)#N#vbQS4rmKb9x)KD;)WON?u4-p~y
zt$<=*F6<EZZjV~I1z0RKIbydPcX2g6Gu8?>YS|r+zOb}mVN^e5`p`dVGdxt`R3G$S
zSzyUbRVUq;sFK^Pm!2DDd9N5+635EN9}8UCQOs;m+edEP{wP&?8niwfPVojh{h%OR
zCq-Iu-SVqzR^XqYz`iloqmPv9vn1jar2L|P!~tH~H-9+IJK$uX1!eIPtm%@fswZhH
zZ|vT*zTPS`jOA!*IR{_JB=jLqQnqJh<3sRG+r(yL(nKasOr3}rc_lNTg}R7HcEVFJ
z$;K@9=9O}Ijitq2|B;LXX&GIFjWwT#S6ye&|HGiBC|RVY9OEq`NYs3*v&!Y7wu5+$
z;E6i`Q2jOCxYTy=0tAvBv-U^5pb9#4m1TX*tSR(?!A&e8A<yCOR@c9x)ndLnEDyEq
zhL6n{2wClD5r~!k%2a}1YW9s2w>Wgd=9X5A?Sio@TgjI+En4<6sZVW=wv=FpbcaXL
zeE%sqBI4yHRJ0>CMpLKtj=gNJE2xzs5sN@kmM3#78Na2cXNeMG;HZ=b5RSN*I?`)S
zIJ2V21jb`>q{RWeR3gOI{fec@;$W-lALJ9L(Ty$Mu;s0FZP6F+`9fkj#}qSI<a3o}
zi@%KFtsbJh5#)7`erCrnVv<nUV;-er{$XStCaKN2qO|Ej|3l~9uZRy}52`{YHuKtu
z=LA>#`&7Sf>WJPW?P!n(;-=wfXX9dg(=UN2sbquvQ1E;&(KQeCiO-8-&r*?=^Bhav
zW1$b@tBS2`I%%VqYtkKZpo(W)_3{it=R(fQUu8<#JH(%A)+s`M!CaarN`0?=sZ{W*
z<ZY0FyX6OGugK$17))cF94bdl0g4Qg1v&ae<vK}4#%CHymuCY<Y}{aj;uB6g=c2_6
zf-K7CvI8luf2-KXRoE^L<ITauW0LifVVl`|smSe6{w_SRGg%xC!|+A-cAd)M{sa@q
zuB<hE_mN;b)<zfL+N&%AZsSfoVq^iOjp8oatR8am<th}l7IZ8h$7NQZ7MYA6F&aXp
z-Vk%cEr{lhUNAR|dfUFJEBqb73kfGY*Edx`p<)f39X;94FR+_F9ZJy|`Yw9Frfw_&
z1NS~kh6L}fJm}i~A(?VBD3YLuaRg%Cs|a%sPES2Xpth=b5N3hq+LGM?WiBPp7|%rt
z#+jAj6}IT<!0@3g8y#21pF)fcC;8vJ&MBo1*?U{XORu!G==y@5-FNWfd3+qb!;G1z
z#L#Dns&)R?i!*_@obRZEmc=WW9RG*0cM9$;XuEx5Co8sX+qP{xE4FRhwrwXXwv!b*
z|JY8}&e`WXs{NkY->iAjRb8_$=1o6O_ZYwN05fZZ)6gXG<CB~Mfqv&M*u6sJHtBgJ
zz0M!xNS;#}CG=Q`K^5={{()qAd3^%*IY>YE0rmnQoEH~Dgz$5BN==ERvfs%0@NPkf
zUIUG}GKX`MF)`xdtc)DIa5&)P;>z7sPd_2QnNBc}Wz3uf-A{PQ0aa9hJqy}erJeLH
zCYw|-u+nDi?}<6&v3H3Wh2qcR{`1piwBm0}_W0ow`VZ%n*WvQUt?AC>fE0D*tHou~
zhObazx9Dq3x*+qyR`wPbvr#PYHgbo@qmXV|I-G26>R=U=6KR%`vKsBRzP*6i)d?~t
z0AqiNGWd63MakLfMC<YQK3AOgP{roCbC$)0%rQ)ONNt0r9M%bpl}>l*WY<leSGnA~
zBZaG)IRZ`t9}C%pvv*B4D)kn8YE*LVrt#(W1_H?3>N_sv6QIH=hCN2tar&Mn+)(6F
zum5OVEKm@(-d(558byrqoV^qrZAqTAl+E^sWJOK)c8lNK!Hy@44+2{i1+F5znFx@P
zBqwJ{WPI#cHaeBTdy1~g(3AlXhYE8`OV}x;N_8=Z*XXC0tdaa5a4l|g5$AnXbUXSS
zX-nDt$J2SY=z)H^@%-lzcl{)&`>_MvQslNnoQc~fgN*rz8I>Jedg(fOVp}q6n5j4G
z6yHN_sGmiP+J$B|9n`lW2@&9KEPMc5(3waQ*k7P?Ql@zFST?JIU050wfi5h06@icF
zGxX0Rt?^5<`~a(gFf|+Qu1p7b0S=TrCuf`4_&V0a#p$l;Yqe4lJ*2Lbyu%8rP%j*>
zI<|@*n*1=yY~{eep;>syK?t}~BS7c=No8Z<iKpog0`I<xD(c@qL}!5D;DigRPv4V(
z^t9v9J=rT;1Xo~10;2`;!?L%Uv3@1vSV>eP2D0-LwXq$W$a?<76kcEQ@hK@<B09T<
z&G1L_cpLiUbQY+btH3s#aasOLMfMV^-Lfp!njAOe@HZtR0H>lWT2)J(t;#S_N4mlc
zu*7?+mfk)vbz7Xj!byP0`RP=}rcK)r&}4X*E&)(0k6`fG-l!i-`4^dck*>I6KYN%+
ziJz<-qRa6eFH!Iq#z}SXi@CJF{-FJ9iuSt7zbejsrh5}}h-pW~gBC*Yew53iq$wOd
zUb0Gn_^E&B14B;v;_W{b5-k5~!@$hQ{NGbyUFmej9f~8)JfXRqfcNz3t#C9TGA_<y
zI-HD{hLH?kSb~211J&4tq)JTU1jTd^loDaDuC6jzyBptMGC;}q<ImCmm9HO?1J?`#
zCttryUI3fJAGjn{OAb2F5&(Gmc-*v@Ra$Z^5^<sj!wd7%c~>8c2ceAx!UlFh+k0J*
zEMQ@!jf3R={Ta(J7LQzV^o@`N$qIFO_p)Yyz)ST45ZjnwK8`_`$+9BBn#pHFxvn7f
zO_r@g_02!D2l)5T8%c>+-9shE@h0`9#-5h^TNO%YyRj_INE_kV!MRU4f!RrjGc{V=
zBl}iYaFiGOPEhbWg4Z_d9GoQa`Zu}}1($xwU#=xV;Nn%;WcSu*w)@s;724-QquLV~
z6zG&Ju1#-}%ptGqTFfP4Mkk*m#a*Ll{2pF6J?HmiV{f!@1Cm*&@+I5yyyp5c4!h=g
zD^#u~v>BQ@+vYeXc=6Iun{Bntu2c{HVcJ9PG9=ffk=cEzzRFa*%(K2d=Tg2c0@m5*
zohn|2m$v?#l3>@RchYEbkMPi4Ku|oHH>c6v&2eLydQ+cBwB8gPt|Wc(XjxZtV&{l$
zoewoFEA6^qcaA-Un@^{3_%8R(J=l?}C1*u-E`@LxwXilfY%o_J-v5iG#=%he3=n?<
z=`RFh7(|;Cr$vbKFQ*UB53e*$ZU~U}TIlg8;eKK>`6n)!NY|zx8oLFgQ+7g13j4R7
z<7awAW&^?Ew;-l23NCNDWKxS_v1ood_%RqS1%_0;!8eq(yG0*6mLN2F4jvf@?J2BT
z200vD@dZdSad}6z$hEBZh|XlC-8?S-68az`SgVqbsJ7>o`A!$rsc7_TXRBcDkb#mn
ziKLt)fzZdbY;o`2I9R^toWk<4gDgvhST&fKo}ER-!AAB)g5C&?W*Pit<eL5gzQ+|=
z-M(wX8KC#BOKLM{ZBXZ9kIkl@m+?VjzK>_=o%=A5xRQu+H*D=V0ox07jFfmyCR5K^
z&N=nH;hEG$@uZ9)FDp|KOazXfOlc}BY`Aj^PS;+t1U05fo{GXWD2M^gloo2TUXI1y
zT8LEIT+cSx8mmczA3}B6G8SdRSYNjrM^?Lbj?xu}pM_5HPum*Rb?Cv4ri`oWa?Q&|
z$K^{JOc3^xRw3c<b+8*HN=a$N(Fr`z8%m{Je1}TKA~*k(kWcFYJaGZ!TlidHKItCR
z+b<a9Gu*2_*n1Nv&e1&IxlhADcFNzZ2(`v8(>WJO>KlkLqaL}uI)v=Epo8;TJ01hf
z`xSr$opd?Qk*#TZOLXX@ScO&|SlDB29J^IzYQym>*{S168EUaNtg(ebi&LVP?L4#I
zv6}g-G^(_`18z#Bf1?mZgI;JF+%><eMtnzWte6E~uos)hbwNBCD^6{!!4h@YHo_Jv
zj@dFsi@HxZqBcjvEu9LPyu0Xw7^qv2{PhL;<9&LW_x6{fLm~<7;LwQm_R7EN3t9HY
z^7;Rq2L@dq(5ZgYEbZP8hd)i)qz4gGNLOamUIKK~QN>l>GiNlhmPz?&kC|2x5jU`w
zxmVG<7jOj4igqh|>2;zP-q1J2!!>8-O-yeC@{e%?@v8iy{#t@8J*9ZmItM*Yd=RVZ
zpI&ChwKp}QN1Hg+rj4@trD0uAf4E_8I-FVcp<`^*;vglFj<JY|;CAq+jRGc%AZGQ`
zV^~BOwa=$ry2G^e6Yh1b<!{6^Y}US_oT+OP-V?#)c97*FWNr9nvl0_Kl-FMGWcAig
z<D`5=`jXvvLVLxIm5Vnj)i<Wln!zQN|HLHS*rLuWs~BFcQ@^?#EqAxT1gYrO@Vz0_
z4iP#+#+fH{3@yZ<>*+nx^~COu{c6CMWTT?o7}<P;GxR}r4Fq)Td@qSpcwcR{{~H#4
zc@~b8(JBX`O$Hg$vM&H448_bl7ue*x{DT{3PXv|~&bIwk<$bAxj{?q27uxG(CfV^a
zickOr2Q#`3T#+gKML!4{YSt539}gL?bnwMHOlTEqfhm47@S|PO@CaMD?tW3XGO$2b
zR=#`k+mV<evcHp`|1p~w=9y?v<lVf;F(4he?mia#YAmZPzE~baf^aoo2@lKr8ShOe
zkxONZ^*h3R4z>DuLupLl^e#ttG))&`OMF0XQI8BQ`juU`_H-rXo$d$)6LMR#Udc)%
zMz}X#>P1eEXPTu74?#{a$+c#+;qZk+Q28D3c*_?UPEM~@^EM%3;O|?Jhv?ns9PyE>
z_<&PjX<`yQ^9gbUmYSnqEV<r8-eRki)c#6wVo8%6gjJn&Z9BtXw5Bpw7eAq2w3dpM
zywE4EtwSuASu!xJsE~x2*~vTFQBm^)?dUa_Uvb~K_i*^0&^E0<KoUh@h^a-=to6l#
zuDMPN9OwdG!=OPa=pESx%Q~l_e<@d^L3S@zE0%O7sr5oD6q}+WRLg;`{01c~m4wf~
zG1k_>X$NWTJ%$gY#$^L}_b;#mqE?QNJP`m>b>&gS`Ja<!QzU_GQA3{i7T<A?oi-J6
z59AgHg|Hfb`wDq!u)@-OI97vO+_n%V``K!%;8B2vOQvqT>iN1OWZ%S2{!*G}rN`{X
zu%Dko^EX==M3$Yiq?T>oHOBipO3y=JrJPBXz~_AB95qJExqN(2n6YfW1j}UaehCPc
zHm&A6i3U+&tgT6v5$*E!<9~zGTp|_yr@G+>pvXx0U*z-uQbaIuaQye`h7oNU`$Klb
zpM%D;vjWCpqHqRQMnvy5*vD1EcDrA;3wn<Y;n$S{{ln#wlWnFFuD_gqM`6#t%sAky
zzTUd-4VQ<1Z4NIf9T;Y)8UDF_HUsJ!1fX!jTeD#Tm65^O?}i!VabaZD@#&3srnq3e
z+X3qT+LCEz0keXh(GS?!AX<i$jl2n_$L@|ewu=cKeMA404bWTHYwh{?9?Een<zE{A
z%KtsxUG3k5>#NPsp`eEg^zpsoPzFlTKmoIW?S14Ar-$czSN?43@qd1egx?s1+a0kb
z9&;v+w5nGT!E!Q11h2+P`O_qY5^!pOVaB`Qo!t)<0>8PmI4Bc473n=8^Q9^$q==*p
z&Hr5O2K}9Zs?;N;o+xSV6}5cAvcM@zy|kdgL!!#0UlgP2lI;n9{4qg=v+n-k!tb>p
zipkD>f$RGft$?nUrn+LLgbz*(JtRsYkmrXCFJiS?Ff|W#rhmZI1;=b_*wh!@0Qz|Z
zRFNd|nwxMoKEC|T%Up?t-A&%4et_HL*5)Bq@mn|psR9$lA%C>DQGgt;b&5y2qQZpb
z{vB`$fg;RA;!qSCtqXA@sSOkxsMRtNqTJF>p*&$hp+3Hu_^7ydFPpbP0UApHti__-
zoG$`Ss8SqV&WBtS+KUPSOKJu=mBON(2eTP9ZQe;Ks-{DcLG@Xd5X)~$l4d{kT(@RV
z??@?DavwAg`He_~%+1_Ux+hyFnVy>2HgRUdwCh5~43YjN?1624@ykM$!oOkl85tCM
z=#+ycv43;OS<E6QkPCyp*m=u)&79Yzg{$<61)tkoXou{OiYn8w!Fcq7v@YR}JCX&f
zUczND*XNu&JEJX2gxL=o@xgSCSuQk#%}K6IC(?#-TRO+2OT-1!1}YWVLdc@d{PR6Q
zb|@2+_Fa{GIq^cz5sQrx^RQ17-Cia~o<&woxuix3xj`Q_7YUrj#9<a|Gj04Bd0~uc
zxGG(V$_J}>o4B-TOA-D`lri1bv;)aav$Zny$9&htKKkx`yWb)LBPH$#%0>$3O7Hff
z>k45&WF}RV8}>I9y-Vn_07S#*JkGh3&L|~?8FdJcv5kgqj8Z}XhI0u$1(10p#B{YI
z9>`!dxD^~^UOIyMYAv_nJa0c%D~UsK9oVe}$Sv!KuNIB6f5z9$PPJ>fv&E!}&`fz5
zSg&cpHNl46U@?0cO$UW^q`AnZL_{^w1&n>m!2C^3Q`u7{4@s$%?IOtQg=YarJH5rv
z$W8@0SzMyHB+|KHQl&d3%knj|t_VDQfx4U#o;u8)<@?j*#=2rHY^klfCcz*jt*07l
zQlhI!wLPa)`iExa_t)bAi2R1Nz?iT@t)D*OyKz@=G=oMWd5b_jI|>zvSaqcht+S6Y
zzH4q`XmZRZ^tp}#m|(I=lEt?_1xOGHXs>bhUl`p&8C&pK*2=4+g$|kb1@|-R0Y+f!
z%sOEpSCqPJ9~~7W$GhQ@dU-Y&mcGESM<hm85;O<^ABbpqnnC&X#yw3iy*38X5%}bj
zo93M^c#R>*8m|U`v~I64_YYS36;kikhDEwnc5YjCZ`0b2A2t}o(B_|ut>-osq5{U+
zKPzLvL^NZ`MEUi<j@L$tl4*y!VTgE-cTZE%mJgCifvS4{Xku0`R38)~O=y(m+@u<O
zC+GrB=f6Tk9=C|VVuy0~O6zzNwf2jDspPKj=JH6vHV<!$4E@&Wt<t8ByYwcNqjR;)
ztBG$T4r5N*$yl_%(_J$v<^#0{_r1)CdbF{{X3PEC(k5<Xzx8xC;){zmlY`{?VNLr1
z3<*(onZ8J>+N-a&Z6As{VXtFPm3acF7}>KVfR2sb#e%tMk?-9($sdcZvxd9kTd2;!
zW3CK()8UM4A(?sWVcM3H&&*>!+;R%=jkt*x4hD@~hP??&%?qXPEMVWt4U<n4W3ebv
zyCOk#qwL}yTf`bUsgKr``aAn~azq!<_U}h=nU*^pqdPDSM<Y3t?_3LN$M2jFuhaQ<
z=Lq}vHQxiNP2g9%A$ddNP9v;PeuQYLEqgSaTxWoi_k!>{o#8_EE<t*T*##v=H@-xq
zobIgkxyAm*)%-r^61|EuNhR;^R3s0k#srVB)n_QDXDIH_OQe3f;el#?j(o;z2`f40
zR~OPr1x+!|Db8gtTm59Zadip~g3QE=L2Pqnx7|<1=8uHDv&QhM+F25A@n==7+@-bj
zS>2pvjf>=WW7<7yUxB6jbhR?<6{GsQLWj+DxfO=21IgUx1IgL<I}8^tgd~nW7tfIu
zWd_DJF^BS_lYi!uL{*78)Dsi^V@M`mvR*4m07DLfaY5-_A6YL?ltI8S{rSy5@EidU
zwEwVsvj0SR{NKCh|DP}F$05b|{}`iK{ww_Bf0O+&q9YfN+m6)x6B3Z2%UQ#qj>HJ|
zO5y}U8ejxS5S<rF#0W$(S<|cqKHWV<v*4uh^s?2mv-Mdws@^bd)O<QuOT^b}X<_ct
zOdL0$PMkv*6P-WPRTb~E8euiCV%X-wiR(*)_nFv`kSJUQp!n><EX?CZ!a#esbtF7<
zb)hyT4rik~%?eG!gqFgc37eSq>v!Cx1kDc-xI7FkgQ@I!pHTgc7%Fl`^f4M!xLf4{
zKhx&C1)-ziHI`q9Mx68053!w1l!?PFjVBfdb@rf1q7jUadi9BJohr~xkXbP2in39P
z!X1HsFUnc-R$AGZi~g%xEumDtVUG&0HXVq;&HDk4Ml!h(6y112iMjko<$d0G)u|V8
zxa)*+A^PpE7K7HOFLp8egmPG#BEM@W?+(|KQ`@;Wcwv&xspl+<#_4mAeI`)?FqSHk
zo8avi=ohREX@Y9sd9|ZGs4<_)K~3MuZm+9rL~beNIIO&|cp#rcM}yy~v*UHw>15AH
zR+Lst$)s_uDLb87J~WhdifdVd#LWt4GO<sn;F32ru(`Fd->?;tW^c62qzxmTP*jTN
zb<I=C#HE65=Qnj$gPTFDVb9l}yUL<G&pN+{8_M*_>37la<mQ8Cd4K2Y6)=t9fN7nj
zj?h|SaEl6WzEglQP0E+;4CWy;Z2aJhfTLmt;%`M__UMvvnJ70ch}tZ|?rEyBw5zK8
z`!`>r93Q1P;qO(7d_S+(dvKqO{mWh4@9%j3GG4U5)Rn;(1pWWMkCQ2~dvSXM;J0nU
zv2VQ7hA}vHcenUI_E#>diQj~{e7?@Fug{;mZtw0C1h##jF0WU%)d~*<czZkhhH(|l
z=-X$s@`lc+wmtYC!2*fKAT=TI&mO}b0u}5T2Li4L2=>0#BUCsjVY(fB3b_7S;a8}J
ztXd94KCT@O;}XXG5+poZzTcln9$stbsL$8i{xK4L9WM<(ROesITYG))64W(p`5Oqq
zuDNThx!WWA#4-E)(DYBnHrF)l<1ADEJ+@cd|FE#%ek{StX$TCrgA8|HbbTS)Ph8vk
z%@-HhT#BcMkNG_r4cT4U4c^wT)_<&HREE_pidXhl5?1C{GMf8%87$=eJ)RHyukQi%
zyL$i1aryf@zn;(CqPF$%zkOb(_M<Ued|pq^pSEG`l%NwJB`X15zjY~cBGf!RUuz6+
zEyC;V@^*3iX#E4UhPb_PYGDo0c+MWjC=_;-0gw}Wx@d~VL~MJ##@fI*5e#Y*^r7RG
zz~4wv9KjoOK9EEqG^p0}Xl>Iw){uLeifT9*%sqn?yW6@Nh1NVX@R&%}D?l<(0<Ht(
zAv&dgOk9S;D~*>>!#fj}S1(h73pAj=1I2%U3Sjku3&ah->~yjXe~&Vb7A|5oZn+}{
zxh@CpWvFNP2u@oY)3^`W>q-vBB7+Ak!ej+zFnT}|b*;kLchU?PL@MDmW~JJY{GAsS
z*|Q=+wS><O5%tax{e_7iLI@SO$3S8Yy2vaNbqh<?LPRaBM~DfrRJibrI8-z{js}3|
zdsD&7nct>1_$z83V>k_EY)D84gu)XL%vuG%1-gLJoydRCpkT=BQ>=Fm+%h1@HA8}`
zJ3k;liXc3`Kr|<esRnZBAsFF6l3`B?9Z?`e&NY>Aw{d`*KPb%s@~VTU4y5q>deQr*
zhy71YJmSf2RJB0#8`QjMt*_)CJTD|i1=M{G@hzkmq&4+|D;m>hH0JnxyoYcV6Y#^n
zAV-72bq&U6v>0D-69bYD?2WbYKJjO^ruszRSX`SUO_DKeuGR5I@i;c8nnaUW+`A7f
zAfhlc2oN`&<SC1@;6qu6`A6pW1aS%=A|5Y7M@!r;kykC{fdp_~+68r%L89O>J>q_N
zQ-_m~T5t{J!*vCSDaCOpZajJXJtB1z(AN#2+u$(F9)$CV#&Wm;e~{%SuSw&)g*PXb
z5-ei{Q*};kW+zJ%2`pCXrZy~Jw24A33CxZySR*kP!AG=L%0+5Z4j!06r9_%WN!tPn
z$x<W)dJa&R@t7<oS`aGqpQHGc!q})#v(_SQ^ScP4A{<qqhafgMiK;jstVqDRBM+)*
ziGp+i8gwfHq>LY}TZbtSXUnvM(7d_YAB(LGLNQ5iqYMW<({%?m0-m4{>%XEoY`|F{
z$G_Lfo<TD?jZsJeCB71nBiHKk!fwp64YAm>>6=J#h<|S>Dg#>dUdOedFwOI^SUgyo
zq{yJmyr9(}>IGXcP5G#dRW=l$5?g>1L=oBO_V}PkaQcuyBUMf1lz4&*0C5mRA+%&R
zWQ1jsk_nuyqKPtmgs8tz{=nJ(6&0LzdxU@GLj|g^)=ox{1l5g`AsYRWP6lMVy_OM<
zUPB>Y7MoLnRWi*3uj3U=xd3nqv~IWAnO%tHECr+RLK~Qp&x*aBUS@Ey!)&CIa9(2)
zDkt;25KV0jtz;lVlmZsSyXL|2s<m1}xd`nW@n(=at*(e4IGV0wH6d(&r>HeOa)8wZ
z>gpJw+0JcS)?y<@4G=Z;WRvChCAPZ!adRT%flET#Nm#n35G_m4g4UZA3RJycBoCvy
zOe`RkxHlzC=dpu)ReGA%mA8bRRW<`{DlDi28^#QvO~84=dxNuDb$Qf>3RN_jBKMkA
zKyH{a2cBO?s%Ca%yRyh23{*{@M&P24uq5Yz&<38`(^FemB){&&o4H>Li?_B^<PIg9
zGG~nb3W2Ii^t_(nLKJKO;4lZA50fY2#s(DM)<vBu59%1j!_^cWsG5-+2bK~#kO>Ev
z(=aa4?oiR>OkVdpao)<;tLZr0YWv6MX@xCH&@p_DMhWLBe$^Buj7j(FD+yA+$%_#w
zEY6~ZYDRwipD$<c{i=Gfda#Cwx`@WGdU1VNy?D*&I#cJ1?EY*_r$Swu&>{pp<l~_L
z`Z^EgVVruN*;j%XDw3(-L3@WaE2g`Bq@Q5Ep2`RPo~h^W@Em+j<vm5TjYSAPVFw@q
zkatK~a33d$p%~8Wx5>;}^J6oT_zOuMF+Kovg!O3y-xMu8)+L|O^&XgXHkPfE85lLE
zdH{2<<fWFS6`PC|4Z8Mf>PM**Vd!uf7u3hN%EnnYqnNrVUgpHto++vB_C1h5=ApV0
z1~%b@9+A*9h{lFcMC1}ERxOb?7`|J!B%HK<?#Ch2pxnU7EA1cB%ligI>_zONra&=f
zSll1d!Tdz5h5k0V@}PJ`yyp|i#K8P_fC)WTi4ZW*E2Jr5lMB>Mju}PCt8IGDvuXO`
z0~o!+d?SwM^ObgSEW9;H!L$w(x8PC}dWQu61PUQCWxOZ|bsg3QCPcSbJoba)M3jQd
z?BoXmGgQTWiRHjK`^_m~*fHBL#x|LMtb)mAPrrJ&LLko;H!sH(I3V$PT|B#oi+E1p
z)G=?CSU@MdpG*VtseK=_F>es7ehCygc5)Uu?*AfEb@CTYcK3`U4acxK7|HTzB-gKr
z{QnNw?z`L*-QC@INu?}GK02r(*FGi34qp~{tfxehRQRwj3l_y;Di^^$NtNXx@?Zy?
z=88C(UwbB<2zq!9P9gZER!<-o%OT_=Y3(qSBf-@?ynw$6hQa-)<{-V$Cyo<mLpOR)
za(kOm2g=ek{^2Jaq$7m+yARCwU?+qx2M4C}*)y#+)HZdjo2Eo)0CASA_7EfjSAwMj
zl|JzD9f1f4E+32`5i`g*==<<N15fHg6Fm|UK`DW3hUr~Q+iDg(_{oZ|R~@aemPx<S
zW;*jGA;NuCPijFC$=cl)AQGfb4qRF9w1El3aiNE7iu)F4zVytk`~^}a8{Q;24CILE
zX^X1B@vO)W@8q>nXbMFk7$cRBx=kVLqzeH8ACEC9o=;}01KxqA=RZk|?4L3{@;&Fh
zd44bDo>Nz7=+Hv<j?Y_dZu<7`c69Q{S~E*kco9~62owl8TPIs+{6);oejOP{%Uu*5
zcSg+J6%j{YR2+TwGrX9ssOKr}yN-G9VmTk<DQ0+)3QN2{_BCa$MaEjcM>rw5`+BG(
zhxTE)R$v1znCw%_u^eDy>|M2O=>cy#Z8X;xkZaL~?1`<iO$kG3x@Uwhi`w1sol`My
zXv_LHHDK(8toi}r1KJ3@2MYGfF;y}PxA4I~e?r^rKXfV_oc}L^!~f-cVrTpBH6t^+
zGWMkHNPRc;8He<63$#tI!biZYQDy_}>)y^j5w@gIxR8?ckq;Ltzxlhy{TjU4qMS2Q
zmxqa?4SJ(5Q`7r(kB+O|+v9Ke{TDnOhj1Lu@9*X?3K&SgV3Y8*ogIOOtUG)@kMFmO
zQzjQr#R*cBZQeQGx0e*0edqVAz~-R}?1AfkAq3{lH?ZQL9WLFRAHN3#P?CVp{9F^}
z)$RI!zUNL)r{){@`Mkutr_vE@IAUD1HXzdreBWPp-y9!Pyx(6$qtbUw^_La)>!QNn
zanIlQcl&%?|LV?ulGkL!xJXH;m@+3h{8Pl*B1KA`n8ES6!Q%V-9mjwe$S&yY9kJaO
z)z@-rv}|GYC~@wslO%w&t0$ps|GcxHVra!M=TG^d<tf?HpOAcsR_PkslNGsuI#A2r
zf(->w%N4Q+!&EVZ0%4GZFl`fXF9G1!Zv}#bbI~FOb+GbYE6EsSA#Gz(m;~sD7XsLc
zcWQE|+bX@|-OxWZNt(tnq<<4W6IC7jiB7>Y=2e6W?;Fa*s5J|7tSX{dtI|xbm?q#D
zj}%rTDcR7TN9uxMltlEK5^Uekib^=)H@ESnXkcKKFu^54#WsnNLBuyrhVF?ix9gbE
z(-78^5*nrjHRZw1A_WPO(Tcr5wWt)iC(o&}`z>aN%PlC5*|jp&hcJC7CtJ`Azo=*q
zt5FBJ+HK5oj-pM-N2TJjF@L^kEL>O!9vd)cgr`%HaKvh5AQyv)4Czc7^5~u)hd*AE
zPiT8oB~egQ4N}2Vg$@;#;w(k>NezyB*>a0gcbUgqh7o@Xi)$af*@%#<&>G;6GncNa
zc_eK`7<Dz$A{3(#q{OR^d#N8wu!zQm>?6Y7t5ZYl+LjODn^si7QvNeyiu8&lK>NB)
z+@jsbT!z$wemRmruelf+ysm6+6GU_zRy;onv_k8|D(EU245Zd@teOu+4^Gr$f;h?j
zn>T9w-Vu*IMS*$+aO)?yAB`JLk#o=h!cwI0CxbZ5OYBd?!bW10;@5r+7dF#8RBT~6
zww6~SK{+?2Amow7ehdotgbtE%_^=QabCt$>p+<+tacHz(#AMciY97P}%5ed63u!*Q
zX@tt%D3u@PIB3OWP?#`Uor4}as%&BDtGR*P_#7A3OCX3^+RpA2v0r*f6nJdi_J?1C
zg%6t12bY}!(+653eg&n6EM2trn1!;BH4hMZN&%ZH98IJ>bzu(^B%4U|9#2-HH6rwT
znrR7+atLMt=Y%F#0~zK81h_VG3$oa(7+cMDsh|u~yO#@9ycML$m4697_dmOb`Q&pU
zBB0i8%y)nYXBj@kI2d<UpjQF`p~`PcKg<wgVdcuSqv&71Nb_GH%?2rV1<^;XaLT{<
z9#*OjdjEbF-OCF)SY#O_waZRCUAM?a24>u(#sOryb5WBhmkocf$l+kK!(={Ry4x(M
z=)q~O#cA>?7}Xk7TarBnB!$ZLZb%w2i@{~44#{rK=P*@HcW2fBf2or2ooK09JDxax
zbF)dz&}qSF1OVJMI%KJp&d3pb+UWUcqBF;wacZ~|IHe!X;B!=(HmB7NztD>0N6hX2
zUa*ciSAt%wS5oD4tqNl)kU2d@qh_=$r28k0(&qa;tdtERil`17GK}!7HTI14l`4p<
z$eHTSNv#fru8yb<s)AaK^A`M(m~SU$%bpjv!bjOfQGmYtbV<YW%|rgCboC1=jI&YC
z%(JYE#$w4w=?R)s$(?zkGCH1Oe9xIy(BEcQlYd5Y?`#QuY}p@YnsC9+kPv7*bJW?T
zIq6+JLVsusxnWk<0k|j7sH;~T-rQY|dL_>ixnu^qHBtMbpk!B(Y%RiCptN&cr9BmZ
zb;eaCMHsvUAt@^FO|;*$*;F;*-vy|yC%bWwemJ59#RxSOzQz#eDc=|gBq!+`&f!;k
zu@lpR5tZufE_eAVW6{kYf_R%I8k;c5EqS^^2!)0ejJSev1Qbk*(m10-Nz3!P-4602
zuF7vnL4pl*Rf$7hWDQRT555G3w6M;3bNP*@+KlZ|s5VICbQ7QAQ?>%sr}nS(y`$7>
zA^Qh>whnOa70_*;K<M6_1a^8!qIK{6>Vnz-hOaJMU*TMfsC-pyI}v>7RT1sR<-|YH
z_y|zq{WWw<*Gu4G2gc~e<ggOw5uyXPT$$Sn8#ew>>mmxB{Y0&{5745CvB)DmC^|B^
zP@x{vTLf|1c0O)HR%moMju5Wmf;;M;uEVHg4I?wzC{IxZT1pM!t~QT+OCjBM(o_H}
zIm=fr^~}LDIIf=NAo99fUCGd^9~FgED74+0HCLOPZJ?d>-7~QKV}KLu6+AYR7j-r=
zBw>?^6d$O$2nAKIYB_92v^xRrLs8eIZ2_hoiB>nTLFQ!p40mF9t|B;PWkom2Oe>9e
zo3N(R?9*HNP1|N^16nejT5Or*<pcA3rVXfKml-G10}w5_YCVzV!tu`}B|-nCqB_*>
zD5yKRaq@kJ1k#-6%<dfS3(c!=Nci)`wcT=S+2vAw0d~dj;cC>pP^s_Kl%#Wsh`oxp
zzb7>@6HTJkH^vm?+VP~R9&P35oqX(`6>{9W#7AgJq<`-{R+&GEN~+obUV&`jJJI0x
z9W2MbhMp29Vi2>f#z$Q(lKR@M+&sO4^w~t97a4Af{#8<XF8)B(wI{Vmj=r0qlcWq<
zqnQ-gO=hY62LBUvn%9onO2Afm*Gg6w7~oL@dITQdEAZldx-PTx5wWczJ_Ki5@;ok=
zMJ}~?xfe5T)ji)*wIcAGa;?7WIIji+IQxiAdK4&q`f<&zC|Rk6xx85J74WHzt)^1s
zncJE|S=*i2qw1ALc|?&7y(m9qa&*|kUgO%Vi!%Op&LZ8cK)<Wdhw}*km;QF^w2nmK
zwYNuGYA$9{%HPk<v!_DeR1G*L_rUrN3Di6Auhr@nD7Cv`x#g7Ql#iAW?S}sI;iN49
zJU+;3IgEFseeSsaoFTp05U)9QThB-MchzIXZm}hpn3U$31%H_e3k>Zse@s4~DaXr{
zQq)NilJah2FzDTCk?Zy!xa96D!>*`;54^4RcoD2%!Hr=T>UQQHRrNA1gP|G@@t2JR
zDl+R>^f)wgD<1>vOy%2^r0x_goCkY!zp(t(<NgkUs)dWJBK}*f1$%4z{e?RSH9e#H
z-~M*9{OZ09tHQJGfOBZot$3ZwT_X&O#pGJ`301jQ?Y$k$`<dTYpE2Pn7E)o2I;n-(
zW*MJ?BxAL{zb71bN~@M}oLt1><eM>crBmpA?$9(oy*Jz$c15=(O+g!d3G~JBS_CN9
z@LqmBx%LFuIMTJs#?~nCE*SVvNqG?s>ZuAAwrqEoBPJ^X;(CuwYa4{hzdIl5eW$GK
zv$=S;59ufxb}k<3xzNCJ^Phdt2cJx8J>-L_x9wuwpUOe2@?YW!a*Ov0w3Esg-bl>o
z`--u)E)qq9)|24sd?(gXG<T;)ZTTjB#tkF>KxKDBJ!}+iaBr&qZLf~G39KT_0k}OJ
zeuJ~b3TyLh%NE}t^sWRbTV5h!>Wx1L8*!3hiIC3M-Y<QRP+5q|{^p#}W02P);&B49
zz-1x3?kNgh^@KL9YjgDVb2vqkqPcQ8?dbD;yx3#IR5mryP69-Ixg1`e28u#oK;cs0
z=c<l<B-N2&WeE8+$1V#b{ndTZU|Hu@gd0lGIH|srRIMhRQQ}}Uk5&qK#wXtDb1b}c
zL1DAL8-&GxDI|M)0!5p?`PWPX16A@Y-5L@u_Un$@!7>uE2$BZwEX|bA3mgR0isJ8I
zi5Uml@B|WWw|-$PuLwGV{tv;#Ir)x_$8ekh3<Gdjn!T%C!qP=>?z>5os%gpVlF6g{
za*Qk)zpF001aBTX$a@ea%M@738U}pK&Uyf%^YFa#Rlm=VUnPKDrQ?3vyXAsfgX#d<
z!Lu3s`K{<S5{S+qI7Nh~uyAwI&oEl7CrS&3;?2mPfxJ)hz{w|E#G5It33aV5Db@st
z&5TNZ5UT<fJlt#V5;x(XLq(Cti*kmlV&bI7wY;Eg5m>QZ@z*mlr3S<kwBnezz4l)e
zMGY8)R(uZ+)Uvg9p{_DY1A)?IrmukvAzTUWoh^4JygkVaA#ALh{EyJ?z{22ojYS;l
z)S$!kF=T1+w>Icwmyxqk?#YH8d;Km&3uNj0j@1xV^^|l{2x>vB^sq4Fv-RR1f)1@b
z>pbhC&l&?WGKJ#Xf>T=Q+ZEwI($EJlkYRA=v(m~Upa*SxQvc#IM1Bvfvb1aFYGLR*
zpOI6p$9C!@`|3tnV>>3jhstxwXk!dU^#N@QAp5c9h-BO~5_wn2#4a-gYvgk7a_L2=
z_;S|B+9&p^k}a2=yOS!@&8Y-{;T|^(@=m$ev1Vc>_#Z0y35zbegZNCZ*|`nvmnzl=
z8T<ZBTLSc$H4wDCBsdB#3Ty&<;G85loeTLQA|0mw+S(d71CxgVB{Bb^59;`TYo>-R
zj4W|a^QByf{nu`|tQ$~2Nf0V>rEPKaN??DL+da6tMYav{|G11adH=H&prC#*%*Hw{
zzkeoow@gtBNw{%;o|LuofzIR2)`3#tUyHug8H!Hdl$x)-eYRC6896*7Jcym0HRo7o
z5k>mca(tMw^py1kr}6E#sYor$vf)3eCeJpgw3$v-$#g0V?nPJs8hc-$OrT(O^NWnn
z1)52DnCRd7XOC#?9I$6^UYq$IW;UYX@fnl|!?s*~rMROY_U<TRP5|(7G}2pp(d=%1
z{A|hD@1jL?J0{pCOfjp|>oE7r#zD~8vtj@bupjKcy08KY4EX21$2M4WHv#)qth>>q
zAqV*m+3=Jv_Koqp{}lKio6`RfiO%$+Wc{Dd|MrGV${x29t?xo(d117`>RP6VmAF*`
z;-oA_zR@?s(U2?v&lDMs`3D@>fj&`JXWUL2ALv%O?&AZyE6<*bPr=>W-R0Zi<xFv8
z8lg6Q3)p}7a|=H5FUDB0dV*2B#c*kO4e-TQs(a`h9bcjf<pcNK=27(}&lrsb$AZ~s
z_Pd`Y$p~!ez$Yv|*8idVOP-mkGJL>G_Vx7paQK%7D^hy?u7K{BliD~kp-hLAju5Na
zQhB|`jld$n{`EdQ68UiYahLJwf)tJfd3pgb^?11QMcu57GI(49PoWw$T|<FOstis}
z7Lh`zzhVOE<KmO<t_7qo8g&TjmKw*BD&34=wRP#fN}ZShC3_m)6tGDXS>EK>w9L}g
zaqBl0!ZtN8Od%|naMJYHLZ<os8-T@Qbr91~m6}I34P7H0$w)1EgXZlU!#Agr#wpo3
zOrMk62nx#s+l~xYh2i)R@G+Bnl1w7+OoxLgi-wJuB{@R!#~w?VmRJelASqC&N)F6Z
zQ=$kBsYuKzzY((;${5Q?wvBPO*2FPJap~@UiJJkTivEqO$6+$k+)m4Qs>eY*^57j;
zpR-)_FT&IumMC2?W)YaZ5WiyGsuhZ&pOp?}mLRRp{V!dq1Hv(E(roR<e@7UG0AcY%
z_`}qvHU8l4-+SISBp9|iOr<>n9a1vu)>zq!W^1INvIpMVs<$Do_SsZ9MZwJ;>MW^>
zn&h)gDpZkw>~c-Y1jWCM56!hnu-KNc&F5up5`<XrPD?WYNq^X<U2@Wm%h<*XGFJ&1
zRk+N{*p>@2)d_b8WILC>dHm9+mB;4D%BUJH{@P`i0TnaGS+L0=Tohj8E}`NYt6OA}
zW0t(*PSrTZXq&^cSZ4k#WQNeJ4CgGRj}yi<E6p^GeRS$dyTNFaAke$xXd+Hm8}td0
zMxOF4RnNew+Iyte!akCmnmc(9-Z(9|L`0=Qjh$-Qk-hYQY$c<nQvO9PRI&MK3%`bf
z-uIvw{*@q#a}Xl)=ZHH<pk5H|bgeM~R>X^(gBRslS$V$2OIom`T2o*;Qc4f_&3*LV
zqLkDq3DahlO3G4iq$o?P(l}KuM9p{`mB3B?FKz3L@vMKk7LV`@Z6ojNxdHk7ryWd)
z0l0hZfG#FE9L(N8gE`!H#8B_y?xVe|&qMv#od5SD3hv6|D}g|&e3HY+-ev!&Mu&-2
zZHVSUb&zN)5()`|HwA@RMB;oYhz10|oCdNyP?%Zu0)yYOk=%qQI4!2o3V^~KY(<fH
zkU<lIrHkRoklcC{k&{~Wm7`xe=&3$nPVsxi|NMm)P0>HUj$c9JMXF5N`LJ9Q8^aCz
zUz#aQud#T8cAY2GIaQ}3GwcY~<-f><tX*cPU3<1Fy$L!dRDMh9b(lHNk^`P!MzuY{
z_>*&fpCP#1`b^9cW<|Jrqzk+)%9=@bDL)z4<@M0lNQyUA7P^pBh;0S{E69H_9*8v_
zGt^6uygWK&GHvKvk`3iCmFM8O+F+D7<xwT$n|9{cewu_(ee#*Wfu>F9V1Zv8>Zn3F
z5t5mZDPrR+(vnWz`>_3ISNl<LGDMWIw|n!p-{c>Mgq&RAETyo&onD7L`~_Z`?<KD$
zjjcK@PPjd9x3;4?bPIz3<czBwB{cRD9m^39v8tV|<eDc8HJQY!$~Z7G4XfU1hTp>A
z3SErv>FyMu%s%)a%r`G5p2|NhXUw#-Xx&GCd5XA@44Vd@k(Ekb^jA@qRAN3GktfUt
zE)_*HpI!~P{yj>w82wx1#z}Gh>*=z*%z}?b=tE$S2zOwY1jg_zOil`=yKB<e6j!(`
zGqc$dedfxH@X}inuUTh%UL<>(CD?4LDs6}}$D;OF8KiA0_Fx=fO6P*SWi_!`KfZO1
z?7la8^CAht$HHx``4cfQ!|-&{zOx3UA_py$M>&R)L_Eqia50l>z7VlW_MH6Fq*d%t
z#l}uInv4Uu07ixo9p@ss2m7eLTCKQ*9(9OK@|nu1u6jtm>PX+0D^Vr}H^*n&%u+Fe
z-VPkJaRez3waPtrwY(*dkcCt=GG*>PWsO(^Lqu3ixu(Uo@m5pcU!QI|QFj=}HNmCz
zQvw%m`!B>%{Do!=(ZF(0gKHWsg##zu_~Ck)2BE^H%bnY`zjrWN>9X#w3j*GYd0>mq
zMPV?(Ak&#Eu0q)ut<=Eb$701Oa~nW^*acFGgLS&yz!df#$~33ntH01FZCoGohJzLE
z{sQoE?=^<|_<O9y^XY<Hn#Nabb;uugDHrBQczdEYt}j(-Oio1%F?-ZDGLr(weo=z`
zHdGY^hPy}u_rtsr0?L}5{YW;U8|kLenpDJkB$g2Oz?N`rC5i-iuj{a)Le0ZR|3-`(
zQYHBcB(};&O9O2bDNYpjpWdJ8)l}51p?x##@c@<ncM(xN0hju&@=LhtthoGNiiL6v
zv!Ql{>tz@MV&C;9d7azML>E<pc^h4ALl2QT3%guOjbJPfjx_0&lx$oCv}BF5#m`-K
zd}-gm+nfXpT4pA47|PH*rx}$7>aJ0$fNFJMqGuRbj!h!@%&93l3%P%hf@xtaqE+Ln
zLi;p=D!y+|ebDrv4c)n5Dx^8^s&DCh+*I39fdc#W^JPG}`>kf*IJov{k>SE--gbJQ
zL-$60uymDI3MD+dIxBE#h{7x>b~Kjb4(dBEu{^c#feV!GSxFB%EhBc(IFji_mr#`<
zJS~Ja^$MbhHBO{Gh7N^7{Z4fJGpQw_BE_&<*8h~WKfReK-Lj;^T-2N^Y9MXFF$orh
z3>mk;<7Bf)^!abqmp{}*Fmp4YgoN$ct2||N)9V|`f|fgnfUxhWBQPpTL;|Pc$yv6L
zRq0N4BsRwwW+sve&!XQ}BbATgLSm}0Dg|#mzCK*F3`9!_%m0HBSGV(dE3Td$=Vhq!
zg8o^JE(&W9&=!DEK;hd%H&5W8Rg^PmN6(rSa(mV5ggCV;ZY~Teazw6V)e!MYb-7#9
zeH{a8v{ChjSbKdbo}c73Jb11=CB4n+LG6$EH`G1#r{d#(i>q@~d*hAgsMB}YM3uac
zjYdVc{xu&G1n1ten0UFK3Eu`s)UEFe<D%xwTC4Ux&LzkE_4C*27elzMWo%t1v2wkS
z)GE->zNlvn?Xw|_w#CfWz2k`GQ*fHvRbBZuQD#i{eIaiOs4xzlO4}V3e19)5Po-35
zcw3*VD7^cAPf7H>Vg&1Q(b+TWvy_ED9iiBXI&Usfs~yWL|0TL|<1Oyf+MT*Oq8fUe
zrD~a<dMYwYYoO%)brEUkB0h9Rje`uH>HKa9+Fs9V+xie~Y$Y{)9JM0Y-(l?dj5=a%
z<y&RkTKy#-wSw#zY3z#Qv6H%HM3(%GJw776D%jtXv`|i|C3B&YTBCR6uH2LsA(K+G
zIq+m-IxHR9rKDG<3je5*DxCx6%YaI)cM0Xrt6TLO*ORv@(604VrOj4uI8J-n8xQ={
zGHZT+P5G-1U$0&={o(*{qV*Rd*(KdZ2lVTJo^}W$JPp=GZ97W!A+Acqw-1|XakvZ}
zCN4)bGOt~x6;Um#e|@F)nXVNvv{OuJUCAL1-C6?9LZaI@?Nh@_Rre?1gfF(nit>0U
z5le=&K8wp+B}*q7KD$izD3ufvriP16((GODM_!2)zFd^G>D)v`|4R+>qplyC*14K0
zwNNgzRiAgzFv8?gc+<8p_^K+lZLM);_Sx(5uDyrvv)Qbs{7{(ZAu`HLLRZh>wc#4a
zSQgNbV<aQ&BwWjr^!=d5Kh5DwzZM1YKB4WXyAfk8CNrkeTBLVX+EO&)6shjADDaZ*
zbpg+D^2=`P^CIe8SB6=QfHcG87j@>ehR843t)}M;A@xQpeTE&msfMbu^%SlqL-z}Q
z-kGjSKU@c%`XqYo5w13CX>sao&js)D#IDBmS-vRr^0#}%ORsMD^pf?~edLvuOw}Ql
zEmFAEc2t@~<;}%mJ3FPF*-y7IzAGga&*o?h?~7>`0=A9N?Yh18&7g{%ubWc_XCwIA
zQ2a~P<A)TF^3S;+;rLEz+|1}^`IYh9)!GnM!gWH<;VO`n_iN!UTORc`_$1+Kv_l)<
zscQRQ$WTwsc*#>gRhO>u;+uiqqx8M3qo7pmh&L#pj$1>+^WQHgnw8!E(2H;~{jYIf
z%v?Xb!2eD!vVt>Zhue8qSARx;I3sZf4mG6RoEEs%Gfy{eWb|hz7|)a={L?S0-Z*ia
zZ^<N<Ju>i4wpnVmTl1GRy#Yq2Pmi}(r}xWm!#iTl<EQ(3Go@GIgMZQo1<lbK(wxO<
z={Hjb1SQJ|0t&CdRI!8cW4plpSBJeyz!sD_g}1fJ@h|wOaeFPy2Ky8#yYFvz_^1K0
zG!6ox$;*H{dj|i`WVBBAId)p5fYn7?qZ3}JkX^<LhdUJfn>!j?JBM{^s(q|}qz<u=
zebrIg*+zQMnSr2&1)gT;Y3cnGyMZu0lWipT;Foa-$=`yDlM(hLs|PA0g#Y+gPG_F5
z=E_-3T6fT-nSakwaxzF-w9HQ;fv5U8fI~=%X||V7lMH*1U%bqrp~zgwW5qu|>UM%g
zelv620hh>k_tihJgn~scY!-?XbqdR5Ad@=|aI0@rS*4vUJ48MvhS2b$*g>0bjX~qC
z;gZXoG1D>&sh8@pDe+!1-E#?)F8s3S5W)_x`kLn)FiY4!uYsl6u%T)pA!%!V4m0lw
zUZwh{XqtD*&c)x}DaHOQsu1_!Jb(6kS$Grikr;fB0@Ef<V19b7xSw?Rm$giCD>F#Z
z4=6`ShD_GZY*W6)Rd}d}%d0Vne@O22U~LQl7@pd`p^4k95N+{$`~0Hs1u6~^5>7=~
z=s1q=XE3<tpXqig&*37>ZFXxg3)O83=!Pszj4_{TT5qmHPz}(1_s^UghPuyiTZZ`3
z&T#*2s*&>ky98l_eWs5V_>&j!o<b_|*ed$;Q<d|uqBeGjk>jP0(nWpwAcMx?Bx9}4
z&5UqleqCpGV&Odjv@M|I5hVn4<r^8p=xXJhI$3Gy=;sKeeTCwn`zRO&SA&lG#=VFS
zB52XlQeRDD=z$kVyYT$F>HTcrl#M|<pV|$J=4{%QpA{U&aAiDM=MC;~p=t*GM&xem
zk6G);vDf0qH@(Ugf+uPN5#Sms`*~j}z^qYfeg8nrSIKGVMJKnR!L%C3!sF>nR=&+f
z!Bc~{)QcQwOO}7~(8u1K)_|c9x$3~PB5NkxWGrB<g3)NCZN@M}Q=?ZFy^O|4&P>fi
zb>l2As^NZNJQ&=#>NwB(hH>XX#(Iju%wq*jHM~z5A$69l8<2v^id#k85iT<%Sq(RA
ziY^6hutYn>+jvZ*gMeinptFl+Hw5a;)(5Wkq=nR%%tI@_TmZsB<E71|XGkO)<U&2D
zJJMx@C6T@g$K}69XloH}aZ*(ns>&Fi?s&x8W8dON!SFG83Ya$(vJCw@yh|q+5K);%
z<Z5C368`5p(}Qw|tT8zX`QeQz+x?g9z9+k?Is1^v$Z2F{`^D>tNbjC9RuEHb<FBoV
z@y7#7>$4Ny{K~f3NecX{VN~Ay7*Qh*sfg_@Do`>38FM{icXTOxsz3|r29sIDl(W?u
zBOr$=TArCK2{^mro4JU~XNi_54^b0F4&r|lxeI^8T{z<#e&sIcrQ=E=gq5%7y0W|-
zwKz>bJj9oxL4exZKjo-<ko73l%hLdvoEFR^kY6`EWubx|5?EaJbf;}nL5LO_hECZi
z<AzseJU+yEQ^h$8bMeiQ-(19(F^{}(SK6V@l2+XRX(Gv#Yy}8taWTf2Mn~V*^bi%O
zD`BUv@XDJv7c;Th=D2mFj%t!9E}d?-ZmngJw>pm4w2<>a@Z5BivkgKWAHwvRt1YV3
zhn!p7=gAqXv(HV4`)QT$81h=PbW#J7;s7wP{0_A?X65&OG;XpI`o2@d{oKY?eRT)}
zYEsAKT@|?zf}T7rAmut@*Ag9V_3f$AlTnzTS<#hk+a#spY>ukV@~OdnHN^>IPsCRx
zG#vX$EyCvmoqvqG5UjeLocj0!rKwY%XuT){ReUZf{HHibQ$O=G$6t(KwoQGCBZv&+
z%G5FcX%Pi44S8^gFeJ8Z9Zg+H7bwBjYc$Nr+X%KRGPf)~62ZU&32WVPU|6>tc^bp|
zH<hdzsmuhO%vYX)qm_+11)m;6bWR3O{q4P1@!TX<4-zqpb-geR1O8K*>bp+ZvMds{
z+~8!<?3`YE90?vV%i$*nE&OfUy-Q_7CFJn^6Yc9jkHMQ%QL{z|?O<g7IsheH5xJyP
z$V&OJ%q>ROe?HEl3>5`nVo@kl;i+38rFwFx4U!^D#AeSVgd8Et0lC3*M6b6twP+}O
z%YwC}n}^<&D+QupeJ$)#x2fi~oQ&K==_Ff=hM3V?eRKSAe?FRDL$7SlR=BaNmuf=i
zsMgdI=snae-4QCjS!gqWadgj$iSIM2W^=iljTGLSZpL!FqfT_(On9k^NtY_BbXzB-
z0_1xQ+3u=2BgvV(sC(zmA3a*LPqdp#r<k`^IY5P*t(}@RZ4tOqto8n0gRz+9M;~{C
z)N+x*=d8)SIN2DCA&Wlp2jxW-VL_a7?^{-=i!g>^cW8y`g$St4?mG<H6>Dm!gZ>Z3
z-YLeGKyBA;+qP}nw$0hLZQHhH+qP}nw(agc|C4i0cCyz_*3C$bq;9HGmArZD`<_Q|
z*Xb7YWKNYaZQJ=+bbs?Cb6p(S2G_EBsJ2PODBO@DeTOcx#GX`IP(3EbDddV{6c3dm
z5@Bky$&m%=)FnHj)8Q)VFL$*;sH3RlnZ0V(>fqq;#f!U0HB_a`w1(K$xCGs^g3Az1
zP|Cpz+`LQe(-vsf<9gVOTpI!;S?Sz5`vkE7WC%^V!EG=@?Kox;N;;E_f9zH*<q*CS
z|7obgyzbTc^S0fp-Le=(H`-}zLhkLh?s6NjCS8SW{APDbm5lAtPJGpj)Cxx%vcXd8
zbka@r4qFawUNtV6E*RVGnRWWI95?_mmCU-o{;Cyjs%uhMdlEHQkG4-#V;5{!p%AeO
zcqx0S4Y*sgSYmmxl8%bHqoMr7l524zoi#cNnYh3ULLdOLsdc8Z-mcnaT2<`8hn`l|
zH!GEUwiWg*r@hO1KrNQY<<1s1&NBjNPljZyEMjG*Y(e!pdrK55e25_wHJLo`Vit#l
zQOGDIGKtUKKjE)Ah=~Y<Y0^le_#^^%)shk=;$d9*@eQcF?{Gr)bcl2=DwNd)8qW_m
z6i!v0<OH%Cou0+D)(izlQ9l~8Fs^C1c##(al?otynXjn8>~OZ#-+p*{YEh-KRHfmi
zjV~WTM}E2tQ<Sg>^Ne4;g*vG&w;7KwwtRc8$dj9G{$Ahr^T%4Ly+FTWR7XKm4b;Jh
zZayr;lz!!XZ68-4V4rAZC963O4T3);<_<#I62N(B+;eALu#vrSzS4z=K49k4Xs5na
z^2pc#Y)keM+%NM`x7he3RS+U#nR62I7HJ)9yayjGRJ<kdezwXE%XX97X~#Fn_sm#d
zVY@TZt;2Gp0>w8>Mi$r$auTpmSloZbNjEW%gI~%qDma%Y=LFr$@6ZK_4L2c(+hF(g
zY{J+?%483@VbaZMY070K5=1NoKrE8fDkfG2Hhw-IUpsE7Yd*$$Ciz~=)uC4Q;ltoC
zl%=<**$?NpS`rhvx~EH`s+w_?;rp%zMz~jC9i33O#!6%nL*$`_>*qXoBfg!}1<&lN
zLV#g1g^5Qn;@ycb4X5Nv*38Qq*YgYo+*QZY-J*SWJf7@@+l?B*uT*cOWIJJ2%~4S&
zRnsj&_dYO1PnX<VuD6HEiK{k!Zm1GAo+cKHQC%zVG6VUl_dvfXV-$o9_cW^F>#1PD
zEA-yBeJ@fy-9$H(Zfhz;yMODW7cV58XJh{Y`w<q~Ycp)7Kojp)I0pQ9hVooqQeU7_
z5#h;ABfz23FinR+p$+9I0mOtwmfVkkMu`;YLEc|qqLPX11DKA5U7|KxvVz^1G?@50
z5ne8fS~m)&c_=HiIxX^?{4p^<(~+|hj#Toc@L6-<`;A(!Lc$Y%Ojs(!%+q5hq&I6k
zXG4|&sFeU!^Wz5_IMaqY@5_6se?0poy<|Rjja@?SRI`1DP3kKPfwdLdgX&*|{p&mO
z11>s9{-5X=9RENV|NoxO|2zH0^xwLCxBs_rFWaEu4G>tDc^@0l&3SKA9~J`44bDAA
z_CM)<maYMR(MZy@0c<4GAZggMUau;3@l;ao_C^l>W=E$_M$TPDP9C36HWE+bLthfB
ziG#B^7?joLjqCe^pt$9UbAD6;QUG6km-l;k^dGd&lp?d>to&cNpP|4iu<+>l8#lZA
ze0;w{VPvf1Kt;WG-wNNmH6FN?1^qD&4Kp)<Wx9eX_~|sIuqd7?|0T!aAE9C3E!DOV
zdrtqsF_M@0_*FC`WdtdgiCf<}LLWs@W$_jH#VC?@FH+mzp}sX(b0gSDGtH!7s{7A1
z+_BU4MbaO^z%V#Qz45WwMmJxw`RKG>FI2b;aerGF%Oq`7qwO`ryr9SSIHt1!le1zg
zWTwEm%`!>?GaVZZHPv8ttnZu1fi&`6Z=Z9U<zWO>G#2q@yp#$wZ#uL3y_Ty8j>g1J
zP}Qn>f|v?fq%Ejc;hbRnXRGBsdd(TEE%(UfNw&T1N98U)<Ts6JG7MB5O17n+Ap`uQ
zAjq3efi7h<WS7I-w_i<+a?)8&Wk1=f?yVgAiF{F+wKFV(B1{6X>Y>!(Hew$%1UM$h
zCY6SH#wK__@eVq~cGY1v8Ie8D&lfK$wsrz1*}hTNdV}ZEFf~`Fk!)VA1N>(a-i$4d
zU=`|^z1d=w$s%Ten%R&hucpe8Xl|~K>YnGY$5ZRHQS+|)3AmIpx95xc1B}4)slCu1
z*>)j2{B;%|jY!Z86pPpA%ky<q#P(=A*?T+(1%XHIH)Ekr0ZTT$s7{Jg+7EcH_gtUl
zewv@(@8kUNR%))nZ?xQzNvIIY_e0CqQcw^6`KkDeZGcN&Tt7k})0#|}!d({mVu0<=
zp@Vry8T4%ZER6#z_+HGaB|RN(XE?$Im`%kY9N~1nVHcfmt<F;0n1~ANUOiZ8WhAeu
zrzVpBFzmNR*<yLOv{Qa@`*T3GX=9jD&$O?9p_jFW%0NYhVq~4Gs5*boU!~celw;vK
zMp5pThc(jxSfl_w)(|Rv!#3uGYv8O#7^G6DOc=J=z&i*YxSC}l8HF-Y5!ru^tGtYA
zYZ@oF*Y8Vf#ivB|c5N5tRL=qd?AqDR1R-`vs7?t-e11VoN2z3WomK1NO10kJqZG^U
z9+?fGfOc0NrdVe4TW0HX5T@V`X&&c~gs9axL}EpJKbx3)^Jq9%XJ&uvrvYw!fxUrF
z%Crx2h&1!RxiY8;x<W<knStkhln&77i!WUEG7aFt3l(>bcsB`ol<XxbfD)>%V6QUU
zn!amBR&N_^;9<q}^?E0%sE8d!60!zX(aaPg>IME1UoFPB*QtYB+<*)Gcw@+pkP`;*
z*=1Q#sN|`y$_N1%N2koWK-}T%+j40}Tc>GQ<~bqZMLHh?u7fO;3NV-gSkSq1>IM$Q
zmwQ%ZQ>9KS2|0y-1kngOY|fGVx7TYmo<ldI?NDIl)}gf<fv0*wZh%ubAvX;oAJzk@
z`#dJbHC@X+{7HL3j{-_0g5&pRcFyk@ydNR3QBd&$q3n-l4qZpHBdj0KS6_Dh7AbQ_
z2#V@DgPGsl5#cMmHHbL#$~q&l>ZpyXEGF2ayC%)6VJFTSK<agxs#7f}n<+m4&6<-W
zPgSqQ1#45uCUwVC0h=pPI2pmEmzydZIeNj}NStCaGdq2$*fY^ShK&4RCR;W%(p>s=
zl}Sn-e)-DM5=*aY`s`N|E7oiHsBhNF*6h>F$et@K3<6oe5MmY+R|f#`{aCHNeBqks
zBCXhIck`(es@SZ^LOKpH-l(qn1@F%X-A@b2nM?#1!LociWHkBp%oJVsMjr#rNah@@
z*V?sPWp5dK-Xp8P$J8g^dMXDcbhyodeQCN8){w~Av3P13nog;K_IQUGg1zki(vSRE
zw4%TjLe2Q6Wi*8wDsYrJ!6Jly4e_3OZiq|qs`eQ2)TunQdMtW{C8XJN#3SLwIJ#}}
zSWl@jgpik*f_3@&Ik*hBYb()J>o);CzC|a!?4jexvs??{kDS;If56^fUTlTuA%B4H
z!e<gLtXC?}g{0R^&xKoOS@}6An|rh`TlGL=ehGMr->s0G>nZB9bjQGj@9{E$_g|dS
zRsTs?((e<qle*4xRoLgcxWXwzUFB4#T8kM-WN1E=AB;*g(bMW^XI2jG5>M%hQ>$#E
z1@tHPZ(H$YyT1F*wpy>-BbytS)`JR)^nxc8Ad~)vgn_~RL{?nH4XEVcBsxw>UJ#8t
zpNU=M`vfpA35`KI5}Ju^L;`fRx7X=ATyF%u0owh}qt&!ePJRMWp|zz0iylShHi=t@
zVWB)J!M>?cmTx_r-EXG8{ZC;A5$+Z)+$EkGCtlF7{Zv}hf$JAzx+Ul0Kr0uZSKJ17
zH<BoI@J9DJlm{(0S)ZwbEqV<D0I~(^0u%RboE?t<1T-&R{;)X0E-jLKbtd=X9>Rdy
z_5o&OcLgbP8$F3&v9Uj7xPjs52)o=qVb1F%XrF%C5~d_2hC=|?k@75BQFQ|Q+QhGa
zW#|tNQ7{5XEU$CyrT*E#%Wcd;HpaQpTy>>yXQBHaHppEC(PJPg1LSBEU2^R{YQdZu
zWygjesM!EP`!U*`Xxa?x`BQIus;K#TzFj%<@6Rs@zb_G~Vz@Rx(3M!dP#&&T!_yYH
zZb1u-0)Swj2Sooq+ME!&DxWhPJOJ%GA*;f_mz9C-d~qnYyFUvX=OgXxU7<<8DOW!%
z*n0U&lgT%0nqV#^lVSEcmp$a)iEX9J(vQ|1qcxmhj0Ud?^HFsH=HZj`I0NAn;S7bD
zjn`C8Ph~t=hJV0F!iD!_6L9x|f4j7vbaq^}=gOJ-1MMw+0gq`@FY1wxK(wh@Z&bp>
z<K!sUDvw$+okeg^1P455>)a|~t<>h#T=}&g*JF`c;nGJu-!phzrI+~|4x<cMwMc4G
zHx;-v%T6fO90FIYNF5b104Cc&ZwjoEEsJ%sm3uf?Ia^}a6+{l^)yPhOAcZ;z-!6`v
z{hDQ=h_UFE$Vv?(rL2H&7F}GKHA~jhu6Agf9@ttZ4Pb8+Ni6eZpnHN~cqN?^fR`7R
z*zmkhQa%1v!{k}h5+y1Pt36_S!1c$~LtL}<0EH(Oss;h2&6J=WL!Gyr-d8!b?qQ2x
zq2+4Y^K?TW{nO6J(lsKN*M#)pJgkwc)i4I7_8II>pK0BgYpJw2&y)O%p)x8rK=AK2
z$M1^k)D|C|I11=a*0Lg9ZM-R8<cYUr4nHc(N&C|FH;UOB>pH^vP9Qwc`yQHo1~c3S
zsB*oB1b_DgME08&X+0n;o4%~kSgc-|#Fh52q(`m2R4h3i@=kCl=Y2G^v^x*)sY0SV
z_0_Pa6>m>77ovTzrD?0UsW0C&ux0LDLSui4f;M_r8{dRjt4=>VvNfQi%@`$ZIsV?D
zW)s1#fUFwuv$x@&M;X7kHqs;2zKYHu?gUUH6|;!)cD>fEr`~uLX<{h9SxDQyAg$un
zTVKM;8)Ru&vj|%j_wI)lR&BKJk1>y6!{ul#k6_0F*>HfF-1PGed)@VjVb$2^b358>
zhui7v%L~oA)E5k<r`1HT(RDLTeI3aEW~%hSo*nC&qGaHk8X@MszavfwUzU2x*#|Ny
zC}4uc%-dzu^FO`@cMz}o`=i1=0~mAUU1k5EA2>vB|EB`{KM~~EnE$&1{91cFmY5x>
zdsY1i9%Y)ejbsi%q1cvnZ`1w5^zEqLCxb7V{_zu;gt+bauZpB<F5pQ%UBm%W^mDgF
zbM{aA(bMz!^X4KvI&cztKUHn1U9UV;Z~R+rBFEX)Q3}#b+4<({`Ki2;Wmv7ccR=#U
z-sIt@zu)!wdYL~n%8bHWwNj}WJe0V-R&MPb>VuaLCvOl?yu&bcwf`dhuCI3@3C{2L
z!}Ynoe7A<JV?PA>^3&yOX`WxgBLCeaJ~`#rRjFTT`7SwwRC@mXH-Bj_|F_L+LnHD?
z$hZeB2}VzqM*@-p2a2*5KK&U#D=}FwK&`Rr540jKm+oPhw1Sn{IRH?ycJ{Fac@(`0
zMz+POGn>zf#}o%v;jOjKsod<uMUkDfcKa8q&O?)VXA>x7w0Z;G4h7c2hO$mpY=XyS
zo9$&8!pRvdi=&CF%4i3wE&|87q8dwE)>2ZV+}etPrk$%PZe9ba{+gxrCP|a2I@q3@
zYnz5u3=3MdD_x`PPpVCL<j!+ACq$I`&2j9-5{i91e<|Y@va_1Y2Vd+oCj@I=V-3DQ
zb(SD6ij(%=0I&k_@i&$+CG%tn?L>xn=g=Z<<+aZw#dtJwcCW#Y?kvifdJ;uB00RRH
zFUbT?!q3l@o8^UZV3p>o#3N$#hEfxC`qY>4ETy$#T;cW6c?cHfgEgxT$!VTO2G`u|
z`7>aewjmTF{jmUVt)cN!H%ZxnkYeqLN5)cYl^0*zRea~Ud$98#Xx&+Vfk%r7)dxXI
z@*3*AHkhdx?slHst@{DF*VLgyvwa~A3BO@*4d%}kjw9snesCYd_?LJVyN2pzjBJ3z
zhzBS9foyMM0h=HC<6M7A0K{<7r$bifaq(ul_B=}|%S<nK#QM2rk|DUY%A3QlT5k0W
z;hWWJNxdR<M<W7j?-?Vo<)S}ot+eZ%fvJt@Ie9b(lD8*pIMzX=Y2n3d(7feA-lxD4
z!{K$F#>arhaW_Q9t<W)&dQL-Pg9houx!ux`xQBP^@0vBiM^&IxZD@?H3d4=8mj<+?
zPJewUKh_tFL7TNbnHD0}!|uIXyj7G`2C$P25h6Tj!Ut!@8(!kCvxjaYxso15vwTW4
z9AmCeGC6n6JM|`XH>;6Atg01OwuKOkCv6h6NzxCeuv(ip3@Fqe0qQYJ3{LxL_Y|0d
zm4<oA&$Vi8G#vi=fY~zcpD8fQDQg%`tV+cxO;6+MWC+UTLd_RPJC5Hw6qN*JV|5IW
zQZL(>f@iMhdg!HCimEq_3aKZjgZT`{20{jmKUhH?XaSufL1`_fN4v3RLfNEJqY7iV
zB=CvC`fy@q-L`1M8v?Oe8sKX(hAN@n4r>=JP)U-*fY51*kuOhJHX5oQW)<N+WH%*E
zl8-qlyTAHC8y*fOlYub*%9vngiODvmCb?QgPHC{Z*aWu_h86&LgBndgk%JvLK9$yH
zVX4B+`!W=&gm%Y%U~ajOr`i&#*6@BhtUjl+MO7*fSnx_MpuxT{KD2W7s8`SnSDD}s
z!OX2lrz?gr9V5qvDTuXY3^6E%rjj1X(?9z<(|>^JT<Tw6)TEF?ciJB#8{5%x_AiHn
zwi1d*3^6riM!apF<;f;Lj0ZSRAi?EplFd=8H!lZlv0SV^EuJ`N{&Dklv-lc<gU#W+
zW>hZZR$0UD?}SASSGLf53lUC7kfN1iC-YTT>tIAmQdbC-!J0}qqmX0WtUL_K<dM79
zDqDb()htSkQM;vL;cRkTE^>;Ld=o+e3*Ay|YHCCOG&E_L?wD|Ka{1d_(6A5KrCX2)
zF~@)&19sPZN4~Ec^g-pf0m91weQvO&>Qe&3*_rNgkXJicih5;49!xyoj$JSR73dJF
zq^2?S?f1j>jkhn?OZz&LkZ_P5^SjrX9MPGO2SG56WJi?K*djed6)4gLsX=2?{GJDt
z2788f+$CQnc|Vz>C=Qj%%`S@c)Dcy3U+`3ozO-~&0Oc-%oMky@4%x(rA;40ZX1l;&
z%VDpi3JmdtFeiqguGm$rRR>c7cd;B~uNT9u#vwT2PXYa#ByuIDg&2hHe(xZ@JgJPC
zCMDyW{;hWB)Y!B^6^sQQvOGprrK58#SM`h_uPmWsa?WByl3zLV=U&fA<7$M?<Tfqs
zOPvaEZvqf=4F`E18lCyBP8c=8n<y<rrzU5~%`*Z7P%Wm^wjC14i+-Eq)4~3!h-$KJ
z5-VKI?3|qB3Ml>%U!HtaZkvHH<jcOIK-|!j2XZj!r8%)(ka-~!@*8b(y{_JQ7?=$J
z#sda0{Q-g{^&t&r>*0fUQZ!Mrw5l>rp7KY-pNiqiGJmc3q=Tlwv;igs4Y|F0o1wGc
z;)|vVPS@Pz$fZi*R8~iziLtg_r6fcK5^^K$&1T^MHnj8&M^0C(hlN<{P68nQY7o8$
z5+dnD|AR~nmzJGyqUQ$uJLkPB737$;8ZUEaOSnq9cwh*oqDJGDOhzbw)_-ux+vk*Z
z%cFYtkD(}3iP^vqGY<oJj(iFx1G|IGuIc<E<kO#_b5J*h6IJ!t(UWQN%!GB0N6J}x
zoZ$@CpS0O>(_O!@fRF5H_gC)m&<OdQKx{W!nhGh1j2P6JF`jV;`RJQ5dDL=X0J8<1
zeGhN9M~mXUX52ZRAkNau=BXE!WVwr4MKE)HZ^I^qQQlQ|)``%JSvRf(X}Jt(@%_XS
zi|181(H}WATou#nc5x3pPR7Rj+%&omQjenh;~#Wx(T0JYZ-?5bID(=;Mf6us+&_jb
z23yb7D667+pu~gAT<qU&n6`$s*u_yu883n$tgX}23t+)N_ngNLs}vLrqsQ4b;)h5h
z+(Ot_Y0=s_IC!|HIyJgWVHYIVnAg)+G43yQjUPHOtV|q2M}ts}g!L{EoYbT{f-02t
zI44hbeNULh@JQ8LDu!4mA7KzA*BdccD!SEOa}%+p<~9FZZ|7KbhOmLCPT$|VygSRx
zm&zght<*`i>(H%-Y>XO*ap0k-^XrHJj(@tFX05mcR=`JO8^B28O=wUkSC3)BL(vT#
zD|7BI#rzKR0!?MSqY=<bTieNS9`KOMc~3?RO2wN9IIKqSS(d~8!h@tj3?J0r&dK7q
z`C7Bzn~sXWS{90zzXU=w-Q|y<x*Z3PvKBWj0a3rmrJaVmw%DZm>9A%K8FWjTXmQE-
ziu60nQM9&QsdX++J`vw2#)_x)cV+?9OBND%t4N1;%ng=j(H4CAD1%a}_^dpHTMu)0
zVivyL^ZJ<t&FS&io$_D3`RD%D9j4^|GJF<#{fq`5=l+&e_4b_B_CXg;1$q7)+QQ$y
zSB9I=Y|TI7yS=67{@P#sKWG0_ZX5mT5vH;=`WUN1i&5~$aod5ig#8kX>Py^op{ZRe
z{slbTJzSn53PsMhnePPU>7~4FbV23%>wy2|Y|8P>E5&#WY60Tzw|f|hIbA&f&Tn%D
z|NT1{nt~M~Xn6evmyO`i<h7U-n(|BffuC=)7x<sTDa(H<aM}L5N%84_1uk*>{Y~xJ
zR)fIM{wE-Y5J?QHfGzVK?HD13F<Dp!K_}<YtMaf}!ZvO5)d&&TM#I{$VRxlO>id(=
z!D4CY<K<#l<wBCt4g`7Ws$G&lI*u<iT+oV)9E52~x}Kk$pU;Vomix?|`AkaSX#XxA
z-}HTcFgOGt0?-bW0}lu7ACNwzF8|Te-s5NYIU%c9<{6Jr&Y$aNdU+n?*L}+`l=RfZ
zP11>~MVJ-(<Yt$B$WdT#0!NYs40n=KEWT${Jc0px0uADRhV&gE>JOr&b5K4&OoUNc
z9O>Jb@S|<ybk;=C$i_Z8hZ8~0=WP2C7pNUPzt90iV)au1vyWgq7hh~~oe;N(!HIKg
zz2Js_-`=hD7LS&hoN*eHV_T4FY=rPwE#$JLO<>$PDYQJ&rt>FIRBG&9QhKBt46`FN
zm|n`D^-Y4BK6sQ(@TMfJYL$mvP|!x{RnS7Xyp_e~tr6^w3Tp5O6+fA6tY(VJZxo9v
zx)n~DJ`+9uBE{7Fjk1Aw-8w?@kiEs9SIiW2K@mUEs3iBEay5*I3AX)?H=zY9GpvTy
zH%C$NEra{Z8>q}T+E+&*Nr{<Hk*Jb#GsIoscemQ8^SDZc`6@#|wXFyivx+?NLLqum
zwx&_ErAni6kvOrIeI_9uH@il9pFEZFwiePwg<PJ_V88?@l)a=$W@ejOkYWrMs0jZp
z#z}~;c#{U>Q=?;gZr*!Dc&TP>9U)AO_DDzFc}+ZikT}8xCw;~His2I)`;=?z^k+&8
z&ejqmF@3E<AE>j_b&(xOP>%J+W&Xl6ID5RoGAzlQ)|{uScsHiO$0G+6d5{)*4@%#U
z@0%+?fL~UzLO#2EZ~AE$GRy!8K>Kn}*@AH33ESWM?e*2p@8`p<|IdJ*Jy;$~=-Xat
z-g~7XpbU`=|AX5kib54xKajeb{Dv^-UVAO_W#8(#r1WSoQ*QK6pJJIa&k~n$!@nH|
za-dYuF@*v^#+Lc69e6H{@(O8@ztos-u=PjnZ>T3}ns2E1AKm@~Reo~vyCtQws=0#w
z@ZMVBq;r8Bpo;^7oRT71p+zQJhUFQ<h&XL)6BH&*0Zf;)Qj(r0`Dh{`N6S==Ccz9k
zp#bXr3ZYaMgKv=lV0uXsiMm*(f$7PndfZJ?3B#Epn?FAv%(I_xjeH5S!KeNX!r0un
z(P#)LAP14EJWCVOq84EEbENILu@gM42ee%3)QHNSa@1*mL~^VJph&!PAtEAHg^gYw
zO#%B&2Qj(@1&pFe0ugg9PAmVYRXh&}htdTS-a|2IRJ?X;4Ee`GF&YgqRC?MsCK2C2
z_jd@XEBw)BuCijqF1P!L6fx$*nKqboaSBoJ0cF{wov!TQ_7Hz+w;K=6Y)F+U%MfNN
z@RQ}$h<yP^x`|?T@7t@=+9<F94?tjPEvlEE;YjP?E_#m6RR!*t)a^K->+bg5ZK_;C
z%ND4Ra;l6h4DL7W8tsb(+@LOGc%h5-eQ&azU0}Ew34v(*-ti96fr0c54Y8JXC0+#?
z8I)mgJWQ439Z3W6I@X&lQqGo}I6Wd+_81#s?wBqXQVdG8WGQ1&f|05)RB-$(=@!ar
zsOeDeSv(AoFrkXS+^RRmJG=M?{=QjI)}-xo&35yykq~p+cR#_vy6XFoBTj2^p(YFm
zjiMTflOJS<knWwCn?OiUVPx2kQjwLU7{*;uShV|1@8R{qn>Qs6WI=U8{bAzCaAkiW
z&Fcb%YXMt^U7}lR0M+R2_na1rw&r_c1hx0E38PhlS;6)(iB%L9ESZ8SoNkzxP=7}q
zzZUC>uZED&KEt$0q_@+FHd-k|!~&+n`x%=&yyV6}bRD8lW7DiNXpozFD*?B3{A7gY
ze?`-13$iXAd}!gmwBq5RVs1ADduk}u@r-^wVyd#!rlOw*JoZ0rzGt9|mk2m*i_&g1
z;uY+VA<BL1?uuePCp@C6PhdTV4La?vqI86dfSRuG3$_iB%1zRhD_bUducK3zi)*yE
zvhkS0BT$IlnXOdi6+)3Q8kia)UXrb7il>3#FT7~5T!)&_IJ}BY<^9uTv<}R&qkxff
zXvPM3oS4xg*j&eN*4B$_(NV2UY4B-2$YCWJ$*Cti0*LGg+b{@S4&t_PtX|ZHU7$3@
zDar}>an43<$Db&_)e=eV@e>1}rI0!oK4PtBn}DcjTWp+_gg^Y)jUcwrFC8f8%vfed
z1sWY-;y8Iak)$9f@C>oUfIU$2H`X*8L!<79c1SGn(P#6IJbsOqOepB-oPaqm5FiE-
zrpZ~H-v{ZnRz8|WHW4Pxw*31befRSYed3L(gj4{&kZOy52+JRD`7I`>SuF8rD~HOV
zX}peWn8Xz|OXxA!1D_?S{d5#l;%3&w=efAsZX|V1AXB=g;`bX|9!kXYL_=cvx_$pL
zo}|8+3}SRpVo6hFMTV(J!yxf4V<uO73Il`J!dK@qyhA*=WuvZB1z_>g-OU0vMQz-Q
zu$}pWojtxSrXWV|{D@lH*NVaG5~pHg9@lj;(PA24{0zmeMrd?99P^IDK|RI-(}gNe
z4`Iqt!h6R>T<`R$VtlCKV&zCL^4JBilT3u!b(ARJ5VeY;sNNK`(0JUdjf+IjOclB8
zg?lGF*?4DPMmMGki^ai#l*dr&EeXL+4Bbv&n$ay#VqVe9+LX?u(i&EQ9ovyISwI!K
z%m?X9{{^PIyNFbnsjzzi=&0pCws~g5jTI>jc`=HXqrohf5r_Bo@HO;BxQvr2_IZob
z0NuhJm<&6T?7&aAdc8J<Ek#+1Ex&osn{2sK-)s-g6FhG{gy5r=Q3-Tms&*1!TzJTF
zEIZO*aMo5Mnp9^W)5iDo5;g;_r<fB*;v+Yc?4Rm!9@T{9Z|d;XdOP-kRg5q(laUWw
zuC>0YC(;G+Epa<zc0}$JF(z)0l^fTV*JfP{n_bE79Y57UEuQac@on-B<%w{U?SJdD
zY=48Hr(SrA88s=zJD2P5CEvwsB9aRu6q?}%KIBUIMv@h%VK^yqsVg02jU5#fI(nb}
zu3ehT2ULHnQq~Jiq|C~m=;AB31J}M8HmKS<!BYrrGO|`K-HI(daFXluIZLWiW*Rcr
z3=)=?7-|;abz$bN2a4*O<Cdn?l-_Sxk>ds|BbjvV^CLnZzh}lckJdO_9+1Xc<zyBQ
zYF2KT&2dqjokhA?G@t02^KU3s)d!OllV9-FD$K0XZD&&B!gHM)trSatf<9WNL3J+?
zieFLGN?n!BD-s3WR1rDVl*6>V*#wR(Z5~Jt?>DYyN9d_u9!W5{6tk4#)_V6&z8g|b
z)UO36(1tvP94?aY#lV8)LY|6Gtv|k6(cL-Raxe6@r?0K30E};^XCxnFz<nrp0j2Gi
zIj=fgjJR8l=H8R?OFiyv2lcK|8Uw}nY`0W+4C-~vZwZedqtWh}wrm$6!<z8B2FI>(
zzUxOiO;TGSB1x&&o>X6TvRzUjhlaQ&0_wH5=Yt=)Z8lw3Np39<6tNZ;Sgd8j*kHI_
zz7znBxE_aJ4(flL!twoR2j^T5*}zVyLnW0FQ0~Vhqr|>vR8^WiT6arnvHH-bCl#(d
zUt=kGVQXK1NPaEU>J}3?LYhc(DoR%?^8#wJczikAoML|@#&byXrdrwr`~vfZero)u
zWX$#-3}P&djQ`c#@-_BkEY3e9roLjkNk9{g)DkoU*8S5A#@KXJJFO!`w7X+n(A7!&
zF8p&vqRQ+|Y77pOfQwtXL?pc;isoQgYXCgHynJn`8$A;EU=<K?=qf$%9I#wopM^qY
zKL7)*f%`Yx*GH~X!MW#OXwI}mI^kYgpUO$iKvD<-i2PE3v_`>vM4+g<e+cj1;mPA@
z|1sVX<$<}h-kzW1A!Mg)&TGGq+LxR6=cmnb6IF-Nrh)y()+9fFr1U03fWeF<hOeL3
z{mFA>Y1iy%WVtT|@3(Iob1`4Z4ae-J{~>rUpxG-C!fA1neHWI|eq}g`A{|2Os&{l9
zDM-HG3=a^A9u>K<7N$E7%{`{@19dJ#)N?C3Mzd@yA$1)X1*?T7Q)rvp2~$E`dVYQt
z`GIUJfoSCJ5-ft5IwqaW6-m4E#+#mNC&*L&W{sps*{`z58GTVsF(t`ql8pl*xd_25
z0X-Yj|1?BLgch*yJ}W72ElM-@#&$qB_72(Ag7=Qh6;q!NN|Wh4joIftg|0^FmuTi>
zrxtqqO0XCN%scW^WY$H_KbM&^Va^V^2azrA4zxar<u>B}#ng37@Hsz>>PL(EI<0EI
zizE$k?EQE@Q*C7sHHTS};U!>i7ebf{$Z>GdeYO6F3Ey07w2Jh~)fJ9PGnT`x{j*bg
z1ZJP+5wbd-U2XFSRgjLjW=KoP6lOi3utzp#s!gmbj~}paS)^jTqXv+%P;3&2mMGSh
zk{tc|H{8dAF*qd9f^nUq3gggB=M$TdC26XEzO&JD$B;OlXJWkG1n-|n!*DqC4#$9}
z9bpAIO~B|VOXsou8sr!u8PI|bvwEYZ0xp)P+@n7&jsYiKE<@>1=~fQ`Y&e^7Vnq`&
z!~lO5s^I6XG>?a>V2>&C4~t98Ls*_A_>O$a%L9ohc5urYqr4@Bsk$P*V~@QT%3@{!
z8trj&7o})D*hd0|rO#3Bxg~8R7rcp?Ck2|zOiEW$Xfcy*awR6h^`ASwj>o1v-ee4p
zlGsR+Y~zvX{NXT?MA{WoEk9f~$cQnTOw2e-ZpiQq;UE+?$c{bE`gwz6iqYs5UUfOy
zd&D@^#0JYvC7s8$WVa?H-J~>6c$eWG`<NY%Ic?_GR=XA`+D7lG^twI5I}`ft!h*Re
zt<}X6)?qxhYRa4Sri3$tteL}iBQ~&&Ph)J7poMESM4}kJ-kC^v2?ODce-E(@L0_D(
zKN(OFFYA{X#2`$tkoXac|9oy$5&!Jt`F8J4&)56ouK6h5&d;^ryT6PAo<*45d1{NP
zQU!KTiPuvGZvn7Vve>?VrsNSwJmP4;>>woNAYi#Q<HvFwS14KtMZEC#kb57GAY-z}
z9ZKmu%kHY9{gGuh5mdtZom;Ndz;kPGUj0{f@5CpJl*TWHoPOwHX_0Di$UQ--_<-Ev
zZlYSi3~4A_Nr>qqf(OlU_CHq3OuT$mbn1g#KFT|5VneV==mLvf8p|ey6CUN!jIao`
z@|ZBO)%05=9+swsu~w=K=SC3;vN4#WqQpiFR<5t#PwL$c#HG<a?s>MLM5r?oXJJ9e
zOmivfgkiEM9V#NSd+5wQgc6Y9$aAWhmX-l5V&=o5LLG_;xBP$)R~5oU9IB?RKA|!P
zg0a}`c>lI>EPLa;K3YnN_tHOSn-Gk?x>g`x_Bbo|kd&^vcBriwZ(bkJ@jOpYiPhI)
zyzMcRor{)qsWZ#FMflvRw=tH3y?1V`nB~CT{Z@<{XGkL|*a$9BY-UxQ6;JGN;mUC|
z(C#RfQ@ir+pGMX!2`*@jLQ}R3W6r-`7NqZ4>E}OMtBg*Dmu>^5dAeF?Ny@aeQRBmh
z4R5CkLB-dXI#>E0g4%!oH0fMG5xjkO%g~Qhe`anyNQDk$2o6I01I~(fU7>MsA(*~y
z(tqn6P`Y)t3u-s+W{PosaKk}x$wDm7I2RrQER;x<-iJ!*)tP_XUqT-?xv?=v1RcLz
zPfQVNf>N2z&9I!9p`<n1|D|06kLw)Dw|_>MbJrbOf9w$9^w1sXzh7mnot&e3L^7|4
zf&fR(8d#QAB%N4phR<*!>7wWl-wj?$ZKdSI%WEbCC2uE>A<?Yl($K?HV<xSbHiMcd
zftpCzobJQvejIVl%|`$OGri{Fq9dvf!JnT=(qaPVom8P6{@nC{EzWV(1KuuZ+f~@2
zylHiOwK<g}P0z{<VQ+dQ5mh^NmhWCdd3{=#`sLrA;U?3vJ>{j)>UGtde5rTU3tH~Z
zS#@-&`uEekGn2G{U2Rs_QP6g}&f4TBDPj<rHx619Ve2VR9rr2xl3m)Zhlg9xY*o+Z
z=~2WqGFh`b*v&Ie=^OOv$9((A{s>YWKi6u5)y479yEnGm9;ZI$R2QRe_;h;!S(qU(
z&&X>@F;mfVb3>;m>`@xED&Tf@z>4D5;eP$4yAVhv@gxB~)^2_g^mqO}Wv1>yao)2C
zbZ5-4<ft1CA5*C#!|G`gx{tvR#kPH8XI|)C^8!=dCo5KnFzN59RSfJf9Bf7Cfm~2U
zK<)*jc&1`p6=nkyt~I=4)?*4-hde8p)Wo4JPrJkRWuExn&kQLj<@l3xO{O9IX$9pC
zr#kU`82+_RSQ~CmW$4e(jEk(B@s|sGXJ<3^kdT*~om;Ib%)UGKvTknT*!r05MqbyJ
zvTT*D<<W|)?cZpYNY6jp3K8+7zHE|=7jE+1^HP*lA)xr$&2MV}6ReWFQYx%AB`<PS
zV}afi%qh+C6j@R(Idw<ts~+m464x~ZZsqjW&5g&|caJ`-k^-XZ+16HxKVADvu9A)y
zZ`EQ#6UrGYuDnP)c)jK}L36Yw^+>%+U)K5DL$m)b`QL%x+lhDdCJw*$_D|pRy9JNX
zjE*EA^H$0-dietBK=0+&wG``dSK^9Jd1&ga+fru&+N!nI*lG*;vc_0i>Nmz)t5@z1
z$_)yRl_wo))96n}xmMw!f$30ps?A*aeK}g36nIvJ_(p+TsNVhSb?67W`bET20~tK4
z7$qYb5O&~hA-&~To=asxbK_;n{o&&S?Q%K+-ev3uI%1OrYkF280QIzDPF5j<Rcs)~
zUM-^K!|Ialafac`x(MuNx5vJD29$%!<14C2tU2EK7yLkTCj7YsyTyEhhuKdTSvAe&
z`SGRT`rU^>*jfBtCUr@jZky~Kt)9|5y&heg>wgGt>}>xVk@5e=j_2fL`v1`JOdM?g
zU8oz*w6Q}OM;zUKr2IH0Rr35O_@Lw>bg<V6j~yBi=@@E=XT-K|a8KxlxY;~a@8%xq
z+2-A3a7Eb5%*egTyr}T|uH2}WV3}zmGW3R_B0+8_i<h-l4KY-DqEbmxkISU|Ry<h#
zQIi_!%vPC@sOA%(8u6MRR~D`C(P7{N$th$xBt64UDOQIjvGbOLr4&DuUS!vpYvqeQ
z57OqHCbLgsbz~&0$yGxRCQZnMEI;=i&QSPRJdiDuRHf?=UNNJa7wKIPg-J3_NsOq!
zpjpmJ4>W^Ci7T(w_8{)$`xoMJLi+1p+_2@fZ<yT{UCJ@;_3~CsNy07XaFmuc6_gCa
zs{(|OI=?O1i2l8Br+Ticq&u{FtkvTZv_5Ez;~u(&q)A<8owQjfk(L{4%$mH_WTcWE
zV$782IqRHB6;@OpNtx~`f#YyV1fog6!tp~#7^20-4S!sVqLHg|(UPRkUhQG3KJ(9u
z+EsJbxs|==94KbUoV5v?e5o^#gnyf&RX(~6rPSL!M<f?ay2UL!Ksv*f>yA<74KAOy
z^OY`?`Utb=up<rReA1d3vwi8Hmk~Y}LyJmk6H3$dZXpvhdq&JFRKBCfXx2pY9O*lM
zSTeLeXc&oIgbzMd-AYxj^tBfpu#o9qSrw)~`C<tL@l|5NG(uRHF2*ZgF2p^aSu$_z
zB)V2QSl?Ei-72}@mCHr0IM6?26&Wvys?K#GDr8p(q8WMLem?h)Yw=A73l|rM9oK%}
zJj>Pj^Rj<dVfblgcBW?FaPDv6!g^t5Lr1}VUG5ECyLj&AH*Y2=H&j^PGEuikOzoT|
z!v={5pFRiWqtc@GBi`XITgHI5t-sU^KL>?>WD13&)%Yt|H|vB^x^dZ5%o82h`*}ea
z9NHYS6~OUr3~qrtI9}Mm>>#6INyA@7zBtGU+x7yc845^+qSOwF-Yluz(Q&m@dMzSB
zn9z9!Js2vBaUO%EP?7=Ofocfk>X)^>orEK%rcaXgV9OKKiY9`k_R=C-dq}bS0^@-{
z8_rSw7)=+D3L>LBYcGI&B8vRu1F$v1y^S1Kk5Vsj3pB~nHjGjWQKgKP9U(LAfoLTg
zV&{7mh#r#1*Gor9X^31Aa)72sD(M3qjRRt8toKFjr}_OrKY>+ubAEH=;oIWw>DHs}
z$uw`I+cGIdt^uX%@h9>&*683_gpNkNA6Wkm&fiZxp-<JhHR;UE$jNHx{=9j-FaA83
zb29t79G)LIk7LOf1{%Bgq?D2neEsneq`S++`VcLp%(v`S7_FeKqR%I;uA+}*LRxsb
z2*h=><rO?7ZKi&`+plwL;n>Neg%9l}mk&#8oSht$&gyo1aved>p))I;bOQZO5u!Ie
zjfyiHZG1!>oztE*vy+SU)$qQNi|;n&^7g(xi)P0>R*!T(<<ze)@A(P}nkij^%aN-F
z#SCGA@<5rOSW+~%Dri&0AlF$p6#ct0z~OYxjPLtTb<1r{#x7;C2SbO}7(zVYlSyi3
z+s0Z{n%bw&upt=lodI+O#SsJxp?(>7z{dfcLD;*d!{(q*fC_h$u%}2U0{$Wk%q(R`
zmt007WIjqq`Uk+fTLgTLL%sv3f_(0;>19{%98qK=LVS)^pP{5D0jvy(0ph|#D>WkN
znnrIrQJ_zvpBX<vf*=nc$b}~rI(vY!$g=>)b6GG5si;XSX#*KP5`puUBGDUD0Ngn6
zgO+Cab0K|M9EbqF-94lR_A!#P$RR2b_8^D_27jy6vp8+yg2>pD9POTcB2rc&ad3#I
z0aWQ;rv!!|H9a(UFp4p@3%^c-3p8e15<$qHejbY7iZuY@(y~U`+LX#++7bdb91E-t
zC{gPOPjjeNg|5HGQIcb$n$YrnO42K3*e>)n=enlcne!T?vQlIeZBkP3gS-JLr(VTD
z$rk{CtwM+`(qY^IRL8?brleSG3i5~VEf=h><zB5~=Fau3;<EC``59MG^wif#qDsi8
zHb++zWvrZ!1#vyT#!u4&bT|GE*Ck35PnJ>IZ5YR_@?W$VtEfM&Rv)VDXRDsw+6%R4
zy>qopN|j1v`6B<ueM>=D2b!#D7^195%Ui};)&j5gWuV@?^<)g7=3WERFLcQ&uaqWX
z+SE?XM2(drP$=2~Fc@}~0lC^QR+@E`ZFR8h_vfqs_&X;h908cnuEs@AH2BYO+y7MF
zbaa-rUEOvL9!NzvQqtI--!6~tessQ|*xuf5e0AKta;>1)RDo^`=P;sdXqYtBPncg)
z6>_md`JxO^3@OGdN)*P67eou;hB%ns_VE4y2~eG7j+0CE?I$07S&wO=5!F{;DtMFN
z3xg35#V#F`j@`Gh{G{aP32u&>7kpoUfL6I6gF1%7D%^DHKqH1TDWizBulo;{0IV*A
zYB->22$dUauPuB?bhuIkaG`*h!!^qEe=&fHR<4A1e-|DD@E5CTfBR(o6&n5rGlf1%
zQwZ}CG1Rg`PW=9+2q4zrozl1jm~4S6=61<P$uF`)1P_T9?pgrh>XV98($f%eC~h5;
znsZ7qE!>KOtE1r4_*T4w86Mf;IU{x2jGoi6Bz7=CB(HOmVGU3hLn3+dT>_LcdzP?H
zgDX!a0bzBqj)_5%Dlvyra|ZAPe%g%4l~*jcuu2^0`<x#6=A_S|k^41&CNYRi4FSt{
z71LBp);O8b0U`nX+gr>q1Q8#kMbK{yXlN9f2{TbG9$$!d(CmfaX6b-liXPy}aDKxb
zC}1H51tNOc{b_UU$QFh|jK}NaTK;z)MWB3|#gWQ&1j#<MN|xoiJ`i2IVzHT9p-h8+
z`635VvzRGzMpHht67-u*R$-lnl-OL#A^T4X@>`%=Q7AMn5mz6+Yq>o;BP3f&nRVn6
zB+XyRK`K?hrBBXLqF1fb(N<6IFeuqhiKPT?DWx9-kOUtHI!R)>MU$JF(SH)3B6acV
zG1nT$Bt=aYukAm^RGU&;HpI&pBzxG=OIwH=IVywd6kn*0e&@yqE#7=buzdJOR&U-<
z=-zs>tLk5z!L(ohaOwAuBl@&x&2cx$NajWi=AcL61CDQdk-u<pbf&j=LDWBo!_42W
z+nvV`gYS7^`-`RWFGtT_US00|U79#~@XTV@xBOCGwE(Zy-}X%UUCmMKx3kyxVeeOs
zKI~b37&E57e0-U+aeA|6k4dX$;%2>Yw{fwvXs3gR-TJxwn_kba-D$JA{2%YjjlI0z
z_wzkIt={~ji}+t`TQ$FX!G8S!unEPzz3~7#-^G}>4E{%(es`Nm$DMJtpRY4<a;JV=
zT)h04KOgr`M}t>WKc|U_7sFO*8=BJkcz2{Il7PwZt8p@%TlPgQgzT4(HpD&%)QsC^
z@!M=Qf31g9Vf0bekCk+W?dZju$ICwnZI;0<MrT3P9TE37R(w`P5)C+L0j<CM%5j#z
zBksxU@mg_y6W4)WZ~vgXwes}<QkfuA9arWvwbN^3jphpDc)>BbG8>C1Vs>IO4gXvQ
z@QVZ*E+iJltnekPh*1D3UaNus>eqrqb*|7Q=fkv0&hoobII#CdFy?Z;rB{t2jw*vg
zpuC^-#7g)8Nj-)j3Bv0?y<)miIFOIR9W=2Y(<dWbz}6ajg(1P_K~U~D@txCipHRyA
z+fS8M%j#RtRS*k1Kpt98rY@%K&I{~|)l6C>+6vLPMkk_KPd<YYH@jPvf9;m--Ki|a
z40Nzpfo4}htlK2ehru3A-dM%cpVE(|iK$2%#j(ZkZm_(VFRm5J{?8xxj1O@9A9&&n
z7~=mB%Q^nJ5C4x<F~)x->wm+6_@7m=L$<rzS_M1Zd;<;~zJR{Ikt~$6YEfHJc~Lic
zud|=u0^!Ky;RNFeHaFy_e2$Xw{Vsi?<2#f1!CE)@S~qz5GH~)Pz|zpwS>QSN#Gk+-
zK`StD0-Axl50|$`MJbbWkH6y55fX8Qzq~$~gS>&Im;%iHzMvdi+^{S{izDv>arnRL
z^kH7GT9a)Dl(qI@y?TC~Ntm3Jfib@$@ITk@&o{f=u@#C+8h!h>wQl#|F$8m!fHDYL
z75kg~gQuOloNE@$l2#-9VsN`XFxE$LQ`H0!+6!7nRJH*mXczq(1Pnhcj?#XK3kbs>
zA-Tl7365oD-tPBxP>RTm=4c4lZJ88Ua(+l1%Mb-#fhZsjnZim914fxP*>zy=R}{i3
zO&9L(Ke1wcj^O~8pWL)nL7>}1oYxP9MEK#2PZkoTGGBQkG2@T19@!&D${QwvIrl)4
zfK1L>>G`B*ntBf=KE<v=g_q-LScXyb)_&X%MDyQFzG8eL*o%bx8TbiIsA4fPyUe>!
z_XyKXb;^mH2amEvl<<L5KI9e?*n+dwMFwbk*~p-oqJEG&7$IFE+#bxH_NYD^>iLV_
zsOPiYDO60U7*vJNlaiHmf|03^4A3tCxx-RHNPu5Jm4_{+g{T|vxtY4c))ub3++wJO
zjlWY=2%`V6Xi)Q1=OIbb*)x-Bln5JOJ(e+1!}`?3Fu{v5Xp7<VrTmbGmc}!~NXi1x
zMyH07sR+&A7AOh*_nn!XJ%*6uQqC<T{798#r<_<nn7^6=XtNkD>pXeF!~t7T<9yio
z^*opTrt^Cq*~tB*dAaq$2+S~))jAV$ia>(raVZpI#+?mZ7WWc;kwmwGY=qg$Bqeh2
z^jKiTflec<=Me9^XJe)y-(QnqZON%;&!kdwE6<5}d`lCXGYMLdi3MYSx^L167{goJ
zXoHeb+~FK1ZA@CnP1VbX98{p3a7fz{Ocm1oM-^@41P@csDpTwg8C~ZKnot9j2xmrV
zCv_yd7VL%%*5)-v>Wkh!4kd0_0W-en4wOh@!f}}O77SXbJZwDoI1lFqbIM!nS6)<3
zu?+7$PmE@~29GUmssSoFj;g=-#f{iL(Hb8Zfi>?eC#!^#GxV=^NoT7E^x}Bf-37I8
zNnneSDDKr3JdfHH$1MAgQZ*9@7->`r_5VW44bcsmXjq_%2z2&UG!_sSLu|aVJYgBe
zwW$hQ57UY%x-J;aZ$+xMuE9uZ2CPTeNcOG&D>T0BNTZl6uFl}dSmgrV(t8u`$e(u}
zhSN@;#5DK6RY*~ulBC|9bn7fsFmF|j>3H!bd!pv~$warV?kpbK@>O^H^id)+hCFK(
zC22`u8D*f*j%sgF`wk&Xn|dP)CH2UzlY2=WPGNL9EtaO$X_A_N+c<n_xKpzIz|@a(
z_)ONET%aB^KD@6o40vzS@{LNCO|PN`e>tXCmoE-GNJcga2incgudxb&mm!#FSW3n-
zOnG(+Nk<gloV}gilE*viG^;4mB^@)nG-XJr(0wHuHMHq*+F}_)1g{X3F^0rs3>u>_
zb1jA}KV(wQhGR5jU80~1j5;F)t(z*}e%P@EufT9@VQW%(DEH))P$HsoK`FN#uge{b
zq{zeyc(Njxh$)xHGGu68O)+aWr;~I;m6f!>Ty|VJ7a5n$hpHa*i@>vFg{oOX{NCi+
zD@wSCHLlLpEyH=gq+sih)i2gJc~eyx4WP6?pzwq4c|%)pePDvt?ijs)-Hwf35|q4Q
z$d7qHF+#1#@dxv7o*3`16Q6i@*_py=QH?))qgwWg0;5sJrAnm<Oj%73uxrB}yyRM_
z+|Jw4MUg=$esj9`467=+xxTUffamE0Teo?O&5W9oMW`rI1FyP!FbIJzFAiCT(;^#v
z_R`b$MkSAXV^H31Os=-QJUp;2rKZuhkTnYoWdLhB=XsWybW2OqUasGM7LD^QOh_hS
zOJ2Gy1Q&MS7D2A2?vTrJsI7*n%EmXrpf70ZfzVVRZ1`3}rE|F1^gizozx6{S>r@G)
z@~Vgu&a7%~Gzf5R6~4q>nyt#UD`l9ZCt`9}^T+1&eaXsG3*j;>;F)Hw-t;xk1X5MF
zxo{MND3>;Char|RfJvdQ$z6i(*3=MU&?Yb794yhheA|;>Ho5i}sTO$ex^^^^(6#<K
zSON-*L2I*V>eG(Q%9_Ep<gnAp!}{1>izDQUsmuCZEg|R>$uI1^hmIPrIumV3kT3zi
zBE>u{JvXJFf{JkXN&Ko;>#S0>0}M`sevnq_X;POjr$~hC-EX%A#qhh*&~*r3&AT*>
z$2#p;zk<^w4O$zc;@tL`b)hZ?v6w~`B`8{z;+5}Z3Wx2v`|fP3PO<JJ<HK(Ol#j@}
zy{9AXnOFwn7^?Q>!Jjz()R*uM^X^(g#If^fiT%^qnY|-y5yEIo4$jC$p#TlnBPe>P
z2taplo6OQ5UN=3eG*^`}BkE>@0OeI-3tCs){(NIs*2cPW(zT<O^jUF@cO$FyzZiR`
zAko1sT6eW=cdxc>+qP}nwr$(CZQI6b+ve$8_dMLHy{qniN-FhN9+FBjGv^rJ&~&_k
zNlvl1jh(T%z0I0H)wS)CyLOrpwwYa$!OLQM^b%&eCm<WdXSRuz8d0w8C@9S{ki~C<
zdQrW5jyQ^3L02Ad<s!3sjIpH2i^79=hJIuoQpq#@7^M0)9+7Rq-F^8!5_F1*h9E1(
z(c9-ZD2j(NRrY$xI4$|dA<K+Y#~pyes>$(Zk(0JFU?{fZ0Wpbg&(pzU=~QvhCsbkA
zZrC{FX9kPr0p*LfZHPo}l))ahW{|Rty=?@l?QO}`soWS0<VZwvxRb`3?Rn?D!Y_(Q
z9m(Rmg!91^F@}s=5y?rIOa$s6`TK3A=59p4i%ibES*u;5al=zTced8`2f)Jnk@~;Y
zFYNz?W5vKq_rKLInv$_Z%+NivwPZF`Lm_MSI*6F&CF%V8pc%j5Q9mTG=O0`<3UOuP
zCoyMN7qGY%6O&5pQI{SzpQ5AN6TDDdoSa>p9&cqPwSP=%e|Nso{dnU(k%)SkJnH@o
zkfHB<WBR<wNHBeJJqRjL496bzV*jkf;-avX{3-eKjl6ePNW8}?g?i?Xi}o|JIvNWs
z#1Q2sL%4NNYs<?sMwErT6T~}M`|e9?3+KzbRA5mJ%+CAyyejzWhYWsX_lHTK5QFE7
z=kxAXaoQzA>OSpNaPr?`uHe)0i4isJy2b)}HI${l5z>NQ6H){Qspb5Og2hUj&RhE}
z5>u$aq^&Ntl>^bZ%j$3F1Pjq|S?(+A{Q)5>KmW|9A#$(V;J$2DG=g5;aa2Q`WkCy3
zRB#4IRr&tgox|7X45R|3CV<uK4TkSb_3_I*?2y@F6#sOC8Fsxe4|nRi&gp?f)Q_y;
z$m}4S8#G<=Y~i29p$*8wWj(Sm%hGUN2?~P#IK~O^hMG|4`zGR1?hAByosH#BRSnY1
zeFSBx!Jh{5ns_V}2&J}!OqL^&Be%+$!<0*usG$$4unnlnFL9^>e0+UlTjPSKff#I8
zT5lI9iWENb))u|EY}{t@UIJRa{T5=NDrQerck$T^5;u7$m>yC$$r$Qem4{wJH_G-W
zdN;}Qo;_~keu%DQkX?kXW{_Eb*U1FxTazjP2=3fAhg>;|{tHF+`wN*F;I)KB^t($+
z_Uyc)GNL1A(o#gdFAE<}uHMywMTsA3$>KCYPvj6$Bk}duxawutL=Aex8*_bm*Qj`k
z7*}?>Vfm0s_G|XKpel{R(xz;yH3@r35K<BIOb}_)u-yZyK1g711q~s1BQI%ef_HR~
zyF~38nHfU%@Y`#)fQXC<$*3X{S#<C><4;^#AD>>Vkp|`y+`{|e%p(P3Hjr5J<3H@-
zyaMNR`DuN{+j~LJD`nNdVvXx-bCPz_k?!F=N{t~A0{4Qz6l?hBtliQiBxt<Y-Vem;
zc<HK{;T}xyQ}jQ@tljTIu7CGHIhz!?spZ+S6uRRoq8&*nA(vtn$wBdXWsaWIyT9Bw
z>MqjPW;JWi(;r);CDG~#Ic@o82rXj@tpkXyc-N~@g)AwrvYQt(QcbT{Ho!DIa8@zz
zkKHu-U1x#O@5!Ag3uE#X#~PkW!g2e3(DEmrB_9jnX<{+@r5Mu|H;L2CCG*W^+UiMd
zwo?L<HZ&eB2XU-y(^IaRA!qultawAJzVcF8$SC`Ir6qNR$G_yYZZ4-Aee|bVt`U4r
zmQ9z}?N=O*9!@siVR8cm>P2fWSTFa&4^n95d;dMii`jCHkT8ondP^Y&7-@{RsMd<p
ziwk?pN)6os(vMKac+$&9xL9^0p^k!?(k?`BxI{1%W&b&1fs11w*!U~%5|z)nQGB@8
zAj#~DmfO76N)N9{`vQ^z8>D2X!0e=P>sV{S-7r@-W?7Ggf~+HIT#myGR?iNhZ6vh{
zK0F9~n(WSpEg|eT5U45MoC}Urx);XnuT5w5uLBR38lamiMCafgG-^j7H(o}qJ>k?%
z+p$Lxc7rOASao`y95ao)^77yNR!|^vzEM5NBLC;#RnbahAet<l9gqXqar~6XiJ3Ev
ztw&J|%^j_D<1t~~m?3oaoaDo;?;Jvn)fxOiCx4PUtS|QjM6W7`wCD^Fhuc1R2=SS%
zlMG=r#aO%;NUOcqsJg<w@^0dbs1P<E<Q<I}-blG1Z(BP9sNkOOV?n7sHt4m#0Rg2n
zp<Yk<LN{N*Si_}&&eU5<e>stDC}N@qq=mb)LvzUYqnu-Co^%EDG0;yQ`4|Qq|7va4
zOUiHn@|F}&W-+B7yp=S{VT9%T6zdd9{hdMAqPC%dcCC`8m|<>m;>Oo;xLoV4W@CMC
zMR5-OwjQpZ@2x($$*dgxzWOy;xCWH+lw;;zdUaHKDqkSH9K$G#Z?CCz#ZA=a&C}in
z!*|s9Z9sa>L)A`6RLo#$ugkTcw#`C?aLY6V=wIMbF260ngef*0yLAzL*S6v!XB>T%
zRc^v@Uj&}6<L7^I6`zgI>L_M_jz*Me)KMv&I^glqz}+=Slf%g5o~n6?^dW~{AQvQN
z?`&`ZVijMHIjWz_7}8BYdb_{q38m>Ou)I^HOPsukTvMD`COocx@Jc@_MfD_o=tK5R
zNto<bk5v(eqWA>om*~GQ&1WNM7;<wSjx5_p!bYS1P5y+^aLQCw<!Ct-=(GJYwViKY
zmU3-59q?DN|J0)92ue;q^y-jhSslNYZI11n=?}(I@?bvL@+>sDktPNv-oOSfmMx9J
zw0=H0Rq~}FV3#K)GL8B2aEc;~**`Bn^|XFSvg9ZBgFE*Kha7ZK^4h3~$j{F}Q|WbU
ztVy_7G0DbvS-RP%6qNfn7(?$hefO-RJzPFtSy{%JbWIr(z2RN$x~qfjkTF%(p0A2#
z0mfHTs7(<A)4BBEYOjOU-J$TNL<d01UUn8QUqHl<X%XXTu5sSH^w2w4y|olvbyPA)
zV<fy?Oi^hBE_)>oRk4@%M-|m8v_r0qu42W36|>;qz-bS5?De6^<uAnGDafY3E_ln~
z*%jD^yqRajntuze#k7hwzolu1GoYnm2V?4b?H0UYebth54reE3CTZvmy3F%BUXzsT
zJVUXNj>EdfVF@N!9hw!ftJ}fncO%7px;MZbbOUUcx6;!ApCF^tp9AJSUhh1o`Y=ea
zC*FsXD4>DHE5PRz840dzbD`pJ?R`Mx!&_Xy(>D31FUB$Eo^#9iPdTvq2XFVJ-Fe)r
zf-43WBuq_dF<}Ae+pOPvOeGZ5o9vF3TFj}hy}!4>ER+3sjbP!gD}Mf1J>ff&hB9jo
zZ&#_z+SOaTrp-4_Q;n9@$2@xeM8fOR2<`2|T5>`2nm5P;f8Al^pf<LcGkSO^^$2ih
zo|a~4j?<Mk%T8vy+U=YfjS<J!mU=g{iW!$M=KXxsB)k0m>;Cx`uHg+{?7!4X>~#O1
z5$ylP0<qFD{=cOX-v3|zfa-Y6CLMJ53FRZuAil5ylpJtcQ6gY>^H}w1`RSB7Z1xXt
zQiE}6N;Zr*J>RKt$yn(LX$Jqx`JA`9^>gJYc4TB+Qi1CSn&|VND>1V`er=+FzQbqR
z>rql6+$k25fH5o7D~@&bRXC|}t}(8$9K>qeg1G23%^O*h^4{mpr@%Py6pTy`!dl<!
z-kQ1q!LfX;6h+UMF5NHnW;taHBIB0}Yh@g8B)j_<uRl-y-9Ik0mA2Mv^cWji-+8Ay
zy4sBQ-3`g}YXk5~q*e%+!TSLudQe*H3A14qlquQkD^*olB$ZxU#2Qg#W?hz8^3j6z
zVedw!wkqsE<ZPjWV7?3nx1(eDe?WFcp4s}DJw*NC%b#d&lBj!kb#i%y;dAVTo3eN_
zzZEi$X1U=EEEE^>Fj1FvG!rVV;k@3qdc&`KX3la_-J56_hNZ#uq^b}C8lV*RM(FPz
zmL!3t9Q=XAq+aw)#W<tgqO$qo=x6SA$7R-svn!@XeSlC__nxWt)@SV|32bJ^Th5_b
zHI%_mZ4KuvL(1L+K)vDlGVZ4HO=~aLY+fdOP2*qy95epWJS{cA)9#y&^PQ%qQ#Ghm
z*0fc(Xd>6D%1yi=-!Sd7-?6;$%5}lma8vJ;zjzOOA_Nw?VS)ZXhpyey-BQ!INO$-x
ziSj~eX>eSHKs|Y-@q4c))j>F;g}fZ%573-ADhs-<C&&5dbu)P>jnb=q!^<}u&7R?{
zM*Nn>O8u)JtQXpG#-EvlYF#hmGs0L==vrt9n-OAB>ME*Sgk_yMhZ~aLsg)zrmaJ7&
zJmOxdhXV-Q7NF)mWbyD@cZv&dI`JCw4=t#1n#Gdp7qOLk6}FzUm((C43t^sJpYKEO
z3azW^-PxQXbE7^#DnQwjBQRiL9D;YnDYD-M>7!UoKP>OifH?}J{_DfQ^dF5+de&cX
z*8kG@T-KPh*<wN7KA|*W7tv2iO_d$R&@=gqcFJ&Dz_%&fXciybY4hC^N>GmfqyhwX
zDiBY7xAh{q8XwvVdw;vgnf~kebmis2^2I|P;V^}x^~C|VCHf{l3u-;qD@LTcJAb;p
z_zj1YD5q^}N(!-O|GczkE%jW%Aq6KL%kuGkV@GNgaIfA*wl~b0lX)4ECa5AzjBT5V
zr=tm`WWeY8Mb`716(1I@aj@k>ZWAHW?1|@I+b70nOB!*e%(m5t<2p`Lsp0ZX$0;UK
zWg%DzTBpm%NhCD|3tPEG5zO9`lkO+F?FhGXWQt&Y-0!S?DwzhhMTK|Ge38o_Cb5n@
zJqo5wBjd19V8|{t)J%})Q^PnXpP(e6rVPLvKdTirlt8eLlr401Mb;EAcH7fGziLsT
z?V@UX(#%G4OYp6eSA31n?2`(|4&(NCN_fp<+(6;Fmsf3B#OEYTcf9m05(jepQo5%a
z=9WrDPj;y!PfkjU&YD2uz<MQ5tIE~zN$0r1a5jj`@|PKv6%!g_DqwMCol;!Z4Y-#_
z4E!06#wrVQ?30NIZF9$jVN=P+IdWP~OF}RfvqR^`McR3YvDtgijJlq;q@yp1uBaIk
zlw>7<HN+=#dMnNkm>r{EN+&M5drP`)#mtaQx@~Rlh0mD;gXOoDt9?R|GDj(7I5}N4
z;(e<Qi=-XUmI|!6Q!S&zhr2y%8PIfxHOq6)4wa0hZEc*Xha;ur<6oKO<;T^{OB{`e
zELOE-a~zxr15uHx8am2<b|3-=LjB-Zo>H#ZoYm9kQ{Ys9M|Yac2mv4b`GpajPY-xq
zF$_rY`3d05g3F-oO)w3DRH^qB^$XxMfMvH<tjOiZsPXv8(k*80ZUE&KL*#TI7V_zb
z?qWNMuTX#&=qL-hIdcs@UKNz?CgB5UvhW<#>&WEtRO$>2PfGBp#M9{JjzWcLr2O`j
zf!?~xppdo(VBytf6e}m4Xqo-gU<b`2yGyVw4Z3Y9|8SiVr?I0>d0h-ZAT$!LM!ADo
z`g@s^{hghJ4M~lPkb=lC0rOSAOF>14*QBxBQ5YDGD#N$j0Sp4w;-NhsK-V@Dld51)
z$5gA2v*{Ds-ljZ>SvA&)3X;>0-Dbk4hi;wfq1uoZ<grwOdW$<OG&aFJ($~8U?euJ5
z+17fya@J&OrDu;YVT?K6npMNfUw=_<&9N}4@Zq#>S?>9%zVeGdKdR=`V_a6uQGTq>
z+;-Yn1yKD+MMMlnMufs}+U^B!%GSBoogo;qPHkq(BKT(+-a)slF(%1>*wH{G$y(5P
zM1xVDnKJni8XEed`<vAVveYHh5wUz>Fhdbku(XtR@QO9nwlGHJBw;*k#;SLCvQaGZ
zXA)29S69_rMW@g2Z~qeTgxCL-r6zxadFQV(QHv%5`>0DERsUHqP`8wZ;<Y^-DvFHR
zZM>l{TcBTLe$WtG)ebq0UgA|6xUKLf9k9pLd(yDjo{Z0Jgon`@hp{!W2G!Cq(N@tx
z3H7RtiX$wFs~-NKrFebuARFyYUOFFl|2P0T7dcEKQbNjtE>8Lcon(!zlUYi|F^ael
zjUYXrFf~ejrv5ib{4JIuSfzt_oPY@4U<{o<q{T=bc}hRZ5|7dxH54NFrfvjyvQa6E
z#LCYo4WcMxs=;BvBA?lIf+5J!&U&cI7{ib+@C)k~eS<L)oCgLw8#vOUgf8?*hN6`V
zS(#xGr8Mk=#T@*qwJ|d>rpi>^4lUX<KFh&khEKyVh=#A>EDouIia>8#Fo_{vZ!f0=
zyAh-^l*~_W0}~Qw9as=Qps`BOtFKUXCN=Dzn~kh&1*<Gw<r?_pLm{o!qrt5!BZ))m
zhBtJsGq0O|J&p(o5&*n7Ko)4yGDZK5zy?HZqiEULn!a5d%|d3}YclC$&85FnRRAXW
zl71fp0H>SBh+4nM4Xitm)`84p5T$?>$Xj3lGn~qbWJ_N-km`chYA|BiJf!t*gieKT
zs$ygR1~~6pz11@bHu~dsWOATe7Uw~h9`WMSaF~o`F15pPsou^UCr#N3T5o+6f&d7=
z(dbYKQ*VD1QbCTq$xX?cM74C!jnyhrV~y1c+F*UK!q;NGHD;!tyvrB`7ht{QK5#h_
zjgEZ(__oVp!yFuq#X4{)Fmtuu#;o<b5Ob_0S`%o!$v#uX*#2OJZoS!lQ@^#r-ZIU4
zvyD0K_r##DFpf)jB%<?PIoXB#+seKtZkf@Wu5X!(ShiaeKb2*J>ICm}6{^8OH0Ii}
zC!mjaN87!l_t&xi?ORCC{2$*!M!NseZd|Q3S+m80xcx!tfiXxoP0{n!=I4FD*+&IK
z3n*n04^0=S?zVEbJ4-IEl-PlYD}$PCS>~*CSZ;8q6provF)O|M{k*@pie{S=)eVAz
z4btQN^%8v>JmM>UP~h6iSJaiixoX?$+qIg(&69guuYnaP+t*9@-4VJPLU4_@#Se??
z;f75GF*5!IcgXn7$b$<K2zNl}Hqh*6DwG+;1;R2mu^Z~^^YsDM4f8sU)UK&8-q^Zq
zR;c`UdwMf3>V!)pvQ$m2j+a)}{o!zWx)K@MhI$^LY>W@-djsTa@g<d#s3=pJJfZw7
zyod@)2nbOXyxU46T*iNFVyv^(cDH9Tg{~M$vP!&P!}?5f5mB_TFIbWypav6u_VmC{
zQyOrna?B_AbD=&z8EGa;Q707Tes@Z}3p1q}0M}*W)k$l-JdL9E`_a=;4P0D!$B^V!
zc8%o6*`KR(Sc8jenUEJZbY4e&@`<mf6)eEnknK+Z`K#L=X_f|l-Hm|tk+P>A6)Y#|
z=@!ID+`{<s!2;GH7YbXzs1yec%~!eB<we4BEFAH2gT$J@MMFfuwAhJi5R_zwnFybs
z_2@E^&wDkJi3ah2U3fBFdCoufhrd7u_j*Ulq;qu4ps}}pCv;5{3UaWiTv)dlp(^ST
z6MFS}b`uJi>c$Lo3I54(GaQUT2>8lM4?k-;yo50XA|eRstXVWoU#JReA|a@ll40o{
zc-g@wYW300ot#k9e#(w^FQKWm>_PLza$t{*#kXy=!8=mgQW76bjEbRV&g`vMUL7k^
zl$2dFkMx9J<@BN4!1k93{m+IfhqY^En}SYJ*i?NuD#6c$G_7iUYxx)WEnlX=;NE*S
z54xOl)zW&|7KfXuwe`4Fz5?G5)kt6(?wYZCRIoz*X(#6Piz+jQ{y}Y)l)s^E>Q0kL
zJYo20^Qh-z`|L8vDX7*;J_m=HS25VZP3Jf-o%A@rYB|K4`|3^j%|tCTQhg$-z^mH7
zH#tGHh-jj~e+EW1j}plj2a8bxzt9KD8tz3oZ8Z<u?EIQsr9^@dhcWLx9TDh*kj3Yc
z;&kyXH$%Jxcp#b2b*Ibr_YksY_YLk6b`1+3JG>GhZ$`IsGhfhb9XXLOq-NhSZZO42
zCAki?!@Q3Sv+io_*EC|FT^{4$DClS1R)L}Tjbe6|JFLMqP^+>$Uk2nB<_^Ge*ctU?
z7g35T!lEGWi`&064^bz<Re-A}-9MBWFR54<kqt-d7G)8IuD?=zi2gM#1~HEvuKAM!
zj!V!RZA;;)m&wgCA&IT94+}B+8`0;TZb{*-^OceAH!|sOh7^o3aVC;bv0yKJrWS=S
zT0>+K)vqsm@CM`MQ-Fs+uHSKmgNJZAkB#C-{%bi2moFyK)in3afxh=I!CZLe!taR@
z*@z)EO>WE&abg4)AG-jm$mRz}NMT!!a{-NqoEioK(9M{LHBiC_2WpthAsveJL6C5E
z)qDTi0BO)k2os0HKQ!83$<qQ{*bYRbFW(;hJK+HH&&o}0*b1s|&gR5F)pztmnIYGK
z<D<V6$^h3mj%Lw^*^0aLBXKxoooCYwRkFlHpMOV`Tvf1L4sODx9qRV{t&`@{t*5FU
zHP5QjiND>lbuYIC!}OTks<64d`xCp`qpNO?92df~6OMrlS*d@i;)h@$?pk60fU5;`
zpmAM6K&ShKdNI&G-LrMJ`f+t4aJMZC3yxhP_a1vvS##(lLQ{QWb~%`JkCvkLOUvz(
zepFlJKJig`Ri|7{jpbity@Wpgm!c0<cUmEZXy|sX-XhP>)0@W+$*g&7sIl@&RS<&$
z8hzY67FJSc)d{!8{0l5IQXl-4NrUKQ*&Ct-Jf(v)Zp+t$&<Hs4=>}xIsibQJA2&k0
z9_3i7i|2|&S2zl}VJpe#&&CEqD+jQhxR9(JVYDdWmjB@C#rAyxToOKh1y@Z@N%R~_
zIgYAm0qIl*tl}9N<rUppV6w~XLcxYf7=5|~hs)yz$UFX;43yJH6`v{DI@#jA#{AdO
zsL-y0mlThR?3&_!@?c&VykP=HG;f!@8D|yKy=W;o|K&2$MQVeBqM49(@OgVwiq74c
z6$eeFIIna1#mi~~P?aL`U=$G5H&M-l@F+lWJUv)am`{8hwp;-ywjE1m3=kxxD(vk6
zES*}{$`qR)KHpV}_jDSN1mRZ_F)SGZ?#LP~e&A)Ubn&2%{!zr2-MT{4M%fj}&hxSJ
zeB?NGQ<)j}i<-H}mT@eHVZZkE_eE>4%FKJa5@x<qMe_0b<D(6VunCTN{wNyA@r$iY
z4izQ@@#Dt_+0{vXdiV%UXXy(m3U%zdJ5Qsfc77%AP0M7*KF>qtjBPN!_qkUhyt9Wd
z<@^;~t;9{heCPYFWpU?%GstReFzqTQMX?6OtTC|crXnv+n}(?4t(dIhc@eZqW0x!4
zPHKk2PCxMGO40@(itUKUFn)B`iz#;UGN65gqi6NkPHs3#shp<!D!t$sR#_PZ<JMVQ
zqeJ-d6drf6Q<^mCfq>FrIm$t|>wO9xgVcz-@j;x+G<YO}8Dmi}Dt6Afy|=Aq{7NFD
zq$LMvO|@2Yk-T!o;H4+Ed61ceVh~6p@+eY}v5>f$XeY3i6O%A{MJzbID#A$iHfl~0
zlVI9%koqFDzJg*9F-#O<3!%ioM!ao(jC|%Sya@yv`ASf_ORdIxK}HEU<9)?S7Lp4+
zU6ld*ts;z7RRKrQhN3rOuBy^lW?9YZ0qM@CnD8(^NX#;lhbc1j!p6NWq#^SIWSncN
zA-gCn#+!#&HnGvyJ>Z6&hE6Tq(ci*4+!YlMb})GCk~T;MvQqLEhQk>&S1eor=TB(e
z_!J;C6y5Bkz>GdR*}LT5$XC07T5kWShi1WRFBp?7`<~}0#rx$+HkEkJA=Od>2GmXI
z=-28*yIb^iwn=*!k}MA6ERw7xYEf4jet7we{RBGAme8oZ46PKEi0_GI_mt!rM5Zxm
zfcE|412$F%Cn{Vj%!6MMz^L(hehHI;6VJv{QXanr<QNKbF3p7*EV}jn%tfHqiVuax
z^)#w4mB!~kTARCEG)Lc_VO&#<PdNfRqDh|fDJ%At=GW9l!S(<Qi#ye@dQ@K!49mMl
zek0?0rj>o#aCED?#&|*t{gPqYA?(flD<5^{n*X~<HLf1DS4hrIk^L6lsub$$u3>mq
z>{|XO9o6?Os2-FwQC;Rqrs<WEP!RhVl74HyR?d)nW^vs~OC|Ap%D9GuW!>?QK3oLN
zFCzO;t(ldP`g>w54v1pNF0xl3ntjiS#TqP0rRJ~RC^mdN^Y{!KpATe0@r+tzF9XBU
zel5Hq^*1EL*!Q{;$0eFd&6;cKLXei+Sx;+9^tra;*7BIgD<1BZ-Sy79?O99*OsB5*
zUqt|Pv%1Orx+D|dNMc@B$Tp!dii|Ba*%f_BG;kv-m@ibx_Yr?9>*CEs;U-7x;?;!Z
zN%{bC%1Z>9(D5{v9{+}x4W%HsEWKbeJW=L~#nho3b9O4z7cn>&B^K~8$E=YKimu0=
zO8=)03jTI7Awk#@HidEwQxvca;%Ok`DP(gV3eq>rfF@{i@r!1a%NDB<aVQ7A?&B)$
zf%MZ<!<@t{e5^v5?0-@-+>@PA5Hl`f<Laf|1^57HpdsvuMYw-gi<~zo-8z`TW}O}x
zEszrT*$q{{P;_SIm4UDpXG6KXJFwmoLux+hXtxm^yd!v*W{s1?w0QD1XM{P1wZJun
z2yi5MDT&yPqp&7;L<*WQmtU?sDJ0ZreSBXwEKeo4!$Xr<IKITK-sTseRdsx<tV-H8
z@Q78riwHv_7opYVq8$EK5>>8F3q$&)W#UQt)Qgx0RB^jTB_`t&s~fJHWL0HZNi@8-
zlMHh-@HlliKxjr+!f!{>*v1y^)9`0M(>^=7!>Gv^Q*kc89Wk*7`vf{3@GwBHV_X-8
zedf>Hh<KY9?7nWV(efKMA%{_F?9)--7m{b_4@ZhT32N^~rwj=7o^;kL=#7ZA1J$s#
z_mYnNh=zZ@JglPqfW@31{x^pYJJWx5)EHS<|2HXNvzBHoaWjHXuI_v+XT*qQr&sW{
zya|{IaZ)JNdZ<_*kQFe$fgB#3mXCMOc}u4?u4^Y4iCF0IqkX&9OYN_z?A<0|akG;3
zrdEe1F1-s}#hv@N1FZ1MzrQQ7U)-4Z8!V-;_%V7GsrpORoW=5rN)mc?`*iJF+}VkP
zM9-%aQ1@E8ie`;4r;6=Wrdjljn8jY)E4<V92zJGB1V#!s1%Ly=IsbmHe96@wZ`zaH
z)oq4y$`Ary3>~e^C5iInoDQrl%0s_=oE)m<53HYY=NczNDb1xnJo4@Nc&^3mQ2_-0
zF@|Kv$ghsj_t5ImfNCf+f5-QR6QHymED}mOoH{I$)6nN5(bDxwtBHCRh=&>8#C^3!
zCi$J>8z=+>jP@n--%-^1n`-Tv`0ystK2)%vHcDMEykgBJn7vV+GNSk%$-Z&Jfh_ly
zR@#<Ux>oWK{=?<U{^c3S5J%nHWeF6&p|~S>GtpEMdYE6$Y%0t>6{FIy0S%wwaeW4r
zByUn!`6z;C>FU>T3Hl`GUVUJLFg4InsBbMmV)35@xn3qwm^;s3KJVh<Ag#ONd1!K_
zn%nH54XCkh^YsF*^eH(u4X}F5XNZ*cldGLwr%>g9`Wqowq4JwtFFx47R!y~{={;qs
z^>vo3?Gl_!Q=E1#cJPbTuVl%|=qQPj8wKY}Q*yFXN$PPC5Y@0#(OA@HEIW7>1|Hjv
z8G)*g0{a-cZ2_IMeNML(iTMy1j)Q$79)S>>*1At|Q@YEyIJ=`Bo9}a62b6|SUQ&|g
zfRr%89esj2u|zY#U2NGiK1rEMY&f7u@%H9F8in3oR5*D6ad_(1Y*l<En|4#(tNiNo
zS+uR|O%kb(G5t$TBB?BORN8CCUA5ZkUGAfeuF^Tofn$|Ti?|;AytE3B-_ml)nKR2i
z3qpnTS-h!HY*W%!Tf4<OBz7UgTX#(-y%MUf7Ro-1hOG~SRZb>4RFSYyE{85r{UqnU
z*_H<tjJww*EM>R>klaMM^YjXl4zDnczWL%wnj)Ag>uU1oCH+ct6Y))tj@C_wr_a#=
zFfLBs8;(Z?$VlFSuFq68{wj8#*Ze8msw66&WQuI0@JP(2oUaAlJ~iO(Os3q8bgl^(
zL(#X_t?@3N?zL&Mm-O^}sHHhERX0%;M1l7I2Rs<Sm(sTHy5sYW>H=P23I>_NKOT(j
zVAnY<+CtZN<Ic<=p02Q(MRpUJactXY(M&0<E2VurMC@9{-W*lTP1+E|VCvlIw&ZuB
zFneAy1Ce*c9EFq6xlN6q4D;X{CofD*uprPFaFRqX;Q7Z5226s#`*4nz!j$2MT;BPX
zkGeuzOri>ngr0w_t(osdlAX}vAhb+kot>7Q8=#G2y01zuq<#P>DXFoytyhMKiJcA8
zpoE~l4qgRjDwwHD&db+3?Q>2J7*`X)K^P}oZn?O)gnJoHCE4-r4y|LW7O@z6RUEIK
zp(LBwbekxKM)LicB~mN(5T<IWH>uqtbC~IRtO*);T#mD{<FL7hobgM$``KR~(XU<@
zeT_BL$3zQRht<c$_0$wpU1pM$Jp+PhDcD`o=;Q~uVbVX0s%orS=Mv2@GpIa8pz@Gq
z^-{$ZBvuQ@G@=?Bjh*=4A=5xoOr67h<oZg1=2rvi;*s=L)@I0gs*y#ZfpmMBduf;e
zl1#Y@lKQNre$?>!HJ11`z7_!(Y=_C8;p1^!)UK2z$KKbs;M-Ycx+n(59P6QCwbIhz
z;S}j|h0njm95fTTy*EWo2!G@>@??K6M|i1+Mw^|!br9q%0$w?G`xNoyO6Gi4JcEaa
zNT!+C{#P*3Z0G?)Sf_Gr4)#Hu;DsSgm7E$qvpWcI`rr#saGj?7Kp8Gz9oo!?PqC(F
zUc!TMsv^q`!*aOsl6HPI>&0e+`}71uCUsQh>$QkbJdsOFW<wMY9t8_7+jOo3`AtN8
za?<@Y1F3ujY)#8cX$Q9<p@wQ_>rtY?$&`oksk3i4Gp%Dwiac7{Vg8N`p9?dHhB`x~
zrDx|3?F!29q?LVB(Hlxfua{g)Bp<h_YBUv8HbV7$-+H9+=ixJaln2ZcV>9C603f2z
zblbp_I9JF>h7c)YQ4)^$9N>*b!Ym^iBL+fX!3M}cLSCL)98uqc;qks_#@q@aE+6V4
z$RzAAkp;=y<JYpRYU)7|-cEvV*ZV#t`xg|i56d(u>>};M+^DfWl5O43!TENBFT-AS
zQ5gM-#F-$+6&?mIB67&4(r$AKC}ffqb@ef-6^Xm=)UYP_x}AIK6MBc1A7Hc5GABS^
zN8FepO<l@E3GHOe2!Tep+{aa9?_URe2*gFZw|6;1U{By>RW;prDSw&JPmJ|=fFS5^
z@&+U>CXolCbqJ!CfA`b%M99G-b2Zqmh_$X}j9WsdchepG4NH^6gw@WDvqE@H#9w)m
z_<n<B%w$wb`{y*qvj56dkts<Cy&Xr-C?n!K`0*`qcK<>d#Bl^{BNSzJKxAWJc1DEv
z^vv(TwZNnQA-B#cVB%4Ds8Q*E36)?Ja#MdBh+LgKU!teXRBc7@aq=$A(!}BM_53`2
zhlb><2jJ6`=j&xz<@T;v#Q1q!p2ea^t(trb^<>pb)}u$;!#R(NM_KUdvqDLwE!Lqd
zS+6{2Vx%bJ;nmUnL-Fhd82Ol;^2}CN&(C(K2naqmk+l&KA@XXbk%<icS|>H;r*vjf
z9@7nf0mx=8Cexn3BL6x^_I;3pTVjh#8#^ipVm2CjRQ-9Me-e5zpQ|ei>jIbfDEJGh
z$XqeAPFDG3yAF0;1`GyulGO@qlJ>c8HN-OO;a-UIf*9kP`?X`1&Ly0$8u{(UU=XB&
zo^~6Tp?3E6&}2G9lkhuE6NKoT|9SrHxP$JTng)2A%P1Ku9Xr;|LK0%FgVxPb;uF?#
z{ejwatb@brUfu(LR?nCI1JT#<bfSkVAQjw0<yjZ&TB0S*5a6wIs~s0%1pMLTa33@B
zY^O^iA5i-*OTb5-OhFbQS1yD&m^54#FCHcp;GR*U56DHTJy5tStR|X`Jd|z7A>t^I
z**qfysl!W--?R(N>7Q}{%@5bHfqV%%Sf4=2_Sla6A5;@Jed6(p1CM@ZB&21E*DE(;
zoPV(`T?ZzcaerDNn>DdhyWXOhD@aQ;1*j4%so45{iB+j)1U`QVi;*L0-LqZY2ZeXf
zO53^s22wYkwS7*SfEecf5@V~opoRb*CEX~c8{%Q>XNzPx@}aVcl<{qr-!|a89H${D
zsGn70=zRK+Web$lhs8t~oeGPYp6-^FaL}wk_TLnA>lMW~8-VNB=O?02lwT3Kb4#C@
zA|EYE#RoN#ga7dOO$2of4})h*EE(@~pbv-|%x8&nKo-bFF^}V&m5sr7$?w=NS%l|L
z&<?YP=q#*<@P)aN)G#9w4`X-UQ~{-l&cd1mLETBw7X2IHA6ZfUwc)`FY*c2PUl>ve
zis2nzG-KO(n>k__pfvm}rfq9?)h`p*-|Nh4+cR9YJ|qRZboU*{#Pj&kKn_5ckmmrt
zlTDn0QAW_ufbrlX_eY!)&4_|suLWa<cxTO$p~F<MWUcM<%;=8B=b3J7Pb($>AUBrk
zeS5v*U{(JGQ<@`?&|$NEnkP?-{hQ}CX#D2^pzmaJ6}U_9l01tN;Gt<fz8H>_xZ=<%
zXrjRZx{RD9$_Lk`IAdjUR4*JrbMRzM8Kd^z<*B$5hk)w-Vg3b4O2STa*TQB_Np0A*
zlQM!|7QpM0XqhN52hiJcPwnbmdKZFAp@P#cpl9Y77#ggZ$VofPH4{>3<}wczAm<~F
zU$H;*@62qc5W||HGzjX5AZ3EQqW}9$rh~TZCVxYM(B8iH8RHTB!7+vC_ttZ5g1`uN
z;7yfa`dd~l!)sXNNpDwKd)CkhJb5-W%Zmr6X?ah3w%nD+)%CzVf^Yxv@a$?~;yi|m
z1b3%l_T#(W%!7DbMcQ99Es^y8Wn{ys8PUQ+ih{%^ROAE%i)kC`R<|GoseMOane60l
zhW+suAr+&B*4|Oh8zx<=A}2U7ugHSlNW5&|(<kPm*+D_Yop#Y$%SQ}G3tPV&?P<e5
zS%80eScnV{ft}8=)Mqdf+F(LaG<+)N|6ThR$5fE;fWic+sI%)@+AA3p+AT@2meupQ
zj4`2%=f%;ZC&8p1?D$c&A~g|8_Qt8_VO(fB$3kMTmH?>p&!`5?F44DD$Rn&n#=BNT
z=-Yryll2^6ER|D)bdC55t13iApxcTuFTk;k5E!Ybr7oJ08Mqok3yw9^W2nxR2kwHa
ze2V%X34jVi&n26dM}2*jM-(*ywmdJ_A7t)}9J@TWrGC0Bv>qwyRvhC7<vNfJmz6^P
z!Z{WHm1?DFUPVmWzuLbo^pv=9*9euppp5HjQrxq64jOCS>$!stqb!%LkdrNF(Q)jg
zEFkWf30Qva)0FIP_1V0%-%r1U9`n4qEUUHpvZ(AxTz;mfle2()u=X<T%KITd!)<-+
z%d89(YH@j9Me5_JD9QibeK+`H*;GjuL+foG_z4DgdwxH&Dgz-*Cr-3308ywAoiudT
zR)nr<F{(+fgpeJiWhPBjP@!Kc<DU(Fh$to4sgliY1|PIwzn?-@x;#T8j&nTNxa~iw
zNbr!=1vr}OE4HCg9T?cnCVPFb4rS-HRq6eE^IYJuCMv>f^|Cs@L-YsCmy-=eRkeJa
zuY!S!28{_<xJ0l!vuZz`j!qB$vhMC|+_iR#Xh1-SLh1LE+Jhds@lb&UgFyejU}e%|
zv*kdG;Z8i9*-d{E4Av?f6}3e-!H;f#+WMTc>i_bSk@5vN-Frn3O1oBfh1wO&wY1z1
z5b74veh<C+cE7-I{kK^c^M4A)85!vRN3`y*VEk_@LeEWYS|PBYHEQLn!zD}?C8b|{
zWEMAYowpn!5gj~+7)|`c#RE#GXUX~?$+{mbd(_eP(WoeQFNMR@5iXkJ+u-2h<Y4NJ
z8jXHLOh9+n=a;~1V88G9Uh0lE;QV1a4J|G$ua9-q(<|3M!V$Uyq}v~apN>&hDFi}r
z!m)pXv9<wm23aw_M7{gHgD0B9dxm%NT>o4F9EHTI8o8(&Zx3c^cX`u{2S2vo@Q|MC
zu4nXJOr1&jeFt`cAH=j?NM54{EdsVDJD*x%PwcMe&vme$31xmN*gmaR)(_QBW;@^R
zvD+mx5?Xay7#ax@(R$plNORq`ldt{#U>owdL!`*<oQN|>C4rl=$A+VPo;EK)2+-0M
zh+2q=q*m1V%m{2v$7m^sR8{Rw5RoR8FCQWxmW{dS;9HW%-!SfwlvZ{y>5S@_B@~ya
zty*31c;Jdbb&+z*G<I5sPri%-C&3z7rlEW+CJEBWiD>X${7;pb!)6m`@OYbu5%RN=
z8eXfixzCnsDT`>r(Mx_AFNr}BTUDM=%1ZZ%v1(!4L|u7gf<2z@ud9P!Y*thgzt9I<
zpU-uz?+{)eUwWicw>MdwcY7Y#@0;9IUhb~kL19yCHG{`uBK#W2fau{dg-up(W#uW7
z3IGq#g(;bz^Wtvsab)+_u@B{l`VMd=EHUQ44I2zYO{(+x@#muDlacv^iAT-;Uh_J_
zf5vBe{6CAmQN@v1T<uaS&gvf$^93h4<)UQ85MLAf3XJBbMo|zeFWgEW_|orCL%+Aw
zhXomLAqQ~Tlkt!E{;Z`522kSd%4yEb%Z#ZOgZo(CK!00!FEiK*`c2xS>06AKw#sSm
zitUe&`wzgeB)|)hr$E})5`c0XY}@y0;91-3CGfwuD*zR-iN{tqhry=@we-cXCLBS$
zz3dxg4P<ypdAdf87j9qKi0`f%eZP{Fdd|T{Wh!=Sk_|eF159;N=D)V!A;~hvr`=6E
zYNp{TTs*H`-Rf0mGDTY<>@qEzA;@(NRhYpw;J$_gW_JcP$vpFn4{=cNBM^r{{<2u1
zO6@B7n}N!lGcY5&e_E+=Q%N!<1m7Wys=Fgh53dKTYJI%xqGItR^?RY2uVw=hL*QSY
zUz*e~sAN7z3}D;hm=i_Sobr;!UmLZ+izdl1`;~`a?eW<%fZizZuxWzA;)Nu{LZRqQ
zMg9RtA?{PQ%V}GppT`ZhB)C0|wruc4u3zGLKiElxT7|zqcCOlvebi*T30b(d;&uBN
z3bwgl1FRpAqwfTE8qE&u+KXInMB2u68)r=r1<wjSr?BU{AzgxRHj>N82M#>PS>u`x
zDoR}!q2{}~=|Wgg#)kL1SqZyOOGU=v^gChi=hq8rEjiF&M>ucbXY|j><D3YOjrPM&
zS6K}umobR!lC-S$t;AE7Y415gB8QD-so+5>;mVTF`AcYMKJi$3R&mH3=YzBjR>%Ws
zNH>c)H5~WBR)UsFNrm$G3?0L(`~^PU>+}gz9%M8ONQslPZ@^c9NjXyo;+<Mt$5)B=
z4!^d9I;*pw&!E!Mu6@deh+vJ#-#2JQ07uBGRtpMUqq(nm1#-r<0&@1`yOJyWNTm<3
z;#2(l)7w<a*3b3{a*~UfeI;X_Fd8ppO+TjQIw3PdHUxwi+M!rx-(BjEtm6uKUHhik
z2O)@lEejCb2Y{T`vc75v_C*eeB(C`eb9nYAP|(ZYCZ2K8K1XNNU&5;-Wr~B)$b3Lf
zR-{w2;$RwplZ%UhQf(u=NVKUyP+yK_@t<nmLj0glx=Tn4nS4_`A>Q4{y<7wf=tA@`
zM>ceH9qRHx9XJT)^juk>GzEVru=j5h3!x}ApZ`6ZH=j#HZ*k?D{D@wxm+?F39$lS4
zOxT@pUA=NziJ~YmEB+0B(vMFDk2NW)n_nC&=JOrTx?lpxAJGlk6K7(2!#ac=v?sVl
z#sZYJpM9@?qtaEd9Ah&;F(uAsaFXyh>!DaQt3QCjak|+?P%hQXD;QvM1M%{yli)e0
zVuD9(8WdyXCTZK!mHqblzRt203dNByJ@Pp@g#pby(KxYcaEoW+Dz7>Ca8-!McZ*S>
zdFTjM3d!Au=MAI2z!9^Qy<Jib;j6iZ(B0gKT>jOCae?ZpmTQ7*+VNb5ZTFCSCJC4~
z^JEmkf>1~?$<k2Z=HY&4)dW*pJX#B47CzqoLMk0c+f;o<>3b!5%y^{K{cyuaU3AK$
z?aCPgVWt|sh18?d%(ioa0;L<vdy~i<7%b%Bc*a@Gq2Uh5nBO=;Bo3n!xG-;6m;(Q5
z1TH_L|3*Tr^G?vCFnGP`YIY?hI=2u<ckBzDx^-)VvVy+!f^Df`?bP(2%taPBm3b6L
zIh7XG<4XHX1J4l@wup$B;fp#;*{e%s6X)rMPJ;+BZfU&<mC{jsUIi)pzNwy5J=2Dc
ziRfPu1Qh`uCk+}se<ie4gxL%P9Fm@$_D3<~F$Rknq)=7N<b*5^E}(svN+&m%2-Zn0
z#RLAlkWJtV{`|iC&lX_f3__%N!`e|A3dm6Co!cJcq&~yH-Xu4g$3|I`upPOmF%8_-
zvmbmlYR-5F@!}(0a5?=)1UH+JX61!UDw@7NLAh(_59yX*>^&ZEo!U=caK<{2AO$BL
zLFR2FRUi0zIp$Hf2;jJ-NT$XY4<HZWjSTQ5T7{0G6m*)O6*Ww*OJ~TsGzKR-KaM97
znIsJkBYZ;yiSG))y-dL*9%uWKiSH&m+NGu^E8E+Cy3;TCZ@G2fkM@^6YPyE#o}Sgt
zqmK@RmV+_P%qi>p8*3GjxpKeW@U@v;1TOX?!Z~#Er5LjGxPOJ1uP({CQQZ#2?E8*c
zXd8JxW|5Q@5R#tu#*mhb!8SYfrE!!C2K_DmYJ+5jRuC8s@ZBZQoO;^=8>%H?82#JX
zuPzTXoeAWMH+4Iky5sNKqRTy+Wgjib1#c%@p~#QD+C)voJiUr@ScbUnrzs0*<3-p-
zyIg@a3VDdkZj#~<Ok>;BYFi3B@7-h0^d=+ZD>x@V7i>cjeYvDRwT>Ru^ow%eXfyaF
z$da6_k{MZ&YwS_B#BbPudt1mah?jG=LeYE=2O<&nVSbjHbStqAOXp8|)tqB!*f{`U
zo9pwTGJ?q_ShWzBdykYtxTYKVgQaiFW8sS@3IlWLi<Uqtwf>M|p?-s&kC<gOVy;6f
zORg@ks~1!OzUcN3o%8*cyuUGV!y#^-zN=@{1Eb*iD;%HmBl$Z75^+R5L607hV$%VF
z$ONZMXiwLpDGoL#wAOnbJP}2kXrcW<HbR}zA6^Jo94Br<i^sN64kQKi=SXP#>0u)X
z_KV`&1=!5KPsDuMyrR+CV3Dw<UK*U8<L!(*)Oil(xdFURHfhWY4$^wAf-DGCqI1oG
zsWRCq)NGn;&KhzgJm7j@p~CU|(VPcr0m$jsKx!udTM{qS7v5)2&zq>xdzOtKBRf_;
zyc$~`w1!jF$uh^cIb76UGuv(2^9IjXE}_zI8BU!S030J0rAh@HU4i&g(Kl7LT&Zfe
zA2glyElH=9*e9H7Q`2w)6_MC4Jj89B{7yRD7S2Sj612J7CIr)12G-9KNt=ZGoj0GS
zALLJzgR8YVUe-BLlp?r@PGl5%lbTMV@=3W|eFRaX^F>M9{%4VsvMrO`*SNTo%SO}3
zBPZ?;51X9}BQ`jytG2L@6X@j_&K06N5u~=T4zl0?sgEcicVpB3E?!LhsBzEJ{yiCr
zZ^*wp0Gbox(QzDu?*=EU*VxSc+ARsE$cF8NmM(GM#%!Zo!enx}A%7Kb%Ej@z^zJ2(
z0Y;5Nm-jS2xLwfGTJe_hg`D4+BWvS&7M@l~r~b0U?g+l!lA&Epmkr#4yV0I<$5&mt
zAG%i+er9<=eY%n!7jFrGT-<)N&VqrY_VruW3zfQnu|Kp3Jvs&>%GO0un@y?Rf1tUw
z;0q<wArB?`EE@Foet8}SZivB<pswwxmOK{6TY>C&3IW;%-k~L&zLx4(DfpjqV*><8
zAb6d6X;KtK)a2~%RVpxm%yQzp_EMn8SCrE{HP7s+78C6r!V5G(nO{~aXzY$iE{WO%
z?j|jaqUG~B?M9;qh0A62zHpO&IjC1*+*^-*0@6V&Aiya5+QFtSu<II4)y5uJi1>W2
z1j%<;PZj<w40-wvj6JbeLOBeK3HLVl_gEZ?)~X*MM^|GXw6WkzyyO?P)6PkEBW(v9
zXEok8J?SRx`~cY_{-OMD|28bF|Doe!ru&T@{+~4MTAGPRE%v_?IC~sZNoHi4=*+GR
zBa<rm<ViSn_gyMpHVJW+<P{WBxR~F+8)ophKi~mH3=ZoR92}fnR)D7$tI?ptSD~40
z-rg^tch?W9(yXlRgxMY5pWh^Rfqg&4eZrPDG85w#?jNsb&mS)WWRbIsJVZe&uWX;2
z2b)|xW@q-HL)Z&@gt)kW#Y=uI^My_6-;`cCX`r~laEp>!j&32Pd2i&iXXKBsp|QQ~
zNah1?p6SI4iA?`Yu%*)^GmaYRP?jB9RT2joE;>Q<sX-4CBb9P{HJqUo>n0S25eP~6
zQSQEWcZeaq8x`#~GBXCGCLY{u8boG^eiFo7)@+uGqIae4Kopy%4A4lk(qUky7ky=3
z!XCdYy#lLE5F!Bz$%{ZF7B%xi#*)Dpo52*P8CW6#e`myO8da#Ell@c7$Rf;_`B>8Z
zfj`W`X<mAe01-nuxQyXfumn#;(8_^jsZ#qLg?z$7c6b1y$^~zzn5;RO0TuAaB?Cp{
z%#Ii5p(;L*=y^uFU$#i$j@pRUB&3)Lw)n$$d>m3E>lx?1AsJ;58Q?f^8A&kB1lQ04
z@G<41ugjfa=2ryE*~}PIHl0N5IMe6ip(bs#BPj{d7#c)@6B`q-;mj-=Ae_k3kzajm
zI&YAhK*8%I1;JA<egQgfj4nVtd?p>3uJchml16ojk3A%B_uGM}&puOVAUBg8nusLi
zUj4c??+p*O*WZhAj*uFA(U$C__Z`_c<#7Q5WB>^Lz{OGo8KytWqn*85);EFS9^N0<
zi`zFpcs4)&xW0}^aSgp+%scOACG?PqUrry}+o9~}*ikUJ2T~0p%*R>z;s<9a7b7A8
zYfKQ6j9|V&x#B!AqV=fP`4~roZpdyf#OWX4ko~!LRFHRrd@0+_kIFLcj2MQN39}=G
zz=2}BWY4&xHp1X;w{r(;$Ef<GNCV`ekTYVYHB_pSftuUNM;600_4?<%v8j=IdSs8n
ztF6q>xA6rG&BR%=F{7oF>&|Zr)F}B{<LG0lx)v|PvT>8PlZ_@sOc-!k!_THwt*<Qr
z_9wl5vG2WJjRhotg;5)98X{8Wpa@AuWCHWU3?YP>*KrqIm_j0RR>Z8aH1!cz1}OJ7
z1oE=y-^O9sf2c@{F^#GH{jL?u#~I*H&un#%_lAU`+k12g7Q{&V%gN=C$VSi?#fDN$
z<PF0ht_kc~8V%cIWax$9Vv9YJN3slSAmB9vWsd&p7w#MLS(`2&l|kJkTyw=$Ir$)h
z10A7P8hx;B0&$Z8LJR<%F)|3SfvR2548@d!K?-P<m%Mzf)k88P?D-1c_uhyUl92^K
zPLur>z~;{4ev%dewj=m}J5w7b8oE)c1?$uAl9c#;w$_#jw()Q=Cnih-s?@oj^f+U$
z=tPtWxFdYvWs83%m^A7tF7|~Jdqcz;N*^jSW_!c#zkzv9Dw7vp0AfhSIN#OS#XBRb
z@2%~G5*5Ft>z3c8VBiX7^eAUYm=K2}mcbyx;uCM7Y5ssYP;iZ?Y_JU#nIzG-ETGbB
zXZ8@TO>%vzEW-AM2$o!64YL2i;yKnu83t&A44c|`Mqk-60K^xXFJC4=y1QSn5T7Fo
zuc=CjBS7DFtEF!ufcQXuj({R1dWHsg=i$;U(xMlG6DP-=2{s9s6TEmC<&ed}fImSh
z1|O|Yhfj$raFIq~NHPS+_xjtEae@%!*9c$3I(v_5_@1$%idA@6B*h4K*fnO1D%uk$
z+a#RO8ef^@O}2l6shRn`3euQ58cKF|B;g;c=uZ`pyjle1@+~xL4W-q&h2aco-*+l>
zwAu#mOg)IF#Y8KKP=ij$^HLcV^N^NFZ=l^U2B))fK`p8<ywYduK$Aq-LW+mhK-&MT
zA&0|H<0lq|NaP4|0evWdqwmexJ*-tNo-BR1G*uB7U^S4$zAoNNVl>`6WOrpnxb(=?
zy5h5+=taGio;<0TXs8w>PTVp*8^t);T#%~9GSkbI9KJOt#jf8RVv+K5h4*3Piu+Er
z8?Iy*zq3Rkzk=op&rf9(F!6sFd#4!DqHSBdY^<_v+qP}nwr$&9W!tuG+qSFjI{(9;
zo9uIw{Wj-F#><$=OdqYc-aeq{vAQ_X2Olv0h0az1qDiLbqblwY(YCrt^&agbK+O3*
zIdRtlFb3>g0U~2EgDeHOIf#`%O}Eh2c#mbO`rzh6hVL~QYk-&zP!=UM5^#GE4U8m$
zaYk8MFoD7B;9}e|Lym2ATli!&v@~cfoD!>R8-Un=X9P~eOAtG&8m$a09I>zAm9p(C
z6b(d*>j>whrf&h0<FQqa!BK-vA%}mjWDk{`BO*;(El^K4;~2XN`K=rVtKc|h#`6TJ
z08*1um3gksm0||6g9z)LuV<!Yj{?vUXxN-|U%y=y`gPGOSeGZxdXDR3J~Iv!+DYBf
zdnK7-<jyHcFw&!Cm{2wD1jPHLYn-U+0cS-_cd-0`t6<<@uCML_wfGe$UGp4=vC$vr
zOgn=gEGOSb+CLsvNuqm-_^*n6C&I}tuOtapnn%pKt3G;`j7PkdgPYVl>-qz&K^0Z%
z2vkH;?)zd53B9?$p0yFSwZBiSqOVQ-u%Nb}U71>BzZzg&r9O%}zy1mH1Y-R}87(ub
zju%~BAqpBPd5qZA6V;ude~ja`#mq^2^&7^T!pDoK*G4T792DJ|kyASDF(lgCbIZCf
zuf~c?QMnu`Q=`yK9xzzZs@y%90un%&5wHSMwx(UCtbL)uoUtC+2TdHi%Gcr2eMYA=
z|F&x3qd7)@o-U0IC%5<1n$M3ZEofjnVDKh&rP@HXFG?`DK3&&5Tm@$lGbz|s7wxtt
zWY~apW7-w`LC1!07+h>s3Lnj8{sw{)Sdqh2Z_fCVy^%_FJ9{!*nzi7COgJjO&o-p;
z-w^gmSsbARqx^g^Oz3JH;-PxZ{>3#s54&MHn1gT=v_tP!7c|~gbat9V$zr8d0V}Kq
z6>D~<wNM>P01>Z5dHj1Ai?XgeZ-&v>bjG+VNA%WC5z|>pnXYs6rRmb5#AXTp1~*Q(
zWp22CfT4>j7ym}U)GDaNNqNfoSzQX1`r+bZNF%M{zf!tWy$(Q`w|uYv^IajiZj6?9
zh8pyFacFwK5vD%D|M8oKg?x6Ywe+`B9gx1Ycjw?(f^vrwEh%N;p4DmYEw1|H=JEGy
z|I_dLCe9IP-S4sbK(vk+>^<*<FJQO$VPV@>((e&Bo8^ME6u{=h^Sr1#-j+a{*4wR<
zDC9`8(L1%0?|#{TE9qcHjC5s;ruz`<5n#BO8{GH6ktOa`$%A3m5{IzHA_MGqSWcqp
zi|n7p=XX5s;ntg~?8mwD^T+j-z+qYa#=yT3;V~#^*pH`XBBE%q;mEqD==&(;DTAV4
z0Pq8Ca+wxwf4Iv8zM4DHB*Krzh=yuLEmGxF%3al0uqU~9^O8&3Ts4x)>L9Es%m`wk
zw-1UE$<FMj=J)4D>F+0?vL4#*qZ4l1Y;trrPdsaqOMNAUYJI$>i&W0z_500VWUOl@
z&z2*JlZ1v{n(Imrh|n*4suJefYXhCFutk^zN5;lsT>7HxRdRHKqoQCv^m?^N8LPb#
zhRUo=W`93t|Mhi;+>+%_Aj^>CZj0v@bPdiLt~OWQ-@SS?cG`5;8D~a^hzQh>7EWfV
zd}{AfMvp4_w&Pe?SKAR1+ZXV^8MohyZP;MW1>d4Nf4hmL`Ruy7bQ+L(zC{sy`>pL6
zB8>UB;clS>rV;$gGF2celw3VjZGh75l!V#9={>eG#lw#pG^YuVx>tmOAnaMj6|4vW
z^`61kEs)VH5wUwg31Uf(y|7y^`l%>O;XoEqgn%5Q$hSn1DNwMZ3YGJ8@zr%~>Hbv7
z`oMws^(?oNw7tmhH6xw!H6<lc39eZI?e%Oi+}=`BqHs}7kZt|iwmpLQgK|9PcXgs$
zF0WYuyCMCNXOfv17B6Ol#I?~vdh}_}<}&il5=<h0c&=dA*zuUTNKF_Sp5dO{T3DYH
zcyY^-elp)!ACFur{XV@9%|oeA`b$U=M{vcwIkl~n2KIoYAakFg?J9AxJ#2#G?~^Lp
z88tEBwpe~GVnqhbvL9i7eR(~M)4DfZbjWQ9>ogpJ)him=0MVNY@4P~$V|zrMsPiP7
zx7ewGhc67&^S*J_gTdzCr-)(TXp|&taYZ=Z<HzFvLi(0@r1ka>HKaL+mG#39?hB!j
znHHsnizN?bo7({{=&j?6`$0b&QCTU<_h69QjDS2>hwToht?oaRDNsS`u=_!O?UykP
zQ;c9X-oo^LN{-W5;3;u^q$VRzrv<B?p7Bbp=LGK6by^7&$#eYkAuHg%Rx#5as;cG;
z!Ia^w{IMu0jJ@29OQk@~e~FSqto*>aF>#LC#eDyI0JHghcy^MS*s=+-wf#YpR|jFm
zwkpzO02uo}pI{Sk10rdj_a!_2^>i*2d;{W|8?CKictqI*-qu{B!-r;e0G6h9AZGij
zIuJ`!nM~Xm&GyQ<Dt$a8^j3`Ap9o3(53DphS};N_kImDe;-6fTMd*QIX3>Ham_!_b
zXU)-q6zYfgKNrohLR1=TkRmMO2IeHC?VKxk=CdaD14LRSxlq-rHbPdb`NgA#v^aI~
zuCB(Q#{~pN)2R~vH&5;9l}JbDGj_c=M`-5w0J4LFn;Bm`hBv>Z$i-y54BOJ4vz1Q3
zsOudt9k!7r<G0T3sGC@|r-T8=Jq9{4p1+s%Kn^ItjPm_*+RuE2{JC*SpRP7i#OTf#
zY6<zPV$LN<rnT5XXSYJ#FAmc(?|(KyGef9N__^3pS)e&>)6HrLVV_GI+^}=Iy=5Hc
z#>IJYHP7K1lrR`sb3c~?xujKeFfOc+@u(c!HCZAvz1Bx_CA?T9j62Yfp*`l0jLnbh
znVqi5eHdAe--CO3j#)?u9GN1*bDqqSb3Y?ocqGqRM9i(y2Jrd^#gn}79Y41^dfT_|
z6?y9|&+<mgL}4d$owiGPehG6YPebORMpPvBM71m2eF1#hYLm%vXJ2%419=he1;Suw
z<%ICI?h?1l$X^ezdP0FEPN<@?Cv-h2wU@0n&QU!Y32yd26*JWqO8ZPmFRWf<9Ri_?
z`+!l>-iz!UHd^WO+?~^o2Yr>9#SKsMqd*cNDsp$zg!~NkMYYGe^{ePHK(zga)gL)B
zMIBrXnM>G*8RT4NmiG0XdP;1m*Enp1wn;-1sN#J$q3)Y1wmn736s<GyXv=WyS}e*b
zd$7xJdKLK9c~zn0iH}bF-*+OwBW1>+G!*17q9+(ePSH-3ov<piNdtd^GmT*)-^NJM
zE|C?{@s7&ZOxkOMG)*>o^vn51-TnEWPum6^`O<vp8n{0|-6D?rhuBWF^*Qb*CA-e8
z=47|Qdxq+3xn>;`4Ho(eU?*Rzvf~LNmL~=awK5&1JC@ru-2>VyR|;7PDCzBNk32cb
zsyx@!%1oX0^}aB13<y{UMb?uVP7W#v@4&(E&yuT90-43)98L9bu~*p;DpiP$E)}5&
zPjJyPXGydFu;8tcZ9q84EkVUQ*`!y+0akPy`3pDv?egR0UPJ?G{%2!(s%uEGjBmkG
zt3V>|9Q~nr((~k{$yNU?4N7|<zmr=fmOF2joHY{g@=XyclknqSyAj1DHkNVBmn9g&
zd4RS$unrYunh+oyqQeTvox}$;@=1ay@6I>(8N5n$`0hvV<#+oCdL9J4i-Ny_s%1HM
z7k&V$Z4S`ut5$Qj!5$E88?rKvACj~WScNMuY%tQz>Q-`5o!~tj>9j^%5vEiY0eEKj
zje_`UGdSH(Wm7AwcLnaYzciC@(68G7L~n~cDC&M_=hTXG(XUC>3o@XL1hzW<#v!S#
z3WE;@`E*)AdLxhMsKcFM-%pncoE_O962PwsCqo6)A<7jhn%eG5)I;2RqHvm{OtqO1
zn9%?R`q!y|ngsAZ4+OgN6^(y^8g0u*Hcm78yu~Ik`*hs$IfkgG<EBf=uY>cDpu{e3
z>ZU~z?P#bS^}0niC{oxwMl1Ek^e<Ux;z3q#<v|6<{x?S;HVv4KKW^y{-2Nz_d?_U2
zFmp$8WEVg$HG#Lw6Mz53NaFPNAz^z<VKG{}7Y>2`q0v~w4J^rYIMdCjWT9H-`7j%1
z7jw`GmWphtd&3@NXgzv6EuwpNhFs=KuZy%%Tmp9NlJTw$TCF18zS6PaGg`YbHz5k(
zp=*ZDVX>uoJ_QJcn6&te4&D>UldHJR@KSMx3-lN3F1plyLZ$EYpJ9k$hQmHK6+)a(
z?3v}D*lvwS)rWVKZoD_;0;w-{@gwj8se#Mp%4L~MN*wU4>jFtxFH-gS(5v$>K-(s&
zTG(q1(m$E<1nCrI8NkxqOGE)SZ7sShzYMIcpL|<oDTpm^vlzTt+Ty!?XO1HW^u%t~
zWcsN+7~{8MGy1KfUA0P?b7(abS&8?I{-7a9kQDst!qMiUPr>mqi1TP-L70Udi`Pli
z`h5=jQ{1~uvfbix?H?9^wrW5KS_%$F+38(JZLHUw41aEY!q7VcH~emSpozyXXh`Ri
zL2vTGW7wk6Qh7%(*W=}}H9kT~nZF}ZNe=v$E9E|i=NXX|YYaG{d_I*FjdhvAZ(wb%
zk1HA3$d9l+Mn%yrI)||6q;>d^j13fLW{KeqhiofeE7rmCQGLYix(y30Xt~|geJav<
z*#9*v33M1T=oh$hg*54Z5&YQyZ^7^X3Dsa{|No&HObqP*ofxR8V29lf>-$<eclazq
zq)1GH5(dyN<-?Ot6G#ktV~HM;rKKGt)g90+u;}N9*UU8|dlfPIvcV5LbYf0^pCs%<
z%wns;Q8ZU^!k?iX-$zPoK33hPFVNmDABfQ@FJy56l)3H#vz_$Le<oGFOeBsOUern=
zzFZvGLd^??D^c{Zi7B3x)r30M(2CMUGkm`X-h4Cy1unZDkc^!UH7H6E@c3dA(sUh#
zab#Ov5aw!?4q<w1M+HiYqcW_e7Md(1Ti%-zFR^7i{8RvewJex=y@}XhySox!-L|~}
z6wOhQTBomNg-HC|s!l>ML5o3rZ39wXYi%Xa&{DX9xFP&Oj)It9Km*Q@mnElHMaK?C
z!Qq7jqL-ICr(Dj4BEH5ejy?!07#NaTj}<dlh@x&TEP&*tXDGh%Dfrhmir9PvOlY9z
z6~|H;e|e`6Oyvc42|wd?4|z#302Zd=q|rS3zRClh{1TPX=t$8M9#(Vwp22#_7S^_i
zh9phDu+}JO)uNKOFguWhb<O`DT<A7|13lrj?@siBr-9$HI}Ge_%PhE6dn=RwqGk(V
zp{jQptej}t{&wPk)w0jg3%*TDV%nz91j=p7-aqlPBRrI_xwlz9*;6aPxEr<<Bwxo+
z{Gbe=L7z$=7)wPdk1qRfNspjp{0CG<Napq;rmP?eOk`{P()Vt`e(+NZ;$F8cT9mpD
z2BI?gapAwhWasO4(9+%4w`T{YmhZ>vP|nZ8?MDl}M@#Rkp@Ut5{qx-U?eXsL?i<nu
zPq%u<RyLf&Q1&m!$x-Qop%Xl7?A`L>zDM6#Z5ZB)-O6Cp9c;XrYIQq8Hu$9S&mGJ!
zE(k%EDBm(ss5S>_)h;%eJ#?8!rThphCB|`B5dKX4z_BARx;63}H7Q@LL}54&0t5}h
z42Pk90=az(2aG)cy;?&e0AasI1CA*HI`y)mGvNrUSjfLP>9_3~1AneRTd39m*V^`|
z4cGv|&E~k-E%@lV5_@wRe+VI8P%h&?{oGFhzZzQ3x|)FrmhBbIQ9;n+w0oeEabQz`
zuR`0X{K?$V`NI_iL>r+K`1v6DoTfp&3#Rhi4u>HJBH@25-_B3{+|lDuo`4&mPw{qp
z`Euqd=YhM9kU^t8J4x$Bw-=!YUoT%TqgL@|c4`D&pXs}}H?39CP6Jy1(NCM+_n%L!
z8=;zP*md*l^!M4lVilBH9bjBVKkB^r;{!`uu7Tf@f-PS0fBH7T<4FMH`MAAUySmhS
zeY!WYVd(U*;q6rOSWCS&A0Mt(QSx)`R8O;{W@)S7?=4;0f86g~UD{>XyUuGX^>%W6
zT7Q;Wr^hmK<>r5Zcyz3yD(qg5SphFj;$%Icj*zY}=2Xcokj;_Zkv}YRc&~_`bwnNI
zsoIJ>3{f1^;|Z6<+3Y6{u}LhD88%R$h@lWHmZ_nJ&_xkL#lx*(!(^*)W7pZ8s>*d0
zu%HF=PsmDI5QfKW8p-=E#)y&5Kq68cQak~W+xsaS_7OsuV?iyZv7wMeQB6Si!6Jwp
z50`M^fW|GP#e2p6uuHzlJW3Tr(~fx~oQo&KO6JCE%~LTB7<2@zz95UX7*Z+xar*sz
zuQ;~yUOwaX4v9RDti2GIbRjOz6cZ2?5bGB$h#nAkb1CXfh3m7RuE&=cJ`kPojc4&x
z$uB2Rzfi(C3cP|j9u5bJ#el?Ri4y$N><|bDvd(?hlLDOoCmZM9&?XG13j@pYivu|>
zvG@ucgBVA*QS2G9Wj{Wp(}V&ZtY3Vy3k*xZ-ipD@^Y6F((S>eWEO`?UDAdru9mwbQ
z{Fc|CAjR8&_ap*CLFJ{~_|$6T*(=MpW+(Fb`KLt?<B$k|A>4R>OMrliF!L11iBewN
zxq=Q92)9%0Uc_5ZvjhPYZm`cW=_C%xG&k)_F>(z|0!Y&^rCLg1qrLX+cqMlj=v#7%
zCG@K%fe&{mo8P|<=Fv`P8dK@auHF3H@|t&LcW$hds;j{<7=t&PlfSfEYRxGvn5hGc
zFk{fjO@j$&O(IBr$%LmjS=tq=I)oKyQ0qZ6mF1Dw5!3C%&7I5sdOD7g_JAWbP0ocH
z^`6_Aq@@`yX0(*=+G~GVM!Jc$C|Tit*1^|Z9-=){Trbt(FrO5Tu3SeK8elQT*m>r8
z;)RiQpWG$i9^Mh_(Gk6%jNmXHrYF(NAE8EEsIk;<`Ud3`p-s11eb6MxYy8#NyAw$>
z&m>E35>|@WSt40cco-ECGObTv_)E};_m?Cp8Ipvq88kDzd%ky*dTz)RRihyrCi8LQ
z#Z^YI<V;Tz)%*lotA!-9tSagbNXGc8><=orsxc#F90K5}hPfmP(!u)_mfoM)<V4ed
zulD|nEz4KI*Ef<9jyzz{yM@HBwdzI^H&{DT7QfNWNEhT+71qC@%}8ha$F4PFhYB*}
zZ9&~vGj|Bd%#kmuwR(%lwb#*9Yq1T#=ALaZn2-)nw<b$_pBr?D3FLZomHNpE{`wMC
zDVF$2!20H>K2)MG<XeXGjL?CSf+V2>3w)2qN4v7;`{daY`V9H(;Xt=HdJ*1U@<+>9
zP?=KBo~^ub<UDZtI^_S_XQ0pI5F51^nYTgukeWW3mXZ6=pkl%qf0=*z^9&zq*Mr%M
z+#bZp*ff&&CSW0J<Z?{{N|Hz$Y37o_rbj6{Nj-Lpt}^Y9f*ne=yQm?v<FAsV$UJs=
z$9eH@7p-1%e>}c%J@OuCJo0pF6Gh&ll=mSA_Lv{8?Vu_kLB^aH;bc;M5y8xyUL@pm
z+-G-owp}`_v^QxNKZk}M-WxN2hlf~q&e5@5U}8DJ#BhQNV<tok`~J~R1Z=~z*qD$m
zp837fWF|J0UIX`ewpT9vdIA?%*3ajI&qGhu2HhgT%)Y;u<9>d#{OnlqXn)yZDj(l=
z-Fmw}_oR?k(R$&2n7?d%?(-U7f4w_-^#aCYaWPjI#o>pH<}>xu0;Q*a9mfoPE6!r;
z_Cor@1tcSWN)?>YGu25F0;h_~^3uJ-qXt5wS9pWym+*(anxOYZBp-T11Z+cuNKaZv
zeoA$GQ#4VH-f;Lldu;xfx{HJDKSv!I+1dZgQAaJAIO0~=?%Ue4L%8S!l7x?e8#wyK
zOW61%z*_?G>>wokdogcG+MibsWfd1G7D|ccQ26e~#cgDYhXikBi>KpVG@tjyth8CX
zo*!XvNJ>tmtK-LBS};8GFDvGqryIm*h1{v*$J(^s?$r&?F5NrB>WDk2H@_jd*C%n(
z11zvzdN{hrC-k-O=%iadjNWm-pw-5)J(GJxkDynemGEqFRSyj{2RXmruTwTp7I(WG
z_$vKDQyc@A5zFX2I4Z8Hi8^^d#7m%_{=8JXwz~DpBR0QJ8!dvJoq%E(t1T9-hi9@f
z8A^}K_9wI^iacqP`OP@^%J5$I6d)0`DB<Ir5b-`aP2rwVTGjGDaYf|%{rbt!UE)5?
z5y2%kq-A@8Hk)B#H?G#otnxGuHoNqW6P5_|6Aa6t%>t(DW#$9C(f+aG-B2yiZ>hC`
zJ~~0!gDs$AEw096eWrTV4=?SPAcLwTk<H;g?jZJd1vZ~R5&&^b#5mptX5I#Ys-z{~
zw2yEB`eIb>wN(C~4YZK_gFx~pFqgUTRT7k7ZhJwj;x_U_5D2cs`qPARK3DScaA5!J
z_930(yqRuU&k$A(UabU&ow0XoDaKY72U31&1VAtaot407q~dGJo%Y$M<G_OAWTRt`
z<<5lHfCX)@A$?@4%({stAUP*5kf!Jdo}YZYPubG6l+{L9HrJ9O;Zl*xQQpg|fo*Hu
z20qe<YA8*7ea>1^4Urb$>2C0GBcFh_>Wz}7X-<4JZST0<^_TE1McL~CvTgzZR4CmO
z;IdX?&RSO_C++!9Od&Yf;?5=3jDhG$pMhUykkJfg4|xda4K)eBaYwB_wA}nV^S5e4
zqwgd2&3LYoVoXX@(E?^^n!1r|VH7mN6Kl&By+3!oc)#z1egKWk<~@6oP!1t&bIt9c
zhYy@7BeIVyWi)q>E8f6w+Oz~O=kpyrE)7y1?MkK3Ia|{M*G|2@wrH5C`|h8~I?bu_
z9o1+TF+-5~tGQ=s`J4JJDQ-EBwj#8Ft~7qDSP(68mw(mOXUd_G8>-fpYdU~KGE>5<
z7Qogm(G9e$20QaUi^CNnin~)Sclh*A4T`3<k|otFKHmd&dOpi{Rr{Eow8{zU+_lnh
zCnc8ch%9Kji|>P26B8YqUs`9KiqV`$B_?S_KD|sXksqnr%z}>u8w$QdXS5X=2|X}a
zZ^S1MZNG1t`>4>`suRwmns0{RRT%%4reo=(9-BZ;Ag!c^YG{y?KRESHI0%mabv*aV
zD0Ll&oPgMBR^e^qao>c66%k6PakbAA-UtP?SGAf!M-6I-ki|?ToZF*f@i`y}QV(w)
z^83_YW`9~}g5T{Ha9Nhfko6vbPoGag15^Gen0O6xm@g<Eu_qKo4Iw#KYBAU-*yC)$
zsk;j=Zv_KRZ@SS+gK5*^L?XAhD|`*mltuVjS*y|7$6^MJ$+VIqC(CHF)Xu3QiAreq
zJ(9~yEsQ%Yrr&C1h~Q?$eAMRZ5K}=$bZrgN-W;fS?!f7Ii$Qt%1mjXXvSNd4BL)+q
z39^c)158Us$ksY`;ZmMx8U05l>ONuSjF6M5ZGN07;6}HSOU4hgE#4Nev5w`13m()c
zj{zK>TrQORC7uhgC=A@*LPSM$;ffx^?^LS1{)+YyD4oIy1?OG_1*cJdkUrh14KE0+
z-R3-cSpN>coe{Mab9bIiXz0Z3v?a2=O+kFmx#VP|(V_?|6Iorcezw%O`3===862$^
zzq0md$SB_Ur&e>X6jQxx;z*-9Vna`xF4o1eM^V%5(&MQ`X9ihD#{WI%aE3y<8T{)m
zXZ^Xx_mG?#(JFb4kS4?kl-N9`(zeAa{adW{1Nlh*r8W7AfrDMdz33bFQOsST>qm(C
zq1mta5p&OAmZ@`E!E$5z+{NIeOR1XvtVpqNTN<pRNdb{e(l|utL!xLy@2M<F8uSNc
zp(*uB2`aJCK)FovXi$HuzldKs>Z|Y;1g76{Z|P-Vzh0HGIdQFZ@{*fCnXq?Cpbks{
zDUgm00(UV3MIP~??Z)vBLnD2^<Pa^<q1!gCq^p6T?3nFA2Ce6&O~I9U>#7=L>oXuK
zV(qP2I?Ev^^>e4=4vjCMMrbR#Gn$cv#@A3ypBAzq7gT<iVQCsL+Kxy->`XwC6k4hz
z^0#v0MbTdK#E1eS`oNPB!Qb01Gi4}4TpCK6-sfigVP^crrG?SqncZ4@g5T2B5cc>7
z075O*fOfrb_)!21o=KEBuMUqw0w6-Ih~!66>rNy^up45q;6*r0a3vu2-re<ou1|Fy
z#WArVSK6s$c~mR3*nz34uH%-(pg_yVI~|G)Di+-2WbI<BnuRO+d0}KcHOv@=WZMN-
zj<Ze5DJK;j$ps2zJNptKy)$(ybleK!W9_zhWRF}4swXSu6t&H<$Aa#A0W}ACCqHi6
zr>}L(dwy<J_$e&2w#vbpO`_7kV#jw_HT!jmmPld1p#ijpWkt`#Fcm_*K2@rBj+Ty1
zBcz3!_-aL8O#a@W>Qt&J&vN7p!^$DY6W7RURmp!WMM=llYE?of>t@6Zqe>gPr!~uB
z<Nt5P#}oW7Dabdo7E0bjrRkbUWy;uwATQh+rDk{<0OFT(WOeF9tSc6r7X*d}6vmH!
zi==*_JrL9&TD*!!v~}U^s-h`6Bt-9^Wk?UfeUrHc2xO69CUI^!+MEU@#=R_6TumnD
z2Vf1(YP6kogY4n4xc1w6TD>f6mSFQhyvv1mrf!6Eoh-TFiuzrEIhh6-x3z0aj0V|o
zvfafH{UoPAw-t%eiiLbbmLwchA>fNEt(MeCEAUFhI+^0X<2L-C<5uRZLRSq9GQT~|
zwz4UM`@ZQ7Sy7>1LQvYeefJRyI1_y6TO<yS+SYJ9=4kCz^yJ5)#=dna6QOi`UhPl9
zXI~i6^u5!klF+(q9#ad_ChjB5Sw|nA?<o8p`M`nqDhCafYfja5D=P81h2)tbNrteN
zq>Y-|@-norE?H?9IqejX4E##887<Az;DTXgs&r7EB+w&dK}Qn&Ub9I3vL&ZgO&4*X
zVDx-CPeBzAsM3;RVp$S(4mP)>anc|hRbE+1ETm3gD$Qw7NFDM{ISW2SZTU0HE{j_X
zm7nSBB*rmtWJz*3Mc9t*X`IP-p{^kBztgFk+DnG&LP8J8k!x9W=0Mv{3ISGrX4B*B
z1O!6-%IqF?*$+XC#&V}h!k8beK$}vcL=eMIc00$B9+0?|zis0e%Y0)g(TFIMy=Lbk
zBHPwd*%j0U+P|ExPnl3|#7vNV<;{!MK#*g=T-;E&jhnZI#X%)~5|<sIzQ*2ClJpQe
z3n!pk^9Nvol%mTNg7K?#ZAfw-V1X5}%fGLH7?V$J%jjkzC2<!_4)I98yxi!|{)SU6
zA@+vZZFrS4NRe(^oV-~mzMZ_kV0&Lo1vYbm{zH+bButG_K`4lOVe#Rfu5WVV3bZ$d
z#|=D|V(_L>u_1RFf%w%FsLnERg@T*BS9OzUqtfA%ed?y2-#;*Y^Gw@TUah>dKKdjQ
zGA9D^xdyuUrfM@+dDJ^|JqF5!#2B#9ZRyMYu_fF!jx`vW#aXzrkvsheShV||E)6&y
z3tM!3dF<C#V-m<jAe>S~X63t!EE_bC9CI5<@BZI13)6%3ht6bly(2B9uJV$!gMbH)
z(FkO|hRUpiGt_kOnIiXUQ^sMbf_CK5c)?A93xEKt$h^qO)MJ^!4vifcX|~B@_|`Y+
z#9M9dw4>rV?Zj4XZ~e8*qvBY2sl?0jWEzk-#g`v*^Uulk>r=M8-7mb0T3gEhtsdw2
z&j3S4X7>MXsB2l1D-OFIVdj~Nb8FbZgSz2Ovl?tIV-elneQj?~OdpgLc`TrCvG;Y9
zbY18riD=)LJ^ebz7PV5n>Z9UKW^!LI2>a*h=IiC=wDP1M(r&DL|2hwNKVV2NvCU$7
zFFbgexy$Dt=8mwq;Vt{#fTDS1bl)$ol7q+M0bKxH01S_lhogkt=->@$o$wbSZ<RbK
zZh&k8>4sx!pW}>ul4I6T&d=}jkt+i>SIZ0sdQHB{NZo`UX*6SOKQ;jOWti1SmQr(Q
zZme#IzZ3GDD_2B@Pp=v?5JFYM_G@2U91k-F-p@mtaHcxa9T9z9+}nLi1yS7nUp~~X
zYYKTe=y}GNT&E_o1woMBsv$!Pgsj(MI5{W{3<6ozyu?wm$O^DHMO>Caj|fB*u?e@e
zQE`zDG}k1x;GTpfRE0Ya&px3Mv(AI>DG&B5o)E_C`q3qe4lLoa_8&-WB7z&=jpwIe
zx*|Qklh5qi;nSC<*0FKz5S_Y?^h%?poJW{3;A5k!N2?ZM>RkJ;*=}V8iFHq{5p<uZ
zU3Nq5U6xtmQUzKo+*SH%KMXZRXKat_Oh?k$GSP00l{aduholFdPb%GGl)@lSiN@GM
z<BWL~x58Yl7Zm7gsb*;A@hP=`7P+Q4hh&X6hn`+Dmfc;?7;MikA;;*j;U8d<lqV#(
z>QUT!NsPN9A?3<R)0FFZlX3+TvolGzdp)2gCtQc=!iIW@9n`T(A^h>jC%sy$L>8#^
zm&ENt8@ab==vhUT1C#4CW=|NzLbLoF7u4RZs<MF3jS}<3#>S&|5)BQL=P6odmY#)%
z1$$W$OBjB)plnEd=@$^W{x(#6Qjmh^3g~6@f#s=rDi%&pgTScthG48ifKl2d5f3;_
zj1$SsJ3@3X{b_WX?B4#3*Y{%y4%$D4WbVV;3{cjQ>7Z}LrJI-JjUC99=URb<$g79e
zNv=Pki3uc~52hQVWF_^&_f4VktM)SQ31-jb44J3T=0vkbZkd(%qM1h3b;_01KFr4|
z9Y&R=Wy{PhI^i0(%lM#&hb)x8eske8w?7@$(=df;3P{8i-{RWL?i=_ITi&VRnC10Y
z(D8B_S(%;Xd$Q^bG1EH}hwdM<N9^6kbiw5c%8d)iwuMud6mwNc>yrcxf==d+XO2-0
zN<F^~J-ZZfO7B`!U)wZL;ADXrjv66JtSTO}xtZQZH_b`jVo8eEqr^!U+S}wuWArGQ
z3A^U84@vaZ#3~XRqmnqU%{j{x?8nBrG^^wvBwndgo_=;Ai@=-epLTP&D8t2Q2K;Z5
zVlq1riU#_F>BMaNy|=9UM?rO0ng}Yg;HE0X=m96~DDTbY`+zd6@(tIU#mILg!XwKu
zh`&30a_$uIr7=XfKlOgOeHYzBid6`hFD$2!r*0zAVLB9fxJNw(0+2z=ETP{$UfF`$
z5GSgJzsw;?mHHe?4;F3|nd<7?hDgDaI)_ck&(xx0Mz?7Fq_d5U2~xYq#n*y=IQ(+~
z4dAHB2g>D(pVIrRt)f!y`j?QnyLfMgTcH%qqplYK)slpH@vPHOSDXru%v&L35-o{i
zU7-<6bjC`zL(B!Y&SPAv4gyt)ftop&h$w5k+O8@WRMD|+?lA^AUkc`*Jyy+NF0Z5y
zgc@RBR4lO1h`y8qXXE6IZ93W>t|vcx<ZS~TSQ57x6Qx*$R8rPOE{A4;9yf6thCfW~
z<SPO>e&?TOlXs#j3z>H<*KK58mwN{<P!fn?HY}XgmDA`3wqjo=)0hMe^b!vE4~G`H
zy=el_R-mBasQW(14egr4abKf~LSvRJJ$}EwzJ4xkVTrO^in7W-8nT)u67;LF+_=AB
zZMVcYW(NS&Su_EimzYF+D>XesZ5(WNrq)_vX#i-p7E}q&PBm^?VGhmJX&wUL?pA@0
z%xK@5v_^>ll41mcPa3S}Ebxv<5fC@taN`#l&VgG)eR%k7NU*gF9TZn8pwSZ~1ZNQs
z$}-9G8uX`)u$uJ>$p9l+L!Xy_G)ig+OY{gy7h>1oFzv0He^L)_tLjQ}3@&b42H;8L
z`x(lNi_CClM1ef%s+g-&2SJt;AwwPs;0gp&E1cu~`wJ&T5tuvrT}br9=U8l8-FnD=
z|LW9t+X=%x*jYo5x8k~Q&+>&lg`$dj&9+c|H-JQPwAUO(QRT1qU=!!93^aTB#e1S(
z%)PU4Iw~z>VA(j=o`|M99@<x*RY;rSe5^>`i^1Joz)fme;XJ<B`!vGBIgL?7j`FjR
zgzvP%lkq(z#V;rgIxi_vVowknE}`Nwr(mL3(k`s3FvdF7%P%KV*%V*eP+dl}5bnD)
z)<%z_sudhv($*eP`96AYY`P5WfM)AnzQfA`2N&XE!0N*YrU2ib>frTR$lIJ&wmrAe
zj52NbOy?AK*PlzZkT@8Glxts$+3B`FAAGvQzme?Rjj>wwA;e+;UT`ttOBT$Hd<bqG
z4vS-fID+d^e--nv>oN0Qn8K!n%!i^)eZGVLqYup^Vfvw2y$Q+6{Ek)?Al#ebfOZ9@
z-SW7ZY|#H)*-5TE&4<Q_<Jqi*|9g-WeQSQb3vDLj^biX1tdtzw9>t;*ZZhc!W3z_2
zJb=ZD;N7W1I6s6<5w4W}$ZZ^?-cYSP@JM3A78*;x#fQ7Z;LH$;^tU<3J*k^t5W`_5
z^w}~XS`fjlp)*P2wocv}#@*R(9l&Ls_&6IBJ)-@oOZdVCazIU~Wbj$PdKz;S;rkH)
zIe>o$y!I2tw*-N&VOP*>c>Vb~{!`t@NEj8Qbw>C{f$rcv{a=gVw-aJ$Y<68$3OoS4
zGD8%IxuOR<(U})0QuNwcM>zP={OZtdn?2F%f{+j&5!*kO(5>bpoga~QLnR8EF@U^Q
zmqp8~4<xn<3R6U}H4jz}oL!Ij$bIIV5NIAnjJx`plMsl(tBSObCj1#8VC`v7i$+Ws
zU-+t{m({X;fO(_E^?CV0V}Qn<E|>91l18jI;GeqNO8d<=?@#Sd^$a3Ly*Ku@xJTuv
ze$zW0X-d=cqH+0l9-rCBUX>7CJ5#|6^hYup=Z{~)&CS6t03HF%{k7Vtjs|#RonOol
z!8bTMfa8)CpZ+g|P(aF=uQWk0+*&X7K9QXM#~B}0vbEb$ID(X&<Z7p{@Gpk|K#b<7
z&Xpe4VkKlh&XJxb#RpLJTW*TxQyb+|f}cfkbFA!*+<p^)imv^;KxNRU6rG%gHYb)K
zic>b`XKXqjh4N*S?9MDCj-FsG$ZA&qY0aOqlXWOSrVxIQ%_?oap6*Y&F#_D@|EvHR
z=;;|a{$uuq>EG&q)$ShFlCaZmN9eg!>naT4XCALONK^<e+av;Za@x}(4a2~41zHnG
z6?%A7(#9Kaop@mf(%g=lHmv?qr*1fYa!6df%$nTHN|W<XLWxaYUfhIFI-n!|RZGNz
zl~JWI;Fq!Elk@9IQ^fMsc?(elx|@738@+w92cK|&ALc>6m*UydP!0$c`~^3G`i}eZ
zgkk<#_=0+h+@y}%pA$kqP86&ESHXQQ#!LbAdK0p72<%|0C#|R>Q!6E~MoV)uK8O!x
zmvJ{Poegj~4dJuY(<Xa`?7Ss=-ib+&PS_h$2ob}~qz>pKXpQ>L^`lPSOVM?S%C1K`
zi$1I|%@x$jsedddINKV;I^&f22UhF3u-14|5OG&pHlg@0mp~QTsZsDlcNXR2KayN#
zUHjKuT;nFD&<=w|5Uvxx^iej?ZWvvlHJ7baVVdcU%o|_C$USQ%b>AgdTRP2}m-<N6
zl`04|gWzcWlhtSxTpA@Jbw_wn50W?+dL|f0NHpk@U3Ihrd5vWJd8|elUaBeZ@~n{x
zEV>ZmY9{J205f)y+iH-P7ScFSJdBKL8RZCf*A6!(Vl;Lro$@`Te~>e?7n6|3=)gf1
zd)EdhSNK4s9W7ze{I6|^V_(7o0csn9OA_$CKq8$h#5wc#I1vdr0e$-!F3Olft0yCm
zf}LCusCWy!gg^baR-;e$zbU%IPyhnMTyB!)xqWy;vW=_cTV@iEwvqD~G&QH@N{0S0
zyC%JVt6YvZf!&+GN=z_562d*EQ@pIa-PscixC}vO^v3u?Rl7!e>DAH9ZnTHdG^DPe
z=kbnqR;Q~!StdMaieawA4cSR3c|@lpe(tka*`A~_wJ8?gy<=s&l;=qs1)L#)qHRxU
zZ^jfc;Z{j>M(TGpSgI60$lq9{erNL7%ls@cU`4R+CDJvA2DvYeUj*qfZj(H|6x#_b
z`wrK9*Izf9-7on;dpx#&j{TeT?6NiDtEJQk{Qx1M&6_AImc!D@uwoL6eDXNPwy0ZU
zpTN%&ZzD#gW>Povn|UPAf{-$<76T2OGZl>y&pD{c|IHR+&b->sCqyP+W^ZYsNP}TV
zx2MzB8RBgYN{|bCE9H=LGzd9sM>!&Y`74`*_}#3Q!KE`Y`e-Hh5Ao!FM|7GQLC5%5
zRk*zj-UsPSAKQ_VW{wD;qe(dJ)F<q!PpDQPmq`Z{&8aIfRMc&k81JFn4mj(?ncy{=
zKjM0&@c<ev)tTl#?y4~NdYP5l_H=H0K8e(juyYh5`_awJ^Smpym;?q2Wn3C^yO<4A
zM-d&J+jS?3Wjz?~&*t!@f`m`EvzsePj0{n?BvUknMF&kz{!*qDMXlAC5{c@wk&-I1
zpIm<WePC9;OcMu8hug5XQiJG4QMJHRzqOiDS_3JeTWhs5WFt{7Q7<zcETDObR5&^n
zUTV3v=U<o`%(k(YE(~?pVC}c5B8D{#q%?bM78|~)vkn6}^TYY~_vW!4AI9#HD{=Rl
z$w8VC3N=TVrHwV+15T`WE<(b?f@^PLj{^CnsrN<`2f3DiqDm`A$U7z?)v&JJbic~S
zVcQIpSovuMv(m)}Xuyfvb;`G`n`s0$7)-wI?*+wg{E+e|cQqm=IQZAki!4Ucj<v_f
z*C0hQ{}N%{n1_%W10(<eot3@(E4ro0Y(NyJpjy;g#kVm;53ZbEu6iny{)|j;V)@9^
zL7E|JBc~568)qyJM&fDFO|+<O97xyU`LQSSMs9>ETp-S@yA`Oz-#G;wX%8_Vp|2)l
zaHZS?L`czfSs73LN)PS1-0Wvf$lH3dAD7jqE~3Na@y2tZ`M#Hu-OMO_@yy&wpG<|%
z0>HBn*Bg!k1#wj&BOzW!!3r}D3>b*oykp}%0B0j|oV5+kk?zy=4F=;gf~na88Gkyu
zOU`W|Xv}Q@`7xOJzKCms3zPy8HY}jdywbW;0A(^cm(h_QVMbFcY!!&|Vj?zWm4-99
z>7L?pJL3eIe^!VhuF_YJv?Q+<mP*(>aX1khuMjw(JSWoA015GU)l#6|cyzit<3jNM
zfy;V>x*HHLKQFO-ci)>4zQ6KwSYP-!Fi1)cJ$2_qIh9{Nb1ASQGN){6joK`VDGyVT
zBUw;SwreX#5Ed1U!Pl<UyZ$WQ(-%Jo;WX^0A2J4i>Bdq7?RPHA+Pz!riPmOcs^x2p
zeE&E_764NiU^%2AF{nw5Ikvb!<tHD~*Oeq)`e;JGxZ!73ThytU^Bpunshfj++F3sY
zIMI|}D8$!<skk~|AFxhVxQMBG8>4ozvf+BMgo&D6LfYIH&C$l1U7dRhX0pqOYjr!V
zl@L@*z_7L8qU#X5boUO|3v*;%aFCGGsah)OXs4nYbo0BMgbPybbYCdUf?(HBG>~yE
z8uZH|S-~IL>uJ<-DgI@b@;bTrcC%NVrO54~lM>Q9Mn4vsfg&oUbAusk8sfO4-zh+2
zl}&Z#(D>_px`(O8ni3YRXVe~h$4wRrvT*QfZ#XI;Nb7(mCQ8N{b>5LQJIyGogs?s4
zu$lq-i<%x?LJB_tBWC76Y!2sgE&nNv>?<mtLdP~Gwx`qXNmO{yB|fpE<^k-sF%9x)
z`JCn1Zh{Qlh+dj1UCvhuXQOv$O~Ubrh4IVhZNsJK7m=p{dQ>8s5=2O16n3YdPFN&;
zq6|BU&(+en>SRQLT=wh+SrWV0-U%!Po*e`;two5*5o-`EEQ4NLN;@h?bIEAEYFb=p
z%(Wx7Ok>tYMW?2JSzK9op-Q3<c(x=f>*XrcH?LavT?lba1OD$~7X4Y@1`l1Smo4bJ
z9J;oV5t_7V{e5$u{5;9V(Y{^89bHQqWS3~$Smk@7dmU=Yvk*a~LwgAKDu)RPBQ|B|
zJGTGBpCOXzvG(%8--azXjL;Nmw7FImN~GyhR-~GnM1T%jki1F0#zz7+lTQsh<k9<l
zxH+6Eg}~5wzI~eIC(C`UC+Bd(L>+Jnd7nqY8b(C*q(rV9Y-0qV@B_{7^7w#<$M<!=
zO(#je-YcZW0otd$Ba`<?6pqg>oA<zUzm{z;R{+v~$60`~LF7gr+`PZ-9F*BQ$6sUd
zjjY+tAEwDR#I4Z3K37j~o*4;aFx1ZS__@TYT22{7?u<gj9iqT!<8J&Rh=My<=|iiN
zaZ$TC9d>>q(?W@1L)0_LT*vm+YN|3bn+&?F*}{D8uao*ZwNAO?7;f9t@-d?m;y}7o
z(~P7_`J>#fk|QNm=w@+HKLJr(9?9Bdw`MSxqIL3#f}Kf_hZK1|L{QP-k`!_}d&_Iq
z5hwlJ;URap6;jMz_3Bn|XCQ-3wPQ-!+Uyv1Q|zW8=S3~MmcX9!+>oO`h(Gl(y%e78
z+?kc)<+OoMNiAUp!?*&ADHJT_%#LNq6%xUDv~zva+na8wC{;m!su`*=ekw`(zP5#X
zW{1eUZd%cvA}Lqi2eayyrc&^JGc;e0sN~XtQ@sKgvh2$|N?oDTe!83ue$bV?Bp65K
zmf)HO0wH$HPh=MI$z~mmHf2zhc?-Syxt=|5@?2xkS;u0_t`+!r;p?#JVPiX`(8kJk
z!Y=1@c&oHaLucUjh^b2YJX7p9%ua^WOcJWJirKq7>_@_L`1jV9F~Lvvcz(>4S7c*)
zldt|3ZQERuv3ZqWNveIBInMh;6{-DRbukmw#=ZD{G}5y0`lUcnJdff-%1k=aooVUA
z$Y*!rMbfOksS>TcHvp6^j+p(m)j!IsqPPj_vHE7Cq@#UZ=)%!9cWgFm;LX%mH&&3+
z-mJshD{-SifTAm{Qb!?lR5Z!avI%_evv+K*dtxP(_GGu#d6dVINchyS``TB_$q=pE
zA=`kb)b&x$JG8L*{*<zJIwj#9xsu~>jklDYDNX6F*~#PBj6CYor~LW-^-~a#bZf^l
z+M`Z7Xc;IOZ|Ya4=9S=<dAvh*t+Dmf9f?{Qj|JS!stSrO&sE-;oyjtF_ZQGFPbc+%
zy$3Kc{O3J@o%O#1LU&_N#$vbL_vp=wVe~IhqvqXP#^)rc^2<8p3;DB$2?rztD8_>a
zvis#xvW}CkZ(reJ)Cb&KJE<Gj|BKc<da_7d><qj7471admy>k|mX*2XljY|V|KJnH
zY%+BO7$B6s?UnBJOH|bO&^|9L7A}e_?B(&vnj{J=#o%Y^n**nflLa0`&-DY$+TUF`
z%`iACJR7{@H;P@k<HWCZl&O51#M}A(5Fici^&v?9ABk7TNU9Rz)}dho`k1hx);C_!
zcd!pcxHo)nE(;7o9z=*wgCcu@X(<infbA8KLRmhEMV6hJY3vrx6;MDdhg#F1ZwB8_
z27l(3-~^){^N8ihW{K2v0&0l{G?yuOdkfS84Sy&aQTPSph>=$cY+ls;ytA=QrmG2A
z)GeC9(HpfoBzBXU{}+d-1o)ohPcX{Nx>7g>4QIsrFgjIbh|vz4n8MvOQKp(7ZsBRP
zyzX{L?&H7^5=(I5cCu5g?qix?%6MD#KxpuC3!y+JNo{j%L+JGEgDc0NSO`#6cUl6=
zl7uW}1DAL1V|EoA$N0nx^XvTs90(H~*8xh*PK!k@>%yF1dB$&RfoYNsYUa=srC1rN
zRrsM(2)?(zZURg4TCP=AG<sKj4N@cnMbi>-6wcV?ySx!_X~YgDSMK*ime_~akK_CK
zq}%DzUrlp4I_>$W9lFfDp#xNwb|FnQOCW!?Gy*LNO^XBLgI12q$|@qoL*puxjW?y9
z2|^0Ipt)AqQg4)VP;}00Rly9Unu&*eTl`q=6zE_%F@L@oofI+0)&Tx4fV)y{6+plS
zt<xpuF&4v7R$B9bGp({;+=@)aA&iHvfB*N;BGwt_vKSj|paA*1xtx>a7$K1ou???3
z%Eku=37Q)*B7pPpP_CmLFx7;LX9nLn8$rZuGh2n+=F#!{xZ4I0axdpyH}tBhV_so5
z`M`ijNH=ph#D2wv*x4&NpvFROu74gaW7gE6=aM7}o$Mgjb9)nz4=Y@g>g(xFFweN&
z3~Pb=l@|2bZ5S#V_D)8K(of9ud<jPL?Vbou^Yz@jrU;|#7O1}y)ehggcxR9GVzTxR
zbL}y}ABgu*F{7bLTH(v(Bl&4Nxroj5MIB^EG%|Zkn+_cyRmMhLAqDi1i19LHsbNcu
zb^&TBr<*cB5OiU4a48$!OSqJ)TQ()bbJNA-g%PD0?(9YwMdpbTLl`9yM6|2VD=Cce
zf?!UOQ-ZnDQIEux##1c+Mcwc#(P&WUA4yyT$XQr|0sf6=E=>+k$x0~`z3Rhpk4#6t
z6?}O#Hb3LwcD$xdtA{QyS8N=u5g5UMBOvN1WOl3_0G(dcT#kyKECO7mzLcPPyJd@V
zRj$Y)V5lj>ukJw)^-viv=@VRtE%M58C}Jiwrw%<&1(MJeTP-2^K1qy-Y))l2q$9!s
zIt(v3q6ajhbQO&750(VS0Y27WN<~beNPL<)W0C|p>WCkl$2{VX@DF5v(1J~XONJ1?
z<7PA^2+WoL3kovc9n5q9i~NLvq&YMq34;8@G2K~baA)SK(A1%}eV;eNKy8m9-imKT
zqG)U@zpjg@5!E!1jN#xzYw0hgOFW^Ma9u`LsgTzto`H8tQk@~DQ%T?fnsqV(_K|E$
zyq&e0Ai>p`<QtJ&p`XM&I#}W$z&MCK=i;1@Rv;k=$D-z`0~CJx4wG$gW1PB?WJ{to
z!x~*MkgiIy1d)Q4H=EvWG${_}<1MQ}<$2U=txLAqd-Q5!f`kU6JTiwi*){}6LT~BO
zhNiu(+A>q$EhFJl1B{mG04L{JMkBBK2XVS)fw_?+WOb)9Vu9<!JyNs8Z-MS|ez#{c
z1&p5r8@8EiIzgo*xoW2M0h@WXq#oPp-LO0D<0UpW(XX|F@05C1V$$^62qVJzNoZX)
zR5tG0k=nBU58laNU+RZV4{p?ZJ@+r<fe?n|P4i-3xDjS9iDOORd!%sq<W2Oe*$1R>
zxqYZJlmW4!UJ|e^({iB|<P=UyUKgb{ET_()5-D+fR+{6RQ=KnXKJeSj9Ksn53+V$b
zW(rRQdR!|HZ`lN5;oHq?_M_v8>Sa^Le}@0|;TjsSbPp2knF1XX>|Bo!2n1AMj3AL+
z_L22ExI^pda{I|j<nU}$rVyT=)B2ogSu2!jho`I~ENc^?q~5-|DY+reVONm?YK59X
z{oxmV%a+m7(Cf!BR=7^kU3H3==^ZjhJ&TaSY$;$7GgMiUUSqZ76QK{1fQ!Uk#*i}y
z%4d=RiAqcV!lAVEm<FyJfrCjO*;#cCf&KgPp;mu?`F-t!N{5&=frH9ZW-+Cizyc|!
z782MPR6y0l736GFii*Mzt>>;Zw;6eG53}e;Qwmy|2m@bne?f;O58atF>H!b*@pZ=I
zqU>}3fgJ)9Y+Wy#+XO~gGl6~{#TNhURK}Wwsy|p#7-*Ga&$~n>q6zw#Eb$Gbjp8q8
zz~TPtTD>hF4)_5o-h5mgHjvTnM7O^$-<0hFwi21=gK1%Y(#FlZGV{(!Y5n}d@3Y8Z
z5htM!_7v3qjZZS-Mo(PuyFcHHqk2AXL=ZrqXmTk@C7=OAcT%3h{SFDayrpJBR#Hz$
z_A$vm00w`UWybtX_6DEwA^G?pY$nglaA5_#;8sxeC*bC}a?{z$LK)$@4j40EKDFUd
zt0qtp&I7^t?kOQt@gRAH?uq;^PPtR^e;9eZvvGUmlvz@_YL9A;yX@<pc2QL*gn>qZ
z+T~f{DN^-n*hwou#0Kh^H1pgtBq71w^jIY7lKA~nwkI+k#a(W=g{_WJ9^?h8v33iD
z+*C0&BuxfDSU-TYkrl+3awz^qeC|sHehSOqR@{2zh{aYl)g!3^Pz=Rc4Ogm;+=3c%
z^UThan`k*^pm^?3gmE<ox<t*T?#iJ^KCa`Y4f^<H;vZo|&yztcn4O{!v?bEH5DfVW
zW6s8#u5!FI!$lk69AsY2(kxqFi=JiMtT?G8R8%UJz}=)x$XAK4Rdg_m134NR+8(yV
zIvFE*JGC7AAI9D(NR()cwk+G`Dcg3{DciPf+dO64wr$(CZ5v(v)BPgueG&IB^D{Ft
z_uPA}HOEAE2~eq_0D<A*KvUAz-i3r!XuG}F0>$9ibP2_zghm8?`kKxSEuw4WpM~R5
zqEs;TLgpZv=qEvKI=@#L3TYUs29O;(i*z!K>!;aLJM>zP>XE)`Jna92j>oESetApo
zR10-M{WJZJes-xNef?&wn;==ZEUj(h1J(AO?lPy%<+wfcxYSoZKyOob9|hgC$t1YO
zs+`0oyR??e*)!H-)8NXxz#V8g=S-cY@+2NxExFbDNd2XSs3I@+x?OgM&G1OGY-rAt
zmC;_Vrwn6{d6s`0c2QNTqBB%d-`jMZOB%g(@4-2AzOKUE{FEjgAzg&R!g>*InLni}
zrWL@g4l1KYJrtf~#155{r9mr+>I@$Na@f5gK`Tj7oHEfS*!QS7hAzZ)qq(Nu4*1)4
z(C9~6piQz(n7M4O*zmmXcqKVT;l@R;G06Nhxnv3;eAn-EsaS~mg;KFaNHK4Q?@z?i
zjq$|LEsAxY<)=&g#7f|4(K7Zc&}swo{IG9>pG*+HToougQLxPRz$>_rWd>VAMbd$&
zWL$26KQ8gC;YLyfv;4&Y=-J%*!vxzFaiv`eJ)wl)<NQ}jFDk$##n6H>w{o*dPE;#F
zKz`NXL?nIQ7t10y`qkv$tEbTF_K;lZMcvlX=GF|BQOt^diy(Mm%EZtdtAxKkG|aU{
zHNT;co@WhF3-3+a)$*C@o6;4`(<c<?N?U>^4O_aG4c_s(a~rc#`R1)}iJXfWiCZl+
z32UvFj#gH;jb7`R-HdDg)P65Sj+TxgcvyMyJmdhKZiom74lJ0}GmjWrFZTGQKn!K#
z_?`-S4)u0>1mC0BXi>T6t%8`i|L~uh$q@2ly5Jyf#)Ml%mNEtwwe$3Z=d?~^5nLi#
zdXDM#<faSq>oifzqKngmkz~FgRB$+McvZQk2de$ia@n+|!}NYN-1YryR^$h>qBdN9
z^#0Ry*i{_etXbIZfZ`OSLt4xFA$bz=DcLG-Ia$fDWX->r2ctM#X!7^Pz06g&oZ;z(
zsJ&!+o!2M_t$1|6m2f@fZ8sK(0+Q{(FQw$pBrQ(p(TDPZkUY`Ctv11!$E~49X0Q%T
zu)};}=K7HWwROJpdGz4d<!zh!7IZ8d$Mz}zDhBud&=Li^;_v-I8oo_1SgtW-O2<bT
zf^@A-vyB47zV4sPTa6o>kIshV=9+8AX(ZQL$N_57=R{b3W;9jZkoZJiKgb9fR@)2U
zWo5q7E_i(WbihsS^wT9gSa)oRz+U04L;{!<^nO}Gzz?>+cmfDj8<d1??f14Dff=Qu
zQWQ{YnvM+JgIuHQ-q5OXK*`RHPz~#dDiCF|z%Yc<l{A|9=Y<bD*nLEU&eh#5emp7B
z<n7lLRY_l5v{|tNo+9*?RfWrV)p<&+oC#&*=H=6N>tH9o7_r--9o>oE10Kl3bgVX~
zVxq3^dm@*&B}Fz)SGGD4tNBdYfTOrX!k<;eEbO(uI$cdaX?$Ev*f_}}mWuRw4JoTN
zKIgTh?z;bU&{{Hm1=uJro2hK-(Yu~dIXdfD*M<z2v7OoR5JqKe1O>#;#S;k%)AGg9
zTQ2)zU&>9YqpRd<QTDH<FBJD{&3x6dS-*L#wIL#3Z_;J*-Np+?*k#!@Y)l__2p?i)
zBS7{*19YBN#5#3SI-)xBZ`+-K#T~D%dq+zYjLSMBFQ&ED(-PbSvXAHn>$aU=)`v22
z%yt#IUDl20`Q1gU*4ScR(T+bek$9wm>x8s5H^rn^T=>qI%`=A&bixh@@X*hC0GIQK
z_}go<Xv2hBDhKs)y`a<|6i}WW9W#DjT%ma}4Kxo@emG_rU77q$KZG?i8c-kSy57ZL
zP6rm8qe2&Vchg*2e)Ao4Q-O@+W!Fa8=k`q0@!<BQp%iq~5~XsLbdDMwE%`}5<x*@|
z<%)czzWqopxoelGbGDm$(L`e(6^@P;2VL7#{S+uwXU8QK*NP53zGJUp)~nC+r`UQE
zn~RebmSbE&<r4a`vH0m&s-%ZZ-1Zc#UG9m(N7W-}=_2~a=FZ)@hD=5#rS|5U_qOxa
zY1ZcJ*S28ei;13U#0=$!b_;_a_=E12|C00c+iGKB{Es6x12g;o)@tigmy9`Jh3THD
zxy8e%B9HR`;P)f$P|qDPuo14gDogqP5iBbqS*y#Mbb`;8LWzGDyCcz%@s)LWe{=bA
z`*2xdSKouEOy~Vh_w$YYL?@=hN~iXt$10BTE#{L^pd6S<!b?=J^r8N_c}<wU!NO4l
zPz1OGK5>mea3i11=|lO(`_ux0=L^s2h3_;`4-u>hOtehipYiqid?sbuwRLq!Kt8OR
zvownm?n{NznCc*qBoh%Ql~`X`L4XXr{R?zqCaZt+`G=WEBua|E5_G*Wn;eKoXF{SD
zI*mHGi7@5O>OJSQa(D3#{9pHNGEcux70#hzucl7RS&N>llADBS0=-eWhX^9uLw#6b
zmbpAwcB|%Y2lx~vTu$890&*E`Qp>^`Z($C@fw8h+j=XUdp>qL`*Ns<xcJ4gTsH}*3
zUfpZ2@;qmgN`F1{<KkRRoJu85PtFeN-{nLgN}df_e@@h?@s?#m_o8>jqgK-D)^yj3
zZ#kI*6;~fSUUjsig>LHD#H28C(SX1vvDtMyn7)MMwQw41aRh2BT$bPfzBk{F`Bg_~
zx&Ty(J#z2pO=+uA{#<6FmQE(1Dr`KsIYqf*Es+(D@O-k(h3H9a9vN~68EFN4KAD8<
zZ(}^I>hJG|I-6+-*!Q-E(e{i^7dqai+ObUK6>Gs^i0M+Q<~PbO3lt#8t0ny}hmcYZ
zrLMvpplJ+^m=0j8WAuCCS)3_49u^`#O-~I$UseLsK<<jvpD!;DhH34w0hiHa2G_^8
zjtW)qdJkmhv~Ra&Ie9*;CnXwzmMmO=U=uoQ;vk62iczf|AAOKtJQE=LsX&iHC+%gm
zgz$s3bH`MY!O$Pi?SE#$jW$EHy2;vNcW}^4QU}?f_0m-uFhX4D73hnEML{hQ+AjgL
zi3p3Lfz<`yPO}uY4U#EB5<_65;P}S|*c5?O$w)dH1Jy^+R{?CAN1GiQ)0$<V>e68^
z_m-+s-sboDo8|nfIt$}I?n~3q)zyLGIbpVmG62z{sr#D!YWhsGx~lrHP-OahjI)oK
zvrfvZkA>YXz9TqzAQTU)a8v%)xMI!@1)x;#Q2S&uvf@)bet51M(5gnXj_g`o7G+0~
zoSg48?g9{s4XDM-e5FxmDRr#(UCvY>3pIk|iqkho%_UZCZ$ate<@Zb2ThxMC>4kIK
zwssHM53P*m`ufZ|zKs+a$R<k;&+E{uc42w9kMS8W4QtT_hjwfot?2_*@pL4EEqCrj
zn+gxB^&NHkVk3x1;kZap9g3^Qdy!rQwe_6uKZ!2v1W=a4Q~g53K&oRziCH|)pupG9
zJs9VFk>nwP-%!q|?s-i=iQO@{?y|Dm>Z|rQ%w%)c0|(j5E!TP|9fUNGno~v<+Mo5&
zlP?=~T<Wh9JE)>3$<D^9r22*{j7;Hc3zw!2yV|&arVeqR0zJ)!KE}0!htt-%B`r=H
zd<(e-S7DBlmR!n@)O%c4b)9bTi@}*RlvozJHTD)j?&OHsX&SY}vTiiMJrQ8E0m}|f
zT~2Dkyk-~+dam5bJoMzhb6uYo?Z*+mqH~2}sF+%6%Oz-;k2EU}DQ8F|E@ZdfdKzY8
z8fFJOSRIS1(sV>_&5k9k`Q(W@a@(dHE-vxyak3lHyxgc3uEX8(aFeN$tV-sUUnU|8
zCuF)BH~J~v71Cc?mFYAxE1&u&&RG_n)B!PF6*4cTFcu`T)8E_;hZt9COrd<19FCIj
z?cVbbO++4yzax|lbhMw6gI1V*^f>=yaL&GIZZ*x4aGz}nj4Q#2aZQV+M>2hycPde6
zR(;Peawih2Et`GFwG4R4=&n>??gXHWy!0-qVQ!cw`tDzWk0w>T6s46<Sk1g$4V2Si
zQgJU+^H<pV>3GQAt4B~_9-sOMl^XLo&hKR0j9eE>aQRP1#~&Gg*o;<n^J{2WRSj_5
zZkG#d@cMdnet|EWERg@#(km<Lf9e9m|EddWV>XBEFkLIEui*KJ6bjIC<oHL#u#?tU
zZLr78r`L*j?LR;Cizb?ss%^<qaM6m%QoOV=g_WUrfq%L$o!anrTzxz1@OsIcr`Ptr
zJ^<?TkLb)wI*<A+ltd=xXZCKomd%=q3$zMZu-h^}Z;qxz$pm#={GwL&a$fi%>?&rF
z+gSD{W#S6NTOn8oQAQ{y?a*ZN$|+o9J%jd5h4~Tlbyxxd6xEuH{2czl<PIckIw=`P
zuqX|#{1iy20t(|tzENDLx2fV}vQ&&`$xq)I8L`X7a?wqSd5jRMXWKvJgz^W`vuh<y
zRz^ZA*SpaYxp0VponY*#5qAU=32x`$rr{z9{+(+MItTcuPcq5Hp!KKgg{=Y<wM^1q
z0rd50*w?X}1Cbv?Ns7kw98L<dK(&(Ebo-e9^6=jIT9>}=OP$fjzlFkw`T5=)0tHf0
zNhB->VE+7O_1gJblDxv0rzZgw)STD^0j@F_0}xOkT{WhiiSC?;xR{-OYu2#tR5jb>
z#96vAUz(wnJ-bTQ5987VVwRoORB##6MFj6c5mDuKfwCi<Cz;4p!T3zPZbFk=&T(4y
zU_eZEUQ-r}vugz%V(lDL__o4cYH?GZ(Qf8FEBiBRd~*19*i~1VO1MTxkKW%eIUO1g
zh%7QNF8RcKnsiBp5tE*=Wf}odM|jiT!jk;CD=aVOvSh&MGC%WLV*rc-x9(ZTVq0T~
zJ#n)hD>b$q_Tq-D0;AHWRO(q7?|I1sg)2^Ly9(1o#37sQv(;R&#QcG7lD*7IoaT@O
zOQJjH-w5W4_i+~VB>gRN925}+S;Qb7)dM?Zh&$w1-kv2=U1gVPbqG9c<D$U;7xtmC
zhg49;{1cvU=%nGRS^qBOUB<Ofm-#v^&C(U{)s{`Op8J<gr+%vTN#&Z=Y_hGj-GKtn
za=BK%Ps$Nv#j7%~pW8fZF_$qsmtnkLRE7}Q@02%zEbiP?OUU$$#`A%Aepd&Ekg=$4
zEp`4nOA=pbD~CBbck%|!)PPy1YpJN-E;HOOk(+1IR{@AswtLz<V4a-4S`?VhGy@xm
zT(a5spiY7Cbp%|5Dz=ay;sL+`e|_e0#)-pdeqxu1^zs1_7!Z1m4T@YXzO+qe;c*$$
zixC6#p3F9?4aOL!Ef3{_FE<I4f+>C|Yr5)nP!VlSL+Au~*<+LO0v-8IB$s{^#-<j$
zs7M2P?&aYy9av#Pv2x@H5(T4O<-DDbC~m<3ZsgG@N=C8Q5|;+m=Nv!cJR8`_*aN}+
zsI!NGXWtxJ>vi!lR?+*j*v`=-=-SEeWS*!<#8+?qH@($CDn-x({M^^7@)zm@1JA4y
zdxvv!Lx(q~zG2Vgc<HpyJ;4AtxVZOTIE`tL^-(#E*@H=-!&b1fh0zBF8ZTK@5ABVZ
zncxrFj{e6w;+REwEMtSyQy3{$`Z12-zInMU(kO2#VX!E7Wgk(r34f_r*<if}nW|)5
zkzc1+Yo#Mvw}Bozt!N?W*2tM~$k7`9X+S`$?>Gy3VNivIQ4h2LGtvkkeat-{K?swZ
zjTI928+c;O@ajL;It~uD|D((Fe*>>D&@=!4<P{cr*8gb~ZA!!17G(_ad%6aLLYFqK
zx}@TwsG|TG$u$TXsRSQi0cis7vmQ)$lD7W$z^0p1;<22dU1L4$#?KG55?Rqv(NUi5
zQ68#6<5E|#R7zxmu8=aRwHs~e%+G1;kmqD_E^y^wHNI|q$F`Oe4!tf<eS}DS5f&da
zfr)e+ti&lHuT>s9xI%%ba93Bfw!54NLh%vSj|=%0Mi|=|)r+Oz9TmW$v<^|3SM+<A
z{ClBqO^0fEcxtIs8Mh>==P|x?Rgv1J!aXcjCs}ueHS>?$AkIQt;<5t;!n%aT_1~DT
z{=%!aOPN4FUn+?K8lnXu|D*KZ!N~N+e=u_n$P8se@1w%Or33Vp=FS))O^omZwQA3f
z!C^~KO}P5i`PSOpOw*gfH~=Ngm;dnJ&-kTq5173f0z8%1wb*r<xhHdR(Je0G`+nv_
zidMxZr?l9W!qTHswe8-e*#s-40l$%l%Uabq=hVDdtTa(ZuJfG@jOmdEpkO8cw!j_R
zozs_}5fUbqst<<=z!Rc3DaH_@!{)<4n}#jnsvy*GQ_B%F$40WNHmz<`4xj9$VnC5J
zl_7{|G-1N6^pRL6#B)kgfnHB)U*$WI7IDDt*(R_5>+_mee+EdXgvm$@Fv>Ft#J3$?
zcS-D+3{<rbN>jBZx;L+#RFnh{j%*=4ujuQx8sHx|YAYlq43(T5pg)%gp8~^C(YL`C
zC`HxIa#y8ps=_3;9ElvHrO893P^@1q!1FoX`TW*j=1ZGFpZI<{+MCbO$(;V#dg7y<
zjxjy)jK;R_`LX=z{4tLS2~}+<IQpXur6xBi;7?aOfe-O#42}8R=cjPsvgRWP-4^X@
z;hKw4X4QwX&4H1Yo&Np6@Y6jNF%8($Z#b8R#sMF`1x2vwGb$253sj0(!(R>-%yq{E
z6baxAKP|!tG?MTe&ql}qF3inE6M;i(7cCTI0x+)2SP;!Gx5z;2-&am~$E`>Z8vrjc
z0nzOz*<ZZY!aVw|8C=;HfU%G@-KU?|lOBYgRfKG&d=DRYmuBAxm5^i)l0cIB;<YC}
znrXC?8a^DHDb8m=?Vo$Sw#=-V4rcD}qlNpEm5IQk%7VS2h9R$iEZKFyg8!_OOyOjT
zikE;9`Xr@1Z!hVdBw+pMyq*9uOZqFo-9pWuutPT#%%1>uT}Sa2*R3iT)+|*qpWi0u
zA^@7M`YI@(;`z9JHhZ{rak{c;Ufxq=bDnHXa*{tRX}#TVT|yL7=}aGvf>T7y6+6?l
zscC0oOwa|J9#wlLqh;ZJw7e{;S@D&tA@WkdPcYW<AZn639R4290jAm9&!z(b@nX=c
zC?jGkqL6?Iuw>@`qnG41mf=@HINt%0t_DQl<44}Zj+a{F#J&`YmYG!|PL!!*S_<l#
zK$*$}8lm8l$nGXb%~=tG18Jb}EQuE=rtk!}h0!Qk6-RW-D9Ka+q!s_BEUnG1!anPc
zTf{!&MHOEMIg(`TJXZ;+vC2%`6GUuPV>l)_dciUc>HMixnA_cS*MvRfj;ArIRZT*2
z*eXXs$8c;<1zs312`*(U60kCZMeUCaY(dUhS<ELWVH9RUEll3b<LHM{L;5E@2O^U!
z?&#1r*B1_%o~r1Woj-<HNh`g|#hL7vZPA|zNq@!d9bn=Fk;+K6KANi-y-*i$vpJh&
zQca-#b`E2JjJWJ?STJcz*w3kwdm|oy4qJHu$0?r`FbjM)q@Sk)oSG;3N!{-j+P{KN
zI)*w3vsH<=J3FZ(&Mm<x!71@<r?jsXkcX^VSkh9yp9=y|i_pRVS=FU%7bsz&$F$9H
z?I#e}>Tv@ffDzuYY)OBQP!*JPRSiX6oG(1PVJwnDYt_*JXzfz!Bk!Mcy3Q3{OL5<f
z*rsI6gy^&qMpe)%NW|!`Ql`5V6i&*8+DW&u4FF`Md_#huK2)P3v6hE={uQK)1$qbl
zhX}ljhDbudTUhR}aLfSEeKQZz&)~^2v?35eWTnj;&IpK|NMyhmG0W$uRfQq+qM&2-
zy?9uS78>B$qpedtDF~?l-V*=BI%_5?wH+=w!EC)tKGqx?R5i@Y7Ui%$B*0v##i$@r
zwCvI^v?^Z`W-b_6doK#ZPws4EKk7x}Jd0KVCsGt-EZK^yUlNrWObGQAhA~dq#SGXF
zt&PhhZP)u#M3tTB`D5KjUy20fiAxoZ+Dot;d2jObomDvLTy-L&tlI#{wf;c5iR$^_
z5u=V?JPA7etb@5G*1ryM%6PmG(%WxP{NiJ3#hKFk;aw;k@#x~k7d!TW&%-$@-ulrO
zTdxsL51ZW;o1&A=#C(02I!ZT0z1%XDMXkNvq9<O2m5eAOA#Q4H^w?;S>(>!0(wV?0
z@w7KeW_E$0wG<z8+0yKz7#?623AYD!N>rWOfztOrd!ZzgIDhT1?bq@BWD?xd%Ys3`
zriW*$Fs1yd$X{)D0N09Uiol&od#XtO%>h1!&d*Cjihm2MRe>V94Rs5O3tWHL9`T!i
zp@A7S8wAGla3!K908R7{REWC|uN}CX%#Y`$nPmXt=QL3cc9rkX)~mcAIoa)WY83Nw
zKE5s&1Y{4;T}@pXm??m`T)ZDCROVEpcakh7PrT8+Z0c~>OseaxJn+bwuxVp}K4|Jy
zl?pNPeRNopUkH=xx<4XJwzZFBBV#|6e%nCU63bi;kXK#c693<L-=a!?Lg&&afBtV&
zV~z>f&cVMD5<-`xs7~<^nO=d2Y&F>O+NYEd1t23zcT}i_dH7IW*fdoEH+<`>B0ol%
zU01kZzoZn$pO3AMqx*rJkMDy^dYrH4Gj1H3?H7digCTkNfj?V2^)3VnZm*}{<f7_O
z-)oxnAncz$#oDdbczhu*B@YWfnbR@-`=3=C&tI6~S3~tVQ!ABLB-&?=Srh5QS6LIP
zwxZYCk*d12tl@QIDX0fHE8%s~B#I;TR)*@$b<`WQ)lR}dP>$tTpNdTG?M`B;dd3~G
z1bmd~;7_tElWG1qx`|YW5LsV8>HxK0OdLKc#FF#jAzhn%gge=^JKBGmwceiYzd~<f
zOnaVsaHw~_t_JQoa9n%>av^)NpF&L|2EHLiJjB}6(6Y|{;btd#<u|yoGjZ`sfO|Og
zC!e<Vu=AbEo{WS6Q~UVe`MSmR_|8EL7KA281Z36!;9=+yIdb`4c0R}+j4oKG#)nn=
z`!55WM-BrnS%L>MaW&yZ0kGVCtZG{8F~_|Y<o4hP=y8hU9s0z>#K(%fBm{61fBQQ8
zj2(5-aC+p-v$!#3Emm2VYU^i8$t#Wne(j?u16l+qNGk*<n0FTb&{^V`ivQ=P@wafw
zPXC_=WL74Q|8Zj|9kaoT@QaO^#ZwExWBLR@16$r^|7#2Ttn#SD%yTsR`6C#*7)wZ@
z9^WM~8%V857-~Dp7K1bI@M;=6kW%vSu~1Us-v=2pezy0uk$W0kC`)pdH`GT#QYk<1
z@v;|#PSI3!A%!kx4x=62?b&{Cwni>Xl6WMUm5p1PQdZf0$Otx~!FjmurMekV5<ks_
zc_nWNb@26g@?k3*pNF2g4$-Jz3kQ-n#tsP&u;pJ#s6Xg7qG*9zhT4c2;fcA^&yhTN
z>W7d!ebfqaYnvijRAR-l7B@t=6f9TCU#2g*F^8tLMw(suPiH(q#(Ch8*C3k4Zz7qX
zmMmOe1SSQtsx!ISsb1T#Yhff{Sk~HE#rCaD>*6opKg~F3Cb(Qw34^_JlD!BLKJ6|I
zqc(W~z-Jc8D(pJ_PWt5;3vDc&x6KTvA-3XYhf9X%*(yUAehkx(<w(t;(jruhelU>)
z5s`|?RQhCqlsXfcqIs?Q7z(zb#dN&&0)Q${q1WUCvqAisg7VRca^J;ZpruJHCP645
z(@*?h{tKEd>sXIh{`I=g<BODyubegEf@~dzYaISa9^qzNW{GgUgB7<6UvD4ZcUKub
z%%JB1-JP6$M>+CJ=)ihH46e^t&)4TiPuoV7`ygG7h}C^EQvAB4NdmNBEmc2zoOpwx
zxnmtrMs;2V+ut*xO8aGIt;ADRd5fEKmr%AOK>=)3!P2ofRt7%0_XG2LyX~OS8F~C1
zv)iwWPuiVyvU2al-a$aG%wUa?+B*K%<w6SBlS3L>hm$E=9{#FyG%kGUmR32ZoW^b$
z-2k$4gyKEegwp&NX>nqwZZaZGV7Q$JJ@)Xkb;!}O^V>gLRl5%}%6qU<`3Q0IAo3f`
ztC-rqKa-eiT+UE|tTjdppUNt^lYHoPo)7nA0V*UsFHaNmLkBR?CSfn02X_EqlkKjL
z-wb0_K_Vu31P~VAW~Z;__KYpjaAwHX6x%5dc}URU5SxpFY*3+9{E4Cp7f)dY%23W{
zGBZN0u%Dl^+KiG6r4HhVCZJOn$@6Wl4{D6+Vcv*zaf%+k`rf|0B>!38np<P*AaJ#^
zv=4cMiz^{lWcGAHOVZFO=RzjiN)zQG&4<@p#5xsU{U>ipR~g*>*xaq!@7;|@bQY;-
z6M{%6(>}?#oGV`RE^)6Q4Q3W{W~lYGD0`ka>69RrzQ=}-!*6w)ShwfjX#@M`@5S;i
zSOu4>4{Vt$!DOZlv~4xtpWHeUQlu(QrBRgoiPO{%>mvAi3^O?+Icx+teFg=vv&lGS
z8=60T2h00taa=JzhzYLx)JiszHMBE$dv(PyM$u!mhU1mClD>(3f7@aLu&wd>wZ?Ou
zYegAbn%qDfH6t`9uejfZC%1y@=vyudmSAkXzGnjjAZbhuD>vIw6xWe6&n>+Z#~7+u
zeSin5f{tFjaSqbix9VaNka+3n*)5ZwM`1CrHspq#+;d*m{C4VE+Ug=y))&d0xFor=
zCnJfFUd{j=D-*YZwWw8eU92}pd#)Bes&V~}e`YxzJ!m{(LUQ#sqf1@oCwO<V)*7C5
z#Ku88F4Zj&#jI8mPQJT|O19#pM1Pv-4*g9K-Vxon%bNeT)m&&=vX$J37jp@Epg!yD
zM`mU?p3t5dB3=NLUsxU38|0!!kGhvga<Tuc+GL8z@<e{vC3-VWm`OEdZAaVueuxc$
znIrJUB6N^-EB3^v_oT8g4JB0jl~w5e@oODkJ8WRsT{Zw_lylw_X1?e8Kpdn>tta$(
zIRK4NPA{S|0HjJ0agu9w(ZB$iG$x8~K@c++`_bvg>whk1kX(<<w`ji*_qz|ch3Ay^
zzZ$bvowl#DJ*5ayo-UkFiVo%~T<IPLhEqojqc3|YcNHglR3c;)FGQM^&j!zBI*N;m
zop}nw9ViEtcQ|Ex<zF|Z#R`<Vamv@%V^FTjJ8PoJ2~c7hr%Ny^kz>IRnie-Px}A>e
zW(_!>|9C3hp9%AFcSZXK)DF|nS&M<=C@Qo18nfx5xAU@TVA8ztyWYbi=~Aw=3c=aw
zs7|K}2^?M$dxmeAPaiCZoxcN+IZsUq<~Y3T62WTQ;n9_c6K-Y1gi#)4HhOTXC}oT}
zbAkzFZvpA1ls!lrBOQ6{s!4RsIrqLmS&?}Vm;V47m5D(8SJmS9FRI1P@;|0ZX-L{`
zvLbZ<sulykezG83D6t;It<5Ds7t3wcX&y~_rh}Qx+Z%3pA+bV?cGa>AkXnWIzsx<a
zhr`+WxB1cIgS}U;*C%m9!(b*o<%iEUKp%oPWMV$z1vLPEtn`zbDZCd2%KeK`bcr$+
z-Q*sw?u=)dzyrj;X93UCgq!e&OB6A4q|*w#GcvAXI5@vAfBbJu8(>eU;>XE@EqGco
zZE(pg%jUUm%uMJHTo=9sv|u-ZA*h(y_nQX<9YQ4j1$I#oxK;Pph8ORrj^b2jxhHrA
zaELS#rebfg7{&n>Ax&cwChL@dek%r)SGcPoA@k{i^S&tzKs$tGYbe7f*_5G|NIh)W
zxFNBr5)LcrTpdTznbh=ZSjJx+{hzxlNkyKw$GT7{Ofm#%8W1lC)x2OIu((?h*}-Fy
ziIV5&!hPLg_yAWbnEniqu#HX$SPiuT)=GtjdelCnq>2kDmkgAIr2f+h#aPPIe(u|a
zgIwBV;t4KtnsHWB&4}-mDHZj+zb}?7H2dEB#lRyfWuF<t8t^@VJ!>DYRMmC93K&M7
z=*0gp8c|LItx%(!&}!*v)iY>LT^w5OToai&MubRRtoP=Y8Lev%gW5i8`-+nhmV-L6
zDpd#UsiqC+bQU5W1=@J2NgGa+2@5Qs8n!ViVemz@6~U=WKzg)dlm)<4&v>QobN9ze
zP<lGGB6Nr2c3r!|d5NF+VCkt!fgFx!y|FP)$cR=z1qCem_8hQgp~2>a`84g+BRixZ
zm84_Ogp!CqRwxl$0#!5QONXqk62DqspvTeB&A>P*j|!k~d7)V~^MyMAFNHA}Ye|G1
zH*vK6DcFnU<P{uP>jRF$EK2|I^jQG^6wb!u_i}sEapWEZeX;m^b16Y}-Fgsm?|PLp
zlk4RZq#^DTb9y;hMUWW=)jwfO2Tn^fz}sm<3$%D)&{zq&aadks1vS`7I?_>GGjQE&
zk(8(J$8Q-o)V$tQK3CWCPb0u6blD(B0XzoK3=7nxqe?bcw{l!FNe`|4>v|>KV^4Do
zwaz%WZyIu1b$4Qk8kam5bHI<AO|lw;^!OYu-JBi*YG>kQ%~@R&Z#)>WvX6e1$%bV)
z*T0@BiHQhSqh*>+GABeeAU%*31Xa}ZS^q;A%AoZAjCC!!kjJRKY&Tmhu56$xz5d8s
z8@fa($1&g;pb;Bd((STrg-Mz?c$|37$u<~R<&5i41zD(TjvTJh;Hit9l+1tSG+$P~
zvnLeD+Eg;PG}fr~vA)bSRO{V8ZTlNE$hG$_ZM9wLv~Q|(dtTM-g5$n+NsFc$svugS
zq4g22sc_{!U^EreMj{O2Jy^NAY1c0a1X31a9fP8eh~J4mpRCaHap9dV01Ii`MD-~v
ziNcv$nArc8s<P8`v(jn?aIC=imTjWO#W%okVlhs4A%!B_=P<l;qxv#xWeioS;%*mO
z*x0|Q0V;k6y$ro>m)9p*i*rX#mR-QC+{V`9vr7n`MPrg94@zdI-^M<dESRsDQVJA_
zYNkDQ^g${07iVNl#XuMuo7x9N_8D32HetM2P>yrTGkW6B9@s>*=gGG<xxLrDYjkoX
zkSIZS=Ct$X;hDr$ZYH`}ZnE<)SAMZ)v6exg-h*nIV~jctt3r{oeR)x&VdfPI1J&;r
zwIt%VT49V<yY!9ecBnBV%U)Srm?_=v<sxnx(GXEffB9++mF3YWTp*uz;(?kejnjKn
z{A+`-xk-#W>G0x)x)wb)7H!-=N6KP1=n@?>DX7&s7tn7|Q$qScJWrQUupf{9%6IsZ
zELfc#@x!oEP62FC5Ht!Sj4|ZT7wP%|-jZ#w-_gVdT=MqSgF20Hv6?JgV6_TyO3YIv
zzCaZbZ(`Kp$S!K35ZM#Vw}>F#-@9k6ydR`nl6=w!<Uqd#Yc5~me=rQKng6GpoTamY
zem%IX<>G2U6~HG>reY7XvTM7VUQT{}I;fI3?PO3f_rX<%+Q*c&n*Md+VsOz^?P5^T
z$F;5v#c%fRj#UTMnkeWjX-P-1or?#cA1GB^+-dZ+_BlTHZl3~9ZXWzxUfjM5MEN-a
zGH(QHJB#9y?kYsD{*qQqU!qh&h7mn|7gJ{a)`;<sjrniuwgC+7V6t`OOWrqlFIr&o
zk<(UVI1S|?g#L_S#Qi%tF)lh;`>YFXpF-&viMo4vwP{OSnYq)$@U@y`-I%BIMfc@K
zXoUrzRe-$X)$|)&C#uNagi~`tFB?|}1i;`!L<?U>E08%YJ<ObJBy$(!{0Fq3oCZ=m
zHW~nnu&;BQraY(OwMxNDO2_XLAK6TOKa0{lj8J8DS<AG=C&c8K@u@vZoR9x-t4^)U
z(+183WajMU{7hXilnlv_F;Z~Y>}fOMoCR}qPyM6&8@E)=eg{hih@^AYt7;f7hGBmx
zw!4e3zb-}Q{#MJjnonii`dHb*>FfO~Ht;t@f6D%w9y9!!+UfCa4J@Fzx&J?~I~()=
z2s_f0v?cycu)k29v`6<ht5+KW$A`(ql!Sh5#@@2RGazF{GJ?M&neLIjTavEdDX{m$
z5#d?bGOGLYd2SdNU!p5}|9G}{|9m@DG5r6U-Gdi4=*WiPiGP&iZLv7ji-P9PAD^Ed
zzY-GlULE(~e?{4s+Sk>q=%GL*dPk1bckyj5-vetW*wS58(E_x)aauKrgxzO>lS<Gx
zWiyBO=k<9ja8fX(KjW4~>s^=57XHTr0|k&~^ff_V1AyxBVI}ky02R{ktr;;{rswPG
zabn=4dx5aLL!^RZNF`wyfkg$FC=BPNJh-`8P5R-L!D|)iHNAR}^TN*gty}tKpXob#
zWsy`3L)$~-9tTE(?u?Z$=^kl{5e7hAQ)LtZLa7=C-$^KzQy_;={ck52ASHy12-@ft
ztgp|<)8f|Ky?Np(Nkw><(3ABmGcC3z`(Ea?CP>ubMJH=1NcuhD+gyPCZJ$RX4K|9A
zFzO%GLCn0|gr;&ZDBx<#TsAQ96(e(Mn=We0sv`U{xH4(BiY=&(0@scGKjA5B1)cao
z^9@Fu^hg1a*=<+GneeCm2`&;wz?0I4!j%&U{MA`aP9k38^G>mDVB|g0Jm&p34@7Dn
z<Bl)!!(hw2N3LWIKUPk)4cviMnb_H)LJ)WxkC@K9R1YZj_-1cY75)jPxP5w6@UiG%
zWpgJ3Gd;eO)^5kYYqDi5G)xP@khM}~X5G{dUv+q0FPk-$iD0;Mf2zzlpF@KCi62ql
zzkF92Y$#K?Ife?8xy<p#^Qn4DEh;MN5jisMK{ikBrpPqC^>{<fN2676e>ZzAE795J
z8=jdGVc^zXdf!3)b<ohWw^}s}>;pm5BX3ec$N+9B?k>FRl#rlIbM`fv;ULg>m8p>7
zI{it+hCJiQM$vwpH|4S|8WTqyyNirTkqQ$c+|ju~g2_t-s*rOQW)|XD`BM6k$hTip
zHGu`z1df621IIKGH;ebo3H#vN9%l|RV`1NGsDq2uwbvNV%mhpH2&M!(Sn+<DEt5k}
z^fTC!{W2ke4>o8Dm$L>bkyCGs%(!!-PH)^Olr@oYTt>0bj`QeOWMaZr0fDQ{NTM_<
z=umvARF;n+d~xB~tm$beq_4wOa~E3_?4=MWcx0DZh{0;JXgiLn3|O$>#tVo;HmiEJ
zkkNY!Su^M4W;<MxhMx7m;+K6}ed`fv(9za%vT^j}?I_eRrqMycdbO}e^H1a<R#yw?
zq|-R+?8n99KCKok0yte|U@ibU&AsSMmo(pPhRu#YJIF4Rzjm1Jk)qZK{hf)PjqulI
zH)3>0#Twuuzf%iXB2Q(Y!$Gde_VM>a1)s9}i|1V7iR;E=-oIJEJQ`nG2p?S`@DbH=
z{e3G+$$oW^15rwX|MHuytn)sk2ldYQ8c!}Xp3GY((2U^71P8eWR1e(qiEU?iet&?t
zC0J(g4d$d3c%@D&5wNA1POUz}{E_*suY*#MIRshH1M)$bqhJ{=gd<aVh9(K|KBl#!
z%niz?PYVI_xnatAb$PX>1r|psHiHHE=&@dkSXJVmg!DX<Keoi%zyzAWC1Y!=ZmO^h
z)QnSNyXK+0wWYa#NG9#wj%SywUTi>^gv;pTBrHHBx>#LR;%RERyy=L5g0T-6<1V$!
z&;$q&E(dMsUL;({;_97soOZnc<*})N)|Do<+K&+m2DYZmi)dHHyIK|s(Ihr>#yvN|
z@&^6AV=Qd~K_#b2MghcRH>wTNP4pE=!<m}!iMMqMOeysf&x<Ni-51vLE$LPB=}v=*
zEA2H|HhvokO;(5v<(<Vwvno0>F^2h^*po}D4GP{<T~;#l_oUb2R?AbWB1N^CvyBE@
z^+SB6YK#^N3$K45&JJ9bU--rsbl>(|{dL@2-(9_)F2oxWXD@R~_#CstG0g;<jY=|<
zuwDUv;YGtT*`iI*!|ADMN&6@R`7u}l;|kLk3aMt66`B+|5g3-H5t)pMc-9-ovUn@d
z0~mo_Hg_q?+Ku-DOe#3DBq%}3sez7YnsUMgdZgZBsKdxNoD_UjE0%K%$Ll_<ptM;B
zzRio>Fu3VK4H(cbyU^J>Mlr(=P~^jiPR-_Ddk^J6;cEDAGI|tQXe3y6cEJ)U4(XHU
zYZm|XLfELrwX)}nxQ9->MpEv-Zix&1^%p!%Kpw;^xM3$Ka`xiTE8&2uA4bmVgEBX@
z)#c1g$kKCL=C#wrPxgW(T2vw#oAJc-aU|!A+=daiu%wkQqROVIluceZuWZ(Mm@?)>
z@QEym4qu*PSvuqCn-|KV)dxwHo=$NXKL$>JNVqZo&}_6cjDBFOcDp6^nlC)Yx9rC?
z{ta#!HeB)m&PL^@HzKn`Z=bT_tg$BMrw7>=nx{mkp=LzdT1w&~c!Bf2;bKbw5SX*y
zl&ViNJ1Zx*pbT<YYF^{~2Qp-5r0+XpwU-EV5&!E$h!v_S3<O#Uf2LW>gap|+7rC2m
zdd`QZmSt=<po48T98d=|VVuTJPMly4LPwIA5ifmyS*yP<qZCU!cMG8z<VirEz4~MO
zEWxCz&|^}|&#5zq`-s81L1{a3Gc({8V(nc-r)GOnA}s1GRjAEfgrT&}6xG-@eapN@
zReeOYL`VYc6y&LDiTQEwkZFm0CN<dy%Y@Wipu)fan*pUg!AEI9f(**s)wowXAV7s0
zZ9G?qYTZakkQVilK6dFGji4zk7M)I@VFwJHW}!(hI)d_AmENc^oHNcU*{}wEk4z)j
z$Mssf-k+QI*u`$^e^p85Ej_J5UFba(0msINWFZo;K~xu>GHtWp(Uc$nU9kUF`d8XR
z;{`dS{AjU*i^l(NKxNu=nRA-P!@Ty?xRrAkxcESAlq9k(?vB=)@Fe-99tVFK<uKmd
zerGdsB=li5r^uXrI=cmz_I~thcn1c!;pK$iD2ZVwD|5#tu6@_xWM;?cVq#`)@r5?)
zED`?*mFXW0&$Y1%l)LiTjWr~!;e$<1+f7*4bJ)D<#rB~l6j3UyO85l$4E1)-bd*Y)
z(D{KTqlP^jo^dyDQ`YM7xT*aK|CVN9j_bb2i8R8ze$R&Jm$ol3EIx}J`xxa<eryg0
z)0%Pg3U)I&gtYFbfE_GoxSmh54*l#^P|YV!n#}y);1!5;d21Ax*#5C*@m1<%LlNQ0
zTj)?*l2u0Ua*O%|`lCdV`I(!xZ18clrL+isBqh0G{uxXpv_dglz#b_uEwv9t=jVZ>
z_Dz&O>2kU}Rh>RAYG-OiRibPvQfAGN8It@fhu~ldPwTYoB1Tr4uurpZoX(Y@?X3+}
zofa314<t&>lr?TjRK{IFP{ap5atntg&d><(VTvdG(5bbA&kty72u6Yoy*Q{w?tg;I
zR9oI>hRZBbh7Iuwu}O53frn`qh2F8m9vo8KAJ5Y*bnztzF7q%K5Jn_7bG{J2kIFDt
z0V2)iBzwc;4*q(v?Ta)Krz1oxngntSOND)bDsUH~PrfRg3T^L=7gXL7(x8$}%-)-H
zt!e_;3`zPfMiH>tWt3GabTb*PB29+GVM<9eWLLjQzZ;9rT1V5s->`M+u*fUgDva9n
zuz=?1mshV73ii~}v3t?oS5h<V@{_$s|6na!)w!;uw1oZu!D|@L{?~0V!+%bpGqTbD
zZ_Sh?O>tYpU*OY+Dmfdrc*7Va8u}ojUPTq)F)Qp2V;iy9j5%D|!r@LyEAMXxvJoE}
zqo47aJhA168npP>03Epd>)!R-)Aa-9Qy-M*c>mth2LEr8{14SwR0!NJ@ktI>$Cu~x
zOJ0Kh)8jZL87WRTnUBk-<bX^m+LJs%_xyFt;)(q!Sn(_{QN*9F;Gtt21?T0rG)2GT
zm@SL<bN7CbHCdVRGENt7r-B!`6wo&Q$t!s-vb3EZ<l=IQTQnyn%j8`ZtI4cL3PA&v
z<Lt3aEgCDyG%|>j5XH=!1B=sn3l?d^0$C}pusGNLb0UaUU+MK5Lq`t10Q*R3TFI49
z1R?_znwIcseYh4%hMSY>`JMe!*lbRUTj-uXJbdDaiD<NNLS0~Ge3Vkwg(RvuPsHKB
z{y6{BNWp_e*Qi|ZY7}ozI1t=LSzBz-JgR1rrJOryC=bI;%AoE?lg#39U=6mQ0evJI
zZ7fH&&f!Y>uvf42gnbTxp=qSTgsYrNk%|*O_HaUO4zvfe3l`)wD@Y#LdxV+8)|3~(
z(pr{PW)X-xIFBINkNhm?bA0qvD?XYhxBgV{-61_HsbAQg6x-78u@m#|?a42ZJo+Y|
z^_aLa&ryY|R8^Ny6G3T8gmmmvvM*fU))rN68d))m23+K-tb)-dEDK9MN^g4#BCg8g
z46xz|rXjh5+d^E`;ExjBp)@lJJtn$|$`JAJuNu&knN&$uh9bsnHcR>Ffl_X5$>bz4
zbG}&<bjqgb;Ne@7HNPo>NG^Q$-#Q{+gAHpRP$?+kQ-ROV#Q>SHl<)}@&sFf9jffnN
z7m(%658C!7j};+KfGFlrXPBLxL@&7of2(nHP|Im4UVl_)yjq3E2QyIf&;xm^r|jVj
zHA*j!4b+gnj+$$QTqK6c1}$`x00cXQ-MflFSm)>m|3<P14QLep*zZD!6K@>3TZMx)
z<<N%uLEB*FpOvj{6xTdZXXsic8j4TmYP#T|tgBEy)F{}Dy6uCJt)p}E?a1bV)p+^x
zcH#Twf!;lv`GIlazK5B|4)ybVGtknJGi)x2Mot_f{v-Cs=i`XJ(Q8V)I)yr&Ii%Fp
zMFhc9`CbO2T)g^w-Ux$38EB~_CrRLw8IhJoH^V3arUQYmU>+m{Wy+FLz0`~Mfl_@x
z1HB@^ULbL8!;~!|mJFWb-F$o}->~@5%X!RvjRd=_8023L{tV{v-X;&`OMN#T<p!2M
zqOII>IaavA2d1OIz%(zt^0%-~NAj|eKEFBFG>Ai+hu$DzTHTA3D%<P~*cF+QMH;s$
z&nT4xan2N`v?h{Qx`Y0yLr)^k=!P<Ce)>q-t)RP#r=nD3=ybfgJlzLD_&;_RR4&5u
zoH!X<Q6;_w5}~Iq*S|`9dn7`zhdoUBU^?c=^4=$1h{5=oEuMrHsdLQAp5k84Eje}x
zO<l)YR>LoT%^8t|9F&d}T$80*0Ej(yYwS3kruEn6=+ih|PoC?>sq!bSrx^wua(=un
z!@brvp&qN!2ZLU@?fa;HinK#k@LhLD+S!1!P|s`!l~mP$jYwFRk>E_$hC)}SY=Bhn
z1N^SGcF3i2Y5|x#F352hu>~faFlUzIcKHiXt8pEK%3L$L{n28$CWW2;N-i+@3CoI*
zxK$0v7J~BRrZ|;($*Mi7Ot8C>J^V>T+08QKZH90xQf>tMQHQE}QL5w}p|1>w^R|#7
z$w91+49)Uz(~y2?tjOca{BJN1oAIx}ltq$#o-6a!0*C8c(-6|2f#d#D&BDSR4+-yQ
zof(<|I;-K+AF0wYjZfPhiS_QRRy{REO6#>93m&3=PslnztAD?L!1V$#jye6F3RD1H
z&H-2*jX$e56KhRH2_fJYddP)W=GXIE)urQV(wUA0qn8dM`fVqzw;C_hsBTkU(B{Qs
zE<&P1I;NKYRoQ8vxkl)p&(It0Ae1mOc8&8&Nx$37QqtL=<40@w)qPk|N4|$mcEf)Z
z@J2fy<d+kl3GlX5Hw;}If;0NsmOBe<X~uoaGv#dC%IfK|<HCOSU73I9J{O{TUdLKV
zkl4+Xo2Ab7xPKJnoc3Rf?zo4FZB4mS$h{2iAeQ8DEpje@o-ikU0>a;j?JBGGZK^ah
z&u(uAyX9Pr%^ofhGWYj?^B?nfqj03-xHf-uj^yj)is4DhOH`njK9?;!8N~!p<IaX$
zz(b?X$wXgG*!$M7vZGhQp-i8zeh0SHi?{-9|B&SH*vquB58fjz!M`mC!jB&X66k20
zwx;1mFn!u|2EnWq*><axw0aOnr#BEfTzMxtpp#fkEe<@T#f+tt(WA_R;30<|pax;F
zmr)EYqCl71l7FzQ;PlpP{=2n8cDd`4lQmU@0~^Q!oNvPlTY=v^47TMV0U4oY&&)y!
z=>qfKi?wr$0S}<Mqw}P>)gW)jfwft_wO|%ovi}EWyDDhv&CbGD-A%m+Tsf7gZ5M}U
zdZod7AJzHNor^mjn$~AL_<rc6+zr@ST~&V&>1`D;>+7*FjVeJ4&*D?7F+P1*nlgm+
zK-pynN7_5hEN$$2kJ}kS9>^5+_uqTT#p?k#qFWqn!yw96V`J}y*G5b9Y8O<E`57;1
zafDX{)ty;C>R*x=gc{7ZvyW80BWn_+$JC-tt%hH~eM&;IAQ%Rn(yRlWY|rRJ|IP0h
zVc_R@yH;14KSAo87>V{2J)8C%R+Ewq;33tY_N5=Ms+6E6i)tMjtzyaI1*5c$XPi*g
z_~7&o7YemzLq>sit@2%Q>^*PKQx$*3n;HbEISP!PDxM6Ij&xi#BsdWXN84^FZ(GpY
zbX-#Mjr?ZAgn8%X^q?p7zZFfz7x{?FRIYB0Hw(hd1t5S(DQcJ6zcRuPStLTT<QX-m
zSoHyN8xZw{wv>#yUxrQXmLxr3@mxL}RWQPmX)tDjqQ%yAl4&xmTe&bUQAgDx;jGSK
z1Gb0hB}tcXGITdDAAz4@F72xDVVO~F2Qfvn<k)mv>Uk6==%@R9+T7<LZZk0rD!qw9
zXT84cFz}_aRxZjX9hXnzG2S=Cl1_6`y8V!qos9)m-K4x+6_vD`a^yhplk5*|;IC8j
znM@$$F_InXm@U|vu1a~YexYKXQ1kC4v3lH6&<kkeGn(jgoXCsUluFP|rm0xs2`1Sl
zG5g?zTlf<0EtK$xjNt8K_0POD6EWGwgic+Hd3E;;&pFJ>4sJw!(Oq0>Ba>9|-xND@
z+d&d4b}hPZxYkSZtI!E2k7d>#yd3lK4g7Jy-qo`a<#+Zrp^aaB+uMv&vO|hAc`Ae1
zBVdsvNw!9;Y*|Qp2^Ndo>Ii93Bkazdl`E9yQ}0_B0wg+DtmKbQ@Ty63Vyv91Qv{ml
z()I`mm&H-g-eHw#zD6JGL@Fc0##LaXpItaI@x~Qbs}ZRT5?BuC8Mt)!+7;i*a-VF`
z@_tEiiRA7WPtC1anhltc&-%>fZYMs^xhVm>zd}^!Wb?uD1$p&O6E<m%ko=W%Mtz#k
z`E3_=c^bGvL8jd`I<Y`KmrmovDboVjUckoSgkGyiQjkh4j=h(0v2EkuS?s5w)`XXP
zj~703-6NL&y2oW?|Ibn{0~6E#G%(bp(H4!vhA{O)#hDmT<c310$J0A4M%0@YJY^*s
zV$2>3xbcr~N4VbYnWLj1x245%lv4-w0zX66SIX=C)6U2J)lQjS{kXB#!Q+h+^3(r7
zx9_fKE&(+FlcYGsc40;~Q!;P`lWtdwqLcdj=C)ckMuHuO07iTN(TtyasqQ|0ggnbc
zJjZlF8v#6&sH%Z`mJ6B3`~B%?1Y9N`oukNL{dG)Dk6b#LX%v(l8HnWMF6wy?lCH4m
zPL<|Pf|*Y%OKq^%jkFWn95^V=&3Cwi0E&v-qc5nA2`c0@f><&a0^NgEcwYs42Kpp{
ze(&c*5kR#-f?QIGCgDQdEUjb)au|~sLaPD+v2>8S;dxL}3F5#k9e2`NX%};ndOy-x
zc%NEW!E#?YD&Ad3EQOiOu_u<H0h89*CwHW#M0A&SrbQS^hB^On&WeeQG2Wi+WeNpn
zjhAJ$?Q$Kljmn&Ll|&B&DZq0aEvYcnXBtbDC(LA-h*!9rf2TgP`N%2in!nzFBZ-8E
zutP@D1Ca%QG>eWPz>7=_^dwR#UpkwiK==02I`L?}D$#LrAmh~ftn8_od1oP&iU1$G
zQCMSi|G>dbKAi(PAuf}mLVSct7g8h_lbtJa7(h+VjK;XUi4y9%%)>Qt$=b5Kfn>{s
zJi&}8(J+xZNZcNFy#zmVg-4@&$(E(l*omFq5UyY$3hs1pPRb(GGKQ&lf}*O}*Y5lC
zDrM#ZP?M_oksMwXdXDfN<*j|v=P#mA4%b^uzFz=i%@jY%``iw2uzw(9Bk5QJBOw7#
zC@NA7>G#A^SPoK*8?KlN1LQNg*w*O{kiv{UEHP*!3PGEAMq|EY8e%PoLJEm4cxtm0
zl`BfE&7&(|<w-5ns*SNdl^$O8It`SFxfs<DOh_wrs%3wrBtN50{5%bf3xq!Urj2~P
z?u9W22$l>gw3fqQb3_{j6HT~dyYFN`;{U_gI|YdnZ9$uD+qP}nwr$%sPTRI^+qP}n
z-KS4$`p(Nt%zr21R#fcgs;8Z?)|Z(yVuizPrPFg9Yu62xza;iNR#Q91@EoYr<A>}Q
z>1z96%)c_K(bGfK?HPR1)YSmSUUT>DOKmzV-3E_Gw;C@J`5+dV3B*m4on!QWBqz$s
zKc#IdMa7+_C0ImpP$<L7zi=O!e&t$!*NIw%l4MHft(AjvH!c4NSh?n0OCn*sE~X?V
zyVTjZHuA_>SQmwTlT2vWjxdcBw=m7jQVmVkAKyVJT|sT<GEUo>92PE#r~F*hiW$DD
zXsc?X<qnYYm07Ut3odLa<9V<Fq=Ldwb*l@Ff+@4pmK&AIAG}L#I1i^<)=g4H!)|p3
z#cg64hQVg^Y)%}lZzy4E3T9>@HX}<|DI{Ag(t1<EJ5vD$IN^3{Y(`(QdC#?b+zDd~
zT(I6#Wp5<1*M&xmA`$4Ih*gsI%@MdS-7|T&y}hJ-F}Yk%;sqgFu=xJw^JAa47vD_7
zn?smTVOjno^x@MlGHg&8%AL7?_F3=_*R9*p2_8QbHNI~{Z2UC!A=t&AZ*zUEoLU3L
zj%}4@tjPIxN;9D=Cg>_fns!Xn+9xjj(CMfHY8u4FGsEtWUuLWaJcSHghRU)EV6(p5
zQdnw+?Y?-qd_6p$ywHE3?h=Qdx7xz*feozKR^@^2HrFrm_R8wu`D09+FJ?JrCH-0C
z7fQXKLR-~%#aB4o3On@;!Yp+r2ylZGTh$JXcZw!-D{;KS1_{5_G0)ibGh+peX9_-}
zr%J$r3z(`6X}~5remR>E97w8N7ojsPy3n(2jk6#EG|1i!yg>#756V^<k3<pX`2=*P
zTfW^c180=oD(rr}P(t7><>@Rim>`9B`Q#SwrOa^v=XGzmWD<S52OEq2X%4Tj71+;-
zd|pAq)QyjNXAtIB_Pd(=?lxgkK;wi=N_qoDHO#^3X(1lskQRY2<PA4_t&nl=`c9K{
z@Y8hgc0Fhz6n&PU9R$7CZCy}v#MjtcJ_Vnstx#iBI`K$rKXE($$G)vz;b{NE3jSG-
z&?q#)m|^dV!NFCk>he<&9%F1XhfN>T^wRNjCGy>!EL!&QMvH;uI_;+p!LDtjdvK!5
zrL1cD4kIAwKY^E#FQjc8R&+zU)p^yu4(TsYSK)9#*jE`^Q)5t46lO0nH5x2Q<L>uB
z)mlfTKU>`bfu^e)r0jooK;Z$a6W+k44-<1I+ME`vsz(-W!W@qMZ?blA8tY-e+r|H;
zS?H)+QzQ(!o7X;BF9Nx?NoNEq$c`(ep!hGmrx9Axs8^m}OdeQAgRLur><x@(CF}h&
zwFXY7GQU7|UTo#lZgq93xH98zi?y_m(fT(kfHfpAsil<7!b6+8%3sKg#}Q4(-9nvd
z(!1~AoBpBFIC0Em@3PXI=swz*Z}lJ?QuCBrimYdu>SIy&nLq#3VnY{l%4?(*2W>sx
z)8d*aGz`JlET_a27v<UIjCBW&l?j}ajFK7{VCOEqhP=89RCiZiu7lrBn>~mpQw>bm
zLHVjUrpYVTK`PE2%G-?rDnmXp<++J%w!mqxauF~`&%L;Ns!CB)%}FM<k!>{#GbxVQ
z;}{MPb*XkXxh5(Gy8c&kbHw<|snjVA7%}Q7q*L<67;SxP7i35wA<*$@V0^c?gyGz-
zbky)4g5VkR8g#@?-s%aw#A7R?trR{_2{MJtXGLHIPE+$P05ko0?A{_cX4b|i;9%4E
z{Y(_^)|Cp|^c#{Wh2t{C^k)+`w+-sbTu`RG$D_IIR^BXx42)78uEE}kMJ8OIx7>^&
z^2x(B?fc(DCZ8~qRro7?Me!KCGhF2sEM5h+8NLx08Mc=<+{%>QiAv;!3iI?v^`Tpu
zwtD&*+~-DoTA45-z^H^DKo*VCfd8sde);A9?>d_CKLwfp_d_pcY2#w*L@#Dz=wd2j
zYHV*}%Et%g?BZl<Xba`BxuPZgAE@~k4^yy*ooC1f*Y7t9tVRYuuNl~l%#9=~!^!;d
zKff;lC)I2%v5gP5(Ihet!#^9ryt=h6>yn3`qldqn9nMkS4H&vTkp_Phnj*LGw2(am
zNG3hk>vi?#vx-rxHSf@{ChExE_I_uRw}<SG2GFMS@Vdgl3vckKG#)?J#JIeEgbfW&
zOPJ-xz2H^Pl=tuX{LYU$*8;`1KT`j=I+_~lgZ~nWPi`n^8l{|{|N6Qi>@{GR?3^d8
z1lacdDf@W57ng2%wCY8+!2@Bltzkr7pQkeRF{h<;svQ(-R7eGFq#1@V0Xs<)P4aQ3
zxZ~tMQtmzT_zl~Ro?Ze9fs{vFm=b1|ixe&VKKfXz^?ZC+hDT(g5L<(S#Xnrhr<1TM
zA%<nixIB(f@O6jtg2Xz`2imyG&v4KM#?Dc+X(+MavjdKNoc`0=KwS>mj%e)LiOFkJ
zCZcYBo@0Lg!d#6;8t4XwhAAsLh|)22$4VNhX5yn6Ckut7KW17+sF2!RyjgXsUjvqt
z-gm)Hz33ZjBm(T3u=Mz<8HsRCem&*@#t)S_+9ff-XZD3Zte@ON(Jv!KqB3esZCadO
zeZpybu3eRcz4!np{%DgvkHuEpBUDeT?wg7l8uI-SsWb?bH1RKrnsD{duA-zvUCAC*
zlYQtc*(?>;<o0bb^@~y}oumER_EA+dZQf%*S4NOdKCCffI@j;Jxzp(M+|wM&t!alX
z*9a4$NHZXPK%1aifMoN_30yS}$1Du-DE#hC65S_KLs2JSB!aBHc}W@#R1#)^_}snX
z8i0knK8#93ZMId$trGDv|1-Ljjzzv_zZJ_?Fit$Zv>pFIvQ#egjdD)PWgFo#KeI<<
z5qefqf>Klbjz0=te);)Nl)h8I7N&4>F2O;`1V(=_(e^lolYk6<>2Tac(i~}&23!hj
z6hmT@#;~3)KgtIW{MMg&kn>5x9oXorB*lFo;@!Z(=qby9x-g7|0c#*82%`X6!Nnq6
zzgbCnqFXTUTlYPaH{ByWmZ<AqWKeBgk7CRVgP64~5BnflG0qwGfF3AeP&cWDf0fuy
zQ5Lhgy3&$D8%lbBa*vS4TMNw9cKiBlE>Z{v8NTz0v-i<oIdos8AZ%R5?Mkzmf7ps%
zaD8B@({Gp^j1$`so+C$!T2T16<rfM~&Ydx5m{`<wa))owGAlVBt^P{*kjsx|D%U`f
z62S6d$;d#56J)S;f6n^lZ+>XCW;K53;*nB~lD15XS3`$~Y`AR$KuemT&_YST!MfQM
zY7f>(qcQ948#cX4lqq^XhfABI6>r_Ohl);<#64}n4mh-J<sxm~Z!Qgd<h&;DhRxZf
z<Ezx>$Gx*;+S|X@(oulZM#i|`fO{(q2zI%1<xNd?-0^WrV28@o;7s8qlc=BWtDmd~
zf=9S^QXddF?qhih!mM_Ea%w~7mMX!d9@hs6vFvtz;L4ZKJ<x|_a)x=>agM%|@J{bC
zChpKvc$Y-zNh>&K=ac**M)CpWKEUaE4Y}fxu6%rE=jk0C#KEJ<hj>4<WMEj@k=A(K
zVR0Ad$Bn*@do0_|HA>l;!&9+Ub{oEBT$^9DK|qu!CY7J)(0_wZJ!JLprZP$RBTq~N
ziJr8*oC^k*^Nvn9CW%e1A#j)Qnb8|)Wf72wjbz7gEX2ju!q;)gta;O>k`=DVJt<-&
zuzWj=h&{V$x)49VHdeJ@A*F{BE-Fx@KJyLR(4^+MOozu<LXn8gAtHFi+tehqPaK+B
zoNyygcn*w=^4JQpfxJs;q+Lpf2%<1oKIk;pzMp77)YKmD!mbuzF=e5~TXaUTXf6uE
zuB`Yez<E~-N=oOZJiI32T`{&yXhF#Eb%qe-Xe-LUY&A|{+78+SP8f!$?<#TxshM`G
zEMOT#yA-!O*o%VQXLQMVLmdrmMBcvn;1ko*na=_AYni0UtVU{BnqD&w#UM8;@(WHh
zxG^EpxiQhZME^t@kiEoowow%@I{`2<U$*qlb+iZ0qcdCHcc+|L(3Uvl`m@gbS)oge
zSa593a`&ieQEhyke8>*=6speNT~kf<8}WgCnfU4%D(X}a^T~(oR!?&53ioUw9pzb{
zH@TZOyaF{UAw{=pug3e^kw07FX)Kl@Z@H_p;kT>7=S0IPNWGqtLx9pk(pa|w34~B~
zZJtAbFJ1K`O$^b^f$ni5rZ8pdPFzm-H;E;vKM`sG$Ais=;?@F$HJR<#czgcEIuf=o
zrr2QWub*@9YNlX!*W6gU4C@U23ve4-9v?Ykk<Q9?5m&*&jsREw9x+zJQ~qa!)qAJB
z{`6D4{bG&Y^2{ZOs|sD>xVRLM^2|Q!SPm!jwzPXo?Xf+^ZKjn_d^#OlEss05ud_+>
zxD<)bND*fgn>w<9TB-b8M43gL6~p8<@?BSH0{H-+53GgTvD=C}e;~D_@*4QViZ^W=
z!yenlHgm2LI7PmJ<)$^hAGi|y1^?<Hs!78IXfaLQbJ4C93*B=p6}y|K3Wwpq<djPw
zV`iU3DD-{5lAU!<Pbp{11YRWT?|!CdR+^90mladJ!w~-LM^foMtu3WIMHaWN*B)&G
zSt~y$GmrWAXeTxQR}sMSe~JJm2B!aA1Z@9CLfzHfGYTTsi2P<kMFE<Tbpd)=*S8g>
zp~=c|vXp&)D$#ExUVCzFp=sPruOmtN`ustm4X+Ev;p4^8!^hKSfm7p9LJM1?4?YC{
z&nAu?7WQ`&*Nmmt2iwOd2^m%gtWQz0>ZAF6`!~rAzuAdLD8r{>2R!Q15(|2aNlZBt
zw0C+okILDBXOW8<xJS6u{I_oW`;(yQT*(8uH!SXT?dfVrAO1*40Gi2<Api&<oWGYh
z1iV}V;|iR}um`L6httRN?{9K~2WuW8TNwtMC5$2{HiqCam~r#`D%v^=J&I@GI5c5c
zbO627eE!^8!d(Rm9e^<1&#d#I^E#-=kgb>8`I9mwJYtq2#4)Kh7EGYPNhCtTAL;0K
z`9#S-P{)=>eRuldr?Y4B$fm)A1-6t`aR*k<L+R3lz_9ruEuBLu`1;*CAPY<$CR%h;
z1Zq0!D~CA_a_)b4^y{#83Nc;~R*+N>R~{Ox;Z%TR4{+ZFS4|ZePxoHgZa7eVTg{qV
zoVrb>M#Cm>I{{xs_TsCey9Ap&(o=(uG15uE(TNGCcbYK5B_+=XkyRCE=IJ;;tOTYV
zLwF!3geM1W572_LV0C={mE>}9z<~LV<EAswVQff0zCh&BsvxBB88CJsaX7Nlg?Kxv
z#?ga;%~=BlqfX;oWRFRGyMOfUm2u)!rk-k$_Om(rdHE;8ZV^~$%fIu=Yl@-aKq}1>
zO?lD(QN!916j^mu<F%4{k_uiH^in>@HL;GJf>w|K%BeAv=cri^Qxqz&<{+C%FXX<Y
zCACm6Z!3HDUUY?^dx)5E!zsrVlU_gL-gwaXIFZbE(K)@$EyawnO)W#HruhJmcSace
z&23APODOF6@N~I_!>mCCN8KWF)}Ob1Zm&j87IRRr<V>sYG-JLdZOfWwd5@l;-VMhX
z&D@#RyJpRlj!u05uLw=TEt-1G9YaTZP@yn6>0}%O%rZ`rW2CJnW1wQgc~OVRi)erf
z1#;&+VYI?i-i~oIHYeV}*)Y+B#&cq-ryOfEB;D%zy66Yv5f$#rKECA+D3fxTF3hz2
zfPki(4?U3~Y_OoV#h8lav&wLr$QR31KBrVu=#SicBTkb6)ZL6Ov>8Kk5Ywjobim84
zA~<uZB{|#DN#Jr?H?M4kdf}T_x+Jm|u@7-$!II5)7+0FecxW$~`SJMz$W<0U+WaT@
zV7ouJ1S%p`eaJs4spER)Oh=`rRGa;nFG@%rJ`~64NI^G14Z1?&6)OD6#z^OWx<Sn7
zx9P4Ei|j`#lhi|V<ra3+OImLJp{39gHZgy|fw`LA_nzS#-Jw%#AT|3Ra9_uW>?C*W
zgJvD45cTTIsb1ZXqJI_T!om4{MFFUzyX=Ii5_8AURW>jIMGN_NwdYHRqV?)0<pC8>
zXe!yunE`-?%n|)^!&(UB1hckth)OpJHH_G~6mO!q{N#pKB76&?t)(LC7zM<i77u&E
z3QbK_D}Hs>X+-T--VIR4-aM4K&&h4GzNh&X$L8rx2@B>?ZBLp7E{N94`>s_}l!LZg
znYu~adz1}uj?#L{ZD{O*yJak;uGHKsyp~PQka{X)YQ2B6?m}R}Y#P0=cQ0nDPuM0d
zSeup=bnSB?`mU%|(7+rKV#2e-VPX^aH6ii9e8IuBALgfq%yVWA`|%T>x01MXWzO2-
z{aHPbCihQs$jzC*OJbuak9G1J{i`}*OXt)aHWU3_bJw@U9LtS{!y`i53|dB$U<0-(
zG_|mg^T&oFXM#U#;ccuaY(K+expA)-l^9;SL;WOej{R>DTI#=%%_O&);1>!jb5rQv
z$Ud@xe3#E1%WsQN*PKw0y?D{R;_wUo<bqD-HTMV5R_^B$fD!DTkA5c%)ZRAR0qk;Q
z^}n4aU%5V^POJGobP0y7!R{;OUQXH($Vk-~_8LJSYa#uL4O)rZ4J1$g8i{Iy0a!8n
z49r1kCCYIhgLTh?Qrk?Qt0;Gn!K2Wm1tslh;L;ei;$!V;^4o_G(wx4*8<zRmIWHNW
zRg{+O!;{?GSl!5ImZW!@+)~DZEk|43WV^4zXyXbbe$w)CEaam_yhWhv71TCn=kp%-
z&3}G9)j$euly5iUI?Qp{ZxzjJ$2C&MY+^E%kt7bAnS;rpu@hcDv=rm4QR{utXD0<c
zvIVY^V?ME$WY%4;IIQVKDbHeeh5<l&_1tamdjaKMo#(hP@lqUj^pBl%VS4zO9GGnT
zTD91HKnR&qQd_B0H7_Xlr*yv3wkM>K@hhI$U6)Edi2KE%xd*BWrOyp!?Exkqv68%o
zHAI_(hpGC}3aij}yWb-ap*nvmDmq7#VkAepJT1yto7i0DmvT5wgx6@_8UU*pTQS(I
zgPk1*3R;fZ3LI(3j<7;naM!4F%wUCY+CN{G73p!iDegi{W8FR9()Bzs#FI;HK37;x
z(;P*Iy271uVAjMpT2zs=q0ljL!?&elvR`BH&HN3XBPz$|NZ(p!e2bo_h26eV3TCQR
z!djH5-+B2q)aMwZS?_3apcxC$*lL5H2S@9bD6*OOD;cL-DsK0vSGq9~dL^~jnJJdI
zWBALdw;4LRD0P-T@}5Iz&m^$bLY$r4U4qX{W^qPl!gu}Q*vs5b5o(P-TXFZLCX;UV
z-p4E7yd@b28xHWH%cBARt8Dn~W%&Qj4w%^4nf|BQ!Hl+a+;J=7Z<3^f9Wu$#Nl_%c
z!9u^=T{GZbcW%~!upS%Ge^B=Kcgjfkh!S&m>ugZ9r5+^ByML7u4W@6_<Xv~Uv-a}%
zwP0N2JXB>}ZTeiI@FZWViFR7o*-C?0D?Pq9e*Y*+T3vE3fGA@MrWV}g^E|v5Czl{h
zpGxHD;g-h?t9z3f!Io%nC$B#%n84+UbKIF0xy`-4pP#ng<e8%AMC%`{KlgvUeZBv^
zDXWTL2eP@Sfffw7_<#IE!4D8J#NA09OgQcPD*8CN0GDfdOk^Sb65RNDwI2H=mP=t0
zWhiSzZDcBH1fFw5>0uCOOGD<Ah$vyzJLy~!GMcTDKOJ7IQi-$-W%f}fMi<X_=Xp^9
z6?-4c(7QT>jnpIAAV^uu%C$T%EzCoHT|Dx)A?1i2(3A}ECW1~ifz_10LDFxBZycp-
z+ob}L&PUAu$WJ)01Zt<kR?}T9ZhhYE7H%_(fd%MF$YRTp+(nV7+on!8K@XCTmMgPA
zY-@TD2Pz_J99LI2y$fE3R}-7#*OsDEE*a;rO+$uya?3tRm*jcvlhD24^_$%BUq1bC
z1+WN*hL<42&Y7AzMIi2%x@FW>o?+*benhMEF%Swa!uWf#XL1}<QAq&E``aCN)vB_S
z$QO!#;)+@{Acb1XW#0;8Dc|H6F=8e7ePqvJ)^bfqRDaDj1qA25s7I}BO3jih$PmuB
z??W1Zv-p+f$d;1&971!2CX4QJb6B!Z8SIb3Xyy4_2D4E7=a8cD&ZTgyW#+UMiuq8=
z#r%J+O4PGLU&E=(;UKGUMrf!lW`&M+&-i~B<w__S_(v8{R^;A=Ga6@2i~s4-9Z^P1
z3GFeEG^C9#*-}~WzKVub8cLmZkn39&q`S#=9P|q?oLHZQu$h}c-tv{7O<CXeP(W31
zDvk|_4E)L5@?Ih@m8uKCG9cr-vMi2lSXQ`1D97S1T`sGNku8Y2OyGb}%$HuIp2&_G
zh}pptIRa02@kUW<U$}0V>uu1s7-GsMR!|ui-ppU^M~wo*CLd=)O*2~naz#)vwmQEt
zfa!$I#lyL|>46Qch0K)B=NpS!nE~hGb~4}RIi2aPYXc?~=!)3Ui@UT;A~wD1-VJ<m
zh+8(()Q2{oBRQ@T=-{8Wqzb6tbgmWWj>%*Zm4!cg{IkWPSO_&hQQIL)ui;cn&j5+v
zrXz}Y=Tb9mAfBOxfCT2yJI1S5;xzBXwa&YI3vm?S>^aO!DXr}V>%|yl#t|w&aT69Q
zXWP$T(y4xxMK8N9;8Lrl!iet1B{fkm)oxfe)INCgTu3WM_f%R>#wDEFOl|x~R30Ag
z(IElo2IzPIdKKC#@l3gQV_M@(-+?gqM$W~hMO|xO+MqTVQ!G$y%Achvtas`s0`HkJ
z*egNCer_CMD5rNgN`kTsCA&;nlR72z-as(_JXF^E)9-p90OX3&CI!9-O*&(;S|YV5
zoj>hq!h&%e5{ZFdyUeD2=^10+2*Xf{$CwWXPT102Y44deS`G(X&=TvrAkRN_d-r_(
zvAzYi_f>}f-Xfia-1wpXM$9Vd`_VL;kvhjkpO@!)J6CfHRDvJs%RFRtc#CDt^V25N
z&?HlhTo9**LY8)Vw#0=legac}@+h2PWY*z9TYo0Rkb!<W_oH)msOX`EF?}g?&&p0}
zfm96}cE(@hBq+sc^6>PTrW#dnec-%nU}g&jMm6sj_gF5or?;2SLt5M}>uNQs0?jt(
zue&xBuL5?~!;zO~Vbk|9@|2`rszXE&Utm?VhgO^vD_W`F?31g%iW!MfS)YG&YfI^1
zK2dey(@I&td>mgKmfRHOsTZzje_T<<O|$@>sWm)u7b~L<a)KvRbR||w9fTacA?OEh
z@oLf;X<_C&Wy-rW@gpSY0u&!$CSOk-9&~7|mlm2kh>aw}tX|f*wC3^Ss4-RsN1V(X
zBwbs~1fGg*dj;bx?hSZCnG20e;VVd?+CCkXCFD`@2?nXOSG<svR{>ZumDQNT+0oVb
zVMuWsuM}eGw_OF8N&{`0)(=L$`Zg4%(Q-SViVm$Srj7Z;74!m96S?H?kkNx7EpTuK
znTrZ><O<A(zS>56EwQ12uo5z86P0WdG7)BY(Q-kkTS6W_Mdy!lPW-o`gBRQv8Zw=h
zA-6{n1XXI#6$n`}bmt$hYP}3&^CyEFIs<~y$jr1@Evx~7x&joVFid|w3piK_6hpdU
z9J;M%Wg0r!`vna~F*zs<$68T80UayOibLs*anzG8pgDzTNl-*i8v((FDL9A%73dY&
zv;AJVW=Q>KY>Hz$;SC{$p{B=f?(0pixSWXmJMv}N6aeHUoP<NA**R`JLuStcZsmG6
zB5NcWA@21B9BK{i5uQA?@yu$oJ*+}3>;AR&{G$sU{8LhLcvX~}m8~)N80p5gZQ-@&
z%!#-e-1%CoAk2JZ3s{sJr@-`ieh}^oYoBf(lp*q*RKMS3axC6oqc5iJxpQPU7Nea2
z6^Ln(Kkif;nOtI8y$*IG%=>I?7UgD`%<LZoHW%z#rK9^@l0CLrM#;A*^b=Y%XoM1r
zSaPx$JFx;$6qt^DhxgGXmZGaiiu4}a#aAl<h+kJ0eWo}P$}4AT!4EVbLKuDf*cqfn
z!nVsR7RomLAQTdZPkQsQCe%gY`TWIdz|Um_>1uJ=*XjP+9NjJ|A5xrH1uSXO6}}c^
zB@kW#O`o@2e=D>$_ueuSIysBT?J--#3d&u6@WLmXi(J1Q7!1Y=8UuxYYz=g7qxE@@
zgUzz1LT_G;`Z_wV&aqnHj%CH}8`QU2p2jzlQ^}H;6}Mh@S=cg)YyxncgW?p2pK!_F
z4#mQblACpSJI2KLotzlBn@QVU92ojc4zKUHdYwL|EH43qGbri|zV~s16p-OZmY6PI
z+W#;$$KE@bh_s+QIQfgTB)1Y(;`$1Rw1Bh`MK=885Jyf#Rw1sy@o^O<N5IELQ;0p_
z?m8?%t#}V{(u(MC8V`zPt!?d~!?y*WBp+7vr8pO90WV~%llC#UNAxWal}F4a>z~P6
z>tC%5(?8C0C<#J>e)n@6qU^mXNI%dX{pLStDNTj>xWrP%x*ii0_;A2oVGR66YF*>=
zTsaLZdRJN6>?iJcFZ1EYn=jvTS>?gaIe*Zp_-J_~-EnRm$!(T3J$+=`HluCIMb|gu
zb!}|0kK+pD=S{|3m0b}_4Uv$X&MwLCwWabj3Bp9R0WVe;w56ERKG{=DU=85SG=!}a
zQ?72JsWVrbdU*uhI6BYNt<L<6JzJHuel%9K9M>LGop9uCIsWU}1Q*u19}k5s8gH|M
z=kCvzt6;K|wiUKy+Fa<itv6{g7-y4+p~ed|2PM#sjN)H&9;9MD@ZaY4e0k6au?F=;
zIn=U>Gr(bf^THlj$BT99^3gYpegCe<zi#R6J7M?697xB)))l>8GA1=2-T9o^MFEpm
zqSBhd^fpmyMQKY+eE4xoYD9frD7LBKO-!ENzRCeN_V!S^ee{);I{mOGr_8`|(^q0T
zq9Z*q?RqGqf*G<0fV-9LE{t<9upWd6>uS@{7Q*JPz1bgA7pLusSwS@Z?&bH4;#b{S
zyNg+RZ-p&2JB>9)J!xxn|NPuS{A)G^YAr`^037yrC&%qmCB?V5m-h>vQ?euEze;3w
z=KoV7GyIPdxkpQ~?mz3b5B2kOcZMO;Fz8Gy54UTeo)(Wbma;-fFNwg?wC_(i^@!zD
z(Obb(8O5Otdo{Z5kIjeMvYiLpx9`Gy=IKI)M(>YBNUwnlbqA~Ebxgu^wL<TYRo{R8
z9m_7b7eKQqg;VoQCER$gQW}?pE}g@M_5wcVACt<rMGStVy5C$oqF4h;^2V@l&#@7?
z+&+FEo@2;Y*_=1{Uo~Iuep_EX)Iky>3SDx}Q^ldTZ}!|jKp;%EI1^(q5pQ2{w^O%i
zsaBZKs%t+!Kbx$yePk3h1QZ>se^8T^m_@cMZ_*#@fRf=PJq&piTq=h2oZ-><KgcgY
z#}2sngfN^)dE*~qh@b@aiC#jKeZYQv3Z_Q=B)<BGe<T!~;K^T($|4XCI*6U@_Uq9T
zf#Dz7p>)ALLnyRLB~*n)Zh#8^`TwH9*oxl|YMJ5tNzNvW@jD`%1{f4cDMfMNycR^)
zdGxTZa(Scp#|d65Ld$#lIQlMh5ip@sj!+P2&639;fha5GO}CMmlqXt|yyFQ_z;im^
zkUZ+N!Xo1jplP(CiciX3*rJ9)Z9R%&5%DZxmHuEE&F8F>No<l$<;q#nKRs>(pv*C!
zdj*PsC{oZL?I`iyLk&8bR%8j)rCn|zA~<%JGj*bjN6E=VoS6k{;DXQ`yOXZ<kCGP3
zO2md3Z^PazVEzvbctx4O!=>&h)3HH_qx;HSt*CiT4Y0EJ4gDyi)u}D<H;g=6yRyDy
z!^IsEM&&&O(>kT8vy1>s5HDT}d|O!bcn!h(m|UVAL~|1`b}~i72u(t*mMT)GAl1bS
zr+%W7B9ipBcfqLRaRk4{6EYo=u^DxHP4=!%kilRRD-3k}ua&>h1sOuw!xAl4nh;({
zfspRFT<Rk{mN?9V%_t8tDy~c}66=G!>EPXCVKbbpFNBtpR%X!4Ch^D7(~=YfB=}%i
zGx)ln?8gLxMLp^ee(_vy(n4X&2lUG2)sb0I+8S;rBJfVjhY3vx_9jqUFXMduT|&5z
z$Z}QOeKGo?I;2|EJvED_s&H>btD);v%_RiD#LeuCv4nEI$2Ltb&6oj<G8S7*WKPpC
za4hybVTY;IO<9m<`mFpSVwLGk{FKMtnM5}kB4w<hs5X{+gE=eHn9POw-Yp>Q1?sSh
zB(c&O)u!nx&VJCX^y<c-i85R<O!?U|oQk+{&Y?*qjOaNuvwSJ8_=nw%Bv(f~a~XSs
z(DKCRhSn5CwyELSj@015W$CQMIrdiuj8J)YP7Il?3@*gNXAL3v9l#V+uBwb2)1g)o
z-era?L(@^Tv{aD!MF+{{ZRZSf43Mq3zO_2{mNWEHAfb6aa9VWre=0$Cl*mDxLTa62
zby@19DvQ_CQLp>E8<&F<^~t2>vAs<(_^flnlB0^cdYxueZKml+wGRYS>!-p!$um`w
zlpp&POJSQ$PRT(5<xgA<^6QGE$}JQLI{;2F3^N=(bB@OyJK0sW3}lzGp1CILiMSob
zxVMUOrvE)HzP>yavPuDn`7~$0FB8x;psB2PQR&n^4h6xz7m2o`<y5wbn^y&p&ddL%
zSlnvNl~)~7D=P2}K)RAbx!#%Q&qp($xB4zaHi@f;U%y}ISBDS#5@1$FOh!AP7FWJ0
z3sj1y&)8C4hq@)nR6u%HHFbOdJaN|2xYYrvz4$>Hyg+QB0$T_4^vxf87hI}FG821i
zOH%(aWXKIVHbtOFv$h~50iEQN!wYdE1V~d!MpA*mp!%Lu)yZXeG2EFZFKa<XqvaM~
zlCLO$I!D5YB+R<&?H-0NAI<$2uwM{l_;<h_WP0k504iNc!}s~_k)oe+Zru&zp)qZ=
z`(t76x=Vg4n26O&7(4#&*<4?0IBW9ma?Rob%S?oJNK>RGobkb;c<U20j_H*tj5P*^
zy@DeUYH#9HwBp5iKG;Ke8n5ht8iq>pTYE+C$7iF-O(Za9uVr;l9TB0bX9riOtEn7>
zxEj~yK9U*(=^9)6D*FDA)y^Xp&j<z?%yvE4UK~R1zd+|H2UM<_skWl*f9c9lKF&K*
z{G_&o+CgzOU}wDf9O0OE|44TDdJhO&O0tdEw(l({PA9KRW8jz^8E8o-(Al{%cjIiD
zNkUga&R#VoZX%a=;r!RawsE0HP?VP?JOGorOF{QapMbSFMM5Rb$I|&bsWGXdrq{zh
zyc4BZOxv1YGxx8)M)eT2nqDfOz)JNNw40v2?#<Pb+}}k2BQ9y<joLkdm=hJ9p0K@U
z&@CM=6KGQhqZCvoVa-H6H_l68l#vzB!c~3=rXZ+@u*>L2Qg)dRM)<j%)ehm0NNzBc
zM6HHW^eB`UCJv;kYp*ET>Hw(6`sAl{R0@V^4Q5j?!`0zmi}BP=x+gU%@i&-wI9UU6
z)v?*I{C<(<>de3*JrH~<-<>><Uj<r0-R@$mbxGS1FXTmh8;+U$q{ts--&Nt~WS5gT
zABzrC*Hwk>eGVyM#}f`I?r$^g2V`vI39wnmj&^meJKDD`e67+J+OouDL;CEMQU0v(
zIUdNni)~o=Vq4>bsqog?`g_?}7swLIbGVYDGw&}~ib+E1wp6m|TsAH$3X;WYZ5a6?
z)Slm>vGL~CFw-9QWM;_`Xc<=|!*2QJ&Y`BxqN1%OQ06YKC~h>Cy8Ek`76eNmjxJiW
z-*I@tX%WH1qXk+<5rMjYIxS`Tz(sv$VxhL%R7B~Gd^V<_?m-x^`OVj7>j(Te2R-F~
zlNk*EZ<+D`A85_V_WvYK{EwZgGg`X#I&Fx)Yjp(kY0rY)B<5klXn}y_waUOyRvL8k
ztMkEpY7NTb))H0yUvIoaUEVEKx?>-vu86}}yC<aKLkAhmL)sZl&~#z(RmN68U5T>u
zmr7D(sBR*%lgALWA1S(WEs}?kt?Dg3Ykk#pWRl{qZY~MdmcWJx7vf4hsNGbm_d>B>
zcNk=6rE!OWVSZjR%H^)*=L#Kn)Ckg7(7pwfZ4=9+NVdL3(wy~2@%IdGRaKBieZ5LL
zi>)9mI+cN0e^A5LQ#Bf{<F)sw>IJUF$79nc$f_<G)b|QF78N!jAlZUCQy}b7ZLP~!
z6l09~#?CojpN#P2ma6~|k=Ye-px=y5$2x|Ik$*iDJ+UK>rnW||i#DMH<B~#MP%6pG
z@Cd-en_;1-B0|c6g-JXsV7{!a4fk~@kZhQ(!j(5@!<UUOjfg<6-T3UHX%(9MgW`?P
zk)%$V&Oa#MidNW!5kqoH$W=?pTS1f@CG*!1$dm(!8(x$tuile#8JVXHWMZ~d44qbW
zuo>xDY$0n__B=r!qq}Sj{*s&Wsk#U2siT13w6eOf2nR~(ZSM~D*rDZLCL<}EIPiK`
z`vkHk$ky5<O;rxvWRXLu&+<)GoauMskhOASDG&|%T+3CYY+DeoRlnacH=D_OS^+83
zx>k`>Z(&X<VIO&V((#sSV;yc7Yh^X7YqTUHb3k;X`XtP<>732oZrBHga{;kuvAKUO
zPO+JqQ3KD`I)TA#Ri`!D-0+HLyedC`cK&hKg?@K^4LyG>`!?tQwe|eK&@7F_Yg}h{
z%kRhI{dyhOIq0+tcO|U8_Qmnd@vEj?{5YLV=-Y2!-c6V~dmJn5zh9n?%a$hIO^%Iz
z_D}wr1xe~{9?rbkq^5xhxh-l1+Yr`^Sf`_sG><l!?_6y?oLpEu)_Q*(n!dR`doroM
z-}v*#Fg(!{Q@Ai=?#SG3n!OlzWA4P<+0%==EAKa&^T*ZlW&bdIf6jKA*tD*?gdy5J
z)VfOBvhF-(Wj(mi+^TYTtsPa@l#Fd;kOU`#OUiO#v)BxE+=f5gtem&X;?8#Vxz=^!
zF9bl&=3kv0<IYy2d#I4Tfc;fW3_s%tC{D|si|t(ch>C#qcsw}e-+40pJ2eshAnU&Q
zbu>f$@zeE|G*Q;o=dweOPj+`l|9SGa?RM>Bj}KQr#_!ILFR%OOqiljY{puC#<IMkS
zUDxMVrltvAf^Jh%Pf651jkqpYV}?d4&Z?(f=+M<smU*n~?C80zy|LRaa(921C6$kn
zNBqE0&)80dMS;nEfL$$#CYjM_!CNl47fJT*f&dIfLa(V>>nV7&02=LCfJtcjqJjt@
z&Y>-67$Ate)8H{aS9{+wd|JLO{k>`TKIomAFT)TZn?!AhlnzAO!p1wBWB19z;@kKX
zg`2p?x2ttHWqe=$1Ug#V*@i7xq&Bc<y-bgkr^LUntiZ`*;H~Z)XjuiU72H92_+iYS
zdr1f6T2BavuUSEzH=b<}TLW1ZBG4@&4@v@OUzg|6{DTGkU_uu{6Ot2iB&mxiKbiVa
zyr{!NDSS=w`*w}M(H)-&fv@cgmcy~TP~hez_w-v(<_b>I#3HN{nbxEI1x`8U93jO3
z#W!~-9wg{(VK_HfhZGx_<iY;I0s!?l=G@Q69eR)~{s*Br`RO^sQRhiooqS?EQ_m7K
z#-%H0E=ak=9v%SN*bx9Oc14SLi9N28F!c&Ml^`9nMF9G(Ef%TOCIltzi637G=hFn2
zt_VI|&Lv^W{tSRlVF<@3@I^Te$lVYCEi&T`fM5gw4ghjIYs|Gy<S9L*MA`>~zkdum
zp)@Oj$%?4dk9z)_Q!Lr7MFQ%A$$lV^Tg_4k+NIPMQV*K~koSB~V*~y;bWH-pN81G2
z4P=z&BG3()S|A0Ni8Qb^b;~7U3U5sC@PIkCc)bt_>4Eh`rfv8axlH_f3dgc{48Sbp
zCH&tEq7`Czb!_Q{^?Ild+d=scquT*`1DGwS2_r3F25L<!P$Ng0khWk=I8{~IOfiqd
z@2|vP%Eg<(tXA>}>rJefONK*RQi>F_XS@MeO28#3b^d`-=HQm%5LZJ|0j;IUJ4t6Q
zt1L$Jhl*8&@CHcEKoa{`gC{LPUn3+6Xk&=E3~<UZn5hiM@iDls;>bTQCE;25ufPAb
zytnSiG%q4Gg#fVFvDWBB<s3MDOBb@JTmu0P1}Iu{(1}_--Y@(F064N3hdB|RA|3Rj
z0i8Q}zb27GRWQIw&F4D^4zer%4^5z9V77GjQ#e>QW34b6de;on)Uj}~ePCTw{6B!8
zI09EdYV_a327NyBW|2g8Z$Pdvp>^;#P33|0<mNu*5T|z(Q2HpK#Ib)-!(0g}7zLaX
z772DSV6A6o)@=wVpneC~lmUCdj^r4SL*Fgs0I@IckFHEgJdipF6Db3pGhzkMX_F6x
z0}u<UF&av{#T?sPNw8R6y*VW?baDZ*f}&c!;QTZienmXmzLC|-s;?#CxbB4@<UQR%
zvrKakm|BX}50X*nEij7py^R9O1o1Pc+_kimMn3(pG#5%cTz`ieM+ll^KnZWY5d|%o
z=0gZ6Cz1nL*B$T+d^Zsf!)uHI%rQi$Qb6$Nj#WLY=g0w$Rej4pmvNkufFgt^y~(|n
zx-0?`(IRWKf{GLD&O)0ey%%rkNF-U*uIAUUT*ekzjm^(tt7DOPZzK%a1qJs_ABRK{
z=TKq;;@I~CJH8bkpP+1q1_u-L4SmU`Q!f}%<LgdSo-3D`!-T_{rImp0;j$V!+WOd*
z_L0$IL#{g@OM;zP7KhAV6c)q}3y2>!FfU9b=J6H<AY^yVBCo=LI{`Mxj|faP4RV6d
zRvM6P931g#vkj)nhk<y0d5G5*r*h}x94}uDNG1R|IX)=RMyjT!+6dk6{s`3-6#;1C
zPZJ<^n2HVQtSwZFgcpp_2KJ7O#H@64T<|^1B-5%?(4uJlhs2;2lnl2m-yt_|GEn?W
zvPZu=u?B0a9zHN_N=ssGs9jSO|JH@87n+IrPak$CFPT<15VI%kbVwS{piD;9_|}`4
zvdS?arKf<@F0c|3Xz}=~tOg*H7N^F;niH7Gm8a-C8IfpRU_kK*6KYrNdQ1us)|p*%
z;zdEYk6tBS9>h&(ijr;Hp9EUQnb(YhwwoBF8ox)9H9$y_`pq)Sdtkt+0*jOTXg5)A
zJCM|yZ0C0@;iJYNH&Ba&JT&NDnNqj*u{h6nyuIF$Ehy+m9{)r@KobD~Oen_2t(aZ{
znU*`4va$I|6w!BwxB`v!#INlM(wB8Y5g`4WB5e{2m3gp0OYCNq{#Bt$&QPI3%Jq_U
zzR6b9igM!hoNMW;>fRrkRl7ef+8DbkkPM_?AOuYjsDatqkyfU*CIwk3QJa!!Ew%6H
zT`vIyH)Qu&Dt7Gj-F&?;ABQ#HgDac62%?sliY#{F6jaWCn0zd)@PB5D0kt0kXgw(&
zPYcGg!inm_cH+fRxxeD6AkUld94blBew@2N_Aosh#xrV&Ak6Fb^|ZQe0Vu*U4lSjM
zbP_VHTy0FX+Qq9fvzYcx3HcR$-PDM|vOPeZb5pJpVE+-FSW3UD*f?QIy{ZT%6!O{5
z<n{+s!>>4j;OMi~#SQ!bgZJs2{x{9c_}_w2*;xL^U{p<Mdu;aqjOMNxdD98QnSB6)
z0HiB`o>%+Z3d;&FC_(>2zQ<3!)v7fTRh=8sdW$;sV){KbXMblNKfhnx|9gHYc5EC>
zXk_T|%R%ucd@oD<T%0#aK@zLFc-ea1lKTyx!1*gk*>XDeco)lW`BR;8ggRL)ev?OA
zsjCY;)-ATw4ehg=!>5|K?@{8a7WM>p(Vg~nSJ%fg0rGVev3(5UY5k04XROH38Lj;U
zIga$_Au)p7mxr^Zq=OI%Eb6p6KqOmUV6-^Uf9+|{;aEPU7g|XJ|M|&FCNU+GHV_3N
zq~*gPyPbs=dW(7<#Odp|>QO@hcswl$lB^Yw4qqMv1jjs^Icy_rFw0={{=xI_Z*gUr
ze{Lk@e+WCKNoa;8{Jq>kZxx!Qe~<zmp-XT4Wp(m>vEilGSp39Cg*zpN*Eu@_5QoyN
z0zu7`(N{KZ3kFvVvyD8I5_(Whms}Ye!w=f(HP~t+D<g*)cY)9-f0epJs4hRj%rgF3
zev@5^msUHt;;;G|!lFppnHg>4$~EK0KC?NwF4r6|WDKU9n`@^wO#^9vMI(V_Ka46@
z3`8R~RqtZ>P!iMTdLFc<KW`d!(Dto0ThrQbL%XxVB>I?ld1+@U4ltUo2(|VGtQQ%m
z<GL?MH_9ygE7MWXvN(Hk5Z<yyFphay_lVh?4qX!_8rLN-VolwL^ck#)R4BKum-5+Y
zi4}>nJgQXt$-F6Wpx<&aw>9z_r9z%j?j%=#!g`g<>kiYit&reMk-<uS3ogcE&u$YY
z3Ycf*AR&EJL@bU-KUd4w8%~kESo$UEdQX^p#4}P8N4vBg-n!Kp!M6kLzNZ5BhWs$>
z^O^F2z^b!!9tWSd0Mu-KVtwVL;zaL~+rISvI>9`IEXJPp(m`iSWoH?8_dUE~j~UhR
znKn0CDXRcD!L!xk@yjIT_k<HQyT6>i(E@qMdD|tKm3vqU3r%;s`UPU7IOtHoE*^ew
zu&18qWkLC0E7({3>pmGPV>u=IbFof$D3Z!(eX(&Pz_=uHMuw6=OyJ7N3Sxz<JlytJ
zrp(z?1O$5QVKRVrHru#L7n|fLlf!J%3w5>X=!TA`YphsNhe6R{mF!7p;jph0OCjQ7
zDo0hEcZuN|894W@wB-ERK#7R!^v4-XqrX}gz7*+rN%h04j7D#`@-f1$bSCU2)=fS&
zEH{dJDaP@usPq$rE|(;_i=;pOZ((D>7D2L`*5A!PVQL8kZ8{mvx&^#_6SSJK^ktI`
zO3nJ>mbTIJ3Ot&>Z7+;ZHa{>XP7ym?Ol&%EJbskaj6Z05;mVWlH9JC;U`OsNsKLDg
zZWhFsvpD(_E2*S4_>m1%{;j5Epe(YsmI#lMoR0Gkei1_e<ZwOpO?8s=VyDTHjdf?I
zUE9wGoid1<(~4?`vpp<AuabkIDD91~U5d<!a`}wboi*&~GX?<@1-RCw6C6W%qtfXw
zdS7PLQYLkR<TMKER)R)^ClBLZE^@&bumukP=ua8arfQo)OI3xqt{eK+JivH{1E-Bo
znfG(<Mc!)l?K8Ik)Y*1F#%67$@|su;(<cVr*PF!pX~owdH436{=l%aaE|KNIG@f-R
z9=Jbi?*q}n;Mu)x&D(5jTJcXh)>)I#aoWLu_wcB{J21jK>72wc7~Jh7>>RgGQmCnZ
z$>X+(>f@yCVIS1EHJ`Rm#sjc-Ers{B`R)L@N4qsE)+O2cDzOD~16z-`JOjDfzd<vw
zG$F1WR@Po~q7pGL@Ve09{E01{lPOzF&x7^QDc0=k{TSOhx+bL_Qo#yUf@EfguM8C6
z(8=~ORG1;9)L$Efm^!G@#oaB1`?S?<OVpD12@q>RRx#(G)moCnuEoCK8rzA)g0m+(
zc;{y<<1(-cl3kiCiFd_R<K|}9HK6{~&LTJ&wI%e762lTU!?InA8CY3q<=G0Bj7APo
z`?CmMUc<F13&lD0673BT$|b0yuu{rbXd@Q%UYnv7%Riqg_JI`%*j5g=L%=f&+`OBQ
zL4i>QDs{v4;2Bd}OgYe+gEjq8K9NR$;q<s%G!C^F;xAIlZa!!{qJ9cVZDE*gyNv7l
zlP;I(Z@NOUOc)|NS#*|m4(>cufw|kh*_V8s2rI-YZn9=^`_HDyE(d^h+-IzjFPS*N
zzWj1bzSj5$^o;Y-?7!V!Obo35$6|?*jpcu1uU0f?V*k_Pdq(*QRD@XJZ-8KsWD9m)
z6WoEyi@L*mQ}FQ}uGH6RLH<CJOXKB`J&yL0@J}Kk8mAZD$zun1*L{w>1f0AZFns8v
z4e&g$SU#S`VrCy8M3tfY2jkl(0f|-@yt6r3YT@L(tE_8-pYr}0Y~5+#d&1HKb8p1h
zS3H9US~y4dRzdClx%5d+%rl(muKwrh>n(?7K2LsBf6z7uEu{8PjbsNw&A0d}`R>Nu
z$1LgjEGaxnDF#uE-8x=L>53@AfvWh9>SweaDB0W*bWt(<v$xd~=DdTBLMHj-ZNb1Y
zlXggv1jS)WJ5&m?IX3Xasc~VO&Q4$*8};NL4L6;B%G~3U8|hP|K~d<@P0H858{fCP
zT1}7?1~wS9ZUvL(KL6`OVAzm>P(VeH`Jp-gF#9h4_T|RV&)oI&YkJy?#SQ!#hUNpb
zmU<WZ*E`!b6Amv23gZYsQAt^&t+BZ~U~Ers+?VHg_g;O>-q9*!lPtty1-U281?c7p
zjlu_H{O5MeN<)(Av62~!Y5t=IQ}O6GA@CIrt5>bQ=tgA7X})(t#43eirgV&Sbupyi
z1uPd&Q<c_5o9COC?l}$eXpmU)8|jh4(L1cQYw{>Sa~z>8l%~Ivs$<6us!z|*Y)e=<
zsFCAStG%rsj7YiebE;a@MWL@vDM?>&MLWvR*e!Lh5SDk+K5`HZ-;wR)bD!t(nsQiQ
zOc52~71@K8T#Md^oYa>H;EsxE7uTlJ7y^-Bk#k{KVsqIzEB@KV-dpYMI;@G)$+Be6
zZ6ne(sD{uydO$m{BllB)%HtUMi{X@xRoI@?f-K|Gg3A`<L2R^t{BjVE(pg7xosOU`
zw%Q$VGr(ko;RK)_Rbz|3kWgnZVzC?MK>Y1N3D7<EHoE*9kt*M|{z6AhQ-E84_SoH3
zjh1$aya5{Lu?n-=(EqkW0GnsQ0+BdH08E`_Dx+*%%M_)U-{W#1_SCD@4M$L&;xSKA
zD;uTGuEt@U92{xDjG!11jvb~pXrQo7zslf|oAcz5ABk&qVQ1W`pP+z3OE#Ez-q31D
zD=G&B!fH;D*!63XJwRD^h{YM-aAl{oT-)-j{3Fq~${gyqg>fTLm{pr(#n!dL%u^@g
z*l)TuX?&$1Fc`{jGjA;UjA?v;&DfUS4w~p0ww5u&;*FoEa$qB4=3WWjdtGs#<spBG
zqv2Mby>+omd+MH<a&@vnRY6Vqlc|9(4!)po)$5kjRBqvtUS)@-svawa$F2OON4^lq
zkiMXpkwjA#IqY;|nND%^TVFy@dh(K*t3!2cbqSSoh0RTQDkEKuR-wH~PL+S}d)X-Z
zG2Cp-yC8IQDMiS%Wf68In1fu2q&)ePJ{!?U(5<#qZb-JFVBhr7MRp8<7EtF^VV#)d
z!**aOiyXS~&pL|?qPtr(JA*a0ve$<8PW1NNp#fxc-NQAPSX<Uc7;lVx3K8ip)a37&
z4lHUs^|Ex4E`}Yo3F_QrW+UxuWJ@I6+G7@+q{}DCor;tDG;X-HMhm#BfSX&iZi*i>
znk&v(hVe>uPJ8P)-^84hI*Hv(v=Is)?qw~(x5r+EWQ?r7iVjt-c;C6}v0XlEnvxLK
z8bB~2T{Nf9N?WTCpWksaYBNf*P9tdi5^Y<xS0^nvGF4Z(gz+O)Otht~{DxSf_?cLW
zcSbSJ5l+3S7<bcYS^u;N9oN}92p?MIL#efG%M8n_G+e59RaA1h%C4ibW)83mY?n+l
zTh{YfsK?o6A7ZtxR*B3WHB)2@P_2DO0VGlWdvz|hf>>qAcM31NiX6q?<5X<z<J(wm
z`aAzc2jNXs*3@Kd6d##GFi`v?kCOc(tQv`(TdG%<n2;0<Ys+!lX4NKH=yYMuC2Q&&
z&ezwR5p*f0@I~!Zn#-%d@^zR|hwb(Z6UOW|LBmh;0Xy^L>BF!-L6Y5v$T~%1e}}oE
zSIFCAtU6hEXLJkVF|qL;xbJMX54i`R%FHMXIE{&p{_LA$?OyQg2i#_@er!Z`%=S&h
zFM#A7nQ%c=4W`PiN^cKNctxybM_WD<kZXrFaBBXz$>M4+0w>j$W4jX_>K~5wgs6v1
z+E2Xc0N+?&COhoa9MM}5AOV*RoP>cH{kRVZ{|i5HuPl(JJfv^ag9&tjxr*TxT|Fzh
zJmpVe|91JW7Z)STf4R7r|F`=26|Je-Ee@ESGwM%ZRZ&IxeIkJ*Z)F2F<?S5<wcyq&
zy3(KD@@mT{w?m$7lqQGkaiozkLQ$rKp*KOeJooUooiXIg0Z9A)ir)$)fM=maatW+v
zGY5b{BA7e=us*$FmDE~bT`eh*P8?qE&a5wR5P9bQROnay?65S)TCP=2i$^Nd%S)&5
z#^ADy-v*@%T&SL-@4GjCx5oPaz{39*WA7AXOTcYwmTlX%ZQHhOyK0whyLQ<&ciFaW
z+jZ;Qj_w=L=a24_4=W>b<;;hCSaYn2Z!m7Zwzj@(^>&hl3mMvAR5=2X!2Eo!4xwX3
z*U`a9MF4oczPY}(d?YDXd)Ip_Wq~VYH7%#h*^WaQe;Uw2xztq1CS}4vmPw+%i=_R%
zAxSP_JQ#oBIFUd<c9aY(oDzbTJ^a5C%iFzYGP}gOrZe0wtDa0q5gMgQQGra!dJMu+
zTqyzi5saQkoDce?$*CK6DWhm8M0Tuar>0$T5dhKD5#Q7yp1OzOm5efC&AQ14Nmja1
z>pd>HTcJ25Gy{5p^Kv&8RSGgx-Pz1#vmfS}_#{T7yqqxs7ibU4AWmW^IWcngf<=)L
zgeun}CokAWHkoYUlG@YF!9xb8b0xa<%%PfK()5;$owgdSwR8vpgGJA;RklrsQ`#$|
zH>i1Y`iN^1`&!4k0_KhCAet^7tl8w0Sg#KsHKA62V`KKp1Y&^acmrCK$R3c>IOA+4
zf^adHQ)M*8Toi=F{f5m<RR-J3>KgVq4tsGi@2(*2`QvI)CSmSp&f^-7?~qyqxWm%0
zLL7qbal4Q(s>-~xf+jVlV$%}n5A#3BA_a9{ekI=IZWF;r%V^V9Jg#-HJ<@w&@Kd`{
zXu3QuPPkHd<v@M*`{Ha7b0Qbci0Fm><_GOwG>GhmgunSvtMv-(67mN<t|nfq7}N;B
zJBJ)r^cyCC2j<iA_MjAEaib<tC?bVaO}fKA{SlSv;d~5Ym*E!FhDgL!Dj{Y?g%u~Y
z)ZVHKVhR2g*ebqi(5ltELzpw{CZHf9@sY|P8a$Z)<{~UEW&V0MrhmMtW%7h3yWs9f
zfg?H$yUWl&1>cN1b2n3F#^Fb#*_l)fSee7RK<%=fMmNl6t^})iP);YJbRMR}0isCo
z8nl2|$U8y@Jfo}tjFY^_0CY#*fGL!;_8YsSbOMno!*6_6bsW4i5=henSvpD+Y=9kZ
zgskEGF8E3Zp1B2>qp$<;6t)3*(9PWdf4cgp_cgAu&RAgz0Xhe*Iwo5R_gR~VJZkUj
zk^)~^`z5?~Yqv<l#<EEri^&fJgV`$&$q|gt(|5NklsA!}b{d#a$*bCMAH=Xnu!Jx#
zoKVJ5xnc31QUq(wo_2gX{gWu}o{O%PFIAH!f%_UYtT5jS-`9Tdd5r8IXv~y8?XDw^
z^;UJ{H4Qr{KBoOM&S?nvTBFUL?6g0__HUoP{I3L&-ik7jj#0WIv;7daD?Y38iqw!F
zQ1_1t!x=dS9mh~`>hWblebb>z$*PYbZ5O$-(GRmPyRqXM00)^L{7X_ilYVvWCAn<n
zMU8em^<`9DMWfhPvX%~8WV1NS^u;N&MD7*F9B0J5!<bp+3<yx1kTQ!&W$h&#ft4YE
zpQ<}R**GW=?+u(qo`<iPJHQ{S3<Ez$?|_?o2mb)X&pZ~0<kFz!H`oP5F~=a6WJMiG
z@pcjQY>~L4T&V?#w#1S3ZS%$z;7<}Ke7lo5(7nDXk44UEK%HY&1&>TX5j^~HH)3j}
zekkZHfX>gB7Q#C3NNs@5(l#&~vbm%HU(`6*)vwDXmP0rOZEB!DQ4gk2iK&%=$3R~d
z;&WA$_twq*17sFt*0ar|nZJ?w?WJv@X1|fMKEIJuUzbsy@4OOfDv^_(4HG=7n=g&{
zq73;8%h_w*(1$_*f4n9@f)qCa+N1ZOoZ3ZZ1g~9r)DE3+b!ndNUv(ouAxO>Hz}z{}
zl)sR1ZdKulFuPDr2cY0*E)4!$e?G>-sTHBLT<VkJ`Dli<_t_WAQR+)Cl&1<;OInn|
z#?24X7P)|SP1r*C*S`;~?xxL3vY!Q(*Rh|ie65?-8;3R6`Pqv{pR+i{oIDN)kz8<H
zuut*-eYR=NIL3eXPMs%ca*95Q$+B#|*tU0?$!i=J9A)`*bAL24(tmYUlO)Q3fFFF;
zF=7Op7I`)ZAx=9dul;7Ek?|XHCcd?MAHe8Ma{dB8G0<E1kJ;*X3HHCOxG-`2PtBmE
z?Ki~!t4LVKj$Ovi-PlMkf?&X{7_e>intH54(#>Llm4CW(a~ajCqf}%<EeVf8n#j$}
z{V<i3IDj97v#q=JYlg=~fmP*zMFS_F2eyly$S-&=S_lqJ5FNO8_HgS*PKw2M{YF5U
zvOjR|K4RY8tGfFN$?z8NRZY6OWPVq`Q|O`<(o;2UK8l0GZ<&+w2hSP!7|7~p?<|cv
zLlpl=gtX}{sGdY4lZbpmc+lLaB2)dsVfkZJx6wojY%H8;N*x?LifdLc;6joZx{?vs
zWfUZqVop-9gj@h#5R7l*F-InzWQ=pW6Pn2={yElvYb7Uo=jnCOqge;?<t)B@EYxZO
zG>VKUrVf+8ru4DP52JQ9G;_jTwxFpuWDBU7b&JX#l=9MbfMpp)3JG))1<Fafufw-R
z196_^mSG|?`~?UUb*Zz3kTylZ*_gVqgAh#`;S5vcy5Z2zNX-hx1*BmJ7Y6FAu7cGP
z(gL@39IEnT$h>l8EW%42RLl|t=!u7YF|>+Jo*{{aVD^2#2wpUm!uJ+3FD3npI9(kq
zqTBB4ln|Jx`iav^;C?aVr)cY(w|~w9`K;Ot7!%N`<c-Kad5Hx<XF0Ngf@Xj6mF&%8
zUM6`<uF56PUHR9(MVrPu-9p)1p+p}Obxssc6PQ*i^7@M6&rI?Y4s~IPS-`Rqz%qxT
zji@KP!GxS{A}L;i^uhid-i8L9Jermx=u$TT%|w}kh4rw=-iy@DdUpre`$ZsW#VS<O
zk>S=qT78nG2dy1W<zZxBjF-9w$1$pt2_-{`6G22u%e13};E6kH8McU`8J&zmAB~ZE
z;@JB}Ub9}E87isxEKMRSLG$3+4}?ExMUKPTCovgVf(_ygy1se818usIVRhV$$mCq4
zfTUQg$n->T#S9#3=A{JKV=OxNwNxKe&phPR^1LfC*|2)hhozmWCuS)$XJo5Z^Htom
zxeQ`35{ksUjIu41O4-Herd0;+85c)IMU^-ZD(^p~dCBr`Cv;1ncVnFdYfNax$ipge
zjxl0STs0Dq#090dz{Xk@VvoK9XdS_+Mp|Q27A0}EL+k@s(4zXK^~n3!<{qF{E%%dL
ze|YDUF3Jz9$Uv>7+6AGCW0-x|^!9I4K5CA*<8Ig{6t4sUPAx^1DvA<6Qa?I3!>pi%
zE)s+oRxJiUOu8&+vZ7bj@DhaQ;wdq^z(x8cDAnWwm%SpaU@FJl0(Wp_DAn3{X02M)
zblAx5$WBSKw_SU;`E<W{A4IEN2FGslYWrj27o(qm|K93Y`+srxbi6m0W}@!o9Yhmo
zsbe8O&uOjPE^p>Xw2o@FYP^IO=A765BW%QX=rL3~5Fgw<95)t)asnQPm7$02f*c`g
zVVGoVGrdp?_7oj`DHe))EwCauvZh72IcY1(q+)YH73;5o#C7imo8B%k`Y8B=?1sUU
zlr|v%?E06h9m#ijg0zx~G$g0Io@*(1j?sbw?o6vZR|+nAVw77;07b#M@>C&@ZJ!df
zSJpUJX?CTgNr!TbA1v|CN_&zN3wby7&xq0i0ybiXbX6PFA}^cl?C77Yh9v{~HC@}@
zD_9=fn78q|6fdJ4hC^pey4iV5!k?u-#u3Qt+BIM3VhJegT7S8S%8op2cqm&n4^y?r
zs};#Zn1UCK1br~`Ga&V(JaaEMo4%6WF~+Mcit3(atzYK<(n-NOmy3VkEV(9R1J^%3
zCypWB-1D+a$1*O<Rt!CfNElK9=ETbqskp&&c{t@X@$x@7A`cOr-Xy;zvu8V>yZHQV
zVvV`GW4r#`!~xi(1xa$x-gB74LEz#Yn%;0!oZ7xW^`=H75LWTLqzoYx7=5TWn(`H0
zpv+ErNkKJOk|Cz{%)>x|GwW=mmr)oG*ww2mW|t*0HHz<|^oBjrn(>*PecTE{>r2zh
zT)DBIM+%yduDXRPk^b&#2lfA2>1A)VsO9(T-kVtoiPS6Fk7k%<49!y2)QrotmreZ~
zrqDM<IcEh~sTHE>vT#JDFjOVqKA5rs*dJc6+~zX2FlE<kq$CFgcP&Z79ylFJDnEJX
zN0qW0N_R%DG4c^?M669DAA!%wF>yL#9H}Gb(=)<w60y|d?Xo2M2X9#lp=&~kv)?-2
zrek@Ka?B?2hM8t&rStKgL?i`3VIY5N%2|tYkq`*k>#!mR+Thsw)Lm@z7;SlLPVrHO
z)nLROnKZtz#PAi3@+&>sH}g+a=6HV{f!<<t#1RcR(@#Ja9^9L=@bp7ccdk8Q&E&Aj
z$3;JrnWVik*2$BBG*eH=ZJM+ke0dr(YAIVhz_DfeBXn`Z$w`)!X~(=}Qng^8h+Sc?
zQqZ)A;IpCs*?^^*w7qH>Dk@rQLj}?fm8Wsqq+c~a(Vf;QN4|gifU}a|tmV!#f;UsK
z*nch1JIvzz;SLo4G>C@JA)?G$E@4wkQ*Z2`V(W3?{N44^3mBAU7zIm+O%=cZY$HT$
z@jn<f6^utxYE=R0BtcX4pf{t77@{0XGwAmKMDmUt5m01UyD9?)Wp#d9W2uompV*Z!
z<4Cr_yK?5p+?W{!#sh6fc8}D-4)bdEe$SRu%*7`}(KO7-87o6-iu+8H!W8=(>X+i9
zA@IQb>r$#Gw~UUfROHXr)YxW>pn1Q%8b}|>IL$4WV#2Kvtt7WUWcW@+D<ey*t{0H)
zz6GCklAUO2jS0XqFj5Kp4}lANBJ_qR<0!OdYxilU-J4iTuU~<ULHB-P3MFRZYVIg3
zb8iRz@D%i>-Y6;7MR5V|oS(X_DrC#^fYU`tnS7-+a~hD{wjlVMZ=oFkx3&~`OHn>8
zYxHPx#j*2IBsReuz>97see@5xp+iZ|{?rxr%Zc1!{A*;MrHLs~DG-nuQ@01vNqf?x
z%iJFwVW!Ep!*>Q*Z$zRg7sN_K-*|IQsd<t2ixu|P?8MrixBH`ml_f)+SK(qHX3bAS
zUj*>Yn+v(&YDSz+8;p|XsFo6kT-m@JNxAHPM|bN?F>_;x6;Y^Iksm-D{lLNhD4u>%
zr~hy9#K`$S?66+_c3A(*dh$cPtW~$E<js%Z&mWH@+F5PbokW~SqGwXU1pvz*KYpd6
zCLGIRUkhA}Q^Ak;(Tu&W6J+~btnE)P*B1xZM%50-N$fq7huc*Mp50dz9QJC<Y=lXw
zIUXO&?cQS@EgknZmJ_Kxr`M}HTf5qnTMa^+p24f;g?-mvHM?;v({){@<gGVt6F5BT
z9Cx01V0J({p&dV;*Q-eLIZkM{?`ZwUozAez-r9~JiJ@rN7{*9T{7n8{ogr{73^Uz6
z<B8~|saE-^gQc5fn~`W^`42;nhpTcz_v=PF;4r%O=3$kNXr5^MDp;8z@oKs!?P*xK
z>ee>4wtRBRQA+5r$4YaZ>j1}*A&n9h3<p5dvNUXsJ_(&yW1XQRvx}j!&21(^O2Jvl
zzgQTe0~}Ij`*XjImV~^w)C{;QFWZCe4eN<QWfCa1@1zW~B4BjMK-~NrO;hWUCMR(U
z@7iBFyLF^wdA#OH<GF3Bt~lX@cw%=)qcsz~Qas{#qPml7C9wIpWrtUO=fInEO*#On
zkYc#vHxi{FDCh1?V>f%ZTjvj-=9@)3OCa(vhXNh0hJu?6eY5ZFk8Ua`B%yZ_OEt|>
z4&+bbG=nX;R(Ls$W4y0RpMiuf_*?pZsAM&TYB0=6IHR78waS7&NqZ%CXOHT>fM!V!
zzL&H$C*iZD)wzsfL%oQr-XK(rOt(p&8pzbv-JK=t!#bkSNB;t4k)$+Q9%2sZm)hg=
zpg<b774EssqQ?+Rj<TcFgl6fw3GT=8(quSu_JN-S<LI67b()U7*k2`uNS5Q86r?SI
zDk9N3(C*pa`d2Nwi?GlgPMiQKngUrt+YEVw@H{eNX}BvVD_FhV2HSWNr_1zFdu|-J
zxOJ#4<>46LMz`r#I1&QMdNZ&*Rgfv1FO4Q`YcWuB+BmMX69=NaOOu8aW@xC}FXwqn
zvrO>~rGaZ*vm?r0jOX<gA;>X33Hr(Wiqqsi!eoQ96(ed^gJ*ciLK2T~%qQ@)sLDxX
zBiglM9s?-*9|A&pqno0ts$6PtfghW`^oJDvV;{L)ixx{L=!t<%Vtw|~Qh{1;)q$GL
z+=2FPxZ0w&pO!p~TDQK}BXnC#^uFzRT70<H501Cf&e71Gg(evjrnE4-R%T*YWW$E1
zAxq=Y-eP{+Xh$pOM)qCCG-od@)11v*5k$CJ>j84(pd-(@EA3)db&PVYw4Eb$bS(K+
zlT$N~GaQ#w_R=yF2RyEMv5DYm#xKFKbj(#Aa)MHfd(7k%*}Aw$7xrk?oxD{mI0?Y?
zF12WMA?rm62{J95o22!Cvz-K-vuYH4rb|892=Zj!v3T~STNqy%ve+EWP>%J%EkDC9
z%n61ZH+%ld#mP9f9vzyBFHBLWOxKBW;%$GIO#an2mfpScck%?Z=)+GrpHNxxDa5L^
z#^n3jGL-c_F*yk34A8hBWp>jpUMs@q$UUyXYmyRKc<U5S9KQG6+_!r2JZ|Q~dQ5$X
zznYvA>`zB^#h3i*>E=0b8q<yF1z&@@W6_y{w0Cc=l#rALS|Cx$xa6@}ltOMp?X3)m
zx~M0mF{a@7I6tMSWY_}42QeZCD$FE3eRQ89ojI{X@Gi9AE=r=UjLw9tQt*Bt1Wo}@
zP+!6lG=W3#J`9lfn?BqhZiSlYzL?Xa-!Gfjn11i*001ggaK>S$2l%6$S<319FT9#E
z_s>5>DP{s3O<z7Jafs<Y2bBbVyf9@Z2Kujwe|soSx(L!8MxqXYC^cjp!2JmaT_iBq
z`Oat%HF`X(6uH4YgjG1G7I5Cm<v%s-v>@F>%HYI5VcW+A>}Rr>GP%%|?Hrda-ub|W
ztyU9pfm=q6Ey|E`0b3+i5qn;W%4!!{y-c0AV=#}$8FL~v(uDRvnu`6?GeKRVK4YVN
z1+k}%zczECJL`#JZM3mHoMs{UIx1qDZZN!Y9hPQC0^FpT(t}y-tf}rO=OvEhFf2?V
z9is3Uqsa=bJ!~I1*yk=|HdUPG7-RQ+EM#k(9{GI+g>HJT@Bo6{V+Ol=aZ`pdfBAs?
zhqfcY_3t#h^F%H{V$Z9qTYdEOqw3x-cEUTV`Fynx9#HKxJKw*bsK4dM5S$}L$@LLG
zIbvQZWfGN^waw~#I2^e};k@~$Mlr~m3g6qo9+nQkG1O&<v(jyB2s*r2t78ZlV9LEF
zW|&_~&9@P4!%Hzyn0ut^7k{lx@!AT+%A~rph+Icwn@drsaD=ftoaNV7vB8EJ&uEJ0
zg?tu1Y3=?h%;g-68PfmcYbXpT+i(q<<NJ>h6BV3Jze;soP!7lpEm2^62RBNGEb=D5
zEA(J(li+OOBHQ?Z8`_y#flN@ZRY}`=%g_|(1vc1JY#rF+`*`N<d4FAh7oN!?a*h`*
z_ae`s!frWEFdLl9d;DArtD`*MZ&`xC1%Dc66^wPY5FFaGI*$eysYb$<Da-)9LdVMN
zUxonsk006e<stYG2K1s^$b1II%gQ<kw)K+n1U;`rp#ywa2Qq8$@A+|-{V<d>&GRpg
zf&TF`sy21_&O8NT=oJQo-rJ}6A#^5ny7M2fM1uW~nV#jpn_4omG5ycF(N`K{b_Zez
z-51p98yMjm-KzBAgni2zfKC&Ar68Q38Jtsu4?kgDdJ>VObV?F2h6Zpb=&tH3_Z=PQ
zW5?3OS^DlXwVh#w^)cvwl2@iKPQ>=$BlJn{<`fT7kfzIypR6BF3X|9OyoODSS-{dp
zocnT~U&KJ86>9XX{BNqgE=SIVxrY@pijBEDw^els4T|9P$UHS<4N*ludiDG_#wt2c
z=88vL4(#b`JN?~r2-hK;>I+Uo<`zG{T+_bmB7-05A+>gA?0;DD+<To1wv>>V$Kl70
zH_}3V-&A-pNqH!9h_O7BxCK$L#04=RYI2+Ki%C{(;@aTac1f$j*!XaQuu>%CFp%3t
zyyzN3WWAdCHMjnumu*NosexRylhA^DxfFIIhT`DI%t{qH`_$_3a}=<kmo2{DD?%My
zWf1RiTwH*%Ql#38bQ{yEiVy4|1@u>v=$<SRjH>xE$^IlQ;rno^_QF;_^<(*8$D2Us
zqO(XHNIb32I?pMs`L6R<8>#m>B1UC`l(&(hg(0mN^otf09V3+r%o!zW4hk6RLV7qc
z7n|!37)A}fb>;zyLk>9l-xT>(-JNZ#rB@W9vFTWp$I5w-ShH#6S>*i_{z4nA>*1ru
z5C+oxkfGP!t(q+J1Iq&7%$f--tkJ_y4{SESSonPJgQelu{$Xec`W``3M;jTX?0Sfm
zlk=y$BUj!Y0W#VMXuFqP3W{|JT}GlBK?3QPDOVKHD;RU~J8!}%DnnYERz?G_E*+L6
zu486%T{6&74E@K`PMm`LruibhKRbE^Ty#`<xWp{MgHvrKA##IgoyySN2wceb>62Ds
z6p5P`=)1e<FfQ-TJR}W2K_J^_ZWFNU;EwnJ%`@`-wpe`ha9>^RrGuciRwH7uU~qAH
zd3ZOd1uf99s`hV-Bm22Lr-8l)w^oXE5{d#NREif@w}Qr3Y32rX&E%}$RCT$E@^n2>
z(Q<F>K2gycac}-DGhs(LMKjI<rGT>}h(2v8e$-khSLI-GZ!#(^cT2`<ok*9F#cb1L
zM&ZlscNY^3*Mv&N#-dOJ)a`{-(N%w;CB`<+fZUe)!TQ;g<1bP4th{krE|KWI#S*Y8
z-=`yc6fGpZ!4A_EwFh3T!%y~9jYwIH%v6GB<tWsUaqn792vI*t_B~D{l)KY+=Og{{
zKFH#6WM;iSA^#I}tG?XKTBK;tc)^RloK7;LGh&Cd{XdYJ3H|b8<>oFDc1^$<`dQk<
z=p=2nD4BhU1aIGml6%O1B=W3r&ECd^D=IS_2B;l$WR4X+!=7e;`Ly{LJ@CD5+-?ZP
z->kZ6*vJDN3-ma~|2i7`?cq{&tLUcABvSfV%J+9L)61yKp&L$0AYoA^N!ngiqvkZr
zJ|^K_Rmp01$9h?nV{LX7vk^kow%U|u45B{5B{d1$)XiO-;gUK0+1`uw@%9?=K$*J<
z-;(@tCg>(b2}5Exa8Dlubqmve%w5g#lvXEw7>B4wRdWL+E091;(?WXd@C^=P6nO{|
z!AzuRF4^p9#AB0Xz8rZuAcmQCE^7>on^oLPhtLij_2IFkg-(~5B2&#E<ufW<tbT=%
zCCPt9^AWjzA#?KEs$Ty`wsrFPCrOUC3D7vdzJ%3<(OAI<h$Y(P8%j@Tk2(Hr&)N3N
zF*;7x5s8)`wse#Wbm{&SrRwjN!#F*3r3gnVWy!Uf?VW`ltn%dqO-uognqqEelhO6$
z9g~2_$xGYIw<BMiV%XIu>!YZf*no1L%a!uP(P46x{En4N6yX7-@P!H`M$JA6%n3Uc
zl_7dtw@Q-0j3Da73m5y_-lKKPe9pR|zLl3G6;JWWHU0)`tTGhld_Tw#N@5d3gop?{
zcecsd;0ddfOFBAA5yboWkm98HRXg+nk`7(TEJ|XTn+r7*ECj=P4{1;;Qr{FLOUtO$
zw++Qj&aC;kmtpc+9k3A+d(lk#A?SN_%ys~uF%tSIbI<qHX(RMg#p7DIzMVwgUAnDs
z9%DK7ZwNrQ$qt}zNwR2r1L@urNlpGtdeMJH>dEIZ{(*kPbU)l!FsIlL_Lcv{L}H#x
zmAO{f1-o2R*ah+&p7fO8ToZL`VZnH~H|p(xTEq(*h9|l(6Is@ItRZh$#l8CHZz>80
z{hyQ6f4Adc<oqAj@~<?<>^8(OdM+qevqF}W*3yO%5U5oHwz2-Jl0Vq~zbp9{FIh2%
zFzkwdFR)Gp|B5($`~Fxyv=>ZuZhZ}0FG-FLBUGmEO$`_i+y?AtDQ+%SFh(QNYCPN)
zzd49guDos@5L2Q8;Ro%Ysk^qJPVU2&EQI0O#hAEnb$+JrAX4S-)8)yjr!r`CSymO`
z;ngu77~=8qeRdT_xJV@z_p>EK6ju+ca*t=(dvJXpesSMgtx?qKAQbBs#*pk%JX8oo
z?7s#ufkKgfOKU4AN=s7%-eLlHlpsw~$J74{z3mhEoQZ2|c%&hAxcdqpjY2S=w?8t(
zh1bhvT1bgAS7;HpPhL)TYo8y4F3x|MlJo|}lyB!`Ank^UO*oX0JIvv@T6;-zU+T@n
zbMwCjGcwJ@*!{VhnC6d{LA2Ew;B=H>+u>u4>(sX@)0|jh#jsHzTGA?4ipZoe#fGhl
zH>nH~@;iZo&&vx$AsRrUSg$oO8p?U$6?RCp;L8upx%Je>WDpUCd}3&eREXmKXMl@)
zkfJPz0n>b}DykO*n*(K|taOl#6snC%0ueCp4Noe0B-=8-gtRx0Nd725WDkyX7QKK(
z3C{ASAS@rIE!&L@!W%KoL0>TGNOT3&9}vlLAqL4E`zA*H8G(_7BcciHTEWMIPggW`
zU1oIH4jY|yxCX%<Uv;d7#*jh;)B4U}kN0vH&iFg*YXfjk7d!h!g9b~ZlP1o0W6ycQ
zj(?n>DTaZEO8$U2H4+E0Z}slCv%}>6FtXEUX~*Zgaqz)vXX*Q}vQqoXFAj_zilM9c
z<H(QAKsq0Z&|IM;;o~Z_o1t9i7ToGNWX|Q9rvxlZ=5y{evHnU0nyaKZvVwq6Q1}Lb
zJs|t4G^QWPeWN^!3opWXN)>@ns~s0_Ul;^d6PutsU2&Vn7_5)G?^BAel29G&08BF?
z=SL_}zRaE4kJpIxtPR{nr++evLSXR|N1l!cHv?#-xgFJiJfF1=Whfs>2;CBl5e$`S
zy3cx&$m)q;cMbg)imCMAC`q{9%;f7?z>V6-dpsUhWO$evZk&DZ6`=RZDwt;L*<anH
zdLq<8j^aS3L+KhPLAo4Yzm`O+ri?0@ND01ZH|kLb+~g*pUm-cq)Hsl;zqqX67^+Y<
zTZ&~%o#x=f{k<jSE1<w5WEpH3i50!#L{W0;?vN52T1DNn$|ag(^~DWOTTtcYf>%Wt
ze||B(gl7?AM{dbpf3JWNL>+(}hY&nmE1s!*+25y54v-V%ghZp<$t?z~R7_?ciZ~g<
zCAplD7zV6OKv9l514|#>gDL~2T93>|T3aRd7dqNBgtX7f_J=E?fVP@I4pX#cEV6K%
zA~IH9Tp#{CzWaJc22)&RfU@)<KX1XrNq@^HdxX-xOj#j)IkBg%wfLeCRN<f3YJeW<
zRCdH&=$Lgl>}kox1?1iZBpul66OW(E!8jaEFrGNi1<*<w8f)l=)A~RVSkM%0#)ZZq
zgm9E16VWKEbLmjul9J|NP1(Ad3KDIkqJ`ot+uF~d&Sf?b0Se&1J7Eeujfu)65y&tZ
z69X{SXx39!@5}Aa_;)GF5&2iOWK{B=jkHwG?N(d<@8o2}*p}D(yV4gWt4-yLb}p)k
zy+s%3nTWkAqWkjDTp=kF(`R~E-Vh;2`$-d{^yJl}py|b3)pT||m*bAw<C{Q1bE`Wl
z32iv8ZfI=A_|eIHX4W89_mVOWHVhXmaTsQ6ipErj8ga@_v-t?400>&P9G7RqT+TC^
z*lhOSkEWz)2Tz@Jy}vVxq!-K!(mZ=jFD7oWl-=-ZNUwZILuL{8ANbO?o=l3jg}vz>
zmOtIGY2P-kI(mUt{klWfwG3IBav%eg@mkrL=7J7r$56&J*@9#+8{T^5OYM(@26n^s
z>%q!?Ww2Z7m-DU}@x`}BiS^okD@Q}=#1C5tb-GMn5z)3D8UiXBWPz(i%3RVuAqv6?
z?NwRDe}qlTDlZn0_pG=mp;eBplEGYe3Aez1N!5(>CP$2;i~sOC0R?K5jOQ;Vu$NfB
zBT{F27E6_i;D`IydWLO&s?Nje0k~J>1c_@>MEv+yL`SfMjybIuAGQ>!uC+86b@qMz
zR$_XIi%Yr}ayeV}Tnx(5erYvdYxdIGQH8~0EiDOoV7@cjm7=Kzz%ta;OXLo_cmfH)
zk!EKpB?G_2sE|T$(Ja3BSt)4uSP43Cg$7!&%qS5HXJb=q`B7TKi76e!(wS}ff^+z#
zGi2H_j8+kb?`&wQSf^~vk3r}PtcfM)3aDLZO@)GQ0k$27ehr{a2VPYmy;Zpsdi%O;
zDnPo+5~@r2)=aLrilz*#@k2;-<&>6|HR%)qb`5FA;qd%R(5m8eSCFTFp|M(mT}HI3
zf?68Vs`mQ5sTBZpRmE0<d_P)HlEf;w^k%e2v=WMq#k8pizp<(a=*vp1;`ElZu6U)M
zw5rlxLyvcfs>v*a8qE2QhNvlM8cXHUYKnDWZI1vy{Zv%6bylAnRev#@S`(jtR2T25
zA|ZQ`eR^<*+CNI8(&T-?W;Xdy#dEE{4oelqbnAWZDQE0DO;C6zqRQz{SD)ji-G7Ag
zJ6K3<Pp-2ocoDoV`~jNfU_bt&AYlCu*lQMc*8dT|atUi97Q6k9PrqQ(EFhgs(+4Op
z0CSBgUOT$o)(N7arM}NA$R^&$KhnOTL)%J{7wma<ftXFR${n?G)P2Xz=jCkg=xOf@
za&eleG$f^SrIz~-KF2pI4tUJACSf{Yz4uqohviH=%WKzeAqRNh$o>!9&igG~@Bn82
zp*YUACh4<n4Nod({sPt6#Q7(hCwLh1uSr+mz*aEq%-&6{zvnKd>vR|7Pbm7=)7{hF
z4txp%qzra1^E%VNq3gUI{C;p!7Ut!fA``u`ts8weef)2!g4FI&?uIEy6Kr;|i$AvW
z4AooH7zPHmhS~)BgxF2f_ObRP8XKR(fM1Mt%!p$sftCJ#KP|+>dHk&JG}lG2TZ-AY
zZvq;jHoMW_7)s=20}w*K{`WqqbmV`z6gm_P)=j-!3QSjBBb1R)VY-SC6yLK8N1Nrd
zl}ej}oYbAgfTTA2gPHNkvdgJas6hWpmy1x|8E^Z`38KZ6o#r;lK%MTDr%51V1rH(G
z&es~yt*0u66l(1gH<W<*;Vj&|`_YLNgO}9^R6iimch<(|r(trPZGRv~t`YmrmkH{L
zF<os6OdPC~c8N^W*CXHK|GYfHJs$A;`FfGR9r*Y1<ZLZKmjH?Z>^KxR$N3*99S^BZ
z(SluPJb~=Zc#d^-31A%aDtH!9Lsyb14Z#*=A^!ehtN2nOO*R;n>UjO@&yhW#%`t#u
zbqK*isj5Z+|3PDg>__=muVON!$&#hr#%~TJ4pdB)DtFsaOr*On=EC%+@<7^wWqQn@
z>Xl<QKH7un{cV8=Uhwaiqv$9sphzX6t3+A!w}s#{27a4z6px6Le3-k`sBaNz#??dt
zmeqk?zRq)UB@<SA4HdaB`J*w@cX->i4C`36I|!Mx267m}Xp1stp9)x1M+U04x62d;
zJe!i-L59FEQdyx}WWQ)iIZDR<bF;%d`m~5SrT2LxT4V}OlyON7guaa*75h-1tYTH!
zIq(j9)d(sz2V5@mgrRXPH3N+ACo4{^hq<hj4&H0IcaJ(=j@V@XYhdT1oUuajHIo@U
z86)ySU$j}YbSYj84F%}Oxn0q!dGdVOjra$>s!v6x7=LYgDG-+E(xQyHj7oW6l4UQI
z8ncx~$|hO|s47vQv|AtKkI|*5=!FGq`gn2CG-{>s8=nyjs_{}(3&;C?s;1(+-$^@D
zB|v()(Yo$Zy##6dyybI$Bc4k|OOciqbI4wy2+)>der7D_Dw84p?)!V?WzK*x7VX>R
zt_KvpOJepSUln~JET4TRSnS039Z>?0{ZxCtk;<7x`;Hl*DyJE)>ESNo5vwD`l)aP_
zMXa-!Je{anNJi-(F@+%A-45c!ST@&z@Ot#AzsWR)oa0xO8>P?0c{|t#hiuoJ8WN%e
zZENN=9wl`-7>lh;0Zr5yRXwg(FMrDBlo2rEsyX@P#g-1!*KtRSvUw0^SZ{=9<$b7Y
zBy_AS2*yD2WI{?uk8^t<(;0{K?p6l11>u3Bd8jg|lv~G*{Cdt5R<YT7B_XcAC5Nc`
zisel66E;I8<fqIzI}mP~l0!j2k`&He+_Y+t;h{6#ISmTGlcKou7|*pP$y<Rl?wVK;
z0cl!;@Mo>9j@Y|U>~IzAr%9Uef&RjZDru?%4N4*qh@Mk!ksHBAy}Vn5<6TCDijflt
ztB%WB&>qtOjDtyeqR#OPsTno-ZDCvBT9nObG$}&q+t5>Us}3r0JZWAv9i+_jKPoQo
z{!VHt9h~&O>`Va^qn?fwYCD&oeVW*q&L{Utj+znbl{vNjVx9+s<5!^v{x8M@CvQ?^
ztx2NEVCVuAT=)1XE(L?8%dR5hKGvkuP>4mPsj4Mm80{<hL2ryR*UHsDM=}naOQ|HB
zgsB{^u_Xa(*@Yc5MsuiPLEoWh+(IyVR3^S_z(X>xZdtg7`VDK2&uDDiLR}_}!L3;&
z$CM9QQ@2iteq!~R0M`aU`2nP%r>-&pK~!Z!A`sMY@}Eamn<!A>9}$5NInlChnmEFI
z`z2z1aZ|6A<y*TTeN1aqoPuoK#Nl$gHpAOO-U+6oQ}Nt_@j|8|;B(u##3aFo<3O7@
zM)l}y{|L`6QY^Ahq(L|*G8a71V9pqGhf$zJusUVwQV?+#Ip*4C@px`#wcE;pc#+&I
z8TLkWX8%edg4@Mlg?oQGUnf(r_73hNG19#)EF2StP=k%zV=9S-yT8_HtPJ<b5YWBW
z&6kiihJE*Id+9(H%w4G&X=|05JZmL^2`!!KNEtH5jB*o0dXQD?OQM4?heGPwqeueG
z=x?q=+L4JPrH{;UhXMxEoKJZI4*-#81k%c@gY1KaiW2BC&`8Zaz_V4kuA&VPHL|l+
zZOCIAfGVrm;qPeGl-tGGGq$i^Abw;VTJ?Mfjl;r)DJo}2l!~g<4JNeHpPV-o<AG)Z
z;$saJr{avUFo!6a)ETd913Rl%6LfiYlJ3K6U1?WyK1n<6DQF_`+*5wdyV9NUoZC<7
zAC^CP%y%QNxQaV@Hr%A7bi6yeLmvsRL`hM90xe&uMOurlXzA(fy8q*`Dp^-ycD(?o
zP@bQCut^5WVUltbIes&NIx~PLlA?VM&cAfk9#al9YCC}MXdpxFVB5iY?aSUQ@*RVF
zTwSY6Z-S_Gdwq+SWk_!?b6r7gK~Tq?)jd>__1|09YJ)F_32|C^V!#yYW~r<V%jtYV
zSr-Ee^GbhOMt;W)YUgQ7VEk<=erQ<VQoVhT<oUmSAGQ8`-q-F->!m{t!N8CN2b7Uj
z-qfu5`Y&0$xoN5WJ6M!%?}<PW0=}Wbo5{)Ja9a`IuSgm1^6?!g+~<<|u_}YMB_>Tz
zOBB*w3xyQ*_vmDUIW~hMYqbFVYsG-^9Zna`&{ySZzZp`{xw#cW2=GIoo9>`HXS*Ik
zh`@f?k1&c()nW1shN=s6CnTD7{o|&ap5k#tHwmM^T?Z{}+9Hb{DzKFgO-!J}`9<tv
z2;x3z|6id`6R_l7L)i@|^NAtH$h;Z8@}h*DyNJ6=f(0B9{Sxy|bLi(+r24sz4o3PH
z6Y6V?Pm__5=ph-c=HTh)kX+m<FoPpg`$17VY$`2?yI`tCh`S*~NQ5V$bQ@0H<>6o7
zndXYYszs>01kpLXf61eTDgdpUx!aefB%sm%0;);DF6fltrmTPddwOTfKa>fFBzqVx
zAt^Zzl?vWbk7_(^dM&9gQ8BAR|86-O2W5dfn*LDLZEt*cFMGS<MBQvM<}BqjmbR6x
zPUyNrdGZ$(N!k97%8%{8D?dh#-;4kI`2TS|yVCl9CKG<gIx(jmUa6u%jmY8tYcwG<
zcb#OEkaAt`{ee5V4BWigOG9e}&}%hC9qHt5DrkUyZmCpC{f5pCpI&bZyP^vUjYlmG
zpL!QOvsdsz)|?tXKl{)5o%GX1%<}20^926<I>xuhH}~xlUJWdWAF$4M-`+a%=Ht!2
z?ni9>6jk)qxlc$H;cXV5N6ukAn*l8XTrcmpPeJ3E)>^)7SrgMT=!iNxtD67A!W}vM
z&FTt3Vb5R1<)Lv;rAx(&N>gc?`!L!@8^Sw#s_J+PPH8Fw>Hy#6!D(%_yj9qw;EZGT
z?K{-_Mm0g3ZF>|_2+ygk@Ae_!MYkok>_{A-`jdjDqo;1^-4Bx#{(xeVlN!cS#iiHc
zztT0u6Wn(fR|iA-=;HB$CZje_A8ApK0ew^_#vx@y3QsR#=LbXdN><o{j@3!^aPe-_
z>r4W2Yllc$ruieSeI=0TX4#a6nb`|(oGdfv-e=v*nnrh9>N#9PIbnhU#HC5(mVjN7
z_zCGSQEFUc0vi&v*NYJtPi1+GWeCxW^CwaTXDC|fQmNU=e{Y6`HpjT;z?`yeK|r@%
zM^B<lX3bJ;BM&;{6!MUhsfil!TvSGz)Bty|#=AZ}mT2z$Tp|{rMpLC9mx|Cvv}RMJ
z`I@CMj8VHGNk4O&Py3`L-c|wJ-$qX8@q4}>-ezjm7ev9KeB(Km)>c~*TGlDeiC}7K
zxp5ec)33KniLP={(!;W7me`a|UFl_aV0ZFYzXG=tAGNB951VcCq?q;<Ci<<omUJM4
zfUbpsGKQ2!?v9VfqfWP7Y%iC1*`vyBo<o5*S>KzURyU&mOv~Q&*gsom5ox-uajl^U
zCOKv6E$G&tAwQ)}$(FZ0#B(1-z$J1QH;<Of;I7+e+EzWE*W959K31y<f2r5Dbu(Xv
zE>jxsC!}`b^|G;}Khr?OR;og6w*0}X$&o8ok#;ykH_MQOYq@g3yV0zJeTmss3-Cn>
zMO#%ef~7kCTTWWL(+9MO92p<&<&!(Mk>l|J(RuW)hz-i44H*GY_?~b=Ie@Ei4)2>}
zHFxC~X2+yuitgy>#DKOg+O>ee9x9EA|NX_basTKD(5-qHt01B5Tn+efz8i3t+(j`G
z-2ycxQEurbwV}#RrlI_#Kc&Ol#XMQv0_TAt@;I_KFH(bNE}oEiPpQeaT*jtKBRWlw
z>t3ZnIJaF>t1$Aw<w<#_5^|3=e1iS<911l>s{z5I2w@u(FQ5cx`)-g+V!=>cbkee=
z`w`(=^sjijdgX=lQPGn_2yxSc7kt2&V6>Gr3(DSo13&y?Rb-WULn=c1?={g)8wF`D
zly{bEKQZF7(V?p>>k1_UhVMf6*=*i=U0ys)RQ5%pU;uWbVKuR51LG<Tj+tdF`;;sD
zi42usBa2h$3=12MsNE+-_c|I}2M|Y#Y1cV~B&It`s){SN8~*;cq?!6XsIGj+WRVkf
zgX{syRa1RT)?B!bFz^)FUIDgi>HTdFZsf~J2^Q=X2H+xi<3`&EJkJNR=U^f1Fs%sQ
zoo8;Gl{b=S&;UEW-GxOCkM~TH&mpQJ#T5_wn0cEVBj4}owaCbKrvPg6SRjiC0DCfT
z8qmnf$)=IP!5j{Ok>W(cN^=4biDF+-M`CRdFf-ZNED~4;hic5gbrTI$FL(brp7y?@
zsmvJRt&gRHgko>ts5c3O2y13<*a@<Qg*1kaP|gtiC^1ZO8j-fMA33F7XB2IibHq_8
zSz1XDrnn`KOmSB%zl+h@V7ETKS)u)~Ta4Q>>ko@(N>s_4)`0MzD4=Wkq#|m0eyCXF
zMsnUSc4A&x6~YvK%QdxmW&->U0V!PxFRX18&5d@68IjxgC-EY3ecBnX89^^BT#p(S
zb28rI)TyFUgx8$-+|gm=HC*^+q5asR(s&+WugF$+0^X28%z5Q=e*Clc*%}$I*j*>F
zdF5dn{CQ=Gtv`=dSFk(4XBlEKi{<WBt4jM|zn3!!d51POB;^SmLLZX_dNAgdI~ZmE
z>se&Ha{*3n^1rui{&_?}GMA7)n}a6jP2*f44FY%xNy?qbww9DX+q9B0<Ru*UHVZ!<
zokk8WBL#I)VHOr@GM`w!Q7F4eTKXA0C1K1oxXvfGi^&BS=ld*qxNkiKB<EaGe*sB3
zf_lcWm<!>>uM30I`A(N#5IK<Mq5tDF2Cu#o`VTa>;*t!kb(?j1WcOHR=pW#>C5)Zn
zNjrwlE}B7J%s<(cUB{-aOwQTdpns@fxe@<RwQ@&yp(wL3Hp6w+q#-xQBy7nXI1#4d
zVKa~HZM#L`)F<y+#W;02c%^r+iVGGQN4A}q0?LV2>V{koZP(9r3#>CchfYxkQ2~mW
zYmF;Iskg0*F!xj8MQR2OV%tDtWh_@=?6aUiawJC31vj`t{sbFl{Zv{h0Gi&>lGrZv
z#%~P$D5^GDZ9{znv(aPX{ioc_$j<hErUw5%Scbn{LsL6*7YhO=R`&nd_TZ70wY~NT
z;#ZA+y+CgfewX`6a~u&w{GcmVpCBwZ=sK1V!I;aw2q9>M)PnyGv$tp@N_j9+O0WQ|
zMy7XjT*g)gv-vX}={RkXKm?Ruq>E%wGFE0YQgU1G-+VK=CA5Y|I9Z{&dTYp~VYgE?
z=QJ&XeV0Ss+3@Pts$fa0yyCw>iuM=u_p-pSDvV)KXz9b{3?$Ci%!^1GTIkrI(@c>t
z^|Oj{vysq-$;mL?ktuY4)nyMfQ2%y7WgAd@FAre_muW*(4#_$p+TV3V(S}z%w@^yd
zcpVafmM6b!!_8y4GJ>08iqa6CIO=qvP9FRfYI4*GS{-&IEsK<g>8#%_rj}&2sF^pn
z7L-;`WGq#i(k(J5kKLk4k482&m6ckFmSQM&WC*qB#+GKdfF9EcN?8=OLMYxCrtAZ@
zvKJs^j<cmTnbOstM%Y-*502&xB#Tag?X*}ej%Rl7o@G5z#Jp6mX?{V6Sz6~%hbDD3
zH@NgLyeNzeHR@SM?T$3;8@`P+YOVwlmOk#mxgAg{c15&XYl=!FOzpwZSdLVEYEKq+
zH4pVym$;rax&C+1Z<uUVOkL)oZKV#n4cUs*EL6MvmAD5VGYp9W9Q&_7y~{MIaDF$8
z);NZ6Zw3_1Iw+`xiH`cYR`r_LHK_iR?gS6NkgYK3F?<iti0Qib4SPwET^-GOfB1=@
z`!3;<m>t^`zx3DZ`YIvAYqL%qd_7LKLfj^f*DXszD{V_xLfmyTi85`Q{lh!m<MD1}
zVCAPf9VSn%wDIls_}%4No}OREA7a3~al$=mq7@P&DK@MZRv#{}xA<=F>ComoHM&p$
znlS>F4)Va*31A#2Y4z^bvhUVDl-^8#CiG;lFk5O~sp?IBY+UsF6Y{&7y53ihMm&G0
zYjO^_s>lNFm$%b{Dir}yd{D?r9V(v|lsCmcS)y=)U)%fmgYV0pr?2y6iqZ4M@N@mn
zk~`&x)6@HXx4+goCO@D~{&4f{$N1yKm8a9g&zUuNGZVJ(uy*E<T9pGt<*b;s`r-55
zllsCYS(=o+lXDl5(;fEqXxG-kgT1XU|52oV`dymF*=kM?&1l1i73254{4dq;<YF)H
z=iR+;8ntci@%{7D!yEeY_oq^V8jU*sq{R}0FE3|?ZbwI8gJ7EwrdIBF)*s8m=ksFb
zfqr{Z{y0hCAPT+l3@51z!M#88orWIgm^c1ic%Z`@<pumg&g>wF*vq!ij2_3)x-bho
zz(vY8K%Ng%Pj`BHTjtn4OnpWlMql197r)l#wKB<`d=|TsCr95^1h>?Sw>?u=KUcQK
zPGD!pm(}xS@33)gW{aN5)8L*_tW8AU-mH*rE3C%ONk@I}^eX*$^N?|hwx}guY{Ncr
zQ)s`KL|QT}k+!%i`5kA(dkCeEgiNHg1*P$XLsk)YRXGx%KPFu$-<wIa2Sg{0Y!s=B
zv<$B-qz&yhG6W5Z7UaXB7ELuk8|;z}v%Qw|ps5$F2<y)X1HYO)lq#_jb!)gwV*wC*
zd^FT|5{)5ey!H5kW&G{LX0^19gpAZIN%NHHXmjyCrHT{vD8p!-d>v*+9cF-LpS2j4
zawyxa93_&dhcI+CfmpCFz;+;I(KE^~$eD#^NC<L3sf>T85QH(CN!1SO{*Q>7<%QoR
z-1A>jv$_J%^(ae{iZGBD^`y7($3D+`?R5X>_Fa<tlLu^PK{6vta7$W7-hp45FhAXP
zKyiH<cJt6+&ap`FV`vg;9eyx(-AuqUXYqV4%`qUG-k<~6a{wfjfO%9}QXml;u$AzP
zRm{C#GT7YFrGNberiS<tgS#$9XN_Cf)M87&YMq;e=?2A=R%62ap_ahOxST8?*DggT
z3+mLI0IZt@R9Ea|FJ!5fv79x?L{ti!5`+U43fNWXRW(vo*gS<?i(p}wkS|Qk{rP*U
zSTkvil{D588cU?ImQ+)+hrf~tW)?(Nm6nL!1@{qaDyy(w3+<RVG!;dDU_p^xbLt~V
zUjT=A`$DQ#I5=g?H6Lx?XrC@K2j(h3ypZ)4G`2#)@L?D}`)<BULgEv_oZ!{k#3u=+
zAji952!cdBesg{cNihM8nHD?xm(_fA?5S9swyopj62NgFWZgPy`=~qjuf-R<{?!=E
z_sIY}J{`~X{^k9~zr4_s74nmv?5UC@d%2REp7!*dJe}u}FO48_xhLZaY;;Ep>oW#R
z;hGseHcfeazlxWMp<An{jY-3~I5RiP$jJiF(h{io1VNDB7gX|#C%(zzRFirDKLMYg
z`V{zV`xKyl`?J7=KFfrTaFqv0LS4sZlFYa79TZsgYRJaKmLk3!2rr;lUSAWO5|};l
zCscJJ-!%DU5E1mB7G;YBc4X}UO-q}TXK3E~wE^T9or+8Mn(Qfbq;{fiCK0-*%9{VG
z;AW6;UTZKPKhvWp>!T&+dl!uFHrjh<X94T3Rfnsv^)xU%^K~1GGrNpw?<}x%bh5p)
zfGb`iJnLiGhK#Lz8zrYydvgZWh!$Mabu&o7l>5KXqm^)I`@k^6+KqL}?`~Ptp)lJ$
zj*U#JbC7jQmHR}p3H(<~!>fsZ>+GQFwXx=}?&Bu2{z^S(2YIe3`08X)Hbw=r2XzWI
zpwONXA&Jkz0-tL1-e?e`*NkrI35wZXP2KQ(){nRCyss=M0Y2%kX&AR7F5u9??nHPE
zM~!&iDI0E<<5~)b$2&wWD}P;SV{f9mxaT^)Z5>!$Zlb=x!)oSJXKmx4b#>nqPG8{A
z3bdc0w{(o;S|z6<#cZ|>xtykRM=tydq7o^dz0soNS%RCk;gV`12x2D0rt;5L=3k|v
zOc&qpOty=QNl+5S!jPI?!j8S}yUE4*c`Po>X07~mX0^0n*M|(7cFr{=TGU^UfyjG!
zZKjGvQMqLF*@_4$8_f3-lt{EW4+YN0-xy%sN(eugV$=|tsm*aDjmN_x)G`&!DOWlz
zRNaWMtf$aU@`DdaI$2V{CXu#_DaW1VCwr|~>yeeV1Y60e394*?GE;N`rtAl6fVK2Y
z30pq>%({|IC}{<!FCnMZ%7;<`9Pw_AzRW})qEc3;Dza9jciUG@LwMCF%y$QoHyr`9
zXqW7pnOtV`zX`LUm8j42GTD>ksYsUsFPE?a`Vz;Hecr<M%9{)_iEi;eK|@#&fCj{q
zrl`e>NNANMD+c|}kfZKUqz&h8YG|8o2S$@m^=P4dx>A;{6njyno1Tycfi<%6D|VbX
zEcRBa;S!sjK&_iQai-5}9otgSq?3h@NV8r*T3C;u^l9J0`!{2smgrTiThU?hqCGvK
zkB`KS^a@lbwbk_2QkA~_2_2~%%$j}W<Jv{F+Sy8s7+Zmx+c?TCJ8stQJWOpoU9#i5
z;1&#sZ(RglEpC4R7(UV&|C8+eHLm+Vvh)8|mW-?%|Fe>i_JkeI2<p}YH5WG$gS2a<
zUAIk!fJhWzLnP>RHUOZJbUf%<_*Do$e`!ahcZppur>rJmS30Jg{E9FCheU_x%IUGa
zdM#e99*>4yU}S8dCEgBOFh0w+<iTHeSpubPrH9x0x9j5N;e#hA`sMXsD&=?VtBZSU
zD$zB8zMhLacDi2fHS5b4@3L#D%(;A;JQWVnH4Nja4E!d94#n2j#odj1M|h3u{oBkZ
zVQ;rz*Za*<et7-(+akwIBq!U?n@5Vjfzx={A_89y{1iPui+ycRiV3(GmU#zzX_o#r
z>;8C_(L|A<^i{WW3T=$~^w9ACV(guQWDB=-+gY}4+ji9~+qP}HYL;!=wr$(CZTr?b
z_uQAe*Ta6w$jE<0WJZjL(X;o~zxGjR^<CCL;#<%Zu>ARNg-540fgsO#ChzNFu(bzp
z7zphosmXj0BvD`IY8SxtwjT|Rl9mpaP<kAvi;xJyUxunYKoNbqt{8qMFE&m7jmBk?
z-T9?qX64$ZYnp#f;r1<qf|&^?P>gC7^mkJwBftR51>JKQx=Y_Oyh@&c2~GPgZ**FH
z^t&pltKH$cUd%ckVq8Sk0mhZA%1Y1uWz=Qk4a*wEFFsd{d2A<_c-2Vc)9~))Ypi|X
zW!1yIMJw_!kcoACNSx%N<Wk`SY)2POT@GG?8DnKkL0dhP4k^=suU>%Gm<G&o9F`rr
z$Wk#<{VGwPs*KRMQ6&0oyJecSq($ZWeB9mY$z88&AAV$R-aXoET0xT2c&r$5$rwX2
zQo+lpUj=xp;_EFe8-W8{!0`-&db|*gO*b~n`>pD9>abT1%L2>6-oodIQCq@ZGdV(f
z4F#*EY&~d0$b5oD^QOj*lo&-!BlgpYP_#k;-@YL@_wl{9dpBMlH_b&W6(T&v4Ld8j
zF%J2}O9xQKf;cn#!)D{P%+YRopY>1c;`w|{3J%{&=}SmI>(RSf?69<Otxg={oKf2S
zubo`PO;rg_ar*ePL6gT636MLQJ7r|2Pw}hu_#*mzzKyWs$mnrA^-^s5EYN5q+%1ub
zgXklgdGY5fYfS(BE-ogGszb(PUCnl@9rc=LjCavd=&+cLMnu#hg!sbGgDtX8vTi+T
zv(G8gn6woXS(*M?Y+<Qy<*WnI_sdCuJ3IspE^<>C6pUR%X>YepVfQ^E!wXr|?Om)F
zp*-gE$V4B5d&p=oGGL-KezckUAg|S>j^k+}BX2ooWzQ#~aY<}0WqG@LE1XHaa)S42
zc0-D%5cjyPbM0)|QP8eLarQ;W>BT!c@gl3&5e_g?G{uo0{Cfo~C*>_x8Qk^o)mW5$
zVtY(kVquylD?k+3t*r>F0P$`-o|bnb$g9{f1#-}gaX9Lt=7MvNm&&DWHaVfX&Iqw;
z=af2XedFat1`Bz;3purWj7~n3PR)9JouoB$-)=$8NK0u_Mgc=5m11X#H|Tr-&@wK%
zgrwML0=|89q!!2{VBEncJ+C8X>FmNc#o&OUjA+rSk|==TPSNBvQBvEsy*Q%?m;j8Y
z3pvh-K?6(H)w&p;Lvs37={t{8irlYbmaj}@-#U&w@Qer^H$%|*Fh|d~!HA|Hp`Kx}
z#OLGuIJvXu+dGY5Q)oQ~{W&;eZc-25Xb*RRX@86e6{sGGvHPrTl_Can`Uz^-&J3^J
z?eXIU&-e9g<%RF3u~J3ogdbjMvs;VRpkK-+o_@a`LqQ<8o;ll|t%Wr30xsf{8j_d<
z0jG!V{+CfgU@edKQk#Ud#7<XDoenqOP*<YH0ZyrzKuAZ0ZsE+En4q15AR)iqiqD~W
zM{aU=AU*gnmekua!1%{UyTOJ2J1$%G)To75nm13;OO1>wrC8zX)}$6>joYWM*Xful
z%DWC+0NM*$W@ZCnC^wuYILoV2Dj_i~Ohd3?zmh3X(Q|;fZK%>S@ww3C_kI@`qu^XM
z6l&VaH|+6&Zg$;zxPb*c3+~?2#CYo{*%z$HS{P<8AlDd6adVG%43)veIaeO!Zd~-E
z2stz-CC5^vvg1eNB*|ud!B9Fea|@id%O<%b<2SX69KU#-JdHebVBjPI|IW!}YSS3F
z8sQ17fWQr%S`1XpO?y8EicL%3g`_8H<-?<GkS+(Dh_t?sERDQfX2_M2e*7af%=s9F
zRQio!$hfR=DJaTIgNFQJ@B*wPBB$YJJh+u|!07*<%VFA(($aj`ebwPH16*bl`PsR9
zwPMZA-^^d=Ydy<-e7&d*kYqv#-~IPN!1Hn&Q6rwVzy=MZ?PySg2>7rP<;v@zk-|IR
zh9MfHN-M!cGIHJi$dHp)?*=Lq%wUsk7Hi=_&$oi@1wK1(<kvfdR^@ATyq@|8@H{d-
zfKZ+sjT_c%7dz-g@3RZg`FeEe+Fatg!zDR{n71X!TZua01(IUo#d-Mphkf)39RQX?
zc&o%N3h(Z>31IaNosZ&y&|)ga5Pi%IL8HEUjCzBrt~0S+@guI4>WP8BX^9cOZ9ZL+
z2_QO(KASY7&S+j@vt)X}x&!X#{KFJDAr|JsWMYP{CZhFZ81T1+q(svX4vdQE@+35I
z^YQ3~7S!s89z*om1jcYsAQhq5*H0xN$@xhVC>p>t5*yDeR}p3X8!SUmcdm#<AlMIX
zxfQyHX>_UtF3`X-I;$^&Ai9h%SYS0dJo;~tcMx8)yN;y+Wsv^ZSDOyd28JOJg*fFZ
zRNp67H%A))daQ+|f%9IUfuxcEB_Ts945O3208wRh=^$uyL_8E>$loxd-g~3-dxR17
zq_KY-mbNsh&$fRb8#tVhN#(BqwO8;$WR&TwBzm~CaV^X3d>%6tsU^C;-KvlJa+V>3
z0Fv@QHsBaZhW>x0C_1Sl;2S@S0eT*oBIGi7fe4W0InYGNUM>28_p_{#^v0jO)<NJy
zu|WjazQ$3=i2t>CVgD~NB}@$eJCf;2Q#=N11m)Mnd-{tqNtGq5X3)UbilqPHB^lqu
z1_HH`jK$%IM#6u8d$Twry?7Flc@ab11qv}}IKPtYdM97&x-xo6sX&8Mow<f_)&Z{Q
zyvE&jg47lJR7Ql$y6RNu$5`U&yy<qQEpl<|>WwI52BnMf1?$`xsFDYuO01xBZhb1u
z##=e>@QJE3PKo}yey^qs`Ip4&mUilw0f_drw6_sbQ)7gp9ez5`w}AEYc)oYOqZ&ZC
z4t`-^8ik0;>GShV>CHb5d16>X4o<l6^=$ZZd~Tb0*(+rr`(|kTvau1(<F0LuL1IOb
z)zDm5<kG)B%Yhaou8X2MFG)euSnSz(ee0jqXrS?oecuxu`9Y?~ZCCfZkt3@07vney
z6}J|X&a1Y2@BWgvRJ<(!Ny5=TH$k|`4ziC!M0KA68Iei32Pz0-s6JcE!==@k5k$Cn
zcb-h!oRkwhj(Qd2S%-zyoqt$Y?^2!V@ppIBq(H!=ynWs~DJwqqD~H<i;c(HI|A^Qy
zvB~NXQ$>^_lC0aqsN><y(mW85L`1$!{2P;`Pl#t9ue|3bdKrf|<nk!uh>(~p)4o^>
zS(w>ooVLpM_Dwkn1SC%+p;<Wjcuc^{+!mxQ4n=`t0{fAHYnMK;y1`&ljEZ?2z92j?
z4v~Y@b4OzV=f|imgRG=(>BP_8-wRz-@-9=sF<E1tLY!5(wAr|lQ0bfpshMCEtxNKL
zArAZyBd_i85!86TZV0QsfvyLpD-9dnLeNsMTt>P^so0WY+PG~!(W=RwJkNy8UtjVy
zZ!mXQHb@t?3nlD5cwW*_7nuSDM*-~&WP<^c#-bjo(GZ<6{CTM0AXZ)jLa5oplbF$;
zDsL3VU7z~*&v(qwLyD{BRl%)=k?>J*hXg&L$B^m-a%+tXSCJZ2+Ge(JjQr-lbLI!K
zLW+E*qbHuT6w&ds6XfKGoU&G?hXgD@Ia!E@hnCCNX5Ae^pIx$pAU*xTFr`|J7>_Pl
z+I-zgCdR?BvW=`b$DoQbZ4zfZ$$t2_MaFod2#2_7c_wHfee^7IK8&GI5((O=oCure
zy}FLh*(e9X(&+3ZGE3#N1bJr8+PM|OHU_P&Hfn)>tWQo`Nn?jW-sNc%aO%0&nH(#P
z2!XpF@kO2|`fy{>DUGokJ&ykBvt8(&nVP=rLX%%gkY;$-_hF?OGStj91VFIYvmIl_
z*tR^;=TnaDyoy!c84a$cQpb@j8x$JKh#NG!?*W5_iJklR-!d9|j;E=My!)6Mi+3#8
zvLO2b`H5%LvnY+@o3V9mI!|)5!vn>RD(9uVRo8b^i3@ychPLi@*E7N~G6;EQ8WQ81
zaI+w58ZNYLc=>crb0@(<7D~h_E%>?@FAU~|5=&=-ioa}=OH1}NQ|8~RxYMm=DQ82J
zhX<wxM_qQM>yZ<~X{|lBdP*@}v7dJyR3mycWpC=uTSp^+k^)0ytSK&-mZDnG6wcZ1
z=fc2jGh0fJaE&H5o*&4u@~a26Di1zl3U(3rrSinGY)DG3drl8eu1!{)3we2-Nb7C{
zlT2fijNn2ZYxfsZ3M6vKAm#>$l<LySOee9HfaOwEu*}M4a>vC^p|G@dQLwh<f460?
zW)Vv$yqwz)@br8dj3{Yn;&oXvzCVtpsvNH02N=0ff^h8>*(4>*qmqN#uACCC+hz}8
z4TI*YylT+2%fz8xA(t4nNmNw7Z?C^%dOlCKdVJJXC~(THk(^d^sWa&C!Sl_hTCDul
zWa_H68n4)b%A`dM68kBF9aIAKRqM8DAu&CfNfhXz0>>-9+s&v0Nj7sj$%AUx&R1?w
zQRr{=!L9Ya8_kFVvd^q8p5P`7Fc5>aGNP#r_IeL8(@-tbA<BW%SY>>)U4Y)}m<`!d
zdCTEB2|<n11@x&m<Nb9C;zIznv{Sgv+kG@lwezaSN~(XRTIxRlbbYd5b<vA{Qc}8a
zBJE~{U4AaVqmlkzU&T5FyB3!ztguLpnG3Gt6E+2Vb*MqWi2Mqx;Ez$ULi$~G3^Tdp
zNbj;(B8S$5N`^!#1Wcbg|5dp7lzNW>GjoeKM427qYhfX5bjjPHSWP>Mu;|JJ`UGQ9
z+d_>h+f6&1V(>1tmHpKf%*k00(L|?4pH`i_4_hk+@MtVkMJTe%`@+gZeHs&X#o=n`
zMG>Hu;J*G>wddi~E(_6@Kbd#J%>X!&Y6YsN>5&8RosOBp>@kudEn+YFnIx~%r@MBJ
zbE_(158`|DnFH&~qgWl@BeQx0ai@CS;Qjl7GuY=6IpTcA%pcz+1elZjoRUgo>Vj4P
zKjTD|Sh{8puKOYIYJA<z*c{xTC2}Wpdp<m5l?~?WU>?J|qE>()<HTiyI_)k&>So{-
zgE)+P*MNQpmSlOgW)D+j6~lBEnJWxynjsuHwcoE0bcFdeEx-DZh~@8x@zNlqKcm<t
zs^2zVVb*u#P?TYdX`enOS)aN#EdLhjF-z2)dBeCr@pc{XYSF3qVOYcmzw-dE8}(OO
z@N*;(mMyQd1TZJ;o>h|<WxPU;vET_yPUlF$adSoanz%W7E}L*o32q)DOcGu%<@*O9
z!5tTXmY@>=SmtKEH3k@v3MWWeRF&DJb1;|$CYmiP0Oz?^C~Eo-5t56oTw)^r6-SCB
z|D690BVh*pT<F{Hth?e=vp{>Qm!gEiy;b?n9ZCk5XVumepCr53`E9r8!}%<%yXMx!
zy&MayN2=nIx~bh1K_fy!50ib#<&VAnb+*8FK8l8Os&ke8?!FY|NUCEQ%PDmrmwr}X
z!h>1(;GuLB>I8jt9qWO7373k#R)D!sxyg(>dG(!!8xBW{G5X4DVCt*M0%N2l;9*47
zzu&WAGN*rVd4GV90g+SYWpe~9?~RdD&g1eJcJe{&K>c`=)1)^tG;yY$W+pZ>QL;4>
zeASMFnMnJ<OSpV=q(JUzf7taY(M;TcDFm2dVvJW~1O&v9C$a%vsMhLXUM{F~W(D56
z?D&X-<j99;<}fyA;^^P&YIQr@$rfu!-sk@IBwx9w;%yArRlY|H&inu{3-sjt*UF0H
zzic*hF#j*xLcgTYe=HFfs#dnFa!l9`uNFW6VTNF*RerYCvet;e!V1jN|2{uqb*b!=
z*OMxp5B~6o)vFA+`~04z4rvL%*0IUnvCiJKz|On{&BVOkhW^VZ<^d)SSA~K7uNE`7
z<7I$Own!;iXXY$O!9to&)YtQ!d4mcjo4hv}&8h~we*veey_$LbP?>UZ=?EztST^xD
zw)8EI+3wEqV`EDeOI^v=!O~0BwkW7hzZMq6ob6nC6UJU(J}6LJ@O_u2%j=lZKPvKI
zMDqFS*yiCSxPdKqj7I<U>x=aJpI)G|{+~xcc{n^>I9pae$7W$?3gW6UTM<~i9oWmU
z3dA)0F#+=`hN>VHbjEE=_)?`Nc->g#ZZ+pD8s*tnD>8zwCM2z#*j@SXI)MZ~t?{R6
zUy-?pZuAktgS?8C7C>ZEF)s%>@<B0x`EvT0-thk1&*KJq=rmkw$nB&g1cR`Es2!y}
zj5qjQ2fLK7u2U<NP()d~XMQ%Nk!I`!MR8N==6&lqoPzi#nzL>0HMq2p#QMmF_$VDB
z=U4%*!$qVW96DWbntuw>_Q0p(s!DoJa!7C8gC8CzTX!9@^1SMFg@=WD*kx#-2#r^7
zMNx3xJrwP7{N!c|q!;ttX?-aUBDq;^1gOKr{8#k>fFeP+<dNp)f%%MstqS8};vA%w
zR#CVvw&FL(W4z54xc|^Vr;uql9R7rGvzRI_R;rdxkE<b7){fVt)<qi9c$OHI3h&L?
z%Y-m4n&(9wEt(pcQS?Rb`3aSR_;VpjnjKJ~UMR#W+S5i0E4r>gvoH7!V&F4pBc?M}
z0`&v;f<&D+D>Q|9^DiRWVZFqcucoUtYOn)Y8q^(`W$)lc(3Fr+H!o7@)fu-23LlRm
z)tOxqa37E6be~Y71qOtn^*z6!)(G46mKg(IPg!(@^PdAX>L6Luj5y}o1pFyu9}(2A
zDK$i`s~ypyXDNsk#CA;v2B7{DJR>*}^euJo1VfKzv;hWkv>^{y7L-WAaSorV;YB=n
zC$*NJWA9gCpvCbs647GCm}bH_xQpeCJdlbWA?6b<ZT2Ls(6(P5a@xH3S>lym*O6J_
z;M9?poARi@X+qm&DUrcF<Qn2>W~2#nZJjnJxAk2Uq_~xYnA?CbtXNhhYT)KiFebMh
z8Q8#1cT~sK4sil@4UXPjEDca!aHfG5MZ=94TGs`Vsl7WyrC6ZY!2!uyvNEk>INtx5
zXiZGk^9?H?nWcTZt`|1=+4gG&xH@B4QnFp32$HZLHor+xJ(#w;O(!{?_rX)(QOL4#
zRi*B1PJtpV?1E3yWQi;`LpZYwv<|Q2D%rGh)Ou$#p~P4I>4o-iu{@|si71w5VMR}B
zqeYYRQKgXW`X>Mgt2RQHn@A|N+y901A>}F!hWSsSSgyjUMmb=S{T2J>c$fL9Gd3}z
zSV&`B%_G@Nb|<}xPzuZjyfc(eWoU+DDnQO*zZ%CT=Dgj)u(vf5tXM}1PUir%)r|Ak
z49*BzYoyxAuh=(?o?3-z$N*tk$Vk%T*$2|r<u8|#Vcq{hJFg7>Tpwof2Zgbs6!Mx2
zfQE-TY9Gy3Lf@&nWAGoi*#=1uV)!^^#d}35nfz?^`{Th8ikjyghj)7f-GhJGO#>{c
z|DMCL0Oud`*(*wA5Pes1`hJQfT6g85Q$MZaMpIS(Izo00#6Y;Vxz^W%t&bJFZ}*p#
z5cLLT!YeQu7h0dzPKD_xk7w_4m@FC`sO^%OhU?kn#uA(u4*4|d+ZlI1-eSmz-V}HD
zx|#*45z3*rYPtZ_Kh00Z&x46VP;}N<i$LvQq8oGp)_prA{ypw1`kjdhc?a!nxA~C%
zX6wohI<X>@0oB^~&ki=Kg{AQ0acH?tz)*(T?A5z?79z1O+iEUpkp*qfENtl-Oa<&K
z_4=}ql@9_1SIDazQ)P<C%O2hDQd513X|YJvXQvkzVYx0OWgE}Vj42b}uT@Ta?wyP_
zdLfnaB1Hv=2~Z~8wv34u^y*5|nRpNKEd=nldBa|Jijq86^^t!Je-apSmKU$VetHdg
zX^$=$@C;w0BHx@kb8jtm78#?dqTkR@f*9}$bo3`j*l8o*T0cXu9&pN3F&>;h@BbO{
z;x;%BdqrMA8uHS}2fB#;h>dvN)e@`XOfp5AM!un!1kFJf@cK6PqtiyJv=a6FLP8uQ
zM!w6R>?2;|V)q@IND(xm2RAW8$qR_6n&&9BtBhLxgpP+qz4@?!mAW_)$9=WnwAbQP
z0(JUs!@X_BjlhG!ZXnpKpUHZw@3(W%8EFU)4adoDNAgU-g1c&ti@)4Hy@d%jn#j~<
zzGqDOhNl1cmZ89M-$Jj`y0Yp#S$}w1FFI`@AX?9QtVibbB}|$37wAHq^9q{^1hb3q
zP9t=R*T5j(q21wQ`d*iu#Xiv~+>sJK<xH4ob<G*(PX-%1Zvnd@A#TlJ<nxd0plnX4
z;HRrBT6dVd1yAv@uN;TBUbpQ<9MFTw*a~K9NOAPGp^glTET|d_Pj=GIfkqm@>5IZ4
zXM$y$D+Y8))G!YCS>4jt>KJ>5%Fd?sH%M<gPx*g;)c?=@a7Gq}|0NWxDQQdGg7EvO
zcd|i5l&7wLl|TimJX`|oR^QGzP>k0G9>omXW4HWy_ga#EhUVb8#3);-^TXCySP)b}
z;aiSSrs45If4qG@|K3{KDmWr2Jta@w+xqak4H(v)cRSsJ4^3=g?)+f)dXTU@u-`f`
zsHz%5vBSGNbNL+tiNwyJUD*IFRD7vD$PKHpO)-{DPUj?Y2t`6OnB{@l0MkC!emrlb
zsB%X9&VU@K@UB|eZmPeXS?C7Uj=V5Cp~9JAeBazi_~<!~1TiD!!`xoKZ3jLoYBb7&
zR3A9{Q6Xy0Z?Y=%m~=SMz;3GQ2_|d1mecG0n%bRt#RzkF?SW~jHBQZ7*tJ0qvS$~c
zVE(mRg~(?#CJp!{LS}~dkYvIM<~~n8-~=BI&#s{3^cAcGsF=M8R0B9;7bT!z4oJ(F
zkmR~Nr92=o2Yx3qF}d!emf$EQihr<ddl{OPoR*`9zPQQ`gjuAN<_ZNQozg+9<^;&z
z$166-h<<-{%>wE4Lq!!fNtb7!JpjBdx9UZM28U*t)7Q62y5AA)wB0zulHxbC_&v34
zvMGsiL51npZApga{r9_KPF_PTvFl*0O*_<s9M{MbwIcf6ueFF;prO8`9|XS)*$8GS
zG1osNUcDf7i}|RdpM+K>EMHzp_KoL>Kuib3)MD&=6=c)GEqyWfxeO|eq+rw$s?lD1
zjk<AtSZ(H<Dqu)%1HIpQISB8aP$0^HJwA4V7+0-Gtht^G3M5<yo~-uii8O2XAQJL+
z0e9SH%IJCJ2fIXRd)q?+obQ0$wA1awwtHzn_wFFY4vKYXkMCeM2FSA)+3q7|&$9J<
zt}N$Y3B2}tyKzx}{bFewzj}sq_`4^(n7kMo0|f5wMta2A0XmIWlQo;vLgU&It(2>_
zyFMUk_8;Rlc6Fr0Jk*w+t4QiVO?$pWiV2x4eda2|gg<xSrk6t)9V<g!5kNa>47sQ!
z+@M*8>64VFMCsTnf_GIbhLD9v)18zvhcZ1Bu7gL@v+OEH9FPjc8B>mSE*-d#z~Xy_
z(I)cQLC8f&@Y?aaddKiId9+|GK9d&-rt!0Pp%h0bAwbk0{42r=_6*1MMbOvT=I)eI
z>|<<WFk5$(1!7$tMXDonoQjtD&c*Vg7xzl2S08m;1PF{FRGLs*V||!q>2<hmNQv~V
z7V|v~4c~Qm7@;U68}%0>))CMg7S*Ir1I<l(NS4^|@*9Kz!rd_#x9jUUNKI9(Hwc#j
z6jbt>pRH_93M&51vXI3}hFc+2#HFrQ^w~iB5Pk9T>r4JsLYIM?tIHCzr@as{;SZTX
z8#ad*CCJbjQGHtYco8K|&2cqUsZF6|RWzq4ta8|ygmkNTF6<YJ*Ci5a$!oi{aavK&
zu3N|<K=(<N*~>vD5?V(7z`YPiZU_!%M&SeSfHKb5A%dcGLw;wRqsk@h6O&ML;4?R5
zzn)xWzVrZJ0m@?mCGisq-<tz)-xN&Ud&3k6?nX$JM=_OSEYb+BOXh)8O5Cr=zB6FW
zd{RR3{@#VO>-#``SO=Zy&g%v;=p%PN3VN|l?E5Q`a>oQfvJ)aMG7!Qa@8~S}>9tJk
z!J>p$4{y-0K#56f@8Dfb7~T(%^s;aIZ~L}Yc^>z1;f$kkA&PZ&4$lwwpiklHzYgO>
zR4j{&f{-D1HZ!f<9HsTKf`j#Ep%VkRJ@R0iYfbRsxN~JB?8|X(IoRq6VIczslJchF
zB@XdNB~GRay>Y(@nk39Y(G<qX&Yh}#%ku6nX#c^>0j;F%ftQR5uqbr3(24HAZy8~Z
zGV{>dj_3Ac;}BU2Vrn~b=}W2zu{6SkB0gbYB}a*)Ur!8{LeKf1d3|CW6+G4Z;D;&d
z-nrWx_G|%CJbZ7zq?3<8jE8sOcVpyY6&^<WE~|jBiqE(e8`dd+PFCcC#v+~9+CD=Q
zO@B-iJ3JgpHP>L#(=ube)ijH<h&Y3kY^4d;;tM?Dh8SaE#PFIMR)Y=fJyX<5O;I=+
z6J^xRYQ@sLxZXif)3xmcZvzRC!R7$R+%1m~O|x+By>yaxw8N?s`rAN~+`<qcBROd(
z_DcUzAtu!-6vlP_EfmTY02_LvagMX6ls1vnuB2%HIJq#8HkU@bn8P~#o#;rTVd2v)
z3XkceJlU_mO&HR~^Prr93*S%Fk?b3xmf}<i+wWK<SNwsn+oNf!Dt?XEPC*2T)`b~T
z+|X&PR+n$M3L&9i+i&w4=I|ZmDD1AWObX%z&^b3UXfSDHvhG)(e7ma5BWN{%q4`Jz
zp@l#E+)SsiZZ_2j@ou0J7E6b-KiCtOZTEnTLQj&>7i6~Zw=`QUXfZd?y@u~ph0w1^
z;-TgW{5yWS8m$A!k&zDrAnsztHRf2`&YdptAgaNUfjwFb&b`NOJ{K-c3t~8inlhA%
z@}MQYY;C}ngXd{7X9cQx<d0P;m~G7}Ts^-o-fT*4Gxiu3qjY#3XOHa91+)4=>H$0x
zV%h#JSS`drHxql!qRiyuy5N3gTvLcp0#f$P!67+n4nd%=uw0~@y<MGL{jcd>KL-wl
z8*<S{_uPC-)W><$A~k=@i7tY3V;u8XL4$p^WyA&Uqlm#AW>w>GCFWiOEU1pBiD;z%
zTAwvtyz~U6Bhoc|^+$k?C`|USOL>W02T<Q})qA>wbyx*v>|FbI>zCNWa3uPX<7wX8
z*UqB>+DU}+q^jn&^Y9BBPg$>gxo-JLkQr{a++y4#15SU8;Gc2>?-A~dEL+R6EKH8w
zmt#*$VU&My)1+#;kB0<hg*ufsU-*Hg*ivGN58TWyb|nE8(ea5w*A?Axzd);>Ls5`v
z!T#*-t75jmM1^U|E|aEmkUy}!UgrfH#va>KQHQm?3wM0|4<@P<n7Rni>()3IsU6oe
zwV}QL#H!n#<P({M=e^j4%DUD3qEnR{-_9nZF5W56-IGm`=2@&H^W>AngS>3^z1{e@
zC+U4~9FK#UDXSv%lC7H6rj$5n_W;h5!tI&YtFQ+Jv1i$o>2*<~md>W6Irz7T2L*74
zi~YpAkf8m5j)b6lQObPec5ceNoYq-U<0qdk-H}wA1D!EI%FL>y`HNM??-}O+op2?+
zF8dxo9mu*)JjCY4)%1O+^A(t)q<3eg_am0_;9ZHRa*%sf#U7iKxf)%z0=8;hr0M$)
zk44mR6@<-P4{anZ3LUlycbBX7ma7u~Cnn^*%%Uj|IiL>tuQh2K=Y}G6242tO>b=K5
zY&+NM6A|zki53uGTdhAY%|9s?GjvPQH3S*D((X6HB<lf9nd<pGnRh*ubA&pY187QS
z8?u==KTw#ClWG60N@V&k;x$HQ*8iP0^y;@tJZwey4YhT$VTE7#trFAGL6@`d06ixL
zTc)z((*XP@N!5*XJ}w^Q36}^-<SkrZE~#S!wrs?ojFjuta!6cnTpV7$hQ1btrV5U;
zkhWtLJ4#x(`eDL$i0$XnvxpE?7Ou~}u17Q}l7?n@eAFbW?hbBSMoQW`mJjcM*B%93
zSEA2cekwgiXUrB9%VNtG$Veb?WjSw*4#CU=JqvGqJl^*NNSADJd72)Zqd^)cJOn$9
zPW5x?0t!@KHD7j?fOsjk0*HaE*&k;bu4?2Lp<kk8SMY8)7xoWc!jjI{oTwQ)LO$JW
zSHIJyQ%Oz&s_55MMZ8|j3V^hkN4DbhKlO1v_PV0GLWFAP#P1G1&NnJbB7U@>s}(fR
zv0PNR^6WosZ|L@6NQo+;xQc}Ex4-U9UY_@7SXLS#pG?u@w)({^qbN{C)b*jQ3}Bu;
zz)SnfMBSNbLxIb@QXdB3*`V&Svk1CoNuli%_Jh1#c=QqSiNg_+kQp5c3#C`;4ZIr*
zM^6}o5oby}yS6M||FWRTfq(spNIGK7?aZBTooG?z336;e=8rhd_dl$uWtLZGkqt^R
z)l~3)U07mK_k6oM*Z5U!M|6BHCRU>e?vy#TTd0Q-5^*QD<SMAc&&1avhdYbkb0*+x
z2-~e^#1BG5kA;n+XP${WZD0?Cx+_Y!&ovJTFqGrqZ(Oe|hrv5m+W8XZxpPc|S1l4>
zBdF3YOLkjgqRN%7VxyCA*6wm*^ELCJ``6Y(R%R3<CsRkdK1`>k^aGcK02M?-@aL1&
zmA9U`s)Q0g&pdK92>hK}ykS;x-KH4^PewT_sjy@zJO=Z!|Gsa&qO?WGS@IukE&wDb
zs{nC#U0Mg{P*chiSC--0nj$jjlAnOdU$`#2U!H#gV7$tR5xnAd+cAoe9eU~(e`Q{-
zvdR?jFa9w;yiRfLI3K@gOaon8mx_8S6<+)6lg64d^5v3OYL@W~*xvK(@PIOk6NpaL
z2-_4J()H55H8S{RUd@(a5tT}ye8*MI)X~w*cs+~*hSgIx$aF8?hNR~DTwa2aBj?jm
zfz}nfrQvYtRjwfW+YoDaRAT%&ccQC~M4MDx`B2GBqJG587I87V3}0G6+ObZlA@%@m
ztUwXXSAUZjZdQ~&{E+WS=;ZQx*`ma&L?dFVL&thDSfsT+`BbQe>+X?g&&4b>Qz%|3
zYMl?I3;IgA8vOOkFN({3T=x2luP2*pq*qs!wGtYS!>*lfeXEwZ+8bkQa=<Z{XUr{_
zovG4hCmi!E)l*oosuo92sP*#$UAXl}(r~1I3@|O$-(FhCNMf8byYOi8c?$&}LuxT<
zi3R5URf;}rY`h~wy-<qn*w2KcAhoUf@*@Y`lJdKdu?kt?IP$k+z9U?F^Lobg=3J7M
zES0E{Y33U|9**PFZ&b#PQFq1y7FTxE(s)L8ZW3OqkCDx>4>M$SCFm_PDe>;;dX&U^
z-s`KKYQ=_bauigc#)`+xC{?s#e_TetiMSYxj5wjHYN^)ZR>j~`Z)=EDd(Ie@tB2c>
z!dGsc+dDxPY?s-SnQI!Iq;^zt;dO5+6q&jIBfV^RZgZ-oslT)`(&rp07{bKv1-#d@
zc6Qvogr6-mK1Ec3Nd$u<kV{zvD>~d|p5iN2B6#c$2nL9W^to3aIri3ocDaa~B7Yo9
zTc<AQd`aYTJ$*6U%8>)Y0`)x5x@Kf+6teNf&_LT_t~WK<U!DM4>8#Uo6MPsIgUcAY
ztU;b>EO|7|61qGcw%fCeui+VN{qfyxx13p`;A|Ii(_=1LMT|HftfM0Q_(m4y^<Hts
zSO$IfE!DQ{2Ta$w!cvkszex=>!=4qVH>h%zm4bA$NY=Qn;Cgte+kob0YWia(Fx(8A
z`9VZrE82~jPqo8g6mBb4x(ms(GW=>rn)E<2_Kv<I!M|D$RX$@L`D4sI{L@6Ug+(wX
zRoDVn4rvWdL0hIW{5nr4TuQ8tad5!?7aYzS88wG3>duga(zw$MuAPMt9!u<OL@M8O
zX}(mZNyC=+=R=pT=k2Y>S|8T9UzzuH{+S>9JApO`Cc^MU++QwRGM<5+!4?Iu55B2N
zBQ=3&L>Y&FTuU3vnD%E5*ZE?W<pbod#@aqI^hjbm99cv>PFbr^>unR&1i|0#VX^i2
zaBB~GCaKefm#G<ap1EQ_B`J*uQ8EKa`PUD6lcJP~VpJ;e%1W;n8*ZXd*KwYH67*uc
zlWEC)PX&xi{#XihA}@@mQK1xvw$66}ztVi%t{<Xx*-HE?-Aq3a85BgSXF7Exz(q!W
z2Kmt>#Kq~ino1*x(HNgg6&<>D1@su)W3J=cldITw6PPX7?*#NvlFj+^Sjt}g54Y*_
z!4DVfNrL7nQg<5MW6{zD^Y@4c;4F}568?=ZSZ;^`$)Y$?vz{C9p**7oKlgY@3>deB
z50#tf(SR-Jb!nto0nX6?87Nm`I8wc${~pQ+(mx!O5tQ13=2I*%u4W9jL4l)9TkOkK
z6D?3Kiv!lkyTd)@$ff~WK6&Yvp`2sUj!5aH5QwF)UZfzw<n0{zZ@6qLFxmZo2~BV3
z4bL<*ee$zotF=+P;{5lTePqaF1mV&DWcTNzP5Vyx2bU`(KkRJnd|aC9Py7|dGyn*6
z!K4WkN}X*W9C^vw<JX%d0u&3VAdt!mxq-Ch)k5VSIMA96VmGZwyUq2emc#XmU2r3+
z*hBVvA8M3WO$DfUj<6f&cQ}AG^pWKfE926CvpVy9+Wif1)ELABR6ypKbYkRm_&35X
zQPhUO`pVV9P?Us+Qbl?26f6HxB82ec54t;oXV9a}-(@nQ-VuKPv{7ZY{Yi2&s*dT8
zY&`h6p&QQ=0ro)F#@ma#?rHM<6u<tC_T(GPqI}&dxgL9QLlmgBzWQ$afU+=DeL@u?
z%Z>)A%L!si-UjxY?(B4Un5Pc=w`G61WIgYHwtcA)KBsOjWshtG+2WJgj>18{`kM$o
z!WP!*bm8@Clx^73yjS`f59aaUZ-5dNUgNH|$?5i#iUFaYtYX_o{><Kh647FS)E>y*
zhNW1!5!egq%}6~oir8<+$zIk-0cgblu`!zxSLeD2a#ezoI?@|JG87Jd-!DN)t}ygD
zi&`XZm&#E2YqPTaN8Pwjvx-R(16qoY4(TDu<Mi@nDM-}s{8W49WD7G@#zr^1T74cv
zu|BJDITYYz7*5fc{JuoA#}vdQR=HbLYB<}#ZlzRgpe7A}Pd2{x^Jl8l-u}N<d<=~L
zlMHA6#l`=(VFCYNUyz7>zkYOoA`Ya|XzZWtm<ZcQ32_H~EDJo)^Y@q6rHLz?<vR;g
z*BIugKJiqja;2;Hl4ASY?CEhw#>h$r8(nfnZhnDY`pOH;9v&v2_<`U}`rpFNqT?sS
zhk>sDpy8otmew>n0KTH1E^w~Ngfg*obRMlxQZEPU4~sc<)dh@YCA2+doACCGgSkw6
zn<B4xcLOmsJ3FU@?-^)lWJjzj966YHYW(f|Hg<NqgRvkf=3wrxAT_J@%44R(Z!50*
zTMrjgJ+{Q2BRU-SQ#*Xyhra*m7+Qe${D&I<vD)7G6t1Vqw=Cjc__Aw%I}#+~byww1
zU}j$2=2Eo`t;p+iEnH325&TBg7|LH#EGf$yH#va0(mfdZB5QSP9db9e=4?AMZ_HtG
z+lzwgXNn^|iq?sXgYGMFgai!R!?%Qdz*c)>r2L84GR|y&xYRt^a?cX1JlcKWQgk!n
z1#g{hyZEK8n#I}2O#d$|9{qn|@uL3&i;GfuK8@8)5H?Xzv14hEv1H0<to7bMhFvx9
z*fuH$CL&>nk>0S8)3c@}eEibl3uaq;-d0^+&HsN|{3a)YnjD-YJ@Ue+jQ9Vg#eEHe
zw9q%OCR1_jQj4w2Oqvvfs}92x=P)^nJ$E=aBKb7hDrBP#D`tSg$LqMFlBXSo{ts*=
z`3qZBkE^A0-jiOZ*7+BBa`OrDRE+XFq)*K1B{}P4cM}RqLZCvMBIeP-&&O(MW$3o4
zrTA~6o*1aPc)LzPY=CSD`AsBg=IEYCiBZmmp^V9%I5tv(#FwxL6`gFo2LNq?H=5Qe
zL)YB3$%Z!~K%aPRhiFwr+qqwNSwECacX1?WzBrbx6q(X>cvq|I!h>rnVeRL#T2UH2
zSs+U#kc`U>*hF1dS6&yVPmP(fc}$V2&`Q*#$@#}}*;~V{ZAY7pW4kl@7SGg}U#cNY
zK;v()0EH(YHoQ>AIf+b<b+f@_?SA4Obt?DG@q#xf7BnbcY&ZRSD4qdwh=^Kx%o#Y7
zO~vYQYvF+Kb;8%GFJX%fBU$=brZG_B*Or`=(VP3N!v>uPXAjbnN$$i?JKgnx(H(g4
zNrlE>v#GS<d^@KK@%PN*{~(r&f7@w-oj;a>Em;qal$MXqLPN_(bkJwN&vfK5P0BQ*
zb5WtsV^zQS;gx8pZ%s(em<2TlH?Meb$lmg-8sg@+OGhHsU4hFO?zeuYlbwBqOrtHp
zk@kYR;{_qw>U#OK!q;=|`C;$4Ms9HMMJGLMa^z?$$catA4d;5@ji=K2v=DtsEzl)J
zbSb(_c8?dUVl-Y!#uSk!A!Hw@O|H+ZAD~I8Zgnb6$m|@akRPc=iqldKugdE(f}W6?
z83R%j6eWe|n~jGw$d0_<_WlDIWOb`FlJlYOqW~*#ge$HZY4DjO`}&IG>K9}2aAIdk
zA{R4%?ZrZFGIXX1PkK&@40$P<4x*Pvq^Af;KeP=^76{^Z4iASI&P_jG;oI7ddkNXG
zxvAu9wC&$pxWD2R*j?B-bCI9=Hw!OR+!I1Zfk<#1*m>iDIN56UhjlqGA)inP)HDON
z*CDs&-Lo@`u-*uMxtjFdeGa%eiYjkvC1M#&LTOk{|4TSQ-RVP226MeNn^H!hEq4{8
zl<7HVASZ=p7w8rLfSNwvbp(~2!yhpe{w(^l@f6x94bnOhH_l6NLIv<0?7CF>yOEk3
z-wic=?z5{zgE5`e%AvPkB{_(<>FQu0{(+<AWcRqiRzXT6u}BNEN;x>lV6#>tuCk26
zA$y3G7Uoy_qzs46k#N7sd-1&|y7dY4^T~UVjGQ5RH5Vf-#3e*w`@Qsu@>P4~<Ty|G
zD5to_jsh9`ZY`{#7UWPiRH~;2JlaSQCtQDvvx2}R4Q#JD7F`-l*nA2=3`~*^n|KB-
zL>KVYZv{Gin$#DHY~RM(U|KXLYte-Ljg2h==$ov9KpXB<tx{SgWiC=N7LnO3k-9f<
zg=pr)Nr@FDHOXDx@yu23g6w|DV6)R8*-m?Khn*`D11Gsmpz=1~fBMZ2(DwcCDZ>Ys
z<M@-3cg{^Gc$m35CNTyHs`GI?SzI_9eu7K%2zSUtT8I4+w))7_pKmmP6PMUz1Iead
zKNK#iX!zEM1FYy{dOhN9T)6Jv4Ecl7xU_{ot0Mcs>;=OBB|%@93x{3N2HKS`h$<&t
z58FlP6?9gY-~Pu#eKL{!fvkm2uqIN)3EqO%@b?uxOf8Pukj@gxq#7L239JRdgQLR#
z31tfM{=P6CA^Foo-TN=vbHABAMlf*WBC*Yzxgt+9!e5GhXo9lQjB|S*TDSci!%Vg<
zMkk;2R3Ryv6ubUEDU(*WG6f{Xk&s=~pqg`t@6J;8y!s{{P0Gj^O0cWkxKyf!czW5;
zl1w&{CI}zEe|gsfy*;u4vNwUfD+M3>eE~Kbw}8FT`^{ZV2$?)zZ-~E|bo_{Vd1U0~
zIA{A`s~C=zjW+(N=q06P#Q}no2GO!)06NABRix5!7Ha@#{QHjEDpcv&-qx4k;ild`
zBX9W)S1XZ+hi5RkNjCxkGbYC~9!_f3;DNu!YLfjUW2%=SJb1`c?$DWdCUr2F7!Txg
z=>AH&!3CXVR_qM8!U>DAH$x>WO5b4Qi6`$BR1wKBo4*vbP}`2=VH;%Vt{cp^L5QU9
z#&G=zKnCCJSNRhqOlnDJ(m&HRN^INa5hqUnnMKX0ezhUG^5LyW!+za5O3YqE%^0@8
zMa>8iO%)+^|2@+(+9+RrujkWGn`0j*{%(~cOxouhB21dV*!<2z&6wtq;2aW(g00=@
z7AJ0}p>Bj&Qlo2x2(5~mR!A_TWt6^cg0al&qD9I$ZFJ@0<Tf;0H1Va)pyBB1>;`75
z`njB`y>T0%u7M;$eN(v<;;N(0gvz3QgK5Gd`tp3wUcpfXqT%fkB36M>9h6Rgn1W;}
z5S64?A$x;k%}I&$GMMFm0816`e=r_hxOloxGUjnhyGa3{jL8;LF6-LrAu(9{<kMpX
z53O3yuhX;OU*vtA9BzP>WXwBWw$z2yJuqh+WjD-<QFM-j5UST|f@S};4aJM56b{=Y
z$+9h@8m^!^ct+uwf<5(=$uX{-L6yukrapA`(~)&@<AR8_n%KQs)ni}Tp_lfc))qC9
zi8A(Mk(S`3Ohc!3EQBM_v$j;>^e86fiq+wxRb-TgPj_-<>tS{&K=ZQvBweTQ_UZ9}
zH@AKBU$+9882-yln2m}4e^>iys>k4Pp!nSA+Ph0at>4vocm8!M>dta1boVdCiv^bj
z3KJoW3<bS=dAqEy7_Yvnh9Dv+YSHAam@!#FYy0R-=y-7rN$FUbS-;G@m6EyA;k^Qf
zBkoQZ3`zdYfmM_4NTKWK>5NlQ_rkd~E?mUah3)bDJp2kEjX^PbK)b@r%wrqScv4z6
zyDn8CUwk7ohWL-git#>xxu0qOOIO#YH9_5(!au5K81G8=M(6t5c?B!KQ4h-{ISXIB
z1FvUu2rO7l4b4R&?61ko(@lr0tFJ{Hs`}K-A*-xiMq_fDLq@&-+V4Fhz@@s03S$^V
zhaT>$Fd!Bz2@TD3=eokfWS8WAof*5V*dl_<ShX%1b5F!OYzV~ofUNG)jR30gxYO;k
z&64Bgw$mX@L08FAmH)`<!;&5RQWzT-vL9{v1ku{<1M4u2BQbC$)1t7bXo}e(F(0p(
z1VL!F;Se|T0Y<fyaVfCW7VUkqf7m|Z6$pda7lQR6Hi0j7i}o%?6q{5O7L#4S<Y?cO
z`axn0PZUT(5?Vk8$AJF=7aqUk4QqI6fg1uZlY_Hgj$CDrCXlxWk<zLeR!+>{_&Lj`
zEuG|}oTKMiA7&bttxzJyo~TA{&zNjbKaxPw{fp=feU;M}nrZAsAR{q~Vm3;BIv}2S
zyM0B*hwC*%ma*Nl11iU=CXMKNx8I1>W;75}kDjL=5=o6KsZ1$?5H!`4`MeE$0jYnS
z&s+dTp)dx9{cYE6icI>8Ra&NmU}Pd*i8#nI6e0ky^xH^ultu_-#h(M#&m+p)K^=Oe
zjuKoqS6=Yy@}0e3n)#tfG%=L}5K}$V_eUkC$&D1R<{a;-f`XDEH@48{^QK^Si4LCg
z3v-fwKnT)TMnt;T4M{&A$@#JW>8<(xI?AaCuH)~6yha{aMbG>6*M6`cG#do{(73v}
zQ7veu6Em8VsZZ5aIf&K^xa%f;BS)!$Vq9#xNtd`-x+MzOG9F~d&_5_Z_NkD;4<0~n
z@@K72Nd*C`E?aI$rgoY~;<yl(T!u=H%A83v$}?q$@!bd~eleS>o}>=hWOA!fSbF?m
zaZDPF9BypIdLRh%k2KIb-+DDyb{7U0=aA<`v4aAVD#cnUpBI5ZSEK48*uEuJ@$~C@
zje-rr@W>N9B5Q+^jii`4>9|cV!_-GK_+4BcaaU|g3H9V*Xi|v|9u{{<^)FZx9xBu+
zg&yTkBFBTO(f^#t6D3k5#aJ6;R*`-*hTw5;XVE&GgiXr`NMbI-Nn~eQ6G4L$v$4;m
z87YCXR$zZmP2Gw@9f6yR@Qa+b&a&oEERnqcd1j}=U}4j7(J2*l`1xxayjQE42iGMl
zQC-~AK8g+#X@R_G`@poaFuZjMSE4n{wlJ6BdN6pQs=WS=@+0h12JPZRl!5pnII%V5
zcvAU9un;om^Ad8y8#AwaepM<ph_F?8oz0*_&Ta85%v-Hkhs=4J4Cz^i8W3!Fls(v6
zYkTo)>$1zXEo%w3qjdKR)ESczQGal0y}y0i)w|C=zm(^5ZCvjToy9z$@sXh=q@QWn
zY7IO6t!c1M7A&8}LcF*bd>5%(kRs)Iala#WNbLbD$~bmWA<kO(?cR)>tb$R@fvt<U
zVZJYJ1fsb%5uekPCgfriA+QRSO|FMaP8jW-_U6c%VqC$L!fR;QAmU693&z{rH+C|n
zRX`i+c>2~NY{;XlI7my93A5O9ycYPg=dE=I)W;-he4PJ|A6zb1X1`tsZjhJRquSB@
zr;7_oMiliRPPGF^q1(CfbM+DL6t9yyNzHm7)^*Rb6sddu)wgp~Pm8u$#nEWbbtllN
zD2=!ni?d2EF1k&mf2<<XqGbpd!A9}rjd-zjOLL~}%@|!355}IxDb<!>cIFaoHl}FF
zDm--zxg?BX30U8!&7f}?Ty&ZhpTj=net;<=c@}nEIwyWUgl1$)+r!cmgiYR*cb7Mo
zao+A;kF7W^m7MK(RxwbdjRlPsh^H3%ml@56;*ph^5~j!uiOj#0B+ZFP7rDa(nNda+
zy|NlTm_^i^oe$PhM$=EN#)S6Cjd}W9|A6j{=nfZUm<=)IoW&wD4yh*-v(mg*7jkvM
zvo6{>erYh6-V%t^8+d7O<o=9%4RhpS?3uM@J9vniSjN5TlxYw%a#%z5aNWIaI4`cX
zZ0uwOG}G(wHb3BZ6`T0YU2L%-g-F}6-3pqg<-M>R3CQ-I^c?u1Zi5%qB`SaK^}Hx<
zp_m)KxWlsExkVq#r2G}5TSyFHt>R;Td)Ukut2w6SEOZ_maxWdH?$~WC(kKAGOe0!S
zQQ<5yN?7`Yfe8Ii%MGt+kudH#RP&@Z7)`KdYtl0_!J`DZ><ShRRxnnBA(hV_qijWW
z>N(LzgU^Z|+5X<1Hpgv@E#yMJ;0kwCW%}u$V056$c{px4hK+kEf-gho;iEW}&J&@7
zX4i=*h;bD-dFs?TlSq)mx%JU>4CFY--l+89w3SsgxLww)3!_rz0)oTRJZWz*)s}{{
zbK6D_Z91F8fL54Jw#GX!hq;~l)RedHSaWjrl*ynwJW$xO4*I}~lXe!Rc$;EF(PmI=
z+Th5AsUR0>ABUCKLSH6J5`T?P%1$B%ldM0*ih}fg#s>=YCVT;}0Jjmf(snTx33D;6
z%1O4DTaTOc&W^gI6Cy|Z<_s=Hbn}3-n`Rwjw46KgbbHaJ0To3IuuEtSqURH<P3K5>
z?uf|Ai0MX;i-V^D63*SCsdUMbdkt_RzgwU!u(VR;ogGHHTO-s+?cn1$OWU4;nt>~^
zZDJ=UHAB10et3d@*w&HVz|nE75DGjE_~yoYsTQ#Aae)KR-1{k_nt_m|+f$c*ScJI)
zG4Y%XerOl3699glXIj#i9Ray~{==X@!~F{?c^q@PvUFXP_@IR}%u-yckTpHrgU~K0
z>0E4VDkN1oj1(aeR8o~V>))`;K;^o^Uus@VE?;;drMs8|Hu02UZ#ceqcI(SNS7^rH
zQXVT-2nl^sP`Uv|Jn^)6QH&EczLc9)Eyn=0pp_}w>>-#&4^e&`XQ{&zo0!31vGYlk
zA5qzh$IK$0fp~645D~$hp`RhbXzIJwg3rFPq*j+QPPgUhwZT|TKeVV@#R|w&)Lbmw
zL5ed$#Xq?(^bUO$d|iM?gge-QWeN0rZm7(W(hTT16x*R4p%K7y@N7#S2h<%@?s6uO
zhY&KUfy^c({J;QmH0??a-7!u}LZQOTPl3Wv<{uAZ*f5;t{8}z=xrzrrkHIs0F^mi6
z^xl03Sy5SYbE53@Umng##=gmr1H|(=aEydjh9*^+96ygqQ0CH+=#?b7n5yoSJX{Jh
zsjYXe0$fd+>JSercDh{LG4;I1ANiOc%hC}O-Dn{mRwdDVT+Q+YxfqzI->XJl`@E=T
zi7<PH^_Dj6Ufn`IrQu@=t7g3{2Mwu?3<iJIFslA;#Ehw(HDPX6>^cZ{t9ILexmhb}
z{{6nS!N=8nYZ2teEmIZYo_;+F4jplssSa?nYSjU_6I(Sr*sS?;cY5IK*>N24l)eWN
zI+8vII$NpZ^(MKNzBA0VHXO2pcXt}Az5{vfHrW9I4f+F8^ZIY<zcxJ?S^j6!lb+>&
zX>`MEi#}j&%&C!Ur;gpbkMrXG=jV5KDfQ1uT?(xbvR>0V?(2szro>H**T{XLFxFUc
zzk{yZl-E1!_3>ih?rZNLLn%QYwJ-GSYXbQ3-vXJP#v(h$pJEcJo1J35Y<9HtY9j4$
zM~XAX$E|rBxEMt0E&xt#J2z&?>|@s<Ngq*zO%ko`tm{~MHID_g0Uqvt(B%Q{&$sUl
zAlX94<HnzJ+pn6Hmk?iENFo8oMH>MK=&+HVZk|`L|A(=22+u5NyLHk*$4SSwZQHih
zF*|x=cWm3XZQJSCw%*t{`Oi7C{~LUh+SgUJ2Ya+@)q3i=7fh_>KDlBt6<G<;SK&9l
z53mp2B=d0)@nHU`My2JJYi!OTo@^PrF0N=6?w>q1T>?1gn6mdCZWzmh_%f^>DFrBe
z3c)rlt+41+p%b5|dt^u2t<zzQplyTj0lK6%Q-v&|)OB$5ZF8J#{?)elT+X*i>r@gQ
zhmz<JHE;JiUIg{M19}VFLk@a{{4$8+Cqlg>IM_Q?BG(iB*)B>l9j}qf8uxtdG><G1
z_edsp(^BCcOPD6abiIS_PKZI&fo(GRyLmS`Gfk-L??fpXGhi9?0kSc?Q+^z@Mt`^h
zj>MHfOWsXM{!XHIy>j6wQ!k2;+S1I2%awk|cVXivnO%Gz_~akR6SdDZ{=1_Kx#f(8
zFS>Wn*zDCtj<ke12l$gVUSlIHd3HKpd|s`s>m3!0(`tj9MjJm`IjiDGI)}XHZS}o;
z0XL?2Wqcv~HVREgf?8YlJRK9NSfbWV(=y<|H?U{r7(y#sA>5pWI7Qa6&RqNt(N?Ux
zAP*#XSwL&NyR_K_=NWSh;xGP!K&1SN>jGsc>Aav3*)0(Xiqv8&vg6H%ch-wW)e)_x
z5Zy|%19ZGzwmQz~LM1Sd6sT!pBqg|PQUbZdAZYFG;L=l}-U`IaJu1^DaNbU*$khac
z1udu8*gRP(dkN?!kQu2`5yS@7&h!D`plY&ka%^cNb+Ybqiq^@50j`C9HA)~ki&M3)
zr$`nd9An?*U)MHyB*?V60F#u&^nTEMV{t^Zp=766Ve~n_K7dNbSXGno!tcfd76(Hr
z{+blpA+G9jZTVMI5cefE9@A%|*;ofDZWZc|k`rwX=s~@Gs_doMtpx)df^AgU(M+bo
zdwOTADr_^K&%=OiX`O{zbZR@cMjfA5`#nyc^^(K6Ic5mTiqmSm!(@vR83q0na4!6l
zJFZlYILrpo`q%*b6cRNnXv*+IWnj4&-qQ<DsYv1>;J2_n#_KQmUo&=VE$ERHBe)<*
z<`<^?0Y|f*HZy`$L%Bdqs6O0M7WjZG{B$syAenHQ1Chj^W_9euQu4Od3SyL)6gIB2
zyEeP2LU1K}QNq@L*D5oqk6Bj>62K5SJ##!dDkE8fE&t*xKdN}iBYE$*M?+%lC3f@Q
z#d2>FrYLlbleLg|w9$&MMjVhN8b@Dr2Vho_jGdZ}t-x+1x%_T+Q<Ux1<)<i7XMe*O
ze!j+Jk<q6!rNb8t37!>tR;Hy0L_!ixY6oC@(%T2%0s@eJb}G+Y;R=t))KzS#R$iyv
zYyQiVg{R?VbGM^Jp^4cqQghep({F`xuU9-U{>5HL*@HP_Y?CZY9V^dbcywn5Y6;o5
zQX=D_6_g}+RUc}jWW()?S~t9+za>QxO&MOVt!!n3WN6-pVUg}2M7E^RQ17p$L}ik_
zbZAVQAkC0!F^NNSst`vJYWS4)xBiSiI2;7fgOFZc-k|L7mR+YR7wc>4$`sY+Qtj8t
zPL0~NMO-LDOW0jT(;jWQsH&&KQaOF!bF^4*Q~-7a+~V%XvyvRTQr|JUjCL4F+9d^y
zo-@wvYd4b0g@9|A;lTT1t5=Z-$a!Toa@!FkPIJrRRLO%Va%1JbNWtb$3Mq?Dji}he
zETo#CwvD5$%V^a(8Ys|+0Ul5s(jTqtX8UN#vYpg<y^3o_G#<5LCDb6^Of>q~&Bhz-
zXa#?%x#N(=8|maql4eA{ajyYvuw583&Feo1i=Nt$ANfO?rG6Sm>d7k0Kr0pX8$|CG
zM`skvD-L-Qoet_FA6lm$6&C3_n~c{p{3{Hur(b@?t+!Kj_RK$sjfXMN_s}NJvI6@w
z;edWB#>Qu5&PCds`IwDk2qR#Ay1OH=zC_cT#u9mECz+;^3nS3M<4axx-a@bHst1mI
zs{Oof4$#l;5N#N*YR?;U@s!cx%td*Q#ydpcSpjU$;s@GC+k)1}>8g0vdcP$5QK!Ha
zo`O_Qsho~7WS!J$Aml*k_^jKMFp1epd9+5BJ3(Qa&T$Ofuqgosu;?Ffd2HGO1t;Zq
zkLB#sLZqT)DOW+KN;Ytgvi<)=3^ub?#<#83Q<kXEx{tdMd_ckUZjABaH3fyPf5H6|
z;=VHvy}Dfg`4g{~B_KdQw&P=&$Vr9c;U15LQPQER#Ez^D<lgQro4@la1+CD~;PL%j
z?x#bDknzu($=teQi*2C-!Zv6|9LT8d8uHnz6L`8_$F14CZOhrz&uj-HFUnS<{x-Z9
zThQ$N4w3DqnE6^KorJ~`(IL?vfI|VzWJ^B9MzUCNloT)@Z*1Rc1ly6E;CwzG+2_mg
zSmtQf4n6DnV^sOxIIJ?f<3t091oY1w6@7sc@VQ*%y=WAAxAwxLc1wOO7KdH0`qkUt
zQxrn%pk4!B$bw2BimoaI`mDKs%37BL=VB%(U%yg+71b=*GY#xafsI2xW(_CMeXNHh
zz-M+uk?VwQm@}?QA3WBXfC2aNOnfP1zA!?b$07ftq+t2qt(451Y^?v4m$ryAVN23>
zb5)~g%Z}_gN*ob0_-iKxazzKSn~V#B$8im0^`Q3U4K*W%ykd1TzDlb8r<QU=<x>Lf
z#6U{`uUEIbN5_lP+|f>;#Np2M<3sMrPbuG|!Vq*sc`Cyu?#>UV=evMJgVP6529lzs
zd%*qbJP$sw>9$dL^<8<y(|lo|6FqwZJIWFxu%4%<BC>B@`Zz1{8Y&}yJCWz};W7j#
zO{&JpAKhZ0kDDYkUD&*EVpbl7h)P0Xw%9I?*y*q=uLSD8zn~s*o=4E@Y93+X#WjCY
zG<D|X&N%3NFT5X%{`w4#OgM6d7+eIIsf0QT@+)xQ$q940Ki$XKw8l$Xfh6*oz9N#w
zoZ1ojbTE>~k!gacnl(@PAR2~=--muMCFnX+zi_KGItY6frWTBDa41J!AKqq4^4DLx
z;gLxn)D`0&tW1V1wP*v*>YV0SwdE~PxfaCCLI3dcWfha`hVXSu5Q{w%s}_)2uTTnO
z4UG$)o?ldHE2|sf^%Nc~{VeLMUrSg%Y7Q;E_Y#7VW&&BeD$tepEb!=?ACp~lPZ+{P
zR$rk<*jP!71CuqImTyxooGow{A5ng1upeiF^z`1#7l4rkxxiX5L1NElQIMvo;RR98
zP)oR58BgHKy|a6p2FWHTdUZ>K^H8oLlXIp0arfCQ>w^LR-PNwyrLP&qzkcRy?rR-e
z{$W~X;bqZDR28S?-|Tr7Ur8P@KLn`ADWO=|RhH}YI%Px4p?EJAkHAf#A?Omtw4Jx&
z<4-a-UdS>rl8ja`EIb9(NVwR)N_k(c=bSw=EcG-6=hpi$PSXfkwN9n#z-6hu6kjiQ
zV|dMbGCN~btVLYhHiGvf(;v@lmuDMSOs=c*3bo5wZCP%~@TE=V_}dTZ`n;~dar3j_
zgvPzCpD;3TU`7DHd+R+cU<x7}HU6<sfD%!_m#3EEH{b(})Pb@}?1GClhvX>k4BOa>
z&}(ekBVF%1+YWOIs#t+#>-IoKg;yzhf!m&Rrr|!`g+j+N?x3{2a7`!gZ{K|V^PjnA
zRp7}zo)y`gro4sCgPWb{0})P-P%zf6qFfql2g8M)6p3PD0IX;ibrA6qzEPO!kPM{&
z#hM&x{d<}PFuM`v&Z>MI%OrMxjdQTl<oPV<VtNEo8CFb<J)m>}Nk9=X4~+KOoFJz#
zy-26$E!~<Se<IdAo_V@(0{mAa-dNExqPOo<qay3*5_qb`Y4)EL3{KQx`a&en9B8n~
zEg_hF@9Fd?r6i(W3k0@ikUEN3v6fQAyPUAAD=BN8A;if1osyKLc-E4I^ulJh@WiP8
zg%MtL2t#=Nd1KGeXQ#FhZIVt0+`04+ZaP-&?pH8;ZDM;w`=MV}e+~R;#YDXNxFGbH
z+lBr{#OdhcANNA+E(nK@M}MqZ|H|B}F#V(QyC#`H*1q4bBfp6$l;B@Gys(m3KVfwl
zpAOz}XxT^SH;enV8$dfdhMe_qzFMO_Y(ya@%P{cJ=?A)T0~yUf8NyK(CTApRqxkuO
z3v`p-y6&8Ijl?Uh9(6^ukFg8x=2BNqjY-#Ma5be)AMi&`Kn|NT4uNE-)Kk3A7R=8g
zyrI8dw|Q?9Z2jSLuL`~YL%{6aCvuw#bnZ@1JbJ1m5M`nNXX*ffOlwNYk71B8q2DF3
z3b+HPs?kmRTcI?wnX73Vt3%Yrvcq+aTgS@vz~zxP*KK>7VmOhi0+JwWF|ILP6{ju{
zav#@->Ne6y{*ga>^J9ujq}+H^8AZGpA2o)1kFKs|BqhVPtwDe=wU6ZOjR$njs|2bD
zD}@*;a1S1|rMi~v3&;D3l$rAok2bqyKVGA0Ua}ap)wLJVzajlz=XrRbwO8fA+uC<`
zbv*IqOD*1`rA;8j<rG&pCg%@=yv-b&bsxC*41907q0DUP7Z+fZ#FPz=_kCL>#?eX2
zloHr-@R_h`rd_q-@#D>Q^a@d2tIUz%ypmm!QN$dZnc^%X(Nuqeh9$F)=b$fkV8W{6
z%9NLYCti-_oH<#_Ix#DKLZZ{#B^?wxSJg+Yr5P$6s8-PzXFJNtP76q@nC7RZNo06z
z;OW3o3?s0+nj5=AgtF(q`A7QcUk^n_e9iK+uHO6gG`)+)zEFg+^Vz)YA+tOwG{I*a
z<pNlN7gZI?66GCiw|Xc=E;s~>QtD@9-3Q*>A*ZP~B@-uJmUtPNRC(p^FQ(nmOSO0r
z@LItxxcfV2M`^#OW!XZ~n&NF<=fi~IlCsb#Z<<s6^K_RcZXd0uho+-iz@?G8IBma#
zNAKui3-q*=J@P77FJGTFyN1^VHa2O_ZGBe<rE%kt7k*T<uNx_daF3k1^lkESwhWlc
zDkeA1Uv3!wi_I$C>JM#`qg{ZJl;1{>u>fDy>N&)z0~frFRlH8yH48ruuXvN`D?v;y
zr@4G95IipPjcB%@P%feK-98F`6>02x+4}U+acN}5XJWit=@%BKPc%Zm87b6pca5<)
zT~8*pL`fs9%vgw>TWR7v+1F^=L=%-%{$@sU#A#AXyup9H-~3*eq^!wn+@NO{F@he7
zhQkm*4&wnihTOBACv!+O>|lgDp4o97H$*MPtOox2%Qu8|QCNrLYgyF18NGQ8+lRt~
zCy*lc+v=05kTKJ<FU5m)5yivof{l38XJ7uL$l|gK=yD;ZFYOiCiswS9E@HnK-LPNX
z)Xzu<4AR5X_#yV2lIJm>KB=rL+{LKc6?HXkb&x##uPWU=Cty(RSWGt;GpNFhH9BYw
zma2oW0mBYnYTe4rErZiY7Qh$1`PaJ3o}ST@)4UpF7P<Llr$W}M^(q*^VU;j7Ves-{
zl(o!Q&6SlJZQt(|%s5nO3BiV07UgzE0gVF3la9g?0S)V4c&}oa<q4CXhT>JJFwq$7
z0H)~<D4YDHwKa#1ZFSFPb3T`yBs-<Z@{8?bDXU>|vpB$%+9u=YH6GamE0f&Q;agXe
z6FS2WB(~*!R>L3jRMW|5si{X*dmV@KUgW40t%>F4q1H|Yyq;lFogoXj`Nvq$FU+<F
zrIE_wY%S^T;lka?3ev7W_s*02$P}~%6K5_falSKH1QyDQtAAy$FK|z2n1RYqgf%UF
zdgvV5+!(Z%PJC9Pm+E)%C#YsvuV)os4kM;<*WMk)KkkLXNawbkk09-f3n=(L-gWET
z#J?i!QdYD)yTE{oVU2q%+pShP7cbtcw4D1J{Ukj!?uA&oyah?jPS3Vo);g4DeG<e5
zx?CkXa_eW^x35#V?dc^Q-Cx#<Rv`qn_)N5sAsVE<#@RZSxn~kA@(QZOCn)(XG|8Nx
zJi%zFl0!^MZqR$~L6fn`Z1A$z#o}Sl(4NKjOH{11_qIz_sWvU_K6?7BqUJ8$1S%`I
zRx@ues^Qfv_9aT-vP5j#CImGG;cO!qo&?QBGOnPf6=^%}z8W;Qg{7QlW9h97?-yNf
zFf_lJUGlvI?ZzdT29zO>s>SVnS;tjGxo&1D;#Bt04vU}i4d;9fuUOzhfqrC-h-1Yg
z#zZ|{`;ZwIj4LjvzlIwN!R?`a&mUyfDK)>s7#G0pO|MXne0>ay&wS@o1i>BL;*E&6
zO;>*dM052G|EOBl>$CRw*C~mWZCQBA#at5k82m{TTEoIItq|a690s|9i*ykB+>VR_
zu-iNSNv!$|MO4g;DAMLRBs13`_#n|ASGI1DJl~fGraEJn?`cl!iO;Jm30;hT5Z5U>
zUyPOifKv5A#cH<Hq4;B(eabheWF-oxL4ojdFpGso<l(><WFRU8>pzR`|DANi%)#~F
zk>RFvq-sdMo0@xm6K($?j4_KFf=2?6WK{qKMBLa4$@^U<Al`fl6t*&SNuP~9&svpb
z(4t4E{iZ8?K7B$F`1rVdeR;go+Z$u7=nb7ewxh_!$mI8%7panm{A&qaG9w`W#qCJX
zs;AK&`~ST>J-sb5O{@R$^4vL~j@Q&T061kv47&iJQwXbnQSvbM)#*PAgAeDIr(=OF
zH=`_;k`E1E9A9s*?vJM`Sy9R2MxI|Nk@JC{cOx0Tez~daSEP}~@>!qn5BKZq0utq>
zwDNj;A$uhPz5?(HK2rHknWAG&>g6a!v$9m4b8Aeezr~dHG^Wwy;DAn>C&UO4P6_WX
zwuZ_1bF$JL%BrG4+YG>T&$@!pMobU-I|+ksPvJ?7_F+bN#yPak{qn}<bU#d-d0|JY
zek%dq&W1^CE@kjkJbn8y!SpaqsYr`CF)zZ?g!~TTP32D>J`L)oj-g`<YtO^Ij@cs@
zKUw9Xr2OJc)p#Gs`aB)&A?i1bK*A)0!BKf4$vJSLac=tl<<X>^z+>fbZlxK%r*r@g
zob;H=2NbW1!dlQcb`@;`s|WU?N;S(9@BEW|rixa0C%9DfxT7n)(Gz9AU4bST)An<E
zG1)#)Rx*v-SvJ5E?|O!CcOsd;@m_m)VFVY^Z&hMdGew4wizeMuv#}mYa>r~zs=!dP
zd3I%|0&u2A$+%z@_Zz2Y%QVgw3_z5L>Np&m4yS3s>H_!m+>+y^apr%g2j}e9J%K3j
z(B1(*n{h&ms<IO;dhD0)iHpR>9AQ@27&s!140%%bK=U^fshgq|L~MsM0g_^%13Y$h
zNpyjV01dp$kq_Z5wHKEEUEaI@&BXj0slejTYbSVnZ@LlFt#lOK8GN({vWxmPucCXe
z9-ONzJM62-U0<g?6hVo<mqCmVrqb<@ZAUXkYWN3_NvreZtkvkpp**zo6(%=NT&{d;
z(3s{;hK!=Nfx7I=aA+ps@r!F=wmO3hbD&+z-rZ5723H@e)CK1y9;K}Xkn{6tek;zd
zsk}?%T%^f02fX1i8u{I4jpw#Y!4re*v>+j{Eyg~8zvqDS&@34?r%8(~j`-GQB2UXY
zLPA?!!C?XreA82K2<K&|DZl-yIk8ICXT%i%{(q-{jhSRp-Bw&3OAX(P9#a5iO|i4F
zNJYJvpyY5?*_dA|ED)i7-+wV^3>=^*8GTOKcRir{n#M2TWk!m+2olB;SO2ytlb0{f
zpT?3>tTrru{`Qt$$^UVCY67}DzoUsO*|nQ}Zj{*c@SVgOU<{3t=vIUqJ3l$Wqv(?N
zgQ8|8inf2Or+0x-{_=}69|qDl6p0IxBW@A6R=WKGaEgvTNNnu~bG-M<pHnzK0V|5a
z^|tu%h~`Jy82u+<rX$v-*0Ux~-uC$LXCW~B9pU0YC6{D9Gj#0D<#$g)S@a=1;RSu6
z9+Yg6d{4}*4~ASH^)wOM3O*uep%LrBv}J5C6HEvByyPR+c#I`?w=9Mz!b}kP0cTRS
zCe#MIFpBa|E~V^%@&OBQl31>KO04yv{C5PsO#xj}E{?r<Y79yG1#2a+L!?*5e@@b9
z=S>0n?sd<%=_3=hu5rQOz&BaexFuNxO2?WC57M@Pik+r#dI2>DTe0qjy$xrs?!8ES
zyk7j%DUa|PH?&JkdnRRF0L^LzJd1}B&9p$q_V0<uM1P_h8##oc7(ZIWG{q~$js?BT
z*Hk)+8FA8=ljM1;DWd$HYNL03Q7thcMp=9ye!X321J@wm!7lY!%H}}5Dkm08|L+A3
zRMz2<U#+xH`tMbEKBIU`5*DH@u93|Xp*MFSLZgI*MI(W3>hpR$KgOXL9PNClTKV`)
z5~hnK!xOzMt4)ZcH16Vqi!vp)oMa`a$0A=Kdy`^0^X24eKS&yrWTi8br;(uxPoXhW
z-vP^?ys^;ZZlFPyu!DD4?vncFeF6N)v<cb@2(M5_ARGv<a8LH7=6#3cF+>o}NhkUi
z*2ocNi3KZHUGyht4v#k`6DjQCOFI=kNAq-6VP110qy$CSAD)zlV8Q*e&wu;m=F^K>
zrAj61JnNCA4hsPXp0oDv%Kh(r{}LI)^Uz5al?yZGGlz_wq^N?xhBM@DBD>i$1mkZ!
zoxV}FVX2ft|Bmf3h@Qs`=qniL%MW^dshk3~Cw>A)0}hXQDYbBc9&zuPF+y^~TZ)Ix
zKhVT!M&2KXJ*7y<nOX3P!!VwI5#sNJ?9e_yRRF3iaEDnxZT^a&Emt;At)SW8L*7!C
ztEyBmB(Zd&PuJEt=2fC?rJvL@hg2t&(D;ZETUN86Ko`I6m|UGc8U1Z_CN*ZPa90s*
z97K(l1+D4-nf*Tc-sr`Dq>#*8hKXosTlXY4vE-vtx<vYW!Tml%j57j7J99aeVT-s4
zPAT)Q=b%meM^b#c-ldF%94ygf8-6gHQl1Tx@Uh1jz@@f5F+O1dN+{$7riJ>h7Oiw@
zg@)EkTvUa$D12GF0Y&8Byr5l3q()mKYV+-w_AA&w=`<HLyI<DnNlm|40H2GUL!I#q
zyY=`EjA<L9h(!k4t@OlMiwF3yDisHI5dr}l<%|80h7{|q5EwbjjMDpQG5sNFZnOym
zW&Qoj1L`=6u7$BI#b-aw%$j`KS!S-9Z75KgNk$H}WS^p%n^M&E2K92`p>rGIvM?o{
z(rK$~Sh;k5OIxOZZZOud;oMZ2sW|B<23j5vdkmgP9G*;Odmhk-K$%32?C7%j>K=>2
zCXjJ3rSfsy{Y5R<T4&;LvQT>x(IcQ>=_=;lumLlrqJ+fs{JES45Sgj!v8!xBjn?vD
zRXlTL4|w#c3Lz@yX8iL_la^y@E(|n0;K3uZTO!s5mKXaMuWUy)>fgih|NXbbTjba5
zaglomdjm$%8QX9Z028mywb8C0Xl?ZQ{D`)lzxO(pUK9u=+N4Upzc5SV0i#E9J!lth
z#QQMJ@z^%a?N4n+3BV_9SVpH@aKf*9-GXjoiOEVFLe@{zw9>mHy_q;PArpwlReK8w
zMM}8rbugy1#g@87^#@TMcq~Lqw+%OP370ls(aDlmmnGa6HBM2!*B1<H89KEMI+~8l
zh}%!)LjqTuC|hg#Ypa9KmQb>YtEW8WNtc$qP))j%ci3c^Rv#P03>S)7A2@}<)qALW
zYQ*iR3I}Zf36ob#L+onkc({hom51@!Dh#?Vx5^E>57j5s=3F9b3lWrhFI$lr$#et0
zdD{9ae-He#?5cpSZBiL4#rm+YvCn)(^=A|_JI+C%pM)9TV|u*x_Y)|R>x5GKkpjkP
zOb%jL9Z4;{SgQcZknLZIz!m4rGs7f;$@4vRLKwka#?t0!C1pZAun2m-_T?UT;&#vs
z#e>0?KLnGgkV>QYJ@W~XmSwW-nVP5e`R+SDB_a<^IOhi|JVpH#``WkqiSjYZ>$_6@
zaejMZ<Mes!_oVUhk{~?MtCC+ba)%8RVHLB*kYpPU@G*;sv>`wIv5JQOo%{7};j7ym
zTiH&8z9Ml`m4(qTktDSJMSa%-T8V1uHdmR>w5~433eyn7^H~b?sAb&g{L%OH@DZ(A
zJz5i2fwgYvVJGfRI-4x-NA|Hm9f&=LS|%4(>0E6SMz#<%4)!BDCLgc%i*Ks@<dP~>
zffAzEldpx{c^`A=s6YN^GCQwph)jO?HgdfF1)}Yb?)4v)GwXj~Dl@UN|5tYPq4rqx
zJu6c8j7lJm<oj^-Ep?Q3`c<0H9%MSsUf^GppKf1Zhb!`XfMj`!fI!Z|ltKEG9&KG2
zlP3v^4q#3-fx>oI!Byt9IM)fk4-l0v_P&gCv$5Dw8akr<?YZIYMN7KYN$+${nRxE-
zq@~2p@x_099?JBti2iq``pmv&YZ8;i94(Zo%cr_KgaXO;rO&y7MN~eO-P?%~L<Qys
zda>FY^Y^{4E&l9J4{Y`yl#%dRY=o#zJ)bYAKx@-L=RB}AQ^w8h9_Jo`kI^uBrQ2}D
z_$6$WJaBc)eIDz8-e<}@lj;-(b>Rbj=`uSfiJ|Qq+Zf4iV<<j3PnJes5lstW>7MyS
z>-4gd9xnomA>S7KAT#fM2;_NpZ?B0vJuRhfbZV8Y;NLFo(Rqtax?7+oe2jt^d~B($
zSFXf&u3K4k?i^-2-uMS%e6)LDxgm-!rJe_<`b1TWHI@ZFZO4(;8$k@%=Hv9$7?Ij&
zBi0(j=c}8#-KEDa5=!Dq`g~2T1M9Wa!xR`N=D*l%id*3d^kMAwPX>>9kGV~(fTw3K
zURb92zk}nJ$3`Q&JgO?qMA@LuZOny{kReqIg5x&j%}u844)ddCx7bi!EH`uInDqy|
zFSs=4pcJ$JBs1Xd&JrnPnq!S>#_F!(PIzd(dj}teK{MMYC$bNmE9DLuk~I+Mi|;bI
zErMMy!2<nSvbpMq3u&JQEVJyD5e-&8?3rN@`^vTm{x;~lA!a^Iig7+r;|xV%NA-8?
z<_fnQMaEa?0F-W^Q8l=^wRyOS_U}$~Ty)O(OTd+5W%&0Jtzc9psUIRb1#5m(thkeS
ziQJj;{<wG6i&u^?h*b&gQ{ctkcaV`VhK6r7+O)xKgz@kkTwXv$S6ZKCmGG#L6S?&1
zR3-Eg-a%!*Sv!_(eCqe(*^$Kz(&}uk&EL^}37*5REVOruP+A$J9=|cA$6ZG#jAQ;G
z0#YxjokrLg_@I~$_xGp}k))&N>trP>2hfUlpxr8WlP-82z~L>|U5w~uJ7(u+bSkWs
z_r*r9d-cwGBSP-6HCqY50UO-P7|Pw8vXq@@X<>RY09P`@$iSl(Ccipj0oP*&&%m`s
zwo)+U%pjhU{3%`874<~EauTG@Q>!nUx1;^*f+z_2pG1F}Bc2v50pIarR#|W7nd4$Q
zqvhKLl?Zx(d_y0+?Nbc`a**%{6!ULp1ee58(BIwwMRbFZFRF&sDKsuB89-|c4W6hV
zPk<C1*6UX?7m-A46O`;;VS7Wq#?Gj>HiFQ-CRrDixKKW|`CYPY9_tTQ=6t4?#AT@?
zsgmpGe~x;aX=keceC&lYu=u2uRji;(n&`x)k2dC+c*vmO&RF1^idP88fdSioA1-IO
zc&_2|ddv$Ktml0{;nW~thWG&C5y|OO>3!u5x%&_ab3)NJZ}ES}_HB(F@*^2cvsHy#
zQ4#s8(C&XeMaoC55P*QfD<Z+T`u{D27Np=zSuy*t68J6jqy%vsu|9?$PHMY}v@m9&
z<)lY`s(F<-McekJNNfnZcm`a0g-Xm}V;t3KUs$&s;Z{z!P2&5f#~3a_y9k*_nQVfJ
zk*LarT39+2jbFRq#z4d1s(m7;xt~0cEF_bjmh43Rtir1qu7^29raM|l8<gSop#a4$
zX&rFudJAy^J{B2sg%>&l02oJZ-aT=*i+iol5Lk-aN0B0MV&r<k_OhdHRyqR#I$ZIF
ziHBZO;Ak-BxqY{wKj8>_v%`KFP0Hc;yDWm{7t^AwUp8Ecx`e?(np+qGBq^|U<s2q<
z&i=i5kReLEW=b?*pEdk*v`}ug`XTirz`lUcR+iu~WFhefEaF)vD<_qTe``!}^M)sp
z7Ov|;31-zyL!9K*4md`q<{$sZeUb1N@;t*y6xpQ!<|^tm+hgRjHniWYq~@a{gW~j=
zF)m6cjX7vp^ja&0zFfhmiIrHlQ_=VS5LDA=;V(5kALMHNjoCSS7UAV3UJp~!Gtbl0
zByLEhUOLPK;|CSDIt<%N^l&FH*huU!bAnindkr4HPJO1NKzo+cP4J5pEXqIkWGUF3
z#XoXOL>zn-BW0{0M$6A(x;zxHL5*Vo7iC8!c~^zExT6OgP^>3`KhuwmWCyOij`q?m
znZB@RB8Pmgt!iBUC|%G+`$O(JJr^#+^~T;knynoHIS}X3yFIcuoY$+~pXY{ltipeE
z%61OgL{6Ely}-wCyJ%d}#&{%4%;N$X-=zDq-c~=X{CS?Oekj5fe+<-+pYXh7`QG#r
zc;6UI@#U%dchapBirF?~%_>c54QjYr54|d?V2i*waxJuH9#*V}O7UeScRU%YKknkM
z{e%{GV}++UW3n@D5}cKz2)?DceCU}AQxbwM66in~(BRq#GX(!4$x78k2j#6ZQbFOZ
zLs3E5IvHFDQ1~9eIk=hrJj|RAZj7kO&uGi?*P#p)I?GcOlY0nLh+{GI_W9<b6vR$W
zWUhu8q7qO7DY<DnT{7q5Esx@<al8xT<MHg|?CWCV`M|*9)trD*#77xa^MURubRC45
zR*ELYeG}1t?#F_>iW+C#%!qQ7c=KXsC~Z-1U!w8p4{E}EPE&mZv98uOa^c`Mf;@f+
z!?0ZI<M~!HN0SoP7nmVF-jv&+=Fh<oErX%riFHKfGhpHkBLyPhSBn~tJ-eYaIY9#x
zydo#OtxMnTu0FiDVH}M)ho3VU;LkA;Sew(eV?rXj!;M%jH@+%Pc(89~PkuM;zrfgn
z7$yp!v}r#e+u^La=7#d@eobRvIYBb);tMe1iH%ZZC#wJ`aFbL3IL9VUamJfhLBybM
zK5uTXKWTuM|1s2@h3!Ao*O}S>%U%_2nHbz}s}J};ArPPpL)7)pu|=TjBvpSm>Q_ZT
z)<X!B13{Q*VWEsZ-cWyk7&LU0m`m6raNXa_B1TkV^iZoeUGO{Cd-~oTogNJZ%w?fP
z_ZmIj*MWVC+)_yTSd=z^^i!T>`<(i?k&$6{Z(SD^sfII5Y~yX8x|Q_aKrz1f?`JJu
zTVY0&FpDeY3;!;?!@Grvme4jgO^wk2`$W8o5#alH5jR^Zx+eF6Z+TtrD9PBi^$=2o
zX6|O0gLZ0v8)4wsMT94fe4x##rQJ$F+#Jd1S!jq}Jovjv8by>|I;L$1!fo8CxLZ?5
zo-BbZU8cpCp0<qmzC#4VSU^gF%8){(rJ!u6s*b8*fxqSJ^|GL*fWM8%;z5<T_z?Lg
zv~Dl>pEjmb!-Mxx>Xab6lu)AE9RF9%Jd3yVHX9i<7#>rWNy4w+w~e$-&B8REa%Ckf
zwo-g}c$xIl*cZj(bE94%l~I(Nw)jd74Q*))1{u8hkNq?%F}9ZTl0y2<;#hfAIO6I$
zL8=^HqVg09!)u}>p_&4c{=G4mC{_6HFLL#2vntO~B|-v98i~AQ9>G#GAmPz5puhI6
z_Y4w4qKm0ik}pv*!?x(^;)*x=#+qYBs5mw|nS;%8>lUzP;W(F<b#JBz*TVxKm-^Ac
z`5c2V5V5(Dx3rsYFo#zzmQI*}AwXL&GL;kdAXAuB#rIPeh`gWXPRd+9m7|;kuKQ$W
zC4WBdp1c17dBsgFD9*XmA69f4)mX9FR>cEw4^l-b#Wb3<>{nI9QBrghl?!n?^mWPI
zUcKv1nJl~TlHwwvq|=zaz6^v-n*5wS!RKm*tr0a_IZTkPKpR5CD0v0+qEa8c>HN(G
z4b6grx58CEvMM?co_9U(|GvdtJq5UodNr0|1+?szgCr-B1;IHI7HLr^<Rf~nA`>XV
z>_{w9+k9u@0(A1oR#aKU-y5;;@S|PD#k*QB*1FQ4&c_^#g0^)PvVVnh3_aWzuEhPT
zJU?AHKjjw^>N(VV{z<W~lTmqWNHj6cBj<_iLYhQZk>o80J$R*DI)T|YI+W%w1bKCn
zDf;92pWF1n2;3251}PjxuBG(t+vOScFf_?jA~;J}lN@rerJxD2OKPGvuFt`adZ<&z
zC^R&hbxaVOyf-H+6^9tAgg*yEg1vlp41Gw>vew<GWnOyqEkMtl7J|{X$$7~HuE+=j
z3TC#`m@)3%QEiG+g}zl}3X0anW|BbN)X)rEknFR^7T8PR^D%RYw>A*Ma2{_tpeY3#
z-?Y4Nv+FbkbeA(m`1*Wzw-$ONBeh^?&BX}mT&52AXsq^rEJQM+Ahp$_<&Wm%DO%Z?
zVBu6ltqGXB1XWK_Mgp&wgYSmp$gusW0nT3@tEDCFnT%_=yf}3vCs9^unW+_M%Jb;+
zUejw0@Nezzs1iIob`u1=9dNp6AqR+u{e5>vAMpIewXUmRtZWiakT|e&FlH=ov&K|W
z$e!N7uG$s&Y7^d}RPNO*SW0ZFMdX8nPOmU5!JeHJ#jbP6dQyXHKbnn()rSRn=Fxv_
zIz$#bj3o=5-Qf7u+|7|x)u<@Qlh{=noWF7!dABzGnI072c2#O7-lo_fi56{j&tgU1
z_ae0FF&vZ*eM=<;AzHfRy1G>&)bRmy6YXx52ax}6lb@TF(eo^DO-nH+%}owuXQ2y3
zHsi8O-0Xv_u;z)7Dm~B@tMFI?1Tu^~!BTUk-zX>*RHxEnoD<5)RgJM0w(%eV97m*)
zwA-w&zCWX}$7lq)(SB`-h3--@!$~j7M_!FZ<l3o!DBlb+R6343gIyQ2a)xem8kC-}
zWh2IsNTz{to;dZ<$0ZzvuT(<781ed4@#J5<*I3`;@k0Z0VKPkM1luSqOst$t;a;|N
zieP}ZnZy~M_HID6af$LDG+_&(0dB{MOu9rIhP`OSWRvT&jepLqBoib0&dyh|0UgIA
zG&y>l+4?LP<z`+8{a6fwMtu9ogQkI1WaWhNDQiIS0HP;*vxWk6^r>ok{bF#k%DQEd
zPpf|TgPQMCC<+77m!+0%SpOJ}c&AlAU0X>(jtXmW3z(dxir&iQgi;N<Dc_k|LBq66
z9MjvqLFT3|l|3BZk!-cJL*tOz01h*E?y+OkqsE0uzVXJx^%wuaCqHSrJ#YV4Sn(3q
zakyJ^sJ){W5I_q|`}U~E2o}3U5iB`74IPEvko^LaG;FwN5`534)XOhDJdrzNtIbYy
znkf7n>#iR~)K+1KB%4p0-V*XE46-s(sRke)y0-}k$<JuoCw2_!eMXA*K=IAK18rRU
zxztn)f@82w4!0vh6ZHayqpClC-FSWeGv;j5%laT!_BTxt5@5fyxThC66x5vhK3^;a
z1&8Om#$R$@FA;%6!r+C)Gj$-(MX%epEH;B5;BMW^?t3HTbXN|%>Azw2KjQ=~zgV;x
zcC1)BaxaYzSKxhiZ)KT0K@&0(^(~5%0E&&ZUZKan3#~E7H=W@|b6%9E9W*jfmO+5>
zC0zohOG#tPY?o`nskZ<#^io~d7dGS>-4!f0yk8M9rS~0#z5*<8YFnuLU?W$HNVZ1v
z+^g+JW_kOA_25(TtY6FPJan7EQ;E>py?pE{xVLhG$#*uzv_SgB1!p0jkK&c8Wq*87
zsCY-rUVnxX=_?-3_IO$gMqV?sx=U;?m}Ay{yDWu9hvP-Hvt~$PicLpqHP0#m>n~&i
zdFMkL96Jmo0s+Z#kFPDNq7{~<$VCwpJ67<`6@v=J0#oI?(AR5cJS$cSZJT1!)Yq0E
zHRE(36{b-Sj1K`i9Z*EFi9QFu$|H5m^$dqTBAeir#F9r`WzOa7dAJE@GS5poD*#iH
zcxZdc72|L3wMSK?q5ay2iwz8&8>}3qzTJJ7h=EG6v7BU0x&5q}%uQQjg9FYt2=llz
z3+YT2#HW7vbpnsN6`_>R+P)YYN?gWkQ?q}>)$mAW;f3~i;tPTct_kixdOB?XDddy&
zzn4hs;&9*azS0)j*(^A&Obr}bB%X7SHnvrrbdi6oZLESBGH-Vi?x#%gbgS_J=qEzS
zWHFKmIRsP_M~);J%N-x{S1(FR#zFH2b?&bVAYX(pD5Tr0?RDSU$RfL&75G~+EI5Xb
zfV4>2OY`&Uxsu=tjZ+_#oZ#Bt4qD&I`bKeyGAmp7clNaTC@yYI<Mg;-Oq(C<aR{P<
zoeOis)_9n-#XgR)R4uER{1o7qfRrEdiv37n#kEY~-TR3FU|SoSSY=*i_tgGc|0I?2
z*In!zziBtW$=STN*iv~(Iw#(bIARMz8H5G=7kX56F=rtV{fyLF@6KyTc=%KcD%5S}
z@W3T{fcxVqJ)hC#J&88u+<=5nvi4S%`2oa_^K8LbwbL?zJ(3uO4gnpUygFf$Vdr49
z+w+voH^lmvQGsqtO?6;;&)yuv<Pi11wvXM`rH}NdrG30>-fGHoO1;!F=57_h<2qE#
z5bvunF_t_6)-+^Zw-T-NGInD7xQ!E4e?Uw(6D^SO;CA1A(Gw|pqw<XF0vz0uUDaF0
zvF$8r!M_(F$=^IFzYt?OO-8jw5cZPZ$X{9Vw&@7imufB$n1Gi#8L5Vv2h!3@jhlCX
zp*;jfUoV}<Ws0^moY>8IR#jbjd(+tu=NQLnlZ2uh`D-r2l79X)v?9DVu;4pd;0ZVG
zPecQ8FbKG-@StUf11ptITc7*PE^B6j6fn-!ybX!Mw|=pj|5)PLqZ~Ww43aX@I>Mfk
z{Y@9P^~gdkd~S75CDc-(|JGK<lBNjoO`oWnpH;`#8@FtotETbNbA};PH1=epPS-=T
zUVfGxNGy4==o4)v-qsg7qKRNNw9!E%tKiKIu=mi@gYU1Q6vmp_g&}(HM;tsIsMSZ3
z{M$0Myq&Iq>xi}*DZpuW&2lO-adduM+<wPCK*O+#K8INq(#-hpr&FoaERmmvWr&ha
zNFi4s==CY*pdH>m;MfKc*ODahmV`cJ+Ah9ZM^m-c#Emt19jou9l(^KyCHX=chUMlj
z0^+KDGYjU+1u8jLP~5B#PItd7%WR(r`&DPRdD0ZsoQJPpy+U~&rHf=;|2OfEeN!UQ
z-M++*x*~9{xHuGm{lFc1(y7fiw+ivjMTLob3T2Sq<Eopv80zalnUFPk<cI)JyC-QF
zZcq^Q9rcON_6$YXRp!a{l;kw2I1vts&v|~Y2Rj7#-Yl)g{<s`X&829m4owKbw$VdR
zvR-r|hy(7(|2`a|OY0$7Ay4%m3n$7)US7|)3l#;_(fOnEZ0g2xn~Pwz#tAo>Qfi={
z-Xbv%x-_yCsSxS@@M-msgF4>eC3R!=$1>N(u5-;QK{ah&uGX+hi3mCgWNK8TSB36c
z1Uz{Bp=IoozdsC_hF>IY+}URW`Vk4ba4%RQ2Jo0~#NI$L;9xOKviPI^YCb?R6O++Y
z*G1DBP9*thbfYkZ)gIeUd?^We4w`4?8JHF|iOur}2%w=VSvdV}XRXXRbO!4vxo8-_
zZ;C@6Taxc=GG6#oQ5AbtWjpGKLGth@)oya49{cb=4@m&M)K8$lMv^apZEj<1b+!mD
zp~f`k^n7pism+fl-6v3Qt@k<(ct!uPnJ}g!waa-NH@`8Gv6Mn(iMjBgq)Zv2l+{2o
z&hFb%HaG+%SD^Op#;E>%#IlZ!@Zze?kffA(wt3;RjG;}81jTWs5G=d4R1bd=R3+6v
zC}OlS14ke7S6<|_H6rDud)iPqqX`E*lI9~Bvnan6k|uCT!$|H1RdZCc=*SnW7D=BY
zR*Mn?Fv33<K3&!-*R^msLemaL%Fu}|Mf&?&^bA}WG(NXtJLC%K^W=9TJLF0TT|w>z
zW7WADJJ=k*5#6q7qZK${Nh)Wa#fdW?yQlJszObmCx<(Tf>CWS6(==k8LQOikV#Xth
zv}m^vlq19y?mzcHmeUOji6F6p!c<eKjO*Ak-@}jB@9Bs1zE6xZ1Jz=i8{V?8DW~OZ
z(3qC1D~qVc`VDRTM%Unm8@Am7amXIh7@W?HS3~_gHZfBiQ=eGV_o%Oeg?tmJY%P9<
zHRIB%qwY7Ndp>`;4LK3^6@=fGor>Zks#a(aA|%3TB~?*nmtVxQgpB<9!XZR3@^z?H
zt&W5wzh`coAvcDEdl`&utaX_rY!XyvTVyuVR4P$*8(`a|9kZ`H&Qe{_Wm@a35W2!}
z9&N~g_#>tStidNxPc$12Sv-5#9|+%>e6Cd|HJuLzbF|Fe1tCr@HYP}rJK{J2`v=><
z`uc2EzBSCMD^e2JqF94q50B4b4IbCwTwp0<S3!wTeuf6ju4DvZTTB3E$LGLd)Ob83
z>XWOXDyRpmN?3hT#8&qg;x{dzESm~0<jK*5F5^7dBT`BQ16>|Q(V_QpoR}Yw^gc*&
zP%b9<C+zD*wKstL;y8tmVPn-8iJ)~&kLvzsm*9u|Q+^#uxn?{<Z4gdXhg03h8Y5E^
z#!wClI>{s9dyov{G~p}^;}wfIos>0{)8E6(SIO_hWgUU9{7&sI;<g_HhQNg{hH(aH
zQ@QtoE#|}!R1QpG4OH{sS+=ir_`H;{q~O`bH5~Xn8v5gq*@wN$;90aoo8Z~?2%X^B
zrq=IQtkWpI7bdOcAP+KO=I(dNfW!Oj*T|?axu8}NU~{bflO1s}hXfv7H%BJ|p4Pc4
zLivbK*;L&t;QXM%@8*w1IyxK*TiSHo&O_^j(#fKPufBdwQ%g&ape2F?v@J>iUB+<p
z`sj1%XNqoixOS1@UFIZBg^#%pk==WNRw>Ft{?91OlEk0$pPn6P`P>5hm!F_{iwmp&
zQL3=B{O<-#c9#FH{IvMZ8OCk<e`yrfgvBqE-5`b_mX%DgI&fLY7nUe*jUg9Q0=^Z`
zR`F9lk!A;*e2r$s9a^+L59BglulzTs69aeW10@muW6UK%N%Q9|QafPLz7QiB;tJ|e
zW{Ii#zFD6iqiw8ieESfYG(jmtKm>t@(*k6KZ6tENz59RuZkgLj$M4Z|4vF{fPOqqz
zkRf*3=3M^<R>ApEj^}4;zT9@g9uzPx$wwMn#G{_s(J=uUcU5HioB_qa(UxH?Q1<}G
zb~^`k!q3KV1xs+=DZ4$1)>|kyYLZy;R?%f}c>%r572zd|DL7OUGzCXOh+a0Dv}eLV
z*bqzfy%CD5`AK2X!d+wrO9gQLI&A%iMMVvCTr9kbzu*%fx+5p^L$R22b&)C-xr;d@
zHE}VN6i4u3;r!iBJ^K0tpHGTAvEi2c6ePGRp(_kM{Ig(4oMjIQiIcFYK5xML%iSEl
z#X{H*bH#HIevVNuux?h_f+I}m$sKd`7|2$>1w~~_LP;P*(CCaTD4&&1{}21e+v66X
zitF)t%8!u5Oz49)V@0pxiaL;KB(`bJf3e)kHdkI`G&f%00g2IAV6d3eQAU%^?2sz{
z281>WK{*oGiWXJR8HJE*T#))`6_QIJ^fY>njq?)yZIq1Z2_x>L^kv0i;WXFka(5+m
z1anASxJF1mvzM`7zAKqSgvlF?A|}{IhMa@k!(tp0J-4)0%#%8Y+C$EPby<xmvi1Za
z@8%gy$7y@zS%+^SbO{(+jOQLi;322+74G!@qxpUZ+yT~zXCxC0fDnZTy!u1~%j$;A
z6V5xGV<TSccx7QG>k)r!26?HzPW@6%&7Z4Z6H=!m;)~daG`9H6$A`8qLn(nO?#Nq3
zf`%L-Kak9YA2Xx)WcT&?vGIjzOYCmz(k&xFIfqoukz<A|Lvq_qGn4ntfvf=@qqxw!
zWHds^U6cyB;ZZ#Lf-+`>I&OZB>d455e}?Z5hyc1VB-2sxX<C1@$Xvn6m?X6fJwLIj
z0q7e#U^OCV!?b2x6s*SARd-15syP+gt;{}Hl!i9?#p-8&Npc)h#V=*$%Hp1N9Kac6
zpMx?Z`ZK~K8u7bIR#Jgs5M~l+I(e=S?H24+uWjQms`{%;muM$q|E$q2`Jnd+SWF%r
z)WMabLAv;*((;_vPYak*nNNI>jWqI3p^?F7d_a5DyYo5=26!^4DVaXvi*v{dZ;44$
zGR%IYgT6H2xUqKdQb!99NZ9Qw8P5GUr#*YCJpnI9fyHzrceRPGgwja#P9cH4ylLK$
ziZ|BB{5(kvWOwZ@9(2!SIvT{{-3gc0gUC@C?I4YjkPM?&m|_XEyBmCwPx^qdNmce?
zns9F(u%HidRNbAOAI;ve8B$k|K%I@(Mv}`XaeVEFd7@uYXQui<GpOm0^|PUZRCMfx
zNvxsQ1TW`IYQQrc!|6T(Phbg>a(H_D(Ln=4UY*)2+*3d~Xzvp?)N%`8@<n@OZa-|H
zv7Auz%B#Xg9Tw^aQPDLOZ*9Eo+hzz$!5J5b@EN34%)Lxy49KhkgBTuSQc=-mfi5qR
zZi8B!D571kO8)}y&fsaXHRWOSm+DYWrsQ0E!31k+A~pJRP-4rX8AO(n{KUGFobS`U
zB?!+LQe9KfXqJl(3qE<%(L%MgX;3(Hp;JyXOgO8hg0NfwE!E9#y883+-iRK1d&X4T
zjY5*dcIkx6hKz3!OfBDp`}G}1Q1P#lZW$huZJjj)`ovRFX-BBIPC8RPUW2tY)0<`}
z$(mE=!o-6`ZfuhdW%}PdX-c4C7zTFVxnf6}VWkixDw2Vrxtfs=IB~CuavH=nq1~~9
z)X~L{VygfX^~~>F%9j#@rhE1*#_C@T27-=0WfwWy@=7W&Es7$B!7yNJL78rH##=~%
zg#!zqNA5R!Iyl-xriFMQqx{t7wG2<I4;rvN$ywU>cDCe6-q=Zus3{nGtv}F*(h4TE
zH6f*A{O{Nu&okr9^RVM`?2e<2oBm8~w#~cl%DYe1nsoBGn)Pm8j(zvj2)=o$li&^M
zAj9pa2|5BPtI05_#!fZvgTYC1AmORYY8~yh=n!it$tn>$qDRmvvW&LEAzEqlL*0mL
z>P!BB3u;=(9s-*=#me%ysoOHyoa}!4`na;?aOFODRR`r@R<^eoci+48#L@keG~21B
z7;%tj`IBa6@iuy18B#BKyJpc`R#H;`R9{r?Kv6&_(blq<Fh{I<=zEa&ri+Z|zfl4k
zDpdfg!>wqORV$XZ4EuZvJne)v;03P$C0Qf#eVe7;cB?@8Xa~6AWRYUypb<89VX8Xb
zn@rPcJPjsU2<ve<j-E{7Y{Qyoy^@ha$|z;QZHqX0sM{2U>j!D*r?baQbpcG%Mh0t5
zue)AJ3x`*0!{3>_k>@}XydE3i4XiDSK10u+Cc)Ly#Y26TgziLIy4K-y@A}(JmKE%_
zR5#2Iifz6Rgg+W~d`>`PoK146jR5m1a93P!S$B~#{9Dub<(c-0Pj)8<x%enf69w`J
zr+KGoYo~i4N7z=Le7M;2WE5*eZ+y7OCOjrJAz;pGt_>GY_L5FW%6S`8b?68o!&^q~
zxj`a(1qh$&E2oVPZ-q+I+`;O6WVHd+0W!|TFRVO=v5fplge^WhQLFK@AF?L>b8kXF
zVcOS(zQKzX2HdkfIxhDjY2Dt;l1+gv5zclsLR$UajOAx0T4^(uW-t8)lw5=%4a@)f
z$kJbxw}|wpi<4d6g^l7_g@AbfjH9d;C1X@?A9+G@{%TP_d4z^iWtJs^uZSzRkw_+Y
zydot(XK-l(KD<(`s~G>y0dgbc_|vj786!C$4mK?7=7baZJZyQWy|HP8#}yI;E7++^
zb?S9tmAeznGTt~Zj2ClQTmxGZ7;n4`uTJelJh);(?xr6U|8Trg4$j06aReyKF9gtb
z8Vj$Rl7>~wvdZbyF=B4g6?r?ircAtPAqNGgZOU_bUWu#mvF@9-2HiW>M1Ei68$>h>
zk(ye{K=(Rny+B$8)5-H(oQ>JL^vFQS^V>pbYM^RX*Mq*SJNNPE%2Epu1R7p)&kIq~
zFPD2q%>7Z%%`gqs@@n+fK?ut(2kd1cL~OK6bIV?lR<@MU%P~!6a)do(x~N^Maq*W?
zr?VsE<~OU-=_RR_@z<cTw#Q7l0nyA@XKU#2xT=S3{|{sD6r)MkhK;su+qUhVwr$(C
zZEM=LZA{~7+n%;<&fj|{|IR*m55796T3IKRR8lKf?n{IpZQV*~IkLOR+Oh4QqJc9b
zTV7(;#Tcz8b;tqYm2OC+sCs|q-S5hI>0{`_Hm|<2x{|%w5CG0{S79=Ah#Za3vF!_!
zgmZb+Kyw9O`EC5o2Z&Ww=%y#W^%W1b#wan<dGBg=nG$AEPn?o24a9fv0d0}+AKL^V
zSB@}Gq8npi7;Ey89`>Xc7M-1OHKebYR;0xUwN=%h&o0^NfAO=&v<G?lO|K?BEQxKP
zuU!+}n)?kv*O}{-|E-~<c-RtNl(^+zH>RKBXN}T%GlAdi@VT1~q5Iw4zyy^e{ueZ5
z;r#yvBCC2inlUOETdTO*G0GD$GchrWTRFS9{;W8;8D-7vEnF>$SeQBgH<$0Wj-A5~
z3;A7V&?wYjLO6T`7vUgmI1qKQ6c~(?4qJl$D9l*CR#}8vqVo6qLqN3d+PyTzwJUBx
z4o)&l4`CeJInQfv!IXOgz_G)#*|8OQv_)obEL%tX46lXISeTd^4a|3>H+8B^z3Les
zmP`?@OLauiJ(QFxQ36a}plmiQGw4g?hS6B60t4?=MYhqSNPO2l2C~3utCfXE7X*W_
zAi_X_>)w(nES<!<8D0vLrZ=>Mj*`;+;s)X>(<YnP>4f28R09Rm)cN-=z5*_dWlYg@
z+7i=-Iip#VPT{FG&{bFRF-{4lQU*FLcY<aG6Wkdn%77=3ZQm7N2p~?ThSIYG5OF3|
zQqF|iW>Iwxa$VMvaMN_t16EuRO6H-`Pp$|~?E3?fE=(NuqC2WUR2rX9Xe=G}haH&F
zUDps6M&~!EIvtXe6+q_Rjj9K|9EQj2Nm)`BwTBT!XdtnH3yUbIss>F6z(Q*pA$97`
zZL4so=*=8d)5IiOUsrYjL_3Uf6u2D~<$3@<J-%9~i`u+Q@5oeYDtZ=*jXJU+7$jTd
zVl+VJIWq&Un5G8byU`HRKrKzotRbzYb0VzXIZuz8F%GuTn9=o5>q=(R1Y73>Wz}?C
z3_JF>UzrF?NqzR;`C4u0%d&B5bY&*dvL-yUkgDlU<5^?c_UJ2OTA+eB0RTQ-%eBX;
zDXl9NNg8~RZy3Pgv$BQiDn^x{SfxCev4F0?sY__>o~7CzIZV;2+7*s>*LSgxTX8k<
zFRwt0;P=HYPotD!{MvL(_QS)FM*T{&Si>J<65JU6-9U6+3s+(e!_rgP0h$`k?C0*c
zw@>ST_ot>STvu}~HwhRq>&s$8Pj?Om7`r!&@fgkF=P~UMG2j$lI#ZnMER#BJzhcSh
z5Vi1Aue4ovQuX{4eWWtGFlub$bqUXnKkCBLYHgHE#vB8Vgtzk?IRimtI~sbum;3s&
z@-JsGAg?|+9X1o18Ds3(1-@ULAxizTgjZumE{&b^dTi-Te1GfBI@HOVujZaCdk9X+
z3ljSA{&OzC0UO{m7Z`@pIMw0vN56a;8R@~L)j)Pfs2W{0t0Ev=+EWf+D75O&HOQT*
zoA~?O@a6MS_l1BscY(B!Lx&K#?iq(WgRQiBx^6nK*lpXCF_i26@7S{t^xCiQRpG;%
z`*VE%;j?Jl;osX{?ziugRbxlbcb8qjd(W$@i>>@|e~aMmr*o$d&qwRmp5@&cVMdIH
zW@0s~ZQt8l55GV@VLRo@iOFy((D3Z17MFm`-PG*m9bM=ms=z&^1rShf+o}fT)swAL
zTwcKXL$b$<uVlao$)oyLCJw7v3R+4oVCoMSr(BrI+BU4I1`OuhNCTK&Nvq+F)7Bsg
z!M0{Thmo1Ixabrtktfgwa1P<%7JklPw=mxe+Bu{nK?e|x-z82!U~E+F)iCrCv0d0Z
z*^5%XC-0I5KF`vs+NE-nQboI|X7M!rdE>);gj6sQlOkacZiKh2O>&K*!Z_{;Wu+sW
zzXxNPFXU71yde4sq%HnqEjvck97GKoY48hM(yqi^*_c5g7CjXy8+AQ>MZ_@w9z<T;
zr+g&w{e$~h{l+$lAVX5VII#hWNr=XszA^{dBA7@d*x?Bi!dqtCibp$)#Vc@j8MLqR
zUzrZjPX<kXlxrWn$x!}3Lx@eckXp^`*iR?0hW>#+ErtU_R$n3G9|H=pATC{q$9QcZ
z&DGu>hBMBv$Cna8*sXncM(6;gdfcVF6t?SM_8yGml1z{mnzWDddx>%-TylP(qPwn~
z9<xrxR04+8Om1Tnp!25Z=N6#1jRm|=ONbvK`uE?uT)Y=If3y@X+Cp4(f1Yr(IZp$Q
zs<z_v*ZVXQ=u7)nS%oyEE!`Pcl58qMa#axv?NVNnfJL^HBw3KST-<#v#3Vw*8Nc|W
zKRU^BYhdKODqEpx!R+JxJuYhiLO9)lp`(|98@^-|xQW1JtAXY(;5)wnA$Q&Yp@kPn
zt24y=4w~GadC-$h68w7lMYefj<Wne<ub`)Jzj{eZ7N@K;S$TJlgugx4uOW4Pz9fB!
z>s>Ut!&F!th9f$kX@YVS0<tI&KsP7gpu6*v<avl+EfDWI=FeM`mrtx)6}~$ldtT^>
zd(yG9i(r9WoMJ*95rMC&ze?E0YsHYy{{f9(k(O;J*h7vkfy(j+=UgS^;MYspdrcmu
ze~i;Jrsy0eNz7mb4E-Hj2XHBEt{^h5vGTIyad$kbdC42g-SWw$q54$heClyE+oH{R
z%Drbb7q%eN5--LH$R2HVLatem&d=kJ*}H;W=5&YrT2-L++WJ#JIaa$*lZ_U$MKkd$
zjv38l$|G)`-wkc!L=%us2LaDWpwm`m{+$4sQ-|tnSjAv(Q6<8!;`Ys9*}*hpNt!8C
z?Mp=wl4Y31kqu3rXg}5Un;FTo_gj@P`ddCH%6kam{lbnV$rgkzB!!*1bLiPEWbbA3
z=7jL%D*Kog3@86~cF4HK@GFde`1HWISVu&(4M708GE*)9VwdhkaO$GAVH&=#U3s6t
zU94a9`dvp)$=P47Cz<{r?Z&TZY11_qHL;@vFv5iRIr?o@T5<x88fU~keoe5+#S&pU
zP3)zH+C-7WH_fVy0PTb;AL9>7NOdTu%)0ua)!;hr_^plC9SZ(KIVF;r#^?DhY|aX(
zIEoUUUWp6L?J5xaE`2aNp{lzhd_3K+5)wWqBCd7}<rp`AC=boDg=}hsOrn$CBL)0R
zFVPR%o&kTdmpbR)WWGL5cxX%l;wB?xe&>a5M(p=U-C!&ByTv<tvwCcXXXyBVz_jEI
zf$fF$_U884dhumG3NWVYKe<?rwL0#$jj5+Rm%R=^$RrN>FQmiD`5%%hW{&^OJ$a=o
z<#5P>@xvEd+HViAWjpL)xWcX3`vi8H{oOQ_PjNMY7$fFBIyp%qX;oBCtBOYdP=r)-
z8$Z+$><dHD@z~|v{yMn6d%AvrbvjAq{M+F3wG#9xJV|NUX(e|8luCZ8A8_^kC1%s&
z%s(-rjXHjGxU1hc{Y44FNIGx~do9Qv&Um2dOJg$j*vyo={)Wo}$@rrL=UG6tCprkc
z{`c@c%bP2<O0)Jl`>OWO+rRhY%3cvn3LBSasiNojy1##CM34!~Oz(7!J8bUjTju5J
zx;*QAKxQ)ad+7A~<wno##9%#DIEK2l6)+_mrH=h04DTT!&XF58Xyqtk>mPd}%lHj#
zRCaiRiCB#mr}oQ8c>Xap_V@f+Uh+Pb$(v-+>~-WRs2Xz^ctvGV-Z)$AvpVof$jEdS
z)7!i0^B>xq$}*^i2?BO%qIRYrtR%g@UkIr0K7d$pdGl!Tmw8!|1|nH|s}L`J_h^4^
zEl9Wu&poT#3mUNMoSHuC7KSp`r8vT<tIL@sS8$_Cv=}9bKIb!~!V<^tT&DeNK`s0Y
zsxZC@KzA>7(fE|h&4THmW!C6UZ}|JJ=3M%8`c#e_fz^l)Q08|P?ltKLS4|=f0n;E2
z4`iSE#Wz@UqPl=_YK`$p1{OcXR+}j@>p*ccLlQz93sRPh#fBs#-ClF2V_Tdf3Rn%`
zc%hSG0UH`3y|UBQU##l@`drgJnY;!sD}24O$dbT0Hfo9BT{GEu8atE5QY_vEL9J{-
zdcnUf2u>Q!!UI8zoU%Ae^prkXSu8%PlFO-^F61<=6}I#2?eUjr1qfXaL{Fa87_}p;
z-HenYommQ1dkP~Y`dHgvTJGd^y4NhXK6Q0{DEe0(#fmTez8R{vwcvV6^sR*%_uTpy
z7$+l(6$=d%0N7EDosN`keV%#cPD70jzq-)Y=!V#XBZQfjXx<k)Rm^lRB5B)(QaT0Y
z96OYa{4Jhgr|{c-%WxA?%U-xuWs#@UsBs7gwIEG2E`AJ{p_sDtGL9O)j5e2MR!2{y
zECFZgrUhlJUsHOhQk=jX_0f@_gYcxmEemf7Li7;xI-}GJSV@kSXOq3dYhep_H<E;G
zlfQaxMLq*9^W?c>g)7rd{oarq+6(+{=#+M@PZn0x;aTlT5KFhhT1aa#VeZ*lFh~}=
zDCVYB<rq|x0BxNqV3v`tH0YNhC00dnS<1UJ&1pn`K&*-xOud2=3;40h_w${qp*_TQ
z4%6^HilVN{B0SQ?S3frGY>+02*V{D!&A8X)go?Jhkvln^$-2Yj4;or<Klg)O?z(wM
zL@A{hs{M?wjz!x#l8IxqeVwn4M%y~SDg719&~&E5#ngpuTe8K7Bz*Tqv~fsXTajl;
zbuIh5P4~vnC#Y+~`2y%&39QRRXvgUfd3zmax_a{IF_7|;Gk)Euzj%ggrKhc$8*MJO
z3s955c>WQ@JHDx1STJ1r+<aVDqG)Ru_^3Z?9xoc6xA_?vDX%w+`k4y~s99PF3S=uk
z9l(D$vlPJXzTfj#m@-$r+@Pi5UVSN<%BOC`Or53DlAwk(d0xo85TW)oUGsx3<{(Td
zmG^e!Om&dZwK|6A=(K^RpM7zJ++b%so9fD^Z9q@0r(O`F_9`)Kxd%KGElM?ueC8#c
z$x9sns_Y%hk?$m7p{Kr?g3shlwAv2`bk7`pyZ+y|y*nKqeK=5!!4b1<1yoaUto_^w
z)&r{9d2IgK*)Y-tm5rBSb`c$=Y*y-e!PaqH1InaRsq26?-WCG-Ygg+do0?x#Z&>mx
zscfVj%r8|^ORLEA3!o@J;_tww8(Q*5Kb%(IWAy|$Y63B7jSl4MjM7`1=r9qm3md<Z
z9h!T&ercU=31m(=&QWAOk}2?0Iw7t5k#Iv?#~l`hEcDsku!x=#ZQ)4P_oOhZ**hd>
z%2%dLy$dsqqZ}T!h3Yyi!kVL0`mEJ`nbF0kLtfr@vN?*+zdn~!2X$fbWBc$wo{)Mo
zgEO-kAdh=U1rq?g)K1FTbWR|YmZI5rbt@t;ac93fsq#`a!b$&>>(}L3AQ(~-oP$Pz
z3u{-OD-r_sgA(x0tNRv#`{MBONfn0S<8_o>Nag5L7h@*-%M~Cda{Ao)Ut)F1p?&qd
zyxu?CiJ8rd17ZEDPh2Gsc`d=6tBY9h9CTbJj;)q1Z;%Ws@Tw1D9pgvEGVjl&${2{!
zj*vsar2~JX+lh~LQg<TmI*&DU@C;4BoMZwKe!qGqMx;8?6=co~B;-Rm86It&|C)sv
zp;?S_)Z`*W>+0kY{JecIx(nWdd*I%kggYmj=DkO!+2`Fg2BEaSYR4A9J4QaJ7Slg!
zMCRZBpv|HY%#d7>Xx+c)CASkWXffk^qyjWo=WQdgj43zyi%C+GL_3xZB&$g0z-*7k
z`h+slIvtS5!;vE&DhCq6mQJ8WpcQ*c!WiK<RTGpMra?HF`2ywefA)uHn#G;n&8Lng
zKsHVgi5f91|3l2u*}#nQQq1n=^FP|_aft%{kQ-i+F+!6^7H?`*eR9!Cen|g>e3{j?
z8fCE?dotIrJ2E^Gj*d3^U`vzTE8r0-p__Nz|JxteA<gWV9jyjpB)u}y_0x5xH2IrI
z@gRCC{xjo)ZmBfjugP+lx_kkbMQU}#hJSjBBDoKL>>hC%e!{wy)cjIAANon!zj7kC
zK`(;=G`otW4KuV=e&D!wgqM3AU@t%tmEsQbc++xHLHVDQxwdPJXv&xo5Es!o7k?sF
zHoM9Vf~=i2kmD2+9s9&#T8ujA8!%X}PAN}Cn%`1kZ%--X^#?30E&`~6W-7wnYrBYT
z;jz<yG!}nAnM3E%c*-X|`FTFqQLOuaW0?QYJaKaUj~M3vtF@Q4*Z!%miXK_j#xi85
z#l9sL0?9==Of~g$4tzb#&)=`kHWiduY%&^(P_ofBy))_6q8ndCr1W~fzCL@pK8Ptw
zL6;91S-e<9y(J|32R6<aC!hsngqEWp!1r~aB+d5ewT~!Oi55Wj@A|cJ77+$70VqY#
z(Z2yK;Xtz|m07y{CzA8>Gqfl8AZbMl`}EFw;2?13pYPYVy!k1iK}+^X+>iE20g!Cy
z3Th7!VpO&+t4pd>84#ZRgdfEH_C@GN3@BvcsOuOjY)a+;!B0;Ek%(koBv<wnDhA<C
z(FlqSlCa8Mi6R~H;f*CYR?d^b4)|m$A2YW_IBp=;BrZK`6qC;AaTeOU_vr3YS)v-K
zfNqu9*BqPzA#g{@JC}LNT}@`>TTtGkALdWR*FW|k>}M92LsvI&8d}bQO(ijK`hCH=
z`h+d}Y7=+of)$3WfX%37uYJVbR+fLftUQ-tG(9Lqfq4Ikq{f%5c(Byrn#aIMzr~g-
z3_&cr6eCI1!oP^V#wSRVO<Sf$t8SWBjaI(lO<>1ko)yaTT^G}tihclxo<wrUA1;U3
zY{vFcB8__-C2q-Eq0Hp}m8vuuOROv>&e#L(PDrrpwwq^g^UZ05V@1!F^Zj(7=c!N-
z*7yy!0h{dkekNF6?-6iSX+BD_?kN8jkzStf&*kRuo&Hsp(IadXyK)8LWzyq8D%G$J
zXZ(@uc4)Rl778eiqaEG_giYe=duoQ!D^Uw-9kDbqo_!f^7VpLqm_%=41FxZ~{v;yg
zRZ3UiDPv2lVA24rlI3xri&>8xep^BC;2|X2w!ig`*j^>B3u`d>B*AAk;*X3i6<oDx
zssiMJP2Z3kIUTH%xFdmEi({Mqu*W_tQ^|atLf<^sCn!Vz4O~+K`>jXF{YjgU@}zeJ
z7b8Kv`*Z5l4~p65(8cC<jKS7V;TdSZ-M*Z}hJ6XIP%J^qJeZ61&shtb4~8YS4Wm1@
z#Q!e3WMwD+kzw#JBaw;%C5yFb=K=_kJs!$dYK<W=G;JPLb&?N((z#mO$^&g@l(x+q
z=_K|~f-p9fmlRfYDv0;So>SB5mS_rF&LpZ&vTIx$KGM&F*#jw!94F)^?%@fyGJMNQ
zSmGF|)$veEo#O3#V4j#&vXn{_kI2gdKvB`0&@`+=pxd$9NN8VunUUQv7fxNE?LOc(
zwPyq<_9ex01<v2R?HdwXv#GQ_2yi5L)%T;s$9Pui%CKbUijj~ljV;uodn3d`bEdPP
z51<`J%N4nJjZ|hbbVk<Aze&E4YRh)YPyy`w<YkASs_XqdxkVJfY)Ya`%`iix#>zb<
z%np$v(bV38b-3|j%Ti?dXy-#Z?$}2?##Jdv@^@!#vTUP$Q3kZ~^lzM;$ut7=c+w(<
zd65S&ZB*2F<tsSrymPv?`lD@dbX*t(%qAV+cnz0$c=7UtA@QsfA)Z`3tRdIWTj4n<
z>I7amXH+}M23U-NwY|k1sRoC5<-7B^_c6Ej0#=c28TSx9cabK&4Tc`}2wth}Oh(ti
zs4%~SRo0c;1_C?gQzr=9xY1ceTQ!`?txm-cjDeW~`*2VTO}cX$9H|!$p|<1XE9W=;
zel_yM3?>VdIVXL>lo>nGx!F}b=fr3s=tLd1CfTS_&x#hB$uSyipqPv`zja7LaN%-D
zG~27TwCqZ+PH1^hc>DPhb6K^?;wC4pbZd+L)RGGZNREgXy;0{J;aZk|o{3glJ+KQw
zRO9uV-;AOT{UKR`FD97zGMmD}q|F<Cov+&SwLYJ?QDn}38hoP(CzA-_fWa%?!v>GF
z&Kx)Hv@0n3x_ANets;@<(<#U|kfqk3^?|^@*vV*|(dAofW!E1mj75GF)Bm*>#9`gD
z&AQ@(DdOnPIALwnTZMl%)<Iu)6ba-&g#)OvZRhgMl#<-_$6`<m^^ZAdUCPrfRfXfF
zF%c#*(9%Pn`>{!w?RfmdD#G$fw@>rR?@qwn13UH!mOvvLGKzbwEovfClm{coy(ePG
z-468+S>x-r7Q3NqQx?A(S=rmES`UojEFGPcZln|e<(!bFreABGnmA`U;ezzOXkUrO
z>d)5Ij_~=Vta2CNy9adW1#K#7B~O(#ewMZMwThPtKQqIGCpyl?>i^x)6%L&SZEJ3r
zXJq9LK?s&=yZY**C7KIY4m07lYLFx7cnVW0aiW6Z_3x$$O2T0ezal*)CsBs5ZN|!u
z!DCA_bKhL!&iBx<O7AFJ&G6FI6FmT)3C_dQRL8IbCbl;SsI(pgVy_HogF0Wpk7iAz
ziX#NRmB@jmtz_mUKnq>)zA-3)HI6CZwIM5l3PEhem4`q#o8;ebZT{r81dny@7j4@L
zD4xjP^+TC(F}#&qnPoYeEBN)9rQdCcy&y+XZgzcB6$$t_8cGlDnQr_*;MEn}jpzv8
z*UNGU#SRr7B>C$6;3p7wj(xi-R1WLwU0Mpv^#C%p3E|X(P=j^~?tZ9|!!9?5U4Ks?
zlWW1S2)Ts4&GxgqhblWHUOQA>(KwTBODV5CDj`Hp)LhQY6DnjKqAprE(xoiqGWW(V
zm=l)X7rhMJFb>$#WByOR{@7!-Rs9i04#^hpJOkJbBQ;wSYxOsCQGSuy;H}2<LSW;@
zr|2tdf$g=w0nZK^A(#+6Wpy8}(>v=lZs<&=y>9djx@tPLns#xWB6bluVW@6la@b9e
zNk{G#3)`G+hQbU?y{YY2HG9*qc4t$(^+aR3$AO!M-T!X9Jw={v?3=PjE|M-yBlwdT
zKj^s*lD;eL&$GJDyM*axyaZ^!G_^-BP4Ml@sypAzo2xJx>rSm}mX|k`17XCFp9lCE
zDaKEdd0~%!N9kYM3abzuJbdx~N)pWn)jmg>ZqlVM$|nq`IV`w3@O@cfP{I^-j&Pu&
z*^6eUbBp#SUJf-}Y+mDtE{l7bF5(r#dUw%;_HakGkCt-Tpc)M+;eQxEX!J}qb_ns^
zIBawnD~Z;C4HJ&=ZEF5%YA95hiEy#tq-H1wWNjoW#h<qx9rBcG@x0=wFzg<0&!jhN
z&Jbz1XwJa%oe*xvi#qWeR-DUtpQ>uh+WAS^8KwI)STPJ!5?FoFp==W|XG=!T3jRVR
zE9G;xqhC^V4fN};2KE1nY3VM4=>+a|>9*;gn7*pR|FbLEO&bgoz;_$mS?p=pZ6o1T
zyxofHC-k#&`dJzL``Xa+*q^wnWA*hKbzjvtA8qM&S4Fh+5Q_pS`GRlh4o(5d5u7vj
ziMFE}zPMVxF9a(n?>4O;_j3<%IQMb$jaB6z5p$6P@DGI^yz58cJ{i>AsJVzscKiv=
zz9t(CaCG{-DfKnkrL_rZL3N7ttLNnO2&Dy5W#2RL9w>@cK^0*f8me1Cyug!MeTaDC
z1EhizU~01NQiE(jw|H8@im)<ZZ=X0xZ&EwEwM)W#TC_{Dmfm1+6QNPL|9v039>^;a
zph0VK<rER-eT`I{Q{I9B$4B*NEHq4mjtC8wpMVStPCrj71Y$2gfA2L8`X<G=i~|=6
zPJLrH4)WKbs{p!*1Aj_AFbR?<+#3rBzJSUKArY2w=SONWL5u{Y{}abF$Nbj=1zcUr
z@fXr5n6E6HWfTDx2;S{bzXljnz^}ku<bkH5FP!V$p7F1r5BL7<^jVjFcl}fO^%C%=
z?8x_@c$w=zav#`O{x{*>6|S7iMcduA!JnN=t2oF@hQ2a_={0X>3vU}hIR`hn-%aWa
zTlwcF9aJD0kTROsacKmBsT5j+!P-rj*-LB6_AbGW0D(X+=}$y&EuAZ&0ndn_<aafh
zAX`_nG<3*M_Cx2#YD}u(>l1)6870XeU69v*wcJ8AUV|z{|NPNbsyEhVFK4wNzI)(g
z`%I4<E?(ouK^Zu@hE}ig{x~$cOK3`>2jeg}daEAL?aTN1Vq^%WXnDYU6)R-YMHt`@
zC78G{sJZKf7Vi5p+275G_^+Z6p@YmJDxl@k|7}A}p--mhF<Fc{ZoUhN9ILSPBGw%Q
z+YyD_FI|~+@o}b}wO>=V0ER-amzNz@iIwrxCNfKmd{<#s9HKF0dY9-EU`!4uvl}L&
zn#=q5NQbWQyA(M4$Ej1(7({0|T^)oD$-h*6hr*pRS6=?8whPkg{0_#LeFjktHYR?Z
zMiN)uO0+mPjb;nDQ1q;}AGyc@Vv2JgtE{>HOHhg-G+jx}Dg!RfIM&WXLtDJK78l_C
zIg(PjVp>Z{3w1@Lu%h$~-pgRQq6o7aBxv%9U<<8Xx*Ocrq@S2@L4DS+zdN7rO<vSO
zhOMYryi>3R70pu7UJ(xLCxq>vP{5J{buX?Mgo+-2rkk@wavPWU1jL;)K%{2uVD4ry
z2i0WXP<PNsFr5&Us8zfqW3>i-70BXS9VZBKnx99DDOI&->yYkN>)f@XNL(@6&lgtP
z6L#FLxl=SjAK*;`Z^(~Q6Amnfmlz)l0zPmz(8lY15SFY6tZKbtLKOqzr8a_4T8zjT
zNl{Il*%D>GU5y~--dqWZ4lRweV!mAXqOEpq0bhfuzP_o6i|0wLnLjpRD`ZNph{m^+
zWd`=_!?Ej4E)dHOn6!30s4P4Ys}o=!D?bFF1|1d<C+#+y-&@(0C6B?LG?pcsSCYwz
zDB01ZJPyc%u%5V&e+x<AL{n<eC5w#JgGZ<pwHSvE@9A`A)qm$e3XPZN;hwl@@TKK6
zHYLeV`8m+VjqchUX64RwdiZIPkOps|Ew9?zvY)@0a}ct%LdrkidCr?nV!!no^9A)W
zd#s^NmBDRcO-0kYF=aEl|HgFv{N2mz6KyD8W3R+{;{RsP6yFv()tbT3gwBtiH3O2H
z@Kf==jF*yD=HTSYBazSSJa0qvj&{Uvz$hx|vTl|Z+Siw@_^=_=#DDh?c!*~e=SoJ|
z*7&<rG1;psMhj>Uy%w!`*~g%RfedT>$dZ9PRZXLKq-d$A>E(ML&xG&Dj{>IGppU<Z
zb+(T}&JUq}VswX)2q?0VLsJ~SoB9`!_D&poD)Zj9Qa0$S&9C>D^bqF3i|~-RLY5Ol
zbTq$p{4tlowuc-(AFZ})nCMJ2{AP^SHIYfcgcW-Tc1gM~hEI=$BWrJ_!p-F^-vN;O
zn^>PhF#kK+)<FguSWWKkUfA<;`pr7jA_Hi1HyjL0+168jzW_h22|++_d}4wo?k16{
zf(EvWcpx446BIV`wEf-XT`Fz+C#%CAUDOk>o>vbF7?VaiD9P~WnbH?GYIw5_8Wfq%
z_1hT@V*wVzNhi3GL4I90f?{lzH9zz>aL()<i<#IXM&AA;)`u~kY!)kF#2?3^3*9(k
zuWhle-!_Bcb(hCJak*VCo86B!i++HWIiB;+#IxKym(#fQ9X0txH<dywF0bwh=<J+C
zge*M{Gw=-5^NipwtdzD}5$e~@obcl(X1~4(bY_hT?RV>brdj+afAnmZqMRy&QT(1f
zTLmu;!*UCgVSC?P^U<p15{VpXj=IMMN-ylmj%weq@9=qe*)+59%yIVm7=0@C!<@uf
z$`zhjNh2iTjA5<6r|fxN?_ugHeZZ)jjNJ`@{`_K$;l~?qQnGcSST(QFhTWrh&wXN@
z-HGvFXTP4>(+lpDMK(L~M)JLiJ(g76cRsRqXFAj&UjIB_&h~ul)wY5N5P0PvZpGNl
zmyag%kvHZEb1*PK<D~koDMV;|Gma~6R{&}F_hKZMVCFNFvP1*Gncygc_8ixx2q|lw
zSq<nV&0q+KDr-#{(l($RotwMO(XhM~jBBxeYs<;4iqS!405*jk9mg++GBxIfP0S3c
ztPP}N2~EzW7skih@d-pO7{9OkPndmC^cvk8o_%pN7vzC&L7J2o%V;XL)i2H5bi&CU
zTePo%cMjM|mo5}bdldMv(PwdD$~r$tOY>i$o!cC=<u-KS6Y>%YzmV7>u=o}fWg`C&
z%0bLFY|GakIg2hkX27v?Pdne`{gKTY&_KeZXU}vaU*Q#JWuA<@aAV%ZdT&_Vx%fo0
z@XzV+n<*4#BBm8tGua*60L6Q{P^lx)KD8McbI%TH=8)XPlW4gX6$h!{5_3^aSbIsZ
z`w-2`cOqQz<G1~blpW<cI$)&4z(1xnKZS^uNI%;gkH22Zup(CNgWsMy6gmwWM*W8q
z0WF$-RAIEU==kV*KvPk8$-!f4L=@W8Mp9-+5FSvyd^DZHD<UNOO;aH=<fSkuL^~}i
zD)j0*!P(g-2L<IG-a!(oM@&Q}SI2*zOB}ceX_s|?B|6~8*|R{`pAJDU`^_C;`ABJS
z+f_KGn^P1q7>L6QJ7c>rr;}R@3e6VxH$vAA(hq8IU0gP$ZL5OoV$h&OP?f&6y%OOd
z4;Phvc=kn9c9q6D(tHmk{NSshf_d{H)sZ;m$#{5@(W|zQuOl|h26&PaF@6H1C_qOt
zQRLPx-Xug*5oXj4q+CSgefNZB*MDCV7QJeyl66!uLcXF?e@y=T63z|skR$pc<_={D
zxl<0>X_ovZr!^Hxm;gijC%JI2!6}l@l!H9A@3?bo8@RfF-g9^hq}r)}*UPIAV13dU
zj3#|S+7SHnr+?$zd@kb-IA)hD|MZR1f){Qdx%s?TKCs=brNpcBQQh&JG4;V0mlj6&
z`muE`YT2lZZ(a}<X(ev!A>r%fwZdvUVh%HPsQC+LC)}Fi>3Wv(%%!LyA$E-D3hQ<P
zbX^_={S)3ej_-SaYj;3yn*t^sUF6|KxBU$%GycUHeoXJ(5?UtPn{vZF=~fqR-qhHY
zr!?z~+|`!3>{OR`k1p?b{#)g#k*k=4mHXGv?K?VcG?Jvk`zczVf8E8+XZ9B$ZqRs?
z(8YZzO1A0Z^joy>@m&W`w~xsL0Bza#-QXMq$>xZzGVgt{ym!uXb9AuBIk%6<_MbqG
z`#;EA4vzn=!m9VDZ-T7z>iQ=R!<K9%lH^GQ4%B^9pQs)GHT5!&Ba_=M@RIuNL$SUp
zS#|arfLGOBtzAUbH(qpj_92z$-{tew^Kl>K+B9g@aQ^sq1@;+nsF3(McbWo@1hMz{
z_Hjn2piDEgN^3wyPGXQA!27*2n*bFr5jc*?C#W4<?o`8*!CbLOcl0y!g!cj;C9Q2{
z88}!Tfc!n%@Be)zZ@%;sgYgMxaJBt$vK??89!G_?@tQKP9vHgl(7zid2-U*0lAnq~
zaQZmfFTZ>3Z(o?+nsp<|0V&JzGr&V%Pes=5PGP2++n`==*^l9@^h)7LjAP_^a7ej6
zaQIHR?>&rD@tSEzI}w144yV3rj7}iY17n658YSFSm`<q|geOyqevNWS3T<~}S2drI
zA_j#kMO`|F7X7;4JHn$zgIIVj;le~wrKLw!MSh6-XI(Jnvh#%3k73po8QdfSjG-wU
zPL0GXiN7Xd&==Dp@9F<$Xm;BZ&!YNgNgGHoYq00vb7C1&pNmQ?now8XJ?sP5EQ_^2
z2DEyp_e%f2_gg5y);|wF;ZazcQ;g7v<y>yknm^v@G_doRXb!&$yx5O!%wS@IZ!z?y
z4Jt!Tz}cw6R2bxf1Tm)^-4Pm08)-CLzB{|H!PPBU*v)qzQoXE{Y;o8`9ZJOfWlPDZ
zTopCF>D%Nmt%|64uT(hsf@~3jtuJbJLeSx}<odSet17XMUt4JWOnX3nsWiyBG|@@u
zUAV{QU5O0!0lhqV#mSSci+XO)NWa$ANerY{CN-2pr^8+_)5fuGd#73%ude<=(f6=<
z4sUC2N)Q)!pcrCAza&pHxbOYMQO_UMYhD#TwFEaCz2siGOn)Sa@o}$CTApub%BGH<
z@DUdv`btKc{Wlj%d;y{CgCyC;ns_2z5UmiR#MC-N;yD9AjSk!>Kl9Bv!E}nt*fh-p
zp`=^G=<k^5xvi(^pa??IvD8vxnW3UQ2kgG8SMSQNSj7a3#s_%-cN>#=g2#nHe7p|H
zN95qVJ-`73&6p+Iq|9-RS&amFb_IaM=<zmP;syZOEOQ!_ClV@}8yv-?QfvCP%!q6{
zb{M<n<_mP_?A7yiYRX(l6QZD3M+a{BkP(B=YmnJ_>iAcWW%w}qk=z^h5tIi$Lis^s
zgYiwRwm%6QdVr22W?`AbhDNdDMx(tz8^(vK`H0^dKLyr#JEJ`~8^f*7wpbzE;BqFp
zf3YHZOy;~3Az3lao>RCxKse~0F5@#tKxNxESXh0)vhYzeI~ml%N1{akVxXcIhmeLk
zvNc*qKr2n_6Acei(2CF@T>uf|$i|}%N(b#2NJm;fGe<G;l8GROu1)DTRi&6~7(Z8e
z^S?({ak65nZC@%sL2Oi2F?I!5g%%<-9yE0_c;r~(u!h{ySOcJ!?2$QpLEG(eg%EIC
zpYT;)!ull1@YcofrV;S0&{m7Vu}0z-ye?swP}U(k;$UQHxP}pPt8tf8>N{yz4Wg;k
zuwlXwdC{6r=|`gpwU`I<3aJBA(f8|}_U=uU)q(B6if+`o6PWdz^t)pSiNPV4L1Kxi
z<FZ0%K-(A3rN^m4`cCG!G}uu>TzC?(@eCojs+QE+&U&S6XD(X-pu_rPqQtiFjR{VO
z8iT=ZFlsXpP00@&%gu~X7##%L^VdS;@nM8>ZAif8s~QVUjLO^Z_BAkQ=Ljz*IId9j
zx^A0Mm|lDHQ}d<M!jC?Hqv)=p<$2#$z6URGc58_m+#414R6IJQHbpHnV`BFT&U=Fc
ziQ#T<mbBU$$g#81=Or1LeY^pb(v@d4rfgPK4J$gFS&H>7MHAe}KP;uXwAw^&x(2oA
z5Cvh~CH3RJ93l0J8h;D6xNV6Us%(_6(`!<(@fljxB*yc$;|^p>N#nT%F?9%RB0FPI
z$lU8dm=?ZaNQg!wV#gqm>*gIz&y+6z=`e8i#p#DvoiV^$SS7ZZNGx0{4MBb~&bj7J
zsGyFi*!?tjFxQ{xuAA&>L-icmm^w%Hk`$q%G9rSZh*(EF%I(P^!%|d@689GJL;`s9
zNXE1l$T9>e`0=c@r<3Z?)w?&rWHaTLJ4)Ob4^sCjAt-WjGs(ED_UwkM+IrGevZx1N
z0><mfBah2(6+xd;1^KtOFsWuZpzY6Kts~!R2Z<kSltC7Sw41)uQMv6l=3o6wVI`Rj
zLEZj<|1brr3e(wLfw#QShlnn~!m~)6k&bDwPzzVy>IP>ZT)uN<CZney!Jc^K5Ug6=
z@^j9>-RRJ8%)clFMU{JV?HQ6{McjpQ^T16wbl9>fmB!fTT|)_tTWN1a$#bs8*t{6I
zVP-h)Fz*mEWgP5q%<+sEo!n8+G3bZeCyA|5gmSSs@EHy)M<14phINt3fi~ywo|G<E
zl8U2<R(l{@L&k5tR2!8_bME|cDZBB&$v)bUh`>)a8xMfAL~w=2P`UJ9U`|k4Cq>&;
zKe{P8rq-zILpfbtE~o4oTyoGP{ah|D;oMgtIq@E0&Mx6YvLW`_fg#E%2hMkdbYYmB
z6Bn;fgMo@4bEZ?53%n(~<$5RIe-mGA<~Y|RV({!UXcDWASJ^`A5D9HrVp)WPd9MYi
zmCU8A{iPryI^rS=%_XCfZ^ri165MmrT|Y*0e_u#l0?daWbjv$Cr0PbHkXCFSyH>iG
zTdH<Sz2nYF5+(5TsU;DNxXM+XEE<K$H)wV8pG5=3e{3iI;gJ(aB~jpPWEHC`YoWgP
zJvjBoRIpUf@#f}-&a!Meh$_W@FKc)-BL7b3Wj_m)UIO*?%0}8%yK2(O75*zJ$?Fc$
ztIDX+tx;FG33l5G=By^Vr!Dkb!j|)m8bN2=IA&eEIA^~@q}8Qu=T8yF-@;tCE%iE+
z<8=h3_S`OZ20AhB3a-^@D^UZdWgJc@aJX9U(t4dUr0Qm?<4GgtjP1qx$zC{1ZeezO
zSnv7Cu4~K@hiZu+rbmslcpl!NgncmX-1JL#Qz$#GMg%j^_{ik!-yarcHzCGYq(Gwr
z@s*0Fd)ueBk!cDdq46+~#}b<J8A-20DA;iFn=za?OP-Szb|IdYfz-Wm&)95XVzZs#
z^Kzb!w@(6+qyF0e3Lemrlr?o%@3vKQ@>(TjHLpPg@a~Gxk){mdZfj??4_{j8wGi@J
zq0%50=u@j_v^C@1<Ffg!gShZo*W9Lp$>jfBahV0hS7CP}Xi@1I`Tf%oFuCiGkYIbc
z1K1OEID&YEnP0^{VCciDsn{NLge8;lE$eJ?*#Qq=sOEF&oE%o+wXN~k2Mz40vRfQ9
zfa92_*WGS#*$59zHN~|IQ!ul+NYE#~7h_a63VXg;RA>cq;0*KM3uxCsXb6to{GFLG
zDZw?^%%7x+ZadXVqfxg@7^GUhh73pd&eh`KN&Z%&9*7*3xKEMgW_9vp6XzQePI7wm
zjv<=G%mL>Jo&U+}K`a!oJC5w!c^ZAW>}=Gt{ZfpMr&rmw{=hj3f56D`3ha79PJ)e(
z3oVvJ&<WjGrGLxrefS3q|IzL9-=OtB*m`!>|II#er6ZGY*oNFU`_q!fDj7}@GlJQK
zRZ#8$5VGeuJ|xQ+UgzfyT>VGrct@S&-Lf>jJ{*qtmG}!I#(-Z6{q5H4?@+tPtb@mH
zk2D@!vosdgjsScxf#W+yj=eQI7V+<NW6uw(*AK`U;l-0ofw9iiN9M=fg?exA)Inv~
z!bJ%_ejGGHK%3)?T}gEBz|qc`MhP-qE8B{rLV`HzlKk)TfWEJ*c*>Q2FE#fhheTN`
zPRo$k4q9JS#Mg;Vl5?0y9&dr$u)5HG(fpKu26+VzxZ=8ER0$<)ps-$K2~a4ro(O1`
z0Knp+6<U&Kfx7gFQ=;P?SC~cFBh{DPd4d(fo~UJ^)TdVCFd@`rEE9b|7;3geiH~bs
zIn(<}tT6YWtN3kBAgpal&SI{U@%_P}zG@uT6gaZ<Bw8%VBb*@R{2Z0EIaQG{Q+cob
zYraEvbw+!4{*Cti0+%`7PoCD<jG$2AZw`APwpd(l-?!P@v+M43+XaCcuBRx2XhZ`6
z#Od6d6en(Z)<bPtwOK`0v6R)|#6z?2SLSH)sdGFFm?Gs5r_fsmJg(isph+q+O6qOf
z)Q3m|MKComK^zUmO72~CR$%%+M9#8Vzd(``BYk!q^DHyvcy5vzE{J+0rbjHDT*0Ht
z%0&f8VP05vG6vKUusc9C>x&W{YoQ?|d+|#%_igs^_+#E^9rd>}VWy)0y4Va3(AG(t
zjApW}vE}Wx1-SlAhQ9yJE=-IeL3zh|R)D`vA(by-IN3^1@PwoxX)fiwHj2)3T%tcv
zk`AZWPbNQxOhyla&RuzJdy0-VqS%h~nuq}rI8TK!wAy4ZRfknz8<||_D4#ykteS=g
zs2^~1v#3(h`lux<MR(zkQbsj~FR!Kcnn~YUv~#Zl7bW(=u@;>ldei<yz-w1A`HfLj
z+HGDjZ^aSkz@)lI*R0-#rt$yGAwWSWgimh!GYd-A>_(qyxbNnUKU4X+s+a}uaH=s}
zt%tuMkV-(rV|9V{Fh0@+rVA@}r8}c61stN2?DB(BI2>6~e<l+1gRuWJG`mP@(B!EV
zYWX$2$$L5v@9~<Kd^bk&LlVI$#A%6tEZ?M}4Wb`qk7co5nCvISOg^b`cDf+rVwT1g
zjKM<EOz>e&XIMPTlycHsYF9<exc~-kc{Ji&+}vNz+Edt~)(f8c#f`?V-;9XN5)!)C
z#uYIY-JfTLK6FUo1#J8#!g5NvJ|n?d*I}B9oY%;(*3lWp!k`y7<QTTWbL1Fo%GYSH
z)>Zr_z^f&>&p$-S`wtNk3w}tNseT=R47;S(XgF~@qp9e2MhH76X^zAL2o5ju7P$8R
zsnaMcZ8`p`ME6Q=YxV791{;nG3_q8KLTYJPVv=aKfx7)yEaKma^!G}30+whC4J5H;
zYevk^1ghXs_$-qPfb@6yfD4Dr)XXB9!cmF5WR}?cJ;b-`p{;YT0|NV9j^w9?%){L2
zzzOpg)L2O#uz_<c?rX|GP3E6ejuxAdPPlKHfBuj**LFBYc*u-JCM(lvUk6WU=qy33
z>}cLfMv7YJh}AeQd<w$wrn(0-JN;yTi({Z5(&`?5L=-H#)@gpKJ9GYLsHUsUfv3Cb
zC}f1b7W-U4F^5wNJiuv&7>-tp*Y8X8gFZ6OnVq5pZPh+57VnOb@!WOYE*6*_7l~Vm
z)vhbgvi8=|?PORR%zM+lWgY2;hh*iM^XSmw{p~sIU<WZ#pj)piL;c`r<|l^((h~tA
za9ik%hhF#kN6AfIK!Y2XBa4wX)fs%;{bJ0W_YOjD3tbn|K+>Mu;6TBpk4_LkbUmu!
zDp?J=9ean?lQPb+S8sPq*%eF?to*1y2iehX^)2>pZu<H{60ykcvL2l(l^d`g#$aBY
z<h>=-mg;rWL38EqB<w*4<N{%5P0WESVueX>g16;1U*_zL*+Yv&x3zv9e?g{{O#Aq>
za`H_E(_wCL5d%NA%i^*!Y-L3~j)zz|Qq_H#|NM424>9%EX<Iju%5nJu#OOunB{x0q
zmHT`&S&YXTlczU*Dfa>&blCWOco7P3SAVe>`Cqc;LyL_f6hryHVOhz?oJ$Ec!>^Yf
zW03db?xx^tIND0B^Ra0}<h3hj`p>+h#g+9Kq~%*o6aow-L!3QckTYQ0*5w9|)*3}%
z3$SJo#$Ka+WW_SAh6Pr8wNC1==J^D|>!w=3*Sc0>WujhH5W7}c!^EH!XWIeN*P)(z
zm-|?M3_RW-$2p%C8rEcF0JT*<U-zpK5`|A+345zs0bW2Z`PW}I+0YTh+18mzh&s)s
z1F{6wne>N2j*T&Y5Bco2k0Q245VAgNTUvX<j85+wWq-HGP_r61H$oyAw4L4WUI~w}
zkk|55F$DB6`BhqR?@C@j43(*fYGdK7pJyvT5v$74Jb`%y;e0aa4Mi~H9JIVMo*+`z
zmmUwDw#L}##S<tiR?XT?f?qp}X3OhX8+r%bTPa5+vlgz}>U3wIkxG~I%~Bg@yFB|4
zu4JXN8USIdI2jMz{3~<qH5DS!&;%e4HBm4OsZ1`_Pb4*IC?Uti4r0iqSrHt7tKIot
z1f_nGc_pRePOYD~v(w;E7zt!dGpQbAo$R$U?)wFdOfFDeRn(rNz4?#@>^dq&E48zd
zB}yV5H7a)N3ChBn^~bCz*3qG(KDn683wQyux2npv4i$vR-uFxO6!QLQ6W!3y%hvVE
ztSWRXbgD`TSKP<VfBkBGa19j^QzLC{ntocEER?i1C{+LB^lx&gEu(?nd#8<jtwf7F
zAXmD~Gg8i85eOVhuleDCf>#n4K)q&Y#g+!APrhWekef=BHb!Bjpu<_Eq#C|D^EJqp
zhSoxW@x<0|E3cx|!4k8m`>2=tu5BS9n7QUMa#05Jvh7Wn{M<v%cDL~+)ta8Y-CA1Y
zPo5sTb%`Qdv#jy-EO8e`n<Za>OHN}eZQOd?<4e!i;)v0lU`zI;d{iK&{_M0-t8`s6
z0$GPezurIS1Y<<4sQD1!lT)-;FWPkABG$rH+~ZhhoRb;gebfolpI=Shv)TzG9ZNP9
zuQu1ES^Pm$P7|uYj)0D2;8InDe{`&Z4*zF&tP%z|dc1M0aVRDx+<4Rl9sURwk~V7v
zUtXOJjS;Wbax|d78J+h{I)WC%(Q9-Oy>{vyuimo4oy&k2j3xD*)3I+5Sf%JB8<6!E
z-TQZSg|!G~bj3L!*Sf;G2vfQOhbVv&T{Ro&k*dJDWUJi2iejVEw<^fE*}F<9+<!y`
ze(%D%9Eapfx*TWlMuoeTI>H8~94EMaRgN93eX>kl^~iIouthwMeIx))Ze);9l%djR
z8IfTL`qC>(g0{o}*kpl?;J-j48`FPK)?8fNME~#gW5~62HFIVZw>5G#`(<Y0U}`2H
z0OR87Y-VH!<GFFAE8{@c3IAhLakVEPT8Z|yaW@>eZUjAS5DM=6zumgi|0E>&&xO@0
ztGaBSkq+nFJ#>nwlHXsYjJ5Rp^bU^{5bF&H{5@BX_r)6iE+0*<!u};B`v&{5w))%w
z4O(^hz8&4S|4ck1>NR|^sapXVqc6Y#n<td;z&e4njGq4Y5CU;_%dKtm952d^&kX&A
zu}BT2i2BKVc<>GQp(j6KybmFCRWiu<y`=`9i<=n{0fKO-O4^#3jN-J&@Veh`lb~OJ
z0?}tiV_5Vh>lypMec#@1F&W&krp_l{4^|y8pQjCC=s^qUOI=oGY%5El3K(X@3eObQ
z(=Aad(bvOj!fu?>A?(cNagMWgalQY=*C%dj*;@F1+s?Bj*M!&|rfFH|?1}T@h61ee
zYjbtx<oD;^BGR<F*EuV;5pR?jHk7%I>^iuvLhlN#Z3GS?3oTVzRG4txmU&m8YfVV!
zPj&_qm7gRQ+@7bGl4l|f`7%llWfKZpSFzyZ+gBV^HVIdWr3-ual!4RTBE`9uwHMI@
z7hs>5PDQJm0_b~R$Pr}Q-0yv`Ha98M^Pr*1_CyMWon(w{#^!bpKORz2ZqFD<EkkVd
zFW^ZfgBlcr<=5$)p5-Xv*asm%R1eOL;dCq>l`%rL(W5tfhN}ShrnFwuj9>0sIH#t#
zSwf)ASSUOYaQhS_(r}0}Y`wS|_FrWtEFsup+3vX`C^{dqrm4sq?BSf}{$EVOV7tGm
zM>%10><;SB9hh!fEQfNOhI;|Okd1LK^BpoQ<A4@Ji91FDnSXP}s#YVr5TUBsYN%^Q
z9(24Wlm03JR}(lH;Sdt1tECaIj1_J(v>qhW&Z@`Ou9?YyYm)0wV~RDM_xQs5eP>4w
zuMA<k<s=#mp*wnVViADglc>!V8{FIqTW)Dlj52e&*c!7NwrPk3Q-2w?6)shy=>yts
zl+4iqZ{vs3ZluiF0pEW7532nPcMr~fNY4-9XHQ|DV+%*meu#9R&whw}7hu<s`p^NN
z@#)tgw~O8Gh|6&2|Fi$xIjud6cfzqEzL)zU*lq;9@pmHi`w{HR8NqLkz=<@cZu^3J
z!^koWNn!Q+-ES<8jl`QPWs51wJ<AI6mQRS4NrfAT!6WV*jB*V60pQVyrwb$O5=BcV
zL3{&og-e<<;miW^dpG4slUy+go^)%(naNw#H7?`Elp70vf%6(mYv!uDw}vnzRw6h{
zTEJq2k`j~yTrV}3`+n!M$o<z})g9GNj!|;6fKj%HHlk7Z86uSHs}V5z681&4fc9MT
zt0rYe6U>Qr=-l%7uBJ6MD19t{f}t?t1`qSH#{)t^MU6}2^zOgv1AlrZ2?cU;QDZ^D
zm_3Z;4N-1_OfbQTz)ewBIdaVoCM{e+@fJNMv9g0n>gC;b4^t&LVEj)G(Zqp+a5%4F
zSjls#2HXEe!X+AB#_d>Oc~_?tYbE4M{`zLS>^wrNxn_$H&V%$WXO9Ig3^}Srok7PC
z1ale*wHdFW_9M_>1p7(AVif+<6hChyVjKKYqiR|z)Z|L@UK*BU{5R*9(()Kd0!6gF
zDq>9q?iy{Bs6-)Aj(r#LZId%?Bc~jRM9S!|yWVk=jWEiDw2NmVw}E<O=@6L`1Z1w!
z!vzBXYgAi&q5eNoV%8y$2K0bx{uKNS<X$w9)4*8yb5?$FF=SNf6uVB{_(D@zGb|T+
z{T=)?RnLV{!4P(~RvUFPL}jL9!`e{5HkbKInYu_6lzFjg@Yi*k7#h|*#)&DQ8kE1C
z4&0hYvU$Yf{&>iS1I1ULtGmNbLdiuif0iWi(uCEFc8qMpTi(07&TVsyEFq~|MldW0
zxic1wL6l>LRX|>r3f1Q;m%r?EfMp<xK!qSQdC$jlN(YR|t$s97gz5;x&s`5BMtDa$
zQ&S9%9F@INyLSrp|1fqA;h6<%wC<#%4m;`Cwr$(CZQHi>$F^<Twr$(Fxp#0U=M2uQ
zM*G>%9_(7R>Ra!M1|1+fRp=^k$-OW)*Ou{pWdA74tEWs+zbkiojLmD3>|llc1Y2z^
ze017y8#k=|Q`}UQb8U|xCx-_L>6skn%GYauveyh2r8Z#u$Lvbh<bcCLHO(qq9CT6t
zlkLRU5h+!~1eV<KShoN{8yLXLUQ*{YwMEVzG>Tb7LT<)gb>rfZISC#t^GHj6=@nwL
zPuhLpf3}?}(w>Z}uRN!of|9pFcW25GubO$G4r+;5!U0gWg=?G#Z;S}V6<tL+yj&6o
zs<OG5EI?##LqGm2gquuv?G#wNQv~sMq{Zs_!evvpGXAxGd(xBYVAgIK9zGY5nyhK;
zW5Y4VN5lj<E&QZ)WfThIn4IV9X+MO-_Ue}BwAtrF<#X?No7<78#!+1t11YCW7sZ2(
zg0pU3jb;%CVjgJ48TZfUC>T3ehR4h8{&4o20UnH<ATq2JIh{4K<ch@8u@SAUQKZPR
zQE0logXhP~5>5npGA?E&K&~Zo7V6&JzXZGcDWn1R{uk7&x98jz7p<?`M=ejd190j{
z$M{f1Xp1UyxlLn47wg9Mtl4k{RJ1>BjvwO0>&|WQ7b(zhENcd=zMz?_(8G!behgN9
zHi}bsX6)LG!O%zKXw@nFDfh~w<nr1!=nBVClMMA=5Al*Dg6SmeNMx7~{7U?5>_voS
zo}0)a_Mx?c)Ar?%(?zlWkj$z4Ft}55A35C?x@m~)4}%q7KR~jb3nh?jnedqm956>_
z#M2<@;MupziLIPs<xK-|Y-h&K6Z#zO)=lF)%T8Cp-{t@9CMr)-HOh43D(D}-S;Z=|
z+gN%;Wjz_dh~1u3Ivv}fe=lo`8hNl6&kLQZXF6mQ&Fila@`K2uPmE!<?SqfiOrDV~
zQkL)Q83yevd9mkztRg#<t<R5|6I7UU&5>OFhSj}W{)^7jHxV=Q0+0L&p_h&0cX5)|
z+kY}-=u5>dQGu#lGy@8;Q7O7E>D?bCj#m;Qd~Tn!gt|1ldoltj9-6N)<jkBB9z)_Y
z9%!@z-j-coU=Uzmk}wH*l%813spBlEOSDL8<(#u+R>0$qHc3jgf{T|#d-OVSv3!;+
z4x>Z2E~L82P3PPYtI<w4P_okCt5ZK|td?rv-=%JvJwmM@{achehpjj}qMwtxgj{Bt
zK8B}S{bTr8r|(ZR6`7KnsDnS$4v7=)ZDn3H>6NMx0H0v!fj}cv3lB9d`=0%>W3D(3
zUhXjinq}%V-&vo@EPcY4o3t-b{z$E+o(U)3y<Ts@9L@XFJ87G-_jCxFm#>TQl6D+&
zh0J~5s9wuDN@_iVXpNrGFDLuWieIp9R+P<UkLicWPe!db;nbylxG*kAF7U<qdhzrw
zfdGy<UO1j)IM;w_c;vS2OtMc4Pu5>;IENK(uMtyi4KOY6Nb^nt17UxT054;FN~F9W
zgW(=?vWL}_(0<Nz0*rAI-*iQ^=RV+8S2e~5yi4dq*d?lvBoEMrK^YN+R4)P-l*<vj
zRm#?JcJ``oD|E_J*QN|Cii^=WDo(!Wb>LJlVR6T(-NkAjs+$#SrzGS#OE~WqCXr|p
z*kupW`=hIfxSx4%%g-Z-CCzG@S2CbhB}Xr(`0`cv?j>lnY3I|Zm$8t{T|D#|%OdZe
zPZT$$2)<oI7)TXCvY;?Cw2E(iFbV~&wDKY|!u*=h7=KQeD0pqzy^1IP#opjNf427E
zd!YL?2##<Khux+d$nc#G^v~7gQ}9nl>x?(wN5>^hjPs&xDwMigT{-s!i;bs%>0(Bm
z#X_3=6%G@Ud=1^@U5ev-ICr;?q>vD|@o~w+w!#kdJNgD|`B0ZN5z~?2nE&fjadm8s
zDhIbJ_LS?3xj6dpgZ-F-9QOU&nS!H7S+y<6t<D-(#LIux<#e?F!EmHwW}y4uboq~&
zyUz^$!^v!Erk<G%&np41fuNlR?rcoam{pkFoX2ka{+1m}*4GF#7(TKrionE_!3@uP
zd3i|@WYIz2{AlxV-SqP0TX#i*=rGgfVd=p!^#+}U!(zhbWeYrdw75Sesf^k`_oFFi
z!q>_AhV?A|#NyY5w0-se%xu0XEoO=d0xQ)p(pSs8j&*^MJPlY5!?^~d(|YgL_E|rr
zY*U^s<e9JZe*S2CcX442<rAX5<4X+7eSJTt=4$30dtG70N(A5bd@p|Oc<(%NUx}zW
z)#>swD<knSR07jgGCvL#E)%Q-P8n(~q4ioQMX+?rd6kN0m6sjKLym69%zqrq0~~}(
ztKq<6Y7NRUxIOc??X$Y~Ve)hdddXK0k>$;>)>!V=`FYKytpF~dbYeVfpV3gQ*v&}o
zA?cK%!?ZFahj~{CmEFt=vAEE(vMXWM6el9F#q4pPwAp;aQL|~bnRqdp41|-=JoJlm
zl4ZL!yUZHKQrG?_6?*LD@b_G{qYINp8n0Fn_og@jx1(43Mk8LxT$@wU!~nPF$-(#a
zY0XmHGzMEebx82HfTRc?t|W~&VQN7ytR<7~a~wW{km)|TNRu24r<7{`A?Qj;%u-+`
zT|nD@oiaqvVitfE>K_v3`%3^zfpS3@TDf!5co$5_Ca<_9#WqYHoh+dAE*8C>oBo2$
zk2r7@PSDD{RmZc2M4J`Yy2lZ;kCmr9p!sgfXa(fR*5GWaW(UHG^@>^2v)ZmqQ9k?T
znTi=r9HR3_YRr+1a=m8q0kM5rJ#Cl^F?nIQ+xvtTx7G?^MDf(NodZIcuNsR+WICv%
z{lpDhk-UF0ismEu^dlA(tqy^N8%f%xNkt+&5mqr8c<4pulDs9C-yo=N`;9CI@b<MQ
z#rV+t;hH#@HQDs((1~`j)iDAl?V~MM=p_c=HiZK>-_=zvgWc{AY){b+-n^$?gFk}(
z3p?y61t=~G^)r?w#_tNu+cvyv0uGP*6a?1FPV`W3(jOV~&i${tbmDEPR^N3ChR2Ag
z43NhURwc-H9;c!~zFq-vW@I$A#Q;}r;C&(<uUKhN=A1c0PzAn0Ix=}eBt_5ZV*hYe
zTqH$I6M4SMB^*Me3>AMt@{*s9WJvNV|I!}!CCP&3F;K^GP$&mKr=GausyHZ9C2Ca2
z4qBAR<PJXSo)Tz~QBB&T6}FADROvx0;}Qiq8*w1@+-9cXFhtoQ`bOq*L|%rHnI_r$
z6xhz~N<Nx@sDSsmEs01}LVg=`j5gW~U#VQNV9@HD6}zS$H<rxpc-wHxG=Xvd7DsJC
z_Q(QSRJ&9ENXK4sivubQ8PR5}=vlO)#Ecph%w}Hw;DhjeL_&8(%^j>vQz0cu??*(J
z)C!|W4Y_FnySO$FeP-kh^>Wlb{#3aH`X^Ci&oe&e(Ws(pN)QQMLRrJlcM1G?Cu@M9
za#EXuz9Z31nf!AeI||i09>^%W%LU0LUvO2t=bNB@X1Ps>KXrwTh{V~O^A_Fl*S|<&
zdn05ajlh5irze!tqHxrFWK#_mr5DF&o+tX=sXNR)%r&A^b_hJuW1FUqDypC|;mKP#
zTZSX2C^7}Z^pO{tBjbs4^m{9Jg#AiL(${PMG=mjJ{IZtG$WU@~KX|!8J7Yn`>`y6S
zL1IgEc(f35*Z5E5v2PziLsF(BZO=qont)zp#R6RF&45Tp&ltLJiDy!ROg+Jdy*@fD
z-Sbog`m+*8Ul#beEM$9+E8iU?R>TP`BoRM1<)0^KDP^C$Pt_-nsvoH#J$i1j8lsN6
zEzzfrbCx$<Kc|}bxr{wP+<4P(!LdyP11zPkeX^Jk9d|h`8U($wKn;`rTrHVtHP6SB
zt;%!e{9!TBUQVG$qP7hec1dT{Sxupb0Ya61V$SHOqgwwG8_McgB-@03gM~KVE|`V(
zqoP?giAIBLNHqv%&@y()P?9=Q>|cI-L%_<qt^E|c&dX2ry#%aL3`#%9h~*s^x7cEb
zPhUc}ega=Qu&KWXG=FI=DL)&oToD5<y6xNL$d=;mMPM0o5iRl|;VlYxSh44U66nP(
ziPB-a0^=8%ErV1VdbS*Fj1S@Mc|P(Yq4}@IYiamL>b0F;yg$`M5`-qk#`d>UGv5Dr
zyQyzx>$rRa1uP#({P*38=|6TWdS(We|7o}SAr3qI->Ud?6<U^3a(-XoeWheIBDwf-
z`+H-XXqnrpad4cy1d49%sO5<qaE1fhlF9?#I85NpDDPd193r@G?1SBqcEGfYkH@mB
zZ6C(r&duvN-z;4G7r7W*B^zsD;NscM>vO-1jJ$v0GE}Fmc;)HgNqw=4gXhBkZN40r
zz{0^QE4HJ|I779%ql9pBc2fxoD2|)rKs$wJ5!zq4zdfGz=}r)GB#Q*B;S|gaxl3F#
zi?V+3bwByMKK!m+EWc9cgEHvvWk+ntYf~18N6kSIGZy79xIV!kYiX*O#dp;ro5>@Z
zJA*O%V}`{pKmW>+aiW#9O&$_EY$7ceqQ-K|khOO`Tr%@O$89f`Li7)qYMerspR0S_
z6bxn1LBxo*>LKxPjPo8+sKzet_7eQ*^74MO``UO+L0WX~y4$-uf+vT*uZ@>KH0S1K
z7|_4mj>EQSF7cGLL7ytFZXszwP7!$vE}_|oCb>v%R)$yCBf)GWUq=xkfY0(FCOyyj
z%1BokkwJM}qP<^!4XtftVrnu#%d4<4h7^EeFnsME8)wM<7%BTg8Ejb_cq=V9T-`=C
z71FhaRc=sT$~ZiHppnuD(5im{IE3)h`wj+YgYOnty=;Cz#1^Zb$j||2S{HvD>m^C#
z_zKTUzglNX@WQhKs>?M+G4OviD14DcX&Qkb1VL40J|43Z2*DyRM8>_f+S7lUJ=A1^
zoagcVRj(t<#>U?K1Z;P4bCRE4`@G+cx0N7+`fiu~I}eJNO!!tc5p-L2<5(W;H+Rpc
zRT#2&k-<~M;y7F+D0%Y~^<pA8zVOu4h+RtT``hfAu%!Kx!(uK?FqrVD(IJ+PVyY+S
z;AJv56dU${>@s%%C4AbY^!_xP$ZF-A>UGDhtN?FpF{`Xx!zU&Ag8f|h)-a*|58f{i
z`!adM-%$Nn&HDB0Jw(Lz%6s#DMzs$zyjr>o4*N-S{%QeyS>><UY0$+A3sG>2tY11k
zmF}iU6m)|Go$)!WIU?jV!EU<~bz-ycc=J_e@C8KhpQrOOI%pz}nUN?vszf|SybsKg
zltk;LxX4&|z5<@2UlB*6#s1d*ngT3C=s>;Dc=u@vYS3Yb5QvO9sF+olbH#CXUi3nY
z2W{#hXn>aUabQ`z{e*A;*g>uDSO*daBU-Q5$@7b;5hx*@jmDoK<AUEvxS7k(YDNUT
zjK|ER9&#{BwN!%ziX*y=esj;4xDCAG!f1v#z6geMu&*bnSJcOF@=!ua@cAW*03KPU
z0)r!y52zL?b*eS2uH#L&dQH(@v7r`O>H^1nD-~*`gFpE^MR`E=6@Q1OPnm$OUy>uB
zsT7=?3&Th4NEL#i0C{GKVH$%W<*Fjfzr9K|6aN%Aa4lO1_;z(8sgdp{`+_DInPpW3
z(hl|iVf&%DRy_CB66QDt+Nv=NDDwXhqT&h@QV-4KvuC-PzD7gf=1J!NR-FdZf|K*W
zjahao-sv314(EDOaGPx*kNjY>6g2QtnE=Z+L_C43UEZ=FY9W^G=qml#f6tN`x52B_
zmL26ftsIxV{qbabAz3zLP=)T(?2Kn&H0=ubXvN5+UCdZU(>X9Y3>f76$_T321_vhz
zZMZZmRiUwTY8;DZvrHEL<+5%e4oetF8(gI5lj*=lvvV_@d50i00P*>2mViiQ{R(7k
zh1Vt-1%m98&wrR#Y$Pz8_0StqnCw`Nx&`Q0%N=V1kgk=u(nvQYrf5__tQOjyR}iUh
zLdZHxb(!p%=n3^$vbU<vh~h5Fhy@0E|JpUX9Qo4O0@=8CY_!b$(k(+-l(M0n3p1i7
zydm;;I3T7(VJAN?0^a^^T(GV6zmT$H6lzrXy0g8%u#{wP=+JYvCQAcc{+S(WTeB0F
zcx}NHt)KNkuphml>J3s<jDV^VO^%sN4XjDgtQJuYCB<yRu505Kn%#)3Fm`(Y>w@5&
zVc(cT(FphO><e~`94Dn_6M7VD%A;UDb^N@YE9w@{-05;LZ>`EAnPO+azNTo=t7rh3
z9)vXb1`48e)uYlTMHwKlD&t!@QMXvYD0PQSJn+EUSMG+vb*#_uE6YYRFtQINpVYVf
zofXfgn<AI|Nx@#p33@g`qdzhnRn-%hMPh~{Nc-lTH2{%yOE~p4(RrLjfO82#^~V3|
zYi<lOEYk0Ai_61D4r+CBiWG_N{{?BYhp)f<L#C$#wr#43v_5jCNmb&Y;qS3~uvV<3
zdd*>O25~@pFwSvj*uWSZA#C2;k|q}e?Bt#rmdG+BpB80#uFw&{Vhp{?YjaWYn^>8y
zJ3gg58CRIy?-AT6VwYO!HB=P|(q|vo+IIh0ky&)ZtFbAv)!puvZ1swWi7u9zyD+_-
zl@D7NwyN9tB^)y;wA;7X%pc#|T<581AY=KP_hGxF7CZwHR&u^$jsnYgy4w0e50fK`
z%Okqvd=f@ydB6rN^wA!Esk)gWQeT|qFeTZ)6|=CRdFD|YChe9(J&G3t55`#|O8yz!
zqC>syUP)0BtQy_D|MQVQAttKs%y}%koMpU0Vn%zaWuXfkSY<Pqh@NRNaWYXRDS~@t
z#q=O`P)twsX%;^ISpDP+;M9u{t*5zE-C@u#4-z;5+;RS^XtW?39Vm~-Oq$Hu&`<`5
z?Iki3<#!QEZoizi;FFPs+QmOPyW5`EVrfX7)-->j8>5>FZEzW%phWT@Xa*aq+a^)~
zQy-m;8~ORe{#V-0n8PVm-EQe#Cg0R9qm`?C8{ZtwbU$T^W6Q%be2k_(@zMFoao^w+
z5=!6$gU7J9*4*7c0mLft%nR#st%n8Rdsu*4R*nnW$4b7_1Q$IG!Dc~YEl;`i#9hVr
zoXT!u>aDBRqQCtu^~$tyA(uxc&LRx)(YCX38M}6MI6Nl?!CwxF9{a2WYXWS7R)D!c
zr0l1qDUS)JB-!)7nk1z?G2xR6$|#(xk{z>@PIRkYqOJ|>g!JxOgGgC<Yh^E;7?eKJ
zxw5RNumkLf%xUvrXR{!NX_duhh_i~za>e|!BD<H)A&^rb1(6L6K0s_ex0}@`Xb3@p
z!@J8yd1T<~msi@*-dm|wlb{(&6+`fA*M!ZLYjNc{lt=s>;(YXYGu=GUki+9DxkRK&
z1tp2mNvURN#t{7jf5hOuJV}TpWMP4H6RDIEjP&N3mW(e>tiesrpNv48x%i@3S^qhl
zjrgI~3u`e0kZ&3FIYikPo3n2Y7rGwOts0^zIqv!RA>r+upB3lN>eWY6VlBu%DY+q9
zn>$AbGDGZgZ^QWgt?<l`JfHnZLdz?^7!h*8xl!fJC;v1Ak~RMW=64fO%f5Uok>inc
zD^|z1eSV?Kk#)eS?fy@9nN%BKY~Em3Do&D!9|X@5z6obUJ7NbBqoe<stP`DKt6i87
zUj#*lB01sP*-qeoqHpy5YyZSy`DFObhPksn$TV}cb`7l3H1!KGD_7V43d-Yr<v`6g
zMOg5pgN^)K_P=UVhW~U@=-F8Qx8kl&4KW9U|EEnmHVVW8fo7sA^Yj-(86=Z_5`+$(
z;RZVNr2~MEPsXRiON7hcLU-^d)}AL=f3v+)UmgG<%grG&EIp&N<$-bY>(dE2IIN#Y
z8QA7bXm}}Lbe&%5-mbfv=I)(yP;B5ifZXZit;IvMz(RQbVn7Qg1|p49MaL0c&K~9I
z-Ngen0=NW06w}5$fpy>O^LUKSHrEXyyNUFB%$L>2)8plCH=Ge3Ldu-AJZ?%J)7Sob
zz|VhGf2~C<$0(Hc^KJU{IHWt;5Q+g*;q(5SX8z{jJ+KfdkZ4)G6jowq0+P+_67?k;
zKN;i?_^I8k>8y!55rT5~OpnD!<kx+aHgmKBBdpX%6$IMVY%bjgdrm#gjU>Af+DzRw
zRi<`q^uO~Lky}!&=fHyZpU(m>QJO06JePxpo?bXeGcZeR7VjwUsHx&2Vg|M&96lCl
z5xIcsj(4Ap>+xJ5v=pu@@Y`q<oKcixeyuPZv7JQ2ly$0SCo&*{o+9;Wf)Il~<Zs;1
zM1`?^|MEP9XxM{caAgmW-#<h?dXft+vOHbV$w+)h#B<X-43Kr!H>?#>!@kQJE+JV7
z`HXgdE06?FxC?f5x`jLJ8CBh>P>-qsvEZj@-2@NAskVTbUn-LFiLXD&ZC^B^YX}(8
z(Sk0L6~v2FI+0wUK~kX<wub}4X^=Wh^aA~GN{V};<KvF|d4S2J5Tw;A)sgK{f6`mq
zAv@c1oTTTgYYvhqnphCgkjmVk2_D-N4&The-AVri8||b~qo2UXt@Q^|Y96u1iYm`X
z(FYm1MZ}K<xc_}CDNHsTx#Tio2o}ziz>(c6J47z30v66snN~;3`&yY*38A2TqO(=u
z$Vih-MZTGWN34K&n>vl{WsV_9XO!gH?h19pVl#ec5Rni(fOxdCXD^A~>m*jw@fQtc
z--y3Co7v-6AHR-3nrHK%-!$~@$&?=fz|e#P+?d+-uk-2aqk)oW6Ij_WB}?U%F-v4M
zJPhb<yQL!$Zz=u(A3aufn4%mxTF$0IYa#xSNRSa7{^-B>3knI`)^eKWUk-k-rF~4{
z<eKEGvs?e#w1M($pHy=ms%5`-C26=fV=qE0MrVTIuxCv#kndiKNqCZqcJ2@b6a;w=
z;bbr}{masm0}6)^YDMq8CWWG@&rarOwo$q{u9FUse_!KAjCoW4uCPE-i_JMyEy|d!
zS@7Z$VDh(<&`|#e7-z_l`sllQ(Rn25Yueo8D@>`ak<g+1iGUZ8o(l<Yj>s&A7lDWC
z?2XX&C!*;o9w8x4HDw4$Dw=IDmDw|f(vz#o-6hJaQFN*l0D>;OjAdAy7dbrt5(v_(
z^-n|;T$dAT_h2CEx&`nm#2XuLB)n%?5r7~xesue~k9N>Z%ZCPCCh#lI+_7Z#9NsDo
znb;^EQ<aE6VUCJ$N%#W5N|ix7Vm>p>Ksk+`6oUlrG)4`Tn@8d0VTGYP&Tli!X&=Ls
z>;=n967c2|{1jD<At2NJf4x<OX(Jv;AQPT^IO+TqXt=PHGfn4>Z5D~tK>xL5v>stK
zR8wU6Z?9(R`BS|nm&2mIuEKJVs{$%>8fj3{>3PCJg-?S9`M0COE}y|~-8WJUr#hRb
zOI5R4o-)=-h01-IrYtQ5J8R>W_`h+d?VhDg3eAlZqNnx{Ez8N~Ysnvm)0an{J85RH
zQx2K-$xQ89i<MSD!}0^VEa>Lm^E(PH=|)!+NbhMp1un%Fs_z-6gGdsU;MDTzO4NtT
z7i9e`X-Gz1T3h`!G0ROeqn<H|JhQ4$#|2zr_2x#=3~O=yY3N#(#)xMAh;^S@ShTe}
zn^C8k21VcMOsGyI78(%RB2tpw+hZpIM*8AT!dSnGOCeb@|JlqX9$+#oHIdOtoj(Cz
z=~Ya}C<sO=I*ODEX(LMY7=o)py`}V23~UF`#nb^G+iDu%+z-^IkWkIWM%ZG7x@poE
zq%;$boTh+x2|+Q4F&k(DOh2gYA`~|2oD=W3?aGyXj0vnywW|}^v8sFU2`1Br%avjv
z)fw>GyL!CrkP!I`d;H=QVFZZ>j~gZ%8@8X}8_H&0v;wW8Ch<gxluQXYdl_`Yt4_4t
zTk9&Kc*5cWk7q+4#BS*lAY}R~5H0#Fx9&*g>L;!M>xpK$hL%P%Gz5*YwUG_i#-eNV
zymY)Sl&PX#f-3(o2dB||c|7I|<>8V&tSHu~B@6Y5l!{dU<wBQ?RVFFh&oA|p_^TZ`
zXo-z5|NiY)I@r?jJP2xndN~VS7U0K+%aDUY#D;|`OLusb{uB*uS#AX1tkF|Wie^1I
zCv=RZlBudpGb8EWVezjXing)F&BcllEM;wjLm8v;Dj5OVZeZuWG6#;Z^KO{Z3Wemk
zaC+-3ZYsFy!?s-HO#-;Zq142aCWLucfoptJ=8h=)azDd$!J@rCW^q9Gl-5ka*PS+C
zA`h+gakADRM40(y+i&JNbsDpzUj~bN`2IkTZ;zRyz4B|2)?m^M0@nEr_Tnr}I#&?P
zomj23YQm^{wi6w)9@K33W%jG6`fV-P@)B&qktWKLK2p#6Xzr0P%EAJl;R$~)!4{-3
z)Z3gKK<w~v_Mu*2)@NI&^4gluC2J7##A`KD&dq{!eg0`rN|cJAztNBRDlSBkZdgQ>
z+XmrA>X(;0fzGwRLd>6i%;dJjWHhq85~R6@#>f2r?@hCGXFNU_s&MV`P$UZw{ZAY{
z$%O#x?|__JHeGTA7Lw|f-Kp)%G)8zSpdfrg<Xy+?Vdh&r<C#$7RI+dp?p4w~&1q_J
zN~NGF(*Bi+s9%GluA@D%_|I>vDnKPdpVC$pK|E1^QOC5~43&CLGfoWwzV1+k<;LYm
z<pbh=b2NYKk@C;&`z0m(qzsl=TH$7@waEdyGe?x_7|-v+!cTEU_H**P<2)lhNzUkJ
zWC457YVDy*?1@5w))9jn0NUz-$tKqLcY(k!$?!A0DR(L=5|LY(H~QRM>Bgdb^Mb;!
zm+RumfARuiKog*I0<+x6STnzp&)F7{7RnM|+?z#JocmwqKe!tVt+gOV4B-02XxE5P
z{p5wlm0IY|BieuM`v6C3*<X1W1dG1d!6wQ~`Mku*@XW?+zZy816n_WhZif;E!OO>=
z>dbGGD!{!q(nU-#%sRsjl6bl&%X5Fxb&WO@kDL3>$=|3i0hg^M8~f@hIzK6VNuLy|
zjpyMLC8njzDYll{mXIY=(2;A_=uN1_C@#9gzbBil>33N2<p_>Eag2Je1@gujsnZGQ
zVwi!7u3kW9XL>3^5}&V5mfc0zFZoV7g*d$*4>T2O4XAAi_yDBBwK3J3_tMXwv;*25
zA}Q$|XJo|AC_{I%wnWL@!tQ>pZ;9Dm(hU;uIy}wDG;JgEpHA)SCQmjqi@%4I9a9b7
zWIn4G7p$xKk;>p-xscG)=^?%1)z8dK={mYEFCbN~Q^%*<UQX!@BA&Y9x)rYez4PiD
za9$#?J`3%BxHK*KHX#BED|2;!+JS$3iRJ=uf{Om7sK>P<EpTb%3@Vc*lR~D1i@JU@
zb9nuF!shSxJZU|^yS!V(TgGiMj0ddC2Hb8F!2qczqEsC}U?3~TYZMJyTIJa&hkGiV
zWrlocj2*)L!g{cP&62ikV?X_vEYQ9g=CLWKE8hIumR*M>>Uk_6+hrRbVen9q4ZN-i
z671_09FYT@VxMjeY3+d5zPEqp(jxl39kwDPx2{Qgkf&JO63D?C0A`Bgk=68mml${3
zdKNBlmC={Q_*hhHTIGx7u&d5c@-4j6KBivi+0ttMTZ-b-q61k)98)j3s#uThdwkZ)
zS#~)v*u0n-E4TBA0{Id&s|jXl!)(Khp;E^iZC$b)zJqIEhAG};TMlaKs!Wa@cHl@h
zugdYfELM$jQsJfHJcRb%r$X#nQS}4II8S*0zY2Xu=KnCE85kJ;x6t+?#%T0m(=E5o
zZS*EYnsMA96&lM-WfpU+2AD&H1LWQSkY}jO)OK-8IFWc_79DJS|G7}i;F^m3-P3b=
zhu8D&<9LY87SmS74Om8-*Xu6WLr_6B?rAO)3pj2?@H!yd+UJY3u<pUF5Kg?5$qV!I
z^0e8_dv3=Jbmh+Pm?Y%JPOCh*O*kF8kU!{WhEeyXVa|<Y-#ZY-6pM1&=k;zFt}a&F
zhsC`SertD<9Z^sz_gM)YY*9HM{$$g*kx?*U_BG)4dLSucyX6IkOY21712qp(haRL0
zU?K%3%88oPE#@%>G22z>Gx@`ovWxAyeDbD|(RN4w++>{7AvY(vR(d^Xuv!}6A+mza
zehN}o((T}NPy@Am1oF@XotBA<XN$OEQSR?5{cu_QB~~MqBmU!fPnsi6m66Y!<i4PX
zMRFsB@6aNkDFYMCp1w+S)Bhoy@zF*edduYPE}JL|>vM3z3MQ_oU{$+m5$;9MLr?bJ
zjx6O@1fo%jv;tQcxKdUS`|gKrloQ|X@K}j(Sc^kD<1?4rg)jH?@vfs!UDrr-U$x*d
z&THs{Pn*=H_%7E=Q6YvJK^G%7fFfxctw)brk6Vdkf#yS<(Qv<+PlOVP)9J;97cQnr
zlhP#or!+{`-z7y()S<a=S$*4~K+4`ABpuGt?nbfQeuSTyw2FV2;vjpP;*hLwv)WqT
zX{6c288_Ir#Xxiu2O6V&<_6$bnU{8!CUvZFI%%KjaTc@4g+neU^c2yGb0{3DWMom#
z-7ma6bMNNrtF?$z@>V?rr=U@L8pScr-HsguDv8L=t~PcgM=90KKvp+URh)9vJZdD6
z0exz`Y^rF#bX}iFNQ~=4Z!COK5&Cn;*MnO=!Dc*8j(Pbecv~OzE|@&x7bCr&9fdU0
zgEk1UCn_Wn`&expfaj3a;!~a>paGs@DEpOQ1M^iWPF_iGsnj1rLdR^nyGYtuIGAO#
zMP>>LUr{%*pN7ci2KIWhZvaI1jw{`ReBeQDic=ig^Cu!>AjF4T+MtTxZwgkklFqcK
z|41fl<Wjt}oL%#3X4hUU#Kr+GC&Qu_`XTFF){SZe1=aB1WfyGe@LyDB++y|xgusG@
z(H!rLGg*l_JXp<4#PK0N+Je^z_=L(MHgdDr!8gpf4^*D<HGn?UpCT@<iGg&=jf>r>
z(0ozuu@ygnn^GPyWR$n>S)Ot3cgb2jZFSR;TzncLi@wej7j%ug3eLXxWTX&DRfX4O
zXVd)<K=T$Z%SCw@l#4DEtzQ@o@3N$zIa*KoMCmsZpDL&Ug#R`(h?LZDtbloaVP(O5
zX=(hL5o(J2JI)fjI1O(f?>$o^LFs0YT$G2iw^HM?3PgKwP@<w0!vIwLuK;{p=vdjD
z$n79<tccw0-bx1=6?9k1PyVoUvNZhb-)&tNzIEXNZ$d^B*2P&Ksv7u;eQQ6}gc`Z8
zuD>WUMX=bo(C!4i5%A1a2-~mPqUkL)+z|CX(e+k+s+{Xx4gt_S6A0)sQ0q?FI68Q~
zQVhC2XE_lCfW2&*@@eOH?ReMT#$_mj3Bz_SnG=pq_NdX!P4nO6D*;oad`Vl3*Glag
z{n>3vEDc;&acSD&Xb#J#%OnH~VT8*{3}VWjNXtM$904um@s1piv*d${?Q)yGi~qc0
zRD8x*6{a&R8Kg=H5;5{s7durH2X!Na;t-gdP3d$`O`t2r)JV{cyydt`rf%q2)_aZ9
zkzeBdJXtbpJ{^69BAo7OaXx=74)@Cu$~%~q`kcYSp^{0kP==ZRZse=R1=JV~C=Ax1
zZ5${~Fs?|OY4NCB4TZTK>V_9xX-o3x#kSV`gaA5KSk5R;_8VuLvr9E`vc12qTIwK_
zaglm4{+yOcuO}Uiqcb`w<s8gKWH47HCkN^hJTwgUy0$^oUvtA+KtuC;bQ5*0S;BKO
z4<@4e7pY4<E)jn*se8lh7ImC~?+!VTTF#&0gK9FwaZma-1Rr)CC+}%<Y_?S68(L(>
z+TY|VKlE5POVEXeduKzk;i#sA2uAd;X`iPTz6pEu=O)>T`u~=%Ay0_VG0(Sr<||1j
zKt~NF$DO+HBP|VFObhVjHX9D0m8WY$*8>+qE>wd9!7ugI6tIRoC)#M1pK0A3^<w}o
zdAM+KeaVvijeX1jiHN|q7s4}Pg2>$j&=8cLx|M|pb8rWHln`o|jwha;fxrX{B}thp
z_ER4%Ilz{jEXEuAbNQhBq0V#8-rh+wk~B|@yz`WYVN{@qJkfbA%4_;j(oy2D2xX>%
zBUof>{rl~Gm?FxTIH8vs3VN7;6DVdF{oUY;;<dVG>VBJZPc|J%ba$ad4ckuEDZeK_
zG(YH>5a^Pu;@kp0BtwWD)*ji8v2@5j7TpfBH5S59hm%hM2dh6sc#Ne?5Yub#1-Ica
z-u1<0aL?t>{XxGe1PEx*;9iUZiZWvUK5Dw&>!#R)A|lf@zN7?zXz5>(3^29dU;Rqa
zjG?bIUkMID`gWK!eHqqsK5Bd6t`KDAqk0^sUI;X++~K>+{RI0fAnB6`G0k~c`pcdZ
zj_^iMH8^`@3xk(mV0S+77E&2ibiy>j>Sa+M!h*~ARuc_+QSxs@B4w;Yy@;Q8cs8HY
ziRXV0dA8M8&~$bJeg#1<hhYulLeAO7wqgy_qVlH|DG1pQN-B0_>24~6Iq536Uet+_
z#8E#d7N=`RBQ1=t1?HW_{5cb^cyA3>$;AFL#R@Ep--E>)yfKRG77<^Qji{Pb77w@V
zJKe|w06`?kPi7@C7%U68jRj~mZH=($M5expO{i~e7|4VsDwBC_;gJi{WcYaq3c*y+
zj$NZCaxs%~FKUvc9j?DBn%Ye8+NFdyIjMVjvY4I)C`10zZD&fw5=06Y0D9!*^`R^d
zrjS<G#%eZa!)QLTu>eun%%S&jaGEJQoXt^k0yGx=naj;Vj*6fTzH<Z5??mQRi}lN7
zq@86C*uy1ZOMxWbzoDkYE4qY1UM(!luSYqPj#^I>NV`g<?N$;npYbrMYm-ip>NH%5
z%yS2)X47#9!ggJ)`c@Sm=*kpOgi~zC9<i41tZ~i{uAeC!2?$1EWP=rocBQGvnB@F0
z(0=Bhcufn)SWl5pjvoS>*L)ZMO=K|6C++TSCFXxcO}U_oYF>&O%1Py=n9<4B0+o<^
zMUo7uF**s42B^1!bDEn}O{%ZV6CNp*@Q+ZSW`el-Q^6&0M0t4H6IYr57Uj-{<_%NF
z+sX$dQY7=nNX{jk+dpcAbY^uqMC#3ncA7G)-tCpqcu_l8BTo1CPDjGq!U!92h~rCL
z3Fqnk5H*R24l&^p+%zA-X-)#qse?RTf=M7VUUUu!kGhGT5|n?(i>bJy0qvUGY6+5-
zcBZJlj)OdgJ^|qbxcIJMp<trG>H&r1Z$Hw$$e)9hBW%js@Y+YrQ6Xh_GHt`43y|O1
z<g64F%e9Ki?u71TN+6IE1CJ$(X0MXm<r7Oh^7r0-|8nh4uEKmnZ>`XrWMuj4?p*Z(
z^hzf$KBD<ET1-*Y87$iRadpDEocorM?C+V;FdNmIGUbQsr#zL_96IR>$<^8*i5EfI
zI^Gq`{dbWYVW*?FtFKBJFKKCGxOB`{Hw<L2-SQSTpRUqpXK+xTSn?FK=C|{!f@qzm
z$Ei1^Oxv;9i{=UDrNmV~4Ln&GP90~awc7yfZRF)yEW^`dCLR00ZJy%uYQ(0B?5?Gs
zQjk^^%@67R&0{1+?c0*l<7rb=vWWJd8`hN^>u9%Emjib{X<Y>S%#TMCqm;YRvfqu4
z*O`jYdR#FP(|OUashbR*ALzr*pQ)P;?F4GgeNG#3T&8DqX`I~A-E7x1*OkK23iJt8
ze<{z2f{KLC((<QHmR-`nxxHR*s&nPR#eeEF4y$>*LUnW{56GknU0?3Et$93O=AD-(
zp&Tw*oLFbo_u$?rQei#rT03x132&(?nLIxPot!Ejal3F#-=>~lLX*0CT!6mUO!99G
z(=b<$C{uCDh&6CZpAan<jkBvAs6RHSO<<UvM|%o-MqEGR4tld3?`Gs)6xZ#zafb$*
zwcw)i8H9rEN%q121t#V5zqS&`OpvPA;7pFxuZNe9UZ6dxQRj2>Rf^e9Ie5nPJuhcD
zq6{^wplm$VL5EbTrnMHJZ~%MQ4Z>A-HZ;S~bbrKE%Ojhwgm&qfzVr=_D)_C3rCH#`
zDJ0!FV`_L-Mqrwj<bte}*m^iYEIxgfY`WxEX}0AQ_Hncejcj#W3N75kV?b8gx<I4^
zHhhPhap8G**O_e$+4xP_lq_a0r$<@0teO_HFkiaoH$+yi&jX~0emiiZ|6pVS5*nS~
z)i1{^U}DGLqRlxXtC;#dnD<QmNguZ&8qj$1jmt;wkS|GJJ6R^l@V>#-C+?R1>ye82
zzucWLGXHNt)5)I#00I_-pLE!R-H~pvq0+6NDd@_St8b^(jz(QEt{NdPY`VtLvd!%(
z;d^d!<?RJu@}hp+U|Drm!NEhem~{3vw$4?T2UtdhZ9zqrws$(B58i8a{Kw3UVG;t^
z!o%J1!@QWd)yv}uO(KG?PV%?QhxD<BY_tezg3g(zja=qY?OSOikhtO+%GpQN9b~k~
z570CvMnwVc!~ecp=gW<R;XFwO#y>JUQmA(Wk=u<wtb>})X8`9WY#V3#d99z!hyL-O
zj|_X%pj+gqj1ublL||%GHo<))v+NF8%*<uQ$yhOKzrr5H-I4p&Orje8a@hpbxEszr
zs?O1H9RY?5QXwHDTQLb)tPU(hQMan~crlN+F$im8@Go410z16b<num?^ALZ^to!;K
zu5nneWWCs$i-+bQjs<B|)rz!&RAo~2-VmohL|jdC)olg!^u;e#k7O2jp5T)eO|#bm
z2X1~$`9X$k$YCjJ*+7q^1VdK&<v9`@1TjE26EYc7GBY99k0Ty=H)2h3*^&LM5`)Gp
z&EyPvMW+vk<osP8Ke@0z46B-??j~_kU+WkgZ{tfZNk!pmg}B(cC&g}`?^|uO``wtB
zb|`iV2$b_u$%beI<#OCSxqnfY7gVFx!Z~&-SMZ3CyqgN2y&4~XpUrqay4q!CgngG(
ztPy#kezmvBfz`4jl?;62Ga_;tFEhrZ<i`3IvzH*P2>#UrodE)8{hf|YKj07LT2(Ix
zZZyHUHsqwdBJx|>!RR<Ddy3aAq}*6kp43W%fp8h!1%jn%(~Nl#F;HDWLdIFLC=&G`
z>nO&6ww+g!URNacLEM_DBblqlJ82kQyI(R$Ar(_0v7|j!qnd&tqp>v9e-C&PQ*y6e
z<~?BVz%W^2HrHmusDJ}y<P?n8C|p@m@P3F|5Y(#TCQG<T0@H+IbOzKH43e&H9n%f3
zoH8nO_SCsHK0n#c_<+WcJ1%6_^UkCt1?eV*Izg<@OPth?_@;xe<%E%-k2-Lj5MQka
zoYO!4yor1@Mr@vpI^?ETxSYwlz;;!rT?vOG6C!5Be}}N)MK$U@OHIwMgqR+4fK~yi
zd4lW){Z9=8U?t&W0{U9y??Fbe--6s{XLqdrtK9<b9+sPvBVHv4$58aico($nU9hK&
zJc=3$vwQ9;v54NhH7|uPt}MU$N!{4rUy0`Z!E+)G$Bg#NVU71xT4*6sn{Bnp_`i5k
z1-$0$R4v(oTR>p8L0ZKHWG*vM&Lf5->94K4Q?e4B#xTjIHx-zHDTL3Y!Rfx`+DrHu
z`)payN4mCyXb|9xx>+k6fx1MgT{p7M#)mxyB*xK(R}r+^Gngg_9~6CjI>5`utwv_!
zY8Xf4v?oqyFq)*f(tJHR3pnGmpa`5Q?^6EkW^@X>UATadbgKLWsbod#DJG*1dmIkI
zm#IT8Ix@HV$AU3OA+8r1$Dq2_{nSXW&6qCY?)b@A+}VcAROYj;7fY)x=0n{&uB~|N
zTH<!>tPER`F_ylJ|7i0!<&hUms)zgzhp{d-;Y?hZ3dZCsKE$|fv`Qx|5KxjJ^JDn{
z^1V@a@?4<p-2KB^mvBn<W!F2bwspaKXS8jDN)U^j!>;;}C$Nk30QEyM0h0P1PvfR{
z-BMJ$FAjiKi@y=gU!d{OlP8Qshk!$+FTM!ZO+4bG(S*@d&z;btsc6{{km6eHVnixV
zqXtR+fi%r2LW>e8$v6#Ak@M5VY9$xatH1bu2UD^QX3R{UZi?R}t|`A$Plo4|dA~j1
zm|{#$DJI}NTk>%)(H#7HBW%>Iq)d&V#?C;|HO;twtXyqK(9SttBXm!aTn0fe-iSna
zr>8^?$qG!iHU+xKBYB<y8?%AU-n=vix9nl0&s-OmuX*T^(rl1B;Skm16q#CSR{cEU
z6@aKbY)EW#PNSx3giVfMNbyt-#4F5mS6*<Y=HCKlS21M{v1i&RtlCKz9}DW78WZc2
z@n4ck&}P=+m}ee(4#Ge+D*Yj#aHh{i&3sADQoEpED<N%xxCJ&(_ZQ31xO{6J1ojX$
znZSJB9c#S+3%uxW?_|&pN&9@Ve>y)`+bvPv+;U}Y3ss66gt}Lv^J!MtHSDsr5u1S@
z3G_i<<xbngJ@={pnUFXIkl?8?k`}+p2%0<}41)<&K>p&HD)EM>hIIm(m83Shm@QOm
zCxJiLt`Z;2u{9l5>R~$K>^HH_uo_uG%YER7$aa@QV-|z^H#YO}%AVPKP-d3T;7S(E
zRecBLeFf6V(60fcA0Sr@+Ws<jvVS2LeokGa(d4u7xa@uI<Yz5&#-twkKFSQb3u0YG
z*{`?DO25urPu9?v3#R3Q1?c<131-c-Z`x1Ut)3@uS(#VcWHnSgI!CPANNZ4=1gFjQ
z@=ky4gfRE}8Xc%I{KesH1sWZyw)CLOTgvR>kzhwn8k<NH@Uy>9l{Le76Y^?rl^&zw
z+uLK^_yLKalEj#^L>FwFQJ(TW5H?K-?5j&@)Mbejox^moHBC9-Sl9>Rnji2hS3}vf
ztMg6b@@-ZprVXk-XTg0boyrSdJin8PD2WL_0*fcck!MCCks*f9`7>`SY%7&R5vmb1
zJUJBSWFTH+#U8>pbKx3<5O)N}<$tHsx}M}f`$m{E?i0_&wQ)L?wTzbHyYt;qKvpKZ
zE(h%i^x>BV-9tgk*S4XV(I2lCv0s2jvsuj}x2IKeVBu7FxDDx4&}1?(^OJc~uHP(^
zS25bVN1VY7CLQI_9J@Ls?3FQufdrphJ6L5ehiC-+p1xMpA*(W|)LBsZ)n}*dbp9r8
z=S2-<v*b^=tRlVj=UDgPU??fU7}bAypL_;9hCnId$P$XpF(&@q@o~zj#YxY$_WpQq
ziUM>cMFS48`L=|tc6ix!A>)XrvgkF#31XGecnhMI1@ROKgiC}%wj9yY%Nz51vU;$i
zR#|>TXn5l&k72KLmwuB^;#Oo4vwGUXCSCz;{4g71#Z6{+3%@Y_7D22#<f*Hw^Qz05
z;L3KVCP}(_J8gvFb0%=f=r&(v{=Id_E|sSQK$~8h-@;raL8e=uYfdY1UsVb{(PBN9
z)%DR;w4$AS(d=IiMgR?-s%Nv^QR@g@oaMmXEIVQQ_-<hoBhH4`*oE)W@lX$I<96Gc
z=bB|xRGnM0qD!^K@EaabR-2}|!eaI`W<z^-5p^=cXa#_gOWe+P)r)F1LjI60)~@D(
z+-u`=b&lR#N;gDDeGDvShEJbmHEhVB`-YQMP5ke}GCy-Mw753<=8&A6|G$EpndN^Q
z<etQ6wKim-UsiD>1<i9)y!BlNF336sV%88y?VRmnw(RL;zxVt^zTV$VOLh)mk(m^!
zyNVdxE#8*?csRT|e7tHe_fOD64vH&Y){*h&wtZ8v{o#nw^V3!ioIXy^%a}|R799s+
z?`V>_Lw{PGM#;p8tES;Ox)03f1KCwM91=~-p{j@hREnEL|58r=z@xfx`2I~l^vri_
zf7=cJJIi3Z{C=puT|QoJ|K332FVHX96wBB1kC^ZN?zqR{#ho9@EHISrIjz=_?%w)z
z*kI~M-}19k{=43?6pFQ8U>y2X(5OF_GQljvE$$x5+A_0Vz(V(nbC<m<3{DSI@&kr?
z@}Jvg!g0y>!7E}Dq8XIBe@e0dI!RYAllL(VlIN=STlao40AzEh5On?*eq9z2yxQ06
zdGkE&-9Dht>F^cT_8{P=5sNRD5Ob<!TR?xK5hGWtyPchT*s3r#&OJ)ja1Yh4Jg9+?
z0W>(SE~@=z+^$THL9C2H-YZj9OWAZ-6p9!z_7_k-A}!QXOyRb+4y{;)Xr<E6NP(+-
zAgy-L$6QJJ>RkGOKU=x#1Lo>Zff0fx9Z<Ccjo!2fgg2gPK?DI%yJSTL>TYpCDKP|L
zWuWx)X9?#pyzSvKam3h!(!S+$dRmDrO&!y;wGfFfc}fjb=WCKN`k@Io%el^&yId(^
zhq`+#tmMr@yf`^`+#f%@&gOi{qt9MylNvIwT`3?)LeF^oQHbmS`n;#zo(zme1~y(p
zAP=H}frrS@P>9H1vuZ`Tzb44F6;<;Z=QJp}lyk$AInh3E$MOGx+8Pf)oHP}ovjy0&
zEW+$7YSNKhiU!6o%wE(<0=Lnl7olWviY!VaFd&ulRt%q{8KblkJBiT`4gnR!A$~5w
zbe{+g*6Nbl8{!kmB^!!jublnmYl%e2`m7^9tdL3m$W*8-1vz2fEg|8=Q6O3H7Yh>l
zQSm#kIfnLJMNz6@)hYTsiU7QU6cIrqHwSV$fJh)!z!-l4lWt^GfZ7ZN5gIwy%!emc
zTaIWW{M0D2BC#2Dm*ufNsUO<hUd-sB>>qLA9Ri0|HK2+^+f@?)l}%j#9|_0vxG#-k
zTQUXA9L&kxM1y!m-IYTsHq9|Gk+Nlb);oGqV#EzXT0uWpwE;p}N{owjdluV~idlzN
z4ai>tV9A*7V?A@#SZF^d&^W)v&nIJDv%zR+%1KQ68b)`@-xjx&FG)8}mBBy$uxPm&
z2D>~ZoA^Ff9dqJHWE!u!Xk(8oGTGyMvHI_v7|BLu${)>0dTykvE9gx&Sk`*0?4G1R
z@g7&YDEPFsil%uBU79w2Dyi3nrO-d5AbET2pk2KKRj=uB_1o@n+LKML%RQ;>A{_(Y
zHZE_nm)Sx*7`C$${H;ok4I?{Y?o9EY5jnf%TiYmjgVE@(LLPf<diSK_wZnQKEpI*l
z)K5K3tB$V$2%0D#MmM2_&PX8Lq2bFYICyIfT<hz0Bq_)TuUWz@qdYn)m3*kt8yYFj
zTKXuEPwJJYItXMyzMr)%<who;Hjh_ZIt+i@Ok&|!dlf7{qG~%OA)k@g{6w>|sFJTi
zWBXfD(=9ZjAw$Oy)o61wZ@b&a6;XZr`?dyhY6f=&KEqX&OR*-+j7;RNT)BcT5i`a1
z4%IJS;i_D!`2;TSn)|QWm5&aQytud*9}YphKIkAPi-_U**<T0Wme{&*O`0P#C2)Uw
z^NR(Va|uo-vb6gCih#NhCpINW*ems?5#rESL6RL)^pL+tNc}?dErc&01vUA!J9;8(
zP7+egz1|Ibx+?y%1FgrG6}^9kTS%oF9iB!<BQn(GN?VfU9ePDEi2${9rx2_+iU>L^
z=NBF6b>WGiYfTeH22<Av=FTCThGfpt%SdYBIQX@67I+lxfl5$#AgH_*O`3W8-0R?W
zX4YYIl5%c~RIjOjwV4b-dqP`3%et3={_N4YVQg#Jo;(4^`;*Uj&n`P*%f^?toC6SX
z?<c^>9k8TRqTt#Sg>nD&C9~tF<{{?M>5@HVWfiAl%r;#$H#<jA@MnU}>J`Sp_S-W)
z&80MzUlmo|t}1Tcho%y4U@(Xg)UQ68BLWNg@75*oK?vw8mH#5Wi!(S)_UbO{^7XeX
z7ZJrcgEI9}TB?5g;WgKQN^2B}Uc>q`q{S-!yd%yC9f~cKXM<wtUXMr@nr5$Aoitk-
zZ^BeHkdU+-Z&-dG0pW=FB>ZmNww^fVA@M4%5k>ii=C5MR#m_TR*1JO8!?Y^O8+axM
zWlSV&^E+p8nLw9RnANU9KC#P_Lvtr2%ICEV9j<04(+m#9LW^R?KJ`4sDBJJh;Q?T%
z0OaWlGDh}m4fg?*F$uBY5PqlQ`XUPTLs!oxit35QQU2UrUo~<R^%Ejbn#1>aQ>M>&
zoTYlN-S+D;ojWH2jbchH;i?NNF~mbj&m8mhY|$H%ZLS%Z-vC7Do-p{iU_Xf;Y@u+G
z4Tp(Xtu9ED)m9>;0vMUF_SM7W9NC&Y_F*AD6<t2Wl3Qtup8v)SopwA+F|KdEwH1d1
zksle9NNZ`LzP!{$l1`EIRD)A`&7+{+<R%$Gif>rMmy9**rLBjk6NvQ7eRQT_sTa?S
z2Ky)i<v%I#JFrZ{SNZPaCOv|t|HIfj$A}KT+rnepw(UK(ZQHiF$Fs+_ZQHhO+cv&A
z=Oi~bxxeH)&6~V!+CSQ+X;;_tthFY0S|XZk+t%&3T`CL|R%jWNiT$=KE~H(MD5KwU
zs~aB23d&G@$!qKoRWH3d&s=-sYx65>yfPMBF66E5t8m10x7vkllt;(J{^4B8Yn8i?
zF2Hwm77594AjX4ZmFZr~1HU1H+6wlm^4aA5Q^q#gnKeX924a|OrY;P+_T25R@K@ZI
zHsZWa+K_ROZ7N#7PG3*RsUic%DQjcJa}=f!gW~`q^TJw0I<AHw3;WYa6m~cwZXBq+
zU9>hSl;9D>(%6osNziX^x+&eyvnkDP?;=lt%D+TadyaE>K5i!7JHoucKEKo8Q;BZ#
z5IL_2-J^N6aAIPIH+=(WW+2G?N1JRE3+zl9&k;#wJ=U8Yk^WVh6DF>_xFA-pomv!k
z!|ZilXZ}jArC{hyIf_rh=^DE@Zsn@GNZyZ5+lpZN2sX7WM3GAx1TJx4{ae|uPG^fP
z^Fu3EV5-(`!HssdU6sD2oh#;va&Kq-rvE9up3bW-K=1yvd~Ibi`)7kiinX(fwV%yy
z1MQXO$JI@&zoU)%EL(8B`TWDcI>U2{Vap|CnbYk_1*8W2!U%b*FD?Srz}2gcXubU>
zf1369*eSy>0dd?NcWkdxCV5M_Lurk4YX_DSS&Zzl^4vFwYQ-&eDW9q3DXbzLMU7^*
zE)tN=wQ3&JEAO}_WbzI4gkEfAwcZ_h5T}!MzT@vQB-#l31+TPT{La29D(Rzing9HT
zP9>FE{LY>sYS8-xTr2&cbHRYm>|JAg1RI_c;|lIceck5j(X|o6VpcuUZJoj*|30fM
zI-(hF4+@Rw)X9~f1tw$F&H*k!4xRNeh_>|_7dq~wN4n`%*W?x|t0{%-8tBZLNauC3
zY&p^Khjpr?AnBefoAwNJ2-f$*#g-r7c78GPe<CfJ+5RsdRAmo)6M8vAOC@I;dRYQS
z28LftD@P~i-!*pj{{wLU_l%}O40a2mZ;zfpF8x_xTPCKXIGR5od9@NSq~)u2Zj}Rk
zcLNTWi|lZgcI!{~PHa2R8Ocz*Y;$lg`So=y4qQ*#TS!M90hW$^(9U5YCuF*g!3S+L
zdlUKLXm!BhE(m-TDc$mNe*%hO#|Yg>Ky_QI_Uj;?D$Fw2Sa?rml?HCyBN?aSrO`o+
za>iP=Se`4gkCANwx<evI3)wr`naruKVpAIw$FN*S29DHVfFNokLoR~z;i!xWjekM2
zf|Y${$2*y3=}gkxC06o&t05*Cz)|WUh>?N0tPf4(R1<Aoc>>9DBw)z=>WwIil2UU&
zJ8De1sDRXaqv=bg7ClB%*`FnmJ#Rx9K_$zM3~gx!5U?;hOD!3g$l_;<iU`s_3W7E1
z9e=OnXI=la3zv-{UEg<Vi8eHjh3QBlrT=YD%(x<UmJA-O49hU#{hO;{b3>$7f=h!K
zp?EDyu$k0s6c-o>Tgp0S0XIe#_`UwZSkGS7HZ?JK)E$E;WSR|KI>Cfb*s)1yMlrSy
z3UazYN3ic`%Y}(4KUqc#yAfxF4zMptp@RlNYTGcRuWwT><~Ss;FXJjw<Q8b9-@Z)S
zrYOiv-<+S=yuzrW_HWcw$u+6XZ6IXK8wz3&EF|@BpmGL3QD~JV5P(Y6I_M=>uxWyE
zguxk+O?h;xaYy-Yd=#5?w2iRSn!MBwqBSCzWhsMF9xwv(s`gB*wT=u(^m04f%_W!l
zYu@UL=>-{oc~e!5BhRliaPLm)!ATUsgZJ}6?}yFDTj3H7&F}5*>!TqQekH;rj1++2
z1LT)4PaiX@G=f$jyz`$AD8=^8?9S26@^yD%bTGu-yuDnqW@~GA-lYF|w|HJS``z3x
z-En8?$~1^4pHe`yuudkQQUx&WV1QvHSfoRcf{0$GpKOp!dIQ}MQG58;rYr(+A*Zk6
zwBTysW%fNspRdzM4v!BvQ(cT+lCU<hw{}225HVJ=wh2K0{bCT{^2ynEPiK}evE~Au
z;S+h1aB((L5h2l0qBEgKR<qOfZt#9qbOu=xEO(13+A64roKSsVFfnoFO>;(-*?HNA
zhi-N-DKiPnCtaf}SG^v)XZ^Zy>->s{u0Dn7+?v`&VwzJAdV9KtByOm;m#!QzBQh-;
zvJi8^jD36Jmhyv%7d4<axh+E<y0C{ole6A)vpsE!U;V=b)6<>N_2qs|U)%%V<Ma77
z5V(@?1J~1;-u3knx~7!6cMj{i_Vet^?#IW2PbXM@k{EHEN_G~N-9z~%lOFW&@B9MV
zH;HZ=K<Lr4wYwdon!Lg4@pEBDx)T1j<us@66}0t;p&!;&7yJSAA2z=tS`I)UJ-{gG
zDlMQRa`rbu)*78TXjX?2L(Q+<FAt+SIz#F~7eiR#{@60#2!Ugm+S0)w0Wx49&Ji7$
zONDy}$+CZ#^ab_(??*M#E6>qrlEs#s1s@hCUMN+rT=bRX5_B6$txwCH@7A^HWcRAs
zF{^qTe3+7S=XDlmSYj+?QVi+<1$H>e&Y*K9o!VpLJcCq?AU)c7ND!4eIq{m^Hg1MA
z=J!l<K#5;ayE}u;%ev$7Owu)`X|TJz)<V&Bd<eTQFK1m~iK)Dp^4mwlB=K-PasO!c
zAA1%m^_YvNh$2tQA6D7&s{r?dPy@XaUaZ}Y8oVy-m7GPK-LCqm$?uwqcnqPklJ{q6
z7O(CM#A5iwf(C$%XB~<|Vnu@D3psd>j99KxaV)da$P8(5yn$|-O-Sy=$~yK2<X00q
zxo|P5UDSUJBA{)SY{yJWRBnf#V|NnmzNbecSI0&4shrViKRHq6RnN2k8)~?{TWIwH
z@>DAvt{vaJ46^1cq+|HQ71E)$hBzqI9#zSxtT!Jgx|~||DC1pnSqnT)OLS4c+Ib0a
zNjkq0RhdE9j+r#UI_iHaR#*#`WCE^Cr`ci~jlMHm+Oa`WwH{>5Cs$N|TJ|S#Qnumd
zpdo#)f~mdIlmt8P!1>qm!|;p<bDcWR@zy%D^ABXw(uroY)075HLBP*8q!LOCl#-NT
zi`Zm(OPsJjc=6{^*?w}I<wk4E9nKN&oXoQDAG!~Z`aw{+V$XvyoME&mT0AYL=KWPk
zBdHP9?wvF#h}WeJW^{eRL+{M#JVSDX(?&XiOZ8+&f#(q9L6YqFuGvW4J#vLLp^BvX
z#qDWD$+iG-;m@E!0~ww60<=282oO3#*MN+yfJ@@#5sbdV{cxmo5&{}a#UPPsZAsq0
zaKlEs(TV~mjrao=mNyIDOYQl>Rb>#;XyIM=aN>l{exzXqW`*1oyLAk|&0gdbNp(Cl
z)ROk)AqYiCxmHJ-aCO+}K3*fdYCu+xk@^i1kl}0E$t1+Hq{aMK_Ltq7B=?Zg0V-Wx
zuCfplmNhpG@3nAg3Q#}<hCqp>aTUX!VX$X_0d1c3@TmIOKyn8Zcwl;SM4_2L-%7#V
ze?>+^<psk-2ii2fMa_*t$m?6WQDW!sZIjJ4VlK>}7DWEVou)*dU9ja-gnGvpltCt+
zZy!l-gfG<75$+uCW;K{DE>&V4_-ZiI-RogCnBx5RIQ+m@a+JY47b7Kr^TLC5Jj|~n
z-&1KVaEBWQrGykf@H+?O#=izG<~+~U^sc=zVmqzXVOu>aWO*fOA(vb1-*Z3%2Z`l-
z#e~ZP-v#=Eh;FdW+tUWw3aSAxvBNVy*odm3hm?`COeS5W$Lnitx+O&zxSmiOqGco@
zeRRxEv*u=gvHFGIIj#-E54~r7l3HMu>bag!MomAM$<B1kE}FygKg02v<yoBAOzwZi
z8$gYUqA(s^RcBxqE)<~M&|fqlsoa`Dm0=@L!La;I_6VPKfY7JTXhy>5e?+Z^;Y`|+
zVMhRm3bn3P1xR%udUbpmiL24Q*Y<EwhRfY#T^!&d7I_5IVwit8%7>-`7v+sZC9Ee=
zG7-j<<cA38)<cW+tM6UL(Y<W`mb)kuf!=F8h?2r}ec;43(#}u2<-|$X<D5WalgKE~
zKEE9hy6&7Kf9S5j0H8nwR&K1MhDj%NE8dNRaTMYqq0~^IOWh51Hx28`fA4$A6{em%
z4^We+<4eRWJ~UkM?0mh>bl-kS%Oox@3Uj=}KwA@9q87}J(+X2lY0K=~0ywtmYZg*u
zM3cE4Q+vSb&GEcR>K;;a*2Ckjg+^Zri9P3m>*JFwZJeA!sw8~z0r~rU81kwEjp}tO
zO0>NzqRG-in~RN!ox7?0hIk&+kNv>=yq@CxCp5^y{vV>8*qQ#9jOKrpu-g!SOV|{D
zu#p!k!SytR{$An2*S5fTW9YldNtnhB;M9+P{-CTGyCrN{mebnA2hvn5J?Yos>D5WS
z-);!t^KNnXY;kuY;b-21;oxm|<MDUT-lI2aHy-Z-B$Bds`fz=Ima=Mk@JvKBk;SJZ
z{B(6@9Ps$3aP=Oe<J{m595~Z=Cwh&-C}Tjm_=eK-O=L}T;lBN2agR^o_<nz0h?*{s
z@g$=BFZ&7A{Hani_V_rrz<bGdj<><@)a|HS<2lJeOXBo9t4~?p&Sb8r%7`3fk!6nU
z40G*6ik}`~UWt-WwF`Iw(1_#{HespoSv;BsNA6SC2)8(62WZVB|4^-0+uid`Dz%w<
zVnIPwj+D0kEuMOzJ>QtH7W|MuDVQ4CM#BAnJ8Lk%9yW^YnRGGBC0gUpT1-5vDmO@x
zY&t0{iKt`PTk^#jD@dj(Yn2l|iLh50E9LXb%9>@u0F#6zW#*X)JyCNt;E%qTlV}Gt
znl}l(8cB17TPn7yj3zk7-kLL&-X5$u$j{c0)Coh=hpnqa5m1<KwBh6Ao5q75`mJU>
zaVR0CJTJFY;|UovV#rd-G0Wp8-6y!mP^h!Se>kerbCE4%E3}bHaO3L)4|wnz4Lv;K
zCvQ6nPmY#ma)jeURU}9v3@u49xeC=?g)B`ce|ip;X;xsQUy6j`RsOLzu`TLqs8_;X
zs*}0zPF0iqkn}|t_2MAnHNdfW58tm%J5#rJgg*?LX&4I@*S5{V99b=xsBc!Tm2Ko0
zY_P&o$D%0Dea^~#T3aD)CKEMH<!+D;aiT_+NtkPe3!#o)#}iRT9V&|SXK6=-uU^LQ
z8%0&%dX39mC<0NKxu+!Rql(w6khm&AS&3u%grY2qrU9^yxmp;{j<@<K<}%jah{r};
z&2u3Ysya4A-d>oJ8OPTfe-ddOY8OPV9Foq&)oqV=z+#2cw7yTNlF@Md@nl)`Ad~j!
zHQP;Nn&iZo&fUUDyz1}wH=SiY#_rE-A#(RCb5e;%2>IYRVG67LbeRnR4KLt|KAV9J
z^AYbEU`;1HWK7~kP4wAhV&f!6g-QqI?usr;6K>oyPfCH=`m&}Ehywd(my<4szhnE|
zbczXsArchY<J0j^FYE&{8wxz~_+;SbhuypD&1!;GtJ6m=OL<+@ri!H2Qmcw-&L8BL
zgX5I9nSIeza((8o876&FjGZ!gPw;MOzkge()T@inEz^Pd#L!EU6ZJ!PtU|`aXjN5H
zBmwC%tv>~$5!gw$eq1V^uEECZQlscn84&5I^pFB)MoBTSL-M4FW3Ac6z?RLnp+2?X
zD}EiPEuH=VSZGcULuT;&qb6&=Ie0QJJIa$}+M?Ie#}O9~*GBdxqP#wQ5?~g1oPNH;
zx49MA)UvbC(eX8J!B;4O?$r{Ix8N^++!VD7#Di1Qf{*B}3X%iuPvD4(<FW8eO@=`T
zxi79U7@Ch(6%SHUnV7>dw(}=A7DZPMwQuvhQFZ_l?<S7OEhKywSeTekzhsEeWNuOQ
zL#kv5Qj%t+b<p)=NK}5V|D_hVCsHL+Udd;ML|-;f?nDN=jud|Wfvwy%zOgS0CJvct
z=&Eqd(qvc}->MpRgDH@up^=%2()}zQTkmHc&}ly74+=bPAt}ab;+vcS$5{U8gnUlu
zM>_)xB)e)N>D1<y(cg`OW^=TfLs6}+vm_c=M8htJmL8h}L1IOgCa<h$1V)qB5=HD|
zT*ESCQw0t-c#IXq0>WfCXf&0xH+NaPndn60e~e7Scq=-~2i-U~$VC281Ik>uy~}Dc
z!_>s4X#+WBPQ{plU+g$fY??uJUSL`pZe8g3mUrR9TF>O*P(?tj$&sO360UukprJ?C
z+2LQ#fouIR$RzEEW_pN5*nKS_bt;^Jr@;vr|3r;e0N~q=sZv=Y{+5`hHtxCJ#f_sM
zm{FkIuyp&E*)H6Qiw+%^&s_r{c#mrqN+`YyDnce(2aZ$u=8d(UchdzMEpkV0CV#$=
zL$~Ricj#Q6y4LHLt=xgs<b_+<&ay~?rD6>|MzzEwT)?~D52=mIW-zi=*Km;M8mE-p
zjXLb*(3FbA%#@^`@NihZvX&ImbgbfzZ%cVKCs7m;X<UprUd!<ac<Nz?-wA70?lk5h
z$j3{aKIRE=C4HxEcBrWFo@_gJ%~5qxMx;z|ir#!X)T(HzOfx{pc&U*T{*d(@jaOQL
z2-pFkqATi>eZYRWx>|v6LlIbD>bg!lbfWX6FM(DtK16!gI!F0<y+q33p~P+(Tqc-~
zhS;!^73|Z%c3)pV2(S+)6oxRLKRJ~Uw5px!JH7O~a!q3zoB3-`AnL2|Lgh!bb|l@D
z+$&Xb`93$4U~oOgJDc-lR-_QjxwCZ{rg-`tK8#dO9F>q=We2a72}i@NF~FV#V_6WJ
zr)yRtF=_J6$SRS*l<*C_oqUdOwn*ZL>%mXccA2cUsY@@@HZ>KOu4(4C(-^KrTn(@i
zHy^9dZ)Za^RdRPEFO)%`){@IOd%b`nrYqSG+zJy!Lgq7)oK(30t1{;i42;!6l1G|l
zS&dyWk&{T^TG6F$))dH5QXLTPJmeDF16_+$S$P_A*FU?<$W?NfvcJ;(PEXu*ab9iY
z%P?Nqs~C<Y<f4zQ*HHu0H#Mz5Pl<uSsD0IHXmV8S<~Pf84t%<>W%{vbbN8rSReSDZ
z)f>6|MiY07&1N-LYKzt%pPO5&8eT(6W_UGcC+(P&n_Z2|>zTs_1!go9xP>DZY7rqJ
z2IN?pcNInufs%+U3u$u#_@QP>bRH%Gi*+vZfi>CnAyIO2U#L|I%$FJk-5=p{=g;Jz
z@}*ybKTjH?KmZvXD=<3!3IOFpLj`qp`d2k2;<_4cOXM?SlYzu8{f<?5Vy&&!!>1W1
zmh<ot2dn(HBBBlG(&Ow-{LYF1yFd4NUqLVa=(hK_SYlL#I2>y~#o_E@O%K6a-yHbm
z%ryz4RcMy+d|OMhZN6}hUmP$HDvI7CIADXfBJX-O7N3)C#^^bhlX~i7(8H|ZKMD`%
z{3!TCpDD-mgHEXP-)|8r)TWOq)3%{9uYo6Yk!q90I)hCobob6;R^?ev=$K<{+H~n1
z08^4&4V-QQ-&juQhtD%Qv1KS+MwbB)mhsHPbea4BT1r>=9l5(`0bWMv1XUt7@7JrK
z6#9Azk6`qJk<yeV>-NXH!Gk~oB23Z3LBJKe{^_CY0TUwKB=UK+Q;M?E$x?&jMk=&L
z%&>%S)T^!~oCBw+(a0HDXEH)z#F)0v4Zl0;FaXxsIo+`OW9uyvmU&)+qz~p{M2H_Y
zEy7NHhc6@1kRT$X^7S~4+AC+-9Pu~@dL4RqgTCZuT!W4br&B`^R+{qu_2Yc$HG`w@
z$F|;8z0J6fc9+FGq6PV?<lMLgAeH{fB78Qr1*%~FosV7BxE3%C?Usvmywj)kwbQ2m
zj(D(IWdF1*QG6?c_IFP0ywRY+P>i9l=4`Kq@c(^n2v0YEn*gycX4{=Ci+Yi^zD%nx
zEvi(1!|EaX*R$ZDh>bO&YeP^VZG<x-w%}8{HNftGFaXh9;lj%pv#0K#^KQi@lzv31
z7Kev8H?g@qFQRv)#wC0G(5-Sp<%;)lztorVvXKUWYGLl2?fcN?^Qsn-UYr(^7C3`E
z>h8j={`ioCbQDDa$YKQ_sdvBv2I*E}$R5Sr41EdfE}oT?ft62I<!m*)FJ09sb?|~8
z39mTxpOH2D|Jy6T^oRL>XZnUcW~)7RRbAuWrjgpJOkyPy2}Yv9#L$Al2_o(&X_#{y
z+xN-Wzpm?cXVb+^YSdeG?W?05?p0^b<Bk7#dvtwvb-kA~xvNsU>)`&u2Jn;nMsMzW
zdbtajNQ|-Vi~0R5Vdebkxt}tZ9;TPg2kzSv8Uy05_P2tb-TM)Y7k6dfNlU1i-@tnI
z370Ckm)hr+bI-`~Hq{>Q`|kN6tGY7An~3saf45U%Qs#j9&IDtgf$T9mM}63)uJfVi
z(>Ar(U^9fH5pq9m$0$%T^H}c^5GkZ_o(Y6GgOFJ?r_IwHmVsP^WI`KsEXyT0`7Q~(
zGA9T0xxM;UhjQv7f`u|1SrT(Sis`fvi7Yd+WBa(sKss{-aRa4?S8`=Ne9uC~v{ReI
z899o#>UsyGVl+Nzz9wox7_)qO!Rwo}m5g}j5GB2eMP`EvxTDt0i@#Lv%4j(LHFLE@
zs0cV3?Jp=X>VzXI^+TeA%DT+;th3!fWepqFFvS4xhBvE$NBO<nDuu%6`dD^hzSMN$
zc6XQ^y`9@nnsr}c?C!zsjNbJ;JXodUC+Zwkt?ld6&Ol2?k9gyfPSZS$(ziO#=erMr
zT=%m1ui!(T%gm-nsD#GuSMqRw05^3)RMA;+Qm+(5CyPUQC@0o`HMWwVZkR<Bj!ZQp
z_J-OO<-1d=td)I_n+cQ%aW+U9-+`cdNSVsp(mOv%hpDBJGfC2xyf7=6AhlDwFYV4g
zigsi>;ia^yNp186+OLZuglrwd2#qS&ziV~0MMAcbGwJnmsB8l0n2waHFhd?ussS%r
zojrQ({XS}d&c6BEQPA)h#}h^*dm3u=ZZVJB6UtbqJ%}DEG&0VfK##0K@`pS{m+$kd
z=)NB?PZN2)U+TqpsC!lmP6?T^(2zi>x4^BopC$9evam`}2#MNnDcwi<&z55}`}df)
zo1Cmo#k)VO0;X&gvK3{z)R?iZqLr6q<`igo2KO>&jrHoBfmaaism^}Rbm_06vhgrG
z+X`cv<ZB<hvms;$Zzll0ah1T>yj;<zZ2mXueovveP0*b-T&~#EEOHO+!AOTah~r_R
zxTQ})iXtjQ8Igpp4@VKrBJ#>2*f?>eu>|uO!-~>93HLd*8KzFpMvL;)GR?|TK?T)M
z3Qu54BrypkWuMhCuY}yOZBc!J7Y+9LUR#=fTr2akZ|2%xF2GtD8S~=g!aPCMH($5*
z;a4LEGZplRY7Cn)EPOK0s(V-F+J1Cl1e|iby-KW8Wi<I{6(`<sQWKwPOD$9r#mjL{
zG7vaJbgy0Ov+OGFj~fem?g}5pMtwwV@6tl9$%kxPv)XtZkN$O?;d;|>c1y=G>|U+T
z0d$>W1Tw5u4u_6vo`{ih9-VG=7%fa=M;ow=gn_%RTv9z<p7Z`JG;kdd?)tXND%t~?
z`$Q$WT7!WM@u!?of>$*p1LWKxvo?sFllfpy3QnjE$In8{+$(^a-jnKV^puWPhGbA7
z3v*RLkf)y(@~NUOiw*D1=|Mzqy?n}do>_87<_XKG(&q7~36=A(K`--`CkJGXqggn<
z;sj)qJ0Wx4H=eErGLP1?Fc~$~=be$17wrwLV%>hOO};HNMog%Sg0$xM3A4tpYWpWI
zTJGG53zNEewV~vkHHLOf*SPLF4$fQ`6RHLrs51BATz<j)&#ic4Hk5BWh4&Y%5oq^t
zoC<u~?`R!(za(H5_r`GO+jqr#l4cTwS51H2KTQ~Vb-Eb+djL(nmL!xahg`?S?qYIW
ztiY{Tz?AYTXVN(|rekn}zQ3B$M7ZyQJwA!iN%1V(i6P{0L&$LVcBv|nmEKM#Njo{9
ztQwk^(%KokNykiU>{dBssNa4=yJ?%~k`G_J6~<V)8?Q<^lL49n|1Qb6&Y_cp!9%fZ
z!++&g`q-J5i5w3^7mfz=Qbi0yOp1$)KB?zuqT9|;jp#E+u}JNRA#uFlE-h!%kxD~;
zTJ|cjp3$JEbxL(`i=6U3pBR{PS{-RCb{<uz?nn?W@&e=Yu5gHnnQcIzVS?uw-AOBe
zI<iL_uN{50UKjX!KKY2*-3;xiMs6(nLjc+Tbv~TAmulllVI4F{6DCJ~Hyh|}LtwD0
zIHYtcLkh5R_76&}SbqNo{I$$Mw1EN95p|>Q(<)2)P%~B#-jV77R4{Jqh)D2gpZ)pQ
z^ez4wxZ3g`14G0j^o?8vn}9G41mOKfxhka&I5@ElpiFRp2L<$aeu8bPWK`mcHh1Oq
zq{c6!YJNL-!b@TxeGwwdOYvwrH;b-}7OT`#3+t}l1XB&Qc7aY~Byc~ONWK@mbE8&*
zU<f%ZczITxbx$MSPDt-?ivcJl-Dv3hcT|sqFSa^GKmWbz&0k(q0;qgS?vatpN$!Zt
zm(v0&v|F0V*A)#Q4GnW;?yqW#Gt*xO(SgQE@At*#3=5x0Orv+V14&NGMO`>jJ(TP9
zPXQcqq+S<Ae6_`i;{b=agLrxlZ*b`)fF7VGp|!&^$vCOrw^UdS@$hxn&I@{pjj^nB
zxW-bueTi#tJgJ4vmiA7PEh@lNhiB~A$(3~{Lzv7a?%|!Y-T2*WT<{c-B7Apy3CqXM
zK2pG^4oq#q3IDrQACHv2B4ZnJncyz)6!rR`gaPcr1S8JiANb>B1d6ZMX(y@p@hK-M
zpgnk|77%Tll6&)>4nTj0kv39)4<p-O(%-;VLVq(=EB<lE&h|{Y`+RdgQhVFMR^r?`
z=l2+w&^A&CO@T1J-c6)cXPAB-p5_tvfy4OTMn*;=)g50<5(HaqFbRS0pOTbb`%{2-
zhvF1~HoOQ|Iju{^ZH6uq2Aj3~T(1t=XNVSS1qjEMy~6yr73%f5A`Ze9#vgnshS+et
z!*gdjNcxW@s?5-+9j9SU{Y`^tf!~udisf2&izdvkN2pFU+Vz}RZ_S~9d}Y)AWsheF
zLs%MX8Ebdj4%g%mN><qhXA!qQR4>Bq<;UA>fuWh9SsgR!uE0=d+FlO|&inukKbXV*
zBVOY8Pk4#re`iE?rTLrB*$UhJQe$t+Bd$ptw@7V)GONSP%5RJLNXQ?r6rZ+$=UVrB
znLIXQ^%WO&9UO<dj!6^w;^CjKv4@X7@eBWLM;1d~I0jXc6nu8IQTPf!mLqvDR6t&l
z$WU6`f?w>%Ed?1;qt-1hDcnut=jOc<O0FEeBTuGxerqhjGjr2l*BQte1*9i>(lO4=
z?z_NA4a}1*({*bC|Ks^_5ja^XwI}n6W&2e#V+YU2t1JKz%;w8f7{p}x{sIrb2d}7!
zaY61(!zIJ_<@Wx3+md~*C?i{!UdX8V$8uhjfXU?mG^8g!{jknR^?=3i^U>!#+xx42
zcLkzq%eWDF)Y>?_lJB>laSDoD6%xpA<PkgCTkVZcwZPORUhXDAVn$iDDnv*<;k8{t
zX#Zkw2{cVG9!erx(Qn(3IIi<A^5Yp7DjGo-_Gs|SoLIaASr&T)N&kxy*_6-`WqDh9
zw0l|AaZz)??6#6ipCEC<UVR(~g^=}}wQ8D_ZJjYUuU}dkeCLGzYGZv;47p+g+O{TJ
zKv7Yhj=E5keT?5bLdZP=7JMDIaFfvYt-(l?sh0~9E}y7P=3d7gQvFu+_uG4s`zV%p
z_zz?3-=TVncJoRK9;%KxVR(^s=8QwX857tZN=6+UX2!1I0{x`eXwG9jSu*`RSi{@K
z^Z?Q5e?Nm-^MgeMC9F^f-tdk9{T0jvD?2az^g@Zt*%r#PR!JUFWC=><37J3a1MmXh
zc+MouYcK|P8YEdRqLU_HVQd@R8v3*t`X5hS1m{YvoFHabl7zspv;Qm~NId_h5W9dP
z1GFiwSQljd#3aGNtqrKi73GfD%lrxSVnpxe(g@@-l`~B6HSC07D}zU<^Qo(pr0W-a
zsu5y2Ze=5NmQRW)OFM~tUV^CKsHG4f&^OnTj360J4H+1*!@R1*g})0J{k=6~G)D?x
zAqDKF1pHj*Cq3Toio@LBK7N7&6sPvAdGEE<m}XFs=2sLH$aD5#MypR(jy%4!LhLH~
z1N<u*&_X?Ct9uCiR3?Zq=P|*|IMH|3RMy^8sXvJppThlI!BK>HAblE}gRP_xwaBKW
ztS}A)xo?W8d^gpI<WZCAJ)#|&qRD`LFIK>bVr6#}Y=rRDFVO5or-q*fYD{2y8ylAz
zz`{1ckc565mOHnz7$GhV7Vbs~`ZuPF*EIVF)gtzAbfX8WL~Mfwsfw!}1}Cyc@`5+P
zkJLva!h{39!QUpHCa(TL3SsC)2PYqN#KM~Dnt8<J8FI!%<6S73nTSjV#U9xYagT$j
z%7w``U+HUjW-?gf>_uu2U+o#z=7;<{n4-~z>y2fO{{SwdqAimcr4pPdzhP&b`gIO$
z|4lR(GD1F0)D*!fj-iVwY4qQ-92!%;@KG$7m(LfCdxbRsf?-&`I;vBZEiGhOvxt7?
zuBcwiXN&Sm@+vsO@|934xUL<M`Xb&$vxdU>>BMqnL$8o<VDwwoImWNjQ*zOmd9luk
zHi&0Lw}DVT<OJ9)k%kAaZYft`*p0TK^0_IvRI%SwS}GI7@{|~HzWO$g3~65gz27QF
zSYcneS`0xKrsjy=VcBWYh}^}CG*@heDL(xve>+^RkKLm<Qbo!ZK!Y-g9Q0KdO%9H^
z5qfH#)2!CpqCf<-ijZtr7Xs2u=0sas)Kb7WGHoI^=`lW{oe1?&0klL6-%4&XY*kw%
z%TOiffPDHA7HQJ}j*{IBkE43<P=y*=wJuv+ex!%jB3yR_xlVbAv~lQ!RiB50qDy&&
zd7xrNzq-RW%QpKI*K{0`9jCg$${#_ZkH|=wLXZeyAE^RWgHyxBnWgOT&F(PQNeVu3
z^e90fX3G5Fmtl7rsREf?MPr{LaJA6-TQL*cQL_Q0Nntbn2&(uGo6otrB!gVfsYYM>
zGg%i<4Lrd5YHY3qW~QV~0<4_6Z35<QgWJKQ-a7r5&6;n*VOh_DvAbbYbs7Scr*Fek
zHk6!rDRbeLev}0mmm=xY!zV|jS=svqo#I2=ucH7X2?`p9TW8ZtL3XHGR}Ay875ubq
z`S`rm%d4mZ+LMn7g{rxvbkW{7R$Hz<@cE&G;}PX*dB--cO}7`lAP@5H&Bq&Yh1*=z
z(_)3Z@rIec#kszbpkZEcb@(51?<H=h{#Wu;a#Y?T5QJBGk(e%*cFIj2fmW2phNNc8
z>JRR@g<O8`3PiE(y&6mO3eL@G4ek(-XUZ#iD34fkN#jut+CVh_&ok_I8el7N>mC>=
zo~H;Mvi0cN-gLI&B`>`6J{t0(s(Nl5;&N0iZnWr!y~P2ADu0E~FxM&(d6#Uu*Set;
zqgkNZnxWKLCT+7_3T;R#Ov)kDNpKHo^4~*Y$rol@12%gOdNt4AKbQls)8sx~ych)s
zol*nRK-}?PUvQY<uZ?^6txbwx9om}tgEAqNVZ0?dpP@#B9*n!L2PCih2CKN55ioI6
zos*>b4|YFU6bE(xtE)l1I#@r%ci~0Ikd}4)4t{T48}!^P+G!3qH}gdYg=`cv_ZJc?
zJFaDgpkxB>*A8eKU);Q{8YH8?PJYK1(Tv3f?o!)`FDJPrtx4{O>I+hf25-JA_O7aL
z>&*mEE)(ZVwK_gkG5v{goYv)=jQAH2*<&N`XES&o9(tn#Gfh7gyeXWT^kE|@x=P$o
zV%BI57Da@P+TiD0mZhZbvO-&rZX7}mLjn-Q+`y_cAM8lUz_42z#H8pi!qkbZZrpg_
z-ry!be%#mh2M)Q;tD|o@Tw*Q@b*)A6H{56|FmGMaLzEi?hg{J^L%Q|ogCe1rS*t3C
z!3s_XrP(7gX(r~mb}nI)o0YDDURL((fEC(rn)8AZ@+n+AoDs1tWU-r07XsM0^IcQH
zoINmRTGy6678Rr`I>py}=SbZy>#O;O`>m@ERASk@T?aG-YH*S+Ks3`p0et2(9T1k!
zKOQ&7?*;yQkKkITc1}#QSZImsQI+{yRw`aP`V&4I$$5Q%J)D104?0%zGL5;qN7iRD
z$1ZL<)q4N{DAW7@6M$l6{(lXgm{{0Y|5t?5mezzVbsOT=2lXRx$^aFb-e1Dr=dSkN
zOQu`8QL0g6)`-a;`r=pX_$|vin=#1x*bT*oh^ON+BCYIipZJm-xtbff9&MDYN_#At
z>}@|jz@6Anbdn$Q$1xy?g5AgOk+Yi_#r(Y)YCmzwqF19gzY#T`@8aG&2!=P{6T9SH
z56lQVlgRlq%EzNK?_lo0gScf4%+ovT+kSkv?~#uhIi?2PH(}DoJ8~llhCjh_$Dv_0
zgUYNAr2{HC2bbv_N^#8j;tweifik$}^p$#ZqQ|O6@3W*zt@$G0f>jW+TKWMsY+rI@
zwec8LFS<|(>A@u{B8l&Q@*I0kA67oY<l)W|`6XNq*81x!4qFMYowta5oDU@pKsJAv
z>@BHC(=5a=`MV7F<E?S^rfNw|GSTDzfm)v2l7QZgP*%zM)z7W;d_O;wpM^!%&PS)m
zm?}jTn{@liZAzYJUZ!h|Q3G0vbgeS}(sPHT;$~VrVpq>tkq7;i)G2HfBvRV$b~d1{
z%c^rn6;+A=tt@s;a__A2eeY4>CnW3dRk$s;h<<ZsK}gnQd~Rd$qOuh$HuAMCm-C`v
zWIceR1|?gjtEjL@kOX?bf0Uw`4U#9&cBqpjD2ZZ9@ot92gYK;!-NDa~#GcQlTcDd>
zr9X${H0ai?{xqRr&Oro`-l1-pMkIt=%$lvob~dh<VHAmRu}LhTMWt>z|04h`xsD^Y
zbW(V1*J+d2AiPSKeD0%>TUN;1^3+ZqBC(U*$*<h%$}OPq@r^&<Vo<8}$fkUseAAkA
zpo(!>6VkcMAWg>dF&dvsJI@?~4i)M)H&@0^YI|n7Aws;k)@^I@#Eaz*bepcyj|7k>
zcA>BWrk3V7R<t;P(Thq+HD2vxCJcQRrKr>(@{ny%rRD7a!p9`BumXXvW~9id98y(|
zGQF*ONn;oK7n_*?nB3$az&eGg3EFeC*WP;I%Z~^Z!NBn5$VmE@&u8`2nL>c|<b~{v
zK*twq$G4sH-Jm(p=-406`}Gi0Vjsv4ZS)1H;GVmX^TYY|{n!#`se11orO&FVh*jWX
zCqc|Y51P(mMsRAcy0n)Kw5DI9!5?38B==Zxow9><p+ybT%TDbLpJU_aaJj+G``6^n
zW8!B=9Qod60<ERL97CA1n<FNHdA9Xm0_2tW2ZAh|1-K~G0zU_>f*1cAmh&3W_}Mg^
zu|sHY3f)<*U}^o}p<R@fC>3h*+C}YKHvbxX*Ze02ojTw?3coAGZ1tsv|H;?UtRl4C
zl%=TBr$M+Kn#NyUk9(5MLvA|m`5)Fzl8LenA@l>lWdF5Lla99mbpzO_#+ag;rbG@o
zT^)3n(y1kOD**LZNQB93yC+Gxcux<sp2_<&vT0X;<scw@gcLF8nWpnzb?b#emLTOx
z48+*+ViwqP7MCFD==Em=BrqEbYs1S!3Q!|ATVTfqI5J>ulZ-)!{lZrF+|fd?!75T5
z{>`&DENhcAkE_$h(v6R-!mJ!jwCd*i^4s%xA@Vsd>vT2aV8g+sv-<@L-L}8~RFwne
zJfI;pi?NCp1tpdP{AF?wpub^#kY$GCEDgQ4V7ZxO%^j&BOA<zHAI(VwIy)m5p--N#
zmr&!-=WWOa<Va#LD~JH-11t}BHQfk=hPDbvS-bX*oD<LV&3Y*evNrM+KdFKHSk)gu
zLyp%N#Js_Yww#w+qi>VYrX7WAd(<nKVrmrK3U2nvwS{9F2Bd$HPz{JLD%sgm)H(O)
zIy(MI{Alx_+K4@>W7*+o<mB=23D2UxEaSfO!nt_{nTJQaj#m-WXY1=YHeKHr@DTYw
zAehPU7*ooti7pBz*0nX>dE)X9AgGnbH@sMR8Gd7yY3C(oC|6oM27;*zUB_zk(?Eg#
zqd?7y!baU{<Mdsv>GWR!^d+s(9<OACnbjb^Iv4?!xIhi(K7|ui&)tt1>r#QJT5r4q
zZ`I$d?S!_-$x<SE-l^(MtUM#6)sJIE-pIw?ga~AF*Se)^KG#pH=a?rttG6e$B|T^6
zd{%xjoI)1Jgl_WZDp1wD*Wp?06_augGYXE`hyBd?NqBtVIQBFg6KaDJSg4OxWz%fO
zQ6@SWxOy$pcAQi+!3V!Ml1rY56w7UUnBXLsvGE8dt^}s~soDar$|YP1<)&M|OiPZ1
ze$HQ?$lo#XTs34a{DS-k6|*>lQskGP``9(KlS&rr>48LgR&@#QUc^gVKxQOaFg2zS
z#_Q0-_HdbSZSKhwbDe35Ti(pE5vAyu^T463;kxx0RCM2};hS0)YVYA7|LBX}3&1m@
zZ^pz{9eRRj<u7{P9d7FMpw^a$&aI&!{z$W>8*Km`w5BugNE(cb@LCQHZ{TTn2$4e<
zcT-Qzl7ElOXLd2T+!B#nvp++*Xq3X0eYe=!4{P6GY;hVy&};)QC$9V$pOwU=EgUwn
zE8<yXy+Y-sMaq%l1&h4h=9NcfeHDPTuT~k#OwCC~M1>;ZalAzkz7Be8trX+pC0H{(
zFov`l5NOP5)`%vs)h4uk;`489+QYGKj@;T;N@p57e9u~cj&ZBOl0*XGdO9)A(EGz7
zLLqlh3C~Rvyxi!uiO)_G+y{N_seE1pJyp>jPtsXv`%TXY8r&-d=QBr0ad_0v0xoDW
zPt}dfw>y|BK^TcrW1wni@LM3IUmn|7erLxI2_o*;H#*5pQt4|LA!kr)uNa-!Ij{%`
zSiI`@(s4ZT5UmF6Pb>Avj_bS~-<<O`$9VPNv*2UFS<5^ee&gFpx00=#dcb>E(JX`f
z3f>%wk-}tl&0qgQYDiDfVnohmJUL3rVMzm5=dB%x#OLq;kK~d#s7r9eCSPW|ayVLC
zzmk>n0;?JVUc|RjeOevfjsW=eT=}(YlXO^SGF_CNM)KN{zEI_qp+1eiW^?^=WT~t)
z5iq~KG6iqDJBp}x(217X1Qs)=FI1~|vqBTM+CI5>XxCzeW{r&MPy{wg5If+1SX!$p
za8Q$5TzKaRt~zjBKuRoz%8k()+K@>?oP8=td7;*ZZ*Ndqe6hZ-TSLO&x}l~aM_G3Z
zE!9Xr74uCgy*^>NIYYT}h78y|zsgpe*Oi1I_kF;u*3qg2D)-8u0PEn(L`nCl0gBD|
zlekVCnNW6_SBnnh-Uu##V-SfUd(w~ri^13kfPs8zI0z|y8cVwD78JzOhaFan?=8m9
zZts8%{|x_$zPrmZ0t&l5<p>HRt5L1rC!puH)^*8Qc*k(D{Q+~3WOPDL=OOza#mKCz
z|8MIaBl~YZ-+wRvf7#@kQg#Py2)`MKr|>i>%sAbD3HbM;&`+x(Hp8>SZRNa9zkh^6
z>Hp>HI$INiY7&H7$KgzSb?1?W_JDDC`?z~}y?U#%sPD0;<7{`s_X5ZN1RjQp!lUtH
zfDUZL?6rtd&fPxsqf3;b=q2-UbzhXoLE+2*B+$FR8$;riw7F9`CCXOJrYya|d4P{F
zS2Zxr9&GKV(f>StY|z6qHSpg5lQcg7|36so=6|rB4&TrgD+uqU5msIbLxsJlS+^{=
zVZXUYd0fu>9k!DU%4t#C9`jiz+oh4HCE0a&EXThJ>u+mlbb~71tJ}hguKrUe9N6_&
zZ`WE=9x_sKr3%z}kEQl0)O-N4)k;ID5EbYsD`Nh0g#)sQBhfaBGm>+pgeMTo6Z_bJ
zS4YW;J@O5wx5teYnKgbd!Y<8A`nUdF6Mv)b8Je`=FxnEhO7vx`u;+iKT%^pV1Iu)+
zMWT(W1bmhlla`;}%eE?&`Km|4IOBlms|It8<v7yET*hvcwl0#@Sh$(Lg%$E=B1nFV
zJ(>=93RxhkhkK$YPl$N5Lu+QP;*i5p7lg=Z2P*$~Z4>8bB)lPMI<7g4ND~svgUl$g
zHOZRF^aqU6vAgk>{4B=W{4^chXwub#(Ka}#eK~;uA1#`IR~9Mo5)G^~*hePGx3Ve=
zG|WdQKTj+l1t`@TM;vEKY_nI}&Owy;3SuvVv@bO-`bNAi-nCqkT_h=HzWDx(TgJ_G
zyBa%qKYAwk1(Xons~CKEgTMcUm;glGO+s?~;CAqRZOWEW>7fXsAhnl)FibZYU5bS!
z8sk|-9aFzWWTvvcEA|CF@({2BM4loD4A&}c$;jcwyNUFML3|O>f8R?Rj(4Qm7n$m)
zk=r!Amm;KD=TztYZrV><7*Q#XkstxLK+$pDjh|CAp<rt}J~>AW();mCmvCz*CehuO
zKSeNkX-rdsB_)(CBXuhPNroKpC(0IYa^n-AN`1}zBjvDn(-lRLLmY@%Jpa;qCz|*b
z;erNK3DqG%D*<X{C^cgw4ROj#S%ZD6t)#}|MTb8>)Af%EODhYTtZOxXB=Z#%)vc;Q
zlRTHum-|RvK89NHi^mhqt?l)lzLZ906$~BY+AdFD@`y7ekQyuaWwisj|5j$t1!|mb
z;_$0(uh7z{{Qct!Wamb`E5qN<-m+_k?~Xh%tzw<m<+57^$Wc%M54?T#4SU<o`!FP(
zlGO=qkL?T5Y*LAApFqC1r{~|YHWn1O`mg1r;m;i#W%(@|W!CTP^ta_<6RuN6RLz~M
zs^ZS|*EaWRP#~h^p}NGVGHd0Jy4OCZ;_|_@inrNXhoeoG?|9ymu1kd83aHkEwU|_4
zrBg#G&30}U1B*#!xt93HL-5yAaIJV|zJks&5Hk_67_E|<HJ5se;N3V^GcJk70?nk)
zIvfxJ+7j$XbjZHI!Q5oG!O7gc!=@(-HE-_A=}8+4eoKSi8Sne0+Iozt8TN=?m(PeZ
zjOui6_ArXsvcBu37!-tF@QJ&Xdv#@W-U8%|f)SOE7XTc)5#nw@^NdAfDNKT~LiS7P
zVo~FJeYGU9+Z~Z73JfwfUcs_l(JZLPuY{z}c(lyB2(t)q>|sO^KYKQqb_{C>Ivu`x
zLXcxg(skAw0i2e<<1}D9uyG$_8M@%+1rH@U4=o~EGp6!r(^DFYYPJxoMrpl!dpL|C
zoG|a)nhPL{Kf2X)NpvHRVv*~$wz*)d=b^KQk5fswI#d$<)FFhq)&ld&T1b%GiX7k}
z&nGNCydwAJQs}th=SJvz!w14sR6hAz$47i%5%(k39u07KvjHAZE+EnF0|9wD+P(*V
z?4(7rIAN3G>~gGSv!XKRGFcb&-z%?a>*)ap9u$AMqv8ys-5Dy@_=>Z@G8FKv;^uik
ze;O<LgKE77ItzV7o1Ng08(((+ptgTfPtMNPAJ#;!_hQo_*7XC;VTx8010j#>gP;N{
zlaDs4Vk%rD*QI>g#PCfZN)__BR1m{>_fWyLK1=}?<rfsoP5`)9(;*e8DY)g^+M{TR
zuZOWvAA3Dq%9}%JtSPoRj>xNSn|^Eom4F>ylAcLXjRU9i4%r-F@9zEafdG;6giKX~
zPY$oosVkl@ONdL(wyu|xLMRA}KiGX&=rET=(ZwlP$H_ZaEUi2SDTAuXNf2AUJfb5l
z5m->jH`t0b<jYez2^g}-fp@Mcv{cx9^mUDW(3%$1@-1)6s^q_a3T0)!v_%0{88$Q{
zu`JM2*WZy%bL*bV*_M!7h-`qEaK{~ZJ3y<e*;^;CXt5Z4Hh4Lr2FjpO<9mZCbqbz#
zA{MUO6RB;of2gCGia8H?SoO{>Z^kvFrXgV*HfT{jZ9A593e(lb&U|{C<#m-%aAGLk
zG-Sd!L*9pKzcjmZ=13v*TRF@Q<AbItNR9t-_z!q*qa$W$KiJ~@o8aj%w>9<};)yK@
zF$bO-&vKgN`#Pa^`j5N}>;EOVls2(7b2cYnV*X#6hFtw-75`6RtCWu+oI-&<SApN{
zY7(H$WVVH%grK8@$N1;Rpe%(}sd*x%JR4-J?@K-nMdI!*&UI|U4&CGJ`~LgxKE)1;
zUe+C0R)^QO3&4lqj9gr^@pvB~XvzNV_3QpcjWVg|B8x6j(W?EwowwV3=XTHFYL5ag
z&_i#3qs?uf@NAfpppx1@L+S%dxU$@s?hCedX?0ipyg#2r(B_ID*uF9S?tZ#`J3l{6
z<b^E^tXtJ9f}OwbXLj`V!68J|(dqO>W}ojzdgyw(Kck#%rVjllNvc>)4arP`8M9l#
z^A5z;q{V@VDUC<D<538{3KbfU?>T)Z#vw@924eVpJ8F~gcf@#TpkktLKzatP4AcGl
zE!r0xp|-{Z!AhX*oPDOO?l+zylJ0I18qm&|I*L@+A>)lg7opDdM^__>I1%G9qpIqS
z*J?lcDt@gR@(m8-WRKGqBgVJGHJ!|QFz}bE8WGo{hFR4HSMF{XRFBg;xA`Ca)S{qR
zvFueQN2$luv}+TDB@hsBji|;Z`)JSd6N$*drU}!SmD3VPZY0u)UILWp$Is+JF5NWt
z>j|uG``IVj<eSqt24FS(=W&oa$W1_J`!T;TH4NGpk}DaCUO6-j8Wx3AQx!L-pz=mF
z2g*0`n<B8DFp129B_5l?GitRoaAU;NamUZyh5v3LZYdq3HH^Cl9e)|yGSfkvh#KuU
z^2#qac;{NP=Zy4H1LnR~d)=ol0y&Qmqz}OvZeRrlt`d6jGSX8kEg>xpKCaNSf{8xV
zl<?jJC1*>iW?FD85t6zxSHRQY4@@^!#?9BeWLbaA*Ej2`ui@&#;2gzg9+&GEGqkUi
zM^&`9dFtlXc{e&1mhF{ltM6%!U!`x0yCdW(VTUs-H!P=mYN_CtQJrOLOpKb{Fxo8<
zKWI{F>zrz=gKDFDLdwfc<17&i!k0o>c3)J_Dg{Ir9>^QfPw-&0jAz}uCMu*zp;D{V
zI3=-JICTIsKw!Dgn5<rm#plZkj1=*RVHEbi=alQJ`=oO#b+M(9YW14ZSu`nY;8e!d
z|2rws_TirElyRk>)~L{%htsN1fl#aL!lGrK=Cv~uvRKw!(l}K=#jq(Gb=iyt;w%V+
zK*$C;h0z4uV_xTOUI61Cq40&u-Jo*;b&lNcft=k^Q-!)0r!+pylRoCuctaFWk9DF`
z6CmJQmXB`05|09B>uvWSUsf*k9+~y#38At0NjOC+HSWEh_BiD|fmPtc^<1&Rb){1>
zfzJTc-D8(Hd7Ca)SE+~`t3FGK!XmjXz(4p94i)KuZ){lx^1;uh$%R^KDN{kJqI%y1
z6b{!ksCjMM1S~66XvKh0D)(Z#0VzABHG#e~-cVaV`x|tP!mVjBOALD2APT;i1cA(J
z<lO7JE^mP`Bv;0=QB<82qOe)PG9Hb%y`|)zVnr=m&LVYI#Q1kh&q0!QwYrf^_(sL7
zcuJDTGx2W-5XQR#5W$UEFL!=U+~IoqO?yR7UbOf}w9u7V{cpww;l$Q7()@U&ZN{{I
z*8(Jgv)$6Z#>4{%dFD#PuK-r#GMk(|uW7I2@~R()BNcd>)a849^X4;xEx|)@+0-IC
z_v+0fU@psL1{Q_A;Q$y^7$y09`~(hVwHh1@Dsx~4gMtAQ4i1Gq<p&<+mET|Mbd%rT
zfeY%PQHe;N)P!M1a?gBcfg#2<Z`lm*SOyqV0#D2%X|}@orC2NU?{4A)8UcIJDg=mG
zQ#d@M0j<L20!0Ddu!{^aHi-BTB_23Z4J*)!=Pc-fycb<Bi0ZK$W>7ejitq4gA2M{V
zVUf#jM`o(}AQO_FpaIUo=srxm%p+lXMq$I5@jeT&49hBHBEu3L1nb{et)P3#pDr0I
zhsh?BN|rUEdtSh7Ru=}zFM4=MHpAmkSn2D6uGe?K`+_#$UPRQPhJ9gpyd)55dMD{W
zYdn;9K%GX?a4GV+$$t@MoybTzuaQ2<r6pHov7tx`Ruu@`ri7KHV@w0R&E`84v*oc|
zIfvvskND|z`nno*$(bS-$|N4yXG8l1qVF#;o8<MC+O{2*v78P@t_({oSjf8)j8Kk%
zp8|Ez#SAo+c++W}j{ZN!-Z8kcuwB<p$9B@OZQC|xY}*~%wr$(CZQEwYNe4S?*IE0V
zs{Ng+^=H(mG5?SEp3k_S>+)LNTq7t|&Q$aKNBb;}x{4kr%{pn2Ke?!^HtrGyqIhil
z%J<yT9K9N5vmnA}Ja3KwOKTV+v}@7OxG2mex4TaKUcwgfdo(?jpux7AP56x)V2tgl
zB3q*$>D*{tXVJo-{_i!;_|E7QFzQ;f80oi|5u@exkoI1&`XT3#H0AR706Maz37N#=
zQ5iQRTTtHh%AA;rWqW&k!$5i9t<joDQC^w2#;N-!nPsW@o!q%P0+!xEd=ZPe=kYG6
zB-vp{ArW&p(egg8J!TRAsVOBA>n_cqq!XuXy!n@D7T&65F&lNO&_lH<D~3CRdN)8O
zMg|yT6)q7-tX_TQ6k7}}mobL${+yPWL%t8U)1kKhFwTKhQK>SZ2;#Kyi-98svkzMi
z>`xQs*ps)-<3Bl{?j3di=7IcUWv$K;Yc%-Ki3@lRV}UZ=2wwX!%KP>iA1iwa%tgMz
zai-QoFQZuA!Ty^&SN?(U!X~-&$GYKn+E{z^B+kGPO~>iN{12iFa2A%ABvP(|hQquI
zPyiDuWYZ8Qo`@75S!2v5as0wF<FqfHCh&Ezi+qN>fET>~4ay0Mbt#;Bm?DompMzeH
z_!N1Y+~_Gv)S?E>;~J-EI+r!SBW^uW#SGtUni9}l`})n3v+^PH%;RHojWx~Eo2_oy
z<VB2->^>eh5ZQ5#kSv2J*dS*Wnx%3B<&x+L*=%xhcW?$)M;@2#vyVFn7}aRIh@NN@
zc28Hl^JO!qQaKESYPpV5d8G9ymGj+zzlTWv2tmMn*8X2E+IVJZ6E0te4wK)JWlPg2
zcuG0SY_kj#xQCyEleh-3wJbw>!~-&`T|YgO4Tb7|DdL}lyDkGpA0syF{9qU-pmSqt
z6p`1L<Qxj;C01C+z?*-vuHb3WSZ+Vt#&=A;fRo%_P4`%(PEEJRMqS}LYCDtK6(3xr
z0@Fq`m-I7_$gN9~<}umMro4%`9%hBjj18{Kix7eTvZ=UTv??`cejbg!bUuh&tUE;U
za`zx4J~rwcpY}qbeL;;zvU&f<T*mf)7mgU&e{_idlljTE=5)*<8*I<4(kqzQGD11N
zKB7L^HqD&YZ%1kmst(T$;g?T@Vw&+5ip_cyvzD`d2g6KM$@pFI@%ZglIhWkrot<2t
zc8s&EyRxjConDs+TZwn&1Uk*DEWiNPa<}j8ms>nQi)-f$h%!2VaNbqkjE5Kd_#DdQ
zAzyA5erfWxiZ9g+mW&cN!ilevC!m!4CzM?j&M#T<<bHX5UrU-U{)950X!5Vi#ea+Z
z@`1z~kWKlGE0E`2FRL%5ZXl@8mqMjgkRvYqUpGB`pEJ#JgGXq=lA4j~!IE13XOf<X
zWIIv@z)Topghh7}fBZfY%g!^TSN9XLkOVA4V)1vmT#hGC=Mi8A0uPDH^$rz-P2=oC
za+<V-es3!CB|6xb1OnS(ai0$yKU$k)N%$Qp`tI2xumtg~Z1T>XfdTyqH#eyVqv!~s
z5li&Bbi9Y%9_ozCiNy=pwP_7^tiqzE=s&eCk@;PdUZM^nP?ATxk1ZR~TS_-H9tO3T
z>x;WR*F9Ovff5+usCp?Ld(7#{kvb81tXnaJkw2htqr3t{Z$$TUxK~bZs9c}5J;{Nv
zC+%+({z^J7b2D@WJy<KXz{-iv%ftK9uI9e(nP%1SSG16DY+)-k{Yoby#DxNK0}C!c
zb&2br{~uD|2ee+60!C1Ar^sjuy;=#*{eXRdd8AG~z|b;!4kmFZP@bH5%2p&WNt?7D
zY<Sacf&duF?%X<Xsa)Cg*6Xc{gH+f9=>nh_8*(C$B9x7KR?QK+18sC2Xpa|&jy?sk
z$AfJFGQ%968}qQqOTjTq;)6Wn*s11_1M!J9GYhKA9{1asB#R$<6{#BN^!s<)j92@w
z)%K9lLoRFw;u_}iZ@QYozrep-_qg=T^6rJKyk*YC^)NXGJq#&;L9EiyYEudKxU%Mw
z0+fw-UwAX*RzqHIjA<I;Z>VO}H6!!%v+G~oK?40<g|PW2_2V>TVp-QAy0`E#4uEnN
z6Fkr>L``S0q=$+r)ae*u%2QzrbD3OLwnE13DoVFiQ3Cvyqn9gi9#a0LsYV~W5_07g
zJkiL6n~B(tkro(d^kXDKG0V;K{8pe@|Iv_CkW+^jcgG;*4`Q=W0;ZSI+QOL41spa<
zajFVh)vK_;0P#$;oX@gZhIX)r5hVsCNA9MDtJ*sXe+_40QNk%X)(dL)9EW`^#-H=d
zfx`&kUhM5ys*#b<(~y0oYI=%G^7bqT_CoNi*tHg{YZVk!@+mSi!zYzpbEAv|GD&VM
zs^fOug*Okxmu?bq)H2}&5H-2*Qr*q9kJ0Zm^$XAZjd7qvq<tYUGX<;=5yTT1H>jw8
zH!6-~ft&V2Al1ibM8rJQ73#`_c>fIKqqhA23Y$um*LaO<;~rtDVOvpPJ$(%p>{f03
z?(pmCFwP+VJi2+J8Xlz@W~F1uzo#-$|9nittp_Q0z%l=WBT4Wi%)k-`QPY%u0!M8_
z<{<yrg(k&LMgvNZh=I624cnXlD+7*-0;x+wL+Toaf|P?+D{ZCOZ-W3eA1kqqI*+qI
z4Nm$zZ9tFT%Lb|TI*fIi83Zq9qBGXTW)}S!#?h#(d)_E92nDWY?gP#-iCEW4Pxf7V
zuX$*hV0xqTs+vr@=@7bM-Q1iMMq=VDp0}ucrKYATz6pK@`ER2Z(Z3d9zmpY8+lQQM
zM9WW};T#K$>z#HmCc9F_B}ol)vGw+Np&hIY=^Go2j&@N`1Fd#BHo(a}`UUxnT8B7^
zhgG#s&HnKXqgd_u6hupGhIQ)F9glr%&dPug<x3u~hYd?H^wbObdit`O4SOOywmY0M
zb24){u`UW;YizD7ZHs4dAleMiP-TXMr>)N3LW3ylD8Y+;Akk?P{BlSnF1y^qN<S49
zh4O$wp+$(`3L@csJwnC&!p~Nk!kte`xYC5!_Bb2e>&T6jbz`;m@AXiDelT|S3$kh$
zYt-EvVP=R{I-X?bfi&??cG4QnD$S9>Lsfuv5gauonZuy&4eqOHR^DG%^=f~z5o$Vz
zE$LnZ;qt3!QnM%FyQnP)9^e@&(bfWu>Z8EEDy&Czn#yXv9tYJPOygS^iH<Tjr9aFH
zlDonafqoG<&*1#PWf$=+Kq|C<9sMjDmq8Skk=6XEWOU=QA9$y@Z++mK<gU3tI-oYc
z>7zJAkzF~g4p5P>9w<9brrNu3XDyze=bq=?sggnFS{59M*H*arz^y77?ff=2POvQ0
z%sg<7iqRG@@Rpq=RK&~!f;I}@(U#<jLIr+nL_GuZs{E>Z)bGqWmYMsa{|&ET`=1U(
zmY?OM|3f$Z|0P2)GBE!)8H$;Kjq`tWT25=}*!_5dy?6gN8EO!41Zy@%836QlRziv6
zRG$fK^IL2m!RElph~Bzg;@_obh+0RA8o+SfRf#lAM(6LJkfW(|W?{0@0%XA0Yb7D`
ziz=LUWveB80!@+#R#B6Jj27Z#w;IgEuoi8w*pe1blxJm%el^cSN>G;LC4x@SbE~(P
z?B`R&E|rxp{w*8q6f|xF={>qnJyCaQ{`$K@H8|0=lqNK&w~i{cdBzhwvczOwGFXK|
zn97~!q<4bsd=g)kK@aC!T>=LK+=O9@%tq=+UMTDY78eEkZ`0w_a%l#=!YVrdd1PN|
zv!<wjWt2%<W)x8Jj@MWjAwu5UL8M6w<YZC#L*Q_CO8}3PS~P1BJsU}6lFPj6TPeH@
z)-b2`fb_5yKDuW)&1bOGtU8C{SU-piFRi(vJQf+;$$(Fi)D&rjT|ql31zBM|I8FC6
zt-zLzVnHw2a(QO0vPla7baDuMlJM}sBUv!j_DB$ivURm|A+qH)GkStTbZ8Hj=0Xj~
z@?3*6ZdvdM%VYCg9l(29-!D)FTGzH3Ru+0X#NrQ}7zrI53Im3if<G8W>4tA;^L?Gi
zD+@ck-9W~$2a43-W2ZK(ZFEPFt!_EIO)fed^>S+Zx-6jMSnAR+B38w5wf!}bY_M}$
zhXn4P(bV*OQ*CNf+T<?XgyDCtvnt-k@aEu1MC@XfR#c`7+eO2ce4gg*ZKjZh=VmoF
z`r58@YZ230^0!p;zi<k^a_8yc*1#%wBSSB;iw6LYf!k@LJb$+LOYY0V`e|yaTIBlW
z=3=lfcSqZYTAH_mvo~L_Pof{&x!ZH;I(sAi?$Q2a<=gs!tZ5`#3PfKR!-jy^d@Di@
zj$`F$<19#sFYmwii_Xe*wS5$3i>zW23=XRW%eR-@{oH?2)~S*1p<M~msKrckLPk$2
zC}Y9lipLTKqq8^DM$c~?AL|DP7s$|`Ps{9IyX5>d=(l!X44K}(^{@NSEIB%#EdTyK
zm%hx-o_^<edp5l8)8GEy)u#J#ImpkP{2g^|Ty#|{ZEPfQ$*@|~G^VqAdbFvuGs%%V
ztvEGvA~Lhnpv;}Mt(i4te;9__`Ph!H{c!YxHkz2_%X-LxkvELr2Kh5229i^UcfuBL
z*x0q&;rlcK5@>Lz*4X>`iH{*{1mdUmI(dBE={nlkq2dHmvy6^ghCi3ATrhBY(-E?s
zs&-pbqt`on5B~gn;OclB7{&+F{Ple__qSfJZWhf6Ap(M~whkS^k*N4eN=MS~X<mPb
zczQ(bhC|oy13?&RCNBqa3$ek}8+VdPJ$)XiJT5P!+m)xctH$-o%gV;phvCfi$y(j1
zp3Ltq4&BYJRPx<fMyR#1<AajXGQp+Fx9R2Ur}Yn0Opwp@)`y44MEV+^o6QQpzm`r>
zNHu?J0%PH}J9xKfgdNrP*E0#Z{(Xd=by=RTRTx#uAH0V9Y#02x_VY(7i7GBf8VfQ9
zZJf;X$%bLKg52KN!w!~-<Q=asK3PMAQj57(9eES@?7;lMV)V}ZIVC6Z<m&IXlNu?X
z_ZUjGMZ71f<$z44C{^RiP3LK`UGR7%1{eHPGgCqkCnBDHof5=xxwg@!wA6M}2Z502
zJJV${4?wQTP9i04$%n%GNUQ(316f=PDYD|di9-dCn-uN<&NievHX=CZCD?+?u|j-{
zJk!ehyu6pj_gJ6ij95N4Z;wrB*=txF5R!(=V)XZ+7jNv#dN>s5unBksOm+);D0o6G
z*npmsY3aZ2oA&{AHHCn0ZPCJ`z{QOL!S&^p3It{qmNqNXQdt+7hmT3cy%64yg;0B>
zBqU`t_^<5^9)ZYLArP^>binbApuJY&XIE`CK6~~wZa&^}?%LvS@4h{|kZc}EJQIi~
zh`ivAKfH2_W05t<9I5@Vn2sf9iaCd?TrC1kgX?k4HwZSiI;Tn`uvx^N@lt6LQdtv*
zgmkma^R%AWV_JqAf7jyB4pzH81CHx8uLPSdqi-;Xmx0@Wf#Ds$>-E-r(YiOk;Wbo!
zqm9<c$}H}!uLX6yQ@d&&;%aok+fvuAZ=}Y4{SwB<R&@<L<&DZlmt3V7x0P$8AA!CE
zAMyBmsffW%GYf5`=3bOarB>Z(#?^AMW@$93TAekbil@cW;$|JqbWcGX``?#9F-(?;
zJVo$mCi5i0LRc)5Ig0=NXipJ5mMJYKM>>_L-s25@TvSpwRz5!Bw-GbPM+$5y!P_yk
z$LvyC0(v^I<LF!{<erLzs93Bl=$;6XPZheS@go^x(ygKG1c8f_L>kOodk;sq&J@H^
zF-=$(C{7l^eFvKP!&v2fKZu6TV<034h`mE5nlbIsH`!spoLa6ei2cLb;8h;57iB>6
zt@5vY4#)VDQG)2b%JkeCkl)d;zen2`*=Mcx(DdKNTqf)qAgMiR6oS%z9ZRs(L5tQ=
z*(7>V1kz%Z6Iq{7&2h`-iTv7U;FpTPbV!W5V*Q|n%`MAhscN=n$SEk-G_n$sSM$db
z{b0Z7Q{9j4P=a_aOk9$w)&#~~E|6CPUe_W&pe8NiU`xrCGB!T<krG$IBpgI;W&*<x
zW)V39wyyFCizjBk5^GsAU!+fiMg|9=4%Q)HpXQV|MS@ML@?kN8O|C->H`=TcK&q-!
z0grDufC2{I07dsd_uSGx7q~&n6XbTXSA~zIQAR048iP<Z>~OxjAGOzoou%-DVsL4h
z{&AFQQL)vsu7RA%X7q9Uv(J;nN-;PUe85O4Bb{%F?BlLQ^8AuZUWrR=QG}voeXl>D
zBh{Stk*U!)4ci23PnJE!tiobe2l4Mt%xzT@Bv}VugAN3hfGX?#@wQz!u@UOx!yS1^
zeMs;VM%i?j*`QGh4up!+$_d;;xfJ%{cf+!u0RhgJ&$GE0`k~`$h-M|=%>7^4wwc_3
zK&r*0wi-=MM`vi&q0qv{sb1mSa5U<no&(yPzuFc8B1K)QUCaqh?<S#VDI!qyk@J8b
z#d~S$7k&v6zKcfca4A&#Q3+Hqun5*Qvqk$h{5u6eP2WlPS~Lhlxq<(Y<`1AfNX48W
ziMO&q<}^UjM*_OLP}4_NUJHUmG0JORISTC~n^3|4wU_+dna&Gbwfp27r^D3;mL2~E
zEXlEOCd3r%3BATz>m(qAk-00pkq?3blV+kCmMT?jl3`7~7w>0r1vd3-oQ-@=dlNWG
z644NeNpPs2M;fHtV?!2_fazdnTNZMT`|e6uuD0%<L~R{7N|BnxvLTa9Z*D%J0%5*6
z<oX<Kip^*k0>~(@RD@hYmuV9fWe4SjAnux%gSW%i=ZC4w#`iNX#p8$hCtrlx^@rgm
zNQ}&lt5%MFKr>qf!nqXMjnJbnx9)l;G<oI=*GJ82VTFBz%fkNZQp$1((@OS)`X-fY
zZ%lJF(@M*K-5c3d$~b%*oy$o?AuU<TD|7>DCy-!1CBVDH>mmD_U*S5JpAZ7^1nvi)
z$$C-ql?u|`geI7Nu3{b4a`%d5I~5ZCYZq#~<K^@T#=S$yEyaO;JS2rs{HBa?CY~8^
zs3#I$Y78_!V|`A0Yzjz=ShD6IIbQF<az`7w^rGNGa(%ui1>FZMIS~mpG<}qE2A{9!
zd#&7wP0gFxZZ#0LK(aab8r}yg3`K1|8B7UKAF^+Avit}yS`QFTLO%O+OZlT89heWI
zc$g2XayuWuuZcJGuO?&MHBeN`O!y4Tr`kFA4p#V@Az#q*A7`t9FD;PUUAIK<WB=%!
zuMg7nF7@u&aW(Q{XcWZJ%ZOo=-4AfnkRLYCFpeg|uop4rbM1P)Ik9G-rk_X@OCH`^
z&Lpk$$xof&b2etAeW=_2r03pguxC}iL`LS)eR21GoepP!ecebvp%*<4QMU}B#al(I
zqTe(x31>IC2&I>_y)IG{e}1dqHw2|Ff4XhU?<3oFIs}qRku<Yb{jRf0AfkHvL(Xpq
z91A3os3tFXB_#ytk5w^55IZN!UBM{c^x`@qJ_B66DQfh-KTzP2_T0EWJUi2eBV@~&
zgJ+lFFF<WOU!VKJSdM)Dr#~939RIgQfsvj4f2JZn{+YaOa{ND&H#Eu%1Psu$aTn+-
zo5*(B=ccvirNQS<1nKbvBgK|OnFJSqJRV8wziK4>06bnV+{e4~{p0KX!OHypO8x%B
zyJr(1pZ`K5^ql_}8iARi%NzU4YXJ*akKQYOW#sPo&2vxQ(~Sku4Xo`?|NcA*Pxi_|
zeGQ{a(SUmSBsnUSzuNnr^TFuCUV81}er@Y6zLTO(<=MjOBcZRim*?w?0TXC+10B2^
zM6iEIZ!d2V1gI(}{yN|Bde`kYtRCK1oc5_Lbf4|QjK*dgRCWU!?2M?%LgBn@N;Co6
zScB)Iw5nGD!OG}n|9T#5<c~aHj!5Y(K0#O95sh*E@$}!S;MUv6`Otbi^bx7qE3HoS
z0~xfPx)nL4%+EJbC_Sg6%R@K$UOFvN-T1%BDD!85UhA&vMaj`(f3_RKUVgDiMW6;4
zk9uP|V>%b!Zw?hy(yFFBjo%y7%DLbGMC2!tCmW`<<QV1R5f)WPCvp2vK#4JFUj?b*
zovWP;{auVVwhT(}bqjT)OfHic6{W^|pvd)#$I3i4ifT;7{j^1jK=^L%ICdNDBLpW+
z{Ls{$88uA^bF{mPn7>Dm*XzCLHA%`Xh#N^(rOP$ohnkC^=YQbcD^bfGniKI)vJ=NZ
zj>YyQt~}%qO5Kl~(U!+oXgN7+ASG*6Wa^6@1P<9-31zX#|IYHl<u(5?oo<o+1ss|2
zClQ>{jEp6-<u$qGvsArBEh}?zj2mqI(6;xJ79NQ{n{iANgGVo}(Gc2*uAflzqV*UT
z?cDu20t<_0LfR}iH7fi5&A5)bJE?oS1(k}vsxOl<T{&z>*gQg~nq1QuQT&pi1Cn-k
zz{{shxVI<wYq^rd-vjn*8M2RXgYZMoas_J2YXi=!X4FoYj{c=-zx``2-~dOUuu`L?
zu0%joSDaYDE~V1>l%S%_;!lhxy#kDw))z${HHFJ&Cs3l*hyywPI2$5mo*me;dQHSW
zTbp=I_!0KrPWF5xz{nqP3o{3kt%Q_yL@;Ny>SJeVNk9dczqn*$M19MM?4-O(6y-1$
zK6m{R3~3}}iBo-auN%a;6&v8-8h}BhT=ofuC=9;#whxa;%yH|7A?N@$WN$Q`=+ebF
z1X;sPAg1uaPY~jC<l)OZ(ByT)0@Wg9%K+H|yNKfkH)rymrae8;-bHQUvWbBrdnC6!
zm^~_j!t*lBF*qq+$~f*>bEbS0^az?$L^&Qp&9JgYmh9`l4p9GySqIp!F>C#dUhz4M
z?xRo;x$LZ6fsUdKv$-oS&8Et5oIM@SqsAGd`6TFyLC%PBOQAtp!ap{s^WbnOwdLNR
zM>6V3TwM8mHJc$}N1@da2#13-V1pRB?^P-!MM3W$0*9vE_*2i#-IQGfEtyu52iuIY
z8edsd&c!F3$!;s8d+pJKS$iKg0Jw6Ol(w!O8Op{yNFCZ6=MMk&^AClJsbZ5$Et9*W
zja<;ARlVtLWHH%y+8dq~z{hnDKbu4E+a+5g-;bAgR@~YI9{~eBe5Nf0D;+xX=Fv)s
zT=Ger^^=OAD?f$Pkta-oZ5jhg3W?%sBn#8%w>K}GU8(4K)r>JfKIUp(fQQ_zce#}$
z=V3868&hjrv9WC=zI3{$CO^?y5MN^C?A2TMTm}I}?u8ltFK(2Orhui9W>_*fEWJ0*
zh82cPwV3-Qw(J=iwjFL0qQRht7mvlTtdSf#_r<WSn0NbF_UM3z7)m7tZ3Yuvwjahx
z_hE?ns;6+|7I;F@8h7@T^u)xOe@rIx@0#2X&_Z0Z$bbVjVznCY!?Ffj=GBfeBmBzL
zO<h?FPTL&dBn%tMsrR*zD(5DC8%yEsBJd8PX5b^oY9*g~w%IJLTm3+w)m(-G1CC`?
zOApf?3Gy=Vz#=Z$5b9@v`cwD*0msj>`EDM`ZWp|anGSC3QVfon(!Pjd%<y`d>DuOW
z4S`fsan<Lz5oB7qwpiy2+d4O}%Fo3snnIZm=h#uG?-Vx(i3{D@>$Ym+q28lrTsHvZ
zi<=EO$E^Ujpo3Wx2;R|r2F#qC!WDRkIu@NeSAn`{APL$b+)fRY7&oGV!UqQ^XaEn1
zY9{t<_dgnK`)vnwI*N#2(iD(l)Q8*hf&FVGQM`!DZ(~Upn(_yL>c#p>pE=GBZqY-P
z4$i{>Y{DYAh&A_Wq*I`T%|<7M736iz#zvb>0Ek099KaIasxq$>&Wd0eoFlMN@h{H8
zf<BMq#zhhcCB^EQN7L2ii7_Hg<37}%rvsN<h|W*J`9Yf$MAJ~F9*szpZ*8)^E`1w4
zbQWDMWGL<ItMgnkSJl#p)u&+jH{+-;1S#i}!~b}6a{hm2bpKr@;$;4x@j8dK#$#~W
zk-Asaju2ReN!xx!2XtC>fW4OZ+lC;7=#N8KFfjAIzTo!LrW@BZM`4xC$f09KL5|!9
z7(eW9_JziluFjSs+(Ls-W}&OX=k~$__wK!F*;})C)Cn;Z4&A+t-o^e{i27%#ef%7Y
z?~MN1h01&T00s;Y3>L(*a%sX&>ubA?_@qV^7f$ETDHmYjJvGfZ=RIZdoxe^`dvo0k
zKd&Fk>!{EcM-W1&$!KBqe6qeU>JCQ<ANd?gx)uk&L*}M}mi8cKq7Ya17O6ilzlLJ)
zeZHTc?DW1!G2s8vJ0F8Eln1*G#=cs@fmF6*_msMhW+~@^_w%TMgO8>sr)>U6U%W8o
z{5n8MO_r~@o2g83S5@>7GcX|l(J0EY>7fFQttW3itI_F|4iN-O27x<aQdrWp9DK0_
zYVS+8?bI>vv-a;?S$w3`mckXMQKp1aJA1QPst&dW^LbR$CuD<v5$H@%fw|M_Dwv~D
zER-@u@@40ejCZsyLzz^NI9ip3PRU4;DUV7%tT%rr%|S-c4hyj^QiPj?!0wMql1z(T
z$N2ib#fNwg3gTCqkvxzjnE_MsVn$}T1ek;2k0fV5>@V@Y$nnYOmXP|)x%PwAPB2fq
zs|kaaey}LhfX(NKt@B;0gbIWwZ{^2KEk$9?)u6G9!olRzNCUkV)h1xUhT|_IDm(O(
zSAgrzP$U;{pk?R<rj(5vp`;b@016}v8NGf*p*G=m8U-!VA+ars_z$f`XH0R}Y2?>K
z%=^aX-YrWX;avqmzSia7?>H2jD*w_p46|S*e8_YE^uL3>?#4}agBQfa9pDU)L>KZ~
zAk3L*`g(UrXAO>4b*?-ryb>HuErQ5)5~L)l&5*5p<B~*)WOAQPV!5_Hxc(pk6A9jx
zZ8G$uC_r`D1gb^K{CJl+XX|XyZb{?CN|@2#VWC|#VDd(P=Hw}!VkQD&$Wd#-*=>MP
z+#T4KX8Ee{J(l$s4=V`YBN9j~yNV3%IwOO@`ZK>6@T^TDZS#TqH!P56#1z^e6Z0Rq
z^ZI~?VY(Ph&Vg}b0l?^<@}zJv%OHdaOLih1l`8<0p~ex{si;wQDx8)<<Fb)Jb9^e5
zMvf{JO0W6DU&Gl6DnnyxqguJ-pnoar`h?ZO?=g#LiT6VaYBK0dbzhKBn2^VD;hAeY
z5wRedtorXf(>Q$M$zy}5jxJbOA+`5l&^?Fcil$OW6^L82DqD4Ogzs>(R9N>3d9$@S
zrj9FAE&J6JaRd^ch{#{9*Jy!72T>p;3s4i#L*CSRC)3`y<`KAZ1-A;SU_Wg10hfy!
zMr5L!%56f3D1t6ZXuEhRx&oE~Vp?dh;lzss0Ggr_rKK7fh<iwpC4)wDI85qDIpf`>
zp-p8D9q!`X-P678{Q)u2+=V+_M-)tyLhew>=;#UD@6Xs88IaSuf~6v_cOqqK$dzaV
zRRrHJ&sUH4C1hDko7}?<2cxE(nrs6s#rhHf$d(lz3vB86F>b_#sonrISxu@-!b%lD
z6Dl?{efO_mQTL3UMHQ=5*mTcqF540!+7L&zvb&=kPOTpN)Da-ZF*RY0520N5@G7EN
zOYyA&#e3@Y`no)Y*qCTKZst#Aw6O`#2njpVhgm8WjHuHd5n+sYjFX!~=wYBE6=`SL
zeDFa6u`<&kS!guIEfyXcbCz=74UH0po#KW^x|d~s_g%9kb4`?NoW7b`xK(vn)Et)n
z-ouNz1P2JX`gF8>WkE1<gjZtZ>?TmygE#BE@k_q+puUP85q!HpgN8#5DJ?}<Lhsie
z)M*i)J=S6C%_UO1OrSc|6|&~&YDop^ofVWAfe<I6>!Ew8?Vt_*15)UW0UG>N+PS}S
zDm|!@DeE(4ENGXVBBAs=!NcybHUt6b2UGtnb<u*5EJDqt3<8?yRJ<3gxk-+m66#f>
zJEfF0hd!`yHKzaMW|<+uVN{@Z#9xkA*mrhCH;hyR!U~elS(e53MR{WG7j{esS>Tc`
zm549)2b_sd8!rwX&sfD=AAT!4;gs6$D**NNoB+Y|#!QzL+|#lQ!tOhitJbf>(|~h8
zrGcoQ@kx-Ew42^HcSR{qjKyqcOMqCxC_+Ju#s@74Sk>hk(1^{`bs%Z|(ndp9QgkoR
zYNI@Pbzmsy8_8>+9Hl)^4e%2Ywl=vHeh!eol3_o(mYNJL$_#l%J1@z<OD)MXy$MuQ
z<_d?IKX)~|IA<qy*?jZ}vG=0}k=E`PGrU*d`?DR>)1s?-wxe9ZOGZImAyo9vDcGLw
zC9YsAh66Lfp5g}J2G2vVr-h?SP`wVD9@D;t0E3xH`Iu}YXQ8c`ai{O9tgV6EQs$Rv
z7M{ukX6FpOklFk4*TOU&y@ubKx&8U+A)f>8TG(HUNtP=Iu!h?PlBxz_8{4<s;)oBw
z6*B+qC2c}&FFteGhdZmuCd>B{G_4c-grXq0*V`Bg;tZjGd_ajROF5^Q40T#}Tm*ga
z<<n3;`9?`B3@zDb<BVS1fl|7|RjIS>zRpq%4qX{-?Zd;jQnD}?aw?(UM%HX&n?(mh
zIyc|xU!ti?MC6#<k|f7PkX@09I+eoSj07UJK2y6`uf^$3!kmP-&GU(-4ONjSJI|3t
z0})!#Wpc{0=_Uky!t0d5sc;260p1u17-!zE<%I2+^v`Tgqv3)(jjF8t5EXMB<q<0y
zeYn>-b=H(9X}g{hwB4vrL57o@!UV7kFBdPId6%dqxLG|9C8QFjZpj~**#o<-ZbHbd
zxs;f+ZWZeoux12Yl)eAu5&%B$YD#*zQ^s4TaxK0-#3+Q$N^Q?7Xq^wAe5u(8(;ee_
zOOO$Dpq?FCce&Dy%_`ZOmr~Dt-mK4x`B>wkjuJgI)|nyF&hKh*iBlh&yLmLSgb2r&
z!Tff#3^l#SZ5C!eC|!d8*FN)eXV~bHmT0-0G>hq=twcS?NVH45BHIj%%L{#`>~O0o
zXgOIG(%lR!Vz*m?hKO<=^PI0K$k?z$>EB%}WPHB?_yvPY-_eec3he|l;H1p<<5F=~
zp%Un_S((l0X>VrPLq#_!JqIcFUz+karwppRlbTHe55JXGLHDbi2q5_zZSt#Cy+V+D
zwZ1`sT7ZM}^vb@HX6#0Eb+YUG@)CH;V`f@5*U`~?m(9e0cU@I^agqaSc4*V4YxW#v
z=}(?a-@D|4uA=zN5D8b+1%)=5`Z3@zT099VEx($9_IISoz<!d-T&f$NF1r1A)VXp~
zDGj2N`-Tg#YYG@-hF5RfDiODV|1ZP3XLK|(^wVz4b%W~yyY6O^U}$!%8(*vP{I0_l
z0P{hJm2kpUDVJ6IY;%6dMU0j3beODmn=@t07}70r&F(5Ei6J2C4j2ia=K@cz<7#p0
z&$MV((OyV`Pv?#5*Mv6)SC6pv+;%AEA4;S`Ms#7gV$Vx99dB#XcJ429{q%L;5zDup
zY7Yx*;rBf`-j@c#I?uwLoq(Qt)5i*xNrSs;OedBWmvbAKZl%eJPo-MVKKuUkyB`*b
z0taWE6(l0$>i#!>$EQ0r7Gh0S&C5yQ6%4`F4a7bwXsM76M8BujR(M~qP?G+myx`s=
zzA^K|te`#S4=zAhgrTAvQCm}2aEs%4PL4N(>kNZ8ppR2Y&KVUMIJK}E?HMni53R$`
zyj@*nud`q9a{d&VqGq?qb)-4!W~*qm1?6uaS-0rpQdAFmG%=kDG9Ktius~VuRJ6Ia
z{czs~MBd+!dw7<esGscX$pXB813OYG;;R4m-0z=uJ~`KLr;B!eFL(mCKlllL)L3ca
z7SLE}5M=D_$`LFb!lAc<-i|22sfPb0x=AmHYV9}=WZ9}xRo<gXusDC&xD12!#1s?+
zs;DV#jrOHoE{5zk!i4Oe`hUvtYz+TlDS+{R$ldSO9Jj?`N9?|&K7o*F9HXN9S>9uQ
z#u%gZw-uBV%t}P~`;8CK&XlHSZYUD2WZl+|Xzub5ZQ=R!OenQOI86<mRf{I4#_?Av
zO}!=G`<?I|a1<IM_p`ePSDK1n%8y^H98$}yM^U`urSY}<H#>*V!hQ=_rf2UwQPfR)
zqe9xEkn#!8^El%=#>`=Bk)3SN<=As~UYoj$XN=xGQQkywC_hr5ZxNA2gFv(!FHV{g
zKaibsa5a=*VK%VIa&lT3sDg_@32Ba_Ro_&!5WLDW5Q3hKZ2rgrtQ&yp)q3vBBA7yA
z;r)VB)FilL_T{X*0{AY~IY)CBJQ``&cqM6<EySr3;3Q7UQ57J&R6eCX3v8&7P41G?
z$_YvyX^9)HE7w!^*5F*<;JPeof$k{aaZK`61A){v4^MpVRCHB~g%pa{#U-r8snw{-
zlisY!Qf`G`OI_X^pZ%z@CJ|I;eKG`7l?~X+P6S#P){nq#<gWz;47x2y8m;r(q85rl
zohzRTS3b+1^)vI61%>kkjWfIDAPa8k>Ar=tA4_`Ax~SA4f6ueczdRd>QyUm0V0foI
zW|-D~wIXrj`1|O(KBD*X_c{bDZ8e^8yy{|zqRAV{W~~}>TZ~CiyqB;<bX`f+z(LAa
zK9K9pq+VDHQ7$xsISG~*wFV)|>ob8Ty9wT-AzezC;5p(P%E9cZ6&AIQa!4Yx<qcUE
z`x!^~s=Q%oM{<TU4mikD#?hBzGWh&S$;kcsYs0;ogh5CtK#O6o<Hb8wp|ViY0VS?<
zBkSQ7L01P|U;$|()#eoLve*tsowGz?@6-!&+M-z&ry82G%!`1WfX3x#vWu;c`-dZL
z{2d&?%ByN-oRE*78h4^RCVz+Y$l3M6>7Zf#lhu2IG=Am7K2kPG=dyhcle3fj&|f?C
z@F+96fM<*OY=jv2*Ci)e1&dkDLT%rV{i6F_CAs>j1U)u@tkNav;MMbfO&_Rq+j>Yw
z;!z5tJzr8BmDNwN)tMQ#N}oo^Q&2_?T)3{ID@J$=Kp1nGU2_bi+)C0G=`2__tF|z^
zph8(4*(@q%L+bF~{6gD|CJhG$etaTU-IYd;=?N8Fr5rwz@E_?k%mu%b{wF+ZcbfH2
z4O#o_{%tx>EfJ+HKPbtlv@4QxQzsS29w-$OHr133ud~sES6*;wz%xUvd%JiZan;pQ
zlUI&zP7UuR-I<nW3<Sijnx>xN#N--mqc1vZ^LV%UqRKwSFAr?zv1hWdIeWq$&b8hA
zqO)j{sBz-jCO6c|yhwv0WIjiwxnxWP9L~n!m)|jfy_)dP2bky!u85~g)VMb0Lz{pL
zfIEgsb_BBrn_eZNf0><Kfxys50G-=QiIhdkMPnCsEE8QsJ42bQ)I>=qH*u$c9L#6Q
zzBr{??4Jn?L^r$_GTy36WH5UYQ9tXf4RGvH9|TTl5Z3BnFV^A3v=Ie~1BZ(&`^z1{
z^8N__E+|dtR=dI4T6%=<X#*%#bu)S2Z#T%FMjHOwhxVg;-{P#V+|Q~fzWB5$#;t=V
z<%TAf@Y3eSkhWNan)xG_%{LmZ+DppmvZbF=AP=o?!Jv!$m--f!av&a=!M=2xDy||S
zr12lFM}=aSjxF@^b}6mZXEMq05Rub%IlZ|<&VQk^Iq<`o_cue+6m9YV9)3^^v))2H
zC4EqSOz<!srv58i3q37aO8u~2O0Xf_L_L+}Htmia&yp?1%c2>uz1^T}j8W}GvS7Yl
zm1>L9phQ#g$_m+1@lEpE4nq+aAnpjMm*i4w?0t)SVG2@UqgmZ6bwro&&72N3F-MH>
ztWjKYCJ}N-S-fi+3to3ZDC50rty1gd1-r80Bx@G}Bg*H;ax6G5Q^Vi)T?}cORdleH
z24b9M2dl~qr2=);xOq3TO}|Pme!v7tLx~{4-Mq?#zU6WO3DXzwCrv#rHp5VnD0X)}
z3v>`Jl&sjg?wqFUWLLnkF4l8ixdIw3l_m8r+{hHlBu#%@5z8<TXmVoZ_;Z0RZrc=s
zQWW9LAdB%hj+Fkb7fzXV*Cp&E^gl6?d)Vtpu=?<tnrJdpHwN0n(JnqiQP={0)I@(Z
z@k-j8j4<8_@O})8LAINHl^GtmjPri=<~9Qqrg09~jY*aO_)eVRDJ8Mk&8_Vn!HY~7
z=8uKbYsg>=D=!Y?F=E^Rh3x!Y#GR>ueh<}HtT?B+NSh<`tKwVku|<R4q>paLM5H`}
zD`(*}v+2hbzvk@1=RW2q${?2Y5~ydxc(-pRoQ#T^R9$tMW?!R0*C4%NRE+sgq0b7M
zUEUsxUz@gOq_5O9*#qYI@5*C3f!g@WaUq%$X|cdU!K!ZEJuqit**4i7+d2jThYU)G
zN8Dg!ZbudR$S)YJzV&6w2H1=9Hn|!&$kh5yiVW}`r<7K&BSM84V(m`B+mLhE;z|ji
zK|oid9WR+s(_Zn*TvAT9&nlS{op&Cse2dr4Zx=H~nGv13@U3*>`;#VQI0Q>mM)<zF
z@fV4}YQ}7_a($sX=huZObykv{8k2w)c@Ebq3+8Bm>EFYZepACg*&}-GjUP1@&{Zf4
z-$E(C2Ss_)nWp74+LY2@NyCA-#Ix;v@W9>*9szfpQA_wqIRqpJ$2^{_r2Cu2q%$wz
z`dQBO^u;1<PHrzvN6q%r;waRDv1g%h0UySPiI`57V1^v#qp`7oPM2PqSoJ>OOF<3J
zak)YqVQX%Xo!5Tx4`b4rs+<wE_Emyf3r`S%<CP~_-1bilit{IM`>Bm7M>{W;69EpG
z!|{`q_7(1fV!`}n-crC}_Wq&o!RAule@rJ#tp9Hl@V}>wOl<!%cKNpEcntPWI+;~F
zg9teR41*zpMj1~E*fO4=AE6HMy&-;o6N)yBk*!B+uDAgo<;|C6O51onRdgKhZwlh^
z@n`Gh@@d22{8uxcugiy*qJ!|fHo<BBpr4YoN-=cxHk3aGU@4u;>@6T=bar^=E5GaJ
z!8Wl7Y4TS1U)6XAtV(*5sGl$1p-h>*Xt#S7MGtZ#Pct9=i@CMEyprgjgUgk)?bRTH
z3ux^Yo=_wZ30(q}n6lrQA6j~6T7P*HfF-dY-Feo$WQYIJd(p?ZmX>LDfpN);W`D~o
zSwLzZev&+9#IsuFZ&6%LE{h4bn|Ig?3GbdYhAC!|oRHrBu8q%6G!UtqXZw5kaCf<Q
z`BT4N7W%2*@9RgS&9o2d?dBf9DNt0wKvat=*uVaL*}q)8J*=?)_{%-4cput?UjGo<
z0z8jTgF%QC!SA7U!$`sz_aUZq|H*VZF#%$u_Wkkdf>Go3rv0>vN=SC5Nc6<$(}7XR
zJjP;xV`rKW*I7gJmptXw!=r@>Y$8wk51Fwyhs4bSVcac6$@8T?2{fE(2^fU(lW0<;
z7csrGpvvqHv}w{<f~p(-1s*+l-{IM90NAm}2{q%HQe3H{8-Q(^+rm5RO3W-)oK8fz
zk?P`@_d_wgu?C{)Dr3$+_L8Z^oE$78-G|$&)0C08o56e(?M))qMf6rQZ@&7L=P=Dc
zpznmIK3mMD(woj*H(@=6dpPAOhKjn_9m;9t%43muItHyp0_>1-XbgU5p%o5rdZyR)
zU;^?lGCK=7$(d?JJ*88>Gka#sUP|gY&)~qu5u$;YK!aHjSXlc{`iXrzyRSNu%|gUK
zxpk0aB%s?Q9tE&{Ww?)>G*8<0B!LEEAEHnZRZtgHYwA#QN>mH>saN?%NE^BU1Xptd
zV?;z#_%~7wP#*8X8qYnby|wIu1h5d;h5c>x9F!+y)max0pji|pAYQF{!SJl)E#1eh
z8F30oqS0+tHH~=AIQi)*==k2im^&m!_<<4<F^heuO8mqjj0rp2c{jAh1WXtgNtZYt
z-qAxv0nQRk_w4kyg%VDhJRi(NjUw!dqK?!Uyrm%s8F_7C`9JTxV9hyw{eZsmyI!e1
z{2I_Xtx=4uu=CC{?T=GiypoOpvN{Wj%A5v`By4327Q!of768I|JC2Z+7^%+S237a*
zLmDa+4)OFdbuC!*wRW-z5YSHNHtlV19K0ckFxebmV!99u8EpbBhmgC?*x}{cWJfUr
zYBk1d0sI<J-nCc>Ql>qS;Jjc<;i;jN#ASQ4A)Q61+n*$q7-V@1VeLuxnM%x==aUe!
z;F@2J1a-|(vC^XXN5v=ADnl$oDvuOOz*AG5Ghn=K=L~Oz`w1+xw8>sy#`t#9W?pQ&
zjkmL<W&?kbvs*qQT$sqdu~u$Y0us`StQl95ouM^b9C$u&o>fhf$7+!Iax?;ktgLPF
z=Xm%D&8VKKJod*&oNAB$@F;LzmS^E2(<pb*oqC|1+KzZh`YS>!owl0gYLwncF9lCA
zUvEsw?i}+RtY6Sx$fHjx_T;oDY&pjspnQQf>Wwke!C3@Ck!JApW7hnYoV#3~b`!tq
z9bi%6WE+-?qPCm3v933MC`$}o&SHIKtyX+c%UIRm(9?)1Hg$=J2ws;QfME`YCw@l(
z;q1V?U5v=Rh>#4eNVbFScf@VPba%FmRDL2+;f$=kkKh%wF{#gnd(n5f9fly-MC696
z&pC8!gUIj>i7vkn$9}Oqv1*Oe<W3llY4{k}2^5rX@8_BGWJuDnhP2!8J_|c3Z_uK#
z4!b;-UL@}usP;TbO6hq%%rt$~4#1AmgJE%*d&}0usQ-u#p8fa85Y?p2Wb_+@dh46d
z*B-&hY1~7kO9g*>7|#zIZS%t4jlQYnFiM9vPMiy>+%D3xJdr_uKtxKp;f!rqMdD~S
z!@(w6R&%Y!CHY|h14qqoQs~{E?1B^5viOZtij<63a{US?uSEUjaOcGZ9;X=k)U?t%
z#HE>@$(uGiMfCHXcrLD#RlVb6DD4Gf0*#Wzqt=LIaTJ?{bHGfE-JY*-XzI54>d@sV
znq-%MT)TyCC^W#uft>DqSEf|pP^hg~Nljrhhlj0PKh`$Ph4M=^%p^kfnB0_ep6z<7
zil<y{I?2@vXTuh6QSe$nUomNvB%(LN`=X2sBBNB^YnJu2(FSLntcN^(MczpRL9=_*
zR5m5g2J()^+Ev3!jb%gTIzo0MtbPbz$B8`7PrTrlhxRc!!nYyhFGUXSd8TO<L6@PB
z@yra%iar`w==3#{SIzPfYb}r_w47T6M`pbA1Jb>;SXZ0qO_za<cRQ|v>%;^-2`|>a
zWG|4n%K((YI=gXCEhB(_QmL(udMz5fm|HPK{!VxrRiu!=H@RU&7II#xi>F!RS9R1F
z4t2&l)b*`?bOcABishDbAS;y}n7d(`{ca#L4eZ8_yrX~B?JbfzF(jIg)ZV0gFI0nG
z*FD)7gHiccjoBPeUr%!XnAisUSO&BEY-1g;y@8tqacg@GtPmgq>j*>05EnA;J`>Mu
zH=b~OC4ee-9s0qsi7IdzRl!!<Aeff&IB33YC<RMPc#*7<H|IKHwBs>hqwWjWgQDUa
zXvsSGre9m~4(zg&WUWQ@H?kS&FV=u50c;_V*&gq*Hg!j5gyoQza2Fni_Vhh=4X0Ky
z)8<W@sF?-?gjet^biO`oF!RjK!tMy%g6ZXY7U#^aYwit3F-yn+(8gMt(^3UhdEV<*
zDqmDILR+#KJf1FpKji^v;5DD^AdC6kpN7oV)@+3ePSBraoKQC2QQ3&*k(qTKQ}IB~
zSWDu>+N;Xj2o$y%*jo0{bH&N)7v??>Bz>nTxjpa#xS0VCcfv{;reg8N<%8%TqazP;
zeUJ7wEL%6?JUrgTAMe6r(*+yQlf1JF76V_|61cd-?HJElFl%9tG{P^J9yt9#f$&Ls
z5EAE2N0pv%l5d!&0uTQGn37oj?-~;m;}7lOf3iS0{C{RR?+`<EV+Y`nAm$@@Ann6*
zv}2TEj`xH={H2J03zrhtM{|~}q0-kXUL?6+Ij=u_%=hQVNu2}ZG!^}!$7dVhcmGAb
zy`{2cW1zGYhHmd%AK!g#bB~%wf>XeP(V4HbowIw2emTV1W8d*QLDy2xLCQY~G;yd@
z@ly68G@2)t6?^_&R^Ipf(al<Up7@;IVBPXGRt}56s4}TX-S!Ky<ef4MP3A{3Xijlg
zcl43D--d6o=70da0{OS^aI0c31P|Nt?bn?g?r&<{!`gb*DWrEWOP_Wv^@MiB6i!59
z6$8=U47tKoR<ba11JlC64lX4m3N%_B1DAyEgmWZ^=031yqkS0fOPfRxN>FC)X>$oj
z(RG_Ygz;B!M;su%C}^t93(bN#qH3ZYSx_d`X&k~HRmy+f1vheAeLfpzLhEIycmaeq
zMj)9?kuF^}%)i|z55~Yd$k170G-8<hJ&2aT0*IDH!%}6Uen;Ib?PiqR!nEI7ks#}@
zt27u<PR^|r#w<V4#fWz0*3!&nSU5GZGNxZqd_2l}z<ih;4te6Byq|ovvV_V`-0L7v
z%3#QeoX1l%95$|zeKIK@7_{gB1Wf6lUamYDm61eJxG8x^fY0LqqXD2f^;o%FvUq3A
zXl3HZgz&@9G^vK*_;J#Ik(OkjIKeHUIB6hm**GomBnSiyV7!yg%=pp-`Y1;%4&rMD
z4DzcbiQ-#`B}p9Wb4ALK^&^~k7aVxPAx|+Z8E5un=^YqI;zc&cB}K4Cva!nDD8j2o
z3Js$mqTxxI;vDmBN#ZCfEK|f$#xMzWBuZF<ohrzyQbLU~B??%=ZF6>VnM%#x>j@MK
zW}|rHmqU%xq636FZ+MHllEPDFjDV93l?4hMCZBPVZ$WHgQ4J>;saV8L&=bgx@`|oX
z=-LhQw6&1k6j2ju7)c@W$FBvFO_f&^z?HY7c)}S?8ufp*h&c?6a7it_bi&E3$e2u+
z+8-nkRybU%22WHfpbp!BQ%U+yL?wceRiuP*ye((~GCG`f#=2)>3~(eqJrtKVCpN3$
z!<=o7IU0A_sD8IZ-w@Tqi7P^{-#I~x#CeI8D%jIcSRbLCVIg~ufwxS8b1o$(cToj$
zuJc%32M*^7=q`2*J6h>cJU-@dGn*EsUHlPYTGXFtJ=r9KCSgfuAF3VUOrDn8^Ng^@
z^w?|@WEy`GfQj{AzQoysw5i{1h#tynzCA>x4o6*2JctHq{$288ufBmw?8Mp+G%TJF
zRVEX59cUw6g%+ZbstyRaVK;}pZ64=-;UU9-Isw&9LOGQDssN7@Q<*>EL!%z|3R>4&
zofccft`CCeJaR}yHF8k`r$}fDa2Qei#1Gp9j>ELXQk4i<C;QQ7PKu>uzv}lWDw}Fm
zjj0r{PA!;(qtawQBTe@9!0P4h=6t0?;q@~yf4+T=vlCO4L}&W#NM)Xn>MgdG?v6+@
zJBcEtZJ@^b0%UA8d&rFG&cdy(3ymi6RqThjR2*a;BI4peSzkPU*N$82(Au-omj|w5
zg+u>k0Y9Dy;h6k#v`lmtIXnN{Q-T{akke63yk059#dgLKHqq&&m-``$P{f-EoJ(#3
zl*+6YNIfJ^_yzlPCRg!x_KNBa$okdx-~c^}Mz{LKWW4K_Y|m>S8`2@W;JYzf?sKCU
zXY660(8KIg1640oP<AgC(ZnNaR=5%Ua#2oq1^e0pp~gn+tic;hFb7DwVxAnaxj88l
z2y97ZN%RHY;2!8UP3m{eVhu-xh%hgY%|<8{8~qiq7J~-3E=)K?4soXnVY~-$8;pB~
z+7{E;Q3hE0Vyl+4U6rdKe1cyWMPIqs5(~N;QB6ybUDo1pR}#ga5|wY3tp7BNB|bTu
zACMp1=N5qVtXfFoIE`lF4~<xQ<xkE6svtTlr`QeWHL*%+766UP8<rkc;wAWH>L15I
zx`ebgmZ(mBnPXbbfIzwn((G2rdjX@)gU&0BTHXP|kO`KmJu0Gu%@eduQtYw$0$Fvc
z74Qz-+hX0s6JV+<y9wQL{{TFiilWDy;*ka^$P*yfii&;jynuTxg(`Rb&W30{1sd1z
zAl@#aKgL@rZiJhk0#)9!+f%AZ2lud)nMt@Kw6_2J*>&IFh@25vXjRR@BsPLjG!Z8r
z>#jM?ek*rc;*d537F-EIYcm&`HsWp3Lsq6Y-HUihDP-?Yzkqs4Q7j%Rm74pHHm>;3
zAXo}YECi%Ya|LLI7ot%&K`lje%Yios5z8tqCiw6{TM~kBFM9RSw#nqBSRz;?b3$59
zA2<KJ%&ucVhQk6JY<CGGmlMHZnr>B@KN0LT7p|(CZ_zkxpk5HEG&R&<Jj-NbC8XQ%
z?I9H|Z-lLZEua1uWA7N;Nx*LV#<p!+6Wg|JV`3*0XJYdo+Y{TiZQJ(AKKtIPQ+3{3
zcXxGl^@r}S-K*B~tlygIMaT_R6&Z<NaYk<Ztwe4bjxP!yFJ&ZdOq2`y0abit2(MDX
zq5IP{%cF(<d0}j=gi2a$E*tqnv1>E7j&u`IW@_WXzq)Bk`_^|RXoN3@7*AXCDD+rw
zhFKG!DtCg=o)TSa$d$DmLf`t^@EIko1!a)73!3aSvaK4)y&qd}#Ie|;6hgzjQ72I!
zu0l85VSfJNpb=yy7vZYBqEhjA6;;30>+w99d?oArIRJnI-{_f{rnwK8P{w^HOn%L^
zj5)eEYw#9Q;_A=4ZFhqp;NbU~xMb`*-JKF+FV#wbKi1irknTkux@s3BXa~QK_<DfT
zIdeKiiQvp^Msn2uRF>&=8vHkL$iaJB(F&R<j?T~*RkME$smBwIXgGnj0cq95mB#K5
zkH4E=il3;M@AH}VVF9LMk%syPS{hV)z;7Y`Q+!Zp#`y3X(;lmt4G*4IgelkXp3$)q
zo9;hqmtX$@7XClgE|WjnWnUbr^PTnu^vcEc5e5r_Y|8S_2FlBi!$Zb(829=uQvMv0
zX*!yX$BC0Tsco>lbTGe-n$yLXwVmzPmJ5KZhlxtt`pWo2@cRvub92F(1xS`u*5O;W
zyYn{#o1R4**>SiF_FH$6!Q)$$p%wUvOYeDo{EhN*1Xqm^`Zfwfca|Tq<G+SE&3?X4
z=eCLSYX<>=TQY77`S<U)oZFfj2Ze8bbTKfNqRql&qaLHT{T#|~!9e4+7AdGBLHX@(
zz?avXSz(4Je0Q%c#)z%1O%ZoJ^kh<9;?l@Wnk{=x8X0R`wr)Z`0Bz4%)iU|8Sf`G?
zj|o_FgQU}HQsy8R%#k@hvBR_g@{2;RL3YYzwx;ejyTPm|l1<go+QZL&nQBu-S-)Q;
zk4XdiBk`uI`&hpq>HZ}}>$q-)0z?MpwDAc|cGpv$`>fotSa-|aN-ke*Tm4k3#fAr4
zG;aN!#?yBf{cl2^rU!XM&X(?TNm2%My$9<m%{q93fUJ!#_NQALAV`iRMbf1KjN<2e
z^tKoidD<Sfe3od?ZFi;`H-wMM4vVqDfMVHk9z?V`g-04*m%>#vuD*jFv1N8rX7r&m
z0k6vvc9B-dBIq+LqTvU<;Tsp<mQVM>5`vkGw8HmHXSB@N2ehw_(3iu&{SG!sOO>8`
zaTr@yWp}#2Qiukw<@Lc;E2Tu{QfyeoapX-WwXD%iLCyGA@0HT9rv2U~B%CNb7qvii
z$Ne=`O%a+U8h7~}43$I+H_<d1%{5SAZ5!CnTA>1n+CTGZi?A{o$h`G4p!aQN5Wn$~
zKJhdQC5#LEse>R`4!TdPx=qIol*U>Nqd(+O7h<?mo)MvonL#pgB7P#s5UBkOjp(=O
z(n<Lg1EXNXT<%b_@P)uK6JAom1QEbX%D(BaMn#*nB47;&xB8}f3IPQ-iI>Xh2=MNs
zRqs=*nwPQd;H=r1*n9FSD9_jy?=Fw+$>3zGhB3Gp;x{Xfm`S3<{l`nBD3D#H|KBcH
z;Us$HtR#jVO+EP$I(5gfgpVX-RG*PsgB?nzm#bnxlgF%wCI+Xm@%&$ph<0V|to?Cd
z%SMx2i8(GY$h@h~6jW-nFxt?<O*~pW*5K;^k!s&Ud>dh_8hCTZ!}u`LuqeQvVu+%@
zp=VTQ6vwUFg{4E+Qe)v=pXKp)dHN^*Na$oV5MT+r^nTQDg(RcbzTf9LNcx1fGCSvj
z3xP^vN<_Tcyy9V8Ld;jtS|&-hAfB76enp;T`U&RC)gWMla;#%ch8jZ1dXiY%(sfsN
zB8Dg!4)zp8=@Ht?4}!ys)+@sX`*lbtV!u#dIxKw}>nccagbq{mIjPkUqn%}$Jrk!B
z+jH%KV;Q&>E(qMHjm>bZ<|Sc0rjkE4N}>nexglwVE>loAxHXBB=f%7Gd%-s%Shv_7
z5uJaRi3bMZ1`%*3-Hb)C6&Mn(z)Q1`NV-b!Nb-be0e`o<F_}+_`L-E(oTuFjwWu5w
ztckEC_zBp!gru-W9>rf8HQ0)B@mqp~Dkq@(9e`cDxrs}b?;vBc5So&KN4c%MsAovn
zO-~SPUL<R&v?jvN4{kV=I&h(mY45&&jRr{@Me#?r^TbOiKZ`)&RNeHHg@l>GG)U}_
zhDNExGeJ4SRThgC(l<fctc8eU1hc8R_55PTPu;pEU=^}2Q~Bll*%h4~_0e;vtP_@u
zU32w}y_o*4hW)BTe{d?-+)Y?IRi_<;*7UZ^=BJqD=2ptVxM{ytvplO%fpOsra&Pz$
z3O_mme`%AUV%nSxK||Si?>A`<&zh78S?$T|W<!JQ=X7+DCLI@E5v1xQc7~)LwngTB
z<Z~~IRNB6~7L^2MP(w{sO;a$mxJcgoV@naY#T!p;kowCdC-pnJUYYihpH-T|qVQu@
zXSo!9wEU}6k|p97)^ULYk3ws1m|VKj;IKem&zRw#$Gb{YSEK-<k*!W-pH7jm;UjbC
z*s6#qpi;p7;#LbsZhW)P8=H(T&a!BhS30))XEiD+CrP{eM$sb$<H(7!xwh%a_*8M-
z+fHbu#Dq^}zq^@x9g9K1>5H-D+Yj?3Z3T`aj>1^^7<b2)a&@oSMVNF$IQs|22)hUW
zvm_RF=qr3*a$#~pgzz~H%-<nNe<${gxi?N>zye<=a8`9PTt~32VGA@vGA@|+LjI`q
zV9SyuB!o6|M6$q+WWRlrSoC0C<k%CR$LtYBDBZ~#X1*;3QFj%M;u9ed7%^pq8{D?i
z;(k>(uaf(G0X00V;ctV^k8OTq*`5`I;c5I@M97k=*xV(7_%6I4i<%92RVFPEVI{~c
zw_crlpGWD#MwT!8dS3eS{W+wd!VJTy%x5{>15_m5NE7+(cmBbk1x7SKS2dD=JM)xg
zS>TBFg*O(%cGqko#Enl+q59TI2K(l`^QxCX>Mlzk1*I5to$t*5F?154#QMercQZQn
z_F7w{BV$*Ct)&QKDjp3PtT&g03reB2Hv%Y7H$Hu#k&zHXXbj^Uf42ctP*BS=!G1n$
zFX_&}OSa&J(dya_Tt6Imd=Ryms-Fo(A&Q6o3kAW2lU<Ap3wrPpyOt1h7d31M6UXEo
zx|ZnG3p54LoMpczvvylRSHoknfP+;_3rgO`Wj^(Hheyy^_CfDx6@pr9&RK~!#MXI0
zbPB-D>?P7Oa3*)Day?CC&i7h0l#6IbP`VA`WWGzOR4?7(<7{b;kdiaCGnswyl8*tb
z3pN*WsZs+8Mtoe%1x=^Y0<3x}s&XEax~sXo!nbqq@cBf&O5UL3Ix78s=hFX@a#cyO
zsu{JAt9vaUIU4<EhyU+6#5p5)*@3TZ`uEUO7<a`t5GGZI@qg4V9BlvBbc&gkmGgh(
zyIlTgmjiaB&gcJ{PKnfEB>b8E)1Jj1rvuRi&IO(hIQlQ~5|K$($I(pJ%o>De+DP@Y
zJyWhBHtO}_J=1@0{#^I!1B3^-GX>!5cOi(|TfW*fz_GwjVGSCFraJUrBWmlUm{3Am
z!xM!iZtG7qe1Z*)!}T9Btt%i*_OmTto#W+xB*8zPI<bodrlwEv;4iNy_d9%hyt)=d
zo)u(cpgCf#<e}J!^F3=PS^{Mp!ia&o9EQ7}tqy5vpNjynO(`6)lZ750J@c!gQ*uwj
z^Cohu91%S!25&`D=4<5eH1fiD1BrF==sX=Ds}vCs!TQ~6&pjT->xm8~se|c)jDxqF
z!j(m-eQ4h9WJ}_P^S2A8l)8`n=2~y>TVH@oLy9r8!7d6&#g|oF)UczC3F={1>8stV
zIpd$4!_1K8J6;~D1IW=V5y3$Orm}Q>r+o9HNU*hc_;-whMuCI6B)KnXK-6Pm$kH!P
z3Ixq78o-1WEgIfzv9xuDl(CL;D-GWqgYUUL8-gn;prsZ`+-EJ8U`~V~I%Ce?Cm@uC
zg*pq#4htWnv-rmeUBivOTD13owaDjFJ&ZyIME=q2C;^fz6Vje=Iy8X?vJWSO)@}Dk
zc`z^VNKusjHZTu+M&u{0qGm`O=te;4DEBaVNl%wLrSeI{)!6mYZdRaJXJgt%TjlhA
zE}>D#Px)|94m5}6vhEpPB>D6HhX3_jM?|w8zT6%>tY;Sc_F+Ts^RDjonCfyAGQ{cY
zew|`VP$WVxE=r|$=jO9!_xz~YZ*lMP3sCMAsLG1k7*|Jf2P);oRp@ylKc#yyI)9dR
z$y9!5J8)O~+hq6SU{6Vs_k2vf{-WmNR}H-9fvVkzyOSNKpHpKW97qv+mXQO|Xhox@
zcTv_knEK9~5-k(DK^uj(03jI#gHUj-Q^;)6KRL(kH;%;z8R%ct6hb5B+VD+G>fQ{Y
zn0hM787+K(B&`CFbr`E`d1sWkI|RtN+>H@y9=XYjUwb`7)DP8(Nh^$Xy0&h!<})_;
znf~1{F8kmvWrLe%#;%V`O7a6+v||2uyPRz~-ppc=_%#oeflmWFjHcU*h7L}zuIRNA
zNQ0mkCkRkY2`qt*D)l<xz4njpf=uQ^SKZ;cK?_owz`f#gEw{2%PE!&xhdxxV)NVar
za!@)}_`GlWZ{)%AZgVM}2@Z;Ghw}YiWvVL+t;?S6fgX`_ni|Uz&-h;QO7g9zjs1by
zDo@mP8nqQq0C?w;?0Z>506Y|s*{bW<q?h#gJqM&1IbcC&O+S#$hfH##q%-MC!c(MK
zo%b{&^i&}t$NXR%&P;riZdry!;rP7kAXjuZ-kxyv4TMLG9dPRVT+*LA@YUI-g$i)7
z**I@Y*{)_T#iYl7m|OSlt&HuQpdqR>c4=W|n9d)mj0*5u%&7(Sh75)L7g3VY@@KbK
zQvCfNcYkcwrVdt@+7$wvLzW8Gi9DRc54a!D*+{D2bCXr5!Q@l3F2T|<Nc2hhjRd!S
z>SYmnUC3zaL0d^8<ZYs>Z}MWlnBy4s^z%AAP4<=qy`cxG9&N5p;_iBuf$>X7X1C3C
zw9etRA~2j3gjD%<2?!R&p?T-IG|!n1jp)so_sOqTMdHc<bg*@S+dTBXfhwXtL?vQw
zy=z@!?o(LWU}W_<4E31x`ZgOyV6VLIZlE?86~jTCEk&L}e2C}@dj#{|(1xmQ?7fEv
zN6G<4(nlWAPl51P_oxa3N$XQf7ynDl;h4T6(GS?)??N~K=|lXB<Ns(N`2Pe3F>x{d
zzrY|CPWJy9UT^US7^M3H3|g*f5j2>CbmlTR_y+}dP#_OVWocL<sMC*_(jgDvP9Uy|
zeAs@5yN*bxqt#fd4x)#%k#Ijw5OLx>o`|!);6E%&SSw^vYEq_PBwk>5v?u0L5Y{V?
zi)tfFY;%(;5pF!6Ri%Hn;8$~yh{I8c=FI-puW)4?BFWYr-KF_{jxI3x{ytl$)I6N0
zBXf=F=BO`Fg&H<V{5GIV&+VL6*~+oxVw}j{2w*K-+qPzN%)`<fwXD$ARDpugH^(Jr
zKM!r>mU?QX&k1hDU0Ng4{-DwxNle<%8DWz`KPNFS<#I(<8DOD3DLiu?=C?u*Gg2lR
z%)R|*#6BCe^uir=8>-)E#uKPYT;?uDr()CTxioqtMYlm=az%nX&`K%Wo_@qI7W@c6
zztP^`qFoL4OhX@24Q3k}L{jsHRiuqTza3&hKj&9mA@1u-*@Y($#X!9EJssC%IxSk-
zqTwxXnZ{ozGtW3)Z|@c}u&i~uB<(+`nB%RxA<bGpEKAW6x>-2JU@F6kc2S}zVRx@g
zgq?0<Z(aL9`?B4{j9eJ?^^=2sdaumv7iq5!t?zpnx+xq&I~rE5eLOueK4(mqlWWzn
zQT%mEVr#GYbOPJ8x^dLCXEs)k_xbPP=q;kUQ+VCWJij;XAGg4%NjSA_A+w85FjnU;
zpRkY+$(!2~Ry<7lr<lIozt#?9PgsL-!juoL>>L|8wr+ozlBLhDhwKX;KZd$?blV*;
zd3*U!RMciLBO3~)1q^&XdH2M{@I2dI-UYs}mna$8)05DS0G<BtPlNlzgvx4rs<o6g
zY9)gj#}3ns*`3~=Z2&U;G>q+T-KrE!3)JcEuh->8j%js9jq$%nkP1`N>L&+(gRc$e
z@>3~mv9aY>{d*xO$3HWNV==c406F|R)!B?)-*2OPVTUypjmZKZ$D1fwpDn@g78Qr9
zXLXPnt-U4V<M+>)KUde>NQ-THgHF%a!`GT=5ps5YyAKe0-JMOv@y)$KObd(S^Cl1W
z-IG^5wf+yTa(yYBxkw{9f7jcK&rupb9zW-&8mrsG*TITIv1`_D%@4EN%e{(Yw(mO^
z&o(c4=5I~zcz`sR!ftRQ`Xu!2$KYNhbd`XPkMmdm>)!b(Yl5|Je}g%kgoF9R@Nley
z?5If8jC!Ifj(FOFs9!UBL~noXKR<LKBG&xNEAhB|l?x7Hg)|c)kv>c0MLwz7$@G!%
zarBYB@U?U#RwsOM7tHF*P*}{C-$FzML2la=nOEy4Mpv+O-Md}vfF`&yeW_{@954Qd
z5AEr&1vhu!kn{ckAOmu25=shs;C|3nl#_wt<<<k)@HOz#p@e`g?;Gx0QK$^TVpC03
zH)(p?xdGWs{b4RxO`sJ7q^!RCvnz2ZqXFRQs7b_~q_qTeL(V{8Oz$08+iT1%uGz(%
zS%U#qI&To&9tSR94IzH0&t*)B=AqgbV>bUbFK0i9ps%Ng^_TBeg^I=YaRGV^fiA7j
zhcB|M?C-jl*H6Z4f==x(@7vAsMFoG~<~Iy|zgrw0`p*wH$GfKwQ%4v=zPz)da{5bG
zF3-!yM*92Fx+IN~k2!e*SdXEK!Nb8*koCL0$<g=4JCy71T}@a4ZuV#+LO!33@`S#!
znxU!`l_xjJ{jZ;ADhN|kBNM+?hhr|EPtqitqCZXWYf=h#D*BSJa(YsN(_n~lgMn&y
zJrD>qI;cOYxi}kc{oHa%8TzA9B}j;&J9z4_Uk;8SRA2apyEjC!Z%q)_5w8$^Br?ka
zytOETgGb~zJyYOO$w;+(<)ka<6MqfV_^$Lo!$w}fWCezk+Cia#704FZlKca576M0Z
z1p>`I-idGSt+@pf0D8CESf9Z%ok0ex_kjryRteZ%^Sec{ZwW|^_u~U&BpvJ+8c(T-
ztA>BD0sYk0ks&`Oo{jU6WGVDj9CTd}CP|oqtUEwE1-^?VXkd4bFp*7rh){*q5;tGw
zd!hDHIZ(vn0z#o?{!on{%cIq-wF0Z-97dqT6v<d}vg*K9Rg?EZEiW%|n^6q0y_jXz
z*pz5t_$6i^n|ztG7YH?&*p$#Rma;m`W;TQuK+xWsYO^lVlna$Q;#*xQNG2_Jo~6#I
zjG;7ualkzVQD91mDR_fu?-so_qAeJNumWDm$IW2JQQeoe8D4rWBm*H&g+G_u+yGK1
zyhCHwoJWc+3BUTXxPmN;LUpqnyBS!DzGe3yhBEKkLrbq#9wzXFVp8+!bs@x^U--TY
z@QGZ@$W#T2UVJZwpxE6f$e6Q^xZB`UZPiM6FH`<?5ce@k3(vv1D8X9t3%z|1ph++U
zXwhRWp|Y>xLWCy}4T6kE7!6fz`dtuf2)2S8ZO;O7!1j&E6Aa@9b&Z8TfP~AO3W|xu
z5J!Ulbpl=)-f$IwPHjcg7>Km2x-STcUC4H^qMev-w@W&(Q#j~ET2>zYBoxs>eqQOr
z-q({T)42|1#9WCQAVFGQ%Y8W%DHN^@7|?wMuBSBS^ARWo5%|3oM5ri8S_~ozO|5_z
zMR9<KZ^5*TOFN@IzF0|B=NZ5eH&TVhjd0Bm5Z+#BiTE6R<hNGm#lH7d9fT`?o)s+y
zHokz2=4ta3M73~3?88saY*B&Ceu)#$o+4E^4p88|ks4xc^<!s}fiNHKp`#WlZ?|{~
zBnrTgQ3UusUHCF$PDAUrk=I*Hn(I0_lPy-$|MaQtMig-xN8Rx-p~u4_i|xOx^fY<)
z!%Phr?j@<Q>IOAgJP}%f1#y=p8-eY4oRw_4V<|ZS0&8J#W!xr4_+$F1N^Y`fNo8wm
z@{b(DkTYm4Y*p%8=7Bgzo5F-;@Zi{Qe8pn6|5!URQPuLsf1F&quy7{CLD8RNHByMN
zfgQ_CtyhlFOV?Mct>PO6J~3l*QPy<wVv~1srqDS`=VhJw$Qt@14LS{2cgB15Qv$Dr
zAy&npMR^jUMcO3g6`TqF4tfC_2WBIl9XsomiO+Vn9W(SyL2{+IDN)_`Q(SZ7{Gf&l
z2m6;|^rr{t&M6i8{YluIEZK4KAYzA@p!BuZJFk?VTm1J1{UmUSa~pqY0o|u@AF+mS
z!O45(Sy+R&nt**-l~nF9JbZU7O(I#f7onPU%)zZb1Q^1)Q7T)dt+LGgyW+U<7p}Dx
zhYSukR4xKYX%NsEQ8ssY3BsP0XUU%UsN{Y=n3vk;d22PnPt014)6^xrohN&87`b5q
zSIv)XZZC}2X{s|{#EwubdV=~Gq`LppFT06|7)`@l2-H=EueYb!x*kwZyHWTO^IK*1
zJK8Hy5)Yp6;h}!$A@bVfE#Ad>Ll0w^&7&I`bGXY<%5J<UIHGJ<m`mm_tA^r^FJ9xA
z58;@Ob<yK_s-s+kSaJd13^0M@cYK`r+~4{>>@hH=HSsS2Q{iFs`foL*);wk!8o{wP
z@VBN?nz{r@t?N7-7n&>2@0dSL>uO=Ib*FYNnT$n#yls>2K+7lQ1usVb`&1fgHuKS`
z?#3U{gkVVth%zHG!Tzt?C8dP225;yPShq0!Wku1;skc$eRT|{9{c4I*P!G?Y)ItpL
zq)?!*Q&6`B=!xq31r+zy2?Q=fRH^Qfl;`+762)OTc2)_uP_V|rmK_{d^-ZVPMEG;%
z4P33iZSX{HgdAzV0$vCCmR@jy?pjJzj%8eR`pAXjl*Oh;)YChTIyWq`>=Y#uaQvBS
zn}mY8$B#nzSy)!WYGmvPN}S#^A*--B1cTyFEp;(Rt;U6*?)Gd!Xd9S8ijYc8ICi@)
zmuCL+aEFuY|29b2xY+(j2Faq<f3NSJ)4oAY#HkI!W5F>c@cp2xn0^XJdZtFhUcM2j
zt2a$+GU9`OR#VC{Wfm(gYstd8A-Q~h#yj%e9VGDToK<vi4%?7}jAZ^aqlnQ2mCAj&
zI9q;j{=8Dq|CKEwBPTIP>Hg^jJ60&gi&dmBID1Vbe`JdZk--k)o&+&4KW-o6;UxI4
zY*D9k*T?-204duPLl4HGZ~6DRzpp3Hr*nA(5RBc2VFS8!`ECti$xUDq#?TwEpoduh
zv_t>%yu)X?S<MxJk#mGm;*jNL4}`eoC4`A$ZX<Y8Zkjo`mV(rW&@Rt5@(WMUA0Ol*
z3m2*F{g!3k6rp&yzz#Abj6vOBW<o%fKp~4sXcI;`398(Z#`E?M+HmM*b+0sar<F2z
zdJ=C9q17=++c_mB4<s;;b_4leYhs00KG3v`*KNgmzh&HJg2^L;V|qpZyXHPL2Y73c
zquZZ;$v{He3%D8xKMFH%RGboZFw@JK&~=S#x3TPsR6+&pFmT;cwtF`rps_$l34-ww
zP4&>R2D_Xn+7|%|@3E^+O!S6@VVoUGy7CpDld&UAK0Q@Kg)WFLwD`{q`Dvo5p408;
z%Tfj^dbgjfkIfXtR!r@}pLb-_0wa5jJDjB7b8;ftBQ;>)ux?BT^EmTy3eJ?hsdy$_
zvOp`uQRhXGhOGzsj_#5uCyW{(*koqZQ5!A_qU0C45ecDnF-gy*FSeA$5qn-EXt0K&
z2*pn^&sgKG-+&GCm#a1A7x^@N<L(0>>@9ArM!1&=pU}kA`Nc+?KKSzvG;C?8VUDB@
zQlnpT132sv0fe*B^?Djm3Co}sh=$FSU3hhx%*@l-_GD&)^k^lw@-?xk2NXdyUg>+H
zyN_^-+6@;h+|<=co*6F<)uqkk*W}X-es0+{xK6gLTXLN>)iR#VdvZ_JW(gfAx`bN>
z6r(hGoxY0c!6dmjsJ^<PLmsiC+fC^7QDPG&db)M6Pm^qehEi@%>dW40O2R1NvE}(u
z04D3^2z^8!6)}km?;Qt4BdG);YZ^N4X7Tb(Bjw9Cm`sf<^N0rR^M9V^<nylTy7)XG
zoh(i3z0I$Z<CAAGwz?MoBq)gRr@RxMB~+9QT)ih(H>Yos(FZdg*JPWh(?x)FPrWl=
z?WB0ygA4rB`+OR#g4D}WSGD#vN4R*5whqFFSV?SF7J~3tQ}qcpDE0?T0hL_kV$rDW
zB)2Hef{3>D)kCg$;S}Mb;>;ClN$z^P@R0e(DHFitPhOYIVJ4^LO96roB?0kP=M1$w
z_Xsn14<LEik#GEm5d-$Z;uYyQYqM#Vg@iflM5c48I6}{)2$p}Z&p;=De?j+TvV<+s
z55W}HFq*bQ+Jwdg*d(`Lk_(!>l-dxALW67q`|fK-hKp|l2G`)JfH;?iU%w>`qGLeW
zGrOHs(=kIAK`Qa@`F@GmmH>t6)3keY&;l$X2du*qG)a>J{6ex4uunj^H>YVJ-8H#d
z7zt*NN+*5xtFcj|Fs*H=x=2kxxERuI2zhKen}40A!^dAHwP`?r4@z)g>@Z`aq0#ah
zH52lxMnjb(EdB8*i7HRq!71LDZz>9IbW0s5up4J{T5F0@Kxn3~Qq;BQIc-`?xpvy$
zAmL;TYF}PtBZ@jTtCX9JP_wfxrW~fS!R}n#kPT;l*Ha>1;S_#vg|yXxx|3ZnDt<&q
z7nT`*bu?eQq{~momO7(cOu0;OnB>j6uM4SdqGN`rUz{3zp|q-+b-b`1XYbUrUd7?E
zTg~WN{o$Ns5k_^0jHg^e>pFA0*h@gjSPn|I-|dK9(5R>Db9$BmrOtIjm|PXDbTwCe
z`aCH+H0y9YyA9vo$Bax|HlQ!XFubpH)d)|cpOFr86m{$@5<Y&Pn6`2j9wbliC;T(;
z9}6HH-WrP^a;P}(<ct7~nUTQ11MJ_Qf8~^D^z*YMLA(+?fx;DavzmW{lqe9Z{wMQ>
zgX#Y;Z~mW*2QxeK|LiBx`BNXc&iUW<kx9hIy^u&?l7P0Jt)>yX+7hw$5`w1h?|=Pa
zRDT=HsAJ$&mE_|4cR6Am#G44dei&?*yN{o{x8ZD@hpU`eTmA^*1pnJa*CqfkkL6=;
z`NRA3S;4DUYqUz*h_(Vgw#Tz_x^r(?`d6INj~Qbo=Gib2_VErnVES6xHsn@&$7PSR
zV(#?EAMZY%-w6@hO7$PXRBihW@u_Ev&#1lYEP?P2P=j>yfg_t-;@TOWIa<fNCiZ3X
zd1>_k7V#Uasf+pPmRJ^{B+b=w(+2Z^($oFfOqCH3h$6mb9Fvi@IQn*xr9a(A^JMfF
zvaA_L#_nCmP#<^U;7N=9R(7sRUky5O>|>~nV-jZ66GDY19d6h}{cS=LP51{^YWTio
z%EY^I@N=zE;LKnPtSfb6<i~PT#ilwBLFyw*I<4!uMlbXX9&e-14i=O8?&Qj$eLl;<
z^0mmT<qpSZ<YZNncZL=FC(w=IgB#dH#<u1oj&2*Xo_lG3h1}_$BoUGUki-hndx@Tr
z;g*^!res<=#ConY3m|_Bjtm(kaPLDR+rp(~k*h@bb(Y#!s8*m<JA7^UeSQ63YFK%^
zY269FULTPN<BZJouz*yV*gn3PKYU*5leOt)w^Y7lkCv8L7$@o5{}FQj0h4vDBcEB$
z1(0@tY9E9np7c}OMl4yVd_f`U2n{Hdxv+t{OUFR_%B<0`T3E5h+RA0-XTer4UX?ej
z=2#q5(9V--*O2Mzm<_X~?*d9jK?c!*N*Y=qh-y_Zw3AZ;mz}#&_dw1N%<6Hdb2`)w
zv`o$aE;X}0;HysLMCHybRa#m`T%R>s+|ksMTx&y{#37@3$7<aw{y7Henns=n(Iq*-
z!vNCHyD_Y{F~#%9FlM%RJdO%)wIngD3H&Eb+jx?s_;l{Yg<XiC;0xf<TL|80<BL)Q
zTO11I@pgG()R@Tk>bft+i+{;vnNE%WD`~D3v@43hMJU~D?2>@yumdAZX+hSw89g)p
zU>!t((-iNG(Hg%U4F=nM+=1SWnqb`EzUMBa;OU&c?H-i<>f2cQ23Ac;-APB^O3=?m
z(^o~WoX&OZHRs);-++S3zNR<$NC_V&w6*(^>akrAI76-PNx1%0`ls^@e?6dE_a{#Z
zSRjw~<B0%T%I$HDnMViq$mio(49+7{!2K=%k?b{HsQ1@_)J(GeZYh}Y0Yfy;eAxy%
z2Vgv$wSXVYm~8$VKZZ4fk!$IIHJf}QbM8LmV)&_;o`YL?N?&MZq!8*_Y=()(QogW0
z{RQzc?EDngR59$RqnX4SncD8VGaScEV=q#)lkfxVp{bcDq+FO`q>8|X>eZbuImjL6
zBVnw*N?ys+tB(5Ueu696qD-d@fDY8KrYCYJgKG97fJU-T#b?@Wg`h~?38F>K7i)W0
za2eVndr3@%>Qiw{U=2xa=1ZBE_oPCG5Q9$Q*16rZMYvM<3`2PRi3{z1_L-?CQ1<Ll
zbVUVT*dh{$5N(GqKty}#C^5WU*`hMstuGG^lA$w6RJl)a9RIORoLeIK>n-qd&oB&X
z4Yo)Ju*r}g4aQcHP7vGYU^8J@c$P}ztU8{Azaf^2L!Nf1V+%;vw)8dtnU~hx44T0m
zJ&b48aNj)9Hly@S0Te9SuzH5&r>AR@^C{S5E;wI_Wh$L;o30tY2iFH8MoffFuO<1V
zPNB;P)Oxk+zzi>!{gywSk6t=HQ(7eS*u@WnK5t=>!m2=S2A~V9Qe-6^k0*HIJr;G}
z?4K#A7ar#uB&%Jnd$z<LXmrn1KA>YjM@LSX)zp`r6<MCqnE6FeT1s8?Hn!dI!v-y^
z?)cCihJL2jD$lX%&Hlj-e62y^F8~%15Mk(s3Y;bHpD>H8u$RQP>6if7xI=0_E00Ri
zp4coaW4T=>X8Q~^3OH5!Bkn!0M#oot7a6ce7V*IJDOKHy^~%qsSh-kPq^@+oq*gSW
zW!AbxO;x`zDCzW&tI!LK#iMnx^KY=QEsj`rTZE2wT|Ym%##xw_xjFi#!6z$0`P)c<
zJz;Nj8w=;w9Not~j#`?dHvl3zk)krh<U3D^1i124-j{d1*syXs*E#RgfT5<bpfU2N
zCxKj;9QXrKGPl&?KmVNm!+ILm&%XWtKK_5YjSqFCVrg69yPj)q2rZ*Um2hX6ZZ4Mn
z1nqyI`Ee?7qgDv%-~LdI$xYgJ?$cyks_lAe`H~SIEs{qi{#h@l!{dkhhpCwZiAL(s
zpe3;e3V}aMYP3H~YG<awhT0MH!wx0gosI|@)v$Uatw}#yYTI)q+}+Yfr#~1*@~Ik#
zK^;%b>R0@-J*wZHE^la75Hb!pHZ3LYqsy<?g#Mnpe;@xU#K~H@&q4eUoL`ef!T;v7
z_{rZNeHZYELe@NVPXUOAD0h`_spm@Stb!*ULOBfIil=tW?wxm<Wq-v&PtSLR&tu_k
zOHDUE%sIXh3_*DUw=tqj{&jG%D)djVxTz0mrsU$CUd~u1s>mwDlv3Xs0n7HAIcMac
zJ-QI&?k()s@5b+xG}=l~1Y(^+K&x8<sj1-2MVp{Lp$>jk%M-q<xyz&5SD#YhlF;*)
zX*;q?7R|@ITvZ{wkhg-cX5pcNRipM_*Bo<h8+NETLuw#1%*%Z1i2$k!Zz`sY<LuAu
zj(D9z#jfcgG~Y)scZV0p!b_CY+KXT<9G50n=uWJNarghyOoD4LLa-ag+R#px3fJHD
z<*igVX7mWAv3@Ln2+zOE-<UGqYfYi3%ABZB>J_Fa!@bdBV`hN~tK0V<bt*I~Jv8NB
zu8)fYD9bgW(m^Lun;T64kB?PP71Vv4$s>`>Td%8mgO)*q4D|d4wP=vlx~!hB{VUiF
zVr(-+6W}r`C*U3Yn-ye~4lI<%Ye*<X2(g9Wh!?`kOHu1D-TWPd7}8TDG7QOEC20Xg
z&<~a7n21HYpKT;NVg$C}f)37%2V$y$8tG&T1{JY)#a0_^LP^tXUPohBGzYYv%%s?W
zo-X%}ZIcr>x7yh4<EYUfx|}8Wf!7S;6j$CMBan4e?~bP}SB8nRQdYp<FSDg=k4NZ(
z){Hy-t@)*)4R+QO7E=FcXp>TEi8WZeI3F?Du5uYarK2Q@`fZ8l$i}Q<ndXqQ{7=%(
zvBqvh&meU^>jW_6?F4{s$ew1S3nQMQNmr@$V6SSYOrfqPJ$u%JGF_r`pe1OR&4*BB
z-n4P~oMz=rDe9jx8m*!@rmxs8Cb()0JTiFiDrg%t#h$EG7z)qVm1&!k?wKvOQuB<E
zZoUV+lsU$uHi>Q<>3n6|QGs^67A{i?mx&CUrWU<UmHIye=j<dsmh>#x6yvF1+zB2P
zCay8Hvn}T+q;whfFyZoNuceJUOa-DMi?z<Xm`<og;00{XV?7&M<uJ(!EltAQ%i>u$
zKW>OyuBvk1*BB6$u~e0}pT*vG1w!?;5Jt=QQMB5QSGFzsVpnm=l!J(zXIK)otG>L3
zdA@w+n5(&)lv8`Qv<@@`0`2<k!oc(Enl_C&e0CBWTq#7<ok7c4!fpr)IC{ZEJ}&;z
zF@MTy#+P@8kL>aC+q;+fPHBE&VPS!+@<icaA_1y;s#=;sZ>w#7S<1EUd(-~EY`+Ax
zAglwGR6Vhg25X`;r?)LN8*!aD1#K>o1vQCJYTD;fMbARBz@EZ|&bax~5N=3(m95fs
zB@@^BET#pT{P22hL8z&y`2I>rx#;;4*5?duR_?Tn-jXz2#Zw@VHtcM0v|*>1X%3#l
zv;d(M`uh0{F4fP{X?C7}vL{A!<6h}BzQmTHivR_q8IXc7T<QtkAC9B6nIh`SdiHg)
zs`!Px30*xD-{E&V+YrYea1Hl-RUR+otYE5T;zAORr0Sf*bL9+ldI9nxB{bAcqU~PJ
zAT+J`?aaA`g3P3$XZd^S*Grnpx$q6dsXLivjAJm@AApC*Rsr%%;*S2Tl6z)&smx1p
zP(wB+(S9z^ZQ!Y82sLAb+@~c4RjrO~Qr(sLZFvsETUOl1Q%rcUWm@fwa}zOQLK)T7
zo05J`>F&-{e61`Gmy_U{pTf<QMyrx8lI5PD5IJdSn~xMDY;>e;a006;bg=tDr8&E2
zS{r2fi|ND7{%2IyDfo+MZ$%x3A!{d4P*T9D-vVU`X2K0f5tex;+;|tRCz$dm^fQS9
ztJlcyT_+%4k4B!NvCySML=i%n{89OVaDL_KJVdw?)N559*X&=9otwzFRjk1xoam=l
zJ=EuEhxw_%>~l(CvH_P8-~bpOf*xw)1Zi;&Cq+1RI~c^`!OttN(Wh<)k_^RAYOWZ7
z5F7P7Ls!yX{E-JqOQ`;CwAX~`o(vSl>lDb}Gsu9T4cy0g7q~|zYoLE40~mkg-Zsi(
zIW49ljF!f)T0#VG$IgV-tj`M!HmzyhteWF1lg3WTtDEc)qfLN<bd62)E7AspQslnD
ziX~X5hyC-j7BgL=$&0<WOZGXPrzFY@ODpS0ib;iGJzwAolersmy#iePD%#-}uX-E#
zDdYBiA3M0H<E{tn`h8kPV(>V^ifzpp7H9Ltu<wr0stI#f*W_d1|5DH14fib+U9_ii
z*8#>1<=+J_1pHjKw*>UR__8|$DS#EeUoHo?CGtk(H=sY7NDi~QzwaN;#|^y8spp)f
zArwwy_ilq_oR2NI)s8u#M;q14(Urp}a;8_4f8z1S?4s;KNOBQM=1||sfA6z$)Ossi
z!YMxKmKn^;%WTh4(&HT}u(|vs{!NK>6pdS_B2_Ooc}$7ujhc06oYdh~I5YOoKe+5<
zEV=1yXO2-&!Nt1KdMm+e)#4A9V=9vC^IKX_&(vprgHdPjJnqF~OOJ83bN!iqW(Sxp
zjeT9}Y7nMjwwMOa1?o3#!Nn@FsoskyD3!ibCsZJ;9beaH01M*qbyUUFKmno0<K&cK
zkBI{QWoG1cw0%IMW%4vGn(90cyJqr*9cZ+U$1Al~!z(Ftwai}sZFoDdI1@fP?szEc
zt3{VQ#;I|Rdx~=aExIZOfWD9sU0SU2Pd5D}Q{}Zhc_s8xw!_Bk7K)@t6(dbo&7iRt
z2HR~jA14N@Bf<N;qIwOi70EMcNg=x`<^KZ#UCSNe(*+B+#rk_jVeZ^fHvHsH)Cz^-
zA_)Rq1j|%IZb7&4R&09At?&~ONcvjzvGcfh>H3cD&E==qgwlm2EdKZGJd3j8)YbG6
zdv;2v4KDs}Hwu1lbLgT)feFVr>TGQu?>a{X^4-<QmXxBl>3P?3jCP+1apxa9nPWd$
z`?8{}ZerS~Q?KVcd?pV1fi5c*mbK0-ohtXjn9jQwkcRh$8?K1#Kb?u1^knIye+ZV?
zB>P89pQbY9ZXD$bOytkt<p}dj=|u44)W}N`!hkJURUHijGZhfjzAQLEE-`~?jJQyj
zqW?V$5LAO{Az*nJ=WQv$Z6Z=#nssc#58xjT-tl7GZ?20Z?t9FwE9t_hK})>9+AUsn
zxL!`T)w+R6pv|ja9Ma}Z;Th!fWK@qSxbxO1;+sQhGzx`Y&vK9<;+nXZuQY%3wq)f?
zCLJu@!K%`Ao(61;V;%x+q;bvL==C_@QRp{}Bgk0E?G&}v@AaKj(DGw>X;$iFFuTHx
zQnsutyNSX$&ml}<uyZmxzB`@$!@o|3B0ENMav5v!;@8}?h&$d`nqbFvGd~`gW*7pj
zfsB^6Qgt68K1T$9>*ZJ@m^1RI&l5{qoV>+c@&zs`^9kfaNgoyZnG0G`^8+0S^GIvw
zmbC0laxI>2G4QaLf$|kRD--jfxGB3E6E<_DUp77_8AJ`utAB;}+y+i_=n6#&MZHo)
zexH`1STm8jYlI4fB-`6J5~!~=zc%X6#6d#Xb2IsDxZ=f~JpIj{NAU^VZvf-?DE56F
z-!#XpS+(Jo`?LZSqZ)5oaKFDk#f`hL)_BXl-zD@Ho~UtOO}i3`l+B}Nl~6o`Etiej
zams(5rM6EZ@Ym0?;k?ba$ddN7^|N6(dyeR3Xf0<I`MEY%u(=n9+FlCQOHj5=KgXti
z1A*+<L;pu9&-@=qXH1;T|05~?Lpu9`{r~s2W(Khq9m5}^-1SWYLwnQ*JEjt$1QmqD
zZvk0HOzO1+J6D7|5*5?hv)e`5IiEM?+tVLhacl{rg<_%+Mo9ep@UzqmB0wL!Bol{{
z6l|WTI5uZ~8*-|s_&k|mkPApJlwZedc%KzYhCWz6m6IQr9HFs__mKV1E%oW${Hb0c
zGF{_ORCasYrcHj>pa08eZ`5T9@I?5j^L_cbf7)Hbf(#AUn3*&?fyMgyen0x13&vu%
zLB_=d?e_L)_xAOmv2BMPHP3nr{`zn}ys-J2;d9FQnqU*X@}g`VYDpQL0_+O#Dcj<d
z$Y|uAcg>IdAQNL9XSARV<;yn*s=h`x*q@j6Y`^;%!?cPVrDQi4tEYESj^gr2+knC;
zK37q<;F*TsOz!JU24C=?8S{4R!$oOxCg7FGwvE8K+!7eOR`wlOI&q>5a0*4(J0_jB
zB83+W(+X_!!dl_68t;k)^jk^7PGHYAAFX;rsrW+gv%4Q{jgKgYQZu>gws~Z$X1Kh?
zJb-5G*B|*+?<mt=UW>bXH1B#xS^q9@pSFo&68>ONXf0W&aCdNOzaHO4C3zHp6Z_rh
zkkUS;B#1KxDcRu)`vuVqWU*8L!=^ICVAN*5VK&ZdA6(w+HNgEBE@XX^8&9)&%An#v
zOth?ZBe7$RfRGsZGA%c|jexOMlA^Zh7sZN~Wb7)jqKH1{vkt8g3+08l4LO!L24e>~
zD$dp#_=QP;l~PbP{p-W(bX1O82DFDQ)CMQ&98gk#gweC&Q|opj$4NQ(-Em)3Y(EzL
zDRzZhu;Vi66zIM{JC9R7)LO;)tFpYjRISY?lpwWmV4c+RKKY*nWiU$2{qrPf)jRuq
z7Q*E?Gtoz{7h4<L8dzo^4)ecUP4fjGZe^ew%!e3ACwH<|^nb8*WydL7NsuJfDQ={q
zOgP-Ddzfng`~x&qOY;;^C?8l-u_5Q%ChUKn*L{N%kZm#fl}!j|QR2u9y$Ek781H^}
zlMlt&8;IM`mGv-ULzBt`{(6?71tDNXoACTFcph0&vcceDE<WHZ>?s)mR!z~ee>#|L
z8pb~mT;7ksJJJ4j<N7zbMhEHTU>66PtTvn!aSfc12gfN^Mu<m5dzLy%4>EE7SX77A
zyv3M2-`;4i0M*17Y7A);(FPL&i6u_t%rxvf`nN>2`SlrTHLzK)?i>q}2TgxOt@Jm~
z+=Jx9=qRhc911YaE2b^<I{|Kj4!YVZ9xflsPOT>SV5sUh!;!!iO8!*OoZTKmg~%D@
z*x!_BzgO*@<_(*19aYIXp$ab0khFt0Zl<}k{yF<BklrYC_!NT59W!yIhvX5{{c;(R
zRHk*_k@G%VVAIn&Bn0&enN;qH_c?oVs!2FUOP=r4`X`>4p=X7Z0`7*16zZp?D$(B=
zGtrS%qSO1<W|Xzwi0z@6lGOVv*NVnvscRm*C{d+VkYKe0Q4>$@4oR+wD94f9y4<4+
zbFfmtW{jq+tU0}Z^3AeC)2Yj#1R7SANJWz_9E~b?`va5m=NkX{{*7c%-@y(qMS)*U
zM<(X=bjCAKR_)7|dy;HW_i~ssBUahjyFyPOwot4x^rBsqN#W8f#+z1y8AJ=HJ0C>#
z%pfE%+ggQ<X>yDX+m66Yk{*8ztdi_qf(_GOuUN3Mx(|lMHmQDf-HHoTzBVRXtg5Y(
z)E{G$L$HTl=MFkdsi6JU497}cLJ`CaNyS*$<nKoktaRfbsbDY#hCV)n4PVetZUBC+
zK5+EVki-uxF`avnr)Ii;;EHZw=3}3Z_Fg<5nnu$4u+hLjX_J;L&LmH6Z}VYs$St)P
zzt>H>{Q8i4>6Ws6H0u4-#Vu1!PQ|$W{?z}G2%Go}i5gATXtF8Jh@X=@5aY`7XI*)^
zc|od5LVnEsM%eX~i@q^!o`2x_PBC`<Y*b~uSi0Qh&vBv%&bMlFrHHqhO|rEaoy<6E
zQ5chkrw>LL?TVuBqDpo!cPh)dQ=qzbI=drfq%0nnj_Q?2vihMEAFaBY&>6{`petHh
zw$rCdYxz&baris<^v(it_GlQdV+3-7VzB<}lIYjMV#0Bn>zs2rFDt<(jRv(bBK5oH
zO=@yUoF~f(<+iNY0O^!d<*3Eq1}uM6y2V^H2Knd@gDe&vM}GQVHs`vIjK4`p4N0y7
z%h3;xqX95dMP(t&)5~%~<CVWYRd^h4Xi8qAlwa?0_Q{|<V{DP|^<1<N0__&mgVuB&
zRF#a=gUV&EgFn==*qeJ1jGnYI<ebLjj{@kyjo=pz%oU;+gej~G!L5~nUCTwH&zn7O
zo7xnJvI;_s+N69_+)(<M$JJ`GkkTEE-7m>3Ke=s1vUo!xJF6en-s8R_?FPOrp+v`+
zjyHT<YoZC*qaJPDy*P97RYMCvT4g5PoyUnIq$BmVeOqSS7`7nZL!Su|VNB96@^MXd
zfgIL$r|O!VN8kG)(_bH&>)Bqp!CVVM5}YvHVPT&~1hLy6efB|ou3yeEJ^ledcyL}v
zg!#~|BRHK~)qYnHUZ;j^kIS@^JwA~8l!vjrPKAnKe@~qwxL|0G*{-M<%tIlhx$M2g
zWB$D=ozd4OHjvKoyj5z@r2-PP&oAD%gK!#Cm=G21QsHl0zZh*&KG?8crzn^ApIdxm
z8sF^Cr%m>1yw0zP#B{A_iWDCqR|ZTUU)Y)d(d-2}+ZJY&En6@w-+a&$$;&Gj&9skG
zNV0Fx720ybS?ErrdL`5&5k&cqdV}>pScjZHk&*w2bvTLpLso5l+%ixSAj%l`)&&v@
zUm(REnGCfTLI5RoLr{-P$Gdy?)Fp^CrPrf(g)qt8V8D2Nd5UfT_{loG`FMHp`Fb%k
zaqTK`@pby;p!vo<G$*iGIyy;1+blf1v439bOWU1#=0g^;g%U`7KRrYux{G^e<k@;I
z^k;#(3QR9rWScP6%mrwq8mPJ5u+4i?lJc&}Wt}o3`ul&LOJOVsbb2&~HM<f)@eE})
zr3!iw?ZSf1TD~=pUT5SM&}>7&lGzd+yc*u8MQmsO`t289Xat6v`khcBB`ImotcEL0
z6AausjERfAPg2nbngO+Jk+k>Z*<RNv9VHiq)GMw(kTc!Cet&-2ta2F64)T+1ulzkF
zx4k8Q3rdBy6e4AVj3(RtZSv;##v)8VLExXbMfN3QgUs_~LyPoj+6(o!=i}P5d5F_1
zn_*gnZs5OgWpAIuY*9X+x~I#9$P_2Yq>nN_+*#$c;lGRkgpF8G_;e&v+937e6Jirq
zI~^2l<+HdcBX@~EG_rE>>#Jdy!MALQ{L7+WFS%;*(uL6aFfj?s<mE0Vn2Na-yr_8n
z;{{@u%XZ2-^msdxy+T?Jw}X;sYm6_-6D9`)UW_1XG;(w<!GRLHk_V^<{;R@|GM;H~
z!+P4u8X!g3&+V1p?=d2Y{XiVT$&A52)kagfkHB0OGzILS>sgGfx}kX!iBn#guGz$t
zlCC+G)QIlT8E2QILJjp6Auyy|oa>Is-E>rC&bG8AeIf!UCo;*~Vl`7YiQCkE+2Zvf
z*2ax7UBLtnGp2;~Sf6dEe>Ly%TBWo|E%#y>@0crnLoASC`II}v>>>oE4BwDieP1>d
z*&_QqSV`!o7xPYrp#1?-8587ZuHAZ%{P6<EvW9!H{G!21f&jvRr533Tj_lv8mYqcg
zxP?<`HGW(q#&BrRB9wc^ic$zNU$Lh!I61lpIV^go@iuel{oJSKj38H^)G*R>5V8J>
z5rG5V7K@G4FihL^Djn?jN?R~R!;|iH$moKaor2F4uTY*cv|I?LYMuPu`DulXoq;6y
zRH)?1xay*I#_)kCkuHBbZghixci{+_1kNzv^)TY_(x<X;)xtI6N>vk${3>D*oaxk=
zIWTv~DY(`~uMajK&D#O7@X4O+F5gwzXOK;>I+}U63edbC4e>FaN{SLK;q(qhn8|k0
zh>olL5F~($VX53P=UI6)8ly5ADMPFv)7Jzi=`m)U3Qvqo7I`{&B2Ey6PqVrGjTWO6
zTk*D$)KNRY?i>iHI-F%sp^%ZTruAzy@r}#VEk|Ton3)!C%|ypN1y_57^?TB{VEu`h
z*c9pf&VvpnXI<GKWILR0`SZ`^Fq2bLBta1ts%NPsPIzqbtTe}2wxPX6`Hd#Ym9Pgr
zT>Cey9<7gY6vJXH=r1)Bw?~uuFA%oAHElUuAE1_;`Y-?u-vNvoI{aiZ)xwzTsm!a)
z0O>*mnzfi$B3iNwCY3QF+Qs{>$tb&`zF4)#pN&(rd&A}~x57-=4=)G4rN?NFKd>1J
zC`n3$-Y<at=+y|JUHmwV#BPx(?3{=koe7A-;A!}Jl?6I&_j>Dz1jz7?EBCe#G-|78
z-$r1k#^?kGp|ZSR|F#dKqAQLH=V!>C*r(*hG^Bn=WzM&soH|0Gt7mfh%dU<?ytQ7_
zZC!_ho(vS`yesGDtLGdKY$)qbc4*Uvz2zoGK!3_Xby>XP+1+H9lK)22(t@pPPamx{
z#i?&rS_%8n*_d9hX?iB3GC>rr%urD@_b*k4EVZ<PRz}H5R}F2(LPFcwN>8>uLTWn+
zneBIj;`1yXJw>D9D}vcUjPb9aFxApyLZmt0$C@2x1U6wwv9UCy;_FxHG2^*<Y+f9M
zQvJDP#~A(Owi-^xMd(a&eF(^YDZUPC0&FiMpa(xkZc01WQgg95$T$(G<2N|q<6h0t
zsS98%V_cF4<7ZgSeb%}#wb|lhbaas57v`n`zJT)JH=&oA@h3Q!z_z9g_OHtFDh_>f
zwwB4>9d8)ZaW$GGVD+n(d5_f@O%`)F)c_Xw_>HSCf=l2M+JO7r0NN^3-Wm^nf?7aN
z64A<gE}u6W14Z^xF5gpq>>8;go`$|jY@^L{-(c^c2$0C%lYb@)*!yfHW9n6B8Xx)V
zs$7%joY~zqqVp`xjs3hDi*u>ZzJ5&qi?MSG&MfSrblkC%j&0kvjgHMPwr$(CZQD*d
zwrzAgnW?F%shYp0{@ZhL&ds~)?04_Ao;CJn)<K^d8y=NvNZ3{x#RO(hZi!DYq#q2a
zqq~+V<q16m)i<@A0n&wPvSp}KS><CPMoZkmYYkh3Jq}J=u)fkdK9dn{>X;z^u#$0+
z9hv^h*GN+?{u_ih^0i!2HOJ=n>bx>qjaUXU?$Kf4@>mTHKp*A$g?d@o0?7Wa5?Wn^
zpr<KUs3hOV?N*njSqPsZ2J=IS!`~-3)73QGJP=;brjr%XCmhczFjBS=-^{RyS!JT+
zBEx{uvtDH^bSdH?yj<0Z;fF~5Q~9LDv&CLtukXoLz2rQT?{MWKFcNOGkH-`}#^*O3
z$i>-v8*snkBPp<ndwrrkOh^|0-^>2Ci<B*}{c)ig;=Ut_5ed46%<NB(yhkw)5YvbU
zaRm1iH!qF2##P6($g*hB|IqW3n(vaAA1eM~<vwSZE>!r+mEJ6#N$*i<UL|-e((yS{
zpUq;aTi6~&S#nMaha8G?QWz7N2auN6z`HD^Fv^N-tn#R62JLF6ovS75<5`3q6M_sQ
zx`p|U-ocT<1}%)?K!A^2$IRK^v?fwzO>bs@e_mD7ybDJ1vG-z5Wits<1UQLfQA=zQ
zS2=|UC+Jwu9iQ=JV*g?#QcdmSt_PS`1Av6IeA2O!kS$!9t@UCWgiUT7(s7Nrm>LB~
z|CrV{7*xvZP=s+?HZSnED^GNi+1XV~4nDC#z7xv*wdz&Un+j%0xCxKx7C#6~#fu_B
zK-vU3KhHmiq9Y$ad*HdeFLhEST+HF7YJUtJ9S9|~h}rHD&_9{2l|3{zeFJ~UXgVR&
zsUosnhEH)#v|j3@s=0bOAvWMG32siqn1xY~jYLA&CY%#Ob^uKf?&&cMjh?Ap-Lq0z
zga2lIYLqJL0vI0UfbNZ*E3Ia|FFdRrFwJJjdo85phr#Q3>1_^E61SYNzDy=ty?In(
z{edOPUaZ#6sT)vy#4`Vvx28Iihfa0Mn!CU*-4?a?E_}Lv+0ZHcH%n2ue9raeB~wQX
z?3P&WhP&aUapw+}_u?ZuX8uIBhGTjA(hGSFW<QeOV?`9^C;+Il&D@|Is4)A5Uh@#e
zIO&y!v%Svd5YX;b@6$4sO5X$zl6|3RpP8zQkj+XBGZ2VP_$|1?WoQr26xDgN6~J((
zIT?WBo8gH|CMC6phKd;&aEq_5Z2;u>B<*o-ZhKUob{Bc<;i#ENHMlo){C2@HD5iO>
zJB&HMekohqnd|IZKMZmfG<>U=s!l(n-Lfsa_!GkG=3TlUctCh<9~eEFV(#gdD!Qp{
z@6J!FsXR6mpJL+sfmO@i!EvO>kw(w@wvK&G*n5H;6|g;2V`XK5pTS3a`?TnIrgfj6
z8_LJCQv`)7%sJ1caBJV{O2ba-UN~z1QVp0}$38xEKRqIT2(3@AC8bH=I4%Mth;PI!
zPmp{fU~#UHxiqpF9NPWtR*@a!9+J0kokq8bsnhK%gH!i$YFlodt6Obcb!|l%Y=-s{
zDyt<|R2A=W%h!kH3dMju`Uj8wP8YmMv7?RS`UFpWqhMHfx|6&ENR6k$roQj^nw3-_
z-(ojQOQaT7Gj&UAUPzBM^L4lF9-5;{9-6#o%7cmdTsV3j`2FEpb12uo@%u65Q2%2i
z&iWr9!fgL#%ICC}bS&}z-r-|Zi*y4I3T!a*3fO2&{owJ&t}PJOepdKTDHN*hn^$;K
z`RgvpQv1?};t|vGzB&KS-Y1_%Cv9z&vS}PO|8RW%J0Q6Q9_<Y{kS&x~gETGrd2;bs
z%Y34L+p%BFk%~+oxr6I-KKYvxCkByBZ~uOn%HGCkPi;gZdmBV=dnT8tcGs@RIVJSf
zj@5opTIcKY0H2;GaVDL85AZLqx98LCbw}C+jAU^nH!2P`F8}-E2^o(fKfT^<gQ4;V
zp8oPUI>~=lWI}n5S$qs86^&i^M8L`*0|;%DBKwg9DNFQ<ZL{2$C>lx~?XdAi_71ux
z0r!gWT|_~Ti$h04lmt;TTlPt}U@G{XJpeL958#50lE#fX{|+)=-7{zAUPvpT`D)SE
zfGL;Md}gdb1N1hy{3HLl%d!N$+ka^lsJ0ggme@VNo!s!tAI{wDPy_p^QSc&y46WE9
zISQaTjtrvaEfx*pW<9pD&#Fs-vVhB{#L<n@++!ZG)3g@*#-%_}m-qFME-ueO(PjpD
zR4aWWBB&h%s3<rPetP1#H8G1HGkzFM??p^;Q>pEW0ODDvQ4;M3_GTFJ%_fv59!vg&
z2A`8GFxMqj&{n4UBh~|+xon&3!mSsM8)f)8bc^B@?yNR`%>?DRdDp)Fst?E}d11cX
zVWuug-f0GAAkvp1VeSUNm_tY{r1LfWVhGWIOZ39u1nciP68XIaPMQQY+Z4cco^n(d
zk(8-VG^&4vnrI1xg5gFWlG6sT#DEp(&ymin>SY@+hvlmAizI+Lnyh6D<VIz?@H8+}
z??W!Xc=5r)E+^`&zo%h!RLhS0V`>exss0mNjP_=0Hhzk(wsX&F$+mGzV+%9I)$xoX
zC=3Pz#ST{V5><YXL`(vuU#7X4Vl5d;xY8!+FwAy&1k<tLs7d!n^23H(DUoNgtYXjZ
z*UXx0*;K$pRXfBg(Q8SR9P_oMxFhD~%<R%k062-eeAQoqWFz=I@rJf`muQ6eieKV8
zUUY`CH0L_#0@P9Xmcv4G>URulx&AKj0SKd$lC+8BIoDrb-ih{gWwX0SBTt9GBc%Q4
zc^T1aUh^%Ith~%237$4>Z6x7q;kV*C3k+|cbdpTOu-R9MQD1&LKvAM)^Woj^5O@Sk
zf8O1za9Him&<-FO&}p0?<erKXD2H$8j?XOMqMkhSvG-d4ZgB1reQN>jdIAaNo3{hb
z27qme!EKa8cRuuEIxJYU_t4p$nyKt5AvjuW#dgzoZ3?{br>vDs{h;`#L-OFu7tBC(
zKnM~926Re%EJir3o*5`(Y_fz?JfayJBv`=rScpPL?=ReE)CpHz!VaOn7o1g7T4%^?
zh676=p_bS>9!WTFkKI<<?5Aj#r54Pq7Gy+@arp~wx(IQO&8k02MiXSkW+-MJWrJ<)
zbkjZD11>~V#e!fHp^CV93QuuN19pWt<s~y6E2xN+)vJ1q*@X*QXD*<jVJ^&KUzTU%
z$`Sw-G5Rko4aU_qQz?D3q}2x=lv@u&yh2m9Ca)(z+RIj%J$Cv`qwd5>t4rgBcWq1E
z_A-BjBf}z$Wzk|IQqV-pV>p#V)dDXootL^Oho&lUUwBl4Z0Pc|2c}z`n9u|`w~yGM
zLE-B?jSHoy)1C#wk6Q=tkcDZ;RyOX;3+*0r(kOwGZ~sgU{plz5et-h6nLZ@}6C)Q?
zr;!6~eh}D9O7YJ+_ORU)nTy#@oI&3&i8w*%CZ2G@vlFo(*|R0*aybzFeCj%-=r$t*
z!!xATQcEi|bX<5{Zw`y(Aj>j}Ul@{FsY)p>1w-|(RBZSc3Vi<T)3uy}Y}CxPG@7Lb
zt_fWh6BZ*j28ba>yDlni&3|K=jtz*o1+l>l+0?21H!Y#@yR<Sx6W3~W#+xgy;P--d
z)_(O%IdAjg^EQxPngz8A$^0NW7b?y`o4wQUWfuASZbClHxsLHtM6ls6Ha%P(nlS<)
z^HR?sL-I{y1Y%UJ_|b;@^y<+FwZKj1kT8VeERK()rZ+iE_@BrEq0yk^%b?_iY~HFS
z!mH4`I5cS<Cr03EsGcUAGhvT45OhH~sry-BxdDmeTlD<B-{Nv<p_nq(kn_Nt!)X<t
zNs-U;L}=}G+d!VC9Yd3N8SlDC?T@{!20|oY#ANwCa@{~z{OG0SEJVF5Bove}1(AF&
zx*63^GS6Msrbjvnb1^bGW&`N;8Mx!G%}NgA{+QHZLrOw9#f|AssUQyV8&qeyIFh)b
zm~awSMTgQ`nRE3I`cSeIvbrBb&>3neG|Sn`m#I@=j{Atk@)KHfTK^<jlv(pBktnvg
z(v~D}*-#W_wo?}+$kTsuVb-$xm!nEdHSoVig`9Lq$sIK_-;C!qpUqAzhc#d^kWIjM
zgLI&j3!g%ke&HlEYVfT&a+sh{p!#E8n8tC=Px^^r?~qV9ZA}C2;EQ!em2um!xg5Dn
zw60*bLO9r0@$_`A>$t_W)5f)=E-$pIzXE&BHKc3<hGt+tDzNY_%RtT2_Yz!|fEnnQ
z7PZd%h_M`a>#PcRh(=)-OHMt?^OL}0#DC~k9}%8H2L-C7e;ESeD3l%c4hh92o@klx
zo;jS`*^lX*@h$u8V}LOx<_|vEivgxt6fKBrb0H`Q&pjA)5huManKQxU3{l?#ozO(t
z=JUar-iqg&f6p9HB%G`ZSP;Q@;;?@?a0>LsOapJg(Y$s?q2Je*6B=wh15gHyB|fdP
zh2A<9PF4mRd(3SwFj!f5Y2xI4_9dU&f+uj5!>GcC5YQ9Z=39lVPQbI9^R_B#OS1(n
zCVAFrnltsFZ`BUeLNpY-O*7F{!4kJjNZikp*SPt^Wt<r6F(=H|#Vr-38@Fo1#{O)Q
zeMowiVguRA(T7`-#%=UpU-f}C3c!p1X%z(XE-!c>LZiv}%-#n}^fA53i-wEN?y=R!
zf(F@@T=ct^dC-C`4oy954q5uu(qE|O7_8rGLfmxgIMm#jqUMe$fwo~kB61oT)lptT
zooE1eWKcd4GS(46?C!1pm+=J)`JgQ0bG}tC98bjCXs`DHWxw|!>&R<+iF=r7TH*ZM
zoqIPqdc<i`QPO9LI&uz^u2iG+D@ak(o5fnGBHC?OW9`k!D7W3Q6|`c+1c0WqjvSy#
zzSRAFV*abbh5Mox?&u@AME&T6ufC(jn2hr$Ag+Zk<KI#f3}P2lF`h){K4*o|`bJ#O
z)VD7xl?43eohjF$?%$Ce6PT?f{?>8Td_m6=5*}u>ATO5UP9o0}qrnUDHwc6$sOuHz
zxl*j8n=zwQLBm~C&1`n2mtV?aNYC+~8kB{QxM~%meuBsNNN#A?>veo_4*ClRu}V=`
zQW0nHV3s_k+rj{l`-RRIS4nsElZuQtF15Wv`o)SDy9$NU&P{^1$EFgzW71dAOlmd#
zMGWN+H*ekcxsx8xeGa`~2OW4jlJFOi4Bjdz%rY70<o$!@lHfWqGc#rEQ5nedE?XN%
z7fanNgw`C4VdGHPcdUJvSq7VxZ%}IS@$3KCzOwy?IWjxPe@7WQ!Wxgk>Ad3mnIkhe
zNRl)I1&W!zwZk1A_q7s)_Jw5!;vRPY43hKFrKtWcn0Q`F!!`k=NR>#2buvn47kRop
z-Cl2Rk0CfXk0iO7z5lUTd@($HdmIZ&CoXDg%q@o2F6!kTul@PzzC}2}^z)7Re*P$$
zBsQOAZldM)c6TbvUDtL@<`nqxPx{lz>!{uGEV{Fcdv#^>+{e=SzTcjNw#Vt7vE+?~
zjB}~TM9*QK35AG^7*o`Q+aFf!iH^>@lXav?=)7^qXry25QzX;Fiq5{?@GvXh9UniX
zEJe9YVDT!UkRmcB&|(m~OaVq*89stO*qLfSIQe~1G?`<-GI#029|3||kS6!c_8VHz
z=BP+9L!v?nK-%bahVJgn)6lOc{`zl$qCI$@lbt$@K;USlUX>VW)=Vaq1%0-}8v~Ap
zzWQ1$MzP*aMSbmP@2vqWhK;IPH5Rp#*wxMDr{Y-ABq1P13)dnZYFL)h85%(*TSMk<
ztnmP#7G(@|Nfyy*w&1~b3oIS^aar!_II*FJt0xIpHNtbWGirxI)hwd5DP0Q&>Zsb<
zvk}Bu?3H>^Bm!mJ;9L%dlQQyJyJGtJ+yei@4Z4g|%7YjH$y4)iip{I|_*zGidFrS!
zT`0Vm>A}P{)8oRdu`*KFu}ULpWTrgR953&<lkP<730Ez1EmOli7}^^ZpwwM-!6I1U
zxPzpsYfy&CZCjk+f|lPp?I)vPnqK@zqZn1jVjG)Qf9Y($NE&4oQh}^QYji)dtiZ}%
zE>8K_->2OOR=IQlo$o%)QS!=7b7)s?AMGF`dwS9dJm9BHxGS*xP;KM9z;nqu>)HU%
z#)N}^6Vlp>kLRQJDtw=k%i{g<hRB~sW}>4XsU&#tILcolk8mSiC0cO6e8gz6VG0OO
z*%+ye%ca0Ls@FuZqL|Ut{J^&z^V%}J%~?UE8)wq$LGWj+3A4yyl<u`ishp!{7LFEr
zKQ>8CONi%tAtMV>Prk`0P%U}m5^DccWizgtrkn}%c`FZ**-bv^d;ScO$-r%%@12O_
zMk<M;I}71svt*L$j9<+)SvqZAQ2Qn-6uUunk^Ih4ypg&_gT;}gKG^+U=zC#pm;}PO
zV9c(>)k@>5T%)-EGnOYGfAA8{e!i9D63LDTg7EDsH7JF>&oK=z=oo_8iuO<M5fj1o
z>+^Lnm+wo58sc3(Ya(O9A{XB0Ekh6V&WX(6O=+9gwn3cR0`1RioHr2tF5E2Q7~8$Q
zhX7qBQq&o_PxxS7)`{%ZUSYJIc|eN<JBbrX788}U&)Kiwh8D5?N@V06q7)EpJxtxa
zeZa@l4O4Rd7bmY*2Hy7X@dQ>nlhLjSd5FR97LrK-;8#a#4QxB2H-#ByI8=_lyBYhh
z-<v6ngECsilIG7e{q>RM(yZgD@i7*dhAIJqA%~nXH;@MUN#YeL2?Y$Vr!6uniaI2j
z#T#E=|C;^0P4is0m<!#P5R*)0vYvj4o)Xm@qWruRrjJL*G*W(PnCyrAMI9b`L2D&2
zAl;M8m4FP0jc$E08}83U)<Sq0W*k{3gUI{S8cG#{+e(8ogaJo;1N}R%^q3=lN-zW&
zCdwv+UJeJW$MJMc)W<evfyuTI_ATDBp&B-2znSYo>_E^B+XANu#)=R9vW@QdOYG^@
zf(iapH&B@OP1y0q!JPrwAjpr^tt3uY^FD=2pL}3EI^!P%<pdp@3$f6#J96uU*6&v6
zW%NL2<V7!ub_&uu?iU?|;Fx@lDb7;yGK%pu%MqEukCa83ysB-=;4H@uOj^AjjaP`@
zk(AQyZ3*v+W!75<JPA4h^|#SHBE!!()M90SW-}FJ8G2}$z-YX|4vLOlnZ&tpDJL0f
zHV&S3mV6Dc7NpDEq~*nS6-EkGD#>&fQ-zOS@;JbvdGJiK$EhO?l&3x6oAJ917t|8G
zEQUe(PRx?8A81nu_x3LrIW-@<taKu_#o%QD=}Loa_<(3K!|HsbLEdTdS<M5nvbIbt
z=GcshMLeUQ1x)d?^@U^5B9@xM$_I7&nUqvYVzUB_lv~fCn*Gw$p^V}qZRavp`h|oq
ziNENtTK7#FhW_k6AU+M(F*P3|4#?o8H@80&1RJ&5lZP5&TiF7dyXa{}i)qIz32iF)
z9n^&z9eY~;Bp$UhQ)|18J9|AzyH_PKU_-h<*LHfF*c7SKZHsk=Qby$(+?Je9ovFpO
z=<iPTbtQ!0agLdKpKXsMCX!VpOt%`3$ij0{_Kjje+Z45rqm|r_Ao09Y599}_&it;Z
zx-Y0ox3RKVOzGNfmYAqLu9$qu3sT+pCwFv#a<9o)#!?X43RrC1?s?{T2|bIs`GRGx
zqhK+L=Nob(UEOcQ=(dqBvLe5!Kfj6FPYaXQVs)~&G8j9f5gEpc((q49`=^Fe?rOK>
zD=Us3>J31+XEdD@8u6jtG<ccm%wH?NMR}PI{G{Q|X}h4xpq@hC4euBXbf3DT2NSL9
zc^+i6s-`a24Na2amgnGS2xl*6U9p4r-2}d3^vbCbxUhPI*)42&_p@w-c_z<YZ+;7e
zUBP1nX%&WdD)WO6LtR5&UPJscb_A+y1)5ab<oqKPmR$|ZH%xBZ@zDd>D$;#gepVit
z2!Dv$7zt~5>CNA$dt|3H>D)bBUHIx2Q3)4)GITb@q#pT`pu6Bni*^*=W1X|MLRw=_
zc&AFFGHt9j>^%VJ2)ppW>Rh~P*#GouFgABD?%|fM+sRI@O_(+{$Kaqs(t0kKS^xp<
z6ibl}RFZ|D&NvUZGsm=?A);QMJ4)%f@PPBI5wvLSBDXgO=PHrd73=e1^;|(VGzq-C
zNR^P_@xwXnT^e+jWZ8&s4P$7%|8)CD#}2NP?yJLcprcj1S6SYp!*Yy}`PmtfW<jS8
z(LoJjrlKe1NK5M7$y8_ILmn%OF8OWq&_oe@4E6)Set6VfU7Vqm!?xUf5I%E=wxP_5
z@!!kqPwu6&A$Wy$G;HPKavE;Qv)@^bNI7dyl3Gobs@cjrmtwfgZfx^?SBcFrmc4x4
zcHjryd1f)D3UQ(YB?U}k0&B&O7dz&EEz`{$@(79*MmdukbXj-&#<k6C1`wG&vNT(V
ziCqZmIqao<mvATjIpR*rbpYg>j;x-gYJ)BW7|7Rl)RJDDsafsZoWS)p*fk{tj~oyo
zHSL3WkqX{Hp30k(S={-a{X_aJQTbiWQO((ljS>f7={>gNF6Wbr$2nQasz*qhF*eJQ
z7HEZXqvJl%fZmD-r*+mYo4q&JHvt+#bL#ubFku=`Tf^`;4?^kS);*i7F~sUTeO@Rl
z37H9}Ira>xeD@cFbLHOcKSQE&)dJ+wIyp>uMIdJ+TVQt<2PM~9|8BOQwT%;j>ekTw
z(_8yM^h#9vh`B!`hxQWN!rs99bzSY*q|AcNx+s;|GuRil*t&ZuMK4d{Q!tMqpECGj
z>N8dYB1S^J7*LNlO&lrfuk!$mBMVX{|ChaPQL!L*CB+NLJh5dOZYHj{zLg<DrrYg-
zH$tq?uYLr-tej;ZVdOgixag^)3t^SDl06%*6Y_C&=dNay%TpD+g0l_jUM?z16U-E>
zBB$he9-y~syOXP9J15`mP^#Rj^eqs{U}rN2Z>(D^yn3?JE^KZiEOose4uhA=j_CwF
zPk+X<!GV+EZp+W}fgr8wVep??W;XW!OX~Um;lXkKA08YV1LuFG%Q(`~vejxw@_DJ{
z=<Bh<9}u1iK17WKT9a*rixUmx0AGjjslzonuvUY%a*zLd_ZcL{os18dNWr0?K;|wi
z<UF+FNVN@-iQokM?hNvE6z5VXKFGjQ?VJKN(HY-YvCm<N`$r{dQREU%%dW-%2~f8r
zQ7Fdq->0J_^$lT1Bi2Ax;DomgWQ^q00Hj9v#;h@TRY%jfSw_BA6Js6M))V6$hjtaI
z#bg!5K({8VT#ua#NR5gNqOLQV*(eXiI2xuxATW<KsPvub`R&KR9B4haO5Yuoi>$;n
zr)w*&fzu$#SHqIHBLAufT$5NN{w+fZAMEIhb0#Q8(c<|lN4#`we9a>GB9|h^IYNw|
z%BAYuP7q+pTH3Zab=a`K(E`Ir8l|#u+u=V|9v)o!SBkj4+J3H!GrfwJR%e8)xg5;+
z=_n=10AvY0<a2Uh5DTtI(kM)ajww#RtXMX_Y;ab<LA|Ysvd&sm*~<7zN|AGcNDi|#
zw60;Iy|Txm7@i(U0-L07BEm$^*uR8od3|hMRVtd~no1cvvZ%5^76gQT`I?#-R@Y>g
zHm#JUzg#HLem7V%d}FAN@WAvMTuvWrqpjv@Dtbk>@X&t+U!>IzWyMQ{mx;&4VNO{~
zP_ye)UWs{XVJc8rx#jJT&<_L2d}D-YIw*5IN$}tL2&%6k4Bpfc{XZ&iZlwxYMTW~H
zxBkHRJ#Sfvm&`+#i-$Ky0Q?_qlX1%f^^cO%NOBIaF+rV(^1NHNs2hm*`*{TZ$J*})
z`vv}VIkShFZ+n+l7l&6eqPC&LuiDRtFLE!xOY+ak@z)PHU1;dfus>smg8l&p_#%!H
zMdae2cTa1Zzr5@jKV83#alX8K7&D(oKX>$aUvh7DW>4^Izcycbb*Fapa)Te%Za!^S
zeiopQF8AW!lV=~FJA5;_@+XOtuB9i5w;Q2fmxGs8{IQm1<tl$_<Q{eSJk=+;vxgcb
zXKZKBRAzVVOCDQ#J9uv7PaY;C8!uiDp>KR&K(K0Kbp?9w<YV#i_in0_caS|Bu@-I(
z6ar0Xc7){xw5eJDz;GcC<&~4;j__Uk=5*&g$#>o;`@<=PNZOCsUSdEQ|I-e-N}kOM
z@F>%x$ZdxRs<9P#ReUb>O2DC-@+tQTfQncORp%##k2q~s;a7x>zxtN>47V&BgOz?F
z$H*7)c6-Vkm~>j=3xPF##a{ecWRD5Q0C@lXKFIyP=Ru;X%QCw-dkRG$*E6KnSN`Fo
z!S}Z42TOy^@@dsEjqa`*!Kg_g0==5Ss3q?Hxe95#@Xhc})4}r%3g{f!d7{s696xF*
z+&j?YoA%|)$(<qB!NTLols!p1(2(wDZIR4zhkPvB;d5IMhPX$F#ebovH#-?ke9ini
zS~UB-7%zPtkLd5_`E&-stQ%i^RR#4L#Hb}xZ@id%7{9G*GV7Ju*3J6jcwf3?;^t%G
z>W55qYc+XUU(JBiqusSd**l#E^3(YM{-!Gvd;@>eQDRZzQDRcERwaLNPQwA%DRe!i
zWz`Xj!mk-2S9-#3;O(PXM#t>}L3@%2p!>C;3;6xz^{oP@5#Y$9@hj6r>yM-z$=*k!
z0tcJGeu_-)JFELkzkj~cM*ieDMbQqs2Xy|+L8O1sRM1JdL!n{_%umj*UC6K1oxM&4
z17)iWpH-HD*Xl{j|IL4wz5E-VgA^2f0$L83?{8qpX}NK0XU4OE>lpW;sxBQNQMqK$
zRy#I&U#cvH_fE)(-xy#TsVod)e{i|i;lcAAwrq*Z^eKVOFbORNsSs!Zm86gM4|u)c
zf?Aa}g~(P!GLs9I^BZ~<#Vu$fFM~uoH60~m^#O9D=N|Hn?bdEWM(XN-*r5UOeI1f}
zTFlZkwh226QB;`-yE)jaKrq58tyHjjlBR3qh#}>WNUOdd8Q*C~u^V25s;gGJ+3qs9
zmgz2*FxMq44J!WYO;@Z(JCUKggdecmH;N>sTO&I#@Lvi*_}EffgzF{BiFt;*GF}IO
zt*UHhF&eRo(yjC#3=@I+Ul~712t)~_uXTrk8X*y@E_Pqw{x}RspscPJ9i*V)<)AX~
zFJM71k20{f9Jkh>ta{U%Tsjn{TO?^-V{VZrU~ZM-4XCyEUM(khn?_phUzGp+$I+{D
zk{>+OwujC&(w(T6lZbCQyr+ExzNa>JxgZ^1NU~NE>T;Vl>7iDVNvxcOiMlJLfe+G3
z=%@{8Xapnz1i?9(*^Ia9DV6EeNZ)|fXIQIITL)v#QuDX{ZB?i)u#+EsutlkLaQNA<
z>tb^=Xv<mt*Qshi5xZYZE;2cC5{(x`;8OZ46sA(rk_z7lJ_=Q~8Ig-kE;?i`-VV&x
znN#)g!JGh60Hv-vDBal%o9B{}oZ_!W5)%kPlhSQqN_xG@6o$M$|F6L$PHERmfxvl?
zVyBw`NOgO1Z$t>muk31FaJ;{k7l=2$8lSdY)#Ps$j5tQvQU|usUEY=W#N@&lH~LXr
zk}qR-My}k+Z^&C=i^J1HW~+pIwbeuPjcLW;)`3=ilRGRT+hHl355r0HDw{9u3*n-!
z<e`?{z{mmYI;b6#%v)%-J8aBI6)`3yE+zM(35mb)Sc{~wX9=TD<3=6E+qmB}yyB64
zcx_K!AIs1cA&BfbVg-)_Qz6cb`f(cD($3&AokZZoMuaq88qktI=sn+pdIegO(z<-j
zfTPTJps+s@Nlv$lnMi!b_==gky$Y0eN|wcJ;rLGkd3nn~V><Np(}w*Jz^aIjV(H_Z
z*#4=hNG<e9fu3RNooI5Tu!%B97m#~8wt>n8%c0;(+zZk3nmT@4ZMme|st{>Ls1|kQ
z#zB^Mm4bebtX3R0lXGxFC=L|bjTGmfZK2(}SW3@;Ad1)zQ$i*{P$&bk(#%CQ1e=3|
zypI-aJ4nAh{#6W^D~lhbKO#jtqsOiba`NZb*qd$}nq7gWvMMo^<|N;g5+ip2D9jv7
ztCKmrbscRw9uHfwitUza04dJ8PH}X-Bx0u8Y(AE;{wMwkK_!mSXz>J_)wx>y16$Q?
zqF=y8KlOuFn#u39f)04)q?=#LOKSSiu?oUsWrV~_aY&Y9kY>g~%uEBAn0rx4-c80;
zoLE8{w~-aw*D01hFdRu8`705B6o^ILpzDN&0r4)bo_A946Axr~gfNt^_1sD%mKXn;
zK?n$e3ErTE5L*jfhrs#(lvqmjwO})=6{B870x&*;YVBaX-85DP2QC6IyW$XuG$6!8
zTjHoC3cQ#e_<YNRNP?_%IA<$L!KsZ)a>fjvRB=QG#0h$#2wi&^`rMUm!bYNb!GAk3
z49CJ8#?a`5a#q^j%>l2t7R=xH2<PnRinv7zNEndU@znfn&)pr-S?U?wNF0ZG?=a)1
z^Dsdpv-foqYP2C*R+Bw~&{_>|ikPVlXw)EB<gYHs4%>6=V=tH3hoU>Xy)C0h%m1l4
zWoQ2nP&9U?|IUXr{KLCBWP|NFQU8ZfMI}{)C%_=Uw5E~QO6NfBLDj)~mH+;YP@I!O
zvvH7Z-0*8OJm*mSoS0Hv?pw~`?f&-T?RG=-z&Mz=&eG$P4cJxU3n~E{P8MAN18(u+
z<!n}*GI{SL9Y0A4;6w9${ZF=39t$THC=s8-y8$xx;g2(&5lgYgMC#frocpg);_61G
zs?pAV&MNn}i>H@g>7uCvnYREp+V8iY@9YSxidYDa$cIz`K`rj=;0yf%fin6mO1kx1
zjP)(CMewCjr}d=WMq>K~o5>PV7AT7;ct~c{G`Ebt3L-NmRhNStxSx7_b9?M#t%QHC
z)v$1%g{q5Yv1lc#bp*z#Gd)5^z!pX}>>Ix?UaG@Qsha>MVnSkwg@!Lx7`JnHZ>28&
z)@73|t&z*bf87T$QpzAS`uAcLKDXo8Pw5QmU(r)aXFjKQKmqX~uJw89iM7QYhtO)%
ziB3RcVnhFe{N_UzYp8kuC3Rr++@C#KT5HyW6th|_EL`Z?B?j)(iszRnk}-G`!b+n?
zkjD1sAp+guxm~;a0fh`ZpT@lbtiZ$oB%azS!A#k(Geng*Z%_}Q5iq>^*yySn25BIA
zb0lr@8S3jSk5n6UE&C$|h`@pY2d7TniY!sO)R#u*+)WfM9rhQTqaEhh<tXS3QZDv&
zacuf3JT=g<!-0K^(aVByjzQal;tU_K9*wKlHxuP+L)H5n_xa{)oFkWXp#o1=?Vc!3
z<8FakD))(zuc~$%_|gdqm7Tmbx-<l2fee%|7~<h<5MfxlmWj#ET`DG`yakvKr*im;
zD<}##r0m2g$sAV0Uc7$DW(BK(*l_(I-MAskAo#w*({5oBVgUmKNaM2)d?@3yO<X8r
z3uA07s!}V6hORfZ)FoAw_|0nO6;;31wc@uWWS+;YM?@zJi7i&5vci&z5SiS1N*uGv
zG2Ch6k@==-#osFgO{4Bop7}=<wUZL;`6+y*UiD%T(qU|ruG%N!l+l|f5=Y<^+Sfh7
zl=v?9jZveZd<3IX3X^E?5MB%<N!ZW)h`0Sr${5^JU9<%hXF<gV%@}v=1@|#G#ygR3
z@9BeuYCnV}+}7C{kENYf&W(Y^yy0Z?3R3p-`GBV+t7C|bFXJSAk1f7J*BSOCV@?v9
zqqe`?9G*~PG-?|QxQcmy0UeIkZHco-+X<Liop&uFA?xe8^9x)zDFcZ_C#a+3hoko4
z3HN#V&g8$8a0+No_@ukx!Q^?@>!#z`c#%=b!J9T2H+TUNO=0@D5%p;u##2AGSZf|D
z1%CD^uVN19sXQLu8K0nP-kK{^H7~xlMfa`ACS<T0lc_e(T<OQ{y7A&e6qpylb_(8A
zE{Q0J_?fZ^^2ABpd%+FI92;jm9a5KFV`t*5=gesrE}{j~{fv>?WeJQXskCEqbIn`t
z-_5YVXxW8E6+A|e>2_G1W@d+Vi+?=Ts~W)@6{%i<Cil3joY}i`AyjKzE#n_OYEzDq
z*J?7}wM4XKC+$|f-4csAK`u91MT)uylvA;y>pm{I7yh4JJ}*(6vptpQ6PuNOi&MBj
z_^N#tGru)zE%V?vTM05@6Rj}=#Cew+2+@-kKZxHfSMHskDT+C-jtg55F?0tS&jX<`
zIe!I4l>BZj`nIj$7oITlv3PZ7*<vN62%^Avf5Hj7g?-Z~?woJGjAT*-wDWPMp^PQ?
zEi-E`SArW-{yXjRZ5|$#DaTttJ6SW9B%@WR$@LXaSjr0)aRZg4s^tth=u99nh8oax
zsu)H&A{m;sU}-jKu!i=BV`AA;o)tQ_rmNE0kqV8dQot7pnW$5meo2EL$`!oDwP{kJ
z;5`}B;2hpplnW<Du#YGvEiKY{u4&rOig8}ZQk~LlH)2|!wwZh%fpJ&at#GhVaW@v$
z<cXPEZpDl}P@VP@rT$E&MoSPAxUJkyVMm|-%iIT9S&`cEk0q55^!<ch*Xb6yz-#rE
z9DqMsIdsefss-bs9zrGUvfWWM1fIR8tPG!~_wGN;olgsEzGKS0)aS%m<uY0ama{m<
z?yc|d=Iz|Oxs^5HeJ{T`Mp+%q<U+XF?O-oJ`xxpi1N=C^g0IX~w;bmY$^e>0%XxOf
zpo~yA#hqg0(ypA*;X%$6HG0$<<WSi95u?8%B&GxC!LO;&C~u6@{H{K}vUvQLOUvDs
zr9}xJ<c{!Cj5NFUyW3x+>Imhbrrg>Or5rNxep33vo9B?yHu3wVY_*myGQDZqe?2!c
zz8FvZ%v2CqhVdH|XFXB;KmG?f|6lGKGsBN*=D+dBY-vf_YPTc(pmp5D-wEJfz!8Ha
zGAuyb*XC%)4Ms3HAukD-9mjU$<E!?cms2z$m_RhQro|i3(IfTHDqkN>_7_7J2N%ym
z8y6Cdwjs$&OHU;GK;gf^L&X4O<Y0|bGV(hmy;}}+v|V_+#hmKUy6}9SZtSk?Qf{Y&
zG(AIG(Tn>n&&jb%;^{n65LnSSAe3&WXI%&Oe0gMedq>w_caJM5;>(;cY(7!82d%j|
za=qCR1WJrXgT~QD0g-19t;$|`!3IA*7Af?g9G`Dz4?b@kGg8e#Po>P?d2BY-jM?i&
zCL{mGHHVWNMXmEd;DoRcXC(f4g}5TF`n@@+4Doge;|Ed|^WCL0{)0^_ywX`Rw>r-I
z{gvk2F!pplksxHOuNIji3JNHR?wE!ZF+~o{^B%mhK|cETDsD*gyxcmLGnhKwozNya
zw7|T(X=0@k{_N~kfGFBc-!-%xYLg=CoT4jQ9=u#)+4{O9%I~4bA8q5bNcV7Q&+dj+
ztVkm3WjT5I*?Q1)&m}^gR$Xg1<sp30*(=6yg=sG3vtd3g2y@aThcFf`j$2YXP6umQ
z`z-sQB<-_(rfMo3l8rY&d#-oTbut6r5G{Ss4Mx0jdp}Wj_F{hDxGaxv?l?oE2LY;O
zT$0ZF>9FOV)9hWGWF}T-|J-)Z=Fa3S?-EtnnMbWy#WJs*C;~ovTvX7Jz)&5oidMr|
zlWaOMVb2!Bd^*>VF(tYuZK_Q4Fi+P}v>c^myZPs5N<A&A#cqYo#6DGSI9j8r)S%Qf
zJO*~^VFAK`R6;A8x0zi>SnbIZyIBX(E(+kPtSBzfY3bz713zJm?(MJd{34@<@ic;&
za$}7QH5!XdhnPy?py!B=xk5PTpBn0ep~J*kJE%C*PrtqE`mzof=a&fwDFct%k@?Ox
zYz+IUM=m#YO~0bNf<=Vy*ZYsuS$-G@(62D!-L_xN;k%`_ua%*`y8SGAcb&$Xk{EP<
zb|=oq2UCDTWHJ9>@yU}Dpmt7{Y^<zSq&?`zPVYi8b){4wn`eZDvJf{Tch_1I+p9{F
zq(fuhX=Q<rj?^;bhz-}?{q-rWVK~oSJagyrrg?*f{@TUOO3_gwM?`~1G$>TEY)A;t
zGx)QQ$F&;7jE=?4ST7sUq!tJbrr)rutqn(_N>ZF6rfC^aZVq8t@KWzB`X)G66>F6k
zlFq=X5VJT>1YW^X+3fT60T=4+FNmM<!ZRJM;188{Hu6WHkZd}6j)dk6S!M8gDf9W}
zGn>gt0(&#EyOw;SiE`tURR)^ID+>!fv5Z}HIad5HXyjoue;%Hpr6BDlbB&kcBy<2N
zR1-*IjirUXp~NVhEe=k0SC^{?n(JJd*8t=b0rp=PH@fGfFq~btO)l+X@bCaE_ajr!
zB$O$^8-pdf>*<dbNAt*#1T2n+-wq{3pfjc&v88E>9wIUML$L4&-&O4FS(QPuocY8<
z!u|&y7<!6;aQFa>!oW=jj7(w*B<3$_uB%(HwQLae3SobpNre%yCj3&eA!K)=GpS69
zv7?d5GHgkVcRbG^k6SYRBXD~QY>AOyTjuIO1|P?^zsx1>v^<n<*{1g=Fy@B@+i#5I
zInTP?g_<IZT1H?=Xq9R&ZYIi_o0omGFQ}9omjmc)(;)4NPyt(!%LO?mb*s$H>s4r&
z%l&{9-DkCO3+**19NN9je}yB*@dJpY`~N`g*%o?7OHVG>uzGMUsvnbez_MI>Sy9TS
zr21g2%lg2XHC^AlUTR-BMCcHJSx7|q4_`mY(2(*Yu0lU>ADd()pvi=Lmv4O5$Vm2f
z9gd0YN%H>Cz7y)1Lk1b!S&1pE468%qSgT+G?Zj=3bmscL9(<ZAl)$0PM5k>0IreT%
z)beef=xhYad1?qzTLs$1io#hdGaV1L4iAYyg0RUPrVSV8oi_Jik2f5DDB%E!QM<8X
zSFW!@JKV(q#o!~3+1F=T4ND|$6vVNFn?5tSu$kiS6mdq@V$96DQu;{q2+05*&uNZ^
z&vbN_rj%N<Vy{+exl*?1^I)n8lg)gO^QHpT3Vn}cZ4cD1Q%CVL^lJk*rZ(3GYCuai
zC3wGi%w^82Ff1#A9fDxogwj3+^1K@_MXxao@=z9D%0SkgZnZq=ubfs8=DfP3Nx*Zw
zTki`<4XVcKtG0|0dgZ`tVKop6|3x?=)2GAPuA2mF{xJ((@dQECEg19e$e5azK8`s%
z8j)iDNk9Dp@M}dB^ey%3RhySP%ryB^6L+JoU)7(+xKrE3lVUWIBhCQ%Y~^q&VM4r8
z`;m4r$|!zss~h#oPLFW>p6aG53;*Tkqjvv^9^E1CAW{4MJkgAyNMG%ih_yZW_KS1G
zX2jOyGaAtNH4>YCdHm|IjPL0cTh$9Rbj)q7O?^<C7Ajl&X4SjmK)KPIN9@b8VVe$C
zT3i!%p4wLrFgsuP7}3c%MyQZ}Sfowqq#3To5iqvajX#KW8)^qpAeBstN4uIbcO{xC
z^twznmpFVYB?O^Os~sA&D4`>prv1f}f@deUr-HmQ=?Kuoc1%S=pG#;JUN%N*+>erO
z1B7`sFWWj|BOP?O`M?95Kk9gzsGC&<L}OKycmtXoH}Y^Kmv5BFNC0wC)bi<ZtT&Yp
zThfs)6nguzne=zhq^}7)qnbS=yjPdxitUd$((a6BsV5o3)WJ5>k-CF6sJp{%ja=rB
z>n@;>ijs{@Hj(7|%e@w?LTh+SvvJ%5NF7>Fj(4I@@bd`XturBp*3kPnl{?Q7wrHy`
zvq5phDD=VCL_kiVa{#|<D1dZl;8uTwA?)Y_e^YbG(IO#Ftw_{2bk0egeoE=%CG5f>
zc%v8llXeTa`LTFLEjZY!VM}a`OgS%?-DdsnC9-I=W;`JS+LAVPH)=e8zC#eDALw=L
zBv~!tM%0blQI!)m=cu`c5=U|2*dx6~0l<PaUr34oju~DuVgXxQa*7_ldxL>ufZhQW
z{5h-fU^>GKyfuNEEQx!p`Vd}CeWuA7O+8^N%-U!zVYgF+WEgbOprvSjN~y(f^n6hC
zPK5=6Q42Pz+>Go1XK_Y*(fAGXMW*DY8^7KvZNZ2$oPPV3anfuv0DI{agqo{;419s-
z!{D0^1~;ds74sDJ+H-TGRY^p0p?BoUVp`_bk+z9_hT?OBlLPBrfe~zhre6QdDAOlE
zL9`{}Flo5sGkywjW4+R2%4UG~;)V<5eP>n=Z?Cy+Qdimbl$L&HlnNa-J%03flTB7=
zX2mkWQX4zwNmNU6L=%Vcq=<3SbWoY8!X}|(OZP{3Ii;KZo8Ec>Wvh)?{v&KlzeB;7
z#2Z<u068h)kqW-4=zF5|F-nObV^dF=j4O5TF96J-BN1lxrI{ki@wlp^ME=Zd(~-*i
z=H^Bf-d<~SBCi_eV}cqd#3{sryO59IGuW<;|I5R7VXenX3h#J)a&wVe_hTUCTLT*C
zH<e`Y*o&F1k`y%L_yz%5ast+Dk=;nDgY;y2Tmj`34V(_8g~gSW<i)NMtR^Ij?9gCt
zX_5*jy7rv+-&4bbZxhP$mb&g5B}u9(Q?-o1<f5hbisT|V7?|I;qD#`ylF{?ER3vD~
z10t(~qtKGU=rZlYLP{nMR0zIbPJ)aUc<x1MDn}Oy#KT1q?&M-VFA<WoK?%Ash44C6
zrkL>PNdwo7UqZ)8K(6S=M55rD+<kpjWi2_VT38H!m*tijaaG!FEdf-~&%Y(z=WA6~
zpl+ihuw}G1>vU@*mZc?3+DKSq4`#)dBarSwDZk`M{PCz7*K0mDO$f{RQ3cz@`d=L!
zsRe0y$C#v48Une|hiMOHrXLwogsQc%`|fI+(&v{)%?Axo8mA9Vw;YN_o@i2uF$;G}
zaLW5Rf1Sq*fcM4s<l$M|v<~Ptp-ea`(_D5qv?*NKe2a&##eCXLXC=tv{nW2z%fdQc
z(Q2=*;QVfovXWl&V}(2Xs5B)4`s8o>iDmPqJV56=G03&^?EAb!Wj+}6<%e%sV)V;@
zx+-um{=XwLCU$0~|MIHve>r%pB@1)~3jJGLJ^VM3AOEL?=eQ~K@|$0TREb8a*u!nz
z2n=|{NJ}jo6(1j8Nj}wAuVm7eeq}aQkB>$>AVTh!<Lk#~!2^Qwm!7=)T&MJ$rh6rQ
zuCMmj7nkhF$qH+)en!=FZ5;pT$MtrP2JQpw)miaQ5BOZ|q_$(FW7S}(OojYnCmA&3
zEG(nUgGYKs&yl*<`{;?DE}5aX()cX?N$c0f%WUd**R5ET>P%^*tKgZW-q-8s<OisE
zsh7lWBb5ca*9X1S9lfvhT1^Y*8&NAovwGQW?v=1<w<GN?H``>m{0M7wltZ|VuK5q!
zfG)yRb{Gv7x;fWCqvb`6&*S5h`_&|H#w`U%VK#PE4O0CACdvjX12u6s(HPTn0qp^C
zRn!N>xR-Q!yCLQDX=aowKe+r4A6iG#97#j!YnfxZqB+MG$nfcspm&eVt2j)|$Z@RH
zVPj9Rdu58EE;KW<Jpw}0vYy^Xjvf1hYc@ntT*fvm+hUu$l@9A1$;<<C8cIr3d|}5i
z?iR+4Kxe+UkLyKaI^P*|rjH}rUh9g$Xwj4QoX_oe3fyuqp2@}xy;vMNlXsSIG%lzk
z6OBKFz`{&ULT8PFXq=uniyiFHD(9(}`*RdBGE}6C*2kZN>LwdjTJWY&Tj*PLy4OP#
zCj$O3@&jcw&}jE9_}BJjX`*fX5bAXAg|2mde`P#kV2;rWMeEc4&=AII-cw09%hl5f
zw_N<brNYB`O}8||Z_`BbukB^43#rTrV!ct3Y1tWG3GRwDh)as$oi{=qwYxH`qR|aU
z(`)snLe3krQSviMjL)H`qXSR3F6L{7X~Z7TOOI;7+{JKcSbzoLpRXk-8CAgb!BTNd
zI@IY|D|a{!Mpf?*dg&eBB8kd6W8o$TGC4cw@L_x>-A<I7vyg51mDr7bPpxUjm{%p{
z8UQ6DCOFgX(TyYF6hKyIg~<(623Dh&-S?6j3v17P3o>qI&T^S{R^9gx2>KZ(Mjb}c
z-@YruyYc!7$VZ`2>$dSwPbl?m#eS?A(+WhJU^MoS>i5dHGv&hyYehe0Sbm{O0B0&F
zJM8n(r+SbjL)vDlJ8*BR2>;Ej*BpQja)5F=vr8}YAuT47CYzMXmzt?lN~5v6S6zX;
zlem`3AMD7dI;=k@|15pZll$CUP$@ml`fQx+sN@!wmo;g`gNloT)HS<$Gx%?X(lsyN
z7U2VllFU}#Pa=)$J*sfe5ae>v2C<hv#EB)uzj*`t;BDehOocVa-2C1U<Z~JNU`NVo
z-WY84&(kcrL`6bjWAQz~j>kwE>7w#e^tZ4vBsi>{pA_3VlQE1Qu?3^-8e+R?6^ip$
zqiT6Y_v+mIrw}Ev;i~AiJQKm$QH_9k-L<6~1loKSJ1l7m;!+ymDGg}~Kn)GBcqOR-
zEtNLvb`;Ad5U0S6Wv1B}SnpAwt5#i50R!>uui<kjM>C0(4@X^lsxJE}sK*YtI8#ez
zjRv61MwDKW7W5N4Z-($Xq!POUX}e{o>i4tydZ$$L<#>7|(TPWc`-j#WBB`E9ch<$y
z(gwK?0rIv|2o$%^xzUYcoCFq2D4&Z5ttj|_3rQ8sR?-X2<%E>5_e2-L2bZgWBZmVT
zhvFm1pP*!|n~99nu7cISO6_%{PZ(jU>Vfsm$Zd4im7-#SZ^q@whQDRU9pnj0o+uA;
z-9ZdnlAzz#1Y<00lOjFdyz!)8j&dA@H)|<o%q{X_Ga+V<vssk$ilw#n(0wV}5eQ3w
z=|3uz%5y#B<24Vgp4et%eq<H}Q0yD&ci{Md7E-I=vP)x)fY@vrm~s|HqJ$!lTEhEl
zcae=sgCzX7B$x@3XR48{%Th75O=<w1YlG)d+}8|NhH>zgo5cnaJF=(*!VL^`gpBk*
z#sifkR*2cI@URh#KKs1on4W?mEN-y%pidsHuk(FFX=&R~ic!Z?AWcy>G4w4ZE|F02
zV>Oovw8RU}w0V%MvpEp)nLO}mo^bTtrl<j#ViEHXwRbsj&Y(u&p|T4V&)+-nm$>RY
zMmGVAb-0SMN{s$IEpa{i*U#YL?B2dLm9qE>#VVI!rQMIK2(8rBx4oJc8gS5k{Ggav
zQslh34v>MrR*oEbyIffCCNU*2p<jxB!>h<KdW=YUAbIpL?9u4&PIW`DRv(E}Ku-3@
z<hc~;N9`C}x&tP;0)GKzuci0I>xsiBSfDgfaDvf~*0F*`w6kVAEhZfM`I4cn?5M{$
z12(J4WP6UJ$qX-+(WIr!OQq!lID&Qhk=S)<9}G6o9S*w6d22WA#8Cn#OuxWfttjp7
za$HRS>v%!$pKjERC3m>P->qt<u64vQp&Xbcfi1@pRAZo}9bCCnf#xS!(oYQ!RHFv3
zrJ4%D7ApaG>Bs3^@?pGBhO*-?%Sm$hVo7H5&2tN7_WrYuVe)3{B9Bw?s>OCox`WOg
zBq6cPck1pk?Ixk)a96j}v-2i}-lp10+a<?9A99i43S7K}Ya2@It*f0_{iNXSD{+J-
zC2v<oZ+qIgJ{-PmpA@q>jK)8H02U65u_nzG7&04(ZvBkjc~_b<u$<3`iQa>s+8Qih
zlTwDRhHJ#ns9>G=Lw`fV^E;ziop8hq!$a6&XYS^FCqI8D-pK3#e&z)httqBO{DR2=
zx|;YB^(dcu!rF?if45QLEZcf-{xTX}H>YiWpz!3mA#z~8+<~>2G!QT2AK01%P2pYM
z8hPZr{T=PY&iap$0j`^&E5k!i*tSx&PJ23>yI$s-X7n#3C3dd;T48Yz5_wqW20?>e
z$Ejj=TRSZBHCN1KS-}r72Zdn*!n`}8SdY6_@^7sC$t@$U&avjQl<nkjl?NtGO-vu4
z>^=~nYws+Kaklc_pl(q6)M9ifd~aDTw;+E{cPJp$b}x2aB&^Uk^9EYVw+=|&XPy{N
zQPP4Qp~p1n#+_GmlW}+3&)H3}IGaO_m-ZAZ84~HAScVm{fD!RtNe0Gyo{)e?9gPNs
z)8&Nt%C&@oBkcjGW!lEp1*Hg`?T&_j#enc9Z#MpVptcMe1@Xo|EJvK(uI@|sbZN3{
z`a45xxinMwN+e$6?FOAKt;RJA@AU%2xdu$F?G96Q>JC;(h@+uEXP2O+9Hhw{#^9@^
zVR@0Ow-jSuG^XG?0OO?eMkz-C)y(POWfwMbtRekB7<<RyT;6T%H<;m!ZQHhO+qV70
zwr$(CjTzf^W-_B0CvVo?@2a!w?Eg8nK0ICBUH#>$zH5xp;~v-bbGRIg@3<VCqvJS8
z$-ylmvug7U8Mkpg?2oQ{9hal5>EYtDqbH9cV!*%*Nb~t^vVfw|C+Jip{XEq>3ZZ|m
zzkRl6!taGD;Ig>;cc{pO_l)F_xp!JWi_grNX1P<(3lZp^SrHv})uy@s!|uTNo{2~G
zr<`6Wd>r3WCo6=5V*rw+A=4sP0t0DHAYjyXFBG9nQpca$@yBOSKLe&|^Eo8%wcg)|
zUzYGV;8Z$9LW@rJu$^2?Un8QlNYrf|D61pF@GQ99O#7JZcvwOh;3rZswTIaBgMzP4
zkt)=!sTT8|{(Lw;+|>?&V_9OYe_M7}4`7YG6geQYe<+>c1Om^vgA|&i;_2)TErDVw
zKZzGN_6y7LclO0$?t0wiS8M755S~xrKG9lei4yfT9eut&U+(l7y|TRivGo1Mi8Bz`
z8(Bf~@cesPk&*42cl)1o$(O%fD~Z`*dfuonAjIZ{Yz2n=f8()#BgTDxIzcmWP7~gL
zg()DlY}7VxUSH%2cjz>VC><fQqvw6*JU=w9_u2V!Q>Fh)y3YCSA@$?4|5n)nhsCo|
zh@ois;nn!{KEm1Z&&DBGHg!na$PL_$=bI>07;|uev`wD|g1l=j&k?6ond-vnvJ3j~
zz`o2S4g5oV7RTOkn_tgo?>>oVFFGK_Y-Z4eh9NofXxn+a&Pa)yoO}j|zN|UZ_kg~v
zXje~xSuT4S#sYi>F1zqbs<da%jC0Q&0MjKmduzvSNq4a$ZPtR#8Gf42|CZbgSB~zW
zS>(?Jegw`v!|Jp`0Y)G$%gj2<+#*@zm@0*ye?5Svl_@4uSf#6_+HsEjsb$i;Nnv<^
z-nUQjtR_c0shm2OFdUypKRGDQNu;iMf^Ug8f%t&9LX4$7otH1hJ=&uQ%8FYSj3I3v
zDL1G)-Y_m2rvu_tyaei0%%V<2KAkNLH{S?Fo8p<N4@j{WKNw_gYa}B6hA*-lFe|k(
zJ*P}+BP%@+@{ha}CIvfa924>W=+SG$@;BLuQp`K0MH!9Q09^XR1`%&W>M_4w0)9R<
z^5^riWoihqO3^p#fiB~c*UBu&i#*YtQ;Yq~89l1KeIe6PFVJ=x>b@gxJ45fl{2*#u
zeHHBuwku>kMJJAlK4Q_ovB;(5!6`_z!6c<_JZTbYmJxMK{wuJn+ZKx4Qs)YO1<kK2
zFFpK4$2(OPT`Skw{kFX)Y@3Zr<6`6<;z7xAT^Kmcm13Facmu$1t|NVi7GQa0{j=#P
z)vz_cKJzjW@U5C$i{<n8ooZKm<-++nl#KuM6mj4+;OIG7J{|4?OEvp!+~||L{ONfx
zF`FIIoEe{c;InO?iW!^Z!g)zxqfB-Hx(gO<d4qwZ1#HY)hQbUo<2|A>oKEUDF`=#$
zN*l-iG9I8ERr5Q|BJR4_*W+`1LK&eH>kV&Qe3@JjGB8s-*i|~!Es?TI%yiCKtF_t2
z>f6jVikcTn&HN<z(EO1^S44(bI5^UcmxZIAR!I5Sz}!)R!o0#higN=4+$u&Y|Mc5I
zFRcCGRZ4qD4GKnwPZ7351^q8#=sUNf5~eEWRCwzy>-w_e`U2MP<JIw4FS#`9-ZoR@
zkp<o~jeAc174L7wS|<HWcFhY?4W=}5j1B;GXJFyMh12s8hJ7W-a$1S^Oh+3Y4O5C&
zf*1wDXMgEp;@+C7i9}wcLg8(Yc+!zhe`S)fMrxE@eUZ_m6sJ(e@(^Yx=9J0s&lSBh
zM{=ZP5U1ccyEVOh+e<dt&(3@E(9VE8)cfwg-5!fg>*PyjdDQYV4s)!(*N=K)=k2L;
zGZSAqaU(HEM0q)G>scy1x+|@_F=%ir;ZRrYjoew*6UJMQtYqakDM|-6c-MgKzm2X}
zSEv256mRljnGMSN{A>o{30Rm{<dvA{4PIYnAAIh|BwQ_=?^=zqYktK*kvdqL4mFSd
z#w?~TxM#C<bIfy^ANy0`E5hSOyh~+8QYZ%Clp@J3a_0(Tj;OpP=e9slc@-@b)I;80
z4nfY6&6>I3WTpeG+ej}p*05&r<fXvu1!Rn~yA<Lfxib)97>a4)myLNyX{*weRO~Q^
zG{S)#4>F;9y7Z=QVK|Isv)ugnGgZ+#$lE{u;TWQ-vr0ZDdv=Ljx0G8k(?aXFL+D^|
zJeVEE4FW-xd#-u_o~-mzNb0CAH?CJo2SXy<D^8&Rtw^F4`9li8^71ZNanU)_^fTH*
z!sm$La6JV0gaJI3$%kyWw%v%}n0|i~LX&@|4IZRtzD*h($<&Y3gdU59MCUbl!#Hwx
zFM5N`Lis`HrM0NEtr&z-nZLB~NGK{WZhG1p`|M`#zNFzy<QFcVe&C3cFz>?oDVe0F
zqk;Ag9mA61B4_q21dwUQn8V7?b|FH74s$$H&9RTLK1`D5hJj@L&UD2s*@Au#G6W^i
zwhje@pr6Qxk?t;r_L1NqQ=-><Y2n~O)%f_Z-~l@ru)$D%50mT!&xMYRr@{^3gYYs3
z*t3WY0<m2;=I$2$bOY$yn>0(vQl`MJoLyg&sX-zidC{FVZ|`}~>zm!y7>GbJ(0<|=
zP{}%rygnir4Hb2A6~W_mE?aT*mV=j0LBFU)(88eN7O26{@W)rwLAXtKswc9HU{?wz
z2MPLh%5OEQxmjs6xEG?`oA<ubAPf!uF{`ou@4a6}c834V8}U1<*%7lN^!zs%QWOon
zAA#U;4fH5Az+O;JkgSZ;<nt^3pw~s(PP35*e9JDus4228(z{Ob_Ix)&Uf+EzZQFiU
zNj6jy1-vxvCMlnu;{!dLid9Ywva}>!&sWBe&*>YgLaZC<OnJ`qu)cu%1h1rFMd;Ew
z{5e2ztFbwly(tkbA220f)+K@V-)7ACn#jz$VITVReK#IXJ<m{y1#U@fkqPW0@{JLL
zo6IUoP4x4kKE!}(6&h+P>Z~52{6uQf-D`|qcsI=a76!-Cvqq%jha?`&72h|hbghQS
z3TBy~yssCGDlDEPC}(vZ*xf7@Fb#68#QEGC98_^wgF8btGOk<!u2kcDI7vsBdkPL>
z#WMiNlAGU0^-hoq%;)vN{_pW@p&+>x>VGKX1G7Q!m4wGU``6%X8r_H(bBrd><};s#
z0+iW?05SovtavoXyc{G1i~*TiMD(<-AzUP5&LV;(2$YA~oou{91b6qqtQ0*ZK%#FY
zXh|`_ja2eUV?u9wu@ikLHcP#t5+-Tl3N&<LQB2<OL}e_c^Q#to`3iA}uDirfy#+H8
z87|XP57ZK#IwzkK2rZ&=6p^V^5ywc;RB$uyy4#!l6mUzC`1b^fj(V{7oNa?Bl{kev
z%Ue`~8dCC{A}+WaZop=LNHahT)j;;cs(`fCx;`CU-Q3ttI9eH;ctNzT$`r#=XDc2H
z>)1X*Hv*4eO#ajigOntuM~O@)5fC(B*+I#yGg0!6x)vTUbBNOoFS}jV$d=<Ijlh*`
zHOeLRj3=>0i_;;axoY-u6a`2~U$U51FT1S93hRuh_i?r*p58)_R41BtcM8J%?5n<0
zJ5YBShwhTvseTO(2<D<1$62WtO_#4EA`sCL`fv^qS2YN(uYA4WtR3_nftB^v)r**3
zS4FpIc}#D<#>;huqns<tHnJAE<WFPZ_}%cQtxov--1KRes0@L&flSWtYwd+TS)eo~
z3n(4g>c1lI)7Y9dh*vj64>HkYi`W%$C8lPkg0hJ2GB<7}x}apHzx<>AcvJ0(b`q=_
zGjNgPE$N>Pt5<}dRvPVI+?4Z|;i@k3OfGOZ9A-G)(2{J=ueXUd4reimFklP;>=GwQ
z4XsC|Sg-)bQU?@c=xs5^R4q0m7+p9C6SFTDG?RHH!#^e^OYr-c$1;$&*?B*!Xj#rP
zG<N|5C>mIva#FJc{rugwhWg9YIVqu{AWEeg6C0$<kbqz!qSfd0ErY_yBce+Q#D3%B
zO;T&LT3ayloObUD8mVB9q#CH6;kS}Jo|emCIY$na9#wSU<+G5&aGN3w8B0#~KuwdS
zr~{cDu@J>>E|{0Jm#)8_7DRDBS)Iy>@Egjy4c2?RpcT==yJ{5)r9hjWQVH2TFn<Ic
zwVi$3xJ<lRAc<ND4)(A_5UJL@n8`W;*QQyZ>?J~3irLSEqx|s_mx`0$QY#h->=7%I
zSO6O550k_6{_x0Q;&>3tA&Mz-J;OzMFMi9`qnd*W9U@a&bG>3q5SlKPB_XDpqixu&
z)~-y6_ijiYdYSB%Z0=g38xb?x(ejJ|yd`8C?_5}#?VL?pp6Du5DPemK@H99xG&Y_m
z%LgqM)Hhn8{?fb}R&!0&vRn9BM;$m~RB&;Y88e&}Q?Sj;Sd|rhQ>jR@5Li|>Pob3P
z+Fl`czLMa>S*1Zkdj1G`ASwEsEl<s`eA>9)nj=(Irc{c*i0kr`qi^1iep_hYKuKl0
z(Q<t!&=qnSdJBC5Q$giWjC^(i^W0Di=2AV0IX@xb8OO=&&*S0oue~|!V;Bk~(Gy>Q
zpOZ>*!`{)JyaK0O9MxDx&G*ktzvRk$Tnu@*w|;$IomD8Nh0{@wriC2n-r)@GUDv_%
z!Gtk#RgTxl{$;7?)--CxX<BbIGZ8q#u#by*J9H7{R>i}QL*446$S!{AH1-&J(Y!eG
z45o{_)Nn5(^bQ5vNga(JOXt2<Oc93;`LfFr5}_DzniWfkV@hC0MY#A0c{eS3N^VVz
zb=rHTFDV_lbU#6Mj5epwoy#6ki=pVc1TnZZI>=(Zau9%+YR_Sh8U3$A?sLhEcJM0s
zAI@;U@4p%0_r2VQ=`9)+Qg!lv)#bE3gnq+o7~G@j*uI`ib}#My(wTZ?>#jrMemP&W
z93{FhH?kv1!`}bMqIw4xNab2=w7Bk}+0h;BdwZ*$b+)X_rxI@SbaIBM?3{4xHrT4%
zd1CKgfnyP3dE_TG<3RtXw8zf&KlU;He{^YV|Nhcr;$;5Ma^kU;uDwnhqR(m_!CczI
zfcwO^<_{9&sK^eKquj7eKzj&2v0E;iE0M4!=wbH>#yySNoITt`7aMyxu%x4_`&*XA
zd<rv>Sk;oOJfW)PbP-5S1S{!mWUEUhxw1<|N(?pOWObrTRgtXJ6)tAdQ{qUabtf7A
zTw7Rn5{pjddH3tNx*;@pR%_&sW`2aS`WQo~IC{qc_2jgW4Jjhub)H$drEXWGe2Yf3
z3T1-jO@ULP2cd{$cEzO@E8}vlHLHw0?_y|%7<IV<QrqCf5;e&(Nwm0~M{p%=gEXHY
zci_^>>oN&oJMNSuWVMilQ1eM{UJ@VKKHNgRvLK+<ZZcF6%e~4l_n<Y`MmBTXSgVGV
z^tx!kgK7?)%0Y=Xk!q;2vJ;PBxZcSQjQ+7@B1URK?XA9c;%=j!p?Ks-sCq)gD*6~U
z$$mTdY|IL(M{8yxQvzL@scuD73e~T3yaK6iWUeA^rID;1(S07t+vr0}w%(T;s)6)u
z>v1}FPWrlXs7=bfNKMM>$j889A=nds4$FE`6q-L6%<!kThUIp%5KP+?Z0t&A7(N?R
z)A1vu##P}Z;$HAlb5yF1RNjvP$rdjqVmUioTZwKos*vzio!6yKx$qdevHavAY-~~a
zMn&e6@|sx}UP-;F`Ha<C$#wz4i|sM1p%jzNMve4kYji{NYGpV?Pr+!&;U#$jG5pO6
z<0GWeU>xQ3AL_~vve?SZV)=uYw~U}aRPm7m6ezZ+kMNck-l8#;MTEznQ|D(Vnp{C9
zz8m#HHY6~eDU<OQ)a%bHqiwBksluMVcjskM_FqqC9nbFv2Tm;>uYYY&<>T~p-`Qz<
z+q*p*x;@c8j!vWc=zI0NU#hF|b7^MPXXAUxLO=FLlKPj2O^gPdsgZ>~UJmYquk+|W
z-ClklA5=X+nXZ9ICxRhP>-{>P&D;NW0Q7ex4?-s@kwH1BWs#&Xgd;VPkH}QTjH>v<
zCZy6Y$4sG!8g!u$*eJ)F=qDP1)X`-|RM}zCO0<4c2bN%f>rn>wDfQ;6j6xYo6Pb4f
zW4z=k5H>V4*rB{GmusQ47XT%n(1G}jEQB}9SMldQ$-T7U+1tgh*V7dM?DP3@bou->
z@l>^(b=gOQw%eoS^Yzz;tMA?Mc4!lSm&c>w-ST$&{Nl~8Thqs<K})ZPhgV;YzgsW6
zr<X@N+x)R9@Ry!w+UvrMUhAt;+KY|l--B$q$9Gcf6!fUes{4_1E7!-1tKU|Dn5VCY
zay}20!E5pxe#>Vi_rBl7swPxhv490VyuGyE-R_7f5eX&`cnEt50ghkz?b;?nWL<CF
zb!q&c&E1)Fb-i7?LJW@Z1HXkBUXF~!yAcFFAmV^hp0eC2G4E3u?fubCYQ}hO4$|3I
zx4ZkGgPCDPwB9dwzkn}+H!05#eqXN6!ngG=F8#4lD)pbyW>Ia8@ULsLn5?I5U#8Ux
z)FC*1i7B<bJwCpisw%TLs9h3dqs1EnnW~Cv7TX=s_tX3MeSYE6B4cf!$J5=dVHh9Q
zZRYQj95%q^pLiKYfg*r$u-`O`7FS0%n@6YqebVRI{_OU0fAIqQ1@L~|M&dG<pi^7T
z5v+UjXzcFw_Hym*-l@NMzREg|QZLij`P|z6)ej`-pH~z{-5q6SP}!sDK5;1<F{gT#
zT?`kh2r*P1WT=9d2qPJ00`WYCyHLA25#(dvKnD669lL%RviG<EC;F!AsAmB!{4<+h
zg9rmvhz}a?PrVj=J*)s@I(v|pX1&QNCI>tqmx&{mU_KBbNLf~}JS52F>f9idUNPR(
zI_0NA;%_|@(58&cYaAdP@F9HWbpc%<h<Z=?bZS$lof0oPTO_^<7k$u^3m?Oo&1Cx@
zJcN6X?V!EjzgT;AySGb`MEB9QnJaM>UJi5I3<tW4T0g;hSKUB)Vb3ff1V3}ku#f)c
zv$R=KhLn!BS*|;i0o|a<{;Vf!owu7`avdD>F<^5r2iY^t17hie1umcIBLJdfhZMjz
z*acr}1O994@6$xMOxW^fFAzU<qzFI*Ls2v`h%<}hT`~xPfYz%Cp|&zC9ZiM}=LAK-
z%7&||l^uG)P<6MtqRJM$5?Hn^DY$A|X-pTi6}bDzuOEE5tG!MoLq0VQ6(R@x87G|`
z2;p(yZ%q;8hqiz$g0pQ%&<P~)B|%3pK~^5yU|wi>$f=iLLKCsggE2f_%G6e^&ZB<T
z$q|fLi(O;@wlEa*7H3+bm}ed0|Ax@EgR0fEWw~jTB()juKwH9D9k93gdm25xYp0yW
zpn-ZKgsi`_$yPmW922iZM}-9#+a^upkM8(6Z}9x+&`Ps)7e!OK^BPAA1HB2Nag1yf
zM#Ep*2j%uQ2#cF`*tH8CL!4Bj5-udXjW?0tBEeR|Rl;7tD2isB{MP|`x2m`W=FbA|
za+8trigBrDatCj}{XQmA#}S7}d@+RsQLIQjZ!w9ep*cQn0x>{7aG(MFPpwFU`H^EI
zFbky{K?obk>?FnYKzXIpGFKiFM)!yz)t%@VfDs|mheZ+eLujqw4$Vr8YfN)c4Pvlz
zct7)A#mhBrxd4<KMrDt*#hyVza6J828K?uvOx4`iFt9bWGtOLqwyHf^8K#yYa`2H2
zSN(Tl11ZIF*@qsZfEdq7gt~~-sv?jK$DS<%U9+`mA4cQx#^=~2i*D#!Rw@u#&Oa5v
zwqw=;Zh%52Us!vGScJS8hwx}sRd!YZy6%vSS=?MtCYaki=qd>(A~^jh5s;CJLe`jG
zz!7bpZ9_4!tu%5JFTCj|d`;+rKY|D*Ed=Hq9Ei5Mt!B$Z+5^xq06pgn7yAD_2hLQd
z!b}~YN+i7_QJK6ySKyFx7HSmx8^lW_i!-^Ly3cdr5JG|YbG8kwWIcGeOBQ~F7OvoF
z8@OP-pL#7g$Oh=<L>ZV=ABw1TA`$F*zp)Rc1p|qbv9Xw0d@$FZbZ<{tX9<HV9|e)W
z9Ss&U5oete+hEab2{<_fD`t)cGVbcFVj}??5*4BzdOkuW=QQX#v{++5JtLuz2N0y{
z%`aZ~1FSSWFg!?<AZEb^O9dl$Y6Tzrfx`uT5Cb(01%l94O2M@aMv;B^qfpx>t!LB5
znJjgaglf#246L_NpQFA<f1d!L&M`b)M2n06&<o=!A5^_a0cp`KcnG`E!w3z61qK`w
z=r1nVUqrCS2tMmDW_GCTI6`LiYU`psmDGh#Ac$wFw~RUy5?D8VwuV*@UmVfnQJ-Z}
zTCM<RV-NW!L26>0!EqiO)3f)Pt)2kR3eMiWU1p03c=$IZp<_&Ca=87l?9Z(4_2Chp
zjCf$si#+ce?F<eDsvM22i;*jMWUgFrL$^~eq%*dr4@GxfIr5fU)Y{+k!V=yX>`mfj
ztAVlOF3I4OYzf0-Wng3=z=OY6ev7jFwn7|Gs}b*G4zyu~tbEYEx52dNJ(I8Wh8kQD
zq*lBOs>9$BX9|M)1EdSBW7RAx5P-UY`@<-OeD9nw4+e0&MmNqk<R&=BL%dYRWvHq6
z1L*Bl*5GIiv_OMJE~f%XfFI)KAc*7{?zH~TUr*zvB;Im)UeEi(r|6SYoUUu8U-)HK
zTSr4#d4+g4zr37rbZ*tZ3|7LZkoSBw`kJ|pK}%+@R^qgZlD7@Lfx)%+cQanN{Z3x7
zy*}8rqd_hA_@UD`SxlFRuUxW9m(G*7N-YJ?^`&_mgHB4{Af{z^Zt^+p=5I83;bUil
zilz*<m@wL4)aVak*LxT>dcMz?afIy$=ZPS0!$<W?&YZ=GiR>%8yI!o4$&KEENQFyQ
zp7>i&;x2Q`lli(|23~eLJZQ!T$9!&|gBSwV-0pK;O7zFyAx}=rzaB<=>|w>8y_mkf
zfXVXeoBpZ7=V1FE&J+KJ*|RYJXWKQ~np1Yh>@eGJ)YtGLsG<pp{`~@1E5MzX1YLzy
zh0Il&E}wpi7Xa7evEeul$$*ZA8tMdbk`ZC6hX>n*#e1v!##hrYH>XKt4(1;3905Ix
zm!DeT8DtRo7>1rNd>`M?Q$qmTokk7fvE$wL6nPCkLiaBqI{0e*?2s`)?fMF*1Rw><
z)%7K~@c*LhMX+rSPhacOcGWutR;e7`vG_;z`+BNBbz5RERE7<WoS<1`{QUGs(9t9s
zekL(e`RDk)GyCX%FiO>QhR&L>!5gtn%QQ1&8Hyf}o2i`XFlVLrUzr4H;Ss}|FKQxB
zE3xTCC;S=#Y$@oGgG*+qt1UJ2^Y_XQji;Rgb_b|$97RC-M$HW;I|sw37}Fg^2PxYn
z>c{zQ!O)@fs5vH5cFE#Iw=Ao>i`wa~Lsl7)4p_2=Q-0$#{+$o6j%CpU+tp4b08jPG
zZcvOPHaaeD4%QDu+^FNQUkqS6>WTQ9JVtETdcSj)c8Ue|rVykI!z0iWLwJ~N@@pra
zpD_nYus}ffKvE-reC2#t%ut+N^*24HV5{HeIA5er5j4kH@-yw{=(_tz*1Un*(MFk(
z3{@yCvt(NaPZsyc7+uyN8oOp2I^hJSElMJFWwV0X`c--PqS649xZb>`EA94<$&2d-
zuI=O|63npc?nCpIDqKlv$O?F5P||0(Wz#6zNhS{aVMRd)53{Bh6l~287EexKi7QP$
zPK%`3u>tn783`zIXf;d8;gPmwA(i<*ddYDpqBB$wvO>|rZpCbzR8aT!%J;(VC`qSR
z-TMBFm5YwrKU%#D^UAiV)`WM)orLAPPTVAm;92AFq?~WPoR>4+CC*ZsT*!S%q@}Pu
zRM9x6)+W@|=Sms10u1ngatmc8D}TA^+=i}y5^bRHM#g-gbaNW`_#NN?>5gZ!<%5{t
zAU}w=nPmI#Bg!18NE4T7`O=o1WYq016`RH>FTRdeXlmW+ujs3bsoBB+tCD?1v)^wh
zxYvgj{bq9%HJn4aUAir@yNqRYNF}_Kg+iH|!1o&16U0o9YKSd+6hE@0)(tTB1C!Ub
zCS8B^x_h9DlI1nc$Q*5b@yXOaT~<Dp`ySg<i6^t{$l!Ktrcgn<mBX<mOeL(FR5M@a
zG_P7;GM{g<;e;8Nb8>!IXy0@5PMooztE$U4lYD>%<_BYi*VxU=gb+YiVs1>pG2P)?
zi(702>gOehNTrN!&KA=XovMb=`qmn!PB7KlnF#kL$ZA_x@nOBMzOluOV63&SYZNw0
zAAci<JC~(H0w@)A&Ux!DEeL9sx1)p4T#$8}9g;-7)XHrexFvo`IC;BveNK`}+P-fS
z?~S^Ih}*4E2ln2c_?S3@_ECs5MmMfDyjy%O=sI+L?TZ2tQAvA?1_UG?yE?GbGPFi<
zU%Tb#iWF}FIqOLnchYhD>~ZQ<!>6OW4sR|Yi~_($w<<yzcBo)}+nHI;G}!`nKH7rw
zAJ{<ykw=7j81dD-o5l2Ac}S7?MOzpQe7>v<fm{q*4CW+UhGGg|%ZGwc>SD!h?|(mC
zy^0aoN$f6{?>BPatq86zMI3H>B%-dwh)7u~4mI?_mv4K^M?1cs3lOX%WDc_9etTC8
z&UjTQKk-)3fqoQ+bgI0ClAyn#w+9{&#5<wcNjlOhxps0(L0p1x36uLY-~#Og%=!m9
z{c=BGHc&Il<tvzO?ZS0;R3HFmwcsAnr6bef2)-e9M9)XzP2~e~hy8v@UEl)fPUsQ?
zlwl8nkbD<9<YFyFd!PK4_c3eP8{0mT<7|>6@W;Hy!xd5B*>7xJ)4Ds2;#MYVT`Ydp
z_<ACzZr9s8@fsk~8!hS#Q6vZte@ll!j){i112xc|>Go0xSAV9xL&SBv<Njlm=V1EZ
zoANA7|Jgk1vgUN`;lCu@l<y!_Qpprj0tg0)wrFQHQ5~qfsN4UNaDRp=^`)YaCw;90
zAFQRC-XTk#6Pt4S<Q+YjwtR=z)>Sj2eFa3Bx?Qpnc#<#bi5N6;B`J-X(pCILeQU~r
zyN2T-i3(Qy;nmpL_Lt(6L%hioVOzcg#mf%*D=GY#bJ{=0zrEEogAQ|#B)|dHvcCr#
zUfv$#Qhv>q{9)4xUG#W*yS(UA#f7v8SdmAQq^KzM^Q|5PEmF|9L-zkODZ~Gj(_8b2
zQ-Y>D#F;rw$(GfO*1(Ld5pYIk-7+UKVq<84<zcX{L#A^RB-yTe!dwCjJydwHJA1P@
zclS8O!*&UfVX#up8=B9k>J=oZR)ToONHf$KkKM~-iH5*zGF&A*_L8vOIB9){8S-M|
zgfm3uu`6un$tI1$i|IU|At=bHP()X$K;Ui89WeY!;@_fZ9oqo3`s|7RYKyJnKO@LM
zfiG7i7s_Yx@mkXUicke#+DyCxk$DGu81Mum65~&VkrFb+!5GwBG(<%guZM9{8@^XG
zL&{c(3Fe&N80Y=q)WMW~wF+RX25qz~@{U{2t!-AP+*)K1xli;mX!VT+OD-{wC|;8;
zBTlbL2<aFx`?0RfxFe8Sp8eJ(8?Cx!X89Iq2wHDf(iPq;F~Brvn@SONNVPJW0)}ej
zqgB-VHv~xp*l`5Ch^f=sg~v&!Xnur$`WNN>joD2yCT4X(B&!arEM<F_8xI7lmidbf
zZ=I84C><XstMpX)_|K@;*h#kb`dZv5X)`k+ky!0=q-L9w2R7MQ4DMty;bxyEG>kmr
zh17=9m%4kA1-cpc9DVp{!x>%1S)dw7wR7OUD_ci`KPg(XQ*-U{p_dntKP}NR6bASs
zOc8OYb2O`&Peoh1!(DSwp~a5NBv!Ev_YtnQ(+}XHt$83xJ_s?~`(Hn!xq&(p9bp~Q
z+r32Y70tQn$8jtoQL0@TO*OBKr!+~hymCqANiI<$O7TKO7IX9o;)*YU%%w@wO~ox~
zM~0vFQ6k7RN#jo)zo7u`7)t3;xX4o}Rs*`Fn*bfR_&lUwr0yvb8yd(31<NX}YK3zi
zUTa5HBSHIQUGxR|PLA;@ZF)^+TAB5Qh1l)3iO|cVj(Jd1D?>a}5p#4Gik@l3l!LNT
z5l!uA<g8H)-WLAIoBnTsS#8qQ$w*a9ZEAIsJbp?vhA{_nFDWKtKeNWZZW3Zflmq6D
z`~;-?a@S=dg76>)JJ9dM_%RZKjRqr6HSoe7v|wE&j&J{cP`zZp(S^^PC)J}w$0X!E
zU33Q}=h5^;m>81DzAy<s=*p8kAGm-z8NQa+{#mL3U#?4i0dFjp(zd+uLc-;~GnM<3
z+41#`6c^f=VXe*P++Z$+b-iN6dQ%H1ZY2$vJGQA%>#YFUe17l~J3>UN)SOS$1_F_x
z@RRe_xGJEg?!2nc6BNp3E$W((KR~fR^zM7?QPE#R>`~l&Dhd{W*ya(-$JFzgBwB+!
za}In%-x>$A0ZD%@j4<W@_z__W<U-oue_#ONS{BZC;b6~k*;Ldj!Q>#o)No9b7{M!_
z2s4+pIkc20)2Ur)!(1tMS+P;Nh;yr`Tmam%Z`J8;hHGmOvpHRH<LkQ+SFlvDrU~pG
z<w_%QL<%2gPt}$}d-!zM&|T3krQq_@k;*!R{j>I0MW%^#sihZ~wADd!KMYV_e-bpu
zfby#(Pen88!prTOO5^r+Up?H$%<~}E!P`iB!ZPIf?=XW=;LA7WL!dJ$GZz9ZJsy7F
z-RIto2-~V^l?<5OyId_xS{sJSS6!A>r4)qDvkU6R$T`+WELYa;aM4(f!j(<{rmcY)
zj;ZrQ0Lrjx2)HvaXpFgD@E1f-7<=MB26xVXDB+ozS^vX{tB%-14ut6s)YpKM@^LCO
zC>EIZo<nf9&48v6v8ap)vbC2lgQzl*R~!9p+_^@J#7B))V@bTQydjQmI=$8?dhDQY
zC+L39wNDHIUyNg9vGZmY2LDFz(BF0YH>ZwEY8z23IO4{|4c_kG^S?>J<i!0a;8Wyc
z9=W3zu9(IY(+j=7X4J=#vZGDLj5X|Or-NJH&EHOqe$DmS;NT3|Vy5I%x0?naf<Zv(
zp}dPtxhGw>NDz$oK|v$2oGp4ZSHKfYN4)t&KoZ*jDABfGU=oz+cKYEZ(Sg-s1dBVq
zJPHoxsK$H_Gi7GMF@SO`W=dfn#5Z`{()HG<``%WrH7}O%7{n?^2=n!+73}*%DfJ`2
z<<5WT=4L2vME|2ubN7R_Q;uc8f>zNO#1=IHo2IbQj!CenH>ndFb<*ka@M*T-^5rj=
zZCnhKiPtwS!XjpK2@@_c%LT)nPGG>a(OP9yQ{IiyxobQ{r%|;5*U1IZ<xNJl=ULzZ
z)oY-9trN><9xO5m#i1vNB#*WMt>D=+?W3=7;V2spI^_f=qxdXm$6lOhrm4}utGZ3u
zREbzVU!FqVt1jB$H;^bpJUYHL4DOa18!(^<f&TFz2;3}jxkF#Kn-yFAgnz`R55-t4
zDJ+>J(IF*8Tr*+4nng|6JVi;;h0Ic*n<=ip`YfYr`1aAgdY4U8iqTk?_(FNstKNi8
z|5xAAAD%SaC&X+c-Rnr~?DOU@now{Qjtke$U9N(!@{l>L5tt7vmwwelJzHaPD^^A4
zRvAV)5NW2=CZ`tPCCaP6pH4Ink#UWx`P$K3#zmP3e@_Nv(nJ!+Te>K9H;SJy7^_f*
zx$>|mTrZcS-)t?$ER&iO`_KEY=ggq9wUx;&5mo>w?un_zlS#Z-y16yu?2Z;@!*X|w
zaE}KS(cBt0Y0vZr3dulbcX@%r+|@U;1-SV?IZ;UY3p!S^;ITFmib1!fBXlMemNYzA
z{$`I9Rm<O6Xhk{2O;6TEE3A}fV~#CS#Wl5z$(qGf*`72}Q<GcXPx_m~U1f4|`W>!j
zELMM$QE#1Geyet_jtsU6s=MhsEHZ__r70fTDO`E@i-ycRz@WNex$kn`<8~;u>jbc>
zJCItDM_}7erWAB|^JPd)G4QSorGt>+ncfvN{fJW85$9YCeeD@~_m^7qpdA>8FO)va
zF)A0cU4Ajs?Tuh#hH`_9T_{zS6Dk;cAFx=I7;YSs7tF$4c`BQRn+s{SBvB<ijOl*O
zALuDIYwc+UP;$`=Z`VxC89!u*v5=b2lg=Q!CmB7VX1gxEujVMZ!h@BqGjA>@TtYG1
zVYy@w{461ZrS*yT5;sahP6ipI4xa`pXCztMMAW8f4cZMQ@4*<VSUSkOpRoT`*85JO
zX+J{+qFoP91;UzyMCF4}Wl1^Vk*(xVG}VL(kVIUQ<ST-_GV^9(TTI+qn=dnHg1dZc
z(u6n@`o57jw~(^qT?u(r?2Us~g!>}sel9AQ+hvI{%jO;rfG!(r7<WHLh7|)|K=nqF
zLu0z0_bxQABU^pJlXHkv6{M!!3sq*y+bU23fxbC<|MsdpGO_x2QS{Krus=-s<KKzy
zo2+J1yc6KH`Q)tb1T9f%*qRrp%=NtxDF3w(z(~Xi*RehB4fSp#9mN;KYhp9QqAJR~
z*3dOgc#Uy$`<gcn`(ZDS(h|42JDK5q{7*?!FucHM8Fydv2$yf#0p93Y)?XAwvv1}S
zLAQZpuDFblx9{**5CRJS6ook%S^js4&CbI7A4TEg@3~j(e@_V2M{*ND#vhFf*fE}<
z9ia^Ky&!yj5sEgB+clz3TV!xR?YT*BlT<xh`7Dj^UlN4LcP&m`mqjfEA@BGP?H^A8
zJqRs)5u6vW27w}|4g9&YeYT}1)#{XUv7}5XoSHY6cXxiy?_Y$h_a4{}TCB$(JLrg3
zB%2*nFuTXq2rNsO;l^C$UKDK-#P{`kF(ub@r3{S0Kdkxsdp~tur@IYFWoUy@?Fd2w
z<LA9Of{qnkPX`lD<d4_;#p%8N@vZSPaeexc1raE#Wi11Wtz{UEU#%*{scy)9IvwM1
zd#sHye=G|vJGK1P>xbJ(p2r{a=1elNrbpx*tq!W0vrm}OnuX{(R~$sxbC$PM+(eke
znla1q8T*pa_#VBR!7*!gIKsGrTX@`Z`kSVp^8zqpAVI$jX@~+gAGGDGG9!Z;F}XpK
z@<CI9bkW9f%`y)i^MnChGDTg)G}ly?$B*hf>k2EU@vVo93t}2N6Dz6n4<m#W?R0}C
zHT6eNqvSZWs@kL?1EMTmWcR$S49F>1*6rg0P-Cli423n{C^MQdaRL|9yWnE-3<)%9
z##D!zNi0&N)P(~9NaN;0^EAM71@)bAb#rFrbLui*;#p)){?aU>_AGXquV9b*)Wa6%
zQ%)u|DWZ5uyJM4>sq1(*9mm?RBP&q!2UhQwvs@+OkVJI^EcQA!gS=kMIURHdw@{6D
zUZK)rA#(1NP0s?|WgiiiCGQ1X`j5_JFsLTcLI}kS61qs?(YFfdEzBu}hUZZ~*jE#Z
zC7crVD#J2j1^+DA$RD_&@@QL%mph=glBhd`R}+qIX3YEgz9_ro*B9mrv8yKB@G6+F
z_{J=^U@g?(Swm9k?j61cNG-n(q?i_k9?fgBl1Q&r5&}dVaq(tn#&$D19R;?~V=BS>
z<;5aMYeKm24f&U)8}v?imI*tiBh^X=ByYML92R{>yp~B{kjs^I`v%0C@Q`iuG;Sm?
z_mew|b`lx~Gs052Ssm#!jN_@loZ9s$0df|<wRy==<Fv>aCOl8))$M7Wv63ORdXzJY
z&d@cyQqWqh)93w><4$jG0XX*Uo_i$DXi+&kIs}R%Jx`r<OJiqEEZUPQKkf>rla8hZ
z>3{=Oq`y=6_?dl9{^dp;p}b^H_QXTrQ5M(|SPg1Gr4B@#9t@`ofx=&n$$MN=JwOgm
zb1DG(C(GNFU`N^>jBAAbmG6PL-2=PstNR$N03PP(uU}Psy-)&6AQbDa+=@}TADKJn
z%gcI_C>SdA9j&|{v6!e$7-f&N>dhhLTNjw&5{c5k6d0%}Y<@g#*tHPJm+)*zhKH$Y
zal^51Jf)^GLSzBSFD;A{bZ4!T)EHICIxJv_DJ31cqhZq}mP|HEuSaH%OD{slUA0F;
z8JjD+B^|A*rDsk#U6on)WLO%ogxsg^cS!IxhODTd8j@Lz5OCPSpJceQuKy4QNIxlM
zKFs}zoWQPsW0cvr?!iMga2|FzARhxeQ8bg;=<|&mm5Y-dO^9Jxk&YX~V6|w`W-oIB
z5u1F9^aln5Dc^Bn#>ZelD7S2GO;Z3RBX+E?IP$^2?EOgvc6icWCd$VPWo&HZ9Y<Ys
z8{de~Jjc?xxf;Z)ZAl-PRlt7DUC@xxPOB)U)xDY&=Tj(Zyf!xLu@5SIU4pzG^t)I#
z_jKM~yDcJA3F+rRV1_z>P$&hYP+V9o`&_(4Pm5yQ5#hLoLqO<<?R!u@yr4TYUzC5o
zs9}p2IY?mY7C!YEtw9?@T;d?o+7lDJ6-*0EPhwmt-Y<Jkirju)7X-oejAHI_h^7p@
z0@S~Vy41h^bVvE-D+vKuOYqMgp+Cm$PacBcOSzy2i^zGw>$a5QU9|#K_+{$$K`lnO
za>4=1k4TpVNqM+jFHYGk6G{(DHf|j1m5A$Vof&ZO#fTD^l7{qz^N{`O3y{YlkIyo+
zhZBvVdO{UhHc_!+E<X$_Kt-Y<P<u7Moo*iu?);UYCJATB;d}B=<HO`8zhZg(*-<ht
z!0+a!y+HtHB=6?q!fkcr31K^Y_TNwC#*xBjt$YBFgDT8}_&~S=G5x@Jn%iZ7sdJ1b
z8~=SJ<fX6?>WPPe$ZdBIh5c1sy$HyW!-;gXGerRsKm(8q0SmaI>moa#j<gA+OOERH
zREQJy*vgGIsnh2~$N614{Wjr=v#C_i-F%0Ib8>@P)Y06noVY1FeU~n5ZR*>%^r;-b
z#mOSq3u|)5UH>IP9H)ami=6>LB6`%}-r=Xa=zG5WqwE~KEVQrfB;r@L_Ke-v;%oGe
z+41{x<o~S-Vd7-}&zLaN-`+_y-$=tZ^#}a=Tr?|C7*MzAONjrTpB<2qa~^;H5~hgI
zl2W4OrRzJIb5l=|Dwc?x5@+rEcpJ6n&)wnA?cGM*`j;>S@b>&c_HU-4{WsGvx!vd8
z@ontq>B8PAE`NWA(sRS}_P*c8YjyAt;^3TkH=i~Fy=%439Y)Eb_QGk~c^?5ZoVb=b
zAl-fdzL)mllAr(c?{LJUy*v5eXErnpr?i(*C=LRAtQW8&eaZLQg%%NLtvTi+A!7Eu
zMGuDaT!e_JXPg_$C#+O(_e`QKXyr~GRTDSqb|jJ}B^R0AZCQkP*sBfo0^q(g(%TFV
z2_%D^AB4ev{GC-d`0`%>s>1k4%R_rH&C!s)kIT5SKS04|qpw-XzkarVylnn{D(ib{
zN=$zHnl~ui>tNKX`#}%8xUmgGLDoYTs1Ic;Z&XgG0;6vhMI*J#u+puh!iS+6joQeQ
z5Fr68G2$>=njS;s3sOhnFgL7@qbufl(4OweR?0<IJ51^sudX)+0q5PZP{*_}C&JaK
zA#_BKeV8aVe3RL$KJ@G+etRP!&OIAVqs?gttZRz&5{#O69?60Vm{QV|KqI>AG`|e#
zME#|;dI6rilYmLF_$RS8wk5?R%?y0e!LkabMPpjMZXA&$qMK|AiVaYPI-?u=kZ8$W
ztzfUgH=e7M-$*SCmP-l83wCsgDU-a9t$83sciA&;Fo6`YK;nCVSAT35UrM7MqbtU)
z@c?|Nk!ZID%<R*`9sJDMd`b~9;Mr_0S4p+m(~}ekFY%HHqOwO&_b!k&=vi8VeXd(7
zU4hfYP#dm}$LM*qU8jy0UY%0gSr2C4(?{7gkr=|!iPpuWpv81(Gh5*AgFP~@_drjY
zn$(a${h7STW5`#wW$l<5R1jdWv+NBqj#<``?>#k$COZR!qEWbSscRXE*v)52>fZ$Q
zC|BD~k!QB1C5VCOMEu!%lg<)qLjB6hYxau*)VJDOhO8Qo;<jYm+VZ$P)!zohB}D)b
zrCge7YEq@Tobc9&L{Hf<o}n|R_pI-#)T)+QGXcj8pX25pbg<QkVxImNg;s%zFvjA{
zSPNzK<CVt3+yos*JRR#U@{!(TQ95HvB}>Me^k>vl0T)bfe9;lYB5<Lm%i<mS5w&|F
zD%Ru5A6I&J+oUnNlsFcv5zdY-swAVVh^!+)k*-*pCS*&gA$c7_lDD&KwAU(3ANqtu
zxXz3Hvxt(e&UJZv=PZO+M{W0$N%oC2G7VwlLzF#A5uDNz<S(Z)!nSVWQ<=%kVcMgt
zl<6JCj_%cKk{+?md3*}@=Ak9WjSO+rzQU~u6V7XZ!b2^>GPQ!0KlRd#WHm1kznrD(
z8?y@0(6+wcRdFz!HO=~zG%Ko^&h3JxvvJ)dOD~;v^10O>to>EJnR{uOT;m4pM=mq*
zaDSUk<p-(iA|cb{o1F>}4OCEm=lZzT)3d)qBNhAFChsj0BZ$Q6?M`B$dc3AL-Zf$?
z!IQM(bb}5mXo&@>)wueS3H$V4D2}mW-h|r1**T}g&k2o>G!Qbc-kqOSWMFaj6=b4y
z>LYa{nr&|8cO#Z*3tik@EYq$>uFBpC9a~q%(wT?(2xNLpfy(XF^F$LDGR+!|<os<y
zu2i)svy2)6qHpLu+wcGfE##LYvlLsTU9+Jclyk|{Y`#CWgE=}W&w65Mjkiqj6~QL`
zuJlC)q>dwf3R}!QCkr!l-~84pWW5Ap<{mYRHU*aHX06U-#?c#1q#N?cYC=46S({<C
zxmZGbADa{+bvz{z`GU+}g)J)&EaoY<ROUpNjHa@2CaLGs${ELzv6SlM#!V{vI9ZI(
z%<PkI<RfPBnVj&vkH`&}D0ekxW=MwjIM)OX=-1M+gbTmJEW<{!Fg3W4OL^^#nCpX)
z6l=xK5Hjcuny!%hD_aNCX$J{%u8DJ*;m76)t@}ubQGhSdd7rF0v-nL77o<);2+G8A
z(yGTQ()V3djwTPi>7rCEQqxENzVaoE1R6xbH}W+Pgj_jD?hu%#R1l|VVN!y|_T+I)
zD<#b@L2vF_ZXAkXsvvLcnRotx;9$mR&P8!yvE!c32FC$=-sK4uNVO*y3=JM06a#sd
zAU~pD<@$*}*C}@=f3BP5dyW^d4O46_keDqKwaF$xV%}~;1K>!NDexE)`sJL#cJeSW
zb3%iK7P<~@Z0BXiHD%C(`NmdxXHkJV*Hs+TDRryIT<FL&wfW$=QBRP<I;6j>1Z&X7
zc1PZ#k}{Aj_%eGuUU9v+$$OQf2Oj#vx5}?jw)Xm6i&nhw9?a6KbcBAtsj<ai?3wG7
z@u65#-Muixiw`;G$80wfZ;DkAZi(?{TQUJG7TV}wE|Oom5bPXQ6B(xjWGT0z)@(J#
z(Mtti2ebkU`_~ZPO950`LJ@Yl`=U4&cE~~ohzEa9|Boq)vl0)(ID}&JNjT393LIfd
zS_Z!xXmN>pkIewwfV<YdEG_ud+!MLZiJ0&oyvdF?vXbg2a4*1$lnZ68LPfw>jPo?p
z3mS7as0YPh^EREa;W+9H#@Rk?8wvX+QJ2R)B|D52Y?FUcq~QHp@Or8T`p&h$wV@cu
zpQ;Yd%X6qkCjLs<HXq+Zc9g8LytZccB(d#Fc}}Hg30kIz$v4YvM5P6QPcoK(zvS?5
zyVM>Hcw!-UXC%VG^#tqbElzt|GgP?pZbn(q{$ZBWllya#Q8D2kV14GL-SIc}06>k(
zrBg~(z?!7=_kmrN_0r?Aw=HcaNk#4GudlKaKI(BggA^kqshYTuY*o>Q+_Y@J2Kg&Y
zoT$UxdI4XvI4p#@mIGakyKIij9{d7!O<XDW+-L;c8Gid@i;~?bX|0j9Q8u6=-(~YN
zj4G`HSS#q;y#xqBQF~@mu-0O+4UR_E9ey+Ze7idaCp(18atH(;PK1MlYmT8{J(G$<
z2vX#4(!_!c7vgd@ur!8Hy6~yly%+1pIQ4uMO9F1PnG3REpMGnCAZQW>EH|$&ut{l%
z(GDD%GE<mU!Vm13-_t<K^<1wgq()e@nKPJA;)QOrrJTRO1)p5_=9$*{@~#eEjwT{w
zLFUC`sXz}#sAJ>fSar|~z#OWRz=c-9E%q5(9nLDMrSUceg)sqawBgHL2_}$EuKhPX
zViTGP#`xSKI%2iui2Bondx6nGA4(2VQemdaUq3iVJB+o`sRt@peQ}9D)UW8HTDlRo
zL3*L9k<=LaJC9HORR^`qXl>Yjd@^~cF4RIq)r93|%uNWPZ_*xE{fQ(f(Noa}((6E2
z8x}Z6-cz-Bfz{d>tnD^LYZG99RbESh6qgaKt*`RHaEsLCRaki#AG#{9ToWorg1}73
z?4Fn{)N90{b77sr9Tep(%{;rMEE5)^re?aQQm2^Y#h2$GjnGa0(ZqIRTUYi|5sCJP
z8M#xw!MR50(cy+UXGDB1z<l<-)PT4O5x6V>7?8w(LlLqKT0%fJj@`3JV9!!3ymLlZ
zePa3D@7b!?!rEQY*3QnW*U7Ynu*Ejp<&UJm7){n2RWG#N%YdEnl`T*it$KX0*i0?{
zJh;(zP$S!lhxuJ-t5%u^#1oYUJ3M_&WT)1mYSP0cv{h||$X?Y8=@<pU>@}bR+iuv6
zKKY=6C$xo>-5AzhBifAioH1vzT+Q2;+2VTEpjUrqq!I(ex~G2e1*Ve3hG(F0kHW}G
zkKez9k`At{JTz0v?G+w<_ghcu@cZ7<UWDo+8FqRmh_8P|2_kQ)7&E@H6-gQ}`_!%3
z$ry;gQ5E(kSCz%%<V3>HtbS`t6UT`a9>_XbxnR-_FQD__6HE`-AG`jm7O}7%v^+UD
z=M5)O`>ZM%Ev?GyArS5T;P7y`0k{gZ(!R(X*1o8$U%ayYRU|ULa!pZHlwrpjPuHxj
zmweC&RA$S_zCb3I-|j%4Uq7Kwb&LL~;OAidAMLRJZ{+~T|DzmWW?=u%dSzMvr5p&o
z`Jje;AnRd(=gh`*itP7CUabiD!D@qUL478axZ0pBZY@#O|MBD#CSItP5?E_%N$75C
z@wDB<^Ns;GVeZk+$(kB3G};xaBc;xo%vOa)Qbk>$kRr_LJ+I_AFmMwPKbP4tXAq*Z
zWT1-Y0Uny={D_0{Vv<Q(t8}h3EZstv;7{!uaYHrT7_y}-sFk1Hei{#wbDGbI!2Y<9
z3}ShmVF|S~6>y}Wbdh2CGlawoRkz?o0S%fYnY8?3R|_triG3APsuN0upaviayK8Ob
zi)z--tE;xPTaJy~fs$k=+#GI6*PGg3^3bXRM`N3c^h^W+Yr{1Ky;yctqWT~tSfi_j
zkoyK+rzBlhLyZDABB=FuhcN)jK94ea+ErFC+Dojx(sE{z#$rVeiK@8R`LBdRH8%Ek
z#pL}mTJwsY?aiHt?w|v!G(AXa6^}M}X_b>6R;7w<s-3D4J#30q$KQ<=pdnTDcCqam
zg&Hx6;*w+`Vf2nk1&UsPA6srrXz3v0_El)<kh@t0@OF+u?8`^ZlRp4ltz8;Ab(7{N
zPTO0@K#z<|Dw51}XRvu2(<#s{TMAu|$D}h_Zmq-PU8B1HrVgyV3vJW2djY$mk46(b
zO!}P8V4JFUvp|tiGOyr?uQ=K2wDhIxY{{wB^;ZVv6O*p=4VQD1$!{NXJ}joIvkhJ5
ztiN7Rb;&0zd~@OD?3wYPjybHoq`zsIZ!9XeWcMhc?B4b^?aS1^KI_x*@bT%@7pIzr
zQv6<Tw{QCfAGYs~9uG|D4V@(47WGTb7oWTUpMQ)Tn>U0-jqvMm=3fQWjIEz_C{HHe
z77zWmzDMV2^l8h;_p`2g>7HAyBz5LYSwC(OKXq${r;iJ_3H6huQ4-M2kYf0X6D(QQ
zAQmNQD&OF1n~gnr8u_()wDF3>{Jd^|ef0vmw0t-|7Fj;t?H>PLG2qeq<@j>GJq``6
zM(vj7!KvAQy&67Vd9-i<oE@6O`?~YscIn~Eh+Q5GYh|gYgm%vMv($FSiE!yu^zr8d
zaC#`EULIXMomg<I)yQw)X4RsUx4y#go0jk9^kHHB{_N6z@!Suk7)_;}%#Qr&1-mcu
z$ithBi7P->p3u*@&AWOjJd*Wc{P@8C5~1EIVQ@Eez|HIN>-+MWu=@5!MQO1~MoE$!
z4s4cO7^Hn2?iEy;M5?k&ULs)%{V<(m@agmJ;{m(?gTsw1hMjZTv~qC5vG~2;k1fN~
zt3y<Yj^BrrO&y|6dTgBTvZ3EKzli(*>JJl-`#+4GQ*bB$+U<jhJ<-IrZCk%MnK+r)
zwmq?J+qP}nwyl%B&v{S1Rr`OjySn<SFT46#-+I>itimI*vAm?Ph0Xi-k-3wF-90n!
ziGfVK_jdl8<E6a?Q=J&|sl9<+2UjP%dUhV2f{#_%u%#>bP7SZ~Oh~D>+`1@ifK64?
zs9T0{6%UVbtac!TJ5z#7gfV%GoMSO-ycW5P&_Rw#F(XRbYF9)}SwQDa%DKn6S&WIa
zGoRA{y&2B@q*zt5NTuOWAwU8RZ0s`dB`1%$!~L%A&*rqAr%G6N6Hc912^!RIC%unZ
zvwTWtYfm5Y!arDtbOi&XTt=QhG-o8d<kC!-Y1XnEfAVa$<Ts+MxEvMJ_-3wRHZ&BM
zSoyC0=vP>(hO5?IUHbbC3$x|k?ebk-RM3=*rt`opv5byq85*^W%Ljv<6k(G%z06`A
zoo83+w|;_M#?*pnyz}70SjMaZ*1&mIhVWYO)p6WbJ;(vyUGQ~r#Mj38t9pUvD^nwS
zf?3Uw7)yibKaXZ;IT1^Gy@!*=yh1;IKl=)6mQV}!N0+wdPa|$ODa9Y43(RU@BqdsB
z%+N;+Gnj60;p>T6U<=4J$u$;(Y!EW$He%Mv-(=^P6Ip44^=Q;aXG_lG?~A#7CcnCa
zlsP5Z+|SHI;_3870zz4qs$!PZH%-zmBx~(W%KpO6iAtm(t*ww!NL_{vkxJONC6ODk
zooB2VUuLYlZmMDFE`a}T09LW|QO0sNX5F3uZ(s3Y!@4$VYIaVfuB`Dn3?*^H!Uw-r
zUt>4tfIi3oWdMQNAB(7qr{{0%(fb`DyvPo_RB(wItS2$;)MBYd38Fk>X_GwA-`Z$P
z3Msidd6B+T(b$l|R1UA%RAVp7P;s##XVn>8Z_Z`BfA@*`7d7j7O~q;QXT$e9SuYpn
z#0+m)T*gkGT#{JUvC@=ce2aAg3PvG6`mCbs<-+o@lzr?XiEAYkgp^-8d44vGFI7)Y
zMR~w4z96OgI(q#z8l>kFvj6uNhCm{8@~@9>9t`reCj}<K+HvV(R8irP-vvunVn@<T
z-IAMRKq=1QC^J@a{H!_&DaH!)5%H9=f`_C6lj#EX{Kb^Ce<TOL*M#H^SR58)8aG{G
z?Lizupw^}*%Z}<h%4tRaRn*FVa{&%7B5C0lqYFj^bXRj|;)cVq@Yh^M<%e@fzUILg
zl>m7>QV80X1v=l7T3(hlRrm{6!c=)bnJyiB+4Sn&pFj7*JdZ-X;qTXCqZAv1)HatO
z@!X4MePGcY_RCIM_)W8*-QQ@)3h{Co=SvVZs%iI&dJP)>tG9O0h)dH;(3XvT(66qw
zsNYo`8xo_+#}EDkSQdQRGER~86S)mC3bl>Hrg_p*l^^rx;1YF`7OI&TTVO>1S-1~6
zCSotfdbAe!iUJSZ`ovc73fFBT+~yPrtJ!WDOmQhtQ&W5zxcma&dgkavQ4HqMY<W=$
z9r)Twz^XUqwrpi}VbO8XAQ(@&b+TXsq|{#iL`rDOaguBfm@N|Me9P-)%)_3l-zNl(
zu49|;QEe^85y<m&00WTFx*$#RpD&(3$ZdEzvA@7G$gp@vSh>$!3?h8~^7(v=$_Ui2
z;Fy1r<B1$TTx%DN6qHhs^-+gYt$TjXL(W*^BwRBQ_jYLnuZIt)PY#v~L)N{F9dqV^
zE|NM8K6rHDQR^{4XeA9c@v{`@;@@H`N;Uu%wx#5p2HaE|g;=hQ*}I;n(yuSO^2N#5
z@v8OZ)~_w(sX9xZqe=qI2m{o0w4t2Y{{%Sm(vcq$k0T#;&E2{Z%s7Gf`A8=&d)GPu
zr|@FNTCC%I0bfMLSN)UR;$Z#<4+sbSe`oO9z#Or{V!N-b+OFY9KDccL4w91y`L5hv
zY+?>xtWZR;oyiaW^xE2EB{n8@jls?aAKcryzTg`2^ZetdRn^#3c2&hpKhzNes9AF-
zy$S5zQrOOA8J8HTNIZY^^=w8FRBLZ^@kLbMJGi!va<1!~p522ty(3vdhu=S_SG^d;
zwfq?m4aT`;Sh#IlbO%7}#L-T*spQ-2)Cn1PJ5V&Wvd~=w7PErR!mRNrq7tw*Y*&@G
z=^iuBTT$(2()qZ-WZ6B)73H@fGY*g@G>Fb7@KbehmVnKhO^vBvtm#s{=&>Qu%?BaS
z4>NPBua$C1W1HROtR({aiC~M+5eH`GM#$97T4gaR8L%qafC^csV$e^55mB+u-&U8G
z_yUq-yG@&LGlSVG!H7t>h&z!T@ql*43YidI|9lvo6Dsv<3`fuaCv}m~fQ74N=vZmx
zrDCwH^p``K#(XTP1?QT(>bd5=bYsp(bcX!s577Hd%Y0l^BP{#CdgYW&PyCiC-X?i+
zO49;k7kg#6{0((Xe46v0Qz2R9Z@Ln|4sE!Zn;Vs#<N7k>l|0pL!KfWTrp=DN6a)T0
z9v_^Yx>_-6jNqc4l6`YA0viVJjq2BbMa^?J#ldlY@Tj<&u-7Lo(~P)N6V^?G6EXf!
zL?%$8Ll05Ev1u;7N$@1mCcP1F7wgmvLE#I#awQ(*u`IN8e9(Gq`!%HHFr{U}05eXJ
zGFX+TUAHA0o06we%q0b@d9c1Sg?Tr~ehH+Fcqi&Uj%hQS%r?MutQG!}GG@hy;mJNJ
zU|Y1`z2CCC4mrV6tQu-h1kyAc705zy5j_Y%T;GO8`2E=>ZzI*A%pv|pZytJ8i`yEl
z4ZHF#t-1IKgSlaN#1HW+As3W_B=rkMNLjnCSXQY9ezV#egGTB!y}TjP4>cbarK{}f
z>(k+@Z-XwDsGIMH@23AmB5Wp4zL-=X^fWSWkue$jDk*)4qqGKd;%;wD5lVBb&5(+C
zu)j*_%43QO3pf~NHZW<@L$<>JUxgApVQ;Pau%{I|j3RO%x%`3Ox55Z25DTD_yBAk|
z(`+Q1$Fp#pQejh06WBv2;3hkdWsE6ySdL+Xu8;!G=`p;t^(WcV*0dNHH-|20@S%+U
zwcW$PC_1P?LPHf%ZwMui3Ts%hKJ>sdi{MA<Nu0#IuFKn-PX_mrH%=*Imhsu?KCGLC
zc--1T#*D52L5nm8_G^2;{x4M_sa<N|eRNaO1__Jh;_7@Q8-&%#5F0}}^g*o8>NRFG
zAq-o&9z&=b!-(76?}zM|aMy6R!P%u^u&`{TIQ2(^2}0jU@Cq~2I`vTv1_ag>%O?8d
z{tSj8a5n`@uk`~0BE2*-1GS!Tw4Ad{g=_sbq|;$N7Ma*w3?VyH>I;JBlK8oO=v4f9
zC$sY1NP1K^0X6lLLecrbb0D%nRY#74li@MtO#(M<%qB{gBuD#pOp+#SYI0DBveRUa
zRT!F(zrmqkg1oh?B+H9dP^`Qwp(V_wxp}_|fx>2~VH+@bUjV0RHVWPH!=_sgzc3GA
zeL}?aPN7PyUoWTFFofy+bkp(}P&lJ?wy^VaVV^)_vgC{V9O^^kydWIClOqRvRE|kZ
zTY0Tw)e{9jck*<*cE+BlRcxtu>mPh~03i!8r;D=7^8?#kWW+X62~hb5gi)zt$|uFo
zkN88v@#}j(s>KaCs7w)m4~-wsp%g_E8Y_A<bCnNUkOz&{LZ8`(&FMjuD+kwc_TaL%
zZ%=kp*E<ayB=wo5UluOd2AK;M^Zfda22_pO3*Lsl{h1Av4AiV}8VEvdl%}bbDpZlo
zgD;Sfo?y2#Cn~cPk=jBDyLafTz@5+$sKC9jpcMKQ+LguB0Z6tF^PIU~p8H$xEIk9?
z211}izhQ=K!<$H!VbvzsM9U)36!+H}U~mxwwPXv#NAFe%r8)$As$<%_fT`7&@)+)R
zj`(PP@k6l^I@$D~iK{2{x02T!T&osxjy>n~!T4y@MZ6B@)?-rNv&Kp2EbR!Ih775h
z8K|k&d8?vj8*{iAuCmp!=2#h(n)_4#9ZPFh#Q+6j0aIoi1$y(q{^PrJF10*N(Il~=
zVmhP5NmvtxOxT}ABvE^uaa#Bx0tC>zqfuJuHvz+QDns-@nz)(PH;_hVJ#O_lvSD_{
z=(=ZX7n4jrH4!%m*%`w2bf+b+L5B?SV})Vqt~26|#y4r_bz_*ON~j*ETXz7VZ-<0}
z><IO!LE7YmigVV=Pb>{CnK#OOq3E4DVkOGJqzm=1NB^v3gR&C)xzkmen`YB{vQgSw
zSxN5wsrx{+^IOeMSshF0s#a~|wW+t^Ss|6T^H#s>01#(1>y#I7o|}$S?$(gH!OkOc
zs@dIk1D)FuWe^)gvXA{zX?rDO1+^6!d@^b(p1UJTDg_shGM$ogap9`dqiXG>DfxY<
zQfT;M8LSU)N=mtfG)1AJ-J^mKxAU5`h{Zi1Mk)>&PCVq=69k&67jNwD2DeeNjhjgM
zdpx!_d$xZZ-v67j@c%McjEoHbopq=~L&}<}34ZfPbsVplGUU5pO&8xb3TAzSrk1Le
z(yjG$_wkF4_QA`=h<nnuJ$ZRioKVEeLn2i3Q+xmRZ0v1lYzn=509tJH^kDBT_c?IC
zOVTbcXMmCfw)kK#>)=V1qG9JK126uk58l`oj`!>dnOvkhc^sbO>qxGfrRPJ-J+Ne6
zJzd^i(I^V`Q~l&9{Z<MdZpHaYRMc1d@c8Q#deOi3qG0x_&xcUNjClR;gm(z3T|%!F
zDIg{9kuCU#;GU;Zqzx1a%|y7tHCqyfwgIGehM~jziE=DH(}RHV4lQZKa5#(AIKq88
z2f#bYs_8**TJ`hw&Ds6g+R)gSb}1w?B%>FDZa~9#Tstf28U*#bxGlHg8am1QOTDw}
z-8qv*%K0;#KLjLc5D^Y!NffbV03TA~HyAIZxr{<Nl1cz~6-_0jD`>i5;4I{k2N+o`
zw&664xlMLt(zR$8D`Gzxc`8^${EK8L4UgwxvZNHS>>0B0%nRJ3!!D-QF7bGlkMb7y
zrbmHr3cB<j;c;?gtXA4YmuWoFYU`oIuEFU$tO~MQyRn?s>8OJn-}qgksw;q}ZNUg<
z@P_x{FLyi3$s;B|>+|!+$B)|Ck+bA*_^r}#5L_dfII*3aT&h&7LteZ4MF}}H^Vq8p
zFZxg2{T`?hdt*6jVH;RZ{5@E2OuC0#YoDFdv@NG~z|LH4$nYs`i5L&6Gb0BhYbK7E
zXnw%GqbT`rM$^S&m_&l)px(cu>qX02PU<6E7i|ELiL-*>)=A&9V@|$-=O7K1AXG!L
z!LHGYTryvNS-UdggzaDN{14uf3%LXWCTV4Gp>($kmCi0CqpD;{UIMIOahba=WnmH>
z&@QyDo&zxhP5x4Vk2EQO`P_c`;Q#?2JY+0~PQzt#Eo-Vj^f(o`39!Gxo}JchTu9Xz
z+Ak)_S?Q9N_uvYsZkX*twXl>evGM`eobIYhL$-BNW&{0pK^%2GB`L~RAnX|Dh$L*g
z(u*a0Lng<$mb%b3dUN4LjVS8bhmX)aDgp6>Y0*YImTL^-vp2X#uJV48PcwA^GCWbk
z6N{iwLYj*}+X>ZYMF%&k$ND3gGmG>WjeEE+vg{=dRtxJj`p8+iF4*T%O&}=a7+XUy
zG@y1VOisRDW|UekXD4Emxh2G7IkGayEyrraB=byrGf>B2&iJDd92sGklPCyjgk{qE
z^pM16nWH$tjpzmn&$1tGU$8d_@n(j<%*EpND=1Fz>%}_w7soix#&Bl?;ZZAEa-W3_
zb@8U?-?lQ6lZlM4{H;zR$c+$Wr`Px?l&?js)pTqCudT903=-{P6@PnUQ!w0e)e$I|
z8XNPA=OOHl71fn5^)wvk`02q*1&s@0*KNBTGct&Gxsyv-7gR0T@P{aZ<R3AjvB;J0
zz9f@!66Dk;;<HW|C0jShh;)v4*}>=+xrLHI3)Wh>jXg6Je^piv)JOEAWX?UNPYFq9
zQeLXyDz@kiHvh`}1xtApYOUaEwIc6&S(9^n-&l;jXCx!<`eY@0U|4sk3F?tD<bYv5
z?u*)f#(^;T)?gbjNqcIR;hqL@T#sL2?CP0owmO~zZ6K&!YMz)@(AZYfLm;5R7UhAv
z%qn0cp!-*iNdLCCcc`X)LN{66e5^G2Y>h{EH<M2Ao`=3BpYCOiz(kq$)`=gND?YQ~
z!7G{SoH8GSYSmp2?g2BZXQ^dxSY>c1)G71K<)S?!<h6N+b;1K^CY6fkz{Jv%FY-9x
z@U93EPP5wIH>JksSnZf2ccOHdhKbj%tE2eTDX$bOx3nyz*}h7j0#EE1n0&9wQ9~mT
zFUj5NUg^p2jX$QaCEE57Lu6Illy9XACX`}j>%NK)ZKPwT*T?8k|MqYj+QSCfC>r_c
zbeFKW_Ka%8={x}loC~V)JW3~>c7%@wWj_8>Gt`gMy$kEqCfj+QrN|a;dwNG{#8ECT
zifaRGAxf0V?;0)B_gf}##W$L!k?~uupK?G?1srlArb;b^<ouPpI`W}Apb2nmLZ97x
zK84(SXIU!0$~~k#DX!dVsroIm7kQYtCFrACwSj#(CuYO4CrmrqY#OVcu<!s~vSPk}
zZRhxFsaA~$XrB8?*Vc6NPflaPkvzf+cmPA^0Z!TC4|VZ_%I2<wIPsmHsXV-7OY;d)
zRAPDs_T;qTvOv!aN6V~A_8x0e-an+d6Y8Y7Rg^FDwJ{n;P3}R-#}<aftB*m+UOVEB
zBt#glLxQFmPZVM5VE9z(n9JO1MOMZX*cV9mnnj>NY4AiCpuVSZHhz0PODI-J$0=(N
z#NWAgqpPD71R$JvH0UbgXZCzNKUelb0mlK6{ukn%@H>Z}oBAXt4wKD_T|6Ffn|;m$
zqQNMdL)i}JqZ?@8YXY@eeVuk^Ts9g`ZH-+{us!asVYlJnci}vBF|$@^UU+D6#-4JE
zmMusV<L3xgDVw9MMRy5nYpW0O&*F<6wh%jR_K!x7`Rkrs_HEsDq9!s^K%4v&8G+!}
zp5{jd*Ngnh){d#Lol?9#bjGEP*_6O^P90QwnZI6VKb}9O2LZ+m=F{H|NiFcTpvk*h
zZ6GaAj{FO0UR$d;UG$Y&T_sQFthD6EOAHb?t}Sr%0M-WgL5z2C3gFR0m>~{@O^EV0
zxq&k<T5?iaUY(#99;E{iN^>g&wHBWeW@~(^r_`7R`Qwew`P!hA)>glv9O7`A&dsBF
z|2CNpax2L40}yBq%O%IOLh(YbGVH1ZEsfVf<z@=;eipLev_m{iCMPz_4}^rWCaZAD
z2*Fc#tMJJu7F6TvLB97=_qNa2^}A2FWJ}eGy_vyb17|kHDr~2uy|d5a+sm~pQGD*R
z+qMG{%IEU~0$)2*Z^G1|M5{R=HAu)b$WoogD|wED&nW%<>$1w2$*l6!FVHercJhDJ
zJna8#KlDGjb0&uWE_c45Hu9ZFMEu4C--Gs5qgnCy@s~^d4<^{iE>-C83xT3Ck$h~f
zqH*EGpITA#wC1!1m*Qk6O8`MOc<yc|No===!n%9s`rQ(!Wcn6H9ao)+9mo(W=;*31
z3qd_E;fxsFT!uI)?o0df@HwPg3a0wNZ;x!~+QuqO-%^P5saMWDi3<frTb=XD0oyuW
z%FUne^WptX0R0$_ES8ZKcCqFsGhER!GvWmY#mz4%yOCv4M5ZG~sT5BULpjNugg<SH
zrLs<_m{+#bp$E4HCeH-oiLzbXA+kK0_og<Zeit=^D|K|o`;$>CoH8@5n&@l72qfi~
z;gv3iG^xkocJ0;a<$@S$Macs*DmSV+SY`ZaK{}$@e<7F-pYaUg_hI-3-@{Ljslt)N
z2V1=yo;CB>L&PIuI0(5lZ?(^m?P$<-M7vT37y4n=BnM!EB-d^rL%IiO)JLB8Nu~|T
zo9vnd^b30iQq7f$D~YQVlrzgIa%6xRBC^p#bkK^yTI#HOrj2`rjYWo*p!BrBE2UZ0
zAzM5fQI6cS6HRwk_0{`kQDq{{%Oz_Hf1cIYbZbS>NowY)euLvZ>seIDzp~5wq{>)6
zt#cq*Ux_v|TW77Og1ymC1h6R?Zu_>j5NiB-$j^Fxib1Zg_s*^_uEqGwY+bx%C748%
zHO$AkOI0xT22vB2ZKOn*ygNrzVt?P;d`f7h@w_nV1^fm(kLWw95295dudZ$OaS!uw
zn29T+>H;4zd0=3ChCPXVaW`_-e`F09RdWWls~ntz5NNBe*C#!Cta{w#wz3!^jpC^R
zkPz+#O54BXTl)l>z6a$`)!M;z)5LkLx+~)*r{CxOjY<J=&W3`P5$v{N<bndhBBSyq
z<I$xOosnUaHdgH#n4*xzIsh@8%n5c9p+Qs={eADk)}ZSUO6>{Raa?^_{43)iB9*{a
zRr=}Z>eRwO&+XLqoSd94Hs=0bx+rz=p*%wCur5B<m6A_Uvap*E5c?E9YQK^;**0On
z8^R2*aO6~QchafJ0}Azb_g{_y<sh1ok!2UjVbqAapc#k=m{7NkJM)<exZuqd70%VE
zk1Ito2g^CwN|CzqUZyZk;`nQRTOa}4QrXv5f^Mp^n}m>~jkAceP*&G2-yt{I#JLTZ
z(u%o(7>`)d)O<8jbUQ5Ta^J-bmh&Jv+@OUXeE@K>*7^k_H2CgRXX$bNqUGdymJ);^
zg+un3Pw=7qGv<-f<oK}3-veKqIb%bD7qNW&<Dl$zz-x+EySek_Lphi@G)~8-qmYUE
z4Tl5Qk!MrJq<bv({!<82CfHFd=Y-lC1+%(1o#9Q<&CyLzbf^*n?Ew;)i$~BW>sa2j
zCRfb0zf`P1>6o>otx!yVCcnnhw}j_{!<KBEXc}tFp7-03xe(()J{gs-HGZd#pCX~A
zR-e(eOWTd&Q8+B!@0#*gmFN#M<;1iGxprI7#PA9myE9+x0K+7nyEcGvC5Q@<ABX-U
z*^OrLZ;Z#ZQdeNX><_g3{#F*XAK43=$+cBPT0*{xy|y`h*e5|?K;Z}pOI}n}h7#l$
zCYK)R0_*)oq*V_4`wr{qHYsY4Lw0f*d#=u;SYvjq0UYHlxMTA?X)L5wmQU{5>rZ!L
z+H9`p^LZ348mF(Lm}Ojnx3peF?pKbXK=B`l?FXQ((&V0)n>QMxX(5?gX_c9bUS9pt
zo&tY`vnD+Ikn+|Z$7*)@es!yn{lnjufu5fJKkT;*tluR?|Gwx%O)VS@?dU`;^c)O@
z4fU-J40(B>?H%k4^(>)XRu|M|{#D`K@vjPRVN?<YJbwCaK<ameH`@Qte}7ka6H7K4
zx-+I!F+)@hH45{@%DUy&x;^tA@0|60KI|m<5A>ArZ%<smS^|A$zvxWA&(3TE$Jf!f
zfByRPm|bP>eDug!T)aJaao*;-z@_<N1q`5F=5Y%ss;O|Tw4_eZCSBaLgTNrLk-}>S
zJZ8S^ZX!B=Q`sN0!CF;j3bIZX`8wKp-`}RO8B}7z{E4dg`>vm&g71NS<L_+gMMb)I
zj`;eTg16ku9Lp~=`_O#)3NwgO11H!;P{5DMYd4pw4Tt;!2szxhxwwKz14TNAB(PLB
zv0~o@R$sJ2P0mgSpAV$D!%~c2f>?w|oS-ob-;R!+a@_BRme5%D!tEFyBfG0kQVnw-
zuCv3nTPK&J8Gh~&bP}}*b~Yzp{yhN9lJ1Z`!p${>y3<gljaNXi38<N{D^NrJ1m)q|
zv3L8agbK!&H<kx2%M`hssnGWEBOwP-Iuu){Czb|DtNmloykP97`TeNxS&cT4k+txM
zivlh>(UH6W^%-|qT!YLz%VKr{%>7sHTjSYq$PJH4;X@v9I$7$LoSP6AsI0k!spoof
zqBJdpleJlt8HDY`pW67hOXtQ8^F(Blkp5Eiu!j{Cg|G*5>*5E?J=8x*nP7R_X}h=|
zsJl`+DbyoggF(C2qha_XypV!!o4~0>gSZv!+N{?e)%$*XZDFhNE48l&%i&0a^7WE;
z`?32m<3Oc7wqYsLR9LRsMWTlKAsC=NBTXZzG*g9@Xmi&?vz_2*y59|Rl|kbh{bQpG
z28{AwshaTe116lto!FXM$FN_1MWgWteo0<U%pHmfBb=z*n-U686(Fg!G5W=HA@*v2
zCY2<dIGlzfCGrJW5mhx`5Ce&E;^A);$qj-ZBB56zJPHMG?Q*_No%XL*t$lNh%V5#@
zZhz-XGTZec7_mypXi5pim&;z`t3)8y;-<~t6IuM}9vi3|L>)x#Wz{qDH7yPdpN3o9
zQkNl($7CZg?dFH+au@$J+(!@c@Qb=EqYyd>#4vPRl!S(DP;v>LV{*kZQXS7z$`i^A
zh_K4$78z5YyA>2=oHGdjn)`b;cSe4NUoD5V6@?$e6-t9SfUM_x@u^7WSIYzIxvh+F
zRF+wub=-DBd|@O&Q>zKG_KLk?6dDl$rP`^&If~5M=2V~IKCkQ^iDFDdbDc-*(T}ZS
zaRj#`;4#fn2~*3V;&6|HipwD%7AU|<7dDWCn00%r6k-jvDe#`G@`M|j(&B=)1y`$g
z#G1|xK7vSzXSsFv>Cpr$+y)js{+yl%w)ZwXmIkIT*``lT_t^AXyL|G+Z{iYd>?oOP
zY-zbxmr~5{Pug95D|I9nsa{#yWe#Bfw&Ij{pNG>LdWtR%v7M1syFoe?2O*x()K1Wb
zV@Kn_@eisY$M4jpe0v>ZWu?$G;_Fdml(u4z=B1u|xOdxH1RgtuyH4)-fdg=enD$K0
z$C7UY^ysbxjP1^en6g^?v`xq-$0Pw)xpR4KYs^%6ZMldcjwf{+n-+<}3^jitba}KM
zX*S8iP=4*b>)v|J@(5u|b+h9o`8@Z|x_r4TWg_S1`5RP2W<7VqVq1o;y3OdC6g(Q7
zEUBoA@<6#aw!6fW;9&m@9g*zX)^z&yi=22$m$DU%d>x1LY|rj`3KtIbZT_E{)}@}@
z9ZM?{sI1`ESoS35h#ar4o?Yw}D2T8ODm)U`LBHE9p_@awD{_}R->G>n(DsFzoyT?0
zGo>0(yPQJH1aBFl)m$bHB&vLM`S#R2|A3_`(_fw`<QraEL*lUkbd0GKA1cE<YEwmq
zC-*4Uk$Z)y;4!M4o)V~#naT{?$}AI*JCAj6i!$}$LPNf3Ts7oZl~wva=~R1H`T9Np
z4xY--x@T?db>|ZFy20DS93cnk?%K@<T#tA!U>`!}EWvKha%ZJ#t9>RCYV}N58=_gO
zRv;BMZKy+k;|0ME_V-2WlEetHsKXtUg<hv*wLy2O2FWt*2E@YBz*j;UvrOfRy9uKQ
zQkAs0QacNgmoGo1^4$;OLc_>%oOW=znDyN$OFi|+%9$1__qwry0BWJBP#KiJ6ESH}
z)qCH#d9M4WuDbSbLdl6a`#(-7hW`wRFmQ1EcQDzB{|#8*=>!S_DkEe3ZVl-K;{Z$f
zzpWuCBZO)Obg_mb!#Nu&&N1Z=%BU+IuM9UQ4>oRG9voCW9ETA$c5ZLHP%b??RomCo
zsUuK5<v-`vrcS1eCkq14ek2`gQZ!?~U+q{$%1XLqBcjpSdf2IRv#E2SV;bI)V~m=#
z!)W$^7TiS+y%;s?*1K?gK0Lnp5}oNx4qn|@W*Bwj9IS~q2guPGPqK`c?lD17o%Acd
zg2_5|Vr>sC*B>U;p|_c1^p#g~f*74qV>%S<wo-m2+V2Bz4WRPd;V^pclkp;hxQ6Yy
zV7rH#NyF~Y(?ErEQ*hQ0A`HK``W7HlLvqo7*)`6H)0R?WM^K)HvRZ5AmBWYvMD=<^
z+xpkeuFsm9`nrNk8A_6C^lE#Vz$8E3jtDxZL0oL~Vt@XgovZRLyb9(URGYcp<9V^E
zMQFm8GR8Aa0cNXFV1jY3m#tdUK!I7Jm^El4&qa@Xsv2l7#)HFN;Op{|?u)aoQWjBw
z`HoVgpy})_1oC4c{g!jWM(P8k9rqtk7q+Oq%WT+XIJ?kStuzH~gF9r$L_A4DG(h&C
zHl!Upvbv=$-tc`$KO+Tn4S%&vla6@AF;48CKAVin9g&D?nFRbWJQSW=*&8ovB?Whc
zzaCJOZIQ)B88XxC+77K8FGFV{-|ym)=?th}8c{f65X1Ojp77O7@zGQ+Wv)!Z`5<ih
z;&|!l?S%)CZs<;xn0A?u8Bl|r+@~_5He$+Dv!V?D=`EoyQEYH+nocEU>mey`68$Bd
zU{OeQmFyI2UGT+eJt7<0WFQ$!3n)eSqC6OjeTCi<1nkb;Qjvh!kKg(>?AzDC-*f|l
zQS*%bCcYtpGCy_vMoWX58HiZu;*kz_F>gZip4%;XCV=};8?<9{Zq?6YX#O@60uvkb
z^EbCMQtI#Ro-{385L<xk`7rjq-h%wS%72r3CN0pnqOff=4J~mWA#IQl8IQDZ|8!D=
z;PBlj1P9wG?Z8VBdRi`@3T@-BbZ&JE!*@AyK1<{TUK+ns*`D~-W#v0$g&Z}P7V06@
zijk}HGk^S^vm!ok0m;IN#hareNT70?+wH4x0f&1Z28vI+@%)q}Rk7Jt9v9=uO1Xf9
zLd*!d-ng1Hnk>vtMM*0{J4+nIF6JRg{-wC{&tWb&knEf0v1+<5T(v`VHdu??74-g)
z_Mm#Oi1IGiBO?Lzc=&ArTf)+7*(qHVHCU=eNqL7vo|G=pGo551xBUqlrL5Mdvy9kC
zB*R5O{4EHuA>SmXR<XLZkl%aea%p~^LZLETg?Z!TT!OIp1%(K7nP^$sgt;p@BLdb5
zRDeEjGc8X{E}NQc9`jN9Pgff3+0+79<OY<JY(js*T@&1}0Fn27HS}nuMa`nE1^t3!
zsW227OYkL0f5|W}`4*e}3Q!_`*P#@~MUCdQZbN{6nf2qWLydnR?v%<l-9b7<M|H&w
zXCp_!+M;?bYY$KA1+Q#j`g5e6UY#_3qDX<md+eurOCCMu2Z>&BS4Ym`I=|3|#7BLk
zyWpT#`TXd>K`^HD0W5QJElc|lSxIB8r<Xm38Wl+oUwXq7nKQ*TvG7lw@1=|A82bg9
ze0|CzL1}$05d=MF-gL;mGY|9rUuy*lzZeD>ntNr3R)1ZEl**5h0#)@sXO-&wS~(iK
z`Z5$ybp-RM1I0K48f$)MSQ@JdEFQ|XF~`0<8zljA-4`(%C5gU#Z<NT?MUC~v)OIXu
zRHaG~^-w^i*iOZLrf&`}gtg(cH29P0#THJs+Mt!-!!&h2*JEXV^ir>5A+_FlYV<-D
z#cz^Rc)}na*8J@41XUOGqmiEKZ07i)ly;&5?u1=T+{HAMgEOz#yt_i3g^RX@^Nm;x
zUDi9GKY15)EjtaVsda6;O3vfHDzEhiy!4;^_d%?&A&9&_n+iV;Gd^_Ud&(rYc%2!?
zVHhB<$<xtrL#_3Ap^Cu_z`l|hFG*3*L>bB2kaGfPB|2))(_u7IlGqICG%*!f)rdJy
zMghzUmKLP@O<D{YCvSh;>iAiK1?OmqUHM<|aZr@3G*411lK-}+jvaYmXXdAJom7qd
zNvXwBWt|(Hv>T;WcDFT$5_4U5`AxG}rfkZ}V6;q1Y0RQlR#s3AG;LSu#zymOy-|xV
z!n+|+Va%)%$UCW>7N1vYx4bahiE_rYfxWUB%dpWXSwzN_S!|KT7SplmzORb<Cs%EG
z2k_RLL=Oi0!<K=zvu2GHNY`^MPkv~9{URIEZF(`$fQyBwahV+f9hwoA8Kg((NQKve
z-InQZOH58q)$HcOcdoxb9&^s(^?aa7Rwp;|%-t8SA^m23oRD;zD}fJ#JZiRy7IS3k
zSth(KnimzB5cb`dv{9e_+_TS??<#;0dUVvcX37pg@O%4@W_$YF(rPZ3-7o`Z6j_h!
zSiIHaYl5Yha4yf6|7>3X)<#sM$W+Vb+-#9jk&{`OU0u_e3UsKN@vGxfd#I>xa4Cyq
zUF24csU>9XsR|g9lJ;)BnTJEL=)=9+$kL}OR9j+wqMBzJMO)8RlOccVt?N3NOLTY5
zV_&ZOb@h9;AE}-UXIwt{QkpAI0v%uTvMp%om|d{?CtsKgC-^@I19}do{|-UV`2P{~
ztp6pw@<7AV8b$2q=cNvVLgzm3^qkhh!)y^G1(X=ALatDB+$H>rzhz<o_6?zXNVB)M
z>_fS+OndjAsu&`{-)2VUtz~6r2PLN~u~;k7!G7MJ_$kG#)k)dx2nxlDoF)f3@+JPM
zzQu;sxdOwwvz*8XnOyDz=;sgu18fU#OTX&mviy2P7;AQe_IYuFs7a9}Ko1_YaQ`kk
zn-4FGbw?XKcfUiAUI4|z?MGvv;5x&&`d7)=vC73|Bc|LDm%Mh1m`86Lhhq?<^d!{B
z8R|BK5R;QO{Z9008iT2YXQ{%A%TdU5aV#ERl)y4;plx`Tb5utoE*bn_m=JQjh%&_>
zo^wdb_LxbQaYekP<rGek)N+8OFbgOU(5x}l?gI*X2<bmedDF#>;(AhB;0N?_UN$b0
zIGkIJEEdHIP{=Co&@R!kWy+Z%LxF3+00kK#n^p&Lmmn!J-Xj}eRg29Q4g2pnkx32D
zqX%8J>pI(L2?%&+@8a%5XZApW_%O;L{)$JKoSS6#owri2fYWBMrKqW(-Mr$=tWAFt
zcCn}cIj%Vv>C{-3JjrbY3CmzD5h&nXG4+R9F({%zYhq1BJ0qVvY={o@Y{grRS3nSq
zOzx=;iRPF$rAb^dwy_cb_jjk4=FG3mI5e;8|7FAK0XT#9mAo;Hz|P!nnd?zpvi}o6
zXyczvpnAi7)1NWuSs#v6{46X)zdQRP1)nq@5@eDuXEyz-$X$6Ftsd~CpRRJeT~;-i
z_4#DR;N|JostLEpHdbdGR#$`}Pl6Zejn{?OH2?<Ajg3bKP?7!4rod!166dJr%=;rO
z#RnoF43bqvsWNOM&;h=<ym%12Yu#$Nak9d{Jl;LuKW>v2P8VKuhHj3~%XGByisBBe
z_o}MA-5sx1cAsvB@)qd6TpC_K9U9o(y&P)AyE?I>bZDc7;;1j00|b=ShV6O0E*~|~
zUR-is&jTX&_c^E%jQ3tvMqKLEGdkG4I@HtmUg*5Ym0oap2R`k-K7l{I%@3?YI<(*n
zTaV4Qr78SO{m~qBXM}6qSAMdd+Pu4UZ_5r|CGyb>LG&eI_DmqM4M?H}NFfi~2qDor
z^^#B=lG^<RSxXW=R~4ZbmhHp)AygZFF_#wRBc4dT3rtplgh*ld!}aJ#TpOXFt*})<
zl%H_+FqFy}b+3O>j0)wT7x(8k++BbfGknzM_4Qz^?*cpL43ZUEXW18yHu2Yp4TK8e
zIcd91A7}U$tq$!}X1gB^JT?OC9#Qhr?zuzu$*r}u$`+3rP6e7!iWRqJW+nYdAFdJi
zlNzq*I$=PR7XcD1BFn4DnQMLBMi(D1r$#2XT6tz<6fQTmw|cu8j`nJam>3@&K4SWw
zgX891&I1E3-DFbL6aek$3>)LgOOxB(_I~14W=m+`wPJ=ql)bNxr}SY#I*pHKi4+Mw
z6F|YY$Y}V^qI@Skv9&nS#I2}R6U@eBYc;}YAP_AuZIMnrl!(5p)>4>ufxD-)tuF3|
z&HzYIxZ%Rttq*3uvm~&33WV8B0e@Aa(~MW#Y3s17s<G?Z8C^k~2Bj)jos6ts+y_*(
zqn4-Jx2%lnF<X;4>?d$#PIH)-x3AkAfLX~Pg)dRjosQfC=-TfC6+2dXcNS2duL;{l
zsvEdEtjubxI_obp`{jkKVw($#xf^qWo0b=k1<XPvO(9eSSo!@{0u$f_p8+6<AOI8;
zu4*(IiRdLh3R+()IjieeLc8jjSCE=sEG?KCleW;`^G*WP=yTSHVK*fGgiM?<7#YLi
z00U7n1!GAgN#jwYQR7lAInJq^JM4cyTH$~_g@3Y2cu}B+DDJjpgeJFnFSZ6lhv{Bi
zvSurZ!zEBDUOEC$FbFlevp61A8x}R0hy)Wo?m&W9kkeg>0<b7EoJX%ap^skin-`Td
z#0Wv0sU{QyyX^Wdq*V=v%?@#SAS)AXMcVlYqA^i3MFL~Zx=A4O)}mcV^tY}3z;ASF
z+W9cH5y%lrBx{fN^vn&yFmoFe;8e27f5h7>Z)HK2^FzWV5JLOWhn3o_I4MIC&dap`
zb7>g)Le1em00mzInW@*H(?bUNYtp<UNs6aJ)neZRWlVqG^ZUSJge0aG9KVAHIcy%H
zBMLAHKvmLfiwnIW=O&_@oc7|2Glq(^5C-IM`hYMRJDZA_<(Ml6L()U3uuN&}Bo`#O
zZbn5;i5XjKzh@>I^F@gb9qE^Fc#Ik%FgR2Y1QhfS6kU|W^kETqnNeC@TPp^HS7wyA
zcMZ#jBS~0plITLu&o+{HU`YI?{^LsIU+f;9BSNh_xY*-O!P#IIQmT#G+p9!k2c<cb
zFfbOu33X!~26zw&5ujO6xU!)wqV___%KY+aSm2w_Z|lG5B;7!3JZ8d|yXuFwj;lpB
zC)hrm<t4HDG~qvgpu~6jE+TR2`0>{<sEM;rBdw=uVr%tq4dZ<k*2@ySi`QR=&fVmi
zMaF0H>=RcF&e_A*TXsC3@sltZ2nuG&sqPb3W3*LZS)C)~G^?A+5f9S6!lR@DbPVPK
z#xMWC?QAuv7s}R2aSLb)@x{3C12Y(ou<ns+DS+~Dv0;iPFLnK%gCL^mHid!@aH9?(
z8h6`*e=XPMcz|pp*12vD#vaSHVcr3EfiASYLf$28xpBOh5OvD!TA5UxOgib^SiIFU
z*-H16(|M#}Va`8jTi*avls#`;sP$X^w8*994}^-sI2{+19Q0|9-PrL35wJ$Pq#bHC
zOBcNE*;Ts*P}<b>nDO%g)4SRCZGTlW(J{pOgu4cNv-<Bl^#Y+4NwRpIbr~*bpy{fg
zd|UB;*!UVztHvw~%fWw6EQMaEiIBqD^~*qKwMt&1N5nkflu=1JUkbxaNKI?C!im?3
zEdm<GVP$@`fzU_dV?7)0RR1YLr?x)+?qW#^7FlVPom4C=o|kPX|9IeZkRgO8<>o>^
zE$DcSwK|ircmD{SN(%577)ulC9bcJGfNf{WG4&Imw6Yq^>5;^4l>FCiHqc2g;wPIn
z$Cht`luR#IggPel5w$ZM8QKdGAY3)7Dj1n!ioK>wczs6GJMh^rHf1)YHsv<O3|(o2
zsUgZKll?jvuDAG{b~AsTRUAyan#R+qxEE5pm=u5uk?78jo~w@RU`HTq(0Tf>A&3}e
ztl-$}Cuy|i_WK11h!{|5hztc}3dDb@q;Eb(uJgWa>yqV&XZ%E7Q_z1yU-0|_GJ5u@
zg2Pp?+GrsVXv-GqUo@Veaqx#th~T5$cuoVBIpFr3wu=iGbv<m#GyP3j$Dc3ehrehH
zI6D5Nvhz;aYn7bmK++=6g3|ElL}VKxH4+5vLlfn`iSw(L;F28Q&k<fPY{a7qU*Y9f
zEd2$^P&^G(Vf?k9zJ$Boj|6^-mHRChI}-`TV=*2;kSvsq*gf=K!+sgdcI=<HNVwb?
zUZb}=fH}Nn+Q?605&b{tD*E%@OlM!f1<(Pk{~%cz=$TmmbKAnq_TK>+57eci2-y%j
z@2Fl8%7+Y9UkC{Dec`AM5*OBU({n8V?!%v7LZeeoZKk@(L2#ovL)(p7SFfex!#h_3
z@H(t-O%r4Cd%A%{)n+bltUz7S@94y`n3&W)^q4_SSAAC+6bU=VQwWkJEPRQ4T)3uB
z68?Lry0@ThvS@l-mCC`jqKc~p9xEAQ!zkEQ-)=seO<%lCxVH^F>n<w;qmJ+d9hqJ=
z>~N7DO+&AzSkgA_NO*u`81eLqmGizx*|_TV!>|0|J8F(nX1pboPs{w~d{xwYQ=|B8
z`{{@S+2K_}4dy^;B&|%A3#>n>KXxp&%txt@dCXxETqh@0nmulxgGk9<Y_;-Z3}XGr
z>ap^t4bpdJTuOrQ%Gc(sJ2PjRXZ^zQV10EFDBQAxgmU{C?iN<7pX43^dOo92)Gmtz
zQEvuVvnTG=TG3>kR45+mS>#|{GLIFn4_xFy#n3YFI3serv|TVnCqaRIF7cD@NP96m
z`tft&C^IUqN!;S*RPjB4%cVRD-24l8NKoHiDHN)->9p)2dXJONkK4u<<GHn9Zm8Yn
z*Sj~;c7A-l@^5P-uFdn!(zWHChPN4dZtVo?3|2mau%b1+-!Po0AYyYw*9<%!gP_9m
zI6QiUje6}szFNFd`Lw7i0|*WvSbc_-r1dyJFW!)K_TeFIay|84xTiSZzqPdak|Ee3
zKCit(09&DsagBkv)FL!pksAAUkNE^;v~Gn-uRI)WbVho2&?deC#~-fF)(%mdxw<5V
zkwqnG0iAHb&B)3q(oI2+l8P<U!o1l-FyC_W?rpr0qO{#Zl%+3+1$u-uOK5W9M21~L
z#!0|wz{iJClG~p7&w6a7q!DJ8GYQ_bIfMHd$jbK#Y3TKYW$w`xZhYn-U4n{J3oB*7
zP)xc^a{zhaM;APeSOgdPId`Vu=xEEh@4_<1h2eeq$lbv(#i{19z0sq(GVyI&a*%Am
zam%~5tRl9b!KtZRkyP$)X_%2sDN}+q&ePu_a`rX1>>-scnS$0}gtYUD=K0YI?lKeu
z>XeJh1(TflEzd6&T29JVa=HPJ9Ud|`tPhn>VnD0u^F_G4cPD2=2dU)EauAsf^@=r>
zE<Och@r#YrKf)M^9hDaA&MjLmv{m#i;v#YN8lVb(8xKv={w8~}e)_98V3&JVB4$o`
z&XQ7nr*8fYI-qtk39Bk2sueghXHV4+;|@}0>XIf)@Vj5eXH@3ZfmvNhVQnKGb*Z%}
zZvapH^DSJ~FU&*O?|>`Wve6V0c2;XEW7(yUKw&0I)<`(tm<J!gPD!h+q)^wP$z<<L
zs~~&0>br^)fU5CKypiIo^`V+Y-?#!dfBSYZm7?dZ1gzyp9u+$BEU<+a++JvvH<E6)
zau<dzgmZg!NCr{8>%AjySTf!<*g{Xy4i`_`0%awgabE@4w{G2(XEcc}p-nX00d73m
zFVr|jQ%Q-Pc6sEg7MOxT27FR(UhkB>O0q;F2@@;yvB}UeWttp#A8NRJ+Y<Z?T7C)p
zE;{soA}D|2bW4Ip#qt8Fgia8Av^qB+i-lBrQ?V1|#rj2#Kx8A6ArPF%JD0B`Lv6h0
z+KVyQaz6i-y}`!h5citBX;wE2ccinx3lNobwbH&<%$A=zo6D9TdX6y3Zzf(C-28<a
z7O5K|+w^fV<%MnwVFAT|(f$h5_(LmCv5ZwPjw^B=vPJm}*-$_{80QY58+%$Czt^9h
zY#r~njrkT%M;Vi(Ctzp<gIeKSHI=7!tp@N%x{Cr2cNqiNM{WvO1gA{aUs$oz`UhcE
zAc-c20@_^EeD7JLkucdT4`Y8y*$l-8sGxgOD!@Y)FP{8%Q>j|3Z=v_2RKCvB_(~Ny
z<I#(y!uqXMxTQ;{0(<aHnbW>&)Oz9l=fiweyC#apm2p!g*(J?a+fx}|Njaovf7@qU
zF&+l@7uW&WQPV$8gnz1!<@hh%P8aG@|JuKI{(JvU6@o$n+ygYijGWy>Ye(fq+3LP3
z^z?-%ITC9z5>;~)liiJrJ>^0iYVDOz)OM9|zKN^7iPxD9#6^`Zui|{Q#W}zy{#i!M
zM?9xi@Wc0X)7!zDSE3}O3cXWK5~!2f+sT71SWGTbm0U83lUrR$wol!2+`Rj~t$wO)
zPOAX<-$Q}_Y8XA>ZKvvK3(w%s$Jj{oB+R_}EcTXxJiT+um2a61ANMT{B7`^nu(JUg
z9Fy<RCsWWAS!XhZ(9O0>Hu9i`-wJ31O1^`9JYV)v8x+bj^L}5%g)Dumw+cB|gAoV_
zO?|F(isQ^3R&fmlzMTJO8!zc7>kaPywFK#D78q*(6|2@OV!6S4D(26;yUyGhXyGOL
zSC3p>>ouyrZVbcfF5x$5*`EHL$3+i!Pw(fa91}5r`hBiFc^9FLEk4{{gyL}gGp>q2
z)#Z=(qPNFuF)5Z8Irpd?L`#B@@Dht0iC<1Yv63d_DwPi|fDjAzMs)88LH$~!%-HQ=
zu>?N8F6W83Dr*rZ8Ub>r%OI==jpkB%3NFE4L=9ZJy)am3@h;vn+;~rS#&YCc!925m
z?*<U*^2Sb2<DIT4cR8^PXpCA;#Ecfssvy4*W}MycjATW}ImYFnDWu<pdB>{GX5xzj
zf6=#|O8{?t;bx5|i)D^h9cBydi;h@^($3lAG{iRCY1&(}-Jk=yCA%O*^p2!L7DB)|
z5Xk9?;2syJ2RU%k1Lz>Y-HuYYKVBQtp#`AMHtqum-X~uqi)k2^mg}$Pb#7Mf2&iIZ
zw31Q+<00_d3lj;j(|jiKvSxvM7$9pUh5is@e>a2F8k3THw&Qwg>l~!jxVoEtT9Ax6
z#O@8M%cCfIv5Y{yT04o)<`oT%lQdRp0O%uFL+eUCF!T9C-|S_mm~+qGQkaY~5W?k1
z+|XmsbG>bESfLw#s)-ZBa-cM($F~ly0u6P-2Vvi2w9k7IkAK#p!KoP-f-P<K(vk=r
z^va5;S$0lLgYeW~?G*`zA;GKvWyuSyrK!XFK{xFp5dO>yu8qHhAC;6FwhhE4^h?EO
z2~MaC$RC|`j4cO-)>!jFe`Ixvq(;o<fPKvHFU2N}VL2Di%&9~;E8USrUImhP1!a(p
z$>~75Z-c#{eMfPKCP+m)+VLe7!<vi>TTP*?N<OUSAC;}|;`=9+LIM+QIsq4j*mjVE
zoX1}EI-gq=^B;_?BxV%EO<LUOIBVv1myB-dxlETTS!kgz8}bHGVQ4-Ez^#f^!(E5a
zok%FM{71Irv_WXTWOmMbh?kJuWMm$!*A~BC!MOS0T>PGTTg+pQR4C7;>vHVG&Y=Ld
zCoHPEve{;}E@bB+%$ZFE*oL#S=fhT1gX%0G;;eoHqSXk!C95?wEl8F&l%T}4oJ~O`
zK<DH_do#Q9V~*OR6Am^iL%6l`)W66@kBsw-5IPkC(p4>`T)|!_VF=t-aj<%>%9UoJ
z$~*&6*MV4`#P!92o)dRdnbkhsf&q`o$xNQXYTR*|2<=Ug`5lHz`IX&$p;|K(8w#xq
zgeu!$CJUJ$7<2*&_>(<)BG_fP>Sjm?W8c#lhFPxypJj*Pdk9aacL||ck+cFuHEE}F
zi*to0Vhr>(#4eyrNvKpUV#UrjFNTw}&oa8Uc0_!w&o{odjWf2iz|Z2w6}5)j%r^Ug
z-_u3qbhXwF%@$H3#7VTq2`$w|M*Ox82VlCj)x<=3Y6y`u(gP|U{s1gZ)z^8G<&O_5
zE4HjdsiVN6jKg`fy9(*Hwde9uj>{%{!mb1PZM3bNhX6Bc34oc24r_8&+T-<-f$waN
z=ir9OuG<s5Mz{;kwdYAKDK1PTQN3Hml0C;|%~_?Je5{wJsRz1~^{A-3%96}$FR`v7
zsP+^x_>@JvUAxl)RR^-q@~KOnYVK_BkE(KKI#`hU#BgVb`JaeV&sCu0FGFw{{v!%2
z`v;Kv6$f4zxoGd07Ib9K@@O2&Z(=S`I(KuqC*QhyE0PAr@SX9RWpnQ~wXX4!!v%1k
zH_wKZ=EDSn+VHr{HoeJ=ARw&AGnt`8`0Mo7I9eNdZFc{sAgSK^`F9ySb&Y{Fl2pzT
zJwOJqf+jTzK?QJak3ujxS%JvrxrjBIu}F_l>uXkTBHOh6=BS^FLVA6n2&?hpY<!tR
z&*l{}^&t)tDzMyoPSXdJ(~qk#!PsgF3~ze;`spSl^z7Rg{mHX+e#~!*rjW`7D~vgc
z1x`2-Nd-<Qx$?Xv73ZM3Zj&YK`L-d~@i6SuE|)HnsyGR+K<hnL=-;`5q+!`{Y5kj^
zebWgUE+bLrEB9ydg_Y@4WP~IYClr^2UwTK+adD1A(Tu$PXhL3Ikb2Dxsz>4PJJH`4
zR@bv^5d1RNs{~0zRti|DQvUd~le)Hm+zsn=DYMqVP#htZTQ((InQK<cO4Cev+m2@@
zF-sMwAn&#2sjn3V{iqH9Mx^R0QOlaYO^S`byT|e^)gKi?_c%|X8#ObA8=OT&zRz{)
z)$RW{#aY?^<6Otc&hlS|tg%L{&YN#8tB&#L@xlgB=`pLhE`jaFx|{HebpNLVp3+L8
z&hn`bAEJdOgnDsOo!sm7I|%IM>h|XHYAfQ(I8I5wdw=r=bqrS2^?w+9#~9IqcHOpZ
z+qUgqZQHhOt+so$ZQHhO+qP|--Z$su?A-g^WGC}a{hq1hskiDGqxURNl^nR2rT^?8
zW`R$S13kT(y`AJF%o+22XTFAuk60ZGkW0tjhYd-$xo$sxBsaVvXrf_(z7Un>7xiin
ze<O%40KoHead0EYlc#%&GMea0ZbL$B4V8Tl6XH{JxALapDCQP^od{v?qfIM)raIb+
z%*!JfYf9!GRMtsG!M({G&1nHfA)N7;@`$isLo_XMn@E+CGkZfCrD#$3)bJn<KVVcq
z&=2EW#&Up^15P93k4ri8pi*oYNUYaUC#EfEk;=ePG(!gQ9xGJJ3e0_*rUwZ<n9_^B
z*HqHmu%|Jb<9Ack-yyEn{e~Gl%`;3AeV&~lP+VQy8Eju7O_k{n{tnK&6yT80J%_EI
zA*tDC*snAyCmf@J<0Pz=nP9OISeq*%5f3>WzrEzRrty|qadN^Snlh*1MhxfK&W&w_
zY(10F5+FA>pPUkkMUTrA+MKe3i&ewVK<xrg_$pDNn@<nx1;xd<f;WU2W=IQYhg$_!
zB_r24IXC*b{mR{>uK@akJ7W&2G+R)sPar0Z*w1is$iHd|AVvZCOSSu(+Eu+8+R7<3
z3%>dKw94veEa;6#Iv@BlDCu)45T$>}Q9>1mn#<NDNm{>AYc}h%nBg0dG$9VfJgrM~
zkUf{w!J*u34PdU}-c;-zQ3+y%xx&FOHw|J$QFq{ODBe(O0VPm@!CN`s8f%gn3<mbs
z4~)SLCqUG04lypDF1WVy$Jo@ANamg<xHlI@qpx3ykNg!;d|$Rh!C0QRe2{0ww7Z<=
z(Q{;yN3^AZJ893@Hu;m%ES@d`O3qg?lH<?UQ;PQ-$3)NqZbpATna+N8Ep;@0(*q3s
zl&qRqqh~-eL0atUUNa*{7=umvlPUVn;(C&EAagv1jY($Syy`-YtF~~l%nlAi^(lHv
zVN~<-nO~mF;)TEVbl8rM#QWLJFPHV3yR<f6Itbh2{RLm{I~ZiB(SdwD>GUY4C(Fn8
zT@1ca)(Pla+p0+OlT~RvJ5dDe@S;cO{sc!a5}R0KMa!x(o`UtfaI_&%ncWe7-r8sV
z8|02v&ZnxLgqHG|f!FvoqvzNoNuUPg3kJuY&@SHk_aJL#5r|WI4;PVWY47=>RKOzS
z%sWVI$@Fyv6KBj6#<Eo)rF{HC6x}yo`q#2SaITrAJa^BLFG<`7S{U9!u!vnLecI#p
zFQVlG2}bRK5-|UWJbw?=@hgIx*V6Bfa3qBDG(YE6_yOf<Ktt>{bYgLWq}@LsJrsdg
z=n|t9L!?4;lbTZ0<p!9_Zc&pG%TCr_k1P!=6=h+}jM?FM**2cJrMTw5wnYA>aag<6
z4D(zxF)<g57?O7vJwgvACSj1UYzyg`p<_9do$28%Au!AhWM?L#FJoC*2s?>SB1QPD
z+G>3>q1(N*>W!_LeznMcW!ivpBi?Nslcbg~b!~|xyQV0RsQ&Bzk~~Kvf#3`uykBbe
zuZV{^-Pb1rnxIvCysu-zYBh1?wze~*;)vBzrhnDn^D4P514EI97`!-mJe>*L1%SDt
z@?Tzk2s^bE^D#UE8Z>xdfFe&mC06yrd3bhrb_ii#)rEx)n~_v<T?&R+_u(MNc2t8J
z^1;zu6Uwvww$jgI^R6}el5e{ARdJW8Oj_4qtPy>ih)Hr|j%3CA8n;EY@?kwGj7Rk>
zwMTX8aa(`oVVgs+MmF^%%?#`TbjiUbuTC{<UgbjMBNpLec_OgrIqM3zQ2VP+8jI%m
zJcZsV$sD<I#Ii3ldogYL#hT&JCUil;8OwhA3QPe7eaDf81Y%0HV&+X1{#&B728OBz
zGYP&G6zl#g1}nnm+;r5JWzJZIQ-He=YLG(X)<2&i@MRv7V7>KmZc99GClyC23#7$3
z561Tv425@)%IJ`XV>DI!B&9prlF6aO)}m1+@Y1$k42+*MATbzKyYyR!AR<;~eeQR?
z8m^6Sdo$GYxemIP_{%I8l0*{6st+!tHiU!S*4V`#I+WyDj^*XQ>}OsgEGLTBT2wsE
z+b$*7LF-Hfdw?z_aQA{+Wf^zD?Z{N~{JO$^?><x{0=;;%Zs>-Jwp#^S^@C4D(xwK`
zWhe&M1e+MefMDY_ANvtUF}D3vTpN$t!dhkvUg;ly)eKu&Z=qe9;pPL+t+%KM1yLnh
zYQ#YAhd}TIn6BA66wD~yJi8MkEYOk92YQ8o9Y&vP$y;y8@8kRZ0sC1ZGKIUCGC`p7
z#y3Qz-t=%=wi*d8ReGMYiXa5`ki{AKIh{{2A3A1a1~uyCeGWQ>v63PXj%y+2qJmwE
z-mQZy?q6+mFQT7ctL2%FSi2WS4~jkKjOZ|S(Kyz*q-h;}leJZ+7{->Wc*-LaS#3J)
zlEuW?-|$X&DT0be1t>Y~_vfn0ip;Tfc)FJe)z0nw83C)g?MM?;&2x<1WGC$)xK~c_
zDDx{T`#9BKAvNICp~#PF_hmVS<w*dBpN|j#Yvth(DaemvSzbgrh5c43Lkrn;I7$BK
za0&?k76oM2-auEhkbRN&pRdkm+0w>R)uu-{e>Y{ybQJ$X|6z-Kzy(Mdp_^-7RZJsh
zKcgmj%JbHPz`+({T>6yGpt=XpHMCC2Z~7WA2p2S5I#Ag6lk#27<IQlCFPd7qUUxXv
z&fA(E1nB-}>4rDmoH(wp<tBrp+Ft!O{aWqJxP`9<Wg4R$ndnl+PvQgi{g5Z;Y324b
zat&flTa^}lcIa$#0A?HxTRG_<<W=K@)2dPwSh2H>CnDjxF@zi`)m5M39;Rw-|90bY
zs6GSd>Cvghpb~za6E(kvYZ7LmI;h`Noc=avw4`o|x<sXg8u7@r?8FT^G|G6L_?SIs
z7_ah9FBVjjkvWg8Vvt0$k|Vd&KjK4_Lknw!eGx7PhAU3)!TxI6(8QlpRz8miMgk+<
zB}<WSYbDKk)|yxBM>&|S^x|22pbH9dTzRer*gb~(>QEEJlYJB@JfzJB?~n(x2@Zi<
zI2714FD}VF6!=youcXScx1v44xoq<)5)Jge_GNsR9(GDJEbm;ZgS#ni1)HD^Q@Z7_
z$}X%4RX(r|v^0o0Y+kurG|_5|cSqN2-W=#oZ-Z=doet?}s!woIVvO@ai#KIH>!;jw
z<r-JM#a0t8R>PXpZuIeXhM_)*f{4SQoY66aA$|xCH~y^V{YL;%`h;l#S!-{faxOo)
zEF|RXNWyhttnb9Z<9pjRd6>4jyk+SjnXQs}O`HqHhOfgI0y#P{hS_*|o#{H{z(4-;
zM9{%$RA5m+qMnnue1;Uvy4bNK2uEkGSHP9Ap)9g!JjE7u6{5cR3u@=mHkjTS>Qt@G
z$P=l%g-iv0FQc>UXB(*PERUX?ozQHElzugNqxrI7%6UktqrdYeS?rz#UH@Wu>s3`a
zev=gc@EiRMY3r3oG=U57xwRzGwiZ{0+r-6s#CF?)C+Hqm%-S?mJ>AXn!3Gd3S;q1K
zTjP4=*a=dlZnDR_Dh7K^){F|rSmTGk8=J%_^qQ}B0yaG1a2Hy~mS2Cm^=E|c<!VNo
z@)zoV(kFNBY&9l7nzMKc7CuwR=*x-Uym96F?Iil)XX*y+p0~0k`zlzmp_p8;xx&3g
zn+eX^&TY)Mb7BkVoj0`==iM-|75x((BjK998PfUzi+hbV8Mwajgyi<p{GsP)`2$RD
z6;S^lBlj;N{{PL5XJ-AMx~*^~qH)@;u4_{9YgHGudiDL%n$7$-kzRkTD#lLhLeD?&
zc_fngcq%$JG3aA?Tf8brB5^1?^<M8U3106{3mYwRJ)Kq$ha$DROA9SK*#LUtzqcuW
zoS(;lAxienHg4p6eUlYAKRfqEe(i<74;tWJZS?+FL2v-!c<&xHu-|TulxyJ@*rXL1
zX36aEQ-k1~1*Sl5!}5`N>we!|(dUrlZQm+us-x#*vIEtXxfA^v-fa;mMRRC8+vq-A
zqBShelC*=K)Y}HOw%9EMxH4UZs%t9un&cMX$Tb8&%+DfB)#?Z8@qV+>?l00`-|ix#
zap`|S{rP+0Y0+~<`8*D`r>~3UKp*x2bT|#BTq`Kr!|x!hD3c=$CspVg)fv4FFm-F9
zyk0I~;AggeUT<P7z2!y1bMe`!=$YY)6j~Ru+&af~_QH)g*%XJVs5Pak>FkC1Hd2Sn
ztfs+akZ{@-R-z;bwse}JzYTRI&4Dgh5Znw6ay1P)Bbk?Q8r^VDXBUCtt=h-C(pA#q
zb(HO`Ab;WjXhV#90Q9w_uO=?WQ`mBF3C&aJQR77~!U=)$ce4rwj88!8>k+&>;^}P1
zhqthWO_64CTHXD<)H7ZN3FZCOMUDQp(>n5^g?JbXT(CkV7sbq2l(^fk<`QGA(`=S-
z?gc0c2+VTP{LH<t?9@1ZQGxHDKLX3Sc7y&aF3n(pcQCm*LeS$xxH(`;;({c_5tvoL
zxV%8IYLViRxQf)PcJ~2=P1t&&Ib7Lp92C`L#g-B_pam>+^x)RQlOTfrRekmqQtXRE
zcQ+7iBr9{GfQ~?sil=0K2?2JNcrj#qKc=!O-cV7(2l#iJ<lvbtD}@2i<?W;?14C2y
zPT8`icmcX(#yMuAPPvK>FdwgRNVxOqIPJ`tLYW*+X}xL@ufXd%h0^5K+I5se|3#8i
zi7a4NyHrUn#mGdvvudGuWXswDCBY7JR&$AB-k~Q)vR3B5lEpC)vEfGv-?EC2?m`>a
zF!o9d8*P5eq|UQKgWH$o)?~(iFYi!oT#xlL%wwzwBI4_|h3XHA6?`Ah_iKK}&3%z2
zY>h9~Yc-DFtw;7E$x)Fq+yOpHlMXmDFUc)T5U^j7BFCc1sQ>npRjC!mqv4lXyS^tn
zV6|k#EAlg;fSb?WgfH0lgGp9yzN)j$XFHM5oC948IKU=_m0+LSCe&BRRiL1vB2d{9
z5_H|w3>df&lg=|~iVEEze#B_p7M{tV_`4@PiM=-)m#erYHz4FaV)7(X^G3Detnx=B
z^Qij-^J4dy*0Q^wOVT9s+8=;j*4Cy4gb!qOiB+?7-8zmW2Zm+Q_Ra0onqW2BM3xz@
zn<zFn$@XMXuk_N8xDPlhTOUjOD-$>m4RAd$N8+Vs8RIv1?z|FJC{!ls@ScR_%W<Z&
zEfl<?IzuB1h^(}-DP$4x^5;FTJx^Y^WfxYFI~0o|_0?t99HPJ^=DO1+lbRl<2dH?;
zvxPB#B|Df#A(RsA22RviaEH3>(g?E6PU}zlX(s)<p7`$>lpO|W`uN8IjWnP<zCF-D
zdgJ74`;pN%1f+=PQf6Z812+4y<h?gFo+~6(YMTjzaHJRX*Pp0GPPXetP*euRmfMz!
z{E})7y)w?FcBb_>kbbVC?E+mih<|!T@8X3f;c>OODGCJz6$#Vv;afd<*27pzsNN$h
zPUs#mvMF3)x|MpHo=V;Ii<d_v-S5Kf@G19j4|V%nkn$q!>&7<H9tXI)I(nUsbay?L
zt3PXd?bY$Zw&Pu9F|K68;M3OWT1%{xH3tz)#aEhJ)T0-|kiUr0H#-$4HYqqikcL_y
zCs&FwG}&v&$@Ly4@r;&V3QVjTOpCd`hn3;ioLU24RBnwQITe{P3OsRJv(~6WP~Jo*
zs&r;itPXLT3k%Yc<b38CqFD=u6NXuiyQxZVLpL*hOb@)&1dgN6<q7)hf8Zk&AM&u^
zex@aPQIenieCHY8ejN$0l)+|@clfGkGTmB+rPL8hCb?ArR;EGeOU@yXDX!0h*@<;x
zW0X|7S>(k?aDV%b82R|E=#^J(Avp3I%ar8HLXOdc4llAaB+aXiHL@}Tv#DA$jb#(=
zT09c79Ya85euoL4*4Z^yz!4|bM)uTLQ*)dC%}j)0xx+vkj_(NB|CrH^{#~SjK0or*
z$Q7urjzybXR#hf;88m`#)t8+pcp{RtXl_V<Wb)0{6l4fHO<UDs=_n}AFnjy(T)13A
zP&@loWvGF!JrUg9afm_n?^a;ma9`}(7HAaaNl+_9wz(B{ITjDRcUfx}3=zd`|Jx0P
zQi?Y6*1k;Sp}ilgN;>>ApiRxhsua6od9-$}3+|COEbQKH8G3YC>h>CAz=-;f`>;E@
ztnICJ=&zHYq-|n7wo9c6&jU!42Ql|f7(*4_#^4DRx$?sW&ZgrICSsn{F6#`y6K;(R
zi|=gTgW}L><F&_dvXNlJoVo4^{@lQ;!C=+b(9GDZA%KcecMK}RU#GIDTrt^`cWQCM
zth6%^G5FKnU?tbRetrLpCa$rjxzdX_emL~h74sGr9Ny<D91YTZB`eaJqir`Yit6bi
z7@Jj7^RHw!SjUOs5YQ5P`8fGWO6xx))`ifYLai+n=MX7TbZNDnZe$au3+<-{P3yT%
z)H*d0fED5)#%oVxjtxpGhKJF#637w<b}l?_^2lybm*v0^#YVPuRz~U~p~F>6r4JOF
z&^gl^ty&h}G3DurL#^_CQkVdZ*AKR7uAEbys$;!&vDf6oDL)6n@>lu2uYdD}nvWTu
z)=FRFJ6Ow&lS^!JoPwj5?yCHW!@0^cMdhZ<k5>K!b9VCDu54s`0?b9E4BU5~N?vEi
zMdQnkMLt$*XP9Z5^W-bUq|&GN!I16^_hV^#CWPhJ6~@=>iqmbTBs-jh#G8_$ipA2W
z25-Uf$ngR$;MFMU>GC+AI!6K-o}4D>9A$c7I35cjPYp@G57E5m#WqL2Z|8B(gko$y
zb-v}`@SdAHp8IG%#hoHv#fif59(ezSiI~@fCpZ$e{X}TSS;-V$_StmJyA!H{8uL&&
zt{cs(&^tHC=jpWZmaj5sbra2sSO7N3C(}2htzq2`tY8ER^e9ttKJP7mh}oZ>TBR)J
zo|XYIAZPQ=G-vHOGz&{^5e#<4!?Mwyc~gNd>!bZ{?6^$pu$s$KgsLK$boYi~)bne)
z@gZPDu2(3>JR8VlK?<~PJ)%{N20uNj>9FZZJ{i)AZN3mUyJcTaR4-d64HJ>iL~CC;
z@;lMHNt;1~mhgAxhnIu0Q96=_N;ysbNN={AZjYxn9EK{`-({snGkQ7vbX*9wJv27~
zZQWiWTEp8O)&(AVw7Ospr&)#@a<TO;<U}}g823b-d3X;MbTK6B)ZM|3cxnz9D#g)R
zw|swhdB3=5k^9kbuaI{bD$V|W`zU$r_d}gNFd`*NE&1g*JZ_ZqOw0dEesJpsNWvt`
zYH&49orvOtifVIokRFO=)GAp}*c8NOy)XMo7&3S7*E}HL0JPNYMA>|v_b1NMoL$O5
zkW_K*c@8T8;T`BI+^l(J*dM3lfxZKrOajk43~1A|w*XuddL5jv@WRh=4smEtY_qE4
z$IU)mz&(;xZxmLse5?<BDpuFCA<uca)N5ZTLecI%0{iZ#Wye+^e#3kgYS~n@BcDW|
z+i|>T5qNIdJl<Tj8U2EFyjO%7&<Y%9ax=BON^|mO&Btqqk%qjF)R`~T8^$w<o^kkd
za|^T`R}<1~YVaA@@~ViXkq06t?~|XV6xOWDr+<{#A82+Es>xXlJSIYF_w*?#iNZ7q
z@g@8KCh!m8F~@kDKtE_r>i`_dCEh1dMM|a#U?Ja3wJNpG;WNi>!CIC`d$@hQ*P*{m
z-~-PHKouHbXyaII`wDf=af;|#>f`CPR?0K*o|@1@&Cll?ekxSAxeYy$kW?PkZW$xD
z%V$$Y4wlH>H!9lbOy~|;B|K8iJo|o2Dl@h-+{C1PS?$cdX9;(VL~zr8P5hIP2`xz3
z(trQ2{+JVGP$sUWGB}Bmr9Gl<6dp3b%{JCh!?Lp_Y#Vg%vgdAqGxgs2zFi{?ObzyZ
zI5IaCnHo`37FBo`v}x13TWV2}05H(Q^per$ye1U?1%rSAI}6{jIS8Vf_RWz9{IlZ=
z$7rEL3NP*&2u{*<PM*Y!fj;IJ*_DrTcuQc8s9u{GP@5y}aT<I!-R<cgadx3S9V3l4
z$q8?<w!`Yb28MS$BZ|_P9yRM3AHl|lQU7p(SC$WW9iF1KwQIL2y)0U|TSL_J-*o-1
z`|es$ty&CaJammTlCFk{NP4yex)9a%y)ZA9OL6PLN@chFrA4Xv(<6vDVWyV^KX`Ve
zt-HiutX9u{upHd`#cGNBW_Wed1YuJFc;uT#TUz>L0akw!9JWOw*cj<f^{yRm(f?95
zUk|59h~aLYH%hScdcfpH8VU5Get}{50KPihlcFBR6^SP}><=nlpR^<ODU$cTSX>-%
z!voUBWW}{Dv*gBBf!RzaWf26vq85W+8I7Bc{!^HM%_8=F>`P8(M0D7JY^N}$pleq#
z#Y3$(xF;55ll2lG_0gqN5sB1U@I7WEa+5h#yi?;&C9eTu4GRCf#$i_(16aeip>FOE
z!OWB32&EzP_WgQ3>+x67zx$7;Lr%v3&*_kf>3>-N==v><9{xWNnRoC>O0)=UeU{lO
z2l!Ympq+rs0Fn}%_#eK)*6>2r-|#qEPg1Y`WJ!)q=c}`Oe|7f#{CJdiY8bpQzkhT!
z6LARN>q~koRY0#QX^`02`Q`og1PTDF=hh9uPuWfLbMsnVGb71`M<~;M__P-Pi^$qn
z9;V1L&7*I4&^boMiD{e`%w@kDLMiP2etup+HeR0Sy0_{isG%E<*RN*kS%B@(aA`~c
zdgQ<gTOq<X^YY+m3y*N>jMkm>x*b9?PxiC*GSd!o2=L4uojKAH4*5B88&tH&Qak#g
z{QPMJ8zrK|gMvx(had`~<}m;WW=gmiH%bidQ@Lw|l&t+MWNaAJ4@KQo=rKMXNGYuC
zb;ZPgg(4=pNGuC^X^yVG7$na)EJH{#>z-I~c|tM4Lg>}cQ8s0N&t$4%Fy(w$NTQUZ
zMgpbMW)M#U?4ZsTWCK1JUQ_V*Cr_f6t`sY}XYa3U%oR<|3)?9kwr2=e6Ca{vxnR#l
zEiuk&c=gPW7a<&@Nv&$eK<oNbkXT{oP`mMIIvW`@I?(Sp3R&g_Z8xtcRiW;3%&%8R
zh0a9Lt?rt2upmzJ4b}lARKRgcw|XXRi!SU;B3=dMJgf`g?!h;mo#d`xA*wDB(bQs~
zglQ;Q%_4*@1`?6^{MMA2p-*^~ERfxFv29;N%0?Dy(w@se(s1~3L$1ziU=TbGaaI~e
zm|^aF$~*752TNx@`t=%)qbr^rRkL*(CU_DmV9=6(ivx_u3SGz#4%Fo24k+NoIb{5E
zw3KY;`__C#DhX+EGzxEfm?p1$I%!2Avi~CH<5pq1BH3-TuSjDrUazNgZ0)U+;<bJ<
z5TM}@{JUWj1@!dr)~hVFto&gBR?MNaO%I<LN`hiTf)ZU*zu;V|Qy?}ZYN<*7Qose6
zhs?n1_X#X71fa5xzeH1%hhgaQDW(>c?0?d2VKe76L+8c285vKAC}ZVtHdGx}-)hTH
z&cABH(j7R+sJOA=^aUjvkt?p-lK~p==sVUe(R169DL_uG@pF^%w^#q|ovfpa^W?QZ
z#1q6uL9@if5;TvRc(L}N;x?xxGovQiISIpZt!U7*lG%M;D0l?g9F@1vtW*b;xojmD
zFv`;0eIgshu-yqMS!pTp%|=_Tqm9p-xbfu+h9_xaSk~k4bDvK9`F41FI6H|upWWr(
zs1T9yXViR2XA*QOoRx&qGWt>7!fimDtS@5J_!BVkNuDnAa-1UCmTYC_+Tf$`#;T2N
z5u}6+cd!(I*z7MBO;F>sDB<E{!74dVi<URlu_nd_c<{ye(Wl_f#y_d}&-{g~JvNke
zhya}eqQFr3pM6oBB;NMOwTn!_308zn2YIQU5%9Mz1d2TxA*<l+>SAuZq&jaof>sfT
zNv(IVt(mQxA(q9wq&Rq4y+mf{oz=HX;%I$>12^5I4{Gmsm4UR@UmEeTnLbp#bi7(h
zS!UeHv8#(*Nu^UVr7aD&4G6XiJi9i`#jZ|py1KOe))1fzFwF_!kIK=Sho1dbg9$OQ
z!`i|=E;g{wvit(AszJMA3+cR}j*FL43!!BNqMOxn8~U{Q4s-KokO)^$4dw@R54-2z
za%sIfNr|+O6A7D8nUXb<bh3<Gn~4LYh^>n5{IRA8t~K_8R9#YBjJgnIa%q{P1pC-S
zQ)Ky;c-~%mBli<Z{_^`$Xe5$CfQSZ!>(HNqEx9w!Y8<t5(T!pP7TzrHDPnZaFz83V
zW5U9;aBcaBBD~iW%z|dsFr*>#6(SuLDN}y8bm;a?e;OLf#@1@P;yjD2tyzy|+MC7e
zwI!hXry=FZ^XkY@tcd!qWe3@EZdOGWD<8u*%t!sGzhZBg9c1K<F$>9B>w`x~(<qDw
zDlaT0+W9YN-5oLN6eAyQw9awsA%v*yQBt92mMsrlNDOug*JjZ6{ed`#Ru`sZ1!pFR
zqkoiHoK}&+e3j=aEC1EAF3+-=<AZxHFZJhsi`5lCtl~~POsp%IuHIHvd8fGai*vJO
zo=xSsCB}8I#??>I=&CZ91lD&`B2qfs9D9j07Y$9>d8D^oVA*0(+8(%ZojiFsU#2a}
z<HDLRr>3^NGrwTS5kjefFDD#K3Z9(?0dJW#W>uv=tzR|}HO3QUe(_<0oO(RnV6%Jj
z$q3j2a9hI+6WVt=@T(%U`OnQj`HfP_eboYltwS?|_Z9%G6eMdC+sXRU<P+|c{WD)N
zJ3bu@A;RLo%t7j6YffrA>Y?*V{(GltCM)B%p*r?WC9~|oHxX5$V37oDZgd3SewK6m
zbbu)NOF@dg8zmWAkSvwjL3>HQqk(tG^03WPJ7~Tku{4oxkUmj1+0<uw+jKL4QiR$X
zZ1e1qs`<KnD1$3WKq~XV&v(Xnxo>tRYAVi7&s64SZ6g|z_Lw$!>z_|>*`gCB!9VNd
zjqY1bqWqdaY~i4CP*!uDDpRblzi~;==Yaxh>-}EFeaGcw1_Vbg5v7{TH~RhEK#*~U
zGR66VRezd2wK?7S&vh@*d3YXHGM#2;M=d1NKRQ4hUflmJCNlG&NX@M4HQd_EmOR}X
zCpnFMDJ3{_3_|lkzPZOIi$aku#mLT={~Kqx&`B%aSPBhYIK|)st47~%Kx!36()X4B
zP7bd!kj)K!`{2pFKjJQQlLIm{gFMrT<8jgzzqkGn;Ao~J8R<{0-GWLVuBx`>Wd_$8
z2XFWylj;z(79@H)SDKZ~RlBQpdRADq8Sg?~UkB80y}GdkY>APCPcXzNo|=_(tnwh%
z*57OvZT~VDT+Gr=Zo5)Naj}@L0+S2f>nwURkRSZ>OKA=;q8lTyo@P6#Y*4Xlk<su~
zaNV&TR>ZzxW1BVvl5#d~h^oRaTUa@W@&VQUar!(xvhf*4c)OPJzl!p)#qBz`7;<g?
zs+0UO7%w>{5UQ)<%n2Cg2H*GrKu#_j{7)%}k>S5JWd9F7n3eH==2pGZ8jB`oN9tNp
zI|WalB5ef>@Kcnj@NbhTunmzLUM0sL{_#yEz}$3FN-Dn+!mU>sCm{}~8&P88`*PSX
zxsh90DXC!xC7;MdQ_a@(<%9I<z0>QvE1n~#24R-W)cxW4ay-_?>a2DLJdxZpxw8w+
zdA}Ed45bdrlhW>1M>uI$?LK6cEYX=i*?Oc53miyV*1$XM+6DDLK3ZAv<#Ry#x|sS$
z+ZF{HGH=or#|vJVub|8SmNeL-@=g~Qi1r9mG)n>d+K4CJPJY{#?|Jt&L5f352dQUg
zp}}GLH!^1s$yVW(RYb}r^~4e3_02S(-Wp*`Hwf%gsD~sifky&+G9x$k_RjnD`ReJ+
zS5Aahk<sRJjAEd&sbuHd3ksg!+_aKUVyx3|LGOq6>-#Y$j>^L8GjbX^syAw`ojHI-
z+RBY77h^>}a+dP8^!YZ1dJBG7AkaJ|(U&+wVm*K@z<PlC3|v0{en!tXGlZM38y!^w
z0j3Z)80`m2?Zh)PkBsxcKV<Ru67WRqR&FLmnYfR+W^%OXeYe*_@0*oaJTSUwY4G39
zRGxfex_?|kNN$0;ny2Ydq1<=<VH;_=TD^j+o^7gAcP1#EULizxF1rqOW&BAx;C_<Y
z`K~<Y`-Eg+iBR<Ho9r|WFl$1!_*t6u=ookcZ^oBqvR%mOLvp^ZEdJZsGiFPCBvZNp
z66VuLj+?7SaS)c91+yiuX#RyW@>iR9H9eE4HRLW%*Pt(FN}%(k2J_vDR3h6$?>c0k
z!8i0+%L#&IievOX(r0rbp55O!e?`-q6+8yFMDcKC$ux@XAQjcjiX8BFfB2rGB!`@-
zrdCA<%H)SH1oR?9GYKahYdPU2^Y~>$#r^y#`%;NV2v!kf7DU9dckO=<e`_BXO@1Og
zL9#ZIUDV=dS{ebNc9tLlReV*I{`j4NoXf)j&08%NmZe8x4$_{QPTVA-PkLfJeU6^a
zW>QYh0Eb>O7N$+!1hb%=)oiq>gNsYG3`w(jAe%BQYD@`4wH#h<i6ii?hineipuo>a
z#J2`KW~mN7MSlvY%_ylh-6`~#z;?5!(!{B%9>3N>v%%^FI58=!cV_l8;Zol?zl{7X
zsc1D+j0{(xK2B-dZ5<LdY63AI0coZNPWckVg5o8`HJdKu2(X%0R6Tqs%-7ummt?y-
zVNGqsEjva+ak<Tufxl&NLJ?b9y>_p&ilIcg%f7ZjNywHcKpdy1Xs3k7Wz0!L3uY>i
z&+GPwhYE9&zmv)T$&BI;fe{4}+QxTm1!5dz+tR3c&`Z3xVJUig%l+W=cWv}Yz2?^f
zUrz`^^@=Tjv}SrDw-CW#K%+Rx-b66j2T9PC-0o1AOfa5EgE@&rP~=RcIipk~B6fSw
zQeQeQZZy-2V5KO^vP@k&TVzV<SIY8K10`Q%wF+pDKDE#WmvdSw$WKwhDju60gb%}4
zSY)j{J1o}Z`u@5y@P`3PRdsA!Shp|Nys6~`#SNFOh5E2!_Xw))Lh~S5BgyRUg)9DR
z$Q>-NsPo8kmS*3i#IbA!abzEj!`@?)ydx^uY9g8*w-Be<up9u@N772fTz26vq7E=Z
z=w|r1N<`CX8b2^vG@;(qfWC!nFuep|gl0houE6zWr9^U&MK@D|TcMGkB_8joykpZE
zr(N{AN9A!{FUo{~qzX5c+vq@+Ps33uLQMOIQ$D*dG<CrUzl{0XtlSgET~rjUPx7*3
zNKO<@$SyWQjcyvd%#3vyryLG-1%JO_R#ZM+gVYCU9fH6?+ChwY5f=;fgcRZa4x%vb
zbi{{ev;a{&LL*4-{2c;ho|B9*FD7%_pZ{+l9B4vFGNSmS?(tUPj2mI>>Lderxd?#(
ze%kj!Iq?BZm32M01BYr&5tB3}tHj^cEvqca?#?JgC>p=*sLg#_Jvr}^WC-3uUOQoU
zPEFi}Ns2Y%JH*g!jyK`cn!YtBW(kprs%){PKWxsexmumAcNs@yVg@2k(=J!@+cY9c
zsj-Ml&S0=)q7|-Wap7pu+fDO{lofS3i^y~o<elQS`vG@nLD1un=+YOW+VZA7g(nH0
zxIcuV<<7Wt6b5e~o~^mco~^s$9`NWauF4OA1_J&zcfEP#Hppn&3iH?o0w%l2NH~Oj
zqLla`;A}X5@_Lta1uQE0bAL4CcL^5BQC`^v{YT>+e*S@c55)CuOXEX;z<u8;aOB1l
zMqIS)`3)R{%dTa#2KVPJjmCX&*Y+{AiqrAoVdG$W!?1DKzYT)z_`yoj&y9WJuI{d6
zxRTHCOWG!H3&Mm&KQ*l;yAj{@!-U{}Zl-pZF=(zlW)Th|5sxl1UR5bXNGF+<;}H}5
z9aDwPR};7C5&3swxLq!0nDu2qf25KW{4oA;j5PTVhIM|T6V67rKBXIV3q{tzt0E`J
z7V0GJ<;t^FOxv`qWpQUdqb2Vhn~vfQ^BYbXW<=n)&{ly_W&7te)>yU(V3b>1<4s!*
z!CR7bYU34dn(UjjR#t*}Y|&ig6~i}U8YF$MJx(%5O=`zzJ;C@fJiCRqB*#h9g7_o7
zi)-J2a8X;X<Y9v%i%Hs$ocR)|^B&bJgWf&l1c$Qpsy+%cRRSRaErz^PBuns#ew!$%
zY<^Id`CQFqiL&QMWgXRgd#Q5CR0^5`1Gv`B62HgDFuhT4r&G%hxEKPsFVjY=8B0|>
z-~-IL$nI+IAnUX<YGR5sr=?uCz;V|7EO@|`E<$`%n$^*}#2~;kzZ{Ge_fD4kygAk6
z=#%@GKt2ZR!Fk76g+$j>uy(kcDVcs3bc95RC1yC2<k@8Toz6!NpP;UPPG4Xn+y|7I
zY%10Q=4ZmQ-0H*(-p(`AhFHW3DiXag6<>fF*_kw)f)%|`#0!sVr;!eqWVfv%%hZeY
zW8#ET9;j8=k$DO|_k>k}T+%Nfzcz=){RV2!bO1K2<7WH<a2B$I00zZ<2aLsyr_Fo0
zPcK3a49Bjy*3>Ex%hIE9$@N18yj8zr==$_DB(D%f_KNl9H}Q(QQiQtL?V6i+raIUq
zs3Ssqy}$W`4gh)$7ut{Lkmqu4avoA5><#t#sGLa+P%zY&vOW+4LC0ZOve`v9aIo7q
zB9kbG4>oQ;l9&jbpo3v}wDP8-FUqvIc4|c1Oi{1$7smOXPVC^hX|>|q#L>bQ<IgX|
zE9&ML|D>0d4{6y>1*nC<pB8Dg*|h3%nOjl@57edrG^+<aXhp`o>cV$X`$@?97}PC6
z6m2;DF?~==w15fn`0Uijj_P5doZKV$AqWQ^DTAbHhE@ARKibV7sjt4lUB7fAKkYko
z`dS3~ebC_fhZ%e5`#*(51}3KeJ1YDC_i=Kv{$Gp=>;Ei49%*UWX^kNIywq^;>&=7T
ztgU5D$Rc8l7Rds0nCqADYYD-}OOs3Dj>Vq->NvklW1NLUt%d$23KX!>!2aXpZgBcC
zeNi9u9t||N1zz!|NPw5&_Kz6~EN4QY6kvo^;TAF39TrI9QqQ6oE~rvsYHbk{7}897
zPYD5Sf~MR>I)@I1IQsUQ0%pw-#~OdMt~Q|<BCCfBeh^-lS0=05U0N<$VnfL&Q~|fh
z@_&%Wl8IpyKkGp9x~RJFbFqB32sJ64ElF@G-90722qFjbhm;n=x{&X3wy*{^JM%PB
zk0(R!H=oRtB?B~R^VI-Pz6><;G6jr5rkUJlz!D%-wFX9cZB#%%1lBr}qy%BTf*v~X
z=_-bcCrAGWi!=ka21dbavkV?L_KbD?T%9vd@TUz|cgY%QzPFh4#k&H2ypHxYK!6Q;
zh+pE$QAI#A#fq&|MvfIL_q7qgJwvXONj;m*fCb}UFawXZ$$bD}tg#}x4>R)_3EqQ+
zj*xkXo{%t@g#p=W`P?B0h(yOTu5fdDhjwZc;}HTFXB4?Xt-R7XQf<z1ajESd(0NQ>
zpl(qHrR&rJ8I~}X+(mNOU3F?`u;ZDO@TaNua~%3jcIY+2;$H*;Z<6L<nhQ%d_ywY5
z%0DoO%X_fln?eLGGeH%OAY!w-L1UE*eZII*co(fHVs)}!LR+)3$ZKZExaSht#zW)l
zFMyn{7F+*f19P^06%REC&zw^W6KjS3*;6)^kbQ$hb@_b1y*^DKj8KrxjP(2*<m8!I
z-f~*W!dztccs?ATjwSpIX|@&1J3*_i(oj}9OmIc&>2$WWdfh$0Ix{Z1>tpor>C&O6
zoCu};`F{Jl&p@9zI5NyVizq=qk%I8RI-(Y&D&&DxLxfWQnSF;V5dPzPZ~XavzyCO7
z_~!WeeNl1%AMr9_Mf+P1+W-yvYW0IVnHYMzFJHmBvf|Ox*6Hlv(b3jf$$7o*7^-ME
zpr_yNfk_J&G}n!;?I+EKmD5~t|J?ulG}Mf+P~4lFD0>chSGwxTb1yb7b}qIqUYCA@
z{Y8V!Nl!6ZrZ}+NG*Gif>gj5CYid8k9v0luAG92(y#hGl8jO@oq=coAgqfPi$%b)J
zjuGmJAfF%172FMVB$&{=US59c9iTtlmrOK7JlqYvww&w?_ssZc((UMeS>X8cXnj9k
zlVs_BGJStur?kZA(fZ)_aO-|QuCz?p89Yv`*n7RdO}XD6?;G=o3}?o&vL*+2E+?!^
z=}r*k&?xI=B^z;iM~RBo3|CAnASyU|xZc~gwe~;wx`)K?$K)WM+>2p&4Gnmo;For-
z!LWFrno?*g3ZzFkGNsn1@3~<Pa^%I}MCrk(e-FJ*4Z))ytPGfiv02<^vMtJ@qJT`u
z8;Q||af%>2$OimX97PJyWa^%(FBXb34f<LcKFu=T|5)h7*%rII{YAbEO({{f+lEA@
zBDw=n<hJ+6(0wQYb%*6c_m$-}F?;#Lbr%P*vV%wjvd(&E&E>J3)_PL`KoypbuB7#r
z8VEX%rp>kzA*Zk6G5KpFzw?_&z$Q`pAm@spO7#A*8a@dJ0T^T^&*h^G8_kN;D<e3z
zyO1wRPH&OxeZTx@_!H|)#gERuEcPM_wHm0E(@7TLvA6zCYL+PS#@FzNU$i~SOAsV^
z9NyNm%D!!zO<%6mEGPAd8?KV+pf{BO;N3R@MUBHYmMA3oFcqg+KX^|GhD57Yl_+=J
zhGR_1cvXmC3Dy1`(7KCR5mi&H{PIPpyq{V;$qZ$TS@8(7>^@T2L$sWSSj86G{zT^i
z4i7ZF9|LAL&>t*NDm+a&2zlwLI4}=|zW~5y@bCb2p#3`0Q!xFs`4X~Q^-pQ?&&A!*
z+TB>j#b(9*E9MRG%eH!U!%K*hkw<iMN8aA$)_v1awwqIrU2o?m+{z=UNX;>CDqbxw
z;T_N+H<@#E85i8}?7PfG`jNk_6tVQ3-{w+k)-%8hG?&iqXFO@!Zi@v{q<}qLqWZ`O
zsuBFF1amo1uWLa)_I9tf!QTw`IlnnY_?3?+JiZI!Y0H6x&UsmYMP+u^F8I~z4!qiB
zHtW<N{iiB*(5bprwo1WN7lq9h{VGOyAZVMua84NH2H#2IoI8umcZunj^YZoOj`H)b
z_VS~oYAUCe#5S#QuWI6lw&>{-F0P!5)6KZ1>utS=f>J}^TWNu+k+5JOCUc8_*0`(7
zs<Z{0?oKsyK=QTNwNtD|sQMAA#D9CIl1QeLq+b*d;-VZD{JJt$bPw0&bXft0-{auQ
zM&5#<JNj;?R9t*jvN5}wce<oivK6TPI`BKkoe)2llaNvUV7Pbj<g)Sks6Frz<ts@A
zzR|&ZEqmFgDb2zoXb7e~kGFLMwwv%J^((NHe6EA}#(^9>Ug>7w+NgDOK^)wBb+Q3#
zsy>S+D!#VWJNg-C$|1ZTD{a&vk?lQS0iBCfuGym2L%J<qE;DSNwuMi|lupJLj;GC?
z%o;sZo8BN3wmM9zvUx_X%0~L+daxpbRd|n@_YiRS%P9gn0GYT^Z4{ql8><HUPI9fw
zFbLY@p_ofC8gD(xu5>EU4Uw*qFFOItMHwSH7$G^#FRm^u=a|tch_BqCCL={`XLhN%
zRH%>qxJBt4(3|cM96xM`B^ZNKHV)K=F+`tYo4sAXp@ISXOT4+Mif#^0n}!91QK;Ox
zuD>x;MfR(4$HW@5e@BOqggc2{OZy2|ZmK{H?wwTu0LiV-<B7NbpgZgL0627Y2zoh0
z6n>fT+oq_0<JdPW=~y=GhoS-a9kv5lt5>l<MtE8`ZMN@WOH!}0OH%2my$hzY^8WYN
zRXF9Ok#SPr9I7<Q!9Bo(YIo$Eii%JJcy1w}Y6tMK5qyga$B0eZQ<XLk^t1qo4qu2F
zUYxiUZaS%ob<?vcjc&d{;yU_)rYtj$`+~b?(mYQr0f;gVxPzf0t)lWgN0Yf96-EfQ
zfkuv18jOVT58*yzdl|wxIRb+p#!+^2(v7EOQbLvP0wO`>&*cUe-Iq;Td56drxtr9O
zbe&!EPNY+|uU+N?tXT2hi(<2ZiGG{tMxCL!@<w<Q(Y{s1MlBLQSJLws0I;<oEQUV)
z3lP|^ww;#Ywr%fY^Va)M*i%NJ`nX52dGWaPQrdpnFN1{_Cc6z-CQ=JDYlGn_H(v;3
zxUF&Ti~gzKbgcrTPP`Maxu<~$5~rFa5eTQ65y%kY8CFu=f$>vi!H`oIa7PESIK9yD
zpy}yDKLjwGzYIgnBM@$?sNAs=Tx|~A0{>=ziNeMn?8t`|$8I=?e%}Ux^2K==S3b&U
zg7IfF7t1d823*I2RIF&J%lY*}xm2<YYpM)@>u3_69i3`qBbMnz5VU{t@^@v(cC?{f
z3NyHe&xsD?_{@G%laLqRNBzD#=C?)>SDhKrO1kuSbPT{5qd}&Kmqe&{<4nb;PyZjn
zDDo|RCIk2On>Num^Z?ok5`O*H%E!5k^(E%2nJDEC;DO`v@PA6PjLiQn&HkT}C)@up
zur6th*^;y&c3n`9x7hYhkp&`Z1LOgewX^!!HGi0X7`L9VfY0>Zo^X$nFt;6%CH)-^
zRW4&{12OWbGfwRM^omQ#!OQ+N&-l!_*Y(HM+rK@3<-hv>deiP@gwQJz88mgezZ~B-
z3KI2S{~kjZEvC~?{BVC4UXWMe#K@EBod1_NO?$YXHfD}x3Mj{{L@vqIo@btmdf+d6
zw>zjw&Cly|33!@#N`J<!-K%+z_t*37P3RjzE-0oiqbRh+)^9h8CkVV$9OHt19HM6G
zXvJ5joA0wtr^z@$X_CwWMq!BTmppQJu6ufGZivh@HyA2l3V+{s>OrJSAeZs;`TkJU
z_K=c;)*c6%yAhmLL5hCd2O`L2Er$Gnp62%2|I@)g?!i=vDq7}h61#`BuwOLWTo8=P
zZyBa(p9~hONnrzj52WtY6e>$1CxJqPG=?=vMGlFE2#PnBazReKgk8p3spzcisGkI|
zMk!w#z^@}u#Du&gERinJ*59QsAW*NcHsAWnGnF*6Vvpl75gM518iH5?Umlptio;X-
zOfi}^K*hx*JVq^k_1a7}U?pMD8T4;_54o64!xr8i6P<-RB?mB?=|IP@Rl{6T*~`Zu
z&@G^o^)E=8C{rT?tVZH@__u?WwDCS%MPqT&m<C{XV`*=D>*_;cO<{sA3Rl1zMcR^S
zdx<g;kr%v^_hJ%F)GdTfzSSiOKq{<y!qcN6(By`6t`@}PaWd;6P0#~j&?0RBFhq?#
zcHhPfaG0!dP;#;0W10eO$}+(b&n6c!N@GR^{hp?P777@b;}gIZLBfYidrR&X?j|aR
ze!>Ge#mwl+q~C`hvO6)2)zu$fKq~ik7%Kpo3RO3vF!x_T#Z)m|X$P}BlZ|?!K2FPI
zY&}x2$rLyxN<WQGJJBDyVOIYo)?wAc`Zq$`KqwD?{f#E@#u~*Fl2D3roxJ&Ct$u78
zetI6yapn015I9nK>9C}Np*LHWKChPLwF;&f1-plGXHapR@YWOt`!xJFFwmO&uj$qk
za(`k#v7SN~;@Fi(XO@n)X$+j@!U2-nPz#NY_EB|yb9n$O`-btzo|l(m691_6oZI56
z7@?9NCucSC(0@jI0b^sh;Wf=!_#0lkk}~#t->Y{C+YX;KF2P}d&u8Fog3Isji`KiS
zd@)Q;hs5L>nQjRP^@=RC23p%>sQ@|EX%6#ps&mgC<ymOk<;=>S2&;E|AYaf5#!T+n
zF}q1d^ll+hT;@UNYv#Hs0_I3ax0Bza?=>zXaRq1d1g*I>X$5;Y3dB*9Wr{~x=KT+Y
z&z72*4cV=iZ*Iz?bKzMrfh%LwYnQ&4vn82=jKJtK%Km-fKsOv3dEo)@JAATAmAEp1
zW2*%jl)h#Y->prEp?@nmp9+6$Zw7=sYHab7Uh?*6*qlRQU2gA?)qjB(`0Ri>p37CL
zXiZ?+g!_IQhZdy(6v{@~)t5FxuGSk<DY0g%;3bbvNFE@SqyUh6Q#Q<{v-+S8^|ROW
z1M65>SW*02bC|Of;EHSg)mn=4AxqIRg@$v@?u|{6*|*a)g0uzDHch!;YfFC6n6o@h
z4SE+xtaW@RGs6=F=w5?Hy5D|oE0@M7ojGnZem=_^A>v-yY5nc3ceFOLl#*`L8V9{1
zXw>OtG|4?ZOTm>&9CZdsqvPH*OM1%yQs`5CMNlTih!{=RrRpaS&<F5!UFvA|#7+~%
z>eU~6F=N*VBT6)K3QviAUX}6dYg1R|=}CxlzA@>^%<Yl~n@CM!{pNDZ7XfhRWO~Db
z$I>{le=FNknq@CBDKB^Z2nV3>3_fC4tkr^n%xvM3P;~POLG_}kEI5l4=G8<@)A@>&
zF5)m<;?mlTjEBW2DPM1Mu2#kBv)<BA1MA>p0V>uHn)~2!Abi3W3=W2kYfR<|Z|kp)
zBh1)^@D`c_iWTlTD_mJ)a%NlrHlo7F?5%T*L0dc?afp+XN{X<Or~##n1;L4tjX){R
zg%4YZk>F?$+cm%kSRScFagrEXSpUHj@$@${sP#8O;WPj`me2Sx)HKXc3~3bY5mcsm
zFa#+Pw(kcOrgc5<C%rhLaW=8zB($--Ys|Z|gochP`0#Rd6O`Z(0day8Ds>0#Q=Jeq
z23}v=F71y(%Eo2LG*Zuw77?Y>eQG`@K8Lxb?e4JM45?0Th{J`;NXaurmTPyU=6PE~
zcRy$q8tXe;DyiIdi!^La_(h(R*w<H&10kS!yXAwBAX7|QbVxSR1s|#)S_AW!UMrOF
za6jY2S$6=3stv5cQ?}{-OD<m^Z%BZw69T(bcV@6nBZ{y|IMAQha9TVppL^RH^`4Z0
zW+~nR4Bve=Z!9)bk$BLXkGLx{myttNyxzctoH9`Ke1z=~DueFG=(btgV*1lMVb~12
zbL)U@>$?3nGzd6wpEYtz)g5;y*R|wN-fcQ?&1~;nWo^kVjwk0O>Q!z3%xA%1U0Nu#
z*pty6m(0>?eD*$4CK)7+zfzVL$$-Ayo<LR+dx(nDE_2=uisF)c=8P%|O~QBD$RCFM
zqSZ9L<U=<;sksI?9o0QROj>yEL&w{$zCK43B|8MiT$*9?1IqTzi(||^2;>U95rq66
z6#74V7^dzhm&95j)&ioJcLks6*J#)LiU$P0bs4g-T&~kj+)z_oW;tbg1wgaVEg;Nm
zvi1~x>n_pJF{Cn6OmPbVGuE>YhdjNLY#0i24QRClFe_3L`J~n@rkxIX5-*=JtM#Fd
z+q~o?uyaw&!BjPV0%?G|kZz+s4h-3oDl0Sce_Ipc9^ic?wr-HPWkF@Mw2(3z1H~xF
z5L;F01?k+gQ*9_6?Xu6ei5GOFUd8*eu3#eLb#8D6hD49iC$*XsOF0FjH3Lsj>k1Cm
z#oR|3xfi;J#AfB$T%@=z<TO*HBctq+>QYzZo^7T1t1w=(PbvjB7a#ogFe}#p^w6JX
z-IFq6wv?I20IkYLsRmm}$?ZxK6eUl4DsG;eurpv_&^G|Xm#=jupV|0IqXGHRjNJaT
zrLV=!ynWL6TWVgKq06toWeJ6kKgH+qWl-H+DkEzXG~|;%xD|E9a|&CviAS5R+NzhP
zq!~vV-Gvy5W%ZcBcuz&_E!T9|Znak{^cAv9)>05oFFQ!OGdk9*m>N%Tjf_pW{0h@h
zN72LcPp2;ciKgH>g(-Y^^6LlR$$s;*jtKJW$erxDjr*Q@{AY`IKOpjnM<hZQJ6Q7N
z#mm#P(SGCH&S~YdU!bocbV@}YEjuDH77Acm6aM}5w^ZT2@(E=byUk~R=F;m#b1oQp
z=c4w#X=Z!sL<o|L%h~+fpbu28ib=eL)=X<C?}4s@S{l>Q&aTsP_y>>Ybn17WfgzGR
zFQU1uY0dRLE5X|O8g&`iwQN}wILKkz6yS%b)k~yLyH}?2-Hx-EbSc0(|0b~C(*2;7
z8VqeZ-?ih<7ph*knTT9ZcVkln4A|W+`k}s9p+<Yt*jb}*TlrI-@~r|YO>@ixEcn!u
zzv7E(r}zKIz{>I;0xK)~|4hKtgeCF&Qt|d$?Gb#^AZ`fuzoR1v#wbH4|I2o_Poz*~
zi;bbNfTB_9P5gG2gUsjc`(NA$*ChA<&5d}0JO%#cMp(#KBmnAb{&FLB7H;S>p=D;U
zc}Lq4A2Hs0{ug8K+?)y2y=x{N+qP}n?%1|%TOHf%*ha^8p4iC~c5FMD{?(kBGgarU
zsrLiyfA?B@?R8yuX*;iwTmw6po6qit9kM}s$A0p$G=?TjhGsce2?0Ie|Mo^8iGcWi
z9UQDu<ts6H!A|^-HzE;*6GD+c^fwpocW!}1zxb<Rn1GHD>Bq09zK;x6Qmf%ZjEdRW
zL`jU{SHfu0*rbL*IoxN{&Wi)f322;8M65(d(oaS{k^!ZRyD7?F0bjZ}6o@o<&;2c)
zYixUB-z!#$7#>yquzw&K1G;pyQaYE?ahBO!g+-}|&#TA?SdFEqGl?^4V4-t)O5c~^
zlDgiDokZOEpqZ%%N=>Y!BwX%$l$e9Mtks7d_~vLmPZq;yeyJn&P%Pg}y49-h&}|ky
zi9{DASMb7lX>-TUuR)S6hB(tr#v5Q}NGz+f`l8=abVeB2Wmw_PKZN?24qoX<Zt@Uf
zQqq24Y1THb@7`Smg=gMr7n+m3ftJW5SY@#`yWU71;{a0Hf5F3jkiqjPm%diro_%0I
z#+5N;bS#Oakgb5B&|adv?CIxA(VZANw`)H|8Nysj9C=gdxTvl}=-U`zo$tkRraTR2
zsb#Zgpc!^>f0CDXe;tlJ%U7b;ikb!0tsK-01%Gy)9PECmCN6+a!0KO6xe{9!qAN>%
zIx54U-Xqqz^l-xd!4%{HkTPdv9pms|WF{uV*z^81sON3Y(Tll@V12xvL(bpXshd>L
zo;(G_Bul$OUK!X#YfN~jTd5@OaDfZ)f!|Ynh$A>}n4TK$bQxX~DSjg;NF?DcI=$WR
zPzl8sW^!JMZ(K=l27bA}y`Q!dTIKbE1NU|4mRUvFB=||US#9MH8q4^#*e<PBZ{+;U
zUN8YYZ}fE&R@(<_CMO1U-i?3e$ILhYeAEd`zSEg69IQY~iET$UMu>%r)s-IfrB7PL
z&BJD9SUamk6~ammdFraJBqs}ZVn?IlqRz9}Ow*8`EQM>dBWz@E!!z(vc}a3_DS^PU
zgT)F`87YV->;rxuxYTLx_4+$fxOYg|qjwc>B|2n{aG|l05xweAbQA7&cldUw*OePT
zm?RQGKxJ`v0nTJaytBHF#4S|&pX}zi5<Fyj?gOt{m*1=S7?Yansyk++8PXYCLdFs~
zkB6^h-(@hLHY=qiMv_c|HB-wbx%^{mh&&0fxA=0?Or6Td#$2y26he51nQZYIp4Hv-
zi`$o#sNh8<se+vo#+}osoZMjJ^fclVIVr2Xlrp6+S3DsDF-sjNMOCm}4Hh`dQT40>
zJc610L-(8mKV8KrHN5j6dYPyetMEGVi4}{NaolE(+<Hj=)n{&HRD`%DKoIW4O}9rn
zS{whpb8|=v4dX^~!+si3g0YKwlOD6}mdPiE{<=c5f=`}go;{tmRDVbY6Za#}ZjdtZ
zg(t>pY29r?O$Qv`1Q9YpRiu|{ttdt*j*_>0NRzrvPM9Y+LwOd#AKh|T5)R^T#a#ah
z5uxL7`2jba6!mME(m}>INf>q<b%)54=*$tQGN35olaBS3O8rk;8-!Gcsr=DQ=0kCr
zn}1YOuDO9Xj~OE5uMZWg&5CZc)g=poIKi*z{k|j2j+TVv1h)xN=aGY%R(-ZmpS7MO
zr9mTu6j<ZTqcyP&d!1TeSN0GdDEIlIkmAYU-Gz3f{q{l+3cD5(sdZT#xdazgBe~f*
zjZ4Ty;Vnel_V5K-87)bd0@j2YC<j$2dOW_A(prvsXk{8nTGfvzpvS*7?A9O3xR7%5
z3?Y-#7zw$gg=d;qB_6~KDI18@jQUk<Ky4DSRInK&4b$Hb2)NpvLczajIeWwm2h@{j
zTe~gWl4AHLHC(Z$7*VaTf7ZP|4XPqkMWFSs6h3sv1xHumv;E>5xg1*V{YR)3zqu6P
zN4cNvxe(EisDZd|B<Av2N~=y4%r)3Z2B_InoMxiWiJ|%Rz=Vf&T~G-vSX=2r$qC&)
zk@pV?{%mVVCKpIVVD%Ulu1mgFg%zt>tiAIVKeIv<y#vrt^rouvl{q7M&V|RTMSc=`
zR%oowJXX_QIZ>uM=HI<g?asw6abu4t#eilr5*+K@RzwGO+LP36tY|*Cmva8pxfeFf
zK&5VS|5)mEP+i9$!i(uD@!L1@wIbtb8s;_8lN-+a+$na@u#zyezqV+78uL|U+U#?@
z=|p8eXiHDdi2bv5>7^T_kK}jTB~GinsIXnPl>poB6=f_|M~%v8Oyae3&_Wv+A1ctY
zmVQZ}4nB!<JkJJR0g!Hx7}Jy}@jk)u=o&BUTexVWu@mbMe0@et_7PI(2)6<f9Igi6
zm70)Tn?f3d<IB~KGk;=EpdBgbUtN`!y=L1%Ec+Bnc)t1m>dZuR3lh|=NlleGebc`}
zt$v8_iF8ZQN(p&LRUEeAh;k*~2mRioZ>F0z^dt@`qLyYG_fxF)V13E*P>zu3@;Lr@
zw$g+<scxwjdj`F<ArUCjCpdiHSVp;p*;xMFP^^D>@lU6v&g3AUqB4#l$sDZpSQI~|
zGgEV{2zT17P9yQ)fR%K=I9J<8A_TORd};ne3sH%L5J{Gt;|m%8o0;GQW17qP-f|X|
z7X+X@X<iDGt-CQbKdHMM>4WE^Z7O6#cENyUclTS|J-Dp(95BQa72Cy;Kkm&+pM+`D
z=j1V+Y}O(<HCXy8!8*}G-~{ZLg-BO!+G4oFFx|#@nc-kH5=DgVa4nYGP+I-Bs_EF{
z>E(*0gzSwjdX^8I`#b=W%-V$Z;&Vn<{aw)ixpsC;V?mPzMm53jB6Sy69Y?=oIK>9M
zNSiWuPvPGbx5Gm3VqWI>(WPZ<i}h%1B+zN^nD=S7vDP%H7C?OB?~iby^Bh1HNab5i
zfkUIR4~RV*=Y8n3FJ=MjYj%sh9$(ijU7W6~Lugr@MiyW0pa26vRR+252pH$`-R}{j
z-R*e;KgVdj-h8?8J*nlbIsdxTp4TyusY3}6RhKY*1~aF|z^lqls8uMip)DZ*^iF8U
z%AEdYk2G2OqBLJ@5<!RJBSk}6za4?LX~IW(Eu7raS}p)i=jHq*nqlU4RL8&oWkUiO
z_E`rOR+UZ`(4|s*t03{tP6a!o&<u5P3Xo26*tLWp8x$;j$!a!c>HBhUUKHgQaA6+7
zde)D1TlIVf{1Q!h#{GRY2gz@|mz2*-3L7Zk^PrQFyz2H&NN|4kVf!)Ka}IYPa5}=T
zf9+KWaYp(l6)WO(gbmg3TXTNe6x$1qFdD*D^zfqyxO7sz1sK{+kt`M+UzLE*ayUut
zh9abUEkC?~#yd|3@dWj>{)F~Nes2pZ&<nfVL3jOvg0xt#`HvkY$A8!YW#-}fuOdE2
zbpB@x6u#$K^94#0g+>K#j(JWW=N}6b@;?!3G73E8-+|Dr)oLDT0s;a&f2#FM!~OQ}
z#+w8`-51K46zd%oh4kR_a_`^~6&(Zfg$b=*;dC5y1bINr5puV`1V7(uICNgSCnH-p
zB?T3}o?Q;_slW^*gBXao16{yp{plIqWB2&s#Pam(lss-&7re_4d}92}AfS(TwgY??
zBpxWKa7Q?EYse@T)OaKaNjs8jCP~T0=n59E;GNxL6=`Z;hen3f;cGOoy$tXqbVfgZ
z%W3aHMiXLSv(*dBrHa-uek*cLCR5nY8Wg2cPQH`w7<4gS25zSj@=(_TyW61)C_KH>
zz!2t|O*%O0P2kGP{74D)8rr&CaKsN|cM5Sm4w+BT*8$qlN(d}-@9)XEWji|2z@3a-
zSTimvtC$(31}E)ruW0rB*l=jCq?CszZ|)wqm0Q0`9WlaoDj}F3jF~yf!~X!gFmua}
zvsK&UJc=gVGaeTM38pUTanfi2q!6tyMa;C8gWqD&0%6XjaFP}>@d2dXa1o0PMnSJx
zpys&2k&VRNZ3N)dFbZU|tIu!!)>H%YlRolx=oQ#-7?wGueEQ^F@Vn|zL*%SXRlD9q
zEz=3-6aUGduMl>{Uc;)q!*P+Asw3Fce8GYB5G(_eN@SzP>+b?QEqA`9!2Uw|Fa;;1
z$KM5oe!jWy{!&$!dzuc347m4r+gNb^0AvDIgCB&wU<9as^P5ofBr<s7ACD($<q9Ho
z-0ZJjtAfuQCfNIjK7WjAL4d7J#&@`ph<H++=~kNeKXK4huYMe1q;xIawoO;Mtme6`
z@}5GN*AR*FpNeu5mtLDubgfj&knGo7C#_mB!B~ebkQsy)=Ca+xN*RQTfZAIKm|(r4
z<E{t4+AYF3wWSMk3alQ=P)Jp)Dg?V$T9*Jyoj)<`tM2s_veqfH(~Ry<;>@EJDYXa}
z1+<~qecjR^$9QNijoX?|&{}+SV2piStQE60YbR=b1r_j7D_aXNDnqD9J_LzU^>9BR
zqfO3O>ZJ|Ra_!b+Pv$2f<9JLB{hKmc=}rdul)T0=4;pCofljhrsKB26-I@-KMW=rB
z#Ofi8BVfGa#OP>mWNIk*ng*tWS!azTQnvx(hFboS%f3}*+Zk%~?HZwUk(*4m8Df^T
z6VYho-x5bx{rPUQ#m68XXy-lc*^HQ4NC$Dz_>#8wWh4B7QMBj`rp5MIN5yZut++Vs
z>ihDY?|^6|=&;car{p&hki4(YgNNhA+54&y@FUwQx5ag9TZ!PbJnaz^=AV?OoLUlN
zp8|!JL7oT~xPnmlO~K@4P@q?6?;kL#6D$^UHh!+`(`3amM;{wk_?>c_5VHk^qYp&B
z+qJKSvDwi@O>$GHmx%Q%j9Kj#Kiu?`N0q1W$-2bp<PAoT{W2Vi|5kRhs;;w=WNTE=
z8%MFp&$76$zPp~Z)`$mF2;X&fC=oPS_NEKep@$J+_fGJbw|Eo#&5ES0$>!bIhn~US
zUKMHzCkRxw<O$BiDJzMXGu%i?NQD&Qhh3u+z))O65xxgKXI|j;T(FkT-nj40ybwQr
zAz$IPcFGl<fj@&cvbsPAxnKR9&n)lNCv114vX#u58o9`x;y>>k+QuV{Cb36$E%duV
zXcQG0G0?fL3^+KW&7k$i<0@3T;#Wbg({S|4P)qZ;uTG+f&B(KHfi6w7Go@8HjMw+B
z+G;KJ!)HJioI<nBhdR?=6h09lQ}LLD03P@#97@j*BG5d$lwNN>W|H@!JjJ1(1y5_m
z>M>LsscV?Wbixx{%v!4Q9WISp%4qIfOzQNhJ0h0WJghjhoL0WpMc^hS-`>;q+~CTU
zwQxQr)4ZOPmRmdAc9xy)U$(#B;O)o*;-a6-`pjGHP1-bB0dXtwH%_S3l>P}ivL-x6
z#`%xZaudU63C9@mI<kW`%8m+_^4v(96o4z3w1{VlqR$LY6++V67G=H>3jzToc3O2Z
zz2E9$l|texD(0(HUTebQ^=Z@`FsV^nf+Q&NX45qjDzzJc%T|ZTfHYMVi5^c%>aSxT
zI%~VG*yVc>hgA(wj(FpAHtd@24X)tlF3QchcJzwr-6IN#pZ@QgBC?YPPR&6?M!wi#
zypU|y>?J$E8Ny*ry>8Gu;@kZfUe=;}pE)6_O0G6m!4;wJ3Lmf@P3V_+AFx`zp6|Xh
z1hyL)U~vS_<}82aMeTDNsx6qpq@Nl?h-5jJ-A*~C2--Ql&l&<!l0C3>sny@Hl9ZH2
z%JiS4i?Bj%^2B$8Az-oUmfbm)wXE}vb%`uBX9|_)ETHYaXFNSw+^5EDlt5)@Wbwq#
zr}fzX?J5F4!R3xwq{<MHG6{WrF*iB77s8zbG*vwGHFw86&a+X!$$*xZ%a-$b-XRkN
z>9*RQuwJMxK!a76O<K=Af2<&>Brjb)siEl4JU8T<o>e;oUH^wc+}Lza*|-C9gLtv%
z>>ih_p_5B=Y>qVHyAwzc{YGgXy-2K;f$hD8wLUKJOC&Mn=h5-93YGH<k7%i(&xAy5
z$)+ebMe8!~^TwbimE`qw7_(O<>qr=<sd-CE{tjEm);N1jlIP%MsJ@D+^;z$DMLfLg
zqn{h|_cY`*q*hEs5Q3nWL5_m0bbI3s3$knv<pU`{ym2P(u@~GRxM9=KJBitW>Z<_E
zWAF$4D_WT64I2$C`m@Gg&m&8%H*Q0~)P8-%N6E0%Z4)t~$qU~+>s+Co<+f)ZF&`v)
z{&k`0y*kJ%QWZff+JEdn{tdJK4^tBhEAxNXf9T02khP=qUf0hNF%QvV8upt6;jEhy
zw`2Rlxxnt{`u+O;CXr}*v1;m~xU^b<iYsy8JlC&zlaA~CcrIe<3-Eq_I&{2_U^IY$
zBV^hM=mXiA`@mv1bbHzdO=bJ{Lpyd@VAv`-R?iiXv@&`#b${G{{d7nL7AzH%g4O-;
z7pC{A(W}-pmO;_DMsbP{I=r9E|ArUfd5AR&QqT1L_Vmo8qvhyVq`Wc`b6=8~GfGv8
zgC9tmKTF!@<&Y!DHkN*hGApCmakXgm-Kqy5JEJBP`EU^Hko+AN%UWNndH31QOSzc-
z8buMwyHDG)vq_CQle^`QQRkLzd&KM-ajBc8;ONz*cvLb=_MU^B79X%u9Uu@$8zO3Z
zTwX)hcHm30D?zuf$-JWXWoBEa8Tg(?1HM*{oI2mK_RbjPoy*jQ4i`)4PL9S?|H~<<
zQYWX%eQO_VT{Mbr(Y}BoZ95S2d}DfKe-<&#qE^jDhx!nE@*eUwQ>nPhTk`dZR&fo>
zi4~0Qw{`gz+P3knA$Cv9_NzcT-$VjWo2NIW%HPU6JIH*L1y`D6ulRkD_a}%LT2z3m
z`TJhKX<UtKVAyV2jB&nvU({FOS;KUVw+6sDYV_})pMIvLgu}3H>+)BKS-aW>@r0=)
zT=#jb)-#gLkyD-rlgIl&;3&<khWEMFjEoWx?FttBfPu{%)0{JW*^S!fx6RYMJMF6=
zrl3@~*)Obymj+l<mh1Fc0St>rbD?g-@D*?io5t$l6pfz);FvR7SonQi6^*}oeUHUI
zFjG!gN7AFhcZiF+Fq5VxX)%HhJbyKaw~80x+~FZ+rPGG5VdU0<POra}GRFDB@L?dj
zlhny5p*|-i&+}-c6krGXdves`wB?nGKj?Td`9&cV$wf;$_9^THygnTU$}N~urX6p7
ze;vqmiWsHc--hIl1~CSH9pAlY1Wq?cmmNRr#hbg@9D88^GZ$%W!Bh`&2#NHco?5A+
z8pzEnwo%A;sPOg^t*hmC#1;QNZWN#90cks(R3s7@aXDF$1&g)^z7zGY&b6)$8QNT$
zO=D6$<)jLT-3pr$5}Q8ds-1xEB$h-p5*|J3(@2^YoKWTYp(2uvxS#MMB#duMs{%;9
zNlDpwXsdYLD)EnOFH&d6dhT25LK#V>pIc`c27zHxv<E$JmUXdHv5FP31Wbb6oQk+z
z>NtEu3P0~v+bAfy3Qnrn8=>su=ZFW+n}F4B+96M|{sY~T;XF->bj&Wf16GGiR{*zc
zGDJ-gg?Nj{j?k(F()u)Kc*|(|B3)UH=ftk``PG$Ne8!S3r!ZkOyb0gtubt?xya<tR
zh4LTcf9}nrg;64`_toS9O;O5xW0$%vsr0qgD(Y1Azg2&>71y(6AvDH}X&)spaF%xx
z|B42&q%cJCs@F^}wDM&3ov6%tuK?LMS#=fUQ9NEVv_CRMXw9NKgFtbgEyny6GXzF!
zYh}$ol$>m9I~E*oTA1Tdfu1Z@{17{dYjOtWbhJ3U?ZloqBuj>>0Nh_84UgA#gpRQo
zePV}aXd4emRH9CfUX$Giw>+dWfO?m?&GyAh;G2Ts&1{lSUti|4J*U6)AYv&wHfEu(
z)O{x3{5)PNVBuyY8Kfsz&yE*fRjh>CNB<O%nQ~fK(2JB;GdmRk(4^C$o~KjkA5`6X
zPP;ZrcPtcKp!VCtjB}5mTL_}+O~BN-L6xfdpQ9`@1Vrjft(i8qJhr}~$`QR-_3I&D
z9fDlOyBK!3@;%>b9uA8B)pPELVFQ^~l&^hvUf6xJ^nXKJMRyB6$L3R=p3sS`&zrIM
z-6|hP)(dr97q^w+njm0%#QiWMMo*m(14j;ZXwF)!IU<r2LHn=jK32+>wM)dKFfAc?
zI@D#=ZScWWw56tb*X$<yYgp}|s^hX~W{n(Mdg$-CMhEBjUp*w+tv2z$k&r*}hN1xA
zhC&Y0t2UpF9LXv}g=_z0zhy)(MnH5R;L$fPUwSMS>qW}qM(BrvuFFwJ3(ggAZ0SdI
z7*U3*lK%d&aY@<YlTigJu99V*C_K6>9p!!n3ksVSdL?;vMb4IGe1d9zK|0Vw(Kl_y
z*DuvFGLHpY4ncm-r;_>jIV>3UBSRn;loGa;etXii-gMrRcxs=n5~>kF#0nD~uzXjX
z;e1gH?U`!8Lwh&AMbU$ZQ*}hlFSat(Tnl=#Q}q<#wT8a>%T!WS%l#%*pBgQdCFW~7
z_|jQiUz7gKiwi;BsY{YKT<c9TqT;~4Lg2&V-b`a&WA8bLXC$^0YC%L(Y@ESBuWi=m
zT7}lBE>l!nJb$%Ts~Xw4EYPZoLi{xwn9W2bq1jfWN<gJ7<Y2UAV`ZlX9iuXdTd=>f
ztCt~-uJ5Z`kdtC{7LAf}cO<D@hfgM=zGvcvV+hehMwCi5%Fm!_Wx2{2OU8+SjKF}Y
zR;O~cmi5{Rc)X9U*?A^F?rvK0RI{z~%g{rS)}ViF0n(&bAA8pe=nhFAb(fA1Hk24M
zKEnag;I8KJ_0xW2a^cgp?^`oM<->p}nVsWnm;sm4u9Ho(LZ+>tPAZ-DI!o`54fET7
zTgv>tZ6-))<D5PrXgzg#zFZIyJ-|w^P7`~U4%UDAl-D&6C0gX{X#8h4_mCUC0g;9h
z$&mDrJ7(jf#Y6Y%qFS$oI0cDj8L*Uhv|U!MM5sVcrz+5L7cqlfnL*cChOqBg5Mw&<
zwL9eIWKey5`&lrM+C%WFR*ms`)RnRP)*_oSt9H1({0mE<wM?dKI%aFO&N@UL)O!9(
z>)6wN029~MREWdeAY_kQZ23@^iGDIo{#mJRdiRvrwEcOB?)t6CQ?_`$hH1K)$c}d*
znf^oj{fqK4BIL8bH?Ydt6Y>5d)BZcpRcB0lqOv3vp|7BI(~<oslt;*-ednamEA4BI
zq}@5AuhWaC>+VATtEJ%3vaFzIipdhjV%cwyvCp(`Kb1%It)H<k7cY~;tzqiA_^#pP
zUaafmB+jLUhhmChTnsz9x0xk13GYL47U$1guWLr1P(cI#E2sSw^s?o}-MxeYRm!Nt
z<(;pGN$9<*HvvtPf_}KPv#F<HNwPhn_mQvXb7=GcJ-(o;=MHT>mf=)uOq{2N_p%{@
zlSTh*#oZ6uK4dF-HJV}Nc>4l$pC3GiTuxW+Vd24&ne9XUT(CVos9;g4i0oorC&&Gw
zNK|tU9eiS4tWzMDlYd|~(yG-8GWB(=u7x}A{;tgtIR|uRY+3LIyO`(m&GL!Y5=UtF
z&4#&nubWI*!XkLkWb+5PU}<L~I1I9N8JH=7j{17eH&Y!8+;rZyz#YG-a<AsFFut|x
zI@J;3mE2vj?dsX157uN_dJ0zh9!;7T-67}dHgvLadsY|K4qm_6$7t_uS+8{7EzMMA
zeq|-vyQ}?ez}HFa!)%h6{n?RK3BNAs2hw!-WE^CNzz{?<z|cQC%#uC%0!6%eH+%#_
zIe;*}lGqEno{qw1M#r`jVJE=^M^Be~Y01hMhw`?vIY3Fa3e4fB(zb)@Wgzp_@IV@D
z*9uycFKcTDc_GMt?r#Tw1L;lULEPPR^9*-v{M5Q?-qmf(dJbq?->MD(X`?7;;s&5A
z7;RR)i&<&#mlhP{2p|`!o=ng1<hdfJ#5nN@eFX0`5KuB)c6hq6l_C+h?RMa`B51yW
z<D*Ku?sOshT$$v~Fc+0&O+G1gej{Qxc_1E$T%bg3@tp<qN^p6QPbK9k$iKY76-w?B
z$|eA=hLp&VsTTvsZ;Bfw`S3H+N8ZR02u5{&(YnxeU>r3ENzav#mM;>4iT^0sd0ZIE
zIxDj<G<u;qk5&~VDtVy)WdQEODNQp6tZ7pOq2X!teHzPJcn(dKbhRLFwz48BVdo_<
z(t3C$NnE(s;kbaIEF{?l!={Mv?IOtXbRFCcMl-VdLt$i@jor(a_aIz!>-Q}34t)`L
zx)PCz|AZlvE_OD+U)OFc5FAV-j6-tgvTtEg@T>f08q$bVN%rQ&-KyJ<2ae7ouaxB=
z!9QpElllD(&eJ@R(;VGy<w!SAQ~UGzi=zK&_q5iO@^kIvuXP8Ta3_+5JTk*B`7&Cq
z+0PAzP;Gm(aUIZ=FCo?d>TzDC$nLE#BEM^Bpa1y$;Q2o@_gFbt|2u?fS$8Vlv;zUS
z`fvQHLngcQkDsf!Yr)F|g7?5R!Ln)tNbkR6lu#eL6+Js}CkE{^>guGEmi|3R4i9&*
zykDR0zCI5#x<+Hrwu4kNt^@Ug2&Lb!$Psj0ok0d|{`h_0-#u&6q?OL{`76oRo*taE
zmpi(C77Z-IH+%%2V@2L<998LN70U<6h2L=of~nrHFM3lQ_;9_=!QKRXjy>-fkT)B=
zs_9_kF|Pb3cAzbIz>GM@5|<Ce*+-4CeQ*XpfL3OSmb`i(*hhQgG#BIzz?HK1hq|k9
zb~n&Nn}RhbH1T(M;ZYviY)Ch+V-Ca<szJ_7Uf%}2Lqyf=&NlXm!i3Mw?#_A9C_2rm
zrc9C~nS9OYisa}LPHsnFp4!YLW!Rp|(^QMgp<ZS`&ZDW(CG=tN)3R!rsNYHG;Giko
zT$>uys6kz*?sS<amSeoj55W5(s8t@Y<H$eV@)1bQO0LLhpO)~knc}U|Qt>nN;A1%i
zHZ5M&llJJAi=<m@xf^iQnK`7bDnqJJA6U<-5c_v?zTIKvl^mYz^xB^p*fO;kj4L{x
z7+}CXF5|wVb(aPbpp(p2CF@xpacws^7A3M%UaR9fPN$!vV?eyNA<t9ah{>wQH4%_+
zZw)TRq8Dzk7)O^}bc>=R-4=HP`fusggIPprI2L!*miK=xWa2Dm2+vy-Qho(13LGrq
z0BF&|Lc*1b%7uTG>r`($cz?DR5o@s0{KFTh$V|HL97LrFNl2)|89DVJW}=T@&5NmS
zi2LRb*b|JdsowHxoYnorMiU0rhvU=I9LZL&WY_*S*#x;+q~NZBz-esb(D~hJeN?{`
z{n_!w0gm;&Iq*n!A0?r0221deX#DOf9^lCCsVsu}H1IgT*J@$ULCk*w6eXvXa=>4=
z7I7|$m5tG5`}~E&YP073{CaB%jfD&+W#T?%8C8YIO~x9St~7|EvvRalC<kQ-(bf7H
zho3EB;RU>@om{~26tJU;T2qMe_+jdFe1LSLf~?3!;7*G>S;#4AX3ILYYvDPBekkG<
z)QUj#O2ay&J}<+RI#y+ZxsFD_Ju*Ky5Hjcb$p!8rfNDuAP-GsbiD#wQ-^=N1Q24Vg
z<t-!dQsGP&2S*9~=@X>Dw_;r15^*WB{tW*drlKMlVdO>KyTJzrKAiKBvecJp4+&6P
zx1TC}vwStAp0ldmR1E(yDj#;(8jlV9Do!|5Z|KE{9$p)^H0xW<(sJO6&22BwYQB^0
z7bSlSYIQ|+!sX;p&RCeyQ!8HMj*`d&zm}fAS~R8?TWi#(F%JvRN_hS8Gdh#8-<y1X
z;Z;=(ws;LZZX+!iR)IT|?NaPhz$0%7bG-I-1W{3T!-;mN0@=GuUu3l(m;qSf!%-V>
zThnm#wCJ6<-)e2nI(Ende6SY)I5bU^V0ihKsRU+2t+Ed;#CM_8JuYOFyv&b)xj+4w
zYck}VAshJm@Fbqy5yku-cg|;qiNZ?+%mgnN<OMgSKP!d{;9R6_R+%?3O!04zX$pJA
zpDj+qWQ;@AyvwT<pDwDg8$^voB~%w9Nz1}g&S69p^l)W@n{7oDh}638)m0PPcg~^_
zi1h4&>H1@5R$VN$Ax`Fik{6`u>k?*`Dn>jGyO#$3O_WYtr=6~*e;@N0L*jl!8Sa$z
z^#%mK&i<g8<0S!Q`u@155ELf%6wUxC6)&9c`*MFhIDGb2Pk;HyYnqU$UC=R=Y}GNQ
z<`*pwheh)3(l;K3b6p7Uu8rD-HjTz_K^rzVN1zMJjlc`!^iH)zyloH)1XFPnx(7|i
zXDiK;dN+<C7lE$CL`R#GpzUkRafAv5BH%#@mo^lR8#ycPKyEi9`X0E1BU5WE`TG~T
zqKb**qeW6csZc-MG<1=%-CT6NJg*#zVV6>ER(hzcUV+u|sx-rY`gcA<JP#ZKdHdkx
zuPH2KwDWyzt@yO(`+FhMp50uNPi3&Z`w>>V_iZBa0UZ6uo%pH5yaxeW2Ud!ii?Eop
zciar6nF5H@Z|~5F2DVXq@bF%!XWg^2E<bTOyh-0v@KZ*KMPuwFI|9W|WctStFG9EF
zLNRCF2YfpsyE9xZdops^b`F90N4+<b(<GTIrWU37e%Qq;pw79MuIMxVD!?h!Y$fK-
z#Jfm=E9b3C3F3MV<?V-VhvoZDP+)WX3jcNX2WGk?^=WIs18&b=TG}e<_=<jmX48~+
zU58nQq~f}LK2NXb%A)P29JA}frbCB(7M<@YceRGb!e->Q>y}|hM4hugXVrM~dj;v@
zNpW;{vs6*^<h-$(vgdGAVfYiluOJJ?__mAlZ>8bJT_7<qCwpi&4MRAChLN#nmfJPY
zferWfyH-t-!ugmHDI_wB69{(9X633dD=qrub8cvBWfN!LC;#(Gv0gj{Cs215*+nNL
zZ=3Q%c`hE-!q~|1<~dz>s+5cI^hfTLXb}fZt240^Fiijrp!?j;<UJ^G?8yLtkm0%(
zs##RG6{k4qVK^$lB)U4kQ!1dH+&IqyF40<;%Bn8AAjXikyl`Vg*V=`D3)G%rQ}J@C
z>O{K<?Z+t_YyIL{$7WYh)lH7Tc*&Ra$Ke5Xr&$v96PKmB*(X+^BZf4ni!G=GsmVzn
zK|)kKA?HvVDFoa@W0lIOgUD5@9_M;drW?w@cDjnSU&W3+2G|?TIF5Vh$>KEl@>++S
zVoW^BV@ni06AmV9`3_}lxFyrzpG0@~o+Fpoa%#!8n&`oW=dbAODiv}z&j`i4g@zNf
z5onRuSyHxjGwO?g;?2YnaLSj)7=$lywVa^CzVJ#am7@JKL|x{Ha7^^P)r(4p4@5nS
z_%V=f&cdg}km3O7vhoi$RL!~whg_UOuJ+fupI<L|4Txp$8=>_9mE-n_C4#13zpn!r
zc&mKm_%)!{nSL*5&&-&mj)o#9kaT{An;+)05B0hoHA`jzI0S0@HR$zqK=<~<Dpx%N
z^{ks4S<FZ~?QCf;UH<ep9AiA9NPxT-|EQ=9IWwd8s2g`A)*c&HJH5FCDxE*+<|Zx~
zdU%O`?7>*)If}vFW;g#~=8bprh&13iD&7SyC1lEex2r$IJ!+M%qS(>Rg86HjO=V|n
z1D;kvpG4-uirC=pE_W%KN4-|Jn{+HH!WSEp%%j60S5~Wr=&npM?mDKf;*DeN)h*#o
zeTF<Lk?^DO@$7|DQb5!Sa*ClW!_!}a$^GqO9o!EK^e`A1{G|f6sX-m~ADizkgyR#&
zzjN97sDC`)IIUWwY@)dwZl@rrd`pFIXYGB~+-u{J@h@-xj45C^HR$`xzUp&&gErM!
z07{E4HTv!7MyPt^$gBAcVaCC^@t@j7Ru+!`V_MC^&CT*(O{;<be#`N=kb7n|ACNLF
zlGOD;M1mya;s0WqI*2HU<fb6qeiNyfxoUdKWzUZ*AZDlgR6Z_OlluOBcqT8ORh(r~
zBvQnsqQehXUz}Qnc?tR3NBmmyd*TOlv9Zr9=Vxz<3eFe*J=8yPNGQFJPjKNY6mrZn
zNQP)FumQTur=Iu7mmQ!;Q&6BFC5?-}W1W{KvYjM^9se$55a6FY@sL0TrR#wg6H+_n
z(d-jNAHHT__TIXG?ekTu1t!uts)4K<8qiAw@IwqSPc;j39rJf$@LQ%7$R19~nT3-2
zwQPir7iS3fPjfvU3W!XBxosNo>4F+8WXh~6K3lKip&ZOXPIr3Y|9Cq*JS`|k#V>`x
z5yG*Duon7!y`<6mpGf3_)PIr4Z(pkqkEiX*Hqq~1vB=StiKEEwl}eJx9TSC^h(Cjs
zutsxC|Jbe2#x{>FME!^Wkv8!vDgwfHXMQ=Q?$xYQ)C%sP{VuY|X#!$@a7%REvEi_n
zYwLigOlX3Y6v5M~pkc>Y{)<Dtz`Yx@URZqIJy~C8I=|5zg`u*(;gVqS$s(1ffm;ju
z?wlsyMOyxYA7C#+ytbW74J)d0ebnZW4+{L31B*AmXnNQ7_W~2&;E<M=BGwhIZi8|k
zhD+I@3~z--F&0LpvF5q)Z<I4As|{h}&u!?KRCjPY$%1UV^aB!A>Cjg3!Ilf#-s~Hp
zZ&CUkM43*-6|s(e8#|e!Y7vj5tb(|<)aDk=thpk2-U-&%bvvQT54gtVp9on}udJ)a
z$JJG7+E;3MP;kGb(cf=}#oGp$%z6^HF)l=s^<DfhB7{?%FsVpU|N19C(Vup%ME=X!
z-aECDv-0F^ZCHy3Y(4#8E0jI9dkc_#CX%(?xzL}4i7N(}Ce`0zY}$cr+cbMv<IiiZ
z*Q{HxJ1pVef90}f%<Caz&ns6`{pKKD_K*$>Yi7M-pYGF|)qzPgA0DJz=Uu6~m2>AP
zxa^v;fg~@61XztMD5<?~a+1zL?>bu~W)@?RN9-lm(D}^R2~vLVLt{rW*P#B;wJGN%
zQtd8I&3a_3MqgL*E_I5l6etoj{A01=K!=CE9h3z_CL52#qc50@wPrZc94enKfe>yx
z)@>e$D_-*mP61J9ZQ2uymri3^FoBZpqbUX4FJJ;Yg36WhP$wIY97s_EYp&6~!Lv`A
zbstAmMv(H}5NBBJXn>XKTdtg~ooUQ{AEBfoPDtf1MMSOt%OW2SW#n%$OY(Y^eJE8`
zhG!?1NKYVndY4J|XHt)aEUlW;e|I?Q9>h-dwrt(6eXe9i)aPO#DKr!!w~`n$c;*1(
zMV59o3EzsU+x=j%y$$EpI_HI)sn837edEs`dg+l;=T%`6Ec24O=79<VQotZG+m5~J
zhaOcmJo#SiWdDl0U+e?d4&UMq(8AU9&w6(AnvfSi6<KQN{J){H2|uQc?tq3fV766z
zl5;+Q5XuNI4Q6p1tY_gUPYEU->%AKiucgIf>^EUGFlmX8mci7kot~P-Rg8|YOXPSe
zUIiu`b}<ABH=3TIf}q#s1C1cVTuPt0)xHyFn~ign1+Nyh?V>%_^od*J*>iQkyzP1k
z9<olHO-f|dYYdvuMvh!NJt)UCTP1|5Y(6?DhX3OzujN<5W~Pod9eDw6iIlFqNe(}_
z27SwV<b5fry~azp1j*mSeBJ4QE;WXlb$*;Ihsr==pSmAA&qa?CcX;Vx6~S0oRd+x^
zhdl;6WP1(AbV!|-&XZN2dh7NcnlwM1_0+26HC9l+jT&r9djt9!^=W}liMlp}McjNQ
z#)>$s!X@gso@w_f=Mdn^s6_WU_l|m_q4PoAFu+5P0rq-{Y=6VL)1tEr+q679y%<S9
zmX6XCUDKpnWyC&hhh_h72VJL>c=w`P{bi<{wsQ2RIG4>gOw$6Z;?Bfu{j<kIO`vQW
z{%Dq|vBsMCkD^_~PiS(t?5oxZcMepa%Wop`;eSg5&L2MH)KuKH3Bsb~rz<p9q-4HR
zI#x@p7X0pl&tR+u2EjF!{44F&T%TC<{a25o>qR==QO+|YQ?7^IcZVq(^2Sg@+(S|a
zrP*w~UgmV?))9-zVHX}bF)owu$x|TmOs#2TzgxSW<t||xUmUI%bU+gYJVRLkn>J}p
z8czfp>U^mPyP7>68Yh|4gb({VC@?`WY!$*h%R(q5z$~Uqon){osev17{~O97ktew_
z{ttpq6_xQL1W*sS-2;~<v??sPJA&;%s53UB89RT|ZpKzEIMG4vu!i^95R1J(@@U3*
z%4h~bK$|vf6WIl7^Ge|C_cch@)Zg!oy>|j>bbMI4OFS)f_)+@}z9uW7$;+ug0sW45
z)HilSk35vK&k6KFTcsu>6V<s~?zr|Een(@Cp~O3%r6!45W5*yCF8*7wie+C5hW7%a
zm4T%(WZU|OwC*!uFE~%HdI-oci-j|YFM2$}>;8_PKXvvHQuzq80=^)k+Oz(%+mMC*
zf7j{%A8ifG|EsOx=Hd9Sia0=5-%+m}#s9hf&;Mv^Tqj<4oud1N1<%-nLLr@DTR1+9
zCeFnx$#51roxc5NZ?06LTWvHJboBAkkFSz5hW4hj*vQpgZfO)D0<BQ<t2lFX;*gr;
z!}-Y>(bO+h^42=l+!jpqt5VdjQ1e}{^ks(>_oPVn!Ls>KB$Kx~waSVXq$smxInGG8
z7w@(-%J#Q7Nb;yxzYMC-l=C+?#HU?1Ig^$!i&+%Snoa1DTYnm0%LB~s>@sDCipSI{
zTI)cLqjKNSOMY%eMo^=nd7@YtRq10rm@{ld+AxpuyJ{)0c6twaCExzGe8@=lrJ7ca
zet@yc3Da5aj2Q86t5zhfwPoOyK{yw#NolcRNynbnKWf}4A@9L6sATOExi3k3vr-a~
zKc2T*IO@>_s3Hi6(IX>B_Scld`tGxsURExUX;iD4^NX@d-%I0~oYPq%7^rlHATa5i
zSDiSn7}m3@7hoy7H%<5C49ZV-PckRKC{I^UnpNrBdu1-KEJsw<&Tp+yy5?Ojm!(^h
z4dCFqVlgPn*VL0?wl8nF_X`6ou7g!-)f46$wg~G^HHH}2e+HSmOumE+s#~=wan0aE
zGS^RCEXC*RbXuLL*UsYCT{3tI1)pwpFliP43fHcW@U+Oru(itcUZLM29&+r8RdfL~
zYm>!i&-lN$d#9UP?K@y4D0;=qCK&7kzXf_n-X3R7hg%bT#a+RpBL|W9{F%XYOg}jU
zYe{5@5P$#Z=u4L!M-?ijFuUACp%i1c@qc(dj;S{ci$@T?8`{7AexYh<Un;~CWq^p!
z?G605k&r_5-CU{SvgVOdW=CV%$e5!LQ$MqvpC=UfeCEaD?3RhS)2CCLeq4?QNC^CD
zcsqJMTejgzdkrnOET@CNV<+Vwl4O}w<z|P!xEIURx_`q%7y|iqbMtp%|NXxC;CN~|
zX5`sWNdM!{M}toF$3fsW1IwMF9JO1gPJSJukDDg}rVu^T%yt106Vpm2T7iGNpa0$2
zR5KFOe&df&z$wMHQ+#iaHj(q~$m!|&o4&io-h2#Bm?<M0v?Dp53SP1!`J@=FSP{u>
zqM5WQJg(^+cIX>6)7Vy5(c=D8^Zn}7y<p{s%WcL;drbWe$U$!yM!Y5yvxN${g}E%x
zAejcy^^Z|<`@6gO{nz(m_3oSXiPJtz!`M}#dW9JUtXZtU_r1-R=k?j@<*pbZ227#6
znt*`7)cDx|{g1n^RnG)NrthAYtL3xfzFuAbm~ieuuvzK8Zhib70X`jk%SW<ZEspup
z-byLQ?4EgZL7kPiw@ic6TRg~`9s|8xA|?d|LsNg27l^m-y{wk8S=w(DI4AQFM;==M
zgt3U}3hWr91Hny=K9P{?4(It!MK=)KJRnoPFr6FKH3fZ5G8)3O#N>DhI^xY_@VtPj
zm*@S~?)LFP8zs3$mc!RpT6mvFEY8<f1O_L~MsDdF>SyaNG)Sh0K9ohKd}^GCbZYRr
zxe%}T;P3}3HUzzridZk@?O*8NgjOSGnxj@&(bXo)ws$%dwizQ7aR!Jyn$#b)(8ie)
z^{2#6dE*cuG#qBu%z5wU{SvC>nYIK~<v}GmZb4vNYtaL-iT>hALeGVv#{2AUi7LsF
zvRo%YyAT;DGd2vNMry3ReW-=v6G=&2vP({hIcdyJkFeL)u-Vp1t518W5MpE_gD6Hi
z+{K{33zhpH(#JtgAr6A}5j7oZel|sp7==G>Dun=lqwiFNv(MoCRI%kGow>3POb0QR
ztFkwRQ7Tp=l`tHGW;U1b1DA(vw)J(MYNg<mwWD`V!R*pDc@~FI$<jerJr{>SE%?(S
zCaX4PS&mhGuSr#?BcK{w&mREU!{n-7o}>pECoYYXW-ER#8-p-kGlCWze7V35Hhibx
zh(TXuPcovLU9=(xU8gwXR-d?;VC9Go4Y2%rLsv&3oP5jqg>d`|YR<;`_=cyZCSjrC
zZCgdJThX+YL>r^`dy<`+8R>5FnY?~bIzNNGj#NkOC9yd45C3R_kI>g-Ygs(xeK(o6
z4jq)hHRQ(3VSYD3er9T2KZ%QIS!l=*)2yPwcT4xLOdUKU=Ov+RfN(>IKX{xAh9|94
zxx<Lm`VSu!6^zYN8copFm2bRcAnvcFda$GoU{J*Z#y_S&oR!uVC#@CUDOtH^>g0E^
z&#H0{ojqvdW->tyqU>#DRj9M`#Z%c4tGd%Iy~}FV1*TH;PrUH4AQcsAw>c%7LYx30
zew1pOCoHb!bjeB|ff8*p;#tM`?fR8>5O{2i?IA(oC2oec1%{FIUaJ>vMf&{5&OY@_
z>f)oY>fltmDulj_K4rE!4xIFaODLs3U~PWN2qx9;L|x1mJcr!8x9PGXrPh^_doT@c
z414iDo66~$#n9gL!Pln8By9p^D!)Y@xus>`t>_yeirsZfwZSPgNa*{NSL}S(eERYB
z!r5DtBzw?OpD7uX4mduPmG)z2XiJ{y8kEf2L^WaQQ@_HIHjqr28kVMs=ArHTG>xq{
zRX8D1QuVW|RB1cq<XD{I^l=9>k}pfBSRIp1kYo%U09tl%rj2~8=^Ay+m;~iNcoiVr
zoUjPCO5;7;TdIrod+fNl6$tRl5#vom57u_MITfg5uV+2h7@Q#|8j|<fb^1N}+id8h
zMOWJk^>5;SIioFVaUJyx7&2RlN%REw!SW;9BEbjZF>u~^7Gu6g^c8xST~6L2QK$%+
zy?~iOd|<4P8UK<qaS8+wwm`MuS<u+4G#DS+KXHM*f_hI3GaTz}8gSyEAUX#Ji%dW-
z_hbZ!lQq7!P{3$Vu}mReMW9N$vH-_Cs>y%cz-XfiLwHG&_V+l+WAAsOO4dd6Ez_Dr
zSux!f=0cBDqstg@JLr*1k^Hius<dki<Ir@{NTUCDWQ1_0EVJPSA+`cAM!LoQSZjE)
z^b%j~9MeOPH{fh{rDh~9c#%l29Hai*pZNeDoR$?uLj=FNkeC`J8y|vJNQv35D%IqM
z?-WXO11UnO)o$QW5e<IH^HB-i>RF2Aj03j_HL3FuyriR|{SQAW#_w2B6QuN@b%v{a
zEv=%FxCS%lR_FocWg&0&pi9bxCbOYg<-%xuH|Ajoh>BMhf%djV#sd4j6Ja6pZ2jyf
ztG&Ez$<oCtp}!RQx@<P}t2nJwK<w6by4Vd!p|*hp`<nyKH3z)d&NxE*2dfML$YO_3
z<@SLJe`Xt^*o`slUL?t$GUS}kA%yUg_sGj&8gNo#^S%;f3><{02V)v>^TM*ZLS*$s
zz-2J?FeHa<yq_|Z%}US}1Jp@@84vWKnt0406#YxA^CS)?O9C6<{8=|!T}mA#1Gd<q
z=6%haR{8)sirS9<N@EZi9&4Vv($qUM2fX53*w~&rn1+&(i>@(;#KCzpmV3e+RyA|D
z2j2~`nviGu+FOr`r+lp&d-HtUR~?DhS9j-cYi9z32O_g%KMOH+d~?{l_2F=;vYR0k
zC^g1&N!A}9{QT?UHgxcw2G1kfy%_alJWu@h>(#wP*12qyFnMv;IM;W9h11xRgejOh
zB}lhUW@H;YN-PyO*8a4Ni;l~^P9m7+95q`4ZiiR%K4mCtoHRk5qG_s11GyfeB#ysC
zeJ*YYZgty;yqK^O^iI7*u;OpGQv@DhVvnKQMq|-J@da-vp`M>T2v2kjqjla}SztID
zziE8WbDa6uN2*HFPZQgRtb6gk@$TY{y3o9-CNuu#-9;#<GXAgn&RfF!KQ(_LqDd9U
zG?hhieMm$|JqDN?aP<RWzs~+cRt;HfN_XZ}M2LAOmm=U4)?BVQE9L`BZ4hztKaSYg
z{y(kz|8n%OuyOu(KnhTM!U2y5`Jbcb4N9~*dV~}P`v?69D|0)HAKtVz^FQc$;+v=}
zTi3H0YodvijHjjY=uPXfig0}e*skEo)HQT|wP#l0I;rcl))R0kjN9@7!H{Oh2UWNx
z{QTDQey(h5=8brlxH39^@V7l<;YtYSW&w;_sNwrBG+|_~7rlS1d@*zCd`DP!NJjn)
zA8^()Rcd44;yd8ew4CiBG9{8>>p~i)GekFim7Kd~mtw$CDIn(|z9}Spi;}vgMv~-_
zO2HijMqh}rsO7zrLqYYSAI3Frs!!2-T$Mw$FN&kM7_3x)W4AzI3vb5!UU%AwN30CA
z7h3>V2IA^+35yA^d4GP@r{9W+YtG|lr^R#J8Q6RM2Kl3UEgVr9qVQ_054Z{}xU5-6
zvHu74`WI)_=Dk2_YoGH`7Ppk1<wWV%b5?1PPb~=v^}5RgGBS*J2I#KkW55NUXTvO)
zIR@#_8VKeO;t)@q3018}<C7NN)C_Z_k}T!otF{cHCiWQsA%be~Ng&~*ZhB=>CN5F^
zzsQ!_@i8zkbgmoQv-!8RQfhdrKbue@2WTMkjs44eX0)gwk*Dd=i*fI!J0m7&pX@Bx
z>tNFzrx(>PHY$uZB*NoI8d)Loo(c->q!T!gD6fvsw8mct)3U`AAjUGDvO;a-NEY1c
ztl!Zib6N^m)60`B#%xD14K=0NQ(+In?M9m{X?U!Ds_EIS7nwR+D<aqI*?PA$)rvHw
zf<!k+b3M*CnfPfK-cxGhi1me5Xfa2hC}f+EP(x{sokdI3&!<98Kx8k(z97q!FqMuN
zimK|=JkO_EEYop>YYl=7#(Dcls%A?B63Iu6a-F_1%afq$_yk0jA<L2)xvHw~1=G9a
zh4`IE`(@wS<%oq%lCTqB7ZOogLF|XTL;r%QiN*WjYSPy{>YJj(z35t_?<_{$FZNf4
zZC@-*6jKTMK@M6X=21{|zQ(6i2^)s*!Ha@Ov*)B{`f6i+EIM&{x1|=kGNV>zx2(#j
zxs_9mWSg=mffIS|rM9GkgOG7c%K{fwH!wXOo7|kbTSSUb;h&_OR2(2jG{BsA=kU_4
zECiTXeAo8NIV@B8gg!duVbUN!;hwqP{L26*7rp4FD$<OWsyz4R@brhlz@=rS7lIwL
zzM=id`LAidV<)s08SVTV-k10$-0{nlb#*dlq_BLBfc6BH7CNV~et9Kzsnm`QZ38}<
zx2`JFoDzGF2~zuc)Kqj|$=zJbEBdSqD4u8;noT!G!5#Z-rOwnqnmf6u0*QIZl%0T6
zJ-r!Hvo;bL$`ldf8EvYnc$a9vb)SRd{3LxQRXSZ`dBSy~J+up2W|apISAhw78a?mS
zVw(&SlwPQo)QuCVV%<dX$3Id=YQ8I1AyKBW!5ubWooUmWSW6-~wj52OeD7_>dNG)e
z)<rwh@iLn{3Q}w9(W8kh5v@~A9_r|VgDlLnwAFUTnTszN2NRf6YoftASPE)eG&*L7
zOh!)xzM&X%*yMTDKU(o~HkxMKFX?AVgYU45*9lXGGX0tJ@CJ(WTkj1waTI#xCz%92
zL)nPHjVY`6I=#ShBUZLWKU-n(lUAyYmyff8FCKcP%`AQqK+rU3+A|sTkwQv0v0-H>
zBXwC@gEd>PU3t?Z-KS<7)0|j?*$;Lg_tBJbUQ9F=fA^dnlN|oG+}cj>&?hc0#-eJ#
zZ;U!i05=$wWk&zTS%Ug;58!DEPaV|>Bv<<W-tlwdh|x4FKE%hjAbY&6ad5aFk&`9(
zQjL1uxbsGCr-dbv8qJw;1SF}vJP4UB8a@BVTbiBie-BvxpIe&qzYJJD{9h4T&;P<B
zri~l5^nxV8t_Yt&Iw)=!#%ad?vj0<45UJ=<DBChAFi(j5BPK>RUh~b?Ki4O&ruIGp
zUfd*r+I!G7XWM-q{Yd2B;1i({yl}+uqR(Ho{}*HL5F|>VE$z15yKURHZQHhO?Y3>(
zwr$(mZQFi*{&*PN6EE(tW;LjYm9=u^mmg2$#f;wBZwrcq_a<*{qi&qt`0P1=q<Z$A
z^_VeV+Ab7MaSN13ljq){U4aMbD;gMwI<NaDd2+qoKhAL)E)-pmy&+m3tDC>rCLtE#
zeEb70BYE+Qj@@3I0bp3-s3#osy?N)|pVJ;(U)$n8+lPO*iK7WH+a)jr4%Iu*&~ZuF
z>U$EBk}|dn*8!$O5O<H-Jw`+?sIXo*bl5#yBXGM+8SE_7b>-+pL&#hV)Q1wDPP*c{
z#XX$vEvVfyUuEaG)hwz~F>@?cRW7(cSR}|hgTq3$$)yM+og$a2v+YNC?riUKs!dao
zKh{sFP3<AM3jlV_?Bin3plZlwB%`0x++bBZIT;s_ZmUw9VW<R6#bWjG0y`)_Z6dc$
zJB^iOtHPftiIJt7a$}`am~X(7@ByT$C`M$e*Lu<whE2OMuj09TEgQk6hnW?eHIb~2
z8)SjDhPKvXvY@6%bNXJ5>UXHUnunbCGhrhGJrk*w@sWR9SRGm}nHbQ;0!!erMw%a|
z65Je)U6y12{hEcg7>Z0b_2+yf&2<wp$UVePQsZJIE)h`FY19gNLkfEWw8kPc6g9@v
zsX_zSUS${Kr*wKL(ZW>nnXYKsr(Ir{#_Ya^@`BKd^$(<WCwr@!ql1a&^YMp4WZRVJ
zLi}W5ECOW3jyAV9Bv^rxCDl{(XffTS<qfCmQ&~3}cM=;H3*ZX9Dc$CwTAn(X(-ocP
zVdF0GA65-kar7-4;%v^kt=V@RM`OTb!QG86e_I3{_LvnNJ^ecFIeJ@KMf%GMO1y4`
z&M0h>Gut1mu?FLg!OCv~>0!hq+q@8@;c#;RMZ*05WR#qpB}pTAzb35(jx@<xKXZL3
zLQxn^N>rOMBW}i|I?V_w3R<$bs;;#<hX=s^9nHH!egNbLA8=3?W7PO^2MwK^9P5{g
zVE~qje0md(q_F{99(Q_<_;Z|!H&0cF@GQw=+-(rcq6?4$jebtDC6<!v0zyt5l`3+u
z0=Of($6!tWC4nOM*D)vsq-=O}wA__x?)u3F!#5-)Rkhfd%5R#ZOK{IocY|{?*EtBs
z3ewZeXbv+d70J=4sP3{+3@h+C>%p><DN5n5=`yuTMxs^*3Ulj-uW9-0C9Nvdsf+a7
zFHoD#LV)jD?df(uo8YdF+*EB2cyrodW+bQ)f8`+n@DTK;C*}c<;Z(U_sQFBRF;#e{
z|K#{n-cbhpwdUja7T{r#erl6rKUBVdL~|6t4GCik2lyF&0Or~aUT`!p2eUBc!8jYK
zA+(Ji%ODuhv=)Y6V`ah4wbsyu&<psyq13oy0s1j+KK9Xb1~^qFlBLkWMC*ly9__I!
zvC-$4JHGj^KFHQ}-H_%pYo-|!+F;0<>Nv}Bf?~;SU=|3~RB`TZ>A3E3!s@kLMM8+t
z71*VTd<E7e$xXLT4YokZ{M`Gks3>9{>;f?&7Y3vz(cecgM#SM0+Q1MnH+N*wk5C8j
z)dc)UHz2qr$vCY+9B45H$7oCy1n8b8nGz2X%p`T*g>(o=qjxqWy8{iFnK7f#1!yX(
zxRySjrENB`OA>GgpvMbHo!Q_~j5x-KlXulSaE>^`sB#^uCipxElLoSaW{k$9$82CN
zw>eB01_JYDb6%du#%B9ps#<#0Z{E0BTfw&Dg)-l86hgSUKq%90g3EyA%l&PI{5#9r
zD(GA4ZN(h0DZ_6168xchyA$5MHN$R_vq^{kpX!A)D8l5N-krly(l+GRYAi3<N8!##
zY&Q~m{xN$17jMy}kC6Q?0qh7=?4y7kI5Il{+D^M|?t)~4`3djq6M(y@AW!?u9d+E@
zU2GXodx!}RHe}sxui2}9vwU98z%8Iy`wL$FYqScS<EQ8boZ$V5cU|rgTP2z~FFk)~
zKU${Qf;Y*fbLfdrR@JV5=+HZ9L~Z+9r)vLlY@+QW<eZJdr?2lP`<ZS2tHo7~pMd?~
z;eq#PV~$vP_ewPE;B1=hP2^weZC8a=<HVTpHox0NV<MB2NB}UK3UF1g9pI#&h6%S2
z6hB~=nVz5jI<Z*k|L4SF;b8t>ZT)|(`~Oq=_<tr=ToonlUxL2v16=%881LVp86(Fu
z!NX4|GDQL*#nZ7aHxsx<%TV#TO8pLz3CAbb*xrNbN6VWH9J~6SxCT#`4-Rr~+^3R+
zueo`hBqXq<`v>di8+9?Ox6PX%1>kGLXXgP|v|Ju`>o4`w(z~H7=cL6mbrV3QyJ@z2
zQRg>7*4i{PLBA722etaTx_N_@F8v?or&ZgNUw4nq_a?F!GX$eAz8MHQ)905Bl5QTM
zK@}c1M4<2M_iq&TL5$B5Hc<&CkvXKyUjj?u5X`9S@2zD(_*HUIv2Y0ZI-s#w_>Ci0
zHhFivO(Qe9Bar(<^9~I3^;^GhH<_ltzLOk8cN>?lBoBtPBx_u54(MF?O;N{-r>AHw
zcTD;YK@34aKfob5qRc$Z^^jF+2#3`gg>QfnL9j7fP|2PbK&Qw35mVsR5#}0xH~)=i
z>EKlV00f3C6P0g?bs|HkAjH4mW9d^-6SRF2j?JTMdw@QOrW%lc%{C@Q_;7*q#NuxY
z6@8lqRtab|!|W~FGP6V}u9lIThy@TIQkIT9OB?347hZ$4L7HAPb5DV$N&U0p!gdTK
z54<MoGc8Tj=31yp$mre^d<B)$NUqI8Tth+1cEJN9!R;(iWaV}`CPN}lT9d<LG$((a
z66n-h`=tQ$V%O@%*n+ud<<e}Ux7j9IOs8TcTSmfSCQ*M2O?z0EH_{1AGPrUqLaEi0
z_{KjeTx`P~*BKdz9hhMTxJ~cPqwO)6$olV$^wFfHiF1hGJKv9`#siAu#y@AVttKCn
z6!{oYi<9(5iKz$-bLi#CbNs0#Fi=JFbOV~yT^|PexrHR#EJ`S#I7xJxMr5HQk4+uQ
zkeZ;VBnzwdnp87s=SPJq;LxehoKG$$cR-Fz16Qp{_8y8%13D9_tm?Ui5+sln&vX+k
zXu1Xt5~_N03Fgsh@^uW(Q9rEy<`OReeW?6f?dBjKK$17t=+soooH?i$j$y)cA#PgW
z*xSZ#oTF@^b(o?lRE0%VT4alsPOczU((L-hBJbE`_Rv?(LzXg-B3(7W)q~EcCP#G9
z9WZ;;=vTt+ZLOhHEk&D+5?6#TUqEJQJQrgYvP)`?nM)C;{4x*yYm?P|&(4h^7|{Hd
zDk$U|_Ah;O1BgdU!`4a6hjP{fuxHJANa6NJsAJA1^_b63)n+n%R0d4#8g$CxMUxDF
z7Tr@*k{7~gD$)yq)}p7_XI0AEc?9q;{8*IMxzK@{jp~i;#^I}-u^v2Ym$i@-w3V!z
zwYfCEn~Cz#Vzw$jfvJG(U-Rzab|9Tvh<Wjr>Z{^b&)A8nrn(qnni;BR2Zi*iZrmp+
zs~WQP&Xw9!8<X1Q%xygJFCl4MBQgWbo7Lwn%fTzF6fC9m1WHTWk7+O9N))u?rP?bD
z$_Xzpa(Ay0ty0pDElqOeMoO>I&@RiQD@Mk56}hx3%~_U3ktWlzT`n(9Ue=|uc(@hB
z7J;;dNwLk%cgw)aC_RdmkF(`;ra2Imbj=;9p*Vow9Qo-@j`z>%2`4)dY`hZx)C@$<
zrY)2-ZKC0pIiiA!uVN`RjO<KVU7_40Otsqz%{XMG@ldrFgh|8o|5oBc0pTk1+pXZV
zqwqr&FQIHfN#5!1m2}eGJ`tRyu`Fk|H@h6_ZdK4g-rnNdF3U0$PrY4jE}py;zRw?S
zaI_56sAYSSi1PlDC}$c_Af!6*>IH5v8&<U$tX=qDDigmviB;L8Q1)tHI{5aiY8oJ}
z#RxxViUO1O2!eJooWutB7O%vRe9cyHAoZLG6dy_>+~9le$=T^}yV83umB!u<Z@T%L
zvXx#4{X}Bkq<O+VQS1O@o<JW>@iLDASnx7A3G5%aX1Ky~K`^(uT%5IF2eg|1?(yuV
zJ|85;A3bbDbaBA$4C5NTg2RQQxle)JZRqMdKS>1apBuw|0qq;=HLf{pX1ac5fAW{=
zN4j7yuGdstagsA<IIU&OZ9j}B*Pl<q*<0qPallW<5*u<R@;)yWFdfGb9!t04(HL?z
zUXvSkAxL&oF~gL?K__QQqNn%mq%uXGc}X-}9r@e`MM0!7Kj&1+at`-J9QwI?5Xb%3
zLB#ey7hH@C^i2P&LDa1&WrxiM)AOM=#lWi`Pp$~pr;fif1n6n?Pis>uk~H;r4gmE1
zL5fFgf!lR;=gOd~6EnPs{Qf{FWwBL4Huok+_j-rVDyz!Ah+3LXk8GHZ_(vrH99Bk^
z!k{Tbw@0p*TM{xPm(2G!aqw3ExqU~5gWKfDEtLLQHW(=K#6BBrlwL?t4Wy@fBA3j;
zp7VP}=D!n<2gMB6+x_WG)_AJuj?@zpd$(Ge+Sz+^F3Jzh=*##Q2%wa&XEzw_Kk#WX
zLWAxM-%k#o?hlV5O?TK_1smC41DQcMd!rA6kPV|SxTU2m+Kn!pEo}C4oKk29`K|=w
z^(~Gyx=Php1%V3C1IB~bN0Rb*RgqT&R9SZHqk$O%$bSMw_K*ohQhUIyXb3|`F`u7Q
z;pg*}6_!YkcUJp9Q0n-|cl4ZV{&Og^pa2K|9TciH(VezwEHjg_lPB0hMDufB6<v7d
zkqzO*Cd`q!lasWOgHkE00gAYGbv*bj|MnNbsxl#{77hwye88kIS0d@huf|j>Q4sv#
z0*6wCVA`8G)6M*tvlJDehWdv}0hiqS=UsdX&%LLz82$SG4T%?5H%ZA#5OFtP3K|9h
zbO^LW=nmy7EtA>A-wG@YCNIN#57(vBs=zB$^*P13DsxH<gy^7A>mIpvZi<Si-Lg?s
zpa$F(cuS;Vrq_6}{HHD#V{e(INCpM4vRAIIXvKDBqKs{B%;?(h0$>nH1;;66?R@Q5
z{FJvgC>Ix9?xYl@1xcMR?PKF|!jh(@#0RRcpiEi4ve5LNnk#2T5AQNkZPhDa>-?1~
zqaB`VuKIaEkwZZ1K>{%7RZBk0u@ff+rsOf$1@zJFo%;q*^+EZP%1G5U7>0G%4leGp
zQKWY%syd!8M$B2A+0LO~4x%X<126<;p-~bu*T#UWKalyEg5z7#%XtI0=Ql2JpC|G7
zj<bY&oJfz$?m|q^fjHZk81$_a=Q<~vBhXcrp21Fhjt$QQtER?+%l%LgH2;hMg!(7D
z45z`N;p60qOmwI|oGY>g&o!}v77A)=kPI-7@QotyRE-rHH4!R!pg>v#=CfrdSS@=k
z{fVs>TGXJi5dygVj%nI%CB99k(_*7?pKvhl@Pz4*!H~q2yH`fk@Bgc$c&46Dqub>$
za{TH!M<R1{p#tk5LxPj^0}zJ+gVivwLklsi>bjK$JR$8M%!m3&L4a=NYYZSuf!Sdb
z2fovCxAbh(#=4URH`$n0;p&jk==fEafejpG01}QF#3aXl$u$(k0Vsacn=cC#7(T3q
z`{oG_%NE+$Y(J|xM8o6>KW22;v4w$0Rcelt%aJM}IIZHajGqy6;IJq!w^|e21_5@H
zQuW<{Axtx34v#bW2tbgU`_u1}k?<|e$V`M!r0PR{7XgD1q@X9aT+<-1xf1sx3#+7u
zD^EK8Vu~>VjxVl-QU&+u)MljIAQIoQOWdxvwSj{G%I5rOOvwyJ!nlWtIO#(ztG_gb
zU}xfVQ)!K!L^jZD8~{KtF|<}K7wcJ;jW3<rv_a|E!5<0`J`{CM0B+XuT8)2S?$6i;
zK?}>&T>Og?xm@9aE%=0&tjkf=<?HptmbNs}aujP0F~VvRxXNjK6U!V*vicAaE_Z1*
zkFh7kVEP!`_J_`cV`Y#W3<J=F2k3ZBD+7xxgF>}yLBeydVKR;(s}5&(&Jg(U6Mq_Z
zD1m|)Gx)x90s5NJ7T21_7d@p+#Oie@w>dhx{Dz+Mc%l$>I{k;vABke)l2mU7CqHo8
zLz(4eidAe-PgyB-YO=S=+Us4hH(OQIR$^5()iuX*BF9!u!1bf-5j$F@wT)|=Z4ety
z5IIDhHj9T%z*g4g4umjhs^gixEMAglu+bWgrKrrMzER{RAsnFs0da^dqEqr+%Vdmt
zBdws~XSHIHLlhkod-uo#Ewz1U?Jm&gw*p!zKCOmI6xZy;E;rn*_FudQl1r}0%WbER
zB~qyC_xzqIz+lEocUgf$Md#8CL^NW+gY7r}!D^1+K3xQ|ksp`*8=3!znlfi{EJO*j
zg@07d*YZaznnLfysATtOb3IjEfd=m3>PGn|wcV{ooOy7Sz!>AU_-<h#Lv7jcTxI~h
ztLrd`Z0JTevV4;ommVX%<J_B6<L?=8cf5~MuG({^o&}>}u)GDyUAyZrZ&Io>zY8ww
zm71+RU_lELg3@|MPmLTvYKkpFQ2fn0j=M_rj2cL%7I#8^{5bZuah&dS=r6^Ro3;+K
zMW~SwE^xRK-Q-2yS1>r;dp%WdL0$Wk*`Mx_c|%WCVl#7DY;#A`as{g{vAs|wUQO)p
zAX6;A*z$Oh)ni%UK{_>@63{3md&Q$uz|NMVoJHUSaB!JZKfr$q#Wvn3@XT$F@^W4K
z0e-am<(s=Yf6i%TxZ0reLPhoIo0{I31@}DFbA+-|Q*)~+9@3KoJMP1(z2|vu^5K>$
zV-AM-NQBi~33Qki`-fHu&@sPSd-g21;5>R0;u#H{+DkTH2{BCZWT>fR_@W&Hb2=O<
z=8#sb4*Qir8Q8H|4%PFsr8k>u=R`IM7C#QIf{gy8Lk89*rf}M2IUMJR+pBlC-K|Z(
zN`CLy2=#M&Rc8>mGC1KYBJmma(knl7mv*Do&fa==CCYl~>w8$Z&x!?mP*iEj-9Vw)
zC&wyVYE>@SG|3UK?UU=T4V`&p;M|=f^y$CuZe~2Ip_&dn&(GEKYs;Ddp<o8`SD5C0
z3n5+)gb3oMB7YB=1nWW{%WsT43Wt!6z{!js0P1u(*Z-<QIR57XmywB`^?#*8hBc*P
zvDy)OW@}IR@H2+l{Qco$yD;(D=#Ehj;nC}z>d-nfg}!|&#8LIOX)LUx-TvZ+BaeI5
zsq3lKAKjb8Pd*GTP7Zc9vPPoy2dGU>o`BwhhyMr|mNrI$!^;ld-`l*67A8&XJp{$Z
zD_VJT`@+3i#Dc>Div6X6!-m`GGs3Er<rdxKmJ-R);mwzaU|#^Iv*<?RedT&R;M4p5
z#)rO0mMac+ad6myI1G6s?t!_t5X^YG#@1Jo1SSmfXb@1PZZM5=a2mNG@M~==Q+Zu$
zIE^_Mg#SyKsU{*&#f2NCw#4|9X_d5KB$fH@x2GV?_H|G2Al9p?=V#&o`*MyVk}%N0
z=#^%nVv|lt6D?x-tlS=D0;<1Z;<Et0^wzinCF8-1OC866zN;$tV6<psQ)`wk*nS_4
z1sy|Pu%wD-0dE@_Fv`jUxv5@6qoon8BnYO`dg@~kHFj2|sAav*u2{|)Pf5F(#5ADZ
ztC`sv+NS7*wL!J#ZqJyFCjr;&PE--0W45<BQl`ysJ2eGcu4=WIDLeLCqEUDDntx<D
z2}87KireqTCI1I9CN4Yt=uV4N)pvdossl@fIq{{XrfR|V_S5ya{R9jLje<6V?P4;f
zq|HJ>6-xdfpG=43tE{xy<=*oz7fd~*iG@>`W3RxH>WLvhcfFu(D`@&*)pf}b@e+xU
zDz1p4EusXlHftMxFdLPK`Z~8fW7Uw(Mpj>4-wwqOuIJ)-HQe%6L_9Xrp>d3_hsIS1
z>?_69{n_Rq9<<@;Lju5|02EH3&Ub5(?cpNfs3JSS%`KDVr`Gq7|Bdg@{hM9xHz-i)
zS8B+_FoX#~$dQ(<rt^LO-eYa?RnR=?DX4JL2F9U)IgUMmToq1k3V7hHWs~A|UF}k3
zFkD%!Eu;EgfI<htX-8UU6X6SEcdZYZ)3P{mIwjUpfy#uo<-T{0h>YhuE8MWgj|MI;
z9UNYyYPC{M2t0}=1?T7H0_6f&?B8{Zlul=O#eA`8!G8duogM2O;#%*>P**S(lHusx
zu~H-|pvO6~=GOzo?x5HEv+aIjkcoL~N;FGQk<wcE_z50@^VTc2AKzzM>Cd2KJZ`xj
z>A<Jt0BW4<C_!Q#_SMKMdo#BFN*FVR9NNGCE8d{UB}0H}ejOzuafhSDMik+$<_$hw
zDmZNvD;ro#3}XH=u&kN{I?DsHK_fu=agqH;*vx<Mkt57}EN$-PC`HtH4i9q}T!mA=
z+fevS`4t3sp6l_d4LZmx-mR3lkJ<tCanr>d)|A>6VH32vmIH_qUu&DjFFFR!xF3er
z7Nnwl*@E}Y9{`1*C>AJIS=*<gcl2`I7(~V3HP=1J5J$4nY3$9EL(F|dm?GX}x=PFD
z+A5lO{nnUZivm*2O+yqx$c<w6=2mF=IWyJjMOtXnkIXIPp_tj{&H^)|wy?L1F9T7d
z(wthE@-+ctvj)og9S`{pg2#c2R~FZ0p_B%C;_nvJWZDLvLBSXQJcRgJ%n*L0?k@Ok
zb>p-2N4K07KrR(3uhX;ny*0lOl324lI|w-vamDkBVa!CpUx;e0wnI&-GHT#;ocuL(
zFKunGjf6kR@KU|x_AXVJoH4Jv(b4rV2?kdV-XXVoe@cR7#)z{E9e<|Ky37Jg2TrQ`
z(suODgxhXk)15J(8;OJ<Ov{b(?&eb?TK<NhKy4{aL99E|mlnV*jy)+4K7Jgf#2u=G
z)IqeNdMYi02m}hA2TVJf^`pzh<DH32!O7&y)DS<Jg8^@}H%}*MM?9VR$t%$iYlAEE
z_1H+bh?uMx(AX5)npkRj!Og*Bj_8me)+o5BT5Q6OxTJ#Aunr5_X+1WQIT(<c(Zqmy
z)Hk(MfsH?|H&U1`2ExqOGW24zPSix_c0~GZ`QU%M7SG6h%)`;t)IRe!gekn##-w=7
z>rjLQ&d8?cv!Y$DsQAMnM8yMIEO`Nz#JKRh297nd$dN-Ur?cKUPe@Tvw4oDSMnKNh
z$eVssY15Kl|MsV<TT}nkd-=lu4D<Hiv0};-)W7sS05+i@eLqTUW2BS3wJf^9z`KDo
zdWLt^*E}6$($`!N_zkUipud|qLW`PUQB9E811LWcp-dDyE-8~<&&rR6<k|av<6xmV
zwB(8zL3#|(c&8ZFk(Vh6%;}%Tx-O>@s5nO{l1cs2m^`b(#p_$bXPiG?6Jl^NU?yrw
zhm{mfh?PhP`T8qy>)aj_f%1;Uvv`8OHbuAnT^RR;?VPP5N3}t|G{~pv<;tVg%Vxb5
zP#>dIpi9h@&f^Na3P<E;U=YoWSCk|w3Fd3R6=jzUcoGFY_ylKYY&E{M)vV?dzRioN
zIEylyCJ1)O**6xQvVk4P&{@ZGv?Z+GQ1poql&Z{I&e3RJ``^*H6$~5cK-PMc7E{vY
z_sef}R%)V{AaQb-W;;qiQzFdSd+~zZYZiA2(#eUycw(Yn1&Oo*jst$@7r-XTk#3Nv
zXB){u!(!LCj`z#FRc*&sh54_2_2RYVY*F3HO9TCILAuV}h|S+pLUh;@2$g%0Et6aU
zrzCfuF;l%N>?QOkO&gy@_EGpG0N!@&Eh~pOj-FA-a~73sLatPVIkm8qFGA0#aXS&0
zl-;Zh49i5Fj+V4No$mv{bzbL`v1U1gY`0I(Pd*$P4RQe<JhHY@R(YkiVZ;+%YQl)o
zwEg=vS+!w=5nC7JhR_}OoCsZ7D2f9(GPL@2+(;clL4dqaOEWbPCB1IAK#iU+S>!pL
zZt|Krvor0g&>Ne~TyD!ZR@Hec8;dsI8@oLbrEJ1YRRVsajI8+QqieP<`}0GilXRL}
z>LXy$+@tPtx7gXE@w;B-!v#paf9)~w5gH7ckpgk}3lOV%3W9EirObj<soI*-%thO{
zw1JY@z8*^^Q*N$Y#NyaJaY1x0t`L*!<ZDSLCRu972m0M9uVOE(l_@r&8<8|eEe2Fh
zFL0KxtpY|}Q#)#OQT>5+^s0q?n0nY(Lw8HG12VojIUX$rE+V-kq3W2JUTUuNR}Qn%
z!%o>#q&v~5GQHNR_R#&;`-=L3koX47DPU!yIuNHO3^6NEX=)iU=_Q}(X~a>kR3x}^
zDPpH;C<SZTfiVHmm+Gd9iSmfvq0vL&>;kx9TLC{5E<eL(&VI_6A3y(_Kx1NH`yYM<
zBNN;Ij!ZGEC1pq44*R<%T4Hdj6D58Ei4xty-Q>4zyrmf@K(8VTO~>!#Am-y+Y9v;1
zadYXs!pYOov5Y5tQy_jq=kUoreDrSgYk&I`wrm)%sAKQ$lI`OY|AtP$iq4__OAoR*
zIeS*?rdY5yo5Lq5N%*b)ef?VQ)(R5i^>-X^ZCex6%}LLj+((>@@-di`k0AD6Td9jG
zm@nMt-fsWZ_xF7&R=P;7BO#>Cr*Rb`wM=qiIkS_}p2eb4J@z?8tIwX<{_ER#e+6h0
zn(Dxq#=5jipY2GE!r6yt%Gb53e>;_(0-|kw{%}z1t(1*`?cI^g+oRVUiTuT!+7100
z$3pr5R}7dKi?9`lljc|j6J7EvS&zoXEpCpxcnakx164dApzy)iyqUZypH#}8-0OGR
z99miSpaj$~o>)f7?mAvdbuTyGQbrd7s=fb(f=$E79$PNY8G)D5Yt6&Uju@OpZ5HYR
zZIhZu6TSq$&j^ct#Z3cA1RCF!ZT|k$?8m{5_HnoX!Mdu+l+*DzU~DUK*wh%u2K-=@
z-52L+f+2qv!iV9_LdS3~(&$=Y{T#P!QT}It(EeqoOVQlB{gP>rD0%nXr7#G-;jSz5
ze|l46=_cn$e(%ITu1==ykUnM}ka5C%vbz=+{YvLx`s<?^QgVv`yl+1;XMeuG-%sp*
z0Qy0GAdwvvxzYMM&&+&D_WZe}`R+Rr%;BFv2Ba7xd=&o9Z{@?UgxD>@!Vx1Aux|SX
zsUoh_r(+q;4X_6@!_6_HU=WUOWQ=~yv#OS#!z-Lpin>E<GupVDybGe}4wm}R>7<|M
zE>2%KKQwPFBK;hc)-P10CxV;ohO877hvNh;VwMr}xY0))SQ?d@o{eo_4wC{P?`U0=
zaw;!3H8F3bkeaHnCzVjH&iV3x2+SMDW<3TyBT@8^z~Mr!7mZgvyx8$Et-FzW+UA5$
z(;U#S`^&b9N&v!p3|j^g=tL5<d!1#Uz--@$ueR?9S8(hk$4gDwoUHaGiM(>M5Q?hU
znd$V}vcLrvHE-zW)T>)e0fn42R+w_YYes^#yD>;nVBt#gM**s&ED(HQLW64o7f$)l
zJMXEs612<FSph!Cds+%I4%KNI544PhGm90=+#kt%z$~nLQ^n1^5{v(^Jas#6b!P+L
zFsx2>H#Y`cYU{seC%B=Z@h4Z_36_&IG|DtKZzkbYd+y2vm%L~#m{DJ0&*3tiXkYiI
zKJ8~V*{jNETF{<IuNcxj$%EmLOQcD1aLGTYTW^0@OEl;IGfm~!cHAp~EI)-c%2!3`
z5S0jDfP93O*(dKwS4yjgQS?ra+*EWw3(!4X)XG!#6v43JLdvaRr;;BVxqny8<2(mh
zN~PY1bYqLC#OOY}&Bs5YSe>i`Qv@NA^KZB>*XKqCaon}^2fBtysgSB`^xp@9XzoBR
zi}3`AD3rcyM9R}3Ytq?6MLoF><UioAC5dL#Mr#5|vlW<3Om^H>NKneI%Utn54&XWq
zzBK{y>Sn2iwgtX1E72L}sx-A!S8g%6GdWIM$z-b6ZarPGDBvbD$%8JHkjpL^?^(a0
z&qSOJn8b^72g6WxxM9``P=8^()}mirYV&jAJ}3yNB<uRmyMUHmF-*tfm#1L!Au<c+
zS7Dj7;hNVIUM9|I+*vb~h3u7k4VndRD3HkqR(xp8w7|gFSo&dW#}0(pgvURK1xVp@
z?~xB4H`F%vwNMik(>vxrc^PB;6SYDm%1%_Eotxz_z@d>3?)tt#Q<fBnn%L$5K%>^(
zdGabf%-6VI@p1E&{8o`zy^hHcE}gt~>%5KdKDRC5>b{Uw<?pLZeI#<-RtTMLcR1{b
z{%55?#w6(fBh)mu>`k6`YC$k4B2{eDc~ult9BkG#_$iNR)ZZo0YU5;@=!@Adyi$2S
zeetnej)*E~%KOaG?xIV!UEZBg7aguVIfuAXD81f{n!lQvXXefgXBSJy)=LvaHkhWr
z)e6lwL}NTT9aF5hc+)Fe%cdt*b5(98ERQ+(OR|Azab`G1zje_H2_}sqOOsGXX?dH-
zJtfnbq?_H2K_Q634yCE#U~6zyjqPsDUZ{)v08z@WwI-)1it^1mywoh9^J+H@uhRr;
zfR;)_M%aJWDJRZZx&KZeVk9d>#l}kqc1nZ`Of)5~GO4fvnrGk+7FvA+Lbr_$Dmz@i
zFE~utIUv7z7sU=y@-zvVDV;1oC}AVTHPM@Y<{4EQx$)}le@~T)!B+beZAT~u=uY~e
z1fLYBRbB*N#-%jBAdrQ^4ITi*usD#;dO=LAsO-faw`AEzTvW`Zty37KX1sS#4f<O+
z(<KR8hg=AvMx0go$NFM~F(8-CMi8sde5<3_ey{&4A2~7wjT;~k=ta(j4d+W%2?Hxc
zk)-r(pXJ1qvx%G=4Ogh-K%^)XDc#oFcFfznoB#3VxC?En4|xNuf57oho_^}VKCh2@
zp<rv`=I(x8si%TD$-09@V1o|k@c#3!;DXcF1#8(u%B9S|w#@<a2T}<I6BCq{Ydqut
z#`!S?zMl34d=n2JGqeU>T?_fP_yH%Qf98krVIJpse>D6ojuS1qmCkK>2d2Pn&diZE
z0$I0l&e}`nXlK-~y{3QLV0%{YKIq$g^W3<HSh$()O{SiJO)^s9U1(G%Ts200PSqG(
zbW{o-&-Z~Du#*iUU3>7HtC!KqDX61>vmN%%imcpGgHpW?AhJDHb@F0NWB7YLQ9Hj1
zloK6l^k9O{k-Pka7yxF5ri-jlfpwN-A<aH`2l@|dqajq!K)>@2;U*poCH&;?PpT-g
zMq6|AN})bKRRYuo{s~4HrSsA$f=67Kydy8&Oc)Xh>vZ?xpD;1c*vVhYcrdG>o?srp
zviRQ#o(Jo`Gw2vI+&4glpQJZE!?5AIVo&c`<>6~FSh<-!B$qZZXPNj2*#P_tHW{@^
z-3HgLXfN-mf^He12uw7CWY~hRik(P~({)++JbO|iH1Mj4w=c$~{p&{TcX0T9cFK=V
z>)erZ?ned&WdMT)++SK4pDJGV*UD~}9<32}GiX)W;Mvwetxi?K24zLZT}elTdH-aG
zB|8Ab;Dn0;tw;%6JF*Ylc#+Aq25#SS;D_Q&q=!ED4@ZpC=?~yups3Y<6K0Hz4F8`n
z`~S(7vj4x?Qr7>axp$=b|71(?=TaX9x`@p}4v_`^64xvOKv}NQ&aKP^64x3Q$E?Ju
z`hVPbg)AvvTy|)9sf~a^BAPvCq%vjAG7VxGLCu4r-TfRA7K_G9X<!b=G_gu%70H^U
zI{6j06?{dXGrLIMnOPVQ5klRMjhz;VZ!+%RI8H|o<w|ai0X4`IRR&7?fP;FxU>ZZ|
zAOoc<S{u8*u8$GZI^Z9q+$xvV|3=ss+6CT=6tK!;@P#)C*bC;^B4dO(DPkP2K1jt#
zSTPX&lSu6X3?ov+7}f?S`YotIm2J%a;z}qPO^!SK8%Y^<ND@^9EzeHnEUDqP9;CJb
zHyvF7Jvn7BEfMRQB57S^<fti**T%Lon?f-ykgY(fnU+V)I$T@X_f~A9i8~Z-ro&G)
zQlppwH!XKSO4Z2hG8DauNZEHBmw}Ee3NpP*LK44(W>J=p(<Vwwh0r-j4+v2EhoLK>
zlN}Jew#MFn5BlCz+LC0oo2V1txfnsE!L@iP8LSgcT@$j9tEx8560NM3KhmOM&*C<X
z?9eRA8o5`!(X7oVK0OEQt5!@4_^sP{=YR@Z&Om=AWNW?mVCfodZ-<^!gaOwgf9=`d
zuQJ8Ozi&}YDm`qq(LEy!#nALZsS7z2tv5W-_X`nvC3qoHxrJ@cEjJ9!E50!T8eOwq
zYZ9s&QL!O;3aq^9+s0TvG$V(QtXC@@)rh2uoK?&ym4s_L&TW?vMXkKG_sY#ET-tB}
z+!05AKb@p2Mf>sL%IxXs$e=4ZX*GlXaPjtd!ejGc>h)#m^u>rX9~ggZJ^MJ@Os)TM
zXXVoC5?*r`Andc?Nr38FL3+$AbZ1{4e{3GSo-MsOraN9VdI42=1<IP<0s@jG{d!P)
zHDkkQa9@l90dxXMMYgNKhbOG?#w8$Z3<WK#AY+n;<RvtTN1YW|HLlA-2_jM;C8R~z
zzmy9T>>;!?G_m8L^*nOXGNx!(H6}uVj1q*O_$H3KL~s~8!3B(t%Hm0d)hzS3S~7+N
zCJA5^H1K1ze^mVpfk*TSySV;)pk%B71!1dRT-3pbfw&JaU%h;gSEjaaq67451PhNe
zE<mN-5K<=)FfENRL`$1qatKNh2K1(?LhK=OH<aI(BZu4V7It<~|Mj--=nLxqeRt{g
zaTB>3IlP(KE4>(-8oAnP_U--s@VIz2YZt3K?VIk)uG97Xu{r#r7P<a3eD*f=^JvQH
z#l)966L^YYDNakx)IMg0)~r27oi**~>yjgT^a%tOQ(KaG@o?r`-`2z3sWWAJ`<)v@
zwCd~}Rb(n~Yp%9Ul|{$VhY&jez+a>b6Wa&JT<Ss=g(07h7$H<FFMnTjS@f+w9*MHc
z^&}=j3vqY5JllLN_hs*apHh6U0GE&D<9=jz51YMnhI<L9=WABs_qm|YbWk>1;`e>A
zE90@;hcxlBe3i*YhNr1o2QvoOW*liU=2nKNMH9_JkXg;>w%Gc>_Hc1xmk%!X9FUo_
zZ{B!8mHIyWo00W?@pkA)YhT0TPS>O9I}Cq<En|(IL=v(BE82v!_$P8Xs-`C+S9@|Y
zYZjThxHu|Tg4XC3nTXQr?)iQI!`O$a7MTe)(z!Nk%J#5%@vuIPlh$g+H>InS<ICi?
z+68XD?wN~pX&>zd8?11!#&dOC!(l<l?YX{*gw*RsCRG7b@kcThIar{%F!z!uhj9*e
z?jfuUTj7v|*IO78{ajKUkk$Qrdw8kl+zc8pFRb_-Lqdm;uj@Ts>lpi7EcHz2E6^fW
zSV3@dj`wKaXdJHAN3x7;MT{n2Sx~^>eXuOI>bHG3JSJeStpwZK7JzK%X$d$7(8oM4
zZa%s4o+r*WIoMn|1(8j)F&X8jHp51p6vLgU*d~RskZ#LXtpOGB>K;{UUaoLDFR&(F
z`hq3lpmbS0B^gtq5Ne?;0Q*_~gWSu02GBOtT-uy!xCFYNsAKiB)|{US4j>VI+6J-O
z11=jo-=o36@)2g641xJod=PNE$X6NW5?}DKI64LDoxLriDF#>8pBnu_>@2^Xu5`CA
z1@5|I5ebV=0wVWtl%wjSeo5}(K$^;T3V8MI)*Kg?26$BhTNx7>HUgZbt`O#u+(by*
z8w5KsAi{ljOUe|E0Ir<yi`&n|!@<MXGRjL~Vz}d@uAI;Z1`)BA6%R*ztRul&oUFI4
zo3oeCg@A{1#Oi~52w<Dlip(^YSb44kG!MaGR@e#>AElaPPfzf<p);f*RM;GJ=C3(g
z|9viO86PpwgyBerd>IhCvY^c|PJG%sylRLhs}!_%-^vNp#UyWaY)|95kHwELImAVv
zj;LaywxWQ)Av*c#cU3R4fyK`>VV^jnB8n(ZL;(f;mGLKXq8(eL+eqtrjxeYG!3n^@
zNrTvdSY8h1B)zXR0O`Glksrn2!aH7^!A*bd{$TeLQJYi*xb;^6Y8CR=vbfx1>w9i8
zN?|=hXOH+Mbu*B~K_^X$V#SW_aI1O-p|y<Gctw=yTfe~}?O45a8e)oE79y&SK#Yfe
z4>!FYZW<k&b~<>?6tG%~kF%tkhTvST@}u)4HGwOK<k<sicrpAW^Bn=fvkvK;z5nc2
zlPC;Bz|9^hRB_<}=5ruIK&u@`nYxi>aEE90b2YT%KuG3lnYmJ-BviVfxP+S$59C|g
zP3aJ8{p4F-$&i}e&NE~>Hcn}i79Eh!TgG*1!#Eg<b66=#-rSUzJ^chLpu~)5EkgZt
z5J)1<y)5KKzRYQq6-|UN4&}+%o7!NXpw_lsGHfAlM2_T+B?Y2<l}l!^_OQ~1X|a#E
z{Df+OCOo6hO2TU;12m&xRmQdNno5o<L+%=yfXV?`vcfjK_);}kxWv;JqvWLI&V)He
zU@~V2&C~Zfddm`nb$Q|Kih}zEd6x_F4kwh|^7bT6{zCjV9r!bdYt`SPIRI>&=!th8
zHvuAY27u<{D$slcx9%h(y^0#`35{Z&p--hbh~|xk4;^7qgM-d`xOaR}IS9UgS>8oY
zh-SG3Ba#gC;2@Gy8PB)X8sb-4jm7XO$X5aXgaj|UXB-WU_mT<FL2DD>=I<e0QucFf
z^FcaF&I{>bV%3qbgi(h0zqmup_|-pLT`7WmtjSMFd5J*~n~TlTsLJIh1l&GKs3^XB
z7l4UkH`p)XuVlTQr#|YrfCorJ=kA{4y$oMDDTO1<>BM<s_lQtH(tHNZrfDoqDW(w=
z0lq!lJB8kQd^~!xe3ZdzVnWT`RtgW69n645GJLI$^DPFafY+PmRHY1Z5Fc2jIO1FB
zDu2JN%a+)d1QtRDhzC`vE9j*c+eEmJoM=OhK#$W!KG@^pHlPP*$ZPT;fntIAJa>7<
zGVU_QGS0FiWl#%!`9Lvj?)`uBy>6Q*e<U9Uv|;|3e`uBg;2uInK7gt74?y}v0b$xl
zMO~N1>V1QU{PW5GZ?cq$>A#R{+5eaK&C^;Ec1FL3?uXi`Io_0MeFWA%dfjq0*l~)#
zox~r31XiFTI17A!U*CueqBqv@2Crj$DNn=SSk46z4O}g_$E)YVr{lv}kaOMdw($J+
z{tDnT@IQF6{CV)-T+YDW(Z}8_oFe65t(k9pQrKt1=lWfzZ<m-oH;}Y%|3Np2d)i2a
z5k?@V0)<}L1J@x6cI<ynoOk2(KDk~;dwV_|iI^^aFUxpCTJJVfQ!{_ImZHg#H(mo3
zh?5EicYAk(@F1G#SMZupR|fY+`QUbKznNz=S1@-9*Rbo>iWoHA43$L9ul|hW@moc%
zg|U@-G!7NyA@w5nx))~bve;>c&QeQOKcpG?&dd@=W?FA_;p>)szukGF=X}1qY_z%!
zaA0%z&+aXUH^`H#a(J>Rf;`3R{$ucqmE<x(f%{mZ^;vmlP#4^yO`KFR&TEjYBCLcH
zr4sm8F_%NT%Ur*_&ptnpQ>>pDCm2GJB%A0e=e{6EMu+B770v&6K-uYdKmO+zIuI38
z$x^9kUDf#Xn0n~Ty<Nc%36d^(`)TUo`rw_!-YLSXo6&8}8l<6(Sv=PuD(KOroxvUW
zd-2}1m+)6NFCd<<O?!*pNqSbPO+U7sjnQ*M>3i5@zWp}x+%@A(FT%Z2g8*FVsu@36
zMMEw<v<i8;9*?0)6q$EVNYVIndWDPt;{Dx5BcJpW>|S9#G<a^`(45*T8$-T1oa@J#
z@%*31?u5J<?jO+n^Jn*un@;J1REaM7iDzOz1FW7+&9DO!EQn|2+Vr4<8^Ul@%%!%t
z6fWp`6plruFFP-SXM{7;i@p-{HN{}F+)jCzvY=&Px1VX^Nz<@Xkh64*hJyx*v|w+p
zqx?gt^}~pdC9;!rc^39u)0mo6u~7ww5E&w($xJY>@aypg;-q%@mqUJa$kfvu?J{6<
zmA6Sw0yx2R*zD7p`j0%8MKQd;PIHhk7A2phcqqs^nN<f$D{w*f@v7vzoHnD3%C#rI
zrJ1IkhX_?{m>%QB6B*XXoiqqh%JSAhjgzW%EE$P)75%K25ulX}s|mC|WNg80mKTXE
zJ4C}gt>uF({q|)45|RG7$ne{ELONAa-tf$5R#(c>jgG<cMSG)md|FAPvTi2c8wjsw
zT3I;PgK3@Aa@BY%;wDy;R;!p%d=_=MOk&Q-LsfM<yH)|xv?^<>L@SSqVo_ai&;qw*
zU^dsAk5Vj$qEdMVXH)^bVn~zpRft)Mhu7)ZEbv#&N=nK#TsRv>;tqC<;+)ZWa%1+>
zXgR#<GE^sSo-ddvPSp}Zby8{Di^jPf6qM8t8g>r5<p5u7=FE=6UeTLv0W8(nvaG<X
zp*Za?9?Cx|7d1==;kcn>IQ>~5$aobi+Q?AJ?-)g9ifjX~jbW`L{#M<*gFK_%e@5_K
zp5bxh?4R*cd2Q0(`f7<pz(1Kt2TTrNa**Zn=o>^XT))W@9(7@~9K^-jz$#a4?|q%D
z)mw$FR%c^3FgCLocXRY*R*`LN!4G6Y_Vm_rqM0DQZ%k4!A*<uC;zxRMsnC;@L&*V*
zN>;ku09Hp3K@;bFDX^^khOp#j0Zp`d1eKxJsaL$=kk~L#k|R?>9x16XO0Uy&WI6AE
z$Jy`6*xe5MgE3fq))^?I4HQ=5s@@l}Lwa3QZIo+&^gzhWzX4TgDfWi?m%;#eP*`i)
z+V&AVA~%!{p`J53W5JoMEAqu|?TII*v4@S*unQoDY>*NUGM2nW_8KAtW%h3-M7w;l
zJI@ZY^rOJv#kz@5BH2RsX96-sV(fV?e%V23Gk#g%VY@#Z`484HNR-{Hdr~~l-V5B9
zq-ER{s!&vom!l^gJDUsv0Gr|#o!PrRJKKdlUk+}yTUWc^_0fCwSCB+sc&!N7iUal!
zJMo`uA$H7E2{<)GZ!Od4TzQSa<Gc~Wcdo=6fc5M?NSU{7Gpup`2pJBA{lzWCX~NXF
zi2*i7{3T31))2S!cGHG{G~2kpw=TX~b7K3o{CEd2QZ}ey#*lfHjdXz`_~`xKCBZPA
z=f_zm*e?`=FlK_p{j7cGyU2u)J%t?ARRY8zJ3#U1-5~Kp?W_cdR->+63(oKo8NI>c
zh`Cnq60@@$@Q^2i>GrhrFwR?l(9@>v!8G@)Z?lX<hrQu72SmD0E_tg`k_<$fh$#w{
zVsSrCWD)RFb=a_B+IvE+mzRlhm!8jWz_Q5oam~Y%t^Oj-APfah(qFXCg%7XwMlYor
zk&!3=Q?<%EbNiDAU-!uF7R}iXARyptZooQt11%^ZKakiBB&dKJ4D?U~KXFpxAYV=~
z8{VB!d;S5l9bsFM+bY0=gJja{p9_)-a??cp0!XV+1)JwKDwQNaP)by++hd1fD=hlY
zW-b&ewO}kN%4WM-K|AmE8Z6?apSL1?c;fz^lz|x&@NH*zhQ)>r9152eX#>TE1#Ih=
z+^gU>j@TBU)-b@yb+9Ze!6<V7@OrqBqVj^^?U*C<+%{j8A)Y*u`tbTV6z^V=4<oBR
zY}V(z8z}?TGgRD9uhpkeBZ!`~@9EHi81x>0YCvhIc<LUsL1<r8oK!&JeScw5_nDDM
zwyeB^r4#|Iu~k_Nn;^duwH2Zjz_3;xG!z6xw8zLIol~+wf13+D%!yqatin_t<Uuh}
zJnAJWgz0M5xMXSQO^bL6)o=8j+@V@IR$<}!kf@lPp~(C+y-9AQ$FkVGZUvd2cT=kD
zeYBp}4tL+aOEtPx9D%KFTXLqfrCF$UQ<0>~vz>5yT@;kVZ&T7Kxr$7SBI;fP=>61?
z2$s*|tiAiKXphc~zDLu6JGGRuJxdP>IR>ydvy5WGug+7|0R2inSjbW1th^Z(+t7D6
zH;Wq3=_Tuk^m?>n>)?NgT^$yVRNZ0aqAfNyw?vxT#zH0ZxA?Fn1cdZAZv?^CyF#pe
zkkXRRWvh%V6>KH&tmcO5jE2d!s-!H^L9!icWe^5AN1AN{!o@b;Wsmr-^4!mhg2}$E
zbd~2b0S#l|tqw7Rlu5kDqf6M5vM*gP5f!Xzq}gd`rN*IWO|zVCh=sZVD4{WiNe>3N
z;Q9CBxUlVz+1F0F?eBc^+Ga#zR_+!cH5$!CZgCqxnUUOG2gU__+y(zIv9_iD^WY8F
z{ULnOhsexaX3S64oWj*<G?o*bZ412h2)i&{RWa>|0h;G|80*M-GJ#B#7Vg4uTQY%7
z)Z(Tgp)Enx+299W3#>}xBnSH9z2sd(d)omeTD_k4s5kaU(xOkr=PGOOC1sAOiZGzB
z0-ZlUl^UPW(%w2(NUIv$Dt^`ylLbi@HJu*a_(TbLVDW-!#~u%z&6K?EFyYy6_e3N{
zx3(b$$(U@}&d1irA2RrmKqsybSsf(rS&7ZWOHQse`!8V63L7(UR^r1pUfL#{qChhW
zo(QY39+k2oFM-vTj^^(b%W79T7B5TzAu{;fmQd(72`_Jvc3i%alXHWP6*){?%AWq+
zI)9-D=@PLDGRft@;iK^M^J8v3EAAGS3`wSEi%1`->nLrlc{AMHlEKUk*r%<uhSED#
z&)&(k%va2&486%aISz=av=!;mvi+Z+=oq5X8p8J7_v|yt!EAeu2<0mUV=VZQ!K!j5
zZI1T{hMIvLSO{I&3pBH=HW;QB`SuyNq79=4>TcAGobhm?(hPl*%OZWel5En*8?S}%
zi>n-``^01Xz9z05yuKqpH;d<1Z)tii`~Ks9gMaclr&axZ*cLl&I&O}8@mmr}>Wa%{
z6Ie#7aUM1Dg&c2!fBZ|qcc6gPvT%Y5Eie2At|pzBwvyTC<Fx=}Exvko+zaUQKJ|2M
zzu_2a)%O3ZIA>=6|8~azcU+#4mHmGQHCg^O^S9U$f5rLZJa3SJ${nOT?2s(BzAe|S
zjjLQ_wmuw(wd}(~D|tM!MnakECeJ!xfphDSn$Ht$^3&O-L~hRxch8o$H>uM-rTVSA
z=i3*^C;!7w-djl{8C)KA|Jl*US*us|{N2-7>U>5R-Rv%~-$p1L$ZyNtBpSE(E!cHx
z{j2gwT(+43b>josAr4UNWB*0d4wBv{HZRZT!%-?$c5ID`^O@CqTyIa;*5^wJVn9(<
zBw49KylbhSPp<?VjQWu`w?}PPOex+E@0ag`Tf61Dl(%dhsQy}$(~4_dzS2raqEpe>
z2T-QLJ~fgRI@H4y+Wl~l=692HlguuLq}4-OsrT+leeOhKXi(YRL!8C>ub(aZx)g{`
zy1ALySOyd13C~wp1_M|5^Y!YA=;r&iQ8kpn??~{PRTD$v65tX+K7k<t6SbYk@{^PH
zlO-bZ&t&j(`#t;{k~OC$x7C6*FKY4wtSD&rV*GgVLjQfZWo#v)Sc)n;g`bOVV|!18
zk5{alf>xFaOTyYQOgT$K&v(||IdmgYPPn=cDy->?(TcIQHIm{^!lfDb`Oc8z9L>{?
zKl1XaQyWQvmALY21`xTR0j$!HR|m$tcdi-W%B|lrZw3fbTRBgmhA^LEL<!s^y@Bx@
z$UF@PP`-wlQ4<-3U<_lNTxKAS`ue~yZ9UK1-m^g@_dYRhn6ij$4N`#jC&h04sdwE*
zLvEt|T>KgKpo?kUeK)IRM+18-<)!Q)p)%ZMM&plV0nLMt2q<TzdOI<S#@dUySPJS*
zR#tNH(E(E^NxXuvS%=0B<v*>fs!cpM4=7t%$RX9b-P32S^nr@vg3`H4F9@jha*bH1
zPinH(pK+zNJ-aJJ0S!d(kI)uR?-p~7GOttHA9Pz`O!Z0gW`wQiIZ~T6>0bY5aK}XK
zH0p=QtVdJC(lldyA5>I{eZLJ~+IPm;<}}fgF?*6ti=;|K4vB5~eHgA5h=4R!u(iN`
zb$Q~U($IdDfd1XqATxgRa2fqyN(&9DwW%ejVmHoI#|^(n*-FE*z$Ry96fr7oRcvci
zMy`OXTTfI@mQj79xlo_6(i@bqMN4yKXRq(}s@FgI!M{cS8sglzc39X4c|eK<y^L+c
zeJBj$z(WYFFOx+@H_$=5?W7fDIo9AFZmu7PP5DrS(~wA;I`%6cV~1LB3EpI@`dOQG
z!&gWCMSNptSH)|<eCxjQs@<H6v@9%+yS-`{&?{mpm_wWVSlkrQ_S7g-(7KU2YxkUE
zTu>fyAJ6^3H!fl#o@N>S3H!eoJBJ|I0<Bp$PTRI^+qP}nHc#8OZQHhO+jjSPed7)P
zJGk?z8tjN#dm;17X@CrThiP=^S7>TMTvhCKXogdAg5&<g;Xx5b|J!~<o~ulnBW_oh
znRF<1W-e!UH}CQ;=0p&5C;dgqfPMcKQRzUg;U4Z}z5O1I<Oh|yyU;U1`-!2tjb)aF
zw<YyG=UWCGUu*|>#|~OF8|+>BF$RV9==Ir1Lzxk4AX9JXUP9_*1!z*2fj9mV(q5Fc
zI2q3X?XuUS-WK(_I`1AMYX>}o>kQXm9><tjj38q%Yv`ANT@XdURYPUL`S+`m94juR
z$U1eanW4avKHG>YC|gNa_SdwmwUfjEA}T>!!qLF|UnMC!Myn%pLyT$HDYs=kXvaMu
zG?T%~pWPEx8C8VTAeL?5xE<%05Ie`Rh0K}e5}VKtE4Fw$$F+^c&+~N@)duv<aW3R#
z`voM{C~uSKSP3KAYD@1hY{U$=+WT~5rZ%cSY)m)qvi6A=zGy1X>={ZU>JmfvLZYLT
zeFnK{_q-<EH6^J`mr#;~*nN#p*+R=&mV^|n_^9OJLzk9>Lpjn;J0z;@c^m!tAeWfB
znHT8F>ImN<6X^nj7^{!~pCrKc?+P%W$M7#0$^(^;GvhxzL&<x+<yA)hz_3b<(hYDU
z*2GpU1Zj$U7ZBHdNGA?b6l7bGac33|L=qOK&}>zb8aA4@4{U9ATXRM{N^4SNA)lGF
z;}ps+!++0ZA>ehw4TLwEL~a8Y;vs14jB}i*jsKFXHw{Bp4MzZXv;vvzNGJr+Ci`d9
zv^vk;uPKb%ro>>91+S)DxlWiLA$zW7zm!FpDxR$%xhsnrWfNxDNIsdV%qtcQVlLV^
zYKt^d&<yylv_*_i*1&IB7CNBNc!+f$IKUbg5zF%*8&z%W-sL>D=l852l&sFb+h$6h
za@k;v8RNKHLWw$IjH#PXD49aeONz#+Ir<c|ObIIxPq{_J2<g-&%kD~=WFMvu*)LSd
zXfbS3HzqM(d|CGApkk589zx0xH`SCnHh|Sz`aU)xb6L;-*w~nrG;}gop2?eOLxm3Z
z0WqV9x$Qn$*RaD*MANs(ub-?;)s6#4cCTwYaF}S;PovtzSZx-_i`GuhsIqGq+Qn#f
z#JNo+%^>c61{^Pm4J({Q6c#5n%XiVbCOw#&PqtvCKU8e-%k137+O+5>WxaQ_tYxy&
zqGTyrw1_lpv}IG%xVNmy2Ko*E$p~{|cn$fJ;SWRo<|dQZ>0D&0Q!9DI%eD?nTagX3
zJMV>dCkp>UW-?)(^*}Hoy{`Epb$64__*=UlYd)*D=~13OMv6zjAF9&6o>b+q-IGwT
zL}~rA9D=LdE5iZ06&yUYqI2nN2;rf0%ES3_UpI6x411Y67}uI8c~>_#wTA>(ftq%Y
zKl~xxLH$ve&EN{xy?eV;g{quI=|a=)7ri*}MNKkvCk*JsAmm$w2B-7Nc!%rN5^4b=
zinvt*c4bNxSN|5rBy8%9r02XYLQ#h4ia7c>wj~BQcnagYa>zahD#3g6UB(X1MD;z;
zvITa+d0%bzCVvmOwu)qVYQg_yOdF)_vP#Z_1w`GaM;o-;eu-NO1}sGAYj-5ZgcDHh
z$db+MRLGD^YuIWhn9Fo(@Rp>p)$hPajx~sUrZ^GW!6l%4Nqj^`Z>_7l@NC#74b38B
z-H{!&T@WahrsOkE*zEI&JMXJc^g%JbPGS-`Y(q1mQVC6K&dh?-<hqnz8WIKj7CA~{
zuuag@mdqi<(H2dw3F&5&loO#`FR348f2ryseWQO5c0wPO#Cca`GtpikBs>n7J;{bl
zOTPOA%*&$>)0x@JG9kkL&;ZM7wDza2VGXPgZlqS{d_s-!oKy;FLY=XLv!LF337Z+X
zhOwZ|!kNrIL?4$d?NkZ|7>$YE%7Y~6>~=H}X<~0v;Q{(=FTsP`RPh(^@mmY;1C$e%
zke-2jp-RnT|4MF*nOIz*fdQM|F40zr#FJ_a#1LaCY%tUPXXe_BVc6M<n&JSnx5d_O
z)?BWEy4rU@Cu?t-iCM(?t(fpd;bTcb7bep#_f(GX0rjzT@+-29ws{lC#WX<vmY>XO
zk)x^IMP7a6aQj$#psYPtVu}b%Ji|7p^^lKol3}>4!exn(VjoW0YEzBHtb-73-y8^N
z5u-6$y%&BdgPGrDzU}KF>)>k_lezhz<*_uJS9>WNUr&853s?Ce<grY^sl8Q5r26uE
zE^`R9Y4Q*X=lP~FUR=LVEvs8qrI@$~pVS({_*V}h<44=4U?KJq%;K(q_V2;*k<Z2u
zqw~LZvn>Bbn##)dzj=qQf4f<t|G#QJ6VT6~?g6X@kZJ*4Q);-gNn9jk8Z&@RmrC%<
zJ!l-iD5aiE)H%wFU!JC1iF|(R2)*sOdAoVszJI>Cr(T|6EDa9czD&*T!@~ZDf=!3n
z-6%*;G<@^Cd9%uMs(<G+6nmg8`KA7~x!=mmExvCbO#e;<3#S_XU56@b61@}->775b
zOL)6yU+j_s=1caq=i|Wrb^UzGl_PRGm3e3OR?w@fd;8;|h5|U=bk8*yAYtp;>(dFv
zmNYk|-rJa>apSi8ujkY86>XyVN~EW7J&V85!KmqG3N*yx+I2*nw3^0LsVqy>1SLZ-
z`yTjBK!nBDN?9|scRoq;l%>>r`=ojHy@TUVBq(F?VuL*gHOyp_H1)xE-^{Vion6kA
z8i<NHYNeYfu+o5vfK3CVQd`x}<M(BZ_WQBZAOw+!qUZsKH)pB}kWE@rj$#l!gk-dJ
zyx-^^`>8zPOH>4?(3)i~|IF+nCsltE#plIhrpnEmwD}YUD6Sc|JwUd*>#uX~aQ$&y
zoKp3IRf8o#u3y!cOv+hsI+%7ZJa6+n)p;m>vx<SlEMxUgeym-K>Fk;6$(37sH&3_+
zE4s3K35xQsm>a9f%$e%%OXt>3-ViqwC{24BSIInXdSLH<S0Pe1z=Q?GsKhKWxFmw3
z`V#xm^I_cS^5TB~U0O%Z15T>V=>X0d^YnvUa>cr1PGw$8LO`4iN9@C$Hn(;Pjd!Me
z&2>-wN-_}aY_u9#MEo8Y{byF;=nk#7Q9CC;jfKv5geGqe%oX<qxTs6LzYCUh>CCqu
zF9{@drxLO(k%(QuXD%;|?Ey<;%+-SvK|k<FS8`3H!Z^sF?Qe`i)m`|(Wr22rkuue`
z-GkcGSb=$hSCn|cbZ;u#z+s^S#kM%LUC^iu=c>{-RigB0#*mX4#5ixx%)u)y=7&<;
zffcp`iL9dW2|LFq`?xkzDhTXjI+<BkJgFoVX_#F|jDf6QF^-daxV(YTmRYGcVYBJP
zs;okbBUM^5SMF>ZVU3!JNw;OAjC+gb4bx%shFi$8<L(-B75hrJ&EPie1|&8tBz`$>
z{ie%sFR+-ebUSRL!;CjMqOW9&RuR_<H80j*_YH=;aC(M6D&Qd-EOb=G^gf}7-T|_u
zD(gw2v%kx0rAsBA7~uFvrrGzEW9^KakgxO3W_hQ3<E>y<qq0JUQb$-RHl5Ip>c$9G
zmKnTRb<b2v1oIPpb89&vl!_M{qSX{~YofL`rB)Y~<r)$=WsJG*f)~>Un(hFWrKSc_
z^~}CJs`TM<_+e`2cUuZi2HO5xr8i*_ZK*=3g~nt*m9~lU2h7Ew@jjDJffP}L5w;<A
zDh@frex|k4Ct;1x;ce`f(KQQ1wW;H-)IXR*Q*c7U;9aC4JLW$}`}RZu)~xD$tTi?2
z%h6v3k#H)^V-3fpu)hG(@VRBv!WUf&sPt-+O8ew@C20dS8>+1SKZl~78KjU>M=5`i
z7Sc4C3B~mn6icv>D9LB1{yZCt>YxtBAVb9^2Fg$O;T3Hx`XXf~e8d_$I#z|k2Djz8
zcua(LTvCT<ccWk6rVYrK-rTjjxw@=nM(YtUn8gg19sLk@YM#4yp?4r~d&fwH8U-zZ
z!uh2AV=uBWf0&`mJsqt6)9&Iv;f-Z`4t|_rP*E(Zqy&m11XiqajPH^7IX~0Dufw_u
zGjIG;_ZP1u(Ow2TqE960>>7V-X--L_g_?5H(Y{f5Qh!_0?9d?w(6F%7DSL(f#pUX3
z!h^BNl69vSUN{-_)G_JQh+i2btt9#DSJEx>BSpP>uS!CVB&!qKiYSV8rz&{w;9hwU
z0(f=_s5WpW8iu-ld=iGrGRrAU?$gIO^_)g!;7s&4LC4ls-ZBRU)Mf3_(TnFq5`B9j
z@0+%jjhtbY^Mtb6m_4H#_(YQ3rcWDU7;aR%xM@nu&uhv>>l$*cDrS2>AHdSceTC(Y
zS{rtj5cnkfQdzIx4A}Cl`&uI9e!Y$vKLckRk~b2J*ST5oz&y!nACo_lY*W2ql9NBO
zutZfj8^vE$3sf?S6>$<1BcPGGggzMSqiX)L!Fp|2LXK6l-?`dw1vvV6m!PDQLtp8<
zjFjpvWCZhCF#oLFlwJp0z7AlV54uF^Y2Dv%2Y!)cpbKTOTe!4zI*s=Dqdu($y29ag
zkY48w+X?<lUK)*hPlDwim<ySJj$w9m8Et+Cz)JB>8%>Rr>nFcHS|;a-fH!@el$#Hl
zeky%8{95sn9k44z>Mlf`)V2dhofOBH9bl*bG?Tp#NFCIe1mtr~-%Y&*K2n1iLN_`n
zC73=FPVg?#Q*C)rxtk<-s?D!@Kmy6-2sW$+@6R0M)FUbz2|%^tl!33eoxiI7o|kUt
z{tn9O8^OBw5#&XvMv{vD;l=YazSXQ5@W=*|H^7i4x50d-ZWr*5AbU4kU@`+{!z2p0
zN|<}-U)KFHy#|n)xMg?Jw=u#6SQjpoG#f`lyC|h6nxiMk(*gzPCJ5$d1vGqL=A0Qd
zDF}9qV6&84ViR{63{hRUup<@q>^OrFZ0~9~$?MP>%Ho{AWseY~F3Ff+C_fKSx7;V4
z?M0Ov`cL&q2_h}8){P`6z*kP;O%AHlazzJIp0rNx!_c!rWLQUf>6V93ar6z6dzA$`
zo<~i290FXOc&0Ofe8f=B+2<6yDm<VWHp^3!?~;)wIU|Wg&Gof84qA|s9(C@rGzPzo
z|Mo}ws><LIGknkV9e`kaJ4ItlHp*#6zb_hP<z8s0stpddQs3S|)*TjW@-C(D?|Y2@
zRz0%+5M#4l0$q(BdUoc|>n_yIIVpnlo%G=uejlSf#l(#P$h%6KlXUN7s);iSRqSez
zG0w<v2;mbT%{*j8<li_@mJKm)PXPZ<zU1pIX2wdTe}o`NjDQ@ke!3%m)dpLkYffty
z#VdzDo29JOp(JTFkA9LJ|3(|+s*I8jN+E^iymktwminl2UUSOlx~*Tk4iL$RH*X(Y
z<^=)opk(R*^cnhHF61~BAgm9+iu(~eY|Q8&P=iZFGRyTfvQPy^(X7QY(g6!a-21iV
z4CW;VwRveh*sJA<JFW1|ReBZ<-YHMl7VqFGPWszAVwB;1$4&2MvZP}}p_Ar>5RW4;
z7Sv8w3o=_oSg2sCyAzwq)ws-Fl6qEpl$7>8Eu`lq7A#jRFfOC<=63gO`;jIS<GIJ>
z(Y&~krrQ>|^_8y+(b@F$DPdUIOm{o0yu#N0Ox(GT(xSFMs5kCW!E6*Two1g<d@pnj
z8yC#yc&@Qv)eiHQrPIR*K4Q0U6Ggk@(UH)lwA+g`0Z2z&KF(>W6!C6b5Hc#Uj_K-J
z9p!Kr^E`{RdA9F8DO40crn}X;Gw&v<70I*38}S61GqR(k=rn`79J_sbDm{2Zzj1V%
zG{4jZRlb!qm^D1#!<0yeLih>>o>KRRsNJEgXykFvR109WUcP_?Ug&f9`mbX$Ivl3r
z9QiK{-Z`d~gtlzgO)FXUrKz}Wv7WR6u_1t;9Y5dnrdgG<|JN3l^}nbk7+4wqC;R)g
z7S}I?2ma=ns-+2v>1gT>1RSWSahAX1&qac=0Ctf$g4+8#25;)o^-cB-!C*kdvxN4H
zy>8ZMX=Bz;*3r}Z&H4LHD9?g%@Z!Y4)0c_jm*7LM<U_%HqKYKC^5Wy}BHgQE;Qnbi
zUIJn~-XtH_&k`J-JjYJHwC~a&nAC;u4pe6Ym`efBuUMjVf{cCVKNOzb7(5(aJYTP$
zds)-@$}E3DEItgWz+wtV(Dtz7qCN=?-T5iCM&zLd^K4(TlWLFBnuK8;)?5AXvag$g
z4d6dMraAFD$QxtpM`Ty226TKaj&NCdH><&^R9(tz*<}S)w(3d&V#lCe7g{@iN77!e
zC{612;Jknn(RTdvR}RmT>|r^jdOlx$2sL04omI`PL=PJ;3oHL<0d5n6SRO+AVbU3V
zOu%aUK0cc9j+YPHQ`@eHl9x|P%rD7rC2kFaszWDX&w)1A)qpZjT(w#A23fbkuPxiJ
zf9?OteJM>0tQo!<euty<Vrqb?CHN$W^P6IbKcxH^Ib~6g>>X>kIKyT2DZtipS$cG^
zlT`Qjn^S))7eE%bTK>E9hbbg>cF-<2=k|xAEf^bmKs?L$-zRMsesOTUFCilA`lFw;
z-p#Kg!;~*Nmw|0qKW?OsSBD-wzd19C*uV#-a|b5%4S(H>1*WL_fnuW!Qp6gllI%J@
z*+qpslSZPe`UmNjzxBZhseH(GI6nRtG;}<7A{B5Z77^zr@yehg$o7TXc+GoinrZ{T
zbKAJUJc>)~*naCW6Eg=JWW;y+_xR?Rvk+Off6PMQJdP?&8x%JWY1Nj*6oC#67_6IL
z{*dDA(_ehUwX`OrV1JI3%W)~U0fhFq_|G_lq4pyNr(v@gcdBskoN7w7#w{k&+rtCf
z+#C%SH;|1(WHEdRa9`7<Ao4>FOQ|I!B;m8)afZV2N&QCRYFV^Meh?E?RV+tmQWkt>
zEuD}OCwnDp!cA0*y^Zg(g;&KMKd`n!Gq)1yP!}7=Uw5lU0!jd=eG6#5Cq0_5-!nMA
z>l;&tv`eNAAtFk|@tuaVRQjJnf`X|+cvEbUWGN>4E@mlM&S)!@=SF~Zuv(?=Dk>ey
zG^;GiA180gH&-8Ov%uM0Rhq@LrZ`<ht8S>+TGmW_7}Zj>Cit+<C=C4=ye0u1VV+JD
z^D?Fvxkl^YVeAZA>Qr}|EeeaM0F9PrbhIBXN#T&i9a<S{2JHGw!;~IB&~}Ne>T%aN
zvl3$+k&c$cxr5!dZfMd=lG`F~2=ychT`>luM0iH%X^dVQRUsn&8Z+HK(Xj?LqDIT6
zV=^zB<_aU7T3zA6CcwcWi4ZK*JE4wYi*T5?s@gP-7FLowLj6jxcRQP4UmwwhHCc?{
zhk2W#1kw}JJEI-R#$6xEY=@`%)85lhM%KJ*T}{M_FdladamUg}(m&&SPO^0pb^cSJ
zAx5PaEHG|6YFj%zuwGKfNhi!xeWqyYpa>G^yykUGJ5k{}pD%SR+8WY6cg#%8X<w-0
z5&e9tjP5KW&PR;4VK?TfT%!-EP4?DeXAO3>kutw-mp^p$*&t(kY@!Q;1aRX>f`^+>
zRF%1OOUHI*ldIjSdMykYl}^?wh`M~~319`st(1(5y{*5bgIl>Bg&xmA0>w3%X}Ifc
zDaE`1jKKe@8EmR)x!}>1T5;DEPq7KyESkF%RUnB~y)ZjkRJD-+3jk0&6+q)PLU$fA
z#-U=urd??jvewYaM9Zl0aAAZtnRyzK+CG@9rc*=QmSKv-s<rae%cwyZHJ}7ntX>T%
zx4POeyY*^z=Ubu`T2_<D2`_^l?!v1aYSAP)qUBj`9Y9__Qx#h3l;bJHyrhX3+lYF|
z3a_kbGPX4=J}1Yq9O<r`vk_Wlh!NM;#8d^tJ4<FqRAt|U8ap|$c7$v(bHSt*S^*a#
z%{2!cpv*NFG|mrrRB_(UthAds+t^pHmGd1-hlyxVL5uD6*?{;9PrD$ULjoN2or?k-
z*-xMVj0GH`0FEiFOahEyjim#Q1w4uX9HGmt0^`>bn}Fhj*`fi*?Z=9(_>#RD(OpgN
zjqazhJvm_c2#=dMidyvozttzd34g^$a>WkN1>kiG{ZJFPxr3*2%k?K@M2Zx=U4k2)
zgQN*0BH>FRM<~m=KsP#XiO=;5jPgsttKtTYOiV${42Pe?w0!*=NCEtOVPgIU)vDPB
zxej`L0@cz5Y{9nnAhqVO_aLxtCmxJ6n2aB94W$V<m0H(t-^FbC16~;E@t1}!ghop$
z2V4twq?f=NBEpL*60I-AfYRlt7E;SqT`rJjA5%}3i?J%MZUPRn8<DlV2u*b{Oa}8N
z`FS~CqL`<%=F?M1jd!NSMI=qitPx0n5*MqhDOD)$_sQaq1spu%iwCEOCA?<_%b8tn
zF?AsP;fU~>l06-?gP_HnS<F?OhUk9F5a7t@&g>0!>k;r0ZDlm~o4AJIOfwtIBcUN5
zq!~j2*vpw;DV|cdmGiH-?)V@e=vUyeO|oo?pDXtO3^MC{3adu1M<2o%w?hf6g9<ID
zRuxzURtZ<zwcOKmI-dBkVF#NHb&4YjSW}J$jLG9b?-4oPK9ZOJDfvQR4!sQoIP73*
z>6H%IYW&-PGpt|?Iu3Q7PqYDQ3IuPRjS;&(ViE@iVH6S;4%~Mo#<d#ayE6>Di-%8X
zii@pS2#K@RI$*%#Rwb+U3*clYPl#N)p#P0KF9s{n=y+a>si&}#a(&~APqX`8v<M)-
z@>y)ZJT5l_JX*snLXxq%1BvoYYd}nkZ-APO-`aP=JsQm)Ny39uWy5rs!}pJruYIWe
z;fu;ItNO#q#qrjW+St@71UEr*<3}IDy8B-n7&eaodr@X&``<$Bu74?lL~Q@LD4&@d
z8NgPmdjrM*jAOynm2Us)J!J(*f0uAvrQh#b#);e0ij`V$t)`%30x8?mo_9v|IV+ib
zY@Uzz2LE2}_VvvVRqFN}9$!3wzXTuU=Dz0Eqrv27`cAGVPg2o-Z^~w~cncfoKkh#~
zcdgsJSm0ei+V6cY>quSDR`<;{3@wEN>gAJc$PfW)AA2qbubOY^*LHUQJm*B7;lQv+
zpG6C+BC$>->Rr2crh<dbCq3r^`0_uw{2G-fKk+GdEoe$kt3XJ69wDjikb);WvW99d
z^s@u^Dqu}Vmp&;IJqrCO7E+kKu6;{&iYMe=kE5mJ8gr6sx@0(!{?)))h)7di><pxh
zobZrz<oNX->dH}CX=~(~JZo3*Vhje#`cpVZ5t*Q)1i;>yF|g!Z)1ynMyzo)cnVZbx
zW{Y0c5w2R~x=m&_@VU9B&k!D}{YR>I*Sgr~nE)aJTHLh%XAJerJ66Yuv?s?X;j_cv
zcj0L4A`BUpO*IR8w{PddyE~RgO^Fu;syB@GRzU$;OmTY*vGRFkvr%qhwjc}?MNhY%
z*^YoILj{83cV;f(g-(rdMM?8u`I;X|&>~_U*FRjDlLDy(W3_&4t&WSV%!V<z03Q*{
z$t4yVM}y9QLOqG*DJU?$I$$>>qr3HEv+L7W?lWSNR3mulP^w)$V%ZQqtDB(Js<%n6
zo1ai0RTHE_W3laNAj<W9JU_H>suYch$sSwV=)hW9r)kG6!=y9TcG{o_TJr+tdH4t2
z*zbts9gai)@FO93wW3t-<o{5^%V&)Q6hxG&S7XN;YB{bRH%M4eU=Ee3`_z>^QWn;=
zwsf<1;Za|`VcgiV(VLiW>#R9XGPSTOcE37n2ltQXfekf*Z*=-r4|f3ys4^=S>o$4S
z;qf_k`ZV;Pc+Ii=k=u;eL%-E&l?~hu#U#P5XXRMNk0M*zE|9yMF3gEF(&x%BqTTvt
z^fA0>yeEmbK#aS_KcY>y$D_qD=<wKianZZkr}P%B7_t#ziFn6KY=xrzq+Tac7xMSi
zK1vLsaL7AGcAppdM6Z@Fn=0_t)vI<B=K9kk&@@dl)T>Fx#S;lvcm=r)iD8Md&Y;C6
zEGwoNi%DVDHzE&;NO}$}b{KL3zgLXFZrr()c}@Qo#T78+J}1z;5GCO|Axn3bJetC0
z0Gr~dLg`ei91zsHTo-_Z%%Cz)Ib$9hJho>Pc#LYGyd&n*c4=i$0mU>2nPxjH?9xUd
z%vC883`a~Xqclhj9>d3F3=l<L!`j%gGWw2G+L>I~jPD9e1<3|8AtkC5l|tkiUI!A6
zq6E9ndo6B^QO^&Bl>`LOYs6VcL_FB6GhtB>92N;e$FP<;{JXd3av}+e+K8Q?k*O#K
zO2@Tg_%-|vA@3o<1AE1Y{bF-X3O`juCW>G?#ygo3edH0s<JyX}+$6#BA9%0#<oNW}
zX~&6kZK>y84*_qRX08Naxki;r6WK4xHBDNeCSQzpIqYoxb`Y}a5nBXZ(G~<r?pKT8
z>SB;7d-m%rIg*_r7u>K#?V1gW6=0h()MqE5d*^})lGv+??>24;4sBZ54{jfFHOW-r
z6vgHxOtrFFAyvKt45J&X;`-nXoC$N-<>u*N?bFOT(3zO^*atH79-;MKT%K$S&$$A@
z7%9WqTP5|PW}P24O~cU<hPGW+{A_v;UQ03ZWSZq+#myM%a&}8dpU4&*j9Pv1;~(2i
z78|aBnRFX20T+Qy*_@FM8!m$D7#psGJm~!y9WH_=oeWo+Stu{pz_*SRoyjAx^R@88
zwk`uQbZ@<H&V`&ve=ST`(z|KreJ}Gh9LRCou+n8gF^&%om`+zEH_J5wG4q$n27}S`
zZl|AZELRFUQW!RtD^kGQ8LL%VzWF*PV+^~l?%$uwp07hfezjV|5Q=~*-^6kaoR_?Y
zFTdINCLy?u`bm&@1-rh6;p#W1;n$+L1g6~M0G8)>6=zvz>*njko(q}RXl*E_|7l9d
zU{cq)L@#txqgtoeyk7eKaIPnn-OZ8|{~mQV)X?jh0Hmzh6F$f=+pPd$Y9}vrf4*>I
zM$xe&(~@@X7P+;egh6L(r(;Ad3E>UeB*&7fgE=B5y84(F`>J$bt7WX?SH)bD#&RnQ
z+q7Rc$G@!Y3rCi5g*D)@>x=g#&Sr@4N1ggzWrVmErDyNf!r+p6i2JA#hxOYwjpfV%
z<;6CyNTzn=vumcIeE@695PrFoLh&(`V7>}5rHiUg<l`vpw{_=M^Lja+^<t5`2&QAv
z)KNVS7SD`JTcTV4-;JPoPkXe5tKoW;5?y<dCDgPHb2$!Vb<%7j%u$oFS6rvYnKBWU
zE%B=2v|Gh|iQhU`;Pk*EQ_$U8EG3;ZBA4s)d&Fgw59!aV^(t=nY2R!s@n}1339qjQ
zrj|-?(6>E(Lp`4lc|fH!Z9HZ|u0AexO|cEL<LTMSb8J94;!q#KQqx3sTAuw<q&_VJ
zV+>1#&<TmCsecmGlB7mj9spX($&jurG%I{^>xp0|Gxju2ZBHMq%mo~Q^xAP1oGcp+
z=XqQQ0whh;?~bx!l()%?kxB3zl*vWdAz=`O2sxC}MP^7gMU<S@@d{9a<qT}`qnRk_
z!`V}s5d%Om;dm2$x0P(iDd^yqQLw)cE|&F~Hao|b#;z%6s2T|wj8Z32b(IN>oI+?3
zeAn3@Xr9vk{uXfJ2tXvy=>$MTUWcAPzO?=ZKt#Or5CCL>4+db{BapDTXA)e5zS};8
zu)G$l0GyU@3|Rzs^w<b_eE&0}(EqSuJ00tUyblD#^2^|aeDEJxnj=S~6CUAJ7ch&q
zup{Bd&8KW(%F#poQPD7mY}>BS9FR6moP!%?6Z}l}*th_+dEwaSCKyBDJO+)PfJ3i<
zG^L~%({<%2C{1qi$7prj^5$rD8<Qz0;M&^ZHB(y3tfYR6pemS>ej{|rDGf~igC_Z?
zwknV#wV?|-mU4^$_pE^v`y1b285-Sx?U>mAuLh5aiRpjS;0<g4a+iOvmX~T%^P>0}
z)Tny?s{YIgsACiwXDyjZ@V`wPJYL|}7qZ^88e69svpwwL{L`I66{YX9O1zQphwDM9
z-fhnAZBLg3H-_nNvZm)hPsAsXQa_N=P|fTpcxfpI&wsuj-vLf@Po5=cj<h7*(BC%a
zIy^g42IawX=OnndFcF~(s59-+%Jpb%TU_0d%zukhMlf$ixk~Tv?(J^<dJZL=XR==W
zK3abkFF&_?Z&QN(jU_u1Mf-zLOMCr#ANBI&rLo)+Qda(>!ejsCFI(|hR+CbR%V=8_
zW3muL+b&61jTZf*ER$F`_*(~EzWy*9Ygc+RZr6*-5sysU8PMi_Uu$|S1y-D%RCBve
zGq5WdD<E!?UX)Bi@V>Y#$O>EyF^*B5#dH%2fC6TdGnc2ggRQr0p{vdn=|)5>rf>uu
zfm(iM_cXg;CNUG#;;p*3$*gFjfiQ*<^1fuH{96Rck(rey667#K7LIVgyxRpy%vAaL
z7N)KereF$g_giJv86={b1matw3(Vf0+-uu+LoO(Mj1q6E{()vBlfJLA_LIsVeIEw^
zZ5vjX5(_m%5VFVV;}(2(MF(E*9CSY10kTS280kqPQXw5lu3!-Z(53@l5s!J<OWEeF
z@W-r;;}qR{GDUV>`X(P2%bh)x&Ve3)|IH<hBxlmAyyeiTB_<YH`u1FO-O;QSo(?*1
zKH~Yz>1$8P#-$9jGD&1x$5%+<_)MLa!~aZ@ZQ((qf&6i&qhz6Ynwh>dI4~2B*;|f7
zLEUykhkt5O*pR`*QQ{DrBTBWQ5t8z!P7arfCL%$xWan`7n_B};^iz&~JI~EELNikB
zG&ddzYnh}gk8G*bi89ommQeb`%IWg{*ZIkVGSh(|js@vkz=7>mv9g;Q5bxIWdyI<E
z-)KgBv!GE3G^;DtHT!5jmbYJ<)r?}L!ZoE<sT)wNMBsT+!HEQ*wd(S5_njFCgT_2f
z)Vl}FU(PQB6B4`=P=!8ejsd{$L+XZbHA##i%35+yW|9fpF!eUqs&!f>X>o}-`lGg7
z0)NPD)%9IuN~IXNu`p0z_INeWQ{%!!!mb9zc6*qgid0_bz)+tUFsprNxDWj#$inQM
zRWA~x<atDsb!*rT&8uEbb0gz}u+=%@Z8$?q_B&pa2%{7%rkzM!R#@G|>SOlQf*e6u
zQ~qfTQ!sptpeN5!O4JZ^E-3yjh<<%Bi6m)F98Rzex^P5+l*oECs9HW^VRPq&s{i+E
zB7X?6D{hm)N8$abuQ6z>e07g#|NI3y7NYZl>~eUXrGV`4oaXfe>AsV0l!T`hHlB~8
zt&dq*vqF3=Pay%4;@zyKm}A!>bj6M5D}8k$&EGTPnH_Rkk)!$EYw)!tIM{qCw1Mu>
z8x}+lUzZE3DTtGGa3Sb86--BMx3o&DTHl~7@`CQDwdkxa+s(43B6Je=Ld&fv{}ybz
z4>NwAiE?h0#fQ9}{9b5H=LqJ3n`s@hYIMO%c5}*KoQy!@sE-qC9%F}}@wrVh#yF2}
zS)Q4(O$h;KAZhlgedI+b-Ol5NbzV^)&QCLaj}YT$A=f21OD8{HEm@^3_(uF#j{p&L
zw9fofSRX#_`t+FN*3=teoKS{GOJQb~j#(@d%O2n|DT(%J5mT3uKYC)Ts^ek!veV_o
z4Z6pDo!Ccibpu-6&_nS>0Qe($tHPcUX%G_p9*7b66nB72!mFHas$At0wNF^twYfp^
z<w{y6gZnqca=;WqT*=Ih|4I<%#GXRn@iNyekDNGubUxY<6vt#V+x|v4z;eSkMat_y
zno<U9cQq|cB{X94g01(v<Jxw89Hc$le2d05U)bcSpgf;3QL;pw8GFou3Hy!`ATEcb
zs9Im~4j$#}6GOY)D)@;V(6nwt>gjI;#gYhWPy?J|4wsAIgju($W<c%^>Wn~4S(i}7
z)mBHWQ8EQ>)o6jEbVNQblyLQsEKiGH?y{LuF)Rg?!U`*I>)B%&BT_y?s@i+jtSJ};
zsS2;&B4QA;PV6>jSy*wByltxmP5BIJ7#^#agtl7KA2<*()yH^B`f7hi8a8u%=i6wa
zfWIzMe@3IUE({cT2zis>kzT(G)yFb$;~{k=I}jsPxB_S@YO~R4*IJLTR}A?EI+W-%
zM6>nA8Fn{}g9o73Rbj<G^5Z#{mOd2s2HcElJ9@}6sf!?hsn)k|Qgj9~xq{TaxbcGV
z<tSeqV#n|(K?k{HIo+5f;PHz#XMk##%t(QyV$?w76&N^^fu}lp8ZSdyPOGa(BoV?x
zbSN-N6@L!Y)VprWF<;xqGP(qk?a*%JSxhw64LzbFP_Hj(e0R7t6{awYqWqZv-=`ka
ze@D?IkpdQ-OYnK!i5AR+EBVcGPV+F-a@S+^bld#W6(?`XqsA(L9p$7Y)<*th5@d-C
z%>Fcv8*Fj$kR-9!40VVcUh%utGogAyd|{94<LfO%2ka9L@tyE4Mv@@|3cZ@-1qE}b
zL2j+LV&{E%t)DXTT9wf6dW-EhB6V-_bwpDsw=p199^{L-OSiPL(S)lrZr$9|(I1X+
zaT~;qA~}?g;+2}XgtXEbo)|7d5L@-j5E>OhK0Yav#ndE*p5u;$(u<rNFU=bo8aHPE
zt9P@C?7PRFQuK&j&DYAi2Y;+s>(p=F#i)m74YsIoCy6sA9%iH3kr|(YHiwPGZ+MDD
z*KNQN(KunLs$CWx+zWjJR~$#cMs?ztB-AkWz|z{M2F39){?O9IsyQVZ`!y+0JCn;l
zJ23O@w2o;`<M)ab#<7i(u)?IE7B72$9r68AT;X{$R!|hq1!IDm`H4+RKw37YT*POP
zA8CSL=ZH{2QsurC{L7guR>j8`m=X&OvZPgj08z^SNugjLdb*Yw0bHu-bp;>g<`0ol
zS<%{WR@JT)%ybMypwlD~Sjw|}NC$Us(IUeDD@2BZnhaMmQ^d?yOW)l@$T#%lNjF#D
zxkP{-Quc(^#wuPHDvEm~U51Rl!@wB8AgQ`#FQ{Zqr(@T<V%BC_KtOYaCz}eEO`86R
zZOi#?Vmcwb?IHygmK>8rn?Zlzou5q`fBORO$3x{JqRWa0$jGU32cxy`<j4m-X=}3=
zv>BjDwiWFi%EFp@q^;|@GaAvLNIR$Oh4T_PvXfR5jh(>-hgBXrua=2lOktOwq{1qz
znJD?`z^)eNA`L{|%nXYwx5<85Qsx`m!i$Xv^<`ErzQc>R!hs>*L9ZVQBsx>a0TRl#
zEqOyw+Pe6VQQ74^4O)5$^PrV*WZ^=#h?q(JSvELZ0Tj+e#~JuBDh6f*T%37oa;~=Y
zyX%8%0>f}>eCSW=)5s-1H=S`bFgR3<)8B)dh>=LFX%2fICKs!8X~-ttZAG|blkoP@
zIoCzQRIGkEcJM3(Vn3`Vgwx)}f9CRZ4xr>_w{naXMWgS%A&6%b*_?SWse$s78iTeK
zu%zD^x37ULPIIy^r%i&YrkfA0-M*NiJJLo=VsYXKBdyq^)AE;_+9}f|nPhk|ML=$e
z$ddu2LXoK9ytp%q0dnzMA`mxHyv26GCZEJ({{gVT07_E)79D`6BXX;Lw1IxtG#g9B
z3oNqQ?xw~Xhe}uQ@HW@dOO4m_9T3nb5Kr!|YV%bG7XJ=Seuo5jm^@7p_L{D9{ORCj
z3%0Zo+4T(Ya$x!lcpg%~Zo*Ta1bF&4?u!V+B&_eLv8tq3?k~>1HRk(+8gX2B+xM|0
z%6L%bbNR7%FZ%L$iji?B@aYlPK;)=x#P{ymUKIj<rVpRLD#C*7FkL2ytJg&{qVp?*
z7z$!xDUWD>o{maNS-W6|G~%UOxx`f)M)2!~um>l?Jv3m8xyl_1VQ;uk%t99X1EOKD
z9Wc1ojVJ9wq*Fhd8sB7u|0<~UwXe9@QoCA|^+ZX4+@BrrLG2A(7Yv3K=@V;m1NUW1
zIVotBEa*EVPVweSq+>)37h?Vq-%=%ff`I+8qVxXqu!)wnMuwYAu)Ta{vc^u{mZABy
z5Ju5_rumN=GfB{wj!IU@_m&02&+cqLp{!^?%iW!8K?OHQ6li(rsFj=hI<)1PCXRfP
zV9dz)b5lC)wxwm1<|@f#|CfpP&p$iLKr_s@$m$B2)~nLDh=xl;hQed_4b(L7?vcw!
zKI(07sm5mTn$NZ>%g%dCJ@2cw+*Y3|Z!OGyNwuf1YP+l?RwNqaGiE%lzBX~co!9<{
ze+i%4L=*<L9f75&&iGv*BRAQ5K1XE(w!xNI&sT3?&o{Vv7dalAp5zJ!pd9B6*nJ6>
zIkfN>s6T%e2agcNiE+XG?5h4`z(hj+QJ2h!h@&JARM2uCZGV>b2O@+M!Zunko7(9&
zP5PfP^ab&<68J@ziEv&{s3Kpn_M*WzUj%&<NXS2ugR$k8)z{8{fT)agi~nt|Vy6Eu
zc{c3-qqnL17d&9x`k%S#*V`0!qPPW+4_MB_7^fB0PQn4=0FTc*#P9z*R*fIbv^Fg<
zg{ElU#y8iqCgi>y=c4((AD>@eF3R0P(x!0GRCRfOUlZPejOGR&3Ki0+f}19_d4GRD
zq@v9f_s?Sa`a2ch8Qpgd?tXk!1q~t$%!z5|(uAdQs_s0X7cWwuKU#m#g#8_iU(vul
z+YZa=`MJK#&0!CpNq5z5hcc7IGP`TN*Vg^%YrSOa{R!;gtY1>0EQ#(#hk|CA=&(Gh
zAn}!Fj79|dfvHI}Y)YAr&;U#aqMQb7UNtA-uPo=6Defh?C<f`C#P-I_We343xmAdu
zU(x<6Fqp*74Q(?nPWUWpoiN^#qMlLL^mnTEmb5AX=JKRgvpeob15t`vf>8U|9PUJh
zUgGU}@@_S?R_>oihBOT{a5b&pzi)~zfc*6OA3~Bx@(H|uZC@`>r}37$@B41f2W@p*
z=+&RWEYenP4C1R5U6kD-w+v*oWWHchoWPD0d<Ca2a=tcSgIp(Bn12xXLT4B9i8+%V
z9Mj+lA)W*+Ke$F_xkR+$xIY)YG>4BO$3uxJo_u{+xw!*2w=F4Z(u0sJgVBf9x#QwE
z5w3HF&YrxR!zA<kF2<3C-ukRy3uEq6%*grqlnal(fJp%+>6Ltpkjw{(LHqRsIhLpz
zRgSDrJRv2o>9y%2kFSbr0qBfXGavU@OzkH`lPCZjO;h<2OS?<e2iPb;Y>aiuFxukl
zyNE&^f&)N^g7>xw<oC=TV$kt5w2LT2RHb43lrCs`M0`POO;|~s>_vRsP)>>yH6}uI
zi!>5U)1Nhg*{1?qt%u}66}@TQNv+U5?%UcH_)p1j8$+(p9zy-T+syclXQ~hq0c}vn
zPL@le($!dfXE}<M|24KJUyR7zre(i!FR)W9EFXTOxF>cwm)Ca6DwbI+HM_fYMoqk$
z%4w=yun!a?zhftsjypXGlhCrxT}fgcR3+&Yxypt}2jZ31L5>*H^K7517d=gR&RaFP
zdNy>Hnq8vqlM9L0rndt3%5y4D!G9#T8P}F`g1N<7q4M-CoU*zhmdhr$ug9{3fW_f$
zXY@!TXDaEbe2`zXOTv?6tXqXMrIm154RXJ)d79`v+dz(tPbL|mekdPVl}s@vh0V*P
zN-PvlUFdEk?UyV)2DoP$C!bjfvfdXzNi}1=oYKjBAS#>gmUA6H;~X*mQrSE_K^9;4
zB-D37VZve8$u^Erb0S76ETs9QvRb6_N5+3h`KGg|hPx1-R;kG}K#D0TxYWsGELF2l
z;9M(F&qiySCtaf|?_0*L+7Ga~6t=G&DzHg*;a<veHr#Qq-t|)qM_=i;8R-lrUy+43
zT13T;{Pw&czjS(1fB$MDGaRQv<Cx=OuadP*s7%Dpn%{SJl-=rTYD(C9xZ+PrKM^uN
zv7J%$C{{a}x1qb{5zUh={!qq8Vxk$hjF8ja0~?#uAG^Fn1n8-*hQzMd{<=d)-KKr4
zp%P4WAGfxZg~&sMtn9eG3KWY*K!jlEi1MDA3!HDyRf!N`C$WHP>}<$(?O8P(BO!+?
z$m_2OnkoBcIGW3;G7f5KR~>826f;T4t+X~T-We0aHsGfv;5iavU=re)*TTkgR>>Ny
zArg(ZiUqMx%hv`87Ozl)H+u*$C5Q%E0Espmz#+vd3!Md<?JtTN;TU}}TNj{>#12ZE
zbt6uO@$P9i_}qVk^`JXOg-|4{?m=?bO?~!~cJyJ8bYrt*ZDRnCk-Rx!$1-Eh8($Hn
z@Hk6;gLGi#4jFLVuP~SU8MecPl90sQ+6)Z#w0YDua(gS5Mqvw$RNzi)^cFO%+5M7V
zbML`z<{*6&M5;^}vMg#x<|kf#!1;IQ$6Vb6rN(z}Qd9`l0!pw=f}a@*hfss$GJ+Gp
zzkGKnv3F&yKn>Y5J3<Rg24q{ZZh&s`z|RKG9En7<LB{!HRk%Rw>O46I(JXCzsM3yX
zTYkQAGYW&fN{acgXui+H#3u141)81;d)=C#8*qO`uB~#SQgYV_^Ni+?{^sWtlJ{xm
z7{}abG=<N6*fbMds;vWH{R?ou_EpdcSD})8Z(L#%_~*WP(5#=@Sx-Z_@T7k!b7K>7
z@I8l!0*&JqvLfQ=@9^Ld%+kxjg2WgWVt@QIA&TqKnvnE5l|Osr6Qj7&$Y+ClN3iI@
z5~*Xsy+VEd8P`<Ty=GBxLHu+d;ClJQY9%VHME9-wV`YrP@uI(pG9Hpy=7xJ4rHxMS
zmQ9ff?pkh9O(5yIpsAduB{j}(+ujulG2WuC=v{yJ5JZVu1f*Z$qRVWZACzE4Mrj1Z
zB_U6-wO*D;1l6vd+Ls8g6-Bv*<$~fMu%0zR>>%?o2{A#&#SYy)wMAQT4A|vieGcI5
zxz~mFS?@ya0C*rbEZyhasNVikAGRPA&gTN*_;27}1RjF|!(>-^#>M_vDGS#qEVx_C
zXnQ5J!z5rBKB)WGWB%svgc`0WtszbP0Ou1~&q9E`i+eGMBOtD#5{G+6avj6oC&tbO
z#1Sy=&Tf?Wo+G2th9Z?acL`>eVIU`2@x6y$(vI<hrH4<K^FhpY@JbPw=JuC5^83so
zjqN@e4{@5BOw7cjj{rRr%H`=MOsSx*xecji`t0pBI@+?^kx7WHY5F7{HyulGe%kDu
zRu1e;oHM737I$WL8v^VMZN_)5yeesQxhFxJ^)GALIm6OI%~ASnHrwH{=vYj^t~pXI
zJ?KO@SE9_6URzYzSb97sh6-p^8LfoR!*)<~##3#(S^#z4Ig#DuC(0|vF8yT-oY6l9
z19Gu<noWA$mBb3&K8>H;;tVz4vt@wRUA1p^C!j2i6y@kYl^*^i0b?pnP^2;#TG|1>
z?(#99;3#n_^$fx!35{UmJn7g3E_k%4A_T6@dGJN|(zZ_{N9LLz&c2#Aq6x7fqE0e_
z?+6iK6eRnjttJx_^teC)IN~M-i&k8A=Ez8Jjl=paXY>c11#;L<q@Q$g>`&!{yr5PJ
z&7d`<^wiQz`>@=~wGGB~<tJ(1sPP8pi@p~;0XrwZ{6bfT7Ms8BfY#OFhb(=jcQ_`D
zOT<l(o)u0Z*Hm|2#p|V6V8srqo7?T7jYB5rlNIo>k2>iq1R~R#1Wudbjxt5y)++O|
z=g|3A*>jU(5E7lbzJ`Gqf_-xU68y^O=Eee)ah@6g_zEqJknkhc^MXG%Q5w9gvx1AR
zaC#Z^5UtV9L|T64)j3&Qk)@Tl_wEN)7wcSELvpnUAhpWa8yts$Pl9v>hzDr}m{n$9
znuqSwttza!Jgdc$R=`Y(hsa<dEP1qD{gMT4miFIbIP^(MXLD-?qvfqWkC{IBZPp#n
zWbhTJM$sKJZdXVXw*nBkg_nkAfal<$1ZWelJ1H=?b{7ltHRqxA>2q-FLtqenl4Gf^
z7G-a%%slr<SQYOJ<hOO1M9Q_*R?xrV8&V`23nI#st-!0+M}f-T9WZM11c%o~?^n4U
zguqM2sEb9DYmXh-k;O1qVNpj~+N`Vqomixm=}(IlI4C+}nH#w5jmdM(jQ4&4ptM*x
zg?%{^791{7bIHrB3bg(vHc6~y*Ef31+J9;B`puU3oeOc*3|U-%v|}<PBO2mdg?tmp
z2+S|9FIj*q9vK!DmY4m#PR75Q#NpOeEMl>Te&6*5w*H^2{{NUp&Be-S;-&(y+Tr>x
zL-FT-DJy63i5HMyG__oQ((ZY=s(`~>94Z|2zF*M+;9*?*eD|@fZI=f--lt4n%Q>v~
zA9ge(^{d|#H}Bm-llj}6I5W7raAFsFJ-pv|t3v+R|FtV;{J+-@1~$h3E$(<)lPmU+
z75?U%>Ip1ml)&r_02;7!=MnJSagA0S<w#K{{_~$m2(pC2(qiWeVK8+jVd!Iy9r@%<
zuB_{>ES+5`omzA@N=#qH&9xZV7ykp9I4<*&27m#xvmIY2-^?!vY!yn6ym-Z1<2N@z
z_B#=Dl&P->>VquO9{03DDyPswz1Ukf`zFrSp8MCl1l%oEvi|S=WmAH5rTC828>aPT
zb+vV|x0En!Au#F|S_2TVqt`cQ0Qe8$$cHS&K47PwpBtZSZ>#d7?le!uG+?C+rVz72
zCS%~3%ZPb?HI1ba3O<wsHF*6*Sg#SfRIf{0C$reYcf!d!Gu>40!!alcbwj)5#4ndA
zIM9e1L~7m1b0R7pHWRM#;7SU2l2xRIq@128Hb^i2e$~JRHAOldVaN;P1?MFyNg6YN
zQSe!|skn-s*AY*2*}648L0bDHdG)OIZ-fWbwfg99&Y(A&ZpDq-+{Z!U?QVlo6&65K
zTuuW>0XF4%YQmQ7!Lbc<f>^+UF{GyHHcWyAi#o}dssI(eCCSa67*;5Q|8JcIUlYiW
zk7vnM+zWQ55t>y?RH^lr`^XCVJO38nf;kdVVFzt;uqLuA+Q2Zdunn9Jtz`DN!m4%0
z^NuJ&Q5nx+VjCupzGUKJlH10p<UY+R2`NCT#9J^cBif^uv#);_buN2{S!1I2aMg~w
ziPA}7nd*kyMr##tJ2I~w+A1aW3scjI&|1JmhPCND@P+Wuv8)bKAyK6m-C<F1WkA}g
z0z7PwDo^3$?@4Kn>o)GFIOhKB0y2=H(0y>^O4YW+E=yTR$!<c+s8nhs#1BMg;czRh
z3Yo<t4?4?2cRDL5R(jAHKP|#wZllPC`K)<3lvlUH<SaH_KhAK$QdUnCh)0V7oN&Ii
zFs>o>!}<!!XwHxMfa!@n?qTs|0lE!;<bj*lP7!MSDYm;xBQ!r@LjA*8^$2$SP&~$g
zV(3p-Dgjr|fVC|Mp7DtJY2O$rJKQL0aIZ*&hy_2ZJA!%P`IpdbKqTJ~O@T&9qW*7s
zx!Q#`FyF)*0U@ALyI>lW$Bk2nTiQwErF8~gn>oei10$IICa4(;(F{pLc*LYP7)4o8
z4&itOhNcEsD;yC<PBNjB1gYb?ccvqe8%@4-Gqzlr?L<^ce8w^j$UEa0QuE2p$=F%;
z+{pA*(|(^ZtOrs3wSG;wa+c}-#Cdd5?itj@ioei-ZGWUJfMMsUCjSbFOmQGWq-aFs
z(g1dZJnh=VjO$K}JXg4UD^|X1)ZA8#&V{SZXiF@uF?oCW0c_EHf4^eL)+jpLs($!f
z_z0@z602NSsg95N;`qBP{vM3XX)2Ww|9@gDQ)7bYTl5keqpzXW0P|6FkxZ^bnKHOq
zS|_6FqAG<b{(5YzGT=KQ%$qt7kT3BjD5ap^3KH=z!&+?u2Wzt7W_9)cTNZ_2{W;At
zc3?%TXm}CMy;U)_PRoZZVsT89*cUH(BD(GW(M$x`?rS6awbGD%lA%9XWccJYJW9LZ
zx*%Lur_a=2W-xcoBn^gDVd8m6_}(CbnMzVoGMJ@F<)`<(Z@?;8RusjMQphRIj8#uR
zVV8=Bs~{zDq@Ub#FS0d@?!ft>&%{ODVKUuv*cf)Amz>|ZD9AZ)ND7HnI6Lt4{4IvK
za8BMyL__voSjsRtxy;c1qCOplIzv;zi#WLuPQ5;Xk1vEY!`{WN+BDZ@#01;CX;|<N
zUcb^SW|faktCyWO=2O-#iy!5Q&oR(#ySH<1SCdsPnI#B#anCd@j-x-M_zb}MdJX^=
za`?PL;Ips75tbLu)p>;z`=&lpv)NZrzl-LN9RQ>x=77zb1q6a@zbAaY%gVUeHUkDG
z!8lz4<K9OpDu;bZ0fA_|5@3!qU7?$}`o-0ir&o0Fk=3TUgw(E%5(!ydyd?q~Vw7b>
zQ|6ZaKgQlER<vO2)}6L(&oyn^wr$(C*R*Zhwr$(Ct+UQe?tgA__epj=j8y7rJXN*y
z-ul<9C)?A<s}W^qoXK<DR2_EcYjjYVUWFeI<1OcBHLtZ7Pjz9NU(8`~O>b7`uGIFT
z=Tyu*cCnyh+EC@MY2l|{^t<Kn;;rGM=EdiK8z!;ESy0{SHswG)m0Q4vZP<KOTfkul
zDBno_KGsBC>v$8-H94L?KZ5*7|HRA3)K3^BTWpkF9=-ixp`7Gz#Id94{VxB-X0UN*
zWVd<V=-K^XH(y?W299pI<B)QOFuJgxJF^xd$x+pHJlcshfQi|0^u~whCZVy6G26b%
z9bm<av0*PZiYpqW9e!o_vGZX_`=t-eS{iK?=vZ~$d0YqH`KwfGyEAseMV9{yFizaD
z|DT>hCPv2pW0m~B0^1qr|3AQXj{h!ZXIe`qR;v}!=e72K1Kaz%j_ggxD1d<9%N7CP
zEY~RJR_+FGG-b0msf=jAZ+dSrb+40+EpRmywQ2OdZgg}VP1&8?LyMefDZ)?~e~SHk
zJCuPa+`@(GH%K8HzN*2Nw~DbUBvLi$Ep-~=7T+$frlW7$|DuO9^$yb)bkAxF)>v3a
z?njkJ(#YGAqSVCQE$U$jUmGbv9RYu61(54h0x2-Q>qmic`l>PtpO>IOEs5GdbIzKe
zSoP)B>)|gfk8Xk~jNri>JTS2fNv>wyYb0M4O28FIgE0v^6%r`!4H<1BrD?9Fa^!9D
zf=KBv^MYvUOX&pDSZ{F=H#M)QkiGFLT(?jNU-Li+Tlz4|q~Tl={TsX2(W4Mh?pQ1a
zwKx<uL>rRC7lbjq7WxY!)6V9fnk?B`eUfLQ3@*FWfCFon{9#7VzsXLcDL6dkhUg$r
zSlA>`b*Kfsd^-OrgCKT@XqKG#G(V>*N`)1osf1n`xe!){F;L|OR;`7{6TWV>Qy7Vr
z5tZAFqIq0nJc(eH3s8~Ia?wJlt%Y6Kbx>VgV^`VF=n_doSRy9YZ%d5%d92^1ybwqc
zU4MA6rm?8s>^3qtROJp-H)<s*<#Q)7v2jiRl<V%C*-r@hxORMCp)4kHr(#*|w@K`<
zF83;ocWRejRCrb5sL;C5#df~4f8@8i6Y90n%vcf5OVP|QG9TJ7tLV5$@v3nFh2$YL
zwrp|}<i3Q|j+j2aOI&j6^L~7H=(jr+T3b_l8o7VoIT^yo<LUC9OYhCo>%q|JkpjSR
z{(jfG|GA%O=Hczm*r~}~s7ySNLg6_909_zf3^6=48&(VS;mhgG_UpjwjomXwvb9GV
z4`VHvjP?5t=o_=-UqG-B%>2ULyPR)6Wx6R@Mo_AFyfRc4p;(M;9#0CzO~WDb%8Nvy
zdXy|5L1Lp7X}u277a1i7$&o3S2NK9b6#j>_NaE?dS1LSX2;e{iu?C>JVpx^|9}kg#
zcnl!Jf)!~r;W7mZhz}sXt)5jWH`5acm0S#r#6TbdXb*UU9}>Xb3Nfvbn6{dfX-dEb
z&Z<g8Hcv!iKEiOF-ybM`jCx>xE`q4eND6wQS>;V4U0Mb(oop;70hQ1o5ecAQ{req`
zR9Yl~K&DzyZ${<`M3OuZ@b74`h@iJ)ewn&%n$}bk4}L^58JnCmQV5!QeT-Tl=^(I&
zB2Yj8LAw%oIWH&;a9suye%}ckt}8sWG@J(qb<gEpFezCBoxi_msV)Kk-<Yqjolwj{
zHKl)$ceO_=@~2p`-OpS)UtQ{*uJ1|uuP4j<$IeN2mS2~*<Lz3zf1_mItY7x^Z|}Eu
zXuEoJrfhU{HKzBetAodv@5lBbFulUs7AEQAQr+I`7gFiPlj#?cR2^}Zsp_kbk^9%v
z%l6+Nu)JT3#MIue_{TXwuQAS?Xie(<sP#WXM^bc*b6J0gfci*VnCcC1fy<Fk$+TGl
z+VPHE`K_gU<mt9$F>kB0iZ`W#HoX>hxwiDBb(QN1w1SxSFm(B(g4zf4$EkvPeNIFI
z2S&&S`F@}6N|oya+*RniF5PPN_<AcfqGv4o5sd`1>%^1>Jd6B`crt)mAfIfP8DVy(
z^k(gHdIG=&?_Uhuz3ctuYJj6-|6#k{Zq5%38K6D9@9llN`!L`q$6!AR;4_(6??!(J
zWar|Dz5%sagV3YkWzRF_-vDkR!*(u%zy}XA-<KEH57zJAHp2#E8($+&t>Pcw4oWeV
zhmCdUc3d0Sv0jZ|Tz_s(EMHtTNZ$X=iqh%oc=Pz}0!u4As_5b1>22+4txk&4c+tPy
zvPV;Hd9kM0*dD7lZ%b}7&B8|;jiXwH$~sUh7p!}X)YBiRZ#slWW^j4FyLa;Xvj?E%
z6nC=>?v#wKTVZ*fv+XN~lef(*?3ljdD^TL^DaS)n%D-L*mNlcb>J#l5>E2z>H~9~%
zmH1?2JbJ$m5?<IUKhT7jnm`xu`m+ps2dUHI*I|*}?cZnq=!5rTwc4l~@xZK*)b?Y^
z)r}xmIi!>1kgNl`<Z83#bMlK3(R{FiS4HT{2*`^QX}TrV^RK~F=-!j!B|za1Ct(OI
zT~HE&y0BQrIZ2CK0uU@&ivij0lQ$HErro3+al@+8lcJQ8=^trRLh-k5e56FD3QV_7
zEb%9`zCx507*D|-M;2Q~Bm5By!PL*S&dyP+I&H5?4JIFC`~tjk+K$YzWyV(6coo)c
z@yQ@+?f2c_m5FZ|I%YEZ<=nEP1-XRyjStyVUZZot{uF^5^wlYCVkGRr5i@RIIRklT
zJ!nr}o6u0r3*kQCzzwS$0wK{TW;Y;BWRh{9rB^=>LP7uLF(unHAySJj19`EK%qp+}
z=s!KN!cK3GZ_Kpl5V%ye=!oyj&1&o@gjoijoa>lbH8wQnWMIpFP>-E7x}!qLtE^L<
zhMPB&K=(EZc^doIUL}|DfNjI@rdZU#%0dBC;y`MXw<le0?iNfc)0|z<cMNj7MfJ7Z
zC&Uz%N8F{~frr4!^!M~lYVzTY)3^^BC~N^6Mpnu=(+}fH?mmmRMX#qMwJh+-aGWpG
zc(srmzbyR{)4v8QxRFR<<$t0@`7su*Ziy~y#tSWrUO{M<J<}-dXq6?`GD&!1w<e8S
z>!2J)&?i@%NhH?j`2DNIV{pPCNoN(hlvp-Adji7}f6Y!n4P&AunB_1-?QJ|9`$eJ&
z{1&AdHG{Or(Ybn#0!okQFYqM}HJOPIM<r5)A0Us>suMOComj7G1e~WJ1nZr&+UQ84
zQO($1qJB)RhDlon?tbZC(z07PYDM$M`SshY{@>pNElS$tmS85DVS5#k3BAG+c|&yE
zYM|o}7H!64&${l-lvb<t+6aD*h~XWIYg74#I9k6`%dzm|$wRj!0(b!xt+f@d)ru4Y
z5uLvx6`r$}K3qB`AbS^Ds+1jfjF)ftYROjLu-oqWexGPlPCwh^#v0FLgIe6|76~PQ
z{ww-W`{Lik8Sig;<0hIE2(Tt+vm%mZDvtj-Fs`I?PPiaxmS`v*Dzr;w1B@-uL``qo
zI3Ddld*`mrufwm5*_<A&eW&@DHZ#!sq==(nM6@Sisqn3-REKPW3u<WcrGii53|t8K
z3HWP`f2UqvZY}^*hl*6_2!zXSWL@ejh3oCaNNjmKN62<$aBK|}s;o%V;J8IYE>T~F
z>j*t}#SP!SzVR(4NTva|cLWtj;KoZ|(}xNA7G1WbpeObGxiS@^8L%q$;bMJ;ir-$v
zYz_zXg2{4*dFs~jtxGO-S6Pvk;tNk^lgc^9jgQX(5SmJ(LP<6U*q{1k>TIyYfft1|
zO=r#|CVbFBrr7_n{^iVKCZiFN#UB~C#MDdFvlJ+R^ab%LOGA$yJuOdT7(g8nK3>`e
zYf-O<t!GIX!jE!A+_P69z8TRseX#0x;h++yXE$CM!S~5v%v%|OUvze|eB3#FWdjcK
z$fR!7paJ8d>eGdxX~%#n#3HudQ`S;xajtgxYL#L2^c&S6rZnW#^h-teHO4Wx&$gA;
z0!>sh2t|ZvH^Sq7$EQ4Yt^S`>FEh)3C;&0D{C5SQWzBJ0Vh)(@PpS*>0yGLmXao%L
zEeT+!Oi|k)4%m*<rLgl~J{Zz^RuvDaD|U{g$rNRzRxzG>>rOweO`lEf9_t=%u8Y9%
ze`TZ8TwL+6WVyP9vvz2J5w0^`Ul+e^tw`M(vGy^|(<7({o{8*MH@Jc*fLe6xTcG}L
zb~LwdjL~Nb>F-5Pa2>*lF@&J$$7|=Fy6?~9*l3mc&rgC{BV=F~-Z;c^<sGrf$5I%_
zs5K2CN0HH*Pa2-dmD}ARMnqK_A#3tb(_+e2^N3l``k$F=xfLYU@2o9_YSoP{&G?J$
zVH;LW+nMdQ#nC53Zwwe-73*AV265tdRrR-yZsJ^>enz@*A&QffB&{J8>Vd>6YBnh%
z3?BGyB6K6$c(0{Xx8~@4{N&vIqS>2^vxn?wudCJ4B(+V$mG&uvN?mk&6hdrii)jTF
zvTk+1Q&)#1rFJjuspFb_Y9zA&Y~<_%1XpuKb@rmEqe+-Z_DAz7avoer%QqY$_~#kW
z009|I&$5n6as~zdIF4lJ0D=cFh#g>o#^jpNydp@^Qc58=my87aiG<R8if}6WmOwLR
zi4M!TqX8tDO|CY4#Nt_)zZKn9WfjDyWee@9aVlyz@gZvbDZx;TRUCKMW0W1JLRn!|
zgcUfLe|rQ}W21x4!V?kgL=c68hHe82)+JX3<1brJOYa?C<9c&j;ip5U#{~MoI3Vmn
zI5-V&b&JFUeUB(U6_O7#J`vf2tZq7BOe3n0iL;xq!|CURsFVeEWasbGU^{!5H0ZFe
z0GL;<oG9nufU59SyzeaSu^^$*huLDfi$JWr8`ODPH=H1B&tN{696%Be;vPRRsP9KF
zTHA84k`x=!C8z6i`z<mq5#!=j4E0{i%TXtxERYCAR?(hFp>$(ESspY>wI61E0*;3?
z)0?JTeBAMn7(OUumtzEI^0DA`rFk<McY9cRb97mMZO~{18EHdqa`iH|1o}q8mh$Q8
z@@`&ZZ%$8UT~|)j4NA$1kPue9PCWSC2MaRZBuh%MbAe&jZNJD7d9B`gNAdx--yo{@
z?wXGdo5w^_XjrN(Jn3BmO3Yw&*<2jt$KD80Brl+dJypLO-5%9em52si$0Woja!)wQ
z!&!!}NDk$(e4{@XieCW5``z!LHRw!dY<fLpragxIXCfl62mDFT{bg;z#qp#PP(ZNv
zN6ZepwS(75v>`Oqy^m3q(Dc^FbC~M6282G&29Y_$vu`7ChDhgioOK=MuG6@*n1idO
zi5<cQy$wZWJ^-GKx#2pWVBdkpF~>)rz<T%_=GkX+1L$DHt~rnQ7hiUij{*hcqIaNg
z>^*k&3=?q1ZilcLQFOk<mt&pSLBxa{_05ymhX}jaBoY=QHd2WZJ+w)Y><AHK5(OY~
z1BII!d?AqLA)Y0X_90+IKiC1*^|tE3@{(IsK`j(ei-f!jUhoc)1LTg8O?s$Snf1#*
zqbsq@qp6pnSP&vCQHB}TRDkyBG1<`A;FLGr4hdgEnT6C7vOK=A&B6ZAv+j}QVH(ji
z5P4|CaI2=IU4F=#KpLA7?QzXR&kHNdu+c*I7EsSa_;yj(7o3YxN#$)JFO=$|R9E2#
zDliLnCn)44FWLoe8Xl-?6d*RtLA1KJpTTuktlp2B2Xj5EbjNg=Y6!-7cb();*8kno
zoA<U<5fq&HP%TS#Bet~j0c;rgLvY;s{iAq$$9FY}2%)<tX;v3UH`|~lGDzXx`tnPh
zx5e?tEm}53C*binyv)2u5kVi%sH3&{XV9<~?K?xZV6`KXF~5DQt~8?l&yxWPjw}U|
ztAOiRyC8B$qTZD~hvQX+Y~vey;w}jx<Izs|PLcP(7BZS${k8kb9Es8Jp{pBj=J@Yp
zfU2zfj+|-I4|9Ztq!#(U99+zzG%emYK8rEh(|^4D|K$YJ6WAGALUD8dPYfL!E7O04
zx0uGBj3K6Y(AC?=HwO!BRc=^%Y0}ErZ24%}ShO^ib6R`<9RxAv7mO=fq;<Q+)21~8
z6F{Il0DO4LP2~3R>GbydJT93YkJL{NTYTk{h0+uMswClcemP1;lCAuABKPZ^iiG1!
z`^KPX8I2$N3;xw|DyIn=EFZ&r@imV9+%quH+-Key6{s${*jj)IB>1ssGT=t)z5i8v
z+uJQc)V%<ysg5o;gB`Gh!ilT{cFz2Fd@XpC04^vg4cmwMTXweBp1g7)QP?N6Qk`1S
z0vW~^TQB#(8f^l?YBo;HG>!SB3hH7G&HzLd%g{W6O2(H;9+N}H_VVcD5y;BoOYnnT
zET53BKPQ<?albpts$MgD2gy`JU2XRdcimjdy+>fM>yL1EcS{o4+Ltm?m)%H3Qwrw_
z^#Pi`kbfcy7@vTa&j&QBzW2v9RfVH`W`(x$P(U1eMcG5N`?pnrqe54RZ^1G$_>?66
zL#fI@`F${X7pAH79|Iq0r_siiTV_2*+|-;E4f7Wf#K;qZ$+%z6GobUtcw>t=Z=~8(
zG7*b&bTuDET~8~BRY711zV6N%6l0FbK(A%=0zwNRZ3ovTQtH&klxPL=b;sl!zduvS
z#F1#UCrM^?mrzJ2$%W%C#<m}>$IC#{$=i9Ts!*WI4ppp+SdNa`J$Hl2dg-p@vNQ^I
zk>49Co-Qu={EQWEi%@Q}CWy+#6j$S(9$IEkO*K>N2OML%=hwJ?!pd@AAl=)~$Qr(x
zjPnFjmTc8bt1+JpZ1+acM@hVDK1kHsY43~~_4e!iUc?~>(K@{})bb())nJ<ir>Uv$
zY}HkNk=R%D)x6>=>*trLx~;CH;Ip(^>Ng3|&6QCI;R>d}tZdLOG}1*}BY$mP0TSSB
zD!sUC>*&{P;{_XWS9#7@U%?4NbiVry8K!OdRpnVoo$&%*4nj`18n}4gz8K$wm<L2=
zDOUM6Zm}_SQ@?2yV25srPOVFGEt0AHVj-}JAovej+|@jcpm1yI^=#QH#gu9r;jMyS
zLY2UaZU=ScdS9<k(-6%mieS7=A3S@vtk<H|Nd6)?7qi`0UST=L+iXKi$P#r3D}H=T
z`1-v6yzC0B%h(fT{d}Kld0NlKvg0#6iTMat{C<2tjc)R>FR;3deWlE3Y%bmg4;SbA
zGeAMPn9<<PjxBzWNi}*sR72i0i#+8R0kvmCVQ43{;$W>O*#IuahMl&`n&WbaChXGp
z3Q>qn=Qqx{e*j#!aQU*G7k0&^VqCF(?imh1hIH|CQQON3b0JjFW=;atf{Y1dJxpwo
zL2etIVINSX)fUd>v5#OrM&5ayvFL0VUG&Y%$yUTFz?(lzYy>R<W8mO(JULNJOu1BR
zoTY|fK5WpgY=xsI++g+I{ACg1nnA&huIq=D7GTUXlYYo#-zwotsB@KT=#6!p{KYsA
zVFhNA+EIKo8TNh~%0wM*5U84QhqcCEy*&8C8O)R$Y9)`c`P_$JUy<jiQ!2*JM7f2B
zlG|l!!M9yRpXrLBh)2}@(UvpGNBKur{k{#-fw(F4LW9o`NW}tn1ieS~_920YiMbB6
zG?}!`GRJ$9A-aL*rBPf>?U4K~gyl-s08f;4d!&uG{MRL?me;3Cm`Ocbv|FA9UqYe=
zws`Emi{)88odJK}VHm>%P8D+`q!bsP9ONh*g5%LvM@X@AlCW;wF%<X7;l`aVOA+T#
ztW5~6T&#_9%jcwTB*8RFUUbM@Z{N8aZG$t$;lu@c)ALN@`-$vR-?1`GRZ(aZ?_<5r
z)VcmtGv9A)((zJ#2Ot(*?<~yZLbmO~Wn$}RKQC-_^N%HAcW;o#ML)%esrto4D<dBX
znK<qpLj^OK6B3(Alssc|rX5_;N%S+j#L;{*>2mSPcuBo{?!*}ZKpx8q<*$?;44|s(
z4AYQkn4qG0O^L=b-Wj5_6=|ErpPDBSzn*1I3Cbvq$DTA%*@r<(eR#4rMaWu@P<Vf@
z$aT59_t${n-7$a7i{Dbyp2D_8TiAO^uQAd|nFl2n25z)K>11A19qPMxR>AmkI0>pQ
zB^;qmB?ZS+FnPe}Wk(omv#gm~h1ji?J&T}Pw2?tgj4e`xLxMQBv?Y<5QSnu*U^d`u
zDdKUs>CiXcb!{tbgPoJS6G<Cl5Z$O3D`et-ovM2ai?OZ>1WS6Dc-p&WzMLnV8zAMy
zfjCMmB^DF%#7`oD#k=)mYZHsRDjSNy$xCiEb!-E%E*Yh+UF*qZvU;oOWJ4g9txVU|
z8c=C1T`Xcyb7MvG{FcP~h+Nsbxg6O$`#B)U1olLh7tMN+m7+M4lM;^ih8ED!c1Dbz
z2Yl5>&Z0UjWX|v>#)1Sdh=9(pA?9K_{;BTeIx9<mgb)ec*3lp1W19_^zoP7gVH%z)
zjM)n)rz%|iW~IBAw2b7Hs)B4MS1bIai{7L)Ro%Hl*;mvXSWj1Olb<zicm^fkB#Zs7
zJ}(D~l4R}9_l@sO+ea5fD>1*Kqd&H@Tlm|waCg`CWD-;~U%);^UN<-e$O|>;4(BHp
zAN=tHzBtB(ca-+lOJo@fag6eXZ)KjUC%RWA^i>$D{FztIGiAZ^bsT0q9(M!2HOO#R
zeaBpz&JtGA5>~XMo~>{}u6&XP+i3Y{u%#%(YCKWwH>#vBLM(Sc^d1K4iFB-a6iNjU
z)sb-Cc3+)EMtP)jiS-=3_N>-B&@X}%Y-jRSrG2h$aG8`0z+uVLjwJ^nUy46c;D1sZ
zEFLH>QX*}Sw|{O781hv;_x^S@o8efppJ7k?(X)UCNx-n@WEyJXRK43jVxo!-r6zkQ
z1!ys431L3o_c+GVLozMnChgdOJ_qHTS19?^i;?tl_huDSPO2Whs38mSZuO})ZY#@X
z=bkN5NiHbI8DW@hvq5tRo6O;tgK!ZC-s2oT2?K^_DR^DVWuLPON{aZ8x&SH8GQ%-Z
zG*S&M9GMpW5d$W0P#S=a1*4C0ivgoQ`;GPNkS%Vrw<nEziEAQ$nX9E}T9!S5uPx<M
z$^t$XR@Z{KRyfCjdFRdZ+unC@UY7OX{k?2h3b~HWrUEXVr-P9$H>+EHbmcb6LfusK
zn}Cz{wb9L(*$3sQYe;F)+f|ikC{aDQuMw0j3ap#w>UO~v^;l-bb8*5Z!Dna5=gGRd
zRpp@4jRev9owxxl{FkROu)i4XbS<S8h%x;LHqyStm3wVKo_xUlnV5lLp=zVAH57YS
zmsRg+NGE!^lB1MR;QoC1(i11^@ARgYW#==Y5=OwEIi0D#fY%;Qid*YKD`y>P(iWa<
za3NE5yn2H6I{@`(fNKvr$8xG3ip2C6t$3GsyX&a=F8{EW%X8ccldJQn+0pLzVT26*
zYsBSG!ehXtG?<`McKC@+4HM!YEglv%)?1J_57K7xuz-^n=<kR^nWo9gsj)c@)dH++
z#mda-$h(dmFv{)L?oF9>Duf-z3FcDesQF8ZH=xys^`r$quUYci?}ciW<mF2sc$vHt
zt*!0iVeT+XqKLr8J<k_abc1fcN~=fH(EdC>7ADty=mFLGG#-=Gj+EBoB<VU-J0HJ2
zL-b&G{g;UQ81N(KVcg|EYkjp2^+zT}4#(@Rz<K&TRiMrV!Cj6*#T|MdJ84$Fgi*Jf
z80cTH6;Kk4|4A9M(9{11Wz6v3aZ{$TC2UF99%|H<z%x#e`a#Jd<jens<gvnTgE?WX
zIXxJB{Zq!{dl^#|)Gp0eLX=X{rt~k-(jFhT5;;9QIy<?(9!&O*1BH$sPd}d$^S}~*
z$wdPHax#+YM6-8(*}pHTi(8y^&VVSQ^C#!+<lMY}43kPRCmxD(dfDYqS=D+|IKUpY
zIFnl6YurKQn6lj%2fR$&_uqGKf87(JJcpIA16Z@FfCbi2xQTW7-GKyv!kK|Uz#(~q
z*(K2pLRe#gtKscTT6j*$${xuVtes&5#K?iAwRXZ0`X7!l!G6Kib_7OX?lr`2Nd7vE
zNa4&fLq+dgCZh1CvYE^JAuVst-|mY0>Q*0>3L#+VpGY0q4m?KFUETd8zoDfujb#Po
zaB0N6uR96qqS*)8za&ACKvI=qmX}ekY?8E?frkHrVCr#|cRzgz>5$ePc_bIue#y$!
z3mV+4x^Yi#JjWe&O8!}3{V;8gBZ2)xXb>WkSuyh#6{|oy<GGI+{yB`5f_48uFC*U%
zMYkPQA6}Uz+eo_GYg)zP)FaNVF$k@+&h8_8wGl+Ga77HZ4g_tNx7WWg6as4EYm)7b
z;RmA*h-Z4mY~?$UFP_>=^W(Rj5R)FXB1pkkTg7P8+m)|3(P&QKFONXk6R(GUQZdy)
zqpEd85*STByxDr*tGAo`)2iUOB%9|}`ulZ{g6}VApp&--Fd*so@z}a}EBBk#|58ut
zpsg9nuBZ=+<e*9wFqB~^H2fLB-OMMEGdT7-^NCGzXB%2{%&7_xe{7#46D7IAe}8`S
zJP&0w025Ezrm-vO=x3Uf-86EKTKy0TAA?0fcSXjEmi7IxdT?icb#2?w)J(Jorr@HH
zP@IjVX*)!rM>aYxsKyAn5BLwzg;8z#*K3My0N)J#6)C-Kb_rIjSH0X9#;dwm{x5j_
zD%RO|-G0t3dEZr%<XJoF1#$C3b19f@7h*CK1x1*se`h%>8aakd`5)#d@pcr(o|yYs
z<FY|<$8&)ioG6E*)+^Cfe<Ad)D1oAmb6mIKjG!b%4HK{Faf*ZQjL8`Nz_!yCYXr#)
zkC#~W7&VIaT!I^om{viX6glZHDVVT=V%U9xhvDSuDqnRx?S#D}x5LA+s85=Ii2g2R
z_z&U~d6!i$gYi$aeUXF#V>>Kp=_JzV3(+3EdIY$jMaqSl9MTk7)eELn1JZUk7nBJ)
z9BwTi-S5q1fJozC1rey`fZYhI^;oS*2$zNx;x8+Tc9fBR7>38ng>o$-wp8S)<$xNQ
zc+X@ZV>GxHfji$5VWgxCfFB6EwN02iQJIbz%x(hnSm?h49?uZ57g6-)W`Di#n(1w_
zbf6WkwoXqw)m+oU$SK-)r^cx~#(jPmN*krnIV$~p3r>q}0%VgYX=}B<FDgGf*c{Rm
z4Vgo=quL=oRLl|e<zPw;6#4DWZGlaoEw+ynt-66|zX)Jr%8NM14yLwKnk}7n5yXe5
z(@s|P-7q`hix=gmF*vySFE7bis8keTct(P@!9e1V$IGL{yStW8gs!*<vz%rrRAze0
zzq(bJ_GL_XW>lC^KLz@q3p{%8$qN9jEo;84C$;VmMEwdGgBR?p!|;sZ6e4dVmqF3a
zVsa7^iDIVW&;xwY+2D!dD5Vh$2^Be`+OuhDiIK<;*Av)>b9K6v38<@dYjF2dYc@|b
z$)ZZ!y5*~t^rAYm%401i$X&DT@y@+A^NZDChFNyg^VYXec|dIDDa^`>k0>vj$;uZ$
z>um8biry`Yo&0omJ23(YZRTwk0hR^ucJjuy9p&k59O<J@;JwORqPt;?<g35%mRk02
z(13@1OIws_1{*kSQA4W)mFTnfUYrwKM|QM&ZE-h&T_)rm564~PAb46rr=AaSy(r2p
zuFQ%~Xuwju%th04lH{+<t}8R{7C4Suw`~xF);rEKfJAb2A#1Q6o{JmF$h>SXiL4@7
zOUwz>!eF2e=y)zYu)>{JtD9OMQfhQ^Ips<xTEAAc;wW7#0)-$&9qS()EJ-L36Kzk@
z>k}UuCW>JH(yL5#S#;EVG*~HyEscUqm>3ZAlWrPbI1~Y!grrQgTd;hy-Xy7pfrXs|
z*q29h3Ya4D!uT~Yzde$;rD+R`tmmvp*w|#l1=m~F&5S~+rnfD0#ofF>W>$_H;m6fK
zq$zG&Bs&lAmpp=0Y+goty-hrgOk)VI0*$x^q@E^g4&B}Q&h`irJ$i2Ka3>Qg|DD#R
zqI_ZzYx$Q&mbqdZ4bZDm?@;6xEZvyyTBMQQ|GYKNrsF-_n5m=9rP{u%LfUDrmYGDk
z|B7W+hPYTfp`;FC@o9XtVF6E>dg3dvl?u-!_s~_C<SI|MSvOQ?-<QV~i(pY}JK(Za
zv9{l?GRz=9>NUs(z7N%|yi<FbY$UIvl*$ei<#pQf&{+Xp0-C^>iqypQMicco7Oh+;
zMU;Xh+Q4WndmShJz*KX_h7gTb_H}ZFP&zXbTpnVhzekKCQ_WNqNMlwGs+o2rv+`az
z!Z!rG&X0MW@&Io~hW98)5VGj1%R>6XD-9}Bl>H>pnt@jH3KGRe<e!o6XRqYMw|ghJ
zaZOoR)0c<3skxc4U{zPWn?&zvlELPwcDD=tthulaqm?bgdHfZB$))}HUN{x7pA=o~
zB>k8liOfMibzQuc<0P0}wcQx~He*)NQ&<muzDL39`zvY}D2E4&GebIFMe3C!WI3P4
zTS^#RMB~+3blq(7W^p@m`K5XOK2zND+_(^xO1ree-`GF<I?#jqfqfanzPlmm<6X&#
zN88xHJzpUUUfI>RS8&!MrpuxzfmosJ!ohvk>DWEFlUP1r<I=e<SAL02r(Jl4Sp66u
zx@EZ*=P)*U+rpTQv@jU_HEUUKg&B+0E@xSPNsi$XCWyo1?eu%1%^eN<FK=MD?e6)a
zKR-PbJj?cLPS@WtKCq6n)MjosYSm%o5B4N6DYn$_^C);ir{JzwDz{mBJrDh;<A5iV
zlDDwe)N`8)`Dy|6x)w`q&&43y4zomX&$z0l!FT#OQuF+>|9o!f{^NA?tlF00)Gq=N
z4sH%?%Z<!D#O&p5mhCZc47SW5e*oafig7hjr#+NVrhS!SCE7}aTDBH>*a~CTRcUFh
zJ@IUcwg&m~mQ&5({Z@zt%ge!sX1VRF+Sa7j9B#0_v{j^w0-swF`j=I-Te)24N3?+<
zx}nN0<XBEorLm7rGL=YKw6<^IoVy2WL#1EsVowSi`@ZyM;|WO#tIsELQ-d|xn8-%S
zHl@v??G<EHxw5Y&n-(L;aHmtvtk-1>fb<fsZo6RDK844M|6-2rqtb3^&en(>JL=`j
zC;Rmcjp<ve?IKld9sI3*clI`2vE2b6N+VBJ`=wKgys~5Q<eBmE5%P7r#o8Y7;*?IV
zxyo>?Rdn~oP(=@1<3}HxsK*MKkC8F?b%e&pP;;(p%fnizKh!{2G`UioL}sX3FN5KE
z68ApT?B&)ln<#b2w_Q|92U$GAsDabkcXcoI0%NK!@=5NyiNc(qSM$;K?%pI(u=AAD
zzztImHRF9zdafGpn@PM|(qg6;FLj0`WySgrJO|kv$$wP3j12#OrOWV-d;I?_I#CO2
zXA?&{QELNd6JZl0J7W`GUMMGLM-u}ZDEE!&e}uqIc7&c&wKMo&(l91(fY5-oaOf+W
z$SpK4O>3_UgU?@hE^o;q60Ml*B5)BniTHmGaPf)#QG&2q_^{P5I<n{>)R?4@{pZiT
zHlYVS()Rf|{bVH3%KhilyH#09`Y)U}0gC36(FeOxyA3|eWA|XjcZE-}NK3Yu5F^Y&
z3LnrO@9bU5+CBFo7uC@B7>}OP@7-EI?lI{Ol%3$*{pH`+hmX5mArp`$f<y2I68Vk&
z++6s4U{pvWp2zS3akj4yx3`}=GrntOS=sumA_n2;tUd@LHjKj%I2{RDXYr+ngu%Dd
z)!2Ni_ZJruWn?+K&vbrP@74J+YmE!B#<b#4^2^FqGUBF}@2INL;pb?SBNsj_jaBo<
zLFYI*n<1<bqZ{~Mxv#pBa2Bfd9P%iWgYgmv?%>@|ecq?6u+WypW$OTjfyF3!7j<Lk
zDkK84izJCGrs72844PxOfz>X&GT#xFee~nj!XWy>qAQO48Ilk4qK;@!01A|O(F?%1
z-q-Il1XUsxTjtb0g?v02D7g7x+4uY}KXX3`?a9L=s;)Z``UUf<G_`@s_ElBrYWoqM
zM8FDYO-66o{n)8SRvPZh)WlS&08uXDqULYPc>EL$UX$=<s+!PhLjiM8ORnZTeHQob
z+-3i{nKrbZwcey)-zOTpBV}pz?I<g}p#|l}<3a?!;Q$zj!*9_8EAdNVpZ=xi5Oo|!
zXBTcv9YW7J3dloGxeK3oIF^K+1IT69W<+xx6<z&tw;wR&lcFxk#fR&tuq5ZT8X_5d
zjkUuzP{Bw#UJMcit@@u1jQxqRdHBaOLx1|myYyc`NW~t{Cj4eJQ%Q+e&)Ww@^T_Y3
zY&X03SN1~Wnd!4+6|giD1Zf6NgLt8XJo9}a<BB+jBc4EIaa3+4u6@m@233sWj{bPq
z)UDc)!q`|f4j`NTLc<EvyA<cz*@~**Vw7UrSm{!*LqK%?4X22>3}_JcTRFFAt0NA~
zyc&n;<-d+JDQEYT4l&GdbTeJb!1qJ4mZQ_weKU8P?r2ilZuMqS)sF`aOU4*=QJ3>#
zEa896QhkV1OF?o>mXr8YCc}((h6xG8POyZ?#jXq~Xw~iQ(#G3rTBgOHH2*>D1skw5
z4{~MyERC|(8N5c&iyx(($u*h$gL&VH#4-f9gwXll!D!B|bbosZ;Z6ynx+-@2wxrUH
zR~a5ASH}{u58h3@S*fQlh>d;|ql*ee*7+%8R-#W(%GyIx6~)iv>74W*c~A*M7cLD;
zr?lv@J6|unNKeFV_EbCLAP*DHn&PQ0$ljL>3C$~rVp#puW+SgW0qr!Y(T85S=r@+A
zx4_vJt#`;FV+eTt-5bm{0q90lJ-`C@ah#(|C-z9`v2;fvKx9=JXPC|Dk1atUGcVAU
zv5gPpM5E-^g`MPGLjHcxT55<FAV7(#{6q~@Y_9K|KZy((W|`PnFJ&xb{ZM(ZB{(JU
zNAb-v)lP^vTlM@yY%%^!5O$RF+EKWmvOM|a(1DoIsK_RDCpIvHe=EwtSg}wi{Jy2v
z>R=Ea?Zs+@y<4GCKUlV{&YD_<`KipH+L(D&wKc<JIf6sKxWbblbJfbSuADOUXI=#!
zPCn+ViCKofQU=xKe(qJd!4Ems8aeaf$K&v|TqL$4ui@5nit6Mp%d9sJ;K5stCGIpO
z!!lCpjP>pGHZR}F5GTzFfqC(kvMOC7?jEGV^F|)`G|hY<spJnSdtk+;L9w6I!hnLQ
zB@Ckn>#l@~W9&mU*p11OzJmWXX0MN$bLOm*G2AHk_QgGd*sC}ZgxWc<F=37+X#GWP
z0wA3{-zWB~soG-ae%rBWWD{DD%Ape9-SS}6j8<#CUMy?m=v0G7iyE!1LzV4pqpHum
zEaA^oC-LU;Z;X;%l%8>%dcx9KNcHJVp@?KTg*dn~Vzg8YDB*gjAUo4y=yfXhLtum2
zfxoGAfe8)bv#}Dnf&scq`aTsZ$CM)K$(X%0y&fKBm2r`02c(KXtPQhkc_0VNeSP?O
zbV&#K(s{-o$KkACwrV+3Mc$p>s8v}glO<S_E>=psnIBsd?|$G&eSOqpPdWsf(Iv0s
z*Nj2jypOyGT`T<|6Fvlo?s8{%O8@cdkSN4HkaD4LM_QAorHsqMhuj&EjATLXXidQ0
zd9q7Ss$|QXXbj@~l21ZN+14UYiY-MXdmyn>QzqW%G${C2DvP!#pfrkmQ?1P*N~#Fp
zeQ9?<$o!(C?%3y}{k5_N{&$69iu*A}Nj?A9b^Ve~*3g#X_2MJQ8O9>!$Hi@yGdn)u
z4?hkEeU8(qltt(zL&;Acm>f0RpQZJ`#D&<<(A}#|T(?&mv{bfR{1p?8p1QK`8NrS0
z0BUtJsqCI{%v@rf%i$<UomJPWlBxd?z;^AfF4DS%PL|QlAV)3s?9LNLPJZUlt4#<m
z$F}Hcj#ic3$Y7g>(CRcv4?OwQeGED=aBPN6(_@d>lsk<VDp^UeKEa-u$v4$Io(i`d
zz0;e^n@At;7u`7dGnMM0k@0tkl(woJoe`7Wlc`TJ+j9G_tj}2Ju@YQFWEU!nP>vKL
z89MED4JiqRvH0+^6D{`GMWaxb?YhmU-dIdJiJHf_l%yuz1YO5{OxhB4KR4hxg#yKO
zKc<<Ql91yw8+G*&%*m1HbXvS*!^mkZHQAQ2SDTmQpxW&HVz`Nv)xoJx;TrE5n|Nbq
z_||ewp1L6%{<`iqDQjQ2lOngl>Drb9e_n9@5z9U=NN$2J9MIfuRFFd&ISQraE=qIV
z&Ck<ba%in3uOJur@bPdr(^Ya$Q7%tY_Q4V_&N}f1l7yb|M%LU!%WvbxVW42$MTjnU
zae{xrS!h`PGhBj+<^PD7|GRgef%$*B+#Kxm|CPcy*3z-l{vYqYK-UqBSt#}Y>)jvv
z=iQ$RBB@0#jWHg(B>H~k8SIoApQF^#QX2q+&3bkp_Ts>q8RZb7JRKk{igOx|lDklZ
zta@}qlBCjE$TQYy&{ymXo=s|sgw5URCFPb_(G%t{&VL#c@ith|lT0fky)lGNYU>R^
z7EYu+;pccAPa~M=pa&n=BdB=eNd7^g7`jMAujeOClq*3JH&@LES=dPBe`~ST)dFT2
z*Rc*zy52}aNjm`~1vfo^M1#^kVMBql-rCjDFRVCdM1|bi_(}ttT0>W$MrpB)R}_k0
zZz>KkWW3B2EU)&1mLXBlL-74$?88c}m8Xf$;7|J6j5y>WhceQKv;BjiWx1-xJZ5@t
zTDMXqJ8bkv2j#EIr_q3b(uHdr67wTBf8DC)r}f%P!pG!Y@P#!~wLtYt{YGF*&nIZ7
zmBdN7QyupvOpd-uOFD~tjH{B6%Hn%Bs!@K?EEsnBN{`{9{)mqNY^;jQB}B_^hKmqp
zSVUN-x?nM<9ASRP>W9~Tby5wca`lR-f{@J=uU7Nif;IS5xX2W+X5Z}Rq$X6gxt_>m
z{&?5uT7&7<^&EQ>C*7DE$uT!o)H`zWyBB%p+7#K-Y2p*ZfvBA3$K>{Bp7`~y(TVYr
zUan7+4a%K*<v0sMY~9kwXNY&TL`RnP3@_J_ZIP{NdhSNoE?+fQA&H*-^$&y9a+M3_
zBhs_^t*w+3E5a<oQKT!TX8giemHW^nH{PRS_49lrM{&{J_n%^GQp^On@lfTkr52==
z&g6$Ub8(-TL$O>x-tW8I*2?q9$Hz#Qw4Sau@AeMw{n(PjbGpUWMYT6@G*@n)UQVsv
zKA^(BFy_;>t>a)r;y0HXL|*$ZbI+e2=W9CekB=|CMfAgGv&&WV)5XVgyIsG_U%!_{
z&+nJ*+>-Or)yv$S+>ypey^`GB--nm8<BQ!zH?qszpAXmX>vvC=_lqNIR6k#)?4EY^
z?3sk7MVQ(aA}WvK(%Opd1Qkx5@*eI&WDc)k$@k-@n+I!7jaoXr+#bE!o&9${?rCYU
zw>dsu-!BM1UiDXS$y-*zoL>g)08%0sP!Q*;zn(|4W6MfO*!g&JdLUgt?@!ai@aT)H
z`%D<H<zh{iCBoBx8~LNTECbs9Mn%;x3$7W8s<z2MlW-B51_z}RaO}5tQUxp_I>{|z
zWu~T@i&KsHDWutz5zq`m1V_ZfLiUzaaVf9uLPuin3_2f%l;PEvc=qW+z_WWoK~wLA
zlzGH=|7DB{2n5X)#yt>Za9ZCFf>bFhh$~Dg3MNgXt^1B>kJn{SUXve`|0y^Xn88Wh
zaR!)`Bd0pRZ!T62V7NH~RIWetJ_cNZnzA3m)2|SdLnP_s*8g=edPuAF;Jr{-3VQJt
zV-x#n7X}0CiHtZ{BT+5lf9X10CYL~2kN01$<jYS{(^qfH)#&VebfTD7xDtijvl|NB
zsR3at32mB09Ttk9VoG}3xa58I<o*7}|8>?!#_i%L;_&-DnZCCf-23sr){U9eCGN*H
z1lrj#Gp9v#x5&-M3k8Ac8I_;IOgsOWqxW&SKRI}K*J{&D%b$IM*=g+HhG)`E|M|N7
z$UMDFe5Ra|H}!OM=oulAYjgHqHS<u!^>@hpoo`2TS$cZAf4X`&xO%tzHuQXXPCO4=
zuDv|8%KW`L8bx*X@64^;bs^c^M{#}M9{!wcT<(PBySwet>h|_|I3BLurdf)2D>Ulb
zN1Aoj-$mTKlKbY^E}M(Th2<6lp1tq0<*w7hT&sq;QU!CR0>)3d?_u0g1LLLsw`CB+
zbQ((&dVr&To974U2T(M^BmfZyc!#V<Ycg%q>cNCb@plCbM?zp)Et+ToAQ8mVmk+2$
z+Y3M>bn8TWRYGAaA&(kYD8RBLWYrA;Og;)l0#nq3EEhnAAZHYlg;o!Q22u<~0`Nv`
zyHq8;Sw&jEZ_wcYg-o*GN&<)|NvZ-G_NO1ff#z49I}|U6p{bin+qAQ&Po;RNT~<Vf
z62@%StN?c7FpeUE#1Qd-tXX2F1fAPGgGCf3vxg#L%2z*;1N)1wRUQC^eHP%r{4lN*
zOq?`<IPdxk=YMpCtTT`Dl}(A;1^k!x6o48|ebXKQOoU_t{s<rk*wM8H#CD4F6yOMi
zmSsV9q^Azr-*t4ye!C68EPCaqmKz|1*Aw645-|7xMu}S;%D-3QzJ~*f!0V9Rvo?MY
zlmGDb7H~Z-sH>U-1Dn9BaW@&-S(;8QZRc~D3nF4m_}h@ilh`XZAD|h)u4PufG{6O1
zXq{n+3w5V#!^}LOaw+H!c}q~<8W~cO<SIl2Oaft$@fZOnqj)G>LVl3(8387zcqnXw
zLC_cm4L@6pk9mNe1RrLk{e9lWldIUjd)8@!kcLfSnk!Gr8K5st>w&ljjjD+pq$mtS
zEmTznAHu{-vN5TC{Z3Ec+fhqjExE7h-84J|+Fvjb6(bNRl<5Fe8qje6)K!bL3p)YF
zWw!u?6+smUunI!oo@vC-4Ors<7r6sS11cm5p1rpv^ef{O?=3sCA51*F&k%Tq|IOwd
zkN7rF@5Wq|5Cp`v;pmBr)OrXg#~ex@+Bs;wyAnVbx?28$G$8eMj;Q`^-avImLkZ1&
z9_gN-54#%q)_38W1MmwPOW1mK+P&=WaFsoQIzm}dGbx;?Fz4Alo+jY)GS25a7qLK-
z9LtbcUbHIL6sFI;GSR?KoSbMxZWQ$ZhtD7+YK0+m!qXrmfY0qBP{d7N<~aO_avXX*
zr@{M|f!4~4)#L^oIWXDXCa@RGwET)({VS%Wcr|>4TnQPhEZD{ug}%w-T9|(EvK$bV
zZ{j{w)tWy!D?tS8U&T?k=Zj82tqXq|m%bEEeJSj^V%W6>FlI}m9o7?bqnLjmSGPPN
z2;*LXeVXY&9kJ>tpkw^D;&U?`4!5>rvfbdJ7;BcnYFz$PUjk7?zBN2a6cQ!%cU97q
zFpRx8fH<{iZaVVoxKDPD<O_^JG!FZnrzGfyxXvRqBNTq#gD^*t6d*N_5q~Kn1K?9Y
zC$b<_q;b58Mj2S%Gw|)?LLwZUOidI&V3!40(lJdIf0{JN<uXlO!LUHXqL><1PL#Wq
z1()@5A=zLLXtMf@hJG{afV?LpwPF}jkotW$1H1)<bvypDplzH_p1}zAkU~aBl0})d
zaP^8MM*4sXy(g`sKQM$G^lf{pkSj>prMo3V-}H*T{tDs1j&$C@{5TT`Nq~f?8^_4Q
zU)DHKuaTAz*Ic@zd*s&}Hc`PK>G18N$rJNR>gI?`<vc(QsfAGx<U0{#-xZ+~GT{7@
z6(hd5g#9-X;U4WVAg+s>mxaXf<!_|*;5Xv_E5bk(o314OWNr(&Uf1H*$!o48wn_(V
z`+}nf)!(%<r_G2#z*9{|x%tLBLj~QT5g$IeyYP@{w|Tz&K*l|lANM5wH9F;Yo1IKs
zFP)uQ|NG%@0)6#0bE7&YoR4!CHxW0Aad=l9Io3+2+x2}=(WLN~TU8+aT7kn+IXNk7
z7$Z_^N5FUCMXBDiE)$Ob9$KqKJINdsTJDmS5b-aFFcNwoS^{#3e3pbfoUyZ2A_DeU
zDT+eLtia$-NM$c3sa;Sj!2tL~qXBSJBC4PV>_brunnU8@E|TgUgmyjdb?NJlKkiNL
zuQy&_mXJoJfYORrG?FPk>cmaPx+u|Zu36(EcATrOq!g2DZYJdik>r1%I`%c)xRoO5
zT?5yLj39+S3d_Ro6T*~4>8;uqK18Fr0Ufu#f$Q%MDEJc809;??j{<5d42S-b)}Ta)
za<7@q3!glX+kU>51P=!58<625z+5)m(ubUeVopR`?8oj<m9jH+dJe654RPImWCcZB
zPWf`{SF}2hDP3Xtz#Y1Ljnd*cIw;aajB*J&`-L{nQZf9GMNQ`a(7tD6`0pUz%bJt6
zRBZ4&rxag6^YP>tU?5=Dxkmuo-y*Jj9h4h29p<}!3X%miPW9P`PYw2em<pkv<B9kb
ziQ#*3kM=hg-@nf%Ge?4ucgVrRUiIK3bj3d^NqCx0_L7iPD+fPs!+v|kAya(KKJt?^
z7iZ^xb81fBs*+35CLZK)b8#zDSI}sSqgmV0gBLPuBh2oW7Cguf-O=%Gy}!=C!^p-9
z{5&2l<!mqI<D5loGjxf@4a9Xx-dW~;Id_4l^-+3>DPSam2wTN6b18-|V-!!GONtfL
zY%oKm(C`9s=TUdsIC^7e4~Ud9fd<riHkgeCP_#8~Q+y%B^QXIJ7BizUxK@$?b<>ZR
z4YVlE$@VD%qvJ*<TmLnmzt3Y!p{UIbtl8DCgW-Pe9&hFDK`Du=puqtUQSg4gj_)oP
ze@_%N%pO_N^%M)6QdSl!5v%A&@To1ac9b8nmG5}o%&O2S)%WN?6NH~<b9-@1$oPCK
z5dbLz_Du7aSKqL!TAv|^%sX4AjKhbq=(6GH)+t#o(RbT$d#O0(Zgn(G8NQaLYRbE9
zD`3s!o$1EUtdW`0l^EE&s<5?o1Whi<;)S{ouf=a0mG!&AAgI5)SUG`v+&>OufFWen
zP0$;Bd5hHemq_APq5Bi$Bg7J1^lWKkRSW`AR(^u8jjYo3lK{$10WE;*vd!JRWf5{B
zb=o{vFUeZ5T$D94b!1d7IqM6oeGFsrFr2?8)Hb#GQe@<UEag^8qMuW0SjER$+CEHY
z@gfIsJjO!BGz4^NZ94TGkw}X$3qiks?cgj(LwipuO|3YGH>vica0uQkqWy%}nI@o`
z{M6W)m}>^*Hk+Z}z^`thOi+%K7aRLJ^P}Plmuso-8GKvSy)-Y}!Wf@q30U6gf*Uke
z{A9>8K|b}#g|LGcf9FcpV2s3l`}e7}(MM3g$<wO;aBc)MaVUQ#<d?;GLMYbs>1jWb
zT2X)7z5Vv@bnKm=5jL^AwJeZi>o)JrLOGiIH^h4dZl$t0Mh2KOiFc-q8aR`oU_kw_
zbV_6HGwB#6KyV9u0P0=W1YUw9uo{{xJTEyQ0$~G0BwXl3b&WsJ-iQ0{hY(U-Rlj5`
zVR?xADk+b5*Yl=!*+Apx>oUfKsigA+)!RIZg*;aAX66|6D4j9|fw0y#mXS%^Tws*o
z*QaTwb3JoB|4ma_O6ck{?Q3{3opwSkS>@869c#y{KeOgKr;AE8b!Katx0zb3;bUZY
zp-Q6xLT_<N)Umh~>C<GFfY5<TAt|pHc^p9{+vln(_$nP(ikT&6K?k2Gnf)#(Cb6~_
zss#g09u?P|ePmfxKEV}jLKc%Hz6dANTQg|J6LDR&rW{{Bak1%QmL0vTtUj(!q3SZ7
zHR_#0Y!S<>0`np3z3>=0tMcf2)<M2D$oDf7Avo1|>g@<K*IpGA>w2V%&M!snI1$o}
zwEV?sO+`9o9d}H+pgA2VJ5Fb<=h2M1QEfRbufC~s8=ScyVM=u%!`3*xbTxS<YSD9h
z^mP$V4We#!kQz|B-g5i7)AmX;%?@=_epYH1xeeF~j-XUj(|M<Ds$Dm-B%OuM)NV$y
zBG<3r)vxg|fx9!!1NPZml!j%b(Le&#8+U4C(131C{?nFkb3Zi<WJ+}>1aHnueEq1h
zS(W}(zE0Y(<DMm$nc;jIL;EgpQlO{4K-1*IlQ$M!J1W}iN%IlNzb#PdDMaSHxH*Nz
zX{pw%m>4m49V8X2-F$5QyFM@j+-=?F!kHz8(X0`9gKM?_s+1KaZFKft_P-c=rx@G9
zu4}NYQ?_l}wr!oVZQHhO+r}x|wr#8H>oorK|90{w`(kC^tn9UuHO74AfG}79nxBU#
z3@d{9jm07RPD(oZxH$DRlgz61eR{^0E|=~se9+SEtmdSjsO(e?a6@)4Jp3t{KxaTi
z$>oeNUT#-&jr4(#3bLJ~xalz^6k$3TcUJ@u0Ysk02X&x%Y`O5!t5tEw#^~uw$2T|F
zEZ7qGG?+uNZ-P#@HJWl18B9j|NQRZ_HXYoYlAqRL_!^#bDx>u~_xUZc^(X!cv#2@(
zBvd3z-DdLXDd?Ui$A#89aKcmk(onOsr}vQKCSXj1QQK*WziY(b@v}54IUF0-!l0sf
zj0NJgTO;t@4ifA)uEJx<U%j82icgU;n7D-Zgi(p&8Yz}$l1Rt>VjpB0pR84|`=?30
zA&=_yn@z#(fG1Ba_>EJQ?w|6>b8;62er<csr4inqPar?R!j8(%hTNe5VgH~tA@^y*
zL8B7OO_Zzc{vcA=&6!o>(KFsjk*08g_VU;>$qFAM88}1f^*Hbd%<Nk>%d%=uO!|u`
zRJV~HYDvtMD4=S@^$Lx$&_t@EYCBv!i5k&e%Jz437q*oE{5(5K=|At(h$Tsd_?_(I
zjG}FMxUvJt?n)reSdgAsTe2+MYB6G4xHv?YBclrV-tm&9EWBCRp#J*)hk;}9f<hPy
zQ_7;T&wwVva}=lUi#!%af%3>bCAc4&SdP~Xh|$8~6&nm#tiiLC63rg=KO@``&o<og
zIVL(9XD_P&;xtJv57blDGE6$A5s}`Q-N5^$^n+<a)HCvchbRW4G&aN`?N0@LnT1k+
zcc<ju<!6fXEl73hBaGWHcU_56Qtp!EW%#Qu>!&<G+4GZf!0Rl*nHrMF^TqOTmdY~2
zWQVzMD9>h-#i-tU_~h67+W2%}RaUeZ>quS|WpSA0CtT=lJXje_LEAFe4brKP|LD24
z6DKAtRsd;I{;+2kwR7t3MZaA$8x!pw-$e9nSvhIPjzDk)6|#EQBpq#j-TJkY25}Gp
zwU(X3UX+W<_mpeaEN=}bys!de%Dp9LDwhcPXM$r*w>+fwKlMqrBoUy_s?f^b;I<~q
zAyGU4V#{uzx=WVxrY;P8t9#c}{r%~vn$z`NVDQPMXo6m=mF!e#{hI|#XPnL_OiosA
zc~Yyz@?&vV?NAe9bv`~qVqrvnBw6%C7hPfU8;)jU^nWd4f3N<>#KXYA@}K%@nHc{&
zL(h`dM9c*%;?~RWfl~%S<tqp@u=mmm+9<8SO+Xb`qJ}ZU+s_d44l>1<#osfN2KkK?
zVb#0Ez*2G6-p`W(x*gpfudYugF?WVZO8wpYyAhFDVA<c|l1xpV-{O+OHm^_jx2tgu
zvyPrw5)QGViCI^v-1Fy%fq5*!QmIYdXylc5TIn|&<$9F1O-}C!;{N%}-`<jsOhw;?
ztC!f^zix`%r;wvbE@;-oMAi*C;xt81sj23tF~l^7*^VM|$nw57^?YFCnr(KZXANby
z_h=I6H{b8kbqxjxR4~#Qa~JsCuzwS9zdzJE#2XavbZ7sOR!e)4nm0wr$B}j)Yz$5f
zlB+0hvW?Y}^8g!9r3KM2IRJ!9BNOD&&du{Hm{*=Dals;l43TNc8#$6sF^^JH1mZfW
zLBe)g%^td8!VJy>w#(J&^@Y<5>|hM<#L+8ZFwymqM3V8r7qp7~do)j^UL$f!mu@hD
zI!xi9Js$MzytpjcwSlUYspT|hBrz|j&s@Oe{<*9EczQZG>NSkA>n~bP_5FI=#}m&_
zW4#9I%Zf|w{q}x4I7;nVG0OgCGK$#+`fX)7rp-%NOew-};QY*=<JT3eoh`hjW=nht
zN<B<{CwqOloUUE^L-CoORDl=dRvTXUpf#JpF^;732iu7NYv*7V2*fFN$F|le)6)6Q
z?oD4>D#Zu>YHf2IKd~i59x5Av@(l}P4-?pUYvohRV92c*?nyQG7$DkTeVq()&496q
z{AvkiTq_%kIb-};rlV53Co|AH;%dM(xo4wbFi8yA2$e^vjrN4eDZsyxCxr^P#RP$Z
zTFIeKybqUAHd|HV0)v%NKhb`Ce(D&#2-w6fiT9yd<|~;D(1E}IquPnnlxhp)QtQ<A
zR-7gcNR2ryBD(bvtxSUQv6?THopppP%cNvQZWpbL;llGiY1kl-d3z!2Mc}f!<gmfF
zh>So^V7coN-r(RFliKEfYu5DnX5AH?&OEiWwLF|oX`8UuzdDUF!?Rt2E!|^|qQnt;
z1lP0>x?0Qk@y$(=SKqqcW{2~`@%E#&N@`XaS8TMQS3>Mxl}R5;tbedMnCdZFSX$Vj
zABC4|JUx9;(seRtd=Lc)Y11nRq^B|rIpAc5AU;py3m6FCSo6VQLP-L2Gp%{Vww2<Y
z>6Fe}i!rv|>@m;yqY|w_N-?A3KR0Izi?m@#uKBi`jZPMP7YVKyu2?oTbn}AovQx=T
zhk4l3iHb<d(V)6ahqfeMqj-{x?QN;Bk~qp&-n7LP7AlIq$zd;TGf9ay%C&>YNc3SS
z`$(uk`jE`F=}><R#-Fe5G1Xg+l@sCKJP;&M+1(4~#-IdNG|YB#^SzkK9(DtYKwz=@
zw-t*J<Y31dFt2$9O(-v$Sn7=9kn&w#<4lslf65(JnM;HbiseT&Shh+c1nmM@E+xoM
z&u<M90S2Pb#)(OXmv(Mz!b()hh7U#AiXjacr`mMbXsn)wAvLx2DP<~=si?jJ^AZno
zYTDw)<g-Rg(SJ98%RS`$7PB7WJPc|Zxd^`4dx97X3d)6^74gwhFBG_EP^egw4jDv)
ziBFK5if=gAb44Z52b<u#lcLv^)u%unVdx`a+bD1mpbJc0S$C;rBLg_pEWbL48QC_O
zjHx*(%Pu74cd<#zOWx#8L++2)IJ)N&I?F?(7;F!LbP!`~p<OR5_FE(zxvB`(HiOl<
zsh_M}C%4b4%fFiC20^#$HP{tPfZ+o{Hv<h;7{rDQRwQpJ)o!iZ|7pQ&oqP0MC9{MR
zuax(w=w6E_-uC4x&Z&5j$zG5(I5aJS#TXrl8(X$M<9Hl4ml8!#Hbj7Ni90Th2zgDR
z{!NLf^foL^&_>|9EV_C)U*0F#7BTd0l<qf*%(A!_s?fFgtP#nXdh(-wJ?VtATbH<t
zD@ila|AL4(NhM-hdu(D&ZjnNvKa@#QmI$qpF4AzHC?-yB?uL-09OR$gw>ecaT!PcQ
z6D0`OUi5GZxp?u@WOvxr^<k23`J~!ZfVuE6$pAe_n{ycsQ?9Uo4`1pt>E7Rg|LU-7
zi(Kizj$6SsO{_Cl4OkWgjI&wS-6MOLt;pmYfv-=!57&S$A3zz0_5d^FA-5$~D*H$w
z%HrxPSPMTT%8E)s{w{fkZ6dx$&-i0DX1z9Fy9)g+ubx#Se#BlWb<VOpuEhbxm4t-)
zK=MKm#!NwO*Yb~cJ-ODU#Y{E!vM1OM82Sz07+iD!Y5?pb<+Fz-b{xzm^{5>D&v^(S
z8!&D~O0vDP)FeFmO6Osm&_rUe^m>)LjQI@PG)}GAR`W(yMn%)w@VtT-B!!H5xd(LK
z)~Ez-xXFtHa<}BAYPI@3d^b$;s3x>W@04j-ii(~VOSsNK96wUI16j=AN|E?v7v2aE
z6(v^Ki+6{m*mkt(xedLQFpaS8Z2F33OWmFcB@9Enq2Uffd_sRxL8$Yz`7;}AIYAf!
z$AT2b@Y0(2^63r>YHn*8;$=~m`U2qa0niMdD_})NW_N~Aij8&S#oeIWmlwY2oriwB
z|CfT8MvyacbI2a_>W`}d9XV9ogfDFU4Ki=|s;goS{&RQEN4WROeP`HLyP2`U54UFe
zDqVha?5Zox530e9+OtB3h=HxL+-3APb-h@|u?(FtT74<Enefg%L49jv18%&>2s3m8
zm?GYtnS^&AQT(rn;P<I{TK?M(VWL#huCpHNK8!r<k>v5;q>|;(yN$dddD+gGuU3D6
z@2=LLhRy+dIW~&&NtmTF$>LUU&6Pmj&HF$I+I}BwBdd@FoPIKx&7M37t_{fIPM>jA
zl-+;GPC|O?f*;M|aflt3K;*^FlGH-xf|Z?O{eor6B?t@Bk*6iz^YJD53VH79=aH#u
ze}3(_q0U-N;>}08Z1h?$VR{nYyC=1tE(6a4NJ1Lh&1M`Do9bDdil6_xBj%dL^-qv<
zRAD~(+O^_9V4ZzRL{sr6w0Nt0y-onkg59BM^2%0H$MuFELnwJAA4}dE3)79GnQe*5
zrk(1?SNIJ-!s`MS(DnuniH;zF<oNxjTCOuA-eUd-P|soJ>pwCqcGmxQ(=2UlW9npv
z&%p8PqyBH?azmP9F<5Ny-7{*BaA_v>>e>JTed2}~qcr`S0a*d7WO(~Od_`@nncd<%
z3kTR99>g}yJ(}bkUU-MMPiOb<k8dF!bpr_MEL}cWe&EDEpkseAF{%CNAqFpE_hu54
z#%~?>#H7MUV~>7tde1%~`|qIX-~1oQD)*SDRaO|mDXJBFs-|s6$=Gwuvy=6EsrbC0
zcHe4Oc5;kJb)akqdiR!m-+ym&qhu_c|Erb@B<|b0eTgll&+`we_nFmXS00@44YB>T
zxpt;BcDN#E0;6U!hnNwv8iT}RMG5#Xqt+uZiA;k0rh!MDzrKBrjpE}ge?l%mPhH*@
zJ^Jjq!YjSvCBPrRa{uKk{TX!I<l+9CP<J5l)|7|eQS#L+MTndcUhV7SgV$mE{i<19
zN)16``q>wqn;I@SjH0XlyQLFHBi8i51g_<jA|3bJ5xp)!nft^01*lo97Z@iP!h>9_
zL$-5UlWVI>YmP1fS-uq2%l|=Q&U#+7WI9VHO|IUIH??eGPVW@~rslGO6n>Qg{ZWV?
z`&Dz>P$5FE>!<t`5n4qB6>o%ST4Z@Y6=GHSVg|pie5}KWW8Dg$2jv1f0|hOPIKCik
z?F8l-VOuP#p%Y~ab(WOuCQ&Fanp-I79s>FJW~02^?265mjra_~6{3V#8$V-3$Rvx5
zAf!42k&9?gtOv?J0G|CNp^Hlp+}nmCJfEF&#FE%Vt!<P9z|13K@2fOqG~e@{f`_J}
zDateqiXG_&Z;v0fE-3em|15W+TxaitjYx`>>HRUXqJm)2LPzgn0?m0KAGjR9|EckW
z&3y1Kb+R*&Kxj8GE3K&VwL+NLre)i8IT-L_J%j~PR*urbI&_b3ExK}VjYzp$9EFM^
zy^=PZzC&k4!ZTBkbN6zoa2|Syk>!v4n9y#N$e_Ym@@fKBxp9gs0YzcBCsGPxDS}fd
z9SJBl^EaNcxzrGFcr0UtM~4b=d4r@GTv6Evja9j2+Y+~vY;WQXrbbt0&Yyp!^BtAV
zPg1Wg@N8T{^RQ)xR92MviHjc)Y#Rtii#WlaG{h-JZ%!9YjMb{eVN5_bF{8+yl8lo1
zDarGW>}25|Ag(e@nb}O-WPus&L7tCEpoBgJgm|!AP26rni7_ycvkhYn=TG||*@y0t
zSDj@LsrwN?KNC<p-Q{{8IBWa(g-Ju*4ZR~CQ2b0h@dc%NOxIoDeOSy9lV&w|@F1LO
z2{(?$o;dV>m&Hch3SW~xj4=U*I3^*?`9Vj6!b0;9yNk6ia4UY;BpOAG`nf9v;9ik0
zNjVh@!0kY+-f*kAwKf_JbnW<zg^trzd9xiBkdVW8Y+yqywV^qr7?qA%gv!#w)4BW0
zt(*;Me{v;qEePr<!wVq!hi_ReX#mNsHdO~I^t;OV&@EZKQyI57XJQB1qCK{rbA|_;
zBr@0x;~cWup1insq!APV9?4@7R5Kx(8%^^AnMO5c0t7VyLSLwxl&m)4o~cU|eQQS@
zZw8uJ#QT2OoB$<93J#8UY5g3?yMn)y0y_aXZ}&&kpZM`t7f|X{(fz)Ys%!iV<5a(Z
zj7x?j3AXKef_)5Rxdz^FMY(g0oPGCEiMAm}M4VcIru1o`t9ee!@a6|*0X_Ppxhz3_
z9kc)^1%-%a$T3V>CN!5<&7ac)^ry(3VLB?^an);v%1UiZ_tpGYC<*k3ztoeC#7L6H
zyA-ij*(k<}?PLK{kb-Eg<0BDGw=li2csnzi{7tzt#+bkpw<%{G^&AW%_GS$KSS)^h
zEne3xhRy}o7Z6=!zhwcDP?Ozl)%0#cz*2xE84S%y+lL;7G<WSNk%PQ4GBtbag-2VF
ziG!=IibGF~N1;5cBbxphU9;@w49ArF%F4QX=RL5l2vqTt=`OHxq{?ntZm<|v*J^d{
zlBuwHPytOdw68sP`*TS97oq$(Vh9DwRIx7}ZOeoF(MRDkyrvOgPxIXH4}q4!SsER9
zYJ8}oUG2(1qO*(-jtQT-Z;%w>+1)4*h|WPat%r{Jm&%~odK9Itip~AK_cY}p#^AR;
z*8pF|KL3~kv?&U>P%^l;+mQb}q@RS9A42{P{y%RG!*Re;VQd|;pbJOq4k4EmNLwi+
zA%erFz*H=n?Pi$%axtVQax*y{Hop`P(27nr%Xer)vM!guq`BG)m8mq8GV=Dq%@U8c
z2VWg5OWCY;H1kwUHOa~CeE0tS3h~E9Xs*H|xHrfJ^wW3|t}4U*6~iG?f1YfII&KXU
zV~(B6xKdxARGWZpgH*rK^0N~N+;MEJ$P6}V2EFA=gB6&dVum?cIUXzcj#B+_?2?FK
z-~zKtNb00kTxnK|BYUIqAfvddzr+EWb%8pY909quzP0l>b8{prs;Kf(mwABf!&~2}
zGCLMD&y-Kd#*{7U%h78aziN?uG$E=>=K`3?>Ui&sHiBI0&yCsj>{za<?^W-4LvP`N
zZTX3y_xy)o089hNYth)sD+hzYl{;}3ywLw#5m9;#Iu>QmIq_q<*24J7x@Bc+U#jDc
z6cQ38Wlel#uX^9qD7xz(GWx3IPN>?rqUdc5T@H$%n9iS0{b(rt?s)cuQaO=$acSy1
zg#D1Be*hXlhZ+785n!cf`rnBFD;wi~MFg%iq--}?;d?&*M>oIM^$HLLYDI>X*NWPM
z!h<3s%fbBfGn+5$Kw_bm(xvDzCXkZ456y8CE41N#VfAvf_x7}Rklyw{3gLFC+v}2t
zSM?2A87RvMr-v=@^4|UNFsJGC&V4bsjCgkN^pj`H;f)W$39Rk2``FLKt$VyLl`h06
zuTQ=B2IC%_$K-v>dgnPOrUuCS_4K|GIHlD}+VN=X(cH`1%lr8u4=TX8Z*UR^L+I-3
z>no0j{*N7x3*mssiSPTYhxc>afyR>MM%Nlp*?O5HK74%^BYu`cEToIh0trP@8A3ro
z1oY_p<t(sKLZFBU$X)e4_dYOPkb*<uO(318bTV4y*ENyie(%mphass8)?*&i7|028
zF?^AWx6@&A7)z&;h#p`u+uYEvkt>IEfDENG@ef`rLgqe%oAk)DS8<80(z)A7c9uS<
z0Z6WBhr0Z|HeQ@Y?bS(1(QeLfD8*n6YK26f{<7j=)k=|#TY}!F;$LRay#<K<6p$1n
z=&4d)lHcayTLQrw#m#VuG09iGJ6j^c*3^h&Q>4F;xyKY+mQL`O;^WC&W;=B$K)6-d
z=0Em~;@P2wgseWkXWM@x=XS=T@yjQ5j%Z?7=7clXV!0F0I9H7tI55c<nSfhTIY`se
ztWMi;nC0<Tk^r0YvQU;#OjnPH?UXP^)~}rRCby{Rp60qH9_5(gyg6m<@)m;iC&D^S
zu=HLMEWy@kmaezI1(Eb`-Ms!mU~pN{&$@0pnUiJ(p{i(74HYrNIqp&u*h!7W7^Y9m
z<l83?l;^?ur<*jIfBm5SGS_I2rR}FRh^RgM;+S;njtbUsA1g8o^41qF5wd!87Nu7%
zp3hT3zm~-xWgC%X9wSBGu{HKQ6vx|y|6yo7P_HSFS7^bT2o*ym2sNsTVNi^sj9?;a
zF#z8~;_lEkDHj@yl1l7!PcN)0MF$R@DrI|yDY@rN(+U(PMU3Y1zuIedRT{Gz;KLda
zF31h=2N+m6Q8CO4e|VOEvr{9p7%dh=M@6cWW5zI5S0x*gz#{4aAG!_;z^tIKm*u;G
z!y;CePlW>e%Qp@{5iiPPLkF`L2*|*Q@e|0Nk<xGl3ofFqkdP;tRs{*3KdR-y%SgGA
zKU{|h&gEr~wSYp2u)xot8xbe9b_2aY@2{RHdW+e3@QEtaE(n;rwgZdiqc`E|O|11O
zH>w#bC!!1vA`|Hz|DB{_8Ctoqzf`x60<p@6hLJp1qiWR4;EZr=B+c#kC`WMsH`w>;
zE<`JJUCHu+2(H}Ad2}RKHIM<{{lrlZuTJZ0B2zz?O2)&D;88!<NGUtA`h4)=rw~RW
zVjaK^P72GO>)IqFajxQT<5DE7{f&eWl2lr{DS44O4P8R@kT-ko_Vby|=k&6oa1rG@
z6o=EW!#6QgRrklhj1ZahQ-lv#RYVshGW=82LOt_rsB@o?)S;NzB8Tvtkk@~lq}4?d
zRq^W1UIfRx?CU<Q@6=<Lw)jc_Xd1DtITH-dO5uX8p|CSCiDgXUn|!yappm{`@PU9H
zj=PQlC!h!75pe4h(x-}AxC8+`zyiV=)f3NCbiABtj7w_l^U5+VhXTK(1EEXTzb8s`
z+BXmZ<qGtG6O7_bA5dqBW-yP1*8wkebr%tY={11H7Uvx7OgX6#tOCaza{ODG21V7t
zCMc#oST*ly!}sqBR28{@*yT4n&{D{sqrg;avUdSAE2-*ueUrCIU&_~{5k9GcaSn_^
zv8Or1mQ9l38mGH3br5svpH|b98BCv5F8HP;K{*lLI;i>Y<1xdG$VQME{6Z^OLz%<3
z=;0|V1<@3D+g4H>?D03dk7{1A^joP}$uo0T(R4|c8^E%kY(A6PzGMZhHO)((y^97*
zEHS^MQMnnW8!ARiCE@vY@Jj)1A99umq<Y0z0{E7+Ek-wRSYeB*YQF%+-gy$hqIW-B
z*1?-g?&O8K>+9r2dkW)Q7<0grarL*{vQn9;28&%$Y*?tqCw|;D8+Sa&)+>0bUt96l
z!zMoe$e~4(XLi$0{j&uytY(y|Qv<q0vBoBd13Dg<5}Uhx353NgyEAvG8QIRJ1JeVX
zX<D{5S`RYKgWkk6_Vd>+C;!=lV!k^N!YX-R*REz7{aCjVVmi>CedA=hdbrFq1X$rM
z?F;JnmD0n6i`V!KI_wI1(0|tc;Jr2UTy6fv7tXS1_8(!5;Xl9{8{>Z&eDFKp{=(Yt
ze0vHm-58;w?bi=*-Lk=N!*Gmxi2RS^4gTkcU`Pb38G31f(Lug4QwG(tRxQctMXt27
zRxWj=G?q;$TPdes{};u8--IT}%rKfZIRK1doND<x_;Q*>sa8a2=hxBOv%Fp$PEz5(
zx+Vd%Xx-j+A#z6A-L6j1rz`0nD<^QqVRcP7Z=8x;lcih$ecm3PTFG?O@m_@sS3{;b
z6$hsJN-xLY%j-S~^QS(6t~s_C9EQ?*CZHpjYV3P9b-rZ>Kt(WZ?srPrH+WO$r1%V1
zhfpgRQ>}A`vVPyP^JzGzqgGH~?~i}8ObJIM)??Y=ESLNBRQMa?IFnQ^SXnf!#^h@e
zTfn~gZ>}x<kU457&^vqS3Ic`-in<J-@4E>S_PtaEJLstx$~VP;pT-wbxw)cEoXvwT
zd0aH(wYTfed?Grj#d^KBkQLe#o1DscdJxf8&jFa&wiPiz_txp0N@@ce?e|P)M|K<J
zX|Y|jKq5toMTojd+SKtu#rKz!Cr&y+9G7dY%)89y>Yh}uS<NOnb-zr-W`OC0$CW*X
z39-=1KRr%Uk&P9|U*|BMh@FPSpiv`-lvyB!&RroyemMd_x+v5%@;BX>h)|5IRg?o$
z8ont}#}QSu_c(DQ-%SWCy78PEqff`$w&rj*KxrQa<H$ele$niqjg|A2-@noq&8HuX
zgg5WwVnpdP&@b6J-%*wWrz6+bbtl)Swj8x^@kLoP;i$QJ(s4p-=#mRoRD>*Dvxl5X
zlTRekt##>ETkCM{7@a4(7usC;-tT%i_VB1N-nNjgz;FyxeFCHb*h}|G_mOcypqub+
zn5rRZr{1V!)z6-KOv5S9HgNb-pCXFnhy^chfW5GY96Na+Nj-8W_`p0y5FD!6EmRC;
zG^+2wCv%dbd9`T$Bb#+?l(dq)kh=c8t_?XpqXc+gQJQjjiAqxlR0>E&O;oATKqx^O
zgsKjm*$p#7#;~s>fK$ZSzIDQ$EKHL?%MmcQH5;HgFBx>FIvfm*`$fVSH!Cn2?Y$w-
zsI|NYtSVjp6y&<;0?<kxA5c?aXAW3=y%pdo0x|4QjfPYS-b5Si{030$sn4`wgud2I
z-AH>V2y_LCiA6KQf~k+wX4*k0?dPN@VjQGLYgT%hKHqA!9E4+Et1b^)XI&5d1q86*
z)>hu1#OvXdYcR4cne@5Ut;02ES0Z~}oTo2wqRp%F=>+_=KQ_j_U$DXsDKX!yE2FYW
zR)SQdXIAFP^N6A^8Hed$W>yf2DP1u#<7F(v!6va6^h#7;`w=SM^*_dHo>8*;x0Q^4
zU7Lr7DM!?6x$7D5XU1dSUSfv*s4MkO(OAIZwH;z-h9FDC@ek`;r7TI`M3wTo{b<S#
zxhg5>`y*ynJ1F#2Y5N8<ff#|-U%h$y4XGK<z_0srFovdw)&g^A$-!$hZ`lvizs@~i
z?~~k^kk^BYA?g7tLE(n85w;Zl677^lemm(^s95cgB?jrNU&Gle*7N`83)x|XA_ZoB
z*?xujD{>q&d&w-g9v3_1^f?d7EZKIn(ti(gdwtP6(rSb2V*2FDzo+B#!#e1aUXq+6
zERs~{6=w0Jj92<)3Wf+m)n_woXE(}Ogz0p~y&v`w0{QR*MD|e6p{=ce%8yd(^OrSi
zbD9c?mFRqb{P3mBn{Fck!ukwR@(}zdm%NlmKy|9H(bJ9!^E_w8NTdeG)S3;VTe$!P
zxrbt=gt+)fHHA!6sc@kYX1P$@xGmg&?2-N<DLb$n>A1yBI5Zc`QK2GoR0^^`pqKpq
zG+t#Y*$_Aewjp63FfV?hs?{GfkG2D3fHg@|Yf5|8F5gBl@!?qp!4u&a<?rZ2_j_P5
z5M|tL#~J_u^3{<Xs@E|L`SO)axjrfAw|T7R*c+X6C68gncUNn4@r-GCNGDaCSl54l
zo#Oydm(swt*~Qb$e-tWaj=U|uxzl&lR9UEyA9>K6Ybyd$sVyo&#fogHtNQXF@_pO`
zUftH>wT-3AD6W&U6$2_!tX<SoL!3O~+;kPHUChWUmLd`(DIsP}y=CQSImC%pM3Q%_
z#juP;4HZZhsYF;7(omNeZ_IqQ$eRMrf_5bTQDU~pxl7cSFO!O1_PrJjOk?bLdQ;9w
z7wDixV5rdZSdy!7^7az%4~SK172|r@nTT}5d6G<o=VEr|4?nRrWNbGW_`MovinP9J
zz)85h-{V6!?F3-pl4K2&QF~AIKG9WQt>cbL0FR)=EA0#@g2n0;ywjrm@G(C{M*b%(
zWcgpqvHt&EMfCLlkE@8Ck>kIb|BkdQ|HoDIb^TW$@9#X&zQd?{oCrLAYzItUUPv~O
zHH>%Mp>x6F9OgWy_>cF@?X{v}lZD#0N}T}c>i+go_rtE;bb32-d|WUxM6YF@An!N_
zG7}yNtn_+qa^L)WSYXlSjhTSDm$K&U`ppqym5iKuzJ>eoPRZ;#Gu)1m^c?}Nf>lSv
zpN!H<?9n1AYD9+&5Qbps;+d(4eAx)~Q4WENL#4*JiTx=3K_x{hmP5r39B=i`^oVxG
zwV)BBsYH727KbY#?w0DPJ^9-v-X!@5*>y8=$#N+sw~9xy-DAaOK7Xt(Q5_u#e<oL)
zAfQ^OIA~KvHuYV#?5y7<k47pfVw=JW;Iv<dPKLUlTImL*HeY#+3c}ANSTR6hM}njP
zWy*LchgpM9W=bB1%&^w1Vh6KX^^k?x;l*frPO~-xQUquedl@mCKT?lN;wNq|Cyi5b
zo~RBo?Wtr%ah^;kd7)m?yLC$75u*%$k=>L{nmv6sc1RfgK~0Kzon|LX*l<RPNMKH~
zDiXYYP1!}9U^0bAz0;Bv=J71z6PcbSkx8m5jH-sOdhlFuhe@+Oxwnl_7b-{<g=*;?
zvY?&NL628No>hh<o}?LTHSy<ml?zdwXmxyh9d1O?0<>4U!$rc^&*9v;A%wVU=>_gg
zw5D&{M|jiBrOEMOtla*ODF`<5Be+pp_Yp7kkj&!Yv7Dzm*+=IG8Qz7F-34QZZ?msf
zsVxsTJkjEan;@<EJXppN#q473;W}_RAB=+e4DXlI^HuN8jA`y)UD}jJHN5GR@3Yta
z`@j0nL;L4bpXbhk6%u5Py^}pT-ftNtaC9dp6Dy0%6<qXCk2GTj6k9ghG-Qo4z_{NQ
z&nH8zk!xyuiB@%FClz&U3|?e!FS&asHBgzwf*bsn<s4$MLNpAXP7*O;uK%o=8xl=j
zt<t=CaeaEa)U^e(d%Hi^f4uL%9Nxa)V;k38(I#&AvW6apVjJ;zKVQF3r#?&9*h(jE
ze%^iFo{k)Ro}M4QaDCm_xIA0gGqcIeO@ma}Qrf$g8e5mP$GNkoRCcmXlsJ4Ny@YH{
zR@6=!7mn`F_b&$>elwAMU8^)>W9*5dd{{NUmMrDMY)&WEc7E>U+++<!4%JKS-uAXS
z6GKZdN+2B|e+|JI5yA)<;r5Z`Mz1<Hd@Pea+Fe@O-5zjqLHle$Yk$z$dOM)q-(O2}
zQ+<7!xpk(2WzcfyXj9?uHycGO5|d7Ha+w-be*C`HfxCQ90W*g9yq?!ChnDs~m%QFQ
zp7$$H8f{Pb=;Ulf#v&JPj^!bw9NW9D!#zEQC*Gd+2bbPHJwS1Mt}@>;BNymU`Cjzy
zd5mdtcOli<R+NbD^W`r5UAut>Y6cotkWf2lkWg#Hm_^;A#?kN4owa12>$~f64Z$L+
zUeN9Zs5NuFd;tLso`N;3>51>ui@r)+R%?1rD(Vlhb{r2NcyAATOeJpHVRjnT4?lhP
zKcM;T-&NH0oFGMrS>+4{)Ir)$yjTlTl>DLk)(Wc6Wv6E_M)$)lkG$f>%;G%xr+eGH
z0-GTA5+7iCYCr6};}NM&b!D%i?#==8{APsXE4LiGHEq(s?`<<+xcY3gcK}V2uW<CT
z<ZS?2w3wFRi;`Q@!CEl;Ra?`?vkXD$)(tEGD;Uc*QI!xJ=iKj=gpO#Ox&kv}LnQ4Y
zJ>}a6`ox5aTeRdgoB(pG#kl+8jH8CRd(zcL%m8Yj`;r0M7y;D4cE{rgghHS(7>!2%
z=K>ak)95rV-xv;{2Dw{WFdqf7Sh5DWzO(y0hCkf1vs^>eA0JW$z(P_lY}E!DBYc4i
zG?36TCj1<*5W=0bFy+~Q*AjCuuLr^_&++j|#B0~XTbc5M{9U<jG;C2I7iye%Psqh*
zq0sQK$(#KzP06K9++seq{Z=*s2zUF<$AvrFH(b180gSG1k<)Q_drh$vXr`cn2fqhk
zC6Q%?bUJmjvdf0c5GLgU*0+2`9y1U5w_Q(Lm#%KME&v+o^RU>6=!d$SaCZT-U_N~J
z*%C1+aPmS2!+|}Gy1g;6;bBjLW@RTGUzmsTGi$!E;w#?SihP^#f%+Khtys$BKYs2!
zjsg5L+$6ucLu~zG5TY+){skdPy*&#X@lO?$H0YwuK(e+X_48;l^r&!6PjyI?Iv!b-
zt`PHJD>zPc7f?cjYT)xHGYQ;9f5zPgV;LC??W#VqhECM72|_}E$CNH_o%kF~cOneV
zrVmbH#+pe0VB1yFjpm$*p^FhL;@|zR&N)G8(Bvqj#&bM!oSxxL$#bzOv9PEv>PbBU
z%1CV!fx`F;8weFiUCrFbGu$aZ1y(>Fu^p%|@TZY6Tqa;~Q7;UJq*<m%37pxTG8W>R
z%jqr_V&qpHS&p3$Vne|jRpL@e(LN8U_d86u2UPGij3`o>Kz-n^-d1+Titll9kUsSI
z34jPo<I>g?(T{O11Y%U(Wlq!5u+Oc3<{BZco*>++T3CYPohu$ZS>9HSh+uJCPkDZF
zeM>iRTs8ZVYe5wTT%SIqYf9M$-?iC>skw~p&MC7TBPrOK3RKg-(`V}Xv#=A|StvE%
z$0#*_u9`M5I$}$7$fnSsbq=NOfBTdPiQ+Jx$0@wbqUarJ1pr3)xuWF>g2ShG0%PTU
zX?D^rsiy}tA-5mNMt?dtkmm7^zclNycLUJ19cnrNi4<?08`{;*OY~Lns;Y^@wei|e
zi;c)U$NSaK6o4lh(lJVslV<`{3HbFZUwih3sfh&{NI)JT8roj_Bjk&}Jnv{in$SP~
z6#QENDIl)>b3bb!BZj*epoDlRr7W(e1Q`|l+%GC$laRYuHiwm}S6WX31H>!Xk}vMB
zy}#aP@e%1s(EzR%2Z+Fw2OoD@ir#f`4bHckZ!h%nC84drD8i@Rpc>HlQhh!sX>}4Y
z(j>?Hd>~rroltm3z&1_2F<1XoGuMb?A>7KgMf{JNrr@4~xbs9!EwXa#u+Xqz>rw;*
z;RqD~J`m#Ge_czD{={3}*0*pOE<D++n{GCi(m2m{P1)0)V(G6rYVoj2<z+tdG~sWI
z{{<(&4AGOLAlxZ{DC{c2Q=n2vK$~WdR9EF^ZK~H3Z!8L3uzP0u_jhYk!_*?RxUN<>
z_}b^t#9$!NypQHps_;5oxHokCGm|Zwa0iUw2AH``Aakog#s>bBb=<M)i-11`tDWPl
z=b{Jd>fflhc#%b);G37gQT>{}!Lkp3p-j)hSdO)TS6eo*k$*#MTxLNWoHgcjbg%I~
zAlh(FxlGTf;~FYns&rIWV>)o@3~%v=PyH$TR)MhTPh@f59YjrRNp7v{y(IRNp-bOi
zXOHLmkC$8PW__KP1e@@*5OlU}mXD6*1^90<UsG@zry;HQFg+I4m~2Lv1zJ2gvHR$X
zL_1Jyi?a1Ns%=}A4@-pygsyPm?|Vlynp&=<)o^3Tb>r0~j1Ycw%cx56Jn>}@tV$A9
zy%1bRdBrgNQh1O}PT*!M-Xt>SQBF}-QRnE<T;x}<O*z}%z14Dg^Z;yXd^foU8?%<{
zFW*YE^*=Pj?;ubg3LQe*K2>5PnOfYj0PW**mmz<hYBvP#BJt2{X}5K=ayxb}6y~nf
z`%!DkHqXG0L40-{R$Z9Nl8bQeva?3EsohG>L6-q@(M662JWSXB46`j><-3C;xh;H*
zf73!@l$<9bjqj}g^@m40QKcIBO>`%3Df&;uoAsAQ|9^Ao8QAInJ8nIs!DV~++e-VQ
zMy>-pI+sNd3fHIJr@MnGjy9+hM>(qFEOPyWhQTYMFkfR-OBbxFDTzJp*+X!2YaKWE
zuy^-l=c@XD$}F&Tdtv+Nh<^eO1%v`Y;YIsz-@RWSbS0U+c<zgfl_TgTW@2ufKCAlf
z(Dd#>M`lz9zg}r_#@Qm~sUMA9dW3ZQ?^2gFFix{y8~l&&eq|PLjEz37E_UIzhRu4{
zP&-{+i<96CBvYSMho;HjM78{JI}B+5HqCN-R!E}(d(}RrKdat&#th+Qa_5`+*!4-#
zO0pMFMsN`!tqTDxp~MV>2saqM?N|lOS;O>V`7`iNlicLINU9-h0}KYP+CN^7o^EgQ
zE5MU~r5r@bw}U%7aQf=NkcMCU#8shtBYcA^gLphsRWIMTI^-FGzumVCpqK)OV1>zX
zMUfPv#5|xZV$Q=qBQv64(Xz>1YJ;DjM?f%INvX*t>2vQ72OG1|UEFw@7r(k(>OSng
zg?|he&{Bd}m!`to<`u#FX7M8^+l7wD?Fug6!@n_p<b}bBO+v87G@g@_5n5ON9C`aF
z+X~vn?+{8P6QAVB^9%pP3Q%37ApY4GdC4K<sFDQbAz6Q5taOu$w^PFe02o&=Gm@bu
z)|icia(bpZsoJN|e?g@dImRLGJB3($(UI|g#l6Mr>qGG|J}_apL!PZnL^8lB`c<E?
zNy4#!FB3PU8K_DrPU|2u66;|pfaL3?H6Y>`4j*#udp~vFi^MONm|ADd(_0Yw>OhIU
z6GpAOEH2%!j#$?KO&tBfsShbnt&a?4fuRToVt#BSqANLJPT^B&VoKq&o=ClC_^7n>
zLK)+L_8Ci6%0evkofQ4MEY$?HIad)S<)!fVGxcZ%S`{14Iq9f;op%Mhf>HHCsHd*k
zu}QPpzNv6(MOX=tHcHndl3CrrH-jVs4{MrD*^t<c$_Oq78wE-pp5HW<oAYi6qE-?*
ztr#6(V`+EpB(MZiygx!Q=4$M17^4S^jzPooZVEI~ExZ}kZ-F{?jZnF7*#UK;E+#D@
z(N0M{)luob7vf@BsF1=d4!EPNHZ59|HG<$;j5ZUq4ReUP@X#z{{9Qp@o@;tFi;0su
zjGX&4(@@|_fgN4{P59+T!K-0I`lWkc3BK5l0`SX(eigD&K1j$c>3qxwuUuA?HfbIR
z-!nPg$~Aof1I(HFNt7v~3S}=lSHWnMdIr>Q2fnHdAaoIxX9?pUg#z6~S6oe6bKIuu
z!x4vb`5Gk}f8DVQ1^=LH+YJ=F4;IKgv;&N%9-O~{$7<ZSV+S=!so!&m=x)RF#bS|)
zw*|oT0=g-JUeP@ge{ybWX^Mf>Qr*1mA6ra~C9^LB)#p=#!@of~<MleG7CPccm7_uy
z{`~@+P^ICDPgC?sCd{$ST67?i0?Z6zDFIFSX2DbFl<Zr`hIo~}eWVbgryL`XUPt>*
zuE&UWW7Ww>E+EhT=|J7aEm!Sh!dz6@29E^oTo;r$nK{4HmlMU`R8q6cu<=nTTA$QR
zLKsoMs#`HWWBFVd#6#iJ^N}ajDqF7)71_e)4DxkFI=_6t!qb@D4|+8UCdj=I-!yYD
zPB#m#*~s^v600K&OU_QAwMvW8h%53K2`A}1K<!DRwYjYaC@|A5U}{5yn;^&EiU8k!
z<qu1GU}<K0Bxg`GkV2rTC`rSN$m=0-(Sxu^!Dk2>nSO|1ivwwjijMeY;-Ty;#L*IS
z2c^|!RSmAa>8@0n$;?&~+h6)QFJnV*NW`zCtrRrB1f|&9r5z&mSqTEtnkHW<*m-H>
z#Jx78!)_cH`9|~cr#N$_%lI}?E&W2!WL3u`$CSat6?TJuKx?9B!I;eQF{t&g??+NZ
z;qrtMwVeN1(qG;y=Tpn<bVmGP$DDvRt;WZVr8%YCZ9ENRR>Zt3(SV;+sR<aZ5WISg
ze7%h8T3c0ZE-9fIm~I6rty7qYD>m$l5E*Fl&N_v)gFP*O>0KMK#%a!;xu|kIm#8_T
zR%8D5KQay(#-qzj@P;$K02QLP!OQeaNqCNC{jgUY&@Ei!2u(JKw4zxzXbYK5TETqH
z*(+VK^Jp<94wt%I5-=a|x?At~&Q8a7_$;nZA4^+!Vk8M4GkVE_0^`ym-(}HkhbAMd
zxlK1An;&jGYw{YhWS&d8^!o?Tzyfcpmup?dH^3xp;Wq0eHQ2BC{L@iW{y4Gwf;m%#
zq=6n*>%s3W*x=F>)K>+E$#@yvaAt25euPR6G9jfM8E5+|o7WIvEQS&3v?G*_H2&r_
zC6`q^$z81r;&<%04Ts(T+G%Y_Gs*m2j!F4y&CEju38xnmDpe6bQ#=K<o2jRx-UL)C
zg9zM+u-CnfFNEY$`yBW#pJ4lzcR1vJz_Ui(-~S`N{;zR4|F?*djrqUC*Z(#1?+AK_
zU$4*p0KjvDuyrU0btA`P_4{WKwj4vTz-YoH>ll9{q0BZXE43;VFA!H3kJ*>6m&c2v
zgWXSII&?R*^#*trM<zEfFk%}<w~#Tz<_r7VV;-}6`8gf7vYE*V<Q?a1;|e!Ax(Ub(
zG-zwb)|b9b4ZRx1a(5ckI1t57D3%kbF{XzFN4yvKFBc=hPF`X|m`Ad8W`Q6?05_uu
z_^iHb>|=2wbBw^*W`j|>T|DiMXF5{Uc=7HO<?qTyTwR1<SqE|J_69ij28)pA5`UMn
z;9<O|!W~KdAwE~u<H`JyGUR*D5>y8m7Q?MSWupnB5Ytmt9Vs)9>y|?3;2HriN0c@4
zS+FNS?$Wi8)hOD{0JAp5awI-~zr3To57fP)E)!%)lzt{f#p!gVFf1}EE92&2i>Fe_
zB{ba)fTujRX+Dj*h&=a*e)Sm%H^{<sopd<>E&UIVQdyz+FSIUB5w5rQ0a7nCe^mS)
z-BB#!9N@V-vp(vBvnB%M!NU(i+5+o*>gl^Ws0CDSJB5${*S86NqfAqF*2qk$FQGga
zYL=MA<5X-Cl?GmVa|k0_M$*cVY?^yFLDv2@$5a-a7Xege8m<Qy90)zcN>hD=v6Rp6
zp|C@l^#5s8pHVmO9vJR)!K|Oi9=6U`iAq%qmwvN(no6_Bu8cqsIiq%c0*ia{Dfr}A
zX)q-dH|=iV<=U$b$s)4+4b~6FRel`4%i&1E`NL+1uMDR&^-4B3R52w31r>mhnd>5c
zpHT}cKowD)l=HA+>H2*Ce7h3>0wy2gGRwz@cMoHE%DfM->!66|^Tp-m``Q?qEXZ%}
zA=>^a(ChPs-a)m~CQjZi>UpzC?G{)PW&<u%4yZHb9b?UWH$;em^yLs$7hytPKU)nu
zRVAw3VMB^=7wmguIWQMQ5WtRV?Gh}EG#*rgADq)b+FNi>_JFTzu<DX1e#v4-Uy*Z1
zJ;Slqj-4`Gxik}3Xf84!A-dm)H18PIJ{*^+`Y5Y!(c`fjWMhPv%Iy|UQmwk(hI`k|
zVY|qD!&)DOYky0><4c&xG7i#8uDXtk>pf(`GA1^+eL4P5ts-?gUSWoLfmvTg`fESt
zta4Oo=ra2Zfy=e5lBcX_Fko$>vft88$U52Cj=;985K6hrHX{t3=)39k+(k?*l5!Q)
z*G0Mnje}KjvKVL?T^adf2dAK>&XYMwTv6^VX^ZGdMcMAfOO`du>gN8d`Ob?>X9XsV
zpI70d?%Wv-3{kW{DXH;gmDemObQLgXo>0rP=QWTO&bf3*In`ctsswJ92G_E!$EOgu
z)uP4f+5HoFj;OiIDAO_XVYERpUC2C%FQTfEib|kLXvU&Rx4}43<UYXVVq9F7q?11|
zHG@I0vLW(Pm(r}nu^TV{iBf@LY3{P2kA+h}>?xzgk`vI#hXX5j7FhJxp$D|5ZW%6N
zMtQ{<!KY2ro9}%{9%M7->j^NHFEtX(8-a`MsozB#26E47wX~&Qy<MYzGJ;~39k5EM
zSe230l-Ou4vu7P{`cM;!TdRJSelQVV46`wq#Lwk_zKUBdT%vQP<M>YTQJEQHXUZiJ
ztPCl3P87z+Szs?}ae95;x{^3gAcFAt$A>mD@1X8ld8kF7tT3BVN=AH#TT)8h(DbDX
zD-=-5dSZOSz{l2;s^f}VuqG>w<jyRBNcqZbuj+~h1tP^GRo0+;$8B7JH>fzLLQI^S
z)tk)DVcH*k?A%@xd#Nlk;oJS$Pqf6-T#ZNmQEz1We9_5pS&F6<tD#_8(ut4lkgAa|
z+BMu-oG5p73V+-w>_9%9G^&esmM(*?NwzkZE(*N`RR*a%lGgd4*PiXjMCyLXtNdP(
zevPVKTT30Y3@u=NCp&j;lXp)X*?PYemfay-A92&XZ!Ujplc%efvy?{8dw@sl8}<%K
z)91Df+e>(2halJXt#ioL?2YSRbV{%3v!T}agj-^Y?}D(a?HgAFThr(6Z{!#Tjjxz{
z;nB_8=NFAHEGTVv>k)X^;(iI&^C!-}=)z9*_ioLP;C9>9jcTnLaN_n~ui2I)l^v_4
zwLaLK8eF(TXq?-?5@5$$S6J|4y<THhdNTLWxjsb#HCZrDzgTQ&D~O-(cK#EO)>zO4
zwf;e!5}<9$K$pAF=)ueAhw>>8GZ=jv@^!B3yWq~x{QQ<8Q{C13UsI{4X4;G;mj)G!
zdfMmOe%t%Ew<giEFONDjs_^}6GlNvmxL&~Yo%uIoH!B&vXZv<%TfVBWCsW#IOm|AI
z-EQyKq$8X&#t(3;lB?o>#7Jhg|7`~SO_2ES^nF8`Q#O|^u>WU-gbGQ*J8%GiL<R@H
z>kEEoWKBd%CGGml-=z0^>WSy5>`4w&h69YXlCoIjxA>Nv?~KRCqrIcAy^}xlhbnc$
zgOA@QMz2A~wuSDqYhzf2*coSEH)lVwnt=y1e|*I)g}>dtabC@GeFTwSp$65^aaoSf
z$mBxA;>6-=0Vl+!@OF24Vsy^GVs7AQPl@-MB_ZT`yMNyEDjWFzTBQXv%yDhbFS;vM
z$t^TlK-5yFCzy;rAkB!>V_SdO!r<w=XP-<wXbb$)sKlNMi&8|#!JQX&4q@Z%pnT(8
zhV8yb&OCfI3Xg3`w820#@S;v)$g~@5>Tntsd`H8JM0Kx@svQdA7fi7^r+hhaZ-`<I
zz_F;e(94bTy)>E>AJ>VivH8))k<etJYHFNO;Y?n8$5_Z7D<ls(S8#V;5<*^u*PD-c
zMVj4zr2Clkc)Umj@5e4wT@FTY)UN)>^Zq~y32^G5tX*EIpQy5bGM^i^s5CnQz?me}
zZAB2*s*KR$KvX{>V&Uka?MG`)1U5!OmXEXqb6jG1LuX-51M!1WTc2MuqQLG_U3ruW
z9b-}(#9#vYFv5MQqe*2m`YZCd!~;4_(c~ll`IbPP?9F;b+g0C5$*w7q3Y`M`H^leX
zI`T953f!PHG}eF<ICA#Vi#HF+h|RzaM^>43w!>G#hxg4US&hqp>lilOG8aQMGr+o=
zbfj*~lM6!Dq5(V<)<gXjq%LPmHi2)xB4`pqkQ6Om=d_C`O^IDp`bK5jJS5&aB~8EP
z8<IXrch<NHrO9BYX`any6W@XUmZ8YMm??g1A}V)44PTq?Do(qB4h&yQUi0X?F`P(a
z<z)6jJN2d5{jFHN-&yAjBo=!)ZIiXP%^uO5(x?Zs6rdl?A9{Om6^bAl{#;FXYp+9q
zfTT$<pYYkZ>nkk>f+B-n%;CMR$S;cUs(^UfsR2n|5$3AZ&Ia{paWtBr`Zzy9vJFqc
zWJ>YXfkM1d#g!yRSPVOk|MciEZLJd{amHS<D3|0jG{wF$tvuh)hDXQ~%BU~M1A@)M
zi_YY_pxhJ%TvZg2GphhjcR@vE@q<a|#Ip^_DQM5w5}<ehW-Dh#;=&=obK;n0O1(J4
zgK9hBc#d&|P2}d|t$#+!?Q&?B6I%@(%S2b7Uzk{uo4L~ZIIz!woXyBvb%hU^6BOF!
z?))bpgT)eLDJcHA<zc|8L0BiY*XglYn?ii%g?GiWk>!JfF$FXocw2+VG1w0dN^s%>
z{akvFvyiBod*AIPTD!S3mS=&wao$+BkdB(=14^u@;rI#LfrKT=gYfE~TuZf!N9YZ*
zTR^2DD?_6xyiwV;4Kn2pRObQ+0tV|$kUoAGb_CBhei%3e0m-`r8-zYuo8xqShf!`;
z1S-)A1h9}XQ4EtJ!%zGp(PFsz3%tqsv$h9tdw?&OT&i$wv9cj+CEY<-^UJ}Ko8!d}
zoLD_JmtauaY7&VR?to&End00I;}6f8`K~(jU>EQ~dOIftcwquL>v2W7EP=|WK~XMq
zNLF0v(lO2Y$xh9)3{`8UCLkBR<D~%l&8w%n8pH<v=O^i2S<!0vhm1k701yjokB?rX
zu!#@7<jw|lGa^O}Z-+)5V@4K%-FLP{vsL9kft;B}8!Y5G%=~!g+_+Oe40rAPfQ|3L
zCw-$I?3x}!vp&CwYA+(@OJ)@0!6^UGHHWNXrRsjL84s(Rw|C=QK5-}_Rn$}oFcaTP
z%nB;F3LcWfWOvwkzM9e`i3S|0Xu(4X&iI-?#I+hmdl_o(2$+)!HtoYVi#`GHj-eiU
zXmOI-9iAT)$@Jw;s3Y9bXVTkBerXFT{hIr%D~Pgej}nwIa4RQ)>sZ-I3|?@qc}_xB
ztzjpdSQQxws|z+ya*9KCnw~~P0Ze2~reH(L4{7%&OA_1@HH#X-TUmoLUm1-ZQPjku
zu#Wag>S-^}K;1Lhp9l$7A=(B#ON!EjeU4(Av9-yHgAxlCxJ0=!SHIagT6;g`C6KFa
zX2Tm~8M#RR2V?IPD@xF|+fLiIZEH>2wr$(CZQHhOtZCb}b=H4!a<a4cx05fabo!$E
zraP%TRrS7OlmHxBPRMjyG!pA-2Dm&~h*DPTY85%R(HbJ*VyJL!R*#iaq$^17vQk`}
z_6t1RnO7CEy#YmiNT(ivB)(hE<akL5frK;%dJ+FznF2;lGE-x$x?rn1kKdfCx#2}o
zczwj1k~dRlyL~|O_iLNd1ZgehW7~kG;k(j=Xs<_p;)2ahSTB{f<{uHN-){T(X5gBU
zZFc3tTm^Rvg_+>5F;JlS|8!`!h@V;ajG=}{<YSZZu2qY}Q%IwG^a(T#g1{NHTdD$B
zp<b6|q2|M`ve`JDFNjaoEL{^g);3e7d^;xgOg5{s1&MlXC-pw?3GiJQMIDn|(O|VA
zhjRyDrQUV?h!YD)(Ymq_u(?OhMn%QBiDAhWoi7^wVMmwBGmV{bgL2wp+A6VQ#vvkE
zz{E&iX7x-#r;)^o|LLRHP^0$p34_Rf7v0Bodj~^dVnw!Z$)B@ln+k~1hb976*~N;o
zwQpA#6tS=W{`040d!DsR?w&*c8+J?kcTCDPuzw^E(=2}4VyKd32Rbpi6>k)kZguUj
zjp8Nd7-eMO=WwJ~K$hqw4E^(OiO5AOs`#meuTW@{D>U(sI7+7g=;9$2hW(o?EZ}S)
z#8Zy07-MEd)oWC_FVqXqbl!&2zRKOMH+GOvRZhmdELXdnl7KG1hO9@)KP|@dLn!`&
zVmvwk+$9_@mE9NN_M92{3=MbehBp~!M*sc>EKuDuaw2}#ybqNVh7Sz`X=8)mTRbt6
zD;gqS>|>N0t))+4jEY{h29^Dc<@x-A{d<lz8HfYr7cv{LEW4>P8(|B0eL@?YWk9Eq
zCZs|x^KOGS7Y=9V5BSX*f%JcjGHm~6lwoFO|L>>}BU%!1*d3_7uXW{zZ+}@{LC+u(
zSZskiJoYw8!U7Qgfd3hnMj&~AeQ&29p-L&;4hPhrMVn~5yY?1Iqlebn;nU{j?e;-)
zWE`e4eS5fn9K7>C@DpV$UpB-b)~r81Y&$+Bm+Tll4U8*RhVX{@X?;^I6C0p0Wti5-
zFu8g%Cm9Ow5AVmRy_31$f7>i>IBnK7>RC?xz^%0acl($c<Ky`oY&{SSUD&MEoB(pi
zY}(H+=}moA3ZNw(7Xy;}u<plD_;rJ%KF$EX+b<o61rw4n$>$Yjr(>!&WQ!NqHkUfX
zF_J+d7=egiLQH1>erQmtbMgfy6oF{DMXG<y$eK&^isGQkeNZht`Z>%~21q>j;GDPk
z%6fyyz?8s5M#t?i>iTq%YIaYlfBR1XM?AZR`c1PhYPUd_?H^sk;qL#MD2O%;)yiVw
z!gbltob(pT=(0T`CMiXur(|5%3{pAiQSNlG{{iFtCvz1w_e29RZ`u!&L2IJ(Bv@1T
zY%$1~`7xPt>li&LmIHxjz3#4ZB$3=FLBVry%_*f^i)n_kAu;AJM%E$)scDy?g>gaE
zenHxsV`S;Hcqf{3;j&5GX&h|tsddSu#s)XjYE?5EpOMK6tA)(g_0}|Nil56*Kog=i
z(!|qvm^U5I@wv5ju*TCHUd|AKmcw`aEGAKmdX0C%ETyFW^O#PXaR_!{236yMMIVWK
zrTKg-nz<nkwi{&X?^gB~XZAhb%CtUs$Do->Qc!X&%fwbeNA@4_FdYTo{;UXZsEGrw
zx9Ae3h8Yh#aj37XLeJpo2P%`Y><RJ_3dGbH@{|%5bIoVLfnDUXo~c*8+c3RbUf!=x
zBrwn-(>;G)AFn%j5+z~mHvJ{0hOd{^<Hw;ehgCKJD3~%$>Sg5o2q2qD&Gc!*ZlHzL
za$jt^T%-uvNxwcHjnIU=Onp%=2X<l<+0%kJL<rU=s4`SM{)a@~)h>qv-lHUHig=94
zbrkv=8FUwFs#(VV#{eQfH$J>X#0j--18fd)oIkOS>DNIs`dm#Og)v(73Xnv3P<5(E
z0HkgV?ln!>oRL=y>_T<(1-X*Of8Ukk<m1AF^RZhfU^z`5(kTx1oeVkc%YsXOfu_0=
zRB>v^qT~F{nL-EL!_AdR>|kPSwRQU~4}%vxUb>px%EV;e9nc-qU^r!qaRuNmT;Rj_
zc^y_})Q8;#*&#>4u!GU5Te)PiRSkn2Q&I@Ij@He#WP1O`>i1<mDfKD8|C(<!U=PIF
z<>BUJNB|zodyed1MFdH2z-`C`Rk7I9`IchTXMBT2qIr;n8lcsTwz(#YL6yq+diLR>
z8oLb=GD*6QrjW~3B&Oj~0&UpOg}Xh|gr$z#86*Fu!eHVb&Fs{ZspSTZ=R=;((t;7B
zQkXTT12J+?nQ(QjLL^QW6#n|`xNUogf%4@?U832H<0M|<MlVJirYb8!<C#@jKfMGk
z1x6qQTX7B;vrh-9IjXZ(_fDL`YO@GihD?Gfa9*q~pC^l-W;)(Zmm>kbyo=jwQij4H
zIO{IfKx_0<g&7R(XIa24?dA5!&sa$23@_od%~=f6ufO508kbO##c*n-UpEr+iTs=Y
z8Q1s5tEQxD@f3C`5O=N^PQ9mJFM&k7`#@}h1rTN{k9VaoLa#^p>}Fb6s@H(h!@}Br
z-yYuM@_Y?Irhd-Li0nzU8O_q{&ddA}?9G=}pR|zkLz@E-#X{1kr7+fvCkL7CEIj;+
zfSo~zU7o7&Lo(_aP-YAV5D2jO7=g`0nc#fKxH9*AY`~#aL52A{^Y{JUbrv*#kuM*q
zei!{Vc0H(6zim9;y1Bi@wN*5&f+$xU3%X5MRVOjk_HPjz8&jJ+AV%>hd2vJ!BAIti
z=ESr-FO`M8G@<0pBLYqlr2<sVXNh6AJs`Y}aUbLbY%q5BhWM#PBjc&kE!ssTVBUZb
zO(MDyhMsTbb{9J7)t`IcV?Vh)(?}3+`6Ppn;3y30;v20`wS7Hvw=<>cG>Ak}5G7x9
zYg}Qm0k*<Q1jrYj)geFa(=pc+Du}1uBdqk8WeKnuDw%ADTFcHoUgQ;LO&Z7y=m7;V
zCo=pgj{7aZ%zW-nMa5$rra$MIm{Jj?S*JP2&SYNYD?5u5oUAt3zuB5LL8C88EXryc
z=0ajrQrQq2Ps~jPRYA$1)_2o<H620XV#xpX6kV|+EWRzs@$Nev%(ZA^SU-Rz0P~^L
zK%9kDCZy9FOOlDcR&P40BvE4VX|aju`XxG$aDHMj{L^itgz22zB$F)utlp&2P;1bZ
zrV{X*P;l;vSHXn>SiIaq1LDCxI*xBkl$TYKM_76pepqysA1zW*jv`b!zV&+yGq2(h
z_AU~>RGHNhOt0MEO#8ikkkKkElG?8$0d3O0GY=ol;N=r(kITlBjjHq=k$TSe3LAU|
zM?$^{;33J*BIPggqMQ({c7{FV={RD1(MOB+9kFo^B0;lWsD5A0)Ov=zg}X@QO4bAq
zN)Y4SW7fw>e}!n<qgKuLz{qzZY{DaaeVe!)CPMN4z_BRr>FD+B)2*slK^8?m$8*|l
zg9%obV=@{v3mQIk&Cet_eE!)1#Oix7?I}(OfFC=vPZYOMGXry)3R0A6$64W6>eo}s
zB}*E`#J>puB9Bc+-n>Sivi0aoUZ=<vi^U8!PWRnng;_v%P3aKR^8tfDb&u<nEQ-K=
z^Agf4Q4yAJTTp9(AEz{Y*Z1wcxuAX%ZDy(~uZx5-<nZG0fd$<NqN>Ucxks={xt8i!
zk7i~NgHzkyRjI&aRsir4|Kjeucs45R+2a*uUEDFKKD&paYq`Y6aQy~6in83ib`sLb
zFB679y(w!y5nBog`s~Ho5Lj13pN%iS%#CT=#(uTn#0R+UYWoWiSnYFkWb&(t5IeYr
z`l~l!VxtQ`m;`Mr218bB`FMcM$f-m<`R=53mnnHlCi?>?-qoT!>b`*WisFHIpeK|i
zT2Kj3S!*n0eW3_~S?o1Ox}Jq-JD;#5jlgI@7bI3SU|`gIc+KiGuHYRkBZVTjtVAn*
zk_wTxKTN(Jl13v&CidywR2X>&O^J<t-|o;dR{<TXukB)6TSb=<NGBPWjN>}lg*Lhi
z5`h<i9l+&QZ_{1&(s}_+Qj4~*Rr7I6M=)SO8{LZ@&yObq9{FMu9iMG~IhynqOQ#_&
zj1K&~gLgft4a+KJbLcq!DIR$jR{M02GH4ueB>sr{+AEY!lOm8_t#;~I<LNUVC3%^}
z>IGv5nu8XkUyv^k?@4n8bM46!??YZ$-%<A$4oLQ#4rhde(OM`eI-~3m!Jg540e^>E
z^2RscE*8a(<IN!0BjyTecHHCsobb!9P(>E?UX2KtsV+sIDEvS+?ytKRkK|F*#kcK3
z{@3=H<?7LK*$?tORbr?nMpzuYZFU+`S@b0Kr1om-(Iu|pETK`E%BoMgSCuwx`tpO-
zj>}FMMnT)F+dNFI7pfFKf0w4pBSNlh<AH}m4J~ULuJ!Kk<)_q{Qz-dH51o(G%ruPT
z>fJxlYm>H2*1f889Y&YSFccbpSMOV=&Kj+p6GZCd?hk$1T<Xhq>qAI4z5G@`_(fjw
zCFsh&Tf6fWlkKWE=Dfi)b&LfnK8f_Pj<=TbITx!?#xK|8M7YJOKy90Qz9PDnz035r
zTCz?Kk0J%*Q@-|P%5>_my@2^*!5RzxgmHG{3M}EpoAIk*3D?7mVMrv|H#|dL5?X7{
zK{M^zm>&J&Nb&Mg^piGp^T|hn;;OxYxDQnx(`@0j?d@PXSD|g|X%0@5jGu&|@S?9<
z_~Rn8Sp01J4N}Fa^+Xrm?e=A@#qLO~>x-IK!#3?dh%FDb%^17DO|WK-?T|^vF7Dde
zcJq$@>fzSS9>)8eb~l@a5y2nlUX%8Zla!LyqHfF0NvpS(L^&I+oQ1<b0E(=hoDKUH
zt5<!`4tetV6OFQM4FZcI#kX9ip@oP|P*$;V89lp&YIc$+)@Apy(It&KowIUyl#y0D
z6%9fMWtpyu8hIYy=(1((e<xGe@IBl-h2Qw}am8Lp%4;Fr0Pod|wzGJ6=5|M-#yd4<
z2xU2Drcg`!gr6S^f6(!&DelSB_S%xGMmx)1I~Qo?KSCO%5+7x5v_178vs=a?2Rafg
z6Wnmm-*o0P>G8WZFez7b?3*oI+gW;CHGbR&Z#P&e&sHL5GD$B!(Odm4k4bIFwe5ib
zaVsvoKBtu)(T-YLX2Cba+F94uqY-R-q7pi2Nfkt`*`3koPBF^87AD00u`O}K**D1H
zT4Wh)!_%KUWV;kfd8kr?Pe!_K&Lg8Wh>x$~GORX@9vAd!bjWHuL(=Qk`12`<RaXvE
zv+M*I!t*ui_S`@zQa+L&YT+HBl`f6Wi43H-g`4#VwkDREa#qUhN<+KeEF?hEbjV+L
zeZPScc6Z+#64%;qTn3HaSicp1fnk9BB>v;ti}Qamll*^2txTM(|D9g#8hawvsO^47
z@7_0`lcup95SDPek(J-p=!tfG0K=Fp40Fg8?)@9(^|;2eB53$dV5(?*L;VjWuTRd~
z)BXPO^Zt;?fpNNu{_xX>3D8Ht;vajgB}*m%X{mJGJ~=<H18sB9p1;tXX^Fa_y;>j2
zWnz%2rT{YlSfCs{SRhylEJPNu@=p`zq-R|cMrqBui=I`quL!e{gOBI8b$&j35o_@1
zCg_x?ff_etHt~o&v?Y-Yl=*4DvIwXO{=%HOi|vHfXBDC;z9o9`rU%SD)q0rbYe|W!
znSg4m2WX-;hqKI@?sVUkTo_>tt_(Vw1+w|)8Dvf(_V-9?@9TK*U^waL+^5g`ujqVb
z(QDy<6xHsU*F-R%G4i?JQPc18kmJ(L8)~|r_&Xqz&7go!eS4+p6ONZ23S;jQkIpH}
zk@!0D^AMa3OM$~-{_52D)ff4(3>imMB>9(#o8iph7#%3)5TCmIbDnZI%x9*{uS7;L
z7pJ92%JwDs%!v6tpqUv+T!LDXRcAYNPX?u}D@1_2dy@4;L0ZQsEj}`J4w%qn(K-$U
zHxL6`VN-|eP{fL`O(wG4=6{sNa2vXx(o{CibANl%sWaZ|q?|#KBNRNOSx3@l8**`B
zK5^a%Glb&(l4eP2TsBsgx~Q+3n(U#Daz2`aGfNY*Evn||78g1O&zZ2vRp!ooH};lo
zB|;e~Yzv3klilZejCTciu93!F38r^&(0IW6&8-oG;^{ghILItyN{i)WJ(J8<(FOMM
zsd0-`B@QoatlW1L=4f1a*2X|eIa7GRqEW$H>JxW}ej+vFP!N%cBgl|vsoLc+f4hz*
z_*e=iYAk~tvm_OAF}W$eHPeO}Pi5I|?(DVB{Astb$Y5qTe6h9t^Le$`tG4}88VGx4
zc{D@NRlzO{FojuzYtm4PKgx9O9070<oNi6yxu2cMvf*}Ji<*s+M5Dm62jhF1J#ulM
zu>xmgDC~>ch@TuYcWi@WsYK`$bIWAPr!ebC&vlAIdTKZ-6BCeutScH>LsBc<vMa6h
zk12s-$i(?*VPf1|->fI%Ym!@;OM12`*qkeN4p(aNrIdnA_Yf-68vh2=TOFjyJggWV
z<FK5jwXhMiBL#3ZLWtm0d3zG!&z#Bo;4CL63`)up&<wljg-7B2OIUa#5lK*Y$245#
zU}$lxINg;dtbjol?4cF`W~N<?b=jdsij|n4+T}`88Dc0p_a^P@RVTtP!KKuaL6}rh
zeOlGaqi<^mPcc}RZF9%QYXDleWd$o<;id_!m??zm5uid9iAB6XU!ff<r<02(*15oU
z?3VucH_@+m>4ZP<@*}#%@y2Pj^TyVdsK!t;!a<B{b|iN?cAXcBEC!P?pgnJ<$WVK0
ze2T`g?!JgaBRk{Vam(UI`S4wvew7HzLo9|#Nb`|T1i8kWuLpn@BL;>frjR?4O2koY
zUJU=0O{P#eAYg9O>)@tbBCwGe`r~^IzJrpxN8_Es#dvl`nT#UO30h~goTejKR!(e_
z#o<LTO+nM4daAwn=8R^swK*YD3bbkKW9^2Yw-A9c`K(9&!B4%CYa*JdX!>OVv*Yl>
zG6KsytnAYbjdBR#2p4=%?Lty71gqrdismPPD@r_JSW$WB0u^AZN{wK`E(`IrS`ERn
zL-8_j`hjzUJp8iFsSxIM^2c9<EGOIH?+ZB5OUVy^wL;qmaLA+j^JD`bf}h?6{BVbe
z=>flA|J5@%RB|mJfMW$8K>Y%~diht}5-$a1u;8*(LVj^$HT9-;t<Q};OXOZx><v}7
zXWPrX*I1#$C@fnmRwjstYkO`@57oOm9tb~3CZd;(gWKRA`f$QH7_%0O<#dtrmufa1
zTv!Rs8WV+&GmUFF`ly>UUdmB7kr6>OP|~PGv6P2?&g)-IG!uUd_DEy)2(J9e-w<L=
zN3B@y4gN35zbPw>H+yr!7x2mA5(~QWuG9&M+B?(<iqhts2RO)-6AcnjBx-78iZdUx
z{R<5<XAT^_$NeHqw5QoDkSc|Sgawiq>`0Z90)HWSqv%YGPxjG_-K+#t^x+^?4uH@-
z{`l?@r5FWH#Q>+xG*b&DBti*#Ouv(b$I=T57w0S!4}0t5pLG_C&OoL9O-7vXO0Vo{
z;Paz;1yd$C8e7ef_NJCh%<=yDx!V-(e06|=5@#dNQCaLOW0>|gPc}9cEh0parm|^i
z8w`X0S8wr<Hpd5CWb*r<^d4-S!B6`5`Xo*ogQWcFCpDMpI`HYLAFFi1FTxCb@HaWc
zWbg55Z%%;!qp5`HV<wI(G*wWtNgXI@<t+kJfH~RX4&l;#@U;e)Mi~_);7l;o??L|G
zqFDTY7R6_J*G{>TRX?0O-Whnjq!m;(E%&oiP!(rN*ir2yD*6zjg4FBWB+YT9{u)%X
zZ%TK^^QpUBDf6mCsMg6~*s`E!%JrzZkK#jactq-$f_PNLu!q!+815wrg~gGS8=?`@
z@!H}dgc9tM?Y_TXJT;$^RlIC@h_l2H6Y}Th^JSaGA9$+mR9&hDBj+BjS-~LW>0466
ziMXy-I=}#euNIc$5#BtiqrA9iU;@g`R-c0yhvKM#^`TH=TEX>-HzogGLKB>Y%iJpL
zNK60rxK`|C;oWfMWR2EPnJ`l1Q7N>Zky@a*JCRhug6=uwE#q|a@P0+%HjN}B^<AB!
zW4+e2IEkbRCF61Oa*I9@oOCNZu+SK*w@B(9ys1W+2%s6fx)HHCGc5v@>eP6qxj3O$
z;-ejj8zb|_Y4?_;#ID-*wV=RulRJ0CNHzzR>Y->@J3kLHIUhbY{^;oF=6;$MZ0eB_
z(KRed7OLQQJ=cZkt4$T6<kaCiY#<pAF!<^28Fh<u&_=<Q*<Afo_yQdvH$y4_Q(I}n
zee7*~Orer6#^xy8m#ORAF|r~MN9xV<S3BrPit_c3Am~8<8sN5Wh_}zA3zcOnZR)m=
zdU3Y1DaN{I#-eY}?;8hqg2*r1)CoWUuTb|r-*B<$-??OS$3NdcapzM*Wr)^=@q%LE
z_&+lD6z{f2Wlyh(@zahq-;82L`}|o&_Jz`+8WiLNy(CnLG9?ulT1E!crw9p5M3YgX
z;p=n-4N?y3auPAzeDAdN>oF`M#e2M@3&fy-^yrrQ>He5K$M8(2m-k^LGqDKZySjEk
zih-mfFnu;R+<G5fLpJLkusWg)q{9lG6oME=sYgs~a7%F!XxEJgkf;{tqJhBbD?$^X
zNy<ktUMEOk3Ou|Vziq`rofIYdTeHoZpE|p<`yO^F^ob8z`CZr+*eL6|APZ%=R#ICX
zfBUqOm(b*L8)NH;M|-T+Ni@i>d1$-B-r&$+R24bwiAzzT!N8Rp*BoAG(WG<HQKJ!7
zX-Ge_0yc6&=UJb`p_o^2O29#niJ32MSMw=e;fGM0(zLR@#{F$w@Wiq6DKl`{_t;JY
zbk8`L^Rcr-)s)TByhB%0>?w)ZTm8bHM>F9jGW4fLN>y8Tw!61kQ@k*j|IoUZvxg5T
zZc|SZ4%jDMDLmR!kbJihpKd#>47^-rov{{&XyTN-p7H#=mT7T<wK#NAp8|*3-O4rD
z!4TOla=WpIgk>0>>g~`9(vu#0J+ttiLgc6R^x_SxKcx0fm_CD>Q@p_<*wOs-OOOe`
z@Ea|T>OCc0{|<y60@AmQgTVY!{o5RLDdWj?yuM!8!jFgBEKRnKWy$Ba^)}<A;%svt
z!eigBR9tq6JP5CysV+A=x2M2E;I?$;996?f<1mvIX=|9n#nX*jprRjAm|SRpm|wyh
zznNscS)f>5E7DA|vgi*fDl7$Ed+tj0vfQrhMOOR0PB!lRT(xGl17Z6Wv!=7W(J?Mh
z1D}O%wrSNErvP}CzG+)-sT#Y%!mPn>fW_2XB>#eE9y8SaPtl&8@jpa+4i?7$ssngi
zOU@3b3%2i5N!sofu!ZQN55P#l3m5y0R=^gX2W3FYQ#`VFRjAKbLgHVoGILkcjzj!#
zzo%O>iipBeT=8(q(5V3296LO|?GCpEvpXV<JFE8_3xF@-J)gL>v+Etepall6H?D72
zk|L)!&)$@H1#2(2FWj3gd>a;!B7h=5D-e!t51{q@T492yhIiVV;@f&!YuUoC!S7xm
z3EbG={{CGYykB2}$XXk|t>?Zq2w@LCuW!%05&z&Rp~9|sfl>X?Z|D*YzXmcmC62$_
z8y@QqtRK;S*01-^Q(N4xFx}AUFH!|gv^<(2M(@cOLhvKP%pED;VmOot9=~w%28=>P
zxwCSMk<A;&7RE8CIksqxUO6iia<pR6ELJdFE`*2win0koY)i`6CzGvl#Rh_ftO|@L
z5nTyAkD=d<NZb<^A2a34$Ey9KMZ1Suvt`bMFp{w;4(<=F1!^#092v_34Z8<vbJI$_
zYGK++ZFB6Y7IupL0A>s`%%D`ioJ%v8nJwl_R#rT^(;k`c?WWq(-x*Y`$^^}T7tXiv
z%mw2@`Q|LvEndGb%@3bZ-MOOI=~J|1w9(G}6^~{Eq3VUv76XSvv<t!MPqf!$P)eC$
zx0DyuxTfa5Qf~Wxw;}Od?uJdaCbz5>Kn{Z*5p0^}Kr|8v|H=^5w*$;{|IM5h{xc&)
z9~z>>=deu-)K;uv(kOZ-?kyx;%wasIxH|silu3kT0n$hYTbU=`#jIf!=2at~gIlA0
z;k-QcNZ|xk3SzuWO^9Zf#h(SXfMvnUPcvnHe^OBMVDVe;G60pb4Y?$Mu5>X(+MXqK
zo<rWSVsMqG3JWq5xW;Ziv=!ID11o6L=mF8i2`O1&Od;QpZdt4-k=PVJP1g6|+r32f
zhWHJ`vm6=ourmV>m?^0sE(Um$ESxE3T6#v&F*kPI-^R#V7?5|TM8JX1*uKd^(v;xI
zk!m_Y;`8X3u`s6GJoy(e4!qNJyl)9)8c$Jjk2D@n#Ub}v4Unvo0Fp_eeWJ$)ax06Z
z`sKnXqP$6eiSX^}Zn}!bGuAfEj*HY7miRLnWG*XGP%_F;g@1+TI!afEjyO%+TVz#g
z;RU0D<b0X)x=@OUZ7G3jN`DK$+<OA^8`A3!ZK7U^)}f+c%^H{*wQHrQgCH=TWy&(G
zQ5JgFw$J#X&#$V~bG~Cysu(+2TJvlpZ<9`mC^1Q;Bj0QW;#UkKpMpU_5JO@$4R@n>
zp2}Y#hiOn?A2PD4X;DR?uSw7(QDRR39oKb^+=F^TXxK(>YRmEUSv4{qG$2bg^^=1t
z{NR-u6YsZT6?2sY2*|}}$?7J_G;=^rwI-^thVl|eHJD#v&Qfe~4`)r{(RjE{$PUGm
zSG0jSNvEh_viTZIlGV*qx(4p>nu=q)y7?BVp-{DVqrNN>PJ0DJZDHNIUoVu(Upla*
zrJ51bCtCE{M;kN;{c#6bg)t0Daw~q%hX=V-ZB&kYg!rqdx06r2-6y_d4}B-<_lpgL
zH~?u)_~xQDlx(+Rg;awUBIq#Sqyh!%S()01>Ge}6UH=}d8}0^JEna!dHCj)z{622v
z&oxf4=Tg>dv-(6rC!WZ3Bvx`PQ3fji>)PjJ$<d9suO`FVfQqr-ib;J!gEE`>px<z`
ziDkW;(Bk9Tpm=)O;wb=sGcx6!q!KVw2;X@U;cZh0a!h~9M}K>K-CvyP%C%XKJR?vH
zmHw@s#$dMa(1Uv_9QKept<K@c_r!1nP2IT+WsM^S;WSNB_$$D#*~%#t3{Yk>ch{*R
z)Ly)P*T*s;G4<(6l24NJ-YyG3O1_|Gcqb@8Z^IMIt;_4l6$FJlL<dtoya7?7!sluJ
zu?Jc{$JyPXP8A1Mx+GfBjf`g$(IKm-Kl)BoVrv&gLf1GznMIn>OoHBK?WajqFkvLE
zMx6?%3|&RA@P12%y`%|==~+}Ntxa?zuB+nQd##t~xxlo5e~X0d;DS1RO;V>BIgkGA
zY$-{yZmg-*`qc4l?alAlbQpTW(K|RBk`MZ}7T4V90w|;_2EL<oVf9kWLV10|QEQVF
z$lTL>DKRnfa7G|~r*+II>CT+wtL5v_VLf0tZrwJCR+x@WY7+h&U+$lcRnmP$VSPm0
z=0gBp<P&dMT29S4w#XTD)4_Dsy#>hDUwL;2wC70C*dWy}0_OF8zsli0OAc{dY&Y}s
z92$WnYdViKWmTNDsWR&zo4#@}t0Qtd%mnPtlWX<PGwO<AWEdOQJKhsG<=>TY>i!rE
zpw9?!&nDdils_^koUZrtVFec+qj3|+yYyj^*7xOEL0IdF$|h~ZXz;pRwH0@e69&9}
z5g*CwqlkXQ67m6qrq@)!;0%l|JH@m6M3a$2%jx&XGlpW0G<k!mD2=fZHbo2;=(s4=
zipHQ$k5hY+{1)jBxSSnpRtlr0P2euH`rLyw_R3PZuh62Y<ai%g`(Q9(vP8T^T6d3h
zm`b3m4dax=nTy!CL-pzUA{}_xIOnteY+B)ivT=!f-FWC86QM36Le4YBicN{OF;~uv
z)HIJi8*?mwi|b|NVK^-r)#5LjCKj#$jXyJ5Q59~G6v6Y{VK1*6i>dji;MrTkna%cz
zRqDav@7i;vn<IL(Tog>x{>=G4qGfX)r4pW71YFCIFfbXVa?iATdZ?`w^q|_<=sn>%
zhwc=c+n{vXULf&+38STRyGWvqxr=`nTz%>grzTHl>$;MPjQ?}>X?Bxc#wgHaE;RI5
z(gp^aZ0D1~K3sY&?Z!p!X#(yT4};I-Jx-|`^MU+Q9p|$Wfe$QA8uOUV!Xu)vSPe%L
z)uv{Z^@dgxOC_0vn!rc@2#K)lF>*w=Zi!6JI>wDIkkcf@<spcpNmTiwxg)owiko!{
zzxTfGqsp!=1XWtwA!X`#4Xy4gteviVH_$j^RN@(r0Q4%9DBcSoMGo?Ilu&O($4yY8
z7QutO6$G061L}cXbaDuON~(683ho$gedVD$Sk(uni@M8*30&Hd4!+I?MRyD~d1*Ex
zhUW9q=YV>R&lea&X&}l@^Z5LaGeap3oh<_&dF(d(VR<(c54uLGW|*-YHc?*{lYgND
zwy^ma5?|VA-iE65rc>c@npbSe=9^QBnj|IS%uD1{8PSx{jukzh_O5bLNpZ=p)wE?}
z_qmi?cTX`HwfwQk*q63MYRWlX{bkC9N+XAlgep>nYQs_4D#e7OpeVD_vYVDbT?`t6
zeyK0jCS*%HxpXpzmc`4Hwxn(<f9ha~Vsf!0kIB0$l&Ljy>4P?PzhsYEveL&&^aR~_
zw{kvc0uVqt&Z^wSyQ*Y)YO(pN?o^2qx7oJ$VL5<Jb5TeC5T2B@yi^5S%EZJxjv#a}
z^O$C4Gt%T57I?Jeq3dT#MA0rli0<WQ!M5&mB$;P{C!?C>SHID4t=Q@ac-`CDHrsLR
z#pmo2<3|d1;S6f*IXXIyt&XNsC?ey@kk3D`4`*z|Zt)S!nJMB6ciSze?`L6?Y#_6W
z8SXQji5cz~vy0jHTb-JFtHM9#X={F(8V%3os&(E*Q|u+&h?#zM{K^-=h}qL7*k9kd
z9G>I1$Pu$`akI!dt_$3Mm)L)Io))k4yqgqql)RV6>$H9nYCiUDa)irLZ1@x`arTY=
z7b8hJ-F<c=^7QM6&s-LsBF@h$gz^>u!!^%bWtW(}0y(n6^n1ST7et-MDfJo?q!}!6
zi;q1^<i5lrUH$<@9|0EyXvCX0xWuQwz~`|08viM0voZcJ$o2m#hKGTJ?f=K{u(SWS
zrav1yoH5v+8@&epzGBesKSODljSw(LC9=RA=K5v)T0{6G?>K)PoJY@5+JAfD&5Bb{
za;}oPU6K2F9d6s1wG$uj_NPy@8(}QQGuq*32*oiial1{inMULWO~ykAQ?kJ&Nl6+5
zCWH>SA=oKb1xVQ^LKTm7#d$dtx`2&Hm&pVplz!3j7Rcy=DVv|>??vv}DE8qFYom6%
zn-Mu+g)_2TPzx7&G$qAzjj-r}@7L?!QmwGj^x_*sC@}KHkr_Es5;Ma$D-T<P$%%%9
zPDo%Ng~sMECY>Tr2Puv=vIphu?U4K3vb!4IVk9!d<w`fA2lQzwW=NUr=Y6rW&#x^a
zIfSo?Ny2LxvAd}%uxkl5ADNa&l<?IEU5WPv_Lsn_xWk_uRKOH0dxc<%HS>Tv>1J?5
z?i7#HWo+anbymSmZVt(W+nUppxy{e-saedgV?`s&pWje1mwMK<;-N5uh7wQ{*{2v`
z2Dg!l$--M^<RAU@S*)l8t~kT|i8XE1g~8B9j9CTRO4;bSwW|L#3a<9V`LoO}ZOeyA
zZVam0lv{sm;a8rpHIx*Do(fHm|J>kstIo~9wht-Jp(PDCzV@B#XMePrFr0VTxiDN$
z*s-Z!&x`t|yYO+MRI_345j}3@%@9_$nMnne-RPQyBe5ydDBmKlww8C2!xG`3Htl!l
zc;X$f$B|+Y8PK*})nY=OEDfuZRvg21gLH&IFxkraoMH7!Nc%jfLuyw~;0#Q^@0{kz
z>G8#|%KFETE%yC<<G16~&JNw5U#`yKaj%smDB;$OB#$=dRbs?NCa(*nl*B^X?g*tg
zVvW$paY^gd=+&{y?cm_lfW=~mo__{iyS8xRmFxXnU%zCLZw6gkj(<mmT^`bsd0-UC
zj2ERJ8wf}E6I>H5C@6g%4+Ncm&U~CrSbjQx{9J}D!dJCzGE@7#-DaVZeL8?*$F@}+
zUP}zs@{S%ZPkcB#HFI$eHND+*EkCM(>(%b^#v~lRS1vL~*AMgnMu}u@ppT7R@}iyE
z4)&DcRt|l^EuSXlNhO<7&LPbp%_UAJ&MHEluYXC;_@7Ul@3Yn7^>F=or1xsu<<_yy
zJUm%{d_G-2EiZ1-<JHBs%5^-X19@8F8MOc%BQWJBmdsaiXZ2z*d?lS{J;Ps~U+yyl
z9su7U`fIAhuZ;YB?G2EQu@Sq@MV-OCTklA-0L}P2rSmQ5Svb=WqvEq}HhHP#8hsez
z{g{7R`oIMJyuWR}{jN`4zsyYNm7R~NZKjol9nOEGotG`@?QHq=ez<(x=zl)o+qLcz
zO=h8wH{X}7F797G?>h&J@lK<*h&q?hPwqok(L1NWsg-c^6$0~k4nT}Qo*O%{{L>iM
z!O68%A#eSbiD_K_5bKlTeeT%heWuQ$OA-&f7Y2etnb~vTh7{g~Q#s7MlLChAyT!?k
zHr9R#%!&cxBKN;<y=U)xPxX<X_HppoCb{l9QNH3g%mej`d4Q0CPGC@uOR~0;z-J>p
z{B~X{(DTzEK^dGR{=;mzwg05D#iuFAJI==FKO?G~kYNMNQi1{s;2}5&uK(I&r@IR?
zfv)2?F;KqD58zW=hGXb356VwyEb+U_x1y?#LSyVREszoq`-|==uw&jU6b1zHs`fGf
z+2SRXvST&qalGiobf``{0fwyap?vZP8Av$d<NG$?U;`6$4FgPK2)m#wbriaeDoej8
z2wgIZrfdnNWY;(Il0K&6S=zVhmX?JLD3`3dYKv?C&4GZrbO{{nh8LhTQ`JR4kw}Ap
zS`1r+`1gIHQ&E+b=AZMaff4~l;-@e<zcyQvN)pvnm6VwrB;S#uo7Ee;<IZ&#+|_=r
z<b~*&8XtKp4cQv6SUJlP_^IiWt<)0FBt5s2L)vC3Dpr{#pKLKpyeW%}$?BPq)e}I+
zD+BVHeO1KPip3kDW+NNKnl)AvuZ))w9_Wp(6umJNkU&ZFh$P>rzq|xb=+(x==>0g?
zzPueJuDnq(W;&x7w+mS)<J2M1OCW_ozpwF!01IF7E51ubCDyA%brk-Lr9&lZzbewo
zKYuGrjIDjH38f_5Xqw<JhpY826w#~x)b1@B>bS21N*u?=E7J=l&VD!j8>t$ZUd}oH
zGoF}U&rELSr(hgRLfadm4{Yb5kj3c(Iobw;2_nB4(KJSg9zETXBd)c9Z)1I!78#FQ
zIib~8H~D@CPOhu;#-2~c&iC$&3zI3>DvRmnStDbV4TX5QUibbmX-3vUI$fQuRdlKh
z@-=TrRG)MTZ<hB%{T*0`;E*(o`vfDVYV3?BiKhh!a;n9*4H~RK9pGvc*xpZ<y9B4<
z%*SabGepX2K(%3(hD>tolm(AJ43*6S)vqJq>T0XCm6C`DfFX#$Q1%Z4c_`hAo>b(O
zfu~;|^<{H34aPzWv}HLk`WGkjK^%I6YFRm3=&<lYWe6-DBGACxc{=pg07y)VvaAvK
z@gOx3$!h?2E4Cs?4=_hj!Z;{i^+VHl`eOl%TcUl@Gv9mxtU&vcUt7dWuFiZgdgpzV
z4;lU))}&cpsh{*HcL+1QD9&h#mc;ql+B?;SFy$TMl<5t>mp`t9u6VaLMR6!I&$$(H
z_ZHQ?c0E5kD7y)GTuA@#!@884O$jfHGHxb$?1a-gcj*7USO@Tcu^0poL>UuEMHp=`
z&U!|jFy5N~HeyY(I_8iPDu*<E0$~ifICCfmPzrPyRWTJ@B-f9+VeF?Ma`E>7+RISU
z>kFgd!TQ&s`s173(;49)QkElE^;99(Hu}T+KS!z*G$a@894)0|t6N{0>4AGy|1khw
zJhBc}-F^fTp>NIRm=i_OUgf~(8wiInR~5~uOmxI4+=E0b|D|A^S2TB3XM_M6ikbS8
zpX7R#Yt}d|f*dZ6df&~@=KmQ3d2~4q2I>g~H7LV249H;~Sw#@XnvvEBf=P`T9KU&m
z;gzpB<}O3@vM~t<H#o{VfNOiB|B-1N;1ry`WZoKpYlT|xodm1F+z9cVx~#|@v_?(q
zPzh`qEz3BIteEId4>In*oWQ^N%_Z9qj0&NPD3@QlS0+UOGJwoL1jiT%Rgk*s8iT{=
zK`ux$F#NYG6=!S$l*b0gA^?{NI~XO|JPe+Hc^aW9832`(esBg7Rv=kXV&TO;#Qn+r
z4>-#BHJ2bn0jEAmQ5lE+AW{J-5TSWAm3bU1-6A}Rz0SbQd5*(hQZC@7QT={%Ojgy<
zLLKURG&D`oR2^!d8s&QsX-fTU!D6-7W1W``jz^uW$MEl1ombpStj{Xd7X?)3qx;!4
zyvl|sDMP>X?fT3*r6G6HGub7EPf|;B=K05rvv=vIuk5oCE9q0k+!D~(D2xoHLtGX>
zVIo?Ey7|ta2|^*;tQ0rB44=)4J-Kc-vDAL-e&W9~9Xfg$lQ=YRRIL32K8R$XXcomy
zhzD?C<1wP?z?itoVw5|`M6JhP07&RFj@J9)k5WKYYYkaPtZcBttcUnqW*R`1^9N8P
zT5(2KrJzG%_;uspoBJVgN$|lR*;`;hNptHPynn)pHn@Qfe}(^5)&-;>)|D}c2<A%<
zQ@2v;!fxzCJy@o2hnqM&>6&-e0=UlH{t@q~W(2lJs5ts8N0Rz;@$FY|Q4_cJ3m(_v
zf$^V0G&?ine`sf9{O>laum5qKaoCZ1Ue%ViNcv@%(QNDG`Q4;U^#4_&%p4{eyc^<*
zmc4s}CrOcINSU~pO^xcqX=scbAR$J(d&4N^`SP7F#<|7Kx$WwO%gK5G!^zF%$r1RO
zzk|?*tYqW(H-eO%(wF7&na8N+ig~D85qbOrJU_6VdI8b1@>}{R17+Xh1|`QQ$1e#G
z+tI>Lx}{Cgb@ZgX?0DjOxli(N=lFU2oaW39TB2HiVd;I>$<4X_c3g_I#il^Ge>_j=
zd<8+GcyME)X7l-cUq8Ly==qdQ30=3Pny8|AH>ftlMYX82u*lOGAJAh<CHX?dpm#JI
z{(GLwC&p+_((glI^tGOYS4)IQlijzuJt1tQ5XZ!0ym6Ql=zX?>o0o~YTfHwMgcX_~
z0SdtrGWyAo!SaSzdW^a?oEUSw<lYNIEY1~6Q`_>I)`!>BYUhlUEIf_iNmo$U`N+z)
zuteg6fon!n=|RagGk?T*L(?cz0+=V-01-8ilI$=}<5@~oS=#@v%-|l_Bc6s-eP-_%
zmNREQirD+)7=f!2k>M&HEG1O$EuyB!=!UGu{Zj2qC-x3qoLgkX>m)Kr_m}jCk1f5Z
zSM)43JgFESj`Hd2SalGmSoGpSk?P@Pbj$$tj6s=Wu|k5clo4{lf^S_z1QFOwG_G3o
zKzRItGe1ZYIiXnJKh*IF2|v$u%1jcF2GmD?y@jJxxyX9`q%s*t5mK4w)C+EQzxSl&
znmj*%jA9WbroT5P@|tG|N0dH{#hv*$g*XbFpf;J?K9JbhLhG$c^J-`NxeG;Cq>*Mb
z^p7}>l@~YA*alwRzPi>Yt-W;~H|Q7_^zXKVb($mU*v|AC>e$PKl6|#PueJUxn9(0b
z)AafdEe7d!q#KvXNcWr^LDS^F6UmHjToYAnJOJ&{3f2x1ovQ$8>`)%1PQcny(^pt>
z?5Z?%4YRbyUOdXNx*1-_wOuO>1AJ2W@aqd{xkIJ8Z5Y?%@t549)Vwkt%wd|~j?{Tx
z6SrxOhET<8)2_1mf$j{x$cok&sKNGoh6~T;cy{16a8T}K*cUS*#Z}*#BGVF(YrSJF
zBE*GJK^df7B1QID7a`d`PmL|^Fi|31zS4q<&3JrZDVd2Ck|v!aBk`wflAp?*M8vXg
z2@2_BByG|@s25<b-4|J06oOm}?3;M-1YA0CYT>eV3UV~AhPzv8Nc1bvy-Wx(`@3U9
z=HQg*?qwGt#m+5t^M2oNKpzPZBJ}8_r(l}8H^gfYYpylJitKUt<d~*Ik1uzDtb=l<
zLwvNtOV|5&>?Fon?(|m#wVf>Y1tpOCOoVJHB0Qr$rVkyk>^gzaj}}tV1yenVS3*_6
z`AYqg%vv%bjcCwb`|V19Pvn(ifT}w@SjM>Aq+qP|4W=kQze5bfo4F3PK>p=c^JLH0
z?awvk%sWS9!rMWSOtN2hKbF!a)Srv(fb^YbcYsni9^HW_-J-$g?H$;Clwv*LsbMHc
zH_5o8FijmB8q7?bl_6W2s~Nfy>F$U=$`Kt|tTE!~0#BKVf1py7|Jlk4O%d4TubqA3
z4hooGRH*6@>YBh&X2C3RwCpCU{VM*VO#3u_==77kOT8Y5^4}W2K#`p;j;=Xk+PH!U
zE{rcdSDjsbyJTeO04GLHuI)d<MvxbcM+hSxsJy<1sJHa$22;%u#Z_OH51mNfXP>p4
zxLM7pS7L0g4%T{Zi5qHQy!?W-hcP6A_ziQ24wR<5|8=+m!_kuMhV*)z1TnLqycboi
zauBdC+V_^zX3`b1jcS|oo{M0c^vBNNHI(0ZEFiW$MKft5nKtnS->aB_-3Xy!O4S0g
z?PRU*Nafs3D~4W`r*9#DpDemT$w#n5HvQ1<Qu1H#GowiLi@P4f{rh7D2BI1~s9O#y
zWS_g2Oot0}7aj4bf_#cvBtWoTXqFW%Bd_h-!^`2;!2qT~EEe1H$p4ICCW?ME6NRpM
zl`H(h!93-}V7}+{3AanWp9u8CusWLW4AJANrsH)=#1lH;NM(oy*%lVSUf6H#St(Sr
zg+!HPd6>$Dlc{L#ERpbyb;%R$O(y2J-I^f{ayMf3tQ9NAOVJ=_Rum@XVuYjs<7dv(
z65qcKPMh8;6E4U{u@r=6K)f<yh`P6NTy(>eT~QtHjq(>IS|D)$E5H47dYrz+o(I+(
z9#x6`Bk?R`xlPFyjw=2dG;AvRF{T+iq#`t<atxQuh+o6Ql&FR@<C>q^%VFnC7AE4h
zvtEfUQp>F{!@6trmeaCEeh`&qk`Z8lB5Ze<lAW0Eq$2zFczp<w>~iUAjpJzfl8H=d
zYT%ClbBy@3lP$KU5REP`e+id&4*tCio!}Ff4f7Z`uRp-CmgE7y<s1!bfE$u6%7HcG
z1Y8lzcH?9FG`rr#wjl!1udsHaPb6f^(uBYW1+cm1r;8^Zl(cdVERZQ)!<N)K((IsM
z9c|77w6VlO^ZVX6Nzg+PPsi--r4}H!4NL?!@<X64Ca!T9`ESgq7y#&CsT~6^6yhtM
zEer2x`j{;D7GAh9BZ+m^S9eWpd-rA81>Rs6erpIaKT5wX_F${MXeNY|`aeNR1O51z
z?Opu2jqtPtqSBkncG1ZnDhs{#nMcXr;*`-o<mOf&)EmBdnM~ubEneY8RSPp!M4RhM
zZ_(z`7KY^O?|6aZe7oefps?)&^arM4uRoH0*asWjE7$bVB-0nO9Gki43V^ZLAO+oL
zWggX;xiLHEraPSLy0%{BCL)I|aeX@iS{{xYp!Y+8(oVlLb_O4QKxj=mn*TA4{0F!K
z<A3d`){wE=<Ur{8Cr?_#nRX0i{^ExQmT0A{Z=-eGytbIXD&esX0DU>VEfQWwBAseP
zX5S7H%vvzV3$qJ-#M%5>UOPEFxp8=W8IVAhR)*T*=4B1WM)R96T(Ck8k9J$)_O<f$
zv!nIWc55HjJhcb)j>T|kb3wTW*@qmYnYYCbW;kRxyn~?@jf%SGOzsFXyfwi0X?}-&
zZD$a(ef@XJ5<{#1=@9p~B0>NQe;7baz}?Kn`1yr!;p#&;W>{WBOSa~DzMLY9za;^n
zUh#CLesnyBHISOiWyTF4W7b#G|74xx){_47@9~{HJC&#i1?awFuo^P2krjx+GW_<7
z`yaTX`S8EMVM_QsqIooK6lS;HSi#vj3VB%c6CkoYOe0{_@fHGvyu?@@H5SL5m<?Ro
zTBF;w$F*8zoj0FbTqSyCw<jC!<{e?Dq*qLjN3H2^59092P-pVe#k<3{wLUNV`dRd=
zJ&NznM`Pli7*WHH0x_VtqYfQD;}!z@KBu~*!0oq!i?m(wbniH)+}_o_KJHszJnMWi
z*6{!{dAn@yq48MwibK{&5|7_fkInBjzE&wAuIhrCcv`nAV%(;7%Ea9w#NHHyN&UK5
z%eHZc0T*n1sc^|+6j?)gsRu<FfGz8wLIcbCkaop-rWMoTfk$7)9!*(!X`Z~L{wzfd
z1HzB)h0Ks66{Ag&k7h<r9!?T!hC>PlgT{pK2#7fWD!uea^W2<qs!O@A<<-^umeShN
z4yGu|*p>*_cx#D^JOo;!K!QGgoiCtBy^;pwB@98fHcsoO+pV<9ZDHS_{SVC-v7goF
zw2Z6UO*)WvyG!VYpNw9yM}mH6)QuDpT%KrNw(%^vJ<~ZOLYM4v>^ckrS)9nE#I;QC
zz)}>dVmK98Vl1s6ZtDr;K_zlH3b*2?HG%g`0b`9Y^uj_MZ(^#yTQqDW^F?ZVRCh%~
zFv04MlPekN2b=?DS<=bwrgs@QnP9x+Z_*sF6M2~4q85#s1BWJJdv$N1Tr+<hCpZc=
zpfnnBAVv%-mf8Bc))QYpw|Rk(n+rK)o7nTBAXU4SKM=mZ8W-$l+^n_8Z4wF;6$kUn
zY1LFEECXZev^1y@Y97_GBvw~Ddkb@`O_0Xhlj)XFjV@C<3z<3-Ys<T(okym3TEt?L
zd$Lwh%OEI#R?5ED^FB3Zjji;mf_N#7u|gqXUvZZts0u^Kd~B#<1)6bZ?#@yI4pEhx
zfz;BN9L&2@!-=!3?(L6`$P?S>4m1kP#<iBzhEV)WJ-oDTuGih=mAJ+esl4bCTR9ID
zM~*ZXR!k71Za&7uwZd4;+W1t<1bR3MZJ_lgHe9=Vq<5e`OyF4eNIj|yDP&aN0xw$6
zma!HXU9e%Xx*&6@D>h~8YJ*PXOXfG^^1sbprX~U^&Y_I8x$ol*7*}TXbeCr-H{PU0
zybjh(@MZ3t3o>yv>y(>_la?)Ajay`AR$^Bw)+ZxfQZjKhe0T~>3q&}T1NaI}Pf0UM
zTHDI=H;@-l0#n21)*Ge0F5jnVhu&9zOx;IQ8KW}j&XPe~jX9+MHf&jQ2BQKd7&nt0
zt*QtniA3iEcBd{(U+BK4L~f10GCz}QiSh}fyFMFP4I9X4>0#<r)YWxMV|^lPy#Q$@
z2{KG{URud0lE}H^GR6AvDH)v1d-cXg|NAdkO-!9gnAG(2Pex-><<dl!F7wA!%=-*(
z|Jn!(=!z$qSxy*;tWZr$KJux4fg^=`?*8M<%f|G7tOFTYIQ~0-?Y72%?f>|FZvV}^
z`3ysti9$#eX!VQj{C2S3R$f$@*^cJ^K;K`yipeyJni`r~Y#Tw<tRsZ@yD(voB-=jc
z&aaE-H~UXNJ8QF5d*e>8cOAssf8V6AVh*_EiihQ2?^-^e!D}?^xAz-0v4f@#-TJq4
z>|j8j2%ZR<k#g+1Nt-d6F`f*<x<*IduBA488qn9kA-Q7i0MR$bbN&3@-ix8tYYL0=
z&R2e3@AscpErMr7IU{m7>co{^$5xd*^(QFr8;n#FtY5F^ZXVCmT$c=LPjVxmKnk0~
zolq5MCOiH{V3>5{M+^?qpzRm;At}(&exbRo@--36zc0)^gB+U?^q0PDK3!#VIOBg3
zIv-G{{=tj7;Z#I3ALZNab(Wbb%M<lnOKiN->JLUdAB()7AJh<WXdM)3wrKez1T88o
z!z9rRMGiDf#l$<ubX8V=1<ku=JaO&A#xB(M=tIn8lr^xM&L@Si1yp_i4LtNJm%7&o
zbm+AT`)gL0J`FgK&Y&}i8I6l3_{ZY}5Am}8^%_7p?GLJ;u{b{FpiV9nE${jmo^@*v
zJ#VkiQ$-K1&34A3>BJP}|HIfh1c?%ETe@u9wr%T_U3JQ~ZQHhO+qP}nwocc3(SwfY
zJGc=!Nept5|M%zKYp<nhHd$x{k{`uH^E_xcqy^y}ARHJf%-Rs*Vl3mjq0+s(3+;+X
zXHcKaU8ThKHu2Fpd+Py*h$lq`*SW|-)(hxU-@y>?_^U@!{Z}PNQdIY}F2W*xlv3S=
z#2q;5f|~;$F|?%W+p8H#3RM6jRvDgab#&D6h^4$D7L1`$=8-KSdzzihOK12nA-aCf
zWWup=>wsx!7=PuAiG7!6dbifysejAPaG-XMcDBcZ<3(*g)_)tGe!#<Qbe3SC_%Zpp
zb+RpGock$Jpf2Y)W(-ySV!fcn?4HB&WKO5M1;16=N{$@G9OmlE*{?-VK$5nB1+AkE
z&|he0X1HWq?Qm8H?Z2oVMvY<m!R)Z|VGf$Xz*c8&iV5APd^F!M>+sV`N+h|@jKW1s
zC?|qPXm;%efbB~rE0_}MoM<pydiU6Phdr5ADl{jyBt}8LO9GNP?#Ri=3tWoKiyNS7
zAd9jZtY;lf<SYfjCI%j?$2>^`^4F`85vO@zg&i<0Dv_W~YH5pD$BT)QVN_gFL`?TM
z(Y-;-NpIdFC7Z~$<PvAqakZAylx(-wo8wPtb4h}EnWU2DNlC;sjVRvb*MhOq_Vz0)
z%ehoxhlDU3cku<&kwUvgDn_gYPQJpUgD1VPdHO;?WB3((F&}Jf@wZO9c|d3#ODzez
z%v6Sz-E-8cr35eh_@O8&d0C%>#Z+YgGlf;?$et*uDLo;!2>O$Wkx?7kl!8SgdR(R-
zCNe7V5SCFbEgNiCumXJK@N;5Zyh5|BU{}`}&+gGH1M~iQENTweawoZO@m-dNElN)E
zxXHe{x;1jnwa8|Wsqyfjh>N;2kuO)d3?D2RLI{-y^2HSv4OLVVj+oEh&e5lHJ{D{Q
zj0H}!R2h<FI)~wpL_Su7{Pw<qN#^kyA38HJ<|i+nNuU>rOKE6)t3z_j;Gj4rRPkLy
zsAvk$)C~0|hugq&M+E~Hk#6YV6V2$8tiTRiA{#+YxuyC7H?(hYSy?sv#m2}Pfx`UM
zLPUgAPi2gs8ae}ZGp**xpj%gq%sbJ8W~d?TK{^(CT#|en8xHklG*&Ak7ZuK>y6T|)
zL1vT*FOIIzW&(Qf(L0QFCR8-WQE!1Y^YC{ncYB>M8TT3HM$5@Dg|!p~8&AdM>U;}G
zH=H8c;|z(*YokX|o3((gd~5VD;6uxBQ-!T2$0*d6^&D96rh%t*I5b}%WGLuSNnqW`
zT`yuAkj%jS27G1gh-Po4EKQ5AYBq-TSA~XL<9*@9kXdF^wno)uj^KKP!p0PTh|bEd
z6<l)ci^QhLAG6<e*vh-Htu(lLwPChhVzuKpIZ9%+VfTEC#FBU%?vzZ$a*dYD^}SgX
zsX4~-^u?BTV-dadv)2u;efmG*f&D)?Z*wyJcUB?Wzs{gI>@dHnrc;b!`l2MyAPJx=
zvTXlD2PsF02(mH4(37QHT?a8{YV#e-TFobhobB{t@%Y_85=G9B)$`M#nV+MXtK>}4
zfBOSQPwp*&zl0uj=f39FM}R5*%<=q~@paE*;p)D6VOEE~J$!NAZR6cQ0R01cL+_IV
zgmc{i;3xDE`oeDVTJ<9IeK}vJ9Oq7fr!&Lc_ow#(d%64Fnqy~2d3!hLFDQ~2Yg=Wp
zM_ixNqE@G{s^g6}m<;a9=+TN*HO+Oi_IOiy=;Q*?#Mu6Z@r2$lLR^a{@paA+KeW5S
z7X;TDo(8Yvgcg0fA&|q0NrwU}P?vgJ2T}Z{4W}a(^ts;dY3~o{OB$)%mF)LX;cH`E
z>#V}%+K1IofapqC^3cry&``lQ$GP<jm8#~$0%BLg==#&J@~_9;FUCTba4H{_tuW=u
z#(_T8TUOtn<8`K~x55WcVIvRZt>XiL+Km*|;V;(d<emMI$3dy-pARXk`cB<f!z9FM
zeDT|cFkcvKQr$PpRx$euwxcO#9ig?W3aSAam7FBPX<UQ3$qX%2mG*_Qf>uYz+0|ZJ
zQ|Ed=F+6#}lMt&E&mW^a9OA@mYZ6!U2VV2?3A}#Y*3y*zu$j5$@l*>7S^Mr;TzW4W
zG7cR#3?VsNDd@MRBjwEunOJI~lf7a7zOBP;Sk>Egt{t^fD+W@RA6eq#$T1QG&Q<hd
zEx5{e6lNa=VUkruqVDl7>*@CGdbdQOMh-hP&*1mj8iNK_<-|{qr3x_j`C0esey_%B
z4wZDaMgND^y1}_`mt%;t$W_;pM*m|-Zu3mQ@40(S<ZV$-Ig=XOAlVxGW-3dbi?N;c
zuSuRR#|S|zAv_i)c+sjxa0V^LR9j*y<N9Uuh!dX^XB}km<!#J;EcR+W=z6;a5f~KL
zswo--4CZ`;t-fD&ExCt2&I~KvQYyQC2+vtAbpIg&Av@x$Y+)s)0@+ax5|wK^f7jx<
zbC>V9nR`0rCt+*j_<Kpax_Q(Q#l8R9_%WIJ<&mqYun8@QMY4y+2t!?68&(XIZK|dQ
zW_`?F(6mH#RqJE9b#o?uM=nB5$CINkc<h!Zn1r=(49l2`*O&g%Uc;Z6n0M~7eo;{t
zl(zxyy1{venoN%J`Pc<fuAd6mq8FVSxmHLDStNQwg^QmyBYcBu1DzH`7HR%A3N1ik
z=(7}eKY_RKB<%_s-2o10$5=(tW0=Q)(r$4-gK-w1v*zLD)OZud6sAlov>T0?a^RF*
zwz7C^RELBmA%FrUl96jBfTHGQV%&`AVbWS+2NU9D30x84NAbdRL9;wJ(yGLVt||T_
zIVXCFzIe{6(1)%${v&voKHsOtk@&gwk7eFZjT!9HE?$lQN9O>md~ec*<afDUwLCxi
zy7;%2(Qmjze9lMLu37<~--CEEcuPZW!QZp(;H%kU_J&@B^@$%LGwe@GUOUgwonDVm
z7xmn@#CuKg+B;67+~@L#BRW3j<*?R<;0076W?8?j8~_54;jt}DfFohRz26ghQNR+Q
zx-m>5QqnL<i|E;q4Xi<<wv)H1NNN>CoqguOx%LUiwf*zU1}1b=+g>S&?kXCU3UXrZ
zz&<AwqFHbq*P3OV&qFa+uRiw6V=FSs+_Xynv0CLxo?Mdh6bncO6t0OzxwA`uhZjhJ
zS8xAQLVoV!(-KE-BR8rBdwZ^5GjS6mOzse+>=v){(826QhXx?pTn+N{<Bt*qJ|SKU
zF7n#t-=uFM&t_SW_`iK39_f#E#n<7q;6TcJ8PJwp#ZMzC!lX{cShmH*HTRPKm9PTZ
zXsz1XRirTzcH+$qJp`(#H!CZiM}44D=pAiIJ7))yXOeU^O;K82T&tm<E2EAbtY$k8
z{lzx4MXeppC><%9QnbXa5mGW#{OL6y%dW{LS{*3_#?)6V+PxAXC`pg>Nfc7X-t;i^
zm%+`YG@+>l)p0uu(9mX<<uEo<2o3fm5;p3&m$FOANqqKAkmN0_b^4mf#d!xV8UY*k
zkZaO2Td*s7yD5p}$~RIufnjE8*ZM{ukIs;$o`1b`81%@1?zNPN2@AGao4dc?(#s+7
z23SjfP1vI<l{{T_PCL!5nRf3GN8S*#LZAMe$~4`pFcSh_1FU&j?sU_$Zy797gqt17
z{EM4IIy+_mgjWX;u5Wks=9<OI{rp_99LMvL=i)1lr5(k&{kQ0Cl^VkCO@G$@xmFBc
zvpfoR#4#xKTz`B+d<5RG5I(ywiubXa`ouwei1WQjyVa!h>t|-ml{G`T;;5g^jI2}s
z=Eo2$84jx=JJ}l|WL7-^8BwZTscvb{086AIL`}+eZYK^&{hRGrRbp*7BxKP30Xu&u
zOO1j=?M)Ax^}$&Y!d@!6Y3afa-eqwK1-_`V(JpBwZ$y?)@K2dTpV+ziYis0o@>u8~
zN%>r@b`I<PPfQ&!&q8io!8F}nUKW>(^`@k$mIn$)V77T~dtQ~e^qUn~O%A&t%4J&)
zvUqOThn2fV43jOaM5^+a54Q@^UhyTqu!Er3e<V8Re<C{PfBD0w!Tn2ggqvsT7AL61
zEKEMYbAa&|OmXx-+6hDu2E@UUZIVCUAzBYHM?Gz)-2LM98s^S}m@{5kqpsAUvw=8#
zTfDrx-JA(nRd!f3aI?8!d%og7kO}x$ylMdY5Fw}ZgY>V81^bub^rWPSx+%V{UaO(2
zATj)aUyysH0HLd@fd7kh%q6c>FT#H|i;ac}>@>GgiTc5<UhQ=Y`>m(=w?<5CyK-m=
zXfqZ4YNTs1M*E3N$xE4q{r%9kb7dC3lF-Liq~8~d@Pbd%^Bd}jutEyN50y(y60{Tx
z8VCZk&$Aswz(SI|3Lxt7IjIdo9u~z%A?_)~P36Ttx_^(F;$~Kf(p-y7M7xkCIyH_A
zR|YnMTaTP50}}jCh20d)L=&&42GIne;Ats}48+a%BK1q}>rN0v$dhtLgY(s;!a?#Y
zX}>L52gJoNid}`m;j`!VUIieN5h0q8CClQqxf$V?JKRIQ_4ENHA8VLn8Y8n^w}pb-
zpGO=C`}jaoYSTik+y*cuaR;06Y;J>rs-p2|IZA(%|I(4;T&@hg&4H+jJdx!o7jpe1
z%V$8AC@!Q0nw4#VsEWCMuLkgUiCMPtU=&<i2)6s6kPJ?<?0j~`2MV}3z|y@O6R<T3
zu9zSx2Cy8eVe2D)B*dYPyk*-7f!>$h_#snK+kv$*gk0_GYZIP=nW5Sgp^hX4xa-mk
zO~nfgNleknP?KZ>tcoyX=51SX2_kU?1rkFpm?8_kqR80w!>CUAQjeaL%~^{I-1{a+
z%Z)zcR@K03A-y4>T#m9njwml6?|sl!ak<#Mt1msNnQ4m0@C|`#T5l9(%XipmapFbH
zarRUum}Xiu!>jcTD|gmCCRtdThsR!W?-7kHS0Q4_-IuB6xxj%_qUgr3!0N8Y{q~uu
zOE6J}2yB|NHjd_?9K$GsIYY2V@`eb?=AGxM`1xY#eE3^V#AaING*M+V2XUv1b?AaU
zv$e*zCPtGJRM1V0GK{8RwXBS;!swsuNw+VLG|p6YYMQ0DvJ@qGp4u)Zszu7%!^wQ%
z%6I`fK8XT($yC^kq`1dZ1l|9REbRsH&Gxa+q(!&ng-^mCcyqnHk9F2Wl@A;Re+XqQ
zFCcxB&EUXeM=Jw1vEG@5m=Q~uXydbIjmcXU9mOBY_`A>bU&>+Hfd_qFQ;p*iiMK8y
zC53Q1vUWN;Hh?)k(B)G2HWP<I1=}N2ZxDyYLT1utJh{OXjw*A}O8s%%fe%tu^I?e|
zYkm<Q!eDlBj2$IX%Cy}{O&I!AoD`qDr1EUerh>^vwz^@tQjFGCnGrWa-f+1|$V3RP
zYnY5ek#rTGT`(&g;bIvlidp52zBptY`)-=?w@(*{s@~hII(l;Ap&B-|C|`6EE$?qC
zoz2h!BckLC3FRL^&Nzx6=NNQ7B#j}=_@33+mJ$`RFO}1WpupNT-{i}zKwhq(n1WFT
z`*7mP9myNrUT1uBPnr6|K1GEHVBU*}qk_pbgu#c_3@j%0y7ZcM<ddC3wKR0VTkd5)
zp2jiNjxiG3)msMfz%<~#`*v_|#-t;Zyn4|SB#x+R8(=$lE&RCpdD_`~OU9T;;0L{e
zuX{7Vk7A}@b$ZY35T^SNhj06Psb=HJy`gA<Sg43mV-<o?$dS5`W9H~f6?^`6+^f$J
z!ANi6+YMgoh?zbKXNI3>sFECiRzL&_I~N0LbeInKJqR1TdZ?h?q*o@z5n^_&T86<k
z_Ic}kBb-+oW7is+QGQOK)EY$(z{5bZPH=WD`A3IS<Z*0EQrU>m<hZimu9I14(IQ$W
z^&I3>gLYQM_z3u9v70t5Jw8HW-n)uQA%*wZ8#S954L|+jq#m;vS(VMh5|4@_!H)1=
zQ_>ZwH(8whUQ^;3vA4Rw`e3zegy+h@n>s%CFC~}V7JR%-S5KsBv^LJwJ&z<l<&byI
zpS{A9XS`WrH~%cu7k9jPVJq^!wY;tM$N-DCo)6c+e$o`PDo;0z_swYAqbT%+K@uzH
z!7ei@@a(4CR_A{5t?mSS&pB~K3>UoUOA?U#Y|EwTHI0NUT7hn-mYQZCV#`Q4ka4N@
z8jX&`ON$^b%CJ6N)&yLXpr>m2th5m*M@Zm(x3j$LPqCOZaOTc?rN^xM(pvUnA_+vW
zk`-&(>r=}d>Easl)gPX@EV8Lmz8=81QBdA>=S;1XHV{fZ*RgP+1UGj%<W&>iVGVx#
zx=1ZiYtRXFg1_-z6k_z5yB5Xi2EmgQsz>#Ra+TMzw#iQL!&nQNv`)lSz4H#BC*`l3
zh~l=!Q=Ly#p3*n9tli2xQ1qDb7O(3{B?GB1J95`-=*YimL?AhB5*fAx4Fh|CW<91v
z)ZX!_p;odn>_iO~x?2%rKz8<79+*VXaokr2nM>i!z34=hj7@=K79=(TM$K&1$m-R+
zt)vYfR+A#CmYY(@>NAj{tHAny59QClbmP=Q0>dsAQqgAzx^L;wkiKi_(1fyW=~gU0
z-YON#Vtc4aKh`QIwo~K&e$&lvEz_l%UR3NJr?+n$KS*mW3wd{`P&NvJl5SMMS&e!Y
z=zo;Yj;vS~Y%#U=?6^rSJ=|gHgx3PjOGT>Fk(tC&^W>^ky6R;WA<dLyMzzucx03t=
zWJ!`9`%jw#j{h*i%E-d@->HJOHTq)?*<iX)skb&^Hs-}qC?)%d1eiM2n_9PLs?W;L
zUYbH*mS*XV$74vjW3tCV1mUsB8z>XxVo}bcY<?`99^D&#w|Ves*s6FiLr$JG4Ua-n
zdc`J-3PYk5W-?BmJs&s0tNX{-CIzaillvdmCzD&`R@~eN#*oEKqSyHtvC`|A439Ah
z+AS~%aye&l8@iC)Z@XfsyT<P9ldo@Ud##+<Ufu<K78L6uw9Ie{kvzGi3RRJ0NT}n@
z#NK0cn)I3yq&p(FX2f`B!GDuR$xEY3e7+G#7dwX3(Mh{0bO=qda}!RhVd>QmZStJ4
zeCK2d+cc=#ui~ePEv;$8#(ZfQF1_oDcnpcT%Q@P*HguGN?;t_qP&{giSg%ZY89*Xr
zXi*3WnBm)ijkBvsgNYh1aHw%D2))C)j>5EtymDWn%%J1bPhI%jjsvvM+07LD+ayM{
zZDmgIjbo?@Zu94%jNPPK-v7K9@q9i%(fKbA`0@u-N48<8n1r-4VhQ!c0{4jFDV;R!
zW~2n6T9^3O8HBBJ00AX_jy3LS!bnaoaiRU{T_N1Cy#Dd+Nzo~pz0thka|4+`Z*`!M
zOJM$1@=vP)|LNE1c6RZM@{AVptWxu&UXFo_I$di0HMeelj?d`6`sT>I9E%4!f_BuV
ztR|!xcTS!oSk(?R)YhqLsybS0FT37Pd~5tzQm{w$<tl@g`75oq8N&&UoZ?3XCty7n
zKnzWC4H93bJ?@c=v!0ONobYs%t))E|2awtI=-hWgM+B$>p|u2fcwxeX=Ro=gE7_1c
zs6DpKJD{<ef<tiE1ddM;a*$xgdn1Es1$nDe&@V%op(P~4ip3Q1%1_4eTK4zlUCc>w
z*-;MZKP_TJ%!VHqu1}@Epqb*5r-0$8qFh3Mx?;*x7_c4_^du3Eu9JiCUlr+cD{XUi
z*nMMPkb-H&*S_w(k!|w`6|j!FlKK)kJ{sR!-W6*#L~XsTsozsFH`>#LLOHi3RG(oR
zP+edr50NgV{AD+F^k^4z6>an2C6X60sg|bi<*)&Vr~KXt>QSmHufTX^98k>Cv667c
z16_`z*g?4m32Zwq+H(}gH<fZ;$}PKFtx(K#x#twB9MIJiOYfRJK8{KZVs`CdI$VHF
z_a*vMw-Y;UofVLp8_>`?%zz~woQTm)Cl?JbE}(P1!-8hPlL|FJO~NCn<k$4`@Apeb
zX#y6BXoyNs-uLTyC`%9Cv;adG&9XZsnr2o_&`>|)1+NFVN(W%PzhZvU7QjK-UAzCL
z$`K=Pc09$2S>D<7&1lneR{;pSe%UIa7bkx5k|cjZSBNYjM5ZLF6o8}`5rPjyhA$XX
zeBarYL}WE95gDln=tc;TGgS|(mqSz0_VUf36~IIoEYNIG$!Y|2?N232l1duxs!*^#
zx_~JT1evpxH2d*j8SKG``gnN(`GNCB68qW2o}qHIWY>q9<D7ESO;!PQZ*r5W%bO3r
zBA+A++Inpp9Zzs=<%LqP%Df^}%0?T*0=pU4zc?xk>b29JT6T86e8EshBRcJT(*l4{
ztBl~*H!_uZ3Ie2d69Awm<!jb8F0q)L4jXF%IRHHxuD3*uDavsPH|I-*T}g12GC16O
zWYZIfHQSzf33`|U^_<SQKE)iaWZQ<%dph@wG@2cDJc3e!!PH!Op*wYnEr90pzM?(;
zXzWbfA@kN65e17Emewvjw$RBwo?GzNKFdnV)ZlWgBb2<@pV!n}xGS|qC!7x1=qyZQ
z8W?kK1%*YzEllM7lVLZlLKV|qn}Q8Oe{m2FH?H22H&7@4+s3jzmVGgyNX)MM{&<M|
z*;t*f0wzP)+Q$3{LM<O2t<K|BBA;IC7yS|;`p}KV+&?pj3u<LPtN9OFAm$GPdzmM+
zsO#OJtodxjt~1r25?_VHh*N$MEn18J$N;AQ-yNEf<-acZ$|P?7GT?u;7z3mNe)!Co
zvu^D6CLWd@txct}?4>vvO1?gV@lB$VDaFmJhbMi>w2&Zg06=@1mp7WFowJ#vr_+Nw
z@~26rc5A%7P6Y5xU!b(?w61Uh>cHENuQ!+A)$<c>NrD>K#Dn9e^nZ7H*!PZM<Jbjm
zfUI`8f!Ym^V&Yxnjw_7wyj<RQ_$D>uUso<U1`}cMM_*=7b=kSFKkH2S%_*^l*_>A5
zkvt=cN2~!01xZR8%1W|f(o?d_h{)wIjKUa?9*{Swz1BR_REAUu*8&UTipEFU7o!yL
zC%j)Iq};{i-Z_tE@e{3rLfo<EBU(KJJscYa7x{#QC*I=)yP0-~e8xf)RCnkV?7r3Z
zJ=Ajak8zz>@`R3fHVaQKtgGX?M`fuK_F*$`LU}NEzaQRN;)Wu+ELm_3i9hg7{F0yL
z#||k2&qMvl3CI~v<tUA+ZG<jC`eKqnZD3_2O$|YLsiW!_Xc$G)R1He+fs*wvtF^4>
z()_Z#9*p-&kb*a;A^n&foveB#a|^fe22796x|)X#p<<Xt-r(~e(zN=?yhmXk38Od$
z<wTVn)FTd&hQ?t#1&;(L2*6!YW*u@Y@=9X;uAb8bJy?@Uw(wn>^P5g|ycksOy-+Hv
zHJX~E;1Bq%7&fjAGk{VTp_M>>DEpHL#%^HQ4Djf4Gi`Ho*#AX&jl)o3-L1aLcnJ{b
zp9+PR0m;7}-rg8LYI0ky$4)oPoqJ1LSf*co8KX+U8M{P`;UgPEKUgB>jOmWjmKv7q
z7F#N-wB!+O<NkZn%<Eu=@E&J0;An;R`yS5<fkU~NPxyoxK)z4y&6V07#kfhG+Ha#i
zNLsE;tWVWT4>Ft>Kczw{L^&5=zdtL8Si)EWE_mcyHwAF&Ir*MxaY7&k46o^0H>fy=
ze&&nD+uA1hyoGB}e9c?xmbVyb$t`Z}gU>VT@+hHGw3{$@ETH4NBGm-3$JWTGa27t-
z?dg^x;9{p7Py`+Nmpq1V;FH3YA10wtc)IJ(t4dwRvH&a!FDIpb>}-fTlX+s`JJ39K
zdwT)|S>%&37odX*F>;-yxRk&vQ~hX+1}}j!@tIsloaZoQY4_*On@J<I(+N>Bl(lYj
zTAZ*dfR`M_K}_P**No1^_zxWS<hzM_0EQQLyw#HWGowsoarKgt1bofYqdXhWbjgG8
zuj_SCD5uHdx4hE7kj&e!fF%c<TEZ#vnZ0B}Pn+X*F|C~d?6@@ntLJEqyT5<qdK{)!
z=S68b5UHk@u%L>+bNG>wq;B9wZ#*0Rq%7v)XM>?<gN(ZsZ+|CkZ{+w$y=@I;9==nm
zY(75A3W;$0HtbC2UR#{^EK7KAyx&}+_)umjjyiMY$r_M2L(?>c&~i?WQ_cxPi*rt9
zLkn^aQ(fB%NMVrwk+Hlb$HESzD<U;Yq;hfs)+ktU3gxVyLyFF)HVbzNUo9ksNf5H7
z%Oi~<JI9&&=byrXpiD^#%Td8zLTVAU6Okfti9)m$mLgDohcePM1c`R{V-{`#2yGTF
zc3@gaIycrII3>f=QAWfH_Kng662LMNRa-<FvH)!!E@?zbs{lz`NUFg*X%;SNfk~?n
zPE$x~F{djirObssk94U=-!ufZ1x;HI_bM$_<<~ujWT`g`30m~U=?SV)kZwjuJ1?m%
zD<y3*z+ONq@6mQYF*}bW7{h00bNsi6l%!n3cl}KBBe!1^m89svjgpoS5wb8$OI*rt
z=^voBueAoziI$%ozjG_&nUDcc)Rn$dEeV;$x%1G_mG1{VDH+zFDgtu!(eJ&cGuD^u
zaLC)_tIN;uk&PTFUs^%Qn6B2h`XTvwROW+biSUgXR{rENbTdnmN0piKe}KUd9(ez0
zb!TV(Z-!?X8UCN)SytBn%E4w?^VU|m9rb%gSHE1h1^=$`mVR>?3NocZ)IbG9*&>oC
zDG0D3b2`sRpvl)e_It+-iOiflQDTWe{LAE~@wRc6_f?A@+3;X5Ao4ByuCX~xTiJrv
zKB>5QNkY2XSeCVgdu;uUw$aixpOh|rSyeAp*j=7qDYf(`Vp7ynN)K6*xk^{TOlANg
z$<gwlbnZ;ueqPeNbw6;*n#3P|cw}Y_F?^*y2-n_%7b0tQZ6x`M;&{GgFW%r%8aQJG
z=1xF)z%0*9^0L3+&|LOTTDp*X!px+iu$&`>gD;p%74tLLx}<*hsLky1Z)0NICJGj}
zzNMqtfye*{NQqhzZzd><<0bkmQHl~S?O}_u2E0brk-a1hC+wD3;}uJclEHAf9kVzf
zbV`O`gzO^6@|(b;QJ<`o#SMu)8u03T0vEAO3n$0o9M}w^jCl*@WbNTc>CsdfW|Eem
z0xT1Po^g;QQ>u`~YzX&4)0P^dK&d_me5TP8pkojuhIM10YoWD0gT<s`z@)uXw@9kF
zSBntcWLpsEoD$wPV(PS{BX}!n4RboM4fXg&>5(MHy<a(P{yyO!``{=|X6~)!EkLKp
zs7L5CG+IOWI_oh}WUzH1%YnS4HL*cg|B~2y5PvCb61}B75^EGMYzJVW7+SVAf1vDD
z4)ew>Jh~?5zQ@%THqQzhgw^(Ot<pHsReMA0SjHqxoU-5Km6E^=@6Q$%8j}cC1gXc;
z0d#+sm)Lr7VA|V(p{Q%cP3w7w=M=>xaQ+{TuJ_f`^*e`uhgNUqJ`HKvG&33~f9|F#
z_m{7%evY5)T-fdfzpql)KMTLfA8l=^Q8BfK4Q}#7mA0Un59hD$$H%=csF&gKcs=f1
zim0`uotmx8ZPl%>oqILN2QTTtBR!rE&W_mGdD3%xa<g0J2Y2pL_uuapH#u~=GH~g3
z#1v-JbQ8}2oedrWBv_(#l-9BFxSxJJ-rRpo_;{T8>rm!VSRn);P9rz)@%Lw=z4F5P
z5wxWH|Bl&9>mcG!^+#55`!9oSLb#!5;VHKt;c^JB1NRUt9{Cf`Wxe}hfJA%r*=<9L
zk^r7gAW@hcs!dFHhoSLm!cx6Jl87fTf}+k`0hhAUQ`yRwE9jSyLGfM7@>D}uP~I)y
zz50G3dml&$yg>HbJsiKe-BLAzFw6TG=pO1JD+N>9nqWn+`?&)v6|us5@S0vAIFCLq
zuI9eNo7%v@1}zvca`W1==={8I9-OX5eh*I#-v^q8DZxBM71zsSAFJ0mF$C{0+tC5A
zdMll4Q^=w#(wlY7bwC;3!Ol`8pD=j?<J&WCD{-P|@qpdmJvjNid9=83cX@qSxiNpX
ztZ{?KvEs!y%@7xC-iIx?<4I&QKQyrJFjB@EQ@@}+tDX|X<K9h!*x6Oq5(TO8EMMYy
z$F;%P9qI3p8X2AVU7;_M3c9@4si~r;4kG!_#x)_zsUrI&(5FEwJqrx%1VYC$^4x6W
z)5V;S-<{gl=B)biduh<{<$nL#TH(?1d2`-{KJJF<_I~}m?|-hZ?x%ham-hPT@QeoJ
z_yCA}*NzVM-+$?Gz-Yi#AoN+xJO$!<jhzS14D?T3eG^_^?t`;M+=ezT4(^Yi&ws`?
zHbO-;4xj(bj~j5L@p0*RX?1B%?cy#Y_T1nb`tf(gJ>3OUnjH7XwEd~llP#l<CV(S6
zU%_M`pa#Md%rEdGFC*E(1rAWzjBuupBLN851cN9&D0UbIL72IKzVrqYuMedsNO1{6
zDC*#%<mDDr_b-3fSU*3JW$>($1&2uMFiEl2@8A?6kmX$_vJHS(TMg{Eg%sh#!T0`6
z41c;W(kK3F738>PHm+7hZYaJt`82>8Y23jkkwal^5J81zQDS8fkriDR+e!z-CV7S~
z(%0PBIq%`pTBq&8&Yim5Y-)C`67WJLSl!wIHP_WC4F`AyEctGXQ^5ABx8Es?9+Kg&
z8la98u&|Jl)WnGy7FPTGF$7vbp)c8$s)f+*2g7pBpb!_HO4BFy0=mpj>ffYRdsdXv
zK?zDB1PgkPXB`?|K%(oQ!-;6&-W~(P5s^TL_2mbQX?!=mAfhOzPs*yG{DJc3BSZ|0
zmDYOu5HuVeF$rx#Tn(MzOxY1`Ugc~i{x1|0z~rpjD3h)N0cvt*N?~d<*}uGsJR+$@
z3PlQiL}Xry@3OqlL(uGda|wFj{~n7n7w2T>a8kkn8C6G&!<GB48(ey=(-t8VEvSy>
zL?Vtm3GHmAu*V0RvZq5>;_^A^H>8$v2x@}^aL7=?o|e+Q(~dNKAv@N`f}(Y*E2!|`
z49ZE|syxn~RWY&<=y@XwVi}Q_N`z`k2^4xzGV{^{bS+kwE{tu(HrrTl6*Yk(Z$P_r
z<^49Ky^gS$6g$Q6NCC0$lK3{kbRbf*PBY-;5|<a@;u|^wOREtBNcAC3!twffrEaoD
zb*T=zX9v`H#y%cTx9WwFPj{e?u296s|5U<r8r7)4Kuw`H$x2j;mR+@53Fzn`f~}KA
z#D0B0wNUA-i&-W}B@3-eSuUvOlNZFh&!)7Ak8~1FmRq$$k7|S+RtbwQ;})F7&$)`6
za1~wJ7^^X*!SY2W@XqWs^27qRtj7c9UQ~7YXtM>>luEc$1R_9$lP3e<)p$h1X<JB(
zt>blJ_Tm&DrFoO~HWB?0y0sMfcO*MZWB3A{$?!?SY)fHk?)F_8x)CbG=Wj#+dq<s$
zkuym@Ev;<L9*QcvWf?<uI_N#KO+!CNKhYngw*1km1E`lAIT7^0UW|`?#NhI?ct@WJ
zlBWno@)m@M4&+_#J2I-F_LGs-9BZGSpH*bHLZGe`S%7Gbi$wtOmSkLy0>f%e6QOdr
zDMr|GFYn9rAW$|P7;A|8hDH@<AnW#SW!!LX@qgH^Z;v{96ML0(No|{Yl~$aS&AhEg
zGSi4+159j11FVUgY8F)gLU&wN)ZjSF{ALWwVVBk5EV@=upEy=jIt8nwKvm|Hs<$+?
zs_vHTOmsId={_H=L_04h&|cQaiiy>RdT>N_a$TYzDjj*y#2^#bC4$Is?@fHjqPJ;x
z5DfL0bqoY+&NEk9lg^12kNLep{I#Xt{(VlR!bl$N0#ZSD7q=WsaU?~`K02Fr%0maS
zSQTe0lbw{h;z-RRTf?QkoV5l#Wlb}T*SS7U&b64B41bKlrQ{w|1_YGsvFv`ui-5t6
z=<7I)AcRYneKD0}E^z=Cwdi@fV=MJ~hMs&nw|Jn)V=18U1`fMKo^eYU)i^&vhg#Wk
zquSKx-zVFIdt^^^0BgdtD@$rR5<%p_rA@0s4PS*T;jOOpGc}^!b(73x=_Av*i+ScD
zmG$c2)MSuU_j2MdE-qY_u1AmuJh}U^77We*W$PUQ93B|z_y*eujaY;$QYfgAZe58|
zo=yGHSWZJ*FLdiE#pxs21IWj<1K>dG05#we#kaVJZ3qfW)G*L$n2*&HkqFMs;yj)T
z(_L2&&|^ekcC}*w;Zq)Ok<~gBz+0xRL*>GtXm+xl&xdkqDsp#p5?@r}d<W2|1Xn;w
z7Nt2Mgs0=I)%lkztU3m6R3`L0YO%)<w$^|I??-LinF3B*A7`y-qr;LplOrf=x-TQ0
zUQm0k$Mg*^)Buk7pSH^!EG+-OJerm9zqFH4=T7`Bh~IouE1yIhBa!rtPDPt`Wj@ao
zO%7`(P$hO>L}(@F;@^>mH<3v3<=QHlb(cT@z1@+}bk1g9>A1?3ZXYTbng$csRd9DZ
z_tBI5109bH;esZJ3b}bI{`@XXnz(x$f|o33<xlzN`X%uR4jbbNa0{RVy2h^tqBGxB
z;2ul#lk`RLvwPBPSi&y-+bG`;GWBtT{rZ_*q~}f#P7c6XbW${9nz<{zS9I2WzkbS;
ztF6iol(<epGx9Q}fm+Eo`$I!n@8p$5l7jZm-60u)hA*ta5q=z$htI?|^s%qBknHqP
zqedU_RbzK&?t6eDFU2wJ+J%43ClD4YZSasnL}8R~P1Y}>x)#*kFQRYDAZ!$PK4zH0
zq6Qjrc;LZDR7c}!Q63?Vn8}I1FavF@95}y^VFk;UH@txewwH@64H*E^c1Na%$ETBN
zOJBOp-^r{QfH<P}x|*u(<}Waywo6IMqxbrJdt?1->1knKeZo8<`h;T+qj1eBhR{9}
z<OtrFtM5hIH_nDS()E>~HikeO`h=;5nVv6%R`U;;xIp8b<*k~C9HBfO|4rb{ePKQs
zG~cMwF3h(G@^o@Hc~${rP(zt3kb%QaIqeGpX@AV(vgnXho~%Eu;jUOjV5ZK+4&9S(
zBx-@8iW`}NDUMS-1`{Bi-?1){lC#YS|0J21_9&geu87xUAD*}$o|pDoC21X2Suom8
zGN~^*&#a2;P$*A=BN7*rv}&A+SE7QAACFVXttI0C15Xn|BCpZ#AeYZq96^IyLQo4~
z?ZYsFU+qPjWiUiuZ~A!Gm`)6eVa>sRAh;wP#&ke~rI{v`RR~Z8rPo{SfxwQW$#$-$
z0;v2ONs5$n!Fl-J8WBnsexTjjje2id1RLNl0#oWHSeqhA(XO9s82^(G^bTpT8F?7A
z>^UhMDVUOUd2HcYu7ucC+ulYX2MhXo;xS6?Y*~(!P}qJf<MIi@=<v}tdBFJAYHoen
z-e}eVTxVI^0|CYeGwzo|#gnWVODfB(ia&I&Qa;YZA<MZk*{mvhh8XM>mSpqSN6HK5
ztLER;C^hO7YSTofQhz~ZA*7aolaY)2vFphM{K`50N}cZWoHs#3r-UlyDjW2!&aByU
z<1NO2Z;G1GD1`&ZE!wF-YLZvOM4>B?gFQTdk4^f<RccCNzGW_^=~}fM0mwontyj9_
zrUQmhp<$W=NU3SSn~eqXKuNZqR@?|fJfZ^8A7lidF@wdl^~V01&1x~8__5xEaj`r-
z`E1oAQfyS#;urQzRy+QwY1}x7;@%;meW}s3_Xss-<S5!#tw3~&rgoGn4ya|J%Fxnm
zfhW_TUz&lu`@jZ6+)&#e?>!V8YKQzp;Os@yaU|Em6LX{G3H=+r_=+WFHti4mBRdFD
z{a$?Q`Td-FD@9RSA&5?f^uv}I&SPelq>d%S!B!s3{A|mX32}EdY5W#CP1XdWI=eI7
zfnJe>4pM@8@sBZAO@SVDx0P3my9h(6HcL^N#Bvk`b-sgBbkcM?ub@DKy}V3(C7Q}&
zt~O0YnddJobxXfa2`aW7{b7W4+o}67*xF>D0rVATJ%vH_)0{cnd{Uu9*kTr+7JjgY
z@?m*|H7fm0<p{_nk>WjV%FTTLN|vf?UH#76S5(I;p+YHE!&Y+)_*0JOh*%m9G`!eI
z?vhSQV$-oG%PLh7W2u6)aX!i%k$0YgPi{uB`ypOTzjInC3_r%j2^Mlj$f@z5;~Xa3
zGB-xgfs=5SjkmKo2rlWEW*=e=C!f(G^}d=TtL*oZBnMr;A`d9fU+UtUJxbP6M>B_B
zl_#pk!p)+vcT8A_C0ZEvSGb#5+%ikTH@yA^9-#wceX-v2ZI3ui4@PdI)z|q{b|LRW
z2IMy^sh16X6PaoB0%(3CJ9Z%2`G|8E4AQ1~+T#g9V(~%VGhDFf7|Z&i*JIjdY2zh`
z)o3hOsOmdO-SRNubhUjd)@d-ga>z-HT1u|p1<6-&BgX$fvV-+M6w=u_{#)&|%>QVo
zyKigL3IhisL?S)_6oGX{;pQ}fcQ^1h3gA4%{eQ*NyQP@ZSCWl|+v=X%ETPYCtOeo<
zOT_P1+uj#WulkMNTV39Io9?JXZWnmDT!H9V{Lv$Mi8H^c(U~{*j<1W{)#Lk(!=o0{
zQL&@9fO75aCddqfdLjK~jO?pBSWygO7>?q1OK%81ukKfbliXwA=uR5@aQIQ_o0+<J
zzH4VzRjNn?i{`@z0fUR8GTZ1~yk94W;8%2Mft|#v2L`8IDr<vn-2TNtmvvcaYL@rg
zAp;Q`{IXS0Ebrvkvjgk_-|N?PfA{yf^>qy%Iu(R*PT(^VR>An6Zy9z!GLM8jGexy>
zn6NAw=J_Jm1g(@4Psm=$k*|b(*l9Rv^IvtyQ8)KRhTwe8QoS%_xu)r>v1Km3n^2^1
z0v|?HdOEztO>Mj$u%wYP;r8gT7ds8?x+3V_`FQ!1cW?07eQixlEew21dVP0_&{x77
z_*_J&ur|Hl3q3c!)M~jx)Sj<y|KJO5aK<wQWFbLj6DwG`as5kgoPB<>69><ZuK(1}
zkX<D|A>ZN5_BPQ!ek$xG8Xg<fCP;Lof?CC9RGaVO$M`rRj#*qqifrjvW~+~x7)f^=
z4J*6qhdB;5dC7A6v$9Dhi&$_XvWe3tF*Vg<6o^wt?Q5&mFc|L%b4G$?YzM8wES;Gq
zutL?Kg}s8Ugq`=u9GKC@xEe(Pt&J&xlv5(})Bx`;+}a7;+L~CkB@bV%Paa(Fle`lU
z_ZM^i%LZK*8hx;?Y|lEBq=+WgbHjqtq}SVSNfZa4Qxhi{)i{v8s5WN@M)v6^FAPJv
zTEr*hUEM-3zDSLwc!}RybcVvpyJW1ZovfhpcH@KRPDzXMtGt>%j!AaKhK7T{SF4#Z
zDA$y)uD(hfPKvM~+=6mi=21LS$Xr?S`@P*9Kb=J_WNG=;niXaYs(7f@n)C`e0|n`t
zm=rpE1F)CXTTJ|CrRoI}X!e{mmsTJWTnN0TMqdFPur))2xKi&Rqf#7EVtzOoaWW&h
zv5U+=$>s>ymG9S<G<FEA(4t-e4t!MTegwecp4oN-rmW{&0!#>Uoz<0}RLyly67Swf
zvWctefrJWw!Ks8CnA#<j9$;cXl?rU8Q?2B{eb#=uWmb{wl({!mA-2c^D&m}a>M2;;
z`xx25amerk7y0WR1)3%}GnMHoAan$#lRm1W-Mj};s9AFyMlIozF#&;MIs6SojB<N5
zwsBsS`vRhWC^@10Z_lKRZXvZ4tl*{EaSwkF3vOcxpT5J|z>cH1deNa?DOoFhS&JdE
zr|{E_n_Yw|*%CCabN8S7>1O5IuX5KWm#VW5D!SEhL&EOLjgpRA7gxlq7IT*PTfl-A
zkNMeaPe0+8upSR~cUEi|+dpJ`H8pOfV22nTYM~geotLpQ83SRf4XerTAdEg>j7q&O
zFeX*7lA*N-rc{F-|87=4Tw?t>;aJ#MZBk}MG8D0@KPcsuAhjVZsc0?xO#~90DGDH)
zae4tFlAv*)EyUs#D@5WO5^+2yadBCrOCpirUvnF4m^!EDYHF`XUq0u(Xpe0gf?mbE
zi(8Lz*nm#Qrns%6j+gZEXljed9<Gl$|H2hOknX={SyF8^z3if$PTabpy*O^UYgvXt
zrI$R+7q_X3)k+N24mC3!y4wt>zuekwpc|=y@&+U?e3A89$h#<SdMmh~{y{V_o(1v^
zk6<BPUPysvtXz2oU661gRI!js)~=ndXpn7xau#aP{>+nj)i;@A?=LCgXV3fOOO)A8
z<%^f06_GVjf-YZNThf?ET$p!1wlc<p)f!$@U{hJeUbHp_<cLzNM8I^%R*sbwjTO;J
zC|FX^>l>ci)cRgNN|u%m{!<mxYH~hOM9RK$vl<Fbf`i2dr+*BIl*wmY1CczGmXO~Z
zK{L}GGwxCr=iIhtf=BgdFj)RYk%wGM@yA|YrHy1_xxI}{i$Wl>DjY+|oY)4Bk(-R^
zkCg7yxk*u-J(XGG1Au~hK~Fqud;-SbKMkr12X}^A$dr`G8Dpj*P^B1)_fg{d0O{(;
z<lZV8p^n_L81rS@Kut~U2uL^%cIbv0;fQgb-ebNwKh%La<Ml{4VXi;Y7llcD@y69b
zbRyo$8WtCZK6I64aLCmQN4av2%Fx;3DXNOKvMDR7{IN9Eu8$2y<r3zTA(M&4Vv2w9
zdhv?NV6Y?<xnfCJ%b^S1$P|+n$P<c^=Yr!379JMF<A3ZahtLHg9P*JN=Or&PNkuGg
zv&6zw712pV3Y9#4H8tfzJz@uDf8f{WtR;lV>{D>43?;E5KR0oS_zwByq5oz}MyQG$
ze;=Oco;?a1{zr<k|33|j{~MUa$jQp|-}NNaWn#B}+tIhRclPMRR7B-+B-(mx<0xm<
z{W}R-8Z7NBp<obmxx3_=In76yn;F+u3xT7|Sc^q?xfBTt+o!Uwy;|G+THR}r=Uq@K
zJg)G$Tm$%6xkZd;R;S@f$QO7$PPe+|({QW2bcM|jbg{R9uAZ*cD0ly34L~u!UCkbZ
z*dMls+0tgI6?MxOd)?R-=w&BZ%HA~+%-GTU`F3}iJTqc}s{N5+^IY2s|JPR+SK3rz
zWq3yYe@9ePE1yCME3;y88zYQ5YJFXDYkVqqxh^Vhc;)>8l3ZpsfGAuhHs(0VCA;wb
zB@<iHs{27;$&9`Mvb?iL83i)|>VM^zA*L{X%|V}ZlOdf!`U7;lIXcE(Kiu61*L%R(
zc3a1uOB?5a+{Svx`-hjOHCXPilYne4vZB<_A`rGk?7`%4NPJXsJt-XKm6k0MPRe96
zeW0x|>%=4znmOx<dB=dNnff=^%<HV?TdkWW*#>nch1RlxWGs!e>w2+k$-elbOq(82
z?dc_t&pdQcFKVFs>gF4no>X&nj1kdGA<rXc;r9A(+q;y3$0K=*nNGWx17UIx_I4ih
zowFcoyj)vXa1AdpiPzsx4JacYtxsq{UucQXT|PyN;lY9izY|rCSV2c=r%tuj_fL`C
zhjatpSJL}bhRKjUaXmjy%zh9TILP%yQOZR4djHd(9n&Bc`32i#1P-ONcS)F-=t~=l
zmnGHXn@Kp@5u#;vB>_+WG`FEn_JQf_r1X4odOblJU@03!zm`?(^3+d9d$HR$Qtphd
zujWu9qDVnoqFas{GiBHrCdYTi-m$vJ8aCCxZrz*$*9kS7d}461I;#;Tf;^8^p+7N2
z1-3@SqW~7e8j|#y<XDQ*&84YTgByy<=kD!QB*zpelJT~YjqxaJ!!0vnAlQ~lQHV4M
zh8l)Zdj1fOkxIiN!7L6p%dwLJYk4r6hk{`JZ)p$*ioYb*?Zp<|1B;UqioXeO%@cH>
zq6WL_9+Qi#Q~^>vYZ^?hWF?29e-Xz+XS2#osB-8%(O0@)*GOgOU9>e$V&~<Dbe(+n
zb2bM}$EU|HgSkX5V!)W!<>_)C16Fl4pf-QrNaVhUjC+vPk{tYwWG0vtkF?AuydO))
z-bCOqPHGDvIKN?@UO(#uUBscbSG43@`dc1S(CGa)pixMkNs`zq7VVAuKxE{n*(KrU
zQ$&fs@I0SEt*|28-1qH1O=DyEGQA|pa07!4aoO!H`CPgnWuXkB6`>ia13`hBmQ^8`
zOccxb%td7dcT(tK4nsgM4nR{r0W`ETQk>62D)xt&@&Im&3D6i5eclAmB$kVgT@vOH
zHly^Bd(DS`sySBEC+$&fFp5Pcc~tGKxZdX;Eei)}eLF~^skvnuVODjj?~8b4Z9d~$
zom$vSXiGWm>cysNMlr6A(lobbrQX5j`Vk^QfaBeLg9k}rLah}ecm|l|y};~rXg^2!
zsh5106lOPf?xdWoA%9v)LMv}tX|rkTr2M+wD$}&ky8ZX`Wy=~*Vm<L0J@5P=Caf7<
zsQT;Q{KR@4yVOLp6Gvk<b21-F4EYCXdG?@Ugcm?$$|fhIOL*~;<SqrO0O3l+%7<7^
zl54U%wKwr?a=lN9S+OljoW{sngfI(5fi>l`b$fn*1)co$z5zHdi>fROJ%_bB?HaQS
zin&ORD0M|z+X|7yk#jzxJ6n|^Mx|ft<chi=?LiFlH^?2uhIZCI-^rN|rsLAVd<fZ`
zLgXJ@4SjetRu1<5n4%%KRVYuG2nrS?zUGpwG{YcY7+G>jKgK3PZ3AByvrKW@D`j+|
zOz!32r50J|>M$Fd_0P-PzeG4+aYS3nZ?DH8tOhDcSmPxK0R%=FeMZ?DI?-xO`{!nw
zbw&?VCDGm?$3*Fv_G%XS%@{yJ3HG#Ojr|!V48lH!2=xjIbux9_YrkcC*VK&+nL5wH
zF1G4A;-T<#^;RTd#V)9rnEf8k5VIulldZ*?@~0IHCe(F2;Sxn`L-As>P4mBm!@_Io
z>1uk#m*MBhD3HSzME{hJ9vzB=e2f%DmEI~T6%`s1HNDt^5Nw_>nbpj-BWKfisp`a1
z>oG**EmLW;ai7S2I|>#b)&Z%gjy?CP7gnq|xy9!xFWPHc$i(r^3s~MM!-!H;`9Y=G
zL&*3=#ZE@6V<Hs7#Ul$9U5=u0T$3FCz`M(jia{0^l9^ge!rnW5oYT4J1+M#xoO~+3
zBtMYjP1Ma(r8&YIOP)uEGW~+pV>>xzJ%1{~cZ6<lZM(7~$?}tdc>$RVdqv2Dx~h1t
zSdX@ON>4^UW9xWaK4P}Yi>L@opF9iJoUj{pLGes6E=9wnl*yJ1=6HR4QiDW9CtBnh
zGY0H4NvcOqn#P&`{w-Ef6>*05hq{;@srvUDo2Q~}Z0__8qR?h?@pmkE9D0@#hblaF
z)c+k68_<B#hFVdagjvxxah?|G#5_n`#7WIzD~rdIlMPMzL>dhgQzSM0C_vabi%>if
z%ArX9N*&$A=8RdAUdw{CB<w<5`Mq41B8jGti!A?&@yy=GwX@5o^$nfjkC6NyanJGp
zk0&v)FtYqtxZl=L|NUu3^2ybmk3|c}A3x|u3`%Z__v4}2tJ-OBhsaus?1OTkO~~!V
zue~7)^^UTljKsj0WeFXoQuTJWc)Ro28~ols*}wact>|Ye4KCi_-zK;N9sL(zIDgg;
z)gOg2ICyyXOnPc`*U>L!Pe-bQmW#XdysQ8fiVMUIq8i-Br9oOpR8LeN%Cf_$6Ew?K
zyKU?N>|u%nPrh|<Jn*^94zG*nhV~K}I!J*dQ4)<0gFa<}wZn(!gI9O&6V?fzC2{+7
z_lc02PEh^TmyS4q)#VxPJ4nCw_EGwJ`}6an*)Q;7xi$elEDn7eGWujj`?v$e=_Ni3
zXG&dtMUWP{WMMZX?rud9bU3iNpO?o=8g8=57o;*?b?DLZ7jimEiG&<%ae+dZr@ji!
z2md5ev?5%zYmq|C4XV^=;|}6O2o-P8@<*#B;DUc}f7{M2vAF_vy^hollH5Jem$y2g
zwa#x&y3u&t5BX>!wGRxrYD2TiCUX@cm5qubHXxu1UrhBdcyi)u!5H#SGV+9?yX`7^
z(wr6wHL(!uVp}-)i<e*NScc3-Heavzl+mrA6The2C{FMYBFDKI4a>C%2ek(+&F14q
zSNGM74mId0mQLxu^Mq(ERhGxjG7eCRZAiNKy~XIcjqOVXu-LrDwfrB%ibAkG4F~25
zeL}ZnY>KYXT8BF!Jq13Uf6usdD}~A_>ckoo2qz#y$!~w4yrJ@?GmGS-XQ`@=O&{L8
zQc>_$nx}C-8;~$hf~`5lB-3QpRoo@a#SWBp8fTQGN@Sl4Z3e4X-Z3Hn#*u$Xs=?=L
z7QGU918FV9zo`Eo#@;DNlxS_!tX;Nk+qP}ncGWJscG<RV+qP}nw!6A}CSoT3b8#}(
zRo-T1tZ(IuM^1oEk2njno3bPkHQLRg!|hy#7sEodEY^*X4^r;iq5fYtEn=egCn>Lj
zxXfLKK3f`MG7@Lz;k=|9%GukMTm3Xy$DH8XVJN_OeB&uPOZ--i0Hv)Yzkp3xLH^xl
z+8~6=o8zu5>jJ=*+kH~>u#yL6USV8?5j)pyZnSEp?vf;DS4q<26l*!LM(~3sjqumj
zH2b|oOkemPfln740apCcCbAlReHqyy7^L~Dc&61;)0Rw1w?cM=2UZMx3dTi+yLSzk
zzTCK|>ER*Fo&j4M+J-bTvaqDzQd$nd!YFsGb9G>6)t+`e*jSL-z+R(dpk7+FQCEio
zt;F6QY-$C1i7f}ZYybW;#ycr9Qtw?+hTPzrDSskkNlLk5sMotd*wqCHE+PIs`F{>v
zd>pOTyjARPTTTW0k8gpJ`&ImLa#LfLi7I*!8Ep{TaB>`1(@!|p-a4_aB+8-KYrf0j
z;S_$Z;o)V1;vz#acfO11uv>t$6fEcU@?)N|5FJ}~-RUW?ZEEb(=|Sq8)Vo}EE3yK*
zZ0)US8ro)GZRhniCxualm?9sOLZY6V#CBn)^<Pe9pOcETo=MX?TcVz#OnPz%m)Ijv
zo<>HZX~V;p7#}P;zABD1)GGtkqDPgQINCcPcv0#JvAPu#oVKWsxk=!hFm^9fNbpY&
z5?G~UeX6YLtX>Em1k6C8D7vM4DK{amA#7J=YvxxUfzY<ljEX_fK=$ndQbF5-GDWHt
zgI$=J<M_Dk`glore7Z472>uSGkrTcM_e<l@=jN`p`&xgLw$_6v3AS>3EGSTiSWYI4
z<+sj}U}ZAVUA6i`b4pX6S5!_$Nrhq}|JBN|QeX#Y$Qo~)gC=K{#Ee$;7IxP3k0<Q4
zocHz**s5E$CfyTxTy9ma|5Y*$3uBM=KV~s^_<&zD)N@)X++)a#Y<yroz7R$sv2Oqz
zJw{>_MV+NBG@ULN&`|yhK6@t1rp;Ca)_q?KfuZ6WoK9gzb*{$uX&=`Zs#z!&KtCF0
zna&3(krTx_3E9;<oc|oCQ&>TiAfSRQj=ZmH`K~v|iRN@emh#5d+v$T7f}1swe%=zH
zg3U3&eyelQIW9cP`R33(lGti2sz_CST#=6$Uwhn4cMez}p9Zd$;MF`0!Yg6FQvK^5
zB0QwD!!inU?j!a9Dw1W!Z*Iy3M)K*?_co3dJU<|td<7-SUB>cECIrwl8q^4<7gxJi
zhPyzk550c|<}0L&Holr{dZ*4Zi#}{Sedw*^lfwaToR0t@yLO1f7GYvJ{n!{%Ku_4V
zc7NyQiHiHBUB~`OGq!~IZY#&3faxD?Lye<Mym+pJ519_)z^gn>b58P*Z2>+~s_F~+
zYcW=qo{D6fy<o)4*0y`1RK`2QRD`>3t~t^v1m5n#a77J6?6{tW#HY4Qw#Yus)~PIp
z)rL8#;XV04$<~r}sYfbbqzFD=h;!hxo;e-N5tkT6+Sa{>PM8liR|Pr}Cr~a$ITn$5
zFm<IpOx~3-70z0VVk)Y<WR)d~SoOv{Kppk-UdWk4mnDRJOQ#ISx+yAs$RJTZFNxEt
z5hhWPX8$OhQd&00QndrX8(q2#UnF?AATtuXNZfBVP?Qy&RJP)!i40Bfo7k|1rML4*
z4;w&zB%bFZ&i)SlOENFq3=LB=vWInRmJ`b7lvT>43+c<!!g!lM-D?H+%`Sbuq;#jI
z5CZB0_idSf(yupl1nlea{L*VF+6tDTcChG$la0sJOm!+4Hr&XF#}G_plXkNI;|^V-
z)e<HqI`hVC)T#+K!xBt{SexXN$AF!OYR|l;LN}PrX9K+g-f9HBBF{a5UY%s&4ZSk{
z&}r1_U7y6DO|>Bvvlm!G`5P7T>6q6YxT0eShGHVGIzk$;cRcShY7IBtL9d=Mb%q9$
z6V<x#q_JExfM40<aBP>1T+6K9NlxL=5fQ;E;2G+yh`K*-`O&rZV|v_gBKT<N3R;Vl
zhYr<dEWU)B)t|im;40P7SeR^-&Jkt^RZ5sN)ASZaY*LdS%cQg@l=MyD4}5tYnW*X6
z8*qKSpCI@~{QJsd>ka&vHZpyi(cw#gHcZ_@%vdfmN<uy23A!N1bDS-qPjUHBTB><G
zA2Zpv5uW}sAIg|x@h{9Jwx0er-rOOp2IkEjllQBR!H)UgKlX_^@ZW-bWXJ&R+%H@)
z!mB+8hq*XZeE7y26+A1bdV`7sm@zXfZG3v018Zbcxamyh4*TGNMAZ%U23Ymm98|NT
zN!;mWF0DU{vyYyEa=gj{Ig^pODdX2HEiu^T;J7|m1v#sW%ZpqglrVElEKM*ed?jUD
zg2hO8E15H#4t_o2igdu!>1$~~DQj?k-(Y^e>*oJyc;le|pNa<~>#q#*e=a&v3u`A6
z2Rcz}11A$<6C*og6JB0OM<)jp0~<)U^)1YaXsouo>YCGnD1K%Q>h6Dju^3n5{IYk3
zwi2*VMpya<U~5yf@iTIFYs48_G#vN6p?@?MkRvKzDk9|Y$auSWxOwolwTWeeHAtYP
zXYfb``0U^H>@|N0=YsS!Lyw>C?-M{Lxn~c;c!%1O@96Iv^DUe>MYWAWv#<gcf|=2v
zAXVa;agAWe-j448UpDjWBz;!?E$a@x{b{_>y!VeSCww?=f~>x1&_#yJyM-`}{X_<@
zgDOBz$A&|(yvop3XSp1$0A1rKAd{iE;1C8R`%Tb^P4s_*AL|sy&eQ~r+ujF2EiqFU
z>wAi7ryInCn+fY+&$RqT{(ZCljZwU_`|0?>I_GINp;Af~7I2d&sw%d-Vi&Fq>G_lo
z10Ysfm%+qg+GQ-J+)1_->uPOMV}gk$P-CeQr!3}%kqaZm$l1d_lSlDB-S?5^zWcm8
zNA=<I)XnT9-8TZ9vyd7}{?{mhpBjX@!q@@y@xTi%Cdq&|hkY`yg6kRC2|kjJ+Yq^2
zye~nP)=+)0a4?h+MUM-YFP~@j%~4NTPe4($kbU!gVe|6%^{|%`#GM2r3d-2^<8{lI
zD=&@3_J^L#{{5-sY+z5v*V3%0(}h7`rfpf0#RE2-!sLu~0zN!KgYjLe4I8-8n7PUu
zk0<3gjBm8$tta<=O0wvl-$&@TI&0qlGxgQBbl><LIxG!mP#vNZcRM<v37VhI9?{54
zytQ@r-MQGF$ud5Z86<*AAz_|@-qtT;L;&Viat7;}S#@4zCgUlGo1Xl^5VOtVonKl(
z@I@RX@7DMt#fd_0XUvWwtH8mf&`He!ecQyN5a^Nw;g%f=3lFN2Q{Xn+B{;vyUb45h
zSgC%1x23$%5RJo!$|6;$1+#~OlN}My9DC>n^L=TCuO~Wlkv>qtKM_GQ^~X5&jzv<b
zSB+`~oy+1lvavj7XIFq(C<p<#*>28OVp31yre&-yGcLYKzNc$@Qm@i9Q#$X+9qEF+
zx=8BkRB<={USL1(#FueS&3AhciQG&5de|DT0iK`!NJ8j+S?vzKmyxjMgy3YORl;A8
z?RXE;HZVcJ8_Cw3!ec_^4U;KQu@I$tAj)nWtx?R4LT~bH&*DKa0=l+oRLwBMH<_EY
z6y-b0?k{Cw6U-={F5aJ9c{bAm<V5d}*`+!Kn{BeOvX3IHRrURU_$DKWpr&{pEy{Ql
zWu!tEw3sIEoI`P!<|^S4wBGd_&mu`uZ7rc1HP>GbCk88va!gk`JCPgNjOO*=H#(G|
zI6AYzh{3_%x+tp1S;|ZKXXt3;nMS(fk3yS_B-cZRG*@HRQ6Ix|jFqivR|GC4iNvDY
z4(m$1RlNe<nxPr)HRa*NaSW5}I!Z}{l%%8a^{=|Faa@KOa(ljM;>&g=`1bRfXXbA^
z33{lyB2CM~D{kb0513wV4>K`%->=_d;YkZ4zV6`V0OA!v4GSl)r=viUAPXgr;?9&7
zoU;BLimI8T{aGY_CbgfRKAf{O&?LQ`X9;?Vlj?>;&}UMx5%?zq6-7?o?6h>p9lT2s
z7<q60{4#xgd~ZsKmA=HU&_{S*B|g(y5a%}wdjn2xYV2pkUk@tvdme-0?aJsBmZrxf
zL=qV&<Don@v&4OShC-&AK>4Z*yu?4JnR&qN+A8EzR47gEVbSAchee-TH8t`(2$*ki
z_hk3r2}O>jyaYPPVMTFO4?8z}WVJECkInDVBiJA$eg3kX`8B%I>{N-##Pe}_G8+%q
zm+E+Oiae9a*q64`u3aQY84a$LG!<Nl@@{M@<e-W%OjS44CU6iE6wV-?@5t?jP$bg7
zJw|Kwh%1aBoT68}wInZL>)egDWF7em?>*8VUl8vGF1kTm#k_o(Ns<u&BTU$aZ&nH#
zS;FjZ7Oh5Iwo|UtVe&9~Xzc3F$xaapxvk|Nz&iu-DA00ZclI&`XHJd7Yyw47uEb53
z#HG+G<0RKQ$K9%-JRS%``{CJE-HN5}p@lklUH#u_spZq$C8h4skx6khoFa^Ays;v>
z)9ET?J1J|{<)X!Bl(YUXrArc%9FQ-w=ZBjw)M{*7Gr_%?(1TX(9$n@=W5#HUU@pCM
z-9b}U!#P-gjkP#pNF+g$kcUX$l&D9rpb*Lo?89t$eV*A2B}x5LtFFh}HVI$^%N{%F
zx1dJLGh;I!lnS_2?I`3rnW$trs4{gU7_%LMzETSvwlSGgNuP9W!~hCe=)evQi9<rq
zTD6oPs%I5ySy;0ESRU`lrz<j+WxP*JPZez2u#SC^g#W2b3v#IO0BSWwwGPqMv#V?&
zJ-9h7FOK=)pITCPRM!7CQ(?JvJfZaRRSR5SCZ@8nhY93!dG+MUgQjz4Flrp*+{e(E
zK)K8>Dl*GkglUxQKZ~5jozNr?54DUcedPIGKiid8Q@orWuP@HY7d9a;0%Q#uW(~+v
zchMC)o63}gp3mS=3p<<QD|S&U=Z$uJt4k^_+KOlNAhX`G!|CMXlx$3r5q&ER9(LhV
z!~A&`_SX8^ge`jl_fSDo$uH;^s&P`dS*QrbS$QrtY3Og7<6hW%aT_5}qdHRxlX<5!
z0}$tcPx8YNboIfy*uam5Z0D)Jd0=%}QT&B9@Mwi|K-8*;C~5!pa+sqYE2ef>F>cBq
zy;h}3M=ZKrcUv*)lOXBq_8PBeRA1kWeR(N!FJ5&k)Y!c>JI+*A8V#S}o_S^@IYKYt
zn&nbhmvh6jNo)|+ng3(m%j$zN=XYFKxw+k@buJKWX%TwdR%{9PyHU4k0K!eEe-W*!
zR`cVEYywbhwwU~(Uj_t7q}^-bUe5P8@fkllj)-bjHyv$j>3r8Lpy{{DO`v05CnT?{
zItE|t0<?JiIH!_sVLvXdwY=NLw9WIE=34o)>DWvnTj2y=CDUwfDn7<(+`Vipm8sGO
z$rM)Tt@xbU^{5>oj~vzNqu{+MdpX<lH<`!qhpmA|5OO%{t$!^U^L>3S8T7b)E%}Zm
z<M%e9pCHf2><z#K^Ct7x{B9i=PF#r|CiyBxsz7<GR$1r!)ytIvz>QMnDY5&)U?jh&
zr3Bfke&CAqr2H#?h>=0Hbdh*&<ai`;180&TxCW6&cS=}VlG?Ml{5@%RK>Xx$xaw=S
ze<mqZ{k>=rKtbB$UTN@p1~GhOeV=UiXFU^z&`av{x=S$8=+nn#t4aTKA+*U@`@u$1
z0;72T64dpT%sBZz?q)v^1w|CALvQzGcnua$eE5B+ZbdT&Wl~;v5mWq53j}eIqM{ds
zP?>Cz!>(#D=U8Mt8^fa6fcNgG#Sme(<4xwK8jd_<%{Llb*KOc@_!;HAgF`!xL(xDJ
ztE2Fraj<AeOV`@5l!|hYwK^T5v}KyW6=tiKPXv-q$<gdmt1EGNcTXgndA8>ua{-r0
z@YqD8?tknOC{02v#MK@q0f3+nT6ZXNJ|-F2?Qb;5t0sYecb-30J(lp~Cc8jZdWAeK
ztqoCAkEmBhwQ<TS28%u2H&WH~{1k&e^jvgLkN@Mm%J@IEMJ9R{`u}QPozfb!C2qm!
zzMwcKCooU{0Hlf%w~_is7d(7F_H4kY^=}eU4x2qDC&ntSvWX`>42)a_ErRdIo5<?)
z)--qEYv}4=NX=CcHduWK8V^?w4w{bqn=xFV3UmN{&L1Y`COD^>;s|qvhE}2sz7^U>
zYjf!iKXnK?7z4T>5(iF{F%moyd~y%cH9X>=MZ(AxLKkS;QtAtmOdB?Z`}_WN3q)qn
z=fh1Q^uV9|3e<$AuW@gvBSH^Mx)ZPe5=t_#!SdDnb^rj7OcQG}=(phpkKw4ox6A#j
z1#Jf7Pj3$oFb?rvI0si}L`*aS5i;I(v6w_DiX?0#dn^@LNoWJSCg{WrT`-hBfucRH
ztj7fy2t#uD-(ur%bE-CUw3=YV_gZb-$=2=VfV5l$f!*TS-JDNKh-E%NUZFQJ-=!IM
z%G_@T?M_@;+;HNoUcSwyJP+Ae{&DEbB1BVMMt}9P!q7z{|6oWnFf7&ubg&l82jbr*
zLZeN2c@)_P@gW=Bens~n?!=p4$i)jzfdcLvHj096t$K)+T?BO)bGY;8*S-#*Cs6Lk
zIc^FE{T}#$Bp7(gss3`GP~IE|oJ_;Q;bHFnF*))zB5e7tefZuB#g?7FkV&Uw^inkh
zAZ~DYckhXT@uNYFbuK~CA?%uQdNHHIbl-^$!nok-IJCG>ptA@iINLp|M1Ke-2}8Mb
zqCzangWWBbpV(hMALHSfyt6Ke*>4;ws?$eOsg1>pkfz1B4}A1pb9bT%x~!GKptpJ4
zwC4xAFX*|MS(+t{hVe)$GMx~Fq&*~*dDYET=?Ze>w>N{Xz?eCANSQqO>tgz1U=nUt
zymz1^*R*JUsm5v91i`+J2=9tY2~TeOvU=FpJNVa*31-mjg-CimQQNk~m8QhMVZtL6
z4v6HeQ|GMHkZ3dMv{q5533n!E#T|=DU&4ISwNcKgs%-Und6feVK3y)oO&zu{&G>j@
zWqsB}!Ai`>!jX}cpiN8F2KDx)>a?<gH@B(_nj$5URP+J~;h*6p?kYr8Qc9W09!Hx_
zf{^=C;luBbhZn*6fs~ja<)IhLm-)=`@04PtAT*VdR;5Zl<uWwY<?+OxOj7I#VNq5=
zOH@TADlS~=P8d@%IQ`Zq(*?Bbaz?sn@%0!w#F8rW*A#0LQ^GDmF@jZPZ_x4upQcS`
z0lak4R0+;OyS*iTEUnqn)oy4{y;T@Zh?W9{rdAw`!F0;<3=nz&laag~V&S#cv?BK9
zVMom+HOj@^@5XhSS>xpE|9tTdZN)rDYK}aNS7=gwaaK`PsB_Q0WfZJ*ZfTF$^El9B
zd4RT+<pGNcc}xqfNp%yOEZdq^vM|mRX#8lohSLzE^53?VgGa;TEelEywB0t+_A=WJ
zKeMMy#)LI=XGT6(9RfE|e6C_^jTBU+szk)5y&#yEnd*Q&_!=7&eH}_M6PqMyTbN4Q
zOKG~2M=2%lsN6!){2G(=L_t7YlBv5ZYuF7OsKb@dr`(v+F4OYE<-8Cx47Cwus==3e
zZ|K8n-KCBZn^-AtqHB*bGd6EBL_G`6k?TZZ!+}&>BDB^ymz0FzNfqMN7YNgV;rG@r
zbJ}hly>7v8Oh}m0;)l2DjLEyG@a>U1u&WHE<Rq`<P+5~iLrPmxRi%cc8=Kk~MGGbK
zDRea#q+FOG4F}@Lul-||e&YUWv51f)e`JzQ4Q7NNws1>o;=z3f;w~XHZv$C-BrZnU
zk97Z&YDz6WF6s~@pis6>sTRnkFat}N_f`tI*ibjKXrH;MBpN=;J}q(KHYO}<CLYL;
zEvG~~qO{VYMuqz46Znn@Do?at<=caEW@OnaYgEBR$sL?RHru8k%(OK3wn!QB;(({C
z8q{YOpZ5Y=^<0s|&)I(i0g24DesjSwPSn?6m%Sy&YCvKZD7j30W3j}Dr8|$yWKJBM
zmOj%zV>1p{R(inXoh2QnT>QycwOw%E0hoQW-IN+e_je3_qr&Y646s=~QhaS<_gMBs
zafPg=b;U{26UC)c4<bKi1;zGd3(0a_kE|UL6Q`N68~2H>AD0o+b~;Qb>HTi$_I*E3
zuPFk#k(UenSgJ|2d@36As1Ag6aOCfWDdk`nV&*zsoX-aR55)6&q_WGnfl2>eHl|BU
z>N9)E=^tgdx=n;^xdf*{2%fyKsf7Y!16#rAZtuPZ2}e%-H8IS~ICicgnx%TA_W6hK
zf=R6%RRU7sPT%^nDo&~ZYwTc7YlrH>3fBOmnRX*jnx_U8P^;YQ`)gzWQYaN8(AC7_
zZ}MCkg-W$@kvn*(eJPhCg7jDUVu?V9>wy8ep4mN;ZEv=~fyTWT`iCw2JX^RYm)Ade
zgus5zld6Q(L(SjudLo)EIYFT(hOHmhoCqH~(1f6Lo(37>MC3&4qVmFd;P<WJBKX+s
zO9t+v=Ev>`wTwg>P!B22%=8FCzwv&rEDQq%H&nV(xYk7|M5PA&@r9+jA0})UoH$Xf
zJ`t#=oF2U@PALbO>b{Ee@`L)I2W1kZgh={w59^<rZ>Kyr;J=VOPIU2ENuj<cEiOjf
z5O*wWbw-=?Q@ZF_Y)TKhYppekvOM<^@_|+_FUyntIh~jr_l|G1mMDl#ph7nEW++Cb
zlNr42Fm4vE+zp6{eS<VL(Ht^L_%hZA?kS8NPSl~P&OGjl1uBU=);!k@13E~NP8Xh_
zw^j^!{M9WFiA1(?r?wQE@v8eqBAP^amd}ZKl?qwX!kCirS_j&>xQh;yUs2=e2+C3m
zLPt}5xiK4r0qr{s6{1woWQao|UA7GCL9cco0^XbapUyn`obWZA#C+AkzG2$M#I@wI
zPLxHg`#p~^$T?#5yFR6_Fj=MfjECo+>|v}~H7@?oqI;*)mL#~xtrz1_t4>3RcBHmc
zUMU3U4Swg;xxDio7*?Dn<*ac@r4&kFuUCI*WE`XhFd|$vlj!xn9lN{)7HknVv10LR
zU?bb3GA~`-0-GL2BeUqdhy|u#;6q?#SB{pPAR}8UdXA-le#+El_*F2emqfu&Gh5_8
z&B{+X8T}x4R&g<kuvC&rf-_Jp3(DjD*~v5|Y!X$%gRfFRCq#ovsSS#WtP2De0T6JK
z6_sdm8U7g{l~FWTAP&$YD9hFFbR!)tZK6VAarE7~uhY}wV=mU>G<)r;Nn&l_nnYiv
zUdg=}ncE+n`H-i(Q3jtQd9iY^XT723XF^S+eGOk-xscA~vv<^Pv?C*?3MY%O>W9ZA
zsoG@*_ph%v>B2=Dz3bZHG$qnAcbBX^C~GbPv(#Ws;VK2c(t+@lni|HCG11Hhx~Q|J
z^O(Yumzo<a>!fM;s#j{pKGhD+T59#p7z8Wr?5~2vO_`yd=bx<LXKZEf{Ny(qnH6jt
z_yZ}KwL4C2@B93@jN{x&C@0q{LGfF8L4a9Zxe)YI0B1rN7d_ttoj-ZT#&Imzle?$J
z`G}#K>I_FaJSJBd&Niy{S&)@jzId@(q||&NnNVDA5e^HhUlpBJ6OYHy?XDZ}3!P_p
zxjfgW<@<d8Y*C~0ZRI5Qd^;Zsd7<m+5#(`yL-3tCno!#%_~JES?fJa@Uh>d^&alrK
zd+D;52)W>?TM4<yH0$d67llO28K8Ybmw3~u5@OQTBM<%es`Iw2m9QEv2Vr76Kq$Bs
z;N_>F4Z5QBQDoPW5s4njBAYd>Z;-{zjHDrat!qsBl?dHOm9e7kQpS{Ja@trm;8NQ3
zBaLFrTGjYstP)g!QRyRI1Lg_!u209-gc-nLN#j<70>-w0`2u2F)?)ofE)bO(<RmIE
zfH9N7B^lQX#%4;xgWqm4>wIl}JZIYOG+l!S_;_-9^1ziIJqO&G=qXnU=EuDxcwL+a
zkU-V7x`FJ)3tEor0hSPKp_r|n_okNgX3~K?iePw^sR^+gX}>*!q5?7&R&@F7FOh}<
zo&p!K))?d;p2C=|<VetnN1yDtN?+RJ9wGX9pXM9d0g2RaL9+y!db!U;#R1OJH^7`P
z=FsoKoUcYFpW+#x<)FJTJr9)~?jN{i)XmfXcyM5*{~xh1M#lfLCv8gO|I!mz)DGu(
zkwrjVa~14e*Oq!WjUT6&z=Udz!Q&b3S5x0U-8UH}juqF=hfrh3Wl;lu;fZyzeIal_
zc-L^Qa3O4HWuOz8sL8=0F9E%9*gsIw0tH}Hz)XtJa5*74w(Mj9C$KU5+LGL;zFeJ8
zXXHTx(0&+z1;78fRUsn9BgH2PqS!`+UN?^Gq>9;cU5D>#`!Qlh3gmt~UPt`Q3?@7i
z<t>jZF(Xx=9gHDv+{A{&xx|3XAY&3VVS5fz-cDm)W`mF-+U0*$K`z+^UH?wPE2SW?
zc7>G3WzI9nw@>}*VQrM>XuGTa7m8AfSj&tO`SfC6aZ1+CwZ-(gQA^XnYY-Xf1<09P
zScmkuyOl+O7K;yq3I!C_FP!22>_r{io%Irz2vCACr&xaX<3KEuS~7t@0U{6QtPPF)
zDfMh=7=g+n=tvmVn%(sA{_X~%q9$4?mR15|yjsg9vaDkWQxIRLj`#lL8yw<s)ObUa
z{*~&<%n6Dw{=$=|NOM#E*#Sg8JExAvoc0a}dPB~bvW_cb;Z(fvD9<acvQ(#1ZPO^u
z$djwfg@2kD#*tNyx`yvc>kFgD7fQ2_agTJ?C5kV}iH}gk_%Fu2RpZ4q#>|Va4ta+@
zX8wW@NAISW|BalZa?q6c;DB&sBIu6KuS78aDwaYbOO=>1R^0d|V-LXaD0cxPX&tmj
zDvFks8$&zTB$bc!jCe-8uU<XlPSZxd{c}D4$50W=?3OiK32=9xPK#HgG9N%t*sl{!
zCpP{HTPaN|8n2Cg);JBZKKKrNr2|?bt^15SXwZM^bgA5aT*HWl`}iNJrnnW8$XcuN
zQ1Rj6sfLk$nAGt6P$6Sr^52#q&@$+TQLd|RJ31mnLDKniQjyPG|KY@KVDI1+@0X{6
zfxU^m7No^N@$|i|6QNfD2w}>LIK)_p#oG_MfscU=pQUQkedj(jg}$h{c99?h%s&F|
zom3gq$gN_YO0P#L$W?D7l)T6wB!o;uA!j|{N2Z4(5ca~E5ThFqWV96&(Z=#K)?5|)
z=_3=y;MK=wjA173kp{S9c`0US$#^swhuIW2F5{#@#;Oa_tVsb5bZXz*_L(Nkx=`Zk
zsT2>wr#587a*>JG93nCgmt`rD;MqGx(HZ?-5_(DiFKy+Wg=CtK&6FCpBZ2=GNhIrz
zr69SAoZTV28_F3E4kQ^$<S_-wsD?H(p!S<Wct%T$0IQ0>j}J!d5>R1i=1=KUG}C`j
z;!RjNg{8athui;b8x&a)A!0)-p|}@6Ks;it9~*vpG5r9?wzM3?ZnRmKeZRUOiA~eP
zTDNf8)x$B$n4Tx6BBrpb35Cq^wK$ZN5+gqd+=1_|qe#SzI(abu(@-S+LxOBwERT=u
zQ$jUK{0{^%lLuHb-1@ixU$MYvlseu&gSnjm<{e>!Q$X|S(3Ep7LzE0A%bbO&oo2T_
zLYe$fIRqJ(A~xWomzZSoN_n(b543gh3;qkzK-Q!5tw#tQHSe2kcvU`}g1`%WE`;sp
z`}VuTl`ml8c8|{w5CXdt6?{5pp`>n=S((X8atC#7o4o=o^AY~yAD`hrJW}NTzq?!Z
z07S*a)QnL+kOiTqwEV_|uqFZj$4xX#E~H%-iRzPc-bc7JNO*05Oym`$!f5jf?41Xh
zvsczLp)CK!iM<UM(o(?NHP|CRuW6gS2vOWGsZh}fZ9z@L5sOna0*oZNUDRiq;SOE4
zg#m&Bo`|rWzlo_4#YIE3WIF+$Oi7iPcYi&i=CWhMd`1+aSUj*7$w2@}D57eic=7ns
zF_U%)6nb6Zs(DOaC?<0hQt%kgQ8S=f$E(eOa(rguA~D}WR#rika`zcOuWMi0-&UBv
z-q$Wd(7%y~$YA@#e)V*v`xK}pb?s7GGE<zS*?$M#*Ia5sf|Q7}+8d`D)4FYg^S7%8
zF5#|%L%MiJr5Mu=DBXGHfSYRxZH~IFxKvg0A?HdJ$eD;Z=Pb7eUc0Jm;bvUzNzScw
zdDW>QZmfT%ozozhkhZ%A&kp)o`<pz9Iq1p-OWvd1k|KPuIMsS7m*<@;*L)_j{;AB<
zC(DsDKGl|8(VczkJWy}Ms^-j=Y@sx3<N$!B@HL}On}S<L1t6!zmaIG4C{NtA=OENp
zmAVCREhwMc$Nx(hNk8W(d;I<cuz5Z(K*6ql$T7@zq-9!Et(P5;WF!Z0xQR(L40wTL
z-JRQpyGOGBDed2Gmebcap1W>`4;PgNxg2NpCu8=QEmf^{dPCZotj^PzC3l9%R+)kw
z>42Wf9C^n!vr$cgvcqM2lO=O~B6Z1NV=Q&aWJ>&zoe&pQO5;d*@z(fz5g;n9=;$~I
z>NR8>(^lnJeMKK&D!s!#O7u5Qi8*9SLulo7;#D;)q=xxi=#tQji&|mPu~SD^0DP3N
z+~XupxczCe=(9GeXI6zysF{^OF0(1SbEj88Orf!E7vQI+sNCfa!pr2mT0u?T@u6#8
z4T%TEW0SV&>4tTM^r>}$qXp5hWc^J;BfbpCr@5hu;!vvgI`7$mClk+ipwQ#-K13I~
z0W-j}tL@OFWT0_q$w#tqqI|hh-!HEc{vSO05tXx%yd^`~uW9}XCav`S^<;(B7+KVd
zY{!&un9;yQ?6&eMby5F}So3UnH&?7v?M8~l$idSj#W-))U&Lcfw*%z>hv-Lo=*vSA
z`ekwPetjqWZ-Z*&S4X2^<bdrN6&C}_-L|N-OnnVp!>Dlsj8Pbc(FLtE+n;Z;Dqrc0
zV$^&~Z)&~O6j~V>VYPS8aQ78;BWtZn@MDgK+XPj5+U7O4sY)ByhZEBm@<*Re78+^X
zC2qpL#@s#x$L*Ud9EnNXO2g^0GRMNewq*`LV{n8q&SKZ5sKU@a*wWjz8ZlU^E_7^j
zl>EOkbvBxU43t@R<lnE39V909^tR+K%!RR}?43mm4@o|WPj}V%LN7@!i4UVMG@*)I
z<#d?xiJy2m^_%+tX+~ma`u~u{l-=!3=;RD7m7HwoWbqm3>3_`t4vtRv3=9k`bpKDP
zgOQ!}ziXPh(z3C`ZbA6I))UCpn}<7&AHe8F4)*I@v;Y8u*+(|F0uR=mZdN>!<;VhL
z>*=0}ZQ(9+l`LA24E+cAymd(%gAE(<z|AD(o+A;-v~31EELN__npiGuEohu1NY9W+
zStl{3CL#SdPZA-gj7N4DY$x07Anh$J+>MlCX`4aQp@}A<PYer*=McFf)O6^Ca$h5?
zi770YducA0M4^<>Q>ae@C4yi~RtzZw;&q*W_P3HePWo0EELk70X$-O$UGhqR(KI7n
zU0S?*nwv9JQ6+?wpu>^|Ayk?Z)2DBG1(g1er=?ARfn`CTKlZvgf=yc#YQf&3G-ATu
zyfk71!ie*59r|6_d2D9ft^>{#*4{FJ74dvB;Fq|Ab7&y0qE2ii-oR<oPGkp(+7eXq
zd=e#TsP6g9Ewy4b-^8|XTxo#`Bg64B(7LIm%^qO&qTiexVY%eKXuf)t?r7nzrWpjG
z6&wg#d}xIs1I2K;5bCQ^b@JeC8FK|1D~6S(2m=T6N3!`YvkmH)eqTE5S}(OqgCca}
z0cRkSd1Cpg%pI*-oHT(dox%!@V3iFkiKMfk;D&ltfO}eec@9b|`ujX6$VX9PJ4HBc
zjboX!pWI2s9|qLJ-HSqJDt6QwqVi~2hbx7x1(Kch3p>Y#$PmZY__Fc$#S5HKPNY|j
zBxg#*^H%3X^(o!*lcIU237d`cRvYj1_jaT)mM`ZCCC55J^sICs<keoJ6@~C>_OQjj
z&L0Za?I1Qe7o=2w<o2?we7Ru`CmPV`CU1}St?RrTecZU2{wN%KGI_bSYEC6P3SHgL
z<f!WPXy8<v4!2;{pgt@`pbb>W%j`cM-%2VcV3yeDS;?k~bj|7%ZL88b)h$77AVtv>
zMZe!4-EKMj{WNX%;PQG|s@MxzD_QEBNSU>N-P<`~I(;+%RXM3x^Kw=Yf>`10&e5T!
zO&Mv3a3-Jk+?ke@h3irG@~9{mp*E*<`iGnRTE#lRXwfip@-Wn1v#IR%+&rX|WfD<R
z(AY+%6RsnvGp<8fFu24I=1uCX&g506o;{hYBH~K_Y4U3NiilRb@#om7(<36NIp32n
zE+`ECPvyIm^pL|X*i0~6iWt?^&C_oo^}tEj%fZr_A#W$f_wyIq`%gnm#nRhQPVnBq
z+4;dy9FOiN^QYU}!rMc_8r?ukNJWe<kDfPB1?|qjpUW7tfw{GPb9NWSXfMH8qqMOP
zw@2QR7e+_126VaTGx_KpBK7x#y<dL1R~<Lc8-2&@Qkyyr-0vgbGxxsuI_f-ke6M)^
z3RxkTcv!jI$1}K~iqMjAuo3LII@l0tlDq9={FL!mkK&l@BLoDG7`o;`=Vd$nn!0`3
z4;i{mdZTVXgdZ<DCA*DR{D=6m_-E8YV{cFkP+8tE!aMrHP-xuq+}?8qoMWxMV5J*D
zvUhU<=?Uif9LoT5N4*A0p%L6A7aK$?v`=sstG<gi5z-#>XICluj}nFKG1g&Lls;*J
z9sc=x>E3CN6Tq}Eeu&hcJz#cPfXff6Kbj^`@@|X%JR9JRtr8I-n<W1jp^j@cV9>0`
zJF9>Ka>4tv=h0yMtEKfr53N;8b*cKMHB8LeDEw$hqWQT=K<NN<x6$`2TB5?Jf<M3T
zewZaVlGq+58X5w&Xlg4}Eli|wu^w4z5Xh?#2W1w<{=HwXdckboKrhBO`)L~U<wiS!
zl&-mkUg=X+OMR8er*P&%{oo~z^g{jkd}BtMg!f`~iFgD*ieD9kiu02jv*WsK)J>r)
z{RuOv<ccpu$6`qphbVLbcNk45Q}-Ki#K&U4DjU?%Zwq>dp(gruoC8jvotTwWxS~IN
zT0fs{U7kz~G%V%6OL%;zj|a*M>H<jq6DLzuVj>(gq|~}d5Z2&u$c1M&KhL({B)i)M
zt_*Ee1q=nMz4xU2>q$mo7_cD>Us%M-<rB|)e_`F$iq1pa8Wo~!pWp_vnZFc2PTSbI
zY%MGWgL}j(Balu_(k7z>4*(kZ`TY+?ZWoSZPM17mpLLvc8_q<j&A-0R$n;Xe{{OdH
zGdXEXq9<&M!4?2uoB#5^%VqXmILH8PMXUS6<7p4GK1*H@nkVPL+Y_og1MCJuFHeDn
z8d`|aj^mbzO6q^=YWqQc`}`ZojP;d-Bc%o}WZ*5+zefnCNl9JW`y5!SJdnvmw5z2?
zbaqES?q``ycIU#8R^Vq|^CKoSPuyK_q7_tUl+F6h1&6&H{dGrZi_}kJYrQy-)qSjO
zo!G92?be091jgdH9F2f@V+SLn;3`Edk&qCh0Y?UIWeA3WvkBi>2Gq@lfJ6TyM+h~-
zorOrT7V9xPn;xCQV>FzNxl4t5O^Gj6gxwO=yGo<p0?tx+tas!gn|bJaG^nN-={6a<
zsG0wo1yGN=E){(NwMATg>2Fzec=$!R(q=8>Ec+qwq|RL<pT((s0^8CyU~_F@zukBM
zfsd|x(1(~`tR<lL!9gXDTn|#TvprY2*95N4Ixl>ZTDS@IO>A|v;`*8pxl8#dynCbU
z`F(V29mstQK|m+mEPH^3?l<BJQ1skcu!%z9HUmCPte(NoXy8?#NPTGxuS6*W(f(k<
zHTVY$${dPd73jm?W4P<nU$IeJ&yuYtH}Zxo!H_0;hvx1MKdr^px(l^vh04D*xGMd+
zk&q?(#H-@)TC)N5nZz^D`}eINc0q;M7Q!F2>ul_Vgj}gRpd{A@i<kL!p$jCp?JzF4
zYk3g}-LEEfY$LP%T}36%MSL_FB|KGBzspFo<mN_f%r&eXxmfCU8D)hbcIa}K*JOs@
zB}9-eG_RZfEJ+suip-r*Kp1GWqe8$q-iZ=_V#$bRTWeN;8bW%%u3!u(c;TQ^XJL~h
zz#5;?&l>(8tvWomDVW{b6X%Fh7{VH?qj3AAFymSTjWNh5RotZC7%5K^@!arLqoNh;
zyCO~Dt_sMiB_LPtD>z>Pr1*jL{hKHXF7WXR9cLX8f-X5x+ag=^kOtnTQP&Ce=uV&*
z2_R$YzocI~FPH@7N-)65eCfu{>$$6Yp-o(zjFDY5r<2g-@}7&u6%IODyobnOoP8ZC
z<|mS&cEwP~o1sr<>JQ_`<Mr<coPf#U==Qz-A+a4pz<cH{aa)s5MqKCV$a5TI;^;tz
zr_fwj^5nyH(7m3&vecw4e#R;zad6hu{tiTP(qg1lr+bLAc!5h-3T2Zr4Rxi)xv~(P
zWn*DuGZpA{BGLUoyyG4Zr?dVk*pA$h*q+#-d_7YqN*@JlOg4ldT>w@p@1sqR60SPq
z7_Cp~XS1DD$BUmH%?T`m)Vd+Vxk+oV^&p*we37VAb%&gXv2#D_#%am0KRTt!ji0-m
z;7ws2z4Avlck+Pnyi~Yl5JplOrpzN*UQX)Cm8(z$#{E9^+z7ZdnHizfV<j|2$heSv
zxy|K}_DaWWeMSv|!{VT9R3narr@DX@#-I3b%qao1r%jLQ7Z@O8a^lh8ETCsVlsH1O
zqFpnfO#;6X!MGy%g=P8IY<d>fS?>h;>h@)e14yrI?*ie6rC!Mlaxo3ixftG~V|F8G
zo7Q^WS#(aWI>p7|OCQl!j9q%pH6~_FYb56C-il;+)hhuJ?g7=#2E61Xl+|T6pqVQg
zlW%fwc@9<v?9t-sYXXK+TE{d5qiRrQnGCQn@+}NM5inkGF^Ll61cD!e8bTEyX}TA`
zT8we`#a=<)8VxT@so|BgJvrcUCNFT2<nacCsByI#47{@xIZ`)#q30#siuMifuqmtN
zKMm&`zooA9_;!Ytklft=@BGcc%JkpGtd@R(>;Jo(<d^F>C!jBaTt>%Cuivo-z|$0L
zCm}1M)W{DD<6-{eox_~I6?bgqdf+@epx&{bCmtt<j91mM!?o4b)9%~u(O~qyQ`Mbk
z*}SfSc+B7N!-Wl5WQx&cp6~PCZWT>`-LCIKmnYr0dBGx|pMxoOVfwKGHS#yG=mIf@
zp$~A37%aI#bG*|wP~5@ZcKv0kxOc-WiZ<lr_;k(;FEg@W8j>+pf3wu?3bkgn%oa_v
z#&zvMrhTpuvxH-kQeQRu57zS!rfBBQ-WSLTTQznQ)6zF67au4MYT7*mjR%z<dFJg{
z7FxHSPLzh_bEZ0L5Gf;%RAyCU2YjmC>ovowAaq$q1(~?29V^Xum;KX2ln~u<%g46f
zV9VrkRq_sgMY>gttiTHvByNCO*WLZGXO{=z7SrS|o0e+>R05&z_$F#3CsG(fc+f6K
z4*<xN4?|7+c1RvYyCac$P$dLPU?};QQG@-Ye)L+#VCuvdh}>Vm*%X6f0*j)GZzX7c
z<^?n#bK96onXJz*2=GnQRDYdFg{>@X_8H)90%?KgL??jr2P|BM0QH7fJ8b_^7&?I~
zI@rB`5eP90-S6#QVU~NxZ;z;tdxbyP0-99Cny~A8TI<-O>ns-Ud|uMU6o|u;Bdc~#
zYcFtfLj^gUs8b^oNAfg-EqHU}Zu+vp9o4=LNK9qN2`JqhVIF_=bz>PMSk|oVHLlh#
zip#)aMk&_HnqmVYz}3JB)WMnIFuSO79ANrxUd`)BbQ>d5*?EGDk$ky81r|ts%M5%R
zZ*-JY^pF}1FetX#`+PYh9S7xZj7+1WNVM1VIOo*(SfJtXFGRLgj)n+wnb~BOW!8@-
zBWgf2vj)2D&p2j=wYhS@Lfmqypw>mcuAS|p&TZ|<fx56T0X7=A4LVv;U!f+X#g!6v
zs2a{`ySB^l5<0(y$uQf==@(k2uR0<x$~8wR+5gEquol`^@~MwgRu9i&sPNL`5PjI)
zY-bt*O^kYWk8Ab#t<$jV-<mU@t~hsyb}cBxgKZmW&zr=JH=wm}WE4j|PB&QLb4PO}
zs<?TsZ!HR$H;wJr)+Nz`AE#t&kA~D{wf$?LU~PPdD&{f*A)uJhrsYNL{dH2Y9X34~
z0lYcf)PLa%8s|=f8uG`&NjKs`QDq=cR;{$(TabBtvbPJkimJ~B4`xM@7jXHa<4yo#
zMDzSct%5}w-XskxMvX0r1tf&slu0s*ReUyT9<IU#ljb%cqs4KXAXuKFQTGQtv5JZ@
zHb3a}Ub_@J9?q&v@M$DCI`pDR;eMb3%fYs%1Z3u~aGoMmx9&<%1VLl%>S}D$Xt9rr
z5R@rF9<c57V9C=dd*`}gFSMqy>R>_VgA6|&SR8#vQtpCrQ_@PsDJ!|Cbpisc+8F9y
zd?3euL=zypAzobfK-Qf3_i!@+VGViaM4ROq(FS1DY!JIzfk}UxER6cpmt)yaH62q(
z{r#va0NKvlge%~x+mRUZIS7KL6>dAAldc*7hWFwk7e-#vW1Nnk=&2%jRTDK~A}lj6
zm_RWO;6-*;e@Wm}Nd6}@Yz$FAz3#UdcJ`FtK*+;udu;eGxkT^R*%pb*;DOO^+q=A*
zSQOZAI^h_SsKNAb`RF*x@r<R_uw2>(F`gH1L7bhCUhs)nhm?U@%7N$?k$|Heo^DkU
zc^lVU7cIF@d`dSxMFDxv&WY!A@y5H9N?waaGcz$@XXl`pcC`Opubs)g*)=&hfQcj*
zm?&lU^L_%OdG)u}t<W~2z2tF}8CBg(5Z^dFatK`Aa>c#_1T$JK$g;pT@<q)44ggJ<
z33W?<|92B_AKxjtvq!EdR&4F-SFj9%KcUv%-z^<;I*MWdf_u1h1HZsVoFt^dJ-B|~
zaSo>P$C5_fZ|^M#n0MBOGTM;`i7g(#1O_hx#yP*TZlU2e9>jyQ0lLx5OVEag_ud>O
z>%jl%u+?1vq%W{CXxKi|zgkzW?e->B9(X1$U|-=-QD4@=wie$DjRIN5!Y*UDDs{N^
z@C?Av@WZsPdSOGKMLW32{A=tQFc`2o9O2A1s*g3T_r6@!<Yhz_Aiv}0Nh?=b6mx0J
zEQq!r7L{BjD~ghNywhveOzkf)?3`k!(OSOXO0c*jDQhhYO?0~IdjyoG{2w<gw*RTP
zF>o;acVwa^jWOF_Y+=_0)jPW&e!5EKKHvn9c>M@JtOoQ}d?q5tZJ!GvusAQD2&*T7
zXPOA+2uGu7aYsZG4{AlW(}QP0MhjZY3mTnEFlGsNz`*ghH_$!*gx;*fg2MqoBFpTi
zueDG6GG>)_i$mP<@aWNDM~d9GF2?OW=qMJx>z`(|X#Q*TUor)b@!O(9o|o&}HtqlA
zQndDGX%ke}e;s{qKxIW;@p?kRP~DW)MJKMt(R40pi+8QMr#{7~$GIjo4#yUv74(;4
zFA5TaGBsMF)W58ni#WIs6uTn}6}#?KrJ5>WSXt6|pBlRHcTWtVv$UW|pr47-O2Bkj
ztC-(3&<K^SO1%!D4lHZEH3}G6on0ol7St9fSPL0p3}#qkZCFb!l30*pJ)v@1>1zZa
zqHsM|w|q7~CCYqY-PEVhdLpp|K7XkyVa0*hK1y$swM4B-{kg01J44}uOt(pqCR{A4
z>r(qY8e;|Q&|=t$oR)~O1&9<q%RW(N8k>>h38F+~=xjpoR%<0mE1c<IMK3QwH|ZN$
zAB=3AuVj2g@R4C0=pQYx_4ss#!hx%6pkWH5{7WARIsO`oc`9GpUO4EFo2{rap`DG$
z5Arr*p_JP^aYdXoBBFTNew3K4vjqpPoUX&)g7!s>jo=~zAd26FX!66S>c7w!&HoFT
zSw@07)i|<YTm{KJH*3<~Jo4+xg}5_V<y~h566F{MvPlr)m(V0_7Qu2GuKx5id6*<}
zsB?4CZ_$0sI0@J}7ZtTZ$*XEAF1_e<bkzu$&S{&ky^@`-e*8s9d>O?>ER(Eljm~Gv
zw4a%{N0e>92}oWSsl<cu)-dv7GpeG>?_||TtInFD&8VHE<F^S&Ii6godptC(I;;)P
z?RP527X0>j_UP*B`ZZ<v)}Q(dS|whFY2rD(yXlgXhp0>LwTID?(<g=PgavV0^YZEH
z=52U&c62(lzI~n<EEx{M$2l2?&W^eW(-wWZwN-~j#)Bs_;x`l8c5AGVk|p*b0%z<j
zCuyisKRHt%L2-#&rn}||YW*c;vQ>L}&pnjpV$JL$uZ;A*Bre3S*o{0<?58*}5*PRA
zYi8Q^Qdco#CZAO)W-8Zm>cHK4kc0saIy_059vK+zn@*0|L=fc<TgPc2io6>a464Q(
zq&i^B`H|(?C$NR?;jq?A1&}=7VUD{sT!NjiEYvD@87`*;<k>>^tdqp2iLOJcgmrGI
z*trYbu&Gfr*|t%OBC{wF0vUw6Fa<{yR0&hGVph$GhFB?kZYZy6WXn&BdJO26+QGz>
zV69*!B#c6sBjEa{O?`Qt>MgoGnhP+(pIrzHTt5GHzEo~f+139(_$wn#h+kF|n3{)@
zMujp9`!MTbcH{OD7jip0MqQ*g!7|zmATvo`G31nFEfNOKvZ!`wlk1tib?RD^YOc+m
zC$D|BXW0akh~dYP9N=_fIOEu&Hz~>u4#m4q+Z`3M&En7W9ZDHL>fogg5s1gm_)YL1
z(Hdz4E1g1NEP$9Tf{A?yNU?vxJJ%kh!fBvKJl)AKncG$?%G(AXv#_XoY(`%>vt9$~
zV0DjZ+*e{RO{+dQ>}W<~b~0lr>3NWjCKfw$=cR4ZhPSxowg^?7`R}kz`)sZgy577;
z?}i0jUPz;zeP)Q>B_KEyb;WRHu|zBJ4e)7WNV?BV`DQ$He!su+!yZvX=(%~7HWnL<
znL($=fzTphUPN(8FinAzAsQ7u`VFd+IFTAw&ew}a9vK_7)lJ}E`eBTt-tkPKjDcN7
zkj999t8<UZ)GjCd@w_V@BJrn-YEiT*|92n|FPgdG(}_V>()ybiRF%XFZIL{ccB^QV
zzGY2G1a!o4X1P^C$!R_{MborsH8f(vTQ|FFk`l_ZIIqCLqZ$wuw-Fy*+1hnIJXM4U
zNRDn|`I0v<{!^oxZln~l*S3HoCf~&ee7A!lou)WuKtJI2o{8NN<`nuYr@sc0!MHp7
z6J|1!Y;psjY*W~cSdIFILbakrg8_=aoJ44SLU_m!OPFTd4NX8dOgS<UNuG`Q$f#Gb
z`Lz)GAsw$4U5W5AZ^LEZG{}VbA}T8nnXpS6e?XI@yWvQ>pKZd6vWRV-k<>m&Uw0*i
z^ujO^{i~Rsc`P8m#qC-G<>VaLh5S&=++>eF;LY|E!~y0awSCt;?rHlHI3us<T?9CL
zmwDM!JnbooZ29Jdj?2M;DZcJ=NCo`OJR#3U+EBEFE1NE)?#rK;@J;}&ocBO3yY7?*
z|5gX%8TaxlDO80__D2y*0e7K^Ugv4!iS<bq1b3vpVdxVz9Z57?J9^s0M0;~=IlX@m
z4F@j^V@E^s*Q{iGL|HkQ?<I>T-6evvlfyQYIs*z>{e2`t)<q_G7WJ7c1>nf~bZ`T2
z5Xhm_jH@+h#Yuj0hnn~feYBa73xe<RsmX#T{uX6N5S!bM(7O)kd%ZDP^uAmeqy8W1
zBR6AwGV3Hd&Cw%2lpN03a#Go~uHYnIfWd^}A+iTcfz(K^J)cqi-~VwJXaApudj_`O
zj^+O<AgZ}xyU7ag{i1fN5MhV`JGKV`fm*x8ZthoXe%{gyCr}Roz376~n)2xlM`1~$
zKy_uA;JgnQSCqt@xv^&ti=xq0WBjy#czJ&~7Q~u_ls#5<ey=0qhJTkrY^zy~9ao!P
zK5=oLVx~wLaR40>r)Y%@yMk+`OT*&cDc}}hpJLwR6l^bc5IxKl{x_K&xmu^&HSUP!
zu<MC2+CG|S_qJk(wudinO`XC&D&<c3+r!8G{$y;}sR(34Uc<$oSjDbdZXZa9x)Lf$
zkubK5x3a>Oj4!F6xjozwGLZ~eQgdPxDzV;wO>NmSFUtxDC5!ccG4@U|qD5`DX4$rF
z+q-Prwrv}`Y}>ZA%eHOXuIf6cbJCsuzjVH2-K?yenXHR<%sIz+hW+I3q<tFXN{vHK
zdwa~xLGyXfL1y0YPR70msu70XKyP0qV~IuE1J9ua-m?dDRMedaO|`WlB}~=!=i6-$
z)peI=J!Ij%CWu(EQ841!hOQ|aQXdpJ2P#q_+xi;Y8HV+KfK1a89)R+0@qCsn1gCDN
zb&iF!^vjm_0!~xh+6v2w{~Vy@eiKvI(C<w2qG+t%@Xu1<&f31hQIMsW*8{ao#5Q^o
z=OnFaD(<5+nEgjVb4-Ly{*dm7V560}Pw$lCycZLBpI=WirfW<scN3s5%ihqXI9H02
zLlfd<kz^b&Q<zqD(goib&nc7<;^rSMtdWrQ#g(|}^)Uv@fE1@`B10b?8gf@duJZ%s
zg(gE&y-tT3+_IS3;nks@(EQY^;VeveOZCN%FLTkT3Zgufv$!-h6G__X`E~q!d=(p6
zeE$^o_;SUSqvO+~&ZV-mT}2M7c&UKcKM6rm84UMTwqjR`sxhgKE~<kcMEkc4QpA(n
zg={KWqB|hYFy}&Fzvwj%#tJhy>(;50SH?ZlC4E2j*YV)9Vf^UXn=~v*)#_1IB6Axo
zk6t!SCs(RUJUD5(Xi{~JOqzs?5MHoBnQyUu!en7x!%n7xIQ}&3LbdH2A$o_PPKGBJ
zK?fsim2M~L%+-)DW3jD)p`n%U(Zvu0LXSS&!=jll0Uw^sKcWP|n#fAuP{yu1G=v|v
z!tZSf1?HwFlZIvm->+o8<~&^gQ7kv+gqVkWIy5+>j!nJ!R~f`obGU7xsvW|B-T$^E
zdIQu5If~#>EEa4yq;E|A5uoA<aNTA52GqzX=yo0t;-Kd-hk-6Q2IN3+0$_MC{mv1o
zLNQO;9DWO{x_)R4tQ>2ih4;*P2RmuVA9SVi&wybUt3vuM)=*&9^?Rellg4oA^p?9B
zs$D&n)<BC<?294HJAt%&_X+=7a<}-HxcE#aCBKH(Mn}$x!L-{{p3Ry@bN4kM^QgIG
z+-cbTV}FZ|TNRUa4r@#C^41P>aXX8i7r0<8e)#c)I%+={ZjdrgMi^PhIW(gWeNsb+
z9kMV!CURex;I7mCXJ|V*SaiKN8o_ii<A~uKtliEGlXH>o*?dFT@hcVzWT7u>+E)EL
zSAH{H^!LuK_iJeL8t&*r=lsXRXU-x`j^6o3?8!?J7|AVn+&}r9u1GgryaR2$Egx+f
z)|-hpT0=<WV-cZz1&7_iKjpYtJD7+`Sg->S!Hicb@*)){4!*JdduCi^0(#lXVMSVr
zsHXjMh#^q4Ms?ix-Q;I46wMDh&!VUEM-U9T0K*k)0QTJX))3=!*D>vUGWq*45q-v-
z%V7NLfR4`&OE>+ea+@IGnRem_0hi_U-Q-?W!7+n5P_*1@3su@P>k#sydj=feqW+5Y
zWxQ-0FVVK_z^8Rn0IMW4Z$f*UiQDMYKKVv{?X^eHc3^Rh>bU*n)?f(C;W1x-*@2a*
zD;qp|Fz1~!qhEb^fm!|Rc6bhgN<_5XG*a(4(FVxx`vahPU<SxRc_I0m7pxF=g$g%F
zOanVgA1rutaYc_95p5ufk(KbTa4VoXD(A0e#E(g#O&<T1AkO1-mNQ5X#J}>4T-i0(
z+E^<MXstibMr{Y-nfo~h%hpC|8p4(Lr+j&IVBb3b==1ojw7R3Pqr!cm_|fwmNO_B$
z)=sW|uL;an-wLFjvjs=DLUHbjz2H=p@>L{=#k2ab(l58JTy%zR>2$vVE+<s2)Y=iv
z6~fJ78;wXJbo$1f#2@%0vEOQ66OHKB-Tf(-$txt6D?Oup=cAwMy0ko65L}wLZo;xz
zXg#bMWMANt{=PB8jfh>~N(;sxfYscrCUO`$A~NH?2hlEEofa-@!Fx(GBVf56#kw{-
zMvYUi^%K)I`Fg)TzZp{8@ci<wH`d!RKHUDkW89)^B^Twj;DFC5ky^}n9yqcb7e<!H
zdR#`ai3O5^4Ui9(y))KLZ-Uqykqw$@fD}x8!pCt>hy9P^k@J7@6POwP%j(4`t)2hn
zitt_0bMWELM-YMY<NgaE(F_AZG27;A9)H{?p487z(;BKI_T$r|aeYne(Mcj#7r>h}
zv);LeIqm)65Xw#N_Om^<RoS?~ZkGD8e%S?~>apJaW>R_^FxV@(U(m>1n`EzfwcWMz
zWlu`V*=2jjq-YtPI`RX&ef1Qg3b-Tx8rrW?TtnZb&{p6ea=>kTM|Vf`xf8o`E%v|9
zXb-tJ_fX$lz78^PrVq_8kGf$iQxedYhq50;|K*G+{tmve{cE%g*l#vKO&Bc;mG{Nh
zJ*tcLF8B&hzl-1m*XKc4sglmyir1gvD7WVW7j9jAHBLX%M0T*w>zXx*AVPL%qvZiX
zO{lXnmf;SEd{1wMj<>uh;QQ3c<cIkt)V@qyAR2GygDu-)TAbcJ45N@=P9zvl1+{!i
zuv&3kyU`6vbD%XuQ)2%vYSU6@EW_dJ`Ndz0m|)kAMW(`7#FBfAb;eL7?^U8<X-usQ
z`(nk-ehUH-&2$9QpwDKa;RAa&ZUO&l{i1i`B1ewSSxX_5<?r0QW`@u$%PZEorT#b%
z51A%SGf>PQ>89g(due!0==|E?WkIQkWfO1H)Ou1z9eN|Syri++7zOS@TEG{GX7C*w
zcZ2(`2*HVydto|SN>>kF^mominPwahU0m(dP4<32QYa(eXQV(<Vnd#P4wAr~Z5N8r
zQx`kVo6ctP(ay*GtZf5VZQxXX&H5;pTjv!Oc1alS6tdjiQ`a1IX*%{N0YoKoH()o?
ze(9IB9NS;K-C*79n7G>8UcHe-T>B!ld#E8Wre;;>bET3dKlHb69tl7{RT6-@i1}kh
zz8}Y*f}E{C-nuHMfHerQC>0X<I;r=9`Elb#Wwu84{b<l$Pq%wt1EsMi=!Y*izVFnR
z$qqmMtzR1>oe$8p&-i%w6`}GFtH>5ZdR>o4T&mIYOn>=X?6p;h1!*YO6^YX&sH!9t
z5Cr-RwxKA$NKB5~482CKoa@q);`<%!ImDA@d;O}Y=Vf1aU55?qmMv1(e;R8*iRgg-
z3_~hLus`3BEe5+FUIvVx5bdye+awdHaa=oXF<}RT<zk?O52L5pRPE%8E0VcA=AK77
zC~%C`3!q_25}h~LVj}&+%AVkP7Dz^s^y-FPdJ@jg??6d~E$bJrm5K&0n@3Y74YH+t
z(%NDp5H}nd6aLm-3G`v<5S&o*;Z!9?F!18U_Gjo91`ssTkY7#r*oKU*9yN#cbk<1+
zr@+ufJQrrR4XmCF3LiGLMbNs>r+)Ca8VA+%0N;-;vDYHLA?lD5r!a1Z)3HGJ8DA_;
zy|6GNQSH~-0>E2Hb``Y{ICM%MIlqUM=;K~;d^(s`56he@Ta*8nV&Ax=%m9YQ<s+FP
zJNPfa@px@8E~1jokb-48y(vE&fNM}lY|)qsXN3$KI{|H+h_*=BCM~$f?$R?jgPvaw
zA?1^4bVM)yRq|322O&muFDHVfm4+VYH%Z?X&y|OLvnoMCa@ODA(#u7r&2?XCi4Y(-
zbwUvGcdmU}2U=bLfhqlPE!A%N+v^uW+0+3Q@+NmP<NK;`?K>Q{laDB%@0uG{+as3A
zo~1OlRomH{>yrz&*MhkNY#oGHVWkDQH!xnzZ=~daKrl{{<G5>&`nR7{$QK=#0gD)F
zmB|dt+5eD5_ee3@`Cldc^O|!V%1u#_wvm{`gbR=-<DSHmZB>R{2mlATdB}D`iWgr@
z&3To%f_9$nA{ZxA&I=p0=5xoCNX8=oygOUegwuu?{OY58N`v|2Dxn<z;yA((cUN$T
zG7<u81JI!tlKrns$(3c@2Qz$rbzQqCIS4@^B&AgJC>WqIg){++)oWI0YijeX>c*(I
zOjuu4YkV;>jPT4OSHo0P?<p^dKg20X3ATfO2|?n2tymzk;Q`aR9_7Z08Y&vRV|7#_
zkO*tL@M`=cLf`zAA>y;JCV0admJ~@geZ4_-pSthkNzriHtN9wh2t;%8=zU~Bk5_JC
zfjB5yh46uP1!X-ih^6u6yp(+Xg7AXXL*T$EA8iQ)g)76yP)hGm`{2{QrW!H;au3>J
z%2{bh%Uuuo3>FR1DIRPEBvD<Ta()*p&fk<a<in*qg8!~UYSir>z=2ZnSTYg}!N#$a
z<VYA8Wj1{5RU2z`@WVWBmV>@p7C0Dbk((<7oPukwn>Pd;VymI~7zA&W<lSBW%-`(R
zAB3R#;wXTzHk{iTS#GRSx`@7(DZbR^oZt~s?=OdF0dZaG-7`u7aqJXVS<k`Wcu=U5
zmwNutS0#0QYL#;udqLT08vOui#TnCpVs~=rLo6>103H9QV-hrebp6u#vP96|?g*#S
zx_^rM8=(ugI>@Tpx_sZx06AmDn$2y9!KzHi7EAjGbH9IS*gzO3W_{v_B`r>FL&bIj
zQW@f%0SQCUpho+&xc_uckTJ9fqGlq*`w&BVQskVD1&XL*wm~!`k@a*Hhv!aYbXI(D
zrcXDhg5Fuhr~HPW<h6-Z5gro@#h1mM)XgslCZ5vx@8Auw(7;smdQnvL>Rqb>boM_P
zX6%B1xbZ)kuu^xTmeftDgsB<$L-&c(c33ltbuX5!hR!PGvNPz+<zTt~fjp?8Y_Gw6
zxK0sh*=0#+LzZMSMqXOUg7Xr&60mZYwHF~Vg1GwZE;iQszlqTq7gjtIE9+4<6hcoQ
z-@O<3UZQ@-pXY_npKLwzRzG(SUxR^f2k+spJwCYs6W<n@PwP&4Jwt#<e8Pfw`V?$U
zEq7la1)tSG0#A&QOf%Z!h_I6>ts=~bA@|%YGwE^6s6D)uW`sCV+xrv;hGsywV|iX}
zP07VMS?_mzM_e4f;+GW{m2F4}lR%dlC3%9@{=0i$Je79NV110Qt}P59%$E%dfRK~<
z#U&QHae>UBz5!adjRE?`!{hK}?^0`dY6;xX=3~3|qpoEJ`OM)Qe-K>vpe8s>QUUse
z`O}4S^5(lz#BM6deC5#+iC8f+HM9F+MuQNnQL#pxyo?pU%+T`Cr8gF3)@olo7pw*S
z`*(({bXer#`hm=$IDNKAPOLvuwiy<RG&<B~e-=QDJB~O&%1a5weK2@4stTX#OR%`Y
z&gshf>DgbHF>en`S=B7KVs#QjIJlmfd%LN=$EnIVeNSPnJx>Sm?E#P5d`J=Jto^)K
zzXA)&+pHMh$y15<fHjK_j4oklPkF5UTNwUOSxTqm94}r_dR<PCP<bGrj@mmqN5*#!
z@B!~PQrE(_kzCb_ob{`^^m1zdwX`iG{5w=F`pt0Y_E!W?T)wACyjp9(>`6Cy*}3n_
zc!FgFRYo$^1fux%i0X9(0QgUeXtbXik~wa&On(ANmV+K_&{Mf7P-DR&RTKsh!H97W
ze)4GG)z<s>o|+heQaS?$k0?5r?8k;<E9a%o8FPBuhdDY{{X$x;s7vY7v-{IBiYeB;
z4ul(w!Tl;N*I+S1Dm0tQ&u<l2Bsx7-W16Nom*NK2P*gvlh*bVE2^p0jK2yY##AItM
zPo##dp)JavJ~()JsoZi=DZot=h4r6J^YKsn&&IvNXjo~7-<})><4nuSW&&GKIS>%J
z9`{8Wsr;QxqhzejfOke5nS1Lj;<Lwzl8~7C@>iDrnyowyH>|@~Kq^#dt3N&W1LdQN
z`*sX+VP9d3D0~dh{rSulxuE7bX9Y%8I?J&%p%3rwK<h9kO>Cw4(Q;IQe1aR|d&6q+
zrn7^Vn6g4x@SPQL7B~YpSmDfs5H*#l;gw<Y!nCs%wS;h=YzjJ0QR7sPVp%cJ#ZMgm
zmw*p#kV=~9OauSs6i>Gf7#AqVq*q29;YOizWT9ghOw&j%N3stnF=a~krWw)2^>*QO
z4jln3zq2!uxn?&>s@c-M%lde%AC=7f7HK+!Du+-P?^j5K2<^=w=NdUudT=MAgaRko
zAJQ?wYjRK~3TCQ*UK#)Jzz?mfSy8$@J``LXhleU`ewLIprb|ujoZqC*;(P%@K(U7X
zhur}a1LOb8Q2l>A)68uD9bIimb0UVi6>)1u?a}etkb@>+NX!soR+<KQdnhOSKn2Mx
z5t>BQf6y;?KrOQ9y%<}^I6ppAreRy9bTnmQhm6zb<?;FM(R0q!v7ecf)#sB55l{S+
zG(un^kr+vJ;qlJn{S2EjdFTWdpSXDWmE-gI^?2d;%osBmGgzTuA(A;5F&Z(N^Y0z)
zox<zY_-b)gtNKlKszaEYHxTd1!VLZn-&p_UMaUOXsU&oZLBlk}trBOnGG-D+Co={1
zryWmO+9;`#ucbCtw6iXK4%>?vb<IYllXe<=gy7K5@L_bJ;(Qw&&0>32a@?v1{oqV#
zSz&sZi4_HDNV4s2fDNKtYV5dI<HXD<d^V6-01NcZq;WG+j`G@-%zfT9stX3%aHm1D
zvVjtSAt_FDWC2ndlSclk7psK9?On#xBQ-`l?w3IjYh`eYXA#qf_9OMV%*c?lgvk=k
z^KwyzCvx*K*JTD1KLRuJp=?nBddU{?(s+F2_nGY{bH+nltRq3+_P1ZQvp!Kp;#<)@
zTu6XMl@7hh1*qa6x~E0&dtNGBWKTjWoU$mrPSeC?h{@%KwHP&bv37DJk*Nyppb4s~
z@&|%OB_@|-YeVQhR*SHqgt0>m<WakYH&U<6rw^w}Efrt`c?T05P+5FFc&0j`47i2c
zP)Q|crGyey86r|UROG#cFa{Kb``j&Yq#btAcorS+TYf}U22B!rpMsi<E%dRWHJbgu
zQL963>Yd3CtrXTJ(o_>!$Bs(wOv7dAa*`9eC*y~lOk)8blsX&7*7DFvBcqxbOa?oR
zg?E{PjVZ_VIEVnSwV6~1{n6yhX)>zEKasg2)YGWeizc7%wy!rEv88_7tD_vGRr3qE
ze0)RTq3Y`Bah4(k%wO*-XQ`!pF`aN4{4M(QZ<S3c?ezGn43cRMx8cM?MjCc2!6d|0
z&_~?pv5yZ|8I{5D3r}*6o3=#`>cudkf-`tJs#0O|a3ym=99A?NA8+*#c6iUVXm}nD
zBuO|bUZX)b$5M%CaAQ{xdH@$d!x*nr9#5gA_dlX-G#|Tox{P2rQZGkztvb`)sEnX+
z$_b$+5mQq6>J9`duil;tMT!;Sy#?TI1Z0a;E|LPd&krWOY(8d_>aOCI<8+ulyj6$h
z4T4;J<L?}onSmT?<wD8J2uQGV1^1N+O9O-KSjwZ<5w-2%*Pb|7xRM)S!OC%{Bax?@
zd3V?}uHvLH>iSF3l9IvOgJ20vaWl;pqav(Nb!#Wz;C_?bMTlD5FTrKr(cd0T0q2<G
z=yH1%9UjMCt*zxhZ<}0k(caz3<U*g-XZ)UzolPn1C)#n=iWYa#CDL}4T=e-9vePew
z$g2ugXTbwg$tO>SODJ&<Nvt|7lC(+LECjrnn=gNUE|fI-Ca8hDgFZ<81V!m8<sYLV
za-8v>5FzrI1jYP)+7En|H-cx2Ibkfo&8OG8bAIAeoFRlOe|3L`86!NzVQKEsBRbC~
zZ{Wmksb6V%ZeJ$@VDC5Y7X(X!{qSGS^PK9qpK_#kNRuUt9RcKO*wrB@W8`V*(}j|A
zR!<urbV%SSs)afE4HjvCpljlO8WOJj6mix?$tNntJlOkx&Dl8*|8P2&2ExJ*rdKkx
z5xK{)hA3fIAkK|ib2w-^6|A2zn|qF~Ijt>a2=n7he8=oErz`Y*b~kN70|P57w)Tlu
z7w-RMY%a2D3f7f~>8t>^6m7L~-l#$+b#-JySJ}hGN0*GLF5Hj+1GH1Pb{33*TJ;q-
zu3lB_@OGQ~et(s+wG&-ISrIxD7@)kks^963jWT0NsoR%J%p8Zq;KmOyIaj&@L34^c
z6!ElBpDcD_3PQ{6(GiJLNq3Ysq!P2-H}^zB&HoLNb5nj~LyZA66w?wd{JEBQRe+o3
zcM8dlCD9IJ;fl?58gHLqcmgOK00BK-%4+gNX~PtdBn~W4W9gcP>vWwCN<qZ*M5q;9
zpVar7T@P33n}$;Ygi7iLI;qobsf9X~3iwE31{$l2El6GfvjDhQipD{$;!i)%{Jy0T
zr1HfV*wQA5kw6CoeKXlp<#3UuaI3uQkxmDCqbp5qhvaINJ+RFVMc=YM%CZI(9|pmX
z+NLb{gA<cMg_TK}R$@{#Z8cthWXl+S3jK-jxiHUY3{e{}-AM}0M43^d?sRYmj5~e@
z#ZOO=_zM9`Opx%>7uR)1J{eL%Xa<E*07_y9X3{Q4b6Eg+Sr(vlHqg!hyR8@C3OLHS
zXVOe-6;3zBlG?BYoU^PG*E3QH#|S2ef*MlCpX-GZc-^f7S&PxuGFqps|5S|I^`LRh
z++JXue@2~<6nL-&zjhnP<%)Ng^A7u`@E{!z<robUR%RZkbE8-(dUwnAsO<W;f-pZZ
z1DlHa*#*sIWz$3t{;=w+#0R%72-je^#YamNZlhJldA8>&;g50XvCc2vN7M%<opLX*
zJu>{c-+c#h4cL?d^wNundf7yO!coL+u|pNw$b6y0df{p;-n3h@)#R|wWkUqKdV`MB
zTn7wewj<Rvq}Fm<7b}HZ)?ltC*eePVI8EcFYmUvX{z0ju^uEIehYXMcAknMou}imU
zKV`ne1L^@!vH3iQwiB#l6Rz$w5G0rUpV8G+Jes#_pX-<{ysl224oPgge&?Y&d}2>0
z^Nu5mCRX%9WPC}tc+oMOqiE4vzbsHXmdVAFP6vGSg$IZ67%`+ouL=(LOedyAEXhRH
zR?w@5jT_F+{e0IyWijYLTI!MOyip!}t7WT~dB&ta-kgHI)M8D?o@lxbo^xh?DS9iJ
zvfM|_DNKqn+7It*a3>;@vUf|&`)>@UPo(~PsOcs@M3Ls8T{do87DP(=#9ZK_cwfBX
zk?5qZMfd+wL0e?IX%@seL;aZ^wo<WV%9GkN?hR_CU&B2;)S&g81C;FSYV0Dt*Fa+c
zoz{0KCULw833(^Rmj*P1RVf$0>OL6ocX$-M&XrS-$4A(F>f1YBw<m;2H5D{y!akIl
z*ta)H{yjY5l-Jd1P8CUpv=VkZiRTxI9XJnME+K84tkA$<HKuIBQ7ls2ob~{ht?xY>
zhV;2_?SXxlRTzBG3gPQXE9l*PsU>D?*Se_SdkL^3v_F8Zyu&^|>ES@p3H(Pj#{54G
zNlXm?<yGdC=8WBC3u4bl?W}_j!vJAeZyY*s%Lg03tj)fVKS@}gOdx<qtxX)+mETS*
zN;|b;lB#MKJ?E&V6KlpE`#`$gXz~Dj5Dss*hj-80X_<$8KenT#`?oKFH{tJPCRjD*
z|GLZ!J#iaQvnx_YTOp=PQMA&A%l&hG{4JX}f*ec^Q7HW476VC)MT|8=5!5+Ye7km3
zZ<Nv+_p&_k8cMDWxN`8tK9jpOYHi(~OS35IJ`&_Yqc7HStB@giURYRq{kaP9ToXUM
znea$NB?>84jWD1uRH&IB^<o(p=-^@Vc*_iv9_!m<n4u!5wPndEfWAk#iJyl#ix~yD
zcE_yGYT6WjMS(n$UKmpNG5_F<|7W15Z~V1JKoZ($K%Wyr&YNj;EAcO#oL#JDqzK%!
z@r&sq=2QK|&MM4J|70YQ4ACs-DH(J2)Tp%S<zLtf*+*IZswJd5reodcf5x!nodIup
zRCT5X;h;v{JJ_MgO~29(V_9$oYCv;>1}4@ld>@q=h%X1%css^O)m80bABDM!EY9BG
z=DemDQCv&kCb$N~s<1djC9_LW2RDLo*ho+&SFVyXwz3Ot(+6_@wKjwI_{F0S6R~LR
zB1F=&65TT2$|HZ7=>aG^AA4^Cv&5Z56SZQ~HBc2J771LdseXIkHb}X!c*~AA_b@37
zJX+vxD67GxgoBAY<!UnD=8bLCL3(Nb&)Rnyv1sg1WhsxG8yqg&!9x~tj47sDR8yEl
zMiJM3!dx+pgTo<*D$t)kPnweG+AIQWNI^tK6CL5oT-PJNEBWCfyE+dJQ7Az=bhZII
zQBZ|Bjwc-NKGCX!tA*p<)*1GnoVPe&Xi&SKuPvF+z%hu6{yY>*cz8ZPc5lCj*5~S}
z^CYE?lzk0ND`_3hYKF3HexT+n?E<siYmt^`r|C1MhSs<n__uHO>8HRW`|m`-Uaa(B
z9<*A0wkwd+Hg)%8;`~GF@h^W4`Wf_95CY*_15;#|tGRiIAy3nplIB~-ja5i=rwg`?
zI3?1`belP_g?j3r-iEe!M6Za|dXv8Ain;cV$7Dl;BHK%GE)(7JkZBZAokwDB3Pro=
zcH|0%1eCBz$Jm49OK*<ORD9V6CExU7hLi@`hP<%h(UnFz=3=#)F1K+I*{9K_6y{#d
zFXJnPBAJU>xQMlr-NRYb{RVM7$vnbcnF+iGB9;43aRt3tB@Ei8+_N0SV~(>C9tD?U
z)YA2jB=kK8bRcm7@yJb0hkP%CMR?6)H&s|bY8{R1yUvK;?0oCNJMhb?1VDklWkhX#
zN@Uy()DVV~dXpw}r6`;8s}-U^5K-Vl4_%mfw$ZU^k~@<Dibt-nX)J-^XpGM8-pHrw
z$%maw1cx$Fu6%t_tU{b%2Gs6=-9-yHIYd)je5GP?E|#c{da^P@+1);6ne*Dh=|Clg
zBKYY7-O|E7O1ot-5ExulGAmkyEjgGH0Nokc8`vS#A<QN*afWo2p=_xdH9d`zv8<jC
zahTuk#0-%IobM~`pLg3^0tLWaU#uR0NXbw0F>R7>|E^*~Cq}e?1GVUrPyKDD;4GGh
zR(v%45fHz;0r?aJ4b6H<ib~C!B}yyW;Uq|NQXf>KAFUU>l%%H)e50@8$}E&=Iy#+`
z4naH|+A~h*F&Bm|R7<q+`!+ZLPYYBbcPAo6z_@aBG9u@oh`nwCM8Z30Ggaidv-qnB
z@e!&M-IQ?W$w$dv0&$jsx}<9^freR4M(8|_7hM$xWFswTY#T(79PAs6_H#uR68BYh
zLB*QFz$CmPFGwmpOp>yY;sv$9G%^jX=esLh4X=C@^CTebwMkc<EKjtJg{O-OI@!K%
zhT5LYG=xw}M{4%DicE~X8WI&jOCQNH^cFOCFE6vrz20A!?#Ipc&;3Cw-_|yzhkHl2
z)46~C5&|nyG+79lu+xsS1vpQ^gtu+&2R4&i#+UVj*TS$H4Bfz^7@N2#fl^aw5!l>G
zD*CqPIXzv2RP?d#!glwarsn~o46$?Z)qv>4#S>rrJ9#c`!_O1{5>^C<w|^lP?6bJQ
z7bH;&TT#|=nNKNayaGrkyP?%tbSY=_Y%)>H+&s53O9^ng$5SHiQB|Dn&6q50_nmY;
zVh7q_a|-p-Ppv~w9*hCjC-*cuin~oov)}vyXofO%y03ZMd_@jUzRXbUM+{1pfSA4&
zM~zDXe*oyjZGbF_e|ETnvNq3v6MPz@2MOP!EaGz(Ufv#Asg}`-EIfv^(uV3jsvLZh
zP^5&dsghq9o(>ZS=hS8U{F#6=-Rce7Mw2KpN9HIvA5Zv)ziQ=5^q`BaK)uyU^>OtH
z=mC5|$q8sX^tQcjoQYd5y0$ULoh<L==6$v#zHfWo33g;ug-D4Pt&d1~+PkXc#S^aP
zMf})43F}{7TQrO2h>!^pXNT&MDm!chNuCL;!GY8ZPoB7g^qi+^0U)i1#aBxMtGZrd
z%Ai^6z1FsqbJ^X8^G6g<XMPAuqzY?iRQke67m+6WX5pr!<(I1_aJ4;6xq!L#>uFMM
zgSj4LW(bX(*{^r&-!eu4JV3BzR9d%S8hZ;>7%KEKr0SBR)Muk+Cg$0x_mm<R0&5m6
za$BD;)*OjdQ-PFSof||6qVTDuXC@x+iEH@L<>vjJc9($3TGX2_l^tbwqgcfW?F)9F
z<b)x;KQ>QGNBC*U+z?kkfbRULC;w4OurU6QFb>mi8veg`MU5i*tf<{O1naN+RQBjc
z3y5mY00{t*#JJW+TI1meSwV!GtSSJU+&H}t8@$}I8BIXk^D7B8sH>F9s;Qw)b-Ll+
z__lSlcX~R=cQj7M*EMv0R3o?b-PP{4nU^&RG878Cz4Ux+NKF^_IRptg)u6yd_To-A
zZBXsq2LD3_QVUtNsYYunuov0Gl;Ci}LGs)1zKT17J<N9C>DogJgFo5Yx%svqYOJS+
zI~1i&{u_UnU7yjHV_4q6O&BDARf42^@dvJq2wG)*9}W_SL3G1=L^y;VPIi)tXrL&*
zyWle=FUZ|7RoSx;&Tpd%-}O`|knp=_Dx*+G<H4PQB#=P7Ipbi%f*g2MAyalefga%Y
z1M1&4yvU@+7A9%>F=-Qf7FHX%{On*;V>C;u?=rAJg2o&aD1<mkvvh(}!-4dHJ=C^u
zAYrMR-b|V)iIQlursp)MZp7dQ216ok-P<LyGoe4CATNZXKl^6yR>WkM?qabCq8{Zv
z^(DbDm)=h4>xnvg7AJ97F~AE7rkOok&cxy8Y4%ExIGU*Ah&Z}mImST);<2zr=Qtn}
zu?pdG;u%>!It8xp#737r)EotIVPr|Lqs&9NW-pM8udzQ?Gqp(`8CbQB2@xm}eFMm)
zL%m+zD1z&8mtEZ|zMv@*n;FWyz}b~UNTi^i9Z<{OqtL^}($>;QVjLo2t>;>|j~}Y7
zxxl3Qj>K5k>{jW=^)Jq?$&NP4Vwy2>1hp@kutF)tAh=1w_zsRey<xCje1E+%7(#bM
zNcG{ZYsPGL_%<r~c<U~A^zwY%e@+p)q}T=)Eo{u1OoDk&D?p-h>Wz$Zh{sXQDglEX
z{4#z%Yqjz3x!Fdj@^F=oMTX$^$gq#0CPRSe<an=H$~*^O>a!fsSaPCXY}A7nIlF$6
zMYHIvhwez2K)>5>OG*g}D=DI&nk*RM(a`!{T+C_TH->cgjY$}H6MfJ1zE%M!U9Sjy
zOM!H92(Gr*5K3wyH-{u>?5cR-i0f_s<MA0+K=_otG|1X`e3gi=2*=+f?Bu^6j1BPc
zG%_Cx$1@@zA1uqsjxPNT37qIW12yEzNFEeK2iL20VG+x0#mA@yD#)NBw^5y%h$)L{
zL}vc)5k&hB$9>3^<tgtTy74cQ9ZDkmo)M?vN<x1viAmctBU1-~SP8)LspJ>IHD=z)
zlTG`$XO!dL?mVf38M{-5qeS?z5bwbptUWaBTk{|zFsLy%ha~CvZ9_sUa!n&BXmOZJ
zT%t@l>2=ogwWc@LRwd}uL2{`mdGR^)&2X|oY8Ini*-Pai^*88bIH8LY_;t^mko*%?
zm*sI}$ow5d#u>@6uK9d2KoP1jk!5q+$5D__m@H_Y4!VdOC#|>}##<D}r*4j)^4A`>
zvrjM;W0;<`Xa2b=i!?9FJ^ba#mtiqGh6lxh@=4%xWUOEGu>$5NdCGiuhYuXy$Y$BD
z%&^5Z7uabtElVB-8at8|hlQ!DT(^OhD|$e0*=S}NzhJ#`C4Q8?axFY6+TtZwRl)aJ
zekGq}cZQqc{Z=s0%d^wzVbkoje=M8@QujMMElttoe76YBiHM9&Nr}wojs`R*EM#en
zVD;U0Qws9BG|a5=n%@GdPmw-^dIk!owpi_k+Sfd7C#Y%u{E=Ik9R{Sh13h3py92BK
zP93q@>zPiMDg|RLO_36{fMSKVOjy4;!72_B6?FQF7PXjHXG<%}-em({XDrW(nIziY
zQ*FoP%Uqo(G5@EhzO>54F1$cNoNNKobBr)mP>kBrY;GBvLS`GyI>RCe?I050I^Muz
zX9e@I49rqaa-gd>-1ineXh6)VaQkFt(rGjKK>on=5%SQ0S`44zM7H$U!^4;`_Hi%o
zIdlm=A&IFV^~cAp8+IYyZ=_Kt19UCUbuq$_k*{_D3Mt2uvYnwcVe%F$cM6}ct8<%b
z-8(jC)#f_awzgH(mX~My=gCvb^#PH6Z!}Zx?D;bZm`vjg`*UPO3E74geO2Xj5`#ik
znt~5MNbmT=0(8ii3=DYyo<cOGtOyT}oUl8Ha@-Cm#r5I#_)Z7pCHHlVxyedu20@aj
zBW_>l><Oi|2&xNukuASDo*W9j%m$V1WPOa(lDN{sIxKnAYPaX@cJIteexu;5uB7h^
z$5Kj88q$O8*}$G+JskBCog$h@;;+T&2qrDhK~2E-o&(ZY%G_(a#9?pv5?j>HR0uwI
z==@0E<KPIKa&1qFIsrzSwKs9Y&q!}h!iC*7FQ{Vslpj&~c5m+tJKpv;WSV`i$3v3)
z^DXBe_gqk3(Wd2kdCH*f4HgcW$sh92vNC#0a$kvo*{~52JM9bDO=!TMtmTET1%KF@
zal3wcchz$|z>imbyjJ=UQ7Wv%2>~BwM2dZN#-+&j5T^W%JVu!I!ev~AtBv%ILkIc`
zPJ<RZc{~}`+MsJ{zG|tt32<gHMw;&Tmr$UInGyC&Vdanc+7L)ZNNh8&=#O?DUPy$L
zipz=N#qoY#i+db8)%=!pXye#fz`e2YlaL<P{hi(qb+*vTMN8*;nPFV5oSS0koZ{{n
zuBrNzl8x`NrqI93BHUq)fB1|kjims!R{7Z*=N2{@@LDE_N1^)`AaWp`Yy0Op1cqpo
zOtFjC=~2Jq>d3xEc&mM|J}Eud)J{QNTaoW{Oj4Ge;UFF<-1bq@Lsi*y20OlMX>j=J
ztcvIz5(Q4V!`o{ku+Y4wvM_vit?<JmSY(RWd-Yx|%Md7+Lg*(soA#KKX327jHQ#Yi
zv3&oa{B+%3Q9pXPQV^W<{vgtMa-zJVo=#CSwuLhp5EQ2HRZuo=!=AAuG+q&$gtpQ8
z^$1b#p{N8xhvz@v$*LkbCRd<jFtZ@4RyrjIXTZk@_*^KHGKNp1tRl>x@DnJAhB+dy
zcRgCE2W1m~=Vs;)64?l<RO+wFn+LFKGauGU8tSlEsV$2=z{rR)m`rh)A5h6lQ#BNF
z4I?KwmMJVU;73s5AAJlsH}&Ful-!+PpZC%(1ik(T@$`$LVflZ=)Bg`<i1Gi#oS8V7
z{>#+ZFJ=h49q~Iyub#hGn=e>u+CJ3SfrFskss{XUmY^NT2AH_+NM>41jLEoo>g$7d
zK&QHB)b_K%l@Peav{~CRB`f_Toe)>hz}6f27Bf@7H?D|KMrEGt-IG036j?~}F!S(Y
zOI#-5du_8wkPOoER<~09mi-jD)9%Wt1|BjrUtWYIQZX#bJg!U`ajIc?ny80D#L}#X
zaDJV?DgV?8K7nAkBG5QI-G*f^G`IrJJuo3{n{=<fN+oGley&H)ay7Lg{XJB}i;;!q
z3t=`m*-DFNvOE_f8W#P=hUF1i=p0+WWA)`v5gA-Y@$4W<#d!n)+&RObVzP97z;QER
zXe{SSKqJfk2RmI)eB9sY6h&ebtDq$8J>^pRWYHqMq^K6k72mnoB;Q@>=y%_0N6Vyk
zd?F#~uh!x@j0#gu!K4n`S-ECAtyzL6V6)IPs|}z=ihQ`JC%FNHkW7?Nu)3&l&L~&v
zq3@9TqBKwnQQlxO(jk2WqeRDZZ2qQ~fmViI93T}&mh?$q0>wm3Cj%mO4ezVT5wgNa
zD#^sibj);CG&@r;Mpn>alR|@Zs4!h{Tm2Mu_*l^3xuw29JAH8%hE-D47@6Rp-KWZM
z!aJg70R-lE?I&1wK)J;iKoy?I#nu5G{4h3*W%BaB{n>#t>IR86o<2WSU$FT%JHv2`
zy%=kOBi~b_RkBiFJvE9Y>wFuHw}jZK7fi43?@22?qTAMqqbm-DTSv1Od1R%!;8v79
z=kle7Y1%DpA(mcFHqY}jaPKE^M`Nu!Exn(<bImiy4?`6`4`;?EYV<S5KbGq!kw+G-
z-WP~2p9&wT&!x{Ft@*v~TT`lbjDP6z$Yb`c6DUQ@3hRVR`^Byr!M1uW{IowC%>B%J
zAn^dH5g*4Ab`u%J-VWSe+f&g5f{ZxoqiqGEgR?z2O<dJueP~gT!(H~o3zXAH<jR8O
zzEBvr`wo@lQz@<~;}OuhHIt|VuH@P+$!hS>BiDkT+Jt3k)=y3`rg4I`84oD8kBG$N
zT(=8-*j)LvvT2XyPwi^`8g}e_b*ky;eimEsr&W7>xf~jYMxI;c@ac1MTz_o+cy)j8
z^CFkz9#mSr{dg`m?M|-7-u8>KJt)Zm8EHh$4!d)@X3n~uzl)UqWY5)4ZZ|4&*wCjo
zKEK^xJ_@<5`Szyi)ufD=J*apwdAqcyb!*XX>uBc`T<~>8?VU-fOpYY_3i`TR{ZnQD
zuifbVImiuHmkHFG;>*s(*VzK%y<qC^{>Nl+$!>u)sokQ*Hv8!POLz)kw}tW0W523A
zKt0yqmiK!*nW@Czsftn4L4K##r8%aKy6c}^Gt&e2Fdz^Z2&W&!!Jhl|j2S?~qyP5J
z=n)hww=M0IqkFV56pjxtLjYEV>wGbfag}9wzvz5e{;u*&)ow$Lugl!;jJGzXrRrko
zE^HdjaA`^{6;<0STGuT#Qf@+H#cu`v-pS3=_0#*V9V?%xCx{OxKa4b|d6yMN5BAwn
zA6X@Y+|G;FfO%mdqPggo^OPx8UPW4t?9LKl1ca?4{24OA+21vAl2vxiNp=hHDFk^5
zusa2z$fDzn5KU)bu`(xs&`>ZQBnBRFh=OB}i%WurZpbrviV&RJ0|L4=80J+AUCG5*
zkv{4XM`bTI34bLdHmxuIk3LRS9=En!+6)JV#bkhfdfFUS9xvBK4pLq{k2ky<EUuhp
z7xje+PI4&TA*YW6VA{V5xIX&L@a=AE?elBx)B4GaHTLn5bS$_9Q1jGqN?kjO)t#H+
z_^U0RND-Wa`E#;NW*mE=EG{T{i$4a)@lg_C!=eR@@fTFEgv|na<uv?*2QcL7Zv{SI
zBUX=FwH`>m38+$!Q`sBjAi4Y#?cEyMbofFj^B_Y|P4WgbpI|GH5sa8;PFs#&d{<+8
zeNQo@fVC$t#<9pkxLdqsU-0J8AMw|%X-IevkuKR?{LYy1W>~%eIW)nX5u}|tf%9J!
zi})6z<}))&2;m^e1(Ilg*dw=GT0R34s7MD#zqA1~?uR-#6=sg~C_Z!I$fE4;)`B4i
z=pQYddzf}L2eldFK3ez=szUQ&urn4Wb_r2w+~YTweE%acLUppf;CjZHCMl6Mlr{pj
zU_zkf<I4j=C}v!86&QTIGv`~7pmw2A<d5@2b7oyGk^^En2fDhsb8|=f9-AZgXm>9X
zVL(OLoEo4H-3WKb`IJpzcxCaH3l_X=8BLfmRx0||<>l+=*5zvhcB2QSp7F!`lC|!O
zlC#LVp-Ve9%TXE*dbh*^60VZR?Vw-k_hnrX?st<+lq=E?Es(1f9ZOs#E?v^WR-62t
zmNrE+q&F7wD+}4Jhq1`-lU{Pb(Z4@B`X+y$I2t45xtXB1c-eFaTS4q0oq@-rJZMhZ
z%cv|||5uOWR6hUx(If_05YocA-;`hsRjyeShWDW@;*a+57??zrus3Lij1Ar`cwLmt
z6!{@QN*}llKBzp>XR3O_k!bppY$NaENO(J?V`2$m0NKZ*UvR_)^@x_rt`@P7#K)-o
zVa#Zq&g+!XwAS#d-f&c#noq@ijhm&n0%W>OpNs1m6g)Jh1D{!J6<9<~D#SdWi4A&T
z7|SvOXfBBXU*>Lbc+qwwLx4Mq1TqyOsuI7EBe~(y_2aGtpY@!L2@4<#nh&0M)!QsB
zuj`1Dlx4~mXOx$;mEk+lBFYz*&6Z{TnucUu)Gr=tE6MiSn)V9isO7>wpZDMf+cxfL
zK6S6;dS{1lb5RK*(_SgGU-M6BT}nU29K{`lO-`J`yn<w5Qt4KRB!)Y31yqBy@1r~^
zM4XmFzw7kGfvUeXG*njQFJOd$2vr<Oj}e)YK)&A5kSDz$mq<GyJjkxf1*7$#X3pT;
zoYFYJAn^z~Xsi-Pw{5c=P$9>^%_&YZGLn&ln;$udE2iF*y&^W`EZ!!uG)gnP=<n5v
z#T`TD@F;O~uUAnj=OT`p(VuK^;(2bYGaKBYXP^trGTjIa@9gl4B=@+8@KNn^!ae~=
z8m5MC_iCg*B`8QL@o~H^9m>0BUF|>P&reZ%k){K>IE3xsAbi(g8_l=bP=PCX$8?4}
zUknf)&mRObZ?*(e*m1hl>{$a^QpAU$h>k*d#dnZc2HB$OrHIQG!7f+^-MJIxt_uhi
z9zh!`)+K*vY%9^x>(?*(+d{+R<ZVv?Ud-T4#9KoXSrltzTk)PrVCNmyWu2M&`0#n>
z5DENO5%*c0fiY*_3<O@V>P?^qMR_-jNL9(!(#<vwjj4mc&{=sEzax)!4VwU?rl)-a
zjrYf2vwHjai1tHH6uSuC0Q$nBrQ7<T9LqWF;E)4snCA}AfFK~-i9Ty%tNr_keXsCY
zCNj!FV*L5%C=^P0prP@jwU*vI3E0Cr=|*T}S^6Xrzb5+Mg^{z847m`<aV-&sYddGq
zi96|#Q-<#HNn=zTxG4oc0C|%83&JzpC#SP2-*vD=^wtJ9Eb2L@j7O$y=M!cVoZJtI
z+Y?hH;|?*ubrcfTINLj2WY3q4X+wP)e$m*~W15>}RxC)&SP_`9P0=HrUG3Z^#le5G
z8?l|7Dw7D_{B8>ypbjFt*4XC$L1clq+T7cITb#^Z9*U(i=opMj(zXnYUM;R{-}?gw
z60i3*Bcw3~*2zXS8J?7b#Tjwy(#A$PT8EAmW;ftg_ZUKZT5=!7T|&GZBOq#c@WXA%
zB7zx|^Z0N10V>AM#{UPw&BVa^KW%7?9KS&H|0csZrOBO$%?@*Yqh@`UIXMR+e#>ZN
zMKne}Qxtj9L83w?dqh-o{{YOJdRU|CWybAVzl|*7Pk@Jy*O2v{bNKju{pkGiUhrrf
zLR`o3$0y6bH|~>OT$hDK1Hb@9S}wO_XInfXG#8mqQ?mHA{<HOf9TyLUV;isw*d2BE
zR-g6u?>_2~SKKS@o51(Raq}<uT>2&|F<>A)mzeMTc4UFxjR{(dRETu+bxfW1^^++5
zScyeS1c9X9AY^8tW)zhd5tl(+m?prNhXU%Nrm+4@i>sv)MGd<2Rw9>cn@kCikL;$D
zB$oHYMeA(f?3=4~lu&-`Q5X_>#EGN&I_YhCs@3_grk4++?R`HsOGh&`7j0JQm^p2g
z?7A1}4e*#DG}az}DJ|a)dChW-<C7$Ye4>J=dS-k77x0c&C_7QlJEapt+pqv3<V2VY
z&i59NvIWAppj_Y)iYB?kC3gw=TXx7MZ+IG%k6dv}t<kE4!57;_LB3fK0d;nAtU76s
z%@~Q4XiI6o*;!|CSFznt97A|n%aaVjA%$Myjn%jj)$K)tjTL;HNK^DP9>TvH`J!Wy
z^D_r6@%esc23Hal@m61YdiitS=e)@7j^gNUqhn9da%Q{)@I4ce^WYG`IXHOpNq-g#
zQX;9<6DCz1rce+b&J`W9d_Cc5Y=~6KGJ$HN_s-qR3>riu#KnkqkWmN<oOn8WzUkEr
z<IAEjK<t4GN^X)|06-Gp=w?ptu{YD~80&n75TeV7$CH7S2Xo=+_?pIl(f?*ogGm<>
zH(_9^s~E7nDz5UD)*<35$UQ!uc~6N6K>d3HP}V&|*Qmp4I*VLMNc6nz$K;XBF_S6?
zu$Nrv<g)}@r%`00!q_A$CK}CkmG=Co9p*bZb%tluwdeVWBTBdCY#Ovi3J5u}lV$4T
z+7;XLXW-DYH_*lfn{k<Z5t+W*2@_@4U@-N7_)`IHod8habVgrA;zXcy0w+%#!=P(7
zUbLI*>ia(ocU9Yz)05q&UDS!SH7yP}S4ukyNB0-c=SUCOmsT6rox+q+@2Sd_RSXqs
z#vV_W(=A=fPq0|-8r7)(wXt(E@(<%XW@7pqPqt`iRd6&-4AZ^L*MzD#3{CFzk@<xH
z#c_ShJm*zO6%DsrJC1nvC*vH;uMPn!q|;GrpvsN+Ag@QcUBCts?N#MaG%@HiwI&wX
zzAQ~xBCW3lXcp}B0NjD<lStOpS8D|YMh?X@JxjZ$E;S#tIw#NL0a!7}EBM*+H|eDD
za-HHmSP2DQva|fy^Mf2>(ZV`+>Kw0Z;^o!Tv-ru;-nET<0?#o|^2vv`aAoR2sWPCW
zD^f5`-)^`^ESU8Szq8|~R5ySPELC6ZTSCHAj!zl^9Koh?G+t#6GxMvNpBV**I{fpP
zX6(q?bs}Q|ofYNd6sE?<z7Wh{+ms#ixUt}++dt-PC?cZ|y`o}&JvU9v2FD?TrA}y)
z82^f?4uEFES#s-Ij6x_#e$1qaJm1xIq(z%V)V&Fli<NWxp_#0rP`31Ze^}6Y6$FMH
zFc?}YKT+L(&|4Vk)Dx>&+dkly4K0|EGz$0RVAyoxvQ<g@GnJSh@~v0#S#+rK6+3>?
zJTj;nDv&T^_dPvYB{ZeD;%B(%OfAzy09@Y6pbJB%mJha8l|su`v7b?Oyv9_XXPp-!
zT}K5^Ep3m@@VBwH&Zos(6%y35)_l1kmabd$WNec{?Yk>pdKvdk6-nuoxW`Yfb}JZ$
zMbe%N6LjfFKY7D>1ww(!)2X`%1_H<FoAi&#s0I!{zSZ4YiMHOw%b?n^x1YvEk!^`F
zsf53an&N$0;Kx>Uxv56V@xIn*cxkaN%NO0f26cjD{*|e2InjPig|pnrdZDd}0~@Q`
z1A`<^5hIdOrTkBz%>}ry1BYWpx5WuyF>JY;E<Si_X9}U~XOT-~5NxZjxUY`ePp&7a
z1qa$@S@f(8*_L4GMputUnaK_mu$8b1?NXC{vC%eGy;>Uf9%%G|@y60<`GAeqKMcPT
zS=J(cM{6fAb>A$aWPi8%cLU&Ujwh*Fw3F0at!XRPm4;o#p7Ns!mH^YiJ*@#Xk*QP7
zJFx<vGR}d|4xePRkNL9D>@VhA>aGF1WxhhiK*FbynLj~!daGfD&i$6|2E@xsH`aq-
z1to|dRDCI#Eqh5KbyxK_Dja{3aH>@6SzeS+5am7`>E~PfO5jcY;&1IWAgOOK0?9&G
z-wrK=>I49VR&L|-NpmqyGy&<cRRcr=eM$utm2PuZVTJc7#KDwC2^6qhL}+dEc|BYM
zyfWRK_-?mrvQc179BPr?8_dvq@P4DMMFk!!Zu=?(=tjmlS)i?Bx}W!*TOM?dwKjrT
ze4iD(<)WU9XHF<4mL8adZn4p-L6L_E26p?NwY;-yE&Xg_SE<<eAf&t;ENuCSbrE;|
z$MO3M$^QS6`<NIwIRC5hyQDc`OUe$n^+9c!u0Q*iLV-Td;GgS0ScBBAj;X2u(qsX8
z;?m^KOuS?<b-6AT_ccl6MT1n4c)ZwRe$CP221%)$TP(fX4v$q%)@^83hOTcm1TYex
z#C|dB|Lp_6|Mh_vCXMeugu_dfv+~0Adhtwsfy>|_fk5_a6@0+QK-#fvn~zcil@5yS
zxux#>G4p};vFnav=GlAjhu`_RIQrmD&;B}Id#4Q7Pd-{#D$b7cZBT{A)OLKNesWp7
zBwdC0+o2Gc2y(2P>7+>-wUo#zui!ENV~QG9E^NP^sX7F0f}_j=RAP!+AjGVRWk90C
zSQrlvwVAFE8MDDsDDDndo$iO@Rgy1Qq4Ks<Z?+DFeHmlMZE7glWHt5q@vwpq!ND0B
ztaN8$CWkN9FQnQo6%V1AW>1OG2kqK*u-bx}k`_Kmg#a6+0bLr0%riEJK*0$?b{JeA
znSJ48*iug)GWZys?ni__!aXb?b)kB}Av?~veEE{K!4?E!d7#CN$S@Ah1soMam$96~
z1z6@H*ltgH44N99E}*?^I&9C0cM~`vGa_rhHspnzT80Uw-x(-{nqb)~662tei8*3Y
z%^E&obMn7mXpN;LYOl%<^pg^NDKUyGV!;cQpMJkpdg4dL%U~5mb~KkpcFTbML$QC%
zOaVFtD-<a@1|Xa181P47OA`q<@@yd2Cx>x0B+~lzpus>yE8NN8IlN)sK0=dA7V2{n
z@LMDmqNu+8UyQwDjA+rewpq2ywr$(HY}>YN+qP}n+-2LgZKLac=}vcYPbViCzgAXe
zva){6f%km|I!pOHHKsUa(m(czcE*DOh;O5umgf1W*lxUI^Jk*BeNzg<h8mg#M$Kvw
znBxN5n^;D`2r~GQF*g$C-FD$&RD=mJr|F<AR0bWls~BNqMl@LARmhvA`Y>dE+>Ziz
z<Rornc^z~nA-ZTJ;93~U=wTl8f>ir_OiDKo+=A3`!ygRKv8WNr1$FsSVimHPkm|t*
zFnU)LnHaG+H`p2$M}=zu2J3QV4hmOR@S$f70J(;MU(YXa=C1+^K&AnLxNcsxAv}?(
zj9$bo#VSUVV$EtYD_a*Ipro16&P)GS+iZ~yoPnajYvavnnL0a$u@%f0!@nC~3I*I!
z=#qu4gx#{zg6RqE;G2U>TH3;xlcixv&(Um4MlDCAb-AT{V&GVFMRJ0TUTfjaX?Q0_
z(abt+5TXLe$_yZC=m}GRT3QmWiitdIgTWUWV4pgf2Tk~0o%}#)2Jl6t@+YIxsb$&Z
zan&GcNdKaiS_Ffj8n-fF!hlvy9d#jX&d?+*IC;fD#IN+x@Dts5zyJk?>f-f|P@$2r
zl}4elfjd4T`6_I&%1XA#z?hSeg=zz#Tx|_~@1(c?aJGkYPZq7Xd1nYJba-NI2ueHN
zU6}ISm~tOi6fZ@7oCav4>LICnjAwqtUAWQ694RNI)!xPMS~+86B!#ecZF7bp>?$xf
zGqzY<Y~9>>2#BebbWPLKQbPf&ff+T+x!i*XnQpoAn5(g)nyglEno6^k=1C&D;$Xxb
zr8wJyddG;0*96#oT>xJ5()+M#X3OCWhpxrEhka*c)=R%fWF{+=L*(VL9wfe^KeME)
zv!mG4Ym{Q3Evg&4BvJN7RI`8B6BF7QYCKMO<DmFC5t!mvNM|0Xv~crJnv;mvXB|e{
z-@A&o3Qf{PJ$2@yo&;jz<<A!zY824u(E>5pKfgzN8KLz-8yT@bMll8r@D)=BCqd;V
zl~aAbqm-<X(7PLLS-zf@-Ahr8AUWovCe0@vl8-t`FuqB=y#N@yw4`t|#aZT6!?z<|
zG<lg3LFbl+Q3nT#FDtHeu0QIj)Z+d5m_MIyw6OH~e>&+Y;;^wP{{1k1dAtbnTFMyw
zNbN|G6p2|zo5wTej!}2>$w}m>@_RcADKhqCvc-Gll!nQo+#?yN9HTRIQSu@_X0q}6
zbX3xD1BscPuydzcCgaZ6pS2P^Vs6-!o_nvA`h^&yOn?A%bMCg%ssv7F6P)V0NMFR&
zY>#*?R4Fbjf3*BIeste%0=b)E78~DDY=4rOZ~Vp!y=&^0@E0^sj%s9gD1F;sCP2#n
z-m8XMOa3|bw$BC&+7L1qN{R2waCc<L^$9ht8<YpZuR^p(=L5}7u_#_*2$B3&xTr-Q
z7wU&fJ2)`J%H93v;b|vUVz^p6-?fj%-S!Dxq_l8Z?Sdj^J3{Q@Fu15$Nb>8A62<Bp
z+^8$zvc*2cV<zj7dN`eQf_>0DUSS7ydPQQ@km9rC^6xX%sgL7q^wV$cN*p$dA(Mur
zc|hD1HSc)51P+kyfG}^iXUC!So6X=jPX>KP*`pSC*s@IVc8&r)>fY6C=>4noW-y>c
zruG>SxOD-K@bA2F_=_X#G^$iVURAb6?U;iQdE>qoXa!-;BCsV2zd-RlkZL6$XL=lq
zYQ+*W4#k2`w%W<oj6Ep|JK`q?M)Zz+lS{*fuI4IfZQ@|6Uja(QbtUAo`yD}mGp69d
zcpbsP%;(blsU7zw;Y)k{2evu_stlq=cTv8Qkqt4sA8)nX>`+==qrGVXcCczU>HN#R
zRAmU%KPu+^xM`M8?`9*_MtNT!$?A1KjGIShj(CHw+p#bkdScd@QR+T=CEbj|bv<ou
zJ-J2inKab=&ic`T;hQW*v-rwc^pxj}412MpH32KXI?_Ucn%I883$&17<&y-n@<)S%
z*xNr?Fmpwn*ZhGTxUWSO^U@4Fo5bCtH3k$}{Xy4;bGVk5b7gz=dhlml?o+~MgB*^P
zvwE77VPknzGu-+GxciOwu5J+QKjt0m|GVsGWM=&D+-Fl7;<ksZs68L5r)GE&#|JWl
z6umWVnCSR?q__INdT0J!Xou5Z_s+x536Zb@aq}W#=Ny>2ad_;_-Bsl4ZLU8Yo=uL9
z&2_8-ObUmz3N)QBbPyW0Z}>s}Tp)>o$o%cS{p*pxqrM&Yo<f#n_{9G2ROiMu0b2!L
z1wIGy%2fp+M~o|8jfnXgJGIv~GY1E$3$TkRCKLy@4cGqc%a-n^Z7+QFqtGQX339CI
zF0ekrKS!#($rI5ZA4Gvgj>@XsPeL<K0o9gy?+?Rx2phuM*W4T_qBXj`c3fK5iCa;f
z=BKw<=f@-NtFs1{{-juH|DCc{CtMaJF(|FD16+9BNJN{lI>b?0E4MXZ=&~mQn@czx
z>l{+hqu{`{gF<<FqlMpW`V;BZ^Ps1G)vdUBN!C?f`)wu|K22INE;)mVBDDy;xe*+M
zb;Ka!Rh}`B>F-W3JGPw2J*Jqt&*PH_81>ycpd+nq4tiQh4hR;=c^W20^*EQ5c==#f
zo$!70CQusA=K92S@?0g7aya=~B4OtUBXj0OF>;au>p$}8^#_f0D(6kCmbUARRL@74
z6etWg<(wY}3~C%}M%fx5A-KZ_v7(X&fy9APb|14}FP{z%JLyqP_mhq+pWk<+U2;X~
z^*-<c?g|4ZD|@k-zR#Vc`cE&X2!%EFhD|owe)^cSVqi>EkPIhV6vKOK52x(zc0N2?
zctkS?vSz`1;V6)!6S-yWQ;8CW!E$c<obRV?7H0wi>?$q*l|q>l6pSJ6TSTJ@*X8wt
zz&C52IheXq>BM7Is3-CvZNvESISqXNI(g$+O&P|@TFx=*6CkExHHuzXU!rS@W+y`I
z6trZ0&!hfp3%P7rt&eJkCTR|nJXP(H5=T_-LwcrEv}L5TL#rnpWeteC%&>$?vT8~I
zZOHukGk=$bLnD5M1`bgUxH(sdyM%yT!l(xNU~PwGD0V}ec*kVY)lBBYtvWs^b-|k+
zB+>=AUWlj?QIO?aV~}Z4{qMs9eO%HSGQF8v%vVcN-;p<jo+SzEhm~F2(68j~DW53#
z85*pOrqo60Ku7n$pjqCQFc4U~4o0wF8-4{~SV$(}W_*;j5V8T!A&=_e@B{;$X>y^0
zcl*U2=W%58#9c$&#?7Pa`@=iEjG#6|phD6_0!}qd^>P2gO<FhTTp@($0?-X))=&>s
zUq^t2C*)ivH$56+3zMW}(5tnBTtNE*JhC!&tZ`v0+m~n-|6Q{pxX*+ghmi}m12RlY
ztQ0{85IxxeBJ@Wk5Fu}w?jI6D6W{^}0zI7BFx`miopa=J{mBwkXd4%3^e~>Wi@K)e
z&d%G<K-9Jv(Rg<HkWbHTK$@=L(Wm*qMH(Jy&<r*lqYB>q>p{9yRdas~U7Y*ONhCcz
zp*^0EYf>s^M}ZIkzL`)L%oxRaTC`V2YBXn5U#=`5S+ZM-9D%M>SFM84NN!o37DAmY
zFiwgz>a(B&d?`$jIV7Kk)UXPb6^dBs7$=*_O2idGeOUVhJ7+@ataA*tiLkIa{YuF8
z6%ZtVg?o!Z-b5kivanl(f7)P?FVdJZF98g$RAv`D$>Cmd2L%MBP!s%qzXeee+iVBp
zNU^Y2Hacw%BUOv>Z>k^DLlet5Dg&n<uGZJv>+Al_hV1ex_ur3OE77&B?#~wluMM0l
zr(~<Iw}<qG4V${styRgHFLlxYXZfb{7YaJ0AV6DB$NDu>9`;ZpuA~btgXQE}T28@B
zxJ)>-3Bh7=b^<D<{YRFJIGB>N#C8#^n07Mk#7Y%VOS7B=kZD=1%~P_)bJ1VI{)(~N
z9Z4KbYTrhCvMvF2-|D|Dv!<-qfAd@wMqnp4mgM(-%E{d_|DGGk#}^ypASvijRfGOz
zj-QTKfqK-Lx2W&Zeq-f`ofvP8D|Ki1wcEN-#N6wiExwIQ|JnDH9cxDaqIbR!OeSrb
zfypaPC=qd-8fUDoShxuy<t#t(7N0Go#l43VuP$Nj{3zO+Ip~xllfXOUDEZo444SXG
z2?WNjki-0DqG-3-ER=!ZG0w=aSXm+IZ94D~g;QkWRLGIN6+9Y-U68Re<~++_6h7)N
zYB*Hl7yM2pHYmrhFu*7IxT7vn>R@64uQu<T{YTy`ZM+gX_AJT!9T$VK!mRwun!J}I
zgL8iujHP+k+CD*+C>}r+OrvDVh0DlY9O`z_{@9CCSK2j2$x|y>UeY=(MK1gsLsmP3
zyi{Or2XBU-POHrFM3+016)~NEpe_AblQsn&V~}~s#+8Xs?iEwm#}Hc*dU3)y-ylcd
zbmCLn4vkS{9)_xGN{jSYLglq1agp&lBVTIqviFugNgNjhL?w;<^ft7VU+|#_*ME1?
zM`T^!{ZGJrv|WqMlX4;c2=w6QRTB$TfT~jeG`aVHzFfx5BXhNMVBBtA|J9j$t^F+f
z9!D%I1PE$~Nt5@LKYTvqyfV$V4^Td?V(vtB`nvkr!zm#hI%x1F8bTM*c_tnxe@B8$
z{rwFJgQCUf_5i*a5AGA8Tc|$#OW_K8zQ2u?Na{6=Ga-+ZyWvD^qy)G?zx@_~nX#Fm
z))G;1qF`pSx&-FBQ>QVgSkWo?P^y&b@MYfiP1KphsYxEGb)Jm0Dm0TpOqf9F{I-v;
zfu2Er&}vYxr2e}9m_%SM@-u%BiNW8cl2+dA-ak5eOjAPdJ*UqzS8H(miru#OJ*8Ky
zO#RJDE`mOpG$uhdHebQ>0#m5}rcU!ltsDj2Nwr6FK^|9JKGjHNp@<)C_=htKX|C$m
zpQytHi{8{M9`bA^|I%@pp_kitdd;FF5ZKEyW!7u}EdX0rRGxP*t)!4ljSsodMr<h!
zmwF3jty{|gXX(#Z8h2_@AlV_4d?Z%A(nn(R%hHf1cfhw$Mp!FLgeI&_tI_k6zw^~6
zpp_Fh7em)@5^xQ$AUV-hLUimo=m+RjdrrP;^JdzbE0NS{88-@n!XTBc+8fx>78_2N
zgBE0D*5=KDXT?dwJpPluA!H$*v?eQM%7#h+ag*4YQ}ddD6sb%jT(cZ@?-TtS3hlp4
zUt<RX0dd|7YR36UE)yjcdF+D}x*swJd7^43_sZD+D4L(REgQb;>=7<)ohQv0PmwQh
zHo7TJHXh_2tZ^!Si^PJzL%Lu#NOOKRiHEqlb^_*d_W*QZB1e^rN8O%Fzx=e&3b(VE
z3h_QkhlEN7Hbcku1J<|JA@H9vIR<*B|J^}i`Y%OMQyP-7hb;&_Gr!I?Q2vFgea=oB
zy%JXsz#XfZKEJELc(-T}7q)PHA{ig=o&>RQrOKuAaIk0oLR^U<N)_)awBpX(zgyki
zzvha!OEW7pB`4%IPmdJ@T(eJ-X#SfHzM`)D>odcbE6eJ&1CL(gW>!$u-kY$_jT>m;
z|DkZwOEm3(NTM^KGq{D~Tw|ggxQ-f*NZs(`c`q^!0IqaTHvN2`MPg+Ke_fs$rYhZE
zTGLRJ=3zUYbjuHi=*(~2Fh69=1xBQnMYX6k-E=vQ|MKnLOXBpjiuDRC#k=|&{@UKr
z8z*>^610v6!?dZz@wLE<(C}DTyQ`JEufj3y-pxZV|I$g)0V^Fb(NOHso(!<djlRNl
zr1CXVRvYJ4UP+8aGVEF@4fT$?%FQ1u?m7Mr^=6X6w@KYkRA^<1otbGS-1wCU7^g@d
zD}k2-1t>47gXJzSmGp@%4$BKiFzIFO2~Y~z_2t-8z!2X=j_)&uE4y`*yk#vDNb;Ne
z3seW_((TSrUX&96ep9<-6+(KND(UC_k#jFhCc&Mv<3Cx}_vFLc<u*5z8#!2*`e!4f
z8)RQ~_5FNzym90F;>z_|urOY<`qk<CM6?=r_xhCo@qRqfXPrHV@JuR*mIELzQ5`dV
zg>)IaAUxk<J>MUH)e^0sh+>^{H*kwTT{=xNC&rNH@7ZKGfC@Y4@kVpawi}*ER?lvD
z@+`Hs%ev=#JrcrJ@17Mk0HWFEhz42Ab+v<<*7kWwFks^J*ssY)EQ?e7CtM%pvUJ+L
zjTnio!z@G}vG*b?c!xz(i+LauDR>%CgEoVP7Sl(J1PnxT^!R{Di<t~SEYl<;npVRr
z`T(h0%We*Q>@W!LHo>eiahUeuPf><N?{*PnVVBe-xRVx*(UuhdzOC6WWwirb$m@s?
z01R-E?fPD;mk+xbBUaUw^ZxsJMfIXfL1_1p$>wIo8RX^?{_G?o&)TKDoYsf{T6kAp
z@ZPP}aqqd+_xoTWHrL1fXCSl_=VD`Ei}sc#L`Ap`oNTi3ZNCEi)}68~-C}3PE+3&L
z9~P#x^s2-ThhFnA<<OwbguUI*?+jdj!Wwb~yKnT5lw8k~xnvt^0=ys?TIs6I?^C4y
zW?OOEU`ut6`)WuYEh*KeOL-WzSu4XT!<wasVs<4jeraW?gMAI`hFgU7m?yu(XK%T&
z`~o-;L?J6YA2T<L&{@VwJacCsvhqFj(;}Wz(>6Sd5FX&*{FKu-ocqfclu)L4=0B%o
z&{;G6nO=9u4~H1t<`HrBj0Z~4EsbK-42C&Z{GIA!4kq>BvMTVaa+W+*biMk3%w(AT
zx4kO$k*iK(Ms~tu^gbc@rG>iHVIhOV0NG5mhFhe~7}cxAyZ#{z<#3{$Yg|76ZX6R1
z%o@aX(R#sNR8R^@FT+2S22E&Rgc14m*!lC)oRA=4o;@Gw!um0Q)>x%$SdR@e<;*^J
zfpbIqQ#2gc1Nt%Mpz`g8jbj89S*vCcgS=#a0SlFKs;uMXb=;5Smw@hRP2DiPk*C>@
z-Xe}e8nESFQZvmeg4J(6iOm`XsSPR<M}wweB)F7pl8Pi6)MS*XOD!gDIT_K)eQ75Z
z?fJ`}b>dSrbCTulM-7J=kvnk{N#`_$E8MV#Iy1-(I)0e*lxOukjA_AxY99X#FLdCh
z>(UY*dE}pGx4UyLD`MTYidO`|?2)NDdkpT*s_iF&<E8JOVg)faW;P2;yY$Ppo~AA-
zJR?W;(w!)z&yP5c-wa$v@{L$Lk5>$TsbicHls|6@5-aj0Nvldm4JbKaH%U8gX(pRD
z9u(3?Bjb?UN(6iKQ$clgObm-2;~p-d<5cfe^0icwVMtVC)29G<_ZpxP9iZg&aSoKq
z*+t_#CW||##Bd9=tt5h6ht9C+eEAKEkjs*)m{?Gi*Fc2KmTM{)WmA<j^=ml8B14O>
zqeNPVl&rMRF3o6|XPGR+0e$UF^~(WP<MPk*f9D2%#r^8MYo>X`qkN>gExKpBeE4PI
z^JtDd9-H3tT<q~fWR@{StBeJ1Vv9xnrR|>v`|+}1tBD2`-16OKcM>4eTF#Lo<V98I
z*d?AfOGQe81#Y?|0*k{T!wxOh5>GcODY%c_wbF8|02}5DY!E^w1Lf-)>g(e3`s3Db
zxk(!gLn@9hqP%~=EM2;J*^Ru0Ac4rqhIZS!v%glHR78O!KU6u-s>XpNW!~@yDC4hu
zzUw_*ACA11)svQ<<mKVWuDTs*vbr3^HrD_=;!7<|YTi%Iy{XY;lyxt>A+6up<Hogr
zb#u;gB5Yr0FEUK`OB)p@@$hp4Uh+;{nNf}gMzCZW>|_=HG>CyZ0a?ii=L*<Gssyfu
za-=g)Jl4Dv>b6zZ0^~u#vnEIx+d$esha4<122-)|NlHi8d7<kobN-a)3xs7tsow`M
zJSRz&ju&O)!2t}vE-*iHEZ{m)AA8x^rHvd_!oja!8daMW^NruGJ$kk8D9=Acygt`f
zTC<B!g&PXq?JQlV9j^5;|JnBGSq~<{Q#M+VTfEvg?kh-6wajq6a1eiBm`tx?JE2&^
z<zkt$MQbBgBD}lQNZIIKkG!cgr_H*;M;1<sU>RDe1k<0K$BL%0n#LBOX>6+MvQfUv
zRaPDaO!5a?g9S#W@kV>6#J8(MhgQYspO!^@yU6S9)f96AK`z3vE$l>uq|p2KMpcKu
zqs)F#+Z#X-ct|l-7e|%yFmfW#Saapm3t-RM?C@pmO=4%gJVb?WQU$2D4{g8+yO*-8
z>~|X1#i`7%Ee9h!Q87pkn3l>ml#UddhK$A3XiInZ*Q}a#d`7LBX!$%0XJh|n=`S3B
z;7pKaXe9`xoGA-3uk~LD0xS==u0?Th`}2X>Tuu)2;7?vzMKMm6<-&)gU03tpjWGEp
zPU|W%jjVFj+e+&rik5*Bt%%HJ%XE*_HI^Ore8nM$-N_HG*JRJS`Vg}!$@L(vU+2e8
zK-285_{M6F>%Q!1yqHNhR6#|yt7Bi%Z)tYuxv92Tvf+eOIA0TZi0)%M--jQK$#>(1
zB@Vt`9dCGFa;4b+F~4Hq_+R-I0}Jzi*D2f5kd8fIh3-C4eQk3@uNI0&1hE1eos<DR
zt&VJO=0?fZ=?JBj%R8;5Bbh*wbopdRG)lz9CTlV0Nr*)j7Gm>d?r>k<<hA9wre@V)
zz<`^2kTg6DNs%o)DkvPrD$HV;dRRF=k<|!HJkWt%T-V^}@-pRE*s5lsmPL2(A5}KI
zw&BKXH#mw&HuXEKIs$mRerVQCaV4CIarPPr#NK#K>D0~j-8_YE;rvI~RJ3Y%z(v^?
zkFae?S&R(N^b4)Vwi;vIIh?*dPbIqp$&}Z{r!dkrU*ABy!+gVS5~&9n(n0(XyGCgK
z-G>F|o+9;uJ!-rE`a#=KjwBh&_kqscLhI5&e=axMo9GZ?kykzc3?pg>Mo(d<KaDe+
zo?#vFOpsC|{=DRnwuz9QvidaZ!OB3}2$PReU-F52bEZAfHg4=#<7tdkU!XVBS0b<w
zCv9`XxM{rNjmMhICp%PP?m9S&*h1<;TOTfN3&o^iOrJ$G8C-H#5HFb7K)f!(aj0c1
z|Je7{unI^O*_<sJoBc!?<c`H_db#K@Cm^~peu1bpQLvI=NM?%0+*BrV<8?*@s)r(s
znx2QP7#gHy;=7!KSh1)<=Y|{rsif3gW3lHwY_qYWZ8dsxHGn;%qIgXEw12VjxM_lr
zif7-=e6<WWUVC1SL05G}Rk;qi@d<gY9w)#4iGBG@PmZ=}M$*EP2D;1n%d8I_4P7Sx
z)dd~7pslza6e<ou1j+XaUQ8`3rl<A|%?>UoK3&|oKdj&%9#-`OOxfzP-GO43Yw(DH
zmg&{KJ$pVhyp{V{D$_Xq!;?^2nMOc4gO)+4H-j4I^tlQ}rFNM3GdFC<jH>9rs`}J+
zJ2aGEc@62@_#>vSq;+NtM^6Qyie-8jef(h7A~vbCOK&$q5M{$cdPbZ40wiXF9h2nD
z-~(VPuTmX)+)W>wYG619S_fp+J_~1ew1ec%CUYDJC7()$T5R<msef$*0cc;}+ya^%
zP~tdsYLPm9qD467tQlWp5?Q0=X5bu?mv~8w)p5n5u%W8n5jb5VBT-Y_N({VdIZ?rL
z!8h&H4K|hnd6|#;a4^kJ<h<ZNQ14Sf_qDf^xg4HAZb*2Tl#)^b|3(lws=!9N&xf%m
zPjDZpwj>^Hm}tf^Q@l9A%-S*z2#0)TE=-+XN)`E?u7X^|2Ju}deBcvT9C-zRJ4?((
zUoS{O^tm^aDPDjT9uI*+5y8US+6I=aF=ne?r2-EGpKNae!$2I0qK9sOP)(j{>+#oN
z_kdE`lKc-3xsZl90=$8`+5r46K0zH7#vlH7Xv{pIuJ+`8mK%7>DoA=6lrtt$$aW<_
zb+PBIXSY~S%XmkqRLY#2rOmTo0bFH~RS#t-VUWM(yPFjo+r51!Ii4LtM1LvFOjmJG
zBWT?Ul=ivi!1dNKBu~WWILyHxDRZ;pbr=K}c+g&AFzuJ60>&|neikqYJ8^bpyS;r@
zP(IR%f4{Ew>)xuwLIK%UTe&194@vxEN(FDxt3)FuMEC(ip?T4%uU!L>TfSks?ErnP
zBlCDyC<NCq$tXNo&})QzOhcuw?Wd32oEWiOj=G|B1f3+ghHHs@%tB7=SgMe?Qr%sx
zl|UCcY<UmKasxv|JlP&fM7)_ZhDV};Ksq+Tg1r=ngk~^BRGg*7F*ae}WIDbR{h0Qg
zlX7;2K|*5|a45SS;&#|xLDyrCy&PVhRfbXsH8v5l-1wkKW3ItofuPwc%W1Y?%wB;6
zq|Gc#N$iW8Nyb23(@HqP`G7dQA?W~R(67xFswo7WFOTaW3@bGz5x7`;06uOVV;PZ{
zwNpz}&m;pmitsQQ*uVGw(GRc5E$y(#o})MEY^s8Kxgq6ttVBFdumW?XStTiNheMjN
zpAhWiwUNL4Ea0UKWftj5-gNWM`=!h=JbhV4v^19xLrA#$->bH8h|2Hl!i;)Gxh^sO
zw&iCDOijQZ70}r(BS({7u?+H%^UD?=G5!g5x!l@6TohkK=<AJrY)gO;is#W#snpp}
z%8eN#CsV#H3dE`Z@n&KA-!}^*1M7cjXVQ@T-P(xwTe+VW1`o|r_h4`aCO5_VYoR`$
zBX5KT*DJs}#DN7^-CsPQWL^?*aC37<SP_A)b5C6xaH~DlN04xA!a2A#w|8uD{hKqn
zB~-h&XzRuTrz8F*3+FcllO@h7*?T^?eJ)NK-|Of%E?UG)9l7!A+_Gi?@#23C_^sSG
zJOt6_(C1!7)Hg|JgLimhw~Rc%-gg5IF>36!_aQr9T6b)WP*oyMU&(gfOKok~c&$n2
zl7IQ&ed=^#^S65kA-d=d{cr5#B!uaECqFQ*nJZsa!K$vZA&|g=;wM$s(8XDGtkBcu
z`Q5qDASBe{=6Pj(wm?Ma=u`V^{Q%1*17&wICd}}$r8nuYBJ}zrtW7&nEh%gVCo(ZD
z2%rj5KVM$FPDHjk(kAj5D7^4V^h95k^^ZdOdJJY;m8?{x!<%o6?o?5VV2pEJTs)An
z-+Z{K(h}5yto1lk*7uBRzAaXmF$LJwNz#p2aolmy#&t{%M<+y*JmY^)($`?%4w~7!
zA^Rs}_d=)_7@Fk5s^LA3$6jyMBg@#>*2Il^6jA$8c6Z4flhyPGnY^)srJC5(cPG9I
zGA6>8uaQF`X0*!aQ%uAWgSaGBxToT@kHa1kLnOALWi$p`$Np-IZ+Lxuf^K8g3yr6n
zaKQg1eGPeE!$T{C*~4`tUWUMMmb`Z+1w@>QXD0~*Y9e(9tA<?avf2g$7nF2vrPiGN
z6Y1G2M#}`Lqqb)L_+u<04;As<Ci37CCJ($7UW#h0$OvitSLCN%&E+0H${|}+34Va5
zS<?tIHv-!Af~FQ4uz4vx#q{vG`-4VPM?GL<`fXl?;^`7^;uBj%!&TwCr_BWVYk5}P
z5;ZF9xr!t*tNYYv>~QLyQ5ejYSFfr0d+ncoG(AIhD{(JWxuU7p-*%I(K(}H~ESqNU
z=Y0do>D*p87+HpCYzQ19u`d-sHPm7B=~y33KeO5pV9QlECuo;f>s@_#ehe^YdaE!(
z+l<T;-tC>QhXG%4zg@HvJ`a*zeg<f~5m7%3ceg=7ALxFA#5$lZ3@jMQaJ2>tLaG_O
z5STXbz%AFGIUqO-x2C2KIz=8dXlhnz9T?+h^A1KeH}1ti;XLlZqA6k^-aG2_6g+<2
zve)?9_jWJHikr}>tvYvZbw-QNfi3s5V`zvH_)H7K6at(e$aN#bU7%2sEF7!FNkho>
zzd%Uv3B9NBpt4n@(i39E))@Ut#vzPK*I%y)SDeD4o`>GDYT#4l&iBIOq&QO@KVN%>
zq@7@MmFGM%jU+fA`lLY=ww(Sa67HNSPPF>txv8~G%%>CCEp(10dX72uGi>ZhkS7U1
zi}_2^kO(?($zkG4S2`;zmQ#VkkgQeovbn=bCJYj0<)n?+<sT8CDQFJ1DG6EHu@sr}
z6k8cH$agBJNF82F<vT*W#a<LNoaitMVH;!ttjFqB53Et<CLInsnF|vPsUfud_AxVV
zCDpi&KY?>>BTQD&YK+^sY9&fdTFvwNtlE^V3iZ<A2VGyy2XQ>R1}aPJQ|m=@4X6~>
zJrx#ym17?6%j4@-XKi(u_ISAFJz57+6g?CA^X^K|{~Ew_X|DT%YEgjPg;|V|MYYiE
zjpM8yqiRY$_B(3WFcaU$qX=R{=N7g5*wC&%Ml-^ny-#E2z6WM~l3fq1n{nv+RMxsM
zM(eTA=$w*gS!)^46iAarlT?>flQ$i3%nO7u^?Iam));hQZxd0n?tTS~Ae)N#f)vk~
zBB8x;{*_7J!jVow@DIw9lcI(UmeOayWpPWz7wN106s<%XW9k0;<$8#_!AV3p^uw6G
z66;S+q!@v*Uh{JYPi8Al*y*u^(c?B^SGGPYc)8M`ny{>f1#x{hYy&vuawHH75NRi2
z4`2vv`uk&L3qhRT`DK>rWTBNa%HwO|YG}@gOZQ2t-t%>q>{B+QLv{-ov#H^oDzlWg
zGN6j31)4*x)WOGsC<i<P3Qbumf_W+jjHiX^Lp5&s+NB)=w7tRxWV4EfIOCIH64y!V
z7YYItyDk}$Qr7Z{_UY*H%EaX+J4&HTAbd@Nv?$Jvj<_WtjrN&XJnI54v%e6FH`PL1
zbGT<uw+y+yV>8BJMbInIgyG_@D!-uRoIO=Q?ep*pgc!&53M-U};($&cWO9qpJms3k
zk)vp;HR8tR<D*f09Q#o31~;49h)5{u+)CA8;ULkszwTBr@935dx~0LdmEM#nL4pjk
z5bdj_g!)N06gf?@!UAJJ12|WT(c0BIIkJnT1^C2)vy^KRs+Fl%jxXhIrjd_>v~GkG
z1WJ}2C!a573e{q5Ns9fQBdiq`=jDoaMH>dqVWdsO=LVHsyh|bu{AOG7<HjaBP}Ld{
zAN3@6&q5QFkXPl!UY(n@5fXBA+XF)>1|0rB&h5d3`V%?dU*S0k$T|wGZYPgPP0Rgp
zSjeU~<JIbX*T}$SvvFNOpwus9RRZ!gniagXEyG5Qmn03FCGtU2q-D;3Fi!I{5W$UP
z!HtBJ@{2J+Jj_58v{w0a$F}ljSq)>wiU{?VTZk*Yw6JKn^6K)r=H3|i-FC)a4AP+t
z<eu(n*u<M*0tCobzlJ!1f#1Ta*LV;OEY6eR7vi!&7+{Vl!_3J^!^B%Vk?Q<6r(BSj
z>+Vo?Hr*bMo(-@+o!ccVs%B+g`e_6rR0tm}pU{N}2EMNCU<;UuD-nwGHTaMX`T{t>
z<S`SL!sg{@2p|bw&VM=#G2|>F8gGk8wZO<&hW;{_Q!M+7OWXgWP8QLy<tK;-7*75L
zVFgK56HxdP0<dQ!PZUp;6THzTQsdo&wAH;ift|zR%V&rB^mw%zdmXELEIs`Sd$f|G
z{$u{a`X43_j0`OQm3H<2nLNDwP99MC9o2`}x@?S0E_Lw9Y8gt|dDtT-iF<9}IYWtm
zc<@LR2-iFwt2(hb{X`Il47wwgt5UWk$vR&;Lrb?d3g@$tw&M=2&)SK+;$CGD1DBm*
z#MFeh7EZ29T@@OJp8O;w%aLhAO=CMbE>YYk@h9=45OJPmkVKG0kdN#7wD%3TI+<Oh
zZc%TyJup<=2AyN6uzh^q7i35c9=wa>F!r*qs}}w^5kuk?mFELv-LxkP9NdqWU9)*{
z;C$V28||5^pss^oX$P9`lgvxH!75mc4*AJ`v2$I#LgO~$9b-%`*toI5p%hDI=sgRZ
zrOe<4=IaMT-Hwij!?(CMg}bD?4R-GN(;YA>w|Qa4zObKs^r=*3Eu9shnBZWI_NHr|
zw)HzX;;gLJ79YUiYU1#sz7Mta$nH!P=<MJ|=O!~!>06yv&G(<i1GwVH)Eg#Q4e!FE
zG<l}pIufXE9B3|@3=68Tvk_RZM<$IZuSsc{U``&;^H{7KcOV{vNE7NHF3##*e+^=@
z4-e>scqEuU!oAld=vw}e)`3%Q_I_JRwYL0#ek5P|MvmM-ejM*<&LsX6tQuz_U0%{S
zsp@CESY|r8WIdw4IN;1DKcP8eD14Rx86!;-@PvMnyy(g`GW92zLv%M~bR$5Y?7)NJ
zk-bE!9t!dj*U#Q;H$(~V3%#RS)Y5xTS=2&wr>w6G^}SGdG|R7QZ3E}!>7AmI%ISCT
zElzj6klNK`us{uzkQ~ojuu_i()1<gFm~;G+DXZ&bn$->k5%tuhd67A0Pi4)TD$4^6
zXqt-D-_|0>WGP5SZN_i@i02`L6O9Lw8dtTpZQ#4-xjj$B3hyT(vJcF%Ml)=#7F5ww
z)V#edQlQj{7=YBc_|IM8%^%l<yeyk|K#w^)C5YN;Dr>L{jeQWN7Us*~78*z7afj6i
zbL84SO2+(7U&PDigM>mcw{K$WM@dr|Z3=f_)NPD>^=L5E>*Q1Jo_pLszRh1Q!eg3P
z1Q49@cyIQ>q{E)=<jiRCp?{uRuianrwVT6>8o1(4uT*}SecAJx`f|!<qII~u=Z*~O
znhBfb-TfCGXS%kzOJe4!eNZ>bk{~!}imz($YSx-z!t0ln<_2b(jhy-6$7lXKnqh6l
zC)9!7+}7JrCZ4p`Vv|qd5gb=4WJbj=C}8QW9n}@`R8;T)|279zvJ<<I&@Jm91yaVa
zVuf#?6r1Q-a~qCsLA1t-4fD>a4#7RSE<fNADZ1(AJ3qKOD}<4VheK=~!0QYfR*ihI
znkz_AsZ$$rT16+c{diW-+oei^dOmJ4`QyM39PvfM;irm$_t-pLuCF=x|FS`u%iC0d
zDwr5OJaUjAufXqA5%It6Y22G-6V3G7m+|XtS7|8rKGRtfJz2yuc&Qyk85bq$61kH`
zmRBTuC4}UF?5P1Ut9^HkV^T@6RP&3wb5(_NqrO#L1+1Qs_zfn%DYt<A6GeW4LQEr@
zlpqiUx?kv6st?xEFj=08!6{+2aTDdOm0S;GR;eyhOd2lP5fNjLrKcRE`B{PO`RcjA
z!M;-gvF-SA`l5G$D-@i(4%3>v|C2DoW4zV<wKDO5$NMetNAD*_F`kWYZB?gpW}tHJ
z&zz|I3q`kP)zmDg0#=Z;lIp6ufR(gSul=rJ>%yvYvjJ_1cSNBCd+e?Xz=+k5B|R@8
zc?cuSSA;wB!G{v8Nd+k~w=5CH9ZgMw-T~Dvlag1o>RV3fpYxlo>-G(AF$EUZbF}oH
zs7#~HjaxP{<z1200)+;5@|>V{fQj?;USO=|zievunZxK(n%t|B;IL!cVHQU0vB**<
z1k}l@qk?bSF?8Vh`p85JCHT-`tb`ialZgpNU}rKjc)VnFXaL=G*8_~63jFXY>r`;T
z3E`GcYiIp^E*UwdCyN9)ns6w>H_{l2+whV%7N~&ZaN7I|GwVM4R|D<)ehSck&RW>l
zxx3*5!{;QU4b(>S7{OBZIKCw?kSEMYONQo8f^6ChN`iV<V`F4T-t#Qk5itJcqyn6*
z#Xe1T<=~n=wyUL;rI}GR)-$_4d6{SpAP><sdzEDO^qM+|ysFK!+=-v_qiA=zk2^fC
zo@9>sVQG_1jB2RmF&<^NoL)YrTk-zb^VD96dI1{4`pZm&a-d~QIu8ev`l4%{=4x|O
zT{(OhWNJQqbx|}RPnS2WF*~Z%tFGrVH%(iZP?QHHU5^`!2Ntv0<xYz!(IPu=G^j(@
z1@aIURO0zBr`c?^S=8y4+cys#xJOsgaGDs(;j~}4ni6Ra&n3fj^xboE)?+_#I!D^*
zM4uglY8(Db;jevgJ0|#U|4MUnBio|Ua9}B^Z6jHM_b5);3}HW>-g1;c9&#M|{==b!
zz90+OllPC098o>Aw(XzYxo{}<r^7-cTx%-Kzj4b$nQE_^r9ga#GR~)H4|s&i!ZXo$
zs_yy~aaXxXV+|CB;ersGaBWNFQ~W;#_7mG6nkZs=*w2S4CBZHL=exV7K{*#qluAlA
z(BY<YB}!7yvSOXimHi(*o9!(GHK_;i!A+a94nDjCQ_J4YyoTked$k)AbsFT$x~t>l
zelwRwN$_v<=~rH0g8H_OEdKNDPf@E3pqhEAO^^nS{P`8MXP>&q&;z$BM;Bojae%$D
z)btN}Sgt>6yy~gMNyQ?lrp(Uh3_BP#0BkZVhJ7^_F08OCYv2}Pnzm;`C-~ZmPYrX=
z0I)Vzw&xzy&GdeteU{V5sA_Y##_`qoJzp+`So*;$EHAm+mlT^{&V?lU!La!OAA-Y&
zrUYF0d_hEe{c^aQUv)7+6eY^~%fZLeOt?+}!OjP^{A3eN4*>X-1?2eOf2NE&KB?jB
z`7q@Ht6dA<@!%*PAjuTP9EpR~$u!@~<*Dzi4GpJMRIzKn@F*w9>HkyKL(lyGaqRs6
z6Kd!={$D~3{eS1CUDC9U#cD<PzWyIVO%Yz#1cB+OJ_7ovWHzQ9{D6yp%Wz*C+M0?-
ziq_U&yN`<xXjfrz4vM+0vmh<MP5T#<G1E|{nJzo&?Rr}+!B?D&IIRAg)v{Jjm$M-e
z0qNWNs3Hcy#8?<In!-h5+-vhfiqv>%;m1*22p9xMcT8S$H!&-DS=OjFEQfUbeFpmh
z2?#+9qdP_Kbk$p?5ji9618o^&v1qul4GRWV9W8YlQfT)$WoTzLLS`P=L<(F*N~T>~
zo)3M-ouIufZ3k&^C3@mAKPa<Qe0&WRNjZ&zg2F**Ib2+%GKX3HfRwQDehPkPjW~V~
z>VvIKJ8awqJxrQd{wQa7IWsMi4U2Qk>1kka-r#Xf_(c+*-pcZLKH>6;=-WK!sc13E
zn-py#bNJE4FbDq}Z=~??eh%h2Cf)}3_V9cz5t)op5Li}}fd(4jp}{iNz;tmb_cZ9(
zBlh3g+@T$R;F%OJghY>)0M$rcpMztqlb?F%E?6s(o_1jIRFCpdPqS!XYtxKegUlhC
z5iQP*28E`<4Q<JQL@0kQ3ryXH(Y@JqU|>yB9i}!}u;d=K2JrqKotlYE_4o;Q05A{^
z{Wt*7B#~uRQ*M^@#NZ#WX+TH>U?N9UVDpBDiMq`y20lIAfj)*KlDIIB9)y~fytN<k
zWj+9NHNQCszyaA?gXN%+Ar8Kt4>)R92?u-p8rd3;9JQuo->2LczHLmJv6`yLL|y$>
zpRcixAUbHefmjuYraCw3O7Z@)IH<d(wlc2wm-EXP)&T&RiTm2uLGDjK)Syb?8URb<
zOlm5bY>pM0BzzzZKlbf2LNHDf239kob;r$d>cdc_Z0SiN=_-))%YM@hS)}?w=)^yt
zGher*+~>rQ5G>h~M%+j)mhQC7HA>m3@u`N1;Pnp!!SUd7wMIO;LE3&dzykZ#fxSPN
zljg`wvhz|0x_oR!N9km;9cYnnZ<XRdC7yR*3)fGFlDs_+7e7830Pi0V+h)tMCSS_=
zz54yBiIs0BLvM!`Mv9#;2Zsh8?^mA_Se;GQq4ha>As{);FtP&^UQC%&Bke|hx2IFR
zrzKJTbE!nth4<HFD5~j2XFb_;w4{*@d7Nk8%iI?|10YGxe}Yc9kDsc0OVK}y!ib1_
zrtAQRzwo~g<zLPp)NPS?a;LwyycJ7%Z%QB3Un)L6CQP4yRPo-w7nXE8uD_aoZr`f7
zbA2y;zCZ644kE|swqi@aLUVbZX7Y-LIITQC9*kXGICw9n5$LCP0~zY24z;;FOV6^U
zHrI%UBOQGyPj3E5y%lt7WV!CpmZ05`sg_(;`q=-?e%e}dyEx~~%@>mTC_Op5bG-8n
z8DD-bJ2f~EZlMYZiR5P}sJ73Kt$(_873xi?2eB(<RpdGAF{l5kybgW4aDVas?PSEj
zGx0G@`!mAJpp#!i5-MO(%k=M2Plg2T(Cb^FGP~deQWC<o!Oje(MNhzn0&?(Csc{9(
zMj#-V68T%=BVrAB+<JoQ?un*A7SsjF8AAwNStm7;)*m8}+F%&D!f`e9L1Z%5C5pP(
z&qS{gjyt!6?YUSXJYZ^y9f6l)6S#9k)Wn?HbP<$GqZXueAz)l5;DXzFrBU<9e4L$C
zDTs0bIMb8=*<J|(kL9mJb&DwmAUHye|I8SXLtKW3_lr;H8**57;=9%ZO_V&~ZiS?}
zN7R5Hw3ffL#U#gDPu6#X<8gkJ`!inT>&bAlqm!MhBV&4#dc5}rA8)*wcB&b9IN>ec
zcJXAgp|FbB{?6mR8L}!?VZdNbCx3%urFxz1%SO(Y{O{0uvP)(dt9=KPX(uhpS@OK3
zRced5Cf@#0vO_xA4QXlayf#0kEv6{O;!hf7d#S|rnGHnV(PLRs%5{XQmGT#miz;$f
zY%UzeY&JKmS)_mrdGn}GZ`rPMeFIPPwC$Z*r~!36N;U&~`09D*f;)ip;3*Zj2v0>B
zx2xZ4HTC5(I<SbpL6-qv3zR}$c7z;V#ko@B#Cg!3UTCCcAEHxSs_d@(CG2&xIhFtR
zh|A(>M!9j_JR=osj0XCn{c~Y}Sb4J#==k??2D6T_rA5z4$R_b1mgZqIey%1g9LcDz
z)O$ZN&&9#XYVQC8NFQcz4#^8;wtR^5;6=J_FYQ|ybZG<z-Qgx6BCixAT{2`qqiwZj
z8`R*JT=2rOrl2*jzp7<5O9v28{loIolnemq8sEJ>)5P3PHD+=Pb4A+}<_iMNnnfJ1
zL}A2u@J7PB6|pcDP1Tgp`7;f*!=47NQS{NqA^)<K_7T6v18TK<<Z=P?$|m{$3A5SB
zF@Eqw`6%MDDIya!sN}sCYSH@{%wB_D+kcQa6!>{Z2O%-NVHpXM%3?0`0UkVuc1&R1
zoDsW{t|MT?$RH^B0mm2QS+Ys!@?7PGz*og~tpFu8#BeNr=2pT-{vh=AQY$z`7vX-&
z)1dkh^wPO)G$V#8FvK0(c1C!6yklf-)W65PU5o6Yh{Sn;^wA?WB#z+VvxnmZnH&8|
zM5VT=<F4p!V*j`+T0uiX>~Z0?o@x_a|9YV_M=-I-FfIW1<)S6zKAzI}%wyAb<Gp_<
zNnf7~qqV5E{_h|XsX-v3BMQM0rWGxZX^N@9R^}-9d+&^V9(EyiPL7Lz4Q+#hM2&H0
zbEblD8Q5^xX44GJ!0h`e1OM+x#Z@IpEK^-IB5YnlW`+JW3BindS<VjnME1@v0-YZi
z*!R?v3>;L_lQcwGvH*Q;Xn-Q%OK#BP7Lkd*J>}5j)fi@4-l$TC`;YxJ82j2Cfx4s%
z2w^74>lVJ&ZN=@jo~|gq@t7-APka*%9uKvrjE+X=`kOxR&e^4bhtK*9nysw=1seSK
z#@pr94{79EgC_Q=9YT6_BwUBC2XrcJ4==4VHNW16=0}wYTG&SsET_;MpE3eJQ$rAu
z<0&HWKp=5YHvY5X;5V+}nmdOl4mYWg%~l|XwWj-I=fqiQ<=MyDY8gsVEkA5pk40T9
zn#!4F$oqZ$!V1FRyoif`o2)4Ty_LohmXWEQrbf|AWhmf40+WkAdsF4=38uUAXV@if
zJq?eYEJd)61{77X9l=-j@@j{Wb+o*N`U@fqiQ=XxBh<cNPppTz>xPx9vQuawAEf4Z
zw%vq7saHQ|3g>bO9u$<WB<lZ(6gkQNiQOBW2>e8ws{uWK-eE)mJ+QJG5apJgd+kKk
z6X7xYsBL+0|2MtIAZL9#s+{9{|H1mi41~c3+;xzB9`}r^!13my;htf``;qd{lyH;=
z-(Ui_6jq1?E>q>+lmq5hen>gt-oWrJg2mlx<U&ZmC+ST2v`fFbT=8xE%82!f8Utam
zJO*#K`h5!@Wl6Z7N^}ILT3of&{%?Hnsdq49=P>=2uFv`X5NNzsYzLtZoPm+1;oJO{
zL;-8E;B^_u<e&C&`{V=45vq1fvh$UmK^7UYXe5P$pvC2sMV6u(0%w?V%f(udavvJ8
z1i1BA3H$1)d?TDtqE-$O_~xBPAYcI}fd<%+SnRf}j!}M2-DhjPB_ZIIA}=H9DE!Mz
z#E(;RH1LlX1gjq(!2*2X&ZI#+DWtOC5CNaXAfq@Qy!Iz${;2BO-yd(`y`_Kv{?1Wb
z!CXHm6uSYXZBPdawD#lz?V0m$O93uGh`J~l7d}P4__8|e2k6iM)$%{()(mXy|66Or
z!N&Ywb`-~Ku-V|dFR0$hiQuOw8`|(A{L<@3_-yQ*BpoF7v#juWjvf?^61Rvwv*-Tc
zXe?;fw~VZ7G<=jMZk1*p987&Q&S#^<^cuXp76bX<ztM^Hn3pvI^pl_B{&V@aB^e$P
zON947D4II2>(^A=Yh-pGpjVjPP5?Mp6`&n?FVQ<ru}`WuKAzUIZH_st64&9ey8iH3
zL3Z44x6d~bQvIFWu_=iNiMU@}wniP7`h0J48q(eQU>1w6<%pJ3Ttc32g>@I!YZou4
zns#m!fOsH@jCpd~AH2|Swp&OQh{{U1gKoEI?r53iFo$g1fcOIZP=Pe}*1QA(Osimm
zL}cZ&-oeG^nO#BSaefxze)X;Tz~Cp+YL}!%be}%@j3Aem{FL!V)W1|KxfgamG7)40
zXGZLQYb>d|XCpLd|JKY;P{d^o$UshoDi(FzP`^E1pBQVycg0Q1I*r-E#09aG>Vhh7
zYpu8YM|uN5mtjw)q&5TAx&*B3E^4QdoPf+>^;r=~t;Sig4So<u;eu@Vt1}YIZGcSj
zfRQrxSjr$tX@$QaP(i5(k|o(ZdAV#-QFZMs+an-1FR?jcdX;v|&3A>|P|7HFOG!Zg
zeWQ9~C}O)>u%~48leI^EhtWeclBGBd>3*u;krcUhY$Vb5)8NnDn2OoL+VAV&5FjPp
zs!WqkIOs3XxyX-^6ZJa+Ph#_z&HT<%Q5D3uW@KKQ0${q`ucM5{9iUXiO{5X@))#Qj
zFJG-N1elP$a2-gNnlHY>t~s#95EaZ~jU##$s|{sR#3#-tmB<(M8#4PxK(vW=lPT~g
zyBPeDLGiU5xe2qv#i4YN!5T|q&!ebxy#Rx^?p!bdzxQl8^uCV2m|e_3l0}1Yyg>6q
z<ib#^eJJjEFgV#3s?sdcp>?CEUAjX<;zDE)#<jpW5@24_*DXVyAW{1vvm=_yzdtHR
zd*Z8WCiszk?yxDRs2X88Erk-ew<itTh}Okc3aoTbJaWS5CI9}~x~bZqk$h1oDf+KG
zMhJlDX9#5z!68L3uFJc2d+?(^{KTW|!jirfZUNL)9<lPZGaUSu1GRt}Bmd#v!^V3L
z{b<{Ieyx9Awf}VB04ad81VFSLT{#hX=p-`6NKREh$!b*%W?>(Goiy&%nFDsZ`p{L3
zMS0f5T9kDK_6fD{Fvd5FjzR96Np2O%n^l5b7rv6G%lg|CYGV1sk_Oa?b~+lEDz;*K
z;rm5PNDQ`u))hywmWJ@$HbrPbM{rO2wqcK2IiG?UEPBv~QdQDK66eBVGQ{-A(~|U2
z6aLcri5)A3XudN6&cH+)5U8pn5z6N{Z*E3qY%JT4O#ak1{bt|K+Y>3*Y;SS=lNJON
zT*t?^>&ww`s>@QzebSK;QGp01l1B_8OUMTN$V6qE&GXw}?}|q2S@iI{c+LkTqLVLf
z#f%s3jiP<-u!5x7TD0%JO8K=HE@1TZoeF)mgPqQ0>8feJ<H?5HWT3Gb7hSx#{g3ID
z5OEq>s3tJ&KeGMH%3NC1DGiaF4XR9_JvnKk1!r&i(c~J{=Pj<Z)_v``_4!C3S5-yb
ziH1_W5(2kL;v^*-q4iBn><q6hvO>n2Xkm$eHPno)IW1;bi6Ga-LCbaQFoxCKoX~>Z
z99Qg;=Noi49$1W@Ct~11DlibA)AwWzATlADytKEJ4L{2q_92g1id2Xu!08B63hKz9
z+S0)n4IEK|s}YAQ&TH&OYa*G{D+2#cq5*C&MU$1k65Pl+o;4lA6XE6HS&7@uh-*ZF
zM<OM&f=1UOUgnKu1d4+|Ji`;^93j>!ny!f#b=YevOxvqmk084Nib3gc?n~%Kd@mNH
z5;n0~pkSm2|9CcFUoHh%RQ0K<YfLqZN>)w&oruHx<eDJRngaW$l{n&hL`>XKpJ)-N
z&Rc2@WCk8g8Z3a3o3Xie((dr%qXn~w5b-AYbC~o#UGjf1_KvZ#gx%Winzps3ZQHhO
z+qP}n_L{bB+qP|gbMNHrd^zVm$?iPqs-%DPpQ^6A$1}!tJ<H0LT>&&6C_BJ-Q3(39
z4NIlBGV?3^0o1*$fb}4rU2Tb%!qm@sRR)Vj%K1#zb+$NHYO%RtW6q}%Nsr_byt&rw
z@8;SLMmYQ)oL$;ce8vy{0EN1a#w$iW4&PCbBN!R4h*yv=l`+9puo#DmQ$Tu#Sy*}q
zUS814{_~dPE+99n@ey(YQ8q%s344SD{U91kB`)83LJOd_PRKowl{^lu>`T;U@R^5_
zYMCWw>Daya3J`}6Si|fC-m_fsFTC>8zKXG6?&PcPDC(?iKLa7?q>Fvd))^3?oBLn5
zaE>G392|;9dWs0VPTmueyN(%fd<Z0`v15oQFlS`W3KqXxOwQ>JvfIjo$*H6?;q$lo
z&?o$ta|w#)x9<pe@Cqyy<bFQCb{d%7!->0PDl#T_o_v<IHT5pE)<00ATBW)|)DH!j
zRwanWt#n7;e2c=Z?<RV{-8gFvnhQoqXmdg42vwvA5S`6w$}<`zadn$GN}HA)A1_z1
zVXG;u)#B5~1RE$qu>DLknqK}_l5*rGwpDGh!>%3+jv*3Ot+}R>Kv&nx=nY&I>9R(|
zk`y~2luZYR-EL@<2r)m~Njh|KDls$m8inX6+Y3Yo3x4&B$AxwjF7SyOnW|Y8HZA@y
zILNPz|ES`kXa7$jfcbyx&lror{0#v!zoY4dbW`iCcrku+gA8$+QEkNRL-sPT@VTFF
z<n?O`tez`k2dkkt<{JM@sB3gk)ayHXTwPwSuW#L6-o-i}NWk4TxqUYMee|Aa)w<4%
zP5`0^SGj%Gzn+IRoIJ1}4J^YS@85Y&D=u{rfWJUbS@~`NFLm8Py97Oc!J}mF7Ec1+
zHxpfk@vaoOTD4~oa2{J1&Z|FnuM^qS<IOdsFyd0TosK!t=kiQkYG>kIJRF<CNZ1_L
zYa<41uQ`vAv9yV&7lki{1b`vsh;LgF)Cx#Y;eP>7CB4&iNB!$mK%w|M;-Vzo4_X!A
zB1eVdx>!}Tj{32*cvw^+k$L}M^XW3n5qL`RPu)D_ynbc<@^|otqj?084Y{9+c_!bX
z$1sxU&Km5&@YoONmSN$UHSD(OR94oi^k$Li@RjslZx1F^cQ*mwXW0iPHPIeb=FnR5
z;JO%5%}&RJfeq|LC~Rr);0x2l^!@l!0rOTMjHwZyx-7qEZjdOp2q41?`{IM@Z|_kE
zx?lB|L&xnwDY4OJ#dOv{%n_p@7wepa81@%^|5dB<qMFLBpwIuE<bO**+A(=KExml$
zI(~z!Ydb;ui@W92et|NvRJ{fP{AMzwANL%NetfW)<o`JuTJY#CdIymN%(T4ZaRM^G
zu0&&nlhwegoS*BNtS86X*l!P>=xo>@YQ1$Tt{K$;>X&yX2gRIYo+5pSVrL3-fg)!1
zqK<;T?af=5At)Ggt}H{$nykEvaXn6sbc~`WjtBG&tR)`Lopk})tsYW_NEGzWAGxg@
zNWG*01735gQz{$kk-Xlew7J^fC2agI<6!@t#YrC5AHR+JQrdp5;*Q<4E9wALa{^Lh
z!t$=Cwss}Zz__mA)Fc&$(X)yuYUq+mRqJ<SJ-|F&?Et1(rD?wl>-U8*jY2A7`98*V
z#31$P0lDNI9xXd~0XFCvv5xo_!2%2XITdeP?CPUeiZ<!P;cgJ`CES%C-xcTeF;6+x
zg4mAmrcwbt`cgk!vJE7ivEHsK>F3O~w_UNi5mL}H&qf_pubuC|yo|~b>bw}*2fyN?
z!&xFGQENXp-1uu4F&HP#-eAh!rL%mTSXZ!cq@iK+cAhgR+&d+`;UrfI3u312dhE#8
zPi4ypGPa$GP$#)uca}I1G&H=f_+*zUMNSov->g>hnmb4hS-;_|oRtP(6Xc1!MStIi
z@`C)xR=e9%B0@5#7R-5zp9ptDaIcSS?<*Q#dKq!r1n5S_*emK4Jh&Q81R2nnMkIPX
z(tg_680qj0jr+q=rrE;FyHyPLcaW_rNJs)Q;-|+2_r|9xSA#zC3J2z>v35HAkuu%p
zp)sSr`C#%SMi0J15-1snLHm>jVzI4-0>V628g8DXge9)yAvY_mi!&RTNGK52;$x3Z
z`$;1<pXXw8DW<=Cp`lt8Vi3JRje54fY<ZY`Oz2;edU8r8vtY$m+O?mmd5DA*a@}&}
zVwAtiggJZ6Q!#2ZAf~_@iC@<!UlGwCg^eU!E2x7cjjZpR`r_7Ssg^46w?+H~m3ms%
zo}p)}QE$<`k3_2$@u<=(aUY3gyrk0GQCPjUcp{L7U`biuj(J+18QhwPzcuO$NMk3q
zXP4EUEG*pxu^CRQXg>@+2H+x?k=>*H{x*x*wT0u9s@${2bia1+d`y#A{Dz(n0jQI<
zg6!D$Vja^xF4abz_=>BW>=PNYvP^U8TupfYZOqsSzce<2Mm=EG_xMUOeie;x#p&3S
zSZ9bP4Pn-^S$j-z%qyWCOggc3W+*rVyDP%2vM*2UnkCV8Xgz`A13)}<#6~?rJatJ(
z&C2ZfF&bM>6aI4D1)?zJ){~{fGPY{K=Y>v2AzQ>~cEc+l#TSEN4_9(1*P@krJJmx#
z)awXZsjmRCEPJD&wO|1VK~)R=bF?NoAZS#Lv^KJ6xAH~J-{Wf^KT2Z-YB<wFW9OV{
z8JiTCD|dM?#a?5DJ=a+S1elv8*HW3bGI2mVtQnL6J;jZHmdGXWjQK|hJWG259)@P=
zSUWo*0w$@=MF4WtENW|M4M9whO&bYBW`tvkL_$9>8!Wu<RzecRAl}JJD$x_(1IZ&}
zqCHde_Z!kp5W>ypR7q5MAH-w!Rlz$8xBahyQ>;%CuJ89YgX^Qr$HWd$t$>CNHqDEh
zF*PAgnUbIG7vCBGg_0>Fk5hdd_8C9rB?nGOoQnibSR~i*wpMoeWcD2eoACbZ)+;z2
z?II>i(T!e#rIsB<1NTt!)L41!%OMe<C%47T{C>q-=kJ}It6IU8K0}Myv&h((?78qr
zC@PoG^+na#Gw(PfP_bmo1f^YjM>!?n$Rgz4PlXH)<N1m^BOpKhGA~<`E(;R2<088~
zgCna#WW7*RnK();gs#Tw(&THy+#6-y0QK~VI487QF+<Cr`Igk2ZLhV{lgiN&F6mnE
z@|bVP!S4!SQEVC#ld;vFpG41C!{j^&?*w%N${`K?LQsOvS2b-t98+{yV?8Rd6js0w
zqEWTs2#1sXSD>98@5J|7SVa+(QhRv@>Pt`OZ;rK&nz6vgKcxrUA4Nm&|Mb6~f&Tyd
z$Y)^wUq14uFr;HQo1eOS1V+{d7!C<>d0ke3Caqkc9~X-}$>2}q*qZhl$!GrvM|{6`
zi^e04#gnboMPoIt$3MPy^NO=~;vU|<AO8YSri3&5AY#Y+Ztrd69tGxgiO=$f`pAhZ
zRR$(#2jF5A0*WuBQsWgY+p)fG&SFM5tGK-LPoacmlR3883s1vKNHe+tiaLs*PJL5z
zv=KXr-E6xKQF2Z>?!I0RCUR|z2v~MVA(<g>CiSurghUH&sp<mWj}7#}3Djy!_Gj|w
zwv}f?f3_LO{C5e!EfAt=`0)a%4C^$^tZK7tZ{%=zJ}+ikTqQyWcDBC{QjA35wo99`
zG&P`-(D3SXE~16jivRoxa>@_!_~0w6;{IF~ErZC1R0UTK&|vv>aOJ^8YZT!Z7ZgMy
z$Mz2Izm|`g2sYFEIPwVWvKk?WOo=Vm-o2T363Sd?9oOumk@Q4cUc5rB{zVXR8+=TT
zD&cSl@|i>7gK8X%$H(`d!7`U>?Qf;18Lh~deM}R&Q}oEKt7Ov}E)dsWmQiyHW;Ep3
z2FMsmPsDh!PcQ)U)xhVeTkxRh!7avS4mzu#+jghcQPcq+kI9gncUk)PP*X&i;@w$s
zw-$zmtC>-|r+DUAV6N4Y#HR(8FqKyM=AeyKVw=ePcj7{JRiHUOfLF@lVI=2wfAWj3
zxRl-Qv@c?}aRg%!A~c1sh>vx6ejv0O4ZUL2+SwAO@RQonPU7B!!hPV=Bqf>LncJc`
z&N_ExjD#Pu7;(?{?EW7f_yw7mLSy~-awp9QI_zN#v4(;B65P-*q#hOVXonA(2yyQ+
z{Lf`WfXFYiOJzBZ&A$Xe4YK*g%kjYa{fTb-dx3PYaF)LczMcy2CO9$LZ`)N`286n@
z#@sh6;2>FC3|_B94}9+w4-5b{$i+#6MGXnagGG{1FcWrs_EO<b!<*#`Tc|VF9TKb)
zXrx~*7JBaeMz^PE=;^W4&1gFXD%{V71$Y1=|2!r6*n;%@?97)CI#+R`N}Rl&Mzx!p
z!FH1R?&<@zbj*HMmHQc~owNyVr3FN!QII#{wS+1dsX*Dv4rE?}G!o|BGf-KA67^b2
zWSUso@=4+F$QVX_Hi~%JapGHqOIth^7Np-gC#*LkoemU5bL!v}J9WIh^xvX(64Tc#
zDB8$t2`_MP55z#}>y%ZCiy*CRY4&c+p7MD{Ka`lj%Mpy3Z$MQe{xOUDb441pVbk*s
zm>m#-6VM_9ZLp3B_pvQ<6G8^*e|K9)w~+pXLFIM5SJ-T<#rY_qhd$#habJi!l$4cy
z7>D_>Aj9|hfb!ZYOltUb_?O&G6k%XHr89rj?F4msxy=uPZTRzM5YgnvD{C92(wG;$
z*Dwx!$T5GC?(K9hIR5Gf>Xdj8<B18@`(qU8@(+y)zqK@v_#}#EL&%+r{3~i)@)V%3
zOL`fMTwXj0<(4Si*2S~T8@vzMmkGI#l#_v|31>Nn@;=Xj_5C%i7SSa&#s#GIkwhra
zcY+Hz*zoHCN#<$5SjLGno9MlfX_s2b#>aSQ6qT&UQ9PFoCtAp@2y8&os8EHu4=fiW
zLMp*5#d;q4B}v!O`{%Ouqd_iqi#%9Bd)41?twY945bsX6Dg<0cQ4){-&xwQ9SqEsb
zk7zkAc`qehhgRTxCW}!Pazu%Q*#I*a$TKQ1%Y=HN6HNUchEXGxyP$T^#P*|p?wG{5
z?Q<d3@2(ZqZXs23ZvwQ+6~ii^O=%Y@Jm*xC->P{P$e@fNHC}u3I<BmJaHg<sbm_1{
z{6=b>+vL|aTR>!SIp&D4Ihkrw;~8eAgp78S)%<O!XUAIwSM|8I3C{79_R-@L7ak>f
z+n_)j7My?v^&XW+x)wAV_KC<(Sv(@~a{c-#Ad^8BGt+){<W};g#GD@jdJ&}j)kIZJ
z0@NG1-CQ?guzLh(AinP|GA6#yMi}m)V@T_q65X4(SQ+N?q(rzeC5GHSEOkmWqKrAk
zrgijjH1NU;K7$V$J&K$y>$rJB!y@yj$Y&mx6}f^k9_Hm=l+q6raHmArH7Ml*wT($Q
zZ~jDX%$vfgvRNX7oPKJ5xh}~oSCD&&r@31baXRM7fv>Nx0a@=hc}8;Q!fc=Ds`4Yn
zxe_$!#vUj=vX@>a>;-m=VjJd}DzbXC!YBL746U__?R%+omW5^rWO>n4`@6O@-|S(w
z$uu`#n4Dp6=I@-n5%ekfqqm!KJQHK4V&g1(*1^F)=_c=rzcQYHE99zWmUFe9ysvYS
ztq8q&5p9?$pb?vPfjq7?z~X!fAxY&n(y2t_g`+W@P~Gsji_#reu?F~y9?*%#k+F|7
ziyqiJ_r=2GXuWsPa_2k-ib?dhGQAu`)DGh;<me8D34(0vG?H83t#5v*1`-k#5I115
zW-mJf+XW`~oI*QmxVMH?35ElVh2*N!=rLxlDNXbE7Dq!$N`*qJL2JRlFUAibvz$wt
zhgWO*t-w3&I`pOMmzs9BW$a*IiELd#SCkUV?n|u!8O&O?oa9}O#UPJF3<1be3#UID
z)B2qF`odF6+#12u`MH`?IShD1fiU*Bb+An4RyjfmhJBE_9m+Q4-AQ|vWu3o#6cu;r
z(^Lg_T`aqFLK<B~SF`p``IbZCBiXU&e=SZeo$T=CT3!!d7ZotDc@nFQ8<^6{&1nGh
z?4-M11^?2RkK!C7<jp%kG7iWP1j!Kdav#2tN4Cmt72^{Y-Lfn?(zAP}HY(=qK3k2>
z<60IjudyU0<2$7Z6BmY#b<!-O?y!ayy!r)=9idW;CtFsgYEDu!eRa)c+$mnjS-ZZD
ze&>>eL8Yl-{J_Vdd%nHhTpzSMERbl8+;gZkY8z2A9U)RY9C*w4P~n!`n>EUEspeXP
zv!S(gM096DXri!=dFN^?NQ3zUO@?d0l~b1pKwA?Z&Lf(H8A>f~3e5`e1|F_cNSdT#
zp_Dv3h~;noZhBE!XF!8gu8U*dIN%3Qq%<p;XwN14xMrEHF|=5i^Jw;alIQL~cMLJU
zp+4Qg+;@6oVY7?cvCllxVQ*+;Y_u5+v&bwZB*+>g(e+hhX**_$uCP$L6|d3VBFXxv
z%yP>5tKiM%W|t`DF-Hf_Cn`}4y|iq0p)I<pFj?V%C<BnWuM}jIbW^&OC74=T-&$q;
zNe<b!ov`-4eeig%Sm2+r5<c@QUAmmS|Eo03xw$Vb8k7^e&NGdhIdTVmuxK!PZ)Ij#
zYvtiGQa&c+z@gfD3@VisK5<l7jIaJ)`~;FpBC5ESW=W#$B+^WTCm}}apqPsZ(Z6o$
zies!i4+;J{RHDX3pTikggSlm~+2d$3lu4?`V{fDkaFeCBM5k+}r%c%PZ(nt(X|?i*
z6UFrhO#S1M^FKz(Uv<I%8>mdr_P+!hZvEnLetibrH#KR7s6!M4<mdiW&~Axre74dr
z%g>8U!mtptS@#c;=0$g654ByE9pZ2ehOwapZ+G|!8e3mmXV+U(0|!$R*y+>cGrP5(
zpR4|#c}KKn?q+R{f5s7=VskrouA4+CR|2&}ED74Oyq+CSrm+1}S^2jAxI@;qIf5QS
zj3Fhs#vN533A|n2H)xl%2;5Yy*axpTK)~fq;l}droKjulp(hw%a$H0qehV{fTUZB;
zk6KJ(4t~6{SKT_C=&Oh>Lwa?<he$R4f#x)3YElvjGTDHtjO!Yap(IQtExE2Jqw{|M
z-l{iSrm&lJK;Mmo_xFPB>KDN?@lT?s*y%A8_}+S)DG0i22+wuJY@JbCgFGT`n&Utj
zm64N$o+)ABsEJ~hRp#_CR^rs2!<+31z0WSJo~_6po2Oju8MgH_ZGI<Xi)xk#g}~Z>
z?7FGVZL+Y}M!m4WS#-qg{mU6x-n_;lQ_$b~L1n=7ucET~lqVR$J=ten)K(#b5uC@@
zwd+1!aFAmki;z*oEo!~bTPxn0SqNUueTP97nwT!~fY=z!`PZuauqRWWhLrjzAxOx*
z32bU<!!(?IVBBrRF2l8ZEp2>S1*_)t*-pzZU)4y1y8$zowpXm54Qu&MTD)kDycSq7
z`W+xU;))Gib&ztk`q;^OgL#wIr~`-t6<fjzW-KVzgcFX!t@W2qFQeL`J@uu~thrLx
zq#e;y>AL8spUSbGqM&T%2%{K-E=_Izv6W}0T<Otug^5iGwK?i;lj*5KmU+wZx9;6I
z@ns)cY%Ma^(x%+j8GNPIb?D5K#f|#3n5r8V?D|<*wzB}I+?l?j-JCcr4vjsms>Pv&
z-;nm`lJ=%l8i|k+E#xzis2AFW1~QYo+ZtRwa?Dye11@<}uo}7ShbMJPYRshBH|4fW
zVsiU%LtY=;`-mg0Tj?t+6{D_v`{~GC1F<0q<cp}ODsqS1oL1Yzw*{@wY9S&16XK~3
zUruKObo`3uxp&t=u$P3vZL4sNF!@VqJ;uzTt;@2z%e(W<)WE^8OW>-Vmv?6nHwyV=
z8D#B6>Cb}qSH*|hgJg$AqS>RHPMnsm=7g3jmP(xyw+IHof%&Gr)F1Bzd=Jd&b%QDR
z53~6Wp1P#61e)%RA8~mlZggQd4}W;;3am@i4%E8z%LjeHk%Xvcqx2$H#c64w0C#&M
zQTR=?ss;?@(#Yd1w;Wy-7V>bed7);lZRx}Mk6Nv;4tG<YMM_L_RfvAqgSHFmW<DvQ
zs@X9!j+3bRiBqo`iCXHy_JoFDRT(aN&QkgT8=^CpMs{1Ub$e3W5PmrgV)0#3P45O*
z+|x&xzs0E-F{zx9+VB$v%#O*UVXs3i8#n@MO~+(uN_5=?y&Fq&xx{OF+SI!Ng+@dP
z4OWN-O#?<362LQIA$TZ+%3__b+IrOQyIk>Hi%mSbI$PWRO|(3n8o_|#I0j{6zFpeM
zvua0qDDWzKf&iT<H{DS;%X9FnFds;Bwe)<oB875trpk7QZLs8ZG(wQSBp=eZ#_2(I
zW6Vv4{heiZcmLS@U;k@d*q&jc7_$fe`3$`LpCL8L7C%k8dg6nG_+xE<CNl8jpfbET
z2gL<?`MNY9)hpQ)lvZ=5--!KNhc#akA`&ZH)G*u!n&acYhtEQU5h<)M1qqzf*(?C4
zV)|236zZNGwWuL>4ZOn=1ldRZ`b-6BZaP=>f>%>u9ffxKS!o2tR;&0|;|BeT%-EEo
z>c9pYgEqp!I#TumP;AD;J2NE)>o5^+{OpU&raQZx@PL%@v+(%kkdOT%aET~-HVY0z
zK~?NyAfp?VT@5DAex8Ur73d-`JmdX#4QE)p3ZLU4?hh430-bfMR0uHX+It!P=&rIu
z>iNkVY3Y=JXigT35*i1)b?fH~X({!2Dm%M(!TJg>yEQO3%btZwVPR|}c2+^E^;ick
z<sFLv??W#Nu`CE$PMVwLIU$m*pi}<fh8;V_UA1a;DeSSlf^XJ}jt`x#WBYTu#=w_=
zsRfwRp!ISM3_$n!p|%Ff5gnJ|eayVE*z+d<=?&%Cb<~fGfdBkt)9{&sp>vpa`5tBT
zQgMMi=$-l26UO)#`*85^u~pLJT=XTSqGQrJwb*RSdXB9{Odkd5^};xCh2kgi;|v!I
z>%O|s$T*4mjChExamVv2*R}cRwM9;w4urheL$pZQ&yboHVkM+W8%`#!bh3b+l((A)
zB#Br#F!NHXNn=~@?69D82AvizX9PCNsm7S#f*YK{1!}4zEKYop@*UUEb=g?qYc8N!
z94}K?3qAJ|5g6#5Og}v8oQ}m`flsh0(6&39=~moMnC_|vPB={Hp^<~>8q4+L-05+S
zj`oJjiD`Z7^%0qf#w#9Y;cjU}D%Z929}wycKi~iO#AW`!8lt6)ZA_ia@EDlc|94Fr
zR~nkX#Tyi#9bNj^4zd({xSNZyKOIOl{@ujep@7o4=j&#(z-E4c{J!42+e8V3@)(9!
zpuzwX>+w!*G+eGYKCe!@p{}g5w5&T^lFqf0=K5#1_jh==0SCT)ceAH~C;{l?u(Z6m
z-(H7V(AA_`5su{c>YrD3E2W;~Sl#fi0=wn#rJfbQ{{<9&4fzEMqwU#d=5m|50lArC
z!BJ+}1BA+z_H=&Ug_s)jxG&qpD<nu^`qHV;&nl87_m<Sw-)C9sExtzkYK5Bi{_FJz
zMLEyfi43z=#sP_Bk+S2~hoIyFkt+3^KmU81MP|3P&m+8MwL3#ONNS=XoAeR~E=%X1
z`kE~JQmdkt$O|;SYN&&qOSw_W@lzizK>eUyWC~Hk^UQP>Ba5HQAZ@;-0cejcu_-b>
zZCqSmeZSB`<7?Rej>MA7K8xbZF)?-I&ERx9iOKM9o3uFGzVqM9*>g(=Gyz<QJ7u;F
zKUKr}A0hfgrmyE4S3+Ys5jq>=8ffuP61v;8A?IrKg`+7d3T0d8^(2W#Llx3wr3i>4
zq`jNaO33?3H86Msk&6^nOSC$%DJyuh-(J4}VcGz)0WyPfKAvqD_ql+0bNp|56Oaz!
zgcNmkL4$Oa?B4CoEhs}9-2SPOjmHb=_Ag7RumvfPI)Tx8zbwNPSHO2BCH6gH#_Hm)
zzg<>-?Q#(Puwi3p3B2kQJb17CiB>FRl?1xw#rxe}q!BgZ-&E^WQ{NLwHCJ_I7HWiC
z5I!<z?$HJ#zpT)hDo%_+Gqf>vGu46ptDaC!AsxDzy9ne4A@@)pjPcA)!yFHO5r6S+
zgS~WjWibDR(Pkp(A&rH5f)FS1vaZn7bMQ!`S8qfd3004YPTQ58`&_QFlIuZEm%ISR
z{gM-c;ds7_jl{(fam<JuW%{xs##@vp2SOILsW6R;R_qFY2Sl80nsCmbMyX+9{Um`4
z<XuTJFC7+D09_L^#K5uRl(q8(VDff&-p3uWNaQon5o+8I47AH&n$d!cC99=aVqNi>
zL08OXNj_7!b&Zf%S9ID*>~9&v8>W`*E1C|~`#!-kRH&`3#HQb$U;5Fws|?r#xWEYH
z5cFcrI95Zdr1?-#A!zl4S>9^Sex9u@%PQNgG@hX1xI(75gIl6x!+NlMdP)3HM2p~J
z<NhcEL~~W3`&#5^-a|I7ZRfN2Zud5+rwGAa@DjO4P@^n<;Td9~(pW*NA}+!pn<@l_
zsGS1^ayNG{9Xu6~-vI;jvdJ`9PC3Ij1fNpAHeoW#w8XqAwvr+#IbVzg6Lz!yZgTLc
z(d9^t253O`7raLQAwty!>d{8#bNE9-Bg}eiE^J7?=(6#t?kf9L3~J4MYj(85&s2l$
zS!bFwyg7mBFC1ui<?w5z-BOqrkm96j;RkC6FE1;u`P$u7%ie6CV$iiSRg1X)IWg{G
zV1Wtpfr46@XY80^fZr+ttONAQ&<@$l`-cr|eOxaw3TiMwl~*f&-VZ2nMIR}yO`MfI
zr&>;#-&nWL;aoLm<c!j>*aW&+Ud=oz<6On>^3M{cUQydJF26knQwcH{io8zkJ)B)T
zrtOGbh3Ozco#9BS;rxo$Pyo7KbYVe0LxbIw*-^0qxv)+x{HkD~t89FHwqHW-j`G)k
zp*GukUoqYHfE3?gs?#0qtnQw<E@w!&TgPGPw$lulH2a5NzjU%%LXlwNnJm-8zwIBv
zv16*thDIFH28HuWz1ccqBlwq=Gy94IStK8|X39=1C6SV^&P2}Vlt6DDg}+XpIOmB*
zmj+a4KDVi6+!QcaC7~=30TG+Rjl%(=dcF%-FjN1-wqlwwc&o~dX55%!t1vwOB!Rd&
z-(VVz5{8P48V(K)PbWGUIys6(XiV!1<~HJnp*Q6P%+g92dVm2dO8IN|AHS?65MBAg
zv-@QG1>Jb#kP??PDj(w1(qefKRh_1J#a!l#Qlx>WHc@%~^MhFPFRQ2`Q$uPVX>O7p
z5=TldEYDeXY}X(|r1{d*eF@)TlDP1)<Q!xM1ZpW+C8xZ$RNNYeb^V17tB514Vw+12
zq7;Tl+J@WM&+v=uMPp$;$eiglSdO4feXH47k2xUa7zlbfPEoL=zpYxetGfKZk_cST
zBJ?7Cxp~+Lr$^+N)WIgtH3(y@Kw~c4v_^+>loD_ur_E*FsdT$9fy{Z~@?K-OWWk$@
za5NCP)j?EATAi~7GD)`?4RZ4A$gw4yBXO)uk=vBTF}34(s22rUFx#3T7ImdjJKfwd
z7dAQ=D@xrXgS|hCXfRS%ncZ8V7uF6$AvcJj`(@;x!E8d|pN63n$wKa&xdrQqdF;hx
zHWZ|o<2r6j|GashD-)Oq^L=!4ng}VbMZ9w~r>(YlW+H&+Wk&+LOubPwsdQTV8Y`0m
zk<d71NPFX3O1(ouHE0(*u%A_3PKXA=2U%iwl(o$zZvF`_il1xJaNTH|k+>Ij_is0)
z(Y!IRN6Rm5_NqFQ88IW;ZXvy|Z$z0%-$G&TgAWjB93A9PCG10~yzY-4w>jkjI5#u4
zX4<tQkpEU)209{<GG?`z*B?olv~$`{q3`ivo+HJ@nROF6-N(-}WTVPBe5fg^yqL-M
z>ZzQrPUl^CRkq7z98jUI(5D;(kW+~`aqpxuh)kw5euY*Y`r#P!{V*-{9$q=)TktXS
zL8I<D+6clB2#@_a`_I)MPtQlkxv8EJ<H>=3J>OnPI%ELMqMbwrIQY9OJ=i?kUOP4G
z?!Uwf1>vq^lIcH-chhvyX_8{=@*&U%Tc+Y1*7F6mrir^VV~ofPd^v7uKanM~T;)=K
zL?)Bwxq<fGnZJVtLH1LrYUa$UOUQ=-PiV~$BOimRp~>6YK?FBib6AYiVSK!Ct@U-p
z=RmC41$1YPj~C5`T2q)^D6x@H-E$RnmAOaA(-rI&IQqxUJ(#}rJumAi|7Pm>@Mz}V
zkB>y8k3hH>$IaNl-dmG|2(WKXS&Vf#B;+E6V7xzz&NAS5D9BL}YLw~TFD}>~%+Ue`
zkrjCc9c*(w$3tv0urm)VgTeN%8B?at$o20L0gHnSAgA6VxxNTFeT3f0=MQg(%lyR+
zxrY)DeDP8Cxj3q1>P8b`OcZ{AA1$ip@_lfM0Fo$o&M^^DPL-3&akU5)6|Sxy-Q=gS
zoM^L0h(57@76Z7MURQ|g|B{qFg<R|r`SulfdEv>|yjhrr(Dvt}S!70P&t~{cC2<-U
z3+}SuTM5l`UGlPtHr#QufvL0PXaQYi&FUB&c4KK7zOyK7i!_2Jb+lx7hB;P&zi@10
z!GijhjXhGqy3QJh5lnyom%phJp=(CcVq&x-W+S;=A$zzCe1$n`P1IqD(_;Lanh%zv
zel>AjNspGW2LzIj`xDdq(`3&v`^`dExr0}nD7Lyb-#`k&=w|;ZaMH8<U%Q3>{}Xuh
zEdMis$NFoP|L+&Ah`E)Mu>-A$mA;d)kg=hykueVsq@$CAvA#8=+xn8GzAaV@!q>Gf
zf1d6<++G4BQ79E4kc(C+0ODe;c3u@e_`La%7QqgBsp9VA%1iv_QFg(GQu8%zvc6=4
ztJeLM9or2|TeFe2l8E}J%s7MT1g3Bw+60A6xmwZZEp>&NL0y(a_2$Lgy4V-TZmIoH
z(q<vo7>SKt{&P@n->MbpuOVLgF-{7dDKGSJ92{V$SP3JHjDW_3Nci&H+V9*7{cl&I
zD^D!R8-Sf4OI3jxlvtm2`8H@7*DgwFnm2Reey>sp9UWyoIBpm<91%bX?H7r^sa3z3
zkr=z+(b^!TY3PO;J4^A-X_6yWL^x>oJ3i=9(K~*GgZ*DNic|q~YLcVyNZbVfm!U(`
zK-SKhFOo^|vM&-9ogyw1hjCIx4oMx5fuZuWp<b3uBj`!44La&cr43LNE%V3`xxP{^
z+gfFK!eZ(HsVA0FfdUH;Vv;liP~v7MkygPpE&4*l6H7Sp@Yo}k$uMwx`YM4I{8*F3
zh!hD;20BNg-Gw}ZZZwiY{wE^#(gN2efO0ZoJ@9;Z!(Z#OQwD==7<;mnwPX>ZcseU|
ztR0y(05iTzM0hi^OIUzJ9W1n5`7kU3D1(T^$Ql?4uiG#v>0g?Q{6X?g1$spt%$SEb
zcaU_lCT<Tt(I%LX=4F(fqme))i63aJW#ylfIqR??(>CJ$N}uVqYQ?xhkUGuv*G*nj
zS)7$rGKxNtBl@dZS&{x`4yKo53oT^RceUFj*>ga=T+#4i;<ruBgnbtyu!&|W#&HNg
z_IA=8uZ6C!E5jXbO{$cO>#3<X4&5(0-OqydlZhqa9uLzvk4f>Dmg|q}4cBI!&TG>P
z*S|tYsFg3`gu#a>We*fj_~&{@MV=+IUzyh(M;pd30^Y>*LJ}$3(0yQ`y<xsS5xj7K
z`q>L^9ye-=VeE5efoZonPWRuipj7&XaSL8_2?O}W+<JbQ&N-C`z?e0T?-_vpOn(so
z;t2u$hzSBm2juAzr{5ff_N?X)fBppX>qJk7!)+sc;^(;Jt#Z>9h^^T}hBeN_O5S>#
zWOL!E<dM#8=Ww+>VQn=me^_zr()>Bat?8Nh>3%ow`gZ$Dt@h;A{c&=vHl2XmnqkG|
z@$sJfS^sit{yw|GUD=^s{yM+6TE57wX@96Va%W}a$<_=)MO3??9e6WNlHAFB;0cOE
z&wg7Y2CGNgda_vN)}5l=ep%USnHYNVaAwi1?0DEKy!amYbn4>b&ZYg;)aQTje&Ol<
z>;iJ#8}Z^xAT|L66BwJfVke=|>F$Z)j|a5XVZx7{0^_6s>!-yPBJfTJzOWN*Z)kvI
zng#Y+#o!eVVXwgvhRnfr&|{et!}|Prvh)`Lh&8JX0hF~Qyn_Xo(?{;tZMMQ{Obg;>
zGZ5P*1b<ejA3+9(-~n|2^@1&tvO3~#13Z)8nCum-_N!iqbB-la|18y|mQ9WHkwUU!
zG2b8RsC0yYb=&%S(ESP7Q~(TaZf$GlSla^pUPQ2*C)AhhR8>5M%+9-Ikw&r8p;oo!
z*>^`d!Bl!Vw}OGALSFhEzSQu2G4XD`POif3)S_AbduXh}PLjbeK0z><w>K`lOe6;Q
z1jv@>YTkPL7H}b)VnDyP2iQdne&PT?VPfXLhcoG!PzMy}Y!d0=e>wxL!AZ7*3@n@L
zQlm1lmg)0V?fJBC{L=i>$g>vFu4;Po!Nq^!Q`b$i-|h0+-t)^~F!_3tLpYrAxuiWd
zok`7=ume4n7k3K4QE6pm4eXwrlEBl2BbM^NJ=CT)sGqTbVN{#vLE50QKxM6X^t!Ko
zvWRcpyWxWv0}PiG!$hX`FN_(P(wwzX>nK06g(HjO8<F0H8I0>C<;RK}&C_!RZzi_8
z2{iHC{Ny&!BT@*h*9Qy~91di033I0{Tg!rg8{eVn;9bEK{1!=up1Q#m{|?#exlvWW
zIa2D*9^0bndtvWHg^$7P!7ImCz=_mFWyPJ|OIdTcv+QXHt(~Jr_7$}w(98;A@&oRU
zJhEY~SUdtWuRG(Bu5!Cl<N%)7B52l3a{c`TVFUHZn}xO*yECafD)nAxIV5#3mI?`T
zM`)#TswH22zH+s4*u30c!y+pMqqLYs#KMs1$Ihjp&jPjp&>)=u2d}9!B4Elr<=UtL
zoFr*?H~cUPe?%Wwocr(}wi9X}87MSn0I2J6bi3Bp&XHOjRX~1lAJ~4>4$x3Sm`j`p
zA$VwUoVhO~a2EV37(V?N$MbZNw_Cxq2+gVCcqT>V!};?p4*BNRD!+JK@5Hp80Gs;3
z<u}>#E}C@_C}4k(n^CzueWP|ZIrK-#s+ses@xE9c{kiBst*NSNDcZxt1dnN*NVN-Y
z2f3gf;wS(F)&uNT;5~rH=#5&@aIhYt3ivWOBVpFSw*rWJrGlyEzdSuA1Z-C2D7;}M
zdXW9Ra=^~~z5Qk}xnBLNAj7GV+I&Fl>)uPl=06A)X5j}WkqCpQ;o|U~GHA<$+I}0E
z_m5&|^gP2}_U-2<sU67f2od7nE87z5`5b+vZ;eMd($vdH{P$RV&fAijGLZ{2%GtBd
zu1cmn<sSq~;x2_85A24@zHe#vjb04`4Mpm8>V?abU{Qs}FQxeCZ!?}dN>GJNU70YB
z;xwp#>fWW7X?%P?HptBy)=osHYW%N8KQ@3u^j(b*2k=|ohzanArk+_Ip&GDUJ&l0r
zZDOVTTXk8!1w?_D--6oM1>)+can}%l!kTAz0s1khpzQT%!$Ch$#9sM3g=k6@ZGPX)
z#INIT=ij{SBN~4uyzaWQA;xi(t#XpF5SJ}Q*3l}+x0R)(XoQgx)qsjm1b3&hj4%j2
zYp&8ZrYJU&6)-98XRTPhb?iQ(_IyNv`;H!MuCvg%qG+<)BvY|^evcakl9JJ<820NT
zS<MKfO)iV;@|L2RF421B2vp1g#Uizv;3?J(>wh<gcUeGN=v$-v>fs!!Q55-tBv+NB
z0EDaUpIcEDzzr(XO8Br!4E`1Q{hi+zRD|${UOtGoOa?WGgGj`Kbq`CB8?ntM0TD|U
z67``j@c)ZV53pZ*Vv*2j1tAJol~BN2PYMcus#5C34Y03s1LSq~Mqnx9<5)=&b#xZv
zRn6jOc4C2q2TP9C1tf}OolPsyfzI;`U<g7FIYJKai56p5K~tj;Q<G@#GjUx9KC<S;
z+>vhE%;O$g5RaeVCvuer-0aeL5d=7SL{Iycyn_ciK7MKYkjSVmW~MZ4)%fI~MS;s7
zrNr}l$z|pmk1z+b;3CIyx?(L035G~dXN_jUM_)dx8SSegs2oT-hd~^ij&{HTfiF_7
zAN~N)0qpzO0{V(Xi>rVgmZ=yjES3S{Nd8mXrdlpFy;!LDzVtZJ7bo8i0XjePp1DH3
z%4QJoM_=&HI9A1T<i7iPm07#0w^RJ|b_qZE+CrRpNKJO*u}OBTMv?jNWm=8$U#t7S
zSI%=Q97g&0SUfQz|J-bR{k#|=&v~nJ+B7}^Bi0+8H}IU80pKA11-HcP61mM4B-Eob
z$8+`Vp8{(rSwFVJX1aa{Pe!J-&L5}4so98pYe-qW)fWWpx1z7*a33AyVP1+9j+>g4
z$d?sD@wc7CBv;=?gg4#HkZ9-sl#v-&|AXI+ne~4MqWEQV{rcSyfBkNcb0T^nRNW&o
zK-M#lf1nQID&e4v9L9|TfW`c>73+I^dJ4$GjU<z-*!r*zn{jg0sFn{Z3J&*D1GzeT
zJlxvf-z%s3BK49>2Np^L`(eU+B=!p$z-p50G_Z4gct7oEiP^E>*rJj&4sKt0cUN$1
zn;_i;_8@{QB(QC}<Xa2ug%8l{Uz6Fue?GjfvhHACf9ayET>HI$$8mhU-dB<*`+dA!
zoWck=97>Xs0QtRq?UuoL!9BiuFhv0K;beqr0;>>o(!l*ti!S|MpIQ(M(7?iV;#imP
zoE#;$b$!SAZc_a+o=z<CutE!rgOH|IA?;Ui1C!!(RKb_BV@ODgs*2x_j+Zq&vETdA
z5ha%Uzt(Zbf%7qepGpn6bC@O;)KkXM^Q$W0ZEb9l9S34@XMSs;rC&qKbKq!C1)~#y
zf6eny|H1!Uujk1l4wqsOUcw5HKtk4P!T|Z7BOcCl7YpY6wUgA$zwI#W+TH7~f1Z%%
z>9rPBQWJ6~4hXRthA=kc`zoc)Ozn50!r(XM5iAL|pMWoG3!WV&lUI8#L1dtgX0P>0
zD&gSztIq@oY0kG<T@0ayQDIqZEsVUp7@WZhtxb_Hwp8X*8+Zd@Yi$VFV+~AASqb|@
zw*$Q=^Q;X99y3I2XDURF&nYMYv4*ZFF*=>6AdpfhvBL~FDJDP)sSYNwb$}sezMd8s
z94&_C(k9YdNNqI0T!1KYhVk2d<d(#2TPQr7&s|NQP6aM31nM0#CJ_pbkZW3hM7L(N
zM3YGE&nev;bjh@aSifK-IVzJ#jjx@kjqe-gwLBE~6eiyndl);G13=G+XQhHqoE6(L
zOh42>+8x0t&Bhoz;2n8V#jW50STu9XUOOBuTgpk%b2q5Uba3?vYxs)IhBMN#!KQmu
z`uY>ZcJk6{$f`7OPK{{8-g42yf(7CN{-+^;M3(vjj!Jd)RP&4Z-DC8=fdz+TD_z}m
zR(^%A7i<-xBE76zua?|UfKm;SaGKy!@u^j-bMkDuqp<m9fgoyv6MgE@%ZIXuI9>3R
zCQZL<L*p3c3v*Cb*2X+TMLMwpxX`Ll<%ILI5(`8cP{$3dGx~UbwB71=>vciX?5{E<
zZ#C^k(@k|%e-X@@MHKl{HJh~@K|o-$fjs-trBo?Z24;1>%0;MMEC&2BG%U)n+omge
zm6)24H<Wg<=0=cM1(}rH6WNCaRh7Tl;Rac~lOL!c%vxDJbeacc?t0YwAx7Wxk~-Fy
zGT~0hH(#GOC-J!g5;idbc%@Y4AS>cw=y!krII=7R6qbQtjJH~d86AYGV7L62zFdI4
zxTgN~cBi}JOPy|5{^oOp1+kT~fnCY6QVRt=@{?PK{xR}EHC90XC`0-Fx<ZCF3ZR6b
z<P;e$QF^olGz%Dk5r0ADhH7}<D*Orpi@c8ZV!?#d0y}~m2XRol311H0-zE4a`A&BJ
zdUOzx+d5wO7FQ`NIm{r=fI?P`RlH4kC_+`H+F?RKNC|Dh_aIG4>MX-PCOmuGJ=m~9
zrOYYNJrxcJCG-;!Q7I%(lMxPcYNbpj1^;){ET!S9sD7)Fe<y5l&9kf$+ZMIn_ALRI
zksT;Trkm8>PP&`gk8?hoRd}2SHo@*9n^j4*wAafbEfqBMYQ=Gq<~9-|g|rtj@meG#
z&uMP1vM#}Fk#8!tUEhVYVL&SmRaij98EH9~ZMw%HJV@p#uuD?vAAT?<N*)ozCLRPE
z`dpOgmfZymD@g|p<sN@;6bhcc^zap9D>6QGGnd*N)mTeS=WFHR@FOiPX&zA`{z*?e
z8_)Z9q%kv^>fW8U$zU{<nq)irTkObir2mt9YSzW=pG%Vp@FwNZP#@5sjNMXSiknIy
zHwirrY6ZAQxys?1iX>*vVhpxyA(gcMzNI#Tl(9!gRHWVD<FVeXrYdgLKTuTUY=l8L
z7MF~pBMLxd04qJx2j5o+EDv0eIOp^3h1H?ZHxwa(_5qmK%5+Vy7S`763!%j4PdBm*
zTa;gdYF7wm>)ev#mRQITW;h3y#YX60n3?Ku95Z&XW@9nU^Y71Zb!J>T<9FVPokvM?
z7_sy4b+Ng3{9;cYqX?DaWy<(F(D6I4dWmS^Z3K5LJy?$2H>-%h>hOpk7_fS(Qp@Bv
ziqbPJuulIjWpA+QlT}h~&A@X^$P?P*`G&VKHI71zuEVmE#~$ds5KX+TV!_vz9Up>X
z1jLq$3@YkJj>97uH&L<UD&z=0U%OzEWImGOYP=-5+#oN-+ho~n$*q2}NH)Qrb=`6`
zD**ty(rKnRP=;P)BhOyglM|6G*BSRjRu$nq*=BD9v&P^&e)9Hhpm|+(T|si1vOMOF
z(9|*OfpNQ3GDS?T@FQ6|wS#fnxNygoNK&+A6+F<Wz2P`;X9G?mbFB~>KdD2zNZ&NV
z0$B0bFOku4Br3eut@iJ?)IdT%tShuq;Jg)Et5v1(a|E)gQO&eGI4knaBICr^jfjZN
z-h`;DJA~;-RV}K)GBND~6y1(}@?1OfL`L+!8TCERpUs5DWixYUJ%?4Yx4mo9wA>Q#
z+Ug>PpYa`DObOPWl|12OQVBN2i!C==r~)`qb>wx$4BOdkFNpq`oXqS#LQH7{4Mf?n
z-nEI5tY_2wJ$0cS{YXsv+jFwhwD0Xa>mJkWby}aGlbeY=3-Qls%yW#4zR0e#CaFlP
z^Wp0wvWq@}K6!9zTTRi3YWU}KFCVh3H;VRz#W*e$EB{^x8|7ol5Sj6S<e+hAsiawd
z?3eROv36y!V1h%Uu4&T>-BDUz&UnP}qv6ipmC)1Z#jnGn5ss%_IKJR+Ax|${ckFYT
zP{$HE#Ww~+qvP<|X*<2mt^M_4FqW)5ChGhC6=9p#D(gukL_p#A`uVW8u;hBBnaM{u
zaifrUCV5g>#U6)*$`u9C;p)K<8`<P`CV({}ePJBa*bLvmI`#msCQ>7yZJ+i-MEkEA
zd0SA8j9^7ZCR=q4u_eeLjM_k1<i#qT&Uo)DSj<8l%LrI#1IpgTed$5?0dJG$GC3O4
zvxwGc!s%L|+H3o+!`N`!;3guemfnXxWux?2u5P&<#p9EMU?}Pfd$U}20tve(Sxno5
zf=6O^EQ`cSoqw>L3A??EVIkcy9y4OIOJ4t$bj!ZUx1nA&E_^hq>wEJGhSD-PwjJ2Z
zZ<qBG0rgv+^{a+YPENS}kE<(VfJmLT`0L2l?c7T4SD%5@wi4X(Pp`imT_Oek;ubib
z3?zlhKVNfEb8DQ#0PrWoy7|Q^Y6NZHrZ;Tc%EQd=Bu#WGjr<~Y`StY`eX*qNWr?!x
z^g{Pk!+W6gTz57L^;&3RORIJt3)g^zP~<_@o1EW96h}vD_ak%IKrS03xe?u3zaJp!
z7}eGP7%bWTuj1?fv)j+k_`id`E@?`}Znh%)SFp4PjoeF6_VO5pnT)#n1K=R;1zww(
zSto)j4hsut^l<SIU$97|Nw5~)!1t%cQ$wLH=ks4g(c#&Ic#hA*@zFx&4R1XU`(Ctz
z)BEFs_+H51&u_G3CM`{>t(3|A_ITeDm#Ba3+$}6N8XnoRjkR_3(L1^e>PLc9P3GWb
ztFjBW559v&_mcDi_wDeyGAwDyds&`f>!;+QwJ&<e<@tJ!{>_#@bFbRs;lXu#M%k!w
zrpZZt(uPLZMNGS9x0pXg;VZq$f<QpzkCxpvh%JJim0D3K5*g8wSbM|{9K3OJvhg$N
z1<TAgj`dT1zBu^WbsKILUB9!M!6fGiIgk^sSd=m3yaD6wrCc01E7cfJ_Y>5J7}Glz
zZOymLvQ<(gbBcG8Ddy-cjeHoPsMd)_H5XzRtsGV}@1If82ez96%e~0&=~k?{Uq)xh
zLbXQDXQb9*Vw)6SPA@YbQvIZq0fx2~h@@F{cys+U@jfpnA3Ahq6+up$paTXzm?&=3
z#~#$mj2GFF<xXekL*&8cAOCFd%bVmlr~-cSEB+PWRX*V1>qnP6JKLl}VeS?1OVZq^
z(3qqOY4V3l^4~9btD+SGuVw7hdMocz8?37yg@q%Du!4Ga`pPYDcSY;kysi6i)9C)3
zLhVQI=D`B5vBtghj}7zbw&RxqLS0y{<sy!Y{Hut=bK|pC6WhBq8}bVabV+ZZ*JBPQ
zplX7?aI_f80GXNZbVl#Fd~fvB5M+))s1eFqIA$2;h@HCdXyNTx2s(9-?-X=Wrr7?f
zxWI>$01nqI3ML%Kd2h3{h13DIfJuiQzSj#eJ#tP5p6+TJN+X+I(?DBFjzo9{AX@@Y
z83cDkz1Bz#M8aP@h`m)Do@I)5|M>H-2+v%TS^+1#-iNKheZ4?Gx`6;}$wQ<|K?~L)
zkTIYv2ix_FZZ^mN)@EjtzSx+)8l;f8$p*qlI{0_^vjQ+heSK2z!D+&x$r!MLWXSan
z(>R_)@vb$|yYkMydp223>0gHF(}PA0)ym>W-zIUbBk4g_?|@<=+5)zl=`JUsHlypy
zUb27?xVB^d2bhD6q#zg)gGxtpe$rM^|2SlYwDkSziGbvrTzpxzMAU0y3Bu*XS0MfI
z^vjW;Mg+5j@S3b_W!G*>&^yHthQt|A@-s+-2Hkjd#Nm|=%Xij;-OdG;TsO#e>iZvf
zq6BilE`l>vIf%9-i~YKZU?=#XMQdn<2pGmV34|kVCx3Nj0{4FN)5)N49J01g@J;pd
zPnu-Kenz7P9kb)$^H}e-tRSh|%#b;%6*7w}9*E1$gjqjHzA6)ogem5C<g3C~nKc}-
zP|#Vt3JJ~=47ioQ)zK_~Xp5gKCOCubN&@l!%HLZ<A;PK$necBP{&d49U0Sl-pzbDq
zjSZ*>Y0*HC+7}b04<;uVA*xCwh-;CHKie1cGRM7Ro3!rXZEjBTxfsIeqdYb$A`K8K
zVM1ao*K61d)2E?=uNpZY@#UMA453y-*ESrN*bz6f$ZVlj6K7&oTTp08_K(?7i<1mA
z5v4<~Bp5MhNha8|KBJW}VW34He^~fil%pzw87i;c({Ih+*I8)t9P3P^iu~AW@&I?q
zm&{*Xr}lBx0WJXifYq2%cq?7>1Fx*t)TwUVQ8#xKA3%|Pl`R`B-r)n7?Lam=d#MMp
zE7n|dTsES<q`*wwe^zCqTF@&_a9uWnZxwDSGCFHSIwb2kd6pIfw1&L;uA3)wk#JV2
zRHn;FK7F{I1qss^$@5lK+4v6E5{&*4H6yHSU~Gb~z{B$xGOC_2K;2%Nhgqk3rpv7%
z1JzD3N!n_%vjBfJvMfLLs0`cthiaq8zPME_AKHEd4hC_^=Y;;4ne{LT=N(JXuCy2$
zE!RyRcJO|Q0|7*5EqyQZw3c@lab+d^@cMkUF*5UY4K>xQsI;cH=%o+agwJ8oJv+xK
z^{GbtMwiSn%bTVqncwizC&df`dMf~vic5UP{}AjZ_*F<Q5ST+jk6`ZS=u|CcTi?{{
z6K$R^qEt$hZ-X3H*x0NSU&h~K0x}i?Ww7g*E3dUVm!FS;yhb)+Z0OIQjQuwjM{T#v
zE6kW`h3JbT^MIi&Isjei)X|tlA-j~r0@S$KAm4ykY4})=Ju3Xxij?gPyuf*1N$v|a
zxeU5c2+<O&&?cQXN@vQwI$hepMw|hnop_AsJVnt7(fOt1^pCS-g8Cx0xefo;Ja}TQ
zCdif&Yk%5qgzHQGVsU}g(TXKS^T&rPxVX~8P@b#rMWnvG?wUx1voANA>+}C%>>Xo7
ziMp=Aw)?hi+qP}nwr$&X_ifwOZQHhOyL<ZmCNuNneKL8HN>1t|mDG<)t$p^|YpsUf
zp6jUMb&f0TJiKvNvg;B$<ueM+8SbL@_1fNXLKZ!R_O9fM#&QGo8(OY?pQ3n+!m`EB
zGy^_c8yFfV3N$brcqo3>j^<Rlb!HC>;ux)J^|PupB`}daZ;NQvjb<{;Z9CMp_32gD
zqtd)6CNMfIKCNtq+oPo?YZ1#i-e@;sLswx=17w2fcFwXti9F|@0NmB$p;XljJnSY)
zTw<BW%Sox-yrKrh#aUYN_fZ>CiLm?qz7<g?rui|_umj2^NfrZiB@jzB#e<qwd>nuA
zYb9tQT4%t-k7R_Dh&Fvsku*W^wM~KoM6p-i;oqZCWv58&n2GCkD*oyxqZ{EE&bQFL
znI3(b3rIk*7aQK&#-kPAZa2@BTk`>H{1XoU)PIPFz1_mkvNTXO!@P__LpxJ=z>a9e
zTgF~FD^x6Y2dx!AjQmJ%*KyaJuBkRFq&4@|AimgrMd`7|*liXb@RhZAi0B<BJHluX
z&Xcn!c5o}AJu@G=?BY$%t5yw3t#X|x_g{70RkaxFto^b!WPo<4WcY@N4MLiGsn3D1
z>R!4bP*>E3CJkG&?N`pD1E$|uk{$+SFrY+iRW6JWF^?PrPG2%nvZSwPm#xw&7LuYK
zTUKt@##@E+57_$y<9<ql(~T?<llheBQm^!9_g6kH#1?W|k~oo^#e9}(2||w?7y3`;
zpFsO@CgS8Do+I%sN&;<>&C{i}bNqn(Y?ZP%Dxz<nT}o%wNecy`nlbxBN@t?eD)ABa
zLr(lyS4VjS@rrhDwnrsxlXemMNH6JT?VvC8lBQ~(C~Zs?de6|v!;3F-1wQ*i$<8<z
z`IRi?b$Bu7^dPQu%qMOI@pc>2Fi42#XstW<s@CM#%WAZXlrAu*)C0gwd8RBaMA^<~
zmc9IBeg^zzOO%T=rGIUzT>=gNsFz~GmJ2h#fX_hWFdA7DyHG(r1pKxXjIN?e%sGIw
zor=`D?^FPa=9fHPR*p73-k&G@l^dqJvDznm`&+IME)czRTm9Mb4r9Zf&s;dF@Zi2i
zf<8B%6^VUljd|H+&*E1QB%3EA+;~HBc`rD8S~n+#`rpC$J&4!1Vy?HLRWCXr>E@Yl
zps~~8ZYvd}mKPmt{%T2K3Vtjj9&MEms>a`?n}(x$a}V6<5sW@9{T}xk|Kalc4VZXY
z5cZ%Kyw|Ye)#TM~^|rsT1Eub$vOZs`twGhn^kmuPO?*{0(_VqIX5t*?nC2H4$w%#;
zg;Js&?o2bGW~eo5*Ki|^3w|NaN=KEx+B-dam%o%&$u1L|tWs6d*=2cCH6Q#50XzQ+
zv-}lHfSQ+m;6ka;TW(x;9kb`c^|wJ$4y16j$jtL9R=RiG-yg?w6&m_cazAD3AZgKE
z)y#IV_tNuQpM@Cy_<^l8&yHyvamikIN&7-I@)V*%g2#W!m|BLuM9%pRy8bDey*SD#
zD_Wb_tRl{n^<SFobx7sE!52igeC=W9;-Bc=FaA4qNaQ@@3%u!m;|JQS5?#~W&~86J
zq~+X8IX}p{M!7KhR;_b6M_A&9=`|cqCySW_cW4|`1z?}A^6|B|kmU#5a`sjIpQ?B!
zde;A~l41MrL!<3BS>b=H_|rBo^-yfbSN=>;Yci|+Rxn-=b1!jVvY5Fd^LQYO-<jRW
z#TxPM<@%K;f#t}3^Bxa4vV^HHQC2lGv=y@%*L7xDch%U;i<g*w=-=M3;hf_^GNNkb
z@w?K)1Ds;wkW0XXcsUDg*bc7Swr$h>YC)ZlflB$+Fh&SU3`z{oV4IY7@D6S56=OGO
zcdPC&o%Vi5xZc{g=hs2Tq<-fIZh`4ujztv{5IbW(^cYDB{&OxS1q^Mi2AkfkY1{id
zRL8BQGFoWJUCq98JjuR_ge(%>>G0)z4#(Meu(tu7QEf!SL57d_wkV5;qN@jVvCSuC
z`Rce}8AV}&kpapIKhnZu2)I1B2deQvBIjR=!gT9}-=T3d4Vr}p3BVk4IPA>o|H&Dk
zD8Tl4=r}<v0u?hRg%B$cS|-e|vYr;b2VGgeq5(A^elf9fFvIwpUA4eulGJj;`G=r0
zK*p+?m%Vqu_aJd>29DCQbv5^_k*kULc$ed{qPs`Sj!)twWAFvuOt~3!-|160nN|{k
z1~us2pSi#pXI|ArY1#}K63@H1<Kb7odGr@CAYP_v#AIG~Uix`LzmQ`8s>0LL#pikD
z-ypU)K5dp;-t<0n;qDtPvoIm~f_J;ccHWn=U$Z1}LmsG>TG}!rT7sM^y_fw%zkC9y
z)RtfEe$tev%Z}P7f#xACNi3~p{`v83X-uMFF&fTcGCl{q=w8(MkvVhx3L!fSydP9y
zHxp@flf5*sv?VeeWwifk(s4R$R;N+pD0@}xkZH`V{RgJj;6I|a`-e8#5kteg53)#Q
zH1yQ?sjZ*mibf6~6R>dxmeFN#MzW9^>rwBS{54k`(X`+`DznsdTcXia*VZlJqUza`
z?ntlgE7NVotd-9IM5|vH-iq~=n+kk65XqIOI#<GU`nf3X4UQF%^rR|`3%d8kN<Dv*
z9A^uQY)S|l6^W<oH_yT<ov+4i^G-{=&2z2qX%M&4kb2X9Ixd?Y$>UN5hUCxX=D2lm
zfpa;?#Az2XN|k&+-=9zC6P`l8E)UNaGVd<1kDOJ%^rHtV9|sx!k-QfV25HwES3re|
zMs-<}v=<(e*H4ni%#iBDbey3|Dv3?!cJZ)}-=?Q(Ewr@G{`Usw8<`0l>q%HL!rSrT
zgLT+n<}{Fa)xuQA7sl-$>n010iplY(emCBc^nd*YiQH${N&(hLbu*aDI*V@YCJYU*
zk)?kY5_tR^$;%b~nUTm#1F5tUdhXKd5`bD&TG0m4e(SbWCo|tP4d(+xe~e>nh*%Pl
z<`q8IAO5}Q2)MY$iWhki39EQ5@3ez#xI%jmsf0U+cu6^xeMOia1ufh|;CP($$lxO2
zMgF>`Ze}|jayZew7tE3z2j_bprbIv`@%FG!UZyCMdMX@<9z%}5D(OZO2o<B(9{aUH
zVx!C_!?dL?Bbu&SJ=A5WESdmT3W_n3xwa2&{$`NYwf7t+r3=U%KYTcElagTHvxb(*
zs`JiUOO0Jq<U3kZt(^YL+?1!CtA?Xva<H~tctPRfP+@=qN>b!i^}B%6{T0GvscLjv
zFg}^7M!073Ck#rF2hPUgFF4-!)>Pu&1&FLd2>{0?4dvc~s2+Cu3#g16ixXCj6}z;G
z>jfa93VF0O(7xl{NlmkEqUGeX8Zp1fJeKF0DZIYV>vNo)30-`?UAi6TPa4WmB|_gr
ztbu;44og?r)`W!B>-n*PcL6)LH1)YTR+y)%>}%k09-@3Mj}Ki@`w$HTeKqAj*KaQ#
zh_E7B7AE8*-z@0M1}|@EP%P{Hb5*neix{_%D}_jg9ui0lnbJA4bX9f5oM;)mW+o`#
z>s0yDXgP<8Q=>|9)OFuHH)iaUc=Dhv&7GNo{E#H#Y$d1^axvg}(53h0;XXRO9F}-*
zlpDTHOU?@cr-&ksk3uqUdmWx@zt|NG++j148GbE%ZyEqCh=ii4!H(@<p?gy5Pcf(=
za4$G@AuvJ-iix%>vqwG>MKR(J#4@<@j}%7gs@Dx!T9<f4>7k4G-3D0FRa|+GTt|v|
z)SkP^I52LW8G_rBs<;`yHg=wQ`bj&9+CsD72q-Y!$;6iEWz8H2L!!W1dVxHRwDHZI
zYbMkX>rj$a3QR?wf9Tkifq>miik1!)Q4k)vjQEWe4u%wFTlId%4CK+g&U@^rO<6wP
zS^Ib-p7J{mtyqqg8%|tSm&egdnX&tXdK0Vw$LEB+3!m2!QFlJu9A1|Z14c;nvWfom
z?3x|%8@#~2VBtUF;4kF;FSRHe>wia+ZfQvUKR9Sk1TmGdhXeuRo@4F5jrpqjIEyMH
z%fVdw{pp)bUOl?xWJ4E-d-x=YIel@!HlVTPZRPy%c6{sY<cGZAjzaBWo7>A8436n1
zaZu72L#8NO=JCD#`7B`N+V-a{dYQ0|`y1}H3GOBYXdAyvpiS!Wb{nf7rw^x(%lfX|
zf#Lh*ZJS|&YZM&KS#u8(%~{_4_3JGSD<`I#Z=K$PBC3Iwbs7uQ%TsEI51uRtM!H$v
z2G>C#<w@RzJ6F9Q)F^sJPP_8<8{J$X!6ou97sa`AdeOmZb@Fv_N>b<NVtYn^is0<}
zeR6XPt4Z}jg#J~sftnQyZS~_nihmMD7e~9`fiLwC+q*&PC|wrWb%YDx7x^}KEvnW+
zyP{M<RN*Pd;JMj5U-uron!}Vx-o<Ai{>jFJdi&%;awiGWH^tnr9GuGe=@FRVBqh&{
zOZw$ndnn$<ca+1w2|1=4Hlm<rd6NFMZ_J0W*@%}RREIUQ&ev%d_N&L2?&C&;;7AWk
zCv-=MJ2vJ_{5-)vz9eCTzUQoQ3mfy7)sDtiLV8?NeU!3TejL~lBUNYGmqPPMebIWF
zdnS-|0#(p|@@Orusk6QBO`1}eaMng8oiD2;vwD}(0ar-PiIPmGTmK{<Ozx7M&}0~B
zAJVi`2FYtx<P^0c)YGnDS_L{%gVy89CNG~Cpzyd4z!mK*82Pov$c)}oFpU9YOq48b
zB*N>#@nEOK)<4IZ`HW^+v7BSLWY8P<Ikvb=m;#t4gj6)P^vM^dw8>-1HEC0P1V^g!
z&bkl4xSYp($T!4fkcDEJGWn(B_OEUAr^SlV#%%J79k_LnKf~k*-!q~u+DfoswT`oM
z<ASG`kPc<Sksy|v(JFs{uzGvf^!Rx8yq!ctNFyI1JoWbM6oH5Si{S<GqfE5${oL~I
zerwikrsVP5L~m(dn%2;A6_C}q1t%<M;$R6)>f1yb%O&0bdF1A>xY7&)yH=F`^enNl
zoTaS3s^BzZVPiV$ge1$4V>jFF2<S63{cGgbyv!R5b$)=)WeePheC-ll%5;Dure&vX
zx1uEyPT|P7V7rN@;7fErP_Ccd3SaDeVsw9Q6kDE^>IySW?Xu(BJweD#161d!7WwJb
zS~k>pUmVf8bB#)cavDF?u10j<kj;}7A@!IK)J{(;Eh<rT*7h$&BB2aK7n!gLUDDLN
zo)KppgXYwKO(nFu(`#n3)sq1dqSVl)w4tt<fOSmQj#k8+V9S&+hre4xfLu09Te6lA
zIR*f;S<o1^K-8njCu9^;*Z~#NO!rE_Tse0$#K8dSG<MTj+q4dNeORIbSwo1tX;zy2
z1U^zzo6H(ZkS0LG^(dcCxGbUH)qDz143<Vbi!VBILRzA+W6HE0KJxId=S9k#qEs_F
zx@KI%e(*PUDm!^R_It~Pg?9b$()nV&^y5@xcU?0?aXAf9qqczpEc(dhI^IGYSgOl=
zxN-JHvPll$5;ubl$r4-7L?d}0Df6ClU?y+Uk&+Te=sI&CVqF~5%C$XZ^_NA3&0g^h
z*ZZ_-b1gi4PmE&-gJ}7gcXa(;SKVIKtZOY@o^4F$PE3*<LDS)I*Vkc^k@>`&i-DSj
z&)#8@f?`fikPd9l{ng!JP3auX00I;e!_Ad<3IecgsJ-bGN<fbn@+wWcf9}HQrOdD?
zK#}A6C3@D<0_<@4Hv5}IDHda{@E@T91JTF*F)#|FbM&qy<DkF%S)KF7g*N_TMx$f=
zwlVu}r!RjKz+^=!g&#OstLtbDX0k0lccPiZ|3V$l4;WH(yJa(^ZnIC53xA)AXdfs)
zB3koI!#9~3*J*X&LD}J@rsT6M*F0Jtp^Fe$)M$IOR*?t_Q_}d+;<W<^l2VAI&S+N!
zNn#aC&OSTQ8GRgd&<|OooBV!J3x`DiQpZl^O4Dgh`A5c(2OcR%{!|Aq)~)pLl}@D}
zTX(dBqhI!+`NGk%^9vfo4RSd57F&n0i&=ZIdT}Grj6+{Jh=>F5HiGZxi@dUh+*TXW
z85aI?Ph->|=ytng`F-hUO5x>nm@vDzoPeHXk9A=aKok~L&E+!cvN9^Y|5+$qE2T-H
z2<Svq)1#d~w6t;xc$86**W3+a*V<};S91@T37rf$U!A0nH!RkTBTvdEowIr$EmOwp
z`q(c4!Mxyr>b(5>0Br~Bzz?zKdN>=w52L3SLv--buvM)$LGK5pTLfn?GZ>16&T*6B
za%|pl6gtyp#4AHsr^Ttt^UOsxHCk9}oBZ2Ln96M)#vii+8tqtX28U64j`xa{+nnUr
z{PxM;@%9w09seU-GO_(HxMXDd4aE7chU<UL75<0(`Az&|;ANf?y8}fJy29-QADxP7
zzq%%%XZ!=?m9T}K`1#>oBpy+uTb(P$3T(dkSGkHTN@Zw&YbDkVF6`l8?@L6%EMuX6
z{Q9mO$-VDR&mIPIOM@U?;jh={_Sa*8lZAbb1Hy^qUc>9J%+pPb<p0P#^_9ar3rPYH
zLJ&giiiUFz5?^f=R!RD>{&5|+r|;pS*+clfOdRCFaU*2i0Eb92nUyo8qMwQ*=eE}r
zq%Tg7#WM9DC97`TxHJ7m{sA`F8>*rfVQ_5dV|qY7Mt-veh`onHb0fSi?<U>qe)vl7
z*Ep>F_4Y#>EHdG%;+vATd7ptaES2(G=}-fHbKWYoAj<R|hb#eu9lnK9BWZj=yD~k(
zM_Cwl6Ll&H!9LW#35KbdQJyO=71Sb6xeN=Y4<9LEBPG;Eb{8!o*3AS&c%l}eSPhPX
zT29;zDj0Hmfix;2GstFh8#6fe4P=0l-H~`BC}%c`;^t?=qw159PVy=V9a_jd9_*^9
zE&TTK2`x5B3aP?U*A1@jV#=YbLoXW4cS5rKd5JPu<K9o6!<<B1u0#c!-0rgxkd(wS
zvuDel&<$<0^QCUew$|iY1;eAJR#zCjMC>!ITZ+gbPOFLuz+x;u1n=Y~>s0t^l;$<4
z)5!4#HIeQXzd%`tx=>DkOM${^a-xv;hFW-$+-fCNsbjvz@|8nNS`o76&g>t2;hErp
z9b4>%yoYWLJ3d=D*lR7<Eecao+Df|#yA2FB9rjG;dI;>;`B~?|;cabX<wBxsU25}e
zW6)g5ouY@msn{c^dI<Vs7Q#waKgau4+<N5B7t_KiwVzrFifjGAHrrSsV1Of>zrI#D
zc3adSwI5mP)u3tvrk8r@`54p}Sm@UIsHV+VY6#8^TX(xf_;C#qQoo&be@II^MVt8b
zuf>ndD#PDp^5f+Nt$+)TM_kBhPZ)OV0ln=#-PrQDAjWuerS0eOkv3q|177_v%LD%M
zbfEeW{N)lppOEQAc$Gm<=-Ue?CRv7U1|a60tqRCAJ!4Ji`@8KDVz*)Lr0{Ww-y-9f
zkFFJB0KGw4myDI7Aqji0n3bXcsOlpAj|NIKF(5U|;Lb`yya5=@4Lxjm0aKw{X6WK<
z#<~eMVk&L&x&`(;J87|b!7G@@t#bbyt-%14fhrB5#D%oS0JYuWKQ)AM^sH%(UuS>R
zD&7wYq{OzYFJ(|18r~xFRMt=AL=p`lyLXBi_xLI7(ZoE$<N7>Re6?4KFm|;id9B>Y
zblbJ|ekBf3?Zw|a%F{`iqMi{Y3eiB@#VlnY#3ipsQj|?zpe88e2_*r$Mg28<l~K;I
zD~8zZlglPp3-|%2bt9~K^W*Lf3#@rVA~&R{bw}k@288j;#O9M|xs~w6N@)UCqjgry
z-+j!!>NS%qif>}jpNTLC22rPz?8d`SuWnO-2l$#C@%JYH6|E|{CBdN~)*=iSfJ~uz
zh7)vAA@p6f2pECjImbY{;JbhM4tF1&g?j|%DBPck480n3y76M%%m{?NRSSP3Lc+@s
zRxNN;<g2~70w{_MbZsE5S)!B&u_G|{4y^+fm-SaD0?MUL6<l%Hp7}}Kq<+OY{@L;;
z*#TQ=Z*z@Dn5kuO0WVXIpbXT>?3oyf@7*eb!lRSo6O23A<$>5JGL;%?Dt{uROuIdb
zy<0MPmhH7GI?A$5j=5c#qLi1JEQYX<>#U?}H5#d)5!F;vQkEwZJ-#X$0*CHbC|wql
zPTuR=s(q_KhLkTl-W|T4w7*u-=DXymFSLU^lGf`Z$K<3etimdtV^yXEZcaL_8VH79
z>oQTj@HBeS<0R2Zhgjo0o5Ixx7#qWMv@&uY#WOaWQ5F5Z#P|%%%Y6orLchS!ORVLf
zYnBR>CT}ZU)O~mg2e5e#F-vG;1&65}CE|Rvm6vqNKYjHzV(uW8p4P*?N<2RtsiYaf
z`L-qZ!9L_m!YfM1=%R85wGKy-beOlj9R>zO$%M#lGUj{Om&*dsSQ&zo1y{j5nPwRN
z2+%jn2gZ>Syhc%ovlBxOH?L}fj!m>7gi57=k<};$6U!fH2|JA05*r@wpB=$O1aOnQ
zSJ>9Yt>)t2-{@d2)59UQTDArDXb!L}T3NLXv(DX8wCA!u&b3-#8N#C6fnBhivog9;
zPtIa-X-rq1e?PKw(7=dM25Pg@d{zVR!RZv?_N(A(E%$cNQwI{qF#HPFlKABT1n{<h
z4lWkp^u7tv0lqQt-NoR0Zfica`3!BzBrdMCeyllw`S>G)p(#wzpu||QsZHAqeuHW9
z?PUv!sChq7MG-vO<v}q61|Um`ib_4c9gOqwWz)r&I_!@0Ni7HAIWWs0<P1_luy6tM
z3i^%O#!o&e^1qjHaytf21TuB^=&)<FY>$H>i~0>U(1374x%0<J&+h^tE%q<<$4oJk
zc9(-S37>Iqs$s{)AO2(?Cj@SA2bbT&c#!XG+<{(WiF$bun+-UdQ|YUo>c<gAG@)I*
zeoVZ8JfrZ#4Rz-olP8-2G2s*oYa8xMxO0!`-a4E;e7fi=mU0zGRnZKCAGBMm%VId`
zULW0Cl7d3!bp{TOo;w^qJv8qJBsP^$4K9T06sik`nzdsOHXtvO{9#AddK6Cql>rgU
z+}O-BIS;a$lEGjtR#f=mp=kmM&X=*u3GAJ#ETKy^a|0$3SrF&V#nY@H4Lbl{WcD6U
zR&_qXX9`XrI#e7ZzD6w%n*G=mNLTs+;XBe#NIf4lHsF)eM8fS6p>ur=aVETkbC{w3
zaxX&}{2=*F{cF7Qb>Y*9H5KM|t}TN<&o(hq#IJcBMadOLNQJvXW-SJ}vr#dOp0!FX
zLRO)t$i6YdWVtik>uR11jxpAGu0@|tnv-@|QItmxZJdl!jNLUV#Zb|tn9qsuA;C+8
z4pVzlR^gsxpXg_pVxO3Mc$&UndHvzYu^zz&!HFU_zx2?o>Wr|L-Wy`#+zEc1l)MwW
z@_u;w*tC8$=sda2*s7$&maSCftzDuxL5Tt!&te|;SdGS+t~(L&j`AK!oy}P8Fl(M4
zV#;eRaQ9(TK-Pk>^8B{*x`6_J)ng<)Yx<9M*MG9uVE!-L^-KT9V&ji*p5ERF$Pjsl
zp1=O=nv7|{L#qDHIxljRZ6u-A<Nd7D_tzV_R13)tS(XjfzdKE%$V0J6LW+lDrGc!@
zfvK;(DUq{b2D{*-sV^g<Ti}tOK%?2rVaUEJ#HlZ(hb^t?;vttHDW{s>q~%_o>tko-
z{sY(mHW1MecAOkkbOm$;cQMRO4%&!b9n4N*PiRj&o>B~3hk!8lJm0pTnE*0_hwdbK
zjN=^Jss$KFVo19-in49(U<)8<RZX4KQrLxUl|}8K+EDLzo$RBjAz3A@F?Vv0UQh2F
z@k8baqs_I%!`@o7cJ?uCiv3Et2*Y8^T&D67O+W=Adq9`pQ#?}}$Vj%D+CE{3>2L71
zcK$fAgg%reGO;0oBt_p)nFYC78~IpKHOR;DlY-=1BpoK8J>%F^GKweO?@-dS{J|d0
zvnCu(Ss{se2z&Q_0lZr3VkIsn`bhz?#)JG6b{ff8W|So9r+6g~E@nE?v!r&>tu!WA
z%tZ8!_)t+rzV?H>Jpz;bv`3lAU)EJtI2ND;lQh`;gnD>63ToFp0_KyZS%ml)4hf06
zcaT!FBqj{pY7N<<O^bwX2Wcl2B@w)QAO5!1YO<#O!yFN<dss|_bJ5lrVWgp$o}`^<
zUE#)Z0-qw-sB>{6fZ`ewBu@prh%J6Fwt2BuM~l|U1yFRONhO+xM@;Uaa7kjy12kAV
z_>AbY%anMQxm!q$ei^Y%;r{9BPNB>leKbf#*N3oW$+<w<J0Ls^M7g;i)NcL~hW8I0
zDXBH!Y(-+Ri2iJFps9aF@cEzO_(n)ZDK*6M$gzm&ZX)3|z#R^e8U<!RoGtnX-PY_2
zZQNJNe!FRRrz4#n_)52Q98YIR#o~Xzw37O^Qvk$v5l<KJ#qgP?kG}>NnOLK!h2C|@
z{0M?hmsua(C87|EQq{7y8jLVwto=6x@Db*FaqKxu%Pp#{*zTv%1z0`G5XHF``9_7L
z0tg#v!N>r#<1RA82z^dmr-j6HRl9Ixuu~p%=yjhOX-8F$M;&<wGUE}0AY7W8{C+a?
zEyI_EyKI@<2TO7s`vM5D#4e1RRM~+G?)gPAE@-Y<@mC_-WWMhm^w$oLG1pR=TEe{!
z`1#JT$^jNvH=7cLnz;13Bzpd?n{kNmJ0HjQ3s01X)a#IYq%=VvbK76Vh-SBrQl#ZD
zAevm?SWRawlfj!ZEAQL)I7dJYkr(d#VT!fIJ04wzQ0v%Ik-NEp>y5i2kvV+6b<W3Q
zezv#((#`|D1oHb_<i_P)Y*c+>z(aj;AaSs=sKo;ln1BJQ7oD8rtOD)35;ygW1=~$?
zEg1S}ce{}|TiI`h3MDx4;8jQqVr-R`(sGK~9!W}hARP(AFrU`{A`c=_D;R<gIw<b`
zyLcin$?o=RFF_dgJ^SGTGu)-~Ph&<wTw@0p-a*}dMYV|JX*+i;-qUEtIMKKx8c{sG
ztxCKH+;k#3+vIo9ux7a|lBJnZpD1<01g^EAr%01xDbP@01L1h%!i1u3w0=z2jHSZ3
z9nPzYR3XuxOfvDOnps1>U$yPCVNzI&c-RyTsjytOMtV4nM@SJ*q@0JAoCd6W{VWhM
z_9ABEAh_O<*{~EzlAWNdq9`#VA(vPHy{XH&Kn;%xa@;SPQqP(plIV_KrV0QI>Unco
z9Z}k$i*j)Be0AMINx7HG$wr&1?8S9GbP<coCW+X|L8j54rW1b+euaD9Qbn;^gmp@~
zdOCFUIOF(lk!F;&F%zGlkDRfRJ-Scv7I9*pf0udKR+$)DGL6JCA=YIH6)jc4R6g##
zSGEwL*pg+bop{uJzy(haLA3$FYfcXV4y!NUeOdFrC>+%C8QLmfvo2jCs|1=RuKRdn
zng)ZD`e0T@rEZG4KDlD2PE;{jFBCJ`b-J5bRiv@Hav?2+Y8Ug`o6h-Pijl}yna4k<
zZ^%wgY=VLeHTyfUeqVEXp9RAW<hOo(^k*pyt^~U5ng<F6`aS|sB8Lgm1h%fPQOWe6
zA^kTtG~V*u99S7`s=)yuzsi`h8M6$H4Pd^?Ag}AkQUf2Vz*LfHU}U5VDVg=CW;Hnh
zP^op-Z{j@k1l@Uq_*TGh5881<Qd+xaf&DDT+A>}aB<Ongq;;eWM0vN*tK>NKW$&QF
zr|n35rfqW>mS!jSG$0m+H;@#Eb{0Z)EHAK_=zB*XyL^xwAS>IT8bASgD#~AK{k^uR
zN>A41tzxslC2`rfHbxh*Tp>Ye*_MKpQ{w%R=Nqml^<l@SzG<z5Oe7dt%}TSKTW29F
z;kd&>X`Fbn6+543(K)>ig;L?rl4?leR!XR1V56)Y_<{)e?fo5$oyDtFo=_1(3&vi&
zP~V1eTCb;4`SkU5nmm}b%Io3I`MJH@-AUR5<D<-9{j4T55E$IL%*jczAP6#lb$i?@
z7q`RvNsbQjO$@bdeo7_3$lDx;Dk*6r{m||s3I(tw7}N!?g$<v!m^gfsQG?}f{wLG3
zT5`P4D$gRhXD)xdiL4|Dr#1oj<n1FtHjv!_cj}TC4!lvGC#+0?43uY%IwBQUAg}y1
zp=Qs?7e$$mEo|iYwm^u9{Ijr8bmR&-U7s==QHZ~jKzV3XAo0<?Z(d{Na94hcnY5sK
z$sb_M0b)9VVZJM;#KSn>2De|?Ba}P&=IJ&1Ipr@6Zn?ZEDf2KB8O&RuczjFUvmYNf
zNrN1I2iE3>MXaC!hCxZ;Y#s4qpY|Pq9VxhU?MdrH1y^1&Mv?y(8r0PqVs*u&hoy^3
zl3+006#rn^`7dcdpz5wqZ1f1`eontRfJlwnb1t6|bteHhW<)z*z-XQszA$H>26trR
z@od_LgPL28ZL_n&1fHu*k7YEk3xDRg=dGK)<(3E&JNp1>^rAC%>Rnsd5Y)f&=Y;BL
z+Uw7iAkUHZ8_I#>yC+xhHTo9XqY<pB2}MU6lZj{x=F21mUu%$ji-2C_2;Gbgm+X;9
zm)<&;k9dLYhEw%xteso#^G%9wLikB&2(sXov9<O32+B=!ztW^Mqb&!Q1BFm{FRD$n
zc=w@%<XK=zvwLQF>tN&2!Zd@_$0*~oR)YY`a{!G?Pz%^vWGx9ga~|Vy(7s$zoIIgt
zhzjDoO%c@>RY?41(CTBrm^Go|GzSjFN8O6(xZc5!-o45U0~!AaFYr0vL;BBpBag_=
zH>|u**zpheUp$_X|5(GZ{RgUvm6hSY>Ye>6@|-q)M=mbZZq4OJAY9RA$}2NgBN!L7
zn0lIdTrzVV&1GeLefZ+>MfknN8;n2rlfc^l@$z2M@@ldDFn7E)dAP54ZA4gfL8WrD
z$?0MCL&xM5F`QZ-L?k0w=Jv4m?pRIzSplXkY=)qXJMr(_+J*%67I+P$mu%{O2&E@P
zNAMm%b&H9*yL?!s{lA%}%sqq^p^g?m-mi{1jWstetW;)~ZSvCvD@q3|*lVy%%_?i$
zPP3%k`U*((xz`U{<5wWhjF&T-VeQWE%&=@1$!MHq87A&Wrb8Fa8S1Jay3hQJ)T9;8
zcIHgedV&d|X~FGrb=E1{JryLbIEoem`@)nyltIY{SYB@Iuk0>N^Ks7OGUw=K7j7$|
zNca2q#F0)r_PA!(Hyg&5FK*b*>A^D-N_OcmkuH_#MNi-u9qQ6*9775LjtL=Zmlo0y
zW&$>k9u)?Wfcwt)r5G<JFe_*!O{*D4&B+@fx2w{t66zf03v4J!YnzjgEr0%i(5n`V
zAZjkT%G1P$g)60}jgiu+-mJbX>TyneF(%g{W<Q^;BD7F=3_L&@lriR=Z`&Sg-%7Ko
zx#}C65=*1foFKr1)p{k=@;-~z7CGg{?+X8q`i%xzGprcfXTEhCrD**L7-;a)@Jbg*
zhC(<{#jqPAc>?k3Q^#ix5du?T7K7f!aMET234^k&7MP(07DrhTDlyALiF#fL;{+O6
zDU_ty%qc*O=H)2C94=V#i4-8F$TrVoW?<6u7LD!mvzO;wrU~+1I2V&LR0gvz9k*h$
z58fFmiADj}G72w>XqP3Clk*0wqmGeJ*r_0mDyVJdWnq#sq#PNzbybOV2c$heFzOLp
z2GxvR=(DUJTP`v}IJpRQg}GqG;6@1K3`B?9&T7>+?@m-ooodHkHw!Gp>AqYB8cP$Q
zv-GT+RyRXC-=1!wliw!ddIXMtKUMK3p1+P=MP9O?p&MtaqUW5l==NgYQpq0(%RAOb
ze9(8nt&}zdTW>bqee<|Mg!TTvL?PXXt_$cH&w&8RJ)ui<tx=WekC2+@_B0|8IZ5}d
z%LwzH9t8e#OlgXIs6@oH<az>tyo;U7cHkaS23@6uu|9Z_Uo$U6aWlsJl_->gEC`x`
zLpVk}i9L&oHm6(M-s_J!iE8*Br&=W=4)<7OJw9xVbJx=?b<31FipodnBUd<_Ny=f)
zKgN92>!Y{<v*5J2nd2Ej62JCRg<tTzW2qZG7MF04CCyO{HoiRC9=^uTA&m<tP^v6W
zy+<3U>hFy6XcbF}2D30lK1@24v<~KH6O!%o-?MM*=dA-hA*WaH*W(A?NKIFj<~upn
z*rKcM=>M7pbvW8LHDJwh%C;)jc*v!%(Ka1UE^tjnk;m9+;KS|!JMFZ6{OV}yxjE%)
zcQHCA#k;_ACo(o(<FtsFIHz#}%w#Cxr*_`}l7?2c6cQtie6<Ov$faMKR_+MJd@UNa
ztIsO0a3%IQ?H`NX<Sa4$4^<j1c97fxCT6i-CLQK`7U3-+XGN-nIG4snrgBP>?1sZ8
zyC^_N!e>=umF^4>GC=f?SqJykO-^AQu|ljqI`%t|lcc_~UuML?_NxS)RNoFXvBT*L
zt_soi`4wldjnECbFIJVhJ9auw9!Pk26e!yNrPmRdtgaxY6-hIm$X_sz$m0_fbk}Zq
zYs?EfE0f|@CdqTgV4ptjH$jT2=3Xx4w{GS$F_u0zBc_?Dx41d;4K=)8G*Dj>PM?2E
zr<zdv5KkzEV~;c-6Y@;#YX@yFN|7(5i0uKJu6dX^{HvCmb<)uuJd>g{%O0L`$9^^{
zKYu~t9H#ttfE-4TQS^xs_DX*QFPaigT`c6QSz;Hg>DYZgyis5mjG{KDS!frnsw>%a
zbC!ic3--2=ig=+6L+KaIM4xVj87;_C(aj??WWVly{U$BVF_R!HP4O7&p03C+PlM73
z;lRG-coj;dfhNlHIIS1sVlZEK-h81`KF9=_`3O#9^a%~md?ab-pI?sLfhf>u`V12q
z>dosnsJ#Jus1Dl0>y`v0eW~ZBrYy;wbWE^l>P0-p^!-eu#9f?QD|?H4t{t(h!TRRy
zXg19nw0Jq^hkSADyN4}kZgW!544a$R2ltlGJklx-xAw_ry!rK`5hsyg^K)ZzIa3%2
zvwkgJd^W-;fWq}~XYs&i3IZtepRnxL-t+(Cf&PCSm>B<G2PSrw|JH#?#}11P)n}#_
zzm9Gz%u#%MqE^HJ--c}qh(?oGL*EUUsLq7L;X+cw1Kas!<)-}*hDJouaw4(Lx9%5y
z<J0h7vYj}XmO1cJ6j1#n7$Oj=9qW-rYROTk(1*rQCMxvX)n!FuhX_+bJzwMMoBygU
zH~5j7@(Wa{+-2ERXC*&rUznN*!4%8A#`{pjm{XT(S|&5mMrHi%uT>-{Om?isN<R2E
zkHzW8TdJ5zkK8acIbJM}9ac$uVyP*z{AS20GOfl$^>>}Y7by+Om>!EjveVc>OIpHH
zs)&3AkVMt`+&@vx`oh0KWwHkbx29pH*ml@jOie6cL-C6Jpe5UKE5FO7*;aYmgRI*?
zJy)?hD+R8=GZcy;%<!)up^~(+IAadI#+$ViC^JS?%#bk(Jwq{&<k+#HVZhpknIR{N
z_r9Ve<={F{DjPAa3r)Fsn4;>OLk>=@;TBu)kBHtgut6Pzq!DX>V|unUc6KZU$QjHE
zL)JR=x84hYg8bM!BSZ`3m}(M~LNO<-agwAjdZwEC-hAMI%^Ip5{Xc8(yJ3<LLCS8E
zlp%$){<6s>C+n2UB$^T>j`ex0gz*R|v_Z-f^$z?^1MC&jhSlcmxfBAm?iWLFNT7Q}
z%-kp<-D}t-iZy!D<vB@!Ol7ldlL~2^j7LMAnj?Yz@g+y<+^3%BC{~-%W}M{JL(AS%
zPz06X6BBdNl=q${`5T&zB*$tZBU9QWd42q38W|HdIN#JAmP||NbmV$KB-9TgzK`?D
z>kT#_N*H^OwXcJ`pG1N4;IDJ<`EW0m&h8#w9u5JOB_F2VTsmDJ_AF_XG1!Wax2l`c
zs0_!a``X@@o$m9Q_wt73-68Mg!Xz}}#>j5c?yh(NN#-=@e@QZHuxNpl4sGzn{nzN<
z2S<US(7Pc55KGYfP_7g9egDuf-y@1OYtURelcSQ7u1|)%?=9DR&wRh1MP5ALz4XFX
zyT50`D8Qa<BkQN%&#HPU@|Af!Il41;Xh|2c0MCmqN6v<p)>a;Nc1BJ3A%-qx=lX$O
z{{Auz%3i!#Q(-OME_-WPkt>WcD79h}EK!+5Q}<I)Q;}1WQ<OwJHb#_xoNH}oyxQOQ
zET1dSdvfV$OQQDMIj+8!eRg~vhL@w#79@U#!72fZ`In1;Asv%q|D>M>#SZ7W&Y)!a
zE%~td@!a&N=()FGV8z3-^Z!e|%9E2#J$7Nk>uS+Q-~H1<gvtyu1(G+qXQ%tn2LxkS
z#J`JV`nrE`kOQD*F}xWoOK&)+&t&TaRDez|fsjS41y+>a6_%mD3;=T(f|#amm;=DO
znvL0MI0X*Y4UoDD4h#f5ao7|Vfe(W{xsW|7Y6fznO6FfjL9h%Mqb>oJ@yv#R?aa?C
z;Ac+27N!GPqx}Hro=12XKAXRX-1p*@9zLk4Z{Q3hUrB)1!WocrilYU_>3-{^kM#SX
zF^l<?E#a!Ur1N0}dNzO<f$o+nS+F0%jJyQ{MihU@?Z)oVajbyWcx;fHtk)=MnEGTt
zmU;lp(JnP9$?aKVN`Ui>)<e>_C2x?JpWj>e=daVt4d3Qx@x`{z(~aHClkvCK%fyu(
zUQbuL@1-yI#m>)T!4Ix-T^q>NnO;rj!^N8i8v}2imN{Y7#iCy5!vr$j*J8GuP+NK%
z6JPt_?4^iXW4f~*T@CHFj9svz?fcN$Qq|3L*x#{F*X7>swvMw}(<I!fpR=&nBe*_&
z$=oFlicf5t1GDtNKDVYW)k20g$n_xbKmT?&B)mdO)iAul=l}e@9f=}GTeh8rkK21r
zUT)_J*=pm-BC(B&$ul}QS$*X=o|BXTE-C>`+`#h!wrL9J7a1jBtXG{QGGr4W4bx{d
zhaLoF$cbOLit9~~5#$OU-dWb7m*A%Li<6E1?4F7cx_5KbL>y1(5y+1?ZM)>@ppSoH
zMf)?lwTc}6I;{ShMBEfBp&kTH<Y4Ep=hUTR&o)G|Ig|*rs=D`PVJ5Qy_==D&MOtN0
zlkKw!>@=3CGKuqvLC@A4P4RzUY~E8eBjZ=Go7bamZp<(<umes5w_S1ARuJJ3#!xLH
z!s7g-En<Hn`XUqtY@7^lFlogDDa06pLZNYs_D&@%+ghW}mX62|_Y>p2m=m0sHC<`r
z#m&TRVj+*PQo))36a%h0E}QwYXLJLyi3^DJ6w<Y*lFf`9)gBM>>ucq(;|D1~X~voi
z`xjrjw1^$o%|#>(ws1I%UMAI13H1okMrxe68y{lk;Nz!fNB@56eYyO=gS!@STaXH`
z<hz8i>mOXZfy#@+d^EeBt?7cE)Z5QDGp0ng?QLS4P=qZ6Y}s>Q#@In6=*y|34&#6k
z-pqha`&~q>f13{5r%#-Ak1&n?VAt{AUmoJBL<F_);tF=qSQN{Fu1CgdsajP!+K~+D
zE;~qBUiC3fuDM{CKG-r=9rvq*ju}-@CLB^MnYWw9Bue2b#4PJzuuZU0UCNmwFvF%A
z64&u#RNBT#=}$!j{A=NK7)7286u#2=gNjX{8J80bD;o{cG8OTGu=?e&VOEKQYB*?<
zR~S3YgRpr$GhJIsq>OY0Sdl<t5!Lnx0|>~I0Mf4EacZB(3PA-9MN$@HeJcbu(ZI(e
z;Ft@I2jMfpgjTTH<tr+bAD(RZ45mKXXIpX)J_9UY5ROp_iyt#V5#X2(S2rKa%e_mT
z#Z-d7nPXz~%{2V&*GA*V`2(7(=ab!{<Ke+Yb*=r?dL@&lI#YjC9XgYjYF(zfhLOqN
zIQ3EiML2_et+YgN*QCEGgtt-!0ib2RZOvJrp$c5dY4FxIEMNkT<+kq~d66yW%Qk&d
zT!_|CU;pIJFIdCS20dl@w5ij5N1t6r0d6@SkSUp8f*Ahd`Pd(?l#Gqsw=5JFv@b({
zFRY{J3E+>U8!S%hXa5hdjmU0XiW@@L8`J*bW}`7`NTV0j1~+dBjD~KJJx(~W&2$~>
z&lp@26}B=$gJu~KSnOyL7BGmvej2#v$XrGI&S>`OG~}f4=5<hWZe!G8b{G{$2J(ye
z2*<xL{F`$DEPWSvTGFF~DzLG$MD96R_>228KC&Dhl)LMJ4sv^5{5YB04R8o@BX_*N
zI62@Bgmur$;~&YIh!Ej!j-9x(*XKcOoPiDTsQ|)<Ch*Bv>vwfW1(gH=6m&wQ^z8S#
ze@5~Bx5eXJ>O4l4+V~7L$-xrYe!vC?TUIFEN&_<3R?L$0wDxby*$Z#Qz^SEJGFK&6
zLC$nls<VfJ?QeloQI4MM>!0)`j=nmI?whiT?kyw~9gui8gW*~a5N!7&v)cAaYuX~H
zZ1>069FDHG50m2jhYly^$c!+%^MhqZmgKhcokmv)^d3?ZdP0IG$WpuFZw`^jJj|Xg
z^!voLXVC0^@U6T)HzM_|#bm_Bve4CY4$jHV@-LmpYwe(dxV#44<l-yLrKDu>6JQ;K
zJAeO3P!w6*`yd`*dzw87qaqr(797NDskPidiP(v)7TvM;l3a<nje9BrWjtX|^bG2D
z01V5aJm=u%mL4UX<KSgFvfyfP{$z8tM7qM@quz@4^%eO2WKf1jOo3M;Tv$D<`gNgP
zlPh4I*sazpNCzXbalWsJY~0?ET)f`pwU~b@`q-9su^w7)V9I>V_^-HtEq5^FKc@ZH
z-M|)mnDU>0C-`n)WM3SQvExO9cP+<d6hMQCg(o;k>%C<UY`_kH8o5iqcm`3o{7X`d
z0*K#^q3hsJ#qG$(DQaza@zYjdBwClkXh81xxU2<+{!>NH#PPocJ-;4MjQ>@4X-jMC
z*MuLzueVn)&J4tq{S!TL%p_o=lFl3#Xz@uZ{;=Ez2%TjXta*X2mrt^YSP`+f`Gr;#
zZ+U$wB6H9C+mt<Zt2XDdr<<d<{q0^MYZ}sSto(Y44rdPz@h5RmrX`06v8vqbq2}XG
zR&wYy`;I};GA3@oEv}NMd-3oNWDl}m=kJ1)8H5}2wK;bobqyk3$PMk5m-Ge31&<fT
z&aM9phprr7@7uK~scA6;cJHv%*Tqs+mfucv6%`G2SZp;CarwtaXVxpYkZK)tjv`_F
zneT_#%fyu4d8?`BQZrgI8H6OGmKKC$eciI#q^9{vLw=JCmRis(HLJA>A|@PAI8gm7
zy7lUAYx@Xze_btDUjsG>K8$$G&*!t1+Lp)HRht`V^I5*7ErekSAId(q8aV}oR)wnd
z{czfYaUet{8uuJ0-!cRgv;$Z?8&m4-*5FNCLiL2ky)L;+dU1ZRhPgTks`YY>jCJwO
z1)_9S4TerbT``rZm#QrJe4ecV4_sC(o;t{8McEMBuE_j3_2ZtKU$JnG6MT5@WP51{
z7i|q3u-n)A$X;i++)eT5THA82G7Q$;ln2W7lieA|?HLW}oQL?#Y)MQ~OF!4b+AfHf
ztQV)07WdZ?(LRoJ3QCqB;jrLWjk?dr^Hr2UWR>cQI-+8^<CNb+q6=ec2vMiNaNl=6
z+SSDFfF${A4ROH{e36|>?_bY;FfL0bp87HHup?K$z;-j)&2vgR(90U_;&sM*Q4l@L
zqi4Pwy-SVKpseIfZ0fV@Du<>#rKpZTwIOj0#h>>fDZ?J0o#gm=0rwj;#0DA2J~V`o
zimUNsTNK80y=&Hw$=$mv9nMu+mGRGlZmyZYF3rp7W;gI5``fh3Kd2jIhn8;l3?tDD
zypMCnJS~rI>>?o=%+nY9UyPE{4yLG5pXgjI#1;U-O>h^9+G!Dhi}+d=pIOx2y&b|-
zqeVvWPMUPyc0nwKBH{4DmcsCP(io{#%yI%@D?s4}QJaWF5ca@}i*!6!^1%Ly3h=uL
zE*R^qAO<{T_U%PD*f5FJEH&DRTZkE6E)wX1TTO(V&^ehvZu(Pqg8MB~Ibm8Ml8t)J
zL~7&vson7a>_n01j9Pg)X0BW_mOK^yYUF6wi7A!FYs^)V8ES95=NSY*6$C*0HPUWk
zP;<vR-%^nT9xl_zV}+*m%nEs4dY-`d;~dAj99Q|JZhANR(`Ip?{n=5E%c(dP=IrXU
z#%)Go>jY!wYP23dC$9QHoR>}-<rRdJqmcf_{w6a{4s`RgoNY|Q5a9M2bA($U7GUw~
z>{_*v1xg-i9QjqJxp5;HtM^248AiFu%>Yespo3JiBCMABXk(_2?c)$({+*`56tiSR
z(laJRoAiSgG8tg&!)&2neYctIlLWR_dY6_XlgUEW62<t>lhU9V6OodxUp<5)wX*Bm
zF(NC+HHE>R9uxbv=^QyTb{JIIA>U0Su@<`P(#$awh}E*|2N_LbXPJ|*aB^d3bZ|`V
zcKC3IbR<%?wj1nP5>Z9R<0LlO)rMjN3L-}~SX=Vp?Ty+GQ_zFLt>Am|p6F`O?x7Bw
zFzt#=%K4*62x%#`o_`tP^#M@s1J_j@j$0!e`#seHF2IcA!TzO4Mf`wZUUg9`tb|Fn
zvXt$BSurFpgOpoHi+c2FBTN<mp`1_-3Qg8fKf5%JGm-P7oNB>FSBHTOUZKe}GL`We
zg(zCaH3R@sx8O@!LxwO^n&$%k$@1nF#ayt;QIvM+Y`6996_y$Ql#YpxTu)bUelJso
zYC|U2paBWu$E(^fClCjahr<)m9~rT#4C7tgzrGOdBx}Z!lvFxlb`_&=8T?$d9siiM
zyq&y?U5mIL*WF(IS%okdHUO?LHMc)roOyq=p(^Ub_i)?5`08LB6)v3RMvBOst(O%z
zx9!JOptSr>uN9tqhBs=~XYvT=O1O7p3OKQ~2xcFq*)&f6%B@6C&nY}{Oo`N_8$lRe
z67bnki8n`xXVW$@xE@*vnl1zIzCyJRB{yNO@B$+}GX{$G4P<_>v;>N3jO{~oJPs~#
zDm9RHX%9^*O>;3>I#svD%#Hc#lm#~yNjrG^BGTh~uLF#CK`G{Ae5bF1yi|kH;d*DM
ziq~~(m1$(8KvR5Cq2ttU;=>Vwd|3ZB^wHK@NF_l*U5lj&ZSrTdj5J;+A>$9&4>2GH
zSS>0owl|QN(KUY4g@WdOhe_3wf5)coH9j3r%{!=#g>4PzVr$7MKvd(YAdb!U@`CnB
zM|m??b-m{1mZm70>>Wrr(Y73@m-iTMJ&^|KvJ&dfK%mbqz2sv}^SFEFOdtp%+Q<ru
zsPgXEQE@?DzHEj$Gb~FSUThs)RfF`T%w^929kD{~&2SGpaXh~)_Tnv}!Uz^&FDuhq
z>tdmFE56*uJ%9;hDfgA}LE=#<f8zc}8b6L&D8XJHrzw5+7=p^A5$Pu>#ULK%*=J!x
z!R_KdOzSTrc0}M~$|H_qYGZv3EaIZ-G#W3xBUb6&*@C07IJkZ(v0H{iz$tMhptj-e
ziw_E34pyW3!<eNrS{!gK#$&CpLR<N$(ka!*GtNaIRmTLB;S{*b3ClFe$;J_wNe|2Y
z2yU*7h9`8|o7al>G9XcNuX>T)Cw_kwcV>lwNFd+*5>A#_JES?CNj@7^&Nv!GkQR$S
zG>0ZP!xD%rWq$S|7Vj<>sD>d>(Z)d^s9fzHh)|SSdJLgRGhK=yz&SeqyD4?rE*AgJ
z+x9;w6}FCm()<&JKnw)|Xw2&fV$t({3ZYnkyb3YTnQ;t(2)awXmCpY+9~6U=IRJTP
zhd9L=JDp#!pse!iS{fg8O-=+w%9S*Jp8(}ep2E?H)a(mDk^oe~DS<J?vHvvj{3<Mm
zf`mj}r$icM1{z%o9$*PUg#rmKzg4BV9*cBUu()Dp!_bV*x|tZKNW(%cGz4Zv0e`FA
zUY)$UA~a+ebz*oC&>F#ta=<dv?lM+fM`MVcx=3v%1mbk|bo<4c7Y%bzJwL&nvZPS8
zK-+={{q4?d?F+~6gLZ#%w{z!%YzGhn3i@l}I&VuaK-zP%o`F4iXUu;Z_{gtXo)lQA
zRY(}lKmZdpN5X10Kk<IpV)a#akWCBr*ZB0xVz%P)Dyqerh1FDtPD|Fzpl|@Ne`ro*
zen}o|JbYXDOPz3YK2rX1)~c(9cQ~yD#Sj@_(B&yXul;Xq^%)9TvIAAmCih0OLq|2#
zGM8D0bujcdVcz{6df<!CIB2lAlL<*IoedSjTnLA@TGRoT+G3l-GY*kzoy%of1BHIn
zz5f!4ugSqb(VJVF-iE{wsJ#x#4Bd6*2olg8HZ_&8$Larzv3HEoHR`%G%e!pbwr$(C
zZQHhO+qP}jF5C95s;)1c(>dMeP3KLX<XJ!Oe`~BU=NQ*eu8BoPY!o#|7L*lGQ8e2=
zjO={^hFqW*STgxYA~RtVSDV22jiy=@U$a9QFs#uoT&~!<8^{gE+MFD+@}m8?+WkT3
z)Z69d{zHMqykzD8_i40uACoMI|CBN#V;0MXKY#E6UFh3J<)ILcaRB}V--i63^S52F
zE2flClomPc3;f2wzfRCD$;rGy1+kF@WSAH0TQ6n|%K?V-k7IMr*bj)xtqpL`+_H*@
zUG|wodI=}re?GO2EB=$<W&Tgda7Kpz)@7Em!|p)%9j87m3<@$=sgL|=MAf&T0rrY2
z%32G<k3dr&=5G{tP4del%*wWKT)#?SW|D`!#&B_K$i>r$YsmG@eR{gTxVgW$4l0=V
zWj)&eHi89k&p*(acUX8h0!(72<Mzqq^{rvnXuCNktd5Q!A9bb4_4Y9A9mB-4@!tUb
z-C+mnFgl7&_=P<xO@%ycAA7-N>_Gm;*vC+A9!+ffxWC;?n(B|8$O;UOab#8#5IGuy
z-7>8xL#}vaH50UV>N9*($im$Ce2}%=&o%?q6Wn0RdT13?<8KA55Ou_#OcvsfF&_WH
zhWyJu@p9(#D|`FeXU(KYY^({^bAnAo0<A|Or=6s)vj=KPsDa$Ht+YY26iYeA?Rd{A
zuAJ(;$L8&F^FAbhK|~}>Gq|kr-pqGM?fg7iB1E1iZ#8H{P9BmcJ|3iTBJ=8RfpLox
zNH{5}oiI@d>#9!5_!<4wfYUt_+Zai<LA*v9@;!BA>-z9<cqMe~OlZ&}EdXzVaB{|w
z9hbnLinFv0QCak$$%GDo1jO9l<r5YUjIxSamR`tzZTe`E=9mwRX^qCi%j)Ch*8O_y
z7sHl^$S>{l?uvv1SJyxz6-Mzd)%#)f>G?3~vTOoAJAO$qRQn|n3ZT{)z6{P#4Q{`*
z*bdV`P4>V&x|+GDW<NFmkd0Wlm?7kmFPH$eF8a5l`JRmu6{_M>vLIJAZj?<Ct3!<9
zj)VAkq$39dJHZ~loD+J~!N;LW07bxZM{9u4;-2sjB3Z88FTi!K2o(%BiEOBBPZ3#%
zD}XnmP1S^IZQi?pne0#>?4C2-%2-b=XuC4(rzGIKR*@SN`iyrSe>x@#Pq;RFM;#|q
zDm&8EBog0wq%>;n+`4_Fpws<f$0`E~Ndf+M9b`R)ScyvW40}xSoTca6ky~TlAT^ZF
zY$)@XC%9-v6qXOHYjF4v`OEM($wy2E5uq~7oV1r71rs*;k9dGvb<B!HK}G57(i2^5
zhmz)AXUqHamc!7Am~+i2a#LJzeH}i`vG8S5@cNAH=i9~{AsHAD2U|V3&ThVgMy%+5
zvN`tKXIlPXqJ<bMRy3k4Tr?Fa+2;5>NR4=XVq;UAY6C}10@k}257N>NT*m#TXe;gC
z*xIK-ffR>A$P&*Yn)yy1ja4bDDxqZY6_^2EaCA~mS4^1O2v@0@#o=*$+BB`^Cf7JY
zvUnx+kozz5;aVIP4aqR|{yi%9n-b118p8N)>GpiZF7bf^Y=X3`PtGqfnTP~w^(Es}
zL`Giz@4q<MNBjzl$c_Byw^kF{2jmmccK(?zJi||x7{;Bo`zAwO4up}s7LPux%rpR!
zj@p{ft%AVfAVC~itL^PVtksx|LL<%!mscy->Mb!c2~$*f&NJ+S;CPZgman9;TT#a1
zXfA$<&2QgCCbuhj(F(I^QeAS>0J2ai4X3F*$HvB!r4Oez)3LlSpS)^?ovoMr1*{hr
z-GDNi;omoLf~@sk<5_A*54IF<_Ud34ix=w6Snm;bY!>nb+!~wy^^LS{fd7^8RmC&R
z5PU)Pv3o>|Ud+6<Xtz(iSdlj_abJN6M?&nZcXFz6myAQ4_21;=H18+y+S2YrUS%1d
z+C<<1DhKDGXil&R4Z#UnN!&!rfGvaBW8<hKF)3U-Rz|K<WOSvZXeIZjMzlh!j-wc5
z#z})tDuOc|A?{c{B8<jU;3*soKV#6p;!>xsp6YC&inU-JzD`hXv6m`@6U9K<<+?WQ
z*M*};RIlKtoGwj@#1xR(V^GbWR<d>ZD_h{srK-27n`IWywgUL1^+zx8_HF*(pCKYr
zYf~cF?Wg<w;`h{LuOasH_EKO+TEmM91tNGJSbs$ShBQLt;{9PMQ}YiJ&1c%zm$I3y
z?A&yLfD($Hw$!sFG=vs!8Kg@OPiF^MzQZIVq#6)=wUGEW@ckDUAzt&@K}tvgmDsb-
zh=h=wGagu_`KCi;t6|TF@IO3<Wv5W0LOb&j%vjOx#>8S4<TC7-`fR@8Ka4fg1k{lK
z1lQ`xvSY>t4xn);l7x#Pzxhj};Kbs^B^e$c^tmqLZQ8&5pv7X@Z<gre&3*U-ivTAQ
zJ)PDKiVa^_CJKc$4U9*wT2UWo^2Ut7!z>ontX`5CK*Sx6g>xg%PaCoUQH;tZzPYNu
z@!5~Dmx=X8S4x%rMAs^>v09f&0Hv8qBafEc9{?I>wTrZBxw@&|k-#XZw1V(t5&v$G
zLL5>ACJ4_6@~NwB{v9m~kP&&XMF$KGEOzO|tDw?e{8vZte&cCHXvp()aJBvoZ>fkV
zJ}2@o@I!r=PNMZs9O%7NB@u$2#S8T?T?TH%8_7gMxLzXm|2iNKzh?D<-fNOS8)SLn
zJnQGTd{hPYSk&~uc+r(;h&00d&*-QV)@xUEmi0=<U$-6y$}61Dzu`@Kovwl{T3uf+
zp&Lt!fdE}vrU}$jW6YswEs3Ab-q=6${4_=nYh2*~b9*#jY^G6e85t`*ZBibnhjgH$
z?4Qdt6bA|#N(n9#QZhU%?6(K>Klaag8{r<(y)b81UIJ4X4vJB%AsYZaNDlt9uq#Nk
zsfwdIaaR9Z_OFj3EPJZ{@n;0}0fM#jbsdY`<d?nnoQXu^H&=qgQBWi<3OX676mHA4
zqQd6p@t_VWo}5n;ffnl0mzsp_?PV0~N+iF}!b910anUw5RWYYCxMQr{xG}ZavY<wd
z^KlC3_!MW1S=JE1C&|<M^LMDj&@HGkdi?j;M>KO@wdjwYQ-P>@XCk-l)viO#8AB?e
zm{hV`rm<@dYm4yPCfDm=|I^+r+eyf<K3QALAgzqBG&54|{xXyzsn2FpU(&o-mf<j;
zA}(FBmm=WfA(DCxhR$X@s}r+_L-nDzywr5ZUZEz_(yPBlc&T%6acV!mdh@~SKZ>5;
zp3VPnT>&FA{eM?iFr&$lM8twTd#dK#4pOv|lv_)ASlQs#qhj1|jEpdXh`>KQcY1D(
z<KuCEh3I~zaOpl)=Ej=$o&NT8e*W<B^iAr(I88}!==sGB$g}r^#tvi2f&oxUB3<Vv
z+n3FbmdT6ufq$yFI5ux5wR`c8dSDScAA`RL;M6RLpD0n3cm^S|m%|6-i`C*Pb=6AK
zuI1P_%<5zBeQgad*PZa?OJ+kvGQ?Q2O9IPSSgut0mjo>bk5~~IJBs!(uP`~UV15<>
z^a0OczKDba>+RQ6?-9QU-A?(ZqaCwlm+J<7mI!q@>}|R!cWlriAk}#@5-8g=HnekT
zkm7=Rm?9_gep#Fk;#!EPxiL$58Z%owWtu(5wadPR8w*9qs&hlYLDy9Hy$PUB^O*8o
zcWy5JDCYY|i>WEer8Eh`q$ZtR=tI5Y=!tECG4}h(LWNEcUN{7+d`A9JKG22do#SlL
zsrpIzjeV#Ba6Ee4WJQ7ioV<{X8~})wfYc(SnZ3S{t$>h303=^;#7^W5*EVe!IDqV-
zNhV@7THcL}K$L!1q}QcDL^$1<A5zjCoj4xpZ_VEjVIUI#GGg~`<+O`Td|&NX^MHl*
zKw8|<ERvl&-W}tQ8YDTr(X7S(`EC-skvo4PT!WN0IE<-`I-#eaL+J>ESHo|Uk5Fb;
znLfR4JEk(~xvTR)WIBYA$MDXRoXNB3>sK~Ls8|jfW>(VagpoV8_QI>}xm}xWmVIpC
zFdilV35?0aAd&>BlA2Px>-KO+4e!R1<fGSQC25Haxd(knAc@F3R~C$vfxKdUL5vX)
zw-(JQl`i5R!owdzh`u*glp5#yNzQlpGajqvWx5ouV45LbJ0?O(JDfJ|n>G@OD5O0t
zT|HaQpz%MXka8@Flr0M_51fT{a@&4*dcQ5Mn9hVSPG_aN-`)>!@W@M}2~x(e_8*_G
z=Z{NqU$qh(UvA)Fz(DP*k}cr#(h5%{@EPz<M@^K?x`xaz3}8{2WTFXjk9q!6rblIv
zJkB;GBedJ+_F6HnAZgI#$XBY>i>0@v65@3G0)!edJUqoD^p5enr1ZCC695D^^fJM6
z#8%5w<O}(Z$;BsJXg9~6#9h!$I#q)-fUq}X%dSt>|7sg|dD5M8YjrxmzkjS>fX6&H
z`aioaY2@io+uk6Md!K3vt7!~YyC<=d_C(#AGk>Y!%i=elWT9eI8489Zp?XdX8!f9<
zyI)D#&>JTuB$9~^4cCbG+fDQ*+o#Qmg8?Nen~W4)Dc+y-L>b7}k$kK?`0BC-;payR
z&Im^T4m!omYLZYXn@{)1-<n0KnslkQH*Csep9tDbBIu@M)TJD#nx%T<5KprP<FO~i
zH+9h|STE^tD8&LeQPp;i{}}S|0gPa5^w|<U*koZUg<7XxsU3qD$S9=~!bH{_lYIVI
z`B>e8_p3UJm!QA8Rh2*=k@5waYQYXwW*^z0QNt5}_>i<|MwjK1Y!dkKDwpy9$i%y2
zO>JK1+9`$a9%bu@MRi&&5sNepR;M?MnWD}J?AD9o&%=Ke7L`#Va2GQcGCOJCGW4X$
z*bJo$GTDvq^xlw8c-rsd#+nV0OP;xjCE3VQ>M|#L=^lvJjt$7iJu*4BnxXgf!Jbb<
zqb%7)yHLbs!`5(>7@Oy6;iv&+P|`#mU1&#kl{EgXiO<F<t_=_J?cp$GFL`8F?Ri@^
zQ93&UNk?*2e^y)}qs=-ktZRTcb6aOvCnvXP1x^;kA=W+}eAgKesG-W(UTaApuG<3y
zeE=S2=q!*{+kt9I%4$%|-ja3Y3TuK0?)XR(+Ef7u3+9~byiw-8=#+5PZ##=d3pCm(
z+@9k;V=l>8hRpI6&R=g?uDTX9*@)?iXGGazc_hi^B>AE%j@D&#7)5x8Dx-QYJEN>y
zbJ9dy-JP7~CIW4ANV7k5ke!zNP?X@!tr*2h5#wua*=zTjk(tlAwpO6mQ_Mw>gUc_d
zlmpT$f>5)dRGb0#*EE|rd;LHi(46SaXpHsg;rGKzUQKN#o?5HSweH<3<z<Ja@25V?
z{p5wc^I;E>7?s#zG`v+`6Jml03g<ztjcGuDpsz$+)+dC=8|`g&p2)#o|FQGw+7|Uk
zmF^Na8`;Pt*zI72S7+d!kcRDNoi&@=pwz8g46eEON9hv6JU0T(_4&P1Hsc^{R9BZA
z95D&`i?{90Ff1?;&5T|CBjS*l)ZB8znr9<~*Qw%4_?S-8V0n+$!+nC|B4PXoBg%M2
zNm|vn$5{uRJo(z3^hS@<tF>C|#0`OdlfGhQ9+(2vn^t9JwK^7zoK`yqthM`ur*%G?
z5*jWcO0u_Nah9-5e8R(pt^NtyYRV6*#xI*C8M-bdy9+S^VBv{tL=bu{hSe$UST*;B
zLgH`5s0+4+%m|*DVuBKKsn-T9?y-Lyw1|n4SB}w%3OVVSf>=Ob!Qv(ntNbM7gv8)m
zKW$CEKoc2(#xWv>&hbxWSjypF|8A)Wc*m})o9vYVV-#f^%lH#2A{p~1SQzgX`|Cn_
z7c=k!ZEKV)vCDWEOeJmA=TULEQF1-S>cVJY1EH`TsKdidABdNKa|~05geZ9B5xU=c
zkeMA1g1ZbUbaZFMU|_1(IRD(QXaH(49GU!uXZDc|VUkR$q+it_7>qk*bU6uoCkWxn
ztqU-UwoIWjuMZx5wda|`Si4`aT2DaPnpeNVR=L$7m0G^>T2DqwBH=w^UII<!Vx)H%
zk{6ohY9AIIgJtb#FV5UlEB;dGYk+}qHuR{Z?Xqc^r`0VV|0&6+mtkz6{_zUI3+bZx
z3G1fkhct+#wiz2=g#u1yMW+Y8jPo++SIeA-8Pu_ul?vDCwtWfGLytF4Z#W^Bk)(Gd
z3X1N1fda-Se*t-v+3(m!SxyX4)0H~AcY{;y%>M~<gTwYoNy~iE#j_<<i@Z>r64e+6
z9$VU>it<&Z$la`Tw^U41S9U4+ilv~)ti^H$@iJsaYsgCQ8)&`ADa9h`jJc>>tNtPm
zk=k!}>?nbCp~yaI!x;4I4X2mV2VQJ=E}FWkqfo&k2YOFZMbB?EE=aJ(+^yau)7Frf
zJ0X>LW$m70I$<3u6eehM&&0|%3a*xuvVls%vegM_W)7)+MGoES(q;Qnl=$fAb}0{~
zSDGP-G6$Z86>ZA--QRjc<4Pwc{*Q8v`9IWU(6e#;S3}vA|EbGhh3>tryK_JjX+|Wu
z00e=$l4b(*q<&d_*(?mx_O$%*sW=sHq?wYyEaoR=vQi>^zb8;c)caO{d$`|!c)9aV
zvSgSpq_6M(>IL^4IQJVkFBlyGA81CN+A2Fvd(93ioeSapMWUY@owb$Nd3%tBkU$C`
z2j)WFKg^N~$Px1-2m_1nX7lp*$#}X-9kRr<Z94J|vH08_x3%Trx+A{4hq@{$#zVJg
zH7i6P{XUXv#k1p0#SW^^>`Wq}9uy11?NT7*c=ArOP=?+Wm<Af~9qx$^|31Z0C8Epx
zQ>ApgJ9DQMNVO`4YGz#o>!_u}C8pB_P!_Z2R=;;ad}u**-ZIW7K?3i|FGm7r^udR)
zIpDfC0(Mf+B5V?+G4?jimTl5;w?du;s0j&S62YTVp!HkGdS1%oy8*6_r?BFbbyK^q
z-5YR5V2?x2?840z?Re5vDPOQV_S+@w#|7H<ScZ+Cj0c*BLZ+rZfYuYR!Dr&Aja$xr
zXtGw->4UR05|p`MhN<(O6g?_5T#?5D=8>jYe4EulI^9ccJIGvWklYZ_&~j#_1G@U0
z^sVC%_wLV}gHmsfCHRH|YDHpZXYU0du7erOxEcrpjKak5xMKrfPhE0@ug~HCQRv65
z=)s6)N<d%XMTrhea0k)3$2?9y+9jN6%U-=ZZ&s};mV1ELa|kw8A)}6qk($aPH_(8W
zi{B3e&p2_~F^;zg=*idH%A)p2m{T-j_jU-TtV}1E(6Cgib9fMupYG`TIq)GNZvm9U
zwp00_2!~D5+!v(CwQ`ZWXC5L+(I+%^mUQ5lz3tXJ=wkR`7IH6i+ON3FB^>HR$Wx7A
zB3(sh)v!{E8?(yT>IJ_h=^%!1ug*0L4|!NEnzyPUPl1B?)ZA4oVE#FgRHi?0`)WGU
zEo=ti-fh;wcEaSIPqC92+@XcUX`+OD1ea^+&Jcu<w4hZiFJmUNo~;OAPG+vi4@STW
z9P3+xH$t0?5!JK~Q-_2Q68hzBaQ-cV7h+8$5~7{0PgpJlFXLp5A4`{E<zJ#&YkW7L
z##pgX<>xxeYcN)1k}274HRiG=mw~j<WcBm)^c0)$$0IQ4=ev5FU_ex5ggHS*jLqK8
z0UFdOjow7C&Wbu<Mkv|G4Ihn3F$NnaO(cc)oTQ-Gr6g2I6b3r83Eb+P!yrSzWbI4j
z8AUKMPforrXOQLRsYY^@Yqw!F=8tGKIm7|0MiPT1>Iq+e2m=1IqHK-f2KIQ629qVx
zSG_U)e$2fyLdOQmbvFfjE7J!egcPpi3z13D1Nm!zu3DdbHR7_Ir)>?!d^MUfWsN&Z
z*1Zp)M<*`I1t>IraHEq{*+H?*UhtqZ30!oCHy)Eg_02%?LyWZ#wSvUQ#Gxj3g7ueW
zY|6gB;A9P|rzM)Ke;*k_v1Q>C(9OH?m2h{3qEh^%bmc1crhK2JPPIRiTTeDnkNFne
zN{tG>XOy7VxPuUt@J~55o=G-H3k*~f6n>PqNyszQT*V7(r1}`g0gd|@hzsJ{4yz+;
zaztY@74iz1>hY1&F3<~9WL=n$R%cPs24E{AS{K(3RU9Bz=Lah185*4vxN7g!2S3m3
zLeTQ=CeXQA!=j2)?x!%(WCa<Q6^UdL{F#vD0+Y-G_D4(w_KJ*IA{es)w%3;_W)-Sr
zS`~h<V^O;{HkxIk7_IKWuF;t#TMnHVUK9~H)aJSXFpc!FRUz6`C=*5%*Z}N5z48;R
zxBiit9!*u?TrM)BE<$EP{Y-H-&&-_awzyD8Dy=Z#D*ySHxZk!c3wgTbSRn^tMqET#
zk#Jq^OBNNGI)-4nObibv!2x*!6jSL+ql|7bFmp|3E`iaXZ74($e`nW$ZN?Ac5gAp~
zN=aSKg2hn=4RGw@CW^6Vu;oSxYHA{DKQYN{eT0oAAS+X`2aRQ0!Q<eR!cYPe(@HVk
zImoWG44Sb)S^?FFlURwrRa6f7d+5a9BC7bm4_Q@D6+8te0KGC5eO*<Mp{U769^18)
z;NxCxmqToxS#VTC34<gG>s`_N(aL6%i}>1)N^gTX7sLXaTC$6k#3dE6jJmgWQOxJz
zz*Aj;B@}Y_tnBA?+e9snxv3r6RXP8zksx=E1Cpqo`r008GuK`XhGX@@#E`@>M2pT)
z4%%$W9<!(lvGtpA$}Bo3x=rl<jM{L`kCsgC`>#>-QMjoYBRS4H4`77}VT@J&sht<r
zajCzcHRy&`v^o*P&SnPo$=YTb`u<y2k{fRs;}k>aS*?H8kIzhEzvVPK^`@MAW!IDD
z{7^~{iF!x-We}t8h1_CS!caNLP~?ER651&%cmM5g8~ayPyz*k5!d)-6liaMOt+6AD
zQ8+JBgvt}p0lUy2{)N=5Wj2TpQN~eUu;%xRF1*{(Ba@2!Gg*{{e0Jm|yu4{n!(8aC
ziN{P;2I-hcQhp*oPGrM~v@|QlgLN!=8;_;oZHNP!*@Hv)uT}Bv>jBlbqqBxuxByS0
zOXFSlm!mDK&b2-RDgORh_h#xOOi0Bwz!qBA+0)W9(aw2SD!PeI)mmc@iKBAp8_lS-
z8)VT7ueJj`(o%FVuzI`Rp&9oshIS)k%$w<~{(GqrgDU>q{rgf%QR?F^?sKcvCW)cM
z32PfcwA+<jKfx4rsFb3156ntee>-An;y<orTq>=?ig9mP_o%6Lh?rZoM|X+!bRMi<
zIDJ-4Ea|ffyYSh3Nv)xCOAF=g=J^LjqkxeAAJxYHj>Z08WEO`1(i~dRoT~i|7;f*@
z&BFx~gbktD8|-aHKsm4N-9*+#l2%b?{^&W2!%`$9QojSTX^b~gL_TsdWn*Z<?t*dr
zdvWykdv{-DXaD7~(0BXcAoIjORS?5yGO?EguUNYO@cw(JBvJ2-buLKOTsl7MD(m9#
zo}XNT7GXk;p>%oMlV1ogMUZjopVwIgecHZX&)98y>isQrMv?X27}C4#>k~IsjnqF}
zbHDWS@%4LuEop?+M8^WN%9dAz@$>VGfRo2C&_M?fmS_L>X7={+Al7GDu(BIcOcES{
z_jEnhR)F<BI0>DYOi{y9sz9o45RF&8BFLqd98oMxT!O9`PY{AWWOVLE_Ky`fB~xU7
z;dvPmRI!QxE8kA_$Nj}J!Okuu(3hZn+_lvSlgLoK@7{*ET_~D`1PBbFIfo))-blIF
z8G=S-9;F#=oF5+ZNm)^d9jv5vMx;r<_Jq+AHXpK$^C$FSD^_v1t!O7v*pBG95tjCX
zRYK2H*w$a|l7k0@GKEgOJ6Vrlt_oS*kXz1VqH0-hJQehiCw?%2wk~FNH7CZm!ahqw
zfQt^mi)c9>K`2l&NDY;hM~`Jmphs8GoHVIgsT%+NJ3=CT)+TkqSRU%q%)BrheK*%2
zme9Js^}e?J)+xIDCU7Q(hJo$D!W*F@8WjQ&aqeHNbwepH3!=(SGs4ckzBl|H&!cN{
z>0`sEP3=7RQ*BW)%*oq>cbLyv?kwBBytW(0GIx!`X7Q%ze9fK0^0B=6d2T^%=%UUA
zqito-6_H!zfbicev+{xhW+;>uC2{71p!wXoYaHq-gMoA6j*^sn3yo^BN))al<Lvit
z#q1zteV6v=vGHW??NWDQ<$T}J11w3tumdC74E~v@YZdn7_5lUF>#xx8gxyLr^umHC
zD|djaXvtS}egs~M4ZlD@Oi8{rvMZH;D7XzN3P$pS69{>%LPQa~yUrYk3=zA+W-4E>
z<~rMTequI_q!9U%U^nC>_C}kGQn{{23Gq!PvxB8tj~0nqjW~;sPBTZPE<la7vN}EU
z$MzieJl_2Cbx^D?9E&j?1_yiN5rVp1cNglftxNhf;-wm9!paqRqpwiwP^E}AU725o
z(#G&rU+t~)yjSx&Paw%RF3X0C+a29@t@EhHb`$?}_5Luo8fM5Ie_eBT)oQWm?V8}M
zst}d5wSpPzTvcDX>l|SqL6OoRUp8a9XPv#e2W*+#Qec~I(P(Z2=1Z)&MV8hNr_Mf_
ziPwN{RfRMFQ94ZHF@_wPt^{#{m*qBiwI!c8c-wWmMn7JGwppixlVfum4OvrcQ#Y+9
zttWH8_0R4y%6cy3t}l&xO1WVX8=s)%Nq?Gyb|Zo`C=Y0%A616}c7v1K7Fxm;ABi3v
z<Rk1XwE|05H3||)>zGtPC5qf)61=X=drWAG?rICw0mii||CUZVKf2aR00aj`U}aLk
z@N&V8cUNd{W0qzr__z94rAK1c?RT!)TYlCy%XM5!#S)Bap8MI^3=22MEQ)J7fGB|k
z*iM%iPR{D61T+X}@g{yU6DhSZ9MR%_t~l0ia*$;NV)ZKzm#JhWKyB-rOauhoRGeXD
z{+)%8VrbWz@b1-htZVFYOUzqy%jVx7n{js0LZvccuHkm$qBoMKGEjywB#G)1<&<H_
zRXtPnY}RqPIxEkWtla-92b|?{+enKO>O`sXxgF4lmz-rT^kBn%9Psf*rRL6^^B)Y<
zJ9^~Q3vkB_Z`GqXnEj1=bvH={?}21)eyyDNzj3Z2dQ|ZoWy`O4LRsamE&opWy5prf
z*IGWDzE8gUwRd-|?`PTvN$s|8-s&9Fd*!Oc+h`w0AhBAICd+qNlXMJbD#HnIx?SyL
z*?4!_`Et+iNwnQFR5Mk>u*fi38Y(S-CSox-xDQafwzM!9(G$T|Gpl@3w`OH~Q?I-)
z=4>isVy#u!s!E@qlAFW?FlnC8jGcQzFXYHuAiXH*ns#Ci+$w$P=2?vd*RmVmcia-C
zdx^Unx(v_(*PJ%ax?Zd|VTac1&K-$b&APV`rqo72+77K*KB~CozJ3u#XtLkZMM+EO
z7?r9zcukWPH<ry;hLRBzwGk7h9(k~q+&|h?odFL}Pp)--EKOqCZy2u~Yz><oXI{<|
zUG1OBd10#G_cJtAj$|asWpG?nVfw)Znk7p^)nD8+tgKKSI`{F8$bgon0T!Y!;Xgp`
zL{CN~TR8U%iX7%<oip2_TtjKQhd4x<sZHw`y(wg$tlKB6bK4?wy??pjbmzsz2y<np
zWgXw1O2LlO&vUHa>o*zMh(R)Q2WMsU5Sn{1v<7+Et`iN8xodh`U)21S^D=v1$~+f4
zYBu@IwvO?n4tLWg$9ww%C?4<|1-7MlV{|?OHRV4o!Hhyy3-lbA51&<V*O<2>ts1t@
z*Lg3^>V8B{Eq(bd3!1kk;Ka$C(;a@&Q>A$0q2eq$kViCV_`}4|X@At}wCrbWwXIy6
z@GRK!r*rx{<M8nFgfUDn)*G~RhLY|nsIQMEfK=9<a+g7vGW)b@OAEYo?;cw{AIfo+
z4ldW7hcey?qQ@jl=n1!^j$K30aIiF$hn=fvTQVf^33>bHI;ES};_y1&d2uW<Z?K6D
zE2yHi{xH<FYie>CO_?1`<rJ15mCQ~Xdmq$y<TcE%bqd1Xq>gm8&f7Kq^4^&y89h1k
zSwQX6h4UZ(?IKfk{e3ngt*QXC>kQZ<_`(q*pIu<p_bos|9+KLzXDkfpH`Yq9<3>M+
zr6MQzfa7SWk^WB-go*8cGd%u(a-q!sPcD>${=alXv~29K+Yr8Q^aS$t7UA|C$M@r@
zfq-4L%fJwa>$eMQ2+%z_HY81Q>^c9~`ugPU-#o}J=y;~8{+X(m)uxT}X3tJD6OZ(U
zXPkxi^%5djR;1T0PF)|K?~4&_29-2a+}D&WeN&d&h)?=$%x3FFFFd+~&EZ+pbe^t#
z$+*)wQH(+uS$vI*T!+(>&l1Eb$Y*t;A%#{@+!bQT)y84xI=t72GQNv+ZoZ0%t94Cv
zp)K&GX}XvAs1rGis&a2i-xC~(Pl6zJY0icrU1*~r6|N9MHWR$)3w1l&x-<v1aVpfD
zB1K-rO;0qaENXZS^{LE3VVqVbi}P@b_Awi%H@*&7;7X&!aWyW)vh0b-RZfhy80l|V
zHFRlznWD{157lH~fNPluCsK>r09VaNjY6oK0*RjucdZp!eu3)8ZBa;!J%O0odF2l@
ztcIy)PyFl<cv1XV2$v>QD@bCScr8+(r%*np2T3}h*JPHsz-$r=I=J}DopmLc6B0)n
zcYu`SGtrVReIn2GxN)S;l2~A)A^C-w%wnd;puZ4R6*&nQxp~<dRbGPZ!mNxe*P<s{
zbAcGSfHmHS${>#s>K}+)kvW;JI?+p4C=lTSE*j#bVTnP+A&1gUSqG)<%wMa@7_Hun
zZN*jzBI{X>kmMwRg36kItyI~x(Bd*wPl<}T%CetCU-zLCE%l2g>OoM}9gBIR6{%b=
ze4*B65WXfW>5Z17{JIeNa{e1v!lP6<m{NJTa_0^&^;KJLO!E5786|zZ;I`{x_wB`T
znX9cmowj``Cnn|daesBdSL(&7mpzSYhl3r*b@sj5bJpu8^)*zrasN>Bx<h&ff3;i4
zmTp)Nik?5hs>FSki|+i>`DyTNy2Hb?D_<Co&)kmDo(f_m?&llr=LUoTq~+cS2J~Tr
zG~5gWLXG?VM^=u!Gm&Vx8*uo<fxuX3s6BW>>`m&)B!{#6%tnFW$p6Fp?rf90ldR_R
z@Oa}NkpK7hMaZ}Jqd`Y^_lMM451*Fro3nA;DPETzuTS^C7vmSJ*QNc3p}DjvG_<LW
zgQq)pQ$LQ|*x6ID?S`o1aK!4YL+aeg1q<0Shm%-SLb9=MdfYrh#d!(07WSK5+F9A1
zx!O)YY}#4axLLF91$l?gbc-~<4+fvlqB;1XA}(66)RoE!elQF$5S$TB{%gs4`iw8~
z;JAfBzc_Tke$Vm(+ktp?C(;<e2>}d7?kf;<uql2&fP_pFegPZjZoMgVf#}o*nBnDA
zg<;aWga!INJ({=4A^PaYrYScyvs=1K7y4<T)#<|xEfY4*AGK}1pJ%&ZrC<d{w7l6l
zIMz0P?|O|6jZoV_(W^z~MwIGRSQ&wNe1krzCB&L)B9`sf$H(`BXA)!cKvuTxsUebL
zPXq6tx3#$`JY71tG$tXRk;nHDL?Lc9-91OX(MQbmey|In8x7T;d;)SbK5k(j@UI}n
z=|J_re8fINLQH@z(?(3}h^Llzad>WYXpe_$O~+d|$4#f(SsSOv1DT<{+}?ac(fKH7
zQ&o`#w_ht)mfsc*++CRKd8=Q{TpV29oIlzq8p)6-e*U<qZ?&vNzNRYYbzrGl1l*om
zkJ>1ra!z52bWs6Eu_U{a1u5+*hEx(%C90N2N35b1M%dE(B^(hdp*-vsZvtAzeX^dv
zeLow}7hNzwe<>Ux6?V#{R|1~?;s8ki7~s!<SO?Dc)USvD6||+F<Q~PZi`t?Cd6Njd
zw;8O@i0rDuSE!r(E5gr@VD;RH!w1WOKCl+J2>_WBgta$dJ!vE`4^oSo0~!GRB6o<)
zhY^w=1t13<#qGz(L9EZuO@Evj8J`!9e<C*r9n_J`#WyEX7ur(8PaJ!6g$8%ZY#4AG
z0htld<Osn}W~8v9SQROe-$?yNQ@$;ADNRfo`vF)oE?&0)f*v2?*HuN2mw)9LMTPHz
zIDj(mLAqp7!?fTDSU3(7xI|NXJan6S@0zMgO4x47ir+@kIvSvIg-cM_Vd}!8k4`Yc
z3gwH$P9aKZA5DOdcHXiEPU~8=*4K%xdC2s_S6X{t;xy?LEnN*?xP4dhNn(qqfGeX+
zW`(b}`VCt_N@9)$WCJhQfTJt*3iLZiU9f^kvz@7eR5M4aV2)4{jVyk3GzixmX<4kq
zYEiT%HY-||{FNj^G0KdMgy`n(k3%KNs6PDCTm~SpuN1JS0vQ;D|Lo_Dq#GFXCgY3B
zHx+rE0^_6N4@&W7_eJG<Z3r36tuAJ9jb#V#{xmrS4pks)HC&e<vy}j^;;laN&(M_L
z>zOndrU%QiL3fo>>h%u-QDd{k`~xu&h{p@QCBLLzy&;+p0Ri4f#KZQ03#jFGGa*3d
z!3Tl^Ww&4jjKG~P{-PzoAHP8)Wgs5~PEEapAr`9&$1BQ#437x=i82HL*h6SnWCK6$
zQ4CrimEK5{ULj||NFm<9TiZ-vGK2q06#@u!V#d%_1OJULrjR-TOt;lr*1l6{gFMLP
zmFH?vIf49Hm)uy(G;5XojLD!=RIN8HBGf%0BGit92IDazl7Tr^JjzhvAZ>}AggN$f
zU1MkWbzx(a-^(Dsi(YOA%`Dx|-{Yo=4+>%{Rs0<tzsg?|Z+`j}tg$K$X~ikH7yTBt
zn?>+<3OD_R+f)_?$UF168D_928zW{XT7OpAW^~Z%0tW!9&IKt5_KN4P%pO3G)Vr-Q
z(V$BeBrD+!wl81wVWu8K!u!Ehn`=6TF=s1+WGiviQM8Mhx!vawFfa~1ey~OlXamfU
zt>pcP*K&S=K8AdJI_2(!y?f_Z0DWs7&P06EiWe_17;2-kF%{SW;SLGQ;^4?op0q%x
zpugR`w6Jw02IeY&CBlefdj>K}4<!f>xILA^p)t5pI|aL0L8yRtGob}lxI8`>$Q)2<
z<Q*A98U?D*5+XPuz2d83;gXLN3;g<2sSkhtkf#5;NwFBo1xq7W2PNF~Y4;|Re^)aJ
zUo=n9u4&Y}%2vwr8eGcr=qhYi!dlkzGHt59&U#77m}5tFZ>G>53dBD{!Lf7Ff<t#M
zL9{hiv}ef;@dnI-!`71VYnr+pR>3ju?bf}1(aqkyZ}yFLDt1N2VSxKBc=6fsJIMrm
ziQ<0TvMGm@2m(7>#&A#k*3&LwBCvS-OT~@Iecf0RF^>A&@WE%Y(a7)!`}cqsC-l>S
z*k{U%aDlkracir7`Uf?v?Aisacb7!*8i=+B{P*{e_5HoIhpRiohc`su<2&^a-MO9>
z1E&Mxw%bMp$rWj7jway}{?S?&bBTV#uvaX>8f=T}UlsL`OuAUw#{Dx%TMy?Iy%Eep
zDeDHp)+4_E_fJogI}tUmc!bEpNE}P(V4EmO*G)uOSl^#&KI(eUKeO`>5AAMzNn}!Z
zK7l`i;fuCAXI#1fFao-ypuT?+t-t)gP-A}<xj#QMSMVc7xkroB{sajtEcXZz8eqhY
z<1DP%@$ej_XTt#;XzXW<+c_A|gFv!~jcpiGg!Yr)Cpj63PO;$T{y;EKybHk%5m)wB
z6cxyX8~tocjo1P(!wRmR6Sx`|$Q$S20?!D(02px1{6Dy&CGe~{Dh4m&qm8)lU?7y4
z@E`)P<5JJ1HMc+LAc5=sDuSeO?cpN)mg(`756;4Xp0*<cv_J#*mw1i%YJcq!zI73o
zT|F@_DrnXHU2uT_Q;}_JMZqkf<JBM3`O-L&xe!2Cd8l~cF~fm~v;)e#V}uJ%|Jd|d
zMqYBbI1z{6cN(EP_Hc@NawFr1WPJntoj03bKI2zKaw?E`0b2)C=>-G^GVK}Z1_hR&
zP%lR8C`Q+zCG4^L#`7M3n)**do|&Eb|5!9*U}t9fuO^q*8d7opQ&#rNhl8UY8iU5;
z*T-M21%6u9w-b>QNm7EH`QiH?FEJv~+M<Tr|Al|5bYkB=cQ@MjwIpeKZ0Y#<>BchO
zX|Tw#)9dRm#fQ*hSt7GVgOenLmGbLH%jaQF^5$auJOD*Z!Tzt8NS2!y{m?A904dTW
znUjZ2p2@&$WF8~xDT$fc%az;p;4Agn^USDt=F$F`SJ#(A%tR$pe{bd8+^4H6_mfR|
zxe4A}&;%CpQ2Kd!N%0K`8RAj^48L=Bv$eZ@x^zpk;;8y49rzFlLQ*4=4KAq_aK>d7
zLqCfm00;q&4L2v($sSK1pIdG|e9Ig+e(9Ba+H&6ajcpQ6niQdRL^4sr2;i&}mLo#L
z|6W-^Zj+i1psC5%>vNu&A&2+JQdq>(yrSZi9n&X1x+Fd!bX3}fQAl}LCrBGPGX*Hf
zLXP2l@VGT;Np<8lyI!&-`Of5LrbiNWcjq5xm&Y~&afLT&)ftvQ)Wz|x0+U6pS<_p*
z3ziC1ozgbyArMJ_DzI_ikE1ALowLxHc}*sMwCW52Y5gB$2jQtvo5_}=sy)m=mA1Hb
zJ$T911M!p7jeoUUGWF>L``Tdes-GmZuN)Dv0@>uurp`X;LNW|2*!_76gyk=V3wYJW
zxM_t?mN%3T{*LJd<Lc-`O#NQ5MItNivKmk-(|WX^H;EI85Kkj<Q`XFS;LmDLyHcIx
zw__8}hd0MLT1W%tejHGv|N14FHymw0V$Etq6*amf7J;ekNga@?=K*Sl#W>7m)bM-o
zx$N9xm&gbrqyvEVVQh#x3b6epjkRNaIDTa35{&}#d=MaXa2h%8qN@h8H}50&5f1K<
zA?6oNj3vO)BBsuP<^A1eE{^ylX1;_Y9E%H~s9fj+2Y;)F!E0FpnZ0Dv4H>(vr+saK
zn6Yl!`)QJj{gqyn-*?!tHbhp$r5qax>kG6WC*4JZbMg~t^!d_|JcJTd($kGg0AS#a
zd<Y?fN;4ls#pV_ru%6}Jg?GWq-Rl-;J_lB+vMEV(YCB7!npU9-mShCUXw5$0ho)4?
zN~QVjD?J^!%YwN%=BB-*fc*I*L|aG#lm-=IoVtj|-|2nui)tpO$!L*O9XW_N7DK>$
zb>VY0NV2UCL|Qh%xaJ{^1f?2Oe>3b6(UF*uk~Cp-VxVxFt}VKAVqWhHi{K`IMSY#8
z@eM}57(&!(2S2t#)Dgmoi{?+jL)@vro7q6to(AZfHGGPBn~@`<?XmbYfv@41d|4-a
z`0Q3A`32mF?2hg;6>Y03%~D}Zab+YjzeBO{#@<6CnNCduA+Z~t{guVXY_yDNxYN}Q
zi!3W>Q8KW_`~MIEH&G@Pb9xs?$(WA5ObWna2XJZ(WE0{dAEud5yZSaN)~rS<j7zO-
znEM`OFvKyooaAZCQzv(G*^av~!<#C}cku+@C+4XMdl(!tXg|aOLt>!^7tq8fNZF<Y
z$PAPMWC<g7RKdGH5p_WWOH&$R0ob4>z-Am3vZ!Q{D{}4ON?nqgUmutHR2i0k9Yzu!
zs_x*`mNTa{0l`*q8PK)P2v!<CPQ9i&!br5^cbZ5Gr5bU|DD6D3hEPxbXw*yncz?8O
z0;z_Vu)O=mu&EFFbcE%&fNrKTjjzM_JA8X9E2=QIMJdn-fIz_f!K-b>*KSe|fw3^9
z!)0NJR^K{M$cf|Y6tMx%2fzlBZeLoH*A^oETVSgSCpqB@@dh@Wi4{zPH+RXS1!gf5
z+xB8@1!wa|F+zs6C(I0|3e^a1Ij>dOG^Ad^4^*q=l*-jgBXX9SsHh1klis+l1Rw+d
z$#+Mlnbe4N{FXo(bvMXAcVmKlkJ`VX?U_R6iS+67BN<BMqmWniBh}4?=^QQAGnfT=
z4J?pmCfV1m>DW|mL4LV#eO)}PXgE{4$c3FR>3n`qJs}4>&?BJY3WmwXf<zF#+ka_#
zo~;I-tc17KE+*GfvC#x`2l<IHt0zT2VUHtcD_5@wfGCMe*Q4)yvV7F$#?w;LP`4ws
zAqEGJ=uiswRqDy6r3}lwtHF*-hv&)WYu%tyVcFL9F|hz$+d-VQSAowiIBN;Lgw#F5
zS*2+pMNfn?2>`J7`fG-3C}a0A$<t2`jVUbisd~Qv_Nv{1>B8s{=Wgglo^{Y{>`+I>
zwt|J%-wnbG9Fm9+;Z~O8E!Le4x4Z)Llfhk$O0g;g_}*6N9(Kn@6))UPi}&Vm+jLB>
zziZ2b*8P?J3#GeD#t-zil`*RuZYS1>EXLp0UU@nu7cyWM2q&c}ISN-KHfzTjFC2Xv
zgrRc?zQ_>>9<$&taCC#-{%<25=%)yDsc20EUd-3#Bi7Yxd&mjz{<1im0sm=mX+`Kj
z47$8RxIMft1avua1ZHb*S6=Nm9`H9CgMR%WG#&5`0N{ida%WyR02%ZZfIkd5(fli2
z?A^omf4YHL=$ZbH8<>%S?Y~>~zW&`E`kx#4wr(CzJ(LN%*FPjchYWUJE2syB2gOE?
zgZbxYem|?wSVNa93xEb*L^lH2jCb}DOG4I9&hg{l-NVP-ezAGuphZ1<&o>L;Zu~z!
z@m}*MW58eq`X1jLU#~<&7(6=f_+-g@gZH+>CSGnzdu||ikUJIsp(2t!s>nKW+@Q?4
zM7X0}Rz0rc_s~<WLu2<r2E3cQ&r3QuMkuNfwtX}97(X9RpSRP3$XH49L$ab80Ei_#
zJ-k5}Ae!iBq{2J?r8d5@K3rd%Jk3?iSp@Pw(8<iAml*Qq!K=26q7WuFfR&V<(vsi@
z`_FxLesV)8ba`3-CZR&;c@qGy_sfeST3S4@W#+u;>oWsVyA6u#(uU&gjnN#vp~pI@
zQ^hQqX2C}VyWcm7J(zz=?P-2+pow%3WyrRS)5@w$M1cg{W2VJ*-A!)V{mkXjZ{2S2
zFp0WVqn-m8AU?^ETpj3K5+l4Fg=hL$s`osn*KB%UBgQ_3?nFE!zsG-yyLDMmSkrS(
z8M_k+_Rbl(fn7W=t|0lIhbE?2(iCBXhd_zfD?oXLGd3bmv6I;oDSo$KU`1aR{Yl*1
znHlwy_bwVH*=s~1o9>%H2%H0jgtt|F@I5VSG1}De43C!K4a<Eb-4d-yDlb%=%H4_-
zlKaftES8_|?$DX#=Gch@(h*Fn+*$Aw5%kV#iAkw4E@w&8O>2ls-SdbC*rAM*Fppk3
z_c+WlKu=7)bdFrM4a*24Km60(sPfSiAo^?uC_-C|;uf%3G+}JAG~wgvXK9WWeh-(k
z7`6erkepCeATQQ~q4KL!Dm{TU*^OWNt$xe12H_Ci>>~_Rp3estX>(>Sz}F#9mJPlz
ztpW#LAbsdpjR|u>!<^pcZ-na#_0YRvNMvs?WNWw~E54}Z#;z$!sCp+(uDzDBV}+1H
zaw{th7+bP1{1*GvGSqPZI?#I}7fdQas7F`yoT*MU<hb-;psbLJ4CXjW^Y|r25y5&V
z6j-fNT1d~18}8OPmst`cMO6T2O4Cnq#)IS5_8z&kXYHP#krkwD(ul}L;Z?Yom1wU8
zvvi(wEKlzfowYk0?yEj(_1C9u_?vYr^SNYTTbI_Cvc>gIY;Dx5{U}V6npa{t)@^<J
z(Vs}r1`F&RMV*uRLI$F-q6Lx|B<KXk>KwG}IT0%%!^IyLIW*5vSX8`TYE;36rWNqg
zG22jXlhzVJh^w;)nl4Xi$bCSN4QJN@lvHoCG}@6NHSxRAKjaeb5<wtyWtoR5)}tuK
z5iCL1cio9P2_~*4Gh7W+j*{`CT#Z%x5oqfx^dGTTmdS^VU6`i-09;%2P6lgl+WbOv
zuDALIYelDJNmj|CYfV%}JSfcU8N)EUXE1}pt*h>$R89h3O>u>5<FeAy|5!>(_*Aii
zUbjyUk2Lwn!h<qOXNvgj098mq9Gc)r6152#l#fo68~x^&5A^2Mcvv7dMO%lnMUib&
z#?}Z)wF{^=RyUFA_EfYTz>k;9H$Zhou9^W3i>b^}EJk>%(Xx7Fs?gC4P?KmIDx)1u
zRbUDfon2l(<6C32p20`qEa-}WOclp**`_&%%E7Vo{IH8K*YMC+t^w&hjykKZC{MEy
zQX|gs&tyYK4(Pia_&|pgwcqo%r|M6Ne$3Q=Y(h(-Pc9x$3Duo4dWrfJvm>bg4%W5X
zqs{K7=*>(WDv$ECDbhk`cnNpPa`>+T6~5uRlEho3>`#U4lo`E%1^ME59b)G>8n{*~
zTD0uFxRQp+jtRf=O2hKd7gcJxW3^grF#$)>*6&#LMC#k`u!V{Ww28+;DyzpZ?MP4W
z-06lc6}kC_O|$i+C1BJ4GlVL_O~2e+TIJfxnFei%K>q4<Bju^eB-`i$bwC?0#yZ}z
ztK=GD6t3i~b-dbLFCNa?*|bI#u(G3S-7UBq4brLZea1bMmOe}_L)G^gdkT!iQQN>;
zgmSdG0NCN^(YD|_g|ZH%1>jXW$Z-T`M+X1+pTDsH@H)5tM5>!4@?ymP7gq766!O6_
zW@(Zz3^wOGNhyieP+RBsr7Wd!UmHnIbko1{TT0e+=K2Ifo_2DeR{B|58h)`NMSTGf
zJuD-bq?;d<-s$fS;9BErO>=3pj(K+GF9E7R6&TC=2n)!8E~~pk<r`bRlFF2GGZ<ts
z5wOch(~AuPTSx*M3>k;>yA#U{JUn2E+Y&`y7B}kz`-`gn%D}=#LBehgJ>*Dpcugbo
zwMR9AYf(@2>Z|2@n+51;KDFez4o}TcrZ2=?onH*DC`vTWS7T|C<dsYEH)r-@LrN+n
zTEA>E(FzbJgJA6r`V96phZz@5xgizWH>=zBpKs!`jb1wp%TC8KW?|>1w)GEomZOI>
zbmdJoyHDBC72Ep}JNag9!v@^^xi#5r%|MKqJjc>k^jT5EH8@$AxGNRwq>72;cp6{c
zW*b391=QS1l4Fl1J5A`P_LLKe-8mX|!B^JdWwVmmS5)6|SYnMsfikw`ynvE?E$(sF
zC8`j{4TW)WD(LE}x~@(~7gr!S$3Jv@W=ryBR+lj?K6)6`t>Kd@>!+AiqsdmuwIVC5
z?CgOan3?jmP|BQGi+2bb1eT`v3CX`5N+HeZ2i;98ZC$ai(zx0owDKKv$4Z#%+uQAu
zP=O_InnE5|7pq*#*fl<^EK?>l)@mS3E`Ku`w|HUB3t#F`l4GB}*B+)CS)?9wD{!DE
zcVeojNVqo)5^j^92dq#ndzUlRXSgLMH@u>6il%DOc1g0KzMl7&hK%y+pEBZj6XQBp
zSfn1-O8vg?J}mpZUpM7Nv}J%{0d{tH|F#SSdoNTP+~2NG(Gx4H@#)%{M4OqO{{W$s
zq5o$g^PhtH^o(r(m6gz=*%6Dy0Y7s}_3=l#DUuqupWk1s6>Qx)W&#z9W;GVO`iF1*
zU_<7m<R$^<NqmU(((ryoU}FChKWwcmt*tDMt>h0CcEHH~^%lTC!3Vu~ZgcZ`fI)Kf
z9bafa9;t}XJT%|tUlC*U-gPL~=`FwK4t58<r{F*Jrz$`UIgUIxXgDts@935_Zu`M=
z=rYHFQNI0TOz+$G^H9RnfNuxeTEem#m0B)=*?y6lgn{fUu{}}phe(NR{5GAd!FaRm
zeY2RspV+~B?6IW>#N?^0d$^dQGb3@=?BSk^G%?r$W}|#Lb7R3{pT~95yA$-$9|miW
zuJ|oU`^2CTb8Q;9@Vsb0m~Fsy|FGrrc4}+MPi`f!90Y?8-6VL;_3d;_$-N*TxWShn
z8TaY#ad&@fDUWBRo8!yusk&Wn5xvl~ZyubG4ZQ}WkoME%Z+!Ovoq>+3D%1PN>FMk5
zQq%|M#bt@S>MLUZ=^ZmKTdn;d$4VV|8UAF>am2%=gOa@j+1iKUvI6NYq#j=!Erneg
zoM<VLv$_OH&-~7IK{0aotKv96FolARXm9Ay>G7iJL&m8^=_#UqK&^ztb!zN=b|T>b
z*fqsM+F>RYUHn(;+P~J$9!V{OoE!^<{tHiSFrtkK15#k>M>b7%39)S{lXA%2cPsBB
zm-8#NtvY`nK!8xir^*XCw3<9ITm+{xM1x%Pos_s;#{LK1(*)St$!>dppSQVN+}BRh
zAjg~xnnF9!5Go#~gWYM}klLu<A#ZdMMoilO#n?LqXA-vU+OajUZQHhO+qN~q#I|kQ
znAo<TaAG@I`TmW+R@L{es<*qV`rhj8vo9QnNR+a)q(EEL`jBJE`bF`(npEeAxU10E
zQ%!+&Fu?bVqvW2@c9yJ`1B+f`PeS9=mZ{WxHq*RY(Gql$x5}LRI&d7H#AkA}XP|<@
zG+Gn>c3ogRNpV6g>X;8fmpT*;ah0;Ug)g)zA(4|@J6%kg{~`6&UVMUwm_!Q|;q#r?
zMAcR`@2HhE9(eyE*)b9_4#tcrdKf-0yiv^SPYtMv`<X=01`;xeP|yt(;|Z^w+6!sC
zTydLo>7N`LVQ1dk5Wv~DnebEK4y@`2C>}+`<4X_hU5mvZ7Eg}DQPSohU(FC-OG1I^
z^{@{*AVF|pB?j8JOWn*xt#ON99dKt;Sm&%7a}k+zEGwMvglM4HgY8NmVpUugw6v!i
zMIGduta<`6j9~vz3#Vwgi*ShdDwzvI0d1s(v)pUo3kEhdi2`GRq`g1=WWZP!##b~<
zW*$)vy}ei0uv!J=T(;}4%GQ~o)#<a3;!ahsK$a{ugU@;qG1VHgIW!J3WO1u@X*9t=
zRj2zL8ta}X>sll0qRt%Ym!xiCB)#Vkd8o&cWMx1(i=K|`eTu0Gxj`quDW3EI$<ko@
zH#e3$wiD7xaQz6b38xX0(Iq1mz0-+FKVa%MMJTIHmm<IsX*<v$UxAC(%>F+_Hj-Fo
zzULSiRCsaMOZDy|5HWEcJ9Bi+3c(KjSZF(kx)e&`llNcG<3%s&THM_MGJfGIR!W>#
z_F%dx0IkFAJ4z}gX*XmE(i29_K!P$!vdJm&HsT1U!NAP%`IF?Tcb5R;?)^KS5#hQ}
zQ{I0eaF_%Gnbz8l4&B<bD0)el)2%e7Lns_lC7+NXkss^=tNf3MO^%7_(cdX>9C5Od
zZVT&pwN|)!qpTG7-KB>V+)`HRho<EtKYP$otwkdhO7S&9D*S8gO!)im(ppAKQ@L?2
zT?a1XpEDgf2t90aJ@G_YVkNnw7cTzVDz%xdmd?hjQ$qb(Y3?k~+yr9oPLQ0bMm1Q4
zupL%fW*&IleSIb2wU3(H`-wJ}4q}5^(~$U~_hfZ`2}|H_kbAku<wv%oR>y%Yy^5~`
zy6|gK;LT9nCiI+AI3hx>3Q~RDw-JZ(d8BFg9XsB(-+PMT01WATWmPL~%+(dKr$GlN
zRvhKd-F*uc@t8@~EjJVJCM`_>s%<0D43&c*;llM??|yT|Mh;PT%7drbHh{=kFmVUJ
zy!bsBZD0t01{8J$-H)CueG5`Bid@G*ufAeUu?BguQbnHkfeez(r}qYq|CDYTdsFvg
z;jcrWNDTQlNhF@4K4nPmfwYqncGvdBy^tAIA#H0SQFH#B+$4=zoQ({ojxk;fq`{@I
zWK;LQj_mdga0o5XITdcF7FdLXf-AN@yREIVuCJXtiaH+lT4>>Gkm&2o#qj1uIN(Z(
zhf_N;64bI6yfl@~N)Z3&<QZEVr+E4M7(2Pk9Q_CnLdmaC35G%GE>UXWiJ7MlCaR4>
zj|60>ZTKOq6YJ9t>Rszlzi=D2kR5<l&BlA|vh@v6pEdZJQ<eGMd|puX7IqkoBWe0J
zHhx57wl)NyIT`;_lq|JPmzdaO#?c<HjfIXsp8Sr>^?rPL=a>%Kt6=`A8N<{dWhV3i
z&l~13mz6|ZR-){)=&mVaDf&aTQ)5wH{cNn`2xse2{rbS}N?5_ar>fWlxz7;I%)>!X
zGt>XbyHx%8>_g9`eV@sm{_@GJ34(%C=Sl;w-4;Q^xBS?@!*9JhCCct=$6!X6y-ru?
z`2HY${)GmLcYRMbE!GJ_o|?`@<D^b?4A*V+e7RTfDr(Ez8BTouHp|CGo3`lRJf(7_
zl7QWN(JzUtSe<3Bqp@iR<kYzJ_6hlXLhfFiHl_#nyG-kC_#M_#V!*6o7MA}7X{pZB
zh~Arb+~oK-Fwu^#)_*L4{HJYq7LNbUHav|p84F;)+cDVhF$;qGOWhk37GyvB0Mv1r
z?=X;yE+x-p`t@Be7Fj%2NIKcdJucKrmQd}{P^p<(wySXW@Obw4c~-_zq9E@HilA`a
zmlLQj`35G1Si{8yWQ3Nz^Ue8rM@y#NseiVhOf8mHw42*I{h9=#KoTep#0Sn9;Ea75
zQAiOvl3(B;g|Xm}me+IintJSWZl=4RSeJGmup8i?h_!yF$D}ImS*EaJCFE%eXHarx
zzC6DHb%}?QbF(TMM~)lKWiq7keoq098So2zWP}Ip+V!IZz`!v*N5o&kbjzV+tc#bm
zJ5+YY{F>`o(1(PwV_EDx>*Uv2KgIHkcm(YnC7m&oF{=I?XUc)(P$m>tgAhpEY|Ot?
zVOP;HS~XqRN5bsq1LBmOMKl9!j}bXcQqAoz>?%H5!5NmSSQBt<hDAF3D`osrZt<mJ
zM11I0;`9`2X|zRY5gssG<QC*t^6B!b@lWJ*$fVjw3d)owJtYn}q`AFG_H=!(Pbst$
ze+$J5kVnFj)bu@@!7Wr?PFSv5m3dAbTcu!y^4ZnG&Gt8jsw@o>Mm*!vGU)XPL&P!s
zirpHgGh9xJZ5hNwy~gN;6ijqVB>Sd2P8|kB0h!&U9zCU`{?)J4M{riajJ@66>%?Ag
z7D>(H9F!f=?&3|-ylZ1mFOMnPeqI4@_YqNx<%EKvi{D?KkqoSuMtD-|Fc$K=0eqiN
z4+OoGE|!gtBT^p4Eh7(|CZSAUro2j)l^po%z6^0yo`*Cntc8hCOf#@>%X5h|9vMcN
zVJmT_*xN}CF(6r%wV8Dgd4OdeA7+Y7dH}X%sahAs_*}5AX(|~2E>c9PNEpqs%GN_(
z6R0=w)tToAKk>|YVi#iKJX)}J7`{6a#{rqDt=~6}=Ad-dX;CY=#!Z+?jVNTZV$n1u
zlEJlEGp1x2G~78&Q|)2Hx5Q^i8#P3!mq^_Kd{0mmI*Pun2+5L1%aEToSfXM~r$gv?
zwVB$MXO`|V(%9wR-0XU#*1sTbfj;aCbYr>h;g37hM4b%=$pO%%F{SQ&t^WhtrLy}@
zYxPA-xvekL-fwKyaUMz~<(IK6|B|BRKZU@d{ROt=^Uk*QnW-t|pimU)pW$3K+hCe1
zj9&8489~SpYy+<HE9eKT7oaWf_98;Ld5OOjuR9;N%>}PyPS0KHE2WL+J=6Xy%OHHF
z{<aRKB>X2a8`m9P0})-Uwu8eDdkBYd$gZ2%^jiqd9jCgLIDXNr>0ze7b6lpK`!qi!
z8z4N)+(jELN3l~WIzgU~=A315>=Z_$oOXRI%huOJ%C>Gyi7nB63re=^CKoZBIx*xy
zb%6JhQMgH#I<bm8DNogra>IM+zKC~xC-%f)6HF$9fep!I`s9#MGY<t7J4;I)4+Fb7
z9=D<$$m$zfO)1cm;AH#qORgdbVcC}0LIz>SY$A(tX}A!MDY_-)qI%wQVtJft@hC+p
z$p$ai{%FqBU4I=o(ay1_+M>6#VuBWO!ZzHX6ax(kJj|pa)k+LjG_WGr6_sMnd*8C^
zDBJL<P&%}7FQx6kxF}Snq|i*?!%6Ur&=ixjcTX@bUrmdU@*j51Ci1QrMh9XdEkfe9
z7GSW$TGN-^Kk%Yu6*c~)PwHs|+nE)%(O@To5(!4$js8~i5(tgRN$|IG_ta5>HtnQz
z;l={Vg$r<Et8*2^ovV@6wksm5mopE!eZpq4fATme6ahrg&$|d6VOg!BcbaX&bDK!&
zC<ST*E^$lmvb)Z<oVDjMHx1qjrLCF3VHLhil+-Uh`_WaQ_elFdIx<m(S}q+-LqYiA
z<Q2E59DcAjn&V`~c35IT9*44AsP{m6LyOs-9eec|pmR$Vamnwi$XzfP_}L5ccG*6!
zkDI(5l~v&AgtY7wdWb4I{27ap7!@5jsCpxDI;MJV2wPhrSzFpTFHs2<iXiCTfJi|4
zU22AT^Gb-;(5#d2cRL~rG-v^_kehxitQFrqpds`@$`I3)y3~P6W7w;T;#J16t2=|Z
zC(#t!cDV%EV%{o(ikMrz>>a%Bj1@HC{;N+q9foj^v$6B27HK)Ar_>zs6uA&%#koDs
z3K^U?crem`nv8Gaz3lE#nDvg>V*6=4_$2NOuG4RUnfLmZ`l?7DFF=_$2LnO>{we8T
zy=sZ<+JjaJMY}>y#y!~PoL8bA0A({tK^D0b`V=Y>x*StwR=?xk`q2IFJcxEPEi%VO
zW`rLMgR3$}(8sn&G|h>W^u5@GVDL5F<cTm7`5ZGLVZKpwTdZ!Wd<0S8%ZGZzs)u)Y
z9`&^Bqe^x?Y@yzUA=!Nb%<SC|K*DW}nm>n5a9TbPIOy|2;jsjVP}gcpyV|V*I=U>n
zHj%vd-2D>OX{i`|W9Y?wAvtNv*&PF6M>?-Zvvp;0VZ95NsoD7e`8Zv_%wO&hCghdx
z5LWXbxC>YI;k^rY@^S0Of+@qP7O~(3&E8Ys;GxRbe6!IIicvfTc<>(t1aXTP@+cDo
zJ=>1?9GhAD%JG@ge5e+eGUnyFjiz`ove>xv{X^$)nU9$EZ00bJ(S}NHam!j!b?W%b
zBYOs_<CL^eZTT53fm)`?``1X!EmG~d$l(lSUzaXd&ztBscfLYMP4n>d*8PCFYwuSs
zew^j@{VR<96T3s3yKW4WIsz_f#Hf=f`i1jja!BEj=HV4ad%-mETPkI=Kd?DX)m#k9
zhRcH-LVqSi(xrLIG&>U3W|5<i>OblF&|5fDLR)U8Eqh+zp#BrOYrd-f2qx)!v5_X4
z9eCBwNZs)%TbGID=1^PLk3()}UR%jO@|RfPJ9BcZtbuj7e;{PM(d{vQC#YLaIHm_*
zo#z0VgxK`!eFDUiRGaNMHZc5+lhoJr$~2A?81pH?o!jpf7(Y*2m418Hag<ovC&7V;
z<s>gL{oxV)-Ox1YJVCzbmxIzhWVL$02jH?QS>HZseSl2Sd`qve|ISa;=#}M^16;Ul
zLa<Hg1}TqLl%^K18#qb-np&A())vP+?u0wKfWfhz`e$Gl$csL1`7l86W|4t3340@G
zP5?muMw~exO)o>Idxl8IYoDu>#?r8{YHt~?X&#yj1C(51uA_pFMGl;v>{x29Y4lRv
ze+%UFSla%_RL1(hM?;udSpPeZ-KEZSED05I@2biRcxDg~NW66FMD3lVOP12Mx{vnd
zlobNcb-)@B2s2SYj{D=2<YwX|JNa)y|2`;7uQvec4+C76-@w1q%Eft5=V<TY^<8Qn
zG4YF9>bIqWJXMf(4*xgrCqPYF&|B}+tavF?KbGIeyY0k5+bgrcGOSRkWM(&8iM7y5
zbP>1yf&L8qVf}V3CQrh9rzkC9FvTG4UO*tgD`Bb{6&Nu3vJ&wA;`=%)V}{pUk121N
zqo|zo^>`xlF+fQ(7m5-HHIe=G_I&o7X5dq4R`aDNNlaw>A(Hp?23U&u7#W01459i}
zV;2HVS^Y>MO4sZnmL3*jDM^_MW-3Lva;N#WW9Oe0)QOF~A9JoDLldt_ch>(B_NKs5
zXa4#`iTV&?m}2+FH^+0fRB1Fd6iO5)hSTU6Tl%YS1L<|#lz7~{((I1aoH&Cc_dOvk
zq0LsG^4>or;(+ppke8W|klgGv_!^Ons-$`y<N6Op@}Tvfz0=fH(ChZfzeG7KMqxlm
zfC||4N_N>lzk$~%$RL#bhu)kKpuJM}5hJ1);m<Hz;!K17HxIXwO~MvyK_>$TvoNhj
z_-d8~7L&;1!TX{dy%F3p1L`*O<&bmQ!xPzkyTKqcZhP8r_ZrjtX<WdaG$@^hFE}$e
zOs!HEir+Y4zlPiJaNW^uGyJKpd`z4Bm%6U<PO2D-y016BmabH&MW|Ek)Po#bq+L*F
zA){^dklIAPK830dIAXW>zN2CmW*a+dG@kpQawUqA17MYyo}B-hOPrcWPekVo_IzQh
z(O%2v@HjV-kgKHiG+~EHa!kqaDQ(wZe@nFi2q!szOx?tP{nzNKX@fs|XG(|z-X)N1
z0TPP4G8-%h`SoD~Q=^^)xvDypi0mH|Yvf^-Jp$d(vI}It+$twrghI{PVb;{U;q>!|
z^26un!PiRl7T7_V-SOi~bw8EG+G;O6$lip^>mfRIn|(zA&^A)p#`sOFBu$YFbH!)T
zRtaz&x5R@t$iD&}9=Q-VAoH%iwZ&F*VDH&i_?)RFB#rq2Y28c^bR%dUy#%hJ1vHxu
z$9`BYmRYp5i^@XttAFdu0D{wX_Ce$WQ~#*=I*6*paPvf;6Gji+W}Z9-5olU&Pzxr#
zpKN8B$Tfo3xki0ku+LM}oa)OwmoSfO_(<0M62Oos69s1#<@Cq_KsTLBkR0OlVZUKV
z45lHz_OZys3}A<gTVLQ@Hj5A97kLG`66)`(&2Tx2t#_j6eA5xst1r#fI@k*r_b4Xt
z4yF>(tHt*(L}6jeCp>9_D=0O)fF;PacpkaGAV>@c`(lzdG5MM7XJ$$1Fq(ZTis;G{
zK&qdSwlnqhb#k`-m^*{f0uPUoDBdfD%;;@qi!X4l5BPMBrBin^{M!zGyQ|(Ib^gPQ
zvqV!qF+@0M!qoa0&#6_Dv!PE`UB$F6X+9Iyx;&?6wXv3u+M}VA^9_Z#>?~~`xD2+A
z&iEYrl{S$88VfZ7BgWVQ@DC>uww*s6Pio2k8I}<k412zCX+fUT2jffV%sFgxbCe|7
z=H66MB4hbAqCM_5QW*a$qhab`N18Z3n#_UZxerCgT83#r-rG<JmemMU=2%_yOag8?
zg1NEo#0+XWGZL?Gbxj#Y@19@i(ndU?*mX%=#4&rYhQRw5rCJbmbF})!n!`C?3Fao^
zLt;M|GwFq<&fjuV7R<LEUZC@SEh8jvgkvK-MJ_<{IXT1V`@0Z%N+0=)-<D;G3iDO_
zx<S<LFc#J?u&r?{jt}mcueQ(LnY`2uzs+7AC)JV=aL>{*chps|CWFwe<f}iwv#+he
zNRY_0e1;Z_QMhGS44hj$eT^LlFYkTfqmd!D^!+K-NUUm3PHej2yD@;;)7pRrWz`<t
z@FLzDL9cTYsk?ryb-$aNPX9JB5mpc-z)TsOvs59<Ex5d6`%M{~Ny*wu=X_o`oDrf#
z36^V#7Yz9ZeWlplK-Qjbw2?-LB%rl}_>X$O)L-BfZHsMV6ytPF=ibHm3<po49<*|q
zGI-O!1romCO<Shrq3)0kO*)3jF@1#}3U`C2#Z5%*@P@*pWR!%3N|YzDv%v+;?4b$}
zvr~}S&9`w%SDV+x?qT@X?6W@E!DLD6sip&x-$7<mRmnZ#Gf;gl!ZT74$z|1IjF$#a
zVBq{pC7#)bB#Cd3W4#QtZN}qBa#<b3g<WOjmSJ{rww7h(5hNbds<6L>;Uf&t?7d=a
zjxZN}Jp(17y=L8REp;U6GvzZ00_wn_6Qha1#PHWhh5w(m|0PGN4^BsCUuw$62qC%)
zZr)c$hj}qwPDcGs#SPd_^*Q?o@1KTi)(B;C;aSnt5<Be@O$D4gB4cAmSELyhS(2(5
zHyaesxsl>^+fm4?Fplt~SN~QEF?`~=b@*ffA7h#9I```YW$cn8`%ng<5F9-pi80r~
zv3XE_V^vSaBv5{BuC?ohWZ=Z=As*?Tt}6nj+Eet5wc{BYNoqWqwxKJbE13koC+~&N
z0^drHL(2s!;KsCL?ago;hQ(=II&qhu?p(d9z7jXtdv3zvh|x~Lrkd$LZI38Av@_il
zAcHj@ZMB78TjkQ7>$BDKRHHPF=YAhRnBM%dFRz8zeCY9D{qA?_vb`$HamvMM==M&R
zzHGnG2_J9!3tX3s=IqZ<;}^3C=#^47bI^v3T+)L8I&t@Hhv)^f@$wj8Ot`<-I!-yZ
z6j-J?dPh<Q2y`}a@?LrfCfz0=guJ?E-N>N^S$D}+EA~Pm9*26(C2ns`A3SA(Hu3N*
z;#yJLmFAV`=t9F|X`G2KC>g-<>ckwaiJwo}Vps!#RC!*h+q{l1_FuW_)0}dvOz>{e
zPRa1fTIX-W;v*l?2fuNI1yn^Oxk$S>0=mNDg|iOkce{ul&{qA?qOX|=+8?WmDNs-S
z_jLNKrfzA*`IAhmjSUwc(BGNVk62g73|+f02G{?sK7KDbH)9BTVxO&EyR5EqO%!#$
zUQ5J?KL5<eh3!A@YugUzWenlIB|jTv(FqnG1src&h+sy-7P$ZN`~J?Pb}S)VS|VRX
zuK5?!av<3&UiHO2fsaFo^3ISRB!vGTGuMwM_<wO9*;(2Ds~+0VFpz^z*RP#(jcftK
zFcIe$AdrGqWUN#AAP+GN451|YgTiO``MzDOUWD=ccv_wjJy#-`aB{A{Uj8h+&TmZP
zXSX*ut}l)&=FWp8oNRo4d4XR<XAF|q&742AFg96**~MLc=?Ypu{8N*k%*7w;@AEbS
z$bnE`2{Z?a0l~e_2F9NM30<+mZ>bMT51U5~F8U5uKcMS7gt9xXug|kO$c&7dY%03B
zuyJmZKmJEkxC8TTP9Fu^AdBonWnA~+Kb<%v3`qhP?j50w00n{r5WAe=BK!x?+sOE@
zEccjizFSK#;oSFr7vesG8T=K}N;nw0$KdOU*)iBjMk!nt4BnBY17DSwy#`q?y)<nb
zBi!g@fIYvAC^_|%5H8Q?Ju8K1iw`$mF_eD+T9Z@fi_W7f;Kl2i6e5LDJVteGrP}E7
zTnj@?L(5+&i=0ow-Zw?hM>@5<wo`X4TTR$ojV%39wT47%k4k$Xw;8C#(QxGRp%)+6
zkK3;=Ql8*b6p3d;LD{l-XQ2&VSuP;{x%&I$9KarM)I9=KVrRiborlvz``MXx{9=S=
zHId~loI%_60yp}ja1^Mrp4Q@#B{G$nRgT4<UteP4zLk?tItjDA(t1RVB(zCr6I;8R
z_se5RWxWW$-_5gYKwe<We_hMeEt)XY%Kdoc<n{CC{^+fjd61W6U2Ho6{r(FCb3Ry{
z3}tE-@L}@i|H6EgiN@|fnvGN=tR1zc6N>i7e=t9Eg)NZo&?Dx?^1{Vg!%N<#U$kRY
zLblKA{Y7rFS{FsZ+;`C?GI|c0#RuM^9`kR~Y$k4%xJd(-@X!{!=pS>|5!}V;Y2!b?
z<S~AgL@T?<_4Qt1Rvaja@jS$gW=Cpa^qQE$x=}1&p+8gdT$yJWpbIEM3M0Nmz62pj
z+64T2X<*DT5ySVA>}uY|4#zZ4c9zK|GC1tsoIKMQ#=0XgQtVU7X4-)*6FzU)<bOk^
zk^r|GEvCg0*;*2qhwn_NYigjr1mommh7*P3cRQZ06`3TvE?~Ueh!HJ=ah9WnBczkM
z2sNg<<qCsQ892aYUriBAMTWP@S08kNp00x#hdb#G{|-NliI?T)jFjaW^rtp|+oeE$
z0HBu0!ORT+9(E(bt$u+k*x-fRf+DTWArO3h0^SXJKOutAKf*QQL6gRXZWdj<n{(@7
zs5w4mXJg^?P(DC^NuCu;i+SxPq@=pJ?QOaea<l4Gm~SZHlNFy+9dJZ(S$h__ZU^93
z;10o0S`s}s;nB)YqaH+1V|)yC9kzT(o=;NyA1b`-mM&BV$cJAq?#rZ@=VMDZ)p1*`
zhDeE00SVH~F^4D%3MO@wL8qbsHVi$ccuP;4SK^@pjcpxKZD}gJ7nm2aTjA!_&BB>g
zdq;0$77-!k?#7P(2!`bnb51@%X%$OV-~Z?hz19)6AJ%-!?5+X@LFVK=!e(!>nL%-;
zr1!>DrneI2o+V<qbcZAnT_V|^J0fLJWi+jS{KWB+0U6k>2e((p_kJDfq{Wyg4`{?#
zLbJ3mm*eHXs}9xzwL*PBWaK+-CM8AobD!|lhs#f8y7_jnu$+?*DAV(%{IzfbUxJ)Y
za9EbmzhRU-`@z3O$Sq%If##aYd@>rM=N>v3%sz%&W5u0uu{Vy$r8wGRp7wEr?8K2j
z?Gs%7IJ(7QknjOh_@FU#1OECqg+>O%$?$aK3HeUtH`ToOju7qEXeonPR{ryDw!~B9
ztRCJVwn<ZY!T&c*9(_n&=$j5KV4)uPd5k-(<FF;Ev9CK>d7f!EVOel^0CqQ&9?`)v
zd|ZFiNv=d!kfxk#6KNI}El;`>X1Z8Qt6k2b4TO94f<rcZ0^?6rab9c^h@7XFg1~Vr
zD=JQ19M3OVp}9;&(!$DZ+*s-?bGK4uV}ue^vuP4LTEn-201!&OtoU13Yi0M8u4cAX
zJ?Arw25!PG*P>Ppjb6kv+?pu9H)dqloVOCWV>Ub)>fF-jVm(9tI<;Nj8c=oD;0F3z
z=FhvTULON1yKt_VDJTHg7>Zu#F!;wnE0(PNr(dr&8@Np^D|rnau|=m>-<5Q)4=0~W
zOR<1A-WTxk=U3=yf^@}en~)+Xw|d}Qv^6r)o5qKa%&O%M7Y3V$oLUp4hcx_<vg;L^
z9$5|id{e+dZQ_j1=!m!}0~_6gogVRJc{bzUjsUeZNpKnb7>9YXiI`u>6=KNKl{l}N
zJW}U$Nefbiv`9kQpk?mbuzr7!nSY=w(}aa(t<H6>6|W$u^MWxrbauwGgzDTE4JH1x
z9*u_X+!qX%{t2yk8Qc9A()Tv@bl^sV!Za9v_gc3Nic08k=H+4GVZUITS&R6#7EYB{
zps%!8+fMBW%PA%TwO(ORh*=L4>Xi=;=eHau6k-w#ZI>=V4UWJDqJjo|37Vo@sH`Z3
z9sf?y(4d8?m*=aW9q{YPjLok(k(uzJ$<7L)LsQ&-oNuv0!Wa)xcaD@5{&{B~bO)FG
z<5>alyA@jkIO(b>LD<|Dw1l+gaOsFxDIWuJjzvtbHCqz22%B0|wR_m;Ojc7d<Bd+~
zQbP<^X`!n?-5u|cIrpvt8@{C30#m`fmfEcm!O)%D@*R3`{3<kzFM{vb@cVPT903#A
zH5hmKOActXhYib0%E>>3k$-l~3Ozcz<Gh=41dX+Z;v<oT(M?)PudCj*W8Zk6RB^3W
zecxYU$LN__jzw3`e)w3zf7@XWajgZf7F%e+bya4;DHdm-PcS0cG=vw#M=nplc!Wk2
zLrc=gG3lFgy7#K0HcL!tJJbW!s4lxwox&V2(3J;O*Xq`!^2o*CwF9CgsX&HW)d@--
zVtv-8ZA!W=tUQ;(l&X^$!m(%A__k#V5*%|t_-Z#`Sah}67?@k0EttY`@6X+7);S33
zBi+6k@u6$ag7$`+C@19HHCaWyI5cd|TI;-|kT4j$j3OUyyn5!AD7#sP=u(z6^#eWQ
zIRnR6Jp%d_(QoYY9}d10Y?c~2SS&hY0WyJY*;&qBKbu%5SCl<(TTLvb$o^+>s}Ien
zyVzC&%x(KTf=&yN+bzlrY+YIDMRo(OhwcPf9c0Gkw(VB>me^IbHr4DJBv$r1LB$>K
z4b)vp@HH-E*uPe<x`VX1BARX2#`iOls#r9v>58Iij%X2S&5@2PVV$uG{GMYS4rryS
z5mDQT7u7zk|E$bCfLc2)wc4R}?AM^B=DCX~C9Sx+?flDxkwbxJnXOSCu9jQv21-~#
zaFc8rol^1ASUY^KqA^Ig>sa?`j?ZBkem$c1Xnq*lJfJSRu-3tmc9~p;$$N2;@X}ZE
z-spJ5v+EbUrMhhOvsT+gDsNoWbJx%-)alhlHqS^AcjyVGwaBy3zw1+tgj?;nm!W{)
zuBuyR_u(SjyRShC@%ZzMSCD4ll01f_9T&kI0i)rN3*CdNm~dE!Dl2iy2j&9YwnI!y
z?a<78O;{Z}HyNr{`^0b7m%|{itEf3GH~kGHw&G6oA44kVf0|fl`|lKl)7X=Vm*n=l
z2Kxk1z@)e3|G3q*PmZ@L*QaZC<jhi}%~OUCzcEM{4f{v<>#j`q-3j1?k-cGm2+)a(
z&uPzbX?YBRNocXj%ZsA~Ne76D-_+y3!^o?W8S_iy1j@d<r7PeCu0QgVfj%_9c5k!f
z1%jM;MQ-7Si==b>8_PYxe(qRue;I$)Z->P-vKo%=9jPIpAS0Tw*|XKSfS$1sM}G|3
zppYg`tf*ioDa~FPYYk|*nKLPv!%^UY2aEu%xUZL)4@c`H0!fgZOLy&4F33m=><EIx
z-7(e!906H^j8`8_3jZVp0BeAV2TZjvL}+@d|If^sw(-ezH}MRS<UJN@hZjO7D?|1C
z&t6v;#^3uD!m*cer2E0Ndk7Z5_{;69xF_(35B_`|OXDvkmPKT9Q+(0lok3460wsDI
zJz~y_``nXgsmO=o7AFF|J@)jGYX|7Xd-RWPmk?xygJM8i=t5hwNxB!bYwF|D1qq6K
z#8>E*@DbNy^#Vg9eCnE`WQm_$5B~*n+(#m*M5zZl{Hoz83LG^f<HW$f?g3~-@;r~^
zRTr|ngiZGOZnAai@iasBAQL$d=9!yK&k75^&5y=18!V0*FzTsxv9Eu=+GnzQB-{V9
zd>0{|e&l>8#29=4ks7x9w$;dNzelt0ssqDMV^;>?22^nS^{Z#7rP}LlhG<z;AeTH~
zxv}vD#}vn*V*kkq$AjIWC(kQ;q6l8)<Exrk$?D&`dE5DXI|3X9&=e5QP5JYDKAp(C
z1V)B5@X8J{QwMzVf8O3*<6B$on_U(brc0Xq`Qc*(kI9*~#CK-jWtOboB@F|5)R0ZR
zY2{2&9qOV`9wO){%Z?OKAI%RX$l}K&!1xFs>3DPMjK8Fa%pMe2=FCa|#iV7y5s?7-
z_Q~AL|MWipy6qVDYr_iZ_((oXUWVw<L@U!wFwjg)!dQeX&=5x!_27YDhin@5Zfxnx
zAVuQ&1{|l$8rIF7wLQI84v8r1x0Xcd)g~&sdw1z{Y76F(>0Q#tNov5J{W|+cDUezl
z`-ehOBYZH()y?7ydLgBT#IvBDQ^^Z5w4camPYd>J{a|=5sK&ERCZ9Kiw-{r$##fT5
zGp8+kT&Z3LnX8<^#|aq)CH|BtwVtt*jR;8?RECfZPJhk*MiTNZFq){4R#%e7Ql`2y
z)DkCXMsih&2isRK6xoVZ;$%B}CuPGwjuU5b%Ef3|2)D0=qULh^&vJQn{YbUvXX~Vj
z>9FEQ#fjAf`2>+gC4U-(GNmihyS~1=6=4l{&0(7J4L+iefpii?zJ=fh!H7Jh@kH&w
z9E^5jP!%6tgAij2=(rVj%1txIY)5<jloOXLo7+D?X7k*#qw<?5{pia$0>3`?T(<D9
z{m5D!8Bm?wqC`5a5qB6z)2p>r7}!n;M^|1`oFYvvgiEHQ_k<%VzKK^SPV!4Dz|BJE
z>y7*H0%X8WQM?PwmIh3X>J&rg@&yJKumC00RQZu<XF6k&6jc}shP!}>c{U3AcH}z8
z#1!W;Z;G@?iXDFFFK-Sv?xQ<+c&09CT&bWrLB30;q-V8pGZoQ<wVC$DMym9+)yjI4
zG~@B?vsHCtDw7VJyAcIi&wTfmV11>Vm)Aj}9r;)E5_Cm4XQUd<uD8^oo>btx^dNi9
z%W(sjM~=`*kw#?Xz4+a$s7c1rA#%m6hG#hD@Ldxlf(>2O!6qhatGa&-{hBJxd}NC4
zeS3qW#I%!|8jkyv12o+=dx^nri7spSM;%zN36add8Tm`_1Sz}eEMmK$<uN7*Nb12h
zr}UaX?Oi40xlcmz^8a?*j>7E38(BRxKy6<3CXHAZtyjMI4}*GDAMXK8M}ijdu6P?Q
zN8XwQxUujP_6T0r4neM!X?kD1;zjLCLicN{PqrYNuD(ZarNn8~11fRY=4}?FKavKd
zs3w|-G0kwpt?LH?D^TX1!ev`_-h@kgr}C1mfq02_9>t6&V=UBdi;x-68*mzdIrWcA
zVNJM(GdY=CS50bn>PKmvXKr+T(z5(<OK3s3NqHx?3l>dgvo>#1oP%Ve-%O0mXdLA`
zWe+t=i=qgDO`w{$_g=(lM>P+IS4do8#84Fy%;rQuT$P(kOUWKfb_<^iLZ>(FDW2O7
z_p!DJCy0eGYe$K}Z3WMoxr@&-6PUAd#sP#0MUnQMp!x7J;iC-gpmSHM2&^OZ@pM0`
zQYpE2nqZ}DoF_=|vYK<Gzb=#2M07VYwGz!dY+PzgXtmaw=M6$Jwlr@}(mk(TV-{UX
z(cfjt1`f>%YYlxJQOeU=;VTA0)=i7k?3`(r6D6sGaVqX6o7B$51U*~TT3!-6JDUTF
zsckE~j!Z&MAJudpA}R-VQR*O#o6#FB9LZq_Y9L))r<zY<99Zc%!hbRJOck>sf`9i7
z*TG%$Q#yjMtzhA=ZTkGVz#t7rz-zX6MiRL=D%d@$MStAVydvCQEOfVCd5OKtL1&ZD
zGSukyCT&yDJHZcEirav>@jgGyOWu?au-e!5jR)xV4C2?i7H}*J9#@W}cyjGs_>+-5
z5ASwyh$gp!!1MWkmxDe?afo*6;kOel2zl>?Nh-YV!zbka8Fl#cKzXr?bT<2OZ~upK
z_`2l$rXM~5VS!>B>C8RiEm?2>2s)+3uzp_Ghl<pq=RF7MZa$<XpM#O$;8@tQmK*2@
zA*?1dv%NPj!CJTD^Xwmta&0htrm;y;-fHsu)>Ug3A`oTq?g2FJmaP~GW`_Mfr}jqA
zz?S3qwmLMgVh6>|9}MT>j)VF6qxd<z68nS+rKfzO^lM#`bInb!z9L1=Ht4ioTO2Ca
z1SZIxZ%8yYR>J=jTUa^%FLUz$n_0!k{QonnI9WLVtHRsz|1hgK5Pq0dO#Oame)Zy{
zVoKboz;DiHXpT-o3!qo@j3Arq#aCSVe|J#NKHq%GwI<K2f7^F&=E66#S)`IEk&5vb
zs~#k~{Q=!~z~fvlkd6dRI{<?xUh<#(ozAIdI;V+aCQ>SnMQP(3HrsxS<sV_~GxKMk
zO#2-|m#XC~5+Ru>=KfA_WJ}W5HhyfJRZ+`7Y*+fOgEhGNi%t@Ub7dsTDcvg*elmWm
zxME&oX{ds=zJy)OjCK~%p-(EhBf$!3g&i{tm1<F^nnxz709%!@rZH8>I+|anMkJD0
zE}N5dBFWTBGNM$CQc~`+CTSTv1F}0hhBUq;Mv$G<hFXpCS5>aME<|<6Q+D-l<y=od
zzOD<2)g0?BGGb*04z$S826-u~<$gpP0#|DbmBd<H7Mv{MvK@oj816d7Zz1NgR{ROV
ztQysYc$7Ayg0zd-nS7BMW_Z{zX6O-a&BG@wa}H~DrPhIVEwNji=)_H1H^${wsEFT3
z8|c8-O(lneA`vB4h$v`MR|Q-_gECP1;fXh7ze>T@yVZ;`4kjTp1kG(+_7{zlkn4Ap
zEw!;$2y7N1X;nJ7iiLrfYc5RjE9G0|)0JytfXXTt`RAf%t6$>>5v|8J%!>+wJ8s`f
zH4&=QspNMmV1|dLVkB;9hAf$B{7XuFIuL&yo$(##mpr0<m$3Ssr&268Li(jACH)wt
z`|ZfF{L*#Fe^pjRW%YlBO_&n3u`VB<Z-B+gFzG4qW|(Y1C-M1yGjEw+Pcn=W1oWiG
z806&5V*Gu1{J49(xHA*=jR*(qj$DsA+yo*bj-E6n4a>lxqlQ&tGX^`CU(qIj$a=PO
z^m+z#-^?S(&wjJ`FOTwTWzWoxVgLkk1FpaQZx@Z7oF3Dhe0@4!&OV(uGx~4>{J*b^
zzaKBsZo`X)6$1J+eFJ*a^?JFuGeYxbM*vN)x1Pu?_$b-^i8Bv!{QfEzF*66Zv?r?+
zzQq&{q%1x7b@U1&bqopyeK~qDqn8XLFN?rv+X9A>0seQUh~I+k4hG)YtksXNFS7xk
zoAB5$2c<?XIeh{=JRhi8LGTGgBS32QgW4+bjohb<{wxAn2EFb5Fpy4PM!xRNKU{Ul
zT>S4hqucGM;dx_-$2JUwlM~pf+ofj*+6KA1rJh}Yrq7M$v}5MeUa7}V$@*~o<-*zj
z&bZZ+i~s%U>;-V~*O<zQWv;2HjRguUC6A#tGpU$;<bHkm<MU|z!WwJ-a~rW~wM@cb
zR)Cd2h!wl~j!3+Hs4a$$qxD!pLnhpY_@XEKF1F5Wf@f8=r?gazt_h+1?|Y}k0cNE`
zYe~3q6Zp_A<W+$z2qIe7?OayNF7k)H*xUmnm9Wg8-0%E9!CR9ZA;P(gGUb^FqjeB)
z-^L&IUjW|^_ZN^zYFb_RJ1tRbVf;``0-1&maHK>KVzoA)6I3J@f_8A^Yl89%#qZ0y
zB77QPb|C)=MH#KdJ|bvfLc*Gp=#ZgMTbf57-=~WoHq0WZM(^?GZ(D-U6Jow};(%?m
z3=K|i)38+%6Z^qPRZJjL_B~NJ1t^a`5x73IRtNA0I^%Y6t*HH2-D{M;cx}5wHwz>l
zX9-<TV|o=TTU2MN|B~IQJt{52H%`h|U1(FB+mrk<OyG$*oS0CMMY}z8KXj>(eW62x
zK+wxaCoq8=fgko-y6HtA3PtUZl^{6z50%?s5rB^QKiwRG3dI|XA@T8OuLw|gj4b8t
zo%@|tQwL<$hp-i4YAOs0a6mn0u-rJvIEQ9%fnX8zr5W%9qd~VWpFM!2m^`=ZU4f*+
ziXZ0i`uBadq1mait=(&6ZVFKfFw&uBwSwzwD+qL&%~1oez-eI`CN_ad8ur8v*yw{P
zM7>Ts!>t2l@9CRo1IZ8hh}{T^h3fHiz;jIYJj3rPF8G8l<oas^o4Ba#`hfLd2<a3C
z2fDeEsdIaB5nlcAZ#v}vVCtrXSrd69<{uO4*XM&xQwPMOlOFBG5LONNNI?$^1u09z
z|JpZ{3!{`O_a24znq$5zvu#1E-wpJ5S07`;<a@!MzDG)P#VPfUbNNj_Xx9p@Ua1&g
zoK+74qBdcLqK^KeA{7T&qpT%Sm3$qpu4O-0`D|QwYa)?|$|II@ATL$6;I0QX?Rb-g
z%}KP46WIrOzF&8Y5w+-1_MM~0ZcryH*^%gx;L=32fpbu8qH8(;gXSdw!qq<X=$*c&
zHXIlO<jF~WaE+jtW~g!q_VTwLmO3Gs%94i(m@xir6sjS_U)SUh6(ZK71*wK4yx4sw
z+>UHGF0KikBuIvx2RM{Y@NbY;QkVpme8zsLIthO|WhAY554e<E=&uWTh{FIT987U4
zs+P#;iB955bTIaDakawIM9tPABS!BE8o{ZkLP+Q#B#JupG>Nj4C>`+mFW}|41*CNc
zR~dD#S{E^TD-?FYxIGHfta@f;ubn?BWWYDcqe<nWt0*aBpU5fI&1$&ebc+aO^4r~$
zIgRp<OCx+OQVMnpDp@z`R<uRNOahKwKVIJUsW^)w#%qF$(hB<iQAwQ20R=5B#Q{=<
zbLcaZ9_NY{QDNKF?1DkkawZSOcESYhLi&*W=sT201?YsxAxTOMXV$PGH;M?<7VS_1
zeb@x3WtF#Pq(9(fE|K8t?Aj?jS(|xOpy<cpv3d}eT!a$e4Usq0GqV;btsy4<`Cx>u
z68>w{=I4==SS=_8(NkEvRueo(gP(IoQ!PlsxRG5$6d^Buy1_;W9uYP~Fbh^cA4kO+
z))%JcxIq4^-+Fo1f%NkS=+tL#oxmpeXgR1l@eb*3@%x@7){hy+ON;Bth;;SA(^ZjL
zQ0}9>J+)cM$uG(d+79<tV>?8R5~7d01-Qzi{_s0YbEIs%&XkFcAW`6^r(RZD;P`ox
z>3NX3qS_GaY)q7!S;~n9wU9FX#oUKR7tiZPsr_mB<@nPp39`g(^vVijS`}5UJck#f
zarb$V&_4lw*e&MVPWHfhxE-BjhRuW^<#OiNAEl~w>24=R2SmOd-JIM$zO0_krX%3L
zq@R=@#k<_W`bVdBJqMaf97ZMPLwA<;kXEIS^#zaqMkMFbYmz;tQhc_S43^E`l|33|
z|MOTb=(Aew)ZPdU%XDkM03AE~?Opkau*c0H&{*<l?o$kP7DCBkAN<x3MeMaG!>Fs3
z-_3oVlj(x$Q3~E|h!55-&pAqbj{C@T0xIkQkOvujPqBdz6$P^&+HND<REBsi>Y~?p
zerh5gtRGw+EE|9QTe5gLI<3wsy(pB=<BmE^z3tLCxYYM#panq666lnc7a%uUA8=%C
zg3seY#{;8U8ETe#YqG1BHt(3D#|~2`lot@nF8{j{Ex#hUB<3R>uOXy7LO5Yv3Idnx
zK<|qq8Eo#NZ_+Y!Xjg{~!_=o8avp`Ih<KT9lk9&<jf9?`5xwcL>WDH#eThY9aYr{D
zQ)>qz86c`3?Si87IcZlSbB?mpiXesgVJ4^s^_JTYQcDjo^W`2}bpYhGJ5F+%8?_xR
z9U`q6_^JWm=bi)H`t`9G|8taE@mJ8gH1Cd;nxMBYq~k;Bl`Lsxp`7l=#BctGXQ8Cz
za!~%K7CNENw9>8XuZMUKj)wg1=Lgw_$mmtmaFvGy#_E}{yOT!S_L$N)ffr}IH~Ay1
z5x5!}xT@57nNbCyHpQZJuqX8H)^I3lUufyhj4Uuv)R>Bxjd^La{<u?Wd5JjUs`anx
z#h|TkpraTG%KwzW*?$=I|Es&r__O@+|6L5?R<^EY&J5zVMy_U}W+o1%X8imxF0RgI
zMs_fs8{69Bv0Lm&J+m68kg_fn2Ox0RH_c`6hv+8uLh?e{Nr<=K1j@yd($q>aRctO}
z1wU5xM$JhRyTXgFGt=kGv!lzi`{^Hkhs2TV^HY(^!|&og>4p5^87i`BrR~kxt2hg)
zmXcXmKS61eQ^5bwru6x-j;z88u_DD&I(?n<Er;g*6mVgmQJoc?bxoSIb?+>AGXN8V
z7uLvr^&ic)?dF=AxAW4ZL4<nDX+Z{ij_dYHW>;4>0fT#vnp0o}ng&-TW4<c>=ykQA
zhp>(f651g&iowzX*j_8%8vOmTf1Q*mB_h==&c7L~ug6K;XrIPv2wB!ec9YMj^}(>r
zN@}OIb&4PLY)|Z@YgB!lw^$a;Ks^wcX85LVk?U|g>Q(>N*F1UFwqW+Dm*0YA;3lp-
zfy(Qc2Kv`#OE88@VecU3|IYh?ED`jOd}@qOlCjrv$}khFTv)r-Zf>UzODba#^FZk&
z2e)=2R5w1-Mw@}_81naxCjl#9dEdDHA|@pur{q1;tC<8R96EI{FnGM0R%SLF&jnI0
z*RSzzs?#uZu~d`e4%G?6Lz4ZS`LCObN>2}r@+c<iQ1@*PDnw1&L$lDoD-5HW&&g+p
zL(m@UH$gkFjX5pv;E+L=xX@~{Da%_sSqld1l3z2SkycvGE9cbSL$2;c96hn}hU-Mp
zj%P*HJsfzf9#S`-%>0R;O(0^|z(N9b;r%6+?V-V<_W!C1KV;@ZSm_VuR1ppA-9xll
zXV1r~&=J@xx%hh<zV&x=zx)o4MS}6Mw?7~6slI?<#`{7LekMFVA3dG{XCCq?!#G?B
z>Cm(b+Cx;D)q)Rr+{j8@Oq#KqhSkwKLsjgAnmzR@64Z`s1<OAZA}T7R?pj%bD(88w
z^5i7D5|-k0#gOuHeldAg*`-ZllV-&Q4Q1R|Qf2gGrNneOONKA>=CQ4fn}t%@DC$)5
z&<59OTUD;H9fxfhhi+no0s!+)=1u<>{hj7gPpcDkt5H+CW~hl-B;a)n3x7vc=p+s6
zH!CWWL?I>T;0|Y`cvteO*`(F?7W?8GX%}e=z%-aq^FbHx0FD=z2}qeMn8Ohq<G99+
z`ENz%(yB}cgE*7i+xqx(ue+P3OAe!!1Y6B@=htO8*b<rB_lZcsR))%c8ES?|j;p`V
zWM!i<4@bGlz8c;~9E(+07>x4kM*B$eQL#ZZ^+kq8qh~Hs3?uK;!d^U7%sbGL<+@!(
z@r~(VA%uie>airI5UVFRuW@LTAWF=Ps3lp@kCfZk=h&qsO3TL&^$hv67#w@m@!Hq4
z_N7hVh$SZ=<N}2cTT$rJI$QC*qwb)oO>9Eb4G#-oWR~Yz-dL3G88HuyaSCUM<{rFQ
zuemBSp<^i|bc)%{Y=CI~^5m{@YSJ=w5Ab@b0SRNtSZS4D8KMwo3u@Fg)1HfmtC{5T
zH^q@lfVj$;If%t)wrej#&DQ6c?|eMWAeEnVyn<2IFI{HE8LRaR7P{gR#gK8o4@>xF
z{;8!~1TJ>7N%oK)Uo99%OLOlHN%w{4sry2C8G0|?UA6{tR=KN0mt~Q9(%(Py?hXW$
zf(r9B%Vo&naY+W;Of5e_k-vGp4lMXszd7CoP_i9?XD(`oa^3jLTBM*pACectJ~|L;
z`snTtY^+^K6Xp02&+Oq5FY}%J!zb4?(fH@~BJ&sU1=<J9+&v9Sv?Cmm%f0sxiZKwj
zd#33z;DKZ*2$n7uZnq&y2xxJHict#X9o(_Rh%>!6Hof@^4`nJ;(6P0}tqXFvDhh_n
zdsIZCF7E_IqLjSz3x>&d?Jb?CVo&hN@BqBkCnoTdc`3e_6qJ8ACJ`xty+=NMND1h}
zmbSg`#KAcJj%C32`a-aw@iQH+f*tsyf^;bSh0gK&aPA#&=W_jcd&pAacK)~%I!>&^
z?c-NE5$275g=Fv*d6g-v6R6rcMtpuo-yBsfO=`Rid15MhkiRr(orLL*bz^b6AmsYA
z<hXZKxGC^g%=yumk_Wh6eYPvne;!@b{&Nn0i0!2~C2W4&LRdei(?W#v0tEW1@td*W
z*bwi3dV4r_{now6W_E1ZK?6HH#iSRDu48%mxcfC&naaxjs~l`UQB{DgJ41g^$hv`5
zhi0*L{NLn4#p{ub)UWrE;uLC)?)4hlOvEur&AkKD-_?WQzTQ1k>gW{fZH%kWt*Xnx
z2dF!@xxdw$#ScKu3}ulo!*^ZQ^dMUORERsO(%PIxgOG*yxL@)c<AzYm<J0_BV5FC3
zYTt~7Ou}gps?fF(7HN|`VS?++wzUkGb}EKUx_A{2n4}StJ$3Ick>g};JtIcFG#PHO
zQw2JPHpdm{Z`=8I>V8fdKd0OohTG)Z!5Dkyu{&mF8GoJ|ai|c;Q7VoZlDf{d$Wa4t
zNAwi~Z`Rhl%EG>>?BpL$R}g!OH1PH^<>U?=UJn&BS!e0YhqQg?UWgBI^zUUG;>C_Q
z^#oY7Vr6zVK7u+ztP-}%W9#^ubg(*-Wr8!#u^5j+8V${v&fLb#x1`en>Je+u9hpY!
zst_VVIcP)qhti{c>52|-FEOjZ9EWn3^Q%Y*?oVGMkShJ%dBP)tR0dm~h2d=}mdrs=
z6}sd}g5|sFbOyN~XOW!U<mXLr9~3UY`(=2fRbjgC$hLdGW}#)P{RtmZB(LcRVh5{g
zZgSN(WWS1G`+rUloc}vPuyFpDI|(>au_PRK`gLQ7tkY29j=(S=m8vBWzFFeC5H8r!
zCx?Guz9AvoOvzo8tk1&Wr-)LJSJya{y<e{L!|NEXcMbeIFtM}m;1I&E```)jwE2Z(
zijD?=DQyg2zTR(UHEQN3?&%S&?4o()dwsm_&!T`CNCd?Ln?bUBIYF$L7K}~d8oOI~
zDR=d(x{_bjkA37&HR4+alEW3g-@eYH=z&(Bp-PO7@ZeVxkTqeFuf6$gaOJAUy6jyC
zgTN^6KtP+C;{_+~=CpFx-qn3-9vYkLAL^xQoXQO7@mI?K2=hVp?lLJ2revHyNU%nh
zkIGK|=*1aYf5lRD!}Aw5@Wmoqvm{^p4vzi0!-@*K#idw3B<5><BK@bh&2;HB!`ST9
zkVk);QM-QHDcr$5lf3*RLA(m>44d_{9mNVp2|cUixOET`*|fy;YK+H-lJ>#%GQ<HF
zn@B3A(wch_qS2bt@!HIHdJ#Ev=sDS3mCFF$=iOw~1f+zXl}4n5Vwy@gVi@bIt_EWq
zSsK^Ups;ilVv8?~?t0D$Gm~+I4lV`L3!-m{uc;)Kv@s7K6FIGmPR4EJ9?qaeK-*PV
zqLtfbt+bPd?6vHc*fE82$}Jy#Di!RgQ2{j9pz!o|=#{Gb$wVAMV$m%tp-IAj$GFDU
z)os_!UNe0AoUM=Iqgi2peZI?F5E`24A@!&OTfP7U{yH#xpK4rIXXj>~k<xBxA9vJi
z4qupY{6ZIK^iFAgMtWk+jN*KBG4E{N<7kc}y=Y2xkR$&E&M$+PP3Sl<T8UKT1e5)L
z7<<Pa-NJ^;wy?{#ZQHhO?y_y$wr$(C&0V%_S9NvLr_;&%oaFg*|ACczjdfjfj(J8Z
z+?trw39iC|p)%G2AC)4xDX8i_C61zi=IKnuNQ#KlCU$Iygh6B@sdv+^?7jEq0zO7T
zmQo%Fu;OaU@HdF6lZ=aVfxz=XxMOx7WX7`LZ^}*-b_Gyfm(WJ6+<2$jWID_m#V|B<
zr!F$Ofno1{_dp%(l>|AgCL(g^?r?-Fz*7xkcRIVM4;9bJ<;2|Y&{Rfro<ZiUvq+DH
z!DxFvJZ6Ui1$kK^BA_=x1V#5|7a)sv+kim!!PcNjYI~Gy{#>2u^%Hz}v`P$`Ho0aU
zyx6=z{0{GN<&@nm=U4C(T9kL9x&@^*OT?HXV&ZHvv!NxC@(LNU&QRI%z1twE<wt=t
znIu0!W@C$(B{PNPx5rsk;($`h2YD*j_gCISw+T+vP@`xk>@~WmB&IxsI(i~utmqKt
zm~I5L3F*~YY@n{>XmZWDoXLC7chv-jWM56YLCP*9%fMKHf>JT1&D$yh4fR6=YKr^f
zZR^ucNicant+?B|fG@FvzWtI3L!)LjwDVB+!kr2C%PuP-F<4_BCtcQtj}iHE7S2m0
zzq1sg-W3O@b#nR;IYR48%0v<WJ)^=m7O=zueoFtdzY+a}VjH43B&*QdE~4U;W>4}i
z<7xM90yjmHc2mPM5mTH+6Mh(>(|RvLO+%dIzvXK7r(JH!!?Va`c_NgxQGSK1R-D1-
zbnDLL*3dF!;Up`#4c#b81H(a}Ot5{<B;;UfT0M2R%Q=vvzNf<0=<h5064;0UKJZw;
zL~IAS*o)(5lQv0lPk7pRPH3o^Z8Sbq^-9oUTbX(bsr$_fO_4H`J|LI-!u~9kF4(eL
z!M`H3Fde4nmaY4^#xafx*f9}7X>>6j-4zJp8rt|TVGZV`((0GJ#a8w3k}iN(4@oc(
z$m0`WdkC@C!InHcIki@Z@0Gi|<k|_wgtVuRSqZ|aMJS%WiN55gk%ly!L0R*;&6j7j
z@rScCAOuNs$Xe+zf>ptI3r%eqM<5S|dxU6o$NeQt^*QWuMkB`*D8Utx4<KxA)NT|I
zhiDrgBefP*BRj6=!sS>vflxA4HE%2pkHs-$-~mOJ5n&%&xW5f&-{Z;^$Q0gED|1rl
zWpSk<U3A9^%0CI>EwS1OIW1Kj@W`#RmJ-HGl^>N|9Qj?H=<OxX=HrF&d#sj_{&b6J
zjr&%Vw3P-<-YS(`u|8lz{J-U^LBE=bIDJs0UK|;_2l0;m9$bHoO)9L-(Dn(?L3NQb
zc50=zg$M3!C-xdOj+aEf1G@rBw`<29k1Q6Bzxl>EGxD^7DX46Q1I>b!`ii)}@79z+
z_D8Sg5wnZ<6;CdG$%n0~)=}zJ76%jG8m?)ofM;j-%eD~j^h~%Xmr7(^LRR|(oRL|s
zYj&3y;+lnhpa^N192JQxw>g00?v>gMjw+(xZ-lHRiOr(#jo11zCP#(3dUA7<&FoY>
zw0j}Mx!dMlAI<)NdO@L&U{9(Z=}-ICK{$?_4#hgxh)<w6&)MfFR)d(qeJeQ^*iR)%
zMnS}*DH@V=A__}9g~v|nq6c-vGlFa}*UBx#7Tfd}_4D~uasm$6wgbHdf5C42Q{ClK
zWgXSIb~Xf<eW4V5tD(035(UepE{7~vF@>o(@h*lh8!ZkG2`PiMF%BPH%>^1a5}Gqn
zJ|>aaW~}&Y!&`hs$T3fcPe0^tWkR?C<7!C6rYAd|o#+y$uZL|!o8_x#EM7eJI8iY{
z;zs8`>YrNOZhwl=#}dV7W@Z?zTv^PyqxG@B4pbi(KhHo>)+xYi*LXL$Z!guz4e4ku
zW$gW}k1thCL-A*1FLh2X$;EFCeIgo#Vn&(ds7+_kyfzVdbS24+V>t9;RS_k<4SLLI
zh`F64<9T+y_NG1bYH(eceSONu@F|<@pesCrO+^rMX16>v@071I&{gu+c5Nt;w*CS{
z6q`Za{y@(CEddG=J&eAyRRI(PR|@@Ls|NTA>+Ie&ulVb#eGMWt!S#dIy_uQfa@6ku
zRztz3o7_3Q4<+#q+%xFw5RNt&K7Fy>oc;0O1J5Rt;k%HnOP?hBO3G%D!ptK$ceb@U
zvR**+?jVMm^pFADmkd=0H!OTaU%D|&{o`%V0#C`t)sbuxKGMEkcn74poHVaUd!4eG
z7!;Jp9;8}1jj}Yg%0QO(oosL-x=MY2QyJycvR`Gfx(wWU6XfsE|AhayUD1ucb8E>q
zo@R2%TJt7~gi{*&D-A5_O?kZ6<t9qvkXY%R#-@`|RCD)o^<SF@<p&5<r7zpWCdDz|
zNzTBaMcV9QVjxDyBHz4Gjx3|%xh?v&gGLe%3Ga>nW^L(+rE9-yqR(NjzIeFN<}Tew
z>%Rv+K)el1ng7UUjEw)2%@|n!JC*qK@97Q^8)EnB@99p)bl5H=@*m}LEtpr6LYqJi
z7^_?_^q(L70Lz3am2)dOJVi-{bR%jyRY~3NuCdA1{~ZgPbyY;8$^wl{jpd{5gCq70
zI20(PL*+*gm>m6^>f0?D9#ZStlawUvrP1U1SzVh8jy(&I1IP_&_hp1~3w{TFhrRSk
z^CsBS>VBEC!}i&eC>f06W}WAie$mV4;~t;BJL`JYQpTbRQBp21^<6O-5uvu!)!9+(
ziE{03dgP{N{C812yT7Tt^`7}wxoudh==fo)pA=;sJg!5LfC)r@>t(37<PpoA59Td>
zlKEA-Si%rM!Uf5D6@NvT5H6EB_!G6zJ)IkM8$xTO-M+HTsK`tMG5p533SPrs6vnG}
zD^Qj|JT+~<4BT6)ip(RyBA88}bMIo~_Tpl6()VwLQ+yEXw?Q97>CzFOZ+8$FWE1_$
zU-)1!=eLiSNj$pVm`an^sh;3Nk|0<-ua}qHRQ<zND<WOSodxu2F7HBw3_x>vs6GoK
z3ze}w*U%H9KTBIt20d48TATS1&UGnaiyT`qPeBKQZ4%F<^hNUi+xG}-r=pmw^6q=%
zQJuhr$URPnt8CDa2VbwE&Lpj)=5?CnOo{vghi$!20|U~xllS2Fo@uxVy5u*o3qt$@
z=Z;UUaDyyF_?BzbC!)IvhDt=csh18c;KqroGZ3{h%U-)<#x|a>rlExlZSzxD_ZR3!
zGpR%};k|38wmU00o%RC&)2C(N+71hja9`fru5#+99MAP&*kRD~#9=}gl-RvU2!_~d
znNz|MCKe6&Y2%P%=o>Y;@Of?`nYCImZmrT<S;51Z`X5xnQ~WBp^J+mrbjkX5QU__H
z7!Jm>x56fNhxF(^Xm%q{%a}A1ve;rolnZlLyz0;Xi@T`%<@xE{vw>RungLsmb_Xg!
zsvg9O+ZyD;weA}tRMEVQm1XryEcRxy0?7_cZag<PE??2lm#1DZ^^EojE|ey}<=D^9
z99r*66K8+~DtmZwDATn*2?(vr51sS={+!)w<WMowkLU_s!*j{ua-%gcq<-p;sUK-N
zj%8yU{(9QYCs7<*&%FY0a-*=_9|#C%F6T?N*u9cV-Uzc{i$dy(bYcXa@gCuCm!Cf$
zTe6u|I6|WSA`a4U_(JrI;|ym*?O}Vi2PC8;sJTt<&a58B*>f#ITqZ*oicvHh#<U+v
zE<nj|5R*O5i!uSMv!`Y<0^t)MHH|6@5Owrn)#aG+l`rVx+J*{&iQX+?#Igze@AY9z
zn(A<|DDn0D+7wL={$=#YW3RqXMNuO!yp~GTkdt*FMB<8CD{<a!L4eVMoGdk<e@*tN
z)%D8h<V#gL<D(6L1sC?rH-mG^{M)S<01hPIFAlP}#E<#xdUWRkOl^%!8$4+9;}h^g
zn=BPt6`>TBK;PWVAtPi`M}lWmD@_FK2DlPFqBJSk-|dS{moxzw?RxT8mQ*p)-FjpI
za~dN-_G+6bh#~N~X10Y;(ka9=1QlKMhn7j&*B6uI?1Zw^>%2meXF?F&1;7tz=UwPi
zv0jjV0lPqTF=}D;UngyXwD}-Q$!C56@ngNwziT|lHH3BO!F8ynTE)>GaP+1@NEW@r
zQw*i3mTQ>VZOG*JYUM3!kQ^>>;?NH6=qgvJ^NR90OI37f&u&cBfkfQ|$rz<dZ*$|I
zlSd~vM<Ogr56%6OWTpK#yAI{{?eMw8ri_>YRSl~vXqvUr+HqmWsd<MG_%ih30xgx!
zWv0tU?fEoV9n$1)BV`N-ZA+$SE0Y!ltddGq{bT?Z8_ggbE@&>5?vsu%)bSX<C66E{
zYz-bnN$I`UX90GiQIuGEMMr>{F>~NsEFwG_=r3itBC7_62XJ*HIkCQX?*@5ov4V_o
z1F(SzB^fNi1t4gB3j&iKmF`T;eE9N|z3}z%(MD<1X+In!K&R!KYE}&`Zj!Rh{Q)f=
z9Uf@{?<cp$OLWAM^6e3_jm$H7rkTDHP?pelpW;cs93-F41tWAUkgdb<;Z4Sfgdrn5
zhk`1rt4tKiavm-6o6gmeyM?C>%HK?`HdW&T@G7ClUiI|BzVNT99xwX(cO*wfu+|^u
z{ZSb}jc39pB-bN)9Lq)Ky_uG9J+8W=5c1MQ4%4aH6Yg2KHd^;#&$FBAS(!U>Qh9@O
z@WoqcnYi-Y5`v+*Gio^iszV1y;^~ge&H5`tb0Azb^u6Kf3!&TG38p)_)&pG0TNWnF
z1T}bRKPI{J>0i?b)L1_Mlm`p8HcfN(^@=Am`}%{-Shy>BV5BW%C9*81W{%}Qzdti4
zrlsHY+yae@^%B8C9P?Q*Y?EX4hCY9=B4YQid!+$r_Yn5f<1;3=eEM%FkfYJ??bYq3
z0S^(|bG`P>T_}X%6ap6@?}^Tw@}55Cd|~;?xg`-YL-dEA5@#>gc;*grUM?f9iA#CE
zqk07oONQ%qF`XO^EYiG5UE&<ITT%=k@hN?FW{2z<J=&D_rJu-U-gYR7Wajh=cSjF#
z(O_qxWVco281k(xcy?&==bBZRo(gC85K8|fIMemx-rFK`7Kk5INNp8IoyZp;?76$)
zqHgv-w0%X?bSUZa8EX<Lt#t<;yz50&Hyf9nF?tsfRk##!Wk|-8g-Q6YmwDqsID6W|
zGJ8td%0fF0Mvw`2=nrShvrT;i)?cIZko=3je`T1z#+TvPglY7xaRZ}F3*njPdxW}v
z3-x)13v(SkrMR;}9*qrm1v8(xvne_&jd3j$oDFv!tkAlH?Vel74LkAa=i!Fis68#X
zz`+S~-5qio`{LP)*(zd~`wpyUWuFop)2eyH{Sh0p0OMTEbGasH{UKoVaWwwm`*?Z_
z|BnR2#PL4~hKcFFYhO<PUlsc;9=MQ$i8v%UbXB*#-!|rD^=Y}8!$khuPblnPL>3SA
zhU5aU+HOP+?3o8F*r&4(iQJwp?d`w3+W>O9AdK5N+r8P)y>p+dW<IB<htUeKQ)geJ
zXR&gcTBEs8zVem;-)`SHubDG=eSdKE-GcT?N3!_bm_Pr#LfoQDeOA92zO~ZL+8Fm7
zya^DALP~G&zr8->-t=~#4&07v(OYhE#zweTY>q6)l_S)Kq;@e1eDCPKZ6vu)0W!G$
zJGwi)XEJ15J;R9=^Nqn8SY;eOB9Je68YB#V{a}p?Cmug0S+u2OhLrdYXO+WJ#Y%~|
zAuM&8=IV<(Y$)QfE6Q>f#I0>5H$EuD?PLP<RWeh*ci);IwWJ@{b*7Er^YKRb`QOvd
zC&g+3AR;wf955AJUrz@lyrP3yuh+swUZ%2qKfK<aZ#oOq>&d^hm&Ac)M`5C1t<tc8
zaRPD~b(9|Jqi(|M<^nni$ghDeeD@aSI!V{150R!TgGxwuv!7q%X`<Wx(<Fa%nk2u(
zbl33g*nP9XxJ5A$j8yvh>G<j?PIeAT%3=g(TxH)<PEf84g?nulLjt1rcJG5;!)QeI
z4k_1-sgtIE?$`vXX`w;CL^Hd3la4%F@5(!Ct(IjH;OgY~)ecUBQUI}2ljb;upUuEE
zkp&3IA#t(Tk|vDsrI$!`KZPM#W|{kbM{Y1}89k8~Q)kzao-w;%#QaMY<R-lOF?D$e
zSs_Yfk>WDkC1?|*0<WQ@SSF$qWL11+Roa5rpTELLBrNh%VqKY%xib_8Y7!+@!Y4sQ
zV@uj9jcktc^drnuxz%q=A@43Wje(4OGs87Mw!(Zf8FC0;Zffko_x0{hxqL)wwO^-P
zyMEfemCUV&PwrWRkO~=wc}ay@V!)S4Ee{T=RKwxfY^UUd*!nvnx!3_tX451P_nNKA
zL2K3hQL4of(j_X28IF;Zm!t8_K||Np@Z^C_qIp%vEdHAJ6@mC{3u3v@+bhhtJ0$81
z@g@QAih1dS0j$?)Aw9}^@jz2xrBqq8{zejeVcCR4%SuZPq_9PKR<{NQEK%eETzyDu
zOVDH|41h}28AAJ-PJVgpj?WD2lAnsgMUilDx{V{Ea`O&{O#eim%+jcr$l<DbS5Oed
z((8Rk+%N;8=QH>K!5zA?xxYFT8~-Q|rHkE?ThI2u(znzaOLi}CVZ&x^g%8c|&X0*y
zj-vKc6BzBpnhgeKqaC~nA+AfeS(;;$%Gzm{8Mw-2>0z=}f4{DWJ!?KQps(@76OzeA
z>JutW*jJ93h%<^^p&Q1ddCOe3+5I)NhEX*)wSC*-Cf{z-0xHdFaae{FeMXHO6UX4#
z4|9m8LY=*+NsYu<E8<huTl#(A6a{JZasuY-9<*_syVtiR^EzSb`!cNQI{+L6?l7n?
zIE~f+1?&T*RMYE5=~noe`RXXm-?REuJwtXtyj+Zue7Fl1&e<G8to5mTD_(@+U{B>@
zDo@+hAKZcuj;SoBZ6|0X4Y+f^idXK+>-|GvwKt@>f0T{b*UQwy4s1&%GUbSwad?zR
zw}w0!rDF6$2P~fH01CZ|t5ffUDlJmVxWf*M!sk@$QFVM;(}-zYW$5<=<-+a8fxs#7
z=dA&!%~|w+64ekL2{_L}v|zO_>uOG9`U46!nA_AipQ!rEvWP%_v-`Su=NLSCz~AuY
z<aKT(bk{V#v+~GDnFIUp0a*}M$6o^(W4UwhHfR}@d9{G^UDs}0nd8NDd8S)~ik7mA
zv5;AeWqSFl9pXBX24z>+oaYNqi!$<!J#o{ntG*laFQ497!2Jk>#+TR?Dm}L5SJB>n
zmVs={bt^z(b2iR0J5-!m?q_4~4i9`fPc0Z>Op={Eov8eWNH==@1X4G;z*)F%)aKil
zZdB6eT{k+vG}ZFf8R8xlZ7+|!Qa<D<$q@#R6g#mAH(V$&)Wz>j-&;Mo0~Zg(i=%xS
zNInpG43f1fdH3m;Y@$F!UK{BY<@a5bX^;kvz-=A=w8iDYyeB3v)HH6LDnUDN22a33
z2D8+G=oETzpBN?!Yt(+1j#b#E?OfI9x3zTw4h^D|4z<k?Xizv^PK*bu=>b_I6eEVK
zt172RB@gxm2KAH#`bxSXpVEws+oYYlv&jC^(5=EuR&4R44sncz(qb8knyN7)*nD)Q
zPWR(Xl%~+X7H3sXo1;gNopr|izqSk0nVSjesZb2ckV@9J@*o-pd8S8G(cZ8ZKqrZN
zUe}z9J99^!1Lj%gS^8y&%`m6iE@sw#FEisq4vO<Ekh;v%Vi-kE;#`>AOPdOE!j_u9
z!E&Y*bmI<u)XhBfNkUQNjfn*q@4q#BF1xgA20lH-omrx?vR`d8f5fcCe`7lr8`M{A
z`R;2A*hS~Z9{=32nE{b1hYw4nwEReHCz&d8AC+CrZ3%_g#_XUOBkIGr1Z9A(q_Pj7
zw%v{|1APm2S^B9yQA&yB(7Na)1qQjK3KtVlHCi_f!kMnZA?LHx@L2o=NoDjWug%DH
z_sq6-=OtI{avCs=NG9h>nB3ebu=1(i^Qr2|H@g$AYGFxVK0fUPEUb6EViV?*8>Rx(
zw5!H2<dTh3A(8J}aBDk%Mc9C?n=_1UATtV6pTbJ9sgQmgoZ`60QXpYuKxQW^y;^QM
zxd1yO4XZ50^Y1j}gV#jx^pZp4=l8q=20LY4xhySv!ZUpPiiLX4a#7lF<#Ya<VbbH^
z|A_S}N?VUeh21+hi7k&yk?o@8%npKU8C}JalT4LQSEp?I!si=Po1#qG&P3h39Y7{`
zI#*eH)J3k6tJ3dneN~JO=7XRD)@Z`$GO9rs?Du*!OO>Qd_j)h`-j=>iLn9i$S1qdM
zRY6431sjSjQKYj!pltlqVdL_;H&=9HyOU+Qj34ku;0S*7;G#|G(f{lqpfvhb0z?`h
z#&kRupristYN~E2gZC<~AC@SdLD3!qlZDOe559+o{xG}#)IS{=v(dC-lFhJnkc87T
z{{#mDr-Hx-z%Qp5!repi%Vu%7fXfhXHBK%=#m`ZU<OueDz{<{P9CafJ|2Zz(fsXnC
z+3;9d{ExWB{2y?Oh2y`=++Y1_U$NTnYHCmM;HUR9{RoUinK1F$upFTr)Y0o-^uX2*
z<96gwG>q8RwiGYoqk*gZb5h6~)NjztIlQnAZ(BTlcD&imncNVl-B`MOaR7YdUeSnQ
zIl0^b_L!k>_(FQUk`y_-a_$%uEuv#bX90Jf>sI0IKzE`06#NFavHB3Ai7-#dBW=TD
z2WE3dWxKZb-tnUez`?PY`SxyCU-)i_i9X+ZD#0mfsv-l0{scx0rs8$(fM&h27Oinn
zI6=F-zVHvWftTCQ9B5(rv40|!dcBUe*80BRLxF4G(rFTa5hVqqhv2<cJ`hrx>%Hdm
zdQv1oZNuUd?nh-Q0bzt9O$>0}8vol99WW~<UF=ybL**6cO1JETM2lKD06mZ&CJ>-5
zH154w><?IO6DcS!5ZD2StSzo>>!~WJeU$D^EpIzt)qipR=Uwp`h556$;eM9#FFZ3~
zmEPDKqXAq~s+94Wy+0ox4GfFXU7DbI1|Nz)9KAbMXsbd|Q<QWmh2W10ep3GZ^{86e
z4fyoun522Ih-3gbKv8@j^|TcZpYL1$WY1nf>o`~7T3`$i`~q$|l{_*y<r*c$bU^jJ
zRnAmNMK0|6{ze{$1~U)T>)Y5(sgbJj4#|UsJ|R$U2!v1g%$N9SX>}KYE8!sMHgXoK
zTR@6!Q?0ayy<RZD!Th23Q3RhKQOiFvFJ*(lP$fSRcv;BUp%?tkr1;l7IHt>;u=&T#
zyHBmL@0kL+#++3NpI+=So&{t}jCf<sxr+kpQ(S~dStd6zdhAJz*L)hDuk$lCLCU1|
zlxIixze*?Y4owyK*>T=A|B}6>*P8pLx#l+^bMuBMHX+4Y>|P<J66m!I5m7XhxXLXe
zzjah(mvmq~<0w%X{Yeni8ObR;t)}v1Q=>TK*h=rnh8q{p?f>B_C+D8PToHiu69co!
zMi*j`UeQx0DFr%+M4}AWC6&bGxM4PSK@W!gvXpv)Nr6lK5!^vi8KiL`c58AoF_;bu
z37Gz42mv{>Vg!tUmW++B{^P?&hXl%7Jq%hRrFnlmygFNPl%6aWqBWgH-ox8+(@J$M
zL-)Rosto-f!1<>zw0IH`+6qHP#If<_&EkWB23H&%<L4S*ymgH%fGaYLxpw^iMQi2(
z0|RA5P6L5@mUN9_i*(}{=ES(WrxkCZ0H4faLsCgV42*YzBU1GI!V&TSjn$oy+D<th
zWfS=KHe|rl*kp~?jQH8CdQs`2Ag7u26n--51)FYEfWjbM^<KHcD~eO2oFr%1dCq+2
zO9E7l(mc$D^n^xdMD=6|5|&H?jmrwo-ec%>36mBA$;sRyKa;};blBwZH2-#7Hh#*z
z;(Qxu!nExn43@$(o!G$E0lk5h1Axho4|49AbCcT4_vKeJjNov6BHoMyBd$T++?Qpw
zgCK$puwE3h${=Q@EaU_F<*pz!-@nHM2bxu6-k8x9QNQISD+Qx2{r&<ABWtsN20ct#
z&G<`1G|m35F;p8Q;QjexmI74m<h=Yl=gML$JfI;W?8Y20T`Ug;cKX_(CkIhb7s~oc
zJ0X$z5y+gp+$(hzTx$VxbA<1>xuL~(5WcTt2zyWK=zT?ow?<kDtclxES&d34@>iVq
z%7>e>(dPKV=hIdV<bloE@!97DqYFg28%5Y(fh}bopbRXx0K;^wt3cZEw2CB6Y>iig
z<dyr1=Py{+e&|<IqFqU<go5e@Qd;KR!Vt(1Q~!Kgxt9GFbc;<2rX~I0?iF)1jmc}&
zMe+qza5R$o6lcpmC6HN(qy&Hh<|OHsA&r!k;j&(6xJ^xNnnl~D4^uC_t4`^Pye}N|
zuX9QujPupdG*O*T8I{R8<Gcfz$w{MiN_z08*#k@DEQHQ7P1B+5f5&5GoXm!wxqF(U
z0_zX@U+Eq}?&)F^rNj;tt)0p%2AQ?pPAZo1J}hb;)zf5mo<J3EcI#G2r8^;d;s-2B
zt9Shrb|{yLt4pUCP%hmm4xmSiQJ(A<u1QcHr)znPN0+Pvme@E`vP-`5x>$=R5@V(j
zpSoCIYG+tLRrs;$dP-mGF@}e0@<#B_mY^f2rYCL;kN6WvPfb}aKvFSE_EOF@v5I@y
zCkn3?QNDSiK_m-irjJ8e4UoU)AD__{Ub<x)X8xrLkqv!%5up{Li|*Tm{)iB*LLr%i
zCH>i7H;Eka2!uWBFo_(<F8tETpm72PRztSanwaNJG<!)iKW+hH;NMpUzQ|zqo^Dk7
z5X}*1OMo@=6(R`4Xt`8(uZbNDo{dV<ikvEwC+oaYW7Lc)*3AT=b;(i?0oHx^2LnTE
zEkD|9^P*YV*$%*f_=mdz;0!>QoWEl&mbxby*$=ZD2Co>pCYH&CyA9wZz>r9G)BDo?
zT_fUlA&h248>^8b+gs;;7EKnqB2hHg_9%@w3EyohKYkDcXUgt;8IA6s>TF(za4{Jm
zE%wHOU0cd!<hFePT(SjyKVi_(u+O?jB*znTaWe_DZ9Y;5XQK$osb*=I6$#{+zE{d>
zaRj3L3V0(CXXxW#_Dcu&+RQ0!YlB5^bxwdrC$KXH+u`;C|L^I?lb5fTsW=r$Qed>3
z`2u}i<om4~MygQo=!CkGXz^o4l#=P`b587vs`$<K>DwnV4_-9H@FT#Z>CM~RJnrtt
zPv=SC63AxouVDpEyTNUWh!HHshr{(Ippf!S4yvAGLiOlcvfaKR>+n8Uz9Z7edojBD
zY}nC~_niHVH5q5$%}JxGsX5(Bb}KlRW<>W|r${WiRV!k?GLo}(-JAj6ASmX!R7qw^
znOmKw+v)KAdf`f>M-esIUG-pbEucLP%=sqip~{HdibHad;_p-E1$Fn(EH;R!RE2Cp
z%6)vlGm3nTIp?1}V^y-9WZ1iSnO3Ny%fkz$MFuM=|5qPTuyP9Km|+wWuqSSYgmYq3
zHes9c--!gAAbM64$4+Lb7S%tAE;Oc3rt8hzg|-wSpazQt7a6{l)hf&^cwf*bP-p|J
zOD&ZnFys_08}-^U(%NZlubMzhnh}<@j>_gYLH7?l$NtCEB?t^4%R>62jENrnrp8lM
z*VvZ)F5m6X!W5{jURk)v%*rvW@<r=9q{iLsvfAenS<M15eb|!^UrIQ02%$bI&egGS
zX}e8xu0c}tpfAjP`fb#2;rjV0+UYm4n(UsnW-4OMqk>~jPI7`II(;q22g>O-vC;Sn
zc#lK|2Pq?uAvu8fkdp7DJ~g_V9o<aVQQ#vloog3^$S)jR{1+AWvUD9GZIjhZN<I6g
zo*OD+J%Mus?j<5I<dNtGy@H*0EiNp=-N{)|zG9jrC(l|goH}FK@vR<{*D25dmd=l!
zi-~L4X<x=T*{tX%*S%bThx5aZ>sh2t5JA=fivQeXCjrBMash=DAATC`D5BqLL1CE{
zs55KBu)nZywtWH8<LPt%BXs^=NdAuuIx_>uf0aT1J>#?GVuS5HQF?_EMxi(dK?gMW
zIRNUmj^8LK9d0SbZQ1#eKj%m=Va6phuLC1%5hbMH^(_$1`DPzqoRv#`Db2NskX2=e
zLH%7C$o0b$|AtBci{{k;&`17V8sPr=peDim)_xUIq#VYZ+`;)-rJIHPKlhZk_f`;Y
z!gpc&9A$64hj90<nWHxD-Fr{6h#-*CSjyk8sXN=g?g<*Zxi+xRMJy?QCY6g|Hp9xz
zk4005G%6plYCrMrnNs*0oemeh7|o&aC2(@@yc+2Ag&s3f)0jn$jGZ;{5qfq|87UK?
zFDuk3<6?=?V1vCjU#2)A-R0l)duv<sMC$ZUsLNp|mR5(z#((=nT@dfkKR&oV-Q9d&
zH{}*6$#Ws?<V01?r9Yo<$hh!k`c-@ckZ?}#7rmuBx!){`)1X>kXbBSFIN!Iez2EFG
zM7<`pj|nd{(eoXROqlu2To{Ui7L+cukZ_<7kTL61E=w&Bc|<_UHX%VoTn>@g6`HM|
zjx?=IKz78;FX1$yyY;QecK0bOA^h>JsfdPsQc+W+C3V^NWY-N?<}TqYP*9aRKurF4
zTi551u>q>0eL6%_F3H2)wn^zy<ko*0W}|8M6x|lhzG>*PN&6827Qj!uGXJm~Kk8@T
z3^U)~N6nzHxx+9GhW+J|#TbrrJFguf8*pv0mHe|24#$CD#sZm^WyltocMrr8LsbYx
zYmwOyEjcRF<B?lCKahZ-cF&qSuS6%H3Zd}jA~g{+t0U2&;w%F$f|8nE+v%yG3Dw(j
z*YlpVcX;yHM)}?|339mLAxO?wj@-NrFIGt=?EgUO;AHhaRuDI=MOnL@xl(RX0!njc
zm{Zhm<(^Ojs19rx>kYwjCE+FlHr)p=p#D6lucB;J9W0;&k=b-UL8PJcwrcO}4<=wT
z{3iZKe1<p6HZ-}Nz4lR`<3vLJ%#bh{C?CBVrn0AGPB;lQLPp8#_$>Hfu=Y=RNW3yM
z(G$mNw6pct!^o?Ir9~AB3)w>J+FQf{&1Xp0mP$EmH8uA&erJe_cCRJ20s3z>fpG>W
zmB0a7c&UVyDF{ZzWG<z0eQS*BOB+@;2pd+Rw3fja2U1=3v}#QZxaDtiP`>w)EQMX{
zKZ+S{Ju+qO-%cDOE=2`2YN$$M9Vr!na`DzW^_d?OVOHy;PDXuFr#BY|4YDI9`*tIe
z5O?)dn}v9l3kR4;rpT3PyR#*cmci${!5ue9N?wy)ifl$Ks_gmH6XBEJo;+75;Mn~r
zMU5r`Lyf5K76*NT3AiPg?Bv~Jhx4me?x*IMz(15U*?mv;#!Idwd7+*9o0!GY7aZ7`
z^4(Hkhs>CI&`pIlnV4)jItr4~+OeuXP1^WAM@e<tjf5HzQlF|PDMq1dG%A0Tcqhv)
zJ~5+)^R5zs&X~}4@Uxd|9P5hGSD)1=b@OP8H|(rm7Vos&rX!3z-o+QBmu^U>mxlLv
za+Frd7U*hrQ+p)%Zz7~~b=6dx6`de*S++~#^#I!&!m;}sfYw;`t#$?_NY>!U!kbFC
zw5@A2Amh2|x(h2NEn5a4bzfm_cjnZMCU%L>-&ZD3ICe1C3(>AFDQ0F<v_*sxs-0RG
zLyO5>d{T~kP_G`kTCcp3>B4#vq*NY({E$5}ar3h2l@qU+X%p3&cmZsVfNRndi_AV7
zw|*EObhAil8E2WVI0%v(VTCs*#c4G@WXp*}qW)d66@BI`4pQ_6$n1>BEJ-qR__(tW
z+X=cq;8dKkdxu&)f880LCy<N#f|;Fj@YK*w9Pl|jt{6_aR!DIWF?)fnoO663ZL_SN
z7M@F|B8={4><1ONoBi84975tW`G<d%g{7z%tl}}MNX-yw_Rg5K;3=^=mu$mLcT!x8
zKDfmwzp~<xcIuK~K;e^WRgd!Jp-AN&=G9{3^1hzcWLtbHpk`8sJZ8Dwi7|<JESU7H
zJsZ)In7|1maBR8pGzzIv+OL^nuX~vcJFf^AQ%s3Q7ILlLw)g$i5Yb$O9h$i?g7IzJ
zHG~|MDfbN25LOZwJJe#iZor`0{<TrF{+1K(Owd*OGw3;TzD3ncnt+k~sG;AxT&+lz
zx8vxz8ul$x`M}%re!5JASSMUrDOMf)bo+1FuuC;dd-lju#0fv;d0ji1iRV653J>~s
z({?XL2C_t@Q2&PV@^xCLJr}!dFqY%IiBfp7&45B8b~4U}RSZEdv-?EvXq{^+N3Cy^
zj!6Mh?*KOuEZ(jAZ@D4c3~p+8ERS%g@jj>dr({X&R?W)CAV{K}wnmDGCjpPD+DpCT
zOWhp{<e6Df%ZXpeVvE{A#4Nj>{h2^WIG83CbYH|z_LOFVo3)DeN(K!_P!@m+SIK=x
z@+&Os0DX~m<EF$rTJfyRmr4n9$3m-l0VLDqR*GDDawhwcTJhiauHtz4kNI$yMQsU~
z{4hbv_cM3}P79%4Fi)G~GF8gO{Z3{%mg?$-0R@*I%r<$8_+!BCrX{6Z&ow2Q50zPy
zGiSNCD`dZ9IPn=^DKMY@l9rZ03#|Y}A%^DnbMjoI0_0TnbA4t>DB5}<lhx<dfWM3!
zYq}s+HwvM$(>ay{fUBF`{bYyq{>c?7MXbZDS$*hAQB-^ps{9=OS!ZBPg1YRaJ&~%g
zd5LkINv?k(<Q!N44F6F49Ni8<=?U}7?*x-kaKTaWD&a8hqZ_B~At8cpv42S2NeE((
z9tt=Oq$NN8jh18FXaAkJyw|LDdTf3TV#vY=^CtHdKgf%gr#|7gAA}kNZ0K?uAzKx1
z##(cSs^-7*kEHzaD?q8D8`$(Vu5-K23vZPh2A+r$vMY`buO6K2HDYyr!q<eXZIFRS
zf`^EHB?*HlGGsO@4TGRs=l86o-Xj<a(kU0-m|xv^rjrvJDmp|MV!qTzV$fGhvX4l+
zr8TJFwh$N{45Kg~EV{_Aqsz)CZKAQrxb*^HahH`v6q^ew0g*9l;aB=S&@a(_r5PI1
zsJ9OkZWobm5s(fS@>ZCQ%A9t%YBMSMffy?CGIN5~JWssDW}*Ph5_})ND^Oqe(&??Q
zw_NMmXn&uaxH)0fnLK{$zi53mD0h}|Wv*;(hmTeZU8Y}BruEo(q#YH>y8M%nnPcZp
zU}C<xF1Kwwb0f2L+<oSzg(q~`XQ?#!8`$_B0!f&B&Pxk~9e&Gw{j)<Ca%G7#vgHd5
zjU*EM9|`XN^>|1d+n73;;s5RuF#lHsH~gD3!)Am2Z4o}gsUuKW0E5Cv$1;IBL^8DE
zm*PK<haUdG<*z78G(R!U<a;PcAgkReByfB4jW4E_PJS%X+5Ug{wj<j|_rFtSqJ?t5
zDKn7L7rauwoMPd@*<ic`MG7D4AD7SS*<57yYrsw5Hk94ZaYVbwebixZ`Cmmc*jwwg
zwXJ*C-rfJ9Hy_;@-=3dmVx~bwHzc2s+V7p(8@V5Dm1hMA>(6mNH1YGt)!5S80DdUL
zb`8kM`|ppJvyH_Yo@G@ISapuy1efWA08x_*nPHdwo{D0Oi2$>>1a;YQsP3D-z*#U}
zs&C`%X9UU7O<uGVa-AaS@y2=Lyezp1nUZcvnle#X=t`__Z*E6Q?O)RKDLV1X^1+ka
zDP$cnbt-6nzJiRqClA~%<cWnoVCvZ$33MlKTkZKeR<(yf@y&g>`@qv62H=+o(@4}D
z^0laKn+7CI<Pp5%arC`rOs6FGIc%SGlWjq*@7@xjx0cIzBM=HLZys^4z^C8;pR&@H
zBY*75^-7;19$YgNv1$xAzbo+5<DDkAU8-6T*59!aoyH09hMJLg7(@nJ*{7hEq>5<y
zi5H@41q%&LPTZBG3>o^Hud%umNr21?biT1Q**E_W5~7rrQve&t-^JRO?J*LSJXB1t
zeCo)MRK{5g*%=8LPd&%ViDJHAE@9v0jv#GA2>$!*PKA?<>*4uE=~Qs|gf}e;mr-Ub
zvyvU%@itVp$I&zD2Q!hG#`(lS7{)xR9kMR=q#e9Aw(3ji_>)mIU(#cqVDmdH`>`-1
zkcx4Yjh|bF^b<ghX41YdzHz~+dEo3l6Rphm>i;w#vz_5I>twW`vr;6z0_2=eU0+#!
z4ILjTwIVg>EsaHb0Rcs``0ihYXmo+XBi23Cv&`ubr+WQvQc#zx$s~=nFQOpMT9{L^
z)QoB_NhMo87aN5e6J6Eaqdc#YQu0(!EYqw|8cV|=&Ox96vuQQSPt9YR3PP4YDR6|!
z>KfJ70N|GPo8@o@9uP1?Tc-yeVYq^R$b;mmYs;hf35W<(eCSQ3UbM8XKjLwcl4B{W
zmEDY@;6CK2g-&7Cf=CImfJI&WBXtZNiK@!}4?Kn2#I!#lb*R*9%btBkJPVeKs<=-h
zuVLShQ~xDKQXJ^?$_Me^;hedP4VN9b#*c`r5`c|;ML&cXaI^e?!*H6^zp}G{BZXf+
zbGkinYZlitGW&WrRtdm_&~y-CP(jR%3^kIN!lcdsOjjAxb3dc6$PM;NEa`7KBedXq
z?HKRD-5H~O8nDt4H%w`#_Exj9Zd?YHs-`~c>qD`on~io+I|S+Hul5t}m!=$TF)BNU
z_Qje7XVdh!n?|stLaCe{aMkc_7#^oqPnka$0gXPky?N|b`Rn&(uR>)mb$m0sI8ipn
zBd5jJ)*XWU{ckyFSR-7)f8Lna)_-M@GM&5m7H=KE17l{sTDK&a-KAX$<|HFHg{A&X
zX#$dYY5eQnM3R{NSUuF6_A4WBf`!)g=o{6Xh8?@(ibi<PzooS`#Fjt^rFbxOYIwQo
z$Z{L;!NF203L0EyQLxcG)W+5}*S~zds7=qcaTudQwUb_RZNgC$;5i5vNn|`aW$UCu
zbA`j-%!72#y1B;`)lK2{D~#*h6C?uZI<G9*BZCP@Yf8Iv1vpZE-*ZQ4<x|;xf7o+3
zYGS2&Yu%PtJr^mYxE6)eShW)mo+H&IX0u6ge5~14VJRuMaeuQZV<dq3$;Uj%IT}rl
z$Ez{X8q@G_Ky=nq6?u;p-a<3wp`$%vHsao|vAk#vbt=Y*CAX=VvK<#q8cV_xR9P27
z#HED>MbZVr8Ub8p3!pG(-z2ms-=N|Ns!5f6=_#H&%OgX$c{j8SN2`75%F9DHUFJ#d
zlFn3wrzkr0Y9IqmW<XbNUY6ZDb$-L8q={V-(7B(HQ$dn^lSRcKrMKRCp&8f7yI9ot
z`4ce`5q5ZW&=Tkf-$o2}(gldk6H$Uw70V8W1mR<i7ygS<b|dc*<+N2SF`@~e(1$T&
zlV@&YTQr>5-qv`lTsf+4m-7uvIsCh7`QdrOn`%thGRTm8BVRfs=U4tm{?M0SVQ{~_
z_dI{1u+<vDm#ShY!A5mb*Ht!D-8wN<ZOVj`4{XY`6$TevRB5U-W+#E^F<1Uq?5eN)
zFWWv~sSgj_Myb!&Cmj3Jq)m)0MBF<8m>~KK!+A}>!iihF!I#|;e+<|(^P0Wqx(@An
z@FQt$OHy-EbEqdldpmXkQH@=_sSPJoz+fpGd-O&K_a$%EWd{aucUnSiFmK>Is2!hm
zssgB+Mli}yja78PvGN({`$s8mQIlTqBM}jR(qop*$CP^(>j4vZB~3JriK({RPI`>!
ztVXC!Tn%VsLqJI!tr$c*_r@5jLQnZ&MnrRV^>oO)rB=+EP2tlV1oo7yL1}~C*&iGI
zL>H6DB?n?j#aUnl1A#oJz!=*|#026wxTOL<0|2gyt$UixW4c7K;|uCmO>3r!HZuWH
z3LIOoPkAY1)pU?@UPu=r2Moe1oH{I$4QHThJ_r}Q9=X=jf$&2=Alty5sBytXZ8$y%
zpFqJqU&L1v0uZvSN?YdZfLo*J?CN3^cv#S&$9Et2Z3Iq$flh=UIB+nUb}Lpmpgo*s
zVEituI{^6YhdP*nk2gH90Ri_I%MW<?89m*&&_VpoS77`snm%`fyf1qA9PRln@PX*K
zB9CLr-7p~B05nxxkiD+g_zP3wZqIbEYK+&*ek;Z?7(Xq^y-&R&B0nra#!w$40QU6`
zl0Iu=xV9~5JF@Ine7dt|*gl+@RFrM}Y}mR%7>{y4E-?P5M&3Mh{WFSl^FEa6BY(59
z2QpenSZ#}m?&mF-E+km1*)eN9)MNv4V-JIpjkud@Xu|g68T`^zFDl0II+r8Dmqm^y
z>*RKu^i<v>Nx5<+)JJ7nx#I7u;-aPf46Yo`zc&mO{Tt~s{*)BI!x$1836{0GBv(n5
zYDg!B$xVlDu;>=C7a0FWk>$Bejg4ZJXoumw<rxlRP=q8C>p(q*aoSAH+XjFaVLtm5
z!%mdM)szLl*oYo~0Dz4ref}eMaj^W)gEL0@|4!b}jX7y^*migQds~YRTd6Lh=r`}T
z#l*vJgYk-bh-~!I6ngn<Z}5sJSZaFT;~k?>BEI+{e%hN2<o)z;|8aZsVRf>bU}ER?
z`m_gl%R8bo!)Wrd1DL>M_xf~xU9DU(J91AAZDJM0CA-n}d3&AqOU1xv;ui&s*}?XY
zLdRfW1k>>OZC>8c#%4cwP+fGBLUN65>)k*2e7=nG={kwNeyk0GYh8##$bxJDL>sZ8
znmy*=H%j~*Nca%=B0;Oy6;e!-WsU$KuH0o#k-0+unb*=aSVY#8PVXIcmELHO+Z|Xz
zFap4v^B3PGVL+FG#C8tq&Ia!61M!y@W34UxDkLXFS~dd48=S4Z?#;#ftt6Ij3K#y9
z1+%QWhlTxlyCCuU?ImtG@hc<E?cMy~e!IPy_FAGU|6sBFX3w=IyW(&(hcDq$qJZv2
zTwWMd{`}xRj|C<ARrr2?Kg>mceyf;^(WrwLa#`()xuJHbtNno@*r)|ZMKm946VWsq
zL7|&lJvS>NJH4ei$N5Q9QOS+nsfsBSF`o@5&=29&k!eGCTuY+@XN!^)p05oX%0Wax
z*r2Y$k0Yqj@^G5xK|w(o1fyTxwcc*k_cAJ&oIs4$u>uR67G8zd?xTpBlRsmO8QV78
z&F0>&li=2g{_eS~fUo)oP)ko2bC={5F`iZN&&U}v3kn_OMP%6fDaIkwmfc;y?j~Oh
z7e%fqx^-T6?P;P}c=UTJ?!Kvj_x`IXNh(<)FXsNbltlP`IDED-cAoCula+{<uxI9*
zlRNr`240S4h=cgB+euyRnAE$*kuP3JQ|zGwAvRL=VF=P~4>C|znA<LeBRjMUv}NMn
zq7@S+D1Rh!1hZv(qBfhE)O-7raqcXl)e+KK1txP%yP&fAe7-XQu{WKB|I~EUc1D-T
z#`@Qw-yJf&H!_oNe73NdLc{8|79Yk5JuM`R8j|AWrr2hhHpH5Xa;@P5xGAV%lOjOn
zM;x0EZ8HmNrkGgIb6<Gr*G*{ewiPjI1V%wMm;TPm)kg5GRqq*sQv~9RKiM1;Pj}UU
z9-d>cptZo}qR9DL6oZpR61iMCJ!dlxHH1V?!_dx-$g3DlLV^Z=gA18*`ch~gP6OA?
zBo{IF<Y~L>gTRG%4}9vH$C{Gc@KG>QH`v-5adfuG9zdS9fC&p%FC_6&UF@^`j}b2w
zjiWGXtSwbL6>YDt?BM0ghf}t99M-j1Q&@MeNpYACy7y4t|AcFwjv>r<&~)_g%v)4;
ze6{;}Qt#gKTP*$Gsy=679*Q)?IENM9-ycEh;fO<9lhf&Vgrd6J?Z>nz!ot(6^t!)y
zRby^a$_zOhkH~WEh<0%Wp<R118Td=r4a*nXQ6)O@UAF$?9Mk!D(S15prvXaX^(d<X
zj>Y~Y%}7?lGHL}Si?xl8W_r!a?KPy1P@7aQtS%IfLe(P^n{t*lOFoC<!Rqa6yV|N>
zLr;|LHE%$i7L$^2#H!<)7B*NOsEJdK<`Jsr5y*&qR+u!hoclKu8RGGHqwf61O=oxQ
z^}50a3}eeSf6?opLnj_H{GPt3w{7MCSQFlMt@2KO=Rp;v&O36nsN@kK-&WgiqD$>7
z)aL-@M|ik$w#8glEwEQBcSTNlvSq?p0tiO0#TF5|SV{_rg5Nb5P)Ib$Y<u`qfdUQL
zoHlrCDaB>*eS$B^FA;|*OSBkH)>aStiu5$|0R}zXfW30Vg=4_3zrFYi(Zm1PSO=tY
zY(ihDr;!-{?S1#}pr;*+f)D3;#1!tLH5c>Ps8KnjvI=UE!8<_%%yux7z8E#u)2KX=
zS0zYC%$U?2oRJ34NVHymU?48pa#Tcc<%>R}RTseXA($R3zL1m~$}`Il`q=1V^w@i8
z*O(&}XDJdIX$ZMP*twq<M7kwXnM~?my{QDZucB8H>N2R;ms!JAdz8!`Kx(D9LKofh
zNlCg+Xw7E{pXLfcnIh9jR36nO-VmUZ5G%6CU;!)LM4YgVG^&pOD}iqx!RT=V+xu!J
z6pjH)e4rlvtuHcn?;u_;ej-I!N%g2_tX}WnsHj_Hv>}X~ux$j6#hF!)qdo(4U2<%u
zoI;7kM3wQnO$R_nVy|?1#Do?k0=sDRamz`iBhPvlN|G3+VW1uv$jy1hZTQ@#n_L8r
zSzlU)ErCXa!IBTb=QvIQcc9ZvN(3%rr3RjzQf&N38tV9vkeaE1N@?M=m@qjrpb=Tk
zU1ZT0{~Zw_jwL}qvlgLW6|#$MK2HD0i$;L*Em*>1ZG(n71IH3ekMQ?}UxLs7bfq0A
z&+OyDWiv#WE8u{1q~t|$^g^R>ZUF4EaZ4ibsN^Xs*CW!DgJ&v2m3{%d^oN?#WiEfN
z;khB?l1qMdZw{IZ;Qh1j1E`_CPZx+Br)Pb{ti$6){-8>bVy`y_6HYPrLl>81qsdiC
zPx_SfKBUi)@{L>)TH*|<1njX6-mWT}ZQFfw7x@Dhk35x=wn-);_}4t{xDv_BazVK9
zNEq0AMTkKWxU_nu+8`ics@+;DH2fOsh|_`omTmcVbugkh8HkRM?!iRRBvqN@0F0o2
zQL^VD)-sV+zH<jy%pGPPl%vqfbbf`DDM58mmg_Btk!#!I%%xj#;P~=uW%0V%>sSd<
z+MUw%a?mOqJ-O4#<Wi@OAlkNFmr!hvUyc07wl?Vgdx!MX8@)_ghI?(?e3$UW*;?%3
zJwMLJMBBQ9o@dvsSc{BP3JrMTC21uXxM4g(d~WbL@lWuFU`#zX!zFRZ|4r`PwUwwE
z0?n1EM0mzJTc2oWv9Q~%YM3uoPx>0w7OL2Ir(Riu2@DCZ>iSw=dp1-ef$la^cf!2J
zckxMue(&nRu#xd1F;nFwk4m+V`%1+1#JZ?fY$KfE>i8}Oc#sNVd(kGfYWHXHYQpqC
zu?`Ez|0AmU|5wE_(EndmEGxr*2SP`h3bt772tF^h_;aZjLTy<X2|}o$0OQ5&ptpMg
z>@@43qWBvQ1xxYJ=HB8T7a#Fd)g)5w<SB^i?7I$_wW5c1V{gZq8i-&h2he+qSR=iJ
zAO|{Cip&BI$xFgVlq8ATWs0mCP2vzl?A3$B&;(f(;56Z0%EHhmleOYFJr6xfeX-sp
zl>0vd=Z_djzh+gZGzBFbwu4UL<p6Bx&Y3i97^sg5sgfyOlSu>i24xWL=z3aGCe+h&
z3x@#owr^>eQst>|^lZnP6l0csFVo-WX*GO146WRNKA=A)A}#uGM%UpstMV9L;8D|=
zXh<6cT`t`a<xaUMf_TJpb!|5qE~g^QI&2*VRENELO4P#Z<Vk$|51Jzi6KlkUQWaN_
za&(ZPpizzPsO$*(`J!HcOJjW?`ngp=zhVWETU1<Q${Kqnd#+%XzfY+YspM7_xm7`|
zGd(+~SW%4Mjlnf&qFz!PE$+ASbr&ql(b5|M21<?PV28?apg_M;HjJr=(8)NysNv~I
zG`wT;lxJc21}dI(fsHk9266UTe>yUknqfVKMK(+daj~o#!3d|;&|=aBvVx3DP3AwL
zC7U`uMpvbvFo088h^PtON*IE4Y#lZ%LgC_^gi~m4MKruU4lzYvwU!};svhdmrTWT1
zWJBx^&Ir`(H$};;U3Haqw~1*uskfDBIVp}vS}C2PeXWcey>xHf%=$D*65@y`5=F=@
zTcvo%`u^1&gRFOMG>u9o3;qrpxr-=T)l8`{#v`xCgV^bdA6>XivV*%Kmg&XuM+0{6
z13#wgrZ=sJf;R|Xz1zj(<=ZZAp{bU>xDzv_wS_*`LFu5-r&6)$@o5dM5H8&4pGEF7
zFg{zc#Q(+EJH?0=wB5RG+qP}nwr$(CZDX~K)wXTh#%kNA|C5}PoxQ*8d^dB>i@KRf
zRqCx8?|6oz^TOxXa`gR8_=~MK?}zka7k2ZF{mj3&Bjbntlf$Q?=zaTX`rzlUy49;_
zNA4UwxS8<hK~-lrR!-{~-7)h%3>p%cK~tVi886JRYa5KiVNUpgH&TKax~%obE>A98
z-7f5ztbDRVO>W$n5nE?Ae-}^Y+UJ;W^^xyIMxx_+M&;Ls!%^{9`(a{g=#77P7vE=F
z$jfQ<=mX?#!OUBrIRahUa%I!V!SnK2`Q&xv)65UREz6`&5cIF}pnL8?h};_he;Pp1
z0YcIOlKGNo$>nugupU1!0-Fqd!ScMb5irNp>PtREC~7wmP|8yF*+Bam$?rP&iG5lK
zQ{@}r6yj10Nl#sg6U<q-`=X2`BM1y8zhe;WK_R9f2n={)0>yc1Fbhx`k^8bJJXIw#
zU(A6`F9U$a64y{owB^`Hd&d!hKtQMa9+SGopLY$0o-F|NH=w+on+(E-FNH&*^=J9|
zZs+X>Iki**6#9ZYMZ?YQp&*TD-c=ud`MB7e$dE}>uO}B5E{$Li>9P{|V_#|sN39A^
zArC<)?Xgrqyg|CQ1yY!602}UARsjj>$>-y&I_6N_KBige!Dur$*w^2mmu1hV$!>FS
zu&(fV@*}ilo_ZFc(zP@9T_)UMfHnRoV5nQ4*$Ve488F`-pKYF;xjFeTo9xq*=Cj$W
zgYsAB3+-N<9_%y;+i;^-)%SnxV$=uMkIcOIa;7m2J)b9UjaWRm-c;iC2}82pP>5yj
zHj|J9HKVk6D;qh5YshfA&pE6#rNtG|mA;10edRgNvX6ztPs(sy#;xSQou@SMaHH)e
zf;K+{5^R(xK~MMqM4Jpw;pf38sbAm-@wV-(%}F%0HvAQ!hK~3D^}@%;33mlMbJ7Pw
zaNtMWkkxA|W(VUYAJ?Z=il~>5vsM0iK!f8W0FkzaP@{_TCd8;A$e3}gO6i6ke^FY4
zofNU9IGzTY>?7D3j5(a9#7^P63Z7?%!c9>Ee1bhxW<lX$?nm#v1-k&*D37`Si=v6R
z-$3mu6%>2>C8prE7I9n8Zp4Pw+lke4^?Z1B)Pd_O@|RUNffC;hcq5`)fN?%wc8L`}
zSN2&c-CnA{qDn%$L%VmDBAwMMLG<q_VwrZil2N&z_jM~s8a2cF^4NG+T+YAi@Ejcb
z#Ky<l6ADRQJ*BgfHEG@frHhg-K=qq-QV2>QmEs$ufT#_b-F8_7C}Cm7HVb4B)okV0
zTH7FkkkBFyIs@?DKMJdDdg)|NoRRVz<7<W(d#3(m0oA5M+l7({92&BhPn?_t2vE#9
za*`$5dzSuee-eag=(9)xeV9sMImHx{#MMrVNb;rYNUl?id0KFQci4XKN;oHkMV9<w
zN{jl}m$&b8?(Hx|b5mV5^XC?}cv6m;gImJ6$3rZ!H+m}BW+@;86HKX9O$A2PDme;Q
zM8D9+%}IBC;s*U)?=$27Ci#f8O7hJ=<Q^TGQocFp6(c!zi*yflFOgH8C#Qt)lH>>d
z1wDm?(VvLsV1PKnQ<Lu4$s7ih)nNb=rx3yS*%H4N>{i!9ptZS8N`QDUkG(CXsgz4B
zO#}j(b95*6d2H!Q)xtoO8c*i82YG{eFMxa_|A>JZ!q&Y*t(U&_(F@Y$5kJC(x)u?P
zEa_~K5fE~zV3aX})O-=pNszuE+!&1}=IBdRAATA}#4fPuR1ep>n0cy?f!n8`JV22K
zz@PAYi73CJh>%gL3IGx8Pq@u&CK;`xcD2S!cr1xBlRRYmd9_{QdCi*ah^1PqJVt>l
zGd*HI!t<ZqLQZpze=Lc;agMry#>an2&<!|3k*4o*c|_Qwa3pGh9G?lQ_y6#Nv~I|=
z6HHAOx4f$te1g78syEipBGT5HU=E3*=ey-;MBoa;#>a*J5%dyszn`jZ@50jU&5hlO
z{V)o@PVmidA^PwGW+}X<eu-^SD3NG+RD}!c7uOV@mv|ALFPEmk^7;qE>lTKmDZI-9
z<-17nn8V+}Yw)koIMH^*hGrf}pkLdu^pHzc0?<T|wW^$cb?14l_I5?lJ)g)}7<dd@
z;w}qtfX<RE=??|I&q^0Rc2|mW=0{%S3y8T50W}4IP88?SK&pby9iLQ0DB>OwAYrhM
z*(56OZ2!baHXD&9Fn8rkZxYm1ohJ$BrkAjjVF++=F{~*Jb2Y%Xv(U=$UxU_Ox~xBK
zH13BYdx;qaDFGv4vbroFna~0JbLw*%6P{&F^c>(at#!GIla##wxQy6rcKvX4=oNI2
z;3`Myw;427UU{Fp1Z4wIBYUVxzpXx3+TwOEZo+pX({8z++i<*fR(jBjeCfqL^kll-
zyQ!^a<*jZ@dg0Vc5eb2oV9H<FwqQ+$kO&cj1<5Bk<`Uhb5(sQe&&|LJXUN<cw!x^b
zz})Qco1cN06{svMy^F*gLFObN>uIsE_yr}GJmIHAn1shgIs;M1%8hk9NRn;$9v$e%
zt5-=-+-*Knsv@Lc!Z9hCxwtUU_efSAB<{f}rTGp)Y0=99zTBPC!)mCKP@CzPRn;zA
z7hFPL?200n0p44bOo6~ixN_2uwy%blX)=%XG?lh_q`&b?y5E}qQ;Ev{pJ*E_O#dAT
zZTdHz{g4ChH=X?mREt)s&@VtNK+4rLp#AEPeKaqv*$GQ{?h8N7lXYZ6Zt-E|b})4$
zjq;Is)VoB*(fMpL&)2=v+ui+B<@7jY{NB>}C_ecg?4<9YVzFp)aDwWv#rR&ipHH%q
z1{d8kCM7G`^rD+QzPnG+;4(Y`-T*DY8T}kUtNEqEig=;-6c@_(YpKm<9{Y1o`q)S$
z3AQt*?%bcJ!&P)mlr*BXS607U{TqJ0?}vpYb*(?J<kZ~?J2~H9tI}PD0HLnGLRnnQ
z`^U@T@0Hq~(xyha{A7Ob56|DUU8n#&2mwplgWu8sXg<FhR%lckQ@|&6p=+KT(17lg
zUQH}Q3HoRf0Vm=vF6s-Cfut{H(tWCV$5NrklmgIcb1mDiS2<syXVjZ49*(e=f*Nxu
z31-Y+!K^VHjp^f(bxb!q7Fn#*s_{*Z7|ph@L=hNtusoZh5LFAD&ham=G$I+5`VUEN
zINCzHWbq52l4B&p#iAFtto`i$&U`9jG%Moe>iy2Y!FP%Z(1fFpM@1G|*Nmq42r}LS
z)ZM!<zx#$0#U+QDyjPG0Dvd1F6FF%M0z{uz0TW6-NO$^rt)f5WY@AvY$YAKJv3q3d
z&+%0vGvth0Rmv`W87@)DoacuR&5-Nf^YI4;nLSXAe!`P8A!PbU09LZWPYzD2#cX5{
z_$=vClOZX~S8xzQP@qB_Oc1e1@eo2qL)wBC|Gp3f&**%$5R4%7quAGe!U<f`NogT`
zF&bDQw#(#Tg{J2DD1N_`VTE7{e*_V-SUHFwVwHY}Wdq0wg2UX+h&}6okq+bbj+&s)
zoJUS6GEckPBTJx=-OpLO`&om{!1O2>xa65#7A8+86wuR8$#vY}1i&COLFpgpf%43}
zaVRUE?p7jcsz=iJqP<KLn-+3)OmaHzv5h;WziL3Qt$rmS-58dIc}f!*r${cndlxOw
z&(!3IQ(L73u77yNwnMfZIhil*Lb)Z-P7J2*8)k~dn^8S>-5wp&C@V^VWchw%LVa`*
zQwuVwzLUD}hq}VM_@y3C*NMN68g~F1rW{HQzIfEsDX_E$w|qvLub@_0jD$*t$D%H1
zAyNoc$IPnps=keRYvSUMiU*^?E#toOAcwKG7d9}gqZn^#yMd#Lx8x+Wr6|}9Fp*Vm
zG$ZJ=2}oto>$2R@vf6#En&~VW7WF8W;(GQ#G##l~-{x2$ShA=W9dW9%Dd`kx1pHB&
zP!&@>>yEk!p{u(TVr;$^wYo=p47*I?89X{eMcmOcdO9(|DcqqINg=gOJ=LlX3q(4e
zwvH^jbQ7f$qYsAMf`Hf*T!+vqB5xLo6vEJtsI8-+i3_s7nxS&IvReRt6s3Ja*GjtB
zHnI9rP48N=(yqEjYkOf{%Svo@EF3I0u%I@+aJ5d&Di;iMaZ=VA3Wl{_$tLNxJTZ=@
z^F>0ds~OH;WC~0FH|Hvdmm{t!KASv;Votx-ruSYLn8%_Z>gTnHxE(dCjmexC@~w#`
z(vl-(P1^N<gLw7aEx~O_dS7NVF|TQn%~v<6&A4jjeAyWI>b9Zajbgtc@=wdxS^#_&
z$}LFgX^p7p^_ykHaKG}sxa1brxXr5Lv>v5fWbjN?vgkx!*X+<Lf_pXTOzYZ}I%<9+
z$e`URwN6beuNd7-uiZ*>vR5w>|0+{<OX$6X2#<7WN*xXT8o3eN^AB7e+Gr%Wl|0t=
zL7=;eEE4+`eGbMu%L<+bhanRDpO8Jm5;B>Fq{~Q{8r%)coexYd4Jb01gV-QQ{9Ss-
z)@h&B;jw&*BMc{W_?HV+i+d=#2fBep&n_p4A{h#v%u(w|A<XF#G9=GPYYa&*{_1iI
zlH->Vd`|X^p_=L{mFIYgGKylEL%S&)(SMU;6)()p;c7+>S<+jNb2;yhw(vB*TnJL{
z-cJ>LQ#%`mZ_Z|xyU+KhbmBfl*+r&gr8qFF<knUb$m-y9t{K6Xk{jPiRW;}v*_Pwm
z$90`P^^#)u;30gOH8C+j(xc9o+0!__5`k8;xX)!J6(q%u?e92hen(lwe(ZG)okWDr
zo!iBeicYd;Op|;=!Y`EUindM3M2Sw;=NeoGr7p#pIH}OEG^?t>J6?Ta2D0KUZ^>7b
zVlC3W+~d!WLS(B%6J`Sn(0o}Dix7Zs0;6}gYT{p*cyVH{du>j?K^;$7coxM&vpFmn
zfIZh(dL#`{tAK4tkxVQD97<dt1AsDY8P@>e)ZmOZngrJT%e?_07X-B@7k_i?CS+JY
zbzzpH`_Ht8Yk_j3+C%?h&us@8`@A!9IKU_2$nGKhHbp2-6;+>zWe6ds;+gy9Fx=a&
z!l$YS35}4y?sF_W_?0_%PJGabzo(PRSx<wJBx?%aG`Wfd`q|q8FNtuW_}`0Nfz$L~
zC-SRWdw5!#YlJ$xPA%C(n&<;9YgAP7F3E1I4n?F1+qte!=IWWe@iI5)YODT{LFU0H
zrR`Xsk=3Tcn#v=_;Zy<;$*}S7OSIolX(4nX(cBCk>Hzojo!e)Rpj2AOd^pX>rNIsj
z4+oHzM&MvuFdpl6<9;#=5j{l+xKLIs$~4e<g;@$q#1p>i_$M|h-HBHkvoTBD+%2UD
zCceiwe9Sed0+1rrO0jlDo6^3cyo0kCg*w6n@@`Wa5Ad`~Gw$w<m%h^RcTwru%^S?x
z3AbqXi9wN2Hd&Pvlwv-s1Rz^6_f%~=?dvX_p1`7@$^YOM6jF{HfbZpY*jZ9U8q03U
zlC3Nd5D|!Bvojcj(avzH_x<?NTwYJ)*=XKpVH`T5<y?M~#c=Z?{n}*_n3YNDgn(gr
zzr**l5XT<_-=oyxEXV=nA_EMYmFqXPy)Usxl)&z~&>WM~+m^J0Gtl>d0Z0O+wkjqa
zJ+CmQ3i_$~=uNt#)#gR{rsF3ox0KI*0x-%4UmqPKNo6~gy&*(P$PpahGXf8&Ulz^(
zIuSfJk4MCrwtfo7ZE7EZ?4L8(px9@^gSxw*fstJWdFX-v&7XL1%me?Arb2RVqr?TM
z_zzbGEx`2cg_eIS#5Jzo<GAXmMI`eJ2aGWv>*kg*s3bN^1?{Of$g`x=Hct}kwxBEL
zYe@4t={jYWoup@p4w!OLUI)#pfy#m+@5aFb>v6%Q{;NUTGX6e)HroEBpl5$c=%aN?
zV{FLQnG%!-!Y=>BFHQNdMQ1%FpAGl}R(6vj|Hm4R`9F9O*#ApObr^fxj+*`cx^51h
zCrsKA7zQvc?=zr%I@mrqC&)yOQ|9}JLA0Um#Y|ET&USqJpV1Zdu>5a=`eJ<g<me>*
z4U}9OCv7#4@24ZuYrvs?|6S=kIW-8gR1V)a&(~?Hv*m5a9%v4AX!`IC9RJf>HWHjV
zgbv|tEyAfi8pw{Khxk5L#a;3t>0M`u_swHZ{|hcS1Oh^f!rQ~=<!G-S2Vo9BblL**
z;ncW>L4@>((pj7ThnXiZ3$Xbp`yMR)kTCSS2t>fgpm)b3>zuj%qHfIsYQEkbQ<K&r
zZ+bf*p85np-4><7L+&e&KocYukYxTYHQfeE6xCIf<)OTo^NAoT<XhuH-E=M4ElU^v
zjl;mq*&w$$hC=~6Q=O0QFfc3IJ$aY?1Kb_I^0PB!@~zrY3^5-Gj%|Sc4vY_!$NHPx
zPADJgIINb1Aev&7x)-srt(Fb3K91z=!nMkI+=t1A>?pq%(yF@$(W&X&R}-2ZwSV+Z
z8%=D8Z^$uGceIjPoTvH|s{r!6*#ey+`%}q)Ak0E{8uLmo5|{wX?u=*^bvTobpQL+;
zg#4iSQEvQFFRN2q6GvnkJDu<(7sge^nVe@wq-sIRGqu)Z9!sP?G|sx<bVs5s#{t_$
zeUIBMwZ5z|c-n2Ix(>EyGM}hq_Pg3pKhBZ;qqdCI=}{ze-y~&?oPPm%_RaTaAm3J`
zD1M%+xYna<!_$5yJ?7}aO|2?6Fx$MlwAX0qOUl?y{xAexG%BRBK&hx1H9BmA^MZB1
z1lMd^2E_iVYuOm@;=*yN6^rvR|Kaz7)nZ#<wHF)zHv0`w8J1X|kF3W6SLu8We9G9%
zKNaL+f*b<?RSM{+^xjKSNfAP}miz=Qw2=G0m=<EC@gLSK7t){027ar^cs{-_^)3=#
z!d}#(LW=6phP)+cEZdkCy&=e+uI5_1q6Q25%ofrhu`zWd)0WjEX>osjBCpe%s+)=V
ziNwm{CoZoXbItY?9Ms_$m8Dmybe;If{q+_`5mG^-c#W#0BuYZ^kcPvYb!O0qBrH=`
z%AeGPB#m`V;cNv_Q3-1sSq=ayWf%H{R29SFUIK&>*(dt-cZqe<a!}WTPVjCyEnf;-
zf-+e9O}t@<Qif0r2me}e$(<p*7m5`IXk?{C5sAdiwUzfM8nLLZd)7u>yfJ%`xamqS
z<TfZ*$Po3B<gcSKUbAdfk)u$w<sl9-ktdqcJJ)^R^^<NTE2R;M=o}_9?G>Hmpv9~a
znlm|KuKv~`@vXm@=~hsu_RvXFRbcVrE-1tl9&&sP%b^qCJrzOj^xLIoMB}B^$!)o&
zdmFlF^_GNe%ktJdHTHT*qQ~E*hL(02b9PB=<mxHzf><!|l*sxkXt{ePWl7e@)TR-u
zrp8lQnMDpqtE9`P<8VoeJ7CNRke_+<NUbr?DYrMFDRc>JLx=2=P<?fq*MQ9Yb;Dy>
zk*K7zBJ+@%#(imP0$wL#W7$HJ9{jQlFVd#r_F`ON@?ZxmL6`u^I?dxwB-^}oNPsid
z8EhTd@uQVta~()-Td5hhlic^FMHUMS42RY!#g$|S3?s&kQ))6vq2WM)6pg<DviS{6
z+yz+2=aMopi&_2q_9r~`c@H6FJttO?@_F=Hr2t@hF)DFPsa5UErQc?F2XIpyyUUKs
z(1AkA4IerI%AWZ^<3@RpD8nN-tAjD8eW4agrIz1{*DNbz?J)FM_!~8hh?0=6Ms!XL
zl&vyGsSE^JiBM6I&kf^!WsI_#266+0rOor%L@i0IcT-SvY=m$f;0$4uF3`l{QgM$g
zUj340%RNbFr4a!w#BHO|SuVS?aYl|S#0|rF6&lm3+FQgW6IxAZp(H-!VV=k>Q26{|
zKX$(-&+M(@dksJg&-bg@^jm-svyt~?)QB*Sucwdm_;mkxg$IrRN1xF1Tku}deiwJY
zg&5)v2D_LQ-tjQY7@$9i_*RuU@S55wuQ2O>!i-g!1!|GV)dc&Pe<IJ^-Tl~{p{*G+
zA`uTZQU)VAx-luetDeAneel3<GMudj_PtTR1HzBeDCyH*5s(InukCp;xze2!*liaN
zj|db_=5gYaix^e{^HEMXQCXTb(l!y^#f}R46XK+E$1bl$=$opq`s=r<uL!$_9H^*o
zLgAUagtOoRPAbXB!U&;fwR-G}tI(j%h-oeuF-~k8r4preXB`@*h;u4iN-`GAwnU$f
zCkZ=WatvLv3AW!s<_Yfbt|Tl^Lj9uL381_ptak`MlKh_JDxWcZqPFn|$jL4k4E=57
z?mV|x3<kVVSg$ZDUr8#b|BlUPN?_$zJAOGvs}n-qLB4_0@=XCWWf!B;e*0?9k{Te|
z=T)5Y4)OBdsQZWJ;SZFZy#>s!Uc50bHRQn;F3iq7CFS{5)Ov?)%6IvytTG-3BAz8x
zo(hnnJoLn!za0`XkXPt1)>?s2J()^DdP{r$lHi3iNxs0jQQNsZI%AgwuQZwxHVvIC
zjy4aWv~#0~w}DaPM`DEcY8zO$IVg+BS4Iab=ywqD)mu|pwk_pN%f;a9w(G5okpw+h
z_yk)Q<=JO&Klar#bJ0HIWWEmH;0FBnkHU|xF!M(#Zbm%gmAGP-$yFmx%^z9xt;km1
z8`kO2u)13@0VUT~L~y0ZID1sh3z&&*xK~k{TXo|TMA>vdu3Jy27iZmcz@m+-r!xu3
zn$W)1WmumE-d;Xo*Tr|O%5Wat-1QAP!Xvq<ZLLT_g<lp*PkC3(f=dWj&T5e2xtm?3
zw_#fuJ$EqScQ#@BEhYrA{$c~=5gKi7C#;=oR7g`qNZc^Wyt(0VZffIsl#60Z?h~@5
z`PLOX*&BID3o(Dj>5zez1Nz_f?0D7(N%ARv3S3j-hPhFb;iw@|oo0w|!c~#xBIWWp
z*x}OTIbeEv(}DkcZ&}KS1tE#S(OE9Lw}nvs0x&NUBx*90dfNwG#^o#dQQ!7ieEf%U
zt0VKqv~rScCmREi8i3eb;mJ<!B>-LRK@v&(@6ugUEy@@c(h39)fn3>#hd*_MLA%T*
ze@edffun*1*{4F3A8*Pjmp}XpneWBZA}|kB_(J{nEj4h(e^{`3+4#Zh{+z|DtDVO^
z1+0J{#*TNp)o#muQ2(zTHO)b7((wsh^&CnTkCHD+`>gydGTU3x^zzKsLDJ+b>vUbs
zx$GfuJR2&;5=n}3LRRiYC^ezB{|w)qo@hw9_fq#ca-F$vjz#GSoE|))bF>H&f>>}f
zHs`EOIVxqd1Oni|LY$&%e6Va=Fp~{Ft=DTw$l7ElM1i-*lHc(bQ#cLDjOe<e$L>_d
zC?iEi2#S1Fpz;wLn>rgz>_@slN4QLXP@%W4HcsI|aFH7tj6M^PQH+?T9A~7w*Vv27
zBqzYX|IsDjGvxdnz|q``OX)f1{5)dt`_biSFcElY*zN`^Hq2z^Qhq;*4cKhpX%ZeC
zzF4obF)ICJ)Tj25GB4xfBp?DZtnoQ|G<lE?K89ReqSd^86@?wnJXzj0p~%zIQQsfu
z#j&7u1~@-AvSE*@;P30vSbQymi1gXL$9hVe8PdjU<j(g`M>CA*f2?nRO&R~Y*Orlu
zll{MPYApYHZ4cQHdOp<8;AzqaF?#~~1d?LW&uXH#qVl4y^PcCvetCM7QcU9X+6}qS
zdlUsMH<*6-$7d(vMcH-X=`Vw)$V!K*^?R}p-`#*c1#jdQdM+(@0272PH~iWBc};9_
z{PdrMHHdGAZ!dl7-k%`|uaJA>0qTHp`8<Gi1a$=b#x(k@{xCn+lN%Ld+-dMx6qW|t
z0mo9K{JZ|#X9u^Z%D*t9T=HQ)*BvtJAuSs1qV$!9;b;4Y33ZB>HQ>n|MVox`)mp$N
zy;JW`o;PEth`i_sYy5Gd9-lQ3Rif<&&V+UsY<pDz1VEmAC9S@)q0u7Io{hPVT%D=f
zG6tE>5z5(drTG${l1cTu)y<FneBpf7;E9ek4wnh<-Os0o1Um$cV;`FsMa1&AC-;Z1
z+s-8|D5ARdBB$6U=M^<pCOJ)PEb;QDl3KBJ(s<~%r0Z0GNY!V4zb}iSiPk#dm>ueL
zByH_b`g7$cHG0GxAYxFZq7yYE!4IlUBZr5sVJBOccE54ngbnh@^ToU5#ght|&|WS&
zWCc-wNgXZ6$mRvf@V_-HR>#^|Rpn(?W`&9@RGL`tDZwb_=hlpWh(*hb3ON=`c5C5q
z08?9w-uo<D{*1-iY*DCMLQhVskx{BAhmB`dG3>6qyIL*0<4oC9Y`0mK2}N|{Z;PP&
zpQSWvo%C#WY^2Uz0n}<q8Y6%7iXk}@n{Ic3Dekt%u6igahAYYlE6-kgZ1wfWku=uT
zO|liKGOqnSD&&n5ePdZ|R_Qf&(qeR=obTAXkFZ#DJ2=X<hc;08C6p=5cl504u2*o^
zI~JWMI*fo}`rP$9q}hkNT~5G_KHt>Lp%q5WM;PFN*rEv%emr1UG>#bv``YW~k6kgK
zxAXH#!lxO~f#m_GAhq_llO%gv^UsK$ka<WNcssx~tA>e<uK~G$+L%z=0CX&=7>Fm*
z1N`=I*0WKaiv#8^_x<aC_H(H@_--{2k8*ER)JYG>RR$AJLJ!nc$oL~5RU3e-oExEZ
z5oaI{bDCzk?p9A6K9O;I*0?Fnc<7eDeF1<6;$V^i(|$`3Kzy_v8`iifI=<=ahOD+a
zYbUhV>*b~$e+g#JZfG*PE^D#j=$0H<m%nWy$`7@Wa2|p8OyNCcUwt2T(7AoRb^zsa
zDBsjJ7<~MQMXz!SAp=;oc<>D@SQS8kqOeX}!NT0$_|??;vPFF39A0#qkyZ2tQ7lR_
zQdf~v4M2+ooq#gsxLGs~z)&BXr)}bLR@odkjv?vUurLNiB2LFlzbxb~V)7^m_=X!1
zlU8@KKPr!dcdnY{)$`CjzNk{wa6K~I;Lzg2K(qoL$Znoc_eyXp2Oh;>v)j3#lSoW)
zm1=;u+p(krA5BJaD)H)Rt#Z_e=z$`p^ihZ?w!)h~*+xib0$?d3FPB|VPQ;~|R}1oO
zW8+JN-&KkX7+5A-TsarCOp@Qr9!&@p`0bsuI!z+r0kF72=%8xqG{tkTa|P>p9NasH
zZH=K@eA((}+7O<fF;=wq0gQj}H?+fQ4n1zOwg_Mmp$FzIRgP<w3AS%hTbBM|BeC4_
zeZE_@2#AwcIt$P?nav{sQ%l(PBsXnTRAL*kC6--=u9j0mAC|jX&axiKEk`ZSS1TkJ
z+8}H*f7al;mLvJ_1{OJdxS0_HwJk8mUYw|2j!>-PZbShj&IA2!VIuZ#L_gir-q-{S
zhO4;U-6Is1tXlB=mc`rEl&b1p4d)+~CYSe{2{q||eN+xPV0qf-z*8i3#T|a{VHTz=
zR0945{9&FvBZ)7q{<ut<{<m8{gZyzr(@%A%+g=czc@;G_9<mGSy(>-otdjg8ME8pu
z-t?G5aU=A*aFXaJST#*0<!>9>S#3VmPDbYE0Ak$3G%bw{(e#0JeHXy5O0!X~?p6<p
zv=1GgSlu}hq}(k=Y6_b2&rl^g2I|r6yFWmq;(5B%587^lnk}diRYt{2xa(*PMd21V
z<1IQ0YVmq=-U@hz?CU_to~^O*<l24U%;t7u%~qlAg<8#t+>Eq_CYct?@b-I}=cc*X
zEH<Z_Li3<J(h^$%=92hno@pNMa6Fex{#<#aU!D-NQa|_wj{MV=Pg{|u%VKT`)mrm#
zaijx!B^pFQUog@f^)mG$f88Id)@z~E-Wus)w^wR^$5vE@G;VfQsx=i=iL6!Wk`k@$
zstyDm*H+@|T~fIWu7)V%s$@ixABX?&Xp)@YPq19*v&qh@{>S}}^?yki|Notd>`edN
ze)p@%aZ|$ezWi!(S^WWT30jFsF6PDE)8+O>oHcIG|F0xhT@V1Ij4w(4mA@_QN(y3Q
zjQtT_aJ;_@?)mz5{P}jjsG1#&jMx8b;KK+1gHQ4WDjACg4=11j8$Qbq-`g!&@#_Zj
zib>I0JUnYGW5?5H^|%0B0xn10d%b{Bh%QBw^+ow@LoPaNoU~o#*;(-TRpd}2*v>vj
z4sY^*JQ6kcsB#y8Y&kT{LB<t|7vCY`b|iA9LU#n>qmxS14lE_y4x_6r4qF$<n-grv
zQtZLZwaDW5j=x(LhKiz`Wok(1-={J+RPESRfso1IG@4CizDyqM8OBjdWV>MVCn?M?
zus_aNI?Mm&*W#u9tBghK=d9&Ky)QSlpBsTDhs75qQe79YduS$5c<VQE%dTq{B82lX
zyoVCJSotOh6}zPEwoSOWMSq*Pr|<plqSAk&XaXWZlhdd^Xk@incDR$I42s1+jQ0O{
z^?94!YdUM5GYRtlenawS2pHfxpa~_;{C;2Ek5AX<Sa$0#g_N+Cq~8Xq79at&rj3Dj
z1<Ilaf3I%PHf5LY4gFVbHGLJI@I*Lk`JKw>iNe*8s04H>QNYh+m}+ZU?){Z)v()^#
zQEk@{xDu!w^76{Coy=E~kQg=)X|<X}gLhOoP>s05F(ayL6k??FY^u*1y*LjJPUrOd
z$K*5sH@XLcEMiY$?=+NNSnW2jQJWaLK7|Iie$*IRWLd~s3<Cx&ej>1M=^vu?foMfo
zCIP6&rm8R9%U%e*>UgBh08qlM<3$vq7Yq{k%6dA!uY90atZ#)g3m3H1O(2^uo~$D+
zmknQ}H6U{|Gk(M0yand;0|(49TZ{#$cAd~QWg+-4@?A7SY!8<mr`}RnRvAOUarj-g
z&=dO3P5;t-ZzATEC{8Vi{8v%GBrG*i#>YW?5zu4{C?;jnoEE8)Xe{A7haP`omX#V<
zV0)`Aooz{bK<B4<sN<BOnRi6g=r+dS-?s(Tb|9`2<KM+hLb1lnws6OyZbTCjk`}h#
zrPRiH7Qked2XY4pr@rww0X%MH1zp{8DkiCdxTa5J951>m95#*>q+V>0%B)@GgZ9dI
z8-xN18vltc^SBGi2@}$COcRn9!~mFKJKev)A}l4?Q#Qg2f`f|LykR)O9Hq}VK<4E(
zIk}kW<i+5qZh2yJJ7|i<Nl5~V1YG)IbM|OK>I;WYP%Wsp2`;*Z02dtX;m!sMQ_2k%
zQBKUUse5fnD}n(@D=JHWvK6I~BrTg}^N2+=RmvjAm4x5(E$@((Wl}d&v}ea43HY#r
z+0!w)5X+q0P_=gPvuIu&;x#PGm%(aVqD*J1<(EoAIJ%89p6W9$UKe9L9=~C9F&HG%
z)<FI}_jGUCbB?`ee#S{Ya|pgzw970lRMx#zdumjSg=AXQMK)Pll+I1%ozEzy^C*8_
zh9N}-QL5LxU>)x$K+QCrYG=n{cFu)a7*o*kj9nqC%&uq)%hqOVV*=6<ZNRj&-KUz<
z0y%sg!=Y9ej(kqhg2hioej!M0LV-T5DX>SFXg^T1c+#9V{R{Le+T~-<Fj2m$wxrG2
zXw^mN89{}&UT_K5xf7MzShlMOXe<8s`k6M@q(ZMtPk~_cWFmYe7y`9E<5WUsn~>0C
z++6}Ee!BDg3|*8PZTGccm22agd6iJU0S52>5k-bj>)Gif1+Yi(B|Zc(4otEcR`=$=
zk|~N9)S+swd*d?UpiTREvx<ElT@RTb$b;UED&SPLlR!W*P~&y!MIO`C3-k@{Gd<4-
z*#782%QblE7BfAsOpP+*B3i@=wxanSgZg)*pDyTx#vSwE5+PbTtXeOEJ2iso{jmJb
zwzheSky-Ja6<()<W|<|QlN{!hzgp`a*lEN{ZPBrpkJ}UtG@1en%}{bx3#DjTT`4oI
z#>so4M+p|x8Vw>z=UYq6%?Z`DERhv*W5jdAL61{yJSrPBO>Yxvqy*mU4q$l=0`RF~
zOfO)SgLn-<8remqY7S-xELK|$?K16M!(}E+JunDkCfvNtj&Wna*(!9vyzb1Mj>~$2
zvxOGK?oC{?KGcndfh5599Lkq-kCe|Tp-jJ2+BDjl+EIBRIrfYGqrdgJx2eB)KAb+h
z=B+W;wxI#TxE|JuFV<1Kev!_IVald*k;f73=XwfTXAJL(l{>(4v3AZI)f)EEBi5@@
z5T*uuSu9iz;K^*!Z-?>B$Dt{Co;An;s-48zY=B@uCtTHV0uyx~Y@KFTZkDfUy%3VY
zjkvFvw&!u{9Z*Ts3`(xzi6uUNbO;Mpg(|U^>ZwFodHu@E!j<K|1~;7|H{)~-laRf8
zhqKpCJ;w#Pq-Fh3#`ru_sv3Rt57$$CAOOQI%ecK~#j|E(rJ(Q!_wW^|qj?JY1UBEr
zKGgHa9qLtaylGt&C@t@euI}J$(`7FSwT;3rmkK{&&w(z;25lt%SIB&}yJ=CI#pY=l
z*0-&{uB2@so4M+wk;15IvvZ@tu<!eG80^yV&VtG)z5W0#PX6GUE?{>6-k94csG}Az
z9jeC4YYIsELjM-klFLWmii9BbOl3T=Z&0Y5CZq|Mz@L7pbI1Aj;Q0YQG9?X&gvUXm
z*aTtb4c|vX1yEH}xQsHq81i*W=|zA{bjJe9yL!#od$9BBj~d6V%{Tv<4`iQWE%r=0
z#RDg1o?j2h3^e+C`LpT&cuJkNc2aZG>UaZVS4nQ~qS<2RV9{;YBJiF6k}$o@G(-pM
zXWGl3Wz(fG(824_7^dL0X$+Y6c>fKNKe&q~)d{(ZCfz#WMweDui(_*9Xg<*e69ssA
zZ6*~v>+Fh-{0MM=O2NCMV|(zuF-rCba%V(z>#Dl(-EAf<{Qmn-aw8%D72M~IsuC$F
zfV=4Vv8p^IBj`k@=bVb_w805=S$IIX%e}&fMdBYd-oRt~A))WhlQe8}`Wlf0Bq7f+
zQuYng^*v}{sPf-wXR!@zBLT6FyH(Xp9M4!xuoW%GIN?>Z)-DaJt+n-7&~^rO+d21M
z8PbKaWk;T0Y<hCGvGUIP7VO<dhEqr@px%v|YY_5Yn~nw@OxyLITYH`*)6spkOWFi_
zMTZ94Tu(2`ZSK!Ag2+G5QGecgc3bFgXt(Edv;SC<ar_5~jq$%^%>ED3ADbPf=ceu!
zUqiLH9KD{Q{&EP=%eua`u&i*LGW_HRUy9dUEGxacWifcwdKyMl>1Z?_Ykcq}n8(M{
z%iI6`s`~6V7un$M_WKIJNAO-hfyd&p9|)po@Amuf_B1-h;`L$Ve}EC+8sEEjt$cjc
z4nE+wxP3(cqvq0m#8KiXyTMt#r2Yv%zqv#?`Rt!v$s&S4lVGX6(|2_*yZI(+udR*2
zthTYF{amQFgEm7-EVE+vmS!~z_B?m&N7mJWW>;fnALGT00mZIeBQMFkV8Lgs{6-73
zDWVE(%g2`)wc)#t6Sw^&R5BIGhSJyK@&iHkFkZai?18Y*`|(=cCT|HUaiux;JP7Fo
zT=E{spXfb3EWLca{9hj`-4Rk2hwvNJc`0!IKORu%5KRneu8ad0gR}l{ze&p$p(jt4
zy~Y8<q{$-|yqZibLe}mK#UKfk(r&r*p1He)HL6_F%hffC6n)D91i8FIKH6Uo4<7}X
zZ||lgk?G>iL3_%tGDg2n+F#^Dt9*z={NFM83uSC|VTD*@7P``@-TZ+k4F7#b&z^y#
zf5;$#XKtB6LR?n)<&r)fI^e<siB~QGZ*RZh8h^9#m3`*QaHtR0UMyk0yzI_q8N-9P
zg&+G)-);dZ6POD_<WU^j=Z9mJEt%uegXpCRgJG5=FHDZC`vmCCktZbjGdE|D9OMj{
zCtI$gEv-`(He4FV-auUv9NOEpPjw<VyxaM^k!M`f=s*yqz%at*PV97p22TDs{ANm5
zqm?$j2np5%-@Ev*AqD5Gs%TrmI_^V}MC1kTYK@~^*!3F}foQ1;>){grLm#3@m0k@(
z@HeERf>u@Hjzid<z!QBcR7*3Ceka#mxu7*fLyZYm+{Pu>U1n3+0ND>^Pr<_dOc$48
z<vUek(18R{=!)Lpgkc=OFKd$eyBO=IYX*PTblcHr7tpS9;-2zQv#eSSs<AnPVtl6J
z!=N`mLQM9d8Yg}j!50*{M4}3zZif{<5rKFMJMq_=BQ-aYV2}YDm)+ZR*pt;@on<(I
zQ?4#atwW)5L^4O6+C(lGUFAX5R1Qt1qD$^CJjLsu+j*U&7LUTTzRUpllgN%rUWY7`
z1uk@!z#w)m{s<_fRqKH`D5763oHN~+22FJ}LFd1^i5I#<yb1xf#;q<G@|lGxB&-sF
zk3w3z+KO{hPU1hvD3Q@}K)}xXUa6PUL%i5ekld2^<*LOJ7{bdczA?T!^Q3}(jD#5O
zp%8O_uz`AF*631~DAmP0dL#C8aQC=$pGz^;*6y0*&2croL(&1~m?pr=^;k}LP%apt
zfECis!4R~WWf)B)SFOUDtm1)4GfdgxOLm}^u+h4SR7`!7TG**8eKalB&;R(`35Q8d
z=zQ}CobxQz3H+<wkj^To);^F@|2Ito4|!RAo;DQ|sUR0C6Ed1&mk*xw^h^YAaVc7+
zRtuvm$2_EcK6;A^{K;bI_>@llGzw#wx-L2TXi@t>TXEkp1#X!ROkNE!k`;6JcV(2z
ze6Y<f4Qxn7O7L_xO#2!Zj&-<kI%8C!tv>fFjwOMF{nY^w+NL7YX-KD!#|yeH;y$<>
zYtn<)^{_S(Bce=<H=>iWLC8_k2gV)RV|JBH_E;6JJ8shxjP+kr8)GZAHsmyox&!yo
z5D1<X$rt=M?x^MCKk#wgsdbqmBbccq-ZjbR01$Dts?D8Z5&Ev;>vPeD)~UxoOx;7b
zR8<{O=Z?LXu2m~&i-(WOxwP@7USX>BEzv32jNO3_Hzr;qBi$V<l<t4L<RGO>wHvdQ
zx~<1*pE0lk#Ogq6%S_5)F6eY_Tw|TvKOE>(2v{1^S87Kc=;C<xeNLCTo#<TD)}mW>
zv=%~h7~j?H!{;xkvt1?JwG6CBgDuc1>y0i1S)fz*W1gk8O`7gDF0rgqclTMCcfB^J
z8qsN+8+C+hI7(4pcxuYfDpW5U!xyOa(#Pc<>$bS9T{7QIgOvj7!L%Upi&R?Ic%U_C
zY^zpJt}~JMI?M37d{00cylj0wxs*z}0S>v!wv2yb5LGJe=Ka3DG2QQ=I@>f1RQGZi
ziW((gtAsAatD$mp3{@>%rYi5ppsFU4uKOz6Y#XL1wL5D~RjnS`wDF{)7CrID51b@q
zepPCwHPh3JWs`~-9gc|P-E=<Wqx82ZJu&FSH2c%Eqi2}{9L9KT`fU;xHDFV=PnvXo
zcF+f*Y1n`3VXU+q)WMy$a%hy%+Sql9C%gm8UWAzetbT25Iv)=rqfzg`{M#_vClzt;
z!|lKkQluCS_NLz~+TqTa6FAl>3Y>6%;bjurgjum`&k8pZ>Mcee*-w1psAj&J68$&U
zcJSG&Cuzkk70X7M?N&>Z1g>wfvOA2)b7T8^bos<euZM~|GWR#+z){5OZn$t62B;f%
z&i|d+1SCL?fOBOpB)jMW>xli#h8}_6iJF{yPrfZV996_n^7Dpkz=4&B0pTr51h4=t
zg=0rXESpK(2D%-_`HZ_O)_O4nUq!DA$Z5>vj=*u%u;bm)I-C1)O0ZhRU2LQhTTlBr
z*V$rb>TXZ+D3z$qrj(b{&bPB5XIMW-1kd-pdC*=>+7K5~^U*NdAmse5S*#svL+aSB
z7)MFTJ8*1W4`VtRy(kcdqMwzaHqq#NSId~F=#prD%o#ki0J^)KjJe`nzWj@IQsa|p
z+R?DSJiY^AVP_mruIf*mm)Xn#JuS?dlbmLE_Oc=D7g=A~?nk`A^SRLY6_yD$C(ET-
zdbZGPr}1at?#I_VaMK=epkjv3M{E?Rft&m)b;qV|%J<bu5;GjgF|p3}>F0sR7GDzp
zp^r-rwGEh6)FlYv+UQ&yqe4j!!|I9Zm{55vHqGfcB)anhQ3tt&{48h{;EKJm=LcX(
z>wf(|7C8+6OY-9X#jeQo|JfCp8UD*AqhoDrdz@C(uWNlqf!<>LA;Ozz6GV)$A~|4+
za>OzLZDEE-4Q?sh8iE>N^4^|PCpXIVPASqGt%P}udA(SUquj|!@0ooE<l~Gm4w42T
zVW}k=Jc%YvCX4Fnh5nFiQ{mX;RL5{xv;Ntv?8xVV@#(0z<)g&xHiPD-DvV+W?8c3R
zN=ar%XI(Z5U{Sl=WvE7vO`Jo4NRI>=u*vE#+*+jv24$=WQT+CbZ-9;Kpo(h>nJqP!
z3u<_kXqHBp93@LL+>9mR1ZA8IXm)1U>I&De+j>cC+Pl8?`uH@Hy{u~8WkIr8COU}b
z(_hyx(bl``&$3H-j6?MB2@s@=C7)7=W_Em=AXIY?#<p4un?`@kP{k&fMbZk|G|F^V
zCV1+a9dVOtUQlpLn(`$LoYn@2Sz3RgoLLa(Fyx|*$kqH_3ZpzMjG3ij-%`f;k|riY
z^E56D$jt_DKjq`X22s3|G-kZnbLO2ur={|%=@{t=TSk9QAvH*2ON1eobk2aiBZcLK
zAkq|9yLK4Sf==JGsZ>LZCtZoDVOUv-9OLkjTGCn3{$FZ&#CaZ-q6)#lifQH=hB88j
zf)mPxj211e8FP`U70R<PKck7{bO-Kx7SSDsN;UkseD)Ss0Bl<00=Bp^>FNzICVfQ|
za%^6bywcfQX2QGenR)eXmceU(s>SZ_Rh2SI8@>Jf_-x-4Q!RG*^u~;C@Z1x+)?qeT
zOFQ+_tyw;=+`^lxgu|3lzG<1}x4mvc@lv)Ew>0tDvLK{z{B%%LQ_0pQ?UkCXnuVH<
z28K{-YS*e*$J?_>`q6wp?)+H&+}vsO!^T!|=pWmk$AA2TuTP%`CT@3IH*w&HyCXNF
z_IYLJ>G<ZO(WfaaA|lvTj1j9nS-}XuM$O_B_>np=mmix?yYEGxPq$YfTO*k;R~+UQ
zNs;CE2jkC|9^~v%QJC;19P~yc6y}n3p?lSGl*+nwAwW)#+q3RdjNXfFe0sT4bB8m0
ze!rU^|GNhxN58jpM<3UY*S9N0j=Ww>f6tE-pBt0s%h$pEgVWx8`W}AmDLOqoyx9Tz
zy3wN5hoU<UD_*G#Mw<D|o~EL7(%vOao~+DVoW%mXUXvmhCOtizczykxysj+WIPt58
z;ZGmO_mjt$bx^O5MLsLnjfea&2A+2YIQQHbnLx#-lB08Ft6LFd87Z0;HcUzBq@brk
zBuYb5Ad!|V*l|h`efY`sgdE3iKek4~rdE2-MVT!=uxg|YDgBWETCmjn^`QV$`}|9~
zhC~G{oM36F5}?y@AkNhaN`N{x6XJ=I0;NcRc~B0q4npjkvrvpsBXzw1CD0h>Sv?YI
z2{=nhM8LTeBZ@x%UbwoIE`y3f0};g(z>wxj+Zzzlkw0$}kkxWd)gAqjAtp$%G$kTI
z{1qb&!1Pi><xAZFC`P_(v3ddd0N)irv`InVzpe3oK|q<cM1W97t`{c~pNdL8mma#<
zZ1F#ztG$9pV5f`b_WHeiYg<IaF*zX=0EN4Sm`LO%2~XNApg*v3&i3;uq@X+_HCajN
zcbBr^ciC^3Z^MnwOMeWBC=M3~K`R&585q;=>Tb|>d9`uq%mzg`M#vu;LBTu+W{#x}
zi4T(xnj$YBrIUFa-_Tv5#ADBR{CN1fGw}JdY4hUcdwrR{S-e|0+R}jJ;q|BY5TXv&
z4;mPsPk>L}oxYzBJlxt<!2Pbm*T>KIS0CJIIbjM()VT!U(c4xf6J=Uks<W=OJNBBU
zQ0GPRbP#H)|7Vh;-Z?QgH7K=~Yj4q>tKkzrF)Hl{EME^K`IM|$(?kvqh8(Yk_x?N^
z5!0y(IBx;ckS_iSAXI4<NDwNPOCw7PS*xz|RoDQ6)o@ze>Pt#;2Y{d+0pT!md(sth
zS<mKNs=m|{xOG4rb9E3k5Q?P5*!1f6FoR&BNKi<KE{@R%Xp8{coAvc1oP%3-M*;$<
z^Hn=DD@93Kaup~CGBi%mN%_-^*q+-i&<^Ip`3D9hm~J==(&W-n!3>=~0G2`*ol(;b
zPzpeZ!i@<otfPD!$SX`a17$NIL!lng3*A(xZ>OI-)V*H?;}asD4$SYU8n&W8Y^k0g
zhb;kQS3J-{E)_#&SmGY8+d!ybAq}~wSwM)6dG!qmp_^pvZK4k?QbSD}5|c+pLcNk{
z8R<_Po>9M!pPZ4>%d@Jx;I<;MFFBYzJX3#pg6$@aAyf5^a;OQ#+uE}`89l>Q;5mHP
zE6Wfrrh$Y&IcOily}Uw^k1<tP&beKIgV*uYZIqs7G8y*T!N;^3i$Zo#yQnh5(;jwy
zKCGd!9=S~pUtIm2>p&V8sdR3VVwvhv1@@EM(x&5<pTCW&f&v;a_ke+!kN64(N3Z0S
z50FFAQ6G~<Xbge*!(3iOk^Oog#?GpckviGj?f{{YAY$6zf}u(y({(R`KMyGdFEEP$
zkPs~`&<<4rDxjA@wZVNKaUo_YX!JL(xT^IzGDLn%Cd@brg}Qq@w_UF%Yeyk}R<Jgw
z<uF3*k&FOAltJ-s2uO=xuf^d*3Xm|{Ze{rh5Utsc7{nFK;^wBb`>E%0KTBqj^s+6c
z;c<0oZ?4KNKbmUZnYL*@`DN$3)y#LR`rPR{h%QClHDzyfTuAk)JGP){UVtXM`<&?O
zky@3TK`qEvRcN7bE>BCW)*GXoH|%CL`%)E%>z=m@Y^!|i3Mdc@C>oBRy}>lQqd=(K
z6(sXDNvt;i?OhQ8P5>Ls<3UY;W{uz(QOKEz+v_)rtJBRf!0NSlxF}@I>5bJrVxz>Q
z-S^>dKj7>60Wb`-BEj6Sb~<`{Dl^k6JbP7@%3&Aol#&WpXKYmgwKIQL)EwqdjPB<T
zTfdU~mZ@`HNQ(1baSv>4TJ9pM?LZ;Rme^vsTA5M?Ec5=G0{^?U8BVleH%|>MaEpbT
z0MYQ%m*0hv*EK=+Y_lXp6~<IYXomSmO^AC{l_aHsuD&uNzhm4&8CqR3g)eHcF;yV-
zb50}SEWE}_0*;*)nIFgwj(#Wi8|wtUZe*}%GSR3QL9@k|R$fKNLBD`-a63E?50i&r
zIQQQfA#~+RKR-Z>j$&RTY!9izYAe9!r^YlYEn=aT(!^#iv|!XL-8I$om_j9dPDI2g
z-u){ui`p0djb~l!8N?*aTFNF1ssFOpW$?&!gJEN>&yksKdo6a&VSy`8(|6d?JGRoL
zy+jSQ?4{Hq*NN8X))piC3!@W0@&ECdfNOzPpH!g@;wy-NHqJc}bcQL{xEwuKz>WCI
z1;<H~Fw({=LxwyGCs)U){A=67Z*!1SuaY_eH&Gm1TFL@Jz9i{G1uhSdXi))svlpc<
z(ICV6exei86^$&-`ucHR0-XR_4&hb=O{qphU0IU#7yqU>HvIJ(3!*kqkuQX3O=Or{
z%)PEW43=IbSwK`lwW6Hr=p(q}SaKkuZ?uZtLE=&1PB8DXD~DFEi<2iG0NbM*nrXB&
zknt^Nh}{W{are)Y`vH)w{7Z5;D2P%gv=%2s<v&?EpDZ4d=fz|Qkb6Qe0Hn|bJqYBk
z1o`|35LX*F(sBqx2{k+~A~F-AJ2@&Nu=`i)TO2D9!Z{=X<-$dj&U~IJ2@d|1-bVrG
z6|nJGKA%F75Wib{00La2sVfbTw{u_eg$6&GLIZ`CC<zKpnt@IS!VgSB8BidgHGB+{
zeqYI9_?T*NAd#T2KU#5-v}&w7{e(&hFe!l7aum&>9!1D%jHo1>q(4Ja4Z2<PAO&PU
zXyJm`?}(VYKVk$<z!gt122!FTvQlI-s!%rvyn1)aPyfV$GpxABQ#aK$g$`gxk%!d5
zICN-6QaP<FI=RKtsMCO0JgToUV(_&KVX1@rk>+cUB1mzKn;N3W34SKRYkdsp2;bCW
z>~00uI7p>u)Hs>VNPqmhq`UO;#tG<-fo(O<)VRBUSC?m*yD&3oNlKOM`VSEK?WOmB
zIw)}bQcC`hhAa-I|1RKk`4^qL$$`-QLH!6koiKa=h5@!Ahht!e{AkNiCYYoQm;A#o
z<A^FM)$(sFZ$_}DMFh1<kJg7Y3Vs)y&u0gB*A0IL34E{`2`sGDKKKuI&Nq}CdNWoG
zf{Zj{*B8e3hBI7N95L=#dx{Ij`|g4!u3m9Hi_jc|pnB;4!`M4E*%ocxnrYj%ZQHhO
z+qP}nwz<=`cG}u$JG1iKs*1W1_f*6i^UGL2V9nO&Xzl5ZqXz@UMRXBW(pX?2a|!UV
zUvh2s+*5W=JMN4$!G>^Y9p~Hc!+0!t5Pw!pXHgB?q(pgjbuNrVbiu|WhM5}Ee<*U6
zXqy-o#WI{OP8_!$60sHT#eyd*m>oiYiI@v+fMznbNChd-Z?YyVrvHm&a@1S_JW&vu
zeiR;k;DmHeA|XcG!p4`^8zl6O&k3CL<q4(sb1yf5+Id6xLOwrRKYz#V=Njq{OXZ^7
zyb`nV^zB@HZhxQ<qwN;JpPWDYr)&G0*K=xC_qyJ21L#{5t!h>Wa`Q5j+!iFtsAlfp
zShK7Uh|XQp@sQvpnN#`syWuC3>2sh^1lhXPS+8gDk~L0Z9*(2EAPR~nN0aFIb8&Nk
z00Au4?9p><u2NtdozrKeg$591sU>`+J|vMLFsG+O(I<o**z>OwyMRB_*3(5vv~63N
z64}u;MaM#|6Z<2q&8w*1D}*W`48@XOmBJOwDWvp!TYl4a_3(hnz`r?CnN=V0&<2S&
zR(<IJ=@Xw@Sy#)XXQeP83NQj~NKN@B0FlKJ>ESgE)i}`!&tNm+{$Ffs&@%<;9yFN9
z(`O<ldR~Hx6G_z?WA8Y~)EDA8AZ(EO2$8CGdielFs(=QSM@odNMumw?+5Kti9w<<y
z(5Ae@1v5fr*SDtvM6K#aDgKVU8i-jJU5viXk~2Q?9@5Sw9FmM?+AQsKSqUyPS%dz_
zlTo3P4a~VAime$e$ihv~kQIa=Ffh5+D34vjTDcWk@N%Va1s3FTQ~x(36F_O)V^lsv
z08*X-V#BJ7%v);YvXJEZEUl^&SH|f9%ZD|A6Y5f|f+`3iMtO$N%p3Td5~+zK3UR<=
zT#cV$nwKCGg*PG#DMztZ7kDaR-u-3Iag0?+z0&a1%}^6a1!)p7nAI*+6cC53oE2&L
zC$G|bFArH`E4GkH*RCav_0qSKTh@1Y4JLR2XrXGoSJPn6D%2^FQAc`Duq<ImWKOZ<
z)hv!Z6mp>E=GlOa`1ICl5x0=?n`_B@A$S2olTgGMw07IY3B!aXx3Zhj5HJgDP-$2w
zT!cELdbc|1%<Wg`6&0kV0i;T6;bQWPNdItqo5K!AWpq;Gwk;Af!)%<{iQ0DSiBeZC
z$lrJ|4m>gr=tcW!tZ}zKN$S!g0V;fI>ee3wVNzfC^y8n}cJGL_g1h$4(Dctgu_f(T
zu0>}ZNY-3?l_uCmlRH#ipjz3kN&jun+WPx|Wo3Z{={g6UoX5`EBlv)+jTYR;qfGS`
z6jB<-PSn)4pE~WhuGLz!e?04ALC4xt{L*IYA$vM+i?!2%v$cA49M(Q^<OoZ@*j4q!
zZR3GG!e;5wjd!DjeAj>{PZW#T)p~nPMuggr&B2WN#d*q4Fr5HF0qGM{tNqw4VO_9)
z&2Ba-Vc!4y9U7A;9R{_t?l*5Lw$=mFCHAk5*<5zuQYUyMIiq<w3ld!3Xm$}m_+N(l
zHF~Z2>cWirmjiyY*jUbT?Y*|MwJBzim3e2lHnu2ILF%JiXKPB9WL&j-ipYajrCtcO
zU2fU}>y{RSA-Arb(VEIss|6{|4TY#1uX${SY|K#f-*|1Q^Uf9{7Z*Zfr@&(Q2F<jF
zKMPB*t~j`$qHP!Ry@LyhuNAp>6F}nij+=2p7A9~m=*2&qbt;lrw;w+6r02sZtqoT)
zn~9I!A1L~<q(aph<mVXgEpa8>efIvYB)oBzl3qPdktN*U1W9$3X%(}B=1DGt8W2lX
z3RR*lpObOYLbro)(pZeJJ9NvS4Cf6e8Pa%>Hs^@+o2{j2LhW5Pn^Gz1*X=zf>~T>h
z<$vw}N(BfLQUCt_aUc`&+*f8MD5qrg!(ca8$K`QszHuEUQ>ZiUOXjZ#Ui$6UnyxeY
z^06}p?tMqsj0Pf;skT}1l*kebU2_i&+S&bGPJ@E>ELjT%G3N5bDH<Mg0}8a(o8Jv3
zjhkHwA*&ih7IDY;u)&nsG^=tKxWofnDkOi#Kr)-a+@sZz)RV){Gm@E0o!;inXzd)y
zYZj@h^G`CDL(Ue4*cA6zG%u%5udXs_P5$WFgD{`@uNI@KFBer(OVRd-0=?2a!wfn8
zWh|^*IQ6Wqt$rprlA$#Zc}I!~R&|r?CQs})=DPBZSjG$dY>3lR1X^{e(W*fMo{pcQ
z^HG(hcKcrRvWwQzw9A)uA|#Y@Z5^p$TP#)GBkSH>o2;2ufBW|Z8`nU@7T>_|got4(
z)XsZGQIB8Yftoi%zNP9be1W=0+J6;X7RLWma9J7thg%vnC+v>c{%^rOX&9%X3s?|n
z=VeOJ2I~Xm1kKDjNqGAaj;f}RYFfEAbxJTrpU}$3i*CXB#yq-xckuFlWwOGlaR5<+
zwciEb1DxO&Y?_}p01R1R_>bCpBq?F~$$b{2XgwIexsTa%@mM+V0DFKnpzPgRAZiq8
z7J0@I`APes{Ma?yOkcp(yCX4L7;G0jY~p_Vd|U~dg;%>vK{NR=pn-lU^bP@Q_!Yy-
z-%@;v{)wMdY7;Gcm12uo_B8xNIZ1KkX-er8dk%5IsHHB(OXDH`@{sX^5}mCSggP3t
zTpW^r_M!ZGp3HP_Nb?MpvVixs83;!~=dX&lN-M3+aw-Z<s#}{*g{tn4IZ{hhaFP`Z
zD9N%;x5c+sOHPC3L8aY@<Va~b%X&yS4~FuPd~4@CDP>W+`l}w7=Nx;23Ws^$nkSt<
zB~fD>9UY(W$J1MLtc+}*1`dqHUyw`OD~ToB7q$c-Nx+<2TG#YMjbA|GF1H6&uaf(f
zuN|cmP&iNDsa2*<tRh~%$^GC1!q=pgA>ig=giY$Z-6wGJvGK_#@YJ049`v7{^GUR*
zU1W&klH5Lg);+p}p}RwL4yXEONeU*8bJ)R)j@fdJ{DX9FQ_1XOD05R^k#~+J%R03L
zYvjB0ij}X9HQ`KNZGBA&En3X1xsjOsDfxSeAS&$jG=z^?8zb++zs47?n(U^(8oP-8
z2@`j>5lLE)UDbtcMPRKBF*`9?CZwmzzA5oBmn$A{w=FQ|&Di9^o7JJs-0X5nBsFWa
zBcM(d35o0Sa_yrU1<LKYcPI(f!!71hrGF<^;FWg^GAKe<y4q0yHK5jBaii7vS{74s
zpAkD>29clBPKI*;96ewlAl75o=?a`SA91M_KPOWoFHC0r0kuFhNxxPoF3$cYgyh=-
zn1ZQ&=oWVgww!T*Pp){KsIqb~*}5d=)TvtzX$QLZ!n0Sj99W>TgY}j&jy)z#(27cT
zDkVU&OBtI{E_31V7V(ENJg}pmgipF<tJ*;gT2aw3om}b4CZ^#e)3{N=C4U8_v%)L_
z1yd$T1*j@%{cSNqE{3iKEp6FZwmF<~k~VYQElQUqu!HWhPe#0je55ol{|mQxq-fYN
zK52|O@z$k94f5Oj1nvOspv6|QJ%}`_B>R{o$BI*P&D|6R=0xSRCa5TceY(n{j@sL?
zP;Ul1R#u-y%KLuL8rYXoe@<GkAlB>}s#G8)8w92VtT)|sFS+4bfJsauL$0&rV|55q
zq}0GQ+l~_BX#Q&3qXJ1#P0Y;&FG5~Y-1ZRBcnXU?Q^b3ABc`Bfm>sLi*1Kofw@}>%
zNr~oUA6z!?OH>6}T^XL~+gS4)oF+m4k$nZ0AEE`~gd@97%-Su`I4R4mWIdR#MFG&S
z*WKUO$NSk$!Ol%5F?hQcAI|Ux0d>@i%xnM6;qKvP|6}N;clCOUYBRc8L(dXsp{GX-
zf!$;S+t3zt%~SJGDz9xqumGnFA2Eb^q8*Y-$_jAfdFRj6!8-v!db%_wC%t8^BvDDP
zPPAsIG#5*6Dx?c;Wv#g!rOAzNy(nLOs`qK2Hg+74i}RjdZ^}iTVcmLdfHp@XPw9;n
zZztkHftT$SC!OEb%tXUIA<LDY?YE8X_hnu(evYqO==wV9bFQxaWG7ZC!!^+4$~2nI
z<7`T`TVBRop<eu8&<#)*>JwN|I*$+jKUKVtysn}-Q~TH;V@JNB!AyMavAUT-DJTO)
z{O?Sf8ev5=$)1SNp)b*n*Pl|FEpA6lX!nh}kwNd9QM>U(Deb~bQr}7-VxCIxms0M!
z>$!ijZ(BF2=+m_jNHwQG+tuu#Wybth!KdceQD2p)S<PFXE>GVntuJh|*K)ZGVBWOo
zcSHr~atHOtBRi2(9*s+Gl=Nvk3u)06?uJ^($K@9t>ouG183lqI(C7{14$T4<H`_*Z
zKfLV|F&3aj7dYEfF}T`{*x%qEBS%jERV!FI{!gvoWd0xanrKVgZ~Z^COvpaZD^LvR
zB?;ZUHuUZ<EJId~ll9|&rchLJ?9iIla{o?XolZTQ&Ac1Ya^>gJ_HniU_O$;1?T9Yp
zg3#dowG;Sha7?d(`B(=HFpBI_&)?mTS1U@t)~sVgo2VnmC*Ik;L>@#TAHW011m*CK
zf?*P35^{v;;?(X`=mTeZSvvi~#IJQCBWMAf>dp1__5Kk|3#6Uw^POS;zS+JO{)0z$
zml*R<`G5wix%->=ApH5W7Xx1jJphv3rT^o$i~l{ZBChvu2b($wW6JAB$CT!AwDYg0
ze3(#13~5v&#iYE`g!#u7HS>&7B4fuaHw%no&WI5iIl6zPIWwFw++;^gXJ8R44M{nF
zzn_cM(cG?CU1aQ7Md#p}52bN5U~OY2SXZDGE0GF+KZ*AQk;Mc6=nQD|C!I=8+GW&1
zU5Vx2^eHtHhD&l{Cq^Yk6@PT#E|yO=%nb102WiHNT^ogzR3XSz0P#)d29c2<rG**u
z;K|e{=L?u@NJ9mX?RKHZs6YrXv0}r}mm=|TEc1QZ^v46^$Y@1W^}PH8(|l{N>J=xP
zD@>F94TsJN&IywHETd<*x|S6*dO@Z*$NR<cfw^iIX}3EfXCm1rgX~~=Fd8emUkmMu
z^>pBO_cEta&hD!D0=u!Ea?X_*oO6s+U@B_`ZCW(`k&uR>T~S?LsV~neDRyi$3-43u
z>yy&DHV8;*QP&@`k;r0{?>>u{31)HQ%fp^<AUT47inaQ$Pe^>}r^z{<SYqE8H2bHQ
z<_@=0YKwvx-d+x+Jn<uD*j>lZ>#5$Fs7f!s8K+pKIOmfn?%3z%sF8Wm1zjHsbrSJl
z`|JhIAS~{rzl;%MDY$o*dJ~%PerlZuP3jU(p{Dc8Ubgw--EwXB=4x)rYPe51dANPJ
zWsKVPKx|BQ)hyLD13Q=jt60L<xyE-4Gej6Mwe;?$=c-s-a7qo>aSWs7G)PfIoSgW+
z`HG~n&tAHp$5D1mUhYAuo<jO|%d}{@KE&O2ggBzej+1<ypc$(l5A$%3)a*HjLrqT1
z;Yj2<YR<D!8iq&8H>>5I3#Z~t>wGzVtDzn40+h3?0yn7|Ogv(glR@`Zrg?f!yFi51
zq(&V%klX1^L5G?8Y;ClcB(>1&D{UT{64$qNBkLsF2=d?SHXS3gVah3g9o?^OP}bc4
z*`TEB=EfE^2F!CzLdD>(Z%{z)l(sAKz**R#j0AUM!|5RjzTRGoP29<C#~wHdxEE94
zZEvS+r6CGF-XR=Kg@E~LMEis{B_z_PHe$ZP5ql@z2zb)p;~rO)pj6#z6z|?o4V<Za
zX-NrgI?=zo9+=wj6ff8}t^vK%_GRS7HoL+NMMjJQvLdz=EnW)_B$71ro;=54FdiaH
z`v8lT9F-QwyN^bwF5)4}+g^(DP-w*HC<Nv23(Am-$B#f1<Y;?bg@%(5u#lAR6E=$w
z;l;sAs>m|x*0{tjFrneV_*(MX3py*pL}$H)-k0s?XDyVaRchE!wUr|D(ay;ptEJfl
ziG}w}F}0&+Pu3K*V1xLs+b{wp*_lRzLpl!ziG5trLuDbP2^pRT6LlQ)k_A*lijUoB
zFXABep(bk=wG+88i)|qe8%xx=8i8>TTHs&%6E2G|Nms(&3Kr@x>myG+#S{m>$$G*?
zo@7nY?&~Q3OTIIM@eme@D{9-_6P*+UuWS4xx7AgBL6l(UXNU62(n9Jx-M?ZIItI2B
zJNG1OE$9Kg$3}__!nX!QLsKT2L7~+UaF#XBE##OI8jfNYaDb76=*W)|^$@1s3>vi-
zv<n$U43b<@*b9#$p2m~oK>6YPp!Ga{S)Wlq{%tiX^!#%~%cE!0$F>r3ziQu`Vyvdd
z+%|Zq84S(X4Qmav?$!$?l9Zl^&3(b}SCJZ|-{4-`<PPb+Ezz%K^-<2u57`xRr@$ga
z1RrIO^??;oZ>~c=pgzqB%7OG)4#HQbAl|TB3we@$`&hDe0x*y-?8-|MNHWv1WEblU
z$dK_C9Mo&ZN48*%`5O7Gx-<u<H?I*~8z1Aym$*X@vWP^!j8_V$5l)FEkv(F{(>_7l
zv7SEImluEp@J}FGRzNj@o{{8jVOay9Zjg9pC1lMxK8ZX8mYs&ZH35O=;Rs`t6x`C{
zE)S~kTDHg<iEs><amN6K=fE3CtSVZ&#^5g_sM<76)|qzw{&H{P%)E!S?ZKnRAnB|J
zDrjr&wL+ffi*4h(^K0J5H^Y|tft+E^oy}p(JhW{blqEWYX=&TsB7w6d5wnE4U-CIm
zKhS;%>}fDC>~_B8v`MSul~%An%b%KG&;eG)YD*v0Mp9C0Z&l`_a)<!UTtECE1uQm_
zyW`P+Y&%>BFL`u4bYY{x(*ZUZ-rFBt+JYB@nmf&>)C<<fFz2o@*Yyb5wlp+ho%)Mw
z8qZ=EVw<2KN;>y}ZBHT+vl@#d`%M5*-gJ+h*l=HJ)qBgytpMpX3;So3`+bIqT_rIR
zjTcFS<3H9%_APP#WKE0~yAd=yTUJKW<e)<lZB-2kz0*xR`mg;^;Qd!|XaC<eO8>uT
z5EIA$?B?}<t?TPR?0Zpu1r3vKs`_;mVE*Qv13FC=*bB-H*vWZI?EEAUX_in(rIlLB
z+r!yuB5Bg7`94Xv^L>4k$nWac-O%^<0_OEVne_1X{x~vy3O(o-zZDiIvx{*ul)v!H
z|9GryaDD5%1#1D{jy?SHuw4Ft$Q~j4h++9y&ff<Mx5#_se~fAFR3FCf-Rb9cO!^Mr
z1c}L@WHtkI=udyk%dSMBhhn**qjY#U<D)z+w(el0Qh080%ea%T6|NK1iS;l&MT?pp
z>n%fKWajLBV$Xd`6Z4X0cs`%ern5?F#87i_Jttg{bg5h(-c>v#E{bWbk@XoLNUCct
zhPcY-=R2bOq@5mI{UhV@fBQyF_3T2^dx{l~7e>J{eJi`ie7nJ+s0d<wNVz;uC4p+B
zd+e^^fNHPkE{?RGmRy9)twT;lxVJaGZjP)%J}|MOd_>SBZ_-O^7-rSa8g2!OYoc4m
zK*e%Zx&D%6QgHbr^N``_RDmU-n%a`dZdb{gUQe=<Gs~=7?-Y-BUQ<a^#m`<0yQPor
zh~OJyFO{S&`3-#3t-u>&*C?7e=+GK-OA}9>s@=_-2<FmJxp~jblBT_Fk2=LbC0<{5
zl{fSuvj_qXNCf&CW{MiFtf`d{Pz5*621yJ0j{=_bZh<UdcwJ=hv!b@EJ|A0Mk)6hZ
zU!<_k01@W-KGD56=&yn$syuZ4{hX{lZ1vcAPaE!YFf3fbYIwhHl4@2R4o6!by%D(b
z>1FELU703+To5`lt4XSJwv2Y}I5kCZyNRSrzHVNwzmlatmxNy%ovMmws;J3Ei2wBy
z@7b%da(v=#4?z%((6t-|f*clYxam0hx+<>D`*9xDc0XVJ*1|60D7Zk|){$0xLH2i9
z>s!WYNgfDqP~h#-3b<$VrO}d$<(q1EURF>=;wca(GLisg@h)uitpCu$lo3|T$u5S1
zt#(bjDYW9KTp1Ckv#Y0CA7(Ngkr)(d19ewYbdi+1yzy=Mdi%f40V$f~K&|F~Kb=Z@
z4GidV;>%y|12tIr^MBplza&;s81<acuv%+#i?gh?DG$q~CZx2{O7oJ5s9iqUjTWKa
zD_zNt9$ikvLvpGGGv)NcY-QA+H%GB%H(9$J!gTMIdSH~#Tv(a#4wIVOP)_3RB{|QH
zE81oN!rv2S+)Z$I;GUq?m_lUZ_K?l}S#m8htW?P5o!}$>$BN=eSo|quo2)p8<^>_T
zhtlR;>d%OBx^)?N;4}JL0Lh<@8U>urjl5(op!DZH;W(a<iE8TFY3oE@D9N{J;r0h;
zf|K2wfHVJ&Ghu7aH1x<vY!beS{gZ0Z)i{1%5--Ws=K1!6k%w$80$R4_(2=)>zlw*U
z0mrb)?`(@IVxw7tH(TkjWwODv+jmC$p^3(N`-!XUvf1!c=S&BCM0AFyyx)%3{a%M0
zHK#T=GELq6&EK!Oomv?n^hg{Nx3Y19YR|->;1-(M!6mWqm@gra80;9x#|_L6*>v4k
zwx`fFO;T+WzPk&|;H@xj$d8{=QQWP`;wq!U`2ZD`k5^Ir5K{ow4?o$J{*2}Y>pdT}
zks)$4UdDa|D8@$DF6Yjg7)`w^FR(RVdR_;ED2ADjIN$wJe*ALpNqB+E6E6yyHbF!B
zx{vSIL1&0!Zs|pw5<5S(^Yju<Y$jn})Z71}a}V@&ElC`t`M2Grw>=Nv&iMRQ+HkgQ
z`x2YWWTx{z&KUlJQ-)Sl%9j8ZSW-W|%2u>h%Tidz-icY>%F=1l9Is_(`WeVpG~cRv
zcPD|ZxY-v=(StU7!MD+FitSk2n+_A?v*pghDy_I7IPxYft(xbfaVMROP@p7qDKF%d
z%vY9>;gtSc-ee_tG&1bkdN})-TUydQSd-JCx;uf0Q+|D$k~)d0%8d&VRpj2##?TFa
zyzufQT1(6{(q+`uZ<Eii5;dVbzdU#DyzP_b;DF9ErA`PQ7+P0G4*&&gmd#h=bOy7g
z)-t$X2H%=9B!Xj3@BScw|MF$cVpzg=<<d4qZ_zYtF#ck!E;x=Ug_${HM?6POENfmc
zA)7*8?2S3cWeQr8Sm5ZMnFC0BlCgHZ^}V{f_+UejZ_2h)kvAfj5*6v}i)%9p{ffM(
zsL6ISvIM&%7MH+Bcg%lXnAhmNza%B<kWF~2SVJBS)r+luVe(qz%Q~HG=UpV2-q!W0
z#rpl&C&Ohqwr&+KM@j8ZgNR#JqgcaBjKdY3m3t7X4xDcPMY(z}^4|EHEM_ssWbqu_
zTo;JWq=e4^;6gR{T&~NKz}?LbRj^1vIz1+x+iM4_bK8)k<lEq}1WG|A-S-9)p-B*8
zDF%4vZbK?JFzo6k7#h$kbUys&QofMt2LM8-67lB|&7O#Ntz|JZ@3ADEPI)mJpI&8M
z>V#0Gz)TAI^ub4$z`SaBWFOr1&cmnD=8LZy4Vn=#b!Ev@<4aVALL}L0Ug(?#;Q8E4
z@=Qxz24nSCuM_IKc|z~xG|}1bqv-@`SS7uMk=JUY%eq^hHdBhXrX=B>g{hVj{47wl
z_(5WTz(lPU928r(;VDNfzlD_`=kh)i@Fu+0CmjqliP8fKaAdFE>ve?)*SlCh-~bYs
zO$!o<wi|zdLwg)!yv)TX)#}+FIsS6kk1pscIeL8k>4QLPVlwmn8dnv7fVPV#4GGU;
zLO=oQPv(zKQ~pqWi=kArvmJr$z_-K!l><DG!wr(d<@pi84T^-SS*IkHK+P?eUncyZ
z2KA#KB{1BoGoqOQKN_ol3vW)EJjP#A!d%L-<Z*phw%Yxcv<F_%XYkU=nZ%Ka74;tL
zO-nn(d!b@kZ6vct#tPRR$KIG_yk_0P14#)H{2M`Tf3809@-QPi#EC)v7qF|JNeX0-
zDkTqt<OC!igZLx_KP2L9k6a{za{wIB$lz&hEL?YM<+dSZhFm0EccXe>-*Th6(J<Y|
z%HoyfYGu43o*NQI`uch_wEcb#6+Z%Ur%J*O6Uk|U@JLyc)dlWf<nkQH&E2HlgFk;c
z?ZJN?o;d!ueCz)+JTY>z{m+J{W3As^$80ctr<9*S)u^KX;30rW)+c~^T^aua@pPB-
zy6pZaNTu0GyBW=iOrV?*IucDSlPA4PNTqGZ%em>hm#=M~LMxg7Qaq|J@4di%gm2g+
z>{#2H0EP%%Yx&#w-ju2PbCLWcq(uI|t*<^S;qnl1-T*HEcAy+SO(0uAzmK?OpVT+S
zx2^IfDX;yjCwXKLNE&QLUHMRcegBmS(C1sxHf_sjl7S4St$=t;8jf^%09}kcje0pZ
zVD#vB|2=rpmUuYZXjORM*HXY3x?QKtaiabuJ|na}(51th)|a>E_E&Rupg0mp*~i5x
z-4+2zpqy*ejo0?qd-q4*M{iozBhl|`|DB$;Tc5wH>R6kpExm@%GG;aOU7$&>k~u53
z7TkCVJugODneQc8Spjo+phzU?lC3KZLj`47RPamDNeLA+=Z$)hUJK)!U>yTcY$Yj`
z#sRi>oiHY}>t!$Ov_1Z&h9Ku0%D~fSt~V#U@JX^bT@T)BA@<dHCq}*y`RG#2Cn{x6
zwW}n5QEtmY%khQX;8CujCYDXaGGw)_5EijV02JFb1S+)cJ(A622g#!S9RmzsJ(>^g
zF*wqV*c9BmD)PD`e)@gsZ>&C|{sNOOa3=jA)r%EIB3yoqN;ONjOrzLf(J894URnQ2
z0|9F|s4F+ku!C8i@Y9Ax**m<6*HsFG2^NztX(Qp1dXjoZ-@omEbLvQH8U12ko+bOC
z^h`fNetCm;TefML-Py3r^Y(KJ=C22vj7&e`zTrpcS34%&XUSJwnR?q>yYAUXL4M_2
zk>1R2erEwQPniwqJoiigF>({Hto%iUYuLjE7+fvcB<!(?`=Bv3X4$!O@nPgxF$YQ)
z;yeKg+dO3HacnoLufze6gzZLC8cL8h7}C(aRtY5+J46YnKx@a6koORVCeUpJQL`>j
zF@$xaUc#`JrX2*Kwlnhxg1O(509)bIopIQ;)oXxX{7K$CA*V)XBMqc;sLY&WjmXwN
zrc!6W9(cg?@Fh;n6CzOdGo5nXs>5W8+zIU79|0-yqVeG?j^3od3Cq+I)HQlfb^;fv
z!>D@o3w(Kh+xMrY`wH^O6+YOo&Ft*Whh(ldT~gM+JlUon`w;aHUO;`+A^i&4x+U@w
zh@)74vfuZf3n9Vk!xs1n--2fVfnShgVZ>oXtEbu=kZ^$oY;(>B{jyai$fVt&{6wB{
zGYTn#FwqJx0NnWX=U9~;o&i*%W62nU#~&&d4Jt^L<B?aXK8u~+&;2@C!*RUjt+2e$
zPe5M9aVBnzf4;9P`^9MJ>}A6AtQKe-L8Of1It*PVC`@y>MM<EZLp#IM{DvuF#ZXv?
zILK_pNPvS5D`q@aG>aD`n9*Hyin1ylPfjGNWCjN(C3~9E7z!PQFnx&)na(5N;qj30
z3D(2O<tlK<+g<}Bk(1D7-@_u6NgE9hR3!BzbE<Y=k;<hGqz+VyeSHpEVD;s4&>kN`
zhAp#Uoe!30!#Eu`%8GM1c<26eil8TzgMO_elLH=kx^gFxk03#lUzjG*=Xx0uRliS$
zIpaA&y_v~XfwLP^{4xf}I#FDkyH=TPI<fLJIj`ekE*}~wjHevFP+}&0mQjwa3b`@f
zxyq5#q(@nrM|glU_~e7U1m~rL!P$)@#CoTpCL9TP+f#!MSl?ruSP{PLhsw=d@?!Pg
zkW*xC^G22>?IpM{qw4yqgZ=lF(c0<z)Cc>QX-V5euROE75|7=?r?*YFzPiBj^?tV9
zDZ(?(3^P<@!0>>QGIUW4B(`0ek}5KlyO%P>ni{5Ai}X&5DiJBZ>V>y5#8oD@8B<r9
z6i1y-=H}_uPN_hL4EmL)<k=o>4hjUbA2D4?OG8xJt5O-_E2*L3_DF)~E27lm?fI!n
zpAy!nt_dTW-tte2)j+uW8(U|Z6G6v(ZE2VmcStCAQSm{>CFJSsA~xM6GClRb)f}KE
zs?VDGm6gyc35gY_rIsX*IiUrL4x@TfJ#7k4O?Bp?lEV~W%lztX$FPzyA36;`lJt)r
z0EP<73Fsp58UjVm^9j*g%3LNHruMKr1S0wb>odsKVIKI0x_^K$w^XgI%hjj_G4y3>
zy*E_=-b8gIdZkx=dNm7tO+X}YH&(=bI`njBr-i5lPHy7Zwh>B@DSzj~Ah_zrZMi0&
z4jybG2-gE@`O+>%Kz-o-B9U@X_CA&Mi){WlY(Sc7la;0pgEiJAbs$~*!^|u(yo5)S
zIG&`eJqQZ?s48CbD~<p+0R^(FDw|EV-R@z(occP(APpCJL1-^Rci!&1Q2K*Iu*a}s
zRo&&TgQ4TmwTq+s#DiuO%-ko+Dnk?|8yXg^=%v7v@WN%{vOi(;Nl%U=fJ3C+H>WLW
z(}*J-X+70=$-ApVnQrnbtIE*XB;)<b)`4D9AJK2|#<sgxiy+C1&)C<JG_m&bQ%uX0
zDbV5R9NmUTdhy|he<x*Ok(rzZ#41OVvg$(v;L@Z<g?Ds5gEkInz-eF%T+-Bmb*O~H
zu;~<q0zY$O<Vly^iU;qmcYPu@YYc^6J4ZPLU$*DNP`8_HUSW+{hljpCQ@j2Ce&6+h
zzT&0@^WVjv^Z&3#m>B=3{K_xEyZLL2{D<I$5OxZqH!w7~(gi=O9oYH*e34tqq$HZn
z1277&2GXTRhn;x&lkgjGzt?*gFaLKRKpqUUREGOsfS2G5TJ9!jsYnf+7-We6?uH-E
zpU0VwwO8&@Xcq9y^lWEQ6_-D-p+$HePOw@ChaVHA1?cw?r|NR@lI&u)yh`$`KYvLh
z_JR`87U;^0|4-%nF!E3i=av0e>CfZy_v4)MTwhX-C17~JAI}>-0thfQybx!Y4xZn)
z@9XoUw|pB;*&ajh*Qei^S|1s#mLS%V7WImS0tXltoYvBcKPv0EB=n?X97*f_O9?K|
zQff<)$fsTp9>3+aYxxe>Te~LJsk9OxrYiMR07&&*IdqcS9cym&yL0dLo9WNG1x8S2
zu5VmZU5+h2Kl=p^i?H!SXUT!?qDhq!Bk9(35=QB*5@_xFhjdfJb!jjl8tS+T;~5@0
zAuFI~{mCDeP+{5(iJ&_Q32KqQj#73-ZqOw;CRm0{Yj(<%(w0+3Wf8_W0+_QgHo}@@
z!Tn;qgqS&PqUi3@UDybVQH$vdwh|Xd4n|%1n=r`Y15K2725#HZ2U#8Jr@n48??O{F
z?7>3{fR=!^w7&CQIVJE+6)16>>lw!mk2{UQ<Z`XZm9rLRwc1a6rSJD|MVGCM(W|RF
zU{+1ISXAj}Q?>0CQhpU4I;gkzY03(F#?|A+FV^Y@_|u9Or@E@|Rd%1bTvGLQVS%@>
zMt#95szD$5M!D(;{L=0qKC_rV7+j#L&*D%(a!!*uM--LC3T{r8Ogajey}keh43=iL
z0R<s-Gium>&e`u|@6O<6r0jrk^zDF)!U;x#7)a5C&Rdn2Jw+uJJZ7wn1DBTqvJ_dA
zCkRu;SPM@X`$1Q;Q#X=Z4dRSii>b8b>AEM370&$JEpKLD^6djDI`fJUrcSkYHKtpr
zbW^8nK6A2RR6ENT-5<ySGeJCw{gG&dt<xTnsl=L6Pm}YNQ088H^2NbX#ypB)1Es-m
zz%~#l^2T-Tml0om=tk&U2rXtRqP8aS-#9Wv@gLp;S!P2N$2XPLx{7&uE^fZ9glL(Q
zi&Quw`?MxzP`b~AJ7;Hj>Z=ttCGD7p5}yoB{Ho8dCTO560GeofTYHLNmP}I2>aD2k
zTU@l=%Ud^`g-zNuIa!P(pQb(F5zc8N=4+M%W3xD`q+qp63#;Q$h?<I#xWAnlLhHz<
zQyoZ{q1JN~S9*7O1Co;IyAE+miaGv>az$awElNKo$=IgJ;B#kDV&EUCK%H+UZ(^hl
zo+f4%nOr|28z<y|vW018a#_ke!2UD7h7>mTz5i~azC%ogIYueL13XZ!gdxAPn$KL!
zwjQfbg4ktCGGoOt5dllaQjlkK)nyfgjdiohiSJc&ntGpiDa_c5qe?WWSlL>zH?+%S
z#w&Zth@Tyj1}rd!5@+ki9nm>@J-^n`GU_em;K)v!zmS_~6m-^AI4L<aD~<Cw&tP`;
zbG3uwA<O23vm=07-iw9x)OUbZ|M2=I8&^6^oKOQTUv5Tu_$vV|#1c6!gXs*G=b_tw
z#K=BIH?m<f9Lq^GgvQtOP*(!4eH_Ato}(ONTy#Xrx68utA%-o@lSWcRt+<##v%qT(
zUL+sYJ~%uSbTJ3N#L}#_jq7HrqH(<8=`r^dvJ+R}^!|+0({jjJmPE?RQ;s_z+a#mu
zy0cxRqnXIa6Lo3o)35ATpmf^3*ZJuh(<X4N${SFGxK`Qyq(Cy-VWrfl4t-29fwNT`
z(nXqHE+j_}&3QVf*6*je5|R%d!rZ&L%|lVj<md~cv4rJoCiP@?8}H7LwqrBnoH=2(
zZ&c~m88qqzi#`fHt^Ai&xiH?eD?ouvSP4GN?c20T*K*b+FzX)+u?@1_!Zoi&DtRD8
zJSs}vt+sMCV4ct|Sy#D~6Qe&3$|LQ*Cs?T|9TgWRju2JPCGf0;5g47<o{W=}tbsqI
zgIM5TjU%vkWSo&hAOI0R1N*RKL2(CImglSBl;2BU=ZMD1kxY^5JdKde*IOG@);~gl
zR~vlbsS<mQ?N}el%8bO0l5{+pDR_AdMEN+818}{pRBGO;5-J9DO^LE}7lTQn!FHb$
zvbB56xZtzOFmk+dUE)NMp&T8nuYtPbCytljTjSdRBmX)2(y}XJz=FI3jDYUqE<gwO
zZg7At+!5G-?$hT(!0XM|DS($Czg2+i*t#n^0(QZ*HCxX)m5hb@xPuK9j4WtX+0edH
z58bg<q}wg!-9$8=<v5P;2vgQ<vIH{OU?tVAwMJG-zG5Nb9LpM^iEnNDxm$h$)qrar
zaX+5%OLI)Zr$01HiZ}s8ZYb;u!+J$F{1Jx{;Azg$-ngnc?-{yzq%|dUrJgt~*u2!Z
zVjwW+*)biikkwQ)0@||TU8!YlfZ}YKxFd*mD@t2V)o$(?R;s1$n_a86bM$yV9)y{K
z3(hPQty1Qch+UXZj_`CKsazS~K%9f@T2ZchUfd8*0?x?d(Q)@y&cQO+3pYC}Tf=Up
z#T_Tbvp?KZvmIaKCuD8Z?RV-tZgu(ki(3|DDeDc#zB95ubi0Z;D_bR}N4!tCq!L&~
z6hHnmu<9`7{ZR4aEVLtl3gY69!2izN-~+gR-v|og{@-JQxb1hTg>j(0J~+z@*)}lC
z3yQb^t{f(3kNyDv7?LW_caMPwTZlQp&MYGB_HeFF02#!k9e|I|G`>6%4#}>amKfG%
z=tDT_wC)f?ib**z_Rb>m)A6`WcDgS%KbHF8qT_h~+Z@8q@xR40{{JCc2KN6KvSnib
zAKE)B+P1%tE#h~be&hdwY>%fM!u~_qO6P(p%>R)KYzsp;J2v5Nbs2n9DgWu4bC6$N
zPfAL5K}-3wI8$pjvG1Ul7v4>%)=UT&GRH(>DTEg#I39Z2eNe7|^>>*xF4)A@zcx7S
zQYlurY=Z<VN-Ei-g7l^%<Ysu8jtrll714N{MRvqqcRT*hDNdFKcx7C+23Xv28|m;5
z$ZtVqU@0Zar<v-GqoRRjq8q9h`lRe~VD%{327zN~0yv_iEKP8UWoRQ@QR}wY{ESYP
zvIA@OTv>xW1Gky-tALahu>ldKX%}p;s%3Mcz_e`}O1Ma4C$c+9*+g>(Rc4li@Zw(^
z0=y(i?3lvr+V>5OaF))J23Rc9jYOWaNm2&Y&NjzJ`>c}lK_uNy5eRRwndH~(5z7&x
z;Dr_=k0B+deAwdblL@^uw7fEc2sX{I>eLM;Ok$PXv~#1}vI^n2MX~u91WQL{n1vu{
z*?}+`&i3_!f+Yi`Fq>rMMpMimR>VTo&e&=mltoJ&%!Xw)#sesR+mR~aI|!FXIM%WW
zreKblG?$8D+qM+KjS_{hNyjXpuAB{&zi&|#4vdFZRI)5sRc#7%Cxv`h+40DS|Cs#5
zBPULs`G=y~5RR}#Z3TC5Fdte<@jx17_RpBdm_}mdaj9FobxP%^U-*sc&R36|tjIc&
z9qlcbjHV(=MAONc?^>L%eUm2Y*IiAZuy$=17uhy|!|qwk?(QB?ph3fuZ+fTv>{rTm
z*$B1P-|naGHr4g;F<g&FUVjTEH`g$>z@=VAC~|aryxV^+*X=9I#swSfu`7()gNC0J
z|3go9za1RE4J_Co^c!OM^JvhSM>hwS*ZY6DzHLt(4>+;SyX?Px+C)m`ef(a}E}DDp
zP!SMy_zGqc$5!{h7AXRJPBoh!#z(&`Os(IqydJjCjpdv5)2WmFJoZ0en)&et3z|=@
zuiyJlUAt|*8a?~@HF9M2>Z3ySKM!5LUJa~0=<C0+gfRs$?2JJ`Z*4A<Y4>q)XUiKq
z=E(f@WKA{EpHfMpC;f-yF4>jRRifKOcTAT`r#q`?bcy5|*-fG=rMpO{neLP>mA--V
zu^nvNc{mk4RO-L!?fsl@`03Thkw=@AfAIa<e)n;gt<K`>$;hqMClm;)hfGtpCs|7q
zm>?>Kgi!Nws16+F&*{hhW69^o?Gw0>OLZS?tuj<k;qUisV~c)c49o~FT4Z&MuyCnC
zLZEDb97=VZ1uJ1EJ>9CpY>1+c@(`{)zzTY|+R#Kaf&p^%PmNkhw61(QhNK3S8}DJL
z_K!!#(<BrvQc-)hb}a(34aNY3ik##hQteyOKvW5B+FQZ<Yh8m>|N6qT3&lVq>YU<Z
z2_uSLvNlllVXUI?qe%#suYyQBRv_HEDAnPzCQ+BaT_f`>8~~~T$l)<w)tw>ZYwbf+
z2T96e3q$3aZBD_6>~qmwob%-mRBu{}ikvF&a<qh~+q-WTxQFV#uf@lkZ<sLeW6?KX
zr_t<F$8R6~F^3oJQ|##B`xbs&-%gHRecn7DH?Knj_cOb5>AS^8baL~+gY-9}L+%el
zcNA8<B{6;3E@k9y9!mF-vj_Xrm#-1<k7pt==N>-&99aE&J-mLr-u!*}wEld&9AgCj
z8_FBo%!8Becv!#`^4B*e8vQcI6UKu}!>^80$O`1oL0+CxXu<2ospN@R^DFeR5;lcJ
zg<EDKvS3_EOb+e9Vvm`$9z>)se&=EpQsZ$w_D>>lP(zBz8Yr)j8lQsa0PXK0ytqhf
z0&kV(!KK^1J*hSff@mDr!nZ?TWq|<d004oPlZPb0gYy8^pplk;&SDf06c7=iitaTg
z0s^KA=+o5*=7X0?HfvsK)&<glBZ3NK10!JnqzeAA38m5+AtnKa$fa*12j+pMcscSG
z7Gp!%5n?kC7Cy|wfk25So@~O$fgrl?dJ&xdGbt#6ya)Ks{B27JqDCeHZSfAAoK9{Y
z5|CuF7Bdc(0Z|0!1fE4NhtqBbjGzSuXxhMQpMpA&@f1b?;%Y|DiI}O4DCCj~Ig9fJ
zbCJXuWM!U`Dc^D2HA>pfGTjrHRiyE1I1g5}^U7(@LKUH)U=Sp$G82Ta^=ZiA@S8F8
zV4cY2bdXZV`3t|SMw7%vT{<h(>8yCIlfsRDJS+O2qmTOg_@a0kUz1B8wJ+R4_=ARA
z#V5gO8stjBdm>p?pg389RP7ov;Dl0@RHezDEY?6_TUxZ3-LSyfT+p`-4&b((gxKv^
z1`-D-jR>}+qN_EV#me5DhbBVKUV2g;CSVrGO2$643q#9!5A1iuwgnBO?S^2Gec1Bo
z5z8j5Dy63?nAk)bB3z{OI+44ygTGj?f`D0Gl|mn6Wr~f;7YKSI9-@e7dvJQwk(i!f
z9Jr97+&^=lzi=lPv_TO>T)?I-lqsqka7uvIu1yL~06yl+t0DIZR8@w)ILPyvzzGBq
z*bLO(q;vy0b15OWGA}H@EfR!=4__27Nt^*)B|3RT8eY8|9R5oMNTNI2#cpX$&7lYc
zE|v>CxD*5^J$ujsH9G((^JU1X7@TC)6se7i!JqUM5Z|8X{*t7_Ie8sF+h3cHd{H_U
z@X=~J=KXp6-4B{H^H{de$~4XIExB_o-D6#%r^pgSZ_EX$50GC3o2T%;<Rt5LLir@%
zZRB|GNbWBrx}N=n|4`NtSiQitK4d}miCa-~<e`q}vn!xBSTA=(Qb|K|*s)SEYPSBF
zi+@`i{19U~m><<2bKR?V<KMTpmh1b;I7N1I$EC*|widfzo#}FWt;Oy99=CzK(4UQ6
z3yA#R3%9z02~=BO@7lMSMQ=WP1cgtiiIV^&xTAVUBIcY+0!Bcd!TG7FSvU=7+O-<j
zGY56^I5ZgMXTOIBNDKnyyR}Fkp|(dMs|#dE#3x@__*Yf_xH)Jyl>E`br$h*xOLbx|
zGBi2Vj(wdG&h-v7MY`;*#cdu-;NZW)XlUwQ>$sR5ic{I_*0C0^B9q4<heVsD<h40I
zIv&*z=>nnb3?Y*DJulJVdLU;IeYn9jth3a~m9?s{5X7$V@@lYQp;CvFTqLWqz#q^z
z^B!)V;0{j0sSie?egP(3UloXSZ7{^O(E!I@9SD0ml(=9Ry@v%jf`grdpM0&_f6GS8
z%=Z{Ngz<?h<)=AE9`~<QKc+J!$Sa?f1{o$<4X4%l6b#X!-I5nQP}b)x{u?=L%OkFa
z^772)087CiJV<->mYS;Y#Pw(IpS<H;G%SFm^2R8)7;<n8FgGpud?`#t>K4VNXwSZg
zf(Fg#b04mSl;o&3M;gl5Yb7raNEm(y&8xOg43hs4Fb^s5e0S8SvS{^YQ>=O)%z(r;
z7%hYnN{O}&K|_GWEdXf)gYuGm-k#=mM5%vC+rPj7{%;p^;1v@Ft}<DbKv`j`>`Ek{
z#tpKn9ZgcZGODm~DSw>nelPErVz%zAyII?|Gp+2I+|yI9{xZwUxm0k$rN<xoz{sHf
zGW_KHy4~IYzt>yBo<5Cf_}w-lFCP9bju9!iEH79tIy3TeaP!yK+lL7sN3RAL`?T_#
zYxOVA_bF!y8j>8tW(1no9X^ijLzZP}KMm@)x$MA0a*LMvo5BUw*gS~u>*AC-#HM!9
zbMXP6mxNmO5Su>zak(4adZ9L!oc14o0BJ3O|29o?GW?f36elb5|7cb6G4@m(PUqE*
zz9K(E(+<1=Lqz$QF#$Wh5400R%wNvGf^N>^H{mH2=BZuP<x}0^xCQ0iD<~uB5>YC5
z$D4!c@#Uv;<?i)T^l_fVIXZm)x*(r__WSo3EM^Wu4AdfzAL5Vg<8!2;>&-nZXH$dH
zgXi#eW|o4H@CZNzATUr5&Ic?3iJ`<${zw4d;OHUoabtOxzM-9X*RdJ_Yykm;4!o52
z`+r|VA`j$%M&sYvem?I%oUbiO^eF~m?p;pBBYzj{SKm7!ZZOxb+;8%y*O$u^Z!~2n
zN`Jj<wwq$Sg_4|sN+f7o%a98W2-bKtvN446CxSdJ1s^AwSPGztIn>SFHoZbi!H|~n
zegP?9k-!NP;CYPg@}b}aI%I$i9_zm`TTYWm_|5hc(9AB;)t2akb2=LkeKGSi?!*FI
z+~@)KkGH$YP^Z7G(l;3CM0S8mm7b~PN+~%v2r?Vdw7JM>WQZNy_e(gl1F@jTh%usm
zMkqc$s#uKCoOvHnT);yxjwAZHS;CO!n_Amooz`piI}Ve;@F@Fl|5Za$OS3%H_9Xct
zk5<fN+bTXaktJISsSXFhr=2E#v6L6fctA$VN=*%;;)cjlL}SaxMp8i%G`p5)M#{qU
zE>U%bh-avR+h@)G7ck?kG}BpC#Y+8&v*JdU_O-l>^aNKcKHo3l%f6^U-)gc~4<6K^
zNl($FUbK7QPVx!5&)%6jz#M>^XpXsSk4h$CHc*Ifjs`sHHdTZc(z(r7V>&pA*nG4J
zcEHjs7c>9zw_N<CHECs902svYMtvD>v;j9jL63cj92La$#%(eE%2k-p*Ps}eX&#q1
zoh<8XriIEblkqx=&VJor3WH@7@lX&O;hv>zOX|fP1pjb>?=hdQmXVCXJRU{>codju
zKHyFY0;>SllClh}ZZxL?TJk+(xNt{$6a}dHwLOHGI74URq+&b4{U+e+k4MSgXW>_(
z+M@9^cQF4h&4v*v?`W|Q@5{7f_h6nWv{bB6P1BW_7IH$6CAM0I&4ndX>4I=Xl0k90
zA|{=dlST6G0X@k;mPt3KQO5j9S9)B5sCRn&LRj#$E++_;ptX3^{Q~E^Fph6L@pWU_
zJ#0*EHv`$riR7pZuepC(lX&J);_caie13@UrLR>`1Rp9(wVp+vlwaz}P`Qf(C2<xB
z>N(}IWm37dPpI82eA{eH^PH*Ks_gSt+uJ3S%CBYi$t(%^L}R|tclNkleV5)+mTOF>
zGAENkG?#Z&pCz%fSj}x4nUyX6Qz1E7G?iATiCADPwVL5A$Cb8o0;8H1117d@oa)hb
z-K6x$c9B|RsMY0I`frVX36e9gy?SCDVm5PGj;0P(T>4pVU?Ii`z+^Q<yj>eae1>=?
zjdlw!6-g|a6LSX6(|V?Ex})I@iI-E5w1Q_2`O}$IucQ`aOFaO_Ni@{kW~rFyA!?T@
zW{Of(H#6f|CU6PYHM|ANQ&dS2n`qLjg56+#ye|DVgzT<30b^*|no%Vv2I&PCw)PuJ
zMvvKeBnyVWT&m<gkSzudmptjHV!5iw*|C@c^eu9p+Vg6YLQ1%V-H>&Y)Le?^oJe+Q
zd_vXoA6263LAL*$tU|<oH;B;<X<5SUJyrQv^@2X1oM}O;5uni5pU@nIY-sTXBwtt{
z#>^bYj3kMrlqMqLoE`lr&8N=gw!em|XhB$zX!ImW*wIsy7G;3j3OdgY!%*qcQ{0oI
zT0(JDE5i+TT&NNR=pc*+IV-p+pmf6!zH^l#;AD?c#pjhIrsaq4eG^1%)k*D3qNs7X
zE}%m|IXUb{Osm+_;6<R*0dlYjhhkb?iO{DJIUjLc&@DkVo~eI&wywCO=L@X_tUT3g
zbTNw<rvGBi<lCdSKs0PGM_8H#tm}B6b;B^=7J|hvi(1e4U<6KCX2=b%4Wj$+xJ({6
z(wWy~9-lz?l^UtNA|A%<(%j&AnRwWXwVCy}ASnThC`l=A61q1~wSvamxi@89`6oO|
zUquqEP@&iR?4T>>Rstlgwqh)&?%S_rC&@@~8|7XhH~j;A54d0W{g=I2lsuE8MF`q&
zMC3jm3k`?)n1LR|CLPYOXw7seaJy3Eg2>arFB9TWVDvmsj)S+c9p+aEg|3HgSQ|mf
zY=y`o3*S+6DIm}jCuC#&8_do|LuS$6?07hOe-=O%9&7kx^K-JjD!kk5Upc^HkaMLE
z{t&9G$UU>O+n~>hJaAP~et9j(%bttohKo9Um|Azf`M?HRxW>_p^v~BbTS%FaT6N}f
z6Y`{lpCJ*0K~Itd?5ui?r0nD77otPmO=DJ<cNM?gHtrd9hyN0Pz7HLKI1xY~f6+~8
zK!h9%`l%VuKU{4X^t>RDb>4^I#VaU%pvedTKhsAC`0Nm|5CDi<p!v~0U^o&HJN#}I
z;)u3D{VW8LyZL^&y3p_Y6lTbH)&~hy=D1V<w0jbNhZ1-B^=N2z&$);52@GuqfYi}G
zGGxGd_S4AI0_mqDgRMfD`Y^xZiUTd{1An+U^S*F_PC&B?Wq^%niamR8t_;em=ny@r
zO!nA^g`DrGr{|h)srxo=Fj36DtT}yk<?<z}My1fJd3Xrn(7xMBpvV>Xu>ij+geM#K
zfJDraq#O+HlNZu<nIgWYN0nL0tbzDaSD@xROZK_p0qbH2;g4S&eTc#X)iEmXFrDBt
z%rhL|WK|Mel0z;|>G>^o*fX9QJ>@;{=Cdtsci^S03?_)`Gyt+#@F94F?b~CLfMuv{
zuNnX$Y2GX?C8T<IRZjS>&>|3~15WCoO<b^45fI=q5xXFa7G=kI95Ag3In&<D_{sAE
zz_q=6<Xh22M$Fz_sYg?SETWjXQU+V69yV4(*-91XR}1x^ZIn~w8Wtko@4e~L*mb+K
z7%3D~fXY_X7)wH@Nd1GhDFEvopJAh@{x8PffjP5qYuAl!+qUhFtq$IxW7{@5wrwXJ
z+qP}n_Q_YfcI{JVuT!;Z{)0KIo;jW|u5sO28;-R2eCxZ(Wp?r}PtUz=Tyk#{YlQcK
z_*L0!6;|uy3%DDz3{v;|gT{Q6iCufyCSX+#Ht*%<Uv#==Y8IP;7d)v8a1Jq_?&0a^
z);*gmbCXA_Yje(?#q^)oMUVwE?=u5iGE-C%78632(yKbi)RXdAj6p{~fgDAD;-!iS
zq-2XY)5CxfybUyX9-Ca9Wr|pkAp2M77e%5EtBR8|-;V&L*%PEV6nz#)k+{_D{3;uZ
zm$9k8%mk~KN>8b}$aEtiv~XTM=H>ErI;erM+W@y1pDp3KMCWsuQR^NuK@IC0StG@j
zD*aW=#sQmm1JJfjnHXBpC{DxvM9a;X42@X8%&girtF>(%pbhGFH%qkMXxD4^`20na
z9SS%hDj8Gl@Ea;jZS<rK&j>eqwZ#1q?^v*ru_uVBkndS@hitIzr8t#|=Mkz>)Yu5{
zsoQJ6z+(K=$+w=o#vVN_W*h%>l~WLRf><DQlym5HlsHSV9|7-I=8Hep0Ka9606|E#
z?wP4U^;jSGS$ytmrL^+5#4u39J1X0b9-1~Vs3aJxu4ZM)AVR|NEk0_5E*)xJF&koP
zoJ3h@Dr<XDnW!kHxQUD*bi318>D`S;;6Vhra@5eN;cxgkz1jUL=YQ^T<zJ6&(v7ie
zQ<x7#0xc7~ABBL29ra!~941<-drh;h$W4Xy$a%UEs&3(t_k3Lq)wxr0%Mc3RR`7)B
z4_`6YdKHdfcdn2Aors$yd3|fB5iy`@>_uScMaOh?d)m1%m!QtBuzY<UwsBKj6kpfh
z5nWW--r5@R?1(Dx)v@>R>}f)}Wp2ST0;+8qbL&_9{&Nz}dmo&{_@Mn33eO98b>t7H
zX)##CPvefknwERrg{N3F)Z)*4dcvy#d)E(UwFYgg#a+aqP-mlz&$V;`J@P`7@Z!2a
z#*YcEv&+5n2+|VEYC)fVT(VyasF-vwi{R+$&%67!25{H?=eqv?)g}Mms2&$9*MC>n
z|G!YZg8zl;$+9H;e^9-%;h$kWw(0&9#tCsPsSMeVa$ADO;e}^pH>Lme^5!(0O=Fi$
zgT2EW+mBH43t}`-91%kZD`@v}{pMXs+W5us2vI6*gmC;hy5{X6XvYoW25v{mZ$5#1
z2l)W`fLr?h_npLT=cM6s$JUJpWf%x}F+3l#Q_Bx<{&X&E;$L(_u@NHkeg1fP^PGpd
zElHVJusdxVdgF%##xvNClFa$|S+hI7+njB&OYSc9Hs|p2<$NM8U<&Rr3X}Mw!B!%Y
zT%v&=a`WF+QYPw@EOYN4z%7CGb*qP=$oOz_{wEnnmLc+t#$J)4{nb&ty+$zxDDiZc
z-}tG-?$C8nhvBunTz!qe^EMrQRl@BBI}l;sReR~{F3^JTy7Mm*kdr$VImG>x_>IXa
z9g<d)gKB4{8LB;7bGl&@P1-;-j;>!vxH%J+OrvxYs`{_>rp*k~F0BK*u9E*3qxXtN
z-Am~yB`8c#w~AS_dW~_|@kjg`Y<|t`SE9I?wQB=0NDj~c=JY_VN8&VcH_r^QytL`x
zG-eQw6!H5z<c=B65LCEQ3clD79%;uovCXG@uo!2c>Z<Ju=M&;+)=f49vl4s&|HIBb
zWK`?_mSsq7#N@l?mVl9_-9JNmB;up>65(rQvuC-(J_V1tbV*-6P4jgvGY{2(A!}G=
zO_YZc-)yPujYTgV{{hVIQ`Zlpih^EG%Qxw%@~V(e&NUw>@#nnjOt<p0);t^V;;j{G
z5J|6~UOoWR$^xv{z&KUhOW2#@K}+W#-chHV_rl(I3MbTk859Bh#1Pn179^lCZ~Lj<
z3N&U3+UXhmY>M~p0xImo9n*T)S|$)VU8)xQIulPOJ|Z9+dq_UA|7i3BZX<aJ41Zjj
zQ#XrlWp|o79&<xzgI6*u=dztfikSHzh402`@u;qpMT)L?{TmY`%wJ{sp&>RMzVtlM
zUwgRl<mbbn%=JD%KFRmO!?cnbQ;jt!65og}0T}{f7qi*+gFzPO$fy;KKGt`Yd#qP0
z;$0)oN38?g@D%@SSCKI=X@{Z9pdTnJ>QD?{CVH8IfWR`)b5s>`hSFA%Zh?d-pSbF%
z)+M=Vv`m*%l_D%y4%Ovjm9m!UW*zr{7-f1-KGXfM;OkPH$bsDn1`w&C8x~|Sum9dB
z{pzTQ9FF6dJzCjzY&B9JhHzN1UJc3v_dlh6($6i9u?9A1MhU{RPhG1YIHNk#y4zNQ
zFyePvP#;I4>_{JVNU^#0TZZ8Zdd*%3K<766vnYOT6h#smQoVXVtd}+`Y8o5%Uv+3W
zzx*qw$(KBdPsRvz133<+D#EiyMahD7v?5r4+pR&th34l{;icV@ANnH*L26GClwlA$
zkNS{8sd<+1QR7jnPNu1q?FBRi9RoKBC|1XcHI<~L&^~AOA71?q6Ke*ht&MT}ozKv8
zh#ZOSiB40bEg({qsV|eAox*l21*tz*E|ZZ$>X!xdv%?mYgWk0bVls9^^q5gtE09s&
z8a%x6&}XP{Az>fDk{&_14hlY#?XBz$rA$=49sB7!Jh7yN2OB>NyWo-c4hmlry$&e8
zkECsxI58DWWflU@a1vccfzZg)E6y$oBA(Vn!_9{{sG%A&I=JsJTvB3`gV_uYyG)GG
zBzHhpAMo+^%f%F8%T;H-<IG>mZ{*%>6@i+5&5XJXgnaqo?m`WmA|({%UXCX)6@-6w
zPrG&}{W}Dvcz4($Shu+It8QFk&NE3PAaGtouI5SGvCB}@eHx)v>C!emB1vmWXssiC
zT(NZNI*t|<wXAN164NXSedW~?JY_vEK&oaM-N_yS_-U61o~!YHj*^OxrX4Cx+ll_H
z8y9pZ*u<`q7yOgiyd>cL9TL-E>Ds0iA<-VCYrcFTez$uQkQ-dv<#Du+hP|r!C{I=i
znNX17`kC+j$B4QlD8XV_dtF5voW1N!<NRcy%|oln2kM#zHgec*E)ut0zy^}7B9_0*
z<r|hMDk|?7dS_$9FB}q$LI{0{_^iQjz2{ijC<)nB+94~$6=ns*aNWhi(w7UD=V%a*
zXs^vWmMb~dDyyu#Sm~Jq?r2|R%^A$W@RGvx7oS;Ma4Ppc#FVxU;1|CCnyR)L`kU((
zP-&R8H@rtJ{)-T9>~nay4@H~0-|lWUa!Eny`FQqbEI-VS?2J7??CJEki{QJk5o6ZB
zxd&DCJnu@-woy;w_m88Ff-cfJb!BLmRs1|}CYacF^BDp>LHn-~Ax|dV(in-S+$=@(
z){7uC|4N~r&)844IbWo@&>7Ih2SK5yT(*9V&m55_paJKmRpiw`qX+%VteLJqjf_pM
zJ>)mp)(z1+`y*eQvY$xh&2-MlO~`ehIX!Y3+kdc?K0zhqJoW$M&zG6=KcI3f|D|mP
zFls}}`9HRqH^>S7I64MDVxc(=2#1JWtyq=Vw0S(1Z(m9KKk1wC#)$F1Wav3G>6YHg
z0bZ|NH;0G24^KCTB#Q>AVoV0^?@d7bI}gyA5bO*Pe&*2Q>oJEhlvBaQ=b|Zl+R|-s
zn_6!_qD3$nRzDM<C=fh?C=i@DHWKq_p|>L^=pRn*B7VsdVB4yk4KfRn4s}w~pJUt2
ze=~41lEo2@jF`r&fNaib++WU!2|Hg8UHE(7;Fi5s`a<gO^VlCsg;*%m^RasI>6t-K
zv5N{;9f~jUc`E_TN}&G20DISNlwi#K+wQ-R!C~3qo<v5-p7I}X1A#Xn#rX4WT#V4Y
z_s)?Y1*4YZk#&D^Q;)O8|Bg~Jl6sm7P)5VVjQyVKhrw0vSuRyPYFIA&_Q3Yx@8<ug
zpJTh6f?GMPAw>N0kMAi&H_D!fb^iMBeEnJAl4Qq*H_c2QKs*?qz8+_~OXusgSH@b#
zsjbJ@DHLloE9|Tshfy(NFvGQ<(Fb>0O24p`Q?kFK#-rXc@ssizyU!Nvi))Kp9#XF~
z(%Wdk?b)cW)R(T7Z@0Y}SfR$=9MFF`yeV~OQk7NxpizR?B&W&d5iCK?*FKZr(&wDZ
zX0rTyU||dm!TEE`y$tWf&}s{$-%WiKCJIyL@XG8%h$#)2A33HmAFxmtwh1dEGA1~F
zlPr4~#s(Y5L+fo6CW9pU<~>;w4>^Uqq7J$3F-i}gp0(=dS5p>15MxDBV0I6JKcSXG
z7}~e)#qHdlpbXPfO>O*Dem9l0^xm4fWaRida!WnH^)~gE)~?f(aLncr9RM``sKQf8
zs76hbd7jEbXo8$&GPP?>*VKjMCdKoktXNk>GC|?P=vBC|zyNe>&oCtCu>z)MK=bv*
zG3TdCg<seT0tQX7Tb<(Rh7D|~)8+(tslba2mCx&9Eb+Oo$v1>=OEP&Zy<mpQT@zeZ
zC6I7LG@rS1CFoY{HbPT`UX@_hi`{m0D9;(pq>g6O=Nt<sDk3n`S)Nz8(xfQi6=24A
zQd$!vak(9TEIla5>~XD7raQsk#4N!jO>@BRC{Yvp)H?n&xjEu#5{adXl!;he_0(Um
z4T(F{8T10>rV5%-99#KN0PNbic&j|wo$v>f0TsdvbJ??!u_{e_t%KZzKl+Tz0T35#
z_4)r~^RQYi$qv-3=%RUxv*@$C;XmU|v_`NRwH-bw2;@DfAtzS%#NwaB#<UU@bibO-
z^o(xT@mw9U{gNm}9Hv@fhsYg_%irD7iHMpl^uF9?HJP@g4$&zm^-fGwX~=$EVi;uU
z2TB8{QIr~wAhfmzNQYkg331-!&%jXw(ZIveQY5fGIGt5VBIks#MRZi{?YIj><d;?p
ziOHcq`h1k(gEEMp-8h3+kHN&}KwzmqG&LAYULPqyiILhg;^y2@6cSY$nRbA0WLs>+
zo}!GwKXZH`IHn(O#dyh0GCD>nPB6_3hO2>iZ~1ha)MdWwP^1%|+qdfF1L2mskk&A+
zKPD`~4YY(!T&R=z$efZ<sY*5HW7tuWNpo-1{4=!?pO`5-y5_{Asm#<2Q;6)LPD$IX
zR?BfOEII*GadB>$43wi+h3O5_t!*^c7cf=3Ip4K5gE-*xrCT{+$hj^9J+Qob=AN>j
zloe|)BFQA3fgeaFX()WiH5$!9LChOYWHpwANxb}}`#cF-QEWyn2t8UL1_iGR+ak17
zm@ivGn@^;LW>#xke_?Ev$4*)nRnp2#VzSk|f0&F`q%3R6es#VnWpDP+2_O2SWpm!m
z0ztiGb*F|N*uBg^0L|8Qx-v6gn3dNte!q6<xA1Y>id@ZP5`8}Db2f1|ur-Ek^Uz*9
z*Oa7qeA=ZWxuq2+EwnK$zq0ICJ5LUX37o9mGsAKhZt-Qy8>qI5oV+5E8rtt9=Oz7W
zH`3(GwbC{=fqKC|Z4lH49z|iOIvJ1Qm{2mzD8<KLf^NW2b>!?0u2u0P$wH~&Kcj24
z)dgY{8dMiC>NCZ^)lby)>{;*MRSY3&&uz8V8E#pYX)+hFPPgrrU9PRfXWGrbeS@)y
z)lLA$sNV?b2z2@$b7<roVd*efF_l!#LtQf$qvzw!uPP8p(#@aw%2kHTnyVi5Bv6gi
z^xyk{8~j*b)*vM=&YciY_QyOn$iF(7j&)?7zQR<Aq$ey(fqlzxXY=R6{%e9A6#{LL
zGSIu*y@{6$+=@+ktSuq53~_QqX~G<7#2RZ0cs#*s8WFT1+AZ5s>1*|(CQ*Q%I3ls3
zM9S!T2gCS`*t=|OwePVOS}?^%u>I(~V*>EZf-(1>*jNF@Fg~qW+mqEKbysye=p{Ga
zfwe309s2f3ixql&02F=$uk%XP)^c97ZAI|<w32Ey_FccfLrsOi@?Q#x0Zhun(Un&L
z_W0-NY}2<>e*HflYd@Gt(Q?&R%GRLkYYh?Gc&Q#?*_AYda_6vxYf(U=PHUSw&E}a-
zSz`7_6vv{<R>FjhjSFO1dUaFQ%e|?zigI1jGfcgXmrgRvhvfl<rb|#w7z^^3(=X6~
z_tNVmxn_?)Tm+mtHjdrdXVvdzH26wHwl9ousz@JccY9zw7n{9ZTAhrHvOoH=EfZP(
zL;f`I(+2qbeSU-#0HXbahL#_*l+*Y5tkzH9>=9<L^J1+OPt)tXe|E|78r;TAIwzfv
zZS*O|^|@?GckEA^uKI(?UoM0J*FZm;t@51t96T#kfVpOdO|Wbl*Y9Fru;b=>DYrC0
z6@{Ebg&x#!xO+wV-1>6H#tEis9Dtdods|5~dRvv`6BbU9e&#>cB`p7UUBb-9_TM!H
zE;av`+331vhLEYB*z^Sm1Ej;!)^7{vMeRYAjsI}w<NH_WL2P~tVom42EC}uqfD+m9
z<0R)VO@(h+kGGChero}II(c1i5ZmIC8q^!uNnhdn+=5Y1DpP%rH;&KR#Wk)j7q7gf
zrMtsd*B#zTR}7G8plo117^ZD^FoIe>1NUg+F3T>s?<W?Q*vIH)$f#78olbvBkfO~a
z!5p7;Go_b#;vRDAEQvv$-_{9i2R~EX-D*h+t?bF{YoconPy=(UaW5xhxR*{TLq1GI
zb1Qj>B4QR}XlIg<Q55w^x%)XsWztCr<ls=N1@7exq~DPyOh;d_+F$}{ho-p))N7cY
z;njmGG;If!PVant`~ucrn`|lA6(fVAnk602^MAHEPQgK_kN1rt!~N!;yGv>^wug1d
zarXK9sUanKB^qF2?Zc^L5vt~by#1E4cYzTSw{lpD{deGuD#X^xkPHwt^H&Q;jIq&?
z^4?fuggjx$%h9PCYG#IDuqk=<+!^f3>g)KGNW5^Whdj`cW7u?rFV!-QO9epL$HTF>
zlneF4KuoFMx8q@>uNf>@y%mtTUysIv;gWU~p=jbO;_GM=!q`D0Xx}pl9W%jMNLOQ~
zZhsN?s2_4tI~t8>9dc?aCZoIMd{96n_P=QR^zVBeJ3*X~PPbrw7vF495=NAbWHbw6
zyHvz6;&TP4=!xG_a~<+Tm11E%x@bG2zQ#fX`wOIc9gGgFW?V|r)E8NUKWnRHfpdY*
zT2JS7aJx8xKy)RT%e!SVtL;CLZ*{so(3;+ueR(rB<QkK!f`_h7w%*wBOeCw`a1o0c
z|2tZjpLebTM}b^GtiVjUBx7rvU!uF3B5wP00!=X1sHl>MOaCd)LaJSiC~{<lg>JBa
zow>L!rYZuC<EXUCR<IDJa@N!w1-Bkei)}8i8^*3}t$IWCuBS!Vp12K5I-f6bML-{J
zk_I}dOn;|wap=dyy+T8L?H1cAE_A$PSCuDL9YEYLgHwB?ct5>d-C(RX)uv+_sG8U{
zM`rz)XMEIIsc|MdsjMz7bsp?ml_D+-jt2W^nBZ%qqcyMSl>KZWQaI=L)d*b+=p?y4
z=!9@WfhV_tKCOZv3{Sbi9>CRe2vsr=;lpW1Zz_x?{oE@_`sVc*Mcz^$Sf*;G3LYEV
zg=xT$hJmecp(-;g@7SEFBc!IsQv%GFleRi%qx1Om`@pxJ&v9(c)MzqH5}X9SH@e((
zT*@|a994O6L?{t+YFy9PYmywMkbZeSTLr$*dN%1V%JF9ekGOt+U$0`sPVej-dS@)2
z8X|t_d1u_}YVPZ57B5vOB*L67)6_~4y$ATTB{p_l7A;O_4wiAzxX7`ew=)Qj2YUT6
z6sugI{m0Atgk@Z_U+%Vb!C`p%DNS$DqO~~jnXglU7_nv3`CgQsc27D|9a=3O8h$^n
zXTL$(aK9YUTL<_g;GZPARaa2)NFQgGpB3@4j#RIJrN1KgsXgcAIV=Av>t2WM{X5N4
zeGI_OkivImv)F>A?SIRi?e7Q)&dcT&CviNmc9*}F>7sAlu*^70Gda{obXvv*oOjKh
zrmS`?rHlMdf;vNnzVAj9jcE}OICQP@nM7yYq}+3SJ%maOGyGRHhbP<z=xPQf5;&iE
zof=r_=?~I^wg7r0&q+^BD1fC9NM`BYDVd^4S1m)s8C-}X9yE3`GoV?Z<saFid<)F%
z01(D`uD`CjaUyp5F&KeAEe)>`6|p0d5GMh9^Tf+mG<f0J2*eD7AHXBXAa$~Y*l^ML
zyV6*^AS1{$6`sR~7Bw>4I%|O7CjD|N=~o5;fPE{4772eOg%<sMr%>`4CHFePQT<br
zk?HINlG2+3?@bq7<W%)1(Tc(A*{5pouQQKfMt4wV=G#xmua@cX|9I@P|KIXcRwh>N
z|Ed=}iQ5`a+swROqc;K%mFGW1erbq4+N_#{f4T@!U6;9H+=|xGGxqRolNo4O&7WQA
z$9pir_9mL6;qx->Pw3;-(#7ZHn&QSZLBq6nb2kor0$kGFb6Tth3EI;T)IYx5zn#N{
znJ%FIMs%z^H$3AiqwXvq_3r^<23N=xc#1Mm$dzm}?jc&y>F@#hVYATl8LS3+%R#go
zqze9lgsOY)m5|ZbeJ5<ki(XZFu(CIhW8`nL=wVU?#94cLuqJ^Lb3<|9QMes#UleB>
zi1%MLvUGIsneA`c?J&xQQ};K7oc-bZvc7lIpX3~jFl(oW-WT2gFp?BBq6Xu+uxzdu
zFM+(+dLt8buz(uiUu~$O^4H<hO8CMDvE7(015jrTL-fobhfy-Rt=(Ew%&1~+<_k<z
za+*U~J{5B~W_|3`WZJ0w*(tevf9KTs_?I1F?$@BDJrnxK<{J!o`a1dwe@k`f@`<V5
zd7)5cQ_uxvUb%gcwjOS3p?Q1MC_8&y5rdFYP~YnsLZ7FDR85hjmPQ?+ne~vd@$TJJ
zq_n3<?=A;)=0U2V|D49U1(}?1K<uDae5i;+&E@1X0scb-ZSj*dHv3r8v%h2-5t~V=
zt{^rT(*E4!nYe)3R?-ET*c^oa;mE}UWwuLT$1#VV5-QZuOYlyr`m?U;05Mt&B9YRk
zWNpr)sda6k-K6p6!~2T~0fw|fK-7U^>r9wNqnfD>b3)KoIrHzhkCCz(%gTfN6FExY
z_X<FIg27S1oS-x28{_Gs{?A?2`Ru36TzdK5avMD{*PUUNQ$FgxLctvf@D*N?R{=e}
z`!xia5tqHd;osCGxsW_>Hc)|Pj*$w<YLZHsb1_+UMC28$2AE`M)hOm=|F&>*kh*f|
zgZpaa0olTgD+bZF5AjL&j{35$50fnm{fJRnQ?YI-f1+yKN_+5YBq#%A3hOB-kZ6jj
z<#R}!@g^PTh*Ge9AI_$_Z;kz^(S@c@nxZhQld#=qZ>3;Mij7eFT*GuaWwqV;>hm@Q
zaU=25Xao|0_I+=EX|sMR^lv>EBNl3w_=|~K_ayu>=<dtY5&0}+;yD7Y1nGI1O!MKv
zT~3GXdxfq_E@$GjxvN|{Cesp)<Yv_=zG6NbDC)93rF0qsV;QTOmuK9{Ew21(Bo+wu
zW5w)cs%fHiPbYnaaS{Y~>VXUGjrnKhA}w~}ic~DD)(z(-V)Ln7Ca#OoU5qTnvomy|
z6vFmD(~OAU-dJD^v1e$GFU!l2-$265)_8{)6Tk#M@6Wp_p*e5A+byD(56u)AOGE*U
zXN$7PjH2}PF^vu$gu~ueZc}*;w^yq<Td3>YL|gMTZ7Zy%l7Gi@^x`M87!!7(3|1c#
z2fcFXbw~d~r@o^Hg<EpC2DtMI40jVzlM?=ON|K4N&IlwoPgplfCwN?O3nfOXb~7T@
zGrOw%0XKSk<g23teCr%XH(FwWDPHujJ~&J#jjMt`_|n*T$OyD`G#6eiZy{V9#>l59
zci)s#w&ZdTPv2U}gt#56-(a0>KHxjaan{s^I{LLp#I*7bzSp(R7L?;_*&i`#q@+2G
z;^}{Fk6DtCQ%;Qvj#%d3Aoq6mS@x-3fG13xp-5ecfz>=G8*z2jEY3d6ezCk9WM5;C
z3HE78_xFqttnbio_sZQ3^Fyc)R%?g0<(ao#@K7Hv22yMt@)SA4wRl9GPh5f}tb=+v
zV0YzDz{i$_z|tP@X)IgEtD;t!3>DIcaw)EHWj%XjQu9BZx%*6p>}lUWzGcd+34wx+
zqQb-Ey<lNTT5TER_&bErscwFh%P1^{dj!CuMSzs#!^|<t4mxTH%*_|GM}2y<x{2*5
zG`bmj3h8MRvx8m150&|17&6r?H{W~p8U3wAO~*YmC9We*FQZocmHel*p1ORi$ecG*
z290Ab07oKMmTm4-ZVT;nNY3zxy#^0e(ZJK3L1@!A3HA;)3gkF`K9i@-A&}rR@qOF?
znUX=0OljV!>`<P&<o9-XP2GH141PL)ex+V{+v^vErYhRWM)oYL0%{kFzVxdB+|uk9
ztM6!lX=6-CW}BHQV74hRqD?i)H=nFcF9%AsopJ8W^Ss#%-TsAa7~v6tzKGMbAu)Tx
z?B{ist^Rh+k_q8g(t?ED3`zx-$ZS3ka3+8vpN2i#p}K*NvObB0qjru<>?4x>e7B?_
zXKhG^oXm`Cl0h9oO;|rdQ)Z41jqVntfZ|6K1_m@CkqBe4!}@1p9|K|;v7rfOp2Z7K
z9ARnM{S~jM4KeIK(uD^y@k%#LakNQ8_XP=;QzZ~4-g3cv68{RHztn{XYjcWKV@M@0
ziyTjz&dxo)L6=#lNe{m|NIO_a&3j9SL6umX5&IRxXc;J4^tK?WiYo#jv(PRp(UOe8
zoRN|XsZmC@)rwAZc<e}2EXSnwj|9Hb>HDVU!6%?s1H1y9KXJ%j!^Clh$9asCA1ph8
z*yjOFiS&B6)-(w8$=l1dv3A_(E%zD=9aZsiem1oqvJn6DJ;ToomPIzzkcrJ8nu4h^
z&`pV{uGN};-7V{2Y^|Q^?OOc1!NUX_1gDp&+m7UxzH-zxOvEsqqs;@D2U_JvY1#(S
z-|ZFug#F&h9TNfAVecxb+rO~LjCt+ogaIY@NX(MQJIMOE=`MvX@VaXhZkh7ve|fOG
znxJb%7h{Ae$3)p3)Bm{p#!e$S(<|rL|7)asXXBOO3hvZO@7KZ6Yy$n%+>h%*7&JH2
z6}eauo7`QJ^?-dTug7w$QOIVS?EYA98r77Ba2nLn(Mi7(cjeMQ)t(4pTRxMd#^d%{
z93NN?AjZ}pRE}F^`gP?Nds2NGjmR+tSNb?a6~eI%)u_m+eAp8iP;3*#GJ(EXhLqDd
z#zKVdY#m=W9Lv0}&qnuj^z2ZZ9cw_{jo7fn#I2H(4h`>RE!`AZw|NjbHq<+MFn)>$
z?fr<NF1|MPTAj6D$c-6f?G32XW{W+aAMJsphhx!o@c2aHx6IkgDML`7Y#B0#>O8=i
zcjH%QmoYqpmq^?42rjq?IaTCoq%_Y6$6(Dm)c)P1N^Kd&sQA^T`|J@joHL=ca<0G6
zho~G(?7g)lG8iYY5K3H*)1Z-2s{KOlQaCrU(M(x|-i0hpNShRSIlg-}HUL~z#5eKO
zkaVBhyddIuc|p3aZdbwJZIht(D}y12AwuQ!>cC(fItw$icr7%>r28LtkCaKpzE)V~
z!;=IQY~k<%DUMMiZya^6l?)QdUoK-vjP-vH$aFCJsks4m5yPD8{jFTz%j<y;#%$%;
zTszyM>kn>I)euURJYy>m*g8fSv#pslopL@xP6EnHN}T5(%5PPkq7sAri_w9=zyBt?
z{4OPOW)0Vd3y&~|H5=7tkG5jzdgC&*_jm4BcRj9dL)BGzYMqU~wC`-4Wp_ksu>57y
z++#Hl!ukuAj=kr(x%Y!vO$-3(57$z;t|z1cU(F%S*RX$JK-H_-=VBf6Du6{<q&&3d
zXT{dnV(7EZPIN9-{*S>ail-?bnIbtd?Gly;1{ZDJVq|~1aAA=k&@P&KIsDemlMkC}
zFI9q4n00|EH&CZ4FuF6(Id}L7i&8KIEm{&D&@1A2aS&V4d51C)t9S<fJR*Y%S{%G~
zB5e-`S?#~aOVwFHd*~Uey2P&Ga#mmnNubb|x?ke#g?8r6*!UB4UL<WU!uW`;O$=3x
z)}-O1b2yfx$izQVGBW4!jzq5faei>yzYpCY;B(DLp*vPXxOaH|{0`T`-rwstX>c7u
z5*b&=XpVd-A{BAc7S9pobmwpKhgK1NPJzk(c!41=8y3|*YPEG2h*f=&0!MfDOd9Ll
z-3V<$+?&I`y;(p8Nlu6VkBS1!ikYnWa3r+^2dJC(#G?Q`(o|{x0+!?~qKh%!>%~J(
zw_jo(OJX_dg!rc<Hia13D)1-un1p%f^f`KGGa&Yqkn^B?vdFrVc2X=K5~BZI92`Ws
zcn6~M1NbnP<-j5MwCY6QT`=u`iUm-dj?KRRcw6B7zq5&0nb`h2{dK37OblKteAjeM
zsUQ<TX#o`rj%9v}IW_}yI{*)a)CIvMb`$UJ`Ipo}fol2LnX>`{|7KA<Rs7z5L1El@
zvls&*p{}6d$6bhreIKr!#p{PNum|xeqohTHi9Jvs30;maov(MivgTXIoP<o6B%wq%
zzjwy51egN9p9fGQ2%jJi{6fIbBZ2*v{Gjx(ZoiHj)6BLdQ$-4%55It1D)@5T?frEh
zXcSVUION$VnRmKByK~3ex!Uw~NGf%tPm?C1Z>3Q=AMnB$|9grd&h-tyaA6>g;tJm0
z@+6D7H3s2B_n5<bwA=aP^9e6F`d5?=b9aeF5gK!+{$DrAAozjpv^o3%b+e#Gx8Ghy
zarBP6FByv{nCzn?+7`oSFq&Q_9SD!BE+|rZC^l7{3tuSgSlc<uSroMF$}xSa7FRYF
z^G58zHCj=V7zvfTRhp6yDQ4`aj^;dy4_17F>xFIQr-kJz7EC7t6Ml0$QJY%iWTQGx
z&c3Lik*rauQRU$_woVw`rL<{5JmjoyjAkF=nNuUO_B#En|4Pdyi;>uW$1Z_5l@eda
zSH3B^XUNiiIVh~&?aMCSZI5NJ#pv)0;}7z&qtc7t!Id>vus&WnNF6Crg`h-d_p(s3
zOc5-gW+YTGG~hz=K}mSv7CE1zOKte-5vV7psVn(P#vi90(FTW1Ku&O}Ae4)(`DBf)
zSszxmN$DF`?PllG`aGmRD2jL5O(>_SMa_E?)meTX>z=d!QwZ%ylXaf{RD_0VTeB=E
zSWwyJ5DjOHsu;*S7a5k(wtijM8gi;2E}lLylzWmTof}h2Td!+eiAn|;2k`&xSDniw
zs2AE5?lSq(+A2WA$iAT#yfjS^8J2Z~*{N8gKy@!Ga6z@7?8_2YiQ#-&suZ5TN6Mig
zU$&xnH%`^mBEVeY(Ys!+>>~Noa|BB1Hf%M?nDtRpyI5i-x<(y0u)H+S{7+GkHNUx&
zZ+dLZXdCCTFHtTmT8GVZ_31H?zHWhYg=AhYRn0ufZ4XB7bxc5q-^RKYLqX%h$+`%0
zOnq9wcGBn<AR4)QrKMU;Glz$N6CRx=e6ly1#meC++V#MyEyT@{w;bsh%CJUhmg9Dz
zFtI#b)}eoq8>p<TW7Z>)cMTX~7^hS-w;5~kpWTmNN!4&ni+?n~`cFS|`-KrKYNZgm
zM+VQG3HZNo$ISh$u?(DZA)N^4)?uu3Wy^;so_1<h+NZ9dq`MPh9{Y3n*|OvOsaey?
z4F{Q_a<40~FnP|D^4WLKOb+2()_s`n=3)EX^C)F!=^8&`w*Aez9`q~jOkKKF%XoDK
z=>`cfP7_qBb?EC{cRMvKQ;F!SlGoqhUi3><hO2}KMrM_Yb5SCfiCjYf;VtfDE-LgC
zE~?rf+`-Xyi1xdtL~g$DtOi{7qKRGm1eVZKSEO}kV+3hOnY02)MibK8_eAcgJ}#AD
zF6B+dqud~5B;!u`aCx8kKm{E(xCA1fQ1tY*#90wfUTQ11$9$P{0u+}bM99Rsf(b0)
z6=T2b8B9N@;$?g9a=<vfBQHWvnQlpgPm|pko|S1@B9QSHE0f?eo-|sqmD)XCq4gDW
zwcC+fe3QGH{Ccy)xPV|IL={-@?i94e6vxw2W|jorp_So#ru7AHp3!BH6Bm@L<sMh9
zW9A6h)t;{aHyCN<9O$+;j0t$O-{guGu$J1Vh8ga9;B=FQBW?B?s<iZ)$ws49PlISI
z8fVOo>y<k%RF&Vqp#kDO8xFHO*|q@r^{1eDg4>$M$-Lo7Za2t#anOMB6Q$*6VDZ(a
zM|Zs-T<1?ZH1c{Rp0;a%z>ws^5xK`4sy_9Yq4EmR@1?zm$uL$5(C{%C?Ihdq+6etx
z@vSI9o|pnSTD&H?N4cr|mCAXF8O!F|03GrAP`9!v!%jQ>-^r@DzK8Di-99sUS}hE@
zi={jBePuI)k2c{rG{l6+OLM^qFX^g$K9yt%(GeTf39giEG!}{avCoH-$#r0xtf^4a
zBuLcZl@zE}+RZPI_HJ*F_x-jRRK>_tK%s}@F`+Lapo7Mk7)7E$ogZH@hv)qzaTeTf
zcwO_W)_`B;Rm_&>1t6gllBA;$=y5vgJF=ecp6a&C48h!!=V}45$ujQ=+NO&l4D;7q
zr=>6K2{*((Y7DdT3<c>&2F|=Mhgi%a>>2tG7;M`URnH>Sp0LftJmp(<?Yrcc_i~k|
zMmkmGP?i?0{x6vbvobupDCR1Lgpj}$LHD?8K>UdUPP8{XSq_y{2c6SbZmzChJJeQf
z_CEYrmjfWK(<`$QjnQDwbuHJN6w+J9=+TX$<1Z2<#6;}Ta2Fa!HvjsFK#J~_A}@8N
zeR1}s`*ua(%Jtq-4}s9MI3S)v^GEVD>}e-<O^OLp(h!I)F3fW!c5{s@FFEwPz1x+f
zyC0$3>z_3kqK2<Zdm>YB3){=wglHzk1AtwBXC6k&Woa`iJ`k2B;qsR2@2L4=`oP*S
zXza3(03O-6&njp)nQVFMD(#{kHIxGIozwpHS+4WhV47XkZb}!$EWM>i*KUJeGVDw|
z1^?R?mU<chIv~RWH!$jt^=jW1Ftvx1VAbU7IB`<ED#;KcZyr-t&1zpIqmIVqbcp3t
z+IZJqPy)LGZLw6YNC4mQD7evXgjX9mnX*LKs+}N8iq+i-Vc9o1C=bXt$`wB-2ldD0
z6R)O>r9oj(dlo7`pU7oibjL#6l<UO&{rty}?{~kv*(mIj$zLedG=CgQUZKB1F+sn5
z@a+9eOZrBMRBtM>ni7C7qWfs$fhR&TNWChlI5ha9&3rs~5z&+y9dHTs1iC3;DJt-e
zzsta4A2khR#GR?BBO``g6i{KUmW6%Fkrct!@_46Mw4vb4h6R((&AxcRz<&KpGmCDx
zjsZqr+yz*htcmGY!18-hAK6d`fpFTO^BhewL|$n2ZKO;aBS(lN_$r|F5&fge+f3UK
zyUWHIO%(?mgYGgoNs#um#ogg6MMks7+^ehae?x4PcNbYk={GS|t#s8f<?`n#Z_Kp)
zLE2{gD+Lo|fAkz|G6aRm;V55|Z3}dqosi{75zkR-zvqg?Y?BxGyex_rljAGG1`pOt
z6|>ckS0<D%;x*2H)6GxyWr~pGY8=OSD^X^E+h<?-vsmF$*rhTxyeY64ose{eD5k@W
zh}YcJyRGQs(y1;nb5*7z81WrE>-1VprR!hJ{I=8|Ampwh#k)Qb({8uLF1PGE8MZwg
zqr{=Rg(J7M^qtc)Haaey;I1!HZcdS)8Sp2{8;@Zmsk83kV~&a-Q&{w`%-*w!TYv)1
zACB(<Be1uI^RIQ0vG*1-_&E}T8<W>-lni2z=ElVFQcciRO8TRI#_zAEM}<FM2mufu
zl+n3!p&%egdLMCw)Wz<&Xbju0s|ohJdpM#*wqfe9Wpugh)u5LBKPKOsa-c$>c|iBD
z8bDQHeC-Fr<v89u)lSQ`q~Q|{Z)}`+?zE-TZlsQvO^SL!E4x{Fk=!3IeZdm^bUc+V
z_-f_z(jgG|?F|v$j>X>-=Pkb=1pyizuUIlS7ZhFRGt*Q`%7r~1VQ$H2W~~}@wLhCt
z%+oK$QdBgh7EDi{)l(E^h7MCy<`r|v3xj1eNu#eSrz^Tog|}vw*aMhMx9;x#*wNoM
zn;E|1O>mF&|AmX|Tl$cLE=r5O&TBKcDRaH2#yD;Eh6=i(x5(lc#@iVIFB%rCATw-$
z@mM?dcG}<pR4U3s`M|2?BFCs22U%4Cc)HcTkSj_>oWC@befCfYB0MvHhnK70+GRXc
zEYPf(Iiy9-<}CQLEYXAx^?LQQudLtEWyBOc^hLk@HCewv*P33X+>p8J)TzR9Xw!H}
zyUdwe-~`PBs8r_-IXCM@SZ=w4bMa0o(IGGGNKx69S&UEH>WPsVIp=RD?Yn3wa${H)
z3hB?5@qKUDVs2Y_FU7}W%sBoq`3%C>K8~{XnYTQus?Yh!AMA`jymt8L>Vrr9wytJz
z+&+p}_5Duzik6uB21VUr5&Vx=2Co00hGXaW@8-%sR|ZR?cI2&h^~^&zeVsB@5Ab2V
z=1czz7B^R0HDXk#G$97CyL%N6!Rb?*b<|WB*NP&Kiih8)aNA#9x0mOW^Xrq0H7K0_
z6KLAhe4UX?$n}Yy3DLOrKY^yTou4Z{oyT}=o!KYBnKDw0zjEk1D?)jx=_Zlbe$A93
zA-Y0<M@jy0Z;)u6jvm0?9Of77Rm%-O69xr;GuU)0vo1TtpTzljcm<G{IrG3sqmemx
z^Vq^)xN4tw7K6&1in7GA72wC=Mk)`{WSH4=rtyeU6h$bD_z~E>K?Gqm%DV0#g>XM3
z`I*AnWumhzr!Co1T2Et=ak_vxW3@N(shfh>R=08*rKxReB-}pfjMJNU(SBM}7I-J~
zuF`W*-&7+zbJauS$5e@vWz6oYU4}?cK+I*O5OYhJVg(@6RjByDbW~hCdn*U`9Bnml
zy?AD}rzLogQj;FttFX?O&zW#TfUQ9EE2C@LgN!gFZiuCC<2Z?C-sx4V!AH}0{-T5i
z5-ny>y+E$`HuWOPJ}G6umIrLrCRpW8J5i&!#bZLb0{lDo8OvKKLTWz~YG>g*nS}Pt
zw48ysPPGu<%8K`u=gq$W;LbMsytXQ!L<@LkDQqFt9XC_Vgr;h5N_a`Mq7U#kgi(B=
zGPe+d$fv;2adk?h8Fq|Evc}{kwh>t~Y#=V${OOu&$z8A6ulST0wR|0}FV{osI8OHL
zp-WvO{Aj>=3!?-HI-&Z834Ftyyd6@0a*30zdKAWu<Rh@&oHk^VTS_*P%-^k$4{zfV
zN!hRBNzB(QAJ@@a1JuguUvn2Zj)Mj+YsK|qZ^}O6?|G-+N;Q`#zABR0L6_;aFBMoV
zw%w`<N&(i{2`TH^*JHcfLs1SRWG#DH{qud-JoPvFc%4Mkp{dBz=FQNJ+(&F?R-`yB
zhT8qbXe?|^l1Gh?JV%QFn{;l{sxx)m5Fd7ccr0n<*WLDdgD7JP5{ARjqYM!}vqk3s
z)7sW&ylon+%}6GC-Fhid4<Gs+%f9w7`S+NNo;fsAWW2iNU&ifvZ5QekFp3Us7|8%K
z%xRg4_^<`p2>)?gw5j2IV$%~3N_5t1CXq?7!-lxH7XE-g(o`VjHjxlH5j12)DfI58
zZ<5Tt=0&0ug*I^+lI;}-pN8+8U|k>gPfTxQrhf6sK=W8iBPqTI<mYxN13~NA<F-yf
zIknyT5dZ!>erj-pL(7GZDmUS$P)|~fPKkvS{~Cvj&kBR9N?TwhwYTCTHNV!e<fKgl
z*{pe+uHCFO1hN?}kztsNGb8e37%ndUkaV}yxEo&QWjc!BG73qIV1o;J>HlGCdAia*
zE{BbItF{^E<sIkcX-&ffPn9N&*WCak?@qdJDtRQbi+Sp@f5+Gs2ZPXH1G-=M$eU`w
z`3=O&4kM=@*DXUmO$#%pn}(A`1Rie=g$-soW?bMo_{dvTBb=EviOlqA^D8?%!F11)
z$&xN>GnR!Kpk7stPHBdR)Ji&GnAhn%sYgsb+UZF_pmqGV<<Wg<Q*Q!B<w9m7gh^CL
zs3w)IwQuWOu+59Bh@9&5tAQ@!_sSFpNG)a^jH}DIhTj{$Sbh%&OCO?D&73j;f+#Ow
z&{Z0-YUjchOTQ`H1A}r50x3vMut+Fh1P9IjC3LpXh*+nw4G!xZ{uSd@e+`;0dHB3-
zd5w2RG7||B`(@TJf5Es;`eD(!YEdk)Br!>`GZL9a{;p9EJ;Mx(x){+Wr?jq877QIm
zW$AArS(uU}Jp|*31HX6(kdXq=d?L{AYZBvX77(^c+fT!w$pt;?9i*Q@$V=o>?re7<
zZ0l~>eDRYLhe?L7UDgQNd>~SR1KaaRlj9%TynouRKW7NWs3rsvl-r)h?9KXYjx;Y=
zJ7=Qp0xA%AV&NruX&|8_*NWk9H(D6*A&7Y9RS|Yoj58&=*mYF;_5kH!<4$A<Lx>DH
z#uQ8xQ08z83~*KsshCXP&<?JoQ5!)XX?fCSJ78WVSwdWzvVi2N`%{lEA+vDlaQ}QH
z%P+8q(vG$*2GLvcQ5S*KDRC6_342#DvdcGa2{{F&ORH7d3jj%JNyomre(V`I^VApi
zEk)1umH%vZyJHKyf!t9T6!&uv*5UjcrOU%w+>-2QQ9M(;W%P6?Z~2@+vQb#yPtXK(
z!$GQLP}E>U;oE3ydc1M+lh=D0fh^(fb%X#l=7%xpzgq4yT)%tlkI%<|-ZbAGC7e%B
zxv%LsVQBFjOcw1e)eB<{LT|!Q#6vI)>%ZSd4@8xExtBc+x~ZvN+~~c8(}xq~RQ2sC
zb}FZSn`XM85fN>3B2~dha9fLD_Ka=f5qezTEIL|MTKiF=#mufHhIH#DqZ4lvP({bB
zARvC4;ilVqyT&s9=H-p^d&zW5XrYfz(6^4x;4kAkHVlVR0No+pNjzG~1*WC#2+!W8
zrtxUW2ZO~I%@YDPa<n`N+7Z^b-m=zf81riuRm5})XikT5z&o#e#fIW0c|>!vk`G!2
zeb!N=4t(ngIa>zrTYl%gJjW7$l}&%4=k>vGlfkl#`}_M{e<-GdD9vzC0QOdKrdy*e
ziNLKRs`zRZKR1$IT>)6_z2O?~zpily4RTa7idKnqH*u6gIQh_zhVpY~S#v5v-@Nbm
z>wD4V$=Sy)hU?qCZGl5Mos6tPuOY&eVj*hm1!>G`P*l@;JYdLGy%jasf;EABD-X}x
zl-t%sCXcIDw3eLG^SdfV|1vl3XmXl;@Kv*!;^swrZF7(q=}0uMBPDqq%e0E@s}T$;
z&W%^8xM-P61NPSnZ{<E8z?H=Jzr@VS*8~Sa0R9yzq|K@t@vsJ`OG*^LboF-$c8ig+
zzWR`h4q$$hX+`Y0qtU}1M1KqpKQ`oZ!tQfk&}51?^xa=CAq(yJ$PHGpwk#4~y|cNI
zg3J_WVg)dtiM{v!ZB_*>({W(RAqo-B-B)AoN0g<-HQKZcq!ePlR@r}n$kr5;hur53
zI+!IdVbc@&u^r4-Z97)jv_?gH#kTun2OU}^Jn4eE)bOs{sozQVU2=A#J7aD01^H|W
zoQ8Y)i7eUZY<}ldBnbF!vy%s2Q^9M~bZ#<6pqSi3PVbYZh|GA|duO)~Ft?;!)m`eH
z;zs@1{2%Wn-fG@eHgrtvtuzMeRaD?mv>v~Ii+;y)-rK~WKQ+9G<B=x}ydsn&jQ4}D
z`5Za!_;y50_R8Q*%-R;%_joFJr8dhxAR{E>{=9!xxdLSE9FHOFt>aT2-NgHNuSRe7
zgS^71%f)~_bk<i8+EHoe(qm)h+GG*B5~q{4(AqT&pp!?~8i4dj>N7qA8`B+f{7T|8
zv3nj^&JdP_qN>WkBOPiWclNnBnApeJ^tCo?ui!RWAGR4raPpvxd=<=%mbUv(WhECk
z+y7ak`2RKiF>$m0AJZQP`+o(sECQ@;wVKhtvj5ood!IrCr5w1EhUN>cP3S<Du<s$P
zV1NbwNgSvm#jdL}{`Q&PO{&h4lsDFb-so|WJk`Ft>EJsK<~&qxh|n1;+G9NAkk+YN
z5N{n1^HQxG&%dl+Q*v&@X-(kbBhEr<j&mtr5?@NfBIMRNqfzoCac%V0B0XY;$Hjy*
zej(T{Qua>G^j`~6>RDxqdGV9Med~`md`F86i@#1{TG*|Ht3wGL>@nWmWwugnY#?^!
zJ|2>%(K+g~*07&jfcnRSut}!R)xFnK3C^O_)Rd;DfeWkIu3ZN^5)&?HohsbEL8l_I
zQU%{i%3<C=7dLxDePAJXkz^oI1YM%sr?@sNB+l@s1pUNsZI;aCr*n_qPi75eY)60~
zJIA~{1e)5{WzAU^Eo087=#Qg<3YwfqCTEVvJ6dE8ehyVafyYPJcgbufI{-nT2gQ>q
zrw9Gh4{$N4wWbTnV{-v(*R+oCx1?SmGv6=|x8^ichb*Yb<UL%$Fg#N&r60PWSi%6M
z1nm98f}2)2{B=%eu^%a8+PZ2MOc3?gn{zU&mP3lhnIf>0%2lnf$`4e7-kLKui7gX8
zS*xyLGk2b(fAycztndN`FP1bZknc_Dt`*bUh+IZ7{sSLUDz6xCF*@8WdWR5?WV=1+
z=$@?gHVSWPYgjMTFPry(&NehqZTJYXZMY@P@qd-Hj6=P)8x4jyXvcsU_ZjV59rshq
zS?-;u%3nfgmVI`nF+6U7NA};4IDh8z`QF~nzmC6kWZx*~PpxJNyLfe9tiG+@ESG(G
zGjimQ7)h8Yx-P!gzGr-&N#GNX6A}htDd4ezfGH7n@S`#H^za9<G)|1a3;WMPKcDM!
zOh4zZXr<DX@2&my`)jJcZgD>x39`HFGHEnE+`nUdzs9lQ!OvKMdLK4lU&k{>;Eb5G
zte_ihgr(9zdKaoC+rQdZU#)!H?44LnM+i^e0t*HWSaMw4Q@;*B_ul&lXAS0xb^|+C
zo&4pV@Hh>_tW!E6tZ(`_{3~`US!A_&^=2uzz2*rgOesSX1|z<9PIbTU4?Yb>9p?Lc
zXV0Or(ite`2aWrO^&gKD##h^4M-Lpr+P+rnM^9I0o9mzDe!*lmr2jDp=FWDmW4oDC
z^=t+E9O^5;1igCKgxL5KH6g)xZHk@=DjTaAAmpj6TnYtChXRyhsTNnrP2*;$X0o4Y
z%of4R8kZOV%+8|p_sf?fmBy}%xp^vBg3|b7R607W(#4m8#nKK-gIPE%5M=J>wCN2(
z(?uL+lAz*y&kJ1OS!3-nOt951(9q;d!NKcpCYpr7_wSv)#9RK@1yyM6bWCJFz6sis
zpNgAHa&%j(yMoGKkFIwzFSD0s7Y$<auf>FluVga{(&j!NmmtJa{5I15G_vQ&=tKGX
z`hfiYJexGIw|_nE^7U?VyT9VfoY;D*IZBvc-Jc&lZIAe}4Hf)!bGtLSd!qVqKtOeO
zDpFJEZvvV{98D$yNE4Ak=Lo+OQhiWlb~HXER9Om+ZZ_o9SPCAJ6L}1^;@l$SGMoRq
zaL1VdODi<OC;Q22BD3_f<2DbuEdi2BhIe~dhX}i%ghD|KN2x2(139@JU@y{By*7^P
z?(p;k4*XE~zA(EzDv20GoSFDQmjxGJCCr(Qeovr!<kD0!OdFx)_xXL1B)fMj`uIRb
zI66uz(&e`HXG-C%v;KZ`%j)aN%--EQa`5Zl^y6*UTvE}t7}#H^RociRFtpOti)Ru)
z*_dZ3L5Ih}Ze)aNl}KG1YB%8~_%c7I@!y8<%2v3_;gMz|-8n=FoISE#q!UOwfoXhw
zvIP1R_gNs*V<UNiiXHjdfL2wAVnJ2f@2hqGZ)*$&2uy`%NBf10-ioWmQ|`l_M8dP%
zXT%#;Jb`V(ZRN$-4)lL~eA1?1NszF&(>`nW+b$o8K(^Grk+z!6(#Cf@LuR|yBZWl5
zDOHl#rQ}BIu@o<vy#<1dJ)1Z~WL5oG%hr08W5Z0vw>Xi=H9ZkVc&A)-_NqflI#FEF
z1x^Y(5qWO!Cg5H`B6W}vUegE&pU-WN%bo@~c|$X_2c|XcgjPQH`vN{8^T!7zDRL#<
z)*kR#p0unIFj7_XC(fK>fFR(6x@28i3l@|{E96st=;5}&DNtcTNn2$(V&cmYY6x2y
zi51G4l9x;%bd%ZG?_mHL`KItbu(Mb-$DL556(NEPDr$}x5*2xhy}6zUi(cG)lq{Bn
z?QaZnRh_Srt6W4;A-C=tMc7&o#=;(!Le>R+@|ypLv3Cj*CD^)k+qP}nwr$(CZQHhO
zthU{&jn%eoci-L@Cr;e6AO47pjC#+gnlopOF}{}~8bS1bn-reZRFiO>0y@&SbT>($
zzd8D+4#3d)P2p$`B;RTa{;~U1XZ3-#cEVA{;TE5iN|Hzg1bxVYPLvx7Ko5$;WwsK?
zu)Ha@5<LexeGYw8go7WU+SwlLTlA1&k%Jr1u*lZFy+=U(!P6~Xqx3E3=a}+KX|^@I
z*2&Avw)x9ea1^5H$Pe}lxlI7cPc;3fcA5em+su278UI;~%)9^Fnq!F@eR9!}ANyZh
z{R%cS?0?n^28w%M-poI}+fJ=dwIr<zc|?EWa{6G#Ope#(t0>547fjB7vNAqsHKr0p
zTTir^2vmVlTjL^aS~LgYNTS2nCY@ck)j-1!)S@5&fMS(Es!??Zv`p#Kkh>WpNmL~5
z(XY~k#9shXqJ5Bpgkh)9BJi23ni3HyLX>K&QJ4U=$G8^41ik{j0G`YA8jLS^?c|~&
z7wPb=X#YSw6hm#9RM!BJAtft!m_JAnbP|W{4HIgJNeR{yU^GRp1yKrNm+nR^#M>Dl
z${EI@x`()gJj?v3^%Ups!n7wr&@dsA`+;PFUC=ueU~ZWX1t6LN`~-!LC%CAj(q}sf
zs2q^K4NwTdLcvSDTm2m=aMbFmh(Tr~-W95Yhswd_*YgWGi}VQQ=%7H*xr1Oii3D+g
zW<^*E&fCW>G#Q<YJP;88!!klV(bE{%4g<wO|D+hho~IwH0X_Br;sihdgcva@M<PzA
z=mMyO;>O4*2-!iEbVfpA7~dI)-s0ckM4tM!BQglzB*e5M0$k+00isu@H$N1@#~Y>Y
z!^t3)q9FjD0maE_7DnNGnx?~~28_Yp>O@+FQd-0!-YK~o#Kb^c+2__|0K5_j-NpVM
zQv7&O=N#$4c4HtSQ4sTF*5v?EJvddXQYb<tr&o<c*1jD8*BiF8G#LhfWM1n|Kghj3
z8c9DSXPub><IWVZ6VWVeT~>Q#kw)82pm+k$_DMW<`>3|79Y29~T7E^E>yYRI*#c2T
z#RXdlW~D}zHsdZOG>#3m5};)f{TtnU?M08$>P3%{d-Y3LdvMy59_B?e(~7xi&FrK`
zKCI}))ko)F8!KsbUOqTp9u(x@;>~SFvc=n-X8l>Z&pAUj|E-UjA4fJ`{%pCp<DcG)
zS-si#?OxvO-fh5cVtZjsrLU{M&tkBdeNZY~V0pJ?Sp7S<!maQMdr4n^dG>Hgu%!u$
z`i0J;J*)*#^G|*$!RY%$)ZvGTwIFOD-iQ7Be32%^rB6)>(1hrUXt7|Yq2cMeOJ$o#
z2JFP@!i~FE(U-V2u=Ym2QQ=`=tJ^f)h0yJLTf(HG8h`VNbY0sFxFLcz5}VZ9HZEsD
zImaN>f55ZXHy!^=<IK$POXXl7us5=T;^F!KdSgaThX2_-_OF(-Ju&+)r%QbrKSHgf
z0v!RPq51AN{3B6&UU~k?5`OE)pW<o7V)Xf+V-g;j91D`D-~ItIF6S5X#M#Es-qYTc
z;=IW(rOVOlixIFd;S-qz9*g&v(uE>Z&tJ};S0Xa>D-2IkqV%Krxqel?!9(Tv_4ojL
zpcuSZKzfMWMeK7{yeZ!Ye|G&f9{Yc?KmJ4M;z9J_yL9*O{eC8B9z1Hcij)d?PHrU8
zNUu2(JYTEg(k^VUIreEAHm7Wbeb@eu*3dSBWjO?&(oWwQO}rf~%S2fb&K=OQ?in)h
z%)m7NCs#K?ou7=dUVouQvq%E4q&8*Nd}E`zlrtL|rx~ZKk!0SKm$XTDT?OCO)+LR^
zm3#aT_aS0BpRJZ~<3eHEV)Ze?Skya@IDfXd3WNNlt%xhz*WRDg4FWCNie==@aFnkv
zb~|<S$kC07?!xJG)k^tf{Cde{J9!rJ5Kh+P<^gVvzj&on15>{C#cIPXrukR990-E7
z9`$1L+;cZ0fz}P3<(z#meLmp)aX_{BW_05|)?se$8;@CQy6Ml!r7_h<WdKP0Tat$?
zM=89wd9g5C?`8do?<%{8=q4BRt4793FY9yBdPygJ(T<=)nAL;lA0e``ap8rnz$ETq
zN^o;XW7qvk7I&#8mTPri%A`kuTAev01wT1Tmmi~bT{+*sHqL-SNLkAFIYXeb94`{_
zsm$%^_%r@il%Y1e-L^E;z}F~`Nn3=NmuFpo={MhPhdk{aY}~F+T@3ZXLz5E@iH^zp
zfB8VrO9U#wAm3}(m+Sk>^W*v9m!^5qk<d%>x5yC8iuIZ&?UQa0#?E^km-U|%tictB
zK-DJ^KC*^ugK-2z5;NT`r6EZ&ZOubRGQTp*pt#WR7o9gJ-sl=J6Lhx}tE605&fR-4
z99&M-Ik)Oj<Tf}pBKKG6L14()hnOXo*M&S2nl>uQ7GaK^&J&)03@@7}sjY*K>3<+z
z<-|U|qVvyFpkyl`TuBk(A~DK7lAz;SppazkH+N)Ch6vb}BdPMEFaANyY+^0vI`yFk
zK&!YDuw4PZFfsxsZkYE<Mck!%uc-0DiqfTf=1Hp%8P7+9wk)XY@9^+Cfx_0HN7=0y
zkP69d8Q>0t25eN}P|g`e5stV{TR`OG^k@aJH<s)3`KHXD)hQrormUw1GVH-erTNbF
z>B-S0R(hJQMt>|W@$M@~{SyJU_i3bH`C){bt@%7;DBqK@!yNHgR{fB0e=^uks!XO`
z>_hE3rD>lcHUlAgCFMF@g3$bz{s6S1DfNr$OUB%RPpD-M<Y0z+;@&6412ATgpp=oS
zXZB>BPK`6jo)m~g5yMuT$Y70f4U<0z0%4Ub`Z`X;Ai-lF!H!aFcw`()A2*tce<wqJ
zNh#e2eeUD>+s6W2YoJ$5n^xdGlB4xcmZAWvX)zF&;FdnV`3uOy^OY(9topM$Y|1O8
z_TQ?vGq{gnb2jZ_h;u{RVs87L{LC!%mKJ}vJ=U$Kys#Q;wnnfkHD|~ohZ@kukZsdc
zsoLP!SN(mGDBwX@+l~bGMc%sd%vs~b^69E4P<caTZScaPf^*93Ty_wmaz<4(NkV$=
zQe|AO^r6Iv%rbmiAg*W-s|ZLJjnw)l^A~0c1hr@BN}@KU*?b@_gn;NC<DLkPv!J$%
zyIPnc!tLetQLuAJ+o;tgJ~Gqf^pUn(F!}@Df*BgEX*a$ZOdpB3)D9XGL&;K4wW`Ew
z8bQYn2U`BRyM2P>dIexR2n8{?zuc6o$2|NeslOgUTj>kxDzo|7aVD3s+vcUz=ar5)
z$z)_AnwzeKDp}!{WB;B9rY-8sqK!}1K}!CK2!&LSdE@ixmy2Tr@v`S==ib&_=k+qT
zqs43r{*9?sN`9&|dP8V-Q99y(6^?G#8UL(q{vc^61DU>P<j|xiaaAv9ZRoIIrD&>K
zf7ON#*NPzLq9=7+ow(+tn5e-zh!=56ES<m8Au}G7huW2Q%B(xpDjr<ApE3om4lWv%
zE9;U`f_iibcsPT!()4Z{RSh_*mdEhgd0tW<xw4Sg%y}8kU}tHJJ5-5<^8bk}W!vSu
z(JLt%?ZMWl@cqfv+36QbF~`z)u1sO}w+5DGhHNVa<1~;zzxr3o6I%_`Bb21ON@eLk
zR;OwKg0K(NSld25b%dU0ccTFaNsK<m=~lx;!D}z1ai($P*~tO4IoCMJ6c&fNQqMOQ
zx7;AT+Blh1x36jL{rk_t;j)+i^^e|C!$WLOh2W(3bCHne;;6q7J=8$t^T-u0#-PqF
zF1O-%M&>B)&q*%Dt(eBJytQZ`^qYSvYRg{H{c#U1B>CCmx`IK9L&;7LtZVNx@XV*`
z(GL|7Hw#V}_hef1;ky4!inV)8fN`61KJruS*#rYA7)~=XvIqFNa1*iwuHH5yFD%k(
zM7!5SIf4C*YTiG@D=?s5&~j-4O%tX4n}~UxU9yO}Kjv!=9r)~+b^hv*E@}x_h7v!U
z6Y>=^e!?8_`8Wtafam|>GyZD{XZ$buU2Loj|HJW?bUgNN3IF&{2`7pw!4qic_iCn_
z(+=)I<wcd1<7D~yQ8;I$x4gvMXm)f?7)xXIi98f{6m#F<+WK<i=*8vBOW-Q!31;5f
zScN-;r_C=gU2r-9ND)2V?sfO~TUgh?6vr%KErKE6>Ct&R0}e_@C7>2C1I**&0Lm21
z6x1}<sM88dys4Yco%~X}e#xRxLRcq&7F-#+JF9Qsf$xN<QEj-h?q2HC+kyMRv$~_T
zGHp5g{`QOTM+g$K=YpC9cH{p(>&5%7sD$gr=(;-?Yp=_CQ-V1agsG@C&BH`!(MaX8
z4jz0}XER`ng;B)%jS))IegMCic3>lh;m-@|9goc+a^YovuVIBIM`B)#hGDD?#9VE@
zPfm+H$tt>xG+Pru;;m2^RnJU@CC;U^XV4*L*AktvZ(3L|^W|yOcQH0jSuH%+O?53S
z$Jb5Z*dVtgq?@GMh!D;*&VH=}jmS`sb+PSO9dU4Q1h0UH3A5?&xOyS>eeu-yfFRUM
zwbRHv%$f6|<2Lvgooy}3#Pi%WH|>Y{+&4d4ps3|$QnynD=egt25Ulf;Gd9o2ErWv9
zeaf`C_)lqiKwczhe)PJ!8MsqTM&S$`U}|YY8tzIkt7-Vu-Q;De&NK8X=fKk!Rah~F
z>rY}z=3g1&6;D)|zYxbtaM8R}2@Tow+sC*M7E5DCo5k|#)&@C>HZrvXC!(-1&6Aw!
zv@2s59bO7e_^gEO)630CiW{)?Cse(cLRaSRtx}a6kVR}ztai%4t*Nh7vFUB!8H5L5
z;3C8W(cD%fbHS(VRwRa_g~xcU^Yo4=w%6cJXq7b_XK|T^a*eQM*#p_I)Ygs_h@kg1
zZ0h}>ZA}oN5pbV6*HJ$>*%Uo;T51T=fCNX67%#ys*ak0l+ggK(9)av}Xp-u@l8JsX
z$=n{w43nO#_jzN+`kyjT(JjF*59uJMe>L{^`gh{=21ky2aJ}Lso!yiPJvwxVn<g>x
z^30RW*+ezyHg%N$@cLp@rImR&$D*e$M}z86!JCry!5rz_fVwr-aBEj-*5a^t!aAk@
zv#g-irV<J(4_mKgRipX`B7m=MoH#%-;aAyG^cttCHey*iU8UA1ox8lj=ZoE5lCN3O
zY*k6uWvVTF7*t%;M7)#)hB|(8hVLB3Ekp|ziHPrqSRu+;Vrj$7XV34qGqMfFIXEah
zhgK>8!T>L<O;R3kH((JBV0Gy`0!A>~Nj!1#dBg|Zgak;f^zL~?Rt}@!svH%)*q@yb
zlL8B<3xj}bkWx3R@cljm6KZtaSipg5j-5;6*nz4M%T5&$huRRN3V0<CsCEz>1i(!;
z6SdC`WIbXDs!_fClME(SJ9Gpm*7_cYCMJ941Yg{?btM~<b@QbgW1ID88?(OSer9}G
zp<-!bqf+I<bbf`Yg}JX#FP&*=Jhs(vSZq)RUQ2CY4yyq(`DPfRUKxDvXqDDCw*A32
z_W`Nv8Zk=<VExvvW=|}3G`x78>U9oXWqV#5oVX;@>d+2^5|B)A09L?4S^YIDNHO3)
z<n1<Z74_@|LIlZSWy0)*8TgNqnoqUX58ecZCOj={5z_=^JBdxlI9?T87w*H<MP|ka
zTHrZAvC;{QfRwCMuUdTvAOW~W#pmof2M<?Ms4?xzm9tz}!J<Gf7e>EHg6;V!>&ZfK
z!Vyke!qY0pIEvPb!yT?3q+5qR6~%em-c>VZ1cyy2bMX@)%y(ISjn}>b4O77cOHQGQ
zW4FjBgoJQ8!{c3N?>$T~Mj3e&Jjlxlg41lM{GGISj4^V|M_^(s1EsojUzRq6@Z}Kv
z5BjYY_4!LpB@Z%Ps_Hb<oo#~N53YVn#0h|lo482@#r3~OzzdYX578{Xel^68a2|a?
zBT+)0k^^9*UKx1n*8ss0zrwZF#^;{CwS!!!WObXezsT~cL!^+3Sp;uK8GoSoQRNFR
zXDNxk6o7JQt8@_@sz3`;lS9Y{q?ET%3?(yO<UMTYx##`pS0XxE9BQUp8B<o7t@r?^
zldNt+Hsk<DsSm7%Zc$LT4ARymzSKSj-%C~S{LUdL2cCtV0v>=oScDm#smpa+WI7zH
zYe~&~2<Nj-jid$GJ9U|q9<WMOlibBu*L^Z@^T^D8W+4rE3efoFAHxpHX~Bk>rZqZA
z-5qT6=+rsGzrbg=%%1*h0b}|vnji-D|5*Xs`o%^5hZQ~hi;L8#l~ANNKrl$!qMXx4
zcc%8DTF>*c`1vVZT5LKbiA!Fy^mf2&l1W6t$ETb+i<gw+=fT~{<Ljf!s&T-gk+biU
z?`J3R4>}YqCWj`72|Bx)Jv%X{oWFHFj4n}$qMz!+^}iS*2aPiam<7Q5zfsXVMZecj
zoTX3NH=%Fahkw?7+qVyLNT5(M*d9c;J{-7yo*#FTW?{bxl%FvAUw!rV^8UO}m)6yw
zlH6bKC(3^Wp+cO-+a)p1%KH4X{rul6$+gjwKV{zMoZdg;E3x0wSWii<6|jn*gn|h)
zYgsT4LZRA|z|NvmvPj9&a^!p(TOgn*`TMvuN+ptBgy2zvM_r5h2jw2bO*|O~Y1FN>
zAdhtAVUL2+xi$iBYsT%`fYQQVl!s^(Kd+KZ-ragLvH<vNKpG~|@8y!h1j)PE)_e>X
z6!(<Fci3*zPE7NLGE6p(_c+-3vfZV6x9P&&g_7_OUz@1HlPFSDzHB_ZJ#|TA=e(Bi
zUGrRkMdt=Ez_)wrq=TdyHlCV`S12ucE#KWHB0`D<nJib|2<Qd&`@wG-c#bZfZYRbu
zilub8wq{b5ftJ*~RcD%3VV|R7VZpMaweHvW^xZ9^6`&8oQU$KtfQ9J0nTkz8ErqA`
zQZpbz?}G6!Zz-Og1Dzr?80B>|b}06)3c^i*Gr?=i-iYrTFf5DBBG1?F__DCTP2=1R
zf8r+6TBLqfpVKC{vslnx^Nr64GY-8tWR)F_IjE(RWl40~evB8gE}Ji&Lflz5D2c~A
z#m%}%M;N(oc7&f1V0+&QLM;Nao3yqY6Vq|zAEo`MHiXfi><`qOag2Wek=9xEc$l#<
z=OmXCBW>Uv-we#Ize`Jv{)A9lS8Ed9X`2yOm-)S{uRFgIu-56uaCT<S{K$3|T~4H4
z>+LWCxUqm|cYcfEFC8O=)XF(vhFKetTNro%%bNM3L&%}{txX>6DYHyO`=pV#ShHZr
ze3N#eTaEQ+3Nz`;p<clj?^FrHHx)eM^!n0IshJnMW~LXK=tiFcIVgNhj!@duQblmU
zBR-3Unv)xLzwlZ?#gZuSGK|wf_5xRdUxbYL6*3b*0+lrNqB}1G&>21B#43j_9e#W~
z<iL60;&p;bN5M}!AFc_oeu>}X(D(mxEF_AEM<Yt6!%4V|L?l=IG>V9uV!}2j-?Ndn
z%1Fw$tdfLL=p6bX2bGrts46Lp%@Rg{1bcjJX;?|<oz{f*<vPsDZSYoEQT3(Uq6opW
z&La!$OXa)6RBTZuA$mQ0Yb^Ddzn(xKvW`H`&+KE`%sZadG_Wd0bDh?#REmQ^B%@dW
zsU)m;GNDo=ibbOlp=ho_>>r=<r<~=Sx3KbgSzQ%<+mu>D<1wzVWbKNbefyO0lN9*M
zdri9^lJajKk_pu*EP?*+D>Ap2Ic(nJbIRUQUCtN?6Gln4PwN)Mnda+)>Xd^Ol}adG
zo9G>MHq7@zgV3;L=PiQMJmIHC*mg#_fccny;Vd4&XeK|g(jUtZx3EIfZyRC4KN-|v
z=GKEoFNFWo<fHcoomO#eOT8hraq}d98W`4V)7kj2d3&I%&V={4F8zj^(+LHdT@$p&
zz7w2~GPi{`{@A-ewiCKXw055O6<?xH6fft%>c$^VIOnMCfKE`?oyUfQ?~msd%|g8W
z+ugx-@V@?VS8?=5N2G}lIHTO2&!=k@6~B}5$I3B{VYldgs^NYN4H%Gnmkrn&B-99G
zfB=61n6dG?d{o=Bp#Z%Fy{jaV)F^{sYOT1JO|U8S2r35Upoa=!S_im5lfXY=I>6OR
zJ|!5eAM4N(zr&q6wkxG}b!PwY^61U;NEdaxBD>+mcGipGb~?6z;oYdjHxsv$d-IsR
z`Gp%tDnJ0^yWAO6XXmcNXb4?d(jwbl-V<UL{>|{yXA-alFb1oDyle$<1*0!Z8#aeV
znu?sM^3#1?uzn#r6|(#XTUnk|(+5QD4)Cyv)&yQNa-}eBcG}Bc{lH#u*35OUTi++G
zCQec~P2B7y-WYb2-Ypu9H}K8StaG&41;82X*1U6|3$Qe}->}f(4mkK$;u`$?;SR|5
zq`dMT^`v7bUqd-xjC?eG0WhGU%HQF%6xM@#a=P~y-WKfKd}h8C6$}h`v$2K|2&3y|
z%5Uq(D9BVMWGE~@ddKF_PPDST3qBYT#W@V$6{CAT&z=`%J9$8e0C3BFMq#ffmszf3
z!R7GAE~k-qJ<sAQxTlQI{kZ2*z;wmg`7}(5_vBn}b{ERTx9!ZWG_g&2&P{!y4L7)q
z{s23ei#3;8`B#0evs-$#x%8pVE#S;$_u8$~kgj?1j^q<;ZH;LZ%lv7}^C96>(aL?5
zjTg7yMV-s}7Q8Sk%Y0gT?K~*RmLHSn!A+2l<(C;R*XdD^il}DXE-^zJY#6?KWL^<(
z`*WKRka^8jF-5n2Cgfk2Lzf|#Z2CQ7RISi9;ul}4AuYNJgFL(Da9rG&-RO+bp=>^@
z2Q{~0(d9j$>*2bH=wmPiCINYauPOeF($lhcwBEJOW>Xh4!aeEUYWt6r_7NqVZx|NE
zxpiGJ<st&Y#{+)osTqH{O%TDP?OrLstj~UAU>b_Py%<VCN}@Df?T-Iu9ss~t<U?Xd
zOUY*pPquv_6;GD^Lw}?Ath96UcRP<lopWE7<E3}0bUP<ww-0`;vrG=U;*G_wezT5~
zJn$5IJ5G6qSIfQC`>##LFR14KKj>!SU}gUw-IBMor{XTykhWgbPr>QvqiK2r!T>k6
za0$8~z1TCB4U$yRl7IAN?kM?4w9}T^Cuo-3q*W514lAj$<ojfO-p`Ki-|oMJ_f0ZX
z4Tc}zE<hdw7WH>sl&dEI7-$XOKYia%s856zT??T()060gcW8a97TF=w!~kLeI-woj
z%@9uGO1P6oAw=_Y=md9UJ6{sFS08(iBL0)MnMpCR*vIGZw`V!<OQr66kN@s_nN9!c
zeVkfNVjdWIyxGzJ1Wx&lxJ2eM6O{A+G=G0O!slIPLNEU%DUX!%`(<}>_mKc{$pQTb
zObvm+A&E*}i<|tdh(gfoSQlYPGI}S4&PZKgc>fU|DjGE|iE34F`iD^UHQHGzwkTSk
zSTxM9_D$ZPutS&z<vi><;JOlen{pR1Dvgt|27SpdYK0*sSmsSzxG}~piZf-%A_+dp
zl9EZdG(Ra5Aqtc69K+3-BE)#+AFwG|iK)e3fkgUAnF6yJShd}w!&=n==X<6gO2W95
z^!aHFXK_8#7SK`3j3=~2PDQ7zpAgS%?p9K>8?`jaq}AQ%a{Dadwl-#6HYs|as>6@<
zc%-`?1k^zY`)sm_jSY%J#q*r&_wwI~WCPDhq@M@gju`X9i-18F_Hv1tgqxRviZ!Lw
z_~xui^CypMshh|jl=$^bfVAT9RG63xaqz9Rl11ys7=oJ`^~4q0YHy6OkPj+PgxPma
zPv!v<p6|Oq%9>+hvHeamC~`~(Ve^KptAD)gqx~+n8c`@ol+98Y=;}nF){9N~Gs<M^
z4$zT)kejVz-S@XH2c@~+Uf@m98RHP9o!BW)f~uO&?$$D*i92oFnghjKk?jQIy{{>C
z&l(23J){*Q7JT%gxj<!F?40o2G5LX*ft2Ix_<K#2GvZZRVX{SF2qX;7h^W9>JXbb!
z(iPjbgl=<MNK;5z#LPs}Q#nuJ-2Ex{j1#gSUr{y*+Pn6KxWCoh;pn`A|FX@{Qe%`7
zkBaooU6#iKX=h~;p7SIAqY`x->i}*K$lK=(jvW*3Not|p$SYd)=uQ*vU4dyi-B|}M
zOSA;$B^|K0W8JcVZ#w)qaOP%W;J%Mz7j&Zc{(j{f<^(neladeMj)Pey7%}f>fVhF~
z3ED}2kphHyQnZ>F_siA_`IR7Z2+>kt9PI(mQojlSidGhDkWq<saKBR*vocM-#2avy
zVq!1nBn1U~#Qn;uryyf-`<t3$0OkW{$U6*@rn+WcJZ4Jfm|~q%W<9R`szCh5R3ss#
z6eqCYH_5f5^fRbjnYScUJLmn^R4Px--Kv(HRm-JLd@YeU$UtE(6sRqgWU4ynRq<SU
zvwt>cI1><pTZ!G`HcesjeQQi$wm*V;4W<c|>Y8KLCe$)moAS}Oz0S6$GeA}k!~Q8p
zuIM)rc?0DZb+j|zgAyhof_g8xK1Tf9gdSFajIgc=JLMm7JqZ(V7NR0uCaPw}89`2+
zln`x1xY=i%nQ4YjG`?@deh1yAUvzHCQ18(bcj>)nB$+ppRJ$VE2<x%zt!@$0i^u1s
zwri7S+Bs+{;y&ZaPnO6}rr=d%AC*!FM}8RuaF;!4;{GvQfBBjlH+MHuc-6{`j@hJx
za1E!M04xcv1$y)(I9CkvYad^|9%QF-(zUQMD=l-0<n1*$?d8wFX<GYG=7mxonJ_<U
zCPoj|r{0K&?M*4KD-hoONs`K0)!rJZ?J6(m9+`n<*1zb=>n()yHreNV&|W=C0<F1D
zl46wQ=qT=E+^oIHLa1#l&AM|ccvV!(-!S2cT%S=L;vl-<-e!c1H6aOy%>F39&8j7c
zhXI_SRfcl6s1Ms{ooz}t7Iqa&aop~u$~>)CB|QslF#$uVsVS4c`Xiy&_*ileCnCwV
z2bS$XUa42Db=z?Y)sz}RK%J1TQhcAy?nmRr7M3N;A8G$bJD5^0dg>Gvvc@wh7%|u&
z#hV?2>PSn`3=w!mPwz+v(i-#&g(l-_DTSMBLZ^&cn)o9kGROIbcaRyWD1y2(#(Esn
zuu?C3-7AaL;C_UqPo{b`csm^Hy@!Tb5r$3mV2{;4bIcrD{)WbD3Wxs1NKrDMpm3}x
zWx>LQA0%3843j76X$CpO{-OI#t_PabV_nus+cMk1UZ@1tklI8wR?Ydy;9B*npA-lN
zdG5iXTqL*A>_cgN>^L{L0NHjU>7Hh8?&QGl$oqV<vTRFHx0>2qu&vlVrH$$GhJx3b
zfpodsTW)DF^*nB0E0I;L(EI7KmU@ipiLLc;`q;De5XyD-UMi7QwGe)vQ?1he!XB@N
zBBZ8~=QXBnum9fw-=dPPrsZ9x+avSuqn30{Grvpsjf&iJXxs^E@L?Ke9sF*`@M78m
zvZtP7TK4+J5z7|Vjve=3MlVUZ<_VPO#hQwvZ?&E%_@?6da)FLR3o6pcbAW(=EJ$v4
zY?}#i{41jxqOchE5+kJwG^Wq|b~X|TrY7QxSlkhM8(SCVElXRM;-9A4*u3hE|7>-m
z;OCV|#(ik6)rbFJ>*NA+tM$fWhVwaZj^JZ~_$%C6&s5SC(gw#Bnf;RC-hGnQ3lFrO
z%$QK~k-6tj3gFIVe77d<yL?yl%!eICe(mq5Mn5#pnA?YE54NRMyZ4RNI*?FHs#>Kd
zQR#}D$nig<6CqOB*+&|Z19VFOgULk8I3X8tiIiCQEHc=!S?pOl7yTrabG*r^bOlub
zxUOh$Decya^&>xO#5KgAl}h#!KOY1a{^5JGieZCRZEf+x!jL+5IGb$`c0zXw)~9I7
ztSGb26J!||r6{6x&WNUJuDTeS6cs#F!$|%K28eJ%$=W$ZDXCb%yyT3ff!$4yRg?c#
zA4?$b>;>Hg=~tUXVbjQQim&><L*<k?ZRNbOR?A>KB9;jWJ(>djaKkmX+>(U{FT|dy
zAseVX9b?kN@90=8p$qdJFV2+LTHeCEZm{x|xF%P@BK?mjwXdMV-RSed&(Ki~L3nRr
z!PlLhDFMP>&F>)N8gd%GsIqc@ftz);?H@g%Nz(^Cn;&WXx$Wvq?nQz9I`sSl7ckaV
z8oS|9pIfC<YMI0;#@HN)l7np6#C)aBy!jfn=sU*j<x*tWnJ3&;LIOqSKf}nBe>O+7
zqHeeG8o2iE70Gc;u_wo?ufnH$!b9}|STv;Iv#C}d7t7LVYR}$CN>pgJm->(z{$}PJ
z(M4eG@1EO%qh<9v7cSSqo#d-VZZ7A@vo2S_cXU_pP~?grqni&-J~-pICOYI%5-+pV
z&q#WmCqy{~QN=sZcHd^8qeB5UryO`6rMIm@bRx0V_@O&ZkasE(c~e@Ww5BP|vhXFw
z=0N39jKmU^@As~n*(H0mCSuX^B_4HDILm%)`53v36Hl(y8l~Fq)DQ0W_WawuJ9v@O
zmt9_XuI3|C3R7SMpDHFYt>LR_om5kZ5C$84^r=h15s(0zMARygPprO0-(FRMPv}Dl
z0TN|jH1_szZ4ptyT{mDP)aO4M=HyUT*X(NVF;|=`KjIfb`75)rS6P>JZ1wG#G3|CZ
zdDe35=<Q*P?6le&QSu&&X7X}c|D7lq==-3(uF&#{_}891BZ&B-2zMZs<DoE4k;bJB
zxt1+vUp{7iK;H?0xvgs>a>eVQvSXpqL0wH+oK6~-O_i`OfIV`YCd;;4(go&~$ll#p
z6}N(@{&W5I8Lx6qdNv1w#GowYBPfLuzft4T+Xhh2*}ar5f?^{N_mZt{Ok@|MbUc~R
zyw=OBgY}_JR2f^2=s6$d;~N|-23qdF_TB9Nn=<?VrOPm}GXBppLYA>7;;7lLxAph%
zEkvO!K`k&W)cL_PHb8u_r!5nvDZ?jz@Wa;Isk$!f`~={RM4=@Wbx6_pzg~vr`+9Zw
z^!+-hI55o=F+ATt4@fG47t42-D(*}|474H_Px8zCdmQOldC|QBoJ&s`A2$}*^Yl|2
zScK*v1Z#zI^kE^a6qkxDVnp7LA0^&xl{Q%auCnYn0l)_uK&KK6yzS)wc<e>)CUV&Z
zAi`#H$-^SJ8qC%caI^*-;6U<0z02Ot#cP~ukFqU!me>QTf2&FW@+Vu=@;HJ$YJ{FY
zqARkrf3M7)f;g|su|!4T)<t6dWLJP_aT*m0hWaa+oNs^qutbfOPy6W~CAY*CW?GDQ
zk*5!qlMbjvQf2RwZ-e@IK*?qbZvEkz&eg9|tC`4ViW-=V+dfK5Z=n!AHOj^h@7O7M
z`*&w9eJ*GnrW2zgpyj(OmqxZIDSsZ@Vka?A(xb&Bhcl_+f`e;t8gs?~2W^@u#A~$m
z*HFScuZ2R|XoYGg{+VVQtL&i@60ZiI>cKu8FI%8BsIBE;M)RAQ{`#^|DUOny?3(Ry
zFur-P-H$$0V}}@^EJUl3!S9i*o|NHWlSA8wM6$=^s_ds$&HJsBWE|TqlLT4pPiCU%
zzHKYcHk#CldiE1%j8vOE*Max;14r@^LJ`gjtsq}rxMT@SoF(5K$3L#w#CQ>V>-GmJ
z)AgRg3Ew8bG)P9nX&k}}w+!CJ@e&4PvaR<#!Q?bC&^&GYgg9WidqmWQn=(IZ(&B->
zk{>$hpzAC3G?0bcIpou*s<oiS2^ZSjNja%fl1w7)45sS->V-PT$uoUnILiI3=;3YO
zHv*7Yavigw_iZ|s{UddJp9(QVqTVWqv%3e`QC<qsyu}(&nIVn&nZMU57U*P^W$NOk
zz7VK?ESiNWUI}72Wdv|DDX@62M|LH82(Hz}ms`aNBd2zVhsIff?Z^fjh~R4))}--H
z2%9~mux6K)8OLR(@Qn49q6+9&u>`gZBT?dX)<iH!__0}M!FP{s6P6H?z;gquHXB<P
z%Q)oNhsoUWpBd=aL87wXwkt|H4EsVmBZ$_PscPO_v(G>&5ZpsXA+u$@l>Zi|WV2ap
zbT~^+z@{{CjF4{QNu=1vO0LL4<dks7A=Q4Uyw^I>FF58nXpO`KCLZ?;o+UR3VXgou
z?^(bX_a58jiA<}=cAXeg8pFJAAg?!6BX(P0^IS9-laMK6`Kp-NM(((DEr;IyMqZDY
z_D8m5MiSCp#H{|sfFWTT*m{RLI%S{Ivg-(K2|Bz+V&|gpQa5R#I$j!SakOpmpe8?F
z-Ds?hryYlYRRD(PzsK8(Gi)vTClP}H^pk?Aj)&4KWcR0m#nFm|Kz1;(7^^#FdwOf)
zbHeERdDyJ?1};8ju4MC=EncFl>7n~CJ!r~&fM<w)`*i!|^h?9D3s^u~))_Cifc1en
zrpeM%Q)ZKWkkJYdo3e0jV0$tO_EbzrY=$p4GO@0zRgSL&*{_3f0EE;0QqEiWtj`bk
z`|H)`ZKs7tz&`8tF`r-G2b^L6g%f;I4a3U${o-X2pZ*7DfJ}KBvz+q|*z=KAcgp^(
zX4X*)@)sAO9ZB#atO;BM$H5%Vzp$-yAveVc4zH6>2qalG2)RF_jJ1DJ=@)Y3LVHDt
z-61N-#b}V^^R+n#SyqcHe#4RMokB;6>ufeiGW^b&3q+Qhqvn{dUqYL%r#;6Yn+O8L
zl)wsn`RNhDcCMKx*%A<De;Dg9jw6%N<}vhZ_K|V9g0pBV!kp#yO0$)d5S<JLJ&3oU
zhuC8wp6W_~5k!a1ofU`F_=4y@)6&{D(Jhexy>ib9?zPejS|R;_+o9UYVgbb30kk#B
zJ*m&<T>c~D<I?qp0YAQD=?ZjnV=@8oCSK-5toMOL=Z{vp1y;vb_lATgyDhXS%p3})
zFC@8d;{biaN3L;K|CaBaE<r$w+-7sQ)tr%mejKucI!_CpMY-52TZT?A*yg;!TlT1R
z+~=O@q}JjB*p&5aq(P(vBv+Hip`q4$pYF^Ei~l;yDpOEBLljoq5g)=h8aGSAi#3cf
zi%a;7p%CL7vuPaw7hwb*b_(y{xqee%Ji3H%9C|S2XQ{JTJr=)l)4p7MvrYN5KwgR;
z6UHq;NR^Ml#}UB28prBzY{5Y?@$=1ri;l_bae1wtk33xg;TH_LXQDWiLqZvOCUI%E
zP>^MNg5gc!gzC-4qtE47Y(CIo9u}F(n)r7{lu`!aK}zJGFMIDH-f9)_K%;lVdB0`V
z+adeVq)Xu*G`3|gDa#m&vHHA;_|Y3|L~eOP7Bh?Gyzcp)7t?__O>yTO!wDgU8rO1!
zE6~kjH`pUJAhD*_n@Q`z{tAgr%>uY<27p3k`neXZHD(uJNi0ToU!IIqYLQ-hI`@#J
zFe6ILsi{yMF7&7%MXKv0+tf+6XM~ven@KN4vM)&MX+2unJ=Dfa<SnmP+8lBx5r|8G
zhu{N)aPv#QjmqoB&k0V%@UaQmqZIG)p$t!3NvKk*f5~gZO%Bji%3bqTZUw8v&34Q7
z8v?9M5XoPivajJ8C(k1;6Q*~Y{o2~*J1Ie7%}Q)A&rMc9x#r|F#mb4nYv&)~eiP>N
ztY~@xP><utyT#<4qm?(c=$sqcHRn-v;^HkRFlXmMwWO+uES&YGlJso;#P*_+9IVBX
zFYW!S3^|#$t02!Rz|*e`i*Vyn<tQ4C*==!|)|;{C=f>;BoT0p#;^OXi9Z<U*707!J
zViNMlECWwnDKN>9cjs$^qO!mP>q65Jl03bnK{^UQ4TU{}@VFmEIpqg@7&XWm3DHJc
za|K$R;wJQ&O(!rrO6|DwUKmQl$IePz%HJupoyCMi49iJWOuI!zW{4G)H_#H^mw5GF
z_tmx69x8mI!ak15JMnx?YF3GX=Lc`l3^u03H^Ivm|8oOluLr=!g<?heraF(jnb+1~
zt)y)UWdkWVPV>fcKw9|+BXh|nZnv5g2u@>U14U@HlVjG2h4hwB?#}?W#ImwOc4;LV
z-|El+-`$Z2dS(Hfe1#L}a#}gPXRTYaO~oVJnCH$Y@fis763g8wc~xJNKw{{brA1Oh
z_kC%aII2xqWebf@EmYPyvao!n{!N`!uKkCYh@UQrQSQ0l{IhFcad)!l5gdLsEq|Yo
zLQ6qVQzgjMO>+UB22hXDy61d=F{)Bf(*YY-Y<Eg1nME%Ri{bXeD9U-RJ)DPW<7BD(
z8dvmAN|~LLV0zZ@&?Td(-nVYCA9LS;TY?ias8>ps?H$oz$oUkQDW5?-`ycD{ue0p?
z=O5thBjcR^+8}cLpA8~2Bj^9joi~Cr6^Gq%$EQEf&%m<tixnT2$oq|uL+=IUr1@W1
z@p1{ZMoBj9tt2lT;HD*&Mh%*k?9#o(mq_`1JUM+lzHUlh&f{1Ir}ukXq+8JNAJA|?
zYcg_>W+@qcKRJKzeH|^2+}%=6btpYJK5q~9*?1|B00aO+1C@vq@p8}@3JgU6V%TnH
z56Ra{hfCNK?Tnj_6-3}&G$_W*lRp2q3t;gePG~m282hiqi}Y}QJg@s&2pKC@->)}>
zzBws|qazbe=Y}2rA4XsA*G*$uYsZYFA7-EThmWP$&oCGm9E|Z=)-uC_-F{VGP3(*)
zeN(XW_#Eti4Pmht%-sgfN-Wu*x5pz0#m5dPW`Xl2Hemm%0!`mpjg7e$B!5Wnvk?jU
z4BFBlslBI@*j-R9$}?Sy!s*1ld;NTGGSr~;AQ|=e10=lqH1tgzql_rgc_xU^J8W-r
zgE6c*Ad&B#gMRgyIHc@y>;X|jf{;cRZ7McVA#oe7;!z^ej7d3d*Y{{o8&^d0M%ze{
zXzvh-7L0b0BJI6n2Yj=ew6k~I-mJPxU}!lzlIbd0lBD4-kR{{<v}6m+(M2`GOg%7s
z#T>uz#^63FDzlum4tUamoGSf(J=77okRB7xnUqE1M(Yv$JV{?$0@GBo#f2Ch1$|T)
zZH2VBNE%ozbi^X)&C-4?EQHO_{eH}H;JC08ivwV4tlitd(ee?G_yz0{m0R9X)>8I3
z&0??95)q2b#BehL*w3rCc_&BwRwWLy&^5{0lpsrl);w)&y=n=;LXG$WD@;sKW(1t4
z!bHX-NIPST$1#wA{4x+&h+T;3#~n(zArgSLm4U7WMdm<Kc>M3LuMPd)Z#oQ_SNR;k
zL{y6SfZ~tH_$W}02!^kI^C^_^v@8&4gv&EYZ?cY)UKtu-{c>>FV=;<@>Un%e7KW?V
zl;uN|*$xN^Xcb`*;bD&WeJ}p=0U=dQB1GNs2zf;4nb>%Se6x|U*Q$KTUtF)V*DD*Z
z58WpW)VFBYLBI#FTTl)JIHCK95>^sh6`_xS8wj|cv1!ZpDGb!gs2H8(ERM+foq$7Y
zlO|Rrn~ZQkG{c+%VJ)`4p+NiUG?I$B8IXqN-)=WWD$pj@X1DLpC-RvOxMn`p8_d~a
zfYq|tT)fzJ0zk@F`M}P*AsO%i$SjnHBh5^5mh-av)w?g!N~~HQr6xGagj>%liD_@&
z9QF@S_Fw`+W{0E=m~&c%5sd1L)*=QoQqU4!0v1LwBIF`iuCbO`-a=BUJt!?A$bA%5
z%Bm3c{xB1H^x{5L{N|E@fJ3RCg1st@z6Y~|Eva#oP_{3rFt_Y(sH>HnJ8XH&dcM7=
zz}?Se1#(TuIMIsJV4T0!-O7)BdmS|jk=V5t5aOBjMIL<8QtIQ}B=F-?XCyL_@YXpi
zKrActZTvGND;W@kFwjPXOu5oHXA?Au#JriNd>0daWoR(E>&8cmZcA)175FA}q7^7j
zc){RMrW?c*l*A*KC+8Yi7V+R|uP;pCq$D7e%|6r#(uqQ;LBZ82F>u-+Qp3>(K@G@!
zijX(jqS$|_QL{lxtG=9kel%(YmLju3;7^7PsttGwDF;$X7tA-OadE41%#WkxDnWTc
zb$)EwfR9P5vtK!&BHOCPqP#~5Ax^8rkW@rK+N*6$7-;g1U9i^F*_b2`eIXa`bdcJA
zQI?ykhF;-!+s*w3D6@l3oKR)$nvEmiSHR7k8c8neeryxDHrEH@ERWLDV9110X2fP4
zNqjkTKgo!|jey9-x#0v7c$%SvTH4e52Q(Yhl8WpHc_lDOHn40x<$^jjAh0Vp9Tb#D
zK7Wlf`f2`H;W{g^9?ew26-i=k!k&@cSLyB{xi?a+{QVQ%g$IPwTwoa5r588`75&t1
z`}O2nms59LEG?vTK_GKca6lV4+NpLQvb;vhT}*8s-KplC13~Ug$u=hTN|by^&FqTD
zS$$}LT}aeyMC07{$IlNhI*)DoLCKJ<{32dmP!GzjVyHz;Y&iiw1n;#j2ujK$+EKRC
zX4*g(K38s96^(IP6S1-9`MX*)9hIz{WE_nw-`?CQ+A?^8_7+jbj?s>>Bh=5HOgQK2
zi!js^hMX(Lx=Yloy}z$cWQ?^0-H5y~X{;zhmm#`nBVIxm*S&yKsf|h8jjyQ9#eG%|
zD5Qkzax-#2iLy>?49~dhl)flIv}*CHI)lF+G9h8NT<TGSCzj{2ty0_cpsg~PZ?~;-
z=#S&ZQvDB}9>)&bRdKXkHkgm5Si=SR-vNO&hmy+QuVqG*-P=K?R#1hnNuou|Tq~rs
zRKt)2iE_I_4$gK<+Km-wpw1|bg$|^Vsls$9C=36VT?&`Xb(7*bhv!#kl(vMaZoFi4
zo#r7_3?m&1^qzi8zD_VB1L;T*dwsj{bb14KW-ZBqA7vZ+Cd$~U)EjXotlI!c>%b)@
z7y879Z@xNh2GGa!$*^DRSg;f&d-?E{Q1v#Ixd-hnuNmr=e6+)|Iwo4nxw=2NdgzQf
zx~h_A86n*n#ve~5Yc>~J7ae*u7Q;2`79kd*%J-R7&}nC|YP?e-?wE946Y$Te;xpf2
zC6+Dbi*x83Zwj|0Ov)FfOB&}HUFo~t(WVn6I#*j`O#T85DJM(dgTfSjsn}t|%n_QX
z5^uFDYw}4CvJ|^ZYt;bijDcf9ymM7cF~2T_qSDm<CkaI-6F&4TOLHjNfhLC9(WnB{
zy>WHwH##`+CDZ7{lQQV|^s0?B742A`(Byh`hsT8{DPYb5i=KayDdR5Y61N^>8Dg`;
z$*$f)K51Yx$8*nuG!rfv^ywyCyTDRRxTCR2GJW6NF7Qryb_vh(s+U&*bwBnm1f>Yv
zFM*bW9}nq9%f2p(I9v%`FBbU2E2Da(*VY8nvXQtmZN3f~0P1(0{I45!eWM5auH2WM
zg+NL6Xk1?EFp`i$+Na(my4q+#SB!O%orsE1??*d?>obr)a#a%A+)XrG3c73X`kbrV
zZLg(vE=p{@NcvaZgBQ;j?y7*IQowRch5~H>mKAkw`&V+8)dkw6E$9xTQXgEeXALar
zVqQMMJ|YK5SAU1~MyQeMJ)(%QTdE@^JqBW$zN(JCW!t`pE29S^dbWtjLNV~VOyr#5
zBYd>@OFd(&AkQG+sCOF>+L=7G$M5L{hUt38)<Z&L4EtYRxhaURl)r0L&XAgh<&!1^
z*@vLUr2D<=IB7nf8+cY#ll;Ye5o^T>BtfOgtGzj3XjZXiiB#R%aZ>yW5_`>jY^U&Q
zwYPAxqogs6ib>$tH{%i*&URHgPC1=(BCgWvQMfI(l!?^jl)TD@eV~69Mxq?B7A!B(
zl4mZLeqcK01BDez+kv|*zDACRBpff%&M4xoOD%|Z)C^nzd-Ij3xs|zwM%GQL$cRw<
zX>ZG;x1tFmWP<%KD<V=M&Qyo$TOP_-UF7a7Y)W^`_=wqchD0?VT<3*f*PoHqGGtQC
zI!Ao;S1?YsMi*WK0<V+eQG96hfr>6Wl5Iwklfb4<##&bH+75MI0L^|@T%?k~s}3?2
zJhGc(k0nuP)`9?nRc)TtN4>I%eCe7|G5r=p%;yJPe3YmzgwE7xWCuc-tN5^O@bmge
zaeELt9x>xo4enIrCu)5Y&*-Mtd+Cu9z#+Kk+C<CJ#-^6>5QkKF)xMQrSj?F~n;1bU
z5qxX&TPpI+lo_#uv62~{k~wX|TGF+njR6P$yq$8d((78p3ZVTge5(-5q(al@%4srC
z1PDv$m+C0a^VVC*ml{{~ya-eb*G8N9(oIO+RJF<L9F<np_sm;BC(D?Bt3-ZxD?_nY
zbsT%bVe&YY?*oIXYekfI__}s4hAFFRv(}l-*&J^@9;Mbnhi4g4+o{aR*>iUh5wA0F
zvrnlac}++{Qeuva@N=^#b(O!_RPUfGAP!w^pk@tUxRH9=>_I?x>d{||iEV-ZJvaMx
zIX2Gl?f^*jG(7WsU{%Sf=$(SLd)9SrH_P9vdko*%18uK0vGY}ZyY}Z0vN1fbx=v~j
z4K^w2frEK}lKSfjIt=~czMrLb!LYY{AFjKCThzM4wqC}41jOC=eQo2e{JvtrMBZZs
z9OcV5v?LmgRu`-2x3S&lz<)%zp20JTe|UtW9dZ`aF3YB{C^tlS?Wc*P;*w=T=01&N
zw>E=TiLf6Bca(w2HLcMlLnS&fnN3RQzV^_uuC@n6zTnwA`C0z!O@{OTgy%4EFfjd(
zBqlxDlJV5+aJ?_TI#!P;G%Mg3Fq?8*f_9i~TZVGMBxShdAAby~A`+?Ae|6XC!R=I$
zM3hleV@*6?o=N5U+jToTex2AHo`^N~?~h;c$%EKPJE_CbGkM)ggP5-G->YXkzjnu=
zRaE_7E7$A&EAMW(?~j>EKO&SkMR(QG2Rl4eJw~^YLrmK%<z4ArSGl!&){TQ-0zf3t
zf?60w$IFk;&yy_DkdjDxZ(01?>ex&BPtV^=ZsO3i_s5;S?FCo{sf+@71W|?K$CsOO
z{%ybPN~a@`(8<IZO>T-Yue};hPV`X8_9!O+OfVI=&D+o*KqyN^3jx4dsvrmfq^qkN
ze)+{z|E~qph>CZFW7VbEGuFxx(HuHtSX=1jJt(EGGrXfLnm?FAqWtwnMtj^7CdoKl
z6HGOorMR`Vw^(LT&%rFWkb15t^Yko6O@7ha%=zb>e6>;+wEt+(EFFn!DX}e72y@t2
zC`GsV>7`f($3l}_Xv&h+gNwXS7bCr-NQz?{{0Nh~o0HwW*AQSO3=al3mn)9>utWk}
zgMV#c?f1AexG9H_b}f-52POV3ygrD1nw4s3PiB7M)ZCq7;X!;btwzBN9-|kIwPpw9
zLnNfc+v-*7uZXHhy5!eIjzDt^)TTT|Yza0b)qXP+*!9RG14R1nq;ioIUoxsw%eW#h
zv;rL?r{d-N?@anyV0%YKrHA{6V^sE9o`uUo{2oMooR!JBzMmL!x(KgY2PiAEfzNUt
zH>;eg{Bg>~jE|Y5#TVhr7KB5C0ky>TRc2wx%@%)^wllDO0y*(TH1=l5+>-{sxzZ_-
z1_xy@@P`Rc<d!{}(Khi*c`5oBpt2tBd7Cy;s)k#%wn>~80M4pNwlt$Y)5PZz;_GIZ
zg>hjM*?2mjQi!D!jsBI^kzAxQ<*%?EZ}n&X?U}FB-M|@IRX)zLc<eH<NSm5FAg?u|
zgdT6&%-XC2uk0r4;F9(RWXg-Qa;#5XFO}M^0MMsh47BqNHS6*=9uRc>1@=+<*AX|{
zxVFD*ao(mlIuPnkCZ^F=9@{{&8O${_=Emee6Y84J#{7IzUktg#s`e;dr*KMIHo97#
zC}HWkbIlrb%4xA|CoiqEKh>!U=LWZVEw7iF37gJ#<$+7(-9<1j2!g-on){gK-~Fu`
zcLS>mO+^JaU#8dr%f0QPc%ip3WuCKMiP{O2o2}JZeez1fi8{3Gqe!5WWAX?Vq?!3>
zPhPcq1EEg6)nJiHnZ)a`T4lY#XA$`}TCQW@31In~uABLVy$)aD7x;MjPZRdY<m|Mp
zS*=c^t<RKyCzhA~c)o+K?~%xMZ8;i@+9(G!F&#WPNCDwk(^5OK6nYA|8sLz_BIJP1
z*_GR}L?-pTRHc@aq{3zFgr}>bIDU4r!(Dye1M<m^E1ie!#MaRSB8mIYjJYd<3u!`i
zr<=GU2Y#Du``yzYaXEJGI_%`Wjw@SZX955(kJU9>qmgU$v5Yv4x}y!nnZB@Mooo5~
z81U3m#H>Y4E`SBY!lUtlIO}Z={Y*XQ^uv`kz)8Y#1!KFC&Q>%vDE73gU^BzQ`cWXu
z6=mM@wodc+7*+N>P`8Z)#9pB?%(g-M>RPl92sw>qDh%BrdCkPlP*V4r8C~u*&VI7`
zVz;jb;Au*B$ONmjii<aBo8rm-OZfG6@BfRjchC_9+O}=WRjX{<wr$(CZQHhO+qP}n
zwz2B1^OBp~<m}w+ub5+u*?Vhp;#pbEK`H!vN-P^bC^SISV#KI=NSejWmO5QN?dKVd
z3__AvER-sVs8sLoAuh+_cUG0-0jJ7BzF<fdE@$F>rL=8l$FQR0cw(5RM=lycX!@l;
z>lIQe!}<`W*nau^i-<feDodTNJ%h!UA4^t>GClDwRgEP6cS~28@SEpn#w-H-_j!6q
zJm%51sV+V<K=80aWs~nGCq1HSj8V{C0E5(wRW5P%@C#!C*(tqxeq?9G&zNMLDm%jw
zEM<RFIWTIg{moP@xH)S^@@<oULUVUxS~Q1<RGmBm?X$MX=?{Sz{6gl)qnSEj!w0U}
zgVxG53yir1lz}>S=zod43nl02Hgkf!!{yK(mV8L^SW|#<pEl&y&OSitlp(OpuHpMe
z|I~QNTljl&aESrjD(4rwrgi)1>N-zu^0Ci{4XffuEB=PBi{*EMP%ttv@*wR5fhU7$
zgPm@qXkEEgse}HBUw*o#_6jNTu9CB~<q9J2N!<X8c{QHr<an!}_ouR*Lo8OZJo%~Q
z7U4WIljM*ybFmCc*7Rx4{W60IofTLULa&^=XJ&5GKq#nB@3IDJ*`x{BMAXd<v4|KH
zhg)RhDFMcyX@AmiCE_ocujtwEk2R+SA3{G|xZE9u0S}9)>M(5KPA*kwtkt<%v&x!s
zZJM`(nCyuK)T&x5O_*r4*p`LJd+F?TeIfO3yX15hMMDJyY-&ZQ?{VQueFOI;h%_IR
zHmMYB{Q-h_&U(78=HLoC#H1eLVg~5S6W~|}P9a+ki&Krk3>b618+0i0Y#}hGw?E-f
zO9lR$=9?QKxMDwoPRV=n9=@4B-l&Xm9`Gkc3!ve1&u0&cQAIP#$M;`oYEf6BsUCKW
zMGv8e@B^S-4elgM_M84T;`HCa+q!#N$HzK2=P%4$*hP3kee6XKs#Plb&N)_k+(T76
zd_B)>u00$YCJS@(@F#Zcsr7hWAbsx{$zPog_u6@tAjF9mb&gVs*RxYtRS-?!YSd`B
zK-}Oi)%<yycNfr{x+8ZrbEz8#keMw7T;f05*apGtTH!<ytdXW6EQN!>r9`R(y`_k1
z9lEPjMGd1XOXnh3-BMs|K&&gg&*#1RbnQ3~Lh@hD(0{FWmO<C6ItAjsIBxa=cHyMT
ztO21kgbdv5x`*X{u6Ln}oXD#Qf|=dedoZnC3ruKDB#73Uv-+GOyLp_BR0o#;-17;d
zEJ|SKEX~n%g4Pi_)WN&k$VK`8R`9;BPPk)RUT#xJ;3<CXKG6`Yg}x|DM){{ytXB9#
zcXJ2Cx!t_`uF!)HZPuG}39viW^>giB*dFUL#nt2zYfXA08L;6L>lcxLZRwl+KF42*
zW}3IuE6gaDwWg|`y|gC3+RjVG$X&>~r~g&M%zg9YS-UD!C^-`>s~JwhxKqoGC~%lV
z)%wW*A5o;%vxzUaNlr6yZNVi~qSRU`R@044R>`}kkgP2wqznXTQgNuKZ}@vJ@W*_M
zrB<@%ah)OZ{nkYVsVYVt?xoQ_RjF0gE7;3>21jj`L09#gWrcCt_}a4F5bmmwlLIn8
z)S<C!s=$iFKt?BV%zQHcB&{n*QtKcvuPcbu82KyD9G22b7~(VJb)PA2W|n5i10{EY
z+9_&AG}}p#gGWhf&XVjd!_YVo_6hHg^+!(M5*{Q!S~|neME>_c{@&bL1OC*&@e5dZ
z3;B;=2G-o5CAc7E6a3N9JyJ1JNzE=2OJwOn5LW6n!PwaA%Y6HME)wBqP>hx4Nzy2b
z;?z3xPZcrNq5%=rXvtIwk*Mm7sj>ak_@$}bAvQhZC5CMBD{eWg>b_O3LwS?>-^KdI
zS5cx3ahhmta|`-wBANj6k@V=d=GAF*Lpq892l4(nidCSpd|^9Ly5sP}2J=$2F+Mzw
zJYk?RvNHW&jM~)4XgS{ou}Cr5&nDHIKSiqAcE@zQKF~K21XIPQ^AFD^$PFg0Ag+9#
z<ra8X8x<wF<J4<bb2H7BPXgT!8n-mKeSaN#Xre7|OP9z>Wq!fuVU0`v^8o$7rNfN$
ztW5va0KKC5KN)T4ojJ94a8VQzML2!<eNs2Y{4Ufcs!qzy+D`LdKZVIe>!`+@;iO*F
z3~A#CJB6_1m;0^!Fgv|jJi0w?8rc6a4u{6+f$xK7{sx{3hy{Vm&ko){eBB>qt&TpN
zP2(+ZYW{Wq%XiO~qe}yl`~Sm8_Ma~i+=m~-PdGx~wVuZAe|w{KraZ^4f<&c}(^=Og
zFP@ji-01$N$LxHu_e|o~<JbNDKsZu%p2EWX{rx57>5rNWV+tYK!T$X*b$>PWmUuoa
z`4ssb_I!V47li&%uuf;$n&%o@-hgM$N(nnFVzN|Z&G%i&($G8Zc@)A|T?JTB^fV24
z4G^(5-i~1~x~C0X?I@CKCE>_oEn{2@lEHNmdepob5V<+E^2t4$Ynq@Id%aIiK1E8^
zWL-FuE0X6?GLF9*4{7J4Qx>RO@FGRY!mFsAPr(pIqFU==9bJUTX9Ueih9#w~_~|2A
z!1kBAxD}*cwg3(e3jT3B*Wn}+blt#O9ATJe)GBz1wzA7KiZP$i4ensU+aO0BhQuai
z$rLhF3N~kj$B8N<-YPpsnNU1psPBZIU{VWc<1iWM!qHR4RJu<Jer4n0q9d3kXPTXA
zxhz7IPf{j)HCTjDenLi0dd#fB6nmsnfC&>568~a#Mowkw7=@4LrWrHI;wZ~fE;>8H
z{?Ylw4@1pN>V4EwFLg0SRo5P)<soLoTUV%5L#YMvicF_=q7Wt~#HXqBD5=n|(W;Pv
zbd>WHQ*X94Y9wrJa18m+lEH%u4XYg<qB;8=C>7ivfZhCTkKYKRz-T3cVer@pEiwOm
z_AO6~v!$w~d6GNBk;;JF*aA~y?oifd%82mwzy(=C<^cRzZfWJKyaWxVMz>Rs<0k|V
zrgzS<AnwyZmHi_1dOzQqQ;l6u%FjPoF-#4_4W65*-;QCvIqozsC?U>33?o$04vdQG
zvDY2sjI~e#juR9fsb8w~C{p^uB?g?t_3{<1#ZO17_W~lS0`k7CD1w~v@{SwzgFIS-
zwnD~smYIsqF;Lni-B}{Wll-Dq`a4KUq5183jML|9BMmr{kZRw2xQJC>U=ah=dHMhm
zIb#!l?VFpEe{0PxI*92*ri!Mb#Ws0T+4L732w@&~%ZSR>WGspp=?l~%LHTr|reHIh
z{^8Vq>?8#Or^==;G7w!<J?K!`X{!vv{cD@TBbXp7&i@#ywc1Wf#lJ%qj;!Qy`d`Je
z24D<lZJa@*w=Iz4oMAHMQuoN*R_@bG`{T|bQQsbjGAPW=aDaj&Zc&E7czCQKrU?ZC
zwT_HI@4OXDe5DwY!StJ=ZY56UtAsa{fV^6pLhJ~>F_ew@Ty!f8?Dy70HuH3DBqnb(
z^*OBg+F}`F$P!?CtZJwbeO0TqIovT=C7%_XY?8TxE&dug<Z=Oj**400tW`2_j<7eC
zmqj<(MR^iqeEDQ11rFSD(Zb&u6x``}PT#~RxFvNtb`Y`gwUJ&)m0-N};v(}{AKCR-
z;R)jkTZ_zTcrEHFhZCu{C+{_E&V_$+XgN2V)=XK<Hc*)fqPceu1iFa{)oJ_;*UsOB
zGwEvZ4h4f`v2=%bB<kOVbIf_qyJqf|4D_MeYG*!u5QWqNszE!8^}A^JZPat1+kfNy
zi3QVc`s8P`|AjvO9Hn1`*M?;dbrp(K2!mO_Q1tqncI}3R=HghP96y<Xc6w=rf%d6w
zWDZ@UWuV-Z+?Frxf`c|-eOXj%rMkCa*8`}5zF*R_q3G)iTR$BZ<u_o~6X|{_T8qEv
zuDH8d8{`3ujHw(ukfhE7lyyToT}q9t0wN$I$iSP>P4_3z0R0z|>Yg=jlGjN$18|Ai
z%wyF=Q)#JS{)f>m2Mw8&E3dyNrUu3Ll5@iIv&jVYf<de7vXzGI`jbdt2>qs#4>xF}
z0>uiW+AUeO4#Kl!nmiH8rs-{TG-PYF9=~HBw`EO}E*B8WENFi4=!YC^Oe8afDkB7J
zJ5HVUDjcciSYoR8t5HeLf#SDOpvj|6KOTKUp^s@~0iH|`{b_Y)(Rpmr+h`GIrL{z`
zHfEu5FnL;ZCWc>qTb%h*rs>|v4di0W5ry~;B3j9Ko6k{9|Gk;<HR6X?px)w~XP1dG
zZHz{t$zkc*gM*#pqc-*MG{l*5c4m9B_3ozHQAVHH%hvW~l2;1|GcGODiNgb!HBD+&
z4BE<2pB}zsKOe7GAY-wGb<%Q6VOT{^V}?Lhca}H*PZ|pa`rQeXg2obuwBBdc<)LPH
zG~dMEUu`OoJ!fIptaP7oglvWN;zT+dcET>1#~p`Yvo#U#NTO2DfBN1DSO^6NQ<O*B
z^#NgC8zZ)_@SiF&(7hiR@{il~J<yO;+?L-MAf#9dD|B5i+Qwk{{neQ3MVaiBrnvi;
z&^Xhj2OTy>BlA$j-lHxfV_w(!t|C!gkC#(|n|X#~SH;Mh`581A!$-4hiuJ=@CeS^c
zqDIYV*I%QAEl3xs4cAvQ!Z+St7Dv#%snah5{RpH<ZPxRfjEZ?w|4RsR>vMpUMfhtl
zR_cO!?Iuzqfd&9oj^L8r`@a!s%idizX?Myqb_6<^Y{|79lL4L<x2u;~4RZS?T~$^h
zyHBDUpIjCtE)-W?#0s_Jf;9+2HRXaAB?%h6MXTdZ6`nGipCUr=WVcmxP|_`#-biw2
zS<H#m%dWl(P`ZwC#tKV58Fsp3x+;Tkr6D6HSyr8W04$~`H~(q6U|{&)s59pOI%uFh
z5lhs8(6jch0Y(}21%wQlx?BV1oh7gbjRywv{7!G<&hh?+qfLlz%&4TH^4JdOLXn{z
zG!TvkMIzgmGkh_%baJ&M$7Unz`$yH&<@580_#njpx5Qu2Oip)l#KPnA^XNGZjUYK_
z4wP3EqUPY}z{|hy^+jU}fQ5KDUY_#gs)-Fzk~!JDyomZ6<zC@_XRT4o<}Tj{ezX`j
z->$++?%<_P@AsSdGI(r*!wxkac;=QeIORSd6;+gezA0cgI&3#55EFHjm4W9X?K!qr
zN7E*26t(4~#@mrgZO>>9SL>{;p7W`qcjD+Moir75b<v!?Zuuz0Gohtyv+rLvC6C;2
zt|GXOU}fg1)0IlRDN3E~?Lx|`iV?@J6O?1i=H|%I8~dVyNDr>2@R|7xE6H`Vyn*a$
zR}w}&=(*yg<QigGt#WntE6Y+3i^X5li;8)Hj-mF9Wgh#qqm8o)dVV6N$GCH98o|p6
zzt+VPL0|#f@Q4Kqc1FmujT~h(7jX>T#$J@)iHPI@&XFk<)&;YSU!t_xYPkeSMJ=<v
zt2OHp;x6OP9jhw4-KQ4Wl3^jbO8Sv#BA3vulKQb_+f<`bjLU}ldMDCVLN{dtg<`NG
zxnoGp_lc_4i1`dT^|KIthi6fk;3lBR<iMF61$gAit#Y7Pmgl8Jr1ZTdJDxB}ReS#3
zaU@KYTCt^RKMVWVY3TzymWj6H<onvG2PO*16Prfjmz%Bdgm(<Q$=#i0Hbbg7=AlR@
zS;<H$B-?ujM=juQ6z8n@$VKjKSh5qPRg3t(AvRmD<=3DI{={>C3E|BsOIj+AIp>M?
ziRwv5DHUW~B65%8M$}?#k66xGJ1xao`kTO`LMsx0XY+l3%BS5K=2mAbBqwNg)rlrk
zl`NB;)kv`e)6D0J>?*Hbh3B~>6Ir^03pRN4deKoI${KkC`F<A+k4mx1rI?(Y4VjCC
z9qDF8IeMY|Ko7k!k&=Y2#mNO_fdMV6e>HMcfZV)l-e1sgYn{)jv>B&z<ZNQ2(G-gf
zB1|O)?3*;1Dig<dBq6APiODM~UC&EN=KR#mGg>;F_-$2KUj&BQalMUyuEtddMrMMC
z=wqpCtxHqaB^!#Aa`Dp->4qA!`q{e8^TFMgThcF=7Vd~El%a_Fa(&au{K$OB`#Zb~
z83hP+gN4-SKQnL%#VA<FRqB%98ORE%fwj0hi;4~m92~G0=#pQ9{E@X{kOw~<wG%M&
zznUSk2!KE3L$E6!n3XdZhz_bTP?yonNHWc`h-g3JF{)PQk;srp@)OW$ic2u@C8^j5
z>s&`&l<8s!$_7KpRNm^&&TkxXPwklQ+JxTq29Q9>nt|bDR)C6NY$yCk?03pfgjAXm
zc6Hws=_E4P37z)Po3~kOs#zm}+t<+Q>@^s~w{L!H#|ibT0M0n0^E<sabBk9a?ewqq
z%iDh(A_l8h(c2Z;Oy+Nb7;*S_hA+Cq+2SL&<FCm*h)%xXUr?P_(Xp0$%XUe-9M#k`
zGGNkwc07hRo1dam4tlJ-hGO@$nj}QCl8&Siu@?z)CXTZwe{$9o+j7&y<cA^K*5C7Y
zXdTNlkiy6X5J8BuVgg)$7oojQL`I;3kpFqnY}3nHU=DM>h3%&pgh~XrG=nnUTa5d~
z+dm7ZEan`zN2DpH=n<SqmosZGn=Yv>nTXRSz~F*5>Y_0(NC0Ikhl>pNEcTBLV{{e1
z(utSz529O}8cuHht}D>Ub~d<xJ@NKV!%yRhs{t$Odhuyxg<mw?$s@Ts@b)IvcBG^=
zHP{qR*!WrBpxUhj*D56wjvPkg!Zh9_^@`HG)ScsShJ!<mLjPQ5WQ7!yy78qU*xy+%
z{Hx%VZV+GYZw)(j9(B?8d-fi@q!3rk^+fxqYW6TkPG>)$>?sFp<$-|{Dm+WGple$3
zY@ghn+E2+fcFLVE;}3Q!?C6Y{ZPApxoU?Yh9yvs4VdwP_^)$&|PS?ty@$$=IV4%@$
zw<n0LEaN;{eo0cXGi~VBXb7vhF=wzy^iCMWVvWI%1`1h`1qMRAe`<SAD>I@;Gr1m-
z6}J?1KCzD{$CM4e^Gs8ID0rM{$5I{yJze>$npIwVO?&{>-iC$Ivz4n*%#0(28Cb;V
zRH_?6u;6CI3?fTGWOOOHSu|i)Z9GOBge{BI`chLnAg=}Nw7jGLhL!75HklKQCE$wk
z3J?z;^2JAMUWx=2@KMu7-4I6BiC`Yz5$AKsnICA-mVy(g<w&m+eMmc5ygr_?JJxFK
zvm!J~l&u4>usvo06=u-439CfIaY;x#Q@-e`@)z04=3OSMfJR$GrVDLlem8E`%7%Go
zj}||f71Tc2sb&=Z%TV8}SAZj+8e@xz%#}F;rTAaj2iuhMP;rx)`5KYkQgD;48f=8Q
zi*WD5{rFn$g5rZd^m;@1qXPmAk2@re^k4QH5iG&W&<WeQKyW3?bfI7@or?bC*SbE|
zXUxa6;%u;RQY*AV2MzGMOX-3-Q58Kkq2uHTang$g86;eK*=l6diCdLRUmLV(pNim>
z8+o8|Zi=+uo0f8Meug?JLgf3Pi;WdoL(FA_3SI5mLuww^`zQ4B#tGhl7N%b)Z#i0c
z+Z+H^;$+d&<0Mw!Y<a_>_m;qbYS=?7{Q6>!s;1j5WbcY_CwRD4H6bZm)k3LZ!!zRx
z@8kqY(APfF@%>5ak?ZU}dzjUSlV*<a4H;FK(FRiIZ-rk$pB$$leWu_EhjJwF?z7dr
z-P{lfbt#7tvH(;8Y9r}faD-7(69&x1AL9<>k0qQdMsXxw$rtyEPr9z{9txdEKU^m?
zqmAiq2{MV^vELenJx#dm7*&{^hIz>7Fe}7*f}Om4dxVvW*?DpGa;&}THT-iTJ<9~j
z4@@V?5|z$H_ida^Ikg~(D-`=W-2pR4Gs8hYGc1~JA~KV$aS#%Mb|tA0x$l&vbnaeL
zJHp!0?o;YcpUDWid1*vvl&EcSphH__pSOD3y?u4IhMLG*#~8rtr(=@}Y09vg)BEH8
zdG^PDB27t@%)iAM`t><ai8A$%ItREoNMsC%Iz;=`MVa6C>yO-?ZweIiyPG9}ELeSd
z^ucoxw<c61`1Qfwm68A2!d8+tVY<xe@h~9{IgNGb+@7qQ=!b|UfUEMm7iDzRAFyj6
z%L)>872HaNv+8o#yv`y)>?=3ja3o9M<CS^rMWxG<R++eYT&<F^=Yu(Q6wRNT=jWg2
zXC?(k%Xe@;oyQ;<2nsZ!p49Agbr5n%@{4fv1z;lU<i`+2g8j+(>`~Ex4nkt@q6PB?
z^+D8Vk$JRNf01`L>OK7a`c=V1OdMTn>Cs=rV99TCDh()$tv5S9%dWlr6|~p3=O^lv
zS?Zr^1HS=Qai5dn#~N|(ir5wffgCAK>_4s?3b!W4!B5}rFgb(BzPt?-sV%&1BY(Zq
z`4wqbGl4-cotCi-l%1jcwTh=gLBpj0BoJ^SoKE6O#4g*(>nB5354-m!0lv^VS|>>(
z8+<U)Q|<A(9q4W-{W}eDPKS1zp#X`sIy7&@3fdT?pnZFCFW+gJwx)I|Szv=sGQ-J+
zZ%=|OB{?Ib9McjtVr+5GRslKP<?)D8_7L0l#)&+G=0=6%A8kjnu+m65!<FWQdHC1w
zr@g1s#yET((GAhB>Sx$TO0o+m6c>9??Pfpab!l|N406wX%E?Vlvaq%aQWJ4v-^Yn`
zos_y)LEAHb?AC2SwH_`*&HM@0H3vgPpK#;*js~yyL(5!2{E!-$I`)j+6TbZTUREgC
zy-d%o)q+yA8TD&NGG!I-jALQ`Y`xLC<Lfd5XlDX$om6dHYyJ*vgRa;sgM>Gzrr~&)
zW_26I0$(*+D=`w7l7^&c+YUH81KDf>2}|@-1<uZlIeea`l`<lUFxkCTl?H{aRhr9i
z@HI|5`JdQ=PL-ieFyF_4p|iw}(@|{9!kgBnbOM`mYEI-BXLThphpKS4RX2Fb6>Y1N
znqN7kxGBL^$9S!=QoE_v?#a(uVjQs+kI&&mLHbuW{iGluLtU=(hCj`m$=!*{Z^zI6
zqhG$GAa2Vt#BBs{O+OZu0}<1z8wk}kN5-lxckRGe9+8N`c(R?9?EJgkqgL$%T^LAD
zci0Kl>Ex2Q4RdM6*<|Kwo5t4V(b^}IrB11O(L2>9!%kB<y6rhZsPAQb6;KsaU+q}y
zbEI8Rsf@fNH$M;ysYmlyoe_$Hv=lfnpL>&EYC*9?)A0eK($H2ysU1%n975M-nsuD9
zxhnrvb_2DpbqBm;nM4~6(K;ie&1V|haO;$=x+tl97JbrL5Yw5j+0oLTZ&{fH{h1Uo
z=IxA?>+AxG?>dK%!#x^rYM-(<C^B-|_PqnhHNU+mQg2~$?q5vux$ssKfj-K~lOZL)
zNesid)5prV(_dptwo=S3jVL)am)^oi)!g(r!mfL0$cVo%&6>X^gq&#dD)IH40+KR{
zbF{ix9l?ErMRM%ig6%iY#zPZmZkm?ch&8DZ;C)l{8@pc_JRCtWc>TWqsOhXe3K@@R
z`MESydSTNuXGM>IH;v|%H6&n3eYZBAZU%^*T^D;OfjvIRIL*B*AVzTQcGb}IF?~Fl
zwokA_KiySvigzOHI1PpQPG4>5M_uS@Sd$#30M-C~jnLeyy4e2t4g(>Fel>kg{33c@
zY*ZXGhF<_h6;G;dn3+s3#!@v<>T@v&^643wK71U$i^*8Omne<tt6Ot%#O>CqbE_%y
z+<}~+AsC?Rs+IOV%ka0e@aeYgc)kpUmn%&^<RetDZD6sBG0_;;QElvs0&Z^gM>}cP
zoTMCes(i?XO@--h$ewEEF@0OX{h$F;9L(*T37ll)ZGe#z@BW)*CHK+tUI)HHHu7yv
zS<jYpJA0b$i~!<x{Tv|L11aEmj5<7Rvou)0je$%~Y}MI+m$fKdzLKFJvzGTnepCzm
z`fkcLR#r7qj8L+Xxgba#Ie2p18B^b;udd)4a6^#RwPNSZw<nhX#9h$D4>U#f)(BMP
zDkbWWzW;?6YhWRk4tBl3OUXLU8AmmQPG?3xS|y?1F4W%Ucm!-sy5p?(stNur@sD=I
z6xK|hnYYQ_;v$=nBLzE74?N>`G4)|F<pm@PBFDe0j`L=+y|%&oQToyzhqtmpNs=0q
z18eCvD$B4VePR1tL!wrq#hopT+2L3D7aAb$bogLQbMgc2z4Ufwh!NOE+rKQd<S$+=
za_RRUs~h@+N2NcpB*ybITTqI1;$0HdWrq|fBg$q+zSD00>}rfjF?Rs1!rmiG^t8%u
z(||FRA8Mx5X})?StYf-#kGUx=Obw<xlH2<TGqAGF>5@ISgHT<zqjaC^D)9@OF!V}Z
zB8BU=P4*i#$)B4*Kqa&%4w`KR_Evf5%FVCmfZ9q2t9q+tQYkfZ%4j7tb5vdR*4Rwi
zux!ch+_p_>A7>ry<eb-d7rx*J+aYP`RD~rI&Z0iYe07VyQvpXW7L5r%>Xe8teCNZO
zobBM_xXwDAaE7rd*2ZKjHPlsX^Dc)R#sjs#K$?EUa{u9C!N|e!f014PA66rl|3hwP
zWM^ajFF#AP?6BKVeBbK!jXnoS=2EDEfF-ocC^Y-)mGWy0kRy8}lG_sr>Jpy!-ng<C
z#2;^0JF7w72N=y#g?zcW)9$bv74I@7E8q7L?U^&nOe-8I64SrTTZk9M>T-+M9%_k+
zFlsWjeVWNzbVZ?vkQ2oghUG@wyGQmkNB2?eB*-(7{al!>oUM$OBl|SB%q=TRjaDI%
zcT4B!6{?Kr2km6MrcvoUGdP=tL&hMcBPydw>6L7Ni7g0gmGq>D3ybU|j9v8iBTHe=
zrOjV0SmlnnO@S>fd3l3C@VhKPq_n5vA-oI|8L;B(mn$}qs-i&;*jTZA9s+yd0)*(d
zJ*1;$5)3h$B8V$|s)6J~JTL*V$XM?q$rgCQl%8F1n<zd%mjxEuWXAF3EI2vYOV5{b
zyGUquMy+WmE||6CIPIW%<!hE8+^rbzF?VOT3|TDlhRu?x_eGMDRPsJ08#OLVC;WL#
znWRdcz~xSz!cwaznAPQ0$rQXKc&FA1yqIqoe4Sn`1d_EU(W@Dflg9Mrt-}KhS>sIc
zjy$FKd1L02+_FAXmnHF<S~4*g)InR@9*RNW)siH7)!l-si~gM0vC3>etWKrZFl8>e
zw#A+jOGGbm#vs&}`7HjVH;wVC*1{Xs83~~5aLPzRE$f6%D5vYItt?bmXhG!!Z<Oqw
z!e;C5DS=OG2!brE($CUk;;yQ@-y)?_*2K6%Rpo>Y1mf>5kRi7>SIE^E%qrvge4JcP
zOxfeW^d5g);{ATBdB<bDqKW1Wjb(egf4#08js4b9MVvz3^qgm^?q&D#@OVAkp1*K9
zc*m>G#j<9-Y)BMdJjnIg_4+h$Y*s7f=yLP$Z0IMJVMB$E8>W`64@uT{KEQM8NR)-$
z(;PI73b|L@GV`gkL+0Xg?=NrzFsB>dm4QbBp4xG1%Vu;5Ox#Q5bj5g_E)0rUVgPgd
zc6++IetwR4J~_TzPb{sxUG1?oW%vHxb+M#=EnzCMZ&b^%-i5*C`19%Y@%`Apdj9yn
zRz2L>U9pXc>E-?X%P7jK9`EKEf-rk;!)}-3@8ja=$Qidn3K6F(TgwB&-tqdRs-tNX
zQirN-6u%M$$7g+bQ8z}`SGuxRW@fi)X7u`K`Es`6R%NHVdU5da`Ep9~%pP_(KiVPK
z63jgW*{pO+Tc@f=68+ck<>un8tcvUc>fC}21dOU+Zx!NTA1)404{_FpU;c&zyF+z0
zRE1zz;0qv<9q9@+QdIHJClK0YJ>LFmR}egn27&!)DiM3d$~1%yHGN1c43NM=b3gdd
zHo|Tv`3??<phRkTfTe~3GIRuUy9Q$r&LS&BBJ3T$y|lbvh*MK%zP|$8sJ+2*=m7L?
zhp+J$o=et_aM3ly%$v<u8ueWyQF{Z%YO77?dcRf%!Gd9E5YP;%nu*25P7IthnOM1C
z_UwRGzldW!j1B!@<IrdHHdkBde*d!c@)d#eY?FZOFbO!>pB$=z9?9<A0>}Bh+VBo~
znMcfkcK}*|PL9G>Q!qb+MrnNy9Y?$F5I_(RR<f0L0WJQ)@LEeA?#&KPIL^kJ`!vsN
zs&19)^b<~<33T?Kr-m2~S=|Ph$E^*pCz7rF5kgX9Ca1nGo8;6#p=CgD>VZpEU=l%x
zW*Z2I)ok<&V1CeOk+SpKT~s8hf$hP_{v>dvbIf5N`j=a!0?@;h#*6C+a+puB@Lgvk
zqGg`Ak?_$p=oMC>TJo>g2^0b#Dkm#NR8#6cEzy5HoxDAPxc&V8>U@74p3R)@eDT$)
z^5NL|{=FCS_4&+O(c|sr;q`R<K8lm?ZQ6^S-4*8geMDu)(^2%WT#LkPo<R?>j5WT)
z)JZTlxnA=@UqI>F7-P1x7M|P^Eq-woE%u!VU!Ro;^RZ|tX{qPeP|v5PmN&PcF_1(p
zN**CM$q@R6??&iNbPuK#^Q8z!MZU1G3di*JD<w$QA|8`xidgQ@y2W-$gedvRwK;la
zP&MdeQB2-Osfb*)UWX5EkJj%ap%JroFuo>)vEbz81?W3jC+O)1po0{$@nY!KgG(?8
zrQ-yc0ElueXC4=fZ3K-l9^^$Nx?Ow4N5l`}3bDvGf+(eX=mrNyyr2|^7a3@%a92nH
zz78TN7f!GgC7|R`Ho${WQVXKTrrhIVtOF+2PLdPT48Y$iqz`T$ipL`A2o=VK5H9Dp
zuX2K4hqRX4gAy#FG+li_<62C3#R>j~q(=BVf}IJQb56fZj?M8M$%72RPUj|tLW)_U
z9hj6Pi!d0Ud^hfw0Mx8{AO_eVKeCDq55jHrLS8dCl&U}g47ddb7#Tvo+}|1*L@Ypw
z=U`tV=Ws8@L(9%?x;HT3=1dZtS+4EN`aEEg#XFr+gS7<SCiN_UT8R&(G9N;D{y3ZJ
z);bTabsj|XJb?PC59L!I!n-c0V4bm+%Hf?cbhTq4dP5@iGRK|VAY#H7BuZ=r@uqmk
zVcswVTQtryM~f;j^)~Cv*Nb_L2Qo%Y47eCQR&Fy09&o=t#7+)06@|&kO#%W0Z@6KA
z28JutdxKT(=>&R$e7On}eTN@_v>1lQ0ZtGPLT!Q<YFmlWkbtImcdzUT5py0LV#HQ%
zW=}!QV0B=;-5}OA>4f<`qEa<lfyKh1ecuWxiGmFId)0qDujrR=jxZu%Hz7Hp@Egn7
z7)&fVH+mzwa0n|Qj$Eac-7mAE2;xADDGvL2YiUsh%G_>c0^G;U-Xeq!Zqb@BD@cbR
z3c<N}7>}SE<f>lI05V^%^**1U5pDv$4u?1_6Hg()i;A!ADAX&tDae5<&UApLm5qru
zi#^5!%LXlq06MaWH>x9Qxb7Z+kD9US3sE9soyIP%2o|Udn{d^Snpp<x0}w|kALP&-
zmS4~)MZrZZ--Re>PXY47;jc(&50;fhFHE6Vl-~?D)N?H?OfGMAH>5+gofH(s+<o(7
zYKV?J|AE9(;YByFoBq&(6J58H_|P&1N~Fi0PKP~>2D^$1y9zrnurJ8@!48kLMz`-G
zQ(vGvc*hejNJmjc7y(XHEl-6<W`ueAB0s?;-=9WJPE0x6h!>I-Q0vdds=~(Z%^PeJ
zwjUKMDT}==yk5l!$>FR8kyqQTu<j6%#hK5KJb>y9jSmr<YOCv}m3O=s>&N4ZW*RRy
z=c_rgbh&Y{b@ayfi{<~1$QW0-7{iyMz`C15+WPmSIO6FgxJtHYuirvDam+jFP|4++
zYR9ypAqHs2%xQ4NN6T8Lj)2v93o+|Pt=VfWH><XL_2!DK<q<~J7s2FEbx9=o;>^Xr
zQ=6}1>Lp3j+~Q%e^Ib44m4-%7)5ao4tfqa`c~|qa9K8KdFnC(S`&bIQ%ma}ug7cB2
zeV!(a0;j{urQJ;!rqW^28z2t`4AqS)g({GOY&8)St4!SVAkFRm$+AYsFjSn9sQ6W=
zcs@0SRz5rcwdzj9f$jiG<D>n=36}5{Q{$)yKUnlueSj#E;HcEsUl^BghTaBHCf4bm
z-K4er-nl-qQzkxZ^%T`?J0t?6^A5a!KYR-%$y58yKQPdIBuDrdb|BfU{<*x2{ZaCJ
zP5MxEoq$mc<Q_32(bRZ!{FsEPRS4Tuv2~`%Gkc65E1s84HNi7niabzwWC9gAlsWc@
z_Zk;@jZ3bZWDJ>j!6my-4qFMCVc`ncVVb;^LoXk`xN)=>ZRF>iRQTYaOuriYtg0ub
z>2p5T7A|>e=;5UO6$b@HG`P6Dd|1Rry!g6)@d_oQR=|aYRvX{NEbdC|Pbcf586x;q
z-w0^5-l6e_Og3rIg>3PwUn0sG?SGilnHZS<H(Qd4iRHhmF#k{RAFB)DAHe$jUx)b(
z6dht=JT+h&`yKW8UVnIY0QONA_T!5pQN=hxrsWuP&YzkkOX}vOVLV|gO~&>vPnVCk
zrvtMK6d?!F^XuJ1>0WT~S5mOZO-o+Vy{I*AueGmFyn@GD_rR!x8H^XuSM1%j%&ds;
zS$<wfph~gqOdq8g-JD@wjO#u1e&L==e+y;I3mvbd#RDu|3nQB%=kLQqogCty(mT--
zR>6+u4}H-2K;>RZwrEc$7l(<W1w#HMcFoJCWLWcN{xHzb<*lWrmkdU5We&ByP7ph~
zdq7!hkV_l)0k-})sP!ygv%l7<aX}k5RIx7~ZeB!JnH|5+p`*!dH7i6pYQ?04E;cob
zkD3${eNWZ3m!wl?skbuEgz4;4Bv&=_uaZ_KIcBv@Ea&{F@~TbaLtRQ23b}LavrFt*
z@xXD3mUfZYC^PED-U_EGEh#O8e~}eIAC(*LTbAn91Vv@MnKAS(&AoVef_QrCQ9Y-e
zvoESp6THq=UbL)x4C$ooAoawtCc3A~5_I)`h9;+TdX;q(z~(+fDi;J9W{K*`yUJ5L
zlQhXoOmky|EMQy}{fb^;X$AvDhI8;Ih<%89BlOkJ1@EAcrqOA-73lKU4MA$uu7@ci
zcxbGqU@4^LsPT7(eQ!wxtk_J(L^o{EbTH$R1TRiOnB6`=K4+higG6m>p6RwV&V8W#
z&@3q|MO~5zT3D=I8aI;!ZCTSSxlKmrha^^kXDK(JBl{?Rfb?>lhD`Y<4$C1p3DbI|
zW8Fw{f!XvjShytHCkLFk>bH@x%8>QBdpqo9BcZspi1a+EmiSNXDkP9Z8m?MMqfo@e
zmPaaO@+-lL)gXV#YV=;875XcgnCJ5K9dia{bi%6JVjEVhB>B6QbDgOd4G(Ik+!Al-
zyM8EW!8C#yUg$NkRJ3ht=*l?YWl{&_`-@R0Qa7jLLS`|mZHrxMM3Mm@lTl?|&KJ9b
ziT4VZ6!=N-j}uWOy7Z9cQ<w%ZnC2v#S^D%5MvD#ilC2R3W0CbVk!kbA{NM3Cqztm-
zaC?0cJ->kS=RarmII|Z4UJK#8bmc(PcaGCa%$lCBtC`+?wHC?3Ol?kx1;jmd|A~}B
zAPuzpgmGe4sjH}wN0vxy4y7WBn}c;#e9LDN1*iq%K2I6BM`kfbdOk%tKx$eC^#BUy
zL3z<@d;sH}noXTIbf5&HuFz-3af?tV%2J~uf`n6ddX}ym6DazJY5;UXh(XSehx}_z
zV~D;|tau7{!E|1)6&Z788+5rsv8cuV#LP1BfT9N391#^E?Is;-DS$0(y7za-+su<k
z6p>r5yBSThFkT<8T9-{C9~ce*fSuJWM|0x$(6rzt17KM_1Zu>pWF87dU3RSC<C<@h
z>RX+iZ+a@-(#c8jdkRz!F?nhDVNN|c*<>&(JRPMbY)V62RXfEeH|!bb_?qRkW8gd}
zS|jU(JrDy&W0S97brc=0*(SYiifW?^8mc?zEwxuiGDmi=ZMsSL-)6p=C8KS!GE=if
zQ~9;QdAa9&bT=t^3W>XF&(Zsm`1YA76xb?zEf+R$YqI2Sm)zm4r)Cy<ABB`t*<n9H
za@;p2DXxE56x}mgS1-Iq;c0WsdC5AGs@Dmlst^V|O`{q(uMA|k3Rrp73y@oIc%G+*
zv)(Qt?AF@dNmidVpfl#2PFIV!c+C2&c9s>jOvV;HQ@qYK6yP)kzaYFx+{{p=;?5^N
zPc|Vo?WvPoTCQ?A$quy22r-79XWrDj+9-@ZsmI=tVhQ}EcxA~*T2U&l8m}2GZ#2tX
z)3w}SFR*^|)J2xJ&iO;J){Rm|9({rm^iz&ctdV87TjAQf4#gN|9xX!5@?LJ4f$lq_
zI{5mfruuqslM3!S)jp_0h2fQG?7RACZ`Zg~06FNf1m7}1eV#rGd6T-4pun|Vx;?zL
z-=<uud7RFP*ka$jw?K*Z^BjBlmB_bwXMPp}<(u2GS-$19e4x#6r$235`;?>L!p0s`
z*?NN8T0`o0yW$Y1rFIVaGO^&BGuNsIZPk)fpBHbM6XJMrRjieHam(9Tw_7;0=vacJ
zJ4*05t-*ee%_R1QC<Zh_fcvWj^0;hN{}P`rteMf%%BhQBUBf-Qg8Fz9pW37(;1?Eg
z<dC4ERj?$>iVDZ7X{}ChvwZBT9mqylQlG?Ibpc;$r|V~faK1rj6*S&N+ymV>)QQjA
zBg;-~50kRK>>~SWo%(8&gs^L2%O?ZA`@oZA*U~oC4mH&xeDo7t%iXpM+A8BQms)Zl
zDaTZO-9bDqNeK#fn_{cLyzYmo;J3#{@AqmO!pTzyPguv+fOIz(KTPu#5-^C)a&<~(
z)Ey%W(zW)klV$U*!b4p{PI;ij7x`I14ReLTA!b<A-aVy+F;G9z9%j{K-&J{pTgbmt
z5w^PF=DI*`m2eA&GE+xsCe=3xt<BJI^Ac|edNv^|!<VHoWfct;enOj_#nl1k&!j$&
zPGz_!fbXhOz3JBL0OAp}6-&%@*%e@q#D$R=i|z5l0huNkiu4uP(4x7%9+$H90doTl
zQ?<duWx&zp@esqwRs*~tPWpnQD}5!EtL0XP9QI<R$#qtC+^%<98Ao|vw1P{gk`C7c
zUX84-6M_TUm6OUW#A%>ySFoo^v_>o1BVllW#FS3`@}NUi7h?v3brX9uW_a4Wdn3~L
z?-{8)u>#Sf+EjOttT8d`k|~YE(_-r4;k}?|fFxO}i)d+i<deJ@4mvKvf-nkxH6Xn8
ze(^qX1%_2gA<J`#@7tA=yW{2k;nk*a;I$d0KnUF6_3@gds|FGYX||wAY}~^8-R9-<
zNjJ-}tq}K}DIM4LyxVqH1yu#Twb4d??257M;IwCYDijh;qJ4V&0>jy}D`At?S9IR`
zZQZVJT=0Tp3^lD%p5%RaKIPNn{XJQ-lkE5ZJw?EWyhlnoe68W-1RqLd-BR<86|4f_
zrbsA$aDDm!z!A3q)*33i`GbnMYChBHGgBiP4B=@U)C5!*qGtoRD$JaWQIWFxHHj5b
z*(-cdk&ux0NX^d5H|UO;g7*~0r=*7dQg?Il^!i*d!G!Ui2DcgbOyw|GNNJ3}Dk3RE
z8jhU!lHz&3-=t8B3C`Le1Kh?thF0>#3k2Opv%+42I!ePbpv(-+vhcR;8}t>iKnchF
z8Jyg40!5;05~WPONV)p=Yl#i}6%?$;Y|kar%tW!{>jfH60_5yTdt}L0=e2UDf~{lo
z!}v3TwR0mK*y1E5NG1vB4<4(ip}|YX)8~Q0gMVW7zuvT&EVT|b=DvH*;dv(B>na7r
z7!dp%1?ROO;%Pog<NSfbbX{1dr%$F_*t@K!g26!btF;cV0K<|?y|h`*+6<S$j@<-)
zVzCzd&2o@u3^J_R9B~)qxEvJA3b|`b{Kv99Tk?=y^t}=qkMZ-Aa}#KbnhFm#s2lz;
z+~}9UmNxx0$sUilrT!r!e>BvVCztNIiJYlJnZit9u{KY#v+{aMjGuzTYkxx567_dG
zm&wbKU)rzV>EH;H$+=l-A(ytgw#`5=tp2iVxG0+d&chBHTo+ao#A9<nKJs7h_%Ize
zt~jInvnP_{2-Db~L5U7U4!V?yrDlaM+iURTe+T-!;GbrY*5~e!!>mTfjAKXt58F^T
zV{uRt!@gnme=ZH}tu4E4cZRLl5qRJA{X<9RO%V`Wel)d(Q?8z#Yn=6gg*R+a99$u~
zw%j;uqB(FLVheEp@mDD`G~Ld9G*dVzxKCM%Qk<NYbBz1Kp{<rREP!!9{5-%Rg}<jf
zA6N=wor>e=Vr~`PI@X#<Ix97Jk~2dH!(5RVWGrXq(}9S;-1&Q2-9BauQ%8}Vl7eC^
zObK_0;1Vo%r0$(iS1Y&H>O7uSYgJ*v$zGG-veuJ2?ZOis+cFkZ@-on7k?w=sdzue7
zpYO6Ql&9f?%v`8T88Dw+CnhKaTo!C{oN5<%Rm>LiSwZDovaBYLIXO~ESB`v^XG7-)
zb>y+o^`Cf~UYIK~Bz0+)uYKdCE{%g189SbQx>*d`wSP48W`Fx_4!Bg<z_8AaUByxm
zT)#&D=97jE9LcnHl+Mj9cMx6-hX&=TBZ%f#)k#<1UG9nVG)ABL*(jDxBhoS44N%;U
zbSJz0nj#9Z!XYvLbc7y?835$gmjn~}%2V}qIse-SCLlrUL>qmjtxrI_nIl@LvnTxq
z88?*r0{Qe>)IlfqLvG_U1W3E57(D*;$sxHu+^Rt+N>LHNkBhGx1$v8{Z&Dak=b@In
zIbOsjbJbD?g#pAYL0i>nb}TpeScUr$%tjW#bV)AZA*KA(ocdj`Xuq~}i@`;_SPnb{
zP=e^!ZBL(Yp5o>fU_zC$IxW+Ikf>@?fSIo2F*@JkHUV7XjOx65Z@|~=X#qas&Yr5n
zrWdkr8UEsxoG-Zj$4!fc@qbbL|LJP~yG?6GvnTe575?@I*)!n8VI1QxKxjaVn-6~{
zb}N)3Xos@e@owLEJRwCyva!b1jugREY5BqTQ&xe=n;cnJUD(ss!9Rt$SqM>mrPtRk
zAYc6VpG0sN7gT<9n4PV-+qk0S$$OVmsFLMOx~V;!y^H7Kfd}Y)jQ%2kGP0<CqD(R7
zIfY@jiL}Ez))<|ouh2a=L<V{N&3vSiv!y({J{)61_C2Vy%;2VVjHo~l3GG}GncYuE
z7kVn9M1HfXHj;<Uac=k$uBYf}q7COWK*;W%=d{-!_Do;c3g)vH<YBdghZelp<3LC6
zI$|QCA}LY%)L_6$jNr5-B+8k+`|NW@A84s8>(KFQ4*FCvLa&4$q;#WhDi>)&0e7P5
z=7Wekju`DO=Aq;=1D?L_S<`>M67d*`i9Z)Jx=FITw=#ucFHJa+6{DxxIao546dK<d
zbyB~c*@)sgtr^af`}&HR?h6tF4GjEMk6DK)8MS%@j&@O2AQ<{1#f%v|8T9O;ta<v0
zn=XpxjI;MU?iHPR)}<e0d#cf8&G<)LMJ@HSC`TE@5~;M0dL6(WE!3uiJE;F|4D%gp
z7z(R)i_;`EU<!6UG-&TK%M%_435+lCM~$F?xJ^PKdzXnpHpVsBwiC4db1cx>`z)L|
zrD{chiryg=S|JlsZ=%t}7u}atnw$qD%8xvHQDl~0K#&w!v_ZhAWBPZny8Uf+D|Ksa
zLw#YJm%@_hVmo?#um^xZ;E^vmii7|Bg1@)Dc)oQNs#}5BcUB9gi|B<FaTo&E)ZPp*
z?Q2A(lGK>lW>yY%Gl)&rt7*-+ugkoUH|yK~nBr|~<lym?4^bxGhipb=96Taqk+Djv
z<jzgUMKlEoy%T)tY%z&Kq1;Tf+OF&F+YZOWCPM=Rc9I3-93z#&z1gHYXNjiHByr|5
z9Vtw5C41;btU8oicd0W~+)Q=~$>hk)RHgFf*RxGI1BP$iQSaRtb!IrugPN5U!WE9I
z0}HOJ$G0y(5|#zhLfB1AwT-37WIy?#mp8@d(~Dd|HVis|b}ZlJnmW2cSs4?jZImIa
z_|?m8$klq_6ZSY`7F8fYJ(-8_!=|*x<!hqfH&3qCnJ}BC)@k<<JhLE;xFw)ml2CJS
zP@KmyKpzS>+i_5gi=yg^FOYZ;!t)phVTWeG@D&Asm_+E_--PD(vC(bOf8+s%o+|W-
zr1QTcoMBnfD+?X;??NETyXVdM{t)7;5sDsswZgH!cYw|J%+yMNlXo2ubzT`CZ|y5n
zE+3j(j(#?CUWF16086R$6qwz6izkT=HlRrB-kvQ7$}jE=RlE!a*6h8Lv`c%*njA;9
z{#z(W9F1+(1p$QKvoaTzo2neZLSG6nb?uXh4_6Ij;(l>cMhBb#cc{PmscYK>+vY1x
z`O!tnvF$vF|MW!ibKf=RTV6WjK`at9<X2S$r=oh`VF65#pzhm%45mhE#|5>H{%Km`
zo>Fjx+9^B?hKNzT;>>5uq7Xbokd!a^Woeo*aYv*=#m*&Z-xw}y3KZx!lul+{0Z4XS
zIU0R(VceD<+E#&+*;H@K4?cA#h4NwqEWB4;LoSpu>m^NDE<Xu7%AJ)jX3NQKk)&;d
zd4UjJ92WiP8B=wbbjTU<5~pdaWqKH@;;@M~2(473P1QYTF_BU_PiJe$&?-pywYs>t
z85D?YXtmOZf5bm`L2m|pH3!pByuhk>!es^+me??m(U|Z!w??n`(wWO0p*{<g#iZzO
z2rnU1yYI)*W<pFDT}@7H7iEP}J6XAQ+;!L}U>Q28uA`uBnQ2<BeEBKI@RP`*guz|E
zg%GHEJL(B$veI<`Od;&s6K^$^pe^Cgv^d>1kfFsvUS3A6nE<TKD#6nM6khget3?2z
z)M9E2uEMazSR3yd8_m}1W`!Z)>Z5hYIlr3Vf*#Wf<u4vZU7~M~DpZIrz})qJRgylD
zt;t+-lfRxS$Q%ptcg!RZv^t(Ll&7!*;reKAQkXqgeq7#Ppj#@MamZ?Xxm=7}KU(RI
z_{glmQn6v-VnjwC0W7$FeBw2<y!^MJ6&7adI9`|-hKB48gE!YAGw4OP;JSsDGDBLd
zlvApnwx5){En#E=t?v0}@^&bcvSn-RZN<pOG*;;2xwwO^G7vuJEn@(SF#tqSU9ax;
zg2}K0J>XsjF@e@Q+OP7>h!g^)Ko<YqdK}oWH%~R)O>Uj8o~&v%rL05bg90~c<s_GC
z<ay7RZdmiLdeJz_hC#HWtm1YuyBZNx65(?D<v@2LRUY~@k{y`0oRzecWx$FZpoh4`
zY5`+G^~KGze^)Um{}bDc#zs)g(fa~0-HH<fihL8KTWU9l6C+Zl!G{Vg6*$k*E&auw
z6+5#1*swm^#~t&l)%5H}=Ea2bTy|T8^k#!wtih!F$@QG=`9|rh^n_UB$$f+la~Vnl
zDV=>mUt%mWXJUM6XN>2r4!)1^ed&h1`e9#x`%7l#f|Yb@n8ovKi4iAjyy+pW1_%O%
z{|m@v6}|Hx-*49cE72frVr%AXj?YNX_}?K?Ml@$)x7h!w{*;bEMNvc)`1<ktiESXS
zZK678UYplnmxo?|;G`xJt(6?G$ucemj-ss<zR^B3Jl~AU-1PEb?d0-g&E5W|;^}I)
z=hhF${0ll06zhu1&k?+{HMgTjr-?PT7C=|ULa3MP2j08%7lYs)tKSYF4(K|s2NYKT
zSHO1+evd^j)bESYmGTK^4sLT(IKToJRWj^n&(4qgVz6Fyp(O8u^!MxW^?6UpXfK}-
z%`LVY2&T)=m$w8QRh$FQZL}Q1_21w0&RZED$x3HxyO{~7zKO+IxalOLEJ8?*Q~CT?
zCG}LGx*jcRrad|lrJHJ~$-zXL7bZLOk?|xqZd;kapUFqZo4GnXE{4Y&tE9d94|>cA
z1aH|j=@J{TD$>cfhNEC`*)OxQBm)Xm43Ff+rTQGXGiBoLK)IUrY|RVdp?Q86)<I>y
zoI+kRydw#XHH)F)!7hqdaa*jUuUS>yMovq`QP=LPBU76E*vZlQ0ZnWx>h`N^myoaw
zqCJCViWGP*ERd*+g3+{tum+0d+krmivOPX6NDk=5x_8A?8L8gZ1eMUOEM5%CCQax?
z29tCQds+Gxcp0fgj7IC6*l{X3Zq!~_rQB-*E?*3bsO0&4MCRkI>?eUHVkFJF>XdEj
z1$H__Px=3kv3HCSC0f^Y+qP|Y&$eyb)@<9hZQHhO+qP}<&Yhc_oMf+@tYoB;N=E&t
z%BZRj?|7dLsd&WzIM(#YL|54^m3w9i-qJ@SMh9dixE@0aki@UFT0fk=_1?~TTRn8p
z#;ibw0R0O*a9w*1>}Vi4#pLB67YmZh35X^{3yT{=0%O^tlxJv9rIf@>Q+)2|bh6HH
zxlI$e^ur%a-I4UGi(`olHqH$Mq(Rk9)d54M-JBqvHfFYKvbNbEbK5+ubyzhreNkd`
zRjXy&zxh*`8gO(<tbs6%e<S<|(dX(F(uJkc7eDR24H<*SB@ItJcEae{-w`moxb(#T
zq67x)Y2rZ#UQiu(ZIoZ!Xm@DW3PWT+jEqYTiJ}6~sS&9u;|+0X-&u0lPUe6ZCBdmG
zvVL6o3^ivIggsn0J`d(!$7iF=P|<2uAcS5^6W7Od5i|`o8c*5@ruA-7EgiyAIYKc&
z1q_3tv#{#%$bNmO7X*slUOMV2p1m|ont_stfP3~ZGMkGf8U9I+i_30(hR*V`Y1z+O
z$Ur=`<*Rv`0lhW(HngDUv_6^%sRnn_olkKD8f{N&E;3KaT3#S=X|?o_HJ4d5)X6_@
za*pwoTqliKDb+d@YMyq^6&qhyEc`2=u_)m!g`(mKDd}UdJ!Tw`XuBJCtQZbe&jb?Y
zV-DTlCt_l%xnI)y<eX5@lWlmuU!&X-J0WCxlmzx<zcIKIe`C2yW*ih81DjkL<Lj7_
zYQ$bGwoleN)a1BnQw&pHbQE~{et3oFbPBAYH42SObZaQ^ch7)ilvMd*!rkNK?kW+Q
zN1!sfNe!e$qq-z1q0G@4Pk6ppeh(F65=<U%ly^PJ`q@!}n^fT#1l2Tnhp0B0(RGg@
z$ozed>T9r0v1azW&0UTydxK&_QI{N=Fp4sXnk_BqBYnLg7BEVMW7*qArEYTC)+n__
z7vkB&qT4^*b6^9}FHcXs2C%MK@T90J>Ybo$hREUWqdd}j@o+GovwxmSnS4jj|AOpc
z3qT&zl-=zR4UdPSkc)?HwZ^>W<M|So0A&OuM20>H$TW$aur;EAUr-}WOTc_-W4H`Z
zHk&gpi>KOU#O22y0V=2u)P2<XTGw$Szc}WXYHCPdzw<TW8A2d37ffI+f|ADR?lQeX
zF(UOLnjePN(A@ud5>r&u9h={b<w3vnIfPkzXM+(w@#jnuzL9&f!=?qV9uq5{B*T1m
z<}Ty!d9a`<CRgK*y$O%nN<ZSnMEo%+S<*_$;UM=`P!qfWIore;ZTILQH$Ej5IEfmq
z>P^+edpv*^*n_7X<`u3&7v+~Kj*ZhJ9^Ru)7|h$=di2?niSlsU^2e}UcBnQf)XKR-
zfUY#k4rL^6Gxa`Old8MeEo3jIt-5x?a`55=licfo?U+gg=`Y+}W?mOiZ+I))fGtEE
zcE;yDn}9i}0)j0cfIl*d8UcIz&#0kY!s;YJmYe5zLI&yJ8{D$>gg`4lbGs3$)t}@T
z6^6%FTSa??jJzg4Obl9=tm`D#Hq0O7q#FS<s3B4oH*i6vKI-yJI@dgL%#;Kg+BZl4
ztfPlfkz7_nDyZAjg)PeThjFWL;8{4sDNivK-Z_2`_Ki3eUhMM%$%f_y!#{pA0TZEr
zdCTzty+}^}_L0+;h^5v5medY;6FB&n`J&%AR?|OrTf8B7wRx%YJ1Uh6jIsIqDcVcD
z7;i{z#k>QoMrt2>!JG`Tv;(bOA;9y{1Bc<5BqB34hdwJ=ex8$mnFrGQ*y4Y1^1-@M
zut+|pFEFlonDzB)e|noJxw2bZ&*0!Ag5%<C5)!Fpb=gaG@4Tp7bU2b=F{+ZNYM;UQ
z>u?6$Q=e1<q^rpTV<(-yAt*N}v#hU03}#op(y>`%p_YEfe1zR-^I8v&@BCW~*zfI3
z11U-<rI1%mjOKxZDrNXsvS=nU?Qg|K=HBV$An_InOG6rXm{iTf5J*oK^L{4-%1FSA
zjElfL>=M8~FijtROj`cbj(6Ege6($0G6DIH<1O9XJvZXyzH?R-Tph>@(tW=*oO?0c
ztvEW`Bc5ETmm_lghOs9|d%h(M1}gtCXke%R-$_3^>wm`wp4J?X*<wZLp8dCRQ@Rn&
zrZ3MwJ!X^NhT#P52xW-(9pU$vP$XhWQPTsL3tZ+^Ub&uCwR<$>y2rWY^Xuv3<>_8{
z3ltd_<Z7qKg@CQ<3zRx=#~HBC?H^FW^Yyf``Rv#uIkK8GfqLwfBKPLbpI`ze4wIk|
zAQawTKvY~*{3I_?jv9U9Ren_V*`01yY%+^9%qoxS-0ki1`juP@wT0~ci^ccsm$#ee
z`;)mW)+kFZ+i~8`@8<Sa?j0<UiI=1uc7(z8e-QV)-}%EW!%L?3sv!3%&pYkut#jBK
z6^}VE!E;^1hLE8i;mb5=9Pds9YCp7xGt2IyoV0&mn$>y4O&Q4AX1JJP>LpUaBb$hH
zyv9m*)IyHy_h(~mv$5VE+>@@>h2b<Azfx+`PeaKA)X%zn0?<0f`pmbXzy^Kn6cWJy
z+IqOmt_T@LHws~yoE{r%x9u36!YV8vsY$npB8_2X8s}XW&xi_Nj4Cx$LeArG=>=yL
zkETF0qOvVgdjPlD4V1Q5c5TTlxr--kvPVm=+mBxAEF1EqLe#}$_PYV_VCz}LjPlOv
zOviDjRD$=e*kjOijNXUMyUjxXQOr#w6{|#xb13eh07#pJ;QGUIS}D7R28taW`7}l`
zMVb>i;N?=&IDFaDyEFKxsI_Rg&GkK_SbI2-f!o_j@<PlvcPn{-*;~2JUVicCQl|p-
z_ldvv(i#8ImDG9mR<S^Mg7}*^)oteXaUD~M{YFdb(Rt=w(FuRwQ9z_??%t+W%5*5N
zxW>J^MWkK2&d1h7giIDoaSt$k4K0;ruQ41FyKErodr?p(6lgP|zvlUbs0XV9WvD;z
zRdNmoN*NG4!>t?~quEb$iv6k<kF*`PJQTHPO%wAER@er((%s439iS4`hcda;XIL7t
zPK|hjlSIq#S?X*iXJHUS9V-UiKqA#E>lGAGB#^bIQ(d`$Cg9ZyFFfn5yEc`79Ax!3
zn9ue?TpH-N87$MZKE!Nrz>_;hrEKvQ-bD9~z%2rL#5f5dQtOcM72GuN&2xczeH~1b
zY^USK!l>Ij+Ata&t-e*5E{>8fOG)H3e1q%a98!s2uX~R;osr#Kc~sD;6y@9<cM3+!
zucm-)^b`A5xuVp)wPqAX=ZwXJRDQpN%NiszPt3=1%HuuBvtucagJiI@td?A)kgR=%
z+h}d__9OE@udH*uj1fhY>Y2JVXmx5Sc##3I2T6la)`pZQm+nwCn0%93hl$mE*FRNv
zJNg*(hMQ8WW-_u_C{0Cwsx9GF8*c@8lQXEowZ=-_WvX*WRI?J>t!UUv_8LxLPGMKi
zA`}<;|5SnNE~CaDzXuy@#XHR)M3Ii{;*GZ#ImRnr?s$1LwuFKixg@7jV)PGcHbyt5
zhu|+?X%%SQ_okq*t|XI*>(2HVfEWA)Hb;bzKV@8nxC-xd$<tyRDwOvd{&-Ojc<tBf
zvJLzjB^tLSnE4DF0|`s#GRQ&ZE7TNT1sT~yIa=2ay-?wj#0EQGQTs9$j8bhzIZ+qa
zbq0OjutweHVG8aIdySY7zE>*WW3?24>}Es~S*;h_F5%(k<fAzRE_}B!%Jq6rX_l$z
zVUyHDRDQ)4P_Hcx9ZU%j{x=$I4Md@nTbPQnJrBi~;~lJt9v6pa7G8}FU}o7a2*hur
zja@~I84}O8OqLf_$rbTbb-jE4EuT+bzVZe){I#_4!ao|>o|4E;OU@Ga)nRx;qs-&L
z#f5QwL+va(r~^qapH0Iw&U}BqC0}#c@}qmpK1D!S5ddtpi#l^ZT?l?}l&pRHQ-sG!
zbDF7?cnk?RzwRPm83W$%P{qa97(QKl##mcoouMK8;YoWexG93aC0T#OwMYfi#?8Q{
zBF6rP%!#df;2-Ak4z5X3f>)?=>5AGMRg^jY0N5-!!3iv?aveTT<=>mCS-kq&-XwY1
zA@^@ewKy0(LlO8K73sM)$#TapzGg8q*Z*A0WtN=M^glB;&6DAVzMCYu>~dAj8}Vf-
zn^-%WBv0=;=c%ffjhif{)vz};p;QMp=k9S5et=Fw&Wvp5QKzhtS(VAa1;e-(ucD#(
z7@Kh>c6{<u<&Zg36tLh;XPEb0wq^!5Z?!&q|8b%UK<M}B43nHjAS`T+x|A`C@6h6(
zjl5q=F$fMBsIbY85`~i{`9|6f%SNlMkc%vUf^vsg`~CBoPiAFetaP&oD>osa!rB?$
zG2M|-;AcOQGyD_nK39<bDcwx1!bI7tbkBJwU)=>njq23vzdbaYUl*S<h6|Gq=@OH5
zM4l~%2e|A#a2#+<<rucTUL2AYg)FaeVw7ekIP{drHV^MrSS9?{x9z!A8Hc_LG((<o
z7r$IY+;5wD*d$T;PktNQx|o|eYQm<3j(4dFq(^_2mw%d<?|m>`k5!hoH_#h?-ER`k
z+c%-IaCaJ@pp_SRiyywhm+mM2a~{wBznz*G7})<ir>0|#=@@J_n4V9{6Hrw|%7`St
zzP=h)oxD~sPihaUj4TJUonQG2TOT7&>)R<X2oH?dq4+U55)p+a+#Xn$H|y7@j!!2i
z<}gS+kkHjy;4|>bUj5`T9|V9wxRBM`!Q9+xk(E;k{Du_{Z{baWU7z1Qf-{IY#5`rc
z??42KVanlijHi#5kGz)Rg&1ev%P;+KB=BR{8VCE&UR~eab*QzBvGR0X@7rVN*IS#Q
z*crDNU{3~pgSXF5T5p285K_1>G9XyKAIzScpR;yWjcqP!YCoz;M$cY(w1#_Ln5h}|
zcm)j=4vflyi-5!h2&rBF1-K8Jzh393F4=NzvmQeQj|I2p;w7I|gH@G{mF;GyB@`%u
zdst5}3fIzv>ncftqu)Y;0#GxRBHc}Ny2naZFRu|ZLb#=oL?Y+j=cg^qHTg5ov<eqx
zoJ5P>3CuB8VP)g8O9S4qnCoFIhtR=}K3Hs&31%6&Wuf(}J#1U$-ZYO(xrr`(xXUxP
z=kub022Xw<1^zA0<)bjqQROwAjf47GSPibGaAcxw2eTs=Sr3QTcZZvE3O&mBu`hq{
zU;U$fEUAl5C1j?DAht`aksK_MwkW9Fk`6S`G31JH$i8Jpw8h)o*piX;XX%IY=dp7q
z@Fu2;t$L68r?ndNRAUKfeMZnSUhOaTPc^K#ln-@RUsa5mFbUg**3p;eP&$v?0@9J%
zUJ@DTC>)jSm*ygP#MBQk#UevQXUQM(u~14>j$JTbU%K1qv(<;#4g|La$u-w$?OMyS
zf!~(<uD9vGk3?uG9FN(+?`Om@4q>@33UT^XQSmCFSAFa+PkJCd@P>8fyeUIlYCCDc
zC1|#itoGYCcU|{fR)U&vA!Yj54^&56=q}qi=>_y>P_H~{%HoxGnQAytk`B#TTzTjV
zTpe%T-wGNPZg`L=N*4Xm-r5g##RVZmdEUE162xj!39uG}(2+FLVlLYoa}WujZ$>X3
zV=vo1ZTmtWx3oGsb)n+@G3X@Qcfl&>%tE38Ww{UJy>H~Z0^u%KtM_;EJ(UzMJ1J5B
zAli33Sw{cu=$V0!zY$y-v~9ojihk10xkqu^8Jf%j5gKBXDtboJfB|h6F<f{-vs_Xh
z|J$Z<0A@*&yZG5Y<bIG|#9N(uFm~r4B{|)Ice&s|_OcS~iG<G#t=K?CUDv~5L8Y#r
z_4!~tB~7=JMNg}3Ikbq<Pp(x`tCGh_PGFOGSQ=G7EBAT`>K0#+lCPF1+F6u*R8JD3
z_BdKOepvk3ooE&Sn{9&^yLB@$wg}}d1M=9QZ-3kUSy92hO!1aTMJI097o3Uw*4V;s
z8pdGXJBW5$*+*~&k`GHo@1oKEmaA$=!1))Fw*qAltAX(41~!hU{VPclCQR2mp05pQ
zQA$H6{2b$lC>bL!k9aU7IW6W8?0*i)lG1d)hg1x$`&7wkHdikGU*5$}pVW+dLv9%k
z?Spjk(c^2av?E3G?4PG*g@!QdGKR+X+l`l~d1AqR4`|?(>0rzU=&4M7%hI4xLy4Ra
z!|T~W|0pH45{*x3pYY(A*^K`&0pa-Hh#E##=Kqpm_>Ud_w;gtS^`9@DCL*Z<e4c*Z
z)eTgfPQV7J257p*G|}U4WSoZ%&&?#BYXD7!)g+RL!qPsW^B=li-J|1K+t%7vK5q+0
z`vGEh=l73epr_!XUxM=agJD33%KeAu&Ewl!MJg#@whtyn3mM;-9Z!$pOE%aX4}fO?
z3(z(nN3b#xrN|OqsB@Kt+=5ryR_~Mhlo3yJ7%77Fyhaqq@6lDg9V1os2>tgzy-~S7
zUftiH)fKf({MfWBegM#^zCSM`dKml$I&Efh)U2tyzs~Q+*R_R9oah&m54W#+bV8H6
z4S+mxfW^&W*?`R`cuT_@1kAq4Fc7<>)ryi%@2DZ5cygT6e!6Cg-Rb@`f2O}#7NU_w
zJ0&<TeKt3<jAiAtWTvsITc0cfmECWeZN)>a7S|~!YZ8>GN+V^(;X;m-No0vAP0t$d
zvuz5s{t~drP#`r~xg4ixqd0$572rO)8i&Mvl%}t(o5L9WIa~J9U4N9po6sq6lh!GO
zu=yPGz9P^yR_HKiQXp}S)HBfDUAZrA#OSWkI~^s*XR`B(8$^BfY{4u_7Dm09q$2Q~
zEexGurw(X5%!d|GH-?AXzUy%9OF|9%3BV8+<<XyAZ(*T%7j(GdOw4!YU@U<YobZ@y
zBaWrl;I#H&OwBiV)^mEQDdsU+fw0}8NQcMt(sR`>eQ%2*eky`y@v5l}=!^YWRuR_X
zuHm2}?8YpvHk%pE2$4$3r<b_aJn)BiaE|@sFFbchgQIYl;}7vT!Plw5);*#nP{s<e
z#h6S4*;63QP5zG2=$Y^BGw3;9%c^Cc;FAJ6j>o7%XEdqIpG{4adDmu4G2_NF=H9F`
z2UR^t9}eY7?>3_h(zuZ>aF<d}0%e!1c#gk7{hc*E;o%WZ#}~A2`l+F7$a-MVb>3~n
zX{go0)DIyd%ds|-etMO<ZyfjG@CgbiY!#xRtItQu(w>P;TDwjME_KLa-syJc{byiR
zgH16I=>!%yu-4OmReZNK&~_g$*cefa`Va5(amtv(oyA`&0x5A|Zzb4t2WKPniN!PC
zEYqB(8H!|3f3T5eW&!(2s3E>!$j7mUhqjUL_CTbSJrV<+i7N)4(=Cc(`Pnf;@1ZZN
zdXi757oN#ij<}EDQ^VtM8MOLlxCx%I-s`3qXOE_4IADP`ck;X`nVHVh$cPyXTBous
z>J?t*mo7H$-0s=8*P)YFH@omhu~WgqI~0&8X}5Hzb_s2+$w#xm0XMd&n2vauiDoF;
z;v%h`&DqYBU_U{`dpwfS@B8DZ2cOZ!$YbphxsskuG3txp>v7iVsf5c6e6?sy!i3Og
zM_~{O%0)4v`e@opyy$4LPMO%9XFXt$!zJd8Q<uG3SqiS()bS6x%F)Vz1`D-@B}<wV
zqg#t>r^^*Dv*&=e;mLIqZlEIszC>5^0UUhg(rRo5mfM}2X+K473CuTw+dh-M;=4l(
zS`^G;CcbF}FGt-Nr@){T9_xZ;lBj{@<I?F3PzxZ;jhkk3?1OraX2MsfOUFdZvR7J!
z6rup?=ccop_22_G!a?=z7OX2azziImqI6~maQ|V*YO>UHgTZ8^sM(DR(nQ@A1*_;+
zZyg>$4rA=WDy&qh2?GL&A{*^%6aJ{s2VJuvUS$GORVcFS7r-?F@=Sx_do`vn98qR9
z1%w<J;_ovr;U7DSe4XgvR$Y%04<orDptJLBM6Z2|Id#p|0I&?3nCZNDX7x?yAtmNB
zzb6VngN7kMfD3*Sv5#>d>pAhtim^Fx*i|sttN^g+X<^Jv<cyWgl|D_)qm0>QoA|bC
zAqETLxY>ca6mRA79|;L#^@tT<Yg+O~*lZWVj}h5ZwQFMb{@y|qt1M^VRun%1Lyi~^
z3VlkEC;oSVk@q!$^Y9^sK#v_C!>PO^?GTn`xggjWmcU|Pm!&%*aST&$dJ8j2!xftE
zh%&-wqr}0o{h-FW1Mg0Sz{`Q>j`RJp`DmN2C(!u7FNsQn5Hg3Lh>LyYBZ$g);w#Y@
zpP1`0L{?%kH<6>fq5iXZB(Za_VoWc4j<i~6oPDbne;*heRkT19H<iMpL_i~!%Q;zo
zEiDei_b?#JBB#R`ph%U0d%&}kk8y)wq`*F-=FCTeG737JHJ_<*s!%Y^zkW9h<a#eo
zop-xcE))T}htG+Ai?KT(SboR1gmaKY7+f=qpv1Ks6QQf1WDA|d6cOEI6q=e$DCuYs
z3sENc6i#_YgVGy(>iA6*0ohh{P_`PCO3(e<equ{=nmpz{0QRDgK%)|iXom?ZvvG%Q
zDR74~S0`8=ltgW1DH9ix&ByPYP@urdhbUqQM5rQh67{2QlzJ17^Jhj`RK*mE9Bbn=
z<d&2nZ*?q|k2(8;?b^MpuAtA(x>veGkt{=V%ocY15zVpv=3A$^lM(5Qn$d?<v2}*k
zX;ZP0-jI|o!e&*uNn8ZOv=GR4!_Tkc^#`83zg8kW_}j^f82C~&2Vw^>w640x{Rsh#
zoJ};u4$xz>m3y%$CsrM8pC-m%CoBT=%lB7xQ4?LLNAkkAd%|r}-=lV`O*QnlH0w+2
zR*!GWA*VD3j0^jyIrW-t=9y-ea5xfBc0M@C<L2{kjqM3j8Ux{xJV%C&K@x?tq+zrP
zXamHzomhqU_Z@a|oH|LSrN=v()p}BG+N2R!-YJ>^zP5J#Xam|>kuRrAxFXC1c{xt+
zVd$@r?#H}4kHBz4;oy+iO+B5wl+JO2$r1RbOVrd{TBuSFF-qv1W#*|NEcip<qp_qp
zI;WKj=t3Y$ALhPm=yEsks}zEM3Im8ZO3L;78q7A}a;Q+wo&a%~p}gt5X!2d(FbF57
zvgPHAKfV$DYt=n8-+g*-c;;(BjPHQKbtv_t0)O!8O_=atEvpcTd}<%o#fNq;zRJWq
zQ#c1l6G%*R+9FaVp-NK+Bq{yJLIV4RBoIJ!+ptJ0Pn`bPnDpTPC3IwnQSa_*R4=oE
zHC$T}jm^W>OFk?b3{R+guv2_g%g=O45_QO7obtfierpM-oH(?YHRk;C15oef_`Q$8
zp1_(rU*9-&LlpE{^2YHt%{@#kGJ}<*#cQjLObuY{RoECGP~nEQaQA67F_DolqLeNF
zs#CY>FOAUN-1)KdkIVq=YuvBE_XbJ%Qf?+2VUDVx@bzO~fxgzq(Yf5(O_Y($N_uU0
zHM+Zkyh3NF=QyeW?OysE-}$^!r6z7_F1+Ce7Z9VJO*T?$$H0~pK3dy*lH|?)WO4Re
zs%6yP^pH63opy<Z*7H&ewAgtjXUHu<-z&xC2?v145S3xIM$Eg}`#BebI<V1Q1)Mz`
zP3-5%ba^eD&B6gc%=;;5O7K>{;XF!!($|{43k>{o4iudI@a=W4nIl*2%-;)ANqJwf
z%LanFdDQyuF1h&WNy__AT*lhF;{JVG&F@G7>AUx;3ZUIK_<OpmRcyl$7sOq(dEId-
zw(RzUsB9Zc?wiW;rFt>Cy6FpWy*|JEpF-4sHp&0HM$X2<_TRYyXh_8#vcmM7s=dKS
zsTP%?(?i4+C4zcbGPD+y7Ic>KIQ{<SFD*8Fi8D%+ob}t&Xra7^6Y(iHyz!3hU*5d+
zd^)HytM4<b<Lvap_R*321{{eNf<qHT1?+#1-i=l!nZ0p7h$vDGryl>t+<N>L^xdQP
z=>fC?UFp#Pwo=$Ae8F1uNcF`1Zf?2Ee8axFCQyj$uje(ObpJknuVzm{N$tzLVcUE!
zZhqe6>Zsr+&yB!~a{9-F^!i*4Km+}WdbUcgvs#+%4Ylpu^(Z&4y20yr=+l?@+mz!M
zENXTQ_P0xZUj<02v43b}-oF-^j|pL;q~$Hfe{Nq<2D{kS(DNrm_S28}eG9Z<&2Igk
zHWsn60)+uY6<(Apf2{#pGT6mi(hS-V@%i-hHQ)06`;tAPc>v(4u#av(V@l_GKMLTA
zs|=dGOs0AL?XltB2$WJQI<|1jrbv`<J(ANLT2!)MmOL1SQJ%a!5iRB8`0BU-9h0(2
z?VG~j9Rzw^UV?L-|8%^p(kDGX2TD6o5>sGBrHnu>KsL|Pk;^J1h<H=VhuKQ?zJTeQ
z!cCIZ%oABziO^C$9}UavJ>u(n?k<Ua8QO!HaRZ{*ix0{HXP$P}4pO4Di+zu?sS9Od
zLHf6qQk#?p$ez9?Erp;F#Yt^ZA?9>LwlHM8V$!M9(Oy%QoFTmkBSB2~8YiTEZ<4Ya
zDXVd1w{qW+XRWx{z8mjV>;vw_7$q(Y-GPst>l`|CbQK3r`{{keixaI_lG@sn|J<`o
zKVTY>a9q9?(}oXwREJMAp}`lj-8u<dtUnyqBbMQ}8mLRm%lct24E*?BOrzgmrvYZ!
zut4Teux2RcDeLiI3KJF_x?-!02x&focx@mefJ{Nr5}MRo!(($)_$CP_1yDI=JK->s
z$^KYMTDJ4eWYRuz<|ruTZxS2k^+6KZ%)a5yOq6^fa^H9zjQ~f7!<}Ek*x(~ZAa?8P
zx<lfI-)<%2DS{F<3WYuhl~Zs+f^<#|4P(S;wF-vF$|@ys)gQ*3I`M-N7rdF0DWs3?
zn4FG_?P-{uF*QuweftqL#9(j_c0E#C(203}z94I8D}sDV{Vv`l<AR2ON3YsDut{OV
zqqUEkQDC!^$oM;Zs_{%RgKN}~t;9EubLb`?ldsYAA}D3q3thtiV0<C;WYdhk{;T%2
zIAy_)bW*m3;X3+lFVO8_>oLDQj*7-)U<(P%`F%J-i1dsN3l9l%`X0)2`o{>eW}}>r
z3?gobVW;lJLi;Bb-GQ_MSMOO#PSFoPYsIcV^>PU$R+<AnbYp{436om8&qaIdzvg81
z;t`I`<%x>2Zajl##@kgF2F#wfFlQmfB80Sb4<V!6>1qhJ(fk!$ZlLxF*si*{GSt%h
zweP>1%-13yQj$SmGh~|fRFub5nhXq&iR!@)=9w=XRf6_7^@lG<f9G1-Ru)!dlKlT7
zQEXa1?pX!ob4!-si>#;XX-`m$IW1ybke;`aMEhHgiw)nnXu~u(|3GPvm_cP7nml5y
zft+?0%o}7K7wH&-tpmBsdPj8;i#<RgiS7r@yqhdJny@IddVAO-<2)AOeAAt1SbYv%
zTaM?Qf8wwj!pZ*0PF-2+PDqYIhBGi3bQuz}1j*2-)m}ZL$+8bNlG=wlFAwUQ7c6s=
zM|SdBpScV%SR)NlI4%j`$7(D%+Q1#&2_gV*NU9OQ-As5v`JI07zrh&75c_ZtQOi@`
zn4p%`8_9s|@YdFo@XIG>I+Ps<)&Tvpj;_61IVCe0P(eptc?X3TI(29QWqs?>zp1`<
z;lUHnK(RD{wB3)9v=ngHYY0P27q7ieXod5_+!HC39bs`ef<gE3MwU585XK3?8!khT
zApeLJ%rI~fnh4pHf5br>2cO>q;+_!>s;(L{KSv?3!uyU<jeY#Upt-s)F=|#84_V|n
z$Dm1m{qJO(;uXKp8E7?+NPo0ijOgQSzA;~RE`|EYF{eh1NGFn<u?%Cj#7v=VgD5-@
z61F?A%Rjfu_-duW<=UefvX)0f+&FL&oU#%HW(_!f!X*PGyxz?nCJd#E?nH)UHeCA^
z=@3(R!8gD>pL$c16Yi$1;JAfV7t+|2r!+WCd>Sc1O=L#Xs6%*sK`nKXX~J^<3#O54
zf~hE}u1lCJT+WEDuSt7X<rbn6ZR<C!8)xKLY#ntZU{U4D7^t7T9=-A}+y}}7khFb2
zUrd+s(4z1ZzaA+a0Es+LW|vc;HW{5xD0w_O#IpG(#lwPxCzAWQXnSs#6OdXS6qlGA
zfNT0hdq9UYmlZv_Z1lzbRB^B4(HlBOqY~_Y{8uvkZxI>;GyQ*8&e=|%{wE@b@w<IQ
z@$sLJAS%b#M<7VI0XwUXY)9om-OhQL^Zpg0)LL14k~2AA$GGaZC($AiiC;3^Wc_7r
zad&6$=JnzrX6Y9W*~?hHg&hN>?$s||bVLS-5;N2FW%lLOxa`6q=NQuz<_P$TbvPL$
z7EI#ezX3Q0W#<M1Iu|$>XdkW5s>YH130-idJ?xH(O`(uNSoaf&;QM<1_ze65rgn*g
z^-_NCAGIc5jVzH>6&5c<;&}DbC|;2E2S%>g`NH!6c~64Ius+7y6?I6`V$|vT>5Hrt
zcf0j3pCuY(l7*D{Sq&2$37Uj8U#sJnLkxW@5-sT9kVj3oFtvv|95rjd?Y4iuc$S`)
zFzZ=%irMV4(CE(33TLXC!byV(sfTSssXO&*HRL==+Endx{p}y3ikvA{f6c@O(Zl8$
zS0${dyTIj17lH1}cI)ctSR#yp@LA1ij)|D=VPhRfW}e!aOhQbtM2wIbILjfcsA!fk
zUJs5>Bb`U32mPOk)>M&A%a%Ge{q~KUNc@kqC+B4Hchw62)G^3<)j%Ad<<nByMn~zB
z7SEfmnL+Lg$3<6ff>p$i<uMLVzYd(?dSmzerqFYl8=n2S@Tl!;mI`mzrA>Vv(eX?G
zh3Q|nn)~ww|9kU4D`3=_UM{x0-DNAQtD(PSA~ESvv0Go?hlF|r`iFs4C22#`v%VR(
z@V;jG#_vz<)vwl3>bL2m3<t1{Q*#NB=YODU9p1BgH;pkmxzTG^NS7fJFei)H^TUx2
zZsiy)#f6u0S!G0dcES2_0`37crPR1j2l*rq*m(fdEIoFRSEp<)udj{@nG{xBH3j&X
zC0s^C7Zyy&(l0|{S~WrrWiQG~t7*n(M59wJlH@JNxqNa6Pe1yMUAbeMXh#g}pVjvg
zDuL3DDoSPXUy&pl3KJU9AQS}<6S3k5#kJHIBv!N)8_=Xbz1*HwWLnm2CLda+<7=~p
zZBHjeoOgj*o>>N?7v4DOsot@qI4&cfOo2((1)>^@PP;(kD6lU1rMAGt2{CWcs<d0l
z{DnV_j!c1zNH7w~csGH>DI`n^Q`(3L>YoERngHX(C(sv5fk?C#zN-`)Lqy0GZMQD%
z>#=GT9}OiZ&|L#a8?3@^K%|CfPMy^|&VD#{UF(_-azO2fh}L6Y!N`k`8QIGLx%`m{
zv0__+$Vsiq?G!az0uzzij;KzL;tjdjqmQ>X#6Vj{Bp))#5A5ZiNSRt0VB?Ho5K^_z
ziqy@VZJArf+(7gLI{6L_FjR6YF>MDcDdV)`aWyp>{JU^Skx6NihGdS8A1bk1^^ZGP
z9{NHJ7#-rfc$F`0<@8nY<aE-e9zB4Ffeyo)hHxQaIGU2Yc>D>$0(${`alwnLVG7r@
zzG#22JPu@^j>DpfYtN<H2-ln(zXbc8(-)`jFW2^03trLhCrMMncCmYtQYPPu!LEYz
zXcPLKsJ|2!Tn6Zhjk$a%{iSN4DBvksY+_drR@G-@-&j-|xO}3QtM}DzrEg)ny%XE;
zOp^1uw>VZBdijz)X&<Zp(qC<DyR&X`Q7ElGT>Vp!{&*jOfz1pF%4*R`=yav@_pL8P
zz@5IidGobOF#E04U0WB$WfLzk?mn_6%=2eNeN{<c9oxH^<Y4beYPx6KY3{{{!%&VL
zPNvkhVB+-sK+%!betBf}dP=RX!Mh3`jE(F>hxr?oa3S26`v~WD_{4_*uJ_$AS2byu
z7_jJ*V}Xf{Y=;GlHMkJ%$H!dba3i%I5%I4F?uegR*eM0L0I^x$klJA@uYZI;JptM&
zeWYZWW!v9_?e_h<u`KqL2Ci?5SM8Er5HQ>w9G_>6mxy1$C04W3|MV$mVEn&4-TxmB
zX-xmeA&r^ozja8nw#9Bo{JGJq=kL{q4w8a14KcEZi0fot1?XW7`{$Cj5APFC*Tb1W
zI2roSB@MHtDTSuAH6e>1ZqKa=FMcpRbapBZ+Yn;ujcl*=z_&utLJ&%20v5DJ(Og_H
z*$~<L8>}haFj<ngUa~0Ks9ez^;Rra<A@Oup5?Ok^CxDQ``A{G1#rXn)B}P-^YqorG
zct|i?Fr;ijSe08j_fiO>63L3zKTVNEgPwmz#b`>&0XIcR2<37$4J)D?ZW3|waar^v
zMBfdyIA6OWx=P*I6sQy|uyJ|>Gi0h)HG^Wi4G6wi5tm|X$Vw`j*DY*3>xjom>e&^6
zrdYfj%C5B5=8lm6h_l8mto;L#+62m3TYWmdgVJH}s`8NmPU@6rcvZ1xM+ruiQ>PfL
zvZ@0Mkx{R^a0_A-OWRGA<ed}SFryQdSQ?@t`V!H0yf_9rOxz#7cx@M&C=01dMQ$0o
z0-+OtI6|^#Bl&M(y_N@~!sP@4eKH)N1lF>8Y7V{NO+gAk35T778Irci2mpKdUmYQ`
zu<{!JJLNGpiD8oKkm<K8%Q<=?1Vz}uK*bc<Ae$Q}raOzdQ2-!`j!?1{bD?mVn0n5p
zA~Gh|g3z+%)-Z2b<?b-sqXiM%NGdcV&kBTM>?i`XM+`vm;0p%`3sMIi%pX<5a$UdO
z$Thrx8e7SV9>Uo3?o&5?#*o+T7ZP||{RHJ>D~DkcK=1xbFFad^XDyMG&~=qinVL$%
zLg3wD|FN9Zh3Uv%!v?Rr@7L?l-<_xpwZnU(6};{2>8r*4<u^W=&p{&(h73GUyC^>M
z-;Ew?Uv9YV(AUDpbp{C+JU5CfpgPpx>@~cXc|s@h>q@Xf{ANBGUg?e1X;%pqj<t}!
zfh35k$I;X`Q`RnCRu@B<pPuj2-XCjdM@$OozMr?7kB=iM3rLH;?uJHrIC*13wK{yT
zvVeoEw_}COm(lTOBM&D>4_=D`?|k#!>=~27^S1n(Lp|PEAI^;I8Bv)tysq>p)3l>#
z&}maWQ}@7VfYjuhKtxr7dU!g$mfMoV$HMumHt*LDnNNc!ch=U1kn`<xI^s3@oam9p
zP08G4sG&m_2R0u3?HS7DoNw-qHqLK@hjuV|{|V(Y5Mt?jok6GblG^y*Y|{XXfqwf`
zn5i#XOq!xCDPzLnj)FuL0-hM$yhcdqE09|UV?u|ZXi?HO7)jOGT@5)n)w;Ewi%`B?
z9hnjSWxk!>=G<G}856cvUz<(68Wa56nXUIDrO)$^n=6YbcOPsSTd~?&S~^5}@tajk
zfE)es?9F`LEI$+ehqr9~b^y;UJA%@L6@6c>3bNmEgTZ&ayUaH#1P+Om;~|m1kc4Uo
z(28#pquE~1m)oV8=;(#J8@Y0(_`g0(p3r78bbz+(l{}p}S<R<=IZSqSWGV657wg8d
zsf-V%#LJP6oXL-qWhk^8o9FJIt>bSW7TlcKJ5m^D@S*3!27BaoypB%EGsBb=Y5|<F
zu@Hc6?$gg18iEE#Qb!<nM94fyelmyf9*-#r1xhUw5!%9J)ANX-EXZ*X@k-ETiWCKJ
z<Ghztmm(M8;Gg>p16m8}@`(Y%>ds+Q&|WB^<)UQ|^eVQUexYSZ4?sRWV9~S;r%<If
z(uIPT%v3&HSg1M$3NO$FM{`n3QuV2AsYR(CP$xM7?WOv(T8Os+(#!!h4zr|K(5z_J
z<f&;qt;u40xb)#k<ar4g$bB(L>?mq`(n3uv(i98wJx9;kFq;Zn2_G2p&qfJxuBON{
z>kKc5RA#2ng?$x41X2L4mOEu(8(?5CWYemcGxmNqgCYstqky1$eh<id0FVRr0<~)L
z4n*S(Dz#@T^8N*+Ym5Myh6;&FO=OPPbI6?iqjc+je*J)EDG6k;R8Xqi0=kp@cxQhq
zI*g9H3O4waa1uFb9Ht=-61vTZmT{CsLA(FjCn_LJDB1^}C*F`F@X64MuA+ts(!o7b
z>SI8zFK~LRnz|Df5Rh?^3(TuUR4pn9+dLxy3el(0(KIFg3W?XShC2R&SJF#vcoIS2
z2_ro50yV?L)0ZV{D`cvvii2b7EF-8yNlytW3*5)dGsDW00`wK0VijN&QgfPV*dbD#
zeBT(e)S??yY@!;3r$o@7N0Oa?s$N4g{of;545~(Ekp~n#-YQAnib7ujhXQ38y;IxG
zJ2B8mfzXbBdQ2ffY_fHc7JZU}_qkJht_dyoHQQJqV(r-Znyb{as6Bv|w7^u@g?LHJ
zrO0$S7k-ywTPb99zGWl{XEVLaqm*zBvyu4YqdWx?#!>EhXdUvJ2LhrzWiV0s$4vav
zIoy?U)1JpbX(<vWNhj7yuqcM8Y<jL=NibFT10>j2a8SghIN*dm?{dtK*or@@Zl8@y
zYz+(qbQ^v<Rm;UJa>wL$(L2f|;RwITQ>NE7q@ig3^ax48X^~+0%CeT7KY5PVFc3~r
zWL(7T6}&8<$|0^cr~A4`x@(swc~(qM;+WDlb;40w^0-`X>VF0wgT!~Zgk&eVWgU-Z
zzm4q*^r#Vs;9OxPu9<bG7@KirmQFpge><vGX-`&<*>IiEEUYYqpv(A8HX1pIYDbW_
z92#XH`B&k(VW4n09t6vwcY*N0n8B^*0s<5ozIp4gRzR3Li@;Dru4V34R7q*$xKljI
zKXF6+L|%@ah_&o5fdwt6aIzpjY<y^0h~a9f5OehRIY+&(rs*$@R>SW@vC&EPOc!KY
zj8s}opOX(>nyl*UTj(f`hHveS#=;pP6g%pPtvgCxKy~!KvKy{!a7&<)gp?iCHzq%1
zr%nmT?Uh$vuT1?S+OnxYN)K5iE|al!@zak5XI?c~v}Z0QG9xIcdtJ38P7<Z+5}H^5
zMalxfvVL?xiRs|Q3)0%4wb?C}bt5b#4gPHMk6R5*%<0&tE>yR0=jszhWVYNAgoV2S
z*C)d0c4RCqufwHS9SamQyI9`TDemoNkb>lzmR=+<dw7kT2b^^O6djy~8(WD!9Q7;A
zl)LXOCLTNgyt3U!eNBuR74r@k!N?we>g)jEaLk_M27{XZ<v+wuNmR&-I|09!hdl+n
zZRXYS?5|$PAIqm(nhl=VRl*7v{3@ThvVY$F;zG(oh8PGv#y}hFr$NuugixD0kn5kd
z-ZS;+`llMCTT)w64O$!&FfpC~Y-V8h+P-KI@B|(P>V@8Z0k~urO#V~6_m6P;zslT<
z|2B~QS7q*hQg9m_4y1n!pCkA%6FYJOh`-~sz}}8Rt)lo?qMMF&V23$_Uq4;?3F#z`
z<szi75aKGNk@px;5Te|SN^rYi9A2E<y!bn~&6r&fDP34Oy>LPJBtMD5g^hvaNV1D}
zUyNTabj1#Bc=ni-%_GP6JbR10Tjo$_0%rndq#kXLA#(-u1@mGWUXt2SyWTcjChp+x
z1knZ{`iV0XKl`tzdT@-?+~%cCb+m9Z*nqfX*M3HYATYE9oAf@d{EYrAPp|Jc^-Xm%
zRNMRLY`Q0ulM3es`HkVv*~=OuGk=R*GMRl`VVX?TjH~HYT{FNznQcvsAfst$-;?H7
zx{Cx6R-Vih@k-l!=p1D-C||%&x=RVSh{xLS9G9%ru&a~f><PzLnrh69ri?_=d7P$r
z-7~sVt3H-IV~eG6ifQ&cD@jzYYnKE&ADKn_U0fLCxHKob=rT-q3$|}Ofei(@mWdb2
zMu}Xiku@_Gq}UYBXCo%o8^&{L!UwJ_wS5}Cm5>)Bo#VLk9hRaH51D8AXIRskF|3uj
z**&ZyxOhJ2KcZEdFmj0vHKp8XKHKoC@fc36ZDpR@@XL3?2~_Bkd7kgx_i3;T>=51_
z9NriRaJ{h<WjXuy8P@iDg)GpDs=@<NV#c_<M5BB6>72Y?Z*BE{?~mb;5I*5ulLHYa
z6k!G@-7FEAM+laEW?$w1P!;JTA->P^gXa^S9uTp!cwxeUxo?;iSJ>>ft3c!L6-uU6
z+4WQKz{|6s;vyNE|2g!o#g0^cL@<|4#PS-Y&`7Z*)GK;Rsr~423SQz19bDbN<Iq~>
zJZ^`M9T1=2x1FxQ$n-MKOKq3%MD!NcluG0W2d}MRWo?|chDvl9YYa&WoBw2&7Lq}x
z{j*#og`5W^Xik;u{TYJT5nf;%Ax}+{)Dl1pw=c2YQ0RPmtEbzr#s=YC!7G-=wV;3y
zcaz0B@Y$yxj^Ed4`9<JL*DZxsm?iL~Y&s!c9I{$j3EUr-8!B1M+}!DIlRL`-o@H3g
zF^I!cUp7Smskdbm0-K?1;-|vTyjWDKz2hVsLrC7L-Im}kt=BbTWMfpNSWLKHJaPZp
zeLzoam6*i65|$&Sv+%?|E?0x|8SkSU$?I$n<0#OwoL2tIJ`<V}M&iV#D1z>-1k#Cw
zcG^(lfSrg_#og05shfVrKdZd9B@MOstPG{Dl@ZC-cneKRq1h~aLPKOIB~>=bTo_%g
zk=rhO;c}8xSmxG9*-CD9&2o%_A}h72wWrJxyj=wa-H8g7%)E(@{yYx4OgslUaNfjE
zKwDI{&Rj*B<!)a~lEvW}1$nA>&UYc-mjk*kq;+If^MDuyU~G1ns(0Cx;69`yLZYat
z!8R()#chC&MjqI~VU-}Xon(y8Rn+C%h(@|HMbiic35TXLFmYO>lAwf{e^sIlCsyIW
zT|6xFy6gePbY1ABXFC>^Az?ze2C2DxX&$6aTfQwXZzfF9^yuxIl~Nf2pVEHyl8_Ay
z4ih*%j-T7X1+o|m$_0`_PsNe!D^3E5)9<;Tzo6#}rw5<wxpY<g;|K?Jp4DtE&pt$2
zmIb5{=CPrxV|=Ir+_-)GHxmiA1i}Xv(xACrKNTd=uKdsM0;joV#j^lCD!U^el=q$`
zPfsB5<_+jmct)w~vMh_S^#d7EuX}3)YL8h0npU#DM^mzd1CT+=0LDa2nGCf2KfarG
zG}2Ikh?7&{9bwdTRnVnzM9g2N3zMzrk{f_lY1Fte^PNw8$9hM#H0V!cT+gBIG&@m{
zLr+LDRILq%t@SzKXX<8a#M$|lWc5W>I<G^V0sRWFhY07A`ba~)7V%a%0cDNqGJKKa
zx=xZON%~gt>+%(Gclm|))YwAMZ0*UTMKf%po+3w?wU=zqiZ(Gzbzfe@3xs#4T1`xm
zSBC06S(f^-99mw@=7W>-y!}Q|&8VM@MS)<m;1r${%+f%pDe#$d(pFjURMM-d#nneS
z0G2gWpu*e<YB&JnA>NhOf1tp&EdCs#Fs~2A)Z<CWf*7pvhu?7g6ppbM4Eu**$vDUx
z=Hm6^sJ)00C$cw*Bn2rl;C*rh(H)BW9@ALH9OloAX|}QtvgoBf{)=b{Bc>5^sC6e_
zVxV$Zq=0;E#4)(*Xb2;=iAzW$@HS&Ac@)$oSV*H#H!#vm5*LUfvXXy*8juDdZm~^O
z!%%(;Opx4`-W+}xaa&=6P6b^70<I!9vbYIq#O4k{|0@JvM*l0ZUQquldOCu*>?1ON
z?#jIF9A(4Is%E{ZThuB=PZZlt;vT}vqf`_MU5Bf-_cV$|>0%;2vP{*WaWkJ6g>UX4
zF?}S11)lR){EKGI`MJ{OQ;NG3$m4c?!cu(vq|V%8Ob*j?8B3Jycht0XEWUK2ThiDQ
z@5E}6L2#q|Okoj+W?ur!i9{$Jr>>Y~Bv#%comq=|PqFGQmh$30FQTkOV`%S9{o*5q
z_!GNS4h|FY6h%jg^r_-@(4ER--AAfxkSD=J86qH;*e0g<O4g^#y50a))8!3BmvENi
z;6;hM0=cPrNpF$SCbt=Js0y#x!-DYCqhRYp%sSxR6(X)o?}=@nt)Hsyq9hpIoG0Cx
z9~4!|tLHX`O2vtyQsZq3&ds?t5k-#SnxybDaT|l)*yTu{R7dGxC&YdP56d}=I?7lY
zg29U_7cXVT=c7Zh6R*svbooHob5O^g5ZnY3M4?AX<HxE`X_K&N0;E@<Yv<#D0s-z|
zJZBTH<x5Z6!p$h&<=7kN+V53D1^Ow&E)EX|MDBRbww#B`qAy7cOaJ?-Ge%<lLO&Uy
zEe63WUBftX3PPp2HrDZE&PEAfiAW4|IA0+w3~3#`T6V1<+~$wY93DM2n-MpSYWCEW
z%x{cA6Vdz5g68m<Vq;u2Ak<HM7%+k9{}|{n|A!t08_Rzuu%Fia|NHr~wRiX!)fDj_
z00MpzEgE^Pv?r=h%1)jeLa)F0k`hTYlP+YBTRd)A@iZB?u|@@@zT`aK?OZ%N-W}>$
zT~Hu9(y_Z_d$7~K0*?iS<G>WA1|DC$-|ll8>^-tx`70|A$M24NT)4M?KzxASAoo=K
z4!gznP$#KVav6qPr&IT5bysK|cJANv5RpMkU{%Y44)tPtcg+pdE6)~2ov?pDukUYq
zJ4brzBeb({+(EHQdV6<^VL=pUM#nKw0ZZ|JrFU<BFv?e4VPC7N0IF!NapnazMkovD
zIZTtEdnhA81<*3o0fz<<d`FtdeoMWf8b&4eue)-8MjhQ}j%G)T#G)KP36pQ;cy@57
zkMhoJlsa@e)rK*1MJ|L5JgoluA`rTOR_4UHBGIwlYpNxJq<5Y$bfuOrNW7u9*G(kw
z@`cfa2-4HQJfPC7pNk<`H-9`!v8+f62Feoh@kplaV2Q*?F-R|$-EDdv<orm7MHCtp
z!7ZanuYjE_BtXa9(po4Pt!=dvi8Ut0ObAukPOvf@s4H6qOj2O|%}O(cQm0(oft4fG
zO@wWn%;Czdv<+)~Uc|cV3BCxF$Q<^rq^vY7=gvU&O;Eeba8JXf>XbDG9HKv^nW_Oa
zR1Wh&xY$3ac+~U-k#skCnV#XDouU&KjUR=dpz|X2N2mRynf}-7kjWKMUt0Vk=h`<s
zwmhXmOlG2n^D7@Xl{sa+(q_#-owZFs&J^-ClO^f7#DuOM5}U9Fn-XADO(P~cjguM=
z2h&6_W0)0rTMic8u<fdSM}s=&i$VhO#GIiREP(my!NoC#!y@64d-jxUiBF!wwAJTK
zMq~|=ab9GkUQ!`9RB!*+L@uFun%1tL6zwUQD)blfmSRfptgr*B^izcc%C!xjuhcX!
zKw%tiR^5S?6;BK>(*q>O5U=O;Man9&V4_xC5g%K-D9D~{;qkW0KhE^(%z^T6=A3xf
zg$TqzL4G3$5)d?=)`sd5Pwytvi()5Z!8;R;sUeIJC_x-W?iJne#5^GA!~#E355|tl
zR42kH0IzAOBT*tylI0F2=S)^~iOkf)kjql=CpOaWFv%CJrX-sBnQZ^b(OS;<RNdKi
z!B(^y1^o9EvQ#SebCjjM!DlWAGrROm)75MWI6Uf`0UFU_M{sNwjCLE+S5Df~d$a0=
zrkARnFCSmrZSt_5C&H=4W56s`*9FWRqFa~^vu-<A?D1<aHFo>*Q!7;AjEkcQXP!a+
zAfEVO8=8CwwYNm-+=eV_qu6hP&-R_mq08Q>iSlq5ZfzmFmIDvxcFNBVS6}{_Cp?9_
zeU=#;Ge4NEK-e=WV>?oA7){uJ4e&UW7f@$OoS~Gnjd?<o$0S?$oSkSG^(=jau&rjS
z{zq1l^lj3AHzfVETgv1nfSzL1uug97nVsJ5Gvj#k>~L61PO~&Sh;VV!s+e)5g0xc*
zVElX-QiLuh#(cviWuB+FJe~5o5!GsEa`Pddkwuz|O=2sOXfMJVme21{Z#N;dRT9P8
zWfVJ%({N$dYewMDH#fN<a1JR!rzpTMm-w1~-b)6?&}}FT(GoO-5t?a$(vdoRyn5W&
z`W(5KS3g4PxctRlYxim3REq^loz$=0>-19|x<B6tvkg44n6V6f7;@H5_gT?SSb%v*
z&~Hx?2aUgmVzWv-)lSaYwKBA-J|x?!dg5q$;d0nwWNGQAs=a;#158NP>8Gs0`XH+|
z3z-y^6YBrV-GgOFt&<dhY6<8nRjMHCqU5Vl?!$QN$5j=%pALtosw~&=Q1^$Ux=Ul?
z(*aAXYT0!3-PaQjUMs5wSKIEHn{W4Dz9!KFJ`yv*m70RUj(oA|RS2Nn2x99S4rGg5
z>I%&(x5Y9*(C3NjrT_mi_72gJM*aGBY`bH#W81c!bjP-B+crC?*tTukw(;fvt#!`i
zyo2`{)aV)2s#W{9_rCAzBCa?u?BWHIGxVFLG|>y;l2O+Lt*}K1e^X;gdilG2u`WT;
zD!L|FG|28JWH>u@cn=7_D4@v&w1_nXTOL;GuhljEGwM8eISMar7Dhu&f^}I~R7?Lv
zb3jtoP_`gJSZ0Y;*JeKd<=(P`b5dltMy2)4Aw)~GpbX;4gi@9&Bw5<DbtvJ1<TEWh
zy~tKIetT(kk9`J+#Bnb&*=8e5>jqgFYuixNUgjz&>6*KhzN1GChvD9%<JJu_7In+g
z;g8`lU!%@mo*lrWfavDQ!YlgnB5p5<&o#>XNZM0cy7LnD(fdm_x<<*rp2zbH6h3(e
z8*?s+Omfczl~8mg)o|!#e98opT6*YIBAZ#&M&?Hx)?YE|bSt(0{J<JwY`54Ih$E)L
zI{YX7Q+=7CIZhwr*x?xMn?DtaP~|_aldS)Don&JEX{Y`F551VVm6Ndpy_nTsCu0#~
zLt7(bK0X*nCkNxd)-Y~sgW97pxa@F0-S7uQ4Gc0x*gq^Yn>lQ8+EK0GoZ$9yd;{OU
zP~~%a($&nIQ5Di;@o19qu?fFjzb_81)_=ac!+XulY*hT-@*n;i-VCJVcR-O?BsU~s
zWyt^I$q?ulsib_xJeU?OWPIU%K0OS4Qv6RXW5w6ue`^_3r}DG5scTc_9mThL*!^f>
zY$qaQ7vEm*m!jnU_w^Fxif*PZOTPr`NU35`+7)xB@iC6g6tJ#CW|f4K<BgfAC)Y{c
zr<HNWr@a@$bBIpCN}pfWFWuk9#Q!d#$?^uP{6z!lsuRNd6hi5b8;hQ%W*YzsueGm5
z|NYE$^z}*;E1c{`Pa_~rM*7QrCz^$GK_)YYg~*Vp@f2@$V4|0TGtE%q)6OGzNCZ5s
zsDIoDvI5{{(!}Yp+t9syw0UBjnX9}`w7Ot}jS_%@@G0>s3|Bq_h50b@t|gptzhHte
z>gF<OFq;>PHEH(uid~D{!#UqU8;+$-KZ(mX@Dd$b?Sx_MO)$hs4=V<O_qqb!Tsc$b
zGZQ`h6-TDx4o|J<4&;j6hTUVfh0HT+?p<i%@Hew@vNZ`!!s<)>c}l!vSY*YZtG?T1
zkafQYRYT@<T{<$30)3ETAgX**!WnLKb`(@iQmMIdOs-s})QmrmZghzBS6e<5zx(_F
zTPE&E=x7UvLtN;kViKx3cvRsGM!MDGHCCmV@<DU!1=FLXmcby1Qw*1*r9$7iUA4Ar
zVd~}^Fnfv`SOK5is>421wUGmjGOhWB^K7Ms8#hJ)4L*~iUx(M&N*u>J+lAE{NFb_P
zGws`KW>eVcw*G>z+#Bd&>0v}GfvqVbRnM*rwg!wISx2=PWO$$ENx47YzcvT*T>Ir<
z0a-Q`wdlEWv<f%g;hNL}>4$OUQO#yj;s+@OjrMuTwcc^Qxjh(W@LP?61p_l2eFaHl
zaWMig`UX2C_)3!1q_@{!pT(kxCJ~vlOjUisIJ>}F5vs1*#dsZ7$t4p}Y8)jftljIU
z^u)b13VA6Y?zGNP{6kV`D!+SrJVvFC{xvGol9`<<^N~>oENOeJ32fa&%G7)udZPL#
zW0>_LK&NvRTgb`jFa<S`NlhBKttkl@gRLNdB)j8LtPi}lz_@BvGv`mfL$e8H(;Q-&
z0d^}$9G7wWCjVKwAsFq8ARL3px7>+ujrAylulzC&`=iPD@xs>8+12^Mt=uArR<!kl
z;nRN!1V&qm$D`zcT={(SeBM8Njz6o>C0_r`IdPoR8nDTxqKRus`vV4EHq}8#`<(Fa
zIkz%sUC32;nv0HRzfnPB6UswF4R>j#=1$&ep{m$f!4hh=J)X%cx}s@al=!%0H^H1T
ztIUOVn-d?rwTvPnuR%nN?u{u)bJ6Rj&y#n4#M1e^5}hX1`IIt4`7u|#s`PFB;TM<!
z999{>ZZ6$WoBufIn~IY1?2d_+yDWB#zM3eoOlZ28ikXXo5@BG^d+PFN1)e_}uaaz&
zTtI%7NYtGRt#;d_<yyc@8+k&9^)UnuU2xoDQBEQ1{SBGk2^)>YWASQ2PiKY;=zXf$
zG74d<YO}lZn!tqj_zAA{lazNkGQ~92Ttjd(l}bBUZ4#b?{L}_xH$f_Abb0b~YqYa@
z?)uAMYf{Q^f^NFH7P4Y<7qF?quAtWvR+cGd1gnGelE|k)D_)33;w8M8hVl5n1ey&g
zMTSmMFDDt=(C9m_9LPH}<qj8T{*rV3W9Cs}SQO@L753G3QD1<(3-p&uTw{b!;u34J
z%-k!aKxzp`q^#zBZ{B#5MuSj<L|sigjMF2kJN-<tg0dD2$w@(L9Q^HRnKVvDlCye2
z)p;@}F*pEFGv}JVGdA~*aj*jbVr9|>LO=(xSGe5|gWM@rGh#sf(Ks)jUnUAlk1|K0
z<-!M|xZ$;&JOx>s6BrEw5`ZDdI-dU39gGLDvxq-1Y{zx!DnnyGgDtbEqh6hJS2)@l
zZw3InA%qnc<BByQ5_1T9Jnid9cu`G{Q&WVUo#nqVJ8da<6g_mY6WAR+swN1*d|dwp
zJ`8R&As-AB6QGakjXw)Pcgq~)d=#O9L!;ODdxBS(EeXuui=8c<Yx=^gz>qfjQga^2
zFcz)S+;Ao6HZ#58A%B~&^u8tirxY}0I)Z~_`#31e|14okS?r)5IQ9r{OgdQb#qU~4
z?~ai+?F^tY6Hh{q77HUKS$?NY$8jeqYLSh}Q!b!EXN=N~agiIEBXc<;S)vv05T2-2
zMJbwC)SP#e%#5k!oFWw?7|Y|z$0lbw0IhN4s86H%+q8Mw%sq3kBA&@zf2p5nx`XDc
zxrkMZ*=7~PMVT_>64wX<rnH<X3oO%O{FJbQva1ae(r>;bjWQaD*Kz8c{eb=04~)b8
z+oEcv*h2?6GhYQkHo1cplmQJcO`QKNV6@jQ0)#oG^oZ0x7*BQ-wTnOuWFqdB+A=h0
z@iMEgJ$VC(vhVChU=)4|=JeK=x^-WH{JD`L@HIoBO?W<p7a=(P!2aDT?V<-=n}4_j
zTm5hBo3x|&OicXa*Iw2VzBCKQu~azF(P<+m7F~%~d!Nm+rJ6;A(QMQ^cGT5cp1Kqa
zVLBvmu1>m11D~C{Q6N5*S4jNMox6DW%X2Rz;Y<ZyQ*-M2xxEHL13Mk}>K6lU`?*+Z
zXMU#}%X6sEFstWkQ|osy?|{Z8J1q*^`vFznz&IaUG{+>JjTLYvgxA5WqHPoZ;+EW*
z(eciPS!6P5_u_=y%_7PSUK!ADlfgWy?Uq4dX!EtvDYoh6dn{w3F8bxfR7qwyZ<M$<
z59IQjLE2L>CG`U(@DQead<d8JV9cs0a~%4tY0hD7T}A%I-jG0?avY9ir}XCe3KQKn
zDV;|=>TGjhy%s8we8~#?n84wa<%X*Rmw$B<i!vHYE5YlxQJgw0ViEJAR8bFLFk%L`
zH&8-+JY0)Ciq4^(<cQ$(=%1<jO-tOb8k9$xt>0+r2cDgZdfUPX`xxpd%c+f3`mz?s
znn1kHp^TwNzR*s?EF8FZ4a4lOF2qYDxxyB_jrn$M6qx;r|H@j)-YzWTY4CBYAZNC0
zxznfMTYi%FEf>+*DY3N<mI!FA3OmOENB?B%^LP<&?jEWrS2D0}svvv!6M;5-aXJei
z{^Nx9qa<V?v^B7R;o<q86b>Wje<?><{Gs>!AbUHXf9O3K{nQxjd4KaR@e$*+A-rK7
zV3-^;gpa-vscaG?E}OW=d#$C3BO8}1l*X?-Dcb4fkDp8B)kU+><pKw<KD8k~1n%_X
zyyj1)K%ffqJili@pNPq@zp`G1e~JkN<K1|^=kxU-!!&;CK-=IPyxJgLg)cFCtYxp-
zcaX2mlh0M|-MgQ1DBv(MI4UJ~9&ewYsTAqIzMPzEh5)ssX{CME9MNeRCQ?pv6qd<V
zw|(1jxr~q*tt7ciiIUHLxfL}K^IB~=0Rl<QA!P9#a?+ru4ghB8KAEqq*rnIC1J+at
zrfei*-MTOkGPpLep=HY%+?78#zH`(w4(Ys!#qDq?0$jo{wgFQ8%c)0a_wV!i@|^~X
z(AHk2NytlUSFr^0yTJY!LvIL*Zh2S7xtq<U`r7|iO)g!X+RVhvqn46?%|iVM&6R{K
zs7*R#s{>~emK7Ah$aG6{B@qWGo_&3Mt~0pxyC41uq^veq_G-VV@U2yg8^9#u^-_Qu
zCVj?mfE+wNSRJf17D&+DQqvE+H|lBBh_^V)5oWkRUMbyX$v+~u$LU8HN|I#Vce*T4
z9r>+oaMEEI<o=ev){;_|trEl3I1+yq{n3nS5d)s_7pDHs@cCxVQEM7IJfBM)-N{jw
zt~th@cni-<;8s}89tH8C);J^?$|bW+-8A2#B;wY~FBy%p8Q%skmmumVzS3&C)R^bu
z#2k^(Hugwz)e-(){<rAVK=Ns37)0!*^A=u*@HIxPv1txiy<73jz+3#czZyPXHKW(4
zqE3t-h1mmP?ehDeR1-Rk8TX%FfD}Qqw1V;`iC`z_^O~_Bbuk!YipDg_Fe%CINK=w2
zDJ<jAdU2E{sUqNN;N;$6$&)kCzNo1btO(<^dxEH(3f&a!kF=4bs?K*Ki=)4g4Oq2Z
zP&&*87lC?VDU@G^GC`$Sk`z#X#ZeM4-IbCd2rPWGS|r&f!@$x);1bRlZ=vySz@mke
zWqx6Y0@P&;BK@<`xsjuEt7%?=iu9>n>oUj(alFnYXE{&PTH%4s*y+<osM@=VXC^90
zAL*UwA|n#ytAhrDE1jbM&%d(aB;^L_ZMm|y98I5rYd<vPMv~CZQq*rnpY(-Ua5>>%
zY`56P@oL9$PC57bDQlb4ysD61;<Ni4l~gWK?l`j6e?WJ97~vu1D`GVK4ha%I8*AW(
zU2N%O;+ea@W12t=@{7bv<dMf>CIVmTbbdGf>bc$#C_k)DLhdpyVYUFw`z#N?K^OYy
z<ZF>e96!pBQ#Bz@IrmH1M6)(HObb+qvcus4%%P1cc@*hMEboL(hXEaZa}$uGN<HZX
z%w2``3xQXcI|7PJEv6uxkld!&gDJBv|4wLwfBZnk3w1>mJk>Lx!8acX(+;ab=GHLz
zx2iQ4+D#I*NvDUi$a+yrmi*ZNbeN+POiJojh|<J$qB>7>`f7aNwB}UY)*OWn-MOF_
zT1R;A%5{b#t7v#9aI}h#Y6g&PJcubLXDipU9WTBA&7tNsX|&=}NHNeJB3d0AxRS*-
zI9_VVxnC`1y`>I082xN857Nl=QyY|@oy=7l6dOn}wV|WEzGo!l%qd7+dA}zl<m?dx
zee4c0oidi-_FJ$*ZXEVGV)D__Q)(Cev(S3xSh?9C9G+Pl&EP;uoP(sRzKv=&Yun2^
zH~gp2xvsn)+*bH)54!GBCob{Xk?Kc8Rb|~c5#k}+3kBv;c>HAJ*L4leg#Mbdr6mX2
z{<6_!mMdON@SlPi@~v>NH6jUVTxvuE!r&69Qb-Bj^I85odSzOEtPWYWZ&m*mALbAN
zHH_{Vq6gk?S)_3zF_al=xy`ezmtpqT)@rX0&WMC78Jfzi;0;`(>F?E!@ziPdQI$FY
z(AzhCy%qviwYK>pg1Id#)vqQ9nFr58Sz0LhhUJpQv!s>Cq*S;DCadf90Pz*aR9W%;
z;^7t5TY%W!7v2Mz@F+a|25u7}(x=Ea4%-(HP^?j~BEcIU#BW9v7g{V<C5@xgElj4c
zjC?vC2(Kx*I^?j}YH`8h^OhrQ@x`FMo2Vthi{OmJC6k66S-igXE&t38+Ntw>D{a2F
zU|C`3Vc1NQ;jK65T+?EZSe@ihAR4!?gihd4J<knS@#~wyQOWJ(t(R(Gc!F3T`uOkc
zJWjF?<>)D;Ip^(m=m3rl3rgCKY$zXC;qbPKZYVN|J>P;(GmX~v_OjE>866fdzM^Ig
zR1bAaMN|!}NRYaDPIe#1CdN3(K|Zmz%~wB(DP5ki8`$v%dDFvI_;I9T_IZcK7cp9Y
z#wlB!`wuw=wQ3&)epQpgS<WinmO1dIRsi<dzf+&XfEO2!kW}HR(jxRFL7H>qXJG<I
z1>t-j)W0~Ns@D=1BihubCH_mKIS!^v&mPo}){Vd?c&?CRmrnT>_WiB2DRQWG4D|4}
z#wa;0wu+);-d-ee#dwp3)1Oz1E>xb;d_QUO`klKflB+_`-yRbt6;E|Kt{t1^)fTt@
zPnNMmS7uf8ukecrO>Z9=jpXcxIW0;oeL3^%^Y2_WEb08mf_G$p622}ixL1!%4sD(0
zuuA{qivJ(rKuoOc|7A!+TRIk(3Z?70dW3%_4JaUVf!bPqb#ZRX>e0F}G@kVd|6-O0
z;r^MO5SoyvUQZW)RmoXR32A_v4?ksoSyGN3@qIbeRot4Kf+s+E{yY`>1X9vhP^oZm
z^aq#m*W;7#^H@^K__V$>HeSi{jpOa=G`wB>e~)E&xj+{R%0-n4f*+C|6dyJZn+)-7
z%{x+uL|{a4oJ#7pKc64B5t*2ET9mc&;1iw2kPw601Cz*$kgsF#)eh$fd;ahzXe3H_
zrAr7-ZpXg}?EyP~;`=faeaGB$81C7&i_;V~NDbMZnO|Z0GLw=dAVf`P0W5XW2^;2k
zO>ca3L9v5fP3l<2z>__%$*P{&D5G3;uLo3<CjRUp=vb|rhTE4Wpz%G|*Z(B0khTj4
z=fjRGn+-o|2#hjbX_Y`rnuK_HiOxA34$?e|0z-usC+?tww<;uIgbAt@<$V(d49`nb
zW#wm3f|Z+=B_D?V1zNurRu7hr;_lBWH&<5bBIWh{flg+9;v2U$)m1>tZs=O-^Ykr$
z#OzEA8g*slH%<v6>vN{L#!n=R1qAz8NLLUB<1{4SEpxuwHZnmVC(4xG;6%g8f!Wfr
z-W4<t%s{!1KtntJs4y^wFn@78wO7Q03eo#gafFtczWIm|u!pj0YP|`*m(qUiI+vN5
zB|)K-Om7o>rJ1dzQccQI&-Ly<Jfme2o%L&-pOnK3gq?Y~J-;v0f4_aeLSf&i;}TH_
zQwQyoUf<IcKmnP&rRej1;^N8)C|gM9kb`TW#D(6zAD{Z0AdAtca(hQcYjMbv6~%y|
zE#Xib2D{BCmC7Hq#2>#cu`vld2aw9{5_O4EyIXk`-||p*4avVx>6$NmmxR*K_YW8R
zl+6fu0f{cM@hnsv0RquXKdBBP^p#4uvp=Ij716Maj+w;f03vjtRUQt^5+>F7ByDW-
z)^G_%aGO2T?Jgp3ij&INRXjyPM(fdX=LgXd>g+|WZr`t*Q14<v`bkIfd#H>tJVg%?
z`A4}737UL4rc-LJ%yvgxguUbJp;GJ?DgjP~TxH7J3EVPkGiQ<T;8^8;1C+<wvA0{L
z$GCW;Tad?x);2|0Y-+ToGGoxFda$JgL0GJ18%VYy7GX)%wbCeM?aEgFfN&(NZPX8B
zN{;BkDCTLxKYe0Y5RAsZ#h6A%UuZB$Pty#=%xY&wRJ5MbbK@d0AHh3KZKStVlB`5e
zb->DuHsOsu<SCUAOU}oXZjXeQZt{1jn>y*&xYZ-zxblG^*ZJ5No+&!}IcOT1dBNTI
z3k)AOb|^L#`1VIuFPRTn+JKOSd@SwyYY=O}A>EK2B1iR(Q_;B68dHcYSgMS|u&*3L
zUT8m)1CC~EA<T=4aLy&-EL@U)G`o%NgzBHeB^RYUHONhHC`tJNz>&tdq0}a+Pv?}3
z^yiA$y8?|@BNmapsJhJQE2<(3eS38cjg?y1pw9LpFO25doomY*@Cx@K_Ob*|E{hGc
zf-5FlP(e-6=RqEP7_mx!Lo@10RIrw+j#s!vY7|XxZ|6GWhd&sB*bK7;BvK;zN#TT(
zb>exU&puwy9}ag7<ymoxebvULEl1}uQkoCEa)g*ch?n=q&(S19wvMkapMr8P2jde*
zOuW-H#$Mzd&PAp7r-IbW$R1>dI%;bp^PN=Y2~)6F&fPOB`gAO#I&+k(MeXe1O9f}|
zEdDGs;Se{jkPT(FC|bI3=u9?|r$xI#uQM)-fDtRVW|Y%uDUHH7OfEOT3+??G%7AQ3
zU()fc2eMl*YX)*_k=FL@_?Rm9qyte<E92x|d^B65PU_6P7h94WFKv=oDMfGoXSoh$
zF}BW-+B&jS{LCd@I#yM3qX_n!Yt!g?OUK=US{fx(F|@92D{ETdpU3A&eXnZ<0Sv9U
zKwBZq#5lx_X}4bst;yRY$5xIcQsI@(Mp8{OB=v}QmmIX(>OAy&02m+;HDX3Rw(J)G
zw;I7ifUJm}*H~z!0-UcTEbo2GW5<vVj1wAO$xT|aV261NtTn_kx`=dHd2Mek&Kmcr
zlSZmb&@$++^9hxWbz|*P+iUgK?+C5Vd>bj40YMLj^JGf()vD(dGCUY`FIn>BE7hea
zQ`8yd^#*d$7W@)6O~nqsjv5y+cHNNKDAgEfP#u7!l!sVrPAlM_?Ox0R;WI58n;@7q
zq?C=fo>rK^3HLTqNP`NQSyUixUd67uLPJE{Ko3skwG@q(5CpE5AU;-4d2_IPh%qm7
zjsI??ol2tzJs+LCdxND5(hj-hdD`a!gK_|kJs@BGbsmi!*$zNuNzw+8)U5hph48ke
zE=-BIvOi9vZ%Q9ciT)r9v<un`ow(Xkv$s(;vuUr8HnTydMO-PKIt$wkk7S%&^*%rg
z+xO4=PV^L)#GDE-*7U>2t6{vfPW@FjMYQd&vuc36kEK)2-dPDdXq$*CXj&kl4hEWv
zTEIbm1<`rxH?g)iR@b%JGmhK8!|2Yg!DQjH7~f;GPM8Az(Q%Bmd3nDo?fvX@{V7MR
zrDxaXH6CLx#Asc;45TcY6eM)Pb`C!%u4mb4bB8wC4|LWN05#jeWy`upqwEfUXTu|u
z&-QquyPH#C*$fS+idL68;XtBaDeC$iN*qt~&%M7t1ZE`ihPjVIWA@TNW*)!7?}e+&
zn~y_t_D(<YcwU3*Y$p+EGAVBsgrb@cK-Cr#R5klPorDzjxx5Xlze5ZQ9ESy9{Doxo
z?ctJ|wn2>D$S&8c<%{?9l;MLMQWN=np0Rf4=DqCe-QH^NKiZg6$w^$Hr_LHj{k%q(
zIxz;18#RaXiC(-s?&F|6>goZRW0x3Jx2cRdo=xl|+Ryz7WAeF<OyU?Pr|O&zcS<oz
z)36+SIc+(f1r{u)9vP=a&iM1!PPM}XQixmlx#A)7rRJBoxC_ODtfA2+tvmxq2`aum
z>odPy5n){G(y2<4j(0;wj?CGM*+kjStQYlhwW$=SwI{qd&XFakX6MElGK1;E{B_6`
zo^o5Nf?&woER7T$kJ*;e6PN-FS|nVTuA7Rw)+fXI(i^~-uZC7zK`Hq`yKR=9nD$L3
zuvDvQ=Q5A>wh3fLu2L#a_XEw7(=yCAZ*7<<=K%Olp6Q6W-J55@a5gi4+v{9J_3<Ba
zs$p=sCUG4@B<~zUq?=|YFNv4WyLA+^A6+ErEUZ!;XRUzO54hT~1N|}J%2M3;^vMIh
zV8-qeWI2P!O#{Z0`T)1x%A{(!I9?~`2^6^$OvYd6a#<=(>aV(m4s5_NyR6&e&i$Nb
z<Bo?WM7gwpS5mgyHd}e&czW-aU6w|P^)1*c*SO6)C2i&mX>nPYiX%rg%0djtCs|pl
z+EaBKg6Ll*59WhM76a@fD{0~7R~lVgyl?z7#D<3dX|!Z!Wc)veNlYA^|DFG75$7lW
z+xobrugFiB5z7snk6d2K7MBUX9e@iVc7c0A$jo{0eBZ)f_peP#N&6O8Yjbe#?kRb1
zzByc;KrS{{{+donHiZ*jO(wTn&X1ns3p^%JOdiAU7xerw!Tk0WCnf}c?2BTSbT~m|
zH=fV=eKuGqGmt4z3`o{>B?xXj2eEz3%**60$xZX5VQR<Lj>m8^4EN3!;XU2`>vit8
z$KJ0$Kgn|b{^gg4hu4R9bFgi*UI7_>pRB%?gq|YQ#GX_=dEXDiH;*?C8?D&A7ShjI
zuebemmME{WhkM-9;po}6LIOL|e;HA?G9l{WK4-kg*te=-%^A0WEifl`Uas~vZH0Ry
z(7ab+B8S*i|KhZ$ud-YWse3fPL~h!V!Md-ke)gb4+2GeZ|LOq-Lwy|W4lX`luTxB9
zm<=Er{M&^hT+JK=_RDO_tjf)Np{_@(cF?=&%~ymvLe%6%8imr+1YSO4HCXy!QYT>2
z?02XAC0@{%>5D$S%(?PNl1(7sJ&?Lwq=tR&O=wyR`csOh?tcL(%I~PXS7MZwk=SO#
zq-V=Q8YugAYY_QW`DldgCiv8KYr4$#VmBi#{Q*h4@ZCh+&#&2X*PvN{0XL?2Q3mcY
z1lhU$9mG&V+zrx{elPZQ3eJpg<#jeVICN<@U&nt?eR2b3yF+Bi-rG!(=a_eFz*z93
z^aRWaphLRh=qbRPF#Vj}QD@zL3n_~Ks70?*#Y6>StD;EivM__kmzh<GLvn0~@Xg_s
z4SKWD(=i8Gd4g+&NmYdytQP}{e!&~lWH6y92oQx&o32`rj>}Mr2*2)xSN?SlnqxUu
zH9qt9NrjuFOm;&`ltLc~U=5EN^8gTJ4CE(|<_Fy^vrq`q^d6<#=VFHhae{wAwac~-
zKRV(!8zZtSL`G@UxyUh`6e!135e(X6IXf4bWmt2Fb4|cAJ=zKP&cpTB$Ahvb@;s=y
z)V7boEKwSj%dnoM$snfl6KmsXZ(ax`RQ`d{5;&Vy+D5&CbtlM|r@-*J`Q4cchT6R`
zb%p&$D-5}+-^G>0Yn`<KMmTY>oOiQu6t>;^DnM`5jQgsOpsJcl0-a(1-^ilGr*&1x
zuYVIr;q;2kb7o7=|17D`KZ2*YDwEA!l21t~M5}Cj+`%J&$38}0D>RB4%71)Rrmz(C
z$<&sOyrj?>(npL)xbAMyXfw%6!T7`!gp44G=hz`1s%=9ST;odB<Gq52DJiBrk?_j0
zdy(&xTDjTQ9&52YF|UE>+&(mNu%<#f+(<_h6)gi$+v9`f$epZgXqYeq05FM+K<q4j
zbg9Ynm=L5##?Ij9^rY#DF4pG>O6HR7xxuU*+BsWVWE60AneI?L+cY0U?(3rM{tKL+
zj~MRLXSc6u>F(HG)wyfd7KB6Ab$gRAHGM&SrwFX<^FX%|g{%Zy0li-X2p@_yzaiYz
z0O6bRWi}Gks~~miOKeFqlr6A>=7Z7d=o}3Uzd4dD!DBEZ4WwG7m+qY+-z0y4eF|!0
zl=M9WafU)_!tMM2;6sV5`!^l|u)-(xVzI@f#9<DB_~*~i6h3KtRvQ8yrKR6;f3bu{
z{zTO(OLznz5QWQKe|=t_>vw%Xz+rw+R0@QK8X#=R^3N5Dp#YbidiprMtQGzRI%0f-
zJ>`A_W*IQnbpK+?M$-b1TWjpS{yW2b$na`QQyALV%ab}es=&u*r4Bi+xqbQvxC{@e
zwwxiTv_@JF2jAXX6CJ-llrOcs<t%sp^xWx9gDRXc2F}6wH@jsJ7@Q>lOlpED$RDiD
zRyll^HezmwQnUd9rU6P!Qe={G9N2epp1RIBh)^fQuDVY67xni3Pvizd2URjb#G-_^
zO1Qqy5EPV26V-hEnx6lz&9>zX^i}SC1Gp_0iGPdAGJU`b@i`)XUQCrmp1wX(&~xNf
zc~tI?B7l;jx_r5YD&WSRn2=oO8oQ1F5fzHzQ|!H7K5w=VxsS5f?_DY_I!2>ckj9Cw
zae+vMnpNc7h`V4z5+D^-^|<@z;9}Q@VmAD}<RD#zMwZi5F+!@43|U%rA!*MREt=zu
zBn){u`j0Dfpdj8}FExBSPaMaQQa;UKzEpFCF&KawdojM>xI2qDQw5Fi!>?^QYkmO?
zp7vnz*}gl&K3kU37wCHajF??>tjS!HsYYvnb9+$7y4Ww}t%ADbp-zC!^-xBcv@k{|
zqob@!;MX{)w|xTIjM-{4n#SB46kv)(bc;vw)j8pis6>YuP8NI_PFVTQ2<waWalY=k
zi{$D(Bb;U`$B<f7ssQrdohTOoeJFAYS5Md!k@_!Vg%UIejajC!J)wkVJP0|Xb>2WW
zhucMJ&Mal!NB>ZOk6P)#Rn+pVjm9IL=oXDaSxl$Dm>y11qrF6wSDtfjR`T8|L={P9
z4Lja>nAO`hI);tL)q4+B^7c+*1Xjzuwtw4&%?c0|vWik)WOFZKturg_$<GF7#z||j
zB?FTOC}v=PCtyz1B9#Ut0W##bZ^~VYSDP}ECm7nDK*Xpw1cqJCO{5sBF)5anDkRa6
zC8dbJZ~zeHpa_$E4=QU`P=9hp8EtUYx(6(F5H@0PqQe9*T&aHIkA$Hf&21E_>o05}
z#P%I(X42q%F=GCuoh($T6*?|+x4J4-t(@VMVRbNjhEzf!+s^Q|M>H@#o9!;4OczCw
z%=52zIV?f6ivC7EI()3c0#h}UPK$G;<+jTnS$zEri21sMT@c_E4<o_p!Djh!T=N0`
zR$)QGrcZS95wktV-YSW4K>0OZrN=2zW(CD|UJS$inw%Qcrn;&yy9&Poe?Pm0()_*k
zswI$dnEd@?=|V`RNRhz4<T3Jzu9=S?s*aeHDaSeBpPp(^_OLWQI$+hJOh3y2C2x2@
zmi=>uLrY4?FCG6XzA1P)F#IMLedk($LX`$7nR&1IP`$=!C$na6@6Vz%s>`pZMP8|o
z%dg%{>Vk?slOSn0#4DzSf8k>5!UH_7ra(`c1ck1vu_GyM#2MGVRRJEAl*%x%eg!*&
z%%x$kPqv^@pU6t6(U=J67rT%`Fg&UnZswW9<R!TBR;bW=4{Y(3uF&<Q1C-U$0FyuU
z3ztr{oIUKT&PDS<%+1oNC!z6fx&N|qm-Zm;mYB<-60%{XWGBtQgCF5dZf-C4JIbXp
z7K;MEr9>Y0z$~TDzY5a*$@aZyDl6m8yyKpzOL&H;RkuK&UXq_q;>7C}0^3`NsOpTU
z!V9aYELPx5llzwCk7_kVHrKk2Y*%xX0d-YFKO}t`qXW%2(IJtuZ(0z)GKe5|+1B%>
zQuQfh!n@>BeMazRy8*NN0B74xP&Esv(p7`OpdcY)&Lh<O?KfBSVzhW;JkZ$eVm(z5
zcE?n?O+6$)al2*?@N8Inc_G*LMT03Z#K?A@PhhKcjD_MJ&wcBMRu1OP(9Y-&(wCy~
z`5u{E4Ui8}MZ2ZbfEfN_A?LebuGO}K6@MsI;&76~B8C&u5Tz$$2j1Wf_R8-E2`>ND
z@a^Wlaje_Dr;2bcU+fp-UMu1oXg#go6Z~{Zc27k6Y`}Y7f2U<*y12p?ai2_d=tPJ<
zls-jWJV{62;7}>s_&S-LnuymKBY}fhzFKTZV=i2?#!odP2MipYf-`U52I`>rCY>Ej
z-DEE-tIjOdm}!AIa<rwJ1>l}bOi<iXMZr@8bH76b@#-7~>gB}|mWGi<zpz(bp{sT~
zO2(enO#1h9pLRJ1Ltf~@lV{Fe4x<mKm2mR=qX&7i)mU%>cRyJ$$yCe&hkAh=7mX5q
z4+pxHfxx#k;R$FmXbVBcZFCVH3}X5F4M}Q5g!Z4NbEf}_#=ywH^gs8&Ol<6&|5Z|N
zQCr(qry1Gj`KM8;KL>dqhdTv;f;&(u1x{M3`C&2SLrUsV$e{xW0f9d(2ArJ5cw9hY
zql>kK0PbYfQC$D7PFHLwwF)e}JLSu2$k-YgK31;KonT{-%Ke|>^F<~s<<zr3bh-t-
zqwlCpEsJqb)5#ODLJj!?W%<ybk1Xil`qRq!cC7J;tiz@}VxUQHHDLQ?51Qdb4!J2{
z1@8i~M2%_$;h`q=e3-_SWlyac2&zT=)5c~0<`vS$wU(`CoG<5_m}HiOE3www!lNOS
zG9WBL(~mTM;oE3veJ_QoaXK;Mi2SU=p<)Zshj6vkB|Y{GO)l|-V@3gR(T?oE2o`g1
z9Rb5h+A>3KG8CbPpd8JTR-ns|%_|%>n_^G?w8LntaKu&$gLA7(<N^;iLc#Q_6sSp%
z{ys7gtW;ZocqXXBu2&^U^F-SnS>((v>%f%;ZY=S7nQeoyJ#Uy<K}JPP>zN^MW$;;W
zk~b^&nMb)E<1_3?O1r?rxzrFq?L?wbAcv+2lScLiqAQMU3Q=68>=C6<((=BLpo$@z
zhV{7_+sWI3c@SWEL!4p2G1CC9{n~0%k7KU!$^JIjGRS^NX@mUOEo9~Gx>JB9d~^r+
znG3!LiIqo)(UZ=*nTm;Ic~1R+fh<vOfAD$bh^{~FxAKSkM7vz-jqp=rjQEhEmHu{>
z(i;bwr4X({Dx;^lch(oYmrh$0a^6;3lzXINGp*s)o+Vy1i{812)H41m6!{40*UA0#
z@b$NA7gje1w{}&??fBtf#o*w0)LR1LaCh`}jQ%cY2p4WHy__07qkuuQ#L+_bC@ZH@
zuC#3O<<0EN-sLSd1dp5J)Anto`%B_t=Hcp#!N>8v?f$kYf0f#Cs_(C1f~`s)FV8Hw
z6#iKL&d=A6+loQ@eOK`cf^Yk-w<oK{7B_FFItl*lnBmg#{aY|?R~<CQS-0Nm!<WWW
zOTB~r;E{?%j{%BmNeg#YJ-w_bU;UiyE>3M+-Kg>V!NFMzcl&VvFQ*<Qe1e)Iet!2m
zsPD~LU`o^Q@kUSy%ELPb+?}@<fx3$$zrBQ$5YQRm1ax00gg6X@Kb<Z;%k+ha)Jq>;
zAIsg}sP37+BZy!YSJ&s~`=86H78Lx_69{FS6-2EiYV2@S4diBBn-plbYM`Bjow2&L
za_H#wvid!RdxDSqUoOnReRJTQ-(HsW)weq|@BkPqn_q+ThpRA&JjIk|Deu4K6|JS6
zHot%m>+G(&-S3RNTxxmTYc)D?bG<*EK74M^&u+vxc)8#H0^`&KnpTW&wZ1n%6^tLQ
zM<xy)wrvi(Zqw>@a=tmdTE=n-S}RqDV(o%VyEK{)@(Pt6TQTj<jzi}Ly`u88CRnAk
zorb#7Ib|dkWKIRkvE~=^1u;NAxO)NN)@kQ}(iCGL*<Jdn{r2GlgVeB$MFjG4bp)Ex
z8uJgYg2_G@kVEVPHn8(svq4~xD%Lgib5C%QM<u*SPtk+x)z9)(-Y=hML<KTecJIz-
zT(-I$IDybEYf@5DX`@vU)ItHywJFr%opou_lH58dWCxik22pD9R5x>Jr(tQ;R^M?b
zq>)Lfa^Fi<F)5!ynN-Ro+nQnC3V)tlRW3o~^<<E+AmmnBS7GBRjk2LzK$6iI!2}KQ
zW%Q~4@Zt2~^-=?pz#0*0^zY~XCsTB}S&uPYYPY7kCIy^9>GXZiMq+C{T!KzXGaTM0
zBd<^|d-c9YQ)mXPdm$k!<1coakpYAo0J8Ny7-@JrmWlE0-zpxoaULp%UR2pfLJM8y
zlBDByq;^#v)-YH1t6mBq)1Z8XzJKuAW{d$DGWOhM3PEeI9r{f}jMO-N+b^9pO!WQ+
zE%DD~{j(T@aoF}RXB>n^(c8uM*2$X7LZx+{$%~{hjVeT-s^xlFLX*+;4dA&#>XL#?
zyDSB2UI$<9{s?|D^LnL&Y};&$r`a@Toq75)-JM-(nVM@q`e+6>`*w>x)C`_m>6gPo
z@i_@jQqjW0Rw`M1UTu4h8nE*$JW8tcnA7!a+~3oyc3NG^L7JRjkREdd&yKs1`Z03b
z5Z>89f|o?}-T8Kvz9PGB!q(T>xy_33EeCarWrI|=+!RjF*DJDzF40Ra&Da2+La}j~
z5~HdmCUwj7ipHDYlaS6#1Ph~H3KYzL=R3GRK|+AqAwEG+%Wg{)mGb<BMq9j`fS?)@
zD_@->1!a)8r;E7x+gfyVoYRz^fW9rJ{CxTHn)6PE#&zZKZG`~j7MF#d1ct>&%Uh87
zUDs9gPcMFF_+I)Ae=eEw^vI^{ozRY#g&7Ji9K@G>Prd@L&>*hF6GibsKn(9*;4rF8
znxks4=?`EDYt7gO+3xV86b29Mwf&U+lq|BK;0uY-0jCfsa~<i8N?PEq3sE4=1T+%0
z@Mi6|b7@$Y>{rrdWHPSf5hh8F@k2YbjY=-tmuQqfLQK%kK3*9r_3rCazY|eDq)SGk
z8bG*!*fO;Li;T@<AWA_$&>^7ggxkWBy3JqPM#1t#@pkBoFA!#aJ+JS@Ca>=lqHN}u
z$_X5`69hV2?!j86X?>3FdE+K97%&HzhYckO5J!**25D7|-&C`_N<ZA9kpF7=z$J_d
zq9{qQ>@UhAWmSgAqs_Nh#$Ag1@?q5kpDJrtz#oFPUp>lN&VS;5ec)g_GNHL9wJeN5
zsB4Ta)HxD`XPRiTwe5id>4MueoHMJIz`2bg28h{5I;w6y(wfOdABd5}A3PAc8Bv{(
zz4tUKcRsgVCTr>Ns1%uc#YE`ZJfjHy`op2XtYxF~TE}}pk_<xP#70a<+~|XDRgY}1
zTZ!I#Ap~R)V>;gc(+-P;dW4ToR}j}p0;pIz=^X-8_OUgetX1}9c<I`03Z?8XL=lg4
zSf?s{*Yy-7t{3z`o%EX3b2VG1D>qElu9#__)05xPX-cz9%O{kw$uf%Bq<N&7q`9P<
zAXB=hBwXYJa+BlWG)VjKWG(tLAy9-l!#mI*DJ&L@5UI+uW*kguv;@bKo^ftHf3*f8
z(;C<LJmeKWdU8co?sGt`_OS59pp(%JtDLHnqm3#;tHMzAQYL3&KZ;Yu;|a27M~5}{
zPF=GX<kb=#gVhl9NcN(MJ_Bd9<ofuDlLktM!Q-LjiYh0{?}x)2!5gg!Y6JTegj_<A
z+kj`EH<6%ysEo+o`f(ts`(k$1BinhNKR+ppWkVU^ncfO~NwSHSw4jfRL7WI#^ys~>
zAUsI+NHE-{NagCj?Sh?v=N(!L=rfIkv7s?G!VN>AJ=^2QB@X9WvuGfjwG|izvF;q6
z{h)X-2hUh(c3fSAhm*ZqK#)h%(2)3AP{ufP_IWjB*$*ILGu=g9T?#W4sh>n_Y8r*3
zacVC{5DBG$>qMT~mgR+iBkf8I&FOazMf*#j#m<9>SB_n-R7M;CeMWa%jlxHZJl7KB
z1WLhc=tLZdh$vO~rG~}B4;PEV3W1PGLw7HwXg)hC9NJH!`j7S~%*wSHVu_tx?lYOI
z0`V#mGwuD#X6b$(<NMOubDfg>LX;r7eEUx$E;G~r6(2Ei{>;7qP2c1acRU8C^|q?|
zq|?|RZgMZP=TBUE?XZ9i+mDP839HtzCh+1YZd)FmU^28#o0fH1E&yvr!XR<bqyC=;
zQ*M`xmxr&12XCi)yhYuFd5wX`Cl4^+&I=|70yeKEkO`?w^>zvWrXw9(SI$i_M=FXw
z20ypgGNLOA?i5fKFbB+yohkZG*Z^|4W6+Rm484D@HMXbZIee9?4po3(4KPJO;d3Z|
z_t)(JYG0PSFu^I{`|0ZOVZ2;iU$eimalX`+!XuBA;dFqs#Ym{_bh%qDSDwJ5=H3K%
zd4NAdzS}3gEj~6pFoy)Pl~Icw7c9m~HIXq#(gTgNY|Ei4o+&t~!qkhX5d@ohyMsd`
zsElJw=o!g0hfkbR2HEFTUZDM#V1sFx7SpLNBC?j(nALbxis?HE>l(QQit4?qhe_4&
zxaCoDhaaXx|DIADYMcznfJsq6*vH=;YeYrDdj;Coq2n<zuf=iVM<Fwl!E~Ni<uvah
zlFPXIl9YHc9|;K_cNxRUE{>ZzI9bh-B3e)U#5UCtij({t$7NGvOfLIo{MQ#B-DiLy
zeyI~T)Eu0|njjiKxy_008-0smN+z=n(sqvN7WhVhH%7g+m<8Y_#{O`cuz;Njx6X!S
zsE1^URG1d^Da9n*bI$I=q<|I@R>8mI&Acw&3f3}(3|@IDI>iyzd9tm3Vr_LV-b@m$
zf|whd<hnX%jw32wWZ0iL{JO`x!4BSs<ypHdwPkyX+*go~SOZ=~tuP{Xk?(z#5#^-C
z4nRw4B|-c*r9~AxFO97*cowLWG}44AnA<;X*Z>F(!T5G&1?s8U=z8}1FYa&|6lS!E
zJM1IM^!n<YsVCiR!rbVTE~nz%uYb8iQ2F}<Gxq|oSile+vc|)!Q_aOioqr&Hcx*~%
zd3te5?Amm5IQ3u!Y#lj~>0zn}JV=R<cx1eV2n1x};KvNf4i*hNT4kwV%Tp<lRU~^N
z3wli54hLYpc*ttQS^1??ioKdK@!j~eEf{BHy^ZzDV9$ZHRfY?@{pV7>$qQN2(0`fh
zlIx8AT25D{BkuSm%2Bnir<RxkH^;@?GzPNVqbr!)G^vRz6%|2nt(le(uGqZEIZ9$M
z^SgWaY7ep*Xc8$S3(*YY(o@YlQa~FSzoM2|iCF|n1f$T>I*u(;-QyndHF7+M8|6)7
zVw*?3C`h3SPNr6%h8uHM2ErUTf;CT6nQ@{IKl}u{C{HV&h0RyU316)6U}YLb2hoV;
zu77;5JZgZ_GH9afEp<{bn9&GH*#4~h{a|$DPLcB%$8scW@W$9BDLc<I*1dp0i<V6s
z!@<;>-+xfz$c*4}y$=kf>Y49#ZY#3Il}=pJn$_By(4bFB@{rfAeZHA#2gxkl!~cY7
zwjY${Vs>1bct0rFt2hRG_V?_WLD~aaTqpgle{N(vMgppFS4Q8P>z}EUd{S@PJ737b
z7=jZ$y}W;Zp~~R$gKyr0?C=e?kSrRL^GMIt9R|1q?yB4#A?sdFE;F;>q>lQ>BG(`u
zU`^f^&Im5T_^aR|Mg!PP>YTk%6Bqi=`}OWfloYdyWtNX0b^B4o?R~C##633^S;n#}
za;B@lPqPo_(C6Ei>v3pKe=8%|RQjab?Y&b?hyl%U^L|<u{M89clx?|)gXu{O&))jj
zA3x^zdV?+C_kN`S>?8_hf?t=|-d5!EMkUWLzu@gV17jC*{K2nxW7X^h!j&6=J@%KM
zm&_j#v`~-sDOQ|LQWjE)>}Sm^4tEb^V%Zl!I-c#=UiM%xRBeJ{nKYYW4@g*=Q`gB=
zpH=r*+`d=Qd2Bv?aOd()dp#xlf@PQ3_+5)C@f*b=Wl9*|xi8NMmAgY!R2K&Tl?@4g
z)TQ!IA;9K`bM&bS!i$<IP91QM)HjB7r{Hrvtd2D^i>;Y(M_j_2{0obU*Mf|+No*-)
zI=8FXS>JB){fxJ-6_hA_&ZJElW?H!dTCIgov_=|p#@r((`*VChaN&9I7W_NDbJg%p
z!Ty+x<lK)KhM)vyP~u#d=&9cZB7b>C>-vuZ0R!4lo_`@<SUTe_p`K|J5>|TnIqtQv
zlOc`k*AnxtM^A5On^Vh}l7$5LA@|Ny>4rg9c;4`|d9^Wh(2#5kSK;27$2GYqio~{c
zH|$LT%Y)_=>z&ZlR8Mw{8Iqh)@^84&XE0#rj6q)ddIu`Mnm)h8kJgF!xz&^5=L#<s
z73>`!B!cnPmI?Xje6pF<l`FMmwwO6)OLnnfy_H+^qd26id{sZ=t#|xHLOXQg0k6yQ
z3x>O3F<T+UMZUXYsn>*vf(dM{;|Y^!bXBLuNSrw-Cr8tb8B9=v0}PJFiWJPu5K>I8
zXP6rmrU_(P`rGcN%rCNY#iw#uP4FJk<EB$uyd1#N6SomPVegi@YtumJ`xxrFyBMX%
z(N%d*QBdGmljU(~d)6FJdzmTrDhU84*7^PPb{BlTMhFnmywn&R*vbWxig4sGG8FiF
zt;|h3IOD2{)LA$ziyhUkatb|$a?rJBkk`KyjI}3pbCRZ;@$UAVc2}WiZweXClZ;j-
z75!Ip<g``d^mZ<><j3D}YZP0VgeV44L{KT6>|!yKD9J)ZbQfvORzDpdvRi=9#K{G%
zu$-G9t|*`W@Ot%5hWiq3%pGMBRGs98<vy1lL(*R*=E5xw9zx|5gGX+i#&|U{+#ndi
zEQ^%PSj9bA&z-26!I0XRHty0c0ts<fe&$gYb=W!(oE$Qe{>3~$?=-TMN;|%c94${G
zf5%pPKjx#rfXys<@-EoB+x|j#nTG;^Ul@rng|rvoHiNts;x_C4wp~VOi(aO^%)=YN
zSrg=jqWO8tR2V&H{P8@Y@x3K^X(nzjtOW01NPEL;aYLBJ5!s=U8>GsTC1JtG`avR#
zI}H4IFw+!km5uU;dKG)6a-OGWhInFf%?}&1`Tk=%UeB`6+jx()kVQSnOU5f><Q?F}
zW{$#+w9j!QDL*B=YFG`7W)$)anH`bf;tC7Be4z{;tF5(I2KoH}I9r0^lpJU2F+*Gw
z>O9C_TmAkEFC)k7mCbloqx=F4FC&B{lQ8H9QGzf>t0nKwafHcRlwR$RY1C0r1cRg5
zqb8gHB?`h>LnLW;5UBEPEAeJ)CILdPXg>L^BRoe8kYj*-5}#|EW!ZS*nQX9W;?))H
z73M>k4r_T;2q#R%eLyhRfLaa)Ek@RC(fA>5?Mk;}RW?P2pRW&EB(ALj&;EX;qF>!A
zRyyizUyydWQi3+Wrh81_-Na2joA8?rrsS{mJoAx1n0EMF9EU&rm0Abt=B@bm-bdKa
zV5gl|rS6~3lQNsQel4FIi9tQs#>9v#q)noZB*^VQ3fCJ|MqN9qdFwi7WnA;^uEg+W
zKI^tmyHZ1Q7Hs#UUBiZj0lb_SQ;VEg*+5x#_<H#f(bo%r{!#lg^CxPCWUb~{XJeHG
znvmMhmE0oQu5W4`wOzP-a%Vwh8^zL<;e=1BWfj7b!DSl0b6jkZsgt6}rx5Kx<HPS{
zwf5C@bq+?oZZ90p+|P1q4_#WW5@x+WVrmz+*0ZxY2@56bbQ}uio#5Kym`OgM>%!gt
z<k6Zk8_G}liX9XDAMx^V!#2xYTZ^`8|7c!PHY#CzDA~L;nxdy|5;H_zd`Aze+S&_f
zu%-d)?s2l();OvPZ~73V{9xcj;i?q~d5*kc$-cVG55LG<TG?L+X7tL_;s7;EPU>7G
zS=O#>s!m@U<eZn^PRQpfkPOLL5tj~#B?-!_hae^1Z`q8l<{fY8$H<kEk;R+AtQI8?
zGX}cqtRfEkKV%lfKOtfQ_E!)F;uA9Qb!AHv2w&`!V;tmm%h*Pn+g-RO&evFjpE&<K
zxAgww?9TijG=7}S|6N!7@+Ul^^Z#adw^lNNf!!`3Arhr^rnpSVHgHa`|4ol5i>lds
zO1nx+#R<bTv=b0SHq!rAa(HbTIh@~|AK#Rp9-<QzDGkum>!O3vlX%gW#CouSmV&lj
zxVtlaoexY}>pzC(Q<E~>LEUi;vUP1^r|yFTXGV*nc5Xvd>kacq26b!*w1p;s?*@(L
zMeTrmp#%(~|FH%~RUOza<?sHul|$)OB8}z0we)#=ad;lzl<8`gqO4A#=fAoMW|c}S
z!$kErZ2x#$K0Mz%c3;%vQ1rJedY*#st3?R{sdy)a>!R+9KNhNtW02+eK=k>1vrZTD
z`2wS%Xl<>7?c=U@!wd#?Ff)cMs3$~#G;plt;bsj#Vp@tIgZL|Wjf#=E1KQV|hjxs-
zf6iDF^xFUi9dxd0ha0y%S&Tlt1af|%S7lzNG!J<=-42N&9S19M+N=N=elDv+5XuwA
ztZp?d_XQyO%d%imo%31=72fRk#KOohtb&|K<6bX#39+FV44gNMId4rNk*@#_uW`hz
z6d&3cfPN{VIKGbV7v1cbMMm+sM!d2Swv)HVh5G1BJYRB&Rc@Qq9ESOxxHsW;%dEfd
z!cHv>q?#V-ek@*EWuk1uDoyN%j9k}#gwzq?t^pYMutgt1#7mudHf*AY)1dTUL#^5&
z5d;E^)BSIz)qaU<lPN#YVRup1IcCUI^QW%zWY@b>m$n4d(JF>=#7QLbwT6a1uEvCD
zB4Kf64{C=ombzJ~H}ZP%wq8HmV)b{*9Y@&C-bkcA?lYxf(xs!OLaFR1D09|zohT42
z3i_At%c9Ib!OJwa=I)MPD~@8IifZY}E%YXpfEFNo6LjOAx0z+4v5&f*0eR^hcvwHz
z)<b<*1W+WIC%&HMS}8K4-K$bxulKi5dS68lps#Qwq@o}fn2tMfcPESzNc7a}hl^l&
zgb#&quwX=L#g_k%v3Cg1q;0!!!;X_qI<{@|iEZ1qZFg+jw%xI9+qRRPy@&gM2j8Sd
zRimooxK~|kUFS;F0|8y~ryI{&1ZIcmUO#cgmhKa&aE8bfH2fD9nTutBX*b9@_R{?v
zD5WeI=U#5K^QT&%_670=HaO<xG{>Ei@6)>Ny?w*G`cCTCS8of#oNbsQR0T+4PL`Pa
zh#MrjoPIsK0?_p~#wA8opvKDyftWguu%E;q6oD1W2JShA3Bg*_nty^HJTcdhdV5WO
zg9us3QFbzCNb{8`kt`9d-#-r_dc8qQWj9^bk9EvJ?{3C<z(fcVhMF9vK=IZnCCxae
zk81DS7^5ZOMe_p3D-}-Yk7&p;P{EM^bU0zb%kM)M#}(jU6qp8EC?M(krex^g)p*`0
z6Ea;x+-7w7RqF(T<`giX&Gh&=5E2r}`4wE|k9^AV5(xWd(X*Lb2I_$(>RQsyNy(m-
z=b;L-2807_YIW)~9pP1cCglRMN&D%|(MMK@=LvY{Ce>Fb)@QH1D4yZ<*n50LB$iT?
zotUNs?^>n?QxYhm`<indGbmc1Eos?Wl}aaW3b^_?V%715jUhDk6vbSpsnh|Ek;fVt
zfcBbsbsr_o>yes@PO{fo$^$8m(5x~Gn~F6{%jYtU6^A~PD@~ad>~i+q?G$^=!|``o
zeca$b={ABBhxne-R*8-n+Bu^Fd2dsbzn%)tj5!Xp6d`f1dr}z@dIg%Eeo1H)fTzWP
z<w(A=bcO@fn7lk6Ukp^+fjX{HBIZh^{cVU(+qV;z?Tn+2Y=KVR4xA^?cCAL9SwDMA
z-pH>GtKJ<Ea|G>YEwEmBpW&}fvp3FG8dMG>!SZ42u#MtwmNwrn1rw%3Qd}pqnvrvF
ztp&&H`;`$dEzoQtr+kWh)j%7T5V?#BWtmZ&n_`^X0mRJC#mBCt2#E5R<AM*`6~JHE
zg1-~JWc-kS_osKtc&|pqFW0hI1i6lHD9+Z*Cz5I}YRx3qDkciG2S?RhIUyZ&lm&>I
zU{RidI{z6H-(+5+8daF$NU_%z`sp~c7NnQsnZl_l?+gbw^ic7`Xz0T)&Lq%cDS0z`
zMKgts#sx>(ehXbHo}ALc4}~2pXh1|cv5fp%ni~06MBNgZZ!H7!ox2x(s?3IFy9*t*
zSB1QE@Rz%#CpxBdQ0L)3q##_NH#d|;w|Lbs;tKIf|K5^;U!CB+BSKa%Y}Ncg@?+>r
zWk{7sy4e9^<IuJvncX%O)OgKSt;E7LY{a;dYMx1fw-p2KZGM6RL%jn+QqWx!N#(R4
z$6)bMX!73v%wOhBnII1(E^W+UZ-L)Q;<e7U)=OS}wev(w8bK^Vw6M^@j%U$=;}3O?
zLA{Ipndf64fGts%QTD_vXCrDBZ5Vlu1?eb#T5cvCad>vq<dx}m8bf=-Pqi!H<`-7w
zwvKTJJOZaAK55WlVcnyKrnd^f%2Q5x;#6RgunoR-OT)g2!RfKI2M_#NjYV)jqrhOJ
zMesT~^+5W!2;Bys7k^i`E}fT#eLtSf6^|#vt?)FJ?Bob1mSzh7WW)AcLP`JsYvXjt
zO1z+}zR=ka+g#Hf52<#>(^})hq%|RLH<DDCLfwafA#c<wGF%gQ|HV^zbsyw8ZFnO4
z_y#_uio$irQP%*^OIk<zK_J|G#EIR8b*Xm<jL|#HXfC4))M!HiPeU>O+8l#Sy0QDF
zgjnW>DS7Oq%qG)s5?B5xX5hQ?gZ}u5tzQ$MEcJ>_Y8Dv~Fy!MzR^W>N=tC!g;zjL1
z%x#?J*`GzuQ~J3MNN~rjO1Yt$tZMdu>zwpOwsFdJ4G)Gb1u=%f?Ojgl4l{Rg011p8
zM>_imzly0bul2!&xnpjmE@@%XWT?@ztOaNrM1pDy?iJ`;F{q<G9gm2a1drC_C6qi-
z1<8WOR)*N3G*W@1qv<xp)NSSYk%6;&2Ixxgdw*R_jCH^acbN@@2g!B@bO3v?@|wbL
zPqEX(XwUsNGZ7!eaTKJj$h(x&XNPdU-*Y$zSAN4q-V`PH)MdRlKJV|_Nt;DY>-d-f
z%IN$){rJOboTu6mMGzEfHen7O;AEmaH9=2?De#MVc<H^(L>MYc-36rUS!{tmr4e^$
zJgtst2|c_Ep7yrz^D@vxi|~VibxIc23?))!q~%rx3hyn7#_{yci<dFlNe=@~R+ly~
zo9EihJ?^|I=<Z%HHJp7i<61k9;a#VI&=%8vO@-Yswm}F&A7@r>xX{Ea>t57w+KS8X
zhTU%~)<hk(>n%F0t>gyI`&QPB2Yl6|%IIOsx(9s6DUZobm5m`6{+#x;R?{#<x`Um;
zbTf9pDiKZG4fZxCLUblWx15A=1Cx{(W4bV4S$dSt^uLujm-kb*@9y15nw<u%#2H7`
zn|2rlj~1^B<F3PZ)x3IP87Tesyc&Ywhe%du|7?g=31sK2t#Oyj3vf>%Y2ZSy&(GsN
zEl0)GI`^3kfNWmQ9?IumhXg&AHtbH@dyF=!H3z=D^g8DzmYJpkuVNU;oM+Gsc+6u3
z7it#gfj*Zo9hOxK5djm8s9C}+=hSD<GwM3V>=j|<J>qg(!^x0iNhCh|)g6PLwhF1!
zO!-%|tj@AYA-yIBRc9RyQCr8Bkteli+!h6M1+!WEx=Wl&QL(hg=YpT^^1h4>)mFE3
zrufE_a6SF&KwZl-;+sxS3(>A&HH;Nsqy0k5Zg+KeZ~Kb$fD8+tj;8}6*mqMNr)cL4
zG_sM{yCxZKpk-~PC*}1|9wlnu`3Kyu%38408xz~73dl@Ro#RQTl2D$t8W~;h6P~qU
z%@5YGRa@_q&t{(2_9x5Eg?Add$BG6Oyk{Q!@lZkJwUAH2vc1Bsk!5ysSWOwsPZ}9a
z?`r}3K|9zM68=5#NoF*$*&Y|+BU_mha39DiwC^d3kc?`f*_;Y)+wuj^y!EJ~M+}ZQ
z-RhXRq#PF<QYV+C>ZwZE!{c<unwwU4j)soHS6e*f)bjt-@juPm|9_$YBO^2Y|2BjF
z*)POmL+U<Pb+YN;3(=2L_5up2&9-O$|EBO4`uYP~fjGpg{a>HRx|_ruBuyq7PwZ*H
z%vH2;j|R?lYB|KAF@e}ut}O^1c<`7IRM`H?{r#c9FQB8}{EV|RdcZ-m^xR&!K0YKx
zj34cXXcA>CJ=8s}UPYf#zlMqYM1d56pV(0SDbgv@hspPTc%FODtn={h-Mg<cgumb?
zu-x$Jam>E2LvvghsWvQ4siBwh%+T18Bn4AUpoFK&_@NI{<_kl4Q>QcEuChOO^kxyl
z)%x=BS%I%#oW+)J0aU>Mt{G<cO+Ro09WDSyXmW0+-Zye-WAGBoY-!S#pLI<J!_WsT
zXNi#k%4%rit}pEy!O8-7Z58A^<b(x<hB|UO4Lznvdt;*%NDtbKG?h2#IIiYWo!uA(
zB;uOQqF7Xkv!WLqMJ3HD=t@jH4Ckiv?+oGE;RA3^A_YNNWw5pm_#S_j=(Vmoe*S-4
zdZzc%k)iKw+`iq>06T*y>4K<@C?|F=`zgEqSJu<4qG8DK$A!MciUVwkuCKk%@o;fN
zzgd70KIDGGhw$Yp3?dn^#b)I52sH%Ia_{F)aO58#3-Amxpllf9>Ce=BL|FVhx418>
zVdf$`jHt(<{u=_E_ps}uVfyCK!yO6tQKD<b1$b+U+nd!iQvzM8Rr-M*)p~+16-@?J
zH76#vVK<LEW`L%@;E+8{?!mUTF54sZOBf4NqKjJC)Gi?Q2skx>fcPR71G5C;SYRys
zt?*uv&=m92E$H#QQ-r=s*(r%*h_kLTnYC60;Lh<z8Vdc=cs$nsH6+GvpxG)aVTr<H
zLFo^e4ftdj$%X#cp>k?WlV2~4c!ZP)1;c7xn7PJoiYa4%a8V{a17IqmNX_31HFwRT
zN+<MAs%8?Fb)Rh63KyX8oOcaRWLWT^os+`BlwZX^O8i@^432oKjDZT9k0icNcs1OG
zt+L+L$afG$KSo0%=PsZy7%SkSh0GD`TE`})zPr1c&@JOj)bSBOZ}-)dtd&vcmjKV~
z{~UAuoyO`oa7ZCh+v+mrYCI-R(e$h$VO~s3kz&+V;Lc*o<3JpFYgioZkGv$<l}_0F
zl5(%6QwC>QTykKw7cka`c$o3e9NAUfJt?X0r-eDYI&rpdt`w!Kx;%6}f4J{g^5c+C
zg8gB_C2q{pP|1pf){2Mo&{Bi*C<P4NHU`})s-aF|XacRz!&GYg)gjiD!JhEWeyPtY
zM$Y9_>1$kV>*kc(?3Z%r@LE<;cD=)Qxs?`GQTE)mt|XZf;O+@Rhj@6|ZRgW_$NyOn
zdlL2~Q?&m(R17s2lBa<QXbG%2{aXNc{qtsh7~$LF8;{lC<)w!{QFc-Q@rL5ov0GW|
z!_z!O_7ykyzLFJ^u8AMe3q$6QfN#7d=Ei#wD+t8g7XC&d=gl5-ziOS19dpL|Q&>VJ
zF^f;1S*pHKlFm-aA|XbKOw*^m6Mr0{nXPLQyna)DJua91xUK)JztS}cAStlA2pSz&
z!TQp?qsUR{wJ8165F3LG9!uRrwtZ#b5vaP%3W45-&B>JK&?mX+kY6)%%u{kfL%OHL
zhBjBKD!{uHG^J83e)7jh2wV!l1C5p)nUaRj-s<_zEb%@Ne_mv%YXcs{9)V>$Nt$6C
z^0?@~QEXaqlj<KonH!^Aor;;PB{V<1Jxi?}X%Q?^V7xSGyu7q<wQH|csksr|ek}da
zd*wf}Qu?qfMAViVDUiH7tgKcZJ+F4+FuZa$?!oBDN(`^ufRYwIC<73()7~^*2mAN~
z^6N7}ke}k$b!e@(k5#syPR&W5rCv$wl|wa#ai=G8v<<B&>`Ktpp*JX?6%W2R8~g9B
z92BkTsOS-Bq1;RvmPJvL!K(OB3=ml;lZKf$d&~AbiSzvi%+Wo;TZE7^j~f9AH{jXy
z)^IuGn~3W&`7qHjt6HNEoVpO;Pz+Q*>*#7?{O()8ulqHS6!`2@)>zGvnju&~&@}x_
zZUSVV9(H|{RnX}(@PHi3B@igX7*9ide$RfBbGU;1eK&A?ey+U$C|$px8Yp73h|IQ=
zBJDe^b&@qR{;`ikzaY*Vm1ePYigQecrspMzvJxu<Vl#z+$S!m-Hx2iPLhdO12`3jH
z(i32;NmI=47Cr=KH~kM7H#_GWK1s5R2cE8!&6~E%kTFgDsyB%pejHp&*94ETcAko6
z`=UgXF``~ja1A$C<r}|P1ZA0W5ZqYAAxuD03Mv=glHf(D{hpZFOkmtvF2wu+A`m*b
zQel3-E<9jH=7wNeF9M2Mu}A&|NSyp8V+SI<Q*N$$0AcKeY!@PFq=@yxXhRqX1Q7tm
zCl{`srvZguI&*s0{6*@yAqB78Ly9{iO9ZL~Ml@_RTqS!<;+%38cqM?VdsfB_msIen
zhVwSKAW8E!EG>6wn<&{8xDb5#4lo)*bfNrds>t)o8OlD7pp;Q@DBf)9#+l;!<$qQb
zyp2U1<-8wtLevAHcrk@SBI(i8DWgn$Q>x~$e-c&H#1$m#eZJfbHiDSC?_9><4DqHS
zJ@~1{SMFpq+B$t+-s47|^{kLNAl8~;s8j7#rlwjXd?m5>H%_|Cds$fQh@Q)(ve&j;
z^a0W*A>p6-NAcRO>JHmZ6l6!=;%x>AzoidKyydFHO&XigGszNe@aBk|&EuH%-_1n(
zRLN<E>SRXYKm24&!6Lv&Cec*Gr8M=RHQ9_^LBsi^AyiCNrn%yObq#wU&H9u>|IzfA
zZZmEGF`Hv#150zBE`v&|m@D%oOpTi6OgH(7OU>5>*W&1b42V0<;~hC+*8ru0i5oq@
z=-sgZ|Jm7(k9(OT^r)sG1R|TbyH4mf|5pg#nPi=3GS+^ic-hsYa7M8XN}BhVeo6~E
zl{+8PL&D}rV3hKGf*wh<FWnZB&^;&&H8`*;Htyx3;%sBdg-NDG>77co&lg0JYjyKK
z<q*q%bYd{E{%^tBOHJ_@%^x}RTq9@m2Y*sOZV=84@+6rCv~BQYT~8*ExCEQ@`<s5y
z+dU%tfp~rNCD#Y_QLT~^wMRqkD0Ca)<L31A^7AyuGz_A@5`+l;x(CP4XXgQp9nQj;
z9!N@(z4P1tWf|sJVE#A?Fw&Oz!0=!zDeLT&tal2R%jPEn96kNZUxYYH%tMsi!{H6`
ztwr4Mm|1t;YaEdWJRk0XP)+vo<@z}%$IZa)KnBZbK=4Rr%rMtah~NDCRx6Dg<Pb}q
z)TA|E8QDHxCe4%Nj68c!bN7{ClF}i>ozP)8s9pt|UGPn<>|j@R&z1A-_0Q>gUuHKr
z`dmY{m=P77n@s1yHLC>F&TIKL!!%+6w`dOz6JZmj4kIK54G|ami=@Vghxg=mXUGir
z9!R~8yPBCe=o|<WyF8Fk;BDWFDMPy=bMeiGTqHd$Jst2WkfVoq5X+JraKfLZzWC01
zVp&&xMW|k|bbFSUSgT3%9>uR)^OvXF&(od+qWmC}oq=6mU+z~txp;D;Bz!o*h)`Z%
zo{y)8aWU3a5Yx<~0p4)Yf-4aK8LWLK@va8dqCRBoCe>U<e58y+GMG5Rn2P6<sp3Dq
zAbp}@WOJA%C8EIq%928&hmS*E|4W+#$3S8saZWA;u+(83;gUr7yVkEkMGqcb8HP{?
zB3)uG>w1jstq^F(SUyl&D5B5Z`|DLdTpey70Kk$~CqE6s?2)c>P`Xt}&0A`0hE_9P
zP99@Z3~1cFKi&sRFHxm+L}!Pc0etEQ{nAWQMIf0fN3sa6eUtD_yhMi=5n&oIML^ck
zv7dHA@#J|-;kxiw*1gm>0CrQ%G9t~CUnOK-5Dmf4RS>{^Q&|xG^!o`n-%zb3>LfkD
zojTS>niOm?F`)PX=~S$~p_7GZ)CCTv9-OXlPX#>q%!D6pxbtJzj{gk~3#ZZnF?BK$
zEF<Dw`HAm$W<tTt219LaGt?YOe)*b$-iOw#-k2HpW`<i9eS0y>vLzLDc|7apKQSNC
z$8;FOc9y7rmAWR@P<RDzBz+}_6EKDKB&Pt@u}rH3Pn*ER1}R8}7XH_u0X_0`t5sy3
zNv`p8)z#eHQU2|ro+Hps{6X9ai~=hJH83h%E3E$aVOicG$Y1t8;d+f)yoaIO#?DD9
zo>;1<#RJOKz-_{6(zzugG>|kfaybTsh@ZO6D|?{UTV+|jySS+;B21(9{`*@+rZx;~
ziaSa_kva#82cPQ3d~^+z=ZCc*aoe;QjFP06q1lENq`-|!sRzcZIAsFx)wh63EQLv9
zs2bPM<(EF75b0D`Raf-5Cl5NNh-PvO==)6^pN1<r=#aIoX&jxyMveQGGnq9CtSl9x
z{Uzl!ih}uc$w9is(-{tvyrokqmH@qnVz?nmxO0|3b4{EI7Myp5=k`op9RAAJ#d+pn
zlk%4bcT-8O{^ur?7W(q_H_cf3=|7{9&fI~HAK>s~9a*RjM?N;M0<M)>&zv4re(^ZI
z8uHKZtWiguSNrlI&_!8h2Z`nGY)kIHVVQP@IAefKh^;8^OB4H+yjWwbsP1#?CR(d2
ziv_HkT5KdP{~Ts=xe{3tsP`c+JXErnSDOZAC<eD<bC)iu=<S@>!_!joo&M8>?_5*f
zFI!EVHl>KW0X^Wc&?*z4m(wb@Cty4$9p}v*FeU-rZ>|@%XGi)4Z@XPH;5<=^BB52H
zu-ao(mEfH)lG#`0JT^dibLBJ<+^yT|5^OY;bkiA>eAK$lCN8Q*YzfO(zI0X8#SYmK
zh={ug&HZC{KgO4+<x=vG@X#9lTieMKJz~=tv8o7KB)=b$#E3O=B#P1Jv8w3_Xmd_s
zGAv$XUpjgD^{nL(8AhR!!ji4VNFjHH&(fPe6}=l(!GWH6mP1G^DEN6%m3EddjyWd4
zAEP85U|O~zS(pHl*n~xJBMbq^bM*XduzE#04ygB#?menCP=`8~Fy;*uS`|1S(OM+@
zWc(FmkDRVIcT};Mu8jDXzxDLdwHsP`r5KYV=_CvRdZezj-g=?TJ$yaDJe}}Yvaj`6
z+?y3K)*q6i>f~BO3nlolN<^}Q_wg4IkTQ=9ZqG;={OtikgJX<X5euPm#RJNDG7>A6
zKAPH+_Kq{0x(D>4<Bm`Ra``F6arY$5NxKHB&<*`ZM&)=c5SFXK0vpBA9(gnF)mPxJ
z<bY^ceFY}tTNTK26V;J`1<>7VkE!$8B7b0EJ5Uk?&BT=F+9VMrEMy8^jr_>YW6~Zj
z1VZDf#3xw4Lf!cL)CsMw*OG-n=^#)1WC^l<i2}~i1;iVP7l{C}7m00+nALxOW#|bj
zEz=d^>uEjK+{ONkG}an6xl(M>W}f>~aZc?ORbe~dFndJc;4_*JV6cm6amYLRi}Rh<
zbewnD3J6CMGg%zf>-Hn_)^IuZK^8oU`RESp)*J;^ISUOla;qqx-qB$A$o;Kma6hSp
z4o7=kzIf0^gXSd)&)b`olfjY~UAwem!)ln;yxz(aQ<7>vvNuI%S?$4oOpml){6_C{
zvJMGj^Q+%WB)^a1iXaijjZ#;`lUSOd>J_dOS^V5o#F`s{BpsTI%{|}`qniUJZq%3a
ze}dnH2OTW33AG$UL{<w#B{J?kCIS_#wbb+SluaT<VgD5_83$G5wYE0;_DnYqxtq%5
zLZmA|_{X7qP@8W@lI3$%+*?h+oHVFC^eHOl&!d>V_`Caa=p$aP_wO$u8^dw7IB4i9
z(*?#ZxXsOAlGxwqWD+Z@AtaIbN=QVD)o|zBwy>qCGjG{LGm4!?u>Ef_Od--%D2A!!
z#%CwR5|iwb$Le<uy&7lv%evt(V@1SkbQ=KFkz%ZE@j86(?G?b2@S2^?9Ccc7<JB5!
z5|&1VtNp0|$2yBP1r@_8u87z>EZm1Pr@T$vEPFBEMg?dQYSmC6hbv-T1;bK7n*}Z5
zQ)oEoAQy840x}(QM9evFwYnFUZkxK7+q>-+H9*X>#N^ihEHHhLpE8_&f@fMt)}Fz;
z9oQX=5p;X=<8zZ+uG=Q@9mY8y1Z~0x6gS7gIO4{4*2CgdfAD!~KGZeaVmLE)8%wrl
zbx2XNjM1m4dwz($`Aav<WN4^*o^ofYc-~|vIeY&0_2>NdMdS}t{ZBo{`X7t}Mvni@
zX71AXiHLR}Zk|)UK@5x=+(K!=GRB+!8krip2%rWbazS;8^=ufqE_v~BYH2ra;_(x)
z2&Yy-k>wEvcxSylxU_fX_;?C)V3?qwGj#c;`E~C+;QeSZ77U<b#Ny}c`E8$LEi(`8
z6A}*5!j!z}<gLYdjNUoeTnxSfAX!r)f00;`*ds{c>r9;hKnC+u`jVxFZCheC$n39l
z$R$OOxA%8lS#E}jwE3?BU#uLREm_}8k*B7($ka@LFYhZIF^C`)>0oBve9m`fcP1cO
zD5b@MrTBs-%A5jSEQRH$sZh+py~s5;r7Y*rG1lfBwlP-9j{;eoF5HRITEkl3zDbVw
ztyOJurEM|*<QJxik{Be(M_h(+IutDbiliP(l-RNA+pBEuCoD)1p+7M&|Ck7m4}2m{
zkgj^Ri4rIZ@33z|DIyT4(^W96#o)g+L`DJ3AvhcVkooi%+1u0UntA#=3jPuMIo4&3
zGMV*%r;rDQ92hug>g#uc1qGH134T-bNRUU!BxLu>M1yVJ@zsULOy4jUu0#Vqd6g^#
zf{}^f*DsU}qwr1s_Ad4i%kR%EA>x^*(0cB^q75f<!=Ki|T1z(W@Fp&7x&o~`T2Hqm
zu@}&T@j4>h6)6J|E*NqG!aD77vbGg3NMj5I8y6E=Yy3)BU~G=vmgs+$>Nw0s`R!#y
z#>A6i)wi3YY3AjIX8FfJr44gSX`cn<IqzPWww9#$F^W;58}Jyw1&MM+!!3(rrR*09
zrR?tMuD!HD0?X5rD9zV~8no?*06X{8AmZxZ;E3&UDnVYIYm9>U*%X5jf;nudNF24K
zAqymoXNutyi#XLDe<Br_#wX+SE+_yX(7~`-I26_Xgn{)rIP;Mwh{*_<zjSh@jNW7H
zCELIc(JaVX1b`9_9z?9Ib3qBbr>TOvJj;>5IuF#$xTOd_VK!We;GRdo1p)Kf+arqD
zlVKIYWxv7qT|&wYq#<^a>e($f<J>MOaQ&CF_fT~p!hVsFI!_T3B@1qq&vlX#<QQ_X
z*F)RN_u2!kM41ewX5~Y}Bo6mWE#k$SmyfWLGH~hf?)(K#>wnUcbny|-(Dy>M)=xl>
zhq`2ZMY!$|=7`8$*Bq`p7+M^sVFelO%DjjmL<F|RKb`uLci!Oa$>jPhtR=zWz+|uQ
zL?_4H`SmLt%SJEtbG&~EVB@%g*+hc$fkaqnGp3h<f4YYO9&}PC6}^O5d5ReYlK>(F
zrYk4l0Bvmm5bEA1gfmaKFNqFPlf8xf`-t3vzcl^z_<McO+J7at3g>q}HKNk#5xYAd
z(!&b=%5?%5zdJp3mZuq@xO2>rM!_b97cb1v93z#8>BB!?!wjK%?{v7oO(=$tE8;k%
zkxBL%&p?$X>k&ao^F|+1_4%e8-L9{CMv7b&=8hrIM=!JxjuvBVZsDmFs)Vy4b`M9I
zJ1Vp=!m|KH3VK7cf%+^}wew|~CF$=}&p0VA0G@fPs1$1$G07;HI_*yKVrO|{)s!)P
z`|J?yzHJjmR0&p<bGQ4`_$S&Ns)O1P)G#Y+tEqke%F5Oe#87nZvC>>#e(7%8(&=eh
zUkR6YVwHHgdXaI*sJKXhGt!~GoN*8HnYq#2bL7(Lixh^cuvN8rP2r<8wiMM#8aJ8j
zXv?PK)*_zmbTX8HD4+O_(&YxWHinxv3X6;|V)XM4(k5EJZ+YzO!a2gYUI|w**qP=Z
zidn4F6tKiQEUtR>!m(2k+nLS0Mp#$pcP$-ud~%LSYkP=~v;FV2&UvPnC%Cqz(NLf!
zPS)T(rTuxp-)^WA`UDDyL)os@a7APFj&>WYUf$Sqh8V`JhfOt4Gn6i`;m5lR-BVl8
zFN2yVmUL|^OjO!Bx%?o=tA}rjT=+&u;F03F{k3E5KOgRoIN+H%)gvFjGDe=_TI13=
zkZEijdI=|KVaDBJP&=}s<is#Rk<5(n`=bJQNz-3ggI9(8xa-ot*n;(ciTQRLR~8W<
z!6r4YMYz9R^6K_{9NzQ#&DMXvl_4*Z3j^QnZ|@i6UnHKPGk8l*r<)hZ%l^Yw56(<9
z-l7*$SNjE!C8L17l%bHt`8*QENl<~FrG~Ftc=<`mnvmb`P$d7PWLsS59hF-k8Bth2
z5`L<U-$XTXR9{8EEpqOaG#+bCAB;J@r8JZ^5`2l6)%g7r`^xGRXw8!pROlhAOzwYM
zkY*{8S`K3Ly^NVr*<w7%ODH-e^f3yW4VPW07)=pHZMgq6Glq62?}^!`@<2B#io@JX
zb7+*2uRG=zA~0vOX!ZH};Hi3-sMtz>7CVLoNmKIRgt5U;z>sOQ+4L5lzrV1`<G?no
zZR#A4cZLHq>-a&~$t0d@fxowEVjOUfWL(J&rKTv6%;?KaUWFs6NF*}jP`j&F@`4^O
ztd>3yqdJIP17?GXh1qN_oOkPIBz6r({{~~b_S|>J01?CWM@mc_CHz?K)~yA$&M$Dq
zUUMv)&_zP%%G1)(?i`hNf2Lv~?1_$t7i{Wm6hjwmwyed-OelS9QoSb2NsD06m2KAh
zPs-T~mdjh+@R$1zVH=_LFdD~Ewx~mO>34_JlBHVH4Via+Wj`B{O{S$YH|<a1y?kLF
zyGo*0N?BL6tP0+#qToCnYph;JTGCO0q6unJ2TJT_8RX`rgPn!m`xx`=oA-C$+{Sn@
zKkOj-8AYqT^q3#%;T@sVDdDb>L$x(v+DM6duJ+cKeLz%J00lTv&H2N=RHA`V_jT3s
z`E+>5)8`*%W?Mlln{=C12(-w<j+LLAYHvkp?RV%OkCLw@1-0_+6P2N?f`BeVnv%C<
z#S;|7B<km-ng_%pSyKvnGPE||styrl?{?<rV|pB?<rQd~=lCY&CFA9?xsT)SRD|^C
z^ZS~$W)9a)Gqc?<i@bOr_fa2wvo^DrrOl<nFub!;bHQCxpRPsKptBUGg-WAexBo&k
zWN-d;RPp@$!&H49fZ8-afw9px!*_gldRBF`jxB2(bH0YPXoSvl9w=DKI^N2h=NG|V
z8~y!;_*Ho|_MbY6?f+K0{D8yR|5uyO7K`n7_q*x^9%Mg8H?SZd@jAGN<&Ng8;_N}9
z;QcoO4!~8sx|1r*Wuq}PsZ;C9?qxEA=>GJ4phWj7dfF8)CnyuP0UB0jT9-5zA`{?0
zpR1Aptf%qs;lujvGuv^&3FkCE6NWz}uZPCFq8|b{*#u}32os#$iyp~LXy$J&N9jq%
zeDK51Kuh-2Tl!j)wy1=!(VsNnad|i9Sohk^AeE`)Gv}+K%hRj-{k_^)7uras@8P(G
z>zfZ+uWv7Cj^1YA;j`v+eAAX@wI031buV;p{<%hx>5hZogq>idNu^xiH+wH7SryKo
zFL@z3fL*njX}k-qtkbLwV!5`DSADI7yhD7j1x2}`SBChs>KEjN3emuo)F?wNj_n8)
zV9d%mRV6=5+sOR~1NWmOG{b?;&#^`u{o%z0WYV1w@x}%jicE}(B*gygWj-X^wDmp6
z#P;=8ke8g^q&C;lJW9tUc!=wZgm)IZ!Thv5?pjISVn3)qJX#eId5vWTCjH8W$72p%
zZxyD8<-a^YMND0&7PJ&(04qH9QbDdkyoX<4*bu%NwYB?oN@S-SMcs;qowkGH;giSr
zgV+}Ey9=M`a>gvlDNO}oQ{cr4fAdIDQe1Nof4o4XEh}z4j8FGwqZL0hLxawrbn}GL
zASasB?`+81iAObYs+%<%KRIjO+oyb4SsPe@A=ex{v7opQb#`<Lz*r7r#`u$lx<$$e
zjEwNxC0r_3Z_KIF&;f*>o705(E)^TZZB}b4B_!6X(W^}PV_Z|Gs5U{vPAIVd4_f3u
zJo7J>!5uLkpS|##MA(jEw<$QYNj^iacTzUJw0;}W;|)I)*htNFU#L)Ad1$_~hqXL~
zxGlT31V^Q4j757|1vvyQGH>$dUJ2Ucadi>Mi-U@wi4eNDCE2{Sd(_O+(Pdnwc>eSH
zN7C*FP}*cE41szyB&VG&O!wXZF!Ru+E*F!afQW5ljy{Sd{k_Ni@EM9WV~v^>Dbryz
z`NvV@bQv2uPX-F&Kp(>rRePhK4s?q{KG7vPC4(u<Wy8n5zYU|5Igm#&wKWz-r&f0W
z@@INA6czWuo2+ZR4~s$NlTy1?Y=@9@NylruzNE4jg>_-4zOJiDL7pLw-bOoC0=v50
z5><9?<};^J9Td<j(nA~s;}-QSA^DO<y028lz~J&CMBTlx;TNGhEI091nEuK}aM!E{
zQDJT9_$bn>bD%$UZBU_&th>7d4yX2guuuoI?zGI4-`)U>aAMc8(M!rc631UIWq&V}
z4%D)5(M%S@6OVa<r3zGgG=pg<WOpR4Xv7#KIor_iW;x}(C9c=iXb=2M_w6!<y8TxQ
z`G+>Ws2BM!UWH2WQBCo?d02mEC#FYn2bz;liQ8<L((C#-!^ZY8fO%%kO`98fIYb4#
zwjLQL?XSGWS@xbC8XAg|S`A!m$w2(>z3TL>|EgTsn8uCaR^jBKZPO;tyyq5%27R`c
z@%}q46P!Mw)OV}2wO7)4SwB>r-&tv`=^;K)S}B~vGHWc1>d5xiMy>*Zsxik}#8u1c
zctd=^#9%m;6k*!Rt!+bXsJ*7((emr^mxEL1LoKcxqi~Xdr&>)JvEvPC=wl-;$qW6A
zB{Q-rd@dcWyyEP-96*(@B%$2oiH4;9@i-nw<ygY#n8{+j>Ki(klv4uhLZ&+w8U?vE
zL^5NwVmJe4Sj(n`kqax%^%PxW*rVt%OdHULt7a9~L6Go=NV=fdO%SgluhHIYV*=OC
zY(s8?o!KVNtHqm|p96}9N8x~F5ZJpiWKRUIBJK0YyV85(2%r+N6mw<B@^Nut*kl`Z
zZYUoz1SER&LNj>&t_D{`lcF?@yBlv2u1Q<A`rL75*rbWEJZwR3`%ss@+W)ThE|mPP
zRu#~tgI95Asf$-J=<;z(#~HsjqSZxkQqB66J`-}eoB)yPBW>4zcRsr8THP6eZUD!j
zywwD=5Fba_j2=x#>j1ein6#5lVXRTjdj2RG@hAvDZkzk(xL>-uAQDRU7<`tZ)884T
zcOueUh81qRu*ol7PBh?~-w_Cn4({hVyLo#ItZ<4qx5;mF6872C)sRoLGs1=W{HMuP
zfO_%8vBi~75p`ma%lz68%;aLEAS^k*V{Lc7IZgPp#{7`{U4d#CjW=THaO(&xC#iN|
z#c(72*`>81U#e{FsE5KVq(PEmiGLRv`1%>Gwo%K2NG%R)>HM~bY93QVgpSX#%CdGg
z=XQ4Gr{XC>#F|sDa_ZA(*T)~4HYbD;FbR2m`dW>DnB3baS86}uN(#=(U}+cXp1eBF
zPd^>6{tc7lP^x;AdBzqmQ6(v%@W~Ij))21>s9SiaH1FYmL<yEZs1hj(FKj}fOaPk_
zi41!+j3x)~2tF|M@Z#&tX-Pd#ym=b4y+D1dJGms?P%g3!Hn_N?eUKWgw%HS-{@C8L
zbHsp?eB*vLLwi&44A4>sZ+A)chAIC@`i9eidyuKRo$XHai;#BzU2r)}y~B8)5>Cl%
z<qnN9r7*!65NEjxtoM8MrT_X?Pwz$9eroCcjpD)@fqfS(e5z?DvpxT3OaR#a1hwcw
z4gb$WpZ)(H`i!jX|C^cBr8#Ot%!br;PIU&6PKwd{V`arkD1bUT3bOT+^=Bs_?0@r0
z*-0}x&?qM(6e*Z62G(&8i6nP^JQLFKX6x|g_`CvjvhPMYyxqMn3(taz`z94;ENNvb
z^aZqgeR;oLb|kLN9VY?ENr1euKd(>MxqU$o&4cn0!>=eGKPL0?x%gcDhRg)F;{oYk
zu7_Kq?%Q+kI)oz7LYlLD!q3WXU!V89vi0;6Z4L0?w==CU(|RZCGq54-bNHvwHBLn*
zf?nw9XbieU+C-BCyk^Ah%y+Mv2k3&HDMNdr9|K&s>iLEDE%9L;x`0lN1${om%wsJ9
za_-a#&Gi06_k0=Xo|C&KcbW!jO$jVIW=eo$V)dOmi7b;rNu1mCQ+@oB)0|(<6{TAh
zrsZ0AvP`=d&QLPP?6GTnNR4o~K0vwhX9q<2ZP-ZRFv{v$q+coAJX+!hQY*SimNk82
z(iE%1hNH`5%qNDl@Ew`;(wkzj`c>u+1#0f(fjKM7cK3&kAn3Q;_G6}p4qkv9pO?7*
z!>!)V9YR&=>t+tG%4|8o*imbweG;jpRs{C)Rl8k*lFq7Q_r1k%ERSgO=6&vq@%>|j
z58R7hpd36$|Lm<E<aUE&IndZx7dT&59VXw{NXk|w&%jt#X|kOD_7UqxmBqxKh>0GP
z0{JGY)4l+z&Nik}k9jBx_DuEMi<>bt`?xo`WX<S$cBhgzgJ^8FNq-B~cV+%~B<WE7
z!XIi#!lN~Ems~NA@&)T@W6gutsgR)t>PdSm`I+F9nFDpNj+Qs_=K*mCcO6%l?9+3N
zBFnYv)Dg+|N|JB@Bp!&Uh@f+$)qybEIi9+x2v*+j<Wf{Ey8J(#VnJgk;Wmp9b7tQM
zXA)+TZZ`!(L_Xu1`p}ykX(!g2SCuQQyir<04@p$(p-IE+(q!P`?DIq3sKDmrr+1ka
z+Oz7K(rn9nHR%&W4a&qBPe!vki=U!*DG^VL3Toh%Wy($WHWo=N($sO>3VwB4b1iTR
zrq*<T*VQ7A<jNiUp%@x;zeBKDb$)SDU#fBG(Xec!^h&Olzs=zz;km@0<9>xSliet%
zfl(KOHFy_ipZBM}OpAs<yU03UFYgopJ~<G`I0+2>%8rlw<HMQ~U5-^ik*r$M5++k&
zK{&JdjX?mKOZYVlQ4gw#yiPIis8D~_7Bi+;TQ(xKTe`YXf?bd?>${DMJ(ubdwDBXW
zY7n%jEfQm5Z19d<37MEBc9Jm|OX8t5Y5#EWGRP0<v-ra{O|Tq=&*N#sV2b<Oyn9F?
z-?pv&(6Tz;fCUvaDd`MQ5wHM4jNL#fMiptV#U}ro`i=Zfz**$_)mLFq{3G7|-!8vT
znOJ@Lxty;P@U}~sp}lFzePrqAY_K6Wkp?0x+g~4wms{=d!}~7jFRB*CdgHyEcAyY0
ziWBg*Q?brJOXZK@3E&+lEnvwT8G8$QM{+$oP>GOp_G22=L6P+n$6j8GydJqXIP8Ng
z5aI>DO*FjyLKhNmIHDk4=Qs;#fBjIu6UUS{7KPAJ1M=MIyw>FMilyy4KeK2y7Jm!$
zZjPFh1W~XrC*=xU2;lVR5VY;xVU{H?7_+vJY!YPp!xNA8+J#%5@D(ojvVMa0nSx(3
zG*g-#TW~DFu2l6aV%USYgnz+8A5W^!gv}06$Rr7tNJZgyHk;y?5V{n$n&h8>uYMrh
zP{rSs?7-g@kw7zBh|>+p9tv$^1P@NFa&8b#;0uxyT=}3k9he!t$zbzabP<%Sm6ve6
zMD-?}QBBq0yzz#*G5@w~<cK*YwhhIfuWtzz#7IKre+~4?b10KJ_Ue+(V;~gF=%z~S
z-|t-IZH|^4YdesIWZr6L`WzLrpU$a$#XFIOzg_$oW#_z78AO%3!&*dVu|2YmNU2tK
z?rAUs$J##unC6wi=Hic)*#HTP2EKN$gbnUt@Hc^kPTH9&RdEc8Ig9^GPn73J5+lJP
zk{k{2k6*RfQ3y0C;aF~GPP{9nr+F%KcT#bbHz!Vxopv&HWMEyGi2RhOExKc;*{-+l
zZ!dq1K5r4yeFIo63wLbIK|MKB&0RU~&TWmNd0L@@pQuN`dHw#?=Mr%#cAB|nx4y`v
zP&?ybAsrI`TslQ3J;vK;M-QoS&ot%%!_*eI&?l?fvVBn2aBa*~(K5=np*aj>`3fVa
z?)1mvt*iR)L>X{GYLwW!fSJ7pdX=sI3KfAq>7GK_x&}4d19}9Rpu<_Z)CvoR<&a!M
z2=c!T6EhQ~D{(l9;C}MzqF+-Hd<20}Q(Va4e|_%2q5PIffLL*+9t0WqT_3TwXywh>
z>?Is3E=&Tifpp-~BKKsf2Sv7PqrBfGp8CW+V|&`eQ&uI<2@<EBAK6Fpvqr`H62+Hc
zIWN0T>ZeP%LSI?*vD$}MmulrWi85Qb{{4nMr1?itP1?$VV^Z+Lz+G{Ve&QL*k!R55
z{-<*;Spy10E+0s1YP0SQ4y5gY5ryz66Nxbo``Vs}68KD9h3EP!l;^uB1-q!&)jbC(
zUsS-df;V}ix1m?wo@zs#ioTR-1&iYHmMxmKA_)>*@H7b!*c&&88@O+FG5z~rqk&3l
zG2Q+SY_X_>r&LUVCz@GbV@-1(^FWn4ci`#Ijr02_jU(Ie43$k{rx)gk$?_XvwN?+c
ztJfu2YeoMW2<5ohu`bd;qL81#4DR0poY|@6%D*I)4UK4YJGgjF-H{N{Y&Zict)&+Y
z!_j1+mJo_uAB23bxu3AkN43s@b66Q8iO%E`KbWWJv11&)^)XB{!8B#=m{EMEUPX_w
zi#mEuG#gCtsXV4d>3SV8Ledo+YWjLe7TDLb<sw_a|5h`5HAX3WMX*=xv;)^B#pfZH
ztUorO$jXh=ERRj@i#2J-TfUQ3&7Z7gO@5T>@~)ql3~OGkY}TarW@fsKm_sO8V3gZr
zC%N)7=R*2WqF}*amsH|IKl^<_7kJ6KYiF_)w4SZqkCM`^26}RhK~+se4o9hobLjqi
ze5he@?G!&efx{mD<ElSfPc@mfY7wVFe%ig|sG0nkdCwU<eBz4ohNbebl?EkpTP>mp
z@c#M6SkoPLr;dI9;VNC(VH;M^qOfvGU#s=Jb3XReaz*i8fSz4}@e~R?A)b(bDY#Q4
zA|TuhP<{!NabbIBlbIyS-V8+UMUx%HTUWmb_vFn$?LZ{<TvSNT4ga-a@78FkV}ox$
z6ZSXoUEINNmgT@7B)1V))Mx0n|3!_Cz*Vbv&ud`GZQF{yU%~2}@}l4kvW`<hvDglO
z!G>(|)`ifWQc!f)b~vnExUmwVoJhb&Pj+xE5l8rM>fV*7au8E}og(k5P_EqH%Qc{R
zRxEo=iZ`F!RsW$ElI8~^JCizei=H!FekhU4l|{eQ{AenwI5_j!`d1m%q`a~Ny0E6|
z456akdU>|rw-*`-jFYXUkA}B3qr6h)Zv8@edf^?AkY3&juHDBmjz1@V_JB!kroWH|
zwJvrO<Yi6Vnou1TI6V^6(^+Ky_qx8Ix3acr-_sZKD2d*L7e<pi^<VckZ_XFO7tp4_
zf9Y9fdY1p2Rp+4pU-j&`4Rssb&nEU8M70RQ88|a$Esu-e1>+6v2<6Cx1!~&2ZcsdV
zn1Zsia>u#Jtczw*zF2sSphR*CFYVo3OG-;kiY@?7s}oYi-TU?JGyleaAV*>^w*jpt
z(N09W`>XTgp0dd0#l0Uv%pBH-+Xvxgeai<O#V(`|16nkx{qrW*ZeZtcFUQ)RZ5UqE
zHcyoI+&la>0usJF-x@R-CAMl!i?8?Rr!Z2#f>26#vF^Pto-W<DciL*p7If^_-8HZ7
z94Ng+YVIuqBw7Dq&9|11@9VHsGY#o2j{ZEt$I~)iiunQCoL*un|7m7Z;IAa1roO>8
zaw!ibW=MI7h<NqjZFweCmY;09W=S3Ynl2zvqX<@>TusiV-J4EH?nd%EZWT`?_ETh*
z0#RQ2k1)&KyeU^+rjFQ=7_`wqzEEz2letv4u|!nB&?SGMBP(Q}L=AF0ZM;pr4sHqv
zUl98r<(}f{cKoHC3Hl!W^54#u#GhYDE0Ul&=OF%QdR|3|;C13G+rLHf_gB;Qc+AVy
zBW*MFQU4%tnbc|VrtxiRy-?faFhk=RTP>M5<LraI$We?wJ7-x^RfbC+M?B)DbgT`E
zsgJ|8iA(NZu(TbnHzPv_U1s0NAm)1vV7)3b{)8J}Tjg*?6EjfP;zuD6o3coG4J8oo
z#UTIm(nyA@2+=plTn(i(MNNl1_KUN$?meN31*-G6JU)D-8xCMDY3>)FS@YIenV{?~
ziqK$}mtuC20Jx3OJh?N}=rUQD2K!N8f~XIUSc%vDnjI#izjUr_6;|yKderDWWy*-a
z3uK{R)qUMBd7O+A#zS$>1Y|M!2&zA`st%q&RXgH&Rx_rO;~O>H&7OqT7FqFi_-*@S
z>071M`y*zz*S=`og&~b*BXEyc+Kv&UHW!@bG`7>756!Q^DeX+iv6K~m^Svc63Qimh
zLUHR$)r1q9PEh+ayWW`<u&~=_DXPS>FcX(I*8~|T|8$&c?OF;mKs^LV;HwO0__wz~
zMmzsimt;GvK9J&LhPCD!4HVlfPOE7ospAwn&TV_@!%o`O>uRH)CC4nTYH(IAtaxzd
z^0F|Tn^y&}nJ2#J5d}soZ$h|XZFx<<k!$cl%ouU;`NOEZD#l1MU7AC0v51*^k8v`E
z-j29fC}OR0b|BAh9sU!`s~{L2xIQ<phNaD=s&LP7UhFz!JPHEJieMf(Hl|TQ7P(F>
z$7^LFGpI8guCks%zn+}f$h52|lZbrXnw+vt?0LEzfz5|%CV!Q%4+FX>Sj1Ne5FoTv
zWceG};nssX;k+gCRY?5{uqVj!Pfh@~HeFvs`2oj_v#MDEb%Yq1#8Ai~?Tp@mj_FTf
zmdY&O(1Q=Gx?5#^S9;lqZSJsM(4Bp6OeV1pNW^bImTR0AMQHrZjRg!AKZFXbe)cj<
zDmh+5H>M5V7D6GV(!+hZYq%m!;LY5mEw{Y9Sru3=|1@`H3NmJI$TAEM4ty4snOi|P
z(7=2+!9A){3gRO{4;{I8ez$sjvVvFor_`qi_^yzBg<r)N=hRegh~4q$*(pW`JwH(2
zm|>>HOkbeJ$>Tzz#_u~lpeD3`!PP6r1B{43aCs>zWQk!+g9#H{h1rM`3nW(QZZQ1&
zSkvgyb3cXWU@aFS?s-Yt)HP#wWpwcE15wP$JzO5e5<k6H+t<ujvK7U>YW}k}CFxtj
zOqf9~*{0=9*a~@`SFqf(7nQdbTFwm|!mS3GpHfa0M{EikR<ejHsY}~6UwG5@@<Mn&
z@)vH3(L<VU*aNpR#s!V*z)fW@dCR;?*|9JG-^dNqdIk!54Atd>RmrL>i5TNp++H%&
zq7&!o;riv!Fm~c$>NWFXWQ$;ZBpViae^PuDr9*D!(`=+~O`eGrbcHpS+Iwy$?U4&4
zjbLh~#n4)Dp1X2fe9vaQ(&W(sXRSZQWYcQpC_j-)E-MRhyzYyzreUeM`>-tQ6oLds
zvVz5zi@7d0p*hp^AZseQ6R(^2B=vZP`WBA033%!{-<GY|Ul<ptX^S+oG~(F?q0!NM
zo4L3xIqzAYbdng+^$eL%_rW~lWM{dmurCADjLn=&rFz0S%@gUGgbc|cDuTfG=;8+=
z=Zp3wp#GNrH>)Mg$dkiJ(sWVwCIiIA^+`>?7L;-8(6`FsGLOdAWeg@6w+=P?@#orV
zmKoVc-8A14u><O+yzv&y)B2<#1Nk$Fx_?#8@!Fnynu03Jq&}-X-SFzwhG8XLy6YE1
zZB!9Smy=zPiG)EEOW!hD&q>=5GVZ&3>W$-6u7C^Y#9}l~<P0WXK`D##*2Pw(^fy*s
zX*OL8JVwQs{$M{+n8A~w_0}nJ?d$>z>f`jHzik#o=!;TXL8xq*#W@acMn*&8`vIcU
z(QaMaGY3FO8pc|7eRV*VzbkFa_?o=94$gLek4|G`;=vl3oO~gHoF-j}K7?MlT9OG+
z1N*c{M`02F_LJ7O_`?|g<*TJW2Mn70jiH)MI!3l3oko83b0NvVR3O{HnMDFqYuY)N
zv21+6F<Hd!2vXYp{%wtCx>nNkNhAe9b>p}fa)u6;zSaxJqF}gdwAA>7t-;asSzIRv
zJ(jiNlR*wpu1$XD`!hHh(3IG6|Hhkl%ztI~w`wgU!{nMQyN+|S2f*B)#uQm7q6zV`
z{+)PEr@^&rz<X8NLed9m(}cIa1X1zmRl46CTMq9}3(vg4jT<o!Ytql$$8}K^hq#)&
zi0aWz3;f|BpvPiVQ;j8>J%b9SqMqb>mK5h8zaO=Xd*Q%ex*1Fv1GmZUhCG@Ap31C+
zhk{_Dz7;~|9o2^7B5+uMDDc8=z5X2YFB-*#0XoL`u+?l?X4%sWC5+LwT;D6DI)}V1
zJW$z*aAa_jVZEIeXIjx4F=PvZkh{S_y@ZInXMXzxCsWH;?%Mc7pzf_?jc4*aMMmb_
zMUPxwXVQbCeeksqL}U44<jFvtI#Ro`Rers=*>o%dd%x7+QEuiT(gS*+u%yW%zpc>u
zt%Bf#6>F`9qSKzYip96RQK(`4@F3t_{Leb4D4bu0v1ZT$0`94Vx098wgLRz;(C6y)
zMI<R{QGzQGd!g^S96eLkw>c5ku})b8bv1H}RR(2iIsGtA3klaBnCk;rR~%e>FQsf<
zM>G93xXIOWjp2TNYobPnI1IS9i+qgr$y{_Od5`0&y*$610H7NpgBB~<TGZwePm5v}
zToq>^ej%5G)>gx3zGU4Fnqj+Gxcjd5^c}~*0VPa}K~&De_p*BV=EtMltq%Ng0wGN<
zP`J&l9H&wC_>JDH30}HO!e8?~b+!;0<nJA;w&yrLMc=-YGG4YKroVg7UjDq-OK1*^
z_~@K+v#=DXs(Kic6la%m;>_<>rGrz}3c_WN=f_53Ek%r`J&rjPjykH&DHr8yNYUID
zwN#yVKp+vRI{9;-jcK+;HGrGd&@9(IjWINeuO8_3kfulVJ&!AmyU}DsA}e!w38F(?
zh}?o;UrFln5%qM-#;rYenLl7m97Lu2fU)-tBpLSTFSA8|B)^)MWW{k}eyDKGGEhUW
zX)3w$Wk{(Os2~_afsZgOC(rKkN&Y{Ky+e>F(Y7sGwr$(CYnN@?wryLxY}>YN+qV7o
zeU0~j#669Z*~!l8WyBhD%mF3}wViTq$YxDDIdR<rpx6F1MxhbVIC%H_l2<U|z!m1>
zP-lUL1zREES3qJS0Ru#;T7{in<a)><M9gaWZ7vwdE(BRGQ^VQ*QWL%qkpO6D33fY9
zoJtD7!wMbw={45huwX`*`j|Dx@E{!4#IGxi0M|8Jyn#Q|${4>!a<%~I9l`U|BT%js
z3^Z~sX$%j($T_?60KVt}A+<LrfBjk9Ofd2;x#3@4XC+pGl(+Tog!42>Dvq;VKECgt
zANc=`W$*tHt26%ZY)3{8X2$<&;XJK5ZM(_wZ{bY&1Y|ItzK<`DQI^fox6?9aQ(rpV
zQi|7d>$gTI5=lg=(NJB)JLYI^`R4F8T}#w?n|-AVU)nlY3aw}sviL6(@ugbnE9gj$
z)MM^!hKdxX;^tcHb=8cr|IXzgs_4(r<b&I&n~%59v3uwNW~5jO7oUbAPvE~7oMo3(
zuW#1|tF!17mj9MSA-=BxRZf|wMs~D!_lIK)L><a*VAbC0*VBXh%gub>$__N}=I7l3
zf**k@0FDcM;O*wLce`~kw|cJLHp_5Z+^DI>xsZf9n#-uG?@p-}n@g%3b=SjgXpQwU
z!o9Z3W(Q-pem9B$joJI_YloFsPj8Zq1?s6r2hzAW#h^uDKx%Ua&)K)f$46>FeOdCT
zF<WPsYhPo%TKDbCWeym4{<H{rGj4oQT(1N_E$uKVA%N52&nuF#C_S(=E}yZGUJS#B
z96Eh%PNr*>3iIA0WNwXXv-?NilA)1BQE;|De&J}gycKoy2;{|Bi%mFfZY5K7u$cS4
zJjJC}gjAQkDuV3u4{3Jib+cNw>-@nt1z<LT8G8W;P(NH{HZG6yl8d2P9})&|`f|k=
z+Pu;|h(ph{2L77d&0h=LEVHbfaQxc}3gMhrLOC+(5aPO5#T;=Cencew+p$~G58SUB
z(yRVk%>%67Ok8h_UiN&PH=SAE<XwhX&(kqGCgQN4#BIY{guxwb-&Pu7;;|0FC>dvn
z`9i}T+IgDcQqPcL^9;G>{zj~0k3xD(Vm*nDy^ti{v*1Q7%tPr&0NSPpVSmPSCY;fQ
zWfC#&H%Jj7bolEBT{~TKBqvfRVgr>2ZK{^kLQ5q^O?W!i(;{F!<}mF8hcTtx#rLhh
z8bqa{69SnURHJTfIhwZfdeZ#0rHov2AbAhx8r2?!r`AK*-(Bd^J)0!tNzyfhZQ*BR
zGw1O^I$4(7a@vdGRTA1v&aMfKa)9=UQiBGovMw)>bOt=#RvEU^M?#H>o^($cMH4=T
z+dTL3V;9>hQ3iy@9(0#knN_zF=ZbJJuFe@JF9e{v?w4!_KqP=1^CU%%gmnx|*9(UM
z(X`-Y+;{c;lFC~5oV0Hldzt%7s#EXlZ{7_U-$=mcB=&3|v#|mVK@P=8oM8fh{TxDI
zj#P$0V$vzgBs90FmJO{4UE`yQ)2d@V<)+l%FUpMxb;1?v;@T3do4DS}745JlvaFfw
zy*d>VE02huZqlrosH%$7XtQC(8OB-i7sAXkhFZ#+*Xfp~<3san@avQECY#B_XA#Z2
zNpnb!Bm&y~y(>Z_AwbJwO$eDOs+!@`SGbztF+(Y6hLvf%)gj_YAQxg<;W5c>c7!cu
zx8YvR!Mq2$v<|d4tlV#PZLN^U2(K6aP*TVm5<KU(nB6P&T>~WCUXA)bX=$m}p;kSl
zIGdoOn$<}|*!uB}wXFFnXMQ(q@l_LN07?4-R$2T`CL?NYypSExB4j8tg@-36^oj)l
zVp{tGupnJj|NM(k4g6I<%K+smah{YWZAZZIG6`e<@*3<AaHCImz`9D@BsTX(t{jP6
zRg5^*kdLDq_1QZc;@>-{#{REkq#X__?77Xth5cb8JEQM~8jSao8HBvKtnL7V3J+H2
z(Bu<<u|a~af|MC~r`{JNr0BUR$T}R#X}S~$&wxD52r0aD_f}Q8mEGPT1Wn5TXKV+~
znIC;wT2_+}Pr|kpMRJ%M%NE<v*+a9Fw)&OYDG)(wfrB-2dkkM8;q)ga&d(*La5#sf
zZE!8>>+UD#t-@Z*!Mru4HONR#*yBDCdSd=2ePhC5{oh6LX}o+fo;()H9sLf1;gp_`
zBvwHQ;53E;_M!5R^`ikXD>xH>Op=jpDjq?LzGO4Sz~mpQlI34FVS2)k){h85Paa_^
zK_8CMdU_Z+9(d~2d9HiVwxv9zoWPQqbuU;fy6u}})ghnxG9Da^WU~vZdJJhx#uZ~8
z9MF;CKXBVDgW68H@G3<iWSbxI4}t8><(+h+$9Aj@3p#SXDx659f}XNIUDgO)D(-oq
zOQHlHrZEZo=s|0#*{^)HNU!SBGdT>B48j-7U}-KgwH5336IXj+BJ@;*)Xk}7r%54R
zK=Q;M<-sqCdBv_T#3JEGijX$c4&@M#E9V9cKio|(>*6s=S`PL1o-9tD>7pmWA-sZy
z$Y*#`td1di`ppE)^FkD)2T<}=LGaQQHld+R?uORIu}tB3?5i8c#0NE^6(uSSz|%Ou
zSq;Vf?e9&kPdulu$`8qvq%cN~T~P2&^B2-f4Z*=)H)*xClc++jd;LaMdfg>K3;x(D
zY$}8<#V7VTR#t9K0ZEG?cb{cLk>>)E41M^`l}NFQH==p;nUO@3<YiN}6%lOLjoXKe
zpKq3#kC=F+F!#M56*hEf3NEnAORPALzu>yhF&qE!*8NXu4jlhzv;X_h|0`H`HgTjE
zwKi}z`D0>aXKcd92j%4KXkuUk<-XCaC1pq53A=Nu_6GkaMVvQ)K!At5UBHcZtmdrZ
z?D~Sv^Y^!4DRUv(Y=nG9hogSf^N#d^m~wK`kN)9dbMg7{+z!8tUe>)3N1M;Dm*PWc
zL67u2pB23{xvq>Z*H8QBorqN3Q|G+6_)k$<;cL!M=Bk-;jwl6I{#m(V?W`=dXW$cv
z1q{Xq{W<8<PKp`6Nl)>+t|hKoUNb7J@{=}g?Z2Yp&m<ZUDMV{;?BD6!oquHE>(UrF
z%1CGwt1Lx1-S5vKrEh);;z|g#KZK;-&)4knYpLH<<w|$SuY|g8LS0ew2+RNwb@V;D
zl|r7Aih6=zVbK!wH~bU7FBu^QQaBlzi!tBR>$%CqDaQ#fbFajX%&^J~Aq28eVF*Y~
zZ9-_1>Ao>9cF>)<S|k48IBf<X3T9@8+p>v^a0v)ihN9-;sN9`m&)}CNH(ff@^#ug5
ziz}Yq5#)#$L4c*F2aEeVbEMnz)2SW%F-SjT<*H)MN%s(NJii!L6LqqWagtne54(+W
z_R8}45QJEmo#u9x>ShX+r&LLY?(ncm#vhxdBkH2pp0Onr%1Wjlb=i~)TCNt8V%m6D
zd^^QI!zsv*8}51ej_WBh+@w?&3he8DEZ%HY#(kv|`V}F3i8<BvuH$n4lJY^SWXR`i
z5M4?-r`v7u@UL6t=oRgr(59dNQA%)iuF~Z4EjgS>t(C;`kX?AAc<Q=GoCbQxS#7YR
z;G(Nsx<$tSXWoLA>B!}H49f1{aig>>KV3MT-B5#M;MQ8BoPHwVyw*x)2kNV7a@&0;
zQ2b;6NW`Vw%<f>&JfjxH$T!|}HIYm*jE2v)qELgdieg(?mXF&%;Eov4qLZNiY_V8j
zIg;XWo^-N&Dppv2lCr$kQy;wmFMIqB9K6iS{;Fsa>!wgaU994{+A|_hd(=}jJ#^zC
zYi=!dOO|K97K6~4zxs`YK3Tb-uzmhT<`+ekyFdnpFPMD7#~xw8Z>U59wBbv3*wY9w
z1XGZ%R-RKVKfESMw94s9LVVLSsw4ar?3Q7ZGdrEi%V_{(^4kF|oiNRsA+0!C6Gr$4
zX^d0);ms7sxCjkzrK4$G&+KCQ=s7C-p;QL=S(4*))3QPnXPVV_KtaXO0))wfSzF$e
zR$}(E#UhxHheyf2)+Ugke5R7+hZ~;C7b?qhwGc5*Wq$rB0a3_4P48ruN!F{%9;wmh
z{hq-(f=eE){=jV{Rg6cHSASA#A%-#mNL`}UGk4}&pFKbS{^Dk1k$HaC?0S;)Q@HS2
z>ZIz=PYSHogcJa0sL4;ljQ`<#AT82cnZ!AzmpE~fT^Sp`Sg*tH!K6jBS1Bv>##GO6
zBTRAhb2b*6Ddx99GkNyniM1FDDz64KmeigG6gG*m*2b6VXif$UEJWe*nF>^lDTKqe
zDMk4<B^*YcF~?g!x1)L=piIJ_ChfjXr|_0S5^nj_+=!F(ns;wy-a)1fE!g{M^v}dX
zSmjfa12u@={jMrfk@kqYR(P$LmWDoLM#q=)PE!~+Ro=0F2(ftA3Qvldz{L8`2|Pt{
z%ow_#e7A9+Fmiq=3zD(!1&D}9acm)hQ!WP{teqGk==8cLPpf9m>D1Nv1cciWOi9t8
zI;N}Xr9S;w>l$Mx)<aMYQf`F|!hq<wiNN6lrj_>Yyb@!QGD0;y^`GN4OT8^fS_2q!
z?vR{)mnY}lNPUXHJ;H1FH3>vn#nPi3!_Hd*-zm;lw*>Libzc$m&rWi^MqvQ3e{uWF
zR3AW@;sN}5zmXd~V<k}@?ida_>rWiEX5i(8%oTo#aacKDSIw-bU3Yk=I)btxCq}hy
zlgEI=REpb?^r`57tJOR8v|?EW*JG_H?Y=cFFf1hI`m8=SN&BF@jK@prQ9(LGuEK2O
z5W$zb4moQ(@zoGR7F>1TQQnMQAhc6B9+Ti&HvhV8mD=r@NE@{Vi0T+g+brWgC-U{K
z&`<234vhoxqLU4$z?=UI%Jw<%h4xhPrR?L@-7S;oE%!Ara6GsFnj>cw;2A3puttMe
za@9nbykp7JXfRXkz0oQu(HGH{p|<wwRkv4R2cGW99;qM@47R`r?OY0Y0$Pr2lR@V}
zSpCx?htiK0=1#AmTHNKF;UYBMsRQ+px=;<G{3jFky!wg(=V+|e*ad)ADz0>5@VQWB
zGR;Gp)sHQ<yzB2~qYGI^@($kWixwQ`(n_)B!chJeJI*x<po*V$-e)BUElPgZ=L?x2
z@?9NuHjTC|r-Ws$dUu|M;p3Pmu{r9_q14A6RBeU}Mn{fY9AmLsz}-Vrg~#kRBK2#h
z07?!`F(R`}^G(8JpRoNM{U7qDirjNLtA@A-6igyxND+9nTES=KThkGpVzHJZ>sy9S
zHPXS!jX8x*ES=Qyb4T7~3}E5DaqJ$28!$nqTe~i}jiTa<zw0rT$P6he);lF7V|;S}
zMNP2`tAAXpZg8D}1i^ujY;)2ppWIWQRgDz~lSY0RO3*W393Jro#b?&4%0_&SiL%ik
zDm64NZhXs)2#f$5NnN57`ELd-K|v6nL!Tb2g6r(CDFX!JKcLeG_Zj6yLC)mg|4d_(
zhS+AA?MQ4k%IKyvAbEn*cfTf3#e&aa6v^+IK^3G*=x^fsfc;XFRL4~2+pG#wmzKWu
zI~M9w>d<cHnET=vj@=TbI4&f@SkZRZqh*;rYjdkYyfRG$DMJYh%CnnHWtO%yk)jI2
zSgk9!XS&e5qn%sSg6;>Uhqccl)OuPp&UaJi+DSF(jB!vZ3-7OKbQ3Pg5D+FT?C9ob
zJSS@Cxtl|9*cS0>X^cc485pV(#?CW$cZzoC81W_Io0CUtNSHd`HYzlvf}G}v)M<&X
z6K@d5v3Np_nx5=Op;y}CAG(u|&vs^m82t<R@UepnQ(N3vuCmE^cu<TglZlXY6Eq4-
z!Tv|y$=Ihk-g`3CELiX0MU_{VEV*J#(F5d6%;KB>%2F7aI+})r+9Be&Q<8`rIy@>*
z-sALjw)eNsPhl(QS$H&r?f~jW(fV*BKNL?q^EC{tiR8Rg!WOR^R(K50G!8euutwv-
zz#uv9P^G8XxlZVY)P`*>fY6dupcebZT<0&9!|`0#7>_DT-eAW#0wwFr`^hU1z8VLm
zn{B%Uz<|LVatiT`K{P>ZY)jOqN}8+Rk6Tj1_j5Br*u+K?t{}wFDb(UfIb-8aXbs~T
zihO#cf$PH0r@~!A@>x=`t3APTS9ExN5L=^dhJf`X)Sjr$LDJgkdIi9zmaOR=fRfDB
z2YQr(KL9O21^vXQ#mod)*c3Q*$!h}+Ozn<Ekt$=}m=teka;D2s0vYmo6JyU)gPFv>
z_2EmS$~R!U;K_4I>JaU`@$o^-1K?c%(^~n`CcT<z_(Tpoqs8-{A2?3C6yznh;GLXk
zgRI`z^7WdAk(J0BxMaoJBLk4?mO8vMcn--_xH(z|PE>++UJu--(I2Lqg>eE$Gb-1*
zzeORS%G<}Fr;luZ-Uf>z^uA~2pX1zp)?AO+CKXj!AoZeBzxrV|_DqQ|d6XiyliZHp
z{0d)iP;DQjyFBFS<>D5XELWryB^{pVVWuoylx(*!1!mb<Kel66?-d?xeJ;HjW*azN
z)BfG}_2*TAaxH>%;?A6bNzUDdaRb5eQbgwSjq$m9jC)RVjZtJIF>`maVAx~dF~^Gh
z;7(JKDB>TPq(A=sJYD<+d<tD4`A-;{neqRpnpF0%H=&m^v{Z7op_e6KWMKG*p&gx^
z|9x>X{J-Ko4i@JB3Scj5YT0Rx{=eeA+x5w*8EZrc`ztQSy}Zz402@$VaTHYQ5)r`y
z?bqGcOx~qXCW)}b*lVzeX~J$Zvl;E)^zC>>@$6e5zVbrjVEXJtL`FBp6~CzB!CM#9
zGNenpp2M*eC>_&s<*TA#$>N$i9wsO+587w3YoX}DMRz7gHl_e-&}tIbeU5njQDqOZ
zs%b?NDR&{EZNh!KeJcNP9id~<`1MgLX($3GldPp3<F$O}`iFkaU6I6zz_ut4bpgbq
zI{*5h<PKmZ^r%{r{nbRFV=$XHZT!O}K~5E%7u=oR3WgwQP4uHJgYUXj?kfD1gVn4{
zY~c_IpJhDirmihQ*P})(c7}qz%Skf~>bW{$z;YLG#J>zHqb$yUCUNWhiz=|_r99lB
z`++c@sP*#|F{RGeK8?3fUY?cP_YGb&+sped>T77#^(9+CA&>4dgm9h8tU|bFlvyFe
z3=<|WKJ76v2B{m;)Z35g!0m(*OX)_<t?pZN*rG67*Gvuq90GeSQmXW;gFkHNd{k>b
z(X3Tdp^K90wDYq6RAdCp3@aONd6Nfqfow~WRg>Uoq(V2mJ6DjfQR7fYOiN<o33t7*
z@T3aB#RZg(TqJ?Stt$H<u>q%tx-H7KhawS{v5bB^gi}qQG!Y7j<9cW3+3S3CTvHc2
z1}#;nBr&*@iZnO`4MeRPFoPH*ccAGiP$-Whc!pHoBDkYHt1ykJ)fFVV4|wHE3ZvkI
z52}lQ6hE1|$<=ND`y*!Ir0L>V9y)$cdw*j55K1vm#OFadrrw{eUE5k;+l%GXCaIx2
zf6l$xdbfHxGNies<ZC;Bmgfx_yLoO8?w)V)(fM!aW6>7rVEB1*bLsmIJSrf~_U^mX
zvF<17P17OXFVTn81)v(}y}Z6FC*O_zIylbsc)fOd(|xsRKz$!~qbNCkxk3a^qt4$C
zDop9iy}ELAr)%lS7b{If8+Uzka<Xv08GhaMd7SEkG2rV5Nppr9{)UR>X64SNd8G&j
zi-8M!HoIXiT*9Jap<<_CiHPE|X<Uwc|2*Fh^mOxZ`>j6nPtnVjq^S`3t^F?bp5@{N
z?@wNy8#@S8$?hH8@u4S-A?(u$@YnSyO^1Ca!EN5Y^3Uu}eVT2%6^(!T+c%xXP+*As
zcC@w@)O|t(Ozg#XBWlnM^cu;V0g4LZG+bxEh)+=={ucUsNa-cw(+L|p()2<nFdZ`6
zpaR`wCb01&N+2zR49|m6vN4C6x$d7FN4cpwE25<#tTLK(@q!?DD{^`}Uy6K?+kB26
zSo^|7_<DQ~{CK~**0;5NXzl)bwzzxwT#-G#eji<T{_0OxlH2)i<j(R*({I}`L%g2_
zbOeTVBxVkk5CwMIlKd#EV2*<=%ND#LK|C*|g>6EYPiQc(5waX8wL%3bV#f>SYveLb
zsZg*HT3QMTbxFujga^)jR&b6#(8J!x(i`g?3^uA1JX@z6d^arL4Kitw{@!gN1938{
zgCm>%h3gO?*;@uBd+xFwK;otHq`6WE!4SyXVl$AWUM4N<IFh?+GgxncAcQ~_d<B*E
zot$XB>NQ-ue(7x-w+I?P&mk-AvV2gD_II}}Q*APb2@5~+XM$M;<OQfs(!2p}gPvI|
z!O;#JSg6Vem<TLjq!L67>i}z`k(CGZZ`n&NgRz%lfx1*gC=84QM~VXQ+Nrz1V0D-S
zC3V5W;74kQ@P>&8&#pRUFR3#2E1+i}(JhE32U&tXQ6xsO6mUw(I&~pb76yr4iD!)j
zC{Empm6y~b9Jh^YqfavCtP#Wc{1m}ilpIos;-G?$t1%bI)U?m%cE)M}UrlqM096)P
zOjp)v(dEdqEDT3Saf)=CeZ%=7IP?=P?P$b5uIrDRpz{T(`-s3lUY?rvJ`~HsCXHFR
zXJYRQAszzOgj_>6kJi_dyF|b{hoP;@gBER#O-*)>9)dkH(_zlj|AP&)cD^>k3Cb(8
zP)qzL{J;=QDT<9{Az?pt<x=?X(Q3#TB<-K6GVOK<EXbz^VsoNOrORWxU<y#i+!m++
zX%vL3=&hI<NvIB<4qWqfJGm_@>J1bCUX{D9$pwLU1j`Sg1TSb)8|XuB@X9mKI>{cr
zI!krPl`VplCQkn*Ap(bBp}h49c1@k35bz;@raO|eL!i61Dzkn697x{OR;;}Bu$b%{
zoqx4NC@h5HC}p+@-NL@NC<4N(jA)UnM+ozRLJ)m?wZsBdAnxPqC2@4<1k{-T1cC*8
zKx5Sr0Ku35H-C9ai`*fY0@1`H8xwXEgRdZSwL~BeL>y7N_}YN=hcRr!%Ex4J=c-L6
z0iLera(_L*96AUAo^t@zza#-9vFjjND^%oddMV=&A|dxG*9f8fo3n|--2}Oofc$@+
z&oj}8j3w9Tf#C>Y1~B*_XjJ3Dn7E?MO92?$4TU{K@rz7^dy$z$K4;N0F(xTMq3NU7
zCL^+^SpzWT8=|Ja?<HQ748qp2H<OV+S_pHO#RQ=Qw*N$PqIu3;UAJDoqQ}XV;D3}h
zMmdbw_QNEh#!w2R;HuV9PQy7_n^L<eC{6=W?bcy-T6HCzHoLb9r+p4_hb+6(+cfEF
z9;Klj7mN5>thKXTX>X7eZ6(`0^eE&j5GP+>WVOqlLC<-ifQ>}q0z2B74?{u{iI5=9
zv-EAU2vv-QLpvkG6!`<KF~+-6rZQQFcJtK-9b*<#nKjSDWN;wz3pLas^46ug)VZ32
zZvc)Lu8#r0U+R#^GpqxuEub`r!D}o)D5_Vy6hj8ojIL{Ju@Nf%F;oyFzwB0oJF6P4
z6DtoxV;9QXba>#D=>v#%6&oHciYK|G7-a|bQLB5jA9W3lC4n2xdlkU3xq}RqODt<O
zngmi#d%ze4^k*E~6ELU1a80}(>H*8kXftp$r&zF}87m%(ts2WUQmzlafTd9>^95od
zR)9Zb8oxZ|A+YtYB<)AA9>ytlgHfq#jDoxa#3rxLLvPB&gsD6Olb49(ftm$4psAC@
zJLaf{1GyLma!kd(k8E}Ytq2Nyz7r5^goc5c1?Um;jz+|*<HOSAuECpmm@`J<mV&Q?
zzt@2-(UDsN)Y;W#%d;Yp5%W*trV-H))t%Q#B?`4#`_*9J4tXAQ>YT8z#qwn~Zp5-p
zqh7hZgnNT;PQLb3s9Tve1bME~RT1(z<beD6d^f*;!AFlmmmfT)yaHla!=F7F{RWYF
zb3jNoKcLJ#IaEqf%K137m?2McA4*zGHsUz1r6d2Qk7`_BKd|;(?#<1WHd{NQr)Mo0
zrDq+#U2E`^9P?wS)X<sVg4?du>0URXG~C!SNGo>D27uZzu($(+rMPZTmf~gnRSk1-
zbNz%Bb8Om7*y`9^oS>v>?j%4dSJ*1sqOeG2s{mYC0<hvfC`7?tb}t$sX2mPN7mpF+
z#w&8=5npnL3BwA5&v`P9JzH#`b_MUt!`R>m63mo6lQ-HIJ4<V{K$)(NX5nkdw4(uC
zE?T-zht<LSZn6f_>&!;J$!U(S*X}-Rozk3gWf}X;)wullx*B|XxG;D1WXZ(Em)Ti{
zA&b)sWtxadHb}-4c69ZDjR$V4bw=#H4D~oLBe#J6TSOxE475l9+6QC5>q<C<sU)C<
zaK%M{#g)PsyA~-d18@fN)+CqdSYQ_oxF^;#qmJxtCB8N0pM$|mWs|f&Ib<e~2%+5(
zPY5Ef4}xIYIZX+e65BJg%w`SJHHjJ%wVn5JW$lcKamF(_BlHDtW2FBdllWLz|93Dk
zBPY{;?O2Z6Wcwe-@)>;a{?r{L2H46>hkzaIi|Ugy3-9sV*ROCGibO;c$z)5G3_f`E
zinb%iZMt37*v+i0EB|$@U1yw`Vm7L5(D2Q*6!;hTIKSXS!F)C-LTH}X=f-zO8WJqO
zm~V3OpD*>F?RzbJJ_L>(z%F2S)V;fYsy);L)C1nKciMNspY4)nCoa3Ut|UoNC~0hW
z;<r1#ua}ocS+daLJF+h<-RsiT)@{Ac0lvh!5jV~Jxux$P7xWcWbAzf||0;P|m))PU
zo-1FA^3=}gW7a&l;N<Prx_r&|ER-{UREn7S4n+d2{N~H(`@n(RL?8`@m7t|6kYP!A
znl3IiAv7qg5Mw%Qs0R$UG@VG&<__i!NJGE}-c#ees7`mnZF2cW4$T+-0{`WS(*X1Z
zQ3549+XQPiVh4UwWd1J=;9iwqx8suIDI|&(sC7YEU)eZ^O?jL)7hbcB_GRy|lA&f;
z?!ZM$S4|VookAKxFrAr9Uq2qZ)M6syWEZ1yNrbFKW2*3Yzdz#@DGA0nIznES<&JeS
z5BkaW6D^&>eHP1Bgp=q%oNIB-*(`SeX=4s2Y7gOKh{g0z!SsT=<5U%^&P#1K9x9_v
za9a0QSpqCp?Jr}PM}g-~h6_)#N*JWBmX}wOqLd_vA?R5-p#Wo+%f9wuUcj^5h)k-7
zk2r37MeDPmICNk*TNw6bAz-~Oeap(VpxWJ;d@)VY-7K3Nv}zU)ZHe44&D!bEHv|r#
zlse#@UC{$K$=-uD<L$)zv{lDE;HN;REN5(vJj3rn$3^#5S`!inMDA^}<u5&sK#|y)
zv{uG4cba9k)X^mUVJ*oe`E44*YQA|1$FVml+jbg?)6WAG8yu9Z%C#iLrtC5ko2@T`
z8*f>m$^L3AeX7uw-gExd1*-3dO0=gO7LGETYmThTR;iV4i^gBiJfIfi@GCHhqTOdj
znTbRA>?(!n(+m5BXpM^MOIli0C7#oU&o`V=-A+}kk)AxSyXKsj`p$-o!dI3d=|Ao-
zj!1fDR#x;?l%Ucv8EZXAD(o}=uwrY*2JwUqMao(D2LmLok)?J+j<IK2sKdtOg0&#E
zZR@eWU89>^9P=8(k$XC)GZ_ps5Bdbf<dGg9x{VH8*n=~uxIfKcK7<#8hrcYR2iRlj
zec}TN2^yqOR%*-R+Nkx)EQ&GZiYza<0Ke+Qhq6yr^k;t-z~Bl1QwsJ#4T&@}D@R1&
z)b!*WwLv?36{sg?SUcJpK?bis++nqY_A!m)rMgsok|6i85DH!mY;(qIBw#8N#Tc4c
zY*0?sK}B8kq=PgpR}99%>D5o`c@}@~3kMep&A`*kH@7r#Oc%BnOCq3BupG1fF2rQ0
zovePMAB}_S=$(gT4hJ#BLl4<|lD9}KW#hE0MI(P*b#!%Kx3q4mRS(Wm4%Z_~oeEXw
z*^oBVbp%icN)*ddZDMO6qaM_7GI%mD*U1^r^ftcR$3uAq59LuwBSQYEv=#O~=S=EB
za(FjZCbI`0GDz+{00C{534HvRLk{mAg&9)Qf)MBCRqRheei^pXlXr#Gg}-K$5c;I0
zNl`wD(Uf<&`z!lPiKa3YmgvToAA1&X)jqp<tsg{Y1NULCMCDQRU975Q^Un18p8oZZ
z=A@RhymfaKI3y4R-K}O1IeIyc&_|m#Yc$$IbK23{Id=(JbJhk%gI0P@Tc&=!mTkd#
zZFqH}Rq{oyG6J-XE2LkhlsZB&yA7{l3*$H*6W`SKQvsrY6Fc~H4AN1ab$nDL@E~h0
zc3B2M`mt8~yD0E<{-zB69;_KuxXtIiJd%A?Hl4>od;z)-UsF|wiG@)bIi1GHXXkOO
z&5iWPpx2hJ1btQ~1OL{D@{41a;j?nxAOgP6ws_WO<fQ7_>XTGvk*R_Tmr6#%DsZ`v
zo{+ql$~*(MgpUW`VWW=6@Ywg954@M?n@2=%2{wF?VN)^bkjDq@&XOso5~l3Ti%?Yg
z*_0<9xfZ>Qmt;=b?E!}hVOwTrsuPq3t^fV>v5tn{<phGmVfis=HnnaEiGywXUVGC`
zIU&A=W4%^Y7xgBsCetsoe}Z1w1pebHr0tT*#>eL+4r11;L_fyp>AFyb!|}^dqwXqR
z%t>JBl?8bCFH?SB?R7rmBMr+u@Z4W5t^B56#QgiKrGt3vIF`1*u<(0C{xu2r8s?aT
z5OH_h`z9=+9Ou2zx;<Qco>0Cg*zJN()kV;6oNJp*SC+Ec!?WyTuL!=P2Z2o9Hc$A#
zb1_es2+mGtU=X-HO<>2Wf!&_3{E$NIS^StnLE!!AHpQx!+p{5a=x)5gLhNZ>%)z#8
z9^??PTftaiZiZ(cPnrnZSslP4>?o=OXD3VW0fkx;Ilx7?sb3c29nS%$LKbrbG#$VB
zPr~vY^9Pa?+gX~B#haZgg6ET+H$e;4y;VUA)8+VBlE4l>^-p#I3;uoET|I!3rGCr|
zf{4Ul`f__zZdyhy+MN%-srG?q-96JKsMDmekM>jcz!K`<+v`+`N&gaKHIH`kEJQE;
zfThp5cm5N*#=`a=bYYyF{~dugtoc8lWSH*Hf1czjL`p?y2J^m4fOgV5+A+$pBbKnK
zUwonJVnx~;+twuE&Q4_{35P{u6W(v$^YhK^!_Dn9HQ^+soc{f%7at&x{9}4^%qAXA
z025egJN}a0UdtF&zPJa%>X?MMq+Q*fcW1l4Ntk#{f<k~&FEIjQQPJor!Bjh1{N&x5
z=<a{~{)(3j3MGV1d<4S5?IONi@76g{oofF`?!A?_r$^`KX-!0&XgzF_NvXi#(`j7i
zTcCf6i-C|bp!vu9>eKhlYJ%l4t2eoc2ZD)(RXX~FQ03om<7$PG_KJTh&qSHUMsMRJ
zl1Tk^4&HIMF(ZY1nn#QYZmL_tUNenZ`qCmX4;z=2(UPIgmZ}MX(kgE?EybqvB)unf
z5EQGr^5{yD_VDaK7MtdFhuBxAWcM!lPZ_zCSxDr*6+DZk$h#eEyb9|bnX4tIVIzRZ
z%|GYz!!e#?3*?sHRTbWwrerzKl%5vZn<N#pe76_`UapBF(;dkeV=G7_BT?-zT6V{q
zU@?oc=6%!by*`ySEp}44Bz%>W?Vbwvf7=AtJ@L8pbc1%t!Y3{n>dA{`oMPLGiv*}p
zawF{=6@)iD59+8bB?zsng%$&a&<lot7{v=phs?Y$fbwmNAre}ZoxAHKpkj3zyE?<F
zF|s6D7-+P%V;P85=ke0CeB7It*JHQ=l^%XM2Qu8W|1OR-5?PIAkyYXD@)Xw_4A0Qx
z*bt+uXEQCSOzcRy3%3v>R(wl1G-4#Umpc-JqDJE$hhoIcM){jl(?~9b{3|_L|HM_}
zkDeu?VV2!MK2@p-F&sB0mCvx1fT3(LBrzd;>L|Wg|LbJFV2b0{u8zD%vI!imz2{LJ
z@z=$;J(eY&d3kuuj@GB@dq@kEgK1rLCO)nO7oK>Jpi@y+rSG|*07Oz<;v&t0JRr_K
zrwAl+E(RrPu_hW;6mFM`GP_BE5<cq7F~R3w9EUA|x78yTa$HHkIQSHFKlSajfIY-p
z$_)!uDK>;mz6qNAvE?x>rC;hxs<v%@k07zpfb>B9j3HS%I<Xj2+K?PzGO;z>=E6E6
z0Y7p$l2$*~60ia$7l&@RPKdpRnn`fj(dOVV)1eQHlKbpjV-4&6(bnqWGpn2NnwmgO
zAJ&|N>`*N;j&<Fr8eN2)A3<!GOj<fq0q6Fow(|RX-<i3|nP>3C7{nhyJexVF#!}jF
z7sxAK5Xopk$71aE(x4X$%+F;tar<*`8g{<rnr3T<T&Pxw0-$YBdoPgwqT?)>FaCUe
zn2_jMd?&hw-4>$IJn@E8N2S~lw;1=TdxN3QsR__F&OKIH{k^PLMZ!z&vW+HX!CO9(
z*m}R>liGl;pd&MZ9fRR4V7j#a;S3Npram@5^yDO(x_&>J{CYM7-GAD$v%WEnnI6=?
zFlmK=NaXCq`C37M5+0&ZR{@hi-XpPT-0%=f(~Wx+S#b+d+cM5{?_zoe4?s&|kWVe+
z3e<ZkUDX#};f~sLK@O4vE8?hAEij;U@Qs^0{BRjC68aiHD09Xn>tL?D=19mmBm^Kh
z0!nzCKr-7{fb{bHQFdK3{z@THmt3BWEG<CQYy&BLs^+`>8BAepl%K+RMnfAkVT*a}
zKWp=5Jk<RNQWg4m^+6sr7SM995@E%8$uRr8qV%b8=<HLvEm*t&Bq*0;fhp`f&O0Za
z+v_^vFFCsj!mPH=8kJjy+`1JoO~3uFYwjXlirSp{Bl?4VNDp=53pLF(opa)O)r0TV
z5sUI(Rt-v4nze3otc#y{l(6SWRjhpz+7v7HQKCL-*sB1y6e(Ebt^42W%VL|HQ{zV~
zu<m6pQAhJZEa9?>zJKH7axbf!8h;l0@k1R7M+EDRwLx9dkitu(awAN4o6eI=Gulm-
zA4Z-9+G1XGjlgO*?ue!u%OzGc=tf6A6PGGu9+V5K*sy)fenE$C5XOD;oM<h9l2s{Z
zw;b;J__Aj|`ZkXd&<G4IOeQl~3TLNi>tACFK|Y?}e2YHv^5h2Wd4suMoNrV5iV`OO
zDv5^?mLi<FIZEm}*@BHfnRoK)zHfS)qwU!CWJF_D-WBqxZcw;kIX2oY(wOZ^S@2L{
zls6qy)c$KMJjv=DjGgs_XMcP>VHdTeYu)V+ki9sVm7s3w38_&z<@SxmQrzaS*EJ^M
z&vlEEQt<=Q@-%8DoRbYOEN!5BGMk!Fxi>hW%CqA0U}rLMBTMNrJ?(N;$rPI4A{H5*
zD+Z8&MCp_hvx#KhK%!`qb9)7!b-Ju10}XbCcQavzw~BxC_P4jAsCvCUc5^dIp&p+G
z!eynJtNF!GT-TYHSe;Hm5CDAFLHE14fw}*inD((4wh1!HhmIXuc>{{P`OZ7@KD#7T
zW%U*#xnVSxxQ%<`cs09M`m9><hcYxL^u}_=D5|AE588C=!F(dmoG(YTKZ<`eH42b<
zb&u$7@<rO%<YeZ&{<K!f#cHTJvCnm|yi<~rML`WqZFE_IbJ4~3(r>2d_H>vZzy=Fn
zJfWV~ecK7yhQy!~p8eCU|EHW4{Hs{)t5XPE8Rn;u>2|REWFu3i-X^_B+XSw)jio%j
z*hPPVRv?uUhRgd@)!Z4r5>+bXsA=ON=8@~zap)U5lmScWKjH#*w*TerWMpIc@5cRY
zO)0yr|2ytev|^cK{L8xlm&dr<*a7Nhy{kSeH}{;#fBy}G(?g<6(5&tT48&VWq;TYT
zoL0h_JkOPo?cK@Nz2(bhl~v^csw`QhN4|p|_Y-``%O3}VC^Yc==K8*mN}jm89pX<`
zwEEHL`By*`7Kep13y=fI19f*dPjnZ4f;jDja<B3rci%m2blSCZ=ShwT3|a!qgUIuH
zwm-YNdt+?+KZ<diwyIoR-P@n;%yl)*`1_<(0{O+K)z!{Bpn%|d9_wrZcArnL569Qf
z3`^KpU@$Q(;U#~gB7$0SqprDq6(BRT8_Mup5E3+$cg)#yF8(k#ve`oy<&%axaf2V>
zJ~*9$IA>3h<kVq@^igZ9xFb0UF$0;l1*suB(=AN{05V}&^VAlG8I+@Teq>r;bF=R-
zUV>{*&HNaF%N^KANp6D{oFNEtM7-+(>9#m2xpZv|uPr9^rV??4w*dlVoEZ>f>I~9*
z533|bH4S`v$bMcSMQCD`YZ2vfG1FYA21c!nF=_+(TgIAb0yb;Mt7f>%^t**z30-yO
z$?(~6lql0cAIcYFV&XTZdmR8X;$wCyagxdHf7?YuV>DoTG-cO3X}KKLH6GHYXVcx_
zw1;<HQnT?Q!#(vpnkjzO5i}YRQaS1x<Ou!~SY_3FCD(5K;D#Oh&P;EtpeoM41>6u1
z=Ky7Y-<#9bdV|A+7)265)i3-hazDNvmO1r0@2X0aTxGIzF8vW?IIl9=`>!hJv%e1p
zfXTdWa4_KC^@qy-?9<`^qfI~e8Jffl2NM6PYqV?wx-<4Fd0bpq;xX&2dj7#!9M*3<
zMUfuph$j!)H}RId(L4;s3juMF3L)c{f$9d3m8)mma#<HAsb#C??aXMmmafD%p0^-`
zw1ZM9+|uYt@ZB6+fwy98YVNr|k7hvZ3^EB>9fsQUXI+DMS)6_KfP6_)*9E#xgnw-g
zft4yWbN~Z63-T~Igbk|GwhS`YJ`sZ8KS~jl30fNn23=(l2{hbNfLhhd&yqia7cHdQ
zO(R|su?D-X9WcZY>x1%9S`FVSx#C3-$A#&4q6jD5xH2i0j8e)?^K6LPuQ?S-EG#38
znS)@T{p-}8el*??TwB{Vqy}0m)inwasa~=4ru$4Z8IW;K<|r^MV9Ej%iyGuR{|;bq
zEn!}2tCxhL^mo5aIVpL~D?LP}44-_VP=m$6gCYSBlsIuqbIWVWh7ua?s3Bnh7fP8u
zhOL@%27k?KWsPiOgM9<(N*$`EU}zpioBrM2&p2&E3GQZrhO<qoIH&Mmk8}++d6Pn5
zk%ye6%tk>#3_ksf8*&jDLhy^9w)yQy7-vE+E%Ecgpzg{t1t{mMa7egtK%VJW+2g;H
z&Up3%Yxq4VzcsqtG>^^3-UggP6l@N3W-;Ttonk|R#ERT_E5=#-mG_D=y8}ZvnX`=z
zE#NorLz?Skf$|A8Yu3YyRlf9}C@J=dlthq?-jRSfqtViAr+~BV5_y3!6Pdout)oLy
z+UMCQ@X^H9*y69yRmEGE<C}Jo=Bdm?8)KjDaMRFr_XOX76i*BnLBju*GCDUp9OF29
z9WH=e#zNh@o$7k8o|x<KeNN^XlU5fkBopYCoZ)so5sPM$@0LDEvYt@uFAR=%;9`sn
zsmw093Zgk|zY#TS(cU$QYz<t`83=WM@3v+ihKB#XT%t-%b>)B_g4*k=KLUJzSHs;K
z6ML%8fsK86Q$oTwP^%kT>GiZtWdhnl;sUxiKf?m(-rEE2xxSG4yTEC8*Qh(t7lu@N
zv9C%MM0Bs#P+sS!@7tkwK~F8sXf`qcEHZ?fdwtec0;pIfq+Rc8){7okw=^}?OzEph
zoO4=fY2QN38s&8#rh9f1fwViuGoVK&XY3X?Ka1iS$-33;*VjUZ=(^Vc>`d%RwU$Fd
z3{ZE3=%BV*QNm@jux6+=)^dxj`;vrEP|h;W2yW`Y4r4#k03UYgp%i1j^Pn3H`k?~g
zDmpJZ#1U%DdOXXE+!CN^Q(Mx0c4nK)2iUA;S*kTd9e-(a)$dMvH=|ZPv9b$4f(A^~
zn3d`$yyarIQ2C)ec}B<?DHoyByv7AePEM05Z>;l<v!5xCJ<E?1rWIy9Wk#dM<QOyk
zQphnkoI-P=bD=rCrxDSlL~MVm<q8^1hCKSY{hH_h$Lovpf3ww09L)cn@OoS8|BEK5
zO89dDiUGSe^9<TBxpxy-6)3IRIQI2Rq<pS`RP5=zxr#S6VB0E5MA`EvC-L~W7v1Ci
z=kDQfKaIUF3VDa3l%C(WlfskWyfu-<e0rFYl%{MlZkPVoJ5f>V!y_L*8Oh%l=llA6
zQ5#>-BO~8Dv_PS_CJqiY1|SAtdOnno^xOPn^SHjXYiGfe9+4QdpxV3u>Yu|+kN@X~
zps`Pd3wB`5z8(&AT)uelvJg%IN|*=so?&Pa!AHAiv^3wRViCr>SWYB&7P?h$rz_)P
zNQ<x=vn{mCQaDKvs+L7h<cdm<)a~68(inJ>(trE7G6hAJtH*LaSk_yMmayi5#;E0t
ziBDp8;Gz0L{lqNZ7mGHvJ)vgyDQ9ta(oMx;4%s`R&}EwPy}~-ARLXC389I#0uEFWk
z=J-_0_oU)#@uB=#DL>Jz)Qj$tPG^6w5pD-<W1cRkR#RrDFbbDQ28Pj+2o@$il&U9#
z^Sr1UMJuWWopU0*zA}{ZAXOdl55$)U6vK+vN0Z>iBtDAZBYCgy?#mkycY!=c#g&K|
z-{#1Q4tzq!V>qd#?&d}3uv-JKp{zO!?xVzEh&o5`6`)72Wbv26Q?Q+y%2qy%2{4^f
z={Z4tOwsewInp>I-Ae%+3+zPS%jcJARB^vfN`%EsmucF4%z0qUM-=|<daE5j-Vgg*
za(=Y2c<1lWHzA*ZNwCX)K{Estz2DdA!{#7A$Fg<azLGd5FUe0qFPhuAsPFWtNPLr>
zMKxo0X*|h@p7h*+(es0dmnqUXC4MWOlc@j1G?9Zt*5*KdMRQ>euXl1Li}-jtV#lmr
zC`oZnMXz{(1NQN1L#J)gdbX>iIp>+8f~|sIrfB*LL7>tg+l0lZXD3WQlo;KNiM-O!
zkT?jw=~o(08ZoEosQUr^{<}@{JmjR^J1&6;P~V<gkLkqDsA+zkhhrKAbLzlEtBPpk
zaW$3E^fb1L==ACKmUusf`+Q{!L++4As;yidsfdH_ltE_m=)#sf=F#^mv$aMqiEg5v
z%6XHebpp#_C<m>Dr>7;>-9WwtFtZBcx>M7=b4lXsLL4~oxL4?WH#1z|ow3eLIk!$r
z92)HAhM0~^6NS)j!4c0ciAw&7h|mO`PIf#e+=fZI{o#=#3kFyWJL{;J1r88Yc>A$j
z?R`s^kPfSFm@DnF;85!AJNb%r(`oT-Td!%M22l5JLB@4fdV9=c`e#lbR>kds1a`?T
z6Q?32MLE{x@>*(O17)aVLy@Y9hpU}PI9l+^4>8q1@y2FZhGTOW+Z>^fFvzZUSh8fC
zgP{wANJwEZ6kUZX!nU{=BZLUy5XH`^Ewi2!oqKENw0CtSiO03bX_{-mtyqt-;a_eS
zLyFy+IMoV?-gY<JEq&T(^Ms*!&116;;3a3gwf>O@7c?MpD>e`WmO_kIPAd);fZzqt
z8TlyBN9HT2VkUZgs?Iz4c&W4Jsu(VTu(B3q?Hytc0;#T)I5vF}-~p^*brJq~!hG9G
zKN{kBn;p%|6ilw!@B<TS8&y<+&_M&g+89H&NDxP>)1?5z8}Z3K%6x^=z9j+dMSy>z
z@p2gUih!iH-lasJ@RdhC;i)k)?Z0sr{P3&_Mk3UJJ1}D?8{|2{FTj27|9FUQ%n;ZJ
zZyu@yr8E0ll4YlxK&X>fq}=Fk5!?{z;U~=cYPHybFfQqe=+JE+DGGOF?v@xb9Z_lS
z21F&c>%|FR_>!!EplzZ9TeMZ}iu#Jmn<mHSWY4P6rJo+etT}zr${Te4aH5sUZK|D3
zxRtApk(C0>#_gm<aq6!<&d>FTCDLd-n~O-Xb=i}m#w|!#1(G0vvS>ZkJ@1t#JP<{M
zeuu<)Zn&4iEF?kr2IsnZ>^}bzncE^-fp^|JfBgLbC4^U4O(5j~tLbJGxMmPhg<<3M
z0ayT_KqQQnG`ectEa}(`U7W5X*YI9}x^2EVT?Rm+Qx8KaDtRqRUcL4r(Wwua1w*Kw
zCAbj<6}_b@rwR9FH7oMc6=a^;za@t(wc`3-aO--yn;vGmpGWHk)w8iA428q7JoDx2
zf{81sLl&!uJrxqFLF}AV4%mI3CA&{TXD--XT8Az%XH+2%NoUz4J1n>D77MyJ;es~V
z_Gj-svPD$aXX*)kYCnGjK&-80cOH?bO>w&Mko|N7iblI%I_9Sz^#TPXYq}?Vnce`A
zLY7XF`rze{f)xFL-ugLw0N(m3e2L!bZ#6e5!XBDXCqaLal7Y+S!RHD!UKHRWAuy-v
zf?o4*HV8G=eITd$-E)UH-zqL(Ujwn0IIvGSw$IS@IV%8PAeI5V;FbxnJ<T{)nd?HK
zob_QJd_igqXgO!~(`td(U~rdg<4=+jU~S6Ha<|&s(bxB$Y{l3xLikY|`rTe##=w$}
z$ceFoROAS;w=3iX*xF9BgxTJ9ZUotyA~1KS^I&oBT+NBm_M=ZEtbuXh7FZCIi`N0!
zo&ERG^4i~3N!Z$rk2PlNLk?&Kz_KClr@1#VaV|le^VrPiB%;x&d@;2X+%A-2z5(q)
zhnRW$a2*SufXahUTCDdXi`3xMjuaNJAu3VHuAO!7HG9%ya57Jy?T1;g`k=2vVR{5G
z_iO{0`Wz^XFw=7Y82lag{_a*gbtnOGaR74F=|^3_r2d^-4<YTtbIdTFRsp7^0b~a%
zGf%d!pSZAbA)3r9_GQg^`ko*y=&~fPST%=k`lnnsZ`#tCXvQ<yF85RLlAv|eQo>U?
zy2sV33iwPhbELAI;9)pS?;{t#e>QNdE{laybnvB(SG+Ch-qJzEbCJvF5^gUqK@J9!
z`_8s^zV_kX-l75zp*g4|Z#zLV{MIdZ1JIE&DxZTEf7;l|p=Rrc(+tbZe|PW!G~4{C
ztMRMUj=g=A`PLY2{j^-Ux7h}(s)cbCoR+q#?wvdfW&i$T&gj}&whK&4WmXRMOJ#n~
zcV*TM&wk(R>KU0E<i^zGhjsBhw->^t4a;+_m(`BWuAF?u>x*VvHV0AM>F>^e(<0QZ
zv~5;o9k)yt$<|rDX=FW|JrQ#pH?WQ)?<-ovKzsYs3wU>*LI^`9e`V|P`>PL}m}bYx
zNOddpfTA+g?+u$c>RyW;tzQPrOr_swV<H;i+_Slxr;cb8<;zf5-|FkpGJdE@M~=!i
zj~*ugDmin{%eZT*xN3_j5!0#=6lkjg%@Y4wmWtARyCOYlO6VcPJp2x1LS1erz=)b>
zfAs!6fDJgQ4g+M-Ope4H#pnuN%hK?LUy$$vP%G^uHzd@BbIOW-lC&o&(E1fbu4Gm-
z|3Eien6eks7qH<`#pHUbAJ8cR++>_U7GtUUwgqz=Mis+$uYc^Nl`x?JDtSThzOjj-
zm4*v1Fl7%lw8XRmn_AC1KYAX%&UplF$F0v^++gd0$Svs-3^$7Y99zxJ!PvT+nHLY&
zh7e<xzi2ht?)E6QKuo6e5?FXh@RAplI_g(#sUNPXtcTzI3JVhh$7XQmgwRB&C&+F7
zickn(I)NxDk+B7o_qdn@z@k!PE*7^VmKV@*)n}7Un62@*<IPvu#+KaW0QMzI)V4pT
z{3GgjagB)!JP8EFoJPx?tbh_#eRS1FUI3dMs4YTX3{WN=09Vi5?vGS>8GKqn%$e3S
zh`8jVd+Q1XwD+XbD>EP!EsA+fcv5;_c|3Ffe5E{grr@5n?Uz{rz>@r0dvjDQJ67PA
zGi{*zqNL3IFCe7%68e8a*sKiyn;>Fh{Rg}M8*Ah7KZNbX4!3hk`30mP643z|5ODP6
z5OAYu%>H*-pR1B?<5e%Y125wA;BajOh5>rC-usUtiimHKxyQ$>M2@fT`<v(cdBOcr
zF#F!p`Q<C}9D3qUV3DAeES!KkY-ugO?5}&0qK2o>9CYzghHmW7>yzzFEfx<i;1kFk
zMc?}};aqqLqRcFX@NJ@Q<d2)lmEwt3#x1E~(Z3XXykG>%?BDn2$tW72U+4Qes_KP*
z7`x#g##Szl!GRHvK<=*m>V2O=O6L$=$Q$1^NDk)6U=mv}Q%)J-q$?tC6eIF+3zs31
zlAEWlcCun;sh*uj(8?{!W;m%zv`^4CE=#&$Zg#P;fDtO}QPXnZad44aRy9o$7nKEX
zGKmNKR@73;Xs*Mo3@$32R7wE%O->PPvSg1|<rdrgKa9Ooj405uuG_Y4+qP}nwr$(C
zZDY3GJ=?Zz+h?xq<lKj|ZnBf{Jf6R*`e8YKLJkudtC0q0I<<mnDfkcYQ5b)AE8Nf`
zLN99mek{%O_7yGIdzM^JFFndot+<jp+DZnA)z<#0lZ@@2YxPv7UyKWz;#4M~ijve}
zjj&t6>NSTe;8)rA9G6Gw(z_xkQaGxv!c~_ntMLWB2rxrsDYkWv$!En})uj+U%-LYx
z?wpLEQ8O@~CZ%KhWGa+tk)uq|)5G@ti0<BWdTO`zFPA}(qwaMoWqviWth<QwoO@<2
z?nY<AkW<vdBX?E+LamwD%FVBMY~!!W%li8L$(*Ds!tG)jF-aD8od$vGx{n}9dPxAA
zw+4)(r^l@`*@IUm2_vXsnN>DfHq!i7O6+mhq`GbxtzivXdk9TYrUR`N@K%aoThaDv
z7p$tOTekP(4uDTVjVOHor~=bZ%NL%e=(_iQi!g(zR*8?lShwe^_+Y6~B=GFD@@eY+
zrRZetYV3Mo)q|ur&nLGg8TWA=6Lp=3xReqmFOTP<Z|sr~)hq58;!?>m^l|adB))VE
zJVmRY;@kLFq55nSdyB4398b2$f=KluwmL<&_Q(DH+Wh7G@$z_9^ohYj0)*%LGCutc
zH0YlVHd%=XAn)hn_53wmpL5v==Zkm$yv07rW-|uOB9u^>NemL>{^>djr{xdbMr*P`
zKNRi~@bOv<4H8JuF^+c0^W2unSILgpmkLKfZQ}B*MWJT5y$hYGixu-EJC41noHj*y
zdX40W>k)Qp(A7ja9-_(4WQLCA%zj~<4%fB6nq-!m*}r|u;d^7<7;rCH|0Y=-6B6jm
zo2D#kC7aI~^mo(|%viTAY1+BET8M7RP|d%{c2WJ$3U3HE=<Y4+j)LX9A(e!(YZ@y_
zVkzv8?*3KXH;ov&hP%Mr3ZE!ewttlY<<OArgb%ezx1Y&Cp>uohG>4r_j>K1}ivsrJ
zqOUuL8+9z74ZF^Z7C8_o7Ma;(<wz~t?u*`bn30{#<Si5gA!C1ratdN^iiUfe+)~Ql
z<%L*=AP6&0cw328Dmw!1h<m|G7R(fQnp?EkidFT>Hcc7kN71CJ0Mj4KeKr=9slUdU
zbF2Bs<NZ$<C)o6ALk=m<(^M^720h(_$sl#1MAL?2fuXTZ`a691Ob8b3u68QMmx~K5
z_^%w^K0t80`YlgM{PESah-{aj&dm#fK$?_h8d6YRNL0Z>%8+MCMq>0`=lxsSSk&(7
zn?4sn2Fc=(j}0Ye0HZ*6xwn4o3h5HYsisFd<sVxLQ*1rOT!4WH6ug4H6txObu^_2Q
zNf+bnK|~&zZB{OjX6nzo=t$Daz4!L4I^AZl!a8`8^XdAXR>OEq*H<2vw=bWKQJF7^
z&1YD#=_1{$HgF$cgUOFNT*Yl8B_6M=pA|uYD0-WZPi;IDinUpxaC!cvkicCBXc&Pi
ztq8^Zo1$ZEh(_N)*i(2;rIzj7ljotr^Q5!8TlFgqabPYv^+@8C0AJe`2fN2dqLTGF
zsis^&F|?op%o;V7`7%wV(V$LGF;(63+=(Sv;SMgJY*g6mZQND&atjz|O39L$L*xqZ
zu%<iiV0@91r>g1aC$58jn>=pfEyd>HEZP&^<zy|K4g~?rKenkN$YRJ@`5^VIg%q&%
z{L%H80pa_t*$x5kvL+kg>($?O0qVoc?ZX!uGeA$kYAtO+0Gq_I2mf4{fVdM{%VzxL
zct)}45&9fo%W}-Qz03&Sbk$4Sl6Lt+Obe?y;c{BtsTWXbaQpUC{0w%rYm5EBnCD!v
zuBv}{c|+Y^&G`dW64wr9Heee&C}xlvc-MS_!Un)XDaB#D0>E-^Nm~O0;N@qQ2P%Fk
z>3IX@;ReG~jQ|&9jvZ2-(XaLbahH>8!8JM`upm1`oOC_;ZYiOqcqdvH9aZuOcSUdx
z0n6s-(wn#=Jgv<>_gY|v>P?nm^2?sV*%pM+OAE@FeyySR+6=5CR(~jjtX&Pmq{@#$
zuP#iSayCzM-?Z63d8a=-i3LhEYuc-fJb{=(S~6CFSK}^TE<RwR2vn0kR;PRW;?<AJ
zThj>$95hu18Yy~qne&k9yP|c*z<o?TClV~+bnLUjOb-k^-m#~lZ8o(;GYwNKVqC3w
zy?Xaqs|xIrjh$VE+&WSy8@g#)@TV8N7SI~sL_>K0pCGJ$i|!-Ee6vt=tp|ZbtPQpR
z4rHlks6{aaBn=aP0q1h4<Orh2xEVhZEmh^bf&s22tAm{&717-M=tjwxQYvfb-{7Iu
z&SX;ShMHXKK5fWw6r?-tA{_aruXuZd9zsjwvadt`@Twc$X@{Jb*X&ja93Y+bI@c+#
z@Q}x;oSKMgNN-K5eLI<ZcUwZ;s<Mlo*L~=wo9OQ{YsR6ITu)4o?D4IvZDOb<aqFMu
z93yc|%!`_?>p-W@4usY}Q)@RK#~7Z};FC8`S<^}GP%mA4W;xq--buDbtCcs-?eD}_
z>XRzdMmfpfRwJ)`v0Ju%krM`eTF)=l@z0(ndo>JqdC0sjX;b-F<+Rzsw%Ns*4J(C3
z=<88xn{KUE9brw#Yn}O*GB;Rxz#CGP?NR37184#<q8`L+2hDl#Z5U%aStreS?$pOM
z)#aYe(YmhhE$f&09blOhw=gKP_*J=n=I{fg1PX>Y=m!8g4|)n453=%XGI{lN($fFT
zk?Fp{E?G`l{Sc&mAkcjv!YzN6^xQ1IT|w_v5Oiw_swM8^xGc)wiXZ82c_ht{YL;HZ
zmOMF~SN&j0U861^BEz96r9n+GgkQ9y!drcc^_x=)4c)u6Up3xOwrC7WXHTTQbGl&&
zx$0uTe>@8CP^Au1q1&kdtogN`)!=n~y89wQuN2#sXo`-V{eTJ}2533(R>U|RYj2zI
z;=-*!NuiEulNtAqmO5qUnO>hd=?2oSBvxiTmy1+l6eju_E09n0v5xunxylbpn9NDF
z0(C^O?(?Mf?}{4;OK?8`Rd&BNqF1-YPakN0=a)Us_gDRRRPScC?6OIzX0S&mlL{kK
zB^61atujqkUT5pm`v5&37c2iOa56FcKNm?RCI*K89dQ0jup(xM>HQzzjQRiu0*-Xo
zEoevUK;=cXmFI1_^`|H;WWkE#ZCky(0^*e+OM4bhMCpr9xxT5--Tr=idii-8nxCi3
zIwCx{`JTnSLr?n+Iv(9~N9+dy?|SWX@9Wiw4$Now`>S>O7Ww6W(7wyhEwKgs2i$?W
z|A!OFA^H$)%y;f_^C9@lKc%fW2A+pcvxo_J7Iq7fiQn>f^HW}44aDx2uJn;ibR$$E
z%oLGww?;qaV?Sf>2yYx)N-;L?C9SLnRWo|3wf5>+!9~HGzq%vEgKe&PkbH;mV3GyC
z5sq@4r@U0SVYH(=DXvwL1{O3kl=Pwhc2!bq_TYo!T5TtP!s_cKJBe8OhV}yWQ;MZt
z#T?pJ+%=Y??TtVwp<DU)r1iRYYDL@7s?wSilb-vz8`ghK*tP6H?MO&bWQSWn%xD`!
zUZwXmlpt+|tRYuLXw1^cdrO7OopFY&7RX<)F^vuNk`Y6OdPGeXN2E=%F4Y>0B~6@7
zIR%$gku3^`Wm0sSGj4In%De1L?};59hdx`Td^N6FvrBu=^jpWNJ4f@fLcL}Ce1ooS
zb)6p_o}!nar`DpH%h-~3zqHFKZ&ach9bCq?XF+TA!TX*D(;0!&<fJsu*V!O#4v)-V
z7Smb-an^=QrrG`Fd+E@11G$?up!rq>p2B<}%C=$(@igrLKjOH$n*o2)spJAfbMe*j
znq{1P^=j;Pt1nccb+U(nex>SDF!31R2G)G`(pzNyt_4V9l130=o6A!{DB#RygE(Tk
z>ZIJ!Sp)aA={i;RW`4BEueBP4b^WwZ63%y3b=5J?P^&o>f!a2=%}zgE=38pe2dM!i
z!8QE6L{-3v8AH{9>Q){%f?XbqB)Lh?#;{E#x9y<ZkoQl^VZxa()E8W){CQtYRe9gL
zDX{fCCccL$!J+OMb@)ERJK8V$SuJMdtclPC_&_Y-oo!^`+P3%kD6h|QkDSG>@R=h#
zaLF>+ebmD9WYJzP_iCFb4hJ$rC2ubtblu9{K*6bXh9nYA5T8cyvl9gsO+Ef{`RuLy
zFOTJ-$LuiamEZ4YWec(gve6D<7EH_MYy5V;JpZL6Ds3}yeT|=i@SWc|_`sbIH-5yr
z>>Jk}li<kFNOF~S4M!AJg!-+26)?S)U*<MU;vvms4M>Q5>R_&u%&>_N0q+4{k=op0
z3MO|cuP|RWEhDztj*KPhON<j_sxSbmDB|4&1r;y7IJb#=5kjjE7l3P+6IqlZL;t<l
z^x)PFW-c~k{a;@VgX+Ir4uxdjT45{hBuz^!qQVZ85;VS~b|_o!l<<BE3Bn(5ihWtl
ztTEUv@XV75jGjU{J4117e~3B?{*Gn&s(y|eE0lYaI*4Qlzj8?;_}NGMO@5_#>f^QQ
zd+DtId9^;n<;vh`-_v&F`ghs)?`caYf)vsE0WULx>k-6P$I>=~kwx|^L^q@i=1YrQ
zTanXs3X5&ns4VGQ{$WW0Tu7Pl2gNr$WY8k%aay*|D^`!^z?<<J@yG(xKHS2;N-}2u
zZP^LPX2y#gmc7*bvZQipVde#F{>#dEcyIEtb4UQ`gu6|mc{i#KcsFY+Gf9R;YLqXL
zgU@JopPb{4oK~yF9)W?B?{oRA=WTdGz#}fMpfmg`%@M+hmA;BKtwl|HF$2?mNj;OP
zGN~Fht8VKmwA2JSe{t!9RIa>UOZpN?FLmmd6FWFj=hyOHwu~p9#mUd$f^6BF`GuI=
zCzJIhSV7S!`BO*Wjlz*7$a8E>w>>o%pmFqXXD$27WxKbQn{)GpdoK$loU?klZF46(
zu}ozJ($<6nGai=%+<w63{va)9FDuqAG&s3TN?=V?lxKL3jGt(oDxiU6Px==}p$G^A
zlEe?;<Z)+yvfzayj5MrPwjxxi>!teHzIbrP^iWWq$lTDBRUr7!0e$KSoNI=1;w_GJ
zsy)0hN|{-5VU>_!&L=Q599?=4@Q$K`LnF!I4K^xHcRs&6M>j_4i_-Fj3+^Oge#efJ
zKfa2M10CkhBDg9xPU@16TM+%iImWv^cx{f15h=A~+h8{>3e<B%*CdooVNhs#b@WU7
z)k5PpLqBZSQOv+_1^ASOLl;CS?w{5o`DWm;D}u7+57M#9Wm8@H^jY}&U5-gl&atkt
z_yI-K*K73p&M@Am-QmMwPxM+f6m)+7aZFgJWIWPV1VIGX#ipY<;I@`8D)`kPW^4Q#
z38_5#W}R&Pd_a9T7y(1l-&{>^2{YSnW`h_^8174~m_CSrW+^F~{8GTLnoeg^bP~t!
z0RWb*k+OZFlLBS>T!UI<1Kn<nw<T1_l?*X`Bekmq$e@+G+)+}K-V^oa8B1TEd6tV9
z=-2G4WIM!MNS0HnRn-cpYB<94d3Uu6F6~XS1)nsUAU$u}8)Uhjy;$~@DuzZuj)hBJ
z{<5%tuI00Cy=xuc3!9IExO?DCdN2E?F>y6mj!cFYGe`YOn}C_jH{Av{(s9%p+1R(T
zbU{;emP5pg!Ho-|2(Yb;I-;}VCwR0p<Fs_~rKl(JghJD)-dR6UmzisD0lL!V5hz+Z
zwXc&As9dorS((WGSZiHB0mH3wFY+2SG@I@2BSWqO2sI}Dmn3i{2A1;4)Ag_}z#Sn5
zy28BA#J<qOkbdAAH|M~iVQ;}{O*xUdTDK1Zh3K3ggSl$ou>!EGMaEL`Agz6+zl2|m
z1*4ka^+~->v^vwBy3p#mw`0p{Q<tTy^p(L~76f%6n>E4BhDc|K?5hL#S_p|&!_Qq#
zc$cwXn*1HQVFG09*1p4_a(d8nl(l}BQ4jzaF)m}V!CCuDu|tPZwIe;ob4*+Q>k7Bo
zihhR<=YxPS;sL2J?Rp#t;0ER9oXOKCNO=#x7lVW!Sxj313K!D_PxyQZ@YVknbC{X`
zAIxE7<@nzTvX3?YN09xhz67s9oVWz`FI0c?1)#%pPdh<5A#)Lj>DM2kRZX(>4|_ZC
zkwRUY)W1x9pZKJsJL80*)$PU2n@Q>ggMUGtgLf~!3SWW``lKE6X9NEU>fFCRy&Kb$
zV)){@7gi#AFn(`8Z0F-CaM}`fgdeGu!o{PZv=g)wyaCSIO~T*XWoz#_xepuTmP8?G
zuo=Nkg`eG@ZQpfcqNWLD-#7Po{_W|${#jRj1`K8MV*m?nv3t8brSl;u4Q=#O5c6Oe
zs_zTG?b`LOR6gYxnMoYA*yaF_ncHY|p=a=snyDPc4R(}t6yz!Sd4D!O)9UrbbtPL_
zztE%NcDe{^WT!`X$z}b_Sj*oyV*kw4AjQR-`fc{wqV36d?wh+y*$~!m3U6@PNAAh9
zg)Bgwm`PEd-PddZi$D*{-Hr_ytI5M!hOtVP+>xt7<L%hbaMjhCTUMZN4AMgQu&h8P
z?N|rr#5+NffP$Wrwm2lLIVy!lK^MNP8E(QHIA3|<66v5KDc?3fmsd2BQ(su0D43lD
zwB*141KT5$;^Yx=zsLR^0E32;&WvkD-D(Lz1c@!wm74Tu&}KCOS!z~|_7pGQu#5It
z+**7+4Q0M<^X)~G7yX;+IuBlvX;{|eO)b3xb?x!YiJ^#X{$|@9lENRY-|MRz-)z*J
zGN^i+Wn(%XkD)o4ktnFkO<FT-nIo2eeAX{XGsyTkFUK?|Z$vi6+R5nV>8t8!U^zip
zui~VDpe*#m;<EvVDG9@|g!%GjHE|HYZ?R=gUfJa02`6FeIxusDA!`<C3=pE;GR^q?
z2{JQ}#i3VGvu!#<W8lvx?ui6ure#CeB~m!|LQK_}6(hkpbJiH!r}|wl=rx@osKE?z
zm9e_)j>AP#I3$zQ!av_vzlAL$2Rtf^7yZe&CT8haQrXlvk}-kI2`;QTBC<}Po2WZS
zJE<$!{<ZARHc3@laN2mGk`yY2gYn!y&#%2D&`)V)aq?MxV1SI$as<Otm#@f=mJ%h=
zdes&3E$vfW`Y#$oaCembRs~ASL+(lrs>;On@6KIa=`&-Yvk(8Ol6NyqLRR726SD~Z
z=a2>{$#3l{sP2;wL3ZLJ4yFfKelG01NeX~GehVytafd_SJzW86XAh*sr=B?EA`#jA
z8|kzEUhfWL_HiE)NE^1p>P0?6sODbzV)d`wpwja%U@81vO$_??ReROvZv8v(%v212
z*HL;9a$kCr(gX4-ZPAZ@>epFMf|S1EQ}Ht+C4c#u{_%Kf1>62w<>I^CGY=A@=3efF
z=MvLQr<S~qUiD=HS#Q1}a99T)a@_0fc0~5_#c4i4Zk;!$U}xjaADs4dN0?x$p-qkF
zb70l2R?<v(L?5G%OxG{HzvU04hHmfmn3m|gNi%LUb)D&e&^!?51?{Y3w^qI0#5p;h
z6jw|T4ERLi0CYd#*B)qb*nT(<2^jLtE3g)#$Dyq(+NoemD~Q+Mf2GRZHtl@qT_OmI
zX15seM_Br0f+f@$U~iQ`9CW`WWfs#KvK7Z5FuND1e8{r^%P@<VaxK6x&~lxvpVvoF
zkdtz~&a+X56334=V&71rB(qZ%UFo}0urK4jb0<+@fDNUw`L+WLC-6#3NM)xYPL}g*
zN2fh3G=F5}{L~U~JL2|v6;ET`IU=k{Ukv%rj4A&H%oUhVzU&qepx+pxeLX0`lw<_P
zgfU~BO;QVQjncL?k)&AD@fdlOP2U11ZRR&3>d*I1jg%Zrnd{cIl<E40VMPHR;0jtV
zg%BnRN{<O2973R{D-{8a@c7hcqkyVeh)xPRHl40GcySN&_H}nTU8piS$VsS`=ay_S
z7sTHS@1tLl1HL*5^vptBA6r5ynjAsU03yJ0wu#g3^ASF;Bzwntz$d~*M49o;>XfcN
zG4KK(dq@0rOkgHjhsaDX0z1|ypNMBx9*TUAD1YH%OM>!N<om(=T)PM!_m%n3dzxNx
zVLp4=DTzlY4=Hr~JQ6n-bMFv+pu1S6HL{>x!$cos0}mk9l&{Z0g9^RRnR<~<De0g@
zUS2sc;^ypL*8IS&8<!-=X-+)fg;Dwi!deRXuR}J=KZP#?fxVFx6b}#m|8Lk1BLn;Y
z4clR7`QJ_8%i4<nN{f;FX6rTf^t|9tl9+MeJSl)mu1i;gqoI?~{&W#Qn0RhixYh{h
zp<jQ0e1=iino7$ZOHf_`M2cS`at>koJ;HWC;$&j&f_#~L{exTBQWa9mKPgnDVjgpg
zWThIVTNd+FsbvlfltN?heQ_QLQGASwHdlfJU~2mOhh2<Il*EE5T;+R#C{;#wK_66N
z8ihJzs6HJ>T=gMg5#WkgQJltXGfN2xzOl^0GD@gO-47j6P>gDQAf1&elY30Z3grqs
zv@BC<1SeLS53tEgq7sFMkjw2Nu;s_fb{fvD=+gb9OOeWUMe(i&Dg<z4#*lm7-&tT0
zTC0F%Vo<JZp?{UYC+9|Z*szz3mIfqj?;5En%H{}7q9~i{dj)Av3_}KEJ?!lc#EA)2
z;++FBE{7Ym4)#FtB91JYkJlo?tnAaoCY0D&Bh%6dF*(%PqY3CswA2a`q?SefNRF*E
z#}QiVfSssJtJnw@D9N*BdBiY)s0Y|D!nY>=HYHREcd$(OVW2}fM(`AsD=|y$6`Ikh
zl1R_&Q?hfu2`-T#E9cS*4`-o<8D@Up)j~9Ok`=P72LT{<q?CF^rdm5C4T$m35~?R=
zGlgg*2&_T5mLmQTCD|~Aw3t+ptvPw=t1H<+@+dRu4-aQmMc^Erh2J-gD}lEdcuNM`
z%qFX4GGwS@hKYdqP_`k4FlZI1Y!8qSX8e*J?k0q}7TI(y!S9YBKd@@KHY3AVn|;QW
zcuk*>N@CjdzolH%p?snd_R9BuHGS{D!s~us9T@Gj*VoB;eY)^;y_H(_WzoxZjab7E
z-`j8ax9qm+_rVQ?r*F6T!{+O@;6#m60tJ+~qw<h&%6*jit{A+jU$;NvpIoy~qf0n)
z;3PbS@PN|1KdqL(HRmHzcG{dZun1}7W^KA-6k>MhT=YsL8Td_nVj~*401I)7Vj!Z5
zrxm?$%!R=YAir3e&}Z>BkR(3*h$WztC)U;WPiWw68-|dLM<NJEN{IhkCj~%Ka#A|7
z1dexgqTXsj5J>!8)5`>T19&?G)))}6Tf2~^X6blHbP%z*$Z3?X5h1;bbqcU(Kod_C
zCst^{*>vl(rfk3!7M!N0)Up+dngDbQkJ3h)vI-rJ0q`YZOiaThg;wPlL{lvUHUY(!
zQ#o=1J+sjaJO%?o1s~py0l(V@zmmlO7&LhE2uS_AQz^K&03mthL+6f#$Z=etnTYlm
z91VKHVuZau_J|?OfBol^|CiOqy+2t@k%+EtW_EY0v`j%5<qrq^-j-<UFlKV6n(apY
z>FMK0HGHz=>JmN6b+qKJdvtVISiJvl?eB{!|9KY~41C@!y_d|xR{yTKzPTqHUtpGE
z7(6SFM8XH)F&D;Te|Wk4CVKSylj>@poxaaw_jloEt$#CG6nb(#7lzBcJNmb&sX=4h
zeN$JCK5g0daZ=lQ^KIAS`cd_}J*xzz#<)0CZwAesF14CvMz7eb(>O&nq`V${&Vmmn
z#j#Y?PftfC#-=kM9jhVeDgXo^J21b)NJcyWgMyY~FW}+*bTW3wG9Pw${!8;aXwuPG
z`tJHJ{m%1u_+($}wS4o>_HtzQ%x*ise$TIi%f|PHAOHF8_oeVJZkm~6m_|qb>}K-g
zyCnCsl3aoo<Z(lCvx`&OvkaI2La7T^X{5}o0^*>hcSlAW|4Z|uCjTYf=Ii6)4Nmaa
zF-9K`m)trrZvNH|Y8Uq`VTVPI@wtzv>_YgB4<}byL;mT&-58SJxhH`u&2gbOT}Z{4
z^(wW=UsjU<UArOtbs#E04GBkdRk#CzQa44=ra!r5^z)edzgr1y@KL1OB8{Xx+t?#1
zD+)4VDAL4Hr3_LmQan<Q^fAK7*H+DqGA|(kK{mWoCi(eaFq`ImZuaF6qb6?iv{o0F
z^Nt1IY`g_Cpw@M1mk693Ed=n+xPG}bX35EqHxDaZ0o2B%PU~$bugAiHI#G>=_yf|{
z4nzg|K%CH5Qc&1F6$}*#@<C<U!X&qpcu^^_Y#8)WXWKtI*u%t#OLBP)iY7J$wdwmg
z>@fhm;*VL<G^GShq44t~bxkQ5IYIS9?C4UK6h);cv?!buTBOfW*H_YDi5;SHf@1E3
z<?KsoBZUwWid&Mb&fC)56=Fu|(BGK_{=rK-T|gEaH2_%rwLm6)YJe%PgU(8~S0L`8
z#wOF(z@d%JqN_c)jW(f2m0Jf4RQ|(;D=cb|qD69cBum>>iaEJ;zBgOV#7`22?*m%C
zMc<VY>kA60^41k%$W}xriEc;#^EjlKq%2ch_?3|6ea0@0a)`3-{=&Cl3_ZLbTz^eP
z-JA9aaNdNQlzF&Jg;R_MHcz@$%mSq>fz91S1mivwqovxOnvsg22*z&{?cC78Ifx8~
z;yUBiGF+S;6RQyrDnt~0&=digY`EFE<-2urJg?$bDmYQR!ZTARK&UVJX{%vqrt*3}
z9_4OF#`Jh9o$~UxrS2ug#TBm#Um^7S3fdu9bmP_{BM1j0lDU;+)4+bwgh32e^kXTI
z;2-F9VD?9?(FLfPu7amW;%SQ*8Hj`|)I#aFaZk`>_wWD!*oNE`S(RINFBZ^@2=0io
zq7)aL6XKIa?3N_CxpoMHiXZG$?d^+!&KmRL#iLGm)F`MuZ_{b1BgHPTVR`Ka`;I@$
zY$pMP78L5_jI~Gx6}OZc2y~FpX=>hkgSzt$KMM_9p@>rpm{WP|SdN%e<~5q6Bh;m$
zLS7f6&T`jsINZwsh{de}q+$S|;^PV^0tw8&G+%#;kyJIyCNwe_)CFN>A+Z8D9ZUxq
zq7Tgf6m=G~mK7ZvybexStj{e0uRv{bL&q`Mr`6z47|MsINV0-OhW|;N#@iW7g`6Mq
zMwsT7fmrLoL=|9iK<yeT*G|`aSs5GiZgq`qlmMAj+!j@Bh=n@O_uO^0=OOQlJSFtU
z%3IvnZ|JFO5;<i!aM@FgDkdpT5wo0)qahHisKG_nuH%z=%8)C&HACm0=6`B_ZT^iF
zFBTumf&;(?0cdXrRS;?rHxhQk1A7pnuyurv2@g3C2E)dotiB07pnxjKTs0IZzkM)<
z0)Dqacd?xPDZiWU8aGOR0lxtG^?d%0o_~dOwLGAltIQi=8F#DmmMRCr;fvjuxndo)
z8(h5jJiFLj-ykZQWmKCkygFrx>=}2tW+@Gb%EMd4YdOGmna$2J2=%&lQ!5r$^iokk
zxk?jz)du?$d*Z4S27^`kq2LZcy7}UD>9%3#>3&9SU!|)IwvelWi6MFJtQ_Xu400ly
z*1x}~0aV`uVSF1;XU!KI6CfrgKuk=4TgoF<%{RY;v3<FWN#I0fP5uYMo@0r3*!=G-
z&EMIk(`e!hBwk&u>Y4dMPj8ty6LP1cq2A+lxWhU^F0U5OUY@mwE8dA_eh|*1gun@|
zEd^_2;;xdHJZis*lgZHlvidXTD38xu%(LnVy{CjZTtr>+0^J;H-YSrok+`vUghl{<
zCV^rE!6oVA8N>o{lSavWUJSs21lQiMoeh4vUNV8;gcqMghQ{h&NHbN)D}$ebUNMgE
z;}s-|y9E1rv@(9KBaOvBj!uYfN~~WnDQ=3S6&t^m8dX2|IZ{D+|8+aV!ukL4v5cvm
zxr+q>69W_b|LO?!&n_SLuSTWkRs9S+%P>(@2VfxJx&?U3HmDc&geBuBW$5S^URg&}
zv5Dpt>Yuz}q;y3+nkGCy7^jyvCl9|@hXw9Gj<~Q>eBe9a@!$VA;s*ZXi1YYi{bH4>
zgwZGORgkKEZ~9ul&xQNHhByb076_gqPmxC)QJ-`l${(A@HPU+a?j6bD2H@NO7~=HD
zhIVKDdH%}Wr9*enTE&VEu%KE2hYpt?WZw3t=kghEyibk{oWz-gxVZ|G;WOI1NYy0B
z?W%f5OoD&1;LBnGOW2?b-8n%ZMX{t7-mR5XdRWFmANr6~wa`|&fc#22U2j=VGJrB=
zUs{>vLdxQ1k4evgtrh&UGxL@q7cI#Ui>8he;1Y(i3(D@l)PBA^eckN}qt-42#o)tu
zkV(M*etAK_BS>Ief|rJX<M{n@e|fw$@3L+R`p1Xipi;&fx@n105l7UH-r>3^9!-*P
zpK9j~=1^QSgm_Y~SR#Ql0Y2~E<_ohwd~2shDE%ewEk^7Ovo`y64TKZA$gG-i?ac==
z*>48mhBEkec^b5CcKXnP8}BJr<Tw5F=i<S_yiOKB%K*jZImVd!vFl{jj%JeXoNSfq
zTOq+1=oEeEHXB9tsOvXrGkxgpH%cA34`4By0SfoRwwZ&lM(;|{X|1Vl(V*?8b(0Jj
zgTuweWb}bzz@EySf$W=0Jq7fVShmMJWokG|h5V<NBI6JL$T8y<L*T2Zc#iQszY-3~
z*pqAlTM9h%h{Ae=G*gbk%H4~EreveT`&Sw1{A-)?)XjX2Xk&uGB(_EZ_LnHqc>k`v
zv#fJWaB&Ta76kdZbs*oFNJGyfOwm}dGE<6*maua)S!M&!#KhjPWd%B}skWo=GI(uZ
zC<G)L7Y2f?Jd1Y>?{LgS5Q>PYEh~IH4fxzstdS92Y^iFSd^B~XIrrjg1den^Y+m=G
z+ZfqOKD(_X26?7BZ=f9-J9J1GQO@L{+cr!NGIkA_HL`3?gczhoXw3+nbCMVo+70@g
zY)HiLJZEi<K5qDgF0$p|4`q^71WuAdR;aaDNTfxvR1g<pI_;v#ShaIB?eV*MsDhbD
zr|);$njz?zwiXj^BAMn{g@i@Z<7V0f1luB9dvbGVHw#3xpHvDmz=&!YXDU<Q;y7xR
z(-Vi9W!b4)Q6na=85MT~;o|f^TNUX&Al!BwVMYmZ)!MnFk~*mlbTi4^`uo8I^|)By
zDb2bwn?*9~nUmb@BhDhpwF8BN8t`)GJQbWu1vgX1zIZjms`MoYsUg^qu<=+L%^dmF
zc&13qXsm1e7j4)Utm`SO%8c`}2sy&_MI}=;mjKh587=CBxglW}_`7viDhpMrOf**A
zD!tc`B%vy(e=VW86N@t|Hx1dSbcic`YvcRH__MR129QXKs@pOL9<Irz7YP)#uD?(N
zzKWL&t0%a?<2JNQio;noAUgsZAlW;a?5yWs6UvduZKnP-=Z5VrE1gvqht!G94@QzF
z>DI*>U3os$e1hF4t2-Uk`^M%T>{^SN*wLSoOT%)?6Rp}7NN+7pXukPTD5RVh(8!KF
zsE{t8*t)rO&{-y<tON;IHV3cxDoP?y<pNxEwX7IsA9lSemF-%-_uj%$oDl@hC{2q*
z9whZn`ZBJ?9WbPzeVN_XYdR|4yaEW9QcNRmj;vODx7IWv)j%Y^>UJ04M2AtY`-rXU
zH92L@I!{PkfBYqtg`5E}m$!_vq!T7Uh?k<TPMJlGsd8J^f)uH(-7&~(Ef!uw?OGC2
zPqJrfy+!GX>B_t?>S+>NP>mk9pISR53LhnQN-w{sx6jIBlYu6}+w_by5|j(}6pgb%
zDN8r0nx}3(qQ=#yKQ43C^_Hnou<%`$FRo(wGj&TqlM@Uzb{EytUb(dqHu480Nim*X
z-}JHdm}}QmxQb^6bIw|3iOf)@wMVfXDC^5B*_)()N4(gTg%X9*{a>RZNH1(u7ZI7f
zx;K;;phAYz&e_C9UQ}!Ci=~{7l1i=X{yHj6;ti%RmL#9jh_yEX5B(T=D|u-bY+*;Y
z6=eH`PJ6~*kgiBk??1{IJd&;wyQatMFcOVKR=MNC3tCs%7I#(-%S{JegHbv<9jf+O
z(lr#V+NI-dD7C-gCUeua&dKd2ZaqV$IcZ^>^HUB`X&Wu#RFj-tBh5NR0?#qdnJ5&_
zXJx&~`aUeAND`WJ+!XH&Q5v->sZO+|Akp+uSy9H-Rs{xX7Mf=N?C?O#+J0&!EqwWk
zZvJ6EEW&{vP0oh1T-8!|z7^yGue;?zGCf~v9~3oR&w`;NY2uc+5u-C!@t@!Z0$n7B
zn5GT@5~3ctn5`O~iG6SHQ)-X}cG@SioLorgilxjdC0$oL>Go4l1<OIERGS|YfI+|x
zS-74I_&nZ(W*1P5`gnlddF?uj#4XIG|58p}s`XgL{(VB~E8mw&6{F~{$NjA%1#c*{
zVE)bvhr34?;`J%J*}qZ+>Ne!U^vfyq)^ZDZ)z%r6qs#C+6+MA3)uKq_aMj6v_3r9y
zUG?@weKD-^-{}2Zs&_Fyr9XxI$kMyJE|T>oO-TFNeQgFm5)huLlVp5uxIV}gIk!1+
ze@rg}x7%S=@ua@7Qm6BaH7}m^lgrV1cH?gq%!YDg?r??S(FvXMSkK;c8wd!V{&gG`
zUi`1*!1`avft{1>e?<<<*phKLoHu#(5%487p&J20Ma_e%U?(SH+hEQX7{;#yb}p|t
zC!fN!DiTqo#OT;~z~@s65?hjqDN*qHV0=CtJbk!3JyaMp&RQCr-M)Ary?bwcdmI)|
z2B8LGk^9%1ceCU(r316fegV#<cgFYaL-xL0RQnEL{g^=Fp&Yoh5IBk)MGkRGZ`1e0
z_wCch^`1Qg&vC#Y5D?mXBK!A`J|C-_{BS(TQ<?v+J8u{HdOL9bxKQs(jJ96}Ev7-y
z`-|<u9)Lp3#@_^_k3eC4zq~$tK39cEbvN8llaMHA>|&Q<8Vd}y+XgX46Iw`W25Q93
z0aHLskb#;eP@_-Ur;f2f`xGh~KItI+0E(%<KR-+^#C?7g-O_a#3T@?L-SGKxITS=Q
z9sz`|K^rjyOYZe~^!GHuh5_e|AU`ez+63QZkBnrT7V`%B$z>oaNWB*XkE@IG4v~FN
zGIgtyf33j8ZW0tV_fa0zXwEsXo;-MAcNn5~Ru>7-?um&`a(7LQ>(}G-7BI*+(x!Y^
z0TRF^CrcDW`w(!p`AB5~>Z-<!Dmg(wa;-D-LcG$Lqsu&*Hlt{X`G%HKmqRea%tW%2
z*1=HLRl{hQ?39fG@H=K#*bbGw#sxO=W>K-VlhcsEU^QP9n&`}#j`JkdRWKAlUfm_y
zP^^iP3DXAKypJ^<QGx(-R3a(_h*EfX1T)PaE&#4`RIw@(<Vzxym0maxk83M}mCZCx
zr$#a;f=`7WCDu&KNEcRS97B4><{j$fA=Oy|!ZxM+Fc&i3vI*ucAU4hXnEXwI45Cw9
z7_MP)5(?MGbq=xA&!=gs^t~rh#8zv58=~Igb8_YNNDeuJR0ht)oRGSL<6~T>_Z+oH
z01mvUBH{sHJGPi!5I$;?nfXS~Y(z|H8;?4*x#VuF7O<Q?Il8)KovegLIMy9}J~#HT
zo8?mqR2G}hOp|Hl7V^M>g5cNfuFARvHOrP|b`pejOZs(EI1Sd=oi`(ssB%nU^Qg!F
zEEAN}nomrQP^9?m$mF5<8Zd9MafNGDbH<shmvwlT?$+sU=Y%dyX=1dj#Bl%R;dSoe
z!A6F0Xv@ZZR5+lVs4wnFiYQo1^2IOX(0Z?w%{ETJX354_x6@d4e7Zq>{7YIds#S_H
zmyiLl@jY8_Mp>Li&(<H(uzf-04k^ZT3_F~aG%j;2`mb4ocsb@&h1`OdSKV(HsV^xa
z!LHOG>9eQP)~p<YstG+l70l^NCeI@`55GhaPDabKS}c2cSGc(6s9z>)iJo%*w7znJ
z`HL#HC@(|Ft6Xez9T8tsv&3?WafGGu@Bz-Z9ZZo4L&E-f-0j&8mhR*mc8GkU=Y)s+
zy!30IO0Lxq;73jF-;X0|?v`I>H-b#lBeh%X)vuQw)e%IiUQ#6uc*HmLM<2(wX-E`H
z3G7@h(mz-U&G5n|w+J*#=lgxa7^9+7n8q}I)Chy&RIhh8v-J7??@!qVfA;tdLGc-w
zEcdtXXnC<v;L={a-m>!MZxmJh5T=MNF(^n5*;;@sbTQvpa1h{uX$Yz;^@l13MN%%U
z?5^%OWU`^ZEK3v+02e3<KegbI6p_RJDi#v)fKqtcdJI3eQ1OJxf6FL4F5eFAKRUc=
zl$@jfK-Z#APA4%SRyIg20nrNMsRvIl9fl+j)4{LUoG?aiVF@5YjCZ%QQ4kN&<}65z
z(83Z@1PKT~ZO7L;>@sE@CHlr$PATFx6Kxd#C6YDD_4EHoOZOcR#jm;rIiW}~2O`8x
zwgawM`>TEb#KSF5rq7$7;+OTsU>9C$Gc)cwC_P``R)<E0bmv%c7VU#aF2;Y83Tq`{
zy+)YrC!A|i6k}YGMwHTe%_XbesC<A94Q8A+VP{;9C*!x#Lz%5BDm088E&2D@t7upy
z6oQpqr!Wv|NJjDfn4GKS5(=QWYayG+VVE&hOI!!HggGxHsdQXbLbuu>xnf53a*~b2
z{>4|q#H<pG`H`!t7F@C?j=7-}ONywvidRXkf7n+0_aspwa>}ElRyJr=ht;}5^RKxs
zu$-#b)U-}>QGK<RjTt1eZEmMP2_AI)l;-I+lUtjJNAZ|cbYu`KZCQKOwMw8}`?W<G
zPnKl~<0FL?Ua_Ln<}>7hxcYZ<`n4RqC0m#0wx<SXkOb^qj%4I8H@@&8uLH5C<$(S?
zt<s-Rr*-A|)(&(aGt^u(q7#%KrQt7Q3-jZY6OCgQZN@kzg5lT>bzPGHYb}6BGu&2A
zAG~?xp-Yk?=_KxxJXgyg*;r)F_L(TlT=VJbMoP_@5XR?^U`?||IuiGN0N}^A>mxX4
zvPwc6Kuj4eSxLE<Qjd&-qIoM(N<4s}92o2NB)L#$rg~FN4X`Yn`WQCOofeOYZ3sbD
z5tc6zNjz~-3G6Zu8aljzy*_GZUds%Ssm%=(JIlLYrqv+^z`Fz$_#Oz`{;R2H3$~8#
ziq0xKdnffmoe0&?R_i2g>t)^X+$W1uD|;vL=Fg)#NRCNyKa?gVA{JNQY*lN{{-xJ)
zCjDh@W|ngN8%r_8<yh+V2^c@l$x0!Q0bN(p0lIwLrEdWb|9z0Ukdwj7$IwT>sqc3k
zVBO$?H2IW1nvTVjj=)oh?dLiA+uqc7n%mxWzIOLrtHpIE(T$=QxKG)|1pRgH5pNcF
zVEG;c1;D+PINN}@LL!x9cSrTsP-RhZX6{2O%0@co7xt+36qUE|$`cy14u^ua%qPQM
zVvwbVy{1fvsa6g%$>plQ^0a{kp{FMf!f4L7N;{oBddG5e9A66qk13M2DBLF9t9oEa
zaO~1DH=UCLdzdGCo9TgWIzNd@%*RiqG}wQ)OwbzRSeIMV#Y2$q{nWx$ACB?7cpgLj
zUU<{hA4rSg4knIK$8CA6ukO1fJ519HPS3x-s5&2x3v*pY)G{@bCRx$j$#~V$Z4Yp%
zrWb9iB%GO>99yBB8K>6GcvS3s&Y@D3vJ7F_gB7lOvvy_60iLL%*&;Z}P6|jJ&VK9$
zZ>se(_?c+dydt7q^!KUpCPB`<Dsa@kOV_O158nCm6Cr4?o@kd?H&fy;Yp6Td(cwWx
z*`T4cbvwDpZyr5HSktgBfSzA<C{?zY@0fYVpKU$z={?A<WiH_F7Lnc=(Sj(&J(Te}
zXHS?>_J;ry{RM2`D07mE!v7eVAB}8ZJ^L{0tJTV-osnrYHD0>)Mz12bD~~4JPz2<Y
zn#Ra0Y};?7zd_ajHo)x4{gmaIOQzT?->N7(_iWW>>yFhZ)oXE`pQsJ8*lq1$p=W7q
z$My>e=Qi~8(Hb&0DleW|gHdCl9QM3t2{^1|pA4F}JwC&oKlfTR<Qz1Oz3fe{4Q|Xe
zTsF$+Zuursm`CEG_z8nN#Br9JJ>A6}ZW_~H&-I|4`kq5#PcRIvU}`fzT#lx?p#B&b
zm@#iD@ONKg9ZP6<D`d61RwbF+ynF^X(~3JbS;0+Feiqksj<h-Y_G(yk9pvf8)wC8H
znx;>38pQt!v6Xgu_yyBTz{<Sg-4p8$NWh>l!%(Bkyp|E`F`93~Hx;9;!!cW-^Ylk?
zLAAwJ<kGma1>zY>YcAsg!ncrO*U7%%>{v?Cfyn;`V-wRZK>H@TJpgYI(e^{@Cc=Hb
zT1Q<KQ@x6uFm+UsHeoTy9owV3=(5tNfb`%N+IkV=X%Msm&@YK!ALW<uT=X%<{@|mX
zl{oxYbY^G$f0k~{%uN5g?&4OR@i?3w*xp&?CFq#{u=4-`^ED=L$NImJN<s#sQv>k#
zUz~@JLNu$|syaO%3-@%XB9W*`aVg{8FO9c{^Zn<`p>TK5Xj3_8>h$_u^gw!!A8HOZ
ztR4-*3`L{&59{|$<kN+{Q&j%`4yCuox6S=_KR=a0L-;^mkT$6MFAG6~=s~nl-;z=L
zIQ+;yTRZRBL)bWX90CP_4Ipl)`25Ak-M*iHk)WXd9;<e)t^Yx0BN_%BWldV8l|Nbn
zQ5>AB5~vPh{ciX7?}#!P@D9FOW$hm!mjV6YId_uf@@MdWjINb#@8DSGFT^0eUOYuI
zT$|lJkk8M*K2!Q^v1&aJay*eU;Pvyv%G>MV{dr=D1hS@_`g}g05%dD)rZoCE#_x{v
ze|38KeefdNHugX!(y@zO>Gg+LqIN9CYaU2;DHi*4r*!g>(nG`N_E{eRC91w2PvcE3
z-XAKPjChcB;+iAl9apt{_W~(rJPHV1i#A$)Ah{0TspMCfh=)0(=dK!}y_ULR$313J
zLKC6*b!S@Lr#M+_!{KqnS6Jj_#G-U3;l_|`wcaJ`4BLy!x+?O&Z77Chv`^PN)95#o
z5~Y~v$#VBC&ci16&_&v~v#$mn64in0mykXNq*urkTdAzH3Q=6)n?ifct%Gf<?K#M|
zWWz>~1Q;$ODjNV~rAdhSOkai>MVb0>)&lPsp^58V8fPS*;QTV2My90PAtelO6ns8V
zAE_rpR1cwsKd#mwk0=NwceGW0D=mMU#o=s|x|t<!S~V=!fdtU4o{dn<<<c_LyO~Tb
z3|l(3dTWiOvJ*cHc1zol8Gqn<>=Rn=`eR&(M_>)n?U!EZ+ANxP5?PvK-niHAX|)Oi
zA^jZ+mAJteATJuUSiXw_dy<_YZ}TDhratrEPRrBG2f;=@clXRC&_l&}6-3cpQ~PFK
z{Feq2eqd2<i?@@g$q+nYmpE;#JYINxf3;i0xXfT;P}-NmiWV9ABp$(apqV^l=g|DK
zzu@vqXV;!nHm|R_<q|`=PdHJciyMsMubs+@dQ9=Aiy#%O8-o*U_vo#H;vAoic9VLR
z87|Z8Co?S1d?$CzudylYu6;)Z55MZb?3{p)-=(SnM0Woug0F(+d77>qA?cvJZ#1xV
z1cSCQ3cHw>T03gH9-byhTpYnX&TJ9GP|l)!oChALVNBRLoNbreX!_b(ZNf{ZV+tKk
ztTK_cSch8$PoHTxx)Q$kcauK!eUuFzPot<=kJQ}ddZ#U~k0Bh!k&bYNuAX6})_Rv?
zl>@0Fu|Cs2c+48^w}yoSF55W?J3|kY1gY6Wm9CcjQzBvqtXxOg2lz;Aj98ROmXb?P
z>#}RYqDIq_QIDt1J~KrkQE2l%yg%SgU=H$)^b@#LFb#j{A!4r{zXYrhFtWgO^sG#I
zHRf-{?}8`!Iv{yApFW$(XVj8NGa9tmH2~o{Ex*TR+BpmZeTOF;iYVLp3bPUxUpj(q
z4vZ5c>V!pMXGl`my9<f~FEgK69G#@;T=LexdtnlPVt{<LRKaW-$4p6~TZfMw#d=NG
z;6{S2$^+<k*IL&K5~}Bj&2x6qTuUwWuqj`W|8ak(7H;!H5;b$zybrolm#3++(LVCD
zjVhrRBt@EI``SQh%(rJcg+9wvz{iCgm1==lK69mR%nD2#DyTV2W*2n;my-b<WHs;6
zOeyCefJrWTfF<0BxZ-T}OS^BknVe}Xkm=D9@I<;$RT_rnHPH?Uj>=C_*q??@JuPxl
z@K>GO-ayE*s<g~}AMx29C-DE`C`Cm@4PU`6A(>CN=7cKz7;Q1<P3{n3UvXg%>QrJZ
zfMZr6jJa!4k>6a%uP0;_swRn|ozHIKl>K%#l_`<1|C=G=-KMy=K;VaETosj~45ck_
zTw;ae;2fn5LSWF0R&waY*S49zXTrS6S3<JE=uCOY9r*|CCqGgc5?rhoK|Z#IGBhKg
zI-u-FLxGZyR*_jjJk{OneHST9R6)Wx237$N<O64EH5~r;{r;PN_lFPz?#&(}&;VwH
zq+8psRNNNB9qn*p_Uj6x!eNVmp+mh;Ch#c*K9C%q26i_MJO`pndfnv{Q%jK^cU5*5
z0uWb1mE0oB0ydGTb-{;!Q%p-Wm*_zy4KWWWm4}Vzzyk{vZ>Ze2td8SCe{$&cc;7j(
zTpU=LP({;jWjHd~2m+;6qEeF6<ugq!9#>}VZ|i8>^i~+(4(7kfCWM`WuP>z%s(fj|
zL0VQKib4U|H~t!X3qFR-lcaP>Cj4$FA+aElD4Hd*_2=h>8vZjdn2&Wga>`hu9#qLo
zMQ;CQQh?g%I~Ir0y?p0LMn5mrZ-T4}TLOXD-3`Uk%&Dw45LUzyGC$ocGI4oq;_|fk
zU_iD>4Hsb+qY}FjKngI~5NT4ZfE2D}%t594mNr5#9#)~-!AEIb_(VUv7(~%ZxZQ4H
z%jH4U%XOfm<>BafWYNn7bCgT#Sj)xoZ6lKIUvtO9=U(zL<@*&6{JYDl)S-Y7>;%yt
zwc?}q$xOvO(bu_ntCGj0W6}y$*2KnCgJ_JrP*16OY=ekt-5L_x@lMg^$&;Fg(sjuK
zj2pFMc?n`MA@KmK_AMgBN{->(A*iD<5ke#4REK%1bDNbwUQ6S_kvZkKE3^#5S3|Jp
z0Lq+2(3qtvYoSyT$)yd>PUGDj;cZ~+FbTk!nC=}!8=$J1<;4})$oDMw+B}<)E;aSc
z?#sUPUiDFdcvy%HXk?JgkRh4Uk0HzGCJL<*b9nT{2^t(_WyRTa%0dVCO*NUT4Aeyt
zb9ewJ&3Ty4^kNfzLTS_C=xCGh6fv(^l6(vcnNXSGl}90=F-O_5w|r)eF6INNzK^=;
z9y%$Wn|hakwR$hg{C+Xb4yM?}h#)`O+VF(m-ufe?9`u-M3p)iHDPENmge6$w;`T5~
z_vp)I?-4iCSEbXZm)^?egEQRW;ZI-GzLaDC;Dca57eA|wZskfSdoQPGgC&`4c0CIY
zeQ&+WJCoM7B+d0pW%s8RO+uRE0@>L!hx{JAGKf9s5}SVYx_K8d9PKP<rS~lK-az9d
z=vTWz)Hy$=9?iAQc0y<g6OuFl_!h-o6+&D90gen$@M`Y{{{Uh{&Ued4CXL!s%;iQM
zC_y^m;Utu~O{jv`)$3rscx>-G026pv@DPSOkZsW*S?>8TN_sVrCdKWU4O-6a(Jm!s
z79zyz;~LEBHZz6*>}{jjUMfp_3Y^rPayo^R$&dv22aDwXGy!)wK1qcs3D&{#eJ1$a
z!5OwL?*l340-@k8m<qO6Fg2z`pZrU@G5`0`CCu!j{+z(TQ~;5qMT*ALmPMlLlo(Q#
z0N-q3f3iOqe_zR64?VS$cJ3eRQGUya8#0g7Bw$w)iW}eSKnNZ@S6@sQopPMuOr>2i
zwPwgts-dJ12cv1JvV$I~*QWH1bSmB5#WLFOTc@PQO|CW%t7Z~h-TG^tGP?EFrx-p8
zV3#5Cl*c96$RBLKYCSsQKe=pNx@));q$KJc0AGdOAguc%9D6sHVqI#jnmgcHL!o7j
zkgSL4XL#lV9Ti~<$fXnp-Jp8$AQ5I(rJ+n6yK9@*PKabJ5aoq^W51NIWf|kP9xTa)
zwLsOqJ5&Q%P%LINT#!Sg5=)zdbPB0?I<65ad`o*0i|gzBn%<-3jNrf|_JfxHRQ4(y
zNWE)guYGW5q2Ntu9{|#EvvVB@hC0f0T3Ms{D*hkF-hoH7s9V=8+qP}nwyj+@cG<RV
z+qP|6yKLLKb#8ZVclw-kzGNk9{eYP@<{Wdp&*P+C4AtWExUAlHiHB^3ZQs9#UTZQ!
zH%-&@OkCUq=Sa%S<U!1%%JlDVX4wFOpEfYX2|_7VXMm};*YQlr50rQOUrXob9q7OP
zS|xxZ>9H9hVoN|3v=O3?>aul<PtQ1OlSw@oQCGDUxU9j)F@FvmjJt8`6hS1zt`dm%
zqweI^qt%zpsK~Mi0Jo^U?96D3;kJ`_8ZM1po#=*F0R5*!aeMQht5n)nOS#v#Nwbvm
zC`B`*2iuP^W;UH++mCdqZbq0ukFl;%*Xa*s^j<PhMkIc86#Aa}O3HeZqS#xnVn;C@
z%u4C*XTmQ$d*j2G*PmEZiEsYSi*b%1*1-Og)tvVX?36tK*bVStW8jumse#${N4-f0
zbx%o7J$`0>bd9gohYn=`{${@~t{TCdX*j|X$&40dO+IpbuHi5hU3~t8D+HhP%iSqS
zuG~ir#^0B8)CY<`W>_2v5ANIE`{rAQk^`B&w5*?#9Ui#88I(<=UTc;q$if=DO+7f$
zZ~BLDN*HI|4FsbUkU9d*s``CA_C#gguH2B_5J&mIINkn9!?-1Bc5sa&gQVA6U)SC_
z({M~<3_5H#^0ja<5uK^v5vi$1#k*B0NiDZcXy{0zuk#B`!Qq3gsmT%fO{7Aj<cIW1
zxP83e`u^Q=7XBa4QNJmY|1;eCe-AG(vi*mw|10h3nE$Qtp?m}~;5fX6r+}$)T?O>G
zFtihu7fn!xoBqJ(Dv}t-k)&QPNEF=&RBI7|JvqtTyKPi+`NO^L&6TAGk3WmWP2L4q
z+ObNHI|nQ2E4UzP#RX1K6>|Ng{c>WbL?t~M4)+W7<-+lPc48Za$Kr_wzyfqa-F@vR
zoI{mWXNgt}nMr^X&DHI3nVHr0*p?om@HYmA4GMMR`+j==5Ht-ghG6rFBLBX6d%B%9
zQ21kEV124Ujfl|E+l#vo3qo84oh?LOw7B&$yOlbKuXD+eUS#&pFXpn->vB0E{j48%
zDEN!*O9oFy9eJJEAaLqh_C&}nVlie+IC_-;*R0HIxIb7QtP(K_Uh$VEX_XS@6TPP-
z8}DZESDd{kFaM)U_yj(V_ly2FCZ+BUnHPk*ii^cF{BV2vl(Gceunz}le4rjEfwN@d
zU$@nHwtT+6<=t*vy$B9MhZ+&<W(do>%KCHuI`?u?1#Q3rES0X8dOk5(D(XdP2C@;3
zt}@14hNnFbjKVN6E{n0a01=!aU5t{FaIcp@pMTnugIC!u=pKgh64I_H%~E>d*k5pa
z+E!Go^Eczu`0W(89bg+uL1!1}Fhs>TW(cO$Ee?y!m?3Sx-n^mE8()hVSozH@3bo5*
zC>aq<XfljB{P{U-HDQHFFV+^}PHR_dD1Dt~-~7WID=fH{9+>hy4Vw6IwIp61QNF*|
z2Uk%qa#!C_N@w1{ly!$L{U3Eeu820@G?(Ng0fWq(soZ!4yxKrd*e~1RNdQ||i~@B+
z)=B-Of$mBCviLR$8P(AI_acVV(v!?3F#YN0RwXFu7u~h;zVei_;!dqPd09&;MnV2_
zcF|aO=uB&04wV#Bj@MlZ1%jd`>(!(I)^)2lyUt(a@JNB`^YAL>Zx;_c;jiFOc`Up$
zz_Z%pxtJHsY+0z+<V{E+zmM<AY2h<A{E!uSX7ukVDo3V_JDOL0;ZL$m&#lKPG)>GL
z`?ckwf;$j&mERLp_=X}h-`(<NGI069lfnHGE>&p0S0f&_8XVdDX+5u&9!_XWS?lDf
z?Pua;MU9TBYyF_)o=P)21#^kx+jmmRenMu;pE>AnDQa}rJ;uRB#iCaqczr0-E0oyd
zw`y;povT6Qzy;J6OG+MAmJyJ>D%`pva2k`&L&wh|%d8{ndXweZFWjWo3#CuF+vij8
z-pwfdMI%a4Hus9t6bGHF2jO&Vm@8=dQd{IH6$!S{2Ox{zE4dfOl@eLQ7hw@YON>?l
z5Ol%-NFtN0f44#9IUJ=EoWc9seZ<X!3oxW&wU+}b;U09X_MGWKSA={!q2qt5*m}?>
zJ;(47gq)<t7#b9j(ZzL9Rp?fr>9%|2;`l+^QlfwoIb71Zvni9oxs}V59MH929#!jY
zZ{Jcos6-3L0NCE?hE3U>Ra?wt*1MNfw@n$#LHAn^?UEn{3lyK{AQir84d2x|uxP>r
zkbguvv=&E%Vo}=#ff&PK&n(H@gKJC+vFDa$_MvpAPkQlnyA^9?)MsChVw_lGHL*J~
zB)gGytFuwfohJX4kzVIjrWc{!Cq!?{>-`#a4@q4QJBuru1+}oP|7#hyE-uR~XboM=
z!xHI1>qK%ef9TFH|4VXlVomH4Qg57}K0CJ|r1nM`UHv+>tXXC2$oOdCJrik!;C}X{
zZRkx|^Q$pMQa`n1>j2c9CaIZQW`6!*>F~(>-%HDt|MyETxaQb{HNW!Z<`x2XM1EeA
z$5?#Dlhl!+$B!glS}Lid+n-%LzdF-&nk3O~cgB79lI%JQ``Im52~4TUSd_}0Ka5+Q
z5F+$8cXuZFWx%~T9HZIwcqX~I$!%ADZw9ph!u*GP_JpCJ-23SpopctJf^e&LI-432
z2w)r16^;%9g7yrS>@L*t#QGccIxE^|=O3qX3!>{x(e>_3zehgH@5`d>*(u$Hb4gMR
zUNZS4r}KAj$L^ipJ$LcmP?R>ndH5+`|NO9-`pL#~MCD9?8qTB=uN=F)1l?XwpsR*4
zC0dnRK6~InY>$JKzaf2?5AEmGd4hkjm1*PK;Ca>y$fC7ZCSHBz%Q4#e+%GEMMx6F~
z?~}E7YN#hRo7@|~fg)jl#2-L`2%zo*Y;f}EIVu!B!SS>-s@0tGuxeKsQ;fJz50oyd
zxQfns09s(c*0)&nfQ6SQN2k!kRH$n?Imz@Bm)Z}sk)Ip`UC^WP$`<TL^hv+|gZDd`
zVQcK@<X~c8^Iu>86<%RvX8kQ||KD22%=jPb>bkWjW3<~5w_eo$6-VcB8sA03U|Cdl
z!Wx^6c9KN;BXUD?kC8+gA741Wc9GmjTUr?A(S{RMY7Bd1QR&^CT@01JxZdz>ecci*
z8>cT0EIgh~lH7uh{sb7!osL8GN1-g1-Yn+&9%jLOZ|}+5(~*ADY;$*=hk8=tm_ke;
zV4&{rZBzLph9ib^7TrzUiM!c7tV!$H)ot4s1A)$=rBUDweuVb=ydU7sV%KyMk2l>V
z)<+n?JG-9m=_#%Sk9bt$hYfh%D(_bJN&BZBJEP6+7_y&}9CA-EJ7pvG_;*PazkGu3
znB(@iPwfP~5ly}Ps`jiF{=BNKr(~3nAQE5U$YxP~a+V)G(vP*_e7)kC9S&_^#;3|a
z@LJ}dzoN()Bq90Y<;rL4wN?pG%@i>chiXE{mrUM+o%P9pr$3zZ$$wOZrjETxyp)_b
zoNgzx##4cMzu*Lp5SKFOHS3Cj%%0pQm}R7xI+8(OFb#3Y5liE?50(VTin}++f@mqN
zF;q-+j?I#O$Dc@V&n{XckN@F{a1ZVHb^1mmM4M~~D)-dbH6@A7EGj+D5j;JV2Jd2p
zfG&P~eqEnHvfcYcgaBnGZ=ap<D`TMb3q&)b)ha>m1%7RR-Okqj`M!N4@2}h4F#yd)
zvDR{9_m;<R_Lr+}UTpnbXC#dVa6Co?TQU}q_mH80zUVtSgNtIwE=zx=jgm6Og{jq~
zVPTew&%$I8Hi3^f#H9V~X$~||XO5&$-V%F7DLZW5BVI{W-7|W%;!iEtczQXzo^*Q2
zKVOvo9IUMWS5=NI6D-PLtD+TBWHhkZ;3>nvF}mC@H^n0sw?a`SfbgJ@n@d$mWr1l?
zL@P|7N|BU*&lO`Q!4Hw~7;mk_94(6tCKw<ZQMFLC^7eId<G!v=<FV64#`GfOK_+mv
zO;BQiXEFR4N~}BS>+X@RYj|-G7-cViq(N?)s}#!_K{7+y((=uUFpS10GmTbQ_R?5|
zKcfl{H?HL#;xQt!Q5I}P)?Q_~?guIgWFTBoVM!@?ngrA)BPCP^pI?m$38<72i6pM6
z2xJ0dQ<F|QLDMa<=6ogBFwVTekn!wc+Aw~9NtOu`ngqL-F30Ib-b32esm9E?3+}kV
z1&LsmbLockd8eN&a=Lg>K_VW^YMDewKCWsbW}#%rMJw0;6yB^M-_ksg-Czn_J$kHS
zv|klobCi)t>RT>3kJ^`fwBsi;0HC&R5B``$A#zHBY+Z3ezeq``wK-VRG7daf9^59H
z4#}JqPSr6**VwRFHbd>LXp+BjUq*3gf>>|sFDiVj07&`Vu#FH&b(OvxSvW~yO207^
zZ$r8<lg})==_~V91rvU4CexJ43gy>;<=&(w>tHXWiA;NZQNNv|sY(a9Vcrfgu=lTh
zT_At{J>crEC*}KLXnxWr^sSi|qJ@$a3M0+ED2JSi*@Yb3Y-0?JY>vCIpStdmP*mHw
z)8p~vW)sMmoipTa0$$1|Xw>#K(*gz3*i+?q=d6UbX*gPxh(VFMI_5A?$fer+_jXm(
z$QJF};wJBALX6=S|C*YsZBwOt7E_{-UV-~C?BB93g$Jp>WQU>|@NL_G!E~&?@Cxt8
zt{<PqAoXASe_Uo|0R1zAMel&y^+O%>A1?T?QCNn)NK5izsBre^7-3By?Ny>m5FHs_
zjHF4}|5)<r?SvfHDVT_5Z%cn2zi=QM!IP|##Fadj87y>~=9xge2V&mp?lrTF@ih^B
zYMlrG4^scxQw3DXKPTULVN<e}Eu5dUekp&|0rnvlaAIzPu}|unlptK(3ICji(?OSq
zPTpQc=dK|xu2u8EvYN|&Xg9J)5%cWsezV}m0fKWg`}p;TZNPVrFjr5vol}eu#}tHI
zp?x*Xjsmi|pQm1iqC?h=MW3dAcFYc2i3qgEiCum$qRW8go<%41R+{(VFxchom%n{3
z35_8EHnf(Cd^6?UMYoyIK3KYkfZxp%I=MI6(qIz_x8^k>ye--K!2H>dx)t?H0p1F%
zHir~$HAN&+I_gsIT-qf<U94y!b=CQzK?s*Zh&%=pM^AHk3FIjjD!`{m^X}0aO#RnZ
z5&wgB6b7KN@6~TYBYs+ZCX@yyR7v%vs3|l@3?El;l9Ycwt@WieFb6m6J|)pvNxZm5
zEt1I71b#Kf>IUPk%|mt!tSNHL8hMp2Cvaj)IO9Job(C)JA-<S-2_sag5Sc9h@d;x(
z865h{bKZ>pC*}K)4DL?F(1+q&Y)q>SzhY$~2aGSH=-R##5e?9Zl#PC*O`8-erGCT<
z%z-1Vo-HVD=|N;%v6grzJ7pibl{zs9c|5MOLLz}_SqW|2MiCvOjG6ECw${w=aG+)@
zHeI5#LlcHigof-c&0S;|!pUG&#7x9XRa`w~cEn$#mc}82mfwuOJ1_C1f-;yp*$K=J
z5iO>rR53y$!gQSJoS=X!jj{wT*5F*XXUc^k^I3Ui5cN?tcq(xQkV>0c7#BuLqZ130
zIoZ=(=a?sJcQW~)>=~U3Y=353<L<cgGi8&!Z&m2~;9f$to%Z$A<gMAZ-#;CrR)Z-W
z`e&&mDzk9$eqg~FcD5!wo%=`?&9!cFk>=<9sI2qr=+vM|WWB(6*l<rV1dDbf6JRwD
zTVQn;+0C8b%j4VW<zXN!2t0`~{|;m1V44Ib4kaNW*z?Qg*YILclu?bLNx%Ia^ur=w
zvli_&HS6aFN7&9-2E|1P8$s$;5OlJES;5PiJ*kZJ8=?sB=`JdhZOg@_v~5~uB)@F@
zL~q@R_Oi}_@D~p0Sr$lteSaV>11B+sZ9KtnWc1xfDo(?tgxflEXEVf^7>xHlSZS93
z!`DjY)Ze4hHjZ|34UI4fY*uq>BGd8Cc>_w80ShHbb6t!b9`Cc&Ggg4Y08JYtfy+5}
zdCI{+cd5xn#(8hiHhEn>uxiu-hpCy=jo&$%Z}%4+8;bogI;X(LwVgii)t;gfTuQq~
zUr|ech2aD055f$CAaOrzVr2~ng_4P5<J<A{EM*Txi+7n7GIow(Mekm?w1inv=U#X$
z%5<f?SDw?~{lUu%c`pLDzx#!2gW<WE<*DR93R()E_dG&!Vq<I(1xV*84ajFm6AJVQ
zZ~IMeixmkZFZH}EQ*gV6bgnfVjFhavYJ3l4`o&R~hBqHqiGNvAZzi?!a)SQ})t>iW
zWlOa)2>5ikcfq&*O|d=u<Q4q!*tvngZTDe+1IyW&=XUtCy?0*rQ9k<vxXpyXDGl`s
z*l4))Vh6CLP}j#a?cSLn5*3Bv>fCreD%Jl<^`nMjTQ7cv+t!+Zw(Ftspv}=rbbKr>
zZM(O{JN&5euJ~OsMmk<RNSuGvOQFubz-1IKT!5j3>$TH0YR{Ukv#g+r73#Accn=vD
zSxT6#M?k&X4ILkPdv4jQG-fHOS<^nhtf6V)Z9sp|>0S7G@sjW)x1A5rTKyvWp{bLQ
z(tY+>wZrN`(G|~OFD9Pj`)0ACccKYC@gYvQiYNqRey-D?z6pUjcx>k8Dl1}>5f(wF
zbbQCGc4|8;Ei)ml-1WYjY*+yK1|E!*EKKhze4{3J#rTxr%X30?_Tw0!#DwHHXr8}c
zr;d~zaTF3!JBv&LLhL(P;9GP`J;HgE>EsNYg{&1Gg7&c!WCa(m8q?^TIOgC@Kh7y;
zJTk4fDLsK;ou>u!&bp126O-vHxy<{14JNE>cV<`t!_NX3<w|B5rhMG}JiSfoe#S6r
z)_=^KeoW}8TZ`-IZPPUTPzyF#*t+fdD$I)dNg0eQIj(j32`=VphYsy1U{E}a<(k}{
zl8I(n*2-S@mH0v2;ae874*1FQq&lz?)%M%S8zwYBu&7RbxhvkDk@z5JqT*+YdkTg4
zMdv`s=O$p)62^JroL_%whwZU$b65urUXtddq199(OhP)HRqdB8|A}uxd|-P{b;yq5
z2bty%puKT+|55OL*eL3ep6JN~&IiW@c_qkk$z=Wtu<{Vzi#HWLfMbK2y}hKI@f`Q2
zF)u3`Jvo9GMv0m_ABV*0z47@g`B3;k+L)yMt<=xfbDv?o#*8YT8A0dD^A%tWTDh+=
zw5yo5GIor=(Iw~Sb+X*BzV2hu+q9UtULzU=dm?K6ieIOaTJb*}L)aMpyIzxtll4E#
z3r%Zt#!|B*&7AyNvw1SaQQ`Fo^u_Qn|C$;#42wZ;#s@zAz_+-eWQx017MHTG2{<|>
z4s-FXh)U&rbB!<3)8FLM^XtN9W!;u#@p*r{D?JJ=$djJs4<{;1X%x!k`sRK=Q5Cm5
zx|hOBSipGymxY8qrx)|cBB%f{B8STP`&eKeSy&z>dS0TJ)SKn&YNF99-Oi#X6(AmH
zUX4YedmeoD=lk<X6fNYp>GqB38~x+$)${pM+)x`WfGwv6n%|%G{qQ2~1qc=F{HOsX
ze*gV2dN!N7BUVxE&_#44VTH(j1fb~Z*1!gV{Kw>o01}D>rsLJsR0lw|od5gnf`#Yp
zD>c|`A1V;Dydw9YRC$nz7=rYPgP8JPF(&fpX)qz<q(m5w@Y|POn{Uv=fVjMOup{Ld
zh0>UWuu;hdFJT40ZlKO`hvNVhYmK4MG516qi~Q{X9-Fqy(gK9ZKM6$s2r#@8)x~E<
zg9V9yY-Oj8oFlm@LVL!n)T=hoasZNpEzwp6^RQ<ibnZ5P-|Y08Ha>GgjycJE7kFIR
z(Wqv5+2uz9R`fKW52$tmK|%5dBoO!;cBCd8-L3A>c9;U|fwF2%A)?O@)H>0enLTq_
zcfL01Bn)t(sQElMLFv2L52zHW!+#*EFY2_2BAMvp)Gl(Gn9QjWzQ!$4B;L)c-N7z$
zern0K!VE|u^5B8;9OUY$hSQruBmw5%o7Z7XY^JQx`=*wvQ3W%?J8b@)JFkb>Uy#!@
zxrHGz7{e6(KwB?*ddKu5s~Q;Uir5GSF@!_RmD@>!Ttzb&e<=nn>o(B>d8I;(Py=eo
z60J4VO{hsm>CayT(<Tipv-cVIKC*n9B|yEpJ6p>T#(4hnOP!G?iszlTwUJREiKz^V
zdYnI7UVx%Fq0|KWvx?i{sHDsCvPRd$)NUub965TLB&Hm&5;>xXg}?0f0?TbO0-MJj
zq0v<vc<?<>G`GnZ<{_+Ec<t-J*@Wpf-PtWlCbk2uguy=~?kOLgsN=-a-&w)=-AQMZ
zO*bxPZ1Q5*7Y#u_RH#`5ErsGqYt28!0wghCDo~kYUaHbg>QR-qedZ#x!1kIjIl1#%
zb=k%GSJ(Z|t=lz}{n#kuxhjiHT1i+6LmI)ai*+||B3(61TRT8?h*&#Nz3&poxxNK2
zNXF?-?1nw>vy9M)4kHD){*BJ$u*Ox80KfaOsxeAqCRJ~4GAyQ|H7K5;;3oKud-X6U
zo7}1_H%Qh}x|%K<mVT3Y1zs%K-GFjJ)->aav=b9uc~GLdDu{b7*{C9aq!CT@_=-*x
zJr?=Wwb9S(x1XApt>ztKsarj8O!w6WlxWNh8%_k9?EPij&H1-W+23NhYX>dD)rpx>
z06QE6RW0^Xo!rB0LBSkyl}=fDSA$p`1ZyMHYAbetSKJ+~x6)B3^B$vb>DNLL<*`wI
z!oDYgNwUn1o9{d;z1O@oCmw{~VuR^P+vcy<|8zxs?w1E@vTb$lrev$=UydVM;fZVm
zs?w=A%{24fQ&!HSG!DhPyzT;d38&?`wn-Pq0*NLTgYQg!p?aI{dM8EL6|PQz49;GD
zzW}v?s=sOiUkV}=k|q$;=ui4q;J)=6t}*P`ywav9p{ht)_0~sf)ql*uI!Pd@WUG+k
zCtp&TlR{C^aw_lhEPUcSExRAtLN%+8oYH-)32E)7Q$Am|?tcn`#ZA=*6{~x!aS+0|
zGN)2!|13chjD^Bf)O|6zgwxsZnT~!BJck1%exeG~4KmniMf~=o01a10I@(0B0F#g`
zQw!KqGhEJef%EK!++?9r8)XcpzFEv6-zgmNXn$71(6b8;mA-?+uC}@7U)e;~Ag-$R
z&Ua$VTQLKGTG-?m>`K0d^DKG6*3CBWjF%|I@Mf7^qj!%&B8-9_zYueNZs91|C((Rx
zv6reQEq$rD`$_E=9clK-vEIH4U_<@mpp+HguKnWT)?jnwRHnA&+^h#}M;^@=hm@}*
zl(>jjU-}#`<zB7*zU78pwiVQAZ)sHNLXhzOjFaD1yMuuz>`3=lb7B*VnB1D4@csL?
zFc3XY91nS@-jS-Tg&@WOcAtaQ?*12qjIQFnp>`}c{J~U1yM3`H+a`oT4YEEoX;t43
zGysW8#zk-$ILe1ocDso9Iym*=U)zO_btOTmdE6ke%qHAdv6|dS9P?^tKNL6ikflp=
z-_@;LiSfm|k3cZV!`<W_Lgu@#^3sGJv)tA|ud@Bo`T;kd5c={B=^{ys8(O(mm;5VQ
z*gC%z^H-V^hYV}w--WQ}hM91(T%qq$vehf8=w`V`BKDTONny8E&S7_klF`m^_`7*Y
z{|~~}iuCfdyJR3`w3`*L?6?cq?8D7bNX&J!TnRtyM|={NDr4!Q2F#@!+9+zkDo#Q0
zIEASD3_(s6A`qI-;b^4@hoN{_cVeN*%Qdrgxwc-uRd>?G($&7>_)k05xSC3Zcc3zO
zlf2(QD*g+LEkqOw+N_MGF6WY$!wBSJH1nypcd>M2XtGU}VY`zobP;!6d=`7yS!c6J
z&24h~Cf!A+f_ZUzJnhN6&p%H$N6|EJmmN##YMl;Ukv{hJl4dOjm{k0Zfk~?u9M+1N
z*yR5(Yl#N}CRmi%Dz6Fz6XXH7ExFDRA12;9`U<PF4-u{}L5x^>N5U!9sV3CwAerDG
zx20{OWr6n4$TL(=*rIpU{P%3BlZXOuH=?|}6op5}+_3EygaS$xi>rfWwIfaTZB$I`
zaOJ!cEFBix_I+qjTrW2rX}>lUEw)_rJQs!5B?E%O@qNvxdt1%ooltpI<~bi}J(MOW
zmZNRb#9@BPon<^w)TO9Dy-+Mrd_+^0l7FSL<-fLdhU=@UD05Ta^`BIaub1S9QNSB6
z+!QC}(7XxSU2+mGFtLGqBIAVf$^uf5V*(PtItf*@q3E4gV?b>~#aJqzcNZote4Wgw
zoolzbY$uZ7%G8w81+^`sXK`4bnLQv#it)|J6(}2BNG_KzUjN*$;a98JAh*vCh)CBg
zoWsTP;&T_;bvaX_fOW9XWE+-_4N_bLejPwq7}+m$+IBJ7{!?osOw_TfyYyfXtArc~
z>_i5Il3;!`Pup@P&eQ;$b15WbSKJu$wwP(25QI}5sO20e(f%WR5NuI^x4}4OWF@04
z%QE3%4HY8WL4D_d0iA6vOl=YG#?BREe&{ZqWs~AiO$%Tns7gmH@3LyVlBYgPrHgH8
zxCpKNm`-X?_3RQ|3N#s=R)=$jj+MHg3&!oi+gC`-SH>Pl_b+LXksa`Z*tp|LTlQ8+
z$(_yoWw!<hl!&64OPhb}VQpW=)o9eHqo{bUh4XI4D@?Ii;dgfYx$tCMS^Z#2(AdKI
zpE-|3ejp~Mk~QB1h|5q|wT<S3o%ci8{TyTUvl2zMy4;uT7Hyd75<@5ND38mJ0VU{S
ztEcVbl@u&{)rRMAjkD6^OARO!%Em$b?aG<Hbr2VaMbS?4>Jy!M?YkXY!TY+XrJ<h_
z-6*}ewMNWa9tC=rXHu-~vlrGCI&ZxXacp-p(X68T%U!jg#0~pv%nr=0-slbP+$c87
z8{N<2;|nm83pH=gO=;koqbz#6I7uaPO#qH+=%divE$aEHnuiANRXHk+ZQL@z+AHyQ
zjTnXRL-k7Q5%nL}3hCfep=I>4%0FixbB;E)dFvkK^?Flt-Pp6GuCe8#>yNl~^a}E=
z3kT<z^vAU5O*#2+)R~??{3^IxQu1AmDf0~vN7aFUB$P%H^soB`_@}PvXl(RItOr21
z;jWpAHj4+P{^0yeUijtW8d<Wpk9{hb!>8P1KM|mB9e#=mX>>iC{tYGUV}5zDmrGAy
zIToDh9IwoJ=>KaolLRdyzssX1xpE7dQ=omi?(?vFwLKque-k^K`U6<b_kjFA?(LcW
zd!mn#jqyJtHEwB6)*P_GZ0*$E!t1COl%XSlj7{5soz?_)Qg={s=}s3s{6s3zYr}7b
z^GcSa1s`av1o9BYN@T$Qz<9kr-ap^nPm1pAr>PF~-QJlXJO(V*?k<%r>j&v;MeH41
z+>ZJ?&N+3@2WLr3rsn^oT%W&E1sCH5ctBX9?!Q_pC`1;aNqP&+XD@<2?i961W7S%8
z9fkRTWoRpi+ut|h{(Qb(1(W+fcqb`QnZm+t5D;r7B-!aK7h+IY8SKKGH7JemVRxTt
z^ZOF<RHY3AqY7JPB$Ij1=pB?L5^crh+#%f@H2Rhr@jE^4@PGVlyl{Gy3Asv4$LSTD
z6)&o!<wV<M2AkO{GwVz+#?1;;>X@@@e8;rciVuU<B}whx>gv-5fFw{}fzh&d<Cwf8
zcDr*9kW5B6mNeUdwMu~uZy%s`4HikFlQcwNZp&ej-G_}6ptk37iZCQi0~nXWjNLE2
zB!KLl<?JHfpz>r-Ll{fcO<_-~3YUqI7fS*NCFR;}rjUd>#2|!ZJ1iA)N5XKN8Oym0
z&=_4ERkPc6exjWfZ&3TPcBK9YGGWh{Uw1R97kZ{o8|N1Co@YSc<OeFjC|oY*V|81y
zPl#@eM}?RQeox-1I?5X#WSXj_5wgg87(725w#R*glkf&`cSX2Cmru$$Rm}5pm4?5&
z5ZR|QJU2>&iJw+Oj7LxpDk2aer*~`y<Sxcl0|XdiP3Zwt&cg#JL^#p7kxyb(UJ1Or
zYR4m@&|e;3INbis)lscJ%;3RNV=sx7%~^=$-yj1udK7nxno>sD=-9}IMW`|CzGEt=
zP;b#o@QPDV>1~4tugGw{z)(~qKa;bkInAgkas;n~`y<J4x`BZ#OMO%mo{*J(;R($5
z)SqK=sZ)Ea%VoARf(h7f63&wk5s#HxthoUdqAw(jDOpsW)!>Gf&0S{Biuxohus&jf
z;Ts_om!bmOUog66vi04)fgq$E*~YCfnVF;y9ic2%;x0L-KSOhzJ2@hLs_4gyF(@P2
zk7+7*oV^f&KPY5zzkQD2Sd%@Uy;xOH|JyAu^hu@EQ7~lX1eKD@Up{b}=OhGmRk@#2
z<37LD8zSX!+_;JE^|s9M`Eq+}-E@=;fkq4M?df?%r=W;nvd#N)1PZnDh5w`HV`pa5
z6_;9ESEsCFm8+lpoR_ZT5(m+Ao(B9cL>ES8<;v7>G@T+c$|KDuKw5=1m>NJs%C;Vr
zWrVMBoL6EH<VuT~NF<U$bMJJywB5U-K8L8?WJT6Wp)5~ViDz!YWb03n$<o49!z3a4
zp;HLwrCQh$Yl^zi$lstYqX6{rZPM`qXnZ_qX7dQOs=^7G$i4Ly>azL7{W6^eManJs
z)bW?QD~e`b6e~f_mojBKj*WVqCjE)RT-(x%VH?>Mz?s$bEwFdna1@xhmka#yqDV1@
z>;U(6fU^!kshGp8HHv2v%cm&-C7y6CKnqn{D$6aHW1Qs?UTxA~CfT8MDX3#4W-;K1
z?!7tmJ68p^H|&EXFBX6*g_g%S$mJL}2oqpe6p@UtB-!Ydmc@U`jg$ZkTTsLh7?HaK
zBI}mk&^S`qM8j}}%E#Y#%EIfdSX;h>l)84Pm>aZ)V|eM7BhC(SO{opapGm=0;XZJF
za|sf$?>(j;@H3_Nt^xlN3zAl;a(U%dI0aLVEmvu~$WuY*qWW^6)V#HWqsa^@xT<SC
zuR0J(tzH2AvLj?QYm86d5Y(0h(;=bE;P><zJBw^)<{2xs-MKYW<g={4;ZONl)?I>~
zQk>c}xsH#fLK|S#efSYW5}+f<&q>+Z!pR|I1$BQri1F5@R?c@%4B`?e<=#_4JDWDM
zPs;?y)<de(_Ge=r@vLF@XG7NWZN7*8hkp_(O$mQQ@dZm}T^XOdG8GF!4DZ#N?;Zaw
zf3=M_&67XLy;Sg0)uC!JSzNk;v;f6x$baSawNT&Dc}YUtC_Ulie+&3_;9VeTWknUQ
zJW1Bq<VDTSe{t5=Abcwq(7_sV_t3WSJx6c0QMOST?I={PDLN5Eh{0BCS&4#Er!yDe
z-o-(zK$S#f{rxkYpK}OE8tUPn+qPfl(ZQKJ=cFG-O^#ZMt8|suWO#-TU*r}oGmtXj
z=t-{j4)0_L6!<^ZCYJwh=VIji4`m6{zp?~uc7&b}%4g82zhis&`VetR90GQ<Ueq2`
z8CeeI+dp%JoQHJd%QeQ6a)7laj*>G!?94@nkKPH3tD%dNgQaHCiOAc2%8Qe&kf*^#
zI|LSU8RLKul?EPfj9+gg#SM>Hr-GEtC8P5@IegQ5ih-q=0?dA5fN-&J{$ecA<{5=B
z-clYYUsrCI>8sjWzhbU`{?A|;5Tk92d$p-McqGi9XX}G#tt<HiIZ&<fSPQlxWqXv$
zY+OIJGmns;F&qp-xFJ0FHh92eIwhPe__^?kd9P-;5w$is5hPKHU7}vK8&_wc$7KF=
z<+|>~O9W$goE6|YECL)DvbiH>&%~9%sS?~V@#l#PB6W&(cL!E4JE^ImzW5c-NoXS>
z<CfG}j*C0|c>B=*HBt6oRwp_BpY$Hy&+VIukTd>7Gg56yt7r@B;wpv_d9>y@u*m8f
zWxOx0`Z=mzX*Ou<DDzgwmiXN1%oJXHyy~hIU@|1L>@*E~G+oom(yk|^)rZU>0A%({
zhWHGg#^?CM>Run8+>Utu9yMTd8<1d?TBKK;Trz-js0@pyYjtyddi~Ja2E0|C0<dLJ
z2nt)m>RUb)h&+P7^40Y+PqR|%=mKfcV;9EE{1)*>`$B(11$-cQ@vBW*3Z?wxW*)go
zb`&z3lEM7N;G5WS>t1@Y!(M>?<sF@xb(~kens}kH=~}iea{_AuH8XFH9};AmvBv++
zF`(BbzjD;c+p<trE(61SCbm&#J{7<i@=+IS-Ij(uW5m!VHdq-ivCBtUdH>ZZvU38n
z2!Suw8q(HpMW`h<4dY)Tl~<;`7zvEWbrqdFgka1mVbqFCrp~7czTT$&MKYVruVhx@
z(!J}Fy#|H?<vj~KydcoZ)AA{2@T%$akQA_iSbdC9ga|^Zo0z@>Ls)34RI^ps`SJr8
zFE_8v86b5n-3_BOa(Xc=t5YRW0hY=!Ni~&KU>pQt`kMq}gRZLT4EF=_Ui1sn?5y?+
zLx~ttVMJUwU?sGciW@w|PqpgNpdyWI%2+#kc2I0dMQA1zNN6c_8y5;ua`aQzHb-v3
zNe<t5VX^fChEUT``GPCtZ=1)D5Hq0Ttei!^66fSY|DiaeaR(OX6dvK_5tJqlXbfbV
zk&)FMd~n9oPyUPKA=SJ~wj8foG&kZQoy#VXe5C2!NT~coL<Q?#ab$=;g00Mj#%<L=
zjVLW|3?Lc!TIo?xR|Xv6Sf=D@BDk<pnoUI+hM+<|1^y&@n#BOeUq&xfTlhJd)+6Wd
z*{#^TIETHpT~C1<jGwl3s)3?JD*Jb<+ATE(bUqxzwAhH!6Rc#}mFR43Z$b;qP}R~C
z>BSnNOaWT(1%=A97(iNPcAYLJZL3NNM2>mV)_<NiHX~qGhx~9}35L$m$HP%LG7lI+
zR=E3m+s8U!Lz!*R=fn-nghY+Kl4tT9f(jo^g?yL2-E9_CR#}jSx~wYQ@yPbC@PI9Q
zZuV2M<bsZ}j(VqkCK%}5Rg1Ke1SlwPx5tW$DaA7leqnbwE9><u5fqiAr{K&aUcKbU
ztmA|)>&A$Sd?=nxzml_E>-;m86DTdfrPaK%`&?pvV*p+kdWMB~K_Y662=mb;t-09P
zH8@sv8h0?sUOqP4=^Pbvvk3Lkzhv`aEAFQKCtR(&Qz)%^EhH^_V_koZQ_vtm-s$g(
zGaYtr6?mcOi5QAt0(6Suu9`fRk7KMP!|1F?x-B>)#$d4U8sYr?8Z=7Zb?yQL4itUw
zUPo6a1MMHG&^}m{tzN!WjDD~kg<LBkvX{vTN4^y46ov=25MEDSKKLgH&=%aBz7?vS
zh*v>T&r6IsUxn!!pkJO$`CMfa*CfyC-4mTrZ<)@veNh7?zWtWYy0UlWA;E%_34UNu
zigfzMQ<oKU!T|Xg>o6sH5b&z<881%k)3yE)*55T%=>~TUw?F^B)=Vr7xQ;aXwtVc2
zk_ZCz@+;o}{PRWxdWB9r3k|!C#<wo{dbT;e#~rJ^gn9F25XhDu3#SQNkw|wQbY;1!
z{F6Pir%Ck7RZ|$|P7?&WkZb`mZ>aX0?-v$fg_A7dTP&aT)S#*@>C~G1xj<Mc41SuH
zXtzR_jNVu@d1;x7^s?yMzbi(j)#Z+8s(7QfyYv9XS$l5Yu7SQBe1it0m?gZoc`Zal
zppF#(yj-|S0wMA|qv%C`?3R1i4I01>XA^hW+2ALYEKIk50D5n7jA-kMn-W<feO@GG
z-`YB>hoQq5IN{QP^PPx)yt%PlmrX~-gT~S;#;KG&>C_t8h%?tYxy*Z{n<&JxtP>%W
z<qt}+7M5v=&U5dPr;UJbgYtV~@%re^t&Z-Qjqd&c^1E`Q6rIzDsE&GKLvXt~Y2_^h
zKC17UcQO8T(Quc?fgMm@Tu^iy-3E8cyFa1rhWA64`zqKLPwk{z^%W#Sr#PEL<Jd7T
zv3K_?O`p8~ibw<AS3lTiYPKPZ7LcWBDxmiux`NYRP&`c6ylq76oC!kO^#y7|J^r6c
zH!Ji1>sa~!gJ5R*e-O;<oc|d`Mr+<yn-k9aMjc`;bv^V{DoI2tQqX_KiJ4mfOq&H)
z-!(Ar_qUs|z67V^8u$H$cfY|dc{Rlqj?6-vu>e;)Bpn|bnqWvJsq>e1`P9JyF}JX?
zN~k7ANV&XDdcFup$8x^tl*24>EL<g#K|rEk<dcf2TcN%!t=ty^IMq-dwQSj(8SS$M
zEg(&Upv7=!A3Dv1Xi2LX@s<JfT$}>o$v?@+3B{;x9i#wO4NA{%2<%{&!k!-k#eo5|
zU%xmPh+ZH|))<D&(60YbiD2;v1Ruo_#w@i=Ejv0=I?~mMW1J?K0xt2XW4XWinA`SN
z*A7!`Hh)7K;u9`VmV`#IWi&{4W+8IlA^a%Vqg05dyL;ROj?I49C}C)5-(i|eDp14P
zP)3Cal{}xZXgGBq^?Yk?ELp@ffa<U)F&T@X#Ye+yvOHG?CyIH}%DI@Wv4m|jU){Xl
z-q=xi5+Z)M4YYu_WN(mkt&I##mksVxLnal`QAhbFP2kd|PXuYq)M{GtP`{%el9h>(
zYa9(TYBj;oY8Wu~NxB~p520b5A!bRBP_Wg$HckijSqP`0QskVXE2{F<qM$89o2i~i
z+QYNHL)BU}EueDwJ`fRwR-s{ijJ1LdgLHL5h`n=UdK>7VNs5iGKLO#LkR?&*ym`5u
zFrAv?{DpI~FkYpmTed<u%tlTz-9|~audh3mWmTxz`D96@#mN{K_Quvnz*7A4EH~Ix
zX4!h~F-B(!47Ooy?2$Fn=-_t$<EC>J9EuP`!hQ4WV&|tnZ=bq>MC}=4RF*F{7r)2r
zt+YZ2FWyyl*_mH%wzTmVTzR)f>fPsTC-(5>#%90d`}uM7S^D*6&gWA9W9Z72Zj`W<
zbQNrqqC1Q*V6_SETvY<f*X6a|oh}>uGWrmBJ$<otz4hwQQ+L0W8p%lZVFQCr_HbwX
zJ?>Y9e{l7>`1o+?_Rp96d5iuav2h-LaS}LNc~u+>*Dwe(N40s5|J{-){n>~RZaN4Z
zoVTyyjkk}E6sw3u!aZ(+U|k?IaG5Rm!Y6G;ZiJbc9LMG7)aTk~2;6NS;TeAxo~VB(
zf}GQB1&8(!u@*`IZ_@j<VDra|WfkychV17`4|WiLcD5wZw)f-dYWDr66aL|CxU|=K
z%g*cpJ~#IB?$PMw<0KZHoBV*k<B!+_{7+Zf4-EQ@Tz%B{%ft1lm&X!r=t4|^N{Y_N
zQBtg2Wmbwk2NrWK&SDu(t|e2pQ?BXgR_#OML*v`}tFkL2cjgr1*q`UPer`{{>)wID
z)s!i{?~tPlUc?XhZgLKf3aH0r;x_;LadrTpxozS-xK4Gr!5de8*&W%=)$2|Q#sNnO
z>51~=IRFxd11oB_KR_thcBoeMAe1Wf13Sk^IP&y8Yf#4F+^>Y%8h}FA%o8|N1@cnd
zVZ4R<h^Udn$*5z92*tj_!h|)ZLtkI`+7Frr_Itii!>ylJ9lhL{SyA*mzBsyTLJI5a
zg)PY<5`0$CF^8n(_zC1H{7=dWZ4EHh#UvDrib&=dl#=dqTr5=`D`mq>dH3$0o@^I!
z?q;<^mCkii;12GpXBf#bg1Y-sf?fG5=pr>?LUS>)D)Ic3c10PeONQGLtfsLEz2TNn
z^VG820;oybR@(LcmehKjGsFx5RwTP%ylu?H6jm65?a;(+$NmmSfKA}+g@^WSarbjb
z1B2o5NviA;9r%n19?=+g0-<hs!aQ??c&7;PPGRnG#&PEa&BY|ZM+pXS`QVC_EVbKZ
zZ$NzrxOgn@*tG}dNdDTJ<p+LM0WoBwhs{4Sl%9e&Y5iqH5Ytk)@+ej_8lzN6MTDCV
z8Bh-5)DBU2GQ3Co&CVi&m!P;@1S6~OB5XrTT>z$#OGE{bMPzu<Rf^nAY1C?do*_wb
zQ3@jbXhb&sVJHps=9=YVDo6q32&Lxj<)X7s;nrolC}YY41xq7zz~9<)v>9Ov7Sv)0
zC1?x9VGO-c1E$bo<~VFu)`duzsnWMa-Zn9OL7yyQ#-?^3*xj(8%hi-;y~AKAf=wr{
zknmE~<#TK)z=*bz{xe%vPH6NUPa?3{a@t0~ds~iB07n2Rqa+s4Yy=?IYK?$Bw=L@S
zk+1+Q5q`3mOOTQgF0HnN1SFoDqb$ft$o3suK`_Rob%ZK(D9|jE2H1OXn8u!G!2v16
z=WHuBdb>*L4cD6lBw7*bxUu%&<Jj|4*p|HGRHr}jPuS&_+G=N%kn6B94rW?Pi&V7b
zcRo!;DOOD)8-Za!3d8I@D`y1M%)W&pdxkui*|WV}xzV%Ysl=>9=D>fUZOpzqcFLW&
z!*%B|Jb*;H6Pm;!s<2UB2xWS!!;4*JH7r+OzRv|jE(CP{&!eq@?ClC>aNn^$>G!7(
zY7J=u={}fRsIj@U*z)w<o6aQ9@-|;yFejqYUPwABqUx&#56(XWg0)XqC?%z|A|kwM
zQ%}o(k2in0eL6zCJn}DXgnteVD?Hk%!nG6m!V%O;%Jq2SwO?da&=V9IBRATyW`Duu
z<=tx9nt9>=zZC!@Ku2)S4G~ma@nmvd{xg^6kmX>;PQ%hR`#WsjPfh(}VM`8h&3;|p
z#ZLI^`?@2Cxm+f?JEy+2mLUx|gSdTy`9XX3<EpSy3}FwrtvK~pd|5&BN*e=6r2-pw
zmHiYg>juy+?KA*C!1I@Ci}02=yj2oT9mP1jWTX}5k_f3bR%`?!zCuJ|LG7%MY&RCt
z%D6~$^*CTpV3zP9wN|r0l3H-9$>y+A4Q1`Es<?nA=nWz>pu^A!?MGoms{jaqF^}^p
zf*_k_0Hjx6=zw=+gzcb^4x4E(F-ou{`c$B`OmpW%07e2udY2+CX!bnQw)ybW_;2Q|
zVPmbWsb%CRPeEDVfi2-^Jw^?oF^5qldJGy;%pxucV1#)95f%W%c>-ACMFys1!ToB$
zl=~qAVrcVz)Uagr*`T<5R$_6aLX+iA2zyV!ibIgCL4Bac(xwi_gfwlrUD+`Ov_s1c
zy@bP*M-e1_K$nmOl>i5E&B2*RD0mKB;tCC^?Gd<z1u)~XZFtR%yu&Cpbs{BMS!7fa
z#7Lloj23fXj>hfX0-dOOv-!forV$7Mv{T3qbT>eYgjNd?NWRq`kB#UGC2!5Zf4lLb
zcTCKd<=0SghgX>KeY1%$bI}gRCewTmrgPq*p4g0-qWIS~L&t)ny|N9T_AI&Cx;?o$
zJ^VRxeA8xy%mUX5^sCL{@R^<gTDKmB<{UlXd;tDAndEa=mzTpBt@d#{EQoz2P&odd
zW9EpLP{lcZJcy0s5>khlr>JvU^wa_8XM`eEd0bWs+^Yc0N+Uo1OzbFQ;Liyj<;NAN
z#@))uxw;ApA_`T#TlMYq8aMX0yU;6*l<qDyVVVzx+txKtN+)2`PqNL!Um*xy+TB@g
za;V1W(@$cBK<nP1#hU`E^f}4ctz>xF1Uu=-FRjpa4*TC|`#%RmIs@uiZ9`UB9I!C1
zZJ%R<N8)&N{0_cvcmA2b!lfh~KroSc=uHo}h*<(91KHRIimej75jKk&ni;%3z}Rm{
zG0>*(qxrkKqRYYNX;$7B%HSMasMBtFh1yWSf^AsyBPc{xA_sS$xHQ3BLw5q-g~ig8
znk<7MHeP<SzkbXlzN+}|kqf>K1kSn#=hs_*udaOes{ZKxPuplVw*MXgVPa+cPbaU&
ze|4eAza6l5(5bqyeJ~6lYfdbJR*)W0PEcQmiM`nmd?lUCD!rwIBN;fXach_Q=XZ_6
zv9no8xt*MCz8s%6^WV8Eu&fNe?=(L=iFb4oSS@Bw00UMTd|%n#9*J@q@9on-B}-X+
z@m^l<%*949d1gOTfM}qs99*DibF6t5am@FWN5#j@<a%lMorMoQA~7ftY;&Tnh0$D}
zkJmHF|7++f>d$8_K7Y>l&(fN@76vxiL6pE*$<B{Q1Uy1i9bKFgMb+ct&QIpI+k1Lq
zb!7cb5iNWX?Qg&_uqEj?;5hypaP&%Q76p<PUcgL1lcXvFBnh9-&8d#+U}WQGC7T_3
zboHvxYiL>7nbl(DiY;xm!Av!&QW$3T?lqYlcHM7%W7M7j$?zKQy~|>iJ<mAy{r+||
zB<}UHl3HPcfPpUerx@UF#9kMI!nB}BEQcc&c?y+|v+5viEwD(bEhx}Ya6)E%$>@#7
zRNHPy6C=-Z&8p}^p(PoKnxY}|360l-7K0489LHbmgY{rlE7ONRZSFQubhAL?LCHPq
zrg>9*>pVw@ge%O|$0hdGN>{Vei$PI|QDf9I-><C-=v1+Nrur+V0nh{0RqM8FPwn3d
zghD<M)r63g1=5^3_~v2|7O<K%S3&b44iVRDj{8dG7=#lMM%L5<lx-(8xFr-D)FX9j
zpX?c|CWZkqU2No6mSt^}Fx~IiNkoAoNG10*pHT+LTvH${8O8d#ymweugeDL(M6uF<
z80bmO3ZS%QL!Z`gX;s5Y`J!{JejxB!*~CF}4?!A3#hEAU5r6)J<M-(PQv_7blXsAW
zUn3|xJbJInmoEgN_nIb|CdT#JkS~m!R9Dsxr-+K%(XNx<*(p-5idQ|W)38RAdp8z#
zG*6e*=rm58?3{QN5FA~)%+r!|PA#=2{3sg`tt%P)gh9bL&80cqRciJKX*4?TC^*R%
zBW9hg<aq<fY)HvGMKxe6uf8%A&>vlGWma*O-8Q2wFPbcr=mN#1sU*0ZlC8*VQLV6)
zsUs*Jn=s^<u!X)c#{o5M*2l!TpF=C~FYpYslz4*M-&2#IyTWev&Gu%~_+vkJb7h&%
zea_yi{fr(h=s881B<ZKB$>;2q=0gTb@<tWL?cL4B>G^B;lzykBn4lN@=q)36aG$1U
zwOAD-JWJK<$XxY`x3MDNFuvkt;t=%=@|0bb#c`ZRG<W8vs9ghULB>LIGsWEjBRQB3
zZJq1KAc^6_Mw1n$-B$Xe$k4y8?sdghbGbK?RVgP77iMkKAXkat%C_NUtAUGakGjIf
zF`~-AS*kiIMb60<$~;bo*8(N&ofQXmO{PA`G(RbkkQ0C<HBA`_CTl6l=L~tJ#$82V
zV<RK@@Flcu1*8_HM(N*)D6*Qu$^*NR=fp49(>XF_1`hNbtrkKLWPs6n7$YyusY<Gy
zBIPuGqJ8aEX8=H}(U2JhhqSj&;|c(6V}C~kz+fq)rjk77%vZo{`NsbiXGVJp#7F|1
zOQn`Rygr#AsmtpaF-=;>GA0>CEk_J;kfGI_j*9*)@8Kpra6>7AbYYn-2RFy%F17rj
zNJC7PmcAI|WWIQX8xk4cJ?DmX{=%$r=mfKd1^J6>s~9iyM+@zyA@;-Ir@pS7WCtQJ
zwHZdUsY=h%$Dz^L1rRXV4nO};!V_>y#jeq`mRDXap(V+(pS7>T`t;!2Vp|iweTBA9
zzeO4}-+cs!*he;CAN0s#5~w`?`l1#JS4*}o6IL)Gy00~RPN}H8{VqsWtVL7f94}hn
zh9h{TkN<>7v%APyaz+vLW0_uqFtuwM%fTD1ZHo3mD?1GYI{n}{Z}TW)(*%$5Z+dgC
zqy2vM<a9F}P#s6?N(;H$8J4u;?sW$hDBuF|h{~Y6ZSO7@eO6Ftgwq)gNBS9v?n)Oi
zR{Xg?qpj+Y=yn=!q1#Y#EPiWPZUu#-t^JaX?`d_)<rHP#r1Q&ym^piE{^iv*t5Vx4
zTJu;&2GHKpYG6BF$|zOf>0_tel<s&E;oLJg$6{=x8fZPdQuCSZ>aoDC?Wo>TQh#}D
zN-|a@u)MhQvgwp|_W2+Tq{=K=lHQ^~CDNJA&az_&R_(elS4C8#;O7+>Cgney__@6V
zoEu5miVM%Tk_7Kc4MBwfW|vNAXf1^g$q_d4a;lbl3hiRJ{`16cIzfN54eSl1sk-r*
zgyOztN4o&*s;oJyi=Yda4qTLQ0#X~Z$2FtVZY<#wEuU$T(hyU%DyKi%#G@<z)4KmC
zC}xelx>+hMq;|;Is*b__nAKp@Lz`EFCakZS02Ye?bY0c-#e-RP{0atzm$^vLIs)3N
zFF$6*Chq{Wqn~8qSuY(YGY~yQE-sK^Elj!uSfEY#8VIW?F`{*{gsz|TR4z7$vNbIn
z((_eEu*xIFG*-<sV}V-K45r&#E}|*&W-bIwbblL~EALIIiL8AmU<b*%Ic#zuwQX?t
zu5U*w+y~%VjuNOmZNY+o75|H|cks<L?*4UaV`|&B-A-+{Pi@<_ZQIsU+t$>cw$o`l
zwcVcXoU?a!@_TdgCO7_pJNd4)?)AB@==(#yu@elQN*(Ioy~Bk6F3p#fu`~vi<`(hZ
zVj^6}SmvC3%b?k}>)ppl^ie_jZn0J$fLFnascyHLq^5nU+WqrS=E;&W-HMuN5<LCY
zKswq=nQ+O}NAPclNO|C+d*kOn3L#I!%soecf*1J4-!8|&S){fO0@qx9E@lmy;5eGd
zb?2riMT7Jy&X=+yn|WKVZqsoli>okn5iE~pUR&{MU$LCAN8au2XghI_Y_X%xDK2bS
z|2nwT-|KqJk1~eAcX0n%sb-sP214ORsET%%g*Y#_2K*FTSr5g4okrA7yjb&JQr4Mz
zH!7-ey+?Ib;PUE!TD?H9=-V;?r>4`nSdXSvH0W+wE|1xFWSrOO9c8IiMPh0@2c5>9
z)x+X1)9UuFsN>RHV>^_p&t(SY>Q3LUH`HUd-4o3$#GNp4Pt>zH<+pE-=@mu#?eX_+
z5N%`^wQJ6@<vLUD?(sz-Njj`0_PeM4v#x%cmXP0j*L=+8HsSbWjRix35O{--0xb!1
z|K}5w+8uaU*BhZc)&NEgA|loJg(?hc77F(&gsg;J01X1?b0`F+gsq7TNi71eWNpXR
znzpQTDkjY!hFWvODRn@=T8>1V?Ahz$TLqRp!@v|+sRk_-{nLl)&`SbZ<`L)!nW_ca
zK2?dBr4Hsp^F>9oQ9x4wed)XQ#rYwh!wFvF``qlt-|h@W4DWe5_qt>Uvhw7s(uJgV
zy_GMc*K&5vUn;G%&AVINa=B*(rDsQE3Flgatq56q9uA(oFk$Akt_LKa{)Ps$US%r4
z&U3$u&KN0MhH(iaeaf%g5!vqeK>x=Yk^6rt<ym-G+5f9)>IHz$0gva$<p<p(a^`*_
zn=eQ>sB{#9u18S!uzG}SHG0OE!S;2NZt~m0RQd%vjm9tM&S9^4uF^l(eu=|NJ2x+v
zOWKEqVPuWgo}Zi`LW%FN$<T<1n8MhI!*>b$4bU`c`)3naGF2P~=>dG7<>Nkx@fINS
zAcT;fJ?;P7=<&|*qVm4=x5>qS@A^%F6beob&yy7Y`8FXy=*x+UZMa}gpV7G{)$ted
zSg~Qgggt2KPyIYk@q;~d|GTNkm)-EDj-iT{5C#Ek)}cUL8fYHUG~tU(Q_{T46|3TS
zdc3CXlco0od2IMx$d^O>{DkB2!s_c5gw7F1^n={cNb(^Uw=lzLs8v)*t#b&z+z^rh
zDAhN~U+?mai$B?JIP0e+;{T%TLcoVLR9s$8-|Yl=q~rmKhuu|Fw4#QEgf54`AyCI&
zgX&H|BMQEd_AWoqtIV-Z`b&@{x1>!X0S(N-6F1fpo?O|NW?iN}L~mY;rZKAX2~QGZ
z6|`Hs6;zpa`R0`o@1-hEhE$9qFCHJOG&OkI1GhJKZwqn0ezdr2wW?-?OQ60;nrMI{
z$l3Cy9bd$(s`JZCJ$z)@N2bB9WI!`oqNc4It<5rU177{UPhX%8I$XQMaOw18sEN*n
zX?*tCo_*KcTS)@mu{y)7U%A}`y%t7H;GyzEzK=aekuJ0MWh&Q*0y!IJB@-rQw6P6=
zW<k=De78Z*lqNOLYsy*4&N)7G=(k)-k(3dFm>>2Bxz?%jsO@=Mr;CPGDW0R`@6dby
z#W~r0e9PJ5NvG}fSM)eM%=GuJBI%t&Jp#-z{y%Ein5kJui<T9XN3)e#I2-^c;K@5T
zMFOg=mEVk!ZmL1p9nA~??w)%(J{;(VOVgA3pguZWlrAvkHN&P0fV4{r%kK0J8h07Z
zs!9yPI~`HtR)hTPI0cZ$fz<2fDO-s`CM_JhGC3`EM@L{&)n@h3smAzMCo9o<`ALA3
zP$MW@lF;=}y?bH0dGQSnT=eKA4#!G@6}1}Fz-BeTbC5VSnlH}~#U*z|Ox}bz9rCY1
ziki*H3Ggnna^Tv9zD#y%!`o<69?S7w60pxt?Wfb6R~n{aB-^RxY5eH`ut@!$w)zFj
zKO@mESZ5)C7`*>Aqk=&k&k7M|ZiwHZ^oAx1|E$SM2`<HDVmw;~G}o=NP*~x9RR}Xy
zB^-OgYEe1VRpL`{o#u_`*d-`FJlzJN2j@`I#~vTe<YR}z5QpX-ye%F{+l7*LL^-3O
z#?5p6r#=>l?Ss#&C&yQdysvde@W7ZVpCb}N`Mc=PJwxYW%C%+mwrlG)K^R4+L8fyW
zxyU+GBZW6+H+Os<R_t6*wXZ718~?9AF)w&NF@B1(g7+9kbfM{oZ_P$}!dau=IEZde
zh^h2ptE!8>@?!;t*E3?EGjXa4K|+Z|;jN7umybj~%h7|+a&tH-<%<kquM$m-HX_Gu
zmN9tV)Vp3ABM>m9PSn`dby^HA3z?IZ%@TK)^D&!@mF@?x5$~@1zz#W|GUQpMBCqS)
ztc*2o9Aeoa_zx=50s@tX1#)oyUC4nWtbSbM=wDuYHX}&;(M9sl+h~R4Rtg*!U~}Kn
zWy&MIPx7wSle7<>xZVctdc|JZMP=S;<wkrl|65`~u6|QMNgV3xW{%+l9D;bgDpr0q
zF}tUi{xoB~Bf$tA-5<R;e(%f+#B;C!c=<gBr4DxqN6n3`-xzNr8MTD&O7Fgd#ieG4
zvH!AjFGcs@gXte?fcD+JX+v?KygRB|rwHsBDDv0cJMTwZv%4xpzwF)RdPt%drr;V*
zFsMaksSq-eMNuBAAkA}9sngpfQ0i*0LP=AP5Xg4;ieSw0uFr$h#B>|>@QPsU>IK6O
z60steW11Y1FpNVDGFU02f41Vh0OErzrc}mY=3h#6#E%hUGVOO%=y)pr0IxB<WtN(D
zXp_4RERleXrS2!CLNjMaYz@Lra4fGE09Z9(1q5wN(1C8;0jdE0UW`YDZY*XGlSLPQ
z3p(Wr5#OKV53lK|syr$q<yq5ySL3-2iLB4Mu7Fuxd`g1&k7wNR?r0ltkC#yE(w1;h
zs(AYZ(Zyzo*c_`r7fO~lfZ0jSeY`=X##-p84_eIuB}aXb%KRZX1a)0LHMh-AyztLz
zW?=hr#K1y)6jx}`?-wJFYF`2bNJe#2L2EX;cXhk!cb?#?oh|ybze>LD^IzqFLE4!r
zoU!tLK{fb(bkpGb?XfvwN~e!oN0&s{i4Nn}7K4i>HFnP8fsE7itNv01i$4$4D$tGs
zhu`9!NNJ~WbNqXU%G?rjmaMCXgHI1$Qd&@uqkf5tIbggFc422oj;KX#)QaG*)?Xna
z@%~$?eJtTn^{2U%uj=U>&G51z9z#4ks^{;9<wBQ%M?iRm!=Ot4-cP`kLs`Fjn$AhX
zk`$3-5>Il}?8G-4dh`za%P_R`l)cQ;t0FJs9tR2A`Hcp|La9)ewMC`RJG1l7pwEf(
zq+}|byP;rlufDe5%FR4r6|RAL?t?2M_Y2faToLv^9_jyq`@zD){$GxL0I~^N+z5Tg
zn$O6wv?|N+*btd=IiQbiquZE10HDA_<Hy%GJYKu)#3EU`3p#<Snd0Uq>YZAfO9*$T
zhd|*fLeWAjPIEAL{O^Yk_`C3(!NT^5?ICC?+d|Km?WgxDF2T2>OUlON`Sv~Fcej8b
z4$M2$HQqoS$T*ea07JMALVrm5RW{R~FF$*}<IngvX!%r$-%UXVp&70OU;bZ*g)IXG
zbIS8FF_ppbao<f}xjq1-j$vMbKjh)`%N?kA3}VB}cc#YNMno|ww;U&2q2O=Rwf8*b
zF){y(zLsqC(r=25ibgIy@Y#=-C`J2Me;JKyuuKr_ozJ$y2YN;ANYCMV_Ea`;swWQ>
zc!(lntfbAey%^ydqO;cd=DJtihPvzI?dc<>zn_=TyASrcL#mr?0EF2h``0^vDKvxz
z4-wv@0)(}{pG_apXKIQ0+LTYs?5g4>kByB3pBY0<WNpAznO#d0o+2vpw)<Q)Dzg6P
z`|Dz4dV@nqTT-eRi;%B^W|&6ZIyvB!9c<HnruxW6<o>7RvIHnuXEs?K=i6pcLa-Lj
zU7P4KTt`*gN`}2U<-^0-=xfOhTu*1i=g5Lln)Qdz1ZL%BOOOjKfXpuOp-)5t8Lo9;
z@CVA{BxY>1{z(-NE$#&G-%8c%_vCkXUuRxOl(*!9$0n*$DlK2%CI=kMp0-ea$NXn^
zPa#^nl516>Bc}0m*g`NF+rjtkL7IzET<NSR+xE6C4*LkpT>qTKR|~$m(9VOVu$fGg
zyd!Eu?&=5LEt8yH8Cxp>XImH2Lw4vr&II8CI%(5{KU;GtM~E5917u~NzLLI#s1i95
zj_9XwGOsu~+KogFg)~w)UEEm~JZZZ$$zw;mV(bL{Y|+Xq^mm6qcb*E5c*W~&pL%|;
z$zGFut494>=LBXTiJ>T(LGGhmM6>kSTKnj^A|*9dFKvrVot_c8P9}*2$B)THwtA|G
zTYLl1ag;E>#wlGn+`Xy%u{sq<!?-y3W=)$5Q7qdiu~8q&!*z{1JgHND{#x0&?vL(R
zD=mE<8*ei4Hf$-xbK~>vLWRohk;pLkCf%7DAnPjRS~+YS!Hju9t#BRprUa#gNo}jj
zQ`5G5bB}3ty70}^M*;In*l5$Jnhgcvh+19S6q|rQf@*7^Ou}`iRR?7Kt2d;knU;4l
zsxPxn2(}uG#GJA7%q2kv+lJt-Os#aT)+M+Wo^!4TT8!>+9h)Qu+0F><1ay7jwppz)
z-R5P~;-{B8=vC}x@ziBmA^jxJo8TSZZ7^2ts3)d8%9fIrd0Tx!dmT%Blj#wR2g3Ou
zVVqJtQa%(~!Xqy6$F}cHEe&=Dj$#*)QgGOPBJ{;5259u}7i?@XsO&V9&bHeGhOWv-
zb;s_72n^395T3*whb$f~7K+Kdhw>I@-ZV0twVqC#w*=`(6~l3xdSs!)wt71GMbx}X
zDl1j=3}Tu2Km<my;cPn8w_8WsqYLwCtXTAz{Bp3$a+~+k<M*~!e3q#o%<-|mNAYs?
z1(I6AmYJzu1T`>Pauc!2tL@fP3PopI`m~+dOys3%ZpUA&>}ujRb=a{+r?;XlHx9?n
zn^q6UO#qFpTJE~}NtD1=g%M~*B$!8ppssflOz3o~`Y1Chn*L9g$=!VF?NCqRpZ~FK
z?_V~?{<48PL;bn@D91v4v%su=L-=yPh_O09U@e!D#9^oreRaD_<BT9H#b~oOhpn^-
z375+5+7v|zvy&ulxb8c*jeVg})S@Hvhx|h;f#sbBgB{Oyi!!Yj(a#>{MH?c*pFN44
zWkhxHQxZAS6KTTUC>t_4zhDs`w%JT&(|3x;5%CTMFj81&9OB3h*M6otl;CX(Y!DNH
zVjm;T*C%f1)Pp-l?AVi5y}*cXzIF|hmed(*a1Cx;;2}#$GMZAs^j!~37#4pg7N=SM
zBHor^K)J8PpMs3WTaHd&&Y~VhN;mXI=-r_#$GOi!D;lSw3)*ikh&=Lkd7gzU6nII#
z6jCv6=}=qwD3Hl$Zz3}YCX}eue{#_r=;qp~X6+hQvCt9)qyPD=&}Ad9Dpt4q_fR82
z4m0&*Zc<^xkrN5hY7`esA&lE9Bv05h2vO{(#D&uANJ-QDA?>Jv5+-t#vR19&o3_~;
z28`9Xw*0|6_hN~D`7}g?ooVcOULPwd`vxI<>-b!@y=?-<=4;X;a}X8<^v|M2_l}FX
z$8q1KQ^ee)U}3hlZmb(9MFyj;d{6KOM2%Yy7Ke$Ue?7BKplB9zXdt0r$i)YNcALHk
zXo^qV16DicEjWh&LPqo5!C`Ual%o<BQ_&|rzt#0=1+S!x1j)vb)oYZ=?-h*w=&K3u
zgRZ(c`TI;w0r?)cQ76rLgR;G|gh_Y*Ojoj+=y12-k5V|wS`c#6i{~x8T5K?SROXJX
zUg1&w!t<Z|p{>FcXbMmGnk{ihWrum}+hc<`<IELtClI5>=FN+t^TVAc8S=6*o`P3t
z`e$==(||SqZ85EG9E*3}SkCb-uv&cH%MBR)qoaHq9(~gXavg&UPknz-@d`dl7shQy
zzUfpC1@$k`2zAT#|8&xGG5=qx`u{_-<o@r3TC>`+4u_oIE6(51dir4!i)RopaQ7~x
z1)Y%X==|s%fs3NIFJuBi@uXxEJ_S|cXuD|Kxx1W%)Lef#XYXAC0}o`XxD<@|f$GzT
zHwa&`+kixVOUq`EA<Int0EGa-|0Uy;{QqQ}J2zAp{g#gW;GR&|)B~R};=fU5$+C~B
z4-_Xe4z4&8@^0_K3p`L+qzpF((Dnz`oW9o1uK2Pr{JD9!E9#o4wJM73v=Fs%Z>&A#
z=u#B=SI`1P`^Is9*hC1<N~Vl~B&Rze&5th$Vx#x17CCN}R;3cfJPoNtm|wEQ8TrSX
z=EEzS`m!U`aHHGqM#pLZq$+rIkSl(RG?Ga8EAmdbP7WE$D|1aIEJpcb>_V)CCmdg{
zt*z|{b^kNePmL{z<s1U=eSElJ*%K!+uPVyGB6EMda{BFkEXvL{oe7!TjylSiL}Ko%
zN&M)XB+il18uH2s@HxJX%T)vO>U_OgOQ^#0@~VHNQ((SOZ$Np+tQU`#8MIFpg~EwW
zULKv0l`>&kJJ&c@*lR$MiOp9zDAu8iP>x5s!TxySz##jQ_Sy_!VKsL-WO+fCSTT9L
zXre*Xy!uI_>7yjew1CVzXmxGq2Zql#i||SP<JPqo3E=#QL_6?$n+<WOE5t`kq#{$D
z=()(lt(o;WX?5fw$14-hG->k(@UT*7PwY%1VioGr*A?=J!Vz<>^#DxJ%1thC8XOx)
zV|6n6iA5Z;ksg_*blormhgmqE35@TD>j7a8vHwe-OvC6xRTNOI&4eCnQhO+Td&GW<
z{@du@tecD4E(|Ve?NnolI_PO@pedgs`R$O{7t@_^s&h9F2mc3vil@LvDEzz%Lglz|
z4Bf@5?5~1c2d+MDFaukBZgWyyQ=3$+;&XDJx_`>ZY&DHx9g=L%CLAKubZ3Cw=8wZ*
zWoxM%j;)Z~zQBo+Yi|j0uC1`nj+6iyF}=sEZVE~*2_L9Tg@Sb8Pe)ifi(6DuRBEkL
zQ*aoL(r~#oI?m{Y#76IH848$l!i#lW`LV3^=~9`t`duCihnxPRbc&*je~bCQtY#LD
zy|X1vGu=BBi@k!Lpjc8*p*SoJf=e^PoG~E<xFmpi%7_I^?#SrG{`J*cu=p%0WZPi_
zW-?-|NjaHlG2gVf+to=060cG%NCPzF`Ny?*>)twHz1c#?!RhIh6K`k8IT)qh!m|`o
zwyJcb9=kb87xHoGpCN41n1vqO3?txqSf#5`l0g54<wIRsJA<O^naGEn$!{(_Bw-HR
ztUOmzQuLq~q^ItkQ=U4na<PzKyG%(8k5=Q(+J{z_MFvL3)Yq^Q3IojTBSB>(gsAq1
zb|v+=1_qj71W%jVd8gy<`?tyE7wr6pvDKa=C^?uzjmTF5NOF6tD984#323v<g?r#R
z#7;U9p0L;G*ber7R)pJT?uI+<Sy<R0N&&)kc4{KG=R-Y61A+5z%QnvfVGi&&^+vv*
z5B$_r2Y{niz7@DlBTbvBzf%}lXLmQw$+~z;+lJ=<h>o|sX0LJL6`D2V6|ky2E%7FN
zfB-8E6b})iB5&d|oqRii=Ea`yN|&%eO^$!9etQljA<6C7T?W#pb0O91IsFcWV!g-p
zTPK#Y7@@tGZd;iLewNl#vwuoCRocl}>o?+;ie_)LDA*+qz-dAy#+`1i3V#5!cOiaJ
zDMxOhl1AyyM_sVF5~lXGw3nebTUDpIQ}XwuX;aS^0bPmAjp*l%vcXzY=q2<Sc?~oL
z*~b$YCq5B+b<_!6VE26qISLezEk3~;^%P<0VcA0nL5=`xyG~9tY?fd5@g=%u*w}hU
z@75p%gr3{R&?_MGJB8VGz+CA(^Y)Z>`1!)rK%+%?rThI6ZQr@U%4hS@mg{mBO3p{g
zqIPw5hygR~Q`*g{>#<{Bk-e%waJ=Que#YNN)h}+F+ZP)=2rlQ?=G8VhYLRMtqSkk1
z_r@J6v}>lCW&5yC$nseeNkMFhwWm<KuyxA>`ZS<8MH^UoBZxmh=Nxal3hJUU>KUGG
z9gUJ+c4m4i*4yTsa%1}{++?F1wA*v*6kKwU*BCIhhqLzu7yvT9)5S;Ukz|>RG;7I>
zSFenXV@mwB-6b5dQF$C!UKDx)jk-qNu1tKShP0&&k19R^W~tz!K{KIDR-HtHHK3R)
z-rx<7a%>+Jy$ur%g-sY)B%W_-WYInQZsDskiJY=2OairMmW*6G0iZZGCQ1MCkYoLy
z>Khhr_Wy1|J^SsO{?76Ap3%KRhnr(~eB)zEouDTe#<qiV!M3P~4!(p+C;slnbXk;g
zM-WQeHgm4tYT*m|5<OpUo!+{?vcvIUn`2<(@Ae@IGjV@FDue(Kp#rTyhqn^;=Pl-x
zmt2Y@of#nn!uxby)sCq^)6GF<L2x0tg>YcmNiBaa;H%wFo{GHMAFN}q*#I0mQ}e+V
zze7As_qy`|y&wJu%7$~fj|Lx-UuSn)rv<8*M3AbRhyWZY{P6yO+;@1`Xsbi0pMQms
zg!+;86awzm&>B8doJa>#zr^<Z{fWAgm=+7;Q$#vfWjLV3XH%O)R)b+pl0_(J=q8z?
zV@oHfWk#E>yH!8iK(cbuXJnRK;PLD46TQmpDg$sC=%?&Bz|Ub{1Q3dG6{-E(mmIFJ
zKHG;*+nc5Htnyq_6ca%bO>Y~76fa9($gxuN>+Y#6w_?>ci(3(tpG3iUuzENF<rw6o
zE~~T4n+&hv+wC4!vy!1CRufD<L^s<5vOo+3FCtAaUU3NAaN_7O3ub?;Wt<x?>AO3$
z&G)evg9EQ$Jz6AjGFLi@$DYR>p>gFXI95L@2AVn){w~ZQ-@EF9!=pII9o9iO|ANfN
z<7K^N8<l@K+2j3sEUSgxwvA1k5b2}0xfg`x<C~*~n)YV!WZV2{U*g0Nl+^(hXNf5%
z$a4oO6C%|~Z5CGAa?I~7Xp!hIMjgS#pgG4JMBA+dz(tFuvRj0v5zwIio*3HR%PR1?
zSfu6Rka#OlQ-arqRgReMKKHu);@n|~fDJC|;yT2qE^Kh3<ds8`(c$EuFjS?OaWJ@z
zjK8feL!5mq*EO1~Hl->8uH07<j88_#IJD7cJ<f1b;+y-qwq95_8*2XWCNE4ai9Ncl
zzmqwZt%WgujY~VUc<WGEf5#TS&;8SEQVB(3{!58-+NZsbmRY2|RS%0eQ)27%ne`v=
zP<m^Cu_Br#=}3<+QXU<wWqTJo$5PCNx=xIdcUEeMl8!^r7%eQ_=`GT6IyW~LVaNF(
z1DlcOO8!OL<-}V4H`{zz1WJem9KqW3zJQ%KW|?G3<}!E6XV)JPzOyFR0U6zc`WaHQ
z+iBYIZH6b*Ox8dAb%sx%zT+xQ>gYsWXr^=#V|H)s$xIGexrXU2o$oJq5nbrV@XKS>
z`2?{x`>J*5e!4nWPFzE;f=2WhKYT~Nh-3K_-hUlHN<*<&6A!aW(=bf3X)H`pPH5=T
zi;-#%S0w(xtST{4sp?`3L;L&Sr~*=!^`pL`BN^eOa5@!@?X!VG7~Ni0K}#$dd|UZ8
z{eWST?^6kM63E4*{(0l`S4LZZDf&7X^>03)72UyV<))z()lAX4ut|(=pSmr!w?u<r
z6c6z5&c}qN^;o+`Yz`@Msur@P4kQc`4zoQlg$C1`Jw!=@Q`AoHIhdO&AKt|-sLmr5
z(&Tm&U;UZoqqEO6)b-+tlHx+yUb#%!C@!-c&e^M83D&l2x$m1L98o_+%DdtN%4jXC
z@0{^>E516yyjr@-%6TFZCHPi_MC=hS0bG``5|E>h>#1iaUUbzcv(xY_94?Z|eqhd8
z8wdbv0ZdbT)-lK&Jw8^m&XR19h6)-I$y&}12waGDF2i&#30A4E-;JM$v7m7Up`Arl
z`Z5wTz6}n8!BwvVmUo_il%llK>L|ed+NkeEm$aYCGk=ZLR)o*ml^cH_(@5iNTQpa(
zGczgJ3yRS{GhIovYA6d6G<tN#hxJ|R^xGEv;yd_bo-0Gv#Oo1TRz#KO#mKFm9r7w|
zmep!rzK1kFbkV!<i{mYMKBtEr6wCJ^=!5zX=kMwd%ioSKpu7=QbZ@ud**rz<)ARAp
z;1CsX?w_M8<rwCYO;%XjMXRo;wi($xPtY3KA8k=#fBoB{+{`oVRI;q7On7S~4ZJ;8
zCTH1=kj_LPPDh%DlLg<G+_0+G`zs~eF8lB)hO=z&`y1|!>@9946RrruSire|3;pup
zntLDjE!+jq%aS`n#CsqVd0dk&%&)PZSA$&YA)qz{^sDXB;WEzWIII8=uY)k1C2I75
z%ieqoEbuMrveYK#F?Oj&*Z|$uhq#ztPEm(KgwX-2J%td@hpuo|N;~URW%@pNGbyO4
zUY-5~GQ?^n^i8F&X)(9FUN>k<KRX-EtF&(R-1gq+CD4uP{ApC&Xf9VpRPh~6>#O&F
zyJcB%oE{<4@SW?(vuxrb-58_TJ~j8p>!dpbx>OC(Tv?~h*tbW%5<S28(Luqu|I)s{
zfIwJQ-BPoi2_-5i_+f+a%HS!yD}or(%^-oSGj@Q3oRhCZ1-e`!Lj=Crp@IJXu>F28
zZjpnmE9~Z7Vj}n6nKy<l{08<x=kZned__@kyHBZA?RoLoV^;tC{EpT(wX(P>r8j51
zlHr*IF2Vvc4BuAU^Lo1d=g!?~o_76I%-^u;DlPsa!i|fa^?#D2nYsSEK66w@HUX~_
zrEk9BM2ML^#sMriG0A;Q*j|6;f7|CaW=R5s#G5C~HHA3HwqnvKlxK}s#*KTm$RZj-
z@C5~R^@#*~EZIC!=sh`l{Bb}65}%o*@Lk=nL58fb_4|<n_@t}w`xu{DRjlKf5(>ck
zw(d&80~J98LAl`^{qLb1e;!7TS)`77L@|XG2*mX+_(t4?uv01vZw4X{QS=skeZH<i
zOCxYyredy+4tkL#M&$FwEIW}{B?XI5VQRAOzgN}@EVfUZk{@vCF7~7zPcvPrQD=1s
zF8u_6UbmD`)sKJzz2c&E%2m81XwxgTWcl^1w&$ykI`6w_SVi^slTqHTQ4OsOS|3E|
z#*_HQApJFbSL{d4<XJZ(BeO>|W+po0H<!SE=_F$}gt-PStpE$M-sq$ar#~s}Q&v~^
zlJir0_fiY3a#Tg3R>&JI((bns$+dm-VY*A_e=Cm2O4P*Try>Vqpn@5w6KvD8*5&Ek
z(l&&qSW{k|m!cyTDy4uyyJ43utU^XyF_I<;-TYzvx_ujW+n_rbgSoDEQZhp(Z?U1)
zfG1I2`!#+))-TX6_|C#c!f9<}`27As5{K6~^um1-D3`TAAAn>)DCkt&`!Mq^dt94$
zdllN+;Q?1y!G4h{fMWeX?iTr1b7!1<jg9s=%r*;G5%07b9V)~6%QTgWtUqB@MVbLb
z5#|ueU$#N~tvY#yD}LT`_05|PdFr=R?hVNFSq*V6o9vll6Ez`u!QC+{rNLZ+f~bI3
zUw+~iPrOW~Be>;6UL?sb3Ogdevj!<R{8LkkXA3wCHN4bB8I~%(DB%LZL+lthzP7&W
zl(|I_5;-=2r(g%J7wJhJ@|SJ5Bw2kD2bqpGDxKuk0@;MmCF|LnGJX+8dYk)Qa%1Sm
zlsspWI%6gl4)7D1O|E*QAp4DWF?;dXFE}%En62Zd(?%6h9`49yg>W=;K-=+h#<kgN
zjTt&66!1y8<M)*luQaQKlO$`OKe?(#t0V>uM`wh<WXkIP*9%1rg6ys}Q%92me8m1(
zkP_|%H@q3*gsMmYkk}0vUVCpRcxDJPQ)x!HiwaU=X4>q|X^TjKh;#@xYQYHE#OC@q
z<m5VEbFydJ(mw3CEW~vf5NxjBbd#6&SKdc$NI!cqthXg^55cEK5CEK0_?l_N`Utx9
z-H)q$TI@)@u|}=^3+`1#H3bT^HY{~`T`pwk1esHIi?^FMm~Klww(~$=Y9gWKvXvee
zpVt0Y$&m-=#!l-v#vzEC;5~P!X4foth?`YFLy$QkI@Mel7@;OExo?)4a}Shta3b8Y
z4qkL!_Ek<KJ16Ga(6E%{r^L#&Ywx9p!<(_y+f1E(U<TFPBnUp0&p+WeAKa2qJM@=+
zLplLEiJsvlEpeK82F>&S^Y~Gr%Cq3bUSoj^SwCe*P6R*MOYqTmHh-297QyKtD$ST$
z0MoUxuSD2t$8J$8B~B4F<;J)#zdA(*-L$zZiW7<$bmubg^i5hLpDmbs4>E&}X)<PR
zPbCoSF>LEsmf)aKX*b#ZF;AJ_HXX6~tcvBKCoNgO^YlTsw(5bz5_>T+<H4PbsYFxE
z+w{HkrKf##^1hA}-1ol1m*>PD|C(|trvP?5y*W8SYyT%VV9OmJ0|N~LdqrpD@3-pV
z_GrC*`nRncY8%~N2WxuW2}GU?-$V)nUBcgaKS(J8R^*&=g>vsKqhNtv;DzvL%PSDk
zkWY(C;}`Rc7=dO_YPhyaXr2mgfRGY^EikDCx9%lppcHkBe8gtAR^!C_^*!7g%fz{;
z=cfMNfXdQB2_GitNd^>dR{W>xG<C)%ApYdwBGc2g|2MKBYdRO0;OTShQ|Y-{@gaZh
z_v^?0b29EMei2`Ca1WdO8Yl$&eCY*`LDoufbJ;SfkJHJ;^>zi5p6?=4isb<QWTl30
zU($rFp<K#kZI{nsSo{FhifhSp!h-fLe1_#q50}@j659aGCWr+MQ@bjs)J?Tca@k#C
z(evzvf9B#=Y=)%QvYkP!Q7Pmuqto>5`r?pV`1K~cYisK7*x%Vwl9uK53MD9bHv4|b
z8MMNS_gz-*B~p%<0`14H?O|Ke$NbK?=IBRSJCiFFLskJtYkBRyLn<P~>8_l9&FZ_t
zRo!;dcA>%RCNpzds_#lBGJnV*(qggtz{;6MD`zFRNVVgdSm~wI8)*3M)Q3}=hCD`m
z%$cO4J1n|mn};NZvrfgj!LDF*>8dBK0=4o5oZl4i9@*<;m9V`_s|y)mFb)aS5R@)~
zZn?%aDfBc;??^B9Q$?j{yH4(&0*3RmW_50QBkFC?_Hn`*^cB%yC8%@cSvavmFYYu2
zx;}7%U)YJS<0MQ$o~PL7*Zm}jVqoJuKqBr~C;`fILc#arvy6V<PiUCj>pgaG6U`B{
z#LcJMm7*^X8lz{4y(BKjMPy>qgFvuA;f-8-w5@^TeNZe^2@@&qcz@Xvz%#MklWlxv
zh_bhCAzn4VX|x-#{;7_rRyR)8Gz%B+RHs<fWl1yFz@n_-bS0o6;h;7E5!oO-LJ@Jk
z@}?gh=zW_0dqjND=7hvfOxco%nQeai5S-BrB_^adlf7ICkWs=G8(cCjSF&iAyqcu2
z$s1Zq8^)`)_wpQlFSigZxj1Qh1r=kDy{AMLM9k0_Umz4vVP)X-Yy7e|(F=lrEo$5q
z_gc0%O|>M=7+gTVpdR||wD!V^{qKc!W9h@FT}Ikj)%iRYSv0*bE-z+FWkN}9LHtlk
zi?{~}Az^w#BQ=6Zu{&Zq#+?qufMZ|Hzk7}!YuPQ%`}#&dF|Ey#XR=E^<NnSKYmvH|
zP~oIOWq*-6{CSOUo=wDIW4Z`pdn2=k+=OR3rc-tVaNBb@2_X7c?_Ys02a5*Wh^c6z
zH^f!v4eUazFYhZiouBCYkZU7Nv;)NNJ5ap$5ZgLhx3#|ykCM8Uhq-MOouoYn1xZTb
z(5mcE29Bux<v0!N&LT6uFD2HBpK`TN%ltB%dU|eAe^+l;R+Uw$X6upSACBlR(FyjY
zIetGHnhG3P7GIEaOqIu{;)vCCUGSP!)UaxBYB6W$1p%0|OCaeb?)F~x+R=fMQ|O2S
z@%sx}vO0)vX)B`Meqy~jnkuR^v=avE(sq~<Os<vslFfXW%3+a7p~ge+rT=JtB79p?
z)fLkq<fKIlE(pOJwUApD98&5M_}l0BKElImmyJiYx;<u7X<Ka5e3s={pjgM8r=HFv
zOU)bne6^s-a<1!dt!N^hF;}N5$_}H&^qAfn?l(LrKV;1NYQ$qfRMIXf<9ji2=jWXk
zRJ7?TvtnO=Vc0up2|e&wyG_4V%`eVmG^9OQiH8iOfX%xYElqkXXSf8}0S_boiJ#Ut
zU%fP~b*Ir1m-<u8ZgeR|j1yAQV0wg`kTSv7eRZ+Ddr&VJ=2ErpB)wkDR*Zi)yk#Pc
zAFUUUZtb51t{c>&Qs*DNQd5%XSb{c&Iyn!?JK!!Q%1fm0xbiJk6jEUY4OSQHCJ>ni
z9R?-Va0@aeizd3JdL~>dbROCq1ZgSC20e1|)?;r4n=497ZkGM)=H07oEM&H{i1Ivo
z2@zx8IOe}*T*%U&lCvHV|D<(jK9>EKu|+o)?e6DDpHkU%)jl+)&vD8!SP-w+)|~P?
zU7s|~$qGgd$gcL7AM{y;zS#FGE>y9~kAK)V#mSW)5T-mhpCoh;yP;`w)PN<)m9x`d
z>M<(mknL=@desqS%3$5&Z!j+5QnlaYa~j4U*Vh$y!9VkDW0aMn1mdvHKmiGv_1vc`
z%cUc>qvASW66%J22yDXZC-N`ArWQ*c{@s64sol8Nvwg9viHEE&!&6LDm^i_5*fzxy
zczO8%-kLY;b9<=Jb-`;2W)Rej#ECYE7BcC17<+N*H7?d(qOa~f-jE;_jHT$?nE!Oc
z>6DN8ClVNnW3LxyeQatY=Q7}>U{oyIw?-0w59kM({J9#V#QS>O<^KrX#h>7>y<A}V
znhx}LnS5!Av2xbKaFr}4K+18~V!wz0T=;+gK=lx{@|h;`Z+<#ni}(T$CFvmiPg6Zk
zj{mEr`2UIc^ZcKPKPSh3hm||hv2oBDNBOukXcF!(B?4~e<!x-xfPs50TY*4woI_aJ
zU<Z+_MyZISQ`0(r`OTkGFww+FTPbtOq2HPSbGs8ZvUn^W8C|Nf#NayB?1N;n=z*J%
zNLe#wGC`qT%c50!$*Aib#7;j}7^N-Cgy=Z5TB#WKQ*xahY@ao~6FwbgKZ>IF(;}R>
zm1KHDXn$yTX_2&vXo4U8I@V$PdBLUu|5L*c7*6Q90S?#zXi5cHpSmdVn^!#*qRRl<
z(v_SqYLt<}rWYDQNn1V2G0ybe%Zs#D1OZ5LWjQ4%BO+Yz^jWXL3&<IR&S4|~10|QG
zv_9gnmClD#Z12h{WLsP<7c4wp)igl>qHUEBl#SG?Aize%lUa?;_UC0yBl^hZ0#1a?
zu!7OEj2K;873Mzybl^b?qjhEOq$C|;@x4D{m#i!RY2Bec%b2p&J%>gEp#?$gcN1l=
zK}(>4&YHXk8)Cvy;iw804K=U-fiY}#oHt!WdKaB_un^+3+IkBj=Ta`Q^%ec`0{PP%
zt4Qi6HvQVjt4X6o&8RmrhVfl1O;PKSARD(Q+dt#*U_`bM3qZM%9fOwD@BpS<xVjQt
zc^#dnn9T6<c41U)L&!q1%%tDI-)0}Wrg=)I%cYV7XJlBn8p>0`;`8~}TRrIBvQ-oD
zorfEdt33fOWSz>(kSv>m?l*f(6Cx~y3^w}Ztwo(SqdCITRA~6;nW%JL%c5Ve9fFGN
zPfmV(QV~u*084?f<&SvWC;#;~+w=n8&zC=En)yNySx2LPh`!!4m?$|zf#}isc=h&7
zOa?-JuaebaJaECj(CR=5qfI5mlm_#RmY3<Xlol)MPvJ~{A#M8dyBgu)h5*0r=f6h>
zrYwkFmY$s50K-HJ`<Bqed*ytRA)_C7{!V*|%R*kSuD8B|d_FvSJ=y$UuN!~fPPay#
z)AzoPy`u;HvyG87V@_8N$}LI+V<uxxXV2aJIeWFsnH}EjU;iHbncmpa+;s^fs#&zT
zXC7}6&_r^1IrwMYc3y+1>UnEDF)a8&g4{rkQ^qU9D`S^xOR*w2cINcy*El)7wG{T{
z?(6h$#r)#T8NinJq+Pz2(l6ha=lej5+~1R#&!B&lPKuA3Gt)%kk!-1UdP)PQaW5`>
z#YD9?(^nwhhg2|V6c>rkw_b`JC~{j+Ah^S9Trl$J)PzRTFh&u5fYalomOxZ=LZU}6
zIWTla<*~~0VtoE->%`JM6zlJKQ}^|8emr{c`j;tXQV}brz(MhF<9uzBsl8vHs5k4A
z^X1a?=5DV6YtLBn`1rc|dHU$jy|X(nkJwDUc`eu=bK0G^_wK!J`Kc@5EQgIZsl|oZ
z+@Gy8t3cjgOIKH)p0S)OCFd;VW%hpDKoRpUrv2nxaePK{e8n(`>MHxUz*CS;e5@xU
z?3*x57r<DIkp|~2@yY?hR&5=7!IFVPMY_j~Xm<@KS{&Mh>^YdsZ3H9}wT*V#&!AFN
zI*>6E)?P|DYD&K^&=iC2+Uy8kN(BYyS;==WKEB1)C2WTwkwe@avl$5D$PN%h2mj;l
zD)$aE*vJ~uMlH-m`i=~?1(?acdMio{Yz)F}W+POpAM$x25lz7Vmyx5y5UhhgMVUxQ
zkkPyw8sDGBdEE{cVB-jP&nUO2tV7Cc{DVILT0TJ^40cy!Wd|SqU&`30S%gQU!PanN
z{|(gR8cIJ$oC+ZYZdnr9$pnh)9~fR|h?Ng*5SfC^3FUrqiwokJ#Td~fOxZ3X=h@6b
zi0EQ;>SX$(4^kv?n%nJcNDx@EfNB;lh^>T{JED1WO|F=iNl0DuED$#4fqN7c;@|Aa
zSeM3)k%OcZ!6R6T5_Z%6p4y|-EqMR9{|L*!wFIT9qL-g+7$2cm#+ghWh44X6m9E40
zVQ|nx9&Q%#K$byh$z8BW6@L@?lh-}K_=ad+X+UZFv~*rN>_u?QjP#D-QX#L`k-vSX
z*2`TA7^#7)t%SLnrloLGZA1qA#O1eJlkW%L--GvIxz@msr57a;X&&Cc*}F#c)?_vy
z-1cO2c(`=>%1P|AQwr4p!{)A->W~z^cckht=AgW>dCJPBdoiU##8sCF<X$ewW`=R1
z$ALrSB}BMo3EZXcYpf)mlh6&AM62zMkbHPV0nL&K&unOq&K8mFOx|120mM+IC{z+}
zsv814$>`3NZA!S&sSngm4N}zqs_7Ac*MoS^n+f|6{$~wsfl$pXA^nCEk}bZEC<Q)3
zA;>Rf@`TXjmmn;o7Vrh+1+%(F=i#5PgfOvu?H_zPdpV!I*ch^NVk3wHv6aNQLu%Wy
zuRxLD1ApgS*#muMsLsz4N%BxOOEn<{#?W&7d23s0-acN+GxKvc6;|z*gnGZ&7Mlp{
z_ts4T#oOQ2aBQIYe?JCR8Ge~&A(29rpa}?x0cx7Wa1{GmE#)ZrOQK*)73t?H_2#tf
z<V`aCK|18~0cL6|VrYqgEBN$Z=ffGA#lOqOY|_D!)#Q*EK1fRw;>&LH=|>X|w_dsP
zKe%_1sJ{L!I?W|G5gPn}6&C{~#1L|z?nFxu?n!IsrU+TH_|r!pPBckV_Ms50Ws=U}
z#;9VP=FE*ZNd5T2q7dl3ROwGJMj|VWF<A2UkQwd1m0V)@l|I5Wo|!E3_d|c}Z>=Hu
z**yq1C`Hon2JE1MTCc3syT%L=5vQ$nZ70lCG37rdC0i_YsD9Zt)rPRLmOlh)C}B-z
zZ<$Owl=JVaT9jTtmaqxCr`ap{W33YFvcmT;0L@{wCCd_4rbO(`$WsZ9!^T<sqrYKz
zdo(0T3A7}Z7&PhCK3cs;iGsjTtv1nq;Xx%MN*zaP@#bcRL4um<60rn;t&nM7Y9h$o
z@_8$)wr_h>;f%yb8*H;FoO`rChBiaNtRPY!u6GbDm^V5|?A*OaB=0e3_ivFn+SU|M
zw`MCk*fVKPx9@cyUJDZcqYCl@|M@|lh-7%9`w<V{3tlkNy?p1fqVXHXZA-ibeFLs!
zOKnj4o%LF>w)_d4zx*w5{M0CL9PjwtAKnet)GQRpin(m|ajK=1(NxIDDnn+Qj>0(&
zlM@1}FZ3K5BP)j&RoE=f6Zx~y`47|klPqL1;5@@!a~+{s;{CKCzg3exL0R|J=JR~*
zwPw!`9DSI5&P2rgZSYD!iV_GHohPTk{X&y^9&jlN8`_2RynSVaiLQb1^@PW)6A4-o
zOFoP<BQt<aaq;3kvA3_ISMcB+HvWJna4yi_`)`;@jHHQ(eZAzGE&79wazzodr~b(D
zvlX;#{Rj4w0xm$BMl(G-@j`~xCgC@{1dEpvBb71$RB~xA4+5X1VtJwil>&WR`<~@e
zdjQ>^&6@{a5plBL`Hn@NMFrE-VoJX!kcsb_L<{d!l2u%qS%gqVTPC(}IrC#0fmi`%
zm|+l>FY$}K6lPheOVhG?M9wDCnn$Bq3`TCs{~>a>6Y-w`;Du<M01x+YKU5mH!;aDt
zyR#{hMlP<shx5Ut*|j|Op9)zn&i@&pV&&xI_^$#~vpU})dmhy9-_y6yBj%1&hA<{o
zGEjdt5ci;5EYZ98BdR5sB0fL7bW<jE$v2u8AuCx^q{qe#N8nj!BQk=J6}u1bE-PBp
zLd+CWFoy5igRX-Q0}d?Ny&HedNa5xOAbffK*2R90f2Nr$8<{`Y-LFRWkwy}L5rC=t
z8PTgAiW!9&B`A70b%WN?-g%Kbju-2pTO|w%ge5q|zC`}=`+SPG2;#Y;+zyxhx_|L{
z8n2ILB3~LqmT>-djrV=7hhastv#laeu}L~V->vO$UCJu9a>SEiN8pt_jBpPz*X<a_
zTCn!Wj0O2GfM~)r6qfSt=R3BZ>JR>ao3hq@VWHLdZf<QJ41p%7#^}+4s>@kmm+Bot
zR5C%+6C7F5G>NreA1|z~3_?P2y1m8{9r<*AaQ#vH(=dY+WL9q|q|+GeOPtd+$z%K4
z?3j^PLR4$}$F7J_`^RQ6)YItPm@)VcYpCeC=dgCv(wodQenO&z$D){$iQzP7GK&6P
zdQ#lgWQ(es$tpuH5D#R3ZW_6CAyfs1evh9!8;R<AqG_12eqFHgMza)U-1IP(9LoVX
zQks4SZC$7lG^PY;Y_AN-TWDZh3?Z!Ym}~MwY0j`RjE1gQY7Zfns%mHm$Y+Bx{6c_O
zOOPeL29g*9?p6BkBXeCAvhEs#wS^9ybQUH_j@0-;2|5e^v2caqoji3e3^Ot89~g~y
z!L8a-whsLJw?cz9X-IZz7*G|*)`6SAOsHpaGdr(%7CxO)LXZoh7qc(N>7z}!s5EV#
zn^MHt;cUIbojp`;g~UiOgG#2K4`r7Fc`{y!y&0*3x@=ZG7h%}V9a2U!uX|G6atsV=
zyI-+4#jvQFGo{#2Ea%)>B7<D|<;p9RFYlzYj|>nrU1cA8q)wKybK`W7+rph8Aa%@W
z3p31LcHg{g&@6R|laeW=ig#+s%ZSCd6Vmp)%9svgnX%cRYe^IyiXL<%w1mm4Q~(?R
zZha?BhpfwEPdZ`xkD$>mZBj@&EvDuvO;}2M7{51Zg*)SBGc0iC>hHz!QyZaTKt%bd
zhRA;(lb;jLa@*{{$ttiOsxfKE$MFd2UT1sJp4Il+ru=5m$cguwBFNE}(gL4Ym(A;}
zcXok8WAwJ~+H^_FK_=08+_slfb$ZZCO8jxEbSmc+#wv058w^D54O=#WL$NlBL4f6Q
zG-qTN9b-5X%JB-Js~4_?j*KF*GQuX8z={;zVR2O6&3lc!1djZw9ptu4ty0vowph>_
zlDNWVBq81heB`pIgm^ww6U&Vq2cxo9bCqjVUGOV9zdoUzX4W=r`vKh{?J^9WMAIcr
zRnryBgDV6C$RbvMj6*fY1@$9iO%{qd1E@A}J2g{YD6->YCKr3NKZn_?=*5x0Nkb#R
z$j~Zhp%Rp4)*g2|UF;Yf<q?VhM-*c*sh^UxR7E_^E`R-!RYy<#p56F4OGbhXwXZ%1
zRb3$}#PKiO@L&BDmQ9X&Yny^p{cF={RThK7m&{*2<qO15?gtMQD81h%k>*U4tGP!Z
zYYe5M>#azhVsU!m{n249J+<tLT4`{lX1UpesKzmjC_8Wii8DE2Mu)<Lyr6YEq!_k3
z<x(yR>u!Xh1LEqgu!>46Co}b0tzGnK2x;eO&J#v-O>=&idBRj0_wQRR%f?M>H<uDs
z1vVv0Rg8;2YD#t9@>;heBXLYJzI6us$VyDZnx*7CO0$q<-+u~UH>rj)=Fk>tnO<(z
zTmuE|4`mK=ebBg@YKdijvG^kYc>LUEX98`~nBM727NAYaziY7JE)ueT4h5mr$^iuS
zy++0Cbv$tUj(r3obep<&X?Rb}Rw8$aHdQTrB3uBD>4wR4dPlLp-}h~8%>d*_L`J`;
zW@`IHVd$@8N$=%o=?TX0s}~#`RoN<NV?rs=^{pb?RyQsGciwa6&L0Yqn@2vDDKotF
zHvL+1d=!7AI;?1W$Xb5^xcwaPV%W(-DrGT9_#tV%Y&}PwS!wyh4c>ISjF-Q{ciN4f
zD0EcujL14bu&|~D-*uWPJ;^SWY!-i!6M$!vrlbl(qnW!@IThR4qbY4tYXw8x?UIIQ
zXsGZ1`bDjmEnKZcEq}Dx-hf@Ciq3Fo&*4ZfYQJ3_uA|as+kENy^&m<37%ZYV%aODZ
z@nH&9){8oV--8J}_zP?94lDKNvdm-hU#e!Qow3w`S|FginwQk@#mw?&boGVgPM@v~
zl3aA_0i4Udj$a8*X^2Z<2d%l*psF}{ytu<yOEBY+w8C2WJJ=o3StUr9XO5+-pQhxN
zQ-$^h1c@oVPFmhFM9t9;TDy?ZmD52Yrda10nx%wjE8<B_u@Cn_`o@W%Ss&i37PO4V
zk^0S600V8Tc$knXf0mv=Y-Wd=TO7%)X|$qtrwfW_wo{7<2A6G<8jbWE8apH9%UZ0L
zLDm@gG3JJXGxf@Ljy(ROU@FY!gvM)=Vu&X^oRYhb7_?hNyLb7@o8o;(k`(*rv(+0=
z?$1)QFiD<^o^y_OFz_#L<fYtLC_9UYM{w&8@|-~65AyNIjNm^GH(zVHC9$b&e>b=a
zEw&X581%}*gLg>#E^@_b=WNmBgdTItY}<YcuP5)g_KDRMGUe8QFov-`RmBsz)(5A8
z{3sJV(G>@ycjDT}@7bweT9SR}WV~H}k9I*lIrmw-^phb)@0jb@-cS7rGD278GiBna
z`sp5{wd}gBXO!X;YLR{}*Dq@jqzjYlz`Fy>dIJ^lPwYL@(o50COQ=S)CUw7F1lp#%
zFIZcoSI)~`V-OuRM6;Smoz*|u-_9>jk+h4~or1~U!Ytxjh3fFBX2_Ne`ZaK4P9`hJ
z6~kJg7a=bknR~1R>E=4D2}J{u`mpD}?D!HoPNR0fR&6vd#vkjbn-Q4$<f8G&78dkG
zmZ>ojut3`YR;5aT-mKaqIMxcxaMD!Ng`OM+ErP>1NS%k2QeZp?vR8r#w~}R%&`%zB
z(y`l06Brtmi}V?}Mad&L5$)4~bl7T%lRCbayJs-H)J){rouIhS1L)in(CEhXEj<VK
zEiIWE`0l*1t0Yx-pBf9+QH<sN>O7)b@V*vSX+C~HN-3b!qAzxNm2xG7<GrZTd?65$
zMP*I>d5P$#ooMX*>b4{4pI)j-dd%VYBI@ik6S2;=cokB}N+JnD6!1e0NqOtJiO@ey
zUbO%w<%6yd(tU<!L$>*a!SS`D8Slm%o!kG0?I0oNx)v+)jf)ypIM$<&$muNt+99;D
z^w9iT2yV!oRS(vvTiGMH(XK(&&l<U5ZE(xyr+xBu#%1^5#f0D69ux6xly<MZ=k}S`
zSpn0JgX-kJf+7|-hknsSPn&Z-`ZqWJk|)s#Pu(aE&wp}bEg%w3%(OoL<klj@FF*B(
zsp+yZi{{zKo6Oo=@RhlQc~yX48ty0r+Z`(K88p|ioBYsgDj}<d=~sH*EsaLnBB>^8
zGm}FbC_jhl8O_`f%!9BQ1Md5F9e-xmw!dq#U>3PPQn9RoP=*eDfun6hhyKU9mFqvi
zp}4vK%aow*q&+zgQt#V$-73S7v<!s}bzVCM_PYAN3RhgY`6}e4uK;Wr5*pyhKe_Jz
zhp~5R4lUf;C}Ssk$9A$~+qP}nwr$(CZQHhO+nw`Ob>H;4I5(@-KbW)L`8;DB$0kjO
z=8DQ`;-JfqP54mX-yfV^eVwIX&JRJABn3@;<e(0*5dBmT=`g<QCn10>9eveI@MIT(
z4CJtU$0mk;)_t}d*nHi?CLNF<@#MXh%hT*ol6Cugg1({8{?`A2KU<_$ti!fxyAlNT
zKyWnf0>a-&zwT#VcCrl-T1CN08}CpJ#A;#G*pmsOc{%fq;n+PdQb<#b-Tv-tLHV4W
zoTDELxA3C2ec3yh%aOnZ5?r>4CnspXeLUY!i6#vvtuQXvd-LV#Zhxs#Qk8N7XL<{K
zTg|tmv(7rbXJ|i6?k}~l#xM%#W59+D7ouO}2Bc|kOsNrGq^mThBz_)${;jMP<U($s
z@J!k~mWW?*U5s5jmhg-dE?*gBF)!8rQPx?PX%Ay>|HpMEu*^3O0Uh8p2;XAm*9#dT
zQ7oIRU2OEYU?UDEj1)M`2^$`fmh#O`<wh%^9d&fJ&dALn=E;u&`ls5MR}|dn&0?Uw
zif<_xa@9dvg}d5Rwms$8Hy$vuH_krZNlW^kOt6&k<kHpddc&Swdj>RmBN|v6iqI}M
zpPd!tOcIwG40B9INIj{ha*|<2LYZ%v(P66b?@4WUJ)5Le2=Y8>CH8UK!FDGI5G-YN
zJ{iJctaXD73`QHXgX|gi44;vf<P43we=<KG=C{`E2pOzk56C3n3Dt!IM-v*!=&JW6
zwuYQ?gj{YVOeFQxnsq{*O22*f&2X=!Ae%|=VL3#C#BTA&H8r7aDFXo43bnWmd&-|<
zgJAe)U_kLHJNMkk`3~xea?9I<FvG|MExfPtfpbSA6%IgLL4~kW;o716j#@K^euhu0
z&cKPCNVnvh&pQJzu7)@Q2&o2@BC}SaL^3&NUuDQwbkKeY9T<M)?nh>Heh^YX0PPCS
zST-W$*t9O-Ei9h1=ltx7rS9qUuxLRsjXhw8_ArXWg_HE)9u(AxU5h_Nw6FBL?QGrK
z;S0V#bFZDqZxuri|4L~+zfQfs-oGDr@I6f}0-Mk8YkIQ#VdXCemL0qiwC9d)e007&
z-@9;`s^osK?>riT%_5hpuxU|;Q_|@ydIgj)*lJcP7s*!5BoRrBmA=E4*EQb^n8{;F
zkpwcNF`l?QF&JP_rVr^O0o^g0+7PUECW_buN|i<*HFbG{Os*}~LH33AHMT|(_e0oZ
zbLPX2V`bibrJ;i8`YkTW#HF0LXab)4%k8ul>3u@_H~64q^EYI53U@Z%X_Dsa$7+hZ
zFQp=1x~O>pE;YZNdYc8G|CYc$L9Q$NT-<{+<53J0p5lV4TbhvXzmMtBnBC_u!Hlhq
zt;Us8LzP3Z45T8wF5{V%9*@W^l$)tc8B=)>Y|PT~2ad^FSVuGNiEBl~E4&I@&=s5F
zkiRy_<FilHugWD~p}SV^zbT6#PY`kVp65l(;sr|iMqv4MY0oPzM=^K2{~K67&_fHt
z&^mg|_I00k`H+l~f90;>>vv4pq2||$Qg^la$hYwV<p^vH<D>#thil-9!5*CuLRMvw
zS9)e~6L`M5tm1H`Kh(h*-g=NXgfUy>Z8G5V&>k0x)|#Irn*cNkbZ|y1s)9Rjg>lHN
z-Zw%1pjQ$EG@iGqh5-tSFY_G_GAL#yYu5ajR_g21q~|+FPNsMC4_gRbthA~Wbsm?X
zNyJe8rAuaLo5=CrwkN+M-nA=7f-PdZW_2F;pPKABX|=$R!k!=u7bdb?MR%D44hVKb
zBqjiY8R&qHhgEo0K>4y4cOE+TO`S5%DrCMEBsFHT^cLS+4ko*Yd-NKU1G%Fs_x{^p
zDmnM-N$XQ_JXACWZUU9OKGQWiz%Ol-djE9ao1{eJj^)S`T~587=8=ud^K8O^BacU<
zN0m9=3-1JPLd>@XXrf$CdVXLLyo86IM+XT5hbQty09c(+U|VxP{)68Cl&DnUlwiaA
zqD^(bwBgAnt7gFbYJ63t8(S_fDkI$(L|uZNfiUb%fyxAq$~MZV;VMpoTi;q%f*57Z
zVUz$bI^Zv8syBc}cd@1~zA~(cLTKxzJTo;*>ga~9qUfp>Gz)@%zy5DIW1grAK#KlJ
zYIHI;SoV(rj$2Coe(W~1gfrTm?}aE$rabTw;ELa@Jr1cvo4qh0ue7QBpG64P5mYlF
z!U9L7zJ^`qH|vGf>YQnr*tCs5cQ-m8%8m=>#lg%P&md<th=$j^-t`+}Z|8StqBhaq
zy(p@pe1I#ItNqTb3*!v^{@M44e#~Ojcw6X&UKo=3N9BG7j^O{2_MjFl1IB=N;jtm`
zULkku4m<%koss?k#B{-I?AD$1#QW7cN0$ebRaoQx#)?lcB>G3{3|)~y14)oF@^%i8
zBKj&5xl2_A$A;EUPML~db2mD2jaC8rbC^sarJrr8ve*HsN=jl5MWQX|v^3n{_$0uJ
z#-9^K@`z0YkL==Z;2I4b_x!D@?x)r{0H{hx7#bxVaeb31p?%SnpYR$vt$joEijP;T
zJ=_S!6`95*m7T0AjX{n6X;l?W5^IK)8iZDCxZB<Qq0E!)MI>Fg-6VR#y6#iNIzcM+
z7JO?L7_zVlNt1r?D*(l@yQ>JHi+0sB{o2blF(+aTm7jX-zqG@El>#03dypwCSC7Tp
zfq{KRtfMB-hWLas9`)Ek+Hxc@b@~L|Zh>C6fyz`V(H-9<zWflDJDSA&vup4v1+#X(
zVkm14kO}BF=5eywiD|vZWU|qHE$iYrO=-!n)<_AWheajUj(SDb3Fg3cL_%=m>zw&5
zxeL1tHT86N0&sKt&Y{?;^QV(z8r;`K<PKnWYoZpm)8m_BW30Q~!b3Bz&EkVq$oSM-
z@;`s3LfgPqz>SaNpF56L9B$ITpYa5(4dU*R_T~#uAMKT9oh2M~pV9qkXDbeTZ}%0H
zSchOa3oQ-WbIBFvN74~G)D+62n9AA9ZX1peG1hbc3pOk!3!*$mxce=?mjN-b-Xipy
zHs%lx0Y+IITdNr+VTc-NcGeF4QNZA<V?-60su<Bv1B&47+T>4Q6-sMmd2gd3;j263
z)l%^ClME@>qiOe|+S90917sB|f%WU8jVEho9U)2V$L{b~*_}@V2yP-*sV}S;BIFp&
zl0WNjf*`oF*knE<)W}DbwbS2~T1PA8e+N4|7^L>C><z!V_$6x_0JQ(ES<CA*8VK3F
z|N9ccR8DNZn-pvaxs1eFpi6t`D0JXk5UW)KV%x)dMtVu)HPdmU84!mYl-T<)x!ES%
zLgJ;X+`JZi2)7!VBJpsUnZ{w4LeiEExEcZDx{=vhb1TV=Ox%Kya=&`fiOKV^(0Lh8
z%i@50&tts&vHoRM<JLos0H>I!XKj~rS&GYICDAr7?38!KKhd>lNL$IF?rdabDgjR@
ziG0Fcnt&lV|8-#=!OpA!Y44+5clo~4bxzW$09I)47bbotiN?xlp5*K{pdiV^Bcmj5
zTyIYoi4pZ17uGs$ezoa9Isq&=IW#`O9)<UR2L=~^D+3ahd~ZEhEgq9`yfwz$rN~3K
zY~H`h5)i6Z*W@9oe{jollWr&v|Ew4Z$UrY2lV3>sX8d-YCU|Iou3GV2bPJDHD`1e?
zmdn&1ptOiYS-PKjhwVo2hV?jlP*2wrv12Vb#0tc+hm4E7C{NW0uE4a7K-<<}t4t{y
z416Q)rxbKaG}}<FGRB#@C<wZuI$%ar&}>wHaQtx!9&Z{*ZdJt9jQP<4JRn@0`I`Q^
z|3OGyW@<Q?(3qm*zRD12=IYh^eB^0B&5r)BTRH20Y4ozP{cmNbEA=t!L*~D~Tlpif
zv8o|5YA@}sb;obp33$gp(|<DaIQH#FAiRRq{7kte1?w1ZB7aoBswTWBdEkW)PUjk~
z#-%5V5l)#Es3Lg92KW|w%!ltjH?I!}qCn5}i|+FcuZZ3g`<0)p>3HJ)C;H~mwWjY5
zW*5D;03hZC#ZQDW#x%Pi^d_EW=x2k|dHe#l=C{%-tGoVN>HS;E_5Jkzku?b;u_yJ)
zviV+;veMymL;fKSP47)B3~jb``x4V}3*;AA=M6yE$viOQ6Kv!1+jr79V%0qoH3<DT
zE+_n#t@t_!t|s1omc$tJzVZF!<RuUD?cL~50WqP4i<p>1e`Re{UhUjOr3mNcMn>55
z$|KyebLcAy>F~+YhnY4w0yd<?kwjeDpIb)+2e=!&;DMp-bMj*J_=P-@3oIutBx6us
z81jGySpwY|yj`vGFZ3SHaqQ+)16+0PEr0f!DgRnppqvfO$BQsUWkwtR?)`RQpdi!6
zN2AZqt}IA{ml5id2Wh@-Jatl_iCDT>94uC08;d+BkE@;s&ba99^exf}R!uxc5)uHz
zm!I|_JTeql2!vX`K4gcoTvu=YGQWhD=@0;?;f>UHa-NN}AO0xChC--atU-`tH))7t
zF4pyolxOK-1%7Znkp(Bk(6%rIhxL7VBa-x-Wf86c`~79}ONlR>v2YN5=$aKyCBo@m
z)gzO>{&-I8Zhfz!Z+>9DSp-VAuv3l_o4vP!ph_tgVdxn#ctK??!@00N_yR)^G8&R3
zKgY9+)G&HCCXEtQ%MgoBPi2QaBp5YRt-X0YSi`=ZW2_qUBr$>;LN5o)8tXa%xDu@j
zV~vz?1hcukkkkJavdtuv2Nme=>d@`O=4g}11cU23p~s$5IU!UKt%%eFyw(uZ+4qPY
z@42%0%fOfw$-d(r2b8I8E}I){3G1<(x@k7ipgDJ)_m+2PNDo6Bg6EMnY8K2eB=nkl
z{ys>vmUu|-vT05zL3iXlm+Qb+AdW|84pg_QIQBVwu27<4-_rs?b4TUXk`PMFK%6>@
z4ueDD7r9xCJx-Q0)slNZkGB$!#oGgE3ej>C3?0B8uS!Eap#71;3S%q&JHZD9jR1Pd
znmkA9TO-3{xAL^n^kiJlKP<^|z`^lg?k*@SL;F>}D;!o}M(f92S_s^zH1F>%q`w^Q
zh6bCISu@`N-`h+yw+>mL`){mWphjPp6@&GXtjmB|e>l+IR)AX{^)CW96|R$#M5SSN
z`dd*u89%7Qd1A~swNxzFP2c5Ao$k4p+myBD`yHTddBmS_t<e+05*~6Fd^c{X)N@*u
zGas?Hv40isf6A<X`mp%$aB<m-QRDeS(X^?7VXOBDm&QaRiD^<43uZ^YrCTIeb3+($
z2Cwu@rN)YwW+~o?<rwVkDLE_V8@@X!>hAjl-5#dfDZ$UJQ*hW2l7%n31#8v`2@fM(
zrd!@-v9+FObK@)2tJ{0<SqOKE=BzXl*i>R49EK+7YW9gFa7Cbp-B|NQ)f+M^IBquB
zTZJd2+ZLTU5G^PmjO(QWdZzLXj&_y1T8>X8?p%op3sAs}5xV#WE3*mvz-O+RCu6xC
zz{f<efT?5!j|zcry6#_Rd6Mdj+k459%-*Ufx`OJ64hGlc{qK?9_b)wPn$Vi0Skpjg
z$2+)5_ub8*geKuK*haO+6wRerWk418g4UJj?s%A@0{>C%Lj^4-A&6vEq`0wR-6za+
zCYf_1dkdF*=hPcXpDni5St@Gh&^Tdgo6nh`R@HQ&h^JRi$pHzSG!@sKqM?jbE4qI8
z%W5E~Uz^EheiRpk1rD~gsE~%HE45vCjel@aAs?Jb4~WHB&*zsDGWy}R6z=SsaYa2b
zt89L`5x`0xu9LrJw|L|T`M@;xK5d;Ktu{H8Qp=4x5ARZ1vKdH%zUPvwx#4dAt}>Ox
zE8^?Z>ow*wH;z6EBj|wTJD*X|jJkvvOo(zFa*55ooGQ2V*q$;J$(lE}aW<cW^{Ft{
z1UM288tW|M;@tNKMpqfk29d4>;vzalmpoJGO@EeHa>)Cq?Y%#;f~&28nmkj<R8k17
zawz<%#XCqmKTi59JO!Mjc4Xrm$U&fVq#wq}-`_1A32E{jVbo_gPWVYVO}vOgbGX^h
zZ&G)V$?{a+Cmue$OOH70`+ahB_(+e|@U9(dn>IePem_V5EU!$h^eTomU-YZ@%)*iM
zrm&z!&h?h(bq3ZhnA5NA8FqnE_A2Xjd_Q<VH>bI6U|O-CTtGB;fs?G<|9)^>_JWhX
zSKj{&$M4kbt6Eh-;40n%v*sne=Z51VLET^5+jMaws$qN9?${jwldy_<GiupQ+`QW7
zpS!8PVe5RaZccT-E!<I^-elr2e)Q!ecc7;D(MfuMJe%V&qlb>nWb485Lf#oMm?BK?
z>&71BC#HjoQS^Kf`R$RoUl}s53I3@*GM;Z=F2YnPj}Y5H{SDL<7t8b5Tbj=llDb4c
z{Q!9Wx##|GLz|WM|KCuir~e;KgI5?6))Y*RJ2j(!bUmYY<dMt7GAQbRokCRcb%%hy
zefsLWF6+2z3Z<@v{D&B`&_6nNaNJ%G@c!szOW(-s=p=Hp?!mHhcK&z+xZ|JEh`O6j
z_5ErFcW=K>_D)R5=H{L9B_zs_w4%RX9xbAG$Zi{fOhIQAd=Fftfw|-!1?MmrA1TiG
z-*@gBpl@0lcO<F^LHuDG|4PHp-=)0ZW)mT5dQ&8gL)<*}u~UVj6yFm`wNz9u3!B~q
zFNJ|iwKm3ZvbyJo=?A<)at5%UeM;sWFt~#|&QTj&z^Us^!V~r&W;VjTy(GQdlLw46
z1|K$xTI@}_;s<mkH&{E(sKr|qrk4R=-Vi(FDedteQB><!pH{g|Z(^R_s46EpY#U^}
z(15zfgB&a&Vh7ce-pJnY_f5;kV1)kYK(WMuJ;=Xe7V_M+zbZS~P&7t&8|aya2{yp!
zCY%zOY6Y6G;L(v)qbezWNgl03%`Wj#M6Cf}!V_gKNkLb|F&{jHb15jw?=}UmRwkiE
zm+zIgOZwk_4Gri6)JvlsfQwSjnd*)K!aDT5kki&2aUC<E(Wflw7HCe3kO~4O)MG)<
z0mkXO%O4t1AChk}(QJexI}@~Al290iF1`GNNH&nyn}Fp`Aya#3i)>&)F~nz!20Hu6
z_6G`jV^=CDGZ!F^gaxM<GX<(1S+%m1Tr2i##RHfMB^wibsbgU)paScuvRDMV1Y0Zz
zUNG-8CDP)y=v>kL1;===YW+|s|0TBX=EE4zl?7mEANbcY&N?2*U?tU@bxDjwQK1-Y
zu9m?FO2WH8iK3jTs25!MJaCHg%OT6F!l~oGvkNDaSRQn?=tS%b#IW^jz88!+oOH#;
zM$S+0KQD&5>$F)zGC?Y}D`B%bP*7$q8__~4@A%B5RAjnWh3+Apoyv`qvYzMV%ix_c
zZM}4*hX}F`zHq`!Y4B&w*dwUn^CcUz)_>M!y(Ns~f*zzsF@_Gak4~!CO#Z8$k8ugB
zp5M;86!!0fn0!d7nKR`@FSHiIRg35_3EMd7rY^6r$_;Ta-opOc&`k%vsPSI^7o*n9
zC>1X=#(k?qbiESqRs^P+c}|sX!PK3mF)Y6(<+MN+u`)h5ZqYVyjeXxydU7T<sJNJr
zLnM%TZRtW{QemXMAOPB9Bblw4U0&svLIHytAVxe@`w&-t3hJqil%E+k?a52IZovns
zyIZB$;t!$x&5OURYIC*%HZJF*fZg$VTtaiPSFjqN3}k9jN4$~9Qk-Zu@kHOU2Iha8
zTY+PEU|N!ikzmkH_9`j#??Olh6rb0+CW3o($wJUk3fJOc{!6saqb(%Ky`wx<k)(r`
z_Q}2snU)FU-;-`d@j_|D`dTUG`y2gcR~OG0(}TrT9Hj=X_tyhi_pELpvHD*jJo)eM
ztJ{x-7{W!pcD#){(fQ8C<jkUlPN}W1VvzlkyU{EENw|JJ1t_WAiyXs*t%Mwm2}r||
z#9VNwT_)|X*}%)x<_U;^^COTZk9bPXQ6X+%z8CYs{Lev+1zzMx3wZ$|o;(p?#s}w`
z(FI3Gt41{|Xda~H!pOT!gV}oGtEIs7awiKb4y6i*eA+gr%Mj9GhY;T~xY-T-6qO|F
z02dtvNXglZy?{7munF*W@>WQ;GG@qfNuz0n+jRD>@FLQ;ex(5HbEs=Qlg>X|a|AV!
zbbT&Y>FU@=BnC}KX*k(h=ZffL$JK#0_NX^>?1Wm{=t)~<&yIv<r@5fwJnOwlUKiOx
zX{(@USb03OZ)Qb73h%*(E*_FGRhh$d#R15CYgeh>QZV310N)zJx7q$8L$^2+CI+s5
ztSEvhrml+|CKQXEn~>_PMyHa-`4_ihPFnEYWi3^vLv<g2>b`ieHo!4k9V~Hkd1#v}
zR{8Na|8BfYBPQ7gdL3O>Wl=RVD^^ucH@|chNp5~wx^v|8q#IAyG1oEK&+GN)|LZ7!
z-g2mu$jES06wz^rFSg<^lV;UP)n*uCbKS~zM#h%aIX->aV9YQkmM}Gp%b5T?oW3Ls
z{cEdxOu>|fmLd7plr9;mAim?a=L9qv1nHeidlyS-)a&`uPn5y~jKaj?zm*?IsS!sP
z8Smba<`B8VVajpvNq^u;87fRYo&tBWVJ15lZS1bYmMs)sy1Ur=&Mi{Zme=hP7FfsT
z_YV?F%y%Yh*@PsKMOex3+D(7U`DLoC+)QR|7*TqpV8`v4Z1(8_PhcZ!bt`=5c}XF_
z<sorOquz@Ls>w4VE;oBQfA|?I={(96g`Jmr>9k6%b>&FEROdMMx~$wk4sMfsgAsyY
zSCaZqWpXUL-o+Eia4&>9>4%NPWYgM~*eDTOyvE3Y%J|4GDVTSWC5*P*a(%~ID>i(l
zFwYyW@IReF_&0w3?05=v!s*d3N%egqj``a7WY)OoS01m|9XO!*@RZ^@dKm1CV1!~K
z)eKq{nNjOF0H;VpAlo1mZUq!F%s@uHK?I&0$<N(Nz0TkerUvCY2nB|!d@Nd`{-{Yx
zBABfND)0SIu^0%)55V-U_W6an^>mK@l5zOiL>=9C?jfuP5%;4DqVM<#9vSqg9@^O_
z5U;t)3yo<{&z8)viLx)`x5PbL_gR|(L*gDOleZ^J?`Nl8=An*+($4eE?I}A->CAN-
z&~f&zhb3RBG|$BaaZ<5@QH*0&SZAF=={-Xz7AB@>_>u7|`hj0L`x5WwxQU%FK6X^|
z$yN8qW2+rxfD5K1pd48M#@&uA(KQ>QIJrPsAUZ8)?IzGTt@wrr4ySBJ4cRy6@h1MW
z1yr{b_S`{te&i|3W?ncN=_)k=*rf~C^c2;#MmC4)8`T8#e3YZL4qHw+wP@aO41t%d
z9X!|YjO15zQ_%3LWhB9esa4)sj}n@%RWix}jZmBqKm)qr$A4XO=>Gq26Z8!9|J%jz
z3bW0cf~EFF<rMrsK4s8fk{TBUZ#&4fzan2exzPQOQ1A)0iH(Lk%@D;hh1^7&P4&@P
zO#<iFtD~FG`(E+gVIa%y;@Q^>;_RRJZ~sCe3u!oBRoIf39-04h$!MXr%ZgW>**~r=
zY-Q#!xw-?LLgp&@J~RmAB1=)EXUhd&$5RXCXsLB1eI-A8grVX8waf+gyY=PydH!BV
zpafw#|Bu!<!HG09SR7j}4iAryF|cp2gN>veW>!e`*AV{qfCPC(z#GSEtzbU;ox*Ei
zwtMXp3n9pL;$=?_29ywH_EsH_`4Tdkl6HF>(Qd@zgQ<TuX&&SV!E96Hjve``?BaC;
zNyI;X&tC$G#hpDY7OTM9yVWhVfnZS?<cP#Q{m8evB`!#_0M_UtZJbj>w=mw+FC`AG
zK+cM6p9+)uOHf5mcN?}+k-q~H*l^QsWe_V0`4lsy+4ii!rZvSc#8gDsXaZ&2wMJSV
zPfHuq>pjKNtNnc<CPE$zF%Tv_Po}gE;-=oT)8Iut-alU16bg*Xp8$kRba@>yYyEf9
zSqCVlMXc#8vzGZUW?u>`Nvdft<v?S48O9Q<zS3>KwzRyI_!PGT5sa)Z-1zVe;r>Uk
zlj#eiys_|0tNFcar3qk%Obz-rE~jP#Qbgi=h}p7(9Y?aN9i>!Kxb01A3`X2Dev0W|
zDSM*_3E6;>Mq_mkT5v~$At_6ZVtd}wLWbW5!zp0sqV$BXfnI_9@O#dt4kZkphC|rd
z-oAb=zJ<Kf9{k)pdpFl}`_&wU`P0JZLFJ+hL-SMaXhB_xfEERy;<%iWyTe{3(+dT`
zG-`-_BOPY4=`VtTMH3tE4@w(e4cA&V)Ea)4a;hOp)`|t=!&@5~<LjL&FTc)7`yNBc
zW1s~oqhO$nF=eBrr<PX=^PId#9La}~4^se^_Mj5|JopD*oWX&P{mr!RUBk}d*3RUW
zWJWB|AeD|38Xq~%P$Gf(0Uk#T-t&73;|nJl$1Rm{)iFv$G_}pN%P7*U??cb(skn$W
zt<{GhJbTkm^)@tX!M|7{&VdAFCHfS~^B`Odj9ba5Ixg&AJ56yiupwtx?bHL{Yl5cP
zPDB8-?Q%VM!5+c_)1T|3o2UZax+9oo(*tB(ZX%~n@v<Bs31<y$3R^)$KnZ6pBO~MZ
zb~K-zl%-kHRlU+BJDvs~b%1@cds868s6tpCh(qEU97)Z;&n#$5zB%xe@;)l-RT(a;
zs-z^k^Jfw{Fd;Q#7XoEs>?lBJyex&ZgWutA$C2y2sz)r#$HyGBN!CL9z6!Q5F%Dax
znfaZ_4R}Ec23Ld}I@rJcq;n5B486>+^}PC#c0$Ex%xidZbV%D4oa&$zoPm-4i|S#~
zAqaG!k=L?P3F0c-ZLgSiNnAi2Oc_#7sG@!na-}LKj6|;taVEd&11Pm?>(t3q|85ot
z=6f|NLoqHQrRftDa>JxB7o5l{<mt_>$R#AvN=)N;+<42G=FZ0<00b7nHYjmcQ&_WL
zN6tnJ;zAL)t7$^|+Cxo6URufvtM_VJ^^gjf`cCW=!K4asdYA`NF-5l*)qUE2%?hf^
zgH13%q#QINW|yp{%+!DUDTnb_UB{ub3-@_#6n~}YD(NI&l8tg@uVM|}d?_z@v-s$3
zMnoQrR;Sr|wvrr*3IQ16YO1>(Cqt0@bK_mdy^@U{iezRjWO-8=I>D7xFc{f_)b+_Y
zMoP@(v9rW3!tzfxT%xFzjKPU(nMSatXkuogp@N#N#XY)=-AGS>ED>Yqi2snAqkq{Z
zOBh`>B$=Imsz)@tzbniH*V{*rq+@`4M7lh;u7lE1Vubq}DjOb~n@o;Tfd{NEzQuv@
zk`W%-x>jD6-p>Zft3Vk~%%G2lvr(j)VaUH>Yk*Kpbjw9nd04n{18a2ED(^PC2mgX(
zHQ1-;^t%+NoiA<%!VJQ6hu&k4+WQVQE}B&P*(n;qq~!vuwifFvkVX&;;<Jy}2uhgP
z_zws!i0tI4W@V%$QVu0ZLo>pL9`DtXL>kzjn&?Y$`qj!(lD$wjNUnk!!Ua~<VkqZ@
znnI$J7X-A>%J_1hHU2oMTJVT17dH(zYkD6_Tn0*IEVFOCSD8`2NO&%_Xm!613dSTs
zF;tnMaE*j8QtoB8JtYLQssx8|f_ZPAe&Hg%>+RTz@3>ZqRWOxd*I219W$(;^52b9u
zoKH$%U_91aA+0ATp~#R@H5EW)ZlG{cHz{L}{BhN*Sh;)A>ITZb7N9i%G>Dl%Bmd+K
zVQBP@08vGDaOtTbYnk+2rsJSVWP}Oy+_)7=lKxsNb~^HPAs{u((EUn9Dg9{xgVQG{
zXc-+M=&AG&ch|5AFU2epn$UTXNB&T@hR<D=tN3f8;JLtW*q}VR_e6b#Qoq3aWXKeC
zlOLzK9*#v0O!<bg5o)bRsXFz^e*eCQx)X&h;iwfFY=N+}=w^0rH0{Cy)-eJ@nG-(+
zR;F~N=eaZOF1<VdH_H*bB-U5F_17;+^7NMPQ~)o^1s+g)RR0f7{`DW$EOmZEx_<c{
z#|;CZgnl`ot2}F&rSl!#TQcy2=jgSwl(wt_rp$;n<#>-zk(EOT8Ao-8^g#OdLEKNS
zvxDXpmv7Xq8L#WfTCtY83cs23b=GjU!Cf^O_Mt%%xBO1N!vtO?jHa)#f-eyqe8Lv+
zVatY@@cqR}&``W@Inmg^d=EqD5NvU-**rtataW^;chv9zoF#$PQp9wLr;g}X#hwQW
zj-=0f))IE>e;%5%KyAB<yBwagLVh135S>w7CLZ5zelu=)CGQ#s2AQ|O=^fUM3EAsn
zGoivAWl&UR;9rY2pU&y7j_`CPW+p-QdNIW&{d@KSNFchpy{N{@8d(ya9WCIAYw9hK
zQcg5St=<RQ-eKeNkLGX}!xV{%$<zCrE4CCOQ>CC_yfZMo^RF$9R8NuJQ9RTa1wtVB
z*r;cxtp;-%k%j7*Y)%<JQ%x^AZS^IVjZCIOyNv}KaP7m8z?$_s9a^d}Z74y?e?r^d
zYzclsj2m)R<~C@axVo>oo($l-c-~20e#ht!fH>Ih)qmZx{tNDoo{^63f9ew~VU9&#
zFyC+M?BTk?r0jz+!Zd!Z1KLZ>w)AC!C9oaQzWod!se}=y6}CV5Pxf6k7m_b`$L**5
zXuckuq-=~#$ov6CoybH{*5>x^f#~eJtJ!s(JB>j0M<=Gn_0jyk8*gKLZQBE9Re|C`
z`Ehxzs@X}jeGTj-0ntX@f1mR2rTE3udkUL>9>2!D_Drqdpl{=K9smFW@1wlt_BsXj
zd_Nqdpk_l-?<9$zrP&?pQ_>a|$)>I72z)=<>Zi!e!enmg#=^gsIW`(9%i{UAUoIR)
z>O(OGEsj<T-<cx8dKcLk0?9JsJ=i39hn)0_-+goKAVMGwK7C;jQFoVYymtc)jP#z2
z-1<bnydJzy-2%aLhx2!pSd{jEHezv7eZ8NVZMx)Fk=WC`FGt(ils9kMd9|Q1%EJ~z
zG}m|-q(1oTr^<CVNlw5%osjil@XEYVRq$;84OHb8S%Qg&tg1-=vSVsN%O7>!$KDP$
z7gn>Poh)6W{hQJBqmOH&_67a<wT(<d+UV9P804SWS4HcU<i~8Qbq5qI+bK#XwxhFS
z|BvqyBk-)o1Y2OHD0bWg`7DrlobB^RbaT{09^DZ6LStt$ypO!FL0m+*%rQ!;APvfN
zkHCt$&8-nW?No7%TewjrjMG78sF81F7lTkwi^(q14eQlH8`r+HZK;CzHuVu?U3Di3
z3%Ht@;>n2$i+hSVD7?>xKYyFSszoBWzp$ecrU)U}2?%{HDtJ~)S$`7ReN#>OB}Hgv
z-WxdVTh%bzlJkc=52OpGCUgaX>D)_$v%YOV+?r6ZmLf$UUH!5mn4u%siXd3bPRNgJ
z+lR(N**&jXfskK&q1tkJ_BK>@yoOD*%`>GYG^@tnGK{YiwsS}ue~s)3;0pO+obh`d
zc2Cc$MICdyM*LUPto70EDl)q;uFy?}U5LWTx?wjQ%ozru(t4a0J%XW(8rA`2LZQV4
zt)Mv=D$#mm15ElGvgH-SH?p(k-$Dtvd<&6Jh9qNoogOdeG+rV;3DRu&6u<77&K+YI
zahpLerX`iF7u!xa=qw59hXF%c5Y(MVaEMqh)enQFcH62B;^^zF_|j<{K1U>u(2R1A
zB1mjroPYBUMPMQj#Ee?bu<}?53F1Iu%u-1lPb-BqU8wb|q!(-C&l9;ZUJSBXINE@D
zlm4O?9ZgmtR8sfC_FG6wgV5C5fY=;$Mb@jU)l6A#L|a*>H&CJ))>DXFP!Hx`B(Lv*
zGaoKbAMuZgTE_otgzeMx{3M#>EQQeVbS7X^->oj0=>l^~_&2CqO9!d}876$_TdUI9
zu{I+6iPFg<$2s~tpTi&?_Kd36*B*539{&WTYpR&U9vDB{ZYK>$)|qrL6t}KC1t(3o
zZn0TTL%{o;Q+1-#E(^tF0N8JobJTK8DhmXfYYEf&-CxZObF5XfnA*lb9TMbe(mKYz
z>DsAW#!Z|?*QT+4dm~jFh>q_jyX8FL64Qg-54}g$0e3+L&-jq*tU>1(vX-siT@)ap
zbBe%x64ft#&tl%?De#8gleZ1pMAf%Io7LX`>L^F_;yY`HJ)lfeg$$J(wX3>fKfFki
zsR?Iult|_{3DV1aiagwrfgH7a&t=TMDSCEn#N6k4-j!PNuwomx4zK!Z&C6^c@N{vh
zTE|(<4-PEwwMB%g&0l=+qE=Ddysw@86rjiZU(Ys=_IF2_QIP{oW%uVrOo}aV;PCz6
zFNVhTZpKH)+vhEbfcOK>v&#+vc`^!;A+O0Fc41ww$KR^0VPKiI(WIsM5sV0xOder^
z<V+nV&7oK1BF8jzDrragINvDVHG=9-M8*^{qc+G1V8|k*g#&SV(@BRWrXwij;28P5
z49RfG{o`Lcgh@ABM^OZ!>i*nqpG3Uy)(oNbzB8<yQ+yx;v53$~6PKxAgC;UT2Jd+@
zFwOfT_&nsHW>p!MoJq@$aPne@MMGK~joj7n%U6S@KW;pf@L_rbHivaNJ%xvM#**1M
zOXl{o6v<c~f#x)H3D#rOIa>Dv3kb~PCBZ3#*@7}n?p(f{mX@mBtoUTb*ajMSoC@-?
zp){o}th7aYVt)Cs3gZhUJ;l59`sN{mW_Yo~a$|`xn1IO#!#1VJF$!61hpuBR&wruo
z>GhYK9+?hi=MBP@F|!m|RwdR7rl89*J}V!YaJYRnfr|Jk|9}YMWVkAbC<JJE=bRE-
zC{{>nzX(yBEvpUF2kB+V)<$AX95!OtuH-dnAy3)_Js1aGPfgO!Y;@6ePYBd+DG^M=
z#1*%HTx_)YdAL~2s72afYLKeFMHSC%ZC&OS=_!~gc9fAE59eLjX>ApY4RmbM4nn$<
z9o$&JUjG63I8d}*t=15)JI4);-J^{JgGwSi7RHz#LTikMvPW<IR}<lU+Z>JxtK!@h
z*BTk=#6Umg0ccn2BGUn5M%%vLGg7#;kg+?NKI|dk=)2*c1%oeaf)U67(|8dnCt`vD
z0rt(!T27SO5X~=@r4%n&4m<QEZEQB}5rP<d+`3_(!(}v^EvA~ITgt)UWdd`B)_jgh
zcbbQWq#NcZ`Xe(~+sDu*r&Kiov@lT$I=L#%_&jG1>=y0I?PF^s!>+>`LBt<T+hrA&
zw}SxVIc=j+O6a}8pPd&|FX0~F2j_)?HWHm2t~+evhyEQP`1j13CBmn6MtXx(*NetO
zLqP!3ithGiz@WC@4T{OEKv<zyfq(eLfnT`e$7j%d?wHp~kj<cJmD!f!ds;s1bD>Rv
z8QXF@YVG=I)$Y8{m^d*DWI*y9lEnN9?EdCo9SyPZ5fZQkGpG0dArVTWJkSuMoj*vS
zp|~Xyql+jolJkJR&)PRWECgHZO$~>ODcEe|VDf#!9~*PxR3zgS81N9-hzE!j815V#
z*ilTD?yXsOZZgr4KJWpcNEUT-alr+0E;F>I{&iKk(TC5QOil|Gb#DzT=;Q(RhM;jr
zoa0q^pCc<If8BGVM3m#+4pORJGNyRWCgJrFJ6V#|uC}G;l{*$;b&woay+R?|T`&0h
zdq^nt(wdrx%sX#hn6n(4I!Y7kw0^IBs~D6%676R^h~u3nQ@|NnxqR-czMNO65lO&^
zCN@$LH+F3s(i2)L`8^o8n~K_<2^S|eR~g{n5||Kx%=Du!qwy=3(?y>IQ94=GOX$bD
z2gajg|HsJ055UUxXwiS&+!+2#qMn)gf1@N$X|zRKw82mPC%Y3fKoEh62O)dWghQJK
zo`oJAi7tZZ`puKN@7}g0M8!4Wr#_Dy(NwECfuE7{?XX{hYjcxh)3ZsO4buP>$?ElO
z2lVc@ptJiZTaf^$tKNV1FtfljlL;+V<Km8gB0oPdYa(%TelO);1kX$8n+>GdjSk{Y
za28fb$3MqWl9D^~kE?^kUFtEJ@EzUMI@gcabJzWC<)p@j6H>DfF`n#7dL54tz24E=
z7&Nn>Xe}|1MAU&SOM5Uxa5wx|c@Wp;5r_2Cl*Yp~|DlEUm}^QexTINjuG2N<#{2sg
z?zh8(;S}+EZn1cr*y^9W?q@fy!eV3yGw)XKM-Qf5HJBSYw9%$cag@JPdxeRL76WD!
z!(2KQKalgFNW3GsNH!kY?w`X&8ZVt-xk0V6iU?$6=&@OcXfi^`TVa;y>)P>*p}EOn
z++dmF8aY0z`tCN{+P6}twHDfQnaci4Ni^u@;>2eM=A23j(M<&;=X(h`@y`GFWVJm^
z->1X_{<CUq3F;sZqqso<vP?@-TLRQQH33<S%sC$6sdJ!<`R(N~V%s<BB&<A?>IQ9T
zqv9tS(Z|GV$$%OyRs>EC@ynLY)0Y6~F7gfXY{ouN=mO#9(Vy`>L>o048M$f=(oc>^
z9=7S+oiI1)!%`?VIXU=lp7Z=?-(JM=iz)(?RxP(mo~J)zXZNOXp5CDGRB*b744dKM
zQ}M1r9JwDneu6<_lNx(OpvyguGc{i)E=sHw>p90tw@x#OLxskLQnA%09k67>J{5CM
zv`+p@;vTf-mISn2F{oM;bXnk<^U;~us9XN$>H>23p&8lcvsF}y@v$fUp`MVxUBel?
z2-k+mamZ)EvQQxujNBq?JS6j)uxcEoC1@I{t@%^J{`QZD)tggD`20365-Jgc2=u*|
z-Dp)8evuyDAZ?ChHA(z>zP*4Q<w=O-B#rH~m}=%lBIWpSO@iZQie8SD8eo<2d`N)w
z<w!z3&=`n5&w^iUVW2)~NRjHkzt;Ty?j8`DlBJrZmQ9#uj#gu&ngR%k<HQZOXrk|5
z#N>+e<~0<s57o{_X=i?%uvAf;D6T!(5+XByVdpp&tM2@!?M3w)bL<Lfq}0voK(bn|
z*C?6jz#aaL;GN-qCD))E3hYRO_45R%G1(LZ010$jSy=>r3$XC!V67toNDa!Q;B}J)
z1Vid;eQ`I%r$431xun)kFScRN?mNlTw&%=7d&Mt97o_!`3topGO5q`EFIJ>BNZBo3
zB3^<ubDkQ`MQQs*&|+OT#h2Ja(U%AbCkEn_=8{D<i{NETvPGMK9W^o{btQikhvvP`
zBpm24mn~Fstgr+1hLe0@taU=73C{51ZCHd5sP7ZMXa(yIo{Jv?=250sywo_ty=|lo
z;|PYud4{hqb;6A_?g@r++RfZ{l3Gew@op@8%2!@l)D49x?;e_Id_?R9x4~a77ue-K
z&PvhpO<Mu*L3uaiDu~(hm-@0Q;4U?Vfk#8yYy!=cSkL&r4t0pmKT~$RD_IiD3i%r+
zNwroK{S2yQoyb7J6NnlSiS3HU{}otk5fTph;~Z0}ltUo9sG5xkY(|ors6Ba<21%GW
zYp~P$M#ZnYU~o*;%>ja#A-8Tz%pIi&WxGTawK6Iye>ml~;BK|V%(VQ#6;&$sM)ocA
z{CGa_;E@(bjR=}DN&9?xzaL!Teie>=e|Z2D4iGq0C7Gk9#TVC;JaQ#0b!6$)rrGeV
zpJR#{#*4GZ3#LK>jiArZ-jlEzJ{-1;TP2Rz)aFoG7Y)yKA|kTAI6xqqSY?B}ID3}2
z6;tsIxsC|64xHI76;BtaH0QAG3H?zc*jQp!*ZXLojJw*hk0b2tu4r?HAuV@gpxSFz
z)tikaWpjZAIQqJ7Cz=+yOkQcudOE^MY_w_{z2u5<kD9?;ok_MDxmlu{xzdY(2*<Im
z9pOxX>?^y)S~jdPMdXk&kZz6<y#hH}l)L;)52s;kQe~y|zHueH6*EFtL56y(t&y{?
z#cL1ySKa)V^&hAsh)ku`T2B=Sg%OmKn0&kSt%F4dT`>xAcrNc?K|U%^_qY12^<_}d
zheS&qZ7;|T7e0~^x}Fl~8+(01PyrIv#}nx_&%~;SiJ?3x6kHKa$~nTcBb^do9g0LM
z0MX{)nZKkJF4inJ<i0a!xWMq%b76#Kvf4WB*?<#89}uJQW@FGw3gTY5akudlCcxGi
zp=ve*!(W~(ftj{IHbkL8$CJ+~x>m0T&t;5oivXI%lLo8xO2Q-RTMf=Bm%iC#TU!{|
zs=Ic3QN+29MFx;hPL=zXYe^q}0PBLX#S~le8uqZ4=|?hEj3BP1as0fUaZ)EOge0b<
z{Y+`#@#;_+XmxZe=PDqvDjPI4I9ux=WvT}n6A-|s4k2%N2B{Gu`4kR_pnoBI8cu=e
zC)Uxl6R*M)Yd3!Q-LG=AL$M5Ae8q$-8crvG#UV;bMB<fWPfivjgX$Nn^+AM{=b}!k
zP7&Rgte)b^1GfqXDN?f3L5JOqtlnzQq$<Yk3U%5=DC{$Nkmlv(X3-|MQ%V^cz_qyS
zr>oA9Ne7`IMx_>rSe@NF8oVIhIH7wAFzhcNp?rBMb_U;n_#*ACQ&EMEXY9n~h^r~*
zKB%AZiGx(g<<+}>L{`CbdkA*jF9+0%o7~u5_b_pMwhjSa>#Q%*@&djJs#NF@`>@ZR
z2}TREE=lnJQ6<zgccAPyn=k#OVgdb&<yyPBt8Ouhvo@|`>e|}Xxa`y+Lv#WS<swWa
zoNEKbC@QZ_ticoCrDOo$L6<vk<@6(UV!-B2{4`=KSl{psVZO4jd=J$^JYKQ$s?!2Z
zfhjy|dwc_{iFvA!rv(qVv~Xrx+-jfq&i(NJBi|nPB}n7a7wX`s*^N;ZsfUr9BUPk&
zADKx8=ge@E6Dxh=iluH161>O)+rFrA&Q?o@PyPi}(;FpBx46%QqB?+2ZHp5-+SkxT
z&_g+*==xP}(t`L!Ib+?%d9ik7S>vDO(0fi<zYn>2z?+ZTbpdJ7qJSY2igU~PFKh3J
zuSiOkJtaLiR!!kx3LQU7zvR~Gve9rl`5>ZiVV|2S#bkEjlKv`ErmqAKqyTCk&C~tu
zUnM(Kmbh^i`mxVncC9}~b7noE<?7~}rL&iT6RL8G_FyUy8wV+7m!Hf=7(L*4g>qCc
z7;IhE1v%f+V58a@G>J|ZDmr6ol#*9P5ra`mkN>naO$Esko^_$_lmBWOuU7Y%Jl>TK
z*F9)f1Affgf<vP8$5OVZ7=hwD#Y4g}_$Kz>@Qt!YNxtkHv>@{m#$2smx@K%*Ze~%`
zA|+!^7=_`{L}>twT=`%Z)DNCY&!N6&2@1iuOEjZuTexX_kaun7)l^0|87*BXAL0EZ
zEoHF%UFwOyU}lek&fp!|@D^1LcD|vt(txtW36w>LQtWI-9K@Wlw?A<67hhpXZkYkd
z_kKd}1%cUS2llmuEvIV!9)U|kx&o!_s5H**?8r~u;o;u-2BTBH&iSv4>#rM<7SBfC
z9Ey|k|HbXmvC^~tPfw1&q4Vfpja1ivF1zUi1Qo3Su)te59=@COry8*ev14X%X}^w*
zR`EiV3EPb{0JhSj;?~9YbGbyUZ;Zp+qyH2^|33%x&;L1~QHfYEFerU#F@j%i25e~L
z$p=<2bx4brJ=MOiUZq1!V52<%UI45?SvfU8tmiiK+hPd1CAyJ*n*Wbrf^$QppXzr4
zb`4>($Nu^KJ&`a8l;V*j@1~cD9x2pUk3@C=GaLEo=v-S$63=V$fqqC9t=$Airk95s
zcu-$4flkFJ&lA5&G)PXE;ZHgx$ov&U&f`)>^+?e(z`%KLv+Hr`1A#))oOwuQ*qZao
zj+-3?BeUz!?!R-&s`i+;zN*yLlK!ad7FK)sS@x_Zn*gdP26=?4!dzw|3qPZ_H2<;s
zW;ygebaA*Ax^(8)wXgwsNS(uCP;KYNa2t&S{<ibbs{7T5SvZTtNfC4>56iO}H6hb2
zO~?D6sQfwvl@?QN8d_XENHt^(&IODrVEi}?U`4scHiET~%IY%rQV{orsXQicf9vWA
znqy3L_%?+UH6%v+e7Ky!&y3P*)2$&A@?fxek}DUuE+}U12Cit2VL)%Ibf!*Gb39Lf
zhHioF$NZOzyU)9av4=e#BhB~7OUW(3f1>%a;;y}WBRK;tB|0B!9HJ=;=|~J>7y^sO
zUGRd@Dh>U&okI(#uqpTc_Ehkr3rjq7!|BOpmTK7h)jD)S&KhZ~)Na28J5uomW-1+E
zEVs5&R1KOzofh%ZI#IUxD+olPnIlR~px(7`D8Af!28rzI=2(4FrcP0tyoV5VQFy4^
zF9HmE#tFmH#Ly2IU9Nz)NdS157zsApI-(;%yzU7(es)3e0MN07z&KK&SK5uDfJoU{
zne+o!C8Pm*hNe^DrNwR>GnF9<afbW;QFs@;n<oImtX;!PKromJ^JYNfKe+%<f<L0X
zW-6Vixs4T^K(g!-+5ZxMT)A3f57vy`9Vu&`nZv0LKepFn5($?#ErUdbHdV*#67O67
z^a5s+q6~hq##T7uV`kF22{d6vJ(=wWbbq&qHjOI?&f26lIT;P{YBaO*#ZMt0136Wt
z71|XOY0V2MS7DIBv&hL{>SCs+!CGJ$nKudZ(<#vL=#5j)O)1aBMI$RwkZcR8I%ZeQ
zO^&$;svgchTgGz`P@3zT*17Pj!g1woE+tx_WE{H>K3Q5vbb;VTwQfV6IQEQl!Lt|`
z*sNi~uiQ_smZmi+|7B??t$sQG?N=BlH%fYFn7UM;XFgCTXhXvoJ+#+rEIp$DoN(#k
zVf_HV!UAls-JSs7D)({#wH_mHaqV%eZsxTkdRh!TVt)2O{F0em{BNw2z<XV6)?)l_
zw$(Rox1O{EY@)_8>b9I~sA9|{GYTNPuhgTtA;z18qy2IsBNl`VzB_NNqiW5qSF&!q
zX(HuMEz7$8oI_^fEq2wu1iBAv7MwxV{x}+kZm|X^;DC^Fa0Osy?cghT(M(MXzR<I=
zvL4fBc6o1mXDAxhs;gp<ZJXN;F-@!WuZV&ARP{BJ&+VaWmn#x~Dy)}27MeX3OcVDF
zM}6_nrVPfm4ZX};6;7EGEo(z-^JTggOiwpD+fixo4oHbmj`mE!y9kMJJHK{*z1O{F
zN6t!rX`B%J!#qokl9TRhhV4dPR3~`gzC`kvRST1kB<B6YP9f*%qAfLd>TYnnl4W42
zJta|yl50gX1r$+1*zW>v`VQY>OBl4dZ$e*bOL!DWxBAjO3bE3ffDSq0%AY=N{XEz>
zLREv)N1=zUDdW<5i5fV@LG8XJX1qSCI)FGtXU%%W4mq-PKj*@;G44mgU}|KXkTyxJ
z|LbDI(ngurkjC1W*+B8La?}Izg~`~0*;DNXu>D%p4bmf{NqmrKzoLy44yS?kfQ^P#
zaYTf3&qP}IRQ9CQaHiDIpm#8bj6L?>vfrIh(0iUBtL^e=k(?gg(baKGx6|>F#go5n
z$(l;`GfK|Zi(uB9LU$xIVqaE_JVJcTNbkNhBNe!q@<KXVt^G6*khGT@S>RiU3{;Xx
zbiY*RM+Bp>m=Gfup|8}!&H;!{NWGvDVA7FJkp9bUv6a6?8no1|Q?|=X^v`|M3+Y_p
ztD^$#pZ|P(goJA)L|{pqZW7>n3XsBqaM)DOX=PB=X(X9-#b#uMIIgg3rKD@+*or)6
z06)cLS`_q%Jxh<j|E=3hD<IkNSkR+DR~N6>drS$v`H+dxf`LgXeBQO{#T#0^CG`X-
zf81=gbFYlD!WdLW3uy!SE%K&eu>`B*k}=c=K#Ge{mv(N7iOF|p^V=-Fcu(Hbd^U-B
zj*HJxD~qcBBF5LFM~hR;<W=lIq6Gk-&`W(U$b<V5)}a2g=~T16t{OviyK%V+JHrlM
z`BxVH0j??87xjM~xy=9Nt%#Q8f6G!$sY}=#vcmm3nip|RqeSD}0r&v3&I|pzuJBrh
zvQZd3(g&u0cqpG5DS5_~${b(%?1Yj@yT@djxxZctWp;A3w{yHb7#!`#8CW^He%yfE
za*t?C(HlMO0md;{yMCPBR~nX0kKE%T8kvQuNqV|A-#&+YQ_%4kc!dF{b+G&*(a`DW
z!PTzSsS|gqBfFBGl$SlEkjNmcV#Dx0dXGMDUVJ#u`fU&3qVzGPm*Ws{$X)D6?(f1k
zS<&(HgzCX35Qilt@I7Kkszj-6Rx`g*&L`$}dLo=TE6qa!=KCT)O0*1pn2!*|)C!GL
za^RCYM1@h?_`Oj^#**J;PY3PTML>)3Sri77*Jj7@TI1;Ggf+i5Fh7*+%GuY*XE!6U
zh1|x(k;vjj<<jJaETM;BU&@f8RL<n+Vk!FnAI9D>xRa>;7LBcmolN|TZ9AFRwr!ge
z+qP}n_QbaB<mNqfPn}zJ-dpwW>gw+5{?H$G)!u8bXFY4lLewvbAJfUJw3)VRhPR9}
z#yHt3S}c|ReVA`MAUy+YaTQ_c7r+l3NZY4zFC#S(MQXwi-9HvZ>)jf8uPP`063Ben
zB$Pu9vQ|_$@UOR3Ha;}_lO5ba$%Kz)(L+BZcoE3BAZ6mJEa*_moNoWU7q1Z}H~eQ3
zt9fgD6C#R(NHym|#MqfL(Tr6eJxPryA!6e7u~VTQKQtG_uBPIEb;!G5OTnwdxl4XY
zb4*OWhhe>lBgKfXRJ2-%mkGws7JH1OhV!$VEnXSngYbK$<tPku+A15?Q|^h9Xptww
zI56rgs^xmXx8_>*Ta(1GIo05u&zENi60EWkmRwjrSjP9;{_XxWhj_`L=QbNg(*fC<
zcq}bJZ73dMs6>hW<z+FeVT5Q7H7k}^+BZIBM${+A>(?BWrMVGm_mJgL7XzbF+j3dB
zad|@=Dw9A>8E8~ODf2!pgZzC_p%-6bw(}5CLO|{8xh3rc1Dj+%g?V7lGN&BK@JeFZ
z@pdJ3L12DuEoCQ7;*yFV#^=A_9U7Xg>A<=1HgMS5?k@eYaM<(V?r^v8^+MeAy-FAM
z+m<CudzT^svpy3!4lZXIFPSdYqAbTljup&Sg?M@3XfUe#iwscf`h%dN(r}JoT;mwa
zRj)-Px|hxhNA~3idAVdg>FVmnl6v5yqKl+8Rn-a4JqqJ|m75IE3k>gDo5En3DW4S+
zgKB5k=m=OkEhzCu8NRAl{HJfJKSWec5qNxc!FhVhnI68dLNaHW_k0z>{?$QeV`%dq
zwnOrEjKVOWlO$yy9fm#0=yW4%*cdeYfigaFnugpG9!!Rs`<V`$+?`dw4Tkh#!aFW%
zO@cO-wVH%1ov{r`EbR7TLK0c3ie%i7Ex#;VuaiAsRr2FD#NCSo&Zwe<_geQrSqM6#
zhHQ=WyN-mA*qj=hKef<awMd2*E%6s@&n3|a8m61)kv%ZFh4SKSN$L=ko>46VK_nCh
z!Vv+e4sY0Q|7OkPzrF7A$i`jz_}vkvmsZ|hXrZlEfeZEeJIhkTAEfUZltpXL`R!LA
zNa2pCcnv_uq|x>I29n{!hmIPqSp>It`_J2(Stv%}x7lE9hh%hNwa)j;S;14{XZm&O
zd+;jKpcFnwR}1Q}{fV6RX>YYyZle^jtvj?+^igI+*zx-f*my2tF-@~X#JMI#pAPI}
zWr{^TL)_2ZP<<hk8VOk;tN;n33ijtw4~L-ql5gR>X~bD2-CKt;1bIdA4o_izEYj41
zxx768o|QDz`lqH$mAK{2hOttH<)LGpCRLAptx+*PbWRK3oxyq68vkt4UE3elDo{51
zod{G=1zX>@%HyE9k;qL;JWKC-G9b`XttnaQ)vb6~HhA+&!c^jw4L*|(i;ae&AZ;B8
zpAp^$43irVSNs-;=lJk*b&WwH@y6Z{32wptT`M207tGIP9I$qNe}Na|%S&;`dHlDK
zIR=^FStZdW%T1tnxz|E2-81C)X(AMoW0nss5&PG2+&d5jOL}3zTn&O`P1dmvhcJVr
zC$7<_fF8=B&1{U`f<{>w;z&A$UARqXy6W~=6{~k0-fu+25Khn~P&rezTcA>=yINT1
zkTE{|Gr~ZFK}f+Y|2^PRrb;)UMJ)G!p_~I3vO%1J8hdLW0dl507YVu3wG-q*S@ta>
zLRoX!1Oc3&b?~Kun2AddNVTAOGcT$-*xn4dN1w|!h-O`leau$&LR8>7%%nn&APjNJ
zR-hg^fk=rM&Xds8I;~dB4oHX>?nd8)Aqi6S0?~KdO0rD(8<l0%7VVyzO0lXTzohcM
zpF+?**FGdd?dg55o1^9%9lm@N<3xw(E!sh0V%(&KZzotVX&8N;>eDjbXRJ9Zn!4wU
zMG>-V$X5#LNQZ(yf`}S^6FkjNLJ2a`UpXlKj7hMAvly%*K^)ihyOQ4l>R$>_$2s!3
z^Ix@G><;L=wVo^KE0fb)XKjB2zkW6)XEqzE7GV_gkpAx6rUrZR^&`u-cU@~fqu_G-
z!Ext1T5bxW>D6<5F$IpIa-y+EFu^}R?=py%W;0Z)Hfw$HImzbZ!fa_R+5uA#0NN7Y
zT>-mB@?5rqb?bRDr~k2#7*g%N6h_qM(QxzdT|TKRo`u*W2g{kJ^o~04$X$*G=e#VU
zc>dhQH#{$+g!jCFKN6;jI=hOX?-=H>FZpva?akBdS{y~xq^yjh&zVPw90a%}YlH1B
z-QGIt#LRwQX|`tLCANApf_!X@HO0?d`a()<b-FL@PI=rMn`rr@uUvGJLgqV9m@tCd
z_GYfPncqa1ZRigsbj^N;3uXAvWDnWxTB7#fum&45$N$;h|NmR<GqC*cy+0%Kk2(MU
zI_N~rt(=S<=tQmbos5Nz4Q-8#d3j+Rog9qytzq2OmNYeOwOW4pJl7D;raTC6XJSpq
zQp16^{l{q^e(%R=AIhJk2DvC&MXd_-c>5U#|MX%!J|jv6!3&>c%HcTH@#qa+A}%pg
z5Qanq$0tKE8T44j9HzfX5w+2xgdG)oghVNlD@mxLO>yTpW!8n_C<-})Uca{kx16Rg
zrC~CexBY-?l{h)6UB3M*O#Os13}o>#I0a_XqJ(1XvMLj-l|eT)L~C@XN?>d>YzrYr
zRKjl4<Qp<6B%o?Wj#)g3R-GZXs7DSUu=XMt%&(EslI+*^W)UTH#4NMSpIn`hE~s)E
zqYTs5I4#FvHvg5b9H4+F<~W-<A`f3~jB7pH`Y@NwkA79rhY-cg?8}eCyolMaCaG;F
zWd`WjVL8-WbXEj_@6QLDBTI=UvC{uDlc1Q`_$Q-50ch%0@h@^z)u{lEo=-8O;lFO8
zixafp&202T8<_*CcObVc9n@Zf8y~*jWtCb42dy{=7@|kESf-_2E*q~(z%5@LjhGwu
zX4somXt5xLmNvtk+UJxpkmwX4*EEB_-=l(SwrMz9C?xTQc5QN^q1!(l+-xa-_YSys
zO3yl?#pCK({;G^2F+6Q#-|Q-HuOP?RGCwvS^xiybbi&%aHGl9sBW7;rM~lhj9^b!(
zdcI_AHI0v^d^jnLf8)mYKcKtjUOHu*4r@vs+^-=T4l8=(V0u~g*id1kZISo(cCP=@
z*^C8w)x4tn@z-3u(Cr6(tN)TavX{i5*Th&LR;QjzD3KkwdA)mdc-*^tary<OezZ>S
z<=Cl1r-I=>665RJ@pU)!=GFzpjKHm{Q<=kym)#+nD?osot$MZ>L$I;UtnIt+bAk|p
zIFk*(U71`$hkl+iz^2Ph(5>02O~B&`!4Y!p)1jqzNPvNKV$jv9ru!H~KYe?-XYp#{
z+}_$g&CdJuxVQ9rYdX<Fmo>Cw^Y@N;vbMkB%Y&2m?R049&ZC+;{N~^=mE}`=xYAwd
zR(MSKFl;`oK8#WZPx|h8sQ+fw;nAY*ETE`hmta73;P?dy9B)@zW`!<{gOU+Y?`#yr
zzClhT;<o%3>Z*WzAcBqDGUw*zW3OkRWZ|LY{Qhw0#^B{^z2@UMA!qq~qM`?WAm;cy
zXDf4X;gD`%#@45e+pX!v`Q^>+bAgX;MOPrLf=P|->%)WHy^9-5Cazi*%DmJ_`y?$Z
zbIGAN!l|C#(|LTv!-H)2!J<_o3t_uvMmMKs219jVH6x<?1C8u$;<qCgyRgnT)kB+c
zZ?&K1{B{jIHecjgDQ?x?WrA>sW&f?=j~SMPa3UUj4usUV!HdJ|ro#)7SNyMJ3SM;@
zG|?pUD#38?;8s?N5NNQ~IK>bNI6-+5Yu{)jDYP#+^4~}cbw)llNRgSkavEylvHxJv
z)^(xeqe3$6iG@v|fXTgckVXSx#AL=8E7YWEwuuL?1b)Mi{zFAE-~fr8{7uGq+oylz
zqAPXr^+j(I1Rw{mU8D#;&IJPD<z^>AJ^obk6GTI6JrD&?1b|#m3W945D$9+Fg6nqT
z=5?eL_<p|gAqWs5DN^(*V%X~R@Oltv0^{_WT>h$stuPA%H`9Yuab*HVU#;~ykjov_
zrjR?)m&;wGq0m#*Zz-$QOe^J-zYL)?uc^0{^M0z{UF(vT1l##w-+WgU@7DUP<$m8q
z<?YtN%Nf8gsKDR{vS`$;q(Vzfi(*7QgwP)6p4Yx}5+ivpBLsbsTP+3BwhYp9loO~L
zb!oa%zXQ_YoKCagl|P_vkmH|#45him0*eeMY*fAvBj!Bepa9AOpz<I7!uH=_M&C{Y
z`g)d%Md|EigUbg7+5+ww%jL|yNpBz%!n_OO@|>gMM&nrwdk?=vACXpPmiRSwM_Vb0
zuI_38P6Se%^oqK!agDxifk@RH)e>%<7H*csdBf?v<9yb8zUaSF23pguyIK@^Wl?eA
za5rp%pA<4hC4IybK|?v=QY1g$ri8r>>ZrH*bU{;iAZ}gJQwh+ZYA9uafmf=EiL;1H
zh55~(9#)Rc*w{A??ER-u<6ExYTdq$rfY`>R8#5{*N&Ui$MI>drkvks^DBloyUF~-t
zd6l;vRWHb#6AG-7cZ4Z=-Rz`&N|S5*g@(&ny=|Q~hK~bYqqf@8cX5`Q2<efMtG<Yi
zsCzyi3h{@$nq0z`Ke&%lBPg|<_htaN&rXv+xHaYTf6lXs^~IEyGCEsD-7OnJ5$X5G
zLrQr4R)O7gmhwr%V;WK;Qpw4ffPvkxT0#f72l-r#!m`M2r7pV99d$s2@cJ6?^a<$S
z3GPY4?)Q{J-uxd=ECIiL+etz+k2yr$0udV4aZRq<QRBk<+Q3J<+WKpq&Ox6=Ur@!J
z@UlgYQvN?Y?2D8!_WUI*WIeupXt`6n5@GX*A9FyeZRda65`zYzQ#W&~@|y`H^H94y
zq$$9?pSeBIua19D0=7c{KG*x-w$fIPQnjGBw)|mSwJ_8P)Ytvv{rn#veflHIHiJHx
zL02}?FQ2iH`bfp_5Vuq`Z0vg`IQ?{u8*&*8X{Frq&(_l1^pjFch#$p2IbIzsz9;ne
zicr2W=AI<S6wC<|P-C{l^sB@sT+U7JJk6F&W_hJe-eyefCsr9%c0=2f%!pKf+tBU|
zmowq$9x_5QF~{hh(XpIQvNfSja>BPmk&1?k+furD<R~;J+siny+#T>$9VG*>U@qH~
zDYWM_O&RI`-cnB+!IV+$WOV`ejJ6JJ^;$QdkHeB%nUkyBMpX?GIAhE_JKcWt(q^~F
z%zwfh@qV@}Vtbn!IkXk`rVyMq#Y`;Sr5t+P4%|7^ak_Abu7DmjyM9FbVqDzvKvn0T
zHNIwIzS5j<$&fzKm&Dg!DlI%kU$k^3Fx@zAZBUM<6CcJbc8RuVY_KO4@!+U4wJt0J
zJMOCC_<mf=j0u1jeb}&KYymPRHV4$6K^u)aYTuL>lPA3L=~Fu7Qy=x^OV_9376C9N
zZj<KuKSpn|)(XiybP~T`$=58zD3X~WV_141B!MbCvzNkk1OIVW1cM|x<G<qzLw<SA
zG^|;nR@fL@Rme9^34<>w_j+!@J4zX8rl8}Z$Y+dQr5LIDUwCpwqDJ@Jg7mrVrMdY(
z@z08_rpBU2e3!)cg99a~qi`y<hU9UXyUrMwvnt9|u~=J}O=N7g6CQr&J#h6pMHo5?
zP>M0E2P?-9I$2uYc2!1;{Wa$eUcU10c;auhT;W;#26Dh(75#7ckDcwm^zAXQ{m&rY
z57*@QpY+yq)f+Q{AQ86q+fs_kc(V!RRYk3KYX`}C%_e2f(QKD*Z|oUh`hnko2^R!Z
zfE^I?+wJ2D{Ez29r1y){^Mk|lKto=jVs_u|?%^@$D2B#YK(fFDBxG)-|G<Q8Xbyg(
zC{+!)d))Hy=){;Q-L+RI+s+|$Ecp-P6#X2{FB%z%=?<a3)zm4++rjvxb)=bg%_6TQ
zunggjBy;=uGLzD?c}i5P(l;D)YyEY-Ke6e1o&I8mx^7o{fojJ7ao*DL0uD@d-VZDf
zo;z^;&GF&&{OrEezVeWrcqouqRAr2x7_7qZ9@r#`bzEM5BCq4Xo;(nbn=TAYeT_T9
z|CM5oerP6N>Xp<TVQ7!Qa~wz&PqUV@?aLuWoWn<Bicm;hUFe5+%Tii5&2Np0$u?sq
z%DG7m8Xhnp3G#0XuL(r{VEC$ltUe>Q71Vpu@Xh5iw~4a0O^{0?Y~QSCAA8h&Nncpy
zbC=_UMwWjhoVt1%m4@c7A?yyZBkLXjS88S~^_7Ydp+hX$+QKP{$+uaF0Ik9i)77Ug
z<>au#=7eq-An01vff*mmN(GuuNQpM6M@uWnF`K_GtlGDLYz4q1{pXbAOm#IwWDtz{
zFrH@Y%S*Oc?w=iy5d#gYPLI?TwXN`RfZuK^mBcWnu%Xvg0cp>X>KYXrXL7#-*&UfV
zp1w{l0A@TpX<3{L8<<%zk)0s#rwy(#W~M@}{oxwN;dx)(DGIB#APGJrz4*65))<tR
zh+7MF6;omEeT1|@^x62g%AVJ~aJfu#Pld=~zs0&$^QSc)U<)brE{q|k^@JFo_%Lmx
zIN@!I$aG1YL5}tQ=Az9C?GZnJg06jfdy!={ITK@(oSFiEvyAYWvb2}-7-veH$vF=0
z{>&T=!#sp)di<waj}X*9+vx&4(V%}7Fu!Fj>7>s`*Q{>alJ#$9m@GFpBTKa`wR8}B
z7}n^AlPex9{IMKAV$0#n!e}YY4$Q4BEgQGwYH(ydqBOpA6h%jCe_09u<>00gE0lA(
z<L0m&IvX)D>1jnqLyrCC+oG#h&8gm%1IpDc;FP1-ebO?<>IHq9BSFKMXAoaLQZ~<C
zE2SwdyRR?~PS3LPRKRfWux+anT7sz!!qBqoDfb4(GCI3=+$7ItIyysDu^EffawIl|
zFT&MehKNmjGYBT;eh4`Mr`~VA(OxM!7`ftCiVQ;|xQhp8{prn7Ah6g;Ayr-d0I?c&
z%atIHnVB-A$6e)iB3uzt{nawG71aa@s}B;^XRGL08`k>^qO*M!K~q$`xy>rHU!wWa
z5k(&3pJ*57tPU2+EyM;l@p8u4Jr`JqNHif7wGGPzu&oNUIBNOMfaDc`(p+MH1r<Rd
zSuBK+=1n9qB>+YFr2;^=3s-5QXZEhH1eHuuYd(VmN+3W{?LVb#l{m}&ps*BYpJFIO
zAba^L8GGNb$a^`g>N)Pp+*%xBbo2s7gKMbHrwO;}8F>Yuvn>pgmTAV`lz0pYtVbqn
zz|%ZUuEY6rI@tW@PorJ{N5+PR<lmU|smbJ$6~J&5n&xzskcd;0BUvZ|k8fFr*H2jc
zGOT4iTWv9iEDE%?(-GI<X<@OV*q;t=2Dg&S1brQ342TqxRcQj8Hp72pB`J?Go^#Ui
zLh$>A{kDnA5}55nwo3{mH{+0OZqHLD>-Fl(?-f%n%UlXfC^pyNNm}e#&r$1+qL3o`
zMN4$kt!*(ZRc|7I6}R{XtxaRfr0MqgcB{tCX@tKvo77V04w(~OjF^+$R02k~O<MHu
zMC2zOvnCEO%js$H|5)abyq&U56)Jv<1>@N_h)Vp?^ML7}We8o^6E#aA{e#z%^9&Oj
z{aLp#M2inv)?3RcQ_SueZgk7gvU~tOYy1#CI^EE15Y58Mcw?t%DoE&9!paH;WA~ds
zis~;{{PrbVqt1ZMed!f3dF{>jJdi{|678aM2<s~~hU50ES*{@trmlq`rgaj+cEpAj
zQT<6W2**@O@W#JtSFhqSAhOMdDAS97q4r`si-bHh(sL{G@QuG&R<`|Y5>dxE$tO@V
z21btsC7KgVi%3o?1#i3&-#Y6{7QNN|nlnfb$EG@NFWJm@s=z)wXV1NigtMrk9iQ@J
ziPP^S(|lUR*&VOFNkjPsxO8mBV?C<v`m9(!#GF^;`C1Ov8~)`Ex{A=-21wWFF8~jU
zBh><FyLC9Q>v%@!kbR~`{NrPYl%7#*b>3YF;tAJLUQ_7IMe$7o`v@;XXV?tw>P{Dt
zu8)0}@yi+2Y^%vf#A%TWeo5PHrpgF&EV1i8YbW}mbIuaW@~Q#@R*40B!C5N!37)pd
zW35s6NM3I6rlaIlH2MtJEk^ClLI9fKvOe7Sn{Co5vF8S#2`ZtW|1ade%5wvxf$G@C
zFKQrZCA?^vvt)^-y)vKex0m;_Ou$csk-2vDx{K|^zIIDijCM^>S>C7*b^YvLzkAA0
zk(brARhNSemz9m`l8ICLMQ0NgcN?KS&hNV-Mt3QiYIRscL5>V@!H+qVnTB=gy)=t5
z&P+Y=kDm~V)}R$Vd_I|?t1?H@JM9AQZ?PJ&d9pB*qCOi8rRp{d6PWd@vuf8BMVhjZ
zjDK&AVB3#j=zr#w@YT7YY+FtZx(vN_QbQMKK~@#IOZGe8pWnHK?mmv-Y9jmehxvq4
zpF7sUP!F(MUiayoVQi(tdg>T%M*=q!30z5;2ziaCsIfYG#E!Q<n<j%VH)jF32#<2q
z>38BwdDPVbOxee64iFKN?~@tZpXxCtY~+{}e6kuQJ2CHTgenIEcFnrJn@er_a@b<d
zA9X!pMumEM3b|WsRdsbE;sb?z&n<Be;KMdq#?#;ZbYGAF?R&2O3S~H0{|}WVWo%>W
zWJbuy{y!|6EouI-#r<DU=3!1$FJ0M-cpNC+3hQ6jg^kzx1t1JVeUQJNKKA1SHD^kU
z%~KS6byNUbnk_=PlG5jqa41C=quZ;s?c44CNq%&QzM^;M?sfs>Az;DwKSJvvAbJ|T
z*BAHK*JQ`V2S-`Hc*ox-BeUDdyi2E1K!1q+M1gp~*}E8!Pr`D^1BP?+Wo7upeKQ^|
z80IYVTeKwkK&RlDk?=pZOLTdCTn7?*ML1EV%v;QevBqMH6kRj_#LLV_N@gM1es1fB
zw8UUzjAV+^WH?h!_p+jNy<MM+*A-wHr(wGPiB~xtoaiSjVukygpOU^@PCo_oU?`*b
zB}2clLr2^+xoFfG^+;q_%!#Gr&8JOIWGb3LPK`PDi;0C=jT~hPBV2(V6jb+G4SC4g
zsoU)8!b`v#GZJyl_xLMHy=ZE?k-*ghR9}iwAeKT>RC1EoI1yuOYUxb8>Om5ADzaX_
zs^{~`<H+NLIWU>XbBh(F&mW82H5Ic#Lg?v8riIye-0Ff1dQ}sYR)eyhOElgWJpP%$
z_Dh@f_oZ)fB+R~9{9#Rwt<P65oo_+PHcck9Y07s{UJix?q2S%^Va~+Qx5(CpszgR}
zn(LBRWb@Mg=mR>Vm9nSc2cmYFZ(pxZh{mTJMIp2K*umtd_qNz{xj>yq!G2sz#%X_j
z(N?M*Ef#tCbKYt5n&o8Irb`YXW19FX(oeO#!8@+qhnI&EL`5l7nFlT5t;j)XUb5=f
z4rxtf3yM71m^?CIKl8m|-a#Z)mdZH7VMRoh9k3ujQgnz;vRovc?{bKO8(|ZAduUNB
ztq4ZrARXZ0n1h(FsI6N1jiWUE1Xb{|2jPwC^uKGA`?)9N6LvX9c2O6MGyA4bH<&8<
z3E)P?Nv>!{qWGQlW8U!whsQ9&x)Y&=E!ja^yt-qdg$;16Cv9Sy<<^!0Xg(B5X_)2B
zUeSyE_-l}xxS5@cs}1<^@D0{h=>+jj#26?XC3YNOayb!YR;Psxdn1gm??@I@udbkl
zDv@v+bjS3(*m?o1rb0|Hv1V|@i0k?dQIlq9b^ggF_`m1o3@C^dqI>_eeL{_yXF?62
z>VJ{&6EWUZv+~+N9&;C&Bx;LH^G$-{sSUSnsTAbJ_GbQ4X~F_B))@o^7!}F|r-h~>
z#gb`bf5(_x{Sx;ydAJJduqBXCPcSPTWr(C$=N2gPAY4g&YIaz>p*e6i0B0x~XHU}l
zceI$s3Y0w!Jt`?-(uykxtjbX^2$n_0ciFe)i!3e$3-QOszxkL<8X}-E*zqjKBsrDM
z<L!OYOS4Sw%y|Rk@w2=P2-aBH(oT}QAOuA?QJ!UXkem}d98NW+LPaIZdX3h?ali%G
z!M(21y77B<GJ_p5C#`xQ(RmaXecvm_JVeQSv_nIXuNh)Lm#qW-1{vVSa+>7CqUDM<
z9k->oa;(38JduH{VI{?y5B9u7&7sDtblNYf7PhqHV%(IVA^$ZES6!CGH8dM0@stUK
zmKJce@SzUb_r-wKT~Q>39%*ZIPjd<|`<%X)B>rf4?_a6<`n8g@9fWtZ+%}=B4jOy3
z{3@MgmAKiXX#8@PHlMD+7FrZrr~suIzpo%JXD!HLAoNI@BIp<rZUA=T|9kSBTo8K&
z`*VpK1BntM{Et%}Ph{F3X94H*`YZ@UCp{7&$2tV+5<L#Nyv&GWzekG~#$%y27&+Rx
z*bmOA=|*78QGG>~b*eys0b=oOm*h@u`^P}5_WtlNHt$}G!kmrZtrQOol+0qT((J<H
zyo~Yknok$7f+MY054`5!Ov_J)SJ@eWsb&QFiGvvh_k~HW>Ws}?+WurZ*Kb~ARbgrq
z3)Q%S_w)qw=gF+a|1=Dt6U*$_Y`yAqLW^^ab>2Dhi~e};uaVq_zbue7PPw`VY$y*B
z4?`%-g3_6{@zSDHhd+a}&*7$zG5Xeeyi?^^xXl^YTnOc>P1Oc0`6Of7I*SyZa)|TJ
z8uVU#II?c2sze=Pw3#p)F}8c^aI!prCT#ocOT>e5(le2$NFE2Eu-BTPkw3KQ45oeV
zOw-3jSwWVa%2H3`F2mzNHGFM(O-iP?FfRb3n~%Wio;NRGKY@Fey)Hv;w%|%FVi^z+
zt^mK4#q~>oQu|bxC7bi@N*gV2_|yCe5aZ><<eXjD<FLt-c?U;ZX$wYy7?5c`!r;=w
zx;@!msvfcS%%qLdS5IC+^mr7e!Fio9&yd)Ry3#4a5!W?$kngIkxf2_KpSAGWjp!P%
z{;9IDuyK#-nXib5wj~7EX+H1pr0!y-<Vb9-a{2U2O**7N@MM}B3BA%>_#QLK3y4*h
z>jpm170(r1Zv_Wq{GqVlw^L-<_?{NXKp%7R*ibwDEUT2x!8sq&3i*2LRR7%@eb8RV
z5oJz`U!6%wbke=23)1K=>-J@2kn<_?iVcCb)aSr&CG9u&{b9L;C-;Ey#m1BC5Y=s4
zLpttwQD}?1IO;`ExRFki2T3r4$Rd@>5j$=4OlL=vonQIs4p`h}>|!__xn{otP5<6t
zuC`OMG`F|o_KwQ~a5T<8FwMYO2SN5^3sx%M6E+oTZ#!3AEEFoIL+`#}mDr>$uP7u=
zARwkp=)o-j1;W)%Z3_Zj+dRxQBY9j(R)1DLbY99oAQ49vqgBMCE9x!bh6Yw(lCU5G
zee$l!-d|!R6{7Zc6PiK4Ijw0!H@2@%164)u;rOf24dFI=WO;MVrUNG<btn4(cUq8b
z(bvsP{!PEPGr%IZ&JN0)x}7r1`dzK>&}yn(@%Tp3gsWsz0yzY#;hbdMhWMjeso6=N
zU7L18X#5#Izh#*aP%~>WzA33=0V<$<gdE_i^{|7vICShk`0E=F>(vHcIn_0~`w0UU
z(8i|-WL<@Z35$g}I%EZt+6z9X2c0)hxY)ivoRf?j+3RtA&=4>XSko2p;KqK-kyN$H
zG0d+1*+skUSF|j5ALZF#Mkcf4UOZo0!HCFs%B{KN_Lo3Y@2CftV3IQp<w_y`IEORN
z9LX}s7w0aud(DI4ijO9jDBk@3u-<9WoKMRc1@n*DOSOtkCMS}2XwfAu70dI-IGmJE
zFaI3dF<MPz^ys9!5fc}{2zrFpzU=H)69UK(8ZMo%dC>Ke265l9Vsn5=EAbBfLj(D+
z7F=>IY<7(jC#9p>S5V%E61L{IK5HjBWLMgygF|9_`FoBWZ58oK{+4*roU;GnCKutp
zjz>4Xj5q8l#jqNBwSpz(EkFzXIuYn)vlznL2w#t>K^iYmd8t`61qD&|ja_+MojvZl
zJF{l3{W5Yc*q^1M4#;iyb-<gq_sA10+`mJkjEgRSoJc&=$zCo{$~%3nZx<v5%)Nw{
zBl{2lGE2`+mr{|!4XU|?jL(bQ*USX9Oh2_*9&RGnhS-e3s;vc`VZJ3AJptAZZ5CF4
zlUhDk%hj(Tl}wf1m!zXt>cwlXG2YSs>Wd?-RI&R{_#FOD1Hml4d8uE;!m5jnm$)w2
zyf9S}NoQojwNg_yI@Glf*Gf5;p63OrhQK;jFXNNA#h|BK)Vz({5feERG|nlo*7fy6
zFK9-TF)O>mO#BFVMMP@NUBEctO9}OqOxzf*ORxF!z`c=!vehTVlTEP2L<#BT)`~0|
zcn^p=DN560cK9Y>`KPFLNWa5@Y5i&a_01I&LV{CsawJaNnp<kI_K3gB6f{^s+J?;j
z?)%{|JLOHwd_G%gW>(Vc)z(g;U`7OKd&uQyOnF4Jl5%=(ERe=O@-3!Ua>zkK9OO%4
zW2S=?SC)!IK?i(rmB9bv8+1~;_1~~K2gCmze96H4!!`a-D#w@V6E^3pNSps@chqMe
z*oB3H`&W&Z+kw7{rj6F6y(x12O~>ganUC+XH#*dHDHtauf51JRnWk^_tV-UovUzL$
z>?OabzRRq3vDNL=OGokzJ{BnIiq4M-zPY|~)q(+-yKz2<C{zlko0y5d_V_OB{eP{>
z-^u^<5{HYS&InAoj$zz2Oc|zpbnRYek-$*h{Ku+%FlFoe>*-#`#6O2_txCd@8ala5
za-zd*OZpyRumj~<lr{>bMus{K!BJ}C@O4km+1VOQ*Z46PR$jwDT7s@>-u7UL<-oPG
z0HNcZnePI77$_dzDekBq@dzfynWB@`{OocZ?s$ZBg>S5VW}p8T)OFt6rycYc2u`Km
zb8g`J1E4HN>5mP{7T8Siy5;@(NAtU<Dov!M**;fbJk4c0U-wl%EC@~HBVPUu|7Htc
z2VIBmnlcmfn&+U)LkY9!c^uyy)D%CX88}iG4vpz@y0`g3W;|LMtbG#8k`A;^@fDsO
z*if;DUm+A^+BF|(*l4m@foL;3EXorRtFxU{S``Y(cujzAQ@9Ta9rdZo9`;xX9NWTm
zWkiva=m#OTq>)@$<UZO*-p=FNrl)j%m!y$){j<S8B_4*stx*WH@``j6@u#FLvny3n
zgj^BXOH~dMGWJ?H`@!}n{hEl$IgjIfiCHF(iZhYMX^r8!ZKVz&ehvQTL9y3o>NO1S
ztdm3StD`2`0w4{iE`xD<6qM5eKjIiPkoxVG5UHR+b|rqiJh?l3;_dd2B&@s3&W1)5
zac@R$5xpu-0bs>Qo1RXP-*Pe(#eGK=r7KWyziv!07gzZ8<=L50-0@K+&EzoqU0mJE
zbS9r>`~lz=$FXfmFAkCwSDJcjV<$B=6Ay>2B>Y^*9q1`7q<+PO>b=!uyRXU(=g8F*
zg<WTGqXxorPqI)i^Qu^7d#%W?gizWY7lB@t9O@G;_aNFq;+V<<`FI1(-0*|pGXwNh
z=>E_N<d}I110Q8~4X5`ew$cdcEL4;b2zRq4Dv_L%wH0^XSJAZv5YY{ZoQTv;7jRt^
zUW!HYc{M75jH*!BnMbqzE-3=+1dJw*Hze&Y#FP_-%GowIK052u$(bmycHjANZy)KT
z4g(%k^Au05Wnl;OmDv5RBH5BCp{&qDa}&APibIadBlkL|`3ryNOOhnxo1#-A;W%(b
zdSt<02;-WX2;t@p-Bd_dS4lm`)7i-jKc@4uakp8TMR+hx#2a)Lx8(*C?lb2${+bcu
z5E;LcFppJ#qcn_~1u=<hr(sEPTBwS07<^@0wTVAi5Q!;@mO0ufF;>JC(G+`XK+Ri;
zn_@`rk=Z%bi+HSn<{VAzfyMWm8iKraPQsBy7OKQ(uG4GXJfNNbPC!o(?6AC4bJsjn
z^C*IE=tdp>Bh^0Dtv*-t22gUWH$k%Y_#_%a6s#Ch;dLXti5w-k+R>j2jZjX4&hj(A
zs9VwxGT0KdC2+#bOXX`+eijp;q(ojB3}kUjYt$kj#vS@CWHBo-eH^XdoSqk>4^Hp$
zIWo;$%c`FvC@J+tqMv_$+yrsLvoGGnWhKrZ4*j>v6%x3vY}MhW3u(3tU)rm-t3#CE
z(9=qUe-w-ann3vHR@m#S68P~uL?9Zv)+Iz=%69s1zM?PCd`{$5PW+76aX_F?v1$Bk
zgMhORAGF90@{CMjCxNONd!GAB_2i*GU{Qv+4CB2#EnEN5iuu+YfR^$eiz6$QbgVEU
zP5qEt5E@Qo9Q=6noqfH1{wfz@^OyDPHFgCU8t1`lYFyQ$TP6C^lh=;W!^pRVW?L;P
zrX;=jOImuwIx5%TFG!kbGW-NR{UBO*PXQe0d}qZwU84T=HGG?wKwwl<GvyS2h|Q%y
z64uRKbTe#9vteh0^nqLbr}+)yCMj#V%?BT8b4`5ps*_N^LAS!^+$+4-(iVr)7jXKV
z?w|j<C@?bo-vkX4)Bh|Dc}b%!`X>$f@%(2|m^%8^2_)brsc50sWu4ffW?9FyjLG$F
zYZV*0lN4G=s|;*-A|8(w$rV7Si2ubtIuMhx@|<$53JOUVpsYi%<;zEr8T(R2g23r{
zBMEJ~xcfOX&=aj-5t^g=;g>kPGqV5f*Yor+Rq98DS}n(3zId7e8B7cz23I|Y?n~;!
z{PoQ2QuL0uML@!z@7IV%tCS*q`S2rb?i4q8d1oo*vB`q;=VoXzXXCd(cMV*rZ(upp
zAa<@WL*S?##T%^0*TPaPw@X{u@o3GC=RdW@lU%5#mKI!X1dkRVa;3RXUukzBsG+lD
z8uN$64Pk4S(L}?uQXlt2W+<0<x@R$0dNpcrt3OivCyC62qYkG%JgcETL)2b)_Hb|5
z?&0bCy8TDFm9@X!&5jJ}!}nz-ht3=%5p8-TaG8jb{o`?a;F|6`vn-zGc>5x9wR4`s
z67t@Hyb>M~lF%<|B7I90&jSb>pF0dvyw>OQaYP^c^|e6aq{~h+%CmfX-qdcaotJ%a
zu$5ax!U1d|q-Cn79&ziCepTSWc#5Benl&J`s<1G&Z>hGn1lNpzZ9=251h*S~K&9e<
zmc#B2?SYpd9W!~grIS)uL`PW+@Igp_S#){EYfG-^PBk6Flz!#kD%b!v;$bj52em(0
z=}(Ro!cKg>1~&`DeqxHR5WxH!{|`u|`LXltp`e&aUIC{@lX7KC1%@3&7Y=%M-<Dqb
z<Wtlz9doqdT|2I$gDUJr#80+@891pXzhim1N^t(Rf`T0SSBpgor)SY>=FxenWw#Wu
zOKU`94m&`i+pK9)WBl4Z$#(ABEWa_epZOn$JSTwr%$!diCT%`JS~tFanf@*z-Zas3
zY{7#(vZ{$;i4$fk38teEgd((sDJfCQSqj1a?WnNWLN{T_dYG;43A1&Xrgd8Roj3ZA
zU^di-ZUI3x=2~HSi(I2Z#iJk&o!qGVqoK0(!7G-9DH9i!RAkQ?^~o}8dYa_XYg86#
z3(qZwmKY#16ny2GD<T=9AzoK5mDy^`!}a(GokSv+=H5~wK)>*sR!s<fJ$Q{yd_=zR
zM?U+MW1#U7EQdmGr+PxLWge$b4^+9#2KbhQ1yy5hgky=^ANyyx?NcZ?Ui8<~J)Nc~
zcm|~W@p&wAZE$UCCrQN{uymISNSx!3&V_A~sv93A&A($zV$5G^5p)NuT|AJJ6At_1
zot7o3fnyySBN~R?+)uN}Be7PAhq>Mo9n<P6c=9THn103Kc>m97BCJC0s6l$0f?^e$
zE>|=XmaZjD%@6G)nFr&_Z%kWL@jZE(+TD<#nUjK`EU<ngiFkSVmXR`#aIzzMNcKnN
zJdr0)-k*0a+sJsFl0SbyxN6Mp%Cb>l^n6#xCagZt)qWg3Ea^a{h}KS?*XJXwpTlkc
zDEGB~;hTl7cOj*z?6V-x?fNuik>bH<qCV5*3BRmdrCZQco&k+o`g{tcF!AfN<OE0^
zd76gjM5Cz8+oX#0_sOE+4moCL@Yxvin<FIZuG%f3Q)QW>u29i>YYl7~<#|XC)uM%8
zS_BOH4l7wM^S_H-4zMQFBHH@u=?&r<XQG!|zq!LQnX&s{*>WVk%aa<<t^#rOJhECs
z*?RL`ft4%`KHX5}AYtw(4;a6t{i>0aKY~)_M|499na6QWmt6<}RdsEN5nS1zK3+)i
zjHy;xjn9G#Lp|6?y@ZPJ5NDv~$+`J1EG@uJ54z33aKe$OS6~^LD99ybtJ&bJScg|Y
z8CJbS#isbIYZOvdZcx^4uSyCTw9Su@=Wm)o9IHOu$dEwA>drlG_mo$rpY2_{^%upq
zYwsJ6xhu}OrF@Xo-i?8OS)Hr{Zf5iCKtAyooR`d!h&CWTV_6Ys*<Xcc5j}B!4H%5{
zV4OT#K1%M>VY^qWN`ed<zaS4HqJ;e2@<zJNDD~ppV{`nZ-(&mr0y1d3b_=n`R^=mO
zEC1Fr0|^l|h+Xml>E9PNUIH4(v=Z}k37&y~aBc*V{wf;H9sbgB**U(C6^N><+JV^B
z<Y{x!D232EJMy?-uDID|eNaUmpVf6$PRasoLcjk)RAeCL#JgCzWMTm@GodxIt_k=M
zz{8>ZgaJ=-dK1!14Gs`yPng2U8E=K+@!_srps7G_`cJ>pA9+i_2m|pcnhZu=)FFEN
zD78B*YVPN<>MF`@RcmAbqRam^X2f?WTOD1p=DyoGKdSgGz;-WJSUy7H1;<^Y!S$m9
zIJvCiI-UB_$J!S;pRf$xHzD&2-N}o4L!q(lrwrL^8nu^1?`5B-Wxpwx5ZDe}E~1#D
zMB~*u##NUe=;zJXNMc3?_6oV^pZQ|I8Z43)pF)b1<v7$H!eozOg~~<8=;v>tv4i@?
zR)>ox7VWbT<@$qygYN8Wz<Ncu2Gi#AaciC*XT-SBj&>j3RzB^HpQ*4(U}iPin^H~^
zg)&|t=$A0wZUVou@V5^X_a(5rt^O4m>w%O(Ug&o+RV)g?*3$DM)V->F%{X6eP+WCB
zl+04KtBg4{E%X1=Lo4B~h(Lg7YqjLG4e^-X@Gc8hk3B(^sj01fSA)iKHG634FfFpV
zh?zcFGWg~yd@JtS9}x<&{1IT$R@IHkVs!DiK+7sd#G8KHP>eGS3%wtZCmR#2Y_4Xk
zFTK<53sr;*k1{AD^U&U==sV4B1hkjqy3PdayRn-=x_hluuBCTlfuF53EFKCdB^)PH
z`-IU>ccVq!+)30ls^n(tY%itY)9W{}gLFLDgQ>%cE0iOG);A?w=gvc1qS{qlRKC1b
zO%H?Hb)1orn?@hj+0PVh@6@|2M(4y;mM?MTX~r`hC%}uT#AU%)#_)C+Dw=rPY!|d|
zJbWxiLll=c?uBY;Xs;06+8-)hm$n|YyFM42rKYbB>(YUCIVQ-YscvS>E4%aPn{v2j
zo^J%#v8?cg!xOVjxLs=<&krRL5ZKSmH&h&|Jzj1_6JX8`^3JGwTM<QDUist2&0T8m
zU%?r^7x0lQVRpXQ2q8jn$4n3JA=J8bcvQ{FFU(y)=|gH|<z01wbs?73pN<88)MMmt
zw>+nOgTh=+S^Rgj_5V?J|L<}b<Btu||6~jPQx1z+XGQA1uDKzA8raR?ul1{{1btZ6
zvlWpQSx`in`sQ;IM*#4WXeK8C@t9dFE>7C9-%kZ<_`Dqt%IfOSUegiOC}dGPVpp<v
z`Emho?Yv;HBVc>i15Jwmyx{x14Rh4LY2PVgPeG=O*v9iYTR}v|o&JdiWQVzV>!a-b
z6-JDF3>B4wOXt5$6QzCX9yr5Qt(5E62u$9OO40T8`n3*A?$4*w)InNR7oI02qLHXX
z%;qh6Mk!umjalXkIH;C>$rekZe3tgm$xB>Y(842aFJUWy2d!=(5s*DYrZD0GUpu>0
zK9dcz+4*&<8IRR3g>!Z0E~+kFm<Ix%bllF!##d%;txs|hR3BVSu3^0WK(WiQL5{$v
zUFiRors2yDb4PS%G2ThP7ZF)L%67swmc~Aa#s9%^+8h6bM{TZaKLy9u^(3NGC9Z-A
z*xC{!CA}YhPMj?-;Mwcai{IV<^L&2!dAOa@hl`uAk@NNTg*==DMwZ2HUXi9~_jdby
zx;}Buqg?Mk9W=}^wU9%gFUWM4gv(JLgj3*b7~5;_d}d@PF#g(!6pdio3N^kfrc+~=
zVx)&_;(tOz5rr1X|E;>0j2H6Rdz}*qZd^V>dwiu~;j|Zd-$|G0QjYbCjWl_mu}%WE
zz_bO?5rc1#z?Pufl-gQ<8&2$So}GWe>@&cIppGa8;5WIl$Y&Xg<om^euOa*AmVBZn
z%M^m|-sh)!50x>!Zj=Inhwou=se?=2P&X;*m^;lC-X~O$4KM+-jasz6Tgy$`TjdnV
zpx2eBmdwTmpkk(XeX;XU1gYAIAd-nKtL$q&R~Mlh(O}Ic9>nD2{1HuD&I{;$mDLhs
z3OJ|8tTe>W>)Kz;lS;d!R&-E1%pJM>xCKCLJ)t^Ip1;6Wr_KlJ)SDW4NxVXP=TLSC
z;NAWy)KVd#S2Ms|R@nV|1yxbHXenx~Cz4~IK)j&ZpRPvb<C}q?onaP5_-z(<3xhyE
z)QVZpbh%bKLgz|<;s4q2_d(D~tEx6c-SU|_CB^LhhrZZRowrU@*^{xLgEariUVf5F
zMsFN<9Sy=i+2|!5_F$)8{u9HJ#K@CkMa?j68*v-XWim-fm$Qe@(w)R0ucyN`!-ZhU
zL4SQ9uePobkyIe_f=wxAFM~;Sfe<<Liykx-EpqDb#<UMaT>lo=ilUYfNllun!|c^P
zW`!I=P#GvgbAcO<q36>#SL9~%T4WtpvM0r`O#-W-72V}b-@oVy+5Vpk3g)`JM*qJH
zO@tz*BGX*10eBYcMg?gP_<hqeTRG{<60PH_={-(Bh{n=Or3_fBcpX+-o~A!`-3A=)
z$40?86qrw5XiKH(Z#+etSWnVuz&@58lW_feSG<3GJzORZ4)NTTwszejPd&4kd<*R`
zmtvrN8KSw#mH*?C%$fd!vO6Y<GF%keV&bX5N?!)YB=WK=@-it^#J^EZSp+R$gll68
zVS?Wu0vrNjG-l+>da=}WM-ay&1t|=0k(i}5hoIqDT%&hV4~3M~1+D3bg%^3pJ&R&)
zU<{)<Y-5D{K@f-iE|J3;^D&{gOarB}QvbkvDm@J1q;G=Cp5$rD&5Djc#M3HEhMC4|
z*hiM`OV@?I3Q8Wc#smWmkR@Aa_Vwb;#B72oN|a=sd+#7AEG8n8d4}wckX0gK_$Rxq
z-!Bs^8DjsCHBm6g5fp?bpZ)SM1MudDK1uj&TA$3rl67g+KAe3N&wv9&wN6_kpzXP3
zVGS3#Qi9^`CsWD+n!YGP21;$yx702d)kAFPGyUj>BelGcf9)N48xub%@6_g$Sy-R7
zTvavjp3#hE)D_v@(`+>Q$D(&AOEt*&5ie5GBO9k?FtsQGs?_P((|$7y5G@j_kg^%3
zEfxOu$g9%C$;tq3OBIlS*Px)AIW)|HO*>AScZ=sABNDjR;xr@0pUULiEkBq6Bw{Nh
zYPT<Eo2x+?w(~A$4PfAXy3|Wa&$buz?MW?YNo_dFAEqulqPQy0a3NFcWv^!s_#cYf
zoIc<>C>pILn{bZxzGa7daTOMA$#*R%TpQmg!^mD{xy{M~qAw7}5Y6!pmWM>geznL&
z_eEd4I0$`AslsvrXcl~|<Tb*GKt99b*@$6-v%sQ&^<<Qsx;ZC2{T!aqH^`~h4u+ja
zbWCWg&P2!Xs~hom!5m?05~bi(wg*G7J;Mj38F5|D-q=-20$h7(Tye$@?rvSO9R5*$
zHb(?IFX#a7M(ybg_C;tD+qs_lK^(_7nyQW?%qA~V9Z7vQR=2G~+W~1;6nq9lGnElu
z;<>0gttQ$AHe*}bLRQ4C!fP$T*%(P4pk1X4F3Jn9ebO8a9uhi+2}=TTEm*IdAraB-
zu|{VGpWYo#zF6Zdwem)8n*3p13`H0(tt)0h3ve|8yD=@^q$Bx}6WQ~>@cEmmw7mu@
z>nRUg)X|oz@q$TPBMr5`+tk)v{!t|_J7^&k+3!0`GsIY(0G@GKpQ073miWA|w5`Q>
z<8T&QFmz3*`YN%NoxL;ZI1FZ7IBKiHgu>KijXf}}9f;_H3~?=L)x*&N>+5)?KbmGn
zuEnj<d`$?PDH<Hh_Z#`6;?&r^t@)hZQL8-oz`NvIcikY+&N5SrW6Ir`%bm=fRqwfI
z4HCQ((}FDdQm|FU!<Az4jf)feGPx%T8_ef?KzCl}`NwFo>u<77ULXa0iPU;FJD3V%
zXA<La*S_)Asb`99*HTYgVWD2tYac;xOr5aUdiT1L&~sfSCJ?;~<Qpo>;iNU&x)oN{
zsGA4NB1_0J+4EVP#+BF2oJaK2DU{{IqXT<s>DH=~-Nu(&vOPPPZ0y2E6;^WdQ|>qO
zM?=w?Bn^+WHm4*`;|0F(MLgxM`OT>U_uZ+w*YTe31Dcr;b@kod$&(E^7Ge1(I$Z*T
z6cvWc2kjS)kDU0c`=4c$Z%}pR>%Zb4rvD-fVrAv{AK{=bY>8+bj=O5r1IX|r<PD%a
z`X5nHY6gxo&Sai`qa}h%L5h&CZ@?eBKgG@ES0GA~3}bp1IAbX*!E`yUBPpAq4FN-p
zCBaD(6D=Y;u+iV(BLW2&RA5Y!)O38w-xnMy5PV`EK*y51BYO=4wx1t`K?7JIrck(O
zd#_^wy}yEqVUES4wsHB8bFA^XjvpabxDb#C39bB5hM@VbUVG@iw{9cpO}BzZZdktF
z9_rt{+45@&4J0QH&ttURfF<b<5AqxJhf5D*bdzm6=!pN-V~wy#h5EicjGQw@ek5|7
za5%*NpotNIzPPGd%>7^B(zswVZKB~#y`~3|qB1}V<;IxjrJr^#1~6FMD768&zrjbr
zSojy6TW-Byjwt^4gm<8yAghOt>opp}V8!QQjQvn50y`j|5tFB)u^ec+(VT^1fAdS4
zry-6$EZz}b-{t88+-+<{`Y_2R!@*b5R+$)4K$89IhXMl7PI^a-!0%$I3Ayj@QT-U<
zDlMhyp#*=U&ztgxP7NtA(Rj^u``s~^<7c99AW5|46fHB*J+=lOI3~CEf@7Li$hzxR
z*n!4<4|#XQrBQ&>Ya0zbZU6z*ds)Tn{-p469KFW2VU4)CeN+Ec<z(PevsUUegxE(m
zBb+-2EX{L;Sc+-Q#Gp)}F=^=(Xb7Rx<dss96G2L)w*)tXkR@qdvM$kIk}~Cnhuntn
zE5%&UIxiR79F!z2S%&To3#|Sc2-r2yO(W1Y#!Y&cQp~-a6|Ye<OIxa%OH-8(TE!Y@
z>xAEkKTIj#TlNW}d1B&|b9r1f1arS}VU|2hgrna>7wpTsNCOdoB|R^937SGey*6zS
zQBLrfI?7)^F5e%l3<;hiI#ETz!X^{d7a$L*jFb~4+6BJL)8uGX{td-t9>&vpm=~CJ
zTSe?qPT$VtSbqz0NFRvonivEMEqBfaE8uAor4H+lM0FAoZ7!TjWtLHnBcjO#oBG#G
zaW>0@fa#T%=~#U4uN=S{kSYvyneG_cRxXpIfN#cwe1P_wJ`08oZZ0E-uPg0#Cw1SY
z-z`Ue<=O0H!&#UIS{Q%gZ*?TDmKym^Yq+UPoX&mOy}CYR_b=09#&8<#(YD6^MZ}Fq
zqu*j_VRZ_VQI%3*B6n1-#~2D8M+}!(v1_Arg${6;c&3)VuBAzYSm?CGg@gfWNq>t4
zp=WHkA~gmowRf3A<}z(9&41>2roBW{T{=NlDx?5v=KJy~1O2}gLc8`8#fE7G9J#H+
zC0r6!o}(+^b?Q$CS-pS|EE%fQhc?fWIC(wN{VxLnwD8rFmT76xT*WWUC|Vo~MKzwF
zUvuTRQ8d-%&#gmTf+-9}bl7#(X_I-k9nDer&#kRtfg3wwe`jyAEV!S`+C@&Xu1o|r
zgo7j7c$)zJYlRhsCk#GPrQQuziKYszZOy1>xvozYoi<w#W`!ygT-A9eGUsdjps00@
zUr44dZN>jWYFI4*O)!Lrx6+)EvtQ-^4`c5TBucb(VOG^G+qP}nwr$(CZQHhO+qP|2
z-KxGXI-)24>xefwgN)1^WMrJ&x%XP@``bJ4upm3g_`=4l%h`T+LXzGPVNy816*K*f
z!X_<XGh1`a3iqnXaI(|>NduF+O^1PROM<2D?N5Jno#{>1EB?sB$Re{@{FTcrI6lAH
z1;Y{%m?BuGPUJ*kZsm&F$EUhLsxW#kTYKRUDp`i%Ydf;(f!gfD|4EBu{rvAGRH+B+
zT@~t;T)KTgnham>!pR2>KI44o?IqeuedKJ7{N1gV;LyikS*gAN92Y&bazc50xwKfq
zLM;2tai<A)uX&JUIq!^>Jz}HV_#$e-eE__Vak9Ez#UW|H^JJ{0TKp{aM|X-Tki63*
zPRiV2<H283Jt-No9*u^bHL}^Q&VDH}xb=%*QrE}0IKz2$d+I~B3EgQm3~HS=XPNCi
zLO1eFb-GEH^3)%Vh(!}iq{6Xj`OJ3QFf(|oP%C&;y74jAHM!Ebv-Y{ZgyYgHVouJ!
zXSfs$<Smn;3|UFPE`V}LRykrE?XXB7r_Su-#l^=}p6L?KIn0FUD69<~-WFhI?y*>L
z3%4y+))>JKvyntt9Lq4(=HnKIKj-{*cRwd)`q^AKa1ld^gQZjJM7&Z^f|IJelSFv!
z;cXCdGLdxUHe7r~c%*G9(WU(*<Ga(<%oj6GRck5{`Ui8_QI;o{cDP>wu<=~?5cA+?
zH!Lv8*`$Tb-0k)0uAUNEsK;9C>F6_yG+kJ(L3h(m>uH<7FJmBhb8Bo)Y1J{{IQKaQ
zuv<5A^+&tD5D=qy-I@`n353FY^9!RTvcM|68nDupVmJFi+_^jVZYX_V2lKhRgPQ&7
z9-e5&O@iq8ZW%U+X2A5O2O+42kVp9e`F6we_4@l-SI1XmLtW<=T5;Z=00^X%fOk>m
zL@?5CvbSV6C)9_*z%M}YCn{~a10Z!ZC7^B!R^BQYGQtGoFRmy<Iahq626j;>3=-2f
zry@N#z!bU7TMrI#@g;0RVW7x6O8#-n0a<OC%7NK?bDsre?dM+OSL=1JajqqiP5Plb
z&^64@j5}q0bem+ecL@QsHl!20nDhgxii=TmNgK4`zSDs+2XoXsli<ChN1Gs1472kR
z{2(DXv+&;0J18EgrbF>*xwIECLe#B$gShc2Dd@i-V!p>lHSOn)4_W=(U}fm$0kEKa
z;pLZfWOFtd^nV6t8Qq+3W$Cr23BgSZQ=LEX+mBO=`N`s5l+8XJ893)4^F}-UaSIGA
z-u>QDMvLO!>*5;`z~E(+zB!FXm6lTipCh$1C%`)*D|l}ky?ss`1Zbfkf3E`qDtvAc
z6ED)|nc0-B2`cZ_7g?ar0)@g=tC|agozQDqzA<^wINe=C_A16=V5z2f5Xp7kc$aOD
zFa1CF)w1}VMg&YbV$xJXiw=gN=VwZAx0>OHYy_)l>C<cV;TbGOu5Sf*DNy<tcc3Vp
zL_%%Dkb0^MB$>{ns#l!<QB~VYyE_mXom0;_btv#?I4!wu5$Hk}dZv{oi4Nb!9Oa%0
zQzf70RjcXs%-S%PS`i3xx(JiFgEXm4i+fV@-Kbm@8tnU^n-T(mI$~C+H54_OI)D&m
zbNI;}J9#EUx+_UzJ|BL-@m2}#WLFuNy(`zQ9Uc_<?U-jOMkQ$;wyHL@CUPhNShkTR
zh?WV)`_7eVu?%r{lx;E{;s^x0LAfyvLz&PC`Q2bem%lu_PEe_<m>+2~R|oQsVTx7k
zdWIC7?=!C?AWlh`EZ%Fe@~|e&dDF`Khi6~NbR^Og<G`lerawU?7sqWJA(~vaXfpYi
zzIwSK6fXzsbw{8F4MOZ1S!@bxKFmPAEO92wCthl!^{l+)PmsOb9dNQWJufZAKZ&P#
zcRVcSYP!g`?_i1CMxCCfDIhJ%7~4W&Gm9{@Wuge)rwaU=@yq1ykRXE{puZt457ZNt
zz0wwAb%tjLgt&;lvZa(I`Y&nOAANttmguVH3+nP=bj~#jwVf9KEls=06u+I%OO6a*
z%2gXu7{0kS08_=7o#(tbpgJJ6Erah>!-<YP)!hM)&IhX&AsEz(*JWm7FkwG%Rz65c
z|Fw2yX8gYsQ&y(`H3@1W_MfD^XHD%GjyX)q{tqUITS_O$duf1eDBPa`OT6NN>G%D;
zx1&sArlwa=>ipy!X+~DiBaoourLW<#xvrtL9Xh*6(*Gb|WcB*wfbiIV(6D1f<y65<
zlhb>DvVR_?{}Eqs?ulko2Img#!pe5}C=TX@^27M6;QuLUfM_6TAbpBQla9z9+GcF*
zJh}(m$0C40=26edW8Xji^7Z}rj5YNaIgcU+L5py{3~~r?0*S!<OnNn2He244&M&d@
zf#lMCk<|;r6>g$6T8@jwOMOX07=2)UPOh^lMkv~B$T|)+-xoQ5pG=dFMYkN0+eBIm
zl=kQ7zDe?tZ4_pYQTEJyDkV2G<0i>&qWA=f5SH*i)GX4uT3jQOm*_^wwWiJK#ooN@
zmyT;R`w)?&(O4}py@xW;eu*a2M7vx)2OS%rzAAE~X$F*CF>mmV)HShkDrD~gp?FFl
zGCyVX0ehK5m>5kAb4+gyzC{~vPcX`%O;BIc<8_p%->R!3WDN7YCP|!@%Y0+rkXz}s
z$tlBLV2qhUkZa+g>a7GYd)LtB>0VHKA}~U>o5vJ(BRg1UAWm9l+9b%JZq!6>FY2|=
zmA&NSqaa~U$vTogNNeJ>>)zE6H>n5s+@VLe!gSE6tUQ%uwEHd!w?f(B6q@6GMM-B&
z8fM$*RUpT9!Wn;9rCR(jYDIY<NyFpjg|)5Kb@8EvE*z)7`F_4Sq4B~8Vrnp9;P(Bz
z)LnetPKmdSppe0zv@>YBwHGosiAryJkqVEbL_&>GQx|ysy{mh(F#l-D5Ff7~i9v1}
zqbt<`yZH|EV=Gvj(&uiX9|&O#2CqY7ohLzzaA`XW^Myf^$qw3DV@KYmB23Rg09vqh
zDQ=N=uSBSJ7($<k-3ZWK6T=!RUUYHRX0M7mwY1_K%Fa&+I=)(c&ZMXq-L-Tn$p6bS
zWrTtr^{tm&W@T}lpvW|U_6V#L_axjH-1+?0sxtVl;#1rfeLUFf0b7@#<uKFjm^QgH
z^UE+r&32+JN-yDg0OoyRXEI|oLo6tsaMX^P&X_jS-7HLI;jb-q&gm-Nm5NjB?+FH=
zgLOF<_5lm6ofI&muT382@kmo%%6nzs4TGM8YYF;FuBY)u61aXNo=TublsE5UIwxSe
z_hk^T<?|}v$NZr-lZ?covEQ;Oe&UP38^Sq^qVRWQt|3FZDK7rWI_;ZV((M~uJk*<E
z8SY;cN@_<3Rw9G}8*uKJz$5<yG;!QF-k%Ay$#?>fa~d}M1Uy_wcf}CrZQAxX|FMh*
z^vNOj3Z#+RlQno&9=$9hiBO8J`q1n)E1EJIIe}d0aE;mJE1g{+aeuWfC^N`~zyx*A
zDmt!Y)1%6#LTHdsf-hJs{-F^c-`c>lN+};{i)y95c<#?MOF_uFIXMl*$$U>Z?L^Va
zv+3VI+O@$JZLI?{{i%|q3wUWHNKx!TB=nYQX3WzgExitn;z+0S?;vB!_7?%JwA9nM
zYyNS(c4_abQ1oOp0>Ppba~g-;bvetrPI17paoVhrP{4ac1zPEGo6k;rVjrtfn}Vv1
zW9o(sMJ1-<btiexs{L*DN+sg!%_+^-HVWYuk@^$_m3kF*;@Z&X8^V*4*8mkjoruyL
z(}iO{Vmq+sfkw9geLxtU4O?78R{|=4qW#YK-7QQ+hO=~>Umevbl7k(!&MoAu`~Dae
zdHrriXn|EsR|KiPa4fI1(9;}B+c~@Zd+|JgYYhdM<#w1D>C#Xdw-dYd4Nu##$Zgb7
zH+l4q1HZG^GX_Qoj)POi_UnvQAuzHgc34c6*=CWriGPf&jDxW#Y;B9-;o+^rb{7J@
zUl-(+VkY_xisqXNQ0x_Xlq+KQIdVL2-8e-fDbgU~(?s_?s-6|-vnLq>rb#-CXs@6#
zj45BPjS8L3L#IgOQ*@qBn^%mTt3chf9@XSSnTdGRMq|B-L*`}(fZw2?Pd@r9ge?VQ
z>D*_{l`)pk@ZFvKS8@E3w`(6WVC2VE20``8y2g>-HFM+Ht26s7OQ3quJSJeuuu3Ol
zrF?o6bDP>u+3S6fs5G;ucZ&7F+lHt`L|hO>#~pTxTc2vPx1H`!;~^HyNXxj7cqJRC
zRcWKTUBV>Ctw#y`)khJ?;@7@82ITnU7Bmj}Cb@Y2x<$^cOC9Qw>th139u={UQ#Sm#
z^?g75{I%Qj4KgTw`wgxYkF6g;*_Gj2qC#{4c2``A`;KL?o!F%CPh&)erL5=U4`Zxn
z_0Dn)UMEg=mAQ(Wzr^4W#N~mDrN|H&o6NVMWCV_EX}SMeU`4GkMO7%H6m^GEeq6tg
zd@xaVhv|FmaULo6lk2)$x|M3JWCg{5CG3MT(Ktqw4_>XNjOW)N(VR*pJxRZ3r2<Uo
z33;<_Zv9;SH$_bu@Lgn>0{N-juAD6?S?FR8P%V(poOH6@s~^gz(meG<S~9?j-EntR
zNT|u8=KlSAV}|<{tB{vvR``TGAK#o&zpYZ;AD}8d)uN>%*Wq=s19ZqBx)Qt-Aq8=J
zbBPJ=zmryhKz8xMO!azVXoOGGs;Z+OjbLO(6g<-oQehYzlb_gSgw{`;Q8QEU=1Fqy
zWjDKOXhlbvHLMIXR|7&ly<KNt<Y;Tu8mF<8j^N5z;t89_mTotN>9=}iV&e_vDKn+`
zYee=7EumYieJ(LLICl{!Q(Kiih)09HZNWg)C`U5YeWI~T*%D1oQY+C;sF^d@K`YZP
ztM;$kL&b23R1?9RUn=IUkXSAPQu4Gzf^y4Hqvw1si(g($@an-!f$q4La<WTZ-EzW9
zx6{1sm-lIlk(h@m-Mhb=CVYE6gI&DreO6u@kvNQ#W|G*h0KwXD1oN+17$L>OHKk$S
z9D`AVnl$Uli0iq2FdMbi{*-ubgGwU9rq!fOJJ&$bguJA<lt$tNX)$S`w|rF6THe8|
ze9L7~)k(!#Vkg7<grSgjocSX-)49BSMW5<gg<IBK%e`z#AqVu~;iPoXILyjkPs|m|
z>tx?}ks;oXy*QGf>4Zi5sUhRqWVU_)q(p5@lp3Yzacc9ou4W;TPsgX!{-K{H+UXH<
z?1IW)mPyEPW?M5ib%Y27#n&MF2s%kEU6UXN$y<P?ok0S{%H2yuhhz&DK)uivM9|GE
zg*iFh+UYu@s`zbJcya`R0=j+vPZiOh5vG%bgkPacBDBbOC}N>ID>_0HkwF)v+PaSL
zWHCs&>A8xr610R#^DyW2jkxykWU7KcVKtK0wc>C+cpeB5PxXLU9Nx<!#8ernP+z<D
z2ojvv#L4Z-xBqSwA_Q@Q>Ra!KKtbnmjn&}b_NF6DIS^6L_4q=l$mwS3oNni}i~J=G
z!1I|uPbPROcN$P&m+wyc9P^Eb_(_XQ5Y}7p7E?est$O6C7R+m>CD+aP$l>4nfEM_d
zCZNvDp7i!5hY$by1)7)r5@y;K&NEHI$7mXhTc7J#W9C%&*0>)l4_xZ^j~+%i>VK_G
zS^lq@kBOd%`G4!H{+llob-?t#{hKeJYbw|Qh61*g{mTV!O#e?V_>2YI#;+eZi2~wF
zdE?g3_<&K{zpzMNK1G(p=e#8Lzuq2SFOOSci)3vce`Irg|2hD9<KOGSvzyOM09vaK
zT&(4l`DKU1qw?9kF9?!{;Dz<^_-0L20mU%^m;uBBY3t^IVk@v0+Q%}#)7%q1v=ueL
zY1%pTplOOhieuRmZr?lgd_S%2=>ju+Z-qUGZEKdX9alI;Wn*s7XeV^WZeyT{P3mE~
zEz<gKg6CBvFc}@)4Uh0>Xc)1<mtNOygIKdKh3?NbXV?w9kjfS=k>ewY)j4c+f{y6t
zPfJ|yCon_K0vDP-w!fD*)jmj+OvPp7{i*3*O}K40@gr@d27sdi9LHxnI3e~gye=0X
zAAQUhPl3*YGbj!hMNyLpRP&Vr7Qe}SPfL}}J5Kw3Oe0o#5|xV=D{AB>vXcm9oDDa{
znBmFA&oINBc_hU=B|OQHj`k(15#^9_n*Een{LP<>F_ts!kvLMYHC%dsm@Hq?Ge3Vi
zzvFu|v3oYn!7+(<S1exB;x+sH5ja~8GP*ez2%oXwP;fIgV&g2Td+c#pTs=`shRd$v
z0?laos7z{#Xu~E=eCD12)e-PW_Getx`fe2%r_MV;N>(`893r(&zb;%dZ<W9~c~EGV
zCYVX?6V47Q174b_ye?$1#w48(t2|VdEv-d6n#&0u4@xTYL$+!t%8Pxx@(GJcw3Fo}
zxFH?b;qz+sz<`e8cL2K(+Gu;1#Oj$f%O^{u<{mB6!l6+h(J<MT$+jTVIz6XXj|i9J
zZpv}@g`bH+6gdr2a1AY_d}<af$sSJotJQr{7PG+l>takd%%t&!U)3r^{#b11nn>-^
zlk8de72X_>`>*G69_iaxQ}gGU)CP}+@-1Rct>oCCH-ZWI?VUSs!avu`^y%Si^_oBd
z$>pXcoBKiL$}2LY46$p2m)fgz+)xRbu659r+Bem(hf+Z*tn(=4Zq>tNSHi9AL{~%X
z&RCj;Au>b(Mo&C_^IZmtP0?EHdKqfxfR`NjbO?_!yDDOpW=3jU212WyyvukKC3SsN
z;q$L9{v5h9x-bP<y9%WVZhR)M=ed#}$EmPswp%tWSF8r-UI5YPl?;cDkX<|ZlKx;7
zV@}bwc&jL#zWQQb$~O5@_vI-|K5}OS-_mU6vDz9u9FDgIe=?xZ6u!ysUVZ=IP_RPL
zXnw9J%a=q!{2gujF8PaC+X%~<MH(hEM#`kuDS{~mO3YBhnRyUDkZ4WBRgQ{n7Mh7c
z6`&a^gL1RoEvF`mrDK{mHrypms;Hrjvn(Oz!^#Ocn5i3CfG_eGL^Wy!lMcg+%I;Ww
zv-3c|4!u3(%0<T~c`EB$PKmvTOw-P=#F5O4ctM8~Yn3)w&HVW62L88SeJ83V)MiZQ
zCuwyp%c``f?6R7Ak0yVjp2Is=EaLTP4(7$4Q!;!#v7SAPrBOM&`i-UQV6{kvmI{q;
zA0w4T$xjM7S~EhY@VtgrMG4m~xQe1nEk<v<DlJg6<%{mppm|4zC-g``0_6_)Je&)D
z^fEYAWjxvaP}5)cZK8G{bIr+*gkD4+S^4RBl0O&RV8po8TMj;i6b!wQ7GvOjRM5Q-
z-?@R+wc>@8Vr#14R;B+C?YC_FhH26=+o?@qN2X!-S%D4?m(t^yX=*6ssa07Hw%QPU
z=l%nuAP-k(_iJBuX}_|tB@7uH&~m?@Pb9qakgOpMFf0c_9DiT#mhkBO@G2f<4HIMP
zSJP)*0sa&%kP+!a>Ef6(GS!S!{7q_Jh^c49OWx++OlS-)spl`kP04siU!-uTQYaE+
zWxf~P)(Erq=R8n4ti?Lm4%pxiPDoVdIOfM&=5x6%M_?`pVC;WFmPjNTsZT9#G)JeO
z07s$*`aO)gqjvEPL^@)PTCh*vjFqj~>6cB^_YK1ycoVK{IoN}`-8A<`zZG_Sl6PJs
zf5XSAqRPfj(xEOD$WuG-;HHqE9wP1nQf@&}^H0}JInzgvmHX^{cs4I}ww#3-Kub?q
z`fF)Z11FUtzgShnWNEd{ojaARk|jt1yV%_RXMFYe!br(?^_udxr?nC)yN?6}le(}c
z5B&!CT%8~_7=M>?R>OcaCWDauoIa^^2$i}RBNKGheLVmaG%+A98BfYbpT`i7tCn`6
zJ418C?uPVuG;%xgvXVnnRaZ%Y$`;OuF=kJ2$4Toeev1;19zk8U%2C8!dvKjA37|_R
zO8HN3E6`cKT>?bhJUTjhX-s5AsEC3kRgh7`Xjb(<5vo+F=eUl-9SxedNJsNy0HT~C
zQ*F!HId$jbRBcCkxW-UzjN;HO)f#O1%nSqzCHIX=VSUN?fjrjXHk%j!EDS!)S`N@^
zlrD*A5qy1-9!xcN6m-pKe1({PIlv!e&{84ea6$zP{1U%J{H}yOJi$0fCLYs|NEYc2
z==8H6Q=&-4QGM{s+E!8!dGhKFIbP*vvXChi5Lt&#q4#CQ?*u+2D)yP4{*F9asZdY5
zp89Ba%QiUji|uk2S*-D#v+%>*RP%)t#VnAW+X1h<sqYw>^k{6jOnQ^h%n&C%GzXkL
z59@IFn_2?J1?n#20*?|Ewa;qPfZn`#E$=nLby?eLA?fX^Ry#Ztsy?+uOCEs*I}Zi_
z9ODl`K_Opbso<u^%p$7>ZX&&+N2p~~&tAejcZ{9gTX7j8F}xJH)M_n6&8vG$sX~Bh
zF5sEJ2?|+ucEH71qk#7~JB5M_f_hoh6%fQ}Y2$O*dWQ6FvwJZCgsCc|KT4L%w!|+C
zf$Wj3iyt#W=lErI@g<)cSG8@P=47!}(!!pF-#!bhvt89YU8#3Q=V`2^Z0c^>nznzX
zKQv=Elx79obOsd%xT?(wZlAc>6LzK0J=Q4-biDn5Rlnl?jR>q$6<O5l{>7sao(ew+
z8v#tU?u%WDCARo8?X(-xJ!HILJJzJjscs8q(U`ZGyK8$4=qO4yj!L%d5+bWH7jEO%
zCxGxLKw1&)ZBnyTQ#oR{VpZX5b|zQavnYF&a_Gs%IhN-;7FNh%|5~~Dbot}Pay-_=
zG(C3$13gJh0}~ye2`(u$j&&-FGfn-d=i)5Kd&?Z4SaRxkY{qx8blKklAW+jL=B<2b
z=0!pMHrAOGMQMp@T4`8``{>~18`^4|7d4@&7pFqx`xK^VxuPLfQLME%#@*=SI8gAu
zE%zcm^lbKrBpm7vzu7KH#x*5nbC+nW_}lnmF_#(KZoK5sE>S0cirD8W)vjj2k%9Yv
zEZmuGbG$AJ#$`Sna0QD%xaGH#QA-jY{S<WjEpP^7$E2<5fyCQ&fi)Q<U2@H>AauJ`
z+ctKy9@YHCkhGY5*-?MewG|cwOh515h<P6e+)Tzds?*;mXzsN?03`f&g*%Y0I5}m8
z7FOfP@y3kaGxqXoO74>r2Aj84%I~G%gkHA9JMR}Vk~65gwu=d%OHXH3_fMyPY&9}k
zG`x*|yfotjI;leC%_CSmN!G|Gk+^$Ta0T!kS?Cp^G1T(&u$FCBbdQn16=inn8Gy3g
zYQYjt(wwSwbJIur-B7$l(kOla-_$goBV#6lA$$i(Rm>?ipt#tc*XPA2wS6Pmm2eVe
zO5DZOna|?5#1g%6X~yZm;413lMBJ2uhG}s~W>=bW>J`GuM*vwbsw8|Oti@i~wVS_@
z;@=C+#BXMsi@v)s#meJ($wysCl0~Z-+(=Ws8PJ@HAdUKwJl47=25+%gJDT`N+3PTh
zr#dirQOp0MQv`?X#s!*O@VW6hYJVs>$%S)AH~fjuHl!<TRj^xYFU?}GgF@E!zz28}
zc2;XmbUr1S)4P83cK9RPu<BwJ*MWIT`D*AxF+S~#$d&rwQy={C`b||ZGGw|*1y?Of
zMfkB$FLeR)FPE8VX+W#7OT#0!Id8`9C}IA1%V&2B$&T`b&6FL3!c|vE(gZN?14aap
z%&m-gUuo-vl3<?8WKIW9Zx<eR@2*ex2Tbm1gZIDwD6sxtFDN5B%m0YJ#G0@*>Ug-V
zbLZ1%{Rb^a7c_3!;kV_vrx`bx7v~DQ#b@F;ep4Dz)7~T0ScG2ex4KXeX>?r`oVxHy
zTuKW&YYUUB110^B??=vVk1T{w{98`~+u8N;Kfa&mSEtXdq~y(mr)W5F3us>4KCtiR
zFtH*aLGs}L*nSQcf(8+Sh(HaRN8}>nM(=XRbUk{9J_LZ~A=_I8`CSIjKJMSz`gp}f
z)g#Z(seGdC=<MM7c%sjP+y=e)e7-*T`TP|Ifq5(}f>{kd|KcsZh1DH3(zpc*-in`9
z-?SGwco?TUDWHE^oXQs5G8@qV3{3g;^~nv5)9FzT;wQ|xY9n;4FX|Nt1hWJEVT9%8
z3kStNqVd%Xf%6%2u)??<K4K?T$Y;;iAHWkg^=AZ3j`8N<=9D0M<><TEB42?8o_rIO
zY=HvGL(s&_NTyE`GM;zCB+3ak!&Q^UZ3@so&(S}Q$Q7>5o=^{812vE*yCvWC9br1W
z36Kbm7!CRN^>v8~CI+v+{P)zM2c_EdQ4K^9Kd09lfJjH8p3*0fD<p9gQi+|=Bn{ya
z4?~jCD?$=XWv($97%f?;201^dRr>5G;N1a7_ml!YQRc|y;kK#d<LKq+W&Pny20KG*
zWR_LcQSJoFUk3ltG1^7Sl?WyK)_;LcmfkHiK05cB#=Iw-iP0QlUNuOXd*Tso5F`f!
z6USqxXV*#gY|IphDc9r75E8OP4K<lGk<x$+s*F5z2TedmgAjRqLz6Qg#8ZWx&wde+
zIulfLyAxwH+38J4)M68_YlS#0jlD!o+8r|!6f<{A=EyWcr75wdwAy?zAYP=E^QLff
z&C2wcc$=LD&r<THpNuln9vEV9C4P)?lA8l99O}hVE37K!#?(EZzbjrkU(I$wp%;O+
z7iOTTZo4u7-w{QvbM>x#k<$=$_n2leRk?-q;k?0s>KgH}LfmXE1dx;GFlRfT6{#=Q
zFMS{)4cQ<?$TkMS<*-t59rO_y7l|6tCn*RrLT!y>K*0&e9SJR%G><ybs|WsYL^s(f
z{5cDDgz|1(1~`G6jG4FpZUSP&5qUlkv~B^Ewnc5uAQh&?tjfIaNNAooBltDd;5||P
zV22UKGU(AU2LMrX<iQ@&A5Qm@+~G%-)TMM0(lkR;EbEcF^g6tUCr|0wZG`a^T<5}a
z>Qs`AP~t5B#X7|x;TRN~APr@2WInl^z`(+BsF6jETg<6(zO|SocXF;!9*rG?;rrnB
zvYitM|46=)-I~pEA+x;;sNHXIB@Avqej%+b!%Ud&t>mgOI5mraleK0jYrO5pyQiVE
z@@~Q{GxpPF2Cy8~EK0`iDt<8<s%y;i)f`qTt%E=In5fD?M%lT(NkLiAzPj$3IAf+H
zd1))_{#$CsxC=PMcf-7&IBwGJNmGq0;y_zPU;3k3_^{UUFH4K))1qwoaZ~d>b9Dz1
ztKhp(jjCax($C^VaJNBuOg04!Mx+MK)hk1|MRZD+`etUaE;a8X<B%!+&4zkI`Bz6)
zZ|a*7C;1SEF?2+(5he~HD>)->BeV<Q_`2+0y(^guU*WCBIOxL7*KM^SB7s0H4Ij<F
zr)f=fcHvuxJzYd+H^o>#D6Rb`CBQEW{(dSs-vRKM+4b`-IW%MumGFb5q`Sxj=7%U4
zkN5A_-!Hx1zmc)Fp1*JblMq4F(@R6o|1{b3x?dMyeL$wrkWU_!a_C{m90$zrj5K|I
zS-1ciRnagm_h2xhfPnMPQ`5@bufVBjcA@TGnzgWxx_p}hf$|&rqI^gEcFmA;C2HAu
zW(!n2H(sD48o$}Ny>f7R8F{M~a`*zXP-0=ftM3)`-LIl8+%rP%+C{X7Xr~g-<wjUV
zG(9NDPD+&A{-kv^{;^cMHS^hGn4M4XwJB=jLLe;$-a&AGYS<OyI1PwEI?v~l94Sw4
zrw=~=Zpr1o#WwXa=RI-8K?Yg%%fBGVvj9L0wkUoLhOV5#2AQ3E^XFB{wPqqcc$Beh
zVWJMu^vS|VGTFqXwAZ0A0oSZ-N})66y221OqwcDST03_z!a(m7qfdzRVMqI9;`A!;
zmW@x$;51+QSq!#yAc1y)jB?X~J#vGw3(v*&WPpKK!a{Gg`JORzfqiTUY}vkAz!uoU
zMr3?xL;av5De<@;$DL_hX|t3gXl^>BdQvcd4TkKvV1t*8njAk&6Nf>fjcFtw-JfDd
zp(#wtnwLRisRlPqQ9IJh(zup_6X(2ip2AZqCON0U+7An(s|AsQ-WIjWia4Q!QSneH
zHiea^1{`LJsBLL9om=KKiTqk!4!lqMSi|l>Va|{i7KM`pGK~enZ1kHFO*14Jov__j
zOgOSH9VqM#U}u@Hp0VJZ%{*NS87BJK6^?Z}GPH6oswOQoFPP?@l5<XgTE@7klU%9o
z*vuJB%VQ^<LF0PI9M~jT$%9TZSu;sm;n0<o+LIY6aJ+S>hEVpI%Ffj`440?`mZip;
z-qjvwy%6a48-AXWJ+D?yJ0;<qdA}5?0V_ufUOt2Zq<lv!eX#-4Mp!3Z+66CTUmJo8
zK4>#M_^_DAxIOs-p>*muIBo$*bES6_aa7W?f$=HJpFH>bqwS_V^ZV!`BThn7-hr^p
zp39YK+aD%-_9HnDRVXo~LOxg6v~9&KSox&C^lFg8CBQCPvx(Q4i`SQbro5q#;Vc@A
z(yZ9m+QqS3s|*RS*P(u~;RLlJrtY3eu7pwHNG1*Fgr=(MnwzYx8@kYtWHQZQagrZj
zy?oZyfCJzy{Q8FOQpQ3<X842wAq3)JHbYexBPi@W5l+<~`u<vP|NLYY1|m$3Ni9G+
z(*cQ|ukr-iR&%Kpyp;6xwd8ii7Ek{ixluXKyQeNkzC49*tnAboobZ^KWZ{`^7m8B{
zA%ww1h*&a&jr{oe&mHJ)nYf?mZb9ntUOU(J?J^%q;#9=sqISTPw4I1^_Y}5;*^X+D
zGaA8DQUm+Z8680hY*G>2tR5B>oJhbHwUUB$oCuqEU8rqyEByH+=vwz{T56X86hRVg
z>pdi;M6dH*k*VYS+6T;h)axWntA*~$K)l2_t?G-fSboXWxx`r3otQ#Vt2^CNCOV~m
zBTL!nvF4m6q}l3~Dv{xDyc#LJ*;lZNl-GC2xO^?_R+T;X<a&jpoV2oF5OGsOIhiZ@
zorIBex^FSr(Ccg5mJZAhT}DRQNf89(*QP+8uJ4eiA4*={WR_X$RmjCsCXnCZMrn&-
z>5yQF{&_TMQZJ?vtr#x|ux%nIy=FQ2$aM3%;i5in&?ym03-ERC_n$-~Na-3!uUMJO
zKCBX+t~klfm=m%ndhDQ~qQ7pisIGv0?by0;X+OzP%Q%ptX_d4wZ?8i^Cy!C=9W~-o
z!)v#lRKq)E)H+gqy6ea|bDavN1UT(-u=#yfgRA_20{_JQcPRM<xJy0ex@c8%K~b@S
zc&mS<1Tp3ZAt|Oo^-!4k_LHwv|E&;OM^fbsq_O064TYBd4aapaCge=XVfgL0*xsR&
z)f?-a!->j@fMyp}ZqS*G_-|V4jcDhVy8GtvNU!iiB>_C$jfmKc=vq_@`{K=yXU-JM
zA%i(f{7)pst)`AHB|g;-qf1)M`nPA+Xd7Ej{>5fk>(M-$=YIRS+;G7zrJGJ1tEzBw
zGp>!;^oqB`m6MZBK-6UAgkn<KVvA5}L?r~QUTU46j<PN7NG)RT*3bib4+h`9#`^J?
zE%y2o^c={?ze%l*gJ14KHZd!!#qk6D3*)_wCi(9l)DRq}|5kRh|C2cXf8fslf3gQX
z^Z!ToU}gE=MA18%Hg;HJ2tT)a4g9^uAkA$}O?FDapx9ipFy!F~T>M%>^yHDL7*oag
zxzW#iZy7v~#F@;VgcQn_@Jdc@PG5{%Omj?LQre0Pv+!x~(e5aOSa(J#fe;+Z6q(18
zDTxyWOH-n$cKZ=<9g&57#e5JmqL6aR!bBWvwuHOnwU#8#igg5};fh=`=3?!yDf6+;
zF(=mC3O9wq(Q|4DGM;*tMj6p@0WEFUF>)L$UFlKpo9Lv2=N*v?^hjz7&|}&n+(*Zq
zwBUx-og+fYvH>t=wKu5DsZGnfi3f*Q)bs)SoFEjW^5@P?5V@D=khhL(M$w$Xix-?M
zPB-e~!HYBKr6Dh!t{cLsmJ$=BrSdQ#X-0$m#2{vdx737yhAqwg;|h_qdGWM)%PGhV
z5te<2*%I1dq_NPNC6t~?jZ$*T#zC;1qk$8$jwnSFxm{hC&?Q}2GO{{G!lY@IN(6R|
z%2}b5s#%cIc_o|7{M7UrucCm4shjEyA(b*CTuR<K_3y@zW0Z-JI%D>ch-o-SE!kH(
z-PAjZ;b!3-R)DL7E)r?>Nrq*Z2k-SD)>`->w);R0Oi0^ef^6v1nuO61D%^1eu&AMF
z<U3BxDOFt%LV!Zc5e%b>MSz5zH>R}Fvp65Ot_mHavGl1E9%^fZ`)dqq2AA#H=hwAX
zhk}(>RirKxD_;tGGsn$q@+Zn;C9h68-VvN6y648t(#tef{pib8mR#;oLlticu;Gek
zk~#DxwlU7l%H^#^J@C*U$2w`^*&j9;5~_<cJ?l~xX^wO$x``>kQv#eGk3(=XTYh@7
znPyEJ-8Q@0d99u<+&MmsGjUH<O%p3zE6!vyds%8TPj1UP->evIcjTy5w^@uD#F)on
zhEUkFC}s`#WHU@br+dtNb9=K{Xs2Gzd4+E7$Rwoi!Nq;tBYsyw>$`NSgY%VocQvJH
z!iEsO{0mwx<;O)b(`0h1rwiOyS$J(rAJBh3|F)@p-hIW{cy+vMEoIAX#jTaOa{q3B
zJN*54Jhy%<`7l}b{eEfmVX-1p!xKG<_<K~y(UYFXYJ_%@%&2G?p7c(T<4gV_l-<50
zA!%mB7kLDi&7H#%Rps#H#nqGEx-?O!A-xH*`F_iehto9Ho#)N-(51;2l@jcH{D78l
zBGiz%8NY1(nzb4hM<Ps@$lWM9P0PP-_=DqvkEk(B6i=j~CnmxwwdsO9l!GrWELPwa
zc;5Yjh}Vu}&&>(jJOdRvm(8=~%;aDi_EHwc@$lf?YF;P0v>^j8&%qB5?1*Xs5FKg_
z&_A+YEmjNsybeED`bFRY7N$p=J-}1=76uOV%KE2NAB{MFJA}|aPO_oEI<#;Oh&SYS
zAPqY|_t7^ZzHrIcQ!PI|-XAZFY*}2Kt@d7U0d^kD+4ARM&W{3G%cOB5DMLq3JKI}|
zM(d<Ef-#j4P5}myI;aCiUZl5iG7YpGsIkVa4NO1t>?!JFB}+<u&Ut^Q8@HwYZ2JB7
zv5apkZ?Cz?)v&zl!I9T$VPa}hOLFA$v_w&TZc)eF!ykH6CtMYzwvxwZv%5DlZclpC
zZRQG%E~_nxZ!1Zw_Qlhgcfh|sTRlK>m5R$>h~{)n?MT9{t}*;HZDPau`RIG@GN8gy
z#xD@H_P(~V!<KqQ3vG(B5Vk-u^;*@KNQv%4_tHZ8)<QZ~OMK#<9)O`O#kcm|ABuJ4
zk^X3Z>fDl(vnPG~ct|K72o$@hDKZka+CBx2NIC%0i3q7=d{!*kWWvZNAxUx<E=Fk-
zG1(VF5L8seW8XYFUF9o`=YgH#Z+AgaHqZHmP+8G%aEB9T@Kb<ifjdKP6r()y`7Yz7
zUdWKTITEN(KDvTPC!#z87yFg&?Ab$`s>T$^z5XZx?>q>57VYMOHMhs!?4A)rt4RPP
z1IHVD5OXICh@C|>KY3vI{sjC=JYst!w8F{wDhh5f<R={moWhGr${h~9ox!o%c|}O;
zE%XZ1&c!@V(QSlnDTEuEGF>VRaXszClpDVsH_9%AFNw`yUGUDn+9b_8jn7}b&(FT+
zm%itHY94kexY?cLw{0AMqlQ~p`TCKOOc2IKftp2xl&lm3@k6KQ2KGEDWXWzm7Pw>p
z?{IJhJ178AfgZ|z(EW&HSDX<Ycm{(`Ul@c71`-m*Kw2<=u)GP=cpf}`1Yr_ICTw5}
z%$3MFE$l#!l0gC~NeK7Bg7?4o0T3DtyZe94pow{MOTvH7*(ZFdL)e#RI(t=tg{0xq
z@_YEQ=yv2SBscADBWlW=mGHn7`F-D&-qJgNJB3r(%Y}87x<cL~!pc2Eor1k6OxO6`
zV9Nh3II&T-*U2YIo&;L^@A4%L@mu|-8m@k7)~j_QZ0poi*r;co5uOuoh;NGLCI8vK
z9)H|#pw@AcBE-j{8AFYlKVzU(V&EKzxR@3l^tmYrl92MfH}GnB3bcofkuk9;J@gV|
zJaf1cmrX$s9Ag8rSL?}tNQgRaz*hNz6SYpP^uPF#hNC`SE=G<=oV%-!x(Au@fJ%#O
zODMPA_HD?ma2)v&liPMdcL&p(U5yT>b@^U9n(u5$_4iicY^pdhqu>R-v8Uh@s*JdY
z8L5Q?_@3ya1#5kG7HCW)v^JbvMfhsPm+KOd-TMqN!3_{L*th}BBcQ+7RSCL^js(G=
zL8$|MZ$x{WqNS(>T^o&Lzx$4<^y>H%ii{3iW(lR_zh8AnZDM@I`T)X$!72({7yrQg
zQBIhk9uug{j2nl+q@s}haK!Vrh*{n~lyW0w$z_#*nXoM$EyGeQ8WT%(4*s{{|FI^n
z?#Fl`qV$jpA1D;_9RHMIUt~f<`0b_|K`1+XtKJg`YFwN`A?Zi_LdF$C2I`87_4rC#
z<>C*hUm4Qy*?|`AKS=;!w0aqC>uXutf3X$+Xs&IY6f+3d3?E3H7E*@5piSWxM>*c^
zFsLE4AprO>`>J%4lWxpN_r^_354Ok%^Ylvo4P|a)?u}u7%k6plXCZRBQ<v+6X<M&R
z?iEbRob0IM&wx|V5S5Z7`6DU857VT+_^Brp%v!~n?1->p29(`}C`f(+fHfewQ%L)#
z>C`TnZNihsA5H#rc;?fUp+e)j&i}`I=&<tYKEEc-rDb-g@z#0LhkacW+4m~HKL>~U
zT?P!wwk;m|`6k|bKZC-yKw7{xs_AzAlyUY`DaSM2ndF6;@|l?u+h6n}JP=U?SPZM7
zRu?jm;R?b;s{5o^IUGZz2a4Pr(yNRt!;MfeKc+mV{f-@f|2g=v@+hc2JTLe6Q#50U
zM7we#j9mgD+gF-ZA*$B1np=w}8K(uC?#+ybt`?VI=_x7kSs5=N2n6Znjuwo9mM<r7
zW^7+@4Q***#N5^POzmt2&5}3HwCL&L`Wsx}4Fb2%yf=u14DCs(r{id27~X(34F=7)
zb3JFQDk}nrG4b|`eD-VssEv}qn@givCMxeO>SF{=<s%Yc)Bu>ib4(6tG1cZ;-Bn6&
zC!)>d^D;Ik?;9*-HzMS}m9reI{{>XS{y)YU%xF#gv+*JJTv0uO2|MZe3mQmXCI<9&
zD6|b=gO08{)noNwpT3a=8rx5aP^4^{|FNK;pdA=P8wkgblG7vO_2JImg~g+ll#BHs
z%Ff>Hl?L+Nzwg_ZH=j8U)>n<VxcPdXBbzM=IRlP6)uwRA_I|s!$c>Y9FDJ+k^;aaw
zjgy5USCB8vk7d4}I7qnPIcb!#wJYd3(=>vcZ}X2ljGlYy+x6wWr@p`1GC-cE9P+MT
zI=c(oBP)M*h-`t&uGpV>?xR3uD%jr}&S=^cK$FI}x_Jgu{=VN@qx*gTwj0Rb$Wu>|
z%akAr%=?G_1$h@gDz*FAVN4eUz)%O6cq!&V(iQ@-REOl11Eax-w)Ix}gpvBQq${<-
ztQtshS5F8tkDLI#ZbUP4<YNUDrEC&T?u_tfuRd)XRj+`jWl+s~)2$M+=7$TXS2|1%
zL&d1ZMSjGq0@_>itVLH)g`Ng`zFZ(Rve<;5tN<cqzny`q<i^`d1yrg6WqO_o>{>wy
zqh>loGSfo|;aP-7UID>|TyMZ2ohlhQ5z|y3BgGWf9xyU8+E%dL_xGEp?F+Jomo={<
zuP8p!!l<ms#ozIolCth0YBI~}Y*NEbL5<}+SwW%9Jep+Za%)IGuT(UZxxy%CrEEvs
zpM}8`4v}~T2xxO4EwDfGqY$Hb(zr+el{3G@@!k|3YM$w6!W9Fs@9UeLa-~e$F=Aa|
z2hCktG&z3cm98X*)<x%XVEs8qFHyy4Gzf)y6}%r-yYRF6`%!8{C@^N;{HGIQ^rD<Q
zJwL)y#!jKy2^BbxeS@&n`f4JLlY{xAq4r{$0ivkcm;k!5pBTdYdQDk<as9I;h%E5i
zRy`SRx(WeR;yuiiM5{c~zP(RG!5I<CHzBapA5IS#;1#(v8#gqB#@^=X@}i?qUqvG^
z@Gs}f$|J11bt^w3B(A5uD$UJ#s@FQE$pGoy@gp?4`7nP*v?9K9xct4v6k`o$9(Qa&
z2{^urm5*$Otp)ZQb+Wgm${)YTXaZZzU~Yjr9ojBtLTfHM(A0lt4K)e84KgcHL{ceb
zG=wy<XLcI`ZR2|_wmPM1plHM;2zC#mYhBJfkYrSXqjboe9jl)xN}W<(#RS*Dx_NE4
z2tX+>W@uPc(Ty{(;IGiAB!i|+hL)9F+-g^5On@VHgRQ_tM^}vw@~q%Wz*UK7_Rq-^
zMmu;|D|Ex5(Xn0_APfD5l+I*{aFFynAQS;|vq6b*B0P?JU<8lPt3(egQHnMRT4879
z9EmoJ7(lSDNHBnZI5ym55)0;4kq{*f@juw#92_0pgPNYd&B9|~_4h#QOBK}&;4!D;
zUkxJXD+gwM=Uq%4EUjDVE<7-0=7V#$p>8+Xs<rAbd>a#cs#d~Ko}TwxW<MG;cY;3a
z9XqQn-eYw0nPOAZ+;2`Ceun4WUxbHyfpKDUTZ&({PzC7nX#?qHhb2UN@x%s3>7f;2
z*`Od`^?#)q$(^#in10-VyH5>Q3r*m$edGrq`?3m_e^FW_fTKg*0trgOAVD;Rc+K=)
z)^}qQbu=!d?qH7fZo_7Pj1iVMtcfLsrgq|$T(@>k>WX&phN&2W;PNCIFn3({@yq{V
z*oFp-UXSWJwHs9lM1f#Z9w`J?6pXetrHXQGt+j6XH1@Jci_m?cpn%~+gY!C$5q*o_
zI@!HDoteMV8_qIF?n`GIGFNkFTJ;6C6IE{WXzFOZJBtbOu;dHX81Xr3hP@xhJVWV1
z<_X+#*pGxb8C^hoS4sF|C%M%i;(}+FMzD3F5ZbbGzh8a6@gB2=+)$|QF#w_(wVsd!
z&r{i)&L$@kutrdW9->viIxIZh0Y+F6FC}>+jChED`t9w#ajY{Zfg8O~AJ;jm1GQTM
zXHv(_7Mp7K{G9Fme%!>DIwq0l#nGxRoqeU%9X2w_aIh>Lih^o2(Cpmw@QX_ws*jiI
zjcgk;BHS4!P!Vl{2;}c0Pq|GPlP(dpj51*{{Ol(!xUL+V1%?Avt$=9^PU0PI<v8iE
zPO|m-)ejP@Z)t(Pzkt<=#_)X&rcfN$daFHp{ggwHI}$CsSKu+T)Tg)DnB$Bim1A`9
zB$Mm2H-VE7qnrqGAZCG!wt?K^x8@vUYM5OU=nmVN@xU=Qq!|?8d=ej%mlu~hm2F(Q
ztt2^gKxarHs0}&O)_BBlc@`w~fep7$USp_e$y_#?)-<s&AUNS`dBHVxec8%vk}nlG
zmQ=J(x)b)TH31Lql1RZ7UxKwL7>dokgl;T%L_>1a6}<520eM&gWL~@U)Ey#ES68Df
zl?7D76XmSzH4e21>SY7NZ~@gwUSC}bEKSyH&UjiA;DEUu)%yZ3o^+B0f|U`J@0n57
zGnD73`_{3)#nGUj@y7+L7<9`UxCe((!srwCM7AaKDO4Nr%!HB5ct;dbtzPxS*d<|-
zb8VP?1)Du*%%(*{erM2T5~J;U228SJO@MPCY=@z)f?*ofh*6HA&UQ6KF&(=3Kv&YG
z#*F#BOR+MIQ<1WqCpof(Q?ew{Ftf3`MgUxAhNl_QGeWiEuXvSurnyQSu4+Zf=eh(9
zghIsp-2v+9TXtufmd=&e(LKi6=TS-xxmA`FEaZR=X0b3WMMdVHtacPhowkmeCCCvU
z?j0;o923zrj@~UygwaN)LBid^gpTNK+cpk7kWPbRXcw8Xb4O#}T&1a!`Rw{pJ))wC
z5=0LeDqKr;CzxfY<7cTPBi0qy<JJrqs_hiOY@7qLZ5xiHxUPW}Bu!o&lBQ=!-klud
zTP;39`aa2PABQzo3U*yt$f9R)$BJkEN9D$_WB+n`$Lkl>#JV}U&~2_V@bbp}p+z8q
zyQ|{xw+`cwH^mH?-LuxIZoT6MB<k_TzPd|cp6Fe|6_0=qL$IfTG4LPV!?u{V!8eXP
zw4h^Tm=*RKVUa7dDoj=%E^x#*Z@e%C4){f970?{FEza@U+u{X>XB&NmT84QISMR8@
ztu<Oa^EM@eY~$2D{}a*!gQx+oKp!&_U@A4nXR8hY(UKUogrGFjTZc4qrput(+cZ{T
z>5`Z?<MAf0vPWkVV+}Ta1A!?h?AD2OmN(!`{nyD%>RV_HzJ97ZY~fvPHTWGHvhaeP
zCDUQH$}8)?rWoZb>L{!7hN~kCw?0lb`e~t>Gr{&kR~Z<zzHcQRK@Bnt6#c25Jg?eA
z-y*Rj;Ax^(tK#=Cc3s3Ms_O1g#q=SEU%S!fTkfFeg+x(#994ratnG+8!w&I0>$+I)
zVg^+G97rdMn>G8dHD|g(H$usV-7YKb?h|3PG%G|aD#J1TU86_iUKiS89|7=Q@AH?s
z;a#gw;TGau3W6cdstX(teYO<M!3m~Bw}wqNiD^-8jx(j%J3RV3=$BjmpkOkfPIU7+
zS6v<6Q2WCtzQ<z#wQdt$An#Tl8U_FUXGi{N&Ot67NKKcHTursd`}-fvZ(~Kta<My`
zpN>H6as=fT+M&}x!SuKRDmxH~%`H%~39OSCcS-HK&Z&xXBbV%|mg>6r+u<CfxFl&h
z6a^v3_g?ovDPpd9-)b=Lyc8CB{VL2->!v><FU*M(X%e|IR`S}8S5WqYgEI@W8)`kZ
z+H{MX-8ngRi<Ikm{$#M$DcrWRVle;k%-w1w-LV%;pH;*EW?stlJ%!@s)V$`@YTrA#
zbFKf}f3aR*x8Ql781v{@o>%o=<;rw&EV;G23J~N4!v^lA09yZGn#|qk;{W{x{FTy@
z{;%~l`~OvGFfq{ok03y-j#w%-#=N?DIOv%wcYr_uiCl*LX}BaBR?YQT?CM{>^}`M0
zhK-sl9=mxW7?IVH&WY17KDh7usjZW%n=-i)xKs{!RovbmcJS|jBfSBK`ThaOfoR0A
zx<1%nPDff8UNmp!e=!+G4_!lY>^)Qm_M!b4Kw=^6Jz6yO!Uy4l9EkTxM??=@Ge&tH
zJp(VXRyg?Q-{tq&KYw4Yu;zfBPG#OPdC&du^z43ps*a^b`rrS!PbCKr`Tcl8@tFfo
ztNYm}9>n?m`o16Cut7Gb#2}Gl*+TP^Va4>*U`IblhvX$%Gd}{7W5s-19z)&x*$lZK
z23Tn8{(k>G3yAjqG9lR0G!4#2Cnj6uCuKh};CVzRa3unND3hC0^RN$v;K1i^2<kuE
zB*=su<T}kRlwTIp)-l7F?LVTnmoA8&sxE|F0~;=a&ntjeIMJ*GGk9$Zq70h;4!vtv
z@u#d*0R7Ayvr$}WPvNwo`qe&@)hbQ#A#vjw=D7v;%_EP+)?s>yX9th06#mI6`!&<1
z7;1K@JFX0CA5-D`TIK0V;evn;_WOS?_}rM#p!NcG&G|g^4Hm^yLwj;4WQOj?fmWNg
z*D!q?+|Aq)8eo5(W{<vRN8*8`jhU|dd93LSV+chivkq#9@+9{<f85O&5mDqg!3V}q
zcl#$J?7}*BlJEkh)Uql9>C_)@ji(RMkqDsr!EHGwj<sNyjtg$Vk1Wg(Kc+U6jF(oI
zy#8PkW;F<U{6s200@5>US8!-jL<{L)n<em)b`Uzsew!9=!Z|-bQ{A6}1PNF9yu51*
z3C(iRSkZ$|q)>5VS+|5AG6Z?kDp;7dkRpi68)9l0Hf}Q>Vamk^QI#AVMsVL|ty-IZ
zD-_RAbUwz*k_Bw%4`+csA2kqV0*9Z417Bqd7?sVpgJkVsZxpOu43g(n?HHhtj=KDb
zit>{PO<~Q|S3fxKpq3H_6RjWKzea_FCjnwB(MHrMd`uJ!j-w73vq&LWS_=|Zeo$SJ
z+7T^FoF+;Rm{o>b?$>dB)@J#nz>PM*UGD#5?2#vx>Wc+8%Rm7Z0UT%gzJnPi8sjja
zWKLYBXX;h@7r!s#dU?zfRQ`ILu$ct7uUT=zb=l}$_ut|Xg7X5eXCW56N_G>z_+fM%
zet%z#^D>{?QiQvii-RDG+e~z{p7@%bhDcSLCWC4G=$7*TP<D>Zf-OyyKJMc_wr$(C
zZS=8i+qP}nwr$(Cr^gfV-k7<w_YbHKS&>;;mFro!e`!#m);Gr;=gN`;5m!<Y(;#p>
zf;%Mhv?+y^`|Bk><;WUz)YN^#oLF(DyJ4T;J(z3z!XC%SajMjyJPSB+{RR>h$Wrzz
z1L<erYG4|dAM2I$ln@SRF-q^vv2D5rg~RZ&<nB$=NWYg1$h+CiyUr>y<+4RDcH7AJ
zz<RzCjoro``DPdS4PEAEz;cGHkoOVaWQhte&#(~b%8-MQT9H<c;oF+hc-bo#kQc?H
zW|t*BtL_%IRQ6>xyNi;M_~TNd0u?PMNLEn7)g7or9Nm8hIlM{{UrUT-ZuF0h9Rs<~
z)7WA}M8f|_R(z?Q1#DFxV$-2gwUHKvtuvw&x}5S3)!KJ8X6>me&|N?W5kOjR-b7s~
zZV!gCDrpI1@hT)%RG~_9!nJCzV`nVywbd;a56*POY9}f1$G`hBTU!bdrPFvai(;x;
z@NqU9KB-7^77bt&GeO5z3kzEvvSUVpgcJwrfZ>+lWqi@KFTFwHh(U(S!asnQle1Dh
zVxLJcxzmDG=NAWceY8zQb0T&ncGVo4-G`PE&e5j)^d`BYJESmRI9$K$Rvh_3`ADGE
z09&sZn33H_u`DwdfiJbFnGI$>sfK@}J1+OjJlf2*SUWp1tlM<Z$e{%jX}2xZq?%J6
zPk;fhVl<Xl%^?`IgsjO#I0RZiqeej!Qs)?nYOS<qNVYibP0+wunQbg*)s!-NZpxEG
zQF|)p^Osy{BT(DNy{R_5wx#=oO1Bj-Fc^Rgtyx&bec5-`<`rdC59obua73(|w)QBS
zbga3YS@zhMu+~;IfUG-kSlIV4^2zf`2o6PIK&maZa}K@$fhwostIOJbD?*lo99!(q
z<!)DbYh8gHunFY~%Lu@}+9}vpw&})2n?w4ndA4L(B(h^s4~;Qh1@tcT&Kqp=I!fxr
z&TQ*|Bg!(_IHGJ_Ze$9RmHAeC@<SO?fCVj;*drM@lVx<|33QlgwduUe-et_%7sgXR
ztsu4Z2`&7t$YxVf;>sj_Qy$Z8kQ59dSBx@FpF`1weV?ivuKL?|wWOaUB+^HzxamNF
z&dw$+FSfZWn&-L0RAKZV*h8jNv|{kgOD7F_;BcZ=NuY-359V1w%n(73(?Q)&XH_>3
z(!Z%#OFT?1^PR~gowJ;g{hctge-sJ?r=aFVcs7}aVx+zX6OqsrF`03iincGpsFX^+
zvszG*U@!t5O2xLapR+0kC#6k*@?7D{%zcooF0y}(-;Ghnwin)ToVrzAF(fdvmOrD}
z&Qa{f>q8OlM&hAF>e>LFRiVROEIW}Y7xA`W*m4GjOH>33(uizQxzF3{6W3)RH$M$W
zFImqPN&#kPqq!$rk;S_Sqg>mFzBo!{#}u{Ox^OT06Kw%WoQfz)YHXOR^z_kCwU^D#
zE%$^SkkSta;u-q}(<f<$b!eWuM8j_=>JBLwiySWi&cpAJmT9+kMd7y)&$7G_OD9<q
zh;byffh!iP<dZz#-lW$elI{phQf$!FE_>^+10Yoq)<me!@qyO3Zf9JiDIHX0Mys3z
zW!DB9np(Mb#eW&%DYe)cpBV4qgrg`1p1dF!))aeUMT~4#y<O&LXL;V3SfRYd)pa=r
zWJAt=j$8>e#FA{Jp<b<`BOd(ZJRWbmA~P_)-ac)!6GDM-IsX8vLkh?!VOGYf>NLV;
zTy^Pr);!fi38!vwkGw`LYubutySV}ZOUkAIlL!{!e+oLFeaZ6Wcf3e^-Bi|&SeDYb
zk|GjLEqdLqlS7N5c)Xmg`)+fFYvs-?Ros*iz?A*adqcMCSR1f&=%u5eIp$f*Dj<I(
zgCwq<OhaMrve(hnNF=x=GUMb0JfO9pdftxjEH=5Y@OIGk9a7a<<7`@RQq#3=NI8g@
z+0wM0m(sr(Y+qle`o#56T+y)+H?^)7FMrwDM&m*Lywt3jEOQ_0#jMjlW$H3+RYo6>
z^?dv+5$yd`Asc!E+7toGJfy=_T#b=$5X{BhL+zmdh1616^3C4o!oJX2IK!Nc$o#x2
zRBiDq>Ad@ziF?|D#+HRt177fA>R^*-MA63=f_gJ92Ij0%qZ_iN*vaIr;INaddu!A{
zaW4^@p?!-MShIUQ5uxk6XKw1~g6yHF4LQaR6^pnd^j<1<Jr$l_l|8-qUgHyYT0uAm
z&R-4{-9`J(fbd5`u^MAS%uER*HS=X~)mXZK5T<jv=6SINJ!+0f7fHB%_DiCv>f3)_
zeaiw~hf)0!D%o76V?CM5B)ZT*`>SCz6}6@&onof&cw9rXWx;cA7{^m_54mRLcrEIL
zs45fFRLcf$0nvO<klJFSLv6k3>Vm}vb~9;%qJ}MvnIXFj5d#HVi(u`qfR;1qC~rLx
zX&|cPg<xB;B3mGNH8;f}zR@D4i&T$EQDI>PM97rF_T?4I?hPdzT$DhP;i-J$1noNM
z9_js%N7n6^@1*xN%Gj4`&3zG>zz|VYFH#Of1z5Uj;tdW4ALE`4F`$VVOh7E@%ah<Q
z8eImAL_W>p!1oXkTF+J}dO3QpAA+u4F*9$4TT_*95y<f<NrsO;Rr5dw?d-aHoUy~@
z{8(RksAg%G_v~P}bTgyw2@i}jZUldQ<Il4rOa^tYO|O)i&Qw7>JG2(TiDF^=bYIA`
zmN9j$wrmM{&*nxEKR!td^ta|e?c9Z{K;TTFPU$O&gpn^dGQ*1C)-E;V3NWcVc_>=+
zTb+52?=8g;!Pweoq`U=`+=Y0(TP?O5&E&cWxJ%7X_&e&E6$r=BC>Qz^Yi~HNQvT^F
z_TG=e&$~Dt2s&;zoICC$T+vl|Tar6({_ej`3FJxXUf+HqBzL0G3n2IrG0DP2dEfjx
zW`tF^b6ua>w49cwTw3xb)P&K+0jqZ0n={ikw#`X$QZavilMN<={A07h{-4@W21b^D
z*KBlYNX8J6LU+&9l=9Ys1h|P;{RLGnP#z^OZRTg@TgtE#{rtcIAImQmbQE)Xl$M<V
zMcy4Wz>J>Nh|B2eZu97Px0`2X-hySO^K{Gd!Nh<27sf10!SboWER6hx9+8qK6`e%k
zBq~_AV}0G6SxxIQxt;?q0-sX$zC{S;(x!Kto)<uO5p#!syL?`aA5cfVAkhDPR&?~t
z@%ec@5@r}sK$ATnA#bePp5Fi~EA}aK>N?s`)yNB3LZ+aOgh_BBGsQLv3T|Z{#E(k|
zW)pdj@*W-{Bean7l@AioB1HHJUJkW!>k>KHITx_mZDtT0{d$m+w^y$&<F-GTmXEw^
z*rTy?{S!Nye?EG29k)h3M1hISar@~Wb^i;M8EaZ3CzZ>v8sQCU0yN$_flH`Xp6Kq$
z>GkRR?tW%MU~w0OgYoO`MbeUwuD_j*oET{L@g`^CB}G@XT=-#tGtF9|vT-@J?N1fm
zuxzrilQi|qShVBIaD*U%>JJfOhS%FNy8mAw^)UM0$!ldsbciJfa7<ha)Wh*#l#AOa
ztG8lCXLgSoI<>zg>B}zi?i1Y!WB~X(_{q4dU+Zos$MX@}Z`i<iC@*9}%o7)pTxE-g
z8Kby9;&kfn6B}uJ(+qmjmxwN%-wxGt5GPxJXK_j3$SC(J6`w+0HS8D*n}OEPEC*e)
zd5~oBzp>A?)b-%qu-tm$N`lqLHmtrVGA;&}sZPBk<-K;wLt7(lWm^BxKucbjRz?yS
zZ}mO~2ig3|cqsgh7Mi=)=NDu-Es=##j+!k!@YlUT#Ec<Kp2iM3_&&(HudJV@rO=IK
z&OsHDtPAfI0~O9$HA_z^x}w{J1=$2DW(<ow7HyU=PF2SteuD=UBpXnK#+l180+BFD
zP6sCN4~{@!{x7ZRTxED^WkbjgwV<Y30uxYk#Ww+G><N*aD3762E0*{90#-bUM82M~
zU?scEGJG^zOO=mvMZQ&O_UX|*yD@oVlFXnug>a6{SRtnHo6tNfVRaz-D@xn=Yy`8s
z#^{Q_#!Csf1My<kN+U#RI-Ilp$123eBdOq+?FDK_O42${K%2YG9SXF{^&u_E7{*me
z%Z>4pJ<}kJMpS^XnZcI>nGX~;r>QD>bJ3LId_`cb!5i+dc$qqKk{nI<N_%%a-4UO)
z5_>%2-7ML0MeBCA>*BT1<UHwx)6cVoyuQ+J?+#0X8p5#hy*yYeah_L6ZZ!DO2@i(q
zD^3*xlG}eOXvV`My4;&-tVAYOoSA=DQ4bZ|kHVaF5PTuO8)NsNK3<iHQ$3`wKhas7
zKB}q)l_XQHFpQYw3ZJO@vCs!U94RAKF90)YC3f^bYON+Q3<&sdY}$a0LN_-Xv{o2%
z02-Eiz^X^_72E>+_N76~UDS+9Y|f38AVuFqNyOibb4%1|pck>Pm8Xu`L)nRJl8YW5
ztO_V-u7N0}-MY*a*AMR(o;YrbMu)78Zc+MPjLRz9!B{(T=W0hLL$_|#9sAsr;<!?k
zgiW1J1z_U9*6ljfn(@J$)6>)5nFsGx*I%t+64`&`MGX&|F+X{L%W6`|a;J=3?$u%R
z>M)BsEsIs{)-X`7s9`czFh7dT^%Wt<KvsBkwQ{7XMbUkEwyH!J?i(GT3ejB`){Xpf
zWcHqEnFHGcu&_fR@2XoQn)qCI4vN0^C$k6~8KMCPk<dNs>VY9Jq!N_=By8iz@_HNA
zB;>!gio!##HmyTOW2H<c>4=g>l_U$cF8u6m24Ewl+S})|GtkVnERB4iWgUjSgX_H2
z*%=Fn*YMRD-Tm2g6<{D2tynHpHM?GO&F9c`#}d7?;^NL9@a>O}iT<=Ytrhdw{@}-^
zxUaD#{Usp|I9w@SvVxRce>iVMJ-RAk#o^dP8yRdibEKuGc0nUI4wl+JQz?}}Ltgx&
zGHM*(NzXQL{s(g$TtXf7$g6Su<p858keNS5>4hqK7O7MUTjcRh?c>1Hl~Rx2aiO+w
z7IsC=G(Xj5QMF6!T_PXyO;HAtjBFa)ZvHJ)6sxZHp%?hF!ZP)TpUx<0YeYbPt@Gtr
z<V>yHE|XB`FM}2I&iT9)O*(F+-2|P4b2Rg`IJZILIP1eV&eNw>k?Mnrvzej}ED_WM
zi9(x_P2{|QY=VsR1N#CE4JQZzb`AK5yGO48O@ynIo;V8+l<IXA9R7|kk<KtEWI*jc
zts&m?o&H{8a1pF$8>;9s!Kk>5OTP3bF)f*2@7CaEX2Xe>C1ml%z7z2U`_axIQVf4!
z9*y3{!(T^RZE~wio$3;i9|##z8jfzd=H+B1nqk^b4rxwCAQvWy^>B`r51t%-;q|al
zKi3@!<)rH(TWJkK>FR}HX}?iwiq*tiJav56lJhH#>joWNccEf{MnWyx`%rsqk6_@6
z!@44rFafSFB*?M9^Ad-KdXw>>AB-##tagS95$bq`9CLvog6IBfCN)m4ap#X2{T1uA
z`t!ulDc_R_J;$SW^;h<U%5%b1i~m#Rn+XP;>$yHEg}oK(3H$*y7@mCKE-K0K`c`(I
zDenP6IE5CwXlWz&7n{B;;Mf4*Gpkp$O^KyE*P(-@OpdCz_*9aj5=?1Eyygz5$2T)w
z(LvcXa7!<e+8*|IsHWZ)b97-BEPS2L*{O`L^(J^hCIb%NB<gKIb)a7g(5R%(WKyk~
z|1gxP@I(mcTbp*f)+=nn%`&^&hNa=l@Kr7Mm!6ve=SATtM9D8&^dUtm-*hQ+2N@T(
zm6*s6N!STWz>W1%?pH3dpYvwH@gjf8vn!R~=`A&cl{&ebZ1KN!)CxwDa8-M&70oL^
zkQooC%%s^msOtX|V3x4tlEVeox?$(*7?r)G4}a*ijVvpNRsu729NVGKNF|G2J$16i
zoXT9kN^-x}te;QAtvsgeXexxl9b7%HDVtz-koFBL41p7JQ<VrOE86XxiBvgg)R8>W
z>*F?(zm=pr(SLlf=Y)C$hp$jbcRMSpmYomPn`Dj`afRBq`{f9h*kfkrmN5heZ67<&
z>-L+gqLs(A(hTiIwq=GnC*S)NBKL*uE&`hc6=<z(;KnMGZ5;(3i7Xe*Alhu=H=MDo
zNAA#sk!6xvTB_PBY{-cuu@Xc704`U~&;7$ejERo!KW}~M>HeipW0Pvv->C}d>n|jC
zzojrFax>o`zzx)2t=A~mux0*4%HfA!5e47IL{qC40~rF#sf6;2BRw{e?|G5$ot!$(
zK3UY{HNXL)PH)wGI%sd8gaRYuz@{er+TP_qI#K+j2V?<}bt=IrCwPoEx>O+Qfb~E^
zkhMN{AOv*;b@gEC#)QB}Yp82&moD|ZX!)gn<FP}%AG9ex&c$e!kB9pPnZrIgG#$Df
zYo3kiTZLF2Fkp)-IUTS5zD*Aa0*i>T?_bhhJ71jURVGmlmx;`J$fgF1H?4UIoydH^
z;JL%>$FOpa>Fi`DmCf5TzT?E|zW9KCNApySSr};-0I5yv^M0SDQ=t}W!R$xc8K@oS
z_u=U7QTAoSM~;)ra-QFfKK%Fh!Cr2NSI?AYq48)4cBl~x=!a`5A4}|G2~D7FN!p=6
zO!KCwxD(Y^8+uyYYj&y}3tK!h^~Q)QCvTTYhk3RimL$@<A~5vCAmuBWKu|KmH+dQ<
zM#ogI?51!=+XnBrF1Qri5ME^8BCvo%dqjk4&FoUD^DQ%~Ha6nRReB9A0*i_0^<~XG
zs%<v{uNbBv{Pm+XSjoF<f{@gJ0F(e&fVNBYY`-?jem5f3EAc_E2(nfRuvRg?H5O8D
ze`H?QOQ4rg0^M6S(014ZNg7GD=c-*>R7F-to7PhUOOicO)(+LHsoA)tswVHMd5{|<
zSH3KvgqLI<cTII+ij$qpEYB)k2E{LGBa{HlnBmK9lF@OwS;d_IA==MLbkm_jaJ(Gd
z%(c%byo@cBUq>h-_9(va7Ed>jUhSPR@QQ!*^RI0RStOAbB>(k8Cv~$x0u!_9QdRqG
z=HO<7IoMk<VCj*hHcet~%VY<hNsF026F*<kJ}&qbW&9{)NLWUb`B849uwkij@*!Vq
z-^OGx(>3$KwJww<pTT!<ZIY=q+LwEmo?;RCRB=|Tl*xOc2))kUzM#<<*IrfIy-~4&
zP77!i5K7&#n_of-NTlBNb}aId2^E`&0oy%+L6tU8L>Xk|gLFK_P|#q-HR*M!MkW+S
z$3@Lr?MSeE-kJ@W1u$ya6(-_Jy?>JNp-h5A!Y$3Fb!?lRk&n#CR`1P|G>(h1!>jQ`
z%tFk3C0hH%Nx|h-BERSZE^Hn@NjxHE;=`7(--6LLNB0+Vab<~M4rAYOM`4}U0kvRS
ze+rlp2szNmaMsB{UYs0$E$P{Ww5bvwZ_S{IQ)Z>lW<-KQ3~JsEy@RdFLv!<~5s$S1
z0F^do`8lLvK3cJXSF^k>QDzLGYxDtpG_fvwrS4ThKIpKF24VmGg2Qj!$PCifLS1uU
z<*!wRDQ>H?6TeH<ChXrpjSZ3gvc>QE(W%*~z(X1>IhJ;2Sv2NmR6M4cm{Vv4D0{WC
zCy@8mfb=fgNCY%S=vhb{)qu41187lLQT5MB=syYV^h|93jwE!X-eyh23ODtH`~@u8
zFiH^*0}eciPrW<o%P<hP=hW!;_2cKJba!f8BQj6JVmq5&u;$dMdLl;aL+k$bnfvy1
ze}6l{7zA2LaMGoz(@h0;bNY(fsM<HT-zcv#qN$s*gWJR+W6?PX)kK0xi@Ou@Dp4Yf
zBFF`J2NL*~%F$DATO!5x$haq@lQ%-46<BqIoO9rzh!iC;Kj!$WfWcmRt$Y@Z+X+Oq
zrc+y3m7Y-tj+V<SoVNXi7HRKQX|mTsLFuy!lA1=I#gA1nD0{IyfUHr6NR(9NW%Sjk
z<+Y2PqaAT<sL2}vVlDZRx-M}LEc=6W1KllNs_R7Uo>7;KYZ*~3#}UXYE^Viy&C3<n
z+1}kK#>_%hNxAwFhYh-#+k7dX1o)nNYP~Brz^l_g*D(_2N*$p>0ssT9jCk0klCwgE
zq2GN-<2Gpa0E|aC_?j)Wy~?c3sCwu+*<|QqlFb2_t_38Ep!`miC~VqL@>%Fm>l}Mz
zxz_QEN=4P1>OtP~48G!|8~fLj;feO~u2{b7%_?N{u!dhX0<}Zn3j*<r1ni<s&`u%L
zn~>?;gxHA!B2v}%n%S{nrC?JzL>GE+MPQy%eWtluVoQ07E+xcE5+;m<i-w@8l!#gp
z6H|MMVG@8@pb_58mE*7HxDYw+%|6aM1xKax<G$NE$25g4)SZezLe%36h*=($>n06t
z^PIDzhvS8Xi2>dYrocs)M&0kv_gL#(P@{k_T}-3T_Ro)%`=h;<s?~1(U!jc_C!t{L
z{3B|;L+{G!iXRQmXB=MPR-8v%P5>q6SHlv3g?u#;(pIfb6tnl8Uhyt!J%Sq;j$ort
zZz<O$3BAZV?jpq-5zu+9;*H5gMVu`t(Zm-V5=;~38innuHS!2y0f|lwb8X3bP;sV9
zz7Xa|RMadgvV%@B;*fRI&X-NgNZCZdnjW(;3Q0!n_mLL&g*8dDDC%D3bh>w$VewN{
zg+WCWPq7i^;szn@=9oeSx66%?KV95q;VBJzC#@O^t^F6R42s^tf4&nRbAJ5@7^oam
z+#PW4W9GEn8r({&$G%E7{7Dy@I<i{DP%@kvshzUJI?UK+ZEsvFiwFtk=S{qkyQ;6g
zSl*cqf1xvRp6tV~|1}xK9DiZC8GyCRpXJqY(!gOUaA@cBvZSx5UP-pB!88^#jbUwp
ztwLQtYwap!fzUD>$?n7RM@psnk$uH#N<~8-p?yu4B{p)*4Kr%#dZJZO>fFx~RTZ78
z7z+b({Dbaf_T6r|z^#U9d8pYNSJ)XfxXsP^LUl)M0Zf9E5N-IP`RNFbt;KCfi_NMm
zNYEnBU>s(QgM>*euI420u!%MhsU5GQv8eJKN>FbcWcblC;=$Q(&)!%{BR5YsS2GtO
z40+_pJ-Sy;VLCu+7{uDa17In<x!M`kyZ-?sjjJu`5kLk|{}&}^(2V!el!*<gt7Q{m
z81U8=w*~ui^V|!Qo^FaTPHndO1E2`=-Ha8=a0yBM#Co1cCwpzg0ekaF=|!tth=%Kk
zvqr{919a<}#4%%7lVE1*NR_CE3)2*sSN5o`=2veVf%b@q!>qYWhpuYvTLuf(X2Tbt
zJE)$|@Ush(%w==Ww&?2Xs=C{wpzjZRN6uUC^$FdzajW(h7n#l8KnCvMy{YQ^?C~%9
zw%HF{2D{tAphY9@DA1AiXAi|M;`rB(UZrJe8%C>bLLcmncsLI{A65vf5}fnw1W1x0
zq${Mo)?>uYHjNf~ZDUY$kJ$Fp%V6xYj(3I3d8bN7Gv&E~*a^G{#EiJhb&Z%$7Dn7^
z##^RuF}6o0O-o6ZC*PxOS47U=fKa)i%Rj(+5RrzWb6R|3x!Npx%2{IpocfP-9^Ftv
zSjB@k$%Z!})z3-agU6yJD+X4+^o>4Li+O|6$s-5sxl_Oo-IfnOi<^H`L5rJ@*cK8g
zr(B8%B63%Gzk#cE6JP(aG|>M80GWaQ-|5tjG$yRCnqjj}C_jNkk%`asclChRWE656
zfYOk$kz25xX1{-O-Q^I7C$^`}W>H#!O-5+L2vAXR>$SW!5AVMgzP`>nSfe5kw|x~S
zK5C!>F&REV(P=aq*Z}mIscw2`dA*Vp)@(Ni{1nWwVNt(d_Et8#O>g%gz-9Sbfowna
z;r2O?(wKkBTQY`iwJW$pJa*<5?qo8>H*A+G8FYRX5_SS!KR<l@mW@xwA2L7axijJN
z@bY;-Fw;s*#*@o9O4-TThgrmHRE==fLF@T^>p5B}!5xur8|MN^61RvtQwnA%9j98Z
z78ub~H$p=R%t#ao-l<gr-m7(#9(ve7r7XywN5-5{;_@h_v~a80h7e4v$~MEG<vGaT
zubE8Sff5N1ucQP(kM)fZMtwXOhb)W|O_N>VvZL}Wpr2@yr92&B0y=D0GYa%;iiD_-
ze$h}>s3~Jx)_C&YI(G`zo%f$|z)Z*y;8-bau+Lt$B*;zA-e<e7TJUPsW2zeez3$-o
zJ;E{Syz9)LbQN!C_TbV^0S<&TjG~N=kQZ(QC6&KxR_f8I2wIttNNFc;_f{M#r#yhK
z48lB9*a_=+6fN&`3{1=jZgT+MRVdB9g?aVm`dfT^A99pQ9!8M~zA%GA$s9hok_H`L
zX%egva&o?c!F$f%?3SW>Ns29~T|KF`^rt7!I(x0*1?*>}ad8Yn(E|fFKVH;&rVuyb
z3E(_`Gu)8Bc;hO$SBQW#MP1x;V{kLgfO6b|q7OCw9e_jzvSd;?6|ZatoAY}+Ym^uk
zn+{M45i8u!GpW6k6#Wnym!9%EFkH$Z8FIS4xN?I&+G{nGF6Czth<zIUv40|ftwe0Q
z@jNSnX||S)r>VZa|KmxsDw1v9JwBwMCU(`q?!Iutq<$!Qr}sl&NB`v2x2DVCkaM7k
z!JR;w#;m@7yl}&Lox;J9j7{)Kzmti7#o}lK0iU9PYo&}s=A3WS?iv6@*3s8v+fpW;
z;8c?lp5?)Q-IiOM=|CE~8r`?;4toMvCO=lFj^mZf-3_{z{i@Ij`iA5fJ97&S40C6C
zVHL#xUB|9B{_>Bpm0-!>Ix}%x6zQ#+QENW9;%7#rR+1dh!#c`(jQ=X%NMxfE@{sb9
zBuj)1jbr6?Pu|63*J)yG1$!wySSaoylK$IYJ%niuI^|o}MpED`mkWRjFd}G0pCnIy
zqcP#1&=*~YcjOcz*6jI3Gy&%sdW&ot*@L}YR5?Trxj-n(F=hK4N%YAofHx{ngsXg=
zAfOpHR^ULBhAClKfZtK-YfBUSd-mmu^C8{WeOa1LfVYGYQeGXH4ZT&$HV$Nzw25!U
z_pb`t%7^t{ix(IW&7f~sebE+e2*PV$z@};9R{iF2>;mp9ypa^f$FP&)xFq{R!t;<(
zwtT{ZkVqbi!N)4VS8+!s&r3L<<q^#KX$8u=3G20bY4eJ~j#w70Ld>9jLX=^hN<6lu
zOVCSQ0F%}6*<NO6wpdVxlCOViC76h9ly@i0FI?xNl)?NAMFiK){F+7L^JAGZj2W3V
z${%=hi)bLF$VfRxFVNa7+(U`2&g}1LzCdK8dJ~329Gy+S19jLz+gb3Sah{%1gWCPM
zPM#IzDwNfp0O#E|olomglEX5GbvcLEc#z5>4F2G|qfGFkos?(*(N=yO7ux?GjDTjR
zeR=A3*8`x<o`EtyaEpg1qIEG*Z@2QzR00N->UL;j{FPBI&#?!<BOBAy&@$IB)8L;v
z2Nz>oDWV@p_jlL?tE!=?A;r>g$a46#g@=D@7o{W0=Al+?eW<_Fgkfvx^?K@c2&#dj
z!lZ)_-vvct#ana1+Qud-TjA^zcZGc`<D5(U*_2cQ^!~oxIHSF~^^!}uQ}H+Z4e3;U
zx>Q>W?QPRhz}wk6OWtUP=y1W4fHv*rG8E~B5rXK)!sevz!`H%QfPU%w-g*dj`p~A*
z(>eki^~9cp@20gcJj+3~9KG-GttPgdgI(0uI9ZXD;Y5}eEeielLp?;oHlf(UmbItq
zZX~t&Ad`Zo=Asmh2;;p9^mTzo@>HF*vSe#LR2<S`MLr|!?`=dz<>4$9So~O%9e3_H
z%0lenEJ~1X6c#nT+Q9b36dorAN0~~EC_=@aD><B+pyZAg$1*<M_o}dEC$6NwPc~fZ
zVN`}z<6F~26{*S+mG#y|XGu>q&Dy_@5`$-<Y`wAW$U+CNZenjzI!olkOlaI|RH-Xg
zxJhH?p%@yA;v4ClF7Qvd7G%VF1x4zw8y#pz%1Hj;!P;W);52LyDo9>XwXs+Al{Sm3
zU&f~e#w^&^lmk=r&WtwSuZ8d;){-^$-%<L3gB0y<@FM0KYK_6oeDVWJU16{-WP)#E
z9CxK_YAov*LUzvrT)3mHY$%`84dip&ZNti$ps(rM!#E==yw8{d;#lfRADV4Y6(vfg
z>sxc2U8H2X835X(P#%ow2wodIQqTX&YNXttBt>6SGYz;Sh+INm@1Xf@<Al+3(ekFP
z-z+!>ZypWj+c&1tdJXKwGU{MoTw+GzgEX%G13mKF$yL`_!whDSr6JwBXkB`~MPBoI
zD)SvxOB(vV;nTw}-9@I#j?z|!xoMKQe*OzAdh>XAUd31CM{LBq-c{vtI-}|9gmg<K
zv-wNM!ft|)@baBJ-dopvIyD2klUu1$5$G^mH1(IP7ZZ`}u6a(KDh!vHS~cTD63*8)
zfOgS_9q?u6QG+uijjijC_L5SCuK<6Q59tTUNhkCB*EpY74rlnHKw1QD2%czok=Wsy
zNGDI2-H+cYEJd5Z%lN^Ki3;(eIj_mcVdh$t)rM0O<}s|=c0+TSzAj|z*GUGQH<h22
z=liDq4$zM&(T#^XqvJ>HYI6-YVyf$aU|sZ?=7uhp_@XQ7b6oGj6m#pN)2t}|D}e5U
zl=ml0QoruL4iVr#pM3sP??lhS@b9c2w$#U>4_M&3J}4i7r|{K>U`)|!nXdraKu`ZG
zUN>z4EB=Q!P%@4}BC+aPiZlSrP#tMSb)zCavs=dg{mjve+p8T=M#i-uGh3%eilRIA
zPIYYcuhuCM0cG*=-Rk{GUBc?XDGF5r0*)5-=fjotgj9~_S*|W5oLV9)2dkW^_!;7;
z+Qy`J6}Ek{F4AR@vrC-PFRF@Ubg?>?M(iQS$K&gb=nPE(B;&hy<NfUD>G8zOiwuMs
zsQm23rsNTbAL7=1{xr~$?yvv+<I&@UyGC@HDm<fE6+O1Ggb1EpJqx&44W#jW@x`ZM
zkvC1KK382>)o_sq&jdpVyhQNWa2H-HM@YQYr{q^i6}7&RR2pjCWCTx_^ef@Ax5(Uj
z3X5nW#PnVJ5`3GJjMqjf(?*j=&N4vze!B?m&x2(K_u)eERuN9LN-1Zxm}$C|HRVO)
zKr_XnI=ef3Mj>@@Rrb49viBLcRtl7UB%7MwG_D|)srBCp4>0j|PSP)*T1w%PCt;d`
zgrO{zPX?YWA&s473~)CM`hzr&CwpS`w_+cya&bYdk)EfGNO?>Mdg(Jb5p=K6v`Uu#
zQaSixd=7&(GNfS~rr@^HrcP=wISmZUVop4{3Jrk%iSBz<y!x?C4NCF`e)o+q6UzQh
z(iOih<_7GixQX)QX}k_QR%e8np{2zEhm=?oTZ1@pxH<zwb!}QDqOrgY@;cUCYIk5}
z-RTBlBeOh4_TtU(?Va|eU!g<MM#5@Zu+x)YYhPbiqc#D9{DgG;P!V$QTZ})*<(Of|
z02>(+5IlCZw(g{_ruGeD*R7{88-UgDIOzraw4GcSnH*`)fDmuq6#K@eA*cdQBxl+i
z8OPymjF_g`MGs>%bjq8=W+PVE-RTxuSVrfv9%lFJmuNQ}KlIgbZxHV<`y97;BX?I#
zn9T-?Xv|5bw<eGm=={<L<vx!u0*~t4FWTwps&Y<IbVs~A;U)D_XJGu<m4pP?dLw`N
zXvGecQCY9Wo&{s{xuGYfT5j86Q7&1?sM<W3zh$*MNq~(2OF0rQugevl&vDMzTvIU~
zst{1lt2t3y{$RL?3<&caBZYIyG%1K1ehylxPoUZTCJ5d<mSRP7C=ybTcycrKG)<Db
z1&g0+i;QCI8iDGgy*J~~&1vIF=%DTM)k@XrU(PQxq|C}j(G0a6wcv1UdnPN(bP2Od
ztdyJn`im(C*1LHGO~o&hurWWGItFVmG)s8BM-u@Ts3@4;VuZTOY;=moVgH$t`GDz7
zyv&bJ`%|{bO0R3UX2~ie<Jd&sU*LwVc9Ph|i?QnRfR-h?z^*y9_DzTV<H+cJ-w$G<
zuhioU2f|Vid@i@574ipPk3ye_zolg*A0#+l7JHg3N4<Xi{aBiZqQqBKK;!aR(9DoL
zhiPsaG+zP9jG$uQqwLx4uRx`#q6<^?9#k#;ZEo%3DFvIp+QxDO#^N_5=Fu6TyB>!7
z2}Al5sm|lEfAqY@s8}^`X)f?El$PsSI7{mIZ%rC{1|I;*Nd++vIi1z`AiF%{1At{{
zX_tfX2Nf6fyh4sHK4@7QBr486%B?Z-V@1|0E=rQw%TnG3BIUbq5oKGE{grX`&PljQ
zC2`yLEj?H^F(V4f>p1^JBPUs1yz#@1W!kv&+?DwpWE{ORcmKRJ<8n`mZ4WJ<vI<hx
zxQZ%1gDC~KrNM-oDqwBuFl0ax2nsSJl_!XEH3qHtbHb!r50a{JhbZxG8FJ^<iXLa?
z*>kSBovPvUL?zrJzTc|5_be6;)<dv8G_Y0YK3@4Lk8s1DE=CT#!G++eh=IK0Vc(Kk
zm|j(mD9x9`UWrh~NnVa4w$h*&h|zl*FIsVjp73tqsV`qHf-H|g6vq?mWp3!iIb}%>
z?rGPz1Gpgiw)-%zEcYtRaY?k4gx@)TJoJf4ImX~Uo2GJ3VUzoZbz6}rYm=XDeE8Q6
z**EfWsGjGNES{oomMUg|R!N(Wv!(a2W0=D?IgndCBQ^_ejBZj)i=A(vM=6kpFL^)k
zFc2y7X1h}Ew<i9m<Oo!pUZan>1SFw{=Oba*<18xI^#7_eL-?!iySUBEflhn|N_=M>
za>+sL3|aAhiGRgZHh$_W15~UqNYMbhG2cCCbn%LDd)i?f*);seO2zOGJv!Zg$H3#^
z(TbQ^IvUy2idgD98VMO0*cck|@IX2^+8gOxLAtIR=m7x!!c)?7)YEeW002t?B>d+k
z82>lRM$zeyqnoV}t)RG&h>f-5e?{BS3I0!r4Lv;@+rJ{vTPb--DXJp*)Zo^XHgUX(
zT@JidzlI5vSz8YnYWBVY3OY6$N1)`bikgvm&9B_=30X*DKMM%wJ^>5!XQ79`{u*_4
zNPit!EQ`?nX)7I5GtQJjEqJNlAokdbCzf>nW_#Io@HyFh;5<2=O%xHFW6G)<uCF#N
zFtQt4hxt>r)i9W;N!E@%ma&AkyCyfYCj~i^PhZmCiBV!U>e=;Pu9L7T(>-RWVq$LI
z)Ib&<yb$01Lv1)~Le4pb{<T@SSL=;aq$tmn#9z0l2uZ1rzJA*yOT(NtbyEq~!tAC!
zZH}vE&eD^Wu=tic3le^%X)y>^HVk@+^NWnduZcvCf`v>4RgKJ@uSm|yMHR3PQY~o4
zR~MlZ!PZ^)hlP9tq;iP{f+=(xBezngk9w835AYKwWEZsC9I$fRw9a2Sf8>{hmL_Nn
zk?*D!01-dH63<Tak((P&&jOKqmYXu9$hgA~aQuC|cU#4Bl7@q1xJo5eQzg?IhF7k{
zBT&e0L&s+PSr7CNH!|CxMBXUAt0#@_Uswi0x$in(lA?3DpSc3%kpfYWggA<fAbXkj
zIcmG$Q+B!F6Gj%ZRHFcTR)V}#wjW9^Ril&$)FHP0IWD5iqfvVlex|OL#Bx2+udI&#
zPmBt(T_Yw_14w`52sX;_JS-9?G-^&|vEt`jh654(tDHs{`LK|{dg)0$1!b^KuyI$d
z;)J^;SN>orG_Ghg^%adYn2FVTOSGy(t5ESs#NpxySl$qW%T<R}j!Dd0?jdHX9n;)7
z8YbbWfN;j%7bh<|71UKIP5;f{-zWX<&3(?1XKnM#AJ{o;LaxN4rxvbpU9T9;(f7e&
zs*po|_)NE~LsFp(KB5t`l{O1FB|K@bS7z?IjGNlHYh8@}YtOEqO_%MvTXEc<+pee1
zg#~ZXPPzG{phcceXUEhBlKJjJ%MM{@EF@1+{dQV=qe3Cs1Mtp5;;x}75J6zGL|b|C
zrx?34N&~|J2JoAXBSUYZH9qu{-R!Rpzhw<Gv#kAY5<H^|$!^D>HI~2m%i>MV`9zRW
zYp;R44$#}bq@r%}?;Q(uz6;t?2p_;2H9M#SRw7Ylbbao2iy{o49p-sjU)Vf=oF5-f
z4gNN9){w#k9l)E(_E)Q<Lt7b1%~fid%AU=atvNldX0@Pm^RiK*8h_`TU~kyeRgE{b
zTFj?KiB8bQVeLc-YX@c|=q?Y$>#E_wVw_uAT%7F;4A!MvvI9fZ6MeAFbSca<c8NdL
zig8C<MO0)bg-#NFgs{y?!2Xy(=-py{%569SG!I3Q0@r3s4yI`nRVJY`HC%*PfY*`*
zh)+s$%?CU8E&JBdfPJuerpvcuqxI+I-GrRh$YYobnVEY+Zb|x}@Z%$p?14iV(^?Yd
z6wsNQiEilH5?%Q!h&oJ^WR|rRwk11hykFlO2Bhp&P!4<C8m|~xfA(grLOg%k8o1H=
z6hcl<vtq;#m~t3DJtTz*bY*1Hq?my0yLz#%+jF6qwEn5!l;wpuFBwgMXFRfGT3=r7
zuBmEpxGmSAy}@N~<;A&LU%`nvFZl+IoX6Jr=S0Z-zZ0R9k+q4V>0baY<Nrvu|Bqbj
zrJ!xmPlx0+Q`71J_n61>WKfSERF;5z{A)R?Yx(dTh=ttjW4C8V+8i5NJ+FxLWHZA-
zcHV8ZxG#d~7j&=}EGLz;6)4z)vP!LtiM~))LG^Mdg3(0sOl^}$p&GwSHWfluIZhZN
z^>ovqijywWK*D?vLd$Q*msjM@qKrdAm~B<g1QU+^zR66ZUhOFzWs~`DQsDhvnqS>R
zr0WXkU(oV74l+<R8q?C(#+2%=PrvjpUP~JrRTFG&lhv63%7mPRzbWk@$B;`>>ifCo
zL<fo$!Ci!sle%0-xKXStg(@B)+Fh{Y392IYBd1|(QAY6Y#G=SEoRSFxt#5@8D#|Cm
zK)flR|8in??G{s!MgFK80$CNyjus;U#OSvdk{XvD;byZchZq2h&wk5jh<KVTKcR~(
zhCIx%Ih#5P%i=`+!igI!C=zWvoi!-H`pum#h4DC_TT&c7GPUeYdomy76(O{^tbp?^
z^to!SS9IrW8bq#@<nv|L;xe5#75@dGmvT1rk8%C$oASS4ZT};#4FA4{>h0<Q0D;1T
zq2M6`0PX;Y|ML?6$JY73*HF5@td0Nn{XaHT)ZZG4X@u4jkQ`ulfz?7`ws-Bc<`_7)
zf9!vX&@>8iXi&dzYZxeSBXKq2<g$W1icls-)#V*IkH2B-DfU*I40{YMSf5HwVhmwy
zU`Mkw893;JQlC0`XzduL(?XRl945QXb@Rb<bK~XRbMkiJho1)Rug})f$?wwoEvD}z
z&27cFDj9w*p0Hr9r(kom|FG|jGc?2{HW?z!B~3IhjJFY1JR-LysA`W-E7(Lc{(ujT
zvl)LsS0#YbuFWx4oyV!*H#3(dXN&<N*e1u*2*w>~bW7q2^||R2dp(DQ+M^}H$${Qb
zmIdq)9`SN5^QMry;7cKx5#(Hl!brCol`8>k&Xk+%EH2NuD!jo@xn^Cf;9KmQZ3!Zi
zY^bL9$FwAXdKSQ|BMp=$y$-<4M2yl*Pn^(pJ&Fjp)fu64%U0|+?5s1aQpScDr=)*A
z0f1UO$Q9O!t8&FrCJD?`YGY!ce$(;@o=V2Ms$*`oimsBjx{8iRsn&p#t#3ZM9ua%M
z+PHak_1#!kvau!DMTSUwQYBNo^Yo;XK>QIu$0>Fx7zd=?nZgc{ARf^YKRS}XcZle`
z{Q_*U4NA9C=)^U-wTzO?5!EM!BawHcKIveMx1tU`E1q?CSpMq}z80Ba_t8mo_J~21
zw8`&s7u8gAY6I=KvADRgy=?gr0DAbJ=o?ktuWSnZ%Fynb`2q*S?nl5X8x{Snk84Is
zURs*TnEh4Fsy^}i`1GEU9cAJR8p^u7egshm14YZTan_cgaJXtX@BM>*_svFET|WEj
zZ4MH{%dVg(sDzo}IG3$YVS!h~KF2ZCjw6yojn)_NI4l7=*@%pqera$YMpu27>!->6
zRhBMSS#%=;6h_evYQ&NJy1Oznt-9~`f=g{hc7FzZWXdWlx7GP}1qt1w;~o+aSaqv5
z=x5f(&YNCUEf@n1r!$YK5VQV{WWTTb)Rz#y7~*Z&xE^)d?4Ikp;5$-w3OLKN7*5Fv
z*8X=9uC+RNTfx+Q{%77J6rxuU7W`_SqNg|>uA-awpB2_F`Z_i8Wo*<xG-{^aa-~=g
zE}p|>ZP79%IcT!a;|Ayg-LoIG`a!KE5Hz7s>UVQ&RMlb<l<x0@SaFl6;T|$$z!Q{v
zbUd8AoGG_P!*M+8+b41xk=?#gboc|=d@rID**6?>K`A#rN7b9$r`+gDS?pW9MC31W
zg((UfV!FLF`n1BXnt^c5BredcN7<|rTU%^P#>F}p6HNHuY4ks+=V?$nGtls&JdZDd
zG8NM>uvyr{Hk&$I#Jnz<Hy6rnY&WC|W8Jq?d`j3+?%eVMap_7kq2`aEX1jbJ3IiuH
z(H25foP?{gL81(|6kW;@#&@R9o{9vV`!63}fIg$R;W7oMN<BV2r%=KEz<qG-Z?r9p
z@M`~@kXZjC=P~{j2K=kk;}ySQ-p>aYeDgwHjOwp5XsJ+9C6=3wH(OrHJ=@XBxenTZ
z$ubsXv%`%Rq<|?ln(}6HoU_(=t&*{uR)G!nq77+5rCfy$matNZSGBBE$04dt4aa^@
zq5-R4guI~#dL_;wT~x{;k}2U_$<9*VBW%D0?v}L5GrPTSLg#_$Q`zI6veGop$PT{j
zS{jq~*WP12{M>Z$2dYnKPXtLT&1JDj&K_JoMZuhBk|?KiQ${WR_7b6g_(eyodPYUx
zBr1Unj7H6g<i)T}JxO{ya~$$P=^7b53XdsESp6))VGDmx!C_GVC)aDBV9xS9pDq6h
zIFSg#b~9t%STur46#VLyoK;>v%lc|By@4AUyzXEW(>UBf-4NP|<hPiJF$9y}+@=-;
z|CrBE+06PCHwOR1x(&f4=R&g#6TK7Jm1tOz&gI$I!s^1}-+yv0poGnt>69IVi=;cP
z!rE&MS$DbDylM(|lf9MyDpX+qs1#CpmCB`q(fXTJa)SL0aAD<&`5&Xo^bb+}|Egqg
zDsXbC=Y93P@&SNgXy{^qkpR)uK>xgjnEsDT(*J8sV_;?fcZLH|s!&RZCMel(B!ci|
z3?dVeS!vDHF)Dp08db|c;{;EBKs;q!f_#WT`1tqHc(H~I_?>a0cxXC+fNrlsD}U8Z
z6q@E6OBxsEEge_aWy{(YKjJ<zGct`$9VQp-oMUeZ4nLeP-alPCT`!*mgA0i8VM5>f
z7XtZ$EZz0`2P#PnJXUl4pqfVZRRlU$+t5e4ySj#6WZoqGytMn|zq3&uB1to^2yyQ3
zGQopKjPr+a@!q-oLosrkY7P|ND0yfhe>}^*jN3p{JiP;4-S7woar0t3Mq>!1nA8tR
z39^r!?RxuzBn6eu*3OmO!h_X2ZTI5euAM*;v3dhXNUs&o7+#7P?##rcTmc&Beb)--
zA?B5gt0|(yV~YkU<I$BF5Q#VXOc{;Nr9`S#%oZ3=K<VVk%hDocinjmyAepBLpyZHC
z{%JH5XJ9T6kE8F2h>rXtplpU=ENGrjMXfkYrPfQA#(<uXKUY`-4ZUc31W3`uMW;SH
z2Cc-UG7jz7(}L&tP+SU%xg1itfz(FB*PuH<RIw-(FNK&>J_`MsBgg}9sZf@57oJO`
z>~HB6PbHmDJ|+@qLUIp@;0r<!5w{YnM+EJLxE5+D(u^lRswsvQ?^{SfArxgA8-X)1
zBw!k30>&_aZ75ra(MB5%|21UL<=VDN(S4Y9_U7^!85rLvZmnlg=b>ZK`JS_O#y#K>
z<U)M<bOht>5sk?7{nWTsbHX8p^|R`SdfQH0bg~q;isBUJZL<bJ3HqlN-dT=^4~FE*
za#Ok*JW9RBa?I?;(%P(s^2%#M>cuH?`8l=#H6CSQY!!BZV*spZ1+(n@st%-D_buxE
zt1dyX#1aW7Q*9gGC)~2j@stEcmy>d6>OoN9<x=95^I>|Fx;Eec{)q4qcJ^b*SJ3eg
zW%WBLRTqdv+yb^Xn7}|iEDll(idKrGg_<2K+3zSPCbPnOhv=pN5t?`$h@e1yWo>(Y
z@vASrZ)I<hFSM83cD<{u7YdHMw<P%9NnO1}@(>B|eEPH{cUQB{g7Z}YvRMFu%HN_C
z;`fmLn3D``2`Q|BeZjXAg<_wFST6;iUmKo(HY0ZUi%s%P{=AcB%Or`3<e$i-E+!wO
ziz5nI`5lRE%#~_xN!*B1o#yo0qnWcS_Y*f#*AdwGtZ;bVJ+@|gtzq2k>%nDY)#&+8
zU}&4bhR0H%kJ|HH8Vris@5a_=Qv~-Ow%-!?-Yzt;T0C}zBm6ZZx2(R4j2X=54rHFT
zE%q@i(VK)eyy@a5(3mvPh`^ge)rg4=Po_5;hC5>0IHcMc0~ifqBYR5C{(2py4=7|9
zALQ9XT4r%|g{xvy66%9EcrGN)zfDb#%$zH80Jsr4f~>kwzhE*UyMRm}DDSo=0D!wm
zeOE<letj#r@&O$yf3DdDeyYL{C7AjN(dEHddc6o~8er;eKyG~Y#`D(oo20A77`XFf
z<WO2sT#C;RB@+!QvMs(0k7za;@|m7q)H$aSM$n8-Y-b3Nog2wuH?f=OY7xNO(?Y%i
z-TX4)y_(?dC`dTpY!2EIDnC$r;CI(``)eddX0nm4LAJD8+b819&ExHb=fbxof_rw;
z!jkX`&8<16A(hx&siZ4*xQNQKyh^b_)`~(f@3_#-#jLDWDNh_Pxn{G$G06=Bi_ueA
zQ(=CAMbM*nVrsfePcM$q<$Th)mS-}P#l>N}De%Q_|I4+0!c=$HFpG=g5GfSree2u%
z2<;l6GIr_7Xqi79)!-F$w6(##s9k8Z9DAd{WPl75jEl-tH)qZ<a5yJo|3;FN>1tdl
z53{Jn$Pw4<%uy}|8tM~i6ll=k6`&L$GWcu_k?DspIOgY<GD&(i8nULG?h0uPz}b~|
zuL0KR=8hTTo&3s1|HsCXWlI=X+$#2R*=b4o{Y$eYe#(e-^3kmB;jg7@`~ZV}mh#gZ
zx=#|%Jye=z9y?zR-x9zt2=G<ky@KD>**%}$R*x!>$GplHD^7BGRNt5s+e0s0Fh_`n
z2S23s)g^9DUO)l$y(WTqdf=YLh*>VgVcODqevL-F8KV4cLaTgz7~%m@R`5b^Oj;Cg
zNRhe5(+a9Gizvz$4ltnZz)c9*qWXgiEllyBUzIW@2AHOJDMbAMNY!zN+$h-jeF|gd
zZ2swy!Tk&Zy4p$`re!rva;$xBNJ3m*l7#F;z4>_%Ad<Lgk***sTdmli{h~%&UXqhH
zz#c_<>zN_844Ll?ND#`DUbm30U{c*asJn|5$_nh5RZ&IHrUnbn@pg4pqzh(GcFzjD
z^rPKguWU~B2TB)kZl9?v)mhdt1}&qT*wEGv{%_9PJ83t!9&E4}HhYS%#<{UTpGO|E
z4o;qxDxy*PudFd?ZcX(>$=PCV2(GeFd_Lkc6qOhgP0>QSgnfm-T>KKfoj5W(t=;7E
zD~;H)#{XJsQ-h2S-kO<oh;&TH<LYVv$>-I%KESKz%8~;0RS(Uh(3J3tp4;C$Vp)Yy
zt<wN*`+gzU(egf+q;q3=xc)M64a}jql6M>eofz1^C*R#8LH@ERj8~h@Q>@evbvD-L
zA71yy0MUEh^dQvLsW?<a6h)V6?}rvxN;k1_UVH+72%q0z$Ncf1s(E1&W2<71>VYKb
zWjbRwa?!_bY9TG7=pLu6XQ}+w&ad%$g}tq%QLs&HD@`^Eb6++uxG|8nEb3kyMM04Z
zt_K!20;`cx2g(o?&{bD6lqe_{lzcICO!iE?3G%U|{orB^T#buwb#MFZzxq<@>QnOe
zrPBLtyl~4-6~e7%!JJ{m5`Lgoy>^W3^BGu{1hP8D>rFH`uZsp+bH-PPXgT}Fji$;8
zu-EGY(9(x=2I_=_QkX~_Qs-ll?~B-Z7)aYjX<nxBM<dVhILXb7e$x%pS$wxa7xh{9
zs?%K^twgw(YQ-jtAj6_%HR<?Iq0D5d7?V+~9C$v^M^_oMpdb>aL?khirvm|$$UT;Z
zeJHg_ST$5-fc+mZH!?~>{OjfJKL$nuQHRo>zSlL4|Anw`46Zb2w+$xN#G05B+sU5T
z$xQ5IVmq1G-mz`lwr$(Cy-#l4d#Y~L`Kr#J{@K-4Z+E{>ueF}F@J>f4C*6jn{AE5L
zb9^H!QN}+zEB90BboMQ6!&0>c^;O9MV;6tYjL|4V*F?HTGBd>Ds|RNx%t^{K5!|M0
zdC*NLc%WCXdN!n?l~Rlk<Le>tP1&&Krx#;2j1E4Hs_$D~vZ3fJ3|elpoN|V*hYwSD
z`Z{^=o_nrlJjx21Gs}-?zNtM4f&e>UrHtO>wMt9zb3zpxmI8a_SLo5V$SS4>+qu=x
z3pd8=-Af3v$WbAahplXE<iCDD&E(|;uYF{E^dS|E>+~|S#0zc5uW>o}f}xmbL?}bG
zh+O*m&7zs_oi04h<l-fk5#!;NYK~**FoKC7nyQ4SNMp|f*wQ$%a<2r<#*G(o>K44m
zE@2I@O1hh!i&`??gWF(lIO`MrACLk%%ARJ$=Nq_96FLQ&0;p+5r8)mJ3la<H3-UtC
z4qC4+8NERu0)0>sfiKq4v>m8kMhm6a*UDv;0b9*>2ct-Te+$}1phO!>Bd!$S4v6X$
z6fj#dz!gbXb_GLBX%4wDpb-UR>9p!=&Nt3`C6-xKIkI7~8U7ZH?I{#|W8ywsZ?I_M
zqx>T#3@;}PM5!(GfT)-7*ieo9$#=r4*~~>eww@Hi0I0}7F@rE`z;uSgBrJfvcMW<-
z*}o^uGURkHj$~D}Pd|u_iZWPM=uuXq!?F-<ncDY;B%0}F{EyW;`~U0*V&nLKqJ#g9
z%L$B=x7lU-K5+Gdu?dOBpKmvw<w%QB@@-m7V-<dRLC2r}w{(AG`hItl?%H0xg0d3%
zcfe*^%Fr_Ihs(y0ybuvd=DK37zsggPu$E;RC0>L|>JxkMCGIaDocYI?cD#wF=@`2|
zl~NRKycP_YUH1^sOG||m=>@OV&D@3_KHfJ~PB%r%gI!I$+Nv}KH^`G81uV=?MldVd
z<sbM|H7qR@=v&8($tva#60`17B_Dq)L~<;isMe_A1SMHS0Fo3k64H?j?(r<7<W~?c
z@;Yl)ExR8L7ugI8lHx!p$#cUM=fhKnEzuDms$M=%+)YLFV*ORxAq2zsoIjc9;8bgC
zmnSWyzU0=B1=XeJm7J<Ke;k`RekC7akU7w2ThSmFDwu>|IU})A0V6yQP^oVui`1O>
zNHqGc6KVu^Z4&GvW_vku9WxIjm(k4p>L1xMNUeSgr!wW^Din|L$@)K{pMaPAf}g~k
zg*S@`K*y2GDk!#!F3e;J3vdc|O;wzS*tth+TmGKO^zegf{4le2yMJCi`Fgu#n68G&
zSC*xUUOjT{e$4^+uj%+B+tntV-^hF6U9y!m9)8otAP-;lVCxePMLr=fN%o#dH2@d&
zKktzLkBa+mr1HNX?f=^j*~>UN!v=vsx*uNy7#MXJ%zs8h+5aaR`lWs0{=fbJ{|hL2
z(T?Dvw6FkZe_A<8Z!u<Rfe|vE9FIn1Ch`5%O+5S!_PZ~6fRB&8Bnu?mz8J$VC3p<X
zYiJeGQ>d6+HFL}o1&k}P3e0+y@&@w+h*0IQ#s+F~?ibLb04qs$!~X#zpWeT)NC9Mv
zM}edBj>p!jh9|`DQm{2~P608$M#|II*Jf=#;WB?fkC&8|nQA$Xd%}@8GLeMlm8vwF
z%46~SzU7zQoWD{;NS~>MkL!v_+!o=0fCUX!(wVZim!yW>^a0QE%E2^#iRB}VXbRPM
zSK`Whz6W3*h)-{}facapqlawDBh+=GfseZeBxTGrOTen^Z~*U%r}n_9*o*li2?_d`
zUDbl>e4vIHQu%-<0!b!Zglx{>2`dqfpu{yX$!$}>=#|72THtr1fLq1i$1EC;0_vm#
zXrwDLScRjp<o{ytH@D*Iw9!a}eNy5Xej=hpn09oSw-d@=F)z&Q0pULKpV0N+!Q~`E
zqUn4RzTZjo9*FJMUB`zV(Y<SNq}9|$22<wlO++2j-gA9~W#*VrL@5ztO0nh_6(r?z
z=Km`INP8#9NrMdcRzx_Gl&h63>Dxj#dI=3}Y7A~kTcoTLuM&u2y(x%d*ABLZgAeMY
z!edLv|6u*ZPe@aWGgrPa_e;%fU0^=7bn4V7YAjwVJ)1vUh&a7Cb#`icb$GCQaDAwD
zTV&q6F&KJ?l#NXL(ZqA<%(wlVvPUvRIK=ovo;TC!hn;hgR7v9e^L)P2KXd2P<@quN
z;gT}ND%lQ_hout6=3OmeZB{MiN{%IY?Wa~k$B)^EfHSithsq^waV?@Yxz<kGk1?mk
zjBPn5>{gTJD#s}Y-2<I%ojtGFjZ3ekkDNE@rzLCcI&%Sf?XOyGX3#=(mCY)qQK(I}
zP2OF~UCe!w%oRhvo@#`3I_~tw4cdK{KPA(0N@+@Fige1@CY`gV%++%<XVY#I?`mQ*
zxLw{A#ZC9FbMK?<NjmFP`Worl)kD}LrPIR0!UNxf^X>AM|IppNnvtLC<M94%mPp3C
zm+|c@|JXa#JE&_MFw`}@b=oxtY6d}p#y*$ciCzE?-a8;l1I59|a7LC^29l(Gty(@-
z8V5ZUrZQ#=<wbUNi@LbFjTN(%6?e3!=BMYU^~R$|e)p$z{yX!e`G(H3%1W44ua$?T
zN}UeTt3`KrcQ^OI15-mPej(oDrF+d)dZa{3sks7$lj!?YN3~b~`;ut@+*wzkc3)-i
zr;P!?pvypaB*8weU3rt@Dj_RIH~y>PDt=2Pee{=6GmCu5Z;rB3^OLzQaim`$U=E1m
zLHrc@VGwu^!o2SyeBGkH_b%Eiii8RLeB$vs;o5(WxcfKVh5ITmBP%;0MUd#`YHoB&
zlLK;B%yLFmt+hxlTUWklMSk_}ekiucI`cVGU$Us&YW+NZoqtF<t3T7^w4-Wce^UBq
zrUnA8pRL2JSG9n)KbJpHo^ZL6(O;)#ZX0Sea;ClW7IUo@oXuhSUZZZwU7+tzKxjE2
zg(<oz<0+M6z?YKAUEB?!-Dy*^<1Gi2)FIyP?e`fvbuRsgbXy<@a-9b8Y*8MAbhb48
zJ<Z5JX74Lbc;}v8N^~FAK5(bK;qC!Ty5Ju+PXO?bE}HwT9gynaV0c>&$Uz7+#*#3u
zcHG5K#$HB$#*^dCcyF^AS3A^Za+g^#4QiSOT+p1gmRN`{ZB|aTwmyh?Qrvqdo^?IB
zT#a?eKYg}7jcn>Z2+X{>Tp4sk-3zS0{l4<=K>0fG?z&fxo4oFv51Z3%xlM&nqnlqG
zw7aysv>VXf<UaC>U@2`5e+hrV;>5_UYBTTOiOIv>Yw)Yh{nEYagL8+(SB%@t(3!#g
z%qz8Lvw#v&uDer`fCN!7%5+9+Ls3&v*^$hY2|v-}m%XhA<<s~zE4NAQlwpP>AkNwN
z8g~o#3il%XE&NPoeTg}11;Fa`=;-{<Ywkt#=qSVcf^YNXZF0-0bF-!EE_4zz!zp8o
zS2zPR1IUBzth2yfSC{XgVe-dB@aTE6eG)aJkJo`G*!h)rj>j`2Q|@0;8Eb>XZxg8_
zpUM5n{K<gH=gH;CWI2<fin&=yUP*DAp_a%x0s;c>W3JU3>!B;2%~P)Rnv%W2$OeKX
z$9kvoQ`1Y+Q?;2tTYuJ=%M9Fx>XB`RvLoXoeF;Vg;0R_29+9PkUksl{9H>BEx4^g|
z`OlYqO{%DLkGp+*DqY?cuiNR7BKgnP{VKUj{+H<yeECQ&yMxeJ^f@^he&)kx>>&W)
zbLD1ok6=ii5H$96Z2`DHXK#+80RP^nZ;GO>1U|QK-XgbduwRx_4BDcscpu{odJK|0
z+}?_syjV`+YO=ImM~>yiJlps7E{JU2o*TCnd#L=PV1WLtey++{6!|=$-S$4$ItTx{
z3GPj9w$qkklf2BDF2ns`5NHZy0$P3vzF51?U|<0+^|d9aL1*u!HKm=&{mdBx=Y-$^
z(*?hT@j&8&bomCxcTcyC?PIOPLRjA=0&JC``32IhM+Xl)1`%91b|XRtD}CvDx+w@C
z+P|qGx*%Oc>-_|K$n3YHgLwzPHu=R0=9F=-+U?{s_ZKW<7SLBcqZb?@?v1`03N6tk
z$h!pA30=nY>y=((-r5rL)h4$Cx*F@FX+{U78Sdk0dQ;!$9(8pM3?&!DyqX2RDMZ9q
zy9FKzt^gjL(hh5dtTC)qDFgzqm-R6==xoQdt@0HXi=Y7neD0Y(L3jlU{PKbZ!>u`*
z&b3ZhoACy}I0KtRzciYr7uG?%jT3A_>|knFO!vXLC&4x%_L+NNtnUMNm>%LXA_Q|B
zwXTm2hIkn{G3A8pZbQr{q3I6rg3~|%b1L9Nuh;-{#ajIaX*fS%9tM7ES#J$#SYFgC
zHqU&l8_EIJ?d}S-`T{+yrWr_3>~t*r&m0JT%TaITV=>c(Pj9KB9sJ2}$_bWDh5K02
zi>xFE*e~C-_Ce$>TSMArQTc~YEPQx6g%tf#H-xm!^E-?4mlw{(U&wua(;P)imbky5
z$Q$V4x0y<q_vPu7JvYexqG>-YnCEDe?2ZvI+&74!3h3c$aF|ET$TlQ?-fC7`Tl}6?
z<mroT=Np>UV8rcL=X2&?K}fuMRLIX+i1^OkPe0`Gowt2DxH3gSko6wO@9Kk@18{GS
zy{bmbrlp2HyT5=Rf;9J~B6$&<K5Me(q{MKZ=Sm8Hc}v#Cp54#673yY&+UEcI)C74J
z=pZ-4djd3;J<MS}pS@{3XffaD3t%rApsDAkH}!n{mZRYT5+CfZ)lEv*G4S}R)lm*T
zN!oVDeR5hOJ~_t*kqBq4Oe#Hm*p`z}AO)r=Wh9>p>gXs0=3b~2n(YmzDscGwBfoI%
z=qh{Ui#)Jlxf4MkfA$&N{U559Y*y^+lHpW_;aMn<V$G3#&yZFX{^z#ey0$oCmhA_s
zk4v^#j%c`jh6s^kTXmD8r|&0NeTXu{a~8aM>vw>5wx}%GH+>50OF(oQK40sN4nZ*1
z%?i>>Nr>)c3D_l&-4qqDu^x;!J<0kjFnmK5H(O}&5(?V3%dW3$!(v+8(xK00T|O73
zrMV{$|3+B!-QJ&(bGHJXa6P2SdRPW(!`cf(@fu=t=?(V1OO{2W-<s1tWW+EZ6^Xeg
zz$MS_ItK-`G>gPU0&V_7EDxo_>;<eJ9Np$GFW&pGBd*4EX^)sp^HL$MUTT1I3^?Uy
zCn}gGldKqUIhzD4jTj%PMyU<ObJ;gYtFQ+o<3aZ>>u;0t9l!nvk%^?yl^jy$S~5!C
z=ft>vIt#{2PpJIq7Zu@hvl)WM@f0LJ=f^jzQgF)W${`D%6NY;KuT|){eyydai9ZR9
zWkwwF)(~G$==@fOz583+{jVazekF95YvN2dt}fnbYmV-3fF~buM$6G}X%Cb*dO>_0
z*)s?f$*4;m0+8d;Mc9w~fnCWY23HbbmZON8x1<2~zv9Nrao9DB2%QtOGo+NRs3zmv
z1Hk6PK*Wit9DY-u6o5M}m?ksXqU+zc&41HJSD<gdbU1`NHyM$b_V^`81olx_Eysho
z_V(DwvR?I$1cv9sdHOhEeb)7Yrt};{JcOr%FJ4%Xm>`oFQNZ5V$)MjBarAHlu^!Kb
zk*m8KJwh4;uJysMI@=JdyAwU4t57`Dq8&Wak@qDK^au5_z_Ioz>7-}>=3zZ6zZQUP
z*N}P@2_K$Jhhlz4BezM3uUGdL&7soHqSBUUq3?=^U-wa!e;q_sz^*T*<7*85$Fwh|
zz}^&Ccc#@f3J4L>l5Co^=yhy5*z4T^vHGtk;1mNSbiYI9wQuB2`<lJwT?EdtF9$du
ziAp!xyZZ$sVxCipdNhKxjE*6C=|k~ciDvyE!qxk73C_aP6TZC78@v<V<k)8d?8brc
z?@7iGud{p?q16q$BU!66kvL*+0>L?+MWOXg#jsza(Z{uXkOb@{Bf%Sar%5Aq{dgv!
zTUmEZQm&ctJ;>2rDbNLsg!v}nzh<_c@RK#}`DZmep}x(FP%{yPlM$$v2!&rQ_^2Lo
zp<X;^Sv_)Y2x%E&vSJNsIZUx~J!VP02>mub$<ZC8)(PX>4AU|Jt-0aP)esydzPavy
z<DSikZ|R2OVU3~>yzs2wwn8z7!`I)*?0{Yt5Cr0M!Z5Exb7#6PfUW7mp7`-!`}5&j
zwjs1$_=Q_lUoiBiGA!p4e1%?ZT|#U*FMDENJNW{5Eo%@O7oY&O{*es(rwsd~yq4m@
ze2)FRyzlD4Ri`c{Q=)4YJ_IeL*!B0MQ7c&__Z=?~As<k=QYt#vinkdA#1sn6R3E_|
z{>*WxDbykql=_rd`V<`JKXXI;MbN22NovR=`{TaFO2kUA;QoS+#b61^Ck<?iA?51_
z>M`L&Cd$1V!jAZq7{=1E#2d!_U<pU0YBBtVD|tIi&627mbvuG%kBcik7&A8Ff<_Ew
z2$3Q|86(7!B157miQ3{PLpC>ps30RH@iRqobeK*-&Nz;HUy3DZoa}BydEdO&hmIUL
ztPep%${8DB$l8*!F4!1Huy10By&_Ba6{5<&f3%Nn$mkmX9?K_3NcJ|2zOTCf^Ol_Z
z7jmp_oXH5u?GM+@c*#_W=m3cbGir5NLh=Yh?%hwR+Hoyuju`jxq*!BzJXa`nDU}rM
z!<<8zLnasr`WCs@AU>f?k~2xn63(dngG2ker1=>3;qZk0p?zq3&WIv7%A^si+q9r$
zkHVgPv$EiJ5!b}_$#@$)172(1zoVp9g=~sJW?U_q>p~pSnj^A?2rE+7R828$uzX?b
zl9yypaqc64f&vA6b&>`t+5+W$yc0aLhJewK$ZUaEoDNJ6$(=Jc`k?qqBck01A|^0B
zO4TvvQ!?@EK^!M`bqW-{sof1&iE6YfM=)<q+#d3^nJvn-8pzUb;%zIaje_kjQmdo{
zF^%8h#Oe{A@QHU!h@UQ^&jfDSW&7q-P4NiioTq>2eZZ4B9J|j;-APjLLUa&*PjyQ`
zYfUT7#}#`fbPfDLb`?7}>OSgRi;#8oQ#S?^<4<yaDa?XP97;NSw1XjX#K4jCew9MB
z_QYkL;>g`8+6mzarfW^}Z(!`)?80orEdPFTd-5ekR`HFR4P;9&n@^dVeLzHFQJl>P
zZ!gUIwjSe{qBT3iv=!wQ&6VWda6`J(@0vhazXek5xF*>)C72233S~J14rfp5OkOW5
z1<Vvd9Q9O(taufSb?jtr)wc=0sp)c{HG7fdnz+=HXoDGyySwz!=+Du1&(Um8x`}pq
zxjTZle~eK`6R9dZs%uaBJNZnDy#sOAmWx`BF9d{pt~=y6k&LbKl6_mb18?=E<`<Ml
zvYpJF+`<Myx6QvRY~soiHm*I<adD~*Y;pzUC!`9<<I*4Vih|IOW)#Kck$}LP{(M(`
zz6F*OYvZ3AR1a2^Lrox41=j3|J&!Ep3}%st`Csv&!-6B03e@vPPJ{&RG@Jl$t`Y8=
znVr8F)V9n-c|#?#AnXQLou?U$kNjJmDfkZ^JhYn9xC^5U2%@5qLNNJh8FFO~-Y7x=
zHyk!k{zuzOCOq`P#TDsis2^97DL<GTnm#M};8Imkd+a65h~tuDS@)modxXaqtW`a<
z&?+9Ae{fIs*kh43He+yzI56!>g&7zAoy?OoK%qKtZX$Aa?pi>e9ct`Kbmw_Osp8h%
z%VL&%QG2sb<z&ei;4&5R&`*ytQuGc7QVtmI-4xHblEsGOe}4TK*(G5H<r(<wU!HGl
zHl?aOyiR75jJsZUcZzHZ6umUuULMPIJoo&mD2hJarAL~_q>q26>SuCMdb=%MI8u*I
z=ASmh#bdqbHhWK5Ja&D^60&6{A>tOEim%%MTHl_qL@f7lUNE0Skz%-^u=xDO;Sgmd
z33mzgh;S@$Yu5?d*7bEK6I<$%ijZ))*3)8t$g&&=_8Hmum4Bt0v&_2If6R*)r5Axb
zcqlmlS&+>FAB(trwzuNUyDQbrF?nM9lzv26oEdJ)wgRO8F#7d!B$Zw`ulT0cZToVB
zIoGcd39@uGQUQheO5Spj6UVRNoiFzhOMtiY2TlRmA#DwXbbzsO3g6S9$@6rPnd3<E
zSB#&*t+cz;N#($6Z`L*T5|LXfkMvD08^D_~$FbrQ#W@QP0jUiFxgnvROV1sVr2iC5
zxTQ#+3xD!bv9FkhqJN0;WwxL_rY3L2bE!)6-$y9SWy`-{yte|v=liMG<|>{OY44Bl
zPJ0N!_(O0T^>0<7d+dMGQikWD+cz#+j&33)!ev8}WIF%UxROz%{?i}oc;d5#SYB90
ze{G)nERDK)59_==l0O0R4b2+8G8dAX|0MsZXJnW#YG_VD!*`zN9Qpeh=5AnyjIHY<
zO1WMT)Yh3<iRqv#<mRl$KsE_!m=NGv8`ows`Ey6<o`OdhglW7UsWmwTjkgU@Vd45d
zrZP5bPGL$~MsJKsDQ6We>(0q}rr4s3u=Uta-P>ZoH#JLpPsSEiI7{1V@;jx2IYXoB
zcywn0_L{VJ#7m@G$zGW_nA-l<$M>gye7=%s%v?ORdSb|3YI0-o!p(zWL%xp!?QytV
zm_19|*viJH=jj0UU$L6?Mmf(2Aij#_C(}2EI7L@TYa@lt<fDW0{RqmuK>wHaQn*SE
zQq@D90N`bkMqANUe#u^;tf>XHtJBLZ83ITwk1;LvpBsO|!{ucMo5^iDKD?jP-q0+U
z_uvrYc!$YO{@j_XmO8|%CJ8oNYQ}flIch9X9H120xUojXZ>;*?Z3$fCwRlENyy+Vo
ze9i+Q%LDGH3J08AEqCo$r++h(QQ-to1M?wY#a20@4=*+=vd&kiwpQ!1tj=^6^8oTx
zzmq~N@ZF+D5{5B3<|SRhRe!W_{ONH1J<tBgan!K+Dky8_EIsa1+>CSRxHhM!ThuCV
z*T&Ujbinr!JO|7B@dHTOZhK>U(tSmqMLEC~O2zqSwVDe<quu31;gq(;b9a*8ep0^F
z>urWJ>r~fExX$cxzb#*pQ@3m9Fpbl)S=3Dy<aRv$<gNRA`pq@TD{OJ@;@+*b$7ntr
zse@>EVAM{@-GAiRXN_%JgSI26eq@PKV`GWu3_WdDhS@*e?oU5HIt6S{m^|*cCSLh0
z!Q){f4m|Mh_mR#%Dwm#Q^=fhy`T=^qR5)(vjsn%<BQKKI5mA#r@|d@dyOz8GFAO|x
zf2iwW!~5ro&&!`8u`r@Hr>17qK9D=abn~k0B#X$xZQYQFVm@NDf<`5D3TWnO(CZ|G
z*DoV8!^m?&roapVzXhPSVoX^*;`e?>kF|pr;mW~_H?XXTRG4#Rz}Z+`37qp0I4{FC
zK0X2atf_w_byRNkd>Av7#I5aL42xd+C@Lp^+sdKrCr{~5OW!b%zp=HyXPTALB9!%c
zn@Hf-&Z@>}+)M9#j@PkEEQIweewDrzE5LnZjirM{B#PyB(L-}srvFK{;v|`PcFt_^
z?q4+~qp{b@;NX(!9g<Njc|&fdrxeS-8D}^(>A&|+n7Ktv1{)}3AlK1RlyOG0buE4w
z%a)5<%rKcWl#rB31>q6<8_yTHMJ4W9jI8%uAALE7;_fg|=<bl-wUFQQWPZcu{V?Ze
z!)BZHYlc@M^4ZS~6_<x`?9+SgSN+!5$alr4ef1jajj%$evwJ||0TSThR$3QvYGm25
z=o9@z_#?ziRc^9#Chz3xgx(R-Qyk#_<Nc1WN&dMife*DAF*X5HUBNJpv@Aq$eze?W
zD0!i`{EL{ZkzKeKzqT0A=P1dgsiXe6Xqa|)X7)B-S9no3nGSd|(?ihWxl%&=cV9!*
zDxZNO86r+uq~7elzv$vejX4~K#asT@F_~?gZQPQSO(Ql%c{0x+pSh6Ffwe<UL%c~s
zepWcv09^goS5BqNupVK_t8r?6mxRSJ!7GgFS;gH0%Kno3r3vk%(JbaGRwue<-_%&b
zQw=(>=koNndmB7|)x0DUY1aIV=Lu|p0Na37n#N9qie|=N3k!`P?bH|H(7CT4RVt#I
z0s^K=B|Lh^AVw}#eBT_sN^T!f>dLIkX<$_L;vxb_bLzy&@7#DcN#FHcu!*GIToC*)
z0#t3FkE1+8Nzn0pC}+%ccRs5uPJZ<C_E=fwwZ7tKI3i<2Bwq3Lg<Oma>KDUm*xcN<
zte>wd^tB`8wJY<7wm+52h%zyeLy_jS5eSteO~2urK$4!!@|dy|@l858U~_-;j4QNB
z={(B`cFr03!I!)#raH&rwzl2#+A{*>W+UIUsmZhwqMp{roH=GCLF~(>;a5LnmYA-R
z>yoQQzPcL$MY7S(HH>0HDlEV}=}7bdYfl{QSrE11dSviw%!4Ip=l97sUaRd@=l&i%
zFOmTGQ})T)A!@TowEqt9W3TE?<*=6|9$7CQjZm0gEv_V^BqFm0qRGOn1ey<K9xIMk
zhfBPKNq6SDldLT`__K8$QJiQT@DG2u{=&B<0=sPKwjwd}bs?sVILxk>GZNTUF|9X?
zRwl-BfzogzJkcUfpT2AV1Ye`*>O+0B?Ey2Py>F1&zflKA$&m_6;i!!lb2xuZ4hj$x
z?`RpJh^fa}NjR@5Xl<?+>G-aa?loTw8w{nt5u5hKDKK&4H4;ta8o7dN|1$7!%}l;B
zRf#}NHCpH@2sR-yTL6_e&D}U2zYi7aes1rI6K=lDolDLOei<A5WK0#i%AU~bAqzT@
zPntZRai2QfU=FIV`;fz|l|$)I!!K}J##d8D2db=N;mNiKWs8;*xkEf4Raw7QPAo%I
zWqW-<dICSe8xg9cUyfJ`F8CLqFU5U%Ww#%p^l{AcVShw~%-Pqcs9m(@i`-yWrr8f9
zQrq*7uHHYHi9Ly>l5~-cMx(lMo)(co*l1tbJz`!Db_1Z7WsTvnWAO7PA{O-=70U-O
ztQ_J5t4A+G;#(ZL1;|EN<;!=JV%X0P+Z7fR%C_jO_)D_1?;kN8C)dj~7o|Yi_D#y5
zTVRZShsx(i%yu3rQ?Dg6ZK=M3@<4AjEh{Tr6Pj?bzCtQ4{~%PzMJiocJdjB_Wco#)
zi|~*#10|RbYq#oMY~MN)d$)j4NoXEp&n(d8t2TCn{*VC_Uo8_Gw(HVdy4JcU)ZhKp
zhO!#^wWcv$7`^Y4A=obxmZ&&PyztW;#$~7))o;UKp0Pq<lQm3U>QIl6&wqKu20{*@
zIXg^QHclJ%);||9`0yKyrAl0X2(HWuhMXtwRt1#)`1bO*$O@%gEXiCmPr4DC^s3`W
zo0{LP#J{nuA{+B1tbYpQQAKb@HCe_lN=Bs-Hg)Q`^_)r+;Z5jqTe1;|F&gcyAUP6r
zgj)T@r$3|9Jb^ryyX#2YxW~3kdigVkeh=l`{T^;aybGw6eOVWl;0Qr5+$czz1m5L$
zYkGQc{@6<s#=i3QQ#s}BTrRnrsIEWIDttGUuXD6e<v#LO*8P!;iDL+XjTW}UHfizw
zwSe-cwsYF7WGOq^I^w~>Q<=Ka<c`=VcReo&G`p{Lews3F{eY_oTkjAJPvm6)R97kE
zd>hH9>b@PZq1F=ZdpxXN9;TiD6|1K&R|Dz}d#gas&z0R!C+4V)EIO>HWu__>gY-}_
z(*js1v0u|6N*%eaZla1#PE%2T?ZQXrbG`jtmh!yG{(WkLiasB<Kynvsr&p~aOS7`C
z=AIp&2A=F2-|(CRFS=!RB~*yvYBGYKmnD4u4H5UEgmL=)BJT%kBlBmyvRP6ZqKrDf
zimhE#R5az0wje00vXwS+p2dU#7ORRaEy~^>^K3d_ijN4$s<bNp!EYLTXCH!Q+qEAQ
zp&BCRtX0e0*J7SDHKv05N!Ui)llQ9qmmW6FvC+Y-ee|CIHGO>N@GyWvF}9>DIlm&3
z3%FV4jgC4Vr2@Fq{IFEhkqn=ay%+Zs;x5|)d~|ap+$Z6hc7U_E!LWqER7gYQl~a2e
z;&Y@S3QW!_rI535*UW_dCaND3`Af&aK2kb518H^pOTL`92aDjq376DTyjY%Pc^G%>
z0Ov4w%Co-{=^z?NnbRowRAd4#J?r9R*Li+zg1P@Y`13sOusHSr;Q{4Xr_{;n?B(&W
zG4<3UKZy<s!JWq}&0%`V{gL*H;IjG1$8f;{Yhi&N<HnNz+#1>NiYmf)3>`f`-vw~P
zo14L0Bc=3%<g=cOFS?@gp#&*&(0B16F{oIq!Ao!MySEk}?s+{TZhG>*Gi4#m)ptwb
zoMiO&cT6Z-U3xThI0xh!RCc`Ufhh;*sAFk;>d*}(3|ZL0Pg@tll*^)5D&Arbj}y%t
z(F<S_=R~X5MV$MC_X)`KIp6JIsKKdBx=32nm}vFp(1d&CwFsSR>Su??;q_ni&--g#
zff0Pk;G`twi8LIkgZA4kN#b){UL0lfcz@W0nebH68-{T1rYh&?)|gT2XhrZV1<565
zZ(PF$k^NL@h)}F$a%KFyv~o5a3&GTw@0$j1=@@1*%^?y-gN?CAxFdOzW5r!RG0by6
zy^nRzK(9snCQ(jlU(#1A!!75elde7$=FQ=hm+XJMi;9xk#1W*Uxxri(^M~y;@?`Kx
zEErqy)Cys-maWx~Dw9i_Au8C0skR3b7FNNwLbNm<nT0O>QhB+liKTy{jE<kqq#xeQ
z264aC<SAki!o31|RPs;|!!B55qK9t~X_z^4#=;}rikbQxWC2pn`bGVj#-bc-te7Qz
zmQ6ht;r?1_q5D6OdApY%Lo&1^40Q1&fp45Ne3Oy(tWE7A@aj!!3FKxz6f=o}a|;Cu
z4X!Chp;jfHm>pu~Dk^38CzL0Y&UqerwdxzDa$w%f63Ao<&(1y20tE+WpoWU{(Wq1J
zwKtznirrhmlZDK&{gNY=l)5+C>oM;rhhO{e8@{38SHR+LC;r^{HG)4l5Y)t9Va|?d
zA!0!E=FvOs{ouwEr3v}P_O%~E^dnl*)m_QniWv7><l6bI%ch~pIXr>rM!SZ^dW?`a
z69(@Ex8NUEAI)Au-f4u>N7Fm=k_N}No(%jILsBOB^7m}7e^E}A`aSY0naN;~TYSV>
znf=8b@Z`T@Wcasq`@r%-M@Zi(+v`3AX@Thd-QeTP4lX58I+ao1F~`Yf*Dsa%<!`f^
zaZU8nL<u|Dn4)_^T#RCjt;q!!i?h$OYk|dl2zSbF95^4n6uQ)cU*-Rl@nclx$?TYR
zmNd$J_es~~<H?3@UDOK?QV06}z%C*mvtAB4bzed9IiH(aKt&Y_N}Z38#<)J<M4+Pm
zU!z<H**kw`W|%%mcpT^z_MSW{GRjSdd=otLO+ND-`LqNj%X@R!3;~sP^uXdzY<>(r
z1UFU!qAeBod1(ZltU}UYb6G<gH2s5Bq?uG}kDP~pv}bIdUhj%^B2HFJB$GW3Hz_s;
z$vz%(!Lm&zKpYm4-~{iGurT7jUP=~DHZ*ORvUQ|Y;IHSdTy5NC{dv8>50Gu|WI#qJ
zZ=5qF6~i&_7M}PVan5jyzlq2*HPe+nkC1eH=Eld~3#-QBpR7;2&AuM&&AF0?iGrR~
z)g;yT!k9r1I_}XVnJm!r#|^F9VpU7-eUp3x5uQK+was`7$;rD(kRkrFPQhTGDs;>r
zUnO@oX+pif3p0tQ0fomCKj|SQ^ewGxwPk;B(Ipi1tXtbn(}bm7aTDjfe`CP1L+^h@
z&sJAaX*2!f`H?Hy;Py~Vf2ZpWItJ!gt-OjmR<k)S_4v1e%@#u;d>4~H&N>OEr6q)B
z_8@|5o!w{O@w}hNr>{vB-LSNAP|@VL^K{J32v@oLlzuyAGRr%tr+rk(t$qzDROIlU
zD({vbYg~7jN?iFTH#{|-pLAPAVfIO({<@UM$a_6ZCVv>UbDM&~!vHBMCoEu?EgWhB
zSw?WPa7a01yxJT4TS%gXImEbxG-dzLOI(1DAMQhvS?>G|&1`=et-gkwM-4=MR5=Y4
z-}3hS#PSpu(K@FvTU=2@QB|6aVM{^I?ZY~r`Rf~F`_TgZh!<tb$-iAkB{HxjgVrcA
zpcKb}<)m5reJ(pdM1RVeypX&U*T0{{++9trD4n1vV{m1u+ULURZS9W@@1~p!uc-Po
zRTpXtIc1MQx*E%#i!y_by<@><=tag^S{+;EkkHkOS+VC!lY)bRX@k{0Mbh<^g(+R`
z0L<%BStaNu`KG?mXrX~Ndur%AZS#-4;o<II`VOhfSf}f|rZv+QQsHD8ei#!f?3=ER
zmhMl>kJn9xOdLliTAx-3KR5*QU-(cwwpmia?DO{=D?dX%b3QSh1nhUOsHpab8C40-
zbfy*wX+JMKhl1d`D(8<z75|EJumoU01tI>>lW)j>ODW)CzdTgKQD!T9w0g9O=;UW7
zBWSdqicm&x2uIulOFElf@gi+Uh@fTzk;Y-C&DeghMn29tC}ak|q+)4x=npI)q?)b*
z9vG{<?0d?^t6Hs%4y~kh-p6z^b=&FV3MJ7S`5*Q-*0MaLKrNqBjEc}VcN>dmt5QHS
ztVb#px0Cl;`lF|PZlRzRUqkV5C{&e-$Q4VQ9yx}X@WUHwMy<=camP_=Gkrt^HfpSf
zST2erhx;f5p!Crhkz$tnS}lyT<kgd*g!@{ljrLrfO2pMlhu%nAG*M!rjq(!Npm<K0
zRJG=W=CF$Ls=ifXN(3cMzq!3_fn9}N-_a5HWK+K+q(0r2IJiI}<;<#e=;Fyxaa!RC
z(I~KOiHMi-LNB&+DWpsDzNH-TOe9AGp8&^@L}lQs2ZHGg>rKVEt&Us8h+{<%Eswrt
zcfF-U`<r*)Ou9{I)&A_Rau(GT^~P=gZuxo1*4agtYU0V2^uTK{8-3l|hrz>A-K~Mm
zEVoeMIs*~<4deG~ky7wH96e1*pfY>j^&fh+KNFO$?`+#j)?7DPJa&RF4m*z^=2S84
zIU4dy><5zvGQkgF_Rcq}Rx#aWR>4}v#WAvn3&U~((OHNZ_}NFSMg2{E_BA^DT#+B7
zn)qgl=UdHqcP@7DcSpK-mfJE3%3hUCT@F5-V^~CIZac592HtFz1&Ga@!~!lO8Q|Xb
zmi3|>G4mJ;@p@H@%ZKq5km;0Y=={{K<K1>Mhs>wV$IQ$EM}ImQvl^#Vyf9DGasEWy
zvfj*TwWlA{5SR>RP7bIr0CYWW@Np_eQgcgEC0Q-7CQ4WA(d^M2V(nucp!X3+B`53-
znqN|ST+Uq9T;86OZ)*1T28W{LCDe&;MRvhXfy+i+lOf-36jra)?FTN9bPJ9SkzdxL
z*Vpf>Ht~N0%oNDp%t!OLjxYGr3YwPUr&=0Kn(rSQ97dP!5tFc+WRjU#cd54dBM}jZ
zQ~m_=pHpjgZT*q66-U+5-v0^r!B?<f3gkOTO<nVvuT#-^hy8~^wn$9Lskbql^2-UL
zByw|n>KmFtoq8_+UYiNTQCANOL}lD_yGUuJdPl4A%IN~X=7~>NMj_<=<ec{$?@EO{
z;C_1U$GtXNYUG@559P<zE3M;?slC@9u9%C5t~7w{%d8>}-qhm@OA8u%rOIKI50;j@
z{?N}7nu@41TjVa{Y5u-B=f_6d(%gg?rG~ttY;({2Tm?WlhL;d%1LF0mQ-;Ex+O$n(
z*y#HRLVj*{H9qFx;c;R78jPhZ=9gg>O@tRWn@~!O(3{aXkd%&ed=-|&(LKMvnKrmc
z*~}$5GLsXVA-gYHH11@0G)#Ol6LpL1?fE5bI0x6)MVzlrO_GoG{2Qf^CER-vSj(W@
z?aU*NKN=Z2k0}6BGIj}B6JXFYaWY~@I(V+j{zZU~jgQha^K~?K7HR6dFYl(#3yL8I
zM;PkM0$iz?7)BYmEv=JQNK?ugnrJ8V6Dl~8Gz;N4hngr<hknoV+dJWvPLePHCo=R&
zlZwcB#M03AI$QzGo>y-)E6Y8>oY!x`F4XGZhv!3rZfX0{|Bh4}FSDHE{=tnbZ%%(s
z4=C>s7sM^Wtwj76#L>^XoK@{3CPm#u?$xn#*0nH&Dh}~p#`HKcK6Io44br3Ca$UMX
z&&hMbw4u{#wOp-WC2Bdk%L8V{4L!uhxV|JM9@akJXW4ulFnS6|iACY2d7r4+SiV-u
zwq-ir@esY@RG#~6A)}DfB0I3H3F#lJkdfcQ_u!s2R||0kYEy$WG`OsVxHxWk)J3Gs
z0NQ0$jH#aIGEZI~jPnKO)4BqS4sIGfH83FZk)h)QHHV^NR&3oLw5jEeWf?=EGq5<M
zDmX{SgWSHTv<zH7VX=>S%~tbcV_o@%TCPf0URYcIyk-}*SG*<?N#eIU3$*Kv{3%z5
z7g#coO_={S`z*-Bv4xR!G?j}f)~JwRbBe)|evn-Ki9&>Mo>g|s?@X#_OdFe!WiaCG
zB#<&;Ja(tV%8|my`1Q1{TDqFMTD;1`)%Buayj%?xX^)@>rla~$^1AMHd5TqkX&l3|
zHPSg?b}Qy(*W<lYSf*q32@D<9UVi#+rT?8O%9}q(q*<e>U7a_#uCpOeli^I|T)U$D
zH}@qco#VKZzhLmvU=m8l3jbaNrMor#*rbf#<)sMnm=pVOuzjnJU%I1A+iFluUB5$o
zCY%6SAw4(7v_Ya{BzUvZq+N$<*NLQjb7mSguMq-7Lj~yh!p@dcri>cbD#vtlzLus)
z^i7yYZY-aD{*#A(O#Aj!gQeamg;hvR2^@g-l6`mSCbbiz6LPnD6?2Q~e9A+4=x_|~
z^qq7)jBal}s>}4pu*ok`sO`+c0&+Rvo^D=je|-~p3?F+mBqWvrA&GjjR1xu&ylJ4l
z$W8*msVuNoJ-t83@60*as%#F`s7XUfY<$N4-c~a*E~#(u#i}1%A4G4#w{4q1gPbbq
z%FdD5`?$+DdzQpse@j$$m}EkE>QG%m0?H*@b6)=(X%4Ef2S+=7^RxCU%s^>n<Wc-J
zGtzj?3Y8K`YK{z0<VtuGHi3tGm)XH>Y&=q(4Zp-IX+CPY-Hqc10{DJjAc6)}HK=Fj
za*u03q}FxMJ$r3L7~Z&MZD*}tC^l0Pz6t%2H@;=*oMKCr<I>3%4UxywzrkDHe4cfF
zIi8vG!>xrqYx)WF3zlwa{qDN=Z-F4RV9W1MNcR-0bZ!pdG=~)=Ya_<Du-v~iW(1^8
zrwPd%hH32C$(`6+e9Of#Tf(#W^odNdp7CBy)g>DLUV4PPhb!qnVrfO~bRvF8wWmcP
zK2r}~?W4kF+Ttb&e7f{4Vdi5k=^hK&ZHS<RoKaU!y--j{(b^G2BQpgzCsxZ^b~S%?
ze-cl{pwQ|RXsi^x%W)S?D%&VMm1U(duqUKNrlL<cO#Dj)G{viX8N7^_s#{c0Svpv*
zSq0m9QZj55u2t1^VHriFqCcq3A4Mt*$wnR#9OW`b=%9J_mT+V95IK3_$Kl`UkGCwx
z)BwWxv75**51Lv3ZZL?1jFOWK{}vL(*v2!An(GlEc~`9qou4Uc65SuL%|fHf_<ku}
zT?gga^-VszDe#TuwF}`1Q$LBc3rkfpJs$fqO)^{w`*Ka9)RB5Us&+j_qR$fs_vG>b
z<Q{0b16m4F3w-*?QipXZz5IO~{t34wKqcsFlD0n2K=QdrTV0A11*K*iN}Uqw^`;o+
z*;QKr4TJV8@%q1L^CxXX()%rmRSEPG66wD!^dfz7En$Y)6BK>65x=U?$k<rG4K(x&
z=y3>0F{=mZ_~=O=aIzwFe5<jM9x;0s=q`m!g_G;hvIN8HV<nI2A%nQ9O_1u!W~zqL
zqm%!TTw8Q2L`{i>mwemA3!h`^0zXp;`KKa8$-&mhzCc%DJrks4WZKdf6fsS+Ru9Kc
zKS3{>6V^&sQQS$RNn2d8F#23<TtPSg!(0Ow-dK|=Fj0BjY@xK2_B1IdwIQr9T1Byn
zy-{UxQOT?)s5LCrjn0;&pzd2n`XFw0Ugnzy&O~|GtFe`sFn^{@_nyXgDN43^Xb>y(
z#d`qI&J3i9a|rvwDQu1;RM;}z3!C|N7?7H3fp^-levX>Qs6|~XV;GZz&DguNEsgjg
zK$;Yy9t2(=5J4fd9vkFFoeOnJ(kr2ny~|7vqskO+Kw)cWU=6uSS<Fv<%?<}S0=@bD
zF#0qxagdptTaZ|LC3RzcdQ}NiV|KcpEnb_b_iW^YIs`|lzq-p}_?>`rHk#@2`?KY*
z=!x&fA6*9J>*p0)ZM8Y+@b|BC@7hE4LNK5^VAZF0-%10h+d)<>EY(Gq#4=k>!ihVR
z{ns5R6l+qwq+6A$I)l>z@?V1sHvC<KQn><SISchQ9jcA6XBjRYM>)Op^a3$;IrbfX
zU!`IBG*~(PU!*12GA5T2cxfKLI?>;DIl4-EN(t4`gG`UG&;qO38`b&I*CKs>9UtZW
zv}{91(4uQ}$ECgLVNBjZNGd%vXWtUBZYW8!D2^9`E?gvtB92tIdxGY+Aat3|)2m!`
zDBmT?1n{e+%ChhH*adq8fz=IC?N+EZWESsln6<r~N>HO3P*}p0WdTS^z~pJPM41ei
z#jHHo&q9Rq+3ZH;VWzSyXO}LyW7gcWyU(@o6?IThRJnkFgNWAbH{(7c#C7(+B|60)
zM29@2-2SVB0zm7Q;nPWnda2>O%dUE%e*DCcWDTN2rg@uz?B{(8!_Y)&eOxgmIFG^P
zPiHOs$XHY)<<JIyzui=(0)v2miNPkyU<Upnh(}J-0IZCp?tf(9Q-bt=ZA?4$Ec;dQ
zx;UZlvx?v{X?s&k&w*>GnymSMMQ+20rcL~Z1vS*3oN|#qIO`{d>fzjyLa4j#OmvA$
zh*HOlW*5c1Gp;-FcE(~t2F+~X#QFC^1%D=)RcB4UX=;;u-zX!xHb6)RCS&hVQ&S(j
zmVV)P1vo=3CCC7aQ*2?{fCrpLeDyv37dDW1=Ic=NHWZc$iQ~N|#yb-QcICest<I>y
z^6E{Cf%a6QkW<+mSPZoL1#KVB+x(U(7BVjfJ1n@sJ?&!%q;fPZr~!|wr+rZyqAl#H
z?L9K}rM0P5#H|#sKba*MUR%>qi;L9&5w;pM8;eeHeUe&@;TD1OnTY0aIu#O9i((rw
z?+<)7Q1y@!P@*dIE8ECGLT$dQ3ERKz3b-Q}gL)dc`8GE8=Km1(1VQjA@7c($C28?j
zj#LNxrWihEi<DhL(Pc0Y2R~4hap-8adN91*)Do7V67ZQ7q6W$=8?BD254hWn6_Avr
z2-MF&QwZnnl!Ln*9Z4vr8tZQ(<(vCGm9R7MWBtY(nqTs`;8*+3RYxy=l}MHW9Z4M&
z1Ee^v0QaS<ORt^9ZswM3-?RSqt^>k|R*e`*^#mH(*tbhu7ajtuzb-$rm{H}8_cEV~
zLd@sYqCf`c_B*^Dghd0{^*^Sjs@zDFMfC?@jaY3}q2CqzY+eBG3~0|>AL>F1?YisY
zNBc|#P(orqSckY#F=Vm-y`Nat7vOM857+&=w*0M0=aL-$1kD)e2O*3OEs8z>V9KHr
zyY_a|VJP0FO$Zh@hC}8C#Pxf0p0eG{XP1+gAF$Nmhv<zUkIiQLk9o^&jn;hw&Dojx
zf$VRh)>^FLeE@!(`n-zYf^a>Z*)*?JiT8h#sw5L{+WoY5mAk@MUzDds4{UGVtnEjx
z`en;fAkT~KfY#%j<J>vLLrN=y`i==NDT|$FA{ZHnqdWi9Up++HbJkyNiIg6+fzQF*
zk>yAQPuOB?qQB5vEsF`eqCFhALz4b^r-g$@4eKp;PE~ie%7|?7Pd}?tIv~MQl|~g_
zmkuKjKA_U%LGo|y^%m~mQIn^->JUPc;ETHas9J;r{F}VtW{}i5ldb4t=WqYG0Df5%
zMXoY=JP>XbY8!maQfa?eQ30d1ZgR_;YSdBL?B|1tN|)MN7aCLUH3oCzWv<;(7T(Ru
zu=>#x--ku>)tFX{w*8{$f=&J_lYUbT=c(=ov}lI)yx)YUa%c<y@0BJ_dDp6VtFl1y
z_nPgDSCx9Xh3z6<rX4dwUT69D1qj`W7x(-ZVcgSQ0{y(<r$L%BkuD0Y6x&aYK7*G#
z3x-upPVM@#S9zY0>)WjTE%N&foK3|AShzbk7T)lt5*u%49ZD^T!4uUDQ4N5N=a04I
ztN9%IT$Zn-Vy1GtJ@SYuGYZ36uJFC!taq@6t8FFx*&py`2|24l%D5}5-aB8G#^(G0
zSGnR6Ay%d*f@zK5&uyZz2XC+VGVOYMB#C*AR2sjW*090@jlNi5jao%PPGM`RvE6;>
z%tB7#_eG6Wr9WF}@Wy7+soi{qzB`2lIdl~Tf@Q@vb9lnC6Uo!RS0I8cPLuQ#XueN~
zH!2199;IR`D<e5Y;7r2QD?#!GV3gVf`E&@8Zz2*Sm}5>%ASttBs-d8jn^zz#G%3~F
zx+Tuf3k?c$i7^YNEmY?lVH#dee2-@f#7o2Tt57EKrBs6)G^f$j6!Hr~pfnVeR&J;e
zlp85cO*{=+#QLtNimnq9Wy2x@S6?xKGqo}@dn)8-k+XGpTJB|Gc*E|1ms)@L{oZGW
zotS<6Q&iev*QYTCp_PJ;r)IkJG^q7TZrfaKc2y|~N?2dNTF}CJcs5m&HYu=F;~8Ce
zuz^W(^0%_EXLZos4BwzR{ZLuhXix%4kMCbmKOX_IVjELmywr#yE!v19`fisZc2!lM
zlF{F4=ERU<I(Fb$NDtLO%8JUIpZ6V>7W7sTITV;`)N912y-iM-RzJM8g{zGvkvQwa
zR<V)^d46f2BEnB1&DeoTO(C=^8Zc$`bM%T^jLkZT4a}@G4V+lB`swR);U!XE!eRL*
zu&WH!w5GqSiCO`^VUxg30jKQ=$t23Ftme!an4GHTF?K=B?Koq+6I#hyoT%7vJ(ohz
zkY#wOeiNxpQ`j9I7sCh>W7<o!{m>UOC~q=Gu$%ogjY25tp$SOro&kX>P2CptVS{4L
zYNAF~Aze(*Fo!@4MeUi~FBS~ap!id$MX%)!i*R1!%>S=n2lgOlRnHEJSz}91qI=_S
z0#&6zl`%5dcM6v3np|_&0R<*#-Ex~_AQ8x^uT}5oH~MItQtjT*RE=U;hMx>e#tAYE
zUA5WeziCqo5)u=}f2JyNN#0?N_jYhfx78MVn2(&>Baw~5hm9SXQN%U)?Ygek=k7|X
zYwWDg*7mB`)yJ|_n2Md;=L+%B(qwXZKIZ4{{#^YwPRsCfsStaDng(X}7LA!bytLRB
zOPnLG+HzNE9ZQBfRYysHQ94aaNoU+n4tf@%qCUI4-E{Z-1j<O`<h*t^Rz0wr8O{_4
z45Lc=o6w6Js0R^3nHWDwOPk8lz=sewdc!15%SfB7#FdwAsex5g{f#ckOkaYd_#1MC
zEpis0g?Z2Y$!b9Et@?pcAQbDf8bij<0SeaUaARzF8fr$GpNo_jehGeXa|lSQ-J|^{
zP@=50^^#Dm@P!zYJYBFvSjI=~T+abN0|x?)pyul8)e8dAfi5uibTqW-Y7AmRkJ0^(
zf|vcVyXt|gCVSw6qKQ62g;-$}-M_#;<uKym9)Ud!Ly9^IDx2b?38HoIb@hI=x$5Zw
zBSsX&NoVM+Tv2r-@|EqoOdVR{?<@Z*AoHHjnf&A65egx9=;Q|f4feLee`WFFX*OF0
zXRN_Sk;dcQuu7{{pnlWqRXp8k9%P*f<nOB!Ib<eXcRkwu$xZvi%k|mEi3}fihqIX_
zfO>gc816mK7bZ@beWG_R2Ko%I*YykY#{&LU*`n?r3)7IXxFow13JxRhA1uU^QD0D+
z^V7>!uHQ9NoM7KkTE~!6#&+mdZAtzt{Ov%-qC7GB+e(vTkmJUKlDecscXnC{+gLlK
zs?<n^JgCuB8tRr9$Hr76bprHn2K3UD%bg3agFK9yVDLMA$Ou?BBgWTuoqiSEB|M?7
zT-Q!n1ifSH<ZR$%Ajd_CE`%&~20=e{c0Nt=I8O3-&;5AjaNez`7!Mh|tf;sg*K)49
zw*0B}C=z|KHc04AC+F(H?M@fT|6hdNQ*>rgw<h2(wry1^w)0hN+qP42QZXvFZQHhO
z+qRvY>T$ZS&fnu)uD#b7dta=px#oOc++pwB`cA%R#iuMIzG%bfi&l|8!e-GOl-GaU
zvg(d8o6IslZiG(9`tMizL0)s6nzYRL#!cE$R|0^=lt^b6y#bxq;O>W+Na&Qn;u<K}
zm8f^eM1t|v&(;gx%uSD|O~<HDWwUjo)lCa)E#`J1Grrl?OYZfudODKgIunHU%O$tN
zCSjLzI@!Y|H<$4E&5nV>jf`>Iy86j<#DC;vlf^!$$vjDuMP|qC5Ua^PXcH6hcFK@#
z3$JFY0ic~X&1PlDyQ2QFr<pgoQVXxR0<Nyn{z7bhC5^b<7PB1_yL!@OS-*#^wSq3S
zawuR?ph?n<#(8NDzmQd|wg_WYGJZ<cxL&lwM!O$3PQz=Hq)^w)xe+d*O{l^)Q(G&(
zXn|h85#-OSLlZp5IAU!Y|8Ly#p^P(PtuRxI7`k3bDQ(&VV}VSS(+Wj1467BxOgi|w
z^6D_Ps-;zxTWfFnbPiV4)Z~$)pLH}66%K_{M+bGQ+%(xHN@JPUNs)<@(#m5^wW)=&
zCZ?n`&UL7n7}L<ygX=31#OYP_7ToQ5l(Maccr;-~O4GGe$14YH6?TdliY-AhVMb{2
z8h!zh(<aHoHTFgXW1N|hofAb7&NbuK*3#Z_%%*tb-3vtr6ty4DcZKXO7W{-f)EYQE
zX_27z`EeRsBgW*vaq+lFb#YQBL<j0Oc1I;`V-SYVcSo(&WgUnWaM(i+f7Dz58gr@5
z)nFQ%QD=Y1L5Eg9sF*M(rwKA4NPxa#X0WGBou$1e_Cj$@E>S&1g=#+_*1>8~oJtuA
z5@z{kP2%O$kRY0vn<<bn3Lr`A9i70M?g-e!_)ULrPKGgkDBbU9^TV-tfac0?B5<g%
zUKw6u(j;I{W2}dh90K7PA^evee*`mQ6auR`u6~p%<UkEbMiI^7fm;%5Bsr+g9(xjF
z+F-P){X*eDrcrGx?NDm@K!%ZQhx1XL`LH&*JA&iR6jQPYh)hEGs?BWJ=AiZLMA)%u
zybhBXOMP!H41U_h!~j>rNSUF6GK~__Tu*$Rp~0AdBq*5iA3P()gg)&TV|31v@K|+G
zq`4i0=DCS)wQrJdtYz6xzH~UUvj5Bq{Lc(X_Wzed`@gXQfr(Ob`;6$K&#xF+5U8}0
z@h>R+Duuiy3lQkDWrldD$)hDQc|UX>wxjpw{aJBMsTFQ_&o><ZHS<31oA9I@EVdJj
ztjl+-HF2}25=>;xaksRksjc4Xk6*2>Z5$;pJ_&TNsx&oawq(&AkJp{pYTwpPJPmK?
zmVN26HWG=vs?0s3EaSbiuS}_>FRts``V_A3bJMLCmsLQ{TV^BmdG8mUJlXd9SOwA`
zZ(~7!*Hqf9<xvv}JK+W8zV{oYPL~N)uR->+RAK9ayNIN6n3>*g;ejjA-jSY9jry2D
zanyU?zb&{t)zG@CFye;pmTOtJA04dz_TMcVwNXkS+I#bmR%_y2dfa*vCxss}ie1n2
zXiHyqbgwa7nKnpmU(5r38WHQQvlp+0tsFyNRE*hZ6{p(F-StC~afI2?io!kPKs8N#
zt0S#Gc9FIDk5@}LX^eRnMf$`zAi?%?;=UJ1Fj0|E%@Ih%7Zm{KBAnzjV@!EPP)=k2
zu^HVV-f`Y2pFckHJmN@Z1?EW7;(r#1%`FtC{Ci&eAQAYZjiznbNzEx5K_X6Hq2^yT
zf5T`5-T3xzLQG1yAwAOq>>0Ktgfo02w-O7>;DoVwT0Em7cB{c_1MzP0ZTySH#}98@
zzR>I;xxN4YGImH@pfXq04wn-n15y^E*iWg~Vn_KuF9*4>rvwCSa+WdzcJWz{^*Ny;
zV|&FN5n+yVZl!MKU&#M`67_@C8^#;A8=Nmv&p&&Jp;vA{d8hclU&k|Dclj*8Tjg~>
z^D-1&tgv3G-X-UEyT0KSQ(hi_YN$DPkAJ^}Ki_TV$d!xS;Kjif-2BHw&(8Ee1WeBV
z%M|{va{m5zGyp1v72CH*z}L@h$Cp0GS09&=lbLr+782H;Q67~4%79c4NCrkW0`3h7
z^A{SNA0bhH$6O!K8hZ#R121Dj^!I06Mh2G_mlcD=2Y^qbF#qSD{13c8D>KV~n;-uR
z#&OZ{M_*}hjq?|a+2kCnMEXyiIu3-wLYhK!ScH+T--4^n{?;b3jX9vC35D`8g_!cN
zr1~QK1B%FSf&wVWt^T@L!r+?8VY&>;xud7osg8=m+a5mG&!3-O2^1_`6sf6ZdZx0-
zqR<wzr~|}$+kT@KB43NayJI50z5DmWQ~00JlWNF)wa_9DIDEgO_JdIc!ejm*3jQ$J
zn<%~ujxXgv4%<M;-u_-_@DDw=Kh^!4_eQdy#b?pC@~c)PSqhqm4f3iv4IFDgvajzw
zdB^Lb;DZZ}yJN|epo0%&|Gjt;n=i>eT?d*+O=9~vT8D9d$FvjLh8Hy3RhlTB&p`Sz
zTIA-7T2OX=qL^aB`o5P6BsYxv?_cx3t2Ayr|MCInS{2nBWh_#sXN3*_VpzFhkk$RI
zX{4;WB{~rR6DYxnh<KB^$O!fvnOzwPmfCeIkBR;rDexk?k5KRtz~m<gO%7Uuf~rL-
z$$`EFQ)T$djn*+0Mmj;;D*_^6cg*Dpx!u>j=l<c#DFHQ?`tWmj<o)2+Yl5ecH?iwz
zcPjqTzws@c@d-0bmhx0mQ;K^g$9RItC^NJA^_At7=2h2K%O{~H(sy4LJ5^gojx?GB
z)qBwDqsETV=%Y(dye|oF?Ck03U94NF7pxaSZ?JFF?sT6_z7e&5bLvoS_?!1_SWYy~
z<<3^l<ZT3PYR=WCPiI>*1oN3QI&<%p17&qVT9c~nY@=y51h$#_I{G@sI<`EPJQi^E
zjWv%=k<Az@k~ZtsK%3udTXh8Woo4SnE~74+F2ny)>boxeHjSItfa@OX!7IUQ!Rumc
z*^NZCYU_NBJ|^FT-qZA43y95RQ?aM!`y=*Ft`ANTE)xzFj@GUn`<z#9(-IjOEDIJc
zqFj}-$t+7a$I^B`x!Sqjfnq?l);{f@$^C~ejdqQWGA@<&ok#A&DyQ7XZF_IMkKyUA
z=^L@O#qFuU-p-lN3XfEeNRRq=-*<-!AU&6@%*Inorv2|f8r5Uxk53OzPlz{&Pf>SM
zkG6M&52}CUJf{D2KY5>jj=p<*jJ+i9`Goy8k4mH9$WB9!LrzRd=47$=-s>9}lt}vF
z5l2B~@q0k`UH5v+*@NSPZu>pw%Noo=eXXX|K&(F<fsd6{((y<4!V0#zbprXY-`EYg
zcEI&6UrqKv9ECW;Ty%%jRBZSG%U$J!5BkA3zA4E)=#3A-J@q}zJq<9*D@|wm%5Z#j
z|AOiX><QE(b>2}yo0=m5Z@hBvkM3j0x9k?w2NBo@!%p^o@-5BR(6^r7x0K&^AoDlJ
zsiZT|<SY4nYoW9JUH%*9Mi2Z;OuceR*ea~C(>cSrYvS8x$%x&!enrtbz&hS~8u0D9
z^0bb)N^MPHeQZs%`epIH^wfD-b%_M5)4D+P;PB`s{uWzLUZJ@-zQ|s!UiYbA2F3^>
zeaN{ObJY=KntV+=P1sksCU8*)d?4AUWyD6mY-=C+@E!9XQy;tUbmP2SJ2yHx+a0zq
zwIPCi%eB+Dc7cD9HJ>!K7;O;l_(0zHgnE$Q_@uw@t2aXpHb!X?sbRW~gm-qRokv1_
zJwtpQyAOkY>)mxheDKlScMo~+buo5<zkJIdeJ6m^_d<V6%Vngi$Gk*L_+Z`WZQyk-
zH@YrvJ@$O_`F^(;@b{kuso{>{b^X4PVk!L7^9_~o9odX4$3bZIYt60T_Hw8h7m1zF
zd}Fz#__lE<GZKiaXS=cWqPSl)lo&aVXT~OLi$C{bZ1g?k9qEPl&iZ6KVPU<TvVpyX
zJy&9BTp&tC%qEUhpf#s97d*#xTA@tmoZKd%Q;1(0nLyPTr&x%o$TCX4T7XR@8J6%6
z$325>4!r=?h2)7*Qjx8bt(vWwt(~o$txlwNT^e2Ky4-$7@c87e<+Z}o|Gah`xW-@E
z{q($+ZRu0-nfs)A$!Ez|`kC`&cd6A_Z+UN3yd+yfpapDmQNFxc`)c&E@>p`IzLZZ_
zN>@nNP1j7<Ue8<7Ixlrz=C-<QY|^x{k+yNNF|I6J(zYr?sAgNjwR~($(^R3W5@CyE
zL$V@VSXl5>b!<GEPm{Cct!h=TuQL^y+)E2&gJ8S2GF|?q!%}YQKG~4w$JS}-zSLrA
zXqjh~XGzOels1EF6ImwOvEO}vGu3_d;dH!|`TNS#^LQ>3&UI>=%eDAzs@*kelgs`1
z5_ssU+g`sX`*umv{JwSdr+xjBz1h3%ali8V_u^JrtMl=aYqRZEoBjExmP+$hyYq+Y
zVWryd#LYFA^TF$s^ylAQ?vD492pq%P5#KJK%_1CGEwfLeN0F)=%vIf<od<76=T`So
z2Lxllm}OwKn(p{*V-fbJ%xC!9;@j~f+{5K#C<%q&76MlhPvOfyW@Gywvte<F)p-X=
zlLg50)5lkUT%O_wB5H<2-_ZGD0N4QCpdIjbNICz;`g(`}NPsQCjnI$%zfUseB=lqi
zIp0(basa2GukM^nh#4)M0DfB0C@P~^SCB0zU$=-1FixZ!bijbe-zX-G9Twl;08=ne
zxErt@7JzF2;QYroqF}B+Vm#Pq;tX#t;1zjI8{!0iLLFj6Y)0^6WFq(r!W;{fHt5U5
zTo0g*@M$%JNyXn4gt;eVO7T!F^iJI~JOhQFDhIFyRR{4ky)ZYA&xJ^<+t2L=TPp@J
zrP(%H2D4Tg(su>vfcy-YNdoIY@B%z>0%k(~Agw6@7XP_Fmrww73b*OLHjNxCum>Z9
zP1>74KM(wyG!}lMlRgoU8RyMlb_{*Shh4`Fd4M{FRWRmBFUaQ#r@0yk;EMJp;uwUu
z$mCtH9MrcZ?h>g|6~u+O;U~n3z>FwO4=+rm4hZJazJTZp`wS6iB@*DOzCMSb2mt4J
zPYmVB`{3vLfe*1lH-ibF_4kHpAg=#FJ^3Bzf&;h-g?K#>Ky+*K&SB{m0Kq(&=q>;h
zGIeW#V46>LR{^A|36y>Be5)0GH^DANAk_bb)$C9qu^I_{Idj3*#H61jf&Qw#k$(0E
zy#UViX*&5rX@Oi?{@X|9WQCzG0OgbpkVNM_CuW&Z!0~H!uCXOZ46mQ*Dgqof+g5|p
zj?^F?2t&OM`@^5}KRXShbItg5oDlmMbo(ZGLp`K^!jkZe6cainLf|x_LU0l^#1N_b
zRcli=?83^o)`|jBa*<NQFpP&KP5y2LXY{9md$pL!fQ+afjIR;ub3Typ)%j0oQScKz
zY{Ryh=m0*iX6O}t&tWbdAUJCfRA<1>6;1*={*9c_jBs*5kl>xv(ZRnFQ?SGF!=3cI
zS~FH7@Q(LBTN697&oCf%nWcjR>iR&uto6yP`IQjunr)PP!M;{~huCu%o{`r+0sKIE
zz$J-qa}JVy9D<u^z^xl&0((v&!b5a$#yBLm={x@{6yD6Hi^MF<Iv+?6;SI@*e`MXJ
zA8*C_6HkxaO!Qw&>%@P%p)}Weyw~*UI-p31Y3h_dLE2V9#DE*e>jH33tB4Q#z22+s
zAo4mH@N_G3z{Omg8AKIdDBSg6njWyqOEQcccO$6JCO$Z?8XK1xJm5jqubDwa&b40e
z8~yGJ-(-RdIwa1O-mNu1(_a<qvfJOFUNyg;wgElfa4`()y)Nq{>FD5tk(@n0q4j^`
zqi;A|h2z`9LhFMOpfl$pc{&<{_!1zS!`1EogM!xt;aa%)MZw*!PfC4fVAf_~p&!1a
z0WlD{a1Hu+`1;`xIgw!2rAEL90Yv<rar_(~&`;WvbTB6=!1|7x90-VYdORhuuPiqg
z#9y`<(C>^KC_$9TV-WtKKtppsAK4iDT@mw2Sm>N~*<{c@Q*$E><Q`qFeRO30=FCYD
z)Ep%Ck<<cTK-2iD3<$`<9Lz6ge;Et_6vc`X#o#lBxf&I*Zx#?l3~M9-z<kb3MGVTL
zQzMGSC;;Am<v{a)R>X4jGgv6*7nU#pW^#Yp)*ilg{wU!X3ak}PtR1?k08HyRxQhtr
zmX@BXR^OmK5OW#~<kQA^05J{e3OCkB*E#IN=_dn74?{S=-ZBGhh#gUq|HclVa10a2
z3J!-3%Zn1nNGqN>wP40>6bN9O*nf)O;qS&|4pQB*tUUkU+6^24aX0CT{&PKSiKYPD
z-32634(O^wK&3MnJ$27!rv?bdNF3hDIpi<p9mfqPf*mpyeHOz_*peQO-UdoUV`T_;
zjAdNFkV3#`6j<6Yh}^=z!$+)Uug|fNGjFd?v)7|q;v+l)$KD6Wom&-vY4_;QBj3;6
zxr5i)1IBg|!9fT(&;Wbey{z>8NuYUCx&<-=_2mXdljHZ=iCpX^CP3sg0O!35#Cogo
zzWxcpx&!Pameak_z<uLU56AilGUaY4#+dQ_y!ZdU0qthxXSiO5H7D-6^)B~?JUI@;
zo`!o^@+G?W_qzK+^y=Zs0-aF?<3>etz(=~b;AhzLQRD|}!$M+b?E-nrAtOFwHt)G+
zms8~jdz-_*aB~dVXT?L21n@`!2<8F5QG2R?_{v)FPodqT&-nh90#B6&-(2*~mI3cT
z>us{&KSJwGBKL<k2OJCc(4pMS0JbDBAy|p4C0rm__1EM08RC6wac^iOBw)>ZIIbYL
zHsHEsz+Ai-R!Tb`?vaHq1E;Scye)d<t{^hk;a)6yFs~qRR^T`-d|f=oF=qIGwP9Jm
z=+^9};XyBEtMG$$HG&-YVcxNB3&3^Q0NN~Vh2Wg50UeDyLU@2<$QvQenRQ6)Y7oyD
z%%k6cRZ9TRod2acKyMS!B@XvS48rrH##QsC6ZNDF{JA0LS}`9~bEa2RU5f05>5}Ph
zH9|ayMLF`9^-sV6>RLRsk1i5F-}mnUG-pP%7Z1}A6qu)x?>HsOZDZtc&H*xzQ1QS8
zaj;fV=($|^0x?KqVFOD1RC#b_GPp2d1xnmOVwx5gU<)FJgI{b21Vn_8?gIQ^05o}y
z@B%1#u5p6Dq5`g@7@Y#7WgxSHRYKreP7@_wqtK-S;$1Fc37I>nvtXtI)&hpJ(53>?
z0@|ZsrvlLe)}zqpEPxdmK~OHQ*Bp9GL}-?g6+e=nY?hK0H&P%+%N&AQxO@h;0j2Y=
zu0VC3jycC#fUc0o?D+|Ta@12k$E<)8fI#qfZu2bg1iFkr@r07wZ%gDlmzo5GC-QCv
zoK?g8U&vXe`Oh@iFMYx^QduFIS=Cy$G`cBtv)tkt&)J_%$Y|J^CqnZxEC;R)NCaY1
zvu><9)}c_L0m4{gV$=S^_`z9X;|ef{!#{g}u0T}JNZY~Ho}{dWZVfMZ@RAwB*syL}
z*usRC@ZUh+3R1g+VC0&Z(`iPA&AK;$R0^e|xREu$*oZOY+nj74P$~MF#<m(0rbc7N
z1sa3+g}~Ib5({x{0WO7KB}p{kKYqWJ3!Clm#`39!GVG|la6UW}^L5Y*bsMnVgG~J;
zTp>;v=iaqsnvm$$m{`bboq&0u;*Rd}*QJ{>+-Z*?;l4${wpq-sXH>blg72U=EKAY0
z9)30FfE<s+?9E`AF$AT<0g&k#_Fm^m^)w6atl6@Hy*ZhT>whl$@mbCAJfI$g_UHKE
zX9hls6Gz|Szu=Yy$Bsz=nNfIMF-6zIn)XaiaMLD9D}v;<4yt0ou^*xJ`C0ZZH`uPo
zC6Ual?Yl|an)MH9UhR4j)aBbQdnxRwPXl-ZuNedbQcxynk7znFoa^sXiR76CCa_r{
zbdIWScT3dm@(sO7>Gg+4e8T@tXYm3dlk<u6C+sJPZnj`(*6EERW>A=Y&-))iGAQ~Q
zFYl~s@5at&$&W~8yuf1ZIGW%NItR^B*5Q8O)Dg^3hM>&C{{%D5_QSNdO5}?<by)D%
z8n^blHL=`rnnSdqK)c0Nx8ad@gq(+YupzB;aGiDV1Ye5$Yc^|*=;_MIO?`tujvRqX
z@Fp}c&vPA9?NhbipF4#a>f>=f0~aM!vYi_PDrRD2)UC~)hUXp<&NbjRAXG8zb>F?&
zu}WHN)~zi@&hC1Jj^A{jB5P-_Pw#XJY=K_}y2f44Y9QGYzHEC<kPN>_L@mB8Ua^-y
zqOCOmvgG2LYT{cq^^3uc`TW%3NBE72bp1qZaSRk39v9n`3%WI38|ld>c39mqAdi1w
zzM8SjegWq&1)hh0%iZ7Bb@uDkx~A@b2xE`$E5+`gV29V^s4^hgRUeZq=41ss1q2-n
zp80xK1}_!vI9>dFo|3q>J*1GNNo3kHi11>BjAr)IOv=xPrk%Y3g^w2*E|nUW6!1ZO
zB1FZt_>}eK)3XvXJ<sLjnYhH*80f0?o&<_5&E#HhCtuzkH*^8C_hJhD3+mW6;+Egf
z^_|61P&?wJy3~^6Z0*uAf~He1{fw_PU;C{VrsjDV9W=hOU{}3zrAdtEX+)Ax82iZ@
zzbx4+V3v|R*X8V)8EuDRY#q>^{4I)4`yj3~^^+P9<e-%$a}2;z93Y)UikcUT_iv~%
zhXiSNHB`azyOTm8$cpp>w7!fmiCP>0kRYA;#8%XX4E8EJG#4eo4CF(!=uDJx{p|5&
zl(fo5(6#FoOtT7@j>0LkIY^6v=8Mz{C}Who8@b)?v(m}wUTnbiY#F`s`Lj8F!Jj;|
znGzgaCU(KCAQ?EaI5p`K6z>$9+1l){S<1kZTqh&N)q}TD;7<wED54*W`v(&09egEO
z<#9uM1am8ubGwQb_}+nnR9E#F*U+BC1dtWw1#+tfIZVIYdW{o~$5GP<NlCW)-237m
zj~FjB(AuZADpoMNep(aIFLBwpCq-kRaTV<5_qUiL5+F~RBAWa~|5R~w$>f^M)+v78
z?t**CkSg!e@=^17*8bE9U>c7xVN3(5^NqQ*7aixU76_KvMe$WdeKK9CS&act)-o|I
zNZ*Zwp~tnp0nsHEE8=6hPnU4KuRhuhJsv*P!T#PA%g&nU_i{yWcQEUY{(kL;nUgr{
zpBTk^UBLKY20AY6bOb52KqRC46nKgG^VO+Y|9REfl;g8+D&atj(<tr!d|0X^Do@Ci
z2svC8(r04ZIk(xFuAkA@tD~D=o4`QBoh;FXckWDStz8Y5z!T(PX3#XuL-S4USSt{)
ztQ6Z+G*Y85Uq2C`_;h}2pWCDjBR)LT>Cpz=Kw#d|KtI0;d1z4eZC)RQqthxZ7D+LA
z*r?*X7&M*s>6W3Ucs}KE)~x1@hoez@F{L0cBw*0iAaQoAA9Cyd<FLIoU-Wpz(^ykC
z!_h1fzOYXNOJgsaO{L`VoW)-Fd$%^B;DC#3Jep`w!RV%n$@NV(2v4c_!4Hp4bK<gQ
zPrVQoXu`H-IPKqUj+pddv&|Whg)KC>hFm#?kD6x_D;|eBkyC4NO2?vUUz%zfp)f36
z88)1OoQ7*C023i4PRG=ik%hXBX;g~rpJw1^mZ2FtDQ%~jCKAlO%2=B^i$fQ5k{=&Y
zwDeYeUiP{8<_H^G;yLmc6eafo=-U2uWk%FbCt?bJRDsS|?63RG;zY!@F^!%BTqx{Q
zE{U_}PQ<QXFYmX-%G>(^he1vujy{9=LpJ6Ru_SYq&<vG>{(gwb%{4mfaQO}X7w=)V
z;^Y~d1r<@ps)0sw)3ggAuwJGdrJrwO$y3;{L2J4KMq6T&d309f=IX=sD)WOr({5x?
zh408&YtwlgB>c5xnWe*Zy)e1Gn>+6SESzfrU+@xtmgTTOvWaOZAG^}Hw~bX}*K)NY
zrYuK5!P>U<0gok_!(a3th0>gF(}oPfBuc2U!2P1Dndvu{lBLO!mC_e!>WV*VZZTne
zM6cIm{YPW8NQS)QwrQd1C~BeTh@SWU5<y!BYCp=?zEfr&)=3-0O4!&S5*K?}FZt=X
z<rGC!i4$d`G6IJVhL9v$12X@9h@0wtNQSEY%ER+&$;LBk?UttiDjiT~#P8)9`i6{P
z*hp=MS~1cwfz!ZHUTYh!uuz7B@mrr!NZy#TTtOu0Ms!u#2SGn65{w+WQ&EPY=N%p2
z9|CD6_`azpHG^bx#{jLM(#<f*$-K#l5>76m$#mDd-k$I29~He%FCN&k)7QdMJgnCP
zp|}gNI((3)VGYub^00mmCk9k$iCCv%)I=T3!)B74v^vG@F;&Q>cJ(D$IBY@CH$^xA
z_m*=v%wB-W*5{DS9J46K&jx7PCfLSp+c7`t@FIZOb57T+gt?uiGO(k}f@9(>zfao|
zM8`8oN!=O-cK>;SjF(ID`-M#{k7K|DYG0Dn=?)t-YEP&D;Y~5_(r<S9)k}~8o4u%(
zqXS+fFHE7~Vq^OeupO;bIL=&!h6k)mt;^_5gNDX!A~B5T`69Cb!<t4{e!b2*sNxw3
zmZOL2yoTtQ=O5Z1veZMrPa9J=oKI43T9Gu8X)@mm&C|(cFXl;lJ1W;d(l1zl&gpdL
z7<|@XmK+-Wkt$&t-)hXOA&wcZGJ0k#gE-d9^&@!Hz_A6NS8N2uQJL!uUq){M>zUkZ
z(TgpA*8EpmglhEC1wZfxu`Xb>_iD%G2Kbj?4{;Ik5pg6FX)oY_*fEA@gx~-P7HTk{
zEZ{64E&k*OuY@c!#cKS9#8-ix;;&NfQr^<jyKPLgj4+(Zs3Xz){J_~gbBxTFZUl57
zYV<(eqCBl#N;vkz&;#qu)&Na++ab-1>a}X+`{ShLrsL}xo7Ga4R=w#a!}Im`-FT<R
zfvst`cWUMP1L=C9T*3XE*c=5XL~~y%XQ8wvF<oL-GUiU=3s(nrb;2JtB<~Wq8`;P`
z&s}L4Dan>ZP-=xx+pQsh8EemgW%tBOg9qzNKu`4e&B!gv(k}ZP$i_9<#@M$37qD5)
zBszF`EHNYbioC_Io2jImfpE}`Gpk-$l<CMEY1)zCF{qP-d{;|dSujRdre@95<up~j
zd<)4jlN%z5gLAhtJnZu@4ZfPoreqw<(Tu;0rrq(ZnR#vhfi+1}pJ2#w6y$ruwU(T$
zdVGkypD*Ke0m`VmFD>^^Wp!e5a`vFaC05ddmy6CY1$|M`=gn4%k|X!bC9nbZfC+7B
z9o{+cF(@%*!AjXfP}1!kcQ`8r^%=8k;lbTg7~8Oq%*NgwZ@(_M{NlDoSN&#@JHT4_
zjQ>}u@7xYp1oUG*tz9f6PFS=|4~`G_x>&gBK;(L1_7&gbsuJ3C<htX_jg(e%)oPc)
z;7pkW#I^G`s$)5J`*Xk~(KkAl2XFpiQsZ>;l|~AmheEySFhNxHmPUitZGKF29gA-?
zc!jlCHbF2WNr-_UH>oIDf?bM6X`d!rT_`P4S>)pBr?=&eak6XrQ-lEDa26D4cCjr&
z9v^P!7-Z;1%--$s6cOLcj%Q9x{hN;={p-A|)Vlqagk-y6jt*SPdf3{lc7=W<H441d
z!+D>Spg4>sjv{r!IyRnvi{Pc8dwV2vD|g9sbHGDfnE`~s0zk2hSc&ruZkMu?KHkIc
z>@P{^wo#|k`SYvUdju!1%Qf)&>arW-pjmq@2PyIDOP%8L%uZWUil?FcqGlLu!Mu(I
z6CJMH5$<UOwXsvU+B+bS)s{dDwN^y$%zjl~J49;7$S!45XfpEKCDYJG|Bvf~LrILk
ziJ0sGgyKJ$&6k?gOcOk*v%OCKNOY=gt9Ys~LGJr_`ngK#Yr6(7_<7}rj$WI^{2T_2
znuA!BN5>pODf{oThxcxu0N+3KTfSI(jZZxd`|YkoKIfBQ!N#8&U>^nSiDEci;!{G{
zeQxSemm?q|I8nW{E+p*$tIDKIib)OsFhPS7ICB^!;4<So7@xP~06(Ow%}sxyHfW_V
z9a!2!tW7nZ;WdRPVK`=@+bgP8I4YUC-t{>Zrj4~@I@J<pSzT(LP6XY8g&auJ;;3i-
zf;p<|8$F!)JvPiQnOLO7;T*()zQlk1u`<l1cKb(%G#s``yqayh&2m*AO$Mh)1aVI9
zO$#mg&vE{NQkV?X>Vi{A6}tblwT81ofhn?ZB)wJ#w_B9r`+fqV%`WGzi(iDzJFUl<
zwAyvbvEPHfN?16>>I+N9u=@%U19)#SY-`SMs91Wa^@A#BF#9;xuBFy%E#hNG>&^~y
z#lPcv-<gxslb~N1>ppN0Sc~%gWZU}1xe#M-a|-b|K!I?-6>YO+iZ;HUIZ{4l+UP$Q
zywFzX2F7}*^*LPjMmkE>y{`4t$t#mg#RKQ8ffX9JwH0bpAo*S-##n39S9@IlZh9$$
zDHou(l<=!aPMkiymuPfi>x`G-j7Dw&5dw~i(G1dm-FzV--WTD+E7W~6KuBp{@@so-
z(rCf^!GfYFf}u%vMiVLhUisVd2I#FwTvGoPksffkFy};~xO%#&^!?Lnb*Y^!Bw-PA
z<HAzxAWh*upi)2(IMgZbyPPDs9nr7X0$dfk4)&O=LF#Ej@{{QHC2&x=Dst2_pGu1O
zor#E(o#za2ZILg%=W7359zWjT3p?O@0J^De3UI<t-0RK>D=BEf<{?J=8NpJXX$llP
z3KYcbf$HNtGgLeyAqcHAq<Ufq#{Hnhhdi+gvG7vR$Dv_i7v_<B(Boo3r0w1^Cz>`A
zE12)F6xSgU^7b+8K;@9$vX>C~3;he*Gra1#nZHG_qU5`I66&2!*HvGr<`PP8J=9m>
zDh`krkj^8$KF<F_U<G^gHAz0%TSvdjTmpg`<#dy_k8fHOC@2WJzXnP+rtHlaw0rQH
zqO{ob&DS>;a_51r+_8&q3sbt|qVQLoB|3Ev=|9c|8y<Mu6VnmeW>6;|3fZMXBAHpZ
zE2-Pa9cL=IOW53ukbb7vKq;XycaV6p(Cwl1DJYA_v2Ud2*u*A)P@x)&E4nzN8S-zO
zl|LvxHd<=z678n%!D;l-?IuG@^(FmfV7mR+Lt%+T>5&2qIFr?8M2njhshhP<v<Lzs
z9!Aa%oc$ZYXmdJUD3uCZ@<>8hc|9!#zE_6uH00{GMS18Om0M^~afPpmuNTC+flPJX
z#c(Fut~Wir+N>8`n;kg4T@Hk9w>6Bf>O$O3GM$z*Fm5PgTl9YSh`2#~sh53Yn(u*Z
zT_=C$<8~gL`~E`y@HhA1R0Ztt;18GmiqHh_A!N%JB?SFaLikdm><#Ymav-Md&;}~l
zk6I$kh*M?xDzb6ODKvHZ-+k9+K|j6J^yvSZ!zvWbEy4a$hqQtoH`hq)38%TCCXEs%
zW7<yuGojf38DAy9`?Tcc`8D>?^J$^R>FR5O_)EHHyijBz=sNdU5I}@dM+Eg~9a|z?
zNzGI>Kz`PI0Vz|!BmDz^Zk4p9Aw?OZ4o4|N**;)4?1JU&z&sFB)ITWI;>;~0I`!Hu
z!tB$*e4OUp`}K6tM2bU#=jImZL#x$fU#-l*6Vn!R?{^-))YTFuov=3T>7u%o<#=7;
zPX24pVKs1kA=*zz79ulTSfZ<Bq<MBX1oh{E&82}?s3n{dE=E$3NU5hJy3SfQJywN%
z*s<Iz^wEqY?FMzc%w?xli<1&}fMbpGOVt+77*uv;6>s{x`*XyLTL;ASz`yAM&yPbF
zUT8lC2_mWNm$gy%T3B;-V)1%YUNpTt%}`0fKj>wsDXcbS`t!xu1|;5T|KguO_s{Jc
zm{0aIY=u%f(%h_j#N_wIj`3zwe*MCBjgu8WoJ-_EXu-Q$ong$DN@ACgBA(ci9`E^w
z4|Bcqs|D<Nq5O)ho|UZq>m?P*rBVhkP-s&%m2!S03;I<Th(5)94;nDi)-!)l;KLi~
z*p>ig3t?n^Lu5@KQ7Ks4RkO2lp{1{V=%-bd=!fx(kZ9=>jD?KRUmDuFonP!im2d?9
z$i8{SyN8Bm?1_skUmhGy1@#`wdo_>DMstChHgdW~K~iOYH$XbRiu$nW8}1nX3Z7!U
z`aAH?oJ}e2dyVtLm-pwmST>Kx-3A)kUU!cc*5gH_)D8StPlRf+hifuD*!z^!@`Y@N
zL3D@cSidbZ4~ZU%)QA?XX6X@#9=!^Lg;|)O%&N2Lc95(jb^-OmpT2UBMF>=%UBMkk
z&G1dxJHnqaNT3?9bkqGU9lz+^6dEVUo4`Hoe;T357fDK;6OGld(n)v-ON%dM9;dp@
zL2-`zK$=Mq0YvaHlauSE67BEtelt<%YlBHXq&E78%o;TPP^ML=hNojI15bxE@tx4I
z$s-H-8YcFlO6{FdtAaeSpiklZcqrJ@m}8!MoZlW+ZY)3F_PK{BQfl3%T>MB0A)Gzx
zh!6^$rgB#%{h{LzHAfdq1|JYr@<;RY@HYGJT()8aHl61P`$cRks)l8iJB&8C8RmWq
zXD!C5VAc<z0c{mRjS-BHaAj^#gOVPSEa(RzUHVu#<h=oQY1bl}Z58Y(_+5urZG#Zb
zM2!(z<O%&9UmQ*y2)|Nd7Nf6eKiAP655)u3u~><N!XBEh#sXW{oV#mh_iR;6TWcdl
z1*Jyb*QP$5`1^afgD}46_Mc|i*a$(_7(LGo=B*40e`hrGvGaz&%b|M-7C{YSpbutz
zC7?W}O8;cZsNPeO3oCs6j8o<-`?5byoG76Ja^sf`nHMl5p)FFgG%=da=7sg&N@rkK
zTL#vrs2W-&(kuJ&v4K%}rJ&7G#qjs6?-0WCX7RrezEvsfQE&1ghJ+tP2g776f3u)H
z-Gu93Y>O<Nd+Nap>!jR(ZpYHUL)W0g4fGVjLMG(dGOLcc*AtHKs<%n96#a@|K?_Hg
z+xxrJha4wX1noE~C}6W|(Cdk{kxAsA-52k6=G~XE0|S4vryGFswJ%K;H^8#d807Rc
z*AM=rD8#)o)2dG<%DQOdX;2m?Uo_A_-mu_5*HHPwY7``*s~pc-yi=400X}CC{g|o}
zs3iplC)JGU{pLX||J}}a1AGN>9ad08aJ!K#g%sPnniI5u@kXp32y_c=5N_IC$eK#D
zal@9p(Xb08^uyseTO>6=NyeO8^R+H}up5M>tVkHgd|_IKl;B7QsPF9&h3js0sh$N!
z{~8dsw0^lIqvx{wGk<r8_q_3T_U7v6C27l~_9ip5N^I7GXgrT6m@kDLZ;gt;9w1VU
zAc9GT+%sp4O{E6!Afcg0hej$zKjcWi^{Fl+#t0WZ{}&xjVXuFiicYoiwMKYbIM>`i
zng0hFngDf7SAt}_${|ztccP{f7LK_iggIJ};S9Q-6|_1w+qHpu(0vW<e5_e)sDZ&^
zOiJ}aO$LeJAdBUYuc1=}H*+LgK~TBP8hZI2JLN6BMv%VFuXNPe65?CR=>>>>4LtVo
zcB+IT$kdJxGKYA+H)$ff-SGZStVo}4Q2O%WkEtGtWsV9O4Y@6(>{_w3zi=*5@D+Pi
zUiZ;5*KxUi?bFrjZPeHt4}dg}S-37263x7MH*1&l5>o#)qvmz)GB$Q#xXkb1>&67^
zn{{&rt-V!4N*hxptBYCCu<)t5!XP+}ssvM3v8UzDBD^0CYj2bTFVTZumT%B1c#Xc`
z@@%5SDWay)WiL<P^2-uKtjCdo*vYLNdn+Y-Jg@1+ANvUp!9I73rS%Ll)O6YH25S@)
zb|aC!DjQW&HucHVxO}5F5}(7s<u04qz5A<7Rk}2;X_QoLm8aeA4qqS-QYYLCWe#?B
zlp4Gq-pz>(XB!J8);?h?F|ivnP|MO@Kr)7dnyWA=njQ3WZImt5Et@NWi4?P3lD5#j
zV-C}jr2zE?&?J@yOBjht2H(J!arI5<fsXS)M^~z#9e*<fw<a&vyF&t(k-HXM*XBf<
z&D{DCI%OjcJR<LA_c1nJzqn|MB`acj{E~3pnzmRkl39N;OKkKFH_?l;!s+z9O<7`&
z2@XC<1JAePdb<kY^!)1wjaMfVH&HmlI$x0hLWIV3dtAimQ8F%MWRfBKb9Yv)AnUm>
zOKq$boUvfp-^|unXC^PbOLhlJ#S)`jmB}&^ccN4j%$<b)?%04!0X-y`M6+O!u;8);
zQjLoFmm8FV6q>LJnjTrs!H#Ju-+Qdy%ej9};${GUXQ23HwTK9WMM7q6y$f3<B1MUX
zL0h9cHqMa6L>lK@#%;}HtEr*?SbpfdX$-qG<a&@5O_M_7Lc<)QV89Jy+&H`elw<$v
zZXdutgj$V7@jZc4LaS%O2u}H))fQP+$ep$58>bnveHt^_r>*<tA~GI}mm8+39c};4
zXHu0%TL0+1R2S!T)q+cu^Z2~<4!|Rv^EZfH<`&d=g^-J*?{7YW5UY5=DM@V<ZpaUH
zJ0WUau!VpRf17cN{3c7>e_e0e>>qS#GvAqlK4!bS(t-pQ=g3iu>YNO=Lj#WXY1)&Z
zianFd<|l-%CD_zK@szVLt(yKPv&Q`B%*qps(!}+lS=?$N6z1Xr`C+|!{ODW9aHU{f
zw}B6Ntj|)ONE_$mafBx)-+r`kzs6HatyvHKNpa7l-hfr4s8m&kVrRA_4F;T;a3PaN
z)|6WEqi<Lf-rv=IR1D#3Les}?xLI$32L&XuysIaCMA%}wv5$tp3ZsCFW8k}?aN!V?
z;;KRcTNf&#A^OBmvaGj1&<zODigDfOZ;0Wf;y5x4VsmLk2okmTjE-AANSxIq-B1@B
zjg<&2W6gDVCxkO5ryQodM2Jm;M{p)IFZl(aonTa8^HG?_9}WUT3|N^NKbuK54+fHN
zyYze#9bRwtpuYCLTekxaYa(*Gub$;?^^uG!?!KitJYOQ6t_&DCEYi|Td&Ic7IV^|K
z+z)a!WdVt;Bo~1(JOLz%SJmBjyFQ*u6ZK;on<-uw)T>GAo|Zv@t00cDGkTqr$-eZU
zeExm>Tf;<zN$lbbghNE2r+kA}1+4{v)_6Z?JVf%^<#VrdBqvDYJegYHAQz0PS${~q
ze!fw8n3_J(m;UA+_y2ybr)Az-wm(ok)p6egXMkSH*}Nc41Is(@I@;05!`GB5@uj8}
zjGg^wtjuYmYnjkC{?&)WD^<E*lSNG8X#8E(EvBQER>zkH@Avmxx8}=o1U5;pg6auf
zER1DPC1}hJhe|2g!bKz@b)42D)<mH#_C<m2Crdz)p3GrN6Q_`idG8w9il^eD?QuC=
zai<f22)OigN&Z{g`sMlBZE#4TI7rTu!*z1i#U3$^*XrnRYBpi?HHLZ{#qWpr_AMS8
z9fg@Plfq@7KL461xvD#n9fSsmwvl`o!&sgQlSpnEsgi}c%0rX5F{?|X`5)f0@WiYg
z2_^QV13;)LPxNwBjkZ~GbH`qXGI}Z<O>(t*r~BYtqMugi`cLnNPb926$#`lSi90hr
zZ|BRiWxY5EFb5Ki50N(sZlLjwraL}dK1&TD(hM&u`Fmc2m7whDY}$wtibzD4xQfU>
zgN$ykUD$7eeUa#)!OiuX4}&<3a@5_u;}`!wmDm8u^!@kcyx0B+M|5|)&x8~~9V)7w
z6+WU*WjaOIcJ5?zwszNQ6Hnndsd_L<Qi0k@U+zr_sg7H$dRckjIyMWOhO%0K!hGah
zVBJTxQ!Y%!1dv=Y!tFyIilw5Cxs>&~4~fs4)u0@`&EfTq%_@mOZ&aH^?pKKx&<@%*
zTxP|WC0IOdy6fZZtg*VwTJv9f$RPOn=|y;eHc|O<$<iGPMhXLL@3k0aph{qZCJt)8
zn#fIwM}z{AwpOFV!r!<vb_~(a&US}2q&BpI=+aP!9)1B77WQSI^8>#PrPugr_+;Dy
zhT8q*Y42)djtcEAJvmW?{b=tM?sH0H2Do4Yx+jV4jLpZl+{*+XeX>seX2u27lA>YF
zTK{1Y^Ykdb%AW?HiuRu3B|kHk;(A%2ru7PE1zpo(6{o=K|JBbjoi#&^JJp7sholMz
z6X)16Zr;{C-7LO~hp^t`Q)vdhsF5fgI>gV0a09s4Cyyf+*77>3hgyEp3aS`~Q^K0?
z8`<QVrDfV)MGAK`(fnrRZujd`g=#JC_G;HuQ2k`n#2S5T_Y4k<=QA2{Ibb?UW0^}e
z2$X_b#;DfTR{zu7w4CL)KbE!9!F*fzaYp1*k<~=r8fUEY^>II*-2I~1c9s0WG>ah-
z+Kw1H9fj++t`8F4GPz@bm#`8($DmoHmp7d^#}oqcg<ICJPE2&95RH6upwm}MKUS8a
zLm}p`ad(PrQhT=)VEtp|#7N1MPDe>2we9y4UUhsJKa!$C-PH-6pCIqA8jp#W^aR^r
zgDzRdY}mk@A9@AqehrIfHYcod)apgakoVdQ3rch+z6Q;S_eeWJ0eEQ9u+*HJ2SM-;
zjv3e4QTK)EE4mt^QMVLWSS$piXin?k1j&#r?kN=Db#+U;G{H4(;sY2FU7F32Ty$AX
zmcn_YxwO?nWA@Q!XE`fl@l@`Eqlbxhx~s{FNau%=W;dVBdhy=)kd_!z8&tNG-1nzX
zX0OZM@+TT%*=F&V>l%7^vlRo<rXi`yrXd6q{v*g!6GF%m>MX)7lK(vDHOz|662@s)
z!OTYf+qQ-xyVIxbKJu4oUf86PU@UWdE?<f~YysK~kqKhDR#ihR=2$|-qn?2<!yhZi
za9E>aFqpMC`-f>THmOZJh4p`>YoX?Ku*Z-vikRb9M?Z{neKmS8ZXg6FwyBw&RbG~w
z71d}r6NW0w_WD1=Z!j#>bKp)}l*d3e#sRKuzoF8G1&*Szodcg}m3jqcqTFN{FnGug
zx=Vzq96+U~vl<I=u_=~oNA`Aw_`Y`=T?-g)5IPQ1kDK&(CV#2RSs6QB<2p?S>OTIV
z8<NMMA+};=NNEs3ETVpMD$wN*lCS019-aX<$VC*oDw}cqi@CJxf?RzSWy`YUS;)Rr
z^JkG<xLAKaK(I{<2K(csqqKrfA1H^SIDoDa#w`+$7Hc$Bdmwxyd?<W8aG>sJEtNuU
zZO0;{ou`?nn@1Hz$&RHJMi!nQmLF~uRyoMdK}ZEG3K9QE@&72L`uX~E^)hWE-)^WY
zg8p;uYW!@L$Ld=5_$5OZEv0>vTQAlY{3yPVZ(7%hq9%p*)A^^(Dw9#2W81GewVc82
z$HxbsrRDW+1wEPX@UA+er_nP>tGyklwow<^s^c3GQRRZd+K@5qwNR$CH7(=(Z!@0b
z+RUxt=z<stk@ID&N}X=+OQi1Q?-wC29>@Lh=_;?s#m4Eb4f;=^H$>wzr47*e3$D;9
zEph}?sNRkRqM~<0!q<Lk9V*`bj3$4E3u=$SVN!^hO)>h}d+2HXh@0dUgRfen1@a~D
z@7V#x535{M4)iL7jN`>Yuwe1y0wNM=PCoE$q#(qwNj-7|{mTnaE*C#Go-my<L8<&{
z;jE^ySShs-gSGF|LL$BSeO;XI#T>ki7PZ~G!*W$RJzeoU%>-G_@BU{Zfft7hY>6Bz
zKsCNPj~(R{xEz7sIy{d0M~-e3U$O5mS}<-LdA*&e9D^7?s#%p%SvboqG18SL%T{WH
z@D}&jFpse5d@=V54$c@bAUqW%MDe*2Jnuoj+=*9^cL~u|@wFLa)PSV;eQ`(1s)-Gj
z%eEd!1F*GXzC%@7K=Iy37v6I$>ecegP_4r7yUMTQriPFq?fib!8OO1P)~IRKhM<x?
zTsxI1m+b`|uWQI@Owm|B{Fh45oT+Ld>pVx=vNP6;o!YT!So?>Rg($;j@vPd-$d%Dy
zYLVgmHZ0S}eqf66oYRwU>q4{jPJOs2{$;JD%_bkDs`Q_;ft)~^tME+Mj+aBR{-+uZ
zvADQ5q>I)5%J8+Xi!9g8tFnlQTcP9v)J(7ay-cV~=BjbuC(>uw6T`CUynW#jjTn$@
zOqzu}Q4xO5N6}3^vgkIC)x|sep&g5^?iD_;)DTyRe9_5H&wB?h1fCYPg84zeI<q00
z4M)ESs-eq(TT52p+3z>KvZ3Ne<!29Hl$b}o)o^H4+cZgzyC?yr5^UEZhqH-eH6mcr
z+gruw`wg!ppyVyyKS<|&=x9^<B^8SEsNRS;J8dT%iU?GqMq@OvH~ThZIELw-6^nOr
zHq&*YcJp7x<d#+P@i)WiwLx{p%DFj;jInUiVMuUHkrP4tz3kpp%echpX?AZllHn6a
z0^6dsKAooZjq{Q(oJ#pZgG04Kf1Ps8&n@Y2cGE(0Bh@hWqv0_2;%}|lqb(nXHOj@b
zqP@xdHSbu_LCcGgsLCH<p$25F*eVU&l19rzxt#cNdo6Cfx_05AYqWD=<9qdgotU)N
zH7=^o@=kLJXphQ*%!4c`;zlQgU6?sU*aTRjX%8?%rc1Jx8lY9gP&LuR4lXR)x&40F
zT}pV+Yi-K#DDwnLAN?%FiXD^&X;0A^;mOynXaa1eS^i2{a-*I~=<GO`4t%(oSr0}6
z0XdG3@}o9Y5yDBuUMah2#F}y7no$hUeRd8A5#AvoG=NyMqy3q}gqDNMqgBaXzL3I>
z%O<a80=ed*3`c7!vcm=QB<>nptN>Rj1peIlosIPeJT;F!#OrUUTOA`37IcTN0i9)h
z8^XFGPXs{)CWsh@{dg-d1R5xr=0{U4kvLat)uNbYk1}k;0Q`C5m4n*Na2R7dKq@ns
zj-gHA$Dje*f~$~P?HHngB`6{$q^ybShch7>0+`Ho7sd<V6~;ofTsjrxKI<2ZkyZJT
z!E_=UjPsC7(y(WR9=Yk9*>ApAnQ<D=Mtja+a^1a0og8(@SlOtswQKfx{~*j7G9+$e
zES^D6-_V%bp4)D39+I{c`^OKOy8=8T6<sOhF}^LzmAsO$ma6+Zq)Ks{r3LHC9a1jM
zL-8$NCYx7FOr_rIBR$0ew(7x^_Thpvq}tJo34Rf3R@yv6!%tR-_9t`5;3mqphK!BX
znol`SV-Xu!b}7cZ=EcMVaxi1-9=K^3XR-!m84(THrlBTy7Ww$cb)1Rc<VB0uS9Nm&
z3*jGv<--06o{#{eSz$d<;}VH}4`l`Sc$I#x4EQ<a5_t&T;<p}Ul&2`Maz)2ll-QB7
zqdsv?2jM2AV23O|gknWGINrXTXPQ3@ZoWnRuMeWIVwdK?4W|%P_sTWMdf;o0f1<nI
zQPy?uxl;IAvmWs95vq$`L=G{V*JvePMCF`A=ey8Y_I6&&yR)m-ZKRD)R~HMX)ApTc
zFyPSfPIOeV%+2a1;{zC-8UL&YpB2A7t@X`sc=Kn9&pvx^yul-ntRdI1VF(>(%mEQn
zL{&H+_PO;0NstEq+RczOUpn`pK`CL>##kX?b3rzmwIad`w~`$18~7!;Rx(38h%d~|
z0>hUl__vX&FR3)fx7t+`=(ZAZ+7DHIx$xE7Rchtc)u{B$W3HFj{uF(Ao=3La`7yk<
zu@|}Qvw7!qn-Qm1^;_{~bkNxvBwu6}c^R)Z2tkCqi59g%0xiI43%j#}%WE;)D=`Z#
zF}olar)ylw6@svwq?f27mhMDj!*!(sTIRZV?!D@A>v+%)=MPUq5f@o|g<%8KnrTtF
zMscCdl;4Faii>RKYSh~=3*>j$lWMTXJ2s^71{RcG`E&oQcC5gv&`?Sdvz|{bGB)bq
zOk$&WET=DxNckwrP1$^=ZtI+r(y#8f0X`ZuWCqIWAeKZpQ(9g3iCtw4qdo4{wdaMO
zbVS^S_Ia!!%aCoAG*sI{OK;<AO{&4x%wi9h>(z4A-hNid@riMl(pFQ5OgL}Z0_gEL
zN$4!LV;NjotpN?HWFGRBX>2NG#P!3bCaAAwu6Dq>q}7Sfum)taxytvXROGr8GU;8=
zYgk7Z<PCXAQ$c>PABbBy2UEUCanRwbhfG?uu{wsqKYslD2|rp`-J0&v+dZk}{&2Ax
z?))-=f?^uXvhmlI_w}F`Os=mR3@Y2x;pE|CnlC*ooA0xJ`o0*9u+?{1^a@fjzXB2W
z2Y{XW$A2;Qj=`CP&EoeR`;Kj6gN<!%ys@2)ZQIEP8{4*RZj6oXJGS-a`JYd3ol|wr
ze3-s4)mPQjRL@Lb-Tlk46Qx%&WGB-&K-g(EAFO*rIn-oR6F{t{fOtHyop-XcxU<fW
zi%(-zk1Ww?cQCsj>OG21kB-gRL!yvXw=C5KY@Oe%oS+hdj6)2+2f>vcTWdgcp@6Qp
zEmcwKEWe5z72u?DPjpcVfLpVn6SZ;4_oUW^RqT?kR(0d?^%YhImXEcK0CjFZ#{;Z9
zD<izwExi}CZL)3GJk+A#Zi|oH&X^t7Q%NIO*IctRnho_Cd577mwphQ(?m*2dzVqIL
z*cqrhTw;cF48Bj{Z@EzXEMlcKisZ`B!58^tqN}tUl*?T1m<JuZPi3~Iy_Ya}6gl$}
zvDMoMX8><moN^)>f?0Ep_8X_tIR9pn*buJp7_VMQJ-=jUSud8zqDo0ILx4P{RZ=?;
z4~t!R@0R{(&vSPe{QXUv@DHKb49oeVmFBd`__DLn=#_wF_pk7?<zWijVOg5%h*;FY
z_9BDbWx9a%T`TDrw~VK4U=n3OybQD9K7aI<RcnBP(DqVoNoOV@)mqew0h0vnbZ+f0
z$ZJ_d2|c|_^=#Ct;t_6(E`}7Bi(71#0t)o0fF|niPHFrkFgzF`1Oh?F`6o&s@kYv3
zmN;S)g3{kWTiY?utUzKfG^8sQ!DP2TQiO+8%|8UX^RIp!+;p?vygp>GC%m$=mkrGm
z>T`J(+7De9vA*1AaC9kbrA5BEKyv%32|x%eVuywbeaAJVq8dkOKONO{E)9YgizOlK
z4DDxG_yMt7!fG1<3F)$HRO>2BVF2ZhTc5svVfp>Y>&>0orK<ku3!Ev*3+0Vqhm={B
zH6v&zbW)iKgtA#Wzhust<>!Czs6&`vyN~?s!$D>G{IBxa2{WS;iqK%KK1gWz%Xwny
zd4~dS`Gc2Bq9RlKm#~6IAjnC~F-}Gs?@vgXeKCB7s>aV92KU59D<Dyb<aE624yse3
zdeO#mkmR<a(+<B+yz6vHP5>=Fev%KoZ@vJ~MdZgcj9(j|U<u#PBtv*AC9vMxRW4pl
zpF*WbTs?JuH<E8E|FFmcOHL;>z6md(z3}81Y@p(fEhC!<7m$3Cbc^kvC(hR=mzv9D
z^_v^inU1Je9AQ`QTy^z&%lmYpdo1WZxGyCN{ixMdrrpxd&)a;4&+#^SJrMRt#B2Rz
zU@>_jXX0!EvsH>&W8Re9HA}kM$G=KRQJu5trz0U#(06YdEzjvanEfWl;wv|e0bwdO
z%<0|@yUFVK@p*)+NK`~*_-6&<hqG--KKVMTQa{zwhWw7V=NTB%^>;o|zEoR6_?LDb
zK;2Y9e<pf=n`4=Dxse(h#`f*w_)1`CzhgVO+;OY_q1Mf&)dR0s(pW^IM+<4}Z5f+L
zL@Sz;GuE>^u52(LU&%E<7)4gBB3cf;%mi0?2hkj-AW{7B`HH&?$!Y5o>6zBPzBALA
zq}%b}%o+J5!*CtdL{cCpZ8_g1RFi705;RjH$I8Y!J3q_XY5^RI*<2jR*_HOu8YkYp
z<S2;gL%u^f-C{3U*pic){t-)#+gKipxFhzX1bJs><Zhku*-%J1_BE>5c%-Js@F-f`
z??)Me8o^5Y{W&JWLrrfCIv4tMoL-^;AL)ihx1S%=`g;am<saXgX9=EFE5Rb4lwmDb
zR&9R~&~$;pm6t>8TW-9q%{)j13#%21Esyv+NZ2Smae24O;oS2P^uB*%o3IwM<x`e6
zMtC`zC1Clplhi^iYS&3&+N;`~T5(~~kk5EwwuV}vYnUx%b0E8Sszzx~xf$6qJ43LM
zn@YQR&z?YEqo1ZJ7j2Yw_(r}~)xgZ6PHMUinyl7PBF5}1U%YeL0EwZi4to6)r#m?6
z8lNk$R_WiZ-Ca)zp%Q%dOToVxO_nRf>Ly7U?V@&vs={wdysWWiRqWzMOfiu(M-F3N
zq<M2Kk=|Se^662hz2bD$;{@wsAMHb>{}@Ynb0yMQk!y^2_#emedWB3)MAO)CQUuHQ
zt{I;vIA6y1aO5J*MpQ0En2|qee<k%4gA<8i$zk$Ym}@jV2<Bv^s?F7nCqNKvG<-df
zq3+}RIq=zeNnPR(f-+uRDwC_fHQKcCUyCXK34NCeucP1|pf5(5u%`3q1gWFJ@KDvn
zziJ8iSwU&%E0PwJC1|lpw6xGtPByCK$g2i8i(?I9;kIZ>+c+wgYU!(ABk)9chvSy1
zTSwsUQUo|j(1bz=szsP2!*OpcV!4c?{*c%PzdTiIY|c=4DJ!E-d)GCzu-I`++Nxe{
zcfnlKN@uD{Qj>vdMVTIGxTYI}sr5aBC)z~|4)pp<^7{gjRVJJI4EH$mZXNeOP_n8r
z-P-t!D|`He^BU9nDTd)1lz`PF13m=TUzrn2x^sYL_I(5g5-HR^`|4j3!D!oUgt9UU
zm3hC4@$Xg)VUJ`hS4MI4K)j`rJ?gxiZd?SK0ocqJ3_hRt)X}uHp&$rigL<6Dvkn8-
z9#VcUhX`8*2_q-(+t;+ZRDRy=W#;1-DOHd4r9W@hYCXeD`^&3{EaGl=pJ<FTtaZ?C
zf*dY3mZs+U+ZQ`&>>_y1g{g$fxS%D5uz_*W#&KZ$`G=0_)|O?nlt5G8s4IJgN0m|@
z)b#m$*+4;L@D-T~WuoPjMn?8!enwDLU_P{AONR8?-htLl|D)64jFQc6-SPO<d9foy
z2W-!cW33YJcl&Jaum<%`0<JlgdQ}r&x@|1*xx_<d?F9)R>P7S95V+t&8{I}tCTdlI
z-#;;8oH1uh27=I%B!+$fn=sk0hkr@n_%rE=W6xz#u{77paF!##Hq`j(+}s_JJT-iZ
z<Cxm@1S|<X>3i2UhFGMiIC)(f(MPqGH@2dI<oq(e|0aY&ySRfRm8Gc(QPu4N)pWy}
zsv<&N!ddXhDBm<<C?tp{j0ZL9{IE1CLCSJ4=>;599*Q3iMbWN}M0ZF>V`2tq_q})i
z>J1X`cVZh=s*t5@0+R#eEAf|H3RNH-7xI+6=NeCAV-Pbix_>O^K5$D$k$H<g#1#cV
zeXjouFT!<)U|hdasAs8lpGMJ#bV`cZfC6bk8V6+>73gheC{fyF3eacaxU+^^n|>2*
z&&G?HP}AcdC{HfB=lj^dExw{`8}TY;sPF{h@A(Aj#vjC2_jS9I`^FHGzoT-sftHoK
z+U5A0!<~Ec?;4zMJ09?1)0q7Q82JsLN%2?xU@#C1rrGx)d_A_w<wiq0+1feSG(CIe
zn$`VnBq|IX2XWko<l;`oTBy6O`bW#h1`<-3i0?PIqro83W2*?}>is@V4ahjP;G269
z+q-V_uQV*eeu1&e_nhmZJg2vv-dRz@cZj`g8E1!svOibq7P2t1ptMSb7LV<kInwYA
z;;ro*YvIE80QDx#j&uH@svdlMB?MVqt3UJ*5IJMA-JPW_iF~N#k}hogo)6A7cnv*S
znVd+(2-$rHPfzuUnFZ(^YnY8*UFWP{A7B-~jg-!!6sXy6&FSYRoqFHF2(Cx>{Vo{S
z-_R9s+e%$Ja@<lfz7!O;mD0Vg^H+_r05Ki3P>n}}<Hi0YVR|xFbveUSBFV8M#S9;!
z;eJZwWRRY}WhgXYSU-oUZ`bf0*3O*<e?cURaxL@d*W#Wn?^naqc}ZgB|GTnsS0daG
zl8``od488|-(r$e$%eE8@BJmn8yb-`4MR$h47^14rtle$>2oS#ze!(X|A<<dKn!F*
zqO}UOx^pw*WKrmCf{U$o3AUOZ%gML}X<<KIxIV8AN-ptmevDznAMLCv$5t2hH?cgg
z7E-kvT<uTrTHB2}#G{)9CgZS7ral$cge9Fb;#2tqZ48X1rhVMc^vEACqiz^RkzPaj
zMGFger7Z_{rOk|+0AH6_n*|1Hwc7<MV?TLK4*~VoLc~v-08hhS0G=qDKw%Hl7$yXv
zd+JKWM%wb<22OB4owQ&xs571G0ZnYphWsU&ord>T>CtGE<X8jn5*-IIJ8Q~Ep_kIP
z9gNwer@8NM3S}!}m+F;IzLe*`L9-=oGA7oA_Pd~?mPGZv4|l=~ecjE5$3L^^<A!I=
z+Yn<<ETgvxK8jzv)#xQV8u2OUaLJ_LVQd6)7ZnHkopH*o$|d_~kxS@QcEcTMo$hBJ
z@P6&In!DIb38&LB2=1*3<3yPT_51pqQIg0jKGBYhe7L_}jEorQ(4FQ_-xyU~8pQ8@
zV=zQE;XLzYFxPlQdjF(<ibL}8jQwzrs<*m%g6N4QY=g6sE+A61OH@fwF+s9ha(T*S
zB&60i=Lh$M;N@pNXxsliaCGT+-xeVA;_SZj*X*(Lt-mzM<YBB6bH11{jn{1G=|K4=
zWlHq76Skjf^P}Zvy;-sE!)gZcbaIZWY)Iz=;_{;Gzj<u4dmbfawA5%EEwreAiMWa;
zX)!hzG~-vvltYd3auLFlE>#C2ddYwM!PG4D&E_R5f~H5(K9?Ldm;GAJ<UcQq8naJe
zwWLo`J6Y|$EpwXqU%I&rCT=w#+<2-?rw9_GtQlprw*t;4U|}jOk{Em(OKnb_*|Dvr
zF6?uCOpA5g6P}Z!<m;fUM{<#JnFHS*QzV%WMwma_KL7b1dT=i$o<dp!e>CB---^>H
znC$YN+2qOYLt(Kiqt-|qT4~k3$@(b=2_#$W?xVXxIm><`K@SgMVzw>`tT)uFDUH*P
ze-h|>fG70~^ghwi*H1zQ?q4^Y65$J0LJ{VuiDkA$ulaRZ?6mHVtV$ne`OL-l9o)Gw
z{9tJthz9eP8zEU5xg>d*wbLG@1_Vv7uo#2Tn_KH*WDPIBqy9@~-YytefiW5Tc(0#A
ziwgcxRIAeLWFQ+}Br8&0bRQM;xPR$zM?^_e7bg>-KshLWX17e{_xaiQNbm2Us=CQ(
zrk5e#oj_&r;wLL5XZXhe{*HV`_Ky?`TiRfnlpxLa_=F6Ljp^oDMp|K&8eLCFU@o}g
z%tCcVxGlEs;)0NWRX$_&S-X)giUanFF63Ex#m^1!6#mCf&aGnFWV%3)5|7Jnuj$WB
zuM3}=gufM%kUm(FZ!m`M-@<HSNe*IyMzUK>V4rD3Ieu+-7k>Cx{cRT2yPLCtk$e9E
zlXi!G+B4|`P;I>Ad&r~=0?9yiJs*$_{yG6fI-*ykP@RM8Uo}XbE^FF{y@@Z<Y|2Vo
zVH-vahB=nFv$##^-*c469z7prbxgHcet4W_o-Q($+t^D_L6@ADYRdyC{`?{BIDbpI
zS=;CLvhH};RxRJ{Ni5%I8hfM0w_cjipGMB?nC^}cv}%!Qu6XWpw^*@1??n3Jx~YmS
zl?3SN5?Os3>v}_1mrBXlqI4MT?)wn%+*_~|K&xi;>+=>!0Qm$w+p3hisn(R!fAhMY
zy&Xwm(3!eCR9MhyEVC9jo)1B44!GwnFM99U+^3<{A9A}wNc1@9sBy_UE@$2H(YWlk
zpC)SRZkTn&^%0*vNW$(MRZFk===l`=WMWxLdFUbgi<}&OqW<SoH}=-pyY0zu;%iWg
zT+sdPQDU`cM52gT%4*q|qXPq_!*bo&x>i+6KW=mKVi3<c3L`#yVBs9k6@#INNSo96
z71|ZEL&Q_Q9f}?E3S6xWdawtKx%)5V$NkwFedy81GS-=#u^k?*U9Y8g)DxsYg~Xe}
zkF8$=bCi`Zn3};lJ;Ovsm*Vc>z88j%b>wFNM-SL%zR~Iw*IMG-^{_#nNlypLci?MX
zRr)*5fgr^Oqb`k2LOTW11xfR-^9HTCm8W^%)$SBO$Ix!8!w;Rl^=m4A0l}ld{QA~I
zx#BD1JC|i@9@m2=CPutdBjR6?PrAB16ILZp>%>0z#{eJRv-@3sa6)y`4_V1GcF@yt
zTW31&VN8qpkdCKT>$<?v-!|h5cLYL+#R-9@cB>%wZmX1b`R{buuOB$4yY-wZ-mk>c
zGZiDtF!F&nz*En1uTz_nA9d##G=BOkSL-^$?pYqL!OYt4*Y9Zp%mVZ8$8o%F6NDvg
zM3r6xVt!`|o6kn78k3&OT(7VhMhZVpi)h!NnveRRC+*?Zp~805AJAUmUp-&P9Sb@Y
zd@VO4J&bROvfZ3NybiZGKmY#mpu9lyi%`6o2pkL?A+!zK;R3nm|5L-G+fetcy*sXt
z8V?=sR@GA`qU$MpJC4kbv{U8NUeV#!dKOu5U+__7;yTT&nu>E;_b9l%IAN_ADeqBv
zUkm9%sL^&WJe)|kVQ*PfThdc`{k@uUy11mP`8b-NF8+$`V>~;4e||sB*lm5iLwPT=
z*<<1er|&sC_qRx2)6-h>w*K$0u#~mrrC)gqfn5CqB?xNErv8!KsX?5yI8Hwyy*5qk
z4<fxx1IW2GR@5Mus@`74N)pu&swrEQ+wgb79rvS8SQ&+%CUC>#>{oQMSo>XSz<`r|
z1QUJQ7}<ObR})_?g3gGqfYe&cCa9wxTXSuf>++fMad|*NMYN#w@@Hs2_J@JJ`NFhW
z7%L`1HJ`Zqvs*|gPhfTEKVqbHE|?nh3!zpk=+@~#NqYCWK=Kk8wY4Jypgix>*7xN{
zYUHH|g}{`Hrlh)t6?rOOh_1*DMgeX{yQEB6Q`cdQES328uT7h?pAV0h6@{)9HpZ^U
zN)<@UeoYI_#;+A>+NQte=GDB2Dr%Fo&(RJWn>5drn%7+u78v#IWzG#Nbo-`spJ~vy
zco{xquY)MQ{5;pXj%sVe(uKJT&}YQ;tFG2trcJ-pnE0)mj_Q8RF;%+4IjtoBN6Rc$
z?dk!7`WjH{9G;`yG129F7Ui<7X)O)DVFCRR#^PwdHIkn83OB4(9eXeiyp?1ohRkkk
zA6s!fj+1ds>=0@Xt~L=zXl|`gvihG3woc0*tT9KhMm!qS&8{S24D`Qo3<p)?$6*cW
zma*P2_t8e2z#T&4>@cn|3YQeCrHz5ZJa>b(<UB0tN_YcSiHlBCniYuWYO4ihCSx=I
zQnjy#1|}oTtYzU~D+|$P{-iTki>DK2tpZ(SwBTaPuTdGwN<)jq_N2-9*|3nlq4F0?
z>;csZE#mGlcpmbI9Imm?890R5vsGU1S{Vh>bIH#JL#r?OD1jwH^&xzjVm+Y!Cbhd;
zt!5w$8#(U6p0w5Ns!K>Ft9Mi@d@*Yov#hzPtg_l0Yb4^S!<`L0{}jyyZS<<dBH71o
z>yR;l2OD`+j5-8)8LD8A(Pp$a`*lZHL79q?vCLOth?tU$1kY7itATdreD;y1#3wO3
z`5@UC6UY0fbqr&&OSUn6B-S5h=;7a3?(g<<MomVd|4f>fE{`oeKeZhi$kjju-*Fm!
z&{5ZMTY4zwIV3rfNa$@u`0z-4n6$%=Lx<}}w;tTA?(<6<r~g`lm5NYSlL(*w{@YU1
z!Zdi&+9^Sg#{!4?C79h!7LLbJP7^goNiDTGm_fnJz4%8|zenCzO?egyLx*X5l;z?J
zQz{F#`w2wXgI^I*WYp&5&T@bB3=0(*UMC0;jn#(3Wg;b5K5&N@Q<Fx1YZ5x4iE*R8
zl-M(-dN5KPdXssgxi2b8r{7IjppzmfC0dDcrRUE^N2ANwYdg_Krx!&;9Vftz1t^xo
z(T=k+pNzNA3~bSiN-M9iJ<nue6VApc^0C%UO>R*rbto8w@miT7EKr7UW!(1d6E<~f
z;k>U@O}S6~<A|RGsVFWU(5I954eBKXPgvs=r{lHI|GS}(0MVPY9@FvFK`{3ZP7nT_
z-eaziHYLNLT0x!Tvb1QAH#Hm0SBfu?Pa?$*dech0?Y&MCqUWKt;-EmVXLX`vHbxpP
zQ64p2i2V1n+1AW)@-9^K?}FxT**spMMqTWz>MLPBgS~4)_kao%<mDl_!lcAE?|R+}
zzi3na#6!p8e7Fr|wFOo$S-B7eau4R|>Gp{i%EHtSX>#l&5DS+0kFW1E_iRoH6a@`d
z%X?f8c9o(tW+zVaf@Q(K39E_vdt^|!#Y)kpa|W(*@CTT$zo-Z|)YeDB8F^$>IV<4{
z<9+MEyVJ5wI9OS4i`_A2D?Ljv#S8)~BT>4nLS1H)nmSTNw$0RZ)%BXwoJv~EvvBi0
zYL1T6n-4{)(Du0Svigki*2H~dv{`uq462RjRMi+{;)(r~k5<ekXjDBzs-<Kkf3utn
zJ{xIqDkH_tM52_m9=(pUpZX4|q6{c!HAt0br?K;Lp0cn?7u_>7z%bKCip0Q;fwJ6*
z>c~v1qa1E*u0Qr373<`!7vk6@l@{B%GVfK8Z~tWa9**#mmJnZxSbw{qhG07!5P*7g
z0wY)nbjw;^lXAG83CD4pwO)_3>8r~@LSssh1WX4$x}KqSVzy(uWuR`wc0yLA*pTDu
z?PF0)0In|vHuX9^8Tf=}1Nej7nN~L%@!1ca<LShlBxm4sW#vSda}u;llVT}M*-i&K
z05_L)68xiUl5fEuwl|69uMM8An(3aSl7S7|D*<@-b6Ft5!BfC4)ZZMU`jvl+3LMTm
ztrk%!Y~RTZE&l~9fh970Z&$2Au^^=gp_k><q&YjL@FaC%y-!cS*H#41%nOCeG#tn%
zBkz1NX#N$h@4*%(eHC3s&tG<r^9!}3a%+ze)#r$X{wsj|oR>q<IG=ize`RKe#Ttj5
z1iVQJ&&bHW;xmzQ<KP)eHfhC_Fe^SA{Q;tS?c6WZ>j&l&cZFY$Kcqf`$hCRtx6=70
zJ2{>5Ij1`aO$x%#cooZ+#%PhBt7qVyLSf&X)KXEeG>&5Uat`t@Tk_Z9f4d0&TDjI+
zk#?Gu(IAR+`F<M0nBL=)d>G_`n;fy1o0RjeISK!sktT`TxunD^tB1Q1f_hfyebU5+
zXxioA*ll`#v)hhUe3YfqY>QpJ{@gLLUd`tBc|29o^vny^0PluwTM5gwWVLAKG&#n9
zyMMW3s@2A#@b@Br!_M^4aMAX5g~sm@KhLMV4O`aKjq>m%&&-Vqovr+UzX_+XO7E^S
zg`%zamqk>5YD!y=ibV@aA*I>Xq4hVFYjmT)MK9$s$R_2*DY500W_$6^?EDg=0#MnB
z3G?W(Bg-c~#%1umEG8v#)VoQ;(9)<t&Ve@>pF#I~B&IljicjF$@ZLin?JDjOTdW2u
z=T;$H8gRHB^efXnhjuf<bSs4Jx0IZ6Xd|<akzB(K-li`HF9G*1LwX(^)zKNRy5xLi
z*nq={ommpuUqux~DO^jjsIp;sabm{9U3)}x3HtOxx2$mfa$LMIJm!Dy(owBRh7i9c
zS=g3wX4j>q)&D)&Qs&`GQOtxXncK4o1S<8HHulQ?wp{d*V<%WNh7-evEmPc&k)PX@
za;}Zi+0_b~Ad5_wJ=rOZyqO8wah8oL`Ik6j#>EpVV2eQzAq00aPzHC`W`$&7WDX}9
zTR4DM7a=xV3;M}Z6c2=gw2fABrzb>Bmlabk@Z<>@8_Z8839w787HdAH*hO$!mzE{3
zpzZ#v=$4Jx61S3%FRe7dMdFH~+BCt0#*ir547USHJm5iNV=f<wp++&pk;iwyC`C0C
zQ~%ShEma)mYQB>QV)o$+rKLrLKfV5;vK+q-vJbZZK{z+UUf(-GrbLnf>A)Xi_#?fG
zj4nb4I@MU(QANBUlx^Iq6q>3{%b<j4r1N`Xs;xO|dR3t-b)z9B6nlm=X3X%vjX*!y
zSh27Ive|!i7{x>!ab<WBaR<exgKtVm|Gx1T<5AG$Qx=drC>fhryBV?3hKq%k8=CyV
zz`*tSGxTR7Iw}>YJaU7B!CVJoY2W9{kG!tikZvSdn57Vns$fP}XZzEiM#hmwn{T*j
zQ4EIOjw(f0eWv3Kw~_`nL|I%NwH`Wq+;lh*7uq0YFK0{;jJoH^R*~BLEu=w05gjhm
z=wv393R?Cx-bv9wVj+>GoGQKmi|Gy_a{QOOnA8Y)a({Jy`z{MZCPTfHkL-jsd=_kC
zd2|GP;4NiP>dyh*1ynY~BO{&Z#n`264XFHD^SsE_<=UU+BNfi~ucFemJ>MD^(68va
zb6HS8Tf3)su5^qFD3`jnzbcsi%?sivp^$H1hc$nD*Nq(M(Oo&ab!*4GY!2|}#8G*H
zh^(>8T{;>P$muC?tDQf(R9lUo+9Fo+^zrFNVL`e_FB=*=LV`E-*%XpQ-t=7c>FS<w
zZJ9ejc!rv0G&xa(<_n0cF|J)g{|OBZ0r4)(bORD=_RZ<}=C9Q0qsIo0>vD%|jr`Q^
zC1OhfZ&;!Y2gR4QDWG7`pl!Y!O9g`=Vj;m0AGW(e79PZ(y{T8UfejxDuBVXGi@ACH
z@c0NvjP%0$l~UECZ}FH<oBg0yQ-+^DH+y=0-<Bk=kW;dTKLLkop3Ked1{-KwvxR?k
z*PgWP-!edr2a?q9T|*_4z2o=HPx*Hadi<&tKrjd#91rjHzjK#@#M|!kBP_&CyOeZ-
z%K7O>lBoK1w?Ho6vi+H8R6y`|1?^<nKS-?!#QY++pBa3_qs6waR{}HwS6OQ3Pmjez
zh4Y3TT2pm5zqs)6Xl*ufwheqrwCtxP8TE_w;Hmu^4w<c{#O%yD4yLnX#%itW`)D%t
zYji#N-UhCFmL|ztqphK;-)W`crjT_p5+cn1nDQ_NF>h!+{yD*c7cHm9T9k1?{QF0f
z9f4ZKR9F^Aks|GTJ}o^}WcsikdHaY~B$8DbU9->++QqGzKX5u_`FpF&7p8;kgPs!U
zVtWJas135zk0HwVtn6$>(I~gDc-8pilaYJ(Z#Cu!=Yej!3JB)}<aYwqX`8>PYHP0e
zx6VGpR#{LtaBU)txYrjlapW2`*fqT=cJ$T+${w<{nUeU&3$-}@0;XS6b;}J2))0<$
zCd;qS3SBX?zO4|tSLeDu^9!CP^CKy7uSL22ssx)Zs>GZI?zZ^XN^3x6A1Qkg?Tr3w
zRu=pj``fdy`J8|&O#PvMoIr2az7q4p4$rp(+t=w%;+d1c*J;%EZs`1G7JmmM@EXNb
znXJDKWNKc~*Ffs|P>p&!HNL9AdeR5>?C&x;ymdl<n4!hwMoMxKq<P+Z+0?B1Eh0?|
zV#nv?1Rs%I(TC*-G^l*ghM6Vz`#C+;WL@PDWM_xw@HJ_C{AG5xWPTf=c;gKtc5=Uw
zY`RGx@Kr%jknKk%u1WG6-ML0i9*bS-H}=}u;^;@dI)uUB`5tyDcoNq$y`u=@9XrtD
z_;lf0^0x8oipk|_nt&)R_;P)Q#4l_>(BbJp&pmZO(D4Z?%x+VW(JyE~l%RWR=PLr7
zv`3Wv3i;R7HpKi*5CKu^r_ytPg}Xt{`^Pc#^M}mN2gj>3#QMn2rc`SC3b|j>fcn)<
zlJV8!=;=7Y2P62S*tXrbZJC8DaiH66v--9j7s3A3(!v!KFsG+N<iq>hx++*s$2_fH
zHR^?PznS1Xg`X>vcQM>Y<Uk4c8)X<ojut&}cyEb-CZ(Sxvv)oGM?YB3_LT$1jRVjv
zap0WTD%tMns?PL@X6zG0A-l=Jpy1ew<aF9>(Q1?PzKF~|+M);F(!64~bA<{d9)3ku
z7zo4LHgz7JBx;4L>MDa4(17PO8u^Xd3YO(63UTJve@{@ocXdQi-8D?BKfZ%Utuwhp
z2UO=-JNOoK1qOU|?Zdp>`ljEv^X_GM3GYrE_NdlN1Vkqo<}OtJ<=Bokr;KLh)t^ll
z^pDi59`BEhYtTTB!w>tJJz>q9VE0PJ7#Pw;SYra$=6)b>Rt<+AntcBIgzz594hoSL
zX&5(R9i4|cR4Kvcz>Xs@MOTSVUkxlpVtWI5pyTRLm)%~9)<D9pBA|+(*&auyU4zIv
z1SO0l3=rma5dsE$XWu@Wp=ST<L*Q9r=tQn_R*}M_2h$y6xuTBDzv<Z)Cd#?#BAJVD
z*<ReIjhwYS+Xm!W&2p4vca~jQUBy~n<4<zzKwmAfM!e*+!K^Pq<5Qj^?X+4`p=wbT
zT1#w1zT`+^h~rG^E2b^0Vle6n9|$S!4;k482VR;7|0dwb4dN6tcwb8yJ>?wW0UmKK
zj2=00*|BDNO`$!nPu5~FS#dX#8GQ?D*}Gax>E)Kmh0<oYpXtn%x6KZvdK76ZgHA8D
z7OCtQqwT;B)Gp<<5PYqKezpdhksfLy)ebi+>J8+@53F%jRtJZ+ySDWqn5p&C0jtNP
zSB7>Fi=+gMnP}u{9a$@d{(j@4KdMFa>4+XsiXNLAstq160^So~k%o!u{#0M1SZWfr
zkxSB9gJf5&MY+aW6k?xKce2kvm~t%@fm4+)W${thRHRZ@6Dd0Gh;$6oQG~`Uz8sr%
zI{?y1ZuzX}H3P=PY^xel9PM){fk1?l+(P(6r_`{@BZoLEzzi1`G{;%34I~_9T)Mw%
z02`%v^>eQCgZ7buo#T7Tryyw>1@He{^`7m&mF8#u&caUe|BnBq>b=jRuGQKP4F4Uk
zPxy0jk}i8zRx^P@H1K3=QFxmMDH^X+jfBb+F_w?k(H{NdPKWly+K}Y#g=kZv1>DoK
z*o=<j%5&ly6@iRhj?33Q+pLnwmIYTwSB}f2k(719?xXd?2WOYhqSlrB(++Q%sf^QS
zmlkesJZOEl(<9>yg#rO0m7@~5N2}3Rc?aK|&-6`io9@jI?yMs}HNjZ73fs@B%a=0T
zqjfvr^uY2bz8d0EBytU=XH<?y%hugSO2041Fydz|Xtmf6*XH%_(Ar02)O-vvr*4PW
zn#uKdTMr}MS=+_1mN~&CrU6kr>aDH|<mZESuQTeDl%b(>dk4Seu9Fx+IY!^J)y_3%
zg`wv#UIK}A!j`oozKi1T2^AIjld%K8!_01d?*BwZq55tB)j7kKVsm<Ha35>dbhvEB
zO_zK!9@iFlF4Z=avvq*Ty*ijF<2y?W6mI8sOGphwM1Ku?k82h?^|e2{R*$pZ?5T}2
zyXN2%3&<!{^7&Mi0?JE&OqPl*s7SG^h)<T`HsxAcO0jD#{&X5sLKst&k!~FP`2?9W
zMEyiB&}ViHClIA@rRg0wy^HRRm9yRO|C<U^xZ?hr+CldQ=8PO&dHf#*^%I{!-_e)q
zScNMn??9$)bnh=~iSPue-%I@jF3^X41=sEW-=>bPp}dLod;d!e9$kxn_5Lpc`%2s!
zOn>mdL_hWw<yY_jB(5pFQNONC<1{uA(cY7JvJ}I7&1`7ERWQnPHM`5UeY*q2S5n)@
zZ2?Wx>Exp~XBeEB%YxCQiQ~0TkbVs}{mQ44c2GK<%h$&i<&N}d`dRtYi%yABRgryY
z>Lk>%6sjop2cBr1=fDZU;#@6JSQA_22r50&sRN`;0rbunV96IUI7ld2OBh+0_b3(V
zCCKRLEti5{W5fL>_JF24{SOi2#W@1WaoBv?+|Ph|mBLL4sz^*Sx3X01HkqN$v}&_w
zVwtFrMV(8flj<zih?~S!nY^=TI$EE0%2#KCzF(PE4kn3ze`ePw3KZ3=1z9!HA@<>v
z=R2EW<{RKe#wW+u74tO}Iy0#b@pTWHHd&Ix^jVkP{#l?_ERm>i_<VngWn$XnU4G4&
zdud%ms*!rhFSZli^O`9hzr1u@^W~kI+j>EB_rxST9dW#2N_h_(|DWv%=YO*&tepR4
zP5xKZhjBya=g7zBC-&Rt=cl1RI2cT*q7Y^;8y@5`zzYKKOaw>vKd;I4-_~UL&c^y*
zM(lsFVP{=%-WcK!TzU&jcX#V)<Zi{(q_~c8!S?zrVBnA<;iT*8A}pi<IvEgRLiSyX
zV(P*|7(Y1_O_;$e`t$SiPwA>#1FK>3zrluv)9XW0NF9Bads4yVpSE9IfBF=UajMSi
z>fWDh+!qJIk_X?JQyRsa?G`3B;wmo*!m0zoWNGP{N_QWkpMj+75TqlS{o+kGE?!<y
zuuxl(C<*F1>^<I84Oavrd43_+yoV_av_x!XNAIv9eed0l5iPl+IJ~C(A}ed#zMp8V
z2Hre;c-?;BTV9EYOPg1+!P;RnVo&KlwUza?B!(ZoeX#tFD5P$R9(_ovdDtO(uw9Gj
z(4&}82b(9LFol7~I#{Grb=C7KEw!s<+TgS#SPMLq!$p=?)L4|Quq}XwWEsON7@cV#
zz&LBP9(?5cI<3m=2ll^3#{1FOqqxdx+9?_lmOpE2P#H!8U*4jpDDU94wY7LH1NmGO
z5h(y#r3)--JE=C2&d?ptX~&dBbMF}Sc0P2LKQAY<KC@Zcr!iw4`ij&Un=3SLtnNg$
zfqLCzt@g8qFT7w%Pmuml&)8?Y-7XU>o-o>7pe`(SX1{@9{dd(=QWdzSewlv!e$t)k
ze#4#8tE|JsL+we-J^Q_yFF3&^TQy5Hhk(htwfFjywsFuk?VjEq!f?>=(r-I_K{g`R
z9#h*U3+tU`-)4tqoi&7Js*YjbEdG)1Uf&F#tappol80$RV}dAx3^JQ|=J>zytmN2Y
zn<#B;jvo6q@l4_$wB<t<ZW27Kndq6yS@T(}Sq)j^SzehiS(n^`uD*Mn+ba8H&+0c_
zH&oA$Hv-R4&$`bL&*;xiDT_(Rc$Z_Y8;9-RhSAXR_9BmQ6EKvqos1`LQhAMtqE~Ql
z<tMP2sg30y!;cm5T4*nxTfd?R4iDG0TGseC`TMd?vv;%id}rPpU&S9!FQ(VcJex<X
zc3rNIuUm$Re+!~-{pqC;qbQ`HmtFo-%d{hHq+t|0*jYFv_04nW;pU-S7AWOPH9|$L
zV7|4eCs^z&@_v7>eERJ~eW|8uYi8Tro`Zs;kTcpY_22g<L*D@gq(_eRt=g`<_ly|?
z2kbef?nSaj$d)@h+*L<i8PD<3<Oqh?mbdDIgycO2{ieomRFo2Bs#-}84;(qcqvi<n
zvTf1Jrz>T>2b$KHUgTci?4R`>9J)CLvzWFock!Q>o(JBzg;5#_bzI?y#tyu<Z_svJ
zb{L3LrY(TxQ|62iDR$6|Tf%K`xeh@N-wtkE_b#LTz7qi$dkp!k-v&E}yq805w6qIt
z{29iR&3Fy0nlsLcysEnvyT`gi96cHf`s-NG#Cgc}$dt*=BpDwZo_9ur6h|X6N#$3#
z7t{9kZgh^Mj~b?yvMzaJxu<!V0zK*DtPWoH8E$TS4+d^tZ$^55?awJP9<^RLZeXLM
zp`*@|Q!y782aO|-yseBJfQX*!Xf9@3A$otEVXb?*@!5Nj*?XjS^mc8&N58Uf-=thP
zU5j*LTt@UzeoXfse1oGAM%K}btK~B^>vxNwnN8!Or4_W0V*2-OVdTJk`zF99>{N2W
zY~kqN(ZY0vzUrsJ`wMg(X(wqrX_KL|sOvytXntrCk{kQ_Qf<-B4JQb^eFnWXTC0*F
zo&h3YJUcm_p+MVFySb%v>BeF_B3TWJwqB`-9-PrnvUUAK{Po*8%<5yyrHx;4SKYhY
zIq!ea7>z#4uP+Z_t4mEaHu?*`X0L?jEiKrM(>6J)*=yOG7Z${u0=fKE{6+k2{M}t<
zUOF8W{N?<W8;eB$Y%MrjPLpn>|De^hGWj?h*U4xz*>vB5isVujb+x>X`(%)HUM7oP
z8LQfyFN?ewH+(D>ig=keKF<5>9<%%&-Y%yvxqDvUs&A(={a)UtZtsu$ynNT5!=?qD
zKQ?clkB;8n$F^eM-m6$C%?@*}vuSyh-Azx@{+iuP$#St?$!<{<9ZSz26i3Nnr%!Rd
zcOC1@bIO@zq-RXAv9o8{@+>&lAB)Xn%jIzR%-uQ6>t}4*@yy*x&qvAmnR`}Sh|a$h
zhovua@Gm?2&2PxHuy4#cJIzNH3(LKyzpzoViL-}z-*2dYr9<CTFm<nfs16<XWPfbm
z?aX(`3%b7q%s=LIU*4t9KW2Bo-j&b0Wp|(7Rh8KBobzY-c`hE)mfXsZm&D{~xt$y;
zvzO2*VCIbRV!BGd{yI%z@@;z@-}3Vq`TL7HCEJ|)`VqZE!Cf9vfcZ-7q;K-zGI2xK
zRdU7VS4$!TZH6*OmGxNvJc0uK+giT@wr}HA4Z@cG(_z0Zf?tdMNWUbu-qxPyk;qYC
zxH4%E!U|qJ?fy>OZ&G}b#V|gMdJYr0(ZIMl6uVK15sHx?uavLiuTa`ima`BCpV#>%
z#BSyKwPH4$?iM+IFk6$}Q0IK+8DLZ3XMiU`Vm>7Mzc9Z~$mt=OCRk_mkiWwL_QSK_
zza#lWr0zg~8DIe>CM^Nw04xX`Fk46sux)_<a)d4v3YdK?vJv0{Y8f6-hx5u6Oi?!j
z3~uJ^#RzT{6yhZ`V7wCyYX)osL86_5TTlR}8!A9=h971x_!Unt@j~0v06-343kW}c
zLI9AXpaJZ_t`0_oDxd)A*J(2+s@o%?R_X!MfgCzezTl1Wm`~Q`x&V~5YaymSi0`;B
zMKfPPuwm~AtSZ~2+2}W5{t|DKGmGWDpU|t604Cv2Fmo`!9VCz+<heEgrR8QM_q_rB
z&A}YGqSp`lToKbfHx}T9LDGEi37xkJ=C3>hlWjEx@J4vFp3#GD&A;g5(t!foRVUvz
zd1Jo|U>ep|Mez@n$ZbJ2kqHlSO@ap>tR7yVgT0;+@wV}XeSw=1akgAXd^a%<Xpz{i
zm_c6MDklN>{YIjL4a(gBUL{IP?p*>NMEqzPd%gA$W?;H;44EXtbD*|Eu3a!8<K)O^
z{yuTsIpsll;s^!{jFzGrIgsb!juzqWZ@uB&MdgMtAF6={h^$HGIpV-}xXp#3t%FCH
zMDXDNPw<x_yZ{4&JK<34+GeChR17?Fa{TS$0Gw6m)|Kt<Vl=JXCc?vi*1?O2_Jnsi
zAy_5@*%0puGyMb)P%}NS3|Cve_#$MZwS@NdH9-QSiG=peL3)uwqOT=%;ono<oiN3=
z(~Q{wqcyNjymSC8VPcatFvr7XNS`>vW%x+Q>?q2aKq3inyP<;_a3*p?+iP2MWiV3_
zVs{*Fe1JW=p+bD_NIa(yu?eGN89dnxlVjW0nfitUbFfF~^*dm+_KB#mw_;TAOECky
zeYj<SzzmWYkXyK<k;HEb1t9bjoQJzlkDE0YnTuzKU>}k{K?piih?N=?5BcgSuPzMk
z%Vkzh(npllpc>~dC|`>%-zyh*lqCdnL<C9^M&Pptmq!~dKsOnI<B~RD&Af#0M)cY4
zK;IvFM!GZ1yyXKd!yiMK2Pnw*>c0G7uV5LCzk_NNqJCC~JqGCGKG6g*%!75`_yC0b
zHbl$=Vs9?<kew5AV3uKgaHxB71A$XXEHf(5S|`20QU6#(s}p~0mTk%)f>R-?{b|*i
zzBcqhh&CZTms~SOF0*zaLFZiK>T5cX(4+1O9SER+y}b`kbiBQP*e(O*gVX45hPW)l
zKk9>cMMi}|IzawK8HouHN6Z~yHb^lRT0+WgVU7UUtFifa5B~%F6DEH9G4Y!YM@Y{&
zxxxUjI=r422=+A;M2F)ljNpjYfd0~v2elIE0rO%|W8D3-58#OIqy#mPF_W#<2WYN^
zSuP_Y8Gx>h;O|X-0BBqJbN9;TapkswO`SQ!f&;4L{PhV&9SEMF(9hT}?C4Sf$ef^V
z{I^sKhQ@7#bs+`xnR%q<Eu`*#NN4*XT_*^BIfK}Bbd~R5g6jsS8@U5iaoCR-n6kb2
zoXdod1fT&t+bd;TjS)!SB_uv42%8C#Q`_7Se3MDUhan`6zex4wL~!S3bg9BH>-ymS
zT^vEa9evr4!CV`N^@Pao!$CH6{yo28{4kL_TY5dOb3d<-XJ9-Bx4;3eYv8U-L?AR|
z-#+jgJU<C21ds?HNRJOxP((obeTgO@Kw!BVa@h;~YEucxCK4Vil?iO056ol*{FVw}
zhb~mi&?1V!Ckyas_UEJUpO*6XU@(vomi<XKngRt%hK0|8WijCfXBb3{FyLg!m6Y<g
zLR-d0m|=zU)<=7iLI7zqxi@2fp9g=s2k-bd9`zGt)Tl-pGq4wnBCHez@=o5h90k&q
z6xW0tHya6DKO?hJ1ZBnye6$xb*AV6F11I;(=|XJM;fn!So+gRw@yxa3zv6^F6X6HR
z@INJ!<m&QoYa?4h2gol%$es30K?mq9231wVY+*w>Y9f1A!*IicujwOq{-+U$Qy~lq
z@WO)Rvjz}ak+}D-BKey^p!|iuSps-j12#2D^(+Ctoc^7feWPst7aV}DN`n%7p>8|?
z%r}r8#8thaz@4ETlED_9zu+R9xo`=7Z?(p&C_?=vz;y`zp#jW&0q|BAcxevMUI(0N
z_v!hkiEjCKYa8YWxjKU_)`7<|`B&?cJoz)UKc&N~%{UmqSb{w_1p3+5fkTQCTU#NF
zIwMqCK<Tyk>)QpeR|yRypQ_Xb3N-tBH^*`JakltZwg5K1hOw6I5I}!0?@FQW%Al-M
zF?W-&dRw3t>x8XmL=M4m>yU!5ybTZr1vs`f5ti$*W*G3#ilE*+5YEP-odVJ%;GdN6
zXXL?n=nTGDhs=mf6#4&d20O>NN`U7xhS2@Fs#+^dunuHd--LcgfE7iI*0k4YzRXPm
zGa&MU?e_=zB0m|%3O#{@0KKz%5a%d>-JH|hdq6MIxlAsh9M~7Z)Rh1*n(PDB*ai@?
zdp(~a5`OYQx_^EOq6JvxA}<L%eFH3GhM}&S0>ZZl{0G`p0O9C_vormyM?%Pi->2Km
z0KLsziz2TqfYGKs2lKDjV>`Di>6;%GV1^K8y=Mjius?cDZSJKk6w{J>E(yjBBt@)7
z)(`ID-3;Zx0_-ndyP5B-L{VI{BA)+%r_qRol*p$>x$APGmWZwyc?)WYexdk(bP<cV
z@Wvk^JY0J?aj0o=xSBsv>dEly$zk9nNNF%a^6W9d`hZeTquV3}iijs*@}$5vMFk9G
zQ$|$E00$`vS{lrI35!zXlc<-x4|4!sICS0@E(9+!H1BlAeqZ?591<@=A`kID%n)Nm
zD!j1a0(v&EUj-2~FuJjhvlNM8R0Xu=bm^hR1-M@*kPy=XZga*lNrTtfa}dxOyB(k3
zf3ZMvb`Qkdh_q0!l@EPK;R$ae$vbm7GjC39DK?$E0eUn?_Qc|k%@%!~C7wY5i4#VZ
zE27@yD-p)p3TZ21YL%{qVdX)+u!7_p6+bD;XZ;EU!ww6EzEEazFTc-B-Wy608H{U0
z6>@z*rYhldcRfQMFOPSR1&)n6B_X$4HM2G#QtJWN<9uf+%O%n>aZ7Qp-?4cq*9KYI
zZWW9o*cKZWK3XGrlu8-E(@Kz<GjcmrmBbes#hJLBcq3|&Ed0kzkU6?1B!BcrY+Hf{
z#(U(-DBpzjSAqSn2mzd0iB_D)zM)#;86`f%2fq*(@Jn13M0KeSh@(8yRzYNm&54cB
z+gygDW>V;jn(a-hKY_1+DHOvz4O{$8<PIJ<g{^4yTjOV3lSYMX9OrmH%$z+K=Aaq2
z>P-CM^Ih9R*S>I`>eLQg_WmQ1LtR_+mPD^G?Z{%(341KCK$z9=ksX}I_?2+Uou7P>
z9{63d*TgZLyQFX0Z`xnTmPjXqePCm1{1x%FM`mEjj_%bzac5NC9c{s4;xEr4*zoCW
z#~GWR2>vV{0p!+0K?~Sr0*J4C$3U$yRB(*qMsl_S<a7t7BWVss#e@3yVK-KUALdOP
zkoZi%`li@d?BFmo@#|l-o#9E^I2SbQ&!AfLQtmLHjPaw7(G4iK0G?FC*IId3Nbe~!
z@}~lKa)ax(pV2A3TuP*Yl9Lk^6SuV^*t5*b*p*WgM#(Jrv&^07mk`mY?MRpVH|)2q
z_Kk|mqjI8Va5gg?6Kg`T1k$k`vl}9I&hh9o+#Co;P6{ZIc^V(TqHASX<zYomuec@|
zYS9OQLNdbYs4wUeZ{wCBiuouJiw?FAioIEe$*;$uDY-3a!e@MO3`)ek(?QVuu_UXF
zq#PIAzk8kNgB1T*+nN}$M&a=<DI{#AUYsO3FQ~6n#^Wo>7Z^sod3$+Hu`y=mY>rB@
z)z=x3W8>dGBX7KQs?R-TTHM^Wnhn!zHR3)OZ0{CY%usG^v*=9Yh+oF5;y-IVF|qb2
zGSbJZr>PGi>iv2EUWilZYK~yuuJtBc#5oD>e&^Dp*lsne8{HbRr#aY;KinMkE`0vs
zRKBoVKiE!RZ3Jvutlg=iPIAH9JGfogw@2!SO?OE4k);W3gXSCQjM3ZsKAsj!5IDlb
zR+4}Hd$sfW_i$(0yq=KGD_CtFfy%Vdxc}Gh&T?Ir;g$7SKl213myV{>Pa3t0U~Z_Q
zKLM$D))s=XdUJS?CJSV&6-{5dxY<ToN;##u9YLdcDtz6FV?k*)&ZhV(E~e@j&1=D_
zpjHfmkk`Wd%O%dK$q5`jgX745yWLzif2{01jcQJ@3MaBGzE^T){P7As08+lZ&~;=n
zfFY4MN8F^kdozWpMTHx4$s33}S)Eb`(_@EnloT1sAXZubU9$6;B({NAAk95_-=KKx
z?SXCHiWv6t47+*aMD4??kb2CZ=|~p0qK7t9F%fgu_)$B6XTqEj4ZU1cD*ZFnAR)lF
z&m=)qp^y+4zT#OxnfbK-!BE!~I|S!jh%X~N4&zrIMV*FyC15SdjMYAPF4d1-zb(^`
zPyZP{mar;wu<e;kvW*XK`7OU5e5<mOt32~vs13J<<ZEk>C(_P#tvZvk8)zPYPs?yd
z1h*BX2C;`@U}M!VrX>Pc2d(i$+o?=75P(v^xcR9X<!)OQgSb(T=`&Yg6M0jy*kGF>
zV@r(G3Fd3;W&kO@%QnLFCEm3=oXC|U*Wc8ue@^N5GkL!;`#!JLB$1B9&<3gYbjWc}
z>l8alx+9urT(!>T#?ylv{XQgznNNR5*M5b$KOJw?jUnmyJ!}0Hj4)E;x6O&4cx!fj
zIuf(9)cv7``*A%oHY#({c`(Loit@z~^(d8`7Kq43(E!S3tNO2$!X`-RV&&d+wC>N%
zl6gnn#5i(RlTWMdzE5CJIOHnKI3aS$1I6%b6(b;!+vw6(o4r!^#+Y3gxFqMBG#oKJ
z0!hb92W1t*2H?j$H4L~%r#l~*o-k@#c*4ny_|(pU=p6&ye;WlLx$pV{$w#rk6_g}0
zTW72VG9o7(m}dir&H$Ptru%>5LwxIbWl|WlYYkTBpn6Xs#w5qW#=y?{Yze~nWm$pT
z9!lI*nS4HYVKfy<j&FeKzNROzqRefX8z+J*eGz_sS6Awhvwd}|8Tl$4vfw(J;1Ql>
zn_FQQ7z^yuGJ8#T%Ww46PFxBe1k;0<Ts1|wGG|uCb>EwdavbC5cA0})_iB4koLi`g
z#2xvtqX|jl`KE_%NJrgsq*}<$(ZjM$Mb7wGB7wX>&icGVg<(#-o8ohE>(_vQdE{my
zi)KWn-vR9c-ZD^5rOfy}&5+1*sn0bflEw-I=$YF;U`;hm5D!HIgMvi!6`k#!e+d3#
zYSkZwWs4yu<l`A+!w(<?pSX6xsf7BeY}^KXjOVT;$D_(<>p4&zPU7}-*@%9$duwte
zU)+1j*n~`QK?~cPmRx5mcRmQF=i{hTzxCqmeni~Z(>>R>H#3aoS4^fxU2)TBIxRFq
z<c0Lex=jlwTpUi7>wj?%WB(UAkr%_&5ahC&S<gYYx$X%b?perf9w)w#ijFPc2%eH;
zH=RBWlpY#vbf^Y#wG7#<qB#gQ%_CX4Zi&Xy@Fx1Om-wo>I)8+Y;1g{<Ex$ijw_7|!
zIh6~3gK#_78X9ib*LyuKTh<`h^6L%k-<=z0HB(4SK#3hNc`kxzVf(4g5%tmShv5m$
zHb-up62MM^J&tFDS}}jN&)lIEb&s~p5I?^rv=EEPznm$t;<yW|QAI+m+*+d=V;+%L
ze+PCdJUW@FZo2J>yy<XPdSkE#ZtJs#=W1qEeUsz}%wW!xnJLIcK(P}-<exyFv>;mw
zJ)cQk(p+41HFw2LEt$0(c%i`p67&K8MocIkD)@WCOq%($UsIB1ID*HK@R(8$3cKQ;
zNC=8FBd=huVo}P$RD>%3f1I6jj3+^~?|1B(9ox2dY-7i^ZQHhO+vbjEeq-CVtv5Hh
z$@}Bpn}^itr0Z0tJLyj4R9Bt)e3Q4S@vC2KPNVVBHyWMKc@m#eZ^e=mOCP`wGhvRC
z8@uQ204KAqi?kig_2~K2(EsQ|_+UELK;UlJtIa}%RuJ=bP1P*75W6W2ysLDu(E!65
z1A>9sJO<%gISjcx;<Ln7Jny{v7EOAs97<Moq?pK2zZO{{Dk>(-;g5>YCF_?X(I>r$
zOh<t3!=jmG_v8+p4;vh$-7%MUqh4zna<})WC)On$YX?)Gp0BF4UT(4+aJGY`>{)H%
z!mR?nvcWQ-DF%=yLYZg=f*WX%$3ZyM?SyDVvF^z$5P&U3Uq{|wTBQaIxYrBk>-4>F
z)?;xe?;5M=xY+HNwX<5Sa~;t4ZO$k0KkI_`pZuFO;UAz1T4=zh*bTXPgdVPN#a9r6
zu{`}##HwZC>Mb9wVi^<#(|BaC(XLDer0r|)%JU%g_94T|sLEi$*8I_&O4Y2G%B&Dl
zD=I5IMtR;IFKr(YemOAtQdL&dYN}~h({wtb!h`(Dv|HJ9t?|W>-YezrLBQkHWw#ve
za%VO{Wy(ZTR$5q8R$8hK+wR#5Ue@)rA9{KTLce-{h@|KJ(NW=c_wf08ICZ>-x<>!E
z74`4^ck*+_3@*}Os%1MaK_u5iCeFDbu>q_8k{$}KNnRJ<%9@AZ=BPlgAcEYXogZRb
z%-iGRJr^hZyM1VM56^k{1}+3m;h!n?@~r;AI`HRzc*e;MwGY2^c4@)$Dd=4EHHmJx
zXyMLz-?(B_&?wK7J(EvmjxK0hKi-x!eZ9wNm2wGdOdHz0u=a)Bq9%~d*q(#_8A85S
z{oeB%MhZN;*DuN$ECvXHws{%1D<Hr}Of_N|jY4@v2Ps~v#d0x4nX(fu1YLzQZQE*8
zZd+*MI=9l?XyBklG(`=7npRQZSd)?MLe|^x-h)davc6zNxiZu(f_Y(YZGx8~_GHu{
zuDkyWasTUXHV8bO=kuyqv#r~85WR}d0La<LOwQX2qko>#X0)~2Rv#-QkHPVWAb<GS
zm-^WZ!+X+k+pVMW2eIo~?vFXuD=k4RvRt}p-PYjK-^q-b5fr2fIMZCVM~OvVd<$9c
zt%nUiyy^8^>~^wF+3hL@LgC!sAxHrQdjlRO;E&Q_*IYE2ZG3g}1i2>0C+7kS$o|)*
zsqi_hfpC(IWuhB`yo&bd_Jcv|_I1m=<}o`y-ZV}FS8aGk(=1fbE_T8c0bq1w9if`h
zsc+A}G(ZFW0^J7d2Bb`-b?fYgn-sbj&st8ocwf#H=WQcBTvxjTLnoAb%4&I-_WQQ!
z9?<lk{4+I5>9uOS+I(v7J^;zB+wSv!-S%>(?S|%Vk(0au5H@69r`?t|S`8k}OokJK
z*IRD;l2d7#Yr~q~58|`)>1<n^R-@d9vc0;MO(>I@eB{UV0k5qzx-rbUZfLQZJr`4@
zEmK_XiFdAtkolFMjJ4QS5#SyXBY{L<I>?leT_}{2+@nCJ%<(vpK!f5bhQ;F5gZ5Tc
zgZbG+tr~g+pi4J6LH{)TgLe#46NefR-uCo}7%Om*%5mq)zU3<y?1S6hlL-|fSVX8&
zRPs>Wg4+Z&sjCzINYM*;L!8-<;gE=$WxuI03Q(Ujx6giH*jKq=ab;)r%c4rID6nHp
zqTf9>n^E?}SMH}qG_Q5Gt-r1hbv9wQIe%!_7e1-m*FMocVEIJ5OWjAjv^)tt@cQJw
z$~h{1h-V6dKrIEE8@JC|KRPBKN^|=}hO4Hu^7ymi^qU;w_&!JyA&(zaU>1KKuo>oM
zhtY_Oa^$QKG%Y_A)UjDLyV!>Lu~^9^Z56>m9kWysOJ3Jyvc||QfjyB#<IEDQ)Pp)#
zATJ1%<IzU^5i1ow7Iw7zcNYwWR@qx9ig>$Hpz-&Q*-;h;bL>}jOHIp_?Zk?)D^)JV
zb?fFt)IShAJH~f@nWv!h<UniIduSNwBvNU!SXGtpM@V0%Z>~dZvnoYc7dy0-06CUH
z<PtW^_&`>ErN%_o>jfKw<xq9c)A%?HlMqEeJ~L(b3|+-~naxBbUN)R5|N1->A5(Pc
zZh~3R0`QM2FkL#;E0u;DJgwG`aWEK!<&XCE*J;*J^doqTuvP<lx?xQSc)2YAPO%oR
zw`4mvRpMIN=Ygx1<g#AUMSUU2+Tx3!kP!~jo@aVuywG?>Jc}AW-=l)-VNgKe09tN?
z0fBs-N5OADgNZuj$!+bgj!Vrh_~TuuNU6w3DO)+Le7YT%=C6v=m(o&uX@P3h_Klk&
zhcQh}{*%VsF5+KqCqHSl!??e7jNt#Z-}wc()7qibdjJ4|5T74X-b+@aj7pgNKp!e;
z3!ho#)fwsAE&i)ZcZ{H<BS~a_mj^-lHMeM0`6A(XQX^R0B{1S=zzf;)r_LN5RNAel
zTt5j<FQqn}10(Cy3sbSrgCQXt9~W@z*t2~*VP!WnTW`D+Lu*pvQ31^aMx=yO0v_D{
zXNIi1Pc`i_1hzC7fADMDue!+m34ITF4=1-3Las-8bs}k{Qgk(*4qcC9;Ii?3L()|^
zJXnl9D1DL`yeG_xKQn7lhm3}2N%Td``*e#QdZrAG;lt?r8Fi5Skal7?NB`n(&cF!y
z<KcCNv)Yqwk)*MZF^<F`G;LeO)fDjdq+^OtO<3FhMr!_y#{}i<0u2lDvSb=KapySY
zcwiD*^1t7F5mnrJ5p~>>W^W-l_~G{R-0Yf8c7;yws(1AJFj3b-GRVGL6sJuU3Z3Fr
z#t3A9d{CP;=b(dK{HD0t&b-%aye{GgOP98}@%xw=3$M*9kVbP?b?5Rtfs@iN|NO1b
zwx71{dpV;0g35Egc~Kw14H8<M+WZ^YE<pS#U$HoJ_+h7~Pu8=9-7Qpa3BfqjtYPIg
zBmIKKr==tzTh^*nu3&7rR?l{(YDVeQcHaiQ#WG-18jD#Nr!NZqKJD~_#l9wO8s+}b
z0M!vBZ3T6Y%BT4y$t?S?M2xnb_&=&4<Xb80?o~O`3TyJ>H~ZnmKUI%3iHj|zfmcec
zZYUSCi0Q|5fg&_Gy9h`cxi=6@MobldXjk(X9ys9Y;XEEpW${~jMHd)88Hp1@jGGeB
zo>NJztBUm1NPsz<x>SrXC!_>zWzM4cfi#roE9VXf9ys@n>IbL=x~vLr9qLmU!^x@>
z>n*bi6fIMc(sw7+Oo$ST#}Z1I=2_>9kxCR4DXWdIkgw-?$5<?|iW%S??QEQP<d&?Y
z|J7gUZN&cUD7e6|k6_o%G^FXfSFdg3&$R9Ot5xhR4hwW`NV-5wzKS8)PsX6Ac%upk
z1R$d)ktadC6jtXpPEm(%lr6=suB3cwKZ-=AxD&B{N|3ekffCgl-P1>Q`LRmd%FJsv
z6F=ltUwO!-HsXpsSM-^+l-80EXqPekj_unIqIBzsE6j|Q3i8_I{ixD|HST9wX7pxB
zuXB$zVPY0ioJg{8N_tGQizw($wBK_eZM}4iVBGty7p39t?p!A!SWnNFa8wV?6*wdF
zS(5iixTaGAWJT|}r~HZ-3m*)wIhjwHKOp2@cgbrg)F4vDrwW{FWxr2d%ffpPpq}1A
z>e8X~|K$<Lgm1gUtE7=RO;Mf3H6O)+Q-wllnd~fyxD|xH6YQ3_y)1`MM@W5kX@BBs
z;lwp-QN5VgSI~}LiYyRsRk3^PiJif+8hnvvmCbGd@3gqM>;A+{ZAWrd-Bsw7A7`~W
zTY+s`1&nIOIltr*>?xnFsiUM+GAG0#d8#kCYnDSkv9l6B!Nr<6<6#&i-w$qSMsM!d
zjGl_I>*~UyqFl{P;h0b+7N76oScF|HEHf+1gmQil7x@_32aUz@cG__oa@t}}2;)>7
zg8e{r6CQg?faNA4rz|KN8k=KTki;s<xaD^YV!q^aLv~fXQR_AP0*LhVtJ`(sp(7>%
z+1uK#;AWRHM`0(pYIjL);mwUMn~vdkb(`xSNDG@07po&CVPaoF#6F9wGAmVCDc8lN
z;iw|4;viDu&jPaYtW&SFKUoo9RO_Q7q)}K*oN<?%$rI2<Xt0AhU?mASDPiU&;bx(s
z;pV0(=<V($9?02EGsxW=4ZmCKj`^Ta6#In=XrsYBp3tj|1REK`(N)msIAAmeOAOvz
z={V{MALqHI8k&ZGS;TR9iFzHHPs&V=wBAYA_5JQc_XMkMQ%?`EtZ*p5^!VjvNxzec
z&FY3EM`H?8FWx~xZaPxy@T4l_qS2OJpGkA>DJsEf+6EpRie;TKbCPo;n}L95&MQvD
zs~ZBjSDtv3i0iLcL0(IRhY*2-!aB+!j)%@V#UYbJHW`zJ$~wX!mb0tzTOLy#Q~8u8
zp8614wYo+{*HZ0Bi;Ui}dY-4qm))Y!F|lh_xwdk78M{TyCU)&=gU;6VG3#$XQAYMJ
z7p~9rGCUg?o~mEL3UwYRviohVDSwiyI-ET=TduzfysElfYwucFE$eJn|F%wdIbNDx
z>tSN28pPD9!oy>vB{Mr2g;8J5d`bCUZzGugtXq%cWuqQSeva4d9~)3+k8;JqoU9ha
zj!=||`Vb7Vg1FCm1$}_$eGS&3XZ^CTXCZg@CpmR>Z_1X@+Pg=MY!GBjd?_K#TrYk@
z*y*ceXazfhZ#fZWhq#A`I&U^}jys71NBl-#(EmF}_d7B7t!hNWc+*OGpZrnH)c%zP
z_+uiWQ-v1D_8mH__A9Ro_~`toJCp6&ShZ=Zs;i<a{C!{lEa${sg1^;mgJ#U1DL!pi
zA`k-zianw0Kpd7KD<|xDH*VO=Xc`L?!hE|P685Iq8~%xeaCl-=ns~tfRX@-k61yMl
zUPc~mB#+UJhAU-!X9P;|mc=F7?3N4eBH@Pi1<#AEr}z@27{3jmt@V^Fyf_UnB<|(+
zBBgbONiH1FdjMp9ld9mG4Z33a4szA;1v@VO*^D&KBuz4YqR)OIgikFXNBEnK>de`0
zO3{~g-aO9#CiX*^<Zu#ayNnqBZR+1q+;(?~_Ijs|<#DLm7`$iet;rV6kR;PW-;M>u
z{lF`)0+kX%wPPBGm;VFzebfRZYm_O8-HirqE%F0ASrFWjpnIp;e8_44gG~wf`H{z+
zozX%m#=DSL+;H)2xnKpEvh@5@@FxPBNkd_Ky;q1!R!r8n2?O)W*oG0W>geJ@1aniO
zbK_%j8Imd!suh%K!EFTTTChU!kDyGs?gSm0?F`-AW?rwNW2Rf`7l#S{JIyoqchpOP
zPggE;62}cA>^F24b3f}Z*c!POio%@_ZvjlOE{|kv(O(}%#^aw!rk9fk*;*D-%N`ys
zsf;Oz@#1d83Yf?kaBOL=d3{*bxy0c8u*k+C!=)+$e1Fdj{FTb1e#9^D1b|7|t^7q`
z3=Qpp*AWm*6pO<ZiiIJE-hg=`;l9!Oleo3EpTw{ZS(#gwXsunDohy2Cm%hNBv071_
z$k2H9Bxax~7RrTup?t(vthlY>5jeIZR&ef1jJxrMKpP3cs|DC5#;~k}_QU~jumTw(
z&TSTe$MHt=Y52R<u_}~j2KO|}HqQX=K32_~%ZgutZi!xrl>qNpuh?_MSH$J(gY9;m
zI9<Ed`2vM(;zI5z33-mG{xYv)YkBHy1mVqNFB*$TM^&Z5dFhKnK6(z=5e=<RFY_F{
z8s5#h`w2tk-VFC{SJW}J=uOA_DTI0`v0m5rZq{1a8V>D9#X2C}DJ(XN{vVAa*QQM^
zUGDel5gkungS){4BdhSRkB~Vdh?Q!N062p@NR}C7I;%zeKl#;9k0opa_y`<0fEdWT
zpp)hh24?3ly*4ujX~#WN#b@a#KrE5i6&;6Yib+;q1Q5W6Yefoekw#rXvC-*t&ffEE
zt1rRfHoJpoKPkHHWdT{Uh5Pp4Xi}C*?08ZoM{zBWm(6p_@PO3DfEu#*4X9Ue_LnRM
zXHcx&uuur`oC-`UE5-^+38p@O3*qKH5c*U`FWBHwNG(CTpA4VSbdWX^h*s|e1X+!x
znOf>|9Qn1Y;;wWf?%sNmxvc#_lI)y9uFpYy>AVLuK3BHeuNEQP9VKHkw6R6t0Fz{$
zKFwMELSt%QH7#2&WWK32b|XD*ePSme$nLO--|~F+yTXmH!sOj%UXtySb*G6(dYwph
zIUhl6HmaLk01=vQ#m;v|!QL}csg=)(rt1uR>a1X9Mi^E{m-AdmFsPTq%vAQfKZ5L7
zr}l*Jy7f<SZb92sBltm9@E(f2yhIyQVJ;w8Lm1LRL=+(qog_n0k5dWKk_R{90VSKo
zWXW7m<itA3{5lq&f%@}2y>A|`&^N)M$)&e$?0+KN(j2z;J`!{@(|L#3Hd!2!=CanX
zp9lKbArVw`-1FSrwSxhgY=RdSd4;Ij*WnEJYV<a!Hzp5l@+-8ADsXdV?tz2@MUCZr
z%db3*F==u6lhr~9cuL61<Z~K?ZzR3Yr3MJS3XKXzzij_8AXR>0{ZO6-05v{5mJ~9I
zUsYBc&F3@k0hbeg@~f9+&M=v=@=uPkid$5L?!>L2?CZ$U&x0aqYDRnq>_{CP6i5q#
z?<M-K6H9rqGx~pIVqryNwebjH!=Kl|1ID3}=K-yz5^`V{GNZCL#zkfdIVI-W7CKws
z1BmN~<d7|RD_<=RK<C>v>#XpDJCw%0r+xh#o+f{6ENiq}g#Ioz7}HN>28-#K_nkIE
zpKH<r+z(+^g9G26*W02|TdLRF+}GEf*Zq39{+U1kc?Zh#$%gq0slBnE)P4TF-xr<0
zn8Aqb(_>29kFJ;%$^^1C*i?8v2(5ut7HP^Yw$0L9#FvyPSb^siSGQ$YA0bn%CZY2B
zY(%a72w&a$Q;cNmmPbT0{MiX-jSt<@LVh&KTe*h*Lf{s+4Yfyj#u?(`7|YBJo;0!P
zLw;lC*z~$abty@^jA?oKZ^#JMm^gL2L8OlIfIcq$=@VeaSqd*hHNc%u(pjI*)<5wf
z_BJxvMS)Q#5IbSQ`6+<4t`O;*QpBo<E7emBXqYRgF+}#_^Lkr)KVP5LQq4Nlbeulc
z%}sX>^&9P{`kVT%1F2)gu<$i{^TggaV8Cx;6B&nGJb;)l2`(6eKnbH@%!rwld1bYq
zWl}3+?`K4`hXy6Z>e|q1?|S2E`KQtH)wzElquVMi(EcWHNPYstCyPBeopuTW;p7mV
zS$TrUC*2Lgk=(cS1^bBbOMiBYXA$USz}!J<osmp)Ol6oF8%*<@`9+_hDt7ldN#_jD
z>m$qqo@eTJ6W130ueqC7FVx<I=X$>c-8%2(s<R81AQsPmJ%a@bvR-XG;^XFVG-iU*
z_~W65D{O@>Vt#vT0S^TN7?#YS(as^*M3eRwB~VUm2`4t^a!8<+y@SQDxE_1`Gsp*~
zIxrkY{@|HeS4yrioq+B+@7#CuR+B#^3Z|ud$F4qfkrvdnJ=+F-8Q0j4p%tmyN=t^^
zoSkh6@=Kl$$EtnH35c?jwVOjrsY$t?1=i5aOTQ6Y+;)*O*Le+wuQXo{h^_G*Ue}L~
zRLkEBIYHh#HB4P}HBDtc472OML9*Rne0ryT;Z*=$F50kf2;9SUe!wKGkk*U65_d?%
z(&`tG?{SrPNYn6j@Cdj<Q;6EW(-#>DdOA8?Uaxx{#^Y!F(o8$-_FAt>tC1shQn_m#
zkCea|l(rO;a@+qv>fz4#QB!dNC_eBzf*+f|WsOREJ?Go-d;wQ6pt{HTI<-<kdc5{~
z7OCj#0`R(4dY1uI$igRcq}(guyMl0S-6LyHkaI0`eL!4(gD?yGh@L#bYnIs&ne;AD
zKq<jj@ZSqfya}Rocwzb-Ziaj|+Wo0DN{ueJfI*ecH18dcN}yCyTsrmU$@A4ApX>FM
zmEEKfKlQIGgHZ1JNq#PvL3;Q|d1|xYp8U^ww|_5fqDe5~%;1Rt`lJBDxm5WwRuWqa
zoQKa9*JcW^#!1RXMU`sO2f#`<H$^^q=3X7&{d$WE-S;ar{dl4?I{yX>!jf~LEH_?Y
z4&&LI7)PO-0I`L;%-OY*=upKg5SHR9{Az5FRf}Dw;r7J@qBQh0>lzEkEV2jEw@cBh
z@H0(cD+vG^`s%A7YBm)P(U}CBNvv{FQX^ZYmf7sDmvkEM#BUQ=M$+)HH6u#()GFCO
zgX4jNglmLzzc=+h^#<H7+0C<TvV3SJF)xvy&7URSo!%ebVe3Wv$`~U=tL2$az%fEb
zrlXe=?&VWSfREeWoY7dr`F$`C11z_4HMQ7#^w?W^%CuB!lzh6qc8{BjVVK54eI0iM
zEqBF!-)Bqxy_~0KR4<KQy<52AVPss{XuaMMD5nHxJgeWnT0;~$!GizT80Ld!lxioR
z7>y)zsRX-5YbBCcFvp~&@aGs!Gd)kxz~43_#1~+3VVqkz)HA(ss7vnx5qG@IiQ6@6
zOYS-w&zK;Sf=?VcTDed6QrV=i5BW3l`{W;oUl0pg3Bp%$!%>I&G_@_hyU8-$m;7O}
zOVOYSf|=MfP9Oe2-xViSpyt06NG;yazqB~OSXJLEF6?gInAh+~7)Gp^G=~3L(g&12
z<tFdKn`Fk(dJ+a_RpR)t$ELt44=%%W_gi+5hRy7wfYwE`h8}zAeGh57LRaNk6<;}D
zV}^Da?YaIi+|p{|6)wRg^2O~70pSC6yE^v-T&C2rZ_(=3zm^_Sr^L4wPd+?|%fu+A
z8JLhd;x~d##ai(EdSqCbHsgU{O%mGyQ-^Bmiy&TMpo4$>7jDJ9jQ=<kxFVoc!HyL@
zywlz8bql@=??vtj;age<Up;`m0=J>zQ+NaVPO~-sbTZ+~-UkO^2$548C6DI#TCJ*<
zZyafXJRE=!9rSsATU!h8eNeu~_m=o7q#yHr`0V=_xwSmv6aP9(75DZsPc<sXpO2MI
z9-bi?(JGhLAe0BblOrV{`mS+O14g>r7Vm=}20zXr*rG|(&mgG&>u|DXhd>q|i;grX
zY>l8@#XGcu@HpU|PWYjxUuWXSF2}dQ9wI-FYdBc{c6xzg6Tz;SS(;gsYe{2r*Y57s
z6IVNi9$Dn2jjK4cc%v+sTq0SOT;cQ&bXVn;?((vu^U*R2Ap+t2nt6#?w&8Ij@xykA
zy!o-?(nBx&rm9W#o0E66XY5mg9pE+U^Ur%Q7o~U3r`&RGy66IW<bdyDPJdaA+wvQR
z9Z@gkF2CU*_!EGnr%gTHY%gs&j+SDs>~%*C^zGypZrX%WF2reFB>tr)k4@?Ck3J|K
zExJP#ej^H8D?SkKYYkq6Q>aGE5W9b4FCpZawyc-EHO-YctrTV3<ecLSkH@icdp;v;
zx+|t9+GMg+p1g`;jqg(CLXufLX>4l#Lt<LPdJ?^yypVKhc8Pu>bi((<e~+t^=v40v
z_SRS*O?XR4Spc~Jdcyt5ejm8de`@<F%s&PDl5I`u1jkdtY`Mf=&%q-;xk?%oolt|c
z6ew+qG=hgjkA+S6g1Q(>lySfPc8{Wb&W=lCtZMfPX(^1lICMtCgt%2;g`y@AD@gQ+
z^olfobdhU=YY*rY#3_wq>QkAW4lhL}e>%;!ZN8sv&vef9q<{fqrKDJTUZG}}kILBX
zbZp*@vPqVrwb`e1V;I)ZAc=>)^3rX6+%4yhVzyI`eH{#FTe_{+w1^TZoqzlRd`Wx@
zt$PhNT@R(~N0ys{(F9K-b0}7=Y2AwHtcj>zLhp3ar^I<T^^Mr#hQK!@l_b<B&Txl8
z1W;j}O%_?y(Irjm$c0f2QNFr1KV*e_n_m4J-cmgW0&4!I0A$_0Q*ZfU#}kr=MR)|f
z;WPLvLtt2-1UAWkq<+5*mRBfk+AVlJ{b%w&m<!Go3*TU=Pu?(4KkydHgsU^#bMB?+
z^x2K?oqM^=I!R~llJ$bMr?1P{ZTy9N{(VBiL@v~v+2FB>1t&Bh`W(!rL@gZuzJK6a
zX)+51)?=28DOL0BTmLe>Z$EosYf>-_8_sFDpp#pls0?%fsSq7T$C0fw89f}0KmZu|
z8bum7?M28uCB~Jaqnj~|r?7-sk)>F=-UO;q+_CZqC>qj$E`w1nHdqnXc4?^o)*!+r
zj$MmQCabbj)l11M#9IKA5ltXyh<?R{G0A2~CIRHLt}@2E)fzo<^GvT{eQWQow#nK`
zOf_xq$%R#w0tI3yUmh}@wJy;cGX7kWbWEb}arS{TFsJmr>R<GQZJM@SO~VyfsKK_7
z?xBo(_r1s!bc@(I`2?NF?J*x{!na-FEFLxVG8WG8%kT9F$i5wb-SEL9iK!~P87z;}
ziQM^&P<Q+Y5)z@Ub9(wD`18I?w56*WihC_zwD50}H-%mvG7X^yyw5lIlkJhi@1@&4
zpvgd3Ib=e4c{fA*+jsUb7jp$+O0saoHj0(0Ah9UGl#k6B%bL6+$rtz7f_dG(@SUyn
z46_kM2FP)u9IYuoIIfQV999*ki<C|pz9)aS!@QxhSvtYK%2C>&?a|EsAM;jq*_%-N
zquRxX=n)6H4}LddP#o%k%KrNSo?W66dc{WVd5&zTYy^ol$BpxS2<C{pxCS}*vikS~
z=OkG>goJhT(}2f$Tn_SxR$O;(G`w&m%YW26O%_vrb(6aT?TexHt5gnTn<})em^?PK
zSAQ5#He<AqhYrtZkQ<X;iFX|SSW>_%%gjn6bMV~XZcnqekM&yqY9q-rVi`iVfc?u-
zlz}TjEr)`mZbh9HkS`|cq@^Q?E~!J{PEQ2=#9%tHbYRHCkAVaSMF7X+5-Qv=YMwC)
zf*R&ag2=cT9MILjsLe1eCR*WS+&K!)7CG}_0d=d`p(tvW6Y`Nw%p7V4kztATM(Ysq
zVEX*bJHdNtORwB|#nN5<*soT9zAa_}K&|UFSuQ@io|CrK{cKHHGHq*>XT+Hv2Lk<8
zrbFzex*+sro>x!9dZ2m-wE$f?>K&FW%bFBPp6QYLj&FOP<@Qtj6$`Wh#sIS;?}yrR
zJr0Cc=Y6u>;7q;o=gyVq1U{cFAxN$%moFwF*m{OeMW7A8=UVVRof>pa6~=pqGiqKd
z$kR^S&--R%K_x8Iv$g=@g)_|(Oe>IM=y~+o*l%C+a+2?fyx5B|WB6K#s}0?pE0SJp
zUXA^G_KU75&^@02-G^88GE#BI#UWcapI2<S{&@FQRX_JS->$7S7~d}|gNP+1R!rt3
z)Is*QCLYCd!Dn*~vJPk&GKD|V|EGP|qQmD${6YmLmHU6QL*IJHgKa~uNF<;Q1LoJ~
zZ#f)*!1K&nSfCdxknBL~+#AMLD8l}L^zVIrmwI>?Xw7ND8ot(O#g~&9kHk2FuKd&T
z@<r}ib0l1Y<TmY(MwJCLOJCY>g`OsD&g$_c+y&YB?Hb>q-zIH(ZF*TQ%Ac8(i%x~7
zaE?jsgCC|+amHnb;NxN`;{v1tTme&>JWKCYp;l}wUj_(`woV%u+@vHFMSH$wTPwv|
zcxC=r|FgZ}*T#Zd<o)+{Mz5wRR2s~SWkU9Stai?9^+!^ejey4>4%2<~+dSnv!(L#V
zFsT2=8TZ?F^^tvYc|<(4_vR6EaQE8F8!a7kb@@1tZ*2KEKBs+b89grC9k7fZAIx!e
zelAC{KEBKb$}@Klue%-g2Cu$<o*8d5O5W20W-l}~5Ro<yELiU3JU24xzkT<JNwR1A
zbHg#U+eZ!(=5B9qykG{Ao*ut~v`%ktL=k4+n@49ehI?1t?cZE!xYy^{@?&voy4z7W
zLhx^IXe96)(V_*CO2H|47D^FEg-A+~MhnQO1#V{Za#RbvGno*?Ml8U?3k6ty76&3f
za0=4CMQJ6V{lxs#7(gC?qSdel>unOI8a+oHG*^T{;SH+t=~|5yK}7@9lGe|ZN{^+l
z@K?_3u?h}YH`vVviV_4dw7wNk_M}uUMIKjzehTzURRH6?-j)Ie`M`a0od-}iIU!3!
zgN20i=|_S_yQzLxR!YX2SWc@jp1g*pqUI$ng=_*dHFV<PLjw~>YE(tMV>j#8naEaw
z(Wz(b6M5_rDh-vcoI0Z}p1j;02tv#q9YvY~IeH3;(!}%9DA=M-dy`;IMW(8X1_iou
z3VM>tFO$ty_HdgzaJotaS*kKWdSX3h<2q87f`+P-P`r4qP-l0wuhm%T{H`(}XcdLG
z5+Jd+JRrLPO>IRDb;458XUrLIlkcwXnw(VW^>d9hl=(4bpt8sCQvB3C0Cb)yZ5@O>
zVL2EYf+(e2q?aR207t^^s-#(^|D_J6E-PxBQQ9)hbY8;RT16ivr3k}mLB~M`pyN!?
z*53Hh%hn+n-6r-bz+X4-L-5mkk?=Vz&Z9ifQ$IG@j4_JT;By?3I^7(0Lmk*q@+7g?
zt~TPTGKC5}UxU><pbRP9c~&{8%L6|fptk%M)7Oj*X~FVR8#NsG#JPCwc`^VL>E^!j
zAyImg-gf7}x9(V>jd)gvwp*~ME^E$%P^%IY{)1R+bjsnx>(Hr1=8&qGJ$c4cXS?TU
zUYW{EOZM(o`!J;=gUYa?i(I~oit?HI<X)K<OhvUUNZrz#N#!gt0GKmW(tM1~ly+#(
z!kwyUKMbd$@!gc5P<caH{Yl|d)tJ5`GD%o*D&92GOjGEVD##St>pigy4J9BbWnSIl
z6qnCCB471rz9<Km-KB=#3ahUd?Lalzr@?^y?`Ser1*=G%AffkK>|ex90ovqdce~b<
zE#>7nAERKB+&}2RCeBW1$z+k`ut{>2O6P^r9NkOKVqZK;u8ESU>ztAUMRXg4pAG=o
z<YfhP7zGen^3Q*~37k*qg0x}MrYesD*_{hwi&q=px7zQNr3RXdgtN^kwKC=B8?y$I
zv3I#m0lL@`<;t};mow|eLqu;9OX3Bz`Gde1^5VyeBQLM2(me4Ns~<R2MV9FDQBvM%
z@Fi>M#{)ILOsa}f@)IGnt>wwVdTgPn^t`HNu`N2HAI~KhS*`h(kBvHxfg{?5tsPy}
z70dJA5(4mdO-UaE8e*Rhn;yTAH^>ZA<h&xwQ3)K$T?-dtd8Eu529Grt7NI61DPu%<
zC8!D=G&b!;hLUEQ%3!8DC9}5!(d1&)B&|g;Kgpag%EAmJUzBHqvF<xxm7Pg@;b`;b
z3IR2~(Uqr)_Xy7iBMNtyd5Zne6tO;>;G4_s8F<V6v|}UJOmAAJ2x^Yyfn19VQ)qIX
za~!qbl7B;m-yRPwu6G|-Ip{ZoeT#>jR7D<MPb^;sPGY=uKZ8J@Keqn-eCKk`MCWAR
zbg{?o2ACRWUY?urF`_n254Z^TU2%=`dqyKCRHa2vRJq8rOSb~*ZRM;_)gns^NvP%<
zCTmMCc8_wB&$z4?>&<2>SHwq9)TC?eQ)k)=No|Vd<WN3URZFki&WrVM-Sv+QCH~J0
zzNfM$o>}i<DymtSUB{kS!wNnU1yX0VMc>lmw_WYY`&04iX?!$SISU#<-nq2!>P>VN
zGh)wSe)(^)qdCuXyT2jqbm&*o{<eiXXB2=`Cy&sL*pfAN)HkEya5JWCscd3&-m#7O
zUFdYplI^k9TLxb2Bcf@OJaeaL*feV-y+4R=MpsaFtT1b^+w6CP;%!WrJ&9i!Siu$N
z#siA(O0J`hpy;Uc+-+O!7OquHrx89AWdB;V(tN9~Z8#^gC60+-$Y)=|9vZ#fDw4*F
z8<DZ%)5o%ifeLBfqP4PNt>L3Gg-_FEo~sNp|2{KdT~&IH<zbX%$uVL<S;JTT6^#Pw
zx}!iHm2~fkmq>~=;wwD4H(I=PN{Uz&#!q!TrToVy)exIPg%|Xt$f=Se{G1D>T=G60
zNbR1)cFNG}tJ8ZCmCyXUi&Qdt%*ppO2_E?rn)rlWP(*OL3O;=NvOhUKNhiow;=w9i
z#+9igkW2M?kWW+fN=u5fC<b+S=F?bC(s*_@(Juy*fwyr#f+{xMsg~HgN#$7SP$xOO
zif#c{;#r*Clcy?UH?iDFvh%8wz$-)Y_Ie{2;P143JW9uMe&Ev*+v($VcLH!<&I2e_
zL}EHAaUMPv!Q*27eRK^#xF5qsS~3SAdp~eAfH2uSJzpmNZp$^`T4?{UbUogoppsSR
zg(Y{E6T&6@9PO&O1pMk%5ufI0#&zlJJifwWbrCJp-k&O>YiY40$<x?nbh3I?-sh~X
zC?#Dq{rRcw7`9pF)PI*9k=QKQnmQ3`-aJb$vs};=-gjgLuR0Brq%y}*TSZK#DRJ%A
zmZ7$|Dr#8gh^^A%_x4p7(eTk#S|tTg7O17Zs#Lavv-++GNo;drs>u7+)-<F&S$fAq
zt?Fjm(zN^RE0^w`Rmv!4bN|c+fPCxJpmI1B`?4OoCSTJiOHxYjjTvosE#e+G$YlJ6
zE^}|at8Ck%^fqeg+a9M{m<ww|z9*8j97`P_wv=4Zd3DtK{;Zm(&))gc1wcq_-U;0K
z(%_S5Gvkwdhu#a!$pRppV-H|Dca-kC9fxJd^K_^Oh(4F^B=4T@Bozp17f5SKoaaNo
zzYmj(Z1A3XQ&khb#O@{MSnnn8eA#PBKE-=>_!57dr(Fpjwqw*1d~vF8gxBu_tE7)!
zHKIu0maC+v6&DE~KkI`Me1h8TFks)UjnYWTjIX`~<0PL5rs+kM4bpeMq(7Abn#2$M
zaVCjx;S`ClR^xOM$v{embM}a8$xElU@7!@*iRFV=UustpZ+cgf9{~cfIfSE_gKxbn
z@i%e@u{&Q{*Mrt7z6<%D+}n5`(P6PKHe;_MU&?*rZ$V7_a;f3tkvsTa;gFIi&wI}f
zBP@~*_}7%E8MC(lH22!*F3jayrtFQz;q&BU_i-xp5%O`qUD`Gycj?uHiUh3orml*H
zjwMYJbXC=q@zcdEk)G6lAC@z%S7!^YR~>Z364{lqw<_qPR2I0kR}$SmP2zLYzI<{$
z7Kd|K<@`0(bQE_f=t}fS)RDD0A_)#%i-*Om54at3SnY$Fm%+_Ps_07pmJ;=@>uB9z
zDO7m2Gf9?J0AQ$62UNDsaNKenwP+M7&?FRbFhJ%|qDv`mpH5UXRW!8$yGH!hP!sS<
zPbv$)9t$WX3ikCIh}!c?N*d~#ek~<NP{Eu7x>1|Knn82-=9u&pWSNM)`E!)iwK_3I
z@o{At%b!$~RnZnvJx*CfitFm<+?|_s0s<E<{Yms3_0_cz`;6?(ttE`LX$cf{&NRcS
z<f;>pN~$aUZ;iMl9NJSaGzx$f6cZMVqe&V}B#oQ3_Q81noha}xwtiNokd5{=DF@9o
zCrEO3%<LikuDa&oUy;*N)dqs`I1Lm@cKB<Ypry!W>)%?JqDA;lT|)z+PmE1fRYgHX
z`d)@fBjL`V-PZ*W(dhtDj-UX>6_k@#hp{S?uSKd+a+dV-kgcJlq@Yeba`#pOVd0m8
z?zn7h06r+tpsTHV(s5jOxJ;n3wFja}sv=tav~ll4nn82xDr;&>g47UDrwzk%(whq?
zAnI^Tb`lnVsQ{|&Xb7+fr%^hnrp(qAm`$raqv94OY0CReR!%7Ua$DZ+LCRDN3L_VQ
znxyRsyv^*x&1nkqKm|w{E!G~?B3l1>lPF1*h7zSAzY}hTYVma9IBZ$H&Fqbx<)d*n
z^FwEitiRzXGs>)4NHOim#!Kn%yYeKAmN<8S^2Z7QbMS**YzqOyqhQL)c{*VCfUbj(
z4%f7Fry`k>B$NdJDHtn5VQT^Q->?N|N!NHSRrv#<He}^+qSC9Ikv0uQb<w7bhG>Rz
zemdeM5T_x&x{5qI{XVZ1j1zs4VqxLOT;vKHDgQoTly2i8(DHiByOZ!$_q(F-!uyhv
zW0TSJ#DlDja789D<EVtfdGfyJ6B2g|ldYBT#Cs!*_{^MaHtigC4Vmd2y4dLnykrtn
z>G842G!}`BG-^|s(J{b;pJn<gaU)qQU;>Xg*6+B1{Af%)9hvD~Z>}r+L4I2)9W!y%
z2${)2EGD@2l$fcJ>Ar+qqLBH%L0^WSPE{t=anA9cNla!+ayr?O?D$AJD+$OkD!{yv
zxl!!6p9&M1Ne`J}b~K!0-jR7EYEmj7^lrx>At^PJm&#Dk<LZS3lPnWbb1L~X&gdV~
zQ5<tgTqt?;-)TUCn9<Q+!1$ziF8?snS=3k#JE0I3kG;cF*+bVv`08C4?26=DSQRWP
z+X`T`>R|#(;;!rg>hwol{Ub6?|ICGRj4&Qz*vQEApv07PEYX2U)Fc6ZhLQOPS%rrS
z)dV4FB=009Gm(>I|7q0rYyU%vL_%@IWQJ4^<n(4+JOXpm=Z{tDL?Rm_TP;X(d3qvK
z!okDFghUW=p(8nPM^G-Zm?Bgf>;uWf_|vYGn{bV{E~g%f`&w(CJm#r7!!AhxxK><8
zG}tMiEq_K+S;}a=bfdz&$V90*TSCb~nfMnGcL^|xYPV?TP6~fx*|L1KX6XRg6}tp$
z_kx$7Qy{s*U)xzCdbNL829i{j#gYcnXsJtbe<9G-bjshj+e`9b(E>}u&}QZJs9^v*
z^%?Lios>xjt|7{c#W$mSXLA}1WDWwjTGe#jv&SKo-Zl|e{8t-5>c#WyWiP6pmKFb`
zlIiYJXeC1G7Y9@2r6s8&s4K_4HA@eut8gz6P%RE$o$d+%8s!ZhJ`FSyEPHRMLZ!KH
z&xpO!I3(rvaUYI*1m(=|3BD#cB;I`RcbKn&oQl!grdq`Om#}I@zN9J%pY;h3y@=Ej
zRQO-k^$A5(7riLd61{p=B+kQ<chgZnVadL4IxM*eKO*@c7mr#(+n~7SY@JpM*cV=O
zFBLkKM7NFPTWU990Z!Y)=A&4Mth_+sD^-Z>6df*|M0b?r`y)=GgVQL%S7!7lEE*+z
zr^iVEq;i?`H-rOR-d1u^aV4_3zEul|$MA8-z9pSVoTFn7{=>T*6?5QA%1P9}_E(-0
z82=Ok1phS`(yIs`!RsLR_lr2o%_$z&$G%_565p0kAUGvv#)0plEx4lDmJK04^kg>b
ztnOpN>2;;DUG;lyFy9w?2xX)cdP+%BKz#>cdqok|kSU|@621>nk`Hww(g-0~Jv*BE
zP9ua%AGrGlh_=5i<0w$>Y>(xv6V7R)TFG<v3BnB~7;^)A!d5hY{taeK!47kj4peb2
z;RAD;59diK?s^=eN%OpI(X_!2VP;kq`A@;2-L`OVJZ~hNTI`j-CAi0=A3Op4c|H8!
zP|E{&m|W`J40tID_?mbw8ZVC>toD8bT5pA-T&$wf+yf~6WuY;z4pbN?UGC3=6N;3p
zuzXd9PvM-WL#ub^s)D*|S_!Yp{uYvYRg|aBal@DEl*ZJNy40FFbkrx!;O@M-?`Mdo
zuKq7={b|YzHu($n;Li^Az1|;zi*V~FJJSr3FKMiGl6qZjecA2n9a-v&a_Wmz)s@7X
z!=nBixfObpCtK+E_I{s|x)Ic;dFqRLS-}@WkqdIdixJJQUK_wYR|ON~e$Y1VLZ&ra
zduQ9K5nK?A1#o{to+zLgQZTNh0C@ISq`{CNM~WE@BwCS}bX-Nf`~|R7ctiq0ci{Kf
zyZ+55ms>CQaeDf3I<w;zv-&IjbaSP@7PA&yo5P<t>X7kNulB6HxuGDfyY{r5{#S!`
zxv^J+snrO_kjABNtAFWmh#ZPTr;0sUSJUN*qz-^%BPepKH({{vH=Jv_QTyV~O8E{L
zr&f0D-A6-zC)PPs&Jo}*`L8KQr@w+RRKCggrFWro?nOGBaAGc|2jV=T2+3o+IHB+!
z4rNh+!EHDX6gXkxgj4>I@79f<0N%rAlHTN-%Dc$B#ydq@AL)Cci0&oRTK(8yo{-HJ
zAJ-6VG#A$U?mKPY9_1eS9u>}{TCE->m#eTb8Dp_1-2BE(I*U&I%w((ahQ)rB0W_|U
z6o($UKdtCSjdNQ|6qjIYnprK2{!OfNXq>s!nHSe9MP7fxRH&^cLzY*~im|Czf9rR{
zRom)i2Rj2aIzyW&mq^rbZ^|1aUxi=QUd3PK)Z2;Tk^vEZ2uX*4caAN~L+wM{V$#WV
zI$O#~Kqde?WErlOay-Rll*!l}LJG~)C|ut@L=~=<<-Xx=t^iA2X&^EL9d;T!M=?h^
zM;>h|cC;TymRj~uEzZ^0Kf!e@bn3&sW+Ikh+So=>;izR!=|Dm;fm}OB4l9`>Wm?{3
z8Zc^QoMPO_#9hN};--{t?(XJYau;@ImlEn_A}u=(JK&U=mSeKE&?L<)oxL}<z@o^F
zH^J22mpUeaMdm)Y!)$IKP3A5gb?c|!@1{#ef_d%OtG}_578S|-XE<%vKZ-G`C8rRi
zU)~-nS)SOCiE3&TgE&OYSU6%={%TyYcH`VJ*1?r>;0$1-#;;vX?`*u_%-L+CWZ`Y0
znU&uj2K$JK#n}!i1nxTe+qmu#f1^(u5j>fxu0#hM%Ob&lVL>tbc#7MVqLmzTW<3!|
z{Y@3g=v37hU2c<JDp_nBPyW>M<cCpoqV%?LlrX6vpOX^dcjs1~R7$yRYCEh!j$BN$
zNnku#WIgW$(n{6MqtK!Wh-<1h^2;q<C=|Y>c}^0}aibx;RoKgxp%JJW3SX06_&p|X
zB}OeHzp>XQzcJG;(yrbvaY|bApfD&RIKmiLTB`i0<k?-Ia4<_OIonh4ai=V@$ToZ~
z@4ep>@eRaAZ|3md5(3#6{tF@SCrth)mhk^T2=q=Kms#h>5PkNEI0c46BYfI`rHDwf
zOcwD+is*oQ3hI!&7<EB&e~&pP0ab1i7x2o#{66+l2#KVR9p9p;)zOyK9@!gXUF@0C
zvc0ozj-;b2QEu7pXy|sObA2^$Zby!FPNvnSrX!C$EudMW?sb<p$@IRG>EzvJaj_L~
z-<q35Or5?my*`h9-QV-HJ@cv6c^X$<eHdOOyB^xiKQ7pOnf%@)|0sCzk`GGlYU`k^
zxRW7Sw#RvgXk^M&T9>H}e-jg)<rAolWU7tL+G{S1L881PvW%ttm}=?y8PeU&G}&}g
zEGx5MgncFITCKhPPOB~W63XuEpx7QfJPz|2cj#TW)(R!wGh;_-*L-Jt_qkHlB!A0(
zCwi_F@q7`3IwqbD%|>QA0NtnpXg6va*XShM(qX+o@8x4md`>es7=JB$Oq4IP&f2nn
zrf+4XH0V}xc8$KE57j3P#L-B^ixwrZkZs116=}|8u|QFVODHM`X2dIrQ<0e~2nLAs
zCHyPUlTb+bxdiq!#YjSqx(vAtbpdJt`pifYEBJsK(p)A$ae#7n!A_E$H48}wt~BU>
zs6NI!iee?2>CAT(xc|M6IAWco{$Tk53<l=*Pw$l8FuzcK!2XviabUbjj&aWabd<<S
z<z}2dv1jwrdRx3TJ(Qo$&z0u?I75CMSU(Q)6Xs&sK;4MkKKh~Qozy+k8>w67m*N-9
z4~3tnQTl-QqVWA^vn}W<YzpFaR&Z<75pLu~Ptq)be4!!z(@ME76Bl3m;n>HA5mHXn
zK0@gVx=ohdfK#fX>v)fD@M=%ZCl5x=p540hvzC0v^SX!XWryM}a?R`fcP1UrL?NFV
zW8=wx(>Ito{xJXVzM<r9XG||^V4>(_O)o>h$iVQ^Hyj+De$Ln#>7|WrOr6XKnAupE
z{;%-Bmzvg&*zNIOSGtVKuMM7*lRMSe|CBAxtt>B<0WsbQ-K#K>_!Jgamc)rv@iH#o
zpFEow@Y}|cDc4yQVIp-9^S`j7eLAqy;vdt?n35mUX(j1ViUAG_9cu1{&a<k0DV<uu
z{E|Zv-sV!u(lX^*#)(bA<ql!7K@G}N0Q5dX0Ko(Sv88oQRpT1>7*DAYECnS+_%4Mh
z=`7J(Ka4+eB?*%YKpkRWhl2uS*omxyfkSDkgh;h23x-Oj01#Znl;>pS90nRPSo}ws
z?szV+vg@Y-njkTC7Nf`&kjW!X$8!LHuxv&n0i$5XpqOJW*J1XT24uG1jg?C}R0YK2
z{j;pxp`|?K2hYYPfx-(Ar&adXQb3wq(PtSDtVbLw;5nW{#Tt^sGL9>Mz^aIo^n}D>
zQYZBmAqx@{1dpp{NQ<bVhOH-B1Rm9jf97O{0H>!8)Tp5d!=omGJ-Z1MdA<YDi~vH>
z^f+h`b9`E(>>#7#biiOWxXUpvvXUfKwd#UD({q?Q5&xZ|O@(X3C`?Rp5icYM?lhI<
z#KeFR)57pV1yFz-BM=IS*Sk3phqy0~;7x{iKn-t3y&z?>w60TDzy?3%H(4YnODQvh
zKa*$=gULGBQ4WuT8rHP%kB>_4H!`lSS{{l={ig&&%MK;d{;!l({YaWg<O3CLqp2o}
z!ml~MiqSZK+#%v3ZxI#)i3<smo!%`{VhHJ@O^fEQ^;U)g{jkI;fB?8Y&3RZpdTmwA
z_f7Hs==r?GojYCncyhJ$zFn)kE$gG(B1hVmZY|T_Uzc&<sPw4FowM$6uL*Y=BBo5n
zAZwx-nE^}{W|U}|EL8u#P4L?EXR0;R+)>NhRG%wFsKrh;OK&>n^&S6)To4Cg5YhO-
zXYJ#T&$p65FNo(aueLT^a-l_zOl=99>2jex3>T>GVi(@z@#Q4wQh6%h*B@V>KlqdD
zfRqT+`7s=CzC;n0F|*_8!3AMw4?-C7@qF|OA99<Ahfw^CWw!ZSbuN6C+SpuH+qLZ@
zk(h#-<-%uL(2R-oF|QIE26Oz<ORKk2lZ`3PWeR8d^~mwr#8cY}@!^V*i;$CDaO>d7
z$|hqllk8X(L_D~N%#?N%woUmeCirD^+h-4bpJeSS)jX^%VLxSZie(7h$N?j0=Q^7g
zGwaXH5%1^y*F&d(WG_z-qU~$miR0zQZ>x;w)O&>7sq@<h!~65n*kv*FOTLWM!q-sM
z<$L6^qa)m<im%6t_N+<X=0=mbFXZ!2Gm&Tr=qtsFL|tgSf=ZwY&vDxRC8CR$0W%-w
z;Fr#W4q`07wu~L(FvV!doyKagDilRH)%M^J^~$+VF>Bg&XAHXweLpohDOk;se)twO
zDM*o*GDr5PMZaWt1JbExjC<8z`_KnZ#P?Hr7aqDGhSJX<h8keR4(ad#@g5P^V4~+*
ze`hNl0uOjVkHgi(<9|thKmK#`Vdb$M4x>N*Ll)z&;W2W2t>7hA&0(8PKWrns?}|6<
zG&^B4Jo9W-1D5f<3zlApo{s;E0SyugGM6`EmT2Vna}eZFVm0Lu?I3878Bus9$|`}W
zoz%vLF66LYLV<}cGm;tv3VGfNW~{M<rl8R^;&Z$UA@L!~UjQ~E%@)!?3wsJ=;57f;
z9X;|@!=|Vj;8{owz7q00Q0oja<tK2axXz>J#~=!jAfNGvCfCN7`gZ0uX2t5TXW|39
zWDgk*D0!xp3K86lZVAFSPp_+E?;$I!_*rd9ai-VduK;xpC|`3gN=t2W8AYe`vsHO!
z#{48Kc>mfo>{P;@99kUH8&{7!4>hTa{CKOFPLK}LxnJpDUK#Tl$RMBPt_8(*&RfMv
z!mqA(mOv2vTVw}Wc`r_9$Q{I*NVBc#x@Flu_c>pvHxctdu}#f0HxO5RKh-_{A7&p~
z*kfsNIm81|)C$z@R1pw8T5Ua~MvQHa&A#F$c9EbE#z9ax3D*4#0nZV_cwn4xWL)}w
z2|PmjYXsY~_j1W#NxZF&OMMD!OnveSJIDC}GJ(;$BXZir$X!hIsO%5r2aJCPn?QAF
zU}#{+awwjp&sLvJoiqHOV4;=KkhNDgNK3i0Lm@TV26m30EH73~&2(=Vw`1c58_)DA
z4!!h$qS)g`mzN>k!?PpJ2j^jbFTeBkx3wkALla(dBqJbJ8#Z~LAkoG$#<oLBDvsVh
zJHN4Ft*oiqv0|wRJM&c3l8)bwKHXVQhAd*%eYIlN=Z|GWl0{E_DyHd<2QJU!AmZId
zSs-&l$KW7uS?3Ar@sHT`E`I+;0-t@!cUHNkGwhsrY|D<erK>%j$L{_ZKQUbYaC$oX
zUyPkoa3xTXhVPBdiS1;<iEU1-iEZ1unb@{%V`AIM#I|i`r?$3gA9kyDA5K4Z^~*VR
z{?@mDu|lNR!JFlaFI&mw$=umyo~;wLIbEJ>EpIFpvyU?smk{8dOa-;*75KQKo`Q)r
z`FHj~{^H;b+36X_)|kANf<8Ur<@nU`-Tn5nyh_H2-A>?>?rh_IjtwRc;r{%#d;#6Q
z<ZOwTk1M#jsgIt%BuAF$v~Y-3KyK)7#Q5mj;XW9d-jtufYncagdec~&SF+ex1ZptU
zUu%Wz0Lh>89cf)z+WGyzLQ2+xLsA`e3ZExJ&njm4qT*UVo4h0&QlG9^N>7Y$ZNo|A
z*Pc=uUM8;=++rAuk1n1jX-83+7Eap-wYx)DmzLBnBox~e$MF;dg>{nVD!n<=(%lkG
z&y1qL*9XuLc@vh}j$4OJy?d7P$Cuu_{p#`AMDhb=om!mE1Yucd4@lmx1sn5b*7)&b
z2}LYg?Gkd8lZQQ(idEX@szlIIYLTQ|I*h3ve@c-1_Js$wMyJ|DDte?49iZ?Ti<;m_
z`q3zUu&Kdr%VSeI>Y|5$*XS??o?dbnBpAROX&!{b8%;W*)n%qTWM%U|(frks(`=@y
z_&Xho*?#IT4Zh23Oz3_T5(kB|$r6s<)PKICPv%X7yqpT|eH>$#^Zsye2VXf%26+6t
zaBNzF^{|V6AgPfptcmrjqZHmscfQ6`@y~EXaS_&o>qdm<>UBQ9cOKRT$25WOe96Bv
z-2OJ(*|u?8n4OY6{}yszND^vBQ-pnjmH~FjNMqogH7#6&wyV~m2;ekSyJUO92SzHD
zn_d1y8{SB2UoFEr`Kh!e*BP?)AQg8%PB*u%TbvmFweqy~;L@H#(d&m~&|kEctGeiN
zDWAPi(;AC-ML6h?=-BB8z!QpGN@^c2vwd~8pkUzSv_Yh*BLx%cQS~4B_kBdM=@R*-
z;T_IFdoWkP(4mtb85f5rylXtzQIFrJ0?RM2+Xch+EvMcxG1iU@<P~zAHt09LDDoc0
z&H_{oTd3;~Hm)VLWo=F_Xams<Quj)m`%F?FyLap!c=?=Y!}8;s{`)H$)Cwu1`?!cr
zBpIU&fy&^pJLqr_#w0t<B9wG@V=V;jOB8=tJtyvr;GN)F#yz=*2a>bFPyvBz_TxgR
zs2E=UtW`AQIx=N5CL!J9C@QmfZhJq+7v`+BF}&7{j!lnrs#aO$l@t;kKQ_xd{=peb
z<60ScLDal%Lo_213Ko}hs0D19SilrS8<T*F=gkhyr5NwdYH6B{bX^xwJ>IBKI2g7y
zlEu~)YW|uKpAqJip4N5--Ef1Sq%8T3*vj8x2W^{*rcNTI%%;ej&VNR<`>ZGXR9bS!
zdnX*PhI2_9rW_1X3W;6sMkFj--WqlR8xot8QJ{eKqis&#j^ZLye3$d=km{Y1z2Xw5
z;c{uh<Tzm^W`!N1N5dfI4{mI24t|~!+uLJk3xzZ<u_2e=#ELR-$(B&hc{arYYBZl)
z9Jxk0f0Px75J~k~83*ea2SzK_Tw~ON%_!dMa28PXT7J`pniwA+i7$E8K{ZS1-x%03
z%iIrtup#cp|J2cW^ZDGaPB#8Ev)xrIkyI~jECjJ%NrC{5ybsAqN*F(jJLX9gRvlr1
zt@A=M+qK``htr@-QISbwP7$w9)XzjFG-aEL?%RiA;IW;6&$0}hAnG;%LH<?2F=cP$
z>MtuMalm&U3&9t7f4;x?n?k#`ujOcFM~<G1-*Xz_;lSn={pMuWz~VOL>vxJjCw_{*
zn?fjfo^ceto_7B4Oz=$+KYn+k^+57S<AwH}+4ms6TrL0G+~48~$)ed_C?w%*En>Ea
zovAm^$Z7CNi$IW}SL?Cb^;VT6qANMxK|J?|?h#n}`SQGwq1OF=)AHeRe^=5lYy9ZQ
zop#Bk@l9boET5b=PBzA3hBh{up+<0ZU}MdInXi~;fbj1@(MGMm#jRfGiI!<A{DK2k
z`HP?mPpdnNkFVpqqw|KeQ!L$@K9%P0taXv!8*P^;V`Q_3J5aA*I-l*lBmL@Dm~lP9
z*_tg9((o}BURzSlPx+1J5h~t%kM|d`>WtLUZ*<>IW)07=m-c?id|;2-tmi$PbrAN(
zpFD?*J~-Wv_Z)eh*qjz*$?`uLt1pUg(zeTbkTsj?LaEw$GQT{)zzgsmkIp1)K9E0r
zsBgkd+pg5<MKT%_H3w*ry!rla`d27FUsk0XS+*K1qOe1XSe=$AT=MgMJ^K5#c^%{R
zTd~cE=_gNs;_gRX=!Q6*;agCh`O~{=VBX~y`eHvy2sE6C*L(+t|1u*PN(QvgX;4%s
zzJ~#0^u<`unS-L>^`IJ>46ID+oLgED7z0$L1mqK~PqG~Eqm0z!&7gh$5*4W&vPqDf
zuC1ZgT#Dm)yaMBq2|rxmPuLvgaX-yb67eCJ<~=Fnd;2B&6Y30;mDurSBJbNC9f{zl
zdl5lpx%Z#U8dGPLFfuC-tY@&g*UU0yVZWw=_H%-i)R+;96@GnURWR9y`+(DkERw;E
z&{DZY*zO-wsI7SL&F5aG$J6?11%d^Al!ZbV<1Hdm;}AiQp5h1K(<P#L$vOK0=5xAz
z9$$?YpzNf0>9-KilCQGeZx2o5yPa8D&4PJm2;FQb3OtA#um0Y%udBJHyo1`uzaSwA
zf?NI<CCtji@jqEOY(M^=?25fFWZ>05Y=s->A2tF4TLFNZBGc2yBnrSF7^7R+1%2-m
z3S@x%9{joufdOFGMfCpPBVhY)E&g9}o0)}+gXKSi_&@Sg=N%r}i3^{czEd8LW{*i(
zIN16SDk!32vQ)}WF%zAv>-MtrEXw!0@%gQ}H0i;@Mcc_FY5rvz`KFfRs%M@}Rurwo
zG)q<%VAI}PM;=h~mz`VFM^8srZ<fIvS1i*LnRonGA05*kPzHgqgp(o&w|oyWN@Fva
zHH{$J5&%?`ufcv}Ki^(%m=Ovb9KpxfxGgIQDM6k-k+yPuag^82;9`x8DD^kpf~-Bt
z>9M)VPIdSiE0j|%{rBZKBZNmPc~aKkt}RnuZKiL*V*O`oJ1!H?Sh~JVFXu>pAxn@j
z4?H{`r^j(%Sj;ea$i*c4S?_%#Y9){ocZ{U>PZ8H1Pl<7ikuI_qZwwvP630IrR;7(f
zTu?q?3n3a{9QII6queoebF;>5K;v<^rd=W$1K}vMHwa|)fwgM0-@fz|1N5RtBhBKp
za^e<j5N=8SY)!9W@N80dfEdk4&%5^}R>5B(kVxk71;r&fBdE7x`w`UYJhNc6&4O&0
zZZTS@b7`-q^hR796W$}%O|lKM4T72aQxzv<C__B5G^RQx2OR@ij#{-ZT`pnVGAj-*
zufZ7aJU4qoZWy}w1Y(IMkoQV%81HhfbM8u>8DDB%oZhecRCm;Ny6PGn{BPLy(Du%T
zFoqt6g!kHqs)o?^<cxxDa_i>YC)`)vH=jFSN?SJ>mix~w9(5l-9|ImOL$;6iN9<*s
z=^QJaI<B5W_9TuHrqnVaQ6FLCMaM(U$SB1-aN`Jma2P5MP@z(ho@pbg1y*EK)L7M8
zm7ZxV*)M4?>8cf-MV|dWOFoO1c8GE4KGr>9u=hLOyy8Fp9Qw$5%YPfDrYSd5xhnNl
zjL1*+q{~K;Pb^6*`FH!wn$4=ImKqit*6xyy>W|E(o3Q040$YB!+_sQ82zY4J>_2#V
zdHn;WdqI0hxG#BbJrUj#-hFO&&Q1_94xH^0#rV;GYQIN4w?V|CWXC2l(d1DZ`1O4Z
z%nYOs04L4Zr!7#|46P%|=mWIcG>4k4SC+a5_1=vgwsuP*pt0m>bQLr8I`RfIcE3x}
zd2|%Ijj~V#NDn2^(P?PZw;4->ef^T8ccQwS+Uq8$+B^Gwx;pPtd8YE@uCvrua=KaN
zRdZ>$QPg(k;yYOAo|uy4Ie2AvZN<GSE><X>#A29j3Kfu<V^sX3A#F);iZ;!y_1C(B
zE%~2B+Jk5ZJ*Ot#X=z!<=v|-kOctT3hG(PqQBbB;P=6od4&jIg{|J{_(w*iW)ZSWo
zrW^LZ`o1+Ug5&4<bJI<J>&SZ(!5W2EU}sflj1TCewsXk9^%?(zfRRVn?OU1S<z?n>
zCYC>GCaMGKC~7~MDb)V^FHsY6M-sK#@r%=)F_@7Tk0stE{-qyFf2yx^{^)eK#(Cwx
znLK@Wi)+W<e{XuzdV-8)37;!-oC?*%Z5l&AhkY7wN8|gOvHzZYG!DuvFk&`RTt95-
zCpJ4j==8MX9eJUNy+gT@yvo?*+bkeHqTO6|=huyX8f1+8LfiKi08$(aO`)S7Ql85_
z^Py6-q2*X)uXz80pC{!aJ~EHJ+o9LPa-+Xg&Q#>hbyvAL8?=hpqC=JL`<X8tFWr`M
zPm80%Tx_g6Iv1}>Ax80Jo2Tgvy<q|6DDKkS7TV@YY^Jw??y_<Ax^z*>qx=-pR^H|b
zNn>l0ehWA9;`ZI_%jC=Kv;PzP6Yh)9uk4HOi{VS|%hb>ROZLn2i`%d3%jFBruiCH9
zukg#`OW)7$YwsU_v!T06U@&Lj?j+`k<8Jt+;63hBp?kybLF?VLTer1}Rc#mJYGd4O
ztM%=1qpv;V3Ugh@jpOTbwcYBn%fW47qobYnilMvx{PN5#WnJgp>Y(P$?yCJ`=c4nE
zK#Pv6*Yw4c-Bp*4x7XPP{hzJ2&d<lOwX{DH?{}X9)1P<yQ8yez8`xe(6H&7qP+QpU
zMm<qo9IG4VI#nuLHC<{SS`S&LYYShM)m2@pJ`HWfui`E)4^5{P3;!ApRrRVrB_0M(
z;TQZWKAT^yAIug)$_1)FOFcy%G8WD%?kbjk)l}>L>MZp%cz8d}JiR|{Je@xcJRLdR
zKI_)&cKkRxOR2eP;Xhd2s`hcaII}WYTGZ8YvUzAyUAnEV>S#V+JzGlF)^fgRvhr(X
zf7Co|I;uu&W8Y4HJ1wd>tDfrS{y3el&`^WcM(q5qQ`YTx{Btk~xn}>9_vq-nysg~t
zmrlj4YEzA+T9UTLbDvs5itJQYhxu*a=)bm&R9@rTT&X2WRoPdIIRVNKpSdkcAD)v4
zDU*cFj7F=uLn--T-ix&T^wV@SUov<3>&5N-?Y`~F-i+VyzqjBvQ93Bye_q4)qW^~G
zf#%6^p69ycxm@{3*ZaD=`Fa<T|H9$6WV8=G?IqEK-t|T`hF=24fFZOZw8GZHZ|ecw
ziMDM4KbvZp<b}MDw@ZN8K!B_v6gUD9^d5+V2_W3&0vu<c0*IFkOu?{$rr`SkPM|Kh
zI{`7iz6XF8KnvW20%b<}YllWat|Qsg0!FfyV}Qtz?u4rXV1%%-@E2B~6fnV@OjyA2
zVKhJ;5Dmm%dfn}UDuucR6J!|P03QEFL3X|qWPBpmhXvdcR8#49LTTIeAR)H!gZf#w
zU%^~0dLF>>A!i6zdl(6Cg?)(|sQ`qPK<K-p5R0Bk@PD3pcMrjRHNY<sdl~@z{*Sml
zf&yP$jU*}iBwPFcc42ivzZXHRQ}pOr^?U%HGokRWLqfQ2yG7opBv+8YwS(A4v|E7J
zv>j&9-pJcs?9fJLi0&gsQGy8$z(yg4p(WOB{}A!*KMt^sBp&Qt)xa%;YdFMb_Z}<Y
zq6iO=$+&Dt*!&L!LOt7W?HAS_Ew*i;t|7F4F1n+50IrtatX<FwK${s8vXM=9lvQFZ
zvWY=;Sew|On^Y*?<6_ijV_w9IPmn*+(A^9a0h@5NO(8FMCu_iGMvSl*g8p+P&aDmM
zp+0n0a{$sc5fF-?RvCu~3?B%%zuf}Z!q333=8CvtuOaPS!)*frf*F9;MgsGO{<l7r
zo*pacdm3w%JB$5ehzlQ(5C=Za%qI@h{SL?__!>_?w%Z6*nDn?7+}yD;fc-cT+}stU
z6W${-EWnDA$2c5=8%oS#<_4J2Lw$9X1G+=5=B~-juuxSBiOv|L!i59UBKCj*1mYm;
z!B&t^GV08q76g0do`xAH9B~9_z^Z7fGQ_e2M5T@BOlzn$qUg5Xfc8HtKR}8&&<*Ab
z4{%oJ1;;kVZ~$A)wTcrIF(zpAx65@Lfn=xf|7`sY1PS6i?}ParoI=^;&Mtrh&Sjto
z?19tT`V!0&<7_J6IQu)Ezrb-0k8!L05pEs)bL@dO%-&m&-u4$kkH$U3;8-){0sL)(
zZrXwc2>ksI^gWQ~0qZ)sjMZ8*!PBj_;Gmk=Zg2!gN8rWn7TcdwD86_`a47y(1HKY)
z&kbS;j%Q|qnKNBT!26Oc7=oj-AB@d*K|8*_K;w?nWdQE+FL0>rHvbgvYjE4#VML=-
zK}_4+aYWBb@GSeUQvXZYKtVo*Z^so#=X5`z0grj$coC;n{z%pW<vCE?xh!zz;b1r?
z<p9P7K_WQp<9kr|e88rcASPibUa%~vSHepl1$?;{`n?|7dIR8kCItavD*&D^un_1&
zHIhWTn_B4ud>$qidH@{{`<D73&RvZP|2zby<1jOhMgagq+rbOm>UIrZ3L6t>)7^Km
z=NfX`3KzQvYY3DR4kTa)$wvb9aKXD40Y0D%VTT^4{w;tV6hJov)Wtco#~BJwnQFBz
zkDCfYVqEUR+D)Y)Vj2Kry#5REe*gRGNw|^;NYt^t)X^i4{!C`4hbTjg%=qldxLgZ+
zKgc5qjy^I3?>o!`w#pQYwVQ}F^h0R<;rB#G4<5^L8<zVHqz+4f-_39N+usc7Ie_;7
zDAQUXV`XFKcFm;zmhSb@B(WMtfbB4N*3GsS9g5u^iSA&)IX#mu@3P<f-VBuQ03!sT
z+XmQ;h3p*`(xMOaiR#YthqNXkv}PncpS=YW-mX>z^JjmqhDh}S8`-=?+a^b(o6!S1
z_5eGshkTU<tWbjFv@pS;a413iT7pDq#E0a9-I&)!xs=qX_v2vR(hvdi5Crv^kO&^4
zq|dVO%Zg69O<d1d@YfAo16_wGRP_3GjoN>~0xf{Anjk&PADASx$A%GQxaz%FjR=c9
zaPPJ!b3Jgrc@)Ru($CcxhBo8Lxa$4_K0?s1V8~V5c+H?Hn>?GDi`{C#wm?;`-(b&{
zVQdY?yYP=?=^q1_wAa<x$MnHJs`L{Ed;S_;H~gv{H@sdLM0u;sO&KIUGMpi!ef~6z
zb@_1|hv8U>h6;F3!!T9{WB(<55?Ch<nnKf$N4&3rI8+C#ZxHHI0lQ@Y`ZV|_&;Zxy
zfxas}@+lzSBmg3NZ6{i+$tb6J|N76K?j(@6H4Uh@Sh}ZXc@+Sq!wBp)1pm?itN#vs
zStfQ(1TixB-`H$3_Dtq!>0hy@QpJW?lAJX7M+!k4BSl((>uv$tyG1O(zwXu)5v}!?
zh;db*+swgrsvCe1LU`v@NcZhXLx24BDgv<ep!e25b}m`0wf>c7f@){xIr=;<xlb;{
z;CTBiE=*6{wX}H4iUiBQ2%cH+j;#o0bfLNYKYsKOXK)|)BW=|LxT`~NHGp@Y_S7u+
zXRQeNx}dC`=|{V66SP<t3TiijUwzvy1M}@i>QE;BQtHuLGw6aFkb&$nDp`8RT_!<(
z4B&sm%k@Dv&_^^S4qd?mv;%XABslv7k@&&u9eSFaE&#+S4igaTK+w>s2FM4%Br&Yr
zv!N;&4GHoBg0i_V!7h=Rq@uo_wNU}i=gYsZ-Tc4-gq(l??wt(5&;lqXXh7S3DI`85
z;B|i<7_A88LIibwCoVy?=LH;x{)K2^1MIU-q4b=5CpLrFM*3&Tx8nf0XL4BEKfV)F
z5t!ivp25smw?_fnRGAJSJt%;dYcl}Bsf!DGwF>Yh2DvXI(!!1daMUux>tPz!ye626
z0zCmh0?xM3?=*nr{5!IqEda<nYZ&D-1rl(+;eoWn!!$%N<8$-`(kcUGrT|L*Spi~%
z9o#P{X2(w$PE3F`J&gM!p7#^SOgAy#HkFE?knI4vh7MbvpBBB5C>Z^%KxKEM9doD{
z{%#ydl$-?TL|?~LP~2G9)L7`3v54!wpxn2>--<ABVg}UsXsnv$jv@*%;)LJ8{ppZA
z6j{C@S977VL5D1(QRErV!xiPK%;Og2v6s?YhRi8?%>!LTWAfxmS#I-&6|wFiPQ%;s
zE{~<=!5>6B73}7rAH+Upl&Qh_qFxKkW&o@xh{6H0D6HQQMW*wQX6#wP_+pq*2WT6y
z5Auu5aWn%qMKfmGj|mzfF$$_mA^uS@W|11<J3}@l+7)S51Na4r@;l9mzDp40fgc+=
zqY%vfIHq8QiV{P2#`nbNj4IBPYWq0GU{&-3rx(I|SeteJp!cE24KY1cJBGi<OY?vA
zF=tNicNSio@nt275?8+0x&eIPd|2LtrB@{APqqQ*h0p2;c=osU+w1j0_PJXLRk62(
za&{lNYcAPzS{V9%nZ5}Q=W93yAe~tsvo+$=i_*@VQ6pCgSIquQjOmHUYbpiU2;1fP
z#Zefuc*a?i+r@#gW_*j6h2W)@3wX&4JSt&Y1Jg#Rso&OO>`Nk5f^wb!(MnuPbtE=|
ztOmEar+cWJW~EG_BkUPtQTC~ic_n5DcmdDKpuhgN&i-boYcU(+k7na573UW7$yrm`
zrYtKboJk%bcmo*Vg)9<b7+C-t>fw|b=f<4YsX}Jx<?2dQ9YZ-5o|$PTxB&!gGGy`u
z#Kz?tl|+$xDZduSOr9wAKBw;;ez<l}Pbv|QdqKzeo(Tq0$c;ET0UlYR-FZP10O6&B
zGIIE10VJQS;43@Pzd}CK1~@e2sMWv+%Sq5@jJ~G*(YzV`n138}F1<IG9TqPW+=2QJ
z{@+&5Y^5++2H8fs#thqHl_48LkAfb!(5W(s&^~sBTRRQcbg4W(rhhj_W!TZL1eaE1
z48Gm_Bu>E$txLc5N7f6r*U!joA~^lyOZ|}cS9-OD;HJMfOxS*#31Gr9jha!RM3fEu
z#eQs#W7%6N;xUuJt=C2Wgy-D=`anI*3{i{Snz1ORRoB4uLU%_DmU`GD^FLA#)}LRu
zVW6nv8%})9Cfxn>Hdl!_^m9L9hrd)xRJ2SzV0e0C39dC*pHX7yNDRY}UQwTp`MQ`S
z(qg^J`Z1uhI#XqpyJVZDv$M(~k}IXxYtph=Rq;wl(3ltz)SaDHbUC}XQInzf)abXt
z+w22>noytUVvYF+@;w?P?Bm&tbmOAF?+6nS|9yI~QTVWd&CDv^CYf5=I+5aggxJI1
zwA$C4B$wYen-?ki8__q;kCcCNrhoHmzL3{r`SdXB{IL4f?NdUD^T3b<b<e<t<nB5m
zH1LWXPy((h-sPIyd=b2>a!xWHP5Z`itucnSAiqdt((oZK=r<4VdY^{fm@-HNbCBO-
z&r)mNrJj(D#HY4#qMq<+e)M~_6)-J$Bx9+trXp^>{B>5}>6P-a8q*o3_wk@+GV08Y
zr2d@=!~^XtO+9IbHY4oHj|z$O_cYP^pF%IkV}d|a_}Xz5;aM!4QA{X5m8aPAkvv*7
zIZoZFD?^H*Sm<-d&UC%qmwMwt65E+#>nQCA8n)#BqHYM#RvSi9z{&~huc<Td8gb;^
zg`u7MQ7PDkI8q}Rh@|4DXo44+66sTHJW6FEnKs8-CiyCg7Hkv&O}<mNUB)9hfrNdE
zFIAWu4QYkRAIbw^)T&Tiqne6z0B0RI5QHsh8#}pQ+d|s^Z?$|TQv0Ok4?ILqDQlHQ
z1+G-Lj5B%2SuR)JwK^joiPvuCB1#I16*FLnq_byM?^Lj+aFuIHXviKl<<F1;>oS={
zvY(JKvO$g17|X;5Z$QrGGk1KJJac^8l8y}`H>FKnJyy9^rpt?)F~v+SL9Re5;~x=;
z%#><4Vem<DcuDn$W7HjKeNgLmWDJF<4k3=gKK_41yse-d&U5YbS>s&lv4+{mxU=y6
z$@-;~yp^!SV&qQ^8&V4nYTemGQ_%7`(HfUW84u~QXO^<0C#iQ_kUFZ56Ul;}6=*~G
zR_T7IhoU3m=g63lD0(=lAaqJU5M$ke%)C<4Rd(!im2)B8eseEEyL2e4in!zz-&(86
z{w#VaekmSIx0U*%9srn2==}!C@)n`b{%HvN(@yxDQKX<Q#3HQIB93!p4y~lg_AfjA
z0+12keiki^z*SeDPP}J|SenXRAo`pjfa=}FaFIi~Tj1tN*(mtR_b}}#r~Uonz;3vM
z_id2K>ClYU%pQI&fAAmI5!;Wj!D^!&I+CNhuDbRKU9)~naC-Cig+$NqYw`X8k^tsu
z8hMRk1UzQlwO*fDKaEgmyUGCTK5!h7@0zDp=tkjD5m782Mjbjcc&{TIn$<AAWWjzS
z-K0}(9U2p-i<_mj>R~@PBjbC&nMkq*<H%GwSFs%`26!YpB7Lr$2#|f4DZDI&J2urH
z%0;~#+HufSsNl)9{;m7jY1mcf5cL)Wc1@EGO<7pn2qbh@7BumYcgDXy+c~S-=CMZZ
z9QI*xzd!B;mxv*3@%S3f2!Hg`?)IHIRt$iSA?$DbUQYCDf7N??udgrs!}wu!33LC5
zQ9U)s>1Hp6QTzG12a=iDuh+_Isf*oDXX@%}7ZHn}|M>|X`cZG*+P&+Gxdf0yo){_r
zwun=!*u8Anbc{L9U9mt)FCO22g2Kee;L1$^!Q$>9Y7C)Cmp3v3^Cv=7Ow%|6!E_j6
zjh<Zt1(s36nmtXwkNAQf?HuHGes76YA!q1LF2OvBFW!}wGL0-Z@l+k9LC~>oo}hIa
ziDqfx;vN+zBVEBUX8yx}L8E`7YT_X`%FJqFnDy{CyHZf$p?Br!AKD(hyC3BRLiaUc
z3z@~KF}V^oVPD0A|LCI&d9x<P-y%ca#vll==m$dz?9WMs+8`1^dzBZ8R#s{oFs$E@
z6QDIif{>gLWeKs;5J&%tclcLg-MLudzf8R!BhqWf%gse#cB*aBr3Olow4wNfQmhkR
z+Kaa+b&N#pQN*zIChOZBO8z_{ZfY)G<1V?6e_i(CX$#>^VunFZ2tNJid^OvNFNaxC
zpf^5^8GL$fy17!w>e@4Y8UqH>Ft0_?Q~Hl$_xB>|B323?PTXi!WpEoy+8wzToA2@v
z=Qp@@RJe+3D)g*ugHn@!3%(O}eVq3+h)rWXsnk38gfP7Oemw7Rmh#(qFZB6+*JIGl
z%)E`cHETw~?`S;>3?jgu{%eQqvDYxgAHQU8cRZS+tDM+Pow9^L5NWeD`R3q!)Jm|$
zg^S+6Shg@vQdY~5Uk4_yY9ywW6R9V1r-4hp{Dnde7KibMz|RmczF+K6(Lx%n)(;*C
zMbA)b6|Qu^K6+SH8*F~Ya$}eOARYk1rj4<#pY@J&h4UFpA7&}DIXx;@vZS+&tqKux
z->uaO;B_FcYHFe%)$P#V$<KFrv0A!Yr2Sb^p_YUsu+=l{+Y(4?pcB^I<F34Lqm80?
z;tnRxO>2dY8o6{c72s=ACRfof2O*DVrD;_U**2l4)@Z&Xys25>+;JZ;_c=8Ee7sQc
z%;G{A6Q?_YS}O#9dSqqGevDF)0)Z(^-QeGF%rUauvikkhM0}SX!4#%GLp{rw<NDX#
zO0+5s7Vx*qlMId<=5H$#L~YR0=rSlsT_5kr#(&KbWx^<pYz&IGx>M;jUTKvuMwz{F
z7>98O;c?DPLOe^w@uo7NAljS#l{~v>KU>9u2YKoqZ1^|P!!w;^A7pGP3UtJ8<4uy&
zNvt!H^f`A2OnVdNX2IG*177;EzH>@?+$n1-P7zIe>;5!?#yaptNNk%&tsB%rPzut`
zN!p}2$#g9z1ms;$TSEyeQnRd!U{4iagNd)7Ns1%(+qRd{HATD#=5Hj8$2_n;;cAmE
z_n)$R@it7?*l3H_EE+9Ab3VY&i{KyFnvPnM=1;lSiAqg&(?Kj|--4>8KXfG1%2VjQ
zUB89*Is`F?o(RJUiNW^CC>4EInHiDr;hek}AV7Q)B=PeTx+`k^9Jkgrnz&hcCZ+v8
zSJTXZoZKfHC(}YkhfUq0TyW@yf*tR`T9;Z?MiX0$Vh2~Bin>hsPBjk4Zu7faHsiFm
zaJ{NR`O+|KTtV(%g-fn~rmTfcj$WylHX=#egj?6&Y-hzCGV3V@rmkO~KlhYh0;@Qb
zLTud|h#{39fgG|Adp2PPz6dH~n^&=vk!lrDHb^Cvv8f6BQD^MFMh%~WnekJsd|Nk{
z!>GHb^YqWTe#sN{mBZIMlB$5dYR}_uu8@LBemhBXY;=5Epi7-S{@UPsvq@FnNLPB$
z0tvH-Ol*I^1Okyz)Fx2;3SmGgH%O7}cqP3*WMR8(?ha+VRnOj{MdM*~IfXfvU;q)#
zw4tF@t<O;eLb63<SwUSiusc$B$1G3|*^A2?sEJ=s>@4(8*x8tCr9!S7(8-WhAzD_L
z5^i4yHp+3b1r=$)#_0FHTKEv@8=sDJQAFej+?}2`h*TSR{V@|s<l%8WPYqN1<7ePA
zop#p7>rk#b)-;{H#!WBWnm5%AFy<#X+(MAOYZ67u=umg0cU=p|4<@xhB+}j2-P4$4
zbnC#oLij^6M2JRs#2&)pnldn~%+3kT>L{qh1)u|PpcR&D%VHE7nzD6RNMS#VTFC56
zE=fM6Qe)!tPL?f~9&@HUT4;!n%KgQSk%>L&c-1+uI#><eLx=YM?91E3&~m#P-+I5y
zn8x?Ccn!-_tM#QoKP%UM__$bJCaY3fJm^f;<8@lSKTmDxw0+_*adK7KAOv^Z1oIT@
zd7FF^(+4*aL@+x3jh&?cQ4#)vwm43miKWqb>g58Cn?S~wwY(C&oj4z=BgUtqOy~hN
z;Po4hO(+CS0h#ritRTzmnWX)$=(iB`H1-xE-=ANJ1}IZ~EZ5eisO%snijf&*dS`!O
zAH;rI2wID=RwpW4ar?T5l(fGOWowC~nI57^v!=1L*tvXQQP&>J3cdBq_^6uX%MlaO
zh=nkd9CTRCqEJCbX|N$hs6-ak!^qM@?+Oq<lrw-NX|#4qEXA$}U`pTFUKFT|^Ry+9
zMvjRv5t4?VWF$ot??@)E_OK5s9jyl9Zu*hrJ+C4dH@#jjfa|RUV|bJajt#FlV>Px<
zqrXPq;NTty2g-aYkdO6BUX!p{2e6Fx;tc8tF}Pmx?T>yr#ray?pI#r-y_(bTlFwmI
ziT<17>YW9Vs+Lt3mv^1|!@ZnS%}RNvXPU}2P_-uTPIS=4BZRO`i;PnJasumPl^fhE
zwx#(#DAi0yL#GV}D&n_$czKU36Im%F`cX3fS;$Jw;Q)nnc8B@b;ETu3@BL=&>YS=!
z%xE`pZLF%dKS_(DntGwU05|B5xFuV!O|J5u5Cbpn1^gjhAW38pWOuK>40d$y_O=HQ
z(>BRcTlSR$OD0HEZD^#VDpE+!+2p(di6qrbxYfBDnj@M<tf&LP4c^yFh~E^&wSN5<
zr;@eM@d#=oXPUw?Y1E`=&*XayUoDbX9~K@0{^PA!9DR6Y1zC#MmMqfZFnJ)4VV^9T
zyQBS4aVe9!n$V)2MgjAmS<?lz6a7kLvpd$ca4FS04<R-!d1r)uz2+fgkZ-I>MYF9=
zQ@MLZy-Tm<{?{;t$-_Ixrt0NpXT|!)oY6xcq}3vwdP_R~Mrua;P1Rs49(3#Yq0jZh
z643;$lLXN?kKe`WONWvB{y02l2f|G1T0<*SiN2~G?{8WNMsX6c+h5Hv^e%1T6BSwG
zMnTut)LP`>(MTMnmf+~-kqT;wdU#Dk8aTHa$2+zCJ0WEJ5tYucJeAOuOH~NRvRbOc
zRn3BiVM=T|VbT3j=->?R%O$xS=F)HVqXN_}%kSuwWOQSM2S@ynEy{ug+PSo`HR{lO
zA+j9R`X#vQa7kulqp#vV`A+4KgIs-{^XVcj(<6+=5hNWICQ<sc2(6NEQG{-d#PWQi
z7ilc4^9_Vz6=XB(A|G9WhCb{gl!24y!YT;TSV$p)RQ0Q2>A@-sf2nQ$aw|2&v9J>P
zSL04`9FiZ+!d!yQ^b~Y;K`8aqA}|Q17KQ3ZA1MH*U3S!eqYhi|NkhmeAe;fKs-<S<
z!h6DlJ`Q?M1FDNCc6%&R+ZEhmFs^9v4J?xu#?pPBGYsDmvxYq~M|Jc!VXu1uCn$P7
zec#mg^Q=o{EXr;D<@2ZS@bf;_)l%J>7cOD;U85QW4y5<tQ3?$Z)BueIU?Ty~V9`_%
zZaZ9(TwJbr45;pC;2-dbA~7gLxtidprGJVHes*zFj<YnqB^DY|SP=ONCH3MmwJ6O<
zL)|-8GX9`c+STCd8zOx(Ttjp6ow1BH7cCXqRp&y(>l0bAk7W#pZyKB%$uIF;E*6)T
z#IjNuFQ^b;vuk<Da6Br0Vb#t*x$4rGv-W7#sxVpM)wj>ODY*GJRtfv-Cm3C_{pDza
zgn$KywYukN->AF~ljA_i-WcNpHaqz!?QKW?_5$8812!wGjc7%la+`nyap#6vkWVra
zZek+l?(wqb_*Wv>j@Tu^HHnWx7bAuGwB(uyzQM#JfAL~jvCin)8n^49wcVTO*8E~z
zUTInq@*)SkY^3$#-}Ka(O0&D+*L223(3n5%g}Y1>&)9xL()}ExSxj2Yri_0rxUv-;
zLU|KRQ>2mAL?Kdsmt|HhBX1LTo;>q|(}VQsQY#U>+$V`&K-u6BB&NSbnWoeb9+rql
z0ewU!oem$=&Af(36NM6<y`eRVI};1t)@4m*LZ{)vxrPUM`JUFEtIyM)tUWmH{m#0k
z2D=W&YSy=s30be=ZYJOw$K;pKdu{933kPlDt;ZcE<IA}B`q#DwFHQTC@|f8e8(RF+
zWpdK^A6rGw`-_*iqnxd)4Kss)H4lsFLU0`T6u!%GCI0tDg|ckU&!Ap;c>*rmo%Pn1
zqh6nurS;794Y7i1W7=+7y}QIOQ%^lmdcnvs%huiq7(4_MUwVXL#Kg#mFtJ%2q@;p}
zr(!)*z2p{%PJE<<3qhhH*m723*LWn6wkh3UVFmZBKPZTj{BsUFnzfVB>CB=xHv%>z
zREjB#k`<!!cX8f<BwDScySQN%XCOkKh-9!YKB2tsU`;I7g!H-0j_MCAMv}XGyLiM2
zZgxp&$K&q-VTQc>AjgD(?<)cc_uUK@?!Kd3Q_CoQ=Uq9WPKhs;@7%z;MyNxEzVCC0
zH{<BPZ%t^nxQUL&uUhH2<CcZB@tK2pT4S%1(cu2nE6&3zqUXO9r#n|*RNn71E3;03
zEjp%`v$7c$5F+U-x+7k`UKYz<eTTN+U0tYH`p(^_+bLit_xY=+xq1heYxdgbvi=}K
zGI`G8RGLwamcPU7gva%1ltbW)hx*w0%VY|kpBSR)CJ1+v{WAHE{@JE61M=<<9WsV6
z0`-b5QV}^+U)(Pp=}jv*ryx|~m_us+jzU?jkh+tggga9moo$Jq^z25B+m?$1db-M+
z%BFjR&uN3*_61XX0>WI?@3LGq_(V7FX@q?CBv4Y@eN?J={6f<|p`MrLjs#!O>}TlE
z8X{XzMBw03)GOM;!FlFWq3Vk1{uVRSp}ZLusNl!kr1NY;>WsyVNfp6IDhDso>?I+X
zkf*3{phNE^w{RhS|FuYa_fP4ZspQBUv>*&E&Cb4fC+`=$81lbmPZ&sn-V+y;lQXL-
zt>e`k)5ZY>75`w+a!f^tdHe1n+^uR`Va3TCZmimyv-921uPF*0)40buKFDAptF4tT
z)v^(;HgkOIO-B$cl?J$@=F}g*A6JbeM6|xeQcPx%Dzr^UwJ>E)2OdN${-JLtco|yQ
zy_NjkM_;uvzm1ZG6MvTeTK+mlnfBdGH3hR`5us-RqHqL?k52C~v_3}Nj9M)`(Jc|j
z9f$A+3bR<QLa%sT#kZH;I;-&lUsq>6_HIJ31K15J*+%4R^Lz3xzAd_qx>Y+EUmj|8
zB2d|6e_9#zz9oz}9OsqEw4&M(UtFxf5gq<+PM%^?ni)h<OQE`!$$10$GK{6iD%$^!
z2RD-X7M3{f-*toA6g_iGcF&);Q8q~5&^Df5-MFv{{)iLLRaKibjN!<8LEstmc_~5}
z{&AK_*F;2cy4!F9gLd1h#Hu$P=Isd@ka(?}YmPWA<1Y8X^A1K3DJGd#FjshEzw>9?
z31gJgyIk|Q`G##C<@h!g?UKWGW1O$a*m7V+sYz^vO2-@MEnRSQPAYucoBKpT7!R*o
zD0<9FubARQU{Q6_TP#?>Qb;3{Hlkqm9Gs<}ESaCcx|beiBZXJp^h|y-#YcifzBfW^
zO!+64Zs&HOX`lY)y8|d@pPWAu7M1gQLYV1qr1U;|AiFhQHDV4W`5BR!w>8Jq@Eb&q
z@3#b4+E9l+8QXMUoQ6oP(IZ%o*Qg0f--Ozj7X{8Se2gxmE{y6QODk5|Y<?-N!yi)|
zM4zO%pQKFC5{g6>KV4W%@%mB~4*mqId|xXvHWi6^9pA4A{#80^OP#|xg8NWDyKP_0
zXJ|3`bB$5zQEF5;m?<W0xITRjMB*#dRpCdd?rR&e2z~XN|Na<V&uWM?6s%64RA~T3
zDw$YvR0&IiI1&Z{%LOHh_?rVW(Ufbv{}q%a&%Bm$*WZVGk%i0R{kZDLv0@yWBj_Kj
z5K^M9DDC<Qdy<*kn7hV*)-@#)sA9RuwNfdj$>0MaZ@=#v*G1RuFbxG{u;|vM&EvWm
zLxqVH)$NL&U^rkt_hQKK`$p7NSkdU?R}FhoKiR|cVjXQ`N+e|_{FI|`tJXEX+>qbb
zeh3wN^0S_6P+7Q|xD|uL)HG`l%$|j$gjYUrr;D|3jzYN>deEd{liEo0&ebXZ)mkcj
zOYw)~B9L%j$^dK>?GV*}811Sj&)aQ1fhaQM+UN$t7qc6td+KN}$~$kWQVEe^H6IQ?
zW3=sOQL_Q5g4U6tUq{Z!r}`1tm%Pu)ONgcW%df<mg==A%GMUhR{L1(z1;m)PQk5pp
zqlNXP*PI(&6<j8#8`7chFH+kWJ=J9D;lzT*p|J<Wv80?N+MisP{vr<8mvv$uzgA?_
zc--MDV+?}Uu_Kl-Ea~d=9=r5IMOg7QChH~Daih4k%sK0=bfuNyo5GL&rr|S4m<``W
zKM`c&rljw3z{l~RVuO~bY1wVH^BVhemK5R65*b?xgKIWNbl$~p4&b-G=4+#Fg<G+<
z#UM_vq-#^dV9yXe$=I>)b!6f^lel_&@I?wqCLmU9B~!=pN5o;3kPWI;aFRY1z#wyf
z-4YsR$*!fl_Xys?fb$r=Og$+QHVg@@``_MW_?hi})FfWn^`@K95=#9W>ul21SixR`
z1~A%C<6qnq31ttDbObTD)75UD^fiq1)<%1eX4|FK7;$CS2}bX;Iyy?iSe;m>^gmb!
zjvfCTvxvZa<56<e`dj6fn%V5OAEBb$yTU#}g5(tY4qFM%Rhz|`M$C>mxT}sf?Eb<@
zA{Uvk9%o~ZH%H(`wbzPsj$m7C1KTa?p5Ui!kEkZ?BgEd;o?dDj-)@1sNdC)oaBn3`
zF6XOosa}oI*I!_`)YqV2=^*O;onr)eRNhU!=P=c#h$!9awDoB-=zFWuKe-+RkENJV
z_;Bhb*rj`qr6&=PjqO4{he7jGG&XD;V?}IHtW6zSJo?XtRlps+HNPev{HR8Qq{a~(
zx`MNMXk8GBVT`S5mpUw1O+IaM&rMLz&`}*F*26VF-iCU>zWFo1nzz!s@=av&cv%qG
zk{`7t<<8J4?3&SQ+0h?grUv&%`7N3)S~aT8^c0=DIWK5O4%4rDNTgSnzO7bziM8gm
zr?KO@>ezw1x2dn#%zKwVm(riiCegGlgF9Kn<-6vQoye|Y)ATT<DbpiXQ|r(tF)`?q
zW)-IuDm*`>hB~E9H1*IHUgJXthF;?lr5KE%uy)ZV;^osRF@Db3Q`X+~HvGZ0&9-iV
zg`N9wYHSNs#CcA1$DaP}&0iu($31X>mm+Q6rHYWUCM4>!{z_a~p{1W5AO&ljRvYWG
za%&yL9vqBe0}CB3(~wx7&>rL6{8s1)i?CyMfX2Cn(_}U^&pdvbqs;JqYZt<khr?DA
z@8*~+C#A?CyADM>;{LwcG*lGZ#X$X?KOdfWKuCPK<r7HtuD2@_Ea~F?%@8e>YK5xR
zU_#rJgzQ?9m{j_sE-SD;Pa7<}-+??k>d4UJ&oWZcuF728#YoeESFUXf-pT0We)fJy
zzK}O&T94eeX@Q!)25Ust;N<pE*uPAoH}5ilE|^>$IZhXGcxwCZ`O43R6RC?Qv=b=|
zU0eqhE!@I-aHT6-n|fM^*F?l53^vJy8FuiV6;fV=VVMUDrq%PAm2YcHEcjbU2xlSs
z1GffRlM@*?5a(a+s6^aSN$Jz6)}~KZTCi|vbB%Y0i&R9HLhHa(8$onKt`L1mQ!e=}
zwo4mpg$kJheU#ibGn2+F0iQW!eCKfVOp+gqgg8tPs^=D;k0A^a(^)<?c3`ZY5Jv<D
z`V2+n-n1YiYInUu%RYjTMm&0<HOkgVVlG?CNV(OxE-JD5>>tU%Rbs<|G5=6T8w;jI
zM>OX+p>53PfxlpyGf*&c$qc(;gwvFqCwzRLB}m30l2x3rAOqHMgtCU2WhJFm<DJf+
zm{ZnUHK$<u2Lom61ez<&A4yN06f^B<Ox!IFerYl~4Ai7}iN5UWk{;O-z2Yx9T4{{y
zjT{={AGBTxVEfRAZ%wP8X3DzR2YZNgF~)qQn_}vF=UE|+#?1J1%9Da+`0-vTuA|ty
z4HXvU#=w*E*6%QA6F-@jSU+CQuaiNdVHhY>b-a|!{YL1#!B8WGqPs6gH_p6!9lMmr
z9k-dx_g4d`oSuoAelf-zbcUp*)1}Zk4ysXuRz8zpD?cdl#Snlm3SwyOGseMGU}t+h
z88GpuvXx(M=)%mZDbe_K39Rul@j{LpVKL8~InS&f>0X-n(z;|;p(|k`I4Srcduh!r
z1bYbvlVcrYG70b|Z6tlA+R$J9*~1w(T1R-6@_lsd@s+P@wQ#)E2es3{934xf<}s||
z!yHojHaFB>aOM~MQvAz4Qr%)zc1^wbg$rcewe*U5R(dz0s=hbx+#}&y6y%el<oE{^
zis;!e$s=Q7u?7aIl%c|iEEjp^&VdqV@U%QbOctTc@{G}r?}x{5E>nUI*|sw~r6<<D
zK2S~_mp_UdeevCFjW*^9@KzRN-#$d#d7dwj_|e<57|GUV3Dw(}<%#Rcm`kf82>DfV
zJ?CGQX->Ot37Uw1p0z3#xog=}fqfi&=;oAUDf+rIVtYer!<v8up8b~LgZCk?6gUD-
zbeXz?A6>NhXay}~AJft<S|SS?^QqMN?OCS^33~HO2<QK*oz8M52_i@vsDCStD;tZ|
z`s4oA>9>DlO<1+Wyw(C=j*R>`x?Mi&?cbYwhb3R8k_AQpPtE28YLTPnOC+R%f$7*m
z&@G-IrY|L2qDHOMS%Gw7^$Beb*fN`JOa^F>K|MA+wsq6A-`+NUoLi3|Yu2u8_Z!Xv
zb2tWy;Nv6By-@TF8e6d1A4AmK$LPlc?x{mz*(LL*216fh9hkzD&3g4LU%^)NJ_tM=
zLl!OQSru>pAu%PRWn3xR6AI!Bc)Wa2yCieQj~>Nd@>xikNm;aUE+b1Lj}fj|uku6J
zN3_+dc{nT;@4wr{{aM6Yd0rS|Wb-9qVg8$RG8LJQ$u-F~$wzOgyU)AC`_xgo9rDyn
z*l*)sCmS?fD)F;-f2K(1uEe&Kn(@k8f3!LQ(Iqvpa(bOZ|2)s&tDK@tW2K-tXA5tu
z#Nac!GA^gf>AIH9Ok+Hmr(@G|tm31mf}YrpU_a%;n?xpPlUx;oCplxRe=&r(;$c*u
zk1^O)&1oc&$C5rb{1Ts_J*|V0#ZlK-ksREa{W1kmroQ1FAc}4NpcG}|%iysWg$9R6
zyLL6oAN2mIyyCU0rV?A2pjML&LHscsse(lPwC$n4pK2qBSbVgdT0z{8GQWZqjJZcq
z=}>17rfpzA`+LkHZ_-kb?|ege@<BB!h@FK#@5?iRfvj)B;I8ji5Jr&0a~?T8n}U4=
z(zo^0iVt=ZZ~OCQXq=4KhYjuyV0|=ZD$z4iMl<b?o7<DWDd^~K0|PyQk;~z0UMd@4
zZOaCNc^Z3a=8(**w44sDb@(+Q_t;_kQW2_4^ulamiHVfHs{DP7n#qcTQu9fHcRR1_
zcXA#cZ+uR2uW?`dnS6-D{!VMnz^7Cn<AE!-wV}uXojRE5YBCH7kUy}^5v1q(Ex7o_
zkXSDTGeVkk*mXkn_-42lW6|;B_L>PAm4Ni$Gf`~2^F1udU&PPNvT<^eXSGYi`zX``
zRSZ>6VqV-nOVMmS&Dqr%X)hgK$}oOjJ3G91x6HWQ$TF4nU2JTZn(f3#VG05RMWq?u
zTFxi!^IifyCI<K8f&3ua5r25;D=^5h-eq(Mjtu9=`OM8?l^Yf7@I`H;%`@!5+t`Db
zn_Y1iH7vF-vb&9lh|f+Y&e`Av@08eC)H^V38G>9g)12w#^&i_Zz6^cALT<jad4a4i
z>4AWjwrcgq&bKzkWJ6VS<s7np#r<3wUKAYvCi3Nk<e!rf>$C(LSX4AyDv8OL{3B}|
z#*su>YB>Ym2}@UdS&2(y#ID%anjz^zl$Wrx5|WF9>}Ek>WubS8UdWOd#Q}M!4kVKE
z0;5`0m$o;jz7@qtNvi3>)JOAKXfxq~9<XAtezp0;EFQJV?g-E-gXK0JOjp*EdhIS}
ztM63@;f{X$)g?Kb>tFqyCnio^l{H_EavGjo9_M!vSONmRH@k=jR0p1-`_Pl<NWQn{
z$BMCYGaDz51>#q^?;FUiz1T_JRlT<q1K2-utQr%~md3Fi1AON>T-8!e%6HrfwTA9H
zemQCsi&Wux(cz9_dTwhzh@Ge2ayi|)sj4`?=ns_)k0b{ljU4t(hBUM8-#!LU8V#@K
zv)|Jo@WT%kd4a-e<GBVdg`i<V#4ug#R^hA4OCW(xe>JS+s$|Qw3UZSB112kJcQ2Om
z1im0mVpnsJ(iGJ6TB2b_+$9<J9ieA|AA}9&htDMl>hlGcrcF=f&SwF9cB(K;v4du#
zwZ@vo8jb$vQ4bO}J#GT$o0h^vF7pRlgbdiN{&ubeY`3+5p=6S|fv+d-ePCl(9STkU
zWeg^1#itd?T;=5$?=+f*Ga6&q5!#>-1hcLKciJdDv2<(FE^d9a7GsxmW#g-)b0i(B
zQ3qWWA0-_BhdZTmccKOa&J`TL{PIX8mMas*I{!}Tb(MWenaO2+lQnhK_qlx1Njjn8
zH(JT_sNfA_C7b>mxJ_LFpq_2Z-`WUw+^&!p{#c!R33I-@1`F<-LzdLmp@x_y{2Ew<
zz`-@Y8rbWfp@8-<@qTZI4plgI5`JO(V<)n88INUulm-wTrQ$vdT|NyNFLqzF1ijAu
z(vo?5Apy^?4Mq4+6<D9<{nW@*iysw+;oC1#6Ghm5kr0|?HZlB_8OVwAc{CQ@fBNA!
z1ypY4B0F}&Q%tL=t4oiyxvd`MOiwb^d_Tg<&(|PTOhymgHedHQg0<P5o>~;06cRdy
zd+@DGCkxn}wguK?R0}50C!M_yi2WLIhrjgMm{{3xC%dN=eB@A`@$vKw#4uhRHMbhl
zvrQ|bZT&b4j_#b1M#WdWqhrYqCAl1<8MJbYMt&ToI!<sUTU#kg+Jqj={$+{5Kgj5E
zU7C7lYj>?G8~k-(dLfXcwcMxO=D9ntV3kAEt()szJx|>YVv&c=u^GOYUv43K-^I+0
z8yb@9svTNZx7ChDW`-suX5cRq9|@_4by8I|a=(|-^nNFaAM7{ENJW>VAe8!1tfs;{
zLN3`p^lqrtZ>zg`qrl*A#|kA$@5=FFlhie&U9#TDd#*p?55xA?1I;lzB;Sq9XZTa_
zKNx$fph|*hO&52EhNf|M8h39T8h3YhcXx-z-QC^Y-QC@t&Bli_b0+S&bLL?lsxq@y
z<eyJjRgtmQ|Lycxi$jBPZp8t13x`wx7W0Q4UQJA7FlC2qAqU?va{GK$$+x8H2;@LI
z*m5IZxJAZ{u_jAFL!^ZkX}HoLl9by9wR!_E5Eng}zklt^l{I1>1d0Mub*i(c;3fSO
z->}v+LYO_PlK+MN8W+f`7j?|!+EyB^iy3g!ablGd^CXuO^*k2DMs%nts`RciG-sIV
zGMFG(oO=xo7Zq$2bF~yb<9ZXvL)_n}VZm<T(-5<yEg<^sZ1p>6Vz1rKTS}ZV{UZH>
zTO98YsU6893#ZK<!JM3ZfowJ5QX&Szv|-ew+RfEVDay63Q>{g<XH~C=nJ_5D9k0L?
zAu67c(S$q**0?{_j!o#Ee8@G%myeIaN9Vm9m=9bB>Zw(A0Y?Jxx+L~uA-O%O{nVFj
zH)@Nd#Hy~e+Mcd=Lh#ZuovjW7jtpI2UAkWeQq(=yraLlNaTt16n8#CV_09YEVo&5W
z)@MsRO$XBr?y;=4e{fJNV<uj+wP1Te8byf8urO}KLlU4k8sphv)Am`hD(L2sQ&IUF
zm%Hqu7+k(-L=~llZ>{+sQ`wud{Hfbb<Iq%J^nC3LJlP(bI%1>AIby}2yK1Ssts!Mb
zcAa<fW(P(RmFLtClrz8kmuS!*fCC(Ofm9vU1a6hrNadGj={m|qENvBKF$0U*YqZG9
z*I-g8of+*F&*Kw0tIP)rGk2<P{8TnS)z#e0;K!<I){|J<lh-gyV+CZ=VK2J5a-L(G
zpz-e-0lf?B0~C#wVhkh1{@M3Q1RBU2(6h`fe=qjF@@hl9X*}b?gO-ah|G68nt_cZq
z#SMyEAaW<d8Xut0i4Hc{Fd`Q10i-?I)C-%6v!!L5xmN<8!vkjtc@nY5O~TYE;bry6
zYY$rQyox!%yP_J>tX!bjGjxM6_lhPJ!YU(QoXzlbWZ9T(>unt_UaYlXb*b?&e~0_j
zDH>YCdg5~ji0z_^C7xV2q&<PBYCDUR#~|eQ<bm_dAK-LnZwgd>vdbl>`Zs>E-})+l
zy&du|eo<n1rf8hL{IXn-z{wgGA6Kv3k~W6H9Wp<>`8ug50=#hUlgiL*$o%t6$kuXQ
zqu`gyK7ty@zQ<q+o=7nDXECwo)S_P#*gjkyhaLsJmU75^iHO_{8@#QI`6rSEA$~(h
z6!j1BsDF@QtlIJP%qB3@Esc^bIr6w5SZ$7GXjGmnmM_bEpc!?+JS*-|uf2h<IkLHh
z+9j!dT-&)(RQ!k|%&1n@@#E!FC<GYAB7k3t^Gz(PGa_bA&KmuDNCBr<7~gLHS~iSr
z29+&x%v4kFDA>*UZbc;KrkqK3B))b})86)~)xKe=6}x=UWQ3^L+dVX=9fO$v+dQj#
zRxK*x4zoUZZS4B6`C%hCYajFDZ5VkB|CsQNQI+1?ppsN9Rq+yQlUA9cpW$Nw9}j>)
z-E;r&Lc^azgEjs20C1m;ah?>0`*w+Fe+|zi!Rw7N@sByZMen}G-SRP_L!glxtV`7!
z781C`1QuGw8%f8@KqbY~<}zp-`43gXu!(}vF~E8k1f-_N7^pYqs~0TsQsyda1bo)&
ze~KjmU1_`&P1HFPok#8O4XtS03AxpN5}S_I?5<tX^iSNwx%hYb(;2P6+m2Qo8gyoP
z=WJ1T(qN74F%3BAfw#*I_i+}=Y4f0u3`xRrwdC>2n65iiL23}luss^JJ8pe@V&>>m
zHmNWrVoxIOhr$I83eL-Ehv=8@DQ6nG!Poq7*{2gK+oKsE3*3nlj-^12)vyL@FdCu!
zcN_CC?qmGr{_cN2?~WULur?VYg_yJNPy~G@k~tp9A^T(;$1SI8(m&xP3qu@FJ1V>u
z6*p$aNc68SpPeuw3W`Ci3v#kKQjj7oK4!ug>6bI~X;)S~#4>fd+P~>dN)UPDj%%Xk
zi$?G#>J@NLs$?;?^H06#&EJAli6>1w`@%W~2r`_bdTee8<{{bliVpk(4Cbjk^3$x7
z%$1C4#t_1OMQ2f5iYP89%$COiMjg}D$gCmhos>|-H5ty1cLvN<Dlv0Ir(Gcsh5=*N
z=J4B0(WDJyTHI~X)zRHqL|dsglFd9ZUaMkWQL#^uw6&3oe|_d>7giSz8xRC!4@$-a
z$OET9UJxK&5R#z+e;_R)vm&t?tJT|KV*i$9kYO;^s2@^|u1QbMUnvr47kMoTDH<*+
zMpWpg*X@~Woof_n5$RcIT~M-k8Fv`lZ;66bcMz?lTi`Nm@{wJB)Z{juI^+XaC`d#~
zAHe&1xs2w;MAmy-sxc<Fv`_n7-hzxP3SxT^61jS3OO?X1=jQ6QIuV8<!L&Rw%5u@F
zoBIX!Y*MD8kf!jWD>b2ADjP*2n~)7wg8sK%k6c<BW2>D_x;?eB;QhS#{sPvXfmz8*
zCncO&0`uRPi9^94?iSD1TYIfo#KdnVbez|?t|!N_?T|XF<K|a*%Vx$p2CS$=CN`L5
zLq@aW8t1UFyaU4`xdsxUk+y=Lg&XvFSHRlCRhtd;iWDRIVe9~6qiNC0QM~&gt9^qq
z;&Mo{E0c8r>U=!uzsQ+Kl{ICBk#M8%m1loTR2Kf(V+YOe;Z8~84@9%u3sn3K6;Kx8
zb;J^MG-fMZxY?kn@w%+*l-TgSS#%!CH7$|GV>27dQ?~5sY5F|sJbP){sTsqm-`_4y
zBGRXUzpnauf4b@5ZLL+OFViJ~u)rxKO?zS@Hws{A+;ZfYY}~4+VB;ZO)bEc|2<GBZ
z!3=K4mGwKTtNVKfQ4ua%g;2LHt*BS-FY!6@4V195c(Bg2)C;iI+%LxFK7$~b<Ul>W
z?)KWPEmI6d)Y|TPxFP$e>T=%b;4pTRg^W1fp<2sCYjjlXNdtXS_41w6>9IX=Ve8$a
z!n+VPV~3ThLJ8*XWnzW2K!T%*VnQ=3MyyzvhHuTBQ^l`o$X98vLucr5aN1CYf_qpf
zn;X}`uooDwxIJqha8`x-$0a(yW_dKz%*%2x_e#2=zNFnn4|c%${7rgIaf_>bimx_#
z{`}Z0nKkBGQo}9ijim4aHtQf^U(0^-R^^e|B1`^58aYU>FJ_KOxTeAq_KC}g1rro?
z8U8zoGUJp@R4l_Y#)g0+c5%Wt+Z)~LHw@W}6A@E$MATp5X}$IzzRkt;_ewN5I-{)g
zt_;n&T28%rHs&R%IM4j})cZngmCKhNbyX^j7M2$DZXQG3ubE7*0{9u~+i`OR%%ND=
zkML3wf*&Y!1XDzilaSZYuNCPQAR)HQ_`~~SoNAtLyO|pgpGi|6vv!$<D+GjJW*kEJ
zmSM|g_<=RdkVcpWaqa|bG1Uz}$?)Je86<3X4A+-@WlX9p-Yavxmmi6|l6@ZcxIcni
zp1}x$XRb}E0~gQMD)n-ZtWAy|5F&*bY`fKJvJXl;sk2!$sRK_I9^NLOp6yJJ6t6mr
zqXK#?-{~8$r&g{-f^L>syj6MMW7S`^uQxNsOIjWEtb4Ugb>w;~{^;aHi)E0R@VP<r
z3HvpIQRY62WpjCsS7v_HpSdC{3AKif_Tdl^pYujH`d(-nuswtU=z?z5isWh(%(~qk
z^179U3|SrK0=wHB&$r+@z@BLZ+8+@c!lDDRs|X}UuzA05A2ETvQAU;Yn~h|6GTt}8
z9bK-|xm~f^_)SlyIUQZE@=c@9Zj9HLJZT`Wy5-_U;El6uiEyPg;835+qqCIZ*;$}v
zicfBE{MB6nv}-=sn;*~9S&ECH&CkVK+-q>5Z|faZrukFh{n%88SF>Bw?=V+-cJt?9
zM|TR622M(y6w?;T*x;llQ(N|IeZ(3FdqkA<MAT&m>bGJRMO?e_0rO2b!)FqOiN%J4
z#~D;RuAODscV=7?wlMF6LO1h-ufNrJWpeWIOJw~`cR4<+NSNkzqs&-)`2_jbIEGl3
zc$S!!xR%)Z_`ZE)d08zi9+hSiQ^)GiJJB^T;DQbElqj*z;hus+gbee_X0^<y8Zp<G
z+u*i><%8#fv-?x~SNmIqZ1o%Zf&D@KgZ;-D8iljE;!KeyV|49B{xPP+6kNk?Mjj~Q
znr`s&LWuw=xLLPaW0dHhvKx!bPPR0Nzf?3N^dGXfvYXD}+1)#>2Yv()P4O<uxNm(7
z6<1!(AXSaNi*T-8N^cnd^j&Vw#kNsoE)T$Qw&|~(;h#pbIkRs;nEnJt>Z6}hotaG<
zL6f!dN(ip~JUFEgj6Joz>5;VsK!6TgJYzw#ZrMu%D8ZrjwKs07{l`I`cGy}4zKic2
zv33+_{dQ&Rb~RmB%$^%k`)qeWd$vHG0c{%X^XPJsW;jo~@vx9p^YHn7(HHPmcqH=h
ziee9I(5|h1@hiAC2s~f#H8yWMKri)Og8LV)nC@SV-kZK|vML1wxmQN#&>!v2e!(PP
zYB`{dW2k0;Vc*bIl-74?><<W9R)ULYus5bnBv$(8+qL~g>!03AwL}Vyrv=}+E_LVL
zJS_1H3mFw>Y*jPF*t8o}q%2sFH7DB3;9QCdBQ#W0aHmYAET<0x8w|+J28*@Sv+QQt
zAsch8zYAc_mnX01=e9ZC-$*=JqTg6>(pho%H8yG_@}FPtAKBZ5<a~3;D!Jj%vdFn<
zBBB(J--7sZRr8*M-U!L9#g4TMZ;4Lu4?JM;b{UEQ#rfOOsR@@0F9Q+ZXJ}LK%oR9M
zUouQ`2r^RxfO$&qn?N|pmzy6yzTuhi!?D{X!Uy#ipxIDnRu{~W&LhM$7b0_$8qwaF
ziG0yKl39Ha>UbHE%bJDYP(rYZO67}zRr0=@lL4#<<#K!;wK*b@cUk}vZVmARR9)nQ
zkpfJ!)ZZLso^tuAdPt68d9Y{>cLvgh7<elsN8l=DXaCf@!nd(7j_<8v>vpvk_vQNW
znKf=O43L6^h35jiCE>IOQY{rGSwT6W57(OiZjhU>N`B=1<v;?8vG14qL5F(2(Lv}x
zlNl%Gfu1q<%6|I*ocQmt{Wdlw@x-Ah`4nh=<+d>r=CtOsJqOY_gr6gKWO44?*pPI;
zS$h*4$c2A)h5>JHdSHK3qQYP&b?j{Wu5&!b_K=yr#qmbhs|A&iDETUNk)f+dOxidl
z?nJUV3Yvx1qIQk)haCS?a!Rz%Q}Cms6^iY9I~ikYOw$rI1koil)>9Uw#M%|}a|VF~
zF$6K=cL&)3Nd}z^ommF6D)wdp+rX{@?RRWbADx^Mp=fnB<03#_4U?(GnS<r~q0?{*
z;T{v<i?@PZJ>&Q7dBgnV_m+EhiH+#!ZZ{R<ur~4ewh5^u`i4<mP_4!0fl$2!?h`W|
zJdXfu<>I?vnSE(cm1pyE<<AJ#fp{m@4Addr$t7Jst~BG95X(t$`ho=DRncV%znwV)
zOrIX!&Q4NwZn;qZ8uOIWb+%;ig1l>lK!IKhY*<nM>W4JX&rOZ_I*A>!?ELEKMy?0`
zllnrQ;FtgC>f|Tz<@@w{e*r$+ogB=J<ivhsdKh<db$lK6{$uo?;T%9VBv;INdwliq
zsw{z2J*_Xy^dG^&5aS~lgHq{I)p<!B+N+4Iq-gUo{YK`5<?XGt=_pd+^PrF*6hZ+X
z;;;F_zv#I^3O;2EzWwIlMJz;_J_~GdiN)ssn$6Qod1JR-O-=Y;f2w~vq`uVxraYgw
zzC3THpWvWT1VBI`{>!cR0`(k;`UbA@|9fVa;7BVnw?citgG~MB3UB-m@FX5hrsO{E
zy}|72A8U#VA4wJg1g5vQUprN;;J2KehyGN-{qwrc1s6om2<rc_!hai5z@5`cri^d$
zpLmIU_JAZI9a$V@t>UY6tNyD*Z6XLL*%fo%xR`g&sq!qoKH|!2E>|sOFN>`_(|LGt
zJfEJMaksI4ThCbW@x{|z>8h^1&UT)urCF7|D7nb7UR}4-9=Jk`WvjN{TyoPs2u-KA
zTHknFe66~$RrA$-NI&+}_@(ixX=D8&aq)mdziGD-lCr_pqP@hn=Hj{GZT)P$*>tys
zlViQJ{+SIpVbg9dYdtf}&{@UZ$=rCq#B4soYT;=;`fAo`ekyY0o?e>y7<0Frd3$XR
z-C6w{e%8x4QO$;=>(14F=CTaVg`_hR<PN^-`=1?njoARq#$3oNgl2xk|Jh%Weq;Np
z+nBcH^`&TP$?-KU<2`M{(uu7)*vo~qPPz(K+}dVA`=0VG^S@{RhvLNs);5c;;0)yR
zgs;5IfMJ{py$X2}RFR6g3LpJfsJt_d<;GoujC^7ht7_8c)>`b6ljp~;qV8jn8r_P;
zrT58;K!sSEEgw99XJ>ag%i{L#(+XuJ0JEk?#k|aW@M4AL#DD(@N0YzngR5qzP^;N<
z@`9b_siWucNi(s1OZQI<tMi53n#?}^8>^(})Phcrin`{*pBiKi`c9v$d7X{DCuL*D
zE58-pg2tueu1}4k0swuDZq4JO&(a0E3!s<nTU8qHzWA+r)p7ImN;kf~C6>fDq!ck;
zX+WiJ3`TcJskK&Zl!ymhtGXO?p#lt}1k!Yc9Z<4;)?xgaOYB7JpG#E`cxM0*OCsT@
zCXi{c@nQqbchk+^%SgyIL07cNs?85sqC1iH(!mU*I9aqDy>Q>^TwzwOREtqa0$YE{
zGn)mA{<1Gr=*<xs#QxK+K6uc&EqwiJic#c@IkG{;#bADKgPGgy`wu7OzD|8R!`W(t
zsnIKJ1TKvs_RxWcs?I8##cCj(?U6yV(J4hMeYyA{bh01fW{EsSebp9s9w2)+S*ViA
zU{$RSHG}D9Z$X<fWw$kHxdtYwQe_Z<(P+P_ojDH&oQ)h$9oN51J>E;AQ-_`Qsc(*E
zV;T$CGJB9pny~M$y{uH<-P)7-T|?gFH>=P&9G-rmI1<T`bJMCWfX-=VAN+ouyW){Z
zZ`Cel_{%)VzA8Ki(_@~{1M`I^$_Y`cK$=UYuB1XrEPp{!e<QP=X}Hz=@6Wve{{`?L
zXh6pOk+c0R!V{xSY>W;l(YUfn!s@8Sf9Spm^Muor#M7izllxlnj}DZ}MN70AM6RKQ
zTwT2h@K2OCHD%==O_^*68w%|mjJf(X^A?e=T#EFZ{=+Ix9+~gVkv$R9>Y4hBj%LE5
z`q@_AKS!h=dn?9>S#gLw-1Ako|EedLCmbWl*1+y7+;-yMzh8Sy932m*;@l7=OpxAS
znR(A`)X_tumvQMt8Fx|6hJ{#~AI~|7M+Dy`67gV%(mUq_47wP*8m{Fc-VE!Ai6lD3
zog8Xp@mVL(1X{Al@{4v52BM}@P$NwN0oWrt(;E^SQcTI5ql!}FH+t{wbH6#0t%=1V
z8ZDg;<MNieS-8x1<tn51D*rUlBfB6V5!tev%GFcl@2?w$80aO)kc~7*XAe`e-aFTp
z=g%j(L|FYSzm2&(8Oy1zQ6GXeI5?kW@(7ajaG8Vc53lba&{|iYDPP*=Z(i;`(lU=M
z3Ud*!c34beO(*g^R2}TwK|F9O$KpYZhKu3)NaPWEq@uHjVPq1X8M^zRIN&(Rvg!_H
z8q}l+Z^$m>ZRB3acKW5ja%M#NmU~Z2t*Lh>At&H9vVrLGqq;{mLMj_H>4M{|GX5}X
z`uJSUv1)IdY@PAle4cKUcjx46fdk1_DpNhdHB$yuf?LbD%&;nnE$!(VGo_>Pg=4Jr
z+mf?@qe+sax^TYOBY1|UV7f<;wH>v=$>i`fJc(sqLgB1ND43PJ{8~BXpKpi-2CvQK
zD-Y)Sh>voX7gbt`AZI_z{oFWN>2bp>zSJ1z{d-n{^O~`Yqso1Z$VsK>iDX(ezd;J|
z{LXK9TjT-m6NLHx6P7<3XN}#4;cvACh0pyz5Ki7VnhS`>H+h*GcYUx6K5SG*xsI;D
z`DKdj**b!T5ROFa9s4%MM69l_isr&C5>=F5x02Qf^rfy7CAjuDz1Op&phV^!?Dun=
zrLwlTwjdzyJ&U@M>!~YhwTh5cJ5;T?SU)=2J2&353g^0xPw<ml^3x6CiC-B$GjDm0
zw?moA)y|Z7lPH$%HKN@;2k)*2?h=Xs)bjiP<o1E2FQm6rRk*!8OcIWaxCf~tWcDWf
zd`|~v{4Zoe%nsdwE(f5OQW3DmiDqpq2|&epzA%ymwi45qM&!E#ajEKP>J_wDN(HGr
zBiKu+^#Km9pd5K*+3M4QczLhn55aLEZ6ABBq~b1;6$y1Mi3JXwV+}gxD30&sp#w#<
zvzsRq5^-+P$HI4_@o!S(_1KvPy~JWsG^w#f12WmKd<sskh#Zoz6k<<NaZ}ZGjxz?o
zhiZEHgx;qh8hHYr7(&}{#xRGXF}MW9^#O!<f+D_;MjeMOF|}$V*qI3l5BS@%AYRRO
z2l2O?-}W-`pMO?4&T@w1SzgEk9=+RrE)mcmsn|&@qY`{j>;D9BJfdYc<~N?mQSB%E
z&^aUVl;{aQ-C`0-M@a5$kVDVWC~{4XKvLOf5PuaJQ}`2<CB!X?%CGXv{(w=m9pK*V
zV%~0lGdw64?EEG!cD802XSo<z2D4VJ0e@BkPU@=;zXFbWW2zfGHNG9|x&3q`fqwFc
zrot)U$tbCYdhcA`C<bNrh2)tKn@n8k1r_CSmqsa}Vy){-@Zv_p$*rZqcVmGJP~<^$
zM=O?GwqsmnJZ(2s#kgrg=sMvja_o3LuAclgv1Vqxja|-RNy7~KjDb(O`3}~3U;5U3
zxUGKVnl-*@UB+(pb;Y5S^pxCy{Cw%M6HeWrlXdUN4&5lF9xmaYt3EDmVzylBrfF|U
zVY1|zeONiSML>Obe?#0rJTWJ6o;fn5)1N-EBWE?2p&x(knKFU!g!h{{{`IeWtHXOS
zr0u|RKucLYGBAPf)>)m}v9z&A<dfdxMnS#4c9LQMU%TQ_N2NKu(NEKL;`glw=PPMb
zwumxhDARuZA=g^EJ1fuFzd84#mDXvmr8A?I%9&yhl!|iby9E_w1F7Xepm{ZP-H)Zf
za^@H5U}-@9b5xSDTuOOuEwhU1e&o3%Z7GfNVv^dDkK5JKcZ2A8d@gxOTSZexb|LxB
zvXZpQcn(=oQau>cQ9URtp&d@AsH$-d(C%(O?9S;(8(dS-Hh?dedPh)6S!Jx8prLhh
zQBC_f^phXHT>6r>nyC&_ncPS8yRTd!{obgYwra1G2BfNBVyy4;DfD2KxX|DZcqplH
z8BBhBNh`;z6KNvx8MY9!T3biuk8UDy!Tn$3&$0IR==#?E0TubXy^_-X0h#YhR$(Yq
zS;;F#*_8wP$1hXV58fY1P@-OkifZyur|=s06&CjsHbplhcUR5&w~QpF)0Z$0KfLPb
zCZQc0h9ojfA4C=$5VB%5O|r#k4cFh$QfC6Y#o3R1;;R!74UMng5Ty>P&5CRf>ZD>|
zqpCh>e7!(VDyLYk)zWPaSuYsaN3az1tp;oF)ib(TQkq^0^|41S`{&_GJz$*0n(VMO
zCes;&HW}MgQjbIi>h1kGk8_9W&A_5f<;+4-rwjKssQxiNhzO^^Op_{z(cnPhRILb-
zpRM{2I#sVVr;S;R#-5p&P}^eQPQ~9xnL3^pUL_FI?NT{zrU^c{UJ^>s7MP{IDh(4d
zd0_sSi22JIlIX8Y2a}h<+(xA>$@I^ee2NM>c+8YnG$HUP9c@fzE$sIa>|G~OlL-bP
zz^F7pO#J?miOvS+6nbzvzcl+5UQKE*&Px8*x_<Rp*xV#3)N(%L{P$Z&1r+;Nqa1si
zbx12OtIOyphlB7ryUj7fYM~HzL#}BNz6-LUIcZ6B+5(jMt|%(399%%aSuT#1tCB~N
zS`3J5%*H<r`jKIT84F6NYqTm!*#<v*{ku+cY^n08thIc`YNCQ#8l@*Ios6NafQhF|
zHx;P#WJiX_+CX7eZHg)Fg)zi(b;=bbnTAxOLL3*Dd;tULR?i<;V=zlh;)BUFUhbE{
znWPdS2%~x$Hs)a8wg@pytQa+M8Aodo45kWr6O?A|SgI~o=Hh2qzMemSI);cunn+8v
zxQm$+xQK6mLR`dcGzUtDWMkh0CQ|*mB^0uR2*fj_Gu9IHhNF?Le^&p9Z0%~S%Os-E
z+KR8gtgTlhXAPCLU{hd$Q)Y{oSB<gBlKtT%yIWrx739duldNjZOySDf_LocIfYaS$
zSLY5jQscOm%|Bv~k2Y00<opg&`3DV8t(-6kJALpv&JlRHTY3o^_!)S-QR0Y#7irom
z!9^_KAx0c7hg;}ffVPXBG5t6!K5B3Q+TFlUY#;iiTdk9SuwW&Iw4e+_iqzf8&4?0n
zs7|fZ)~t~qli;wcl|EbiA#c@=Hd=TAR^vQALPRlP)sAJY1&e^Bt1eINRWA&2aNfQ!
zyr0ZHC{{S(Ol{{>j&w{+^dCvM5{CIB$(kE2-?#7f+#V&hf6yx$3G~0ZzOFMAddUO5
zu%dtQpQm=DZNCTO)xeZx9E4yC62$|>o+pONy5jSQwsE9N=Euy8)b=GW5QfnQ<k{kW
zMn8+q*9m7ni^?a7-^eW{Au6BYL=1-f>Evg^#ZEj2N|UBn$gJT?)BlVbE*Kk?3o8-L
z_hwuZ<Q`6#>t@l<znVk)Lu#W-oE@RV904A7(g0Hj?pZJh_9JZ9*P9f(j6lqpEX#>F
zjXvCgFgYT`jTBo-A{K!ZowEAT)WDBe595j|3+wqXi7&x0vH-m9&_cHF)6574wz%*o
zM{l_uV|zZkNmM?G;l{@1^_=)fsi#W_rN!1CkUws<emE<51@t;ntqrHk_OdPa!E`kT
zah{L!cm8wH?D3wD=L=BTe$70R(vfNR3~6KQjSF{hlNr}pBi3}GlXgj)B0<5XMacuk
zE)tXFh(qrawPwclV^K-Yw+wsh7%>`a8J(7^_tjMMekrVRTH~RCG(E<vBArF`3!*k`
zQLV|H6XFS0`kRgS+2=;xQ40qpa7a0X`|W283+XV9!;}YHdyD2pJ@<AF_k_pCf?lld
zC^RR6UP<mKnwik<p<W@BYoA$l!T}mGYOCRglAAM0t84Kzj4Q|`w5kTzA-s87;aP)P
z;Yp5U-Q!x}%14kDqe-g{ZD_jg)q~UyMdMeRN%VW(F`<KFk}0Vv^mpizDQZcrkK1nF
zdq9QZ1-3-`vancsQYsC2u1Ma@V7y-;k+zyZh0dRy7+ys37Fr`CUPt1onuJ$Ze@u~p
z$H>pU|J~G$Uxw9Ca~nvRh`kHflVDdW;}j(VYmLzfd~m~&xn1@LRO5gQl3wMGp{lZG
zxCFk)WQep;1bgT&S$*{78+gO~+DhCDXSPrMkKaHo+=|bQe#5U2{S|3|T{BH&=u-6F
zcM-${H>2u9N2cuZcIQdb*AIprDE^C}9G8Kw$v)Zi-5$tywO(K6eqXje-^spMuD#F}
zu|)zn;4>$%g%jAv^M1kOrQ6%{*gMg6I|J6eg1F;NRpjG}q$4z#qHw$K+P;W&vd67`
zEBnm^?B#bFXxZ4kIoZ8|c(<l`pF#y*6x^zU=?FYx#q_ph6DajKLu3;K+4>x4J!OD(
z+ta@n`JF(#2m6^qzQ5q|JUp^JVS~!r2G+O^<iNU}chQ|I+WOTr4X8rChx_qk-im>J
z%JmY!0KGwbD0XSUV&H&Vn85q0tn(B|o|i%SCkaql9-pq)L3y|R<Mt7P3nhXcpBb84
zV0QZm*Pc9R4>NQu4A2@B86KG67wQD8B6qv>mw6Jz!*A_T_RdAWwhjKBKdYM=`WpuL
z7c>X!7N_d=0*vo@jr_@R>voMC$clIUC-LogR~6FhcmeU;6vC?=l<#Sb?!3eacu0}H
zB>P?0$ucm4?IqpY<rsb51$O*vub}m`!0|R7nZ1l)4T-=e0C{Qv`i8UIN3NIGRkvo_
z5#ze8?fSZdYpJ6qwBvbkYYCPB#qI=8|49baHE7R8=>Y)K`PJ<4^kHqM$=itxy<IX;
zWmcxiz1Kx|ix1k;*&)&;(%YfAs|vA6yQ>Wb=)9jM5nFVH<8XlRQ4e)3WG}xt4a$I+
zglnyXHTk8N-j@hsiht=|d%Tx&oJZM<A%_yjO@IW?u{d9WFDa<fdx)Ca8)^?40-`M-
z+M5be1#OcW+QQ!Z1m+{y>kNjZGqD)>U`H1Sex<nxzE!K+OX+}r75peFzQj^dHWg<T
zXj#3Tt3#q!h>r%gKQLvX^=*2D0J|E>{05<gXQ@)N4N~{;Ye|jy(mF^G6x<OdE6@kj
zyD6pA`(f4-gZND^)-MOkZfQefsbYQCTD4krXBkfQ+D55n^I_a$bqj_U%txm8M@mbX
z>g2#tNBYvlhTe+i#gg{a?XpL!6^_<A%$9ALY6oQJ_t)5Ese&6PuL8e6rBZQV%5IDc
z#(7`}%efru!fq}`<tblgbq;Q0XS2*Ewl{-&uYcO(ssCwK9T^M8?Hd9AWXE^z?n7gT
zx6OgPAs1}TmAL|AD*)du*xU9{+l1A;JAetA;k+7@yiS^nht*{39}Tt+VTCdq0aFF0
zA-9sOW5x$|;cSLByE<FV24^O48RAlk)tj!L<<}I}tkTPAT`L;Om039evkJoj2WD~C
z20$w{hY}H(`p5d}$uH-4X9=3vpwa&DNe~8q0sDXVd}se}p6?ti|Bq}!=Vd|LeU6{Q
zcb<GA0J$<py|9mNm(x-qI8ag~ZsuR*7Ou7|_8pay0bsbVSL}BL<CiWiXvzd{&ilHX
ztDEol-n`W30IH}%$K+3#P{5+HzS3v=8U;K1Lo?TvkmLI=A-r>`n?0%5F3SpMj^dUh
zi*AeN5n1oM@-iJ+=T@Z;*E!lws|?FAyPQV=(L0;Q`{#(o3rBir4_j=>Q!!^$ecq?O
zr`OfRXASAq#scz{IqvJv_JL;W?!sfoWMl7v{IUrbb~mIkAWw;tSK84-b6Lm3;CzXG
zdU&0AS5?djhOcVJXCFe~ifyRjYE}9PTU(Fp2><K!K%HkK^G9M-kF)b>^OTgwo4$+p
z&1>UMPl!Ibd9rkgIbQZ#Bw%L$!USvn&QE*E`vw36$a`1$xNU?k4qK%fEiLcW?kMGL
zu>2-!yV+Y~>Cy%8FOaf4^}I5R^75j;1x)0Sd)`!hZaUpbuIo}wRHh9diYFgR2#rff
zrpMu)#T-gNQc+Baqbo?XlT1lKnhU1KCl@l2|BF8+{i6{_Yc7$HN=mYa?gf>z%M66b
z>7xK5@b#JA!tw>m-eP+N%<TTx5bby94wf%c_V%0BGPAGfg_*P80K~`{Dgn~+4Vm6i
zeTVM;Tj-mzj_ws9drQUFck(S9DtpWME&RPEQqHd9_nHF~K-~Wx+IRdG9w_+^i5#L*
z5yy(7x(i2)k|+q{#}nU&VZ@6ShVc;p|3VJm4?;&t{-YU*II2t?NI7r>I(Q-J#?0(Z
z+;Mtw@`e5$Ik`Oo;&$Wa{Fflf__s7?KG;BjI@YE>YK^(u0RjzqSYg!q*p?kX^NIKG
z<iHsuO*v^!2?^$Tqw!>1*WFKs*OL$}TB-4AEUkq7gvtG8+fZ#XaxC<WCt906Kqur3
z`7cG(k))wot>tteaJ+30L)VO?;i3QafnRaQpoY%1akQzZP{j9_$kq>5T#143S08|K
zN%aMCx7+0~vkk{q^ItdqwC5|IpFjBT#CyEbe$KZT-?PW^qfP(ccPHonxjPvNnV1<_
z|C5Hu^o>SjWMKc#(Kku)e^U|}8JU?l{$H>S(J7E}i`H1D60m1BnDN)wHiaEUm}&-C
z$weL9!tILfg-Y%5XecO!iVA{4>5;bQZ(C`erlzYOIXh=3nWk7=zYwk2s4T<C<nX~l
z>g?+r;lPcQju2WoKr=IWE;BPXPo12&2vbRJZ8^EJ@KMg;0{nt~=i>aq==6}8)ZteH
z^Qj>a<Xjvf?CwBXfx&GML7beRo0*wHzBEF){UGY_u5qjaNqD&kpy1kbXGVn6I&yJU
zc$Y-eI|QInD$*eIL`0y={F9&xG6IyN<6sDKO@3>jS@^^n>KtH=L@I=Y-v{hiQTZ+m
zsSOO!p`oGB9l`%-xFIZPrU&6}LHaI1Gz3%$7xByxf$B^GQ~j_&RrAYJClEaDxQiSC
z90Tx&T0byQWH7O~A1Ifvdizk#P(q-{bD-;z)F95Lgg{y2B>o?2puVioy+hMau^l~?
zzFc7bUkuH)^#L55ntojZdYXQSqo^=)DXRLR?)_?@n0gmE5tuR6anPRduJCaM!AF7p
z1ZA2ahea%)n8-gqm2)b7U>qVi8oNSrepbdGlw^dmP54(C6UxX4FiigVF!iY?q?|p^
zyV1e*#j6JYxep3>c^iX*Xl{N{4{6F^4yFHjxDTb2@F8^~81z{)i3tOKYY%4+fBOLL
zM-P%)l7u%h1Hz><-OEq7B764-(T8su#nf-=54He*9OCUB;?~GgKOT&q;Fg~cUr)eJ
z&TLyN$TA_#&t5b$c(8)6iw{AB#+Qeez4tsqSwF^;0iZo7hi~`ii_t5C{xV8!!Aj4$
zFD_$fRAP2{HtvKj*%x?G5m7b>0h^5u2u=$MKZRdpN-P#3VK3tC+twUE9kAn_zwj4k
zZqNMNP!ZZH6d1~DLqsUxLp%IUx1bNWvdsYM6PpX={?m-m?{4RXeedv~(E~W}^wkCF
zoBF;J@>P=Zb@KkjPU6lkHp0(wkOTKMK$j^HqvLDNgV2a`sn4lIgiHhWSyRq;I}@z}
zsS&uD`6-p8hlwmj4B71bUJ_5J>K;cis{sztQ2*u_^T`(q{TxHcAP2q=_v(HEvf9eQ
z`89yQJUza*r(^HUQN@pPt6$-ROa<K>p!TK3Z^P%~2Z>>^7yM}lM3(geA*c@JZG!Un
zWZVWpa$$e~=z+OW&LHc7`{#f0DEiodgpw`x!vS*vAfIt}u%Bd(5bM3P`d_02#pk+h
z3j<Bhv7leSL5|;JGb8^!%kF{%eF)DJZutJ3`O5XhezqOG$D*;B{>t~o-tnFM{F{06
ztk#qK+R?M)aPkKI!8p`I_SJKV1>}CC^RE8V?>YR^9nggX%696~0AnX~rQA3$PeL;Y
z-ub&uKJnhAIDA2S$fwDG+`yFEhfm4RziH2^k-GoTfKs15e~(^}U&!8T4qG~dKL7Yg
zMBT7IN}2Smy(gsa%7O%|b*T4M3RX|YVCsAkSNqQXdH8x78hV<QEDe5|>9|<wgdiw)
zb#`=kJy;q1;y8GR{5+O9C5plU@{{*mbtK_D5;?9CxZV#K-#$MO>~tY>BQy%-<OHjQ
z2d(!YS$CBFBYxJ~_=WW*-cpn^J$THFZ0*|7FC?K~K82*?pDk6?dDEx98E44z?_Jva
zqT#C<m27`YYHHbXC*$OIBYBI)+zELY6d%To6zij6zI^SOxXs&kJMbKmldwDZMyYth
z)gFCto^r6~_EhoK>a<4fCt}Xl{ATr$F929<5J4xyzD!!`lwRG-WEyffM)DJfjGB2n
zo&xZ=_9E8ViTu1c{#$MK18Lt)(kn`(?cld*XG*zKt9FNYPQtHvIpXZSXqovp@tZ<w
zbl-a_8g=^g;_w#xZq+yjz7^O2&Q@dQThuZ`GmzHqSnP%s@tzbJ&bSae@>u<tb3wUS
z4B>!?ZqLK3e>uuyr5gNy@&ph{*clh^@|<h>ojj1uJV{3=Ze{v>i$DU&Qb$^QNNE)b
zlS}wR#Z$NcR`n`*p#Q}Vduw%jpZ;hMMYZKc6BNLfftAw90$=M3lYrVw`ZFtaGmc>5
zj|R@}8av51xjO%H3$OXQmP);qQ3F9+zx1j=_EABzfT$Fjm=VbQi!+;8#Hn}|S8Hrv
z*)8!6$g>FsnV6Wq)H^`un?Yl|J?q`qCPnaU<i51fAL$qQluV)isjM6K55U8;Rj~|R
zLUIk$mJYai1tjh6GGrzq;*QSZ#_hwJxV*iz!WcUq5T{5$MTtEUO~>GvM1}4M!<6s@
ztO!5ZE+Jyt&yI&>f+I46=EY)$@xVdxS4ngWH%ii&$`NPsNOzwP(mYzgmfx6au(>Ky
z_a{lF3D*VV&gNzME4HASoqgAbE3sHuwuTe$s3r696Egk<zo?VZ{pc-$F>hhdCjTaB
zZRKJWryJQ)sWBDfGn^oEGbe118+ERNRwb<(H8er8Vq4}?VK6&G>x2>UihW_WoL8}~
z)lD%-TmL4DF(&>p12_qfy@eX6xo52`H1?Os8AXZSySv1A$i6WOX4B#(HU!sL?h|eA
z8*N;z3h^(JH*Br(`MUnrFm9ilqe?I?wo+EOo3M1rE7kn1S^M}-`gxE!k2q1oTNmP^
z>gJ_8Ip-8|PH2cmyg5Jrvrd)vnW^~u@X*67lh8Q%bq|2c>@dp_4elu=V<~NPs%FV9
zLyCPd2EuRX>psBj0E?Z@XpgGNT^xgr02$e`GaJ;KezYj1o!^%*xFNR383kO-&)Fk6
zs2}xK?@xO-VSy4QhG{*UxTq>q(9W#1Q88mrti9A>zj?G7%K=ZSUL$V8iKBEpVH7Is
z%g`0|_Li(HC9&?@JaFT*to8)_-C80cjY%<xy5$o_teB(ruG05oN7Z<A^WDNJmJdA5
zv~vi3pZ2Q3Q}Xs3Y#rmx50Qquu~t+7cCo3HTLXjZ1|$6@_B?wU{bIo^;H5vTYMF}-
z8YDT%L>q=kl?^Yy5x;L4-gp^I2QeQq_oz2T+Bxq-BXYQD>euNq1LzgbX`lOohY6q5
z<rYCw8C9D5QqQP%P1CBHc;*}K%o(nE=MY1$%CQ4>TyEKPYd+>6ZFpCcT@Vrfa&*hP
z5T)J|9sE_FB3?tKuvaqK-bSs6q=axz<dpnQ*JBkHj^&>>)72?33J7$i?sp80mK1a&
zn=UTauniV0w_cPM?4W6Bo39uyD+FJtc2+3VhL5;0_>qZTwD{Wm+cjOq62&%^W!RdV
zXJ*N;CB|6=T89@71Dy+Bt*97pUtXRbUics*$+Bdqgr$Icr;IWMOy)UL`(V-+YgGWT
z&GC~m){OlX5YVdFQu%|KatMQ!xOsju(tMd|RM(5$FodNmT&0Ge59E?L3<L6oPWkP(
zuaPi^>^$$nq6iBq0<)iUl3(EzMOF?kt7^s<xAph$qfvbn0MS4R+s3&bfppf$$c52F
zg%d6t(U1%fN@DUCc|B||<%!s9sdQJiKTV_5p*V><PCq7@HYHU(?{TGmIlgCskSrYB
zUNFe^o4#X@vD5Q*rMHLyW2E{AS<Py;Gdgf(JiWn>vUpT~p9`R^lBubBi!O|*j-0lO
z8IW+%j)bq5pVnD5{<hgrTW2o@mIjIci`1YK4KtAvzO6evNkZ7`MOJ?Fw1SPQC(6vE
z6EhZLLt;v#gcufdXDy4&F07M6Fv#B9dIgQxnIH&!pVO|BEVzTQt>7W1KiH4>wHUbp
z{sgcf9uQDp7~i9=ASI|dO;8Z`Ht9^BM?KsVx6OtrSedlP{=NO=GttL33^CIl0_wOT
zy=K(1s@PLvs@Zp%Q!@U@KeaU~&y<uUMNnF)M=1^C&G>nxr9ve1CZBV>P$ppNb{~;8
zD}&)P_x2mh`IAyi*2tCH+SH*9MkSnZK^&JhfPz)faz8SJ^QsTV2x{#+k&hg3UOrRO
z&y~;B9&8;(gJ`GRfrVTP7l3^y`Jo^C&TIJQB2zW4Mm_nXO2?`looCD>O`veCc`Hn;
z&}oneHP3Ms+c#1I2RGx)=N5tvh`1GC;f|c3g!X0Cnl3A_in7&#vA}`0cCm6yys<C&
zW{tQZ0<PIS#j^HYiJF_$w(S%)HJD3h-$)Xk!4O2TLzATNY*u!=5Fajr&m`Xha}PQ&
zu|%06Io4E(@Z9!Pe#%|2q473iF<<ldk0eaYYS*^k^rO2+%^c8+*f2Dk&wl`mXy?Po
z%}@i|Hd_{ap9|3Du5F8<0X+=h^{;?iS}<|0Inuq<H6>`Y^EmGi6l3_!$JyjFT8K`#
z@IvOiz9%sUk8crxIBF?jLboMbdDrkBpOJsq*vZ>jdR&)X4=(0?i=yAAz6zEUYeC6h
zWhboR_P4lfB;{zcffl<gViN<(a{;XG;=;5uezYV}_o&cgDB@xySWIcDgbu1$?Ga%c
z$F{THNIT!DKGoC*=Whbb@b5$u3~f^`t<7~z?0rp38P<2f06$8oV;|Zc#ZyLhq$87G
zskwzQ`RPZiMjHZ3t~DorE=R7CAK@rtVKKo{y9=^fL%&!!o!r50&Iqo-pQE9y3rF9&
zN>MklNwe`xI4P|kVlKrS+?V)WTT5zVY|15R=2sy8{CbR1kG>xD%|WY}KY79EI`XPU
zR7x~Jk3e}(-0T;nxjWL^S-O}xor{HY;0^{kjWsVj*y-dAmf<+RfktcUS~wvU>HYy*
z*)eS$Cl{qiVsodC$FnKgNHT*XC#hNJ48iKNacAKs!4v!Xh#%2IybeP$6#<3atND0L
z;6vy(m_^~S5tj~ut2K&%gc~U-w~ml8NUQrH{1d0c_%-aFSwfqgCjWg5d5>!-psi3{
z?2q(|lE{@`O%P#39+uIZc40pi97G6CNlV^MHw{a#Cqci+%PCHuxg)kMfVBXdhiU+E
z>4isn)rrh%>!mmw3aG-$4qVW`va9NVOPD@TEp(&GB5%#`4=>cgwd;ta@bsW>K=HOE
zsc`n?<4?~*z{_{F#aT?Bl{-qH-Z#a+;4~$@j^)nEthdguiasn69*nm|n-F1<8Bu{J
zn?_l(bg&sz+<U2Wz~GapUeV38YqT1moE~VHBmOaojY+x`nMAlK&x$(dz1>xmfH~`X
zySk0RL|`JNz6$OA_%wjwSa&00n!&44Y|SJZCypN$JU3zT<EtVoo)#BsGv0Z2W6?EJ
zOctp{lW{Y`p;BG9)xwLuLBIZk8XGk#H}`Y$jTVyE&avZyF5M+>P5ypO35}*{IqL7e
z;6s*Htc1#xHz-W^b+~7n&3aha?J~RJlwnn@rjR2BlHD@iUJAZ8)j+YX06wsx@B$1&
z>(^DwQ{(cP#d1!DXgg^+L_p+uq4$CfK%wVv4>G{)pMxU~+Vk953hNIy*PgG>QID(M
zlOto+bBwbPzNlNzuySS}H@;Ix6cV34WK47w18Gwd#lGa}?f#Y}e3h7)qfw+|oA&J}
zXllaQ1YMu>Pqm$`g{T(QEGiQVO23gzJbrk@xlZueQ%iaugkV69a^^;VjYb8Wyt6mN
zKgJe|PAC=^xj_=68d1&a+aQ%|$({${eyDxDOeqKRwXzKo-5}Jw5bgALae{w<`ukqV
zXY%pBUq73ncR@1J`Fxb#N#j?0B3Hta^D_d@Wp_Kdsj;N}yBc01;=0C(^yJQrr^<iS
zRGnVh(A*sdJZ^VL6?p%XfiM>;A7|j}*u^z-<X|h)lJ}v^Yr^*F5fhPR>E|ge8-kUi
z3~(5Z7Q7eE_eI`Xe0S%SZ$tHi5RnUn+ct82UnJqppb8@v4wIoINCA>ga4BW*>!mLe
z%Jl!{Nj4@?J?hd=yN71}Em1wOhevDC%U6&Udu>^t0lRCdP%{2U#U1Hl%4z)6e4HWz
z%m|Ga@N><>r|+a}^K>KT^gKEyIqE(nz!0ugmF6k5)A{lvRlu57sdc5;IRB~U9}F#-
zx9AR*l>9)0rI<l?3u<hzQ&t5ABF1frJ+1ed^!3$<s%j;uWYrHbLd5Ktg?*R}-z_0q
z3^%R$p`D<=`mJlM`o%{wu-2PN0yd;79jQwh3h9~8#}*z(=E(&_P27BrJ^Q*%-tlj=
zp}*D;lMB7S`Zp<!lm#a;B`JoIPXy{RT<K2G2S{(zFpJ%X4C4MENl6_sovkw=Rd?O2
zYx;|fT2$nx50hYZmI`YbZ-k3=DuRY~REY4U$QtwzmgeCX&{Fxn@r-nPdIdzewQ~tp
zfiUtt_q*2k_EjeqYz-%HPzL=m<FDQ*HJCzMdz>GpYqKsaoY|3R$lVSpUs2vubKZ0a
zFqfT(^Q@BX9FE-Jl0-QBJIY~RLGLoor~rCv=X-tprpeJ~Ct)IQ5?(xIo!$^63;?g;
zQAF%;GiX{1AD5timrv>4w%w5X<F!#!bD?yA(*(7`X;6C$XrGhK!RgGw*|EA>4Q*!>
zuj1$cy;Bz*okWx$)b0m;-9f}y2UdOOeVj7%OY~pzke3gF8JO7aiES1@QRLw18eepF
zQF{3TuiS;zV!PB4HUOl7jIBQ}MM`ayNGrSD+3E6da<ow8(pUQq=!lV<)GB%fsh^|{
z-D1*)nz@KFMnK&E{>keu8$)5%s$ue>Pm7V~8<8ov;Jn)+z$rEUT};EKLwr%M{dKt{
ziRN9MF?-EApDMN7T3LDvms`dL<@$H1J}ax_3dl#eXz_rQcHK~7epEt<T9|%*%_kg1
zvJaq(95{BOtGVTwge;ipXSVN(`Mu9QJrzLTm(yuqYbMH(%idmBgJPoNePn#zqG&6j
zAR^y%jWgihk%eK$9$7at<heYmZ-=+EZy_(%|8hu^pmgM;ys(Rnl3ur*dH6GyvMK~G
z3wa8CGEM8d8h0Q02`!>7@+@Vwp*~~gZ&4duQL_*QgYYXFPBDJuLbsi<a+GWH=}Ev=
zUub4Zvc<et(l~J1G9-j@50JF>`p!pOAZh+!NY9t;?)eOJqu6mu6_KW$x0QUDetN~?
z;IHmgz>GccAa<*1kLYhJw$!0Y6QpjgkznHt+i#T3<lD@w0&|`hcoMxKkIa9SYfJ_3
zNE`n$64H{2L{v=$bos++m*~^$>+gH6QC^H&s2m)Lu`9E(qTUB{dA<w@#y68dsq~>g
zWnJ|T6{6(zg<+uQ)Z{IRJzKjtT??5aFc2q_X;YgABn+cXePX=E(Y$z`e4Zr)o2v+Z
zq*i@29zS@$v=AIV1kCHY)R|*WoVENN9q$5UA39DI8LzmLP6QVWVYg#wEC2TTa2ksx
zgBb7BtjfWS%W#gtZPhBUTuZYypjJJ;j@aLz$=e5U*T${UQRMN4{*@RT83B;SHpGa!
zKi23VZA3cjLVFYtG@K#gu1D}n&=z|}SV9=YTBcc~J9ryNRCt*n2V=DLXP)flZ+{M!
zc6gN&c95I=>7mQI)(FOGM`00p0P6=w6$ao2U5_A*hRmaOZas2s^O!z3zoreM!q5)a
zK<`a2(!U!4e1s855)H`=n44a@1s-P??x-K%2I@L={DdW0vt8_24?e3%Xk6i~K*2_-
z#B(D{ir(`OxN$H9v#0XJsxghNgv|wCy5)c%>*IyY$;ympKzWQ98lPi&&cC=g({DVX
zsj@JGyZ#4}ZL>)nAzK6G$6(?nL7O$}VHcStumY-?Hh}l<^4D?A6XysycB^bZXNxJ;
z-tblpFYel3f5sSW^ghtg#1yH-4;I!CbMtNW23b*?KZmNgjl)doro%0270gq3%clnO
zNt?8EHofGo8aX7J;*}C9@r)xu`jKdK&d?z>jR!T=@EnhbxZ#Vg=(QHo9p2-Y=yM(C
z*7FL)Ibxh|<hEDg*kz`OOg@I+&gna<!hd{6LeFy2cqM)i=~;(z1<i75nwPiqE^#p9
zdBja=hj9~4a8%oFP6&c13amhrGPCL9yPcfyb_kSbMQ@x9SM4(9hv#ANN2^jrtIvHI
z`IG5VE1+M>kG6H1FRj)7p-R$7>HO)%&|Df|`D2mVl&D@lln5?sf$dNKxB#`7Zf27n
ztF5?j*!jR;WGrenPP6N>-koXU4AD_v9Fk*vhG{^SzN!lT$g;v@ywj;hmI2O`;i@L7
z;N#jJ4LI{7tZ$|BE5x56sTe`Ha>?|L(+*Jf7i8!kIJ~zMt2N+JWh1X?_%uV_aWZ;;
z;__Rfp5tAv40c3xw3cH}ukPe0)dBl}5%Y`}4ULaH&v<XZJZD|O3;yRj_{c<ewv&zI
ztRowJnqk^?ppJXn&u77&yJFBtKdd@kP2kSOUi3{?0qPYC9Pn&_y5(B(03uUn=EU3C
zxFCf&K<DpE;JY@BY!92U2&_h`P3qhKVC<WMGy%F~+qP}n?rCG%n6_=(_Oxx=wr$(C
z&F{|MxDj_F_J7#ttf+dcbLwPPW-AmgAOA#z`LmR@QGi$)7^+5@hJWe#oT#ITcZg{V
zy6hk0tzN3@A>Lp=c!Sp~uJF*b1ioP<(9`iLo_CfqCT~Rp#|Q)eO<`+5gSTZ!W~byx
z={ip=oP^AR@vZVc`Dnd7E%$kEQW$JMY_G3)Fi;Z7sphl*u}3EIJXRtmv~u&?KGy@*
zOy5P$b){%4dO0Y;N(O|VxR!q&_wzTTZXN}m&dprrWT5P0d&Ek=FF&-a&2-RVq3H#b
zcU!~t5sycujN;}U|8e@WB|plU*uYUz`F8BA6>S~58hj2R(&K`b^-qCG4w!pOb)@T5
z6olo>OZ$Gf3Gfa_GVijTMP-rp)WMWc1B|S9KqME0@)Kx{oo<{v95bO_5jxrPU+c)Y
zTbm9R)zNuN$yqm`E20ag4)f#S72PIXl-A#EPE{tZ`<aJ}OJgB%NK;;(dm|~d$yU|c
z_k9?#!{#tfBiSRmDPAm25kb^6yWCk<>;Ts+(d1K(*@{^Zok`K^t=}4`lK+}?z>e>p
zk;e7rf^l>K5)Hfj;nd^k?!VTo6u1p1_m`96!bPD>TdMJIXa=_jOm(kb@5Dk!cR1xY
z%kff&#KZoHoW&eLtIdxx7P=Q(VO{s0OWFhE+v+;po~9kd->@*x)uP1h#F=A3{pQ+Y
z5~0^@mJwZyQpw>SA&jb})07VgA{xlKzT-ILE`N|H|DNVieeS0nvoS`t{w~LMMC6uU
za^NIVGL#stQGCBjsMW>jL&HMt(be_ReGfTYC|*4=<VNNA8p_9k-HJ7ik<R?IXqiR)
zepm|+BXo-OvygxD4&0Pc!YFre832yW2eqNeuowUP9j4O>`9^C2{`PjqTe~Zp{RpW;
z(#s;&jaN;0MI6WO-%F;GYAQ|tzL<s{v>Cx`$|KC$I7O>No{+)u<^(1U8b&S%9I;{U
z(BYd72kX~YkAueOgvaerjoo`7OOQug=yoP`F4LDwh4kgogpJF}bVYNGfHED-jj|{<
zZ$W$LzE<8(6S2GPmW?UcLrl68xc0cAEaZ64r*zkda3fjwYvea1WHO%E^JCP_%RmQu
z`%;8Gk#kCI1=AGUIkl!%ua?xk|CKZAUkNRiY3@RTTk$l;!1bjWV*4^lcY&%Q@Edbg
zymztheDLN?cG~enCMT$ivCXq=oPf$Ye@lc&!w#EJX2_X`SkrFLMu9V2-jE(D<RBJ|
zBRkYrJY$l;LKeM4BkMSyy42TjAqRrHOQ>>*U3Q(k<1!t@@44j^3SUp6+y(sjl<V+%
z`zzo5T<UMxMeZ-6sJEI<TNjpB@L2;)VnH$HY5Na8p1~7rQ-||z4dp*~CXIZ4j(3hV
zYxQ!`{L&^%%F3_u?lepYqAsn}2J4ym65jHV#uK?in3C)5xMp0FOMmf1JxK6<NrekN
zTVgPh^jU@CPiFuh@!QXxQHmwNq1DZTCjq6LcybX>udFm~%G;$~Jb~YqZtlviI;tRD
z#$A_3a?;H^2`N$ZL(Yj@RI%AOp;>ca%!6HQ)=L<Px>YetlecY8744f{O6nN=?+F0i
z=jscSOO`OkjC0JTNKL_CaB|F;Zb><e=2X70PQs3pOx>h$mw8p|qlcs0Q><sG1R5(Y
zudZ}84Qvs1xZ^P<ATCf%+I?9T^R39fe==&hhHbX3Wm}^-;%@f$<X4j^5pw*T0@OC!
zvk>3~R$JU3A7mckAejRKO+=rao=BS_=%nUf%AH_|3pGRn`bx7{AI8}yC0d%`*%PS$
zX8&D+iwq$24x7eJTCQ;4SaYD~fyQin#-3ilTM5HhzT|CTo1iircvE?YD&MT{<*VgS
z$|tCym}qiS#$ok4_RZV@QrM}T;M!^@!Q=G9p!G#^T6ziiO0ss96Y4<Dy{m?$@0^~l
zDLvmpx<IMg&l#=N04dXBI`^JP6peAh$Nc8cTdIJ7CZtR*W5_qh?O$rBk1IcnfNgbk
zoIECbU*EG@Ectox=a8h*803NY^Dey+yH3NgG~US;W3<?3wx=Iu0?(0vX8GGogQ6CQ
zb70wxM<?X8?ZL1HIrxIQyQHubD^C1!4%jbf!oST%Ske|u1MSnaOQ<hV5z!Zdb<zh1
zQba{E{&<aQ*Xqiww3`xmvFzm*ROv*iHCQ8bnC3gR0LqearpJE#$(!$9@&?5;QfzD_
z(+;>Jbq1BxxYUUzVO*3Df0S>#w>5aJif7<n;ZoP^V7rrQ#O)tP)9pa$Bcf^X@Gz!O
zjK3@7Dx(W*M??Ni8ln8_A>t~$Nw?EhwX(L{VQ$2n%L8fkz?X4a;SGJsXyp0&0Hm>r
z6@me@TA=aS>)*SmlR)?J72(f_BPHQ8eCpEYO-1bf)9i&;3C<r35=Y@=2{)>$yV5;m
z&@1Z+<<Q8X+N|(GOWkSSx9#RCZOAKEMH;8xplH{B)92{;htuWZ7{5nga_3f5f*ka?
zs?wrJyyqAa)sq_V`)mfW&E32cSnF}LOveKXZ@GmJhQf=@F%fit=~lu_8-}i}v$(39
z^}IlkS(&3=T-7;T`nnmDTo>9RSnbqDVq+<RDRKXH+#nwK@Q~#puWakQb2_@(d7KzO
zu60g|c_Q=RJEl8+1`w;?t5>nr$E-~%7%kF!)nWEwGWpDy^GZgvXQ&cJH0&j6MS94y
zEc}#INCG;nSj&wBrjHQwl%5>cuxc7+HkL8i;!*+P3CyC!k`m(IG$tbf`ZCSPSX$JF
z1sEf@JEePek379eOe#@?Q99>Q2uix~a_E#vpI#&{zJ4Sw?#^5cRLU&Li49VM>_wxh
z0p9hu;!z2aa};vEKs}`-^}Cute1Tlj@~hxfP{sXep#sr(_`2Az=(Ll9C`I-LeOJ(e
zYS%f0o~k-iyPR>k#1$_IV`_($Qw6NZYfeE?G-E92Zl=i^aa~1wyYWPY5!DH0GhQ0y
zLpW6Ft2y8|xkpn2sq2(b9ub(0@=2pqME9MAPAoP-Ju4yFvgxN5q}ZI5gAXJbux==R
zrx{nJMp@Sg3&FvS_V%4#3a^Q%{fL6&P>YbTaN*tRa^bq8%~3IdI>{_q$(EKGF*{%s
z+#_@=zF&8pzRw`+J9_T6*%)(E>G(Q5tJUk|zAr_i1&pDZh|v!ebrs$l>J{@$z=m=y
zy@4K2a?D&-&G{NPkX|0Ci>3uc1~F`w0JjDQLF}!o3IY6y{nD!25&T7q>Az*ff~(9i
z8#<9K>#&W<dzEX;8=Az>E4J=-vZT<9Qz0uhlGCZc1!dNN{FsY;+~={Lo*2pD6r)4c
zKNfy58w)~K)D^T152^J}-t<b3fiP&6U>}fu1}*(4_=9_OE3Q;PI8*ehl(MHw_ofvr
zA=qoO9k0I(>DlO)8KJ26r{baug3eQV@;@CI^xLW%PAy`|Z^Ki7UfOLx_+ab@FGtgv
zcWWx6)wH{vU7Uvn+jC^Y#sUvmQWehSrj;ZlwB*<up>FYoP33qcBn`NJWt3Y_EgbQj
zUjd;TGnL*AmofhmJ`wdVB0)}xH-Tm9M*vI=gvWU}Q}QY^6*vI8YGn@aK`xop<~R?*
z9;OJl)z*ni>2Cj4B^Sq?NPExxA$Y{Uy9~PgR1&cdgt<ynEWX5uJEK>A>WL^rf(vs6
zo9O)s-VL9rWI^&rkLXKCSe2jvV*1Tn+##TD9=#izk3Y9|TCja1C*4CO%vKDUdGm;N
z6Q7Pth2~yirmB_B%-uPtdrVhR^fK$T@h3QFdU*;v7z=3)wT<3Y0N+iFJc{9V)JX1}
zc851CL!gD@n8J!}qIEQJp6Jio(z5`|nRAp(y3AywvRQe*N30@+>OPZAO4UYAyyxkz
zAO&9qc|Sn$!;{;);gZI*&{HRVoNBW<kn}}XqzBi%Cj-jDT0<Lrjw6t}cH5<=$kslN
z0<Sq;&TP*)y2VSS0<o#*%YIgC8;m-fI#WnC>|H6lpja|4I}Zid0lX;6)^xjHvO#a|
z&f|Q=tF@$|==e?Cv#D!n&|^hE3*WQ*U6tPKU=e|AMxGi76}MbXmVdkw{?x{{2f;NI
zRAiH7Hj~hku+R(Vc{Vu*<z0jPwwhVg?GZ^1nq~No=)H;!j%}^<`)&u3S|+eN@q?uy
zBz2-`!Ei=I=U$vAUj^HSNz(exxnN9Sqz7V`Ki`kwTueC)^@K&4ld%#BG{P}h*Jzmw
zF<S%|WnI(N*tCIaUFnKDwjaqx<8jf0VE%WYhF7yU!wO}f*kY$cSTf@z&on%nnk5qu
z79?S&WR8<)Q+Eu78YZF9JKjuNKA6uR9T{zQ$<F>T8HOe6xr`y52;9V-J_ajMt-w&i
z04j=KkFr@Go65svHW>Z9{Rq%A*&Sq3)8}^E@1LTIMX?B#G5HpkWLsN-imj)?v(aB@
zJ7Fh%p${^jQWAw)XvNmqnhBV^-~x8)!-VPqE_N?{xlAVsjGEpkvlmlkI)486RiY|l
z^rhk_Z`<HBoG4puI|hRD@8Jsz*PEKXtjwO7){{6Fs`0I|_ozD@{5M2GLiy_)bGjb;
z_AaX`rgg9*<Wp6Wz65z)nKl-6Z#O_<5!TFMC=sA#nheTSQT`88LdPHO*U(15Eydem
zCaTEhyU$+{@;krB$s~UNPEpH$k(1O0ub1xvA%wu?)BqHK1+<QreL_m3apnv@d7GiI
z>d<0mbw>UP8}OQvRse4>=Tt{2D(n_GsI$}W{#shv0_OJAU>_PCJKdDge>`wlxMOX|
z`A`)80BtY@&#a^%XDdD4i~d`YJ`sN}mCpUUUBURd%L>75$YUthdpLrZ_tm%^jR$<}
z0VJo^34$OQB(`Lf2}g$m!lF6$SNHEcuIC>j@{V`yn(il6)CWsd(NbvNZRUHGKO8QI
za>C+x0)P8@vE?I&BY6|r>ZaJ)r+(jj1fKgwL$%X{{#Irv<b_i>5Zo`!Z>JXcSdYNn
z_E%|>)#qM2Xk&Z(+X}&NC=MCvwuzZIv{boC<q{mC{SDI`LK+-Y&7B!&$d2UEyvf!v
zKphg{Fg)BuIqOfv{UK>AJUV?(l;;0c-8!YDu45Kf?U`vCW&Scm0wVb@v{h7a$%H`7
zS!V7cM9CM|-7Xg5c2T;bMH2D|96jge2CFg{v3_uI2ShB25z;Ut9`!+y5rt^%G4jHT
zbXV3?!tCkDo4TXrq6~LgYM8NNxppG<BT^bX$}6K9w4{p6EZyQS++N6s{i3$W3G|7^
zwRsQm8~qhi4Mt^KcJ*A!49czA;D*M$51ev|N7f0M9l6ft3-unp_wg_VC(fO_u(eIP
zM5H`)?_nyoT>U}$6E52-Pa?tn(-3`A>?w&CXN0~dDRU%K-jUDTS6s$-u|^!ufXf?7
zF^WzRO`Fycxb*fJyN@<HWIGrcnNFcXbg_|XUJ^a8TAsd%$Zg5nLu5;%l&q&@w@Czn
z0;(L<>Xop4oZ>WEnKy2_2=Dl%S8YUR5i4W(hHE&A%;Uew{RmM~;IiFS$<<;C#f*FJ
zJgj0p=cIPM%e1G81632lX+xf`Np%(@G;v4c8@6*(gB1DtJgKttIuy+>mJzF($EX}?
z1ulLtpl8TB8>^p$0M?OMROr#@#8K6)pYW!ASV-Bkb*ZmLJL9?pv&b?QoxmL>s7=$-
z3wx){o*O2Fhb;uA5%U&h;N)VIcla$V#avd%#N<C}q!?dy_G6uTO<n<r=dt&D5_c}(
zZYD74|1?(FZC4s}h#m=o7yAD4H5)n^p-4d(g|jdg%gI=_)$Uh|*F0HjLNw#9FiXGs
z;*m2wZ;oGoIr0X9H1cNKuyy*`XPYC#7kxa<10P$+{^l_gOJ(F4Tz}kF(mA&?xzU`{
z*yMn-JH5k;T!_`mr8F#=%TdL)0E0_hEz_9vn@jsN-aUK_@i|~FzT27*MO#)s$<qsh
zpz<?fCE5Q@;mS@<hGU6moK+pB+Pd&E)sTpcU>zI9G3PmZsZTjw6|ZdHBn=tFQ>mq4
z!`Lsnc)}?OR3IKZ<b1rH*dITtV$&$T!8u(z2A8BD(kd01z(a*sK&Zal%a3m8ClmP*
zd22^w-TF~BSZ4DP8brTPs8eIqf12Rn96EbX?og$?ja%{5JFA&{Ajn!`R{6^RwB6{m
zcKRu3W@{lU`0HA2J2{zl``v>ItE>1xdB``CsRHK#{M!gGdk=){x<yU1b>M9>`@8Y;
z);i_SQWzT2oE=ujYeH65<_EE%3Rh)d4et<+9jIvWcKd=XeGD77(AO7wE!N&SOG+?1
zyvmxcZfjzm@g7_vjxW}Ed1YDx61$4K>7tnJtP<^2E$YE0BN^&ITlIElq_ZGtaB&%p
z(QojF#j*^|?MGH`$MyY!4DFze*L$O1waeaNN5{0^FxS{6mDrFEA-|m)u@-!$r%BS=
z&v1Z3KN^d~Zx$!_WQtZL)dDzb_RzJjWIrYKu0K+!FN8cMWH0}H25D;F5>F8aA0?r#
zCyH8+=T2T?vaTOl4UARyi7InXM(>OqQ+v-*AbTwB%aN$_QN2Q}$eXi*H9E5;Ism=)
z)uO4TU`ki?ol-c-DD+692$^;&5&^d<i=8^#NGu7qx-l4A9_Fy|NH(MnmQF4-T%#%J
zPk(o`s*S9x%N2fP@jaMQYy^Zed_7l8kSNuS_@?CTa83_mC62<ei{??>HK3PZH~waD
z3#OFc9nz09)U4H(Lo#ctR<~js(@<D7b{k#3e~H!}Pzw$N%=hY6`}~nDdsveN+(bSW
zD6RKv#NOyLI2HhyP#lLc@4v|Z92@|E(oh?`{-b7@i}`<QmYLbu{=a6KgPD`<e@d1)
znc2AiuVR^qQQX4X*~F1i+}gm|MAXE{&e-JFFBm6hM-u}Z82626XYfC0){7j()Md_t
zfkg0~&cCHfod+8_`%pUT=&~13Xo!iSP!@<m7Vtq9=!vL_!i&O#@2c<lW<R?CelxGx
z{AO;q?WQL=Rwo&mK`OKUdccuQLFGiq=Du-23NU#^)z&l!Kp-LD{({7Y2nZrv8{jXa
zGE)X{A|8G5P~`x71>yca^S{>wSkd>3;v_&gISfE?Oh9B5;3%meAi%<cc}IFNl1XTL
z1Z%Je;O4MEY;jO&kpC+6!Zaqho4u$}&HE*zKkNWJ&~K^8$8Q)&s<r_Idnhd6Ifz1h
zgPJs4Vtq`1U=)zBd>=q^p!5iGv}+;?+SAJm22iIb6p<oA!<ILY9oGhO9^#{aewV&N
zpZow2ejV)<Co8Zaf`2~g)?w>5kVWhZM4>*=-+pj}0AfBS<Qym}IPgAg1SW-5kh6|{
zq~7diU*0?5pC<?)6rk_y8~LYMQTw4<eI~F#XBSXX?|mppfBr3Sh&h#c6v5BDcOZSR
zr&7|jVbF*ONUwl>27TVYgS8X=K<FW#Kun=~gq{RQXj|~t#CyPQ@4prA%y48<laUq3
ziz;d10$zxIG<Ny9U?BRNw$<)ljkA#@?ScWhLHu~Y&4Aw^7v~h_!Tz0I153Bx4*hB2
zfO_t}IUvQnqPVs+CZH1p|93$=aK94HCzqga5FvqSEQo*~TS-oT{!lPFBwNJ6FQiW|
z|89LCwh2^>GeB-%ulx`Z5)e#S9<VDDBv;^D?iCj6)Xyt6t2f#fq(1PFIU*R)&)3)Y
zEDR|l70Uj>XY6O~I}8nV#?GR`;rH~Los1&#Iq2OfBsB0_2$`4=Rx?Zl=+JP0f1ls!
zdnc%0J-}ocbP(_d@{6*VXMg}Y?n@SAvAZ|i{d*pKH8*=7;_DLvpG<s;320SdqZKR?
zggA|H4&VXm8}bDR2Cz;7O1=RMlEaI`$Cy*kumC^(I@+z_*H}>T$sF@3U2{_A&44#M
zHPp3Rot03b{hS&=bVXr*+P8urSKHUNcv~3E79_-SP$6+luh`Qn8q>MX9#k5^0_e|U
zGw@xIe;;6g2^Z2~q>G9mP2)RIKP}|rHF*LG3gyG;F<5jMNFNR6N*ZnmN=EVz+8*LS
z6&2daqx2?_5HJ`@G&HdO892yQJk>!iDtH7i*ftX)=m5BBtZFdtm#81;fuFw~=$E{x
zFc#5SMBkskq3?Nq8UPjKA=HovVZZ`h$yV=7F98JB6eI+(uRwILWEgKZg5A%LR6u{G
zpO3m^aMLT(32pwkgb7dW^k=>lOR2<wM|~sAbs}zk>EuyvNMy}LmTY;z%!)MddA-uo
ztyRYT)%D67i1_$zj?c^71KP%%wH0eCy0{>I+3UsgXIJlpy#3-`J7JZ)iEa7BR(UNK
zwc8fyo_8auU?3|Vp|l_^F|kY_>P|u2x>CM+F%$E8#mtQZmB1Hy#BY;Jw(isHxHHRJ
zZD>LcZ}MydugoRcN2`+->y1)5W>48o#XhjnQ8^QmgJUHZt)dO}ZnK!r#aE$P!vnYT
zYgdwOFdo|a$gh80WmiUg*hSnc`ih9<j!9XDBxaqJ*mc+b_3^fYahV6X@V-}2R$VNc
zr-o;j;>4UQL8Uh!BN{VFZKZl_EijHBahp|g6RR$fCFu-PG0`(nB8)Z4@m%u#<Z#g3
zV3#etVt?N=y>!Kdcye@KAd}W**Achfg6=70$Hsu?*X+7PXaS4Srdp!@!~H$~URh^X
zNcQKqrL?9CJqi64=EVTDd?CrL<KI?UN#0{meVdiT)9!UG+E=a}+kSzv20k9m%aA+!
zueT3LFY|^Vmr~D(2tRw1&Aii#LlqUC?s7dipjm%D=b4!MsAro6UWmbjT*HzM49sp!
z&5I(se%j|Z#>=1>*d3)LvMD~f9}r8{v?O?FJf|jMTmp17Gi^#(a<ssbaN6BI75dZm
zcR7OCs}NjF%q3pe6}4Z(`oeU%zq)#1%?UqP25AbgYR?l<q%ULwTl{O|r=r{o{uvfp
zC6M`K)}>!Lt+ypaFBX`V+H%jtVoz8WASZX2oWBG#Kk8NtqcHv*QO$w(+Ll7BSDuPy
zp)H~A|K#(_tOxpQ^QG80S-NljM?_oB-G=qNMs*r(aA@90mAkorWxg?drybPbqA)(8
zZ)w$QW9!bA%<@AfxHlWG0;CbY{Kmu`G#-iHN}N+~&c%aQbfpcs4gxI7DguV>rFDi^
zaJ<42GurOayR?p~yripf{uu}B$W$2p<iKmv@0IjUec`5E9;&_X`eHVKWT)f1cA`q$
zEjqBa^I;&fx@>bud}NoFD)NDs%{PeOKD{rW-4j;cX-KXpa;o<W<@s#$+a1R`vI%+H
zBmW)lqX!ncMb7UCjeeN#+_vWL`L9P^u;a3)fyeW0ZKX)WK=RWVBk+pheR#&H^67B`
zh@dV<ZPRi4$ovu7Ji&iAMg$*?QLos>t(6Xx-D8E#Yevp2#P!D9Pgk5$;QhYbunt4V
z4T-1NkM`UT{ek3Iv<qcw%9?o4=`L>Wv2hv!4}~Ip7N4QB0OP?COa+P(bIfq#V-gY~
z>6k0r-^5NdAXfGs*7%07>ygQ>DGp3ZO9A@IT}GpCAp)b?;S}@tmJJDyy8hjcUCZt?
z^6|AZCoF6-wvzR(dgP~k`G3Y(=#6N03XBr(Q#QjBNgbULzi54bGlx`%b}^pVKbWN-
zzL*%}yN&Qj-xJ;^g`iSIoV5)upN`OAp}&yT3sA{)Z(9_5Ch}<Q8r$W!u#vn=BJ%G%
zM_`l*n}|C_UE&W+fLVMTCo-LBKM7I@ke168&Sq18WOk2~Cu}JFLN?)ys`ckm8!V1~
z8K5vM$;R4riDbH{{1!Vi1a|U++m(s$OjW5>)y%sp#g$kxyu}aa;3&uBk@;PqerBZS
z#CZifizK6GTI~TVA&Fsz;fXT46k8*<K@MVRT>X`CZ>|>oFSCUJM;+P$vloqN+_g&w
zVcL7#WJLQHSz8M?8l|eqvBo8YuuthULX=4M!}{OS-u++Natd4gU4?X1Url+@+a7k7
z`hDgEf3@fAWOrw2GSFY#KGJ#`BnI5;qTxNt<Z<?@G!Hof{_4->6JxV&%bKzF)O()p
zi)tFUi6&QF2`zqKk=G`jD@s*Qm)wVdvHnY3Td(>|eE`yT+hSkmS(p)pV)N}=p&%wN
zqM|mjy$!PMe6WZ;vKp*<949gyO=>>z9&ABWmrEWBoW#i01DCEmvn||4g0{AmSel$w
zg(*6&rCD5w8K@I~etgu-{dh;z%gKV+)Z66XS4`^~Dn~zI*4(C>3TZ}`yCL{1mO#h9
zAxW!gsgKjIXLjm6BmmmviGK1J;{TFqw<~(jtMT1A;;kpk`DNO}L@oa>uc4@SKz29a
zSX|ECottOlCV~&?Ba2w>y*`i{iPx83-ghGw((5Pa?Fl%*J4GWzBk}+^-DWKf!uy(O
zT+1|Xxlg|{7mrS`?RmQama4@!ZTFz`MR@XS{dhAkxfv9jd|CLqB(^K%31xgbp=2BF
zgjC0f@tSp$t=_dw>E#RkOoG2@HO=}M$e2u{`4;5d7Ces7G{GnyQzLd3-7$MMV56&{
zxn7ph6g?}a4o8m{#tEalbK8q@kj@Np=4<DZ>Oluc7Ww*7lCgL~oj+|iW1_iY_>EpG
zxlRQhL}%q?LArM*t5wHi0%Qxt118;EYiG&e=;ia;cp%boA&yBvY7>6Ne{#=}Nq!i~
zdGYKc(F$w}#N0((;x#yZ(%h7Q=mz$6J8q;RPHKJH9>1bhT$F#Ho6aXTb=b}|O!A^&
zR~-s+=r*drzfA((w!1O)v=IeNT3&VmYx~yXxrt5SLGR7*GlQiZ@gjQ)i3wG%qQc>r
zQa$zkcXZ=&7XR12SJjio^}ct++t;x~tk<+1ytu%?3}q6s-b3w@qZZjE5A(@M9+BSr
z5nmk<9!5{1ZTM8Lf-qR**h;!y{)g^6XTCYl->F)%Wa)-dN)=jFK2%Y%Q{z;dc;}mH
zk-+?SS|Ke?Z&hUiTv0IN4e5rdr19oV0x0(v>W{U`4Qf3o_e{xLi6YYXjR<pH8cS|o
ztCbH_$qiH<F6}#mPHXU_9Sc(^?9%T1ZOD`HOk==HT>13n%K+{=zC$7YnChLvDsG?>
z-3$-j1eIv_%=G+|fx=efG)Us`&VqDc!K>N`EZMLHPtIFcf|?Vj*wDD4kBXgRw2LkS
zeLMy~(Y1A~0qrZUdUwFmb<=MIs-QGImJjL^yWsKF^f0PAZ*zW~6v)$lZ>1INtD_w2
zi3WIO{^6brN+zq=waw|Q5|1_$^}RgxL&xX?pRS^L#AKUjDQ*r&SkB6sODA_w+rb!z
zMtsK5P1Gm^l4yG}VeyiLFGmhct!9$r3XvK61apR>G<Kt6+L&2KnHnw#*~G!v+WJ}X
z-}7tF(jC>yW$Gfi)CjEdR%cEIQ5tu+58lWI(|-vc*#{ermukNIL%h1P>K-QjImF6y
z2Cj8KgONlyyhgZMtQ_kSm!ua5cq2Zlw<qib^jc|97GTz6u0~Kg#T|)G$LJl*;-nOi
zy(fQNX-T2b%91j5xDxtQ%4wBQVyvQtU$25_8e$riR>1Q4>60{qIckR0Mn>8E=-Z)}
z2O9KNltWl!SXDoIz@K7y?f`<V<38-&Uy0(%RohQ=kXJMM`#Op91eLUoN#{sjUCy{&
ztk<p!fnC0c&r6W_X>lG>el~(2EPbw8CQC7Mb2usvAIKY@$4&U;J@0I$<(R#U+infT
zN%ReLUC6BN2Xl&@3vr;C3*5j7Nzl6goDUlE{FQ$P8KR+VmT4AViEHqU!tu#$w@TNH
zDKxk`Di?A8D}*~WaI3*#4Mj_W#8h`QmhI_k7Y4Q(0iX+H>hhA{)SkLrn`H$=bm?Fz
zS|cKFl|n3h_G7Ti`$8`Fj7xfo_?Y45elc$!wnN}*una&J5e3<RQthbGZMd+$Y(r_?
ze9b%@*4u7iQ+HuqY`2setFHE5CTc}jl&N9yd|sJxaj%(_lM6wEw-tr+s=i9GVuK%u
zJ1B8KY;SE!^-L&1rjBdcy}#d0ziUw<2vsRyR~n8*c8qO=9%l6nvovWY><2Ey#$)#>
z>zpgTE<>uKlwKwEJiA2BOVXoRgadRwUpP&!vVf9r`2+Yy4e(0&>N&?|rDo<7Vzn>e
z$gS}O#xhIHLQeAwDuhJ(8?&D^nEF~2=oZPElH&1rO{XVj24fqm<t=@{tP2#B6BuZP
zC#AnYO9hvl&TAkV_xCkC5;WP+TSVj6rSRDcatxh^00ohN{^^rjvtW(0%6q}?sBB|r
zP8HkyQ2A%&i1Z$4pOJTTU>s=xu+}W18pfY_IJ)K#J?<8qVBP{kJ<)JCO}Tue{CFN5
zy5`N)deX3S0o}4B-qh|M-DtP9ip!;x4Pm`{w$^?v5q*%tT%2dBl^hni=wpWWoD|2U
zdKrZR2YplTI|h+|Eh7HXJt(24mI$mUy+ppwBBezCN}_=pmwWFa#G$TL5Py-aGYPKB
zqLPQsNIyB1FC)9qQk9a^(G6(OGic?kg5mqw7ZNVWJT-gl@!aDfc*TxQ|KRe-nvj{%
z2x1N{0%1_1-_elV?GDTlnu7;VMi+yh=DKNNWzRo>aVj&(8DQumBbQMCja~Fo#Bw%n
zk!8$LoS~+kjl<`T-`lj{yJ%$3;~vM1gFh%&Kxfdbb0ZZM$&8<OAa4g!y>X~7t74|+
z7F@&^&gug7xN5&B)Zn@PgZv8Bn=xG^$`wq{TWM%wSHs6%f*W41ppfVLI)o#>b^iBB
zjO4|YHmBYROBd0^lxukaB!`vB2=48V3NLYP>5Yd1sYvIpaN$(JEodQpvEuvKpT;i9
z&?_(1=-U)+{#*%!sBFj!+j(kOtc~T2L|Jw)e$qzd80MMk^<u98y+NXa-r<+D2p+(*
z`Sr`sL^N*E@R(%uk^}|5+eR4<h|E;G++eH3a#Q%nF=s*OaDIo52QKLT4MlzYUa&~V
zM2KSs9W^2j0tlli0f$YxkoO7OZgsi}<!^i!>8>W-dH2zBM9z2mL_;|NOXf{kC6_xD
zo--0I^0E)c52*=Sb&IoPN>c9PbOutvCI7PKv8e!kxen4g55_)|#Cr3BtIRiQt(iKV
z@clM#m`$HhW4l&^Qn-pDY6<1J-_2~N4)avH(Jld>vgyb*-GzIN2#U)um-|c+S{(9y
zgDb1p>ek_T<y{QLq1HcB_VsEUx>QEn9v54Z!dH|pZ`7IV`Q$MM(g=a)zy76iJCF1{
z$zef{<MU8mEA99xNO^2VbIr+Q3oep8HBn~Nx*UiS+sNzL#f1pYVly}y8isf0Bmsh*
z-y3b%Jk>dQ6l-78C-)BRLVx&D1P3ogrto|soh})jb>q5H20<xGJLYwK+LtzJe{)oq
zlVof0b*I9u^y{U93lK@LM@^B*jxS~lYe_5dCUADsh*4KMwBWLe8T{FPy?XYzByxM+
z<5D9iek_4Tkx8AUkkHXF63LmCrcG6|v~klisQ^#HC3ESih^zqJ`gLPPYEHWa_=zbm
zvG;jY@lw8>MoFE<9j(QdTU-V5p%e?-hcM(UlQ-gcF?jF9x2YOfBMy!<4Mk{OsM(+0
zR|8kVJ;4fG3@sEXY?PL;zba>;z;A%fpkREJg|&8{AJ4-hwhcMQ{gD>ieO)0kV6D|9
zOTu_Rzu)px%d%uD>_s>>Sa+U(7iRrjiy0^vs7ON@?Hpb#DMd<6L2eVJ{aPrQZoXuw
z*()vH`-If1<iGB%IYA+`uRuN5p0v?SzAUwOqv4i3r66#BdS4&1+Nt8?hB>kuO%qpa
zuaq4&bXyWnOeQPUQFI5^IVf8_lBg)fQ&J4WDVyYEiPA1cs94F~W?vRc{^hClwnCgy
z8I|bFhCgX)W{Ui}LtgJwD{oqw)L4sZCYSBif+o2VXZ~?R`k?$&awkS{N>gh9Mkn~i
zMYVa_StMn3!1g1EvPWG^dB3Nt$>JS7ha9D@1Su;%nuJA+y>SE`X6+8=vjlZd5t;_s
zb)Gk5q;}DWWfVF()%P-36Wf@rpWUl#d!1Z@iJ%~Pr>wz@`fe)BWYt&35m!>Ll-4|q
z?<_VRuVHRBFJ3~s>M}<vn_PX1-@J6E+OEweA45NB%ugihSs|=r%}GGHjuy6vg7f(M
z@u^ZGPXR9*-~(=J=NWCo7I6MRO4C+nmi4QhDaIhQ(>zcI$Z77hQ8b-C|IyWfhc~1z
z9CH2qiLqIR(q))q7TnI^U`?w2uM9!c-cr*t85%REp)$TC*HF1pBVpwOe>cuj%QO6+
zZHBQT?6|{FGKM{Rf?~c-6utSb8Jc{qxrZ_EPCfkqFi4LnK4Bf+W9)0ODh*6qB2F;2
zRB8juUrQ}B!Ze3fGXF9;Cep-3`k7>vF7sL^U5L9xt2i}NO|(^TyAFKt{Sz8O*sZ7X
zsskBQj_dYWo;k@N2#miYFG5`$loNHsj&2QdEuGZEpE9$$wbwu=_9=M<-CDmf4nBXW
z8&Mr)Y^;BuT{tuL22KCXsq;~P`UH(1Z4tABG#GZJw1{>r9?449&-I}(O_0k;CbleS
z<zAB&s^7SjlG1Y`R3#$>9bH$L#o}WLg-@d4u)lGQ`9z4z(vr!M*PN_-!1Q0!i?AeT
zv`maIf1K1+k|(dcH4G@oYh@j0Cv&?ksuUQBZqotBI(KkydV&~OERqbJ*7_sYoM#y5
zFK!7LbC&5*vU$G05|ZM@zZpEX&SlOUl|kAWPe>)ef?=W-7QNXUJ_KK&@lmir&FjIe
z_31&C!T7ln)zHKadFo4xX969`!@bpXrJ-ft6q8qt>$^Way}wLPOw46FkNts<oSG-r
z`ladLq|}d+f{sPr!eQbmgJY{S8@5=5h8q-8Ed71G|D7FX+9H<Og2NG+JUJB;l6V8D
zg@!^_weSrRMrwz>;oh;?)FeN#b7Nz=)=Z9RZnaPx_<bFhfqj7+&Rl43%pdfc1iF}S
zZ#eqLTQJbKk~a}=-Iw}Q<9-_6I;hd2=%cCLgtNa3-Fss$pOIr{z;wK8#dMT9p`PTH
zMT4!kAzWEv;IqV)4HCGvAxRC-VG2l#%l2DzT-V`F3Z<<tOw)7d?@_!>Gwt~AWqU@t
zAA7snwCNcm>>Gy3Rav_jEd2|b38bB~zd_Lmj@?y`XW3#om<sHmHx0LExhpy*tGWzH
z+50M{P6suXl0sQh!LY*DA>l44R>i@|NjZ|2J2-cSiO?lCm8x-xU7H-(;Br}c*N>%q
zyo<i8XpbkG*@D&Z!KaUb`WvYvkLx2CL6cn|?HtIin_^dIcE$37n;>=|pQXEU&VnNh
zG~WU#6Jy5i*I9H+jEaaKvVPGC+?~IIUE&rvGhTtE9g#wMCAB{N=@%K$zA1c&qPN|w
znv8iy-68c&5`;Io`F~Moz8b)XXNtMvO%<fprm33^3>w!g?PekuW;Wa05e1*I)_Kwz
z!+fQcFgv?<bR+w16Lm>Bznw$-<-H9m42<huynf+h^RYc&NTJ7m3;Dpu!!TP&o31%{
zfpayKAIz$nNJD3f*H$WQSGES^_`vvvIW2TmpDrhBVJF)NQF$%$Jm(O1Z<emvoyaC^
z%6wZAg*Gw#OU7JYPSiInJiycTSZx#91@Q(Pa7{D`N`9qka!yjcz&|fWQ8kh8s-g2;
zzUSOt$Bd)7cA~*w6ZR~SfOp>L?3*A+grBh)LsZ-oM7mCrVy5M?-Nss3_J1}S!8yY~
z<}xsj<{craL(E%nfZUJ99%{kGdg(7=l5CiW*;E=0cIMt*(M*-Qs4I!ByA4_F6kQR<
zKxqKPsfmJH*9PB&jnbJp*gsTDo4l|#pKauKF=$JTzcSq(H_fXY-aN6syzj-)pGD(~
zI8rY8%~ql&Ew@$!pD+dz2(Y?A^bs4MB9Bj?A^$F?M#{-;j0|sg%zDLsXD)`sVPBsC
z8&c^pTg5LZM;sj!OV>GlDrt-RqB|0wMABN3^Ky|opzU}}B!w(kX7G5<j?~Lo%8R+Z
zE-EEw63(;{Dr*yq^oeR{Ua5f9l!Y1`iq?gNdx_zkLb!&FFK(4fGdga((OFy2(c~64
zAjHu0PBMVUbhb7fgVq%+ahCey$P;B(rov|EsJs`YJXo7&1=nHOaYkRAl1)|n22fD&
zHsh)!3z>Unnz}&Rfg7pTDi8$H@v35E&?6KSQeYwi^GwtxFPyg;CFwFck1*DiGWw*;
zpOrH6=vB6n+X^C~9>Kq_*)2Z>&-iwY4hfH4YIX@Qm|0CNbS6epJ7p~XlkB9}e8KMA
zJa0t;N&kQ|$+eCfK&QPe@s~`_y^U*sXN%!ma$DCcoV?5=pi?fD78#E^V(U!OiY^$`
zb;umAa0TApM+~}!`o3QCx_h}7hZ%_IsY_rC(zWtCRq{g3vOjc)@ZT7kEuUdWuBvu)
z?^>r>A>&Xv00>1R&b9tCzxdDRh{(>+5{8fO|FMYwJdxP{n<o+rGy8w?L}Fp#X5;wZ
zcq09M0aw7AS_zFh=*JIaZR_L(gT9fohj4RC-Rox;v|rt0=+_DA)Upp_eQzQA>}|cs
zl0r`}O?~QAt-Zm9K@mfTKt|;y%?r-1W(9>MWn~onBPSObg55VYH8M6e1tTw%Z@F&`
z|250Q7s4|)KRz&%^2&&)D3O=XYA!o9hkZn4W(FJQS`SfI2dvKJq0#1^o(^0!HN^*T
zXmUCN56AA%OanT?0$y-!5u|j7QRMYPlz?S*GMoDac!$d4C;_g!yR)AMpdujE1M_HN
z;ckE&!DUYeaOB}f!&d{-xQ<{5n#BTs3r*YITv^Xf4jrAGjF_7r4w;{tQ%H&++Be#B
zKvIEb`^VG>qW0Y;06&Cs^!cTZK*<9uu(#TOqzVnsEUiz^AwvPtt7~C_+GGn7Gqg|$
zN&uO=`9~(v3HB41e=?e_`*F{K{%qp`!IDn;Eq?QUxRJna`US#AM_*fC!Q1fDSx*C%
zV5|uG2L{s>Hg>Y|_sy*wwHD?#C*o7>%q`CiY|KF&-NibK_y>oh^dbG(@4a+lac!z|
zb2W5ssQb|&UA$$><A~bK2;W#=1G34^-t)UeZL|l&3HrF{f4UuPZfJIDeEvX8*Idg;
z`(YRy8qWDWy)w`bPDTD=M>vP_mCMeW$Lb%Ho*5OL2@2#41hgGJo$hOwcVU3|H9Gd-
zh5aWeHnlsOzYpyX#N6lr8q^2k&gA$UI+VSuGw_G|x6PYgpqT-1nmQIQ=wE+RjFvmU
zhM<&y5`f{{JFjLNsB@ao;V6jy&-ZO_uCoxTsiB$G_5J8~E~%-K8X&KXMELbK0buoY
z*FS^5H{3S@qpyFg4^hX|=mgp(Na+2;IqS!JmwN%wP+@C&YwhS|ky_8#0QR-J?}SnK
z{85g$?3D=m*nJI!xa-v#+YErffy(<~1>~l2r$r0-ss9M<2KovA0N??)bb#7NfFJtr
zHgzq4QRz?WARu;oZe(ENaF-x2z|{dl2Z5V+4!!=nw+5%rZ`+(ZFCrvr8qh4ojT^`i
z5nhdLK8LNohMm3+Xtv!aGuTI9Q*5&5-~g;LeV?)kdZ}xzfuR2@cD@39c~qt+03*<C
zek0ns!vip$M!$5pY(dQV_uPM^s^FZ;ryl-6%ZMvQ?=2ZEBxT>+{QMj!0UHX@0g!vW
zKSM3_+;?s|Bop(-26vD!P+o=)P>E~IT`yurB4{7kgV(v<5VZcsU055COo3lGZXel8
zm^+Zn0f3yr6ew|s&=8o(=S!pth+<q|2tq&ML*xpG;#>f1f#fkvP}3NIBpJ0u2x1Wr
z5Fw(60g&6qVGRER3!eUgN&f{-`~wxg|ACf%P(@<jhyhLJU&L`3)Bm;TS?>VYV*W=Y
zk82QsayGBe($vBUXdD2YFn$_ESg-_h{+Wc6viZ_Gc$3Y8e5?)5b1-Ci56bhD@n4zo
z<MnR=`w@!6+XJKt@n-yl<h`VQ_wO3Nz4~Smxi#+U#pK~WQ9?CUzUx6*77pt{T@)Vc
zK`H4!V+LNd{xCKK4r$_UV5NUGla>iy?zSHG(gPni{;Xn7=vfG6*S{x16fKM#*VY6%
zuKn9u=jtcQI*;Vq+TZk_%K+8RkNm0-%C%5z`gZQG#sHL!OaOHP`k#hxLgP>T+b1<d
zU(SVnMlkj(d4ReQICsY8w+V#Fr!pgLnSmUJsUIN^EY*SheHD?<OsHS-XXZ5EG|7+f
zbN3s_qwU{eC+*+tL3syT0AolC7eHmwuEfzjW}wvb7bM^`g2mOr8LZm?;N9xKQwz8g
z=q2_SbjxG@I_Sz~)~hqIvETG*hU&o`yx|gf^0?@xpdEDYA*gypytnwCc2BNNmu4wO
ziSnUC@l45nAjG?AT*0X6a=i%e^{U%tL6XKxL&o(<_v-}DmA&gACCZkt0QoG5{|X3i
z9lM82BQYSe#LS9|SVu&ML9?sa*vW;qO-dU^(Ow|7dBE3M9!_sq1LM1()S8%?77wcE
z#9QV1v#!NaFSMzcJD_~D_}k=-mtgO3yF!5zA{0_>?gOfm6tqYO)XgtGjHfW-5gsMH
zzPD0IZ(PiGhQX0vPU4m2Bb?aV`k~M|o8rAk^=HBR^g%r}n5RB!a*Jr5E%JEiF#u`W
z7yMDA!bomMH+bAp6^q&U&|iNbJJHiFBgA-Ca~LB(*;zgO@92W{?b1kNluTsOjHn+I
zd$)F1pVj+Jggn2PS5u`0l_;!=3(H0f7rraSbbPFZf~`RN^J=ozMVtn_D*nx%M-G+#
z0rZ?b3_n>m4h}N4s3j?O<C2f>05@>WR^6Pd+IBU3d_t;Jyb2PV!udiz&I{G9#m8R$
zoJJ;iA5OW-2$W~%#`Uh}8pqwGWc<2%+Pm^Q>&AVqZFRw@h=xOOB(r2Z7jZC&L#CGC
zL{&PwSDzkYny@!R5LLQd<rjB+fBqG_RfrLYTTMKqXZNdjN4R)|xUg2h4YW>0_;Bg3
zZHp1^xKB`EVf*->YvDR_v3|XbZSi1_C$aQugPwE@r@6`?2Fy+q$ZSX()}a9tH(>?`
z2_dZEtuSDtbRGz5BxMVxGp1i06|%;M9o)PLrX>)5FKG4E>O?>(@ukKBk8c^%aD=xs
z%0TJrr6Rd+(`VCigqN$3AkSaVqd$G|Zt$r%MMI@+NWAcRtezZjjEL7w<XLD3=7Fe>
z)HAZ`jG5a7M_RO|WZ-r~-nPE|41Up{`N_dst{>f@!%CejV^vt#C5gjjh1|*=OLLof
zgM1xhW3eThvb1Mg>R}1%Eca^F6j-UF-PnBM72GvaMv@HVO}U62-?W2I@#kFxeRmeR
zgg=ly?46|_`^y2<56j||`MzulZ51&X1B<-e48I2h>*LLmC`^gy4@JtPraDkJMXjq;
z5W01f6umzfXX8!PMS|mzT)z>SHYKp;&S>So4`b<0ot34{S@)<Yah{v2247s$4dv5K
z_T%tx0tZWA6I^0>hB_(^TDMcqCxVcBWSr%pczM?({Ga|TsaW0M%TFh}e{Nh}{jE(P
zyKtigrhq4d|M8}{YWJuBr;5VsdfJSwNi!Z{sc)iv^mficRlj8`teE4EuL8P#Og3HF
z_JzXXdBSzLX%Lm>=ZKxmQwoIwraZkl0Q*Il&cS&HXh`yiNhfZ#Z0OM`oa=2NUKNu)
zTsJ*~omls=Goj)=2y1<atKi8WJ7CB54+yi!L?vKNC9gj+Dcbn8VG4uV;gX@ekIumB
zBlF%~3G1u=NIUytzdv%gt?}KFrvNmvFypqIl^f)0Heq|tn%Q+{Ad200;^bZ;-FBA(
zDjaBwXB}I*NMW42m&J72O+LMqhoQs(kcK?ngI|=$Wq9f?7tOd(u%$vj#-1jNeLee1
zm)q5#Pegm#;0V%Gpzc62TJC8Hwswb}5obM4QcDisSgXEeiB(Se^lsGiwQIOFiJ4}=
z3Lv<bo0YA)OJ!0x-vDbzd>ES>5VexB6KQMIg<2t)4~&k@<JcRFHANQQTST4jdxj@D
zLbC|@d%vIai|?y>?{TiNVmnFj7d^1};@nuFL=S)@ogV_uBV&Urz#!kR8lZ+iu4==f
zn^hbA7WAb_SlXq2v-@|s*uIjOyZjmdSn_r-TSVsTJ4jXfI95c%C7K?N({8<X2W$p<
z6mL;ehd-rq>e;-L+yuj_OF##(1miYXHk!ygV^mmLeH&$ed}l!SFABk1FI#rPV(kSV
z9=psX?dB0Yp)q(_f+GA~Mz0&ktBd_&`+-Us*(~)|lIyW}IqYrryLQDa(DvRcL*h#-
zL34ijTj8eW>>BIW4=)Pb5q1%_cpBN#bB?N^)P(qqw+rZ4SjxMsk6tyB>yd-qI<2Yx
z8sS3C>ZR%qotQZ*+*HY6l9ZL@ei`iRyEsTia@<mQkl1xJ!btG@jh9%zyMI^`3eQsf
zUHdFCq8<TE$cRG$em{wv@}}lfEv?8-zHBB<pJkP-ur;O`W+xw0?ar{67W1&0lB&ZN
z5rQnHz|!n@Fl?EDdg`*$Dg!%@!Rpikwuvit^AHAFZga(RE&2S{Jo#sF;r`F2V3wxT
zv&{&HAj8y$qDq4Poa@IjX(3|S$}ZsL3XKG$IRV=Mb3mN$rdk@VY(aK&eel&9K|Q&)
zNM=C#i>q{mM4ND8YzVZUlZhPMmyLKoT~c>#zxE~ZHw4InXMfV-wgG7v(hO+N%5*7(
zf>J_xOv-`7I@LH${)I;6$l#1mLw~~q-7tTxYB4|hL!tzK^tD%h3f*&%;5meL0VEdB
zfsi-tU>qL4^=z`@;u4PEQxdo`^@a+*vNkXjx3hH%Hcn(qRD#qJSlGe@1i0ndj6B>1
znPQu$#^^%#T>=|ej8v=M1#~i(?c>)Uhrh-?X<(z=Rde$8H(*Iyj_A7Wb=)}uiTjTf
z^mC^b=SGt>FvH7;aJ{LR)mzH(o!O0Cxgwe<GQ#4Qxow|aQEguR;+^`xOXfdSuEIZw
z{?V2rpDa;DiX?_U?ra6;=(qO2IRfN&5qCtXH8dl;^2yuPtl~Z>5n7@f27yO(a+(n2
zx@yn)Xvh$K9I?%|SG;^uM!T@}IDv#O9iRhX_?wWVs}zvv29~?Vu3}vftuFd(H11BI
zPgn2^$DEoO&&((x?RDBrI0hD**5vj6FnjdSMRAItnNy`P(~-}*N#y^LCfK;$C&HB)
zK#CvmeVFn)7c$AnA%?+d{S)@`F@fSc%z{&-nOMZ|$uhm4PkeZNF5uIg{V?fO-v6%C
za}1d{h~yF=H7nLq%&_cXvD?C*y%F=V%FPjkxtgWsr8P>#F3Rl^z3^9S#0QQZ;xwpf
zY49RaQ+2l13Dcn;T;JgCizaJ^?s>(h5~5oC+zVNMonBvQbWJ(<gH5k_vBvZ{pzL6b
zW=0a+as-dks33D7KEPFgO967rWO#%}Oo<fdOV@V{6Ge_@5X6{No>lg9`&E6oED=<~
z#>q1Egf*33wLqG(M5-Vw&J%O9AgF}^8-G>5k~pGf$VCc^Xi^VH-Zl;~FcOq+Gttcm
zpEh4HhNEO0_E!?gm@t!=_xh>LhrbVKnx0nP6Cga7oqy(FV7%(m7zv2bFZ(EaCqYmr
zXga4`SJdb2Qsf#IG($zp%1PC)L8brXS-BEUu~_z~;x5Ct2WojoIMsDh29t-5znukX
z##B`XvVo+7c4NowNbR`R9t>+b87v;t78iMTIP+VEDiOis*FrX%9$s4c+X2GobIT@L
z6@~VEY{=LT^h4tMGPBR+jyF)cn^{(w4#}lkBPav}FMdjw3>4CBf77Mm_pa_T47E8S
z_r1V(m*rbsYkD2psDO7<C6<^iqJ_r(?y}$6$|at#nPQizVmhM^Pv-I)AFR0=?HlvM
z<hfwwaivMPJi}Khf!Xi*LbuaFL{STwJHyOQ{}lr+Mp*q_4nrJcZa^!|S4?39<}HDE
zE!yD))1}u*TG#B7*qXF&L0!4CnGu*Q*#7}BK+eC5+D)(a<N(SGRKp^a@8sg8@e-2J
zHefXC@7MC#pv#@tm3J*NtjU=v+^tAnMALZvc@FpW#~E6EtbApsg!fO&dyqyx3iRof
zng2YZv8u{wwt(X{nu)s?cG7!)@BZ<PF)kDq<rqidE;L;`^31YoTLM#crEL<4p4{*}
z2v2#r`q9ZW@q1qW&rde~##StLHFSG^uwkQn>rnpP;zwad=aKZHl@3?El%X%07{^k1
z8U3@VP>$`NmK%Bkhu*fHl!0~zuJM~KjwbKzg-#}BE2}b1=S0V`U}{R{n>C<s?L%j;
ztYC73#s@`mh*7B*b>((&7tm1PP%&>;zUT#1OwB1I)mX1iOnL|qTo>zka)?I-;Chn3
z296k?t}%~TsYKYvwu}6NFfVd+LI`R9Yyd4imhfs7+15PzQQUl%@<0eHC<Ks)u~CU{
zko0w1m?x4X)$Wt9Y=WU`imtLV!U>9JXB8D5ZTFoTUnqn46y9z~f?cbs(bj4w$4*sa
z=W{!2VuxVaiHy4gTW`U}kiK&fVMWH_mr3s_jRlI&%z2{%d*X4zRV|5(=d5}gj&%jR
z+0=xRg5|)2CB@Ozx>>%eyPpFK--3N?G%^nvF6!krMaDgy44jO7zGOu$zVtdZ31jLf
zIJGfBq-1SNeC`ZJQHU4sDNC_Fi(E7R^wGcw{ML|&J3qaVoO<nYP*05wqsk%u`|Qp6
z_LV3!Y1x}=XJpfj<trLkv`}jH<I0E#;Y{Nm2V=1lZ-?#xKA$9OWk`Fikag$xT4PM~
zomVi$g1c)HAZ*_95rZ>J%ugmfrhzyk{?Md^l^3fgrd_Xfn8T?=yNvN1jNsU@e#q^U
z0`89#XyNwbZ9~WXjdB;fGf9uOFFoYOD*ng#93-XjSuM;-5nD^bAtcESzB^)YGi2O+
zh<@KLzrS5MK@n!_9aRgMiT=us`yQ|=^l)&z;Q5X(T3<Jz&CwTV3Xn{w3TJYDM(Djd
z;1n(4`e+u@ocb_?&x`W#L4b$+=H11>F%EeKZ6jyg&EufCBgzd)5F9bEobO-+*?(H&
z2V=2`@7>2&X7SX`aX7I2HDQ*WhJt8!i|PpVQU}Di(Wa`h#0u1T`+c)~d2lxm-SzoT
znJZ4L4mM}6LRpMI3uGD|Z~44#Zd9Au15rb}vU_5S<a5twQ8!>Ztd2MOOg7xRrOd?z
z27-z{7wE?8Is#PVyL^yZ&ZGjLPBU0naVyD?3c{e$UXOn0z*VQf&&3<Nym&OLDF;|>
zKz;%JBB>t^$1fWT=nM_^Zqrnp%~5avL4D}CHg#m<<=P(8EALF3J$jLDj0z@}Rx^qh
zkD$-zY&E;akn#11osgoar~z7{_~ASa7x8}ev=$E>QbZw}x-!&h$h3xndgaM}Tx?mA
zl$8=XmBBs&LFU}II^1acnvF!jMf^(XFfj#|`dEygug<c%Z@BSseNk%nfz=6<Ft=g-
zN#wg(=*NUHd6j_(<9^oJ=avOWbjHxtZLT@d7(LMsIh&CO)tCaSj$ft>k}okctsN?;
zS#?)C)8v8pUu6ywnd=Bogo>u^S(5KA5cHx`FOe~^&<qrPRo;jNWLDpV#2@x#@G~+f
z>*%h+^Bg~sv%-<<EJQK6&)c(gfE{mXseLPV2!gJsg)qo*1eDx~AXFr%arY1Wu=_%5
z?}SM3fFxdUYi(6zBy6ihcQwkV92aGq6(7#(F%epY9lY4!L)3a8bNEw6T{VcTj2mw=
zDl5reoVUYm_-ob_wzx4mLPB<Ld9Q%f&0%<v^nk&yp+rp@W%u*7_`@%deeQ4^Q;bXI
z#bmH8Vv@K=*^A7B5XSYs-Obq21&G_BP3Hjxybe&<{e$KnGOOQTCHVNY-m)h{o_%}0
zFUuO&;bze=u2q>FJ6{TktIkFGjE>p{^1eUqSeH{+dtNiA8~Qr!Cx|JE%CpOSYFNG8
z6A~x8P_~`EOCmkmv?5#0zK$1G$$9jaMxc4Smiz>T4}<ALpE&KRCM3pB<3EyouA86n
zl7E)e`YXPCYwhsk<N#WC0yJ1-x+?jBl-%v6M<iG}H#R$y9HDo0V<%a*fB`37KQsh%
zoHG)}IJea>kwxwS48PGsQf78DIQ(`t&SyFzJ^`c*sE^BjexBT@Tq=(Cl%cic;Z`?0
zW5V08^@XXChI0O&dy?N{k`31ay`NIAjjFbY(hv}xWy83qYM4`oKYLHi9CauV(X2a6
z?xIR0Im0C@WJBZ$xYri6{D6*JLi$;#zxPAnEdlj~E~RQ(foC^V#Ag)e+b4ug4?Prm
z0eAhV?lLv#*4VD~Z=4FSVU{13wz-8ugo5{X#h_cG!Qu#XVy->J5*It)*=Q8%lf@fF
z8oniayV0OY8El|b^ja7idK>mTwKFhEe^s?7Wg8R`rrh%oZa>A|nh9irfeC@=O15LV
zjdh;FLvTk(y1Jno#ir$IIG3EzzgObF#$j=FQt)pMQU$MDwwmKeG0}8-OqPtP2{N(L
zu;<!!@KB!SA0XOvM-aGX$aMV-_55)H)2CXi`OMQeZT7lqf`P^Y;=)FKex15pHa&S;
z!z{7{{MB(>fT;$hlafV3)kj?SqK)m_dXWkR0i|4%M|7T+XnrD`(ZhCAVKg0_h#Qtn
z9{gQ9fS@8S6E07jbBEAc!!e8>hW+(kZde~rSM_t6FIB2*72%VRk_WuyM#yAV{$`?I
zH$tj6L$$FiDpc_GcU35~>@!c_K{Pao%r|`5N4!w3*@B9XblP1b*&m{6)+0KP1e#!J
zDMtfTxyrbG{ldPGrA}}3cUmX)a1&!S5ZC@_Z!(vDYAo!|A;_h+jD9oPrNtKTc8<cQ
zH=a&lM*qQV-FRO8eawkpy~Pcze1K&-@$MTjEpe;_ZqQ{4LvinN-2<ARryWPoV8KdR
zok=+)s<>am(=cT(X5del{-XSQKiOtRA?EL^40GE{Dh8s6V1Zl?K-Kho>>_@n0t9<|
z=i0`#>1t<H4F!Wxg6UuEj0*1L?(y#XXcsZ}noYPl<s)X?G9NMZ&zS2-0_N?mJqtC0
zevOV;@D9Lew9#EcrO4E^bw1ClLb>giI5~u=mUoy3ayl*@UCJg<@@i;B^tGkM0k+Lb
z^Xp|z1f#;kAY|TUAPGOQ&3!skk>}<9mWNGc26b4t4yz`#Mzb@kY(z2jv}706=%0EZ
z7{Q6O-8bM%lM1I04_{U?py|dj#zZ=@>1~x(Obx^(XTn*NF6;P+(6o_2G;q_MMZhd~
zV@GCgKc#v|(RaBE&Y+w_Em%N1{s=c+d@{(&@!7oX;Jn3Kijl*m+dxS%?`}Sj#^9nP
zMM-I(;gI_I-5q-Z2ToF*UQy6ma3H4KdX?R0LRgNEv<5<B$`W$LNE^NWuXCej+c8J>
zU8YOx5Vnx7U>zL&EGQ=B`OR^sGJVJBI<oK{O%8D1KXf&a_X<<tejesLB+)qDfFJ+z
zD2~<8FV%d`c~Ac&rP@s%7X3Qzk+B$dmH;l9Y;V!L_oi-ao~4UpO}V`OO`>;|x48#Z
zA^(t^Y%l_rM2xo>ED{YYgn|a*9i}19c7~jL!pq<QslIH=qYy!l2|Dc~Z`eHZIjb(0
zJM~EmyFEgsB6k}t0V9+v2Vs`eh!OuyOr5T*QYiB|?R|sHY0?5+qp%>5kzvWMbSC!T
zRTQ@p7X>Q$Tfe{cFzR{#)F)qDWjZW!*Iv71v4XTEv(R<eSC4>)qnranjAd`z>;xJ?
zmgH~ML?wCe^dG#<T_eRfYDuvaoyn*!Lw7rx!x_Op=8B)esz^_0j3kYg;>jb4KxuDH
zIfguJZIcYyoi<^g<dbFxX}(BI_cGjH(&Rvq>uDix)_7aQ0+dVoMW8s|1ke=Bj>bf7
z4Y@sO%ULIYublZS?{sGai8Af4j46|%z@?;Z1=^W%WjwK_-o}mKN58YpQw0h?d#MjA
zQ<~W!H%b|nrt&Sn>`e9k9NOQqmFm#ne3zcqjet1jjY#fp8jwreBivp;-SMsJanXeE
z`s4BZ4VOEDrP?ib6{51}Os7!3NRl6%CQncF^J)hfY<p{v5@^bLXBNC0y5_4tMQOmt
zE+Bqt)a}iOjeYWGpJl>UPLm}6Y<234;VQ_j+eKV@twhhV-7zj1UH;}j#bW@7{{mIM
zTdt5Qi~$ClflsNMvb*3JqgbS{M$O@+(3?dPzYiBHK>|7L6dq{cVEyC%>EL@1_lE70
zWA$YxPYgH@Z|clyGpC9qz=e0R9xh?!`#olnbUBle%(rhvO<_x$%7SZa{5PhQmTzLM
z(6X>DTs-!*Ek~nRQxQMz;-8RlKdkGCcB(CNg*8G&f>j2+QcS$xwT0eBIGj|uO_M6t
zA=PG6BgES0tDf41lxXh>6aa#ws9PhqYYEHD)XGQ=AsAEfRtVm-&Mt;4ZcBHq6UX0E
zP{4*i4C(p0JeJ^^*hmMO(63{rNZh?0YBRCZa;I1hB@Z&!a};1|n$PkBv2X1}EP*xX
zz6-z5pfWFa0%LgGd#O-_O^>t4nVQH+l*gfc_OcV55E+<wgU{nl32|FF4V~X#pqYbV
z9}$sYpKDls`7E55+L0&hH?z(}?K95HaN_(zEMal^UP&nPqLjW)a14(df$sKm!u|3C
zWvo*foFBz5Wa$L(qt7yJfpFg$MCv3RryH=iw$}mjfPjO-WIy?H#<Qh@z$TjoY;NxQ
zndItAz&Em`v=c?eMrH{w2VoYUsA2{UEUJsPDh6-xTLKYOuxH*z_2Zy-pAWyutHSsm
zwY&YonQDwg$=+WTXB@gd>C!wBVkMkk#VIAe$))ESm8>$GOmkW1f{9&JW?sOw+5I)A
zpY>@zDY3Xu_!(niuA;-QYBXQFX(>F_G@{bZZpC*B2bEA}#Yh;YxpMVYQ>Q08hq^G!
zf#0J`Y5<lDQ9NQ1Kjzk}LDag%<MMEdKXK}{H2+Hj=C-jx0fUC3cPuuojAQYxd624F
zFcf13o4;&cM;tj!MgeQVVvyQP(Y~>W0fIH&gSnxnvznXD1GL+vYV`E+oTV5pMh2b}
zjW)E*bRb&${cwZ{0ikFkanwkX?&Kg3TrFqpsRJ)8hDtQ|<H+&aeia2KSk#tWJ4*`9
zs|B9`bV#($E3%Wm;d*E8ZM}h@H=pG-n8*rNyRA_4Q4gM!Sl+{04M9`CS`nG*ivRp1
zViLb}?9V3QLY%BxlZ_}dQp23S$Wa5EWWT#g>aMeJrOdVO=b7SR3V^U;EZ#!}PX0th
zVTPQ}+7P+r{H?HPuB&B@ol%GfPRy2J_yCCaQm)2{U@9rX7YMfdc(g?GG{-a*=NO4q
zua6u;IixvIok>WSyodE)2)71946;vO0<nzBiLC=s?pxkub}lW1Qz70TixSx&zc8=c
zP)5pXj4I7m>L2p48>(l*JvG{uaI$vTnS>*tm^c(9b`59PNGG>e$5T^^4}oi>Pw;i{
zg3BBkUTM-F2&8PM7?UgbIR^zT-<ZmtPtVgV=)2c(AeA?Z(DE6w;|hrmX;I4rda!L3
zDL=<Sj^TU1;Ep{FrcktPv&&>fQizgnGCSZSx}G8ltFEzS&3mNb9Ht}LM1B#!nZZ|V
zW2LZskcGwKC_qJC1f-Yb0!-*f*!h@EKM%lSS;wn<@7YwVCeMkRV?91xglJKbURitD
zs;Q}`NeB;(ZKVRvPP#VE>ZPI4_K^cVhR7yD%Sw1`r-U?7viTN>z1E;DHX)f=h!BP5
zdOBG24s7dzLHwe-r*r5sbo$ZH@J2WlgX3TvQKxur`9mR*u|q(n+>(ZF<v0-qp<&w#
zDHC%v2il8ob6PS|dOKO($o{4!`9$Y+pn|UHyTbg1;aFrEom|q2k-PI2X#*GUNZ~K;
zMtrQ)=aL5-$3@o=_jyfwWPlC6V+hWS*y{s%px)CNB<z0E2Wx|UXh%t5-Wa`+p5=Z@
zql4T?oY>kcv5R)CqdwR1&~#HKu`TGG`4~uFieKRqZ>(qsWOUER)`{vCkmXbDH(@Sr
z@JB_S5kHyLK7@n@#-E)p^K-y{YnOt1WJj#Q2=9FtN|&_39CJbxN*-P_&I&G6Y(IS#
z9Iu~*206C+c|qD&>o(h*i-nqkX#!bl=<I-Uxmu$e=NPYw!_Qu1THgNI>P9w}Rae~1
zYiX5O*^L^KVf-}4{XK>X`Aw*=7pbV*n<=^0_GM+cyxTz<lXz_dH6wyTATGB@2_?7P
zWH>mVgpR=?^PTYmsfvj~O;!m?zlh!V`=F&cW5g>)_1Q=p-~+Xm*U$R45ncV-*3Bid
zRHCfqR2zaO;6n4q0=BSPnHc@8go{pPUL+T6zt6o0Jk#{$%cAddlB>nvae#jg2@L+!
z6`eRGo|tY3z*T5AN+ImYF9>64=F<w>RFc<6%_QX1_J%uUpCdNc@;{i@uXiK#ctuKa
zM!`aGSm8s}KC#yS*xZgfATQC=@{Xhtb*mo!==S^D_x6ly9xj`8mG#M>rS<xC<-H5*
z5$s_Hjr>V@^%%a5A~%IOTYsU@+yR3%Vv-j<r~atHpdY%l0#4LUOCDpTe%a=Dcl^#V
z$&cA%hdVN?cP)#C9tf&oJH9)&A-ctc<#O(5ukfmS-TlNWkmM)FgI&jae1l<U`5&nq
ze(AZnd09yB@zl|4vfZ2H%WQ+|#Z0>11~{Y*X6#&4r*WUj8o&QlsQt>qMb&xi+`TgN
zv#$cf*T=gTnk+MKvH*bl#^X>*IZzP8#TMfSv!r44U8~3^{+N?w4KZx%g9eRE;dz01
z_vpc4ilc-c4ne+1^%87hLaJ7k^A>S(-R69CvUXQC+`guYG;cu}_37O`(L24BQQ&8%
zQ<^VEy>H91dUjONzE%)G&x-Y0)hW_8x5to&ugP|#(na*7(O;)fER*iPOQ!28CY7Sx
z$rw_Nxo2X&V6Uo%XwLLq+7P9Q@3Q&9QVEUd2z@Z=AuI?a_V|fwP*}fZe{jX6T<Vyh
za9(6QNvtA#$w@8nxv~rhbJ{f6Z?4!P+;ibfauE=2Hprvr(#sB8QESj$m^MVDX1_r?
zthJmpBEsuZ@{-6EZA2hd_<H=)!jD!TjS_||1VM=<a#x?<amB+afE6OI*lnl(iS>pt
z3wtdZEJAjas3HMT{8$V>_q5W4uw06%ZIV2UpgO2#6inCkCZT5>(QK6g&AQWCio-VH
z9y7fu&at+!#f%0f8o71p`NPiCHY>An4kUf;FYG*H9~{{e<CNZahAoMSA(ZkIuetbq
z$Ghj(u-Wl(?x%A(Y!l8Osc!R4<icso_}sS6;<D)E*Vi7CM!4}zQnqv`dwCZ2)a)lG
zTKG9W7PV^9xE)W{zR!h|pYc(!Y83ur$#r#8!j{}4h0zHXN~l5^`>PD*Y4x>C(q#nk
z#WA4uT64<i(=oj7L|@BO>;0`}iAebv5tecnf6F&RugEm)#PuW^EcikMR+C}Q&c@sC
zx!e*&>f@+;p=e)V`J`Op`XELg2Cr))h+R~o0cVqG7y+#p5U@M5xGwioYNheS#=JJ@
zv`8GBZ8rmp`Hu51Y-4B5N?0nI5trzQx0m~J@Zr4v<gatHuO^l;L*9DORcgD>pR^~j
zH7@2n$j&?x9P{7y@3LEgPJ8oW*b^{DTDr5ur&@TIx<bwvqy8pDQFI2;saUH^KSXoR
zKizwvcp?;YI(XK3u#)01PvA7;TO%ruzueN=zR^#Y$u=PGrgb<n2KTCFdAp7Lk}LD$
z5LN57*+pqfdC3(O;K!yxroXn4+dw#9$_(O=5_m2haD6^4{em7M=z<~r6~+*c-BO65
zLHoAv!0e_9v7q1Q^9imxA5M`4L*Lk7ZSF|TKxIP8WjbHirLC2q&yf8+5l1$G^=2L6
zHkpr!a!5BqmiWlvtdE9`l2QwdS6HJf@^aC4mHcmd)X)t6c5Ll50HtYuH!TVYR=Wr*
zL`unD>O(U6uU?!)C=De0lP~=Wx@DU-QlGx^)aqpY9Edo#n|Y+qi^f{d_}M+`fvoW>
z;$$-Nb$TojnKnEW@!)e^s$NJdtJ@;EydFRai~$AG?kc@jTT&6}Dpyvar7gF|X<e>1
zSC*eo8rBvka^!#z&esmhH3Po0xsxT>p<$VXG=iJ{#=Nf^?cBf$HtlHo@`mgaUeR#1
z->wmhi;J{+Y|vo%L#Xs4)=$KMqSRA~9)IG-0P8ocC+10_iR>cmVU=`Z7iyMzVUnI3
zT%+7+wtk$)W&>V`TO9i`ksym9$3mo1xJa?CeA4M&lYz!nQs6Bo#q698mN2zr2^3-2
z-tteYdwYhTCbjhpL#q)6>3TS}7(<1*pW0pA2sV2=9ZjsHM-#Ei(RZe7(^A3`hbjp0
zyUhR${d)2@w|6EGhsu~r>S>q^4nGl8ASIC}Wfu0|Y-7Dc``aWlEFeyi;m9I8a7r{Q
zzf|3hZi_D~y%4T6k7ykR(1W#o+yOQcv^(&-ELMSWX-anPjf?uToTGEk;lpeqLWqac
zao4judgXv^x-NTNhdtd=TzD;w7lrVW8!US6ABuAsAr|oSFDW<yhQOxth?d=zz2b%G
z^0!jfw>5u&URCuvNUsx{j4ou=rFNPkNRN_^bUss*fSc3`t;$j;WZon?XsFn96b<il
zKI=gziJD<nW%%wF?|kYT^X$FAM;6NrYA6kC6kc%rOPyH3=h5{wLE&iNsh}C9Z+6gZ
z$fN+$;4#C;k+?#kRzl?<<A}i$?R`tXmu<=!sU)qv<*-kdH2PAfWt+U}<=mGg!TGi*
z-qy~LkwX0YZ5qm>Xp1ygM<-S~0G~Mm^JBTF&rnv4`nEsW6b9Ba<WD9mxDS=O$+iqx
zfI$uYODpU%oXbK5>ZdBZdEEhE5N1s#_AwTAZzM&`Z!akGNzbiEGf4<QPUGXT$EJ+N
zFWIWJqaPaC@Mfd$Esx>fOHYOsiYSXO6%qsY@qrur+fL$`)R|8R-MRwD;+n%#;)E-#
zc;S4E$vU*DWM_A)i?W{8VMcv#l|OWraACw)TeL1%zU`%B$fd}tPG08?MJyj~U?zCI
z$pLR;USmcQGCWF2cE0vp!d`VNM5)YfixJ)=QI@`I=2Pl0xY1(S7X6tYyXLgITzNsf
zLGJfeMB9wD@!QD4OqGT>c3suoFX9a>s$;#f#xWPCH?)$@KEi#MJC8F6zD^1XFGJ{n
zlmLy^DCm!4*LE6S;C)ayyo9KFXq@r3?=+ri^*A*wWj^uaI_xef>+tt3Qy$^*M6f*i
zXibWetmSWZ4+{^;3x(%rA6Ffn+q4fJCs6k#Xy@9<j+*%%Vl4YgY;C9I<k2eiJD*kp
zyDWltV-MfYD!QmbvQx*Tbc<@(BuYO6K2B?8sA@)or`urb7PAQ;A)(+_WS<P0I)Pyg
zpv%`Q#y^sNg3}%sc|@n!I>5IE>>$j7Q{p@Ofj_LGhc??BXEdBTTcqtb5EpAh>!^!Z
zQDfRDFJ{y&(oLN{T-F6$C}7jOQ)^(&m3~Jx^$|WqWX(;y$(oxM0j9p6zQ7m_(7Iak
z4t5C*PQ~^Raj5osKZg63oK4J$hJ(LJ3GDQ_tZVfzzwX2dag#^9Cl4Z2wFIjpKek@l
zC}cFgtLf0mo}6}@!$qh$(Fj)0GMkQK!3X<6>+vMLpKNIyVOe<^|H0oVVK>8_ylKS_
zyuv1vOQd@PAIyZj?aH(;8pCYzCr}|1-Q=fUU`MW3l=Vaddt<DB`Ac3S<XHs`-w!Sw
z17o3R;Cx^kLL*MjB2MQ=GzF$yB0um}wg;k@7B@q<xX)7W6tt;h+WgXroM;3|c9L7O
zXJ!--dl9^yf+c)li>_-J$Ekv5YnrL$enDX_h1V!jc#?WBtj<246CRC~qHl>lx~kK#
zbi6BBmDnEwL=Ov>_XpBv3zK4H4Di75CuDNIaW@f&P78WSe+)N?)h6wq)PTX(<HwuR
z6if6{&hL{qsf8IdsS=)SY|3$UF+f!J+IX9N+|M`zzu;Q%!<#b%gT#vh&)HEo)-DsZ
z92}{n3uGSnHr#y+hhLzkqtI28IxlZ-Ek>`by&fLQ*YQj?Ko3oChIExV?;r><fG2iQ
zId0;4Gx1nJV7L2?Xn|a=Qu5-SgeS9gKE_4Q`g$+6e9RH~qvKsJ5@$EErLgwu6QU$W
z5I&2Ba66?sdeE9#yVCZ8&yvY==DGCbN7M#+LMij(xlbQt;}$}TX)|*uB<nXnocH$C
zP%lX}gDDZ#FMCdpJ9;d6-S&M<!J@|A+c=>Hi{`$6^DJ?WdFL_~a+qiyJj4wRH9rU`
z%X07|c&|&U%w9B_5j?9C_V$85lU<Eh`0bOoof6o{sbm$9IpHk-3SVp*kL3!lbWBk5
z#8G;Ipy#56QnT&NDxj3t6;7Y}aV7PqdK~WQKmfj^&DN7}qlYNGenDK`XzQWZk$Pqq
z%({b=An{_Xvdv*VHMhE^b*A$>cBT#X{1#;9pVm9U9WUSS-+~XL$Oo&MXn45}#<kxX
zy-1=JdYAgbN2>UjnfiKO!M)PL$n%>6AHouFSCE?wm2~TogqTmE2)k7<zhYXMwf8uj
z+?HBB6hs=dvAmGR4I2QgBg^Q0G)bp*`eA8@hRG|Qe)8Vvx%U#VMl;~i4(w1#-}k7f
zXXwRH$(0Cgz7f+8C(!cAk1v(IlY`yIDCy=6Zzw*q3MwMh-rta=O9Vd#*KJ{zn6Qj(
z7#2SQ)F?;ty?N*JD}+a$%raqbT_t2<(tSL-V>pUPf057AF{rY=hW$Rf2-XvVCbuV-
z*CjviQx-8;xp5gM&F8Nt{{wBd%<#um;WzZpOnw1kv|SpJn<n*(2@8z%!{=EJH#~h)
zhilLYOKD6l;KN$1-K9$fES<&;O~Wk8jb^c8rX4jKSo#;K-zz8|z9eO1Oy>(G7}}z<
z&&W*asU48SH@3)&9>`aJ7aRRrh2M9dH%^unz$qpZgM$1s{yU>&;y6cDqw^rggs5-V
z%<84n_tgZ`n6ie&`V<iX$xs-Ka)m8z?^pahyZwq@$SrX9lDP287*A|c#r!TgfAb9O
zUyVuNsEat@)qL15Is>{vvHoE`H(bNOyEv2^!FI-)(D&kHtB<<Y8Ye^dS(9=H2~0Eg
zyMhRG@XLu`4BST>d{r(SAgTOife}bOWCCj7c*I{Be5MSg6!`IFy5`Hc1uw6P>{T>}
z%SA~x&#kwEs4o1xSt=}!v3)eO$l4qE_}#5?>(dS7eT=p(XAy7>@wc##^`8=sa>`7?
zkAu$=*F#jYInH_Mlbn3UT$w|5Z|sa{oveL|k7C?yv%(KG3j@orUEdrbR)iwp#LCI;
z)SH1xcUTy)C#NsPi0ys>L^s6Qvx0&KcpNU}TPqvww%U5xIz1#D?<C2-9gzF15-+M5
zEXK<*ChYMYLzYxK4P{&hPQQzhnKA!x9ky{Qa)Pr=No72cCv9G+Bl6UD<U83|aGCSm
zrBv)jS@rcc3hOMZfpi*@T)BYNVi*o)QH@RRbjqbS-~2H_Mkke=Y2ZFU<yFPeJs7QT
zWEp9s+Q<Xfd#c3DTWU@iSb3))Q?_^U;ZmIFn(`NFDbiCNBqf~3rN&|1d<8<y&QP~<
zyWa(&0Lx8%S4`7FOr^FSmeVz6t<LMBd^c&}5z>5GtsM0L{=-ATM9;L==Zdmk>m`c(
zz>@aaZ>dml)}qR-#Rxk2zm78XLquk{>khtuk13P)S3B5s_9X5c`dB%1q$1}_DDd+*
z>nB_VdxqTjaiEG%*$=AT_T<ljsn-tQTKrJ=tOmLISiXE)91!ToKaqCttM4UvK#iW9
zudxIUwVI_>LeQ{+Whzu~ZGY41*Z&F5rD^VLFaS&Ha6^;MB%k6{mnJ!(X4YYc=M0|u
z5Z$WDljF9`mTcGNDN!(OLlj$}m+(faVFks1qq#4jVRWhZP!zCs;}5O?Plbwy+|#Hk
zqkL}umJ~1-8C_l?EK-;y34QY3h2%2{x-@ZoLo8XpqIuV*<H~EagW9#lxh2ClucAw|
z^oEy@0>c%Sif?i@>h71Q9g8MItObwHV30zY6qBE*WaK%jWEl`FY1QS)#~Iw+9Ss=P
zlaoI?XwtSK=w)AZY?}Ka8?L)NCt|)7jgr+>w{f@O)+y=1=1*&6@kAW27UVtSr2VK3
z-K=p`B+iV1@AKIR*^_qrpygS<jx6{=FIQH*Fg|SLea9<R!eKf>N|H0SwPI1)kTN31
z`qH(STkHl4%9wJcol{3EHSYNKS8h${V*_r-1;IEX$?TrGBgsKZqk8)<P9xVlS9TeF
zUt=6?A9HKozVx+X+C;jlR)GZ`<%rNo5Ohu=a!v>QY_;TN>~et_FUe9_AZdEU>sDuz
z>xhM3H<8bujs(xrrkM`@RoT`5BuExJE$!eVtmsImIC~pem&-<>%VMo^_&J!4KnBTw
z<9zQ2Dom^nS$@`2UuaFK`&z!|M)s!997X^u0q#M!TAxLY*z)$EMxd}LBofK_OfC2|
zxoxTdk<)!TP+4d{mt-<X%|y;|Y%kKB@?gn9m87TC(fOv&QNr-7jH_F~$y+~P6p&Ip
zE$S6wEP*KiF&n&)t1T@HG|;V6NQ|GqF687)R1UgwtN$wKKwRIUn{=aw>ZX*Z*^Lps
zGz0G1*FrodY3Wz7OGNJNz`z&R)|3+o2%n|FoZg3VK$H(H_sp}&OFYiEAUo8SyR$sQ
zraP}1sz)EVL?x&<+SOYXVA-ch81+;ye)t*EJh9UnQny<s0c@CJAi<$XpQK+_iX<Gt
zjqptk&tVddo<b6N#~Yw=-}Y!hSIzpS+P=(3q~ftyWsz(9?t2D44>x+1v!*%3u1fxC
z)nL}-vAoBbRHat=cwNEsXW{o+Zntrf8xjvpA80v>+&AgEmX^2=18*|nK8~)h=I4_N
zWi4?9Zz$B`xp1JNX%Xs4p0^1d6(X^Vg(Z1*Q7*t#_FNI9GbxYh&zWuHr%}N$H<Z=>
zyp(nK_GnCNwoQ=@4A=hElSFxrz_M#Xh;IAkXzQ7RO_fMxL->2FXWApf-uzBzaC?@b
zMO{f<a#Fc0-KXlA@o8-&jxe1iW9F}?RIUxQW#)64v@y9kS;S;iye}El4+uE%MFZlf
zO9Sm&3pnqnV4;It&Z1Z6#51Z3k8MUrFYFQ?r{J{LZyq$N$-)!?C=WE_XcOWkxg1_z
z<=WkO9;&k!j$k@+?s1Q_b1xZpdan?H#jEG8b^G^TLQ}a%U0_2bp%Fpf+<_%yeFz1l
z$;mX<6$;a1g^U7U6+P9Gk<jqjZt{@}ERv+IQJ88d6S{FJbf(Pjyp5Eq6<ty+6HW?O
zXN-PW9kkFUQEsH@qKGLE_>!<eEcn=~HWg6GetGq`c{4SiFgeM>b8&X;)AY&Qo_v}k
zg1g{}!7`5(3WjVDrkBoVGrYO=9V<I4GH>!<b1Z>>AKvuV@RgOUD}(&@oI~ws`sL#5
zOm>f%8vvKiC;d35BVcH7p&@~z(B+~wF#4x6OUS&u#&8S&ZNMgX=$$3Nv)B-lq!Mvz
zzN7F&cKKvA%DX<)&df%PMdo_J{udDI3Sy2tMT6!A7G+G%uC%8}KndpM9_dTVvC5~c
z`tGr|j4k%rL5X|<Kdq=_cn#kd3Q0~%Z5?HflF7REyQg@`cQQ#>nV#L-9s`8Il?`)3
zXbIRG-M0l_Ch~CCCNShv?%%ATc5G!xps@ts6nHi1rD(FHSC|%9W0@@aYI+WyMRpcH
zdCm}|Wkw6_1C+*;JOtSH3;U%s?g-Qnr?<7e%?b5Rd|YMgtgJj->UPUznH`gy6=Fx9
zB?s)5!+JaeMS?bdUFDsxS+n?xF{G{B;!<X_(y>!-GX{MwgDUbClh)VGuKiU1DNm=y
zTO%K>_;JahvMgo`q4bj3dy7dzu!Jjt3$bbh_H;9jetl^3UB}U0OwF7KcoSku<C(|7
z^lV?=kRQbLe2Q2YWK^5Gpgy+6s20E1U?}%E)(4`<4vJA~@F;QijSc5UwJ`B7@L+Vs
zE15l|7Rqc@@L_EBYYg1g+;~|2#9S3(Nl%ECS{-Mv1=E;vTT>@{S4FCLuK>M8Y5ewq
zgm<wB_2oth!6eH)cez8(GPxsV`F25ymY&5Xs?XKuH{rMo`z}ax$#`n2cr_wYtM8|N
zG@_=zqhVx$oxt>bQ5cEUeF5Yd9-}z1+y9KU`3#@fRZz3c-OdR3Xyc@FD~scF9xm(?
ztC<Y+ETr{@eNV1J$x27px|5>w>iPwRT(}1}vxYFS*1YD}t;Gi(sUjm0XT4V)TC79c
zOE%WS%YS{KK9j1+M`1=G>WiH*+%*N&eIozm4=P7<#+>ps$3D!r%kzLz_EEx06O(Sb
zX@6`4Q<YoJa;{)uUIMgDb8kkoK6J{{pg^lRPu`u8b;98>9q|oUab%jx1z{_oNO01(
zx$he&UJbN!qh91tnbDqpTCr<f@s;1G$baFZH^inM<D%ZjtAulu-!k9m4oRh<O{=Qt
z*UnIFY&HgLRCO%%T70yPtHL+w9Reg-MsSHN+ZDe$EyCjCS-rSYG(nK{#IW95iu3KY
zoMTMcNT5=9&Lvr_SA<iY!@!W9&;WCtOSnAWGhnTf!+dCGs0&bS5Ks>j8e?iY5X<p|
z{x-~>lC%>wh1-2H!IUhe=SpJmF6&ensxFD1Xs26PE6$4SgE4V5<L+tg)BX%p`Sgw1
zS1N#2&1{!-uEPlzUq+@qOJHG>bc+mPsQP~LVk_&eE+>Y1^=ZC?l}d@^J0lhtBVMCj
zoEWPVk=mQ10ZV=Hv#FkHIrq;HsqF?Sn^2Qh(iT^J61D{yCQA)!HO)pU?=yT35{%T;
zT<_J${TrEb8Ga04BzTSoQ&E(B-Pwbk+gC3USd2`p<ag-urOaR|!XjKaYi3Wtj>WCh
z&vZP(=EYKtFmpTD>XAW?M8Z2ZX7#gk=HL1G>v+|M`s~6+aQitB@?NPk01pogCO<vA
z;8M<q(pV||-U*h5okjbErDH*jZOXrc3Qx!mM;v-E=?OyM5UzYrZ!wvS)gTb!@aNz+
z*MU6CSGW{gSE=OEohI~N*1C!P_X;Vt4eh(>0U^0T!zZCf4b}U7T)`?BJg5g-d_P%X
z<4V9~EWoT%iud54k3)2AYPyWgwbLh6$3$AvHyHi9m}Ias3Qo%+uEn`Rd?IDNrb0=`
z_tKU^x7SzR6WA*Ed(Rm)<Zq|9lw~X$5lexq?-WSz3+shGB$ZEx6^6ktcZH)g6cM*P
zVaN}cqVj&R*mL;oq&2587d4DSkDeLqhmJTc#)gpL(WZx0c!}Gpmsy`}ImppVOgjt5
zwOx0NT<3ow-O@seSl_rGZb*J4A^=s-?IYKj@?NAPA03HyBdy{_q@SGY5ib-bXC)uJ
zB3&`fX7zbuv2c!KGSMH7;(=<9XsVSF-tlLwuj;|)ee<?QA}RjB=K9Xa=E)(nd^Sw&
z@#J3@3uO>8@;!0Fp<{hf(~}NK!`sDJ^<+zw#3E|u%J3!ebu}3S{l?QyVYWbJBEH!+
zlrTgANms&V86IZET$dn;wOGfNsfDx*nU9UR)h$aAl)=nzM-?#o@2a|ZYgvh3Nd<O8
z+v<{$$|oH@wzZYIB&VP^UF)4LN8CN6yDi{miASO~LP@bs5Ml<jDNmM8b2tt}gYyhh
z?)J5(sq?7d_RB5F#$N(2msmby{)`ifc|0<pw!6FCAlx4tW6kvCbHKjI`HTr;8RaMD
zf!;)hyW4qf`-7Daw+@W9rs8YiQ(nu`)w&p}+Mx@l$cL^od{stKj@AOC=I*qr;V$T;
zO9joTZ)ck4;5o0{Jd0Auoc5EX+%UD~eMr~k^`<+ha^6?bIS(kt9n}JjsqQaFB`CIz
z=tu<5?#(sE4onNopr3Me5y^CIoLy9T#ElL|8TtyF6!ZG`9}K%bB06fhcnAr`174^S
zV%Zh9fqT#Ho%QU@bg(-Mo7A#TL7EldYevHt%0%fk7Z=|~!9AexTcFyzE+X{wYnXar
z6(0Es$25QSmY#YQvZYB51nJm3_-aZYMZ9r*7yc?^B1fc`v3qncl-yT`vuzhqicP3W
z=~9wO;%|wnt4mut=smQ>4yj9E{wpn{n3X5-c(Gsky_1UXjn9pw8gJW@tvb366A^Y|
z>{5`r;zC$rbt9RNS<VdB9LMe`HZ)nU$x*z%5qX{}ig~5C#t+Qb7J{TO`|ql8;i~zC
zTT6;Rv^bFii?#P@zYZ`bq#e&L-X<M9dLg(52I~ucFm1$hns+Q7tu2UR$4g7k7do@T
zkHBU3b0J0+-#mne5)#bjIbSI!#cHByx#|h)4nHa(3A-2)%*Bm(TXJnWvxQOB!)FR-
zi#TLjSnCtQS?rctV@-LLqOWzX5eZS4pjJm-lg;#WxwK^He9XfoY+I_SR&Sl}^bYNm
zhXBa7Wpv32A`v0yD4J#0V;CW$1`So5#ilUb<K7MAUmj7Ofen7a%gL-S;D^nDOe!I$
zgM;z(a?1c5gb_sVy>R@%UF4rGjcE4pK`<vpSs6jSiZChAas?Yca)s-m-`-I`AJg;6
zKfLdU<1Wqeu-rF3;jgf9IX1-)G%%wjX6>B_s{WF`FDd)pURX-9(bmE6jmZ=r?{@xi
zf}`b|giT&^)%A(jG|^~4B!bsvP<aofsq18jfAc{;RaSC!PrJc*D*^eSGvPhOAnf4Q
zcn3GLE2;Hje6^&lcRp2!$FXipNLgJc*f*}%*b1e`t(NSdLwQlXa0no3m$}27z1|ZF
z>u#16!T9{wyHBb=#9fi(gW|+Eo$}P$Fu4Ao0NxiN=wqrKG#?E|cr!@k3ag4Rq~k$K
z-}=^7l<EEIucTRt1>Mdxi7nC(QdF#4q)*;widouk7F~Z=1q15Ff_mk|)Q)I2kU#EM
z>v7*(xugioji}&8G+z{*J_ku8R|dh0gWUHxof~{!#3`;<vo#q8)z+b+S9-siaHT<5
zvrq-3lg@-5A6P51qvY&STxTMpJK>@-WeTylPr8lk<njg=HKCO}>22qcxRbLiQNOKP
zT73IXRF)KD6grBNSiYKo#Yv?!)?sl?ZcyWpGFO!cDb#tOw5WaMHV~jS)9m&qK~&$(
z5FD_PiA!Cj58h3GkY{5T3f_(OBf(%X9u!y5Fr&&sMrSK~8@8r0SzGowrjnI!O)&*3
z*oW)VRkX|}Xr_MCn_a~)<)xJOnUbWF`q>P5by?S6U3b}xO~PzDMj*8T&~S3faj<-0
z11G`Uui&w0u>W^`GN=C)YwPA{Xj2az;m9EyL+b$SR7xv!A%4l&xl^Pe0^$hvZJhkM
zj2kj2WnSsrbIMM)L=`TkpNVeQi_9C_rcVD9Ar1(TvW-k?T4(C)@Tl0z1QmfN!uRH5
zKR2u9U$&z5e5j7b{h1quB7+=btH%etH)oYFJjZ6~aB?9ZtA_!l<;cAZwNwXY;~O&o
z7uG(M#g+aSj&ByD)~%iptdEPe*BKep`;08LjDC~)>=RCI(F*Z!EkU~|#V7BaxT4&u
zs}Tip<tiKlX1t;?*;-v^#CuYW^rNTTs29f+Gajw}#V(-f<DmRD^xh5aX&dQa`pv$K
zLJ<uvOZ1wvQ~P`L`^_Bj^04_v8s`xmnDDy1u61lYA!@=CFL=W{gnV?yA|)9vtimrb
z5Q~-{xnFqWkD;NJe4A4Wr>rbqa^u)%V-xnrqMXe^3MgKhtaTir!tT3IFWpuz9fr{Y
zo9chM{R{0BqZUo5vPR-SP(h0&?(d7Htqo?H!b8MFbfG<;z~qceEvOEFZTOp)no$gf
z7v{2jNBI4QxMj_*=7+HE*zJ8swACo=c!D-nDVUo&zqrrT^d$eTe|8a<dSc@RivSgz
zrmg{+F$EzRYkN4`qqll;Y_YSj|L#}}wjx<xS0xFd-u807mf`sk%*}wJkJC=WZqqLr
z`5*6NHH!u97pLbLEvgM~o*s8gc9}Y$5g|}>H=Ts>@a}Pvw0c0HtMM65Wt$;~uq}{W
zaY68F*|D%AsnKIgKvllexb}pJTq2`_=?J)>#ZkqG`6LqGAdf+-Rr$WpNca!)`ztT{
zwzsc&4m-p3Wd7{;ca6w?C^g>avrUcn&Z+oCq8$h^Xde0QXf}R`ROyKEV{p~?KXSa~
zy$~>(ik_rV2GmM+>(ieVqze~|s_WeJi4FxVzD8s?NKSE80xE8dMw`V;<Q>+HxXH%`
z-boWQtx#5@<T)elH?ds$oRWTOp=qnUqAi`ZmF%Q@vpQRX|EIu-m$fz2@F?0^1mdn|
zc~z6_b^fqPh^X^kV}>uh3YC!cABJ8>#YA%n_6Zk^XzL?yb=Lu#4<exW!bSZ&Z=)kk
zyif;E{27MEP!bj8fk%$6V@Jq|t1Uun3xal0)6kgR#gq*+D44400UA>%dv{q>&Njp=
zBCK{@E%ozw`(MrpQ?C>E(#PH&N{B$JF>8jz-&oR^P5Z-;<I0Vu>^M2|?o`TJ--ViA
zr_o0+#{5U8b)JUmV%(e$+({9OfY5+W(S%Op4;;gtKb1+aRMq%RGT)kha%0=LJg&wm
zoI%WY<631^z3OP}El$J;eNENc18P%ls^C}@B=t(yV?HjL-Y;vZ7pGRcj|e19ahZDo
zqiyPJIBmlfIH1jP(k7jBq!hqck2qlKbuf0ylYoKYM0cz4s%L>Y{Ydi?R>}XJjLWNU
zx_uP;Q2OJQSfdVREt!R5I;n|Q*2@+fZdc+(@dVxAM+fiDi-ivqR-QtS0T$XQ_ys|v
zpp2~pe+T=dKqg|2%F=Wu;!zk!#1&xQ{&qym>9JtWr?iV1jhwaVXsC_^Su=hR?xOM1
z^ffwZj#>6{GMA|o2I~;Cj(HR#gfUY-UitdvK&s)@bo^3U3gSVk(pTus)!hzIN<#1U
zVKi&TQs3|Bf}BHeT+{lk5rtpdk=?=Lw~6-;?W{8mIzu|MJo00WeS~rifD>5oq|ZiQ
zI->uukl8Kkyd-6jlj=zhrMMIZz2ZM=+{K>{6<v>uL_P%aZrd?&bT!s|+AZ56g-K&1
z9j}n%xt;^z7ob7Bs}YPFG5+_z(6!ezgohle*!}5|%ZGYB>%39sSupVjU2jk_w~+LJ
zXWQ_H0#Qn61v#`YXIEw9j!z$?R%%;PcH02d5N<+Tt3bKG)}M54b7+}>fFZCuuB^EO
zg+t4WwD-;x_eI9>D`Z8##@*3j5IB1|)v@Ci=-XhKv3YXAMvhBaf-sjIuSYtU0Q<@c
zeM?K^rpovbrNn4ddf&el+<3#gsq|J_Vm|V}z?31eOEGUtfg)SAJoH2=Lrbr%M(RgN
z2!uxL*(tHLo1I8WUu|o@zf#$Qr{oJF9=&-<UIl{$JjX;~E(-(`&(p%K$HF<jAbS<j
zXEmWVM&KX*xt&M}OWAO7H+5i%*J8@DTpL+)OuNByy%4SV`6M0k3kztVGSawU1o&-&
z(%;1Ch>6%(IX3mj<4l2pRQKh64o#8m)3?`1@^m!h+7j4QzbLfMA75}Oloav&_)(GX
zaU*|Y{UqySf4hrS-?43eK|<t8>nxpwQlO-~kHj^AizI*=Y}n7g*eXq8X+xMd31ym;
z9wfmDYIKyfq!<E}06(X&km))SeILD1%Z#;X-zNiz7TKHf@>2M#A7*;2i%YaiHKY7(
zC9r1s3LdM6j9WVXnEKXzZP=>5UTU9P4dp{G0q47K(62|}Q*Z2XigDb!Lv3RNLBDhM
zc099vJU07f{APnQyX~c1E{V<onVb8;T*%_sT{y*yFfco+HLVMjb4vnjy(p^POJNyn
z(Pe^Ie>b_qSUGw^cSpPQ)tW>_FOhOo!Xj0;zi}Ap{{9FJ(dRQi&dABPUA5pR_&wm4
zlKZGnIwt{VT3#3V00HfFdZd;wli?hoJv@6ouD4(BzZj$2G#trx*OHYt31U?37*CCq
zzY-{S-iZ-p;9pp*@3E4ZeGwI&^%IvuM;g0}6^x_o6c#e72?(vt6Z6A)!Q&%*t$&>p
z<xh%j(T%Lg@7$pA;*MJ^VZQSJ?0}dgBCOgr<0UuM4=x*({xy5{a&b+)fHh9rjxgK_
z%=HfSONT}B$*+L`iylwpFoXOD;cDh%vJda++vw?BaLnG9%@5{PRS8e(%_$3JO=i{F
z0@vHzo*A6W7YnzkAo-_E&VYtzY_{lNzL35zJXbASJryQjOSXh11M|f)46a+(!y|_a
zj=oSnt{H8X%T=Q8VSSd0K`3kwbk16+g^U7q_xEy}K#Rar8w(eAUuZ?0VcST3sdrSc
z|C+>pHkfDwVva6((S;lJHq0^I(joOU5dX3AawfmHoUo6F;CMLh-i|NGk8d|yvSEmk
zaszO<+zAO?%g$%ead^lk5HWE<&`JHSb^PA{Kfy${J%X@1_f4?^eVxn=h+J!IGs$rA
zSo%TzM$+73Py<pSg=9Tq#(5H`=Uqy{Mz|w%OyB0FKJ^U|T`VYoSs^Qm^`)A}YZ_Ip
z1VM=Ub%?oaJbX}s5H7!X4Kb1yDNuy}XYJ`yzZSQyyIS2`iG>o8_@z@jD_x1qhb{l9
zqkK$?*2%Q$sW6m=x?<hc2f+X}O^)horM18XphT&7@m3~GddvY6KC0$WTHb40d&*S;
zMRS!M+_$7$FQEp#4?U!t4x<egp8Ep;^gOovbEe&e0+@XmtI!3~Y=cQ`1iPW;3LM>|
zo>+~csNyy&?P(9XY(nSP4O6J!@?TU$@@qsh26J#5?FdLoj5H8b>uGXhDQ`MJ(oBn%
zSFRm+fMPGfG&JYlLsC#GA1o|uDev7@8Ikn3Vvq{dijz;egKbO@O5~d{0rjFdD}-Q%
zVoRZ0=(IHVMHU68DtmyF6D@-^JZ(hetV`J$19sbj{jqBis{A2e(R5P6{aCNP>^Txy
zPhQaBlW`WUXt7Lw;J}|apZJTFVJQ&b3X+W5HTKAEP0cKnp-o|ijDygPvGSJ$jw*le
zwln^M-vt^D!|pDfFS~_`7V&jOdIs+qt|ip&I11V62{GCK6l|Jq7|w)h6|ClV-lhd{
zMtNB?DN86iC1VE-{#z8x1rz5IW1qh>vx4^tWLs5nT##MDj8obq0QGl8Q47{fjhf)8
z88}~Xkv!hGI#>dgCj8rwZa4*9MQHq4HI0mI@a(M{UmyHDqA`PaYTLq}t|&No3F&bI
znrjQ!B=u~XsX0<|dNrV2uXM<KTPZdpso-9JQ_P(*S@DisLS5OGL~jr+bW=Kxtyp8s
zLaSZy2mRr8HBu%XAvJOh+?l<WTWLSr6ej+19TnFP|MiR``#LRmCi$NdDb-yXsG#<S
z72xS;@l)JkeOlpf?xe7bVSupvUH+!k62gR0b@U7Lo>^7CM6MK<^`^%lq_7|D=&g>>
z&$-!K34YtldvlDrtxd#SCtmhOZ|t#7*Kbv7db~}6?C8_a_&XtYmap}(+1dZpDlZ(=
ztmQJ6P{X*hwxJ|fo9Kf3G<xiIHr=$NPO%Ar?J|{gwWazjgO>OJ>M7_!ykDV5TfpIj
z;L??CH?UF6@*|YdLu-50C90m!g{Q+)R&zRL1>6*rFpZ#M!slkZ)#?LHywVc#q5e0+
zCto^x(E58pr1}5fgG^4N8w6Kl9*@iQylxa{7T?7X`xO*0^fADz%P|5AD^))HfbU?H
z_cC;Uz1Z>MY*;vQrw7@gwG_^cg<`Z@8jZHxlD_A;nT&+?`MJS$p_p~HTpxc{7d_2_
zWXx%prn28K5My{ho`Nvj9erVzOJEqt?;4WjL@<b3jGaTXDB-psw{6?DZQHhO+qP}n
zwr!kk>ulTV(;d<A2KNqL{eYTOR8)N{a%H-#PCfURS+Z(mIXzw_!(6bzts~3XRIF7w
zqtmsFXVg8%gj^<)brfwWeVm<>t*cCdY|x-&C;{V?ln;{?0Fm+AOcD5EuYfsoporYT
zCn_M)U<<*cK$;RXUxbnzH<1dc75em(^F#<(vr0F9QKE3K^ZZ+$ptOyWEm`;YI%s@N
zcwyFCWnB=1h2$HFg>`U`x@wzg`;4xl%^R;AfVF%BlBL)YHj8xvrs+QyQ(I$CUs%c%
z{8tVsX#~;f#z!7UeGiwYdovhk7R~EUK8l3&3(rG_xmlM7_UYe_>L^JWL#~j=@oU21
zyK;zWA&yLfon*WOzW^;R?Q1m?$qY*(TIBhD&NNlnZ}Rm@NuLBOWigmbFz6tE9{Lb*
z%D*xXMJVSz6$|k#KS?5MT6LG#Z1dUCtX0h`8jwJ|6nY4z+~V6zYbxsyQV3!c0q)>(
z5>H^F16%{g(;_WaUmIeQUc&=gI?k_0JzvFHXl7tW=D~kV8_)_8u9DKs(4)xDL~fKz
z><0n@>h9YT;L2RuoJNa0XN(SObrEu<AD|PXivnbu`|C%XmJ_qa9Oak)i=>m(IX$Q4
zCA8R&Eh##GY(5vMlQL~8&NPUNFxM}6XHhpAvMH8C0i*gp5lGxdE)muQUD2CujhsQ8
z6T1bRW!gXdy5*H4o%CD&5@>qzqA{4Z>^t$6j(|d?Znn(Pm{`v8Rp>1O-hCc*vpS$r
z@TW_r&RlV}i437yVSLnqsl52HaWeXrXK5b9(nYOlV>>#KEFKlHTlvqX9$rrcaoRUi
z<@YyezUiA<ezr_>=eHexm;^;rTa}Ch3i!2WC!ant+XF*O0*!}-i}i6mmlWTE`waWv
zNYm*r`o$Myn?RlvDpk7TzjEA1cSm!7Q@TN}=G=)5{uGV%>kO6<Cx#^=E3hDg^DSpU
zl<smYGCVkM4k5GR1yP4kL8+Eo=B8h(SH2~ACf7nt-0y$66E?4R2D+BtBV1&fPa1WC
zld5>TVj<MYNz^B1ZlmabX$E2qW3h*3&7Lq@X9c2nQ72>f2=ybQ>qg1$qo-L3GC};7
z>b-q2X1Dfja>?1h;Y8WHvWAVNx1ikms?CCD*~&8%8{H0m!FIf~ZP^A(!WMX!uMxK3
zywFX-V=Z-Y^Lpe$_W(`UJy)yo^>x8=^v=8d`;_+Du6C{*e@+$0P{+wJX?KN__{3V<
zeep(_`I&M9ZZTKV*5<ir;R0Dn4=Zesvun4hY{H?`6uJHxYugX3_=xE7yF-KdS+@1+
z>UOCT6&`XU{O$@!1-l04@i;q)sq?Le!Rb;EogO<@ebg~buIQE>ji<r8vx=A~aI^h%
zLI7TO9b)a%x$Wafko18{+HRDB0Y&Uh@%$H#Apfop21VQ@PtC}p{yV77a1qLZ3Q;=n
z?>W6()5i~thb$S@wt^EJ5CYSy?K}-r=j+0OpJ+?UaQ+z6^2E69$+%`FW&QSuvP7E=
zN3HeO=<1i!bAWlc0to;&r?YK2K@vO25YO5$nr9IDDxvfAIazi@^-<U7-{Ju2A%QeZ
zXg2)SiWO!7A+!t=1PX^U-Z^rC2t&tiiOJ^M3^X@zD!nb|cz$N}YN|eCI$~j8=E2SR
ze29hDyZ^l-7Rk8}1pfUfK)WwL>uK8J?a$^L4x=33w%fu+V2IvrvzIKmf5=;sxI~l=
z8LZx=*b6efQEXF!b@qLd1}XL!7$meAt|wP_=e;<HG|J-av-WDbgAGvtWb-15;*GSP
z(t61`K!X;=uUgHrfJZ*+!=ARh`inSartPTXuX2DOB!VWkEL!kz4Vv5nodL9E^lpE=
z)QNoEq!~$UFLT`Y_%q%Rd-)9vW?V~KR&vNk^0ViOlOIQNCHf_ry#~)M_vUGc+!0M5
z$Q*l9h%Sddm9nYdF_^rYJjC=Q@JxQfoQ-3!qqmN`(FiDW1|bF=2epZ30+zGBs=BDC
z4K9~9D@g>i7s)ys!y+}4MsQB32W|y;ziM+`Kg}iBju2h&OD)@+k)I56HLAND)<`&^
z6Z1s;Vqz-Y4xvEq!&Dj+%EBy^1p&)5g{cpx%#082FKsC_om8EnM0%$5N7pBW%!%FZ
zsNVlfWtsQm2<?eG^g;63BCz4BK#baP322CI9@Ql~j=w;e>G=(V^E%p(>-tyhOth{;
zMa-i2ETMG=1MmaSIBI;V&>PxN1i<nZRTLNrUkSSL2VD=HX+!|Za)!T$mBZt7d2?ld
zR8B!)6<(e)H5*u^e<puW`(}w}&Orx{_T};qUVYNy#F)M7>N;sXUzb(dp@o_F$^xu;
z5A2aQR0JXVYuZo$VhRmWdUuE&u4Xunm@b9raITn`k<wld{ZO<DqjpT5`s2EU@M!2T
z`VCAEbD@&+wl&_5GQLn<DIIog{ZN29(T-v~OI7ze;1o14!A_RoqmBPSxfC;4i38Z3
zGMCD!Us#t+ND^#zUYb~TO<z~P#-@R<DE)>#?9MOB`$}wh{W`AmsOV4fIe}D)e-i8R
zyq8YAfy~X$4#@OOXxpYlg5|6Y7kR_TwhS6dE&wJ=nRHsSQzuY@#_+}vA?YQ^z}Vsh
zj<WY3Du?6P;UO0uPTom=gy0{m1L=bTtxrW`a1T^UOVNdzEKRmDT3xZaTRgsg$TvA*
z2<fVyZYwH?zPu~*CB!~g7RuRteRF{ZkwYSa%^K+tpB-B$?w^wS#M`RVFYYz@X%17w
zR|&w1&^AODX|@V92?kORe;&eM{)%z4g(>Uc@{^;#a0_;5g%do}@bK^t?TxsQ>M{K%
zM_)s5_c>Nna{f+aoz&B0svCYRzDf_cZBmvhPsFS*my0E(qe$x8+*q6f<4H*_&OjwF
zXl+->x7TlGu!(a8qE9roF%bLwOZ?galER_l^fQWt=FJl>srj5_+)M<+urLIcPbG;7
zmY(u$Zm<A^M?fY_S)1`&5aiL{A~vT-wbhMey;`1xuQuUO?bh8W{cWi!hRFb7OWu19
zs{;qm0bFTML*SVdnDppdvHFsYtGPCLCXA36(UuSlb$xv~=GTbb`dY}@dsfZ<96%r|
z3r}wjSterao~BtCcpZ3jCKobDDVbyfl1xLF5lMv|UK`*e9#>J&!WKQH+_uSW663Pb
zbR~8piVOw+-3rb6pH^r_Hm3in&T??FaQtU=mXU#hk>P(Mg^q0jRnUn<qmAkYg9Wi5
z=;;1;g@9pMlr(|`A#X$8z!8E3fro>*L(mQGuxd1jo1M<h^!}>%s_VLb)Lb{c6c(zi
z86`Hjuz*i;2^G-b)LaDtP@<xO;sJ1Tqf>KplVicgvSC=ke@MrO6+=5Yf^rLi{F)M+
z009Z*n=T3bZO)A%1}XS!2EV@#<ZuV&bO+hg49u~y0s3S7_b(Ad0$5itEg)45fT%dc
z1Zu3<)Z*kasEsYKuiRt4IzTTdaDcEtK>Rzrdq4>A1gH&Q2!Kj}0(Qc%vlw&%ufX&o
z5KNHQk2-{xAvVQ|WJ2uh@Nh8d-~FkJzecoE12A_0!nOce1nLRV+!IK5-G%{F3xH4N
z<~0^90Jhi-m-$Y#;?x!33!oq#XzRhW0tjp7@az$!6QFlFI0dBzFbd8<f<G*3Kg$7_
zcWX8PI5jx_;h)o=?gR|S@a6@fp~b<-2#1hi8o)IJF9ZQuHB8h`v?-VXAsRnyAe-G>
ze4oHOfd<|M@6Chwslb3uNLT>-&cXhzXI2M5Zb4m5T|w9W?GrEDGmj~z1ZhffaBu_=
zE2!7*zDow^38Rneyc<8;HMk>-z=z-H4M4&)wSMe|=cglOz<}J{fhZ-vOrB_k-^9&e
zf`B|%Up_itR{#mj08f90!miytY~{%p@TKDR4$t6xdUJ7d0o3MY1AGS63eNah?9~gn
zApl$*f#1D;weR;JVuOJVAlm|fW(5BjKy30a=8qCo^WQ%{=nmKeie1fb9|1J{{(OI%
zJ;=~Yi%0aH@t^u%rOV0+X$j8<oX^Mo{*jd8>I3y~zW=}4@AdriDO!MiZ@>HRTMGZz
z-~8R*geu@c|JAV{c>&OfRN{c(?|Hnu@Yj64?cXBMZ+-G%fVW$es<?R_Fu=dRg?)_N
z@ZF=2!AHOE06z}zzeoRW^UHqmYcJSj_wwRjasIpT>%SrZ-q3rd4>qsj7R>iAzk2lO
zDDbnU0{?VISsg$#Xb1OOtr`H7KQ9bq^ey9v24|<{cEf*SpssnRolwe!VP*J8p7_%o
zeTMU503orni2C&z{dhGx_ru?HR^8m-{ZrM!XMZJ+y72rCQs__f$_~bhzu~_|2;At<
z^5q+w!_xy$5DmS6e4U9f9|68Ud+*bhR?zN$hx<>5?HIz>4Up&k0;<8;Rq(CHjL#n+
z``rE)o`e9vp!p5^{!46r4dP4tq1%XU+F#HPAbYl-06qZri+%#yp!pl!2kz_+-iHhu
z``CVB6x6`k1wiwc8XcTGVD>!UC=c4zi!AY|f61ALJm3eKFT3Lxm@^-_=NtHEZG>i9
zM<-AG1AH-m{2l*md>jVE3%W@lJKwPd0jjMQ(yq5MiQw5_0}EA*N3whI%mkB`Ni^_f
zLZ(MkdSPl#OW}irdvBJ|+UiqPb?KeV-1B9~yE(0BF8MPmRLE^DYv}8HW*PNLrRepU
zx(f43EW@W4Mbc>yDIE{PwTU9fzQQQ&53d%>I-d)xu5%<_%NxGwJBqND>Q=w1?~D?e
z4{3pSM;)Xs)9u>s;k%?vdiM<idJCd0PGVJ-U@6JAUv%2OVr26s`+9%Gmh2zIXSqdQ
zhs!5d3KVs*ir-WgcBj%CBfvktc9EZgW6<`_oILly(|Jv(VJX>ItIpZ8kHE-MVrMrw
zV25@o!$zi^ApJ34>w_ToiEmZW9!HrE3=>R0+Hs5ZMhS)>Q}>bFc7)rus2u#T@=tO2
z$cx(B<ZHOfcJJz@&1d65`7Y^YX^BKZ!N%A?tH*ffxM2=Q3Cq^}PN=GAdLIX3WnAGz
zVsc@DaIbEB#T_&$vbO8_7m+r9H(nx0P@M9CEbH~I#xp*h(@0QXcYWe%E-yZa60d}&
z7#gPB-Ord0OlrtUPKw(;GIuV48vZI42kHcOu@q;D;Fq{-k&Ja}v?8M7!PP08a1BSg
zbl}9~3~a06&#)!DkrWvCz})61-QDHA!4mO2uUaJU;aVMd!cfzCX{w|CP2i{v?}z>+
zbvl?Qdo?N;3RV-r42><g)%NHpOSNJ{sxjO86Iq{KNjj`}CT8LN@X~76!RI6;#Y=&Y
z)Hk|ya)Yo#gt2%yfE&-X+=L(D&-_NotVIg$zDtf?%I1vrYg!?eLv1qbr&2<&o*==N
z>%sr2`(j}rV;i<t0AagN!3w@pTj4#PY!gE$0$?SlCVrp-GI!iaZkcFt_o4V`xK~>u
z-XNqyaNlu$yp<jANUXkDY*A91l-aoWxE`btvS%81B{^nB{b_wvmeRSnAY*gYS|3uU
zkk;5+vy{tj#}hey8zuVA7|r4E=v5Spfu#*uokDq?I9ZM6XI7UWWpK$8p&m*hKRe^k
zj(TC=ee@ubus%AL1j+huG+iz{RfFTnNX5*P(3XqD)nVkD1YfZzSRT<>@hG5J)L{}o
zaC!2S?<3;pb#e466hXG*N`>_!wo1?}+|%bRA75qMXG51P=hEEBT81rJ!}e!NQVK)P
zBD(;eM%wx7F;R6Re8gy8y8aCmBK#l_Y?T>m^2v(~)+&IGDX^6KG&Gv>)32nw6j@&D
zyNp_+OenKIqDS^UCP8lu>TZLkESnxMAE3b0(1_rg$X&jMhOjRf!X9yu=|f5t0$KH!
z8$nf|wp&_tqc9^p>wLx_iBNfExbJC>@9ECm!SVZb3>EBh-CbXS96{86b3q(qnHY78
zB6prW-vN4+8u(%M-x2PLGz-jFf^uNju#Vm(F6R9P0oapIk$|B&hn2=1nw3AHkIs+2
zbz%9r`mw`dPIi7IQ<V|m&j)a9-69qYsVRqzKdUTTVO!)X0t`ZIAqj%{QKJV?E^v_2
z5*M{MU<z9`5*ftU@0b{r_@2KOAHBD5)p1<T3@cqI9vAwP1tR8>K*o{w{iQ?#=AWt$
zb<9^`12R;<zv6x|Q=A^kK-Iz-aL1jqLtMOT$2sm+ki_9woICgzFW&2hbl5RomL)Ku
zL~^hEIjmuxAOxWdXOP?kXg+2R!fh`wDJvJlGI`Shj&h<1JS1FI66(~yoW+C#R%zJs
z&1@uM*q8#@ePqnvJ1e?<pn3AM+Y1ABnH~*c&JrTS1{`T-H!{<anyB3!=B(4hk1OR*
zS^#IT&TX<4w=R6otK=7cy~KUl)#KM%ug#x|U8|?0Ynk+Pxq`H%&GIzy=Iv(pwIqXl
z!7QsId?g_8x*MU0F+MDe#&SSz2Y7WQVn~}G{VI{sp=lPM*HIh!Z<rHpReW>KmV=@a
z=#K?dDYLFgx36PYb|IN;s0gO6;Lp!>FZW4B3(?Nh`+o(XMG%aw8(oKSh%vC2C`N<9
zh+>X~kQHDI=-AgyD;6yN;G#m;ePIjWfS9l=Bc0l62v@H$#pb_v1uNuE%<oINqV0kH
z)vYFC@JM91)|Xu_J1Pyk%r>rcV-t7D*iC5zGXWhTU1?iMgP-G|RdmZzZuVQPgFEjZ
zNtiFm4@<=A#=#Uf4K5(C&tC0swxB)O%Z<O}MRPuFzyYKv_E+&%I{G-@5iQXpw<kM%
zCN_?*O*t+&L2w?Sf`zVO))IFPiLrk2Le{P(W9G&Xb?RsDyJ7knvkH|PN76!nm{PQ!
z6eW~}BTsICS0#88F7Bzv>R(O@E1#M8w0)0D`{#&Ad;7kwQOru>x=1&4k8uMdpOm{N
zb%xZjb|J5G7{B(eiP-J>;fPI*_CSn@<K`^?Df}iaRRGi*v47rS6Ab2(v|H+Ue$_cn
zhvQ>LcIs!-R2C4-g#^O4pXoHln5p7jO1Vn<))BoR+ct1LI=e5`stVPVnpJ{U>J+mI
zBDHT4l9DRvC~79x?uu`%OKOJ@B_z4_x;HgvShbX8t*2P?Z(LQq*=}pN>u12uq`?sT
zsv^wjk7iBQ-rsRHk)X77iQ4td)~e>N_ZS=B;?@L3sr}_ldLTS*+#QXwK>!fpxxf<f
zi$C@oP#b$?4r4{=w3SWcw-j85;P1)tv1jT=lA(x4Fh3lvM-g;y9%wNXgYF}PZU~h4
zfVObB^`Lz@WE=Nadk8uzMq$G?_NzB#2*{CvrT3+;(HmeXEm008NH55%_S=vwWcBZM
zx4Q52ZKCnTr4X4dZR<W%0SC#!*FC9f_l~NYWZ^aKL^8FH-unkcpM!pki{rCrP|qi7
zV>Vh%YsWb^sh@Lu1N($}o6-2TmiSc+Tox&iha~d*$}|$(54Su?7&QS1NJq={D3<3)
zK!RpXs~ki(FttiY*_nvZ*bQ;Q`Bkz`fqt@#I$pwH{>`mF)y;aG`g&U`uD-$#y@W2@
zVjDyImT0=J?a&17UiJ+~GkLFYrBr0XLNA%A8Ig?Fs#fixUAgjIY>+@|x((q&{Ify!
zM}A&&l_=XM|F6GikPH7A_eDeUx~6YTGc9oa#(|*pT^N)yc<zM3Y(tqb7pYn^KBzH~
z0ll^d26e$B$;D+4rVZfrQ6lw?q;7A2Nm878Mzwk0V!?~HO-$qg4@gGY=2eTj8&?cG
zi^LIY&qzO0+u2>HxxN?}IGp5qd=@*Y!>99m6?Rm=E1FgG-V#$*9PbICJ$9g(YZZmu
zsTls;+|yC%#byEJFuJ&-pt$!lx#29;F)b7_F|kCyN$YUyL_!kU>WaPipJd#S?JTJN
z`IsgxaF$$9FoeFM>CsLxRJ{?w<*YOwS3EK1o&!A+Vz*f4Ov4q&9auf@EQ73J$QTm(
z2Y8Yj1`}D_3sKoSS{ySd1KO!<-Ua5@L1t7TCV?JT>0DyEznE^vF*yu4<rHIN>RYGM
z!r?;Zb}+1w$TvDmOjB@bn3@fE!~8m)uBTZeH8{Q>kQZnY$ZgQ4^SwU=q_(@e3MX52
zD=;+q9|47rCtKyTU8a6b<AQ44Q==Gg9@E2Ir&HEtN`@r%Y#sN4_%~4bq8sgc8Y@$Z
z-K9@@aYWO1U;SmA(!+#}JS?$f$&qQMcmKBdk;!oLhUtkV);QUR+@3#zIL<n1kCN9K
zCVPjmlyIP|D~?*m)<*suxy0EZORaobFfbjhDJ#!RekTF?<h(UQ-&zg9EaP+ly<z6#
z{ISE3J89phJ?ox+oiGiKcC1JSL8X~?c{XzZX+@MT8ykb?1L;R26+o41e0i#KCa+>&
z<_o(l^SENWvDp2t(GH!YWD^l1_l5cr?EsU3PNwkDtgVh(kcrGHN`qFHz}9W6<g-&X
za+ItusI7jZY5^qKan|f>;;Q~7Ub$QSzBYb(vQ?M$YN@l?W|^LC)U+=ugh2Ha&2O$k
z;Lda2Q_@S_w-&nKAnQDx48rSBKY&ZG9|&vz)+>~H5NeJ8%Tp%4xv9@&`EdGx9Zu%T
zy-IV|pMm6NIB*yUHgN#F9Rd~K7aA5SuU|xN<y-!Mr)zyC8(&oVahk?y+)hbbS(l&f
z+G5PQMsgFVxR(LbFRpA2!|Wn)%hgyh5g_DyHY%=?f8f2M!x(d1nIIwFpY6$`A=Rds
z>3Om86#M<A&XOg^@Bc*3CKBXv)MESFbDY-lZ`xfoI;G^Srm9v`@5|;KL3l#_<UE@D
zJZ{6CU$=g<s~=>KqIj^3E5}`?+fmcy$*y=Pl9_e!rA&-~nL8eXx+_kfMBAg%6+Mu@
zlaZTcN-~h=nPrSVyG!-#WG7;2!(E@?di-W}SW4mh7}RUpE7u<2^qs{xwSRhlL?TKO
zQ@HrgKsmHtG;v3f%t{rsfl9tZZQk29r(7&&anth7+Y3v)pyu4OtLuGYE8Jip_b&(V
z0cfa`0VHJJqga)O6fH&9q9VT#9K*fVt#AIyHJS5t(TsD>(2hTSg1X;?>?*AAPW>)E
z{LOG4f}}o4T8F%!4ei*6*$mbLn9>e+TZteE%xc;Cd=7;<UM(X)^}K~V(6u9_Qa6vX
zZtM%-^lus&`NZ5@r{F%-6ja*0=#{%EEVsF8Ri!fF=arLQHd6VXcK9|mgIK#Q5T*jR
zna2ib)MH*TpM>$!ItI#zcOg%c0q>hTsFSS@Rl5w0Pr~!mh|rxJ-y0sb=QFBUc|7{p
zWege=b8%OWY9<tSjfoE1Lh7%PJUIo@Buc~P(9c8>J8!Zir^irOf51vjnoDR^?{sG!
z)M9~&fntKWaKkF5qoZF4zbP$Tdt**McgY8WY;2P9y$h9-2`0}PDWa60@-Kl+i_!Z0
zuu=<|#N3bXgy-}3u7)zE^U3QT`4u+xh?cqO@J<@5&{8BaiN2@N>D76K%A?BDpnaaj
zI2w7oL1Ln>8W+8>cC|>KG0zRO>_d*#y5@K;E_Qk9FBcDOb86?><dKOGEWg8L09VzG
z1&XU-?OH_>v8_qZoQTu0T1o?{!-Fh=XB<$He$;h+7Sn7<M+RozlYvPTT|Io3yI@p>
zwrylk|DS+2jRiegC&<N)H1Em!ZqYQ@<5lP2QX!q@T3uBq-|^RE@7J@q_9c@?+bu@C
zd|j5ljaCzdA>!kZW!SvDty~A)b3YGxTc>Sl-kQXFYKKh6*+$A&+Lx<;KqaDAS>_?+
zEqWNZkz`=5rHBDuiXV~MQFL4lAlBrM%J4tLv}Mt|Do+0n+sN@Sv;IgDt3OWA&u^Qg
z0M#!WJCNk!ix#gLSx=6QvEU%@i@?Up<;9%Pff|ZKz)4@t0>=Nd?OaZ|K3d^cLr*tz
z+=KmIQ6n6Alls0GTWofOf~NY+;vt7<W$!@cGc@_xiNcYtOVAJbMd)&hQ6UN4mG0~9
zz?`q1Ss02`u70fNtGXC}ag60=6jMCS+y1LhT1}<ly_b3%A)XcCd<f_2kZ+T_Ec?tR
zN_apAM?Q~mK~xXuuH1+fL@ABRx~<YMHF)mEu?<ZsG|~ds6Jjt}J1*&Y5w5}d^A44Z
z3a$9YnF0RM*truB6zP+EO}w9kM-@88U1@9O+*CEAJf<Zm1x3C{F5AruK7AT~_<Lv<
z-IG*rBc&*-Z*siVcQD2jM7O*NywV}$Ic^rz0|}H!fj+{G#y@D7i?4h<VQx)?*9`gc
zuEEhbem5V+arYf7Px_WTIYkP|#4L&qFpmmh$FsgrvKQ$EOVQO3?|v!y4X;<7R#aH>
z+>5c9WL!7HYTyWtLOWLSGh<I@$tUjhkSi>qjO69t@owuqoUhcwMEI}Sq#xTnm|8>{
z!vgQ2n*49vhIqg#Z8V{`$1v5<m|?58W3ZZFpnC6nX%VN?ykD<!YcfXjvU<&OtXo@E
zQplQoRWY$!>(?TjM4_HW8IkNziA~egN;VZv4!1G|0%7olBV#BqXZlI%x+p8N9FbLH
z*bAJY*YtXf5mPO5)L2Tv_DWQ#!_6O~11gRr-l$Gq<v-ayBr1c_J7m#1&k~M6lT_GK
z<yWmpnd32egNVx-IJF)=liopsv~=uA^_RUO$&T;QIx_M6sO<yMX!S_&t>zYxL3||o
zXk;N?=elM>legq(_~BkM6VdaW+lJt3FL7h9rn8QDQi%M!7_#D}rO7v!^)c}{nC3+s
z5^(m;bTs?kQX}5sqtG=-YP7f@(`>{4mLaeoCLCrp)Iob010Sz*;Qls1SA5&?48#Cn
z={!43iV|07cEoGss8sY43X@!|7V<KTZLq|^WRminc1~9W`rbvp^*nP9CP)DF0CWJ~
zY;>PW*OpKKKRM5#HjcOHTOsED4wWdqv9;GOu>QV7{zIG%oDD+U<+NIUQ|f+JC53Lo
z!O|)pP7Q}-!<L4oA8CZ(^l0X?;{XQ<tFUv#z`-0MdD?o4-(!7N9P^o&Co>h0mFKs1
z5MN&4YZspzo3GNR#2fyY*-P1&W4jgiy4e`P9v*h*ZV8MgdBdfxewuf>ny)BP>@3oJ
zBTR{GH{nkWMC>|iIO8+ee~E|R{-lpx!MCoh+^RLt0O9!@0`F+98nQRALPMxyPpE3B
zY=i)`lcB$(@M~nm?D{lau$AZPT{|}&%DFBnthz1_(#gwo6=NHkNvjrT#EDUm^rOVo
z)B-ZV_i?=&$n}%Ohs98t((jBtH(?)<^$4*ImYhH%4`vjI<#hJ5lFA20Nh;3{LJIY3
zj$*KSMi(xW4&Ob&ZJB)|4ckFwte2T+0^MO&>d4EWmnw24`=6=q2Hpp0EL{%x%BxfB
zlw&vjU|f?HgUB&%VArK0@#6I7P>XdMKI^|fH!kz&R$EW-<am}P1ARCD6|zhNOq?ZN
z$xnL4ixO$^wyoVe7F3VM=Jg3p6kbqJ{PGmeXknp~De6kUfsUPTE@Mw~BrZfbp%U*I
z{Hd9FiKkGDY@nl&^oiT72%H35$Yu63yhoX>4!j!L=5_I-a%xw5C>&|3rm}%#fzge{
zM<P@+<ynUoK}>ZHEKC&=PLdTYT&sPPa+r>l%m{#Qm9W7`TKcTLfYiB7T$$MKDVpj|
zGld}Nijt%$F*M&Hb<3b)Ye3DL3C3o~Jkw%pEoobJinA4xuLL7+DoW>W<bc(uHH^X7
zY8<*xT2y!d78;T=Hl*mM?@W|IzFHbhB*4DP5-F1Cx3BC1)By88Jc%C~Y(_bdI1&_v
zOSi5T*l?uzp`!Y`Daso~#T0{24YSrMBgb7}MB#@K;-oR^)*Ipw5v-BW4w1VF#Vw57
ztFT_$Kb0H9U_KlmN56u%0l&>vm=Z74ABr|iKAMxDa<M0l%LOT)B@46^W&62Zy`@kB
zm<1|I{$gmkMvUu_N&=vg5Z%U2>~!cFHgGs|c6vCxpG3@jn1<U`OBoN5xX;XMQo4p)
z#Q1Rgp+<y$5BG4YWGaV*9gu*Br<;K$Lk8W_G{gwDF$L~IhQ7M|STI8E>kmZo)m%jh
z#jGaBK(t^GFkEgNV(QwxB3~iAdwvi8{d#vK@ftwY<b+c*-Ogpf^(-8P!#4WksT_kJ
zK}I(?fpHoR(7conwO${Yii*=tVg>_NT-}W{>B%3e5qbSbx&FEvtQWuclAN#GO-p?8
zt~DN`Au}FhdVk%nhXvQ+tUZ0jH}h^oxhUuk#4{S}Rg_sDg6xRQOKG!4w;GkgV~Bxd
zXM_JNZRDj=m$7h%tKwME5T2duoa+djH_cJK=K8$Xo0@VIg<u2u9)00tvw~iQ8)Vo`
z#zd$qno&IMO|s|HH`ZAIupio6f5wH`>Wzol(QN5VspPFQItO;j=NzE5Vr)s<SyYMR
zkN&iK=sGNpKoD-Cw(SA>2-;Z_YVqDU7C+=MwK2HkuC6L#YUDde!NI60nG4fX34Iof
zKcXQP`aSensUX2X`<R#(>^O4@ajlUsdGy1vg??k%qy4yLFW1m158c-Ax{OSh{*l%<
z1sjRzVBnBZeaLIFSjh(Qbs<9yMbGlV4~Y8*Z{*=!e3oU^E`X%uqaKr2PpqBY*>U&{
z+<>7^sa#tJ`;0e^<;lxn>S-s4V2_FXl5J%JUwI$uP;oW-+<NlEza4*`KOez>Wl%pO
z#a86FX*m0F3>k4nX$+zqUd%;nhKllfa)Vc-Ppc5dHFpw}e+pUIn($Lok{Fe&q?1k9
z^;luojUb1`u<dnChE}^Rd})097&p?jWgV(hS!A|}>BeBO^i7;3C@CQE(d4GDIn)kx
zkP6@IAI{<S5uop7k9u2$=IIi*fOl$&3{Hf(Sg`ukS=?QZU_=Q+D68pWvY(88c$Dvw
zZ~&xy$vru1&sb{?18|4uu{aIoCHNI`e(a`O)4VFt4`R#7_H`21=I-{HE@#})wL9VK
zegZNP$$27`5)fLxh%~RO2Yqo=HO>i=0(a*yJNZH7>sO{|P2GC>@eMo4`g6F2&MN*w
z7~vHuo?1b5W3*)mpthHL-n<sugf_r6>T{c`+}$&LGUc|!s8n~|!34b7gxWXkv|Z-p
zqb{nsxJHat_HmKa$v(Jq@8%~sggNH-p_OHOD)4g_w`u{T+dRb%C!Igi7SnR}Z5qpx
zZJA^up+^un(*ayQ?QW*l_wC3=4F>DMdapN?+5w0$5oxJIpg0q(XGj_02ftsEWA9gz
zSkC4dZ1P;roW```Gi0MU?Amzf;Um9XZu6u_{~0dPTb%tQZ|lP+=|tGUUljiG7#M+$
z$4oA7a%)zZsVO9O-!IZPui!dLXn`SXN=5bl7c!9S>>gI+`>o0qU$ckGORhGVD{yaz
z4qxSg6K-;8uo5QL0m{w8N9Nfg{zgoNPP{)=FumQDTcxb`5!JC}^Qn5%YS-QEYiqTX
z<HNg$3-WKbYje4U1hF#ZQqAZ(f#}Q=rGQrJzPHMLC+vc54{g|$$|R~Z+hciir~B8h
z8E^Ne<dGX;zELV7^4%rE__ovUzuM!^K;=p&Dx!3;!5SN!0;jC-i@c8_pw+2+mR@~1
zIbRt0K<C98fr~QzGwIF%sF2v<nykhTKg=$pEnYLcfGX+<Z5lAv0-e{*FH9HOcW7i9
z7{a+xKiAEJnc~a2!Gvh;&CK5>+y1bjk<q~Vdw^^Nh+&t;4oWvo@1u!gT1N+l6f`e7
z*ot+eI*v;sqbw6fxX9T!$G@T1Ays!VgDT7Fy4im_J6MH~lUEz1FonsU<TaTKmtzyz
z7R{ArM@TkP)A*;g^Uwi>+u2eW&z1as>D|#74M)NcfEvs#<Mnc{OBAAlop><v-UFDJ
zh0gH1)9oP=J|+^SV)-7LfN_c4;;T63n<rCX#kRd1oXs8D?~nfJG6wT<7dId1(19-c
z0y3lhYhCTXcZct@0{zHiB?Azgj?=o7`&ktFR-Kkk{(=-ASMA1vy&mUnSJ;|9<%F=T
zYrV-iBiblPqZ10ObikD82GHDAzM^x)Nsg2nGRYxWyM|d+2KK?4g;lj^IWVqAf#la|
za15Z7r{|u*KA3(}VtPGQGcEDX44TkqJ^6L@Xb449dA5oXpC_6;xyG{|-v&<s&Zqw{
z)Y#^y=ZQLw^HevU02{iP6{6twK-ysl<*%i`?UX=}e(|iM@F-#a?ge?C^DMY<W_!U}
zT@g{;Ae{kf5CrJd%Q-e)9p9?^c1?GzG%}UiF+I9w8rcNgxN8v>vK6sc@vq1b>PQLU
z5P2htb-(wsL0WanOV}1rM)S~STyR#*MW8X*;Pp~t@^6(Zi<h)}55>bZZdQx0E&}Jp
z$4R_^AmEaC^jI8RNsczHE~!<#7Z(I(ig)5EMclB9b@$GarowoD`UOz8Nu=Z^IophO
zDdD@dHUEn#X6HnazA=IS<RWkl^7fr*vw&x4&Eq{D4Rg;g`5GELtu*)iw`*(oTIx&n
z*+rU&HqGpY?q~-T-?K`fGEzR&OuY@phllG(j+r*7AaITRqgQn|^TvWa{tM^{&sQ9R
zG%&YpOTQ({M?cO94ZsML+j2%c`bEqTqQ%H<hiFBN&u7El%}0lX5H8f`a*m`4#AUZu
zjz$_0RZOF80Zmkp#!T7Fh)x^9Ak|5tYIUN#sB^=4XnY;KF~m9XxD0Of&cSsq1>0f`
z%&!As97!4qkmhvxxBWY;g721qM2`Wc+gpdJij04T9!J}y3~k8>B%og2eCv7J`7=U}
zGkUtl$W11L(STvBtcg9ys^90hgW#*%u8TD*X^rV@jmrnSa^K5PdRcDmkmxKVD=!Wo
zOIfx_X;yK%eX=d1_6>I{nR5^p><u>_tAPPDEz8MPjf33!1~yYwY6q}BTzvAMwgo0K
zkb2Fm0Z-yEBt7;|5~o4u*0{*JG&Z2gvf9T896bw0u!4JaKO<M|nfje_{Ry`4*NIGg
z>Y4TYcM74xu9!{z@<%#WCt4rPEMrC?+Azr)jKt<AziP0%qg7F#dP?_oKdIOdzp~2?
zW7a;GaOZnr^NqG-<yQB@(CV*0yBiMRf_a#YWu-Yq9R`sGh;U3nrvw$;B2}8Z(o0G!
zn)~b3fW@F`VZg+#MtTg_K2W2?iZNpR@6FgzRl%~`RP3&Mbf!xOopTHip<JL<mRdIo
zSvp-=F^7HGLStEUF0lxczH-V-<kA}Bvg9ptpIhyKmcLrMGYt2oI?r|<sAn5|59r-N
zB@!wt`dV<9E^&;K3h)W7HvHOwrfEz4{&>V^t|gt;s!F^~I_M^Coqn8&SINGIKTOe;
znLG-e%cg7zi%NK3n~vl9YWC$Z6-J;prmqdbBoAivu+r7gFV-fo6x@5cXYC*t^2dAz
zeA)O|6ruK>`t=>NRd0l9sa0^kQ()8!8w+m7BU*3w$Kt3C(wkyBJ83P^<dj`9*fk=R
zG<f^7m^He8kr8xp5jcf6x%wW!P}Dz{J>mV}`P9q&rc;xkFv3K#8*#W!P#F8gu8Uv(
zh7h(wjShq4n4+dVuqf13?Q#uvCv}N%b@ppGZMa1UB0F^3x?+gF5Nnp#U-*qC(SI+0
zm6~Tu%S_CBr@WB=TC5RK?{!DIG<XjQ#t;CT<17{D-S>W+3Dk)-Y)oo2<o8{|;Qr7K
z-c2RGH4driOc_PTI(0nqF;IjC5C^v0;?h3~Pgz^{TAk?7h8F-?CsN)}JwU9#z@UxO
zA8gY+i%No$heF3<6lVF#K$c!(#g6a#x|bw>ZmYTMVJ2Q#+Hsfo^<@}m8jr{R{K(q6
zpTIKf>J&<kTARR}lM|<LtLiiuHe)PY%$B_LWe{Y7Flmk)hE3hsXsq}^<J~CTebH~K
zaVcfhOi>zweCH{8qlO$dVrOQ%ES|b%o5qCsZ3tCS1G2qq0OffzpO?M#uJq6Ww@jap
zO-UvFc%S;EcdleZW_i|;Sf?n5guD!#kd*s=)vpuuy<-YEd$A10bN7?tEDm?|{EM_B
zTG2Xc7#_RmEA7*h!V3wyE$qteUq*MiPbDOg@u;9|adFK0;jgwhy1@xL>-yx?xyqQF
z9Tw)TTxBV$ci9s;TIf<K*C)2}HD&J7H<;&gw|0Cn(RO(NH$SE#(r5TEWAZ`Kop4`~
z2eMPy+voaJ&uJl(ne}lW#R@kdQ%9CJl;oH;&)BohUt;l$I3E`tuzG{c#(Vj#5ezj&
z#>{ft8h}AJdEK*n?QE_EQ$ucCXY1AV<?o)_)T;NY$R-d8i0(CZN@QmQDC+Oo->E4s
z3Hk~h!k9xDbh4pqd^uwM$x23&gpsz5Z!AGOz<lyRMH*G06xjyuH;g|hbV*PT#ZmIG
zK*T%4A%_(6?s!XGU&3N3b{Ysy0RvA4G_*xpVYQr9uTWKj$`ubV=FWn-*0!fX7P<Ps
zd(aH1Iy5zrbB5|^(*iNqYy(SN8ygII;4&<WMl3Y~4MVXzK;9+yPL~g~R}T-fM6u$m
zLlC`+;&eDb_->k4uy16Ac>Si1SUy?6pO{%U>%DYQ5TgU5xP*QUQtbATBire|6}xo9
zD{fR>wm1gu|G`hOvCIAMFdqAVf$>-v|381nM8L|)%<!Ku9vcTI<Nssv2<XKuZCp&9
z=*4UdT}(wxjqOcL`S_rmU7Sn}ZJ|6iV_m>mwyr<hSCxvb?2FuL*IBE~QlfgJh>{4z
zghuMBR+YSOcQ6{Q(zIGtuZt*XwW?D#TUPjO^Uq$t^7lUT9{b48e4O<^KepG-y52ir
zV?<RADZDu#LPsK`A^=eVE({vlXaFEmkbr?mC*R*>5HQr=^sSuum^n6)1BIx*=)qlC
zaDoOeSqLzC@G(L`{(5u*k(2-+H9;XcK?x8DAVH-3ZXt%M0ayt1ATj`Gz!U@v4e?IC
zuPEO5S%efLgXy2%5(^+ufg~iPrz5}P!YQ~07b7Y_AOVCK<m}tMMJYs34`6_b4lw%t
zN)c8ZhYWU5M?`<Vy+sIa^NcE3h?^232EYRyfiMJl4=wIGVC<tC1mKOp{(xs<aN-Y)
zKpwn{+s8N!ya_Bg0O%b=1q>zTy+P2Ez{G*=M+3*Sx&nUOI2_}f@9@JB1@zy55fGBy
z=ezie{z`=i{J@0?6jHFeLoC4ux`1#H=o|p}Ce&e!p%ViI5E%C35Z2{U`0$Ki2RVeU
zAL07B!vO}Uzyc!h4D|1N3N1po$CPJ~4?Uxc-0NH3wbw+3R2FY<?@Ppxzt{h;AXspq
z_+R@S`jcG-iM$g4{~+NYNMsDtBQVgrJ1Ub54MYF?A54eV6a9&wLyQBQ0I;8u+(`i#
z;018duSkBYe<Jg{x!W7tv*HOF0B}#?9sq;~!i#VSIQpFc<{{j5006rNg@<|D3-gyg
zMnVF>0Ez-!AI3$Lc<^sJT6GBLXYzZq2%#T<ia5&y0eHK7&0qHA={2Azsovwi=|7{k
zs{~wHU_<x`e&c6jWlf<c0MJ)fQ38WVNJvQom^gqB{GEHi|MEBftz``$__=_6lNYir
z0fr&~{0av2_w+xy`-w&i@~@3S*xw_4Qt+S~1`PUD*u$a#4mspU`rU)R<39c!`_$L`
z(O>z!4HM9-tLvW)>c9RC$2<&kc>bIQlsBb^@uMA}_;Mon-CD)@F;5+Iz!;jo?yud%
zfbip0NsRM|ik6ay1}y#ki<n0cLXTq>Fd&fogK&PUxo_P;l;8lILVmaw9UKn?^gA84
z$5jFSqk2Cge-#gT7^yEfb9?ZUbnr&{>5iWzK>_t*jZJE50*JtZbtoQ34x*Pd_z2wJ
z#?WzV-v>N0nm_@=2NE!j1K}W0IO!j&sf9%l7ccbAQK&#y=KwxK&-~smCD6NI;RD~F
z8Cd_@{^k97;GoZu;^T$|stV&=2W!piSvPe`+{IMnVoClzBBbPUoC%9rrtUoCM!CpA
z?tMcC$FfBX*yMP#p!~=&`F&@{zweBz-kblFNP%v#lQ_-i_i-5Ez(l#9@%_Q#iYyrD
zS%p<D1lg575I}sfSxwz;7}q=IvjPud)oq^S^L8#R=)`qU>vdh9mu_~+)HI!KON(uo
zbqy|~W&)Q1`jEDBD8w(SEnaqRs@Gq=;wP2&SJ~IL(ci%!uPnTnnMz}v+vEiM))*$X
z>=2+kjK#8b!{=bNxd-WXR%R2xo0iK}%-8|29g+Dxz{}9-5w7_e6TULk*{wtw>79((
zRon0s>ks5J#kGo58TFHy*6#5$d7M13ffO`BvH11uZ61sq1Ma+@7S7j|)V)=WC`0yH
zpt9(>-oh|@)tKVkxvGM+BIBda8;h<W{532-MZ#f+rZkc4BzybohMt#qM^I&Qz_Tb`
zTy^Fy#NZ%fcvE>yeiPC*(A9e$Swb5kxVemEso{*4{!!3o`6R)*b6T~bp0{icUCs!V
zOjPQqciX2@WlDNKvq#LSqDE{j^1H^O8>p{{eVZjjXxb>;F2N?lHny#&JllMA=X6i3
z?m*fYj~62l&rL$53gNhEue|t$g3vX;iR(s+d;)19%o>yA@})2+SS^Xm%FUX7e(YHf
zT2)j8$Pt9y0^}60Qz4DdRWIZ7+=8}h@^pu0QNF1dZvg-EQF~M;JFwv&XT`6><0^MO
z^(xMUowf9V)XEG}RJb|1ViE?Z;Tlz2H>}q3d}Ft_ieYsKp+ex#*V^cgI6N*cTQ%l^
ziBP%cy=WLo)F$k*6x%NVMX|9BDpezWIuu1UC4ePXW%?BYsY?7zY<ZAu;_xEgTX*mV
zW0z1gfO$9d-CS}x9{}iq_0x7cRLYTvcZIvKEr?#E61m5C1S15$Oc@;UGF~P5$z;#T
zZ3WQ>SEy{XLa0legPfJT-pf4s^tr9sAM`U$&cm;SHY)$ac$Qj5M%=mgN_zj0XfiM8
zQ4-+0_N>;8!tmq?uY8#;^acNIiqRmQ%1kt3Mc*a-?%m~2qzy!OZ|s$qZylB-X+vBy
zTn#i{CIwpEX4-r=@!;@WYpM%38=mvjjLnh^a%JHfRj#XC&L#19xIto?EoAOItoBnS
zWa`)haiF$#rTbX6Q7x)4=9?b0L)6k_m?z~F^NS6q+fElsGU++;ssiPSeBYUiss<2S
zbozgGry=bNT6kfutBqSKtbyeypi0|?L3ciYOeA^9hp<sBk?B^+E3q^yaI<(`)8`~?
zwUb)|p#o>pkSCaP^9M$KlJ69r-Ie-~y()g8O`P-RLY_fh{9=zYa@b^KnC!Ye6^ghv
z-us(^c@-GEL%IUOxlx;V=`1>aa}UD0F2NfY!~^U7V(tBYyzh=jb9noYh@bRkQX<rB
zghkA*3tA+2$>z;u;4sFDX8H9yH$#ZRz|v9M%v_74OMRkw?by$ed2S!%HJ^nrO>q)Z
z!wAu~ALyZ29m2QtV&#@XhpJ22j~UjW&(@))IHKZ1_uXT2Cw*rNU>a-ooRjzwa;<!+
ztEI;s$r1*_JE0TNpA7v*&x^@Xt&`X~#K>KZ@^1dgAk46qbd#;{#Z#(ttx}JQrPkL*
zODY=`-Rfo^dB}K36J;VV+()A}`!7Qn3UmA|1`SI#iTq<cC?rGhy%TIwQY@uc%2|gd
z@_QpbyBmZJ11%w6tOm<1YPkL`3tKh%PU4BP@@ciIa@yCcW~bu!7CZK3Dzk*U1y9!R
z`_Mler3=U9I%?{eQ+YCA=d-XIKsm+1gP_XR%LnkoG_fMOp>w#n56FdF$cc?e@XfW%
zHNv;Kt{vx7YwMmle+;PN$Qd8ZgtvWG7`H~%8IwZ}Q4xXVkrH9mZmNCNk=D0{E~umr
z8NA>3pW1O$z3~FwIZd`sbr|?ovGoZ!AJjyR;fqtAsrt;xyX;``ojyAC`dAKZW?57m
z&pagCKPdd*vpP4{gnF1!<)0z}JiK90zj=9M=))PQgfP{eEQ6?p+23-Wt(mjc6?j#g
zR-@g<p3cam1E+U-DffkRBke^l5Kzsh$6>sF5phfj5e^?YyrjgyMb|Tg__Ic>y}j;I
zQMq#s^_s!kVB9_i)vHUDGHN~*bfh0#-4L!yi)H983-i#IXwlc@$7h91-P77C$sMaV
z+}O{uWM7hd_|nrK_N@F#bOI`}JV{l-e01m_h9nIRXt26lJPkZj@t?#ezDgwrUW!9&
z4<jM`G=8?8aDQ%H&^ljDmiuaw@?F_{VovsNaWlE;a4`7smr@UmKW|DfPiIJW+9#HW
zjJ%5f>HxfSofs@6(mqzu@wB%&WKTzq+uWDBrlo9&Q^WU$R!cPr+(hTFC!I{e<95^t
zMYxI9nl*93v={m6pzSR4oam#-R9*(fzz=)t)!}%<V7Q|zyFZMm7e?1}B`tv5nK*MD
zZ?V_*nP<{?S|HXI<_@VB8jK&bt%rg|&s)%Q#XlkIi}prcTgOV_(L8?Eb)T`-@%|A&
zqPJ^wG`Em#y0*O}`sH+gZU;jv%yXvpi2N(u9NCqQP1@7GM|)qy<%u!VSH<neZWg&L
zzsBXj<TEhzK^!<zQ6?}K?dLs9oz+C8SiEK|y~3^)IqPxT9l6zx8qB+ae0@DZbP$}G
zcg``E6eO|99X+ON{5A7=RP|ysxEv|(K{Y9m@JPJTjd(PrWyo=)cpByuRdXUIV*Gp<
zj?K$DaxUj=>nBWfUn}<Ml8?<tZ@3|iV`z_jZmcQ0OD)%0;Rd=r9%*WUj}zxZ$6Vk2
zQ|UF;r#={1xg7+yy)o`kQ0bG#Qf-b+L!XDXF+MAxAe&#-uR-a;3*6&-g4iTTPV4HU
zpICcW))R4}M7O#e)v0C$Mzwm_fe`y<zKDeZcs=uCXIk;Hks>jAKQdtO_mIt6`~hmX
zdFEuPAppVaJN*h?IYN83Q*LEGnM>7k#R0j6L;0Xg(317?2M7*{X!-QmZjsU&C_D9B
zw;1)L&`+xRT;`D-(mMV1!DWtQE)1DtJ98#Tz4NPp75cl=<RV=DTYk+d^Hf7~PtI?&
zZ{Lf=j~m^M!i#JN<N7vLn|q+UjJu?j*h!MoPn=_^rR9ls14h3qrg5uyXHcNDjvVF#
z<*&H8e3E>pQ>a%3$mZs(^#ywWkaaXK#=QBGVM}@6j;>euYLh#;CXR+%OWj>S`#!w#
zl-7-^f=z|23sXnT4t7mP(p8fC*65iulowZ3dALi|&$vLU^S!HP*=Y9)0SNEjb=JFf
z#}OrBH}|nMWMB1Ny*zuPkS7_jY?rEg(|(j^&5`YMRWBoHwKmGqoIDYSr0k?l1#m+r
z>`k+H7DAxesGI?&GV!XRQbx}Ba=rY)5lgHy9hbrQttd<-_@VQT6LRx~gBb;79URvB
zoHo`^CDL&=Rw6+2wN$`An?7GqO&^p3c)rqY-5Z^+FX@vVOTgL2r1o&l&D#J-e0w=6
z?sfL=`H$@OV`3%`WGC_PwM@^Ny#J|tbzS^2s(B_nr2bd{&WSG1quJ<ll=9BX6_rt_
zOdn4Gf0}$wC=#jI-In`^k>)HkGcZy^nT=2Ac-f6$wdsXs%FCJ751~}`p^4B8K}P7^
zRuOeIOr(t4T`x!C^T7i-VPEWp1sa{#`X*{HJ=HE#VsX|BN^iMa$+NSbXoDz7nsfA7
z)=6pmd-%WN11OVA1dUvANPw}KsaqI6-4O%W4X^zCX|k$LE?BQw7occ}Ko~);^gj>t
zUi!&cNF>YV>EA2?(th*nZSoc1HmfoHbIn@?9-o~cf8H#18Am8L$ZA?BV0v1!Rs8*J
zT|le&k+$_LbYME7*8WQ;6S>qWx>8GEhhzClB8*!{EV_C8d?c~8RcA+y(M$V4BV|=p
zPLI)|Rz4QeH`jZC(~;@O3u0IS6R>-h7KmxPb&F=lweF6y!FfTtb*_C5U54Jo9g0|u
zikFtt-2%hW$YKo2rfyN)(Zp^V3;zZ0HzD!l-G8D!wkT)FlGNx%2uw`$zV|kU_>cP|
zS)Hj%t@AaR5>-?}7s=8ZsEYS(pGe)VhB)3V#Q($CIRpz9Em`#0wr$(CZQHhO+qP}j
zvu)P1ZFl`iPks-2atE1lh9~0O+`0GqWtJ@Bs0K`n(bnH|5Pj0_8#TS0y1IZe%b={5
z&&ig_-|mRnLJZ{ZgX1AZTMF5qypGqL($9M4Bh@`WN~6n0bZ+*&edX)uj@y?pEbLx5
z^5h}Xk0MN=V-%J8!?MVc`%0X;euxkASMxm#S6dekZZQI$2_4zhP2)74{$lHE1qj!6
zx-=+*O}yo+VXQ5xV+7Dpr1I^-mI)PX-aWDp+u-Re7J)Ex{$YNCd+t1z0zo3*Apc_?
zt6?$~Eges9WAqYV9TgbGsl}e9gLr~z(+@68JYoj=Mr8^rVO2b)t5uK3c?ge$6}w4m
z69uq(P!|hFvyYc1F{&<z7eRA$mm99sWT#xaR%T+m75kHJqtY@x>;0S<kbYmX+Vc@B
zV~*A?R^}6p`(b0e`+Hx)m~2qzTb(pzPP7`Mz9T!J&pSna)KAOm+9uWa>0se`Sn}zk
zxci7$Tk~|e60S^>NObZbW_4j2o*q5{KEl)elr6a;2M)LNtTA(!Nbk{iIQ!5$<V|t_
z;fSTRa^kC{4|{+wUb?oQNPK08bv0`@?@}CPEI_<3)^)QH<9hU@uCrYl@33Y)Yp<=7
z$0w;t&G}YreEe7{YSq3O*`Q*RV)N=C!I{PVuM?qs(d~8b=^l-+Q3|zal%ZeMGI???
zJll9D!wHJw48wDLMCM5|5A6yF5%S<}LvC|O>j3*w+r<pUT0|Dc@N_qPH;kQ9o1x(-
za%i>$PRh$qprbK)dl21OxxoO$;r&ZZ@gMZpuSA`7cGJSX=DUz~ld_FGMptu?&LI6U
z<QY-~dF+u<U&r#55dDAg3#;gWN9YPMs4IAMb@j-^w_=j{@`^C(;U%Y4Wj<475+T1S
zusfJ%#sPZ|P?b+>h%rQs_WlRbixEnk8EeNyz%Ji4{*ImIRy2mVvq$236^bm~B#_9p
zEMGl$p(fD=;sTk(ycJ@3$~HUM*^V9}c9mpCAES)5`2GVQ-RLJQ51F4ee%X}d?|)+m
zfl9N*(S7EDcVAWbiyHDXXva2}P$J;uu)oJRbe$4znsm)-MkUwOyimmpLEw8DCj}j0
zq&O+Vc)GofgwN9B8|&%<<l=5+`Jczbo9LtY_@qxP)RruH%BbNj(hIeslf?5@eQVTB
ztU`|wr|*9oL~u~Bq|O$*C>f6Uxs$h5-f3j^U+?4$PvT=XyY<wu52t%Ep9{cx5M^FM
zoW2r;O_x7KU|N;(9qfkeVF__lHKPB9);-_Pe_h6E)@b(iB?S5{7O<=QNQqq7Z1G&`
zXTEU}QRY5F4T|Sx%{M|eHQ`IQOonsVU`l3kni}thCct>meI4z;bZCy%uPDf6`283k
zC|TS4m)k+R$r76U!_Y&`yA4f8ekVJ=>|AH=l2rC2EGZvBE3i|n+F5IE`YLyp+9u!s
z<RNs++EnicS(zxF2RD7~bk<DV%6`icJRr~}+FEa=%{B$T(#W{Zk~R_XhIgbgRX^*+
z=tlY$#X5HS@NEP$wZ5w3J$iRCx~)M&M9bwoP3TA!Y&fGCusfR2(dYqkvp$CpQ4sSb
zren)t9>+XlpypYgM&x{fXg)z7p$4K*>fI)~Pxx)fIz`m#y16_n+r(Vz#jN)6B1}NU
zJ;nT)jueEq2t6*5GJ#NP`QC{(3pr8N7&O}(*M(|xx595K@)`RBcY$ojSA@z@XV>e_
zPgxkddN<!vq|e5*>)6-$SnwFaK;3$p)Qfl@4OE*SXesOfbt$8YMitpe=^g1aj7>Fr
zTxfFpx4Vk}Q92Jig@en%OA(*-c{m46JC!gOxbXTAdTwsaB-ReUHO^V1D<i3u28mwO
z3`Tw{kNgx=<9Y^0EeB_G8c74a0gjR)n0BHUy=(C>uXK414!JU@g|EG@kcAYbF11}q
z;Qhuhjj{LM8JpBH>z%T_)}_ej=__jck)KZ#v{jFsi7L5K+t~5P)A|kgH$&6mQV{uO
z_|Ihp8ixAn1rT{y+4htv4Y%hu;j~&#J+wQYO+NL$NPNgy9r6q#dl1X@G*(Aw{@EUb
z<$x37hCE7DXtt&4GaSdwoI%)4DD1!S?MQ@@h)spog_X`lhkQG!Eva_w_l3Mz{)Tl6
zy2)L^aanRgy6$SckrYzMRg`2**6*TKTp^}u@7j<(cw7BVPN`S4`$~8IDxK;>;yheJ
zU(TgRHNLcMk`-uE-@yz{b_MOw=}vB|ORT8JDFsRvvcFmiopU0AiB3=#wJY=B8qUg!
zkTTsXnRl@l7k62iu;IyWh|_PwcH`=I+i|iJWBYS+c~TYJQs7;wYAxM`wKQB|KE8wM
z?^j1zMM>$<vxRZlF2*?fd0?>4=RogyLnRO|&I|qX@s)*{cmI!$yDy2Q;iK_5LD9d2
zAjuq~=x4r|l$Z7NqkY*w;6;!lKEtJ!k3+kjC<9bY7cGNK+bbCO-5RG%kB|mOkE7M0
z_xt8WoQ2qII-BHy(vDC``4^1~o@Ad!jOq-`Q9ZZ(@mu`<--E|ivFZ}nEYKL(h^S3L
zTGTTgDgd<x)0Pi70JJOC^t#1!_wXmdRL9vYjh=lrnQGh=%_C;=<^XEeE?yb=OnVGf
zZB7F6dDbR82o0IAQ#RZ_)!&*|*r=Gnrs*J#;iwI@z2L2S-lJ**tTlpXv!_7t-Zcjf
zbWcVQKBB>)a~=<;Z$8>IA*||~(K=(({YD^(Ha6h0cvE~G<ncLusg6i3Sy=^SKS>Ge
zA{9qocy4c!jfCZxnFq3p)tiJnyS<j@iGk=oYi7;LJO>aLt8;D47eer?o<V>h23@p6
z+44H1dDd1qk2^S3HcF{#_@#73zdYRuqsr?k&Yr&%J}rQ#N(2<txp*eRe<$s5M>_d7
zpVKrLvYGSyWc6;Wz%z3nh8F4mU`8);xA}4x8PaZ#w7|NHXbK;z;YmBe%ydsw?;m)?
zksOSN3+f`TT4qI;KXo_FlK<^a%U%oiyg3=`?JKR6>a8!sk4YA~e6>YHZG+dO3hsF<
zJxv$e<#>Cd&<ttUB>0*)lAkBhU$7F!%~Z@@2aj(5w_%Z~rIJ1H_y%n2X~btKQYTAU
z%{j|UJ-MTxI=8lzYzLd-Q7Lg!QYyi;F2ST!e<Ao8gK@dXbo|Dkbp_?}j#<v2IIc4&
zaxn)&+O27KU362!F08$l5G=z#={6Fz*F(L;un(fT6Uk^)c3pzn_?6OAPq7o5gRHpW
zDYaBjMTxP;i#eUcAn+2Kf?dunsX45`Y+iedarHt~)`cHLDpOq4eQQnlxB@IQP}4P^
zW5xqpkb<;FQbCm;NTZ!yJuIKVKa>=0f(V`@m2Jb%r)W1R|KTusqFWsvD?MWd)EDW;
zn~EijQh4Z?iwNEQe$qjlv3g-6S(tHRO(Nyi&69Stm7K;@tsNZ{)b4=Zg$|wWcl2-m
ziQD)!V72JK0B+|$_(4|2@rtmZpd+XrqRHY9Q(aRX_yB2t)3($1M|fUfd)T|#o8Q_>
z$4&^jeSpQ6rXyp9dfC_KFXXCTKPOf?<E+gE$~7u88hCE5Gi%rB%cF;smg~PCp;`IQ
zjKhhb`)EI)(J@h9(rQHa(}_4RoFtQ&5*?^74li4l1ulqe^^dyK*<rrFY^m<{`lruX
zNIn(2*G7*EE7BPW+nXO*_UMR?A|v!9J|YJ+lsZitE<;K0f&Un$zFgG1|MG^u_*C#s
zbUzIJ7L*!UsCs=*A7(yxLh1$*nWH3CpZ`8LL_y5@24CiFST0p^+IKWTS^6+zSRU1S
zPpZkpwCsS?n7nLz{6VqF-g-`}-{wYmKNeJe&oK6?>f)l^5CwZwn;rQD6pS)j{=d~V
z&i_`|n3(=oP-7xsWM}7O`)_%Tk&}^;`Tvu35%UVFfFrdu27O5P2bT?Ddq+2@yPLM5
zg9eYK1Kb@5^5zEa&ra^&8$ynBXS@l|-P`J}_KKkI=O>p}|1xiY$=83F^r92EG670(
zZw6v&WN3T=9YMh`?190-+2O&#FcHa0IG`<{Ut6)FWe~3J0GzA$pM8QG7#6_(4HB6F
zdp{_+7XVFlZ2)TSfYj*t)a>|B(1Fpx!B2c~aC5$ZkVx!toPq(61lI=eE&|0UZVoT5
zV4B(id&{3kL;+(lXoKV9W3xAJZh>XEGZ-f3MsSLZfZ6~zd(0Ub8$c^KGXjG3M}MJt
z2#vr$aX!=2H#avjWNvgaaI<MbEieG?2<uP<P%eO6-GDQI{*Yi4S?hqmr7@6^NCkGF
zPVecuGk<`t><(Z+{#oAy%>=yD+i~k&P6W&cw4DO7Vt5255G{V`&ENew832BK*#OkY
zR)51jUvK(>0<V5|ZH&xL3{EYKZXui6K{Ntv0s%-VJ;@rtGXM>Y?0?dWBWt6Fw)~LU
zkrQeo&$Ea2QnLX{L{tID?16r1=SHS6k09I(od7p}NE6@d8SKkxEbZxT49vk>z`F^)
zRr5i{F)d!)uJWdTyEL)A!QH)Y{ewZrcefAd5l`ykrm<cK$khvChWmAQWhDG<(gYX+
zl0(C@bHiZ)Tz~+0r)x5Pq3UnWA-=p5pXnd7|M@$`!NmhcvLOoK$AR8nfG4&_hG8K9
zxVwRTdwx6b`a^~%p%}n30|Cnfq!V}-^Uv(f3P$?ht{-v*@c?+!+INjX8@`^O_vFFs
zuT5_Z>@VNPzHdKGe<(@`=qTlW=|_Hbk&&J606dtRfG{>VGy!UAwrc=x+Tr<rv5)&d
zev7~DtD7Lt_F5aiASsOP_W=w{-_yg|hd$%s+Wp%>XZpp%(C+#vQE+#?Fd+)wmp(pd
za%jx#_cVV0QSbM(e}138^%Z~iw0?ilNRF+nKVO+<_AY<<r^wt0wCDV}ds>^Q-)w16
zaCXCnetx^G0DoC0p#+>0s&{^MQvtJkV#6zXkF}$jo17RL9loaEoDzXN0cKJK4VbI_
zSr`3p4rs$^$MGtz{sMhmj(Bd33_kd;?v~9?pIoatdUiMZ(`I(`#1g#SPJKrQ+6%w*
zyGm?kb7uUq#@giQ03e&2Bg>bxhj|do1+ZUkt!DxA{@l0&Femo{zu7|nm2HL}82p2I
zUT=MJ0KyRcWBP&a2$(VQgTNl3`2~LukU{!`Kp&v_&_6r_WRUhDTw`zh8NLoU&Ey{e
zJV5gqzS}hQUD$s0;D1KjJ&5l$RrmI<ME#xa@$dWluR-;n#^!$-KmQumU;j#3|0%8h
zr}V9V;H%C2PTvd5*Bn_}yn+vXo85kn?^V5W)j#}>-rMW&yZ*ZhI(_mx(C#tdPk<e;
zae)6;HueDjBTDv*eJxwG!ha`Ae1`u<w)pX{F?@pmrq*$tPwiJd^-{0*)%>H@v%??K
z&$<2!ya(m}M&Hz*dSU?H@^vcME9v;<jATE7v!@3d?r&l4OTGL4@9<atla4*2?f}`Z
zj=wVSPq01U9v=D~KXc!?^EY^(XU8}2-ciS|U>iFp-mmCCQH1Bm(aFaCg6B8TUaRM4
z^cPc{S>61!-Shhs)i-<P_inGBGEe}%0CRD%Cm7Q|!BtjaOE;Awk^Atb1K`T@B9{&O
z@XcKq#G(JFWns>vwPyOCt+%q2PmNOeQcjeyy!+}o)cQ78tBFfKPFswr5?GpPspEZ3
zdwc`rc5nFT@k|z>Q)|Mg#3yw2pK*dD@wN*gbcXhGg<^qUaqWqYtV7ulUD9z1tew)$
zMHSU_@(}6KD$k}eC|eeM#hugsuF|>f%81`y{5-szqAYm|(q)@?q+7jj3kL&FXXK3d
zJPszo6?(h(4FPC0;Rv53?QcN+vd)aK_avnYx7^qh+jj&hnQS$?Y_<YQd-g)TE6gFf
z7N<=+dhh8+Z2#EaN5>JRvI}iRTnO|I<C2Ey)7Xus?X&AZ)fMc<B?F%efR45e=x3+|
z+xf&R^*I~bfz(!?5=+1+T4TXxFfa9<EY|^0W;&Sm^k-A(zEsK1f>{#a_!lxv-HLOS
z^7ZRUq!(#lVs9Fxwd&IAJ0fM;SQq`piN~NTpUd06kvto-XmE<qkMaSr*`MMV!y->8
zsA$?f&%ZH6joF!=o~ee-D%cYVw)C`ps?56?*dRrOX}tC+(pFE%yswzD`|nyciKe(q
zyc)lwbp5|UAFSJJ9igKq(b408utHuraV|rGj=mb<uZE2Z=~Gum3$Hl4o6md5lvK-+
zDZG2I#R=q5mqV3PjP8ZGO)HKF=Ewe#zJ)M#LGZx3RvQO7?T^e~Z*=tY<u)J#xu_!X
z1hp`j-E(I9prL*j^n7QlCuGWM-?QSpOX03T#0?9`c=Q?Dkz}S`kuS|L#NdR&jsw~!
zh=VW%mEqhzWmWNmf#t}mgw0rz`mx`TuD}9ARA`k&QpXCbC`aacYDM`koqj{wMH3^?
z0PnQ$t~)gFh{x$gqTNQjs@OBH1NOu{5Z?3hEoEAIM|ARCYdxvK+m0&zs(-N^ASw&=
zUdXm1Zu2r?k=6@(@FO?UzoX^FSFwhK88RxoV-ImXgvJ#3^WFL`mY{TghGX7T2K`6@
zhsGL9-!)b)e+)+`r8&n&Y8p@P7^qtq!j+mTmXkEjmom|_AF(1Qzh+}J>NkQ-FIo)T
zzNiDo+xX04N(ukE=4d1xVjU3E|KMqVFddD0FJvBXaTUlgfwY5_aF?UrL-d9<9)Rr|
z-<#$DR3^V3%*K2FRu5(bD4!fL7|GydbP!yTbdH@G=kd+$7mGd|<70KwM?7HyOHTP5
zUkmzqdaVy4$YW&Yhh4pavY}+8pVG?q<>VA|X_yj~-dlFYuEN$P^{1^AENPXW$W^!C
zit(e)b*+`3_JOm7i$AOl(o0bqJ4>!n42W;}F0c3<t2zcVxqEG#qODP%U`$W1G7-!%
z%NoRvvncySC3u>Znea@J?FM+D2vyt87GQPST3AmkV@rF4q|x^T*DYaUUwG|B(#<Ug
zlD#8aq>$(Fp4qt33irwUIk&)m0Ianw=S~s2tVCyjdV4o3>?e{>hKpucN>EhaoFusU
z;gl&{^9ib&+!@?|l<T#tO%&G(R?_>P-Vq0%3CT|>0L7y}w6O^cOlU)(X_f5>gto`J
zge2os&8^VSmFGaIBQ%$s;oXV^rF^M6bbgF9^nDlri$9$0{LVoSDcE+`G7-79iQ*lv
zmUZi>2e_@B<Fc6*FOE~P(oIDRrBLX9G_#zCqK@W}<qUe?(DLEaK|0)jH6Rd-KIGUx
zz?j7lp8jLC>WRTv2R3L^!g29Rf7&lL#sea+r{MrQyTS6OU`|3!-1=LGFO15@5cid%
zv$4NjzB^<U@68N1j!?&xtWD|V(fXo3&Dr7$8cy^T<Et%y+W<~1UWBw$UM7UMX^}F;
z*iuJkwLiF2kb!g#y;G<y&0K|E`bHJ>)bLMpz7WSqYlOvPkrL1&Fkao@C|bwjwvSx{
zG{^ezV90#IM|7EZeeNU$y>C0{<P`5HM@+aVuGg5eEj4$t3OedOTGJf_+=X{({<&i}
zB_=eIkdl@?Yjaud0#USc>)Ev7RrH@s=VSg?h<nnSU`=+(0_M&5T>U%sD3EK#>ZzqG
z@=S&(RIN<{#3U;I;qj_c)G<n}<8PlTk#EC3mISy|C+-T`S%ldyK<c!Zt3j}ST+<-+
zDCeH|Bng*xYi0-DmZkXa22%^?w#H3FTXm9Gb4!Xh)J#9NG6u4q<@QG`+pPrP)>J@x
z*BqV*qLm+C-_QClL|5`mxO;e)6&{Q2xmkA;;+c@x4xP6@S$m>cJ}7xt8X$3!aVY>!
z1G*PN3Z3#toP{)y-AzhON^Kb3z;dhc#I(xg>Rxgz@vIc`u-l~Qgo;<Nq+}FJVPy`D
zuP0C(z1L1o(DR+Qmnx^X&yTx|yAODWW_QUT=)!On1clw~;9(QWTIx_=EdhCLz{pG`
zba%-55ujAiaOgfbxyJ~mnqv^==?k(;13`J+$b9b|R&l3by)sLpHe8oD7y6G1<`}W4
zf4tW~U{p8?VT(OBVra*%HZps;is)ONHx8*cpo5Gg*X9R;?eHPU%&g+#-&%;Q){Ww^
zVJ0l~KTS*KNF_E+gipnBlE9}50j5K?U(f&kAX9Ndc-J5nkM5sV9UN!mvh@fg<%#`S
z;^Sr-A#HW?=p&=m2;9}|gK*&s5BE^x$cyj{U9OKOuxix5y4Lx-x;>70I5SF)&1ERe
zs{Dtn!ehIq_6r!M=xY@n8%!&YN2Vz=aNjcJo6DW}UQ7dMM5FJV7kAG=${7Hpb)EL(
zx;x=`M}}r~b57@U4b{D;xJs3IFq7QbPOffzJzoSHJ*!s>4CR%is)5rn0;Vw<&wk7#
z3%HQaPj*iq3~rzJY1P(`O|DKyB6nc!1JElL;%JJz(^CKbASmwdmCSZDl^6zfrVUC*
zSd-w4rh+Jb&tGm^Sq_H8Rx}$Jyg4nDv4(6Sra6pMHL(oa2Ifzlg7;5N(@&-Zd=O2b
zCZjWFKk!)_?F|Hl(?VVxX}Zvw9#_N<uDp3)?IUY3oKENIBJU=N1*<Hy_rDs~G6oZP
zCbCr-LQ=Q^wDS}BHF>Q)E9nAHoY{_2-2h)#?DWrkR2{%Zn35QdB=YgGSK;uy#Y{{a
zhxJ3q@2Gl}yo>nr+GttUjQ#!Y4qq5u2LKkIuA{08t8#jaL)uJ(F*N(}G^$IXTi;KG
zk!u3c!Iv5H=buG}9K0kUB73Ry(p08-0sMj+ir9)$OkH#293BRgl?a`{iq{`iNVgT3
z0hbaZwPTkpL4^kd&M#jo^_eip)dOY*i|F3)XJvr|)czmmzh9M`Mv<P<RI8mwc!Y?$
zs!P}B(fc#3)1sh6N7rZNeMX}qW1{i#<#VU=rDqG8cYcG*cb|!Axq8tZZLFe@h3@F$
zF#7J1qt=vJP>b23*u$lhOG&O*O~J5<@*+hU4Nz#hC)x+Ip%FB@*a-A=<3>KYd5He6
z_nnid{l^Y+pzLOw$~<kgLQ|%Y?%1vER?$>FWCCKj5W-f|`Y$cO98R%*2(lv@z+VOL
zu7hDn^<a(msIM#=Gie8DS(i+}oK0z`bc9H%Ud6&J4*#B%*@CDRgfPLl6~PTxj7(k4
z*rQjH28IrSE+j7C=O!K2IR2=23QNF&ji=JO*~h@PJ~KZ+OTHeg(#`EdT3r1?CtJ>E
ztkqVu7k;N&hJ(U@YB+UHlf2=%Lewx{Kl`{xUW?{e^j1=t$x<Ai;zatB+{BmKk#0PU
z5<Qyor61Db`^nz?OJW#O6QFOO<4phjSqU)C9)<Qa&2A)i(hdFrO@syz2AvcG=d%^*
zjG3o_6SSle9rYw+KS&1^J!J<v^t-@R(%3pJj1EYTq{WlJF}P~?Ik^>k6*YCGx57Jv
zkuQg5xCREAJY&?<XcP`|OFdyrpwd@H$*<o7`$&CpufFp;MGS4?^D2{RCcZ}B-A=u7
z<GZzQS6gJ7gjKQ+4vhoSB0j9mPoWXmK#PX>{Bn^2y4;YlM{Ly_!kqkSpjz4Z=+O`@
zlkE^GG1F}r`2%AOvX%5^*YO^Hko+r(F*RLmHC2T4#kO@yLfl(w{Ru1~aFJ{3TR?Vf
zvH}b$Y|Av)r<Ko?<P?pks&-t4aBMv)L(y05t?Q^cHqo2FJhSq~e&Jv`pEWK?9tsol
zdqS<YK?;2}Fd0Lb%>Q<F)_fWHLKgDlN`CI0IGi?r{G_Yw?5JjnvczjFca11J>{}DU
z(2ClaVs%Kpk{4a1K>GJ0`F0gL6Qy08#!_oa=FE}FBr<P_VPR^XPfabkH1Uz0j8~2|
zi*vjfHWmD*#3Qf|rh!6??L%Slm>|2bRKvP)rA8Oip<q_tpFR1DBY8~g{ZbXPcVIEd
zJ2XMHWEtQKh8*x;&)BVE`)YM;<XBWZ2{N(=4Nf`gL6(FWB~_G#Hn4E2l<2ikqYL|n
znXR1=)A-^7w1~p!IrngStWDGatFa#Gf@fqAAH!>xy~4mSU6<lc%X<waBkbEXq_0He
zNmjU&;Phj?VQ_R4h!lpBDYopw+<MhW6u3JGg!|`GU`z9C$S#Cj&%M`LaHJWys9+(7
zWm-YN>)%UcClUMMsuU~>eD*MLhC_|TEb|Uuzw(~0JMEz)&PYk~FC;TPn(;~f{p>Xh
zJ91XC;$JP`qvL^<=jp=zr5%?Qz5JQ&mnZ(gCSMjBve4Y5VS$s)8*DYwi#p_%quM?s
z0vd$;@#vEb<bSbMznq$~w5P72NfULVCKZR#SPgEo^;y4|3Bm23!e2Tqa4bP03DI!n
zskO59eSf5Re~yZlPpT@*HDi3`!6s&=K4&;*#(pLQHTxY}8uz}}?%=tlVyd4AsI)m4
zC?Rk@4I9nANHG#2_|e_d2&c_&o5Tzr1?`PagUKfZW}O&)NoZwxYQmADx9f8QhWbud
zDns2IJm1tE#DsJNvERAP91xm?1ccssP9nT7AL_kH{vtrK_j1BCqGj6e(ko-tMFukL
zlQI`OUiN^K1gE;qa8e77_6TH=7XuC@Q84it-!kq8sW-~Siobz+RBfY4bnL8TZ*8Ct
zlj@!Gk;ArCdyfi3tfDjn{jxnmM}i&Qf*f?~R=MZcIe90+6U;on!pnZX&bY38$w{8w
zmz+U(|Lq_rm9nF)54lWHw%)tQbB2}1EpIE>S^8ebXtg^y>WYP~WwZf|?(|Ji+B&OL
z=xfu7`QpU(>_@Oa4!@&8>T8X5`myEjU<q){flVSIuToKECt>W%Xtf0!C65dU1rJ18
z>w@bvEpiS*cKSmufrsZs=GJq%iZ&z)Q)V1DvF9<G3E~}jpm*P=yszyLg4pbCN`1vU
zzl$H%z@p`YAH68h?JR?XVoZ1nd<d%&6yoL>fqcQ*ua05xN?%k4GUgQ}wZu5m>;U`N
zR-=gj%2wJ&V@;Rc;~I8rdl~p>H+_R0_f5FlPd6GLORnXs#jM1+emPOnW$m?LX=aEq
zU56ZDGCn39>f8sJr4vyTww(Ekkw)?|5upaBUZ`C9*}!0}_OjgD6B4yRjNBD5h_RNw
z@9$O1!zu*_O?Hc>F)o>8`y#XX`^TgrA<D(RN6C|!8mU_jp6Odp`fM4?((n*6t)7F?
zb)imd&J>dEgD%#YQvzQYD?C0)Z8?HZAd(or^XEKpd1959pZq8l|8+7oG0@3|2^?)L
zEkb7`kS(C_rtQ*KQ}btoJz^SF1SZA&{3#>x#2U5FUU-9KodTTZ(L5IEvPN6m$>;m-
zDM+4$J=#!QaI=4}Fl|+}*`A*1BA8>s7P{&~2JN9(U|A1}7jv*2;+5_DgN6<&BVXcb
zhd5o#j5IX#_QJErnE7lq$eXxD(Ygt$K^f^4LK7I1&BVJdYA6Ly<#MvF82fQ+QUi*C
z?7c2x;jaE|o~sWirXPP$Q`oA>{*>)(U03NJ9awxEyPl0o5G^VVVh7+oL1I+IRu46t
z^G*&s3U$nZ{!ZIwIwj|(Xq*_xGXkp8!Za=ZK=fd$Pwbd@V*vu~+|hx9^@H;~sSED5
zi2|F^Uz3UlO^?i7)`Ab#&f<5L9!V9v0w-U4dn!@_j-36tNHNCqnQ^TIXb0(=q@<+y
zh-TzxbfBgC7G0C22=ms!5!+^T__x>lp@tGDbNwifvuQEW$zgA<?d<XriYS+wvv06s
zg;$+wXjU0ek%LYB*Q+lop>-sdN^GBA==p~E{d?(W(6oiCM$;Yyou$dILg-QhBI~C2
zU=gGTjBRqrIqgS)EQo#-u5~vmkfEg=WB^)Sx+V;GdmV)Fg0=^I=~b`=XhN{?Rut^q
z0}y^M9YZN$9+MLv<S%t^a8FKm(o5`Jutfz}Ey4HBn1&&!pH@^|9&8GH?Oh9&r4xC|
zJ{<6Q9EBa7KeawbS~imuIff{xhfeIWYmdPrNSH2zg2|b_Zw-d7$*~C3jxK76fyh=p
z#8s)O_QyK7_j*W`{T$W(`_?TZU@<EFAVafQdhPmr$rHNPLrrCnm7ibEj&@Cm{1rXv
zq@c{|kE?gTX75~{NfRJOJ7XNmFyxt;5Wz7jDF*Jl35@U<mchs?Q-?2O{5cn*JXo9~
z1Qb^nn<*~aa8hkeOO)X6z4jdKl&EXrVs<yYyVvnV#wO%0Qcbu)4zyt>RJ*yrbHuCk
zuFzD=vzOWC6rmi&I|6uV`&%RXa;vF-B$FSrMIOE~uX^CN1VEIemZC6~D<k2P6H!e)
zJ&k3mD5Gp%-bNdyv$p9`$#$hMwl`qq3D3~Cnbowy<$CY!d69_dAOL^yntI4sI$7eE
zsQZT|F?4t{w{9GcQ<k#O_c0b^O45bY*%#liTGXEpp@KzuZ2gfzVV`bV$!z#{u4Jdt
zResBg70VLKbJL4qzE@o!qy<HB<?@Nt08jJ${o%+Za1WrTG%{+S-k*$Rm75ZI2zZ1J
zS;iDDPjKlU{>OM;#k$lyurJoeg3)>)*>Y)3`Sb!<58poCutB|#e;CA_gW0sv%4nlh
z|0#QH{62BQnZU+Zq%`?3bwnt^g>RhI(lUpR6-y^K_qiqcZF#$7GfFys>}ZzeHwitK
z#5>kBmqD~_qDKn&Jal%!Y5hN4y4(d*w-LTwKAl!~FY9_q1tW7B-|M9k^?iWWHk{^3
z7=>buPR4Yg7QG`RjtoKPS`ksxWhF=~c0vix#<pc?R&t6e?zaWz4pO_mj#tR~`17Bj
zX}272IX5Wi-Mh)4&mBvrG|z7+H>Dbb0sFS*TQ6o&g(Q2GLAh;M?Dqg5Az-SYjH`A2
zhG0*S>b)zoy742Y9e1B@LWtczTIvOQs`OeM)%eM7rN?#|<0EU|!9>`o=a!;0Gi}JJ
z;j<LT&xXo)<GP_`PMY}QNaFp0?zLl2Ps8sX1;VYla;%r&h`aAB->lMC);Fmr8x7&k
zg`PbwghUkxy31pT$dMh(k0Tc!8yJZ7q8Tvdnr#y!pBEKJvl}RS_A`5IJ2s;+?L)AM
zA~_UV2e)f-r|4|!3Y^fxMh@Gm`BM6KU6pv$m{gRU@f7mo@FGCRb0%ZehgWclt&2U$
z9yi^V%vAgq5XPRmG#a!P?S)=*%R)Yk3T8a5mUa_JqvNyFYTVqg84e_`@HYnOVZq{R
z$|AOEijFgtUU00UYOX;Jk$EO|Wq{k1FsRA{Jxvz{)*lHHcKA%ElRrJj9e%DDmMixe
zx^GEfQW9f$vS&2vyR2SPA(@gbMe)6?2&;|@&3Fjv4`UXzJnmtu-FPq(qe#wd45$}>
zW8>mYRwx~ET_w^I$77$`Hm;>}Dp^4;;E+h5N70B-+VyR++FbKfH`+NDbX?slt9Y9U
zpe!KFrA!2-*?LZSEuK!J_ean?pMB&x6><rq>aN~9Z@loUKd^#bPXX#0xq8{fHb<gA
zqXyMm(2Qvfh{RZ<A;ocivi4T*GIuo(*MY6Mb}K8B2$?78iotrvzr$gUZ~xLC0JS?g
zTy)>s6Ga!jL~!vX^JIXD2RVF5#SG*C(G^v+`SdnsQA434XDpY$I;yCFWFaT4p<EUU
zK@={uh(9vC-$))WanAq9HX)m{kzUE*@@h}tzJnnNEyDAEDohEF=cSk9$Yv+;b5}_}
zU}{$J-qz}Er~pXAT7P7j-en=;c79wqK(<gG<%yx0c$;ifO<a+~`*<)1M4WvLi$0v#
zIS1Xht@|2bz@7c`ivtC9w`13G$&DCCpXB-Sa0*XoI`B^xLvAyfs!O=f4$5*hqJxKf
z(O>ZQZdKziu}PYHzaid;uU!Q@d<MYt26p8Xapk%Vvpq@z(48_g^DSV>Rc$4eu3HZQ
zLQ(SyM6#OG&Y2Ybba9IJd$-VS*)<yO@zaLvypc084KfsXbMj)>{|n1TOXKy`vweo^
ze?+yZY@nIL@0aJC?%K53@5Yr5QBiVZJ!3KY!=*eSaaes@B|>x)>BWp6Uo|FRJS8P?
zX?ty&?n@4mbGneEsBWvDu$L<;Wt5LiJ<#ewSx}P)>jN*i+&V((y=IhBF&rY0>rMN!
zJ7R4V=>d648nGnFRU}d&pw=j4*0_eH=X9`SnCQF$pvW7-ervy|O!I&xN(H~<%w`Ee
zt?EyN`H0w>1&pB-yynLqoBpkHvWYxhHxVFeT13$?_ShMTRGc5miZPok6Ewq-sej5d
za$F*@KNe(9``M~b^sd<jBIuy-*NK!vSb%Lk^Jt#{(~#){M}TqT>MY8#@7_j5coPaz
zAnZpa2x<u|)f^h@U=!tf4-)B!*%GlXPrbYl$`1i|Zjs8-uy6By^}wKz@^^f*O8p@g
z30yubE$cj}@#SKPW)OTvkB+C>{@@8zt%<(?F}+SqaknuQtMSJOYcoZ7r;^^;u5vOM
z23l8kUGD7aGlo;_pqe-~Sln5f*#7VC`>dDYMe(+r9gVRY7Y+N`f$rTa5vdS5zZ@0A
zHaTn(>d(79HVNk-$5ZhYCsAvZOxigu7WLJ^VvmToBhIy3zk4`)Otu$AeQCPC$E`yp
z%f&&fb&X-`^mck@XyBIi{DH9*8h{sxux1A&zc7u_2tJv7MMTGHU*Tn+S4gck2(VF@
zdM}S^{;QC6$_v95elR()FMVYT+nV-1i!m|{Icq7&NJ*I&TCwfK-_3!0m5pjw1~o^1
zuEi;Y(YJWGDPPw^j`P0M&jUyX=#!`=8>fHSuvD~-nBvmpAGv*Ir!-u)6aA3O;9NRY
zfxBzp;{(J$S{0sfCT;NswRHv&H8Tr2BHl4-Iw5(d&3$nO5qc+zE76F2as!a2x$rvV
zX>TX8JE7<lhkRZnK3M~(r4f?gh9eD&Z_HJ-AG2S&E(71KiNAtzqHu}{%Q6}>@V`9S
zL<coh(b{FIY?q}|Y-^z>7b?n9e@H#I5ZBIxPv(O26YwPF+lTxDoqGCoMIe?NxMjDt
zp;EGSlT%{-zI`}(Le+>~TP?#tHbi{AC>NPV9gmUKr3y`iEz5t!ASNpBm6q4ud8C+f
z$l$T@hm|jSx0^YkA;O3(qqD=j98*C?Hr|Ltw_$CzwMU>_&`e>&D4W(pPNz;VLQxL;
zJR+O|d%7~7o<%es%lH0(DFDc|K~v*`F#f4g@lio`8bWZaSWgRcbhk-dZvnJi{s;mf
z2p?nHSt6V%_Jb&=P0rd{8Y^`R73QoyY<}9ZQ{_zXR;)x7CPS(!E6Zw?$PxI*oSPV}
z7xZa&H7<!9tc;mE%EYSF){<I$KAY79$bm;Vzgh-AVlt2D8j3<)mviE#DM67C+^U#`
zqH|?9Z*g95QQp&uNe82B_Volq$3i4do4;==35rjFh4Gsh_E*{>%g&Q)(!Zd}JKmo*
zMN5{TG$U_}{ba+tpvq!bB%XL(G+d?4h~ug}R=p!-I4XCXCX4DjbjPSTU0olS#7&uc
z8P}&+PTvU{Vhe7GMjVI4*)c{gK9ua*OF|c5R`|>kR-*1@eZ`M}Amy-jP|33gzv=TA
z6ndDseOJ4P0L6M3cE>pPHGqF)HV3)wUddr_J|LGMKAF8<WGTXG5LPny49D{1o_NF@
zK($mf4Sn1s?($uk)fg&r{+lUU*mYRQNcjhDBdf6(a?X-N08+|$|D3GTRctOM$M%^?
z{N2M_I;y1Hg=K5oP0WtU`XMQy6;)EX1R0Wq5}8KyQ|LH|wpA`RfHHtSee~Mb2XB|d
zIe$Y?Y~?NzINqa|(-G0l5BFGNfg;c|$8e&}A;ssRQU1KRR5x&98darae3rJ0bL(8v
zFavThaiPhr;c&EwsCM@UXD5}>Dm^xwX9o7c`Ha5RJPbGQGlf@l{@fV&E1kn1ZV8@R
zUV7f;_N(Y*L34?qM__#f;uM1a0t*#q#2bn0_|5*WB5E+oogfo(MCt?PEt*x$q%MCQ
zlNS*#W0R`<#n6Ten_2Ef?m!RG+%@ZVCfT{BvIjRhq5#4l`6~lX#>@$cUh*P>d`<Ox
zr-1z}5+S2E1Pfym@x<OMlC+i_Yw(JshhFN%$%gLraX$>Y7O;Ad?V!ywEmAF)*1wkD
zwDT-faR8QpCWq38@5!wjr*|gGnaQIE*;qv!jA;d-g%_ilXFHT#tZ40mey}7pEiOrj
zIo!)srYT7ZyTj_uJ?UPA&Sq@ppvd%bS5mw>!t1Xy6TmZAnPEHW{o5Qj_8{<sMZ4$1
zkTaHj1dX8O0FOeLuIXKLc>Q+x6rJf3s!`+kZ|$NE9OI$4WyluURvi2Kl|vkkuw9v-
zfPWW7D9r5{DaShY5>f6l^V3}QcSLF*s0cXy+7dhrbi>jO3mRR7X9LQ!Nq@>zREV!G
z-8+FgaXXZ%g5IQU6wP0jx~;kkJ=_t1TjK$4Z&I59!U^4;WO(^jHp>S=U5LoQq?}hZ
z;k)WrNn2UQ-4iu*{y@~UhT~8t9)5#FRf!HUiM3>WE>kD$5w(r?SL`OED{;j~TqDaT
z8I!_fR!9@hcm!l=j7AJW!Xq^S<tGpH`=gwQoDOt00m9+yEM`uRY#CO_A^Tyk%5YQv
zq1PgRkIH%1<OI<_k)K9&d$`ZwnR<CnqgEse>c^AyUjp^wCWSWXA|XF94n9B-18z%e
zGQR9z!^GYqv=5OOfV+dk*(@Qo2vycZ+nndvCe#ZH%=zL-=#tWhE~(r<Bn0b^Ii?rv
z3Rw)#s_(~h*qhv+m>-j^cWSWHv4%}7^LLHUZ|k>Pun!?{`7No*i8oifx~6}SeiCMg
z#X{_fVZy0BUk-)c>KfPux}?@wejlI%tO;$(Xp2#Ic9WOyK^8}i;-LCXI`-s#OP0Cc
zE_79`XD3uHc9clevp^3HZ>><^T=1H_zO64ta144E=aOyoyH7a+_nL=O!8huWoj%3C
zYS&=~Da1dhSy-D+k0pHz4ZhyGE)W0iLG@A3rzv4TE&03*?RZ>myS&_xi(Bx&Cajap
z_&JD5u{b&|@|D>T?>1~Phvk)h#HP*AZOJ_$tGr&My7URC=ApQrBw|A%C>mi(=p<lB
z!v3NeJ<3_V9e+3{rQdyiV^&-{^wrrq_EOgN(kTr_R8RDcK>Q9iTg~L-TpOHQP;6)6
zCSER`<S{YIS#F58eOA@g#dhgKM<`t;BceKD#hOK?1Y!8hCzs2TnS;uR(I};+Zl86a
z0uiHpF}TLs4oeT4V+ro37cv?P>&*Yo%hJjX#Ej!4wCNgP*}#OYyB1ugi-&8`<7^?H
zw)<oFNfwCBm6~TZP)SVr!CMQT8_7mD_}f^Akm24voBOl;4`(+#6<=d#%??2lQ$q9@
za?0>yRRf(l);!sx_!naC{9Q&YG!Z9)(O%f(L$BF}2V^L-YXvXJp9Tts5M{{~WqiV)
z`zu<iY-51s+9=t&m{7Z9hm(9KYcc8&9nLT(0&CZx8_yH$B&*@7+X}*QTgmF%X2>V!
zq@J~VT%j-l3zesWoX03N1_6uf6X<;(zuQn=Np0#0OQiI$CrYZ{XPRQ}4T*s<A#*$S
zSdZay*<T2daMB~%q|G&UBB`V7#gi@+*QwN^Tb?wG+kA&5(}wh-I)io6sl}PKv|LQ^
zb#ULX0{LlP(4GZ>n`rmGoiq%tFxyHEL?3L(IA;QjoYOw0`lYKFwY>_!_hU`~u8?N8
zDG~w9m}sN3fJKLWxDb1aadxGrHCN;wG9rBux*o&@_6kI@uK15CUMaCQKXG`2pW5i@
zgeB`n__~vw=Y23PyV_w$);}lp%Y=&aQ<;WTfa&ojnCwgz+fA<2G*1@18LcI%IX-X>
zJCTo0?>}JHLsgk(;ldFz2O*Y)FG@<I=-RJvt1q+CDoW-NdPkCRTxP4c1ma1;9<2ll
zo?5gahlFg-+quJQM@<ioTXQd9#ABDL+R$reancBEmOa1R@^3b#kz3<tm8ms~5?wIs
z$7m1gp!c=$3Vr%Aj{$Prju+iptV@_Y@n**rMwEjm*i(-1nN%)J2$;yov~bUHMK3?q
zw4c3K$Km?T_M7e%!Z=+HqV0gms3#hChKv_-sj+~6wLrq$$%7cMi!4U!+AkAsn$`gk
z6)*~Bl9|UTNz;@mEDkjaF<DGl%r~9lzS5_Xu}kCfi18&@g47H%fev6Ki-{{#e?p4t
zeHj4U3U_W!E_{%GZ@UBc-va*%+h7kDG~$=VXbB9MkEU?o1>01p!;heI`BD#xr5;}^
zuCGXiWMgu_S&%;)uBc}7mkyNPb~y+{sBfRlKUkPmmu4q#z`Ru-XRJjY!{^jBaEcbL
zjxrW;y1<?&BsQW6$!Hx*44tOT@{f!Af1&3-EljPP<Yb4Yoy;@J<G<bB)x^nVVtg6c
zuRqEAcr$vgyfii1OG2#KGtERUlD`p6@I%V$E-Hi@)3StgtH|fQGVzj31(qFZx6P<%
z^bvYbDQbw#Q5<0*U|MQTbW?PfHVSqMG|Uq+<6<!0$Z?tqMEmnFnxGyXl0aH9()z0r
zjec$#bK-n9%;tAq?{jc1**~F+HFLIr+W4O-Ed_!}wha?1`u7_SI4D`!<?DoiNzUIi
z@KR>u6QJxaDWft<b(kUP-{wv@h04)fv~6>uCU_tt8#q~T<zj7VVgc8&b=o^@dp&0f
z=VBU^eqN#R5QE=OcVs@0yeO;!dRydy<dOmamb6)9Ged>!gPQTLc^;A5+EGowCk4Kw
z6PNrLk#Rjv%CEf-3tk8mHAbz<wrIxH;@K=osicI6=|?cp7}dWTqx?*|x<Bx>;{)js
z9$;dG2Aw*!X?uR^pLwL-`t_Kxk3n+A2oSIP4$XiG`Isd*HnR20IkY(`rOm048qNdB
zE7t}Q*fUzzz!<?y+SJ<-^Mh7qNsF%!^(U?vSedZeI46|}Y2<FJ@I|~^yMQmyVv`3S
z^p_1Js~09N{qkK5l%LCo+YpQ5!-{IwC?(`-phubeUMLGzOvo4{xCH%D*D-XOo7hUP
zrZ6>C<G?ujw_q-6ArEZ&V(%6~Vc_6br6~|wv@NV81w5(kNkS>?hC$=OG9d(Vl~9o<
zd(97Bj==>&-B1$LI8`WUHD!}T7ZCiQY!C8sw86;lw2~w|w>7*mqU&8A#|o038-e+8
z#~%D%ilelM2XRH#gh4uq!}BaruSY>~fTfqJ6;vKvm7{SInJu_7WfE_m*U+X|^@ShE
z^8c($sBLQSR4r8Dx+oN)LN$Qj%6r_9rW1KgpEmtbu3}Yj#Zd{>ex}%~+r-{nVSy>`
zV<*MNz>g`q|B-lJn*^sYKiiLst-DN8*mBz>oA4c_$FmE6(cJ$>%gC<bQVVqqA+kU5
z(NRue#?&&w6TMF-fx3uHyFVv25?%4hV1Hhi5M<6*GDY+@lqWAn>q({;rn3F05^x~r
z@7UOYWV?_5jdH2eZvyC?!UyG`?1JGIk*(f|LHJsbebQ7OC{(oTih$l40Mb-uYw|tM
zDhR#WiC@`JNJ%MxuN?}+HTr|uEh0UI0BNWl;**jpPsO{&&FPVT+ZxDyG=k^~t>D3`
zc)%vY+!O_}-8ih@k!3~rF5&W5r=<vG+8!Z9hHd9B$Mg<vTe0;)2P_e?Xh!Rq?~VwI
zaTvb>&X<aye*PITpTdb+Lf+quEKp;F`CL@SOX`<u+}`2oo<*Wc4e5H7KMZ61t2M^=
z3@Oe}y<9+A%ov`R6Svh+6MY&sH&E=h8E6JJR1<yleUAlpday;ZBQ5<ZuXW_xd9p~E
zaMBW0uyw;HiBV52JV<6dgm5P>HB7Y@e)(sY@jYE?yZRYoCto;Ao6?Wl(&Px~p&aJh
ze{D8~`5r=YGvrLFoToDggKrn2D`y@7J^c4?H@hO+HCI{1bA0ebQd{}DWEm;bOB=uF
zLO4VYjMM!)0+cVPWJc8kIM{;^ygPQNRCi4k9hgLSLCEiu<hjP|2gw>FR3-F=7I9tT
zIa4Yr`4hjRnufZ<Qg(A1p+0nYbn#wv!(ubO%!_sRGTsZF6`!q*B*80g*D`H=)0_FX
zxJ6RtC<GF$ShQxzW6#tZYw>NEi*%|bhsp+M(lOd0wp!GdV9UkyRT{)^@phoBirx&o
zHm-mL0pA4$0sqZ4SX$<DS)J*rCau3IxbZ1cos^B~XGR>N8airV7^MF~8l=8l0_Shu
z9B1^^EQhMXPMXkq0~6EFa)iX`Jw;0``|b?F<b8ziBL#MYHd#;=%knRDx*?D3OzSpB
zBQq&?2dcDJ!!FQ*`Z=S*?K5hi^GvyY@$vhMU?YSl@^OMXA#-I2@WpQ+7<9?<q`R}e
zhGG}WmeXK|u0SH5hwj-}AS)kwG9U(8YX1}F)MQ<}Z?OHFtTowq3q)ud?!P^Ps)354
zHeb?pT%XeS@a#g)U+NNE75P1DZJl1?vu#}n14hij!tLJBi&sA=ce>NAeQf=$1}Nr*
zZ|Se<F1=2m*Gus^8^e(*Ygo?gD6jKFW&fdW%Xbe+{rF|CHNtJU!YtE?&de_v-|McY
zVK5C}qpfbjTQjaf%F4VzCl)+#Yct~SHYR4C2x-YU<t<L^o^pE%zW5{{L%5GK01D3z
z{y4?sTNghRP{Gb6Ji8L2kEjp|iP`aI1M{DgM1(>&38w-|(1z*WzdJ98#13E>S5xXx
z>S4(S0b%xW@ENHmtwTaZaxauNeA(~1k>rWovgIkC?_Tf3J!mp4lEe2Xw~FNNyH?t-
z`RW-~@8S|>Z>(M#t%`pVZ?$NDvtydDaa!rYJ@tyNFeC}{-LGC0tW8Z5-T1n0#O(3L
z)U!PP3;gEBi7f5|B-R=VnHh6&(kIrXxLfATj!*UXe;dy01BJ9%!j0@12Bvi{1NmmV
z`EyEfL)>mY>azGPqmc&Kn0-mvV>lDqqELb+(u--8rD^I2CwgV1oLVhmXI1hSmtzEl
z8Drayzk*c@AM(cDTzj<i=y2_<!xKXOL$29P&2MC+ESSTBGIFuXG^pvA-g_QITcQm@
zeTKo7JcMjj0<_o93e9Jr*&#?exp46FUa9`5oi_QwRfe{VtGX0lh#3ORnSM$E9xeYc
z1LT!_TeO@J36`fT=I@#mTeQ9$eRm1Ub9xD8ZC@<?jIdRpkv@;=L=OaV-la@F&(wir
zeEbZv?{|;;KOd)1R*CMS#=c_vpI0xwxg}I%KZ=33Xr${ji-lSgy7C>>1mS@yje%_m
z4c@5OU{S9>oJs&5khe#7EKidkie!kW>{x%y>e3MIxBNbDmHgWb7dfVKlFiVIr>iql
zH9|gL^{&~s9EOcXZ`0k*-$3Q0E4{ax;oh7<C@PtVWGX+WaUb}i$jLwL@}Bf+&j0Qv
z+G6UI1Rb1YW8E#_{+qMg&%#pnfXqnWdPk3)lk`51GFccStQ47krmRhAc8A0*`kYa}
zruu-rgrb>maL(MaM%6<4$_ptV$bg(tb@sfAK%e*;PKvmngAU9#leF3V+sNfqJ3EHn
zyOO^A$OF=O9t+NK)avPL%`J%6?4}&G!_~`iBZU^Tj4ESu$}%LovQ~@k!Y*O!_WJwo
zYfoj4#5A*}U;r&1FbKSh_hE2#++}QXta(5%Vvp{2!4v|1+69wCOrvFU*F_o=H?iuy
zM<DhzK$I`!_JDmMy$VXLSH(pbGNIe(+)%C#O^u=C&n;C+;n{Z{@!0HFPWu<f@8Z@W
zz%_0h4V>I3Zr(guRE(`_39lDn^9tFgtkv@pY(cfEXO4wl4?0N`+GT!&kA&FNZa&7|
zxN!cLF|nEAshXxQSw<*vH_lq|BahkD#s7@My80cM-=K3eJLLPAxXeyUKx54O{()O<
zAtk5J;0!4O!rOK$33+Y7J%nKI&?Q<@MlVyFpQ4DR3TkKqB1$!}A+NxuBtF(ny+X<H
zy<2RR5a1W1^@;F&)0Og6ZR1FX4ZU<`yGKe5pW!fS*lss^$>?FwJPAda>EC5G#d?fS
zh7hWr@L<ur&#R5B^GVev=;c=P&fqoYUX@%HWX@Z}iRGpFAHvSDM;9&JvSr&gcNuTl
zwr$(Cwad0`+qSEA*|yDllJ5R?ll~7YSx;u>9Akp=56KZ>t+oioB5~h7{84RF1a7>|
zW4GyHefNF~6QxVGD6te`P>v4moXn$%Ta4suZN~sTG0AO-@Q9tH7wnc0L+hT@&`D!A
zfR>%)+mvs0PB1phf&$G?Fs=_-+9xN{zJuqi1@U7(Ag`i8M?e#f*K>4RIU|cGx_r$)
zsovYQTBsmWHHqHYC7sbHH;N}9`|Sv>Zwe(*ZWkPR&JbYZ1^$wtG(nP(){u?r$Ttpa
zPM0*B_VE>je}u3P$CRl^EwWIPJw?Zoge~bHCX93vbJq*FG_mxH)Fc=aCDELgz)Z+`
zk$a3R>NH<akeN<Z2gEa*hIDLAGxiNMickS<CEmQPfifLyWs;U0v_vLy^DGN*+<~0~
zMp{P>1uIh=qNvVvx3;imj9dcnm$yiCbUSIASH6#q#x;vqhlqF$Kma`z8s3;hZyqjA
zjYEmqtcoHWe8)pQ=<%N*K|hvB%#@R}wCWnR$U;a$wG|K=)+PJ$yoLoec^SWjjM#DJ
z!LQ94q^xW3C0@$^9$*LJ(X@qr@lR|W^gX$Y_KE#c(9@{q>i$99ciM;Kq>+BWS95{P
zgOv*x)GxPaRfp_$;Mt_Z-ZkQx{$0RU)blHNd414uSwYO#mOKHtsaUh`fX5eST9y;p
zHMQ66+E(P{B!Kn8xMI3>E(LkG(E<`|u!d}UMi#XF&6nYZQ=Jy8OCUps$qiKGlUu_r
zgS729hN81gsV2!Fup5>SSqJ`H!Bsc+7j^1jHb)ZTDm5^T{4n3VeS#zCR(uFDmon;`
zCYXT1oZq`faUa<kywkp>LNUu>XUQtD8z4Mg)_DK()8mMXP^|Mke#63XIY#PF9eB|`
zM$h5=U5egkvfZgnwY1a{JT|M6K)1=&cXuF7Ow)T|S*`!i4D@qas)NbVdiKZ4%cVU^
zK)uwvYErZqQYl0o{dI+l;S4D!|IXpOsDEtAH{Wr>Vn$@=q@XiOkD=V5!7l_7*B68K
zXXt$81`;%FUG}Xs9?5_=liIklwX{M|=+kq5h4IGp9O>O7Nw%f~#B(EPLuN|C|3%fo
zofB{$G=RaYjlUWrJn$3puj}%uYtT$r6~68reYjmPYb|B0t%I@tf=CpulkzCTGqqh^
zoKN%>2)~g8-*=elsvrUTN&Vit2al*qSX3o*FXAqRqy;6VEVxiMHoVqqA5)Z)J_dpO
z=}x^eFMf_SN$z9mZ*_*35E<<1WKhv^-{I?Irj)C$-`<~9LcSv--!9qC^&@oxB)Rh{
zL*rdXFNQzQ$ox5m2&_2!Uk$F8=t(E;8H>)of$;}*@?vxGMV)RAWTt%<((|?|2xs!{
z|2^>|0-}FOGG|T^1yuW*Y@`VrlF%2B9ye9S|MYU?lVS@+NRV2j?#@%Zf(b<?X$%bS
zY%mb<jNGO3ckJm9_2-TE+W&Q)rKjXlJw5lenu=05ccVgT*N%pSxUs0ABjdo%aKQ&f
zh#lJ+G&U{03le<a`D#9v4HX6kUqav(mPP{?0os)bl$Yb4azKQAhZ`!f?FL#;8ynf%
z^9<SmJ}0MwESH>pO^Czc-^C^FUg|KS4<dX+g>~dVLY|cY-m~F>FC3T^bSI#dy+pCx
zD+|WCh@QR&gN~af%0>Znj@AB@ZRIJfd1k}x{bYJ#(%{7HjY?)A?cJ-l>mLSn(*?pg
zfTvKi;6%}jmfa_D6ct4+Pg}fkx>mE|vCkZ4rD7$t>%XR|=e?27WoMnP4_oL(mPz@U
ze1deM@liBcZBse1)xi+5V$`$Cu#0dJpWQuHylL*SY1BJTBpl=c{qwT%YE);J1bwAl
zy~WHB7YH5SvZY#KYs{Dj);0hmhN36cfo~7t#Bkh$d0tLk*3qO?9`-5@-A`kNfC@$2
zE}}cK->|^nRqwK-qS$kXe+x5>_vombj^Ou45PiNt#rr^HURPfDt)B<kFcjAt?9jv+
z2$^yv=Q<j}=`6b#+3%taWhr{pSg}8y)x7EO9Y0MrPlj>EK2GA}@i*oW!J!lh0dX@s
z8uC?L*Gk!tSDc-n$#GD2r}7&;v9lkE&O`6Fa5%mE7a@Orqb@>f12hGazP?Fv^5upN
z^tIDAr1-|&y^5XVudU4f_+WwzsiVVmDO}RCp%%zyAY%8%>aPoAUi86_5ZAB|6;ylQ
zUZ7D^V{rh{5m}GnY)CEPlSderBUzxoyy2Gd1t|iFTOMiu=T9FM=e|Avs)1I_3%OCM
zqO3-eqY<u>9RK_)(v=tx_uOZs;^#K_q{CcJxJ#f8y48H%q-6__DU*Z!IK_Ep3$v<p
zHy)VP6RTI)zJpOAHvu+HPFgW^LB~LkC2`7KXN;i{j35Xdf!VR&12OiU@wKk459jQJ
z(|0s%fL?Fv^5YQ>M14!7;jwkw7U=IyQY~8+SP7d?KVT|!SWEDSNg9UW^MYp^#yl^{
znihQZEobzx>SBnYg+g>ZQMfhnEPm{k(T_XlR1Up7Cqa89GB}id@<E(5mla)wz!aQ(
zRTKtM-;0_!#E)$-PF2%HDEwS6EbX<-{l9<YEyPZDcx|v93D=7%vjFZMIR@>Lsx3?w
ziTTzJzj(UWr;Br0a;&x9Cx-{1ia2beK2!4mkggW5CS5#No|%;q^?uI=f@ngP#XCAB
zXO}T+o1<=D^ZQTCtu@8S7}*$_hsLhw@+U@QR-`wMc%2uSZ$h%0{lAtK#;Cj0?>+8s
zdI96lKZ;V{-TGUw+o8#jp0%R}C;#8}^if@UpG2l(^~UZ~gw=r!RnEV0zO}MkPG?{)
zwEvV_@(%r+2+90NT)h-kGD)-jpU6|hNS5q-g{J~6A#~r$1tn~y%ygpMkK%4QTK}EB
z!Zd~cusRH$B+oW1Y$3#i8l`SeIt1N&Q#NcrB4ijL_6+@N+g$JDi|AG<2f8(skcrh~
zhxq|Q(FQgD0@k7gP*Ob+!39Gs#?Rv@>2l#K<xVlj`>@vOM^>a0?A>x9XdKzU@$7Mm
zn9Omh4L18;Q2o^l@W`7A<&jIW|NI7a7VX;}L!ez$b%)UM))WCAsA<H){m-_N-ugnW
zovTa!D<lqqc+kz`(`-O>IWH~XQh?ZTz7zbm@R=Tya48(NCKe3pqPRAYxpT~vOg<z+
z0OY*8)$46rs$8<x?x-aV?;1j+47f&cu5a*_7>%QFH3WG5I8`LXK1V2*<FByy61E0E
z!4Zz2u@inVzmm+R+wuQA8@76E2gJNkZhl!MJ2P?SISa{kI`dXKCxRZhG)hpcGi#}H
zFxXfIX+}6g85Zyb`}3~b>dXminsbt5gq5Z#iiuq0TwTlFEs1#Gc8j}$Mr)ZAZybP3
zb8qQDhHhKbbXaOP>nDzGuv0oJx)=Gbzww`)m=NXxAo%Br4-x32WMJE)34AG2eXX=c
zrTekakOjEqT&7%7cj+HC)kJbE$qd=1yu2&pJ3_uVSw5(I`4&Z7dsfgb2qL9BYKJ}K
z142TU?)VKB3J_U=PylM3Y7+djda+IQs<6zZ`+s=_FvBtmWG)HX-IOvajK&eOR>+@S
z5uKzpIz-=LQ_AyU!9B`i5BPN{8c6oN%><3DUkdq(c?eB9*~@MKi$!fWKV)LweD{VA
zyy+G1cF7O(cq;M5Y^=FliO)Y80O87aQovGJ@X!Ne()I*o#!KQL#G7xHo2}~}PAl__
z@;mxeg`FcVDLJH|w?;|l&|!5=%_9<DvzNKSnBgf4>hemuPUp|IO(;1-Mk0OH9vuo#
zh2*B@Hu$-rgIvPD(a-1O*qAt3k-(i`syK63Mkr$j<M!+kUa@xPA7g8YF(VD7a6lee
z;+%f0G!6rq!E}dztcFtxszkCOuL%==X_$Q?#Y5hk6D03~0Pd48mE2rE{#6O`Wp1!E
zlj_&-X&5^afU1E-AbWl>{MBP}Qk+um#k^teOIi!BrO2~LX5SB8^c@3fLR_{m4?&E>
zcOodKp?u)VaGtf($~EVcL%o1y_pGQ!-*fjT->fiGG~nPWART%b*f$+A3)6xz>srj5
z174cw2cd*yxx}8_%odGI#x19z%l|X)dj)%EbbR$%5l#;wtkfOl=vI;~7=djC|Cx9c
zg{<Sbm+Or0@MQcka#`O=;G(^cVdpTZSIjlTZ^h6oQa)&#;-e&9Uq|U)wePZ6$=})+
zD$c9`v&GHtWWo0cZ9v>_L3kLQM)->hMt~~8g%<$SH1nA|N}s1BMbBV{iLCki3ylqZ
z^5Q_4N+>^_VNn0s4_tqZy@`p*9XxK^?!HIKwet+5Vi7%5d)<mx`l*XeF<(Vl7}L)=
z;Sh?w5Ra>G8#;N5)X%ScOa5on-{vt-upt{BOq>;*Xy3E??Mt0r#-tLmc5v8n|ETWs
z2kGbi8;9=D4yK0o?T9oSXl@^#`M8rPN?=(1pHCAgEm0TU@)>B+O@*(p3*kS?J4Aob
z#Z?e%*;$s=b!<AE(tZ`PBT^&l)QWy&%pP0gh1u0T6{_^Gj%YWp{H&To`8GdCWTq1-
zAz1>b!oEl~d%`((+PH~&1grfO%|?)UwKAexV>=qt2HcsZJ5`reL8%N_>mW3~u_yc?
zaIfiJHlEL=up}rAdm?SDMw>4qelHjt4@ByW3<whm+w&b1kz~49r7os1_h=JFrfqX5
zlR9`|c3Vy{vlPh{cs75@@57zuewbD1j516Am*ZIgF=fN$(0{cC=lrs%41H1ru-*aD
zIwH*x!a(hjvPdKi7E`>__~M_ke@X!X4__vsrhhJgd^lW8&r<VmW38je0!&<dLB$yJ
z=LKge3gY{WXe*sY6Rj3Rb>e15GdwNxg9&DYll#BzR&IkZDXV<~5lHa|!JEiFR2Pdz
z=Ey`aiK`_39!J;*GEv0GtNy1scUuJjO|5t}q#Jv$-kxcCskjN$G;k|o7P37-%3n#t
zKyW98U#Enni%zw=F`|TOHxKRJ!p^r5=pn$Y=><E;JA@`-kij{H5VF@<Urxw8)`Biz
zmJVOR^#Z4>_}k8in36{}DJtYyC^~<kmjE}R#kSG1%QabaCAlrIH+lMt+{{hcS%OkW
zKL}pJpxw6wH-nO2y4r*JYuP&(y@2B<@;*hZ-NC+!soklh{L%mL`==!p;?RqiK8HGM
zG`U?+gik~mHhAFwH@B3^5vWBZ+TYdI#F8uZ;6ERx2f=<+@@Aj-GHxN`i^*2mubWq(
zD+y1yEY!eRMR=e~#9qpO1SBI;#eKlql3Zx58MnIKeD;_kQI?+pPQA=)r{IWGz!8IE
z6gp#(W*p$aK0euyaU56|mLZV=&7e`dCwS)=JmUvKHvSP+N>i*t;2U6Jw$LXdC+=qK
zIdtb6N~&6`QWK}|jELpgRp*n)fFEm$fYV5Dlc%2?M1)WEQCHh1{<@z-Z1&=7M9$7@
z7L#1_?Et!-DIf!#Nf8z%mOl~b;t1Zkw_m60ORO3`sz~R`ck@e<br4wW_9=Jc{X}v9
z6JlGpMD1G?__BlxU$`wo#)0o8i+SOMaf{XHG-mZ+2aOIbI@w-L97FO`U6Y;V5pbUN
zR(SKerAy70=%RN1+e)ntFm-o1-Lfu=qJ7h`kKyC$9}Ba8$wTf;sq7z|A3ue&D?Ko#
z436+7wq9LyUJ|y{4<~Sz-Tt{jN<xumk#St{@Q`H-BdUwX9+LJAm;BS?#5!1@4E8{w
zg^c=9c8TN;Ys3QDQ3s?R2~xj%H&PzV&cpmV%%f(d9!+y1EWVWta*cIn?2#j_#O=-c
z2@TjL-aDkbmeU}M77Lf)|099y-J227TD9A6Jru*|2KSw4Ee(#*;iI{@#}+O`vzJ#m
z2@$ARXLq!NhKrg$j9NI#`7W3F)S0;~V%<*bOmro)*Fa!_?jA{VAkfo{P0S>FPIT6o
z4I06K&#}|}AwbLz3;r>0zvN1eFNx(2$}EfL-hVlai1}$2iK%LHAj95{;R3k)8`jk9
z-G1Vhxc>AQy%!L`3?=<jX%}l*awL^$mvd;qybZ2aA4Xd@RR1H|X6!GHDp}nlWu4;0
zi|ieY7un9Fmq^}U&AOSWT5)9(OBtHj#L+j7t6rq5b)6943PCkxxTYL1+U4j0j#xP(
zbY+7_)jKPW#%ZJY06BUXvJulZdJ@R>eV<hEXNN`|ffd$Y)CNtfA*iNIS>iOY)R?1B
zsJ464q^uMp=(OxldCb)<<{VAYixQ<YxYIR7|B(9UeU}7fJ<LwVIsw$lVT-aA(IDX*
zmj3iz1<Lz+$O$i%NGY?c6&sqVAs0J=oT7~jru-BfiBb=-%`U=q2Ec@@h_?F|gtr79
zGB2-8J&90UpG$L(J(9Sqq##p-j0q<J{{DBe(_b)zDFi|=;i%4{KseN}4+*n4;!|Gl
z9rX)Rq%?frJP<JObPq#9;o)$(a_A&6IR<O9$Wg}30lQIGmIA@3SChUep&=?HD4XPC
zbc3ChD7hnAjH;i5oP}1P?!}9NlJO0xXLU~kD2F-w1i>XUwpLqwWKt_`PJKGIFUmZ8
zjYqLR+<Fc<GdF{xS$>lo2ns)K|Cg;>NSpePrfC0VLW5&q;oJPCGRb4vv^7TCm((5q
zSd_e9jl9=@FKH$g4kOK;gK+$BGAFT+3CrgCa2VL#Wc`#8{Z)=6|82ClHwFlJ!{0xg
zK|0dH@_Tt{-~SwSxatb8rnuM*!qSZUQC+Bcl5x5ASCAGNU2q?;49MB2vwdCK8^-1&
zQ}Ed!`LKudGRd+$gEjY(-K9CenR1zdE&f{DyTL-Q$VUy8Dqei6^DCr<5<>lGvCs`x
z0`7G^d2)Mc?6i@!yz{|aTT=IzGQue?yfY8I)rsH1<Bhq7(blu>Q;hisNXaqkTxcVn
znJwfl%|Cok<y@tJVCCU9k|k;EG^uP<DCN4Im)i{gL^4Pc)j@uXJh2rmWP;XWvwz_E
zYT1c}&;mNISOP)x&}6d@p7wZr?m}>vIYsVY1OiZx^wl5jRxFX!jgkGkr9Hv<30It_
zq6<_u(rZxb+@>6poCu84g4nXHP!zl1o?vXezh#;Pqg$S%-%V@1N=T0hel|GmA~fF|
zwYyP=Hn+{g$EohWrnKxOI1@E{(`^g%R<nM-=_8d7$dP5-EmXK65=crFCYStPTL4Qd
z^Xq;F$XMt$cQ8D;n<O&F7|29|67W}av6PdyAR<XAo>nG$y;aRtPDiC{keEh~N$IL2
zm!W*|%M8Js-IYcIq0y6oW1fzu*Cfku?&9|M3d}2lj;EkQlu8f6ASy~Na0XyDI_wi*
zH<2JRr(Q=iqf`M?u8AJMgf870X{}-@^xMp>vFyXQQJGM-Toh_->cnG7tugwN{}qn)
zpFxI^$lk~bhKJ|>60n$vm^fJf9}eq3I2IElJK%qTV=*(c{U7<v|Ia<s46cy9jzJd<
z4nGj2E$ra#?hZ}R90<V`IEoW)FaQv6fRZC>2MZ!=@9MI4aQDc`=wNZb`Bm>xR{`v5
zdgt`KZa7&~ur_Ou(_PO9I=vuK(NN-sBME3d11KQ_f`hYSf`egmQq^$bTEM@-i<T~e
zva<(XkP3cDB)9_u6tE~jLk2x6UyuS4qv!&XaQi1HhbI_@2Y~ku_K?017j`6}h(KMy
zwLw(YfuvrL;#s(AntOu7QP(ygU|=8f<bi_0NduG>;`5H<E`cPu<5AZ@pn%Ik3fPFa
zHAG<Z=>(>BK;VMBzNWx~BjC|4h=ylxPEJN4nw`y>+#4~BO@P}23)liyiOl1ZXvR_Q
zv>659Eg(OXv2nTR1U4bAz9?%=?ZI6^xdMS0`!Q^QgE)rl+68C@<$yFC0#XZM1h#>}
z-jk{yC<DRo)NO&XZL<BlJ}+PF#SO*@(*iLtUfd7_A|SDh;Ojv)gMcoa{@5tEQ^5j3
zHGIi}ZE$T4-GJ%_8ffQpM&F5_2n1voQvgAszNPV)!&26RE@#YO+k9%1eA=_u|Cth^
zExEqF5<n_v`MCWu>6asd)mZ0U`@Op2j_AQ1zN`yI$1<{f)($C&0_{QoyE+3=N`8+5
z37-EQs1{TJVt|B%>?2|aI)Vb~&|a(mb*sDH2l^Nqf5#1@zPop-cL&lK5DD<<S0h^E
zcVRHkr-1@;bpn3(_-@<ZrF4;#3m{(u`HzjI8BBck>k&RcRK-76f5IKO4IGA7zkUej
z@b&ZgIfcchm->RF9p``M|CFYyr7x_lm+~$@;=fs3e6<a%SIvV5q(-lY0Cajwd>HQg
z_O&^)3;q`O#;*WR^j3HDGgU+~JV`+GRtBMQ|5G&dT-RXmW^IZE{2QBZhL~lF6tJJS
zz<#iITp#f5@Wap0PpJO$d-Btk_PcrZYcJ$P-`eV*j{Q^r>pv~PU0*Y%54J|tC17Y>
zV1=lSF#fCbEZ|?;_-7pD%ILLEO_U0$l^_wxdb4i|$}JVNE0`A0uK=6HZ+^zF8_>}P
zE-ZlaGP}1M2OogGQvcg9fH8Z3e<kP`_I!be3e_tAi$D$AP~Y~|<Dt{-1C*bHIuDA!
z;<fe{b8iHCQ$afZlduThz&v<O*iapa(UT8|#+&kK7ylTnek#DGpnvx)4+t_qcN@Xv
z8<7O+Fr6EiZYR)St`z-=F#FT~YmVaAb`Q8-J0OQv=c?MqaN!s95vJe3TRVNR?{7Dd
zo3uC8)px-43qKDZ5cior!n*->0G_VF&^<dO#wHA$_COynJ(j`k!4rEwIOtUXhTJcD
zY#L!dLZIrieSkPpzuoVydAb|q8Y%R9njFDw-@Jd-QDDiQ!R5hMoOe3Ac|88xE{>sZ
zKtXN(uaLp9BgC&VCV__5Y^#8ImOoZ~gaR(*Jb#9x(?BE$_@>rRnvq7M0{f@`R{!KB
zw9^Ob+iJ8%JI6p{_)o(~J@Ciedt4e0m=mOC{+!&pHk7}I_H)5`r%>W@jH|PROSYRb
zLq@_s>y?W2;}`Hr%F&0X0Mjz*3Hct|HQ38WI7}4hWJEphU26p)oC5&vZTEj9;a-Am
z8<9M0;xnAt&7|rZGI3Jlf2nH^ehyD7&;H?W_&;SRwy8x?{jf3rjl<gzJYO$n58ea!
zDr9GBocK(mKM$BzsB$Doe+QvAY#yT}Nvuq$XgL3%-ZXHp438Xh+??)HN|nS-pYeSv
zZoTW-tF;%}Q}AL+H+N)fM_5MeA|;UO?IA=jdP-iC{Xv%V_MOl_%XIvuPn3VhP9nDq
zM9;Dfr>VjF;sQ}x)jQ{6{0w#`y+qvu<+a4z5%V!s+?H0Rv~vA=5L<%cu^@l_kCr%#
zNK-LVD!r{%)h21;9Ax}YJ^)$<G>f<IL#?h0XKZArmwu&cuQ=>-r?dB2jj%`PUe1x(
znq7`U6L;t?)U$x~q1b$6UN8iMH~v7{CY5tl9^`c{K#)Wam{O^YEZf=>56>Pfw`B-1
z9{<72Gf(6*d4-_rTO*@Cqt$f?fZzjjr0mW7o8jV<Ep9VJXPrtg9iKz7iIducCR|6Y
z174W}MQg^dM@?>!ME#Jj@>lyXhx5F2;#4u&em9%`5}qCA&cmBDeafv|yK)I@DyvGe
zrGapOt6lkWmlPdqnb^a!qsi_4xhfg@r?%&2#<4=XT*dyWlI=JAs-Cd!o3D+ds%%_Z
zj0jDqM!vNuqmOGo)v2VhDr#VH38P1s(1yC*-I!V}xu9e&PxV@52Oau87m%AUVNY3m
zmqY~BRsMKbX{FX@7T}#5+k+*W)eDGwn{4iAlY2c%5Q$Wg{^UAMIfc>rd|KP%QvE&S
zib=DrP$ZTY=M_k2X9iz*A;tgq)Tw{5-bf{E@>Znk@tqi(Wio@OX}TeTY2n#t4kUSj
zF2%w=Cf!ir3of9C?E8@@#rHT>R`E97W}S<T(w5BhTxdEgrYg<AAQIFYiPM7BzeGC~
z5q|L{guC8(`uXHJt;?s4y7-;K3dr^V)sf>)iI)WMM%E63i)&b>XGA*p=H5<8pj)CB
z69!l{1fvGRw9zw5{x+8Vp}0Q+<U+LBu>FHnE0DtT%1#+Wp1_pSIJTKH<E|`D3ENA~
zA8~NG*83W7@JTtV`*IcE5Z8L4C_|%=Je(t0i~7(Tm(}i4oVuqNZqKm@UGzxff8Vr@
zgtJ8?H#CUO?D?fg6dagPJc`MDAX{v=c2`k$XG!W^ws!7Lem3lISw(DVZ7n2`?M_>h
z`HJRui~|~SA-c@iFC#Puxxpdf$xb>lFPz`E25;x2l`-s1zJ2DCo21X~+H$YpT6CD}
zuK)wU%r90hIOa(sl0;UFy<(s>i%rFf`-#EBYx7+2)T)G8ma0d>bE1p?JAQrB)y6<3
z!<#zNmsd@(Q{l&TNQN73_qwM;C6~e0Z39Kh=Hq@TRdPm+(j%rYv53Q_YNDe;{k|}@
zF(ppERek{plj=)IoU;)tBYI6Y_du?dA>5iRnPsZYQW-5P7Zf)3Zu8bwn!}NAH(C>F
zW(UT(ifO;05F}S4NH0AsY>{JRc!ldw!Xc6~bZqdi(;ECIc+_!(l9q_hJTR8LF|HD@
zz8?1O`XqUGI-h2GAFFc=1_CQq__JT7T4a1VH?PEXU-<lAv+j`_s_`XP+xbfkRW`LG
zGr`4IYiB<D<9NAVq?aI#`#3g$G$ry_W1Xwwd*>5gPWNS;fLihI<l`pjpV)S_U>ol_
z+|1ytgep7!jo8J!qMkRqD(Nbn(U+0ysQ~6BNJr5uKiae$8|p4$P7|R5ty#=+mcOax
zqPzq?>cDi)(m+m}(+q?WDn#8{Wp*y2)!v8b#~BZ~tNLL6$bVnv$7~tlcvz`-Hp{nH
zWIS!l4?*<EoJlI%wPZH-`9W7Pn&+Pts&0N8FXPLXvf${+Cuv%=8Ox5-W&K87tW!I~
z7ZdRN@aF7X7jg|ZWvYm$Qs_hqXeP?ZO9p$%m1G>oqtt10s582JQVeNyPX6r%4ct{4
ziv41XAaKUw06^%mZ-N6>FR?#4vR4png~H&KdyK)HC&cwl?Y&jul)EeL$#}Dm#T(;h
z(g|=Yx#fNQ^qc<3X4VDF%Sfp}yUzC5yufjEpiYL^ZxNu?DXq}etEiTVA93&U(BhD6
zr0@Vd5X2|nNs9J2Fm`=P?37VxK#6D0q(;sp6W(nunMX<JES{%>|7teIeWk&~=^dgh
zqaLxdPl~#dVz67A)L*k!5ygCT=#!!&W#Ay3UxpnZx#i_po23|AS@Y<&Lzr#n2Y+ZB
z;1QfkfaLht9@t+V>O#pG@rz_KZF(i->jkpbl^V&CS3|wGOQNL$0&0tLhq;2^O40Km
z_1${jCnX0|x)<<6wF!zQw`_0rFlnz^G<X6|^eZQ1>x-&rc#^hRvABLfP}^j3WO<#z
zrx(eH?pz~D9Qt@U#y^_M6V4*RTDB%@8n;^9THkda#t1uq7Fyw2S4?ZLollHQ6r)t0
zxkr-zCVh)lA+ie)$jS{y2EGQ=1zE>-zCVY8z6WraC%Hp%&ZaI{#$)CZYAY&AIk!EA
zl#Z(~^Ztz~a*U12VH|vPjs(^HXW}PPKKTf{NL3j?cz!@`UM;Ck-0DMcM=WK*<aE(e
zMlxxU<&6{Ud9jd~DB@c?aNjF!+{)>)*|G|ewFG!-5_a%%Cd~?Tq2t8+nG6s&`ciRY
z#PM*O2A|<cDwf63&F>mgRwDL)U}5L!gxst!O#OUq5v3|CmbP>pb%vZUWE0NSslldq
zZUZxe1!|_cRe%=}G7+X=QD})+kjxPhz4tq;Vr1zHR?N-fW~v>?<msFssk2$%8<gYW
zqhB{)3cdaN(4t;7*5wC>mN;j7c_)U*IfDVFvo@=fj<{-GVLvw<5~u0I>uoJ&xp0j;
zmDaR{?9Gvw1L{!f72){Ial93dNa>cC3cM2MMp$IE?aE7Ub?P9-QXZr4*-I()$Hjry
zc4V3e@h9DhbLl;FMpB9~8iG*Co=Er4jOQjWU2C$3>$513-BFI#xJZZ-xX((vgWXi!
zUk`2n4s8gN=nRRLraKUi{gi(`Q)7v(U7&C6N>`tzrKJuDwLw?%mpLAUcag|B+Av_E
zR>#JSC8je$F7D1YYMI{L7WaU|@S}%oKqmK!m?Ma<Wu$Rdg?L9tRxK-78Tjk22q$b<
z|58>>nv%f{_ssb=R0a9b?JGx|z6<5Dq7;!fIG;Q_@v1gl;o*Vf5MAT*@LFNU-Xiy4
zm6JBy#fCa2tN-M7u^Pi$^eTJ#>MU8fb@#ZQ!HFVXh2~$3jVy!W_lg*m*a{EeYFP)A
z7UM#Q_GZ=@RRHWX>`LLkS~Y8hlQ?Q|Q@<SM7W=d0JhxqAhlRiH276EsmX+DW-noR1
z0Ar*=NAFlZGRQj8tS0{#EKi<Vyfx;(B}kI`sx%P4V#-EP(vEK6QdwR1$>}>rZ=nf&
z!8c>RsWV7qho0p3wZPxQJoVREZlBOSy-ByjPB7fVXb;eelH^%Hpee+PAU#&xRsLk?
zJ<Vz{Thf}fZk|7DN4lBsO1MX>-a()&DGxfv%bo_ga%^9(a&`atxK>+8Q;tK!YVdO3
zoM%mN^m!<`C(49R<ry|RXXS|PQ;{>2a73eOKe8EOub1>Pr9FI~(gy`f5>rc~??{>j
zOqU*gA8mgljG}^(f*Oo&kW8+t6Ko`hxU5_>`B)Rg%193GP@&rhrZ8?hUSD4=0U+FM
zHa+-%chhUaI(d0}+aZ){Gu0$LQ=db95cp0}+;c31M8)1*S6rCz7i$p_r%O<Xa;KYw
z<+Blq8{M)>gapNVPIi-pS^m?kG5h(WJYf@;JdLjCSy{D_5yUVv4W&dLnw8P(&=c>y
zA3R(eCR{PkGqUzyXbt6z$rDq-?tWrM?)b*sb@PlYOWT0*SlLxL@2od{)?BKBlV04{
zg^jOf{}rZ&#OStK4H`i(()p5#Q9?9tnZ1e#-`}mqo{ieN2WaXEjKEE;fI%D{=Wt(s
zhf?V;LE$|16vOtkJW?5ht)hf?z1;NLb}fIR6fE9rO@@56pv}LW=6bhSt~3jVDWKO`
z8FDT@tn7bH6*-`FpN%S2YbotAxL?^n2B~qGH~lBbnnpo*H9314|MH4tFJ*o1TOy*N
z#k4bq7eIfM%9{7_rdT(FAD1IDBYM-%L1~bkdnJbbu`N&MU%p)}NVJ;MP-P_ki!6CC
zCzA#xBNKd5qd)qucaH2V5;>N+s2q#)t&kSI(PbL);%%i<K=Y@8K$q5V?4_)9Xv`q7
zj<feLnr>j1+0N=we!VC}u{q*AFUzM7X(NDyTD3Rj?RoLr+BKgRG~6feP670%yOdPO
zMbUkb+xFvjEMsjjgQOw0I6ea&)4#kXlGVK14>G_#o(y{X{?H}-P@}i+O>8)VvaA`y
z8CJ6v!u-=4Mu8y+W0eTvRlP*yUq>%uo(;C#sI{d6HI6_}X4aJ(f>rnA*PRz)zJvYf
z%X<d7LA#LA`DE7Z;V}=rA{QS&H+SpiTpv;2GH|rgGbN7C#xJ<LFaWqNg3GY}-Sms4
zf%!UCwAaS-n{PxnvwZm~ppdB|4}<v>eed782Wad}6tM7&0eY4Z=*mpMtCn0Gfq-5>
zWmurk!chccD3KCc29wePRRJ)bph!Geo&MIr0~ILrhwFzpQ*`u8f2SUHXdCf=WTM$O
zxi{b&*uQ7mHnv92q{_#~>qEDRlP5OZA?GMl`hRWaO4PE0jIe9O*pGOYrY|}#hn-i@
z>GsH)(aumrb8`VB_#}Zk#10@HKiE#dCSsSJDI;4s=KV^Ne-PtJT?>*NPtr0lR61_P
z=1@SV!&M2(2G!PUeA?4cmtH<#J69`?qfi{B+RkfKa>NxWsGgr)2z+iBtS>L%xvYw0
z7{4e=Eq4*zl&QZsA{Ep<gs$<UOIdr~F^IcSl>jampN_H>JAhBol)n2x8kbDXWTR?N
zFH2lU{~fdhL@g1-qZr6<^cqT(`w|gsi;Dh49R8Iw>cn1*{eiC57q2@&C0rTPxSWG)
zXx+(7sW=gYp8e@edyg%$&cxI-&KbIG(|-rrp9(hXn<j=)z{bZ*4~kQ~=~PmjNH|^q
zHwD%Y3!C@FEz%701;Ww_eyx388ce$f?L0>g1%0lslVICJgM*FviS4t8{BykpPlltk
z*LLHtP<RR24<u`pBP?yPL*}zEZ@F=9!8BU&O`kM<+2`ScBUpw&^{}aVNwBIOmCJMS
z5DOy<mF#{DvTgKi;H3;K9{5SbX(#l8RG&&RWKCY?i`F;0onj&LcMtkKMOT`X<0@G7
z;TB=-_`u~73<_++kJ?Zm;_Zcg({+h1Py6Dh>t@6LBBE7I+XO4Z^oG}#GRplJ>Jbln
zdwOCm{dMzZ4d7jk{&vv%PhOalF87qCLpZo;Tz0>Q&R-l2ZZZ9R=&B>y(2&fXiHLU&
zA{Kj1bZyv;`S@MG^APq_4y7KZEGFzv`Z^!#=}&adxf<^}WJ}?NiF=6$*tQgt{{L33
zYbC>VwvI*L&7Myd{&1`P(B5e}CkIMY6jrl|AjhtJ2P8y#D3OmoaV;FF$K;1vp!FPe
zZ3AYu8uJc?^T*Jt)>#q?oUw19a^bB`$O?I8Lo=Q1KUrS+`#3?~U=_yh2DOlhB*F*E
zOWHPiDT78M7*Se;_PCK4nQ^(y)*f6#wheYM5Q@_?h~)MmVX>{;VUJu(oc_5h8y~Yl
zS`uCRGdi{8ND|!%QZLl^Wf(&4R!UtmCvOsn>nyxtx6=V*>+RsY>HA^F!uxAx#{3W_
zEY6!-5?XoJpwtq1=H7OJ@~Q!ZP>#MkFRC`LQd`Gi_8%@EoyNBK=f2)yiH*?Ej<Iql
z0gOF&UW#@!Kc(Z3v<J$Bb|D(yt^L3ypjDGpy7D*p2MLWCSpC%}u(ny6Mzi||b9$#n
z_YkW;qx7gnpNyp(Ulouuz5f`jPSgaaA7%Md`AjyHi$rHIF7g4MvALSRjRatJB4eG)
z<EAmC5zU9^yI!;_m}${OTg(9BCk=hR`-n?t?6#2yA4LiLX~`@;S%4+JqYCi4dI{wv
zUUrYaUkV{6`yfN>+**ozY9Y|y84-&&r9b~gJeTCCUs@qKupjvw>%M-6-I{yuuW|~C
zzdW;l+NgZd9Iq|k_G7W_TLtvp`WTVB!mn5Vs7^_61CkcIF@B>`IYA?xR2p;4#qCd-
zCeBCA$la53S^T$QzTbP0(cKZ$pGXwxypzJ_TBM>xHi0Ow{_!YgNG+nyDeUjJdf;p^
zejop+e#X?{6_nI=Polv-vIVmE{rxiDiG=p{O)!?_R=OqcmNlVNo0yrwjYOG4WENHU
zd%C4gw6)wk$6d*SVq}jk`eZZBrjjKzl3VKRmA6wDq=w*PnoQA?goE?c;3EcGk~8QH
z)z$1x=IzE<9YKJU=zNI6>JyR|svO5otsmQm^HH9o`YLpDv0{;Rv_EM=LXeORhAhkh
zUNiE!Y8?=*C=OH_GzGSsWOh0v!R#AG;U3pf|5^26A$7wdRRdgb(00NB`PlrdP!dOV
zW#qelAs>EzB-bhZV>^5^ioTnWHH@_y%h$Go^SN%eMf9-o+t0Q&OOSv%YNlB9DHI1u
z55r!mvSg#mqneJzLs(#gyo$jUpPs-?vZdxdD6p_-uX|LEMr~-*thtaS^kRR~@Ig<h
zY>vt@$EI8p({EAuY?GWq0(Efm*dK)aAy%V@;V**o)BOcA1VnoA%%V_ZzBzo%<i3=S
z{LzMeHgH%p@TX1_u3*c1iS>1*TtFm8<`&B9yM5y(9`ON1gG+hMClX66nVPT_lYZnT
z)XP=&7RhUb0v0R-jnfBe)H<fT$LvlquHE&7r%TRDm+fe=1_{#aF^HZuweN{(ax@Ia
z0DSB6sBE{BDHUJ)F!c41TMnv5qt`d(x5o#Ve;Hz=0hg4&XJVZo>A6b;??00_kN-56
zYen0M1k|9~{WZnjOhs9fyF!4xUHGI)42GUZZ+xEWeX*{xVHVk|;`DL@284Pz)%_Ut
zkkwRY#+!GEOHrs5NTgDuGDy@nCrRHYF0K8M-&tH+c9E`S;-*~O{+lUXk!jp8s8p>j
z?PT)KokbZya463u&B%CZA?bFBm46wFp7!rq^95`IG5GXMls2ROVy#LcemQ_%!tx#l
zLLahKZAcv5T&YdowOC#I*5a0p&MrM2^W3$NWR69JX#t32J6^ih9qWFG^l&v;Mz2zl
zGo=;R<0R?`atWD>99mlIk5p1-L-@yMep4@a|Dz|)^JcCQJWog&&0{<x{0B`Fiaf)`
zv9+z02wbWp#?%s|8<Gp{7K&P2H2gS$<4iP_#a$apsxlsy&DnFK+C)n!R*e+RZmHMZ
zF71C=;d=R7hK6s(Z3PbkJ-Ph3^F&QfWUAz+=(TJ57aO|jz!%d~1y+5d!k4T&z_;ux
zjqFyG!evme&zqRXgx^rFBVml?%j6{?Q9cy^&%WbUbRMN`Zj9)g<;ZsjSyBeRdz3T9
z#)bo39KpsUq@vKZUyVv7)$TP5Y+|PY3FK_nHcxw->jg6DXR2eG(d~P4NtnOqeCav;
z?vUt<Bnf6pmn@56V@!i%v8Om$55v>&Y4B?L*c;gQprf-1$_$WPk3R?c)fys6JX{j;
zUkQU9(m*eKZ<qpOHAf0!@o%p+bpa!-Cz<D%NAVN`UTO9`W16V5E{fj{4kc&L#mSfu
zs9U_}<#T!I&n>Fw+Jh9%RM!SP%8pyK$$_+08>^d!fJtET6{ksQ4r=>s>Y<8Ic$U5m
z<cSsr0*PK9iDk?@z_jcXv)CYM<t7VOx^KA7^BOx@DPrdfL`E#-b#$fqf;+P*ObM2n
zCKWu^Q^?Z9g;@u|qvIySvb5$-!m2yW4sa~!6Gi=Cedns|7Z%MV2Vu&8w7HmPCH#9t
zcYj0mTWqwFw^4R%Gcs@TWy8WGia*62{sl^{QKWV-9nJU38LxZC9L!|xElx5w(2Gr%
zKo7RzbnF-L)h*y*;0_L`fNoq8!>rJ)Sh%b9H2=l)P7qRYvRu-qg<eI;{Ybi~|0<?L
zM`QwFf0$g9h&F{M{yxQdbXBB1#8xfrwaDl|uj)UubdwfWu4HA9o$D4xq~y8{9ra$Y
zu9N&X+FccKN?@<|cbIQqccIsI!<y1Z37oR&a%GH!;_lTNA*5=8ipKMzvs6Hx((($I
zM1<Gdi@zoJnkWI`$5Bzt(#v&#dtRYUgWaQ;mKpcB;fk!tXh+KkaA%FB&7*d&essmd
zrBtiV4W+xn3Z$vcKJ91ms-}{@_p%bkLdqrHOV#nxkok(KTcm5)0nvj7bx-cNGHo&}
zb>IB3330z-&9!z?an|e8<|jhdzBNrA^k@tyVYp>yxL2LdR-dVU(ZhV_iLc|p$0rF7
zP0$_r!`}N9*II}?-p)ee*=I%Js{_iv=>GH89q`L>?rwi*WG{SrRg2^GnXua6aJS{1
zLSfSe&k3(&Vo@GvlC-$l(L?48(&;~L`cFP#j=Av!7J+<nnIbXXq2OndPI93P=iNnI
zK-KlP2dca$LSpow1|{e7g$2P3(Fqx4N?sh9b=TZm*BWMV(KNL(?j1BNVrA&SqKJ&R
z0;4RcUFEN7TcWQ%HwDN}Xu`|!fGN*3VbsX;Qk;Vxg~*&fO<PWBPaR{E=ZUl!5*ad9
zn%W!lGE}zB+D53h>`4V$8Bct0oPwa;bj{oKhFD<CH3fl{hE;OZK)e3~j(Zt<?{vYi
z*{aI~+k>uJ;o&xi85f#ltuNl`Bwk4aTPo7_S!`7sviJ=DuZt9`QA0&Wi`)Y?BxmMM
zw}pI(dx%caSL*nHTNRalh>=FS-UO3T&W7gOH{T<AXl=~e73O%2JG*ILn16@G%b|8#
z3&_Y3r1b=|2N|(gbv;NH{5I#TkS0l{M8qye0V&>t5Xl}`!0Moed@Pph)(A&DA$@?O
z4RCV9s&d&|E~C|v8BAYNFLS3XyJ0IC^@|2jZPb5zL6>odJo<&b#<4K%Hq0p(&Uy*&
zl|8xr#lPL%PS$}fvd1QP#t}D^y?pl4Iws<^4&}vMRaVq=r|@__c&&(;tdi(U29Sk)
zi?Krosr?Gm%M)oxKx&asN}OcIg#FZ$GAa-Yvdi3fkHxI5Xzrrj7a~Hiu=8ebKO^3B
ziea}*pQD=K&&Y?MZesP#h?%-sxiy4i2M`7x^;0G=`;iSBDn=4)vT>@V$#iXNYC~c*
zdv>~WT#+i(f6`(A&?*idIWuN<Ov4A~2j(>Ec%1n=7J+!c_b+_9NoF7bg_+v|$Q^A%
zJQI4za11ibG&G~XUbtQF-J~h`%uv>EcKh<-RF@$4NR8*osg?rQymGWCUFFb4(T7vf
zuULsaSG#(YaHKpip6OzJgqZ>mRncfUE>jd0gfXUC5~#la`e4*|_5`o(y0YV0j%;?k
zCx30YjnF`uyK%sKNpwyavuhDD-$X)xVTU9_OGP<8jKGUJrAc-lFM=eSw6TR4b}%$e
zcJA<fcfX2tm-=;X$GAs=FGiMUDq-3XaPYeeM2s6*n|Z5%6Twmgp3#Mk1YLyE8osiG
z!IZ2n9Df9c$heEbn4WwBX?OXUGK7<i1;Zb_B0h|!uuZcA&r5S$Knki8?lU|B{!|?I
z3)3LT<@1ZS<@HZnPtN~CH@u+fAeiA{z6WM9jlI9&9sUkb<FZ?@(X$SmTRm61orFdq
zWY5&Ds<^mz-=-0nP6j;Kj~{tnv4K=Zl{@Q7-q#a(thDEVKwpTNY1hT&AE@AyiGqt9
z(X3d!+3sc9-ev|LVSy^*-xLi#dXW}qj7pt0*+3i)!QH`HfhJWT(RN}LUm46ZX;zkm
zKdBx6Jx}T0j>qj{mSu=bSNW7#OvLatg4Yg#pUd^=Y!vuv#IeQKY>i7sy`S8G?v*%4
z)%O$0Szl2Ull%6Gc?f;r<}bR$SO&_hZcxZ^blVtksaQk$!?yaMRW=H$yfZyZ0gY2J
zs$<9%B`FQfjyaG<!9kd#U!T3ax|#CH;ef6^=uPEtLlUV;Vn0$jTt#oZ6zjsUhjUrl
zFCQVsQe-v(=~nCKI<=kn#d#Lne_FOln@t#M%FXvDXf)qNn|C?(U&Q3^_J5}#Gyhk)
zLRwuP_PPU?z2a5lw85frX;wX-yp(x=TLWKyr7!2&e^JPbT@9pclT?MrT}VUaeQ-R{
z#a!26;9%Zl$up{Hq7*}eFtNPiDELfsLqfpnPIuYh6?*F5z9S8w?4~}f`s11toOw?b
zZZ=MSQa;ijn+JShfj&dV=<~1Bf|{!k9!oR(QDWE-g*lXcBZECHfsi60al3@eJ7MFG
zJ<uIsG>ue$eb@sxJ+;+N=)$9~*}@))e_cc$AwH=V?kT&-l=9p{h5%^6ABU>h{EiGI
z<+%c<xOMY-njouE!nAu<*(I;{%=A`GU$V*9GKM9O^z7aaV=%_{gyZf%tvZO~j0(u}
z!!vwR&(ygWs@%B>70)m)Y#4{aPCro|ZRZL4Ii)t#lYaB*T^(J~^Y?&^p}lKi5A5)h
z=7juU+%4BmZdCv;-89H+JmpU<rroVVMrjlan14kC11Zp>j9F*T%+r-{ShCsO<v*$w
z*9+p0YW2EMjZ+iw%%tQ<Z|z5n7kR=L2Bq|&RM^inL}@qQHOpn%DRz^iUB~nhn$@5}
z1Vl*BkosXBCJkgcm<(^-Z8bsu7~LuVg}P$KKmD?)q&s8-DUcNydFi<^U}NZ3TZqRb
zS8-4iVkA96VsB{H7?#laihK49N$ry7e8mFfgPv(VGy*F4U?<nE{s{$i`cn4?q#Q0Y
zopPNGsqQE`N9@rTpbZ1qs~W8B#)2q|thF#(RIVrnL{dF_{<>7%|C9>dk+9J7)ag|O
zMQ-6-ibvW#X?UI`{MOvRJq;w>=ZfaaGDGz0TyUSXWX%UIpRmSM)}_7S{#$jdHc~Rn
z)ZPhEbE9l(4ab#wk<KQzMN#f$kNbQ)w>O3N&B&kM<I&Hhh{D}qJzPYu$e+ToCJjzt
z9U8+BH)H5{y>fHB=t=Mg=~llYFxFAw93d{4t@GO9CuT>Z<QCnv$0xw{LpXSg_ywoQ
z;np?oevxExg1p1)Y|uX7wlHwWL5j{9jogs>Kl4#2nAMwVc8B{LNY_m7SfL<xuzq22
zhw$StLVqWii$VXMlVA!Eig<iHCUmtHvbLd#&P8%?X#bNLgFytpSN~1dqPS?+V!*wh
zLs209cjQ=aT8*S5g3GjDhYy%@l<v5$Lv2!1x+vMb8v6462(Fzw-(=SR*QW}w`xtm@
zRuIKnH7n%r<HTMj8dM@l>vk!)I?qZmBh_h^kQ-}OM9IMi^aM`nS-4E0%8?mkmYs>;
zw!TPw$!d7GAcR&pMzy_$DfH%JDYy0gWc$|48hxZS2j#<0>@ehl$F=>$asuBCgRYoA
z?pCGOYpX4xsaB3K148Vm33{t&T6J=FZ&)1CyG+-rz7!^E^dCt%=w}}LS_iJJ1RTa2
zKrJ|sa8726qf{g;3l97agmG2ZZ{iN*>V&1O?l7$>SeA<F3nn~sc(7%AD^~s9*3(^3
z&UdDkI`vOFW>*bC#p+;k^F6JT1U9avxZE?32WzfPJvlEE<iH$Q8Nw5?-I8XqFl=kf
zhwg1`7&{AaRw+C80W9k>u4!^Caq1i4GopFChmAP`T?c0AwQ<_$Ay|z!_N2;oW0p!u
zPAe=o)8Y*>2>NI0OiM{TvH1`cn>}Vl<@yuonBD;&-8<}XIL{ZD|M|JjpheyJgt`ww
z!9IG<ArmT9;qpVzxEC4yn_O&kB5$eejr{Rv<v7_L=_GYU7-1t~?<C|ZJ-BC=YlDy-
zYJ3npGYN6s8EXYA|Jgm4g<>4#BXHWO_!b%`<dY@AsFa5h_0h+suS=|2lC&5eE$wK>
zEAJkyzIdDKRk5ImJTXU%G!9Ps=VvtSI(J`+{n>X@cB9#wXDhen&h)qbX(pf#DY6eJ
zZkMR1DL8T#;tZ1j$oi>96_@V9O#w;p9seQo0hCEQ$Lv!p7A~&=W-kcp%5rI&7}rsv
z7=?fu?1&Yg=sW;A^1IvU`;Dn`fwJUw0Of*T982yCkp<yiqwGqT`A&lBy%HH+TwZ^t
zfFLge95m80XeF_Mr&AVD=VAu7ekmdq_KH}%xj_554iP!<uml6e6blKZw9B%q*SG@q
zxHxrYyCZppmJv>#4YfwQbu}Pu7_P4JCVPaH>w7FRh-p1l|FoFe7dsd%JHFzd4CP&@
zmH*Z#CWfqf6rB{=4V$$&PFYis@;SAd=hiHxwYbGm1TzN~BE6oVaL$pqoY60A1W^j|
zk2JJJ1>2pTU!8)-x1}!OZCkqyfBn+j{gKl$?A1o@wsXG0PnI5&Jv~s`N5MtNv6b3%
zGUI@<X@bRP`rnY&6K9g_wo>YtyQ(?h#txT*ckvwQrexV=x~JnRpBeQDO6ZZJ>8!|!
zT3x!Lg3F_M&&YBiktjT!#>QXj*b)VDwRL<y3pWbpm4@7jdi_&&sm*9n8TYHVqWrvp
zhtGpH6EFv8KGZ}OMA=Nj$!4D<^_MQw?gADA!JA;_Je@S_Q+8NQr!Ed=u*6E|)mJ-a
zV-@l7?~Q2}cvXJKggq&(OinFVNq!Ykt?#2Nm7<Imx(wcHf)1+@3ETBGYqm0yUeM)-
z`DG7wUPPVe>&|yen=8ODsp)NUMm4-mlYh>N<1ntJROBg5$uzAe6fmZ4=9tVwX!B0h
zn0tK%ubRoIS+9F)LI}?|Q;ivFd$ih`EWL3}WUNH<@m>4<Qqrbmk8##|KP6McoOzbl
zZBLUGWSUh(^`+zvv>!5ay<wI#XuhefMF?xt_y_EE$F5bkv)f|*@=%d7GnU}()Uro3
zH=ta1X3=&SKmmk@YiVp6TbBTqxzNSXy8mJ99h-DvqHW8vZQHi%DSMY~+qP|6yNq47
zZQHhO-ZyS^^p|rYx_?AwMy|QWm>1|%w!$v9oJXf6XZ_%3LR%OsT-kA*xphXK_7fXO
z@l9q|{(dD*43iC*<(=_kl&2#fbc3^LWB4}}zmS{%{%dl^G&fft)FdUrF=bRJkp&x|
z*<MKIAho<eSfG#%x8!64s`pT3Q>Feg%C=M=<O<z&(QsU@=#ENNg2@1AIoprGwX9c;
z`F-##U0pJ~7#pDwcbeh`hdI_G$$}2|JUp3WG*^X4UH;?HL4irnA2_WM0u!+Bj#Uv>
zRxrAH1!NeGSqB^!{TV%PCAbg^6Gmqrz>m*pLgo(5W*N^QEZ2K!!X(K4R$%>Eo5Xrz
zdsO_vAl2lRNstJlorNN2N??0$SBjTeq#$d+h^Vo93*6;a>HqK(Uo+s1OlC!2MZ-eY
zH*aMPAEvArr@<vvW3_tcTB!tGbyO~XilcflKVma#Y3ur&^(<2OZ9I+m*Hc)+UnwR+
z%z8u_NE22<!0k#qc*=t+l4<QNE*I}ve2aB~>Sk=UC$0VJ4SSV;FLs^8cc1v&Q#;Qi
z@j#2~{X-iLNh+Amb9$HI!*8y4Dr>ZQ&Ih=sewY!1cbar{ZCxgtV9N(_jA{e_sk@6y
zssntajk6Mcvii>(X1(bLwek#SWQSiu3NJIrudkm9@HOLb8n-!jM?-PXg;yjAHK985
zR8Zxw%WnjhstFy{jC@*{pH96x%qmf^fK$h}S+(8_<&bgI#%Gb8Uhs#2R|+7&b1#XH
zYIE26gpy&_Rk!<T1PPe8P>)Ia$39Xwh$@*d;BP++6;<trgfN^}LoCHyke<E~>445h
zKeFXN5&w&4KRUCH7JHsF<s~z^P9u8^^Z6-3`z&|6<8}dYM?H8Y@Sx(=m$%b!x$UjV
zy%>JrKnB*FVr^E`@X3KI8&Dv}EsQ@S>;Zd)eSm+VcfyY?XTru=OcSh+D#ySaEW})y
zsnzcuR#8SOf02;Bi<o>?xaLMc06T5q8m+UvNZVSW_EAt?6KWK2@x(CAH~e!frTkFg
z%$FEG!GDPFZ%=hw1|z(`*V9bN>Ifn-oQqyXTT(oH>n2BaL$Q-b#uzEYu>~2v=ge1U
zYmwGc=U#iUcJt1wK>EVD@kZVm9=*`29V!bqkaZG|&+|hJufHd}llnKZ@7l3oC}?}?
zztyy>K&+|*NpT9b*($}+_ET&#*euq{6Qd~n_jc619%FM<z(4ENfJtN`Ks3d98Qi;Q
zh%Rx;+n^B2cP@5U`UWoJQJM5>@bs{*F5y)pR`mvf+r@ust7w%gNP)5P$n{uw%xOfz
z7H_gXJnNW(cU{-#rvvNwo`)#K2CeYzGpe6YZ$y{S53jG5Y(dTKg}nzucljb6$ibik
z7^h^`%J>E@+=DB%SqYb#@A=rv@{<V?uC>3!l*5yPq=TsAP^cvs@n;34Qyh863+FA*
zl3wwcT*;^yF+0pclZ6t5du!29HR@_v=_<7NWQW~wUj9x?__tle*WRm66^4>xb9WW=
zb0P2T<TPsI)i)QQ!nfHE;Qt-BXQvt(pTe}QBBO_ni~^4?=pGrfOFGaRWK2<c_@%Eu
zBM|s=eO^rzH(DZRU&RMPv!z?HN5koFv{e{-Ng%2yp^{ztO2dzAB?6E)bVZmggAp9l
zd3_t|c#+FKhE{|qg?s?-cyoKNWH*J&fo*|terPem9m>Xfu(=x5TGoOax$Kn&h__Rt
z3MpBc`)-tlHVA#+mFY7)3ZX=$BhcS50^bQ=xj@~)jpl~-^|Jc4RXq$$4%VTr5?iK$
zHHWvRb!vqBROB+5!c2NL^0OB-_$wsK!B*lo0iS%>qyHT&VEtdf0yb{e{|6;tA>!a<
z<NBXm0VgLr%m3}%Ci?&43f#a|(6lxgwzrA77O0qAWF*5~Bv<?3m_Y~Uh^WaW!=WVI
zNw;H*iMCOsp`_FI1TMID-)G%#wO(s9t<7#vW4{5PzB}Dv&w*(StyT0>IK@!GByH~=
z0S5t;I6T5YL<#|rP)DQ2rpv)ZIfDKQd}^RbTks&kQNN6zs9a+(fkQ{tNRpiJSZIGQ
zZou|Xf$pHAA0s0`1A+XA6t3(guypW<?2m!#!7c89R16goa0xC^Gom^1j1a|%<lnua
z_v-e+?jj>29ljAD$9RJGkZ@_K1P5V&A-GB9o9dlF&IW9Vh(ca_l^i64kfAUg_4MTB
z<q2rbISH_~ewZmS`%r^`KwSjwh}s~>Fz%um_9{|Y&p4S0_|yUjunt}zS3=i``KtGb
zfKVKQ)&l<8jRCor(?cN$=E2XYDneRx1pM{SS^nj60RHK~0U|`)^<Vf#|E5I-``}%h
z7=bvsG6H`N6TAj?L!v|yR94J|+z;IY(Garr=hotG?@zc_BTz%4q_r*{yd9_pifY&Z
zq9yC?U29+OBW8f&B;br#WDM)-T_9rA1h)DFJGp`jBH6n0Tc$(|47N#Vy_x>(td3L2
zpj>*L-9P|oZvGVOU)c^-Cxmfyg{&C+Jq`@H`{}<Df(k%Ff<i+^2M^EzCSt^8x~)C(
z+F|&+E8e5e)Zpi@kC_46had%fgjf$s?0@I8UBxE|1h$2Lbo<#0|GTq)e-CU!%mEd^
zK1KjC{7K1+0<r$RPRffI?g9KKKyv>cIG`8stC>mMYziC9<N15|)7jqTWc4xe;S}$O
z`V%iDhj0RZcYFd0()J!2AXyCp4oor@`rF(175=-s<zJ+wme6Yj`je~(ZUqS%@m&zm
z)!A$5@l^sH*E<r5dUuPOH;}*s0nGK2*iEefWfA;!@cE1Jdz<i^ed2HW?mza^d#UK+
z?(|`M{AT(iP^C75M0wkLC5hfint)mWEHVTB)T;|G<R7L1Y69!%{$u|SW&nm8yFtOb
z>{o|mNEZ13ykR+r$fnO9?8zN5;XIc}Apv6w{^2qcm}nE?ixJmp#Re}?J`%S<xmT9x
zi0b=O61*IK>ScB69u^w3YR`ezp3mekk(W4tk<f|DR!Fbc6@KqHZ~$ox2`DK~pbr6J
z^mAF>5CFtDwv6w2Dfz}pLItrW2GIY#3Ix)_JSqZq1z9Kl5qkd-{&jh^$6D9Z1H(f2
zpv2SUVjMtyk^f5-unGzG3;Gou70Zj4iFmJVC7I`c>)%UGlz=$~VQ+|GMv8o)fV<(Q
zNiTCd-p$YOoanmA6q~&14!m>YCmWJGrJ0J{1tE|w>$fY?&{fgrNYUu=yW-yaj_>9n
zQ|NXXyNej@kvo2^*7c640*d<n+2HyO*Y?`>Yy$Y)?)bn4<1XYZa;t1G>+Fa|u{=)F
z_j;lBTz*bxhz*;zv-Yn+Q5*bK<Qc{<BDBR(VVlk}&gB*v3<=#c>+_U+*^c)3xcVYp
z-;B(ERXrr!fA^0?rn+O-JN0*EKWtR08NFQ*n-sSGGZ^3}&cQQ&Lfe{LD9wMWty9Ox
zbh}Ka>F`TWtbW152|bJOn@tljBV%D~A3J&`47sc3#zK^m>lffdTVGn7H~=zJETy?@
zb)ST%NV*4Qw02&XtX<5qoax(_?2x8fr~lFujF*efC|4%HkoB7S+sb9WQ`g$wAd{e8
z0>Uc8=Ixg%HUqAg2{`{i=4rgOI1q|91@q0nI#ri7e6<DU%^>9DznIjwOdO}l(|xJe
z!#qqoGPd#I<8H@(z2hFq!wT0`hEf=IY(=*9gGGUDrzvf`tjb!QUKLX}++yh<0*AE{
zrbmAjapDa5wqRF3IK>JH1wZ>hLT)Z$U|iQz&CZ>|hVv;vMJA6j8j-*7;V`RxZgbA(
zd55GK8MA`=tgdrbW(h}upw{i#Xwzx??FQSM9WTEp{?otps`*;Eee+C5?LplOR+jEa
zRd;|D<wNQ{V{U(S{NBNI%G0l}22tm&%LjQAhd!<BWX>}{d(r&TX|_qYt=p?ZV7+B-
zd}>R~Pt68jl1<TX<ux_&vFzJ^QH1baGl!iE7{^6jK#ASlbQi0JCbq)lvWK@6Gv3c|
zoUN9Daq-BD`?!RKMJkDZFJqxih@45LrG7h(kY(oI&3}Fj&m(`s@6cLNa^<F%PGz68
zTFlJ{MB^rMMtQdRaeM0W_G(IA9Pw3QrQE9GccSbj-1MjI+h_4fv1*f0qe@<#&Ft7t
z-ctvIC<S)+>Z47fAAG(aj`8uRw+DCTCMle_r)sKY_->x??pc$NN(7J7QNMrlqOsm5
zVnfCtL+ky3Jz`&x4=ck~$Mk8BYE>wQGVVTBZBM4F$X(`pH~`f+>#;s9ugd8Dy*fZl
zJ%G8uBWzBeksRrXObYPikwf!UO98I2c8F{0sbp_OGr<1~BiGR46V)mA4|8Lk{6Me<
z&4s(D+R%HEhz+8L-+SQO5ok8!vztuhD$^=V=u1a;U@kHcx$4bg{mWpZCuQxk+85o6
z=0$V~^+RJZwgV|}oXY6AzER?G?J<6BUScB^3LN&fcb}mvn*Q465;>OG!LR9W_qHpb
zZbL3x{p<0ru~&3<YlCLJa^kR(ef}_)z3Ar8fTy#6nmB7cW9e4--pxEj{499*Amq^@
zY2$onfS+u3$h99H^(om&w6!ZfhLtPZG8wHCJ>g>c3gN+Oycx75XH!yv3-qd(0FkHP
zjj=GEu^)yhnP2Q;ib2+aSyREXFmWgoMI302CxvdPI&>{dSiqY`>a7BxGhL$4&wSZO
zU)S!%khdt-vfwmwQ`Mz8vT<SLL6YV?g>{U)-#`$G_b(tgtnDC!TKkRPa86El>Skh_
z+gH59nb!=^lK1xc-PQj(Q-+cndh$S=({zAOzh<j&yApw94uZg%>dQU?Ee(&d`Q<KO
ziyzIU?>1)RwnDIet~LzS?HMPH<w@hzMql4K6#37U8Fi3TZO5)~AyMh_S?7CYQ6)ag
z30d6+T_*nhaEslQ=(?Migh9Ja6^7DEBn&adv|Bs#zU8QLNIVZ-;>>iro}ZnI(NnwP
zktFo!t>0VA&1m(JYJITGqGm1E=%MXUK|xl2an=C@Um`s)w5L9_Rvf|!tNzlcb4);+
znos@ZeLHQ!k}K1(`FgF>V|Cv>CfgvC;5D4raZpcyksnIUs58>8`^jaxL8Etp>7RKP
zOSt9`7t8_fz@IMR*oaX4BvwhjDQt<herD8k<4&2>Xe+8XFh<zUZJ*ls=K3&Q{~355
zS`pc)@@=kl?Ai<`IPMm^!9UYlW64DP)AwR0Nqvl$zP6eC77*LrooY1D#?&s3J|a~i
z2<4j4tZD%8Rg0CbstAdRRh!FFgz)UI!BQ;RPloB(empf=6=3GG4D~do8{_B6BCOQU
zg|KkqCq4QyZ1ZC)`8|6$=XX%1L7UX9eOzOqd;I!7R9FyD^z`aa1Z!oGF8hYkBUNj5
zNZl$BR!8wlvIkYwkM)pB!pZwyTQ<{GjfxhMg}9KmqmmsmXUe1Ye~ajiH*@=w;vp;o
z6t?BU?ZH)J*Mn^A$Ld>Q7KeLH>CLz?mUhhauCf3e!xUG>netHBH946EpE;2d;X--Q
zr1mVnu_&FsJjJERq%-XTK-*+W3g)@o!AR5V>m+3P)PF6c9O%?fYA?+bhpx|=%WO4b
zw|+>pkN#ZUeizvOsy{hs;wtz2*3c#QsfLXTY*LN^<H1JaTm$b>g3m&_LmE#9nom`d
zr5KBLTCZq6Uy=G=VywuKvSEPr#<J)&zFCI;DtPsWWD`5%ax~ka140U`L#~P2XWR1D
zWsC)W?i-=zmr0tGl@exbJrih7fQ%%nZMmIzeja+<1{mu*2vjNAAz^oOsoWN3*D;^N
zs+#k#x1Z^K-uaVKE_+&vaWjE*X{qj~IJDc>pq$bfR*02?-T2@22js-giuGwVUF?_l
z;oQm6`_Vh>i7+OTWJzb0MUqp}Xj&DdZ_3f}C0i#J9Xt)Wi?US6e5F32{rh%X&Zei5
zK{9I)`?d%&4@^Wg#R$p_{`7|z=vt~jGLbJU5a5s~)N9SXarfMaexxq<2;@_He=dy5
zSp-;T8LhYQs|Jc-eie6KX%tc|_~pgrYMj%%U;kK*u>^<V_HU-{-K~ut1)BG;{|gnc
zR#Kt1(zoE=d}!8eVXV=?AGJt5L6n+G*ttCHjF9l@H>pJIeQ~VJpDJTf!{8{9)JSSj
z7K3_HoS8Bx-$F%1f5JT<+LsJ%6UvY6@Tlji<oL%ui3slO#gzHPBA1j9!Bh?a%BKWL
zMlkOso4rP$cqa70+2?tt`iN{{^<A)JLBq?)g(ao==lgr(n7)Zgu`olko8JaF4})g8
zfaBRF*prSq-J^mIjkJ>9K=Vx+n-L@580z;+OpCQmSq3K2>_<g7)<g)q9?=>loF8v{
zdy!#CafbrdA`U9_=7injKyyC+0Y1cA5WieKj)4VRP0xxc*Sv{)A$TN+Lf*Zh%Mw)%
zb?R0Y8xchdH1Fn52^X-`L10H|V3mJvEVT6~LkQXRcjr;*jCTG0F#ZXi=)?AXhN*s(
zxnUcR&i&S4CQU2Bts1j5z0BNeV);UXL6RkAeKf7Mb|G{A3vl~BF0(l;q4%RpLcB^;
zPhRDa3ntk0TVa%~R!FBg@<hoamy_rgH05wugRe4q=fJ9F$%#=zx<ignmv_rac~m-p
zALd;3a)dBv@ku8q=q$UGK<7}h1()CcilZ<q&2EI%-88o-S5%(a^yW;mn%ecYi@B}f
zw=T9-1xhfv83%aYosF1DZ2Xgv|I&0YI6~O1Gs&BI78ZObo_Fx9ah8%Qd~5#65<>Bm
zXi%fg%eyc7_GQ&88O|>^lfAgQ{@z)uY{3Tjwa6(!4~y`m>HkORHaz$aJl)hD1cmwo
zf#V)H?+z7@?K1H<IN{HV!>*yqCF417KQdXrw(FZXyUFHNUPYuE>e!){b%`2cW_(s8
znjPj9oefQ+7zZRN!iO30akOX$eZo;^g~wCm=ra(C@%mH(#ier9a=xSO8HD(E3mOf1
zq9*n-{Oig7*eS&!yIMrikyYaG?He=JP-eKRYtsO7?0@vNqiAuuRa%eRI`z)>3Mf;m
zn^3)<g@$5hRFf~rqrF_lP)Tc?lQO5k=zduFQm*#7lO-2hRyn<+zD>H$Xu=_(dHBfo
z>8JHu%@cz3Y%2MBv3ml(w;$>T(>g}tee6K@{BxUSr~9~JxCk+eRZ~->P}tQX5B?I8
zI~t?x{?CKM7k2~EMlY7J#n4d!t;a57_k(+BYekZp{5#8ekrLd0{!{OlKgx_X988rI
z>=oyQK%^*pZajEs!`Uy1`xlQ8cc65I9FQ4(D#A*;lsO0Gycyuiuo&V@15nEh*X;3H
zVtO<=mc2;BCY|P!$s9PHkp;2z+r8fI3~I0|%tB@`ryq$I+Tu>$>=%Ub<(mR1<@Cfp
z$={wtq%AXIygHRk(QCC_17wNE1T&b?;r*!NYh+>S>!Cg;BDo6h2(10J&CEh^A`BP0
zCRMu}=FhStwE3F2Cbcm}TwWgsw+(pok|)cydt!TB>klk7ujlbis-u@UbI;uk7?WDM
zgqmsfm3V!SENTZ;$HeaHW3CTONi<=(W|+^%B;Nk(?)&dYEspD(P;3|cdw4bc`uZ#v
zrFFx)Hp)U)`uzm9jt3|(e*iPK@H%FC%c{tZkQ#M(K6}BODDY*^t0p^8?LHPv$J_d?
z)Cxzn<Z8yu%vY;1lJxUHJ8CX58BS3k(e=uKCT;XK{EXC}C3zb13q#qY%e$GhBle4W
zA`U<e=S*5s|0M|-B+lCT$1fEYayHlGjkZ;$^@og&2A^Eu-mujUP?2R813TPO5Lm_l
z4Vca<K~ryoWR1AE#6Se~?~s^pQeG`OAW})N)jX!a_?nvYAa&Nw@?K-m<^3t+-jj^D
zVW~L|30-qWKwOlQZ?Pzm)s3;qf3d3vGPkp;pc0MNq&=VS0N?wM+Q}Onf;AOI*7pwO
z2+sAmg@ZgnTUlzlly@qx%9fsHHrtLbFF~e+(jPeU-w*9P!^Do1h-*MnL8;B7d;1W<
z6MJvl4t$)QEM8rq&EJjeC`CA$<|}gQPFgv>4v=2VDsb@h0U1-VV5~zH`P3%#SwW3s
zq;gs6gRX>QpIEk~v*<zEzj3m0ts(6vQ}~n`B}cmJniQA$7@#)v&;*44&7Y0Ph0||6
za0<70v8}n)dd?U$42f`4dW=DB${XsqY!r({fiAS(gm&%2h+CFfbU9xX2AP;v+NU7w
zk|_8iRhBc~Qg;gRQo54YQzy4Q9`~(Pakicf+rhj6Y{9@J%0?D*MSqj-If~Rv6_M<_
z;|Y*=!ane)`*N`ejp`y3mv|)C`b+mCo)>RgVj9UaR&1n)S8kXak@lX9)u~Hf(N6Gj
z&8>L8F|d@dayaxM!XUemjc5(8LKA$&K)JpowtYg)LiD<w_}^}ly*g2olTK6Lq8mj=
zXCUS?gHB8T*<_SG#Na#v2NwUI(w&w=vASPLIr}D#UQgX>r2*Y-ihfxKKSSTH5ey`T
zGOwNY1XVVQE;BMKLqp|OelLp_^DO!eU`i{w&mBuxjhTt@NNlcZu5c!H=4`TWt?LV~
z^*n!nMJY96Eg;`U312Puwv111-PrM+X9=Y}X?YA=T9<>Ykwi7|l=JpWADE0-`q8~l
z8Dt5Cf-OhCbjl}i+fW?(Q(O6|zL#WSQZWT9UVg^LhVPW`FImg}&P|{+snXcLm2Nqu
z%<VW>q&$>iS?;;~7|MxR^Gb8zZgZ=B>*})1MFbySIER}(XJA}zgb1&xMf0Y+d8=hF
zY`zAlwEPk9wr5_b++>Eqvb!?gNKxFJPHZIarJn!J+8#GP6mS`T3@HRxF$Zl&GD=(s
zrd_E_Z8VI+&NZ}L1pm5XHwzCFXULi$jUnV~#&iv0|9h>vpW3i-g!cTC!GQk7E>^yJ
zeQW}P3Oo)Q=w#<E@p844&|)G||Li!LWYBXZ;k2qyWl*Fy6?$KN5VG{ln|iBC9C;3&
zMzj$>V=h>^K8&aZof8afp+byKY3FDK;zVIC5wh6~`EY@=o)1CLCXC@rV?$UC>C#ik
zG1(4nc!6Y7t#P%D$$7}UxCvVjhsj0%2;HpT(}_-<QqA5b$oz1P{K>ORANXv`ERxOS
z!<Omg^VTF1Ds^F4z15JY&CIbmNOf+ocqE^Q4msYe?-LH0f4|jaU~t)+)&RBHpI3uk
zUcEjQV>=x4pc4THmd668iDprfM`F^d4<cr{%#U(Bc}X6pe~o0Tp-zYO*!%36;ZpBd
zHoMX6hLSD6)4p<jWs#g!Vd}aKxr~4r!ikH&Esl=9ngaM6-A&bAY*|XG$_!`sblKWn
z7_jOQnxjeH<XT!+`S3L3xN3Z|hgkRDjd!y0^*2$l-XPK+*3gA#kIb?#juu|G&3|{c
zRNAs2BUE8D#w&cAK%JWQum*ouK(qW==Eh>2#BtYAQ~hdX;&WwNq6tQ2vRz1jF(gg{
z79|VAqeNU0y`yq2MUboPyjCqb+hh<D2M1g0STc~}Pd9HT%!WXaojxLlPL(No7T*oF
zsE;Cxr`LvC!#+#H^ej#2d&3s-)urnN%H$HnghCFzh<Frc&A{mnEX=ZRMZi!$Q#_~W
z{!4=?&G~r?8lu154&;T6zJ-{<VNmN;*uW51m|~nAz6#m5>xB=y16lEbZ#ghygsM*Z
zjt$6?E`r2AEch(IsA!Nw*UB5?uKQJPOe<!`6f)+@vP?RiQ@}y`&W+(Vmua?p{ou{R
z2`c>R!jLOj5d^VIYU&OW$Y_|)(l~b2n32W%svTw<)9YO0RuHt_YS%i1n}%xmK~vn0
z-ik*))*e;u4C}_FtGcX2mR`8qw(6(}4K!|zh@y(YvBf7PW=&Kh70@y1`RZP%vyE99
z&`H{C5OBl@d<P@(ju$COhE^JJEu}{G7z8AW@%K8`aeq(J#`(h8nr9|F$#(5ETgFLL
zw+h@tKTGI$1ZJvfV*kvcF&|yi2I?`a)lti8Qh8q4nO=?rWq68H2NB$;{P|Y(={i~o
za*qlgARf~Q#9cLXG+Eb~VmU7EW14AQ?*?i<|479bja@o>p1I1*vy+=t)Sr;%kc6@A
zRTEV-t4Tk5rGAS&=LXmD6Zszx6R<#A;Ln7d0p{=<TV2_$*~()l<q^AcJJh1}*XHdf
znEYCnScrZXv!3Zy2bpcl97dO9=jo?IdBhF{c5bIR&YKvI(dWPI0=A8u-LuWEmN)ZH
z3s=UVUV)Kav2ADf(I`Zs`RQ_{NWv`73zfWSu`GV?U?5tMw87?l&#YXv@u7Vce<%UV
zcK@8-8+mQB75l_1HY=Se&u0VAN~u69NxJ5W%SN@y%A{{aKM&qCa<>;qdcZEC(2)z;
z+|`&mt7cSm&x+S|k*%ACAui;4zG-|aA6|IO!gJb|34m8gu2UF>Zc@(L!*%<5BOm`G
z(YcBjBX8_V(Z*_T#g44gomx)qaQCpFuqaA)w43`F7f3EwC<6>$uUcV<`?T4}rBj|#
zr<?_na4fg>R@pi-LV4TAL0x09rn!M>PnW}QPv{$ftkvz6ct&A3BRrzn{c4h}2c~zQ
z1jpuCdfPPYNWiZV^cYbiZtC?qoYGfW*uP@IX+Gox9d$2^(XE}NaHfN;c7x^`KG;YG
z>=4<LWolWhqkhcJy?8*<KC#+U>2%mAEDsSC+yU8n`VqZa;*;Tfv$<$zjKkvn7Y<{q
zFCtE7w2@@Ra*1uqBXRI8$vLk0xw3NglQRHOT=`RzPDzcrOcsq(q9$b+@yG->rc~aA
zze4ORt?(uk&eFBQRbS-^TejSx*fafaZ;p$fca`DJy4|vyFa-&@T_Y>G<P+#u&rISq
z{TGu@T}sRTNd~5CLeRTumw_xjreweg#y3rklhBOzdM!Hn@bezj>!`F)Js3F5&*d^j
zq&m|{5OR>xhT`lT;%SaDbo5{t9C6AlnoG~Frk~GGfFc+`Bz&SSL0!@@yuQG0X#D4W
z>1U_OwS@O<fMhy*!i-NpH^@i)8{R7A<B@1!Krog`AETo>;vchc4F^ilNAl)&^}{^g
z+r!@IM9v01OA3$_Pl!8mjX7CK&gl!4c6lL3_~~H?BlU6>8M#I0ZV*{6WJqMqlVV-~
zsHQesf$6Q<)z%CXcg|5cCSPl;=<6(}%A={eKK*WqFS06#0+^JLry0vmn!FG(G!*F;
zOix{HhKmUi2_f3WKs~VbK48>)Jey2Nd+pC>SW;NSan?~7B_#M0C2L&k%=(6=?Lkb%
z%P`4Oq0J(2Nnjsx?#Cz=%MrwYwX%G;l*xZvx)_lM|0owj99#Yy+HHnm1VVvJ7nIdV
z%#$)4B3sUPPNNH<Y#K&ad3Z<#l8wwWAkV0^;~s}U()xI`j}mH$Ag0J%_-}v-5Ufdy
ziPY(6`FRN#JJh@MQ_;1GS7lHa=WDhYUble$1>0AKnzFyDB-62Cj^z7WUv+sh%^Kv%
zfg!<{<rdc#S^Yt*Z!GQe;;RRd%lUr45jliuIJi1_bxsOEb&>LQjfw2HCbmVnFc0{1
zHVB+g0`t>MSy#~Z2Of6?fzyW8Xr|ar*z6&ONIHJp9E(-`VAsfg6mxRmQ@*Uy+G?`8
z7xyf#NU#%Ln0>_n>0KuEA!o#1MyT2q@!N^+iphOdf;vATtYNilXraT8IO&opaOH<s
zDn8=`t>@%vTV`4#e2#Ppqm&b{Ml!y|+jVpUoz`hJ17zTlBMbRC%lf&}iNgm!yk(;~
zw1B4NdR{_6bt0hQ$y&DVo!_gjR7oI`M)q=))B+mMR3?ktL3)fxSCmjr^%=8_ItJn6
zQ$SwuIyfswd>QIxtfZ>BU6Hx8HwxoIrwAMNFQ*T4Sx|hw^^!%%-eIJDBt#>*(z^-7
zVX?h&-e+R{8Yu{mZrXz&x&Fd-XJT<<rekGzpUNqU@k<@p15O{sFQy-`6^~n}pXzLy
z&KF5JkKcK<n>o&xubPviDr3@)&sQi*H&#Bhup{h>6=wnbrYr~GTX3S`o8k_Gws;uI
zE=!oF`zdEQ(^GMlV@li`QX=tsBN;5>H?x*A5f%PfpL}kw3eJhK^Ssu!7<_F#Hf+Y{
zJEz(6bZ!1rTCvm?x*3iJbCLP#gyj}GgOKhH@ua)u2jh7_o(7vti|9{(P9+(iqx2Jg
zQ>T7HfbxaOYKgE0d4zrY1(L6#82aC%Yxe&oy5?Zx{(qxuE>;$%{~29#voZbukFHU#
zHUZ?!JjKQ+B;E5w+uQW)2EoX<G4s?^K|ukMwA8su(&>`a#PADBNl`@8oW7T@KR>_D
zJ-1gm*Pb=JwY+((HQEp(AW?FXkmWciffzurLB)X+9pWg0NRpyRU?>m{j+uS)H**3C
zK<U_M$M6y)#lIpoF+sv}SbxBSM#|4gkOk=p-H9~7pfJ)QWF&(H1r3oTFBrwhq;bkY
z?n8NjUBLsjB*B(&`$}S--~{_~b>}ju{P6?b@jL?zsHvfUc25Mw5UF5ppoD;2i7hb8
zAt?$;hwv?8TtY<Vfq$JGB}TwwUs5no9v&Z2Lp(fSNfc+C_`U=MxHiBoz@EbjeG26I
z^j!kIkPy%8je>v}{zk%XeIl-5L5q9}DJT#@IE3>GBky2=5{96nfe#Hq;8<S(J@Xet
z2&B0FDC|N0XyOoQAbquV3RL@Z2a)>Wis0cV667I?0^#k$h4gP12>P<)3pEL35kd)Z
zeJh2ud6P1~An8MG;U5GjzU6X(m6uvTg&4ejkI!r(gTIP;!+H&F`!l0`978CZsD>o7
zEDVB133`M29hZSf2;noUA|ickt8s!YAR<4FVnax9_4_G0JlvzPi1l^*{LP~Aw=y8=
z6UbmigGU1m6hu)XB3goldUkg~{y5$u>qGs}hWY_Fen5nN5cL8v#DbYa;)08M2L|&L
z@hK2wn}EbYdCCv+-vu&YnuETHfZ!3#-G}NE*kNWudl|)F^hDbt^gs;3=3_#7x_o>c
zhhWu~lMM0h6S(NjrZuswsI|5mdCk7tB_Ss-7GRPT)BvF(qJsbl>i_RGfu15h+SU6x
z`Ne$QEe8hsF}M5c&u2wqAo+KR5RIw%Gdk=xpVjen)^#8Fhc}1-g>_V6FV}x$2Zs#8
zX~-w=cXsf1^XQlJvA64s(fD`wz5Knc^EYkho&3kY7zwe>yD#5q3hnbRGj2ij&=vSW
zuj+JxKV3b;67<dC{jQQM8(1TD6qNaEZ#2}aTJWbxF5I4lM7jXPSuKIVR_7qH0%$Sk
z*D5P08c?v{FZtkARuS$A_5Fy;MJrU4W{H2eiWnD3uHSV|T?Q9YkgyO4P{xR{olzeV
zlNopJ-=E%794fL1!9r#VLAY)tBDhG9uZyw_GKDeZ^El}Ez+b_kKw<xcFIZ$qV<bkx
zKjuFx$iOiDdl~n?xJV;Fl7veF0bmMHfrfXSL^R-N!at|@U}z6RU1Ct~R9E>5g?j>u
zOT8>e!$XfQulGNr-B=1^xEHYB@l9gPP-f9T9Q|Z_*9iA+?-HPgxYt2JKN2e$cyoSZ
z{x(d*P$7jsKtHNx5`<_J)K9nUeNTI}pSSJ^FwfyUAZ8Saaqfg|D|{B*6>26As2aFQ
z_=j4o@`uP!SwhVpIf=+*k1E{}8lw2L6tC`zK6<L%&c#fVZrDxR8WMNq)Z=yUd~X8`
z^A;}E*i8JwpMG!4Ek?G8u*i)8kpo_LCF*tSzMNa)S8nt*c#MQnEWvyisJ-qNKh69l
z&V+>&t8lCThMFT8r3IjeGDoy_8F5Z0Ts^pjX5->942X|Pc<v|(1$h3G2ody{ElMAO
zhFlFrqJe4BLksZdLgg@WBEQ@i?3vURn)mY;elU6QN?NjtMcdJow;2%oZ*7z@i61#b
zrp0=l9l<1q`~l)qd}#I|ys36QJ<jL79<ql-{<@e?ZKSSUVSJw)py78MTH$06lvwvQ
zMDrF(rYli_XTQ3}tR+j4zgV!Zi5LyBcHA;bN##HAn{f;9V!528;G8V|2_`OiP)k9g
z&L*zF$+Xk8K>RZnoEpiA`Ew+P@k#F?WzHeUs}rLYS4{aEG)5h%KiN|CLyQVa@Y`P}
z8)8d9<PaPecaO@5I{)g`W5Jo8#Ehrz+dI87tH;YF3TfB+p1g8>G`J|B^DM#J&_GFi
z6b~CtF;$f8(U7D>kyoea&V*3Ehk-KO?pjtQAlI@R+gR!G*rk0Y(Z!5T?Fe%#@qXm>
z#WZRV#5lV;<(-pg8|fXHjv51+%8rohAca}}Dy7{_%>8vG<ey9xuVr^Y-F&+lco^dV
ziqhEKY_xG6f*{|feW2~V`W+d}_=2DLoLgyH$|pI{UHg0jIaDL<RY#fYjBig0{~>Es
zcApI+Aaup`+Db(u`RS{YV^LTtmWP8vP)5V7fjtP~BPzww81>zm29uqJDe)|4TGL?M
zzxXF#&51s3JBj>kgzdj()V2%;4v~=opv5g`Yl`z#KJo%iZua)Dn(nkbYzwC)m@_4E
zc>ApPPWVLE_tz^bx~7=f?!PTsODOKpyR{mbsjiAS)p~dz0gSTqGl>B%8Bexae}RQI
zs~Fe)5>9W7+>c-C0+K#B?eVyARH(rro|#Vi>r61^adP_NduRTcJ3zVym+_$GgsBLF
z(ka-E)1-c=K@V4c2Mt2!o3<w3IILr|!a!%cc9k4@ZjQV?T$bhSFuG$Py79kcV;RW8
zfpL~+#OA>SpIu`b?PQ#Jd?5`0evBl)RZiD|oWU&rzBDmJk6O3HgV)09;pzNAy<fNM
z&V5p?vIQnmg{)orUYGXGba$iQvG|}pcp}VNHSHxi<IPTZ%vZgkvE9YM9gg<-G5$@{
zzsa~(Lm8)@e(DUe+&y)7@Q;alfIX5B?CfcdVLwHkfW;oJx+$IzP_1|l@^US8QS6DK
zHKwRSADgD8cv9cq26ZCvm5VNt9Xmu9YN1mv71M}&3=DdE0-I&H9P~a4D=KO?b-w96
zUtZ!dzk%qp0q`lAy}pQws+>Ye1Vjdj&X}k&n(5NEI*P`RR6jw40Bfj9Y!r0)y=EWP
z*7?u%Y-}A$R!l|?B-5C7eS#{4mw5zJ4iW{#Xprz6HQIKEBPMU`M4#*JH;%T?PX7BA
z;UDU}Go7+&-gaCM1DQSm^;7FHQm0QKQ!nEh*S&pzxS8(cD6v>oYlux7-5#-gfnI4j
zidn@z^#SM5%t;`CwF5O>J#D-Ii&6jtq>xhv4|-Y-F3m|gZ2+)6UzRu;M>-zSu7bL<
z<y%r%41PZVUbc6Gu-RmZ)wBmLiwVd_o;AXB^ZWcP&z`uyUjXKcte7yQHgn`7$bQcm
zDXLFFUWc@eitvbq847E(EW8x+BpXbANL(YcCR`p#I<($xL*l-+c&%7Z{uCdQ*OHhV
z*(Z(yuXw)7J{}Fe#tgb{`EbaKRaMFSw{~9HNJRFUl)X0Q^i<M^x<$-{%np}-H99zG
zM}vPs4X$T!6-~%9oZ+D#P?%d)2Wa;FxcyI)vUVDP{>_2|XR-O4H(}d%(ZwH&xJI#m
z#vP%UQ!ZDAS9cI|^AO}klxyc=CFENA@@6k)k2L0@sod(4cb*&KI|`7rZ-*~zt8I(b
z#@piP@B4V4BRE?HaaIWEJ!Za57h|8SdLWp0#NRk8-}{l<?2XXVUm@!!|2DM5TBuK1
z!g(dC=FW`l<{-o&OsMe3?I7o{J-p7Lnp!w%>-35QA4}w}pD}<e@Rv{x+s@``>$j*4
zdrc`sgXq{FEYYKs5_;<0_=(;^OrbI}khqT@%K@(miK!!XE>&`ljN<2(1bEY&n+8#P
zpDD^tuz4-lukc?99bjdDI@aYi!Z+5YqDq;<)}WXn@k;uS?jv^!B8}dxF>HGwC-;3f
zW(OFb!<#{VFSUE`dQiQx(P%-9pwl~;skUb4_jWRTSYG7NAFAD|hLjA>>Hbq`0dV(S
zY5`V6&xJWEuqR~)Jd0!v)iy*Fn54&OQ~xb3;+LZ_42fF7Z)!cxJDJGZk}uBa@s2ro
z5}aPz9|k_^#F)4Tr{g!fCq0z%ztl7z^<>DO?QHN{asXV99nJ~at$eP$&;k;fN9wL6
z)GP_$DxxgF5BK+S5BF{2xdH1JN{#MIIBzH15TySBqQSF{PJH|eLRZt^wxx0?xbVuh
zaK5iobPX@b;IQ-6s#c)3#O0A;0pgQ$IbS?pYc{@@V_>V(J!8p&Ek)rM9Wr4BhYj-Z
z5U;9X+x0zPa|LYoSIs`R70q6M29Rt|+YH{6yzG%D*)6S`JrunT;J8%MQc*?fxh?1^
z=>JpJ!?^gy7^k$KO&K=E8{eq@8tq=#Iv!X#dx)PB8QIWDV80%p_yKvV!R@=n(n9Lr
zO55kVewZjMjGh_-kK>k%bC6aaeNcwRIQv!3ndgO=ora(G_f_t}%Hq<NJ;Y-E{qQo`
z>`2bUC=)il`$1h<q3X~u+Nm9mvCRbbgKSDiOCj4>KN@W9Ih9Qaq~UB8lXK|yxmN;T
zrn%OjLsNEI7LLo0$fN~dJ5UH2%9@OghlaeM^sYWcM-TN>l{gt?sW%5E)GJ{76!{cr
zbi^3<tVN}}A~-(W1ya?fLCfcOMvu--wNdJVUa;}yi6xczCf*wn7ln5cZMXBIOp;a9
zXc0MA@nh^QN#3T)YQUL0w^^Mf+;I3=lu=;EiV*a}l4D2ig-vhuxVjT!bNZ2h`k-X{
zcL<YCVD9ts5j(@g<?`v46qmd_nm{<9coLqO60}(;CQl0fnq}J#c)*CMTUtP<oC3Sm
z9$&oyBzW8{;rRsU3Ju6qk{H*)5V$mR$7+%wELjx+q-HWle?R`j9plEr7`Jo<Za)LO
zz%oDY6+RuEf@X~EMi)jU>^17*Fd_h~jDQKx10i|i7)FUj>N^&<EK;s}>hbNmf>dvD
zgRSQEF3Ry20ib6Xn6WV8p-9+Aq3Au2`1PYMoZY|qP|rSY7nlU3uuceRt)XR}O5i(z
zg}Zb{Yt~$NYGdCHu0aiW^@U3V&op&)X}Zwuth=KpSO5ZfQBDHZd>B^Q%qbG#S1|WU
zl0d`_$jqc*nQJeN3dA!^)E&Kc_}vPH&KDpeHFD2Jh79_8QMq;S+<M_w(U{cINqgTv
z-#_!-LJgr2_7<*yC^mBJycah!8L`6pe6{1k_ojyg<LV2G8qf-_8fzdNP|q|XX#qaP
z=AK`>(uW-TEHM((zPPsaJ4!=?;`vUj1w*ek7^U-d*!8^}JTS_dx<T2-n<)<G7Q)YG
zHgQM*Bz6mLEQc$~0G=ZHbv(iZ@DB=`wSA1LW%h98WAbxjsl<ycWbtt#x*l0G+R4h1
zIw)svb-k%K{|I$)5v#@!j^HHTwoAWW_3g!Xo6XRzHu-&s&N47<p?UYXbc}e~y<`pV
zBU>ub_h0h|@M$T)=(HK8F%9YP20r(`JM3Fx$Or1f&C!(}toxXRYR`oPKhJ|B?*Mu;
ztYP-$^Z^xdKTX}Lbvyv{JQZ0~fo+qzHg%S2ylO0s+ur&zIWTmFsJhk7o*zfP!?>1R
zgL#;U{H=VI%D;{a=+B<C!rL@{Vx{uu;t!k~&mUYp93g#X?2r1TviJHGYjuHCL$A6e
zq5PKW_K+F&m)sBeYdf6m|0df^9r&gN8XI~(yaVB;+C~3b%cZmkMFVp?bFTyK0o*>F
z^Lxp{OAb9oCn0Q>O|;6#F*H#iaZ2YJ)b_jXZ_+KI%xL+7v&uvJ&fwF#Ir=1PN2%jQ
zcYU|eh6BB3@R6P*sAj6K5P#Hu4$0QV3yEFP>)-EKJiBxMqzUt#cx*DKqz?9gAXZra
zsyUxevbr9R!o~5+WI2lKx*Mddxsn+;y{8`0a|1wRY(XeYX|uw0N@lw)J@^FRe|7Ez
zxO!IR%9NS2oIL+J$DyKrTEsb(6pIr49r=CMFuai!7PoKh?1M=cz;^9^PnsR^#?L`W
zUdl<afR>z*VD8(X64_Oa3BH1PFnOZB-ZUC*!(>jN?c<r~TllvR59^8<b8hz16hYm@
zG^w!lj2(zLtT$EdwJ;>?vX)hXAP|1BO1nGOKR(%-^ss*fL&~C^vP$qk{<)3FqSmpa
z7uT$PQ?lSH`DZOFh~U|8vDrSmelP!AK>Pp<!we~L>U>=kUEdclN~z2N7<+Wdz{9h?
z?QFThMTi1i7hs-1GM;2_t*~sUzqHG`P#2?&D#F|1O2ZKFAsR08_}4YFPz3l$<ztr%
zm{=wi+B6~%-lFq(|32fDPaA0Kzl|XgE|g8d%)&<v_eOVrRfa_E1^UhxZFd5NYB?Xj
z_vhO&52WG9pv|H#akn;`?cm2!mRzh(T0U-5nw}_U^P-WREc|q3yYrg@WaKzHn+@z-
zgUQxM6W0?+ZhZ#MsAUZY!p3PG)V^u=y0!pTtPfnQrmv1yy3NRH&*O#mTj{oR3+(*l
z?{b{(7W4%0q(SW3aW(*q(@$?`+qqz9LR~h`LD*+5d_Ls?0neU9W>+o{$0y1(i*j4S
zoE4{6zmb;W=-Lb1eMX$AX@3b`q<K@NG5Ynb+sv;C%(Gk_vJhAOi(I;%1M)nyeXiHy
zhTv4BK$U$}xC4$dZF%eruYPYfJ5p+^^?dWJ)P>|yU>3(ouhwAOp%VDc`_WMqV>I}i
zFIh5b<iz;p96Q{+TBlguo&UVlXigz&eL7@k{jL*EQo%Ozr^UPVdy-Y|^6B#pf-o{J
zKyAjwUWNn|tYf4X+D9r=X?1fu6=8(D$sNuP_OE}Z{UGYdSoSv^QM@@>s7%8n0V({x
zxPTw$5l6#M((bVAN2iTeZq0&zQ*@<b5hRRLTYGK2&a$%XW?^}qld4n8SMGB@JMH$f
zZOS$oh0Q+2vT}wRvfaLwMOsl;_~vdg=T4;814#WC-SHU*n)MhXTDLNB!_zz|tao>!
zVcuq~&s${K@zlINO?fMp6|M?S3qQ%oLmfq+{-?&2$cHXX^<LFJ@P9<d@Bab6t_O%`
z$9(bejl4I(xPa2(IE$>&Z4hDpU}>YBp(&CtYLVJy{Ib>2%mn&v3$3tn8RlU{x|26S
zuO6!&(})Y%q#iN{!kdNrp00+2K#^}*5`S1brl;UHX`#1_l!6)Y8+iAKz^v{G<!p<f
zgsN;>H33v>kBLKaeo~lQev-${|6DMya@gLF-J7NW&xP-<HbzBkfHw$XB*A1U`f6iQ
zN-EOM4C<dzL<Bt_7Xn^v@uifMN{E>Wt|hosp$J4!5OacRrhCO@gP%hkNQ7Rzd{iY=
zS?RT{=p<V-Mu{GV{|tbBA67wCN#Gnm^9<DePMxSG1&RaYTGM5`Wr5Kzn<bEOLvDaf
zqApaY3j)74yB2Vi$`pUg%i9d9AgBXc1Zih1z#j-#Omh|C!UOh-etjKS^=eD%atB?U
z$UhRgkuYp^kh`fv0;@}rOgUW<ldk*gaErJ&G<ik3^vh|a5X(id*Z+A<ub-o#N`|!}
zj9uItSky#Ng5_7Lp0Z@U{Q=uO!}XF@ZxyLJ?L6c%jrem|9dRZfQgCoH1KwQ&74W6v
zPY|Cn-4;<hxd~p)fA099j1m-X9fIYu<)FmRnWN-q5u?(31Ef4Lyd^21A|RDA5T56G
z4}D4`QMVM_CpO8vQD#QU%bOjOWl|)~?8utdd~K38<UAZ++daFjEgO?-{M>mrF{?(c
zN1O@fP)ZqIZdSw&4A=P(uaqi^N8`fl{KEIb<Bu?<R+4am(&)n8VDJ;UI5H(%;UESW
zz=f4hHj(-MGfkgvB+~(ZW47&@zwsp9kTmJjNp~-|?@HlokIN{V?UEruQb3=2$8Ns!
zvpeVaEHs-R4d#rzia0m9Ud?vDlI||^Hjx<K!0}q!ZbqH&8VVRVAu*$y!pxxDkD$As
zhO|r`LFmo<7U<|*u$`#NZIhQa4A91-Q_2Nx6dTFWD9{nO_%7&zfV}%CV#|7kTh7Xp
zZFtdcZ;)84*4Og0{$X4E36@>9lXgh5U$Q!?Vrt8!%;UOzFl<r&$CdXtL@PP_p<@mO
z!$V7EirP7s;6u3k({&2lxL94(RN9zplO@eQ1bWeE)F@5w_Rg#>$dwlJ#yg0rykR>|
zhm*Z*II*~B;k9+EjwmU}Mzn?4IyQC38!b`MmXD;9X{ekXxJ&Ranq)^$s+>9J1Bn}r
zagnu1fKQA9`|sGatFOmrBjiBrH}+zRM&f0Cf`g64d*xvA8|m^yxKj8$Ual22L7Fo;
zZb82nO&)p$#T3PXn$GTw1X~0iHG|h&P022(ScCqQ)-KospsXAngq;~v61#qD(kGvC
zfuFePsEwS5RkI^IqNLkF(Nr5(yo>IjZnk8oyic?8a;a{R+S7cK_dK#cpxwCO_&Ko2
z;h}uz`c^N+gHRy46rbGT1SD#j&eytg$C_=Aa{(Ka6l+Ow$mikSl>FdZ|46MX7{RW$
zf=-dB8LF7<-AQloemGxRBv?L_c4d`jNk>3#Cq|)I0!%$#<9dJX8&ipN4U9X$`-<#p
z`k5-;=`8we>AI=^@A*9~GCqfR<Qx1cCZo5tvUW@C-4%Q@xqhQ&KHk+dMfji~g9LYN
z;66+0Y7$#Jq7*1e*)C{tA!M~S9EQBRN+FIdtyez}>>eUw6`_-nq6%$m1@{$U$0d#t
z<Op;FJABO%fyIofp{h(z=zCHM^b{QZzQeQ``q?g_G7GbJm_(T(i@DV3>V5^f5uC<V
z?fsq2q7hOg%}uM12lV(OMds|yz`s*=j;19`C6>BQ%mTx>GSTOmIqVjMCIJg$1k*w~
zzYgcw-$6XPE+hiEfKwWxc}_^VFEs|%gns`3s}A>RbCbEn7u{@Op78kHh?uiVQ1D-P
z#GM4C*{GUuuSlWkIzsNs{dpLs?i_OS1|1;`Z_1av4@sJj11Gfm37liLBF;)Fl^Eg`
zjlY{IrTIZKGXkLI=Nmu4IFRcHTHeBQNNQa%=fd@tft+kH>9%0+TAki5H3~q|W>%W?
zT3;^+GmD!f%7gDm*#oyT-$$_q>tVo~ghh=~&__N5`Y*le?+&+uot~H!TFokpCG92c
z{%cQT0(HGGt{~p6?0XLJ^=jkvJOKR*^p?#>4xd-0E53yp!X)!j`3*l7So-}enZr^u
z1s(p(x+JU0-kWcd3EO-nOmZhyfsgr`abSHIL<^hZ&L-2MZ(t&%)wo_ti**Ujy*k8K
z6*RdCulW@sX7S{aTOEtC`>r#o+f#OGbCHy=gRbIO8ghghF7Dt1P5ClIndVn^yp^mW
z-ebK7k>s3YJqhRD2|Z$o9jP%hS?fix>_W@P<@UQdE4vY;;`{Z<4sao3?H*FW9kGZO
zhFg^#%_^QO9i{))V+{4es-e2Pl&H`CvfCr4+7S{dNAhR*k{0|44Qoz{iG#+Mpy=9-
zUEIkqC`dJVTR$fEVY>3cyn6Ag{mG>?rn8PCzK+e(K+W+w>|(yC>S*l5-7V1=t@9bg
zS-47Lq~s9c>VDBbjjKPoDBdY!!B$^GBHDWSV$)z#n!eOLVn=U1^po*}EVTUA(_tZi
zJ>hTr!{v3Gb9B~i;HH0A%F!`;o<dYws(yZE48Nba<_t**D?PF!Zr41$X(%S2pr-nY
zMk%Y!uXHYx9C-Y^y|6GTZ+JEAHE1%mfj+@jV9jpWvh0^lk0gTffD76%6F+(&9o0e{
zR4~DUVp!obG2$SARV5g^+__kGulN!x=K<r4v=QH&tTe0L(cOCY4@oxOY{oLD=QYR6
zK+@Ag)nRDT6x&t8S8DjW1q5H6@st;GGD5|NhmOm;$+Q9Zb>mqanbGe9xS}mO7UuYJ
z)}B0j%EkmE8D_>UT{@Wa>hMI?%ei<G|D_=x*}o70b*#CWL{3QdKJ(BF3+zIa0;(s*
z5@lG~UxG3B|1kDWO~P<Nmu1<uZQHhO+qP}nwrxLU+qP|2ztJ&0H~mG-T%KQWGBWqt
z3$%Wap6HuXrA|g+n5~bDtR_x7jLK^_3tnf4jvJ;85evqpoVe>mWsz>CC_pQt+V4k-
z^(1yuFRF_7#_`ZqWc{+*d*En2iKXB~vV46eQl6?V*zhIPE><~wX84XOTM!5wh(!Fl
z%gd0*`^b-LPOtc6`bO@_St;sNL9(GaIFyiMsaJAx^ih=>o-)aR%shzt=Ur)L`;29@
zCRCKI?YA~a>UB@|bzn$vS|16M$5Fh`in`zG0o{pWbsh*MQJut3YGk(F!bIX_-d6nv
z_NgQLgHNMjOvfWidWq6n#hPHsQO^qsV^gu+DGO9T&Zmaj_vj#UHJ)#0J!)0&=o#SZ
zXf7K2@_CWdVp7H!caf}J+$P`pk+f59V)u7;sK2+4GKAb0SQ^W`Cg;DBZqk!NlQSIs
z1k0|#$W2Vb^lXLSck<-n%flvh8(1?R^W;zADU&W3qDxah_6!q?4#+)G<ZAv?gwBm3
zyqn^GMzzu`EOZ+mcxfYAm{4yT(+T-}?1DaC>Tc~mDQ;pr)o)b~0Ub^Afdu3Sz+a-Y
z5jFxIuJ#2-;ot2jUSCT7X2GS4Yj)mbdy+5NReM)cz)~n|4ZuRFBxN1I<r_rdsCnGw
zjAIrq^Y?FCV@jo<3H@4TO+C1`PNqX?dWO_5Ny3r&I7r*pYLYl?exlzusDD{_wUVZu
zU8e;Xm>nJfjGLCL`qy|qo?WY(-T(4y!;*1cc$W~<2NJ7l1dr-G+i0W5%Q@BDVsR}D
z1twxYL%=Q2e#t{gU)_`>Zau<SF&gePJE-kY5~rP85i;)#({x>cJ;Ac4RdF_zZXy=V
z_zhuu{HRK()%Htfy#K3|3?j#d;g@$W(v3%ZC{OUk%Z6g`zD;azWR3Z4-&1#feOWOl
zEMF(rrUNKA5uIJJb6!HYeD8V?W&#BlazJk62cR#ZofAXwkg8?N=in!u$=G*_Iq^&l
zR4@M2OQ&p2-_O`Y!1P|^W#W?O@LXXY>2t89z^H$o*LtWD-((ip+mn2rQFQ(49g=O6
z6?0<_B)ItQ3fFEungsCK4n>H)N9Auu_Xh5ytThrfUfB^0+;pS9uBV+-xKxdr+QJPH
zl&|C^oSRtnwth}ECx-$wb#CcfCMPxAb3;Yqz{nN~XGe0$E5emMyWK8@!nAK;YX3ku
zCm7mp4RFrZ%gMfiPMWs0fFgNe^;$DBR|TbywN(CI5Dw(|=!}Z0>aAIR-}W>dfpJf1
zypp-S;vJ>FC!?~XS&O@|&U@CXE*6jBVKp6&-dSU(=iF|T@%51B1Fdr18{fa0Vwhzp
z^-opQ=RGb{>wO861zQ%8AgIS5KiH}@-;blVZ~6&q%_``PYvYQ5+i~sqGt+`umpt+z
zE&ngjzm0T(NJkG{fv2Y`2;;<4uXh1bwYXhxEm2NV+sopEWA4u0=%Ul)`D_OXC(9Bw
za#|hM8aVO_SIe}LGkr;Z@Q^O@Srq{Y4f;*=JAwHPzR0f!>HTP0)RPyO<Jt2hf&JoN
zM7GyY+j%t)AhXU{nuN=J>G>Hm<YZK3@hk|z$*sC`4t-KCMKkdz^To3GMRtR)Gcu~@
z^V0be(3T1<hx$en<9#bU1<^a^Cm&>WAH}Hl^L{~B`Y0g2*h%x1-T9oP!Bd7rQn&r!
zv<ogDKg{71c1MR>^>$$bJNUMzbXUD2U0nSTC)Dg@$Cy!0vs1h#`I%GWg~I6NK-$XQ
z*eSb+&fL9n?<uulwPfTQNdLP+zJLOe;q$K}zd8bNGBvG~k!iXTP4$;-gFjOGmqVBM
zS98mgWC$l_TGdwK;a>F?15PfgNRHj+Xn^&#x|VOAv|R0PW9SY249=O3r^+1CL{*It
z59SQn8L=;}l+Fp_AFnvlLe+UXPdhLwvA0o_jZ8jXGj_IEkkF7&5A0TED;*}nl+9%<
zH(|T|JV>Q&$aPO$mD|pSbvYb<arO_{Zv*AB*Y%kllY34FLF5PXVs7#VNot&Keo@6r
zw)nqq$DFy>Z4*5O2z;Q|+r7OmxK%0fOXI8KWxMiPB5PU+QM^|zt}lYP3uH4I|6Q`*
z^Xt|&!E8ebv?OkK$5+iy1XM=-tR>Y`y=PNjz8-ERg(n8LKh4-f-3FGwE2Ut3vEvw^
z&NKGXO_k2Kr1p~KSH<WT7a-z8pXC)4(vunvVro%J{04cgEUH#7WWB7x7g-<KCEXVN
zM}gJGF1!X2t8$2qXn2=q9_E*>l_vhymx70lj+L8xG9+%Da&OaFSN5D6^om=9gCm;5
z-F4Mpu(s0*ia-pz8)&TZ?YV2hrEyE>C6W(6lb>}5{zItU#!rbg9{e^^o}u(|Mn{Gs
zQ}dm{>H2Y1zu%~vwzeamt=jA^W;tg+PTGyYA3@Z2NnOarGKx{sO&9vOJlm7Iog!Mz
zf|k`v3n{@J2)oHMKrr#&)O9Rd@ZQJq!1ax8v-H|CCR|NM9h`3O1A?4LI{8FT6O<d|
zm-ME{C%F~5x2H}~svsVSRcz)bjDDmJYz7pXNX-(zRvHaNgzYLKACp9q5SLaSqTGQi
z+>1lmo|{5UF;eDJl1i`VpC&bTyNFM(b}ON|_L?s`{r<DBgs8dt$6<YD*9|P~GwbLc
zGWkbMNv~p9XkDnlS1M!AqqJLpwggwWM-+sJ?rn{pmE8woz1-T63C)17mOrEO(*K~#
zLG?*4n*ua(sTc>nHjdT+mCPNBc}&ztyX=9nl-l(;YWc62Q|&C`uPKjio4h#i|Ncr~
z8fuocS*%{xaVQj_C|OqfbzRi9v7w!c5v)J(<uq!L|2Zth{(pz1n3x&=k23T>fhZFz
z^Z$AM*9r3f<faH1nK_x+|98{J|5wpR2dDyy7dE|&yL&Ui4k7K%4k;2eO=rXKBJQH@
zqC`2tA}uW}(xL<fDIx8NuDs{$moM`h9CO3m-NiR>jEh!R&f+S9bOO*Y=H$fG2r*QG
z>Kg0m3E=(xz4OEUy&(hB<%aOTf__IBG2n#kZr)+dXMLn5I|2ffunE9I20_;dW`Pl4
znEey*d#I>KsL02AfOiiMkzcfsj!2;6fV|gL09U#IsA0_GxPKJTjt`E_?XAIs-`_uo
z14biI`>2SBhws9;1UDc~&Fw&3z{`OO|4Brri%}6^=0i0$f(!Wm82>#^!$G+so}9iu
zJ{|&bbvg@f+mLE-0_eRea0kGR;~t+vHi3OhVHN<o0R5z5p@R_%%+2op$cNDlgI$8U
z00SBTtnWe{U&ARsIJALv0vhZBc1>CV$fEyuz8<vlhjIY;rGg7ckAA~<_P6|FGP(N>
zXPTdcIy?q)d2MuW16^Ma76gz5Wu>cNPXPvSY5OIG0O=Ms_!N*Y;Ki_jM1CiBArKJM
z!UhO2eOLFpRkM4gQ*ftq_r9;k>GaPGOx4`o&_X*r25}1OGX9Q|pWTHu0j=(af773F
zbqVv}{Q8mISm)N>`j#5n917Lm+t^(JE}?l#hk+S<kKG6=006{86chy10~kjIcx-Dr
z`aKbt8v}g?e}czI6As9S9tJf4mIg*Kum=wM8Mt=__X+|~V24)-*z3pqD`oKT1YCVp
zQwSg(U){w#YInteY5GVGd)b@W#WolO#^%=pn5(b%=gS-<o5nhXdHtaM6_4&LWj#q{
zDcSgw`oJ%Xff>f$ADbWt+DAb=2E2cKgaQf)0RaH~YkiUj{hj@VSAn+pQ+N4ms<^fR
z28{U51gY`zCqMjHGims4WvdDJ^O_GBIK>1p_~8Ae_W<#zG1&9?%TM2jPyOrP_%l8E
z<NE!>P7DIa_y*5Ekq`WX$3HiLfBldKVw=NLShuSXHo!jo?k@@};78ZOHL<-rc<$FR
z{ZAiCyg1v|z<GA;J2~?7EWf?0`)?Q9q}CL^wdu<_?bi-$8hm{hvjo(!{m<3iK#Pax
zZz#AnFO~SZtz$^*6&xC5tLE=B)x9l9<4;SF4^R+beg^hD7%&?EgB0NYDD<X^b^I$H
z5!}I<Q@DTu6@X?8AOLn4_DJ0rG!$UFslR}q(h#71+g=J}fc6ibbQpSH?OXN%xFgkH
z83@2`We;UKYLq{P0X{(ciXC)m=_mX!R5cR&+{XF>ti=iBA6*OVvp!Ft|I;r{d4Hx)
zzXu(i9fEyAHMou2?`ZyD&c7&tho&}=zp+i_-MyH<tRPSjmvGme-5o@>NVmU7dWiit
zkggwSgiqiA&3E)s@F4@lpkH#g-~Ro-&%@AATgYsQd3%Ue4ZDoTdhR#uU``;<t)U{P
zu+2YeZoUjvYzPdP2d6*#O`nEq8-HGDg7!bdi0>hX$A-_`ZDkC*JDBdw*M`t9Y?Om9
z&_R(M!|>NNkzZp_G-3?4(7#q%petMqqc3ZBf81~XAoO4GFxYK-%RqmF@w==+814;j
z?~Xq=P+*<`1#k5GyABUe!@L|rdWie>WPOP7*Ld(PTo{)>^(}yHx`ry@8pFGldZ%wV
znS?)hwCmpBTvqt|elBS17U=b%AAh32&_f_^CxO+1zuH%NPj-muwC?cIuYRwIDd;6f
zV68vL*<V1z7YE>t0KwfqVp<^Rm7$M4*D%|A`s20KpQ!<Vx1GtM8Q8RCd~wXQ;k{>k
zdwPgt*rq`ZtpJV4Q0^Y+yB<$!*^7}6Qy^30W^S$t4+bfK$>)pK{9G)nwnK4sOz10i
z6XW_8G-_fqv~S9dID4{Q{pZuL8gmL%frDf4c2hY{XmP)sx6L4n8bXvdWuf|__6PZ-
z)=EcOnFN^(WmxoEeW;g*yH!70&vUmu-~T2Y2cQ()g9TFsp`+5LZg@na*b5A{!s1c-
zdnAZ$87AK6l$UZm1fv_|I#aHrasPLlY!Z{$x?10HTr8gvf3=cLee!p*poE-p^-wy0
z8q<s%4o>yb5!^UQv3;qw|4eduJ<F6qmbeQ-7C&}v78S&v8GFSOYQk(F%FGTu8^wS`
z4Fq|6?S5jWeaXdd!Pwa~tfZT9M>o`9n4?_Q9n}ld$y|O8g8WF3umGin-Y4asG$lQU
zn;Ni&ot$v23^9AxS1cQ`qCRd+^!b)u92y?)&W$kzXc?oX)1wX_znWYD2|<Yp(wK_D
zvxZB?*O1_ArVj<at%$Z&Pa}F);j;^U<H`_2jI0l12_(gJ*F%B9$7rU3)to-SY@UJ^
zy_e(?@toF1eup**a)NwGYU1rh>&}#kBbJ{cz<%38yx;O(R^1MnSm&cf^rgY(4-OEx
z@zj;p>DC})`qANSQRUF!Pat(N`GZrI{?fzMd!x&@Ux7AhCKqm#GvQ`sTJt~Kvqd_=
znxLO76MY{H7k=`i@5bYm^MhW2)5U%XoQ9(N(LmYn-62yu(Pha%#1~-|XZz<ZdGUf<
zrz1zpx^v-36-;~b{#SQKa(gdoR$aY9T%;{HswlW+@bsYS?Bq}_Q+2S2;#b))fGb?|
zS_|8sqqbS}eg+R&I?=p4kdBO7){7jDoz+gMiNFZ^O!R#y2YY_Sc+(_@;UTD|2qfXY
zlrjOZId)Jd+772d&CQmh>wwZ#bjc{VSq#H%qx<aJR9!iGFUqxuZEMPo)i7ULy2NQ=
zR1kZfRuA=lr)-NT-}}U{7nS^QJlm|SE~P2Zs`kQ>OnzO8es~K}`X)IX{CXxOr2bW}
z1MRl6YhfQ{wMa!Nx8eV!FMkqqERF*@%a?+;`+<Xeon5Uf1f^?}Y4hFG+)`H3IHDaK
z$5PsIA<I<XB_D|-VDSaV;@Z>12=^YnH?k`MUL_dXu45?+YFtX=L1}X)m(1{xM(vnI
z)2jVEM79tdVAD#PjI`s6`2CF0N|c{gS|%xt!0l)IcywB&oR5uxa+%DwyEok3yG_!$
z-s4Z_HH8~%@*eIohwzkW<f{s7j!r_FGd}B`f4<x4>05yVzV%V41|lwY&IWfSSW_xU
z@e3)X!WL0|?c0X2&eLF5%kdjcx5Og{VF+OB)p)j^`b#BVSxXk8N>5g!<7=fHLf~!F
z)!q9G>VFq<@?^P+okOlGiL?7OqEf19&BF!bn5j)?6ZvJB(xSIcBm|J`ErEAa)%q5g
zge$DZAGIdQXqwgV6dzqp8oN&6pZcVWA~7L9%DAiQOScV<%DFT-59A(@t~>@i>YDE(
zzOv2j4^t!!w|0nRtFemB3YQJjlr%)B{x!SOo`R$vHl`BH9ec(RhKYG`|2Vl$M287a
z%&4HZg*p(*=g5~FQfu|x0&YchR4S2TQ39Eje_t$k>0^c1ZW05a@NJqV8ro*n9K^1(
z@(4BS^D`axh`_54Q+3KNJkXNy@V9TgMdz>$r8jH$dbBAYr>C}(CiA&9AUJ&aSy1lf
z&V^~dZkzjF)qXhK*wm&}qQ*K{^)iPhP&;FMsC%qbe3i##9hQj4uP8x(2Bbu(GK^)R
ziSW7QK35YD>Y!oAA!Jb(O)%k3wqI?fO^ozzpeMvBG|R)#$~fxJ&dEH4{=10eroFY5
zIXLLumekfZtJ!ez(gzC)Tt9=zW*w7Wg3Tr&)a`l>HH!4Uc*8Su=vu&71TxnQVmf@M
z7R{Tb*7Vx4ux~=TD(AY;pVd)E%5o}}aR38-7oCBLl|Y>~YZy~9-8vfJSJ46_*Ub@E
zYr{6@pZiiZv{nfl%w1>f`m5OFp8+3e_1#>iAcWlcP%>`lVdlhH1+JN9>s?R&&rKo3
z!POjs7NpF4a}Sd2ZzbmWnwegFejU#qL2>~!jnEo*-lAK#;GIlSzqJig&+R=Tg`H=9
z{Q`62LAMyeVf3+2p1?v``rZ+tI3DY>30A_Dfd6nR7oB$*lD%4UqVY*JDYeKP_8Gh?
z)qDPW$xH@vG$!TR@`};d))M)9on?%BA(i0Y)S+%QqqFo^8lKc#i+Kf%kHdaK5b)qF
zByg+*Z@PU*MVYp=KGk*hh5{^V#WARSN1t|BJ35<PMj#>$!OF;KB~|KtQ&o+AHuF)F
znO;%%ah<PL6XUxGPR=C-?X=GbF|#sugh#_silyKQ0i%FNkxB1Vepy~mRoI0}Oxm%U
zWc?yGED=IuIBG}_UWdyHlO3~%Pf5eKl?-VyzV2?LaUs4l&uOHYm(O&hZ8Cf$n9%3{
zeu}aSh|V6_4gcy66+G?tuJ)0l^CivwkbN(d&m;7=JZvhj6@>gcP?7|c70B{IRvpsm
zE*(L6DIA{Pv8SV$j3b1uPqa%N*hHMxVuJm~%Npl?feH%TFxh8_UyOE~&GvvMl}y$f
zFKzs`=R4K3WXAC4ODxt{1V)2WWhHDWaol~YX5dHSS1xfkrt{)LGbn+pFuEwbny-iN
zD0Dm3yE)%;{H0j?=h@+}-Gak2Gp{VSji1f)C{8!Sh%)DDuRT{9&Q5GSUxdYHlSJPp
zbyw2*;wJcS=OtMRK9tPCi*WZg`BHS5zO`rrOFvj3HPV#LbjLw2ZYt4)g<YzrM#us+
z^7Z*)>4&@!hG#iCo-lUh2Q2B87X)OX@2q=JIQW;h6ls+o*XOzENcb}wCMT@FZ<EBG
z7Qa0emsFllo=?+fFAtXf@IR@eD7};$Z-h4h)_NXK24<iRT%j4a7bz0^F&k?=jy8m2
z#S*_#FVV9bh0U4JnLuWTVf(Rf_x&xmTE7#VhTM(F7blVbxN9?7IrDMRQ%&^8MvLPb
z7*xa)HvJo{oKB#E1WtC(K(hL3;+`}CB&J_klPVHSd<4{aUz^L4qRmKxTZg7ecm}6j
zY<EI&<YFXy0`tMJpenuXR;x$44x1jcP>w=fXhucPMBk7X4QufEKr}r6oqbM2<iLQW
zjj~vgz0R{Knj~)OY3nBKtBN2BH{D%57_x<!qT-G+0*FsX%pekCC7;}Kli#0|YKgiB
zh`O8UOtqH6v4$=$-cbXZif)M5#q^Q1R;h+`ujCchPLc;i4a@Sb`7>-jsj}x<n}v#y
z;_?{0>gxgKIo+vHmM;>oMbc+U2$#%l<hPDRRJ3$1khbfH>5o_0S!DKU%}Ysh0<#(9
zueb<yB10S3q`g>C`c5~}Q0_*N4f2)MG8+W)!|)FgrOW^dIHQ{t{~=tSgv;uM=>^Kx
zZ+n#uIH6dTF9bE%%yggh+}v1cWxD1<ZhU<i1NP|)*V6r0bCTLJ8M}iDoc!$T%<(lf
zp3J12=uZAgZQJ9^f-kw&T71vTT;P;P(4+7G!S}qG82{;jDmBUe=WlWU`&@_3`W>?`
zP}Z;34~6&b2GQP>B@&QZUrCdpLi1La1h{k0$vNtqd&Md8KsqiY>SSlmm=xzkg|%0!
zL@4@j>s+mZpd{xa+o*C1;L5~~O}yZKQTZc~;f7+u+r9;8KiCZoM`sRhwTKW_>F%n%
zX~;JVKX1Y>#;|^0scg?4v9PfN;<y79k^Zt=(lJ3CqBRv{#qQM8lTaP=80vmMohnCL
z?v15K7I|`Od-9WPS5cNh^VG}E@&yLn&FI%lX1QzyBTFKw<kDDAs}`Gy^EvAo4s=&A
zXb}2V?zl?*j?|=V(wY}9EuXO~+sfWZvfeX@r#>~)H%$QyJ#WS#6opOpFpj)+C4p}4
zrCP?gRI`_Mel8A`+-uodSR6`Tnr%&W6m+Qf8nh{amUu_5sFfup+ZOs(RjYKBez4@m
zh3c<XV+i67c)82`n_{IT17zHR{AWB^lax2R=iq<*>(j*eoR6IoD^Ek9Wd(RCs&253
zsW3T&LhR^bAO*u7caNI5rRkD{n$ipQc7fSWQ9XYwb6QR|h0u|rLKD(iCe`{hQbl-j
z=3Q;czC-)?ZKjskcfPF*Tv|`pZCPd^hnMvIlR;$^?+nzLtix0Dpa8yEx&7^2)AO6K
zh|c+RH{xU~cRIa$#uSSa3bs63A~UH20vI*fE_VhS5=Z_Mwqym$rxe`;E@-s%c2CG%
zlgnsX+?z9LD~nlTfj*MiR`d;CrbEsBL4Tml$o4=jP|#d)YTH)>@<~tIrk#v*P-7QG
zas+&>{?|k~LVVimm{%xGqP^LHGa~}F&F!Y5fOEYME+!5x=6}o6RYc-kt&JA6-PHrv
zaUx7JiQL%Sy@xa1UZ+A(>$P6FmESDXSd&go!RZ?(xILHOEGhbxR&7HujEQeYJuLpx
zv5y1=Z@3rp#AK@={N(!e=8rs|6om3-TV(IT2g;<;q3y%FX&VywI<bFV_p|2n<|$8(
zapv0=NvS?^$2A|u%53PiuHqNj+M?`y@)~VdG#KeWRTK}C7hJ8DPMw|^+w47CbT)3L
zRU6i)J4W)y>~b;%%u{VyRVojgBQWhScf*kseBpkRsXPs7hwVlq%qR6l;{%)-<yqmz
zCd%*%MVIB{oAfj2$*pv|(;|3|&KE@fQ_{!x3Izf3fosU7FN?WYS$-wsKE0%0M8!Ly
zs28wTsFTKZoVIL!h}J(m(p$E2mdIsYUfg2J!W?Z^3FOMoJLj#}qv(FH9cHa(x3GF%
zT6lgFB)7RY%J%jxkH#CEPrE8;mt%cZ>K)7d3q^MdxiJw2^|#*^+ptJPDaVy&lKn!>
z?PVp?YSt*tt|F)xqy~&?r!rB%+RJ))(VEl9cYdjPV#-LiHNL%;H6peX*_$mZg)G-=
zm#Azt#%4UpvT;L!4_ShRMs&iGb{EqDLW?IbE9R%F*omuL#X%n87JK&zG9;5Li>if1
zmqls1!UbZ*lGI6$6(52s4K$mTc#(IX;@Ho+&*sgTmX2?P_h|Bl^-Vl?K~T<TL#J55
zq5*JRP<9lTPV>=G2H&)>eB#+e`<%nVq7nS`5O)O+)6P!?XG9XNBlGm?zHYhtkIF_q
zrJtD>@QT7tj6^>2n<eQUnd4&GM&k!eEtKaACt@LyI$GJ2MoFM9UZjcubkk_=N%j%y
zpJm*whLnfQVvO+_@RUHYcg8_GK`Vtwe21=xHd};gO%XeRX6w51f>M_4pF#}IPcsIZ
zpFf=6(a~2UL5NhX2q8laLbWPngcK;~2b01J>M|>@7mpt@Jh2+l;M6~U!Rt3wc3xun
z=4<O0KczMX!6xxi8sP3xK)nTgYub2`x+o#gud1py8xD9*p9>q8m=8k<ceh_x;)c5j
zs7o{QsxC)*Z$VksI)IHI|K^7p2>`#Bh=OYyC&6B^z8BN7*zfKtLyfSWcZ8d*(W%0H
z-1=kSN{TK~hbN%6sk=7pE{b9HA!C?H#QLI)AN{k}myv_fGWyCiTAL|y{w@K!8{kBD
zo~FQ}fhk*fY`lR3etfIOKfV;>E=n#6*akZYH>77qOrj{raq2n_L)gpDxyW`(mNF4y
z&Ee$xig=mAZ{b`X`v`T0J?YT|qP;&KL%kI8R)C_<i6|QN9xqLY*$$p+%2}!u+Y%*G
zgS|<8Q@cAtlpH&~SS<rQ&Gu1wj}of-L))wd&Whl-%<-HMJ1ZAB0MBIp!2=Hx>A*vu
z;C82><T0sT3qJB*CutA^nJUT{BtjRn(MDRDZ$Z15hUzL#@L6mcAw|{ootCU1!~y5m
zBuieLS;-PcB8U%Zk1ZBW(uFV<bDcK0k1d;$bao8;&OJ}<`7UZ8ss(Fs*L3&~9*jk&
z3`}|nNj~rzgH76`*vZli($TB<S2YF)rJ3TAU$a|yp`9(Kjp0hhaX#bH_a7+(IaacU
zl6%%L3V}p9j;;{bq7R4O8}UceN&5*<%S#9%39+&67lcwbxxR++m9Md>l9AEdIXOsP
zaU2NAjLL8;HZ!IEiNj~ZkogN|ggb*12~UMRBFz4q19Fx{nJ*r}0kb6=`hty~2arQF
zvb_ylq`!nM*!P(j**!=Mj<?U=x4iG$xFN_IP0`GIE#SX)$5F|~zvs>F%pAus5+D2m
zE4{4tvSv!v#uu$qde^9=R2{6ytHJ5&i-)DP)_N9!jLz^ETo|?O<X!SFvnvqbnIyh<
zPTEoYfX*+B2$2;NFog!N|1>N32*StO(D`ZZoiC&+s@co$>v^~uaa;rv`*$A!`B0-3
zqYEe$#0cmJd_^Q8FrMgf#P|Yfj8LUC3C$6Qbtm};{l?{fSFBX}Z4OoSzQvswJ&r8y
zc9LN`2U<R%H^}K&Gq-->gvCcLF)Wy)LTGH~>CH=}*Vuen`aoI_2vLU=Yu2%VvOd2F
z`F6>kAD4C(S*Wag7N3V#%>KpnG_?An7*f6~<`*G49jf>3xIB7^tR7cf^tI4)y>2+}
z2GeyKe;1AhQfzTTN15qYi;X%NRMLc51>Y30t!Mv~6QNM8F7A4UVbywPVP>*I7d4x#
zE!o4|JJ0W)uGp_)p*tsNHV9V&zn3VVnh|ktkJc<Oh&%k)W871yTerl_*`y`F$O{Qd
zI6P0?w^;5Fu~#7m`4-6;>g6>T?WJK&p6tQ%#z^i}ZZ4azrD`+Lw^fa4*acDd^gB&+
z?%GCc@hpc{FD{oVhMd~NzDO26qBc3TTsBQ$T{Cw~Kyl9yr+zTT7mHcP(r*s?d7ow$
zcQ(dVWya?rPXn{JMekL6)o#oqZ7q@}U+1(h>*V8GrgVy#Akf1~9rpxOU*lTRuh{_8
zm+~l$B>u{na?TmMnb;v)xAl_v$@ShW>;Hsy6Yo#?OdScU{b8EJ_mtR(QU$Wa=t<s|
zPb94>Hk(^g?5KrKRO{k-jnflwl`Zftm5=E%bM44;DKA6bIOspNgdxAVr)k5cpDi}Z
zl|OQ^R+PE^s^)x<1}|S;RFAm!HMC2JWt;L!uo;>jbK~-;b{2dqEba&QQ%>a&cXE+h
zuc7T*)-?AG{UJZzYYe_XXNMYhsX?nmq2x<Y_E!I~#Qg84ctK(3svbCInNYmOdz+4H
zi2X|XuKIl9J&T8d=|qqci2WK^+J0v$Kc7*#0CA-~Ro~2Sw>4(bPu$V*Qw?24VV}VR
zRD!7dUi5_|d`KI97s}+ANae-1P*4Q}xzD@b_rFwbz+@vV)BaZkeVrT>p$X|BjAA1e
z+q(|YXQGgQq<xvHQ3(6xQn-rOHlFR!>zZK}d^+=l{058f-DYQv)*NolVpVMBExQ^z
zdN~AgKW+p*6k-WxvMH<Hjt>X#egc{Up6aebgyh+5Y7c(=#O5#?XqPjqmBU`qol}K&
z;O$~B40e<UAD<;Fu5MLU|JmClYwU#VzI3f$evbjOnj`c1bdT|FWZiAya%iDv1s-%7
zs#bfosA#&&Lpy-3KACxQN!#3M0&@F%GvPJjyQ20MUO%)F)-ut20x^4?>ef=79nCl=
z5hrP3RA<CAFNND$W@8w7{q@iLWA{ch`CMuZvJ{{GY=j}p)Cz19e|&&6_DUNL@?xIm
zZS9%`HogARq&HC~?RS?GpK9O{II8QPc-rjMPtfwxTON{gjOb_k2|5w-iu4l}X~>mR
ziZifwFF)&3Yb36FaV0DEN)TzXM3dKdlZHeh=UTwfNc)?`=j{nNvnCd;yoL7kXJ*|Q
z5tPjBifWXUDTIO?=Ov&om~r`0_N~?$!3}#B?n}xL6X9`(FwQN6C3x+gtJ^0nd&P*8
zs;_j=%8^#kBDK3A`C$*#W%flzJ9$F_)`(EF<ITrKXh}iTCy}?PgXcxa!wf18T&Md<
zW9VaT#?gy%oMmI+HMi@Pu4#2!&?NLi^k8CP=RvPS|C%e+rPKOTNr#t@PHxv+E(C^s
z$_4fi1v^dSdJP4-2cpw|>3Hws?@mdRL*H|w+P0|Ds#dTG=Vk^8bqBWfM{&jyfCoRB
zuuAy}E}o;4nh_EO_GOYSb=2oKfH7^UPT4OvnV((*v(K$#GQ8?9cXnNl+Hal|BTBn6
zgumU_wDm-4ycfk+l45v)CqNsX|AI{{VQ67Ca45#aFsT<BHIUxw;gxDyIK}p;^?6WJ
z(o&IPDmL7HxX2OKJYIW<qY2)EB2E&!y^P-KwkO&;I}u>)rbo51C{vDr7q*6B*_B8I
zvI0%ZuArNxn`+)X{Np82xR9ok%!O7IFj39<&gy9S97zJG>1x9F9HX2Pihk|9^XlL{
zu80#=^!u_K!hby9BptT%96g3V*#LP$x+yO@>4`6mkl)FrfSPtI<Q{kuO6IF`;l6iX
z#hcnO9sAYHG>Xrt2<%Gs#!<Xh(E9u;pe=jvj2`HExTxZoN-z`QJ5LBVh@ACSA|xHW
zZpywDW@=J=M3doLW4PuMEqup#I+|hi%{vegsCQCLymXcMtl~Ek;vnw}`p$k2P7B@=
zsm+Slh7*K0m`h_dD_xUrqJ_6`O5JgSYlq?Fbbe|<yurKwtor}N3@Rkv{^><n)Te$r
z1CaMS(R-4By#PEHEPO-`8=I=6ChYt$)(b0Kn}9XN)v~u_nX4+*G(MQLO1)}}Go9+5
zii(6q!n@xL>2{`Ga7v&uH>P{(uk-UX{gy!KG)Z3euxgsBN-nSnAU;n|k^~oyY(tt}
z*zLDAFq5HeD7ue<t8|#-GlBA|s6&D>HL}Z2|6Hv8_(bsqnaprn`ev%u5*>-;N~g8w
zt6HN-Aj4J|6mm8|nb`4N+d;T5$MksbY!Z?Rb0_sa9(~RAt{Jo#z0Iu?O2<LYSVBg8
zkzeP~pSO~-q-lJzZ~)z+!mUcCD(b+7oI<9MA~KWF58J8I4&1<@04&MHv_Bo~2;P;b
z54)QtE=vxCnSEsu2cv#4YMC)cy3nUr$c3{;PlCvE<RP27=v(=h1;zL5@aCoc8C|ug
zxxGZ*Q1Ga-EStL!ug}-;qPR=2CRx{9`P}Km&0NfEq693ZgOV)Z`Y<KYUaS28slwr1
z9GTNI)l(nih<EO^-)3`lI~X4Y;{((T?zPv^9D^`QX35sSf(iM)Q;slDM07zmMsjzM
zUo6{y7VDlFzo5px>YV~pmrm9Mr=5?Dm8NrtXNtvAw=m$*!gix5mzcaM7d^Mw;OcmH
zwJCYM!ijlUmWlDrv^8&1gn1ZT=XNiadlN)s^YB`DEs8|zEE#*uOq^ZcS<1uc{=4lS
zhgkMNJ9njc7<y%nd;{{S@Ik{QRXb&7M7<vncXUJBKX98%KJ8Q#*KYc}_CY)Kb>)7C
zqvVUO7L<-;kssC#N6b`UgafDCXEDDj1Dkmw1DNbCD*WQ1BFu9y$UPokUm%;3MEx}l
zC5Yj%w!Xc`$GpX&YeBINu|Anf4_hnueDkQeX2FsY=Zwv3nhiz7uAr#IAZb<?8~<;m
zsr_&O#Pk{k8wtt0>Rgs|kBlvz`P&GH!H-wZH(ezkO#JB8!bol7;p<{gm&)D7A33A&
zWVu&$!_4E}eT57<Pmv7c7cuCem+%^@IAH!vIhL7Esp3u;E?X)zxOX>7cSp;w)^S>m
zUq8z_)ojG0liCI!Bs)nZ2yk0{es<*)tlZ08++0Q5>uqwoz*Bjm#vK=%EGGhD$Xehh
zOVtKKEaeuwOHwE!D}bI?I>`liQFyozR2DtXvYiE#sh=>FNLPK@)hohZ<tO{(u62N}
zHq}}dY+{06Wx?+O2&wQf>@HnP)$sA)8IO;7cKZyg&8n;Sk5pHuJjQ8HqA31ry2p=}
zebE;(*7`x@h_HonI-Yox_h{}w6+L<!ts9}2c6!;HY7f99>2>XE)6tS?&iNqg^)-0t
zWQs9O<rMFk0m3t&tH3bzZQZ5y*;?yz;@kwfQ#<;h1$thsJg-qUAlj-6KR4D?S#2pU
zXfS9;ysUkbQ|5!j0NuJbS2Xut=V4ocX5rK=51&9^C9X#5B;xsvLN#AZZcxU_!E=vo
zpV^b6vst3Q#~v8wEh#q(bd0J^Z(}ZLqD3Ut|BXUy$;j)|cw_h<pOTt1-q+MxWU3O9
z6hc5FKMFy_WtO3R7+(cH2EIt{%ggypNuDsi@gUZ&1<SuA``j+N#HEbc_yVuAxq>3}
z&63gh-3zcBtlb}9FJy$$H(@9dVTz`Q7d*8+5S;x+eszm<16Zsl12h%P%ETV`2^jWb
z&GGcfJI^h!=EZ-I?I>FXyN!90(GR2qg}Qem?Dz8I)vm~&mupwb?J7q0j~b(*TDxfW
zj_f3EJVT77#eoXDnT9qNkKWEiLR~A?^4_xwIH&!nX~!Hl$`p;)iJ~`kk8vB^&>D`D
z&HfZf<Pc729yi5T>f~C-lOz=nb_09&Rh>(3KX^ue-hK544sXDoKOlPz5(e!lT;NYv
z%iZSnPRy(n$Z!3<tw?{7#%4^K35xKC{d(h#8EbWl4%jlEZlk>6N0_cS!>8i=UB5Zr
zmI4u)%~`6bJ%zu@N)j|(`*x&pxLt2eC8b=i=6u0-8|q+N@3lgw6NYE?saL=8&PHP9
zG!&TsEVUPMW?g4g6~tNfr91=kFM+PQaRWsFmoS?;hAZ|mpezHMCJxgCnzoblP^V5d
zOZymjPuWWVb+!s?iKHAMB54}49k##e6S{VlH$6v@;BwmcwtKQ*4Ts!bVct;;{;8>>
zVx=9;?JLO+Zq-Xilp{$yjQ48E6ADZn1K#P{3OxZz@0t^!E}sbOaj*qGFX&0CyFb^6
z%OTf_&Bt{D!xwH%s3-%H>?b>yp+NshX^{l+|F!hnbYNK&E6w5r5x(=#2ijd!;*5ga
zSc!eEEoaYK|1qYW|0Dh~*U<Ly6(2FXxnBqN!Y3Tx!hfy8?^oYG;O`C=hoc|Ub$Zuh
zyDIw+3Mq3mqF>Uni)%Vhv)aSnV$xj7YF9u$Tw9y9xtd3`O0ZU(9VY7%)|SoR(cme%
zoxHgde<|ygs`n$Wdsgx&<x@-pBbqKA1okPJ*xK;PqzU=p)X?3VJjSI8o5n755V;p}
zWU}O)*&2APVP`QNf%HgkS)CmTo26gVu$=0(S9IC0QJ-RJV?aOK(JFeEWLON12;IyB
zWUCa>C`T#AfJjUYw&PE4(NtjiE@->R8WXVUSoBTh@~RrV{VhNT?8yGxL`Yv(0j!Zr
zNR}KV`8fB&6Qijl&;7@ZT!P^96q%P$R)00d?DXFTKL~{Caq5K-zh^jKW(n65O<ig>
zw#2-~&NwHOBEg}h*u^O&s5;4<3F3(+g)*Xhkm-~cyLbK4*oo$mZbVSS+mL=R4`ld7
zSiV^A-`{+Z&ZLDd_G!!(CECq$vkwbl*UcOAog9jH65QA!JK1TG3a`>&;ta}1i4RbJ
zs7UHNDlF>fW@3WSQhF>&2#4k(PEP0=%Lfa`041HOuqWq;<BFk8mJ~#xX)q<!9!sBc
z*hRZf(SlSJLdT|>;ADP^Wpy(4faXT3g&;8qf+nPGM6O+iK$j%nrV+YQ^M~f!(raO>
z4c#8wYktUF+M9JTwXKYP1`>pk7(+>^aFL=Y-(0?oAHVwO`SRDbl%nqZ2FGE_F3;UR
zvByW`8~H(ZXLr>oS6s8uK-#3~JqSb>Ow|9btcZit%L^(9WBoN>eiI4Y%NhFHmQ<qA
z+mY`bXk>hA$o%<oi6>8h8+4KT)^RHCG*4S+y0Y+G%h9m=orx=oV@fSGRq(NSZxk#J
z>ErFo<~OUMI>}PQXAot33*A^I-F{%sSn<pKy^hUi>GD-v)&KblwbBu)s6IgYI_iT@
zfRB*$rhKWg82z9?Tp>f&$21gb;w=nC@{0%7H8~J=!-m_(LkmX{Zsq{-+jK<CabH8k
zLDG|BA78vKeL6lP1LFaqmdZ*pI4V&>Ct)kWRrA#(L{vwYt2Q607vyolOQod;^+@fP
zQ=dK3Oka1?Z@57x-cQHa2^PRti`}OEc%QXEy?6=ky+o$w*t~~|`9L?knnj)8Fe1CF
zgkz#kB1b{{OY(f{pdx5!qsTB~tEvW5k8~ta#PAdPskES|mNpwYxQjSDTW|b$oTYj*
zT3?jgvVxgE`}D#a`l^6Sw-hSWW0T3XX8F}(HgU($MZi_G^$9#KUF>(zlur!PS>UGa
ziJZnK)XH%vkn>{r@F@?7qojsSqZ<AYx&#`1L$orUAX!~h2%kPgW#)Y(TQa85+FuhA
z@w`S!cb5`e9BsgpU!&-A&uis8(c@PU>&~eqrc_RfE2LG)9^;+q2RqJ-@F=BQ(;p>f
zrNml>=`6VOlaWoA{buFq5TBP}TG!l|sHwL4NnvJ;y{)o)vWWFC_(eHs_!D;wYTxVp
zJV^G%gaCNo4X^FKE>_PBRYWt>*<asr0fzs{j2iX!0`K6NRTp-S$_qy31QQ1yKs%kv
zH`>*%HEv*;qtA#(9!Vel^zLGu$J{dT60jOcyOCdVrE&RHv#1~odk(Q2<ASnZmlVcE
z<?dgI+SQD|hABTw?-5JO&$i}WvA-0PgI=(Q9e$Jt<3vuk2n!!NcDpaFLl1>Y+mnlY
z^&ryFG-+F?06&Br`I;|AxPi1lYS25wqP%`kJUaNSge>B~fY=HA#~j*d(W_ve;Nx>!
z+yeQo=wr&rUQ^AN#r2|Bc~P;Gzo^QZ>Na(?bLj9M&(dYgHnXh9YQbz%T9Q6zT)<5C
zw6vKFBF%?fEmdf4<fUMEO#p;h^(y=JyBB?C^mxtyy>#X%gS=*gFzm=a^$z2E8&bW5
zQ7#4;bV8su;<~WYbEAy$$Jy!(m~4+b^IymAk^Fl%DgcOvPW}f$$^wR4uK02>66K<|
zo8gNrC7<uIsRezs|MSGMH%^ZuUmV7n6cYZkRUoAwbcNTmjL~=!nxFFZT_U(i?fr)f
zHJ>)kkTzxcnI)+fG`ba(Kvm&yc8w8s0pj)(<P)O!ATW096v0HT5SfX*Yk6uGW4pJ)
zP-V@0>R?uLb%#1apD@rG-ff1)MY$&K!simi&*3v2J%g>}YVqoZYpMJnHfruXTQk=w
z+39WK%z{{<UfoI{{vE^pFd7S{yU*Pv8g&>F!pZ+_7V#X-+<c24Q-4HIZs81Zyf!V;
z4un}EuO;eu(#Sf=J9SB>EIlof`OIde`38`z0=>WU5PEX5d#(yorPsBed7wpeHS~?e
zi=_|q6cFc~x761w*-RElF}7NrV#3OWTUkRNaEXQ<aK8OP5LITKx69>Wp64wZ2Fx6=
zAVmb_>A_T7Acvh12x%XC+Gp{~5j_~#1NU#u11y9$q>$pyWfmO3mC^d;UDd)sJ1M`l
z*eiSy1}(o2Rp;9A1p85oQe&+p9hQaW5QX16L<va!Ev<;EnmFNNSJ#@U_=Vmwfj2ab
zmI?^t{scIPc^Jpdx~<@SHyJCY+=SisL!*VLCL@tVzW_5C^+jJ}?U|QGLe0kgWetg$
zKKJxste?D)+euw9#EC)PSC+HTw=XR{N-L}}P68MNmh@Q_ybmYluA`79ykR&r>(QaV
zr{wcgjA}@{xOAie5>lUXBVGsFJ#_M!-8sS^k2zxi#vggxTmZBCpJFGrXy<j?^ehTd
zTYSbS*~>-ltd-->;8^!bT=!N_{b}l^rku7$9Efl|%>o3~^KH8-oC`N{9RHdNk55Ip
zHhx-42TGKXw8nW5vjsg*Wxgk90gJzYH7nCEM7-yZ#fu?YM02(AiyVE3+_7G+UjA-{
z_|j?V+^@{YNVU&C@|)%H5<P#N+Z*{=8|B0nqo3B;NAAu5dRJ9RmSt6hDfDsZ`&(oa
z@OK!_RHaaw_=I!jTbmN;X*rdvdNOa`C3f;2XW9#2VPp5MxwnY}T#QVD7;!nb7xhN*
zP)XEd$Q9iit#`&OGNq5GKW=s2iL(@qiCWv_9BkF?)!P#lpBAuyAx)B?$wq2SN;wDG
z;(Fy*)-cJu>HHUOTiLbx+Jk3!%_77t=o3FjLjOZ6T-(%Hh7Z%49Z0Jd+JYjz_$prq
z(s;tYg6`s4ou~p|oQr<}rY^&QVEKt~K`M<)$8FygZq7enXMeDfwh1)-&ic^mez<=(
zhw1Q)=aqy_-Z{rmY*g4tg-${x|H<D9{9c|`i#Dh+a##&xAw?h(zjb6pUT+<pQ+y|7
z`9g_ZLq}y}5BN01AA`W7pS;dXC(PY%*2aoN+{`{QO{s7-Q)DP5?gg5*lEGRN??UT=
z=g6ddLw<xuEsX7+*KnT!AYQxk?8CN<(fJHN1`-u@su$vFlj5*lb@lWt&*qCiaJVtB
zkC3ct&^NJ-?ojCw!E_o@v@p*qCmTl*aTMg~Jn+sEkuajYok&4?DUiEa%L*`+*Z$qY
zT1BVH?kN@Ocebh6UM2cZ(-A=Y1|l1&*==RktA;qp2H*Ja1=}jAPAy*f2z*VyBoA36
z)jc65e7D4D9HY8BuST0D0((w{4jgTyGe-bf$fO(zM^N+QYMpDLy_PDo`J|6p*UZo1
zam!q$VV`iDP+p<Fp6jO|Th$|`I}Vr=Fw1so>@lN0^9!S!xSsOxA`c|Puk^mUeAb{e
zfWwP}@#^fxR-tsP0L}H%u;fxjQ4|WXZUy_)?qFLIc%GI?A6wJnI98s)3r0ly08NtF
z4q;7c@YNe86Mjd+j6VNB==t`&Af0_RMAi%ZAT3<ZWqHI5WjR-ZJs$|R^iV%}xl*(x
zjTJsGH<6MTy5{fgpU{l<4OQ=O+Y6?^<-F<F8%_Kqj**vLf-x%h294kzC3_tn$LXu8
zDc4BAO}LbEE!z;0?+bD&m%c6eI^!#W01=b>!be9V_q74n5a<H>Y~LF9F_w!_hHNJk
zq;;@wp+fcXd%f2)4&O*uWV~JE;gqL&<RvtzwRo+I>Kpi)Y$;>3m!~jJieVeUp%@Uz
zvk~Ik@B=>O+~Ic%Y;938rhjy7?uIT_1Ha3)U=iOUxD9azb?9d#QlqIAj9>GST=F!#
z=AH>Y96Z96GzBokKDw*Dv(yN>U{^qHGEGq{N)%Rdq&=}k&9a~}z14R|H37MbcP%<0
zb|6z&_eokkM|WE8mg(|OjP{IW3ne;0f5Z6D4#~*ucTH+Se`-V+G44I|zLFBV!KigI
z+@r;%(okAz1_ypKMz(Uq$j(|A*J{YVK1Gh&I45U;_NEXXzK5;awi{hchV<0kgN%OA
z(jyOYXVNRdqmno}!dB=1=TG;pEa9QFxZ+6sZ(!AgNPfSHQQfzkaO97IRx!}G$R3?$
zfGqin`KY&Fb8Y`Gr_;wfDP_&qykK?1{4dmApviJPq0%EG<=cvaV4N|rDoJ77|448I
z;nJFRP3mudu$~vkad^0gTRzr5;nbhczCMr}TkFXxji!2}VHNla5cW1=%JyI)Zr=#N
zT*jIBR)J}im(gpQH;hg5y-VtC)>O>jnFWuE*a6BA=%7j%A%D8V<yT;I?5?mv(8Pp-
zAPk?KZ3eERui`wQw<sv}*jzuuTBesM``L(9r4584;63<ISku8pp?l#{^84pHt#}oD
zQl@60$5JVAj@E##8`H93v#A^ZJ!!y0UYqoG9nONSyX~jNHqj2z$lmh<(UmJ^e}T*7
zlUjrWT-=e=Acx3nQoL`wZT0Kkt3Qe;tx19Mv=6=BEj`yn9CaDtOAF-nzvNzc&XuQ|
zbebCVM%G7x(-#P^QZ{`B*w(eLFxV1g=YnSr%>;WcUyQUrNg}geVpxKsI{s>zHJA{o
zkE05oIlstf2IofxEIrp^c6d0o0wi_fTvUL|cB{Xlp_PbsiT>dr)xfJ!qGtS({Y2iv
z!95oCjA7&cssIBE1s{(?N99VWTzL5rC%P$JWy%=OL`6VXwbiyEtk{D1%npE_<ZGGZ
z8ak&N1&yO<EGbA81mB9-+Ytue>7|GRM0dvw8UNe@-ZoMcfA4q9Hl=-G<wY%)zLr}{
zf?QN1OR7Sb(0sQceeUm}?oxOvJS3?ae6zSoVP@}mfSfd?Y$aj|ReZrHu(P^Xstb?^
za%JZxV5ITS38piPSY51Aj`X)@izLUOXRcXRC36<=!y>xe6RmN2`34Z7;Hv!-GyEnz
zrR11P(j|^d=1pr(Xmg+Ji>k&2_=S-nQ%AmG9v^(tR&KqXL6p4|BSGCQ_v3m^anRhE
ziLb+>124?%mOHY(?Q8<9dg_}^``UFmOJ7{CGqt#YYG$=8*<_vUJbnzmy4>N~V41x|
z9;J#^ggSb84Fw{Os^z>Rk=ht3qpB^6o-Xy@f_SgGY!rrwe2rIxky^9#nCe8;QuLjQ
zGgiJC^3h-t`uBwjMSDcr5*@TSP?zj*0sX1Rxwsa}-N<rSlW}$VXAZ<%YO`KOOWfB-
zp5=wbLmek93&O3CMXcTdGJj=Y{9?4F8M6@W;#)0xo!mo3W@Fzb+^O1{;$S!Uw5Cur
zmF)mVs}uhmy%ui5Kw!P8eV5AeXb}aLUTbPp^z%u@HM%Gigt)%qtR9JSZ3hJl#XoC?
zADUqZ8gxP>i8c3g;}w)DVNYKIn%=ZN#z#62*<GL;y~|hawgr{H66-=aekIi)#!O7z
zp5sN9e?AUC*shZ39EOA2IO;0kOF^MBYuDY8tiHr~<3)EAjNzggA$KriZ&RXIuY_&}
zv2>>`cy@XVjhiNj$N>_@D+bWJIw_BT0HWRO1OGGR#`%9kZj3C9Z2xZ~I2#ko|G>Dh
zGcmIKZ^Z5YkIY)ddW(%h!kv;38xjy4gA3#$$t<DH(JX+_ETXVGMUGNbAi)l$L|KTW
z1hhn103=21LyY%~chB?BzxGeRc{T4x$DMc0S^sS}J6RTekQ`o|U!<Zy5ra?AuK}f?
z0+bOEbilySk3l0MBY6oR(oy6a6xkVzNM|R(B4zo{Nx?-_fKfvRGV0%_WufALue-Q_
z0tNy~G7?HUP%t1NgT|x1cuO@(0mzHMARw*ZJ|9-71M#Vf2)B1Pk?jq`I+On%&^7~}
zfP#_|&NDYog1b-$k)Z-g0qDp&VNL_8p@KL-P#7U1#+`rZep=g*K~5<M2oEpsN1>eE
zw}twaHxI!8a%3a$Gol@YIy?x@yC}wd7#HC$!x8x@2+VbYoxcObCfE_At0;iR03xJt
z!FB`UEd|FQ>Hr!z!DUuJ95c|%nC4SFFvR<n4S=BEzrT_%>Q8kdl1q066fj|KPGQAe
zi4<Cc|Harj1!)qr>$W*<+qP}nwx(^{wr$&<w(ag|+s4;6&g_VN5OMyCeOXaAl~EU&
zk?Vcd6YL@i#7&*`iOB#98Bny>hg%qDXJNy;1RCWM!avOYxx<C3n79gBpc&-9;2hjR
zoXceQX(u^mfC}zcI*`}I3|$xD<|JIeXgA#VTpr>g*55F8Lw8%=3{TlYG5Jy-3Wp5d
z_9g~$b3=Oz6zbwAQ9<){926SyGhl;+DTqo-MNUrqi%=IR1Y=r$Ps0n;-h9};c#i>d
zs(*JC>l$|Lw-n)LB0J<2L1R76(!X4v?l#hQ@6YBf0Ijir5J+N>v0o&F4q5C|!HqVi
z;jMX;eHZN<JR`7i0|l(V&+n(BxqlKSPMpi@zq+jZKsLw`Vp3HBpN5|V*+C#1Ft=#P
zV6UWNqCi4ONKk^h%80i=?rA@_KiMDlsu)q9*7ke;$W_HQ^8rT$&}+>f*<oV_ReRso
zMjiqF8uQ}Dr7=M5KT5q^ny9ct`d~kOPd_|IKja?<THgkfKLEJOb{_6;cITfzKLlnK
zD1R3|7$9nrwwcx{=Y@^UL7#T(#3KZfYeEA^wsXHXtKtHU%0=)^qt!U`3m^qFj}YQq
zRD`^U_BthDr*GopzU{#OgpnYJ&Wm*s`SUq6DE+?mUvChBaPH?<-D-gQuoeDoezEJC
zs>I0pxq1r~83nAUp;5uXgjvl*q@>iZfyh(p5VC6T2@3K_$@&Wi<TuwlD7%Hmfq^aB
zFR%hT%~?qf2pfzmKP>;l--7*6kxuR(aG%x0)7#5vJaK)Yf!njYm%GjJf*wS_hZmDA
z*0vcxvKmg3sXfURdRqsqVPt9(p~$z9iZJUcaSeFng37Ck>7iF{?=Jz1E&}KR*qM*U
z2itsD_soEC*gOTbsnI=3_j<~!yB-hs_u3(JN0d0T7Cdx?Xx+m1tco#VL)$QOH=cEi
zdt%*g6^~-0uj6?2HaL|D+kPlY&LR^T)4}0M^Hw{9NOr?R&B4d)43$x(uB_57oPQMH
z+mQ?#%e`hE*ZFYCdxVz0*x#;>p}Cw#?v+rZxLUd96KMucJO9H*3y|z(KTl@1<9`@7
zR7&VoqM4KAa4FuqNK*G!y>jqyYTUsm7uhp-)B0y4YJij*7}WKq9_z01Nv2mcy&-I|
zn1@hpKvF--txnNy>%qd2rsxC!MK2we;yswdF^uYbp(nJHVidH2QFu@G#@UmKIw?#>
zPBGDOoMK@7D&y<K0#<!Bcv>GX8D~YR=t1d(O>MTx%e|G^yZ!eNzELB`Cp5~^*$jxu
zqxa<dE)|-5BJ+gRT^?1la}BZCuO|=qI+J;3gmK((s%EixCzs5Qgq?KyGyzi|fQ*J%
z@zJ7}^86>6x$=^53|i-~GRz49L-PJ&GqFm)`nbKmsS<<u%PsQjt=%amEoZ2K0s|!$
zArD(N%%^{uZoiMqoWB1OYQi)li3Z`u?WApVRg0dY)w|c$?fc;s2t5U`TqQnFAUczC
zEPH86mvdN_oX~IDf#SIJQ+%_zJrVzYQF5uAM)d`m9oDuF7s^&<*?>pP*On?4{LTZ9
z!XiMZQkLq)hTfg#oN^&9zFf(}avTTsp2Iv#Ggpy43_zMm8md!Lb+p;f#2-2FQ^?e~
zJZX1n=#s`X^y2GCq>_(}ytVF%6DVc9^u#*|3xh=}ptkEv%$-09>2lVGsaltrpkmmg
zVsB;O$5YmUcE>v;gC6o1@I&VypL`$cuH817{5T(o`IHU17_G~>^^j}Lvq)*MON+hS
zhG-V|dUqURD1IB!a{3$|Gjf0FwRQOQEOd4q_u54-c|qU3AWHnPwD)*54;*I%5*ub!
z$KrdFU!~hsf8}zSTTJWh!V@_5%Qc@F^laM$zb_r%L$7sb?f6dD<UyCayY;1YWKp=u
zoXFtWb91{gtlqMPSH19yDwlj2AZ91M#}xGa7x7Lx3}MH0^Vj$HiSg$Q&;p6<Fp(T6
zo{*M?{Mz{m2VD0S!>j^|lzNN^=BHJ(lDWiWgtZlLErLRLu;w=CTX1Nu<3BH*D|32Y
z`6aKOl;H`p%K}I+C^awss%YBHY1QrB%AI}8rzT>QMM=dLE^A8TbF$(MtG_t^-w#Bi
zWD)`7%8$>c)D8vV>iMjY-Ysl(!VI_r3mUd2C_mXFa~;za$u0oN<U3t!Q)!~`np0fJ
zxQ4eq17KT0W_JwKWwO(%1`W){zDprp_<`Rn0G{Vk>XNxKE*1#g+E9v+L*0BpyHE+d
zlt=1`m#InnCvmuZ+(aC`>rVvzF5M(cflrQkvFQQKQHCZ|$-AFd0d9Rx8jL>`kb+io
z^x&?>R8V)f{2Q^alEa72A_zw%CS?0vM!WkEQC^zIvT3|3{Y%f<lNU)8_zK$=U|PKP
zFJH(|Qezq62I1qiT4{_qg&P+(Nock`Y3G#Zmut_}5j^!X8t^vj1@WS!toz5J7dw{u
zxVf7^{aMbLo3Dv&yyEbSk+!{&$WXPCb1jXTYZY&~)xCWLyfM7d%elITaWeJIAHto5
zr6l2T^SBImqNbKCmyi~6J0y!yU{3m&osSmbKZj~1nxB5<WBsz(E~Nu@tyJ|7wX4Vo
zSWw-!Tg@>BuZ2pB`Cp}Z&F?tUH2D2}dK;G2&F<n40`$nra!%D1cr=>}7-&%02L6Lv
zR$XyOjr%(71Zr|7%X+a_#=o*!{bC??tJu}5Rnf7bIySuxQU`aWx;I&QpX_iepBH1C
zysL&6l2Jzg*xHVr{4sp~xG-qf@jaOo57^?&LWfe-keDcc34Wd#(eO-=r532rMh5ZF
zEQo4~8j@Je5)eIuI73j`c9>1^V;ne=Vjt9`(c&}imnrWU7_C^~OA=m+WZ8^S`eCxY
zip<GhZ6`r`Ql|zuWIS#0vy0DaC#g0M21tsvPB#L){$6e|Mb}WErUSrS92s!J(qS|`
z*398VJ4Ve-Ec2Mi7^sEvZ$BvF`qUU<YoO_av?~ua7(?MU_{J*b&<|Su5GTl3XOejd
z;P592WJ1FWAp}X$Ob^)jH9S1h9zlg#hNHvHs5=+1*iFl@z5-Vj1daH~#g+%Zzk=wf
zlh);LEpotCJf7(=#*{}W9oT~n^f#JFyS5*5<FWa69#xQJ35n>G=TRMFn?-uDVfJ6X
zCBD{cL(<&r9V;Ot2o`4ucG0G~P}y7SlqOg;lH?dudGV8lR|HUz@SDu4FNS{$*ev^D
z*VRgM>z)I6D7jTABlH*ME=+E<H@S=O&geIA;pCGT{u&l%3^8ox9hKt+esFGjcwlBO
zdK^(tG)>$$$<|Haq`JQHU9P}<xr83J>o=iN0XN2Ez1j5AgQ|-2SIV5%jIZp@stde|
zH>0(9*H;Vg2gI6PWIV-r-6K)$pGjfQCU<~)7B90OyFo)wDt`@~TmModV&U6y{csgd
ztGC64T*4nZ9o^b~cq4v+xk+=S(x#9qoG&L`0TIh3H9RV|nS^*>ap_O9Y5cf!K48)E
zqV|Y>Qcb9}>v!<xJ`G%+`{?NX+P4Btzh3@q4~GWiV4b@62s3S6$SSoG;PHH{NU}Ua
z1}m@0Gpw7CMLGRDEH)Q8B#nBDd3cZZ9TJ|O!+zR)k|jzt=AX_HXAa{BucgBgcL^>e
zPqQKU5G2)tx-RFNpNEffErmqR4uefy_;a+ts>k0$@yN^!AK5{NKn;Ab2<S4~&Is%Z
z{E;ol31O6(nU+5#**w~5O2pSdSXAg6$V5M0g>AIY(rPa)#52^qD{-{F`-12x`Sk%@
zy^Ozz^nQ(C&wULRyiI005tlC9|CNt1#!vcsdU)!KzyqWoHEZ0d`Wn$CM*x$l8WU3%
z7eijfo^yDtld4dp(@b@X(1NG)w2m`}YI}BJ%pgFgB^H0foJji8B>iku_oSJ;=&t-8
zCAPu|?OJn6n%Nd)Z_!6k(U4hUDf!%RFSsM%3J=;!QM-wqPJkzEjjLC`n<OlQjZgy6
zZ_){On}q!1%U?=lthM<oSTlt7YpNDXUp4(iJ+XoQkHJhLK*i%pjHb__Hc^6}sRrAS
z&!_zTyW2n3?X*jSPIaHYJf#P3|KIL5bf<LcexmlybJ*hA3j2(Mfq~2U=x}K~tjF^!
zH?NQDkhxsvj1>ovEqTKO$=B_8{XBxWx@}${b8Ib~ibT_X$Xoo*esR*G_oQI#c$uUR
zp`$0z!}I_<t1QXYd3=3D`o7k_GO*I&At*|EIOVxZF+$k=g%(vvhX0dnq@uT1D|H+u
zoN72$9>YPP3R;l`8K0gFRM>Kb9(|Sr@k9*`D<+x^^BUx+g+=r_<RnlIa<1-cW@nxh
zmU1$EBb&eHLiekRf>ZP^if-zcRYV}aikoPAB*|e1)icUcrlbHt;o8EN8X%_c)htfC
zRBF3vX6g5=U4(xg#(Yy2l{s{l&jz~W8CunlTdM46BVs#5cP;OtR{q#JGWK$sLH=)*
z>@!Qk=D$E3e)p662Z<+iR2KQM$02Mc%{!!oD~4S2EC+dNn528ecU#k}x2=85#tSB`
zk<eJjq_0JCwL4tyr_4)7B6W~jvxPEUG&1pdx%~cra4Crs8i%4X)-#WOgRr0(#ID3J
z7CB)1h@*n3kGhj_7FqMm?4Ol45oxg;Pa#1?q=)eD3U9vU^)^*9dF=j{UsK>2k4duZ
zTIFSPiQTXEI@>PP$j4R!nI~yO=zbPTMgH|+WA_wt!^$kX@5?FG8K1}DtE9}#kZ5_Y
z{$7aY1rg+oIlYiJb~VrMC{DD}O_3?<8mP&?bh}pD(zlI~Qtrp_nYX)%XMl4lkp(fw
zuBdh?Y7V}<Lh6GqNzN_LG4G%w?v6H9>!A&L0=RR0?vS-=sAM9+&(YwCjK~!ej%7=j
z9T!z4AuKFKtwS`UX9_`Htty1X5g6%ya??0xFz;d6#KF=klFleYn{dV+YYKqJdO5%Y
zgNDW5$wqAQLbq@nj&Ij?`GTrUG!o((WY83BBx45US6w4sAD%+fSlpn0(@STxmenpL
zH6>}E5LJ!xXC7cVGd0*nL2rgfv);M(O=3G1;_7z#t<WLIQve~Vn}$N6=;Wgb?NxC~
z-jf(OT+KvTlm*$OO4RHDU289<=H2qoB#s*f*A1ZAdG2c85ET>UtK4%J3+L0jYhaXg
zGO?ABhZu!+^gyj_$|8JMkx2g95g&0)g{jR>t+YF<H5adytlC$-CK4MCf}U|Eu}9E%
zuO`>te4D{QsO>}mw!?DL9__%LAFzh=Ls1V+0!z<kaTIu)P6{oKg02+B-<cFV*varI
z_jse^YxE=r7(O31dcY6by+bS$q!T0AhdER-ALzpVq0V&}t}}M<jg+uN4={P2mWRFw
zLD?J{ml(VeiZlu2Xn|kbf{$3d+Dh{VN#Z5F878%Z52%mOx|6{A$PAH(L^=>M|LStt
z&d%-DBHu4<y$>G2l?h{#2|j$19lBrI-ZS^x7TGjnRh}}?U|<!XMA@UAQ50)SpdQ0q
z;?V%7(SY5E)RHf_XU61CaDUDE^BBi>W8bh$^QZT=ui10IsYDK+%^R!t**SpfM+>MK
zbG?jb-l8UVn&hvph*~w=^_4W?I-m#Xy!E%j@4HNHfO`Db5@NKLOtwmYH9{S2aQ&aY
zX9vPZP1h(xRZ(-e(#9DE1o}EW*ZK)>AupZ^O-qA>_Bzuyk2Xzfc&+TYP%l04nn#F=
z*SL!mWuz6<=mgVL`SxN=*;xey{<U?haqaZ`I5{!VtmXhqzgdxWY1dw_13Cq_o=dPI
zXMU;J=vCnq9?=><;hhzeoXf|oL=~Dj{RIN!z&C60C97p&xfQ04*7K;{X!WJ=S=X-m
zIF$Q_%vmMyIeFB*&Q_$b%hr}pzn8@~v$Divwn=NwX-E79lPvKJw3~Ht>+VebeKBkq
zK}iwyV)|R{)5!u2udL&%jS&eh@C<9Mist7~_baPXHJ#bB%{yIziPQwrHGSPCN4m9k
ze={MQ3GcA2cSG-jMdg>~zw5!?f|s(ZQSgt-T`LdCOdRs1NAKQSqj4PEWT)>mvet)H
zw=%20LQ%Y;RJ1w%Uc1Kx$HsGaOWZ`DL;b~*qm~`^nJ>VO!q;|X+u=UvxT#e{g@a@D
zk(IH_3EiT{%vrTsfypZJysLn2i}sj~dW|l&ZI3RnGECnaZ>xvXR|9UX?8UmP(&IOH
z{uK!@KT#dG{QIEjf(BV?L92Begr!@z9z5(Nd_A&1?X?>qRw1i*djAj~p@Q$8cf~nw
zMhat!_SU)uAX;yGp9`!Ci>=1pJa6qUm-}viGz#VS@o1L4Fxx9y&~o~72IB;=J%Y^T
z<=gyQ8vFVAA#P;jaWN7Gf#1GaG`B7aHNG$!&)Im*H<Z2Og+?E-DZD(GVYjEZ5vdp#
z2XW+v=q1v~qb{y=v89Tzfw`q5ovgIljV&3nQIOr5+Z6km$)Yd*c>lV06s8-By1)B=
z%`GhD(S;t9{zs+4GnZZ<UAFyN8WQ4|v_JS!4!uW`$))rAk0pC^p}jhD9C1&IK*5B4
zvw?)qQEHXBORsu&b@eRhPIp)~fI6<(WoksF?xn}$g5yR|U+An6{aViHHw{%KOT%{o
zZAX`lj4>YSyo)<X2xTIhd>lrhRX6S%yVgx&kNVm@v6EtM#u=N;&m+ohLr-ZAiPQuW
zZEkvin0ux>wM~S!H<Rk2mGq*FSbzX=JV_@gNa4R-Ao_u86;83pt__mRKw&+LR|c-F
z&j`O(S$|$Re;_5(YTr)?-|%sLV2MIW*!kLpUOIZj4tTrL*}bhp4}EM>$=z*>yREE&
z>4tPHx$@~m4nh0V&Q#;WggLoX?&k0MpdEqhXldX*k3!O}1((+<*XCGaci+s0x+cNR
z)0=+CeB*AM&LRbT3WyB^-AaXEfl@;Q@albE1>Dd+++m-Rkcyi$r^;NZ<ZX}Qw(Jqw
z1WkzO98P1`-}CxUIB|8(4O!)f%p9->zF<cF-cP!VX9BA6>Gco3AQBK}*UX`?h-+^x
zNy-lD7{~aL-L17<YihyzHHqhlGjCmg^oNW6#;~Gklg)8Z73B4W6*B_o{aTdABAL-P
zbDIrQ3KJPiPwQUw*bd*OtQlM+Dg9NG3U7BUm>_}rl^Xo#>Wp4ybgef;vR60;VnL|<
z)6$#UjXn8j`=hG{XBAXuwqf6_C@QOWxv_5qv-aU{uN~l|){91BEMxmruGjA=NwM+n
zQ^O8le0#B+?fx&MeLB+EkVMA_g{qIS$gkUsC&<nX@wxU0j11j_{;<@mk$Eh#rjLh<
z-%XI-wg!0`=y02*aaWve3Ybw8Zum~VpGw^V#l!N|r0TPi--7DM1dT)3!3NDsE*-iw
zI?MuH+H|W^EZz^a!&tq}aj*lmrli*G7`=apmEh}suEaXF8`Pk>oW;8BhG^_F_p6Xl
zdR^2RRTgpvxrbex>u34$|CWu$OHSn6dTT4Qw}lR&a<t{A|KK0+cN~&g=D+5#0w^nf
zzV!};D6F@8|K^QU=A>ST6E3(pPpj_lD9l?s+`nAEW@bcnaFq6hM<gZp7wV{4yn6?^
zz_=UhMl8*(c-3j1b<wBICV2$W$4EG?eXj82T>#RVrqgT$kwRijk1gF{HdnDr$Zh2(
zPeOV#{m?$#o@&L)7uc{qNU7s7;p!#M>sqQw2RAM<o!F}k4(4o%td6@y7u8VJvuXc$
z)Y(%DzE()a*5(qHc|oPUwc+?^fQskM!kZ|CAzvQk*yi>ev1{MV=Xo-Z&(tUp9xyNA
z8>toTJEcqVZ!7Q$1RZyvI%DH$8O@pCyHpcydL$b>9|Y|4enw9ocMaQT?ur?KR3%a%
z-XPpj!Ha}>b{)zmzcGGw2K_S$Pr|7Y0QM;fw8i5PfO(zWoskXfk}}8k^zj%+qs#Pw
zW5|p$l3i4LnCt(prAkDL;bi8*!&y$X#e0!mt>reo8nlYIE7`;E7GpQI^hQ2p@s#>p
zJ+qa^Q$&08SdZXoEMA*A)A2E*R@nLIl_PlVlaSt$ZK0P}g~*Tsq)b7Kg|#9NyMVLm
zRfz8$4$M`I9fl(HzkAYgR6w{8js$k=D)S`YwEgQZok{njP3v=c?x1<W<Th(?7?LZx
zw^p-|xwUCuH@7fov_-6Lp?Y1#^+wiy9gbS9-VUD!*rnr5q3>9$Z57)xGt;d)50g|(
z6op4zM&5KUcu5>~hSR@O@0gY@ylZd!887@Md4<GiFS)xWgGVN=?f3(?NZOF0sZ}aD
zWLmx9&g?)FUcDLK$~2!8Z&=3m)$vwxYuyB2JP7-mUuMIAXXQJavl%tx-;;)&aX_#-
zO7_C|etc9CG!1+`@Ubzj<K4;DZSZSC57i^Dfe4zNgYxL}wdMXmANtxB)u(%8^1Dpt
z{b&I{yJqn^<ecQ0-%+y4CDF75Ik4;v3XN!Dp?VRni&F@ND!54{W!=4rz@SHaI!L~p
z`H+dbt>vD>cgLL!noObmb}8TPNvpPUf~!^z<<RwN!$&RIQWY#Zt*o=kb<N|WMK_SL
z3|(d|s<LQC3boYS$CY5DgfuC0bdqLaz!+S-;1bSA#ttPpP6$$=Lr1;a8C_TRVTw8S
zz0n;GCtC;HEhc1&BvF?;J5nm%2Q|8@XFL7Q*yX97hVqXRV>Tiev5y8(2ial`2Ty(A
z8QLxrX9LEK9I2L7rve*g!ErvkpGI({Z6m3kUn|SPJ~^_5*ENOZw40PWmB6<q)P4UQ
zt84Pp!L@XyzKw;n(~)L#MGn7bHs`w-QlIuyE#)(+kQ3JvgU`Q11HnZQ^hMLXc8l{R
zpAU!#KKQP5Ay31rtd(+1J8t}IBhKz(@%c=6`0xFg){FL9TOBKzcfH;C&IIrN^5)c9
zT=|-lf`nJ~+REh6m&8CZrF6h-oy(`FlCqjYF4#k~Av1A=KJP1eS;yG`(ABMqEA%Yg
z(7m%`;96VR%WcY;Ll>Y%@wEOaf586GlG_xYuz!xdSc)DGL)?viv3~JaYy^95Dyx;B
zh5N2&uMP{kTZc8jG#$(NQnfnj=W(Ej&hvb3-Z~cwAP5)j<TO<KS>OSn?cFm{i2NMW
zNx?-*HlE!xVS)Mx_cCOg#R?ilQ80Zh2B{NtY-#pAH5TS0B^JbMvhxEvNpX?)p8zEj
zBh!BXlq_7F{|8XAadEQz_rw1q;biAx;rbr}N(faI>&1UXbJeU!3G?2P0rN_UJ3Nv~
zkfeekq=Lc_Lg#8;3?RhP>BK$MJ&=Eu{wWszDwTXH&ArX~_j<2(x?i_G?*OJ*SO5>R
zEZ(;G@V+ESVwfA~by4DmV*P>%w1QG&<GRQ|Ad#R#|CFKl_{Lx>Sl{WHF=MDVVL}B;
zO+PRNN58_29NA4N(Zd&oih;6jW`X|oGSO1Oq$Pm|0vQn|Oz#S|WYY>_T?7UIy95Bq
zu|mfY9V!WTd2|us;x@K3`}cvoUV{Ps#=@f7-`xiiBi({Cgb)LEDbSSQMAle<zkp>F
z-Xz-7{4YkC?mld!XCfxX$Hy0_Nct<n&AFY75cXE|Hy&`91MJoeC_LyV@jrzp@xFfs
z@Pr4@XdL7;K^M{n{w0)aFsQ>X^AJ4T&;hab;@=J&Ky{q`k|H<-ryvqvz}7#&kf7h(
zI6#<?-wK_5MgC;M1%41B*H`nB7?hJZflfecBsu#)d}_9YD=;*$0Y$(20u;?*u`v*y
z0*G@8|8VvLaz%ivqJ9BU#_btAGP>D8V$lAI5aRqPqI{=;yGc<~WF%unhLSL3GWo|S
zV;q7v_v5y89u>@U6)hA}00uDLV9;hi(O}ctVoPA54zA(Wlt17Dnd9G$T>Xtekdorc
z%EHh<?%@4BSGZ#Sr5w+0fj@3R{+W)jpT2tY^g!OYQ0)SP_;KC|Z$<_94nQk2csGEL
zKD;0O#J~X{%t3gVhEN_tMUx+svCREv|CRQWE~0LrG{T(cexUjKO9A^A$NQi{rP-aI
ze*YJZCEbs@ytv4Dg3+J5?Vjjokk=?lAW;+25=SSrWMD}t8IaGP++%^ypOb&k5JBv-
zm1AB2icRs^B4C;Ty8VAc${%YZFrc5hg6L5xEcm@2i8pmZ$Z>dg&>#QHANTPex#zy7
z&%TtO-G8I0?rz`LCqLIe0$YltgX<p*K=Knenp>)81&+;u@Aef~pL3*_{f3b4NB-Vy
z4T@VXlts7>&=|lnKtjm=SCBE(gD~)m62Gr;f6mST>U&n4B#I8ctiSFqgM*_%0>7ui
zakzOXwj{uVt=H@^a5+?d>Z*g9M27s}@-an`fn1$~y^;eRuu*~qfqH*$r3~)r_Z`C_
zf)~we>OumIEfD;UcO&^&R(Es^prnVsek<4FBv|je-|Xn>*bf7=!8(lm=>O!oaBjia
zNwBj-{z-=M(?r|)>`_h}Q8Dq1urRD-BcAQ1!3{dj^*Xt%kjp829(byQtGRj?tonYe
zo?nm>Kp)5j4C$-mX~mzaEIBuNQ0p1o+(aEceQ};S1nh<s+H-cNOUJ$nf6(m&#BPix
z%_lJ|*xb{KR;l>3`W(M`(wHp^o1p}Nb#s>OOBxQ0Ql7Nf9Y|z4J1`i2%+y*K)pux>
zb>LWMbPJ5d3lHaQ*~XcD?|+Rjv-kYtk{s1tG-^Hq2A8&>CR&tuEIsk$60?HB2rME#
zX=Zr8;$P^GG|$zXa4DRLm{Z*IfV9})%}kHue8*sgL83GX;3<g*>q*-y1ytHQo*Hxu
zwfOxIx!&&QDLj3BNjE(M`46sbydpxkJqybLzTA9MBA=!}#mXPHGD$5;<j!ElR^Y2U
z?)gnX>BXqs(+l)jQ*v3qEST(Z-A;Jp&_jXw6a86Lc?u+s4tJ_>)R%=st5RDpEP7bI
zY`i?jS2-`GP5P#@&)<~iVbpp8TRk-RJ?WbLUS^zqr0OJGsAj^*=F}5*m(!9qv;sX9
z@g|QdI>r0&t35&d{<^B%ZiSHCE-V*;9-kRL8RtePKv5mkY@#EEYD}nWb6a+%m8vl?
zCNW^F+)OBk$G3Ecgz)10)lBO9D4SqUjaLq3lC7_0s%_Ou=+Tn3l!-IBW{=O;(Mfoj
zDY|xn6Z$h3vZqwMmeFKVPG2Q!v%EM%N#Y(=tRei$>`8JIJ&}YxPLZbSNh~kgPhmE!
z#2We0G?WiPt<R*PbZz3eMsT1fvy-%NZi+k_G5}xSmMSQx*>Sq;+1i#?vq^DWAO0dV
z;`hsKVS>pP23HCu{$X5LAAow{SGp#ul_mdoWji+FOM2dkM>0q#r)N}3t39REHneup
z(|8w-YpsK+rLXTV?E_0}`4#WZt=3HdxKQtBQEi?=fT(9_<!@2<%i9@cHuijWHu>xt
zvUw#lmeY(6^`t2h?7%Uf)$wsujJ+CfzAkQRjlV;~=eJN-2WE#Ys_3YeTdL<eW7QEA
zlRFhTrF20Y2wCW_R3%gLmR{!!Cg>32*)-AQZ~Z1Uq4{<}Vm$zQf9S_nw~)}BZGSzp
zD<`tQJhdDH>~GT!QVB2>X%+GhsqDBV8I6(Q3ex75*OFuYp_M<i`&uE<40vPop^@ap
z_gP#M_s5WZojM3GG7lLmm3?r0NBWsIUTyP3x9Dz;6HOW#=h)AvUOD2yJkyxl)oo-q
z`#gR+u6VAaRb7d+Y^wE*mL2@6<&Phh(LQ%19FMS(-N*TDo5^zY=^5x7S^_ljOs`vy
zWzc0gxj54c4os6`LfV;Fm21kA*%T-)H9B0)NRz-5(sB)*)$|qkozOWb0;RE6M}_@e
zv~f-z55o2|7Q{OGQ!BquDlN?x@jt=_64T+&q()aa?B^`ZoLVYr{>9<o<KO>;fu3@X
z<A}}ZXwV$5Zi5*uKm3kjuo^!erhs9WA(o?J*eSDUAQH(R+d01daT7+YNpR@=i*#&A
zm#7pvOj|3NnnGV|fHz={zX<gQt?gF|h7qPjCcn7i{8o*SW+20pA*~%bpB%y#&p@+X
z*DiL*>JZ5>N7*j_1gXUQ)?@F>BJ2m@adWRU?jw{ra2TCRF5yO;)ACKL3X9tw#Ya`|
z^vaH)GzTo^zbl~Pz~h?^g}SG&m3QT%3wKMM(|8T7RfH)4G3g6c8>R=V<ILXyOh+J%
z^^aucs(HwAC7mY%*TDt*A=4LhVufRR8C16c2vev>tglR}FsdBEe93-oBW@}kqj7r^
z70mwVQdu$u4!0|d6M|*#L)ZCM8H4x3WC}Gq;{w{ZvHSoMlz*Mn5pFPc;ID9Cjc!k5
zl?(oVdu27L(^YqFxt+aG5e=A-M84Qm2?MRA8>T3s2x>mB_ctX3vPaFS=vaDg6q{Au
z5+cz^4KT2$kK5R)Blcsm_9qFq$IE{Xd}z{<X@%xFp0+S5hd=gX51>iW_r<Ak2#EH&
zoZoG0_sT^;2eXtUDIUiIj(UyqvBw&SFNw4sVrK`g5O!0le>n+A$-EkzcDL&Iyf7*I
z!9DxF3=&x2cj_JK0)a`yT2lpz^wQW(0}P<5%xARIC{aVIcVJ{Y*(?cgD2et+<??vl
z(tPAN(l*5}PP3O<AeN*;T{fQ6Jypt^q?9VzM`p!V$9MUd0;qIP&n#yf-5kGC(9c~1
z{kIg;#kh?}kXyyohNwj@Kax}cpu-wv=0cs1Jh=|NrwPyu&3CC7XQ}jXu@I4>GM_QD
z6qners0QqZk}s}!wqRu#0Nu5BaK&HxN-_;@6qZ{8_fu-Dd!L_cEU}P!^>VQ2mepN(
zr4u+09kJf8I5a^!x09Oz-OFmO*N*Xb?}rVCdGw1qwCc4ImxjNEX%G#uN-I7g?4Rfa
zCPIsNde7HS0Y%5y8Gq`1p4Cc^&<ad}0^h|_h_(<qt2P!Xz{bGXUOFyoOPgO{I^^x5
zniJUxql7Iwu`$=vkYke7uE@uyLN0}E>s9|`c$FV@amGTLK?{{iKj~EcsiS0d%8c97
zqu_#8G1)W3%xaswE-;QoG0r4-<r?xLM4;{l1fyH+@N#9>SXV;e`D;1@x>8!ZDX01-
zwGD9U8!H~3RA1<YJV!1?JKOA)PnKZ0KUUrEgQ~LzXDABITKDVxw3F>i;#Sq9q`78=
z5a@+_n<iQj?wsDz_2J=@El{-{`O0_8&m#;cOK~l_@U!ME%h_xT{wh1oiw$b8ph{9U
zE6x+Bw+TwjjNmvQNEx~^?j6RbEo`QLuiFoVDfOiOLtWy2V*nL-?_sp!b?<yqc7{=v
zx#@RJ`!`>L3^xL)52Mb$b|G5ZFnXKZ1a=w6D*T8rHi;=Ak7%3jGGzFL>3Lo%7*aJc
z`IW*1KXR?VUc`+c%P}NKXqkgDbQX&3rDDa_21zuHRvFuJV)#$z=3w@}-BOd};zK#!
z?GYbU!+cDjBe$NoGXZs*lo3pR<HtEaHYKBL<TaG!-(23;eoz8m3F}|J|3?4vwzOql
zD5i485~7w`wUsxe_2Yv2Z-?s@lhnW)x}MSU(Kf(l|JkKmK$z@<;)Jtt!SM@L7sD0+
zv*q0I+d{3=R?XkWEJp6_&hYjS0*QYNc6VoWNek=w9bjB6it;10&i+k3GS=@$v{1ow
z5+zBHb95f=hccafQTikZQL-3_K^|Dqcu0QRsmFKu0;L`%m8(-U8|camVBIz?F-U!y
zM+zYM7jg%hx}AvxbvSMM6l_k%r*7bQTs50*gP~_xA+mQJ>3O?%aAVLfY=z0cPaLPi
zJ1ue8h=|~Uvo8WaZ=%AnKTiBQ=YJDh5P(p+ICH<RUx8dOm^!^d%L>hzM!eF)Zqb;U
z<7!~hq?JGagCBCfgtX$^px{6$!%wAk$OU-rWqM)s-+uDbVErRZlMde*E>Nv7wk|r=
zZie4df@(MYL;WflcVg*hQr%}ReroC8bh<R(b1OtsNa{hR#N=+<+LWYaDLA4gUn+e#
z#L&~F=^07V`-keF#Lpaa=Q@zD>AXm?jLnoO&Wq@gnl1FRHe>Xzr@OiR`fksfzhzBg
za-*_2%s8cGNxAT-E(c5xx!S{szOf8JPVh)kAS#*c=97z_cYg%hVxig2T}^wtxt2M|
zH9|qdkK|fnXi#ujmx-OeaRAG-X{8#0z=Mj$pL6l`8^7|>yOUl4RcmzF*qTcgwv&)x
zFOpY(H6x<3bIDjw@*;7{->YZUD9D%n=`xbhsdqNh!-jx!U_cRh07pJl0;_MD)zT3&
zU!TN?RJ}_3MveWobG^3-9wo~GE8YqSNdEwZ{q!jZ9aC+E^dyBY3O`L3=njO`T<6g#
z(8G96_O@7}LKhS@@56Oq<#ee%=Ri8})RtXceYv<WFf-T#9|s1Gul4Yx&Ya?#a8)j$
z^(A(dAxl1iVy!J*9Eo_>$`}I{zuQUzfTn(e_H3GrIp9qx>l=v$v{Z~Q%nzwR^*dIq
z!zM_PwMRw8O`V~HiqLSsw@p@<eX;m>)9g1_>%Z?WL0PZ}QYrT~Q+FXKD6818g3_uS
zF$BH0ZNbh3L?BjGUTOWJ#iC*}XLT-|^+S_>OZKKbnZUst5N?*8mX?q|_Z~dma9^<V
zXdHtc_K1BB-;lzJPE=nZzDAfz!`So+d&}2y+fF~cgeC4v1JU(g|2Ql#yS_G>E=_~%
z+h#F={Bs++VbV1dG#D2il>0m<e*LolFLZKFz)w^^#%AoN2n?rl>{Bl$U&hEt%g)5o
zm&%bzm@EawE3k_c@mIC>rJO}p#tnNe$^w{smb)5**DeW@d*0PlWO0Gx)xz&8TxWCg
z#`V2wwxG#CSv`X$*cYUlwnyeT%o?(OjU0|F<~S{vf=vRu2~o~TGAUe46gg@sEoT{@
zKK)$)majQP)WJu-#KKR%;X(HETKtd3<3C;0o9FopuLK;1qzeMJ4$UyT2KC=fd^qmU
zd4N>jVLI<br9z@DW`>fP*_8pFYE35h+4<yD&H`P|Wl<jVA@IY$Gli_A|NQJuuC>7f
z9OcCi?ca&5$uFYgK|qXvt-OG&HVdA9QqfOQ?t5{q;+n#mkN<0r%rr&mrMB#TJ%6>A
z4syDs<UNz$uB7&pl_4Sao&Qg*`TSwvYljS^RQ5m*jI%GB>dh9pVXxa}|K93`xnvVy
zVq{{}q~!zq5Ht=%^^MmQfu9FKJ6TUha^huGo1-NBMcVIDmY#}#?$hCJ*<64tNA-Sq
z>9}=DoNib62za&Y7-{^>$D$Q2e{XB4GDY(NRL&?;_vWj*-}`~d**@1?Qi8Sah|j3J
zS5%%Gt&|3F&e+q=U@W{lr^J&{rc#yTm^P{r`5=$k=y5XBI-eBXM79LDj@Gv+UQTLs
z?@2QUlP7u_m+#W|&mj7v2qGBE`uZb0&1^~CMn%N?3C7oP7y<)xT{Yb*`Eg7Yonp0T
zW;X+an@Na%@u#!X6a*Q<x8W{o;PHn-NZT6<xAW=&qIreQx*t9X^q=k~iDacegva-^
z48i50ko@_=7j5g-xU3iwp#?Lnc+sE7*v?(ElX!aLOF~o~xH1-0mx5y#nRkC{IFdC|
zh|3{UF$$X;674(2*1Ss(;=D(xV1@a8X?O+9UdUlw>ZLYFk;)%PKMSl@fVpFNWSprp
z++$j+2-3yfw(Cldh~SB0A3H>>d~KE<mh~}O+-+2DBSp>XeD14Ff`eoF5~b6opMFIH
zne>in0FW<|3yXVfqNyaoADKk-fhJ9E!uMw){t9#X1s|V&Tx9%{bio$8%mJ0|iHFPO
z+1vu#2bv_Qq>B}Zt&bIsMt#m6)fQlkAW#hqJDAw7@=E<wL+psLCH`%_kXKMbNLub=
zWRu@oq?zx2e{CIqTwZ$X;&AKHR8aG*a<u_AMN+Q>yY?WLg5c21GFaO=0W^zPv?&S8
zypYB{1AtULbS1ZOpST5SMwbB22ZNCd{5#0;vGQ=<ES;`>`JfmvbGg&pjniRnQpUW_
zLC7t0)}EW1<i(^#3V}5|Gv`%kqg9>`-)Z*RH{{7ka53&BcZ3oGB_=^DrW{4;$5SqH
z%GBMUL9Q&x>J)O_fTSmu9N*Io!c-~W+l9<8@<eMn6l2UW*}&<5Y(Rpl%$#;jtN3U#
z3{bPhCxT`q>O{-JxS%=ES_`txcrfmm0%{<8$Df4udkE~()Um_n4$!`k!$p3!-n8%b
z$z!pB6|$s-s|><I9)$`KiCLhp5~dwirnB$P58QgK?9ayo9;E1Ku*=AVy8*R!1|v};
zC1)F<+mp2}+KKuZV)ft8hYSPpcq@50i-%M4Y}CmmlgBUG3rBf>7JH0K+&l;Z>*i=O
z^6-MSL->i-u$2bDwiAap04k>cZtnd?hx{ziY0meMxjcYL%<tlT@$pyiw~MRw2N_{2
z{5?MDOPqUpn?f-SHi))K;CvhD`sJ1KD-NjF?p|Hn?!%ct91lFX^R9l;M_ua(O5_Ko
zwsAn%BXOIXDH(Dr_oJuhKoIW4hj9a=nxn7?(sd38Prq$DQ1L%lZnt!MR1E;xb0aR6
zm*DycZ8!-U70icO4BD*0RCzl`xOZM8Gh7yW3P_F|Bua2q1)NNR9CE*3xhqihuBnJE
zK(BIGbhk8ND*IK)36n)<#5{DN0h%$~y2bxAji6U(hsMfi)0w+-%KctrziQU=gQR&b
z%Ol_tAJzQvnOtZY#YjYx6rhJ><nI~91@kzf+w<Srgw<CwQpmMBkk(4yE<HtgsE9>1
zoNS?Z7gv3%S{pfc``guS7&c+KVVK{1`3TSu{)66SeGksklUlGs)3V>jZ%68Cm+6Eb
zdxi&t6nj~C&SwY2%sbq}o8S9D3WyKlp|LKr;1Z@^(w_O$s<bjr4f(3=71-50nOV(`
z^Hq?m^*Y-tuE$dM{A_VyFbjM>3{B~>^ml}xiT<Z|uqEF`q8F1ScjIinOYGoQxVr4x
zAaKz)7)n4?Ry%WoHF|HAw*Q@n0@Q0kfekD<K}+f^T=Q1es(ph<kU~CAZ$n+ymgsNw
zx3UkuQ6U;(6sK$gN&f`4rpg9AedH@CVZ!1>XkJ-FSEwd%)C@%>a7ZgkCTgZbjbRV^
zo98jEsw^vJIR#2=U5_yF#4m}hsN$0#c2vSgnNR_Mo{E3I;_VxZ#Ow$D*fP-wjlU{s
zVw;E_$DyQJ55Ovk)Q;Lh2mR*`-lE%B@V<8G-hytcBsY&D3SqOM?WL#6Lf3T$-RT^_
zji?&23R4s4O~vsoQ|nJEdwevBv*=_LF640cK%Uxvv!W!G$WeGX^-CSnCfCg{;94ZJ
z|53Fs2ja2is$yUiWSd`iccMNu7XPR+%ii~%xx>>GNo<_kz7C$@H*WKsu@zBi3C6Y6
zGQDf>pKPSK_<K%l$fuMoAw-O`KBUXa6o!!}WJ<-ZIoHVL$dIs3(ZzT(Wzn};wovrU
zVrU}^bU4QFaZ{Rdx;Z@F5N*uh<aAn1<yORTOD!<3Y1^+3!~bW>87IokCNp6ruEj@x
zMwg`RZvEY9bGhg1_yeQldwCkvd6~Yn#<uF~OAYI)ZSI09O*fMT9F-XB)I>mw0n0P!
z@wHk^L|gBPrx})lzH_cwTcN1gnyfRZx)Ge|drV1uPA_JI<4`rzJ=z-HOD7#GOX4)j
zO|I$>L#cH3vQ0*oVMo|VjS+P;O-I<MJuT18$37WM++R9=Ngr0ye#tggeRl(!vPg3Y
zw|4UBd)^|z*rLg9HOxj5eWlKVTv|N3?ZZgQma}+syVEr;O>Z7WEHC^||DBNr38In4
z;(4+~!5|f)6+zoS%Jz%o!o^KcE{kciha+1DUVk{v!}+fZO3nEhm}PB=&k)U)uj#;7
za4ABwAlcL)y`TR!&lH2GYJ&6rnHiW^w|hmrea^nG`ql4ei)^n6Dcj#1fLZ*Xfkqnd
z8E4t5;42~{jvAuw#ur}1n%cAqK!p3)G%d+}8cL>C_=LFWb{<@Ge_Hgac`@=Px#da2
z`<*3}y`&|sp(cO6%e}3`D0Z^L$8$z~K||3vR&UQV<Br=+yIpGABIqDh^WW|_3GmiP
zC;0fsCq}Qh3g6xgv{4(eJ3|M-anZKd<>dP<jbdXqqBu{_TaWKN*1Q`vSA@^l)lIxj
zqZd~*lM`wYW9OjSB%JdnR9MX|^c&Ic?YlB}o`np(zjbk!vlRi5mbEWwdY5(+?1MOY
zO7{3RUgKq2ZKi#naT-BVlSZ9fQ;Gno7Pm7N%9^g5nmlM*L`O(TjCU8Ko#Obm_l$8C
zQ5rA&!EJjvMs@7hV@cs`Z?E=N+1Dw{$la+87Pko&DDPVI@#i9&nQD|-5yfjOf}_M<
zStENC9|4infRn|yxts0-#wGn$IByL>_mcRYk70OQW-YTkiELAfaFEoSiV|{Dm-|R#
z>Uxpo#t9Gs|G}??>|4CW%4WF?bOJJkJGJXjEZNb~VAZ~soR{Why&q8T4tK==Oym9`
zHbQ$ND`;Nc|7LDXgiP!l|Ch#bvT|_z|1^$^jg$F*M&nw*RaIlr*=e_>N%$xUi5OiZ
z-P>H?82b@f7-6>wiO33Epd`anLJ0Mw!zKT{lJqa`bp(9<wD#K1Xq(UaxdFbehK36&
zhlot!n!%=q3LxR_<O?7L#pT8n5rM$&oj?M4+ht`t5kp8`88FkO5MbRy`wNJF3<*IY
zf)OQU(L)3`DhCRHOgSWgI2ZzS1VeNb!63nIfdU8paR&E@fwU5MhOrAqp%lY}2I|^n
znHkL)IEPo@17#mxXaYu}(E90UX<r}PcmyY*V8b;73xL=V2jQ<mrY*poz|Q(>2!DsH
z`iK5TY84^O5o>Sf<>loRlCR5QY_H|KR0)6pGlZD`)fKdV^V_<=J`DJRkm5FfmeVm4
zsM8H`RiGVy`qw^iP%yA$Kavr6SSK;xR>8_|PM}QOg5nbBd52#jKbY0uOdim$*6cub
z@;Co}^zj!BBG~&TmWe62v(qo39zsQ(e`RP$u=5ET$D$s@c_4y{mj+>k5*nC)369`{
zID=>tfA!Ho03VgM01=u5`?sE2froRF=yK^1#r0_)y`%l&GUcjhg`J*&f)#5&_^whw
z2Z7Kb8oL>JGH-@g%z#)rWf=Gc()#CZIixk4W!0H-cm%G7^dlV#dGHIP8Jq;@Q_v6?
zn4bt}4IJpXrRne;O@DO?>r-dq3r5^+@AMGX`PUj@HqeKVP5-!nE`de@4`x6>SKyb=
zx60ivGCc=JKekaQkY<oAk>ZwrF^OTr3;|Y$59ct-Z&*>McpRY5r;pPaR4E2)%%z1(
zgNNOZV`U{3^;CAFpT#=?ZcEDyivIB6Y=VBG={dN)6Cr-EP!Ax#AH0)%;Lq((dsP&$
zw`$J4KBUwFcE*6)eEfu=kAHQ4eHH`vs}l_%-#o$GuoI??0mlXx*+6fB>f;bEKM%m4
zZf`$_U-gtf92Y<KLicuYaed3Q|3ZWWY~xS|XHV(CNt1SvM_LsF31h%t1}eU{`6}w5
zIwD&+-x}2+KtyvQ&{okUoDduY6bS$Xc*oRmkDyvMB1oF9e`#}Xv~k=xL`qGR6X>_6
z5g{+&z@K6f{oGX0`|5)z-8FkOOi$IHTn%_D@aAt<U*1m;AVggF`$2GGsnMZ1gxh`b
zyh^sU2aY~q-5f<skaG~h=QD_gzyU}9S~Wq$ezFIccM>Sj_M`7au(w2)4-I4xhsqxk
zs5_8<QU1gPka09Fu83+-r*yZUy)8jtpmZ|GZvr3$<bdFs{<z;b!~Xr<{saD;4;*n<
zSumDwm)}1FuL*3z1>-v5gqimW`{NC9gu@IsL<tZd95F8LL4w=5KfnoQ@^*2M5zU8v
zy<1z7sPIkSas&R%Brxv5!hQ>143B-IcE`WMNuu9<@y2sM>3+UJf(G(_(`=nfWh^&i
zJTEh3`KwDSc{|j=8PU_WcE_+>#qCKj`g>m*mq(?RF`f|H8tQ#F2xvK1aoV{2P0zCZ
zIZY^+@)n=>Nit$T$kTm^dL)vVJ&jIw(_zeIi<xxOWPCe@UJ&X=zXpXdb~&V|f1;=t
z0q61HpkCSCZfUo}gaQEN0D%iYVkK1|F)3@&@l+^<!N$l$FocnZG!8o)qH^e!o(^$T
zC?%{ppPyJv!qlT3t}QV*JNshsZBKNMbx`L5z$b?!DW!T+v6|jdyr?D)SW+(a08BCu
zKM7RwG2Wt?0Mt?5_!CvW0XdN&%uO*U7ZB9HYcDcF#EQs!*a#`%*4bGS5FQyHqF8l3
zJ&e1|k#IZqVVC4q0g6JtK?zRi)pY+F(C_Fj=_DA?tX9!`$vP3hZvtSkSV&yG5D1<D
zt_|ZRb2y?MxYRoutL-tmTNVnj&UVPbu$h0Y)o!QauxHMutJDuate~a?i+6ond9*kZ
z3w^|7KNU|B4-*I8+gL-3bKz17q%0MH7I0{@@V;DPUKDO7Y%v)sGYzP#@FX7Hp;4Q4
z5^1$&InC)XwN@r#5nF$m7~wu<5A;6NuGqo(dPWn}Dpoc_e%d)%k3M~w9BJldQERXb
z5sZeSJ;#J^ToikR5T3s0wU|b`v0A0heQHY;;7YUQ!CPXO*D`{80o|o@)nU@chZ$c$
zhAJsljk)S*>5d*&`l!f-KJVmr>!C}W=Oxh?lAo+TLChT`dfWHC=yr|>HOlU{D{K5)
z%w5*!6fjeTVvuAQoj{ZW{oc-#2@+VV8}r@yb@H`zTgu@hPo!{KE-!;o(rZ{&=p78b
z`04h%?CK5Rb|2CUFB1mS8Fbxa?ssX6l7OB9t=UnL(*ty?{`Wel`p6ucz?`mN<PsIK
ztp(m(@BaMQDB2oFf%<703uKQh*+qr_-~k*41o3C(AemAr_?TUj)l~KfO&=knGy=^#
z+|~$rHZugG4VC)Zlb(yHM9-8=Mk`%ghnoV%q=-s*&w+OazUGNOs~7DTZet7d5mbMw
zFz(%HZEYtf-wLAApRpX7aaceTHw!ivbVW9o>5+=fip$*43Fi`vS>YBK)bs4aW<atk
zg^^ixmd7-oEl17Zbd+k}`)u|hq_obpAN0Yj)|uCV3(|)Jar+02Ja{d?oje>YqD5uV
zx;C%%L#Vv;p?7y^3I!W?{EJ5)iQk8NVB4;N$9tuJ{l&~iCYEHA6UPoC3=i}%uV09l
zqjl(w-Eu$Y59brT@BN+@c0@$t%l)EOp*<{<LDT8M`XI$;*S9xWdniLhloqs+@=gG0
z?Q;u+Ss}=*n~J4)3j5B^7;+-DLB*Y9ITZsI2`s}!ilJ^I{g}ZoNS$5WkV1)i1S0}~
zbDqN{Zfl;uk5EEpfh@Hg>&;zA{k}-TFkE5&6_W<FGthT5KM7OMQOR)7@jw#VfL8~Q
zQbT<GorImkDOrI%m!h$%wP){Mu#}W2=E-Gu3bvSEZyi>zoEx%WjKcAeOB0JIS^{8B
zD%&gTVX=``4o6|6s^FecEjXSmyTJYOO@bn(eYD=@1-wyDwY^i?C~qIH$Q-G0*f}8#
z>Mp7PG9t>U=S=#Vv1G2Z{f$&vE>ORj1conJB7%4H6?uhjD`R3SO^%F@`hFdhK!c;6
zNm;-xz6zrF6+|Y^K+$!<k88&1sx}h<EjnoC5`7LQ!u9fcF$X?G5#9tCKR=`*BToaY
z-uWkdfm~ui4Rv-pn>rpXN5N{2{L@Z0Acf4*QiCBZB5)_e4;nZ)&s`v?!^lZ+o6@&9
z&f<JT=G8;)AWpN6j2kso&ceK3_BhYX+95-z_SAmjqI~TzIDE_e8i}V`hhPu@nEtHw
z3+AZo*8X}*yP7r%w{vZUzhaifhJ+dCL)tEq*}8EUG&9?dSF<C+UAiMYHQvLk+TNzo
z!Tyv~>C8}L3E^kdMZr$1S#7V)X*s!PC4KhnP&$p~I&?X%&u>wsbS!FKQ%w%tVF+J|
zIaQJdF{Vy_AFs1BH?wW@`Jg_orgceQ!G%Yb(kHtsF^MbX@Qpr)5}@_Rl@bt4fAgGL
zY}zCqW|TxOe{@DiJxOL`lHX~D=63I64j5Jo<z_T{M$-6SjGaS}DABrRtLl_(+qQYi
zwr$(CZQHhO+qP}1&!At&i|9MJIm(Ef?2O34{&%kLTLN|TLv=Z0UFsbIt4W)V-f`nm
zG(T=U($;-{^R#UI{K<9hlkhwv)i)x($;hIqUI8aZfh#J&t@XX63kkD9iqGWB!MUGE
zg&Y2?*x}Wd7FI5s!}ad&Rv^jz6Kv;+<Iz##ZhB#5HMqOka<j_0#$*+3v0!Ph!JY^t
zzH7iexxGUQK@xneOrpDHiusQOLAjWFPKj`h_c?Nns(Q$hIbX|TgqOx_*!v6gvN-RU
z+i`ggwVr(1Vpf&28~WWjv2=$nFu7>#5Yd#)0?)3z-e5q>-xPS)DY>m<fu96wd*tAB
zFl5XWit<r_$@2R4$KL8>!1VjV!W@f7(w`GoTA?7ZGlOH~sYN928bYSkm+VT_QY4z9
zr)aU~4Lm$jML5M1`O=GJtDP9X(?S>Zjsdhkxme4;&l62Q=O`y0w1k(?#c>CMY*i+3
zefjK>8ppAY1N85gj<$_!)?Pf@jI``sw(;+PY?RtF_Oy@dw+(dIFHQANg!lf;r$Ts9
zBI{Vr%M3)=t#gBe3=DBsG5*f*6(v`)b20kl;k*3$xbSNLz=e@uUH*+VEm6nG)YW=X
zdM;%V?$ylb&?D~%<5Rq)((Tb?u(GbHg#vx7iXzXH&`nI`xYI)xfq?x<mvR8UCsYko
z2pUQ#1F0O8f}ZMJFT70@KytSZY9Pa*YpoKjS#Db7XvqPJJf_PlYcj%k%|ZW>yF=f$
z-@F}SHjQj=)GDrI=qtM;M@wXMLg;01C(UNt=J~6JCECYYb>2bX&=b9wqqM7VaSY#j
zztUF#-%ZA`;Mdarfm12l5Lul03_|@4JE`JFZ5QQL&Xu%mC#Izrq(ADXx4fi!8@UKa
z4L!+7XM==W4`!cfT#PeAa8IOEoKTtww_2>PZ{RTfG~@B~wVQ5K@wBMHC3=I0aDggF
zWFfnzb?5$;p_GpFwNw0X2?Jn*F5m9ycWw!ZPVQt>YQiOgBrdCJ3DteDHIs?dlKVTB
z8yBBCtqaw=YjhlTfR%p}omkA-E{SF=J1W1(aXwGkldpe&Chj^!bP82()rFpryevXz
z8`65nLH>4-&74UeiU-@FSI&<V^Bs6PRF9knm^+T!fiMnUq(-X3Mmvwk4%@^7oL#6F
z#N}J$B|F!V79~0<XN~EeqA=_XWA3!(a_*MnHQJcLm({pBw{g5MCyIT~E0IsmuACp7
z3ooctac^~+sZidr*wyRvEMZq8J|@`qLZnfi6|w282=u_1XnCfE1eX@(|K4Vs$Zk-2
zL$VkKA=GkCdPjUw_AA;k%bsP=bMnUjZlwh3D;Fg6dqkStpwFmufs+CJLgnH^XHvCs
zM&cw4coQ3y#=O%mpQ^P(2pf&fhpU1R2Nkw?_9scj-RZ4S0EBmb%NT*38Em9>VgV0I
z>bN94e(<=B1U{Fmr5A8^N|`~IcW+WipU9}Uco{NvVPh9m;W6)hhIE)!K6NRbG9HT>
zWH!adD|#SzR&UUK(K@E4`;Kuo4x(8l`}fhIt?#Do-A2-1novWXeJgE2T{Qml6ZUhi
zk31&l@^hiD#?|`tIP*wDYg<R}G%g(c#ollJmHh`jPyXBBu6(bc@{b81Mg*KI%dgRp
z6i|<Q*JF3ihSjZM7AR!rHBm+Zt@GDI)he-MA#gS<5+jA9lktQrv$Ce!TV_QIbL2%J
zCY!}3DL2b>dJ_9&i_`a20fdh>BGpR|$l7cQ6LZ(O`G&>x&YJTpb*pgk$tc`|3=0Ef
z5xDTg=cYtjBE)yw69<oDW9EYq=3%7n3yjrpL1K<39N(~3LO|0dTNINy1cs}w$j&kh
z<iV=&#?9GB)V(OzXRpIw^0*lfDWN#*N3*1d18l_;+YN3h`n<0@{%b4!eMlx9nL`{2
zb$od%c^I8Huci~~k;{cWa$)32Q`tB3Ufi7SUv`-WHvidOf?}7?Hb;Hn=kGAsh#2ir
z1qmHz0gOJLit%7UR8A?wq*tqIlM-6H<5Rov4OOR(zLQqa6SkmbiNfV~%o+3dFKdoG
z!T|bnoRS=h{envFMQfrE_lELxQ-Ie@zgBgjvt)isd9S1l8MPF{GNL1?Mv!gZw5lfd
zn&$CI+UN+|iP^aVhn(QGUVA+6w{_EBFV}ukmqem1e{K1|JAgumeG%WfJ=zB~!@olQ
zp<jLfM6UzIQ6PpkKXPdia7czxDZ}|7f*(hW3stx!JTHc}K7DDEy%8d$ex|U!A8tt2
zvEzWE)wD+QWGyosMi)DR8ED43e^}p0`f;D<nGw;NrccioF?(Q^XH^0<5;i;rQM->!
z|K)<<B*9^q+k52kHc`;mz;4+$+`CE^lq)rBZs~~%r=`9PRy-Z9fJKzOHd6vR!hs;O
z6P&z$%+mVJDrWN^gW18_(OFhBi)V=^F0nThtTGawp>3+^;4WX(dZ%1uItnubbWgq8
z+V!7$7~!PNqbuD?e6!)`Cfi_Y5x&J&=t5%`X$mM$M0)5V8*7^EK&`~9AUk#v@^8o0
zo^#Hvx2dy{v3X&`a?XJ}YRwceG?`ZxgmrUgof=2#*uvj*dMHB{fJq4ID$BJEyF=#N
zpuV^5yqy~ybTD7^1*uy~y7wQld0!wMOg<vyNdow(Z>_!$mPSg*#-7<Lb{oID>RbqL
zDj0WSu9&e%!`@K&+cvynEjJ<q3aE1QUczkzP>K|m>cA(ASp}^b)I^Cs3*17xYY+d9
zGkrN&96Jw8rkJ5d2ZD$+sg=|JG3MQd8GK3*oyuY+XbR!w)7sO2$=!*anBop$YE_`w
zrGx%1gWQy-EVM~co@75RP<o`_H|rBl=L;P394&96%kclJQp(3aQbG%^936L-W?Hnx
zytY}6VnCP`)O0#!0Vbz3By<^>x|>HJ91zbyBfpw;3I}<nb{p@RSsKdQWr0Tbp+w|x
zot!5Dybem`otKf=`{=a7De0vFCa*}xxzX0_ktyP_C2L)&+bB?wfG^=H>GnkNYV+x?
z+76oK;uuYoaCpj33&b{A5qhp1$D7d-lvjdkFZPJJD}4cG9Yv1a4CTM)np<*C+I1Vu
z3tgU0FK8NXru&m8Qw~jV56;!*gU2=U{4a27rN0bHg!K7_r7sV1LmU6=@wJ|4rLa5v
zA>{lcN(}b%<_&-=(GUbT8o{q`WVMLKK~-<LruFut<(k4{fTIPX^s;sEQ<T=wQb_$x
zhxY=PcC6_Kx2=7DK8ArHYT#g)-frr6I@?EQLdNbXRm0J<RjGi^J>Y6)s3&;nO18y4
zK0x}-zAaJkbCYe-D&HJ2*qjcw0Y?R0(@t{!?GZ8=Y2M&o39SZt)5;8np%2&`^ANu=
z_ab=H2WcxY6B9E@9?6tGd9?}40)|DSFEJn@$){{`@7j;<GY`g}LtsKa&BC9OoVPZs
zXjWL>r_BQ+9AD<qXy$pG^&<AA!XGLh`&LBXy@#WqUOyfw-~p8M2wVJvxO<b2rXBR0
z+uI#?^U%!>`RleEo$woS$HISV-v4QI*YfZhE#eX}g(K0QrZ>u7h}~SD_?3>yPtoJD
zy_$hBlZ>HmAviqSyh)kZQ3qu0LR4q6o{4y<S<t?UsOkAH3>YF2_}P5}=i$}p9;Ktk
zgFMm^bXy7~T|Q=@#G-b-4E@)f!7Dn2L<kSai4kW%EGVyD!Eeyntw2BcJUS$%{oH1x
z=hgr0g_LGjcoM0Gkw_@IlHsQKMgq|dyU_*085C!LQKJ_L<Y20)q3(vFVI#u2<&*jg
z`#WyijFRBA7E4Zb#KFar*&3eztB7?$;I+Zo{an&Pey30-i<d%NB2F$3ws2+yadhI*
zbvj$Bjf;)|y3#t;J%hkQXI!MHNJtaCCXpDzMrcFQI^Ks33~iB`xGAhHHfa+CQ5*sO
zarzd2BIC6uo^Im4XJc7DXy#*zStW~Nm#_RJtzx~|?OMH<%u{<LD#K#m%TILB2Z+6W
zSGifrxVArkM%TsPzd)6wcCI1eouc0PwS}G`f#$FAO9d6(R?(_{a-+j-lwSDCeKvn|
z%yI<X*GqCe#lGIePco{bn>z9!G<oZ69*}&F{mc0ndQOQzA<mN7j4Io9HBZQC;%kP#
zPf&QOTTE6W;gBuhD^}<e+6&sB?8kCfD%%<ld?q%2C&26<T|rFYWneyFnb2pBp7XU#
z^pt&P?e0@b#=X?z*YZeLzh$V1cZpV^u-b6oYOCekR!vpEy0W@E|HIg*`%mt=4v`S6
zm)AC)6WaNye?|+;pqBH%?dw9#jQwWzR_3L(w{@ol4`Yn$fyft#qgOgzFahB5h}XCC
zARzI@{e?>HaFNCWo#tj~&HeD9gYZ_DKtDh4<U^R2Qogn;3k&A!RZTAynlgbq4P#>&
zo+<Gwav7{P%53-k2rUT=1(IBKf>0dT*jmh-!}J36B&I=J+y-d&J0HYRU}<(%QNyCS
z>mO_sPmHR79|%`}%Usn48H=G-P-TB_9aT;XDn-xlJ56xF+L5F9r3j}$!_l(Yj><!#
zgnXhKNB+KvxWQBg%15`m3v#e~?JSivkMu4Tk3Ch1B&RL5Pw`L&tGE=}`f87*xZ4MT
zPOD{(qDUzNi{mAkGJWfj;6RR&vViA^egZAcL+`X=<~_D!t?*}+``Tlby)sfSaE^Oh
zBMEM40Y{b&U{=LMvi%CTQ^s1n*p1?S53yW5zBAfpnP52C_n)fg@vha<&XL?gkjz8#
zb<6efq&ey>J0Q`huCa{cqcc2OO(y568}n~3E!rf<(Nve%s3#RCF%0KW!F;<W$px&=
z=Q8)Q-mAMQYIu`ml6GupC{jkwn(FFrSnnf{U5ni&93WY`W9KCG7$zSR8sd|2%E{Fa
zV&+WoQfZWC7rO;AOBzG^lo(B_-R|l$9yO^C8(#!ZdRq`G?!76}f(UU%^~f?S)}WJT
zst}z*^`ulZV5PqmHn+D5CrYD!#MrTko^ZU1`xi*!JTvt<Yh{e;EJ(hlBn#`<wT_(r
z+9lBw4bGlHJ*_C+vQy}hcU*|YgUUl~z(UB-`X0+z5u0|TI4_8#*eI5-w4o{SK1Ho`
zBllDUhGbETRqPF^hf`40WfWO&=5v4FTdH<d#Dx8qyztZp;h7=0o@wXc+D#6bp0Dz@
zuPN96n%DNPr=K`;eh|lb!ur{6oHE0fShcmMhH+I~cpfd9RdgEFP)v=s-De$a^{n?g
z`mp{JR|MB88K(ZqSgDpvw}Zx_y50zVau(7UZUY1cbjM_T>6m~hLdScp7O}pTyJFCi
zzNA^@f>RDRSUWG(%NFm~<JsQf)W03~^!0Cn^6JD&JmjZrrxW??E4^-C0(D3o4wTf;
zPo!q9XPKC&_!u!~)dUy6H$|ExVn80aBAzO@etv&U{Gl=h-BawEFJgEy!)rOjF^eTz
zPkZ%=_tZobUY_%y)y-YCq7aC*ynmZnH0WLSzg-5sa9S~SMMHT!o@#-Tj~RYY4+YKL
zd+zHp^~dJYHUk-EWik)t*{CYDz(r^WmEqT*K@!HOv>4XVMSd6rL-xSABbF*S-F(Q6
z&Le|WRij3zw&`coBqnY~HEh_7tb$&2f97*kygy7=pVe_)V;hb*6{}{2^Z<qdIY>`J
zsphy`DWyv~TzAI3=(<OtrlVMDdYM<~SEfp{N92-ZekCU%kyTpr<t{(lA>1&nc+t0W
z&(-@p6&)T3jrs(p$5{-luuW(|f!Zu!I2_0+EOwJNY6|r`RAd}TH{+OY#i#?Q{FZnC
z+hpRSw3{2yuI@E~zRx@IdeqdVejtJchvSh$Rq<B!IMfA`YTp@kmd^Vfk%^UDsNTf7
zKx@rTPwQ~)ELD<U7~pzIjE{?BNr>DxC*9wIJIJneWkwetwz?jY6S&i%Qt>vfG{sgL
z(6E4(To-4`8@JX|f$)TNEhNc2x?l;cdoapNd*`wH%7d%S)0o8Y-0XAkX52bvC9`x*
zb*S`rJ(|>{8L*~`dbH8w_o~r=+UoPPQ^phNgjxVaeK|Vpg7&q&^3~NhkXp8sp_Mak
zQ-;Zn?OSK=b7)$r3QJ}CLg`~E#(&Xx*8DEK4o#T<M4-AzOhVP>1Iq2tm7JBBD#KhZ
z_OPlXnd))}FOBp-SG~4e{PNMATvzeU%c(Dw8c&MmO-N>1p)_q>Z!Qn0Woc96q27hB
zCh8p_NWk5l$J=Xny<i}x|4_UyPShSNMc~*<)}cdRNj|F=KPyXC$rH${qVjegXKkcj
zOXhOdb6Evu*@oLde*@Ck@F<6OP0CJpAjyT|R@I^`lcb{~>K8gE_qZT$`;y>zHf!kG
z@8C`+tSj})Vw@u~uWQ%#x2#yG65Xy(|5IW98?gg~RzW3#aBu!`7C4_ZJRUF-oLG{$
zu81wejoVL!7qlSpJk?>7HQ)_K*FLzgmQk1hJ;~fbqe0f_xr8jN7(JTHXseJU2}M|e
z)oHngZ9DD-B4u&D++$tR5&lq0+^gZj{p^_O{FPsDeDJ5t%GTmZdX9Vq7N#URAWcu4
zHJ6W6F85j1`&*FxV(lh4H>F(IBe3OyY#L?9ggT+z^%Oq1;=%@MyCR1DGUO|fHz@_=
zC+}dqmOz?}#1*ptb=PazqdW=`C1@}-_8%1a4q0y*Dy?PKYRgaaFmIwTC|CD{G+I(k
zh}L(u<m<)(Fti>qUfR+%|3p6ZPBYJ{Ew=wAvE&5{gXcBCM$@X-=$Uz#d>@!lRe)8B
z^^iSFWw9Cr9f0kk8o&u6$Lu~e4M(S{*wRFb*S79ceWxcML%AGIk_hizA@=uI;)`El
z2;l3}v1>Ek`xgsER`4pHb%$vbcw?@q8;SHVK@<97x!`nGpu<Y&vQptO<PBLjAo5$t
z(<luwd3QIg1oGOIRNJ=n@*`4mVJEP`V{{v*=Npp(=F%KpO*}C+N`;@B?7;f!@2F}5
z2ZGjo>+5JEQaGdxT7z!m-YCO#AF9%pcO^mKv6=i9J<SUuWu7d0&?uHgIckcZ<9m=}
zc!0u+VUpe{_~g$Om`&}CyKfEZ<}z~4;qZ$wLJcyrw61ET`v}32N)XpwfNFN{je_xw
zWN#I|_oEv@oxulxio<Pw>y+Tm=ef>z&XJ_q<3Q#d%XG|JBKxcN)QY`n*ht~_TL#9q
z37GwLvgJOD^IK^*{b9v`nIO@AD_?8uhDUS4M53t675CXqOyaRE!uw@ECT+W_oi2(9
ziMf(%y0|#zsYL<vA6mapFrI>V+V29`OA_ZVYt=1HpjXN%G3y5SaB%+sks*f`Kx=eV
z*}`y%x_74QSXg<YD!`<F(y4Kl3g0~~(>XI~zLIj4mV#<_fmL6frS4LBB@rUlV-f1x
zWoxlK)ziO<sJ`&d5!*yB1+EbR?D0&1SCgaoeSvQ5h8%(`)F>PdESw$rw$ld*(|?bM
zT89EGN&Yr0Vk*j>|CQF4{4q;riD)-`SE$(^-bhZ4+EL^32tTj$%UH0{Yx&bhe7uCp
zRbyb>eDPjc85Wsthe+KuuS(c4Dt~Fr7ViuR!UWUA_14a|rX$=B7KI#3=Qn^b$4rfY
z(iMF3qB0j`{t|8r<H>qZ)6>Gvun)r&3TJIIM#BcKNPk*HU?oc&_Gu$<nU>O9e>{aa
z8>}LdipK7{+41b1prW33oZZ-O{?*trj8QB%(gI+<Upvr!1}Nc<SI)gmS8A*_g=d8@
zHtZ{yF!WC{AV5HE*hYCr7V#R%pPT0u1K|C_`C{1~;WTxa8vgL3)27w`11QT0eOs?T
zK7X*c*UOJLJhs&{9lL5+O!p<U9h1nw5G|_ifiXM+2%o58h-_vBHgdHML_%t*{G>op
zN^RVAR%0gZ&7cbJ<OX5cb!fOe{T>Toc`fX?ueCYQ!uPDhRfge0cCui8(9R|j(;H_z
za3uB9p6u8Qb$!G_r{z|WqypG!)K$e>ljD&w0^uTLvlh_NhA^mpgwjdX!za`a$F1^=
z*6)PHZ+Qt;V~pW_14D^@0#N?6Z{UWM^H<p;grRH9w!x7Gn`;x)-)lIE#p9cMe)k$J
zcMm%8ukQ@}zY=he7RMrP`IIXecMT0xCB(Fm&1vIib{)8^4twQ#Y<=J|`ImS;T&YWb
z9`3-3Byeg;kuhOfX#jcGK<?9zwsGnVypPs#(K>HajE|4ThPB6VH^^7x1Ewns#oUEa
zD@BzeCf1V%yJ1QJNW7-_Q&*)vQh8TLLazAw+jZ?6SBa_9xgXkHIYkrB2?5!(eZo4@
z8Es42nJ(nR??P6_ak~Wtg$(hR{2Z_+CyjH}-EcGgq#`n90#Xo!T)vSG8f?SC?+!Js
z!;PdA-=~RW$_&bGM+5{$+q=r8<S{)$9@}VknGpxP&nLGt+)%NlGN|UOhbef7%B1#Y
zDa1wZNDBPZg{s|$^?I=AE#%xwJ1ANvzxc;TS6lB+R!t}sjfv0UR6_U0gB$SPTt|KC
zU6Xt%+loMqUc?*|=C>&;YH6f(3$_o}Ven9jx+f|I`DFAWoAEzU6G6O>*msw4dKrmT
zc#=D?o4T+pR28T_oZ@;6YS@Z;Q|wKK0)ypLoLG_81C#i!He;=y2xzt%Y{NZX2~1TP
zulM^9>aPO+e_!i!uare#4cIvk4OdQL{U>9brT7Zx+Aluzn{H~H&CVb#GJoSI{xCu^
z_7p)sIHa^X5V1HaIod+O$jSx*`R9<krV<RE{}WhG7V*^5_}RMBbh9kL2(kiW&(;@z
zCb3Sh!tGXs8(!MwEZ4%y)`o4CHM_K%lITWAS8rd2`_YMwAO)>v!{+iF5@9NIou_*f
zBU>w8!sfKml}nB$aW&4(X_vml#IGd<_{ObnQJmN9D4VcW%cJv%icpF?*T-jmZUP{1
z43Gx=t^48!-A&3rIm?W>^BBK$j`?h4kF^0KB@I6Hf{^Q8?|HC<&E-{8oJB~`wN6?O
zKFE121l|sV(f+{1&V2vyLI^}QZ=>=im_0CwoZ(wDEmCBgV?dJApj~IM>;A|nHTSJj
zwgu&~<xukCt{=Uren+pt6{nl(9=sR6-HRC-hN7#h^Y=ily$LpH8q#2s$MHb35kruv
z3Q<d|l>NK&&E79G1f&7|!3!2HqTIaM1N-{fV{(X5Ni<fkp!BG$7~eb-RQQi1El~u8
zIc@+fOY?XuX|f=KnuI7zHl0U@&g~91($+mza2EZj4fOu>C|q~8ciZ{FJIZZj`OknC
z%YOj8nAqt5U)=)Zzw7@8yqFjm=>I#wOA&%S$rBCESz1w|TrQ=xfZRm5r4k)f9{p2g
zWQyx#dWy^N1jo?as1l1rf`ZaHazTVkT{3LUR9AomQG%pn{QyhSxpViGd-rwfvFnlV
z^lRolclY-z=0n3%)LM~q6ph$XPY5fv9|MRBd)j5s9spK!XwMF~siOlasIKqUr>O&K
z>`)MU?ame)5w3{PAeRFr9n?WjuhS1?FC0K;7QS~FKJ*Vbdse?IeyAyUn&KY<&bmK;
ztWNvz@v3DG>;m26@Q@gQeXs#xGD!W{=+_EZ5@FLC_;G~_2l`+s2Y~;r&|C?1Z!Di5
zAMr~P0>+Ruzz(KfEnhAyzu_Gnc6L@s5ei-a94XJwD-I2O!QXtMPlggJt1yJW@{f#E
z7$2715j{LXG8m8n7*v5jy>1qv-#I=U{4OlJ#wL)97JLrK<_u7iU;8NF9r{<SgX8*Q
zFuogUx<sM<nieHCbiWV49|M0soORSdU;+&QloHuL<<z|YSX!7O0qTlcboxb36#Pt1
z040V&4j(T}xC25jXgmXH$03<i1inTPDKh&tD|nH&v-6)R0$eeqUeA9$3~fL3qTr!E
zt?2Mlrc#S}P;F|a=#cP!D<_WV^-$l7r(lGBv@v09b7TN|005x#X=~?DSpn^PQt*7&
zsE+W#4jfv31SNnG_@LxJf}s!P{`h`(z<K0PxBj{`%(!rK1eic?0Y30iKkC|7pcsuU
zZa>`Q;CDme$Dv?$i@kVcw)F#N*l8$VNW4r2Rq5G{)s>qk920y2CMH?<_)40pXmcwB
zVa^gu1VEq%foH3nd^&qvpBS|~C>fVGwkXy5pkQl^k^)~IyWzky6oQ5}*3AC8wrk@E
zml(jzJ`=l8rICf<d~d&=5GD#gnJ=Z4U$XDNjYJK8!x`b0Q$RmNGRA^cd+I^KejpD&
z5OyG@t8QRop}<$ZBGm{LZcI}8d@Hf1fp13S;6<B);L#jZIT(E8JK?&HG<uk!yxrWu
z+B|*}hw_2*?DYcR2Mi$x!<r&sfhI!`dJObn2&Dj8fP_Hk*ti}Z#2%mkD+U`ZhSu{S
zb*y~U_g3{ucqxFeeI8S4z07}+?%WQpGt%qZI&WX5C<FE2#ROYf?af5yJ#9o!W4D;K
ziI#j6rRU-)KC!u?9IjjBpuH~P#TjvN2z9uZ3_j}ePv_1mk8i;-6xV-jD(!wcb#A12
zjoK=nZGM14Y*SZuGLCy=yAL-Z!FgDPg>iX>(r#KW3}K74*t@EA6m_Wd?xYXXc_j~F
zIMQ`(>G>pVyMA#S)fzZ&;+g__c<t+roczzz1Ed-?OP*nZS1i6}H!HvplgGUa$BUDJ
zw^xm4Yc0h8W{+Ebd>Xg04w`;Q%RHlAR=3=j170k5M$#aUATbk$J{A%&d%Y{i^lJKm
zw<#`twkglCtKcEb9dBl^mcyt2410G@MQ2lT`RW=AK$pGlmMs!)pS0M@9xR`9peVam
zb~~Z!nzufQLt61R$i@tXkG!!Z-mUkHm)7Q*m|%Ac=6qBOT6Q@xQDQiXp&9!^39`17
zX-IGFbda?)mKX%(dwm7<mlhuF#JKPt9vcl=Q1Fl+`CW@{w?u8TD4ZGiR<7K2_A?9%
zf~ayj<~F4$5m#(&h}Bd`MT#)TKCcC<^Z34}R>mR^$a}FLB*#x!+?^nQa*AHjqsy#T
z`uGy$VdJ^v8&f7!{T1(6VUX%}{`U{_;&J>H%z3o-^*HZFe}G%57u)U8_|~|zwCJuO
zrb(8vLwnsM-|jbU4xZ`8#3|xHr~7Q)&8@n;0qj(vACYDI!=2pluxOJ=C3F@z2aYq{
zH3+@$sOS=vK9RJW%cgV(Nyq|mm#&_V$e8)1Psb(uxwtEVhu1ivz8Qy?v0yey`UoBD
zG7j3vELwJ9F^PI(b9n8~sz-@P${(!>qV|Jt+33p`IAoV?r+D+&#DS7w4ay;J1{=p)
zT}naOI!=b9_YP5t);J*er7T9ak=iwbAn~LKA5nLQ^43RAEHd}%MhWz-w+>Rr`N3xj
z1{K6WXHwuukjoy0Bh>CTIrg$3pMkekflnd04ga9!CTZ6s!w0ZnM30;F&g<sEj8|A-
z%XFTLJK3w5`=p|xNu1St+xtO`)Y)4q`e<+w7U0pHMFjg@u}}Y4R%ZsKt}PT<Z6Xib
z%5sC#q3&DOj%hfU%W!VBB&OvL3Sn3#A17^za>Zmm6{qSMuPVpmm+|nD(-~vn)ykBi
zO^Dap+B$om)W=|tx)~uqs+q?rNS*WBCpxDS%x@@p{R(?P1Xf$JP6n%FtB&d`V<id0
z(ECbhl;l~N+IE9Z31{Elv7@n;W-=rr?SPm46;tt|bW@Jr3$f@cj+a&4qefbyPU&LJ
zKYzGynu092BNpnZ8k)~69WZXOLvL+x#%NW28rnx*!xnpFL<Q*{?%(msh_+^w5(=ux
zBPpQp5^a|m7jK&|u*0Zh<#oEaSU+7ZZ56P(H66GmAtM!2HdjSlN;xIeVlq>&fAi2k
z{17&ID3*B+9Q#_8?nkV(3wraSjFtneb@V^YmmbPM*ZzG|x-J)`5v(#M?*S9R=Ce9T
zqDgQgUfAcG&JWQz!<K=8OdXYTziA1c1W#(`;!!y-c#>5ki<i)FkwV%5aRK2cUmM1s
z#Ii%&_6>BAsjq*X7zZC3_^KwSc6p3RJ$2S{8+v!M9J@fJ#}sd(Y40_j9=hmxIu6^h
z+%7_zevXT_W!;L;i6Ux67x2P2G9Mi)jVV8GL~&=V!-u`0;&k*$9b33n`rWG+!jQ2b
zBPw1JfOClJWXsY{Dc~|;$9S(idL`x}qHp2wdW-h{b@e7!tbfs3w41FFv5f`?u*(dq
zGt7P(|HY-dsHYn_NEWjgnh3bWQlZ5Hcdv8tS}+|qN9=w%QzVV8<h#9_rw|whw$1N0
zNpjy|`CWBT0`#FIC{#b5SmoMC)#179fMYqp4v{*TN{4kQ+PbS&6=k5SusJ@q=I<(u
z35kk1KA$W-qg8S+6FGnNa{@jLDbwLD#^sMIPslCn^iQO<OY*`~;gSC*l&GoLcJ1am
zQRjZE?iTWvXs&5r<|<VAaZ<P)*Zg#IrAbYcaTYKwPIabm%+03$DTABrl!IPwf;F&-
zTL1%^(Cw0{c0vRtyVM$|GUrmY2XH<p-{RtVwAH<YS*YW3U{ibfGaO6S1%c^7r`%kd
zU!y^pbBzZ{dB}Gw3CqZd;3HE?A$%a+{}b}*T?%rJL|hydGh_|bq?tH<x@>Oc<h-Qm
z;N|Eno`lQ5h~e#?1aF9DkWA`JZm8uA(3!v5++m)R;ibQnXQXOl?|<cC8c4D$q9v%~
z9f|uzMEdU)Q(v(?Z4)I`JR&I4(aQbInzOfd&gyB4zSV}`-LW@`0|zU_EeCq!T?z*N
z9hgo6!`vT2P-)b-VZPBac^JBMk#*Rg#<Ql%T}dQqYPxos-iiiRKSC8=uoY)dM&MK{
zapxViC;2v5sR3(oKgKn1cojZoboDGf5L^oO;KN!>6I-jKhiScE3yEF7ht|qGBApWK
zrtDHrnq9PRHrJBKG1WBI7O{G<OKv1<)jMe)B430Z9L*h4wh1)-)TCOK8yaU}R&1U~
zjb_h@5aTM8r>1&s>N(iV)+3akjq@q)btCOW+<Zukvykej3Q^cJwZcW{VCOXW^x~g)
zecRsK<Xc1qI0rX#EX#)n7|MwonIK7Un3v|8zH96aFYH7jC@0rSOko|ju(C~^64P_W
zL`F!o2h>)SPrcW2op{UrJ*V6JYD2E1?x3GiX!=al=FLhX*J@Ha=#UnV%qgM&^2f6n
ztQ?g*s+9{l3pgq&RX<KYyHL1%R&0JlIlDuiMf1hod3l18<N$f@G~Ynq^jPwKEvyZ+
zHMF^5^YR`dGq>vBK5CUWEo`|}*AjmtM1Xu1YM1Juw8h`nu{Bud-8jK+{hli4dfcC-
zfaNrY<iu-+BYAqp*iU8KSAmt{(OOYk8e!E$wnfMU$E1y!jcBhVHN|WB@RkFYZ?j*O
zaH#IA1*>kM;Ufh0zHA%Zptx*K!Q&l#-B<$^8LX69MI*s1K%v@l+3nas&Qr~sZ>^(t
zjO<pp5nZ_J1^k_+u=>w~3EO`fOql-1O~v>xhQj_o)~f&21INh1!th@XCQ$52=2$kP
z%8c{;267%K`DsXpUTR15GRI2`E+11~Q=@OZCwfzioK<s_F01*<Nz16}2pAjHbpm54
znATMmnxWW|`O>G~CS@((zn34smNl1|ukE>mpIft@d?6p1f2On6)zGhE@`0Ez(=DL9
z{%Ctu10-9>lqj4F2@c{5^#z~m73`y?jf&DQz2E`XaHAvct%-=?#}3NvksJYKaOA0P
zEL2&QEB!51u-*YN;MS}|h!9-sONH*DM?OJ<(XK~lfD7LPx9a~SqBltg$cT<kE+@oc
z7aimd0?F?~L$up7`w~GN0`W(O@&`GHl|bKg1P(yggcK%5jL$dY#W)mZ&qxD8Al&l<
zNP~Mk&|{550`-w&=AMj_Vg}~N4&sXhGuQ`07zzZ0wL?ISdbo!sm<VwP=*0$R^zJJ`
zN6&_NNr#~aacl%+X1w-pZ?|$M09Gazq(()>`IBz|91XA!js}{!<vkYQ=1<`#*TM~d
z@W8;07ZC_giyQ(*&Kc_0*~sT-7vT8xMvaR|w3qk}t|}8xNEax(4rAPDFh5UExrcQ8
zZ7t+R4~5s6#`xJFPfjX;bP{C;Y#cf%UKh-eq=y`iGB6eGNWd=i<tN++)R&jBSk4Ru
zFbc$;Gn_v*2^Mas7o12MY7d$agiZ|_bV#fRfe3r(I|R&&4E6_57F-hX=lpGraKX$k
zx(*m0I_9t3LFffLbm;8H+u>uJ+Z_kWDlUxNaqnyCbz9dO`UHwExM~;O@Mz??)BPI#
z<XiXbFi0dAgb6CN-bkJXKu57wz0w!)d)v$>M3-UhUpB!pE!QPDl5{h6&?*!0y+uz^
zVjCH>5l1dVAl{ySxWQrQKZkG|yCTMN>TjVhzP5gZ$Gs4`I>)}|zq>H)Q0-d5n`-;N
zaBRrYcEgQ;HUhL$pRRqTcUPdF16D%l-?ns$0Q;KzQ6hG`aB%ybMECTbOoV#&CdGw)
z`d)avGtPqQ=$bw3P9Y?IZroy0MF$^wOz8OjA-3pPF{)xna%1GAeTsjh<nRIa(PPPd
z;i*9L+$ITvjiteo0D_<ANvH{8Lv3RZQ0`%O)+3aI_ubNjuijw{*GWWGNOZF}E~}^*
zvARTB<t)>`EIjs2|JZEU3;uj+UFXWC=Zf9B*qnL9{e1O&pm!fxTPuZ;vRI7bg}a}(
zWbiDborZ=!XOg;9aCM;{euAIFbzNSz-Y|-;BV-_*{?QO#%$<;Gk}}MN-kHF=uhVty
z`C7-A32Km^W>-(PyO!%<9N*Z`N{;gy=lE4<mn?3qs*Ex3%l&%V7_3shA)0;6s9nJD
zrs+-1@Co=zFXiH_Kt1Hi=DT>?wks@-d99JbXTmmfBI^FyaNB(vKI$IP|BCqw7AsnR
zdJ#(Z3U1pScO88Vclo{#W4=3Hny+JF`t|gM&5Sc`Xe=*+x;@cN&`H!NVY|C@)|P}i
zV8TrkhnUmD`@vbjd~WUqc<gp|NHB^|tLAknZf)x){$jj%g6GRLUfsd1yKG;}Z0c!%
zdO5xKO*Elt<Xr9x{w!<uDdX(puKjP<Y5DoH#*?%7gObiklS=j{5phZ1zc*x}@g0G+
z;9t(+m0XtxYKF$4!m&xi<)}sT7hXS4CZ(wS!n^7c)_3}2nk!^{<UUABQfB{dX8!pg
zDxH{&ulswoBZrx_W!dOSq1A#;Y|VtV{{;<irDkTU+Gnybe}O_ITj{;4xBKm7cGTDD
zX#&COoTt@-a9vmVKu;AV?b{*thr5<K@c1Hu-uScNZP+;<%!t%t5dY9sapCBT*a$L{
znguG&Pa)}#NP@P6fo}Da2eGrggYk~1ot`*78~LcR?ox8Nc>;GO=xof_!*I9BrD4<7
zpZ1_Z51J4_UmO=}Tl>T$<9J7yk6d=a=3NW1t_WGFvAs;=XNgI#8aS?Qh#l{6RqrgF
zg`aG}DR^RMjNXPTYqNxtrj?mb!_`cf4TwMG3(w!A<wR8S7AhlOj%-7n7rN8f(a7vC
zYbi9?6`I8k9yyAZN7A=W%gdV|$-2?O$%Omsp$Wk)B0sps&J}D(HQyc|0)oOD<M)bZ
zud-`cZG4+(+a|bo+lMs+D;UJW2`M}_C*05$9mjH7<?gL*_=~Si#-YmY%I?7$dJ^YO
z&OemJyz5Q=-wzPc8J@zGkYutSoLp^6f(b0Dl~=vu<H`gnc%jlOs4`gWbKL^HQ&8UK
zMmdI`%^L!-CS;0FNk4Xe+J+aiwQ%R#;ZppDyYMnxzN%hslB%BL?PA{&*`Fw-bmUNh
zw=wa0#Z~Sx!gSv%M@`FG6H%OPy~8)F+4b}8mp)cWthRhTbC#Xd-WQMi?8nlkG1>Cv
z+!MZa@7C89^5XW#mqyb!5dCQiW!)0<hk~jzl$m69KH%!&V>a~WL{o~!i4mg}z^;~)
z#Nxdty<t1n64&g-WQ)!9<de8a9%K=%GGhvsPkAS&Lpbjj68bV|J`qnw-nEq#rmKQ=
zhAYY6;(Lso%a7(@SP}xOT;0mME-G79ig~|_i?h`+HS4Na3f6X+JjV&A?z)XXJ36tY
z#()LL`$|d$tX)dS#%4d=s!<S^un4s*a_n0w+1O;p)M4FLIpb1}5O}qp+KtOhnzJ_U
zYDwuT8m)7YhzbG?lf)NBVw|<!h8B-+*u(b0t7aBh#kZ8omcxjd&X<!sj8m_{@Te3V
z%7z>liS>V6C7tNhY?f$wIM%jP({_B*p2*X-bu)KWUF+|Z0ZLDg7*@X*(uK2EYHzQS
z#AqP17+rI7YFg3niLW(FS<Q|V6m|AJSrYZT%6aUquo_pV&Im#~Toil*_mxV#O(YOX
zF|yw!KgFQd=sTWG6d&9)#*WSmka?8F-w&_-72jWfE9NnE4L%C=%obsJPGSgJsg}dH
z+|7g4Jr_E<lkslO#I_fjGv2fA?Hu-Lzl%drOVczjQG&bnYq|@YSCWnBh4hYIE!FX^
zoOo%IZ`K<O^uJ_9GEGyk6}G1rd6d^bDDJ@;rbdodcCuI9kAo%tG{020c4~mx@>Z0I
z?VNNiG|~^8UaZ9nOJD>Asm6~-KYbuq(mELn^HaGy>mRun2KJ~*kqz|dZgUM<TmU98
z<zCfx;d&nM3(KY+=cNyJ<P2d~fV4i01U!bLJu>&0hZ$JOPCwssco`b``^d9ddmH50
zd=Ykg*(qE?Z9tYcZO*1#Jx8%K#_`moylHUjEo?t17jBZNY1+u=x@9XHp3+sCu0^LV
zcqb&R)(rwfaE!)$hr^^kw;MgvaeB^Ixs?vR>fbH!VAgy!(P4%5cyC}Nd91&TAXJC!
z#N0)*2o@B?gJ<inmXxfJY_Z}iZ(DXXomY<#HXcie9rrhWvaR)5{(=foyzBp`ohCio
z|H?Qkx!W1j$r@NFI$6`n;4{$E(}|cnI6D2iV`ccir{MU^Z1n$~3$<Ev-2qFq!FyHp
zsSpRYe5H!SU;v?@EV#wYyh<CImpL3YoB`YpU@qn4(pJS4_bS0}B^sl(p`wyX_C_6K
z+zrRNm}<~Bgi=EjBikui+M@KWyqkhp7V~l=I2>&};E2jnOEhF6p-MVAx?Vfe+1PRE
z71EsQ)QRnq3QgeOlj5!T@(0qSX=u8~cmc_ciBmm|1y!RmO?b+OBTYtSWNu{@G)fN7
z*!80b#(UB5c2FD#W1c`b?i1On@_@r((#EU)O}UkO?AtMT7)64%2HhF~>VVR#iZ};q
zTW6GJ2x|CCN=#FA!14N%uvKFf0VnE!XI9nuQne-X3Up=$T^;N~*W~aNSMl<_!|w#v
zD9U=;#C$auapi{e8}(^UiJkTWBUO@!61z#|2<T~z6&zm_s!MdYiE!t=a`f2;v>^wh
z$%w|Glxf;Ww0;KRm|k#ddDwXOA@s^DXy16P8i)Ce<8Jge6PXr2eJr&UZ6v6CZ9NhU
z9Sdf2RSwH_NY1+BZ0psLG<DhG4o<9}DXgC;v_+m%b;S|%lk2jz_+rLkBsZ|bbun7R
zTg!lxS+ze>BL)d8d}{n(7=@;MOIG}~-uB{yx^6##;x5djlNzk7+m9{m@RGNbE8K>4
z@yZ9<)Q#hhrafx`BuBGlZGnf&QmAN7Rh1K=241HuzNk=6qSURk9dZ88Crq6%pS}5>
zMTY4#Z(~NgAC2W>B?m*$)dtOIv&Xk<j*}WE;wbV6J76LL`On{DHvIFc>r+1b<z+#s
z52o$Zd)w!@pNi?uS1A2S^m{bx*4FE(0{5D_KuNFRS1ZDBNoza;)IGROG&foU?03Hr
ze2=F64WV@*ZcHLb>xDeG@L}FZQ@QTdcy(vPc-lr0Jm>z#x(jd-vZr7q2RyT#Re0#`
z2_qJ!8>VwzGG;%xoll-FTu!qwlNX6v`@_6b{JGqL5LRJvicqPiS^KNhnq0JfME1m|
z0?x848tVM<z0&=R)B98Wnj`)=f5~++-CAGFAt%SUO!vp@6;F_oldm^pP)VGDnwtDp
zKDdh-^{xyQAKZia{V&H-t~)SJHKs?06W35h&agH6H(C1Aj(M*aAEoT)fJSLo<Y_Mz
zEDK{&N}uai-C<3$AvBgwN!wr_+QUz(K-_H_IS#95k?y&sH*+;B&!f6L|DLj*+o@y~
zriF@*WQ8jxnWpb0lcG2sk0YhHN>+slEU3S#zmx~s-?0B?JFBsSa1ew;tzwW_$3%Pq
z%whxGI{4h`{wmb=Zn0jB8*%g5n4~W5UwbiHhx#tDMSFbhAC4!s_F}XBmhbL#SBx?}
z8!72dQ|{7oaAtV%E-rgl4iB^U=2nI_PMUDjHF+rQ@>+(du5vf1qktYutJlh$Je#?_
z+j+5AP*&?qXJ)ou>0WeKowD1gb7RC{Zl>rRTzr*!I5~V(q_-mVNPT@=X8a>owuif*
zC!$nhM^h8~?YxR*K&RZ;e;t0)IV()1JDoIJVE6J&TDtUHR+N2C9BOoAccMbO)E66K
zb`7TCC)>K}jPE=io{rv{dbIuCP=lNW#kdrJeODJ;@`LD@RMyAe7K0eR*0o?vAJ0=S
zN<#z3g&!L|7;u;sQ#wB&xP49BT3ZWGO3?rQeq4gSX}9=3-Cdt=j`nhMdVlSlS$FQ9
zdbs9F+M27gKaQB8mAqD_tV(u*Rl9CvTf=mEnl2SydQ6D?h-*t-PL{lHeREjwPSc%8
zeweVU!Jf3ze5*3i6&CI<kJwpzF>GPi#GY0X>C+XG<eK|;4uBs%hez(OxEUJsuxdG2
zhECnI<ggloH*>0CXhW}pRTjP{jMzY(Sr#tu58D|wv1ZjmPvk6)@7*}?qR-ffKBGaa
zC$(p+1nwL2P;O4&*w|dW+cA!G3=4HW*mQY4KN3D@=FV~jSeN$VcJuvQt6$dIhjwCW
zZZ$s*Y~dyyv^uSNrFD{aZ)&$nH_tgy=^!|1v)b6dBwk-{7N*wXHdvLCaT(5|>`R<h
zOWKzdTWO`K20@L}3?JTDNq2pYMEcF&px)(kD#R|;dsO9=m0qNC%J-J<E;~9`{K(NQ
zNj}>?mwFWIRI*)Edz9`}!d(<rDD!l5*DMEjn4L6tD5WB`bYhnAGac9$PF#^W7uQ@c
z`m$JGxHBoQT9o}1{wNT!RM^sjHe<3hreiv1$g^FX$V4v-N!w&pYShCybYwy=T|dih
zKm3{sKI^h5FJ6Xcl&{}Ku;o;`dzYT925)D7{3GgkD`y3tg}-eFiEU&PyWHHqZe8&-
z66ack+w8BP54?Wt1CMImw(zJvu9#gEml|r4I3z#a1;sWL$6`liV`Oz&1VXF{-qGhF
z%-@$D2L(tsXpDHWWq#Uoq<$##n+Zg>6k|$dW$<`>mL}?%*ut>4+P?Y5^k2eruGiF~
zvy7LhVYaG*#wFa6+S=AppcfB5A3Zd*em`w`T~oY&AD$N#SB`SIkm6Ux{?&}`63vEm
zX~vIQpSaCSbSh-7QkHgL%h280@J6>4=*Wm9)?-SYJR9qmZO|w`?sj?<@xuI?bj#e7
zT$J==Jv7|gZIwa%wvL@pm1}Ls+^CgLPuFcau6%9ZIlie+i#%UtJab!Pvf#yletnr1
z{H#imU3$!XntT=R?s9uNDm_VcVxwBPZ1^fpe|%cV=*l~m=G9gR7;V!W_lQ3YhN6|;
z?x0e*1{Vt>vK+^pk7iN$2v5w!b!r>X>8_mvb~ZQIu6Vy4+&A(`H2Xg5yPjrOep(+|
zOYO;oq?pn!gq~9UCLb&uS0t4CoMJyGJSy_^XhgRj*C~p5Pj>H~%7j}l;A%y61T)4N
zdpV@Tn)=$mx>nP&^WFP?Ijm2)Dz&~H#I)*7I+39*dz}QzaLglfNsAPY^EJlgKEYT9
zW_C)>Bp%8R+H*L;wEt$zWK1qNw5T^~a%E^Gqc-Eh{f!}WjI&D0+)tbKa%9VZvlZ1{
zYU@E?wwNj}y;@xFnrs!GhSrk{Y4wW0*4c?xYdav>#%*0ZX~19}NL3z&`p4Nen#_|`
znkyovNVV`7y8ldHUd)HkU2DGoyxpCA-!HDDx)38yCeyXLS{2()os4pa52#YURSaty
zTY)#)-oM{{oeiI7s-)Fbz9H*xNJ`f5P<-XlzhGU9U1nJdpm*Dv{C+>Ro{HjW61cn&
z4k;uVMjVSuy?_OB`O>uEwd#Lch_c^A2m<dBW{Es|@pf6AByLW_`!n6XXcfLz!-&%D
zaQnXgq%Nw8mU1=X*B5jxAFu2xuh(Rq4tO}L?QnY^f9R6{g0?Qn&(Gyv9D;s%C%!%I
z%E-jzwA{I-uQPcZ99RpsJFRTgVvE8B<+=`<-eJuhzS>MpJy=*AzL;KWP@ca~UX0su
zW}>A)pSsf!uUbCau&3$HPDs?k>36z>?Veb$wPn;9O(N7ESv&%2vBj)zT78phHWF5A
zC$8Z2xqd^pPqSLpIol*x>*!y?%0GOR+hkgA=v{s%&@4*UIE5vw&gyAuHSXw+zQD(7
zaMf8gZa<VM*XSox;q*CId+(ZboXh%<cm$|B9|{_{RA=s(V7?e%z%pElG<Hn0dQ-_)
zc_g}m)oI^avbLw-&Gb$<(?}NyK6S46SUooFf~&SpB*X4^?9f>`K7W!LxQeY=DeYRf
zWbO==K7;IbWn^@5_+5J}Z=H0UOU0`^{ASc;YU!@1PDkY~Du_?2w13hVJE)FmeDh*Y
z%|w08Ea7l)6?{UJeO1Te3!(KK+t5M!v(S!H)gC+AZm^t!5#6i2JmG4B5D5XrgV2~0
z2&5*YpUvzghil*q8-5amivY*MdM2O}QiID2N9m>BQ9`t&F)SVj4aO7Gh$Y1HBrfCA
zA3}puzohwhBZlKs$jK5TWRhK_8<Se5kN2hc6DOuI3ff|LOm2sS5{=u1P@}eLc+Fb;
z8Nh^>2M-;FWx{33wuciLzyn<;@UtBg5b8<*s8@)gq{3lA64Ak!&tz%Ib*{2s>(_U@
z2PEzLXArLwF&NG8W0?;a<B2D>BFVR~P9036FL1YHI0Q)mLBy&f4LCz~V@J#s{=*8z
z1;%H8P`}czk6p*%7Sl)==BReqler$X6FlU8wbviq2SR&PP&Po)HSv(VdQgUI)(!I=
z=SC-TAH<XI&MN_d0OFS#Gy>csp+=3;pRn>B7iYxE`!soCgaba0$o~k;M?hCX{6-Jb
z1<B(qaOWODh(llk@qn1cA!ctbTaV+v2ZLH`brDL8FA73t0Lm`q*Tru^h?M4k8Uzl2
zkW03SBtDX(4j>dHB~&CdhNWK;mW5=_AC`RhdO2f5OkRt6oD2crheTS<A0~rHNFq+r
z3bB){B^xX3-b`M5y-RWh4ut4t7~qUZ2?SXt4Y8&$>y@w<jR8S3h}Ad@?L|TKZuF)=
zY7e($kZCA;c?3+W5Se{b1CmY;MF_56F9NhB9;y>Y9Q5mhoQWfrMOp`iPsl2pCca0d
za}E6$GY~@_ABiU({T|7*sNW8BEy`>seJ>b5jOb<>ctvd8K}C^|7&W7HuJKx?Lkb}C
zc~77Kkg8@97*!IQG2W9<e2wW3M3Ez6;|vCTf_!%3ObP`5K~cJWO{H!qkv;sP`0NPq
zOAw9LP(~02`mp$nzJN*B%{f`w$&8>5#ibPc6g!*%vmfLQch-a;Jf5{I@f{DcN2tK9
z-JSz+a4Z+vJ+?qmT$3a0^G5t_IIwZ{CZQF{c32^B4?TZe&saWcr{54Fd|<vUFGRw)
zcHEx^C;<39Xh1Q-1RZ9~AbkF0r}<(ZF+ytG{W63Y%;LFDa4Uq6NO~=x7GiI9wzS<^
zE*G=};9n`b$)K+6I9a<RCw2$UB_Yx~$^+%ANMb;U3mE}!6F1?IE(~+}LDcn#6h`22
zPq!->VvKhr<>g}l06-=L0KF1uw284$^}EjpTY(S!d-gSt@_Z>Vcy}89pk6=K{fh1Z
z-MO+mD9d1|7D$#~J7N*!n`7?DAR`CTzG1@=!;m}WzDmDzSYJ5egueg7*gFRI5;c3H
zvF#n(*s*Qfwr$%^c5K`F$F}Wc$F_ES^PF4fz4ya;t4`HSRnLd1>RH`A-K$r3|GLLv
z3z>N-bP++A#bgZV^d#}kK<ap&gpJ5_nFDJK@Nj;k_Oly_p)`&fu4t!^U;V(@5j%<<
zih4dFh&ld_O3prTR4@f$MwyZEK;q?Hk;K0R$UXjClmML<li2WX1@;UiL7pv{iK9J<
zF5KT4NU9znn1GV-zkC6Bx88qWZrLlcc18Bxhy-(<NNcE2K_W>%y_PiNof%O*z`YBS
zyf_J81V9)a1<XJs2+8~Df^koH;|BQ%-9WOb6)wX`1MCydkBdk>3HCsd`EbdIyt8?E
z&r9Q-l+1`9g@c64M~Rli50gnDP4W`N91I>gDBeO6`VHsG{=5?jc0c^eoOSpC-Hd<N
zAg|ZhkjMk3B7r?ReJ2{HB-7MFSTMp{a2k}<(n)I`j2Xoi6w3T9{OKYD@Fjayr<EZU
zNNSc$0!jxWC27xr1H!sAPC5%8iYQ7-LFLjc9|)>TsnQzIj8P!rm}^0D6nu=vyPj@n
z525CS7uqnBqX?Rg97UPY3~82^cs~@9C#5&43Yk8V%3e(Gx?kW&oe?=KQ6(w1AiqWT
z?JaulOz^N285g$!=f9gv3LiwGl~|xaqNkRO<CICb2@fe=HU!C{m?F${I2yACSE~02
zwHV+fI=P%fgsw%UlBkp5_?NInj3Go!W-`?m^lv^k-kQ+nXZ48j9it@xprp9Yl>L;@
zVv-7SLCh54!+I~mNyR}5LRh8Egtt=n{HBD0D`Y0R6G~YUjn{h;Ox4MT(X{5X0kG`k
zg-8%}#T*cH2!j-6NZyYPoe5yFYdir7^ps<}MTSCF$_WJ1ll(&HJsikj9}4I)@)uxR
zmEa_hwE>gO0j>O4XZWD^4)qWr!!(dVHvFUjMp%hf;NEmdNO+ERI_R+m!P@$X-Wqh~
zy$&kC7DE&G;|MT+DWncA_{To#?YrFV{$T&0UdU|XKFI5hy1+(g2$Z%UCt5$a7_>n+
z0UrmnCDEbGkpIMqW4wN*8*#s?jDQieA~lpbM_16v51vCpaTbgujDi3rrjo;F!rvW|
zzwD0z^z<g>#FOZ1V+-gZw3-plH9&o=LSx?9Om0MxT1l!yIX*-1K0~5jL!w-$f)NoJ
z5yQWjIbVd00>Cne_RR<j38KZai84c2mE4Kt?&hVKQERRu!E7~}L#F#N9r5m9okTdq
zDLiTS+qO^h+1mi|pZuVcv{wI9E{l<o?f*M9{lDk3nE$(?laPs%>3;{Cm7}FwMcRh!
zyQ@F%-&d_dkOUJ$1a_a4ke<+%04Ypw*3V(+j+f-0r&J|3J(ZK&?!Ec<Xv?nR{8Xjl
zk#T$dO|7+-R{N;o$*ft?UYnrCPRSIfBuzG@m63p*X7jJ%*XmkWOpm7Zrh>nc_1(($
zcLnLA;UizEr2S1@Tjer$`k7i*>okDtoIFd?*+eGmX@f2E+QwxwBFdd6)Pp4TxSk6Z
z<whBnHGFz|Y2XsR<H5;dXWGMUQ3xlEv6Iz(mIYFy<8Fj!tiaVi|8m>PbQ>Vsk+K!k
z`(me>X;T~LUKKmtLinO4ewC$=w_SZfzCT!bNU|fM8!4qTN#1=;4e<Wn?Ut>~!thtM
zRM~8nlN?(j!n_$7#;ZDw0dd@IMt>Q(+^|r;8PV$MudlFtfUJ>`)Y042%1Nr1?qGG}
zSJkPg8UG`4Zj--3;qp?WxmnPL$9D-mUbJyE7}E9hIvdjHs<csHms4|HKdZ8SwUR+Q
z?Pifn@5(NgD}6j5tI>u$GS)PjvPtCq<&d+Z(?n#-E_SZ)0_CYp=e@?G{{x9*zt${w
zxW04&m-kcnm+KeTHA?T4m|@717kQ#u!OT6>RPksT%6oA>7s<0~F-s|LG}~plK905d
zHRbIAWZvV6sGD)ZyRlO{6|=kfGoEYttDnb-E&EEwj(HZ(x#7by%aaG^jnL-SZqC-*
zsuNY+>3FvI_dLrqm+j4?W#PK;6yE7&0e$@0<$P(1ZX93D*d=4h?+o6N=7FYgSsLb%
zTy4~4PSG-S|7+;qH6AQHUoHOszTV^g-*<cU`M&-w&NAqK?9MXu`o*rLUDXlzf7k6E
z?;afx@P9r`?!MUEj%vR=!UzDxs22kBu))GL-)RDZf1QG-=J|iz4svHuhGK?t+w%4$
z0AbW+&W3{XT*FGR*b%P3Ns`-!ROk)phLhK-2ZT}ZqMarTOxgrsZ$YFv%a}pX!jc#<
z(iYdFjF=x{Cght4TK=I1E(b^QOqS#~V3MjbBua;yeE||QSc2NRbtUT4aDtDz#72{3
zm3YNPwW{YFfL&Lzbbyz_Z9jaXLNfqXXpV&9s&|$-K5wE82oP`p2HsQ9BgKx!2*8Rk
z5LwvdBxfQ9b1s2Gl3}?w`C;0WN3cufwz6QAeJV$ooXe=o%CKpe7Asab5Te(aQ$HWr
zr7H@C`I!C~Eg1p?59Sfu@H|3JoXzVMRDePBU<Bv@7nOm)5IFc;V%zS{D&P=kTo5@L
znbjQbl&v}2cs~&VmTrKB3n*#*8x{x{UengVs2sw%oqKqQSg}R~sybvwjZ<%#(2^B{
zsz_nnIv3&;rEf-bltU`{2A?(nNW||=j<m}`hNiAX5PGtNVVyUiC#T^c0y<S)Sg@WF
z3ISqkc*6v{{ZKo~BlOiYU9G2c&soOj0zm5$m+u08+Fuhq9#LscQkrs;P{_!%B35;h
z9)kc2!H}4mE4T;#YU#Y$EP&?8T#U3LvO6_SMc(roEyVT`4K6{WQ0?igZz%ve#;70~
zFNGH|Dk-PhS7G%Q<#a&#sf?UIA@C<+KDe0loPskk2(?8dam+b6TaV87*Pp+K;}xi@
zKGNpKqrVu<9E^ZbSA44ej>Bg<%(ijD%qZ5UG-boB6gv=$z-dQ%hK!j9;U>qhSq<|r
zr-+yvQ~ZK9{kNbZu8e0hjc3|gCD@=gSBoq*zb5B&&&*)sa=HXjxaMqT*v5ybe7oz6
zrrTe9B9rriZSUA9EXV8YZla0>Co4uB=DIR3z!x9Wd8ESv!OieR!KpdJHhI}hncA{X
zt6-i48hG*tqnu&@3om!6Vnt&cPm*{EQa=Ip7l-0~v4-e*R|tsEj`!VRayFFb`60G~
z)j-glIbHfHi@<m~<i^oujnF44#e8$>oAihrbfUtH`M4yc0u7dCRphD7CAL!|TOdje
z8B^EQuZB2kZSUt(c*mOx-L|G|u{UfmL{QhRxEQQf_&5cs$^vRtlCGAeMHM-x##A;Y
zlrT;?@Cz*TB}ZxvQ>qt4mHpErsFa}nh)+K`6~&ZtC*o=gl|*9|#hXYgVhf>=Ov$DG
z$dPee3Nhd3D~OR<xEv1OBd-gQIpL%nLK?{gW6~>Gxyb`O3f%1OdV}kPQU+Ol>h;>d
zoVd%t45^y@n=WIX(F)U{Owx(9y^u>N<^(PxE<A2de#}6}(KkuoIY~Lr8w-=`eWwXz
zGoT}C@t>z+uGDiZYKC7e$=PsKn3|B>!RqvDl1_p=?Fv1ypg|;;RDc)}`<t^kPS05^
zOm)f;bjH1?^=nu}4%N#%K-{IZflxX#?!yBZg-$!?z{W^%q?eb(HwOxLp-<Ryy7TU3
z^zk^ZxlRwK5Z%6xq9&7D5Rai`U16kF&$Qy$ruhzJHcQJ&RJz5893!SCEAo9@6NbZe
z5T|?p2fXzys2qUWd1=AUOh#HCk2#v#Z7YVGPJE4(v9|D7?AV4cHfCllgLUCTVi@9_
zIsTB3X#=!n8&Sc){PDS~Dm<Ikyw!1I`p&$>K$?u!JijCt@F$ZpSCZ<DUJmUKDW)IQ
z?r+yKGmZ8#aZdE|D1Bld3NR%8IF%(nYQ}b-s}mqcWC4>aT=e}hk%7;ax@1-X+-Tn;
zkO&Yl-hElXP_bb@(`ip$6y}ZSFwF(i1lrjNtBvL?$N@SjwsFPm>xDA4*YF32py^QV
z%p&K9rV2$rIpXUc;Re(&pXx1l=wCCz!l~}OsK15!EHi;Y+%9K~F8wJP8amk2sr)q!
zSqikE6Q3plDh6o16qNM;x{XGXk(m`8ZEVSMo3XX~G0eMaBzYxE_9hCSG{&*%JG>(o
z-A9$KN|^nW2N8e7-WR8|*}F&nkUsJCH2l^L<9;2JH|kK;>F<louQimzEhqb4hIui=
zOMUXSH%vwS9*%v1!oITl$s7x4p0e(*^rZ7>hdY=YR@eLZ)AeuOv2|hZWbuhpd<*8V
zBKw|}mQKFJ({0k#;yZyvj_uD$<lA%QKzQ-<EMGnP=!n<S*37Yyl7Rj9K1nMxT^nk9
zUyDwJSi?(Vq0afg#@a4C|8_hqu~vVX#%dK>(j@i8_cT(0YzY+0!dwf|#+QbP0>e>9
zP34;yoAt8}^`w%{{nxbJp0E2e^_Q=3`Mh4A?G*+A|F`ji?T<sg!r8LA-mc4+u=md0
ztJNrOu=mD^W!neWul1}tI;zaU@%TLbuO|JQYkxZWcT6Tpif=8So~k;Y(_zqU6quqt
zqZ;ozuGPyf^3$T3Dh<Dnhe`k5+IZJ%nHuLoCNg>1`Pd3jHLrP%NoM6L>cORb4{Is<
zH)(GxDLa|@z>4TOCt4Sm)L@>3=sBL;!+xzW<!r}f^nS+1LfzSvn~fU&9+6;uQ4*aH
zVLmJ>{l}C0YUb_JS;)6VQGY9W+3i#Rp(y${R9;qBRCulk2;D-(QD)^Q@~bn#s|(T7
zH<eSb<Y$+J=O^O#%Kmtn8WKF2WzhB2IQxtRVdYaF!p9AR>X|3Q+6{o=U1dOuUrr?r
zb3$bu_4!-I$oBe%c1fE<eOFD*wCgNDa+~`;Iqpe_8Nt3y4N5HxPct&^<$821?oFel
z1vNm5=t~KDX{FG3)xBu<u7*ZZy$u-pVp6r2U`xW*+kZi^POL)?j9twOX8ScV#kv3y
z@QBv2ZcKBf_2gIQCB2-**D$cEcHW(W){F<!!9`EshNrA>6(2masw9FhIu*Pg_4+rQ
zY>P#7ReOHWqj>dmjl~tidrPs_b;Ub}%oW4e=t%1Xls<_|+SewuPBhJXo=o)QYiPn1
zW6~GH8D=6ps5Z_g50D*{#2e??jW1HHIK-1)lU~Ia$60dvSXrz?!M+_It7Kj&j)zGD
zWBHAUt_H&SKL+<q|BK+B<Ntu;5;C%L{_kRY-*ov2(stzG7uTq+fM1e#@<E0MzdcN+
z415ALAC3Cf=9t;|dh)1K?n0$9He6^riF1klh>ohxQ#TXiQjaV`dhZjhxwRULkaAJ+
zR36J~@JyinC;M0O9*Dqaml3Y=)VY=xs<gPP{q3BVzS@t9o@SThRvuE~vTwyx+zw;p
ztHu(gD>Mu)loChXOcr_se&cjg6KO!5g&6ebrXW=#)qwXu@4)w`OeoIF+f^_iZB;Ye
zE20=DK2mbNi1rtzrnPweAwH@fu>d@Z4ip+HB2UHmQ3rvyq!Vc<4}kPrV8Vjo7#Bd*
zwlg5b*y)o)*@e;W1*1f<OF|yy&P0jIi_5BbajDW~8U`9bxCnzZ?32tD?Q|A!3|uoz
zLiyd~BHI8Ntff<rm|7%IbNyzKfIo|ZXuqkoNO-@FivMo{;#W#H8a$Kg2Q0QpYL9>e
zL35!X#$fY2q4kmQJRy!bEz;I68WeyU?HnQdS+tXc)KiIt0@JxF3#IRMRA%v`3k&7a
z;(We>ym;^cFCH{;2?G)>hjZp3(sEB9Fv|o3Fl$2;9IpP^39A2JU}kh&a#`J$X!5DY
zZGjP)#Hi2WXx2YLBt#qz<&6wp+e&)GP^r1}KX;tW6_5hnoPs_fAdj2~O#K4(yt;vs
zgR6oDv%iq02z657>u!|b>u!_a`y;F>(c$ImYj76e@vQfGfAM1f{dVAs_kOV3^NEPa
zarwE&(Chn+C(m*Ec~;lkLon0K|NYKC_NM%LRz{e+JWLg3K6CWv`{=#xyXVFv!M_(F
zjHlKr_Koc#O`5-@;@=~^Rg~<qG1-V-R$fXon`)A9(SadalDVVX9ZzJQm-D4)7uzOU
z66jY`PE{q^U3_?FrIqq-<Z<P_xoV3prM2y>s_I*%mD%!ax4J9umYQnS4GZEvTAuN`
z@q@D}c|)a#N+-9Es;aGWiu~|#MfSdOHK%&fN#&A|Dr@fTwW7XiZ1`s7R9PW`vL=Fu
z>PclA<Oi)ueFw~VL&6ED-ZEr?r+YFdi*L8kh@REmnq1MGCj&v~R8a!1TaqjNzB6)%
zG!L|dHYy(TY&xf2SHn?0+0|XQ`nq_3W07b>?HQRz3eWR~Oqeedk3Mpyn^u<DsEk{h
zhuZK(3F)W|^YJ*@NNN#iskzqSqx9xnA!CW+D$B2^<bvtRvVSHMEE!+A7tv~(S@d!3
zhbKdZy5^Cl-9^`}u5O!%DI5V=T3c!tHdO@FcMeCcUv{~!8V5B}3+0VF?T(3^F!n!D
z+Wk4-;BbeOm&%!bO0UYyKw%WT@Y^q%?Q#bds;Ve})OvX_LyF>BbrM60m$W<SeUipB
zDn<V49e>*@O5C^Vl;_6@g9+#K`QbUUMeKR&v+>M|OET8lvp!vPR8vc%Vx-NDm1C<`
zo+?XFD|$1R#SwQ(V)}}RdHI>AkO~P*>;ZO4_b4l+?m?0zh2trcFv0$bT;ntOjOGM6
zqf=^16Nh>2EHy6Y?$W?ahG!8<!6Zx$9#<K63UlIW8VING!inXPhr`JvD>Zi+*^{T%
z;hbd{-_U5ghBwKH&bhNuWVYpzPPX%2tkj{8vb5t28E@4gRCo1Ulklup857DXrifcu
zJCB@nSrsorU;0Y<W%i2XnV;5r!R)hpWM1qJAci}4I_@PAi}NI<=|3cUg@sF3mevxU
zy!q(6LgbwKsIogx4&IG0Sr?s)E062w;v1Yw(dU_!|M_!S3OhAfWwC0gj;Zh}$hP8I
z@qTVuYXJ_gy!x#23r`FD_NixwS8m5i#m)j=B#lBF=km!7Dp$APC>gK)lIpr=(M<BC
zKK*@HigI4;of3bWR;<|eiOX|){1B9u`kb8?SJqCr6qv7iba1Y$y;@{xS=CuZR_;tG
zd*I6Q$&#<!>+57FOEl$vT69(83jm^I6EnYq@!TdR?ce&T70WqaS<|d9sn|@jliT9r
z`lDHjg_nB^;OxxCi+~#W&C2V%w5;iDntJ}xg_YNJaj^>+8&5$UQpH_h8mEJEVVXwQ
zon7|1I*GKh+9kOI##Y5iMf0rmBD}QOt;~tM^r)?*`8@L82a~qa;p5y*)olm7a<(_R
zxWS(Fv#WP$w#G`^y321~x<%Fqh|cy^ltEQDl#LqrQI~F6P`gw4S0s&v^V~CBW~>qF
z-M4=2#7@4rKc+>lyVDGtW6#j6nCsPacGJd>mb9icahaF=mxQCa%q?fBQ3@_#ba6Yz
z#K?qJyDB*p)`5OdT$PIIHHXjc!|2638W&8<CUH7&&+etta_DVOm5L{6be%ey3#1e2
z>s-dq7Q*ZzT`Wta<=SjcQzg>WI)=|u;kQxFQd8_4J3NKi#XC&rKKaWH>*)SxkNf{?
zLioR>(>OT(Pd1H_jg#$v)qLQaHe0E~4ma}hfOY~dRo%-MhC~@!D@9epkRXZ?xFF=D
z{q`@#jHlz?VkG*|;&>ykB+t~}8SHHy0hp_UcY`C~lL|73f3L3zkLOKrU!Du=$`c{_
zlPbKNe^0>*^LpFl(wUv}4Dy0Q&eVH?5$O_o6zjRUS5O)86o?RZO>N{jP-Bia05oE6
zz$EzBB~D|14_XmC#$6vo73?Ozf_ke{erKa9SQ(AC>|hFxHeng~RbQ!7L4g|iz91-F
z{r4#kqR3MKBI$h}gmY8=rolD-nt>tg5(JHhLGaUKAB1~9Y$O-JaLMM@!UZW0lC>%*
zovyyZBp-SVWFffT(*VVy8tOAxkhr-Y?$X~Kx(8{^0Qvp6qaOBc>RMkra0#AQGf)j^
zed75F2(fEyF=+7u6VXN+=&ZH(_i2zCVbKP=T7Cn0-FRz(r$O*~^ushp%YL<hTAm&F
z2Aj44xYNs-8etw^PPcI;m3H8Y&~q$0<uqy&2sjTTjY8iKF>Ic%N<6fh!+F@zflj%j
zWI<5vvbn}seG@?6nun471&_PXoi*b(7P&gz)`t>o7aZ<I7|v!-Y|Q|U3VMa%f_SRn
zV)|=kyq1<NU1$AGrp|YQO+#mWqyoUSZs2O#*uR}pv)8;iUXQ`|h`Te%=8+E3SKkzP
zb$@pd&fnAX`8s^jTJ-WBPC%Dr{kVBzum9!MiJx~d`a^W%%YNTpq<!ziApQH<F*L%2
z!*r=N==rF(=e|OwxT#zE77=TPMoBFy6vV=irf3^g!N{TGACyGxN=l02t9oP?5aXoC
zv<a&2DR|1R-V7>aMAlHdclbQVMwz4trOCsiZ<LdEFgK`@oj${F?=(r3YzEu;OS#Kc
z3wQ%v!7B7p?WWmhx0)W`KdjpC?h-fY^ZCE<Z>znZ-T2GwL;i8C$gib&s_@g6vUbGI
zPhmr4wy4Qx6`1s>A^!E4tV1dgZKJ~ETC+q}4%?-+w>6@)sR2T1AK+wXTwyyUn8xnH
zxp*X%7W^MmEuBoE2iM)0E0IZ~(S`Z3+2{;<)ebZlM`286-AWh6)BP}dvkr|jqq$Ic
z-7%Mj1^8g&wYh1C(dEA!aB+sh>^QjU!3V!Na><|anJCoq7ax3xd%EgBAa^?_x>Alq
zS(!WaG#+vR`WpEcO25+zE%h4yl$M0?9ZM|2`V!ld8%;E{R0<U==v8H(bWSwslwxL1
zRN$b`P;01?syGxaBu&u$EK|TU{9LVOXtWe!hp%v(v08<T_uLguN9NhHlzMQj-KkaS
zCXFt9M@K<bMy*;0w#yOKHfIA!3J~?jrNlJCkxSI|_?Fy!jpqSce<gmtDX=X=;=_Gc
zD*FPu$BegkE$c&X?Xo#0zkakax5qZZL>j0%7%q1oAS2y9-~z$n5-6ummC}ARS%<PU
z_+Q3EwuW|Q5DQ)L4S!1A&#>U;KXWrt*A|}hBZSd3nqwt~H4wqH`({Su(J8bLnjHF7
z0_5+cm{5&<vnt0!QW!zBU8K>IRAu1ICR2d^EszJRX!kqqoSIaOR*5P)KAEX>?P1)w
z@TvnkY7VC3OxUo`Ju81i>I}~oez^C+<uA1q8-I-}#KEuIrC1nq^lEApYJj4oOTz|1
z*r(y{-1Kw!^4IT6#P=|}R{rYea6|t!c9*74!~eW3?E7CjHhAaY+I1vAmp0&Jzm3|7
zSMJu+3I?Y>w2Xdp-d0k)f>9e$Mf1Z*c8e3j%i!{-8v0Cj-XTg)SXCWwxyh+gLHPKd
zl@lU$dvw*fVLI*tgasRCHF>z9lLSu3`dvL2e%Jw#?Ws|zbc&q^&VT8T%30Vv5nqXS
zmPR-H;HJY<s}Dv1;fv-Mm7Dyf7ls+^h^ED))Sf@!ZCCeGjdE#Y;M?J;*VSk;yTA#-
zd5|0jeS(@bZKME%LDiu6MZQ`t{Oaoqy2xIIwCYcqA{I7vv!bLoJF2$owf&1j+$y#C
zEcMY9{tUB<Ns*8ycqL5uM3|kti@8yi)PJ-_xADL1ZFlY}!>;Z}k(S|gr(L{Dssz>9
zU04^dQfeV|x93u1)c>PVjRiFS(I^xbQ(<)`-3k}7(;ts#-l25nG<O8AJL6KaaFqY!
z`u?0bl}rkzvAIzF$FPFvd4E1t0-X|E0k>|;pgpy;du!@%?&?p`j;H#_Q8==}@6pg#
z35KJ97ACj}D25Zkuo!eE2$Rz~QKkcL@^4Tod!MEM*%fm9Z%&rj{<}Yqkco+v{r@eK
z{wblhaWQpb5VJ9KF%>a2wl^{5<AZT_aWXZuh4I+%?Ut?^mT1gj@9o`(+lE+5n{o#U
z0vZex4>DNG{e{gE(z+G!<WNQS@5I}ehCP|vL!&3YQd!21i5crx!+yqzrhi=C1CRg4
z2WR}3^KpUyNnUa9$%kb3*VJ3`!yWT%p3_X;yuZ~1Kgsvfj^Epd<<SSl_A=t<>g(du
zca1gRv-yKL{-JCpZ+%0)#;ev(U%p%J1*7|7Nn@PfoNG~j<^93)OL@h&JteRDUCIdJ
z?S+=#>qOJF7vHf}<&ogc{e+0$eD(9q`c&VX6O}zDn4k5kI_(S9pXsyd)oKAH&Q*AY
z#WUn<iFv~F-|QRndBe4BCa=qU-rLFi@mtQ9ryFua@6#+xDS^`CJCFELtf=}qV`Ux1
z^fT@Seob*%;zyeqt=4OesQb(S*OU7qCZ+Rd>vwUJ-}NtkD$mz9?V)koZRPkp6|V<x
z(LELMdmigINn4rTs`IL6P2O<op!0+3-Sc<+(M)+;;x6^F-gkMqvs>&7B*Z?r4_<#K
z&W<{3we~5W`yIEhL#`F<EItdIvQ=Wfv2KD~W&*cA`z<Wp>bv7N=)s4as~f|7M;6O0
z_6Ondhw?-SbZ5*bey2BVpL&iu<L&R=yq~9M!_F6){x728dVW%Lqi?pKR99HJ@fv{U
zoew=m<+HqY*KYUh6R)@(?zcyuupMscn=W;l8c3BR`-UaP#?^t9Iq_Sc$JET3J<bSJ
zD{Na{aU9~+eb<H}DXHA7>uOW{i;NB07J9FlgR<5ogdG~WDqAytaecMT-!;8YJBdws
zMexceiXXU+5;@%-2+EuHl&;&0{u|@!e{TwU#TjLXmc3F=Qsac*i%K8mzwh<G9=^ZY
z{6BX6KMfJ__`fgszY*VP-s`@P82rC4{68R#2ELzOdcW9V(Y_9LKc2onM-fey{6DJv
zpFR^@cn{P4{~_6S)`c(WBhF*=;46LqBYJ%&I=Kn|^O^7vkI!_n%Dhs&&m1?)Z07Gd
zbYlMc+T@AsB*x{JOkel?FYo*L<@-L~8vFZ=;q6`KLiPK=|E=zAx)*Ej=jc~FpRSP6
zr|6ycZD{wV;>El9qIY+(XP24&EneX3?J4QbdikL6h2#5F{4;v%0JCLO^m6C>7Fa*z
zn>W~RzWapVrw`Nijp6zV(|7Zu2(p=B<*JvXO*h2aUjpzv-1>>R+8P~kg&3*P8?okx
z+@LE`Z7-t3|0W9XLav1v@WmRpr+bZ#KgaBi<2PINN^Ami^22Yzf42rcA~vjG;J*hs
zXJ|}`wmJGJ0?*b*`=?+`XC`Y%XU6-gB?D6jfXV_joULO>_P-PEfy#Oy_&W7&``RD#
zN&G#{?<H`)yZQMV;F8F^{2wkY9~c%&qnpouxXjka&G7S`b3wlLA=!GHXNc=^oyQS>
z6AvVrWr)*J@j<uI=KlO1pJnixv8VF&lbl6_^xa7e#_H6#`$Q-3b!GUs`*Fno@znc4
z@SWolyY%AdVz~Su0RnU-&+jY4e;sZp7&iaN&+6K*WcWL_O_U(MP4C!ZGy~h9%;Xav
zsK*|O=ek2WKnmzO{Q%1!MNI3mcf`B$O%u{MjyB|pqQ?1*JRa5x%6*w!f#h(jYUEu<
zMf2cO0_3--!JUP0*hpRoG^V)ki1+j-5Era@YvE`a*)5iwm!*`ZGv*e$A0NU>-pfHK
zj54Ymy)vL1%IxCBw3yddu_VoBiSMzL_VkaZU+#+sADCj(F-Q0{XMVhz=PqG+Ub2EA
zjqxM-BNVXqI_wx3tnd*}Qi2gG;pCP=JnwL|=LHcKK0MGFj+{;WEW9W|Yg%TM9kzI`
zWS#&l=gplHR(h0H;UFXLs53Cku0oPUs+oO)m<pDR;TAzdgg|=~-Q$Cz^-t*$-?NAH
zXtkEMye19u;0JWM%MK!veB_`Al%VcZU`xn5uy0g_346rCC{`$0g!5Z&p@#+!A;kP7
zcxMY#gJ9JY{%F8yW5duv%1d7u-R^;wF>3NTYf#6c^pvp;v87v+YK>>=x<Owh&JewK
zR7)x~s7SSTC)US9$Qx@YIgrjU39%}^EWt8?lZ{$A?+LL^ZsR4W<GH0Az*Bb4hum#(
zN!K*u;)_%>(H>i-&y~Ber^ro+mpAxejR6T}6~y!%&p0Yj#GLU$L+X3?>!J5+)BoQl
z|Ca=L*Z12lUq~da1>>vDrLl;xkXB6Agi&vXDfnrIu$a)5PiSt>NDHW_funwIdN6jb
zGom%-2^<00U6)#_Xn@W;xx&fh3Kh2e)g-6%-jlEYU@k=+mTFV-XYnj0NNQGJ5{QrX
z`gi8>QkbO7O)0@tNlnEp3I1xbT!`d%W}e@ewiw2*dS@2C2Ra$9MvDmFpW-bcU^f19
z0KxAWl`wqTo`j{qQIngi7wNvUqG(aiJ_F)8^UK&;{b-xaXA@Zb2LsF!Zvts9t7JG4
zspOivf|lrKu&3=I17=ArJmztvw(4ImkP8!}<4qEje1HDtWCq&Y(aD9oh0uCfB3%Vg
z1)TyLjd_EMDlcNJY?7@KF)jGnxRQ9Q!Ksl=70VBa;lZ>W=oK7>ticx0X;KI(rIBtR
zB3-!H_R;E+Otult$p0Gr#J@rLO&mhG{K!U5NZ)JDVH!cOaoXtb&9rD8dh`{hyeKdE
zQH59+%M?X$OLjqi9hh4z%fej~8&7Gq%lromY1_;=R(aPTGO#9T%`ZY&R%nw$+6}?J
zuw#_fC5jgtM0hKBWa!0pi#^*1)sahNNpBubiF1S(;Y=ExhV)qlJ<-|HH$K`YsUXuC
zy@1TA%5{eqm`>|m`XOzEi**qMr=&%`=V&a8!{hr?IN{V^3d0?Z)C@U??p$eIdF~(D
z9cgn3*`s@?SWB+Aj!&jD=7pq66#DFp+_+`DbOtk(i|hfnC$#e=PEy2@O932~)w!^=
zxs=t7BTHyA5h9l+NuimYYr!26UEvK2so3!4iHORd_ZQFG99GdQ$cCTYh_6Y5eL^Af
z5Bg{mKJg{7BIgX-3m}5I_@bD=ifSdDa?hG0&#Ra%Ii1WB=fV+xmxq{>n$q<z_p;dK
zD-R~-^-0=P7pY7sP^yMOW8yn+OhnCso)AY-5Z`0q#;4D`T9d6(*xQz}Jh#kWXJ^ku
zYU%ox)KcD(cR<0m4SnATANByam#^4DMBNJ?iktvY0itVF2pP8jRC~Kd*CgKpW;Wik
zM`nO!T|P)jGcTz@%WuB0ma|XqX@vmN2H3zF;=f&v6>RxK8llBOqt1J3cMD&AWBGyV
zN>da*k<@c(d^d(mSxbr-QposPnYTMd#FLi4D4LlC@U?XL*{@gJ_&Zl(i$ynPZR#6_
z(Ik+>i9qxv<OXyWt?90ANd6=a@g*5A2tkRG)HWmkVz=0Z_Lqo&={AGi@bNb|<c_m;
z3-~>C2s(?SVzsyMP-rD3mIta!a_Ab*E&x=h<)kBGprc*FV=`7n>hb-m9M)V)m#z&#
zRp5Zg0hjMBFL+=?SWOB&wL}Ld*1Z~GOLC8nTR;hFlk8k2ExBI7JGlV^Q)2iLfy@@~
z{>O7zGY8224f_r1=pXhf+~bum=<Q=Y(z;AO@RfwD^3b2+KeR19I{~8XCPS-#Au4Q1
z26);AR{u#rq2G1_%Ib9a7RUnu)s_$Jca?4(KvKrZkJDx#eFK^nu)j1i1+!p(;y8#5
zKCM!P47<fJF}d6azdo=wsSoIB3FpnO#N*9gOan6tIZIq%Y`)N=CC_-+*!*kizI~pW
zi~E}BA*|(cYpF0CESw};PCo8LNJ-%<bjgpX>=Zyx9s`x8Lp1Cfv#jv3o8M<oQ=o~?
zI5`rj2pp`p-UDVlC^^@dhV&2=OYvr2AMvw+HPd`d3N-LD*g~Lf)E%4l06e0FwdEEA
zAq?rpgpKtASy2Xv(V1^>LKt82O5!@+A_VFFdrtuYQwNm2rQUvurE8BlrD6=+sl)~x
zpGo9vZh=iCbnDfyV$^ypc4%cSj>+O#!NReUu~%HGiS9!DI|N_MKZQ<_10@UVC~p`g
zp3F2f5B+)^OG?9q8^^yGF%;0GL4qQ-A@J`Z6AiGW)Mc${f3)@3Zs&9N4`nI4zQ2g%
zNGUo(K1tzH@>@BNre4s(WT+gIR)L$=qk1&)JS%O<FVs^Tiw6#+&y#1od2o6WeChXI
zNY>FCSfj?sp(P%rnrL(nxzomWnCDt4y_%|NrkPpatwt1|VQ&*ik%1Iwya6P3Ypvco
z$3@-1D3HIxMDS;_6j@yFyhkMx?5dZ=0116&VDm_udQt!w?^Ey1uC2i#EXKg0^2Wf;
zL<}=ZZyg<Kh^239@W!A@&exxr8THt1brNtD!-LP85UMD`Nthan`^zIAl~+aHAYPY1
z3`Rk)JS@g>#y4(~<5F8q27v_&X?0|_Eg1rE11uminZ>jh?KY1!E05qJdI{xlBI`v(
zw4N>OBra6C4nprL=@fMSn(NQuQ|iz;a4vO)Zmdqr;%g!U!;GX(PZjcAK+;A}SrbC&
z8?#=8lo}B*YWa6=5tj3+T(!Y^^Y)FzpFz{5@b*}C-1VBMZd|Xvy?G2<ks~OblPu`<
z90~tecMnvLi<%Rf!PgoKlqN=WjRX|Izh)`E2=#Vn>-G>Pfy<eJKEOO4aoEXqpil}0
zm?~rp2`RJV8tmsLfJD}KlSoL`i`LteRy{ORZ!O(*uqTb%3%K#QWyS{NM7sk%{?We@
zN$-#jywBzp78{qmzgwjkKtcz#8-_ZtUgtp+3v5NHu7F+tvVwH&Ys^pMbG)9E>Fq_X
z$8<@KQA{Ps-^tV9sG2~TP5-^p<rD0+LyB!-YSM9BrW-Je(eD|HNDiO5`XOe-K&#ls
zUI<qihn$h#^$d5al)2{@@Ce-6RIN0|f@Io=@&YhPzqaIX8`cy+Dw?$&xqy(Q3?^x?
zvVAyk*g{Jh;0iE}jRCcECnt34@!u7ttDrg1ZlK)7I&+F26bKnJ?c4+9ON;C@Em^ZK
zOyGJhh_LmU`e|lMP$YR{b5Ij{pEWXZy>{$*Z_a@qlPaUzvK^9c3Xl=yIirS70nU>0
z)2A83D%UD}<W)MWZ%sehf!H52v?z9kQ8`83>{{<>yu^ja{k;Lc5d~hdZKsd9l#)9F
zwf~w_`S+orp=!h%a^oNgz0Ujxtf(fDbp|3WLU?bi>jJd*!}DA3Ui@)hEdltCBg7`3
zgtyM*3Vo%(BI+_1R_|&WndvS6_|#m@aHF3OTF<Db)xEuArWa=Xrh2JkCmERJQy@5v
zm=?nBCrHK;Izx%d9T+EYw96`Q9ECjUBK?RO#t_6So;Rx0M}2MN@1vK@FuJuPz;`MM
z!~WmT6GYAdI#xqN6n!8bVhM@bb~6Q!5f%sT#37F165nTlNei5(MNC@16VS~(S>)V&
z5VKG~k{_|rwOYA1fwhhR7<(K`=^v%0;>=B{Wx(ebYCHwd$eg0fw$54~?uOoPjvd=(
zUEVgaw2w1D{A27DYt?XKcq4ja53v@`;Gftac35Vg3uErj_2Xhd1b(}Q%F%;gaC5pb
zM44|R1tOBfCd>IhT1K6cE6uv@`#I-bfqe`v`2`IV3l3B7owvCO6|jYsB+3>$rXsu-
zoJX5^!Fr)%z*i&%Slb*vW;W8v0ZYnE^1_20rT&4N<-SSMT?9X5oyR6A_NmTChldO)
z%2mTs(2M#A0sW^rgD#|%H%)Ptd<YnVKyWErL&HLlY!+DXm7t68rp~`(5&cITcT+;D
zyZF@lQc{JaHmPj_U?OLv<#|*jB`9g%XWqcbvVnS)M`Tb24b}V4FfrofTwxcHzK|TC
zn5_dsjKl^ElmbF%=Ij@+V^V{l4nqWlgGk2OX+0Lr3Q*frkf=IXue2)wg&#RpGg_X)
zrXh*Uz~3!)B=Nw|`RKlPBS$xWfssx#$G`Q-CaqN}mu}yB+$u`x>=HQ^Flri;GoMLh
zb;Nr=GBqTg!KB6nG*w?qz|+J;hhV3rqq%buSB&DMb3r=q=G(G1lsfJaKu2oZKZHz;
zn2oZCHQ2RyI&L*+EWDz8SLG0^opNm1IZA=#VhN9Rf*_^Uqx5w6)A<H<O0Og<h_^Z_
zj4ySLlz`Fj*+<$a0^q|jS_e~#TVgB=LQ&5?JVZ`*!UNDPq-~IR8S$v?w2dW0qKzGX
zN5%!=viZMVB`iTAKdpf#34NsbE^K&vcNxt~D8V3D`n6y^;(8YuGf55}sE`CB{(v40
z0Y;&L-#-_slybKB8ahV7X<&iVgXco#q+7Yqzdi$u%hrqwbslsIPHUK>N!qHFVRS{T
zhWEP;-=R?&LxJ*E28555Dz$ZziGDjs!Qx{PV?AYSFQDaeMR=fg{qnG}A-2QD%r?bN
zD_w08IT8!8^(JdEV=Tz88VZ;as5X@@oGxl1nU6k>P(rSI!?9f|yh(-fS$7S{unFIK
zCP%Mcj}*szLt@O${UQ@{>3zL8+%%w+pY`(>*c%s%P(Jl4+)B1hbGdQi)s;?#B45VH
zQNzXW%*(q&DrmroY&vRgtJ&#gFc=Qwn*c}~68S#3#6Rp&L~|?Y=N4w?IcBLV-acQH
zK@rK58v73th6heXhHckTEDKMh_ZdoWbY|e)vhv*`l6v=CW%Jr(Ue|@^v1pP><64(-
zLNI)7YFFpAUJdyG)_?>e?tJ(Ok)M~p3l(8Amx(P7{Py7&6n;sDFLrVY>Ghfs&J*se
zP@8ZaY&PYc&8$%VT?hM@_Da0i*f9R^%BqKSM^_-%(pDpRuXzKFgqawE?_(txymxBQ
zSrTzdGD3Xm^tWt@wx4E`jh2I5(Y5+dpF)H9%g*5-k%DCA!sQG9m9pHKUti>Q#H^-=
zKs;5P{*JZ;2CL*9iqDX0BxgpXJw1U(d9!%LUjs6D<~=q9=nRv6XRlH?7)6VcAsYrb
zBd>&6V6WjVhROA|qjb^>#|@ZNpH-951(lbZEa7^{iFA$@VD)uQStg(HcB0$zoIei(
z-23g_L*ri}dlq@)Nc;%lajmeTg;YekA{8JmvRA=Uy>^rc8ZKc^SOduq2uueMS^H#`
zXux~}(XF?e&rgZIM5-iWOGb@!U^<@Kt3f|qU=e@e;#ip*e;<AZL}D+I%gS{#{T{XU
zh&=#v-avO!xzHpi$a{g0>>C6IMUPRL!a=6Lx(lM9N^E3NL_u=QJvbxvOoqPc^@BoH
zxs#k0VlL;cQkIPC+GSX_9&AcTH(rWucZR2c%Z!^?)*c@GTj93DZ=Sq2x1^66m>`<(
za&4=U{O|=^3OZL|=HFjq0oF9+j{fji-3kPMvcW6Ery`%4fdziN*Xr9)oG**Zn9~8|
zJ5o}1eFhH*_rSfD9DvZ;!TV)|+k^%gXCWn()YkXqJ3>*HQyS~-7DVc{a79;IBRiJZ
z%mpXILbelX`(f3d^n1fZ2uz~W3#;L!W{ge~tO?IGNk-G6Y8#C5ugm_$z$BokkN9&d
zjNlxPj;@(H|BpHfC_<Gru1>HpzPQui`kL*|4Y*oRnFu5emZAZ9UDx0zuqdxYOaZ@x
zD+}|#EsG|U#bnP6RC1FMpFVj2INeAPOMprBUo2FSgjNQKYOp!YeZ#qV2ZyCiRh`IG
zYYhbz6&bK3TGLir=`OJmq0z2P6Vktf)cV?hGaSjFBnrWW`F6NbZIq&s`Ykd`QaBb+
zJ|-%wmw#9b<Q!EFx~w5!jh^wFv+_adMSv5XxH)g|;8~8}r=+#BNoqoyU)@;JXw}6i
z6JZ$19H_@;N*}^lpBwL6B=maF_E|$-7NOyppUdH~=)SLIaGt2(Kf;_yh}$JXbs^4?
zU7^$Q*xF411NFu4VNguDv`uJbOMN>XN~cH)xNq3IoV}v1jWabl76}U@T`+tR?H&zH
z^&0&eBJ<r3NDSTV9U+4rvqj_Hb7UswKK}&bDR}cR@Uo0Sx0Aq`Lv9<=n_f$^fJurR
zL*G6qMnYo2jg$dD5y0gh4Z5d3%a^0qs2kad+qr&)#}O`U{=F-(jwr7F!VxFj@0kzv
zOCRrBu2&hqs{0?v1bi7VzTj59=cY~}7?v1`j2&7Pr;f?$8;fVwNQTTWVrC?Xzn0gI
z-eK4c(~%xnD@&WnHI7KEJ>Uu64>u=x&DO=cGoUi+fusoa2T&Uat*q40q9!a@3Q}de
zH=`s1O4elN%GR$2DXsQ>7r>cBYf2|UDDHfjf58LAf1xs2Mn)1^c<(B~1cMxt%16i7
z4~(9;`81=>T9OICu_haXGeyjJ<1#Xg2&2Z?Fs|9N*Ns4C74IRWDa=0G6hB-5r`QPS
z&~pTo@;i3QqAbjfAtZ9?6)Zjiqj-ahKsjh}LTgOAP_No9Cyf$?xyu<ySv1=aHc;fl
zSJs597Bi%yD!|xn4W@IMh+`@#g{m*!+wr1G>Y4Vx-w-Gj2kymC!10=f7DYdDKBL`Q
zCEjX6XQG?8+?RF_W#6hs@;0%iWE^(YD8(i>L>(p4&RtpYwW<mgW(Nd~aWqgaz{C@<
zV@f*_v9L&@G&~llFFB5O6&o;4MY6MW5%6lF_$xAUkIagb7RhuDCwOfk21r68IL^oS
z>FZ?_WyO!f5n_kz)v>c^e{sP7{75cUElK*FP)6@ZL)-KfftE~us}zcj6Eto&CSCgY
zGDmXkFQv!(l7%e$3KJ8Gm?g+0DB|gj2>S2`7HfoPkh#m|T_wvj@|%$TjJ&rKcfmEF
zQ0fdHM+a`?yk5U`3YvQUjo`LLK~a0DrN%Uut4E@2S0Uv@R&V>@xiv`0>4XaA{`)@X
zR~CDAUoiK0w<s*DkL2XL%Jm(s^e;#LaUVUi?Hq1dS26V2pbPQ}j&R#lzn@@5g)B{8
zxY*haQg1Xjzgli8(%6J~<M(I$<}0C1h;NQfx+0W9HEmn%z};ToOQ1vpGnlF8q}C1H
zBkh8Uk?+yJGVS!i_P;o`sc|oT1vdxT-cd=UIfuqELP^L-sY3|Yj!w%cV}T~<+;d2u
z-aJK9VVG_CY0HAXToQU@1MfF^jYH|a_3KbLT%rWi0P}o0QW(^oRe+e<${t{J{Iv-*
z6zEptY&C1hIX7t!=n0i!M{|_P_uUt?WF^S;usk8S>v2yMxImMb$hnZZzeOU`$NAU6
z;gjToV!Se5y6E85U9*n$PH~85n)BY;fsu}kC$$)m>=<%kf;TS4;t6WnzL6x&b!MC4
zl4J;1BSCP}wq@F(`VI{tC54{MA4ObJ%{tN)Z<vq;gW@pZCI*FWQ&;z6v#@>#CLyTw
z>Z?9_gdsUC`BF78sUmH)-QKz}4Cr>rkj`I=T#y$%1yW3~O-O@i6>2MnlP|E?z}m!2
z<8n|IJUtdrI#$0$2ke$0ENNWb`2)&XGx!Yr0;tN->i*(W^q24P=WLIxo-su}vg$eB
z?wPM*<{^}BwfE+naw9+hHv|`!TOK%W&g9I-V>K4>Vp8nEcH6^1i7JxKVrMSOm+^kS
z;^!6XNpI}^13|4$-~I#&baK*S78+mHcU!h+Z<N&nrrsW~8$F3LY@j1>v1HyF1Qy<3
zZq9RHPGplO?hU^sK2up)^9Xu|htBJF8G<9|({P1*0WA(PSw;Rd$&)olY}3zOVRyX!
zx(LT@*hA=`pbo-}c?|Hulyb>d169nAmAb9-JHKD>4Q?$9_WdP%heacM+%*b4oPZo2
z&sakZ>eT%L&!I7vV6hZ3&)yGnhQA6I>fVUMj&xlQY1UakMnTFW-(cuW9n?F&(n@oY
z9fT(cUN8g)4ZioloBK)}EEEPxV`b`R2lJ>rA~*`0AUEq^O6JE4^l(ZDJ|x$y#mAR9
zEt-=~wFCdR%`7RqzX6jH`ltEOp)MW%H12oLfi5w*1J}Vh!4kEr@b`^l(!(QGggVvK
zgU9B9$1deDQJytMnxVo)BPJnmKRqo1Bpiw^o^AU&)|W%E5X`tQUTQHQBLM6vSPE3p
zNNplx#1`*h&*VBt1$@|zUXZ!fb91~~diynK$tFeqYM+kgN~0J=iBS5vV)@q-TGk}F
zo*;Zuu!9nnqB_2Jr~O#54Z$Iz`UE?j^vEjcNMm)xCsaOO^lZF$9BPUJ{M6Wi!xT)_
zt21b3VM3Ki5~w5a3FTCdh#`lZPiA4gU}1Zp^@d3EXDmA;MiYSJva+Kz;zfkNX3WDt
z&{?P69!Kez+5r_W_?tCN)vie~wuh-TNF%0>lrjD{NEe^Uc$LIYhEF&hu80UkoHW$%
z1jjq~1x_-qkEdl-p|j=}_A{>>evBWDrWOuYqDw6f>tDrF76i$qI>RJxD<=vpIHfae
zJGPIKE;y10dNC}u5s~AF?&MID5-;-&)O+Mcyrr##c$b9f<eN2GPo-(1WZ;3znc*|{
zL(2r%Qh2!+?1h?AJTmvxh%Z_IZd&KS49Y?4XKV#3Cu8<p1sn00o0R^+5U~J-O~|1W
zZe_-cN2yKRf;g>!FDz_Tt7~oj%NgQa14E+ws8qq4JmG8iUImjzhVeSvrm~O|c#8?G
zzrh2FhLth*1K14CK%o(#BEd1wqnk32K2t|hHf2=Uqw{(}mL!?bu>+T%vj!A?P1fj%
zAwedkb(OQJ3jStv!SuT1wY%mYj=khPCto}ZAy%Yu1&eC3@=**FEsh3q8;^|gwGcVe
z&SNn>SyQhROQoS#ijIrRc3?Bf!0^2KgEHQEOzVOi5elp0p&Sd}u4Z-I5Fd-i;Nr5-
zHOex9=oD#03iZ8&Z3vO<vg!SHUBKbByZOFhcFW=syfei&h1AMjfw70$s^VY1ix&ry
zvqFw7n&8u9)^8%|!h{478>eo`4lQ&Z8kB!z?0W9sn{`D?`M6?*>QWl#IVOOVM5iNT
zgOg;1fM^E=Z=+Q*wrf^9W7Pv=lu}MbK40#%#WSek0IS2=8hg2r3+!<H9P%xo{*k+X
z`n85VkUNEO{^t(u?AQOS0mvziI^s+x8c15&mv1(QF{<y+<ncFoZDpU&fkKDS27gQ@
zUfu`$l+AweD)NjjY|8bvaICN`x-ZgMR?s#-nsX##&5xPnp^+{vii%^p$rdBAT8$;D
zWsris57vfjhnzX@-H}?4Cn?EnKLh@G5CfGtkLAFT$F%y#k2|6Xt!R<qQ-p8RcSuDb
z`4|ntq$c=ZuoTB<ORrNss+5pV?Vi;icHIgj)d<NEw_L0R460%a6j|ejHQ*AX=68hX
zx$8-n*}OU+5Hydh8|m&o&*C8Xh98UkYbUhrZUjxTUY?LxQb*OcpHv2nI;v8n=_$L4
zn=u$2E)1>}LB@g(AwyoyXZtma&vO^UC^z${-7D<e)gA`;Il^OH@c*8YbZHN3Q7fnI
zK1kr6Z|uMeQK~edLaios-g9VfxbkpuQi9dXlk#i@s#tVteJhqweh=;l*w2r9=bviN
z<KxnEJ6Xa#9Z^O8OhB!B|GXMSWK_&ClWX*K_AAo89IgGl1K_hsCM}Wb(hKCwws5rG
zz_>`!Yq4rQ>Hm|lXj*@gN9_=cP8<wv^-hPeR0wpzKk|a@zFI-INEP8i{~dq2t@De2
z)!cRog3ZPRgk>E-c(Wxlbr1rCBvkga<@H*dE=jzm8Nk+^;OBLSe%!moX%{0j|3`Kw
zWN{O4>QqJwr&|vS4Z`-Kp5(dHD}IcL<FQou((vbgSTzYm!wx=BDPgZl;p1LaU?_QD
zQ;-eeQ@hc1RrFOZD$ETA{5^nGq<;(ARm^6+{bDHt?NDJCymn2vaZH5wH?OAeapn3;
zXNyhDgiI+9nuS}DPF|>v&PR(091;kCsU(g!SE7kpMUq3^=O|Lwo`Os6=qPZQP|rCl
zz&TtsVrvstt_|(dfEy{~>X~D&54xIcEc+P7O9$NcA3qEgC8TG+3g_)I10bvX9bz)F
z4OiLH*E6BEW%{tZ3pEfS>Q5?(>-DZMBNF0TVUF9b!jUgLdlYov9w?Qn`qSqOUW2j1
zGxfGnL_~Xis!xKohk||iHUoSR!(2O{UEhqc`e#atqr-mKWL2<GrwU1>R#4Kk5MguY
zZ5GmV(ZaRsZtoY@1gtEKVmLx2JzYahzUj=xXZg&I8oZ<~)sZhOGs|o~Tz^K8$tJSs
z8bNL`AY|7WMT3mz|6%McfZ}R`w$Ts*B*7Dcy9IaG5Ik6LcXxLW!C@B(?(XjH?y~q|
zi_79}|GwW>^;g}hd#i5Mo<1|(eRkyZ^vpRs`?Teo`EsQ5KDL%c+#ucO`h+rY)kSRm
z1Yhu*+Ly7G=F=z?E4U~F(}UyrdO~F!yWvR_`NU86*o6q5t@jazcq3S^7;N(rx*F{I
zeIHma{JpquFD)_Z7XA1}nzjk_7*hgNh7x`UP>|z6YJ8&GlxNX^82r>-t8n}r5>D5U
z703SHZOTWFY_mEle_Pz6zlu^UN@M)=I{15TENR7_Gxv?*A7SBd=H-2Ny}o{jBEhV`
z*9tofy70H>E+gq})}+8X!8izhp+YbRXUd#c48Pg<l~bfSc_@-WQ3l#hAKTyjsx83%
zPei!l#G7AOCg`ijXe-usP6(0(EgN!6@g}-!nkG|B+zaB^O1=ZZ9wq4#H3FBbAStRM
zN*dlnyG*5YEML2`fI*}jQGE0nz8!|KJpTlZO<hr`ZK{OBjf<?`)IQPpQ5NjeV|i5!
z#L~b>3N{J6R}&v#n76v3F0X+s*j6^iEcvSPAODl8u_zAiTKjiQ<|ExeG4ff?o1fqR
zNf**-{+e3>628#T=y5+E14;h;bXU<L{VfIyDCR(SzRr-K6)EPn_yfuB5&5L!=ldl%
zze00byp|z4+)f(F)(}eO;NE?<)$ZWM(L>E`D0^sfpKll8=XSrfcjnRKLs8Byy=<O~
zi?L~1>Anj(I*FIFLOA&Oop^0wN_JYZJU@SPV-%P~yegvQydw}1OI>|!(5_f2@@mdC
zDmVlRp6szQq?Q*YDAtq>NN4QJkPj5om(xEx`p=8pGPWWO2Z7phZx+=EekjIWnLtUU
z(*42A?z+@05sq-$(l(6=XO4CN3a_?N6H?4w?@5i;M2R&-S6R|IJ!|u}lM0@%lX`9z
zb(9VY(T#@^R*l;0)1lG6#uE)P)`vP`Q|7jpUyBTd)stU;rdD2b7s!0YB>cR7c<AJj
z)MpUZ?M5}fRw(jl3KwJQJdIi&$1m)d+p^BA+jmdw_)bOVuVs1nV*4Mq0oq+GTv_t!
zWis?5ja!zLKd(|m-fq&}jO~%uRkC05dEWn<a=M`R1sD?WK)U6em{_>(&UJYIOeW-q
z^rB5#Z;gNl9eQ(99YRxerv!B<GfRZXV)NPvFjc`gnyfd6N1UI=F;O4?jR>9q7hhf<
zk7>S3aR^!v#(t#2sL%qRGGXZqr_7jBD~IfGx9L-TzrCiRW=d8ykA*V4)0|&5=ZP<r
z<zzRs=UzT{YZLJ2nOrV&`}%IDECFhI0wK%kLj#DjO3*?T<1W8pQ0*h4m}LN`JPQ&}
z%zg(az(ODXv!3zFX2yAu{EZ2rDAOYWc-S}+P<%=_?!)26@1BWE5;iFKrT-L*;%oj6
z-A3(8o0{oT0~sxg;Gswb>ahxWy*7AR`eX4rk!UtC2?>M_sYOp=@I(80&xcd^OYRtI
zvCou;vku{05|eGqTYi^u$SIF@D@f^Y*FMb_i=_nh(;wT(D<g@#lN`lMGFQ`nYf1ZU
z)ufi`q&ARx0c{d{LqS)g+{@Z%g=9}rdT>a#$ZAc0kX^Bw0-cvhsk=OabmN&Rwbo){
zD=BUj!`ml(>F=uO0Yd!u4!UXfW}W3`EHpx{rK$%aTTs5{&^rflBSJR@1B~)pwY!z-
z__|e;KNRuH&^23>&+C33zReha`^EYBE3^2^&M5#=Cw2fF9kOp))+|w7gpB;GgyIlu
zXpjWsZ{jCQx6ll!<CYMl45UxcnyW%Mbj(*l9oiFukMC?UKthbdX}}J<r$v3?egvXr
zv}Y;O!R0lEx|^|JRUI^Ki(g9t5h?BG^<#<19q_uc(H9XR`VFYs$g=y7R!a>|pohO4
z#;Y*M+u^F2V=$}vNa=B#tPYqySHeBGCV#?h>RTypIQ^v~U9eOi=UCRz+!cJa>d_G{
zq?%`a!n(OSc9@uwJ0+d#50>*A{)QLzZgTHWoq#b5p~?%9Q~A8QKgV=xLKlj6kPF@|
z+w8=}iZ3F>>VJC%O2St*`u2iFj|vh5WxE5)9ZNxuzKsd@juQZ#LRznF_Sm??GVAr=
zK3el>Kfa6^+Du{@WLt!YX)F0v$60z+g!SJ~r*4x9`*x}(FFHgxw-8Hd+Qb_y^jyL*
z&K7@#W7Tn`OOq*8y8zLL$Co3JxYGL<gvDzMNpIA-p#W@!|KOOftS|H-94WRO@%^$6
z+1a|<5%{<`9APehzh{^UxW#fv&l{CPL!$em)cv;Q4@eIcoC<$$$dN)FQasxTzi7E3
z&2G;mE{n~)ce<U$H2U!O9U}H;uhwJ~UGFtWziEkkwr`B~ucJM|weIZ?_BCC9y#c$U
z%l1Wv7@45Me(JjqMEj<~WLCx1nvJwVqemlxTA7{D$3+rJOGVmU3V(P=%wv6Hgak9I
z9<Ng1XNS8q9XYg1PZ8G=ZMLUK=i2F+vcH>4&XW4^o7S003&iq%cHYUeUA*bG>P4t?
ztzQ}FxF}}c`D(yX6S7jTC18DK(hww<&ObbyTcFj?ocX0-Sld?gL|(hAcG-R7O}7+q
ztkP_fsvjuzgb+1N;^t+nYNmB<Sg=F?wW2k_S#r3(HFfhs_jatrkAfmE6RK7hN7Y10
z{>mDkmc&cG#Wq|)9?&#bqF?sKq0+&4r46Au2p*t^97IXo7Q=L-I-*7PIQ`ox-=O?8
zZE@}byZgL3fjvXqVb(_fgCKFiqtrm*YkWLEjnA_8C-{%>p;G7dQckKP1I+UHHKD;I
zb6ov~6@r}mzh^HV^U2iM=hv5abYkHoo!~)Mo}lF{{_uad3SJzn=(Oi9e$(5IyVu#w
z0RkR&pSUY;rP>Vo<<1LNCw>$BEp}1paYjei3gtt(%0>pXZRiDrhOqyw{A!tBME^vn
z*RqEe(LYTB?GZDLl0mqgr*)$O7m^mz8r8pBKM9%Rggoh)oK14!Yqw1M1^g0pLM(B!
zn5<dKTAN+|qNEYsXplv0B^VLDcnWi_p~(&)CYI=ONQ&^T$7-E=5kG;&kl`d&Eh-8!
z)9i#{Bvu<&8+&|%DA;-tkxXf|B+udHo^GKv*-Gh5K1It-WmwG>3<|?XbP5m|sc~}&
zEBs-NCG#i%MhO1IOK+gV#q1W1agiN^hO4DcFzU99ZbRCUPP~ZVKhBbzx+s~)vD@7?
z8ERu!X&AFwWmrlgy@siO4%w!zO49sg|E2XO4E>*CmO%t-7oSPYXINI@n#>i;?SzMp
zQ8;kqezFx#+O6<3)R0B+I-xCtCeA*7tKIvgk<+oZ=-tvtBI@fHY=2PhJ`$r#tL~?h
zBx~F(!^TWELIZ}j10zYkfK_GIPo0~ix*e=lMUa7lKs?@)e?>Q-qEj!v$7BEr<))%n
zL$Ym>A^(ik5#r`=NjV;J^{QrcM$O?*f?z75_Nbk8Tl@o+pLFWqi!d^6k)?ILiZJWG
zP_<@o4EOeRhFB*rxbRfruYdpinnCNt8^cnl%k$~bp{d%XWXjr>DbK>LeRojE_#0^A
z#W{<eR4BLv>4i=Wd|WeBqlWLf02OU+*M8dn7XL}I@SyZtrC5To_Yi)yQ?oiL^#ChC
zVR1h6;-Ym!-=AprgG+!DH0MXa;6<=!f^BYphJox-<ex+18x=v=+D?wVb5lMn_55Kn
z!FrOTK!HL(27BA)I>SLa=IiheWzH4mW`Dz}E8>N?gDx;wzOl7!u*RHvI8)bUCFAMD
zPN|KncSuqSKzRQ{in0zT-EVJ^$FD>r^P_I8dMQ<fl~AAjXo$-PQlI_=>y<OZWmLN;
zg$?Es3~llKOuhoS3_^T{Oo|E)DLmxG5YoHiN~>RVQb$E?JL@OKE=sv0jR!1#wSwLk
zlnE=7Pv3WvhmL%mPqEvC@kbUc)T}1MM{MvUj!YfHlAkGB0r8?<J*0f&SA5kOk*8{z
zql4tpQ{pl>T6ok+pnH{|`cQhgnf`DM1+TF0o!mh-KaQp@ei}2RkDTr_rIY@Y;KzKt
ze-C;*s(0U-k|5F@AjKedO!0L4(0pdNGqv@_%?cMx_PErWQE&alym}f_q8o0!QBbx)
z-*Ll?WunMX;Q_y~V^MmXU*}E8y^HwQ%z=D&i*bq?8a;zkiztEM8@rHI%9*jxBNV{6
zMVW*gobs&>0d8Dkxx+)pXBM}C>P|!&GP`RcIM+NR;{SfyZgwLxl%d`3W|2f#?87Ue
zSEXrF(v)pYf2Bmc)Fjk3=tiSB$Ka6HwpD&s>NDd~kv4Cxllz@i+yq;O_=v8P4Uxim
z<%MIJt=4xZzYPBJyclK$CR9$$YM-hjE;&$VENn5K+yf9uos;uEYhWDXEC*fhIBo9`
z14HXnZu<@ee*Q=)wm!|8?$N{bykE)PDN{_4J@63+RXMgZRmK~)WtPFsElnqdQDhja
zZEnx;AW9y!5AxDsRP-k;%8ULfIJMUKxJSwxq5?VM8uU%laj_bT0sQ+IJAFOE#(P(-
z-M2mZAv#K-`A|tFBMxPsbuK5uYBYh!3%a5FSp>06+bF2xl&@t$LL)!UYdjp`^sA_B
zn^q1au>YHC0KJcke^ZA`awJ`-G)wFs4DV>BCiLUnCDIv@<a+jLDslKl7S*0}@?HNg
zb)Z^j;H2I{3hsSJtvB7o>0P#KD!<Ei>ki2r7(7c{XcF_4SlTk%GtF_f;p?Gb1ew&t
z@7q8cXrZ$y_SQt4$d5o&)D)UYc81Oc@krdJ3A*&2KeTPtr#IJp9jK9!!RJPqWBxwA
z9DE?I%V>gc?IWn9y1K5>dG?1NApMHxVKOCv&~%U0vnpYL(;qFDQRN8sRcSfB?=&aN
ziYg4O0a;5G8W|p5jvoa1(cU#ireloxYayjU4O^9zwdSSoUUn{QMFP|VKFelFiB1w%
z3+s3CeBw_3l5dF4_HFQUFt{tz#3~}?$54vFy_J@qTn0+ZFXM<`%~qTbWECYRj%yF~
zH{Z!JRS2%;8@L4TcYcKRe<Jby28+&!J16AsH9w2h!E<|-2)8&};&)=fK02Z~RCIHA
zb1fpUK2m1us_xQqxhn+Kz(OMY#6SCU4mE$&j}Oi9$amC*uOamAf{xM`x1S8-gwv_t
z*qzFd@Ql~9c9P%KRd3zs1DPmBsiLvW)YZ4E&e02LdiN5xqv%Y0`rMkY5uXS@(`_zM
znYx!7JsK-yUZb2xt#USVPCH$nF3#OQPS>V-YjS^hUVEuI<c?GK{l34H4xLYKy7Mgm
z#Xc#`OQbV_hN_RYu=OGMSNzB*;HHPg_j|VU$?qZ&vSoZbq0<j>jT_OzU7Sg;bF@yJ
znj2KsLd^_S<DO&x6ue$d6H%IYigv%ID+$X{k0JaB<p&4ABjbq`xjJp&QCpy$ergwa
zwEF&#ML+xR_Wp4FBZnxN_}=sNm2bXIhuy=XzN8|n)Mj25UHHmhE;Z(al)LAp>0%@9
zEIOzl&EMF&a~EqJatF`yN*z=C#hTRqZJlJoG37Erc6G|gKgAH2uZq24q2fNPG)|mp
zkA%B3Jodf5MzDV~oXh8P__lr-(<NZ`(;HX_(@DcA(#D->y41qC<um~HO$@6w!8^b4
zLC8qUk04f3NRDn}oUHC4Grc8ta&Atmrp}DV@C<O(F8(Ex$7DD+>~@lvnj%=fZH}1}
zA;<gjagcXkzVKQkBB{F3--&ReobbPp@DjR36rHNG)+{k-G$%mB)k@r3Lg=Z_9#gT)
zCucjw2s84)9^=7WxKFp0c5!pv(FV0{LI>@dhtTzj`6yn~%-N7YYwG1)U&eAJLSVwg
z2*LHg6Z&4)iRkN4JD0<c>!Ylz!qk01Sn#z=(z9>ZUc=sm{P8qjw>sj4553i0CiD9(
zMUM8X@uIcO{FRRCq`ns)4{`F(St}6jzvtYrX#9z_>+okyoGTRu6V5>vG}7R<x5ddV
zDwYmTXH_14=lFhYATnPF9`CWm$^@liXptoisV*D6NV>&!$Is1`e|x(e+lK2-g&8Q(
zSM~IKGlc#SV35<+*WWS9F!w8cxyrqzp>x6Ia+qhK#rP?0HXzYuO2T0lGyIz+u6MfM
z<2>ygLt%ldN8hAHZp^zZw0l?5AvNo8nd2FzuK>Pk$@?Zly)K)BKsLGicRak&e|A7t
zr(eK-xuqnrWE^dRR`H`w`B|uiN7!;o)1AEl!SC;;kc;qRRH-_1s1oiTphH9ThaZ*#
zD43-6`7YRHyVRX>@5@s@P<Z#bag`OJ_{E*FuNfDl6x~)GmMWeQjM#DdwNAI{_Q@lr
z@{CKTSlKDNCmGAWujS#w$hsDhE_1&Ot&GpOE3x@<L!(qb0J`G|d~Y5ze#jAZlv;Is
zDwO@HvM8EtXrtVvls+%~Dw0PX`G?Y)|Eg@-9m`9Al%I@?H2xTEyjv+_rc-l@jfK}o
zG!oc4kif5!9PJUzm*TbT?5S?+gqW|&Dvpm>R%H8Zj6O${{K0LK_}gI}I`C+b#gqP1
zCT}>koB-64#BPVhyzjf{x<AR}C?(YcKJ|X=Xhz=ZKzslEE#6FV>enxEk#Z)1l-8MJ
zi-@w_;{0R;65B`laZDe^lCrDITO&dze@0R<^N`W=JvdZ}k!LClX_4&Bu<pCA!rsG0
zL>~V5NRt9+t>Yn<CF&l)O|Mkcagy;u!A@k1x4r=pxqwHDTQWv2zv=ym`Ke`&t6-z%
zL-TmElwVa(Pr5O8^QFUIWta*TrPqPejJvyld}hTOy)C%7AEBk&6^1`?;OAa|@lUJA
zP`#HA&eR>4t%^HT6KfgPHX?+gm3*>{nx!yeS%XDKWckg+epY3T&1BM{9U<&OCphlu
z0Un2fY3ZbH$mh@KBsK*++Sxlj!aH4kaAoC>j*w+AA)^sP(s3zS$BAieWJz6)bd-W|
z22)q7j?)~`Y*<R1h{FCW^tJ~_yT>u1xTfyY<kTVK!_V0C7Y=FuU-OsF4nx0O+Xw|*
z(!>CdKZ=eL;)Ho`S3Q}+(Ku4`*c@EdJNMSrwFv;?*(+1FYY%0o7u$9IwbJor3b*ZR
zdUNuwbRwS{5j>aVA6tY7GWbSZspA43TF_)xiMTDG5=*no@ot9|Tt@2WU-Ypj-s!f8
zyF*VPECBDNhiZ<B9>wWJO)pR3j@3Z9uk5dOKG(Hn+caRxEs3Wwy{?(0h-LmiC_f+1
zczL<L7mds;K=JqLn%}<hhK<}1+q(bA`-SAVtt=*GQrwtE+>%nI{d(GjE*-e~-7qm7
ztGaSW0_~w?UCv?hwA_*_9&Da}nrlBPhFz!d^wFY4R~QKjbsWVgZCiK5=&{OlKgkn_
zJri<GCKWb&*^i0sNGSq_k|VS(o8MV}^%j5p&QrTGTqI|rOBeF_P7n}NnV$E@eA?n`
zLqo`0(cAUe>%iMBu<QAS1OW2u9PE1PdcDkgyIAacOG!x_qEaHQOz!JEkT863oN3P>
z#(4p)Ldc~og_zZfOa+-9mm2q1!E;v+ZY2KGBsfKotPAovpK&K5&l`_uWW~!M!es5I
z{7zp?MlkVlO<t3*Kj0V9R8u@QXV#R?euw}QU&rGE`I8f|duu+lMq6*`_4u_}^42w-
z>VcJezHd@t1E-`M57@%EE<lXT<+K$3?x|GV+}r2p_6KxwCTBi9Yr8yxs+r~j!KDu%
zSh$8ibJ9#_C;D$$Z;wg-&!@sKk~Q~PK;b8c@v=l3G)56Wfza8<Z?63IqHZ^Omebxe
zS^h$A?G~vUA#tIj&NRM$oc<lws27KOy!u1WZ7%HqDquzx3>)UqwcLZD3!n<@V9xO@
z?*zC4JF4tG%DVt^!00Ac<};6%6m<3#Q)eJ+JcID_&c@Tu9r|{r7m7FWL&Er=I)B<3
zs~OwAI$N9Ot$d(QRLVGvJz@f_GHD!NW<()2rgrwHs!MWcM^71lNf=LL<fnzQn#Dz{
z$Kg$!)#=0J%N|skTqEn;NscFW;`g}hBZKrKYxUlDC7w|!CXfWC%)N_W*jdHZ^j9P>
zWgjX{){#B>k@m~08*{z;he!IVC!d14B;QV^)iNP{wT#yjrOH!pk0?#$lRvpbn|u#P
zJo-rL5fc{{>et@rudiT|GcDA+o2BQ+$r{At<4v8QQ`>XnFCEfUsF~L130`CLu7p)P
zWil^&i^~h*tK613J>sj+LaQ)h-MBmTFqH+dK3dh#GCsAn;e9(kW*0`_gWPGb;!8c1
z|Kn-&;v3(;JXT6?{-P%lCpkMNbvt}zX)DNBq+i-tQ3Q^r>4R||ymt)N-EqI8J=kPG
z?yjiCazp3Gq;7_<BCn^~tvv}G4=pUaBqH{~{w2B+5t4)d$$an(OI}J;{m)wVfv*NY
zx6ZHFGF69*W2J`Z36A!nBci*u!}NYR`kZ?+zr6`AQ?308E_07P@Q$DY?D`TXSS3Sa
z!`H*{u^}`_e&7uD?T;n+`^=m>WN$}Fr57=_Mf$xPqRXNWw#6O?uJ|C)8{1;f;tmul
z=HpYzWeWQ)TH-@y3J`_;KZh?${uE-|;ee#rt=J|!S)hP9i*Q!S^9uJ>-R!0H#`x4P
zcbsDuC4o!B(6=l5x7V)Mmn!R`tf{i)#?Sk%&po$ojmq~TYhGEgR*JO>5zFe!s%<l~
zokukH4U0nRX<0EcKE2Nv`+ZJt56o|`RsOH@xqkX0zEF0MP!bxB-#-uA7Skcfk(Gs`
zfM3=E2(QZT=Q*!lLx`2S%|Yv9IFUeQrCq`)YbYphe|!4)NTY&l+elX%D9NbPAkkJm
z<Z8>wFn*!CKg9mp?Cz>n;fP32vq6HzuWmPvx9{r57hLG!+Hh7L3)nBkg~sj`Ew;tu
zSCy3=*-G$fDP%e7mA%>ub?!ub7&ouj<39ES<3BvLc4m(S8&s7c*XIv`&~?I1q>ziR
zrE0GOX@CX1N-Z_puCC<*f25@Dyhix_Pn<!&cUhVF@y~qSb#;pnuZK}QGa7h_)zy1J
zQS{oXokKqD6LzUrJ><v3dsjU%Wx~fo<4>)vfL43Qjxy^3q2ev)1g$_N^5aTk7wr*U
zb@j-4JoQGj6zL&Y;9#%2?0dTJ3cQnUuJx;jk&CKT+=9ve;<<CibydXlhrAoAB<>Qa
z@pLU@VAqM`mvjW*x(6I2fAGWTu4{wf6mY*%XbIi127oXMv{zG~t`anBYkj$q3T~b3
zCjyoo4_70d`SXnm-)4Ca?p`M2b#`C=Mx-ad&0iYLVPX_=y?J2*R?P_wwFqWz-Yg((
zl*_)R5<6RzjX3Q1HXV-M<WR1BrmG)Pt!ff7fprI@U9GcQd8UF3?Is5cY{cJ0p=9V<
zY>(b@d5HhEWSM2WNlbSdpLf4h;tR?a+~2&p4XxE}$a(U)o>DFParZpBdn1zwbf+dM
z+xi(E#jl2c8f3e0xYjW=0&x92667chQb8XiD&xHs{Tr1)cCDkKk7`C@&u%FZo;8HT
zv3Ehr8+|_-mS$h6mJXcY#)(M15UmVSZra~(eLU2-@_<@F&kZo{KJPS7d$<<h3s`j0
z^=C%R2D(QOZi8AXLsY9nu-z%cLjBq=R`?PqopLebS6%wcVzLQ`AnWawLYdk=H{%yp
zy6O-W)G0HcQZm^O?m{WKLSQxha_^11LGZrSD3-^7=gm!&vZT}kmh_@8qkP)LErw=y
zjS9tRkcD8akOqU`4!%?rUx))L-Fz`?TnWam)8*{8nyjAIbLEIk<gLi%Ecs1*5;pfc
zgNQtY&Z>)D@BS3hlKS`>p+ya?Nbe6=;ziRAjQkAWhTs4B#aOC)J~x;9POt!*EJPv3
zCN>|6HmhYYc2dybpnoy@J|Qw`p^q&h3A;i5Vs`Liut?yp^E6s6G905pnH6)}j)A%t
z1sPGPAhht^@0mz4ouH5CV8yi(j#~D%2FaD6{8>@aY)vJg1Coi$-QO1-V@tu=<%(G^
zO44MS8=A7RB<!1CH?z4h4&SCL*WYx~I<eOy=T8;e-~TmfM)>?L2%363#hJ*J8}ROZ
z0hs>4x#C{ynNfxy)+ZpV{;EJ|v9&6=fRmZ=SI}O8cw8=@S9^uG<xK90o|b=bKg*Wy
zDnXjtSF&15Zp>}&BkG@6R0g^APc8nHocl6Jxo%m^aofIqLth7;NP8K!It|JHmY@`w
zNMgb@<~)<@$aG?#U6n{--u<zfjdIIIJ~5kb6Z&X2Bn~Up_vgjRc|$j|L5h!<o+wz@
zZT#BvhobO9cMyf$fn|~Mzw|k#qOhEIEQEe|l@(%n9uqQ=5$0Z&GYVVn4HQm;*Ry}#
zY2R-@S>9^-i1oxvKXI+jmdzxh-O+p!bJMTzL3B(;``%x4zkXs(q2sJwPFSUyMXlG@
zu=lke;_Z9nVNkyVIdYTCo#HHZLX23EoP)QMx!uVU5#w9bSn&1;_PuM9zS>XiqnGs2
zb4;dYn6P{N2T{?`qBCz^@MVk%cu7)@KFjFDma=!i!qVQ3-^UV5MJLKjl4{;72=<MM
zOXAx0)zhNXznuM(lkiS5{TZXuRX?(eNFMgKIGrYbeQ93DW8-$PsJj?b|KygB_g+y+
zPj9oDXVGb=+B2g(%ei_#BR1VRUyy}v-+;QnhP?h>0CpctxG=qZGQ%-c{=E277(i9p
z>xT=Wz5jx!73^lT$Tg>BTCubC57zPYkDK*=q>q5CXD;@?(mRh_2$7JdKy);;!t;>`
zqlXa#@(6uxKRNu9!9oPdME<MN)5>?cncZfBB~12Pb8{WN^j4zoT5{hk>}^^(h1@$F
zX=&&dvvW>!iDJhmG-DN$xsa6Ms}<}h2`0XCQ)VCW>d;Ttq8y*o#&OP~V~LMiH1EZn
zhbeQRl9vVM(j47)E_BqNYh|_K*rwG3j8Re-i;~xeko+oZ>an@$mS~y+=4bgZ+lTHR
zDO;MOpp^xVIDyelQDM!DdQdo@B~77R8$19Cl7baJ+9yfXL*?|`-I$P|Za9_vtpoeD
zPkn|PG}=+oxnD<!Op0B+Axg>T|BbQ5PgJoQ5{iOQpe6a=5nS<`X48cl#aKTHMF<Kq
zS42c7Lzaw0QL95mYl#JEw&Hx)2aX7Q#S(n+=r0>Rvl1A$w?ltZc-k>5(99F|E#yps
z@@BRbI+iftM-U}vY!gud66e$AL3ykYd(9IRBe1QBE93gU8n1b~{iBx0&|yJv2`=G8
zcfQao)meJ)RKt$-g-{~4?WeofQFhpWG8{No>EnC<+}SeArlRJVImotc3ZJR5WLZ=H
z1R`@`z7OK3yfJv+SBxt-7NW@BovBLY-G8>_6+&@;0M5ue8Q+=WOe#Xi*#q39Utm@~
zh%Jz&)oHJav6FpoN{kRNalo3+2I@1mJs&tvc;5A-I*54j291rPDp~v_sN{twuT4Pi
zpTycYE)ON#IQ9vZ!oEx5g}9API)?dJIbaD_qALqnx9^q58+f$fl~xWssYOV(Psw<j
zv98JfXSsy~VqwqQqV>w>aOMUo(y`Os#c`7!Bw|an`p2n;D%B{u2nDOXkN`!M^uN}V
zEwmSo@{OM&%p0!d;QqAi^MapxVP2=Ua_YB?qxxg_lZEHbQ9-mr){J7<%idlA!gXR=
z683V7r)6x0h$I!sg=b|*BtIwvEAKQ0mY3v0AW9%yo}XZKq9i#e?zbq+n`P(0_bqoC
zfh#t>djgV8q44FR>cR%&%~iCV?>+cGb7%|4P*s6Bvb*ze306UB_pf9TBX;Cs4ZB}|
z86AK`?r|SSHjBSR#|o%r%HK;E7ce6jMJF_4>2Lij__Cqs%ZNcjsQN4V?6ed69s$%J
z7_*g{ge_jKpo4laOcY1cfx6gUFSUDEfWSW8f%S_+g6*JqUP{sZYiW-$BoWR<4jDFi
z$T|8kFydq3xu_Tk3_6w=MDXAlA+CHQiO~%*b9O(u<VC+2$@hJ_X*ms+M&>CM<1Y-z
z`xSN8JY6%7ea5com7X*5g++r`HW3q+`{Sa?NNB9G+CsuEKadi`YNwq3e%Yx#aczyR
zy;mB@$U_)WN2HxzJ4dJ;us84!E!f@T^vE!Vbq1S5hW6!fsDsar6Qkik*wc#PTj6m!
zbMxDrIPAg63Ywry!S$h8qv2d!qJbE3Rz7?Y@5gGk$NdoNGC3a|M43qH$T|mdA(s;C
zhreHmZLFx<?Y)haA;p;a&a41@U{mvk__N)hBIe4|{Q(<A8J^@s^;o2%%HP`o9=gvO
zO@C=wYoYmhcrhlLp-M=h(dhehb%WL;sMvh~RzSqE#Q={LqKQ@%zzoBx;C^R%xHUV^
zEB(T=?|anXzMMNh@?8tM?6eqkeKduBD;4>8d$yGR$f|t`>9o;<gsMmqxLAt5gQ`-g
zrxn07fCLqr*Ifvw5J5-;rExGN1xz`%lPR-rbsk5{1S;O33|GC&vvB!V*#hUSL%>p+
z8LJzhcoVgDLcyUFCedurTzF6G%VyzKK4A+HEtG_l03A@f+uVlbzn1`gRP(h_l2hbq
z<JFCb8<B(-(ir%vsN!EV=xNP=3dl<ZdQ?2kU)*aaCc^VD!}571<nr!+1=iTso*C#=
zyA`90Yno5>#i9`$RGhK4VG7>Qr+oL6oqaNCu+`5uzSqQp1oqCPf&}%c&X)finnF=H
z1~qC_DHRX-XO>clf9e5>gftkHEpEKHjCMxF))zz=Ibwk@g&BLD!f=d-ZpprX%>_l?
z>!B5!Vws!Nj%MAM<aMB;ss)1&Y6dFN51c}XRSQE9A3v9mfltHlQ{RVLGc?}b1NJE1
zwta;VS;CG#^OUld449+OW1km(CCoj`dk?Te{@D2*Z~*f;E#W5NU8a_d+2~m<%En)Q
zArOB09Af$oVC8~QMKQTrmZ0C*_peRQqdv1xq&mQ7R4zI^O#UFuYDjW@5t~a5K^&BP
zrUKq=&bmllCHdR21)pMdP1%N^6g@)q4NKMdWBcRq;=!FO<e86r<58y&Q!{S!6*Ep`
z?N_;kf2DWVd~(VGz{ChPuT(2Wx4eF{>3<O3!j~cV&vpmHU@Mz}jmHl~6ih({11RD}
zi-Z>qrbOX{dWt^1W&x_8qNMTR8=)8;BN~zq`sO=p+l93~UrGigoIAy{#{b^g+S^1_
z+OS6EKAzM=!q|6YYr-1S6|i*Ujz4Z3Y?s?yx%D%mt~OB&Sua<-ZF4@(5h`k_Hhr|Y
zz_;Z$k`4qi+DNVi+D&^AU#wg=oU=yf=AMf(I=bmK56<TD*S!a<;3X~ih#8sab}L36
zqCj6jlbhYQJ^vUBnN+$J`6)>5lVdeZH~YS7f6$(U^8NWzCbUnJalw_Cm<!uqK}8E4
zg~t0@!&Y*UEr{P7=7j<dM+z&O)|xtMN}etYF@nm3o@n`?r3z@Hw3CSBVz|35jTFu0
zK73-47cZZqtavX237iK#e29b)NNc~bl!|Am64~*7-3?q79w7A~g7VxXkDW`y2kHkX
zx)45hnmJ%4grc*TIC?=@RUp^zBu@j$d|J&>^*&}rE?YhC*_q=LF?@5N^d;OFo3%1g
z?pCx0wp7<A=R`o7KK$EW>t*kr*z1&MEY8^bx}@d6-8QCeP4UY0G(XI(VTBEI9)`1x
zsg?-43qC07^Bt!S&V2zQu)%nP?|&Q=)H`|iiI%bOmM4WRQLrV)f}jUyT0W~Ai_{5T
zq9K`bc_I9#$VW~CmB_78S30FA)@Doh$j*p8t|N9Iu-l_gI+p*jF+Dt%ll%dmixoUk
z1D=b$_V=1#-<3Fm1HgUrNoD+2T2)`))cxvFAN{udVGe%$vrnY~@RRREk;TB@PFLzH
zOdkVz+vw6f)2tNCMh|$Rn*(->?7Y|TdBLUm|D6q?&0dI6bKQzBQDW~TZ}w5r2r38h
zTKUy5)KCdQQ_WldNeIvgT@jni;UyiRPR)pzMwI1&d0wB@Twshfxn?z=S*dnlLBX)-
zW-+m-66(NwUc!bCk1`rxLm<FQg<oD8z>@v&H~Sf8K~5|t_y@)N!#_cx05QYOOn~R$
zlZIA(qpzMP&{G?ga9Rcc8<(szr8x>-vZYRDH$k6b`UUq?I6o-A%~uI7bDZ_l*ugyZ
ztsgL$>X>I=Lt}Q646W|tqMGux2P*lz!3?*93NV4dj(9asWOc>r82i37?W3WWkRFwq
z76(ieVg)aGqtU+LQ!LBzL;KEBOR@J)o)?e~9W?L#+?h?0>K$Uuz_?f41kIz&H9;cE
zPz*vNvEzi;GBA2At)6?5_VlfXXbTMK3atS+j`*8ufdxybF7uM_{A`=ccJMZ~=E8FO
zZ(Hcuv+cg$X7w;HEZ|YSB*?Aao;bsvNFM7OEO0p;=ELj`K(#M4u*L1fm%9fS^fjpu
z4)Z*D!zp$fgg7XGZOQ_NXyf*&fUz*Pq;o_Zw)CDraE~GdpFKa1RX(W>ACFc4%8p;N
zu<!3rL<1Z~wT?(c(YQ@#*?E}8*I*>0{dZ)12NpXp`(YmoDjL4QWi9CIaFS%Y;NsRt
zNp1`RNL(jBwhv&+z0dOE&XwMzo}m~r4FE{+LcBG#P&D01t~8y}*=3%Z5wmponvAnn
z^_u5Rsb)e>Hfiu*L#;8$)Y#m)vl$$Fpu%{9IpEXpGSdlQ@b|H(GZxHV02C^y*#~eN
z+fsD)VHPWf3u@t`#@=u1ISkwG3d)nHxDdKqK*{_IFxnPNbcikh8JM+zZo-HJT7d+1
zwDQ1$1`J_XOJ<0#((DfL{H$HD^pW}SXW|ZFty7gmB67U_)x1}k)AO<vnHDnmU$uey
z@Gcy=wsT;SPRRuXqM4o7V7as2FHkRtlEHkKssoc1Ul13xyLbjt^g$xpLfF+{{-7u3
zu$yMm5qVKIsF}ai65~)^E{LWAmazOBh0hw>0G}t};?v4SKd|b9t#KXfW-yN{67QZR
z$VrX_2N(<^z<1y<>e0S?Ag1^w);$webAZ;1SA#tfx){BLQ;?@Dd7aYQ&)`;b03J#g
z-wJggX3bplm7PKPRZ}-IbaJ6}QDbS?B1Ci+)>`~@o;i#F{!X>a&EK_A>HD8`8_fF@
z|8wnB4A`zLk%2>n74GA6wFS`$Qr(0XHX9E<?h~8_-fNRk5`3#g?;cfjc%Nqm&_=}N
zbpFR91ji!0J~nIgEnEg(`G*=2E~m^nW!u-As1Qjc!ig`>8$=Eq#+j;FcKxf{oQb#7
z;sJRN|3d<J0x0B((H08{z<yJ}le%Q3-S<OS|8f;xlUZMeqF2oSUBf$Y<f)K^wq+vG
z2%=@ldkpT^EGcB1!yaPYPJj@QF<*dU-Qqp=@T{cH$aw=^rAVSi?Qz5;R{3|5GEQUG
zA&0CD+-5xJ89^|}L4O0&?ZH8KV+P`jVK~n7jT;&Dy?F*K4gh%QXpw+&Qp49Ph`T$u
zgtc8*O@F;R#`p$;y6RV+*(xYx*MNgsC0BN18h}XQtEGo{D^mtCJ*RybM}PD{n9T=0
zv-?ffr}`l-b)ubqRD(L4_Bw_{K)1@@C=%V!%zK<ax-=gMWwEN7lrycG`;M)13$NdG
zIG$Jy5mm%&*=$h!T-*{^hWIed?u5zbZbgkzc!ld25YBOg#D4?gLUHBS`Hf2U99C8_
zO*i<cPs1vb2d$_G;&X<q$Oyo)#ti^Dk&tN6Jttq*pw)e<na`7!Nj0YqD^bFLhM>N4
z{`ApJDui2KQ11#a>p*;vFrmh)b#J(@EXf+MR^!z?SSa7ukp$3zObxZ-4sT7$NS2q0
zl@%OB5ZE%Z{q>jBj*R+XXB`{cZOiD?DFW|&;JZP9-lUNCa8Djj;WkPDopAZLsf=$+
z)JG(q$j?s1+GN_yI)~}&O5+ks;lt5n%9kt6{m4g}I}}MR>9U1-XTq1eKxjAO(r=Qu
z1(5xs?f6sABsy#HCE0~_diXRMaF9uGY_Ymzi|{utm)^h!%?+s4`;q`i2Jo-x9fqMQ
z4QlLb{=3I#W^3fdgE$-gxR0}|aOqb8P6Zwbj;eF3@#S9GHCN4IrsX;?3N~i7(GdxL
z@gTaHf$!EYIwH;jkSY+{@e6hP+6$cQ$w;D;g>J3ONTLX!HtF%@HrW$hx<nCvw|5eQ
zm^7cwVlu{VHOfd{zjvZ_6S(3H;N9tYkkVaIiFTOz5JAM^bHpcCsMUO6I`zdLYC3dT
z(*jrv&nY;hPBTB*<^UUCMXmREmOE!U=>Qcx0n2PYYyW}>94jeUfz=9yt4iR|G9RWw
z*K`%1T!nQ0AUL)~*7Tb9)~zxVcpcAkFS?UU_J*x6Y)XKKV|m_TA34|<5*41wVm&G{
zG+93h@(%7an|kWRLCl1nI23-#vUC!|6$VoTVCQdb^O@NUW~CAGp3J5QeHRWE$1lrZ
z&1GMF&{Qm#sF%M*D+YiHG!!`&X=Q4T%wWQrq%96L{=CAJ<9|JI+ZlQ$B^S9kn&c#S
zftF7Qy}Pa~D%B@@d{<S89{W0yQo-VXx}BbdOfzoyU@vb>`2V{=7AIN4z&xr$mZF;7
z3{4(J7orp>SSBYv<8%ybJO~PhzDN*ry@yQSokdC*n0B65^SW%@4EBatnU7;8!zjAM
z+<tW-O+``8O5VeO&XD8~cI({l(<soqbA|=>SpGgAGQ-}AoM;34ip_s~y<E#;?DEv*
z1y3IlQ3cTkXT&lmKL9Nt-L%+kGW%)I_1>s`E;2o~38i}y4-^ci5kd5~5Ke>7$)`~J
z14$2BGNM83PecRS6=!Fnci7HLFj+)SmgT!dqZ9MWpx&u$_aYdFdYR)L<VU^3Or!vW
zi~F6-eiW=ZK#H@Ge%F0V=M4fJf>H!ZAykl(jVn7U$ixT1shr(j+2QfwSz0$rY3euQ
zfFGL>PMYZC_Np8p8JP<ebiMpva(`!^Z-lvh_&1{;hiC&<8p3!Bnr~#UMP-mLL<m+6
z>w^z|Kmo3P3ao~OF;gv*T0ZUI$Z}v4(3Q^EXHUl*)4%H39vRb5|6z5uW@~mMasHw6
zc0;!2O#ks$Ov&4+ec6m{^K}zw^}6-;&c7M?>CPYG_3&!%&g)q=l{%)mr=4k28i}^<
zD?BVxcI#>w?Apn+<om4o<o|HomFN26KlQwF758ML`yP5+s~eHjdfy4yuxy&YC*~16
zzeMI=Wb%0uG(cZ{@cwnn3B~5$fWK|XV6L-uyC&*F@}ai$dKftGeik%XG<toZdBS1-
z_~;9kAAI)48=KSc>w1FNJ-<MKQ#H@TEaEHn_)KTU{+^+<Z2EqN5#vdRD|niEh>4Py
zcl*`awa<mamvWs9KD;!?XVQ$!rcYD@I9I~ZDE=L9n+HyP>LcU)y|KSS`&2K5_d{bT
ze(aA!pcFsn$DxfB{e`CV%$qy!lLrRo#l;KB)4TK}^lRL%jW@@?o#)UYg2!6^BpDa?
z;3r~DK=(CUZF}G;w{7AEdl!q<>wTD>S+x(%(3$2c>SJ%r7)~oN=bl)TDdF;oag4qd
z(zih|miF6Uxz)$!c;i?H<HZA05xK~o<3_Mn_VnRNo4LO;E#3z>ZhHUvD(6}cxnss=
zg6*=Jc)y<&1MV0he%il?6kfAN1{LAJ)n)s7E!!e**b8sn4c8v8*q1Wq>kp3ECkDH>
zPa80l9o;QzcKbISJmNbRx4rH4n7*&aG3+=A@clOe(|=omX=2{rW}1HLJYT6rb$i=7
zi#7e)`LX?$6-9rc#e!XbuTi!?$@f6xVQhTCarea(=i2mRr^Qicc@K`DmFV^fbGx1O
z_S(dFr{od-LkIrqH*^13>#=4UW_=e;+}+(sCicAf4GxEoFT1M@?Ay-kJ;!#&NP>Pj
z;M+I2Jm=!E|2Je}HcpoRi%k5PS;W=F!rqC7mRZHp#l{qNR(3VAGBtLAr9@$X!!9tH
ztf7jj2C2KHiv_8Lr-P}Jlc^c0gQ2mtp}8q7^KU166IWwXn3n5*YK*K*EKD4<%rd5)
z?)FY5&NQ^2|Mz<=?d`;1aK@xGVti~YZ0sz&u!EC@mxWb}mRZi;<o{Eq;$-OHU~0lF
zVQAxQ%KTeJT!Tr%)yC$hp`D$n6Df^~g{3no?DM~+Gmxs8Iyu9NBK@Cgz{IeM%UZgb
zlCrU|ax;sW8d)0J(bB@uluhkS?2W8GvvR;H^M8AT`TrTRfP)8y#O&nkLdwa>%PebZ
zXAbKCZZ0;K|6dUcqBCTwZkkgv_Fw2Ht226NFJ3zX6n|h$ZOe1>9mdWC{d(@9>2As&
z4SK!X(&D4?IG&MUd-&DRFRCQdE|4z;ojq0*0sw2@R)vfh^>2W+5_h}z#dR5X`;y0b
zk50bVd{fmItbl~haUdqIizWcrf>u`GQ?B?Vd{FBrX8l?>D^!~voA4UzDWP^tLPDRe
zlR{DuRdsL+Ug_G71kPFsy3!8e8H4l(@Jty=OvhXPS%61R%+niAk;=@|U4`}kIUQ+-
z3qZ>G!8#L`pqm;Z+=<1D8fU@2#fKW$!_Ben<m+2{Clg7-sZ@ep8A-#G*U@Pnz;AH*
z(XBqyvGEkjw^uUdOEgz`_PvyK|DlN=yoUL}C%jT)4006~(&CM{nj7l`C?(ickyHS#
z#BXnBRxklwcfR1M`p2cknaM7YHE3+{W@Qm9zd-B-9>~0&cTIyK{p<VRZTQ{KHD&qT
z0fl{^^o8H<lZ4-jrkuncKz@Q0drXpVj~7lwN0V!65c#W&Z|!rg=0r$mC>x%?%#Y+x
zjQ;SCxvN1kc+LcN7^nE*-EX6H(UN}=zeXD9SlW{#pUq#%?-F^(5A|jEt#8fccZ~d8
zbjoy(b3c(=2<@ujZXwR(nBotUn<~NgPeRwmp10K;?BQ=@u+-!|@+`WVwA4gf=Vcrl
z;jb+x8rse{<i8=aKW@K@X*$oib@)ET@0aoR3byxuo&yTky$8Q;Z*;xA*3^#dgV(0v
z@H&+7T)L~wHgez#9_0nrEIZ?)-FM9C%dOgKgwD=&lRVcA4{2`e>O*ygJom}$VOg}&
z4nrN#`ah9HE8a7K=je`Q@l^K}&rti;%yltXJCsQ>3s=(kh-YiK+ZwOr)7^=kV)lvl
z49vaURedJS>>D%Ml_&XKDwfl3&p9U6>mM+*h0g@~&Foi1Wu{N${EKsJGu$;6w9?%#
zb#3XJMjPKYZB4Bk>x8k^nE01h?C>x21vrR0(%tW8TTf3i+}u`|NlNXSGL9x4huqw5
z%L`T~FATJrG9Wp>TvvGqMJun``*OKXP$O5Wxq~`>;oX`EsDKUa_eAUXP9}VcZTi;7
zi<V9%N?i<=(m<L5p&jA!MKDrCxRl1w&c2K4SuE>K@nIn2RBos5I32{DTxDrqEBG9q
zG17Cydnzl$;qIL2evep~emJC=NnDnCG<C|nzR-YYZi7A*-+48S!Pqv^r4jnJor4|n
zx}DR8xy<Q?LFXF@(kFs#Twf}Nn!xktISRlt=eMe&nH05p8g}XEHhjzm4MZU5=?)Xq
zun;d#6I`V%U;;nS8#z~X8UyLiHI8^2(N^c}=R+4{I|dgnG>+JB@BeO@ihge?>ssQg
zvnRlY)Sw0~F4@5I&p?~PHo;t!wN=>u?kCkEr@lny`xhFvD&O?_y2?0@NZnLC;E#wK
zC-ZmL3_5-<RuARRC%N*J+KEtT9as|Cn>`<;D>je8w)i{axdE_PPYC%t3<>VHLv0Ow
z@frh}H9CG|oRET$<~4r@%){GcT#SW;%3H4>nkYy6XJqu4_EF-O?r!k)aUTzNroQlf
zcU@*#el8yI#}6#)+2ET)J+4+pb=S?$Hzn9C)H)?1oT)N;IlY%o*yOx#JE}EVGrG{9
z;8TV!T=*?kqlqR(8;9205SA_0I+p8=4c7Fi;<RWm_729b!w&dfE*DCI#y~>+#vFFP
z)ov@IM#JuqYSgb>e{GYig>B%+Dm)A@rDnORU8PpL&4g^>y^rfr%dss9c!*_IyJs|3
z5O0xF-Aj6;2`5@@GJt6dXH%wZ603VP(1<44Lb~%Wk8-*tso5ogO!-z`g@{MLK|8Y1
z3)y26KazfPvML+1y!CngNJ?YH>jLY1YX$3!cY*s?Z)CfB|4QC`P}?j}SLmmlF7^6C
z&T3LY9PnMw?na6Y84`<1h^_)8tz~Ps8I}+%XBVy|n`uhVk9a7s+Lm+S1@))Jx7un%
zt#cW0D9g2f+OD8pH=*Uea8j#^@K|+Zc_YgxQ0rd>)E)a4GgTr~^=yD}*x4}JBVtzr
zErE0r>wJCchml@&03>XRrl4KrsI^fMU#%hJuN3w~AgQReUxu%yIypH;I|kQAJ4A#H
ziZGTjM{Q0ved*CLiP>doK-0IxK05+x;<D9{a3_#0{OMUEoUhck^TyGAqDD9%{if|B
zrR!p(s~Gd9?YDs62<_-|;UHy_7c$-D-<rM%?G>|JaAylC3d_G2`LW8FDGR61tj4PC
zlRsbDSQrQ>7y0(d%6ua`M#gb@SojPx6qKs%ITq4N@|}gHUlB3d@K!3APS!sH>3-Uw
zpNaon<J)Na$V+UpJiVM88ADR|sf_PwFwpheS>xoOWc#XDFYyJ`E$~A27=@Vtk)QJd
zx_}3?!96+Y<s$XEup`)<p<4&@^^GB-!Iycs|EDCu)^3|ExDDnSdA|*=sB&*GZI=bo
z!mN+{ryXIk8m1GZdp-;9PesMY7P-bo@|{rW1-hT>e2`ejI`&Mvj4e|_q*_i}!W1T3
zLRvCgc-R@fG^2F~P`RitkgM3p>v1gCy}HnI&K4RpLcM8=ocBPz35MATFJU7GoIPX_
z3S-ja0ms98UFIG;7_av;y4!5uuVasQ1~eh8B*UaT_y3U_*kyg#ORyy7nc092f?iNp
z34>f%yYY7XNVAksVANQULiz%xUf4LC+*fiK#`Vm{Sc0xDZZapx;$5M<%Q1_J1*mpQ
zFoF~RQ&$PIt=O%Z*kq4~O`nO!YJjGO{}Hk*IR`wmFwTEOonv_!U3QOaVB=FStor{$
zfUuk8rL@6y{YR+clZSWD-8A*)>2t{GS!!S*jKTlV4C5LeQ*45_V#@=XaJY=#uUB+E
zjUWCx$A8v!#U#J`1Lyda{4Q?ZRA)dc=TmG4$4TW_bHoUz|1B!E0$HuP9@rI6eVGX+
zIXbN5KDM9^9@d+2Su_<L7|#6^+k{3CVw*f~3qST&jS2_uA2V}ZLBz06YlIu06>IM>
zY+n>>*EX*O^Fc@!41r-ySCQ^tFYNyFNNRZgT9Uln7qvK;7jAwmw&d&b2KHlI=M~JW
z(;I3ys(%(Y&q~K?9fN2+@>}jdZ~@MImwDJCaa-JdgzSxWD)O?EV_79v`iC<_gQ+E;
z8K()jEq9?FCWN6M<f7JiDJ_h4gfvzMhBNfzQgJn9kaNZm{S2+h*C&}}^-6k{JE^x%
zgY0GcS04t7!#$q5wf<3W|7U|CBS+-hr}3bl{j1+2yTaT<_l-wF#1Szspyzkb9```5
zkw~TY=pga-jQkbxcHfMwrYyS1(r)W>?u3HKSnuqFL!%w}og<^2O!TeLo4-u_UmIPk
z=Eiouj!pi3(T!Rg5!0*tTHFRpwM6Ac#x?;?!eYHIJFr&Y)gKLTWiuWQa0S?S*P#07
zV4iKpMKEm>Ng6&|zUXPU{|G<UX$^BHv@nKIZd8;m2wrfoT0cJz?~-_XP<)VXKmX^k
ziZJ(MS#>U~>l#@07n_gP&@tx)qUrMQsRmbc(7^w?aB>OaZw(!hu}Cwek+G7x<FQc?
zx*4WD{L^fq3dlgQ`a0hg1w<<+b0JoN-D&%!R2RR3&7%fjRP4mlyk2xX5VcS|77}6^
zv2H7B?y=mx+|0BM*2NqRhn0<zwP08m#+@Put9@x2j4w!(i*td-RCP$c-c#8T>Cps$
z^)FFI^m;XF{Fd#1c3z!aY5VSzH=Yro2~^O*Fuj(u0ZO#oSQ?joLc~6p>#QNbA_oR4
zSm}nfyJ9JjUE<<CxRG7LJctiL(8X|5C{!IL5J-CBXf*l>bC$55^aZf_F<}~tZGVT2
z5n#8}07AM+l&*gCe!U01U$?u>RiH8e)=k_M6j|@r``eXp&H?YF++h9?vm?OWA+Z?Q
zV+j7wM-QYXJr}tDL)cq~Mb-3i<4Z2xvGfvBg0M(Px*#3W(zP^_(!GFm2qH*`pfpH#
z2}nyyHwZ{K()Bys_w&5Jx325`V?WoNnK?6aX6D<o=N$A(`RmIu0}T*jF)jbIL+lT}
zt2I(IV+QJ&jCM1kFdi=hO)xppQIRt_WgJaEtF5(N)g2h?HVdfAkL?h%SEUJ0kf&r)
z4LIJDTJtvTERwA^PwRc1c-i8Em9P<HwoAn1B%Y<SgWaOwgtwIInJ4<~>CK0@S_>?%
zx}ifsaFyk1xv&j-Gf{e?XZCbh&J&so0dJs;fCjxe(Lg#3=}9rmv7Euu!MMDTor`HC
z_<^(_le51cOc@%WsS}3feJpxN#Hji***jH)O(-188{D%~bcSE`4EQ9NvN|F0@i%bd
z-XvA4UvcQ~eD`jiCA-`wG<N1UwsgafrCz>zLVUbNJFBMC3J8<M4Dv+c?3u7W3kS4l
zhbsKO3(FV2z2EJjyaY%4_M1N8UY>8B^qTxm43%ycr9oOoc(Q#=B>b1(+Ylak?)2uh
zxq3`$Buq~(#hw^p1N^f12_lh>)RkrRS*m~*aa|mxxs;W=`i_|BcU1xboy4V<@S@c<
z)T1xQ!VPf0&Co#t3dql4+JW!DV_lySF%BOc>Q)B56_)Nbkh_F2R*|deJQqF}0A>a*
z0>DyVIMS2gl^n_e?Kh7x!SY#^IpkIX%mEGa=ZiK`VcUm_{RKV3kuNTtoq?F2C>xgZ
zKhr)aLazi|S$P7}yK{i~q!b;H;@Xsq$Di!W-U5B0RoK(gL|Cp)$3hYkD`O1c(~Q!e
zZ+?is?M+O4_aZ{<@{cdp19jsj)D5}iyMF*P9<R08U?s4DWr5O$(&?shk}(2i2Q-7A
zYW$QlDq9-fbUr{0O3QP|x<y^23I(L)z;I>UvC}d>=$iNN4aPqI;iBu7I2f$IYr7mP
zYqXBeiklZC#QR4Wzu9D|<}+u%&!t4+vJ+=NKhw$B2tQNLTFA_h%z_81cRq1y)#_Np
zK%a@z&u3~JdBmx>p+KJI0+4sK#?QoN5<@J^FiHooXky_9?Pn%sxIB*1enn-#&WjL7
zXjqlh39-zl(i|y{GS-=1&@|TBKY<0LY?$&RKW6vYu2>p8lWE?0mg1;ueZxUlMYs6I
zb;OeRKxAZ9mH1%NEIq=r#Uy<9KSGE3AVsaTvv<c-hMbw_GYxN7Z*)H9fYqulU3#p!
zyR9Gr17S1U7D#H)adl++CG2?RYK=qIIP^d!ruo^|Pl&ZHCmhX~W_Jq{0HzCv0*=yu
zkO4wv>nA2X^)k0ExQcmYq5Vj=Loll+vqKPRx>1gApA_aGANCQSR_!I@RuOnC>7P_@
zQRq=`*w0t8{g=Y{d=p9~isutpWj;fa;@Th2Dv9(SR=mc%xYAp*YtaTZM9Wdf(ic!G
zmJX_2>_6tkcBomur0JpLk~O}OUgh=%SMkUiKXA&S!wHc}W?5tWT`}O%m^XpJ8}f=z
zEI82}7<D$)ANfRiQOeD`&R{Ade84Lx7MW%=hriv!9CQIWI>TuH&(EN^9H1_T7kmLT
zBe*9F0D0Okf8=zk7t^%?Ci->unWlRxj|n^ZgAixt3Ght=rX>jsOiO*JZxpccj#*`m
z=iiG1{reW?L%i|7Riywm5>CmOc^uu&6fg^77UO@4izvl3GZ&6TpJE!VI~!e*FqZ}W
ztQ?==`TVT{klP0%4nqhrL9oN9i2V>k!hygp^n8ilE4+9*OlCoFWDr4vL@0GOu{It)
zS}`&tt&(nzj$Zs2;pswc7U@N)Q^x}=eQK%wg?`DMmcj8KwA|p<LheP!0e2#xTh{(0
zh_}nkH3>7%6JL{wZnL$@qFA)361&y+za0$unqv&WfE21P^G!_<9y;IO4w(^R=J7ky
zx3YMN%k`8m{Jy`lpD+smhamv=1|DFdzq9G0QMoz#&k|7hZBaSk&>_`pD#VdyM2Cx%
zC!aDoQi!0NndV}_J=F`XN&vXW_O0d0pYM^DE4NpNS87JaLbQ+nHq9cW3ChES=3GrH
zVtP?X3}h`nch$x8&R;YEmr@3@Yi)m5`IuY<`T@vl1qvw3V|vAyZKx6&I;N~~Y3;RR
zd*`13d6D$aR9>W+<H<ondeL~zkqR4@%PNyk(3~n?MQDY*9!ehJu3xv0v<zluX2^9O
z@jsntUp-G2DSI~ETdeB{9Ifu=y!h>2c-GziT|d`#Cniu4Tu^FBvl#Hjuq^YR#n%&b
z<jWU~`(qL@fZzVuo;Esg`dM;axeFxgTe-`w;oDwyUS~b2wdK3$N74iVBF0i{?B8S`
z;0ch4j(j({Ot}M;4B>LQpBeOAlOM{O2@^kn!d%;%_q9RFA8Ui0gdEWHcz_?kt|US)
z%elvy7+8KNBtNq!*mk28WTRSKH`OHzA5`czcOA(MMGDz@p9hU<1Lk7-PZFRT@DULc
zc!NVKg0^M^I-~as`zj|1RRf3djdP|4RzkL5*hd%g6DXr}oCrj<Zph1i-a@PvC6^`2
z!IXoe7KJ&Oy%r@FNtkT55Nk&h+-Sd?mqc`kDOB<nDO7L!2)((T8T_J+#_%0ApxOeK
zS`e=4!=)EBs&S@`{}@Iyh952J;<AMz{~WUKgAHJZ`ph9Y=4@(z?+H=r{G(<nRk^@_
z+W@neKT~fSXK#3jZDF_(P02(_%_6;c57h%kN+pSvaQ-|-uKTpxqRG7(#CZHYvWJ8T
ztn(v`NVuL!0FVjc84H4@7Hh6aI8{&W(HgfQwJ15%OpD8F@)3aAV;}Wjn3U|K`2@x$
zc5<eYj_^8XQ2&UJhQ`XktPTJC+mZ;Q%J6q0ADdDn;m=POqZCYw{wE9n9cMH|xW1ST
zU_2vFI(x4fnjue*EgJZkT{Hbs5*p&#Htfj<!onzDI^$ZVBzPO!0Z?l2Py!w36>H-j
z=r!aava+CWqOh_kWyT`qd1r+C04Eg)!wyQ;#V(2YjcHd&*Ay?gd`B8g_qn#Asm><e
zcNDCPr}3(D@NK@MH)ftGUfx5gs_MQ-A<_c`gX4<P?l+ZY6E<3aWnLafYCN#ar+n6e
zKVk;Z#sHvt*LzY8ktmN@RlqWn4e)9Q`qn*%xk9ooJ_vt4$^gi@Du+s#S52P5YIv1(
zG-lurfvqe7*AK;~e8ZaZq&%L|d66H-ptdW2_~p~QF3~V0wBGf8W;MPml1%eLi+(-Y
zNdw``iTv1q2nZ!hQrnfm5I}T#8In4EeWyt3L?NN0AD$uMN-x6=3^QC*Ogq`{5}tZ6
z{9Ebin@Ypf1kw_Xv;^UIiB=Xvt2tDuC(mqU*BjsQGqYv!5BAO*$6HzKyZmL0CU=Jo
z_C_Yi*PcM^Evho|O@Uq1c5EfPt}34+a&7djKy{07#$?hgBtodc7%J1Or|aM306z@i
zBY01K#2S9jv7!=Jui2SS_y+)5WFAnhTvVpH@1#+qv&p@i^uZ!7fHE<B(OdeO^0-}T
z)3Zh<;**#qV`y(5u_aDIBJ)o&K*u})GUp{zrPkKU*mH=4FKKW%q!%qeY=yaZGcR4-
z&dbFdi#3oj9V|y!&0-D+{;9?Xjee>%;{#kTE4+Zt)=`bmMS!|!G?|)N9yj}BS%uza
zpH$o1&ltxmv6UeKb#a&Qcd@&L9H1mWKrxA86`=o=YXV6K;J^B|*Dn~4k?a6D_`cEf
zuL}I5`*%=Bw;43uY?U`&h4Ia?*IP0swm>RQVBo5q52MSt1k7?+098+r7EXevnOUPP
z-PToj#^DjP%crRvDjJ+w!YukVXsuzeaOjCQ+_mUb_4C^dI|4xY9TO?~`1Ae=mzR-G
zOCPc1w)aWE`y=bizo5-0Y0mp!RjFH}C8GX5+1C9KK<O_vr6MNb?`NWM7O4@^sQLkM
zaWc~A>Arf97c<=UPWwfjU+;p{Uy1G#_!)=eKn|gcIlVubUdS%<d`gb$awFT;l>0q4
zZNJ1-uIBkK9>U-L>k7I1PgPk@PX<W0B~I2DjWWG|xDAt*K{LU|l4lhv%H#QhFcpT`
zP8R=lg2caiGRmWD`}2w4YW-DyhO1U&mHJozVZ8A!Kyzo<9x1$53(wjg!=oQP{X#&v
z<wJ9S(z{kXy#&Za#tHD~ySsSpgB$Bt!(*oG42w3ZMnm`#LBqtUgh#y%eY+R5377G7
z!gVR7il)5ONo}TF5{aP+7xEPAw_*txm4p^%?<L;-`D}6hH~9>$(zDvLcaUa*(bwQi
zeCc+KCY|u-#-hyT;=w$lZ<DLkncx!UOp_ngd4nE#<#`c7?_$wUmKK?l&ubTAo&-{W
zu<_Fr8bZCNzEN3#c$o4UIaL_Mx76(36^u1}QkaQe1G^tQG4*mXwTo{Fw3R#ypjq#l
zb4bbaoa60t{fKSh#Ao(?V?8RpS)ygyU-Im|Lzsk){fo2Uy|=sMjpuk_iMtq;AL|Hu
zjTh45TLKbYd@t8;T}Bm#P-dd<NGd6t1bU5ky6anBZJj-{CKNpQs{hWQrOc|btYJAE
z2WBjS8y-+fc$!mR|I}QdWZ`^X;mj@vW2V>{XU&zEV9k{S+k)rIW_(L96ZBg1yd>jW
zSlP9{K;Op4o^<=3gY-3)>;50&By|VpR0|7{8OP$)>6O>bd)DrAE}1&F-%2&lZ6}7e
z-(;5mw0wGW_$2+D+%D5(VbtjFtGlMnj5^O>G#~cAPK7MVl&LjG8~<>oy#2k)+Nicf
z$eXaM5RddDxXUbV(`gUDzcU1<DINIBYFC0=z!vyt`<8hP%a3T?+m*cR(q7ivr44n}
zi|^z0{={x}|MHzdgiTMlfwI}EsPUlV+WraupYItOjS}x@>mrr52CtBdC2VeV9M1E3
zJ~^s8xQ(@_^Hao%ggfVM*Qxi5S)%v;lAd>RLHjp^-3K>Ia$O~vmzn2oJ*k=37vnR{
z>Gw%JJ~|m4-a1LGzGyD%rD9GUSDc4DL7A_+FL^O$P7XAgeeWE$4bFyozUyrLZuHLN
z{<zC<ZDw<Sp#nc^2qdpmtudGN3&TkKn@XTOYE8?laA1vizqnMje|)K=rgr<K5`Grm
z<nqz~l>~{{FlkG%&yOIsaWn$4tj|tRC!cWA!mW^<7Hrq&9+|VZV#F8U#SdIRCu;AM
zURewF8=bzhk89e_oR)6BvsAEHxiRY77(I}>B{^$2OpiQxK6q@)@;x)<WNWc$aU*9^
z_=U%{F5^rq?Cr?iOw}8m+iYw3gPXo8poLATH+O&|czr2X%I;2^6G-GV=34PLl<i&f
zCoJh4`cwBFgewK6^&I%pJRcN1&kVlhpPm1>;-A(ycY90IIM;N3aDGK{wsU(w#SnNu
zcA$9wN2}QXRKT{V_>IxTPLYw$Z6o#6M04q+JNeW^l5FefrVjP3Vz!Cx-^Z~5xxdQz
zul|#{s%UO#8}m5UB^Or8-<{N5KMK*I{wH%_u5-dFT%z0Mo4DPazaC%|2i}~u-k)`Z
zG6#Bwt??`lKMp+kxzv0tRCY7?za)2#O^s~7J&twHttjQ6OzLhN)zzWiQ8YKQ?M#w2
z{J+9llB}`q8h)_Tlbwt7*1N-Sr@&UNeaeoLpIXMLKObv)TXL2u|4Aymd8B#ztQ2_2
znw=sOoz=#)sfotYmBvQI6P<xSQ4Qle+ltwSwjUnHzR6WD<^P!^+b~+7Lw%~4ZEQRK
zIMydu-6^`{`~05M8ACwoZNISNpDQX1BO|KD*$kf=(NjCm?TP*I;)Oz;Ce1RBeB-^B
zE&p3V%U$3uxvvf1&V=$racYsy)6<&B`}+xgG3onsr{}jfiROpdpH}WK-n9l?@3-C@
zo)8A!^fP?F{^-yw_2Fc}?|eA$`tsfLyGx}vm_v>W37?)^INfdW0MokAdi!_p;~_=h
zb;te9Sx=@kYT)&lYOCm3SK$5rz2E)qahKEg`+$w@rog+My65-S)>`~at(B*NC^}uc
zkN@wnbJ+hDE!NerwBk{4w6OFt;DrL=WC1=S9!*;x%YV~6ItDQ4Qz*=cN5jd<?IDj}
z7-&EPxWmfnpDGWrc;HgnKuqX`%nO*LpnwoeQbrOc!!IPs`waF}K$1^@|HTVo0VxUS
z|GSFOf8*3o`Gxsl{{*W4A72w@Y4J3)H1Izk{<!EX`gc(n9LDz|oyv#C3i)MN4bxGc
z|B}N8U-wNJhWYqw#s*z_{;(C57zsvJMiSRwY@C?Q5PA&KM=(=w5@oH3h(Cyi)|0z|
z=Lb`IOEqUA*SiP4d#*!IUi*4o+-xX)rcWB42We-`j&*eObcl)l5EBz`ezNc5W5}bx
zV%0V_)?qD@q>C!Uv#j#%LrTb3!ox=%vASY1&zDud#pu#@fr#XHSu1p-B%vwO+z1%i
z3pjfP0pb@~2mD#ZON2TZGn}OY1(86eO{`)cpORLfdERtse@TLyUqBwkmi$6_VW)$I
zsQ(d#qB;#zq^<<4M4wNdG0%ZW8AhH%JSI5OqX?$@O#j{K8#9GGjMU;Gi?t6^xU(An
zhGRWg46)Z)Eqo%0QXzr(PKGN<T)}}*lz}-nn)}jsHghz>v0<tRMRN}4vQm>5yfD=$
z%dC02IR$1`3{Lo?3UH?wyD``?{Kjd)rN>Z{tQa}RdcVMhNF;j=FUgmKzs9PNBim_H
zK}>YAvM-|6MNSaW<m(brM=UT4%2p8bH<<7E5lpdnS&c|3A6&;H9@{YIoqNG6=<1Vj
z2C$)mQWd20jIX5aQxKH(O3DYK(ns{09zCy)W(4Fbz<)_-m<8cS63+Q-?0N^kRMjES
zZbA1`buI9VPQ?gYSj#0A?9KjmZUMPx;Mc4!U4)O4|0WFa60ty5)6N=ELp-)^XRb=J
z!{+gkDRS16&|?YYNga|s&XMVvseIr|@|eXVoGE$NmvjuLqT~!_zf@v;`2KkXLDY1^
zfvPe?+FtX!nd_j=mB<=(3$0x+=~y_c+;vRBtr}vXQ=grW`h1~-pQY^`<Z{89<v@;P
ztWt%HYWMnArz2@ctUu;;02eHB2A<IVok4Y%7}h*%U4GC6_U-j!t(beLr_eclRPX#I
zv;BLfR1Dp_9ab6O7U~LO#6|P4>5?bmS^Q%jM9o7@y-uG<7{wWo+Ha@7@@e4_U3zGq
z9v0d8>_4q;>Y)NoXP|~mrz4R$-u~0+^Nh~wTTTwX>GOqjj$H*0uP5CH5+VNg)G*A7
z;!jK={!6!<yYrIDG6JP6dFL&%>P~JUx=&&5xk+wi{uiV(or>(+Xp<zbp?MP{-7Zcw
zhngOIPuOG}rsq1VHAm#sG^itLh#r3}c-sh^(%fdsd&fhiA8{n}hck;kVLnWTw(z-w
zh{Cj_a@>achc?fMF5)q(k2ju?6{{Y9)GYq)dXkO}$#a;m?VXbpKT4KKP;Xk2Q{MUJ
z1`He=vA~ur3C!Db=@_~?r6S(xbGZl=5~b|kf2O5ZB9_^<+jNP6nL>pt6nK!p;1sd=
zuXty-3J7Xk1+s;Z1&L+)-S)tJNf~DH$WA;jk{!lm_%B>9(y_%KlE1(gB);-opT9CV
z6Q1=t?15I;F_CbDdMAPdGsqO-*U^Yz4@0UwB))Ql1vw)W;bdJ{L<-6@g73d_evPal
z(Ij=>4yY@zVy1Yh{+LWvd5LJ-n1MAY1=E^c#vy_Q@m{t#8BZf(YYv)%;FM>61#G3q
zbQVOfnf@iu#FSZ<nYP=bY{5%bk4wz~4DZh7NL$zS5LCo$-oF$IbP7-?$(T^2kPIa|
zK<h>P$-%EXt5w7~l7rqLW>b}daMTn4SDacifg_PCjmH8<coEXKTuuj_Bt1tL%`Q@g
z#8<k4V4kJ><*cye5k<Yqx2Wl7T|OglJiQEaG8#!z*zpc+?lc~-WF_l<@)hLO#OC1}
zegI**5p`H^V0FE84u`a&XLOJ?wHYB^bcV4nqAhe@ym2Pb*hK+$VT*FyCp5By5DA1?
z=M@JIT4!%U)w{1E5T;kk%F$bcauG1hj25z*6~f5h@tRJBgyOO>u+81P0(?Qq7?XyV
zC8dIAEHMAtCbEUd<4+F@G3BcUdnrQ+L;KwMVMLUy1t|pc-+UP*(N<!tTtEoWWI$Uc
zWQk?SBMJo?ughK`_`pt61YbVxA|g_Z5ihR_ML@9Fx>W_NpFXTsj1!49X%{<@%C1O-
zY!nQd3|Sf)!i;i(eugY{4*5h8#KV;{_XNA7ft><#mF#p>SH|S%ys&z+n3IAt8d1UQ
zI;pM<rs}+)S^Jiof?XMLL^`1N0!Fh&v}(|yrvjr~BR1do{q1WCAx^{*{_keV3L-LM
zbErs}CKu*QM0O`9M<deLzRPgiRZEQvV=BTHnykr%_6Jx(ZLL*B#BO}X6#I58K&l(R
zwRH6R>=3(RpB|B&qP&hgid`MIvsnd}PU!2Hf;!EaAPgVjkNzkZ*kCTiGLNbwA+4jI
zt}fnx3Du>nK#d|Pj7_PH7R+GTh34(Oms3Y=YKo#W1s{z6LvFT28*0tq#>hUG+;0Fs
zjfN)697$|4Crc}D?Ss)e|9U@VCqO+neJ4Jy8_(Hc`ZlFoin<HD%m@p>7tNhvhcI`S
z@mS+@x$sIVufB+qKuzqt(0IeD?ndZQ<_FyN{47gK;Xw$BPp_t3cDt~u{RP5P7s+lS
z>{lQYL}cd*#}!H-V%D`w9S;=LS$!Ut3qcQ#JPMqkdI3X!OEN|;o`!$`>tBuXy(<F{
zg~++FYSkmnQ7Q#>@?V`Fyad$qD>^R(-NQ7;ak@yy;vR%y{c}s7w6nG^WHpTMBXWtx
ziXMfKA*egs$uuRYVg9c-&2vNW5H}W{m>L7<`+hmC1$+4_@A2P696j#XQhiuC&9DCm
zM3{r7NmOXt0?R5nRp*B&#4=0w5^Dax6xUw(55*lvcxyZ?8^38T9HW3GFN0Xl<G3t@
zrjzKI5`lJ*9pZ+wH|-ySOL+I~UqQIW{5p7r+c_Q5Ltoj=F^P6=I_wweRPzlM6`K{A
zazz6W-MqeqxRK;8W7n$@*%+tRZ(y>yAm4IpBR--kS$2<L(a!gg_?o+@``Z0?xR@@J
zL|wjF>5w~1e1P;?fF1ck;6Wx5o_*r?8`{oi$0J(h(R;-p<G#Evyg_xGp4;_zow>8u
z=@pMmN&PkHpBu<j<=^D15%bRKC;tk@->%88bZC6vLrjeIJUQMM;GUcEoO<TN4q8VG
zmrso2w{YuFL%&3Pa0jHLVf`tGpRg#Bt9Ak-%&30i@r^a6^eeE;Zu22q9yvnP#PlrG
z;PR_ofVjnjlr*dlSE_iv6rs1a8BFZ{o8GR=ugrMZ?n|#tGZtH(bixv@?6C|BjNpfb
zfVxGwNj=zi>(O)TO6S$!_JH5?yHnlM{ni3%|CY>TDp5B#n*LMEvSRX55D<_uBt0-;
z$Ar*CO1!gZGKAu|N)(XbZzF2$<$E{oM|Us)+LzSnlP~pCmjQaBH@z_?1Ig~m5SxB8
zeul#Q_bWQ@6P{8O*;N@~Q#@lWRN2KY{5G$q0?e4oX#KY7#t}ukz0-=n-ZhNXq19W}
zDm4{J7`bxZno>!&9LTqn({2e4^tP?Jsxw`arfw!{p!pW~Hm-E5Da0`A90P3wwU7s5
zM_J(4zT6C7#tl2cm^Cg(tA|hGx^KT4zh<UyL$V#BF1CR$P@gXoqQaqAsO?bnx52zb
z1Wa(22-IIt%+BCvM1-_Z6qaC?csva_XD{l{fI9EBS2pgJ((k<8w2M>l-@u>rqFrWz
z$by;F@IJL+D8f1FF@M5^3qVQBBK0UmaDf7pnk<5O^tv_&H9~4QvkmHC7Djcj;0S>T
zoUI<MvCYVt&@F_11b06~VuTR04eEszCK=|f6{wJIfd%|MF$r5XC~DOdRSY$;Z7)bQ
z2h7{n7b4Ho#t|aV&=wgi|GaH0s29|x5Y!88D@67ZwP7QBdE3nHd$HQu?|RwW81H%^
zZ9%ua&)R<9$b*-m0TxWlBmNeA%Se9<&ShSI3z1$>Diqfugf<@c7%ssM3e7@izh++l
zb#ZW2z_$|QebXHpmAg#Mj`k-DRV#RB8M6R&A`5doSj3rh9<@$V+y)ezh0PyAF@iG-
z7u^J9EE8;^dbW#?;QPRN>Itvl&mo`?Q!Km?;zN)a`nxt3wY(nWm>A~$i%dVV`e&C?
zF4P!#?sr%KOg=#A1B5+*@dF4x0Ko&Sg8}&d0LcKTn*z7}14*|oCn*>4lQOc(DLp6m
zSxh*}Y^gm1Y+BOsTVy9*3v}d1rqqOa|1&gVk)g3ci^nY7)7(_^(|d{~vPusv_0!R(
z@?0E5qO%t+gkn+0*m;d!(+7?hJd!`=SNP0&Jx0EM)UB_*-0g<Hr0xRB%vZOgrpik&
z13=-%t;Jics<6x1yfNd^-*x=;iurt;n3V1a+n5yDFk6ch)&-!w1xtDcQFU|lRM|9J
z`!pbbR2=9BSX^i?f3PxhFGauOmo^F<djKvMG={<lSIamj(79`Q5xC5?7ID7FjN5m1
zN2a)hwTJ&JxWy3?*1B1}`{sy|)-%*Jj;n9K^_TvQ17mOU(5~4+^3Xy;Hn-py#dQex
zag)qY-Mr&&ph?SeH$~svPxGE$t64-@WhmId!YA6SRojMs$?KI;R$KDi$4*icLc+|k
zkMkGw7d(ec%h$``<?v;^Pa*dx&rvCtdzMLV&~M0Y@NQ_Eu}a%`+IZT_+RECM94~x8
zXrJ3~z0eqzac_{%u|PpkEc6f}BAg%aXAz(TD2{Xpga~&9F6N9xkH>ls=aWY%gyQFf
z;KgHC!UcO#8lZ%0A*gCtQE-?u>PMQ2)^n1%&ytm0SPO`^j<C&Ih?~K#XWjIHBdoR-
zz>1{>#G1ncwgO*in@7ky!p3VMb_RGg+})ED*9+WdO%L@U0skO}K};pL><mS=eKccr
zm)abnzybE{Kh3;X`A@q<b^qQUHcRosA^J=0zx7mH?xxTnhv>#_G#4O~W$=w82n$^j
z3VDlEBf|V4i5`Kj4yEoy>Jouhg3wry(pd>Qj`a5jAN3!{R^n8}+?ODz<-sDtpRwH(
z#b`FAy`GG09r$G>E{4MM8>3O!>X$ZAKTQUh|DP<RF38vIo=>6p6|TmmQt`PBx=I#R
zHS*O6#`fS%QH?ImKXhjUT39aC`@2-QAU;CqGq56&t;<tCe5W_}KgiGAlvWuj_o<()
z*w?~cb&H-i`X1L^L_wJ^CRnnnEJw3K?(Dqxj(%o|`P$ba@ZIh~!mW5^tK6&e-oKyg
zQg?W2xMX+DI_dudm(FqYRA;?+#^3Y?^J~!ccfWLfw$lX7EGhB59#T)#ipK-m;@kbw
zX56UfnUPj*bf#uWNOg8-$}_bqS!?HFPN;nrZ&uc8f3;8q4e@QnLnS;@9I%?z4&wAJ
zs0{xHXnY|RI<|u6d^InU#I<H#WVWO)Q$TBcYpk4(Uy0Jr7PEz1VY$=p{rW;HyGTE}
zE0ES=t{7WkO9&iBwk(^lKk_YV`_kfpeB_&9#SoChS#C@C`)oh>I%{rMHBoyG&Dq9@
z=x^$67mZesl`k5||14}k@TXRgp>J<ZxS=nJ?f9;DxZN*%!of4vdcL>*eTz9WFTdO6
zM|gvt4(AcI9HV@1BfJO!5ur8@d_RKkp-KXw#otg=>0ovud=klm8PC?@E(Dw_0_~^`
z<r&c-f}rdN9LL$$-8if3{iVl#b|hLSl7(|rl0shKdW&%U8JY!ZT}B@`vH2&4H<aO-
zBxA4O^+Z{XE{b!R<h(3R0gAo=B|Azkd}{gwV0Ku1{(k?3pnfE;eb4r5rtjLLl6p`2
z$g2Iptz}B{A8DfEeIDtoLU8f|Oo1%=dGw@psyoM19o+3U_bXSYu^iTOwl^2}q_6*f
z4&FxR+%IMxSb8@r3Eo7n(;KreK1Dt`87O$w)U-^$iRo{O^*7DBoozga0S_su@r8d8
zyt>Q?5ge}UIS@sqckW7v%3jwysizda#8MtrqAttx%C&b%^YiT#;~DQK5ZSL6-@7V|
z0(isn62a_#)*b=tK6EFp)tdml5QchSx=YZ#kA6b04ifbo@c_d-r3PJZ^tgk%M0J}(
zx<sEfWOwV<b!;p|3Jfn@HB)CF6BxV!J9r*C^_YErdjD-!3z;v94W+S2+t4g5>RO=t
zG~WjsuDgpw_}$Q1EVMUfAk<7hA2jCcS1)iN;PE3sg<k^3FHih33$dg2vo;C#_o4GI
zL%r|%<g|;UTT4%lPU#`FnwFm#7&N$F?k}$q-0cBlOxv61A8s98M1b{xzBee^8tFdF
z+VTyz2EJ5@cLrm7R=?M+3;8PQRpZ(tc7H@TQHl#R%hUFS?(JNAIUU{{75YV8h@t4h
z`sJxKcR%m6WGPTWg<+l*{&*%vhjjfiqpld1wpadmI`!*hN3+Qp2H{ENH9c8NP`>xw
ziU6lis!sp)=}$s*5W`84j|MDo88=um%e%#2ja+j5G(n&R2aR$4bk|oK&(eJ3gp>Mh
z`V`n&ix`DTk}LuYTR$b|ndlCF7LP@{ep<HlN(NOPLt*_i^FWP@DhG)dZ)I%FQ@})L
z5u$u@)mMPfgAxMbk{u+xe!3%S_jO_r<B1w3&Y-BhRL!F-GD9?(Wf69vg9|KyLQs<5
zn-Y$XB#T8DLIa}a#wvgBbq)-==wI%PVx<B4kM&|`na(Oez&4oNS>)A09wR6ZS`YsA
zVAIO^xFp-ZiUSC<?Zt@sq6&?j7}ixg5d@QY8D;ZNqx|Xw3q;`Hvog0JW_LTfm$azr
zGKL>$eRpM|^lvKP0l2lVvdOC^`gKzfh|Mzu8GgBTqQqOMaBY8{*3IK~viE~+sjFte
z8L84VT^+5`v{^-iSS(xAJTIuDL6f>9X)iVc!_yX?odk%$_u^ku{`N=&8SgQzKVjNq
zntCF*mzQ?Va1)f=i*5_b^?+G_G~8cLIcHcez*bn~MlYz0)@e$q+TlNV9?+1MHS@%H
zFZpXI_3DUYs(0Y-l${1a5j@xXUL^*hTi0Jx+~l4$+x$GSG>RJO+;qB}=Y7zegF`zj
zQ)<LN=fHb_7W=_yd~QO}LGU2(wi01e8*SC(32*Wa^wcM>uZ`Ad!m0xv(`fy+qp7H?
z+z4;3>x)sH&U61zP9j&GtL2G(r1$;R>!K0LLP3JCJ%aRi2Rm9p?oI49*CT#Xi}};+
zZ8b(Wa~Zsu_2k-OD?Lppz)Us=1x#H3R};3J;W$xf*^-1|p`+FLFmP#e{g<PqTz&{J
zR(arPV1jti4@n$(3@T`!pPg!rE+E<*r&gkY7yqsy4dcN$l6jZafwIwmLQvzoRBQas
zpjXW=)lj1b@&n~!jj|vgg5Id1aJIWUqdCLgyneb&vNq|XW+y?YIf1J)-I_x1m&5&$
z8{Nq1a2>ML=R2l83MF@4Z3w{#_CE5-2rl)&hLG6!tr4JRdJ&d5GDFbM2d&rjYIe|W
zO+hh8#uIO&_jcj!5HC{+Cn}v?MhFIp1&OkYcuh4#DATiU#L)<z>0Z0mIFQP_L@7~-
z<EM8nkXhqQPHan;FsW`SKH@hyryV4*v*st>=_k09tILjsmQqD;CXq>_j-gnRJ?=vb
zPdkO2c>l&NlF4y563!z35qpY$ayh#y(A9d%(#C*(mxu9<c~m5<gn7`X=<<_TVRtI1
zZ1TvCu541>34oKqjcb3qaE+ad%Kga=Iozm(D0laF5<8F9Px>viKcJ4b#L4eOR0CjB
zVeI%_qV-i1ry5$k72PQ?z>J&HCH|#|M>Q|%-Amt_^yIl4fy;20(g%b}3{?*8=wzmd
zZCe)nPB68D%yuBP0ff*gpO2Tx>je~LlfIy|67!)e|C5piHEzX{^D1Ja^*Q5g2vZ4R
zn$s$!HZ9-MxkSEBL^a@~cJj`MA_M`?Fm;u3`F%BvvpdKjlR01?EMXk0aB?eEWvr5V
z-yf(${;x_zgg)R$xINy<a1jJw#+H(VM4+~>pHBrAb^jw2iHLB&4jw!$(@2CYgIw0n
z-*rTFg;>O!23$L-aS^iw@y8<qOoX&^DoYf-k}yUG51Ld9n9wYOf$elGiQ*h&e<JEz
zLQNTrSVk9KrY-<IT0dVYo+!op5TfCXrj^!uo@!SXqOWFo_l2OVND{vuV|(EFS$0u9
zHg%iB$?!!m;xvnjKP^xp2%(ZiSDg~LCr9=g!vQX&=N5F0^yPRXGSF&nvbSI9Qk|Ev
zi~>OwUtc<x6xe#$Q+TNR-+R5WWZPLFg7?2(aKrv#Lg@2N-<zwUNQz>TzP$-j+B7^v
zay&2nwm+=}1s;p*_uD+h5zfU}by0@JHnr5C5o>|o8xx^2k9JVPs|rW*omao?1*ln&
z@+W~~nsza-iy^ht;6pp|saHpC0(3b@>65@7o@(^4lCO)LIN$Qp?xh+54Te1Z;Wir6
z6;8zx#E@O-Wgz>@SeQL|?yG3fcFEUG&O%+_T{iibfq!eMO_=5CYoij2FDmuT?(VNO
zYLY-@uUs5zaG;}hv6oJ73$TBbOa`T|-!}|%PGZd)JJ1zI=oI6x-xs5i|FW*>yKz@z
zb0+`&$^ax-Qk!}|{O5#ovgFO{VY_^cn3jTR@exZDaZjzau+Q4G`g^s=FP@x}Xl~$P
ze_Y+4ns{ZeuDO>NnmAWxx3Sv#v<(fq@+KH@z0NWx&`*I!aEB-d$x$4da2HJkMe*%#
zAEDAy94g&afgjNvR=1rfkoyNQV4*KkK;lbq0+-Q#z~eExS&Eqc(Y7&uaP!_K6~zL^
z1ckIg)>BP#7!;*bZ@C1&>p}GnA&W;L;V!SzRnd?Xkw+DVj-iGtYNumVN}}}Ew6QrB
zGXVw$u#5g=k&)>yKmM#KN|~-eE|Wzvj(kmV(a2EP=dOn0h~!kmuXc9|HpoI5*I%9!
zshI|B`0aX%qbFr)lDM1bu@v>wU%N}P@ZFHCS&#OCR~?~UwbS9=$X8SBV4Mf--;PcI
zU)sD_<DS@-C_}-$x$xAcMJH@~2B~ceye8Ck&eS<(U*0dZSitGNHR#iv@03ZD$<h1E
zDrgpTe_0phjoNn_Q5Ex<;+W?7Q8{$!ypY15!k_DK!<_T$3{x66CJL8c=X)7ieZ2Cv
zk7pAE4HOBliRma4Uc;zO`w%G@HTPP3<+d<_&P2)4DhNTj1#vpX7(5u`v_(b<I>jW&
z%^IwXl!$JCE(Ei_oFGW2_+?K_N_k`lxKH&)Xzo;0457op0oO~~o#iry2GTiQB)Dk~
zY8a*5#%Pule~X14hp9x9Q|!8!$h<Y^%<H0|pxrIgO~UAD$<pbq(KT=Xg~2%)EP@+(
zgN(we^XivE80Rtl;AWZ!hF+^1;i^Txm-S0Z!(Y3zMh4Gd7nP3UeLk!zA~D)?`jh_S
zZw8%>GW-V8RWcv+n_n~1J<><rp=juGCT^fus9blU=6yGYl#H&gEl9_54__`1{oA4)
z{d@N7K1*k(%y$}Lfs`Jqwk#RF^Ioa$(Q&W$G$`)j(7G#`OAsNhcUQj~+ZW2i+L3x3
z@>pYpH#w-Dn3ogpLcjO*L4g9%vSV8TQJXkYJ%Eb>tO3WUkSD_UDk+9l7hi{0@DV};
zW`dJA>-IJkX{UooBst}4;|aBjm@u3fWwp{mzul!GwxIvvSwXmDLH4+p(PG@{@VL^G
zpL@1axgzzLQQT`|33eK&ib15Cq}supYF3OUUd`>}{rJlsLYBGjbc7^pLT@xP)5ekd
zYG7*QTw__QWX*%G<w0+9L#=aP;&|00ID(Q^KUa#>@suR2H)ni)Q#-qWwvdHu6MEGS
z-K>evyse$ZM_%6~1@tb;&3U7Xfe725K`P@-p65Q2ZHGYC4dn#NSf<{FUVSX?V}6nV
zLpm$<)^x@eDCySa%&GoDC@8Av*<gL_Md!IhS#XOoRBD_cYe;r`mQ>QTW@grU(P|f?
zp~PPE_JvY!<6BOp-xO(A&1uUM@6n*6Bw^fULmKZxDRzq$_~2=|TDT=%FDv-t@kgvV
z#-yCsikZMb#ZnYbboLr8yr1yzp&nxURKJtst+4%*5A7oDxxWl+G=!dI6w)bjR}g<J
zHQ3ZjM_d&^YJuJaD6Zz5G%>EYz6_IGaV=DiEXAt-_X_AWJ5=J<tIJ?IV7shOYALe_
zLcm?4w<$`kWe+Z>e5;}~T|I95B~Uk4-(J5FjmQ=|AvsL4EVQO!96ce~P5CkP=6!?U
zHBmhwueELGkNwu1N4)mf9m%Y1DWS3Ai!7@Yf+vB0?V_3j#E}nrQ!sL0FTlc#=d}(g
z;OtnN&FQ*5J1AyCN7t=aeIL~h9r-c#QENQoXCuDw_z&|tnF_3|xuk43BmxaaaeF5s
z@Jr?Ad=M>>)30Les4YWFfwK3KUu6Q18MNDt`cSr@U0<Ki73y-zS8cWkkT?8#2JQIj
z8CZhI>+;*`@JYh0<X*25<y_a@oue-USX)&};D9!a0Z`ZLF>$!hRCshsa6bT^NtB;l
zXS?D>++ZPiZkKV3YQ@ffJb3&_&ghe?tl0(BQ+Q@G6~DgWX1~^4C#yP-Ll&JIvMH-;
zg+!vrBO?p9I+)&hc;!vtms!UVun%|X%F*>A_(j&z%614D4JZVVmbLY9Pqbr@FkS-I
z_7~ARW$JtAkUw$G{ORY}&n#V5sm6!Q7sMt9ebM;S<*K|1qu)(yC&H0dCwCO*YeM8N
zet8?Fdj^Il3z(0Il@2mbHAm?RG1$2{UFdKJy*@EB<Y&MbZTP@&x~}dd(Y+F|q_c;Z
z$z!oJ_H64hnD@O?Bqu=noJ3Nb2f@1BUNrA}^t5{0r>R!dCw<iF=RP<o5X!`V140Zj
zPl+)NZs^VTWU8-9Atg?11T%-1oJXI%(>Y-mIH;^2EGvZF4SqyfOXER-2%2D2pCwV8
zcNq(GfAZ!^?P>J>kXBYzU)G{tqbu~z{iKe2U)p1cc^(*X9VjQYr`)?bMJ~xZC)G3k
zVgZLc$mhfpls5Rf>mZ1AgmA=~-hchbjy`<dmuvr<Ast@T3>D@{%cr!hKjeyoO)^(q
zwQ0Ry<Uk$kGyX>EE}=R)qt^2){xyjW?{!G$?|=2<Gfcr#y8>4Pie!E~4n!L2d}-@h
zAr1JZzt)HYf_Q}b=zPC9HoYz7-NorBomMgm)$?2YO3_jN_*+vH%5ScA)f#>ke<Y`l
z3x&z6+-~It=THqD8;>L`=T)5a((*0EDrOb~zrm<vHJ<sHBRr~Rjp0gks#3KBPQvoy
z#vkR6KQtni+>#uQxMR05Iv#qtFMYF_1uB3E8YSua9XuX~oR*G@ml}N;CQmc^GDB_*
zq_ji3-icPDry2cTN~+a<B+mB6`XzP5YCe&^q1iO~q>%tBe^aC75l>*VOZ<8f74!Hl
z@KRYY-t=R`CWMe@SHK?5PHFl61Dq61tOVA*F#5ekF53KfzGXB$`<R^7_|)Jgy|j?d
z!?=RENvf)tF*-2+dwc2O-)%++=7pz#)zL70Q8L})1p1h!7YHrIOThMs;W7yE87V$L
z@j+PPMycz;Ly7KqDnVWn1zLW6#N#Ie#4!nix$dI9qe6puiSWK6(WFzDJtS4WO5#Oe
zxIpQpgZ~W($fM~&+i7QD+mGlhkka3Ze@_U+JH%Bw6-Bb{^kSraWoF=o_3YrHjJY&_
zhwx(e>=<RCZ+=+gHr7TessU~o6*Y{X1M271i<>~K;AgvonnT<frB-q2{LaT8pB{fH
zb3#8|hj^9P(ZF$Sz$)vIdr$3$&;i~)lEAfjx438z{Twb-PuK|j3#SQiX8E)m@q(@f
zj{ZI`sAc%a<KFvT(zSVY(C(NhCfJ3yvrGHu9T{n|e2(0@mkC$waCIk(3McoG)AyVd
zp+l5uw{_dN=PK%<JO&@R-PXHE_Gjk<O>r}R&!)BtPd?B0Xr7}I%)m%Be`*N(AkxuT
zg+*i)H!N3xiamU|%{5AE5~YK8IuyTy*D$2E&DBkNf1Y2|)mJ5rbhb|M3e&LWSie}0
zNy}QK+Rpg0s3!=|sls!;IX9R<?-=6U<sx04(EBbNaK<rVM*?XUd0dG;jOZfTSFZXF
z-O;0hr#&?k?I^AyQNqI3G#?ion-D#1=ERWqdgXu}*E2_TFkF73tm3FIwfI8YCf!an
zdsr{zk>Pa$=i$LOZ0NjR^rP_>O>!*HFlXy}k4cP?4PcNWTwnHFhIKu|%&Z|EleXBJ
zB)9fNfAype!lG<E!hC>CK?`hil_Ge{({a%krB&g{hqwOP(D|8^M}{t5T)PTryIDi5
z?Qb95tg+dhN8s=*1E0G3@F-w+Pk#V>ff6WagjOcTb*Xf4Sl-EVTBv}9Roz)aPaXp2
zjR1W~t`{$h0%A=Fc+R7iQ-dNP;FSD#cSk!0u&dD#+*{&|Me?9XXk7HO>Q{x(nfKQ8
zt=>4jm{icg_a@iw+BqSH@wR<VcUcJGwvz0fCV^&7kazM;jAqSc7UWqFu8rUqjV$4}
zKS58y;(914Jez$xNWI|C-%uyu3sXSs2A53#xQb^p))DobfZww|E?eBZ*>3gmNV-l%
z%+`)ygU{$o><2G1AOv2HpgI~HSH2M6@zsvqX{Q^(*bjCbvFl^FPNiN#+dp>?_!d=C
zyy8CszG??S&|jEBVnVPFiT>&yn`{)d5|1ba`lVE(XDKt!H%JT=JaW2Ki)CvkBPSbv
zv;%3XmQ~VI))}Dm(mS5w(~_Qa-4!DQ1ti~;Fl%3+-H6K7pTL#w4$!rix4hZmzfR<d
z9Q&TnX>wP7W?uIOys8(T^p<_C&$LN9#d=y#Y~nHBb9WhpWn?c3i+r|5H565?`>-ig
zAczZH_a;za^xk0iiR0NQ@ceAevDKR0UEmq#16;*9Z-gdmkQM4Hi}iG#FlD6{m+JjI
z#hF-HYp^OTD*te$*u|efPV$-gxA2b)9X-V!1Fh)9jfpkG$8nq+HXd=p8|>}@{Ti48
zZ9_FHbtsb=b^1TEQMId4+{ENcPhKi{H+T?k=z7MvZ^*j_3~LzQegbx7Gx#qszMdM_
z{{T?aOv(GomOD&mgAnv1<6SU}cPPxkYLUCf*Mj3<y!3#!VJM$#FE;bI{DWobQ^LR>
zG%2}$&s1dqCFCaPUGB}?x%O||8~8J<@(=enUpN6*G!`Z!l<aMg4Urkg^Sdl+xnOps
zQygs=eW&ivvQM1<cmUrAb!ZrT)^Q+_2iZaaCT19qC_LXUG-B-qj3#ojATcF6TdhPt
zH4u&@2Yc-}8+~g6X+0d<pfXhF&vIJscM&y}pKbVRW`(Na*}9MF9Sd>GG}K8mwe_||
z12km@V_%sUF141eaI0AtHa=p#!M2k2mEve}u&s+DF9ydVnX6Q*gI<mxZ1>8v6Fo*y
z^xKV`dq%GqS}iFzQCdui=Bujv&I`{ZHXlrIw-91pf-tT<ax=RdAeNYIidTQ`gF&qL
zlO<&Zc)P4Be(y(YCEq<?vW{ZSW)YcsO$Nblla~pW{<qwAB%tywGS`>4TziD{=i!F=
z?Cw#C`On;_9c=?IUQWm&Aw-t8N(`EKJ1^D$@K26@J^`%Pm&E(yDDgY|zeewlIVO&U
zU#f5Oca2KW<wMp#t;})=W?1RY*0yr;?vm}itm5A5ALd9lw_UnPHvTeh`#INtZvrrU
zITSM}^sb!zUsWI^B(TL>Fb$LMnox=W7k={Hz-q}WKr_27bSjQ`o3EWUI=WQ*b=n;n
znUCeE??4sFQ}DH$FDfukr{1IJ@>FeGt7`lAL}A=TV(6Gie{VqZR{Y`LJWDa*LZ&wv
zh=ncj5i2Uyo6ns4AUx5oMoncgnZhy0elPb2mdml7!8FBj_32D+(uA;(f=)HOjkm2)
zlV4P(XAGOyPVau?OE;D$FeRD;r)(Yb>$|REg27d`Nj!G=tlT{(*HM=s)qW9gbs2!i
z-2|0c-cH4F-oS~l%D-ib<?2GQFxrz(M+*^*K;$EYx1Yc-lM`??V}Sut$%2IR$K96C
z3d6bNF(Z>Rj#{-6&`04>mQ8%*N*Jk7-2X(LQP|OCve0u3JWd%3A1w1rC>r1)bf){D
zFPP}wV4@8&vKhpS%$|Y^)SJz;SlVe|Y`1Zbfb6(@`Sok++EvsD?NCU0<7z{!Mr>wU
zFG&R^@#1{4U$%8dWpgLrM)i1`OLYHw;!Iere`fGf>ldL9*EK#qC*S$Xrhi*ZbhdK-
zGZE97z%%X%0{t`V_ov%{KD(>xQQS)woxQsJ#WK)NwnCj36X5oiCK5xJ%>CHwLWZ3Q
z>YpliP4<e}e~WN$Ts}W9|KzJ4#W<gk|A+rZUgoW+RDCv}Q3z)|*TMKf$M~B#j1EJ8
zX#QL)=&f8kXfW<@8>=jiXdf##?g!UK&ATZmCUjn%sqG7Mvxp+frt4b`xh>=_1LG($
z6P?*mLdToMo8U&NLgo)5QuWM!!ar)V(XD(esCy*bks5OB?rW&iZdx|N2QRgFE<Ckx
zMve+dvSxnaBPHZEU9}vA8S3*M@$O=s#*J)XxdL4m5S#WTu<qJp2yt6qzR<@@kDDi9
z9Y%M3pCp=_=J@9Ihe^WRIR1P#0y<k(RcVpYn+$PKLfkCZ#&BTP9e<>?sN9y{W>;77
zlTy52QLod6%(D7~E4DTn;c3h{!awquRay%<_>;WOH002dPlS4W-pO5jt)aHiPqzp#
z&l}SvwZ8(X6mNlpk#og?pMLRY+$9^_YPaxBs!r{n|B~c}Sq<;vq79_ZP@<%8RPj&%
zuYlGuF0ssp<2fJn9*erxzr3QC0juSaN|MN9azM#*f|QA@g_rz{Z*BKd{a$BL^P_rN
zDL(%y@IRu-=uttWfTJ+pQg|s{XZVR4mLgp6qm?V9x4Hy4U0Z-U#kJAnSv~EphFNU{
zsAePRMch-&ul5{+d`3X?ftVUE(!3{7L?j>E)Ec&)+eT72Q&#IaUMz8VfZc#d2JXI|
z<IQ`!rgl)hQ$t~X)Yc;4BgB(!V~14F-v|};q49Q>1Q>r3nUagwd52_m%~(kmq;~RE
zr>&YFl_kgTZFT9tZTd*|qHqti?9iZcp9KkXV4kmCN8qER1(Ao>3LJ1!%})*#1tt5Q
zf1wy#D|Up@^+<`>SY714Pt_Qjt4gh1g^RTnJ2cOBKY(l+0C5%b-_y6LrfSU_GIKEJ
ze<x1yV%~0kc<X2lx|th7Wv|JwYakRGmlRws?Ov6glsO2Kn5(aief<DBbL=&-5_1=o
zCHp7bp2sHR<gAXABifIRV9Ki1yZA`whw>M50$KBKOzsr94e&X2kL<dWXBRO4X(1Lw
zSNF@l34dIFKQU7iEMITbBiM_;r~fqp;VYo*-u7tvBXJr`y}BeQ=&+-xw>vMpyaz;O
z-z?oZ*6g56H;7|hae4k$D%u3IH^#cpy-03a%f;E9rw0>qpJy{i{obnG&y)VK$6-}-
ziiX#>huwdxG-M_%P<+IUxm*6{j~n~?zh`;?6Y0(KdrJXq%<lkmdI-xRkgPlSRr5Tp
zCQ3GOOV6@W{pp~e5_Y?=3xIDSz+KO+&<2OnZ5r-r^9eM9td3N`3bdBTzsajk1UFFE
z*BpkxUC^;NtpbIXsVQ|0L>4Zld1>Z{g!Lsw@nv82Ia31b8pyOj`G4-tk;#@Mi|c%9
zFBkWZes%?w7#qU|9bI$(+--{5`O%Go+<kx(G~8J~-|$RL><d1eqaolAu>%eU0e{z-
zcoVpg*Cq0xZPi;}x`6dRUS|>h4Ursi@D4DUs6@*OXV&X32F}#P1TpqU(`q+`o08!r
zel{}}M@<}H|7O|4&9k*Mu4}BKyT}muX@&K2`fs(^)bqkk3?kdHsbj}x4H(P3bmkxt
zQX)Ml+0$3O#tbwiSm8*TE?8kp>M!`qnzT%inY(Q75qFSEdgJETbRGHnNcr@L>PXJ?
z$Yjr@FWt|o`(l2jE1xt+TBWD;3U?gSOYEz<id*wR51mMD1qH&NDmvJ$TTzbq4cb(s
zI(ci>=u`L$&QQU65>|CH4PBpRcjW>9=0s)jCRGxwut?l>LUdX+{J9Oh-o|kigjrVw
za~ZyvlMIklA!6Q0R{_}yCcLg9U+%VQ`7`KI^idEgpPo`R)3$iTD9Vo6$NI*>uJ{y$
zKC@fdER_3N@58Kgp^E)%bN>^*bON6(<WX%T&BN#)Zi*C6zo-<p@jiH_sROAMCJ5U^
za-;|^QyQ5eBV!*ogT|<I{yp9iWRX(c{QRnbVk|79^2YQoi0w03MOR`W7KRy=wiDP)
z_TGWX#g@!)tbYRh<C4{X{72>XNZ?GSc0VY6y~xo<;s2hmS_>iT_q{Wjko@;_OU`k6
z^i1C4hx0AL8fvEAqJd?WMp^MZJMw~O0<|q^A3|`{uuIVfxBUl?zhAj<`=Wg~N^~VN
zsaDuVTAh}H1(MI5_U!y{$18=#5NXX#RMjDv{i3m&qJzr|So{p0lp|d8$`wYx%)FnA
zX*c<1C5rnzwgJyRxiy8%*EMI})l$EH@%L9wyv{(55ez5g`Txh)S3p(OH4OuoknZkA
zx<eY2kd*F534u!}NM5*fNJxp4N{6(Bbi<`XP(Usn0@4llKj8Dc-~0XRU;kRHIs43>
znLUSl_TIB+&z$|uhDJS@5azd)Q-}J_!I0nibr0xy>-Xf@^XeVziz?&{kvZ6*CFLi;
zu^TG`Yfa6$47CYd)tL;z2)2UYz!DkS4{Nrk-<E%k<%9)8y7_D;1t++fhXq@?m5l>C
z>g!PPc!0YJZuVdY2p{acI`$qA!XBnt-ahAvk@<@%KBv({AL#Qyw!-H5-0qRRHsfy;
zvot1n$-M&ez_haEvmFpj<-P|q5a-mzb5|a<rnFL|OYSseRO8-Adp1AIUg(+$?9a|5
zQF_f#>>t=G|A6h!f>Bq3CEVU2WUxOM_#)5Q&ejYdtT1oxpiSzM_Dp;9$?9q_y+g@~
zU#je*T(*plSzGgIDRvr_Z^>X8yaBU^>^b`11U!oIlqgt+*lC^GiG7fV-A7#Qezeq|
z^Ss^{`1q@R7bez_ENktOZi(bXjw%dlDiAxXzi#w)j3OYAlMKpbdnPyep)dFWEa<Y%
zi3TK0-o(ZqUw>WSrb>#JfRkP&B}~e%^~T|m)$7ACvjm5WKTTsq?;%_LitlFKa1Nr4
z`xLR|AA(`~w_>FNy73WYAT}EOSI7mF_`?|LEOaKwC6GBNzHDhvx`tHpR+l0Si8v_2
z!1PwRE&nG~d*0H=SCHaPMYmw3_MOF#!w`u!MM?RVwV>sUR2MY$YucZ9=`vEa!1zK3
z|Gw;Ys}~0Zngd~p?@UDKknrUy)?^W}RQM<^an@k^*kE^lOgfX>(*i0j$g_2;zBmcT
zrPL*^`7v6d2u@Er1MpP~r=a(3X8l^wH^V3N1W4LeBcq3iMQ6QWK2P-$S%8QTiE>?m
z2XXaBkC-GKd&Ng`&IwO#J$*e84}TqO%S-6+)UP%jhMtix(|gF5x&7qboU^DO_xqSN
zhLE|;kd2s&;KZnQ+#R0d^^8Wqn<i{29hEc5eR3|S9Js-6!#jBxq+hWoVktc_v;@m6
z=u%FR(0kt%*Z3PretqBS%M_NsZ7ny`u)x4IjnK^K#~vEh?9!T$ef*LEmkE`o4gME+
z`s)-0o(%#s4rp%?`L8#HA~Zb3g}pJXx@nY#th>#`<T4~qW}c%tT8>w@$n}MA{t%yT
z=3Kq<NRRtGlsiaMG8BjOGNUW45t)m_|FiF#2k4D>qtG_br)SJ;MQIGWR`Xb{Q^f+|
z!&{%ElbaqFtDSPDoqI$nrUfF&^Nw#MnfK<K*5Kh8^qHZXbj%zN@J-88$mQKpF&c~^
z`bz9~`QF@ivf^trbbnN%X#&7;W9-+EC<zfIK$c9Jis!)X8j&Zjzy?tT?~jjWOkJ!|
zZQV&Q7A~J3mPMfIQU7^?Gz#|t*Ht1O-jiCC<Hn)-keTCR0lyp=O6%wOx+gi2@RV$;
z+&*G>uYDtF^MzjAt2Hd<i#u^?W(KAmezJuouLaY5m)kYC&d1)`%z6#9H%>95^KT?A
z^xL<PN^MlI`HskeWZn1Ii1c|S%izR0Iy_9~%dj<Ru5;?$K0fzNq5xjMa`>Md!`Q4o
z)(|tM{(nIe+J%3}PsV;$$!s;pV3agnDOJdiD-0Zt1K@nJQ<APs7$#TQRVa{rK!6%-
zMSya*&5Rmo0B$Z}?F^QUM6P1M>jq)mEocX^G)?P_m;eJ&?u;A%kTb(YzhjiAzU(5%
zRE?Y14(i1v{;d$NS`NMcXg@L_+L$5--3`?Z#SPsJ`N-97z3uS|kv7(sU_os>z(Q=m
zJCqmbuL^<n%~c=)@xBSzt`H|Mv3NN=4e_~GBo}K6S)*2DkUTFKl>qB_9)&NrVwB`|
z(lH9KAF1TA8|=emEDg_qz`gnBVnjWqcw<gsAA^e7?i}sN(jS-U8t@;L^PfsP753>J
zBS~i}Xw}1?5rYB@(hrj|jzQ8%3KG6ZK7-xu96rE?aOLKS)6i!&^j>8Abc!)zPCWs~
zu$i36(uZ9Bw=KS(Ldo~hW!lNlk&_=i?o;%Vx(BPD>1-B08e9MVQ<_F_Jia1utP(BI
zz9Aip9{DCRJ>r3PBsiiNNdTX@tnw;m8_flO{EEOBe-b&Ib{#oZ{BdX3V=kJDuc>lF
zT`a#eMkm_Sx@fqr<8q40Eb$K3_|t9zP%HNT5Fqdl{sH6o2Ww-FVHN(RkJ<+mGo`u*
z6v^dkx=i7s?a(eKF8_d<RJjbrV~Ore#UY9A?k)oeVFZPHBw@s`!8wNPj7_@dEN1_$
zHTL4O%Oa_a)%`D|yn!aWV;cl_?E#=_OFN#168*;Doc(b79;y9aK1ESF990aju7iIA
zoV2?!P)f*Uh6qWuFMM?lDMYV55c>6Qk<i|(jMU}{rMf4<%ns<-j!-stBMKEs`pM&X
z+=J>)zQ8AOPy)Mj!(a2u+5@G4lkP5V3SEGKtUK5Tjx@l~jfL2%n07>oMABhM4yG8g
zyuq_@6al7sxR;`5G$~zx66`q=9voc8_qHqCorXBp`StIkEO(Isv^TjV;_|_yw^rJ9
zVYv^?K=!#P4Q2a<vU=}4v{6dr^GR<%DUr@};A&%J$i<P~PUQdmkV_{FF&V@1VyK{<
z2Mrsnu1yiq<`T~%o9Ch(hg?|^c$q$fi`71+-^2jrV#Go)Aj~bpiUuSppL+2-IL6Ra
zoRoPRU4>k`Amml5Tu*&81$NO2$~L8kIc)~eQ(syQpvTAm*#G*eb5WOP;$=d490Npx
znh~i>wEqu!$V-O)e?e}$QaY)*n!e$Abub_G5>2_bZQhSD*DMyd&0O<Oq|*V$YWKt}
z?uKaZe~qcfqcOmuy;fZ2(FI>JbcK(Yl7eDq!N}EZ&m5je^OXolENUAzR0XqAbF|4X
zi>X|rtN!As|Fk5{^9-#L=$LMLU{1I6#E5s)52}L+sq4!ebX0#d;AS8zy%aVYhuNK?
zM=iFe#VJ7usVS=5$*o3|#M<NH@VI=ZH)_+@Ntv#wT@wX^3Zc6*y%jTbbV*%w0U%4!
zeR~j^=$`4gZjT9Bqi8sig-h;Wcq#>kE<BjDMZe3*7G={6<%Lv?bnspOVL<gU8pmjm
z=V+Y;dQuPpQb^h6T#kM@@>?j0GE|%E@|~RizH^N@QE2deo&agL(RfaPTpY0+lL|_P
zD<_#9(v5yU$kE_<K%6HLvl=qexE|uAeYXzXtTbySb&;v)G0R2W|JKleq8CD)i#3m8
zZ*biBnCIi9=8QaU5!$QvmLS!jKLi#3D{r9Ib_cz_+~MrcRKQ&V0pu<GyQg+lgyEU|
z-8*1~b8Iwj=vD%i&z+_~#)s%gM(g-a!^Rz1w+Yixb+0LNp=g|B`&RN;v8LnW$((sT
zsBKS_%NinoQ@q2URmbkHgy7r(K5Q9-;~P;PvIik@{Bk*Hv>cNN{8*z%l=Eu2B=Y%*
z>mo&1FN8l3^v!6)H~PF9!hrqYFGxTkaAd?yxIwEmimdr$DWx5>Swk?YY&8iibx=yU
zttJ&@#t_UnrtO(%L&jr?Ke)M0ou2VU)<Pt(^BJfahEYto9o|`u+>s8TqynH*1EPcN
zP^InKBR@0!P}Yq67<v2!asZhZ)yG6n3lotc<9ktZDQJ4e0Ejhms9G)xJc*FNK0{jP
zK1Mk<hBH*CQB$jtg(H`6t3!~vG#v~v(ra>I4E!&xX&7cYNf2L8(&mubl&+g0VpdG%
zvX<?9kT7L~%&LmsOo!7W3}C7Or_O1asGA&=-s(P0A&x=HOgM<aur@HeINN5$0^-?o
zO*xg{km3>D`y&9yW?;tKw8^c&AMMsX^|z`w_7L@4Sw@;q2QbhD<VYdY&Fj3FsV`d6
z)j>eh@_a3XsQvw$Hs@TNO=dmFQ!wp3QeC@akMyE8pMEY?Lpu+c8u%On(S=9|+Zc|C
z9G3%VH<x@K^vXc$IL~(v3D}>VM-%J#>w+IUr3S57p?aAfFw?&hc@Qp#pN<6NUC5#1
z!6XVqpw)IhRzQT>vmTayMXuW`t<p%c$tE}wz7I+ils!TGzb<_-;sq^b=V!jP6q8nm
zazq%TOz!&>L>TcR?)&qEeZt0T0`zZz90pqFo+BQUA!H&X6ZOw%@m%iv;shA+Aou-#
zLX3Et|GM<mYCcFhcg!kc`Z!r^a6Q!EtAw5{w+xqOe(PwSCV%>-T@@LB7rgHPw1Mlc
zZfJB+Q)q&NRv;q6Nd2KOQoKNQP&jI%5uhfLE&mK<$;JS1&gJKdAGQJg_adRnZ^eCj
zBcU7b#C^^qZ8Ox5G6^t7DctvC0c|g`_9~ZCWOXs@5;mg5@2EH4iqloNvcFh2y%>Ii
zUMQl7O1G~EJSDBWx^osr`S}X#HkPB0nW}u~S6-!kR3Z#526wPT>Pq?!{qIq%ZxOaS
zuhpsIDNFlS4~;g=R5z=tWnFuML_XP9ZBbahJy-)k{vc|^kq`tH-FBoB_k*5>d04=(
zf#%(+?Yy{Buq}!BbJqB0aHF(<P@{v>(K>F)Xxkx)=gRJ7TxOM96u;k&FH`UUz=*s&
zw@1K)<~M|wj)uA4{VQqPzVxd>uuN|pAP4}$l1eeQQsU2d=!H(>t5(ycy964gdLwOP
z#Q~7yzw>&tl>XW11CLmSzJBn~ExaL#PvvANQdTM4)=vCiP#vUx8b7|=un^rw_enEx
ziGuQN;!*>20W4%m%yMymcMpt6%|vyJubb-5*Nd8L@I8iPmH~XBL#M=trvvT!hAGf>
z^^0Vid-DSrG(SQ7F!sBA>#$Mf<z|?|A$QX$#j};>Lv&x5(Z7ij38>Zb=#gKSVQE`&
zN%B}KAOShgq19k@uG_W3X<l+W<g9%jl%94>@g{v^ltKAci2&O=qx&=oWH=EF^ePDG
zFhEBE;$^_pdg(%&<l7}c&Q(e-p)j&J?{^dvd1?oYzq2NsS6gR$nDXO0=M-S?6#glA
z@ZPb4JO-iO>n(PRM;o_Fh{k}#ka8*qqdt4iCJrw=#`!QQ@I}lQm92P8RT@Z&RT{cr
zP!p&%Qj8I)HR^(jNc{!aXnB6aeVK3kOmM`$gKaemm#Z9ks1n9REN@3%c?gH4n03X8
z%p3g$cU)>{{gp~zzb_AMIZgQGfsTzQDQWGD6pKatjhR0>7uAQ8IXhGpK(ko_%f~WP
z7N7AMRRbvTS|#3@eJfo2@B3D9Q(sa(92vBw%H62~=yRt^ovGYt*I4ftB%jJ~3}=9O
zAaj@~0GH?cm<_Jw9aDEj;&~liu9}jU{XDJb>osQqpw)1ZdqC7y0DJiAsbN|P!wl7h
z`G?gRE@N|Qu@`^Ziqan1n4v8fW@_=H5Dhn46sP0wqXhzit`HRh^yZ<9?nx(_2O&ri
z8R8hp{i9>mvhnNsBx9u%?eE14CN)=;?U`gTh%chjH_2T`3<I@qB;-(3P*oXF>>%n1
zXbnZZd>G>LB@&itxp*G(bR0-hkU_2Z3>DDd+%q^j`2EpV&@&8D$fV8nOZC+a`%gKy
zDE9$Ut|6v!o^fKqoSv=HW-U`7UfSgid!{xD;)|a2e!kwvfwXLgzxkid1Rr|xk5~N2
z3_8~3nFBU4{UDe_7h7rl_`{C_<=75|gcIq*qptbYS0%`!uA|kDe~WEI+?uukHq~bT
zvCs9G`f9!Hr<~{s6#VCoQqJwf91FJV{(C|taI6JmHwNb3i-74DU;0s|lg^EaWYv~m
zC?bapB`q3z$WU92wE4wRMSwgcVpM7x1Y)Rb#DxI@)obpk9M*%HRd3PnnK=m4SGll(
z1N&675u0vq!J+LvxWD!TkKv2XS3JQeHU0cCtTj*8rOFJ^l!lnPxao$PYJ%nqu;$ZQ
zH)<XUMWfW@`F(EziJR*Z`@f&&si6>v;j7``R~vaTaKYU=WYhgph+(5fZ1_TUv<O5y
z+8b$HFSE8dU9!|sR#5VXe`9)OGzO!_hrg6PFX7zSoNy6eWs_SqdsiTBW3Coe2@p`W
zZpJuJqap;TZ)-&Z?<jk&VK<kP&`V6q`Mi5nhTjlgIF<9LJ&D*{D4dnO={YsC=IpZ}
z8Q~fDyOJlt#CMi2@kD?^f!6hK%6W4}!p%FR*NAN_dCl2}nF$@lc@iB^VkvRP0rqF>
zgV#~1PhyB`#m7$ueZV0y>^mfy`|!txKN*KH{ep>!f<nmHc@tc}baK23BV#Y}2*F^F
z^io*XN>~cb>cljb23jH!XWU@7I|(hzPM9Y42-LCplEnHdnvfmml=zg8eTPnSRys5H
z2d0E9K(>0A_scAW+$0uY4$0h9P@D+ptds>b@g90*S}ffWM|_IH4(LJ6-o&3!6{KLG
zg+4TW2@+9?A^tKg&3M4ha*Haasj@jhBKzwk85NX$pZk_;rkP$kr@QMPOg}~{eYlp`
z_=~>qCzt@4DVVN}7<XB|sF9AR*c=Ec@lE<{z>IUBX6k{oQIsJV+-1h+T?{2jz?zUp
z=?AGhF6tq9xp)wAARDjy4sX=G1pJ0Cwb2KWXR5{YQRwKVyg>U2aRq>Juq|KQnElky
z(2RGyyY5Qk6`ZUA{oY_Ovl^d5zIP}CZ@v{0e~gtbv~e9_vfvG5pFTkEAQ;m(1Trmr
zfDmU23XDt7Ue*(hcMy%C-FK>7O%Gr#D}gg<dMpQAEZLJkRk;S;+8SF=wMuYNIv7wl
z88ukESli9GYMc~)8K_)iNmyc{W}rl7R)`~Q88g;#y*x;v@C<mV1WVc?6&xmcs>%h>
z<0ed#-kN+#I8Wh*@Nu!c30o~+O$Tz!-WwcKfIM0inhd=(Ye*8JNe{ny5O1<!r!|U3
z)KW-Tj^AMW8jx&Rj0@@S+lQHbnMYss&qq!OBegS<4<}6;3d;b*zX^L;K9%vi#$wxo
zd^+KXbfV=`VQ=ZDw?W&pjh|GRujFLqf&to->wDVOa*j84uX<Xb2%B7A#w#v6;NbU)
zV~;Czj9v9J^$Sq%pJ>0$z>9eOB}|G1uJ|&RCE92-h#UcGh&;!>$jpJF2qy9#q2=xA
zRdX~Dk&Fe_5AA{6O>Xj74}@mUq^4qkY(N$Qya{>qe$;tWs~-#)t07Ve_*h6Y?xrP;
zf#*R}j*J{AKzf8v;Cp?f+n6~ZEkZBYNgL!OPwRkkk~?H3o2@<C)bbVqHNXR6kH=gD
z0l&dMV*ZglIT(MzlB^giDObM%BogP+I5M9G2l68|2re|t&+awJb|Z>F1z6SA{3yp9
zwSJJr&4RhE-ZGX-^Emoo_qWmsH<W_XUA1yTzaZbPpNo6B=^79X4BHL}0hw*NgE(|U
z@D2ckGL$wQRwiZhIb*J`rZXN0u>>lmbtRlr63Hqqzfjp!Z%DDjcf^$XTg670aP!^{
z!XPxSE}}U2eBTKj?YXwq{{sCNI1)LGr9zYjI_It}rULCyN~!dRXy<~u=qf9*Z%5!i
zaj=GR+X0ZUN4a3FI{3V#EKiU6^}1+#9HlvDn2eAQKNRCHunQ^}cv=#q9P_aU$7INA
z<X_Hv_It3ojyfT}VvxIZw2k&3Rn{5DQnsFoDbZ{iN&k4naDb74AKgvQ?}bk>f?<b{
zyU40d`ufcTg$HjRE{D~y;4|Z_-vO6mpvTC&f_Tv<?13f%<eKnR&X6yp8K4-x4h+T8
z52}+e7WMd8ZXw<D4QHG^w~Dz?XPOKM-peHzvLpnQ<dOR!rvDasO9qU;xoD(s*bj!_
zV7NL)v)A%+faON;`!9mA6V5?IL2`lkSY@lEFSg~*g5Vv<e&L|kSQtC_c$Ifs!kcjX
zEg6P3j5-5B(~tH8x5z(a9c=U*h?$3xqh!Z67$z}d&p}e+4cN()fTU%7mFwUFgi~Eo
zAw4@1KJYkyK|WspG4gvEfI0*BMJ7mkNNe*D^z^VYrTLYb`BuE-C5`$sE=r{`yR_Z3
z<-@x*(7*0BD?vp=R29@;28Y3hB~Tzd`aZ=mv)xkyg0Jk}i$EmpNy${pR{XyCGk5hX
z>(`7|p``gjmRF&Us{YrJUstVsTOtc_vJo5{IG>I|k%HoOIEr9-iUc^ml4@*M4~*s0
zfW1d>RM7zz-nf~aPdf<{vR#!R`DYc^Z2H?b(k0uON9DR3H;+=@T?V%0F#l9S9MAZa
z)4c5zD1xkjQ8+(h;T8U*e?#HsMI;Eu((ED>#~61Yb7I68*3NPY@^=>2#yH81_WJJA
z@)9JhK%6NO1*>g*$<+<Me!U6N+zdH)+(k)GOmtDL%elfnpd)7tOm4v$k0JjYJT%jv
zE%s<ENOa``1Kz1i1Gbp%?uO_1AZ-I0e*5FMJIp>fTJVR>A%4ZLo%COlLFuzQup98*
z;5;ge_7;u5n-1fbhb;?#=G-VWf5o!tgquV<rM>jfFuEm_m%)<v)!l>g^^whk!wheq
zku<G_!<s$3zjJ%D0`<&hA5y+;b0IjIyeOASD(3Yq@5O!1`^nc5uY})4$gaBsGXPaP
zx2KHTNqWh#dv!=R;Nvr}wGa$UTQ`<=c^Iay(F^qDTQ;h-dCWwu(bKBcS>A5>SUzop
zaC=u7?b%e#-y98^yQF^8<VJ8jjqhX7nrL{*ZP{p3c2lk~juc_Wepe;7TQ~p2{zCb&
z$B%oS9-GFP9uM-EHC_~$?6yiB!iB@c`mJ-HM_Sz47XdAe#W25Y!A^jRC}WM66DAvt
zCI#*=yd96zGi{>MP+iowYsBvNn0dR#OARykNd0C5G_Hh6(4D_JUQTUI{4$(}fUa4k
zPHmAcuJtG*C%OJhhzLPz^XgGnaw&qn%8TnIbeEv!8O(}(2SrxdZ2=UR^{Wa0v<q4M
zK?IziCfGw9t1lPjts?sO!ufSszawI8Om=Ii^ENxdjn%WFjrW#C-#Z@};eVg}VVvJp
z{j|+4C2)ZACQ1L2wbV<*V9Tj&qTVRqt?KEhMapdl=M7Z<5~K7O-(ZWrY$Dj(>o;Xl
zh?6+>MlQO|&|~6SxmY&w6pt(4<*q3OylqDOLdNy&dIjp*E`0};#>2iU7Yt!VcCW4{
z4LF9mgs$y84Y=NL-gziR#39a?{P}@H2ML+)8a@x=;Q~^@g<x8XPA(tZmVG7JZgivH
zTSv)~^^#P$n3SRfS!9?FqLozN&vhOwS>(+h#Z+P0Vo*r7&m#BYX}dvuK~vQ5vG%6q
z-;s)f?cK<1c>L~=C=d$`d=UwOelFJD7JSm)g!~=1(iQRwR7Rcq18ErDL6o64xa*<!
z7VJPNlSUkEf-1Ki2@RctUoIFK3rE`u@(c8eVHKGXf8Qz>*-lt40ofSO4DvJi+X3!@
zQXVBk7ZN3UKfhc!@)rJsz+CJ+S-Cu*s1gL>1I@^pfTSoAw(ltzV8JmDz0X{lsm0J6
z1lr?}fhdYUn@ITRKK$*_V6}Z3i+{?<bU2|Hf_zUxq#vH0J2qpWO<BvBe3FYoVZ}<$
zvt-h6q>U_moTuZn&%M@q!&OCz_LPL<SuV>_(vX@U-$3)~Tp`QPiS=Z^vou>Fx@3RK
zlzwd{#V8zB!Q@E2Lwmg>gD7X9{2yeVWm2H(CF^+V+k*MNK<;0Wca;CBjnwn_nU<$-
zV}38-AM{q*JyUJ|W%5;%Ta?&8;{T~-O7W*p{=ZoS0oCu@shL$X)Y+q!rT(K7PDJ_C
zjwA8o<M1TDFrvh#p<Du=x+960pT=^WP<<!9_|kpO6Y)tpBv~QBJ2eog)E_$(g=>)-
zF}fHv^Sujw_Q3}$q2!LA$yckEZ%@9q8_eSUwi4Sx_I@Iaxtfki)koLoznt|@?EGDo
z3yHS`QUv_LPRLz1U<64NL4AvYZ6BVsv=M4#g-n9cOLD136wp>bBJG0dsvn-+w9(@V
z%P|Wp@&H4MS=jgFD+lkgd9(#E;TO62?Je~A2+7LB1)0OFsveEV4B43_<Yow6!v_!9
z3+u8o0=f?31`Gk?Rot@fbG|;C7Q=lirqzu-8OM&$dA9F9yDoz_^ICn*Y#Hihhd;mR
zc`v7}h>UASzE-Fu{*;xl7Rt?@$6rNs7<Bp5=*M+3=7nZ@o}heiR8}70+2i()A0j_k
zMf?4wz3ARczKYv->eQ~mbk|B>6J$J!@_@raq0)@an62vjRv{R~Qt9xH&^-3Xl2u8!
zooC9QLvjeNf%^u5L^J_S{udz<#2R4Hwbsj5Rg{Tr4g&Hqm{{tWAXoL;5Iy#8>G4bt
z#8scI(4{)qR>AsH6l^SO+kmH6=T$U7>yFzk4bUs8o2kAkZKL`vU=v~ptiyb;?9zSE
zNzxlWY=6CWC>u({bXDiKVOlO2MHOS~=U?*5_Q~8nwT$Iy5RlJ^s^9OP-J0H)FDX4x
zv<+*bBJDHCZ)&UO&He(7t7MxN;puVyZ4J#^88?|l`Jl59W>!{BIcViv^6gmi;HT&;
z-ZyLfov;8f5?&{xI8I?M(d&XHm}Qi~+xEZH;%OIeQN%yLaHm93F!z)KGew2}j$O$&
zPo7^K7_Sn7*&M>j9KsYz{7#9h4&N9(%eQ$*wHuD-p!O<gudSPN7fkp_cfs-}WraC}
zw*kd)iOMX38cgik;nDy~oKtK_-fnGV%V@&JhYL!D(fOc{F7Y&nD;Mw0mDtvh?-S8c
ze=*<R%u-IpOSE<%Evw2En-17kw(W4*a@+s;L9hOJ;ed)Vyjiq6?W0CJl+hJu7gAU3
zdX9)msYhWBrx18t!?=X&ZW~FhI8U+Mo8S=fK=5$_V+anRoyw)+1pOl{`(oGCpvOnt
zl$;S+4CjnYxLK&dc!|!Dtcu)}HUk7=;cxSsMiJ8vVZw^S)Bxg6Xm=%}AWot+AR(6G
zQ@gfC$(4e*Fo`0ZQeX<5gyK_=Hv5CXNKQqmyB3clsB{<jGGozsgXR7<&W1N`6G^5R
zLf@3Vy?(nm8*<mwr%^yMT}O-FnX|Vo={n8=MhH&1zx4I@pjfnKDw}Sg)&kiw+;<5V
z>=4PI7E9rqx*fO(&yPNTZgglacq}!I7@O(VrQ<HwIB~vxCPjKkq?Wo2V@~z3jlxkw
zHU62r?VTB&(pc3U$BM_}-i7;~GqqtUT%A<SzW8=brA$4AwS@bbeH5o!h~=lAiH~|Y
z5_G>4dug)(va0UPeB0vIbtCSSW1{@j2i1=gSVOo8?V=d=KA2YNz280xV*kp}Pn17z
zT+<oA%r);@*(u%1nd(v8dJ|YizKGLmH(y`d=@QC0e^$^LN6DF*p?h;u5r1N|EJH);
z+N$@_!>32*MM;pCMC#!a<a4b2V6If&f=e;!DhTV~wLZ7G>WSsm@Bx+n6kEBrX^qWR
zIP+ezo^_GUmO%-kx78?~Z(CkJeZ`{6i*Gsz7H1?!kn)goiob|Ipt_TPAv(XW_*U|D
zH|MdyVoB#vCYI+z&RFRRe*OAR{xmE_I?iKm`ZMJCpnwqlN-BR-?iS6Hob2d)XDixJ
z)|TSi_*+}<S)4g*+Y-c%Z0yY4P>KElFdS4)Y!$1XQtkw|RQ|_uNbY(jj-FD<HI3-T
zf2>OiMt!YFnyG#wYfi5w1V(qteKZ}k)YcnPqmBPV578s*Tten?7#Tl}6gP%CJcdi2
z3py+gT>sPkm1hinR~r%o#^mjKWt6y{cG^v52<N=B&JwpjdyO1av|$1^fK=F3@vRtM
z_XNg1FlyTRpp06x?yc0=WX7fV>D;&>+@*ZcDn63`YB%-49Mx$ZUVXPxvz%>?Etq4D
zU!AE@d>eE-XJik0iFz^e$f_nJhxB@fliBc)PSw}ovgVtd(mkpS{sK*w>)g0w6f4kR
zzeXo3<+hCCMSbD*rCW>`dlxt5^L$x98cigz=USb31lQcm&r9XkYc?P_HO2`Sz-bL~
z6E&S5q*D&R0kdNqYt4bRCN$y206*8!k0XYTAK}TeUMyc7(CCJ$?Y<K0J>>sf(vev#
zv{IGKP=orOwI2{iDd2~TDytUr!*nd`u<(!a(c_F3WU&FNh?@}RKq}1(cefg4KiPBm
z<CKTT(_pd@d0e}XWFCOEo)Cc|++NS?d@EDwn)<4y?1TLR_wVOFB<4V)kkA@N#){|0
zVz4IT@57fP?_07CPx@B662PgUHQqa(`7Mhm%I)!s_H*r{3HEE+_WMC$1L37sB4xpK
zm3aS7pD%;UD{;Se;c-B3DlueKE}Wk$=A6e!T9fZ+Ubs=l&LmyLrM$My32-IYG{?qk
zmlsEkSbA{(id3ET8NU5;iNQW4EB^7e1M>u5vnU9xa04ZNhJb%dEB8m?Tw<|<lVpYq
zY1dUB7P=fr^P?_e2T=zym=uTCfbpCdu<N)={d;|7AIPF2kOC>fmz3cfpHk2rLhr_7
zas2YTHwfQP_x2#Z`y))AcE^)I>rv#Dbw%NVue>MT1|`PnL#cyWtgmu3%Xpiiyqn?J
zU_#yw`-bZ3QCYXXwg*D{6$W4rE&sQY1dR*Ot9>L_1(`KV#LetRDHzu#x>WV+$~5=^
z-p{MFeG~e{XAYsy9YV-*ctC}@xheO<I9hXPADC-R_H<@>n7)&_0k;A!a@<MngUwYq
zjroLLlbapv#4Rb3nuQE=H;_`jsIZLI&*X)<H*MncpA|THxC|m-IkEYgY4(M%>Ph<_
zTh%*W#;=>WCCyW_cww4x_JzHNuHa?vKVboIJClNI!L_u9Z9Abx7CDDGW~6Q`u)lo9
zf4cHUA_t`LCQbW}OA0U9*t|PjXR)txe{#;KtblpxF<uN-3)%m8U%jJl{JMZ!Qa3g0
zAxs}`X7rkuTar9A3mY~Pwq^p~64Xk=Y|g}J<eBulP#FxmeD06B(OK<j{iCsR=U?KJ
zKDQ#!6M}*h`;#&|iw8O8!&!~)S&DA{zrM++H%qjnZMCTXvG?UWa%<#SlRZ*9XnkqP
zb}2ZXR)M6ImgNQ0vnV(e3`mV?^_O0gy*jDliAj@arTP4Y$ugWr*X;sRDonh9#Jpfr
z@JZUkCiPBxBHjuWql?|~&4pyde5+6QfW0Pf#Je;9O{}azE=gm_6C95*+)}GooHXtq
zV1sU!xXxBQWzdsKG@U_2sczM;zh+=r$ZjR}*Me)l8^^L{(4ZCj%cE3ty4|pVO8S4_
zTAGNI3WOzf^y%Od2MrW!))wNS^XC#(SWSdV9FtNUt#AoPPhMoIQ$jrP5OIRBEIp!-
zZ*6ZVH7HL!m3JM4cv_}&*CBz|lRGoQzyW%<W{!+s#3hF_nL*c+5tczD!<Hd10zWXR
zNyF98&mX|pu3fYZ*k`#I*&VhG*;v2ZB&b^O@!^c~W4|aJ!qxPmPO`hOeK5H?^?7?k
z*o(DKxehp9Ghh3G+?HsiG{SNOOTOsrmJRkKV-+F-q~hpD>fdy8>DO|SzprUvBi(ML
zDBXYhtbFi^#luGZMherHs$r+mM)uB}^A{aHKxX=GF?e_7Ao+QWeer&jWqhHZl_979
zp^el6uCT5}smB3)(DGx;Ey*tQ<{#UMJ*`|l=gB2UiGHIB{d5(l{#=Qk3qUwOI7n%l
zd1ob?A+o$~yeCc;2NY@zLEO*$OLgi11w<cq={NQl^mGmmenDu8Di8}RV*lrVv>D>s
zfJDCZteNjg%I^}hyy{05DKR{9-=9>M`~n&>c+$j4x>sVxQfgwB;`Ca-<cEHvk7NFL
zpGhE!)<@u9=jnsYw?x_;{wH#w0U6bTrSd1hk803as$*O+_$2#INWZRq3+J?0soDGo
zECKg?o#5K1J!71tiluru<vIfQH~PA;&6;8}uVrq7g3dAT-sdAlPhLzdtA73V?Z>fG
z0}5Y&)@l1l%EGG~??cHYw@&t@Z&pY$XhGQMOW!(97FB10*v%0E`||!48@byyDNhsi
zIiDC5lNjgg*rhnd>+?Q>k*#ohn(#a-Pi89Bi!Rf7Sp6=?6uZ@yydu9dUV>UHkT;?E
z^@7ZdqL9>4yym>lbb#7Tl5bNRQHjn{{aN4h-&WEuA#s(MqQJ?RSaVjAJ7MOPhZrl^
zZ3N{zkgQ7d(~dxKB#b2jns8=qJmn<{y>@qT<P`8<PtSi=<JH>OcN1|!#6mIu6>inH
zyM@jyhzCq(@)7svN6@2>*iB0)W}(j7Dq)s8QK@_#qz&87ym&^?B}l1s6J|YJbUyeS
z7=Nz|^zUkXpU9xKZzD+fS2^%oU^gJQ?4$qcj25PpU9>+ovz!0S_gRhE^CfOs*sDsq
zuZniJoTPy_hBh_0m4u@54hN4!%tT4PH1QQc+N!NEk~=hK(eBK3v=^TPBV&KX{?DI9
z^j&!`W~eHK@4P1`PFfsg4NUU2!&n?dT5Nst&gjqI$z<j4#E}4fR1_gal8a3B5i{c>
z+K3`oZO4<Mr?_)52kX0In7{{e$aAT_bdfe-oxw%Tp5jInd=7R7gIqg#8gTHJG@bgd
zpO<;l)3uxX{O2WegFkcXzlobYWo-_(@Zpxim_tmymk+@qvP)oqqEfQ-GsvT33O4+B
z?xw5w8^4Eao?^Avs^NfFB=&7D3MFo}V*(kpmLji*fd_qnzkyX`{%XPKJuB2Nleg;7
zfZ;E<N`gEc&Y@KT2DEC0lwnoA1Ak8qceMA@Gts~NWz?%Mc=6Q!o@@^%DrNEo{AnzG
z?>KGpcLsg*VnLM@`4Su*`vgv?9wo_?m7>#mlWgqpxnO^qd@Igjk4Cf*)Sq${#vl)z
zFf9b+z^39<+a<6<Us5%r3ex8d0x}#0RB`~?U&tN!9u5gOP&<mecKGnrek<eaNMR`8
zeyOxe(1A|S15c*qL~%ZC^j;X6M+rQWBl=<pAO4D^U~ijUY|sD&B1zzuRTqodq9FXw
z)s|5hXTJ8DzSB8$<sbu9jasKmOTSyUgu?dSL&AanlvqYVN<}-t@P0d-!~qJXcLI;u
zY$Mfs3EsyF2vFME2<D%E=G6wS=pA-ACIiWM`QbSiFAkKr1uR%?t<-zl(_@1LC_%P@
z=)HD0u>%wo?*#bSY+2NMRnlV<`6-#L1$$@zRq8dH?k5N`=N&S<)$XZC@e+9;=dYQx
zh<gcCTq~Xg1L^6p^8A!gtGo7|D)83jg<EFFx;R>^tM>v`e^u)3wu>zrpul@4K*MG$
zuHFkyk6q%Y>;j~A+2KU|^Zd5gYQ05i>RV1u=fj6S%qDl_z3WrQbdew!lnDQ?t`B<_
zEM^q)5W*tYniS6~@7mCnUQ&uzHq8|gu@Exc*BgBV9^J?WI?UIu?WxLR7K0zCAFQdn
zY3;o`d1rC2td*Uvt(;$4d;6WrQ__TVfl==!un17G`wf2I^FBC3H<~57W{0CVNFR)+
znb(_@6E|G)aiT5J9W-y!FHieeqvt;zh1N-DB{A5c0&N637jx_-FJ=abAwUGG78Ejz
zJRj+Op4O4?sCDh*{&#JAw&d$nHqb-XyL~#(Y)vylz5jr&i`WNibV=Un<+;&IG69*;
z2e)??QDnp7LC@eMCQ~(eag{T_h3VQBZ0Z~;ORTl>c5xyn^|rR5ut!=WWviUw+|fi`
zd0Mtu6Xxyf;^dzkW#Av`sHCpzbcw1Ll2DlcXc+29r+#FWfiWqSPIh*DJ9U07_W2>5
zx?wwV!fowwVZN+6=>62WXx{tOTd!+3xfsV*b+sUq%gF=Vq=dqn;sa;5ZZhE0qg}$3
z1s8&}^%f2wcdh!oUN;=df<^(2aK{cU>_C&+Nq39ag}rBs6u(u2G8iTWb5bvkK}Jzm
zZaXph9Q-`1ju(K2CasewS%K(8nW@{yT~p`9lU47J3oR>I-tIBWLVA=Fj*le~!u{Tt
zubi&GwT^4i#|-=GPX74i(?loQPG%IfLr-ecp=5+HXN|EFUS53QiI6sU2z3D#TTsbc
znz?sWh>KyPHKd)e&tM$Q_Qa|4uwEn9z)8*soJ`<@^3UE=Ax>aMYr>hKfX{+%B)^w$
zAxhsv8|2RQn~Wn5IC3;HUt<hu53+0Wr<SXEvTvlYcM&iOyyUpdLK^aq!=}z77U|Xa
zB)bm|_(HZ-d5<)yD?Bb|>!!}xO>J6C2@u^25#Nd(Z`O<m&W_)p#xG$?MPL*=UeWh)
zYArD~hgkR|X5i7p=2tM@R-`x~sN#R+6=2CWWbWa!RJ3SlAT~z(bjMHKLh6c(!O|A*
zyxR1x>O=Id#0m>eU8?j_Z=^OC(PZ2JEKFyz?$pR0e<xItVJ~p}lwL&x|9c;&a{EHV
zRXlY5GCsc`O}FYel$o;pL91o$&2)mQM$m1%O^Q%EG_d$otq$?h>GHhLW=o<76ny$Z
zck+q=$gdr~soi8(X$#7|dNUS8_jr!}Og1gBt>NlMcYz+E2WxH<5DC~w`}END*b>S8
zY_c$j^=DZlR^e6t7y`v;A_&*+msU)m>e*sAta!^0v7@M)kH8xHgg{nl$|1ao-F)#_
z7VYxJbsT|X_bZ591lKLz;Cv#8{1*y78MhI^b%?oXq=VteULw9thm)2e_r*KQn8J}Q
zG$rnr=&F4CRvqt~D`iA9ZX%50F|hTFbK65|$$Tm+g+y*Te;D;hw6rIo(Q$XjwW7Ac
z)h!5qh0t@78PuNVlUJdskx^6;U_AzL-!To5Ur$uYg`-&_VJBH6+=fgB%9mir8UdvS
z^6nOGBtdTRGxEGTh8wgkK9wN(OY8|tpzfGFuMI%C^F{i}&tpHM)~Lm1*PX1m2JcAA
zlkqC{_B{62E9Mnn0Ytz{t>mS!NsJPXL-NQ&o54IHH|vF;A9nEV+XJjg7CxC7fqbpr
zfIvWt44!jNBX~#1A844^zCna?_l=ZDz{z6IbH=WiDSt;luZXg+BEMQXuZx99YCTtI
zQDoI45w~U{E;ooslOdg7E$wSL%AHX2N@_fX;~Ced!;-DcRI{KDH&2P;k2e;?ag=la
z+0G)!dUvIJwDfy?5K|lz1J0cZrRrDU*_>@2M`7z&a0nB1h`@6QEtxXAm>c(?ol?|0
zo!;_U;q+1<_fkB*wG34QQ{o6Uj3;L}e|@5B2&dZo&_I~A>%BxbA93eu2)<Akv*>X4
zL9f^G>}rcTrIRFI2LD>ue>{%u#*h^FA~>awn1V#Gm1-FeIf*Xq*k|=hO2pe=(BLiH
zF!ubL@=CE>AKD|{G56rN*j{FKD}EiWu`w@z!ko_00)4#88HlHMO_aCjprIi3yH6%T
zp`6qT6SZtc2K{L7S9IN2>_9WD6_MhLwVB?K&79P2pJP<zKW+299~gWCL#22>A6!md
z*ImqVv&lsJnN6&`0q^vMiSbI^(7)&fuI28k@i8x@9$~u4;m|drF;V|{I=s05!$C%e
zp{rrPH;Z6p{M8rYfMyI9@mm^;b~WfjqOy?bN^~Yl%(t>C!_P`ChQ#rYuQMsh56=~U
zZ<X*&ikXrf!G@^0VxTv?B0*`Foci`3;Lxe-3d9?!1<XetFt63rUF4QVdtXo3kOY?}
z;(ax3@kK<T4Yuys;2>N)6GFl4K_o~3ve1M&I@<T?9K8sV`6{hA?of*Y*ra!6Xul#d
zRW4yhE~N=vs#YPnkbkCp$gEz_b764k71uYErc&6Kjwj-289|5mHkL!7TtdcQb{@?`
z{#{0`d#1i0HGVGh!kA`9c^@k|mD3L|P3LI!&*LNCeHgDIUvqH%z9ynoUr)ST=cIO9
zQ8kESpe)l~R+3hY`8=)o%~t$!vAbBwdxPTVCTp*)^`B-KG@e+cG!^K(V(I1II!;Fr
zc||mr;ENTmH;z6Aq3E;MZI5!sVaaVxa<XcaCJUGAOw>MA{jZyCMfm`u><olwkgz^l
zt#)JWAgxttvPAi%ww=>EgJNrwd<VyrPx<=v(dqzOoocB}nZ|J20aHd&E>n@dtEOqb
zsAI~{JpI8mg8&Z7lP9Hm56gA(YoDe}aOyw3zexq2gZQCrb9zS5Z?5s8+^I+9m;42K
zPnBpVIKz}nNtl7#l3-P(lX1>z<x)NR^1~~&={!3&R-uRQ2ZrHe8ogdag~b<VvXLit
zdGM`9Tt6MyR;mxra|~=J9qo4^86;CIf7rzeQm0$0vgg_CWgIy-@`^LxF?|rua}spD
zWGPaW*TDLhG%uG8P2}MnYS31P4j__BFcGDX@+SVVYFwCe<UKK4u9~P3{Zhvcs3ZB2
z*Utv5469XN4BELk{V*rfh`Ou6JII@7TN9;ZR$LW*Sg+csG3cV>`p)A+@?0MAi$et`
z2?9}#X!nT$3CB%XUUB8$nmdkr;@*~Y;2_W=@j9B)$O1cwhp++fRHYF2v_kXZwcpo!
z=DT~~={eDMrCGp)!hRX%Qh7Jvbk7!v6zA`?J>inH7VB<Lr{K)bg?wn&caucjQ-I&P
zP46KLcK=>lc7=Decr;++dT5yt(BtR@k?3C9UnL7m<v;)Hg89YQl$K(wmZw)xr2YNW
z#6^!oEgcTbQ?1^q-mCQW|GND8o^IkQ+C!E#hK8=v*kxL0+NAFmV~Sx0{hfQh_VvR)
z5^q83i}=sMreNy5cGj`dN*<YU-Ob<^@Nkwn3y8bP-a3$t?K<tcXs}TK^K(S;+Cq$j
zMDOy0TE6fL>8Ct7@d10E<#Ob%P=B(I$poKb)@tPjpzIPm{ENao-3@6l??yHnIICCy
z;>+g9ZwLg^)oTN)ci!1s#wJSQVd%oxWg>Z2=h)9>x&|+CxPHI$TBM95OcD@%S`whM
zE5uSh+ObD$n?rGR2xhADzjP7iJeTP>>zP$DJ{_pmO1&4Avws;nGlTnS_uB)V&3EcK
zc9$%Z?7giDP8Y2e-F_-f@iGAd^ILTD7#F9~sBLH`Pw*4Duk3!X!({G+@OgYoGzmP`
z5a#^ZN~(7F+s1UM@$-w5Ow<j7;Dn9ua(c@NYnj{|b`tC`3FDCg4_*(ycP6)6XG&b3
zvWRyx9X6PL))%jxAuPC4Iy3b)9*eTCN(7Hc7;a>idTMS+e5TnzmECC0<Wl3Y<ah!S
z%$eKDNtiWf-^fH-N0nv&*%&=_*fZKaxSTENx^`}ql>P2FGjD@Mn!Q&-U_>HlBeTtO
zWKp7UD|2Y$v84B-x}jf@O4tm^5|=}&zM2!B|3kkeaWV45{^$3pL;64olf>K1cRnK*
zlI-Ua>q7@tn)9B^*~Dj=4IACk?9~$PqfveUDls=<!?hq4am}4k*iM7CzI>}7nB%vW
z8HHp}Ci$Ch`9~`1|HEwyH_I|rMTj8`YwK}eLxkS!xjYIw(%}|I0$f=lfY)7dXN+qW
z<Y!G!6Q0sWpf)`|7zjiLSc%bu*=i&AEum(HGjCmISVp>VEwLmkie!fQf`2!s3T*+W
z>d|xMK)Lng>9EF-HW<@J=y#%0sPC8A+qtxla*3OAslUiMKMGp55(XplDp22ThT*$*
zO?%z$3>#grncaC>={o<0gdDdE`QY)vdFE9&KFCt<wuD;vfV^+=kKXk89}4iq*P-b5
z76~e)2$aM=vybJp)ua48NbwIU_${W_%R=B(!lN{W76H6CcZXPE+qqx~Tl&1C#ggm!
zMvFFs<Gn*F%HBL#r91(w*k@L;47SR0CAM}H%v}_ez5V=3;R0LnRj%qKw(f=9H2YRB
z4Id#^{dY*^dwm}DP69G}+LR8ILaB*&DI9y3`NwTy$!w)4Ryxuf`nsjr_5_OGW@&6#
ziuU+b+QqWkKBijvna<tk!`}-PDC|wwwFMM5lwY@<r&#Hw!0*izw?(D=2J#NZu-)Cd
z?%V78258J`OZzEqzd@ikPQW$R-8PmQP|lO-Z40_dEYO;;?H^%=_qdo$=Y0T&=}MhG
zfh>O*nU)kMG;8ZxOD0NRw3BULaw42#y^i#2$LmP0a%d&__rfEPebp}BEG{+t&K*Pm
zjfe+AfRGlisWs(TGed#J!+&jeqU@+`AWnN{q1;oQfes<h9fFU>uw2IQ+`A<|V%v1z
zqBT57gQE-Q9m$@PFH7^@xl9Diz7Ybh%SfT)zt+-SI0m}d<OTgrfx|K3e8|ig_FZUC
zop-tN=;XyXwflO{bH0fS+1omw?Z`8%3PYck1LoX)->@03W{ahh!6L);md@YiX(L*<
zHo^o!RWne3KeXv@9#%nLL{p=`8OFA(Y1a9y5}XYdiQPhCssKdDwsh(U;zLALxt?GC
zPp)Z;;8U#HG?W&bofOlGd0$Vf$Tx1%tMdIV0_QOIy%Z+^XQ(;LM;muW;8Ef4w>HPM
zXmN3>lW$nkqO%KB>-qJ}a`jpWoJLojJ&OEx#@7G&-ztQmU)anQU?$mas&53{MA%rO
zE^?h)bev(e8n(8WD>==@y^$aGa|yJcGaDY%{JuRsI$gG!%9B)=#@V8`W&hZXW1l4p
z(c<>@j6sD@-WPM?+o8)dp=u@-jY<7|>-DUiqtXsaS*@HmbVEm0*$hYYdEckmYsK2|
z@?^F0E;NekZne2pUJom$zZjA7TrU=E!_M3}@`ErOnd<yr4<h@J=p0C2X_u{Yq@MX=
zq`4yGP5AH?XZl8mRK$9tfL6+dVsTyBqm1QgsfzU|;Xl~+*)m7K*9Qam`BvXB4HG+N
zs{q?g>=ROA>svxWoOOHO=;D%`v(1iHGhQ4tcgc)?e1kXa>5#2>WcJ~VR^X4MO}54n
zA@hgz3&GJeK!8}m_s<b4(hj0pHE%M8Ctm}4Bxe*(N#*|~XdQZYC#*GRqay#i?(q08
z#TOPdkFrN}-LhGaxZl4R@UJ*FRD5|LT$k+}qpbJ7a8N37-B0IteWSgj#w67P&+i?+
zvq<*=Pg#|;I!-QabaSqEn^FcHfb+x8if;#SPouuh`yP`Mi~iS@{}Jb{JQ_{J-$;%^
zxNv4LcV^h(l&`I5Mg<U*gjM&~Ie6!Q4)H8L>97;lh5%lBs=qG#f85bTX5pq_+6>*_
zur@Lt%V4s9B9i`R=wfXSnIU*D3#BeTu`T@kW)VX}8cck*54jQ3$;piQShO>ZP&NMX
z<*4p&v6d+P1_2_IW|eZw>lU9l9Oe*u!s|%5vV%Z1h((%dix4aJrAm3uBf1Z&uR;}>
zQ7IW`B9E2xWSMRVx8e!x@@N2QP{Q}|JF4Z1meB<c0tzvNCb4>txNPjd#PQpN5;9NF
zC*D5bQ5p$uQ0CxIQY~LZym!zlVLqm0?0-<9geuE)PS_FO@Bi3@`Bm|&^Fmvpl31`q
zC=Rn71!LC(N~L~&g>XW_*aT|;bx6^OzM=bD_9A|qgi`5TOd*9(DfYgRSPJXz2h2(o
z0-ce6QFr`r@Md&>Z?v(QylumlqG0Ur$A!c0i4HZreK{(L4wFqqU(};3B%~F)z|OiS
z0XOiS{e|MJ^P8_FR$nx@%?j(mcdDQieQy3zBP)Zd_A;Ab;~tcnLFDHt4~AS9GE6pt
zM5yk6>z7bPFvVLriGSxPy-vs39Fw3GYhm9zKR2EnFim-CgN%#9Xn^>|e>WMVGF8jl
zOzGV8Fiy15cE^wbOKah7Vxsv71EaYrFxK$~`qg@qe(87>If!-eks6LSALa)JoWF`b
zsH0239-rYpiEI9MgarMc1B&c+m%Gq2^j|3QI5mTui)s3;e5L-D1{bYmAwfToT{+6r
z^ilas!z~SP+sdc{jqNY0K}+&lPNp}YaT3p|6Byp({!yK|9@;7(v&K1Dzl-m;sIqa`
zbv4$7J?SnCFTvVM*=g=fvMNlD^bE#{ldg}+U;4$;z`U)jDbV=UP4)Jr{qzZpb2UxB
zpRY8*(txq8Od;?e0T@2~V3s=C8>kO{YSHgBePaGnh}HkiS^`we{LAVaV0>~}?Q~<0
zkynvt`I;f^r)B`_FQ_4aNx>cBtiL05?lt@W%J@(7q=&GQ%QV_e6_-~V%L@-aeX;z0
z^SL@2>-pW2?4{|?^OZJO8o0NWEe9I++*C(j*~ct(JWU;zCDI_wqlh=Kxvsq={J|s^
zJ<s9sxj*io0!3Pr6vtuirsSsv&boyGWl$|gQbB(m6Hjz4$HXU<nEQfGE=MwxeDnlg
z(nq7zpT07*yM%fD-T#~BA7UwvOBoKVe2ya~%IFB)xg0+N9-+eiOq*kZCr7O%|K0YZ
z*(XXQJ?3SqE<p!xgr8(E;YSN|9p?9^F<C~(atMCvk7Q#0w}2MCTfDmx#PyR|C6VDt
z2uT1XspGS}Cy8WF=t$mZugWK~_5fSPu^d>fD&fluQ>QMapX^gfjCEE`1ZsUY)DWP^
zAS(5+U1MBlhFcb3|B~Q+o2E-nW)zQTJKHSI$0u%1SB40d)IEscEao>LiY@rRLNe~h
zI%R89rz>@P;ULsOYsvqf<Q1-G3_!*X5T*Oc-B&#Ir_49@U*%*&6r#tC-}RPFK<}OX
z{@MQNK*XVo%}@=S&3V%DD$3Laxatb~BJ*}egTo@uq?(Y$mdNJACHEhE$@>}=L^$^=
zY<5{j3R1dH;@i&sf;rRK4{#><{1cD?4<5&>(<QuKnK}*AJrJ(-Sza{zpy3jh-m+7v
zrJr}2Rbe9?qAp<L#5`sbO{OcAZt~%0>+yk+Tg;J0t!{J1rydd>V5oZ%^Uy?n{EnhF
z{T_uX%X<D$N!{JnX^m^eu$5|<9_o-Sp*!`#rkjMhfI^@r5~vlhi6#U@CS6PoOc`9i
zcZg=y<$j-Zn+z|c5ba5O*s6YDAI+z0m(J8hLRrwy|0Mk3g=V?k?KcE7)mpC}vJG3)
z|7qkpgPPj9wtyiTLh(`pgl<O(AvBQ=Vo>QIAc2Gs3`j2`RjMG;f&~Q$A|gmsnu0`#
zv=|9WRjw$Agd!!8>P0}>d!pX^-g)Qy@y$GwWcKX6_Das2eb!pfv%)Ohf!f=1FwffU
zq(5%s4#0HSXXFA~TrBl8*!v9_*WVssp`WtfC;?d79)5D(zRrGd@uBw(_)kas(p+^!
z4~CZRo0PDW$LT?X$<Ni6tbRCBs%u_FX+FaLBE(FxohTQ_NNo~uh1k6xw_^@pEMHyh
z+pM-`D03feDY7>Fx+4l-ll&<g%*r;226tT<-UdwQc4hZCA5t3j`E9O>m$l8I^#rr|
z;1NNuyO1YAdbM70`?wMc8@`G6{WZ;z|EjHs4&2qM)!tb0u90#n5a`7U0kJxwS=`%t
z#w|_+<I1gy9cBG4sw6r1&bFFTZw~~<TGoVcDL*QI@CCzP0yYpW8BN&;-X|Tgv+^$N
zbNnzprFeducFJ?Lj!b+2G?b7Fr+Z57qa&qKv`sIqBWjzp95l3Ppl7XGB8$k~DTd}V
z@|p}~;4kn9h8<-51XR>B!f;1qt_J;+<isnac1k<aA_b3v+Mjivrzgr2U(Q$bZ7w`r
zkZVg&rjun#8A#!jNMwu<6X`0D*SA?7Cz-G2m1{@OkR-kY%uA_tqynq2<B!}LXyU*K
z>NlZvy!pV2jwHMsR3P@J(VMIDX%fTgvV%`NSCp9{xe|0e$%&Sqh7YgcTZ9E-fgzGY
zYjHCYlyVShk${Il?cJRx=qULMY$1VIHIv4x192?lul&L(6x7#2j}@ZN)}K2Z@{-6X
zuuhm@2wQr*Pm@WB%EpK--=m&B!c<Aa2qt~Hr@D<H7?N&vAx9wOq13np!=D&!rI<uE
z;eT!xta|cI%y+UKNar0ZG~L&3ZSwQWfb5n%tK<`lnK{)c-P6)cmDlEH<H-=Pj<`^W
zvdp-{5hf)l+kQl$?mz@!Q5iPwfM8Nwv+bWNJV-RM0ws_kygEyukn}^i-Gy7i)!+=9
z2nDY{;<{wkKizY%eA%5>RH)u<_+Z}1Y9xW&%BPbI4uJ@di^(xnAd%HNwOeltGmk$n
ziZQlQNg%uM>eOxzNnfZ%L;&wTJI6cfx>c~&Glpx01dLpna<;mvp%*r`i?&W4*dA6n
z8W_*AZ12uXEyTp!8Sq+ed$u4hCr1N_YgPs>CVwC<(pt~53R3oD1rbKxsBr!j8$f*R
z*WxyYoa?z{w=-DR9rk!#@W6)aV<&!HJV1N{qK()1aGxz)LG>aJH)$3ANp^d&yxGqB
z@|eHxhPJ<}!sB;^*hRlHQ$(}b5#iGm!dMe&x$~cjBNZNfTJRPN<(dE}e=1C0M*JeU
z!bf)2lxFE~>3&}Rnd$71GllCVO^O6JT@qJe6P1P^P{tzsYZj#e%_1#1{QcAO`!c3Z
z$PO=~?o@zQyO16!yVMg4&lDV$en1dU>_B?F^7m0Kh#$-50ecYPHts2V)f@1q*i$RV
zG@7qT-R$V`B0lPr8I^k2VRJSGu8wxMW>aI-w2EItPZNdc*!-{rx`t_j)DK`{ih+7e
z-WA!=H&7gL1x0_9r!4Xjs!gP-i8i#5^A`JpqcuL?a#E+=9{IDwwr=`LIN>F9H&IfJ
zo+@>`qh{M>qKmPM`r~~35R^l7);6W)4fHKMR{ETJN!_!cs2h5bIHhh{2<((iobX_;
z+@AfAXsPaZ`?+VO$Nn+zPgb(SReK#3`=QT?OwDg1vk$TYUx=TQ57HLuvY+C?cFy~x
zaQEI>nW|xWUrMP{<lIETHiLNNq%C`;t2*MgyZGZ2&6@_#uAfop^ezFx?An}u5YoAt
zfEMBh-UfF!_KLUvV^|~5!%zjNb@VSuthpP?#3u#q5Tfro06v2~3Tm)#vtBRp-jsAO
z`Z_EZJuqu4?m$0t1gyytb$bZJe=XB}YF9JAv+v45g7~@ozp=(Jj|Z)fd^C2#*JIFX
zDn)_{(Po9PS4IIotDgnoKG?PelxD?!!+GR11jz>^h#~Y79`x{aD?Is{e3?QBP<VC`
zs8s-gKNa)m?ly*U@in=ezr6RX><e3GPZAPX=rPClzIfSV#{=!XzPxufV-}mp<*>^f
z&||5A7P3vS$P7rN19vbpYDo!xw!JJHovf!kZmUT8J+8kq+P=-H@OX6a=8s0{v+Yyo
zL|&K0pugxUzv;DLuUv1E!=wV&)h?qrGW&}+-%I}nPlEwyqIAw2FiSOFw=VOQlyXn6
zE88_(&*kQ384G;|6(!}VqdaZfn^nIwIHf;zvS9E+8&OR1#ZHub&MM#m1@%Kj>CTbi
zJ^H|`6Qq+-4iAg%WPiF{a#VSvS9|__S-bgCEc?@zmo4qv09Yi?aDM8FohQa_a(Kh@
zu(JTCxm>m1pQ}xI*39lZ^X+o=BGH*|(3nyari=Gw`>j5xehml+{i|SHjS4Rq@5H99
zTUIl8KKNJ4wZg66j7Z9yEC@H)#gR>>2l3u;z=Z#SvH^)xo_aYwj)7R2LbKD*ZKp3h
zbz($B{gegvaMh$sHU&qA4M@jBZNOeL=e~x@BmuPxN9KdB#)fJ}<bjAc&g_YP&fkg_
z;K^3ee45H%SELfQnM3HE3ZwF<*?pI&i6*)EXlk0$=JX>V;dYa_!c6};(+HM(Wi&K=
zM28lCYaW6%RXuh9FIf`vZhYy={{lRE{Gs35{nz?OuHyQy9|hMFFd>4MEW7O{D{qWy
zPE6669IbZF5>wdkRDt0W&6~5lTGOKK90%?gN3^|R(gr(x3~P7-yWbpp86C?qVz0L~
zNeMb^1O0Vn9DGknUUccwI9RF2g1vsl@JdeiwvzK1<4d{YOL5~%S>r<y2Pz*aRqo&O
z8@Szbj%JZBAdznJkp&8?389qG7?cY!x9b}19~j)E#a{c33evifFMGX-^zzwusnkeT
zb)0`Q-;iSD|3Fr5HNltiZ8+At#)QZ1V5IEs)dV1%73<k~t^aSR9%qv5Csns{yyJJ(
zn=bw^Tx3MdoM69(7hdmQy83@0|Eg+5%%gQj@}nq|MK1IgfpunbQ#DMEbvtLC30uD3
zHQXeQnKx=Wj>VxH&3R;nBlQ)31Mta%{2c}S?p)BXs?J&$oGN;<NuqIagZ4Rth9%Lu
zQe&!>!^+nf&fQKaLL}O3HZ344My)G=!k%ElfZ*QJ4Fj-LikweO&dGMzn8x1JJJ~b<
zjagj8W^y4!imBENlckB2XB(4FV4cjcs%F?YGi;lwF)nZfAA_Z^h1xk=%-@i0he;K^
z``1;W@5a%{^(ge%(8@i+p31tC`6{6EO2tzV9DJ$lUiuy!^fa|-@cw+yEt7>v_EGp}
z&i;dX79tFI0fqsiaPp%(=KB9BgQ!%AUueDE_vb<!gUHo^#iN`Be%C>MwK$-G7Kf4w
z-=<xeHvz3SPs4~9nO_qj?`RBU$9iN7w`Bu)L4YnTy9@=A<TBE#&1&c6$s+NAyRryl
z#jqtP+FtgTLW-rN!M$brjTO5t1ChU8;#$?Uz`6bRdIy8RvnH3;fn*ps?U(%J4K4$1
z3;<oa*PxgJrf>+HQCrnx^H<NOuJ%M>?D}tYj2_Nab43!U?yrJcGDZ>)wu0=l1ZMZR
z1nuulJlB}KaHIYQ2{@IIPpL>mz;%+qJ`tFO0wUMQ6`4SF=M8SzKAIR(p1dGS;n+9*
zA!v@Polb;wgKF;xlq$krCHr{<wcHp>%wvGm(*=kUu#UrsG9fof!)x`8bE;_jCmOv}
zce~)0JzrX>_dAskxiPLt3f28yFy8eC5e&0XMF7s(s5Bzw7*5@#74imBnQ6VWvDOHO
zn6$d_>}duF@M!F<_I|2+R&dMx(bkpKz9mxKF+NY^AQjoQFEoR@>8r}pimfp?A*ZbS
zq`<pu0Ut7~;|wB~;40ZvP`}Ce=uA=`6IN&?T#0N0eajLUvt7vDH6aV@&<!{;N!Dtk
zA~%nOZG4Km;^^!V6>O|Op7`;N;G`B^+4*llE031Rfq6vO24;&_gFLt$<9^65N!TuY
zEa>pi&fuA+@bAso58y9VO7vTC29BSF!7PPeVjvduUsAw_Bi`&E<wf4B8Zti#b8cv)
zC=BwB*K)o((j=eeI6BK~*~42NB=qy(l&*PDv1JR-cZ&yv`V~gtPl+6g?#9RJv?2{J
z{d$3AK&xssl)oALI0Pt{$fnNAmCWh~5~c~M_wE9flS|-AmY|;H5PD0)F;?(J)jjJZ
zww2&(9<2l54@gNA&2q@gnvpm@-Ee+nZo@?{CqFGgFey>qwV=MC2_aE+jM2<5&Ij)k
z^|SG=GOljsXYj(s1pFj#G`7YICSmNC%i~~!c4j;+{9i!u8+?8?{gwZnWg}UI53***
zzW*bC7uQf27%s-|2aFoyoB7SQ!TAr_EjJXV=j=q8i<p7nVmyA5de`d>FKFI$OfbsX
z;b$XMg|E^Ii`!^$J}Dl?3qM2<M?oj-mvI-RPMuIUzuoezZ)1kz)D<5%mkyettcl8#
z(1&e*nNQ~V?_Mb_hbU`-pPfH`58OQF7pxh!Rh|`bUN)fF(mf0Ky2|`oWs)E6dbrbB
z8+{u?(T>t9vvaAPa`$KFw(Rtxwv+6ay&CJE`~{94t=&&qa|(D>Y!k1;-mV30G#|pv
zx}2GN(<FWhS1M56Buz3`|B#P3U$Eg0+CpNZuOrrH3Nz7%;+&;#8kpZITwl#8G5kJw
zXvI$^spJNtB1)X(_#oYPYmZ#{2^c39C0A|<oBNuI=DcO}TIjhHA~r{@8O4IrqjNco
z;>>9~f~DTT4aCOkTpBw6U5WLDwqlj(w;yn6Xw18kO1%v9;M)?z4IjOMI|%vpV|oKu
z5#h0xddjD61#Zqoh&3hk83CQf0k)#S{iX(Pv!QkQuU_*49ds{;lZmOr72T%iqo0+$
z8Kx1wc*hKeKW6D>E}wg>51l>g=G$sAW833qkKt)EUd?x*qkXMSxO%pE`A~mxhoox#
z+R5h6{R3+*(Kz2Z-<gZ$FDg5eWGidOYA?{<XY~&(xaeYi+k9st{NXjGV6Fg!4hn+(
z7&yhk-5EG^S1DjdlPc8M=`(b+<h#UyL|im@Ht}s(?Q%<HX)XWNegj@{Yo9{kpiA(Y
zm+3=2y0~-S-uI1w0+cThZ`S&1`4D=&U##O?OkwlcnP_eQ=(&l;xqfcvY(U#m>~6or
zPIr;Ewckw?e2uhWPy=v;ECamT*UNlCB0>qQVS&vT|Nj@ScVAQsT~>SI7H5;T<Sil{
u8X#u_JVVcioW=WdKLvXDKtj*sL(WHpoFzguv@}s_>S~a^d(EtmL;eG6<Igex

literal 0
HcmV?d00001

diff --git a/previews/PR3525/api/JuMP.Containers/index.html b/previews/PR3525/api/JuMP.Containers/index.html
new file mode 100644
index 00000000000..3cca26de370
--- /dev/null
+++ b/previews/PR3525/api/JuMP.Containers/index.html
@@ -0,0 +1,192 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>JuMP.Containers · JuMP</title><meta name="title" content="JuMP.Containers · JuMP"/><meta property="og:title" content="JuMP.Containers · JuMP"/><meta property="twitter:title" content="JuMP.Containers · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox" checked/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../JuMP/">JuMP</a></li><li class="is-active"><a class="tocitem" href>JuMP.Containers</a><ul class="internal"><li><a class="tocitem" href="#DenseAxisArray"><span><code>DenseAxisArray</code></span></a></li><li><a class="tocitem" href="#SparseAxisArray"><span><code>SparseAxisArray</code></span></a></li><li><a class="tocitem" href="#Containers.@container"><span><code>Containers.@container</code></span></a></li><li><a class="tocitem" href="#Containers.container"><span><code>Containers.container</code></span></a></li><li><a class="tocitem" href="#Containers.rowtable"><span><code>Containers.rowtable</code></span></a></li><li><a class="tocitem" href="#Containers.default_container"><span><code>Containers.default_container</code></span></a></li><li><a class="tocitem" href="#Containers.nested"><span><code>Containers.nested</code></span></a></li><li><a class="tocitem" href="#Containers.vectorized_product"><span><code>Containers.vectorized_product</code></span></a></li><li><a class="tocitem" href="#Containers.build_ref_sets"><span><code>Containers.build_ref_sets</code></span></a></li><li><a class="tocitem" href="#Containers.container_code"><span><code>Containers.container_code</code></span></a></li><li><a class="tocitem" href="#Containers.AutoContainerType"><span><code>Containers.AutoContainerType</code></span></a></li><li><a class="tocitem" href="#Containers.NestedIterator"><span><code>Containers.NestedIterator</code></span></a></li><li><a class="tocitem" href="#Containers.VectorizedProductIterator"><span><code>Containers.VectorizedProductIterator</code></span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>JuMP.Containers</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>JuMP.Containers</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="DocumenterReference_JuMP.Containers"><a class="docs-heading-anchor" href="#DocumenterReference_JuMP.Containers">JuMP.Containers</a><a id="DocumenterReference_JuMP.Containers-1"></a><a class="docs-heading-anchor-permalink" href="#DocumenterReference_JuMP.Containers" title="Permalink"></a></h1><p>This page lists the public API of <code>JuMP.Containers</code>.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This page is an unstructured list of the JuMP.Containers API. For a more structured overview, read the Manual or Tutorial parts of this documentation.</p></div></div><p>Load all of the public the API into the current scope with:</p><pre><code class="language-julia hljs">using JuMP.Containers</code></pre><p>Alternatively, load only the module with:</p><pre><code class="language-julia hljs">import JuMP.Containers</code></pre><p>and then prefix all calls with <code>JuMP.Containers.</code> to create <code>JuMP.Containers.&lt;NAME&gt;</code>.</p><h2 id="DenseAxisArray"><a class="docs-heading-anchor" href="#DenseAxisArray"><code>DenseAxisArray</code></a><a id="DenseAxisArray-1"></a><a class="docs-heading-anchor-permalink" href="#DenseAxisArray" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.DenseAxisArray" href="#JuMP.Containers.DenseAxisArray"><code>JuMP.Containers.DenseAxisArray</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">DenseAxisArray(data::Array{T, N}, axes...) where {T, N}</code></pre><p>Construct a JuMP array with the underlying data specified by the <code>data</code> array and the given axes. Exactly <code>N</code> axes must be provided, and their lengths must match <code>size(data)</code> in the corresponding dimensions.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; array = Containers.DenseAxisArray([1 2; 3 4], [:a, :b], 2:3)
+2-dimensional DenseAxisArray{Int64,2,...} with index sets:
+    Dimension 1, [:a, :b]
+    Dimension 2, 2:3
+And data, a 2×2 Matrix{Int64}:
+ 1  2
+ 3  4
+
+julia&gt; array[:b, 3]
+4</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/DenseAxisArray.jl#L194-L215">source</a></section><section><div><pre><code class="language-julia hljs">DenseAxisArray{T}(undef, axes...) where T</code></pre><p>Construct an uninitialized DenseAxisArray with element-type <code>T</code> indexed over the given axes.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; array = Containers.DenseAxisArray{Float64}(undef, [:a, :b], 1:2);
+
+julia&gt; fill!(array, 1.0)
+2-dimensional DenseAxisArray{Float64,2,...} with index sets:
+    Dimension 1, [:a, :b]
+    Dimension 2, 1:2
+And data, a 2×2 Matrix{Float64}:
+ 1.0  1.0
+ 1.0  1.0
+
+julia&gt; array[:a, 2] = 5.0
+5.0
+
+julia&gt; array[:a, 2]
+5.0
+
+julia&gt; array
+2-dimensional DenseAxisArray{Float64,2,...} with index sets:
+    Dimension 1, [:a, :b]
+    Dimension 2, 1:2
+And data, a 2×2 Matrix{Float64}:
+ 1.0  5.0
+ 1.0  1.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/DenseAxisArray.jl#L232-L265">source</a></section></article><h2 id="SparseAxisArray"><a class="docs-heading-anchor" href="#SparseAxisArray"><code>SparseAxisArray</code></a><a id="SparseAxisArray-1"></a><a class="docs-heading-anchor-permalink" href="#SparseAxisArray" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.SparseAxisArray" href="#JuMP.Containers.SparseAxisArray"><code>JuMP.Containers.SparseAxisArray</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct SparseAxisArray{T,N,K&lt;:NTuple{N, Any}} &lt;: AbstractArray{T,N}
+    data::Dict{K,T}
+end</code></pre><p><code>N</code>-dimensional array with elements of type <code>T</code> where only a subset of the entries are defined. The entries with indices <code>idx = (i1, i2, ..., iN)</code> in <code>keys(data)</code> has value <code>data[idx]</code>. Note that as opposed to <code>SparseArrays.AbstractSparseArray</code>, the missing entries are not assumed to be <code>zero(T)</code>, they are simply not part of the array. This means that the result of <code>map(f, sa::SparseAxisArray)</code> or <code>f.(sa::SparseAxisArray)</code> has the same sparsity structure than <code>sa</code> even if <code>f(zero(T))</code> is not zero.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; dict = Dict((:a, 2) =&gt; 1.0, (:a, 3) =&gt; 2.0, (:b, 3) =&gt; 3.0)
+Dict{Tuple{Symbol, Int64}, Float64} with 3 entries:
+  (:a, 3) =&gt; 2.0
+  (:b, 3) =&gt; 3.0
+  (:a, 2) =&gt; 1.0
+
+julia&gt; array = Containers.SparseAxisArray(dict)
+SparseAxisArray{Float64, 2, Tuple{Symbol, Int64}} with 3 entries:
+  [a, 2]  =  1.0
+  [a, 3]  =  2.0
+  [b, 3]  =  3.0
+
+julia&gt; array[:b, 3]
+3.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/SparseAxisArray.jl#L6-L37">source</a></section></article><h2 id="Containers.@container"><a class="docs-heading-anchor" href="#Containers.@container"><code>Containers.@container</code></a><a id="Containers.@container-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.@container" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.@container" href="#JuMP.Containers.@container"><code>JuMP.Containers.@container</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@container([i=..., j=..., ...], expr[, container = :Auto])</code></pre><p>Create a container with indices <code>i</code>, <code>j</code>, ... and values given by <code>expr</code> that may depend on the value of the indices.</p><pre><code class="nohighlight hljs">@container(ref[i=..., j=..., ...], expr[, container = :Auto])</code></pre><p>Same as above but the container is assigned to the variable of name <code>ref</code>.</p><p>The type of container can be controlled by the <code>container</code> keyword.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>When the index set is explicitly given as <code>1:n</code> for any expression <code>n</code>, it is transformed to <code>Base.OneTo(n)</code> before being given to <a href="#JuMP.Containers.container"><code>container</code></a>.</p></div></div><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; Containers.@container([i = 1:3, j = 1:3], i + j)
+3×3 Matrix{Int64}:
+ 2  3  4
+ 3  4  5
+ 4  5  6
+
+julia&gt; I = 1:3
+1:3
+
+julia&gt; Containers.@container(x[i = I, j = I], i + j);
+
+julia&gt; x
+2-dimensional DenseAxisArray{Int64,2,...} with index sets:
+    Dimension 1, 1:3
+    Dimension 2, 1:3
+And data, a 3×3 Matrix{Int64}:
+ 2  3  4
+ 3  4  5
+ 4  5  6
+
+julia&gt; Containers.@container([i = 2:3, j = 1:3], i + j)
+2-dimensional DenseAxisArray{Int64,2,...} with index sets:
+    Dimension 1, 2:3
+    Dimension 2, Base.OneTo(3)
+And data, a 2×3 Matrix{Int64}:
+ 3  4  5
+ 4  5  6
+
+julia&gt; Containers.@container([i = 1:3, j = 1:3; i &lt;= j], i + j)
+SparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 6 entries:
+  [1, 1]  =  2
+  [1, 2]  =  3
+  [1, 3]  =  4
+  [2, 2]  =  4
+  [2, 3]  =  5
+  [3, 3]  =  6</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/macro.jl#L320-L376">source</a></section></article><h2 id="Containers.container"><a class="docs-heading-anchor" href="#Containers.container"><code>Containers.container</code></a><a id="Containers.container-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.container" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.container" href="#JuMP.Containers.container"><code>JuMP.Containers.container</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">container(f::Function, indices[[, ::Type{C} = AutoContainerType], names])</code></pre><p>Create a container of type <code>C</code> with index names <code>names</code>, indices <code>indices</code> and values at given indices given by <code>f</code>.</p><p>If the method with <code>names</code> is not specialized on <code>Type{C}</code>, it falls back to calling  <code>container(f, indices, c)</code> for backwards compatibility with containers not supporting index names.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; Containers.container((i, j) -&gt; i + j, Containers.vectorized_product(Base.OneTo(3), Base.OneTo(3)))
+3×3 Matrix{Int64}:
+ 2  3  4
+ 3  4  5
+ 4  5  6
+
+julia&gt; Containers.container((i, j) -&gt; i + j, Containers.vectorized_product(1:3, 1:3))
+2-dimensional DenseAxisArray{Int64,2,...} with index sets:
+    Dimension 1, 1:3
+    Dimension 2, 1:3
+And data, a 3×3 Matrix{Int64}:
+ 2  3  4
+ 3  4  5
+ 4  5  6
+
+julia&gt; Containers.container((i, j) -&gt; i + j, Containers.vectorized_product(2:3, Base.OneTo(3)))
+2-dimensional DenseAxisArray{Int64,2,...} with index sets:
+    Dimension 1, 2:3
+    Dimension 2, Base.OneTo(3)
+And data, a 2×3 Matrix{Int64}:
+ 3  4  5
+ 4  5  6
+
+julia&gt; Containers.container((i, j) -&gt; i + j, Containers.nested(() -&gt; 1:3, i -&gt; i:3, condition = (i, j) -&gt; isodd(i) || isodd(j)))
+SparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 5 entries:
+  [1, 1]  =  2
+  [1, 2]  =  3
+  [1, 3]  =  4
+  [2, 3]  =  5
+  [3, 3]  =  6</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/container.jl#L25-L69">source</a></section></article><h2 id="Containers.rowtable"><a class="docs-heading-anchor" href="#Containers.rowtable"><code>Containers.rowtable</code></a><a id="Containers.rowtable-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.rowtable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.rowtable" href="#JuMP.Containers.rowtable"><code>JuMP.Containers.rowtable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">rowtable([f::Function=identity,] x; [header::Vector{Symbol} = Symbol[]])</code></pre><p>Applies the function <code>f</code> to all elements of the variable container <code>x</code>, returning the result as a <code>Vector</code> of <code>NamedTuple</code>s, where <code>header</code> is a vector containing the corresponding axis names.</p><p>If <code>x</code> is an <code>N</code>-dimensional array, there must be <code>N+1</code> names, so that the last name corresponds to the result of <code>f(x[i])</code>.</p><p>If <code>header</code> is left empty, then the default header is <code>[:x1, :x2, ..., :xN, :y]</code>.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>A <code>Vector</code> of <code>NamedTuple</code>s implements the <a href="https://github.com/JuliaData/Tables.jl">Tables.jl</a> interface, and so the result can be used as input for any function that consumes a &#39;Tables.jl&#39; compatible source.</p></div></div><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2, j=i:2] &gt;= 0, start = i+j);
+
+julia&gt; Containers.rowtable(start_value, x; header = [:i, :j, :start])
+3-element Vector{NamedTuple{(:i, :j, :start), Tuple{Int64, Int64, Float64}}}:
+ (i = 1, j = 2, start = 3.0)
+ (i = 1, j = 1, start = 2.0)
+ (i = 2, j = 2, start = 4.0)
+
+julia&gt; Containers.rowtable(x)
+3-element Vector{NamedTuple{(:x1, :x2, :y), Tuple{Int64, Int64, VariableRef}}}:
+ (x1 = 1, x2 = 2, y = x[1,2])
+ (x1 = 1, x2 = 1, y = x[1,1])
+ (x1 = 2, x2 = 2, y = x[2,2])</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/tables.jl#L12-L48">source</a></section></article><h2 id="Containers.default_container"><a class="docs-heading-anchor" href="#Containers.default_container"><code>Containers.default_container</code></a><a id="Containers.default_container-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.default_container" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.default_container" href="#JuMP.Containers.default_container"><code>JuMP.Containers.default_container</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">default_container(indices)</code></pre><p>If <code>indices</code> is a <a href="#JuMP.Containers.NestedIterator"><code>NestedIterator</code></a>, return a <a href="#JuMP.Containers.SparseAxisArray"><code>SparseAxisArray</code></a>. Otherwise, <code>indices</code> should be a <code>VectorizedProductIterator</code> and the function returns <code>Array</code> if all iterators of the product are <code>Base.OneTo</code> and retunrs <a href="#JuMP.Containers.DenseAxisArray"><code>DenseAxisArray</code></a> otherwise.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/container.jl#L6-L14">source</a></section></article><h2 id="Containers.nested"><a class="docs-heading-anchor" href="#Containers.nested"><code>Containers.nested</code></a><a id="Containers.nested-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.nested" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.nested" href="#JuMP.Containers.nested"><code>JuMP.Containers.nested</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">nested(iterators...; condition = (args...) -&gt; true)</code></pre><p>Create a <a href="#JuMP.Containers.NestedIterator"><code>NestedIterator</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; iterator = Containers.nested(
+           () -&gt; 1:2,
+           (i,) -&gt; [&quot;A&quot;, &quot;B&quot;];
+           condition = (i, j) -&gt; isodd(i) || j == &quot;B&quot;,
+       );
+
+julia&gt; collect(iterator)
+3-element Vector{Tuple{Int64, String}}:
+ (1, &quot;A&quot;)
+ (1, &quot;B&quot;)
+ (2, &quot;B&quot;)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/nested_iterator.jl#L51-L71">source</a></section></article><h2 id="Containers.vectorized_product"><a class="docs-heading-anchor" href="#Containers.vectorized_product"><code>Containers.vectorized_product</code></a><a id="Containers.vectorized_product-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.vectorized_product" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.vectorized_product" href="#JuMP.Containers.vectorized_product"><code>JuMP.Containers.vectorized_product</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">vectorized_product(iterators...)</code></pre><p>Created a <a href="#JuMP.Containers.VectorizedProductIterator"><code>VectorizedProductIterator</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; iterator = Containers.vectorized_product(1:2, [&quot;A&quot;, &quot;B&quot;]);
+
+julia&gt; collect(iterator)
+2×2 Matrix{Tuple{Int64, String}}:
+ (1, &quot;A&quot;)  (1, &quot;B&quot;)
+ (2, &quot;A&quot;)  (2, &quot;B&quot;)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/vectorized_product_iterator.jl#L56-L71">source</a></section></article><h2 id="Containers.build_ref_sets"><a class="docs-heading-anchor" href="#Containers.build_ref_sets"><code>Containers.build_ref_sets</code></a><a id="Containers.build_ref_sets-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.build_ref_sets" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.build_ref_sets" href="#JuMP.Containers.build_ref_sets"><code>JuMP.Containers.build_ref_sets</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">build_ref_sets(_error::Function, expr)</code></pre><p>Helper function for macros to construct container objects.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This function is for advanced users implementing JuMP extensions. See <a href="#JuMP.Containers.container_code"><code>container_code</code></a> for more details.</p></div></div><p><strong>Arguments</strong></p><ul><li><code>_error</code>: a function that takes a <code>String</code> and throws an error, potentially annotating the input string with extra information such as from which macro it was thrown from. Use <code>error</code> if you do not want a modified error message.</li><li><code>expr</code>: an <code>Expr</code> that specifies the container, e.g., <code>:(x[i = 1:3, [:red, :blue], k = S; i + k &lt;= 6])</code></li></ul><p><strong>Returns</strong></p><ol><li><code>index_vars</code>: a <code>Vector{Any}</code> of names for the index variables, e.g., <code>[:i, gensym(), :k]</code>. These may also be expressions, like <code>:((i, j))</code> from a call like <code>:(x[(i, j) in S])</code>.</li><li><code>indices</code>: an iterator over the indices, for example, <a href="#JuMP.Containers.NestedIterator"><code>Containers.NestedIterator</code></a></li></ol><p><strong>Example</strong></p><p>See <a href="#JuMP.Containers.container_code"><code>container_code</code></a> for a worked example.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/macro.jl#L182-L210">source</a></section></article><h2 id="Containers.container_code"><a class="docs-heading-anchor" href="#Containers.container_code"><code>Containers.container_code</code></a><a id="Containers.container_code-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.container_code" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.container_code" href="#JuMP.Containers.container_code"><code>JuMP.Containers.container_code</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">container_code(
+    index_vars::Vector{Any},
+    indices::Expr,
+    code,
+    requested_container::Union{Symbol,Expr},
+)</code></pre><p>Used in macros to construct a call to <a href="#JuMP.Containers.container"><code>container</code></a>. This should be used in conjunction with <a href="#JuMP.Containers.build_ref_sets"><code>build_ref_sets</code></a>.</p><p><strong>Arguments</strong></p><ul><li><code>index_vars::Vector{Any}</code>: a vector of names for the indices of the container. These may also be expressions, like <code>:((i, j))</code> from a call like <code>:(x[(i, j) in S])</code>.</li><li><code>indices::Expr</code>: an expression that evaluates to an iterator of the indices.</li><li><code>code</code>: an expression or literal constant for the value to be stored in the container as a function of the named <code>index_vars</code>.</li><li><code>requested_container</code>: passed to the third argument of <a href="#JuMP.Containers.container"><code>container</code></a>. For built-in JuMP types, choose one of <code>:Array</code>, <code>:DenseAxisArray</code>, <code>:SparseAxisArray</code>, or <code>:Auto</code>. For a user-defined container, this expression must evaluate to the correct type.</li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>In most cases, you should <code>esc(code)</code> before passing it to <code>container_code</code>.</p></div></div><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; macro foo(ref_sets, code)
+           index_vars, indices = Containers.build_ref_sets(error, ref_sets)
+           return Containers.container_code(
+               index_vars,
+               indices,
+               esc(code),
+               :Auto,
+            )
+       end
+@foo (macro with 1 method)
+
+julia&gt; @foo(x[i=1:2, j=[&quot;A&quot;, &quot;B&quot;]], j^i)
+2-dimensional DenseAxisArray{String,2,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+    Dimension 2, [&quot;A&quot;, &quot;B&quot;]
+And data, a 2×2 Matrix{String}:
+ &quot;A&quot;   &quot;B&quot;
+ &quot;AA&quot;  &quot;BB&quot;</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/macro.jl#L241-L290">source</a></section></article><h2 id="Containers.AutoContainerType"><a class="docs-heading-anchor" href="#Containers.AutoContainerType"><code>Containers.AutoContainerType</code></a><a id="Containers.AutoContainerType-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.AutoContainerType" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.AutoContainerType" href="#JuMP.Containers.AutoContainerType"><code>JuMP.Containers.AutoContainerType</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AutoContainerType</code></pre><p>Pass <code>AutoContainerType</code> to <a href="#JuMP.Containers.container"><code>container</code></a> to let the container type be chosen based on the type of the indices using <a href="#JuMP.Containers.default_container"><code>default_container</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/container.jl#L17-L22">source</a></section></article><h2 id="Containers.NestedIterator"><a class="docs-heading-anchor" href="#Containers.NestedIterator"><code>Containers.NestedIterator</code></a><a id="Containers.NestedIterator-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.NestedIterator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.NestedIterator" href="#JuMP.Containers.NestedIterator"><code>JuMP.Containers.NestedIterator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct NestedIterator{T}
+    iterators::T # Tuple of functions
+    condition::Function
+end</code></pre><p>Iterators over the tuples that are produced by a nested for loop.</p><p>Construct a <code>NestedIterator</code> using <a href="#JuMP.Containers.nested"><code>nested</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; iterators = (() -&gt; 1:2, (i,) -&gt; [&quot;A&quot;, &quot;B&quot;]);
+
+julia&gt; condition = (i, j) -&gt; isodd(i) || j == &quot;B&quot;;
+
+julia&gt; x = Containers.NestedIterator(iterators, condition);
+
+julia&gt; for (i, j) in x
+           println((i, j))
+       end
+(1, &quot;A&quot;)
+(1, &quot;B&quot;)
+(2, &quot;B&quot;)</code></pre><p>is the same as</p><pre><code class="language-julia-repl hljs">julia&gt; for i in iterators[1]()
+           for j in iterators[2](i)
+               if condition(i, j)
+                   println((i, j))
+               end
+           end
+       end
+(1, &quot;A&quot;)
+(1, &quot;B&quot;)
+(2, &quot;B&quot;)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/nested_iterator.jl#L6-L45">source</a></section></article><h2 id="Containers.VectorizedProductIterator"><a class="docs-heading-anchor" href="#Containers.VectorizedProductIterator"><code>Containers.VectorizedProductIterator</code></a><a id="Containers.VectorizedProductIterator-1"></a><a class="docs-heading-anchor-permalink" href="#Containers.VectorizedProductIterator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Containers.VectorizedProductIterator" href="#JuMP.Containers.VectorizedProductIterator"><code>JuMP.Containers.VectorizedProductIterator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct VectorizedProductIterator{T}
+    prod::Iterators.ProductIterator{T}
+end</code></pre><p>A wrapper type for <code>Iterators.ProuctIterator</code> that discards shape information and returns a <code>Vector</code>.</p><p>Construct a <code>VectorizedProductIterator</code> using <a href="#JuMP.Containers.vectorized_product"><code>vectorized_product</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/Containers/vectorized_product_iterator.jl#L31-L40">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../JuMP/">« JuMP</a><a class="docs-footer-nextpage" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/api/JuMP/index.html b/previews/PR3525/api/JuMP/index.html
new file mode 100644
index 00000000000..8f6e605af30
--- /dev/null
+++ b/previews/PR3525/api/JuMP/index.html
@@ -0,0 +1,2074 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>JuMP · JuMP</title><meta name="title" content="JuMP · JuMP"/><meta property="og:title" content="JuMP · JuMP"/><meta property="twitter:title" content="JuMP · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox" checked/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>JuMP</a><ul class="internal"><li><a class="tocitem" href="#@NLconstraint"><span><code>@NLconstraint</code></span></a></li><li><a class="tocitem" href="#@NLconstraints"><span><code>@NLconstraints</code></span></a></li><li><a class="tocitem" href="#@NLexpression"><span><code>@NLexpression</code></span></a></li><li><a class="tocitem" href="#@NLexpressions"><span><code>@NLexpressions</code></span></a></li><li><a class="tocitem" href="#@NLobjective"><span><code>@NLobjective</code></span></a></li><li><a class="tocitem" href="#@NLparameter"><span><code>@NLparameter</code></span></a></li><li><a class="tocitem" href="#@NLparameters"><span><code>@NLparameters</code></span></a></li><li><a class="tocitem" href="#@build_constraint"><span><code>@build_constraint</code></span></a></li><li><a class="tocitem" href="#@constraint"><span><code>@constraint</code></span></a></li><li><a class="tocitem" href="#@constraints"><span><code>@constraints</code></span></a></li><li><a class="tocitem" href="#@expression"><span><code>@expression</code></span></a></li><li><a class="tocitem" href="#@expressions"><span><code>@expressions</code></span></a></li><li><a class="tocitem" href="#@objective"><span><code>@objective</code></span></a></li><li><a class="tocitem" href="#@operator"><span><code>@operator</code></span></a></li><li><a class="tocitem" href="#@variable"><span><code>@variable</code></span></a></li><li><a class="tocitem" href="#@variables"><span><code>@variables</code></span></a></li><li><a class="tocitem" href="#add_bridge"><span><code>add_bridge</code></span></a></li><li><a class="tocitem" href="#add_constraint"><span><code>add_constraint</code></span></a></li><li><a class="tocitem" href="#add_nonlinear_constraint"><span><code>add_nonlinear_constraint</code></span></a></li><li><a class="tocitem" href="#add_nonlinear_expression"><span><code>add_nonlinear_expression</code></span></a></li><li><a class="tocitem" href="#add_nonlinear_operator"><span><code>add_nonlinear_operator</code></span></a></li><li><a class="tocitem" href="#add_nonlinear_parameter"><span><code>add_nonlinear_parameter</code></span></a></li><li><a class="tocitem" href="#add_to_expression!"><span><code>add_to_expression!</code></span></a></li><li><a class="tocitem" href="#add_to_function_constant"><span><code>add_to_function_constant</code></span></a></li><li><a class="tocitem" href="#add_variable"><span><code>add_variable</code></span></a></li><li><a class="tocitem" href="#all_constraints"><span><code>all_constraints</code></span></a></li><li><a class="tocitem" href="#all_nonlinear_constraints"><span><code>all_nonlinear_constraints</code></span></a></li><li><a class="tocitem" href="#all_variables"><span><code>all_variables</code></span></a></li><li><a class="tocitem" href="#anonymous_name"><span><code>anonymous_name</code></span></a></li><li><a class="tocitem" href="#backend"><span><code>backend</code></span></a></li><li><a class="tocitem" href="#barrier_iterations"><span><code>barrier_iterations</code></span></a></li><li><a class="tocitem" href="#bridge_constraints"><span><code>bridge_constraints</code></span></a></li><li><a class="tocitem" href="#build_constraint"><span><code>build_constraint</code></span></a></li><li><a class="tocitem" href="#build_variable"><span><code>build_variable</code></span></a></li><li><a class="tocitem" href="#callback_node_status"><span><code>callback_node_status</code></span></a></li><li><a class="tocitem" href="#callback_value"><span><code>callback_value</code></span></a></li><li><a class="tocitem" href="#check_belongs_to_model"><span><code>check_belongs_to_model</code></span></a></li><li><a class="tocitem" href="#coefficient"><span><code>coefficient</code></span></a></li><li><a class="tocitem" href="#compute_conflict!"><span><code>compute_conflict!</code></span></a></li><li><a class="tocitem" href="#constant"><span><code>constant</code></span></a></li><li><a class="tocitem" href="#constraint_by_name"><span><code>constraint_by_name</code></span></a></li><li><a class="tocitem" href="#constraint_object"><span><code>constraint_object</code></span></a></li><li><a class="tocitem" href="#constraint_ref_with_index"><span><code>constraint_ref_with_index</code></span></a></li><li><a class="tocitem" href="#constraint_string"><span><code>constraint_string</code></span></a></li><li><a class="tocitem" href="#constraints_string"><span><code>constraints_string</code></span></a></li><li><a class="tocitem" href="#copy_conflict"><span><code>copy_conflict</code></span></a></li><li><a class="tocitem" href="#copy_extension_data"><span><code>copy_extension_data</code></span></a></li><li><a class="tocitem" href="#copy_model"><span><code>copy_model</code></span></a></li><li><a class="tocitem" href="#delete"><span><code>delete</code></span></a></li><li><a class="tocitem" href="#delete_lower_bound"><span><code>delete_lower_bound</code></span></a></li><li><a class="tocitem" href="#delete_upper_bound"><span><code>delete_upper_bound</code></span></a></li><li><a class="tocitem" href="#direct_generic_model"><span><code>direct_generic_model</code></span></a></li><li><a class="tocitem" href="#direct_model"><span><code>direct_model</code></span></a></li><li><a class="tocitem" href="#drop_zeros!"><span><code>drop_zeros!</code></span></a></li><li><a class="tocitem" href="#dual"><span><code>dual</code></span></a></li><li><a class="tocitem" href="#dual_objective_value"><span><code>dual_objective_value</code></span></a></li><li><a class="tocitem" href="#dual_shape"><span><code>dual_shape</code></span></a></li><li><a class="tocitem" href="#dual_start_value"><span><code>dual_start_value</code></span></a></li><li><a class="tocitem" href="#dual_status"><span><code>dual_status</code></span></a></li><li><a class="tocitem" href="#error_if_direct_mode"><span><code>error_if_direct_mode</code></span></a></li><li><a class="tocitem" href="#fix"><span><code>fix</code></span></a></li><li><a class="tocitem" href="#fix_discrete_variables"><span><code>fix_discrete_variables</code></span></a></li><li><a class="tocitem" href="#fix_value"><span><code>fix_value</code></span></a></li><li><a class="tocitem" href="#flatten!"><span><code>flatten!</code></span></a></li><li><a class="tocitem" href="#function_string"><span><code>function_string</code></span></a></li><li><a class="tocitem" href="#get_attribute"><span><code>get_attribute</code></span></a></li><li><a class="tocitem" href="#get_optimizer_attribute"><span><code>get_optimizer_attribute</code></span></a></li><li><a class="tocitem" href="#has_duals"><span><code>has_duals</code></span></a></li><li><a class="tocitem" href="#has_lower_bound"><span><code>has_lower_bound</code></span></a></li><li><a class="tocitem" href="#has_start_value"><span><code>has_start_value</code></span></a></li><li><a class="tocitem" href="#has_upper_bound"><span><code>has_upper_bound</code></span></a></li><li><a class="tocitem" href="#has_values"><span><code>has_values</code></span></a></li><li><a class="tocitem" href="#in_set_string"><span><code>in_set_string</code></span></a></li><li><a class="tocitem" href="#index"><span><code>index</code></span></a></li><li><a class="tocitem" href="#is_binary"><span><code>is_binary</code></span></a></li><li><a class="tocitem" href="#is_fixed"><span><code>is_fixed</code></span></a></li><li><a class="tocitem" href="#is_integer"><span><code>is_integer</code></span></a></li><li><a class="tocitem" href="#is_parameter"><span><code>is_parameter</code></span></a></li><li><a class="tocitem" href="#is_valid"><span><code>is_valid</code></span></a></li><li><a class="tocitem" href="#isequal_canonical"><span><code>isequal_canonical</code></span></a></li><li><a class="tocitem" href="#jump_function"><span><code>jump_function</code></span></a></li><li><a class="tocitem" href="#jump_function_type"><span><code>jump_function_type</code></span></a></li><li><a class="tocitem" href="#latex_formulation"><span><code>latex_formulation</code></span></a></li><li><a class="tocitem" href="#linear_terms"><span><code>linear_terms</code></span></a></li><li><a class="tocitem" href="#list_of_constraint_types"><span><code>list_of_constraint_types</code></span></a></li><li><a class="tocitem" href="#lower_bound"><span><code>lower_bound</code></span></a></li><li><a class="tocitem" href="#lp_sensitivity_report"><span><code>lp_sensitivity_report</code></span></a></li><li><a class="tocitem" href="#map_coefficients"><span><code>map_coefficients</code></span></a></li><li><a class="tocitem" href="#map_coefficients_inplace!"><span><code>map_coefficients_inplace!</code></span></a></li><li><a class="tocitem" href="#mode"><span><code>mode</code></span></a></li><li><a class="tocitem" href="#model_convert"><span><code>model_convert</code></span></a></li><li><a class="tocitem" href="#model_string"><span><code>model_string</code></span></a></li><li><a class="tocitem" href="#moi_function"><span><code>moi_function</code></span></a></li><li><a class="tocitem" href="#moi_function_type"><span><code>moi_function_type</code></span></a></li><li><a class="tocitem" href="#moi_set"><span><code>moi_set</code></span></a></li><li><a class="tocitem" href="#name"><span><code>name</code></span></a></li><li><a class="tocitem" href="#node_count"><span><code>node_count</code></span></a></li><li><a class="tocitem" href="#nonlinear_constraint_string"><span><code>nonlinear_constraint_string</code></span></a></li><li><a class="tocitem" href="#nonlinear_dual_start_value"><span><code>nonlinear_dual_start_value</code></span></a></li><li><a class="tocitem" href="#nonlinear_expr_string"><span><code>nonlinear_expr_string</code></span></a></li><li><a class="tocitem" href="#nonlinear_model"><span><code>nonlinear_model</code></span></a></li><li><a class="tocitem" href="#normalized_coefficient"><span><code>normalized_coefficient</code></span></a></li><li><a class="tocitem" href="#normalized_rhs"><span><code>normalized_rhs</code></span></a></li><li><a class="tocitem" href="#num_constraints"><span><code>num_constraints</code></span></a></li><li><a class="tocitem" href="#num_nonlinear_constraints"><span><code>num_nonlinear_constraints</code></span></a></li><li><a class="tocitem" href="#num_variables"><span><code>num_variables</code></span></a></li><li><a class="tocitem" href="#object_dictionary"><span><code>object_dictionary</code></span></a></li><li><a class="tocitem" href="#objective_bound"><span><code>objective_bound</code></span></a></li><li><a class="tocitem" href="#objective_function"><span><code>objective_function</code></span></a></li><li><a class="tocitem" href="#objective_function_string"><span><code>objective_function_string</code></span></a></li><li><a class="tocitem" href="#objective_function_type"><span><code>objective_function_type</code></span></a></li><li><a class="tocitem" href="#objective_sense"><span><code>objective_sense</code></span></a></li><li><a class="tocitem" href="#objective_value"><span><code>objective_value</code></span></a></li><li><a class="tocitem" href="#op_ifelse"><span><code>op_ifelse</code></span></a></li><li><a class="tocitem" href="#operator_to_set"><span><code>operator_to_set</code></span></a></li><li><a class="tocitem" href="#operator_warn"><span><code>operator_warn</code></span></a></li><li><a class="tocitem" href="#optimize!"><span><code>optimize!</code></span></a></li><li><a class="tocitem" href="#optimizer_index"><span><code>optimizer_index</code></span></a></li><li><a class="tocitem" href="#optimizer_with_attributes"><span><code>optimizer_with_attributes</code></span></a></li><li><a class="tocitem" href="#owner_model"><span><code>owner_model</code></span></a></li><li><a class="tocitem" href="#parameter_value"><span><code>parameter_value</code></span></a></li><li><a class="tocitem" href="#parse_constraint"><span><code>parse_constraint</code></span></a></li><li><a class="tocitem" href="#parse_constraint_call"><span><code>parse_constraint_call</code></span></a></li><li><a class="tocitem" href="#parse_constraint_head"><span><code>parse_constraint_head</code></span></a></li><li><a class="tocitem" href="#parse_one_operator_variable"><span><code>parse_one_operator_variable</code></span></a></li><li><a class="tocitem" href="#parse_ternary_variable"><span><code>parse_ternary_variable</code></span></a></li><li><a class="tocitem" href="#parse_variable"><span><code>parse_variable</code></span></a></li><li><a class="tocitem" href="#primal_feasibility_report"><span><code>primal_feasibility_report</code></span></a></li><li><a class="tocitem" href="#primal_status"><span><code>primal_status</code></span></a></li><li><a class="tocitem" href="#print_active_bridges"><span><code>print_active_bridges</code></span></a></li><li><a class="tocitem" href="#print_bridge_graph"><span><code>print_bridge_graph</code></span></a></li><li><a class="tocitem" href="#quad_terms"><span><code>quad_terms</code></span></a></li><li><a class="tocitem" href="#raw_status"><span><code>raw_status</code></span></a></li><li><a class="tocitem" href="#read_from_file"><span><code>read_from_file</code></span></a></li><li><a class="tocitem" href="#reduced_cost"><span><code>reduced_cost</code></span></a></li><li><a class="tocitem" href="#register"><span><code>register</code></span></a></li><li><a class="tocitem" href="#relative_gap"><span><code>relative_gap</code></span></a></li><li><a class="tocitem" href="#relax_integrality"><span><code>relax_integrality</code></span></a></li><li><a class="tocitem" href="#relax_with_penalty!"><span><code>relax_with_penalty!</code></span></a></li><li><a class="tocitem" href="#remove_bridge"><span><code>remove_bridge</code></span></a></li><li><a class="tocitem" href="#reshape_set"><span><code>reshape_set</code></span></a></li><li><a class="tocitem" href="#reshape_vector"><span><code>reshape_vector</code></span></a></li><li><a class="tocitem" href="#result_count"><span><code>result_count</code></span></a></li><li><a class="tocitem" href="#reverse_sense"><span><code>reverse_sense</code></span></a></li><li><a class="tocitem" href="#set_attribute"><span><code>set_attribute</code></span></a></li><li><a class="tocitem" href="#set_attributes"><span><code>set_attributes</code></span></a></li><li><a class="tocitem" href="#set_binary"><span><code>set_binary</code></span></a></li><li><a class="tocitem" href="#set_dual_start_value"><span><code>set_dual_start_value</code></span></a></li><li><a class="tocitem" href="#set_integer"><span><code>set_integer</code></span></a></li><li><a class="tocitem" href="#set_lower_bound"><span><code>set_lower_bound</code></span></a></li><li><a class="tocitem" href="#set_name"><span><code>set_name</code></span></a></li><li><a class="tocitem" href="#set_nonlinear_dual_start_value"><span><code>set_nonlinear_dual_start_value</code></span></a></li><li><a class="tocitem" href="#set_nonlinear_objective"><span><code>set_nonlinear_objective</code></span></a></li><li><a class="tocitem" href="#set_normalized_coefficient"><span><code>set_normalized_coefficient</code></span></a></li><li><a class="tocitem" href="#set_normalized_coefficients"><span><code>set_normalized_coefficients</code></span></a></li><li><a class="tocitem" href="#set_normalized_rhs"><span><code>set_normalized_rhs</code></span></a></li><li><a class="tocitem" href="#set_objective"><span><code>set_objective</code></span></a></li><li><a class="tocitem" href="#set_objective_coefficient"><span><code>set_objective_coefficient</code></span></a></li><li><a class="tocitem" href="#set_objective_function"><span><code>set_objective_function</code></span></a></li><li><a class="tocitem" href="#set_objective_sense"><span><code>set_objective_sense</code></span></a></li><li><a class="tocitem" href="#set_optimize_hook"><span><code>set_optimize_hook</code></span></a></li><li><a class="tocitem" href="#set_optimizer"><span><code>set_optimizer</code></span></a></li><li><a class="tocitem" href="#set_optimizer_attribute"><span><code>set_optimizer_attribute</code></span></a></li><li><a class="tocitem" href="#set_optimizer_attributes"><span><code>set_optimizer_attributes</code></span></a></li><li><a class="tocitem" href="#set_parameter_value"><span><code>set_parameter_value</code></span></a></li><li><a class="tocitem" href="#set_silent"><span><code>set_silent</code></span></a></li><li><a class="tocitem" href="#set_start_value"><span><code>set_start_value</code></span></a></li><li><a class="tocitem" href="#set_start_values"><span><code>set_start_values</code></span></a></li><li><a class="tocitem" href="#set_string_names_on_creation"><span><code>set_string_names_on_creation</code></span></a></li><li><a class="tocitem" href="#set_time_limit_sec"><span><code>set_time_limit_sec</code></span></a></li><li><a class="tocitem" href="#set_upper_bound"><span><code>set_upper_bound</code></span></a></li><li><a class="tocitem" href="#set_value"><span><code>set_value</code></span></a></li><li><a class="tocitem" href="#shadow_price"><span><code>shadow_price</code></span></a></li><li><a class="tocitem" href="#shape"><span><code>shape</code></span></a></li><li><a class="tocitem" href="#show_backend_summary"><span><code>show_backend_summary</code></span></a></li><li><a class="tocitem" href="#show_constraints_summary"><span><code>show_constraints_summary</code></span></a></li><li><a class="tocitem" href="#show_objective_function_summary"><span><code>show_objective_function_summary</code></span></a></li><li><a class="tocitem" href="#simplex_iterations"><span><code>simplex_iterations</code></span></a></li><li><a class="tocitem" href="#solution_summary"><span><code>solution_summary</code></span></a></li><li><a class="tocitem" href="#solve_time"><span><code>solve_time</code></span></a></li><li><a class="tocitem" href="#solver_name"><span><code>solver_name</code></span></a></li><li><a class="tocitem" href="#start_value"><span><code>start_value</code></span></a></li><li><a class="tocitem" href="#termination_status"><span><code>termination_status</code></span></a></li><li><a class="tocitem" href="#time_limit_sec"><span><code>time_limit_sec</code></span></a></li><li><a class="tocitem" href="#triangle_vec"><span><code>triangle_vec</code></span></a></li><li><a class="tocitem" href="#unfix"><span><code>unfix</code></span></a></li><li><a class="tocitem" href="#unregister"><span><code>unregister</code></span></a></li><li><a class="tocitem" href="#unsafe_backend"><span><code>unsafe_backend</code></span></a></li><li><a class="tocitem" href="#unset_binary"><span><code>unset_binary</code></span></a></li><li><a class="tocitem" href="#unset_integer"><span><code>unset_integer</code></span></a></li><li><a class="tocitem" href="#unset_silent"><span><code>unset_silent</code></span></a></li><li><a class="tocitem" href="#unset_time_limit_sec"><span><code>unset_time_limit_sec</code></span></a></li><li><a class="tocitem" href="#upper_bound"><span><code>upper_bound</code></span></a></li><li><a class="tocitem" href="#value"><span><code>value</code></span></a></li><li><a class="tocitem" href="#value_type"><span><code>value_type</code></span></a></li><li><a class="tocitem" href="#variable_by_name"><span><code>variable_by_name</code></span></a></li><li><a class="tocitem" href="#variable_ref_type"><span><code>variable_ref_type</code></span></a></li><li><a class="tocitem" href="#vectorize"><span><code>vectorize</code></span></a></li><li><a class="tocitem" href="#write_to_file"><span><code>write_to_file</code></span></a></li><li><a class="tocitem" href="#AbstractConstraint"><span><code>AbstractConstraint</code></span></a></li><li><a class="tocitem" href="#AbstractJuMPScalar"><span><code>AbstractJuMPScalar</code></span></a></li><li><a class="tocitem" href="#AbstractModel"><span><code>AbstractModel</code></span></a></li><li><a class="tocitem" href="#AbstractScalarSet"><span><code>AbstractScalarSet</code></span></a></li><li><a class="tocitem" href="#AbstractShape"><span><code>AbstractShape</code></span></a></li><li><a class="tocitem" href="#AbstractVariable"><span><code>AbstractVariable</code></span></a></li><li><a class="tocitem" href="#AbstractVariableRef"><span><code>AbstractVariableRef</code></span></a></li><li><a class="tocitem" href="#AbstractVectorSet"><span><code>AbstractVectorSet</code></span></a></li><li><a class="tocitem" href="#AffExpr"><span><code>AffExpr</code></span></a></li><li><a class="tocitem" href="#BinaryRef"><span><code>BinaryRef</code></span></a></li><li><a class="tocitem" href="#BridgeableConstraint"><span><code>BridgeableConstraint</code></span></a></li><li><a class="tocitem" href="#ComplexPlane"><span><code>ComplexPlane</code></span></a></li><li><a class="tocitem" href="#ComplexVariable"><span><code>ComplexVariable</code></span></a></li><li><a class="tocitem" href="#ConstraintNotOwned"><span><code>ConstraintNotOwned</code></span></a></li><li><a class="tocitem" href="#ConstraintRef"><span><code>ConstraintRef</code></span></a></li><li><a class="tocitem" href="#FixRef"><span><code>FixRef</code></span></a></li><li><a class="tocitem" href="#GenericAffExpr"><span><code>GenericAffExpr</code></span></a></li><li><a class="tocitem" href="#GenericModel"><span><code>GenericModel</code></span></a></li><li><a class="tocitem" href="#GenericNonlinearExpr"><span><code>GenericNonlinearExpr</code></span></a></li><li><a class="tocitem" href="#GenericQuadExpr"><span><code>GenericQuadExpr</code></span></a></li><li><a class="tocitem" href="#GenericReferenceMap"><span><code>GenericReferenceMap</code></span></a></li><li><a class="tocitem" href="#GenericVariableRef"><span><code>GenericVariableRef</code></span></a></li><li><a class="tocitem" href="#HermitianMatrixShape"><span><code>HermitianMatrixShape</code></span></a></li><li><a class="tocitem" href="#HermitianMatrixSpace"><span><code>HermitianMatrixSpace</code></span></a></li><li><a class="tocitem" href="#HermitianPSDCone"><span><code>HermitianPSDCone</code></span></a></li><li><a class="tocitem" href="#IntegerRef"><span><code>IntegerRef</code></span></a></li><li><a class="tocitem" href="#LinearTermIterator"><span><code>LinearTermIterator</code></span></a></li><li><a class="tocitem" href="#LowerBoundRef"><span><code>LowerBoundRef</code></span></a></li><li><a class="tocitem" href="#Model"><span><code>Model</code></span></a></li><li><a class="tocitem" href="#ModelMode"><span><code>ModelMode</code></span></a></li><li><a class="tocitem" href="#NLPEvaluator"><span><code>NLPEvaluator</code></span></a></li><li><a class="tocitem" href="#NoOptimizer"><span><code>NoOptimizer</code></span></a></li><li><a class="tocitem" href="#NonlinearConstraintIndex"><span><code>NonlinearConstraintIndex</code></span></a></li><li><a class="tocitem" href="#NonlinearConstraintRef"><span><code>NonlinearConstraintRef</code></span></a></li><li><a class="tocitem" href="#NonlinearExpr"><span><code>NonlinearExpr</code></span></a></li><li><a class="tocitem" href="#NonlinearExpression"><span><code>NonlinearExpression</code></span></a></li><li><a class="tocitem" href="#NonlinearOperator"><span><code>NonlinearOperator</code></span></a></li><li><a class="tocitem" href="#NonlinearParameter"><span><code>NonlinearParameter</code></span></a></li><li><a class="tocitem" href="#Nonnegatives"><span><code>Nonnegatives</code></span></a></li><li><a class="tocitem" href="#Nonpositives"><span><code>Nonpositives</code></span></a></li><li><a class="tocitem" href="#OptimizationSense"><span><code>OptimizationSense</code></span></a></li><li><a class="tocitem" href="#OptimizeNotCalled"><span><code>OptimizeNotCalled</code></span></a></li><li><a class="tocitem" href="#PSDCone"><span><code>PSDCone</code></span></a></li><li><a class="tocitem" href="#Parameter"><span><code>Parameter</code></span></a></li><li><a class="tocitem" href="#ParameterRef"><span><code>ParameterRef</code></span></a></li><li><a class="tocitem" href="#QuadExpr"><span><code>QuadExpr</code></span></a></li><li><a class="tocitem" href="#QuadTermIterator"><span><code>QuadTermIterator</code></span></a></li><li><a class="tocitem" href="#ReferenceMap"><span><code>ReferenceMap</code></span></a></li><li><a class="tocitem" href="#ResultStatusCode"><span><code>ResultStatusCode</code></span></a></li><li><a class="tocitem" href="#RotatedSecondOrderCone"><span><code>RotatedSecondOrderCone</code></span></a></li><li><a class="tocitem" href="#SOS1"><span><code>SOS1</code></span></a></li><li><a class="tocitem" href="#SOS2"><span><code>SOS2</code></span></a></li><li><a class="tocitem" href="#ScalarConstraint"><span><code>ScalarConstraint</code></span></a></li><li><a class="tocitem" href="#ScalarShape"><span><code>ScalarShape</code></span></a></li><li><a class="tocitem" href="#ScalarVariable"><span><code>ScalarVariable</code></span></a></li><li><a class="tocitem" href="#SecondOrderCone"><span><code>SecondOrderCone</code></span></a></li><li><a class="tocitem" href="#Semicontinuous"><span><code>Semicontinuous</code></span></a></li><li><a class="tocitem" href="#Semiinteger"><span><code>Semiinteger</code></span></a></li><li><a class="tocitem" href="#SensitivityReport"><span><code>SensitivityReport</code></span></a></li><li><a class="tocitem" href="#SkewSymmetricMatrixShape"><span><code>SkewSymmetricMatrixShape</code></span></a></li><li><a class="tocitem" href="#SkewSymmetricMatrixSpace"><span><code>SkewSymmetricMatrixSpace</code></span></a></li><li><a class="tocitem" href="#SquareMatrixShape"><span><code>SquareMatrixShape</code></span></a></li><li><a class="tocitem" href="#SymmetricMatrixShape"><span><code>SymmetricMatrixShape</code></span></a></li><li><a class="tocitem" href="#SymmetricMatrixSpace"><span><code>SymmetricMatrixSpace</code></span></a></li><li><a class="tocitem" href="#TerminationStatusCode"><span><code>TerminationStatusCode</code></span></a></li><li><a class="tocitem" href="#UnorderedPair"><span><code>UnorderedPair</code></span></a></li><li><a class="tocitem" href="#UpperBoundRef"><span><code>UpperBoundRef</code></span></a></li><li><a class="tocitem" href="#VariableConstrainedOnCreation"><span><code>VariableConstrainedOnCreation</code></span></a></li><li><a class="tocitem" href="#VariableInfo"><span><code>VariableInfo</code></span></a></li><li><a class="tocitem" href="#VariableNotOwned"><span><code>VariableNotOwned</code></span></a></li><li><a class="tocitem" href="#VariableRef"><span><code>VariableRef</code></span></a></li><li><a class="tocitem" href="#VariablesConstrainedOnCreation"><span><code>VariablesConstrainedOnCreation</code></span></a></li><li><a class="tocitem" href="#VectorConstraint"><span><code>VectorConstraint</code></span></a></li><li><a class="tocitem" href="#VectorShape"><span><code>VectorShape</code></span></a></li><li><a class="tocitem" href="#Zeros"><span><code>Zeros</code></span></a></li><li><a class="tocitem" href="#ALMOST_DUAL_INFEASIBLE"><span><code>ALMOST_DUAL_INFEASIBLE</code></span></a></li><li><a class="tocitem" href="#ALMOST_INFEASIBLE"><span><code>ALMOST_INFEASIBLE</code></span></a></li><li><a class="tocitem" href="#ALMOST_LOCALLY_SOLVED"><span><code>ALMOST_LOCALLY_SOLVED</code></span></a></li><li><a class="tocitem" href="#ALMOST_OPTIMAL"><span><code>ALMOST_OPTIMAL</code></span></a></li><li><a class="tocitem" href="#AUTOMATIC"><span><code>AUTOMATIC</code></span></a></li><li><a class="tocitem" href="#DIRECT"><span><code>DIRECT</code></span></a></li><li><a class="tocitem" href="#DUAL_INFEASIBLE"><span><code>DUAL_INFEASIBLE</code></span></a></li><li><a class="tocitem" href="#FEASIBILITY_SENSE"><span><code>FEASIBILITY_SENSE</code></span></a></li><li><a class="tocitem" href="#FEASIBLE_POINT"><span><code>FEASIBLE_POINT</code></span></a></li><li><a class="tocitem" href="#INFEASIBILITY_CERTIFICATE"><span><code>INFEASIBILITY_CERTIFICATE</code></span></a></li><li><a class="tocitem" href="#INFEASIBLE"><span><code>INFEASIBLE</code></span></a></li><li><a class="tocitem" href="#INFEASIBLE_OR_UNBOUNDED"><span><code>INFEASIBLE_OR_UNBOUNDED</code></span></a></li><li><a class="tocitem" href="#INFEASIBLE_POINT"><span><code>INFEASIBLE_POINT</code></span></a></li><li><a class="tocitem" href="#INTERRUPTED"><span><code>INTERRUPTED</code></span></a></li><li><a class="tocitem" href="#INVALID_MODEL"><span><code>INVALID_MODEL</code></span></a></li><li><a class="tocitem" href="#INVALID_OPTION"><span><code>INVALID_OPTION</code></span></a></li><li><a class="tocitem" href="#ITERATION_LIMIT"><span><code>ITERATION_LIMIT</code></span></a></li><li><a class="tocitem" href="#LOCALLY_INFEASIBLE"><span><code>LOCALLY_INFEASIBLE</code></span></a></li><li><a class="tocitem" href="#LOCALLY_SOLVED"><span><code>LOCALLY_SOLVED</code></span></a></li><li><a class="tocitem" href="#MANUAL"><span><code>MANUAL</code></span></a></li><li><a class="tocitem" href="#MAX_SENSE"><span><code>MAX_SENSE</code></span></a></li><li><a class="tocitem" href="#MEMORY_LIMIT"><span><code>MEMORY_LIMIT</code></span></a></li><li><a class="tocitem" href="#MIN_SENSE"><span><code>MIN_SENSE</code></span></a></li><li><a class="tocitem" href="#NEARLY_FEASIBLE_POINT"><span><code>NEARLY_FEASIBLE_POINT</code></span></a></li><li><a class="tocitem" href="#NEARLY_INFEASIBILITY_CERTIFICATE"><span><code>NEARLY_INFEASIBILITY_CERTIFICATE</code></span></a></li><li><a class="tocitem" href="#NEARLY_REDUCTION_CERTIFICATE"><span><code>NEARLY_REDUCTION_CERTIFICATE</code></span></a></li><li><a class="tocitem" href="#NODE_LIMIT"><span><code>NODE_LIMIT</code></span></a></li><li><a class="tocitem" href="#NORM_LIMIT"><span><code>NORM_LIMIT</code></span></a></li><li><a class="tocitem" href="#NO_SOLUTION"><span><code>NO_SOLUTION</code></span></a></li><li><a class="tocitem" href="#NUMERICAL_ERROR"><span><code>NUMERICAL_ERROR</code></span></a></li><li><a class="tocitem" href="#OBJECTIVE_LIMIT"><span><code>OBJECTIVE_LIMIT</code></span></a></li><li><a class="tocitem" href="#OPTIMAL"><span><code>OPTIMAL</code></span></a></li><li><a class="tocitem" href="#OPTIMIZE_NOT_CALLED"><span><code>OPTIMIZE_NOT_CALLED</code></span></a></li><li><a class="tocitem" href="#OTHER_ERROR"><span><code>OTHER_ERROR</code></span></a></li><li><a class="tocitem" href="#OTHER_LIMIT"><span><code>OTHER_LIMIT</code></span></a></li><li><a class="tocitem" href="#OTHER_RESULT_STATUS"><span><code>OTHER_RESULT_STATUS</code></span></a></li><li><a class="tocitem" href="#REDUCTION_CERTIFICATE"><span><code>REDUCTION_CERTIFICATE</code></span></a></li><li><a class="tocitem" href="#SLOW_PROGRESS"><span><code>SLOW_PROGRESS</code></span></a></li><li><a class="tocitem" href="#SOLUTION_LIMIT"><span><code>SOLUTION_LIMIT</code></span></a></li><li><a class="tocitem" href="#TIME_LIMIT"><span><code>TIME_LIMIT</code></span></a></li><li><a class="tocitem" href="#UNKNOWN_RESULT_STATUS"><span><code>UNKNOWN_RESULT_STATUS</code></span></a></li><li><a class="tocitem" href="#op_and"><span><code>op_and</code></span></a></li><li><a class="tocitem" href="#op_equal_to"><span><code>op_equal_to</code></span></a></li><li><a class="tocitem" href="#op_greater_than_or_equal_to"><span><code>op_greater_than_or_equal_to</code></span></a></li><li><a class="tocitem" href="#op_less_than_or_equal_to"><span><code>op_less_than_or_equal_to</code></span></a></li><li><a class="tocitem" href="#op_or"><span><code>op_or</code></span></a></li><li><a class="tocitem" href="#op_strictly_greater_than"><span><code>op_strictly_greater_than</code></span></a></li><li><a class="tocitem" href="#op_strictly_less_than"><span><code>op_strictly_less_than</code></span></a></li><li><a class="tocitem" href="#Base.empty!(::GenericModel)"><span><code>Base.empty!(::GenericModel)</code></span></a></li><li><a class="tocitem" href="#Base.isempty(::GenericModel)"><span><code>Base.isempty(::GenericModel)</code></span></a></li><li><a class="tocitem" href="#Base.copy(::AbstractModel)"><span><code>Base.copy(::AbstractModel)</code></span></a></li><li><a class="tocitem" href="#Base.write(::IO,-::GenericModel;-::MOI.FileFormats.FileFormat)"><span><code>Base.write(::IO, ::GenericModel; ::MOI.FileFormats.FileFormat)</code></span></a></li><li><a class="tocitem" href="#MOI.Utilities.reset_optimizer(::GenericModel)"><span><code>MOI.Utilities.reset_optimizer(::GenericModel)</code></span></a></li><li><a class="tocitem" href="#MOI.Utilities.drop_optimizer(::GenericModel)"><span><code>MOI.Utilities.drop_optimizer(::GenericModel)</code></span></a></li><li><a class="tocitem" href="#MOI.Utilities.attach_optimizer(::GenericModel)"><span><code>MOI.Utilities.attach_optimizer(::GenericModel)</code></span></a></li></ul></li><li><a class="tocitem" href="../JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>JuMP</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>JuMP</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="DocumenterReference_JuMP"><a class="docs-heading-anchor" href="#DocumenterReference_JuMP">JuMP</a><a id="DocumenterReference_JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#DocumenterReference_JuMP" title="Permalink"></a></h1><p>This page lists the public API of <code>JuMP</code>.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This page is an unstructured list of the JuMP API. For a more structured overview, read the Manual or Tutorial parts of this documentation.</p></div></div><p>Load all of the public the API into the current scope with:</p><pre><code class="language-julia hljs">using JuMP</code></pre><p>Alternatively, load only the module with:</p><pre><code class="language-julia hljs">import JuMP</code></pre><p>and then prefix all calls with <code>JuMP.</code> to create <code>JuMP.&lt;NAME&gt;</code>.</p><h2 id="@NLconstraint"><a class="docs-heading-anchor" href="#@NLconstraint"><code>@NLconstraint</code></a><a id="@NLconstraint-1"></a><a class="docs-heading-anchor-permalink" href="#@NLconstraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@NLconstraint" href="#JuMP.@NLconstraint"><code>JuMP.@NLconstraint</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@NLconstraint(model::GenericModel, expr)</code></pre><p>Add a constraint described by the nonlinear expression <code>expr</code>. See also <a href="#JuMP.@constraint"><code>@constraint</code></a>. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @NLconstraint(model, sin(x) &lt;= 1)
+sin(x) - 1.0 ≤ 0
+
+julia&gt; @NLconstraint(model, [i = 1:3], sin(i * x) &lt;= 1 / i)
+3-element Vector{NonlinearConstraintRef{ScalarShape}}:
+ (sin(1.0 * x) - 1.0 / 1.0) - 0.0 ≤ 0
+ (sin(2.0 * x) - 1.0 / 2.0) - 0.0 ≤ 0
+ (sin(3.0 * x) - 1.0 / 3.0) - 0.0 ≤ 0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L3092-L3113">source</a></section></article><h2 id="@NLconstraints"><a class="docs-heading-anchor" href="#@NLconstraints"><code>@NLconstraints</code></a><a id="@NLconstraints-1"></a><a class="docs-heading-anchor-permalink" href="#@NLconstraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@NLconstraints" href="#JuMP.@NLconstraints"><code>JuMP.@NLconstraints</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@NLconstraints(model, args...)</code></pre><p>Adds multiple nonlinear constraints to model at once, in the same fashion as the <a href="#JuMP.@NLconstraint"><code>@NLconstraint</code></a> macro.</p><p>The model must be the first argument, and multiple constraints can be added on multiple lines wrapped in a <code>begin ... end</code> block.</p><p>The macro returns a tuple containing the constraints that were defined.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @variable(model, t);
+
+julia&gt; @variable(model, z[1:2]);
+
+julia&gt; a = [4, 5];
+
+julia&gt; @NLconstraints(model, begin
+           t &gt;= sqrt(x^2 + y^2)
+           [i = 1:2], z[i] &lt;= log(a[i])
+       end)
+((t - sqrt(x ^ 2.0 + y ^ 2.0)) - 0.0 ≥ 0, NonlinearConstraintRef{ScalarShape}[(z[1] - log(4.0)) - 0.0 ≤ 0, (z[2] - log(5.0)) - 0.0 ≤ 0])</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1692-L1724">source</a></section></article><h2 id="@NLexpression"><a class="docs-heading-anchor" href="#@NLexpression"><code>@NLexpression</code></a><a id="@NLexpression-1"></a><a class="docs-heading-anchor-permalink" href="#@NLexpression" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@NLexpression" href="#JuMP.@NLexpression"><code>JuMP.@NLexpression</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@NLexpression(args...)</code></pre><p>Efficiently build a nonlinear expression which can then be inserted in other nonlinear constraints and the objective. See also [<code>@expression</code>]. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; @NLexpression(model, my_expr, sin(x)^2 + cos(x^2))
+subexpression[1]: sin(x) ^ 2.0 + cos(x ^ 2.0)
+
+julia&gt; @NLconstraint(model, my_expr + y &gt;= 5)
+(subexpression[1] + y) - 5.0 ≥ 0
+
+julia&gt; @NLobjective(model, Min, my_expr)</code></pre><p>Indexing over sets and anonymous expressions are also supported:</p><pre><code class="language-julia-repl hljs">julia&gt; @NLexpression(model, my_expr_1[i=1:3], sin(i * x))
+3-element Vector{NonlinearExpression}:
+ subexpression[2]: sin(1.0 * x)
+ subexpression[3]: sin(2.0 * x)
+ subexpression[4]: sin(3.0 * x)
+
+julia&gt; my_expr_2 = @NLexpression(model, log(1 + sum(exp(my_expr_1[i]) for i in 1:2)))
+subexpression[5]: log(1.0 + (exp(subexpression[2]) + exp(subexpression[3])))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L3165-L3200">source</a></section></article><h2 id="@NLexpressions"><a class="docs-heading-anchor" href="#@NLexpressions"><code>@NLexpressions</code></a><a id="@NLexpressions-1"></a><a class="docs-heading-anchor-permalink" href="#@NLexpressions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@NLexpressions" href="#JuMP.@NLexpressions"><code>JuMP.@NLexpressions</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@NLexpressions(model, args...)</code></pre><p>Adds multiple nonlinear expressions to model at once, in the same fashion as the <a href="#JuMP.@NLexpression"><code>@NLexpression</code></a> macro.</p><p>The model must be the first argument, and multiple expressions can be added on multiple lines wrapped in a <code>begin ... end</code> block.</p><p>The macro returns a tuple containing the expressions that were defined.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @variable(model, z[1:2]);
+
+julia&gt; a = [4, 5];
+
+julia&gt; @NLexpressions(model, begin
+           my_expr, sqrt(x^2 + y^2)
+           my_expr_1[i = 1:2], log(a[i]) - z[i]
+       end)
+(subexpression[1]: sqrt(x ^ 2.0 + y ^ 2.0), NonlinearExpression[subexpression[2]: log(4.0) - z[1], subexpression[3]: log(5.0) - z[2]])</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1726-L1756">source</a></section></article><h2 id="@NLobjective"><a class="docs-heading-anchor" href="#@NLobjective"><code>@NLobjective</code></a><a id="@NLobjective-1"></a><a class="docs-heading-anchor-permalink" href="#@NLobjective" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@NLobjective" href="#JuMP.@NLobjective"><code>JuMP.@NLobjective</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@NLobjective(model, sense, expression)</code></pre><p>Add a nonlinear objective to <code>model</code> with optimization sense <code>sense</code>. <code>sense</code> must be <code>Max</code> or <code>Min</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @NLobjective(model, Max, 2x + 1 + sin(x))
+
+julia&gt; print(model)
+Max 2.0 * x + 1.0 + sin(x)
+Subject to</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L3057-L3077">source</a></section></article><h2 id="@NLparameter"><a class="docs-heading-anchor" href="#@NLparameter"><code>@NLparameter</code></a><a id="@NLparameter-1"></a><a class="docs-heading-anchor-permalink" href="#@NLparameter" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@NLparameter" href="#JuMP.@NLparameter"><code>JuMP.@NLparameter</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@NLparameter(model, param == value)</code></pre><p>Create and return a nonlinear parameter <code>param</code> attached to the model <code>model</code> with initial value set to <code>value</code>. Nonlinear parameters may be used only in nonlinear expressions.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @NLparameter(model, x == 10)
+x == 10.0
+
+julia&gt; value(x)
+10.0</code></pre><pre><code class="nohighlight hljs">@NLparameter(model, value = param_value)</code></pre><p>Create and return an anonymous nonlinear parameter <code>param</code> attached to the model <code>model</code> with initial value set to <code>param_value</code>. Nonlinear parameters may be used only in nonlinear expressions.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x = @NLparameter(model, value = 10)
+parameter[1] == 10.0
+
+julia&gt; value(x)
+10.0</code></pre><pre><code class="nohighlight hljs">@NLparameter(model, param_collection[...] == value_expr)</code></pre><p>Create and return a collection of nonlinear parameters <code>param_collection</code> attached to the model <code>model</code> with initial value set to <code>value_expr</code> (may depend on index sets). Uses the same syntax for specifying index sets as <a href="#JuMP.@variable"><code>@variable</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @NLparameter(model, y[i = 1:3] == 2 * i)
+3-element Vector{NonlinearParameter}:
+ parameter[1] == 2.0
+ parameter[2] == 4.0
+ parameter[3] == 6.0
+
+julia&gt; value(y[2])
+4.0</code></pre><pre><code class="nohighlight hljs">@NLparameter(model, [...] == value_expr)</code></pre><p>Create and return an anonymous collection of nonlinear parameters attached to the model <code>model</code> with initial value set to <code>value_expr</code> (may depend on index sets). Uses the same syntax for specifying index sets as <a href="#JuMP.@variable"><code>@variable</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; y = @NLparameter(model, [i = 1:3] == 2 * i)
+3-element Vector{NonlinearParameter}:
+ parameter[1] == 2.0
+ parameter[2] == 4.0
+ parameter[3] == 6.0
+
+julia&gt; value(y[2])
+4.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L3248-L3327">source</a></section></article><h2 id="@NLparameters"><a class="docs-heading-anchor" href="#@NLparameters"><code>@NLparameters</code></a><a id="@NLparameters-1"></a><a class="docs-heading-anchor-permalink" href="#@NLparameters" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@NLparameters" href="#JuMP.@NLparameters"><code>JuMP.@NLparameters</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs"> @NLparameters(model, args...)</code></pre><p>Create and return multiple nonlinear parameters attached to model <code>model</code>, in the same fashion as <a href="#JuMP.@NLparameter"><code>@NLparameter</code></a> macro.</p><p>The model must be the first argument, and multiple parameters can be added on multiple lines wrapped in a <code>begin ... end</code> block. Distinct parameters need to be placed on separate lines as in the following example.</p><p>The macro returns a tuple containing the parameters that were defined.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @NLparameters(model, begin
+           x == 10
+           b == 156
+       end);
+
+julia&gt; value(x)
+10.0</code></pre><pre><code class="nohighlight hljs"></code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1665-L1690">source</a></section></article><h2 id="@build_constraint"><a class="docs-heading-anchor" href="#@build_constraint"><code>@build_constraint</code></a><a id="@build_constraint-1"></a><a class="docs-heading-anchor-permalink" href="#@build_constraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@build_constraint" href="#JuMP.@build_constraint"><code>JuMP.@build_constraint</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@build_constraint(constraint_expr)</code></pre><p>Constructs a <code>ScalarConstraint</code> or <code>VectorConstraint</code> using the same machinery as <a href="#JuMP.@constraint"><code>@constraint</code></a> but without adding the constraint to a model.</p><p>Constraints using broadcast operators like <code>x .&lt;= 1</code> are also supported and will create arrays of <code>ScalarConstraint</code> or <code>VectorConstraint</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @build_constraint(2x &gt;= 1)
+ScalarConstraint{AffExpr, MathOptInterface.GreaterThan{Float64}}(2 x, MathOptInterface.GreaterThan{Float64}(1.0))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1443-L1462">source</a></section></article><h2 id="@constraint"><a class="docs-heading-anchor" href="#@constraint"><code>@constraint</code></a><a id="@constraint-1"></a><a class="docs-heading-anchor-permalink" href="#@constraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@constraint" href="#JuMP.@constraint"><code>JuMP.@constraint</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@constraint(m::GenericModel, expr, kw_args...)</code></pre><p>Add a constraint described by the expression <code>expr</code>.</p><pre><code class="nohighlight hljs">@constraint(m::GenericModel, ref[i=..., j=..., ...], expr, kw_args...)</code></pre><p>Add a group of constraints described by the expression <code>expr</code> parametrized by <code>i</code>, <code>j</code>, ...</p><p>The expression <code>expr</code> can either be</p><ul><li>of the form <code>func in set</code> constraining the function <code>func</code> to belong to the set <code>set</code> which is either a <a href="../../moi/reference/standard_form/#MathOptInterface.AbstractSet"><code>MOI.AbstractSet</code></a> or one of the JuMP shortcuts <a href="#JuMP.SecondOrderCone"><code>SecondOrderCone</code></a>, <a href="#JuMP.RotatedSecondOrderCone"><code>RotatedSecondOrderCone</code></a> and <a href="#JuMP.PSDCone"><code>PSDCone</code></a>, e.g. <code>@constraint(model, [1, x-1, y-2] in SecondOrderCone())</code> constrains the norm of <code>[x-1, y-2]</code> be less than 1;</li><li>of the form <code>a sign b</code>, where <code>sign</code> is one of <code>==</code>, <code>≥</code>, <code>&gt;=</code>, <code>≤</code> and <code>&lt;=</code> building the single constraint enforcing the comparison to hold for the expression <code>a</code> and <code>b</code>, e.g. <code>@constraint(m, x^2 + y^2 == 1)</code> constrains <code>x</code> and <code>y</code> to lie on the unit circle;</li><li>of the form <code>a ≤ b ≤ c</code> or <code>a ≥ b ≥ c</code> (where <code>≤</code> and <code>&lt;=</code> (resp. <code>≥</code> and <code>&gt;=</code>) can be used interchangeably) constraining the paired the expression <code>b</code> to lie between <code>a</code> and <code>c</code>;</li><li>of the forms <code>@constraint(m, a .sign b)</code> or <code>@constraint(m, a .sign b .sign c)</code> which broadcast the constraint creation to each element of the vectors.</li></ul><p>The recognized keyword arguments in <code>kw_args</code> are the following:</p><ul><li><code>base_name</code>: Sets the name prefix used to generate constraint names. It corresponds to the constraint name for scalar constraints, otherwise, the constraint names are set to <code>base_name[...]</code> for each index <code>...</code> of the axes <code>axes</code>.</li><li><code>container</code>: Specify the container type.</li><li><code>set_string_name::Bool = true</code>: control whether to set the <a href="../../moi/reference/constraints/#MathOptInterface.ConstraintName"><code>MOI.ConstraintName</code></a> attribute. Passing <code>set_string_name = false</code> can improve performance.</li></ul><p><strong>Note for extending the constraint macro</strong></p><p>Each constraint will be created using <code>add_constraint(m, build_constraint(_error, func, set))</code> where</p><ul><li><code>_error</code> is an error function showing the constraint call in addition to the error message given as argument,</li><li><code>func</code> is the expression that is constrained</li><li>and <code>set</code> is the set in which it is constrained to belong.</li></ul><p>For <code>expr</code> of the first type (i.e. <code>@constraint(m, func in set)</code>), <code>func</code> and <code>set</code> are passed unchanged to <code>build_constraint</code> but for the other types, they are determined from the expressions and signs. For instance, <code>@constraint(m, x^2 + y^2 == 1)</code> is transformed into <code>add_constraint(m, build_constraint(_error, x^2 + y^2, MOI.EqualTo(1.0)))</code>.</p><p>To extend JuMP to accept new constraints of this form, it is necessary to add the corresponding methods to <code>build_constraint</code>. Note that this will likely mean that either <code>func</code> or <code>set</code> will be some custom type, rather than e.g. a <code>Symbol</code>, since we will likely want to dispatch on the type of the function or set appearing in the constraint.</p><p>For extensions that need to create constraints with more information than just <code>func</code> and <code>set</code>, an additional positional argument can be specified to <code>@constraint</code> that will then be passed on <code>build_constraint</code>. Hence, we can enable this syntax by defining extensions of <code>build_constraint(_error, func, set, my_arg; kw_args...)</code>. This produces the user syntax: <code>@constraint(model, ref[...], expr, my_arg, kw_args...)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1371-L1438">source</a></section></article><h2 id="@constraints"><a class="docs-heading-anchor" href="#@constraints"><code>@constraints</code></a><a id="@constraints-1"></a><a class="docs-heading-anchor-permalink" href="#@constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@constraints" href="#JuMP.@constraints"><code>JuMP.@constraints</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@constraints(model, args...)</code></pre><p>Adds groups of constraints at once, in the same fashion as the <a href="#JuMP.@constraint"><code>@constraint</code></a> macro.</p><p>The model must be the first argument, and multiple constraints can be added on multiple lines wrapped in a <code>begin ... end</code> block.</p><p>The macro returns a tuple containing the constraints that were defined.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, w);
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @variable(model, z[1:3]);
+
+julia&gt; @constraints(model, begin
+           x &gt;= 1
+           y - w &lt;= 2
+           sum_to_one[i=1:3], z[i] + y == 1
+       end);
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ sum_to_one[1] : y + z[1] = 1
+ sum_to_one[2] : y + z[2] = 1
+ sum_to_one[3] : y + z[3] = 1
+ x ≥ 1
+ -w + y ≤ 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1563-L1602">source</a></section></article><h2 id="@expression"><a class="docs-heading-anchor" href="#@expression"><code>@expression</code></a><a id="@expression-1"></a><a class="docs-heading-anchor-permalink" href="#@expression" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@expression" href="#JuMP.@expression"><code>JuMP.@expression</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@expression(args...)</code></pre><p>Efficiently builds a linear or quadratic expression but does not add to model immediately. Instead, returns the expression which can then be inserted in other constraints.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:5]);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @variable(model, z);
+
+julia&gt; @expression(model, shared, sum(i * x[i] for i in 1:5))
+x[1] + 2 x[2] + 3 x[3] + 4 x[4] + 5 x[5]
+
+julia&gt; @constraint(model, shared + y &gt;= 5)
+x[1] + 2 x[2] + 3 x[3] + 4 x[4] + 5 x[5] + y ≥ 5
+
+julia&gt; @constraint(model, shared + z &lt;= 10)
+x[1] + 2 x[2] + 3 x[3] + 4 x[4] + 5 x[5] + z ≤ 10</code></pre><p>The <code>ref</code> accepts index sets in the same way as <code>@variable</code>, and those indices can be used in the construction of the expressions:</p><pre><code class="language-julia-repl hljs">julia&gt; @expression(model, expr[i = 1:3], i * sum(x[j] for j in 1:3))
+3-element Vector{AffExpr}:
+ x[1] + x[2] + x[3]
+ 2 x[1] + 2 x[2] + 2 x[3]
+ 3 x[1] + 3 x[2] + 3 x[3]</code></pre><p>Anonymous syntax is also supported:</p><pre><code class="language-julia-repl hljs">julia&gt; expr = @expression(model, [i in 1:3], i * sum(x[j] for j in 1:3))
+3-element Vector{AffExpr}:
+ x[1] + x[2] + x[3]
+ 2 x[1] + 2 x[2] + 2 x[3]
+ 3 x[1] + 3 x[2] + 3 x[3]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1870-L1918">source</a></section></article><h2 id="@expressions"><a class="docs-heading-anchor" href="#@expressions"><code>@expressions</code></a><a id="@expressions-1"></a><a class="docs-heading-anchor-permalink" href="#@expressions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@expressions" href="#JuMP.@expressions"><code>JuMP.@expressions</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@expressions(model, args...)</code></pre><p>Adds multiple expressions to model at once, in the same fashion as the <a href="#JuMP.@expression"><code>@expression</code></a> macro.</p><p>The model must be the first argument, and multiple expressions can be added on multiple lines wrapped in a <code>begin ... end</code> block.</p><p>The macro returns a tuple containing the expressions that were defined.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @variable(model, z[1:2]);
+
+julia&gt; a = [4, 5];
+
+julia&gt; @expressions(model, begin
+           my_expr, x^2 + y^2
+           my_expr_1[i = 1:2], a[i] - z[i]
+       end)
+(x² + y², AffExpr[-z[1] + 4, -z[2] + 5])</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1633-L1663">source</a></section></article><h2 id="@objective"><a class="docs-heading-anchor" href="#@objective"><code>@objective</code></a><a id="@objective-1"></a><a class="docs-heading-anchor-permalink" href="#@objective" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@objective" href="#JuMP.@objective"><code>JuMP.@objective</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@objective(model::GenericModel, sense, func)</code></pre><p>Set the objective sense to <code>sense</code> and objective function to <code>func</code>. The objective sense can be either <code>Min</code>, <code>Max</code>, <code>MOI.MIN_SENSE</code>, <code>MOI.MAX_SENSE</code> or <code>MOI.FEASIBILITY_SENSE</code>; see <a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense</code></a>.</p><p>In order to set the sense programmatically, i.e., when <code>sense</code> is a Julia variable whose value is the sense, one of the three <code>MOI.ObjectiveSense</code> values should be used.</p><p><strong>Example</strong></p><p>To minimize the value of the variable <code>x</code>, do as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @objective(model, Min, x)
+x</code></pre><p>To maximize the value of the affine expression <code>2x - 1</code>, do as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; @objective(model, Max, 2x - 1)
+2 x - 1</code></pre><p>To set a quadratic objective and set the objective sense programmatically, do as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; sense = MIN_SENSE
+MIN_SENSE::OptimizationSense = 0
+
+julia&gt; @objective(model, sense, x^2 - 2x + 1)
+x² - 2 x + 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1803-L1842">source</a></section></article><h2 id="@operator"><a class="docs-heading-anchor" href="#@operator"><code>@operator</code></a><a id="@operator-1"></a><a class="docs-heading-anchor-permalink" href="#@operator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@operator" href="#JuMP.@operator"><code>JuMP.@operator</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@operator(model, operator, dim, f[, ∇f[, ∇²f]])</code></pre><p>Add the nonlinear operator <code>operator</code> in <code>model</code> with <code>dim</code> arguments, and create a new <a href="#JuMP.NonlinearOperator"><code>NonlinearOperator</code></a> object called <code>operator</code> in the current scope.</p><p>The function <code>f</code> evaluates the operator and must return a scalar.</p><p>The optional function <code>∇f</code> evaluates the first derivative, and the optional function <code>∇²f</code> evaluates the second derivative.</p><p><code>∇²f</code> may be provided only if <code>∇f</code> is also provided.</p><p><strong>Univariate syntax</strong></p><p>If <code>dim == 1</code>, then the method signatures of each function must be:</p><ul><li><code>f(::T)::T where {T&lt;:Real}</code></li><li><code>∇f(::T)::T where {T&lt;:Real}</code></li><li><code>∇²f(::T)::T where {T&lt;:Real}</code></li></ul><p><strong>Multivariate syntax</strong></p><p>If <code>dim &gt; 1</code>, then the method signatures of each function must be:</p><ul><li><code>f(x::T...)::T where {T&lt;:Real}</code></li><li><code>∇f(g::AbstractVector{T}, x::T...)::Nothing where {T&lt;:Real}</code></li><li><code>∇²f(H::AbstractMatrix{T}, x::T...)::Nothing where {T&lt;:Real}</code></li></ul><p>Where the gradient vector <code>g</code> and Hessian matrix <code>H</code> are filled in-place. For the Hessian, you must fill in the non-zero lower-triangular entries only. Setting an off-diagonal upper-triangular element may error.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; f(x::Float64) = x^2
+f (generic function with 1 method)
+
+julia&gt; ∇f(x::Float64) = 2 * x
+∇f (generic function with 1 method)
+
+julia&gt; ∇²f(x::Float64) = 2.0
+∇²f (generic function with 1 method)
+
+julia&gt; @operator(model, op_f, 1, f, ∇f, ∇²f)
+NonlinearOperator(f, :op_f)
+
+julia&gt; @objective(model, Min, op_f(x))
+op_f(x)
+
+julia&gt; op_f(2.0)
+4.0
+
+julia&gt; model[:op_f]
+NonlinearOperator(f, :op_f)
+
+julia&gt; model[:op_f](x)
+op_f(x)</code></pre><p><strong>Non-macro version</strong></p><p>This macro is provided as helpful syntax that matches the style of the rest of the JuMP macros. However, you may also add operators outside the macro using <a href="#JuMP.add_nonlinear_operator"><code>add_nonlinear_operator</code></a>. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; f(x) = x^2
+f (generic function with 1 method)
+
+julia&gt; @operator(model, op_f, 1, f)
+NonlinearOperator(f, :op_f)</code></pre><p>is equivalent to</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; f(x) = x^2
+f (generic function with 1 method)
+
+julia&gt; op_f = model[:op_f] = add_nonlinear_operator(model, 1, f; name = :op_f)
+NonlinearOperator(f, :op_f)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp_expr.jl#L987-L1079">source</a></section></article><h2 id="@variable"><a class="docs-heading-anchor" href="#@variable"><code>@variable</code></a><a id="@variable-1"></a><a class="docs-heading-anchor-permalink" href="#@variable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@variable" href="#JuMP.@variable"><code>JuMP.@variable</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@variable(model, expr, args..., kw_args...)</code></pre><p>Add a variable to the model <code>model</code> described by the expression <code>expr</code>, the positional arguments <code>args</code> and the keyword arguments <code>kw_args</code>.</p><p><strong>Anonymous and named variables</strong></p><p><code>expr</code> must be one of the forms:</p><ul><li>Omitted, like <code>@variable(model)</code>, which creates an anonymous variable</li><li>A single symbol like <code>@variable(model, x)</code></li><li>A container expression like <code>@variable(model, x[i=1:3])</code></li><li>An anoymous container expression like <code>@variable(model, [i=1:3])</code></li></ul><p><strong>Bounds</strong></p><p>In addition, the expression can have bounds, such as:</p><ul><li><code>@variable(model, x &gt;= 0)</code></li><li><code>@variable(model, x &lt;= 0)</code></li><li><code>@variable(model, x == 0)</code></li><li><code>@variable(model, 0 &lt;= x &lt;= 1)</code></li></ul><p>and bounds can depend on the indices of the container expressions:</p><ul><li><code>@variable(model, -i &lt;= x[i=1:3] &lt;= i)</code></li></ul><p><strong>Sets</strong></p><p>You can explicitly specify the set to which the variable belongs:</p><ul><li><code>@variable(model, x in MOI.Interval(0.0, 1.0))</code></li></ul><p>For more information on this syntax, read <a href="../../moi/tutorials/implementing/#Variables-constrained-on-creation">Variables constrained on creation</a>.</p><p><strong>Positional arguments</strong></p><p>The recognized positional arguments in <code>args</code> are the following:</p><ul><li><code>Bin</code>: restricts the variable to the <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a> set, that is, <code>{0, 1}</code>. For example, <code>@variable(model, x, Bin)</code>. Note: you cannot use <code>@variable(model, Bin)</code>, use the <code>binary</code> keyword instead.</li><li><code>Int</code>: restricts the variable to the set of integers, that is, ..., -2, -1,  0, 1, 2, ... For example, <code>@variable(model, x, Int)</code>. Note: you cannot use  <code>@variable(model, Int)</code>, use the <code>integer</code> keyword instead.</li><li><code>Symmetric</code>: Only available when creating a square matrix of variables, i.e., when <code>expr</code> is of the form <code>varname[1:n,1:n]</code> or <code>varname[i=1:n,j=1:n]</code>, it creates a symmetric matrix of variables.</li><li><code>PSD</code>: A restrictive extension to <code>Symmetric</code> which constraints a square matrix of variables to <code>Symmetric</code> and constrains to be positive semidefinite.</li></ul><p><strong>Keyword arguments</strong></p><p>Four keyword arguments are useful in all cases:</p><ul><li><code>base_name</code>: Sets the name prefix used to generate variable names. It corresponds to the variable name for scalar variable, otherwise, the variable names are set to <code>base_name[...]</code> for each index <code>...</code> of the axes <code>axes</code>.</li><li><code>start::Float64</code>: specify the value passed to <code>set_start_value</code> for each variable</li><li><code>container</code>: specify the container type. See <a href="../../manual/variables/#variable_forcing">Forcing the container type</a> for more information.</li><li><code>set_string_name::Bool = true</code>: control whether to set the <a href="../../moi/reference/variables/#MathOptInterface.VariableName"><code>MOI.VariableName</code></a> attribute. Passing <code>set_string_name = false</code> can improve performance.</li></ul><p>Other keyword arguments are needed to disambiguate sitations with anonymous variables:</p><ul><li><code>lower_bound::Float64</code>: an alternative to <code>x &gt;= lb</code>, sets the value of the variable lower bound.</li><li><code>upper_bound::Float64</code>: an alternative to <code>x &lt;= ub</code>, sets the value of the variable upper bound.</li><li><code>binary::Bool</code>: an alternative to passing <code>Bin</code>, sets whether the variable is binary or not.</li><li><code>integer::Bool</code>: an alternative to passing <code>Int</code>, sets whether the variable is integer or not.</li><li><code>set::MOI.AbstractSet</code>: an alternative to using <code>x in set</code></li><li><code>variable_type</code>: used by JuMP extensions. See <a href="../../developers/extensions/#extend_variable_macro">Extend <code>@variable</code></a> for more information.</li></ul><p><strong>Example</strong></p><p>The following are equivalent ways of creating a variable <code>x</code> of name <code>x</code> with lower bound 0:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 0)
+x</code></pre><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, lower_bound = 0)
+x</code></pre><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x = @variable(model, base_name = &quot;x&quot;, lower_bound = 0)
+x</code></pre><p>Other examples:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:3] &lt;= i, Int, start = sqrt(i), lower_bound = -i)
+3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+
+julia&gt; @variable(model, y[i=1:3], container = DenseAxisArray, set = MOI.ZeroOne())
+1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, Base.OneTo(3)
+And data, a 3-element Vector{VariableRef}:
+ y[1]
+ y[2]
+ y[3]
+
+julia&gt; @variable(model, z[i=1:3], set_string_name = false)
+3-element Vector{VariableRef}:
+ _[7]
+ _[8]
+ _[9]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L2706-L2841">source</a></section></article><h2 id="@variables"><a class="docs-heading-anchor" href="#@variables"><code>@variables</code></a><a id="@variables-1"></a><a class="docs-heading-anchor-permalink" href="#@variables" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.@variables" href="#JuMP.@variables"><code>JuMP.@variables</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@variables(model, args...)</code></pre><p>Adds multiple variables to model at once, in the same fashion as the <a href="#JuMP.@variable"><code>@variable</code></a> macro.</p><p>The model must be the first argument, and multiple variables can be added on multiple lines wrapped in a <code>begin ... end</code> block.</p><p>The macro returns a tuple containing the variables that were defined.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variables(model, begin
+           x
+           y[i = 1:2] &gt;= 0, (start = i)
+           z, Bin, (start = 0, base_name = &quot;Z&quot;)
+       end)
+(x, VariableRef[y[1], y[2]], Z)</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Keyword arguments must be contained within parentheses (refer to the example above).</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1604-L1631">source</a></section></article><h2 id="add_bridge"><a class="docs-heading-anchor" href="#add_bridge"><code>add_bridge</code></a><a id="add_bridge-1"></a><a class="docs-heading-anchor-permalink" href="#add_bridge" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_bridge" href="#JuMP.add_bridge"><code>JuMP.add_bridge</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_bridge(
+    model::GenericModel{T},
+    BT::Type{&lt;:MOI.Bridges.AbstractBridge};
+    coefficient_type::Type{S} = T,
+) where {T,S}</code></pre><p>Add <code>BT{T}</code> to the list of bridges that can be used to transform unsupported constraints into an equivalent formulation using only constraints supported by the optimizer.</p><p>See also: <a href="#JuMP.remove_bridge"><code>remove_bridge</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; add_bridge(model, MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge)
+
+julia&gt; add_bridge(
+           model,
+           MOI.Bridges.Constraint.NumberConversionBridge;
+           coefficient_type = Complex{Float64}
+       )</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L611-L637">source</a></section></article><h2 id="add_constraint"><a class="docs-heading-anchor" href="#add_constraint"><code>add_constraint</code></a><a id="add_constraint-1"></a><a class="docs-heading-anchor-permalink" href="#add_constraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_constraint" href="#JuMP.add_constraint"><code>JuMP.add_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_constraint(model::GenericModel, con::AbstractConstraint, name::String=&quot;&quot;)</code></pre><p>Add a constraint <code>con</code> to <code>Model model</code> and sets its name.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L691-L695">source</a></section></article><h2 id="add_nonlinear_constraint"><a class="docs-heading-anchor" href="#add_nonlinear_constraint"><code>add_nonlinear_constraint</code></a><a id="add_nonlinear_constraint-1"></a><a class="docs-heading-anchor-permalink" href="#add_nonlinear_constraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_nonlinear_constraint" href="#JuMP.add_nonlinear_constraint"><code>JuMP.add_nonlinear_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_nonlinear_constraint(model::Model, expr::Expr)</code></pre><p>Add a nonlinear constraint described by the Julia expression <code>ex</code> to <code>model</code>.</p><p>This function is most useful if the expression <code>ex</code> is generated programmatically, and you cannot use <a href="#JuMP.@NLconstraint"><code>@NLconstraint</code></a>.</p><p><strong>Notes</strong></p><ul><li>You must interpolate the variables directly into the expression <code>expr</code>.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; add_nonlinear_constraint(model, :($(x) + $(x)^2 &lt;= 1))
+(x + x ^ 2.0) - 1.0 ≤ 0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L443-L465">source</a></section></article><h2 id="add_nonlinear_expression"><a class="docs-heading-anchor" href="#add_nonlinear_expression"><code>add_nonlinear_expression</code></a><a id="add_nonlinear_expression-1"></a><a class="docs-heading-anchor-permalink" href="#add_nonlinear_expression" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_nonlinear_expression" href="#JuMP.add_nonlinear_expression"><code>JuMP.add_nonlinear_expression</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_nonlinear_expression(model::Model, expr::Expr)</code></pre><p>Add a nonlinear expression <code>expr</code> to <code>model</code>.</p><p>This function is most useful if the expression <code>expr</code> is generated programmatically, and you cannot use <a href="#JuMP.@NLexpression"><code>@NLexpression</code></a>.</p><p><strong>Notes</strong></p><ul><li>You must interpolate the variables directly into the expression <code>expr</code>.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; add_nonlinear_expression(model, :($(x) + $(x)^2))
+subexpression[1]: x + x ^ 2.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L309-L331">source</a></section></article><h2 id="add_nonlinear_operator"><a class="docs-heading-anchor" href="#add_nonlinear_operator"><code>add_nonlinear_operator</code></a><a id="add_nonlinear_operator-1"></a><a class="docs-heading-anchor-permalink" href="#add_nonlinear_operator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_nonlinear_operator" href="#JuMP.add_nonlinear_operator"><code>JuMP.add_nonlinear_operator</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_nonlinear_operator(
+    model::Model,
+    dim::Int,
+    f::Function,
+    [∇f::Function,]
+    [∇²f::Function];
+    [name::Symbol = Symbol(f),]
+)</code></pre><p>Add a new nonlinear operator with <code>dim</code> input arguments to <code>model</code> and associate it with the name <code>name</code>.</p><p>The function <code>f</code> evaluates the operator and must return a scalar.</p><p>The optional function <code>∇f</code> evaluates the first derivative, and the optional function <code>∇²f</code> evaluates the second derivative.</p><p><code>∇²f</code> may be provided only if <code>∇f</code> is also provided.</p><p><strong>Univariate syntax</strong></p><p>If <code>dim == 1</code>, then the method signatures of each function must be:</p><ul><li><code>f(::T)::T where {T&lt;:Real}</code></li><li><code>∇f(::T)::T where {T&lt;:Real}</code></li><li><code>∇²f(::T)::T where {T&lt;:Real}</code></li></ul><p><strong>Multivariate syntax</strong></p><p>If <code>dim &gt; 1</code>, then the method signatures of each function must be:</p><ul><li><code>f(x::T...)::T where {T&lt;:Real}</code></li><li><code>∇f(g::AbstractVector{T}, x::T...)::Nothing where {T&lt;:Real}</code></li><li><code>∇²f(H::AbstractMatrix{T}, x::T...)::Nothing where {T&lt;:Real}</code></li></ul><p>Where the gradient vector <code>g</code> and Hessian matrix <code>H</code> are filled in-place. For the Hessian, you must fill in the non-zero lower-triangular entries only. Setting an off-diagonal upper-triangular element may error.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; f(x::Float64) = x^2
+f (generic function with 1 method)
+
+julia&gt; ∇f(x::Float64) = 2 * x
+∇f (generic function with 1 method)
+
+julia&gt; ∇²f(x::Float64) = 2.0
+∇²f (generic function with 1 method)
+
+julia&gt; op_f = add_nonlinear_operator(model, 1, f, ∇f, ∇²f)
+NonlinearOperator(f, :f)
+
+julia&gt; @objective(model, Min, op_f(x))
+f(x)
+
+julia&gt; op_f(2.0)
+4.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp_expr.jl#L865-L931">source</a></section></article><h2 id="add_nonlinear_parameter"><a class="docs-heading-anchor" href="#add_nonlinear_parameter"><code>add_nonlinear_parameter</code></a><a id="add_nonlinear_parameter-1"></a><a class="docs-heading-anchor-permalink" href="#add_nonlinear_parameter" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_nonlinear_parameter" href="#JuMP.add_nonlinear_parameter"><code>JuMP.add_nonlinear_parameter</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_nonlinear_parameter(model::Model, value::Real)</code></pre><p>Add an anonymous parameter to the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L210-L214">source</a></section></article><h2 id="add_to_expression!"><a class="docs-heading-anchor" href="#add_to_expression!"><code>add_to_expression!</code></a><a id="add_to_expression!-1"></a><a class="docs-heading-anchor-permalink" href="#add_to_expression!" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_to_expression!" href="#JuMP.add_to_expression!"><code>JuMP.add_to_expression!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_to_expression!(expression, terms...)</code></pre><p>Updates <code>expression</code> <em>in place</em> to <code>expression + (*)(terms...)</code>. This is typically much more efficient than <code>expression += (*)(terms...)</code>. For example, <code>add_to_expression!(expression, a, b)</code> produces the same result as <code>expression += a*b</code>, and <code>add_to_expression!(expression, a)</code> produces the same result as <code>expression += a</code>.</p><p>Only a few methods are defined, mostly for internal use, and only for the cases when (1) they can be implemented efficiently and (2) <code>expression</code> is capable of storing the result. For example, <code>add_to_expression!(::AffExpr, ::GenericVariableRef, ::GenericVariableRef)</code> is not defined because a <code>GenericAffExpr</code> cannot store the product of two variables.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L411-L425">source</a></section></article><h2 id="add_to_function_constant"><a class="docs-heading-anchor" href="#add_to_function_constant"><code>add_to_function_constant</code></a><a id="add_to_function_constant-1"></a><a class="docs-heading-anchor-permalink" href="#add_to_function_constant" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_to_function_constant" href="#JuMP.add_to_function_constant"><code>JuMP.add_to_function_constant</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_to_function_constant(constraint::ConstraintRef, value)</code></pre><p>Add <code>value</code> to the function constant term.</p><p>Note that for scalar constraints, JuMP will aggregate all constant terms onto the right-hand side of the constraint so instead of modifying the function, the set will be translated by <code>-value</code>. For example, given a constraint <code>2x &lt;= 3</code>, <code>add_to_function_constant(c, 4)</code> will modify it to <code>2x &lt;= -1</code>.</p><p><strong>Example</strong></p><p>For scalar constraints, the set is translated by <code>-value</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con, 0 &lt;= 2x - 1 &lt;= 2)
+con : 2 x ∈ [1, 3]
+
+julia&gt; add_to_function_constant(con, 4)
+
+julia&gt; con
+con : 2 x ∈ [-3, -1]</code></pre><p>For vector constraints, the constant is added to the function:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @constraint(model, con, [x + y, x, y] in SecondOrderCone())
+con : [x + y, x, y] ∈ MathOptInterface.SecondOrderCone(3)
+
+julia&gt; add_to_function_constant(con, [1, 2, 2])
+
+julia&gt; con
+con : [x + y + 1, x + 2, y + 2] ∈ MathOptInterface.SecondOrderCone(3)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L921-L965">source</a></section></article><h2 id="add_variable"><a class="docs-heading-anchor" href="#add_variable"><code>add_variable</code></a><a id="add_variable-1"></a><a class="docs-heading-anchor-permalink" href="#add_variable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.add_variable" href="#JuMP.add_variable"><code>JuMP.add_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_variable(m::GenericModel, v::AbstractVariable, name::String=&quot;&quot;)</code></pre><p>Add a variable <code>v</code> to <code>Model m</code> and sets its name.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1717-L1721">source</a></section></article><h2 id="all_constraints"><a class="docs-heading-anchor" href="#all_constraints"><code>all_constraints</code></a><a id="all_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#all_constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.all_constraints" href="#JuMP.all_constraints"><code>JuMP.all_constraints</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">all_constraints(model::GenericModel, function_type, set_type)::Vector{&lt;:ConstraintRef}</code></pre><p>Return a list of all constraints currently in the model where the function has type <code>function_type</code> and the set has type <code>set_type</code>. The constraints are ordered by creation time.</p><p>See also <a href="#JuMP.list_of_constraint_types"><code>list_of_constraint_types</code></a> and <a href="#JuMP.num_constraints"><code>num_constraints</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 0, Bin);
+
+julia&gt; @constraint(model, 2x &lt;= 1);
+
+julia&gt; all_constraints(model, VariableRef, MOI.GreaterThan{Float64})
+1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:
+ x ≥ 0
+
+julia&gt; all_constraints(model, VariableRef, MOI.ZeroOne)
+1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}, ScalarShape}}:
+ x binary
+
+julia&gt; all_constraints(model, AffExpr, MOI.LessThan{Float64})
+1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ 2 x ≤ 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1239-L1269">source</a></section><section><div><pre><code class="language-julia hljs">all_constraints(
+    model::GenericModel;
+    include_variable_in_set_constraints::Bool,
+)::Vector{ConstraintRef}</code></pre><p>Return a list of all constraints in <code>model</code>.</p><p>If <code>include_variable_in_set_constraints == true</code>, then <code>VariableRef</code> constraints such as <code>VariableRef</code>-in-<code>Integer</code> are included. To return only the structural constraints (e.g., the rows in the constraint matrix of a linear program), pass <code>include_variable_in_set_constraints = false</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 0, Int);
+
+julia&gt; @constraint(model, 2x &lt;= 1);
+
+julia&gt; @NLconstraint(model, x^2 &lt;= 1);
+
+julia&gt; all_constraints(model; include_variable_in_set_constraints = true)
+4-element Vector{ConstraintRef}:
+ 2 x ≤ 1
+ x ≥ 0
+ x integer
+ x ^ 2.0 - 1.0 ≤ 0
+
+julia&gt; all_constraints(model; include_variable_in_set_constraints = false)
+2-element Vector{ConstraintRef}:
+ 2 x ≤ 1
+ x ^ 2.0 - 1.0 ≤ 0</code></pre><p><strong>Performance considerations</strong></p><p>Note that this function is type-unstable because it returns an abstractly typed vector. If performance is a problem, consider using <a href="#JuMP.list_of_constraint_types"><code>list_of_constraint_types</code></a> and a function barrier. See the <a href="../../developers/extensions/#Performance-tips-for-extensions">Performance tips for extensions</a> section of the documentation for more details.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1378-L1421">source</a></section></article><h2 id="all_nonlinear_constraints"><a class="docs-heading-anchor" href="#all_nonlinear_constraints"><code>all_nonlinear_constraints</code></a><a id="all_nonlinear_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#all_nonlinear_constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.all_nonlinear_constraints" href="#JuMP.all_nonlinear_constraints"><code>JuMP.all_nonlinear_constraints</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">all_nonlinear_constraints(model::GenericModel)</code></pre><p>Return a vector of all nonlinear constraint references in the model in the order they were added to the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L512-L517">source</a></section></article><h2 id="all_variables"><a class="docs-heading-anchor" href="#all_variables"><code>all_variables</code></a><a id="all_variables-1"></a><a class="docs-heading-anchor-permalink" href="#all_variables" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.all_variables" href="#JuMP.all_variables"><code>JuMP.all_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">all_variables(model::GenericModel{T})::Vector{GenericVariableRef{T}} where {T}</code></pre><p>Returns a list of all variables currently in the model. The variables are ordered by creation time.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, y);
+
+julia&gt; all_variables(model)
+2-element Vector{VariableRef}:
+ x
+ y</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L2131-L2151">source</a></section></article><h2 id="anonymous_name"><a class="docs-heading-anchor" href="#anonymous_name"><code>anonymous_name</code></a><a id="anonymous_name-1"></a><a class="docs-heading-anchor-permalink" href="#anonymous_name" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.anonymous_name" href="#JuMP.anonymous_name"><code>JuMP.anonymous_name</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">anonymous_name(::MIME, x::AbstractVariableRef)</code></pre><p>The name to use for an anonymous variable <code>x</code> when printing.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L547-L551">source</a></section></article><h2 id="backend"><a class="docs-heading-anchor" href="#backend"><code>backend</code></a><a id="backend-1"></a><a class="docs-heading-anchor-permalink" href="#backend" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.backend" href="#JuMP.backend"><code>JuMP.backend</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">backend(model::GenericModel)</code></pre><p>Return the lower-level MathOptInterface model that sits underneath JuMP. This model depends on which operating mode JuMP is in (see <a href="#JuMP.mode"><code>mode</code></a>).</p><ul><li>If JuMP is in <code>DIRECT</code> mode (i.e., the model was created using <a href="#JuMP.direct_model"><code>direct_model</code></a>), the backend will be the optimizer passed to <a href="#JuMP.direct_model"><code>direct_model</code></a>.</li><li>If JuMP is in <code>MANUAL</code> or <code>AUTOMATIC</code> mode, the backend is a <code>MOI.Utilities.CachingOptimizer</code>.</li></ul><p><strong>This function should only be used by advanced users looking to access low-level MathOptInterface or solver-specific functionality.</strong></p><p><strong>Notes</strong></p><p>If JuMP is not in <code>DIRECT</code> mode, the type returned by <code>backend</code> may change between any JuMP releases. Therefore, only use the public API exposed by MathOptInterface, and do not access internal fields. If you require access to the innermost optimizer, see <a href="#JuMP.unsafe_backend"><code>unsafe_backend</code></a>. Alternatively, use <a href="#JuMP.direct_model"><code>direct_model</code></a> to create a JuMP model in <code>DIRECT</code> mode.</p><p>See also: <a href="#JuMP.unsafe_backend"><code>unsafe_backend</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L385-L409">source</a></section></article><h2 id="barrier_iterations"><a class="docs-heading-anchor" href="#barrier_iterations"><code>barrier_iterations</code></a><a id="barrier_iterations-1"></a><a class="docs-heading-anchor-permalink" href="#barrier_iterations" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.barrier_iterations" href="#JuMP.barrier_iterations"><code>JuMP.barrier_iterations</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">barrier_iterations(model::GenericModel)</code></pre><p>Gets the cumulative number of barrier iterations during the most recent optimization.</p><p>Solvers must implement <code>MOI.BarrierIterations()</code> to use this function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L600-L607">source</a></section></article><h2 id="bridge_constraints"><a class="docs-heading-anchor" href="#bridge_constraints"><code>bridge_constraints</code></a><a id="bridge_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#bridge_constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.bridge_constraints" href="#JuMP.bridge_constraints"><code>JuMP.bridge_constraints</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridge_constraints(model::GenericModel)</code></pre><p>When in direct mode, return <code>false</code>.</p><p>When in manual or automatic mode, return a <code>Bool</code> indicating whether the optimizer is set and unsupported constraints are automatically bridged to equivalent supported constraints when an appropriate transformation is available.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L569-L578">source</a></section></article><h2 id="build_constraint"><a class="docs-heading-anchor" href="#build_constraint"><code>build_constraint</code></a><a id="build_constraint-1"></a><a class="docs-heading-anchor-permalink" href="#build_constraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.build_constraint" href="#JuMP.build_constraint"><code>JuMP.build_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">build_constraint(
+    _error::Function,
+    Q::LinearAlgebra.Symmetric{V, M},
+    ::PSDCone,
+) where {V&lt;:AbstractJuMPScalar,M&lt;:AbstractMatrix{V}}</code></pre><p>Return a <code>VectorConstraint</code> of shape <a href="#JuMP.SymmetricMatrixShape"><code>SymmetricMatrixShape</code></a> constraining the matrix <code>Q</code> to be positive semidefinite.</p><p>This function is used by the <a href="#JuMP.@constraint"><code>@constraint</code></a> macros as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; import LinearAlgebra
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2]);
+
+julia&gt; @constraint(model, LinearAlgebra.Symmetric(Q) in PSDCone())
+[Q[1,1]  Q[1,2];
+ Q[1,2]  Q[2,2]] ∈ PSDCone()</code></pre><p>The form above is usually used when the entries of <code>Q</code> are affine or quadratic expressions, but it can also be used when the entries are variables to get the reference of the semidefinite constraint, e.g.,</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2], Symmetric)
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ Q[1,1]  Q[1,2]
+ Q[1,2]  Q[2,2]
+
+julia&gt; @constraint(model, Q in PSDCone())
+[Q[1,1]  Q[1,2];
+ Q[1,2]  Q[2,2]] ∈ PSDCone()</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L435-L473">source</a></section><section><div><pre><code class="language-julia hljs">build_constraint(
+    _error::Function,
+    Q::AbstractMatrix{&lt;:AbstractJuMPScalar},
+    ::PSDCone,
+)</code></pre><p>Return a <code>VectorConstraint</code> of shape <a href="#JuMP.SquareMatrixShape"><code>SquareMatrixShape</code></a> constraining the matrix <code>Q</code> to be symmetric and positive semidefinite.</p><p>This function is used by the <a href="#JuMP.@constraint"><code>@constraint</code></a> macro as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2]);
+
+julia&gt; @constraint(model, Q in PSDCone())
+[Q[1,1]  Q[1,2];
+ Q[2,1]  Q[2,2]] ∈ PSDCone()</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L483-L503">source</a></section><section><div><pre><code class="language-julia hljs">build_constraint(
+    _error::Function,
+    Q::LinearAlgebra.Hermitian{V,M},
+    ::HermitianPSDCone,
+) where {V&lt;:AbstractJuMPScalar,M&lt;:AbstractMatrix{V}}</code></pre><p>Return a <code>VectorConstraint</code> of shape <a href="#JuMP.HermitianMatrixShape"><code>HermitianMatrixShape</code></a> constraining the matrix <code>Q</code> to be Hermitian positive semidefinite.</p><p>This function is used by the <a href="#JuMP.@constraint"><code>@constraint</code></a> macros as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; import LinearAlgebra
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2]);
+
+julia&gt; @constraint(model, LinearAlgebra.Hermitian(Q) in HermitianPSDCone())
+[Q[1,1]  Q[1,2];
+ Q[1,2]  Q[2,2]] ∈ HermitianPSDCone()</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L643-L665">source</a></section><section><div><pre><code class="language-julia hljs">build_constraint(
+    _error::Function,
+    f::AbstractVector{&lt;:AbstractJuMPScalar},
+    ::Nonnegatives,
+    extra::Union{MOI.AbstractVectorSet,AbstractVectorSet},
+)</code></pre><p>A helper method that re-writes</p><pre><code class="language-julia hljs">@constraint(model, X &gt;= Y, extra)</code></pre><p>into</p><pre><code class="language-julia hljs">@constraint(model, X - Y in extra)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L44-L60">source</a></section><section><div><pre><code class="language-julia hljs">build_constraint(
+    _error::Function,
+    f::AbstractVector{&lt;:AbstractJuMPScalar},
+    ::Nonpositives,
+    extra::Union{MOI.AbstractVectorSet,AbstractVectorSet},
+)</code></pre><p>A helper method that re-writes</p><pre><code class="language-julia hljs">@constraint(model, Y &lt;= X, extra)</code></pre><p>into</p><pre><code class="language-julia hljs">@constraint(model, X - Y in extra)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L70-L86">source</a></section></article><h2 id="build_variable"><a class="docs-heading-anchor" href="#build_variable"><code>build_variable</code></a><a id="build_variable-1"></a><a class="docs-heading-anchor-permalink" href="#build_variable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.build_variable" href="#JuMP.build_variable"><code>JuMP.build_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">build_variable(
+    _error::Function,
+    info::VariableInfo,
+    args...;
+    kwargs...,
+)</code></pre><p>Return a new <a href="#JuMP.AbstractVariable"><code>AbstractVariable</code></a> object.</p><p>This method should only be implemented by developers creating JuMP extensions. It should never be called by users of JuMP.</p><p><strong>Arguments</strong></p><ul><li><code>_error</code>: a function to call instead of <code>error</code>. <code>_error</code> annotates the error message with additional information for the user.</li><li><code>info</code>: an instance of <a href="#JuMP.VariableInfo"><code>VariableInfo</code></a>. This has a variety of fields relating to the variable such as <code>info.lower_bound</code> and <code>info.binary</code>.</li><li><code>args</code>: optional additional positional arguments for extending the <a href="#JuMP.@variable"><code>@variable</code></a> macro.</li><li><code>kwargs</code>: optional keyword arguments for extending the <a href="#JuMP.@variable"><code>@variable</code></a> macro.</li></ul><p>See also: <a href="#JuMP.@variable"><code>@variable</code></a></p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Extensions should define a method with ONE positional argument to dispatch the call to a different method. Creating an extension that relies on multiple positional arguments leads to <code>MethodError</code>s if the user passes the arguments in the wrong order.</p></div></div><p><strong>Example</strong></p><pre><code class="language-julia hljs">@variable(model, x, Foo)</code></pre><p>will call</p><pre><code class="language-julia hljs">build_variable(_error::Function, info::VariableInfo, ::Type{Foo})</code></pre><p>Passing special-case positional arguments such as <code>Bin</code>, <code>Int</code>, and <code>PSD</code> is okay, along with keyword arguments:</p><pre><code class="language-julia hljs">@variable(model, x, Int, Foo(), mykwarg = true)
+# or
+@variable(model, x, Foo(), Int, mykwarg = true)</code></pre><p>will call</p><pre><code class="language-julia hljs">build_variable(_error::Function, info::VariableInfo, ::Foo; mykwarg)</code></pre><p>and <code>info.integer</code> will be true.</p><p>Note that the order of the positional arguments does not matter.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1977-L2033">source</a></section><section><div><pre><code class="language-julia hljs">build_variable(_error::Function, variables, ::SymmetricMatrixSpace)</code></pre><p>Return a <code>VariablesConstrainedOnCreation</code> of shape <a href="#JuMP.SymmetricMatrixShape"><code>SymmetricMatrixShape</code></a> creating variables in <code>MOI.Reals</code>, i.e. &quot;free&quot; variables unless they are constrained after their creation.</p><p>This function is used by the <a href="#JuMP.@variable"><code>@variable</code></a> macro as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2], Symmetric)
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ Q[1,1]  Q[1,2]
+ Q[1,2]  Q[2,2]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L302-L318">source</a></section><section><div><pre><code class="language-julia hljs">build_variable(_error::Function, variables, ::SkewSymmetricMatrixSpace)</code></pre><p>Return a <code>VariablesConstrainedOnCreation</code> of shape <a href="#JuMP.SkewSymmetricMatrixShape"><code>SkewSymmetricMatrixShape</code></a> creating variables in <code>MOI.Reals</code>, i.e. &quot;free&quot; variables unless they are constrained after their creation.</p><p>This function is used by the <a href="#JuMP.@variable"><code>@variable</code></a> macro as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2] in SkewSymmetricMatrixSpace())
+2×2 Matrix{AffExpr}:
+ 0        Q[1,2]
+ -Q[1,2]  0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L334-L350">source</a></section><section><div><pre><code class="language-julia hljs">build_variable(_error::Function, variables, ::HermitianMatrixSpace)</code></pre><p>Return a <code>VariablesConstrainedOnCreation</code> of shape <a href="#JuMP.HermitianMatrixShape"><code>HermitianMatrixShape</code></a> creating variables in <code>MOI.Reals</code>, i.e. &quot;free&quot; variables unless they are constrained after their creation.</p><p>This function is used by the <a href="#JuMP.@variable"><code>@variable</code></a> macro as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2] in HermitianMatrixSpace())
+2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(Q[1,1])                    real(Q[1,2]) + imag(Q[1,2]) im
+ real(Q[1,2]) - imag(Q[1,2]) im  real(Q[2,2])</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L366-L382">source</a></section><section><div><pre><code class="language-julia hljs">build_variable(_error::Function, variables, ::PSDCone)</code></pre><p>Return a <code>VariablesConstrainedOnCreation</code> of shape <a href="#JuMP.SymmetricMatrixShape"><code>SymmetricMatrixShape</code></a> constraining the variables to be positive semidefinite.</p><p>This function is used by the <a href="#JuMP.@variable"><code>@variable</code></a> macro as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2], PSD)
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ Q[1,1]  Q[1,2]
+ Q[1,2]  Q[2,2]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L400-L415">source</a></section></article><h2 id="callback_node_status"><a class="docs-heading-anchor" href="#callback_node_status"><code>callback_node_status</code></a><a id="callback_node_status-1"></a><a class="docs-heading-anchor-permalink" href="#callback_node_status" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.callback_node_status" href="#JuMP.callback_node_status"><code>JuMP.callback_node_status</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">callback_node_status(cb_data, model::GenericModel)</code></pre><p>Return an <a href="../../moi/reference/callbacks/#MathOptInterface.CallbackNodeStatusCode"><code>MOI.CallbackNodeStatusCode</code></a> enum, indicating if the current primal solution available from <a href="#JuMP.callback_value"><code>callback_value</code></a> is integer feasible.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/callbacks.jl#L11-L16">source</a></section></article><h2 id="callback_value"><a class="docs-heading-anchor" href="#callback_value"><code>callback_value</code></a><a id="callback_value-1"></a><a class="docs-heading-anchor-permalink" href="#callback_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.callback_value" href="#JuMP.callback_value"><code>JuMP.callback_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">callback_value(cb_data, x::GenericVariableRef)</code></pre><p>Return the primal solution of a variable inside a callback.</p><p><code>cb_data</code> is the argument to the callback function, and the type is dependent on the solver.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/callbacks.jl#L30-L37">source</a></section><section><div><pre><code class="language-julia hljs">callback_value(cb_data, expr::Union{GenericAffExpr, GenericQuadExpr})</code></pre><p>Return the primal solution of an affine or quadratic expression inside a callback by getting the value for each variable appearing in the expression.</p><p><code>cb_data</code> is the argument to the callback function, and the type is dependent on the solver.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/callbacks.jl#L55-L63">source</a></section></article><h2 id="check_belongs_to_model"><a class="docs-heading-anchor" href="#check_belongs_to_model"><code>check_belongs_to_model</code></a><a id="check_belongs_to_model-1"></a><a class="docs-heading-anchor-permalink" href="#check_belongs_to_model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.check_belongs_to_model" href="#JuMP.check_belongs_to_model"><code>JuMP.check_belongs_to_model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">check_belongs_to_model(func::AbstractJuMPScalar, model::AbstractModel)</code></pre><p>Throw <a href="#JuMP.VariableNotOwned"><code>VariableNotOwned</code></a> if the <a href="#JuMP.owner_model"><code>owner_model</code></a> of one of the variables of the function <code>func</code> is not <code>model</code>.</p><pre><code class="nohighlight hljs">check_belongs_to_model(constraint::AbstractConstraint, model::AbstractModel)</code></pre><p>Throw <a href="#JuMP.VariableNotOwned"><code>VariableNotOwned</code></a> if the <a href="#JuMP.owner_model"><code>owner_model</code></a> of one of the variables of the constraint <code>constraint</code> is not <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L321-L331">source</a></section></article><h2 id="coefficient"><a class="docs-heading-anchor" href="#coefficient"><code>coefficient</code></a><a id="coefficient-1"></a><a class="docs-heading-anchor-permalink" href="#coefficient" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.coefficient" href="#JuMP.coefficient"><code>JuMP.coefficient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">coefficient(v1::GenericVariableRef{T}, v2::GenericVariableRef{T}) where {T}</code></pre><p>Return <code>one(T)</code> if <code>v1 == v2</code>, and <code>zero(T)</code> otherwise.</p><p>This is a fallback for other <a href="#JuMP.coefficient"><code>coefficient</code></a> methods to simplify code in which the expression may be a single variable.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L360-L367">source</a></section><section><div><pre><code class="language-julia hljs">coefficient(a::GenericAffExpr{C,V}, v::V) where {C,V}</code></pre><p>Return the coefficient associated with variable <code>v</code> in the affine expression <code>a</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L247-L251">source</a></section><section><div><pre><code class="language-julia hljs">coefficient(a::GenericAffExpr{C,V}, v1::V, v2::V) where {C,V}</code></pre><p>Return the coefficient associated with the term <code>v1 * v2</code> in the quadratic expression <code>a</code>.</p><p>Note that <code>coefficient(a, v1, v2)</code> is the same as <code>coefficient(a, v2, v1)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L148-L154">source</a></section><section><div><pre><code class="language-julia hljs">coefficient(a::GenericQuadExpr{C,V}, v::V) where {C,V}</code></pre><p>Return the coefficient associated with variable <code>v</code> in the affine component of <code>a</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L159-L163">source</a></section></article><h2 id="compute_conflict!"><a class="docs-heading-anchor" href="#compute_conflict!"><code>compute_conflict!</code></a><a id="compute_conflict!-1"></a><a class="docs-heading-anchor-permalink" href="#compute_conflict!" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.compute_conflict!" href="#JuMP.compute_conflict!"><code>JuMP.compute_conflict!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">compute_conflict!(model::GenericModel)</code></pre><p>Compute a conflict if the model is infeasible. If an optimizer has not been set yet (see <a href="#JuMP.set_optimizer"><code>set_optimizer</code></a>), a <a href="#JuMP.NoOptimizer"><code>NoOptimizer</code></a> error is thrown.</p><p>The status of the conflict can be checked with the <code>MOI.ConflictStatus</code> model attribute. Then, the status for each constraint can be queried with the <code>MOI.ConstraintConflictStatus</code> attribute.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L478-L488">source</a></section></article><h2 id="constant"><a class="docs-heading-anchor" href="#constant"><code>constant</code></a><a id="constant-1"></a><a class="docs-heading-anchor-permalink" href="#constant" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.constant" href="#JuMP.constant"><code>JuMP.constant</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constant(aff::GenericAffExpr{C, V})::C</code></pre><p>Return the constant of the affine expression.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L361-L365">source</a></section><section><div><pre><code class="language-julia hljs">constant(aff::GenericQuadExpr{C, V})::C</code></pre><p>Return the constant of the quadratic expression.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L259-L263">source</a></section></article><h2 id="constraint_by_name"><a class="docs-heading-anchor" href="#constraint_by_name"><code>constraint_by_name</code></a><a id="constraint_by_name-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_by_name" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.constraint_by_name" href="#JuMP.constraint_by_name"><code>JuMP.constraint_by_name</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constraint_by_name(model::AbstractModel,
+                   name::String)::Union{ConstraintRef, Nothing}</code></pre><p>Return the reference of the constraint with name attribute <code>name</code> or <code>Nothing</code> if no constraint has this name attribute. Throws an error if several constraints have <code>name</code> as their name attribute.</p><pre><code class="nohighlight hljs">constraint_by_name(model::AbstractModel,
+                   name::String,
+                   F::Type{&lt;:Union{AbstractJuMPScalar,
+                                   Vector{&lt;:AbstractJuMPScalar},
+                                   MOI.AbstactFunction}},
+                   S::Type{&lt;:MOI.AbstractSet})::Union{ConstraintRef, Nothing}</code></pre><p>Similar to the method above, except that it throws an error if the constraint is not an <code>F</code>-in-<code>S</code> contraint where <code>F</code> is either the JuMP or MOI type of the function, and <code>S</code> is the MOI type of the set. This method is recommended if you know the type of the function and set since its returned type can be inferred while for the method above (i.e. without <code>F</code> and <code>S</code>), the exact return type of the constraint index cannot be inferred.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, con, x^2 == 1)
+con : x² = 1
+
+julia&gt; constraint_by_name(model, &quot;kon&quot;)
+
+julia&gt; constraint_by_name(model, &quot;con&quot;)
+con : x² = 1
+
+julia&gt; constraint_by_name(model, &quot;con&quot;, AffExpr, MOI.EqualTo{Float64})
+
+julia&gt; constraint_by_name(model, &quot;con&quot;, QuadExpr, MOI.EqualTo{Float64})
+con : x² = 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L268-L309">source</a></section></article><h2 id="constraint_object"><a class="docs-heading-anchor" href="#constraint_object"><code>constraint_object</code></a><a id="constraint_object-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_object" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.constraint_object" href="#JuMP.constraint_object"><code>JuMP.constraint_object</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constraint_object(con_ref::ConstraintRef)</code></pre><p>Return the underlying constraint data for the constraint referenced by <code>ref</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L570-L574">source</a></section></article><h2 id="constraint_ref_with_index"><a class="docs-heading-anchor" href="#constraint_ref_with_index"><code>constraint_ref_with_index</code></a><a id="constraint_ref_with_index-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_ref_with_index" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.constraint_ref_with_index" href="#JuMP.constraint_ref_with_index"><code>JuMP.constraint_ref_with_index</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constraint_ref_with_index(model::AbstractModel, index::MOI.ConstraintIndex)</code></pre><p>Return a <code>ConstraintRef</code> of <code>model</code> corresponding to <code>index</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L343-L347">source</a></section></article><h2 id="constraint_string"><a class="docs-heading-anchor" href="#constraint_string"><code>constraint_string</code></a><a id="constraint_string-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.constraint_string" href="#JuMP.constraint_string"><code>JuMP.constraint_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constraint_string(
+    mode::MIME,
+    ref::ConstraintRef;
+    in_math_mode::Bool = false)</code></pre><p>Return a string representation of the constraint <code>ref</code>, given the <code>mode</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L792-L799">source</a></section></article><h2 id="constraints_string"><a class="docs-heading-anchor" href="#constraints_string"><code>constraints_string</code></a><a id="constraints_string-1"></a><a class="docs-heading-anchor-permalink" href="#constraints_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.constraints_string" href="#JuMP.constraints_string"><code>JuMP.constraints_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constraints_string(mode, model::AbstractModel)::Vector{String}</code></pre><p>Return a list of <code>String</code>s describing each constraint of the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L414-L418">source</a></section></article><h2 id="copy_conflict"><a class="docs-heading-anchor" href="#copy_conflict"><code>copy_conflict</code></a><a id="copy_conflict-1"></a><a class="docs-heading-anchor-permalink" href="#copy_conflict" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.copy_conflict" href="#JuMP.copy_conflict"><code>JuMP.copy_conflict</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">copy_conflict(model::GenericModel)</code></pre><p>Return a copy of the current conflict for the model <code>model</code> and a <a href="#JuMP.GenericReferenceMap"><code>GenericReferenceMap</code></a> that can be used to obtain the variable and constraint reference of the new model corresponding to a given <code>model</code>&#39;s reference.</p><p>This is a convenience function that provides a filtering function for <a href="#JuMP.copy_model"><code>copy_model</code></a>.</p><p><strong>Note</strong></p><p>Model copy is not supported in <code>DIRECT</code> mode, i.e. when a model is constructed using the <a href="#JuMP.direct_model"><code>direct_model</code></a> constructor instead of the <a href="#JuMP.Model"><code>Model</code></a> constructor. Moreover, independently on whether an optimizer was provided at model construction, the new model will have no optimizer, i.e., an optimizer will have to be provided to the new model in the <a href="#JuMP.optimize!"><code>optimize!</code></a> call.</p><p><strong>Example</strong></p><p>In the following example, a model <code>model</code> is constructed with a variable <code>x</code> and two constraints <code>c1</code> and <code>c2</code>. This model has no solution, as the two constraints are mutually exclusive. The solver is asked to compute a conflict with <a href="#JuMP.compute_conflict!"><code>compute_conflict!</code></a>. The parts of <code>model</code> participating in the conflict are then copied into a model <code>iis_model</code>.</p><pre><code class="language-julia hljs">julia&gt; using JuMP
+
+julia&gt; import Gurobi
+
+julia&gt; model = Model(Gurobi.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 0)
+x
+
+julia&gt; @constraint(model, c1, x &gt;= 2)
+c1 : x ≥ 2
+
+julia&gt; @constraint(model, c2, x &lt;= 1)
+c2 : x ≤ 1
+
+julia&gt; optimize!(model)
+
+julia&gt; compute_conflict!(model)
+
+julia&gt; if get_attribute(model, MOI.ConflictStatus()) == MOI.CONFLICT_FOUND
+           iis_model, reference_map = copy_conflict(model)
+           print(iis_model)
+       end
+Feasibility
+Subject to
+ c1 : x ≥ 2
+ c2 : x ≤ 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/copy.jl#L247-L305">source</a></section></article><h2 id="copy_extension_data"><a class="docs-heading-anchor" href="#copy_extension_data"><code>copy_extension_data</code></a><a id="copy_extension_data-1"></a><a class="docs-heading-anchor-permalink" href="#copy_extension_data" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.copy_extension_data" href="#JuMP.copy_extension_data"><code>JuMP.copy_extension_data</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">copy_extension_data(data, new_model::AbstractModel, model::AbstractModel)</code></pre><p>Return a copy of the extension data <code>data</code> of the model <code>model</code> to the extension data of the new model <code>new_model</code>.</p><p>A method should be added for any JuMP extension storing data in the <code>ext</code> field.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Do not engage in type piracy by implementing this method for types of <code>data</code> that you did not define! JuMP extensions should store types that they define in <code>model.ext</code>, rather than regular Julia types.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/copy.jl#L6-L18">source</a></section></article><h2 id="copy_model"><a class="docs-heading-anchor" href="#copy_model"><code>copy_model</code></a><a id="copy_model-1"></a><a class="docs-heading-anchor-permalink" href="#copy_model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.copy_model" href="#JuMP.copy_model"><code>JuMP.copy_model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">copy_model(model::GenericModel; filter_constraints::Union{Nothing, Function}=nothing)</code></pre><p>Return a copy of the model <code>model</code> and a <a href="#JuMP.GenericReferenceMap"><code>GenericReferenceMap</code></a> that can be used to obtain the variable and constraint reference of the new model corresponding to a given <code>model</code>&#39;s reference. A <a href="#Base.copy-Tuple{AbstractModel}"><code>Base.copy(::AbstractModel)</code></a> method has also been implemented, it is similar to <code>copy_model</code> but does not return the reference map.</p><p>If the <code>filter_constraints</code> argument is given, only the constraints for which this function returns <code>true</code> will be copied. This function is given a constraint reference as argument.</p><p><strong>Note</strong></p><p>Model copy is not supported in <code>DIRECT</code> mode, i.e. when a model is constructed using the <a href="#JuMP.direct_model"><code>direct_model</code></a> constructor instead of the <a href="#JuMP.Model"><code>Model</code></a> constructor. Moreover, independently on whether an optimizer was provided at model construction, the new model will have no optimizer, i.e., an optimizer will have to be provided to the new model in the <a href="#JuMP.optimize!"><code>optimize!</code></a> call.</p><p><strong>Example</strong></p><p>In the following example, a model <code>model</code> is constructed with a variable <code>x</code> and a constraint <code>cref</code>. It is then copied into a model <code>new_model</code> with the new references assigned to <code>x_new</code> and <code>cref_new</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, cref, x == 2)
+cref : x = 2
+
+julia&gt; new_model, reference_map = copy_model(model);
+
+julia&gt; x_new = reference_map[x]
+x
+
+julia&gt; cref_new = reference_map[cref]
+cref : x = 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/copy.jl#L93-L137">source</a></section></article><h2 id="delete"><a class="docs-heading-anchor" href="#delete"><code>delete</code></a><a id="delete-1"></a><a class="docs-heading-anchor-permalink" href="#delete" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.delete" href="#JuMP.delete"><code>JuMP.delete</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">delete(model::GenericModel, con_ref::ConstraintRef)</code></pre><p>Delete the constraint associated with <code>constraint_ref</code> from the model <code>model</code>.</p><p>Note that <code>delete</code> does not unregister the name from the model, so adding a new constraint of the same name will throw an error. Use <a href="#JuMP.unregister"><code>unregister</code></a> to unregister the name after deletion.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, c, 2x &lt;= 1)
+c : 2 x ≤ 1
+
+julia&gt; delete(model, c)
+
+julia&gt; unregister(model, :c)
+
+julia&gt; print(model)
+Feasibility
+Subject to
+
+julia&gt; model[:c]
+ERROR: KeyError: key :c not found
+Stacktrace:
+[...]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L367-L399">source</a></section><section><div><pre><code class="language-julia hljs">delete(model::GenericModel, con_refs::Vector{&lt;:ConstraintRef})</code></pre><p>Delete the constraints associated with <code>con_refs</code> from the model <code>model</code>. Solvers may implement specialized methods for deleting multiple constraints of the same concrete type, i.e., when <code>isconcretetype(eltype(con_refs))</code>. These may be more efficient than repeatedly calling the single constraint delete method.</p><p>See also: <a href="#JuMP.unregister"><code>unregister</code></a></p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L411-L421">source</a></section><section><div><pre><code class="language-julia hljs">delete(model::GenericModel, variable_ref::GenericVariableRef)</code></pre><p>Delete the variable associated with <code>variable_ref</code> from the model <code>model</code>.</p><p>Note that <code>delete</code> does not unregister the name from the model, so adding a new variable of the same name will throw an error. Use <a href="#JuMP.unregister"><code>unregister</code></a> to unregister the name after deletion.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; delete(model, x)
+
+julia&gt; unregister(model, :x)
+
+julia&gt; print(model)
+Feasibility
+Subject to
+
+julia&gt; model[:x]
+ERROR: KeyError: key :x not found
+Stacktrace:
+[...]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L391-L421">source</a></section><section><div><pre><code class="language-julia hljs">delete(model::GenericModel, variable_refs::Vector{&lt;:GenericVariableRef})</code></pre><p>Delete the variables associated with <code>variable_refs</code> from the model <code>model</code>. Solvers may implement methods for deleting multiple variables that are more efficient than repeatedly calling the single variable delete method.</p><p>See also: <a href="#JuMP.unregister"><code>unregister</code></a></p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L434-L442">source</a></section><section><div><pre><code class="language-julia hljs">delete(model::Model, c::NonlinearConstraintRef)</code></pre><p>Delete the nonlinear constraint <code>c</code> from <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L487-L491">source</a></section></article><h2 id="delete_lower_bound"><a class="docs-heading-anchor" href="#delete_lower_bound"><code>delete_lower_bound</code></a><a id="delete_lower_bound-1"></a><a class="docs-heading-anchor-permalink" href="#delete_lower_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.delete_lower_bound" href="#JuMP.delete_lower_bound"><code>JuMP.delete_lower_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">delete_lower_bound(v::GenericVariableRef)</code></pre><p>Delete the lower bound constraint of a variable.</p><p>See also <a href="#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, <a href="#JuMP.has_lower_bound"><code>has_lower_bound</code></a>, <a href="#JuMP.lower_bound"><code>lower_bound</code></a>, <a href="#JuMP.set_lower_bound"><code>set_lower_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 1.0);
+
+julia&gt; has_lower_bound(x)
+true
+
+julia&gt; delete_lower_bound(x)
+
+julia&gt; has_lower_bound(x)
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L779-L802">source</a></section></article><h2 id="delete_upper_bound"><a class="docs-heading-anchor" href="#delete_upper_bound"><code>delete_upper_bound</code></a><a id="delete_upper_bound-1"></a><a class="docs-heading-anchor-permalink" href="#delete_upper_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.delete_upper_bound" href="#JuMP.delete_upper_bound"><code>JuMP.delete_upper_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">delete_upper_bound(v::GenericVariableRef)</code></pre><p>Delete the upper bound constraint of a variable.</p><p>Errors if one does not exist.</p><p>See also <a href="#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, <a href="#JuMP.has_upper_bound"><code>has_upper_bound</code></a>, <a href="#JuMP.upper_bound"><code>upper_bound</code></a>, <a href="#JuMP.set_upper_bound"><code>set_upper_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &lt;= 1.0);
+
+julia&gt; has_upper_bound(x)
+true
+
+julia&gt; delete_upper_bound(x)
+
+julia&gt; has_upper_bound(x)
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L950-L975">source</a></section></article><h2 id="direct_generic_model"><a class="docs-heading-anchor" href="#direct_generic_model"><code>direct_generic_model</code></a><a id="direct_generic_model-1"></a><a class="docs-heading-anchor-permalink" href="#direct_generic_model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.direct_generic_model" href="#JuMP.direct_generic_model"><code>JuMP.direct_generic_model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">direct_generic_model(
+    value_type::Type{T},
+    backend::MOI.ModelLike;
+) where {T&lt;:Real}</code></pre><p>Return a new JuMP model using <a href="#JuMP.backend"><code>backend</code></a> to store the model and solve it.</p><p>As opposed to the <a href="#JuMP.Model"><code>Model</code></a> constructor, no cache of the model is stored outside of <a href="#JuMP.backend"><code>backend</code></a> and no bridges are automatically applied to <a href="#JuMP.backend"><code>backend</code></a>.</p><p><strong>Notes</strong></p><p>The absence of a cache reduces the memory footprint but, it is important to bear in mind the following implications of creating models using this <em>direct</em> mode:</p><ul><li>When <a href="#JuMP.backend"><code>backend</code></a> does not support an operation, such as modifying constraints or adding variables/constraints after solving, an error is thrown. For models created using the <a href="#JuMP.Model"><code>Model</code></a> constructor, such situations can be dealt with by storing the modifications in a cache and loading them into the optimizer when <code>optimize!</code> is called.</li><li>No constraint bridging is supported by default.</li><li>The optimizer used cannot be changed the model is constructed.</li><li>The model created cannot be copied.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L188-L213">source</a></section><section><div><pre><code class="language-julia hljs">direct_generic_model(::Type{T}, factory::MOI.OptimizerWithAttributes)</code></pre><p>Create a <a href="#JuMP.direct_generic_model"><code>direct_generic_model</code></a> using <code>factory</code>, a <code>MOI.OptimizerWithAttributes</code> object created by <a href="#JuMP.optimizer_with_attributes"><code>optimizer_with_attributes</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; optimizer = optimizer_with_attributes(
+           HiGHS.Optimizer,
+           &quot;presolve&quot; =&gt; &quot;off&quot;,
+           MOI.Silent() =&gt; true,
+       );
+
+julia&gt; model = direct_generic_model(Float64, optimizer)
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: DIRECT
+Solver name: HiGHS</code></pre><p>is equivalent to:</p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = direct_generic_model(Float64, HiGHS.Optimizer())
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: DIRECT
+Solver name: HiGHS
+
+julia&gt; set_attribute(model, &quot;presolve&quot;, &quot;off&quot;)
+
+julia&gt; set_attribute(model, MOI.Silent(), true)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L233-L272">source</a></section></article><h2 id="direct_model"><a class="docs-heading-anchor" href="#direct_model"><code>direct_model</code></a><a id="direct_model-1"></a><a class="docs-heading-anchor-permalink" href="#direct_model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.direct_model" href="#JuMP.direct_model"><code>JuMP.direct_model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">direct_model(backend::MOI.ModelLike)</code></pre><p>Return a new JuMP model using <a href="#JuMP.backend"><code>backend</code></a> to store the model and solve it.</p><p>As opposed to the <a href="#JuMP.Model"><code>Model</code></a> constructor, no cache of the model is stored outside of <a href="#JuMP.backend"><code>backend</code></a> and no bridges are automatically applied to <a href="#JuMP.backend"><code>backend</code></a>.</p><p><strong>Notes</strong></p><p>The absence of a cache reduces the memory footprint but, it is important to bear in mind the following implications of creating models using this <em>direct</em> mode:</p><ul><li>When <a href="#JuMP.backend"><code>backend</code></a> does not support an operation, such as modifying constraints or adding variables/constraints after solving, an error is thrown. For models created using the <a href="#JuMP.Model"><code>Model</code></a> constructor, such situations can be dealt with by storing the modifications in a cache and loading them into the optimizer when <code>optimize!</code> is called.</li><li>No constraint bridging is supported by default.</li><li>The optimizer used cannot be changed the model is constructed.</li><li>The model created cannot be copied.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L313-L335">source</a></section><section><div><pre><code class="language-julia hljs">direct_model(factory::MOI.OptimizerWithAttributes)</code></pre><p>Create a <a href="#JuMP.direct_model"><code>direct_model</code></a> using <code>factory</code>, a <code>MOI.OptimizerWithAttributes</code> object created by <a href="#JuMP.optimizer_with_attributes"><code>optimizer_with_attributes</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; optimizer = optimizer_with_attributes(
+           HiGHS.Optimizer,
+           &quot;presolve&quot; =&gt; &quot;off&quot;,
+           MOI.Silent() =&gt; true,
+       );
+
+julia&gt; model = direct_model(optimizer)
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: DIRECT
+Solver name: HiGHS</code></pre><p>is equivalent to:</p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = direct_model(HiGHS.Optimizer())
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: DIRECT
+Solver name: HiGHS
+
+julia&gt; set_attribute(model, &quot;presolve&quot;, &quot;off&quot;)
+
+julia&gt; set_attribute(model, MOI.Silent(), true)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L338-L377">source</a></section></article><h2 id="drop_zeros!"><a class="docs-heading-anchor" href="#drop_zeros!"><code>drop_zeros!</code></a><a id="drop_zeros!-1"></a><a class="docs-heading-anchor-permalink" href="#drop_zeros!" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.drop_zeros!" href="#JuMP.drop_zeros!"><code>JuMP.drop_zeros!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">drop_zeros!(expr::GenericAffExpr)</code></pre><p>Remove terms in the affine expression with <code>0</code> coefficients.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L255-L259">source</a></section><section><div><pre><code class="language-julia hljs">drop_zeros!(expr::GenericQuadExpr)</code></pre><p>Remove terms in the quadratic expression with <code>0</code> coefficients.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L166-L170">source</a></section></article><h2 id="dual"><a class="docs-heading-anchor" href="#dual"><code>dual</code></a><a id="dual-1"></a><a class="docs-heading-anchor-permalink" href="#dual" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.dual" href="#JuMP.dual"><code>JuMP.dual</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dual(con_ref::ConstraintRef; result::Int = 1)</code></pre><p>Return the dual value of constraint <code>con_ref</code> associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>Use <code>has_dual</code> to check if a result exists before asking for values.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>, <a href="#JuMP.shadow_price"><code>shadow_price</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1041-L1050">source</a></section><section><div><pre><code class="language-julia hljs">dual(c::NonlinearConstraintRef)</code></pre><p>Return the dual of the nonlinear constraint <code>c</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L562-L566">source</a></section></article><h2 id="dual_objective_value"><a class="docs-heading-anchor" href="#dual_objective_value"><code>dual_objective_value</code></a><a id="dual_objective_value-1"></a><a class="docs-heading-anchor-permalink" href="#dual_objective_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.dual_objective_value" href="#JuMP.dual_objective_value"><code>JuMP.dual_objective_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dual_objective_value(model::GenericModel; result::Int = 1)</code></pre><p>Return the value of the objective of the dual problem associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>Throws <code>MOI.UnsupportedAttribute{MOI.DualObjectiveValue}</code> if the solver does not support this attribute.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L57-L67">source</a></section></article><h2 id="dual_shape"><a class="docs-heading-anchor" href="#dual_shape"><code>dual_shape</code></a><a id="dual_shape-1"></a><a class="docs-heading-anchor-permalink" href="#dual_shape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.dual_shape" href="#JuMP.dual_shape"><code>JuMP.dual_shape</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dual_shape(shape::AbstractShape)::AbstractShape</code></pre><p>Returns the shape of the dual space of the space of objects of shape <code>shape</code>. By default, the <code>dual_shape</code> of a shape is itself. See the examples section below for an example for which this is not the case.</p><p><strong>Example</strong></p><p>Consider polynomial constraints for which the dual is moment constraints and moment constraints for which the dual is polynomial constraints. Shapes for polynomials can be defined as follows:</p><pre><code class="language-julia hljs">struct Polynomial
+    coefficients::Vector{Float64}
+    monomials::Vector{Monomial}
+end
+struct PolynomialShape &lt;: AbstractShape
+    monomials::Vector{Monomial}
+end
+JuMP.reshape_vector(x::Vector, shape::PolynomialShape) = Polynomial(x, shape.monomials)</code></pre><p>and a shape for moments can be defined as follows:</p><pre><code class="language-julia hljs">struct Moments
+    coefficients::Vector{Float64}
+    monomials::Vector{Monomial}
+end
+struct MomentsShape &lt;: AbstractShape
+    monomials::Vector{Monomial}
+end
+JuMP.reshape_vector(x::Vector, shape::MomentsShape) = Moments(x, shape.monomials)</code></pre><p>Then <code>dual_shape</code> allows the definition of the shape of the dual of polynomial and moment constraints:</p><pre><code class="language-julia hljs">dual_shape(shape::PolynomialShape) = MomentsShape(shape.monomials)
+dual_shape(shape::MomentsShape) = PolynomialShape(shape.monomials)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/shapes.jl#L19-L58">source</a></section></article><h2 id="dual_start_value"><a class="docs-heading-anchor" href="#dual_start_value"><code>dual_start_value</code></a><a id="dual_start_value-1"></a><a class="docs-heading-anchor-permalink" href="#dual_start_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.dual_start_value" href="#JuMP.dual_start_value"><code>JuMP.dual_start_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dual_start_value(con_ref::ConstraintRef)</code></pre><p>Return the dual start value (MOI attribute <code>ConstraintDualStart</code>) of the constraint <code>con_ref</code>.</p><p>Note: If no dual start value has been set, <code>dual_start_value</code> will return <code>nothing</code>.</p><p>See also <a href="#JuMP.set_dual_start_value"><code>set_dual_start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L71-L81">source</a></section></article><h2 id="dual_status"><a class="docs-heading-anchor" href="#dual_status"><code>dual_status</code></a><a id="dual_status-1"></a><a class="docs-heading-anchor-permalink" href="#dual_status" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.dual_status" href="#JuMP.dual_status"><code>JuMP.dual_status</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dual_status(model::GenericModel; result::Int = 1)</code></pre><p>Return a <a href="../../moi/reference/models/#MathOptInterface.ResultStatusCode"><code>MOI.ResultStatusCode</code></a> describing the status of the most recent dual solution of the solver (i.e., the <a href="../../moi/reference/models/#MathOptInterface.DualStatus"><code>MOI.DualStatus</code></a> attribute) associated with the result index <code>result</code>.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L563-L571">source</a></section></article><h2 id="error_if_direct_mode"><a class="docs-heading-anchor" href="#error_if_direct_mode"><code>error_if_direct_mode</code></a><a id="error_if_direct_mode-1"></a><a class="docs-heading-anchor-permalink" href="#error_if_direct_mode" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.error_if_direct_mode" href="#JuMP.error_if_direct_mode"><code>JuMP.error_if_direct_mode</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">error_if_direct_mode(model::GenericModel, func::Symbol)</code></pre><p>Errors if <code>model</code> is in direct mode during a call from the function named <code>func</code>.</p><p>Used internally within JuMP, or by JuMP extensions who do not want to support models in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L250-L258">source</a></section></article><h2 id="fix"><a class="docs-heading-anchor" href="#fix"><code>fix</code></a><a id="fix-1"></a><a class="docs-heading-anchor-permalink" href="#fix" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.fix" href="#JuMP.fix"><code>JuMP.fix</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">fix(v::GenericVariableRef, value::Number; force::Bool = false)</code></pre><p>Fix a variable to a value. Update the fixing constraint if one exists, otherwise create a new one.</p><p>If the variable already has variable bounds and <code>force=false</code>, calling <code>fix</code> will throw an error. If <code>force=true</code>, existing variable bounds will be deleted, and the fixing constraint will be added. Note a variable will have no bounds after a call to <a href="#JuMP.unfix"><code>unfix</code></a>.</p><p>See also <a href="#JuMP.FixRef"><code>FixRef</code></a>, <a href="#JuMP.is_fixed"><code>is_fixed</code></a>, <a href="#JuMP.fix_value"><code>fix_value</code></a>, <a href="#JuMP.unfix"><code>unfix</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; is_fixed(x)
+false
+
+julia&gt; fix(x, 1.0)
+
+julia&gt; is_fixed(x)
+true</code></pre><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, 0 &lt;= x &lt;= 1);
+
+julia&gt; is_fixed(x)
+false
+
+julia&gt; fix(x, 1.0; force = true)
+
+julia&gt; is_fixed(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1045-L1088">source</a></section></article><h2 id="fix_discrete_variables"><a class="docs-heading-anchor" href="#fix_discrete_variables"><code>fix_discrete_variables</code></a><a id="fix_discrete_variables-1"></a><a class="docs-heading-anchor-permalink" href="#fix_discrete_variables" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.fix_discrete_variables" href="#JuMP.fix_discrete_variables"><code>JuMP.fix_discrete_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">fix_discrete_variables([var_value::Function = value,] model::GenericModel)</code></pre><p>Modifies <code>model</code> to convert all binary and integer variables to continuous variables with fixed bounds of <code>var_value(x)</code>.</p><p><strong>Return</strong></p><p>Returns a function that can be called without any arguments to restore the original model. The behavior of this function is undefined if additional changes are made to the affected variables in the meantime.</p><p><strong>Notes</strong></p><ul><li>An error is thrown if semi-continuous or semi-integer constraints are present (support may be added for these in the future).</li><li>All other constraints are ignored (left in place). This includes discrete constraints like SOS and indicator constraints.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Bin, start = 1);
+
+julia&gt; @variable(model, 1 &lt;= y &lt;= 10, Int, start = 2);
+
+julia&gt; @objective(model, Min, x + y);
+
+julia&gt; undo_relax = fix_discrete_variables(start_value, model);
+
+julia&gt; print(model)
+Min x + y
+Subject to
+ x = 1
+ y = 2
+
+julia&gt; undo_relax()
+
+julia&gt; print(model)
+Min x + y
+Subject to
+ y ≥ 1
+ y ≤ 10
+ y integer
+ x binary</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L2265-L2313">source</a></section></article><h2 id="fix_value"><a class="docs-heading-anchor" href="#fix_value"><code>fix_value</code></a><a id="fix_value-1"></a><a class="docs-heading-anchor-permalink" href="#fix_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.fix_value" href="#JuMP.fix_value"><code>JuMP.fix_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">fix_value(v::GenericVariableRef)</code></pre><p>Return the value to which a variable is fixed.</p><p>Error if one does not exist.</p><p>See also <a href="#JuMP.FixRef"><code>FixRef</code></a>, <a href="#JuMP.is_fixed"><code>is_fixed</code></a>, <a href="#JuMP.fix"><code>fix</code></a>, <a href="#JuMP.unfix"><code>unfix</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x == 1);
+
+julia&gt; fix_value(x)
+1.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1163-L1182">source</a></section></article><h2 id="flatten!"><a class="docs-heading-anchor" href="#flatten!"><code>flatten!</code></a><a id="flatten!-1"></a><a class="docs-heading-anchor-permalink" href="#flatten!" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.flatten!" href="#JuMP.flatten!"><code>JuMP.flatten!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">flatten!(expr::GenericNonlinearExpr)</code></pre><p>Flatten a nonlinear expression in-place by lifting nested <code>+</code> and <code>*</code> nodes into a single n-ary operation.</p><p><strong>Motivation</strong></p><p>Nonlinear expressions created using operator overloading can be deeply nested and unbalanced. For example, <code>prod(x for i in 1:4)</code> creates <code>*(x, *(x, *(x, x)))</code> instead of the more preferable <code>*(x, x, x, x)</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; y = prod(x for i in 1:4)
+((x²) * x) * x
+
+julia&gt; flatten!(y)
+(x²) * x * x
+
+julia&gt; flatten!(sin(prod(x for i in 1:4)))
+sin((x²) * x * x)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp_expr.jl#L406-L435">source</a></section></article><h2 id="function_string"><a class="docs-heading-anchor" href="#function_string"><code>function_string</code></a><a id="function_string-1"></a><a class="docs-heading-anchor-permalink" href="#function_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.function_string" href="#JuMP.function_string"><code>JuMP.function_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">function_string(
+    mode::MIME,
+    func::Union{JuMP.AbstractJuMPScalar,Vector{&lt;:JuMP.AbstractJuMPScalar}},
+)</code></pre><p>Return a <code>String</code> representing the function <code>func</code> using print mode <code>mode</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L562-L569">source</a></section></article><h2 id="get_attribute"><a class="docs-heading-anchor" href="#get_attribute"><code>get_attribute</code></a><a id="get_attribute-1"></a><a class="docs-heading-anchor-permalink" href="#get_attribute" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.get_attribute" href="#JuMP.get_attribute"><code>JuMP.get_attribute</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">get_attribute(model::GenericModel, attr::MOI.AbstractModelAttribute)
+get_attribute(x::GenericVariableRef, attr::MOI.AbstractVariableAttribute)
+get_attribute(cr::ConstraintRef, attr::MOI.AbstractConstraintAttribute)</code></pre><p>Get the value of a solver-specifc attribute <code>attr</code>.</p><p>This is equivalent to calling <a href="../../moi/reference/models/#MathOptInterface.get"><code>MOI.get</code></a> with the associated MOI model and, for variables and constraints, with the associated <a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> or <a href="../../moi/reference/constraints/#MathOptInterface.ConstraintIndex"><code>MOI.ConstraintIndex</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, c, 2 * x &lt;= 1)
+c : 2 x ≤ 1
+
+julia&gt; get_attribute(model, MOI.Name())
+&quot;&quot;
+
+julia&gt; get_attribute(x, MOI.VariableName())
+&quot;x&quot;
+
+julia&gt; get_attribute(c, MOI.ConstraintName())
+&quot;c&quot;</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L760-L791">source</a></section><section><div><pre><code class="language-julia hljs">get_attribute(
+    model::Union{GenericModel,MOI.OptimizerWithAttributes},
+    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},
+)</code></pre><p>Get the value of a solver-specifc attribute <code>attr</code>.</p><p>This is equivalent to calling <a href="../../moi/reference/models/#MathOptInterface.get"><code>MOI.get</code></a> with the associated MOI model.</p><p>If <code>attr</code> is an <code>AbstractString</code>, it is converted to <a href="../../moi/reference/models/#MathOptInterface.RawOptimizerAttribute"><code>MOI.RawOptimizerAttribute</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; opt = optimizer_with_attributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; true);
+
+julia&gt; model = Model(opt);
+
+julia&gt; get_attribute(model, &quot;output_flag&quot;)
+true
+
+julia&gt; get_attribute(model, MOI.RawOptimizerAttribute(&quot;output_flag&quot;))
+true
+
+julia&gt; get_attribute(opt, &quot;output_flag&quot;)
+true
+
+julia&gt; get_attribute(opt, MOI.RawOptimizerAttribute(&quot;output_flag&quot;))
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L807-L841">source</a></section></article><h2 id="get_optimizer_attribute"><a class="docs-heading-anchor" href="#get_optimizer_attribute"><code>get_optimizer_attribute</code></a><a id="get_optimizer_attribute-1"></a><a class="docs-heading-anchor-permalink" href="#get_optimizer_attribute" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.get_optimizer_attribute" href="#JuMP.get_optimizer_attribute"><code>JuMP.get_optimizer_attribute</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">get_optimizer_attribute(
+    model::Union{GenericModel,MOI.OptimizerWithAttributes},
+    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},
+)</code></pre><p>Return the value associated with the solver-specific attribute <code>attr</code>.</p><p>If <code>attr</code> is an <code>AbstractString</code>, this is equivalent to <code>get_optimizer_attribute(model, MOI.RawOptimizerAttribute(name))</code>.</p><div class="admonition is-compat"><header class="admonition-header">Compat</header><div class="admonition-body"><p>This method will remain in all v1.X releases of JuMP, but it may be removed in a future v2.0 release. We recommend using <a href="#JuMP.get_attribute"><code>get_attribute</code></a> instead.</p></div></div><p>See also: <a href="#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a>, <a href="#JuMP.set_optimizer_attributes"><code>set_optimizer_attributes</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; model = Model(Ipopt.Optimizer);
+
+julia&gt; get_optimizer_attribute(model, MOI.Silent())
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L124-L151">source</a></section></article><h2 id="has_duals"><a class="docs-heading-anchor" href="#has_duals"><code>has_duals</code></a><a id="has_duals-1"></a><a class="docs-heading-anchor-permalink" href="#has_duals" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.has_duals" href="#JuMP.has_duals"><code>JuMP.has_duals</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">has_duals(model::GenericModel; result::Int = 1)</code></pre><p>Return <code>true</code> if the solver has a dual solution in result index <code>result</code> available to query, otherwise return <code>false</code>.</p><p>See also <a href="#JuMP.dual"><code>dual</code></a>, <a href="#JuMP.shadow_price"><code>shadow_price</code></a>, and <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1029-L1036">source</a></section></article><h2 id="has_lower_bound"><a class="docs-heading-anchor" href="#has_lower_bound"><code>has_lower_bound</code></a><a id="has_lower_bound-1"></a><a class="docs-heading-anchor-permalink" href="#has_lower_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.has_lower_bound" href="#JuMP.has_lower_bound"><code>JuMP.has_lower_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">has_lower_bound(v::GenericVariableRef)</code></pre><p>Return <code>true</code> if <code>v</code> has a lower bound. If <code>true</code>, the lower bound can be queried with <a href="#JuMP.lower_bound"><code>lower_bound</code></a>.</p><p>See also <a href="#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, <a href="#JuMP.lower_bound"><code>lower_bound</code></a>, <a href="#JuMP.set_lower_bound"><code>set_lower_bound</code></a>, <a href="#JuMP.delete_lower_bound"><code>delete_lower_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 1.0);
+
+julia&gt; has_lower_bound(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L661-L680">source</a></section></article><h2 id="has_start_value"><a class="docs-heading-anchor" href="#has_start_value"><code>has_start_value</code></a><a id="has_start_value-1"></a><a class="docs-heading-anchor-permalink" href="#has_start_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.has_start_value" href="#JuMP.has_start_value"><code>JuMP.has_start_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">has_start_value(variable::AbstractVariableRef)</code></pre><p>Return <code>true</code> if the variable has a start value set otherwise return <code>false</code>.</p><p>See also <a href="#JuMP.set_start_value"><code>set_start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1645-L1651">source</a></section></article><h2 id="has_upper_bound"><a class="docs-heading-anchor" href="#has_upper_bound"><code>has_upper_bound</code></a><a id="has_upper_bound-1"></a><a class="docs-heading-anchor-permalink" href="#has_upper_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.has_upper_bound" href="#JuMP.has_upper_bound"><code>JuMP.has_upper_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">has_upper_bound(v::GenericVariableRef)</code></pre><p>Return <code>true</code> if <code>v</code> has a upper bound. If <code>true</code>, the upper bound can be queried with <a href="#JuMP.upper_bound"><code>upper_bound</code></a>.</p><p>See also <a href="#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, <a href="#JuMP.upper_bound"><code>upper_bound</code></a>, <a href="#JuMP.set_upper_bound"><code>set_upper_bound</code></a>, <a href="#JuMP.delete_upper_bound"><code>delete_upper_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &lt;= 1.0);
+
+julia&gt; has_upper_bound(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L834-L853">source</a></section></article><h2 id="has_values"><a class="docs-heading-anchor" href="#has_values"><code>has_values</code></a><a id="has_values-1"></a><a class="docs-heading-anchor-permalink" href="#has_values" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.has_values" href="#JuMP.has_values"><code>JuMP.has_values</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">has_values(model::GenericModel; result::Int = 1)</code></pre><p>Return <code>true</code> if the solver has a primal solution in result index <code>result</code> available to query, otherwise return <code>false</code>.</p><p>See also <a href="#JuMP.value"><code>value</code></a> and <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1705-L1712">source</a></section></article><h2 id="in_set_string"><a class="docs-heading-anchor" href="#in_set_string"><code>in_set_string</code></a><a id="in_set_string-1"></a><a class="docs-heading-anchor-permalink" href="#in_set_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.in_set_string" href="#JuMP.in_set_string"><code>JuMP.in_set_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">in_set_string(mode::MIME, set)</code></pre><p>Return a <code>String</code> representing the membership to the set <code>set</code> using print mode <code>mode</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L722-L727">source</a></section></article><h2 id="index"><a class="docs-heading-anchor" href="#index"><code>index</code></a><a id="index-1"></a><a class="docs-heading-anchor-permalink" href="#index" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.index" href="#JuMP.index"><code>JuMP.index</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">index(cr::ConstraintRef)::MOI.ConstraintIndex</code></pre><p>Return the index of the constraint that corresponds to <code>cr</code> in the MOI backend.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L32-L36">source</a></section><section><div><pre><code class="language-julia hljs">index(v::GenericVariableRef)::MOI.VariableIndex</code></pre><p>Return the index of the variable that corresponds to <code>v</code> in the MOI backend.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L479-L483">source</a></section><section><div><pre><code class="language-julia hljs">index(p::NonlinearParameter)::MOI.Nonlinear.ParameterIndex</code></pre><p>Return the index of the nonlinear parameter associated with <code>p</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L221-L225">source</a></section><section><div><pre><code class="language-julia hljs">index(ex::NonlinearExpression)::MOI.Nonlinear.ExpressionIndex</code></pre><p>Return the index of the nonlinear expression associated with <code>ex</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L302-L306">source</a></section></article><h2 id="is_binary"><a class="docs-heading-anchor" href="#is_binary"><code>is_binary</code></a><a id="is_binary-1"></a><a class="docs-heading-anchor-permalink" href="#is_binary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.is_binary" href="#JuMP.is_binary"><code>JuMP.is_binary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_binary(v::GenericVariableRef)</code></pre><p>Return <code>true</code> if <code>v</code> is constrained to be binary.</p><p>See also <a href="#JuMP.BinaryRef"><code>BinaryRef</code></a>, <a href="#JuMP.set_binary"><code>set_binary</code></a>, <a href="#JuMP.unset_binary"><code>unset_binary</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Bin);
+
+julia&gt; is_binary(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1355-L1372">source</a></section></article><h2 id="is_fixed"><a class="docs-heading-anchor" href="#is_fixed"><code>is_fixed</code></a><a id="is_fixed-1"></a><a class="docs-heading-anchor-permalink" href="#is_fixed" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.is_fixed" href="#JuMP.is_fixed"><code>JuMP.is_fixed</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_fixed(v::GenericVariableRef)</code></pre><p>Return <code>true</code> if <code>v</code> is a fixed variable. If <code>true</code>, the fixed value can be queried with <a href="#JuMP.fix_value"><code>fix_value</code></a>.</p><p>See also <a href="#JuMP.FixRef"><code>FixRef</code></a>, <a href="#JuMP.fix_value"><code>fix_value</code></a>, <a href="#JuMP.fix"><code>fix</code></a>, <a href="#JuMP.unfix"><code>unfix</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; is_fixed(x)
+false
+
+julia&gt; fix(x, 1.0)
+
+julia&gt; is_fixed(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1009-L1032">source</a></section></article><h2 id="is_integer"><a class="docs-heading-anchor" href="#is_integer"><code>is_integer</code></a><a id="is_integer-1"></a><a class="docs-heading-anchor-permalink" href="#is_integer" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.is_integer" href="#JuMP.is_integer"><code>JuMP.is_integer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_integer(v::GenericVariableRef)</code></pre><p>Return <code>true</code> if <code>v</code> is constrained to be integer.</p><p>See also <a href="#JuMP.IntegerRef"><code>IntegerRef</code></a>, <a href="#JuMP.set_integer"><code>set_integer</code></a>, <a href="#JuMP.unset_integer"><code>unset_integer</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; is_integer(x)
+false
+
+julia&gt; set_integer(x)
+
+julia&gt; is_integer(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1218-L1240">source</a></section></article><h2 id="is_parameter"><a class="docs-heading-anchor" href="#is_parameter"><code>is_parameter</code></a><a id="is_parameter-1"></a><a class="docs-heading-anchor-permalink" href="#is_parameter" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.is_parameter" href="#JuMP.is_parameter"><code>JuMP.is_parameter</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_parameter(x::GenericVariableRef)::Bool</code></pre><p>Return <code>true</code> if <code>x</code> is constrained to be a parameter.</p><p>See also <a href="#JuMP.ParameterRef"><code>ParameterRef</code></a>, <a href="#JuMP.set_parameter_value"><code>set_parameter_value</code></a>, <a href="#JuMP.parameter_value"><code>parameter_value</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, p in Parameter(2))
+p
+
+julia&gt; is_parameter(p)
+true
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; is_parameter(x)
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1527-L1552">source</a></section></article><h2 id="is_valid"><a class="docs-heading-anchor" href="#is_valid"><code>is_valid</code></a><a id="is_valid-1"></a><a class="docs-heading-anchor-permalink" href="#is_valid" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.is_valid" href="#JuMP.is_valid"><code>JuMP.is_valid</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_valid(model::GenericModel, con_ref::ConstraintRef{&lt;:AbstractModel})</code></pre><p>Return <code>true</code> if <code>constraint_ref</code> refers to a valid constraint in <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L435-L439">source</a></section><section><div><pre><code class="language-julia hljs">is_valid(model::GenericModel, variable_ref::GenericVariableRef)</code></pre><p>Return <code>true</code> if <code>variable</code> refers to a valid variable in <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L458-L462">source</a></section><section><div><pre><code class="language-julia hljs">is_valid(model::Model, c::NonlinearConstraintRef)</code></pre><p>Return <code>true</code> if <code>c</code> refers to a valid nonlinear constraint in <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L473-L477">source</a></section></article><h2 id="isequal_canonical"><a class="docs-heading-anchor" href="#isequal_canonical"><code>isequal_canonical</code></a><a id="isequal_canonical-1"></a><a class="docs-heading-anchor-permalink" href="#isequal_canonical" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.isequal_canonical" href="#JuMP.isequal_canonical"><code>JuMP.isequal_canonical</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">isequal_canonical(
+    aff::GenericAffExpr{C,V},
+    other::GenericAffExpr{C,V}
+) where {C,V}</code></pre><p>Return <code>true</code> if <code>aff</code> is equal to <code>other</code> after dropping zeros and disregarding the order. Mainly useful for testing.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L540-L548">source</a></section></article><h2 id="jump_function"><a class="docs-heading-anchor" href="#jump_function"><code>jump_function</code></a><a id="jump_function-1"></a><a class="docs-heading-anchor-permalink" href="#jump_function" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.jump_function" href="#JuMP.jump_function"><code>JuMP.jump_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">jump_function(x)</code></pre><p>Given an MathOptInterface object <code>x</code>, return the JuMP equivalent.</p><p>See also: <a href="#JuMP.moi_function"><code>moi_function</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L682-L688">source</a></section></article><h2 id="jump_function_type"><a class="docs-heading-anchor" href="#jump_function_type"><code>jump_function_type</code></a><a id="jump_function_type-1"></a><a class="docs-heading-anchor-permalink" href="#jump_function_type" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.jump_function_type" href="#JuMP.jump_function_type"><code>JuMP.jump_function_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">jump_function_type(::Type{T}) where {T}</code></pre><p>Given an MathOptInterface object type <code>T</code>, return the JuMP equivalent.</p><p>See also: <a href="#JuMP.moi_function_type"><code>moi_function_type</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L691-L697">source</a></section></article><h2 id="latex_formulation"><a class="docs-heading-anchor" href="#latex_formulation"><code>latex_formulation</code></a><a id="latex_formulation-1"></a><a class="docs-heading-anchor-permalink" href="#latex_formulation" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.latex_formulation" href="#JuMP.latex_formulation"><code>JuMP.latex_formulation</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">latex_formulation(model::AbstractModel)</code></pre><p>Wrap <code>model</code> in a type so that it can be pretty-printed as <code>text/latex</code> in a notebook like IJulia, or in Documenter.</p><p>To render the model, end the cell with <code>latex_formulation(model)</code>, or call <code>display(latex_formulation(model))</code> in to force the display of the model from inside a function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L17-L26">source</a></section></article><h2 id="linear_terms"><a class="docs-heading-anchor" href="#linear_terms"><code>linear_terms</code></a><a id="linear_terms-1"></a><a class="docs-heading-anchor-permalink" href="#linear_terms" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.linear_terms" href="#JuMP.linear_terms"><code>JuMP.linear_terms</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">linear_terms(aff::GenericAffExpr{C, V})</code></pre><p>Provides an iterator over coefficient-variable tuples <code>(a_i::C, x_i::V)</code> in the linear part of the affine expression.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L378-L383">source</a></section><section><div><pre><code class="language-julia hljs">linear_terms(quad::GenericQuadExpr{C, V})</code></pre><p>Provides an iterator over tuples <code>(coefficient::C, variable::V)</code> in the linear part of the quadratic expression.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L266-L271">source</a></section></article><h2 id="list_of_constraint_types"><a class="docs-heading-anchor" href="#list_of_constraint_types"><code>list_of_constraint_types</code></a><a id="list_of_constraint_types-1"></a><a class="docs-heading-anchor-permalink" href="#list_of_constraint_types" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.list_of_constraint_types" href="#JuMP.list_of_constraint_types"><code>JuMP.list_of_constraint_types</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">list_of_constraint_types(model::GenericModel)::Vector{Tuple{Type,Type}}</code></pre><p>Return a list of tuples of the form <code>(F, S)</code> where <code>F</code> is a JuMP function type and <code>S</code> is an MOI set type such that <code>all_constraints(model, F, S)</code> returns a nonempty list.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 0, Bin);
+
+julia&gt; @constraint(model, 2x &lt;= 1);
+
+julia&gt; list_of_constraint_types(model)
+3-element Vector{Tuple{Type, Type}}:
+ (AffExpr, MathOptInterface.LessThan{Float64})
+ (VariableRef, MathOptInterface.GreaterThan{Float64})
+ (VariableRef, MathOptInterface.ZeroOne)</code></pre><p><strong>Performance considerations</strong></p><p>Iterating over the list of function and set types is a type-unstable operation. Consider using a function barrier. See the <a href="../../developers/extensions/#Performance-tips-for-extensions">Performance tips for extensions</a> section of the documentation for more details.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1301-L1329">source</a></section></article><h2 id="lower_bound"><a class="docs-heading-anchor" href="#lower_bound"><code>lower_bound</code></a><a id="lower_bound-1"></a><a class="docs-heading-anchor-permalink" href="#lower_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.lower_bound" href="#JuMP.lower_bound"><code>JuMP.lower_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">lower_bound(v::GenericVariableRef)</code></pre><p>Return the lower bound of a variable. Error if one does not exist.</p><p>See also <a href="#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, <a href="#JuMP.has_lower_bound"><code>has_lower_bound</code></a>, <a href="#JuMP.set_lower_bound"><code>set_lower_bound</code></a>, <a href="#JuMP.delete_lower_bound"><code>delete_lower_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 1.0);
+
+julia&gt; lower_bound(x)
+1.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L808-L826">source</a></section></article><h2 id="lp_sensitivity_report"><a class="docs-heading-anchor" href="#lp_sensitivity_report"><code>lp_sensitivity_report</code></a><a id="lp_sensitivity_report-1"></a><a class="docs-heading-anchor-permalink" href="#lp_sensitivity_report" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.lp_sensitivity_report" href="#JuMP.lp_sensitivity_report"><code>JuMP.lp_sensitivity_report</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">lp_sensitivity_report(model::GenericModel{T}; atol::T = Base.rtoldefault(T))::SensitivityReport{T} where {T}</code></pre><p>Given a linear program <code>model</code> with a current optimal basis, return a <a href="#JuMP.SensitivityReport"><code>SensitivityReport</code></a> object, which maps:</p><ul><li>Every variable reference to a tuple <code>(d_lo, d_hi)::Tuple{T,T}</code>, explaining how much the objective coefficient of the corresponding variable can change by, such that the original basis remains optimal.</li><li>Every constraint reference to a tuple <code>(d_lo, d_hi)::Tuple{T,T}</code>, explaining how much the right-hand side of the corresponding constraint can change by, such that the basis remains optimal.</li></ul><p>Both tuples are relative, rather than absolute. So given a objective coefficient of <code>1.0</code> and a tuple <code>(-0.5, 0.5)</code>, the objective coefficient can range between <code>1.0 - 0.5</code> an <code>1.0 + 0.5</code>.</p><p><code>atol</code> is the primal/dual optimality tolerance, and should match the tolerance of the solver used to compute the basis.</p><p>Note: interval constraints are NOT supported.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, -1 &lt;= x &lt;= 2)
+x
+
+julia&gt; @objective(model, Min, x)
+x
+
+julia&gt; optimize!(model)
+
+julia&gt; report = lp_sensitivity_report(model; atol = 1e-7);
+
+julia&gt; dx_lo, dx_hi = report[x]
+(-1.0, Inf)
+
+julia&gt; println(
+           &quot;The objective coefficient of `x` can decrease by $dx_lo or &quot; *
+           &quot;increase by $dx_hi.&quot;
+       )
+The objective coefficient of `x` can decrease by -1.0 or increase by Inf.
+
+julia&gt; dRHS_lo, dRHS_hi = report[LowerBoundRef(x)]
+(-Inf, 3.0)
+
+julia&gt; println(
+           &quot;The lower bound of `x` can decrease by $dRHS_lo or increase &quot; *
+           &quot;by $dRHS_hi.&quot;
+       )
+The lower bound of `x` can decrease by -Inf or increase by 3.0.</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/lp_sensitivity2.jl#L19-L78">source</a></section></article><h2 id="map_coefficients"><a class="docs-heading-anchor" href="#map_coefficients"><code>map_coefficients</code></a><a id="map_coefficients-1"></a><a class="docs-heading-anchor-permalink" href="#map_coefficients" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.map_coefficients" href="#JuMP.map_coefficients"><code>JuMP.map_coefficients</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">map_coefficients(f::Function, a::GenericAffExpr)</code></pre><p>Apply <code>f</code> to the coefficients and constant term of an <a href="#JuMP.GenericAffExpr"><code>GenericAffExpr</code></a> <code>a</code> and return a new expression.</p><p>See also: <a href="#JuMP.map_coefficients_inplace!"><code>map_coefficients_inplace!</code></a></p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; a = GenericAffExpr(1.0, x =&gt; 1.0)
+x + 1
+
+julia&gt; map_coefficients(c -&gt; 2 * c, a)
+2 x + 2
+
+julia&gt; a
+x + 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L305-L329">source</a></section><section><div><pre><code class="language-julia hljs">map_coefficients(f::Function, a::GenericQuadExpr)</code></pre><p>Apply <code>f</code> to the coefficients and constant term of an <a href="#JuMP.GenericQuadExpr"><code>GenericQuadExpr</code></a> <code>a</code> and return a new expression.</p><p>See also: <a href="#JuMP.map_coefficients_inplace!"><code>map_coefficients_inplace!</code></a></p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; a = @expression(model, x^2 + x + 1)
+x² + x + 1
+
+julia&gt; map_coefficients(c -&gt; 2 * c, a)
+2 x² + 2 x + 2
+
+julia&gt; a
+x² + x + 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L211-L235">source</a></section></article><h2 id="map_coefficients_inplace!"><a class="docs-heading-anchor" href="#map_coefficients_inplace!"><code>map_coefficients_inplace!</code></a><a id="map_coefficients_inplace!-1"></a><a class="docs-heading-anchor-permalink" href="#map_coefficients_inplace!" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.map_coefficients_inplace!" href="#JuMP.map_coefficients_inplace!"><code>JuMP.map_coefficients_inplace!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">map_coefficients_inplace!(f::Function, a::GenericAffExpr)</code></pre><p>Apply <code>f</code> to the coefficients and constant term of an <a href="#JuMP.GenericAffExpr"><code>GenericAffExpr</code></a> <code>a</code> and update them in-place.</p><p>See also: <a href="#JuMP.map_coefficients"><code>map_coefficients</code></a></p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; a = GenericAffExpr(1.0, x =&gt; 1.0)
+x + 1
+
+julia&gt; map_coefficients_inplace!(c -&gt; 2 * c, a)
+2 x + 2
+
+julia&gt; a
+2 x + 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L271-L295">source</a></section><section><div><pre><code class="language-julia hljs">map_coefficients_inplace!(f::Function, a::GenericQuadExpr)</code></pre><p>Apply <code>f</code> to the coefficients and constant term of an <a href="#JuMP.GenericQuadExpr"><code>GenericQuadExpr</code></a> <code>a</code> and update them in-place.</p><p>See also: <a href="#JuMP.map_coefficients"><code>map_coefficients</code></a></p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; a = @expression(model, x^2 + x + 1)
+x² + x + 1
+
+julia&gt; map_coefficients_inplace!(c -&gt; 2 * c, a)
+2 x² + 2 x + 2
+
+julia&gt; a
+2 x² + 2 x + 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L177-L201">source</a></section></article><h2 id="mode"><a class="docs-heading-anchor" href="#mode"><code>mode</code></a><a id="mode-1"></a><a class="docs-heading-anchor-permalink" href="#mode" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.mode" href="#JuMP.mode"><code>JuMP.mode</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">mode(model::GenericModel)</code></pre><p>Return the <a href="#JuMP.ModelMode"><code>ModelMode</code></a> (<a href="#JuMP.DIRECT"><code>DIRECT</code></a>, <a href="#JuMP.AUTOMATIC"><code>AUTOMATIC</code></a>, or <a href="#JuMP.MANUAL"><code>MANUAL</code></a>) of <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L528-L533">source</a></section></article><h2 id="model_convert"><a class="docs-heading-anchor" href="#model_convert"><code>model_convert</code></a><a id="model_convert-1"></a><a class="docs-heading-anchor-permalink" href="#model_convert" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.model_convert" href="#JuMP.model_convert"><code>JuMP.model_convert</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">model_convert(
+    model::AbstractModel,
+    rhs::Union{
+        AbstractConstraint,
+        Number,
+        AbstractJuMPScalar,
+        MOI.AbstractSet,
+    },
+)</code></pre><p>Convert the coefficients and constants of functions and sets in the <code>rhs</code> to the coefficient type <code>value_type(typeof(model))</code>.</p><p><strong>Purpose</strong></p><p>Creating and adding a constraint is a two-step process. The first step calls <a href="#JuMP.build_constraint"><code>build_constraint</code></a>, and the result of that is passed to <a href="#JuMP.add_constraint"><code>add_constraint</code></a>.</p><p>However, because <a href="#JuMP.build_constraint"><code>build_constraint</code></a> does not take the <code>model</code> as an argument, the coefficients and constants of the function or set might be different than <code>value_type(typeof(model))</code>.</p><p>Therefore, the result of <a href="#JuMP.build_constraint"><code>build_constraint</code></a> is converted in a call to <code>model_convert</code> before the result is passed to <a href="#JuMP.add_constraint"><code>add_constraint</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L1126-L1152">source</a></section></article><h2 id="model_string"><a class="docs-heading-anchor" href="#model_string"><code>model_string</code></a><a id="model_string-1"></a><a class="docs-heading-anchor-permalink" href="#model_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.model_string" href="#JuMP.model_string"><code>JuMP.model_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">model_string(mode::MIME, model::AbstractModel)</code></pre><p>Return a <code>String</code> representation of <code>model</code> given the <code>mode</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L355-L359">source</a></section></article><h2 id="moi_function"><a class="docs-heading-anchor" href="#moi_function"><code>moi_function</code></a><a id="moi_function-1"></a><a class="docs-heading-anchor-permalink" href="#moi_function" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.moi_function" href="#JuMP.moi_function"><code>JuMP.moi_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">moi_function(x)</code></pre><p>Given a JuMP object <code>x</code>, return the MathOptInterface equivalent.</p><p>See also: <a href="#JuMP.jump_function"><code>jump_function</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L643-L649">source</a></section></article><h2 id="moi_function_type"><a class="docs-heading-anchor" href="#moi_function_type"><code>moi_function_type</code></a><a id="moi_function_type-1"></a><a class="docs-heading-anchor-permalink" href="#moi_function_type" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.moi_function_type" href="#JuMP.moi_function_type"><code>JuMP.moi_function_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">moi_function_type(::Type{T}) where {T}</code></pre><p>Given a JuMP object type <code>T</code>, return the MathOptInterface equivalent.</p><p>See also: <a href="#JuMP.jump_function_type"><code>jump_function_type</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L673-L679">source</a></section></article><h2 id="moi_set"><a class="docs-heading-anchor" href="#moi_set"><code>moi_set</code></a><a id="moi_set-1"></a><a class="docs-heading-anchor-permalink" href="#moi_set" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.moi_set" href="#JuMP.moi_set"><code>JuMP.moi_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">moi_set(constraint::AbstractConstraint)</code></pre><p>Return the set of the constraint <code>constraint</code> in the function-in-set form as a <code>MathOptInterface.AbstractSet</code>.</p><pre><code class="nohighlight hljs">moi_set(s::AbstractVectorSet, dim::Int)</code></pre><p>Returns the MOI set of dimension <code>dim</code> corresponding to the JuMP set <code>s</code>.</p><pre><code class="nohighlight hljs">moi_set(s::AbstractScalarSet)</code></pre><p>Returns the MOI set corresponding to the JuMP set <code>s</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L554-L567">source</a></section></article><h2 id="name"><a class="docs-heading-anchor" href="#name"><code>name</code></a><a id="name-1"></a><a class="docs-heading-anchor-permalink" href="#name" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.name" href="#JuMP.name"><code>JuMP.name</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">name(con_ref::ConstraintRef)</code></pre><p>Get a constraint&#39;s name attribute.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L234-L238">source</a></section><section><div><pre><code class="language-julia hljs">name(v::GenericVariableRef)::String</code></pre><p>Get a variable&#39;s name attribute.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L521-L525">source</a></section><section><div><pre><code class="language-julia hljs">name(model::AbstractModel)</code></pre><p>Return the <a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name</code></a> attribute of <code>model</code>&#39;s <a href="#JuMP.backend"><code>backend</code></a>, or a default if empty.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L178-L183">source</a></section></article><h2 id="node_count"><a class="docs-heading-anchor" href="#node_count"><code>node_count</code></a><a id="node_count-1"></a><a class="docs-heading-anchor-permalink" href="#node_count" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.node_count" href="#JuMP.node_count"><code>JuMP.node_count</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">node_count(model::GenericModel)</code></pre><p>Gets the total number of branch-and-bound nodes explored during the most recent optimization in a Mixed Integer Program.</p><p>Solvers must implement <code>MOI.NodeCount()</code> to use this function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L612-L619">source</a></section></article><h2 id="nonlinear_constraint_string"><a class="docs-heading-anchor" href="#nonlinear_constraint_string"><code>nonlinear_constraint_string</code></a><a id="nonlinear_constraint_string-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_constraint_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.nonlinear_constraint_string" href="#JuMP.nonlinear_constraint_string"><code>JuMP.nonlinear_constraint_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">nonlinear_constraint_string(
+    model::GenericModel,
+    mode::MIME,
+    c::_NonlinearConstraint,
+)</code></pre><p>Return a string representation of the nonlinear constraint <code>c</code> belonging to <code>model</code>, given the <code>mode</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L387-L396">source</a></section></article><h2 id="nonlinear_dual_start_value"><a class="docs-heading-anchor" href="#nonlinear_dual_start_value"><code>nonlinear_dual_start_value</code></a><a id="nonlinear_dual_start_value-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_dual_start_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.nonlinear_dual_start_value" href="#JuMP.nonlinear_dual_start_value"><code>JuMP.nonlinear_dual_start_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">nonlinear_dual_start_value(model::Model)</code></pre><p>Return the current value of the MOI attribute <a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlockDualStart"><code>MOI.NLPBlockDualStart</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L581-L585">source</a></section></article><h2 id="nonlinear_expr_string"><a class="docs-heading-anchor" href="#nonlinear_expr_string"><code>nonlinear_expr_string</code></a><a id="nonlinear_expr_string-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_expr_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.nonlinear_expr_string" href="#JuMP.nonlinear_expr_string"><code>JuMP.nonlinear_expr_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">nonlinear_expr_string(
+    model::GenericModel,
+    mode::MIME,
+    c::MOI.Nonlinear.Expression,
+)</code></pre><p>Return a string representation of the nonlinear expression <code>c</code> belonging to <code>model</code>, given the <code>mode</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L431-L440">source</a></section></article><h2 id="nonlinear_model"><a class="docs-heading-anchor" href="#nonlinear_model"><code>nonlinear_model</code></a><a id="nonlinear_model-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.nonlinear_model" href="#JuMP.nonlinear_model"><code>JuMP.nonlinear_model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">nonlinear_model(
+    model::GenericModel;
+    force::Bool = false,
+)::Union{MOI.Nonlinear.Model,Nothing}</code></pre><p>If <code>model</code> has nonlinear components, return a <a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Model"><code>MOI.Nonlinear.Model</code></a>, otherwise return <code>nothing</code>.</p><p>If <code>force</code>, always return a <a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Model"><code>MOI.Nonlinear.Model</code></a>, and if one does not exist for the model, create an empty one.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L6-L17">source</a></section></article><h2 id="normalized_coefficient"><a class="docs-heading-anchor" href="#normalized_coefficient"><code>normalized_coefficient</code></a><a id="normalized_coefficient-1"></a><a class="docs-heading-anchor-permalink" href="#normalized_coefficient" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.normalized_coefficient" href="#JuMP.normalized_coefficient"><code>JuMP.normalized_coefficient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">normalized_coefficient(con_ref::ConstraintRef, variable::GenericVariableRef)</code></pre><p>Return the coefficient associated with <code>variable</code> in <code>constraint</code> after JuMP has normalized the constraint into its standard form. See also <a href="#JuMP.set_normalized_coefficient"><code>set_normalized_coefficient</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L813-L819">source</a></section></article><h2 id="normalized_rhs"><a class="docs-heading-anchor" href="#normalized_rhs"><code>normalized_rhs</code></a><a id="normalized_rhs-1"></a><a class="docs-heading-anchor-permalink" href="#normalized_rhs" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.normalized_rhs" href="#JuMP.normalized_rhs"><code>JuMP.normalized_rhs</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">normalized_rhs(con_ref::ConstraintRef)</code></pre><p>Return the right-hand side term of <code>con_ref</code> after JuMP has converted the constraint into its normalized form. See also <a href="#JuMP.set_normalized_rhs"><code>set_normalized_rhs</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L873-L878">source</a></section></article><h2 id="num_constraints"><a class="docs-heading-anchor" href="#num_constraints"><code>num_constraints</code></a><a id="num_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#num_constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.num_constraints" href="#JuMP.num_constraints"><code>JuMP.num_constraints</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">num_constraints(model::GenericModel, function_type, set_type)::Int64</code></pre><p>Return the number of constraints currently in the model where the function has type <code>function_type</code> and the set has type <code>set_type</code>.</p><p>See also <a href="#JuMP.list_of_constraint_types"><code>list_of_constraint_types</code></a> and <a href="#JuMP.all_constraints"><code>all_constraints</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 0, Bin);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @constraint(model, y in MOI.GreaterThan(1.0));
+
+julia&gt; @constraint(model, y &lt;= 1.0);
+
+julia&gt; @constraint(model, 2x &lt;= 1);
+
+julia&gt; num_constraints(model, VariableRef, MOI.GreaterThan{Float64})
+2
+
+julia&gt; num_constraints(model, VariableRef, MOI.ZeroOne)
+1
+
+julia&gt; num_constraints(model, AffExpr, MOI.LessThan{Float64})
+2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1192-L1224">source</a></section><section><div><pre><code class="language-julia hljs">num_constraints(model::GenericModel; count_variable_in_set_constraints::Bool)</code></pre><p>Return the number of constraints in <code>model</code>.</p><p>If <code>count_variable_in_set_constraints == true</code>, then <code>VariableRef</code> constraints such as <code>VariableRef</code>-in-<code>Integer</code> are included. To count only the number of structural constraints (e.g., the rows in the constraint matrix of a linear program), pass <code>count_variable_in_set_constraints = false</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 0, Int);
+
+julia&gt; @constraint(model, 2x &lt;= 1);
+
+julia&gt; num_constraints(model; count_variable_in_set_constraints = true)
+3
+
+julia&gt; num_constraints(model; count_variable_in_set_constraints = false)
+1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1339-L1364">source</a></section></article><h2 id="num_nonlinear_constraints"><a class="docs-heading-anchor" href="#num_nonlinear_constraints"><code>num_nonlinear_constraints</code></a><a id="num_nonlinear_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#num_nonlinear_constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.num_nonlinear_constraints" href="#JuMP.num_nonlinear_constraints"><code>JuMP.num_nonlinear_constraints</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">num_nonlinear_constraints(model::GenericModel)</code></pre><p>Returns the number of nonlinear constraints associated with the <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L499-L503">source</a></section></article><h2 id="num_variables"><a class="docs-heading-anchor" href="#num_variables"><code>num_variables</code></a><a id="num_variables-1"></a><a class="docs-heading-anchor-permalink" href="#num_variables" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.num_variables" href="#JuMP.num_variables"><code>JuMP.num_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">num_variables(model::GenericModel)::Int64</code></pre><p>Returns number of variables in <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L6-L10">source</a></section></article><h2 id="object_dictionary"><a class="docs-heading-anchor" href="#object_dictionary"><code>object_dictionary</code></a><a id="object_dictionary-1"></a><a class="docs-heading-anchor-permalink" href="#object_dictionary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.object_dictionary" href="#JuMP.object_dictionary"><code>JuMP.object_dictionary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">object_dictionary(model::GenericModel)</code></pre><p>Return the dictionary that maps the symbol name of a variable, constraint, or expression to the corresponding object.</p><p>Objects are registered to a specific symbol in the macros. For example, <code>@variable(model, x[1:2, 1:2])</code> registers the array of variables <code>x</code> to the symbol <code>:x</code>.</p><p>This method should be defined for any subtype of <code>AbstractModel</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L843-L854">source</a></section></article><h2 id="objective_bound"><a class="docs-heading-anchor" href="#objective_bound"><code>objective_bound</code></a><a id="objective_bound-1"></a><a class="docs-heading-anchor-permalink" href="#objective_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.objective_bound" href="#JuMP.objective_bound"><code>JuMP.objective_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">objective_bound(model::GenericModel)</code></pre><p>Return the best known bound on the optimal objective value after a call to <code>optimize!(model)</code>.</p><p>For scalar-valued objectives, this function returns a <code>Float64</code>. For vector-valued objectives, it returns a <code>Vector{Float64}</code>.</p><p>In the case of a vector-valued objective, this returns the <em>ideal point</em>, that is, the point obtained if each objective was optimized independently.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L23-L34">source</a></section></article><h2 id="objective_function"><a class="docs-heading-anchor" href="#objective_function"><code>objective_function</code></a><a id="objective_function-1"></a><a class="docs-heading-anchor-permalink" href="#objective_function" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.objective_function" href="#JuMP.objective_function"><code>JuMP.objective_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">objective_function(
+    model::GenericModel,
+    T::Type = objective_function_type(model),
+)</code></pre><p>Return an object of type <code>T</code> representing the objective function.</p><p>Error if the objective is not convertible to type <code>T</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @objective(model, Min, 2x + 1)
+2 x + 1
+
+julia&gt; objective_function(model, AffExpr)
+2 x + 1
+
+julia&gt; objective_function(model, QuadExpr)
+2 x + 1
+
+julia&gt; typeof(objective_function(model, QuadExpr))
+QuadExpr (alias for GenericQuadExpr{Float64, GenericVariableRef{Float64}})</code></pre><p>We see with the last two commands that even if the objective function is affine, as it is convertible to a quadratic function, it can be queried as a quadratic function and the result is quadratic.</p><p>However, it is not convertible to a variable.</p><pre><code class="language-julia-repl hljs">julia&gt; objective_function(model, VariableRef)
+ERROR: InexactError: convert(MathOptInterface.VariableIndex, 1.0 + 2.0 MOI.VariableIndex(1))
+[...]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L191-L231">source</a></section></article><h2 id="objective_function_string"><a class="docs-heading-anchor" href="#objective_function_string"><code>objective_function_string</code></a><a id="objective_function_string-1"></a><a class="docs-heading-anchor-permalink" href="#objective_function_string" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.objective_function_string" href="#JuMP.objective_function_string"><code>JuMP.objective_function_string</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">objective_function_string(mode, model::AbstractModel)::String</code></pre><p>Return a <code>String</code> describing the objective function of the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L367-L371">source</a></section></article><h2 id="objective_function_type"><a class="docs-heading-anchor" href="#objective_function_type"><code>objective_function_type</code></a><a id="objective_function_type-1"></a><a class="docs-heading-anchor-permalink" href="#objective_function_type" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.objective_function_type" href="#JuMP.objective_function_type"><code>JuMP.objective_function_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">objective_function_type(model::GenericModel)::AbstractJuMPScalar</code></pre><p>Return the type of the objective function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L179-L183">source</a></section></article><h2 id="objective_sense"><a class="docs-heading-anchor" href="#objective_sense"><code>objective_sense</code></a><a id="objective_sense-1"></a><a class="docs-heading-anchor-permalink" href="#objective_sense" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.objective_sense" href="#JuMP.objective_sense"><code>JuMP.objective_sense</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">objective_sense(model::GenericModel)::MOI.OptimizationSense</code></pre><p>Return the objective sense.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L75-L79">source</a></section></article><h2 id="objective_value"><a class="docs-heading-anchor" href="#objective_value"><code>objective_value</code></a><a id="objective_value-1"></a><a class="docs-heading-anchor-permalink" href="#objective_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.objective_value" href="#JuMP.objective_value"><code>JuMP.objective_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">objective_value(model::GenericModel; result::Int = 1)</code></pre><p>Return the objective value associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>For scalar-valued objectives, this function returns a <code>Float64</code>. For vector-valued objectives, it returns a <code>Vector{Float64}</code>.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L39-L49">source</a></section></article><h2 id="op_ifelse"><a class="docs-heading-anchor" href="#op_ifelse"><code>op_ifelse</code></a><a id="op_ifelse-1"></a><a class="docs-heading-anchor-permalink" href="#op_ifelse" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_ifelse" href="#JuMP.op_ifelse"><code>JuMP.op_ifelse</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">op_ifelse(a, x, y)</code></pre><p>A function that falls back to <code>ifelse(a, x, y)</code>, but when called with a JuMP variables or expression in the first argument, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_ifelse(true, 1.0, 2.0)
+1.0
+
+julia&gt; op_ifelse(x, 1.0, 2.0)
+ifelse(x, 1.0, 2.0)
+
+julia&gt; op_ifelse(true, x, 2.0)
+x</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L484-L507">source</a></section></article><h2 id="operator_to_set"><a class="docs-heading-anchor" href="#operator_to_set"><code>operator_to_set</code></a><a id="operator_to_set-1"></a><a class="docs-heading-anchor-permalink" href="#operator_to_set" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.operator_to_set" href="#JuMP.operator_to_set"><code>JuMP.operator_to_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">operator_to_set(_error::Function, ::Val{sense_symbol})</code></pre><p>Converts a sense symbol to a set <code>set</code> such that <code>@constraint(model, func sense_symbol 0)</code> is equivalent to <code>@constraint(model, func in set)</code> for any <code>func::AbstractJuMPScalar</code>.</p><p><strong>Example</strong></p><p>Once a custom set is defined you can directly create a JuMP constraint with it:</p><pre><code class="language-julia-repl hljs">julia&gt; struct CustomSet{T} &lt;: MOI.AbstractScalarSet
+           value::T
+       end
+
+julia&gt; Base.copy(x::CustomSet) = CustomSet(x.value)
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; cref = @constraint(model, x in CustomSet(1.0))
+x ∈ CustomSet{Float64}(1.0)</code></pre><p>However, there might be an appropriate sign that could be used in order to provide a more convenient syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; JuMP.operator_to_set(::Function, ::Val{:⊰}) = CustomSet(0.0)
+
+julia&gt; MOIU.supports_shift_constant(::Type{&lt;:CustomSet}) = true
+
+julia&gt; MOIU.shift_constant(set::CustomSet, value) = CustomSet(set.value + value)
+
+julia&gt; cref = @constraint(model, x ⊰ 1)
+x ∈ CustomSet{Float64}(1.0)</code></pre><p>Note that the whole function is first moved to the right-hand side, then the sign is transformed into a set with zero constant and finally the constant is moved to the set with <code>MOIU.shift_constant</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L163-L204">source</a></section></article><h2 id="operator_warn"><a class="docs-heading-anchor" href="#operator_warn"><code>operator_warn</code></a><a id="operator_warn-1"></a><a class="docs-heading-anchor-permalink" href="#operator_warn" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.operator_warn" href="#JuMP.operator_warn"><code>JuMP.operator_warn</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">operator_warn(model::AbstractModel)
+operator_warn(model::GenericModel)</code></pre><p>This function is called on the model whenever two affine expressions are added together without using <code>destructive_add!</code>, and at least one of the two expressions has more than 50 terms.</p><p>For the case of <code>Model</code>, if this function is called more than 20,000 times then a warning is generated once.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/operators.jl#L263-L273">source</a></section></article><h2 id="optimize!"><a class="docs-heading-anchor" href="#optimize!"><code>optimize!</code></a><a id="optimize!-1"></a><a class="docs-heading-anchor-permalink" href="#optimize!" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.optimize!" href="#JuMP.optimize!"><code>JuMP.optimize!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">optimize!(
+    model::GenericModel;
+    ignore_optimize_hook = (model.optimize_hook === nothing),
+    _differentiation_backend::MOI.Nonlinear.AbstractAutomaticDifferentiation =
+        MOI.Nonlinear.SparseReverseMode(),
+    kwargs...,
+)</code></pre><p>Optimize the model.</p><p>If an optimizer has not been set yet (see <a href="#JuMP.set_optimizer"><code>set_optimizer</code></a>), a <a href="#JuMP.NoOptimizer"><code>NoOptimizer</code></a> error is thrown.</p><p>If <code>ignore_optimize_hook == true</code>, the optimize hook is ignored and the model is solved as if the hook was not set. Keyword arguments <code>kwargs</code> are passed to the <code>optimize_hook</code>. An error is thrown if <code>optimize_hook</code> is <code>nothing</code> and keyword arguments are provided.</p><p><strong>Experimental features</strong></p><p>These features may change or be removed in any future version of JuMP.</p><p>Pass <code>_differentiation_backend</code> to set the <a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation"><code>MOI.Nonlinear.AbstractAutomaticDifferentiation</code></a> backend used to compute derivatives of nonlinear programs.</p><p>If you require only <code>:ExprGraph</code>, it is more efficient to pass <code>_differentiation_backend = MOI.Nonlinear.ExprGraphOnly()</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L379-L408">source</a></section></article><h2 id="optimizer_index"><a class="docs-heading-anchor" href="#optimizer_index"><code>optimizer_index</code></a><a id="optimizer_index-1"></a><a class="docs-heading-anchor-permalink" href="#optimizer_index" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.optimizer_index" href="#JuMP.optimizer_index"><code>JuMP.optimizer_index</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">optimizer_index(x::GenericVariableRef)::MOI.VariableIndex
+optimizer_index(x::ConstraintRef{&lt;:GenericModel})::MOI.ConstraintIndex</code></pre><p>Return the index that corresponds to <code>x</code> in the optimizer model.</p><p>Throws <a href="#JuMP.NoOptimizer"><code>NoOptimizer</code></a> if no optimizer is set, and throws an <code>ErrorException</code> if the optimizer is set but is not attached.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L1063-L1071">source</a></section></article><h2 id="optimizer_with_attributes"><a class="docs-heading-anchor" href="#optimizer_with_attributes"><code>optimizer_with_attributes</code></a><a id="optimizer_with_attributes-1"></a><a class="docs-heading-anchor-permalink" href="#optimizer_with_attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.optimizer_with_attributes" href="#JuMP.optimizer_with_attributes"><code>JuMP.optimizer_with_attributes</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">optimizer_with_attributes(optimizer_constructor, attrs::Pair...)</code></pre><p>Groups an optimizer constructor with the list of attributes <code>attrs</code>. Note that it is equivalent to <code>MOI.OptimizerWithAttributes</code>.</p><p>When provided to the <code>Model</code> constructor or to <a href="#JuMP.set_optimizer"><code>set_optimizer</code></a>, it creates an optimizer by calling <code>optimizer_constructor()</code>, and then sets the attributes using <a href="#JuMP.set_attribute"><code>set_attribute</code></a>.</p><p>See also: <a href="#JuMP.set_attribute"><code>set_attribute</code></a>, <a href="#JuMP.get_attribute"><code>get_attribute</code></a>.</p><p><strong>Note</strong></p><p>The string names of the attributes are specific to each solver. One should consult the solver&#39;s documentation to find the attributes of interest.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; optimizer = optimizer_with_attributes(
+           HiGHS.Optimizer, &quot;presolve&quot; =&gt; &quot;off&quot;, MOI.Silent() =&gt; true,
+       );
+
+julia&gt; model = Model(optimizer);</code></pre><p>is equivalent to:</p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_attribute(model, &quot;presolve&quot;, &quot;off&quot;)
+
+julia&gt; set_attribute(model, MOI.Silent(), true)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L6-L46">source</a></section></article><h2 id="owner_model"><a class="docs-heading-anchor" href="#owner_model"><code>owner_model</code></a><a id="owner_model-1"></a><a class="docs-heading-anchor-permalink" href="#owner_model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.owner_model" href="#JuMP.owner_model"><code>JuMP.owner_model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">owner_model(s::AbstractJuMPScalar)</code></pre><p>Return the model owning the scalar <code>s</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1042-L1046">source</a></section></article><h2 id="parameter_value"><a class="docs-heading-anchor" href="#parameter_value"><code>parameter_value</code></a><a id="parameter_value-1"></a><a class="docs-heading-anchor-permalink" href="#parameter_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.parameter_value" href="#JuMP.parameter_value"><code>JuMP.parameter_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parameter_value(x::GenericVariableRef)</code></pre><p>Return the value of the parameter <code>x</code>.</p><p>Errors if <code>x</code> is not a parameter.</p><p>See also <a href="#JuMP.ParameterRef"><code>ParameterRef</code></a>, <a href="#JuMP.is_parameter"><code>is_parameter</code></a>, <a href="#JuMP.set_parameter_value"><code>set_parameter_value</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, p in Parameter(2))
+p
+
+julia&gt; parameter_value(p)
+2.0
+
+julia&gt; set_parameter_value(p, 2.5)
+
+julia&gt; parameter_value(p)
+2.5</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1593-L1619">source</a></section></article><h2 id="parse_constraint"><a class="docs-heading-anchor" href="#parse_constraint"><code>parse_constraint</code></a><a id="parse_constraint-1"></a><a class="docs-heading-anchor-permalink" href="#parse_constraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.parse_constraint" href="#JuMP.parse_constraint"><code>JuMP.parse_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parse_constraint(_error::Function, expr::Expr)</code></pre><p>The entry-point for all constraint-related parsing.</p><p><strong>Arguments</strong></p><ul><li>The <code>_error</code> function is passed everywhere to provide better error messages</li><li><code>expr</code> comes from the <code>@constraint</code> macro. There are two possibilities:<ul><li><code>@constraint(model, expr)</code></li><li><code>@constraint(model, name[args], expr)</code></li></ul>In both cases, <code>expr</code> is the main component of the constraint.</li></ul><p><strong>Supported syntax</strong></p><p>JuMP currently supports the following <code>expr</code> objects:</p><ul><li><code>lhs &lt;= rhs</code></li><li><code>lhs == rhs</code></li><li><code>lhs &gt;= rhs</code></li><li><code>l &lt;= body &lt;= u</code></li><li><code>u &gt;= body &gt;= l</code></li><li><code>lhs ⟂ rhs</code></li><li><code>lhs in rhs</code></li><li><code>lhs ∈ rhs</code></li><li><code>z =&gt; {constraint}</code></li><li><code>!z =&gt; {constraint}</code></li></ul><p>as well as all broadcasted variants.</p><p><strong>Extensions</strong></p><p>The infrastructure behind <code>parse_constraint</code> is extendable. See <a href="#JuMP.parse_constraint_head"><code>parse_constraint_head</code></a> and <a href="#JuMP.parse_constraint_call"><code>parse_constraint_call</code></a> for details.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L386-L418">source</a></section></article><h2 id="parse_constraint_call"><a class="docs-heading-anchor" href="#parse_constraint_call"><code>parse_constraint_call</code></a><a id="parse_constraint_call-1"></a><a class="docs-heading-anchor-permalink" href="#parse_constraint_call" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.parse_constraint_call" href="#JuMP.parse_constraint_call"><code>JuMP.parse_constraint_call</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parse_constraint_call(
+    _error::Function,
+    is_vectorized::Bool,
+    ::Val{op},
+    args...,
+)</code></pre><p>Implement this method to intercept the parsing of a <code>:call</code> expression with operator <code>op</code>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Extending the constraint macro at parse time is an advanced operation and has the potential to interfere with existing JuMP syntax. Please discuss with the <a href="https://gitter.im/JuliaOpt/jump-dev">developer chatroom</a> before publishing any code that implements these methods.</p></div></div><p><strong>Arguments</strong></p><ul><li><code>_error</code>: a function that accepts a <code>String</code> and throws the string as an error, along with some descriptive information of the macro from which it was thrown.</li><li><code>is_vectorized</code>: a boolean to indicate if <code>op</code> should be broadcast or not</li><li><code>op</code>: the first element of the <code>.args</code> field of the <code>Expr</code> to intercept</li><li><code>args...</code>: the <code>.args</code> field of the <code>Expr</code>.</li></ul><p><strong>Returns</strong></p><p>This function must return:</p><ul><li><code>parse_code::Expr</code>: an expression containing any setup or rewriting code that needs to be called before <code>build_constraint</code></li><li><code>build_code::Expr</code>: an expression that calls <code>build_constraint(</code> or <code>build_constraint.(</code> depending on <code>is_vectorized</code>.</li></ul><p>See also: <a href="#JuMP.parse_constraint_head"><code>parse_constraint_head</code></a>, <a href="#JuMP.build_constraint"><code>build_constraint</code></a></p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L773-L809">source</a></section><section><div><pre><code class="language-julia hljs">parse_constraint_call(
+    _error::Function,
+    vectorized::Bool,
+    ::Val{op},
+    lhs,
+    rhs,
+) where {op}</code></pre><p>Fallback handler for binary operators. These might be infix operators like <code>@constraint(model, lhs op rhs)</code>, or normal operators like <code>@constraint(model, op(lhs, rhs))</code>.</p><p>In both cases, we rewrite as <code>lhs - rhs in operator_to_set(_error, op)</code>.</p><p>See <a href="#JuMP.operator_to_set"><code>operator_to_set</code></a> for details.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L842-L858">source</a></section></article><h2 id="parse_constraint_head"><a class="docs-heading-anchor" href="#parse_constraint_head"><code>parse_constraint_head</code></a><a id="parse_constraint_head-1"></a><a class="docs-heading-anchor-permalink" href="#parse_constraint_head" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.parse_constraint_head" href="#JuMP.parse_constraint_head"><code>JuMP.parse_constraint_head</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parse_constraint_head(_error::Function, ::Val{head}, args...)</code></pre><p>Implement this method to intercept the parsing of an expression with head <code>head</code>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Extending the constraint macro at parse time is an advanced operation and has the potential to interfere with existing JuMP syntax. Please discuss with the <a href="https://gitter.im/JuliaOpt/jump-dev">developer chatroom</a> before publishing any code that implements these methods.</p></div></div><p><strong>Arguments</strong></p><ul><li><code>_error</code>: a function that accepts a <code>String</code> and throws the string as an error, along with some descriptive information of the macro from which it was thrown.</li><li><code>head</code>: the <code>.head</code> field of the <code>Expr</code> to intercept</li><li><code>args...</code>: the <code>.args</code> field of the <code>Expr</code>.</li></ul><p><strong>Returns</strong></p><p>This function must return:</p><ul><li><code>is_vectorized::Bool</code>: whether the expression represents a broadcasted expression like <code>x .&lt;= 1</code></li><li><code>parse_code::Expr</code>: an expression containing any setup or rewriting code that needs to be called before <code>build_constraint</code></li><li><code>build_code::Expr</code>: an expression that calls <code>build_constraint(</code> or <code>build_constraint.(</code> depending on <code>is_vectorized</code>.</li></ul><p><strong>Existing implementations</strong></p><p>JuMP currently implements:</p><ul><li><code>::Val{:call}</code>, which forwards calls to <a href="#JuMP.parse_constraint_call"><code>parse_constraint_call</code></a></li><li><code>::Val{:comparison}</code>, which handles the special case of <code>l &lt;= body &lt;= u</code>.</li></ul><p>See also: <a href="#JuMP.parse_constraint_call"><code>parse_constraint_call</code></a>, <a href="#JuMP.build_constraint"><code>build_constraint</code></a></p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L423-L462">source</a></section></article><h2 id="parse_one_operator_variable"><a class="docs-heading-anchor" href="#parse_one_operator_variable"><code>parse_one_operator_variable</code></a><a id="parse_one_operator_variable-1"></a><a class="docs-heading-anchor-permalink" href="#parse_one_operator_variable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.parse_one_operator_variable" href="#JuMP.parse_one_operator_variable"><code>JuMP.parse_one_operator_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parse_one_operator_variable(_error::Function, infoexpr::_VariableInfoExpr, sense::Val{S}, value) where S</code></pre><p>Update <code>infoexr</code> for a variable expression in the <code>@variable</code> macro of the form <code>variable name S value</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L2276-L2280">source</a></section></article><h2 id="parse_ternary_variable"><a class="docs-heading-anchor" href="#parse_ternary_variable"><code>parse_ternary_variable</code></a><a id="parse_ternary_variable-1"></a><a class="docs-heading-anchor-permalink" href="#parse_ternary_variable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.parse_ternary_variable" href="#JuMP.parse_ternary_variable"><code>JuMP.parse_ternary_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parse_ternary_variable(_error, variable_info, lhs_sense, lhs, rhs_sense, rhs)</code></pre><p>A hook for JuMP extensiosn to intercept the parsing of a <code>:comparison</code> expression, which has the form <code>lhs lhs_sense variable rhs_sense rhs</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L2405-L2410">source</a></section></article><h2 id="parse_variable"><a class="docs-heading-anchor" href="#parse_variable"><code>parse_variable</code></a><a id="parse_variable-1"></a><a class="docs-heading-anchor-permalink" href="#parse_variable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.parse_variable" href="#JuMP.parse_variable"><code>JuMP.parse_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parse_variable(_error::Function, ::_VariableInfoExpr, args...)</code></pre><p>A hook for extensions to intercept the parsing of inequality constraints in the <a href="#JuMP.@variable"><code>@variable</code></a> macro.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L2262-L2267">source</a></section></article><h2 id="primal_feasibility_report"><a class="docs-heading-anchor" href="#primal_feasibility_report"><code>primal_feasibility_report</code></a><a id="primal_feasibility_report-1"></a><a class="docs-heading-anchor-permalink" href="#primal_feasibility_report" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.primal_feasibility_report" href="#JuMP.primal_feasibility_report"><code>JuMP.primal_feasibility_report</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">primal_feasibility_report(
+    model::GenericModel{T},
+    point::AbstractDict{GenericVariableRef{T},T} = _last_primal_solution(model),
+    atol::T = zero(T),
+    skip_missing::Bool = false,
+)::Dict{Any,T}</code></pre><p>Given a dictionary <code>point</code>, which maps variables to primal values, return a dictionary whose keys are the constraints with an infeasibility greater than the supplied tolerance <code>atol</code>. The value corresponding to each key is the respective infeasibility. Infeasibility is defined as the distance between the primal value of the constraint (see <code>MOI.ConstraintPrimal</code>) and the nearest point by Euclidean distance in the corresponding set.</p><p><strong>Notes</strong></p><ul><li>If <code>skip_missing = true</code>, constraints containing variables that are not in <code>point</code> will be ignored.</li><li>If <code>skip_missing = false</code> and a partial primal solution is provided, an error will be thrown.</li><li>If no point is provided, the primal solution from the last time the model was solved is used.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, 0.5 &lt;= x &lt;= 1);
+
+julia&gt; primal_feasibility_report(model, Dict(x =&gt; 0.2))
+Dict{Any, Float64} with 1 entry:
+  x ≥ 0.5 =&gt; 0.3</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/feasibility_checker.jl#L16-L51">source</a></section><section><div><pre><code class="language-julia hljs">primal_feasibility_report(
+    point::Function,
+    model::GenericModel{T};
+    atol::T = zero(T),
+    skip_missing::Bool = false,
+) where {T}</code></pre><p>A form of <code>primal_feasibility_report</code> where a function is passed as the first argument instead of a dictionary as the second argument.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, 0.5 &lt;= x &lt;= 1, start = 1.3);
+
+julia&gt; primal_feasibility_report(model) do v
+           return start_value(v)
+       end
+Dict{Any, Float64} with 1 entry:
+  x ≤ 1 =&gt; 0.3</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/feasibility_checker.jl#L74-L98">source</a></section></article><h2 id="primal_status"><a class="docs-heading-anchor" href="#primal_status"><code>primal_status</code></a><a id="primal_status-1"></a><a class="docs-heading-anchor-permalink" href="#primal_status" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.primal_status" href="#JuMP.primal_status"><code>JuMP.primal_status</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">primal_status(model::GenericModel; result::Int = 1)</code></pre><p>Return a <a href="../../moi/reference/models/#MathOptInterface.ResultStatusCode"><code>MOI.ResultStatusCode</code></a> describing the status of the most recent primal solution of the solver (i.e., the <a href="../../moi/reference/models/#MathOptInterface.PrimalStatus"><code>MOI.PrimalStatus</code></a> attribute) associated with the result index <code>result</code>.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L550-L558">source</a></section></article><h2 id="print_active_bridges"><a class="docs-heading-anchor" href="#print_active_bridges"><code>print_active_bridges</code></a><a id="print_active_bridges-1"></a><a class="docs-heading-anchor-permalink" href="#print_active_bridges" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.print_active_bridges" href="#JuMP.print_active_bridges"><code>JuMP.print_active_bridges</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">print_active_bridges([io::IO = stdout,] model::GenericModel)</code></pre><p>Print a list of the variable, constraint, and objective bridges that are currently used in the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L733-L738">source</a></section><section><div><pre><code class="language-julia hljs">print_active_bridges([io::IO = stdout,] model::GenericModel, ::Type{F}) where {F}</code></pre><p>Print a list of bridges required for an objective function of type <code>F</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L745-L749">source</a></section><section><div><pre><code class="language-julia hljs">print_active_bridges(
+    [io::IO = stdout,]
+    model::GenericModel,
+    F::Type,
+    S::Type{&lt;:MOI.AbstractSet},
+)</code></pre><p>Print a list of bridges required for a constraint of type <code>F</code>-in-<code>S</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L756-L765">source</a></section><section><div><pre><code class="language-julia hljs">print_active_bridges(
+    [io::IO = stdout,]
+    model::GenericModel,
+    S::Type{&lt;:MOI.AbstractSet},
+)</code></pre><p>Print a list of bridges required to add a variable constrained to the set <code>S</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L777-L785">source</a></section></article><h2 id="print_bridge_graph"><a class="docs-heading-anchor" href="#print_bridge_graph"><code>print_bridge_graph</code></a><a id="print_bridge_graph-1"></a><a class="docs-heading-anchor-permalink" href="#print_bridge_graph" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.print_bridge_graph" href="#JuMP.print_bridge_graph"><code>JuMP.print_bridge_graph</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs"> print_bridge_graph([io::IO,] model::GenericModel)</code></pre><p>Print the hyper-graph containing all variable, constraint, and objective types that could be obtained by bridging the variables, constraints, and objectives that are present in the model.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This function is intended for advanced users. If you want to see only the bridges that are currently used, use <a href="#JuMP.print_active_bridges"><code>print_active_bridges</code></a> instead.</p></div></div><p><strong>Explanation of output</strong></p><p>Each node in the hyper-graph corresponds to a variable, constraint, or objective type.</p><ul><li>Variable nodes are indicated by <code>[ ]</code></li><li>Constraint nodes are indicated by <code>( )</code></li><li>Objective nodes are indicated by <code>| |</code></li></ul><p>The number inside each pair of brackets is an index of the node in the hyper-graph.</p><p>Note that this hyper-graph is the full list of possible transformations. When the bridged model is created, we select the shortest hyper-path(s) from this graph, so many nodes may be un-used.</p><p>For more information, see Legat, B., Dowson, O., Garcia, J., and Lubin, M. (2020).  &quot;MathOptInterface: a data structure for mathematical optimization problems.&quot; URL: <a href="https://arxiv.org/abs/2002.03447">https://arxiv.org/abs/2002.03447</a></p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L696-L724">source</a></section></article><h2 id="quad_terms"><a class="docs-heading-anchor" href="#quad_terms"><code>quad_terms</code></a><a id="quad_terms-1"></a><a class="docs-heading-anchor-permalink" href="#quad_terms" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.quad_terms" href="#JuMP.quad_terms"><code>JuMP.quad_terms</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">quad_terms(quad::GenericQuadExpr{C, V})</code></pre><p>Provides an iterator over tuples <code>(coefficient::C, var_1::V, var_2::V)</code> in the quadratic part of the quadratic expression.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L284-L289">source</a></section></article><h2 id="raw_status"><a class="docs-heading-anchor" href="#raw_status"><code>raw_status</code></a><a id="raw_status-1"></a><a class="docs-heading-anchor-permalink" href="#raw_status" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.raw_status" href="#JuMP.raw_status"><code>JuMP.raw_status</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">raw_status(model::GenericModel)</code></pre><p>Return the reason why the solver stopped in its own words (i.e., the MathOptInterface model attribute <code>RawStatusString</code>).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L527-L532">source</a></section></article><h2 id="read_from_file"><a class="docs-heading-anchor" href="#read_from_file"><code>read_from_file</code></a><a id="read_from_file-1"></a><a class="docs-heading-anchor-permalink" href="#read_from_file" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.read_from_file" href="#JuMP.read_from_file"><code>JuMP.read_from_file</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">read_from_file(
+    filename::String;
+    format::MOI.FileFormats.FileFormat = MOI.FileFormats.FORMAT_AUTOMATIC,
+    kwargs...,
+)</code></pre><p>Return a JuMP model read from <code>filename</code> in the format <code>format</code>.</p><p>If the filename ends in <code>.gz</code>, it will be uncompressed using Gzip. If the filename ends in <code>.bz2</code>, it will be uncompressed using BZip2.</p><p>Other <code>kwargs</code> are passed to the <code>Model</code> constructor of the chosen format.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/file_formats.jl#L91-L104">source</a></section></article><h2 id="reduced_cost"><a class="docs-heading-anchor" href="#reduced_cost"><code>reduced_cost</code></a><a id="reduced_cost-1"></a><a class="docs-heading-anchor-permalink" href="#reduced_cost" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.reduced_cost" href="#JuMP.reduced_cost"><code>JuMP.reduced_cost</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">reduced_cost(x::GenericVariableRef{T})::T where {T}</code></pre><p>Return the reduced cost associated with variable <code>x</code>.</p><p>Equivalent to querying the shadow price of the active variable bound (if one exists and is active).</p><p>See also: <a href="#JuMP.shadow_price"><code>shadow_price</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L2099-L2108">source</a></section></article><h2 id="register"><a class="docs-heading-anchor" href="#register"><code>register</code></a><a id="register-1"></a><a class="docs-heading-anchor-permalink" href="#register" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.register" href="#JuMP.register"><code>JuMP.register</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">register(
+    model::Model,
+    op::Symbol,
+    dimension::Integer,
+    f::Function;
+    autodiff:Bool = false,
+)</code></pre><p>Register the user-defined function <code>f</code> that takes <code>dimension</code> arguments in <code>model</code> as the symbol <code>op</code>.</p><p>The function <code>f</code> must support all subtypes of <code>Real</code> as arguments. Do not assume that the inputs are <code>Float64</code>.</p><p><strong>Notes</strong></p><ul><li>For this method, you must explicitly set <code>autodiff = true</code>, because no user-provided gradient function <code>∇f</code> is given.</li><li>Second-derivative information is only computed if <code>dimension == 1</code>.</li><li><code>op</code> does not have to be the same symbol as <code>f</code>, but it is generally more readable if it is.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; f(x::T) where {T&lt;:Real} = x^2
+f (generic function with 1 method)
+
+julia&gt; register(model, :foo, 1, f; autodiff = true)
+
+julia&gt; @NLobjective(model, Min, foo(x))</code></pre><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; g(x::T, y::T) where {T&lt;:Real} = x * y
+g (generic function with 1 method)
+
+julia&gt; register(model, :g, 2, g; autodiff = true)
+
+julia&gt; @NLobjective(model, Min, g(x[1], x[2]))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L654-L708">source</a></section><section><div><pre><code class="language-julia hljs">register(
+    model::Model,
+    s::Symbol,
+    dimension::Integer,
+    f::Function,
+    ∇f::Function;
+    autodiff:Bool = false,
+)</code></pre><p>Register the user-defined function <code>f</code> that takes <code>dimension</code> arguments in <code>model</code> as the symbol <code>s</code>. In addition, provide a gradient function <code>∇f</code>.</p><p>The functions <code>f</code>and <code>∇f</code> must support all subtypes of <code>Real</code> as arguments. Do not assume that the inputs are <code>Float64</code>.</p><p><strong>Notes</strong></p><ul><li>If the function <code>f</code> is univariate (i.e., <code>dimension == 1</code>), <code>∇f</code> must return a number which represents the first-order derivative of the function <code>f</code>.</li><li>If the function <code>f</code> is multi-variate, <code>∇f</code> must have a signature matching <code>∇f(g::AbstractVector{T}, args::T...) where {T&lt;:Real}</code>, where the first argument is a vector <code>g</code> that is modified in-place with the gradient.</li><li>If <code>autodiff = true</code> and <code>dimension == 1</code>, use automatic differentiation to compute the second-order derivative information. If <code>autodiff = false</code>, only first-order derivative information will be used.</li><li><code>s</code> does not have to be the same symbol as <code>f</code>, but it is generally more readable if it is.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; f(x::T) where {T&lt;:Real} = x^2
+f (generic function with 1 method)
+
+julia&gt; ∇f(x::T) where {T&lt;:Real} = 2 * x
+∇f (generic function with 1 method)
+
+julia&gt; register(model, :foo, 1, f, ∇f; autodiff = true)
+
+julia&gt; @NLobjective(model, Min, foo(x))</code></pre><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; g(x::T, y::T) where {T&lt;:Real} = x * y
+g (generic function with 1 method)
+
+julia&gt; function ∇g(g::AbstractVector{T}, x::T, y::T) where {T&lt;:Real}
+           g[1] = y
+           g[2] = x
+           return
+       end
+∇g (generic function with 1 method)
+
+julia&gt; register(model, :g, 2, g, ∇g)
+
+julia&gt; @NLobjective(model, Min, g(x[1], x[2]))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L724-L794">source</a></section><section><div><pre><code class="language-julia hljs">register(
+    model::Model,
+    s::Symbol,
+    dimension::Integer,
+    f::Function,
+    ∇f::Function,
+    ∇²f::Function,
+)</code></pre><p>Register the user-defined function <code>f</code> that takes <code>dimension</code> arguments in <code>model</code> as the symbol <code>s</code>. In addition, provide a gradient function <code>∇f</code> and a hessian function <code>∇²f</code>.</p><p><code>∇f</code> and <code>∇²f</code> must return numbers corresponding to the first- and second-order derivatives of the function <code>f</code> respectively.</p><p><strong>Notes</strong></p><ul><li>Because automatic differentiation is not used, you can assume the inputs are all <code>Float64</code>.</li><li>This method will throw an error if <code>dimension &gt; 1</code>.</li><li><code>s</code> does not have to be the same symbol as <code>f</code>, but it is generally more readable if it is.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; f(x::Float64) = x^2
+f (generic function with 1 method)
+
+julia&gt; ∇f(x::Float64) = 2 * x
+∇f (generic function with 1 method)
+
+julia&gt; ∇²f(x::Float64) = 2.0
+∇²f (generic function with 1 method)
+
+julia&gt; register(model, :foo, 1, f, ∇f, ∇²f)
+
+julia&gt; @NLobjective(model, Min, foo(x))
+</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L820-L867">source</a></section></article><h2 id="relative_gap"><a class="docs-heading-anchor" href="#relative_gap"><code>relative_gap</code></a><a id="relative_gap-1"></a><a class="docs-heading-anchor-permalink" href="#relative_gap" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.relative_gap" href="#JuMP.relative_gap"><code>JuMP.relative_gap</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">relative_gap(model::GenericModel)</code></pre><p>Return the final relative optimality gap after a call to <code>optimize!(model)</code>. Exact value depends upon implementation of MathOptInterface.RelativeGap() by the particular solver used for optimization.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L12-L18">source</a></section></article><h2 id="relax_integrality"><a class="docs-heading-anchor" href="#relax_integrality"><code>relax_integrality</code></a><a id="relax_integrality-1"></a><a class="docs-heading-anchor-permalink" href="#relax_integrality" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.relax_integrality" href="#JuMP.relax_integrality"><code>JuMP.relax_integrality</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">relax_integrality(model::GenericModel)</code></pre><p>Modifies <code>model</code> to &quot;relax&quot; all binary and integrality constraints on variables. Specifically,</p><ul><li>Binary constraints are deleted, and variable bounds are tightened if necessary to ensure the variable is constrained to the interval <span>$[0, 1]$</span>.</li><li>Integrality constraints are deleted without modifying variable bounds.</li><li>An error is thrown if semi-continuous or semi-integer constraints are present (support may be added for these in the future).</li><li>All other constraints are ignored (left in place). This includes discrete constraints like SOS and indicator constraints.</li></ul><p>Returns a function that can be called without any arguments to restore the original model. The behavior of this function is undefined if additional changes are made to the affected variables in the meantime.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Bin);
+
+julia&gt; @variable(model, 1 &lt;= y &lt;= 10, Int);
+
+julia&gt; @objective(model, Min, x + y);
+
+julia&gt; undo_relax = relax_integrality(model);
+
+julia&gt; print(model)
+Min x + y
+Subject to
+ x ≥ 0
+ y ≥ 1
+ x ≤ 1
+ y ≤ 10
+
+julia&gt; undo_relax()
+
+julia&gt; print(model)
+Min x + y
+Subject to
+ y ≥ 1
+ y ≤ 10
+ y integer
+ x binary</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L2211-L2260">source</a></section></article><h2 id="relax_with_penalty!"><a class="docs-heading-anchor" href="#relax_with_penalty!"><code>relax_with_penalty!</code></a><a id="relax_with_penalty!-1"></a><a class="docs-heading-anchor-permalink" href="#relax_with_penalty!" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.relax_with_penalty!" href="#JuMP.relax_with_penalty!"><code>JuMP.relax_with_penalty!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">relax_with_penalty!(
+    model::GenericModel{T},
+    [penalties::Dict{ConstraintRef,T}];
+    [default::Union{Nothing,Real} = nothing,]
+) where {T}</code></pre><p>Destructively modify the model in-place to create a penalized relaxation of the constraints.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This is a destructive routine that modifies the model in-place. If you don&#39;t want to modify the original model, use <a href="#JuMP.copy_model"><code>copy_model</code></a> to create a copy before calling <a href="#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a>.</p></div></div><p><strong>Reformulation</strong></p><p>See <a href="../../moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>MOI.Utilities.ScalarPenaltyRelaxation</code></a> for details of the reformulation.</p><p>For each constraint <code>ci</code>, the penalty passed to <a href="../../moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>MOI.Utilities.ScalarPenaltyRelaxation</code></a> is <code>get(penalties, ci, default)</code>. If the value is <code>nothing</code>, because <code>ci</code> does not exist in <code>penalties</code> and <code>default = nothing</code>, then the constraint is skipped.</p><p><strong>Return value</strong></p><p>This function returns a <code>Dict{ConstraintRef,AffExpr}</code> that maps each constraint index to the corresponding <code>y + z</code> as an <a href="#JuMP.AffExpr"><code>AffExpr</code></a>. In an optimal solution, query the value of these functions to compute the violation of each constraint.</p><p><strong>Relax a subset of constraints</strong></p><p>To relax a subset of constraints, pass a <code>penalties</code> dictionary and set <code>default = nothing</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; function new_model()
+           model = Model()
+           @variable(model, x)
+           @objective(model, Max, 2x + 1)
+           @constraint(model, c1, 2x - 1 &lt;= -2)
+           @constraint(model, c2, 3x &gt;= 0)
+           return model
+       end
+new_model (generic function with 1 method)
+
+julia&gt; model_1 = new_model();
+
+julia&gt; penalty_map = relax_with_penalty!(model_1; default = 2.0);
+
+julia&gt; penalty_map[model_1[:c1]]
+_[3]
+
+julia&gt; penalty_map[model_1[:c2]]
+_[2]
+
+julia&gt; print(model_1)
+Max 2 x - 2 _[2] - 2 _[3] + 1
+Subject to
+ c2 : 3 x + _[2] ≥ 0
+ c1 : 2 x - _[3] ≤ -1
+ _[2] ≥ 0
+ _[3] ≥ 0
+
+julia&gt; model_2 = new_model();
+
+julia&gt; relax_with_penalty!(model_2, Dict(model_2[:c2] =&gt; 3.0))
+Dict{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.GreaterThan{Float64}}, ScalarShape}, AffExpr} with 1 entry:
+  c2 : 3 x + _[2] ≥ 0 =&gt; _[2]
+
+julia&gt; print(model_2)
+Max 2 x - 3 _[2] + 1
+Subject to
+ c2 : 3 x + _[2] ≥ 0
+ c1 : 2 x ≤ -1
+ _[2] ≥ 0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1436-L1517">source</a></section></article><h2 id="remove_bridge"><a class="docs-heading-anchor" href="#remove_bridge"><code>remove_bridge</code></a><a id="remove_bridge-1"></a><a class="docs-heading-anchor-permalink" href="#remove_bridge" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.remove_bridge" href="#JuMP.remove_bridge"><code>JuMP.remove_bridge</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">remove_bridge(
+    model::GenericModel{S},
+    BT::Type{&lt;:MOI.Bridges.AbstractBridge};
+    coefficient_type::Type{T} = S,
+) where {S,T}</code></pre><p>Remove <code>BT{T}</code> from the list of bridges that can be used to transform unsupported constraints into an equivalent formulation using only constraints supported by the optimizer.</p><p>See also: <a href="#JuMP.add_bridge"><code>add_bridge</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; add_bridge(model, MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge)
+
+julia&gt; remove_bridge(model, MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge)
+
+julia&gt; add_bridge(
+           model,
+           MOI.Bridges.Constraint.NumberConversionBridge;
+           coefficient_type = Complex{Float64},
+       )
+
+julia&gt; remove_bridge(
+           model,
+           MOI.Bridges.Constraint.NumberConversionBridge;
+           coefficient_type = Complex{Float64},
+       )</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L648-L682">source</a></section></article><h2 id="reshape_set"><a class="docs-heading-anchor" href="#reshape_set"><code>reshape_set</code></a><a id="reshape_set-1"></a><a class="docs-heading-anchor-permalink" href="#reshape_set" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.reshape_set" href="#JuMP.reshape_set"><code>JuMP.reshape_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">reshape_set(vectorized_set::MOI.AbstractSet, shape::AbstractShape)</code></pre><p>Return a set in its original shape <code>shape</code> given its vectorized form <code>vectorized_form</code>.</p><p><strong>Example</strong></p><p>Given a <a href="#JuMP.SymmetricMatrixShape"><code>SymmetricMatrixShape</code></a> of vectorized form <code>[1, 2, 3] in MOI.PositiveSemidefinieConeTriangle(2)</code>, the following code returns the set of the original constraint <code>Symmetric(Matrix[1 2; 2 3]) in PSDCone()</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; reshape_set(MOI.PositiveSemidefiniteConeTriangle(2), SymmetricMatrixShape(2))
+PSDCone()</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/shapes.jl#L61-L78">source</a></section></article><h2 id="reshape_vector"><a class="docs-heading-anchor" href="#reshape_vector"><code>reshape_vector</code></a><a id="reshape_vector-1"></a><a class="docs-heading-anchor-permalink" href="#reshape_vector" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.reshape_vector" href="#JuMP.reshape_vector"><code>JuMP.reshape_vector</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">reshape_vector(vectorized_form::Vector, shape::AbstractShape)</code></pre><p>Return an object in its original shape <code>shape</code> given its vectorized form <code>vectorized_form</code>.</p><p><strong>Example</strong></p><p>Given a <a href="#JuMP.SymmetricMatrixShape"><code>SymmetricMatrixShape</code></a> of vectorized form <code>[1, 2, 3]</code>, the following code returns the matrix <code>Symmetric(Matrix[1 2; 2 3])</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; reshape_vector([1, 2, 3], SymmetricMatrixShape(2))
+2×2 LinearAlgebra.Symmetric{Int64, Matrix{Int64}}:
+ 1  2
+ 2  3</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/shapes.jl#L81-L98">source</a></section></article><h2 id="result_count"><a class="docs-heading-anchor" href="#result_count"><code>result_count</code></a><a id="result_count-1"></a><a class="docs-heading-anchor-permalink" href="#result_count" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.result_count" href="#JuMP.result_count"><code>JuMP.result_count</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">result_count(model::GenericModel)</code></pre><p>Return the number of results available to query after a call to <a href="#JuMP.optimize!"><code>optimize!</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L514-L519">source</a></section></article><h2 id="reverse_sense"><a class="docs-heading-anchor" href="#reverse_sense"><code>reverse_sense</code></a><a id="reverse_sense-1"></a><a class="docs-heading-anchor-permalink" href="#reverse_sense" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.reverse_sense" href="#JuMP.reverse_sense"><code>JuMP.reverse_sense</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">reverse_sense(::Val{T}) where {T}</code></pre><p>Given an (in)equality symbol <code>T</code>, return a new <code>Val</code> object with the opposite (in)equality symbol.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L2249-L2254">source</a></section></article><h2 id="set_attribute"><a class="docs-heading-anchor" href="#set_attribute"><code>set_attribute</code></a><a id="set_attribute-1"></a><a class="docs-heading-anchor-permalink" href="#set_attribute" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_attribute" href="#JuMP.set_attribute"><code>JuMP.set_attribute</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_attribute(model::GenericModel, attr::MOI.AbstractModelAttribute, value)
+set_attribute(x::GenericVariableRef, attr::MOI.AbstractVariableAttribute, value)
+set_attribute(cr::ConstraintRef, attr::MOI.AbstractConstraintAttribute, value)</code></pre><p>Set the value of a solver-specifc attribute <code>attr</code> to <code>value</code>.</p><p>This is equivalent to calling <a href="../../moi/reference/models/#MathOptInterface.set"><code>MOI.set</code></a> with the associated MOI model and, for variables and constraints, with the associated <a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> or <a href="../../moi/reference/constraints/#MathOptInterface.ConstraintIndex"><code>MOI.ConstraintIndex</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, c, 2 * x &lt;= 1)
+c : 2 x ≤ 1
+
+julia&gt; set_attribute(model, MOI.Name(), &quot;model_new&quot;)
+
+julia&gt; set_attribute(x, MOI.VariableName(), &quot;x_new&quot;)
+
+julia&gt; set_attribute(c, MOI.ConstraintName(), &quot;c_new&quot;)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L864-L892">source</a></section><section><div><pre><code class="language-julia hljs">set_attribute(
+    model::Union{GenericModel,MOI.OptimizerWithAttributes},
+    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},
+    value,
+)</code></pre><p>Set the value of a solver-specifc attribute <code>attr</code> to <code>value</code>.</p><p>This is equivalent to calling <a href="../../moi/reference/models/#MathOptInterface.set"><code>MOI.set</code></a> with the associated MOI model.</p><p>If <code>attr</code> is an <code>AbstractString</code>, it is converted to <a href="../../moi/reference/models/#MathOptInterface.RawOptimizerAttribute"><code>MOI.RawOptimizerAttribute</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; opt = optimizer_with_attributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; false);
+
+julia&gt; model = Model(opt);
+
+julia&gt; set_attribute(model, &quot;output_flag&quot;, false)
+
+julia&gt; set_attribute(model, MOI.RawOptimizerAttribute(&quot;output_flag&quot;), true)
+
+julia&gt; set_attribute(opt, &quot;output_flag&quot;, true)
+
+julia&gt; set_attribute(opt, MOI.RawOptimizerAttribute(&quot;output_flag&quot;), false)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L920-L951">source</a></section></article><h2 id="set_attributes"><a class="docs-heading-anchor" href="#set_attributes"><code>set_attributes</code></a><a id="set_attributes-1"></a><a class="docs-heading-anchor-permalink" href="#set_attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_attributes" href="#JuMP.set_attributes"><code>JuMP.set_attributes</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_attributes(
+    destination::Union{
+        GenericModel,
+        MOI.OptimizerWithAttributes,
+        GenericVariableRef,
+        ConstraintRef,
+    },
+    pairs::Pair...,
+)</code></pre><p>Given a list of <code>attribute =&gt; value</code> pairs, calls <code>set_attribute(destination, attribute, value)</code> for each pair.</p><p>See also: <a href="#JuMP.set_attribute"><code>set_attribute</code></a>, <a href="#JuMP.get_attribute"><code>get_attribute</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; model = Model(Ipopt.Optimizer);
+
+julia&gt; set_attributes(model, &quot;tol&quot; =&gt; 1e-4, &quot;max_iter&quot; =&gt; 100)</code></pre><p>is equivalent to:</p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; model = Model(Ipopt.Optimizer);
+
+julia&gt; set_attribute(model, &quot;tol&quot;, 1e-4)
+
+julia&gt; set_attribute(model, &quot;max_iter&quot;, 100)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L980-L1015">source</a></section></article><h2 id="set_binary"><a class="docs-heading-anchor" href="#set_binary"><code>set_binary</code></a><a id="set_binary-1"></a><a class="docs-heading-anchor-permalink" href="#set_binary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_binary" href="#JuMP.set_binary"><code>JuMP.set_binary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_binary(v::GenericVariableRef)</code></pre><p>Add a constraint on the variable <code>v</code> that it must take values in the set <span>$\{0,1\}$</span>.</p><p>See also <a href="#JuMP.BinaryRef"><code>BinaryRef</code></a>, <a href="#JuMP.is_binary"><code>is_binary</code></a>, <a href="#JuMP.unset_binary"><code>unset_binary</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; is_binary(x)
+false
+
+julia&gt; set_binary(x)
+
+julia&gt; is_binary(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1385-L1408">source</a></section></article><h2 id="set_dual_start_value"><a class="docs-heading-anchor" href="#set_dual_start_value"><code>set_dual_start_value</code></a><a id="set_dual_start_value-1"></a><a class="docs-heading-anchor-permalink" href="#set_dual_start_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_dual_start_value" href="#JuMP.set_dual_start_value"><code>JuMP.set_dual_start_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_dual_start_value(con_ref::ConstraintRef, value)</code></pre><p>Set the dual start value (MOI attribute <code>ConstraintDualStart</code>) of the constraint <code>con_ref</code> to <code>value</code>. To remove a dual start value set it to <code>nothing</code>.</p><p>See also <a href="#JuMP.dual_start_value"><code>dual_start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L104-L111">source</a></section></article><h2 id="set_integer"><a class="docs-heading-anchor" href="#set_integer"><code>set_integer</code></a><a id="set_integer-1"></a><a class="docs-heading-anchor-permalink" href="#set_integer" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_integer" href="#JuMP.set_integer"><code>JuMP.set_integer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_integer(variable_ref::GenericVariableRef)</code></pre><p>Add an integrality constraint on the variable <code>variable_ref</code>.</p><p>See also <a href="#JuMP.IntegerRef"><code>IntegerRef</code></a>, <a href="#JuMP.is_integer"><code>is_integer</code></a>, <a href="#JuMP.unset_integer"><code>unset_integer</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; is_integer(x)
+false
+
+julia&gt; set_integer(x)
+
+julia&gt; is_integer(x)
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1253-L1275">source</a></section></article><h2 id="set_lower_bound"><a class="docs-heading-anchor" href="#set_lower_bound"><code>set_lower_bound</code></a><a id="set_lower_bound-1"></a><a class="docs-heading-anchor-permalink" href="#set_lower_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_lower_bound" href="#JuMP.set_lower_bound"><code>JuMP.set_lower_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_lower_bound(v::GenericVariableRef, lower::Number)</code></pre><p>Set the lower bound of a variable. If one does not exist, create a new lower bound constraint.</p><p>See also <a href="#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, <a href="#JuMP.has_lower_bound"><code>has_lower_bound</code></a>, <a href="#JuMP.lower_bound"><code>lower_bound</code></a>, <a href="#JuMP.delete_lower_bound"><code>delete_lower_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 1.0);
+
+julia&gt; lower_bound(x)
+1.0
+
+julia&gt; set_lower_bound(x, 2.0)
+
+julia&gt; lower_bound(x)
+2.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L697-L721">source</a></section></article><h2 id="set_name"><a class="docs-heading-anchor" href="#set_name"><code>set_name</code></a><a id="set_name-1"></a><a class="docs-heading-anchor-permalink" href="#set_name" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_name" href="#JuMP.set_name"><code>JuMP.set_name</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_name(con_ref::ConstraintRef, s::AbstractString)</code></pre><p>Set a constraint&#39;s name attribute.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L256-L260">source</a></section><section><div><pre><code class="language-julia hljs">set_name(v::GenericVariableRef, s::AbstractString)</code></pre><p>Set a variable&#39;s name attribute.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L534-L538">source</a></section></article><h2 id="set_nonlinear_dual_start_value"><a class="docs-heading-anchor" href="#set_nonlinear_dual_start_value"><code>set_nonlinear_dual_start_value</code></a><a id="set_nonlinear_dual_start_value-1"></a><a class="docs-heading-anchor-permalink" href="#set_nonlinear_dual_start_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_nonlinear_dual_start_value" href="#JuMP.set_nonlinear_dual_start_value"><code>JuMP.set_nonlinear_dual_start_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_nonlinear_dual_start_value(
+    model::Model,
+    start::Union{Nothing,Vector{Float64}},
+)</code></pre><p>Set the value of the MOI attribute <a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlockDualStart"><code>MOI.NLPBlockDualStart</code></a>.</p><p>The start vector corresponds to the Lagrangian duals of the nonlinear constraints, in the order given by <a href="#JuMP.all_nonlinear_constraints"><code>all_nonlinear_constraints</code></a>. That is, you must pass a single start vector corresponding to all of the nonlinear constraints in a single function call; you cannot set the dual start value of nonlinear constraints one-by-one. The example below demonstrates how to use <a href="#JuMP.all_nonlinear_constraints"><code>all_nonlinear_constraints</code></a> to create a mapping between the nonlinear constraint references and the start vector.</p><p>Pass <code>nothing</code> to unset a previous start.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; nl1 = @NLconstraint(model, x[1] &lt;= sqrt(x[2]));
+
+julia&gt; nl2 = @NLconstraint(model, x[1] &gt;= exp(x[2]));
+
+julia&gt; start = Dict(nl1 =&gt; -1.0, nl2 =&gt; 1.0);
+
+julia&gt; start_vector = [start[con] for con in all_nonlinear_constraints(model)]
+2-element Vector{Float64}:
+ -1.0
+  1.0
+
+julia&gt; set_nonlinear_dual_start_value(model, start_vector)
+
+julia&gt; nonlinear_dual_start_value(model)
+2-element Vector{Float64}:
+ -1.0
+  1.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L590-L633">source</a></section></article><h2 id="set_nonlinear_objective"><a class="docs-heading-anchor" href="#set_nonlinear_objective"><code>set_nonlinear_objective</code></a><a id="set_nonlinear_objective-1"></a><a class="docs-heading-anchor-permalink" href="#set_nonlinear_objective" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_nonlinear_objective" href="#JuMP.set_nonlinear_objective"><code>JuMP.set_nonlinear_objective</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_nonlinear_objective(
+    model::Model,
+    sense::MOI.OptimizationSense,
+    expr::Expr,
+)</code></pre><p>Set the nonlinear objective of <code>model</code> to the expression <code>expr</code>, with the optimization sense <code>sense</code>.</p><p>This function is most useful if the expression <code>expr</code> is generated programmatically, and you cannot use <a href="#JuMP.@NLobjective"><code>@NLobjective</code></a>.</p><p><strong>Notes</strong></p><ul><li>You must interpolate the variables directly into the expression <code>expr</code>.</li><li>You must use <code>MIN_SENSE</code> or <code>MAX_SENSE</code> instead of <code>Min</code> and <code>Max</code>.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; set_nonlinear_objective(model, MIN_SENSE, :($(x) + $(x)^2))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L136-L163">source</a></section></article><h2 id="set_normalized_coefficient"><a class="docs-heading-anchor" href="#set_normalized_coefficient"><code>set_normalized_coefficient</code></a><a id="set_normalized_coefficient-1"></a><a class="docs-heading-anchor-permalink" href="#set_normalized_coefficient" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_normalized_coefficient" href="#JuMP.set_normalized_coefficient"><code>JuMP.set_normalized_coefficient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_normalized_coefficient(con_ref::ConstraintRef, variable::GenericVariableRef, value)</code></pre><p>Set the coefficient of <code>variable</code> in the constraint <code>constraint</code> to <code>value</code>.</p><p>Note that prior to this step, JuMP will aggregate multiple terms containing the same variable. For example, given a constraint <code>2x + 3x &lt;= 2</code>, <code>set_normalized_coefficient(con, x, 4)</code> will create the constraint <code>4x &lt;= 2</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, con, 2x + 3x &lt;= 2)
+con : 5 x ≤ 2
+
+julia&gt; set_normalized_coefficient(con, x, 4)
+
+julia&gt; con
+con : 4 x ≤ 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L726-L749">source</a></section></article><h2 id="set_normalized_coefficients"><a class="docs-heading-anchor" href="#set_normalized_coefficients"><code>set_normalized_coefficients</code></a><a id="set_normalized_coefficients-1"></a><a class="docs-heading-anchor-permalink" href="#set_normalized_coefficients" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_normalized_coefficients" href="#JuMP.set_normalized_coefficients"><code>JuMP.set_normalized_coefficients</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_normalized_coefficients(
+    con_ref::ConstraintRef,
+    variable,
+    new_coefficients::Vector{Tuple{Int64,T}},
+)</code></pre><p>Set the coefficients of <code>variable</code> in the constraint <code>con_ref</code> to <code>new_coefficients</code>, where each element in <code>new_coefficients</code> is a tuple which maps the row to a new coefficient.</p><p>Note that prior to this step, during constraint creation, JuMP will aggregate multiple terms containing the same variable.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, con, [2x + 3x, 4x] in MOI.Nonnegatives(2))
+con : [5 x, 4 x] ∈ MathOptInterface.Nonnegatives(2)
+
+julia&gt; set_normalized_coefficients(con, x, [(1, 2.0), (2, 5.0)])
+
+julia&gt; con
+con : [2 x, 5 x] ∈ MathOptInterface.Nonnegatives(2)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L769-L797">source</a></section></article><h2 id="set_normalized_rhs"><a class="docs-heading-anchor" href="#set_normalized_rhs"><code>set_normalized_rhs</code></a><a id="set_normalized_rhs-1"></a><a class="docs-heading-anchor-permalink" href="#set_normalized_rhs" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_normalized_rhs" href="#JuMP.set_normalized_rhs"><code>JuMP.set_normalized_rhs</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_normalized_rhs(con_ref::ConstraintRef, value)</code></pre><p>Set the right-hand side term of <code>constraint</code> to <code>value</code>.</p><p>Note that prior to this step, JuMP will aggregate all constant terms onto the right-hand side of the constraint. For example, given a constraint <code>2x + 1 &lt;= 2</code>, <code>set_normalized_rhs(con, 4)</code> will create the constraint <code>2x &lt;= 4</code>, not <code>2x + 1 &lt;= 4</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con, 2x + 1 &lt;= 2)
+con : 2 x ≤ 1
+
+julia&gt; set_normalized_rhs(con, 4)
+
+julia&gt; con
+con : 2 x ≤ 4</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L832-L855">source</a></section></article><h2 id="set_objective"><a class="docs-heading-anchor" href="#set_objective"><code>set_objective</code></a><a id="set_objective-1"></a><a class="docs-heading-anchor-permalink" href="#set_objective" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_objective" href="#JuMP.set_objective"><code>JuMP.set_objective</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_objective(model::AbstractModel, sense::MOI.OptimizationSense, func)</code></pre><p>The functional equivalent of the <a href="#JuMP.@objective"><code>@objective</code></a> macro.</p><p>Sets the objective sense and objective function simultaneously, and is equivalent to calling <a href="#JuMP.set_objective_sense"><code>set_objective_sense</code></a> and <a href="#JuMP.set_objective_function"><code>set_objective_function</code></a> separately.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; set_objective(model, MIN_SENSE, x)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L154-L173">source</a></section></article><h2 id="set_objective_coefficient"><a class="docs-heading-anchor" href="#set_objective_coefficient"><code>set_objective_coefficient</code></a><a id="set_objective_coefficient-1"></a><a class="docs-heading-anchor-permalink" href="#set_objective_coefficient" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_objective_coefficient" href="#JuMP.set_objective_coefficient"><code>JuMP.set_objective_coefficient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_objective_coefficient(model::GenericModel, variable::GenericVariableRef, coefficient::Real)</code></pre><p>Set the linear objective coefficient associated with <code>Variable</code> to <code>coefficient</code>.</p><p>Note: this function will throw an error if a nonlinear objective is set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L249-L255">source</a></section></article><h2 id="set_objective_function"><a class="docs-heading-anchor" href="#set_objective_function"><code>set_objective_function</code></a><a id="set_objective_function-1"></a><a class="docs-heading-anchor-permalink" href="#set_objective_function" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_objective_function" href="#JuMP.set_objective_function"><code>JuMP.set_objective_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_objective_function(model::GenericModel, func::MOI.AbstractFunction)
+set_objective_function(model::GenericModel, func::AbstractJuMPScalar)
+set_objective_function(model::GenericModel, func::Real)
+set_objective_function(model::GenericModel, func::Vector{&lt;:AbstractJuMPScalar})</code></pre><p>Sets the objective function of the model to the given function. See <a href="#JuMP.set_objective_sense"><code>set_objective_sense</code></a> to set the objective sense. These are low-level functions; the recommended way to set the objective is with the <a href="#JuMP.@objective"><code>@objective</code></a> macro.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L96-L106">source</a></section></article><h2 id="set_objective_sense"><a class="docs-heading-anchor" href="#set_objective_sense"><code>set_objective_sense</code></a><a id="set_objective_sense-1"></a><a class="docs-heading-anchor-permalink" href="#set_objective_sense" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_objective_sense" href="#JuMP.set_objective_sense"><code>JuMP.set_objective_sense</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_objective_sense(model::GenericModel, sense::MOI.OptimizationSense)</code></pre><p>Sets the objective sense of the model to the given sense. See <a href="#JuMP.set_objective_function"><code>set_objective_function</code></a> to set the objective function. These are low-level functions; the recommended way to set the objective is with the <a href="#JuMP.@objective"><code>@objective</code></a> macro.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/objective.jl#L84-L91">source</a></section></article><h2 id="set_optimize_hook"><a class="docs-heading-anchor" href="#set_optimize_hook"><code>set_optimize_hook</code></a><a id="set_optimize_hook-1"></a><a class="docs-heading-anchor-permalink" href="#set_optimize_hook" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_optimize_hook" href="#JuMP.set_optimize_hook"><code>JuMP.set_optimize_hook</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_optimize_hook(model::GenericModel, f::Union{Function,Nothing})</code></pre><p>Set the function <code>f</code> as the optimize hook for <code>model</code>.</p><p><code>f</code> should have a signature <code>f(model::GenericModel; kwargs...)</code>, where the <code>kwargs</code> are those passed to <a href="#JuMP.optimize!"><code>optimize!</code></a>.</p><p><strong>Notes</strong></p><ul><li>The optimize hook should generally modify the model, or some external state in some way, and then call <code>optimize!(model; ignore_optimize_hook = true)</code> to optimize the problem, bypassing the hook.</li><li>Use <code>set_optimize_hook(model, nothing)</code> to unset an optimize hook.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; function my_hook(model::Model; kwargs...)
+           println(kwargs)
+           println(&quot;Calling with `ignore_optimize_hook = true`&quot;)
+           optimize!(model; ignore_optimize_hook = true)
+           return
+       end
+my_hook (generic function with 1 method)
+
+julia&gt; set_optimize_hook(model, my_hook)
+my_hook (generic function with 1 method)
+
+julia&gt; optimize!(model; test_arg = true)
+Base.Pairs{Symbol, Bool, Tuple{Symbol}, NamedTuple{(:test_arg,), Tuple{Bool}}}(:test_arg =&gt; 1)
+Calling with `ignore_optimize_hook = true`
+ERROR: NoOptimizer()
+[...]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L951-L988">source</a></section></article><h2 id="set_optimizer"><a class="docs-heading-anchor" href="#set_optimizer"><code>set_optimizer</code></a><a id="set_optimizer-1"></a><a class="docs-heading-anchor-permalink" href="#set_optimizer" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_optimizer" href="#JuMP.set_optimizer"><code>JuMP.set_optimizer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_optimizer(
+    model::GenericModel,
+    optimizer_factory;
+    add_bridges::Bool = true,
+)</code></pre><p>Creates an empty <code>MathOptInterface.AbstractOptimizer</code> instance by calling <code>optimizer_factory()</code> and sets it as the optimizer of <code>model</code>. Specifically, <code>optimizer_factory</code> must be callable with zero arguments and return an empty <code>MathOptInterface.AbstractOptimizer</code>.</p><p>If <code>add_bridges</code> is true, constraints and objectives that are not supported by the optimizer are automatically bridged to equivalent supported formulation. Passing <code>add_bridges = false</code> can improve performance if the solver natively supports all of the elements in <code>model</code>.</p><p>See <a href="#JuMP.set_attribute"><code>set_attribute</code></a> for setting solver-specific parameters of the optimizer.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = Model();
+
+julia&gt; set_optimizer(model, () -&gt; HiGHS.Optimizer())
+
+julia&gt; set_optimizer(model, HiGHS.Optimizer; add_bridges = false)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L326-L357">source</a></section></article><h2 id="set_optimizer_attribute"><a class="docs-heading-anchor" href="#set_optimizer_attribute"><code>set_optimizer_attribute</code></a><a id="set_optimizer_attribute-1"></a><a class="docs-heading-anchor-permalink" href="#set_optimizer_attribute" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_optimizer_attribute" href="#JuMP.set_optimizer_attribute"><code>JuMP.set_optimizer_attribute</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_optimizer_attribute(
+    model::Union{GenericModel,MOI.OptimizerWithAttributes},
+    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},
+    value,
+)</code></pre><p>Set the solver-specific attribute <code>attr</code> in <code>model</code> to <code>value</code>.</p><p>If <code>attr</code> is an <code>AbstractString</code>, this is equivalent to <code>set_optimizer_attribute(model, MOI.RawOptimizerAttribute(name), value)</code>.</p><div class="admonition is-compat"><header class="admonition-header">Compat</header><div class="admonition-body"><p>This method will remain in all v1.X releases of JuMP, but it may be removed in a future v2.0 release. We recommend using <a href="#JuMP.set_attribute"><code>set_attribute</code></a> instead.</p></div></div><p>See also: <a href="#JuMP.set_optimizer_attributes"><code>set_optimizer_attributes</code></a>, <a href="#JuMP.get_optimizer_attribute"><code>get_optimizer_attribute</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; set_optimizer_attribute(model, MOI.Silent(), true)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L51-L76">source</a></section></article><h2 id="set_optimizer_attributes"><a class="docs-heading-anchor" href="#set_optimizer_attributes"><code>set_optimizer_attributes</code></a><a id="set_optimizer_attributes-1"></a><a class="docs-heading-anchor-permalink" href="#set_optimizer_attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_optimizer_attributes" href="#JuMP.set_optimizer_attributes"><code>JuMP.set_optimizer_attributes</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_optimizer_attributes(
+    model::Union{GenericModel,MOI.OptimizerWithAttributes},
+    pairs::Pair...,
+)</code></pre><p>Given a list of <code>attribute =&gt; value</code> pairs, calls <code>set_optimizer_attribute(model, attribute, value)</code> for each pair.</p><div class="admonition is-compat"><header class="admonition-header">Compat</header><div class="admonition-body"><p>This method will remain in all v1.X releases of JuMP, but it may be removed in a future v2.0 release. We recommend using <a href="#JuMP.set_attributes"><code>set_attributes</code></a> instead.</p></div></div><p>See also: <a href="#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a>, <a href="#JuMP.get_optimizer_attribute"><code>get_optimizer_attribute</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; model = Model(Ipopt.Optimizer);
+
+julia&gt; set_optimizer_attributes(model, &quot;tol&quot; =&gt; 1e-4, &quot;max_iter&quot; =&gt; 100)</code></pre><p>is equivalent to:</p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; model = Model(Ipopt.Optimizer);
+
+julia&gt; set_optimizer_attribute(model, &quot;tol&quot;, 1e-4)
+
+julia&gt; set_optimizer_attribute(model, &quot;max_iter&quot;, 100)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L79-L113">source</a></section></article><h2 id="set_parameter_value"><a class="docs-heading-anchor" href="#set_parameter_value"><code>set_parameter_value</code></a><a id="set_parameter_value-1"></a><a class="docs-heading-anchor-permalink" href="#set_parameter_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_parameter_value" href="#JuMP.set_parameter_value"><code>JuMP.set_parameter_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_parameter_value(x::GenericVariableRef, value)</code></pre><p>Update the parameter constraint on the variable <code>x</code> to <code>value</code>.</p><p>Errors if <code>x</code> is not a parameter.</p><p>See also <a href="#JuMP.ParameterRef"><code>ParameterRef</code></a>, <a href="#JuMP.is_parameter"><code>is_parameter</code></a>, <a href="#JuMP.parameter_value"><code>parameter_value</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, p in Parameter(2))
+p
+
+julia&gt; parameter_value(p)
+2.0
+
+julia&gt; set_parameter_value(p, 2.5)
+
+julia&gt; parameter_value(p)
+2.5</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1557-L1583">source</a></section></article><h2 id="set_silent"><a class="docs-heading-anchor" href="#set_silent"><code>set_silent</code></a><a id="set_silent-1"></a><a class="docs-heading-anchor-permalink" href="#set_silent" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_silent" href="#JuMP.set_silent"><code>JuMP.set_silent</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_silent(model::GenericModel)</code></pre><p>Takes precedence over any other attribute controlling verbosity and requires the solver to produce no output.</p><p>See also: <a href="#JuMP.unset_silent"><code>unset_silent</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L154-L161">source</a></section></article><h2 id="set_start_value"><a class="docs-heading-anchor" href="#set_start_value"><code>set_start_value</code></a><a id="set_start_value-1"></a><a class="docs-heading-anchor-permalink" href="#set_start_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_start_value" href="#JuMP.set_start_value"><code>JuMP.set_start_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_start_value(con_ref::ConstraintRef, value)</code></pre><p>Set the primal start value (<a href="../../moi/reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>MOI.ConstraintPrimalStart</code></a>) of the constraint <code>con_ref</code> to <code>value</code>. To remove a primal start value set it to <code>nothing</code>.</p><p>See also <a href="#JuMP.start_value"><code>start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L158-L166">source</a></section><section><div><pre><code class="language-julia hljs">set_start_value(variable::GenericVariableRef, value::Union{Real,Nothing})</code></pre><p>Set the start value (MOI attribute <code>VariablePrimalStart</code>) of the <code>variable</code> to <code>value</code>.</p><p>Pass <code>nothing</code> to unset the start value.</p><p>Note: <code>VariablePrimalStart</code>s are sometimes called &quot;MIP-starts&quot; or &quot;warmstarts&quot;.</p><p>See also <a href="#JuMP.start_value"><code>start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1657-L1668">source</a></section></article><h2 id="set_start_values"><a class="docs-heading-anchor" href="#set_start_values"><code>set_start_values</code></a><a id="set_start_values-1"></a><a class="docs-heading-anchor-permalink" href="#set_start_values" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_start_values" href="#JuMP.set_start_values"><code>JuMP.set_start_values</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_start_values(
+    model::GenericModel;
+    variable_primal_start::Union{Nothing,Function} = value,
+    constraint_primal_start::Union{Nothing,Function} = value,
+    constraint_dual_start::Union{Nothing,Function} = dual,
+    nonlinear_dual_start::Union{Nothing,Function} = nonlinear_dual_start_value,
+)</code></pre><p>Set the primal and dual starting values in <code>model</code> using the functions provided.</p><p>If any keyword argument is <code>nothing</code>, the corresponding start value is skipped.</p><p>If the optimizer does not support setting the starting value, the value will be skipped.</p><p><strong><code>variable_primal_start</code></strong></p><p>This function controls the primal starting solution for the variables. It is equivalent to calling <a href="#JuMP.set_start_value"><code>set_start_value</code></a> for each variable, or setting the <a href="../../moi/reference/variables/#MathOptInterface.VariablePrimalStart"><code>MOI.VariablePrimalStart</code></a> attribute.</p><p>If it is a function, it must have the form <code>variable_primal_start(x::VariableRef)</code> that maps each variable <code>x</code> to the starting primal value.</p><p>The default is <a href="#JuMP.value"><code>value</code></a>.</p><p><strong><code>constraint_primal_start</code></strong></p><p>This function controls the primal starting solution for the constraints. It is equivalent to calling <a href="#JuMP.set_start_value"><code>set_start_value</code></a> for each constraint, or setting the <a href="../../moi/reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>MOI.ConstraintPrimalStart</code></a> attribute.</p><p>If it is a function, it must have the form <code>constraint_primal_start(ci::ConstraintRef)</code> that maps each constraint <code>ci</code> to the starting primal value.</p><p>The default is <a href="#JuMP.value"><code>value</code></a>.</p><p><strong><code>constraint_dual_start</code></strong></p><p>This function controls the dual starting solution for the constraints. It is equivalent to calling <a href="#JuMP.set_dual_start_value"><code>set_dual_start_value</code></a> for each constraint, or setting the <a href="../../moi/reference/constraints/#MathOptInterface.ConstraintDualStart"><code>MOI.ConstraintDualStart</code></a> attribute.</p><p>If it is a function, it must have the form <code>constraint_dual_start(ci::ConstraintRef)</code> that maps each constraint <code>ci</code> to the starting dual value.</p><p>The default is <a href="#JuMP.dual"><code>dual</code></a>.</p><p><strong><code>nonlinear_dual_start</code></strong></p><p>This function controls the dual starting solution for the nonlinear constraints It is equivalent to calling <a href="#JuMP.set_nonlinear_dual_start_value"><code>set_nonlinear_dual_start_value</code></a>.</p><p>If it is a function, it must have the form <code>nonlinear_dual_start(model::GenericModel)</code> that returns a vector corresponding to the dual start of the constraints.</p><p>The default is <a href="#JuMP.nonlinear_dual_start_value"><code>nonlinear_dual_start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L1082-L1140">source</a></section></article><h2 id="set_string_names_on_creation"><a class="docs-heading-anchor" href="#set_string_names_on_creation"><code>set_string_names_on_creation</code></a><a id="set_string_names_on_creation-1"></a><a class="docs-heading-anchor-permalink" href="#set_string_names_on_creation" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_string_names_on_creation" href="#JuMP.set_string_names_on_creation"><code>JuMP.set_string_names_on_creation</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_string_names_on_creation(model::GenericModel, value::Bool)</code></pre><p>Set the default argument of the <code>set_string_name</code> keyword in the <a href="#JuMP.@variable"><code>@variable</code></a> and <a href="#JuMP.@constraint"><code>@constraint</code></a> macros to <code>value</code>. This is used to determine whether to assign <code>String</code> names to all variables and constraints in <code>model</code>.</p><p>By default, <code>value</code> is <code>true</code>. However, for larger models calling <code>set_string_names_on_creation(model, false)</code> can improve performance at the cost of reducing the readability of printing and solver log messages.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L540-L551">source</a></section></article><h2 id="set_time_limit_sec"><a class="docs-heading-anchor" href="#set_time_limit_sec"><code>set_time_limit_sec</code></a><a id="set_time_limit_sec-1"></a><a class="docs-heading-anchor-permalink" href="#set_time_limit_sec" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_time_limit_sec" href="#JuMP.set_time_limit_sec"><code>JuMP.set_time_limit_sec</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_time_limit_sec(model::GenericModel, limit::Float64)</code></pre><p>Set the time limit (in seconds) of the solver.</p><p>Can be unset using <a href="#JuMP.unset_time_limit_sec"><code>unset_time_limit_sec</code></a> or with <code>limit</code> set to <code>nothing</code>.</p><p>See also: <a href="#JuMP.unset_time_limit_sec"><code>unset_time_limit_sec</code></a>, <a href="#JuMP.time_limit_sec"><code>time_limit_sec</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L178-L187">source</a></section></article><h2 id="set_upper_bound"><a class="docs-heading-anchor" href="#set_upper_bound"><code>set_upper_bound</code></a><a id="set_upper_bound-1"></a><a class="docs-heading-anchor-permalink" href="#set_upper_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_upper_bound" href="#JuMP.set_upper_bound"><code>JuMP.set_upper_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_upper_bound(v::GenericVariableRef, upper::Number)</code></pre><p>Set the upper bound of a variable. If one does not exist, create an upper bound constraint.</p><p>See also <a href="#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, <a href="#JuMP.has_upper_bound"><code>has_upper_bound</code></a>, <a href="#JuMP.upper_bound"><code>upper_bound</code></a>, <a href="#JuMP.delete_upper_bound"><code>delete_upper_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &lt;= 1.0);
+
+julia&gt; upper_bound(x)
+1.0
+
+julia&gt; set_upper_bound(x, 2.0)
+
+julia&gt; upper_bound(x)
+2.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L868-L892">source</a></section></article><h2 id="set_value"><a class="docs-heading-anchor" href="#set_value"><code>set_value</code></a><a id="set_value-1"></a><a class="docs-heading-anchor-permalink" href="#set_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.set_value" href="#JuMP.set_value"><code>JuMP.set_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_value(p::NonlinearParameter, v::Number)</code></pre><p>Store the value <code>v</code> in the nonlinear parameter <code>p</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @NLparameter(model, p == 0)
+p == 0.0
+
+julia&gt; set_value(p, 5)
+5
+
+julia&gt; value(p)
+5.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L250-L269">source</a></section></article><h2 id="shadow_price"><a class="docs-heading-anchor" href="#shadow_price"><code>shadow_price</code></a><a id="shadow_price-1"></a><a class="docs-heading-anchor-permalink" href="#shadow_price" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.shadow_price" href="#JuMP.shadow_price"><code>JuMP.shadow_price</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">shadow_price(con_ref::ConstraintRef)</code></pre><p>Return the change in the objective from an infinitesimal relaxation of the constraint.</p><p>This value is computed from <a href="#JuMP.dual"><code>dual</code></a> and can be queried only when <code>has_duals</code> is <code>true</code> and the objective sense is <code>MIN_SENSE</code> or <code>MAX_SENSE</code> (not <code>FEASIBILITY_SENSE</code>). For linear constraints, the shadow prices differ at most in sign from the <code>dual</code> value depending on the objective sense.</p><p>See also <a href="#JuMP.reduced_cost"><code>reduced_cost</code></a>.</p><p><strong>Notes</strong></p><ul><li>The function simply translates signs from <code>dual</code> and does not validate the conditions needed to guarantee the sensitivity interpretation of the shadow price. The caller is responsible, e.g., for checking whether the solver converged to an optimal primal-dual pair or a proof of infeasibility.</li><li>The computation is based on the current objective sense of the model. If this has changed since the last solve, the results will be incorrect.</li><li>Relaxation of equality constraints (and hence the shadow price) is defined based on which sense of the equality constraint is active.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1069-L1092">source</a></section></article><h2 id="shape"><a class="docs-heading-anchor" href="#shape"><code>shape</code></a><a id="shape-1"></a><a class="docs-heading-anchor-permalink" href="#shape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.shape" href="#JuMP.shape"><code>JuMP.shape</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">shape(c::AbstractConstraint)::AbstractShape</code></pre><p>Return the shape of the constraint <code>c</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/shapes.jl#L101-L105">source</a></section></article><h2 id="show_backend_summary"><a class="docs-heading-anchor" href="#show_backend_summary"><code>show_backend_summary</code></a><a id="show_backend_summary-1"></a><a class="docs-heading-anchor-permalink" href="#show_backend_summary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.show_backend_summary" href="#JuMP.show_backend_summary"><code>JuMP.show_backend_summary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">show_backend_summary(io::IO, model::GenericModel)</code></pre><p>Print a summary of the optimizer backing <code>model</code>.</p><p><code>AbstractModel</code>s should implement this method.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L262-L268">source</a></section></article><h2 id="show_constraints_summary"><a class="docs-heading-anchor" href="#show_constraints_summary"><code>show_constraints_summary</code></a><a id="show_constraints_summary-1"></a><a class="docs-heading-anchor-permalink" href="#show_constraints_summary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.show_constraints_summary" href="#JuMP.show_constraints_summary"><code>JuMP.show_constraints_summary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">show_constraints_summary(io::IO, model::AbstractModel)</code></pre><p>Write to <code>io</code> a summary of the number of constraints.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L245-L249">source</a></section></article><h2 id="show_objective_function_summary"><a class="docs-heading-anchor" href="#show_objective_function_summary"><code>show_objective_function_summary</code></a><a id="show_objective_function_summary-1"></a><a class="docs-heading-anchor-permalink" href="#show_objective_function_summary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.show_objective_function_summary" href="#JuMP.show_objective_function_summary"><code>JuMP.show_objective_function_summary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">show_objective_function_summary(io::IO, model::AbstractModel)</code></pre><p>Write to <code>io</code> a summary of the objective function type.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/print.jl#L229-L233">source</a></section></article><h2 id="simplex_iterations"><a class="docs-heading-anchor" href="#simplex_iterations"><code>simplex_iterations</code></a><a id="simplex_iterations-1"></a><a class="docs-heading-anchor-permalink" href="#simplex_iterations" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.simplex_iterations" href="#JuMP.simplex_iterations"><code>JuMP.simplex_iterations</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">simplex_iterations(model::GenericModel)</code></pre><p>Gets the cumulative number of simplex iterations during the most-recent optimization.</p><p>Solvers must implement <code>MOI.SimplexIterations()</code> to use this function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L588-L595">source</a></section></article><h2 id="solution_summary"><a class="docs-heading-anchor" href="#solution_summary"><code>solution_summary</code></a><a id="solution_summary-1"></a><a class="docs-heading-anchor-permalink" href="#solution_summary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.solution_summary" href="#JuMP.solution_summary"><code>JuMP.solution_summary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">solution_summary(model::GenericModel; result::Int = 1, verbose::Bool = false)</code></pre><p>Return a struct that can be used print a summary of the solution in result <code>result</code>.</p><p>If <code>verbose=true</code>, write out the primal solution for every variable and the dual solution for every constraint, excluding those with empty names.</p><p><strong>Example</strong></p><p>When called at the REPL, the summary is automatically printed:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; solution_summary(model)
+* Solver : No optimizer attached.
+
+* Status
+  Result count       : 0
+  Termination status : OPTIMIZE_NOT_CALLED
+  Message from the solver:
+  &quot;optimize not called&quot;
+
+* Candidate solution (result #1)
+  Primal status      : NO_SOLUTION
+  Dual status        : NO_SOLUTION
+
+* Work counters</code></pre><p>Use <code>print</code> to force the printing of the summary from inside a function:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; function foo(model)
+           print(solution_summary(model))
+           return
+       end
+foo (generic function with 1 method)
+
+julia&gt; foo(model)
+* Solver : No optimizer attached.
+
+* Status
+  Result count       : 0
+  Termination status : OPTIMIZE_NOT_CALLED
+  Message from the solver:
+  &quot;optimize not called&quot;
+
+* Candidate solution (result #1)
+  Primal status      : NO_SOLUTION
+  Dual status        : NO_SOLUTION
+
+* Work counters</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/solution_summary.jl#L38-L94">source</a></section></article><h2 id="solve_time"><a class="docs-heading-anchor" href="#solve_time"><code>solve_time</code></a><a id="solve_time-1"></a><a class="docs-heading-anchor-permalink" href="#solve_time" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.solve_time" href="#JuMP.solve_time"><code>JuMP.solve_time</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">solve_time(model::GenericModel)</code></pre><p>If available, returns the solve time reported by the solver. Returns &quot;ArgumentError: ModelLike of type <code>Solver.Optimizer</code> does not support accessing the attribute MathOptInterface.SolveTimeSec()&quot; if the attribute is not implemented.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L576-L583">source</a></section></article><h2 id="solver_name"><a class="docs-heading-anchor" href="#solver_name"><code>solver_name</code></a><a id="solver_name-1"></a><a class="docs-heading-anchor-permalink" href="#solver_name" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.solver_name" href="#JuMP.solver_name"><code>JuMP.solver_name</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">solver_name(model::GenericModel)</code></pre><p>If available, returns the <code>SolverName</code> property of the underlying optimizer.</p><p>Returns <code>&quot;No optimizer attached&quot;</code> in <code>AUTOMATIC</code> or <code>MANUAL</code> modes when no optimizer is attached.</p><p>Returns <code>&quot;SolverName() attribute not implemented by the optimizer.&quot;</code> if the attribute is not implemented.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L232-L242">source</a></section></article><h2 id="start_value"><a class="docs-heading-anchor" href="#start_value"><code>start_value</code></a><a id="start_value-1"></a><a class="docs-heading-anchor-permalink" href="#start_value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.start_value" href="#JuMP.start_value"><code>JuMP.start_value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">start_value(con_ref::ConstraintRef)</code></pre><p>Return the primal start value (<a href="../../moi/reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>MOI.ConstraintPrimalStart</code></a>) of the constraint <code>con_ref</code>.</p><p>Note: If no primal start value has been set, <code>start_value</code> will return <code>nothing</code>.</p><p>See also <a href="#JuMP.set_start_value"><code>set_start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L214-L224">source</a></section><section><div><pre><code class="language-julia hljs">start_value(v::GenericVariableRef)</code></pre><p>Return the start value (MOI attribute <code>VariablePrimalStart</code>) of the variable <code>v</code>.</p><p>Note: <code>VariablePrimalStart</code>s are sometimes called &quot;MIP-starts&quot; or &quot;warmstarts&quot;.</p><p>See also <a href="#JuMP.set_start_value"><code>set_start_value</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1631-L1640">source</a></section></article><h2 id="termination_status"><a class="docs-heading-anchor" href="#termination_status"><code>termination_status</code></a><a id="termination_status-1"></a><a class="docs-heading-anchor-permalink" href="#termination_status" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.termination_status" href="#JuMP.termination_status"><code>JuMP.termination_status</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">termination_status(model::GenericModel)</code></pre><p>Return a <a href="../../moi/reference/models/#MathOptInterface.TerminationStatusCode"><code>MOI.TerminationStatusCode</code></a> describing why the solver stopped (i.e., the <a href="../../moi/reference/models/#MathOptInterface.TerminationStatus"><code>MOI.TerminationStatus</code></a> attribute).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L497-L502">source</a></section></article><h2 id="time_limit_sec"><a class="docs-heading-anchor" href="#time_limit_sec"><code>time_limit_sec</code></a><a id="time_limit_sec-1"></a><a class="docs-heading-anchor-permalink" href="#time_limit_sec" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.time_limit_sec" href="#JuMP.time_limit_sec"><code>JuMP.time_limit_sec</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">time_limit_sec(model::GenericModel)</code></pre><p>Return the time limit (in seconds) of the <code>model</code>.</p><p>Returns <code>nothing</code> if unset.</p><p>See also: <a href="#JuMP.set_time_limit_sec"><code>set_time_limit_sec</code></a>, <a href="#JuMP.unset_time_limit_sec"><code>unset_time_limit_sec</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L207-L215">source</a></section></article><h2 id="triangle_vec"><a class="docs-heading-anchor" href="#triangle_vec"><code>triangle_vec</code></a><a id="triangle_vec-1"></a><a class="docs-heading-anchor-permalink" href="#triangle_vec" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.triangle_vec" href="#JuMP.triangle_vec"><code>JuMP.triangle_vec</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">triangle_vec(matrix::Matrix)</code></pre><p>Return the upper triangle of a matrix concatenated into a vector in the order required by JuMP and MathOptInterface for <code>Triangle</code> sets.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, X[1:3, 1:3], Symmetric);
+
+julia&gt; @variable(model, t)
+t
+
+julia&gt; @constraint(model, [t; triangle_vec(X)] in MOI.RootDetConeTriangle(3))
+[t, X[1,1], X[1,2], X[2,2], X[1,3], X[2,3], X[3,3]] ∈ MathOptInterface.RootDetConeTriangle(3)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L174-L193">source</a></section></article><h2 id="unfix"><a class="docs-heading-anchor" href="#unfix"><code>unfix</code></a><a id="unfix-1"></a><a class="docs-heading-anchor-permalink" href="#unfix" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.unfix" href="#JuMP.unfix"><code>JuMP.unfix</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unfix(v::GenericVariableRef)</code></pre><p>Delete the fixing constraint of a variable.</p><p>Error if one does not exist.</p><p>See also <a href="#JuMP.FixRef"><code>FixRef</code></a>, <a href="#JuMP.is_fixed"><code>is_fixed</code></a>, <a href="#JuMP.fix_value"><code>fix_value</code></a>, <a href="#JuMP.fix"><code>fix</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x == 1);
+
+julia&gt; is_fixed(x)
+true
+
+julia&gt; unfix(x)
+
+julia&gt; is_fixed(x)
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1132-L1157">source</a></section></article><h2 id="unregister"><a class="docs-heading-anchor" href="#unregister"><code>unregister</code></a><a id="unregister-1"></a><a class="docs-heading-anchor-permalink" href="#unregister" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.unregister" href="#JuMP.unregister"><code>JuMP.unregister</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unregister(model::GenericModel, key::Symbol)</code></pre><p>Unregister the name <code>key</code> from <code>model</code> so that a new variable, constraint, or expression can be created with the same key.</p><p>Note that this will not delete the object <code>model[key]</code>; it will just remove the reference at <code>model[key]</code>. To delete the object, use <a href="#JuMP.delete"><code>delete</code></a> as well.</p><p>See also: <a href="#JuMP.delete"><code>delete</code></a>, <a href="#JuMP.object_dictionary"><code>object_dictionary</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, x)
+ERROR: An object of name x is already attached to this model. If this
+    is intended, consider using the anonymous construction syntax, e.g.,
+    `x = @variable(model, [1:N], ...)` where the name of the object does
+    not appear inside the macro.
+
+    Alternatively, use `unregister(model, :x)` to first unregister
+    the existing name from the model. Note that this will not delete the
+    object; it will just remove the reference at `model[:x]`.
+
+Stacktrace:
+[...]
+
+julia&gt; num_variables(model)
+1
+
+julia&gt; unregister(model, :x)
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; num_variables(model)
+2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L857-L900">source</a></section></article><h2 id="unsafe_backend"><a class="docs-heading-anchor" href="#unsafe_backend"><code>unsafe_backend</code></a><a id="unsafe_backend-1"></a><a class="docs-heading-anchor-permalink" href="#unsafe_backend" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.unsafe_backend" href="#JuMP.unsafe_backend"><code>JuMP.unsafe_backend</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unsafe_backend(model::GenericModel)</code></pre><p>Return the innermost optimizer associated with the JuMP model <code>model</code>.</p><p><strong>This function should only be used by advanced users looking to access low-level solver-specific functionality. It has a high-risk of incorrect usage. We strongly suggest you use the alternative suggested below.</strong></p><p>See also: <a href="#JuMP.backend"><code>backend</code></a>.</p><p><strong>Unsafe behavior</strong></p><p>This function is unsafe for two main reasons.</p><p>First, the formulation and order of variables and constraints in the unsafe backend may be different to the variables and constraints in <code>model</code>. This can happen because of bridges, or because the solver requires the variables or constraints in a specific order. In addition, the variable or constraint index returned by <a href="#JuMP.index"><code>index</code></a> at the JuMP level may be different to the index of the corresponding variable or constraint in the <code>unsafe_backend</code>. There is no solution to this. Use the alternative suggested below instead.</p><p>Second, the <code>unsafe_backend</code> may be empty, or lack some modifications made to the JuMP model. Thus, before calling <code>unsafe_backend</code> you should first call <a href="../../moi/submodules/Utilities/reference/#MathOptInterface.Utilities.attach_optimizer"><code>MOI.Utilities.attach_optimizer</code></a> to ensure that the backend is synchronized with the JuMP model.</p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = Model(HiGHS.Optimizer)
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS
+
+julia&gt; MOI.Utilities.attach_optimizer(model)
+
+julia&gt; inner = unsafe_backend(model)
+A HiGHS model with 0 columns and 0 rows.</code></pre><p>Moreover, if you modify the JuMP model, the reference you have to the backend (i.e., <code>inner</code> in the example above) may be out-dated, and you should call <a href="../../moi/submodules/Utilities/reference/#MathOptInterface.Utilities.attach_optimizer"><code>MOI.Utilities.attach_optimizer</code></a> again.</p><p>This function is also unsafe in the reverse direction: if you modify the unsafe backend, e.g., by adding a new constraint to <code>inner</code>, the changes may be silently discarded by JuMP when the JuMP <code>model</code> is modified or solved.</p><p><strong>Alternative</strong></p><p>Instead of <code>unsafe_backend</code>, create a model using <a href="#JuMP.direct_model"><code>direct_model</code></a> and call <a href="#JuMP.backend"><code>backend</code></a> instead.</p><p>For example, instead of:</p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 0)
+x
+
+julia&gt; MOI.Utilities.attach_optimizer(model)
+
+julia&gt; highs = unsafe_backend(model)
+A HiGHS model with 1 columns and 0 rows.</code></pre><p>Use:</p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; model = direct_model(HiGHS.Optimizer());
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 0)
+x
+
+julia&gt; highs = backend(model)  # No need to call `attach_optimizer`.
+A HiGHS model with 1 columns and 0 rows.</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L412-L503">source</a></section></article><h2 id="unset_binary"><a class="docs-heading-anchor" href="#unset_binary"><code>unset_binary</code></a><a id="unset_binary-1"></a><a class="docs-heading-anchor-permalink" href="#unset_binary" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.unset_binary" href="#JuMP.unset_binary"><code>JuMP.unset_binary</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unset_binary(variable_ref::GenericVariableRef)</code></pre><p>Remove the binary constraint on the variable <code>variable_ref</code>.</p><p>See also <a href="#JuMP.BinaryRef"><code>BinaryRef</code></a>, <a href="#JuMP.is_binary"><code>is_binary</code></a>, <a href="#JuMP.set_binary"><code>set_binary</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Bin);
+
+julia&gt; is_binary(x)
+true
+
+julia&gt; unset_binary(x)
+
+julia&gt; is_binary(x)
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1429-L1451">source</a></section></article><h2 id="unset_integer"><a class="docs-heading-anchor" href="#unset_integer"><code>unset_integer</code></a><a id="unset_integer-1"></a><a class="docs-heading-anchor-permalink" href="#unset_integer" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.unset_integer" href="#JuMP.unset_integer"><code>JuMP.unset_integer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unset_integer(variable_ref::GenericVariableRef)</code></pre><p>Remove the integrality constraint on the variable <code>variable_ref</code>.</p><p>Errors if one does not exist.</p><p>See also <a href="#JuMP.IntegerRef"><code>IntegerRef</code></a>, <a href="#JuMP.is_integer"><code>is_integer</code></a>, <a href="#JuMP.set_integer"><code>set_integer</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Int);
+
+julia&gt; is_integer(x)
+true
+
+julia&gt; unset_integer(x)
+
+julia&gt; is_integer(x)
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1296-L1320">source</a></section></article><h2 id="unset_silent"><a class="docs-heading-anchor" href="#unset_silent"><code>unset_silent</code></a><a id="unset_silent-1"></a><a class="docs-heading-anchor-permalink" href="#unset_silent" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.unset_silent" href="#JuMP.unset_silent"><code>JuMP.unset_silent</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unset_silent(model::GenericModel)</code></pre><p>Neutralize the effect of the <code>set_silent</code> function and let the solver attributes control the verbosity.</p><p>See also: <a href="#JuMP.set_silent"><code>set_silent</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L166-L173">source</a></section></article><h2 id="unset_time_limit_sec"><a class="docs-heading-anchor" href="#unset_time_limit_sec"><code>unset_time_limit_sec</code></a><a id="unset_time_limit_sec-1"></a><a class="docs-heading-anchor-permalink" href="#unset_time_limit_sec" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.unset_time_limit_sec" href="#JuMP.unset_time_limit_sec"><code>JuMP.unset_time_limit_sec</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unset_time_limit_sec(model::GenericModel)</code></pre><p>Unset the time limit of the solver.</p><p>See also: <a href="#JuMP.set_time_limit_sec"><code>set_time_limit_sec</code></a>, <a href="#JuMP.time_limit_sec"><code>time_limit_sec</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L196-L202">source</a></section></article><h2 id="upper_bound"><a class="docs-heading-anchor" href="#upper_bound"><code>upper_bound</code></a><a id="upper_bound-1"></a><a class="docs-heading-anchor-permalink" href="#upper_bound" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.upper_bound" href="#JuMP.upper_bound"><code>JuMP.upper_bound</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">upper_bound(v::GenericVariableRef)</code></pre><p>Return the upper bound of a variable.</p><p>Error if one does not exist.</p><p>See also <a href="#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, <a href="#JuMP.has_upper_bound"><code>has_upper_bound</code></a>, <a href="#JuMP.set_upper_bound"><code>set_upper_bound</code></a>, <a href="#JuMP.delete_upper_bound"><code>delete_upper_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &lt;= 1.0);
+
+julia&gt; upper_bound(x)
+1.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L981-L1001">source</a></section></article><h2 id="value"><a class="docs-heading-anchor" href="#value"><code>value</code></a><a id="value-1"></a><a class="docs-heading-anchor-permalink" href="#value" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.value" href="#JuMP.value"><code>JuMP.value</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">value(con_ref::ConstraintRef; result::Int = 1)</code></pre><p>Return the primal value of constraint <code>con_ref</code> associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>That is, if <code>con_ref</code> is the reference of a constraint <code>func</code>-in-<code>set</code>, it returns the value of <code>func</code> evaluated at the value of the variables (given by <a href="#JuMP.value"><code>value(::GenericVariableRef)</code></a>).</p><p>Use <a href="#JuMP.has_values"><code>has_values</code></a> to check if a result exists before asking for values.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p><p><strong>Note</strong></p><p>For scalar constraints, the constant is moved to the <code>set</code> so it is not taken into account in the primal value of the constraint. For instance, the constraint <code>@constraint(model, 2x + 3y + 1 == 5)</code> is transformed into <code>2x + 3y</code>-in-<code>MOI.EqualTo(4)</code> so the value returned by this function is the evaluation of <code>2x + 3y</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L978-L999">source</a></section><section><div><pre><code class="language-julia hljs">value(var_value::Function, con_ref::ConstraintRef)</code></pre><p>Evaluate the primal value of the constraint <code>con_ref</code> using <code>var_value(v)</code> as the value for each variable <code>v</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L1007-L1012">source</a></section><section><div><pre><code class="language-julia hljs">value(v::GenericVariableRef; result = 1)</code></pre><p>Return the value of variable <code>v</code> associated with result index <code>result</code> of the most-recent returned by the solver.</p><p>Use <a href="#JuMP.has_values"><code>has_values</code></a> to check if a result exists before asking for values.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1682-L1691">source</a></section><section><div><pre><code class="language-julia hljs">value(var_value::Function, v::GenericVariableRef)</code></pre><p>Evaluate the value of the variable <code>v</code> as <code>var_value(v)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1696-L1700">source</a></section><section><div><pre><code class="language-julia hljs">value(var_value::Function, ex::GenericAffExpr)</code></pre><p>Evaluate <code>ex</code> using <code>var_value(v)</code> as the value for each variable <code>v</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L346-L350">source</a></section><section><div><pre><code class="language-julia hljs">value(v::GenericAffExpr; result::Int = 1)</code></pre><p>Return the value of the <code>GenericAffExpr</code> <code>v</code> associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L609-L616">source</a></section><section><div><pre><code class="language-julia hljs">value(var_value::Function, ex::GenericQuadExpr)</code></pre><p>Evaluate <code>ex</code> using <code>var_value(v)</code> as the value for each variable <code>v</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L717-L721">source</a></section><section><div><pre><code class="language-julia hljs">value(v::GenericQuadExpr; result::Int = 1)</code></pre><p>Return the value of the <code>GenericQuadExpr</code> <code>v</code> associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>Replaces <code>getvalue</code> for most use cases.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L738-L747">source</a></section><section><div><pre><code class="language-julia hljs">value(p::NonlinearParameter)</code></pre><p>Return the current value stored in the nonlinear parameter <code>p</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @NLparameter(model, p == 10)
+p == 10.0
+
+julia&gt; value(p)
+10.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L228-L244">source</a></section><section><div><pre><code class="language-julia hljs">value(ex::NonlinearExpression; result::Int = 1)</code></pre><p>Return the value of the <code>NonlinearExpression</code> <code>ex</code> associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>Replaces <code>getvalue</code> for most use cases.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L338-L347">source</a></section><section><div><pre><code class="language-julia hljs">value(var_value::Function, ex::NonlinearExpression)</code></pre><p>Evaluate <code>ex</code> using <code>var_value(v)</code> as the value for each variable <code>v</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L378-L382">source</a></section><section><div><pre><code class="language-julia hljs">value(c::NonlinearConstraintRef; result::Int = 1)</code></pre><p>Return the value of the <code>NonlinearConstraintRef</code> <code>c</code> associated with result index <code>result</code> of the most-recent solution returned by the solver.</p><p>See also: <a href="#JuMP.result_count"><code>result_count</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L529-L536">source</a></section><section><div><pre><code class="language-julia hljs">value(var_value::Function, c::NonlinearConstraintRef)</code></pre><p>Evaluate <code>c</code> using <code>var_value(v)</code> as the value for each variable <code>v</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L543-L547">source</a></section></article><h2 id="value_type"><a class="docs-heading-anchor" href="#value_type"><code>value_type</code></a><a id="value_type-1"></a><a class="docs-heading-anchor-permalink" href="#value_type" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.value_type" href="#JuMP.value_type"><code>JuMP.value_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">value_type(::Type{&lt;:Union{AbstractModel,AbstractVariableRef}})</code></pre><p>Return the return type of <a href="#JuMP.value"><code>value</code></a> for variables of that model. It defaults to <code>Float64</code> if it is not implemented.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L77-L82">source</a></section></article><h2 id="variable_by_name"><a class="docs-heading-anchor" href="#variable_by_name"><code>variable_by_name</code></a><a id="variable_by_name-1"></a><a class="docs-heading-anchor-permalink" href="#variable_by_name" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.variable_by_name" href="#JuMP.variable_by_name"><code>JuMP.variable_by_name</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">variable_by_name(
+    model::AbstractModel,
+    name::String,
+)::Union{AbstractVariableRef,Nothing}</code></pre><p>Returns the reference of the variable with name attribute <code>name</code> or <code>Nothing</code> if no variable has this name attribute. Throws an error if several variables have <code>name</code> as their name attribute.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; variable_by_name(model, &quot;x&quot;)
+x
+
+julia&gt; @variable(model, base_name=&quot;x&quot;)
+x
+
+julia&gt; variable_by_name(model, &quot;x&quot;)
+ERROR: Multiple variables have the name x.
+Stacktrace:
+ [1] error(::String) at ./error.jl:33
+ [2] get(::MOIU.Model{Float64}, ::Type{MathOptInterface.VariableIndex}, ::String) at /home/blegat/.julia/dev/MathOptInterface/src/Utilities/model.jl:222
+ [3] get at /home/blegat/.julia/dev/MathOptInterface/src/Utilities/universalfallback.jl:201 [inlined]
+ [4] get(::MathOptInterface.Utilities.CachingOptimizer{MathOptInterface.AbstractOptimizer,MathOptInterface.Utilities.UniversalFallback{MOIU.Model{Float64}}}, ::Type{MathOptInterface.VariableIndex}, ::String) at /home/blegat/.julia/dev/MathOptInterface/src/Utilities/cachingoptimizer.jl:490
+ [5] variable_by_name(::GenericModel, ::String) at /home/blegat/.julia/dev/JuMP/src/variables.jl:268
+ [6] top-level scope at none:0
+
+julia&gt; var = @variable(model, base_name=&quot;y&quot;)
+y
+
+julia&gt; variable_by_name(model, &quot;y&quot;)
+y
+
+julia&gt; set_name(var, &quot;z&quot;)
+
+julia&gt; variable_by_name(model, &quot;y&quot;)
+
+julia&gt; variable_by_name(model, &quot;z&quot;)
+z
+
+julia&gt; @variable(model, u[1:2])
+2-element Vector{VariableRef}:
+ u[1]
+ u[2]
+
+julia&gt; variable_by_name(model, &quot;u[2]&quot;)
+u[2]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L544-L599">source</a></section></article><h2 id="variable_ref_type"><a class="docs-heading-anchor" href="#variable_ref_type"><code>variable_ref_type</code></a><a id="variable_ref_type-1"></a><a class="docs-heading-anchor-permalink" href="#variable_ref_type" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.variable_ref_type" href="#JuMP.variable_ref_type"><code>JuMP.variable_ref_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">variable_ref_type(::Union{F,Type{F}}) where {F}</code></pre><p>A helper function used internally by JuMP and some JuMP extensions. Returns the variable type associated with the model or expression type <code>F</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L219-L224">source</a></section></article><h2 id="vectorize"><a class="docs-heading-anchor" href="#vectorize"><code>vectorize</code></a><a id="vectorize-1"></a><a class="docs-heading-anchor-permalink" href="#vectorize" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.vectorize" href="#JuMP.vectorize"><code>JuMP.vectorize</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">vectorize(matrix::AbstractMatrix, ::Shape)</code></pre><p>Convert the <code>matrix</code> into a vector according to <code>Shape</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L11-L15">source</a></section></article><h2 id="write_to_file"><a class="docs-heading-anchor" href="#write_to_file"><code>write_to_file</code></a><a id="write_to_file-1"></a><a class="docs-heading-anchor-permalink" href="#write_to_file" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.write_to_file" href="#JuMP.write_to_file"><code>JuMP.write_to_file</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">write_to_file(
+    model::GenericModel,
+    filename::String;
+    format::MOI.FileFormats.FileFormat = MOI.FileFormats.FORMAT_AUTOMATIC,
+    kwargs...,
+)</code></pre><p>Write the JuMP model <code>model</code> to <code>filename</code> in the format <code>format</code>.</p><p>If the filename ends in <code>.gz</code>, it will be compressed using Gzip. If the filename ends in <code>.bz2</code>, it will be compressed using BZip2.</p><p>Other <code>kwargs</code> are passed to the <code>Model</code> constructor of the chosen format.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/file_formats.jl#L31-L45">source</a></section></article><h2 id="AbstractConstraint"><a class="docs-heading-anchor" href="#AbstractConstraint"><code>AbstractConstraint</code></a><a id="AbstractConstraint-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractConstraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractConstraint" href="#JuMP.AbstractConstraint"><code>JuMP.AbstractConstraint</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractConstraint</code></pre><p>An abstract base type for all constraint types. <code>AbstractConstraint</code>s store the function and set directly, unlike <a href="#JuMP.ConstraintRef"><code>ConstraintRef</code></a>s that are merely references to constraints stored in a model. <code>AbstractConstraint</code>s do not need to be attached to a model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L448-L455">source</a></section></article><h2 id="AbstractJuMPScalar"><a class="docs-heading-anchor" href="#AbstractJuMPScalar"><code>AbstractJuMPScalar</code></a><a id="AbstractJuMPScalar-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractJuMPScalar" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractJuMPScalar" href="#JuMP.AbstractJuMPScalar"><code>JuMP.AbstractJuMPScalar</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractJuMPScalar &lt;: MutableArithmetics.AbstractMutable</code></pre><p>Abstract base type for all scalar types</p><p>The subtyping of <code>AbstractMutable</code> will allow calls of some <code>Base</code> functions to be redirected to a method in MA that handles type promotion more carefully (e.g. the promotion in sparse matrix products in SparseArrays usually does not work for JuMP types) and exploits the mutability of <code>AffExpr</code> and <code>QuadExpr</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L991-L1000">source</a></section></article><h2 id="AbstractModel"><a class="docs-heading-anchor" href="#AbstractModel"><code>AbstractModel</code></a><a id="AbstractModel-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractModel" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractModel" href="#JuMP.AbstractModel"><code>JuMP.AbstractModel</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractModel</code></pre><p>An abstract type that should be subtyped for users creating JuMP extensions.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L70-L74">source</a></section></article><h2 id="AbstractScalarSet"><a class="docs-heading-anchor" href="#AbstractScalarSet"><code>AbstractScalarSet</code></a><a id="AbstractScalarSet-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractScalarSet" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractScalarSet" href="#JuMP.AbstractScalarSet"><code>JuMP.AbstractScalarSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractScalarSet</code></pre><p>An abstract type for defining new scalar sets in JuMP.</p><p>Implement <code>moi_set(::AbstractScalarSet)</code> to convert the type into an MOI set.</p><p>See also: <a href="#JuMP.moi_set"><code>moi_set</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L249-L257">source</a></section></article><h2 id="AbstractShape"><a class="docs-heading-anchor" href="#AbstractShape"><code>AbstractShape</code></a><a id="AbstractShape-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractShape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractShape" href="#JuMP.AbstractShape"><code>JuMP.AbstractShape</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractShape</code></pre><p>Abstract vectorizable shape. Given a flat vector form of an object of shape <code>shape</code>, the original object can be obtained by <a href="#JuMP.reshape_vector"><code>reshape_vector</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/shapes.jl#L11-L16">source</a></section></article><h2 id="AbstractVariable"><a class="docs-heading-anchor" href="#AbstractVariable"><code>AbstractVariable</code></a><a id="AbstractVariable-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractVariable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractVariable" href="#JuMP.AbstractVariable"><code>JuMP.AbstractVariable</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractVariable</code></pre><p>Variable returned by <a href="#JuMP.build_variable"><code>build_variable</code></a>. It represents a variable that has not been added yet to any model. It can be added to a given <code>model</code> with <a href="#JuMP.add_variable"><code>add_variable</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L15-L21">source</a></section></article><h2 id="AbstractVariableRef"><a class="docs-heading-anchor" href="#AbstractVariableRef"><code>AbstractVariableRef</code></a><a id="AbstractVariableRef-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractVariableRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractVariableRef" href="#JuMP.AbstractVariableRef"><code>JuMP.AbstractVariableRef</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractVariableRef</code></pre><p>Variable returned by <a href="#JuMP.add_variable"><code>add_variable</code></a>. Affine (resp. quadratic) operations with variables of type <code>V&lt;:AbstractVariableRef</code> and coefficients of type <code>T</code>     create a <code>GenericAffExpr{T,V}</code> (resp. <code>GenericQuadExpr{T,V}</code>).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L210-L216">source</a></section></article><h2 id="AbstractVectorSet"><a class="docs-heading-anchor" href="#AbstractVectorSet"><code>AbstractVectorSet</code></a><a id="AbstractVectorSet-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractVectorSet" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AbstractVectorSet" href="#JuMP.AbstractVectorSet"><code>JuMP.AbstractVectorSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractVectorSet</code></pre><p>An abstract type for defining new sets in JuMP.</p><p>Implement <code>moi_set(::AbstractVectorSet, dim::Int)</code> to convert the type into an MOI set.</p><p>See also: <a href="#JuMP.moi_set"><code>moi_set</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L11-L20">source</a></section></article><h2 id="AffExpr"><a class="docs-heading-anchor" href="#AffExpr"><code>AffExpr</code></a><a id="AffExpr-1"></a><a class="docs-heading-anchor-permalink" href="#AffExpr" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AffExpr" href="#JuMP.AffExpr"><code>JuMP.AffExpr</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AffExpr</code></pre><p>Alias for <code>GenericAffExpr{Float64,VariableRef}</code>, the specific <a href="#JuMP.GenericAffExpr"><code>GenericAffExpr</code></a> used by JuMP.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L585-L590">source</a></section></article><h2 id="BinaryRef"><a class="docs-heading-anchor" href="#BinaryRef"><code>BinaryRef</code></a><a id="BinaryRef-1"></a><a class="docs-heading-anchor-permalink" href="#BinaryRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.BinaryRef" href="#JuMP.BinaryRef"><code>JuMP.BinaryRef</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">BinaryRef(v::GenericVariableRef)</code></pre><p>Return a constraint reference to the constraint constraining <code>v</code> to be binary. Errors if one does not exist.</p><p>See also <a href="#JuMP.is_binary"><code>is_binary</code></a>, <a href="#JuMP.set_binary"><code>set_binary</code></a>, <a href="#JuMP.unset_binary"><code>unset_binary</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Bin);
+
+julia&gt; BinaryRef(x)
+x binary</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1457-L1475">source</a></section></article><h2 id="BridgeableConstraint"><a class="docs-heading-anchor" href="#BridgeableConstraint"><code>BridgeableConstraint</code></a><a id="BridgeableConstraint-1"></a><a class="docs-heading-anchor-permalink" href="#BridgeableConstraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.BridgeableConstraint" href="#JuMP.BridgeableConstraint"><code>JuMP.BridgeableConstraint</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">BridgeableConstraint(
+    constraint::C,
+    bridge_type::B;
+    coefficient_type::Type{T} = Float64,
+) where {C&lt;:AbstractConstraint,B&lt;:Type{&lt;:MOI.Bridges.AbstractBridge},T}</code></pre><p>An <a href="#JuMP.AbstractConstraint"><code>AbstractConstraint</code></a> representinng that <code>constraint</code> that can be bridged by the bridge of type <code>bridge_type{coefficient_type}</code>.</p><p>Adding a <code>BridgeableConstraint</code> to a model is equivalent to:</p><pre><code class="language-julia hljs">add_bridge(model, bridge_type; coefficient_type = coefficient_type)
+add_constraint(model, constraint)</code></pre><p><strong>Example</strong></p><p>Given a new scalar set type <code>CustomSet</code> with a bridge <code>CustomBridge</code> that can bridge <code>F</code>-in-<code>CustomSet</code> constraints, when the user does:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x)
+@constraint(model, x + 1 in CustomSet())
+optimize!(model)</code></pre><p>with an optimizer that does not support <code>F</code>-in-<code>CustomSet</code> constraints, the constraint will not be bridged unless they first call <code>add_bridge(model, CustomBridge)</code>.</p><p>In order to automatically add the <code>CustomBridge</code> to any model to which an <code>F</code>-in-<code>CustomSet</code> is added, add the following method:</p><pre><code class="language-julia hljs">function JuMP.build_constraint(
+    _error::Function,
+    func::AbstractJuMPScalar,
+    set::CustomSet,
+)
+    constraint = ScalarConstraint(func, set)
+    return BridgeableConstraint(constraint, CustomBridge)
+end</code></pre><p><strong>Note</strong></p><p>JuMP extensions should extend <code>JuMP.build_constraint</code> only if they also defined <code>CustomSet</code>, for three reasons:</p><ol><li>It is problematic if multiple extensions overload the same JuMP method.</li><li>A missing method will not inform the users that they forgot to load the extension module defining the <code>build_constraint</code> method.</li><li>Defining a method where neither the function nor any of the argument types are defined in the package is called <a href="https://docs.julialang.org/en/v1/manual/style-guide/index.html#Avoid-type-piracy-1"><em>type piracy</em></a> and is discouraged in the Julia style guide.</li></ol></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L458-L512">source</a></section></article><h2 id="ComplexPlane"><a class="docs-heading-anchor" href="#ComplexPlane"><code>ComplexPlane</code></a><a id="ComplexPlane-1"></a><a class="docs-heading-anchor-permalink" href="#ComplexPlane" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ComplexPlane" href="#JuMP.ComplexPlane"><code>JuMP.ComplexPlane</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ComplexPlane</code></pre><p>Complex plane object that can be used to create a complex variable in the <a href="#JuMP.@variable"><code>@variable</code></a> macro.</p><p><strong>Example</strong></p><p>Consider the following example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x in ComplexPlane())
+real(x) + imag(x) im
+
+julia&gt; all_variables(model)
+2-element Vector{VariableRef}:
+ real(x)
+ imag(x)</code></pre><p>We see in the output of the last command that two real variables were created. The Julia variable <code>x</code> binds to an affine expression in terms of these two variables that parametrize the complex plane.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1965-L1990">source</a></section></article><h2 id="ComplexVariable"><a class="docs-heading-anchor" href="#ComplexVariable"><code>ComplexVariable</code></a><a id="ComplexVariable-1"></a><a class="docs-heading-anchor-permalink" href="#ComplexVariable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ComplexVariable" href="#JuMP.ComplexVariable"><code>JuMP.ComplexVariable</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ComplexVariable{S,T,U,V} &lt;: AbstractVariable</code></pre><p>A struct used when adding complex variables.</p><p>See also: <a href="#JuMP.ComplexPlane"><code>ComplexPlane</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1993-L1999">source</a></section></article><h2 id="ConstraintNotOwned"><a class="docs-heading-anchor" href="#ConstraintNotOwned"><code>ConstraintNotOwned</code></a><a id="ConstraintNotOwned-1"></a><a class="docs-heading-anchor-permalink" href="#ConstraintNotOwned" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ConstraintNotOwned" href="#JuMP.ConstraintNotOwned"><code>JuMP.ConstraintNotOwned</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ConstraintNotOwned{C &lt;: ConstraintRef} &lt;: Exception
+    constraint_ref::C
+end</code></pre><p>The constraint <code>constraint_ref</code> was used in a model different to <code>owner_model(constraint_ref)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L39-L46">source</a></section></article><h2 id="ConstraintRef"><a class="docs-heading-anchor" href="#ConstraintRef"><code>ConstraintRef</code></a><a id="ConstraintRef-1"></a><a class="docs-heading-anchor-permalink" href="#ConstraintRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ConstraintRef" href="#JuMP.ConstraintRef"><code>JuMP.ConstraintRef</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintRef</code></pre><p>Holds a reference to the model and the corresponding MOI.ConstraintIndex.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L11-L15">source</a></section></article><h2 id="FixRef"><a class="docs-heading-anchor" href="#FixRef"><code>FixRef</code></a><a id="FixRef-1"></a><a class="docs-heading-anchor-permalink" href="#FixRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.FixRef" href="#JuMP.FixRef"><code>JuMP.FixRef</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">FixRef(v::GenericVariableRef)</code></pre><p>Return a constraint reference to the constraint fixing the value of <code>v</code>.</p><p>Errors if one does not exist.</p><p>See also <a href="#JuMP.is_fixed"><code>is_fixed</code></a>, <a href="#JuMP.fix_value"><code>fix_value</code></a>, <a href="#JuMP.fix"><code>fix</code></a>, <a href="#JuMP.unfix"><code>unfix</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x == 1);
+
+julia&gt; FixRef(x)
+x = 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1188-L1208">source</a></section></article><h2 id="GenericAffExpr"><a class="docs-heading-anchor" href="#GenericAffExpr"><code>GenericAffExpr</code></a><a id="GenericAffExpr-1"></a><a class="docs-heading-anchor-permalink" href="#GenericAffExpr" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.GenericAffExpr" href="#JuMP.GenericAffExpr"><code>JuMP.GenericAffExpr</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct GenericAffExpr{CoefType,VarType} &lt;: AbstractJuMPScalar
+    constant::CoefType
+    terms::OrderedDict{VarType,CoefType}
+end</code></pre><p>An expression type representing an affine expression of the form: <span>$\sum a_i x_i + c$</span>.</p><p><strong>Fields</strong></p><ul><li><code>.constant</code>: the constant <code>c</code> in the expression.</li><li><code>.terms</code>: an <code>OrderedDict</code>, with keys of <code>VarType</code> and values of <code>CoefType</code> describing the sparse vector <code>a</code>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L105-L119">source</a></section></article><h2 id="GenericModel"><a class="docs-heading-anchor" href="#GenericModel"><code>GenericModel</code></a><a id="GenericModel-1"></a><a class="docs-heading-anchor-permalink" href="#GenericModel" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.GenericModel" href="#JuMP.GenericModel"><code>JuMP.GenericModel</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GenericModel{T}(
+    [optimizer_factory;]
+    add_bridges::Bool = true,
+) where {T&lt;:Real}</code></pre><p>Create a new instance of a JuMP model.</p><p>If <code>optimizer_factory</code> is provided, the model is initialized with the optimizer returned by <code>MOI.instantiate(optimizer_factory)</code>.</p><p>If <code>optimizer_factory</code> is not provided, use <a href="#JuMP.set_optimizer"><code>set_optimizer</code></a> to set the optimizer before calling <a href="#JuMP.optimize!"><code>optimize!</code></a>.</p><p>If <code>add_bridges</code>, JuMP adds a <a href="../../moi/submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a> to automatically reformulate the problem into a form supported by the optimizer.</p><p><strong>Value type <code>T</code></strong></p><p>Passing a type other than <code>Float64</code> as the value type <code>T</code> is an advanced operation. The value type must match that expected by the chosen optimizer. Consult the optimizers documentation for details.</p><p>If not documented, assume that the optimizer supports only <code>Float64</code>.</p><p>Choosing an unsupported value type will throw an <a href="../../moi/reference/errors/#MathOptInterface.UnsupportedConstraint"><code>MOI.UnsupportedConstraint</code></a> or an <a href="../../moi/reference/errors/#MathOptInterface.UnsupportedAttribute"><code>MOI.UnsupportedAttribute</code></a> error, the timing of which (during the model construction or during a call to <a href="#JuMP.optimize!"><code>optimize!</code></a>) depends on how the solver is interfaced to JuMP.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = GenericModel{BigFloat}();
+
+julia&gt; typeof(model)
+GenericModel{BigFloat}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L136-L174">source</a></section></article><h2 id="GenericNonlinearExpr"><a class="docs-heading-anchor" href="#GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a><a id="GenericNonlinearExpr-1"></a><a class="docs-heading-anchor-permalink" href="#GenericNonlinearExpr" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.GenericNonlinearExpr" href="#JuMP.GenericNonlinearExpr"><code>JuMP.GenericNonlinearExpr</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GenericNonlinearExpr{V}(head::Symbol, args::Vector{Any})
+GenericNonlinearExpr{V}(head::Symbol, args::Any...)</code></pre><p>The scalar-valued nonlinear function <code>head(args...)</code>, represented as a symbolic expression tree, with the call operator <code>head</code> and ordered arguments in <code>args</code>.</p><p><code>V</code> is the type of <a href="#JuMP.AbstractVariableRef"><code>AbstractVariableRef</code></a> present in the expression, and is used to help dispatch JuMP extensions.</p><p><strong><code>head</code></strong></p><p>The <code>head::Symbol</code> must be an operator supported by the model.</p><p>The default list of supported univariate operators is given by:</p><ul><li><a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS"><code>MOI.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS</code></a></li></ul><p>and the default list of supported multivariate operators is given by:</p><ul><li><a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS"><code>MOI.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS</code></a></li></ul><p>Additional operators can be add using <a href="#JuMP.@operator"><code>@operator</code></a>.</p><p>See the full list of operators supported by a <a href="../../moi/reference/models/#MathOptInterface.ModelLike"><code>MOI.ModelLike</code></a> by querying the <a href="../../moi/reference/models/#MathOptInterface.ListOfSupportedNonlinearOperators"><code>MOI.ListOfSupportedNonlinearOperators</code></a> attribute.</p><p><strong><code>args</code></strong></p><p>The vector <code>args</code> contains the arguments to the nonlinear function. If the operator is univariate, it must contain one element. Otherwise, it may contain multiple elements.</p><p>Given a subtype of <a href="#JuMP.AbstractVariableRef"><code>AbstractVariableRef</code></a>, <code>V</code>, for <code>GenericNonlinearExpr{V}</code>, each element must be one of the following:</p><ul><li>A constant value of type <code>&lt;:Real</code></li><li>A <code>V</code></li><li>A <a href="#JuMP.GenericAffExpr"><code>GenericAffExpr{T,V}</code></a></li><li>A <a href="#JuMP.GenericQuadExpr"><code>GenericQuadExpr{T,V}</code></a></li><li>A <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr{V}</code></a></li></ul><p>where <code>T&lt;:Real</code> and <code>T == value_type(V)</code>.</p><p><strong>Unsupported operators</strong></p><p>If the optimizer does not support <code>head</code>, an <a href="../../moi/reference/errors/#MathOptInterface.UnsupportedNonlinearOperator"><code>MOI.UnsupportedNonlinearOperator</code></a> error will be thrown.</p><p>There is no guarantee about when this error will be thrown; it may be thrown when the function is first added to the model, or it may be thrown when <a href="#JuMP.optimize!"><code>optimize!</code></a> is called.</p><p><strong>Example</strong></p><p>To represent the function <span>$f(x) = sin(x)^2$</span>, do:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; f = sin(x)^2
+sin(x) ^ 2.0
+
+julia&gt; f = GenericNonlinearExpr{VariableRef}(
+           :^,
+           GenericNonlinearExpr{VariableRef}(:sin, x),
+           2.0,
+       )
+sin(x) ^ 2.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp_expr.jl#L6-L79">source</a></section></article><h2 id="GenericQuadExpr"><a class="docs-heading-anchor" href="#GenericQuadExpr"><code>GenericQuadExpr</code></a><a id="GenericQuadExpr-1"></a><a class="docs-heading-anchor-permalink" href="#GenericQuadExpr" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.GenericQuadExpr" href="#JuMP.GenericQuadExpr"><code>JuMP.GenericQuadExpr</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct GenericQuadExpr{CoefType,VarType} &lt;: AbstractJuMPScalar
+    aff::GenericAffExpr{CoefType,VarType}
+    terms::OrderedDict{UnorderedPair{VarType}, CoefType}
+end</code></pre><p>An expression type representing an quadratic expression of the form: <span>$\sum q_{i,j} x_i x_j + \sum a_i x_i + c$</span>.</p><p><strong>Fields</strong></p><ul><li><code>.aff</code>: an <a href="#JuMP.GenericAffExpr"><code>GenericAffExpr</code></a> representing the affine portion of the expression.</li><li><code>.terms</code>: an <code>OrderedDict</code>, with keys of <code>UnorderedPair{VarType}</code> and values of <code>CoefType</code>, describing the sparse list of terms <code>q</code>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L33-L48">source</a></section></article><h2 id="GenericReferenceMap"><a class="docs-heading-anchor" href="#GenericReferenceMap"><code>GenericReferenceMap</code></a><a id="GenericReferenceMap-1"></a><a class="docs-heading-anchor-permalink" href="#GenericReferenceMap" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.GenericReferenceMap" href="#JuMP.GenericReferenceMap"><code>JuMP.GenericReferenceMap</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GenericReferenceMap{T}</code></pre><p>Mapping between variable and constraint reference of a model and its copy. The reference of the copied model can be obtained by indexing the map with the reference of the corresponding reference of the original model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/copy.jl#L31-L37">source</a></section></article><h2 id="GenericVariableRef"><a class="docs-heading-anchor" href="#GenericVariableRef"><code>GenericVariableRef</code></a><a id="GenericVariableRef-1"></a><a class="docs-heading-anchor-permalink" href="#GenericVariableRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.GenericVariableRef" href="#JuMP.GenericVariableRef"><code>JuMP.GenericVariableRef</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GenericVariableRef{T} &lt;: AbstractVariableRef</code></pre><p>Holds a reference to the model and the corresponding MOI.VariableIndex.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L237-L241">source</a></section></article><h2 id="HermitianMatrixShape"><a class="docs-heading-anchor" href="#HermitianMatrixShape"><code>HermitianMatrixShape</code></a><a id="HermitianMatrixShape-1"></a><a class="docs-heading-anchor-permalink" href="#HermitianMatrixShape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.HermitianMatrixShape" href="#JuMP.HermitianMatrixShape"><code>JuMP.HermitianMatrixShape</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HermitianMatrixShape</code></pre><p>Shape object for a Hermitian square matrix of <code>side_dimension</code> rows and columns. The vectorized form corresponds to <a href="../../moi/reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>MOI.HermitianPositiveSemidefiniteConeTriangle</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L554-L560">source</a></section></article><h2 id="HermitianMatrixSpace"><a class="docs-heading-anchor" href="#HermitianMatrixSpace"><code>HermitianMatrixSpace</code></a><a id="HermitianMatrixSpace-1"></a><a class="docs-heading-anchor-permalink" href="#HermitianMatrixSpace" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.HermitianMatrixSpace" href="#JuMP.HermitianMatrixSpace"><code>JuMP.HermitianMatrixSpace</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HermitianMatrixSpace()</code></pre><p>Use in the <a href="#JuMP.@variable"><code>@variable</code></a> macro to constrain a matrix of variables to be hermitian.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2] in HermitianMatrixSpace())
+2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(Q[1,1])                    real(Q[1,2]) + imag(Q[1,2]) im
+ real(Q[1,2]) - imag(Q[1,2]) im  real(Q[2,2])</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L56-L72">source</a></section></article><h2 id="HermitianPSDCone"><a class="docs-heading-anchor" href="#HermitianPSDCone"><code>HermitianPSDCone</code></a><a id="HermitianPSDCone-1"></a><a class="docs-heading-anchor-permalink" href="#HermitianPSDCone" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.HermitianPSDCone" href="#JuMP.HermitianPSDCone"><code>JuMP.HermitianPSDCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HermitianPSDCone</code></pre><p>Hermitian positive semidefinite cone object that can be used to create a Hermitian positive semidefinite square matrix in the <a href="#JuMP.@variable"><code>@variable</code></a> and <a href="#JuMP.@constraint"><code>@constraint</code></a> macros.</p><p><strong>Example</strong></p><p>Consider the following example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, H[1:3, 1:3] in HermitianPSDCone())
+3×3 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(H[1,1])                    …  real(H[1,3]) + imag(H[1,3]) im
+ real(H[1,2]) - imag(H[1,2]) im     real(H[2,3]) + imag(H[2,3]) im
+ real(H[1,3]) - imag(H[1,3]) im     real(H[3,3])
+
+julia&gt; all_variables(model)
+9-element Vector{VariableRef}:
+ real(H[1,1])
+ real(H[1,2])
+ real(H[2,2])
+ real(H[1,3])
+ real(H[2,3])
+ real(H[3,3])
+ imag(H[1,2])
+ imag(H[1,3])
+ imag(H[2,3])
+
+julia&gt; all_constraints(model, Vector{VariableRef}, MOI.HermitianPositiveSemidefiniteConeTriangle)
+1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.HermitianPositiveSemidefiniteConeTriangle}}}:
+ [real(H[1,1]), real(H[1,2]), real(H[2,2]), real(H[1,3]), real(H[2,3]), real(H[3,3]), imag(H[1,2]), imag(H[1,3]), imag(H[2,3])] ∈ MathOptInterface.HermitianPositiveSemidefiniteConeTriangle(3)</code></pre><p>We see in the output of the last commands that 9 real variables were created. The matrix <code>H</code> contrains affine expressions in terms of these 9 variables that parametrize a Hermitian matrix.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L513-L551">source</a></section></article><h2 id="IntegerRef"><a class="docs-heading-anchor" href="#IntegerRef"><code>IntegerRef</code></a><a id="IntegerRef-1"></a><a class="docs-heading-anchor-permalink" href="#IntegerRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.IntegerRef" href="#JuMP.IntegerRef"><code>JuMP.IntegerRef</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">IntegerRef(v::GenericVariableRef)</code></pre><p>Return a constraint reference to the constraint constraining <code>v</code> to be integer.</p><p>Errors if one does not exist.</p><p>See also <a href="#JuMP.is_integer"><code>is_integer</code></a>, <a href="#JuMP.set_integer"><code>set_integer</code></a>, <a href="#JuMP.unset_integer"><code>unset_integer</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Int);
+
+julia&gt; IntegerRef(x)
+x integer</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1326-L1345">source</a></section></article><h2 id="LinearTermIterator"><a class="docs-heading-anchor" href="#LinearTermIterator"><code>LinearTermIterator</code></a><a id="LinearTermIterator-1"></a><a class="docs-heading-anchor-permalink" href="#LinearTermIterator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.LinearTermIterator" href="#JuMP.LinearTermIterator"><code>JuMP.LinearTermIterator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LinearTermIterator{GAE&lt;:GenericAffExpr}</code></pre><p>A struct that implements the <code>iterate</code> protocol in order to iterate over tuples of <code>(coefficient, variable)</code> in the <code>GenericAffExpr</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/aff_expr.jl#L368-L373">source</a></section></article><h2 id="LowerBoundRef"><a class="docs-heading-anchor" href="#LowerBoundRef"><code>LowerBoundRef</code></a><a id="LowerBoundRef-1"></a><a class="docs-heading-anchor-permalink" href="#LowerBoundRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.LowerBoundRef" href="#JuMP.LowerBoundRef"><code>JuMP.LowerBoundRef</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">LowerBoundRef(v::GenericVariableRef)</code></pre><p>Return a constraint reference to the lower bound constraint of <code>v</code>.</p><p>Errors if one does not exist.</p><p>See also <a href="#JuMP.has_lower_bound"><code>has_lower_bound</code></a>, <a href="#JuMP.lower_bound"><code>lower_bound</code></a>, <a href="#JuMP.set_lower_bound"><code>set_lower_bound</code></a>, <a href="#JuMP.delete_lower_bound"><code>delete_lower_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 1.0);
+
+julia&gt; LowerBoundRef(x)
+x ≥ 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L751-L771">source</a></section></article><h2 id="Model"><a class="docs-heading-anchor" href="#Model"><code>Model</code></a><a id="Model-1"></a><a class="docs-heading-anchor-permalink" href="#Model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Model" href="#JuMP.Model"><code>JuMP.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model([optimizer_factory;] add_bridges::Bool = true)</code></pre><p>Create a new instance of a JuMP model.</p><p>If <code>optimizer_factory</code> is provided, the model is initialized with thhe optimizer returned by <code>MOI.instantiate(optimizer_factory)</code>.</p><p>If <code>optimizer_factory</code> is not provided, use <a href="#JuMP.set_optimizer"><code>set_optimizer</code></a> to set the optimizer before calling <a href="#JuMP.optimize!"><code>optimize!</code></a>.</p><p>If <code>add_bridges</code>, JuMP adds a <a href="../../moi/submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a> to automatically reformulate the problem into a form supported by the optimizer.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; model = Model(Ipopt.Optimizer);
+
+julia&gt; solver_name(model)
+&quot;Ipopt&quot;
+
+julia&gt; import HiGHS
+
+julia&gt; import MultiObjectiveAlgorithms as MOA
+
+julia&gt; model = Model(() -&gt; MOA.Optimizer(HiGHS.Optimizer); add_bridges = false);</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L280-L310">source</a></section></article><h2 id="ModelMode"><a class="docs-heading-anchor" href="#ModelMode"><code>ModelMode</code></a><a id="ModelMode-1"></a><a class="docs-heading-anchor-permalink" href="#ModelMode" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ModelMode" href="#JuMP.ModelMode"><code>JuMP.ModelMode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ModelMode</code></pre><p>An enum to describe the state of the CachingOptimizer inside a JuMP model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L50-L54">source</a></section></article><h2 id="NLPEvaluator"><a class="docs-heading-anchor" href="#NLPEvaluator"><code>NLPEvaluator</code></a><a id="NLPEvaluator-1"></a><a class="docs-heading-anchor-permalink" href="#NLPEvaluator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NLPEvaluator" href="#JuMP.NLPEvaluator"><code>JuMP.NLPEvaluator</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">NLPEvaluator(
+    model::Model,
+    _differentiation_backend::MOI.Nonlinear.AbstractAutomaticDifferentiation =
+        MOI.Nonlinear.SparseReverseMode(),
+)</code></pre><p>Return an <a href="../../moi/reference/nonlinear/#MathOptInterface.AbstractNLPEvaluator"><code>MOI.AbstractNLPEvaluator</code></a> constructed from <code>model</code></p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Before using, you must initialize the evaluator using <a href="../../moi/reference/nonlinear/#MathOptInterface.initialize"><code>MOI.initialize</code></a>.</p></div></div><p><strong>Experimental</strong></p><p>These features may change or be removed in any future version of JuMP.</p><p>Pass <code>_differentiation_backend</code> to specify the differentiation backend used to compute derivatives.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L881-L900">source</a></section></article><h2 id="NoOptimizer"><a class="docs-heading-anchor" href="#NoOptimizer"><code>NoOptimizer</code></a><a id="NoOptimizer-1"></a><a class="docs-heading-anchor-permalink" href="#NoOptimizer" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NoOptimizer" href="#JuMP.NoOptimizer"><code>JuMP.NoOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct NoOptimizer &lt;: Exception end</code></pre><p>No optimizer is set. The optimizer can be provided to the <a href="#JuMP.Model"><code>Model</code></a> constructor or by calling <a href="#JuMP.set_optimizer"><code>set_optimizer</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L640-L645">source</a></section></article><h2 id="NonlinearConstraintIndex"><a class="docs-heading-anchor" href="#NonlinearConstraintIndex"><code>NonlinearConstraintIndex</code></a><a id="NonlinearConstraintIndex-1"></a><a class="docs-heading-anchor-permalink" href="#NonlinearConstraintIndex" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NonlinearConstraintIndex" href="#JuMP.NonlinearConstraintIndex"><code>JuMP.NonlinearConstraintIndex</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintIndex</code></pre><p>An index to a nonlinear constraint that is returned by <a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.add_constraint"><code>add_constraint</code></a>.</p><p>Given <code>data::Model</code> and <code>c::ConstraintIndex</code>, use <code>data[c]</code> to retrieve the corresponding <a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Constraint"><code>Constraint</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L133-L140">source</a></section></article><h2 id="NonlinearConstraintRef"><a class="docs-heading-anchor" href="#NonlinearConstraintRef"><code>NonlinearConstraintRef</code></a><a id="NonlinearConstraintRef-1"></a><a class="docs-heading-anchor-permalink" href="#NonlinearConstraintRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NonlinearConstraintRef" href="#JuMP.NonlinearConstraintRef"><code>JuMP.NonlinearConstraintRef</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonlinearConstraintRef</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L402-L404">source</a></section></article><h2 id="NonlinearExpr"><a class="docs-heading-anchor" href="#NonlinearExpr"><code>NonlinearExpr</code></a><a id="NonlinearExpr-1"></a><a class="docs-heading-anchor-permalink" href="#NonlinearExpr" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NonlinearExpr" href="#JuMP.NonlinearExpr"><code>JuMP.NonlinearExpr</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonlinearExpr</code></pre><p>Alias for <code>GenericNonlinearExpr{VariableRef}</code>, the specific <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a> used by JuMP.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp_expr.jl#L128-L133">source</a></section></article><h2 id="NonlinearExpression"><a class="docs-heading-anchor" href="#NonlinearExpression"><code>NonlinearExpression</code></a><a id="NonlinearExpression-1"></a><a class="docs-heading-anchor-permalink" href="#NonlinearExpression" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NonlinearExpression" href="#JuMP.NonlinearExpression"><code>JuMP.NonlinearExpression</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonlinearExpression &lt;: AbstractJuMPScalar</code></pre><p>A struct to represent a nonlinear expression.</p><p>Create an expression using <a href="#JuMP.@NLexpression"><code>@NLexpression</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L280-L286">source</a></section></article><h2 id="NonlinearOperator"><a class="docs-heading-anchor" href="#NonlinearOperator"><code>NonlinearOperator</code></a><a id="NonlinearOperator-1"></a><a class="docs-heading-anchor-permalink" href="#NonlinearOperator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NonlinearOperator" href="#JuMP.NonlinearOperator"><code>JuMP.NonlinearOperator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonlinearOperator(func::Function, head::Symbol)</code></pre><p>A callable struct (functor) representing a function named <code>head</code>.</p><p>When called with <a href="#JuMP.AbstractJuMPScalar"><code>AbstractJuMPScalar</code></a>s, the struct returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p>When called with non-JuMP types, the struct returns the evaluation of <code>func(args...)</code>.</p><p>Unless <code>head</code> is special-cased by the optimizer, the operator must have already been added to the model using <a href="#JuMP.add_nonlinear_operator"><code>add_nonlinear_operator</code></a> or <a href="#JuMP.@operator"><code>@operator</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; f(x::Float64) = x^2
+f (generic function with 1 method)
+
+julia&gt; ∇f(x::Float64) = 2 * x
+∇f (generic function with 1 method)
+
+julia&gt; ∇²f(x::Float64) = 2.0
+∇²f (generic function with 1 method)
+
+julia&gt; @operator(model, op_f, 1, f, ∇f, ∇²f)
+NonlinearOperator(f, :op_f)
+
+julia&gt; bar = NonlinearOperator(f, :op_f)
+NonlinearOperator(f, :op_f)
+
+julia&gt; @objective(model, Min, bar(x))
+op_f(x)
+
+julia&gt; bar(2.0)
+4.0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp_expr.jl#L802-L846">source</a></section></article><h2 id="NonlinearParameter"><a class="docs-heading-anchor" href="#NonlinearParameter"><code>NonlinearParameter</code></a><a id="NonlinearParameter-1"></a><a class="docs-heading-anchor-permalink" href="#NonlinearParameter" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NonlinearParameter" href="#JuMP.NonlinearParameter"><code>JuMP.NonlinearParameter</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonlinearParameter &lt;: AbstractJuMPScalar</code></pre><p>A struct to represent a nonlinear parameter.</p><p>Create a parameter using <a href="#JuMP.@NLparameter"><code>@NLparameter</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/nlp.jl#L188-L194">source</a></section></article><h2 id="Nonnegatives"><a class="docs-heading-anchor" href="#Nonnegatives"><code>Nonnegatives</code></a><a id="Nonnegatives-1"></a><a class="docs-heading-anchor-permalink" href="#Nonnegatives" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Nonnegatives" href="#JuMP.Nonnegatives"><code>JuMP.Nonnegatives</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Nonnegatives()</code></pre><p>The JuMP equivalent of the <a href="../../moi/reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a> set, in which the dimension is inferred from the corresponding function.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; @constraint(model, x in Nonnegatives())
+[x[1], x[2]] ∈ MathOptInterface.Nonnegatives(2)
+
+julia&gt; A = [1 2; 3 4];
+
+julia&gt; b = [5, 6];
+
+julia&gt; @constraint(model, A * x &gt;= b)
+[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonnegatives(2)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L235-L261">source</a></section></article><h2 id="Nonpositives"><a class="docs-heading-anchor" href="#Nonpositives"><code>Nonpositives</code></a><a id="Nonpositives-1"></a><a class="docs-heading-anchor-permalink" href="#Nonpositives" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Nonpositives" href="#JuMP.Nonpositives"><code>JuMP.Nonpositives</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Nonpositives()</code></pre><p>The JuMP equivalent of the <a href="../../moi/reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives</code></a> set, in which the dimension is inferred from the corresponding function.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; @constraint(model, x in Nonpositives())
+[x[1], x[2]] ∈ MathOptInterface.Nonpositives(2)
+
+julia&gt; A = [1 2; 3 4];
+
+julia&gt; b = [5, 6];
+
+julia&gt; @constraint(model, A * x &lt;= b)
+[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonpositives(2)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L264-L290">source</a></section></article><h2 id="OptimizationSense"><a class="docs-heading-anchor" href="#OptimizationSense"><code>OptimizationSense</code></a><a id="OptimizationSense-1"></a><a class="docs-heading-anchor-permalink" href="#OptimizationSense" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OptimizationSense" href="#JuMP.OptimizationSense"><code>JuMP.OptimizationSense</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">OptimizationSense</code></pre><p>An enum for the value of the <a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>ObjectiveSense</code></a> attribute.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.MIN_SENSE"><code>MIN_SENSE</code></a>: the goal is to minimize the objective function</li><li><a href="../../moi/reference/models/#MathOptInterface.MAX_SENSE"><code>MAX_SENSE</code></a>: the goal is to maximize the objective function</li><li><a href="../../moi/reference/models/#MathOptInterface.FEASIBILITY_SENSE"><code>FEASIBILITY_SENSE</code></a>: the model does not have an objective function</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1098-L1110">source</a></section></article><h2 id="OptimizeNotCalled"><a class="docs-heading-anchor" href="#OptimizeNotCalled"><code>OptimizeNotCalled</code></a><a id="OptimizeNotCalled-1"></a><a class="docs-heading-anchor-permalink" href="#OptimizeNotCalled" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OptimizeNotCalled" href="#JuMP.OptimizeNotCalled"><code>JuMP.OptimizeNotCalled</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct OptimizeNotCalled &lt;: Exception end</code></pre><p>A result attribute cannot be queried before <a href="#JuMP.optimize!"><code>optimize!</code></a> is called.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L633-L637">source</a></section></article><h2 id="PSDCone"><a class="docs-heading-anchor" href="#PSDCone"><code>PSDCone</code></a><a id="PSDCone-1"></a><a class="docs-heading-anchor-permalink" href="#PSDCone" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.PSDCone" href="#JuMP.PSDCone"><code>JuMP.PSDCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">PSDCone</code></pre><p>Positive semidefinite cone object that can be used to constrain a square matrix to be positive semidefinite in the <a href="#JuMP.@constraint"><code>@constraint</code></a> macro. If the matrix has type <code>Symmetric</code> then the columns vectorization (the vector obtained by concatenating the columns) of its upper triangular part is constrained to belong to the <code>MOI.PositiveSemidefiniteConeTriangle</code> set, otherwise its column vectorization is constrained to belong to the <code>MOI.PositiveSemidefiniteConeSquare</code> set.</p><p><strong>Example</strong></p><p>Consider the following example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; a = [ x 2x
+            2x  x];
+
+julia&gt; b = [1 2
+            2 4];
+
+julia&gt; cref = @constraint(model, a &gt;= b, PSDCone())
+[x - 1    2 x - 2;
+ 2 x - 2  x - 4] ∈ PSDCone()
+
+julia&gt; jump_function(constraint_object(cref))
+4-element Vector{AffExpr}:
+ x - 1
+ 2 x - 2
+ 2 x - 2
+ x - 4
+
+julia&gt; moi_set(constraint_object(cref))
+MathOptInterface.PositiveSemidefiniteConeSquare(2)</code></pre><p>We see in the output of the last command that the vectorization of the matrix is constrained to belong to the <code>PositiveSemidefiniteConeSquare</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; using LinearAlgebra # For Symmetric
+
+julia&gt; cref = @constraint(model, Symmetric(a - b) in PSDCone())
+[x - 1    2 x - 2;
+ 2 x - 2  x - 4] ∈ PSDCone()
+
+julia&gt; jump_function(constraint_object(cref))
+3-element Vector{AffExpr}:
+ x - 1
+ 2 x - 2
+ x - 4
+
+julia&gt; moi_set(constraint_object(cref))
+MathOptInterface.PositiveSemidefiniteConeTriangle(2)</code></pre><p>As we see in the output of the last command, the vectorization of only the upper triangular part of the matrix is constrained to belong to the <code>PositiveSemidefiniteConeSquare</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L75-L137">source</a></section></article><h2 id="Parameter"><a class="docs-heading-anchor" href="#Parameter"><code>Parameter</code></a><a id="Parameter-1"></a><a class="docs-heading-anchor-permalink" href="#Parameter" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Parameter" href="#JuMP.Parameter"><code>JuMP.Parameter</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Parameter(value)</code></pre><p>A short-cut for the <a href="../../moi/reference/standard_form/#MathOptInterface.Parameter"><code>MOI.Parameter</code></a> set.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x in Parameter(2))
+x
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ x ∈ MathOptInterface.Parameter{Float64}(2.0)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L364-L382">source</a></section></article><h2 id="ParameterRef"><a class="docs-heading-anchor" href="#ParameterRef"><code>ParameterRef</code></a><a id="ParameterRef-1"></a><a class="docs-heading-anchor-permalink" href="#ParameterRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ParameterRef" href="#JuMP.ParameterRef"><code>JuMP.ParameterRef</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">ParameterRef(x::GenericVariableRef)</code></pre><p>Return a constraint reference to the constraint constraining <code>x</code> to be a parameter.</p><p>Errors if one does not exist.</p><p>See also <a href="#JuMP.is_parameter"><code>is_parameter</code></a>, <a href="#JuMP.set_parameter_value"><code>set_parameter_value</code></a>, <a href="#JuMP.parameter_value"><code>parameter_value</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, p in Parameter(2))
+p
+
+julia&gt; ParameterRef(p)
+p ∈ MathOptInterface.Parameter{Float64}(2.0)
+
+julia&gt; @variable(model, x);
+
+julia&gt; ParameterRef(x)
+ERROR: Variable x is not a parameter.
+Stacktrace:
+[...]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1485-L1514">source</a></section></article><h2 id="QuadExpr"><a class="docs-heading-anchor" href="#QuadExpr"><code>QuadExpr</code></a><a id="QuadExpr-1"></a><a class="docs-heading-anchor-permalink" href="#QuadExpr" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.QuadExpr" href="#JuMP.QuadExpr"><code>JuMP.QuadExpr</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">QuadExpr</code></pre><p>An alias for <code>GenericQuadExpr{Float64,VariableRef}</code>, the specific <a href="#JuMP.GenericQuadExpr"><code>GenericQuadExpr</code></a> used by JuMP.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L542-L547">source</a></section></article><h2 id="QuadTermIterator"><a class="docs-heading-anchor" href="#QuadTermIterator"><code>QuadTermIterator</code></a><a id="QuadTermIterator-1"></a><a class="docs-heading-anchor-permalink" href="#QuadTermIterator" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.QuadTermIterator" href="#JuMP.QuadTermIterator"><code>JuMP.QuadTermIterator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">QuadTermIterator{GQE&lt;:GenericQuadExpr}</code></pre><p>A struct that implements the <code>iterate</code> protocol in order to iterate over tuples of <code>(coefficient, variable, variable)</code> in the <code>GenericQuadExpr</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L274-L279">source</a></section></article><h2 id="ReferenceMap"><a class="docs-heading-anchor" href="#ReferenceMap"><code>ReferenceMap</code></a><a id="ReferenceMap-1"></a><a class="docs-heading-anchor-permalink" href="#ReferenceMap" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ReferenceMap" href="#JuMP.ReferenceMap"><code>JuMP.ReferenceMap</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GenericReferenceMap{T}</code></pre><p>Mapping between variable and constraint reference of a model and its copy. The reference of the copied model can be obtained by indexing the map with the reference of the corresponding reference of the original model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/copy.jl#L31-L37">source</a></section></article><h2 id="ResultStatusCode"><a class="docs-heading-anchor" href="#ResultStatusCode"><code>ResultStatusCode</code></a><a id="ResultStatusCode-1"></a><a class="docs-heading-anchor-permalink" href="#ResultStatusCode" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ResultStatusCode" href="#JuMP.ResultStatusCode"><code>JuMP.ResultStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ResultStatusCode</code></pre><p>An Enum of possible values for the <a href="../../moi/reference/models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> and <a href="../../moi/reference/models/#MathOptInterface.DualStatus"><code>DualStatus</code></a> attributes.</p><p>The values indicate how to interpret the result vector.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.NO_SOLUTION"><code>NO_SOLUTION</code></a>: the result vector is empty.</li><li><a href="../../moi/reference/models/#MathOptInterface.FEASIBLE_POINT"><code>FEASIBLE_POINT</code></a>: the result vector is a feasible point.</li><li><a href="../../moi/reference/models/#MathOptInterface.NEARLY_FEASIBLE_POINT"><code>NEARLY_FEASIBLE_POINT</code></a>: the result vector is feasible if some constraint tolerances are relaxed.</li><li><a href="../../moi/reference/models/#MathOptInterface.INFEASIBLE_POINT"><code>INFEASIBLE_POINT</code></a>: the result vector is an infeasible point.</li><li><a href="../../moi/reference/models/#MathOptInterface.INFEASIBILITY_CERTIFICATE"><code>INFEASIBILITY_CERTIFICATE</code></a>: the result vector is an infeasibility certificate. If the <code>PrimalStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the primal result vector is a certificate of dual infeasibility. If the <code>DualStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the dual result vector is a proof of primal infeasibility.</li><li><a href="../../moi/reference/models/#MathOptInterface.NEARLY_INFEASIBILITY_CERTIFICATE"><code>NEARLY_INFEASIBILITY_CERTIFICATE</code></a>: the result satisfies a relaxed criterion for a certificate of infeasibility.</li><li><a href="../../moi/reference/models/#MathOptInterface.REDUCTION_CERTIFICATE"><code>REDUCTION_CERTIFICATE</code></a>: the result vector is an ill-posed certificate; see <a href="https://arxiv.org/abs/1408.4685">this article</a> for details. If the <code>PrimalStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the primal result vector is a proof that the dual problem is ill-posed. If the <code>DualStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the dual result vector is a proof that the primal is ill-posed.</li><li><a href="../../moi/reference/models/#MathOptInterface.NEARLY_REDUCTION_CERTIFICATE"><code>NEARLY_REDUCTION_CERTIFICATE</code></a>: the result satisfies a relaxed criterion for an ill-posed certificate.</li><li><a href="../../moi/reference/models/#MathOptInterface.UNKNOWN_RESULT_STATUS"><code>UNKNOWN_RESULT_STATUS</code></a>: the result vector contains a solution with an unknown interpretation.</li><li><a href="../../moi/reference/models/#MathOptInterface.OTHER_RESULT_STATUS"><code>OTHER_RESULT_STATUS</code></a>: the result vector contains a solution with an interpretation not covered by one of the statuses defined above</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2080">source</a></section></article><h2 id="RotatedSecondOrderCone"><a class="docs-heading-anchor" href="#RotatedSecondOrderCone"><code>RotatedSecondOrderCone</code></a><a id="RotatedSecondOrderCone-1"></a><a class="docs-heading-anchor-permalink" href="#RotatedSecondOrderCone" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.RotatedSecondOrderCone" href="#JuMP.RotatedSecondOrderCone"><code>JuMP.RotatedSecondOrderCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RotatedSecondOrderCone</code></pre><p>Rotated second order cone object that can be used to constrain the square of the euclidean norm of a vector <code>x</code> to be less than or equal to <span>$2tu$</span> where <code>t</code> and <code>u</code> are nonnegative scalars. This is a shortcut for the <code>MOI.RotatedSecondOrderCone</code>.</p><p><strong>Example</strong></p><p>The following constrains <span>$\|(x-1, x-2)\|^2_2 \le 2tx$</span> and <span>$t, x \ge 0$</span>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, t)
+t
+
+julia&gt; @constraint(model, [t, x, x-1, x-2] in RotatedSecondOrderCone())
+[t, x, x - 1, x - 2] ∈ MathOptInterface.RotatedSecondOrderCone(4)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L157-L180">source</a></section></article><h2 id="SOS1"><a class="docs-heading-anchor" href="#SOS1"><code>SOS1</code></a><a id="SOS1-1"></a><a class="docs-heading-anchor-permalink" href="#SOS1" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SOS1" href="#JuMP.SOS1"><code>JuMP.SOS1</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOS1</code></pre><p>SOS1 (Special Ordered Sets type 1) object than can be used to constrain a vector <code>x</code> to a set where at most 1 variable can take a non-zero value, all others being at 0. The <code>weights</code>, when specified, induce an ordering of the variables; as such, they should be unique values. The <em>k</em>th element in the set corresponds to the <em>k</em>th weight in <code>weights</code>. See <a href="http://lpsolve.sourceforge.net/5.5/SOS.htm">here</a> for a description of SOS constraints and their potential uses. This is a shortcut for the <code>MathOptInterface.SOS1</code> set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L184-L195">source</a></section></article><h2 id="SOS2"><a class="docs-heading-anchor" href="#SOS2"><code>SOS2</code></a><a id="SOS2-1"></a><a class="docs-heading-anchor-permalink" href="#SOS2" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SOS2" href="#JuMP.SOS2"><code>JuMP.SOS2</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOS2</code></pre><p>SOS1 (Special Ordered Sets type 2) object than can be used to constrain a vector <code>x</code> to a set where at most 2 variables can take a non-zero value, all others being at 0. In addition, if two are non-zero these must be consecutive in their ordering. The <code>weights</code> induce an ordering of the variables; as such, they should be unique values. The <em>k</em>th element in the set corresponds to the <em>k</em>th weight in <code>weights</code>. See <a href="http://lpsolve.sourceforge.net/5.5/SOS.htm">here</a> for a description of SOS constraints and their potential uses. This is a shortcut for the <code>MathOptInterface.SOS2</code> set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L215-L227">source</a></section></article><h2 id="ScalarConstraint"><a class="docs-heading-anchor" href="#ScalarConstraint"><code>ScalarConstraint</code></a><a id="ScalarConstraint-1"></a><a class="docs-heading-anchor-permalink" href="#ScalarConstraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ScalarConstraint" href="#JuMP.ScalarConstraint"><code>JuMP.ScalarConstraint</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ScalarConstraint</code></pre><p>The data for a scalar constraint. The <code>func</code> field contains a JuMP object representing the function and the <code>set</code> field contains the MOI set. See also the <a href="../../moi/manual/constraints/#Constraints">documentation</a> on JuMP&#39;s representation of constraints for more background.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L577-L584">source</a></section></article><h2 id="ScalarShape"><a class="docs-heading-anchor" href="#ScalarShape"><code>ScalarShape</code></a><a id="ScalarShape-1"></a><a class="docs-heading-anchor-permalink" href="#ScalarShape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ScalarShape" href="#JuMP.ScalarShape"><code>JuMP.ScalarShape</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarShape</code></pre><p>Shape of scalar constraints.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/shapes.jl#L108-L112">source</a></section></article><h2 id="ScalarVariable"><a class="docs-heading-anchor" href="#ScalarVariable"><code>ScalarVariable</code></a><a id="ScalarVariable-1"></a><a class="docs-heading-anchor-permalink" href="#ScalarVariable" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ScalarVariable" href="#JuMP.ScalarVariable"><code>JuMP.ScalarVariable</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarVariable{S,T,U,V} &lt;: AbstractVariable</code></pre><p>A struct used when adding variables.</p><p>See also: <a href="#JuMP.add_variable"><code>add_variable</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L199-L205">source</a></section></article><h2 id="SecondOrderCone"><a class="docs-heading-anchor" href="#SecondOrderCone"><code>SecondOrderCone</code></a><a id="SecondOrderCone-1"></a><a class="docs-heading-anchor-permalink" href="#SecondOrderCone" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SecondOrderCone" href="#JuMP.SecondOrderCone"><code>JuMP.SecondOrderCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SecondOrderCone</code></pre><p>Second order cone object that can be used to constrain the euclidean norm of a vector <code>x</code> to be less than or equal to a nonnegative scalar <code>t</code>. This is a shortcut for the <code>MOI.SecondOrderCone</code>.</p><p><strong>Example</strong></p><p>The following constrains <span>$\|(x-1, x-2)\|_2 \le t$</span> and <span>$t \ge 0$</span>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, t)
+t
+
+julia&gt; @constraint(model, [t, x-1, x-2] in SecondOrderCone())
+[t, x - 1, x - 2] ∈ MathOptInterface.SecondOrderCone(3)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L131-L153">source</a></section></article><h2 id="Semicontinuous"><a class="docs-heading-anchor" href="#Semicontinuous"><code>Semicontinuous</code></a><a id="Semicontinuous-1"></a><a class="docs-heading-anchor-permalink" href="#Semicontinuous" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Semicontinuous" href="#JuMP.Semicontinuous"><code>JuMP.Semicontinuous</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Semicontinuous(lower, upper)</code></pre><p>A short-cut for the <a href="../../moi/reference/standard_form/#MathOptInterface.Semicontinuous"><code>MOI.Semicontinuous</code></a> set.</p><p>This short-cut is useful because it automatically promotes <code>lower</code> and <code>upper</code> to the same type, and converts them into the element type supported by the JuMP model.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x in Semicontinuous(1, 2))
+x
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ x ∈ MathOptInterface.Semicontinuous{Int64}(1, 2)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L292-L314">source</a></section></article><h2 id="Semiinteger"><a class="docs-heading-anchor" href="#Semiinteger"><code>Semiinteger</code></a><a id="Semiinteger-1"></a><a class="docs-heading-anchor-permalink" href="#Semiinteger" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Semiinteger" href="#JuMP.Semiinteger"><code>JuMP.Semiinteger</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Semiinteger(lower, upper)</code></pre><p>A short-cut for the <a href="../../moi/reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger</code></a> set.</p><p>This short-cut is useful because it automatically promotes <code>lower</code> and <code>upper</code> to the same type, and converts them into the element type supported by the JuMP model.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x in Semiinteger(3, 5))
+x
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ x ∈ MathOptInterface.Semiinteger{Int64}(3, 5)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sets.jl#L328-L350">source</a></section></article><h2 id="SensitivityReport"><a class="docs-heading-anchor" href="#SensitivityReport"><code>SensitivityReport</code></a><a id="SensitivityReport-1"></a><a class="docs-heading-anchor-permalink" href="#SensitivityReport" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SensitivityReport" href="#JuMP.SensitivityReport"><code>JuMP.SensitivityReport</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SensitivityReport</code></pre><p>See <a href="#JuMP.lp_sensitivity_report"><code>lp_sensitivity_report</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/lp_sensitivity2.jl#L6-L10">source</a></section></article><h2 id="SkewSymmetricMatrixShape"><a class="docs-heading-anchor" href="#SkewSymmetricMatrixShape"><code>SkewSymmetricMatrixShape</code></a><a id="SkewSymmetricMatrixShape-1"></a><a class="docs-heading-anchor-permalink" href="#SkewSymmetricMatrixShape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SkewSymmetricMatrixShape" href="#JuMP.SkewSymmetricMatrixShape"><code>JuMP.SkewSymmetricMatrixShape</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SkewSymmetricMatrixShape</code></pre><p>Shape object for a skew symmetric square matrix of <code>side_dimension</code> rows and columns. The vectorized form contains the entries of the upper-right triangular part of the matrix (without the diagonal) given column by column (or equivalently, the entries of the lower-left triangular part given row by row). The diagonal is zero.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L201-L209">source</a></section></article><h2 id="SkewSymmetricMatrixSpace"><a class="docs-heading-anchor" href="#SkewSymmetricMatrixSpace"><code>SkewSymmetricMatrixSpace</code></a><a id="SkewSymmetricMatrixSpace-1"></a><a class="docs-heading-anchor-permalink" href="#SkewSymmetricMatrixSpace" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SkewSymmetricMatrixSpace" href="#JuMP.SkewSymmetricMatrixSpace"><code>JuMP.SkewSymmetricMatrixSpace</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SkewSymmetricMatrixSpace()</code></pre><p>Use in the <a href="#JuMP.@variable"><code>@variable</code></a> macro to constrain a matrix of variables to be skew-symmetric.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2] in SkewSymmetricMatrixSpace())
+2×2 Matrix{AffExpr}:
+ 0        Q[1,2]
+ -Q[1,2]  0</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L37-L53">source</a></section></article><h2 id="SquareMatrixShape"><a class="docs-heading-anchor" href="#SquareMatrixShape"><code>SquareMatrixShape</code></a><a id="SquareMatrixShape-1"></a><a class="docs-heading-anchor-permalink" href="#SquareMatrixShape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SquareMatrixShape" href="#JuMP.SquareMatrixShape"><code>JuMP.SquareMatrixShape</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SquareMatrixShape</code></pre><p>Shape object for a square matrix of <code>side_dimension</code> rows and columns. The vectorized form contains the entries of the the matrix given column by column (or equivalently, the entries of the lower-left triangular part given row by row).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L240-L247">source</a></section></article><h2 id="SymmetricMatrixShape"><a class="docs-heading-anchor" href="#SymmetricMatrixShape"><code>SymmetricMatrixShape</code></a><a id="SymmetricMatrixShape-1"></a><a class="docs-heading-anchor-permalink" href="#SymmetricMatrixShape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SymmetricMatrixShape" href="#JuMP.SymmetricMatrixShape"><code>JuMP.SymmetricMatrixShape</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SymmetricMatrixShape</code></pre><p>Shape object for a symmetric square matrix of <code>side_dimension</code> rows and columns. The vectorized form contains the entries of the upper-right triangular part of the matrix given column by column (or equivalently, the entries of the lower-left triangular part given row by row).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L140-L147">source</a></section></article><h2 id="SymmetricMatrixSpace"><a class="docs-heading-anchor" href="#SymmetricMatrixSpace"><code>SymmetricMatrixSpace</code></a><a id="SymmetricMatrixSpace-1"></a><a class="docs-heading-anchor-permalink" href="#SymmetricMatrixSpace" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SymmetricMatrixSpace" href="#JuMP.SymmetricMatrixSpace"><code>JuMP.SymmetricMatrixSpace</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SymmetricMatrixSpace()</code></pre><p>Use in the <a href="#JuMP.@variable"><code>@variable</code></a> macro to constrain a matrix of variables to be symmetric.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, Q[1:2, 1:2] in SymmetricMatrixSpace())
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ Q[1,1]  Q[1,2]
+ Q[1,2]  Q[2,2]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/sd.jl#L18-L34">source</a></section></article><h2 id="TerminationStatusCode"><a class="docs-heading-anchor" href="#TerminationStatusCode"><code>TerminationStatusCode</code></a><a id="TerminationStatusCode-1"></a><a class="docs-heading-anchor-permalink" href="#TerminationStatusCode" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.TerminationStatusCode" href="#JuMP.TerminationStatusCode"><code>JuMP.TerminationStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">TerminationStatusCode</code></pre><p>An Enum of possible values for the <code>TerminationStatus</code> attribute. This attribute is meant to explain the reason why the optimizer stopped executing in the most recent call to <a href="../../moi/reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a>.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.OPTIMIZE_NOT_CALLED"><code>OPTIMIZE_NOT_CALLED</code></a>: The algorithm has not started.</li><li><a href="../../moi/reference/models/#MathOptInterface.OPTIMAL"><code>OPTIMAL</code></a>: The algorithm found a globally optimal solution.</li><li><a href="../../moi/reference/models/#MathOptInterface.INFEASIBLE"><code>INFEASIBLE</code></a>: The algorithm concluded that no feasible solution exists.</li><li><a href="../../moi/reference/models/#MathOptInterface.DUAL_INFEASIBLE"><code>DUAL_INFEASIBLE</code></a>: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.</li><li><a href="../../moi/reference/models/#MathOptInterface.LOCALLY_SOLVED"><code>LOCALLY_SOLVED</code></a>: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.</li><li><a href="../../moi/reference/models/#MathOptInterface.LOCALLY_INFEASIBLE"><code>LOCALLY_INFEASIBLE</code></a>: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.</li><li><a href="../../moi/reference/models/#MathOptInterface.INFEASIBLE_OR_UNBOUNDED"><code>INFEASIBLE_OR_UNBOUNDED</code></a>: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.</li><li><a href="../../moi/reference/models/#MathOptInterface.ALMOST_OPTIMAL"><code>ALMOST_OPTIMAL</code></a>: The algorithm found a globally optimal solution to relaxed tolerances.</li><li><a href="../../moi/reference/models/#MathOptInterface.ALMOST_INFEASIBLE"><code>ALMOST_INFEASIBLE</code></a>: The algorithm concluded that no feasible solution exists within relaxed tolerances.</li><li><a href="../../moi/reference/models/#MathOptInterface.ALMOST_DUAL_INFEASIBLE"><code>ALMOST_DUAL_INFEASIBLE</code></a>: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.</li><li><a href="../../moi/reference/models/#MathOptInterface.ALMOST_LOCALLY_SOLVED"><code>ALMOST_LOCALLY_SOLVED</code></a>: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.</li><li><a href="../../moi/reference/models/#MathOptInterface.ITERATION_LIMIT"><code>ITERATION_LIMIT</code></a>: An iterative algorithm stopped after conducting the maximum number of iterations.</li><li><a href="../../moi/reference/models/#MathOptInterface.TIME_LIMIT"><code>TIME_LIMIT</code></a>: The algorithm stopped after a user-specified computation time.</li><li><a href="../../moi/reference/models/#MathOptInterface.NODE_LIMIT"><code>NODE_LIMIT</code></a>: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.</li><li><a href="../../moi/reference/models/#MathOptInterface.SOLUTION_LIMIT"><code>SOLUTION_LIMIT</code></a>: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.</li><li><a href="../../moi/reference/models/#MathOptInterface.MEMORY_LIMIT"><code>MEMORY_LIMIT</code></a>: The algorithm stopped because it ran out of memory.</li><li><a href="../../moi/reference/models/#MathOptInterface.OBJECTIVE_LIMIT"><code>OBJECTIVE_LIMIT</code></a>: The algorithm stopped because it found a solution better than a minimum limit set by the user.</li><li><a href="../../moi/reference/models/#MathOptInterface.NORM_LIMIT"><code>NORM_LIMIT</code></a>: The algorithm stopped because the norm of an iterate became too large.</li><li><a href="../../moi/reference/models/#MathOptInterface.OTHER_LIMIT"><code>OTHER_LIMIT</code></a>: The algorithm stopped due to a limit not covered by one of the <code>_LIMIT_</code> statuses above.</li><li><a href="../../moi/reference/models/#MathOptInterface.SLOW_PROGRESS"><code>SLOW_PROGRESS</code></a>: The algorithm stopped because it was unable to continue making progress towards the solution.</li><li><a href="../../moi/reference/models/#MathOptInterface.NUMERICAL_ERROR"><code>NUMERICAL_ERROR</code></a>: The algorithm stopped because it encountered unrecoverable numerical error.</li><li><a href="../../moi/reference/models/#MathOptInterface.INVALID_MODEL"><code>INVALID_MODEL</code></a>: The algorithm stopped because the model is invalid.</li><li><a href="../../moi/reference/models/#MathOptInterface.INVALID_OPTION"><code>INVALID_OPTION</code></a>: The algorithm stopped because it was provided an invalid option.</li><li><a href="../../moi/reference/models/#MathOptInterface.INTERRUPTED"><code>INTERRUPTED</code></a>: The algorithm stopped because of an interrupt signal.</li><li><a href="../../moi/reference/models/#MathOptInterface.OTHER_ERROR"><code>OTHER_ERROR</code></a>: The algorithm stopped because of an error not covered by one of the statuses defined above.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L2009">source</a></section></article><h2 id="UnorderedPair"><a class="docs-heading-anchor" href="#UnorderedPair"><code>UnorderedPair</code></a><a id="UnorderedPair-1"></a><a class="docs-heading-anchor-permalink" href="#UnorderedPair" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.UnorderedPair" href="#JuMP.UnorderedPair"><code>JuMP.UnorderedPair</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UnorderedPair(a::T, b::T)</code></pre><p>A wrapper type used by <a href="#JuMP.GenericQuadExpr"><code>GenericQuadExpr</code></a> with fields <code>.a</code> and <code>.b</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/quad_expr.jl#L18-L22">source</a></section></article><h2 id="UpperBoundRef"><a class="docs-heading-anchor" href="#UpperBoundRef"><code>UpperBoundRef</code></a><a id="UpperBoundRef-1"></a><a class="docs-heading-anchor-permalink" href="#UpperBoundRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.UpperBoundRef" href="#JuMP.UpperBoundRef"><code>JuMP.UpperBoundRef</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">UpperBoundRef(v::GenericVariableRef)</code></pre><p>Return a constraint reference to the upper bound constraint of <code>v</code>.</p><p>Errors if one does not exist.</p><p>See also <a href="#JuMP.has_upper_bound"><code>has_upper_bound</code></a>, <a href="#JuMP.upper_bound"><code>upper_bound</code></a>, <a href="#JuMP.set_upper_bound"><code>set_upper_bound</code></a>, <a href="#JuMP.delete_upper_bound"><code>delete_upper_bound</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &lt;= 1.0);
+
+julia&gt; UpperBoundRef(x)
+x ≤ 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L922-L942">source</a></section></article><h2 id="VariableConstrainedOnCreation"><a class="docs-heading-anchor" href="#VariableConstrainedOnCreation"><code>VariableConstrainedOnCreation</code></a><a id="VariableConstrainedOnCreation-1"></a><a class="docs-heading-anchor-permalink" href="#VariableConstrainedOnCreation" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.VariableConstrainedOnCreation" href="#JuMP.VariableConstrainedOnCreation"><code>JuMP.VariableConstrainedOnCreation</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariableConstrainedOnCreation &lt;: AbstractVariable</code></pre><p>Variable <code>scalar_variables</code> constrained to belong to <code>set</code>.</p><p>Adding this variable can be understood as doing:</p><pre><code class="language-julia hljs">function JuMP.add_variable(
+    model::GenericModel,
+    variable::VariableConstrainedOnCreation,
+    names,
+)
+    var_ref = add_variable(model, variable.scalar_variable, name)
+    add_constraint(model, VectorConstraint(var_ref, variable.set))
+    return var_ref
+end</code></pre><p>but adds the variables with <code>MOI.add_constrained_variable(model, variable.set)</code> instead. See <a href="https://jump.dev/MathOptInterface.jl/v0.9.3/apireference/#Variables-1">the MOI documentation</a> for the difference between adding the variables with <code>MOI.add_constrained_variable</code> and adding them with <code>MOI.add_variable</code> and adding the constraint separately.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1790-L1811">source</a></section></article><h2 id="VariableInfo"><a class="docs-heading-anchor" href="#VariableInfo"><code>VariableInfo</code></a><a id="VariableInfo-1"></a><a class="docs-heading-anchor-permalink" href="#VariableInfo" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.VariableInfo" href="#JuMP.VariableInfo"><code>JuMP.VariableInfo</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariableInfo{S,T,U,V}</code></pre><p>A struct by JuMP internally when creating variables. This may also be used by JuMP extensions to create new types of variables.</p><p>See also: <a href="#JuMP.ScalarVariable"><code>ScalarVariable</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L127-L134">source</a></section></article><h2 id="VariableNotOwned"><a class="docs-heading-anchor" href="#VariableNotOwned"><code>VariableNotOwned</code></a><a id="VariableNotOwned-1"></a><a class="docs-heading-anchor-permalink" href="#VariableNotOwned" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.VariableNotOwned" href="#JuMP.VariableNotOwned"><code>JuMP.VariableNotOwned</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct VariableNotOwned{V&lt;:AbstractVariableRef} &lt;: Exception
+    variable::V
+end</code></pre><p>The variable <code>variable</code> was used in a model different to <code>owner_model(variable)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L274-L281">source</a></section></article><h2 id="VariableRef"><a class="docs-heading-anchor" href="#VariableRef"><code>VariableRef</code></a><a id="VariableRef-1"></a><a class="docs-heading-anchor-permalink" href="#VariableRef" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.VariableRef" href="#JuMP.VariableRef"><code>JuMP.VariableRef</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GenericVariableRef{T} &lt;: AbstractVariableRef</code></pre><p>Holds a reference to the model and the corresponding MOI.VariableIndex.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L237-L241">source</a></section></article><h2 id="VariablesConstrainedOnCreation"><a class="docs-heading-anchor" href="#VariablesConstrainedOnCreation"><code>VariablesConstrainedOnCreation</code></a><a id="VariablesConstrainedOnCreation-1"></a><a class="docs-heading-anchor-permalink" href="#VariablesConstrainedOnCreation" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.VariablesConstrainedOnCreation" href="#JuMP.VariablesConstrainedOnCreation"><code>JuMP.VariablesConstrainedOnCreation</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariablesConstrainedOnCreation &lt;: AbstractVariable</code></pre><p>Vector of variables <code>scalar_variables</code> constrained to belong to <code>set</code>. Adding this variable can be thought as doing:</p><pre><code class="language-julia hljs">function JuMP.add_variable(
+    model::GenericModel,
+    variable::VariablesConstrainedOnCreation,
+    names,
+)
+    v_names = vectorize(names, variable.shape)
+    var_refs = add_variable.(model, variable.scalar_variables, v_names)
+    add_constraint(model, VectorConstraint(var_refs, variable.set))
+    return reshape_vector(var_refs, variable.shape)
+end</code></pre><p>but adds the variables with <code>MOI.add_constrained_variables(model, variable.set)</code> instead. See <a href="https://jump.dev/MathOptInterface.jl/v0.9.3/apireference/#Variables-1">the MOI documentation</a> for the difference between adding the variables with <code>MOI.add_constrained_variables</code> and adding them with <code>MOI.add_variables</code> and adding the constraint separately.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/variables.jl#L1866-L1887">source</a></section></article><h2 id="VectorConstraint"><a class="docs-heading-anchor" href="#VectorConstraint"><code>VectorConstraint</code></a><a id="VectorConstraint-1"></a><a class="docs-heading-anchor-permalink" href="#VectorConstraint" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.VectorConstraint" href="#JuMP.VectorConstraint"><code>JuMP.VectorConstraint</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct VectorConstraint</code></pre><p>The data for a vector constraint. The <code>func</code> field contains a JuMP object representing the function and the <code>set</code> field contains the MOI set. The <code>shape</code> field contains an <a href="#JuMP.AbstractShape"><code>AbstractShape</code></a> matching the form in which the constraint was constructed (e.g., by using matrices or flat vectors). See also the <a href="../../moi/manual/constraints/#Constraints">documentation</a> on JuMP&#39;s representation of constraints.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/constraints.jl#L611-L620">source</a></section></article><h2 id="VectorShape"><a class="docs-heading-anchor" href="#VectorShape"><code>VectorShape</code></a><a id="VectorShape-1"></a><a class="docs-heading-anchor-permalink" href="#VectorShape" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.VectorShape" href="#JuMP.VectorShape"><code>JuMP.VectorShape</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorShape</code></pre><p>Vector for which the vectorized form corresponds exactly to the vector given.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/shapes.jl#L116-L120">source</a></section></article><h2 id="Zeros"><a class="docs-heading-anchor" href="#Zeros"><code>Zeros</code></a><a id="Zeros-1"></a><a class="docs-heading-anchor-permalink" href="#Zeros" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.Zeros" href="#JuMP.Zeros"><code>JuMP.Zeros</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Zeros()</code></pre><p>The JuMP equivalent of the <a href="../../moi/reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a> set, in which the dimension is inferred from the corresponding function.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; @constraint(model, x in Zeros())
+[x[1], x[2]] ∈ MathOptInterface.Zeros(2)
+
+julia&gt; A = [1 2; 3 4];
+
+julia&gt; b = [5, 6];
+
+julia&gt; @constraint(model, A * x == b)
+[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Zeros(2)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L293-L319">source</a></section></article><h2 id="ALMOST_DUAL_INFEASIBLE"><a class="docs-heading-anchor" href="#ALMOST_DUAL_INFEASIBLE"><code>ALMOST_DUAL_INFEASIBLE</code></a><a id="ALMOST_DUAL_INFEASIBLE-1"></a><a class="docs-heading-anchor-permalink" href="#ALMOST_DUAL_INFEASIBLE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ALMOST_DUAL_INFEASIBLE" href="#JuMP.ALMOST_DUAL_INFEASIBLE"><code>JuMP.ALMOST_DUAL_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_DUAL_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_DUAL_INFEASIBLE</code>: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="ALMOST_INFEASIBLE"><a class="docs-heading-anchor" href="#ALMOST_INFEASIBLE"><code>ALMOST_INFEASIBLE</code></a><a id="ALMOST_INFEASIBLE-1"></a><a class="docs-heading-anchor-permalink" href="#ALMOST_INFEASIBLE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ALMOST_INFEASIBLE" href="#JuMP.ALMOST_INFEASIBLE"><code>JuMP.ALMOST_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_INFEASIBLE</code>: The algorithm concluded that no feasible solution exists within relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="ALMOST_LOCALLY_SOLVED"><a class="docs-heading-anchor" href="#ALMOST_LOCALLY_SOLVED"><code>ALMOST_LOCALLY_SOLVED</code></a><a id="ALMOST_LOCALLY_SOLVED-1"></a><a class="docs-heading-anchor-permalink" href="#ALMOST_LOCALLY_SOLVED" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ALMOST_LOCALLY_SOLVED" href="#JuMP.ALMOST_LOCALLY_SOLVED"><code>JuMP.ALMOST_LOCALLY_SOLVED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_LOCALLY_SOLVED::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_LOCALLY_SOLVED</code>: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="ALMOST_OPTIMAL"><a class="docs-heading-anchor" href="#ALMOST_OPTIMAL"><code>ALMOST_OPTIMAL</code></a><a id="ALMOST_OPTIMAL-1"></a><a class="docs-heading-anchor-permalink" href="#ALMOST_OPTIMAL" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ALMOST_OPTIMAL" href="#JuMP.ALMOST_OPTIMAL"><code>JuMP.ALMOST_OPTIMAL</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_OPTIMAL::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_OPTIMAL</code>: The algorithm found a globally optimal solution to relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="AUTOMATIC"><a class="docs-heading-anchor" href="#AUTOMATIC"><code>AUTOMATIC</code></a><a id="AUTOMATIC-1"></a><a class="docs-heading-anchor-permalink" href="#AUTOMATIC" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.AUTOMATIC" href="#JuMP.AUTOMATIC"><code>JuMP.AUTOMATIC</code></a> — <span class="docstring-category">Constant</span></header><section><div><p><code>moi_backend</code> field holds a CachingOptimizer in AUTOMATIC mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L57">source</a></section></article><h2 id="DIRECT"><a class="docs-heading-anchor" href="#DIRECT"><code>DIRECT</code></a><a id="DIRECT-1"></a><a class="docs-heading-anchor-permalink" href="#DIRECT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.DIRECT" href="#JuMP.DIRECT"><code>JuMP.DIRECT</code></a> — <span class="docstring-category">Constant</span></header><section><div><p><code>moi_backend</code> field holds an AbstractOptimizer. No extra copy of the model is stored. The <code>moi_backend</code> must support <code>add_constraint</code> etc.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L64">source</a></section></article><h2 id="DUAL_INFEASIBLE"><a class="docs-heading-anchor" href="#DUAL_INFEASIBLE"><code>DUAL_INFEASIBLE</code></a><a id="DUAL_INFEASIBLE-1"></a><a class="docs-heading-anchor-permalink" href="#DUAL_INFEASIBLE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.DUAL_INFEASIBLE" href="#JuMP.DUAL_INFEASIBLE"><code>JuMP.DUAL_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">DUAL_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>DUAL_INFEASIBLE</code>: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="FEASIBILITY_SENSE"><a class="docs-heading-anchor" href="#FEASIBILITY_SENSE"><code>FEASIBILITY_SENSE</code></a><a id="FEASIBILITY_SENSE-1"></a><a class="docs-heading-anchor-permalink" href="#FEASIBILITY_SENSE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.FEASIBILITY_SENSE" href="#JuMP.FEASIBILITY_SENSE"><code>JuMP.FEASIBILITY_SENSE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">FEASIBILITY_SENSE::OptimizationSense</code></pre><p>An instance of the <a href="#JuMP.OptimizationSense"><code>OptimizationSense</code></a> enum.</p><p><code>FEASIBILITY_SENSE</code>: the model does not have an objective function</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="FEASIBLE_POINT"><a class="docs-heading-anchor" href="#FEASIBLE_POINT"><code>FEASIBLE_POINT</code></a><a id="FEASIBLE_POINT-1"></a><a class="docs-heading-anchor-permalink" href="#FEASIBLE_POINT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.FEASIBLE_POINT" href="#JuMP.FEASIBLE_POINT"><code>JuMP.FEASIBLE_POINT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">FEASIBLE_POINT::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>FEASIBLE_POINT</code>: the result vector is a feasible point.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="INFEASIBILITY_CERTIFICATE"><a class="docs-heading-anchor" href="#INFEASIBILITY_CERTIFICATE"><code>INFEASIBILITY_CERTIFICATE</code></a><a id="INFEASIBILITY_CERTIFICATE-1"></a><a class="docs-heading-anchor-permalink" href="#INFEASIBILITY_CERTIFICATE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.INFEASIBILITY_CERTIFICATE" href="#JuMP.INFEASIBILITY_CERTIFICATE"><code>JuMP.INFEASIBILITY_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBILITY_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>INFEASIBILITY_CERTIFICATE</code>: the result vector is an infeasibility certificate. If the <code>PrimalStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the primal result vector is a certificate of dual infeasibility. If the <code>DualStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the dual result vector is a proof of primal infeasibility.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="INFEASIBLE"><a class="docs-heading-anchor" href="#INFEASIBLE"><code>INFEASIBLE</code></a><a id="INFEASIBLE-1"></a><a class="docs-heading-anchor-permalink" href="#INFEASIBLE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.INFEASIBLE" href="#JuMP.INFEASIBLE"><code>JuMP.INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INFEASIBLE</code>: The algorithm concluded that no feasible solution exists.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="INFEASIBLE_OR_UNBOUNDED"><a class="docs-heading-anchor" href="#INFEASIBLE_OR_UNBOUNDED"><code>INFEASIBLE_OR_UNBOUNDED</code></a><a id="INFEASIBLE_OR_UNBOUNDED-1"></a><a class="docs-heading-anchor-permalink" href="#INFEASIBLE_OR_UNBOUNDED" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.INFEASIBLE_OR_UNBOUNDED" href="#JuMP.INFEASIBLE_OR_UNBOUNDED"><code>JuMP.INFEASIBLE_OR_UNBOUNDED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBLE_OR_UNBOUNDED::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INFEASIBLE_OR_UNBOUNDED</code>: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="INFEASIBLE_POINT"><a class="docs-heading-anchor" href="#INFEASIBLE_POINT"><code>INFEASIBLE_POINT</code></a><a id="INFEASIBLE_POINT-1"></a><a class="docs-heading-anchor-permalink" href="#INFEASIBLE_POINT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.INFEASIBLE_POINT" href="#JuMP.INFEASIBLE_POINT"><code>JuMP.INFEASIBLE_POINT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBLE_POINT::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>INFEASIBLE_POINT</code>: the result vector is an infeasible point.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="INTERRUPTED"><a class="docs-heading-anchor" href="#INTERRUPTED"><code>INTERRUPTED</code></a><a id="INTERRUPTED-1"></a><a class="docs-heading-anchor-permalink" href="#INTERRUPTED" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.INTERRUPTED" href="#JuMP.INTERRUPTED"><code>JuMP.INTERRUPTED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INTERRUPTED::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INTERRUPTED</code>: The algorithm stopped because of an interrupt signal.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="INVALID_MODEL"><a class="docs-heading-anchor" href="#INVALID_MODEL"><code>INVALID_MODEL</code></a><a id="INVALID_MODEL-1"></a><a class="docs-heading-anchor-permalink" href="#INVALID_MODEL" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.INVALID_MODEL" href="#JuMP.INVALID_MODEL"><code>JuMP.INVALID_MODEL</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INVALID_MODEL::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INVALID_MODEL</code>: The algorithm stopped because the model is invalid.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="INVALID_OPTION"><a class="docs-heading-anchor" href="#INVALID_OPTION"><code>INVALID_OPTION</code></a><a id="INVALID_OPTION-1"></a><a class="docs-heading-anchor-permalink" href="#INVALID_OPTION" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.INVALID_OPTION" href="#JuMP.INVALID_OPTION"><code>JuMP.INVALID_OPTION</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INVALID_OPTION::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INVALID_OPTION</code>: The algorithm stopped because it was provided an invalid option.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="ITERATION_LIMIT"><a class="docs-heading-anchor" href="#ITERATION_LIMIT"><code>ITERATION_LIMIT</code></a><a id="ITERATION_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#ITERATION_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.ITERATION_LIMIT" href="#JuMP.ITERATION_LIMIT"><code>JuMP.ITERATION_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ITERATION_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ITERATION_LIMIT</code>: An iterative algorithm stopped after conducting the maximum number of iterations.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="LOCALLY_INFEASIBLE"><a class="docs-heading-anchor" href="#LOCALLY_INFEASIBLE"><code>LOCALLY_INFEASIBLE</code></a><a id="LOCALLY_INFEASIBLE-1"></a><a class="docs-heading-anchor-permalink" href="#LOCALLY_INFEASIBLE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.LOCALLY_INFEASIBLE" href="#JuMP.LOCALLY_INFEASIBLE"><code>JuMP.LOCALLY_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">LOCALLY_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>LOCALLY_INFEASIBLE</code>: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="LOCALLY_SOLVED"><a class="docs-heading-anchor" href="#LOCALLY_SOLVED"><code>LOCALLY_SOLVED</code></a><a id="LOCALLY_SOLVED-1"></a><a class="docs-heading-anchor-permalink" href="#LOCALLY_SOLVED" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.LOCALLY_SOLVED" href="#JuMP.LOCALLY_SOLVED"><code>JuMP.LOCALLY_SOLVED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">LOCALLY_SOLVED::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>LOCALLY_SOLVED</code>: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="MANUAL"><a class="docs-heading-anchor" href="#MANUAL"><code>MANUAL</code></a><a id="MANUAL-1"></a><a class="docs-heading-anchor-permalink" href="#MANUAL" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.MANUAL" href="#JuMP.MANUAL"><code>JuMP.MANUAL</code></a> — <span class="docstring-category">Constant</span></header><section><div><p><code>moi_backend</code> field holds a CachingOptimizer in MANUAL mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L62">source</a></section></article><h2 id="MAX_SENSE"><a class="docs-heading-anchor" href="#MAX_SENSE"><code>MAX_SENSE</code></a><a id="MAX_SENSE-1"></a><a class="docs-heading-anchor-permalink" href="#MAX_SENSE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.MAX_SENSE" href="#JuMP.MAX_SENSE"><code>JuMP.MAX_SENSE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">MAX_SENSE::OptimizationSense</code></pre><p>An instance of the <a href="#JuMP.OptimizationSense"><code>OptimizationSense</code></a> enum.</p><p><code>MAX_SENSE</code>: the goal is to maximize the objective function</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="MEMORY_LIMIT"><a class="docs-heading-anchor" href="#MEMORY_LIMIT"><code>MEMORY_LIMIT</code></a><a id="MEMORY_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#MEMORY_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.MEMORY_LIMIT" href="#JuMP.MEMORY_LIMIT"><code>JuMP.MEMORY_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">MEMORY_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>MEMORY_LIMIT</code>: The algorithm stopped because it ran out of memory.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="MIN_SENSE"><a class="docs-heading-anchor" href="#MIN_SENSE"><code>MIN_SENSE</code></a><a id="MIN_SENSE-1"></a><a class="docs-heading-anchor-permalink" href="#MIN_SENSE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.MIN_SENSE" href="#JuMP.MIN_SENSE"><code>JuMP.MIN_SENSE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">MIN_SENSE::OptimizationSense</code></pre><p>An instance of the <a href="#JuMP.OptimizationSense"><code>OptimizationSense</code></a> enum.</p><p><code>MIN_SENSE</code>: the goal is to minimize the objective function</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="NEARLY_FEASIBLE_POINT"><a class="docs-heading-anchor" href="#NEARLY_FEASIBLE_POINT"><code>NEARLY_FEASIBLE_POINT</code></a><a id="NEARLY_FEASIBLE_POINT-1"></a><a class="docs-heading-anchor-permalink" href="#NEARLY_FEASIBLE_POINT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NEARLY_FEASIBLE_POINT" href="#JuMP.NEARLY_FEASIBLE_POINT"><code>JuMP.NEARLY_FEASIBLE_POINT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NEARLY_FEASIBLE_POINT::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NEARLY_FEASIBLE_POINT</code>: the result vector is feasible if some constraint tolerances are relaxed.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="NEARLY_INFEASIBILITY_CERTIFICATE"><a class="docs-heading-anchor" href="#NEARLY_INFEASIBILITY_CERTIFICATE"><code>NEARLY_INFEASIBILITY_CERTIFICATE</code></a><a id="NEARLY_INFEASIBILITY_CERTIFICATE-1"></a><a class="docs-heading-anchor-permalink" href="#NEARLY_INFEASIBILITY_CERTIFICATE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NEARLY_INFEASIBILITY_CERTIFICATE" href="#JuMP.NEARLY_INFEASIBILITY_CERTIFICATE"><code>JuMP.NEARLY_INFEASIBILITY_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NEARLY_INFEASIBILITY_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NEARLY_INFEASIBILITY_CERTIFICATE</code>: the result satisfies a relaxed criterion for a certificate of infeasibility.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="NEARLY_REDUCTION_CERTIFICATE"><a class="docs-heading-anchor" href="#NEARLY_REDUCTION_CERTIFICATE"><code>NEARLY_REDUCTION_CERTIFICATE</code></a><a id="NEARLY_REDUCTION_CERTIFICATE-1"></a><a class="docs-heading-anchor-permalink" href="#NEARLY_REDUCTION_CERTIFICATE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NEARLY_REDUCTION_CERTIFICATE" href="#JuMP.NEARLY_REDUCTION_CERTIFICATE"><code>JuMP.NEARLY_REDUCTION_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NEARLY_REDUCTION_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NEARLY_REDUCTION_CERTIFICATE</code>: the result satisfies a relaxed criterion for an ill-posed certificate.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="NODE_LIMIT"><a class="docs-heading-anchor" href="#NODE_LIMIT"><code>NODE_LIMIT</code></a><a id="NODE_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#NODE_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NODE_LIMIT" href="#JuMP.NODE_LIMIT"><code>JuMP.NODE_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NODE_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>NODE_LIMIT</code>: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="NORM_LIMIT"><a class="docs-heading-anchor" href="#NORM_LIMIT"><code>NORM_LIMIT</code></a><a id="NORM_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#NORM_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NORM_LIMIT" href="#JuMP.NORM_LIMIT"><code>JuMP.NORM_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NORM_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>NORM_LIMIT</code>: The algorithm stopped because the norm of an iterate became too large.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="NO_SOLUTION"><a class="docs-heading-anchor" href="#NO_SOLUTION"><code>NO_SOLUTION</code></a><a id="NO_SOLUTION-1"></a><a class="docs-heading-anchor-permalink" href="#NO_SOLUTION" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NO_SOLUTION" href="#JuMP.NO_SOLUTION"><code>JuMP.NO_SOLUTION</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NO_SOLUTION::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NO_SOLUTION</code>: the result vector is empty.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="NUMERICAL_ERROR"><a class="docs-heading-anchor" href="#NUMERICAL_ERROR"><code>NUMERICAL_ERROR</code></a><a id="NUMERICAL_ERROR-1"></a><a class="docs-heading-anchor-permalink" href="#NUMERICAL_ERROR" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.NUMERICAL_ERROR" href="#JuMP.NUMERICAL_ERROR"><code>JuMP.NUMERICAL_ERROR</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NUMERICAL_ERROR::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>NUMERICAL_ERROR</code>: The algorithm stopped because it encountered unrecoverable numerical error.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="OBJECTIVE_LIMIT"><a class="docs-heading-anchor" href="#OBJECTIVE_LIMIT"><code>OBJECTIVE_LIMIT</code></a><a id="OBJECTIVE_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#OBJECTIVE_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OBJECTIVE_LIMIT" href="#JuMP.OBJECTIVE_LIMIT"><code>JuMP.OBJECTIVE_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OBJECTIVE_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OBJECTIVE_LIMIT</code>: The algorithm stopped because it found a solution better than a minimum limit set by the user.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="OPTIMAL"><a class="docs-heading-anchor" href="#OPTIMAL"><code>OPTIMAL</code></a><a id="OPTIMAL-1"></a><a class="docs-heading-anchor-permalink" href="#OPTIMAL" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OPTIMAL" href="#JuMP.OPTIMAL"><code>JuMP.OPTIMAL</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OPTIMAL::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OPTIMAL</code>: The algorithm found a globally optimal solution.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="OPTIMIZE_NOT_CALLED"><a class="docs-heading-anchor" href="#OPTIMIZE_NOT_CALLED"><code>OPTIMIZE_NOT_CALLED</code></a><a id="OPTIMIZE_NOT_CALLED-1"></a><a class="docs-heading-anchor-permalink" href="#OPTIMIZE_NOT_CALLED" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OPTIMIZE_NOT_CALLED" href="#JuMP.OPTIMIZE_NOT_CALLED"><code>JuMP.OPTIMIZE_NOT_CALLED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OPTIMIZE_NOT_CALLED::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OPTIMIZE_NOT_CALLED</code>: The algorithm has not started.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="OTHER_ERROR"><a class="docs-heading-anchor" href="#OTHER_ERROR"><code>OTHER_ERROR</code></a><a id="OTHER_ERROR-1"></a><a class="docs-heading-anchor-permalink" href="#OTHER_ERROR" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OTHER_ERROR" href="#JuMP.OTHER_ERROR"><code>JuMP.OTHER_ERROR</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OTHER_ERROR::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OTHER_ERROR</code>: The algorithm stopped because of an error not covered by one of the statuses defined above.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="OTHER_LIMIT"><a class="docs-heading-anchor" href="#OTHER_LIMIT"><code>OTHER_LIMIT</code></a><a id="OTHER_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#OTHER_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OTHER_LIMIT" href="#JuMP.OTHER_LIMIT"><code>JuMP.OTHER_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OTHER_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OTHER_LIMIT</code>: The algorithm stopped due to a limit not covered by one of the <code>_LIMIT_</code> statuses above.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="OTHER_RESULT_STATUS"><a class="docs-heading-anchor" href="#OTHER_RESULT_STATUS"><code>OTHER_RESULT_STATUS</code></a><a id="OTHER_RESULT_STATUS-1"></a><a class="docs-heading-anchor-permalink" href="#OTHER_RESULT_STATUS" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.OTHER_RESULT_STATUS" href="#JuMP.OTHER_RESULT_STATUS"><code>JuMP.OTHER_RESULT_STATUS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OTHER_RESULT_STATUS::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>OTHER_RESULT_STATUS</code>: the result vector contains a solution with an interpretation not covered by one of the statuses defined above</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="REDUCTION_CERTIFICATE"><a class="docs-heading-anchor" href="#REDUCTION_CERTIFICATE"><code>REDUCTION_CERTIFICATE</code></a><a id="REDUCTION_CERTIFICATE-1"></a><a class="docs-heading-anchor-permalink" href="#REDUCTION_CERTIFICATE" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.REDUCTION_CERTIFICATE" href="#JuMP.REDUCTION_CERTIFICATE"><code>JuMP.REDUCTION_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">REDUCTION_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>REDUCTION_CERTIFICATE</code>: the result vector is an ill-posed certificate; see <a href="https://arxiv.org/abs/1408.4685">this article</a> for details. If the <code>PrimalStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the primal result vector is a proof that the dual problem is ill-posed. If the <code>DualStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the dual result vector is a proof that the primal is ill-posed.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="SLOW_PROGRESS"><a class="docs-heading-anchor" href="#SLOW_PROGRESS"><code>SLOW_PROGRESS</code></a><a id="SLOW_PROGRESS-1"></a><a class="docs-heading-anchor-permalink" href="#SLOW_PROGRESS" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SLOW_PROGRESS" href="#JuMP.SLOW_PROGRESS"><code>JuMP.SLOW_PROGRESS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">SLOW_PROGRESS::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>SLOW_PROGRESS</code>: The algorithm stopped because it was unable to continue making progress towards the solution.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="SOLUTION_LIMIT"><a class="docs-heading-anchor" href="#SOLUTION_LIMIT"><code>SOLUTION_LIMIT</code></a><a id="SOLUTION_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#SOLUTION_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.SOLUTION_LIMIT" href="#JuMP.SOLUTION_LIMIT"><code>JuMP.SOLUTION_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">SOLUTION_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>SOLUTION_LIMIT</code>: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="TIME_LIMIT"><a class="docs-heading-anchor" href="#TIME_LIMIT"><code>TIME_LIMIT</code></a><a id="TIME_LIMIT-1"></a><a class="docs-heading-anchor-permalink" href="#TIME_LIMIT" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.TIME_LIMIT" href="#JuMP.TIME_LIMIT"><code>JuMP.TIME_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">TIME_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#JuMP.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>TIME_LIMIT</code>: The algorithm stopped after a user-specified computation time.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="UNKNOWN_RESULT_STATUS"><a class="docs-heading-anchor" href="#UNKNOWN_RESULT_STATUS"><code>UNKNOWN_RESULT_STATUS</code></a><a id="UNKNOWN_RESULT_STATUS-1"></a><a class="docs-heading-anchor-permalink" href="#UNKNOWN_RESULT_STATUS" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.UNKNOWN_RESULT_STATUS" href="#JuMP.UNKNOWN_RESULT_STATUS"><code>JuMP.UNKNOWN_RESULT_STATUS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">UNKNOWN_RESULT_STATUS::ResultStatusCode</code></pre><p>An instance of the <a href="#JuMP.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>UNKNOWN_RESULT_STATUS</code>: the result vector contains a solution with an unknown interpretation.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L1124-L1130">source</a></section></article><h2 id="op_and"><a class="docs-heading-anchor" href="#op_and"><code>op_and</code></a><a id="op_and-1"></a><a class="docs-heading-anchor-permalink" href="#op_and" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_and" href="#JuMP.op_and"><code>JuMP.op_and</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">op_and(x, y)</code></pre><p>A function that falls back to <code>x &amp; y</code>, but when called with JuMP variables or expressions, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_and(true, false)
+false
+
+julia&gt; op_and(true, x)
+true &amp;&amp; x</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L515-L534">source</a></section></article><h2 id="op_equal_to"><a class="docs-heading-anchor" href="#op_equal_to"><code>op_equal_to</code></a><a id="op_equal_to-1"></a><a class="docs-heading-anchor-permalink" href="#op_equal_to" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_equal_to" href="#JuMP.op_equal_to"><code>JuMP.op_equal_to</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">op_equal_to(x, y)</code></pre><p>A function that falls back to <code>x == y</code>, but when called with JuMP variables or expressions, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_equal_to(2, 2)
+true
+
+julia&gt; op_equal_to(x, 2)
+x == 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L651-L670">source</a></section></article><h2 id="op_greater_than_or_equal_to"><a class="docs-heading-anchor" href="#op_greater_than_or_equal_to"><code>op_greater_than_or_equal_to</code></a><a id="op_greater_than_or_equal_to-1"></a><a class="docs-heading-anchor-permalink" href="#op_greater_than_or_equal_to" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_greater_than_or_equal_to" href="#JuMP.op_greater_than_or_equal_to"><code>JuMP.op_greater_than_or_equal_to</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">op_greater_than_or_equal_to(x, y)</code></pre><p>A function that falls back to <code>x &gt;= y</code>, but when called with JuMP variables or expressions, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_greater_than_or_equal_to(2, 2)
+true
+
+julia&gt; op_greater_than_or_equal_to(x, 2)
+x &gt;= 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L629-L648">source</a></section></article><h2 id="op_less_than_or_equal_to"><a class="docs-heading-anchor" href="#op_less_than_or_equal_to"><code>op_less_than_or_equal_to</code></a><a id="op_less_than_or_equal_to-1"></a><a class="docs-heading-anchor-permalink" href="#op_less_than_or_equal_to" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_less_than_or_equal_to" href="#JuMP.op_less_than_or_equal_to"><code>JuMP.op_less_than_or_equal_to</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">op_less_than_or_equal_to(x, y)</code></pre><p>A function that falls back to <code>x &lt;= y</code>, but when called with JuMP variables or expressions, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_less_than_or_equal_to(2, 2)
+true
+
+julia&gt; op_less_than_or_equal_to(x, 2)
+x &lt;= 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L607-L626">source</a></section></article><h2 id="op_or"><a class="docs-heading-anchor" href="#op_or"><code>op_or</code></a><a id="op_or-1"></a><a class="docs-heading-anchor-permalink" href="#op_or" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_or" href="#JuMP.op_or"><code>JuMP.op_or</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">op_or(x, y)</code></pre><p>A function that falls back to <code>x | y</code>, but when called with JuMP variables or expressions, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_or(true, false)
+true
+
+julia&gt; op_or(true, x)
+true || x</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L539-L558">source</a></section></article><h2 id="op_strictly_greater_than"><a class="docs-heading-anchor" href="#op_strictly_greater_than"><code>op_strictly_greater_than</code></a><a id="op_strictly_greater_than-1"></a><a class="docs-heading-anchor-permalink" href="#op_strictly_greater_than" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_strictly_greater_than" href="#JuMP.op_strictly_greater_than"><code>JuMP.op_strictly_greater_than</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">op_strictly_greater_than(x, y)</code></pre><p>A function that falls back to <code>x &gt; y</code>, but when called with JuMP variables or expressions, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_strictly_greater_than(1, 2)
+false
+
+julia&gt; op_strictly_greater_than(x, 2)
+x &gt; 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L585-L604">source</a></section></article><h2 id="op_strictly_less_than"><a class="docs-heading-anchor" href="#op_strictly_less_than"><code>op_strictly_less_than</code></a><a id="op_strictly_less_than-1"></a><a class="docs-heading-anchor-permalink" href="#op_strictly_less_than" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="JuMP.op_strictly_less_than" href="#JuMP.op_strictly_less_than"><code>JuMP.op_strictly_less_than</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">op_strictly_less_than(x, y)</code></pre><p>A function that falls back to <code>x &lt; y</code>, but when called with JuMP variables or expressions, returns a <a href="#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; op_strictly_less_than(1, 2)
+true
+
+julia&gt; op_strictly_less_than(x, 2)
+x &lt; 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/macros.jl#L563-L582">source</a></section></article><h2 id="Base.empty!(::GenericModel)"><a class="docs-heading-anchor" href="#Base.empty!(::GenericModel)"><code>Base.empty!(::GenericModel)</code></a><a id="Base.empty!(::GenericModel)-1"></a><a class="docs-heading-anchor-permalink" href="#Base.empty!(::GenericModel)" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="Base.empty!-Tuple{GenericModel}" href="#Base.empty!-Tuple{GenericModel}"><code>Base.empty!</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">empty!(model::GenericModel)::GenericModel</code></pre><p>Empty the model, that is, remove all variables, constraints and model attributes but not optimizer attributes. Always return the argument.</p><p>Note: removes extensions data.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L800-L807">source</a></section></article><h2 id="Base.isempty(::GenericModel)"><a class="docs-heading-anchor" href="#Base.isempty(::GenericModel)"><code>Base.isempty(::GenericModel)</code></a><a id="Base.isempty(::GenericModel)-1"></a><a class="docs-heading-anchor-permalink" href="#Base.isempty(::GenericModel)" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="Base.isempty-Tuple{GenericModel}" href="#Base.isempty-Tuple{GenericModel}"><code>Base.isempty</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">isempty(model::GenericModel)</code></pre><p>Verifies whether the model is empty, that is, whether the MOI backend is empty and whether the model is in the same state as at its creation apart from optimizer attributes.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/JuMP.jl#L827-L833">source</a></section></article><h2 id="Base.copy(::AbstractModel)"><a class="docs-heading-anchor" href="#Base.copy(::AbstractModel)"><code>Base.copy(::AbstractModel)</code></a><a id="Base.copy(::AbstractModel)-1"></a><a class="docs-heading-anchor-permalink" href="#Base.copy(::AbstractModel)" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="Base.copy-Tuple{AbstractModel}" href="#Base.copy-Tuple{AbstractModel}"><code>Base.copy</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">copy(model::AbstractModel)</code></pre><p>Return a copy of the model <code>model</code>. It is similar to <a href="#JuMP.copy_model"><code>copy_model</code></a> except that it does not return the mapping between the references of <code>model</code> and its copy.</p><p><strong>Note</strong></p><p>Model copy is not supported in <code>DIRECT</code> mode, i.e. when a model is constructed using the <a href="#JuMP.direct_model"><code>direct_model</code></a> constructor instead of the <a href="#JuMP.Model"><code>Model</code></a> constructor. Moreover, independently on whether an optimizer was provided at model construction, the new model will have no optimizer, i.e., an optimizer will have to be provided to the new model in the <a href="#JuMP.optimize!"><code>optimize!</code></a> call.</p><p><strong>Example</strong></p><p>In the following example, a model <code>model</code> is constructed with a variable <code>x</code> and a constraint <code>cref</code>. It is then copied into a model <code>new_model</code> with the new references assigned to <code>x_new</code> and <code>cref_new</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, cref, x == 2)
+cref : x = 2
+
+julia&gt; new_model = copy(model);
+
+julia&gt; x_new = model[:x]
+x
+
+julia&gt; cref_new = model[:cref]
+cref : x = 2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/copy.jl#L203-L241">source</a></section></article><h2 id="Base.write(::IO,-::GenericModel;-::MOI.FileFormats.FileFormat)"><a class="docs-heading-anchor" href="#Base.write(::IO,-::GenericModel;-::MOI.FileFormats.FileFormat)"><code>Base.write(::IO, ::GenericModel; ::MOI.FileFormats.FileFormat)</code></a><a id="Base.write(::IO,-::GenericModel;-::MOI.FileFormats.FileFormat)-1"></a><a class="docs-heading-anchor-permalink" href="#Base.write(::IO,-::GenericModel;-::MOI.FileFormats.FileFormat)" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="Base.write-Tuple{IO, GenericModel}" href="#Base.write-Tuple{IO, GenericModel}"><code>Base.write</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">Base.write(
+    io::IO,
+    model::GenericModel;
+    format::MOI.FileFormats.FileFormat = MOI.FileFormats.FORMAT_MOF,
+    kwargs...,
+)</code></pre><p>Write the JuMP model <code>model</code> to <code>io</code> in the format <code>format</code>.</p><p>Other <code>kwargs</code> are passed to the <code>Model</code> constructor of the chosen format.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/file_formats.jl#L63-L74">source</a></section></article><h2 id="MOI.Utilities.reset_optimizer(::GenericModel)"><a class="docs-heading-anchor" href="#MOI.Utilities.reset_optimizer(::GenericModel)"><code>MOI.Utilities.reset_optimizer(::GenericModel)</code></a><a id="MOI.Utilities.reset_optimizer(::GenericModel)-1"></a><a class="docs-heading-anchor-permalink" href="#MOI.Utilities.reset_optimizer(::GenericModel)" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.reset_optimizer-Tuple{GenericModel}" href="#MathOptInterface.Utilities.reset_optimizer-Tuple{GenericModel}"><code>MathOptInterface.Utilities.reset_optimizer</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOIU.reset_optimizer(model::GenericModel)</code></pre><p>Call <code>MOIU.reset_optimizer</code> on the backend of <code>model</code>.</p><p>Cannot be called in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L285-L291">source</a></section></article><h2 id="MOI.Utilities.drop_optimizer(::GenericModel)"><a class="docs-heading-anchor" href="#MOI.Utilities.drop_optimizer(::GenericModel)"><code>MOI.Utilities.drop_optimizer(::GenericModel)</code></a><a id="MOI.Utilities.drop_optimizer(::GenericModel)-1"></a><a class="docs-heading-anchor-permalink" href="#MOI.Utilities.drop_optimizer(::GenericModel)" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.drop_optimizer-Tuple{GenericModel}" href="#MathOptInterface.Utilities.drop_optimizer-Tuple{GenericModel}"><code>MathOptInterface.Utilities.drop_optimizer</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOIU.drop_optimizer(model::GenericModel)</code></pre><p>Call <code>MOIU.drop_optimizer</code> on the backend of <code>model</code>.</p><p>Cannot be called in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L300-L306">source</a></section></article><h2 id="MOI.Utilities.attach_optimizer(::GenericModel)"><a class="docs-heading-anchor" href="#MOI.Utilities.attach_optimizer(::GenericModel)"><code>MOI.Utilities.attach_optimizer(::GenericModel)</code></a><a id="MOI.Utilities.attach_optimizer(::GenericModel)-1"></a><a class="docs-heading-anchor-permalink" href="#MOI.Utilities.attach_optimizer(::GenericModel)" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.attach_optimizer-Tuple{GenericModel}" href="#MathOptInterface.Utilities.attach_optimizer-Tuple{GenericModel}"><code>MathOptInterface.Utilities.attach_optimizer</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOIU.attach_optimizer(model::GenericModel)</code></pre><p>Call <code>MOIU.attach_optimizer</code> on the backend of <code>model</code>.</p><p>Cannot be called in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L313-L319">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../manual/nlp/">« Nonlinear Modeling (Legacy)</a><a class="docs-footer-nextpage" href="../JuMP.Containers/">JuMP.Containers »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/assets/case9mod.png b/previews/PR3525/assets/case9mod.png
new file mode 100644
index 0000000000000000000000000000000000000000..a1ad127bb155e7962d476c247412f0d46a953854
GIT binary patch
literal 234660
zcmd>mbyU@97cL+mA>F8SNFyLhN_Qv?QYwvvfTW~^fC7pjDJU({90W<R=!QchN=PH6
z^xX$_W<HlQ>;8Gyy0ezY<C$N)x%ac5{p>gVl7=Dy9xWak8XAGJlAIPA8V)7=p~u01
zU#2&)Lg6oTS1m;uw8D;4bMTKCO9N#qb#*jO_>O~yjgF0W40Q<_T#rur^E*0xr$xg$
z`W+2T5B=m{-%Zia{C$0xEgI&}YpmdJH1z1hWHfYL-CuvB9ZN_5`-$lozploiOh5M5
zcY4&FcO?0~!e6+~N`|gzXuQm*Kj@xW7v0g&Fz{`44BQOV)kLp5Iq+V!aJpv6>*?T(
zx(iL*Qxv{BSh`(h@N~H0=ql<d!E|(mD11l#%*VuVbcvh21e1aKB?eh17fS|VUVdJF
zCP_R71_p5#3oB7AIfb9M!@ne$Y~0+OMfvzVJUn<k1bLlYtohE0h=}m<3-AdD@W2&3
zu3nCAS3P+gU73G8<kxfLEM2d=*gCt}Iyo|+o_qD0(=9g%CMML2{`&7nJKb!p{(h6A
z>(6dM2l-I{hwm&eKi^-^hFir^zlv&Ex;ovsg?hZMqph2yfcVkHzyJLA+kV`o>0)aM
z?M2-tdG_Zm|NQOu>t&rBoLwwkT~XJc`+5C8fBXIVOEyk!PB7Lkw%3&%-7H<;=AUo-
z)xdwh;^(u(`B204kKy|<H%Gt1<VfO)^Zhl^l6ZZ1^Jmb|q|lV*q;)*e7stFl>h>NS
zepMAz&lKSxP`Ug_#qzFFZcN0(l1e!%+{c;9a-rBP$`(y(xQr22tg`YP>PbqR1Yf@G
z3{On#6jpyr>>RQloZa63JdQw``YhGbEu=(Ct{{*`yGtW4SIqVp0?;sU{_W$Cr})sH
z-X-`yF2UK0Mw7+4O-L+-hK~8?4~Ag4g7H$5%)efNh8{}%{lveu0-7g<16RnT$AtVx
zOHMGZ()@cP&;p4uPA~>bG2;JQGof|VG9Ly1Ydjd?3Y<V0%6}g*9RA=2>;LE{rVLy`
zj1j{9FJpze)f^|V^FR8D6AV`{V&WP8>$Kr8G!SE?{YO8EaTptjWpKz#|2$zw?GIqY
zlsxtyuOOBQU<@WcW5M_z0|pBN`#)as|IEU`lAU@GBGrZEw=<XHRKLIVf!^P^<XXsq
z$81Mx9H;Jctt^!{p#+_x3om3s@Vw9|`Ao&V)~2?WMvw(pkUsIJ{y3SpNg3j&$+-L9
zGr?GMt31NJ8xV@erk=F1(W~hXFPco!DKxI4lkgt({Wc~aL4MXUp3ls0*fz&uuz2#Z
znP1i6!LGU_?ted<wLwPV=FQ6&^s)4cOnv)x^>tque0`|Od-Fd1Bc3Z2mopSBTOQpU
zuDJR6z4RSygiZIg{^z=EIqJ#6v=_T3yA|nE^D)KG|6x5#Md1*)wmgVxiTp6?G*Rd4
z)qROJu?WL|e-p_ZE$-4O@$JS$-59k^sPs$6F`<-8MWzyhp6T-UXW9}4oj*1d)gKtK
zA_}`3L+~Rhd4`|s=2yASiX!a>2A_t}+08Qs{%L_r;bGmi%p#6`u7w&h73w?|RV*Qv
z^ycHVxcr?tBFFX3d~X$*)|Q(;Xf7wa^3rjs>**aq?WYbS6}v0-ht(@}3*@<a1xdmh
z#rUxbf7*l$x)^UuwOExUILKiOg(T?2nPA9<o~T;y5QD*XS~o})^B7RTc^64_UWAdX
z+I?xhC#&^}Q|+{9z9IhS_xD{bUGDy|;cM{Rhvze~FJcbjzb4Y1Zz~Q6B`B<3epTnY
zoAmWg5Eee+N#Vt;$EFdYe(8$HPx2Xx(Y(f$#rZVsscsLs!eswA%MZyJ;^SBt9`ABC
z5C;|2=sZD3Z)fbfaR-*?=xFur?KOJ8@0051$1Gf)v6X*a!=f_&$oZZ)2tW8wj|*gh
z!DYp5n)bslj(ITA*q5dkI9BcP(ze%XXJc-rFv&f=J!YXV@4Cx1HEyHw+^cUwvy0s?
z516*h#9;nmfuLb1;}N&2kjAw3GBj`p3o*DxZs2Y#kB!h>?7IKq9!+e+_dwnoUFp(e
zuWq$x2x7gy6!RzRtQ3w`^5PM{R(KJZ=kG15)3y}UV(yEvnv%8?T4GE#*Db`@3*9#7
z`_;GMx1mhsKYEFnOn3?ldX=~YJSFo(V)b`ke*c3#`@*XCb?Uv@>WGliR5GVmMhLqZ
z(oo_PSby48e+`f=87vNmzR@Pdz|&&xZi|DLBR?eikn<RAzyxyY<i<2BGu#}!8YZHz
zo2QkfaGw^puYvWCma-D_u4R?fg*GejUZqtYo)rdIp+xj`cKw{y8@l=?me^>EdzC#|
zDse6?Wh64jj^uw@wRah$xVbP0b4u00;NDmudBW}(*q(fz=(K2G-D1Eo!Y?B)i&A3^
zjE*&SguJ-+h?+tFPbT?SQ(RC@sk!4>^9AGBal+9Ghh!6tWTr=aX5$^o+y!P|Izmn`
zvS2sq{;5B_=+_K+i3!V$p&{HW40L=E0Rw!r=j<Mt<QkM*^WE7faV6CdD<if;7Mhe<
zwFh%e%KUL%pMvMyCUwzBH}8-*OLQ-QXLay;^icJg1)Y%!2O^Ov;%c?Vx$#;bDh{ow
zz4cDbN^D-aKW*m*Fi`^JEH3iq^}f4X(=BCdDL9^3U~6WEmAo2*abe}H$Fm&&lOqgl
zgT>qO;-ZJ96m$l&tCHmk=1hCC$LP%uLz>!^&XZsv<G2lbI#}-ial^ld&16}F?-_sz
zJDw3;h_{(CP&t;ZiZ1Kc!oWTHO0^4rYH<@b^n9obUtG2Z*6CX=g>7!q)q`QMW3rSI
z^wB|odR4<A6+`^yA?x=;V^Ua`kG&(U5m*BSS6<%Ub`evYvuaPOa{1%Sq!>PfRT3D-
z5|_FLZ)2HxubW6^P-b<1fiZ%D%YEvB;wu4>Keh7DH&(kBEHKgv*<30%m)2(s`_uZs
zu`muW(0gAUa9l|(z|;Pkqj8)U`bv{ShQj@8{x6`(8YvgPMqfNQV_M_c{Y?Jcou`+c
z%AicQ-{IcgW@qOz*rTOkgmouf|7hjS&x2e!EPpgS6=)^gSMInG;_?T_$fmx_ICvcJ
z4e4l{Pkn1~@M}i2<W3Ro8+vfALq653??mqWaa8?rJ@2T#Ej_@$d*=p)Ib+6+C7LH(
zNm7%IAtv%Iaa^-4;m(6KUhCleADy|(v^K1m^!O6&_Ks9B#oP$(KlWaZk|CaAD6Gto
z3%u8WI{^`;yg2}fLeiTAjH4bCzWMq^;}s4gu&_7luYZWT)t`-ZpU(RWV}V;{$owPc
z)M?^BuGB$TYVRE7n_r-&KmXO8uyEOrF|vd=<9JP8m2<Jbmu;7?Tr6!R6#c?by_c8s
z#QXF8^WGoAmIN|M{$z0P%EKH19BGz8-;!x+)M#E(Ogel0{oeLExWE<9<ja{#Buh7p
zYrQu=M~P|do>Z(I-L*%G#h*lwYMTRis6W`r7t#4N>e`7QQfB@G_`}qM7p(3)qRi4`
zFbgHF+nML$(d#X>Y<;xq06TKL#%p1qV5HPClCm?;q}m;RqBzh{o)$3)f3{j+gxFc{
zycla6V!G$lm)qUUo_2eo;N>eT?7@ET^C?3$US%c>g&Tk|Or}u)pk<OjGl0_DsMMkv
zUVuk_#(Cr7#p9=(rJ5dI7OdZ%Iq93DabXg(fQh1N6<o38;okcFljqrld*0|@5a6@^
zZGZ%L!O}{YpguYbZ#{;d+Aoh*wLa!M+@4MIKivNYriHUDH|-&-D%*C-fqcqdzDPD3
z7b(^EG=L@uSj^rfCK))&!dpu(CnP`L6Vyq@*H0@BJa&AZ=R3<n=Vq(N%DC^|*AG6v
z#>T@L24j<Za)a%a*WN`q)qGXB5G&!cb;|Q&pnLn9Fd`!T3wD0dbP|it%R^lHRM%<(
zup{0_)0^e%@r<b3-YQh+Grj%wou}`{=1r^kQ-A|J#=T~dX8ubB(pR+AQbiH^MW#E`
z%cJ^-&(<g-W1zp_LVm;IrE)M+!=gd(ky|#YfOp;m9(kPpb}?7V^eOk{QA4pDwymH2
z+=@To+Il8pq*cU8#igf~cC2=!@V+aV<GXJY{!aDrg$#43_;TzP@|@H7=<OLoiH%FO
zQK_puL2i?UCe_fsFfvW??vEG3!6z#?McF4Z_ZnoK^N}+9NF@0^9mZvRvJOna1i3rw
zaJlVur2l}Sbrs%Qzg(<Gi_dv0KL!mJwZD0|W0d&%*nJuyo33>0&L>nRZV%dX-%R-J
z^0=o!|355aWSuL^F|GBse&RI)b5`_}<NiXW6}yn*=!Gr5b7qr+yFm|s+q*O<kVSI)
znjy_q%@K9~?div0hPWc8<<eB*IJ2B%^`yEr0SRtAD~m1q7Wdg#aj5=otpB+Wt#qr5
zm{=R_5!*(mSxrxCn|3>tJ&V%g?aBY@PJDZDuw=$n50-OfWA+SXpxOyK@j*g7dv3+p
zJ4`Qi1m!n;D~ftjLPPMWLTgltUX8GBvLdg{GpZ_d^r3~CME5W&M%}xuEScsz8A8=|
z{9tFH2r0*z@id_Bec%Z&`RA^u?YOq`yiCBqu58nJf<Q9yilW9UMbzRO!~gZTf!rj}
zy`d=hF!WDbt}}m)u|AwXj_<#OK$gD>_btj&eJpH0(DpWC+<Sp%_eOBJKMDJxT~t?!
zb54+WFu%{%4MD$j%?AI2^)zR^2kYm|%_&E^pBor$k6k&a+g%o<FNL_m3xCR5V|Qzb
zAd-}}H}ajdV3TpR-WOI6B-KcsqW#<%C5H*7@tZO^bVyEx=loNgq9x+DKemdmH8B%<
z&h=z%;-9zhDXQDC>*Y8<>MYwNFs|khaJ4WZr7IgBJaU*M`?!kC%qj(8vkcq+Is~Z-
zTyW!_(Wnoq%supM&9r6i%NZOvcq5!C)N7PR=ihNV>tw4X#*mmJLymUrP=7Er(J`^$
z^mCt$eO_rK3-c(DS0}vpYwr<_<I+Ev%Sl@-W2^Pubzf)XGOMeRP<;KU`DL&>+vP_?
zQCJr@x@6BX)ojlsZ-uC6J`oq0LUc;*zA({pmAmG@G?urz|83%>{UxJ%wlA+^hUxwH
zN*cd0J-_rcXe$NRP$6;Hc67V{#Y0YdN78WLs%(LTJd48RVj|sa)yKTUw}$G3U6P8N
zE$BO#{@1KppfK`14@}gDbPK&Xm{1KH{m93`k-SOClB*5J*iK~i3#j=u1s_cw;ZTf$
zz@f)Bdj?rRu&`GB)I(KK_r=^XLv_~(a*k(+zV@s4a0yPJb_|}f*fhfYfx8%vSAk@R
zxAwRQ;_4ghmBtH<MR@Kr&*US?sVYVWpX+k%^2-?6k5+fsrMc-xx#sd*61_EVp<&x_
zNi5Aq<a8{VNC{XmuENSjT?Kp7s)11til)vtn+r($NSuA7&PasKL7=b%;cU8ZF@i6|
zRc{&=%gMsb=0k{u&)vw7<&n1*SPEuCvt{ajIpLSzpZ};rQ(kGEA}Zhw$DbB7>uG6j
zE>%$`-RP&4F~Z`PZ1h_*?e7*sB>2F-;@CqozirFl#vcY+#`W0~ybj_<@DJoyYUZEk
zlTXuM^sC9AR!<T-TO?ntA%8)~*^p5+CqU9UC(X}0xH0UsdxFUBkab$3+c=x#_7qw5
z(y-d6FB#Dlu1uNjZA97%*F)KsRPNCTW&`+VTd(~#-rLZbbk1H)qcV6#)aL7{rSfG?
z=e1_`YL^R-r0V_lg~YRig(JEGQVz;Quf4_L;in+|u>1hW*SheQBoFQUE|{uK6FsE(
z_Ec<A#j}+)MS9<j&(Rm{W~2`f_M<O+!f{pR#;t2CIdd8Ne<1<`GYaB0(oJT+EoJ&R
z^@*e4POZzdio`acpy6!wlLfpP&aspNc*uAFG6~vvw-d|HPnx}Lu1~5hxGRxkz9?T-
zk=YozLu?aiw;t>rVvKI@Q10MDii20byIjpzzgL<!xm>*>JeP~>%ajq1dm9~DGzNzJ
zY<!&EDasYJfb|ff@a(VHJjJf>X{-~zK7lo*cyo5dv06R(dG$b<i(SG+|AXt%64|Te
z+4E-<>o1k7=t2meQU*B1uIc2oYqn$$1lF+)@>C|76P7Ap9v96_=V~sc5}kM;uI>$?
z+wix+@r1W20=S#$Jpju471|L8+h=n~!Q5#iMB)GD!f#Um>B8mI7)REslE914r(<Do
z4BNXnRzp;oFO}=8c|!oaSY=_rX0{`S=miZLn^_5Wzd2>$`E8Z&R5#a(tai7_l#XAe
z44R2c_<F4N-3tjHL~8}t3z}V0QS9z<MnkTVQ@8pq&*(G-f68Q=psVFi9Eu%8*ds@r
z>V3g?R9D<po+PuoOpDaHSE+E5{F8^oz4!>1>A|ohZ-MLYPwXr3C+^%ze>r21Ec6|+
zd^~6A?%>JqI@1<T#%_lsS6A+?VR3kt{`C$itBUx+_s<Zc^+mOGh?6>4Uat4^)z<7_
zy*)q5C>v_i7+7ak>Y^4UwtRurxb{Zn1~pqjFwIzrpt=#6aOzW&>f`N0!DQ$Mc<eVC
zLw@slL#U8)+;o*=^y)=yz|zdAwrQgJR~rC-=QDmAb&W2&R#nzg+qf}lWZfx#iMXv=
zCBpUT595NIqF%qlk3YFs);X}-ofO<td{vewF5_K6-*@|>+5zI+*3V-&lXr+D6C`{p
z@?O`ZIl9z@gO^Ea13%rKS;qb}6?cQx3wR_%UuV>5u?_Q*A`W^dE)-6D7qXQ9iB=}z
zoPE4+LC>xEzMIpPgy&Y|oJutl_@%Y++E}n1o-Eght#H1O_0<r<$eZ3o94jM4)YVfq
z(u%s-U}UI|xAdlcwx4a$CL`?21+0;6*Z*8&L5MWtEg*2J(h}O|i1khHknE>}BJRUC
zf@2zi6ZQ*n5bF7Dh^n|?h~C$&pBtU@#zuv`^+fy2qu{Tf;~FJ5y=B>qs&r{wdih)%
zmZ)GsA%M}>L5xhZKU|Cu=cF9Y-Qb0mz7V_vyF-iJ<%sCcM8OYJ7xFU_-WZT>C1`&A
z^iIs}3-LB5fgOjEU5iD(Rn0J#K3A*#loW|OQu58ivz;l!!S*wct@aa^vaeS?a98r6
zupcVT;`-ooW(y#SJrWP0-^&fY?ON#BzVHrX7FMH|js=H0_Vp8Z6r6)`LiPh{)tDn9
zLuneCv*Q3?cY^KDKM-4~=}ox6^}%+GG4?vUu~@`oo+|^A3o2r1lmzo{#rkHSe~>)U
zqd(|KW&4}qMT2NNVl;dX*Q4@Dpg}zT+nhJ$w+hQTY6fV}-ZWy@FnHxQ=d4~9M%?Hy
zH-BVf-XP9z?J=BS^t^(ya3duyf&)`LGy>MZH*KwPN1O|bj7@5N%x$;^xHa(&kgSPr
z*`!9J%5SiDy4~=s-}$e^p0QbOS5uB!Ko*Wmoi{_~Zrb;BLG-$K&FI~i6xWUtu*ej1
zbsy8{5#FuwJj3V)CyVkG-Ez+!Bd~gVz(TV8a7cS@(bzXGWxu&oe7#M;vLzu@>|}4)
zp6~OOpn*66*0{Uk6YsXmZ7efmPU;VCzD*!}uYckFd#w|{8Co<jw0+@oCp<ha3cf9^
zVdH31_Ma8S@3$$u(~Bpc?^rcWnRTVyhNRTQsM-=@8n6&q0(+4cw)M+wFAId56Ea@7
zQbzmOGmMzdYl_6RfYdvGO!7pcyLw=Fjq8mjTCKP|6U=eX&jj~R-Cl1$e|cfUmNJMk
zy0=GYYwYuU5O}8HIz3wl9rNQwCf09G+vQ4~e(<^u=bJh1taYYqe;n<dZb6UvzGgoc
z2ezlO1$0){qP7w_uHmk^Bn;MRxK7H`HlJf2IR(e`noY*2Sp0@CFs}p{P?m7}Hilbf
zt=bpH-4){eh}l~nIp2PM;sy2Naq$7tFN`Za_5S{`MOgiOb>FYK+K~QO>*BYMdN=qc
z^s8fVy2FV@T{oe_bnwKN!%bvyAYhF2R>s4(p;qr@+sN0ubG(Bk)qoSpwJo`HGNtX)
z%_Jdyt2XN0ZNR@3?Cu(-S%TwL;QbynoO`F0f01UJ=;GdLBwt-u(HM}lofb)NT{|VL
zdsM#Yy9X!I?`$r$J-m5;*^cjYfka#b&Lv*h%DlgCWv?@^mE$$o<0&Gw-nUwa$4}$f
zWQ>w9D?lVhKl{X|FgD|{F=EIXLf#s=;R^$%6lT>~-sfdj`94jVwU5WM#gjEKVuqYv
z3aA^H4}W#yfy1}?{36lnTVjdB*0S_`)lyC6`;6Lw8z&ggyx)N!+J5k3!OQWU2nj~|
zwzxc{LerDT9yOy!3fjb56aN1CfsalN9}a1U7>S*kaydWlGpB#A0@39$lG20Rgnfe4
zYi48DBpvFl>R_Yq&(}Y1hIy-Z%wwV;>dR1>_4eAdiNv#_PJF%^bL%HJ-_vY8<yxw-
zei}6(F0gcMtxd4gW{zjcGvnRs^0A<Un_an=nSCRjUd?4wKNe|uT2vev^cy)GxCa3c
z#nXv!OZkT8=h{PLtu6-TETIEK19eY^i?v@jL&{>*Z+GdOUEk_6j-ZtnH>XlWH~X|l
z5cab?Ge5C}5;uE*ERLXchnukXExAudGK!BYpQYubx*O>XP52-Bg0*u$TkGyXpDnyJ
zRwQn>&#D@)BP(<xorZ#fOYfeMWHUw~BWA4$-6X5quIbJzja1S8sWUIcZ+?=qe(4rV
zvnWUDUMMK4i?GvWMx`L^@q%VpN8Lvp*m_x>c+Ybbs9O+n=|1nv(WH0%qJAMQ{(6*%
z!kw!-x^?bLMmhHs4mQ?-R<Q3$^n>1HdI{4uuMT`{X+9__UI_f118Z^xl5U&`IHn?H
z_Zf?;+^tfleHKf!v(?UqF_O_<`~q1AL*hCJ?!1#lER+wKG!FVtl`@7~*x^11*suH)
z=lwxs3)`;u1Tu1aSWtL=I%)n|FIZXOc~)HVuCn{5MAiWHdI%pYS)t}FkhtMgyQzn}
zI8??SD8}X<Zh9Hn^kiSjsEx}`*M8td39`9++I!mg^5<MC><lSk4@bzw^q;1sAVXMD
zqk_sgV+XR8cSz?X`+F*pv|f>kDZmX^hdvgXmEtl46H;+=Ae%PH`nuJ{FRoeVO4|^O
zWJ7xtgxd|jn*e|Wt#2RPwVZB2GA3U)EIG-Un@`c#{y}r)ESC<&n#*7Pw8VC%4BE%5
z8NW6_dS=Fz_}tePuA(z`wh<6M6+$pL!*AS(xK%u!L#QthxM#4%^)g>~YX6b8MtAVB
zeYwFm19dD2TlJ=5Ga3D961E^EpMe*K-qprMh4x*Y;^BH%diIVvm|SzgNNfu4=r7f$
z75)wgwA}Um&gMdDhXZuKsolD~Eu|%)Llgi~^{oXf!dhI1^F<TJ8h*ggVV|IOZa;?{
zeB+rUk<qCDRJ5+b%Ym2#`?nps)|7rBCYnu*{~JW2VW5cZx24AP;V&fG!VUB;NT?@t
zJ)@1_D=@7!YPGRH>rk6-hX)*mi4)>^KywM%56kCwz|t#yQf3f1bm3v)Gbr<?eyts$
z^5xs(1q$2U7s}P4#I)xfZ)mV4QVZF{=q*NMJ%lJh^bNMrIi+LlCB~n$X?8FAbnv{t
zf71GRNL(K43c_Ix{;}z!cxZnye8$-(^;CgPrf>1m-p4H0EZ0re9)_4#{fRwUepSn_
z7D7+Ze45w3)%B%2Bh7cquvdGSMDDZHgL>&S|AQ)RQ+jjt3piC8mvb~y9f4l*HP_Kv
zFt+V`jrdwxl|eeRMNyG<U+DLdFg+THnzV?;w&pm4czsR$a+}M{WkwJ31GXajubq6L
zU-s?DEEZa^xl)%i23vrSI{6jepIG4aDV%*cjInok7dc*QXbH&{ic$-V#E5u>N|o2f
z5>EwPE&-8(!`kcBA{M;5`i4l+7meVnt$r-nvenX7UM{^C27PPjXJWWgucS)_0xu-B
z+qnnV4T~F&O^wMIj-B<}^HgWx-**8z$sUWx=Qx&Yi-yf4^J%-%AYqxVO#0nN%J!=F
z;(4yvU8iB64LnYVb3QmG#O!J>9_erR4ks98_8zjPCcZ7b*4F`jeJPUURqQsW<y7ZV
zShLo`{=}1MiX(RJ$ilV1$zdTWXT+4}ad?cQ`epE4$_50-LTT?hqAhWxGYE|s#*MRz
z{UFuZc2xkEXY~_p3O{0VaY$S2xQ)ManOcoV+)(h7p>yn?A&exfDvdSO;4`iH&Um<Y
z>gHP#*s(}bwM5_VZsQU!MfIg``^GWdPA*ps%VtYp4cA$lerz@n{4h5Yh0F)!dUz+s
zMF_SKTj?S83og|L6}}-$*37@BDWsHOCK(NT;+46%6y`@GEURdBpUPnIWYt}Jhlh~y
zA>uYHmGfbP)UL_9C%4y><xz>O=~r(hH;$%vYn?rfaF{qq;s*7K0#4zBazIWt_>Xhn
z0^BWY{o*^XQ!7^_@8`F_sbB39J7D!GqMl1GU*3v|kNi5XvZf5zT@m}~<PlrOY+%fj
zWl5u?c-<YF>Ux)lhMf!rkF^b*bgYyPqrJs1LP8pMupNG>K?LDCtD&vX+7B53t|>bc
zSfIk4Y8d|u!_Weq0f#j<XR@am=-un6qHDF==!~{qH1Qa_1^lOolasxzW9`%n#Jx5{
z{l+7rt~x1ZYVttYG)?!h{l#%Ce98nxEl-KU%?*T3WR)JSI@?`V(G{a2&b3lQ_x;Pg
z5Q^dIqZl@$(@IXu_HU+=9bYNGv6;bIb$*=5K5@|b#Px%-eMjc3SLqgS)nJur8w7t9
zw!QD~(d*=C$?wv9iKY4cJ@93tvj{D{aL&eMv%1?=BIRWNvENK%g=PU-lZaE7aQ7O>
zm!KRP&&1Ny32J^RGK~M}$kp^v;^0D$XZMwYZ|SSih)oV_eGVqmDDDhZa~dkOgzUx4
zqtH{Y+?VVKW7Qr(f&f^V2VC7Jhav{;=I>M#)&g@SC>j582LS<#i4Sg4lS6ZU8Fa(_
zI7d{`L>VwP9#)fIVvE$d%imtY3+q}Ov3-{4o12=wC>z^RQB>^3|Hq9a&&@S;oEcxx
zRU&?x)uAgraLYh=d*-xZ?DfA;q6%pst&B4d9Tl?Tz5l^>&O*PH`n9_%;<x5?mb|Y<
z*mlW0z3rM{Aya+YCio*zDv%_&(&vkTl4OL@c{kChR!+xXvFd!1df<ioD#M;RxbhW)
za^dJJagY_v>;plHP`~S*d~=Ji+ns8d!~zJ|9j9;;ZrwQ`{O#DSt1Duvxh*F32mP#m
z$A^$JylY9)y5`4OD#)I5-EdL~K&P|oe80XK0gd|R(SFFjoTq|mhx6JG%k(6vki)Q$
zJ6r8cjmL_qwkcQ1;G0Wq6LAURUZtrSR3%Gqrh1dqGeWMhsU;Lc79FG)wue(q)gALu
zw5rs?Hxk}oe6C+qSQ3mAd+(vYxITe5?ihkZriM(~m;!s3|D`B$4u9XSx8wLQ{Yau!
z*4aCgZ^G<1i*+0r=}JFXccjd<An7g?o;qzct3SA#GWB~t+#GVw;XwxR+u{~pOjOqm
zA1JrWC~#&MoJSU!t?;b;0&K#H8C<WwcXzyT_cU3l|Il9uyxU0jjltrpeXR(ahZV-c
z`f+D2U;7Ys`f`y<X}}EY2^xV+VAf~qE|U%Oy>3%gT3dd==MPy#W1@@)(M^w?2Pm6n
zI{IDhWHHKKJ=WuWwb7$`?qY-A0d=p@Zaz3q+LI^6Jy)~Ulb@MpjY%$c#?{#{@-@d5
zIX8u|ll|{DXCQ(Z8S-BlvAt$kwRWp%%|K5_0jXMIcKd6{H_{^{vgjcHP~>%DNP3oz
zu|3{Uz4x<hiW7)dA={p}5LO*P+899<B+?J6liXH<#7129a-?FKq;L7dd#3w!lc6WC
z4cVfyS*Ya0jD?M9qr=OwEwHm}x4$(}4KvSjdO^?Px%7G0SW^&KrB$v7=^8frFIQ$Z
zUXVf{sn&!QvPkLGYv#KdM=9;S(XEKUNnjHi^_oo`7brXxSMJ*A(3Q$%cD0)4_bl9+
zAtYTY#~~WgU>jL0<B89r;Tf*bu730ZBTLC8nCFo?!LPvTIq7Sfqb;*SDoXnup^iuk
z^`#b&u<c<HyEvUsJrw8D8!>CgJbli7z~*cgK5wNI{&}y~N*BFC<GlOQ8yJ;b7g@t-
zTdy$(KT>-B@Wz7ie)7yc!H%vJQi%~*+!Fo=w~-75FDH{STp=A`3cz}caF#PWBzmzu
z#zZyHYoWVPpzyxsFbe5sr=;xSh26kP{LPPUGedA?XGYc<q4J0otCTT5D<N*CsN8YP
zh|9dXv4SzKoCUbvktOXdZ$Y++Ba56Irm4$xK^A9=Yc(S0Ln%$^{Ypsr?()y8NJk{t
z#}c?IL*C>WwyydlIa|54ENBAQHB!C<e!sZC(Mx-w8&B}TLbRnDiwgl`=lm7x&gM&r
zo{BeR^pZOh4STAR&Ue-M`V<_`32A)d`6>JXvmzyrk&s=z&3waM5@w+hsm}!NsASm(
z$$O$>v+t<P%3Bv`Iz$(^#Xe-ZZgAarlJ=Y0MXC;R3{3G9#htgYuNy{OxKAfB{8fis
zmGMPxMPa%&Ku){9+{xOOqrUlEd0?R-6-rzlStRjQF1d<|*Z2Z2t^Slr<f>eqvD2tM
zb^u4ltx0uJ=5WX&n)+=|)7A~FE;a-n1L+Of^4VwD4sT4TMNoc5rA2Tw2)a7?LZ9JI
zYJPO=4E>F<SGpxW4hz21ar)l$d=^pJO=ryX`(GILl6OT-C!co`4pZ4!9O{+a$Tsty
zV-;Kf6v+n~v(nV-Jyi!3!UVs|u40H#^fC3av!wak(!JehS@o_c>bzQnmZ+Gh*0%Qa
zK_foj;I_AzaUcwj^6n8s>csW14R`XP#eJG8@#c4*H`+q*C<HVM_l+`m*BqlQZQa#?
z;BA{uRM#mm+5~zM1>a{nQc<}<uagg-Fk;?+?dg<in20ti(Vm)r?k;c5s*JVZ4oKBF
zPG?M_sPD-z`sc+JW|H4O-f7nzb{gG&M!}&a7pnvL&!Vrmt^>)Qkm9a8Jn)Gv%3rSi
zX6pJy1#^aL(T?hORYD+OJj++-+9wHJ&5*<xCEU!hjdH`HD3QU|b+eoAD7K!SFW^fD
z!Pxqn+6xOuwsFZli{*3-V=xO<G}*ND@zWMkMB=Mbv4n|T?2E8T1ameTZ}mOp(Z1ji
z4p7~mCJ_W5wjL`Gv5G`?eh~q-{_dlT$4)Y4rU<!&%*`3ATPFgY`DAM1@W7+eEMt7O
z?-7>z-J5GBvi4Yb<Xv)@sXLwi2N^3jzAcZ{K2APGU|uPFW#C|PhbWrkqYCB<^o!y5
zeo3RGg-h)l!D1=9q^)Z^yGfrtW|N$2D49_Ljj6|a{aQQ~S6}KMQ!@H6(s9=I$YS20
zlhh3l%6opMY_kSZ{WIw0Br<^JP{|;lg8bYOcdpVZpKp#iGrgJ@O{!@=cs{0%oNHeI
zN!)vb3w9;4L=rkgFU*&@ifW$W3KWV6T;_|;3W#sDW)3z;*J0Ba^?<b0w=s`2x=Ki#
z?(Esb?!?4a85a!j3PG0N1YsAqhLBbQf3k#;G?C$JTOQvHDVfE+m^onN=Yc*ymm~D_
zwHICNXBCTnVB&6^`i-Mv_=xH?MhQ{4<q$26=BymIs=YprN@P*n6A7{B2^DVr!_?Uc
z70#YBiPnuFYWqL9H=`C(TG!T1`gcVRQQ8_JtbvsK!1{D3s_V{?WxR2Yvw1+d{-B46
z2WTeQ%7?hKkGb?a953Vb0$c3Ss7W!|h<*%vkccA<QV;TH(mHcovRo-jU+(IBK5kU3
zYE}s1K;01!`x8=k@CyBOvTKQ|-lG{~xeb<Ds_N$Ffd*|m=i7_s?oi>KO$WMkpg4#5
zM^gwAo1TN)V`urz?V+J727u#^y=Vt1`S<nbL`Yg+EfiL77>a%jIOYfXgQjZ{4E;rB
z69*s1fX^l2{d$B=NRna)^HR*3ly0_x#$mX&@SWD;8}PcbA8s6SsT71!tjS<0cV0Za
zP0?lOa6$k8mru=^Db}x6^90KPj^6v|-kXk=*czl(eQ&x5R4l?Hv?~NOkT0!2y*1VS
zs5+xEtC1-1Zj+*mN?rQ){WJ;4I1ZCPbzd5u{{$?Et$QTfabs{ENzr<s1$C<1em^JY
z%U8Zx@=`YWoPBm1kpFgZJS&FmvsJTC0M;eHr%rjZHr)#OQ^UiPg*CBh5@wT4oz1_e
zuEjz4M1edG?$X$pgD-`v<OMbXZSIoMfF+<565?{d&;g%A0DkH;1r@#{4lfgBoqj-m
zlo9Um(-0DGHtKP&idRGm7eH*;74R_L7%6iYVOwZgJG_Yo6lMQN7cEoo!fMzlmnM<f
z6zx}Tw)!+HljXL(K&3BD(L<&O>|vI@&^18TMii}8savXD_sT#h=ChzIpx8=qYk5?p
zS*6j0w**LOAeOJ!QF?m=eT#5;@Jgy)Q7G!!$I=C(+Bb!HvUt06pn}k9Z>(i9CQ5no
zW^u*%gS1I@(Vd<o_tlSgh(Nz<a^(cUY0(X!JMUttn5EDNT3yJ%DqPzcwE2z_1b1JE
z#pX;=#?8LxcN(0~&aiz^|HYgXHDA9C`N*5;ysCAL*&-~zX*`G_^FL(`WB4#Y+4Kol
z33qM8^SC7{ZiDo296VMK1UHVXLZ15a<!ge+mGI30<xR~bJ8i|gP5;P;>61N_&}rih
z^GhkVs&UXeA@p9@E;KODIQi5I!?O2Pmt+kh(&m}Jr1Y0^_U33tmvneWj=8k*%+P2h
zcRkzsDyi1A_HEq9$(YC;Y<B1ib|S&Dde?rA1!R#&g+wm+rfd7z$>ak_J;9cHqr0>(
zPlBAVswLSIxY<-%9j$7i&J+>tJ6b?3^?j26aLV<biC40)!?QxOspl6@)-@I?e)>%{
z#gKM{9p^qIv&F}KUD42ZTY8GMoyR#o57`1br+X#Ht^s#KklJvA{&pKvFJI3A2&j_v
z9h7sM8N5P;m^v&FcEl_bdCGy6^r-|eer@B6V0VoH_Yv=DLlVsp=6)l2k<Bu|EIe#k
z^y(Vj7|Vs{Q<;S>apmUH5ftU!!+x=`hTuis3j_J+m+Y3lC30fq2`UMxVDE<_(-WKu
zNt7{dcM|P4DXT))%ClpqHrw4^N!fOb9PC7h9z4@V*Eb&Qiwz5h6>=*gn>Lp|uFIcg
z5i75+-pEWT13??x>!Pru<>_;2qNU{e2fHlh6^>Zf&3^rb$UMFWb=sN->(@0BgX=s{
zQ4Ge(IO{k1gXKZLff`gg<KgxR5b!GSu_9;WxLa-W1~R<>1|ICM_}k+qf-c=AE)wV}
zJ9LU#@;XojWL$<^6GX##faPz>d;GMoE^FbZt~8z5#{QsnM=uabYXAZCWd>b(wh*y%
zN?EtY5#Nnau$$}#cibiHz*7_nVjhJD_yzu-Y~B?^jx+d12;1?El?k&2LtD|O+EUEQ
zCa>IRE;ysGxvc4@mXwSBb?#OJhmIo^?c4Er0d+P>Ug?2&1=QK-GDy}E`ZbEEv51Gi
zVN!JBk3^GbZcOS*;Q<&``#YOHFgoWv(~<4!sS%Q?-t&2&56T|DO+d9G5b}l<fX}wR
zYh}gTxK3gj15hiu^O)5cpKXV1C^Ao5vGsZgPYW()5NrCo5FupdjiJ)Z$yx8}K>Q-6
zPTSR$BypO1K12h3pA*UFrhWw^YU?0ze|}#y<=mrdHR3L3>lz>UUWoi(5h5`^`T;Sa
zL0C_U1!lS5dw_!FTshgNZL0OKinJ)kHdi8N%-Wlk=8Zy$A7W0q{BI`ipRrO9N3lPD
zvV8vUyZ#@_KL!zy{61X&Z1?X^!2p4|jFecs)t{8ZKLX5{6e!|p-oAP6cgFDVu}7x>
zfnxe;)v|wYjFJYZZKw7sY5&=W|G7Lc01paex}viFqr8QM9h&RnXZ$<o_4gP{Nnt^y
zj?IUj|ERNhh?04WTrB*P=K9C*F{FcT7PI`i?SB;fG^0eQDfIe(QbPV1IpRw&{usk0
zLjPWQ6gUNiMM@pf`M;?d{xLE*xiJ3dO9n~*UIHcMf|B{pe(3*uN`C)_sRQF5;LhFt
z@3mO}pDq_X>~fQT%+N<OpKr@xz862ei~P1Ua{M?3&eSt0FiN0_Sp(cU^Xbic+ux&$
zPCOW0&wi{cQV9Lm%EsYv2CD-KPSGh20TL+&eN0;68{dIqm}%(+o%&z24LWKWnx9s2
z{WaNvhh({zGu{KBI9N#_hxcYTgc6(tP4eU9&JOLr*V$ABnh)cY`{CD(Uk{VQ41`!$
z?m{bQrkZdA!FUV!#gIw7ll==x2a=mp$}m*NDg3q351Cg%tU3UolG`{>Y6|3UgajBk
z7F7+PXgvplmGu~q=Tw7)tc~%XMH4+7umwdCRaC#sL5yP{SWXalcULVy)Ku;=-2zrE
zdzZ129s}nsa~F(RqM)^?z7F_XU@9SIjL5-g`Kw_#3^4-elmaR=tQ3DW&G3x#htnR5
zKumSqK@ohw)tFOCvP45UqJnwLBMz%QsYEdD*Y_Ge#{6yXl<LvSh<ltmPXF3Vf#rB^
z3j<3bRHn|3hJf(Y<<96n*D*>9NM;O>YQq?AoC0dWBLeiYCfdY!l-$(o{pA0U#WTFa
zXp>Pod#3-db`7z0y>v1YB)JM$R!y#{g7A|UF+WBuyQc8x<1KyUnD2>B5dJ+@f%kCu
z@fn#HN6CM!45fTDGLoK__I6?=od6D;@RpmWe~n^5fEm1PRW;uGuhuV;6nn^-Gs$di
z;A;j2@HCKPUb*vk+D0lK^>kIypbvkIz#fG;Q&4?fUzAi2hMKfeQVia&m4%L}#RsK2
zeaTwPzg`#<f>9WbgNksEMpkMZuQHoq)Mur3F(~FGJvMFBFPmx+2g6$WT+EYt;{UQt
zF3lXWzB`eHP?`ZD^5zrPAB7l!r${TmJd1qdf8aG62)akmtF=H}G~Knno{NKNknXiU
zgQEF*8|M4-`L0ylV+=;~IIk{yRR6)?4oQuYE0=rN1A`>{RS`HAalh5n(Y-LPa*^=f
zs08_1Bss^m7H+e;z4Bg-5ny6r>$*S?m3;drg$)}TvWaG%lOfQ=xk&_b5E~P4?hI5X
z8wHt(`2!&YMBby%wN;bGF*xk`S>#8G0@P@<kH51taAT&K=LuiI1ms@uu?bLnrNf9Y
zQZNqc?m%?pjqVDHlo=f;elYdj?0-WjrY3Uoykp;jIj#I9Z>UZ&k{Kpv{!s@f6@|x7
z!l)<~^?|S=KmuzLa$aVy+&Hv}gT_9W6CSObKpa49Odp*&nK2eaVo2N_lL!MrpuPZD
zLlkthX^_GJp)L(J#0lFQbG_%C#$S*G!B}!Ul@0=J8EkN~=yOmKh02Xxx(Z>qb((Jl
z+_8sJ^PMJYm-w+rRt*DAlibG6_Qw1~>8vjO1zAM@`S?v{7xW(sCJ=g0ao6#LIdeh?
zaf3#dGHC)N{h^cz;XEmm@$|P6J)_R2vSEZBDvA(lp#s)T2(i*3EjWDPHLuza2oN|*
z@nyWW+6pPsSyW3b#HhK;H09{(>4NEE0$ZbM{Eqm=A8+PB>6l3sIKRpU7~=TlYFUS3
zLH^SP!i15}$5CZ_2hB#h!lT~<pBQw<=4t2HjaX#mfHvwqk(+0aGP`A^@+jSpf$YPz
zQN+CeNNjU|Ys6`ogz?rbZ89)X?5iDOE5-OSw=Ew1Km-hTVZ?(5^IzMN$>O}8Q#E1=
ztOwo%B64jDIo#uNy$xXmC%1Z3`6!?ql=Ctav?xZ?nN&Jo)mIKBHfH!iUrdz;=xr|y
zBl2w<QS2i{=c$KMXn}5cCr(}H8r7&~Fm`^8^>=;U)EyA|-u^Oe?8Gd4)cM0h+{Iyp
zY8e5Y|K815-MZUHdYJ5FM*)$`WT2ebIGBD<ETeNC@5r4@#e-pS$*=buPRFysYlHwz
zZ9&K~-h}ub_D_4XCJm9G<c15{aFZzx`WLyR)r%O_9xIN$+4r}Ug-<29^Ls3hg0wdw
zg_;bKnRbK4PdRRXc^Y;AiiN+L`I^{VG1z>AxuD4wy`a`Mlo<HW5{qUitaEuK>*e_a
zW_|*)hJuPJMDPq)oD-j17j*D`6bD}FhagPZd{2r>7*za@W3Of=6N1doC>`|xQ~4t>
zhVXcXTW1`aMbMGt!G8NYXZ>PaQ=5+caD6fNV2ms_Duq)4?PO{v`)srFyDttxp-4}m
zi5QdHT+h3Xa4wTt?_1gjQ?Z)Dwm)|}rh?SxX%J=!wqI4ml$FK^+`xg&-Y}}9O<ZGd
zn^eZ~(@HXUqNW|0gYHNIfNOd5w(T&~#uQb}&{iW*`cujlca54k*x-q()H3|!KW@Wd
z`#4N;6_g?9_6+UP4E9*#pqK%*PV$HARH&EzSb4Wmwu7klRHm>%FmdL1+lLCsQe{9X
zkdb(YP@PTpGrrmlt*0^}Y8n{fU@zL3Ya@+*odr<!99T7P4(>qOO3WWmTcmPUm!lSO
znSw-F*<p#73?>Exzw<|_Uc77u>R<WpUn?pA@qjTn;Fit!IqUDMllgiDrOBvj;Y4&a
zv5SRnz+tDX-3TUrd@|(7@3?3LgyV=HEybPDFl)dmlAj?CaPP+aF|)*fRS<EaY6-z5
zBz(sdb{bzaa+!p3#7k#b>WpOK1_Fv>+c_1UdNv)IZk(t9d!`_Fx$m1uvpc=wMnO_~
zc!S!+b^|Y(d2tMfRK)@4pjJQ~hz!?4;#7Po&2e5^n?L`iE#>bE=A*st4)AZF$46B<
z)_0}i)5s{Di1`urxF|I|xto4i{3>+THJ4GG6GMctfjl_n3mYBj(MX}EK}u{JQx#DM
zp~XiozPQ~p>6?*|?zWsr|GjRp;9DbUVen5?aT6}Ox*W<`wPn~TVzMz&6OiI$8iR%y
zJzHa6y9}JyUwZ-vrJ&O*-aX$?z(xJBqPyH2Uwmuu`WA1C-HjhbjWLhPUqR#yh2A1i
zg`}-@GVTWijG4wry32pHF&K=?mCiXeIcjlF=lF|)(P2`&cr}{?S0JE1tLe8=`>hl|
zsGH%Z!}tg!2UcC=o#BRk$gWWJRdg4c;{&#8M?5IKMi0vz-#y><C#;afHFjxc{rn-?
zgOks&bPA9p5!8yMEz2WtrV6A`6fkj6HUsDZf*FHr_W8S<QS%%qj;f0ZOQpNVRu*%a
zf9LFq(OIW)apZStpu#}=_1J@0i2BDuB*!ggmWg1<!wxJ4z=5Os=9jUNf+3-Eb1_eD
zB9<|)79J!>7P5Vq9|(nj16;9%YJ@*A4ue7<m>jp2@iB6%A6rug*Jf8vo1*$3>ljr6
zAwKs-N;nn-yYBX@C#UxjBj8>>o=>Lyfg(TVGysx?w7}tRwCvJdXK?kqUgzf(80s;{
zZ?CJD1wSwv2LWR|e@l@jT7f(YP&D}fp71&(X}XFAwQ!Rits{zNkQbkBi8E<8E4S%}
z`uL@msb2)sC(=sDX`-Yyv!w2FcO#=3apWj%dNP|wt6WnB5O)1Lkm;b7xy$|jCzBb(
z0=Y4z+Qcx2tCTWy?%o~ey+Bw{C``_!x7frS4abT6swV%6SLu!dc{Kr#6d5uDvKc^p
zPEiTO8O=o;T=}0xFVdiGr|5gq9_xNKIEMZ7>0B#7+ynhjfE6X@Ht33L{-tYr%jNu*
zL=m2d3O7C#1h3%$cV6rv-9;V@YK&?;=U=aVjjC8&JRrEeaQ78<UF5Mrsh`!FPP1TC
zP+2!9H75t93fQKk*OziHG4kx!OkBL^zrS7v@ewd3VB{3fokx{dwQz(VVMVDzd>f`l
zP-%XH%CLM!l*@$j*02*a4;Su#+9qjAN7E)2xZ7FqWAd9wVFnC)NW>NXs!=6}irD-1
z6d9lm_k+ahSz)PUUQZ{g&j|9LAdm`6h-4BnTXg3gQ`T3ZQmSUDVuTQ~MaDPqJMT#e
zAK7(rZ0*16yw_A3P8u`C$LZ4t7&~#}(@c94&3s;m+}vRQBEZio7p7BSz-qw4vZ_3~
zTeJ0GhE?aePPpw135&>4yiKeV2pI~K!_#g?IpH{qcZvI=vQ2D>QF$mJgRLF=U}3pS
z^V9xxB2DzrtL<K7;0+F0oWR9_fb({!LQu;0=YpLle7}LL%o~gXj9~vE^m0HqnNr9b
zyli>T1Ip*xjkn@|N;EO}KB~#_x1Ty^M=o|`eBQE}uLPk)RFG|YcT41lu_l&cphmf@
z?r5(OJW7F<kwQ)THxyB3HJ)GB4iC$~e?lOpmwQRN6%$2$eRULeZek3pd{PIlvM`03
z=p5A~h_9+a83G)>zH2z$-20PGLSetUB8zUVh`B_ZSs_M7k`NgaMt>|RB{RStuMspF
zt>nQh%2lU+n1OOgld7CNcfQ3B<reCrU3PIn+b!U|@dco!UNUu<eRw>@e0RnN(8hW<
zpdSzBx<2a@MqG2f%6+?OPB&j~=<{*T+sjq63U$F{KSEkeQ>h|VFq&hFcY&^k0`sei
z!42QU2-FH-5os+bJN&pT6VfJn-YHe85Oz7^OeQ<kL4?_g%bt=xZNv2K?car;qykYV
z;s%ExqTGcXBd7<X+oy-mWW&3m=4TK@^)O4jkP^NzT&`@5$WnvSohKkb0jY4gJ+k8w
z(c@Le3uOSRokKEafS2yo&4$RP@D!me-j{xz&5W7&J<uI^OM@wRK2N)E6QSI+R$^_}
zpq!6g1jWx2mIYc_{CZ%VBdPf4d<Z60c7S#idl(J-W~vbUV`B^yb`YLUs&1cbCQ&Mb
zLB^<N$Q()=L4}@GtYA!3cOnh+;@yJ%qoMvu`btRmosEo_-KqM0D48@!`_~tW>bEL1
z$R1>Cqz>on^T~^!MnGX*%awa6!B|3^*ZJ0;m!zoj)w*@0h>YW1gnA<wsHDz9cgH@}
zsPSusDL+e4XE<5mpn_#ivfZa_QiHm-_|s5+Ir4DF3RNLGIWFH%*Ej6IboSIfat8zj
zXW&ePF%A`b1DTbtbNEjT5U*P5lgRv7j;i)c&7@lQ6U~~8_BF3>4IQ?LGh!Grp=8PH
zo87y_kRY>;7pIOsa%(XTu#QWgDne<!EZ_H?6F^Z<?&}LsSx+LMYf;;eU&VGMx58Nv
zV(XxDzIeC?HBum0L@r|Tye)FLRaVR+vZI75T9V+r1jRD7GxSe+yu)f?v$z*YK#)Yz
zB9?<HE;Vz3;!^oq7UMvOF-LSw0E17eJmf=n$jQ6KBQ*Fq9nO4Fnc(J;$)ay(;DEC*
zVvEi`@%(TLm2&nu2+KZg>;pu(iHFJhOuPE=)39}uG^m{vzw5i!Xv%nj_hgvGjOR76
z;~OiS*b+fv(oq8mP_s2`KSR&2QuF;&B#qYz#|V_Qj3I*_3yiBD)#=wsf{&Z>x+5Iy
z45giehX&C}lFOB!J=1u%qaSGBiD#dOIKrW@`FPS<VuYV>?71FuUT%Y8^Ck{_cZnUr
zwnzFh==SQoCHQX>25>Te7Lk4(7U!;js!zdB&xq>}gxxobnAE`u31*^X$M;m|Z?C=v
zx@O8aH!f1>>mk(Q3nNh`>4XM=32>)=f;rFFpv-x`fCE*Nh$)MLTIU77hiUV_iqref
zW3DL;32$a0%kdR4t)oy)Y8iF1g82kVYYT&)x1%~Y7i$NoWQz|=96u#6l6hw>1H$ie
zig<KED?EG5cAq90<u;>Ir$Lc|C@Rp+AAt8W85bsg1X)WtU--9Fj~tgDb)%p3NuIFy
zYpAD5)&g#0>LCvXxp7GHy`nsStH2CUD84YM7Uq0_eL~1_Y9qTs3&*P&j@Ge$b;Tpv
zC`Y@YKI!^xv<TE*LMJw;xIsfkg3%Q%#F&a=Npgh0p(s%dRAH2MsuFyx*;ezFywJIP
z+zv<jWLDCqu3C9_@})%3)Y;@D=`d<VSM>BtHVbb;r4uz(q^XGh;$R`;yo<t7J*a6>
zq0>kOPhm1FnO%1R6|O0V`aL>*hu6F>+V7{TKRh>ieQm1w!O7i*V?@*jwQ3|)#guFU
zBYgUW#@kTpfoJFkq8TfTXK=<#i5pZKhuxL|4iPO74c{0fBQsQBi!nyKg?Mx^+*$5&
z-~fK7LoK^p)N1N4gAdlU3Qz{9Fx~>eV7Xny`nQ4s>+}9!hH<g-4%gETS&HMm@!s9h
zkJF3mW+SUuhX_7SN|0SG!Ce0tK}kl(_loll1BfK-Nh5vs#_Kjj*gP+sCN9spc8IVy
zvU>Tb92)koNw0PDegIf%suKohIt(^I7C(=qCEv>Di_&h2;+Y#n+VW!<WA#bQ;*6-m
z^T2exN;W3B({zN-k1feCP@*HJh98@HbHwMLIXa;s<|V4}J)7{+_@pBl8egz=xEyZx
z$yX%xI|OnYdjOrcp*~{s<uX`!FkPCGP>l<)_7F6+etTaP+r__3)YqG0wai&rjl8iB
zpMpA-M)5fJx2Tf)_N$AyOV@7<R3mjE7ko)dWNjMa(7ekTp&S`swm`S_(yE;{mW=ml
zXXrTLm0E8^eghmnu-Gzr3Bzd39uT35_|BJ%;U*(8sQK=ZQJ8dou9J67gU6`6DYG^l
zDqlYVK)AJr;8_O>b(!b73!&H|f9wHN$*GB2G}b4{jz_7Xrf3ku$b1zJXgcd~3TP@|
zV~%*$Vk~UJQ#OQ~9$^O|*9m`~aq|&X%Xm&%mayc75QFs=B^+I}m^Mrj-ZtRgcJ_nP
z>{~L;w#DmH{bME>Sde2M_^#fFQP`boVCtfCVZ`sg2|Yw(7t+n@{U$54x>(q-q=E0a
zI2)O1c7|Eu9>3XhNb$4ptxDHkl4c@k4gZoEtKkw)G(S|9Lx?(8$XWeht(A`;kMudC
zMKek-#Y=hXTA=%^@bvD+99gLbdoN7w=p)oo7*fSp33rJX53+9Mk(%SlJZH&6QGG`s
zXl}*+n&I4e&GpkzS#(PLLK!hiGCb>RjU}p{auzKdAY?7@Fh+MOfrA?wI2AYn)lXa6
z1a#sC3w<LZH?vavUD{YLPYCHN^C$=bA6?yxpos#?0a;kUXxT2^9eNxHsFVvgpKX0~
z=*d(DvKkQ{RHp<uuFGF=BPqdM$KBKk84KdaiBXAJQb}JOxg3)ZSvLWnUXbPMc%{2i
zx9fO;$!8`ol)g>xc;VU>VKaM*bK_;j)6<+EMMukRW1FF_(@bqRd9}!DrM5{L%*2`;
z#K&Ge-9m!O=stKdsiL<Mf?_h+AEOwGKu|5j9q>Z(*>4@kgC*Y5xl}w}w@cinkX>?r
zL%Wu#KyR>nYSiITMepUiTM!N-KZ$!kH6^%NYoGQtdoHh>3qfY=PM2+1W;Kk(n0VZk
zfDw$mREeM9bq3@1<EPzXC?}A5cu893dS{IQ3c^7tGN#Y4_BT_8DP+a?^HE2b0D<}r
zF!!+PQdkYYnBe|FE+o4z5g>=<tm4>1cH;zr<#A?!$WJh59)Nza`P~UYlqY~H6!j7S
zr-^QrZ+o&-+c^r1Y<dcV{m)f9zC{HWF-{mTP5t~e?wne05Uu-wSLootircd3wK#Ib
zp5PQWI`y?bj{B?t-S@Gk*m79@i27g%fxWGDmqt|=#P*j}WNDzFa%wT!G6ZL-YB+`~
ztmD(o2*T=$t@gNCk0-9c=^JI6F&~czUm2@L`h5q<=x`W~CfAH(nA@f}9K;q$yBxhY
zxJA3e$$>5(fw-VGmG7qaZ3%e9@%4vK*c{Xn1>~8xJ9J@z>nF(CvX*PUfog7FaQeEr
zmtGfY-=W`$?d)hia2uB7G~!gDO-j2eiY&R`#kWP=2qE8T(q1}aU8LKWE=G&9v3ajK
z9jvbrw!cX@6$>lC6(7YXN%`URy)e|aZ;=2=qQ*K<$MJDuQ!%n@+IRGrKT=Thn+G&o
z-~~PX5v~)XSU1d57d1d+EPHNCG~6<ztmVK!uPc-oRlds;LSJ~XRQM2%zd#lT`fZq2
zj8_wFr{KAUwRI|G50Iafyj|(LFmt#!D8iv^Uz>#U)DFo5>7i_i#7$;diAl%~!Lc%W
z&viIX+qHcJDM=Kuhr4bZB5dw3*Tz7DmX_XKZDTEOEBDeX`AAB}I3XXa9|ekr(jeJ*
zsh`)A9#tD#Q8l0e9Kc(zmHuX(m&4ZT!}Uq+JW42%Vw>A}cj=}5!!!w!$=Ao}+pLL)
zqXI~}cyYGeFYWJsv&IX0{X<zutkFOt10iJxJ7OCqe1XQt0n``yDN&YK7C6#@GsdGh
zR}c2K;k1)b1;RRrIC9NuUL=L2hv1W^zD(O5xZ;E$B2)FDsyrWy`|0^5SuAf30i_jr
ztJREWM<t2k0q#$_mZ_3#^15kxPEwR=ZyYacM(%o0ncXfC5eiqN^Lm>Zch@q}9QjIl
zQWROAv~-*el_Z`jK%r}3x}QpiIM>In>xGBT<^<SN1q0NW$vIYVMEKq{fmvGA<KTdb
z3D$ch`Z5r^m%IpoqQ308nwk~Rek8ltkR-p@Y1QLfp4+!4IBB!mV93N4zw*TN8R*je
zAPoL8Tx04(rD$@OE#AsIjQCky3Y{uK#E4B*k`qoxSpo_e6?8ywjC>#tDk`AiufV~1
zd@m}N{|{a79gp?<zK^3@-AHyKao_fq8D-qIjLhsPkrf#wBb$<BWt3fJ8QC+TR7QkC
zW{9#^B>Q_lQSaC1@q7IKdcPm9hj>1(=XIUec^>C+94FxD;my5AWkGFWoMZ>dK~`8b
z^b^VvHPxzpZ<)hg*I38{yttQ<JV25fzye0Mvl`*t0CnXkQ15kr79sOmPWTfq&)X_v
zQghA1s9Y)1!U&&EuA?1)8cWDiRvB8Z<N@@!Qs&5V|L!pkgq}9G@Q_}#OAdEt=%#K~
zoJ7~<_plFNUI~OEO-(p^_LFX9safU?0D^nmB_Qn<CB%AV$=QGH=^O2nTnE*DOtff#
zm_xti^i4`gn_VVQJEsSvxsiGb3=-yF{c1<703XPT+IA^4q=Q+Q`#Cv^JJt_betrS3
zqJDsEZwPKbbXa(k-}Xe5eU$G<OOZ1BP402>0Ejw_cacgM7(J#`6s0T#R20kWzi<Pm
zAm#>Ka}px-Cp(OwhB#h`E)Z5OLRfhoebk7oC@+%NKh$b!VjFKu9GWppm59m{3jM5s
zm4FgAJJEZj^dgW!DxMw4z2y!x{j@q}3!><=pNxtv&&T$f1siKgCr8kKuiXJtz#E2$
zv*3fJ<)(Kp^z-e`s@-Q+3{~J~$|$MY(eAuP^qT=`X>3E`hdrkvPvRzEpC#91;g+S}
z#;uj+=iLn@#vH)~I83!`%zM|OwLe_14LfCd1|rKwf#ocxfzQ2Z&h_%uGBG>64Q$s&
zX2?R1XU4=}k(S2|HbigC&T+XY&-cmQ2}cX4ItmO%M1NlX_*gI@xDW_R+X)>ryX*lk
zdvoeQMPGL^cXGEWibYdq`=&LI@n%G*(-e3%>I@LspgljONcqT^39W)(MQb-Izc5BU
z_=oXodhWGaTgUdpq7-2fp*;d!VnkV+B1O1by5iO^X1#al9VAZ_XCy!0(3nsKTS?H1
zXKd&8c?r1eVdhZbU-1#1F*iz34k04&;7Rj%`(q}J3!PAxT0t$chYlPx?%2$u4fU7r
zx{)>(FQ1O12?`Da!<K?iMZd^y&pAi|u!<P=9n(#5>$laETGiXr?&Z$$=q!Y)B>1jy
zJPEP7=e&8`_~I0zCOoUS(di=i`+1&*J?)X+l6S?l1vAl+_ehTOI!$<GpQ|JupTBa!
zS5IU#&1v+_2DdG@pZ?L$)mP}_guOd!y>63^JV&H_o&uWbp6uDsHB2~ZdF(A)DK5i%
z`tjB9BC_OCUs10G#bd}8^DBV}|A0C;?hFHH7@(JUSr#-3B80LFd{6GKEqwd7BiuOt
zXukzp4~2Rw;n)lod<_>VTU}245{6gB20_6(I3>r|o=bEDRK06Hf(Ymd0FvX^LfLGn
z*I#`5^7|Rj@G{Z(QF;-TnRg#to~C@J6YCHjA9QONHm?ELG(59gyA>qlpekfoePPNQ
zS;Iu55hVdYGk9GkM$YyZeFs|7=a5THB^Q3&WAd>{?IusKCyw&&$p9@jR_4S`_jSg=
zJt`Q)?<{f1zn{MkOHKl<n_M7q0^2p>NE}8JWm<zl%wasboD%jhy#sf}@6|~N6%$KB
z)hfihPS;BOhWrf_j0I+hQAL?i(cu8~h<>8E(m^1<&XrGF+EK;EkYQ~M-@H!e$+z0p
zlw4A<9!~?>XhBJpOXX}T)21z9?&IR*M@}0hpQK$c<>qHxl4ubBI1~|jjgj|KwTPvF
zsmDZYC{wVcm5}IeeRu1I&f}6%O5d%ed-f!oKs~J@-ZUtgQQB{{oxtx^2m%1*G)2Qm
zWD^yB7H0gs@@FLJt|tRH&L`(?-XlYOl{k0d`&M*CTVYm&SJ0y=Uh}rl1O_yY<r6(W
z^cSXPA!|sysY+>lzL{o1r2F@5-!H`KLh0o-nZvH!KPeN6$31_fho5}`0i@s~>R6dr
zJ6L92HqY@c)Rq2twKw(oP1to{6PZScN|tIaTOI=V4Z7Xy;I~Zj3u=llv`$}gv?(sr
z@)>7@my<Lu+Z&`GgC?AqfvrdZT+lx{&j0!ETR{r~?A=QBGakKFK(WX(Yob;n&O!tf
z6T3<*mvU384IwM&ARz;8Ej_rkj%PMmfqX`K{}To1e~hr|+=3Qq!cGtUJY<Zjqs(#}
z-WivgG`8I=fJYH}*WS(-{2DEcO5VK|8|VUSl3P+bkJvR#CGMX>ZfRrI53luSShx(Y
z|3tAQ&oI%<T`XYwObeR3_#I=n_n(v^7`*{fwBciM?TK|FHHzobRr0-F)8bABMzhzf
z?OsRS{sM^5oXNf`s}m74sFlx+eLcr42^Cy{5Cb#09(*oTnYAs#Cy8W)&o$-Ta|!%P
zH8pr!PyBG;wfB`L{}W`EAm;F8vJbe!ZC94>OSt|B%C!VA10gy>P%OK$3}&bq@s+pM
zoPtXvzirwZw0_GTe}YIF45VZ)z2m&hgt+-mZhs~qBo-gaZu=;Tt?!Qg;q*R?4$M>7
zsTzfHe)NxH#?46JpvauIg}sA^AMYnj(O_w(1b#Y+NU=u1JCYpW1UTj4%32Xp1pCD3
z!UYr^&k+~<3R}c`^>GEXD(K}prg^gJVi9$)_hU=&#^^$n?xy#%q-}U*7qkojUIC(Q
z`@164!b?cUW|SOn@mh=ezby-mGP_nc5<p^6CHAugwUFMGCVR+B=BR@RWr42g=IK<Y
zH$<bV30uH2fTX^I!LI2ke6SxKtXsK#gZng3(0<N^q&<vE(ggw&(oSl!uw(|xp@wXs
zc_P%vzT<2c5B1zh*sr_AV6~fd!7pbx(lAc&<A9<5;k)FB)Iu?RMc-$fY0f9ldtIH3
z?a5atHDIX$B8Uy1earf}e-6ijLI40@eIFAa@oOrBWLPud$3hs}AW(pTAln}6K~zw7
z_lvsMz57l1;ZFJjd;qfc@;d=r+jf=X9LMP?oR8iy_<Wyb`%5Lqe;T!#R%Y$k*b2B_
ztZe-9T(|6Ev4j^GDcPgi8O{>z+PMvd>|^6E{dTC*PEn{!Q&AixypKn6nh;x`to9kf
z84!gi*se;S1VuOI$7us;?#5#E_;)cKb$*}M(@*^*q2*`z{Ahc|oR~R#x(~84(O!*J
z*5@w<FGkoQErs8j;oFy3^PMw1h}sPUa?A71O8!i8H4W=mJ=4rV9sh&jYc~kPF1!?E
zH*&aO%J?`U)c%^E(x2KL=m(a08NJMV-_=V0sN`jrOn;YA$(q@|LeHI0MD98}^JM(*
zhi8T-`Q(og$Wl7x*FUEtlC?YKfE1PFi^P)=w=Il*fBSJE!7$E2j<>bPDYN8PI{cr_
z8h*J8)z;X_2=#@;*#fdS325$Tov)!x5WfR<2&v!$k?_3G<JYxbqlKsT&sTt1`)4h>
z#_lx`w@Ups!MAY59dOfM(h)q$Y(QJc@+Mrly9z)&{gwNIy}`E9yx?;=6MlrzJ&lKB
z_$`QVzH|rnCkjrhhr2ICCr=#urQLbt>+v4H67R_j@G`h&4TKUII|2J0|7*V==dP5^
zK_gp21ip}1#ow9-x3GYKRax~&-rbjqv`X=*I?VOK(n#cCm&|G(VS`O1a%Bc^L+rBr
zMA=$w1YwK5;HI089(=%Ru9ho7ql7XS^3Cg{?FKtMEciSNWMt)~1K*RW#_`(HyMYSo
zF7Gw#>ajn@>hGmzZhJNa5F!S0ruIAHSBa`S7OWrS`qUcOe2vv&dUva;ViCN(-4+{;
z4BWEw8is)AsrKXig}6I<EB1O<FM}ZVVemF2t1hHX-b?Rjhe=X>Hf(Fsudes|{O+^<
z&1PWg99&^M0Vr?NK-Av7I6?y*_zoGbS5{53(1U^Qef0JMl{L|a%OVf;_cFO4Di9#;
zRtU%?faWhktMC_SSGG@nG1P}|sh6XLu=>{}?lNgm=|&2U3uJ-Pb086m_{t%LXTJWW
zn*4HTJS1tzk3mY(sni0F5S{6E@gpGXRji$YPHq{W!@_WFdYrm7{yX3)PMx&49pugZ
zzSMBV<8*M@(Ko<KI`JT)iuLmS>wj`gfrpe}(NlfP-rGb}Q0B!VkQ>WVVlU-E!TkH_
zqn@ljw=A(M<z(_%mu^?|_LsN9?dnb2wU?AqppYC~h?k0j*~yvvHsQ)kbGTp~*OtbS
z4E)XBnd@U>gqq`SS7Y(@$h7yxT~a}q&yXxndT-^(Q0O-S$aZ`7wQX0%eeX#P9;mAA
z9t4!MpKWbFuW7ur1SKr62el7zdfqTT`F5Ibr>P>G?IV5F-7hH)ozH@E-=OgWpxXlB
zzLjRUOu!sGdPf=p=NR%g$7_QxVKlqwwsIz0MLkz+K6pUlm_S6`J-E8vWoLjGmzYf$
zH+&QoY%($PeKu1j{RnI{^PN&jVY}Yu*lRtIUmO8+><k!A%RQZuS<)8B#seC`A~;*z
z0T$;E?aj3Xrqyr1vQ85rcY=QsG)@(HTi5ZZT8Y;b+2h!r;C(0-0^;&~`thqf;P84$
z?dRHTpYU>7!PI-GPMrZjWHBF`G0K0F!4@d9^XICDXor_NI9w6kuE|Mf8RcK$ssjaF
z!P8Ts4PFB8(nLKpnxOaks)y?=lHEb$!9dU<a{ypIV5KMCnd{Dz%51|D5oPSAkM%Ql
z)NBUlny<7j0m-pYm}Zi!DynjZY!;IKuGcMz2zf`;`NKF!0yDw4?mx+e0(p*EZFMbD
zSuq<rb1M@9jFo=s!&L;w+Vcnttc9c5_NTG<sC$})(MGnD<Mw3!Z!mSw6&4}Lqh0?H
z!Njz_ZtlB%>$7Y7MdP1R%WP9J=Kx&-*-`;8<$tGhZ-aGB=I#~dZ%a@jRWRv9IK=nv
zQ~>trL=2YNdPLSuPvLG%h`Aj9CH%9SlrbLmeeBwt>hkHiHRTD&^NThW)ov&%rAT{M
zUox9IGwjYX&IU}-G_L0lqvYd?guy#PPS@;_63ZP)xB-0__qVGyA(%;Afp~ixVVp3n
zQW9e#<p~OUqaogxF?hN&((GN_C9rRWW;A(uN_L>g+L54b-AJW7UbUM3ZY?=J{K(N0
z<g3lV*3j&2ZSTiQO<#P%ohCa4U6C5y><bqbFA4bZFsDAhM|9hKF2<gqM%Ctp|8^u)
zE-|jZflT&gS?Gl&Z~*D2w6m_Aczib1YhEJz@vpjsV#H>`HBLbfe^;8@p}*Amtl|R-
zC`p^<2O~8n5)Ng&G^tdZ)rb&fB`+3x^xRqcAw>e&{kUDB>{k$Be-A&{<0ASu5g1&$
z6}vv?4b4;#Zev|M+L<GH2sUWj0mO2U9tJT#+1ia1Q(KBp`)T|IB(M4c|K2dJmvvhL
zj_I$Y2Uc_0*=KgC+hc=<dwPwPj3XuiD|@*0(O6sA>urJ%f`x(}3LM2OU7|H@?lbcJ
zNxUQFM(R0gOp<SxF<$h6vE)|?>(D;GH{sqo;7_Hi1;<2l%u23lI7CPSj0PgPZffYb
z7(t#R8qwTGD@&zJeOqj%cP63;I1Vo%nq}S5qLNtBp6Yn|7JfRDrWBO^r^;ALqn4dN
zoU+zW?@G+xxEy5St@ev1se$!k?x_+MUwh!wNMCt_`iXx5l)=FHQ*u86)}>Cm7mywg
z7<~Kzr61%*`M1TBN`PZ?=g(p!TeAYD0tJQJx$_CuJ=8%66oXLm_IASy=)dJCyoR$K
zdqSXy*TQIbv$^O9&V?=iWo^n%2upYeQw~LnQO&FVk|}T-%BVGGj?tXet8$s3+@w^h
zorXTvh1saB*VE92)klAotouQ}1yQAX3i=h#)k%rmHx)Zk$z8b<0HTW0ayx@xijra4
z`-Qt1RXa(4at3epUYe$@J+lD-jPmbwXl}0TBo8G5jlwrbO>bD8w`1|~v9~lk(C+y7
z`*6S7cc}2t23^f-FMl+xP4MnPAu9On%t4k|;cP;Y(UF|=aG~AVE5vda*4W9AF0-X7
zKwMel$w+JHV=ChK_4CmquK})^Z}SjWKkwR2B_db<mhNzDhmHh%En$I%SOIgh-0B(M
zOYfp3_0;bGz<yqza^7W5j$vN!+OL`JT<e!#Gf%}PKW9Xwa*TQKZlG&*l6qRwCJ$Kc
z5i-W>7m}fx^sJGbzFV7h=K_bz#9mG-VINu#l{B*-;z7dWI3-h`^%1d|1@p)@22U#E
zlo_NWvAjC)6RN{saoXWO@@yW;(P$Y98>|p@Myq1g64d{bZ`Al5e<&g>y~;$*_keb@
zEJguyju7<KRvNEGKJ@@S<5zWHGwWWg_rqDjL5T9+l2+s8!M|c*w5PY1@DC54if(+L
zCgc^XFW6PnY-a}iWtxziQenfC^gy%Q+1&j$8B#v|aD%xdH~IwG>Q4pr79tv#IT?oU
zESdR_dCc;<uTksj%a>^uuhAvdmtM&MUh{bOT&?K1=Y@hhdZj|i!fu)O(Zbh0qMHkb
zgyzcpM0=Ix;YF&25c%mZx{lmC!+qTJ!Dpc?N*3x5F==J4vP0jVG;bT7SCQWPz$ky;
z-OX=5XkYIHj(rg6K;QmJP4Jusx-NE7&yixJ$pqf5uLIg>W7C8dMq|r+5>e+>m#|>h
z8|SRDuRyP^@>llrVZ7?)79JYAFUDs*PnkWSJ>~jQ@W)Giwo<7)U*F1M$lrh0>O9tu
zlzPwq@v<PF>38o#YBTgxe0)-|JA+*(Cp^_!eIA&8_AUKfJbYzMx?D^<q_FZ~#U00q
z@eXB0SCS+n;{+69?RKu?CEH?Ygd(&h+LSL1NkIzpbyH6G=*Kd8H)C+4cg;7`?LJOV
zEG%JNHv8C3Jfzk*cID-dO2)u9dx-~SM>zT#bAV3~=%Y^_N4Xf_1YSS(t>s(-MZix-
z6QBws7xKQHf1NAj&qZq>^!~6f$Nw<gDj4^%8ykdpc{fF3f179juA5E13G#z2pYBU0
zXT25@qZ#;?sjkvqD}O9D1i|=uMYD00imr?4I^d$lukUy?9Wev%`yVl;aD8Gw9mWV(
zNOMEesNe}{n$MSjI_NYx<sZ@qZ5;S~D-8)Xbl;|DbGeo)o=WN+d@CK=D2WTjl%8Ce
zVY^$jP{D}$?^Q7VhN#*5=zh88x-|(4j}tAJ$U@+I1Y~J*bBy*^?orM#pMd0N)r<&W
z!@%#h)qnq63c5MZzH_;SZm=B%tr=jCnYS}(!u9a_iBzX)rT|{CJsU7=AATwznuug_
zfm1QbYh`kjSn%A?)81wvpwDK|XrOP?T3^=n;tt}$`6LihHzMefXmh=2r-FIuu~TOy
zml~-65mEaSg><a)38wPIe%x_gE$g;3h`;342Y3c$fj%veSA;a*l_BXLBSK#%J0}&X
zagF$&yyOYAFTr&B=RCscpw-y{I)}~$JMKYEE{<1;HM_a&Hojzh70+%>tSGs@3#FNR
zNbXW`Nn&k$Is{7FNl|(02O!nxNDC-;+=X7RNLTugBq(#hhHa@_Ew}H>i;EE2FCnsK
z_@7$BrslMIK=Y}43_jWQ`=8Z&<W;&ZtD2%LRba$LFZmS!Kj_lwT+B~tt4q-Ba^l`^
z<Kb=`4g)&}+dOt=8!hbkz+sPPvcj&~xZTR`9Zlo!ogVHQvWBHLQ8RO)rq^jdg+UC(
zM3O?%0B^JcFI1etqO*AdX>~mB4*(&U3pKZTe*K@N?`0o~r@!2xaB$a3!{OKL0IkNR
zrwIYp30+GAR}fo)RJOnJH3{xCjTg|+tOX?Z=Xx2-8_3(D(;b>qKJ?6+-`M??j-}gv
zYYVY-gEr=O@{9FPnvhf=CM0HA=P&x&8vlZ<&&KXEKN`x^aEXH$2P6lC+&9vmX-kBj
zO|2I-4ZxWaAhO#Q`>8}2RngZwz4{CB3<1%~Y*!}9_akD*_5lx+i!m;D)HYfEa}Z*d
zyX5Gq!ffUbn5BJj){on63&Y#c4^06bXL)p&$RNO4Gu?kcEp*ZA&~(|+w%ISG?e09;
zuMBVttabw&k6v=2jaxha-m!T8mr2N5iQRU)s@*}N7h({$>`Vf-7Tu<wVLSr3Gfisw
zF1-e;$Np@C1y>1cgm<kQ_E)@$!lh}ZO2grGp?_BUX4daIo^gCR`=q%BbfJ2m`5MQY
ze?G(V?KbQtu!iphwRtX$G)w&WbxVLgPOF`GJ`Qw{50pw#?+)~sxJY0I6Fxs<tF}zQ
zYZ>zjE2=<I2|OL;uoI))8KO}%s;8eI-I2R;{~`Ly<6fe}=tWy>AMDJ$#H%RpCQ|y`
z^_iG&ydP4K$eQztK6s=O>WN}N_EL>{;4B7YIa<_QNFgXBV1}9Te(C6k!ypXaViQ}v
zH;BFnoQ$v1aKb=`iX17Qv=V56{))r?om}E_sa}-g3XIaJ8h&<DdKYjg<87JW?UExe
zo^O7?k)TG!O_LxJ)fA(o$E<^8W4o$IMIUA>C-Q>xnGHE<LeA0H!wQE&m0~k8sBNP&
zwq7rIZO6??%EM?liM3x_+V%@@zuNo$uKk%((rada$-~NY(|=iwi@*{hF7P%Xg2h9W
znP^WQNdF2Ze29#O@Z9nj-LSiU9|&d5Fki`@jVhf)<M0T`{ctwu!Z@0Q=a!h~au~H2
zD{>s<DnU57NU0=V$RjPLz{8}JH{NRR|1=&dKm`|p-x6{;91S`d&)K36P6{E{xM@mM
zn!RFw)jOENV27RUO`e~!fWr=L0977byU$BAwNzYR1sS-G7Rx$w<mJA{3M}r$m?pxf
zV7Jtnq;l{qcqln}zp}gzb+$w2X$ai08N^Cw&xhdv0;H748@vg`QS}V`dl)?Yy@x@I
z;770!&`~@~Y6BSLa&t7B<984dX?CsyYk91QDMtfFTg6^CYH))#&~!>MG!=k-R3!>F
z4j9@X;3^pk7Xx1zb+8Vz<d^w)+<qiPDarepqC2&oX{Zv}kPX4)itoQ&cGiD>N<LR{
zxgo%rO#9IO-$$0N^u{-z#t@3A(MSqxf^G8YD*+7X!8nj1GJtKA3DwKI!B&?|7>|^9
zsx6K$QAZ1i#2@X*DX|z@l!KEX6@d{s<&}5XgD0pYf-f<@ZgulKvmdQov{R?}6VwXh
zz%xaXwQ=ATkur#Dy<Y!HX3E<{?uUoM;Q4D&x&sNd7<e#r3Cq~gkjOv0z+PH#8FCun
zWO)N5-l2q`acB{9UjG7S30Fy}$)|vhQ^71ZwokBT%3<S2-~T?w3BL^93Dbx&Y$aYA
z7b?x5tS7-fZ?s9hLx6BzdHRr+Ao$Y2&Z3Gtv(70&1v>{Gfn0>3fJH8xL8AP6pPW)q
z&=0L10!2ZWi1x%YFM-78s4<2(qh~O*_GIDw3E+l$km%A!k5ND~$oQB#9=RqLSvL!a
z%6VohX~u(hGGwne*5~~sef12%W8ec5>_6JxPBo_Z18)a(tyvhxrNj^KcclFNhgRbY
zhg82+fEV34&t|rGUX${Vq1wc>pwj^mEpMq+pl%TG6CUP9wGjaFoK9KO+<T<ZvJ~))
za{nFn-aHFU-o%-emDnJM{H49c4Z}JC|DX(v;tHfdF9VhfouO!@5g;)+&mDPz_-M`A
ztRS4)88IJlQ5j-(F(q)3b&|i{OgIXePQ*YM{ogZ#tZpMczB;-{?0XpPzdvhv6yHig
z%kEoK8O{8+O9gEwg70{ALwWV7dj`sAZEl$gcGW5%cR02Jh^zG^ZiwEJ4BtxD5R#jV
z(J-D1vv+`zhlO7+4!>S-2mkV){l`Z>XEM5pqTyktFd!@p6l2do&$Twx0zf0HfHyeb
zWrl)W=n`B80(6~`vnw4WF0#mADpFm3G*q^Q%>P+=Je0)~JoX^N6z})P4)#7Csw);R
z95xSx)H>ey8yOQ%tW%!+;kwU|4+Gu2^dz1PF&C@h(W2rH3;pjC;-O68tD4duY~a1$
zWvhax9Mm|9aF<e?<hcZ~;DXm*JBUn1IUe;830O~ZoMlgoqyN1ae4=_}Xyd<y#9_A4
zlHfaOsi1;}5LhF1Psc(ovf=&Q9Mo~n51N$fzXx<=Yo-Ef)~9`N-v9rX?GllV#w!pG
zxt+IC_*!)J9?ex#Ees2pf6QZYC|hq<R3H{CbKe}u6N^pB2R{y1Vj6#Z@Y9z6-}A7_
z;ZeykJBg$IEfywn1Jx7~N&Xq`J+EUE&5u6lQh6@{F5!BTN&1dEa0eYpi_x!^-k$`C
zsxY;t^g1g)TImq}|6UP~fZ318KLY=Mi;1hIToh-1wK1J5&aA+W0o%fqCmq#w+3akJ
z7IYqyQG^YLf|&pJ<9Mhs%6d=?5sCh1A#K%Q@!i*tb86M=*4>)X*VI^1RODY4(XYOy
zsS!r|y&BUfPJWl{fA4|kUx_-6ZM2dj{oilK^;6bUV9j1TYoJZauivk*5{ZS>;I|0a
zrAKD7$30-A4Fcfvj7<k7&i}OA|2Y(Duu!iIbAtB22C|=3Ob8g2ku-+e;U#U*aEVnb
zWa@MVpJDaE_2au68y!^u3F`l|uP9-4uvnol|M?0TSbPcQQip*aew<PeIusc5A0Mmy
zf%1OHVx>5m6DZ@&pBbqF@S{lHfPs>OVim>|8f5wkd&mFZ8;9YsqvZXOAaU@0BMVTb
z%0X}!jf7w}7^&U>nFku1v0+AZXKzk|y@w=+h1_Wj6}!4&olh+dBJ+Gi@%PHw|JCvE
zEC0{2z->^7!|)-=a5Pd?Zd@9UH>ifsDs0s<{sJ_2n2Hk*UqO}bJ|_sHE=@$T@4bG=
zLGlKY$Pf~_NseUve?bc$&6)@Y)iV9N!oTOm^N*p7WRZjGn57np2YipZ3IaWRV2TT*
z<8#x}&SQ;mUsZ~7#6YK!{^eU)SY}OkM_!S|N({99IF2i|oxuM0{UzZDM}K%&@}HN|
z@rPL8j73~E^Cf{@066bZLgNH@`CUj5;TjeLd?BPlI*Z#`Hstx#wZi{wKrRmWZWBkR
z{<G6D>nP~1{W#0qc+#I^z4plBN3b4!3jO|qG2j3Kzwh%PTx($L59aF+qtX#PD%%K1
zN>ZYNhDdLvIe<LyS9|1to;g|rwpnqf^M6hQj+-1gJdipHrclU^-c^DGGP?WyI`BJR
zZ!Nciup-`Mp+z+%kGcg=IT`oGv}quxj)56<B`NFoVDetVQE>BF)p~~h-_C$VLJmiz
z$bbK~ke?q13qd$q^<D>D<hR+6F-hlr`3w;E2|yK6q`U^@VcDTjjpsHf*aC9#7F=;>
z^?=r5rS-MSHG4o^$PDcw3;s!gS3IDvrlItIx91pJBv(|v`cevx?SBUn52ce_4})!F
z0Ay6D2(>Q-1q48LnHv`gzKnvn!t%5ss4Y#CHP23E0NhW_t)~(&dqlN7uWhwwHh_fR
z9qu#WqO5=t1!3Xq3tRyo1%xYPZ=rfT6F9{m)X_-34wN>FP`tq4|D9exoDg~>IP$1X
z62br0C9(zOq_BwB7B7&>i@(&bT}oYGs0rB91kC7yr*Z4`ae~GWzj3L`??bCU?jMPT
z!v*lAdjX}cVg!jCxY1bH{W*I4zp=2H1-6aHpg76@{v$?|$R>YKV>dSpZOowAMuI97
zu)2@%kRkh{e>jg&3&qm`nv((DP}FZmztEc*yd<4BX<EdDN>^Yo>|J(R6y$@X|L1;S
zg`AlAgE-3nC~KNu9XLs}Lj9T-<LBGBmYw9^H@5lt*ttw%NpM#$yRA9AO(%}xCp~y^
zTr?V73^X2$D2Fq28f_TRQRu1bkd4~&zO9EY4o&nVa()vBFhav{B_q;9WN^|4xk9Vf
z<FrjMw{rUURSia%dlP)R>nm(&dnUa1B>TTw#h$=8yT9i4%bwzumyA~5`K?Y#d5VDQ
zF!U?kZBaE7+=caTso_~=DJx|DmgE>(IiWOXRc`W*hZmi_5Or7n?&hOK2zkwy(@gSr
zSQVo2v&L16nK-S}t`kMo6rvm^K6_Q}AP$S$NGmrx!?*qKC&qLWabg!C^I{pACba`T
z>c;wo)m&83ThMm6LWm(Y%TVT!;%&A~hh72<vhA#cS(#lkIo4X9#X90FXXg5Xh^QPv
z$e>wrf@`{27i`%9$!Kzi7i9U>e){0dK35w-T+J@>5)sqW8NJz9Q0)<@<BW>koQ!>?
z28th(T*LCX%nuSK@1nJ!s|43N+e_>A0GulNvYZN$RQtnFKgYCU^1St38HW&E?UYqd
zs_^M85&Mipm&Q0P$YWi1*Rt^%00veqA?XtQdvG`~X=jQc#Z_jY_yYGvxj1tYo)N)4
z;QaB_q0Z%0Xh`|Nc$xR|S3!#dLE+4QQj&W4%P{z7w=TTw8sM!ItqlMzNw#%W<(v{j
zoQtj#^~+KkHFed&VLNXvT4{O|xrzfHrTmbO`_KdI2=KDVugTv5a&>+HMdtcN-R}f0
zrkqB_8AHL)E_nk#adkw4v~&f3e+Pub`OSo21Qx)D`orv-+oero_Pr1nxpcF>M&%Y%
zPlLpv&_bOEE?#IS8fv}}RP_6o*+z=vJGdb393Wk=y%q@X$>tO48_u0jalDvYKQG7k
z)42snsp+QJ2AS!n=fz=3(H|tQEpo0{WSPvt>d0(<OC9@|HF%3hh&6xeC!{g~jaB&H
zZ*n@CN_2DH57^Op5JJff3Fbuc?)Le}?{apl>N0;Lexn-h>9MkxEesSO>k)PtZ*n)$
zc|z`wpe^?jP9Ap|YrGG*wpxyN<%|?R8bG^w(Oy%kd93i+zxUY%l@VyOh2>dHeFJ(1
zg4T0Lvb7vGZZJr>VAASB*vCCqU^P``utWO1EIR8?oDhe?aX2OfHws0?TJEPCF++rA
zCAcSlpQ1x;3VxF$SS&!S0|G5LAF%bD^}iZQPff&@;Y-uiOS*Ra{$`ft;pWQE=g*wU
zmt$p5W+rI$l^OG7m++{#7JRa28$<{R{Vg0HicohIMgMLjE@s3Z?iigAYZqZSx*698
zxhI7B{`;PJIzc&KKepc^3GhBj{`UffL}42gsbxuQCH_4j(fA6RMu?Q(-Uu1TFj&Sb
z^_soRTB=AvM^HJgkoke}1`Pfwi)6bvL38EY_2@5Ecc}SmSGqvL)(z%YCRa-=o1-&6
z+aC2tLrHLlG8%_#ormse*O0z1GN+@aY%Zu)Ve7@X9@}R2@oy}?)!iwJjioke1CEVJ
zX}*wQ41}a6jIlzACb`kd<^S|3(MAm*&l7p%+xJWVW@;E`0v81>C6reeq<+q(De0c=
zoq;_dZ_P6+dI}sZ37+>c046_1rq|pJBMEqlJGi$R!4rwOd#AU_I8J{&o#_^28)eJs
zr?wF&?3vljx@z1Z!moPXE6Ex9YWYobgH^)Cact>tAR{P^i~Z6x3RhV-rM@MfnwO=n
z+L?{}gpiF`T#3^?tF-<GD3WSRuD3pQC!a}}clq(r1Il(~f+9s7k|vdtZTIQED$P(w
znK(5*OCxvJR(ky?;xTV^HK92|i!yaNFEq$5<n+aqe*a1t(M0l4C<C~JB4jSECh9ZN
zqSkM6H(7$*e=E!S>|{ErHx0L*7I9-!SyMpdv5!@#{m)-dhFUpV=)S~aSFick3&6tR
zROAHC>q`1j1#F-!ghb|C!wrK6vvd}GA&hD7*<1rAZfKp=38s^QT1`Z%iQ!prUhd1+
z)o=lWO@fO^b_aVL(Pg=ii9VTzuVwg7FpmdfFHEnAa(vBUvz@w|<o&zvxShtDY3es?
zO>qWMEmO&0+vwYgU6KlPw3@jmfn@O(s8--@Lh4bW=t|T7H=8AoM>4l33c4i_uz~r?
z^RYmb5h@Z$O7c$V67Zu1MTh=gy=fwc`nP}s`2~D|?9R;+`EzCjl?N&H6jP5Y7cptN
zhkiO~b9)vpbw>%tXG_&u{fGM~OC!5s+>|2u(mI~DKsp4ICeG!Z+IdK`EX6uf^*>Qn
zd{%P%27qu4?BD9{YkGRmR%D+WIQO~SqK!LcBBKI4y9-;@{WHCl&R2&5CRmOW*sG2O
zPW*)|_|P}!2m9VEaAxp^qd9`|6z(81CLO8`ylAZSLfRZi;j>kR62Bv&>-t-t`PWlu
zv5!OAs7Yc3D#dnytEhU-O=QX}3ixP?8JbK42cuYHNYzM;)Z%zM<$3@qjs3%&+Qn)^
z1pm#ro!hk1Y3zgqy**M0^@nkYo3YtG0Y3mBTUi#U7k2I*YZ}1aHrZ}=Cr^~XyZPzj
z<8&1UDYnrfm9`}p-4#DX3k50yt@#A^lUhW7vxQeENdJK`3TaXNcL%x_)yz%ZFGG{A
z7#=Z4fZ$5MH`v%z+#5boR?%)}p^7LsV4wv5$zuKE<1cO~{!3Z{F5uRA%Y@R+?-(7k
zR8+APhq3oX40tUaPgGjTnRG0d;ruA}6k8grSZ4Hl>p;hs&k~s4K_`^&`-MK2hqGi?
z{hcmPtam0@)oi$Wv}Keh<@j!HOY*rjh{Wk@{6@!AV>qL<N>1d(Z{wb1-Xb}UtDWYM
z!^y71v@MhmI3hD4nA`r9zJ3%TM#`usnk(aJcf65_ks>wZ<B~75LLG^%1Y_PQy3q#G
z{oie7@v#2Pw>#;<ymJ8>ceqUFAFCQ00p(t?z2o?%VR}rHdn~aYiBSv95(t9XU!OS9
zM{R;|Lu%|<--OCjAkkp$QQE6Hq3+jUwxc?gLQXiB*6SuFA=vDmcC?vDG+kN_r$Xbb
zS>U>?trBwodez2iG@pUbmCgko^_(-)2Pw!Xd)g}nIH*GR>b^QFbfmulBtW#v)1_l4
z5cVp(qu_8=Txvs_Y#;==rw`p71Hpe0@k`0S&qYTY^f<xs@P4)?2ul;nu9=c&?jg0y
zDq<;hj;p`?crgt4cn@hH$Ct1Tv_I5#$gO8@)NckY@6n*bNq<w#Z1B`3u$2puGmjv`
zuISF5N*g=|BR}&&Cd5sju=aH~9*EO9gpNZFU_>p?u)CWb9kh$ohjf%TN%R?i1TnX9
zDYpTaV+e4v`MUEMHQl{CG(-vl8;Wp4W+t!X&BIZ|1qlSMQW4+<^U}O-2Y|x#J8JzF
zyYMfkJmed;;|0ta<L?sufws$8NAIvNA(}AJC{~h1$Sl~Ui^V7`Ecwbo{V8|;GEmlx
zn~-_Nl*ERt9D$&+lOr)&Zj}U{p*#^*;|N=!i)x2l56G6NS~#%ED2LEmNJwe2QQ1Kv
z$;X)7!igI0J&M-n<{D@KK0KiMQIV<fTSsr^UMxJN7^1s4SlY!Bw=wkEyJkyr+^ho7
zWA_hy?Fiz(ZE)erl33aSs7L-CWzn&xePIx?R9XIU>D&Ss@+`q6{C>HFfgjzxx#)EK
zAYR`<jaxpaaiAlMAr;3CQRl0W7*zlr62uM(S|_K|Ih@9uC<J@u;&_SsgBcd?_xE=}
z+W;{axbo?{1J3FXk|BLddN0h()^ie2-=1Q=V}99DT^%!b2Y~bURXt+%Y1DB=iN2|w
zb3mmEx?TlicjB<Jbsh?5R7rML_9K8ldk3KQ4S`3L??ys5t>YZ?WSax{mcZ7~XiSzI
zCb~F}l3KU~(P4@4VU(&VxrOC3I`+Lc2ucJq@*^=6(}3-Md*$K!iY3<k8Z?~B;$pQ5
zKe&H-Cge$=TAX=AZSErs6W4fU&GpBFwfeAx?)M3*VpVlCHc@NA39;^m+lK30=vw`5
z3y_*G-n=yAbXAPW`;}o=fer~iu~Ww8%WI8P$ru;yZ4{8)E|i-FOQ@a`ouu2pFv)q5
z3)2f|wy_3gL_OruPv=0XW~HK)f~%EFUA*Z~qkFR-z`0bH7A}tK4D*ML9Xkf{9xg$h
z8vmh|<{w6~D1oooenQ+SO-|_jCOX4rpr!A5<rau>_~@cPTXWNh=_Xnrn!{SJwfD6a
ztkv(K`37z&@r_cd8aBwcXw&mN!hDZqqC<^y(H^!22;T77Btxt?ZVMFO<ZTT_EYc@w
z&X{EfK{vJ~G!1h&k09fFk`ve7@mLh~H*{4fY!-K49BI5yHSwy>eQ68;5<_o=jvs+w
zIAMD(#O3f<6DNBm#$6gqeGo73nF}G4Mtl?AAM&t#f3XI-4RFOQK>vyZJFjimmEbp)
zpbj$;WizR#ScmPo>=CCy>L0Ab>3WBg0kcW_bXjtPC*BjP-LbZOj&HufE!y=-Flgs=
zituw)bAsZWRu)+wkGT0ywLU459x>OABv>`i&52Yn{$2k-l*MO?2I0<{-L(p^x_y_)
zr=Rz@OO+=oHXEQkfL@l!@ZR1-YIT1<%E(_6!k;A5X`+GJ#xhS{8bot{D^D@MQu+CL
zYXS%?k=Rk-fM{ohR>sGzWgjJiW@3a_NCJ)u>Xyl=&G=>FuwRgSx4JKa`@+^t{VuUk
zSBoA`AYV^^rIS#(5*2OpF%hV=D-$?HH2~f$Th$pj`kgWJ@WEy|2Ot`1y|5MUrRd66
zI%yUuEIyR_fH45NUqY3lc<67wFBDl1$wJJoEjTZMUDs;P(?UAp6zq3dh?r9MIIoHO
z14mT<oCQ^Afp+$VQ*@M&N5qbLKEv5j)CWZxvkF%-RaM*uJ3(j!gykv&?8W3YE6Dbw
z!T2O?_)9y)3XbVQrfIl+QyUmI5@>RjCYe&*$rNrm$dI}o)`JC$S}eB}-Cddm9}oge
zx!vL)_a%VDA`?{u<7u~EiqRiLK#NCy#jAF)Ml#=AUapX9xhqY)?h-P9W*R){MyTI2
z`&WP<sD8`r3(GYGOZ(WKXP|L{juMqP+*<@^tjvOY^G&jAUX&Mkk!}5yfdbkCl+ad^
zm{D0%@z?o%z`oJdl*$wl?LYrhTPGa7EY!KcNi~gydk?9wEP`;5NgV-q8M-_M3NgpQ
zrUzDG>f1#3o_F7a5|U>~F9RAr9IQ$QtBX^j=m?L5pu;og?;`DS6A8J{q(f}4VOWyq
zG3^yTe2q!?uRx}J*^$o328iTMV1gqWO6w%&yFba_U4~VLLVF{=9PlDt&?=dO0Ho5s
z%`-5132_F@6Wt{-aw0aS>O7_Y#BR*^`zMK?40P@<Yk&8TgdNuSTky*R4E<&F#Cx+C
zPwfHT>z0-OYs~riqB)~~Aszb5WdIgM=(j9pj0})Tt#lx|Z3zWIa#YL>BefXlz#<+Y
zlEn5dG|_r-PoYuM=Ne+6g@l9_^=&mxZ$5<2mzfatp`-2?tgf;7Jn%k2MseZn6zeB!
zs6R_uhl>zv*vT{e1=y$HbLMj798d`lo#CXuyN_^X1VOUQ^^~^9)mauj^HGR@Z<C^_
z`~i%#J0;<`b0OH#v9};OWl3e2E?`R9WjsX8t?8X5G`><sZG32q<r`G`=4Y`~A<XPS
zIQ7qUF-c+$32(+eY;zxnc?st4YUZm+zBkml6V|15C>6iuF(|j@Ot8_YIQl9($tCnB
zx#Iq*z!?&agpl_<(P*x8l0^WI19&Dw)F0$MuYu|a66+OHl~#HLQwor<S8RoqNF%bD
z`32BT?1C#`6)FCDd4|rAZCD!rzP#j1K^3t3bCQYl)+ZR)whi~z&hw2_*1-I{GGi;)
z{}u45JIE1Y#m7}-Sedl~bpG{@%38fz{GxkM69&cl72}#|1cc5CF$dj>OJ|DzK!_OC
zYyz7#T_!p1F9wit3SV)!N2j=po$2h&Cf8U1EL1gb!qJBg^-)rPpv$gJcdM<(Nj(*f
z!N{Joh&|_jst7;9-eL%KwHVUrWgH>b@>eqE!zS<AKyQ5Jci`oZRxV}8(Sls1AWaED
zTn6sVH>ql*S1N6ss9gQqFD;~u)VPFyktO|{=St42({In}yr&`c2aj>X<5p$90ClBH
zJP`Z$n4<W67N(%f_H_C6m8>0*XOOROI#c`(EKLarn=|(^n{ZkLsw!}4R}0ux?_+6o
zGM&8ve;)F*l%c98Y4}v~tHQDRCpXhJbU&o3d!RT!JOKQSdR<MyK~&6?*^|@M2JMfC
z0PSQ@J<n<#0~J_sB3^pn{$_J$pg^!O00;fp4inkvngCoCBh4-326z9WaQ!~UTz?Nm
z{aT5H8vFsoeYPrdQ=X05lgz@i%8T5gRpV_!BDvrcgua!d`u2%{U|T#9vXWe--d4gB
zK-IN1)va{WI|IPA%0FJj#HlivasCn5Bi-8`F@I%|p5`5oOc?+M0Wc!jZ#RaDtbv2U
zB+C-8Rjd9j>LP^}TDh0P|KZQk#4%1~R$5dE(<HStk+w@Z-F}9JnL~31TLTHi%zc!w
z*C&mW9+^7fA_WBU<D0Vm4Qmdc49q+x5JP^_lW}Kmm>D#$|N7ZVrx@Y1%;z=}_4i?A
zHf*q!VEqGzOhYmkch@84tEva2LE;Jtsq0MC)kl5=CR|c?qb)1n8faC4#!Jh{zS#QH
zd=JY(SLlrbgqhQA*OuRpIm_l4mrVmPD!P}(TJ~>i$Ati)ok8u3_>XBC7_%@+xMYYm
z!FaFt7=K6tI4gdG#zpGtOR&Koqvr~$$xeBu34r&=qx*m2=^?@VfFvXDXbjwEk5vEk
zfgR<7dh4?&U?yk^)`<-;muG_HNB<&?-UXsK8Yd?})dfG^eY*E_xF+OGFB64ZxWmm-
zY@yH)#=gH5{Y#nC=Ei}FpYp*c5^VJ$P4fm|<11NLs8`1Z_*qN>0e2PX%SpF-j0)yM
zoo{w3(sHerA1@p9v`i(3oybA86hRL(K3kyWTQ3=`4MHC527?fwzTFj345;6w(#qjs
z`>53)w)^tW!mes4#{6`CVUx$ZeC3147Zm6xghvWit3q)Q8=@(&44>&IIGtoeD5={H
zU@Th=oO3MA4G$pgrM&wFo5r4&mngR@!56n_5beNsYx=6|>{r3t=$7y_2WI5<Fu9*J
ziZV3(xXATK8Gpzwmfb+?OvE;B9$m~c$ws%?AX6B*qk-chhm1hj+b2e^FT5#v0wTQr
zt176IjoA(w)EE)e_`Ph0<58oWceoQf<b_K*kB-CK7ZwsZku;kFu7-*(L4RisC)zmN
zY3>LTO9v@{91^&V(HK2a{R_Z1G9XuYU72__c8oRp)D?gs9oby&Lb4M%s=HW0Fi>#=
z{GbZ9jsn%e;w)EO?duC53C6COUJZyO$FHJCq~R`PdRyWZZ3wncU_V&t?8tV|_{N}6
z#I*c*NjNd`V9iyOe<=t%S%3gCakxa5izGlS&xqo0_sjn#29xL?Kz$Jg(+!vnwO+c_
zeG!Q-d=-OZ`poheGXmo`(KNCr17mtW)6WjoRKN`jvhNw_qU2FxuL4BWWVqvN?hS)o
zXwBT8suLE78UKE`$>(wsBah+7w@Lv!<tdmnKYduBzs`3Tu?vEhT2syGd9E)9`Az`R
z5guv3q$(0m@&+=myd#Q^B6R<Gz(t5ANTXDB{P~|FZdFMVuE$P_%|aG$@X9M!!4Efv
zy1rY=aV~(tXiUlO{41XDtW)?d*o;#HIi7KC<v<;VIvI&o$!YJ*t5}fAFx2WsESg+}
z65ne0H5kJUR@f)=&lD67BW7ghFLj<c0OYI*9F<Ut-g-(XpQ$ZLbFN+ueza9<T-m)O
zNcpCUI3|X=4($gVJOXV>t*t<4*15luDHGM_e+o!lNHJA{Y`r{A`pBuXH#@SM^5ZW}
zL+<+JwP6LN#5NIht#&@{R*eDEr61BPC{${G2g1i}JtbR(U&OQ}Ps}xiLdL*l&?9)E
zmmO=vHk7eLs+I?LavIUix&Ce;W*g@xVQBx9Hk+6agjjmjQvZY8fx;hxoFd7wf^7`6
zY7kN{rHF-$3d;|{E(frQ1^u23=*lO$3bDWl_yYiu%|Lnn$v;Ek%o!)~Uy*l_maWMd
ze~fMGX(Px!4ip6aerIp%SNUA>m?{+D3wI!<GD*63oYnG^*scZ^!~m(ArSjjZ%AW-p
z-wC>BnP+0*X#d35tXyFv2*v1NaU(-frH=<d+s|iG<&8L~8UF%fr#O@QNBv{SwO)S(
z1{*Th&yl00e;Z66Y`>Sm>2*Xc%Bh_BvRmB&SS5F=eXkEDaSo_y=e##=&K}uiR>ZzT
z-5j{6Lpq^Sm_jW*0~K}kj|6*Y6J3Sxy@v<;k5D=ZGeh7T^axXOl~Oc8vy^eb!R`2W
za3%2huGnH*ugU;`deX9pDT6^ksuI->{c<Q$Z_R$~%r`Bv2oA6UB90fF*=i)~cg)#+
z-2tt5z$!B?jXGKkfc}Liy;!UxvqH>NKnxutVD1C$DX8)voP)Ffvn8koI8+sse=8{k
z0=TSnM)W?LQs@m`yq2qsKI;#ZBfC!zAIfP{1M$=0e}G$8NIm@aGQUD-CRO;(jXs9?
z7q7?%VOwMn`$Mec7F|1e<5kqT+m17*qZ_?O>izWtBOL)c%s4r{Yo^j*N&R9H3isPD
zU&N;(9p+G;0w<m*Y-BYGQXT$HKzpY7uS?y4IbuvQeUy2J4^ASv7J)c3>gHH<MmYxw
zfrmh}Ld5%nB^{>#!rAp&&s7*YZ=Vef8fC0iX>=}LwElfOe?5*4fC%P7_zVh?R`aeG
zgJhrvR&K)MdMtVh2W|4RxSdT(goXh1m33RfgxNAQq|FO2%Yt1pve`2z@IC5n7H~xt
z@A7;nYU&X3B^a}c{{5gBAR7eC%<?Lu&RAHID+jeR0xJrc@RUah#=8CoL;@8S0^-0e
z$R;?S<og8AQ6hE^axH^5KWpd*;`x`dq*V@vwtK@|o|<?t%(Vmn^!LKI1%2jhF6e34
zhOVWyPt0A+RHf<t$>a5VcKHLHkYz{>&sHi`BqmY}MO~w1Ca}?49Y0YH0<;Lp<q*bW
zd6a)wPMiRggzW&rC^!f)qZ~Ib<P;WTkmJ)6#*4m7aPVW1$59aMyP#)prE~LK9WZm^
z`7arRBnH_gwyHQY0q~EX2bUt?k<ZFO+U)~S1L8NAD+`y+6GIVo-Pcj=)}K=HPp^bp
zSoWTsEB$;8W<m0Q@5iFh8HeyQ$2z0~fYw{pYX^>`03JibVhV!%el9|wfIIG7dr5&h
z&UzcNBETJJFx%RqSc3kooNi)aY!l%q$W1Qf2(JFiS0n2Y!vdH@0PLa0xolu}0NLnv
zyz@|c6FIbup(?0FY&V&tuhZ&qhl9XF`Tp6w!`O#N!(Qp~?vxu81J!fYzYo1z+yDW6
zA=B1dG}jQ(cvF=8a|%W%n%7#0t=yqg4`tIpzQRFffC}*wVm53MWuPDsapusYp$(#9
zes;i8!jz$}4^}Qg4d6O=r2om&4#7i*kJkG{8b?lX(4nZM)jF#{+k5@z{9x5Dpnvu9
zJHX<xAjD6KuXQ!C_ih11GH5^fND7Q3K>%JE+XO5<0#!ATXq%kLy8>z7&$h&$Kh&By
zhbs9Ga!oqm8mJr=Nlk?mFY@yS&=1F5leNhXU<WM%WB_C|$3lNu>&mCZGYc_Xx_Y*-
zd7E$qE^4&^sY4IT{;Xwtl2NXSi&i0hkRAtcF<-E3XTL>;nqHN$fi@eYmVVnKOiC(r
z8Y7T*-2qAU^3P8<Li5b)%slGO%i7}hdwUo|co=lg9mVwg4}?#F2;G^a_>bxut$G<^
z5Ta4Fmk|dnc0Al~8}ppiHq{LU2s_wwM?=r6pA9JTZs6rp2N*2`j%g%tobB3m%Q;E~
z0K%ht@dI8UCxr|?5$2U6Ta)erV%Si<9>;6cB?5ooK^__?WhL|n9Ri#d8mwmvftQ;#
z2d((R2PYq5i1!bux@#4NbN<=?kT?}S0_a0_70*JMoG%~z*kE}aa&%)-0@M}`p_?J?
zSwM*$S{iF8?UgBXE<pHbq4V^LoKw4RcWEGA?%ngrC6&b#C3q_3G`V2nmh<lI5<edk
zgUDU0BOgv6^1JrZM+_(G?Cts4YYH7=;b}JEDb7RyH+#Z^>%Tnxl%3^ow<^*YMV*=k
z3Gk6)428!g!VAGCF*COeUhdC5zhze-d8^`IHjUMTi#gjQFv%EMt>evgn}{?EhHC&K
zB0gh)U#0g7LS8<EQfXE#n3g>!j*PXM%6iq(Kd}a&!vCI%9BImc`O$kQiVi)MQ1wK^
zxdXLbnX1QmJBhcU|2T;d>uo=00jp<narA9Z;zuI5;w8ZGZvT&Tfb{y)4yEF&FK*0C
zFX(T^!ejk7I_)?Bryj6m<7{G+j3i-zeM^xyiL1BGPR`;x4Ap@!DdO;|D5;m({g>=X
zAY1=E1h=w<`)F*+20ikRS4|cE=59JFm$&{bA)e>qhBMg^TL1<C{&F__<N68|i0^g=
zT|aT*h1up<0L&B4B;zQG=HiGvU-V38m_#bXZ@wH@Ak^`l9Yb!=VT8U{8S*c8QJ^r%
z7&`Ln=ch9lT_;*r;X>e9Dg5_DTbd(K5SEe7F5ff*6Hh?cwgC~N$FIV`;d=k<L^I(P
zK)~Fn|K*S|RFB1-EnsU6zUu+%TrE)DLUW>)+kpf&rG%4x*kA52;}d=H4v?a{H2gcv
z(E>qVo>)xCRD!qmrA=8tKdDuYZOTqBX2#!SO&jS12&Wo)J#z_2?;0d|6)-S6SN5Ra
z{3jI$fx9VWWY<<zK^`zhm~}~M`w{yWUyN2AK}@KTyz^o&oScVr!iP6yb<TH5za@gz
zl7f7&x}5Y0J5=dl?4L_Rp`!xZoEU#+?WTtdA^>#UdIqt{hUzx-YjmF8uJ*%KQvqal
zPx{lcWhu|Xx4PW~`vN2IC;YcycP`-YHy}5$R7DDWONG<)JG7EU<S_)uguWevlcWcz
zm@M`*7<kvN?m~W$<piD$h{NmQpdxuR`J#+JY)N}Q<o3&&hAx!n+Ps5ePz(@bBPl(7
zQwLz8>KWhSEi&rW3GgmFA6zg+KtQK}fg%u?iIp%;cXk6PJBW>Bne%t0|E*RM!d81p
zE}bf1Amxz`=~(yLWanF{(4&9*l&a`EP}YXN^M<a-%s7~f&4Ps^#0+u;0@9OWde`}T
zX=NSrgji9|ygr}lsN__w$6?}yC+PF@X(wc~$WP23IgG+qLUz-g+)Mf%W{9^#oI_?M
zHk_hP*e`|EDG+WRT)D7IL%9g>PvyCU88MQO%JQQjhql6f2<Sh!_39`fE(%cR8G4Xw
z6hXUzj8XC{oJur5e&74|;itU1&hO^A^CU$D&=cNPknY@KN|V;+G_;e1qX}2h>13dn
zf)$p?7%d%nP#RT-vN2b~P(<nM5omHp@nI+B<$d?|NeBh79mL@uu>@XRej)9>(b!af
zNJ`yG)~SMbB!GOfPr+mF=N4)FVYs+m_6QVY>O7ghS!eP(acF#e?~I|LVfv`(JDj?5
z=(}<L@=Y1yJ9O^rICD=>4-j9ml-DYM()@_y8!#G?1jvb8+-bz`<6!`_Dn=1!A7Gz5
zLs5r>lf5?|F(H5o!X1ooSf;1^9RT};)<gM?GDX4tQ>PA`K`@%dH~h;nQIYV6{sh#O
z6To>XEdX1u70BbO{LN>{2~Y$YxNuPz<5mQ3EDe^8>)#AX<)V?ifYo$YBn6Dl64NyM
zuEmrvLRIb;DW`y&l7ilb)DzO{>QJ7Ram+=ZlKXq!9{O2qR)?c46&43j9xK|%5jcpP
zAI&KJw^h>2k~ZXTEOTq_NXu<;D0ya&G=Fb4lIFmeTdozEKY4qjY9bEqh7D7PZ!hi<
zVI_1+puPi|O*8HAg*at7=xpVO;Q2>Vh=bT1#?%7k*yv=u%iY_P9S#2i@;AywO9gkx
zADcr)2AZQu<bYzME|rPHfFu83CJJk!UymmA;;%UzDP;tP|5J@V7ywZLAvh?^&+Qo}
zMF8nE5Ltl?{|8V9$do(*q<?0;db<^_<jF(X5y{D1={-<)*h4iC6ph$8T8VYkiJi?%
zr-Sy@zkBpGn;%QSjqLa7CtwvYe^=qT9L@x9Rq4j1WZ+Xm1OtpP9wG!L1{L$B96&BH
zkv=lX2i0z9tTVy5oJvx}S`KATZ=TUFD8EY&RSdZymq#8fgcB7AwV%jQLe(fPq#nGe
z^Ho94J=m*@?i5_Du#;dR3@aRy3FhxzwRuVxY)T%JC3hu(g(Z5o0p!Rh5;_IR4=PwC
ziW4F>ku|h6T2JM`8ih~$nS#>P+gsSk6x2!(^x$iu;_6ehQj8=CcgZj!eM1CAvy+^{
zt_V`jzqwZRI}m{eOfq+D{ZamfREYOJ*bNA@%37wid?+9{E9MFE9t_)0#RxQUqgq2@
zh($QjR;1h+JSl_3ZCoILHj~#L<L?1-xWnX+;i$=zNm<giKN0>3E!T(J^?sP36dcG+
zR`@ufgn6jPFTSi@&MYSaSG|J#)PmTC`;`1O4o?0p;KQxRA{S~Sfj<mEY$ny=jR=BN
z8i0M#<L}$|In9_#)CH;4FQD`Lwql?Z1ZA}`1@<kGP2g8VVrBQG;F3vrg#JHDIED@t
z2N1J^Z$HIJ34;OrGOj|%%$p}ebur|oa0^gBmhvj6%;+OW?4zJLZA38|1P_)(-uaCb
zEQ`MTl&=&e!f|r>sN^L*haqPsfcncGX1S2sK;s4>=|CX?m-0b$W1_^c1dwsO)ANL^
z+X>w2{}x#>RG^QnFzfc+oI5k4+T-a!Tnc&(i+eb@e5KakPNctr_TyJ<<eMl9xe($J
z*jq3F!dfG?(2VJv_aCH)EK<4*>3%4ib-KQvfR%OeU%3YW%8>F<aa?GmXyEt^U~caR
zZGLnp?KIVp)}ezWT7dxaO<i6`wg}M)FkzWf5-IkCWkw_B-1dQ$!9T+9D8qmk3YNE;
zqo)P2Nhe0&66I#epNQxt0t(0rX*Uk(N5{$00{);)0`lfWl52YrHhkw>>c2B1%?f9R
zj6;!oG+^YDhb+(D4n_0(gYVhcTQV7uZQwizTQs71m~j{}qNai#uueZ&6JRIe(7!U-
znGTsUxxaeL-GmOoX+_)}B-%lqe&Qi$l3-rz7YGv}R*1B+4%~K*F-o5BkA&-;O)XBZ
zz*Hjc$$>^;QNkZkAY_C&1sfz%uiWO0*(#T-+DG*@`O!$YRTPXSc>|+n-bz_29ofGh
zq4ZOI4B+2YlcjUO=%<#bCq<=Z8DzVn|2*+~P1`a##a?sqbt#yh{V}qZnY>H;hUl8C
z^O)%-VnGj%!tMdC;z#`-W3@oWEHUxbDfXBw|2dFgAt~-&672U4L_LV%3`PEU-C)h_
z8+EoP>^I;JvKMZ5-^&{?8bsHJU#&xP_ROX%929^D`DTyUwj{M5Qouy(y#i)tR*73E
z{EEI%^b*2D6^N_jDZ^JyG4FW>Cx1i5sx6`04k{RbU@95e3%<VsAr}HpqEw!t)Iq*p
z&w-MpqX2+yrR~hOKMEBm^yK-`{l!-g89XNZ4~ZJ5L&Qm3A@-Y#xgr^9sDqc6Ap&+b
zrOMRJ+7V}5;hcvW0%~+43M@ZV_RdeXE~O_41Ff}pxKzc+XTGJ%AkVM}r8xr4%~%|?
zt?a+#TuDy__!_Q0xCZuDeeond1I3mw3kV&TVYrG$D%1{-!~SX(;2~fbfFq>SZu*dP
z_E+54<4a-WaM;RwWK51C)M+N=6~O*9A`Q3)(hRc5x(EI8-pC|7D0Gj>0J$5oj?BHi
z$&&V^#ifG_2GMrI*mhSqZvTSApU`@)?4)Opv$H6}(RZ1XgJDap%=CWZ=cWw;HRgXn
zURH{Z2kVaoug=CmIn|=K_cZlvPT)^J1PP};iva4(`@7vN3h2lS5<8j!T{e}eC$01l
zKb2=z@4h7%qj4iLyGGE98{TSjhLKr9Exj>-FMU0zlJ2OCfToU1`&lE<LYS+EfHMR+
z-{RAhQT-|4oKJ%B(?jh~l#gM^)ZgmsJc?TFJ5b?D27>@mz66^OD#u4(h#q3ebx|>$
zC7!F($cVRx#zf91MQkoBqFDTB1D^g-1}WMcz{z$9mq1tF6iMyy36?0e#x2gUOu(@9
znwwLUH`KTf@>#kIxNu~0o!@h;au6hoh7{OS=n>#fMjmh{fgAd+s}i)tZ9Zw*5eyQP
zo3ddarQfQH;b+P}2Df<d;=3-z$rW(&(fetN(^z+zL%;}t;e?qVwW1CcGlZQwr|%*5
zYK{}KNN2PpC>P)!Ewhs<<`f9{7g=62W2oeygka`we$Xx&w5j^3ko$XcKpJPrl~Y{b
z0}jAiQGcK@`T#EBMpO0wsQU77DF6O@GmXqp8e3(lX6%HdEG0F@o_#42r4Xr*Y}pz@
zNM%ilNJ@($*&>BDX%iBqA}y$-K9rRDoqMF``@63Gc&<-9%-r|;{aViJobz%dycM{w
zSy<Wz9^!;g^J7>|oyECbYp9N*V8*RiF#**|<4I32RXHlk6>cSp*mQjP(sw8PV-;$2
zvtKF)|8fL0h#|6gj9!546$bX`z7joy7z(iM-)ZKsMkt~;CQx%!eoH#l`!2Bp(Uk@4
ze|1tm41B#YxBwW5oEOV;h5mcgYzzjqV##kYAy0J%!Vi~JAgq6$B;e(yC`QAh=!({J
zyo-gIrUhR9^vM{WvY5EP+1h@;u{y*fS$T2rnjMEfwx3O(8HQf%{~nDna?+`tU7lI%
z)I)By9>H)9u*1pE(U|6PV&*7iHh~Q%tyr!ZjO!I(!$}sxxXI?1$&V2ndCzZqG162z
zKmXg_Jv>Z(#&UAA_baeLx9P&j9fY0-|H`#Lbg+mG-ZiH4lIFzDILfQW&%M~aj26W8
z%L}{>liCs{7*b2_DYmAw;2Bmiuj3-ZNdw_Laya3_%EP=C)Iay5fO=PL@7nwp>alas
za&zBVfdb(P1&K@3vxST!T>%%SaKIiua#$OLs(ywdbqbL0&XYG<M+ftPuQug%LiYcn
z#|Y&bELyeId!SDLi7gg}dzP!>K+8bgy%)WM(@qGFDU>A=!c>nD?U}qP_{~XOStwa6
zpcYQIYP|V6$~lDzFEwH##)tj+4pv0XDg3ZV=`ZwAM^#Mq1MNrJv90eFMuhiJ32vBk
z=*kDkqhYkPg<TDj3Bk`ga<ZQD_bMsL=gSr2cc{X2vo(G5R?}<nAGGUUJdedNrb#Ui
zG<}s17{KUuo`Us!ISD236_a0&eJ@=n{b$i3U_TX5#eL+%JlmmL3`qU&Gd3}-Rw}U)
zKHz8BE_tjGLXW?GYAn^WfXf)93?K6CZSw#gxMfvKzrQ~9mFN;_ZdNi~z_n8AniY8K
z&iwEhKlR5`-YajCOM$NxVKYE8lAh*pGh92w(Fq>y)#}V&bJ{Ut>2BH?md<oH&RD?6
z)@KNj0rJ`<Z{5~{6Ln+XtpELX5u&h45o^dex?=ZN%oQ&rU|R7erswwh6ah9sSG!Al
zYX06^)F9}TY_+CI7E39D>~4$tfQNin#s&p8xE1#v=7MjrdArkM?t$ro(_`(=vrtB%
z_ep&Fh~mBvtRb7_*`#UHognP~c_A*bboEr?T|mV|9vV*L^cVrQMD(fk+^?RXNrtBA
z5Bk0B?O#oy6P^5x^7#{L&JWAlAc4&1F<G0ujwI5!XL<0)GltJotvL~EMmDEIfPQ3x
zrS4BHREq{n-u|djvGotRVr;tqcP%iyCC(*Vjfw{EvOq7k?oEjX%%s)fqE84ul)&-0
zBj)I`iI_<+<S3$6?zkBpH}J7O`|4GVJBe)8%l!&#XW#1H>BoXR4O8BB5Wj)dOjG=k
zOxE}f&!yCh8AAiW$SPX`ls@Op|JyX4WxD}egk4G}+;|EJrGrwc^|=uX1tD_MymPAQ
z#s-h3ffrUaRXg&KAJfMGrVv~SV1sI{W7y+q&vP4;CC%(zTQQ0?G-Dle1#mHEE5?TJ
z-rmcC24z|EE`u(W({=Q6k>&&8>!IgCZKajbc0Zqy5>0_ED>1n`M#ijK9BQb4aeO6{
zVG;`<wk6~7yOkQ!a}Q()K-*ONA;tz|;g5VuMXP7giI$YUmr@?!`EA<FJs7=o&e}Ze
zUZ2vIT@Tvp5hUoQHNaO-&(+!*X_WVUL7O~bM@yJnY%#rs!j5|efr2)Fb$|x-!zVl`
zg_5$V2J#QRW&RIP2#rUUMg0mzd-Ut%`W8*w(iXIbor=SJ+M-kCdOz(KzC+Ti?VsQo
z-Hblw1@{H`2<d?ARuSr17YRrO_<Y6Y8T^;Fu{$q&S);t!(JZDlba=h*CF1$w$a`GD
z;POlm_AUf#U(9(j1&V4MW`Rl<*LTk;vvFq-Al${d2YCPQs7074V;82_z7geh6bH3G
z&8i5JC%RrVV3{)rFdZ96S}Mj>Yi=K;*UD3NEWGSg#eZ+ed8fPAxVRFf^#BWpPt5e8
zn!?k2&?o(?V|qk@i8O(6V<^5%nqWRFB=I`WbfJ^=A{8eE!z-B>iZ$=-yM0sx1N*hz
zZrr<Z)|`Qs-S_5CLcCP=Sk1Uii7+e2Ekw`c_2x1X$$8A~P7l3Leu;xPD;h1s>bqX$
zQX*&l&?(IpU9m%PDwXB{BbnVuEW&WE3)B4kUg@al6^8e;EN*3vrY<i{aM@qIJ^dI3
zn-g@Q_G#+;@6FxIfNOzB-;(Dyo*5k|V*_D?(c&Z*E46zXvi+|DYistXiJsg*B^<SW
z>O=2ESWEt~pk=cd#S~Prfv65U`(dM7hruEDYY~Snf(>le?HwiW%p4uf4PgUvmlv4+
z0M}x=FIDj`fY#(shRv$h&*+TZSe15ZfyY>3K1+q}-`9b0{&y5O$#J%K=DDkcF^L5b
zG_T~iG<QN?$jxP~2kU|QoVnV(PjFt63;Xc_lmh9&Stq@?p@_r!O6^HVS>e0$5daSe
z{6~1`%K~ABcpYf>&SU?xt&pY=Lxq>S->*wLjl%6el=-%;4Te^|beuD13aSD^6m063
z<XeuON85z$<9h4;#La>^Y&5Am))Hif?Q<tTPzEbi+UU(Rk1RqmN=yy^&$A#Xec9A2
z3`DueZ~c#DD#A`L`1?UW(o;mPcg<Qlh`70VRYGo;SE0C74;%#U;%OPq^mK@Vv51Dt
zsS!uUI}0j;_c*KwX9glkimrOdy*th#DKks3<StmOz+lS5V|Z3HF6}C0%F~YsPr4Vz
zYYh|9I4<<@@1Ok+yp6HB;H>%Y^D*c%iJ5Nm7B^L5?20wq+|C4pG6cMSR6?#`Un(u0
zo283SP$u4)fs_6lesPuD+bBp0c>nNI_RN)7F#{%C;f<gv`47X4XrTFbWI%!iMl_)*
z9X|UP8<X8zkFvkg_l1u_^NKtrdwK4s8DOzC4OiSz3^)Pu(ZcBlukReD>D$@^Xz*Q7
zjH`<{k4_B%dz@}#_ZF0d=M^qDj;|py8*%6~4s$jLV|cQCc~63yVu9!iHR@BBU6!~;
zCs$-`A^rss&JC%*fFko3Y+i~Do&#IHmvd`MOWK=zS)Zf%=9ijVpeDveUz^O!thAsG
zeOvrCu<q^Q^_w>x6qlj=GA5AYLo3J6Y&|vOJ(i(yp90Zfp%HvjT3i*EpaG~Sq^>`r
z;6&b&P@uhcV|w1x*xK9|PNhVn687sVyXA1})Fb#-Fmh?35gG0egqZObRAP?_bcLV#
z;<z#~*0>1K0uXjdwQA_3{tje+8Q^ZI2Ms!CNE<HG4bUexc4*%IKQ0&r3jUx7V;9r2
zXA4i+z&i@y7>Qb_K&q3dMyraN2CZsg5m`L^!dwkzwHg(Cd-)$0x1+vRhb4<N9aen~
z?$&1dyGMC{a1FVKfp~TZbI*2oY_i^Im=I+L!qagvqWUAg7ER~~R6odOpA^-&A5)y&
zLoD0CT=-4ve1VPe{prW&91s(UnXL#W%Vv*E%3c+2Yaq^M-MznGY~hxy+d@+=9t6uZ
zC3S<C#e!whU=7?RT63L<u~H4aXH1v>`Jgbu$j<qJGN%at4mg86&a*N{(tkn|RkX=R
zUDt6TIRcn@+2-mvr^QU7`;><tPa4s@6Nxi=1~UaxOTd1eY8PzW!6hSLBW||Gd&Jyy
z(C{iiJCtqZmpe7K{wp>fi{RMaMh|g8?|KpOLc~q;f!h<(G(%F?kLk1K=R`LbzCY#O
z7@L2FK=d_>oBOim%x81Yx42P0IwicA8rMiH-2FGri(sQ23|ph_>2Ihm`h_22RTOs{
zgS$MxR2EKovqT$Gs~0s9+d`u3M7UEMMkF-{d#lieT;2w_Zio;HZ%&9iqRyNlsdfJ}
zCW0{_4@1T6p%N{%EC2>QF9Qi6c+dEN0{7z(HzI=F1NA-<ke}gETspW_Y*=Y~f1i({
zK+K02V%6OO<;2y0uTGxG7-rL}#`|UL%Yce<Kfbs*%PbRNN#b!Y2(>UmYIcfjeLvW%
z%|sLr!3@s$GSGJC3hv`gxkTwMc>`*-@j<Dhm&W~rxOalXF?U11LU56m>xc`vsL<|C
z9_o2dqKNHHMer*6N~@eIKY1utrf#6tbr_pb2sY;AlQm~;ybBQKh#*A5G13z278kR}
zEU8410aE<zqpP|ST)fgtt(5y>SR888;xlOOPnTAz0UjYRFm7mwPL|GpGK@HU91=p*
z-A~P9EU2Jg9#Y&Z!)B*riVi$&U-up`8C=iW$?cHvuM-g-j1HUp5<C=)tTcDNHtfrf
zQ+fUsBof4_>B2T;OKZn5j>)`oGij3;+n1dpW?IQOb`>-t5S?!lznimwS{?F|m}_7O
zR8$N9Jnk`MY@tDS8i9)pW*4yq43wosg3f%7#uV*%5DicsFW(|`OLrLew>hQ-&Ezdk
z>yjV9omIjuP^`Wkn)i9Y;ct1spfiWD?Rx@co6x%GrBEDC6Gy~yq@4Tp`H@4wk}N0B
zrr4BYbbKmMQMHaO2|3lOOk9&>Ie!o=H}1S8cG^Itd-smg$!FE0O~%=-8*UW%T-|pN
z&N(K5=WR|jVx@}g6<j_YX<3bpAmaK@%n1AT@$79>xJym^AS*Iv$5VFG?!?rFNEc5D
z<4hAc$$Vc2FP{W|QwP)#Zm=4Q<<vG6-<NCN_*&#vtLXuJwr;~_fEabaDO(d{Uf_!<
zF?(L_1}OFggX+_T+EdhTq!zSVK@T*awx4(OL+~FOz&tWZRPP}5*|`{*nCP$73`J7K
zw=G5oCF)IWw@m+Zx_vsCSEsCPiXui&pr>rSMiv*J`Q5XN`xs^@<hyDB^Vihumh(P8
zBb~m5>vpmdX2{1>eQSw85G9kZ$Oqtb=>^n6J2a;7qzl}F7d$yf*eIhtihNW7l6@iQ
zgwCEtR=LD_YNa%xhEmfMftcEtCse4x<roCRqM>TDAv8({T2-E&wSSKqeOPMBU;}{d
z&<B59d&twTI-PFlWMT%^Y}1NCr@bGx?+}cD&113Z$LIjs-9^@7t^3!fl86Ii2RO66
zhhTc&wmM|Py%=BOh!QywRH%Hu31X4P)>(w7{3%w#^Xu$Sxv(r;TT=QP9HC0-gH$2W
z#A?PFeUN-ymLxw;9|kJ30i5vFnuKnt_8$qlQ+61QkJ^T-owqL<HgjV{luQe-ee<A;
zYl{7gzRkEm)BmytInpmrweBcl@-6Nqct!0u&&EMXSnJ%f`@@B*U#FSl+`P^@GMTWK
z7FYiwHmgVnZ<|=L+xUP?3|~SLLZ#_!O4T}-<Y0$~QUKS(zzru<8&Hd_n}TQ=(`f`C
zc0D;YIu8L8aWZlS+pURC<&Tcx71)56bNWJU)(d3=nb?lle237lW>vfGtX6Q%Y8D79
z+7&iozvGUixStQV-IyGJAbqG4E)!x4eY{>lGoO(_#_vt~a9i0V7F!lO*991k!B^nI
z+`Ta~)G=NyjsBoFhznS;*)S!J;d|=&JEAAWHUq$>s9@m}WBZ*xMUpk9Hia@oVrH&Q
z%)n4~3(?GC7J+p{R&9fXl%sm*NihR?d{l;C;4x)S+m+wDcX1)Jv2;%OdsnoLA+I9e
zTkAdM=T=2A<^;Qj|G_I@@p2EA_bRrl0P)^=42;3lWNQ4azL=gZ`lVXJ!aEI1w-A%?
zX|q-U{N~LOIzXE|#Ww|`QP~mA;O%=GoD8CQNUwwmt3d%6D-Xtn>TcJuq!Q+6?=+6B
z`g<AV^9!3nE(Xuw(QkoO=MVUDPa2&{M#F=l=6bc-k7X#L6j#TIIFEkQk+#TNskddG
z(3==y3w)n=i2(b}nV$8<PTLlgU^M`iO<cPIjOW>1F~4b380rbxo}oBBd|o7Pspd@@
ziqzOdPUE{;v+GFDbNr$z=rfk($j3rK`yWUMsRKko{olI>2I^~OkjN(#7zz&vZ|C>{
z<LL6K)B=1Q(uqa<dJe#SPw8A9YgG972AYlQfSrVW;$_&e78z<(pNMB$i(Gd-(4>7y
zUjGLim?vj!4WbL#A7{-EX`dvZu<i>3DY#xB^K7_FU>mB3o?YajsvxMp@Cp9A(s<W?
zQI0&ND`LjJ6)g|ix``M=+9T}$`RX$@i3wxdJ%W4}CX;|W&D;7e;I)h;y8&f@R&SD)
zk-A0~XOEVPSk`2i&fCD-d2Y=@7$X(m61^pp*#P%xy~pK6e_g2<`4rKc)~kwz-oC9W
zGNh8`QGE;;9aLhHPRPR#JbU21Ig?3ed$U0$dye9Bhi|u-8Q9A_lwW47D%TegmfCEl
zB>yD#17`AW$tKk`$NA9zu1UBjn=83B$3|4<Lvvq33wm{3QJEMxfkm_{$ofP7@%~ZX
zs<ZwXu0H^_+njz&XaaA!%O_5G0j$eP4_p4}Zw(pE=(+YwSsg_S@3a$Is}bL&d(>7E
ztDe|>u{0~HCH3O_tk>tD3e&a_x$VTDhmiplUtX+S6F4HI@KI&*119Vwh*~u$-QF4x
zgOMdS%LtB|i>iBKZa&JCiTNnfu?0bqV7`eNEDCj@ODy&FhhkE8rO&N7Ytr`Kd@dX4
z&^vE_<j7-~g2(GLZfCm}&F5YtwzV6O2rz%DHYrD^+$Hk%A6mcA_rmnw8{D7TZ&P#^
zkL!*lDyFv;!+%U>64Vu${4;mt7ban;&^>NeBjTB9PCbgd(XkL{BQVgB{&F~Rf!+z6
z_YrW5yl<>FJp7;*MOb2p<mr<rLPA{Whz>#z`U@z>Y#hI8T?4V_?d7@6b1ugkTk3Hu
z*4X2mO4ao@jn;J&TC$c|M@ejk1wpiTfwht3W5_ft1rN}*L>d7nax>O(&~P7n5EvPq
zA{tnL=Kr*h;^(Vc%m2P)RdN{hDqt(onV%8Ltv2O7GZ6LU*=%hb3juhFKqEx92X-x0
zqo&?&PJzp%neO7<jTo$E4iGL4s35I2%*x^YJC710Y3CS&fs4G<w{X-3qDLJjDMYvU
zrO9tJvLeU9<K47`c|T_vp=(QB)}rU=H>V+cRylNBhV^y-)jYDR|4SRcDVP8DBMNQ1
zW$DB=0EgX9B{+m2-@e|U_)@~tFhSoW_RSRr)eRX6>{RhZcMI2EL%)yl#c#lHmr<7#
z9ybsFTXOE9_L~{6@;N;#eA7%=E0PU+;w636cM!k}@7jvZI<cik>D@E$HLsHSQ{sPL
z<Rph>A>^NTlvuTnKy*Z6<)qhBv?9&BuL*VGy$q$3ip${`p8T~~M9bW+bl#_RUpx${
z6OY!GxzeTI*x!rIf7#kU&A-u_I+J_1dB|r?dvHXb_-!jRxb+&Vv^FURTRC#OTY?4I
zeba4-95k+&(&0$Uq-{b;2MLq9$B^-s$=hk24@k*3WAE^b-4m;Ly$Sa%5Y$?|MZz$w
zaWxNKDz#Zv1_Ix!2=D!%b1cm$K42&9c-HP#lF}Tz7c=GFDnI_K5<o9?@qOLbWk*BY
zNAz0IaSTr*eV1dCtTL4z%yK?;Snz}f6YRP7_?|*b39Ovc-ri^kD;w5w5d<xMo5K$s
z-v@9*e2*8I@TsV^8T+cg5;URor4ff@VxlV99}{|Z>Q{rDMqmuGy>@*lbL2)dL@KL1
z=Hd<gph>p<)gGbGOlr^x|F+#ezOIxgt)gXmqV^DyBJ`;MkP;Q8ltmhe#Z?Atev9vr
zzD#s(&|F`n^@4r>?8+{$(h-wMOxjeJjLg-Ze!O6?=$2fX=&dHvgn1^ED~zUC3}~rJ
z=_lua40#v;bpLacS6}h8a3wd7eA@cgjH1bj?CjB9<+st0Pw)9sa3<YHdk68g%V>&M
z2QdZNR#bbheXY0)(?LEK?D=OF#F`Rr%@Zw?Pv-+IcS!|^xn5J{U_J(1aDh&Jx^=Js
zb%iN<^R&la$vV;)RCoiGy{)iYjk<iI-QF5EdE14&c43c4XsyXtTI}{KOiY$T%jy1e
z$_U^MDukjNMg-cDVod3N>;3(Cu5p(6Bi|eDY@%|^F-*?we$I7CBbyLH3-GbDDhz_(
zSL|d?{-iQmlYJ<-I?!xd1<%@PeHE<2(P^SpE3+k*Nqc*)XthhV4j&y9d$6v-XGz4E
z|5IM(*>&{p2Kh(NZ+vRVBp749-PNGP)`E`Att3f*MEo}ft!w)FC*lN?BCOC8wSN_W
zU98QKeYuilf=}lE!T_A!o!Cml=Ph9)My4m2bFcZDo$Ku;6-t-PdQd54{=EKSQ4U*B
zDe>`wr$U=hJS(-|aV!c)$ky5(3R#!j`Ww?B?^U#6bQRSqD1RzEE@JmozW3Xqzwjd1
z)6nW}1R*GB=o)!Fp^>69$lw-e9CGqp{y4sfM=7uRT<>CL^}^)tV3E9_gKUJu?nPx0
z$xuSiqgL2XcHS7XVGan`qM)6sbp1>LU;iSx3G~t{Zyuzch4<+2;M;RRP}f>i=I<~d
zW}z=T(q2qcn8Ia~>QQ2rH_g2g1`a&{r=zyn{;r6;Yp8FUGtXl**pXT5v?G%w#@@`v
z^m5hgMH{y7>bbarE9%0bb{0)BNga?nty?maf0FTLXT?Ae0(9yN;~W0k)jza?56LtF
zN}ne|oNO%MYoZpl@G@Jio1>qiGMWDx4W#C+sc{KOy64&(d_%n_=_`1tyyw_ZjddGs
z{jNKIgcN@>HvffJ*rh~^+hz*68CO??>gSn{>|>2lV@DrKUOT$kka*32*+HpB3*Uud
zh#sH3wMgz6L+V}Kr+N#E#f9u}O4t_L<lq<*Fd%2<`?%m1CIk0B$(X%wozgGyk%zt(
z{QUZX??nr~lj(fY_u8vS9q1z!N)}wt&AeZwa(!kW!EVV^!-^Cm!wT*dcOehuIvzKE
z*C*aBbYdAd_Z7~I28*c<2RC4W)onRqBXGI*V~?EvWpCkkHw2pMn)n++OjQ@UbVAyv
zgcjhODF0uuA4vF5COZyj;NF|Qpuk+b!(LAL<Y+t~Ee}!B1@qc)7a(AHw%LV4K?a-I
zHdKecm`PZ!g$6d%^Z5MH{e(p7AAC?v!CWaG>1L(x1Z`Vd-QjsvUWa7ZD98FLX=@Ka
z10`nCj#9uQzhxaP8?0_8PFVvE&mx5|%_TRXP9gv4kB8b35vJ+N4=7+t+g|w9teI|0
z-An65hV<qC3HMHwgeB1r@sC`{W!;CtEFKj5+)3nfB7ovQ0?dg5q~6z|upu36h_3m<
zej*dL%cPqYG(Nj0)JxFjHuq4D4x>@@_R;ofH(h_e_y}8n_aXQLX17}bAk+cUT6o-O
zDp{6Lt4pfZoO8IFp_E?IZUE?j(M9QgzyABN;~xbl{5Q|FAei;fQSm;$8B0v&=w$Mc
z_VT46FX{IM42a0kV?ajnYoGlJH$>@d%?q$Q%cNyTiw(g+dG&I1;}E82(gz}=OgGE}
zM_{p>n10h62cfjyKPT$r0Zwg1C38-Zr7#&~(0vhGXlgJ1e721Xv#@B@ue|ljoX1~s
z1QiN`@o|REZN{x7oEurzv}Uj%)JWeSeu(ONp*0n$R2OP*ZS_DG9nYo(0T6M&Wf)7O
zq=2R3``CR6{wTMBJBdfP2k9pS680q+;CIOh3*!#|=yX5SW&3GC4MY;4IS`@=ZuE=h
z5z;}iM%X^_*Ot^lWJID;GB^>z1U_c_8@@TP1fuq{Tq<sCVfR#ic<5XG77CR@q@1s#
z$Cd~ARGUcjo!oW39jgw0!KZ$+6!ou5MJTNsJso<7%Lat{^*9~3?<`{lzp2mYPF_#j
zVEfNEccffJ)Xu^B(O_TJJfRF;rll+Wrpe<x&l&1AU-(x?<=>4&#&{dyHj~iiEX$Qz
z8PQ2_o4AFVKC={M4e$xQf(wT=2@^F!d+?ava&Z^0D~w|j=2I)q8a>Y))j6jsk2JhC
z;TW)@gL4;ehPVIYK%-*};U*La0M`Ao_qpqC-@M2{%OOu-$fElEft8MuqjqdJJkZrv
zWvIe30voSHGi-p{!pYRgVFI?vE?$pV{fE-Hk1<7%0?PdopLWICln6TruQD;Ye}d_p
zQ6=T4(j_f<TyW`~l6xyG(&%K9*gzoB49gXVQ;Qe%-8pn9RGDxxC1$aSPS+x?oVGs1
z2O9BYF-oQ2nw^J#I6%EwvCkBDv&mogM}k|Q@ySeAc7DXL{!#w&ne3nQN2;Z38=(k2
ze>P9UGs!|A`#U%kN}FYWGObKC6J~bv`h5R6|K=sbK*6iDUR1*u@6_lTG~|jcwT`XF
z!rg_|!K!@y*K$mk2*$+|u~Mz3Us}>iXXvVy)ZY?&8EGB0Tkm(-Y*3AKy>f#NkAf=K
z2ZTusprDTn^ULVADo?w&fE2GBqynh6@R=q4{AD^X^V7&v-O31|*MeHOCj=uu-1P}|
z=^E&c6@P&h+JdltCKnJT9wh`4k(WG~3+eW=52S^uoOQZX;NpHVxuO~g0sOyAm30AN
zK0Y3?KE@;DfBosucRiND@EM6vk(k$J^Wb@>q5R3}9P}Vwv7$HhS%Q4vs$2oN;v#Aj
z4BvF5_HGunC7M={;`V=4$)2RDs?b|L$86`RoaK6gr@S8k(VXzx7bSjNQmRQU+Hh$0
z&V_nA(p*P$q?=0dR~uz(s{&VoRv9o}zQ1`!dhnsHuy`PxwQCJ+R{DH;mYyD6v|;U0
zRIhX3%RDG{eszc7zpwYp8ecDHW9UT5GKt&ftVac5(sf|Urwbh?D5M0Oe7OTpNW$;E
z!}L_bI^SZ)(F`=XTW)dqqqUiemnu#?mCUb{2jecE68N4lf+QN&-CsR^L=f~dOucNM
z?Ij2az2s@j72B<7q!78LiRPk0>q)FJnojSm2|BMJ@tZ@MM#TQmeZi3TDW|r=Agph`
zrS6>ArZpIWU95^dA3w_ziVLW?S{=+D8D^-iU;DfQOlzUfMj*LpgLrRlSEL1`4ZA6P
zuO2x2*GK}09r-CZj9ujc8$&M$Go2_S$wjUzJA*n_(s;cN?<Uqa4^2;~Rz<u%D|Iy~
zQ1m-pQ(rdUZLKdy<HD}#J4G)(#8)N8D7R%onCU_>=B)3QFE$}spP%sNpQVj(1$GQ?
zh>rL10P>JyID$0(nD5RT`5PA$O85MU;;nytO;s{sMT6L2OA-Q(iOoUE?kp2Q&xbk0
z;`0N~Rm8TD2NQbk#-yB9qY6D+t$Jq8g#;Ivks$)NCw;Nn7qGbw#)-kEUsQ>W1lQ(t
zLixu{S731#udjO8xK)v9Pi(dU>L;5BK)Ctd+v8~j$135;xl6~-md{yBI#`~HpcwVo
zcLW718ri<pRUax3H6?8;e4vKQkO)&AhOiF6Fw0N$6{n}ajP3C?3x0oFp2qa2lp9Ow
z<l5?3(wb)8da-qGjILky>w;axN4Ugd45jTvhl*-4u<_Ep?fKfv<N6;<)_z;4_~k3j
z+HBLbPy3+Er8EjbbOC(q^2`+%(9zBD0o#spMW!i_@v@zI+S;%ho%xHWqt)_q*5OH$
z34xSH_usfR(?<K{6OX(zNfHXO`8Ln_kN0Y&FNxml=<_Hk*r3K-+{Ye}9g^}dsDhm#
z8$wz4;<EHtpjZ_~1Wrh}4b=<zSKX+2HC{hcw!pu~lIpAGX+b4Th#Hqv{`%=$<vTAj
z+&O+PCn`{^rT(^XjBd#R-pOh)&4fjY>8{rdV<D7b_2jyOkDoA_bCAu!O)&0SDffUF
z-@f}fy2qGuNFau6aLRQyWlnK2p9T%~FIH-Qd?Td;Fr9fJ%fEBldbt3#sihO*M)i@W
z5L`}fVXnCt4nG6EdJLIvhSg&61gd4bn!v5+6CH_07GoKN>;l3pxsA13gRsQ3mb=9e
z_P_pOsA^?`d+NgmEvDiv!h&XIov!xUM7lS5wXf~f&@W(7-!on4{t=EnpjS%t3!RAg
zO09l11dVh$U-YEyHpr{iyuNhFPw(xbT$d#)BwAw#CET2yQW#aRRb{UU@T3c#sKG41
zolxK2slH3!CTwtIr#_u;$BhUqF-xNu4juqF+z|Ik#X4-qc@u6XClM;TpQ}Gy8yq(=
zSelr|sf_2!Vx2DkKyPA{cWO<8<eWttq@E@Xg9oShCN*e3WXMBR>velFe`w}UOFT1g
zyGTs<^ELkK#lMjC#1$7X8C4@_MHLM*H*9=cck212UyY*5YF|>STLiSDQ-17;3oo0?
zl_@NhV|*rs*sj0!Cv5kBKMeAvgdhsq;*@cy_rFK0stQU_;wv_%pDk3OVzEQCSfpaV
zY9t^nja5f{!JEm<O8mzQijrX&`d)uvS~mLlw!n9nxR~UZN?HN0amth*Ifs-a4#Md{
z-L0Od%IghYh9wo6v10+Ql%1ldcZFim|7@#F+RPcyCwa`YOIi`xRzTY;*}bV?cN(Tu
z8=eLqDYWvQY_;?i7<&wj?pLX<2>WLLyT)AY%FRE0Uv;F%v}D-ER~UsJ(?>g0wKE77
zkpj5I^E7=j{(Z*ZO#l8CrqRnKk=>52=M9meJ<U6QJKC>l5X0<l^+MxC#G*xCzT@AX
z3HP2oF;|Bika;+<sa<f~d2OjBRhC^k3#~z;WqstuaSXR!P`C8X*ZJ!kBQuP_1mPJ0
zYI9{a=uF*&rfB3coeH<T-2#UNVjhH|Lw|ex^%fNe!5e^?*G-#>O4i({q-8+JUsr8f
zX*)wRZ8iK7f|GJ)ft&7qtB@Z6Dq-Z(ec8hcK{#{$UMWNJkn`c%gn2hFQJ%;4jB_#;
zI;tgzaEL;EDJ3J>)Y`G5^?;W5->D0IGtr-3CrZdq&ph3Mt_NWrRNTL(p)aB5Mlv-_
z<Y9^m^{ht7v+vpgZ%jW7Lejh)1HzV#PqzQHTc%5phuDYhRv9nQUbU3cxR}1f2kx6O
z2i~4Le{nfD_r*sXdO--PfwWG~OUW{(D1LJfwR-=Pp^#Pn(^4C8>#84atod*HPI3$M
z{t59Mi*bOZmk$qxyy7C#DD#Uc<x<5@m9_m#U$(&*Ua0s>vrsQoY#O;@;a>R1Yw!Cb
zDuS=5p_#Gm+52^Li!~zUdrLw-zg!F0B=0`4-7x*hlwwmOhjUfgl5cj|45J>(kCioB
zQ_!j$Z)7g@ZmDnEOI8o{QJj8|^K)Lv^`Fs_38I3fH2vcnpI7c!(!bLBq*vG^gRv*-
z&fci^mh(vu392LgG6l;9uo!ni=kQyhU&uysn(K3Ok4MK8e5fAy8j&fa4=yX5M5$78
zNBwuL^<rEq+%TS~aNHZ~(sJdhZb1iA(@;Waza|lP18Vip-tM=zX4uB>mp{gtZ|HUZ
zh;5;%&?q5{KulJsJ>QhbU)M9R3HIB^CE%5X5A@=5RL^wjxM|zjEr$zGMZd_ey}X&*
zns_5#eaU~d+hB>(3<}YnEvFN^MI`f$ci%1NI<OD;4>(+p^_wRI;p3YaV%%CZKHPS~
zKztQsd+RHF*3@<YmdF3_v@N<Ye=@2wQkazPtaabzv<)(P3zV4ornI&qYuJ!(I0Tnd
zGc(8K-?2C4%ZOXAoZhJDG{gFdgM5*TwsS10Lp!uFf)Q%l$XRK82HavK+wk{6*MFN0
z#!k8>0CNai+;%OXZ5e6`%+4!+?swZ4BdsXIhTX5l`aB)!*c<5Z*Z6&f5(KK1LjA*!
zR^w6pBW*)JKBQ6b!6z?TdFOzK2o?!8!#&Xc?P7PNk@=bcE#Go4@=(YxXv*Gwm8_07
z?M*pCe+Zh(r)B)FDfvtATIEO~4O9_!MvM1!6z_}{N``Vi%2tdG4<)l_g`?EJ4v>GE
zo?_k1aa9?%w+sE_-V@%o0c-9;Q1N62To9msSvxtc<2Ll)Jv$=rK2VIT_T{H7MdgJ{
ztMt*Z2!y+awnP^{Jx$SRCajCfF!cS6{Zz|s_MhAKvxIEd>e`xH5F2O~vX`2WAKSXR
zNO-l3X+mM3>E6O&o0!<v7cW#9W3Hc<e6Bh^dOuR9`+bhCyxQ*;&G0SXK8_w+vg^^y
z^B$e!U$TsAsajghmVj+pt27R#gLc7^^c(H_oc^$jMoZzLrEs25d)90OXaXMp-FjR#
z+w*HxyUI&ANz~<ZLNMR(#1oEtuZ4MaEnDvNiAb!wyal5^=IVK?{UDci!G?|Pb_|j*
z@h7@ogwypK@{a)5Y&wNE4n5aVhOA{aKwHb_LaYZfoP<0&molkI9R6(GjkP4BNBdJ*
z>eT)_-46V6b8qx&nsF$m9Lg`qdp`o#{~#Se@I&C{wqTKX{YA@YDkg5-726znO8#uj
zBKvf1DLe~CEC!%ymg@!Z6(qn;Ju{5T<pw2aY<Sbw6D+bU_YX-vKeyMd(!}@x&n!;X
zWfjsRYtcN4w$s{;-)G!kJXET6kX<iP@T)=2`nZb+KmI=lW!=K=k4o03j@u~_U(H`N
z8YMa=;$9HIu=64KfX95gB_H)PE0zHTLL7sjUy?$N;_!h6@{<5XgYcc2M{nFm=00W?
zht~fppF$Hh1YNp*TF={JvmOw#aFpBK_s`SSvZP8RM0(DP%YVgF+17Mho<ggh9(Z;4
zj(x|c$&-ZPZ>O{KHG^Vd_1KSRYvJVA>fs^hfiQylO*#oHb7#KwAGPLuEGuzKWyNAD
zx=fRz|69wA%`sZ7^(bfA-ke*=NLl4JQ}2o+RK5GZ(=H36evL`n`!9>8iO9QLi&$p>
z_lo!%*P<$dFQ$#AU&P1R+d~(LxwfFPmQP9@=Ov3mnC6<a#xfKYxd$kn3h2*NAy5AS
zPy4Mnu`aXa5;aat`X=0iui;67{DJ~l9b`~4&{i}@#mVM(EOcp)(Q^WNH*x0v1e`*a
z8@1_?nwrVR&U*E#-)rbG!(fW_?`#YuhI=pAUDqfh+|R;&G&^b@f4>p@okKjC;WSQE
zD_+#Sw%v0l&+{!Bf`j3sEZ8SjqIqAswbd!2vsZFHSqR2U((k8VlP0sVPiKMJG*&Lp
z9S42-sTrg_rekfbdrUE8``mry@_@Q;8a@+oIj+<Cc@4)K$B$LIj{gDmFp#7^qZX(u
zu6gObv1JciX#EPV*GcSE_hQ0#AP8f7b}7SCFp**kHW&9JuN=t9C4LkSjgnOlDk1!7
z$6ntOaG&cxKX0+SaWt7XMpyazhC0L;>g5#k{^E(&{#B$kbl+Fkc8q*(r&CVwJydV)
z_n&>l8Wuw;eE;4b551+$Nw02Z{i)rwZNi=9ThoK{8yMJVq7&F}v+dyFll*+7iNX)s
zES(zuA*EnV<}Xww!#Tpi<58wEyg1bRgj%d?F1)%bSi8t!A93%bt$I)^0dCtp(w`cr
ziOnU16ymfDuRWs?j9?&6;&Ux6*U;%PDLA+FF3X)9?`CU$fcH+2^QI&ITH8D6#b1VN
zCa;{DZ2sPXpp-fRTN_V04lm_e1LolJONVz>14j8Y!l+g3%$R(4!GoPoV%?ne+W-Hz
zx?|$qsNKf-#>5>~WBTLbLFhRk9Y7G4P|bDcuUva6UDixG8v)1B<2wIEetmixAfk>;
zEp_)hS0`Wt5x_%yQR2_=LQ>t~Q+jWHh1mxCT9UDC`z!#44P6PMO0AeM?RwL+=>ch%
zL$6GJUf{%Z>PA2P75A9(6ju+;gr&Q;;wM}e_P>AZX-P)sGU8o!?ytvTK{s2AE>*m8
zYp&z7c}n$9KcbrweL@yV=xL-=m@_^&fRgSZayjHk$;FM=)l^0CkcmI_dnhC1g5xj8
zKPP?c%BFz!8RP$)V8RZktPxh4jY%f{ZH_)`)qsCatG>cFyOxXcu^91D`!NVJNfb8H
zQo?7@$z*|~Wu!SFoMKGGEc~gI83q!DU+NrxVEYn`?OM9Fcx2hf0bd+o#qu~9`g>7h
zeocUtX#KuvIXt2;mLMi$%Q%x!iib9q$TORM%1i#dkf^T+T<8<V_F~V*{@0}oz+|Ze
zk-U}Nu$W?DT@(4V?4u7mbbRFOuf;|OB-U`lrl5TOg5U{Z|I%iM1H?DYvFriBEw1gG
z)1Q!C1rRbpQu|_gwLVAN0*(-d_H!M6ISVz2#&B*0-f`5YSH^SJ5x)9Cr-lH@115cg
z0o~$xcP*!0qAE%6KqRcKBc3k{zswtrC<{JQFe+G$XF%8Iy}B{1lscQ=O4==~b=16m
zgU|AYShpJSZ#5(-Rb1xpl#v_0FuMOf{?(NDiT+yAI4;=D=^B^}Q6iiOTTfdqR2>qs
z5EJ?(k)V#<7;v4F+<xiEOQ<}*`_H!ARN0WXF?;FOLQgmWEI*l1ZvQu$k`@SaZ{ujM
zTxVt`4Vc<V5R6bkrt7~xMj=%b^O<rR|GA{L^otwG-&!-|Dd8^`9Xw2Nv1DD}{pshk
zpi$Mruf$6n{zD_IQ=nX4wBf3{MCoBY?AOI8=l7q9q4Dcpj7F43?`Bc_Blgnt!;Gmh
z+swa-GZ!r{j{vU|6FwoGe~0WnWn}4EYrn`SX0W$D8P6N5MLN;zdTv*;D@44oM<yza
zi}{aw<apqOaA{WAQfB9esGa4qJwD{|2ix$V&igSl6^OjPM+8U#S)*QhR`Kv$4G(IP
z^1agn2n%d56DCLvZ`aT*ObN=6@>SA;dQd{M1^eW=0Tc%+iF5q+;J}5Faj#W}<p_Q-
z)*S%?b~2_uJ&HLx6e!ltiaKeW0kfxzNZ|wKA~AJCb4<+17v=L`UHumMsb+k(-EQJ_
zjphB(pLq!}0OH7XZV?zX5K~Oc5Pqs*Kjtm{2a~8PXBo1njR`_BZ-jjX_32a~X+I#N
z^861#9%ij8;0F6I0seoljF(pL!1J>KIm&apV&e#qVR`ez7WiD5JTVtgf2%wj$l@%0
z#x6Zi>xD@~V^)c)zJ4^$H*3|$ak-xD7tYQjzOPOqH!M6k{~cI-60qCMZ1hFjYu0fD
zBG5v`aULy{X2d-P+?M>?bv>{Ca1u6B?MGZ}mZvw|jB~iv%8uqn2o5E}h*dKG?{k|i
zTM8gQVhG+YFXpZ<G)+bNS_yv)q-tmXa;dac)c$?0*D1ets{R?>>L?N;fAMqdAJa=-
zcyn;nm{l7^qeQw%{Y`0E6h!qZSfwq~fuINGPsze^N#)h+)Ylk}0p>z>?p??@txUzw
z-?>AKVqFLes}dN|mFF!FI)~yla4ntj_l)5)TDy;qbU032UT|LMh~~E0Wg*hjpU$0|
zq~2>@jh1_@q{5v(NAmb4yxQv<+TZ6se-g{M^_}%*^y#N_agY2IbY)3ilb=v${Wc#H
zB8DAga$@*UIf@0rcQw|`0e9(CT#OREVV{`L+!%WE#yOicVECW9YERU%Ex2KT5hw7p
z3*)9jm|+h0#O=5lzGc>ISq>!)=bW8w=)N6En-gLRPMWHYgW%l~L<8K#sJb(cbgMqR
zu3a)iXXW+Wk151UUB3<Cqs!S|mm4OgM<Y28xcq_r+Ip}~Ftf=$ky}zeY|n`Cq7qx(
zA~m;ojZFIrYO-bS@`^J^jyWLXHOWsUp9CGmt<21~2_g697aQ|zppGT=LjqhaLzo~T
z#F+40d=lUMg~H)Pb&i@dc(T6r%6YSgnHIrDWZ9mL<fY89tVQ9(O(b0x>O?F;-+Oal
zy49;Dp8aPtUa5Ka9h?wKsHao>elyZ!&B`F;!OZ7gL&CA65ZnE)Jf{k`oZol<$Q=D6
zhxI04bMU+W<RS2S){8TRWs#s(ZfcB0cwaO5V7|q${&yAeKhOKEB7#AwYBbIagZ2>|
zA){&c&)PVw(^B##`|o?&D<M>p;Kr{HF|C>beW}0d-uoygh|6<8vGNzLRP+BH2$fV9
zcBZIGXdwN)b5cE0pPPD%8g*Due_&BW&pMwRDGLToou8Rf{@`hA+`S*qG~0T$Xga7D
z@O<h|4BDx3UksU)acJ*E54zNX8Fu%P)A3P4g^7h@Y4VoJpI~euv`mWpb@=6#U3YYZ
zjye&Xf`eAwyZALkKP1gN^7RS4Jdja`m@Lb`oreiWp7Q5X1^bb<-%Aw;JAESWx<R_K
z3PrMMs<fE^Z<SE}ox(SlV}{1%+!l!@vrC^&y!JM$e?lpzHC!ukb$rEEw#GZtAOp}R
zb~&I}VTJ*gB-SdGnK2E?Xc!QQ7`uHFOs<s)-QEjv@ZC+gYn!MsEulwwM6A%PdVWUw
zTL`8_-|d<IYhqBF6kQ4nlf3pa^zbQa4t7c7BPkx)p3J{xFN~%h2i>syuBf<ovEIPA
z96jSp?9YOw%vi~J*0SgymXapkReav2-O?VaM?5;}{Ag57GFPs;tO7wJq3<9O18hRY
zNjI-5An<)7N~$;Nv>JZY%J4Osce>3rd>AR#6(TL$KX2})W6<?<_jb=YQ%8LG_+*&>
zt~YGDO&HcWEOhOJRqKpJ=qVvr0c?Kg!ZPQB$|*$kfQS?@SLpg|ORA7p(|lA4WW&ik
zIFqijA)W|<F_vz~Xqfz@+UCbdHP-Z@=};AAGwEP!6PJtBD}Y!}wxG7s*mMlH<&WVn
zIZ3Z@`R-%Ooz!Kl(#$Z<u>4yY$<hVMVG`$ys(SmAik$WlhV~%Q27h^N<Zen6mSuhM
z@c3H(kh(05A*{}QfH!U9`dU#nyE)4}TN(As$tyE0nD|+OGMq-kg48J5WD$*rxt~r9
z{5FtGNXJ0KE;P@2D1GDdEp>}%YVVl@o}VNiCCO$D3bR;y9_ZI<uH%4;?k0jp(UE!a
zRPI#J3N@PyrRZah@X~rj=EP;Bd~2Dzez5&)Z?Fx#9|Z(64fy#d>j!;t>*_^52=u|5
z46<LK*Nc-R5wE#aAR0WMyEdcdED54Q#2=^VdI$?_$h}~8-*)S1(Sv-460&~Nb}fl%
z@^ix)!Z3(N5dAOe*J*kz0ACGCLwMNjR7IZy7A}wg9u^_i+#@aC`*(NIi-Heu?@L(N
z!ULiojd-32Vt~v0Xp4I%8u7-b5_Tca%mog1AiSm@q{l+FGCmsc`@@Q|KQ-UBJsPjr
zFL0uNcaF}^UtvbU^Gyjhe`XM0ThOA7*S7oUHyu$@F<rpvyoAa0!0hI4avTNM>3(4}
zQ{BmP2Cq^UT{|iQMmvf(=B@};SK%)+YX~Qw`5+MX^~>HQYLoYyw*wFWv@of~@7~L4
zw{2P$@CK6k`M(RX<iyY25hTqXH>aziJKIvb@x+MoE!vT&-RermnniYVcxl*PJYyqo
znA;f7MdylhzKIDg@TZG-XjI5|d>b!$-0Z0z&%$I?3wM+PBx6a#o932Fx@X1T-ZyZ2
z{jVP2bo(}osn6V?@I&)c<2zLATiL@mH@D56cxM!*f6S+0Q##RB?xn2=9p>48A&q}_
zwie1<v<-o!%~5L}Pz^Cc%-ty_LvJ`1Dj3Tk?DTHmrc?BV1c~*VX9htJHScqYIEkHI
zyD%_=l{mkN--@`&w&KBLTTyc~F*0b5!YtcfTvTZj7gJ1MZt8BDrLQpA+>F;ZF6XJ0
z*N8cuU3+c8CZqa3M8A)0v>W3y-Lk&y$|35|p6%oW#`czcQrX<ZZ0PX_1167jwhk4N
z4nTbY(qth+L8D3FLpNjZl@FN|n=H0MW+R^&%AP>~Hw(jiNW>4qA0F~Ek%QM|ex)-=
z+H^FXQPVoDLbB=GL+<Ig3HvW1Wz)OEKc7Wo#hFJ9>~A5h$?z;}npTV1jiPOVsX&}R
zkaHu-BYG_6iSd1{D~v=SJ!4_f&8fFQ_&PIEytfuj{w5ar1ij_BKcZx_Qa6&P+qTI!
z_FT!+T_i^9g?IYFPeUZWk7f+BG}+?^wlldcGSA(~%R<SmN7D!(O!bB-)J%gdxA)IM
zaT<G!r}g6?D;7>ju`k7j;x5M2(_>q9?!NLJ-t&2&B@(0eJ13<aLAUTopSsak(>NPg
z?E?EtR;2PdrSLdjR0*|L(UB!uat`l4iSd-}lXwBub<<D<Z!|8#MBm9h7WUir>kd6&
z^!CH_O=g+hby8vK<WB@&sNOkWV1B;a696zkT^D3D=zK0fK7q7r$BUX#PoJ09BG*3;
z!0_r0JvQqS|F-btxr=Rd7NesNc#TH~_d&H|Q|$pG5HH0lyeTef_IKxW78=2ba(}S^
z?Xkhq)32r09HDZ7jmGvwF~Q19H2q!IgFj`9c?ZnE`}{3436AZV<N0Xv4VTCc!_%z2
zWaA$&90K0JbPz0Hm1CmPR`sF@^wmZbd$Cw(<&1s}e$=M|5EPw$88_VQ9th?r<(U8b
zP@)LZ;mCT$Qu_kbLsC&ub2uCBwfgZqKfA>p)|MGcyxKV(MyH7bYDykoyibyGk}kD7
z>#Qga&nK!&O&~4E6CTVAnLA0!=u+f)^XS7GbP|(_;au864T`?am<>S~0W}PLh-@Ty
zwGp7Vl$lM{-5ZQqi#(VQf8!}G(oOq>@Ar_A|FmpGvOf4(S7&Sr*fI@~-#)HOB?2tk
zmZ<x$;O=8GXnG1s%d~z21x>RPA|X;<b1C>V1YX%y2pZB$iZp}Li~j4jQEF}88hS+2
z#*j5|od)08${Efw&iI5tm(2yHQYtl-r=SZ9c4u~-dOG_QaSAx;1XkLyZQW9)&cLIz
z8vF;Ew@6GSWBEF-K`u@*TYebelA$CtaE`YT=DxrqU0+LS>@W}Dub>a|kp<G^<{!=S
z6niSU%-r*Cn$ZGQ*FFl$`X{H%=M2!}x`_RgyIwS_P327ng;3_W+M8Ghyq!z@N7U!m
zhVw4&_)wuDd5m5@0RJ~#NW?<!Z=5T}1kfVeGxy$wA|G5WL|0v5qR>_Q3bNQ~kW9AT
zS+1+>`DL*ToryJ^8m9I}vAs9;eV|j;>eFqagfZbtSf14SBT+<wzMqpo`Csh=`!fk?
zOX~56PBz&}IzMk`VMF`Iua)*~KuVo_H(eKdJ$`XAA>X05JgWXZ7?2#}5KBbG?S^i!
z%UJ$?rSjerw^h!!kNp%h23+wz5fkq^;*}6Mr3OztgUbJhW-5l-tQZqE6SXfTdnV2U
z>;B)#SV8j+anidxx*|;#k0BVr^b6uEHb-<_lfcQ~n2DA((ZV7fxfhn>PVM)C)Z$L}
zT4grkkO>*s^+$k<$x?Y9I2%}wytmwDctLa)ue3U)Vb6Vofn_L^5^g<KQi4Y@`qf7^
zR=sN61VzdcX-Qu<&*yW1ibEbn%8a$DZ@jAtX4+azo+_=EgSDIQ=r$}yVu`il{?)7r
zl_5E9ZW~sO#&V&PxOyl@vwR6EGfq7klk39N#q?+<HOg~-*n7CAtcUKT*y+-nQuuDv
zr!f9qNR~*;y@Y4hRtpS{;&6*oMwVR6i`4b;@)L<Uy$1sy-IqKEiNBE~gZZ+{f+pcN
z@^&#16WlBb#DO63L^0hfMu=^w@4R!DD+4otgxrEj1rag$+)zh59*YZNpHb9$@%gRd
z%(JQg+7%Kre@+?bLUNyon3ZGl_iLa2ro|+tqxrbzHj}86O#Xsl)bSsZNrG(|gG4On
zc+yrtx(l!XYJ?OaEzDg|CZPaVr_d8xAmzByf%#kf3~}jWFc)mu*wkv(P(nCe<ZrXh
zKclgF;mP{4jf<KePPPRfO<8{|Ae#DvekQm*#+|^O<B7QR=$iI>=TCFPkI}K@LX9=*
zn#x7u5HS`EBj6TrPAnNIn9Y42bI8DRygN&K(d~QGV}fbGFVj^>Hzi}lVVx=+=kf)(
zLVWV8J9t@Ux^XOFZ-jKHZkdh<Ra@}y;Xf|-4?*Oi3=7Oxu#(b0QB>1vir7_8q7Xn{
zu7=5?&ChLh!x!Cn3)ATQ8v*bp*6S)H;-xHkyms}fF{{?p8tx;?qy-MPZ-o#{qY>65
zgen3yyYb~VR9O-_TbFDm71OsyjLJZ#QO-gQtg~5&p33+?pJ8P6Yxr9=;hxo}Qs8Pm
zLtDiFYahaA4wY7}ZF?yni#drA;7Z?m7&aMlgSIUBS#19Gs+>ZZAZ~!TC(MdJ3}J5b
zu&}IYDH>PQmgWL+^&%_=5$#8VQupUEsS>6=o`%u)vo#ljfgCV^N7iR`_g_WIL()6-
z#NM?LKhR&8G!a+?NC0=Kvg6@Pc}nN39*TlC>y;#TY=W7|_&~?2^@27Bl(`S#7Ofan
z9^eKIQ^Wa;->f~HyK4N=vOVGnHekfm7CFt7Wvy~uh%5!NSzAZ0_>wIx<$VgzJ|5-<
zm(?umt)mW4PY?Qc;w>|UTdb?11`Sqpjvv4&<&<3GOa4>#dx4aRDY=*7|KqD??rNu1
zqrLqHiils%LkqR;A(mM?mb?U-{8qGmm7;U`neBE|>1cxIhP-{7N802dJmx+#1mm-2
zlW{~~`uqc!s*)-Q2ATyTQ}qVt9<9Ios%g;|{Jd%@>JtT$iExWNWupDL!#b0r^pUzt
z0~0-^IL<3*{U(wkUm?y3pMPWaWPHgJ;qU?xYDO5(I(|v`8MH(=nl5Cjp%d^9s5;sT
z&6WTy_86Nn6dD@D<Q&ro8FDE+oOT$~E|`B!;?ej%)kitslET65L;QMSv9~6jV^8a^
zP$53Wt0t{^RWsJV+KYf>@Qj;0<B8V9Fbn+F6Z{9JuKln`Gzu`nB^8&g^$%e$bJAp1
zS70H>=p}^R<}WU*4<wA7Hv2QN2x>y7?pbYDdRG#bHyzj0dWZHYjOVa$S>T613W1!s
z9$0mSuug;BHI&<XAC=q^{ClO1cwPydQIb3HZA@vOHmA;^#~XjaR`5srjoRwo;;<ji
z5})KcU&ao@wvhS-R_?vuV4%%~5B+(fVNGON-Xh1(&C5i<@uRuTtOIK6WRc~Ydes}U
zqp|_hUi8uE34g%{Mdr*r^1J4&MTb-j#{4L?uLmmSTTWX_-nT>?+Ty(vH3)(g7C+^l
zYT6OGepx8-W`=1bi|r7ec?I%LTXO&Y%7MegD9I_d9w}GOMEvUh9>Hy?(Q)B(E3G{!
zd-VY^811>bd$Ky3K_0)N=2V)$B;#)Y@+1>->3At|=C9MXZ1K2^Q@2vt$bHyV9OD+#
z0)a-h_h6pYTKAkW+<8!j9M0IBKw)5ENndVp=QBuVenJmqeCF#07vrsCFq}F2Z;db{
zo3BtA*7(6nqCJ}%r=O~eFL#so3>P8)ZW?<R^fKM}fGLaWf^r12qO!AjA>6x2#3TbU
zN*-dcR%j4vft5ZQ`Z1K&*AArMXrPWVa*G4t`?BjaPKdlEhzM~<o<ak!xAspd?Q=jo
zJ7?)u;ngOF2rv(wG209OTo(dI<gF)qw8)@ejsr$d^;<3a3P?A)zJC`*m)4*7i2%QT
z{n!-}kl!KO<3PluypW%RA1ESh4CLNS@7I+*KqQ@y=B!oi;B&iQF#Y**^}!qjy4e(H
z7*xBSFLoQd$N0?vKv6mOYRPok1(2XzLDzV}+$@T`lBQ@YZErSzdt-B&=C-vSYJ@PX
zRUFhU7b`=vK_HtsD|XvCD^?$1fkX>mljG7x{I*gRB|k?7WTvvOiz!OchDegt47FHR
zrdfRhCD7P6n_e7nax}NTeC9#-wc_G;^vSaHP51&Sz(d;_En{oF_=tSNBL87>2DV`|
zQn5g4#Rphje9M#RmOW8kwzdjWQ73^8Rc3S3t{R+J*kCuk#LNj!Y7PM>R_pS9r<)#M
zv8)(7-<Zk>2@4+6ri!5Y9TZ`Civ+{ef{NqcR&%F_$An=$tJ9d_IULsOEU`mDhsn=G
zj7kvFYLThvN3z=<B8e`K5uGK}^2+XE=%;HI_LW2v%<`w9qvhQlI`r!Lq8s(f@8OA|
zc<iFEjl*ut)tuiV9t}LO@rcO!4IwPp5r^H~4$NJ$C3(A@KQkIOWSH8SI6fQ$glVzI
zUF{}2lepORXn{x8T8F>5AB*6#1WnBHb!8sd&2)I$NI(pYiJ?3~Q`WH)pVODx;f9sU
zkcxZYF}dFO;`K}aI<1?u6=5S(qm7SpbroERIVg0IWdmBIb&08KSlbRn7=|qg6hE7C
z?blcH$ouGFY&7+&Y=R}Daf1)cUCb^weiMl1^yr(_oXKfzl~7CD`rQ9+?vOgOh^BLh
z6IC#p?z_!0*Nt62v*o)x?tbMpIjI$hBJ|FRZF7{@N*puYRQR7!^WwkVnzU&T@EhRR
zjRjXm_G<9J-{i369Dc$#qeDWDUVbC#<*0_ZY3eoOZv~3vbE{F|g$R!!qa*Gq)<A$g
zEjSK1^S|EWJH16<9l~rKn!517qSHjTAGXzb)j(y(GfX{mPlbTquzX+11;1H*)?_t>
zQkW1i!9x<vus=M7s9lrhXj<0Q<f1FaXG6e!;yCgA(t>mvDKC4I>g~O(SQz?vzes&y
z>Sq_MK&*(IEq=o=q*xrNYr2p!IhNbjR_|jWzaana=-;1hh8Mr<&`^5@MaqI_U{g?Z
zV4cT~Z5QTkW-HpCPVSCJ9>hPk=T1L6vp8DJs{nPrt%;T4LgV*@2{RbXL_9*(J1$Pe
z5_ZF7I{Qf3T-xHf*Nx(_AyjMjnhsXC(n%>YztYL2F=u9lXWo4#Tl<%ZpmT(Ub^(JC
zCH}#?C6i}l9%+YQvWSA9^4K%0I6B`)x(FM?9N?x(5?>!I1Pg)_Ny{%RQrp26iHY=+
zGqFPXyN--T;s<tkc!c`rnCP&f4K2KtOAc7B7dp)PFn^^88)JkZ1;ETJf<GSdJ@L_s
z3{S#$Ae`D4nlcBik7-@lcs=Uhy_2Ji$Yu$wQ@b+Y(DsEga@pmagHrV&9RtsON9g`b
z5Mys|l&xTpm^?)*#-b&TH5K(9=~-l(!hOWx9iU1BkSOElJmVd4%QjpM^*8a1JAbTf
zZrn+F%n2j>rR8AG@}OoAbwY$W6NGRdb~B=*o7hL-l{}T3{il)YQi2c<f&JI&ZyPiv
zKa|h6q>hVM)pk5w%B<fd{`+3=b#19twCWsA)Ti0X4vMJ5)cDw=+`GAtl5$cc$OQ`n
z6(W%st1wM28`#p|!RD-!D~_Yj2s>lHqcQYDhpN4RwyQa@6CnOU&*;-Xe#_%!9SJYR
z%7stKCtefbt%psfPDH@1!EmP@?_qAyttCxJXJFTUkF%9c7o9EKb3!HPPmi@%_Rn?R
z`BMD2T_Kc_@9F6qYF}=BbduYxAEL}(73G5F7aho5O=z-PFkI;hyxeo1O?5l-e%AC>
z+?v{>szKU9EZ4U}GRCHTxO(71+_n)TO^)iiSuE?I)J?53mF-Q0{nbH@Cbm8p#LqZr
z+p)}+I&2f4SJpkZZcK49p|E?90?{+3J|6t6Je3B4P5TG?wx4+nq4b+>`~&Nw8gdZ_
zzqKa1$D-!>Sx6^y0BzijeG7Qy^laan92PKJe6&S$=pdtq8QIwsDi=2JUqRA)@&dlw
z=pUs@G{j^{eCXgFj}HWMnjbD^_G}>p0KT-c{Ye%?(-2GnAg%0fmBP@+K-=>#cEvKJ
z5}g6YU9AQf$0KQ}oXBY%a&8d4n++xh^6I`r@m`|r##q{or(%x+)T2?}BeL3x#tIn0
z@}xRemNiKYx=hPZb9L_4>A6{o6$f4)%s-hTi&-2}x1@H|Jss{!ZOC1=i|`HL9&~^w
zBVX5o+bU~cT3603$V}^fD3FDn5h=R4e~uptJC~;&Rd+qu38s+)s1QCJXWn3J^68bQ
z877HtMQnf;cDAU~s&S-@sc!?YwP#5&E1FzDJbG?%l;|U7Ipqc5wutD=iSbnOVo)pM
z<S8FsUPI9_C~F__pY!*%J3VCS%F&>qNypj?o?UnObkpCM+-S+OvjVgE)V43K0^vL0
z19vKp%+ExvVzdjeW$eot=~v+s*!BfT<KhbcbPB5noF-7LfS%rsn;x)#^z*BY+)##^
zPUi?u%p*hHoAK8|zpy9cTLR}>oDs(s<If#S#~cNp*d`|DUx@l|Q1|<4`U)b|EvKa1
zqVCtZ9cuHg5Za;+aY+*(4J@*j9B~l{>dtd9XNELW`a_@ATl#ouq!X2-W{Ngg)AgV&
zRi4o;N(L%pDBS^!eziF!`9it>8vmTtte^KWoqzD=L65_QvvX)V$vi863%yibJ@6h`
zURCEoZ2o7uaN>Nd5ACShw^xA_BHY3>Y}zcU7Z%8^s)$Rvkc`A>bnOm)r`bMC<F<nO
zy4_Blig3*ICM*bh_fFe(6Xo{k71Z<ARW|Y@OP<U;90=9BlD0pjW}Wp?*}pPQgJ_jf
zIMb+@t*D{sFQC5I#LLSoTq+a|xF&yg%wo5~v#=$qIWYb61Do6}Y{C)|F2RPo>gY<`
zYgcT+v(N>}C0%H+MSAMi_2@-|)#-S_c@!r)U^SE&ZtN6Ges;YFTM=uUKk}a^vPGMc
zp!q(Z)`Q1ZGHn2Cqh2y|^!Gwb{0vhOyz$A;CPQFD*+3#OkL?2|tO}#KL}9srBh6Mz
znE5G@NZoJb2YxC{eVWvb_#TUm=W6;mxXQn1zVWV;8;`OnJx#v=v=dn_-VyaTR@1yr
zH*;2;OG$aV6<;9R7+}rlmLneajymw!E!#90%~zYw!Xy9zVPOqYMjd)+OhQ3}nJfbz
z%H-?4e~4k^_3oS^6-{Z<d!hOjWbrFTizW=sPrl$SI!gCnH4;h4A<<oaib~gs680A(
zd{+{#fj*H4*3VS<r!I^}<MAQi!m#<I#YRXg8lN4KJopNs23N7-^#5b(y9252+yBqu
z;2iVVn`7^py*J4?WM$7J4I^bc_Ffs0lF^Waq@tu_m6ePN4G|@xD3wBf*T;Q7_wW19
z^UwW=&w5|i^;*|lBmCGqulV1lXQ(O|YVj)g!D>7gtVCeZ97OkzWZ_2O|M@?fp45XS
zg!;lqV=A??F1~kLWDj_Qrw{%PwY0!*C+@A3^=K*5oX6wkY7V@xhM1NHVkERk2F=(v
z_ahEWUK-=*@@B}K+ylC{Sju1)cqtsP4K<}t`(>2jof>Zp)j<vsHluzp+Hd~3pr4>;
zKl&+rJu!<hf-ZmssLq9Fra~W}1sy<_or92e`~SYhd{X5$%pKq1upOBw<SB$9)~Z6y
z+F|IDega?;0TW6|l-5{DLnlB8+Kt<?v2*A4nj$B0#o(8+1HKrELl@5t|MyUnFhu98
zkRYYp495bY%MaKXklkdG22zL!4~3pN8GXMQSzNd^5haT2&0?pcegxtnToybJkZex=
z*97=rz=AHx;{iVftld9H<{*YR_8sI8rOeBLhj{o28lw<kH4Y<BySk(Rs4UI)upjRL
z3X5lQJ;poH8pHpw@L!RX;yN$^SPV*CgT3fN!sR<x1h3Jq{df&Kt6s{#Ig|tFTfnk3
zys8#~P1@J2D4Il3yLUjb^CXA+U>;!1>J;w(%~eHMe>ljY^{*KN{C^MafnN(<Qq=b5
zEv0VtnaE^ujLbXP?^84JTdT11?|{s^j-(_CoJ%?R@6Hw|_vJ*kp_||PN#yj<O@yJl
znqvnl;D=&B#n!J#s~EAZQ7%An2AQLxM-7mjPz}$L)G^;aKFxOL$hf7(Xm%t5de<|c
zVtLxGv&SSOWv*3w1>{|#ckZvU&5`z#9Wkv6$H3YTq3oY5eq7{ECAnRB*MWU)PRtbf
zDp(?nmS7%j@`#+NE^R4lt(8Z<Bs>>INaA%^SPP@bLKsjktCLNa0IDe@NoCtXJJ{ol
zMT_CBMr7844NWXEI0Hm*9cK|*1v((1^f@MaxoftUO8@s&<_RI$)zGW*EEBrE!ui8r
z%LC~7H5SCj3m>zoHqC?MrW|7Y;*iWp1K-pK_;#u$K_~%Xhm){AFn!4)!h*aXQ*eA9
zV*OZhXZ<Dg&g>z9P#S>SNh|sW9k+NovXiln?gvEe?a%hWEKnHvlL3A1*76>H2>C=$
zp;Hj%MGY=TM%^Ds!GAGAmVp>@Shq`3Cs$$q*5L585C9IX2B<i|l{KO<iQU)juL?fx
z|Bqq0lp4Tb?=A88Pl=Mz=G2`NZ(p_4ejvknW?-<@Ex!LOQk-o1zMW4#0Efb7;I4AD
zogo_RpuOw08M|dus8@)oI#zCAHlJU+6g07WuPpKgFeE}|hcCzg!bD5Qd~7U=B@0`p
zObr&|0qKa4OApE6yKeC=YLUNK1;sHuV7!D*#fOk={8K)>0(w6!%rfHvveXE~b2aMI
z+AzjJ<|@4XUvXp-T(@{+*uX>W$TdP%Y*n{tcMt5BFM+Z0+>bOr4taMM7$9K6e6+%~
zaT?|`P-%6Jlhc;Q;a)U{<PY=>MSnk|{kIO!`6&9w(RL!Bs8C2!m6!e}-~v2$t-gs^
zog^7HOGZDpfWiWAQK>LVp%vK{3zC(`b&wdU<KK@zkUN^59NL2GY8B3mGJ=yRwm;7U
zsIScel=x%NOe$(!e)=rkdzu|;U<`H}bQD@2S=g6uec|prrem1?b7SeC?2e!5Oon9K
zbmZ(?Chf#qoIlgfmGi+HPhCV(bEwclNGQNNBIbc@0lcTi%~Q&+4%r<C`bsHt-|`s1
z37XuiOXsDqiNIMwon<uRUswa7^V7|Q-Q*nuMiM*X-2}S%EC;h<cCyeSpmU1)kXU-?
z)g4l-SJ_u?XTz7+?eFeYhJ#gr&-xkNg7O4y_1vj9Z;*B>o<b^_TCaepVJVLorfkov
z?Emk;C83-|0X|`HYpXs0emn4(FQFdRQ2@P$>SQT2a&id2`}EnZJNS@YV5c(hEbWEx
zA#ju<j2H-EV38So%F3cN68WtB2(XBNaj*xl6VBXmiP)mK)Z+)Y(5)mN)K4ZZmgb!7
zcNtJ4GmoZ`+TtTqdk&NCc+U$mKF(6qdjlj!cCP}WdpIcX)d?|1Kj0&Ba<xj5%4W7=
z^tmHg)dSxApe@5JHmPIhjRr9ik`=HD`yw-2Qbh&0uHF!j>yryYnihx4=QY*1!eFkV
z+(MrQvCOmYsM-K1Z=cZLpr%vPb0MxC;6Udd)fwkaK%iO5Yfcd}`d%fF*}iNRB<r!g
zp`jH?xlQ#~!&J(J0TL4={|5{484w`B1&F<BNo6GVk;ZOMMh~JEWcF3;uU=j{D;xu*
zU^)d4>IX3To19=_8JmHsxp&81(owRpXB`-R)6UW#xp*s76gl3!<0wMP5t-lVsq_go
z_t~5J-<<RCcQE7EY@0;>gjxbn2ucuM(IexGe+yVeEl5niU>n42@L%Yuqp9<5L)_wT
zpny{%O^a|se5rWi$hCyszvG%Mid}U@U;qVFP8VIQ)YJViRObuNASxwNURX!xwH+P9
zV~N%t78E6E5p9;elvlgnl}v*@lxExmpDpn)s+#^=K+?l9R^i(iaZZKLii)dX-%q;D
zdyF?sbFGW?Uz7h9H%s#g-tHPIt0fpM|0&2Xg;0}Xv1>z-cnz5r5h{rad`Qo1o&>j%
z67O&5viPDnfj6EyPX;dP)J{3T6`$H~-E^FvMW$5(3X}7y9;#AF@DJdY5GHU5g+6fK
zm5vKXR6KMqaxi4tEbkmbr^%POfF`(xLGH1<KfR*kuo0t}<`6I1tLzi^4i`-iN1jCn
zv1Ki0;#x(1Vz_QS8yCjt`n89F&jcP0=FiX72l>YsT=ZFG`-lsN%2k;wbeGKA3N7;{
zU1FkiFD1pBjBf&<YAU99OBjVHJr6@&^8DNfq+di-q9`xzj7!EGA}Q~e^q8G_2Ait&
z&O*a%K)`)EroJXv4=o{~NnkTr^Gyng_zpN=S|<&ahqC({`HW!#B#<!$*Iaz}DcPOr
z_@G9PjM4<$Igr($Qt<Moe%zndI^5!CB>Y>ApWyu}r+CA)5Dt~)r{@Ox(^8BEqZMlK
zIv$gDcR3#)i!1vjj^Pr8h%sb;<KP<sTrpHF&V`vVe*2%5sLmH)hWrwxW2L;t(5+6H
zbRg#lN+Za!{s<qFah?O&?1b>ewI|PcSB@j|iE8u--*SNinie*xGnW_A7JU*=8am16
zVj1f#x?`YcB<H6%|0CW3Wo5_XIlFUUUcWqsc={I4v&Y{(_3~YYSD`K(1tu3z+aCB-
z*`3*a>|GOp$al?&W8UmzyHu#1-cg5`s`r{(tQNYV{fI<1Y$>Yh@EJAlL+>3Ua#$|L
zet^&l#%i)w5M6q-;B<f&`tx6Cu*c?qji-<f7CiF2gyk+NvX+C;-*GvY=%31@S+Ctc
z&LHal^K(_Ea}DgBrnn=b58*yAvZ5O*%ba@by133TEaf0^7!C)zz1p}2yl)yVy0{q%
zg)S(lp1?Ty^p@h^U-2naw$j5oVb@7;lmehg_kxpn0v(p|1aps>mG5S4prn0B`jZ<7
zm<m*qG3wEo{zvJp>6;M~ABuIJ1(*idiXRg~BvWfLD+!OeW`k?=hm*7wcRh6bak@L#
z$~JNi@f~Y7vv*JtuH&AGYB>i7Dxk8mUY&d;^a1q4-zEQX_>g6gz1k5_K*Eq09r{-m
zIaigS&@^vLw|o`DnXS#wHG|X)vg^UJ*kt2Zd~eSJ7X5HZYQ7lc^&^$Kyz+^@I%FRR
zMgYlH^OZL*bZJiRl!7m|pHP_J3Qe5FFNMtgxgIVdh93`Mien+82|k462*Yc+0{afr
zC-Gciwe8x>PP#S)JLzV<Qy@$X#!9&qSf1+3F99QXBwTKbE1ixaKHi_pPEn0Z2Npy>
z1X)@Y=dorU$-F&xi>w9e_r@B}lfViSrDa2KF5iY(Zd1wVsAVnxqv-zol>l$XaRmTU
z?age^ig$NgSk7>>KCg29n3iLt^|ROu^`af+A=;+en@UmlFAsPk^*m7&MR&=+%`Dn$
zV$g<Q-W?2OyMXi%OQ;=!!%)^QO<$EyykSX6E}8xk$;#TL<y`o!_Q-YLX?hRx`!LXO
z{ZN|fP2uxl2=RCn`&;?&R~M6)GzMlj4?lM;IQ9LAVYe_}4sE&y*M&lT%S%ZvDNuY$
z6udmu?HOs1(bs~(pUW@gjHlTT(4ARmWQ%dhb`n<14y#!BKyUG8R3X=!SyUNnW|<U9
zY~!|U8z-&}ggH5;V%)qnN-kU0NB+U5Z!nYEdCF}I^!Z<Y#y^24L%pdyACUL;zJzTh
z!9Xzd@Z-YRo^*?xTv0q51=ivCXjP&+58WY`k+as<Xm@|lo~}<DrF$u~$qL56u2JUS
z#^odZUH2;Lr!z=Pcd{>|VLZGywyEiWa9UW7MA$pz9porY+4Gfa=82>-U&H$5*ZVGn
zw3C}~a(`d}`QDnIKSOFDuj*0JMI(fZkVW4Y6AlAV((Xi8;etq#*qcT8c(%-L(WWrZ
zil9&Ys6Labm;_Xkp!R8cIvEWXCmOEYwQ0e@Lz`~hUTl9GFRZ@^g`pbOn2?g|bU>Y&
za^gREwhW$Z`vPcZ=FKD0kNk5O?uKkr#Dg0af;+~SBoRV})qnf(MV%;?u`nlMx*v$x
zSV@bt+v%dexwbW-UwcnY7->o+J;;bso#_kHo*PZMe1C5`gXO~@p1rA%`}{FL<d7wF
z=fq{JJw=?iFv)Zr$hFKJR7~Me-}M_)n_81UDJ4J0-Sp#Ypfly-y4A_o>26y#?)Otb
z#z|Im;Ml}f2IX&;G(stY{4ZF}klFipg_RAmJbFt0mi(m5VNjZ{izQrQh0JS)xACNd
z02%6OQe3}ysga821<2R?GMJ73m=`>Sa{a%WX@JL(eHkQ}xTqnZoI((5yM?Y{Xm+Kd
zr0kTGc5ZA~rQuSeZ5=cp0)VDaK`G<@+Bm%O)aYL0>>ZRGtS>l)<90R?n)MX+q|aO+
zD!guMDvMNbX(o&PDm!|Xq>2N%3u}D1P#(j60UsOiz1G_ZGruo9ws&&hV{VJz(ekY9
zPn6GsA{$30p<9*|(LX5d{0xe-1oiclr^>{qS|$%#T+(~k*p^HQVGfsB4WcT$OTC$4
zy#Nz%kf}Fn>qWnviqfPnHimucvW$<V=_~1q2eCUQ+nRIvw0-Oa!<J1Q>G7p~zl6y=
z9GtQa-1D&1SZi6gqoqE*ujKvO9_sQEvnI0-l28_8mE+Pg>@D<gbm;hR#NP%yx>PPt
z>{Ppsn6nFRq(xdj>jC+ZC|;2Qy278pUFQt>tkZtPd|ebr^Iq1MU&xZgkc^Goj&r)g
zW-uGxWu}&uqP=MhQvGA5(UCf^I;{`@PuUIf>-Pcaw=N<j{EZn(YSRv9@S3Ul#3l#|
zwab#PmnwgojM>f2t2;Q+!GHUdXzM9O`K|i26rDQHv~`hYZ-{L4lur81YZQ``3{mmh
zmRsmNmA^H3%lBMc=;ChY35Z0edIjIZTzV*I)&8fWg1;QB{~#oZfc~}M%-9Bhk`m#d
zpyycKPWalap-Y$XJ;B*cP*;<~yy*pAFcTZ4ytCGf;Hg()h}G6BG*hp^jtQ4r`EdG1
zpBz+Dc9Zmh4vNehv-8KA-1^4Of<mq8)i8K>O+|W-MIVkZNy>n>3g(FuKy!j=r8xPY
z23;FB`}qot7xqa4czhn|losRPW2w%n<0GG5^Iw%U>(^*_mYS3E8<g3Ks$mnfm|5S<
z<p3_65$!_uP6?&4MdDoaC16-jO8WyCG-MIDoFi*?o{1O>K~v*lKb9RAp00te@PB4W
zkU@caSLES-7}H>TGS&B=Fh5xB%+q0Bp5S6HaN{juvWoG#i#wQ-R|(@pO=H6MmfKT6
z<IGA>*jHk9APedaKq2w>3#cAM6ynhB!4U)<7H%kL{Ym#yjD-<ZYtvhhq5@&a`OaQa
zF}0tv=nvrLDLYKp<Lw0S`w}KR_dKOAz!7|yXG&`nlvR9bvvS17-|Gkor{bGavWRL=
zMM3c)d7#cs(k-2ItA*)Y5<pkcn$jlh2MTB}s4$0b53>3cCP^`-ms?>oz8WSqI9Hxn
zr<@H-{4=9$v3XC}azJtTKUw-ER)e$qF*H?NOupS<&77UqAFr%>{0t&&fSyjuyt#<3
zz6WKU{Y6=h-&*nygl&n}`31(;%}g>PjR!pDq6ntTap2xtd%CW+F8E6OVtF-}a-Kq9
z4Hvibho6Y~4U0lQKQLUO-sFybXMilIyM^;Ik_D+nV!XmoA@;jTvxF?|THn78#aU_!
znDb2Ux{t91y}Nh(b(-w`@%I)&Z&LBII4tKe2XQu8&;mVf=mhC6I8p2|fE3*W7#apB
zrgh8=J^cXR+OLzM8DJy)0_faBB_UWV$QX2e{RxTZE%qIvhHlJ0a;jMJiM@81<U@$K
z0vyxi&%uhTnnqKGkT@C9`4Z+Ke_$?V2>%Gx3@CiWNSSgqpPCy|yu_Y`+nu3VVpWRC
z-CQ&W#m|QT#$U_}tyX76GZ)#X;p3lmrof`j?QG{+))G?v?%#l=XOy;l&N~w`fo&6O
zkzpl&97-__2FR-QZ?3igs?}eN$vX;ZA7zw7KCUD>YQ5?_0o~;&4!6Cdrx4TpGxtzk
z2ksUsi|u~i^XHe@X!SE*;%DCkYRv{zT)m)1P=kFr=lF+DYqKET)An()?!*E10(XkQ
zYCNLbYSlIZB({}G7z)0kP2}{o7mSe7oD<}MYiLlj_^0(bW{sjw2Hv}8LH3b}I9e^J
zO~MQ4B<G{=-!|$~`IN3XK3jnviF|;f-{I<|Q?XXjQ8j*y;m&LnC)AAa_SbCWG7Sq8
z8H+-8%ncB=UlRu64zKOgXXH`A?%O@*!Q7cxcfSLGkdAQcCuEEZ;G+ITCN6bBoa%eT
z-+s)jphq21NW(hY-Xb~MUa4it0Ua6?VlgST0%FV;0IH1sF#V%lR_>Et+B0jDm&FXg
z*;x+S(-I_EN03}~B<$+3CHkQ`oM8O;X?f+uOh@J;$sN<gvShHfK7`~SxH|K=1F<z^
ztAJuJq<*xBei$zQUh>v;4b>Y?ILMkM&Fg2}L8aUleMx9Ai6_)zZ|Y^HXr?8IF<bJ_
zo|Ehq`ta<M;O8xD{e;Y4Ap!1}X0coQ)67ogk#2*IE%gqo062T_dK6gazQy-|5#&tl
z^Y15V@^lLHOjcMw3%iIz{|bUSridM4Ft&DHWtCe+sG?XK)MrDNb@LO!wSiWf`GWd+
z8z$9W*fF2FguLmK8NXu-)!U&5cJZQn$oh|LQ=7#g{xoKoXWBJqy=R4X{sw;p?sFNl
z^m)jmi{o&mR|tE1q<=GR8vX`Qd*x*6vN#W-KcVB3K<ZG)VNTm0o&Tl*u-NbSPB%ZP
zs`CTS<xX7n=$84RW7kV^F<wAnGVVnBDIAr^5W`(3cY<&bo@Yh(v-QbwSm!Xm3DS$M
z%&v+`xk&nAQz6W)bA-=*kh+JD_vxw*`{%~dQ6Rpx0d9<H+Ga2`YR_S^6!s9vnTWSv
z1Z>yhLXs=iY&YY%%5RgcrTd6l4uvUbv5HZi%Vw6pc|v5~jFrMY%=@OP6z}}up1c(9
z>#vWQpV%_K*2y76+OJnC;ICM=s~-H)O3j!PMu_aJK`aMrlUCd|sC*`Yq#gb_cl6q8
zDP;Bm_MKqW3F}zuq6pI7Xtr?Ya~+@X;ogJh&#R4TCXy8R9?wxn!j25C%j;ffD)irv
zWdkBE47a1-ezD+g-!g}E8Lk&=#%};UA-l`;-XIhS)f31^{8x6sF#piKsY{SDOKwh9
zBTtByJ3>KsEB!X2OzgYImWaLDJydzk<u28rQ2@6<%TYqxKE42*D>Qr53sBT9E#^1%
zVvDL)3^SImS(YqMB<vce&ET2Ve2_?zJA#=v6E4SQq<R!W&H!0o3b5CVKvj7RjY6CJ
z@2wh+m3O&NVySy?gA>~1%M)l(*GK*VHvSSwsRAxv@moQs7;Wui)5|%MJ)=At@C@9_
zy$k-*fuQ}Yg&E1kx7LnWhV=N&Qs_3O9^Po6HI(C19{V+7Px=Oh-X#?)&xJG@yYiBb
zg45%|#M>U)@>skR9jkT?X8R8vw#N!6*W}=mE_I#Gg<RMQa)8X*KSYD}+OnXazC^k0
zAqG2r0s_0D65woo?8)}w<^Efqr?UcH+a2J(WTmVVD-Om4)lY_^$^v`$1PFmOZ(}OI
z?U4Ta)j_TBk3&fBvjXvvrbiurYWinsWZE{Z$Vqv_5%TM@V|<cD5StMT89FX$XR}XC
zw|odF(f+_?PAiYtvhrH5LHL@|3F)JGT~TsCKj(Ekamd2R8lfB+kjXnak+MoYi?BL_
zl?<e?Er5?D>r#~-U3_ZDC@EjGUS#?7;NB4+LLqLx&mk;{=GJw~`rPNZ&&2iX)F<^k
zu;}bQA&y5~!`!JMp^ATn$bwO;%{c-uht?w>;0^2*-URGFrU3y-z2_q}BQ=!BXVAY7
z7@zSv0+=ZzkeuAWU5b+xikf~z#Y8Dt7W=Z%L&i_W?XGN?XEnL)e<ox_X3`VtvoKJ%
zEFjFj$CL?=XQcG(B+cnV&mR6gk+>^UzuE7YB)}sn0$C<DVklew1ryU#ZPkQWYdqT*
zhXCWrp)2ATgRUkc;004{Am*=t*>Ml$+bksmQDgUeRMIY^QYkV0kv~QuvCdv2&HFd%
z>#1jr&7;(AO5<V^uSZxJky<2c;|+=wP+621zSm7VWoGeEQZA2Ehfau)i)x|Q_6(Kg
z(n5=zC2LSKOElApS`-jA4tC$?T$h1@t~!#GxC~yWyFV=qKI!s(Aj%nQkc6D7{^N!W
zK~YEp7#5#yw2I|=YCopj3IuVPxWmA18|x5Ixy?rx;Iszw&Eq=-X1E_7$9J^ZQAav3
zIFl6AQ2}or-wjf}ZHM5RbcRLxPu*V;Ta3C%A_{gjCa`AZN)h0GAS!qL_qq*0c-|XN
zvasc+;`l;0cMvxth!rx*q_P#nV<bV?fcu*aUs|E}vk<Rr2_)#lfQ%&%Rf9btOmma2
z=S836Ps&e>SR*oirQ+<&WvqG+kkZ5cp@J8y&R%}DO7qD{S>DJ@Wf<tW9$jyl3mM&G
zSnW}?Rmh{Yfxt5M)_JbI(k_8ftTl10W|Oif8_fpklQg-<z5{>V=ToP|Y64xls`V8R
z<?lk@^#?<tz2}@DT`=|4CwunR-+vj8`tdf{i1aP;g(a6l%aixEUEiOj%E1c@cOO{2
z6c<bVy1JE~JBm7eAzUEz(uZu7*+WF;kuxxC+ZKs4WQDxa;!GT^MR=5B6q4-Om+wPK
zU{x(_WoS3|pAbtLNGSki;zR0X7OcaB3F^`pMAG>f#;VK}V|4r3Qz!c`)~T;biZJjv
zb)E_;0BH&$5kLJr6_LLaz&%g87zv=1WK1$!6f&~M&WX^oy<l8Bp2pV!ee5lZ7iju6
zkQ#RTK)!hM@qfl`FJ^KM{bAE3IWW{pK9s7EXLCPpME+_51?WR#lj|Kz-gB1`Pb5fC
z2!R*kSG#!`ga0ruDfWi1>!wQ&f@)LS=_1JN0<@^rj25(wR;64NbapJQXu@I80@J7|
z;=n88K#Wk-q?BWl8PXN<%!<J6McsCClDPDIwfL<usw=ff+?pkhIe`c(C8sw^NX>n)
zZT-^>(a`neD+jaOP>t~Z_me=A5tV7ZtkoRTdyp$m#Im%7FEUzw78+|T^x_m0xXL*9
zfE0#>HorF^v@Gc4sCuEP@f#oOiPft7TuK7^3EtrORU4|z@k&rkh3A}}Blf-!$E+Fc
zP@3GpwE@Z369|6ge?ZDN!UdP?{%}Ks=cT*9aLE@2Mi9F-f~MiRr*IZ6k5oQT3CZ=B
zl|}zKN2w^v(C_=M+`bAK0DeFvI&ry?o*;>t9v0lr=6?rJ`HAOl^Pr10(u_$tw6^ID
zl@@-zsjvd_yw-kzw66DEtypUv9gru^qXq6jh}P}vT6}5<jF^&-h{XyLgn|2a*&Y-0
zFgsbM|Nd=88D-l=p}&ZNh5nu6cAOK?V~X+ZgCJf$HTq~T1_P^(?hr)w^lI{RE3`7*
zZ+&1H@?cMKSyw7@ZyRPp>`KWV3&t!;s<WM<zgz^|B$Mg3|3rcjsE?Sd^OX-mc7$0U
zZ=BF(8R3H+Bq%~gX`p3sHB;>#Y=RaiT{sr~TNy}ZB;IX+jb~Pf>T_77LLi3`2#T8f
zBMKj3xwn6K&oWf0dx_V|s}G6;UMAorlJPBS+Gjj3=bo@J*{ZAMvw8Fd&q0g1T<NMX
zMR7fV6+2;p?7ye-H5LPQU@198Vj1pa_!N~&$JHn#y<R(OBCqWqZ~TEW=w1UV{m8=g
zqPc3?s(~G&({{)lM3V@eD6sWjk2C8G5ng6noo`2DiR~Vc4hn^#T&EAWUoX1^t(fsY
z&pZW1Pkr_-*Gy7mr)Y+r@Z0F96QB`<E-Q`6u&tmshkN1_N5_Ez?s?eM&=)8y{Kq<`
zeLsR7C5l2O(lh9HhuH0C1)8ox6tYG34GhpOAgBT7)M<9Kbg(eVz1G5CG`HXMb9pQA
zXBJ6oKT^r>jY3+~IJ^e1eyqRo98SajYgpy#Q8<OQ&}_o-kYb))qIy#mK+rua#Z*3X
za*5h;*OVq9*S81n06Xd(ijT-FY&B6Bj1dUB&S7;DlmH^e$odV$s<OvW2^)C>unE+C
ziGbejc6j9V!K=ZklR8D6?FCf6kxJU^#wsi0P`TI5Lt5#k?E}<VVXt_We8dzO(D)1S
z-pZhfG#TuNaIExC7ZV31T^=F3;ff{gD0CD(^N_Fns_mtj%lld9NU!3s=CEQXy|vAf
zd`+t5+58nbSCKhVrRV`|Dm<J$EG$}E>{KV7B-1bzd>6{lFU;V7&W&A<Sckm(PWxcK
zG+Sm(dTH-e3pHJ;%MVm_&xFK<cq*~@AD45eE0AYMJ@2R}7fInmhkO`Q=TTv=g=bL*
z65m@kyy(j3VX^3bLI3#h+Y<`yprL?F1eM41Ub-4SQ?C3mU5iI9TB0JB09Z@<<C2yH
zoK?4TQBk*jPr|&S?YusoOYK-$S_TiU7r~R5DL}7erc$Tlw6;Ymkr(Fwv*0e9M9wMH
zi3;N`<5+vGczVS}CTBuou!R+#j2oWpZ@cK0Ow$7%#p6n0`V1`Pfg2Uv-++?cIVJbt
zOz@vr)zI%RuG)2Bm-v%XgJFY*xCgwQAD((!l@u^X@%=dzB0kD##B&S&>Xi5_>tA3$
ze8o0R`Od9kx!Y&2r_Myd^2p(GNQm{>v7A(npTwU_vmdz1y4r@5n~fHJ)uqIE<-|%I
z_K6k6OCd-d`(_0}%_^UB<l01h*t&Ojd<Z1Y{pkT&6xk}{+)6}KeYN8Kup9=~N3<f*
zBUHEJ{Dn)O&eCo+r2Kjhfh~E0`d>890-_K3Bg_I(aAJOcGqh(<SrbA#RG1fX@$*x8
zd3PklM&e#aq_*y}@KoqYJHQVx;mf3erlC}Tj(Z+`gU@13=9f$sFm^rM`Yw;=BZW^1
zVH!#ZGJP_}!Xz>9ft4M>l{?305q7KpDk52ojJ^f&7la7)l#YXa`#67|g+2g?<EPxO
z?C9WTgBq6gHIx1}$yVxlG_DZ!GPa-qQ@!{qMx<Bd`11@4Ye_LmGxt6SNtf7r`@2n0
z)RdySMF;UWQlu+dmckOnQ+Bt}E3fX*O+@ae20+K@^1duc8%oHlGJ(>}k6xfNTIbfh
z3!jkIlY6Z+&IUV_IkfT@e6^p?8(!eid-MucWXm2>erp4h_a5Qb9$iq0_@2jw&Q1w?
zG#$HtkaVN%uRT=vEjYU0WG1;Lqh7W_iK{;@S~lr(YpW!r;2kdhY6uTW2<^3LadZQ?
z&$O`8pvM`vz2}S_&k}8JZOQlk^5(a0kd^Fg0m2F#kWaouhi=1hBQJ`gAvRRZ-=W^R
zb?r~&70AG&VWgEzUTmNIB3Jf8Soh-Pt>mv%?+*c55`5+%sK>B6=uJ7H+q%-(T<6kW
z)3&$^vmTx`_kAkUa^YLT7dqK-5cHf|1qrc8#DhKdZS5>dXA@iqqBJ7G_!i26s}UHD
zH*uhQX=am0%=jKx6S<}G3leaeZ6HModNpK?oKG2qM4%_gH%tLI<?c;VVs?Nv9e%2;
z*=6_DpFf<eD1K|^4NdVY=%gOfr$Nrw*;@s$n6C*+Oxg*35?$f-sW8yZT|BOj+AmwT
zbcQj^3uZ{MjWFgDTi<It;nF!IQ2V}vQp9^J<vN!tlg=uQ%huNNji$B36-s_-BC9lQ
zhGj2m>N$qVkDqp5V;gV&gmun-XTq3|X)vBeR%&da%~Ge1ykgpTTa7cE*t`Vh-_P&Z
zr*46z=a-jxNHSo<4ipKyJ7>W9t4LQ;3juG)>c#2SDq?$BfP;LWeuE`B*f<{_9NW*F
zc;T!|^qSSPoB9862tLB7OP?Ttv0rL)yG%1&Ts(37z0Di|0?$*6(dc$3<VE1c%-SC{
zn!B=BhRv+SF|1|LYi~^6(|LG83R{g!*f*WwC^?W9i5(uQ;S3|q)_wvh*1sASNGOC!
zJ_OabEO2(DOJX>@AhPILh5A>J#ngq*Z=TAq=v9ZPxx|pAaj|oOeQa!O7UUQTi;8yN
zK3dIAIB$SFE3~F1%>FE|U&=?R(j)sn7h+Uj`8?`lV#x}v*b9jw4`Hj=euJ(sKp_kb
zh1f>zrYr%EK6KVc0tu3F=lnsOyO5!7418*V2$x|?H0NG*!qTAIWXaApHYRdsQj9aV
z%32=~kr<AbJcr*x@&3#)*eHxju=n>Hl@=MR`_+X8Y^1Bbf0Ynr8R0ApIIVMFq1lG$
z3P)4J9m2KO9tPuUpvlSRP@ETfVE3ta!F#XI!beho$08Wib`w_)s2>?66um}MCpTRO
z5!Sq8T9sL^Yx&iPM8V_a#tfBic4>!#p}BN=R+l{i6J86QlJ?o1$0G11U1(~2W6aO-
z6z|T-5Oo3AYS3CkerU5KzXuTM;K-4R5QiM}<oVmv`yUqmW;}rF?@O_@eP%BS1A}RS
zth*~zRFR|h@YWL&6`2hrAD#|LYf|C!ELhdGx~TEZTjO|{-b=gEw31c+sHTSwwC~sf
zYky9YKA?cUV;1rSMl<I0M(r`>N^^fz6Z$J-8Q*XqdhSj5p)^wJBYUXX8+~kFjku09
zk<N0jJ1HQsV`mq=L1qO45`(lJ0F9$vwR{}NmW&)Bc$YNIi|mj2W|atk-eDi@YorU*
z>phBzr0?LvO%<YMB>JyZKhm_eE#fcv_N45Pji!G+w(u`ROZf}_jV`dU+4~KVEC%(y
zqVEejW-^whJL4XV(0`1|>T`p(GU?eU2<YLAcotp<K6<0*0Jexz`6sg_Fn5vuv;-CV
zeg@WA7DuYu^wMGCe|d@@#(loikf0N<QLrQ+pHdeK&BdYE=()aN9<_$MFjqf0D_=?t
zkPS5Xd=uz5rxekVWy?kZqilw_88_y&YpHJ-`)L^5*h>4d*`#U?3B8Jh^|3BKhC;v$
zC-{XhBNUY0u-G>c{L-;WR(;=P<W*|`%Zj|+rK2!Ex|E~D%QDU(swV3zp*cz--prS1
zBFjaXcv+3)U0bo!&I>2riyQ^<Pg&VN4jfwwi!)c_d;s-3{xqzzQiSOgU1#O*XGN&9
zPlC4O?0KP<J^QYRTrL*PM)BgS!sgVVyc^wG7W{{=Lo7#ptI=xIY~Nb$<RPXZ$j&54
zJh-k#Uw;$xar|Y6<)iqW4W=hDA7HPW%R%P6L0b%u`iX{JX*~ZXU`UD{!NA3sF}e8t
zfuJPfFRTpFB3Bn8t6=Yh+Zo0ojI4{#@*5wTp6|v_ynMwNIrR3lhKQ3M_0yv)SH`j4
zU)o8c=?g?m-3ywau0E`0FO4!*4LF&uq=Dl6AZ9`C#LSLd@lsyXGt@|md$InxpF5SH
zZ*Umh6>4!y;--rvl0*p%zl<D$kEW1eQ{n`Yl3?P2j(nvxpeq>6vE~Sw?*18LRsF@k
zU|R#>7c<LQEN1VUtiq9nc*;i$B&>M~Lk>=M7M=cub$Q%Js~{`0`7nb=Rhp&VfUF<;
z#BUC@Ti{9E<h5-a=gfzaFk!oQhDq=!Z7p+rR4Qeu=~bJH(iJn<O&KEtp&DvCr@4S9
zBDqu0iY!ky?Y2<9naCTjoh1!is}6(PwXS&%b^`puS6zj=;6y;VeTBLJ`Z;Y}h<Yw<
z`((iho=jo9|KV4C)|k=SG(q1X4Arvmrz>)M!6YP!4;RnO6w<bp<&fWs!t^4_Dh#0~
zM!nw>a*?y8{ooCU>lA@}$17=9{lAEyC-DaRWlZud^(-}3zE0nkfLUyWlGG;~J0je=
z3m{45D<L0sPoA~Z{bh>Pi=N1f(pHXA=A6_04XFhO2vs`TNl&>$6~Ki-l0cWe@s8YA
zA0T`Lb-mG(gY8pP<B92qG3v5An+hCG%`+RbU6d<Gs3Ja!WQ2v*$CP|aZ|3PYfNUN>
zVg!ch&hw*=|F@ll7IOYHGbTbeEo{KbDu1!qJuSWIDsV*zTMis^i5YDrW-SK9opVr`
zxgN}ZxEnkv@36ChBqq!rrmzr4FHVBiiWbv=nr8r;qZQ_eks|tcll7)diKLw`NZ7}<
zfSccOM38gx9ZbOwx-UV{{HtPTdl>h{HWX`eVLQI#uV9+Yila6h_JOn(3g9uYE`Zu!
z=T{5F;E*q3X0>-GZ1pq0tes9gso=@ESIr0}6q`Ixu4xEx#2zCUiS_z*9P53iZh3;t
zivR|jxn>xl<B>u&;Mtl6@v^@&XazuCZ{+N?=kg&hVrH|`+=a2)2(^)MQG`j<myx(e
zfC48ZKsIAVFT6TFR+Fj%#s&VL-o2%e69xx0Sh%@TPo`2*D1m8Ju7XVmH<~g#u<*zg
zcGPBY55ZQn949Le;^CXncI-yr)h)<3F^abo4z-9HJ;acaZBcv1=ewkPXlY$xY|lY!
z6SB}dik4y!hxdgmAfTPQX%(P0fcS8h+&Z0kh_6)PUq=FH{5ro7=lR(+0xUwdJYXjR
z##&>X7yc|X3!WllpYcV12YCi1H(ADHc2>g~uqXJol^J}uSDa4fRXKvl*cG0A@DZJj
zq=@`ZV7*gUF6X8enK3>rZeYZA#0bkgl%V>ak|#0jgvIzbucP6~m<|HUYYWLTON1K!
zn115;Kqpa*>Q3!BIXCh4WGXBX4CgchtTS=HJofw-v(glR2ljS2JDAC8Zbtzt`){}b
z7=Ga9C$rfjCaFXpobAzr2lkAnEC4Y0cbbj}$BGM`%LtSJq7d*AE#)jW8K+^m9UVfb
z=E$P2ukYD1nq8m~)<nR^wo=rLzQs>51wpG(Z~dxYi!!?cM~ouHYIb$W#q4aMxPE<u
z@5Rn5kp<znvf;Zx4;#NUoh;>hq7nl{4yadCzg8~lrmG8vRjt!a716s{kf$-kZB*u4
zBDRajs@I{Y?o>zY`Zp<%2Co6>T7X(W@SFVos!uAh=Kms~DaojhmUUEAMWQ;=MmoCS
z2nu0<^<k<<Z>2<{cpplT)HTlrc1s_-K+!V{?y?IP!>3|2bjI&s5EW{bK_a9B&?Q1&
z-t_~_j~zKEYpjCEcdykD`7{Q;FE7Tk#s>kdI!x_|HtIT#aC7O7ilS_4Z225=JofCh
z&#PZ+lx_5V#mnXaSQ`i7zCoO7@K3&jZWyq%b~$1XD-~QC0|tcMiBuZ3EDhBP?>Cgh
zPA7kPUI^b58guvLyS$B??`zXDnXc-WVKfa?>k&<SBn6~*AW}*s_v*(3d{d=@m)>T-
z2*2jw+jO|H9z)cK3_!F+A-};mMxax&6iKO-sWhTsB#`V^*H%o}l0J=tc?<yDA@Niq
zb~FV~dtvug#IG+Nvb&%#v2uko;l{x>B>yW!ClL}QtHX2!Xvf&HKU(Z4RLFA@uZ4mZ
z0)!5n`AiI$Lq8$n$P-q~wIzPqL-%cjL%>cr#RSu(yyg0L2jU(VLWfRq7LL2ypeuys
z?Q%TDeQRiSUSmI57()b0=&9L7Uf;IUfUC{zX1`!B!^<jrr!!F69FugBw7g<-k>?C%
zOaM7k+>H=VE-+uCGjoa(Rax53hdZ_@t8iThh`TXm*R+Mg2U5|3OcAt+aGQNSR-kRJ
z!P{@p^Jx&(q_{BPdd|mtbch+9YVEq_29Wi6wc~?>-rLDS;LtJnSOXk_peB>)9e5oO
z+?C#wZ7+3m#mF1rzL2<rG2uJdmzDNct)%v<s0EyG89qtV-XPTNKh66d9a$9nZ9oE!
zM=O9%GH9B|Brx98X-Zt;K1&QUx1g()S*)f=NjiWnI~oP)7NJk__yKY`llh0)xkx;v
zS;?eZ&p$8b+d^h2&siQglrTa2yI<1fi=$t&KzkGl<|g7%z%1PEeF1fMS-Cl>4AUV`
zRG}ygg|lB}wOd^8oOZ-=%Gbjw#nE2^074;I<6x0xB(gFjvXsaI(*pSGEoN);^^7{M
z=jgy>mth+D{i*uDC)g1RfF&!y&<m3;^M#wQP37$`V=cud!8|<jsJpj`Vp!9{Dt+Tr
z7(3Yu`q#j9xR`ky`4MXqzui7a5oi|l?=LWPmFzeRUrUpTzn(V)b%OG&*eO<bo|ngK
z9C?NpEKz^yu*x7?tCrlHiTQebsu?^4ms5>0FqQfVdCpG?gpA?t+diZ8eUg{kaVcXa
zK<Byy-(4hY3PO~dPQL3@J{CFE<#$dD4Lc3tbcLAvFNMIL9)HtjXU51Ry=m7~Vfk;&
zIriAAR-EBfT}9GTzC;1wc4U5m+m+xziuw|e?S+FVeg1X~(fd&2III#K-ZL<2lp$IS
zBAdy~f&NQr%azRmS}p+zb!Mzak#+GuUN}QFE!_u>;lxiV5Oi(<btgm=E&5GBGwgC5
zSUa`Kj&q-uHt1k3IKzZPlb?R~cJVzA_!^L5(7*iqT;#A2tnwKWdJ3w$pP{_4;Gz82
zrvj-(V8=?E?}q^*U@hc>aXic3Xalvpj;Iy>(W--$Q9y@Fn8qpuv}~yZHeduX(jn!d
z2_X=~;GojG<=#m3Xa#}=5cEGpG!dj6LbQv(1=aA+IKNVLe<hV-LlRFJ{oQd&LAeY}
z&(^2GB_kk}zvL%rAgPg$x^Q&PRVI3F{v@0Od+avt46DH>7E9NguvzLmo5qIE4zs)v
zM6hlAgBHoKt_`2Y-tYmi^QO+>yuA?W4pG#*7ohz6)GE3r8wBxs=dOKMghyr0V$kvu
zf*|6-j|t+5bq9P@nB_TnhHO8u6C7e5EV7+-BT`g@_SSxqNBsy45K!-N+ZK_gELi%z
zd<xEA0HrR3LH#w)4TX>|U1;h$0)Kd!*Vqfck6Qm*@)CYn=%~V8u4*EN&lricL_{)m
ztdOV&WA_y(<>N*|d>BZ_1aAV-$MOBrW!BhXp%KQ2db?%Iy%b0AM<Ib&0`cEJT62q1
z{4T7u`gcfsizCNj;$!VO80nHM0#iy1AAIgx2lx(#B;_~V{Iq&27EBb=-#5T#P_4q2
zLj%^OpKpdr2`nT9SsmX16mYRYJzx|?rjii)MC2HibKL+`v4(l(z$<=uJdwr_h<Z{o
zHI{yOr8&TdnRvR^Q}|o=ZP|KM^(;LS-eEp<{ZrD?d)T*?B3laAB8+KX$#`y-ak&WB
z{^ch8W>_K@3oe`p5oB%$rKumCeJmdw?bI(pK?j!Ve|4B4)aV8kp`Tt(^oKY`xbKy?
zB6;1gnn@dSXEyhHAD8dFjlTvShG1Fyo-F)(^f7v^M1Z?~ei}1QL(>=*H+j!C?EOu3
zpYx8=fZU5S0@7BPCey)qW(*mAnQ#pFbO-2d!u7h$P8>2UbY*<G3D(#1Odr&6Q$?r-
zs^*D76J|@R3mUfo1N`~?@%`FPz>u>J!DHVa_^D_PS%JOTf5WbpdOjf)2mVt=G)7!l
zl&>&D`7!{T-r|pGoc0nQLRb=!9IZ!&Dn^Gi*m9pF?44;O5yWtZ)y$qH^S)1^x}@up
zY_=_?mxZ@mwV<%Er(OR7oi#@;vGgQV%(X%!UIpR{p}-n$qqkK}>11b+eh2(vaF-jP
zbH>UjeP2h|r*|J<G7X?PyCZa-iK-L%T~ozma4Z3we`8{WUQmzEwB&ew-jDR$S56i(
zkT>9Rk<K*jA~>yLm#gh5o__3xEC!cE6ia_MzpWWXh2ZkV(Q<=-u0#NR@K757|FaL%
zGcb!mmbq|hD{}Y#q3&jcK7&)3fo)(5Oncrs22U68Oy0GsxY<&l_fmd6+I;u8R8Jyy
z&#7Wm>A@^cg~l5XL=08_c5w5sckx=X8e!0Xzb@6uQOi(V$he|#8G~#UIrkfk?Ak04
zpeOcbX3j$<IdGcWaL#!S4j>DGR5nRS^IGe`cWe{S)j&JsUbT?|8Ae#0MDQpeUX3u1
z=y_UNWK3ifpFPCEI3XxX<GoRmd{WCXTnH`OQ*EQ{R1QEXZ_OWo7LEc{4{-<<PiFAH
zT{@V{5p-4!Iyi2L217X9)75hjMhH&yk>B8T*-Ht@Q-1-i4!rgXa%|lY7M!vcy6+OQ
zYh<boN%+?$9^-bj|9w!}KyoS@^`42iipV;T?EuLOp{Q>8KLMY{vYH76UzE01>HaFH
zd~?3&qJ9c;|1p(humLi2iC&*6W7S!^M?xG9SPpcGO-))yBRcv1b<iv_zKo3YfX_&f
z1yZ0DWNGSN{)9Op>NGFFG$41MD}2I}0rTKg2Fl1v@iW*}J99K4As|G|BvSzqfvO3U
z91ktngLLOojwYnVgJZ%5q-~J-r4C0|ckW26@Lw|szre+uTQRtL69feN?9E6-$dv^V
z7J<WYQNc(vK(w=UX!D}V2$Bon(u{h{d)$<hiaCSiM(IYuil<|xMHI4tZI~__Nu>x#
z*^5TbIze28H|1<9ygU|!HjGWB4o;>UfLMPni-@FQu79koo|xGljJ}0VKa6y^&;>a2
z%oJ7X#ln9w3kzziy}v5*3k6$tX6IOwPria^b!5nDra~{2n{Ri)Ea2Y?zACetJ{YF<
z)1_wMt8hZX8^epRMuVTK6N#MAhHQ=WX#P{I90axoTa?tPH9JY0J6yu<>f$Jr0cyXA
zp@ZW1Tj8V<(_2X~U)RLBRgU0S7At!UMpWL^g8H+I{H6UI!odF}%(jP6zjJGw|2^-D
z2z)_2ae(yLVQ7YjF$dFaG7ASN0yRPQBiuk^oppr4J7WWsZh51oiQe54CJp;RTYl;$
zor80lAg`|9u{gK4YLO0gsb|hL9;|>@4T<&wiNXovIsG`uk5zgNo1OTja$sT*+X4Ga
zIA8`j=b7O$tWZJW1<e-u_>U<{eve1vtI9EUf8k{VNmsB&<`Ld}B>{gyv(?YQ`kft+
zNsdg83_6E5_`}$H=Ci0DNX3zrXa}BOI}cYST7`o)fKKY|WvVYX*1Q0ay|>4QHY=lt
z6%^kh-kK2~8OpS-D!3P$BZf-$L6>KBD4K<5%|x=(>NvzK+5B6KYLTCTOt%H}dnSA;
zK?K%3@YBXz-d|+qDhXa3B<}ghmpP!AkpL!`6#Z%3lYPblR&r3)kQj)VPrLke)f&tb
zpb@;E+QN=S_LIXXwmG^B8Ctqe4nnFl+=lAM4;JMS*aQB}?uUgwtnzZ>?C!1jFS#aW
zd-Far8j~Gd1mY81iovAVeuF?f#L~=(_V<O{Z22&bE}U|2p@`$ywcWqz3#P~rpUzB{
zkeBwGZ!+5GezFRGG5#yEu!Ee0DR30Hz0adU4Zs1H<PT>uo}Gl)YCZEmkN6=xgpZ>c
z#>D+!`^k48qW>y4JQZ0}l>90TFMwcO7GNL7fs1flu2{}=7QSyd;mrm<fKMTYkzRgW
zH+rMa<(9CBh0a`aDn{4{^<7fu%=jME?ZfJM?|)ZfQ3bAc$*W>nBH}Dx&Df5EFcKaZ
zhfoMOa$su@;dr5E=!BmP^v7A}pVvErz&}De;0db%e<x_K37NN!WVJ5<t_6h;60|MI
z_Lu_VEof2iOUrc{H2;AO1x5#QC%BWVaa~@&n150(%xQ(bPjN@FGsBNNfY}^5Ih+V2
zs!yQRoXbDRL@h`wIer6F%0*-VF<=)#9xVqPzCZYUAn+R|u)NKKHmJQz>NZ4`_qHBq
zXN+EJ4KVErDOiv~&Snu3n`B1MOVf&iLovYgS+ssqw7$-Wn>vT8B&P6Jd4Gf9RQFB>
zkkkO0+O3f?`*r5pu2GGoy3iP8D%@1I!j^@7i)$<=-la%{-jDo%PRUW3E7Beb{ej~S
zl*8#zA0a<K+_^s>MImTG^(=qCXoi8JNjh?BVCB(lBe2R~8DSl`HkC<kmSFLPB?CWj
zFvf+Iy+6pL{yZES<g2T|2>_8s<Q)13Z|M*v<vUp2yuyO(2%hsBS{CsfpZ9?)%99`+
z4Jo@1?h`AOzWO4me#d`A0bEFyozQjN!?e*sMqd0yso+cQ-qC^lo%<JCf^mckaiEQH
zjdMu_gy`v%^3#;C5FFv&KKBRyN7K5Br5%p~BC##=<6Rs)n}lP#_ZHzZ6jt~RTAYml
zaF)Dn^3s26RE|k9ngz5D<P5PIVG(TBMoD=f(S~7wm)U?_mP;y~f%o*sVeV%>wRJJD
z>aSo-1xVJ)7(~vD*GhzAK&TCGTOP{QaI-Ff)I2@<?7b({7)XIYn|On)2)~ao#v{*&
zM4jx(HxfldpksztatE;@52e$iG><mFPiv-ZV4Uc3Zzd{YtHgc5NvoO*$(2Ch<^2Aj
z$~tH$M7qy^31(6Nf_}W1<H@!ZICcKHRjm9(%53=MkBmOwcfr+xe|+Oi6_x%1AW1KB
zfiW(?5IwH~y{@&z&|y<BQU_sgiG0*eAHpf>Zt1`W2_~|XPl0T3+77d><)@Ct+Ensu
zm64$oG)rDRNS)84S`XL9=u}>hIJ01%JT&A~s&k?@Q1zkRV)(xD?n&Nf!ianW-Sz=)
zD~D!w!6@bQs%}gMy9|fDh__jR*a41b4<XGq86{rw^t@`(FHIh(oIrAA!Q{rwjKgH3
z=xe}?1tN-C0o2Ul{=QsW#<v#Ji^waeIgZ!refAT?5s3?-(OwX#Q<$m<)9L*}WgI3A
z75f57wYzh;9nyW4)-ub%8xm9vO<$QN*_3BQ1gwW^4w&mU%=I}uuj)4btv}~cod*lg
z&x7z`JcB2>tn|NVc59T>L1N$m*Jc;o1B!DOkVv34LA1ZRsg-6;!YNCoI8Ccxu(EY$
z%v^A;3#3r{0W*k<0esctN$H#dP3MfoUFP@e;Lmx)q7nB;<pRq_2R*ZcqQK~#p{A)T
z;jq#z^&Qo7?Y^4iM)zWf{Q_2~#kKI=*msI7ly4_E&W_dRLujB7W>D0!*n;0;Ob7$U
z14W?CFc%}2lX-^9=6c8olDKz2Sy1SXeM39?KfmbAUJcvb-NgwF=zMS}PEjttRb$^@
zKyUg{h({$@KG)%zA&;HC=-FFK{)~40b|#h|HLLw!dT&0|d-==#(dP!;H(Q>+X0{-v
zc=2y-<DH*8g28UTW<JMxe1#OV)B2oo#~t%swc6L#4Nz!I4OTMvbCIoH5C`#zqvMR$
zkLL%U^6@~ZxmO~)j1#iRY&{v~x%3l<Y^OJdo%th>ZJ-|%Xgx@pSQ8dvnpS}R-DLYe
zJPm?nf7rQ$*guLCV9~l0#F14f+Aof|2QWF*7+;?n=HtP!6H*F@p1&T$Cl~!TViF;B
zI;uO#Py(D7LieRpQ$Tb%Y;o309Q<_rYYmLXgTm$`8m~noR{>YL@uq;nwA!!yhm4WO
z0f&?ANCToBI)Y7<r9HCw6F7jUW^6y4Pd0$&qe4>!k6c<et)%rJh?s}nENgTjpEHY(
z2hW4;H8ps*s=u|ek3pOGv8I{Vq0u#0lenLSL5t}DY9Dhy(HlNh7-L`RzVWmV-{4AA
zj=76l#o#3Sl9pp`G&Jm}7zb#c`t<<zZ?m)>RO6L4zzRJVy%i28+12H2IFIh0sm&3#
z%rpps<)f$HP(A1MP<jp3Z_4e<#I<Jt#9KSFy%r@Oj}Wj+XBNAI1T7ia9=zfU{&k-n
z^$od?gn~+nKlOk5Me@tSCx1PMm#SXjX1OsX8iO|=oF#Q0tCQ@8?aT1Mt{Z;{y>j#h
zx;oLo3A5pF29Q6VsAh0+HopdeO;!ux$Z#m6qjt-atuVfvqjF%et~~AS^0^c&lToT0
z`?KS;Wit&9Q$Z-inFvcFGh>YDkI(koq;~;#_p`i8!Jj3q%=Cd5p%M0sgKuQ&BqFW@
zs|ZjVcLK-F2=WwJ)6x=u0kJx|I5v*u5}RCxn(BVK%go-wAlc60*r6YajC>mS=qWf3
z9K5f?&uw=z^@V<{E1CpWVWr!*u+%@^lJ3ZJ$ZVK+DkZsoPRlLTdh@CP%93#uLRW7q
zG9VWnhpE{<v<(7+jZ;A|te6?IGeb)mC?14yBt6Q(J#0ZD2ZU*)wZZ^S80BB$&F&nK
ziG;k`ozXMp$#8E}8sENPbqChKag=qY;aF(|bB1WM{{V}wo+AzugWi!mIIc!}e@c8H
zh}Fm<?{<uW&-XD-DSHz`kLB^UTV*5XN%40f;l#QmV)bbqW4e(O1k7~m7M*s*)k(g*
z(bQRgNpQJ~E&6?Z>s4kPJZqixURL9G){C0a#2UVqC0TO2ISa8PQPN11@1qFGfJn7@
zo9np@@44^q75v<}o|-{aa6_ohpr<#}7p1TD)p?BSZ$gsm#U}Pm9<Jv1Lcjg_Lu72k
zylM4a*pM9Z{Q)5`|3Kkx=|>8lBUeO^E$4ajIiwr+^a@7KqW7hnDN}`sh6k<;VLx5J
zATV?NOZr772c~ZhSIu@3*u51Gj2jSSC<fcB(ejO(iS<t*^+dgwsfB`<6alfIuK=ru
zLFA}O()jJRkt2Zo)Q{B=i{h{F=|qnHIH6^GblXttlHQ_#De0LR8{do(DzS<-<TKrJ
zN$yb8`g#Ukty&ADe0`{tgM`AtSE;#z(19H|v{WK@h-orIus8eSY6?3spSxr_+;J=h
zST-2SgU`7$K9JwXoMC6c?*Bu>sQvcJrXfgG9PP?C4Rg{mU2^b7!VZ3K4}&~q_xq(?
zhX$9ihN_f(gAR?a*MU_!a(LW;oJ@+oj@<V@#+?w(c>6x<ZAX)8Vztp>@!V4!M8_%c
z#Mn|&rqgrRxR=V)?l8848R|#4sj*W@5?oib&Dlqix=k^%flmWPIK)Z~QHdlRT-Fp-
z?$@QY_5S<U1UMy4hq(hUtWWvn?*IVZ$WT9V?@zLfaSO*pTli6#`>|zK1!S=LH;xbT
zeDyu)b&<_+%4vV@N_i{u8llSI`nZ6!W8uR_f#XuAP2<&Gi~e49A9yDg(@Hg>`|Aw=
zERd?;2R!@X&X&0Ai<_|a{#yE(=k2PpD!!aV&ak-2=jKD6rXp^tYpK5NcX7ud{#*aV
z{UvdRSOFNU=oC7t+zRAsed?tzS8Jc`6#x0<?bh>yws+tyJPi*~;o-*kDGuz<cA2`S
zZXq7kI!1-U5eNNCa1;nEdB97Fq6iYh?>wuq4+f0MjZ$!+kZZ-3DuH`GERH>HY%(mD
z`t(lKKHf?FD;4>VYGMy4(M`a<c8WhcbvbQN&2Pn<o>%&ORI+yL(eBPykV{>Ixij>g
z6fb=Tss8j05zyO$*8REOA>CBs7Yuv!t-x`cP828M_s1`~^_<O}f|B8H0CBBLP#k&9
zbF(%<v|h0how>WcsQ>n}Vy=OIB4DI2^ueaL_S;pycOhlg$0gWbZRU6SvPAHaVo$K}
zybP%W)ML<W`ccB8zojyi6=rP9lvnx8e?C~6cG@=Zx;XLujb^a9!L%k$Xw3&H5Payr
zvfuU>pc1%mZz@ee#?znBM?YS=fq|6%fz(dJgZVsd2e0C5K|XP1<yMJMaexfI3GEZ2
zwViH?Kj3uT-qi0+Q~O7reMQt1+_c+kJxs95%&9#xW&qMUKyKi0NhwBrb%ljcXm@(x
z!LJ*n#5BSB?|D0E-(r_O-jI0o)9Mve(J4^04TgTse!#C!bXK@Ba$#Q}<9f2iiQzLi
z6>$6{(=0x%xRw@7D86s9^tmcaE_ieU`fv{!<xCU>mH>}#?|%d^DkS48jL7!h06ask
z;DuA;`z<}HZgnqzm}vCSeorHia!8DsB0A;DhB8PS9_-w<tDR#1a39=GC1ZNb+5r`4
z*3sFM5%rA_>L(K)G93V1<=vdk!~}VDA<V9Oc7kO5jqT81Z^HBAzB~S1hs*o)_r@t&
zJuN53zCeomi}kJ#|Hj*1)Ow-U3M^V2Xn*r{zGc)CroIP>6jYNQU$Ft)FqGnLlKZ7g
z5cVP|RVp0@*t>gS8oheo|G5xoKBTKUy_WrM^uCLBxw`OJ`pBXyl6eeTt`0K^NzI(j
z_y;oqEQ!ZBiLnRv-*eFO_<d(3+0KkXw~FW~5+i;b<4yrNL8EWKLXh%Z8HpP$*~*;I
zsvaoub|2bTo=cH}Z(-ISZv64pA4X}QT^$m6wpKyZNOWX{l?$H+v4<6mwU1L>h-s{U
z@%7c6A%Z0Siclj&TptI4|5Ga-tiO{1WwjaP#U6bj9pDD>K25<9eOyqK&J2j1`y^BQ
zSx%f8dBgVl5TtCA!46i;?;z4wVh#KE9x-`XnXc&l`BV$@qTU*r5|%&UkdA8892sr@
za6Qn<ISEosI2B;Zs5Wm%Xe>_1Gf|>s&vN#6kixIjj*DPF9BGHdqT-<GN?YyMRBZy&
z@wje2GRTS=E1JFyPJdm6l^3=t3LR@klcNGL+^?3z@k^CA{%}uRwT@Z6Gx=PE@YYJ>
zoeB0r#LGj2gCJRs-Wic1R*`}^d2DYMoAa2f0x-%z40$*t`hOpe1_h6vp5J47U1vG3
z&h2c4&bkIldF~gM9q{GAJ_YkDswZC5vonpk>skGAGDws!`5pW9;jiLHGkX1iU`cCJ
z==vUKeYLE4lL%~iMOL%&*y<NxnjE>K7>1Mj(<zD%{+u)Y{m6J+_YAZfe|)Pihn*HT
z*>ST^YJTwL{tbw4k+>AwVw+#tW0^(z(hgvP>+Qq2!6z;4n-Wt_IfUL)ERRS8IP7{&
z7<gbdc5KnoTQ4u?%2-!hD%}`vk?uLO`pVJo|FQSp@mTlo`*=iFA|WXwdy|!QaYeRd
zT|`E*x2&>9s7Pdltjo&YLS&RQ?3opkvXu&n@I7Bf_x=97Kfk|!|9t+rA9vDqy<X38
zj^jMe<IFE_P^?03j(L1QP$uoW8HW`s2$kHL`&GB{Yff1XQkgfxtqPdHi#_T0>hAFk
zrg}gKO5b>NiqCM|!XEvRZshTjuS*Tjhoc*K+F!CYAK<k#wNK79m(Z7m@#X4>{jNEx
z>Lt2DbK&>sYEz&4#(Y(-n*)pvz1d;aD+4F+k1OxXGxz2>K~7m?iaKuExSlGRwg26(
z?092jM&fn>Es?3NnQk|}@KKG+HemaF2k*_NZoWS4Wa#-NuK2ea6S-m_bk(BZ6wT9N
zD~nW~Kk3O^do9v$U{;oP>k|<#_BCft^7Tg5({K)zh;UYk2BNyVgZmRvw^mskdb95N
z$uBe}8pVIvJ*l|N%MhEw><5VVO9w0lzU%t4tps55t}T||4HWlPXBoqbZ8Vm1Sh;`Y
za@dlim3rim<hSQPf#Wq8+3IFC>9Z>E6M#0|_=}SG)mPM|zF*Ad)zV@vw_g7vC(Fc~
zW`gz4<$tyv(E%#e%#Dj?17?yIvD(xwUkEE^`i@m0ig&9sx?9?Y-w%(uLM%K9P~iy~
z!Zm^>8LOPeuNyyd0$xCrZ*-HL?lXj>E3C-O=`hwm8w(}d;`R^l;2rPEqpNG6I%q3!
zVI>4(Yb<}gy$$1Oyf9x<wc#TS7YX7|5c9g<ATlnG+)cuSfY9>1M}>Y^ad%R=DFUys
ze5MSjP3s1plRti!0dMhe)O1cClpgWB`X9dBzP51o2h62H)L85^Heo|jrZ3vKZ{(-u
z_P<fdnhay;FS>dYqL!VF!N$7`>pMTugfDgmXLlZG45(mmC-ZVrd%JWt?;^H*o>0XQ
z)K-poOD}ySnsq9Y@Z~mpB7FoV+E!kU@-8mAAeo2@c>Mf${U%r?EM7dA2?%P&ZpFD=
z3JeC;EKGeEUge8HPN)<9in}T%58~`)h)U<Vb**o0Kf3A&&9Ha;Ht5TZAOC@EiLEzC
zk%2;I@oa<m3n)#kWQ6nelGz|h7kmaf0DtAan)|VpWSzWw19YS2lma=0EFbfQpS<{@
z!i!D+Y$;8O7sNuIr&!p)&nfxKb0#GxX!3%DNhfPzfFS)sU1dJ|{IkINfa|%S^Q*~g
z;Y+djyV92%7N9!*weK}3pz8i0-@+m~DO;%x4-g5zdG4HcQ6MIkO=Qr~;0!3}zD8$<
zKxd=LV$7#hJG<^qKActVXVZ`EKzlBiK}pZ>M#zEy<##Wr@nV?M0Y#FMcAXRe2Pl2A
zyPCh{g2-&2w7PrrJWM441Z~OK-2ff4kyjX~KkuC5m^1$;-LT10p0rELLyvS^(A9d^
zYvNtL26z(G!=`L<Qaqq02eI$Cdk}5OladZ-DS0pS6b4I(H%~ei4<ue>q{3%fy*sde
zddsy22HwYg<e=&bH|x|uPx!4Ctp}0tb>GGvd<hZ=-OV^Ch?-J7cwc9g;1Ql(pLa3$
z5tx6A%@Bq<$>2{hZ=2O<?L_VR0!4Q)n0L0Qmtv{cr<0swPiQwA#AUseKEArGd-5#=
z8#lV1>RVft1lU&RwpN!HAX0*>7<7tj+z+KKlGwmYY4eLH^|c3x8B;-L$1QysNSp!=
zW7SNs1Ghs93|s-OyxsU1&`A{>L)PR)NV<2Ol9DD)Tvtz|oBV@;@2X}!!WTQ-nTVyh
zzZV7((>_<N*Eo4BCh(-Q*Q@*!wOjRX#qHc>OMoW>X*;DR;d8l-<c~16JFf4FzOc>^
zhI4H31k6{ieVGg87(V^-EL2+64nDl~4sD3IEb4ZjfG3q8uVGE;R+Y;GZqvI*WE+08
za;HxiEV9~cnxMN(<Bp^QH5w#UZqgM^7fiGQaJPD?f0Y{KGwTARD-l_@ubQCHQStNE
zY}3lpg~rE~0Sbci-f}_5Rspxr2V!`TXs6|IG~*cLz)GIXOBK4-6RG!mbwkt!C<;Mw
zP+y_vE>uPzD+8<nCO?l4``n`oq9l;dDFK%Pr!zEF9l`_1B8@Cob;`eiAw;?1KqAhD
z>!l+C04i{1x$820Tj!0^30i6241>k@dowJUH>4|SHt?;VS5rJ9j<R1GAI=xX)18Cm
zzHLR8MiRL!u~FfUxQ_e{%$Og1e%PrR`d*vDs4&44&*{Xlb2ZV^J6&uE%LPpkEb4e^
zrr(MiaG}fedYT(w-OC(IUFy#Jjs<s^8@p(`QI-`;ck{yy#aS8$_Ad{mKYM*(m^k(<
zOiOB|TW+3Nttx79>rm-UJX~h%gn03UOyb6HCcizI)N(-4cW5ZqlIcE@9s$QBD>tCl
z#~mD?CJyTvUCiM@{n_o3$01$@`7FHiS0IR{AH-E-0stb6VRoVg2(7E_t<)N;itf+5
z8P&I%HE#-U@{Vwx#ZsI^LOZR1e7_v-HZOqPVV<*%MNc=XFRqx8RXyUi8yx(sHTDaf
zH!+>Y)q2@ct||A9v`s);2=Nx@iypA{M&AVIF5|eX>wH^;AmSFZ^A$5gxWAJpC4dB4
z*2SKK(P~T|8F03qctN-#>i8j-{{+|Aq=dlMS19daAgKXHT%KP;@y+^6elW;MB|q2X
z#zuU`me(ku-C5XZg^%#T0uCa=+x!(G69C5nsAyAh+<g^$3)_PYM3y7D2~etRf|PmU
z%=m^^w!QAewKFomKgWP}nHSgtqxvEvRl$4c5n}892YyfU=4%!dGbDoZhH~a3s_h0l
z<J!d?ivIY2f{a1uyfn(M|4M=H=GrhpA+CT*L(zDb7bDi?`r6e>ZS??h*;b=+0fIsb
z!dB7@baAS5-T3=P^EiF@&z39c(y2*<gm;So5*Xbx(NR<i36))F2cDAb4Mu$XIJ7?J
z@$v=6VSe0HeDH5i)Co4|Rv`sV^P0~qUg99k?<A~^1lP^6aB2uTJ%$@RayW2*q2fis
zK}6}}=n-D<FBBT}RH4(v81#*oq@spnDtDc}mMq$YhrxUCoclw~S}OMtBGfd*sq=F)
z(8d^xmn=MyF5P~iiMsn_7N`v9k;P-!KF4aSmrZCfUf_((2g&CQc@oT{@`;QkZXDLs
zTA%~0zSV48G4hRHT~@h)MYtp|W#$8!uls-$!_i1)RNrq$d3Wao7{VKK_>H?W?d7a%
z*lkR~>~D}^f&&4eIhSCjkG9t>WXtm(L@g*mCXsy;<F)`%Gub8lDT4%MJRd&nw~*H9
z^sb+$ot{r}5tc9y-o^3#-Ce3UR%qBvZzjVJr6q$7Xra}noCKuoo-YXxk|_=0riYLV
z#<je)S7|O%yGrW5dDrRKdGD#cwfMw`HL;x+E@J4Pr_TIDP`m_=`WHkOCeLq`y^aOi
z&v=~G&{F|e(rYA?PzRDZM0HDBwiI|YLr!yT(Jo@#@<m5jgNw8r`PIOE*9v?687kBl
zvKZ1;MC%=KkC4(3<-&9!`fZ#wF`h;g_58Ag_o&s^)*wQE39ZwTakq>%-*;WxI{Y5j
zE6)|r^Oljca;i?{TGgndHtN)uq}~bD!Euodj7Zt)Ass4yLi!r#1T`C+FM&$~1@Y25
zw8KSyti`T9O`N2O5>xkugvqsrhYY`2irIuRCa{Y5D4Iw-)!%dY<aGfV(Z%$baaqD*
z`^Ndv^g8XuJ1?Slts`GxA(@f8gTP|eS_VNp+_OINN~HCWZCvM9{7kC1BeXR^5ek7n
zUpfXS!U0PUaB<@fZ19S+!HeX6`tUtQfthr+{}2`cae-DTx%OZ_tn~|Uq<PU)KpX){
zny`Y)_k##%v$BN`Z<LjMx=Vzj5KZ;LktKY0{_<;jyMg>{>!KgiKsOe2tEoT2#KDp)
zO$pbE)c01({VCNcUk<$|F4TDj<w0ct9eZG`Dengn!+e|CAcC7??kDp;^{&TIENkDC
zV1t3hIybo@);&^|AZ8LF*ZP%v)u|qnNp1K|;jzxF8xY{qzs;XkgKi@zsZecDyF$Je
zIE!1%qd-r5R%~=oh;8t}%AIkMJIY8V7f6F2`}<<o0@Z_v8=hq0;)h>g9;lP?L6%Hi
zi%;r~F-?LnjmxRDR`U*D(-}PSw}mjcJD)k@jt-Q%xxoe6xo^ejMV{MI>c7GeA+1-=
z491pC2OW5ZHMRF}!Bxj+^6or_o6r~>f5gmY8yVjN*+$K{g&+;&3qd#Nn$zSV|HSC)
zM0Tz4{{ctu!Mg&T?h9Xi-s@ZCg%`J?pDEBkqqY=Wz(o+z4BQRc{ABt3M~h+aZ8zcB
zbaQpPC}TV2QBNz7>?dFc9|;MMPe8#IW7y$di9{n>^oXEdoRS%5b~a9FF>iu$3zrVp
zYga#(3OewgQI5F9<S#Aaji^V2n5kpBO`g3HF}q=Q3V~^L&?Y1pw7Q&uh&1EM7*A)9
zlQ8PBj*OJWV@&!W#LHt(UjRD4@a8FnTYIv21;jV=)Kdllu(oeA(7+PLxS2!Qao+XR
zNQ}|#z}FAx3K(vNv2y%+h#RK=fv!ODWVeIHN!hX3IQ4l*l8EcMxkB@8sfp`s1s;7(
z?x}uV$H@y=d^W#{8bOJ7@%9TlXN#TfEjh^-bE?t#eYubFdcRIMH<QA5gxh5DP*yY{
zLy@3;E!BVf_dMrMkc77BeVU)~G%TawZitWsWJ-JtY<(Bdjfdj`FPVx`8M<syXqVjp
z|Epp%m!`}@+M)uHbf3~U+Xe*-PL)X&B?=2q07P-{K4AziL<2+5)iZSakSlM~ZUT+Y
z4fxT<nVzgYry}FBDu2%Hl#?qoy`Kkyzr{><rm}CrQPQn^lW@Tfu6to?1uf5)LY~5^
zD&VXJVQSPHf5h@Cio42gI&-{UZWrpBWq9)A{@kR9%*yrPh#Tj|j95$3W^z)!^0k~t
zsR%l#brxV4$arsvhLv=Ly1-8aDkqoKC@I{0F3hAR{!64A8NlsjFypt$u+x3>34+*x
zq&Tf>R@66;6q1cMRbRx-k0~spn{Ud<KLWH$UG~{-_Y~`*(78O}5MrHR<cRmGpgk=y
zCGxi!M6P`hAy<5%o-I|o{waJlQ!^w<RzEgG)XhB7?wmYhaqT>pg5k3(2^S9qD&Z2N
zYW*H~aR+0U>}GT9_;2_uei;!&k{Lk8h21cRXD@AMAmdHGU7*OkeMI;J<Qud+esAb*
z23yW_-&Tz~im^J$`{dl0VwaM0UxQl6(M?GK+c*;9zw`>Pq_b3~4~talV<e$bBuApf
zW4>QmKlAsqnTQd6)gJ9J8^BB)<JkL)oS;4l8=qr$8@NI>a1KEH^e1>M<Hir~8kc?$
zEwh{rK4kJiAm_=~`awh22$xU}Su8q3<$W+;B@9plk#Pg%sMskbO#+l#+wu#;S%qGY
zP7u@%Jw}%=;{tTX?xAD8d8E8oiyDJkgO>q`lISt;y?W^?pZb|V4IENG9g`y#D_T$F
zNvUq~3V;*{^cj+>P%Wu_6BroNI@9^od`Nk0OZf2x&ghp7ZEA4Any8&E&h?tC0dU=A
z%(&(;%MlDk*OmqeRVD<<P=l^40FUw$EVOe&lTnFzRSn91kgSe8vF*2!oY$N&FiKfo
zIdiG;0Xs+sz9>vj!5Xy*fn&UP0tPF*1o`z?Be^`fmyh`L65j#8jN7h~OK+~xw$Ab7
zJ;|y=_WE*8R21E%WT;fFG_!~r;-&43zW%ya{Yt>VP;*-7^M}|wIn7nzm_4Zpswn+2
zZ=L+ak(ii^t8%vW%%|iWmqbz;0rZ%k2$S;g@3<w+?10;H<by&F<F=#oTr-Y%j=A++
znrEe>74JS4r!)`1f?kz0DsJ|D(dnL%8e=?BY7QLI3C(SNcRuZbU){v+#6iYj?^}DH
zlxf%UbE+math~ClT^q+4-8#mjz(*~|n1XM37B6H6WtSWO$j%6##n2UwAA%FGEWcw<
z%RQ-NJ@X0pX9YYLvYs+^=y8c}bz-Wi+_Ny|OqI>a<*vC*@$|Y}-)}-D&g5|CHyA&2
z9Q-w518YbyReoa4$-=0E=>r@oX~VI{iM;g8y5pwaD*IB`vba*-?789}cLCBNbMP%N
zP`9I0x0M`wU!m15D{ZD#|CP}>2O05ay=(2_H;85xy4TiKQ>Ju8jkZ3?DLUGolc%Np
zHoy&SQ5#!@lHtZ1#&Of3DwwQqg<RFy-%?u1spRDxz09F!=@)l#P0%>ECiWHyPCObU
zhp629oH-&?yZKRDQ5Db(Ab@e!JYLjI!Mt`o-NAQL6Hj)Eh66<r)(q~+8ILldU(E4p
zvnw^T150IWxL^M8=^P|WU?f*Y*WVl1<kIp#?tZ^-YFlZccA6iU^KAv4y4J$7s`(}<
zxKv(fPsrEnbj>Mn@O)S`K%fj_t=}Ohjk_7sp?`sU;L0qoY|3k46Mnj9Dmap4<vL;)
zx~zQM`IMz;NSm}w52N1ps+wqI-Yn9lvcRX&d9E$f9^;d|*m(!MozE@J^2w<Fo7*p@
zN{}i53?|TM7hM#(;oH}ndA{E5u0b}`IcrH4%5H6BebS^AJ1%~66yQ?%%jU_@m_~Yk
zfJOpyk<rFd)TeORLIZDw^|G&8thWO|iU^8K(KBL2U6mPJj~=SpD9^JB>tzU`yG`Xm
zWyutPHXwRcMjjNM6Y?nGI;E_EZ?c#^9{Dm4mZc29M!`95x$4Xf*Zg~!(rYietJ}|@
zt|AijaYahvyd7$F>7QNZ6Ef5mz&_NX(nN6y*Y(kKFAeDYInyii{pK$`3XKYVP2omN
zOO_x1nktW>NYFuS5|8zwN*cp$uU9!b@?MXiSjrReD$GxxD!u16&+2*I)aqH)AC3k@
z<g!WyZ7JX1y_eqB;?bLWnS9kb>tQ-=4YtJ2%O;bnGQ+mXYamZsQM`3^R%!~B-ATc9
z^Utcki5DI@N~xev;!CV{>}i$rhFj<DpQiPTyFZ}BVIZ0%$?!twl8q=b_^v>zI6r4h
zO{Shn?lRL^gmWdIAg<+7nFKB&-sPsa7Ny%qpjKt*l#{m^x6pQ*YkTgCgI!?=l1N1c
zc%7PnnpIDi-aWB?0{yW|FZrf*^0hsV4PMi&aE5-W*a|&g`@?Ie-zPJY;`LrYql9AL
zn3O=3Kd&U=I+2$yy#c`$%Ex);6ke-yeTtEGw(kq)q}qD@53h~cgBzb;+n7+)H(9Wx
z*^0pipQ3fe66mX>r%O5YoSYt8p#*9{>L6;WqS)il0FZ#(*9W%r>P3a$b83AfW4+j?
zYC%<Lgqqe)#~<@||D`!_0iRc!Z>q6VKJWy-W^&yU_+g>7jW;qYimK-hk2xSUR{1J3
z{SXUqz!kV&A4*C@yyjHXafvp5#1Hchwb{|sZ9hfm&*n50`qs<=K(9NCN9z5nUC@8q
zXRvmDbNX@fuG0@t{RqIHmaOVvdudE-jrX9@&1VgrbHj<DbhGTXLvQYG^o+pN%P%~&
zvl-zp9b=GyIt&-e0h-l9^*W43;ty7MR6TQkEe@YpGo%bicK!3^c|mWSdgGPV7lRdr
zvEdxEv8#-Bpr@$*86BZ?dD)(H>?mYDqd*KtwP~Hwfqb2?29og_bb=d6?sC0K@j|w9
zv4WJixQ`BMsQ!b04o2b1d7k>@Q(ogYA&i{cUU;r^VKUjZ^ZVtKWCNXA*#u)&OYvWy
z@1T;7ZkNp7%^#4w{c*+W_3BfR+i{+NEtfN5{cmmz=J43oiddKm?P&P&${#;MMxPuP
zQOXA=M-46a5ppT3G#%X^*Xu;HJMyao>u-$gXB!0cx3D0LU!3@m4O&@*+~@m2^2HV>
zVnb^PjZX&PAB~6Sm#O;S>)c;b$wE<d$OXNE!+NKCmZ?70cZ4NVm%XRd=0P&Az0@`>
zjN%m12ldAt={FmVJ9fhPCFL5nQGqInT<X_SE%Ny7uY{Efxl453<{8*v@>ICOWEo0K
zE=wq4s(uS6#C1Shj=#Rh$wr;$f>8hlBld^*%?#IKUdU>%wO~a;ISxdk*PnYODgb8q
zG~DnR#g9ZCoompPvk{K1%l~=jh0ScEX<c@L9H|2bM}{eN=a_Df!A`K8ELog5A>LK=
z?e`TlGR{6aL6<f0B4WuzPg>nMgj+_!WBhz*Ag#$2P}57$^7Odg&Rwj#n$Y{0!*2D6
z87pm^A3)_DVs0f*Dy9)xeD;mW!yoP(IJKd4)&d1yV~@vrAT_y0k~<Q3<n&y&=A{jf
z9MdPA6+9^aNWV{@%bS#+GlZ|Rk~=2VC`d}B!3ku$AQfS5w)|#<ZL>w$X5+nnwy?(L
z_7T!#D}AqR>a7*elYtt6!BWMfW2COQY}j^dFV4Q?hRL~3yWRrM?NjwHOQ*2?Xs^VF
z#G(&~O*x}JKBrv8$hpRrNUdZy;?<^wHt77+Kd(KLU02lCJw)iupXjh~IjB`iO#Fuq
zuPn||yJH|!9T*#MYX@G9pXY5hY6MI$j47AeBviE=F(<bJxWgaTO5ZAG%PHsP%CXM`
z!PCp9D`&bD5C`szw}M{+5HJ@ePCp7pcIDnQ7)r~S|1kjahE!4-Jq_)SkC4w-!u5)F
zUHu-fu0j^xew&9>8%W|#p;q~?6uBQX%EtFQDWjgd(3^@!0Q$gfCqH|eCq%$K$A#WJ
zPo}^#BfE-%FH!M|x||%T>@TX#GfQL2LSLmiwSuCC1}&8FWC7$1ek{O#0(V|Hhhe~r
zb2wR72>+@ZiEh337s)Wz-;e8M#)MbHqB-6fi``g)N{!96_y!9J)7wx;Vs$jJKZS!Q
zMF@hWYz^nb0J%4w_ya0^3zOHb;N#%ovV9~zYv?g-k$PER<-(908S1mAUEy6=8oVp$
z$X*GLMQ9(q=rlt^E2D9K5|Y~0(9@3Ht|+Hbb6zNH<OevC=+XJl;jIp-q~pj_#vXEj
z!h|iiK~&xey+hYVzw14DAGZ0jbzT#uuaUlUxqieeBip3r#stY}wX2Yb+j!5NaQOtN
zr(XHB82)t-Pp<uKYaOKA{p743ZB0~^ua#+Lm^}A2tOP0BKng{nm8o}O7zC13-3~0r
znUCn8jko%BHdxCPyWSs1H%;Jr5dv&)GW=#%JiG@0Cc`8xlTr{?I|w%zakgA^!p@Si
z-3@<DrDvFKO2BJS+5L94#y%=yw>lA9G~!9!4^qo1PyFm1he}`c({xYqd>QXXwL0O}
zHs)0qxPI=O;=#}lygl4V$NuI74ZK9)vsgc=;}N4;L)!eV&S76_qie)Ph;;hIDo5|?
za~qR>xhQyQK~0X?7^BT?Tz5?pEH7SQm-MFE#Ak}KOh7Mkb6x{8Bj#>I#U-SGhD_`f
zH4Y(1rm;v^1jX`^@78#UW{V3xGr?G=cNw9nDV$`{K?Gh;9_@L}tfGKuhh)z`12CY7
z;6gJcXro`}vYStT#95p;{GFIyCvJ1LvQNE&``%;G+ef8tm4{iugQw12=?M>*@5*g#
zG%k(<;flr2Uar}G+N_6t(}m|I7ei%_Z9JIunmdxO{;E~NN>AHatH&n!@e$>kHKbIV
zmC;nuC7o2$+yHW7?(CX$jjxu1c!73gh}Tc2rhj5^`&ahl6h1MfMz$C!t4>)Bba{A<
zhQoyu%?*yA7tN+1%-?V<xdMX=p;2wf72F$~yFUQH0eqv$^r29jqwl8Wf`FB8jy(<>
zRer1~@qBQ8JeR>Ub=`Y4&C+cn_1@J5{eeHlRNW=w9N(@owk7uex>ENzXiGzt;Nu8?
z3?YjNqzOMdhRtYj#^f*tQS(8j)(H?u>~Lf0ZAqwk)+an~q}Bref>K^%`h%ELe8<@U
zbTXcB@*fafJ9F-OuwSbDNLH=goXK~PPDtZ7X;5klPWeWB_7-38@~ab<b7LM4hkWIt
zjcwvbAC^U^b5Tu@o6(BcRsxTp#(K7-8LA_(EgXDT&6ttIGZeSgJ^f@{m{9Qbz(a)K
z6Pn~dQzDp;Fon`AIo=!1Ier*NYKi)WOe&&%=aNsD71V@KT_EQ51RQDlnn<Ole1b6c
zs!=^6@J&7iT<X>0*GUMP40X~@B)wv2=3%HmULs3~OqvS70hNKR1(Th8pz5LSRGr5$
zVm7iSl}2h{!nznE+(<HKl%DUe=MLS*suwtoBw2@aB45IwWR9HD*I0r318f-)=rj0k
znQ$K_0W+Ga>3|{!h3wr-_9O&@fPC>dV{j8YY!L)87I+}4DF`)CmNUgajErH78m-m{
z@OxPn9~kYvKJihG<{?nSNMVh8QU%Q_FqFv-PJrQpZJ`z*87b(Opz#9a+-we6DEAnK
zhor#x>KpG#&RXBCZT&thpSmKo06FflIz`LZeAt(vfNBP{PIB~E%o;$<lk!rO)|X;K
zm&dhI-&ftTGdZu#iofLTD}?gXmP28(<^%CSlspbqvR<$F08hs!55odb$H{5cEdH2b
zCyihOa0VHpsBn61EaW5IchT!;uBhO%gwsAN>m`6-LHPP`mnB)Qihg)_;I;7g+psGB
zV0iL{$!TS}k8`bdLvN5#*>(9DYpUbVzdX7M4B4kwj*Z8)%mPzNWb~eeyVa?&yv3|D
z`5o9~)iBRj4fv1Th&V7_zz7BA&j!#GIQ_8qaq_v+c`#l!ckj(#v-u4v_uTb*zz>Su
zg^~?{itr=XS`x}-m_O*|V>;oOQ%ig>v#SWy3&$1dJ+r_Md|%>7nba&hg!_UT+5uZw
z;|(c(tp|3mY&pP~$`|ja7UA5%rYSWXIve~ssX}o#4&cXyCxB_A8Sx_Z#hEJKfIXWg
zY!f|5T1pSqKMXp11_YzBj!k#w;yM(=%otkkk5c0imF^Z<DbV%AL~4nHQPEpFEp#ex
z5^f1cumC83fy7qg68EVG5>?(`X`QNKI!PmBpqnq0OPPZ<Yi1KEy-6PpW4E~&WITW4
zIY?Im%S_`Uj9!`@tE(<IYUy14J(C7&b9HBctM&?m|DUqg63>B76ebRg7$;z(s{6<`
z>+5{QOa>2Cw*$TmdF4(nP|>t3j)qsFsb?TM8RfE0Yl$d@6k~UbhamT}C*2DUx2ZHP
z^+P|$(Nlt-fmf%yqk;M`xDHHJ%=yU^m!-~KcvJKyeh@YhoG3-5uW0xoTunS+CH0qE
zv;lsRWI^x$=LV_t-7jfAkQxJb&`ww9qDJ~%K#d=Zr<1N44SYP}4QsHv^zP^=w9;d|
zSYuu)y}{PqktnsQVIsLmkpN?Kp!7IcVp5{gpU;linrfs5i=)wki%$VF13uYbw3)r`
zBm5fx1z;HH$epRoAm*Aadm`<$cI&u<iX3@zM}A$UAf3E~I>P+Cj~n(aKcEAEN7|_$
z^wiqCF=W!~P0eA2LCEojGyV`SH-ua{6u<`f!#0J+`Ed@E=ju_VEv=M6FksuGe;Rir
zV>Zh2Zw1`Z){4+Y={yEN{m%AVdXG0;gp3~v7bhAUArJQW@)=lpnzxCr|LUo*M4j9Q
zJS7DKliG`l^8o9!%Hs6_CHZ2<>9RZIv4<4NYN#qTP##vBt?4Bryu_{wBXmE7D7&dN
z{m#lqtAjv1sbn|600or%#>hBMU1MfmY|fqhXJd{2J4z&yKjRXV-aVwe+M5-938w!c
zvJrZkRCo>@c@0<|Wir%C6rCBNQK~qc0C{v~`AUk);m{186Y<qLI{IguYV5<n&F?Ni
z0Awo5=XQ@jVNihA=6rTTuBJLN5Lfg1(ax*HG{KG+$VvgVPEA(nO!u}rhHJxtRs7w0
zy5iakT-UE4oa6fB6M=zdq!)}S$;@E0WcI=<6#g~e<v2Y=I$&DGF7lG|H4YxqY1}H3
zbR|So3sQRM6pQ6I86Rg{QX}FB%?cveV6kowEFXf(uh_;sWaGjl)>45scO0yz5Fqd#
z4X0@joJV>OWqE#3BbW{@2Q>&F5B-3i*I=V)Na4th(DH4V1!-~4(AaP4&2$+#&%^yH
zpThy&-Pfu^XYsDy25bN<Zxp^xpmqs3Y{^#TucOuci;PF`Y3O}dG*@*Z%K4nY?LEiq
z_wIkdAPhVqBNddr;i)din#m(SU)dsKEAT^?+Kc@+!NWYLC&_+W%`P`;AphGnmd^}Y
zQY@DPskIwdp=!ei{oLsV0_+|*A^yo)1zf7w^%sz532R^>jDF@4va@xaOk3MMcvf;A
zP*bAc92z&i#OW#MHvq|I+9)KP@GK1}Q0MtTcyS2EVOTfmE*6;IP^J<&8yp1`TE#to
zLm0e6>^S}esjR}A84xAA{J}sAQmK{is6%x~3b3Wf?`8%Fb{3cy4_|}EHJwbc-rHss
zSB4ZohI{po9;aP<U4><bi!5Oni8_$CprMC0trBL=X1D0QGptCUeVl1q4^e5gBJb?)
zcxv5zU*1B#Vm7hK!doD?1DyT}S{YNPJ5>ZYtjD|n{;tdrd~0(kj5a?gQ|m;&WqQy%
z)APB3_?8WVgEk~jB!UvZ3Bwc>5MdA#(#986lWY!UI(3S#8J_+lMK?MPs*9RbJw`uI
zNaXdX)*9JK>C66l`Fhv&kSm=)!ep0paR?3fh<=Rtc>Vq#;!2gszCgq1-0yBg|L+cY
zfB1*tSc(sJZ9V5Vp`S;=r0IEq<D$x|k{$}nV-baf*7kQ|bidwXPj$8>Gj!{bCq6qn
z=buF$g!;h<3K047$L9{EdO>yKJ-CVJ$y{9dr1l@`ns`FW5Yz}#c3TGkgQ1N+KW>ui
zhCcfE1oZz%LBw%mnD+FA-A!TafIKDZH$gz(-X_5h|1e=`lb~g?LN(g()?|n}yIa?t
zT=Jm$VhZyG<wvP11#T)JKWo%XxX8-<^l5SV&xNz@pdS#om<trx*0|#~vszgZ0y4Et
zP|eT5k4<{3Q~AEauLBZq%`7U+lY!?L=sr+t)&&)fGUrLz3UeN_wng&T*M$l0HoLJ#
zzR(VIH<;I4StOZFtKy0kD}l?iRXw-ecqGc^QOX3*no;Qi=P-CLK|i*4bC7c-O2}~|
zYO6+95z&atIZ->4c>^V@)u@BQ*LBJI`o<*nd|Mb1mGcjLID5l}TTpm1^6n1_uQZVi
zMnFoX@L@Ufm8dcxgFSrYHqDKF3e3lf`{MMwwb6?d>rnHhdg?wt(3ldVF2T8x>CeuB
z^-m}s3P33*CMu1z@2w*vh=q_-w6^V^2jmP~r&tH7m9gg+DsRozuY+x=C*rGIdM<iW
z(8U{-K7>A%<U3VZ6{=1T=n8a!0$rXFIFr(=x<SlTxdKjkK|wa75zxJtpr-rSe*mLr
z!xC1Hw%!CFv3I=V4^TuNi;}_x;?n?A170zq6z6!&tl_><2}r|tS^xb}O!VM5LgadU
zcA{|y3ZF|c7<U}+-m8+rp7nicW25h{aEs%%Ud*EN1s8yL?d?&BPJsrKPvQ!<?j=1M
zIw#tFSQDU|qad;JG}23s8U~cgThM+qP9hJ27SoevY`BEM#83!dfnpR7jQ7qUd_^Yf
zn{a#Y@)HvY5T6a<qrL-a;jYai#M8b8k}8J@j#YMDdP*#)CTPabkVi)=`?hq-Gwaeh
z85>+6C2zumARZjhE2|tYJaL#&3e^t&6&esH+eq8_^QK)-qOdfCJhsUev2W*mj{;<3
z%f>7M@SC1Zgp%Dk1|;0wJ^6<M838dR>>c=QxF1ke1qjjsjNLw7K-`WL{npNm5Snvf
z&l2Wm+w;GN1*lmu7<=ZLQ?LINbOx6mH)l=Yfy%{C$X)Y@7oc4lk;AKNg>0O%%JM1h
zVf6>hGnB%{lTbrE<}Cp&&XwHIzEts&!Zg0pYUBk`2Yc|fsVvv4_XcB6+Ch>)%bJ<g
z4!%D>GUG_Ab#hqkxE)OdSKvann)F=Zt7p4;798+v-~{oewOp9?7LMgAy6X~BsAcC|
zd$9V-9BT5&gI_!;2F4h4U5(nnM7|e!Gn$opVU&CNT}+Pi<V*OCZOUBZ`VHu_$zVvL
zAF<7C1uEf3o+n9Nz&~cNp&GO*gr>rWkA#2zS{nTYkxi1aNb*rt_=j_B*sEa*wFpZO
zW)ym)Yy$TlLu6ROs`^MQ?L6k2#@zylFvMM=&_mfR2lq59`jCl*)|~B)!o8<nSHVKW
z3k@d6zQG&+7`WgfuzgZP(W#EQ_*e}fAg$JEtIUHl+L1pOmy&_Uvc|I1PX)vf1Ly?r
zKk-2vKZ5H@={2Maq&c+*2%r6?AN;r<k-vlr@{&O-G=#}4RrW1LbgOgYgv25$M{u>F
z>pk=}*-I@LzjOIr4JDatOU6sScBFK+dmx&saqr76FJ(_Q?fqc+X%n<H!%$OS-<mVY
zk@&#JW*+Qm(VBVxRAuNLd_~<ffn@l@jj08OX|N(o5(_n#r`zI7po<E!=ewJam>rzK
zMZvkzwFrlZm*23)d3|}(gZL@9$mbB+8Zh?G1LaQ!x|%)pR*ecU1(tABb()9$7Jx5U
z%5IYr7eGP%)M)}?htYR>A;6;TfEU#4kRl~nK?U)F)r(pJYg4{O;e2Rsb+b|fB+SO4
z8z}GQhcjjT__A_@Amwwv?O|vrIRNLU;(0e+nUS;;2&>7-X4>&S0|iPi8-z=hQ)(_%
zF%8l$^ymGAq{l)m>mx7-s`Z|K3gf<>R|&ZR-e$AeX152CwZMfKe#6&C4zvK+F0Fl=
z`1ZKuO27O`qpThQmB=S|ABQ<`ynhNr)(8{*vg;vQm5E<g-vefnS^J%sB_VOI4G$JO
zPzO@cC@#IQsoeq`x=9fi4rA#C?&fTOSZpkn#+Rdd-|s{F=-k5_hU&BM&~6Sh`31!;
z9YnpKK`GD^mV`mVO(_%41={^<w6*_WVgBg7(S~}92r1CU$VR=XbgG?u0^Tp-of2d*
z&b}0U(D9-~lJxt!A<Q02)R_YAZZ)j<y#S_%P+N9yA=4TE2$)6*w0p7yL!$#nYED;<
z5y$qknN&$=BidYDIBCPX+ZeOE<l9hU_}I1ZQWIg8L4aU_vswZzQKkNok<6-`fq+le
zJ<v1qx-2orGN?utGk;79IP%^ta9m9Q##>7ccna}=Hv}a|7PElUv<kxLC0DD}eQPm7
z$iQl?4)Yn$uK?s_b3Ju>{#{{3TbK5fFU_LGb)vHz^VrlWOJ1r^x;BPF%FLJWD3#}j
z$}*3PC;l8@ztHRpiJ$p&utTo>YrqtW8@Ihec$dO>faU83kZzDKGcn1_gFB#Pg9Ee8
zSH4Vr+z(@_0#ftFfE3~32X$)+mQBVXlK{Lb$FX^dBxvQUo4%EV@JtL_>f;!cBX^!V
zL_96m#6sIUbivertzFQ0xdniXP8f~Fqni&DV<J&&S60k~SmRS#K+@Uk>$4dJ6Zvlt
z-1QkN{I>l>F<f4)Ji3E>S?YVR9r^uyM{^n(ftl>qWeEDpWyCiY@*2|-4vK#0E-iU?
zo&6!s+HO=hb9V14NK@RJ2668{ooBietx`Rqbu9^Ko#wiX!ok?ZwZt){p)YSDPR~4M
zSmaSCjv-yn+PFyu{>0o?tJ*O_=&nnHiSKIMR`HP&ue1nXV5HAByO6B$Srf=zgA}9q
z93L$@YM6o%sD;||GN&(!B(kd~TX*d;CwCK+b;AJipe}y$JAmB7sR)R~gZ`;voz68J
z8$pP$cm0V9Jc<ScKxz(W7hi&?LmhXxCS>^|)lLi^rj}2VkrC*NL8qXnHqwQ-SgSqW
z&_y{mB&8ae>;>1pljlj)?egISWm+4iYg(4#GX>|=#-rC!&?Ewk`0AcY3-9RN%wczR
zTw!&j8_{O~13(_o@3@AtR<_AHSojVc&a@h>aQGa;Tjb7PXnUzuASyMLEXKsLPHY#>
z8=+dXdGQ9Xt34@h(zi;nIA-iyyR>L)*+2_6-?+}B2sQJPgujx#xhm&U(i6#4&jQGl
z^My>=Y_jeJ6cAU3aizI-iqm%*l@Sa&qwvKd&xn6A)&85u5}JntdSP#Bm2+UQn(RvI
zs~{Y7{m{ykA#NFc1a};7Q6jsOS|yfa(m~sjcVNW#d&|j=EZUK2hw3rQuJn|7T(*Y~
z)h{M~fyrppYxi~T3?E~Iy=kLwAorl^NS5*}-+C1GX(Xny@8UJ6$*(2aE`|f|!2>+x
z(if`Xu$6mHPiF2Gw2?czei1hBG9TWZz5FDg1Rt8QmqDv%<W4R@xxQ!Wb*C^72@*=4
zE8XrJNRNq~x+Oc+hYh%?OHb*+8$yse>}s7;+XRX?OWH$~HZdrh5kO@G9Yl1-MVl`N
zy=G8@?x#X`xMGF?Em=k@Y!JfC$3E%dEBVf&ZKDpr#B_2PE^jdo_-6|d#MiGu;>_v4
zTSyTKQBp#tIJ9B|!;=#U^}CRgVA)_qc4tw&kF_@l*dIP}v~;Fm<126VhwEMGOL90Z
zhY(l|x2o(+QtA&y{&aiMa<ezb`gAja+O}psv<8+4L-OKzwwuE#1-kZ^!a*$9bKGll
zL(hs9f0tDPDQdxhNAZ~7y*i~mmIV0B9pXP-)l56|Y>J_~=lu}wF#9rqV3*RQ$HsTS
zN`n+54pOLkPBi<wp*}-!g!@QT+~s)bWEfg5INC4Q1gSOBhGwbGcv=SNHNYD`;bjY+
zRWFpos06`qzDjD{p8*rnZ@HS~{`u`gKq&7ugd;Q@0yBPX$SAyfD=}9_D+lgcEcSyM
z{vzh>b#uEc%+<6<P{;p}T&^&c=wW%<!`-o{;EIjm{A;T<fh+U}G2k*Av>+s%-em&=
zddP=6-|XFm`ZZ)#5FL#qKu}L=sz5^?ZL~1Ye3CO5OYmjBz!)-9C<xWPcVSyi{yJ=x
z_H+WVWU@C+`N;#;RQc=Opr46t<D$;A7hp=i#%`KGd3vg@1WE_(_ZX}z@@)k`Y44~U
zAu-AD8?4(EII$Ho2~`EgNZ*9jIPs_kxoS&d$5pt_WPfJfB=eBQqrv|#k-r4ul_AM1
z={bXHNV`>{s5a+!d8@mo$GimY)8eYtx4rwcZuGIZ?rXWL7sur-ga!NS_Vl`zB4U|0
zG=^=;`v-nOE#D(0kp3Ju&9&YM6XzHBAkaK8W#oO%;o^%T*o?k*d$=bQsYOX+7J5_A
z4>JGZSw+~9HwpM}f8MYaPz2v0RYGsm0{G^zQ^=z?r4yK|mwy9X;Lz2ty{%+pYM56h
zsAeStv}#FZBvoIc>p_As^2MJiLRJrE&S|wiuyAU9YJpUscV7VvCK=Khu$<wnwEq_J
z==$46^}gV2bH2v-iAz6VUd4~KRstEJ{&NVW9GIpZiF2OcFP$G-SsQh-dr<x%H@;T(
z*nDq6)|$-rdS<Ij0xRazXmN@+)UJ`achh5s&Qr>EbGu&#rOTqV7f?CxOFin~53qXi
zr*|8{!F<#0Wqo{k6V<)v<PCZgOU{v7BJE;xau#6PlF$GLhEs7{+E&<6c7jgPR;T1Q
z^Y@d`XmCo#@3}rV$U;NIaZO^@o)JOJib(*I`oBEXFmtKXgtXrA5$FcA(~w;|L+VXh
z9AJlS0%fh-(m-(|RH^GH0)#$aduV|dCH+1VmAzfv#s<<Hw$Ny1Sp?{%dZHj}^Y)^p
zC+#rDwbUQnr+ON<S4#k^xF|zoWao`T$^&BcA*eTwqv9o$KIId{4hxytk!gt~wX{CL
z_Z4VGb*`B~B)WEw?6jyoBuZ<^M@K=x>t=+-Gw9L-ZRP06FrGcc4S54aFpU1i_=<ZV
zWI!Vxz1Qqn$@DuC>JSs??5qxIx(Mg-J5Pq39GROQVE2U>tP_Tvd_~avHNb?ISNbkX
zc0ES2qvB60hGw;g3qDnkxg!IoRZhDAhc&C=l0otXP^@=LrPVEm${(!7P@LCWd8}+}
z?FY*z3T3lronkSIv_z-aytEQpSzgvcK^RoIoIEZvqY#$rSK{P?=T!<+uX4+v5sGT9
zR5iuUE%wOImKurhy@(86M;fBck1-@KO!to4M}9;GtrhbV$wX*FK9*#Dwhm&#Mg{y>
zOsaVH{q<>)kLVZ`{R+@z4#6`Nsi&gXsK-MkXjgxS;-A~X`wpEBh#^(uI>*(W%|%t4
z({O7WBue4hew%2N**MM>fq{g!L)(w{52U~xYvdM15Q&MMCFac>SPiRO26ZFI8hmGr
zLh^tOamGe@Iw!6Pf^Fm|CTZ)|-=xylv^E9fV`*tU1!W{n>Ou$%d4Ue_{3<lHZ0!zR
z+1GtQ@)u~~lO^Q5ktn9Lm(WCWVcFfhLhaaNG!1AZ)Y(>PvxYn2xIPEqw<o!hi%U|u
zQs*aHDDDZH+j@9RKcbrI4}o+?7@(LPpys0_ZHCs$&@3_#p+h7nm)dpG6^jqaLA8g<
zM!*TRHe`l0kR%=*)!4e<##h(wsrpM8Kl>%$6E50$-mf24h(Y`P@<Wu**+s5)wG%iY
zSO9!_s+~+W`9OM;W$dZ2x;%yys-?fdgF?fa{x9x0cg?p*?^f6iF5+ya83pC$qPvc?
zfGqB9Sd;;Df2tQPwDjx}8X2RG;#l1O4N&VM%I2ahz#fa@mB;Uv1pK%S<V{VY@Z=5{
z9cqA4OSu-1+7a@#ZLX|XO|kPT^l}-1Nq#4Fg80jh_<Uc{arODaqJk2Y5f5E%*JTxs
zk`@r_m~PkH&H!k%6C>ZuvuMlFE&!Wu^+dPF)p%3CblM*bJF&LS^YD@Ucq|t!pulbp
z!SCFty9pH1gMfCI0VSR%vDFHd_vs`>si_BaG-_0<#kAL$A3eQt<gfpUJd}wvdWuo(
zZJEzmtj^T@MH>c{JICk*d7N$eFQ3M7!rk~dXYpfXq3j#f-=5|H1SIC$HK6PRT<^=M
z)mSfN0CE3t_y!?&NPUoT`uD^5G#M<#-#wxVscz!59M%~}4{9Dh4prPI6epo2=A5@?
zU~rOSZ61(A+5E5<G+U3HF#CfZwSn$C1sbu_{acsnH0nvMen442Op4QbGB^)ZK8vnD
zL)!R<B}wt_qX86$fmV6;dSR>yqp_an?d8P7zd|#IZzx71aQy=CX7)BH0Ifm@I1ek|
z34*;=baXV{%4c;8o23sVtwh_c+$AhFzZTh+@cT`lL-+`Z+zfbD($F8Y01?V3h6C+t
zhL{LZ6#+7CzJ@KeXhfR-_A{uFd3hAN)LF8T`V(`ow<Fe~K#w2Q+BgX`hsG@r`Fp}#
z^0(6fc6WTBU;hk4CC3m6@l8@<nQ*MFLYGg<-B?*|NDnL_4qv$e66yi2b$wG~b>AAU
z8np=3eYvvMDHwQM!<eUqj4}wRCK$yq_Y%!B7bV+M%G~|j?6VIUC{f`koXu*D`y{rX
zE`{z!vDQ6I-k0Yyu4Y^$LZTlEF6?{_6s<M#Z=h6=7P#R-DE_9b2iArqNeMx?PkdUY
z+W;iz%A@IqF^@Z*Sg&iKZy$!9CB+EdZpOo?<bj$Ks82UP5qK3&lv}6{e9)C(#&oSM
z$C_OBqq5av^ZJt61yh|~9-&RoaQP6%olDUSSL)mSh99LC3@R>4L=+nN1(IKDroayC
zWeGh3&_!;pin643KED7uY470CQ6=qmydpSAG-DG-FbiV^$=*9^$)O`>kb1-Bl>r7r
zi?d5`#gu{q=ojY$%Z*zNS4M&L0I4%5FLnEgkYBif-%{BC3~wa(a}`FPUHlI9bPEve
z@=|@KhzQLv5YG}d<l8P}zrbl#%T_HhC8+&Gdk$-^s4w$TmG9Zb9!~rWnEjXpl{Elt
z{jzH8#TVYZriPxL7iw_j(zFGd&*V!>5u46yXW?8X=H>X5-L(WwPqEKt0Y`G7mIN@R
z+-&ZVh;=s>!pS5AQr-_vIa$c>eEY=dR&|DJFP0{}c)aHKbW&XS2Og|n9gt)mf!{{|
zz^7E@@c8p(<Y}I@1km*?CR{LY2AU<mO!dZG_YH;CuX%%b<sFjTCuLeXtg~XPc^czc
zcn--ON3=ddnQw8ly2RJE)hC;RF2C%IM{KI&1!5ZOI8b#Jvzv1&I8tYQ=Duu_N|%LG
z*grwOKMuvI-#fH7g7+*Pi`+GE^+J~IrN6R{6CrsW5{a_oA}s?C%ptl5x@&2qicuRz
zPyhE<kqTf*f(%gAL^hJlyyC+KYUJ0Oz9FL=?2nL_Bc(h_2nX!7$ZPp*5uME~=n&WW
z#8V+jYN1i)XwiMp?20TC=h@{Q|8(Vh&0iSz4m6taIIc<~mMMGf(SWMnTq?5r%aH+F
zN54YX(N*Ytda$%S*_`IPrVml94B%rk8nBGIJ7<KDV1bhzAou`8nEy=GCj#2!^WXpg
z+DlA)NDIyAAyl31G3=50mH=x(m9|Lx)oA;_!lwUBrO|?T{cXMK-wShCkV<SusQF@&
zFJqJ^4%LVVH>X)c^!TjYRS<p$u^j{bw<*=a+Hr{;*Aajz5Q=s++kj;a7f3;h!S&Oz
z>s{#>;K+b(4u?Zvy#PQ#**<d&xd1+VkQB_EyxT0diwZJM;zFc3S94=(aqI&iYA21{
z9-{7u;x<rlfOgmHcMx2Bo0_-sWsSuFZVaO~BFd<lFASmb%(Kvi#OjRmFebDQ0Bi44
z14qvf4?*Yi{@NbG@h8?$A5I$#$=oAkFe>Bbf1g!Z%NYNVzfl9x5l!SZSl6{^6}A&L
zM%fX^Ha`y<r-<T~5dvz<wsBpBsuvXFX4{s5MhUfZWO@@I1jSolA#<9{|KQkoy1vMi
zfbGTOM^eroo1k&lNbXFAQ{c^m?+>&`c(D=di?q8ADRA*5bacJ}ncAf(jn%iHP(AD_
zIPm^GmrgF^(-yG^pS8K6MGa4oRRfxRIokB{v@D>1Pm}s1FY@ZO2wVyt*Y=U7Wg`Tm
z3GdpWr442<^0vkEpXc7*oY-Ltz9wwFe<1_9@RiIz|0H=e`8Pl#tieaR2EoSt`Y_c8
z=|>U>ynzN>^In4kBXuBDWro9zpemJ<>=K4zuUO(PdERR*?;eMF(~>HOC>iWxxywaR
zLY*+CJ$CfqCuqiuCeU<5u&_oW0FEtaOdxwiTETbsBe0zq5nW#6vPyA(gqlmIlaT%f
zq`S3JI^DfG8Dbgv(M*|t2^V-iEK$sgKQ>N$Zf!Y9#$dFeBSVRQ6cViL#VazP^a93b
z^%hh&UEckX%8_7v6Nwb>b^5K~#I4oUP7cM;kEF|Gf{Q9Hw4)bof#Uq0okB<Bud}9^
zJA3S7NR4=%VpI1)^B*bEr;@2En@Ml4n^kx{WReFypdH&Kx$n0TLkB?4D|swH`w$j{
zw>=pmgfsDx5GqM5P$wj2p&{4nVHA>CdkqC2Ba$5;W)zF7?R#lp68tB686_2X+Og&=
zw`89jxnwZbzNcTUGmL#<_t(u!0zMIV@LJHd1K*(hmJ$f5I6ItOU_Nbl8e#hHt-n$f
z`iIL&mRY<Zo_$mM*9zaD<S?qOtQTQbt6)b}%2ksa(7kB3)CW@I{yz<ZMD`#5v-d-|
z#<9o6Kw4(WYz!LJ0PkNEPjsl0B~;l^ocz^JTQ0T#HRkb18XRqfh;!_l{oVibqpSg~
z0K+>XPUBL8$wTZgQ4D7BuzCN6b2_x&?TUPr3<<gH^3Q?Ep@Z+;f2?TUZpgdfUiFMG
zK-veb6^`EO_m?1N23aP#ou~0&l<QE+{~W*yIDj%HbPq@m-fQ<<q8W#A!BT&Pj?DT&
z0F^&@j5@3((~N@a+JLbDiy^R@L)Z(457@o1CQV%c?f>y{|5xd5_ag)C$!8%00_bp&
z4#pUJ!K6$ijRU;N3tYNUcs(RA;x5aoAF^K_)L_>20{Hp1>qXC;`p->4R}oq>z66)O
zIN}b26~5O08V7e(Gkr&$UfO3Js#`@yb!$ila1)sn1ruHV{dhC*cs_oCVp9A2xA&fh
zkjMM*?59~I%JSPf(&#&v9S{2AGH$coAOtYtG`z$QaRQV%5UvpSkTWDbWVFI>jsFE+
z`Ue_r%Upaw{+}y&DDX20j1q8y>>aZIiktPo?Sg1lxhDkL3#wm!A0%Q=j)I2kxE<dj
zkZUj=w;9aUoPq=$dT1}tUTyz-o&WHFo$u1SW434Kca1&rpSS9Pi2M=UTw<D<P&#{=
z?Z+r@4<3R+u#LcaTL8TKfrTk3fbBn*4=b99*~IABvA<#c-V#6&gVwYg<oV%?<GC=c
z&8FbUogfr~)sX&vn9S=h80tq3dFh>krgQ&!W(n8z5Vaqe47#+ZT45%o(sC3?bkeLM
zwM$i_*9}9Uw{8D17@!WmJSjgJwf7Sx)<EbF{gtpf0t4!~=umI=>a1vcT-wVYgYnqZ
zlXDFS!I+ZP8N*}>u!_RhU)=>=H?GgvcctOyP8_T`M7Xy9j&KjY4zFAvZAMZozTJBR
zsvORc=;jwc2U;0K+1v)$bHq>pdpj*>{_Efxq74Drc@-BE@sF+Fec7x#0k2uDd1SLD
z3Fl0k^!E)CdKp{;<sZBDQ-3Yg!3X`l)E_SZDnTIpk9q#P6bA>9aUG6kkbi`@+`&%{
zw(Or@l>#p3a=`1q*7xtn|Fs@?2o-g?<eU}4uKphzwExEd0?54SyAfjcKcD;`{}=cX
z><Zzw$n?<z-}~Rcy37dY@i1+{|Nb4wmy;t+&@UQaC+`<K{<p`WhvM)pv>Da^*OTDj
z1@S`{?D}AW&|a<Wzt;q8&TX(>WY&KV=zo6^><|>8cW7Nh|NHL_*69{<tZ^bR|FIJP
z@!o_8h4G__S?B+}<Nx|CGT1~ZY}wyZ=iiO_yG}1oGJ=Ktzr%8{?t2UKe~0CNy{`Xv
zSpNG0{hwv|-_`_v_J4=ve_NRUvn>CATb9p1N{=1H2~99_*r{$w)L+6LY6T$Pm<`-C
z(AJfS6rfSe?m`Qw2Uj3bTRGpy8JTc>8A%!YNu<xA-pd_QNXYF!-kBouKKjSvXxjk|
zM9T<5o_v!F4|d6NjNW*U82l}|8q7u)paOaQYE<@~M}M0iu(QKsxq8&MVjG6F`MUC)
zL6T17$Yg~ax_7_#lc4LKXv~e{Zz1?LtZG&WVsElyzBL#t1iHXS@oQEslwvzcmcJ{_
z6m;$J+wj!jyPNnpa(0P=QRqORA+tFz*h|FWuiIz+7?>J@j<L)`)zXp9b7iy)kt;Qw
z$*jn?q74X>ogoRT6w_4S%R>Wi5yB)=GhpZem(NN{lzPX(2Ovb4m2k5v3C8(C)(qX5
zlG}Br;g15m^yfL1p&7r>Qy4<DK)2(|ztdV6jY2__5=NqoiDy=HN)*7H-3FB8w&2EE
zuQLphMc(w*2?%mJB0z})38a=YoA6H>VmX3b&0#aV;oG)u#hy8_zbfWh*8na7@Z3eW
zS7-wvnS=gt;)f&Q`^>&WxW;$@iO5H|?|)i)z&v{rptYCfpy56048{KE+#;_5ZjG8d
zPDCh(=2qT`DE%89{by9%6p6c*a9aofgz`4NpHrGHYWvYzlxH9S?bWset%X)QJ?Dk~
z0ZH#c&=y6AsoKZu%ke>bTFSDWrb3VOg4ks6iFOtsneM);5|BoHCShz6lmZgxtzLaU
z;3uj;Os#aU?bU_7;5{m^$tuma3djZ(8nE7%k`Bl_cxWtOaM>zxg6x7Id*8QD42MN!
zLG-Xbfxyq+3%@%7v)_V;hk@d(1r3S2+~(b9z-5dAbyjFodVE4tJQ(uI0}VIiJOn``
z%tH96WB)0n0uk9Tgca%!TO;IfP*ftP355Ddf-At(ZHUVF-}gf^UVzY(0<X<-FJ@Ot
zZ33})XkX#C@@G$R(*htky4q<ict7lVN0AFDBeX=dk8|YPb6D?(r|3_}V{|@(20P>z
zqgZex!+V|%8eau<0G6eMv4dZ8!C%a1MSw=YXJYF7*A5sJHtZ0sFY$@9-$(vAxj-PI
z0}nvobQTaD4!(t4DiBy4UpiBu-V5+3G7A2Mq+%!Gei#iID&Y&g_}jPlDc^m#kU0x2
z`~j(KgxB}amJ%kQ0d2clX>F?BN`DWL%|6HuK5@WfJV>x7-u=ixa}ofS%wcq*WeIIi
z6f=SUhX{iGbAiioEmF)7-VO>*xVNH{4-W0RLO{;ZhG&!qHX`w40!;3Ks>k*B&@&fq
z)3^U!foNhV4`$u^{%N`&Mx{88{KFEgfh6j#B<Jb9#D&ol7ePCRhk<XX16^DHIu@K!
z+<JU<eaL<RNJaft=u6h#nMF4t>BJvrDzPJbr}+79<01gqiTV|gzCO9ME29r@jmy&w
zBbHq~V3#+bF#7Lf(EuOuY2n@b{cDgKkk9xB@U+ftI9z=IO?_TULhYxy=rtl81d&;t
zx7?w`h*BQB3n*j5XQ;Ox**{v*0VyrHpdb&}kb3WfVbcAJZeAT=r1Kku?Z+Ga{XH0t
zZ|-yS1Fs>h_i6Vpg?rb4e;FhReSz#YE&T4EQxSCx7rMaoH1THC$j*o%tr>1&cAdxf
z4+Wr^glB(GFYf(0P>k^CY0-B3Y>?#!rl@et-zQzh1+5zE@#oLPFL~Q%BqDtUdoTXB
z7LvEZ)8+Vke~#vLUYXUTk`?gAaNPzh>z}I#&#^}H3dWBC9wSZH`Pg+o@eE{v{ZtYZ
ztJ??KkKuy%0%3xJVGTf`)c)Futb1tq3(%gQPhJEJ1`b6k8+)hC0ahRgnJ9~PlJDM<
zlQRUujJyQWfnuXg;I|6>^CFk<K(hkaxK^&^VCS_MU#R`-cr(Fb@ZGmZWMaYke<p(B
z{0ywA$fh<m^FM~khV|NvpcYdmKvzsf>S;0iMYBK_#Ny!!*6kf(##CGepo0Nqo6MWP
zX$oAYPtV+|5E%^Qvga{7Lo*6YsT)4SFCRyWe_&4UBFN<}<#M|H<z%^Lb<TFUJX<*^
zIcu!q^!}nR$Djf1YoGK1j1L;^1pdC>p};~|&9}8@ukJsDCoV|Ug0(8?*#4KL0O>}M
zAbp!Cl;iq!^Vc^989yH{3Fr3*Zh;X12BIK~ioHLxT)YM>5z9K{0DPz^`-jwG!=V8K
zq=BcAc%oX7`|&?SULc}Dw%%99wD+#>sKsFX9x`IXxXx`Z8ZPBOm$X;Ai9SvY_%gt@
z&CWS&b-=^>9}Y+s4`^%8JbzkN%prCjX>r(nL8YsK&~OUl?S{m+0EN}m5qFRwN*3Gu
z-v9cHMRr_Z;Fds1L)N2NVIPBG%*I1*Y?+y&>s!zp7yj!K0?^cOrp%8=4IEt0CK$cA
z3%ge!Q2Rd*;fk{vju^V>&XlBS#$@UJd;3Zi{MU1q;r*_ZdsH+~9|Zt4#W(rZr+aWI
zFE4Hn9keuNJwWJP$dvd8@foe5K?m;9ToXQo8b_aQ?@Y1~4e8&H{ZRvkQF-KC-9G#n
zs6$=@o7`clo**y@1$%@yxxbko^#LHWlF<EYEdeo2f?m(zw3Nc8e=2QrJl!z;=ZyIF
zgwr2@wPB+1xfBtxoBPZBKTahDhk(!eGOV|_B!_O{6}XlsYJ@}&wvLlS1IF?I`!VTo
z^&9?w9Sf+gz}YOMY1?;#_Bi-tP${uk8-eRMAp5UVjHB2@pCbk^Gl1iad%YrXZ~n(p
z1e44jZvqL*c1I8ku>8kC;4pVXI0Bt85-~s`$<8f*n0`RRN`(!*^~6m6@8#OrF44+v
z1+$%&Mfdl1DRBluLCs^f`^A|bFfQnyP=)d1C0NY!=imP=)WPLj2RZ+aPJ~eXkA(m&
zl|@4Z2riRhrR{Lo_KzA;g@hHLJV|n}0)c?+0<1eRYzH+W&}&FoWWMHqa1AAJLFYt#
zCfT+9yOz$dM)$sJ1S&`{`SWLG)O%a3Oi@}v1H_X;7>fGuBhgP;`i2K4IBx?JO{wGO
z$K3sgXgUf1+~>wAF~Yr7cu8yzv<jFNGR;fOd7AidK%|TX?uxj}pGJmF=t|xO1}qcr
zy93Gtno$mH#3>4BPu_jnVK(gN!vO!fdYEJ>JuE1$<r+N}LU#}fh7l%1NP{>O0=r0S
z;eW-M=io%#Js!3XV81;Q2+bLf>+^|afd%$YK!OakLAiVfR@6k5!__R-|HsvpfJ6Cq
z@0w<6&@z_Dma%0iDWO%yk}aWRElWkIq)<W&Bcg>QN~!Q`qg0X>C8P~4R8q7cOQj7(
zl=$xFof-69*Z=yie?QE;@AI7XKKD82Fsf+0Loy|x-`i+@Ewe;_=&32LU?o=*rs1G_
zW6ev$96?enyj+BD06F!Ry2uay0V1B72OXofln^Cj&t@dvEBR6Yx*yh550Xl*It>$a
z0K^|7I+ae%h9$QAFj-mTGW8(8<x83d+79W#uC|Af9&lh}2ByJwl<&gO0kUl8xop8p
zV37g*PGuw-UrBW6RagWHD>id{mNN$_-hrw3XWuSm{^fHkMX@~E55va(6^j3tED~TY
zY`7nbEb!PM(Pr)rH()J8euUP3C%Y$me+WLd5rS!PTyG0svdy^AysaQwuUz)$SK%Oc
zA2YcEoh;%uGj6-AL{B$1lUe?33Xu{-156wRYfZ5nPqa`SFkzt-+#+WRw(Nz1-z~m#
z*+GKKgPp*n3)Gw5$Xz$M|8&3>p=s%#Tjfgy6jBicv3}dur?BOh{CX#s3A<|>BNwyT
z_gG9L+J3u}BL7*Kv*HuXELWIm)2>zoJoqzoLZV9O_YU{N{)syAjVdRN!tG8eGoLx^
z)V!rEZ!Dux_-oqhKTm#tRZ%<h<yNO=ZTFp!s+zUue}xo&Z)=zt(s|jPMt9YY|Kd%y
z7u15ETUGnnibiL$*qm?<%g(rU1t)ySP#Qf)wKdwXryji_duGOtH<)#wCitZkdfd$B
zWQ<O9@N>a<;r6w{>q6u92$2sM!hI&Ju9z6~X~9wQ89JI%yvq`ktFi3y7k0=5tCL^B
ze`n@evMa~H+)1uKJZsaMy$s&h58+lZ)fIEiO7*CJvsJlhx_W%vVD{*s_q){YJM;gT
z<rB?mcQZpv#a--Ss)>uf9z}g9OZ|MfnDOdE@9vUkVeg@#j7Ij;O+38m+nW%uPh1`?
zvf;nzc6uiiC;MW+qnD@4_8SWx%ZwwGu4i>Yk@~lWkV@_HjTJ^C)?KY$sWHlT9Pa?q
z(rCwKZL-o=!bHy>P-E?nOBd|nJwV1SJiyll2K#w0X<G-HcRogXyuYQtC{BWZfN9FK
z(t`UKRPr5>%=eP#Z|3vPB+&|AzWCv~9P)_qtGU^C>XE-`YJl{7!(4vT;?vKQ<jC3K
zhvH<nFxbv>r&fEtTro#OaI9I2I6}*kCfSHlS23I!DP}B(&9wDcG8q%ye}CrmgtxJz
zZ~WuDkAuG69A7_Y?Yh(auOESc4?M&;jS00ba<;;heBiadijULrvOw#|#9upav4CV$
z6@0kPe|_~7yneOgo2}IOpP&ICi8~y1ZPXVehi!I(+xwVSH0yKLwBXjdu6VESgP(p<
zT4Z+4u4&X2vDa}ynlDEb5f}Q;<!9{T<UDAD<>`B#+PnC`gHR^#Q0#n%szxrqc(Ga(
zFVC4vC&xxQt`gRQ*kQJJ-2Q#}-j=+RZZ!8qFX~s!IQaMW*=dm#w77jdRj{I!UK{rn
z-7IKZv0d$XhZ66hv{Ry}0S5-yH(?tthVAC+UY@>?FBivXR?M%FfwI)|WF!@Vml}vJ
z@Ox)eJ#<(DCnzs94vFSgi>^_Pr&IiP_uosACL@vMSo&Ut`g;po_Lo)T3Wr_qtXYdW
zA5TBUu2bQEk2MX=5<T~)7l^Jp`enynwSMh6iQ`{nVoklI47=jZ8#!}%eP3q`FL|HN
zdvg}sZVA_`5Id;eb=@AJP}$ji+>G}V32R)xr^uRF)b&r}VsJ+zTC`&<)oSeqUH=ua
zaoENRhcwZ4Nsd=3(1o=fUgbP}p$luIMcS2k>uHb&Er~4xrDyqtCBB#LLDuER^grei
zk)PPql)NHt*S?O)CoB~C((8!7&bE?NqYMd0h6Y+>H!!#ci~4&+;RX(hV?a;%t`qv@
z(gU=KXNK;5f0&Y(aL!bEB5U`hxRplamGUH#v1x}4@wPOO7z0l3;fJd&j_~xfB&lfV
z2)PPOxK3`z?3wW^hVo>HlSXt<_^zVh3c4jvYy*Gzq1H9)^}fnPvhV@kV<{2%a|%SA
z0AMWM_ol4pU5)1`h``z*?NfOoaG(hE%PsihyU}WW3Ht^vDwL_RMWGh``i<XJI>tSj
z#<!h4u$_&Yzk1n_Z|RmKt)Z_((siPb8*CXpe$I^AANvh>&l#`4HQF)!D{S-BTCY`J
z0|tOTRF!TxoH~xQsWikqvBSnl>v$d8RJE&L&N#Oi3G(sF!cOm-`#a;Ea_C(CXPQHQ
z=3F>(lv3Vbw6NdheS4mdo3_$=z&uOPk*U4|&!mjAM@suRwucMvkcr@*=`l#(w}aNj
z<lF7&Tu=latoAMFZeFI{G+_8ova=D}D}*%e_fKooR%<)SlRg$yuWl9Z|39gfjkzn0
zQh$rr(r0b%EG$`lF`|uUez;x=jc7$CLmPGYGRIsj)HPy{CM%lv(%)!gw?%Dl=d~G+
z5bmy+#1}5+$q5?(A^rOFTy6B`dt;sM_B)ddxSTFa-mk3|QSp#yeY}2qAi5T9W)ExN
z-#ERXHR|-~(~VBjFnCwDb_Y!1N!$?bV3li`+=}kXTFi9yPi_>Uzqb@f6<eMQRPY`b
zPZs;y@21xHersyP5xTvP_iGjop1*&7!W@Onk=20x(ARl#R{m{)*%jiV&wl&0w?TW<
z5^g2*ADjAR9Qt9W+Bf9V(uUHom2}sM@qv$~BJGJAxbfe#(&HtcVG+R@u{0c^O!<eQ
zi1qA@OTIj_*h_9uSiKcAkh!41`!C1$$6XTKOa+Sz%F9)Rib?56PG?~5*Upp6?06E-
zPJzT%FO_nktnO6Ch3oHk&$vW-SHg#l9^`XO==!fbcPPirP#XQ!&8*+y4w6@RdyqGx
zVp;D!m;EY!a_sLI|DP<r(H$jjQSk617lRXlGsqs?&E+f1UZJi-d-iI@Y$V5&cAW;N
z_5<u)`43W7%&C(6Kc>s5ZzxfW_RW22HkIpDJD@FP#9TeuG<_|36`ycUfO6Btu)m*Q
z`(9o!EMW@I46`|&ssv2kD>qZ;_;Cye124DVnErG6Yy&RbO)Gr*zJ_XxnTwcwk)6a;
z`J;>$%okuQD~q_4a(U|Fa~(H#=MaZFMmu4$TpeQ(uj{<P^c^G@SrX6nE%`^RMk7RL
zv^mHfSP*1;|9~Klh{VyWF5PH90al4SoQWORGvh)l7n+|3^J>7|mPBxJ<_91pLXn`{
z*fD^w5Hb_t?3}|+h~wWJ4f=7c=K$w?hRr6_7Y2?_M`^<w{|8YEL^VsX=iL(x{nD1!
zGCbku41>RtiCOPM9#ef!5ysewo~Z#jXvgST>H{J!E&2*#KW6rq>}2blXr1E{{A=+R
zw=-6My*ovlm>=joS`9xvKl5PW9rvgO3l<D|{y0vKQ+vUmZ(BgO$~ll?%|F+j&|BTe
zw`t^I?3B1<ARxS2jQD8P$Z;aP-+_dSbv@Z!aQXId*~N=o(XMS4X?;G<i~1?Y1l~fX
z*pCuq_5Gao_20pQ+jjHo)rn6V7bNkGS`r6|fi0>8$pL|Zf$A!EP54S}XAE;GHu;7I
zb%3EX43%g@qWwO&^w6-of-4H=1kn=;FGNa7zXg-4x^%$={?Yx$)qP5lrQ$>@l_^ju
zz@&~7<O-^$1B?(`@NO|*!sUd&0P=%OZ?m3LZ7VH^BDn2%`@~h}A19LMlwp%Fdq&IV
z!ar}j0_uYo2|}%sBSrAx<LqycS>Koy({G2%CP4JBj-Bp7?#{xtj>!G6F>_cCWDP<$
z4%qug!g6>mTu0K+=JHfBp~G9>VHe*hc1YS8XI#tvKpj8`BO$gAU3qPa(Q=F2MW(6o
z9-&>Cu*G1qb(>Rqk|y+DzN#oLztlKo!ocM>H#g5{0i6|ai`YP{FiAvZy2I$%nq*q0
z%M(XG{`vjI-L<UxTuT0Fw}wmf12N_v`n%;48g%~ebTURmAJH{XnD7BwoA7$Jw_%7h
z-}LE8;NcZdn^W(??Gv-;J#`BWKU+J_dHlUtZp!<J+wO;F2xb9*{U?G3DXlLttU&8d
z#{n%IiL3N>_6}4bTA0h2i}CntFXT;#JIhe)w*mP~l`w7yY!~=z@AADS+wwF!Z72R#
z@9s}^>YuHfxfso@?VOz!kX_}vmh8Uq5y)}3*sE>7-i{b0BN1p5<UH`@t@wZ_kLBXj
zmkUh5t7XyaVC)8|&>ju~Ry5g6T6%qo_Th&Ic*325Q4GfCRScr3fA+Z5-IMa@xuej_
znwb!`<-vdqcQOz@Kh675+xNB(9^x?i7&-$F@?s>!lhA{j&`=q2ldLXdA0y04na5au
zXVml_sNOJsYQ~N?=K^pfJ;hUhc%^#xhb+6bAc<c(^k!3%xP&b+=159PLx$SX4aToT
z6bH;mx0pqzJ(_8bfma$@(0)jZ`aX7r5x#yj?Zz?QJM_`N!BU{JZ02#9e<b;hkQ`zT
zc*P4RZ^6<49&Gj#79*AupEFAdwougI7cXsE!}Z$HHyZZp$CKGk9`(t*!(t=m0#J(0
zF_)%#s74B+)_(<Au@$|us@J5+Qr73=UX0_3(GYH^uu82riCW0w!w$K}m-*W5E2OUI
zwbsDPggi!`WHKs$gl%a5@!6?|$Z%3Wf3i5v|9Td^^G#qm#vPjN_9`EAe404E$|OB)
za(q7FB;<ad5shZXLj^xi^{YFx(TQChGX)qV66&VSlXdv&S)+m%pD<a9sPGW(3*mfd
z85qg7J+I6LlHx`;(z$VLcHdtG7TMp3rgr`B(-I8%7hr`8a5T=NVhCoI5j&Ti<$USC
zfTvQ0UH9&cG2A+8dRPAuigVEtCBLC)=v0^QI=lG^v0X4ep;-Hp5IKI16&1BA6X{1y
zbI(8w^NxX_pX}QqadUpTMKAAgIcs4GTu(I0P&$-7oLk%1`-^OV%vbN89@H;pIDxv1
z6LGHK!fSS8AkL)jrd>TYR?WN*DUyLJtFmb1dDiTe+(xr91ZrRLe^;BgET!@Z{AS5M
z4iaAs3mE^y!8k_n`2XacGBk8WT}Knrdka^VQWQUiQcozt)S8BkBwxYh^m|9{e~9=c
zHhDj$66j~*7{2!nZTO3%FJ@+biWIl87)Uzk&&0Z#rx(sd_sI;`9Fb}Lrj$Ya=&Oh$
zqse{fay)_m`&Xh(^*RQ$4VVD+1Wu0VGIX&bz|?s5<q=;kom~JJc#E&%c|=({84roP
zdE&b|t$`p-!qyXQ>PW%PKKhIgMq1_G*6d@(KYu_DikXT$O<E_(+#)pDqaZ)vH7@Db
z+wa@2J>Iur7*A3-Y&9B52BDy`2NrZLAFyH;uwwIE7X*@!9)S!&Ejj=fkqtI;1mOaG
zw_5y{ydHeyFGNBT>*H3M@($M&(j(Bub&}WndjF(u`IZ;Fr2T7Af0)d+Qz6@Qf1y8F
z=LE<kHYxh&EiO-!qGaaZq|avuCnqEflxBR$Dp|iE4Z*`l)6%z30o6OXi=?%JykfT+
z^e<mePnmp_g{3EI%~+!sx8oR!A_HNR@TB;yhn|9JPc|u!4bC~uOS-Kl<B!`FhCC<g
z$n6wCkE$Nx2>+&Da5(mjoBH41@&Gl#_Z!k`e^*^AE+|9F=drS1uy-rsGxl!M6(J9R
zAU0=RU(v8JZ+tI5mcgh{o}C#=%M{tzQmK*0sKngJxo8=xygY@?`{4MUZ~`)A_bChT
zUML^UV@Nc%9Cd2H_;DbRP(NFQY^tG><kO^$LSUD<O0ZxOtMwjJ^7KsH$Is9Gs;`T8
z{-&31zkssa3=&`&Ez&yGI_pGo)Mn{<!n`<2JsDqrTe6ah;E@6}F4+rMunGxC;&A<u
zf_Imtl`cqHk6aP~+H~2FQhv&raSDH|rDs5CkCjp@hRqT8>z`9vnA*^veWkskVMXI7
z_*U<@W!jMlGNrZ@tTd;j=bsedV>~B^p}4UvtU<Q+V@1Oo9*_dHPZt1I$-(VXC;LyO
z(G8;ieWH38b@xliJPsJ}y(iWf32Xx=&6h@UDf?bvRX-0E_NewB+;RkP5-a2%Q?CNe
z2Udaj+gX~d9Yp1B5Zpk|QOO3+1AA1pS9;ku31mHrgiwzb9iiZ_lPd8&uV?gBzrk?I
z!3Vd3hULuT_0}x5*7S`gX?4F#5Jtvr7d}syK|Q=}9Yzk3M)`)|!6DR4MuKW&;oq<8
z30gQ){t*v8k#GP=k}+|uJFx(EQd~rCU2kt7{vw!8=EEuDt@viIWLFVEez9dJ**C{y
zz+`okyz_faDTNUWf>%`fwsf?N9SHpbgjN3NtiN8r6Z1mTj*ecrfcn&bFs6(>ML>Xm
zEA+J~v|;?|prqB#m5XH=o6y3d^A|(B$o|+<r>6Lmmu1_{fysD(K2w^?S#Qx3V!Mf5
zx>nF}J~Xy5kNgVJ#~vnusUXC0-4nvRcMH%PjtzW8z>N*hsI&_KAJpmav{m;Pn%vnA
zq_ICaWZS`w+}vy2(2SVLW#y<qp3j{phHHDa=7<Q9?aGC9$QQ#96>1KPe197%_IXCK
zIpvM<;{eqSb=(nTU()F;Q9&@td55oWFNkQTY(^qIK~^K$Y23QJrh@|rbvUP)f!%*|
zFz+_G84w*=*CjYt&UW(mH-o4UfcumF4sj(}`5CuLxnH#C%*AwEJn0NUN+B;zof;=?
z2+TCY+0>poof{&N>z_s4Vrky_`K3A#9I#Ca&xQz&ZmG~n8C@TYMvBqca%ka6UK-C`
zDgt;bv|GiNTqFgq>jDJEDrU7)+)-2rgujQ6MXU;D1u#$Z?ozkYyw5B;i^q2S`hF5+
z5@LAhRCH-!?0}w!7ol(S=BCe%{?7TqZ7_7n{7f@M3gY%hxGa@~QZ19?eT`mW7&VA&
zwKvAB9>ROMaL#B(6&52Z@0+7D;dqa#X1{;!#)Xx<3@trK{`#4uGxC!pCwFP{QGe(`
z-@b$yoQUR2JG-?{d>n$#Iokm=fqqlnuWRk<bqYr9*VhD4BJ-1u9rmY*R{7`K8OQYF
zEbv2huWNgQFpu!PWs0Ub-$IsT!pPZ3+0>AOh)>|Oe}A`rYlAHE-Az`*WOtIMB0n(U
z0-yp-<nG;+l$1kQao>O8-w+@LT=SxWU2@@ADUU>L>c1DBWy9{Ce1c$Sm^tF{Ht7jp
z2&>82^~#jSlkbt4zzQ;Q&hfJJ?m~CgNcma0%QD9e3hoF*g~`{kr4zQy>c5wGDQH`M
zR^lAui_!!T6>j$*hzilCkjj?n7jJde*kbnXCZ}1Hl+c-8OJQUiL{tpOYx*P?L7BRt
zllSWnXmt<xcaEnw7j>2k1f=>QA~pFaYRF|TXYhrp(HNg0^-zTZR1hQ+q{5u+Uc#`4
zqW)XL#~XR>7^Oc0osJERciw_blD!#npSMIuh#U!UX$&<>nY%4iZZ(?tkInw0@c5$u
z!6xA^P7$N1NHe0xol$f735JeForj4|-lPO2Vpz;XNYgEv%!3{&C9?WM2`0<=L{WP5
z_#5bQd+<<K$y6W!hadn2Rr9a%1pqJ==FJ7Bth;JCP?n$zD`F}4A=EMzZBFIog<yDT
zfAH5h53jF1QI*O%k?bQXCkun*$_EAcET3DP_Cpy?0ICpP`To6ifuxkUa~<SH@%y9V
zhZ8yTvhYL*3F!BN3N}Aq!^)#o>N7L>YDN!G%Eldh!*u__;eEU^?q5#pAcbfU4_|wY
z)D~d(QlAy=6ri{g^VxZs?WnALC{Ojs>OMS*5^MDWyn5TqS0gFeSwO=I+eZKsB-UH@
zzLgbVA^1&K<qg<|8V#7rX)+I{I-mEDS41QbOd>1ivLpo<bFG;wOIZ5VUI!2=%n)B~
z$&Vz6{>19ne9XQxde1IfjG_@#JK0j?@|`DG@{<#^{iXtzm-9tR^bbtMu5jcwo-QM0
zJ<Ln0tWvwq&)j^{@D9V;Mp5qTu=sU!xe**V*|A;wU4JOSVpoK6E26RqP<6qB0;;N>
z=Os|e?FEonQREzZ1kr$!-JEtUGo&T2(Pnul=}}{UZgPeoRd?c;Q-2}3-DzyxA2~}C
z74D*0?jxV}OuP#fQL+Pg#_`xhWkG_%t`OxmUUaCRXGRA6&u-V{<-$X0tC66b%v?8#
z`tIB!Y!JYH8fVL?Tiw1PlzqJMMuIX=PMD3Ro`>9msxi}TMNVl!dH<_<+2iq3J}XV6
z6gZsI{y5)U&fj$F>h;4GO%;HwY0k7#vq(@aT96}8tm=8hvuhHLh%~b2uU$?=zhVNn
z5F>$p)rN)T-y1BtAFmgoL9(K<a&ZVrpD_TZCC>1<INyy+;QAMF^yx&P8qut3R|x%5
z2>qBo)53K(`COusN11SwAL30$?<$=T)-mAghr{dWTvuO2>1{3pqwn5PRkf+hO~L0I
z@b#Ru23qN6>9=sEWRcQ?)K5aZk4CWtNR`%?Zt~CgVO-P%e-NZ<SLhyrU&Mp0QDt==
zaRrS5RW)?KR%Y4{s;ae+6vJ?dE#*AkFjb~6A}~nd?-v)FCv4iq$Frb(70^iN-{cAB
zB3XP8jQtSO>*6k-M@2d)se!oy9cvhzo05uPqqZwXi}m2?onC*~!h>?yc7agbBL_m+
z)X7<Ug1&*=HbdnW0jB^9&P)m4D9A5nigwk!EYQb*isY;XNdmF*!u2U$p07NWT)l-Q
zyZhSb%>_S>LDAF)d3^O$;_+J-!~p_5SYEW3`b^GoTw8mtYk(@bo+Bdp7{F0F`gdb(
zuAaifVS=xv3xFxD_wWmr1-4>T7L1c7u+YfKyA#R%0-4X=-wp^+vtb>=$!Pf?=g<|h
zyC~&CG&olPO!>d7KXgfe9dk!Q^1U`JJV~XQ8bS!qa4ij<%cLT+r3RD>F`NmpT)->Z
zVA5W$kSOOH&NVQccH(AZsHAqOFu)rFFoOfE-!9qPj|;Hv47n%bW}U^brlb$p=#ih7
zm$!XIhz>b%2>f7dD_~Om8;cM;`zu_2;2i19{wPtY0-hA6_<`50hw6ODk@Jt>>eA-|
z{KXuJS#c)^P|2aR4oQpNmJYyDy>0u#vVO%J=lk4TTd3VOzi*V(zY$rw>4gJ|2N$hR
zfPh1Qxo>B@s0ZX}&Q@&nP?)j7(^|h;Rz_fyxs#CJcr6>-NLhzbEQ_m;6aq>VJ$^qs
zG&nEy9&6!ub`JPnz@?HLit~AQf0}6BtvQrw;m#5ws1y>v&0B-pwO{qy72KR5VORxy
z>-iOy*oByYRQ!L(Hc_B3JoQa{O3i`~X5>*zE^&Ckq&Ll@=e^Xv8Si7{Y|UE8DU+9w
z6DIA~7?!Ya*PwV6Q#KU6*349)q6iHk3_8?{+lX=qNX>&vc=TtYc!#Yh^xs1M^l9%}
z3EtVHbwKydUkgwrmnmUd2gk4t<HX*3<=+$B5+rc7upC06&smpkyzip?r8j;0nVB~(
z@gHk28^%su07{`-+CAgb;r`@UjfT#4In?>B^}FZ0_5%a|6nJU~ER~F&Z#f4MV}3l2
z=Xxe((W}#mK_Ui-c|LvcH03`!1KK0GVJjv23~I~($*arm9yxLZ8n>{>@`xBuL<y!t
znzAv4b$O1ygMi}_FNzP?J1J6^Qa|GJPmPe@XjfaX&0gk}pmsWxc23NKV$lSG{D1C7
z*-ZYoPsPE-`cKy+U!iMS2XlE7P;5iRiu-$a4T_c(X~>hmKq<XHB181V{Pt0ncc^j@
zC;T5Sg@k<?A|kNelJc-0HWbl^*VGTI9R-))!$k5Yi8*@f2eU!2;w>^h$|0TtuSj~m
zXCsI-Z*OiM#08-<v)oC*hc(7c$e&&QUYW~}XBHvDwMcsHNvSc16fWxskxD3du+^8D
zNTwzd9#sIzD&T=+Q}#XO**30vn!t+NZ6G%f$_FtaB6F8LAMm&!Q5la_oRHFUMXNtw
z83kOxhk4V8d=5x7gh53}G=s_!LuoE7i{8*qEDf5|+tuJwVl|p~7U5zMsnoryM-`ED
z9uTzCY4BM;K+O&=ubIH7&#;i!!=Q+FI>eQ(F%@7P6<ARhdH=U8g~Trsv+z5)6(kGa
zf2dPXHAg0pC=xoi6@xy}4*t|W0mhrQkX(Gi^vUEIPO7ywr>uX6D91k^sr`fgdObT{
zl<Or_k3z~#IPz%;$&`?Va}Gn@%kqCWgFL~_Vd_qw4wN@b=QRPM083~*eDs)K_8>?;
zp-<rxrg{v(&=nj!|M2&+6vlz)FSzG}buM5|SqzG@nYr_bwI}s$sNvNnzn6Ufdj&yO
z#;r^>Doi|0rU|BF#PO`tje~Gh7Rqr1<-ov+Qq3<>tRT}8*PZP=i5XlCwNgWh%6!{a
z+tWOKzPV?yKv0h|v(T|sy4hwz*#3v5BTw?KV>c<~<&M=;AlhRgj3RRI{k_s{dnNl(
zQN-=`>~4|QXu9Yn7*t9~*>1}DC7Qy{tB9;Sg1=NxEP{u0otgR>1r}roSxes_js$Jt
zO1s%1e#MWdqbIE7cQB{qZgiw*Uh8$b+qHpT5zRKk0qTk6@zMr0p!_4ql=Ec=cId4<
z*pm`SC|8)yFSFjC{|B(q!w61sw5R3)K`AD?)db8XBQy;Aei%kOIU<x#Ynr%?k0#|h
zq#e-Qwxo0^<)lVQy;^<dSTEM3M(YkAC8LWmOh|rzeto^UDeE0#(AKe6yQZ%SKIGS4
z{xs?0W<i?FVT;f(4fTyvL3>hiDeARx@C3o*c3{Is`Qd45J#oJ<(Wjt&pKN+0nduSC
z<>&SvuZK^LjFSzeQ8vlC4#c?e)BWB+LbrcRkA4EWJb>TQ#R;`8OMsw{z$VC^#RB+{
z9lwB!y3IFe)YNNd7bKZnsY8l2(qj~m$a2G;JBCEB3*@BF^9q+~&CmipwGN6Ika3e^
zZ*L#47p~zga>fVaxA=BVYae0QbsN=MyKY_ZlRy;t+45py$RR-@PS$q<F1_>35ez;f
z3$n0?)PAqMTE3^%^&!AIOl{55>}wS}d3M@pzO@-G9K3(wh@o#O=FuqDD`*X=bV&{$
zUAuHPYF2$;YkRfa1A&=*M}Xd+St8K#=0d7wY@MyrnLHjf8xnjg^tMJH41^r|BO_T-
zaQ_TXfH^4$>1ysaEhmCn5?LfOpXuPT(=E{>^wTM9*>E_v*+PJ%#Jg}SR>-_tTz&fI
zz@M-yWQJ;OKff@5$_*Qr)WMkl?)_C52El6DB9p_fX)zN`>QIIITV1|A2>Y`WSKL_1
zQxqMoG!ez;50xo5<9w<J#hp$PHME~NXghk`6n4JTbl(E0-(s!>&-hOaCD9Ps3LFZs
z?#L9iK14~rwH2tKxU4;YrDpTRn|wd*fKV|j;T{Dql;m=5&SxgML`|3Z3icO#mNZ7D
z8yuW=>zN=`vSMdrf*b*GP^$r6;ub8v(0paqpc3T&ec*`sOn*O=-8$nHO^iK^U#+gT
z5!yeGimZsvo0A!BBMnBNbwT9fv8>V<$j`E$EON~3c+|Nq*2GZGK!Jt_`U8Y((VeA~
zGz6UhW4{A^K|A+#@M_Q`7;+F^8c?cw$USuRy08->!?vMp?RI0MOH&MwR!S(K*~bF@
zym>Q^TvwD>Fy`ElKE#e4##(1fseo!IO<IyUTe32Ldwl&Tt60j|G&&<0c(6xp6L~JM
zBS92s$11oRj<!GKs7EE=eUM>;+9^crX4BfOe8vAK|Mf}Xtox-!bRR&(i4V8dt*u+b
z|7*?zI`k@^goIzDUu=HadF^h?nb|x9<`j-Wy(uR}m*{VgAt^(?MH0Tuqp5Fmai2hP
zkfYS0tCY<DTO{*{AE$6`60yD4Jb`?TdSua!&lhA{Mr`}jsdd-ax*tB6#ziX+sy3Su
z2R?Rr_l%@1U<wkj_xYNXiTK*xN_Gb%JF<7xIykvqxicDT{_bH?3kg5L^%HjneTBe+
z7{l7Ay-Au1?zsVk<gwz|V+sdw6uXWbXw0mSX-PJ-1zA5k{wxO!Rf3Wuj@Y!d^2pYw
zCmU7=Q+A4_o&e}!zvJ_z<a^St(9pBm2|mcHnsaWv@(2FAoh{c5JH>IDfHOMe?g~oN
zo%rrt^T!56h$@`9qqi6Oq6qlm*k<<?XFNCdU)aea?8e&ube@!IIVEJc!+#?GPeNY+
z3IZ5#32l%Tq+Fb|O*Hf`lu55T`t-(Hr)>@+3Vm`~qG$1l&}D=X@U3I|7JNyfwI>b~
z$La@DGGBlFbKbcklQ*d&1!gEj`~Gv2C@aR9iHj<`^n|DG9$m%&Qpu4bDj!3kSLCoc
z`V3=Pqrku)sB|Vh{^>=@Fx?8#mR<3@>E8ItM)x#%trm1N-@xR}-@m@6{wYnN1ely*
z1q~5kAv%?>)~q=f^e9KZQ1$8xiEGp~au*?ANL-V*fFQ-HV|~`FIf7Z>NJ?Cvb*JbP
z!NaeyooJ|^)q@u<*E`M0Wj=)}GbMnLq$ht~L&c==5?wDYZr^lQ)LcbyrugHWNvs<z
zV{PM)79{}*0lW~<8+3j3JDMrVM7$<coR;iBf8Sq(^I9_8Zt9uH4<;kfNCwM3WqaGi
zS-ZM<jF;iKUfm&Y<H<wCA1T5)_I~K=NzEL!irs&=GQxm(f8v<ujZ*lykb3-Z&JvPU
zX*P$F-vY(=B_NsR=9RTeA3lz-q{vA4!0dz~T&+K5!e?!)5#qnNDx!nLG#<{cC*iZe
zft_!-8A1EI)FmcC`$(!ujtKES=5o(MI`VSH#<kn$Ep0Ga%Ky<~9PeW843b`Rm6*Q3
z%BBOu#dK8X$kJp%jXCWN4RQ>7{RS@V;bYHE>aoFv9Y`rEZP^>gNA~W3`?#^)B$%IU
zLo|-i2Hkpd+I{{rMIbCz7)HHsDo4bvrTzp0BjU}wQ<8XK1$lBuZNhY_Sa}0Qlk>e|
zsr=*rCoUYu*xvgi70VfNxkBUkRAdO14UKROfXlpeU3=bPZei5=!&&X2JRbDFEfMi!
zxCH;k?Bz%17jno+kTYZX+$=)K^Xq$G@Aq|u<?{TSLgx>wM(Be+^&4qLnuW`Qzzq8P
zu&`(Tu**7>B_t2Y5hi(*9J(fjhe`A88575~eI$|BG$!rlmz;;d{gC_Zo+2H{Gg#zH
z7uI28HfW31FI3O<9JG9)CkQ64!X4&1JnA6N5kHd)S?|Gs&M)XzY}PrzqMn$1?>746
zyv78?G1+s=I~S9uWOM2Yp~Jlojj7BtL;{1<ej(Pb#L3Q?puLLsJGlN%p~wYEB-B8~
zp<vs!oBaL*A5ssVn7)*%u1kx>;I3>RAbs%zZ&upMMw53iII~P<7O;Z)b1E?7=l4p6
z&9nC2u!FiWqz+JG?^9H=5G8RP@7{@#*2XrEPm<@O4!lae;-W_()OgufNI^r=Wd47z
z0Z>;lclu@uY*S`vdQ6oy+}?v$QJKThmw15|US39v5R1SgxtW=1DP?@(P;nP|E<0=J
z$MOs32$43em4>*1G|LGaE}*zw)S>ap1Ha)kLt_u|!faeaH0Myzn+8ehY+$Hyo~?y_
zzxv+2lG{r~bL7M0<+)hw^$GmQ5}Ubv=5&<k$6Sm{F1EgJ3V)negd$)Fh|!UKS=|Q=
z;)if0K)(Ja=laLwff;krxF;;~Ik>gQ3?#b;N$`Slw_aGz2r+hi1P682=uxs)pWDnn
zHRYfpc^x&p&aB&^lc1`TmR$8kwSw^9_7DlOTTVI!v4t-4?vT^GXKdHpfV0F*-lrjt
z-QU4QpCL?}RFEjJ>>*(UUOzyyZuYL3ZZlSkTbK}{+G0bCVKw9EEt6OUH=I9+&f%l&
zs0d?|hU1|Rq3Q_Ugh5)Z1K(u6KgWce>wOYh5AT>kn<vJsO?U;8#BN>XKt($n?)q(?
zmoy!{?!t4t`668kZTT0|dW7g)gr<aOuTGB9W2m-H87W+&3`7ce(8b0X&+r8VkiSo;
zq19B^0dgq!(sk#rR3WrIqz^0Qds>sizwrzVF6uNyB9VtIZ8`$F@81T)-W4J*|HKgA
zNmfu17Qx}Y7><uQA5^T(eFNLq=vS&@&2}Ie{kHCh9%OmS#1q^9ysjPnC-2k|E~w%@
zCpcZ&zCMbsXvu3TQ26`N9Ftua%BI!_@HB@#k%j@hGm!JI&|kJWlGHj0AE6|z$nfuT
z!+YmTj75xJp?QxsE%^*VPAjbNzH3wyA2Jgcr(u);Mu+<y?EwSlCIZRX{67U8DfM|o
zI-?$wn9DsAD5xCw<*G3So6L}B-`#_aGUG^xg|q~zJ%s&8sl9s_{TP|sS$ar$8W$rj
zJR_pEj`m3iMcdQ1sdN43gyo|81qe1$G7^Kwl_p5Kn^u9^&|j+yPy^h*@is&1ZZ7Fd
zuZUrha;Mh<_D_XKO=Hz_6%kYX-dxY}Rwr3;F5xSjUw%Q&ZWi~fStMlFLD~4zcBb_V
zLmGBfjrp!{uJIg0(eU#Gd!800kTm^3EVGTFq4l<58gy1MJ3u7jEnW7~==&;z#3)LF
z8Ou!Z&9OXs#Ehid@fuuwb4Yt*@**Fq$lQ?5H_@o#nnfO+d^C5ydKx1=a*tmby^Jt0
zb~@;r)8xlG=ZV%MQfT>T^42z-bDojQu5_*Z!P7WLy0LPx<C1k_FSi{#PQFB*3(kXj
zo=7oFOD@}M)g~|FaE>_)=dAxexYJa{P=wAkixfM5EwkB<BuURA;_qxa!1HrHNgN;3
z8K5jQoT7G=ivG7VoMT2$&|)IUnhP9KhX<(iW6vBH`dJxK_$U2Xrnk$MACGnw;l@`<
zMIOyc(&fcnTj;t<RH*n&jH$&WdePtmmr@zQyWf76%qJlqQ;YH%3wn*Z-HV6~2oGK{
z(qkhh^TKZzsf-#?MO`IAf^xBS-A=bs#&3~pd}jUDRERZD3#`PxA*G%lhJArsdB<J>
zglsYs6Qd1)Qpz4r;F&5?HozWTl{{tRF13e8#?*5m3MQUSl9A}0xzaGx56BC`K@<n`
zCg&){FXUz`EQU8(5r9e5q$<xf6N30@*q^@pl;}B9hI{?w?{Pz8nNcoQ;u^`}l93Tw
z8J}SBRB@e*_pqD?>1TsXuY;*bwviq5Yn2AlE>#g>7q&u~=VP*%k!F#&m{}{Q`M<wQ
zDh@s|z^A*=cp@G>+9Urt@8l#*Ah+W_`$(c{$DFOCA}iYSG-3R${md*Yj!5y@#YcHj
zVAR-*HKxHBp71rs?$tg0OGlZJhP%lQZmS&rodzUaCs$SB3Y%PIl@~XJyPJbOj86cB
zWTaEf5?xdF7rPv6*eI(AGV&=A8^;UfPEMK<-0B(2)JTzyD{7(7MOa2!YufC!n52b+
zvmZjYJNHPlF<TV}qG2Svv*-x@9mo}kknW$enb{<9lWjVR2ls$NRu;=dOQu$f*W*2$
z`;6kr%u>lncmnS^%RUk7ED!7al@ln!%@M)*wBGxW!IBXLvpbHcJzI}^){SK|`)h9H
zPJMn50mv?vIqL3oT`mzQR)O!e$NI1D+LA^y^}PrMmK}U%$~>vbs+smRCjRX8+srH#
zmMiPD^<gpFwf!6|yI^7z01BHb8`n#{rB74$l;W3;GNSPz28#_2aM~D90%NHNv&i=s
zZ%D7S2FsOnYa}GPv2?D^c*eu<Hc7!60H@u1Co-&TFOeV#y0J(xoR!>&XAx9QPJ&cc
zg4eDwqc0L8ro->XYPm-WS~yHLgL_so5=i`}%a8TvFI#HDk2a)5iEY5T#U~bi&)9au
zhidWJ=4q%e)v0lZaD#+#T7ZzIzvHolPO@2@%9dBzuNaDqK;Fl2vxMZ(6oe9M_W89{
zI*DUIJoPbBZz($>1r#2+v9s!)1JDkuw6mhiiX(;2-*`Na&fG{|lT<zCn)DY9kmUH7
zs=o+G@jXpH>&n7lghQD>Nri-4gQ%~v>iKhK;{3EV=8>e(2`Y)U6t`N;ClXp>mLban
zV7(R=#W7u$j-5Hy4WHlBd95(^vK;&p!Q?0d&uJrKb?x*@Q+PY1^Dqkz{yUP~>kz*{
zjee)nctoSR?}Xh}P!}Ph^Z+z4d})Rv0uOFQSRe2?$MNHPX9r_N&WqLTeMlQ&<xo}1
zVLfBVJ3(wh+-8P+aP;iyYV((hZ1v#WWH$z>s`8?dHOdgYgxc`Y-<>ia2O2~H6A>bQ
z8QUWKdrWy}Vdn{DqG?~q6h&cESBc=<{!?ZX(gw_qk5R_VOhiT`Rmm60dK0q&trJb#
z7PYGZ*|IzJJR-jO#r|<aX`6r#WQb=y5JUfogK{wvac}cTF|ESP(-Y(0FpNZuZ*-dR
zl)QTEjOOGOx=LP~L$LV`LcnmfwpZk&k%q%{MFC)`o)fQ#$(|+sL8#IW#|r%W!3aV(
zE<I-PK7<dr+UXNf2Z(5`Y+Sv4_@n9aMCPD@r>0$4Nc_D3!si&#Zv=gnDUg#_6Q_*-
zRdO+D6KJmJhySQZ+1IpqnvAG$Ue<&p4e}|dCHR9id$`O_2S3g@0L?k(X>gG9iMS8n
z2FDwzL&Ps!E<$8<fsy{?{&-%gF<=r-dUAei@eEuEv>?rqw(ND`brEBG`{!8E=5S7J
z3#sVpfG8W^#z+{?eZ%xIEqT1iQa2UP(w^s+!^3wz36U}y>D-bl-yDkiD(3+Msz>(B
zr$V&GX)za?Km_Vx?S%{PPOv~p7$?yxAN%vZz8Ke7WbqYt58B;xMU2y>-Zb+3?iRY^
zJJDsRdVdXx-vV-uC5*0>iVZBi60TlM>uRQ0#x;1;aY&uynotfXApA)<d~T2k90yp3
zi~YV=oqr9is_L709<L{yrJjvIfXI!WT^%-%76Q{*K?~wpBP&uePKeT90CwD7(0@G&
zxBE|d`<$H1TiV2HY{1*s-r5F|bRUms=SkfrCNL-)a}T|)TqF+fOX{>E;higAjWQiX
zeL+9wS$R=iC5dGVAE!UX(aGGR-RJq;5V4Aw_%kmO){cj<G7gZ6tcb|Qi+oV@p9uj@
zTS-FZ6F-3pXwn>FnL(zlJB-l5?x4jq{QY!aE_>tM=eKzB&`L`et;Yhrz0XI+G6@Nq
z)OBh!5$=%Ki<1_NPrA3newtDR_2A%FKZs+Fau>Vv!KDrm_=0Z{WS-NUa<gyL=MkzE
z+TJU%*5JLxfe8zC6N9QR<YVL`eh9+`)Xze~7CKesGj$hnl#P`;V%^wW>|DS$wblN{
zgSht~bv2P!&n;?FrfD(p_d_3;16k^CcZYM0Wg-sT*MRo0+P~j6laQM;k&$qdz47}y
zHxc8BZ+RIdXpwqK)<<XT7aB@i2d8V)6&=<!;wC%yMFn!*0dK9NwxL)^jkLZ_)#3%R
zj=}>1Hu2M-PwV+)8}=-w?o0%5mEx;P<o4dYKx7NEb)`cHr#7gmnH@-*_&p{~R55Um
zO1du**`q-<tE-<l1@X`8D&jEGO>$s5m(Y-&&DV}~iI$8!f3-%Fq8;05LT(pnT`Lu{
zaDbs#%s16}Q9sIzT~2NjyoLq()hgxU3A1_P6@L}7&8$t2JF(L&601k4;73aYA)B{C
zsu=s94}<9?Z5J}4i`<6nXrivujnD*#2N{+8uqxv<=jbk~!wPIN9)87e!F@^+2T8P4
zyMpH+6A2P_^s`?T93<pYEO*&%)mC^v>_#_uV%hJ8Z~<b@A?Ea0<~rI&|J6j-0d6`T
zE^uR6{aNApp3W4s1*CP-0M=Jw%}_(2iE=U0NvgXnYf;)7sYs08y3nq}M_Pu`)=D6f
zN}l2W9+j}F;6ozJM3=O)jCw*Z{u*J|oH_y=UN-Q7uBlLr0+_#+&!0a<5w#=3yq@!Y
zJe4NTCNW?arqRtoxi4tW11)-)GD~0dX@uSJCViwCuFAj)WK>_m=x!%!O#0N7jsNJq
zb)j*yMT)Vf!ToRsUM9QlG_NYsGC55&{oCE)>hXvauVY77LnK}rb$1>KsfhsGA8w@R
zaTha;m5qT3H^kRpNgyo<@e&y|jQR(Ly=@axd=v6Hg1VH?IS%;B#n;X@xTTBUjw2Ql
z?e8MSV`EHUUj&`Gm2RF*W|05~^1hvrAuQQkcH@SpCwU(dz=rw8+LW8g8S<jU>MkPZ
zwWapYammP&`;rGu1L2&dVOC7OS>)n-qYj#Ila{d1?0n!(h{a(ZgjI4B(smuLe7S2F
zl3TDJM-AnwvwGKxZ8jp^P*`XpZ77DLO?w10;_bdIP%ZvA@tNQ(oe@<u*rwcM%C3$e
zMLS$-vx@&}eqnYd3u3fmUczz{2BA8m$8XYU1bE1!IN|(yw?NX`HCnO0$r_T@JU++>
z0!h+S>NX!RaGS6aNX)-<tuWwnmeFAyVC2aJ%_;WNL~&pn?I5{3!u??Bu0S+gCpO(!
zM5+DHGlp8KJ1qHpolO#5-!>Pfyb)dY?4%`DBV;rLk6*wxB<9^0z)H8J2Vp#CT!eMo
zw$0!eI?Qn}DBWqjKa36NXj{htAlb%)?u8leXA)(k!>!0H4_$nlp39c-HChloC?aF|
zSW=gW1zD$Lx0<EN;V=7va=MH*xkOZOk{{gNC9a2)H=-2k)FF47Gkv$dZ?hFKt|Sz0
zf#@SD1JBGx+4$`H;Sci>+C9?VPv@nA{v=n$xKP%IA7B@}5>;How3^83oqx&mc0}3F
zJCm7YL)B^xYr^~=h&wCi!yMI*3G3Qn78&ZY#m|NcE#u@<4%|H@7}-IL{NzImo^29`
zG$S&4d(r^%B}XXb!EGmP*k^F&gq43Gx(HKvj=o{ks~L8ek}kN=k&XNT5hU+<5ywmy
zU52XM24<ooq@6b|#)k}U`Fwt<WjJRYc?AM$q@E(FkrhcTTG76d8_7J)SDllSHV|}v
zvuU#K>I8YwMeeja-llcy!~VSXKIA5O;3?0Pp%|~OsG@flTm}`Ka6wO#_Ji9b)<MFT
zXpR_7rSRt&M*5uj>v<gav`csena}HgenA2^@4!v=QBdRgJ;Nn&c)p-2;|X*t71q@Z
z8H5~1TM<5lVUh=MCvF1#T7N$+039MT%cA9%FUN-V$`}Z;swnYSsFtSo{I|rCWlciy
z^Mol9o3Iw*c8jPI1~siQ+^Hs3k{yn;CV$2vUg&vG3RbaeUtiZguRyQ159@bg8Uo`c
zyNgVnfty74oluhwIE<5Epzf!QKQx_-ExD_(0Vyd9wr?hUdj4Zh0bi2}NK@{HSZ24o
zarI^niU0o;_HEe;A)++0s0`XYF$KhJGy=k%MCe<0(%pfgk#P`{nc%JacRouL<p)4s
z+$34i3!sN7*}eiX5|XbnB@$ocydzw!SSEa{bvWm5|7|KDCr3pG=90tZ=bq@vIm0-f
zm5Jn&z%++RBF(!}otw^L=h8V2BE?~U*7rT0%+zZn71DujuMn{rn3Qm!@LaX1qU|2F
zmz~I`d@dgSL)nWVTp1A`(~PMZ51LHtFlo~R%Xw6*TOVJYRz7E%+OGNJ@MFp0@4dND
zh?POY3J}qU$%yv&O1MKN>kqx+>r)(gd?#={lUoJ4smLJoPv9NtjAeVYuBta$FCc7=
zD<qXjI{5((#fFPa02uB46ezQ>ks<?$TJglaUn1;EqVzSW*fCy;I32Va2EYu5Sh^pl
z*C?{U?I*zD{VP?j^F@qjk9-SQsTxUB+12;W^AUYOZ;&oc)gqC)odvmNwLAcZ7J2c{
zb$K7lVVlzky)uXRlbNowGp{1hVAK;%g=FManH$H4aK)e$i5s3SILv7#nE}y<N2k;k
zX;)amn9zmUiPEWqIEvii!ak-x?1FTA$%=@0o$K|;>`BAg)T*bGn2k$Fq6xpY55u3g
z6`WDo0!A5}XFoE+Gi^bj>?KDNd=5>J4k8h5fGn;z<K5gKyDep(>yh|6lBtkEraoE+
zKeuZD>nk>sU6)v%S;8;Us2`zXp(k;dbCf}!ldvSWdQHT5zhot9s}>7@VCt<ICd#fz
zfE7NtajPQ}8CsOH*#VwlQ3wr*E83eLip*XjJ~6e3bo8Tc{vZSa8YT=q;2}%7fCPCH
zh%6kCRE`8IVf0KS<ggfxZa7IY><np7!D*`dXDE#UV<=jmdFJ>LG$h<^_L=1^nJvPf
z2lA2d&M8U<=P?4q%DvBa<fv`HW>PY>C;=<rZ*6Zskm{vvF3lHzZz(b>!TyReP>!5F
zYYFdOSJv*Fgl3~~&I(9@OoJbaG>#Ok>!n;ACFOp-?BP+Bcx?iCg29>m!#3<UfFY7-
zK^}p|n!{uhn{{O@vi2pmw#uX(4$gat{|{Bnw}Jp-#?%Is01rQL4xqa#B2yS^F`3Vx
z(}6sik~;r#O$3Y2hz*-bV(uBAP9FFY5&!&DB1)>U5PF})e19R9^ObOUQS=A-`Z3&t
z>ls?iJU4?~JTPBJSY-(p*x3El@Z!haL>(dJ-YB(QsX~=<;dPB81WszfJPbPz4_mnJ
z^Q44GD&6_>$!!Cm7BRM&BNB5~Fbja+jKfLGr7bg$#9>gnnb;YfK1BRw1i!0J203d*
zt14-nc|la?-1)1!R~_)!LCEC+iUisW`r-`KVR^j@&&OF!j_(=O1LOi@MqEpctPKm{
z=$C<4l+<G)Qj7rTaBW@*ah!wR`aH4Q=!n?(cGJuAoof_WCE0}NiWW@O+aLt|ZB#QX
z-H^jK<S45^h=^)#*9p*Uty;3)ZF_gq)NX~Ix)G%M!|NCrWYQOn9S2$|`i9{eGDokU
z0ws0MV=O~!f7ZPn20N%*MjpBux%(zHea>G7U}T`MS452emd29Q1q!x1@50J??C&t<
zq-kvoB9ESNB&Nb2caKF#1)M1Gkq33o^WYAa`rjKn^)-rBi1n$UBW(fg7{+owJuSNw
z-j}<PUi{euSW5~b<N}>HFAF-aWM2uGWg_L7h$g%VVQ$~*_6{F8u2s9QT|524^V=U1
zTRK}$$ZijBShDvCNQ;eg>reeR#f37090U!du~A*+ZlfADieav6u)4EL9g67kI$kAZ
zAJ%S}(Bqf~yri=8aMacsjWb8vQ)+uo>s+abXhq@MK`1?5WZfe8cc^VVEeCZkpEE_M
zVZ>ogb1<R#Z)x0be&57Mga$X<142?&F=O4|k02htMx#+psMFz#V+?Q9kGQUsjVWJ+
z2<qQ0U(8E0$ob41`wH>Icu?)F+Uq}y0)+PZ@Wg#(cFjGq*Ui2wi42e<d)b^)BFFUF
z<Mc=b>95?xul$J0lGpS}4e$H9J~ibg43!D{OAJZwAM8)~z9JTasG|-*bY=N*d2!j<
zLB7Ia*7xiAG4BwrtB^|Au(0OCt1ys(YDchhA`t8o^x)hW{Q*{!rG6DA=F-cTU-+h&
z43N=PYTRO#qaDqmkc8F-g}&=@KlJDG4m{0=hW2+KAJZ6~A@y=0Ss0N&?4W)aP#l1W
z1IX|VuQ`T5gzW;j7&yE$EOykz^I0>jF%1?S{F-@|vrpGz@Ga`Xx$@KRFYyIWwrU)N
zUU@%%><d>;XepL|`<q8vDM!F6T-yqQ^ip*G7soNr-AtSC;nw^$v3r-@jOh|S0e1l-
zaOwUM9?L+BiGf=TIR#O6ybw2gTC4v9c`5rmtXW$u`rrMQh_km?#Sa(VnP~5{`O=Ro
zV@Y0x@nKhEP1i*k@K{bP3l~pNy0&3BndS}vm)<>NCpt!IKlK2mD#@SJMrt?Q8Rd4P
zYwpK2FDV+x6i?11#mJp?9)DY$&fTgRK?0=;VN!TwB+SS4Wx(}Ve2YN@DRU(5wotAL
zYt~96A0+p08KD)@%^SSHD%n9oTQm<pHmPEiPW>`W?(+yT76oIgE62FH*Wl)I*}0FJ
zNQo=Af6QUpep>z5S@qY7lDQ;-WA6YIRZ{n#1)r&dQGK9kEjN!GEpzpIClagP-!BU<
zypeUc4n^R(Q%d#9_l2r?6M+(A_K;20WU4AKgd2w(d827LaG{ebgyIbL<NezSJkuvL
z84_3gt9ZH0pzNjOCZ~ORE6L!MjfF*Kq2|-~C#-M1cbeTyX2*`ve(Mh!KN+O}c#1bL
zNdhwm(&HblSp?P#&ZRic+20Y{#VwF;0O;^Cg}pYDG!~30dQ1{x{Ag|vGA{IQ?Td~$
zBGS_iChR@!?V<{@YxL{2Ff1k83XW_884b42=AUv;?h<_4+l_uI93(zx&FHh0PM%$#
z?MRR$>V}9hLn4A(TGs|<&;{g#6GuzEyB~;AO!rJ~x6q4UK9SURU)r6E?nJ~`@`*o)
zTAqXMEV&0ZE4-_&lJ5Pt8RsQufTxLOUB|qwsD_iKj5jSy+yt!15^mA8!oTZKXUTd_
zdO0vCsB2FX;4x!!W8sG8Z85lFKmsdM)*bw1SxTB+r4eJ6lo(s=7{b*PL6E6ZIQ&X;
ziknf|af$>BBHCE<87P@e$7rtp(kg%~cDHQsE74`BKHp1r3R-yybs<boC`6e%DG}p|
z^+8TR|EHJQokuJGWAGb%2KvdpMPb&LZO)qU+FW7VMhu<;o|eX?mAbKtS8fLKcwKTQ
z545GXsAUHz8~5aF{X1Oi#eqAUKis*$-oI>FxzG}}X{=-fx=~kvA=wsTZ!>eAz#TLo
zf{>s7lTKRM@I4;gWlrydgTO+M7*4Wnq86mXLpFI|I+c>*7&^q!W)N8$QLJCsA$!i(
z*3Zw+rgwJN@h#uF52W^-^ksfcS&zp5T))WUb)5ze+=vEdld&caIWv$Dnh}n&-1gaf
zOIP#01{aTrl=tb}L`YXL%A=AaMtK}~i7;iBSU%9V`V{pxq53hYVSitSeMOJL9A)QY
z@XQtk_75jB&+P~xbcL@NpyV_Cn5+=XXD{N$``2YR=@BTdUe?x?<!I8xr|8(RNT}3U
z_)gMmXKb=r2B~n%bpvE^5BSj%?4F%N(iCFE+V!4LXhnZ17=xjZlR19G{K$o`=-_^S
z+XA+7@$}7Gw`Fb1hPnlv6uHUn`DkvqrZ$jA=fu00>?y)%$fLfdUESqj<HO|39!lgK
z`~arSc#H<PEbL*~R#uhqG=N0Om|t1bt~HmC^0))6A6<8XE_Ul>+*do)J<<<tF_eY>
z^9%je9oOKy;Q+UwIh=VjcGK4yJ$fNBWAdX%wn`I)cFH^=a(P*hL_@ncmlz3!-00KR
zu}pcFZ|D-+SzCU<mK3px;5JIPwIK7$vbi4u(^g4Pn4QVv=aB>wwap$!`4J#8KOpmZ
z;oOgLx&rLg#Ufw6bjLtoFo@Q?@P<hk^OaezaGN|wD(0XO6!u0@eHjgyXr1HCi;QL7
z3cLv^Lr^zEu??N$22Y;EG57x`4+_H4R)F4jWeeMrE*{<Z%(Bg{B7)P^ZlAMt6Dr)0
z74Ad8%`Bg5(D$34-dKcllo`XvjF&4gQkYipK~}?2dXh_yO1|fZv>l^`=S4WU$Ir>b
zpoYGV!rqrqZnC+uF~mO!8M0u=VK^s~UTnDaK{vV~!xr?;_nV^)Y7pV}VtAEiNOfPA
zLf?tZOpOq;d=MTaBU?9Rym0$8hs5L?3FBWm^qVRtZ5hpBwi1%75Y^&`dwM%O`mQvs
zJv(<)?MI=%!bJ04ghVfdA?<G#QkAHPu{$uYuNI~`3^Os?77iOel5j~H(!ptI9r2K~
z$1wy(U61rbP&Kf4k`o1)MAIIjw^>HNfb>oW-(1*L9`^f8SPlBR&W#`Uu8hpBb3^6d
z;e-5e^2}K9qWmS)T)^A6)8sZ%zzyudsUzDTsQtmD#XdMAo4IN|t```JLeGq|7(CC-
zNxMpgj0<0@QB$c<7nxl>T7H)9XsOXAZi?d%&LpW=`S#AsNer~z1ORJ~`dah5<MXxl
zGt=K}xU(;;rvdC|jqZiP_xcD&28)IpE!(n}-2q9~-rpH)S?sxK$~MsBk**-1)pW$6
z7fBOUk_(8uRbXM}h7Dc+x4Z~U^kZ8UD$yv=i_GI9QtR_!#;Y%m)MBQuP|m-2MTyir
zJbK#?oq9n{r>I=Qkuts&Wso|AJ4=|V*~*K;bDnL;$lLZ4_-z1!3&tsjsxig6WV#tL
z_;yTfw6SSJ&^oGbX#q@iV`;(zG)h|sjQHOA%%NjPLZ3x4Z^ozkdAM1CFYA6!CH-=f
zZ+!f(Q44dm?F}s=a#_ekoee)?Rq{t6zxw3}f$(A@5W&x4ehdNyMuX*E5(o;{gOMa6
zr5HDsT-4jYJT+JNi^)WX=i~*v0sLBFEx%@3={~A8<52E|^29kd-y(?5cYL8bV(K=2
zAec)d4PgibpQmCq3g#TYCsGvl=UUhwyw7_EVY`FgLj6}{{%;EYfYyrg*LsR41yP;_
z%5H@b=bAVh<)qeqZ+8qIwc8RgqT>AsbyiI&T8qjmqFCj*EVLVa+CA|CAy_B8>`CqW
z5ti^LU)4}^3FO6hqMrbmdPo|OTgQ)|t(5y{tl_A_fCn?^7#kzx5J?%B`}-6WmauMY
z{ysi5!fEp<;=$Q=;_wkirjP-|qlhm`-=#PDqG-BtTa$KJcir8-C|VK9F!D)N<098r
zIo*1q?%UCc{9$G6^n>*gtz|XXo$9pmEF*p7M87!lc{uNfG_;DP0Ylj3*!vSAKBnnl
zV>ejDDr4@Q6AwmEaN6dWjG;kOqJ1u(Ska>5IlGAv?S_Is6&Y6k7QC=KRqK)zH`lB_
zV`pKpzAGY5;>M11SFK4?$DM*Q#xcLOXU4j+2yE|c%xpSmx6tS7;|n~j4`r;ybtv19
zS+BeMz)d71g}-inM-Wp+!+M}+q)=179yvz(f2<qa25H1L1!x!d2&(o@vPF(6qMeTs
zzOyENfN!*$IMj`KCE+Ky{W25mt|%81uni;VbSI_!H&U+C=GB^W=MydqMIEl#6F<kk
z_&Q)Bj2mAjBx=pC2!K$7*bX4uzTox`hJAVRh``HkV;KT=bR^9&4E^cTu|uV#<^5Sn
z`b~D$=rYRslYa55$1&3wK~>z`-#)XIDvuVlyfv>s_iOXJkX4V(OP>xiIr9))+ZDxN
zDBni{r63rTF@&olLgfv=>qR;ow14kY5Q-=w?PF+zGv94Dhp8GFtnOn}G?}SY*A+v>
z6|ixs=hej7@yB7~LY=>ZtKzVa3=BT^B(bFx^w?b&r{stjXC8d##Sa(sgh>tYh_0n_
zaZ3fbZpV9k0GNHmGTipP!AVH?)ARCXZkKSSW)$#ZGVX7X|0>I?4hXR*RLB=(L;n&K
z12AVzX}oZaVk_d<ODz{*O0Tt8hm1bwO66!s^DfSLKAlHtLr0_xgJ5mJ3xZAll99(_
z7z;gt(UQK}IU{dF)Kce2T(m)nyT$7$DLW4ItyAcA@w==3+1bj)#5a#&d}TRzYZ-V?
zR#ZWC2Fs4!ST$|pO?-2Rc}pYYHjr$jKK0=qgQlQI-(dNFgC2Nb`ZqKKl$x|@GTgUy
zAksOP=udhZG>|Wx4EJTyQ1k-?%k<dsET%&AH>i;gZ|8>f_MDxjIujQ2<<h_fM}1!r
z^q+zTD46L#ARtKc{<9uQm_p;C(=szxqs_tcxah@2?5%FTXe~|J2fan-`<aXUT3_0D
z1d$zf9oCRdWlA>tdOi@@oRDO^_cxkp7C&CmR1kSqWOrx=qkqy74rLt)_7FEpr*@;X
zhq`-Z2A#-N`=iM>zlhCRdSKPviKH+axziIenbk;obO+k^(}t3QcA~uKnx<ZKO*k5y
zJzAJ)*imWk$~wDUOpaf$qC2V~d6l}m580}W=znHR3|#2KNH3LgG><M?vC03%WagP;
zK{YKGHd36|7BT>Aj04)v;nHP9*I4wSHniB_{880Ng2SWjb*%7B_Ts-56my;g0gHtM
zGH9$K+_3{=)*swqsif#9e7M)|@dx!{c8Gy|<_p4V+b1yNvhmVB_8ZXi>*!VpBD2#m
zjh<x)K!H!Kmckgn8$$f|ke5K#+WLNvwmY1q{8rCpeqrsUYUi4XJ4NH!Ia2v0WzL!S
zGflZ@7(fe@<Q3&toj}R*I4Mm&!AOD`?n0YQgTbPm)$4GvwIc(F;S`;nWK2sJ2B7bs
zW54q{5CqVsu8h2_Z_aZrgbYbY+hnL|n-Iwj(DCNG*M#MoaSd%7jne)}MP*vdZNGnJ
z-ng$wP6S?8DE(bA`rkRj=s0)c3iLACb(QkL;!Ag8f0<Hc^ovaBb<I`pjz{xO774FT
zxB8<L3*@Fj_X=AutpycIZxXBE>Rtfu7jIw@A^=7i&Oa{dB!uB3K$-j@i%xEJ)J&a4
z=;5#!S`B*jNiDk<F|9~Ty)kBYwc69Pw&Hio$ynKYWI2Si_PqU_+qRZ6-)l0HORUVM
z{-<Q58Y3D=8#|R{`tsKPNz0vugw<oRcMlB=VvJ2vQTH68yj*2>xT&aJMA45z+gjUo
z+d_BF6yGwdr}>`WCfiM$UR*POV(sLy>F=M~JJDloZ;D{6Q_h|8V?hXt5`uONHK-VE
zRPlS>VcSRaO0J(W+2+B<|6~fYVpvX(#zv7E*xmE@ZOl1&)yKF0UVS2ceM2OB+$drE
zTKFNFwPJAQ<2)9pX!cMiwy~*>5}0HTPlv??0C{#twEy!i#+kw>7*;<3VjC5td*iR8
z^F!Cns^MrzI}@N0y(xRIh+q{*$&K*87Q$TS`?m-cP@VT(C8qdF=90G)Gh+O~-RJL&
zr*?FVp(h)+4moVk4(E)g|0f;edFJeE%tinKFYDr4VGWB<w7v4Pt47axc?skkIjk|T
zij7RmF=jpb?G(w5tc%d0^HzPZTAk|<Zd^$<dityM(T`0bb0WVS|AzYYMm-CO2$8$*
z7Bh=+w<|hX5uWPm<WSy;Oo^(yHFFgf6YZl*#8hBId0>t@UFi*)%rKgFs!3HN_r1{1
zhOdx#{n_FYKQ|^<Ad_xBd%`j*qbbCm88K0WrpUZt%SfRP8UMpcrv<T#v7K@}@)zSV
zLjF;hJCJAQly1y;IUf6o6q!3CXt<L2%|lPK+P4dR#Qii{vGrN?c;81aN_(FAVZRv2
zm<zT@8vXr^*?28g=eIo8d->>y-orVL-&?><T7s!^PKDe5BHq8#aZs`~55nv=+obT5
zu+c{hU;G&TXox6h2ya<amdHn!gXlh~)}4(hi0UJheBb3)+s%kVQT^qW=mq9QV@dUW
zOb)xXn*2$-x}DehYX{n2so9WSN?!9heRt(Cf)#TJHpuu^W<14^*38T(-z(qCgD_M*
z2!#o<aqiMv`>fp4(q|pdyPmQdtM%l>s@p%W-`e`K&^1NxYVQjtkMsBT@0N_M+Im^{
z=fbIB3LQU>Haji%f7QQNiEWobAlO9L>cz6t{f^RBJzjZI7<Q{-g+XWh8SBH(&N@uo
zGRE7SECd3gdeZ(K^5CDBPX#oU)sdTIHqH;%h}>El=#JHw7MzDqk75l$vsan^)@tuV
zhaYE4ViJOV)ac~A1G+miR}VYyv;_@nr~G{@#^iy$V6Ei6AQuIeM2Ec*+ZTK?<Zt$A
zT$r}{`Dsjf_Mc^kMQzx3n2w!SAzy3@Z2nf5jQO-W^v@SCV+}Ai_NzgnYTCtGE8Sax
zA8gD|piZ;f!)Q3Q%4Z+#X=t|YzKWwVI9}BaC<D&ZQjW+#|7{lxob~5uP4NJkcC^{N
zjIelm)UYCd&oPEkp>wjKMT7EPMS)y}i*E2HoVEOriHjJLI#%rg2N;T$>Sq=4^2dcB
z#?pFf*ksznyso3P^>4?^^zxWG&LQ=M%fap=D;W?0IAFh%N(a_^{e)C)2NwOc=F0r`
z_;`mvZ=;w6ljU>Tbofhiuz=_#4cm7Xvm<6cL<aTWVNtFI{fD%LogK3nZJuj2)%|v;
z?%g|_Y}{je&?b2$YKJ`81BDXuS@;K|QLGM(c%m^ju@yHTHCH*}>$!a9Zp{iMn;*@O
zmt0xn@MXPzl>AYb)@N2%TzxC&>M!z+VylPjER7%zicHR1L>npDZ$zSsFq-RPW!dQ_
zK$$_Gp55~`J4q-X>fKy}SNM@(_o{7QhM6p#AYvS&J=w%$2)9mnN)+qOTJv&ILePbp
zoIuA=b2u6yzO_Hpo}RK_^UJ$ie^7oTF_pgGn~bR54!eCyW2_EMme+{9HOg`Iv`3m|
z2#3Otb8FC6+Ua=suizqv&`L$n1gQb4Tu{sB?y=scc9p*E(e~4XX%{9jVYeJZud0tc
zI+ZL1oGaxieEWk`%r`|s>SLOg#I49!vhe8c{~VY`%#L|0H}F3+emWP_>#KfmYmUx6
z!ey@Pl6Wsn2=!2a4evDVPd|EpZ;#78^~n;miC*txltk4{opA;)N?o_lts_$I!HR}%
z^;z4;+O;sJg{83~G4%M)l`}v4O)>L02kRdbC2s%R^06Wv18a3tBv=><C*;EF8b>s!
zR|3zP`ZCHL&)Uf#By;RVI~;W7(XBiW-=IY?S>eX4+?pim?QjIn#U(y=d<%npzDk%7
zI^=oW!T*u<<?&GO?faQV4Jt!YDh$e6DYQs3m}s*^5h*IWED4E{(I|B!6%`_~rLtrx
zA)!sl60(#+sfgkfDv96q9_sXbfB&4<={ZlI`7HPQUa#xAuZzJxo09@dpsL~Ic~gPf
zera@APL}(StA%iRxtT$3aFJEKA|W2&(OjsbhrkoBC8f)9)AjIP4e4_h{P&`Po<*DU
z6O#i?NTHDM?d@$hMdt*?b6iWsr)iQ+S30uHMV#>M#z)~tj#Ir(<pUH8DYQt_G;js{
z2CYm5j)%41q0#cQ8B-)y<%6RlG;6soi_t_jE<~+5@CCMZM01MYB~3OqZRNQ%R_$=o
z^mnbprchf08$c2*pOve9!*aF#pI$uQL*9XsK#;x21)Y$dV{EoeV4xr`LRb5mgbi@$
zx@>IdjYD4fo@(j(6}9W)7?H~gHN}Vgb55Mrp^KPRE9yNPp11hp1U|=+c<dVF`M7xd
zz{WOoigt^Zs=69-n0}O0fPZ1qy>qu=e=D+x8?58PT}vNqYNyw5v~gZ574`h0^{e|i
zyB$L*I3cuS69mM9->y{Qr_$G&5o(|MmvR{cR~6}5C(1=dA}=4GuX-?>F^iv{_T7T{
z@8p2b5{D`hCtNzY>dii9>KTzNRcY|NLPKfm3WD=rpmFtdn4QgN0U1(CVgZg44>d@>
zL_~+4Is1^_1a?1f?w)68qq&BSqBRnCc;rCNI;kLvk0=Kta`+?YkQ|-M;%QLkf3xLl
zO7+2IMSI^bpxH{thtl|2ix+~I8WBp<lL-n&@~Q<fJo`2b@>Dm>5*y1tjHoAmCP{jc
zy`Fs=nQR)tks{u7UEIW`5pi%ap;ceMJwGlIwAiV3Ufp~h`V?MXDrFwR_C0K1oP6b;
z*V1lEj(w#ifEx`l$=XzUd2Ap>2qF{hxtUB3`q~kXR(Pg<`g#s1cS9Ze!GM#U67cmU
zB9u$`&5(<M(?dKy-~O;t#+(r(;eQ(fMlzF214jp!>(IZ8Fz6JXN=XDMS?a`{G$#R|
zOk))|xdSWc>>?UzVB-;UXxL|4h*c?%L&DFyoNhf^phE}oPE?2f#edBT4g40}JTN%)
zO?K9FSnnba)SOE6(kUiv8ur^QixH8Qrq2m7hMx`zjonE+2=_4jm;v8!%+Lw=d9V4a
zG*gsQl(o-<c-}3@Axvf@V%Z}{yChO*XM0(&EsJ)~+hFe$4kL-C=RKYnF&QotepXF|
zn!X-YS12|}hu&kolVAG^G~d0BR<DIqVNXB|p~wHT#vCp)hC-FLHB}9CRS-KG%bNh1
zD-MI&r&4P-emPoJpm+$``~dpJypkgc-Dv_%QZ$HN5Gf^7_aicN;v0(SIPn5@Raktu
z8ZWQ(*vtnNi*D=C0a5?>u33HNc#uWbSu54VracMCLg?HoE#1bioew!<KJAe5WQPlQ
zV=O#nkni(ORh&wFn8nFC%a^&H7AiI76vNTMg6gK6rBijadGHOnPN~l%kV#d}JLJ6x
z-`HEda|(Vcd@J&U`>Tr(q;`siOlA4vJD07pc{n)AOVj>7Dm?r<=rh8N*iB51xQ0(h
zj_$>m(+P3!G;is2C&M6;5b5JGJu7PYDKzTlHVC^&)1~4Y7kzS@9u&SzeDa@9W-+rS
zM?A&Aj^zs-ib~nwEdG2QIvEXdypy>3F%GUNtvx@CoZ?y$4_@hmAO=#Ulh%n{yw^kL
z@Xt^T=apJ<J}hqhl1xwsZtgjN5+sJLAVXo}7)9^~Gte6luy?%G;wfN39}CANmypAM
z|5{L^gi~Y~7)pC>@w08C^}&C6E$9b5#5+my$u+D#YaSKp;YMl7d~;eEXic-(+NvyK
zRZX}6_2<wtf1IBaB?tt&fil9|&Suhd<dv0<=-}hxMYEX#C@DHC&YOJLox_;x+s0PN
zV~&|fFL8^nFRDrh2KG`Tgxb-yT=uE};{%wI>oNt79#e{j#@;zGLoiUGPGX${+2`?(
zNAd||&_vh{xZ{p%au=7UQuwu-@?6gEJApDnI4IoIbvE(-+4JF<Zq9R>3@5eF1DA>6
zk*Bw#b-Of1L`McU>$qtC#NXRV!})Nt*#fiouh2qFfqs$uYZFmBo!>95n!L0hA8VF#
zUmBCzd;P0dO49-whtbhx;$ivc-ogUo1{eGDu@;N$RD{hBDP|~e7M~2eL0+kps13i=
z{h8+z5J8>EGoVYO&(Lu`Se)sq|H%A`aRMfZ#l0%x<6DP$sM7PunouOoQ9OckmhE{I
zCT@RLr-oM~h|DG~Y$AH2e@e4z@=ky7UiPOk+K5-fGQU~-Ruex@7lD}ZJ`Dk)oKQ4H
zn}|{&U(hWnr;i;6Cc}i%4eg&S8}|8iWjzk7MMC;|507e?w6#ktkfOC07l%&cp_VPg
z+sw~YPDEN!PqkvxNHf~`k?G4!ijP&1#Qu-9$r}v)3vkMT*g#)C7bz8K2A`b0boKYP
zBNyXNKm4ocaO6*b2oWOL-Jps6t{*`rw<`2;ZK&S0d@Um=T_`16i3-;2&Vu|<ULNXY
zXgl*Cu?>Y`GS5(eHBF%i0ra7oI?`5p1Z9hHu6$1Vx}hJr(MnX}jVuJUn{pgZiQdIp
zgHvP~?yt<a#+NXV%DpIVn_?n<-w4mBK<APW(4}1gSF$Xv=<+~RjaGB=#A>>ahf!!)
zVHsM_G}gWuC@Z{QSV6&)B7{o+^efYv=x-DV;~e|pw^5YZ`%7Jh_>csgSuu(ycKdP+
z;E0eNh()^bE79ApUd!Nr$&gtVH#eAZecLU=>65B1mZHm_jrxI_fyBbFZ;?Am)wySH
zwTS$t%N%>5{^9&rx7P4xA}=-6*bu3h{J~xAe0<|Nx!BF=mP-rm{oWz0f$`^uk#1G{
zlMoL~@9hs=mcCQGce5V_^~2kLl#3n1w$Or9&PExK0v>4`NjXOwLnLtaQYjfssvQfy
zw@x7Z(^-X-dD|@LKA&5DpwwKkZ_<!;V|5o*-2A)N<2S<J`+lz?{cgrQ9l=-pbq`pT
zZg^*h-#v+p?WF$d1*!teqHp&<Ob(^l^J}wdm_Zc6Q`fniNE~J`MX6d=yhG#}ZKYvK
zvVL>nYuiP#CJGggj`&lAD2Jj{VV%VXZ;Q&-ZrC>|`ef5h*KM<BBT}&+{;qS{Ci5IR
ziOkINjOGvQfN7QYO)GAoxYoq8-wXGfgz7-*;f%F-{frO~$OkFVwS@%N`5bXB7G_Y%
zTKOvS2D%W*!sF)X;}l*Xi6+K~RhP6{#pbbTrFWVNx<-+0#*AR4SSsXv>{Z>gqryK_
z#(a1x)_E9ZoA@ou^&0Cq;leW`$E~PDi{f<nWHU1RM&3;15S#Bcqa*-W$X7Md7_y)D
zch{RTd>BD0()5t<txs?D3J0~l+@Sl2xX2D1+O@AwL?enx@$gDJN|%adQH0QT=^{`7
zowtG&#%KAojb%G<#RiMGwfiPv71qgh1LqbV%{}bzaBX7lp9(oiE<8PR2^Jd`#-6>6
zX1bwQ(8zK(!}s{tW-%ceYi;kBcfaJvT={yW2-YGi$+t+7!nc_1I2XaIeWoM!#)^l4
z9NLCu^s<r%#RK9Qp^LR?L1+0arcd%agy?cyRmx#?!=7J*RYh{S<O1Aa0VTb}%q9N_
z?2wiFioubqy~tJb%+LMdOin~bgqBh(OyhmmMf4QDo!jR2uY*?C4DuJ5yAZjBfA%e+
z&{*byOisMDKRQ3VK0wA*%HaiBy09ZG%!;<ddEcUO2U~ubndkg`{ka9VvDtaS%`WHl
z>QS-oj#UBgY|TDX9Udn_s=yW4*Qybe1V{naeE3?4vb-Ijca_+A)KX((k^oTTjMY3e
zQ|{pw0+&+^QN*5Ck-i}3_|%ZUc!sf}fQCRBY>O;h=<j^?76D_xcZA6T3@YC}p>x*w
zBQwI8%$@n<Zj7b6-4*1%0xgKX3sDNnb8mf}I|{sd=$z$MdrbQ71|mzcy6wqfd>=jY
zOc&2$<BoXJV{nvIjU}x}H6#+_k2zvnmsrla(hn3LTF{q)f!@|pfX#lQycl4`Jko4z
zCK|b?jGlCjyKU?);_Gl+m`Ht9XQ}7cqkGNZ<BwC=Wp*L@Nl1J+FRoNg$)IyE_ms5N
z?)y4$+5P{#bIZ0+3}n$@pTfhY9WDESd;y4`X0L1%Lh$3y?ZJQfP!;<VHL<^#Xxpz$
zH&u@@FG4Q#p0Xn#>e-Q*(Q@Cxk|P`9q}QCd@^x^Z6tcsBrJ63KGlJq3sMUs*$GJD0
zy$F`pu+E8&%yvF9e*0B^5$}PEL|S`L$!4^EGd=KJWnQ5lREHK~`!HmQUa#WYlBhTJ
zXU`vp&9>q5{c5KtHq-08g@7R%syUm}OpbO&m>MDjaik_AH#FuJ8<Ug2_9KNxwqk~a
zZ0qzvVy2@rkVm1=yAVnneg3bV6P*5DU3yQ|Ir&XxKTcj<>H{HE*SnDe*=2nYWMa&p
z*aM0R)aUc!YBAvWi6)=@FP2)!Z=<)B??a1V;r4>fe92%QBr0zJ7|jetm^<W&ro=4l
z@7u8GHp@D^pQGP0r*g0k??vQ9*D-!>*m%Yw{!nPlAIclAy7yLnF7B6h2ca8XF0mAp
z8tCeA0_sb}<A2GhE~A!8G!ixz47<zxhRXa-y-_TX?Af2Q)y#}}2Q|#j7n^O0MSncc
zp5S*q6SQ{4&Mj!AXdZU>$TfQIH}I6pg5U%8Om|g$LWAkb1cS=CE9@$zJ3c4fSh(D!
zp$8Kqt#;1zU9Bqq=y%?9eTf{Uu{kj^tl##ic$$r~N#BtYm|<qHu7MFvd$EWJF8g;o
zr9=j9jBFQs^5OkOBQYWD3*UNu>ylH3)lMctXbu4vx;w1%<*m%Kyrn(I7j{0p(~vrZ
zIbhnAA7o9OS@{DA1M>w6zRf3RF0|adZnj!p+eXFC1g%P?xzL$C5tVgqlipZ*$wM`n
zQFHI-(8q*kibEOG?$w7%cjq@Ny+1-6APKq0<IpwLAr};=C+&c)zMg(5blZ-PsPjqA
zbf9aJ+!}I~^2-R=7Wb1B@Z!_);?+*KO)<J#NC>vj2MPAc)t0y*9%P(mGT&U3LzU4$
zK9On=lalodT~uuJVrW;k+(|yhGthr&2O3jx>i@hKR68BtMSY-Df6cu5xaA5YKC#pz
zDYHo*jC<Q9LQtPPGJQ+@sCUZLT)i5U@a>Y4ht}2UEC)zI_un^<U!Qh-k(;JOC&?gH
z`um6I=r~8`kHmq+_-&N?{AP0sg?lwN6B`I8YtWA*oF%6*jN1RtgJ~>duH5zsCTFfw
z>Z|KWhsaTW60t5V51nn(#v!sdZiPp@&QH^o%iQc<bvI7m$16De>&k7O)c7?g&seKD
z;lL05$~^z7_{;LnQcVYE?a}U{ewUUEU9s^DTt_+Uz*qZ2g^OF4Y;M?p6cZWEdu*2*
z`f{^te}pA`A!6R>sL9;3_s%blak7}Bc_U%qUsKo0X{@ccJ?FD&X%s(G>$AFk3m&SU
z&-jAOucz{Wx&SvG7$O{&D=43o31s@B9o9FRw^0LjA8mh>hyX(}M5p45bMh;69-Deo
zEE|kI>+vf8WWF<LxEEOE74)MSJ#W84!#7`Gx5W#5o1f>}o-Gr3YIG=-s8%+4%a2bx
zm`q-^Z8m1op2b?7-K_Pkl5eJ2wMt&>oib$Y@Pa9z_!&_&<H=Ci>2{GI(LI}2E;*Fg
zAbm39;n@~6Jj;SA&qNyPQ8mp^ceHQ*$}EZGWz(sGPt|$x2yG`}P_I|)6NNz)O<>bj
z-{q_~+ZkYT_1cr5aA#VOCR6+xlbU`cX>uCr|L1<?|Ja!S>N>Yu9R{m+?<x*-qc#}{
zUA<Uq%E1!#Epw0V@X7HzQIq7d?AD|~ltP8mewW#Ok%4i`j6HXMaGLn_l?_${i20c0
z@UFjYTP%1bUYQRF)r~35`PNQn)N9fbw~eCnP`7)=V@cs#@4+v8`b|@guDxXHUYBfD
zWdxiu01Whz7Fh8bo`8ev(1F{ux>75PRMr&LvX-?uLbF+8WznrjC3A>mUOY1}@bS3|
zu$NA1S5kkRS+$=1T27>>X9!tyy(d-Hr1W-Vc&+8bm}Rcf+xoI&G|7sz1qol15<b&9
zlSKrcDmAgj-Fly%4JsSpwf6r?ZyeVC;v(_zQ|cM&=4)n^Bik}PpI_bb*38h};fuAC
zRr*r5maXj+Wa*xtjNI#Hgv4LiL=spj&8OET2he%CGyJ6@#iH^veVU7!Jn{M3$akeW
z&BI5~joVL}9Gb5&AGojn4sEp!deq+j>2d3%XzSH&uG-EDvU#zWALE!<0Er-t<u}-q
zHvRT%gQB#;ug8_<1(@G0>^HSgE_?x9(2ECn_Z4a{pKTItz7-kxWM1XMX2mU6MV`tI
zw@Ob^?ABfraQ5Zz2P|jVK<Dnnyv3V%t6Kj>zaT33ZJkRIT6gbH<&PZHsGEen77I+0
z*7{9-sHu1|Q|;q(RMmCpX7>!Tj8393{WkmEp%0|{b}WjH)NL9!b+a`my#~M9+4x3Q
zC^w?ldQAF($(sQn)mY9m^~Z%Hi;Ju{jQoeMAc;!m31D{slW^Be+UHK;txB~3_uh1c
z9BZeu1J2Fa$<2vnD4(rv&Ff$L7Jbi8E00%b%`5_#!R>0-<hz%rt5H9^YdKRhF|=UI
z&OafAy_M$y-7!jB?qJQvAJ@>KscL%ixOdY^HrNaGBZWN?k&hLA=+$0JtJ%0_qw}i=
zw$mC{p1_E9(_Du|acfpU&SX0`IemY*qaaL)s%=51X_x1ze<CM!2fFla$;&=qY*8YR
z>s0q}8f(@nUotcJR9bzr<lZ~a3F=d|votcceMI-~2OjP_nYyH{OjohJJ7-%%@33W_
z^MlXT4<r_v#hZ6Gq#L~T<KF{)t3g+R$fXA12X{<<HDWB_*DNG4l<Q&)4{|#=;icHH
zxlM-xH>+hCFnNN8WUif$sxmiec71u%?oGVAc!5i%R^DK13qnoA0t=P1_NwY;u|FWU
zE_^Hg!xR+PmpLYuOPB1e-1s_gq7aJrrHFwl(zb7+B6n}Mpu1sid|hpW44gI^Uw?JI
zp(@SldzqZAOYh&Up=3c94&9{nYQSm>M&@v-hn*Hlm7X0r(OUiB-hRFz`@6|}q4xxK
zAlp#=x%6;b%iApXeVt3zP4{+l{+u4aW|r4smU3ax#;0CErZuBvKV$Ru{F2o&<G$mI
zbo^vm&)o#Y!{{o(%~WQo&QRQ(mCC=lVdV*+m5q*8+JEyHQ2bXUNFoQD;X{fK*?ud!
zk&cSW+FjNtuBRy3N`NYTU-7_geJL7HStrzGeorpxvys*7>{Ae<QE0wER|o0WJdV{%
zQ|NK%$cf>V(w-?2ojxT1B7dfE*SQ6oS}q<dI2jioTvW6L+OqRa8Y)&7+H3b`CAm1a
zBfrY)p0`M8M#plTuxPh~ojcbx*6YC6eOme3d!(0~TewZl&TaHi{h?Z<X$$X7b2%-k
z`s(w>Ii0=P4YVg&7J}1^`X1<g>mRt&nRM3hOZ=QFg?{SsC>>))8PP<3*$@{SI~^OM
zv_05a$Km}rMmSl0+fDU26DaC4YowRqXc_84>1m(bDz*}}9rz-nt(q0?itT++Xi2?v
z{YmbseeXA#ck=}}6*|A)PqI>d(V`wVd956`h<vwLXl9jWoXsCb3K_7eFZ)`T=y8c_
zUawX*V{6ZUPqGyRWXt^OsD&OQg9<lofG!a8(G|0cy{;o$#?0Xbu&g?(`JlMOw|`Fh
zIPWxiAnDtpWMGcoXmkky_(>#cX-X7l8m5kJrD^v)pOK(S5`#sUo|gwJx=IH!%XxF(
z90ctu!m3g11K0n--+Q7-h|b&l`nX`m1=W?oWY-HQsKD=;p!qf+^Q36bxAQj-G=2zH
zv3oMtv@iST543Gu0t@c#GFs@Ex?OIsh10CaLLGb_ot0{tHg1{a23?|FFCYBQ7}(co
z@@q=b9ol}lB0@?2-ot_Yrh=kPhbuO(?g-aa+0$TttgyIA{v|t3lkD(_9Azl~nn#yX
z?ztf+uI3nSA#B>(S{_8-aG#(77*rHaoHl7P(=R^~h_a2zfkV}VBY|j4m9Ez1L+j0H
zIvF{Pc@I}ZX^+ra{(bwX`$KLHfB0wdd8_An^nP?mkge{|_Af@Z;IN~#CCnn*_IL_^
z4;-GQxgS~@87Z6j3$`Sjzb$T>V{h;({zkX)SklR0mp}{_Yna@O{I6aAaU{P&P}cF$
zz&Vu(PB%+_$k>e-ZozTy7Ylqkb58n}?|_^pBLDgk+bc&e9eDh-BBdCFOd4dcM-%mR
zztm3WhFV6rDGrS1lMQdt1+WCFB1k$Pfi3vGu7=I%)Qy`MsNsqn#21)dzYzDJ5Hp9!
z26xaJ3)0)PZuWdzE5<NF;$Sd}BJIYlJ-NB3$rtIK-AzIYPE1X{TXUyunCPzT@rMYB
zFM*~lxL57hS+*&;1L4O89WEzi`fRzndGAPnlj#e*&dQ|bK#6PfFNhrQ0F2gYT{W;V
zPX!YZXM4>{tAgTJNuuxh>qHH=z|yC*e0n0Zpb`f5ofmyU3?<P~s<`&~8lCFj26jO&
z#(m9D$CFQ;wmP@X`(s+_Y%$F*xA!!w?A!L}c15D*`UiVHidfd%OY+0itFpe{;9^s-
zj9OjE!8obQaFR66M&@B-wMRX};x|&SnGPKhcDbx2Z@IsmHRFPq%Se2a8KB*bk{13(
z0Tui#<0+kBVSe(`t@iK&486BHqNDMCMGj6a0>@H^&3=o8t;Xg_%yJwJS~IWRit&~y
z5h;cSTe&*2)Ys2Iuj1FfpkjAIeK#D{@Zi2p-F+86Zwe1Qx4-62z@2`{FrH`k5;FQD
z3+5;Brky?PxY?!mD(@Z4@5Sq`yM2?73F=-J9(O+Uy>8;sVA55mv)JhL6{K83digV=
z?8Td~0>Bewua^AqPcXnAEt@qC!IgV$td6*#z`?+ukiaZAJo>Equ7!^t6t1z&l9=JJ
zY>7xvYX<F?_5?PAPo7@dBIvCmZ3X<Y_H3rjPv>Kdxn4Ee@SKNijtO^#xpc{WSYotI
zdxN<~^|_U!-K%zzU+p-bF#^#ZuOsV}3oTyFQaKe7{ZenJj}y@4Jcq49GFb;Ud}(Ev
zqOry3nbZg?xzC-sZE@YLqa^Gt+=k-8XObaPupi!sv?Y)O4G-F`YukLbtFJ6r;1iu%
z*@9aRo#Rak#6U7{uMyC{CB!V5NwT->i9$!Gh<T#GrT4{kyS&9{_vI~REs=e$pY49m
zUmtPBx-Z+P!1d*_R~u?8o-VplWM4U(B=*QlW+KUSeo}pX&bo|k8!S7U$K4F0n?ibD
zad)y(PY_BAhCued(stjxN7gMjeSoCy`B0D8vWoWAGhxNbGt0TKmQzJuN6vBwD84dy
zsmrS_qT8&agqZd1{5d+cxAJSiLpUx;i@t1s<VK>J@mDpB2kXnQ%aV8tl(~;Skht`j
z=>$I#Q2R<g^7cf#>6-?R2c|U08eY88!>}CLplrwg%+*4ZdUb{+QpBWDO-MdPoaq3r
zaktw`x$_I6U)9{rwREc*aL+p<W_u;!sV>fL?r3_3d1f4Cn#!JY_m0TNPxg!G_<<>m
z=Sm|+(Bzq$5OwbiMym_`iMnQRB4vJdVwjux&TGRt(4$g>JlAJ-OkZ6MVr_d}{=)d5
z`i~HmVl5}@p2~h<bn~NjBMw}%xA1xI>4SAa=eHZ&O1|ie#CCm`=#xk#hsK1^KnDu2
z=#QbZLiF65s&9@=uc@#o@5ZjYP3QZq*P}y}*Fsgg<?F+i_LqL(k8VhPT*DT!sO)hj
zYvy^w8%LKq*L$o9702p?y-GQ%(G<Vp`x~X?Wy&rr*^LZa%z)or&ww`m?&*D}P(Y2`
z7VW2EXrpK|S-Mlb5PkA9p|snD=Q8NjEBo}tYt9YOZpP`=S9~7DQiG2L$;qReVSmnz
z>#TO*q+Y~>e8SGBy{~F(Wo|6EhVM~T5SWthWlb>c2E}#{x{F{S{1zw-4yZYcD#l+|
z#qzg(?_CTy?Dv8Z?}}@7Pv0z$pn&F{)jvW`f(KlL<NrNJ#b999Wa%lAtBcS>8D3NM
z%;5-~Gq<#M0r$Az^$#EP?wo_cpDvPd0;~oeIrg(~NO)mI3Y`N0NAw^XgpW??CTdHu
z1I@giU4#CPkGSS67dL<P{iV>Kzx?vLPwaiW`3D+YKuSo7cJthlX7YfgxZ)qimdqX7
zmp|Aj&JRE#k)ni7t+wQ=+-+@4YAwCydf->Vlwa;Clp?i%!u4F<hhwJECAslK>pE5O
zYsa^KTJ9~+2qy*U<85|Z>i_^uoI@zGN2iim1-hW4WOy{X^zt!ysR`)dpb_15CC)7X
zbC8U?BSRDVH5|9kqe?lNKW+S_dA%(;@{NIVUj0#t1n-pT8|#rHzPi0HHSKJEylAe(
zogIT#=iQDuiI#fDY!rm1+=3;ty#-XV|Fv+xd!wQ2$H9}UAEF2KT+VXuq!YLI%HF~-
zC+F$;m{XeK{$&f}58uzFg=%hT8$Nqw#t9X%{1o#UH{V6?AUuo>5%Q<k<;?vjaN%zY
z-*ahC<lP_d=a`F9zpO#i(uXu!v<PnR!!_MjBn*ow*w|5kt~MBpXzH9URw-;W+AA^W
zYN12VuGrurTw}0k;jMIY6sl9o5gGfl()UfTY%!~wc>9I>$9aoZEWC0Wpp8fM^(R&B
zOg$uumfm`vo4X&)X#;$WS%k|v*j4vv&&ptQ9mt|(OlvZ}oGg6pxeUOdmqbbV((3J_
z0bib-*dX$6EXokoy5y?6?k_eEzMWzrGGKR^|Kj1+{=5D^BkX+_P17$C#Y9By(t17W
z9Jy^kLh787(9C|W@`+oKe76WxKrJplbQG1XFg#hKxeEYX??niJcCUoq5gkKb5vBMT
zYrkfPX4H>htaKC%q{tPYH4?j8J}LQ7Lxr2!t){1kmQT;~sL`I>>@BZv%<O*aR9i3M
z3rz8&lXLb%ui>S}?ur5a9Y{d`IrNX_rs>9KTpKrE@pMzl(>pqCNWLINC-T4!u_9Ht
zCZV-Puy%m$2EjG&nrq0=@eAb(lcq}XOydr;+R4zfrSbeTsj>Q-lzfry&()JZf5$jv
z&95dpHO}Fww6a06VtmV_0f8NZY@pFnRnu&Y7NQ1!h^+bg#wpe_;q)z|A4c6<oFnk!
zx^Y6#_G6wQys{fFv_)&7U)F`?d`WgYbyhhaJ$JkWQ?b}4g|F_X?FAyHuRgoX`+9@o
zrk;qAUrR<spQmp#?{>6nT5wDzd^;>19YS8|RsA)(w_YWgOo{ld1>m#I$4Ka1gGi7~
z+_{>=BOC_n$Gws!U>}^3f3iNEPz8>#JYlS5c#(O)0(g!!E*b(rPLNL3f@OxLG}|@I
z$9GOC`ZlL`?TB-Gtu8nk{qg|1Be^JO++P(8d?gO|?#tLY2j2U{na%jHBhEl1Ec=Ud
ze(mOE&h?$}^}E`;Jgi!u>L)~Pu(MGay}9BaA-lMVx~k{xlb22xN`7}-rFZc{Mz{+b
z`yoqF9IrkIa39uK%Qk*Esq+)}+^(pq_kX4h%qJAU%xAM{3fJcE1!M~#O?$<UY%A`h
z`;tI}lKJQSp!2V)r_Nb+;Q;T4bC)E<44u{prLX(4uC&7KCSXwSyKK-ET+U;$>Aoz#
znT3&`Vszz*jl2UmTxX-SXbs*^&D`?>qT>ub7V7n5kRBr!RHt!zBu<eh^sCXas56_L
zKlfX{m8T9fexKoN&+VxC`DIQL8|Rq;gNdIR2554i5y>Y><6`RambSX$NvYcaH2Glp
ztS*U~>z6N_)os@ZIfOjbSQ|GX=N(eXMJbkQ>oTzt`emi_?K%6l)f|Y<$NXz!Su?A6
zq06tSb~?*mnHKZ{4X3Sst5Sl0=M3X%gK!$^qaqh*N06e*UV5OJ5t^NEj@}b%#)`U0
zFLbYd%D3v7k^$d&JKqN|S3=j>SOMy}?dq3fAh*i^2<eDqsO-kvzL!kKK`?rznl$w6
z^L9a<{j$mD8B{4R^GKctLdGv_@1S{P|GMtp1BgU<ywa=RT@Losa&P{H*BAKSL0*`W
z@|o4W)2GpurFrz}-OWeK1$lWT^Gz6DJx$M+e(E?DP<w0&@6*HY?XNONAG-`0i#4h`
zS>#Yi_HH2gQS0lQ2R{bRA6I%bNzVq3!G|OPDtA3mLO_&M;p;LdE{=l1q#QgxrRKrq
zYw-6|NXfw+PCky5p-N5H44ax8_<5jPHhxERjM_vDHg8GjEj!I$<*KilHH^G6YUOc(
z0=u*!KZDqn&+Ac_Z0*@48|_pRV0tX*JO$mg4iw~1#qy*ZptlTcOCY~|Q+b@C@eep+
z6N!P1M-vari|py(h+Ih1p0B^D;oHXOcU!5Hg=E?1B>>b=p?d1=1^)D`kkna$t8@;2
z^HnaB7L*WAssR>(*mPRC=%G(3!PjYTjU1%XQu@ebL0P(#t6wmqsoS@|xAhY5*4l?1
zB_6cS_Fd_$B=eUyT`NP>4%1^ywg^p$c3<o*JTsSh!p-hua`K>o!K<K9gS4x`US1Wo
z^^kwyG5FdUWl?CS*_+0g0bKDV0Zi?ymOf84@VAt|GZGUahl-r{ou^Ue(l8yyeTv_t
z2OFgvBefPzZG?sa{nYYQ-5g!|v=wcEU{rVXd#Q!t#u;t8ND>bEZt4Vnkh%kI^L^S*
zP37yMG(ya5(~msG1IUQqlWcU+esuDFV3qVzEHAQd2@PaxTpb8gn;&0zZqT&avZ0(f
z#s2hG<Z{jh7HN|&dN2B!_5m5&OK*7x`984Ix~?{#B>VjaW$mo_<nP?>zPoh$>Q7Iz
z&K4LsAl{6&e5_AP^o^Jj^Ui}x5LH~?@qSlDgu?XG?GAe{+>oN7F=Q?JG8Ll{0_CA0
zfe)!*$-x8K5D?;3@eqvgAce88J2cgA#*D{1k1-rAJap(LOOpa3c-KcNK96s+lR^I@
zYSZFilszwyWrEsUohh)xXX&d3FSorfOa(~KsoU6ahfPcVULU*UNmzB6ap#z4F%?kP
ztId?JtzoKb&etB+-s_+7b8mE1PU^e<OJ}}hJW79R2sk59T%36c!se#C_#Qk|_g2eW
zC(5~RMG#yx&KB56)((q@!4oJ4L=h*y3dCD4ND&%?1gZ$tOuTRsHzI?e@A*pjxj`+;
z1#*X2*FJ1P8US;q+Uu2y_<B8)J+ZAJcDBZ^ypMp#OSu_*s<+$%=uwWFt;>8R-b6I1
z4P1<noE>r>XuebwXd)t?eq&GU`LC^O*Y91mw>o)rQst1%`?~^WX(wDB7*bV4&~t+B
zyvE))tW<G|P;Fnm>cmknvY)b_1eLVzP2{4~LL!g_pDt)2Am^g#_!~r{1l*#Psc4LB
z@iV7k)td=#(rQ4Bhkx}RZagG<r$_f<WXq>C-by9Mt0lxf)`grRuR6WM%`Il-Ke?Ul
zE@tPlj)Qg7X*n?!i@1GwBlD7q8wLU(1>&_ScUNThD_MhR5M`%gTkz86tnDiPFOk50
z09BzUjS{(Z`PqZZPUkOF5-xvi;&%AGphnN)?BiJbBuBYTH)aH(nWo3VMz?RDUun#g
z<9CSx*ct4L0UcE`yN@@9<@3+K%w;yDlt2-8jUGqs)5zfRY(|p7BQ}C%TS*W>w%Fdf
z>HJ)=(bY^dB|Z5hsj>==B%Y7jvBmC{;g$=h;p#;_pY488@GCk$EvpzfWh;M2&nwn<
zi<=WI=DwoLS+eUsIKP#K=A}`$H_eVC$(Zx=s~P)ciqoF9W5&?<s3}gOZkRpjApMpT
z*ksD9s3gz&QLXB9ZZ~XOmq_4MO=Hz)#jWRkTI7A^Du*$TZ_QHDdEzILe=MSwp9)l>
z@9b+UwPmgSvR<A_gy&MgIFa^Oea0|mmIzWNBVjrAChucdPI#DRUi_Tw^$#xE9(!J<
z%&Y7_Ao0rg`5ku$z0#I*&nLbr*c4-%Q<}cap=xg7SD>X-(!BZ`paqsXL8!3o^^>7v
zn_bMhT5ExtcXszKyVLr^u^D&m@*uTC>*C%+{q5IHl21f^_FR1lRbyz`G1Wyh4WYi<
zh*wV53>|uH%3CJ6`-7DBQ8CeCmxm$v1Cs)Tz;SBoJ=8sfMvtboaQw)JYu7Lb0;!Me
z20S>x`HGItF0}VMcii1Ua_F?Q$pr(-afiCircc47++al8eIX+eVIWH7f!eigxmDSV
z3;eI?%T$hp&k)hIIkmQKLe=}Eb!Qjd>9390wL14)J=kPvS(COi3M)O%FMSPOsj;Mb
z`j-e{b1P9fmY$@#YGiWqtqae?$VnMifLroFKC^IJ!=*a|Mr1FD5|A0#1ED8M9kmko
zBX~<+Wn<Q<559-Zs>oQLrO!OoP4=```R1jt-JCCcSeJMtymoC~**d^}pGraDqN^c2
z1`u?TXViFp^_yTdIth;GgiZNy;gyl&HCiG(cE>;YAQk>()c_`Z&7NGd`6hl7RVUYh
z*z?b#f-5LoqHVX7O{~nL4T3jwpp8A*>VQan@nt&F&Y7PbI%Ztlc~0Zhd;?D-(xy(M
zREjy?4E!0o?&GGZC`>wAnHsI9Od?JTCDe3#gLoaIz?nF01@o}u1s!>jlLTJD3_&$U
z`}cvJBSMA%&%(-`6ruVAWb-H`zn96gITjDr?the9yRzC}R`l(5YLHSEIwhKl882J?
zqSAQntkycQhV-|L=B9XsxGO0Q>(o8<5p6}uYlKd*s>e`Ey!OP(^SO@Cq(9S_?P(&M
zim=&B*H2&O!>j52v-<Ez-RtSu*}Hg6hMGEkF1XB>c2&&uE#Eq88U^JJIxwV5Tb1%&
zAg@hgI%ksH&3HQXSJ#q#!$TFdx7=HsoO`&K%8)SluxGa$LPbDpk!3P;zVsstAs`wI
zwjWCx8el}q=C<xLZR|xZ>I6|I?2Ohip2Gg@_HyQ?$68D0$3N1_%fZwIjSkC%<4kO`
z>55(~Z=TUY00<jJeJXY;x+Kf4$-OWB3>p}7G`7Xtwxh)W*hiV-;=>~c(;aQ~Qbp`A
zZCVl|VnskTyWaLMUwc^J%DagRvHaW3evw&xXvoVy(t{!-Y&B`ko(mD==m)-8NWCl@
z@89i{R0bi0#_JYu85}1m8O>SO?iUR{cwXt|FvJ>NRSd=?xM#VoC$}~1OC)V)Y@@Kd
z`x6adT~mR<xW$o~fXATu|61&{uszwsX8(e5f4bbxf*{}<1u^(pD~!S3yo4*L>u9X?
z@s3G)6*dHRT|{%Vi$gjUZR)~f)+jvveR{~u2(_}Ok|CE5kB%(ZDe~_r>OK7u%yLFY
zEFTYn^)gq@u?7^JYa1uLPhUJESNOEcimG9@{j?DT3OP_e?nS;OT-f>-t9`()9H=Lw
zKFytqFB9fqmQ>f;j0U$7HT=hs#oO&31~HFPfP5)*+KK-L5dc7S_OKy8bZ^1VByL9e
zIN&^K76vlaV6q|ay&dXdvUw#X?1_~MP`p_dFzgE$60Hf`4Z^6b+{VR=$kQB!QuFT-
zLZHO8Q$NjGVoz+Zi8%7>oU9xn5F_6r9?OIRiLUD=nK4o>QTfOWq7C@NN-*eiI-}=U
z4KMIeSnuF6F05i60Q!;>@pnpaTA1zptfm{gkIj^uR;3OaO;z6IR6sM8*!c<OmYvp{
zEYjjk1mC`_e&6i@aQspPQS_Hf0xW}>ocx)MXbS?!m;bzpjDZ;Rr3hP>eIAG@V<K))
zGV&}NpmyBV0{89zzB`_`;&@jh<T&+cLC;8L`b^09b8AznA~m5#<jD<v%-?%}ugdy%
z7!YqHV&s{eDUasv3&qr8lc414i%e=_0xCu{3LXFB93s;yaUfWcK4^+?s5R@^@kw<l
z2>67|Ko5yvwkU9bYF(+x!B>8{+5_lIl8@XuMm_{?SVJc0OeMC=C^}R49&VuNQ0BfS
zfNW`kzPhh9VTvRVbrr!ug$tE-wt@$ra-GgWUt|m)%JibXF_eM@K}j9b@6Wv3V6>lK
z+ye}NL03Sb0P&%x&-Vv2u$%Iq1++=f0>{_KK|eT&e2HGFNJ#v7^;YElk3$$!gBedR
zEV-98PN!tnQ+K7(BeuPctYsWH+cYt6a6f2K1f)evMzq?id`-gp1z@{h&4T%7z}br4
zn$4v%Q7<dt*3~jK62UORD(MjVhF*SchJeljA%DD5X$tDLMHbmfnoVd;H^nGK?pBrG
z(~`d=S@cRMJ|c(oC-4--V&^GBH}^E}T9#-5b#`SWJGKuEMSt@GN-&9K8<EFpOJgD|
zf>neSoUB!Jpx6j~sqvGlk+Ps1VZ3}<NI0@h+vTw;;!c!_`ruX=5+w=mA899(+8CLa
z!;q&Yph=j)zlYmhW?3N}4LLkv7L2V!y96*D`-%_;RzBQD<x+;gSYdN|#ucQ$nu8nT
zPX0n`z9(h#Yc$nd1POJNN#%`5d5scZ-P2Zfv}uy~;$>7RpDlOdPn`y^0Z(a5SU5)m
z=aM=jX+b=b10ZDOT+w$-YBJMUBq#y+SAu<Hm-E~1<=^>p@YAA)pyU(|a>t!4g7Db)
zif<gx#X}v$h@tGYZPO`St;4uvMo1rk)bIOjSG&Ea3NlV!9|?o`k)Pu=m6)kcJZkVk
zC!j2kVT;~N|1-+&FK<?E`2qN(*QcH2`-Y?-U6RTc69O{_dGA-T&ga)=O40O=aIGhh
z%)rNhnMKnfbP|kS9dd1&J<9`jhC*%u?e!t5`g98I`<8g1-+UU?re=9{nfM!hc{kAA
zSg2?Y9YIi4$@$6#tUhR@bna5;%~XY+$uSYU37BpRv24nTnNor=hcK2Gnn6mMi?pWJ
z#z-NlAT|N<cN|%h2ov{ULm4I;^N~S@=z^^_?y}Z9fJeUO3HP_%pBpr!mGE)}7Uh>r
z(Zb&(%t>T8p?_$*fq~$eQfJc0M+?-a2nn(-tUk2Tg1&0#%HG+q^a$o@=`}BIy1MND
zfpJvc^0Ugymbv8_WDP~3*}Cl3WMZ-`W0>rad81f9MTm^<A#o}YB$jDbAVV!3=hb--
zN+*T;cKZxVMj9mYm{eeUmo{3xjn!}<2Yz^AO~!VO!Xt>DFMu-J*}E=#7|7<AOF!}U
z_xvKma~Ok~Tg!J8Wy|kLs(W&)cKdt*L}_S<oLm6l{~H4Qr<b!}y&L;i5XNQs+IZ9s
zBpY<vRUR%S^f&|*8vy>J*@Ckc2QC=@rZwm)3C~|YYx1jVKX6Se?au^b&_LP+83!>?
zY#vA8&!aapsXuqhf<5Iv)|a>eS91K2?@aPb>4C9AJCQ;W2{^av52qSF(K%PQTmyRg
zRE3T*oV_vy6Y;wM{Cs-73Y?8WdzE+CKHvc3PO(Speo^uFUE{ow5+N{pnu-qW>)cCt
z1KAADh@vVtE&c~dy{*vM6k-t(q}!x>Y+7FL_oNmp`m}NLqjMNm%ZHF&#_{*}G}o<e
zg(Q=P>kBo1qc>lGjzl|Z+f)d_t@?KpxAbJ0GOPQW#Du_(oZ&Ya{aZkPbhT_Jabct3
z$MJwsoqNg1Ae<Nev<$91I#};e%YDL$^wa;{<Y=Ox67aNR<s;WR`!pmAq8Ttc8-Vjt
zN<Tiow*PZW^Bxm;Tq-?_H*Doc`CmCsgV-%0-NL5vYx{n!TwdwxKk#vlD0Qr$WHK`F
z!+R~bVa->Lp2nnVuF=4R_{fIe9?*l2h!vw5wp0Vq%X++N)Dssp_X;SPXY6@-6GCAD
zF0Z%rm7=3}7rf*Pnk1C-P?1;ZanLNxr`)!RJW@<3=71v9=(9$-@OV$c8p#NDLe!zl
zTa}DRNPZ5bp)^E|>D|7vo?Jm%!5GvJO+FXFLr|5&{<s4J&&@%owCmmfM8l_N7FZaN
zksR<%`}G+BuR|C~yzrOjI=y_-Y{N>bDK3nlG3LjhSKdFt#1r7Ub9u~I@6HnE*kK)y
zNsHf2Xa=PjnuBkI?UM57n2q?KdoF<~<rc%)Ahq-fbh?)!C%NSJQ9-pHL1l;6vo%n0
zX#)1r5;%Z`ueYa}Bx5W~;`#c}+fH?|rgaMP)_2?lZ~J7A<qO0)kZl#%a_3jU$1*P3
zvnMaU_eP^}Iz(M8=s^M6T=Mu|FhOQKo##-@a#I>HWH_H%q**#)d@o*tG>O8s>!DeV
z9`i7HmzlhAzHj)_S^e0y_EOHQ=mn6;pEUPY!0Ae2HP~aJ*|(;R1<bkVYg<8lf#7m-
z(wH19vQGy5eCL-at-~WEGu2BX#rc?~!vqQ+b!)l&c)R{J-wL#)4kCprK2+E|MO#~n
zhD~g2v5>frir}PhD<}9#^LRLFEZ1nvxq%b<&o-&gRhW|sim7y<eB3p{$PgrV!p^5;
zgwmd@n|{mEWB-e5slq|@mhQD6T_aN<11By6Yy#z9-k#om3=W3<N?W?Qk$thU%pC;#
zRW!%h%;UBgh1G$=X=|A2^#l!An_ri0uU{d&hb9ULhDj8mrxFHI9FX3mXz!nl9g4N2
z|FikQ6>b1q-v)fR)2o6!C(Vr@<s=@bQ}@|%5nfKTi`eM_M(}f@SBve)@(!MRisvR<
zf5t|0q?&0f3BLcW2w_Lyso%g`?FFs$`xUPFB_Z$BS#PzO4COtQY(6YEWt828a8gTL
z^|gk9;p^&=*$m`9V>rae0`haUg(i_b=$O2Hh3W+K;cP)wAoK*#v2MA(90cs96?h^)
z&axkq(zpiICJe{*9sm1w@1T5~E#suwup(daE^r+g;mXKPzs#>bKwk9<aCrz1I+#J;
z!`=+FLgLb%QX^y%{02UvW9N)EsZ0G8p*a!Z7GLKQ@dlPTiR|F|A1x^hFa`<|`=I$s
z<7bUUWe0dhUp+2%pPfYQ>HXGDV<G7>nHjg98%w$WSdG?C*_xT#ZC9#zc9h3YmVO$u
zyce!wnT+&rvIXhWDf}*NxRYs`pB~x2=6HJQ?$p$R^%3$rmyO}=NQg%MwVX$Tz7yP`
zC)avzNu<<Eg1Ls21GLcS+IH^CrZW*^S5G7%wwK!D{sSIriWd|-1%et{@;_d^cJPtG
zF|G?3`$rZPI~Q#yOFQyQ<+EuBZ^&076-*z_R$4RWDfodAP4JL$CciHs;JCA!kSNY7
z(zOq$RN4pf<8NT$Aw7s5U!f2BE50^)wdMRAS$b)}isN)nc|e>PnWO&2|6+tsM<%Z0
zF;S?8OA_ojqJ#W``+WF^I-RD^C_@ze_#_BC#A&M+sBvLu)_7BW*1Po(E3^scbIP1$
z3p|KNzo6-VvsCWSE9Krp-(4n+B4ow=Z2xL$XmU&1K0-lTT_o?thr9f4&GU24VfsWN
zns(bhl30x00NslC2M8smjIt|8{C<2^4uq%AuVUzb!o_@;mdTsTq^4+#EkV7w$ALvN
zW0YZ`iNi5Qr`V!whVQp$?Hm7lre->BICQSM%A9n+`lG*?XNd8)oz_St-GDNZp3cft
z+zl*e780+YGeXD;!1&M}jLTHEr-KU!2ymq=DxC56%R{9|5xo_JNF)=KXZsLI!Cyo{
zcRx6hAl?JDhzt4PonVeXl#KGr(-Y7ucqlDTJ~1_$aY1SAU!2jcgl`EFRm!_i%Ck`@
z*G31JWmx^d0$ZYU!2Qj?eTJ{1&EcQcpieL;<O8C5g;}kd|JL@u=cU=MLzZ^l%JaC>
zMnDz>S>K^>#|8g5jPdEo%l?mRMrIY%MRfR6=!Gm<>F0J0QX%$dxo2(s%ajTmZ*)a^
zcyTp2-=kteJ<pB#v}p~;PQE5Z#9#UggvpY1c;a_0Qk;P9J-TD<UuTksSk5Xi4+5sK
zFnJ1;Tv9;3{rBxSG<U6~hK}J6M9~sLX-D_=?bG8RsGf&QOqTy+pEQmx=;Ec-Y}(*k
zeMEHl_7#hG61lIbJP{Ul;Qj|x`hm=B;^0BWG8kOz-rX!kL&^Y_)wX~t(#IIPT=;bM
zq^Yx!5FMNo(g)48mlkA43~JY(9W0I>_c&c9jx-`;+3qz!`tskOBuyk^0_L35p}((+
zb@5AHv-|OHH#&AvswH0intw)G7mBoU%LF3G!(Nv9mHVIn0@Whrr#0Q3Y?ir06XvEf
zt#8;>hrUYg184l^@LnbO@E-A5{8Aoy{1N{7pL)u7o(GbS0u|}+a0x<n=p)6|!o4LL
zP2$@2nDz9r;7k|qx=%AQjOkg|qaQX63?d!+nczgCAP~sh33n<Yd;cf?J<y%nW6c{N
zz{>B=C=5ND%|NwNI-bpqSwB#YsBHGzv1(4}trj~!mfL{C^*A|U4!IbO0M|;_*wPS`
zX5cVGg19Qw2h5VyeE8(0Rw92`T7-2BJOLb<nrFw_OH&#GnnLK*OI9KUpxt9%B^*wT
z)%r<@Z>Q;m7$TNR5qcOV?y~IbRu4OEn9r%kAX9F%<9^K!!;(vE8iZ$sEa;y5pE+?k
zA-|n5Go0jOgs%O1B1*k@$%T_nFBq3I!iy$1{+)VoRtO>RBLypaRE1tNoWmF593=i#
zUyA=rcv#>sfl@&>&4;D~oES}e?T5>rgr7ft0VX;O9?%Vot|Fx3AC)2TsRuCc(yW7T
z7z*e8T^^O6fRS44OcH`WDbZk~*8eUiQ-%RGpLXBp7yD%Yb2E}O6j^$qvra!UY!;lQ
zK;)VZ6p0co9y;wBF0%7ZHsRx!EwC9i{0>$i3D*ZtHp~d}co%8@9_({ctBifM2|Vsp
z<mL{OrAhn6noYK1=IPRh20v{8zUdNA70Du4>+#ChZg-o6F_n}Mq9;*Un{+l43K-gC
zkFLS;@}P0`sj-LVY%7t&IZRwbARb_2kJK+&iC+W@exp>!hB!3oCs4%rn(m5znGu3m
zqIW+0$nO{hA4(EZEJu3krKQY`BY-n@9m!@yEZY9!x5Y8Gks4^j87^|hh}UpMI~7is
z5R>c#5nZtPlI+AhFWzSL`zzovIyTvq1oDt0Xb_3gZmQ(!mbX<+f-{7I+Lo+_@(j{?
z!oE*hwEPGTwK48bOeBP?Mo>P6_%RnR5#}s>Y2P2tGd5vB<cfs-xPcRNYV{R$A?{Zn
zZ!NLo;o#e$r=W7jT#s5!5qcf)(-ovF;GCl3jDNp>@?Q8!K>||Jq5C01aVK8{T!g(>
zH*v2<H$YUoeBhNBERDlrR!ojM?F;21I`y96U!fwK0<@sF9ji~4PSyk>4=+CS^9-+4
zwxrG^(q8e!xIbue`9&hHQXa|NL4mFJAwph|>S|W{Ef_GKB&_A@m`|Lv2dZHGCl^b(
z-5}ouxWJ>fi`?2eJMh=fot4o`1+hQpoDsclt`u#_yq^CaDC7!2CX4P4d#Zj%kbJk2
zxGIp*mpc4mv(n2?!%o$Mk}dG~B{J-{r?g&IgmU#0#Dn4^BY*|(T)z*^hu;^q=EnvS
zB=Dn!g_lEV5Q(dHTK-8tWSV*hU5D;@)!;4nI}4*9c8UkzfbPrl)0~xoOoq;p30v_7
z$~ClAk`sSljlA=Tq>}A?mbh<b$YTG&JsMovnJpyeYxl~NugElkiOqY!nosVq8Y~>5
zh;AU*i%J^cj(dCTZy*Z+R;gZX?eB7Q?DC8!B*5GjV9f#&e6Q@E1gJcXJo3Bbz1&A8
zzYa8^=iXho|Ah6V_ri5I6TtqtVwxHjD+Dq+^b^X5;$hIZuY$!>xnJTw!I~@*;MOXE
zd~-p}M3z9n?*mM#4xKU>6<_$@wIuk*J;lJ;m#3yeya=N36+;S+;Irb@f(VmB-(|p6
zETfu~jI2orrCA7)030qUu#%+PV>i`y9N)8N)lTSzJ#k~8`%3^gM|FPJ{Xp~JI-UzR
zas!3_7(u_Oz}w|!)h|fuys)58nl=}V${Tx@8fyG^Z!`{*oX7VYzF-KI6C+d_U=y-{
zCiOjcF>vx^-Ur2w&Rx&8;18%bf9`+qBF`mfR*20pMt#Qimpi)h!dL&h?Ds9=<vpX9
zZ`b3%CTpf?cS0Hhh3{qd6M3#<`J#-2J(9_zNj_2bd^i|jHXh}s;CWH;(#pU!+L&Yc
z{`soyc0Oa(*l$z$dCgN4W~3iHO|#z^igN3SpP)mHMFDCoT)~&{eLO!K*W9EaV=ez-
z_m=wK6DH7&OES<90wX)hf&5H=SKZW8e-&8&b1wS~+n_L8i5JY(k$Am*LlvkT<5#0I
zdEi6VOmbL3`za-*Lsz-q;lcfcY6z=qC!W+LC&gNz>FyF0LDGY?f)H-8#Hv{C-HtsA
zBa@U{Wa$N36Hebcj|i^Dmv+^h#!bTiC;I3UbBC6Z!7-J`Aw-x+gTxrU$iFi`|LG*z
zmt1|&eW}1JEp*>akNpv>DcJynD5`%?ye~S61&!`iN$B1T*)B6B3M1Yh>!qkt^EDO!
z3(_o8#x`!90}}9QUa9Cy@_NMnItX2Ccxaywu?RWJ*vA_F@zr3|4M0WWNXI22#$Qv`
za-N?R5M+y~6JTYT%WL*z-GgF&(XF~0q?rSYZy`bo<Dm2}3O1v!!GzlmWEtuuaNk#5
zdfVRe-}j?Y_1i*Z!t%jiFgUD#N+K4^lE>x)xiz=kt^f0z$|S8fiQz_u*WCTaQMP1z
z<7MK^AQJe_=;;3Qr&Rw79MtV4RUuhGXD7>VaAFP0Xmc1zEPXrfE%xwxXvr7G<&t#t
z``VI`=ck>=O=5LHh8Muz<4%!5BPT7GS57hnOOF=l(1ScqcM(YsIv?J?^xi13c?jL6
zWReiOKTfPL&cC+K$X=DSM}T~cO8Wy&`kkkIp*R65JX@{&V(_0Y$xtRCJ2>_%@jViJ
zNi(_UFMS)5?^79~f8~<2^Gv_LKA_bl0X@?0pl-wm4BWoBfFN9w=6RqE^^#5ex3U^H
z#sIMcjjj3n{xSuI4Yl9905VQh_GeH~JPHJ7`^;Fq^M%Cc;tq9;wis=vk#)_B3Pn@M
z*K6mBslQJO#sXBLM-Dlc3v5N1pyIPJl}%+&qf|+J&9RoEVd(j%e(KwSbKG0WRYGY?
zBOU*}dK@5|jx3<X?T|41uOguPup?BYn=I&8H>cGCYLNAH@oMgexJzKW-;`LR#JZmH
z_Aoiff&qW^b9aSK7J(7md2ahhhl%+3c`^jMgjK+=iOX2~@a#&i#YvD0aTV~8T{7UI
zFyxGx-V0ep9Xip%F?9tt&Q^Qz?D5i%d#KlLV?5?I(IL7JGbpA$?$`GaD_^+o1u0_>
zF2uE<a<F21%Qi+3nBk}^AR%>VP1oH2UYQ3{)S4T+_3K*m#e^a#+RCQK{$#<-09bc<
zrj`TICt4-!*pmM*kRA?(qH~bD1(hcJj$3#NW&`igw_a;d;!>M2{d|ooEyENQ^;pxN
zSt18x`re^aGZn9n@qxle5rP~7h!pu9XEu1gmcq*G|8B{aDh!`F%odiKHDwF}Mc!}`
z-ctCt4a~{p5D<OSHfU^t+Wn!0n8DP)&49xjA+*z-o0Wr(`#6Z+eu__*Ji|db<TAaB
z_{7P`!BZ}N%Htd0b8451uUPn>LxL5emH9kDcXEz8gbIKLbKVEU3_6gaV}(k<bc}%9
z%{e3F_^6Zm6}@a7lCY9V-BXLDm5}&3EFL=f@BToxEWUj{Ad}B?Q6M2Fve1y6I_4I_
zZ%kNzTtzmCoNtCPNg)t!vvU~rnbw#lu?yn7{-{SX8EfI7JDYAKcHN)|k)4ES6-Q~p
zIJnvdctVKP1Q{)&z9L4lN_fS`D~Ua!v}i5VSu%cx<Xn!Q`I6R=hP#IkWBz(4_cRGi
zB@Q4lAmJfeAKyQnsF57|Qb;i{W3T)n6667>b<F=d-EeOtGeN(CFVZ>JUFsO*;t<+E
ze(ttbfc$$2seg!Ezfbf6+6~z;sKt0p>PF^Et%HtO1^^R(6tgbTM82)l>YPL_1^r%r
zmM?PAZ#~Vs4cCt+9i_rGL(kAf7Bsum<?6%Rv-Ez>OO<r;Ce?+nSmdOrNP`zO-RGxe
z+Z9lj2PK}2J@I#KocG_7KE4Y*uoO#?-j2hgDmR{}L9Jtfyf5#y>i+T9jlKj1HBbpY
zue%9EIivt-+<Wg!g-YR%F-<aJ%=|2Qu-*1WV}^nN_TdfIqfxWvjNvbOmg;Rt5N^v6
zx?1trhEB^vGL>=r{cP3y;C#jYX<wC?lUM%Hvqp@(e+^W~SRs&r8A@f$@=*X>Okqtz
zMfX0SeHEU-gg~;&9`?O^iX{#**UiuEm7=d_quA=hre!pyj%Lea+M*t@+t!%-+){M;
zUCJEspntik6>f|iC3)J$C>1*(paJGRRc15&6j|J$2?3_*jmOIhPiWN)YXkV7xOnSB
z?igt7Y=T?!r^Wg^5|w9{pxNT`{AwN=?4+2y?G*i2-hrP3X0BG@s{(m%MfC3whxt%_
z=`Lo?YKCUetUSWMgD{_te^8sCPOOi1Bk){Oo#f14nJO(D-wRX7#J1cjP`M*&PAxYB
z0dgbc4{{@|<l)IVL;x0uTRYzdqSpfiLybz>Q0tVKH)BZ>o#wdCB$$EjHLKeZG8Jo2
zOH0kdmTcUQr&P(zFJqzsFjW?^sgUwudr0&_G7r5ur^t`cbsR+xh;L{UQdmp~FYp7w
z*8*j(=3D`yzKY!t7(ChcCULu!^rdW!7BRE06&nEOv=v%tSNIkKdCz35<0%f88w2Ly
zhj;Owc=&mjAosNR@Is{(CT<Ok{T3&!JV$SUJgq)qpyZ7*D_Zq?ouAV_WNiX77tPfQ
z7>2r!YxN?m|Na-u;31F|tv+#Q;($A_(WGw)ZCn%+I~FG9;y1sp8$WEWJeU7q>kQZ^
zWc(3Q-nWr00T>pOF`EAyQML0m>D!iTdN{vMx4QiWv!y&va5Hv7k|$V6*>-c!rpt3H
zjWl!{T2EuH<np+%leU;E+~|s*N+q~R;#-7782nN}Gh?M7ixI~g@woDHDP(hYbPlsW
zA~ipZZrV)igE!O%@zL&Jr&%uj=6}DTe0MlIN~`9|mWH^_DVI)@%p;1{7rY86Z>|DL
zNHhwmO@(hl(Z=|)AU^2{zj&6$vzxBl{8hDY!vu-QTkJ~jjpDz)AHb5Kr(+#jb|-BW
zcpMp}4?pkSVnas!Od!goKI|SdZ0zXSD3x@74avb6rfdw1n*~o^RX@w88kMr)+=Aw7
z=E{ZdA>9_awDk3utO%o@g8%RXvk0qu^%D44w%m)B%rBYgIW|lu!Jbe_ZV2|0hp-U!
zB?85wK+!h=&gO5F@D51|(TGWT0Skf>x%apSL_>goMacGYLw_18r0lr^ROHa7w=P$u
zj%WQasbG4ZQk5_8pAby4UH~Vd3^9iFhfEkm2g+UY^7E5D24h~YU9@I%DR5IW_J_WZ
z)X1aXlI0eU<@|W3+2`Y+`HCA&LQvH4bDInBF)c}kUijI0{k2(9roh_OL~OYLb-d-F
zd+6kP%zR$^s*7YVQ2y@vfID-(E%YG|M_Y>ndT<H}%YZrmHOfj-j`udd1q_SirPm@l
zm=O2qHDp?kg2Uc4J`&|@W&G>B=6LJWB?7v{Z=4Zo6F8u<zMc!oW0v3zmNvwb!_{F{
zGD9s^Pnw!~7!iG+w-<B~GBn5FDjX9Vcq>c~*9Rrs8Txqb>e+Y-m&~(C6jzJLT)mrw
ze)?L%lJ9R>!B75OzhvN_WKf6GWOY2M3jDvhHFhoE)6_Nw9SUU1^X|aDfcvUc{{5!0
z#$mDAB)mik;H~Y`wd6kM|5&1Ic)GXw(m-G>Zy&v}_pKhJ(Ydv!jzaj;m1`c?^Ua@j
zgk&7T<O<^^fn=RjIQa{DND-E;3<%)x^?5qR)oAp0eQX7_%W4e5L8o0R0je3qkS(40
zoAs9pr0bEq0^qEJQm0<ba46!QIt##3q?AI{NYhSaQ5%1B21xx_N5ikRV8_WOdCo+^
zRf~758RRA$ob@)?rHQr33SMikCPoLFB@*8XFjjH6{uoyXG2%LqUEwqIW|L3BKhxva
zu1(UMvQy0)PJ+bvVjxU*-k;lx3Y48hG>GO+PL6)OaoN1)+0RLk@gN6xZsX}q+=Sz=
zE|Rn&%mQh;u74k-adMZNDptp8G;X;I9Lt|iy3qz>)W^z3Nzqp&Rd|RXGLhk%`DGl}
z!9=G8@a7*MRZ$-H88ZIq&TzxfAH)9r-m&uIfEKiGO$^RlbpF+Me2FyPSP9JGcDnYM
zFwkFGc7e<xN=CdB?^NVlT6f;}zjeTHXYj`48AUysW2y~0vRP=d5ukapc3N@R`Yt-R
zXptiJHKx%GoBF%S(>6-Ay1O-cxqIBQO6)$gdPKyq$3Q<5*x`Id7@otOu5&WJnwe;_
zlEXb*A-vypB;Nc}QNuk*o{Bh>h6kL8$`%G=>4huNse1W_ECj#{D~zta;rb$XT^jFL
z^+Yvf+}U!xRzuNah1V?n4f)MDlEK28#`=WpJA5Kw8CAc2y_;K(FwLl7y@0n8-DCjm
zvMd6KiknwK_<z{BcHN7X5`ei5q(IiJw410yu0gvL=qb|((fX6UpN8g;ReYFHa;iB(
z#GVxR@S*1~8n5GQ1n9vwF2>v`GpHKT4WCB(-wW2OFI=~?TGFu^)PS;KZNo)#P7>Eq
zEOcS<PMz6bCx0vstG!VyHd&mQ+ZS1U>7r-*cHyHzt_yiTeb}X(hA;SyIJ||HxoZn6
z00xcdZbBP9e+(7Ipv&?YOuRqx_R{bK?>&BXf{!l+L8y?BMZfEO9(7>}a*yOeCFIBM
z%2UMwp$Mn-yvXZ@JIbA0mu<JfU-icqjJCHMK(*)lQ`PE?#k|g3|3)>mX}E4-SuX#{
zt$S-;_+a~7k`{58c*j!tAydWz<S7_rhtMTVADHu~?~&x7H1WUUWF$p$NFNQwZ=YV0
z4F%js3g@$p6ITB!s-ySB<NrI!&7{jW-UP_!wbHI5vZQ{;=@FE6$YEt|(?Uv*rCHKr
zxKm<IdqB?v!}+m?v9tCk+a}sE;?_%Nz&JfJ#ktDj|Bt;ljmx=fzlW8g0VO441EoTR
zCXq@)bCCvVGBi^vN~PgU=13zAR5YM8NAqCTq*0|L5{(MY<A3e5a9!8$_xwN4%jd=O
z#^<`->inMjyZ5n=V;yU)gSG#hyuI8hZ&`8{jhQTkd3IdNKo)jD(*Rj25~nA~U6DVW
z#W0>gndZT({Jn|{z!zUZMqZ{|x164zUAROog*O^M&fudZ?>vpL9`LEC*86=nBCgo+
zi_fsSF!;kxI217J^5`|6B#d5VLU?g|X(1()!IYV4n31TB9e`2kQNJC!-HSgsHJu=X
zdc5P(GQW2LjJv?XZ5!EvCdrA~`~FbZsL+CcP!5qqxP*;vjFSKJ`Rf;U;!NR<kU?Mq
zNueVjh<DT-85dl!a$qwDM0%?<ufx47-$9?2InV$7VJ6;z$fffx$!pQ!hy`s3PWBKn
z%<4yuHjU05;=We4gV~O!bujHg42t~_2`d5h!gvJY2Yu7Q5WySeKK}f4@TC6xg$_K_
zNJ&|fBgU4wnaC<Hr4_+_@>I9UM;_#cp+!42hi&c@J%}2;kgqPi3UoO18+X8<d+1#)
zI_R!X83h{pU|#hE0VQ*o-9T5*4-93f1xjjVdNXP;S<9d!-%dNyTeTIQ!HEW$UEr2l
zze{v+DNcY3e9|@pX%$8#0gZ^QR@E^elyw^!d1u6*VBhJ0f@TVSc-G%A4{-$;Hr=PF
z*JCOz_@;Go+6PB-lH`_O<rGcAaql@uMSc;KIUW3p&pG5EpX+{72mjUkqjU8lAv__&
zZDAXiJH(j-PgJ|>5C+3QKFNeQ@I)Hq!WnxQxv0X*=ow&egK7VAYye9}%OmodLEX+*
zZCR%=Rt33<^R+^78eLlq9p~kg_T(KnI2^{gQGXL$-k`ueXd+e43nPE)j|G$9%#?BI
z6=z=Yk6X3>ww%^ENaP-WV*&VD$+%?QCePNp8di?T9&_ThDN5B6k0A-XIX{U`&$rvU
z6Y=PzqrKDtg$)Rc36ZM^joiLk^~k4FJxJB<zp`Xt)VWDq?I8pz`D640$?m#eS?KON
z)fyb*Cwzzz6`bdSoK+qjWx!ts1HTdc4=vA#(Yfx@cGWSLpT2^hVPV2WroA0C@qV9E
z0{b3$bZ6RrF!)r6_j~;-%YeG^+aExDcbN><T`CA2?)z8>=4omT$80D9KYGB?n>dBD
z9;9*C$((*sze_M&%1$FOIyoE;-5Gu$Ej>qeym<IN{O@X#Trln9#iJ0@cs4Hf4Z3CG
zcZ-@ZUr}VIdyC4RD2umuE{KPEr<qum@P>vA>+W@A`>MEl+o^^_5<?^a?2(st%Q$sz
zG=A+bZv*Y{ICmxz4K;BDz6Ub<C)*0D9D4GzJTe=DYzSI|mvnmU-uc6@rZN+l8!@Gq
zV9H;kxntbyU3eYKeT=4(8!{;(CC0`L4@dIQMB}enM|H~(BmsX)V;^3>AX6D`axN#l
zQ6{4$+M3UOx#gz7X4*Wz>%mgtO}$sH)#Zi?C=e6zy0rK96|sHAW0J=`pGmq$-3TB9
zQvwo`N5g>w3$>YmRUIJ1_c+Xha}r)L=F^CMmL;UJ!EI(KoCzqer}u;SP=Z8e=t~Nf
z@$`jUdeZHKE^P)|4phxs$1cD4(-pRNZL&-Zk~`|}$%M_RjGG!pjvzHYtcwo36uJuW
zZ|3O`Io6`8r|C`$oMB$Y?%ld7Nn$RCQ{Iy}Dx;XEm2lhz5asV~i0X?<3KwDuR==>T
zxLzgZcztcPxQd2p-#(YAiBa?372kwo51*`SY>HdQ_dV}d#Y(%ps?nT8YXkXCZ~^OF
z$K1C2Oe!;VNZ&i=;npD(4;lCYRVAy~wSmjlzR}Kln2#HH=*}0pc|-e<cp$@T#AoET
z-&N0br=|!Gk@9KkY*ZTrihJ-wvsB?Na1~;5N0S6{roMSj#Y3u6aewf39-M$L(~GaP
zuhY?!e$c}{KT*dQ?XnvUyMpL<j&ejyvdE4^-Pw5k_rxM%CK4TmE#tqW(tVovsBk$2
zXfte6+|brE|6yX;o1`6L78M!8f{PUbXy^bn>_p#laP|AMH>utrv<9V}4GZ^cXv+WE
zQ}Nt4J<a;~T37u}uy;+ePJSZe_#BR(KQJsPkS?vreEE5ttz;jy(Xxt0xRne(IcBu$
z#KZPPDb;hN#&}P@W4K7i{Pkr|(2=g_TYa7ZgMP>0^K^2dN@sV^8@u-VV!sw9g;jio
z(nKVyl~#K%^HMlZ@Y|}Z?_PDfc3;IIm_Q`_Y<YpXP~^U)*&DT$mHh5w^oH5jGu5<y
zz8z^&4qh$);qw#zVC_$-uej2O3AJv>{P@9(9mc<~N0KkTEB``y*tt9uy!vA!JZF14
z*W|v+{tjxF0bH9&Y}UCjkiN4}Y7vO2FdR~QGLpH*dZ~TT#0emqam8T9)dx|DVs1-d
zz%!Yf@x-~NC0VjAE#$RoE<LsNlC}YfPLJ!nSc&#+wjT@f(*(D#=GuJ3$6@s=Cu7zA
zM!&N{l%1bH+JpsG9+lVF&S`XNAJBm#P2z3@b|`W6NkEg8uyR6f-Ddmk1wc;uA67Sf
zn(IYF&=uM4k`urSWc_?XSi$@a?0O&VGc_vkTQcb4Z8e)tXIOxaaiuKt?Opn-2V5Z1
zcqjvb-ww;8+;{cq+HY=e^A+p49VEtnl8x80N@tQm`fh9R;d7T+>V0m!F^>*Yd2l;R
zmRjkurEuqziCi);M-CH18W}V7dp-=B1ddlAPy-|1KB>1Y2P{wi@A|-u>Oq@YvW6Jm
zZ;*0m4tqcP4xCR6ef;`?@SYr^&*PTkG`L8jmwXWL=u1EAEP5Ce4dhiDI}R@i=ZWR$
z3jS6kt7}%iiZD|$Ycd;&amStf6J7qxV-HmhcjoJudb-k<B*dL@S=|IwX23vcP|%Iz
z9?pNDs<e)L*IHI}3PxTQE!x|((mO7+yf*+=xG{Dg3)~W{hMGqC$~CB;Zj9J@hjVD+
z&H>o0Jc<UlqPX)y^*WpEXgbeuI1eTXo-jKA*zPpcdu*qSmn`^dRJT$5p(zlSxB_zt
zwa^&|D-jq56$_;(F$k(BeP7qKOWlg;6R6c`emV!4TRRfsH~1Fm2z-w#*o=`!3f*0`
zaDlkGy0dcx3chUQ<;0n07fm2TakvoOqvcy%ejcl_Xm=X-?D^mlrCv8Z{lm|NOHBLa
zmQr^AGQ}%fBu1A7l-vm5Q$EKO%NBV*h<*7LrLhAHLccbi_cdGaMn<!+{_T{<I}7)r
zY;^wmb_{1(^nDwB<_zlqu6IX2_QjvBa7!MPG@~v8D^0q*;|<C8u^bI*bAu(E%V&rg
z0flqs{`@lS`_$v&6{n`<ZNw|*P!8=k(lOBu-)eUh+yv-U#L9D=`)o^6IfBkK_Ea~S
z#;MI@&stur3Q>#_=<P7m)W?wg?!t)MwR`vKtPBaDvtMJA9)I{7^&*$=R~l>Y?^{bd
zENNE4o*tGh@*YY<2E!m#YL!X;*vDJsC9|GOdzsySn<v*-NmoZ~LajxH;WnzN72iQy
z?&zTm%f%DNRh~sDGRLXkuwso{NGfWy`h_N&UB<wg@K}1%{U{3F%K6s~7zsuHu`$rf
zAl9;^=m;75h;JS>>QYwKr|7<G{Yop>fg$!CZ+d1ZzOYA;AmaeOrC)Yqyq85@ZXle!
zq-z}FB!4i!pRctvIG3;Zh)r1V+rj=C+xHo7H1y(3_w*0-5P0jXh&)5_H_thzxM?Dp
zEI;T7TOM_mHBMM|ZMoP>S>1N@Z#%+llI(TTLYwI_st@oj;YFZ-apzVwrmcLL_jK~l
z66mRV(Tvsw6dKgmZP!QsDWRu{H1?*5QJMJ>CUoGQumEk$%wK+#7#}(1=WmKK(xjHf
zkQMQ+onpgU_|aZI=zMy5Y7)CV`cYi_c9S!o=yNYP{AAZ^eyyz8rQ#E{V5xzvV9&=@
zJ$*_z)E3u%ei_v(HliX7pg8I*HkJ3#X-NLOn^-XczPz=v(vZgD!gn|e4TQV}>poaa
zTJw$!MhSm_t8OB{yn1OXeeC9x`df7q=Rh-<C#UFU#7p{D&s`W6k}xjE%&O4;t0umW
zVMMP;!ph(>>-q!tS(z@z<vMj#enN*m&K{I`0kvDKA<MJAMz9Acnc`aBmd@2EjFBQg
zzbH>WPfoPKC0VCv?jHFD#&}qjfottx!>CPc**m7a97eI7uYs3gZ;=nAULJIKlU;W5
zZ71BV(sIcVk5y1LAzGGPKM9Gw+f7=q-wQpbfr>Q_ZVo9N`ONHQAA=C4V^j9Bp*qPB
ziky69hFKhoj)_%)K|C6x)bd1I$6nrbuHL2Q6UFUqQ66UIHE?tQ>Kt(m2?_Rqbgy5{
z1lmT*D*LNU3}B7@cU{p9oxh+Bn;I{%z!$BF*Mu6!^2pdvP?*DhLBRM}IdI)&>6Ndj
zRV&#ITuV&lzo*V$k+&Tc#>3lg@}j0zDMsjJJeqhg7l=U8@ELId{hGyq<vd%9Xw1|r
zo4tWyqkNSc2}O}%DP>8ju;oWbVX2jKB|ib*az-0Chvw#-Y*s@a)EHjF^cmkD-nLa;
zy@-FYx5pwPFGUh@Vr~p)oq*<@RVDkwzp^sv?UNUkk~ZNGKQW3iTJc=>zpk=WUtg|J
z`Q#FIX!)HTO)h52CPtXK*EMFm_~Uj@Bxq&4VeeRv_?E1qf)Laf(w{PCt^<o)b5W92
zomPJF(RxlnE%_R64!PF>{H1uv;8eA~kN2b6A7Qlm?)Y}xa74Z|pQ}e~YddVVrGNRB
zCuu1#pbmPh84fgG;XG5Qm=^U{Gq9U3oee_vG?U_qJ0E^{!^2DGo_{ds)Pok~X~4Z1
z6>-uk-S2Itc7kG}c1?hpBpNNH&?%hVKdkZN7CL2klM<aG8K%Xw0*k#x{P9xsj9p@p
zT?=VXNd+bKt#mDcAh>WvhiDAU7BTk0=;`;po?<x>j*0K~x<a1K=EQZKfQ8fzC?kYm
zHt?#8LX+K6yYKVn+|sTf$-#Hmjo+*%az0h%ui`hHXy@X1O@NIzC{b){dK2>n@Alcm
z?b*Qd0a>oO*kf<Dn8sZ3y^SrieF{N&2Hr{^P^|j1ylkp7ToWh{CbiS7GqehDnK3;w
zzE)NE_)^P$7Z{vm-rXmt>iB;Co}(@hM;b(mZ*dlBaO!=}{~>eWvA^R_G(|dW)O@?y
zGcg(j^CnyetRHYB4D`MTJ|q3SX(g;!twx(2-;RFLd@*oyrc@;5PdqM*t(Vp2vMhK#
zC>zpwT}W`U2r)my$N$0R^MF3Xvgy$RWnxnpK}B1>Z#_Zgg0c&fSiRRQA5s>i!q$I}
zl8gxJ$Tj{TlQ+-|B)r&0cv`5t%wFFpZrd~6X$*DC;JMuDy*45806V6%S5HkRj1=R{
za83O(u&A}({r(~PW^wzv>G^HmdR|qr-It&%+<wgQ;e|!3!@O9Z!UBhg`M)hd&ALw{
zYueb0jl_6rpf8e}D#i;cq{TN;mnj<920>7mC+pD(`4d4(Wf_LB`&l=#UfUaAtEfH2
zx<2W|5F|2sD<9cd_dSAPUcNFl?(+P-|IBEF&DvbMW3VwdR~XeXZ^U`j-7o5nDH3S)
zs!#fUHw(Kp7I*Vp#(PaQx4X$2j(1T)wq#go*ZVHf(*&c+#+of;(aunovMb70G=w%D
z*7ygr+>6E~Z))CMcliS3)`ZMfXZA=43GHdU*v12CX4<+tt=Nt%;1#4a;=bw$iS}U)
z;^;#s=&PVQr>^e2-%W}=G#>Sr{Liak3nhAOIhD=E3+1DUOO8y_;cY|QH2f!OzR4X*
zo%)D2cc?oWmivD`L3j=i+sYgkccSHCE^I4SNZGFX#w-3O%AVyf@@m%eJ0#bbJC-1o
zk+CT;p}GbB$x-<aV$FKhwI#D(Md0?!-XrE!-qOHC$g0j(*V8=XZZ)lrCoO{Ks;z}q
zd~K(fT_aGGAjc-TzOZr4A{B|R`b;fL8giYJJA;h<;j0tg=$n2LJyQO|PGNYZR&Xb^
z=TaboZ0e~)1AM`2_rd3x3;YubiN@s4`W~ps(pfH_kWV^R>LjMVjsbCQgS!H@XiuOP
zK{{D4?Lz|NyYT>`p#{C+M}0K)PKX+|HQLss-k^VM_tEyr4yT@5`OEVrzz_(t+lp~f
zx)`?N{ltP0TBv&oc02RF6u5Sfkn+K#UC#|wa&K7CPh_dRD(QRf8$C$tE_>4Anvk%w
z^($=LdIC3mISI&XzT`dXWKMLS6!wU)qVrHxN=xZZ3fNv+`gZ%2`*w^v#%<f%7WA#L
zWx@V;%<9C7u}1y@Nnj7h%YbXGUZ6{Rfe6KoZPw6dFC|(WYW%?}n#yy<vWOHv=^cwU
z34KG{!dfjb^D_7Y5AN=vIPHRypS)e1Up=_I*~GiM2s>1WG5rDDhm!WI-<r*bQW;ql
zE~#H|#b$}}N^8qO2~pco>`@)-!-P@BuTNVRoE<fk*lfFgsGqnwj4IqU(|_f=yiI%@
zT2CBi*?3u8e_@%c+dg4^smG3LMJ@k5aDMM8ZM#XEh~v)E6$cJ8Z9%jxljVC!lO4Ax
zQqHqQTot(^ikWir<LB)PMqjqnJ<t|#3nVfTFZtuNKCDLqqq#tn5Nr(-f5a8QP<+m>
zv8x(@o9tM}+DCB@^{e0u9#ebbUB$91N%T7D$KC{BK=-oulXk%c-vmASaTjl^BSr%s
z9l>+7G(Pnn0OmB;tKYMScv`np-{}_|0f)gSh?_|(>f_9XkwH99!gs%)-x*}Is~e!B
z8KP@SsL{Q{qkv*8ti@~X>K+FFmai#dZ?Ud&f+bx&9rhN+CoO49)K5O$lsq{K0CSMz
zg|c~+bScz_q|12j=>5<SJkJ?wAZrIk)VrzyHW#Rrx0$)8rV7~JT(46iD@}TxGrRpF
zb+=a<6?@^S4&2w`%__o|FkE<~>`YHsvleoN`hApe+?VeX8MHsbG;r&^MRA*{`SLfc
z;MuA9^23Y!{pWNiz3e{R_aV1l;<I1<r)Y8AFv=1E4fgd11Q8OWGyLqrafq5UF7$*w
zUUf!tve|C>bMibgw%k-q@7mXc0Lwl;$U|HJqH$~mfsulA2%BRzK$2(e567)>W`{jO
zt&xdf&=39PzUS7WI;gXavU-l*Tc!<Mhn`*REg`T=i*IVrHnD_dFYYVylN;yW9A2Mg
z2k)cT#zIo-@kdJrZzwXP?)XYh6ozo6_Ab<O$AlF;$26KOox)QmZ`a7aMU(o3-OEwy
zYLM}Jhg7_p6Ay~zu1h%!ln>e>y*JGns(}XZ^^3ZvmQK&S^i~=h-zMhYw0^9xC(|Y#
zmdR|lqw+~l&Vwr9Jk)bAz71(7lSqo)M_E0GE$aQ5^BcbQBc6q(F`*HkDK)p0$;2I8
zn>Y+!OfGB>IrO`Y-HP@09#28<J8v!7vBM?`hH)>q)A0_+0KP|+sqMXGQjzQDPDZO+
zKrOJ6d|zH?U9T#iMqk(|VXO9XPxSkpP|h>~f~`!l4?fK(G>IGbVSP*w0N!KeVRg1g
zFS05C0FVu#o)U7}T6bU*N6Tqr-Rm0RBzJ3sBB}l~rD4^;hC*e&I7yX&L^Fk>$9uxc
zqn@}NIc~n57|nv96J8ErJZuZuDiHk9Gw?oFRz#HTzvR`u^6H*Y1Ep7M-I<-G2CnUn
zJQ*o*_uF=tpE2+DRyQfT_9?r&Kw=&)ZpHRxQ30A|IM=u4ZJQY@LZL60$pQ=-ae0x-
z3rWb6kb^Xo2R&=#)7NzeG}Dtm)3hTSoSVA*(I(6N_6~a)-bR+0`=Nd7E+T}_{xF4_
zpAdJ9j+PAVzrW(ztI_fbjVp0Y7&@IR&GA1T+Z$Wl^^?%ccoyV-)f;J6Z#FHO6PUpB
zc!^=&nX0<404TG*=X85#)KAQ{Sf5dlN=6`3ens^PF>GT=Fn@7DdB<$1A5rw8$=2k`
z_@FNQ)ymHL%W=D$k}$!Plb;l{L-9>{`UWi7QmjJZjHrD*l?*9BZWU@o24cLK)tq5D
zr*e-uMyOknc&M0M1UyP@jZcUBS;j0!cRBllRLsSeu6QN~)d{giG&eZafl5L$Jjwg>
zLW8!~2M3k(M!_DTZJWRCxyKs~=lum|UFuzCd}c^sn7xggI*-2z;0IyRa@+U3&qpPD
zz0FzVFp>3&Sg1<*N<s7v8&7du;S{yC{%m+w$BsF4FbMa0;xTjFxzuYdnt@e;VfST8
zoWMcaq_P}7g-c@sD-`BO47>muP~$hXGi<}MevqF-=g~8<6Z`5NV()wQe?lgY*zgTQ
zs&1>qTHb+B>jv*G@o%oUzMWL_cbiE{&=+SKBsA`(nN5^g`!Ay~h?KF+4PS8piRgz~
z!9!_6xlAke*`dll;M7&yi7vqAa0>lxT~)trgBD3=X`8l6z=}I^(_wGzicqm{VDlWY
znR!Y7xE%7<o*d~v?AEJ#dL#DQP_16yso*>Mv{1#SKMo2sg=|B2aGL#<A9pgpZg4}<
z1GGH)@PFR(oQ|*Zlmu8q#u^D98i>5&iKylb+6~w|HBzhE6)H=$Twhpq2m3j^3N=Rs
zEky-bsP`Kt@uKSNkKS%%i+iAR9MMO;QA9A0Se<CspihNcqW<_xeI@9KaK%L!h2{2x
z!u6s-O)LPVMz5Of>oAFgmZ{Wu(dTCvf~l{9LMzl`b%c`pfV5ce?vvL(Xi;`6I`}Fe
zlApl!En_SN9-nrNd0{=kl<l!m_d*gZaMM<pi_3Pj4kyZa!NK<v8fLoBl?unv2`;rw
zo}PdyzuOMifvi)V0ouuhKIk#O8jAglS#VJAUNhI<eoZ1le&2mK5p|EBZ1wGdsGBgI
zzB-^6cq&e@qfbkFWCJRAi}liu8_9X&x#X3$Gwg{_?&)qx2y}U_H4;vW;(lbUe3^m2
zGBqM)9}Qkw#3ge-yq47<!TlZ70L>{czkvHu0-V{@BYgjaDOyD47bsTC^(&iX#<;%K
z+SFBI`v|TM`*0i~60heY2wdxfMR)F+L}vef;Y>B*Ly96r!7yH!12J~IIJt1wuEI8S
z%E%1Z=qK@J84Mlu;`EHCo{Z%njNb%%2i~I^a$}-KIRhXO*3>Vk@rMsZFA%!RC#EQ2
zJ8<$)*Xryvk(tNBl{gY;?C!9ro>5{a5$S~0u(tK|z~WL_bcdCLH?6_8FOYWD4ln<v
zKSh8bHjMk^-AtW0sJtEppt8%o?k}#n3Ai3Au5MSk)}QP-smjQhA>RSm2`$5iN?x7U
zt9e1mmkAo7I9@XTQMEiGtzr}g(i|K0wZg9)LNj7J5px54$#@W}A^<m;iy{%bMyPTy
z9z{wm@8f&R$vwH7)^*r$xvhkaPG7X;oQM2dbt?B3uC@8{&~Ir{Rmy(uYld}Ea8=z`
zyWLje6a}T)ovrpqpydb?+%#4q(66t~uuSr}euQ$J_-XlsWkL);V8vqnlj6>$zPFND
zYs_*gGcHf?Mr+u;{39=za(6rkKC!UmcLJ=8d*;sBM8@s|SC<N7!hH=UhFRZ9@~G*9
zvtO$Vc%tt&AAKoqT`PD}@5YLSd4n99p=Gsu0vB9aD*P}tzu@z28JFr%!9+W6S0go(
z`QoC|uU#6*4&q@o^g75;?A}1A=p+(_t*MOH<eqfvY1r)UqNTOdNF`kIkvW&(VsCRP
z3mAp?-afD<3-396iM9DI6kb=@g&Zb>{aQR8!o99hl>{9n-#O{5Z1-NJQlH+Ry|T<`
zGALQ!dujHT`Z~=Crj>#&X)nL5hzuV)!x6;riBRBx1hsHD?it?}Jnsv}5`72z64ABo
zwB(A}f_CDqThQzQ#bJ0)KM**HTr8vD;k(7vbnN+)F;60XB4A;%MNG$g7|mKn>Rw$Z
z<Y8JrQ-qMf4d~8}HY_{g#rXkZh5Gesx1LA4QYG82u#I<=n9Q;dq4Q2yiYJ=l$nxsC
zyf*xGVTlXlA}T%lF93;q`l{F0t=r|=0YR@(9#7nDR6aVGe`$nL74<s@$2XUc-Vm9x
zU2xyQ#x%2M9tIjzz9<ic&xK|Q(RY!XWP|BN)~<Jtx)06YMzt0F$4>*GV&(B^rCe0v
zr2gB0tE5sPY~Sw}SZDoHXA5m=ScKAt9Mct%D~73z%NuI+L8pof(qS3!+oXm1dw%~x
z>0smWVKsVtHJ8dBRC-5?crP`qmtncDT><mVfB(DRcL?&7`%CZ>r~Q0aQp-A}0{E~|
zgc#loOWRw_ZqvWd&l-lLvNFPwy2ZTY4QW+;Qaq<UY969dKMQPA-^iaoF72UFV5l0Y
z&9b28L|+6+1rgSy;=hKT_qUe)&p#%MclQ8N{~+0|wkyda@-#BS!1Iwjg|_0vU~dzu
z{=7HemuB<MEc8FW438}BgSMJmk7pm5H9;QDdr&sg085$e6T!In#NGxXsr~orGALfg
zHxhKwY9<a^h=dHs%5J41`=~k@-%Ru+f4@8RzgVAu2>s%OBJ=-zg5O``EM8dgMV8wl
z)GhwsU-0|4@cLw3v;1G4XZGLYI)pl`Kf1>3|M`PHFU5;{vRL?^%kcXlX8sOl1H5$H
zvZtu}^8fWoyyTmaVt<yrONOKU=ez#?pBVn_)B5b9#!V`n=zsmRONa2*n)hVPoX3AZ
z#OxWpNJ^)X%FaJY_U~W+_jB!mc3*0`wDj-a{P({`fEszTaBnlB?$!VG(|ByL<Lp;f
z5%uT)`VUBs3{?HouK3db{`$Ww@LU+D&UUTA-@p0qe*wHuV30lAacYKL_rHA_C5w~*
z|NmzB&ujSq)Mla7+Dk=KH$TXeFyGCB5C-%i{?;Irhh@&~+Q6*P0^o05Fyk>ZvukI%
zT@x%`->HZExgNIv*%y|?&4$mg6}~?hy_bvX=BcKA`)kOt&^n7jRWl~#A3wv=pWQAK
zYXJeuy$!oVMR|Nwz0UhS<$vRXjf=gr<=AT|Due*OwTQzEsq>Maq*1SC#MaG_6rx-7
z*s|&~Zp4$lW>hNn??ZQl9RSu^=9vtbfo?Hk1l>g=>~u0rnNuJ0H{d1$fOs5aFe%&S
zhSMaa;o5q#6-vW+d;|+7gTtl<($anq5R?MfU6bxYeha|(B|cP*+Bf|DS*(}Tk*FTu
zQacmsSi=!Z8;;#47i)kP59pfbnK_em5<=V;=CetlRgI)3UdjJyQawo3(pcq3!B+{r
z8YmH*0dLJinVxXy$#>k;|M?gO5u}7!EN(LaXcHm{)oW3nr<%6pi@8Y}ZXb}K(rk`w
zyeXt@4om|dz1fWbENGR2+7-UyPoI*S4$Xk^ma`Fvmu_kvS@DFmpw}Z0OG|dgjNr!b
zoh2~AnE<B?%y+Bz9e@@w8}(mKI9~E6k-+`Vvb?;I#8L|B5O-R;7!nP$_bV^%v9Z|!
z)8R7*#gZdE#;kJ7XIS+~(bU`gcL+>n+{EKY#S6(%IgY^ccpeG-JPx#Y4d>i9$I1&H
zdYovq#3I`$*9m3grt%>6!zAthmp=*KHs1Zr+s5}?hTn6+n@V)vM^i5q0$4RSP4JbA
zRM@4;SXIv4P+h4sDWQMfQYZo!ygo`r2ubWjQ1_l;4_~?=-<p%~!{Ej1X0vREcC~}&
zAruKb&^Yg(-$UM){PeMCHtOc4$m8QF>uO2dR2BjM89Vo|$-2Y5KHwv~7m&tw07;{F
z#qV60x)!X0c+u2G4OxgIxOiA+wz7ORQ`S0l0iMO)j-J?5P9Hh|=@?1O(BWr8GdLhL
z*~~t(?qWwsS`)rtknDUmrmTWg6lDShmsumib5Xo?i$IrVr?>DQQgpyRZ0H@%%J=3_
z8t4$b*X<$@#*G8CpA4LJh5;+e)*Erjo^8;fI&S`ce=xdQ@cK_$PsfQ*ZoZgLK6!*5
z^)N?>B129(Y)f;#_knJkv|@%UKYKla^Dl2ZLm$Q)&%zOEm8CYvpNazVx{ZD!D5>ju
zwouYp0h<P=Xxp>@<F1C63PFX{rKo4)=^9*H8+{{)vgp!9f;zGBK5uKCK9K{8+4Q54
zi?nS2uRnbw-WjKep=D$S2p&V`c_lao_L*<PrIHga%&zgwH`s(PzmT4a9$`2W2{>2S
zPLKnY!suiFp$<eUGAHD;MVZ3v^uFKkOBlg#M(lu<iXeFi-(9Mqfd&{6R_ZGRk+nS)
zi4v8At6sWaxpVu`f2I9ogVOw^xBtc4<ar`!lk7G^#;DQX$SG>vv+XfhL;u$=u7S;J
z&sIeO8$-C)J0qhEOMmwqYL@+DeXd6AQ55j2lEo?s)%^E$W`a{YT-O$tPE8{YlP|t>
zj|Pz=`0%CRM`6F#q`o9Vb@a9WTJU4iG=tadW-{k(;Q5Qz(XqlOlE-;wsNsJWL6R&2
zzk(ndGEXZoBbYr_`?236L{kmR|JyQXVcYuLRa{BJhYVQ;+!o6HE*Kp1ZQBsMOzi7s
z<Ws-*?qh&RhtADrWq-jdFZpG3B51hc5*t@!G=5=;;GCqblc@g#EVn0`8vn|Ajw1@r
zL@X?z0y(v7*vrWdc+Lbq?|Q?_tS|kz`*H}+G28kt<VIHVU$2mre*eXW>*iQbFHbr^
zA0Uy9otPf&!RRrYcZu-)-k_gi3W*0==3$$duspXn9VRsK!<ZmFw1apIL5Ta<W(MDw
z*=4Nx*q!%=8|}YQ-%<;rLTkQ-iXT?z6D_D^t}xdo(g4C$aSA~T66D)maHM_%zl!50
zjwCvtXJw0E_(W!x!ndQ0A0=|UGNX|8mV~k!TEBM!4x28Ppla+(mK}9}^pOnndwMJ?
zC{}Kwky`|J3otXjMxm0$=;`E>pPr&qVT9xYLKgz><b%=T0SFvzATa<zxN2%LSd1C!
za}r>l20FM@ug$_0;`d>1?6j~wg<%4XB4{=e&;-!0c#PD-#B)+d%-}lL4p(rPLe}ea
z0ZG>z5!fj3jgR`p$(qgwvBRyEcU>AYx=Z^OSiw60!0yA7?Mpt^KUT=2^e-CRl_yAj
zaUC3ZiTQ7@Z{ItI($65ciI^_qHryJo3Z2BoMXS*<asneLTO`G_*c;|WTPZ%cnLfa-
z_{tg?OyN#4-)sxo^|l|ESt}ytK4v>w>|R^O4>=rI)}rakyoR*sT;~Ez8?0-G88w{t
zl>4nXJYB(wlTcUmoRbs?UdSF#^DymzwR;!Ui~Nn4v!`xRPhvqUQ3V3zZAmk`ny)AJ
z2m$H=%`Z&mJz6i2wkj#wV~yQ8DobCG&Gge%L0A3e^*Ea*^8V)#QNu=d;wD4JcVBHT
z<;Ps3L#iyom=xdVw|&?ingAq4*J0Ng)a3B!NRXVThvOZsMb7Urn_j=K^wv;Gd6dLW
z_FtPM*q^S6S}s5u5k-6awqmXfqwF_yP#&G)yz*n~k8{8+F?}T!(Fe4n8tt0Vi!9tp
z*$mbpt*R$^en)ZEUb00e8Q1-sk{Eh-D;uPZE8K$SQQ~$u`c;;BM0;}0t~qO;-O$9Z
z$m!aNRz>62(lUm-CR3Uh2vD)j-N-*PWQ4vW6MVW6o!)G4`=X%etVpE2oexlM?7<VL
zsdgsCaC;&iCMun?ssUqA*zuU;(G6J&42()X{eWv0t=d`{A%`v}_LI($dD^>pH_}2P
za|rt+xLDI70pMU)En`GdJbm1$J78~j$b@YqhTG_RA;O!31gcW|C|RV!qzcVi%;@$|
zf48W#+>z~f$mF?#4Y;@JnabTmx5cdjE|gS!IR_ke0wAj~`i1bPk+`b2_6tCoZS{3}
z(U2iznV%MD4S6b2t63CFmCIhTMxfp|Ig;f9JSxh|gU1$r#O9@8ZRl^fsLNKq-x9a(
zv3JOA7g$i0wSXid$PR{-koFRNrL${%zyg*p-4M#}dx{0fc2#0ak<GrqWi*qO$zSFZ
z(X(R!_Vv1+<9<7a^g#p@H7SVLpeIt!zAt6(FN2;!5qbf3m+EodZG{EQgazNH&);J}
z6%kTnjp99Y>nzV&f6BKg|Msp_+@d1l4caeDIbG2fNjEDs7m3z$5x3B*DYzYmIMuZw
zyfNHMA?v~zW0v?u@*AoEM)AIn<QYaC@ndOaX)A~|WISh?zg^ys3#&Oc4W@E@FBrGw
zKP2fk_^u3A{g#q*nSqZusOK%5d}1Z+e9z4dqKGohGQ5u45;h@eSLux}g*|8JaJ$N8
z89qYM;21lamOM`QrhrEu0vG#+qaZHS)M8W8Sre~UTheG)=#kI%D%9vNRAk4Ddv560
z!dU09*hwf39-5(jJ#bq2W6BX3YXG6<W>ub(-6l{ZUQIn15&UUI*DDv^732%uA;`g-
zS1R#{wz@QvK_9d1+eUwV3tu{R+W6|?&~hy%gV$3KRb(FOqF?@NNug0+Q1HdZqyuYw
zSOr)~=54TJ`UR&+?RZbc^7VR2dB}g3X-hQRj9Mj10jbX>U@yLWPa_}BE8zv_Ror`d
zWjSaUd+B43Okv=2#!sCaVwNJ_1zV&S4(OlDI7XBR)QhimIpXylmmtRPuVEe)^hw?^
zF<4kl?)&`ZV%-UW#?uZWF+=EDT<zQaY-_DwU#t}=?au5$LTtgs56Kqg1|3S47_$!b
z8Z-^IiTZ|zu`;jt*(K~ya|SnM*m6^JjzS)whV(VkMn~CAEG2Olx~3;zt#%ePBl9WM
z=JJ27E8jqq{2=<qGMY!p+rVjkaf2Yf_H8e-HDX^KYdPc4d*y`o;TQz<t;Kx#n_>%-
z8Vth5pwOrgBI9J^0{-xeS`Xv<aCv0$I;r<5^?ocgxwiM(ir6Pk;!jM!y!J1-wz%yR
zGIpuK=t8F@L(v7<^GQJjkcsUUfAz)v1jCTwiK09(cG$eSDlA;3EI8$Cw(F}rTvH{l
z%qk>-yP^2A)*cNF-ZG4I2L3R0;}R&48_-#>iu1OHo_7#+ZoJy)$E`_AH|}r4eep1L
zi=>UR);MeV?qxD(PC|Xxqf)}SR4b$nZG-YRO<yAv#Z-e1D=;>fNiRvcQbi1m3+lj~
z+VS<kPA0TeFQyNJ|H4@9OUhTQKb}~MTw!g}>*6@UWFrKy(-%-Zp8no^ke+|{2$1)I
z_SeT`Q{I}SnC?t}kZfS!d|>pYnz||P<qKmJ&bx#*f(e$qbkj06u2;N1(ht~PwJBaR
zDHsW8kuH~AX#ev?pnLOvJ3xAmq30q&WGN_Kc`M(*NtwCL=cjD0w2wjrEs~MVKjsSS
z^L+tf8^kO4A)83NS}*D?bWy*uWVePYzc-kS2O{<gS$@g$hz-GiUryh#@OS1&xsQ~e
ztIki<P)hFDYgYBGl|0S9`I%}73Jhbi97h?72OWaVtzb*N{dGBCm6qIV4r*8Ej4`&~
zH}1`=73WZ<MNfj=na|;Y?)o!(qz5+iE7L>0ycHOR_94B{eMy^N=J9Nyl;yRx^-A@&
z%MTrV6^9G;oYmx5qRy!dEf0j2UYFd|)k^MX=`)mqjNNlUYl(9f+??5Pdfua*$0`C-
z6dB+mlka$e5C*mBh2TfZX`Y;Bcx$G=O%mZEH{4wL!)i+^@1gs>!60fb`dH>61E-d{
ze#o*v>MKy2F*nyZaDn$l)&R@<b|2@Gq84bq59`Fx^G;D(S&qu5JBNH}gz})c8hazV
zc~w&6<xDnzU6Ja~N~@F4vN@VxSGT5VKcPK!Y<&HgM(){B2km<!*KT(|t(`B_!P1H*
z@TnCrzZR}o{?57|Innms%!m5{E}V;q__Jx`GZ4W1h6I*{jC&$E8Zf@O|N0=Xfe0<T
zq5G?v`kR8qHbThuwP+c$^5$B<9|}~?&D7$NzAepDonkwHq!%)-+IFe~1iHs=kDiPV
zrrJE+|5&e@rIb-x`8i#Xr^4nrj@fj43YAY2!!CS|*y)I2$S>X`9IxnHKhRy@2R!Iu
zt(0K9zvDl!n*IPJ-l99o>4tZ=_2S5P_QN6LLj_%2?8O5Q-V{?T3gACVCiRXXWo!tK
z4q75t%hJ9+gWUHr8v!vtW4W2c3u9Hr1*^9WV4#NN#Wd!IaK;Vu=PigZ@mb;|c|9kG
zp#M+ocxZU(+Fw0>8}<AHTBi;z_tUy>r^VU%fn+TNSNJH1Bq%5DlV42Tp)v%N67$=#
zdG#o)Aa3`W0ZmDL`wN_5OKnx(_rV%25Tp<3Wy4W}L&R`v%EA7y=PZNgA|pw~>-7zo
z=b5R+H9P4hM8LzMkkOpFT1nnZg$)`c@N^egkA%C0r~Gm}z()_4!TFBaKKu%#cN>zr
zMPT2~KK)M7Ui8A9VoR=Yld+E#LFaJ?if*J4uIK+-It2I;^8K%kN#Ae7(O6R|>imVE
zq@w1_)t80gd$Gh3{CBJM>W3#re#9jz;Cfh+vf{+&i(^AOsC}cq5PhpCP5zG<v+upK
zPi*Q%XxcQ!7ib)0uFqP}%r4ma@UYl=I_M6Or87AusAS$wu>?CPvrIYSDd~6fr+Z9K
zI$LNTKDT)79$<toE_2HaH5|owZKc~`d5R*EzmwJ?Z5)#wgIkwWEjLqmR|OG%-7(s4
z;s)`rm098zI{>dT`lkEN=kgMP_rahJrAqA*x@dEsyCl<+sKC$Y;MnU$Zk&WQ@3d$X
zx3$qZ4V|30@^$(p!S$-iN#78Ui!IuU5(HB)G05#jG+pA6FU@B$aF^CCx-gJ<l(V6F
zP1KIBKn0Vu3dKv@S$JczYcQyaZ|J{EMQEr2_KhOYye%%QZ$3!|vF&<dk90mXysAN8
zn^jztwDTl`<G@kjZV)RRe5(7gz`eYMl^M7F#WT`qBLC~QhZ;Nt*^_ghariy-z#$<@
zvtiSH*YAu<ekc_wz!NE<z5eo9x_hy=tUz&?yFH0<Ls(Q)nVsC^K;Te|L#Z(8o}Q0|
zo~EkSzvCv2AE~$-=BleXgVuWY8AwczJfWnQ!sJ-oNe>g7O~FVyb^BoYe8Q>HE5JmX
zq=BPDEe$0kDRJ`XA|QzzL&_O%?mJuD6^jaK!6<h=T>xcjQkc&P6fEygM?KNoJcgb<
ze8N2mqpq$y#SWfGsmO^-Hn0wf&6keO-}W$MQ?<gLoh#;l9K;cU>d~{<n?6;d#v#c-
zWy9nASv(YLLg?sRN%O&Gb^!!H`r_HyuE{o7yov?qfPcxsXOW@;MuFB!Z4<C7i6-ax
z1!IObExCcK)dWfFzlH0{Z--V6_QLd9(0*fq&Q8~?iC#(JCF2&>4IJBdvKiGXa*Cy)
zE6c&S;PX5t=hV}H|5L3V#cQ$jIksYV<e)jzgjVDD{ya=aej=}wEG>MW`?pwbp+2hx
zVP5Zf4?$sJ=Y~~lmOi3OXf_SI8ns*2NCmZW3x+)(S$~yXXbT{1x{Wj|T98d}Q?rSW
zKpsYVFV~pRo#jgzEqV1l<>xyniW`%ZPc|O(e<^F&SKAQIwE@k}bl4QdYev4?NP%So
zzfm8%To!bIf!`-67K_Ha?xU6+1W|!j^h+(dbU^R8y~E83a_1v9R5)%XgIkD<=kVF<
zT+42ByFCY$nv`-sQUU}v9M(5wGP?|yor;RhsJ=D}?>T0n#b6}Svum<aR9%%^L8`cd
z@&?2^nQ0j&{K$-}DT{ILX+e{MJYzOzopq+aaHdv_zn^ctPs6Mv(?&-3-Br1vtu3ng
z%xzuwOX%xSJWX!k&gYADP7Dz-dYE<fstrU?Pay3EB7citb?HAVR#EI)mbRLKmtKaV
zr>$-zY98N$H(gyzXZFcy#r<|PM3d@$kJxbM$@=b^Y~?&cHGDfkGkk2b-Ec??IFQM7
z#5t@hxy8}Gj_){e=2Dq9<4zd#7iXzYie4p5d-&jL9TNxO!P?+U3wT5Yec0BEHhK^{
z?@O5`s6X6)z7e>0{?a({-Jl*OwspW+58F=3bODcqQ0NrQ#lh?N!a$6qTpc$Q|CoE5
zZ{sd^PuWGT#G0@H>vf#a7FvZUMRpsZon}#G8C>JrEz@HB;&t6Y(G$o~Yc*`5+Yc^b
zWkM>U0C3{tnYM*{<4g)Y!Ys%Di4!fSYSaP7``M2sqPwu^wd-DEg*egK)UNdqUT1;(
zz>nUnsRQDg&K;G}6%;69t4ktPUdA0{{;?z2iaR8_Kzbzs8X;Ul4lyy%xt*4alVRfJ
zehkQaK0JzxHc7`9Rl;fP)OAy35xhTX4KC{>Zx8LB7rOt%`z$8NUEvCvn_-%uEqA8B
z6^(;f_3(G%HN)Iy6_!v4J=jdoe_re#w4`B@aQ1~>$ln`QF^=E|4va18>uuQ~b9*63
zfl7a+gmuts8-9=bk*`;_5(}mxF+7~iFTqX*S_aq|*etYi_HE2*zW5>&=ez?qa?>O6
zkGk2cJu6D8*0dU|&O}}PBgfJGlU3E-7no0A_w7Ce6z#z37-h~mXxV%%LEQfSKGsg+
zpAaV5Rm0UaYF7wbP&6!W98Oaknj$Jpq7~@#&;9}t#tdDaA2(IMXvzJMDq2q7`&)3+
z`pfRcti_95y47PMD`7pV{iLuFTX0&QChE1VN^rE!uNb*=ap;X=k@0Gs2anrwYwHkP
z3Z|#r;zh6UA#aj<?9qApA(Hh&rP=Qn5~jD^{y*(l8EmZPqb%aSAF2$Z45rgToT)e@
zab>w>Mk4<nyX^LmlScghyCzS5nD>pE$nYSMnakoY0_?Ml7K{qV-q%4wfeGsh!q@Kc
z9U)Ve?!lfp!MO~Lf~qx9&bNDx)BRBQL1IC3eqlrf?1hda$FO&rfS{^YM=gDFd^r4n
z=BLl9?*UY)gX9NH7bd<Aot5nzO?fWqUwh&(2zBGPdSy20T9*0~C%MS$TS<b$k5!%M
z3O?t#A!ebGt9w-2O^l`;KuV{8>A1YGBM4P`kcP1{T80dO8&$UOmDFN|WaIpxWzVDy
z9kmtSJd4<3hKZ789CHf~g#V+KW?{fvE84ZdjT7UGqTbA3aizPAh%5+#3EkkLzVqZs
z`pQpA;z_`VVY~BQ_Ptxh<XVzg4U89j*eiS}d6_({AbLyA1YcaZatP9z@Ll`kHFltW
zeB^P|TIL`j{-+~64{De!%TNfqqZbCS<@<Vyh0~+!E0S5gr$W@W9LrZY4Xkbc9l4ib
z$PhxuAUgYRV*l)?TD~|>z34b{5|F@sj}2@gz(Ft^NLp0@Bb{0&c;CLGhMi5Umc>JE
zM$h+$`_RfsDKZn!{Hb=Fm+o3A(rZ9`ZwL#VAp-?l@wxd;x#B-0``Mi4vII$}fko}S
z-&xTrDKq9R#Au{f7*i19?or=$e~4|DQ~I(}0~+O>LKb?ct&bjGa<m<+E;V=Lb?P$;
zQ&JqZ?w}M^ZJGOXMAl4n;cLu&`iA>;gGW?$HmI{nX@VE;Zbsck@`QF+1DO_HUoz3~
zd;#*-P-@=#mvAw$1{1V6_~o{Oo!i+@z^S{6jOm^1#KxR{E<EE{EVKiy$<2-Z*<u?P
zNJ>_qxe~$7A5QDhc^s0unQL$!b}?{$|BkLIie>d`x5{JJ2|c-!-%z05Cjc2osLG%?
zvLAD?il>+itR4<mf-r}O^SHOdBmoc2h)>BEgWL>3NMI_|)m%QQ??{+u)M)t??z3Z^
z@x;pg($+BMq>ydv`x6Y0TPWeMKd#wNz={!Fq>XmmJ67*48mBd|ypOA&*isu!qMN%y
z^0=ks;L_}tv!jR_I`>A?U03hnou6-%<a`JwST}%o)}+BKMWx?NGI;wi!jEJ)mq^FT
ztFm``OxNh`mGGaV9`}wdDSC<3j}7tmxSe2CvUH4+@#liMtASMVGk2)eye|g6N}K1J
z?OS^3E4L4q+$|K^*s{s@Y@r9yv%bdj`0}c*`O;!k1%@Zn4UNXgx3`VgwsO+cD^f_h
z^1?!KlX=-P0gp;}HwvFfHZmwxV*cD+w`}`~v=)CL_nHer-z86@E1~RTBY!vJ#_PI$
z<y#WRwz0c1&8O#5Q|ktHf64Jvy<leh<Cz9BA|WYpv_`V~u^1vieQEt<$1E>(b3RBd
zxJh`7*Tapwm{xB>q^cBEq-G2P2nuI^Ms=<uMQhj45N#qb^w90jNk{FWrU8cZaDs5x
z;htNOT;!VC4wLG7zJt3tBh29((A}J!{wYB#`YCQ(yR1(HSyiDoe|zV|cf(Ra%qISD
z{K9@gh$F1N=4qG#lC|6+I4O)oPGeySd`Td;XfU9@Qz{Aq6pAL>g3HW%`(1u{;$5*i
zY2dx9OE)?b(GgwX4`X=?(_0clLQkMh6Wl9(f`pVw1?tN=iw0q_085zdnpP8<^K=7q
zQ#(DOk`@0XA0F2HkbesY+Zp5v<;EQBT$O9WT|ey>W4i;(Jdi)P7LHo$0q6!Uv)7HX
zc3i$>A76^cq-#2xzwM!#XKld@J1<%uHmNbOu+41rO^tfPG{%&1><wG5PKj-MQ)S-N
zvP1i+BXdahc18`~;T4zV+1z`ghG;TY?nC>7r0at}Lkji_Eq)Jb@1xza#7XTfyR$};
z3Vo#bXj9_>)%>p<pHGuY_=kcxu}=emqK{KUCGEs-FMuhRcLVGY69)+&8oDOcvgwy`
z@QVbbK&Z;JD&}4mw`*xmR66SA@U7xe+58g!gixk9AfO@Jt^`{wP#Rb|(4q~7M9(<O
z#LAT6m-<Ntnbxr78ng`>l7*E|Nsd3Nl4XY;Iv~Bl*t6G2I}hk49zo`XTj2Vm&7|8r
zb7WtUyrGBkGGRN@+>UnPan@@ur}KX9yE8Z5`08`rhX<4y$$Y<I^a^0^WQlckrGM3U
zDDTV<^tGET81^qJ1@1FS#eGJ!agyaCx99IEGNDD9Swp8UtpDYd+vn$(SKQk6W&z?U
zhvlaUk>j7|huLjM=!8x~DurL-jfwon9pb8{h?yX3*MA8BCfAjFr8~R()g-V6BNN+S
zU|FLlNk11Xev#Pii>kMm^DTps&bp=d%Yke`N4ni(Vg!uaoJCc2XkEyf7I`6vSlz53
zDxCTU+OO(-|FnXOng|#)dG_2u1z%_a97c?Ls~w^Uf1w~uy2kPLTo?sERB@+$d(YWD
zC`{z>@@LB8wJdoHspVZHa|}a-=vT+ZIfz}x9F<!KNvm?w?~cRTw-M6ms39MI%sHK~
zu6n)WPw+!l`JmqJq<73uy#vhv=T+JT2HcmG8AU>lWj*ksm@B^;emOAsGI&h%rC3L4
zKh%-kJta`fB2LHGC5b<i-~5o+ImVR|PP^@gZSj<MudBINUKT$R6oZ<}n?Z2igWzFo
z`WWwb@3y-QB~G1|B~!Ls&Zh%r4(pHMK3KMQ`u6_o-0o@S<tx5S^bq;#8j@5WseSWB
zA?$e6&Ew)?^3&Gty!q$bnr{`H?aUX6jnhuc;d%1#rW>>?u?xylGnK2OJ&tq6LXgG!
zm#XLGNsrpO1zKh7%WQ@W2Y=&za=3ntl3#oue0m08M2|{*qEsWgU5#j0fZ;Old+fS{
zjHD%~5}bJ8rGc{&FZX0rT%F#aWT38goim%eTb14l_kp9$og28zcUIALA+VkgFg+Mw
z$2T{&?OnL`PS;1=ee9Q9F-GN*H2uCKoU{cTl|L)TXm+Q!2c3I0It<zh(=x&1t%{4-
z#gzE22PAJzrJHms5B3YWON^z^_xm<cI=zj5|1C7!cupagk3QrQ6}z;*)-|Q;HH?nL
zmpKh~V7s~6?_E-$bF#NOZmJ1#KW;KHj*_dU9<XSzTt|?4dQxsI_MD6Rd@Ti;iPh7@
zkhWTV-M(|EvOjj_c_i91(I=fHPF3ZglMB5Zp*UI5C31)LZY;0(G!edlK4|lyXbNA>
z=G|+>HDi&oVLnC#S|K5-)tZ!bd%Zo<YhvggU5>Gn-MNLIuPWc(0G*@@v`qTzrAw0a
zWYMS9a^-rlfIURAbcaCIj&9Gej4xrbskco^vxGvTG|@24co@F!n^j4Wx!Ch}ccAf;
z=PAa{U!ul&ZN9a#{)fkVo+5NF%JWno^g!QO{Jh4|PV@mf<SuEfyq4Mr=LFd5=|q1o
zy4MrNx4Nr|^Df~6jS#-J?DX8IaY+5;lOwxVYpOpw?XkZ~(Fc`WmlV5j@j;kbXNy}^
zD;8gnk3zykEtAutdP*wxj{v{>5%0B*U+_wl(4w(>C7$iiuI5ios7BR;26X;DCUE|M
z<@>Faoz?Ub<-4El*J3>|ovOXH__Cm-`5oQlMWvzAj7T+15w%t_v2X^hJb2j3z3EE2
z-`Jf@*gL8-yz{Yp+F<WMWk_6#0ZB#4+EEyNZ(0hl%hMf|JEb1Ju*mTC@mTcO&Y<)|
zGhDN=gWqXBb5(zt_|@vk3{ePG-g!0v+qnaB+~+x-?T()2vI+?F*q<urs;T{{A8*U*
zCfnoguIZb6pGv0f9Rz&3=dqg=j>>VX*Vb@6J8fP;5V0&gWv=`z7}(mylyal0Ztz<d
zcsf0F2%q1Y13mmhx1D|%R~dGw*1Fcff6tG(b@%U#Q2eQGZX*ZfRhxwMQ{<Pg=GdyW
z{<NO?ZC0~^t9gN;@(GMqO9KNlij+;GzR<5fRvMzd+)zO#+$zq^k!^j4$+vqq&b|8~
z=E9j_wPL=grAM8fo~(7;sVxRQO43J(KLLzg)Z3UnHWH>9<qRq`KTb{}&cy*&BArWR
z;j&Sjj}kgaP>y`hLW*VZG8*-&!R|2<q+ksLHu&v)1bMOYnZq!mF5k!;wAG>_*e|W8
zV{p2XKI7K8bom=gowSOIF0T?)VjWwqVE5p)&yt8sy{u*_2W02DvmdDRY>S}3dEok^
zeC_0zMJ1*W-oh_As^~(<UE5Rbj@^%*1PBUlU5%&YVJAmbu!^duokO|Am~vy^^ZVJX
z(4btDN5I~Lncg{ilRQU~jUVwx-wrBGGN0jGojD}0+5o3V?V2j_Z$(q*Hz*|!GJBxQ
zJy=n&M6$BbV6mXm_?MJ!K)EGDi6^}SU+h#BqgQ@${NqW~>Nwc3mQPnZH%H%@MWCXA
z(X)KG5;AjvD$3(DESCSu2L6OPt%2&cey83M$`IrnSAQeb{SS!?9lBCO=nz^ySi?WP
zphZ1D*aXPn^s$9m2Mc3ztGnprWTbn_1Ng5l6{>#K>>IFcBjmEmKcX&q?-#yh<g|0c
zRC)eSm#?7~K?~5J4FR~jzxjhL)$;a|HZs@<#aX#FPXKM1!lf9=dVsKktd(I4$T*lA
zE9;VZOZ-`5XlC@odAun&E9V!o%jg4#>E{x9VGj#fkRFm>Q#@SsTD)6(@3|&_rc)V>
zI_bSdzARGC<iR;_=e<`Yq*Lh@3td*|e@U&P_=>zD<Yv{`0l_$H#|F1%!6`dyjMku$
zj70P8l|^Eh9H|=AxaIM@8MIL`*GqT93g`BFDDev{FHi#K${z%hQNo^=DloiWuM*kD
z?FiPexTtcT4`s+Ik)VPdxp(rFxRMxzwsR1|MpsgIsyxp@=$Dmr{mzA>(RqybBn(T+
zZ*5-Nugh*wnarnhZh6*KSSetKTXI&G4vuk8;|}Wb><Si^e*E3i<r<g<@RQUpK8M=p
z724E;Vb@V*$O-QAsoqVe`8qN78oQF#R^H12O5?1NB5SE)%NDn|0R{q2@wlRYpxBq*
zR)Nj=fDVL~_efh{hj)wFQuzwT{9(ka7z)za-X|J@<SKYw47$i7ZrN3b{DS6Y17J5O
z)kxs@@zJ{>NNo~@RfI$onJm5Di2l)3E#fJ9cEpBp-%XLHb)10dd3x~mpJ|jZ&Ysm2
zEIM6{Jr6QfLwPyANu>6eZog0AH4^yrv+1uQr9L40oMw!WbvLN`4ER6Mq)_&83+MU+
z9l%v}zqF!IZkMA=6{3G|^HJdCZ43?+T6(VhpNH?5uqr$8wrb+kO`ZKdK&Dv9^JJ6d
zw$=0*|I>nDRWhJJcZz9!4$$3#Z*%2`3R?PLbUZS?p*OFUG%6b|l$>FN=y4PRtbNxn
z<3f0K#uR5OBg22)1c+LjU(Wk09>=kQKMjx1Nu7pl>ck&}QSw);`uGXJX28W(9Q|yv
zn>j<nKcurh<J^MfTMW;S{~F^5EI5F@V$*n^pvLI&Zg$ekegM{n{UiJ5U3)f<Y5mFS
z0pj?Ox3?^!bk-9GpQ32cimhi5<_ScPpnpH!BS>g}i2@)4BIor~n&3|ymD?zR0-C6-
z{b1>`v@I0y^3UgR5Sr+J_T1sfz$~MbtQ84S5D6@k;QV74!g?vg%h-)bgmA+p4oFh!
z-(}2g|5>M5M~v2s@5)Cq)Iy%G@k;@7+|Qs^Eh=p<4}StH5-^squaCAG|Lri4*FlR?
zZ;e+OKNTsZ{;@_ls+H2>=!ve0q;2!EGreeikvnfg-_4B&x11D}SiE#A&_pn*<*aM(
z<D1bN8(4*4PVECwC(Y1~#6Ot{Y%}G-jTn4HK+w&=zvNpi6ZV#Cf>H6@*#3oik9A*n
zR^8i6h>sR;xw3ACB?z7be+>Oj!$3BG+KbPutT%L2?`z2)>;bUCmUWYQBB@J+Ka|#x
z%Z9Gh>QBK-m*nl`z=dw49aN~4`Xm&i#IFq|us;XpedbFD13h2^?EqSw^ICN9#PHi|
zB>o6MC@5}J0p;)Iprk;i!LVIO;7_s7n#Dw`5<G!YT*M$lu>U>8-YYPY9X3b1pPrMC
zICkh>HkndJ3go$cTAj%k7c8$v;4U)y+j_$4$e+R9@u?3?|JU9@F<W1D4#Nn0u#1vx
z^d9VQN9h4RLUzKPa?8c~up6Nz17dYrpJ5ijOW;$d(GB0LYLcZ%w#)2-+$Cle;<6Io
z|I!v)r<qsna6MHSwG$*0v?N6QaVZI%sf65JZ%25QU<%4AOqg{+ZAYQ_pysgNqfm8s
z_q|k;W-{A#rp%i|`^E&n`KRZeX5K%4>2*k6jvvj{J%KFHvy0Xj;Kg=Jwc}>b?P~OO
zq7gn3<6A=7TnTG|VBylgxn|GuEyqsw9{CrMpG^pCApe<(@iQOCYDIyx8D-~M5SvaJ
zZAC~baRD>W*`YV*SYC!S-W7ldY`MBRA$7%Nn6VDdh2h}8NYe60==QJsYC2;sKO61?
zXhBtSfv4Ac-NcS7+z6#a+Fkj!#C7HEk)3>zN?(%euu)#zz)ZJZl)Uft$jt!)AL$3l
zO;6M$QSC_ntjXoYrb`u$%*-R7{RjtI5Ux7mJ&c7kym=fFUtJM0uQ=h3x*ap3oFXF|
z#Lu_UWNX|}9wHn*#U-xEzxxWi(+)xz`28ALmApt`j<N@`KKp6%iu9nZMRn4Hgm%hE
z$BLx|-oB`>vb!1w{EEmxQA@Pum<^4Rusd)-O(6Tm`~CgnBn+~O(k$I_<xiqUK5dQg
zrlT)OBO&|qe1q)!p3_r^eat`3DarXAbIJ#0a|$;arYu%rp8L^{@9`I)q>f>Ko8+8&
z^PkoD!G_hy%d-5-_-B<SMiGL~iV=JaZzh7$YrIKRJI)u2A;&y{-Xm^2E*Nt$f5J7Q
zV+O|Me_x^?;)lY%yZ@L26hn^&do5mnZ_|?|&I<pv7h5(!L@HL=+_W+1S`n^d^0?3f
zhp*s3&ARQ#^Gm328-oAIMJCzc(k)C>@%r!m?Z0HJ=TzQOp@$gLw!xzkSP6_!2}kxh
z%bMX$cY9T!bQ#Bq93Sk4&&GPkJ|H9A7}ip4%f0Q?>jVFqFGkU#yLEn+96oz6*YHL0
zpaCEQzO?nYtLxRX0Mo`e8j@ms<yk4qEK<Le-n0WWjtigBf#;7)3;AKt=9kW@6K3{D
zt2GlWkw@VL?C^pmBV-b_@g;7<^7%{7*B8K2V!+}T?5y4oPodvN_VL_#3wU0ol`H&>
z2y-YrNb(rGpNah0aufze%O7_!oLPnWT~obkxMVNLs^bHlfm`@R3mP^BwegaD0t9iD
z%$Z8rICN%l;(+GfG@5Za`n_>!5!&d&v`FzXdzAN)`#s}ClM`(Hhjb8=w;z7$Lu@?A
zSEr>L>id|?z5*3qrh5;>*4xn|fN^Bgzh4*pZ^jL&{x`orXU+Oe2#t7v+|Tn}Pn?BZ
z0Iz9`FV8I%Ne`mBcBgoCjPg;}Y#?HggZQnGD5=%@d%SSH%F~$t=SdkUxV8&{&AkP)
z7PmzCAcRR#Ca2$=HI*0i*=GttF9J_hBIkI7%16@S0)zSee;yQ>2e7N(U!LS0TOGfu
z0hr1R@OAG+eAK+XEfW*YocPT!ikQ$w!1>U0H0g1)F1Q*Z4b!_D|5;H!SpQ5c{}XbE
zJ7)F?#L{n9xJm74kauZhXAJj+w@Ei4pnyzY4wiC4`(rj4nV_2-u!blS{{78q_Ik0>
zdNTeE42wA$GOPkUl0d?JkZp~!0`R>!-Zt)va1|y7K7X&N7BuuHKtMJx4JkJ~g3DOt
zTB}O-@u7e2`l2Va8;^Zy+N=5M_gzd$;EAF?24wN{m)DCOPfYqIytqmbd22x)*5wpH
z|JL8H9x?dpa76^;rF&r2Dkkxlj0vPD8#3$}6PnpKM|sWYyLvmZ!gHF4!nhY$*R1nk
z&9*2C-bf*KWp9&lC&X!l<^!@GMUq8=)W1&VKia9k_op<4jC(=DQwGOBJ0?WT;PLT;
z5Lppr)|!nsK+2BYgN*K+`v+vd8IYm@+aYi7Z?K-g{Z&i_^Z&+@O6Cth+B=9}f=L>B
zAZUi_*&fOb*h1^i38`5*VFRO^*~AzU%>L<dlkIlb0n0Ca7dQMdK|+n6NcZ~ujYO{b
zQh!d<8TD2Wr9eoIfU(7gRAM@gKHre&<00b`215=C76u}SPQ19UT~MLJTLU#i_P;e|
z6YKD_cA61We?AJ44{a;laNFI-PC34Td<e&dXanOe(9*uyYl>vR;3<}1*VmAY%6G_9
zz!ve)o*5o9dmc@YtM>m<h+t{uzYtjiZYbw&ep|;k05R1OvdD;SykaF=TA)v`AFP56
zdHFRCQRB1dG$Tot{P+9+!(aRzO%yd+QDzgeFi1LO#JuEwCj*jS@^eaeXd$|{tBfi}
zBg@vd@4ZcOouLn`EsFoO{V}cVcYovW4k(<S%Ej7Ygv+LzWnmKC=uP)w+lE3-$7ex7
z!2u)rel&u`q(afvL5na5xaGfpEen3_M1K%<R3CL4;O--B)P>*iBEnRS9m^$%As`wR
z>eyHTg$S?nGwxO)&pmD{f#ru`+fiaX_%GR^jOdN(Upr!mdGhQMEMBtpbXU+j(X_Vt
zP>GCxy2yQrTd`;H$>@`e*q3`fbMOIT0>oP_ZXK?ZXa9c4UwBBF|HIy!hf}$JVZf(-
zsEt!?GL(vqWTufs*%TEKMM<U%r_vx2qG%T-WeCwgNraL@ln{v$4U(}WNrRz8kurVv
zQ{Ij9`>t=euJ6z9zr)Gi@B2K@TF+Ydy4Ssi0$P}qgJMXq5q8V_cHvk(>Bjaf&dan)
zwN%IMvSkIp4LQ1&iSzKV{u{XJ3m)kWogLF)=(!1foVf%0TV{^>K!j}Q2Kay+WkF5Z
z1+Dz13DWuzXdqw&=#k{dN%s;2zE7X4vC#YVs``!|Ib<X6XDx)w?hglyx;SNaBX}fN
zc<0~mwT!=PY?^S=d%UIGBQ7LD-%$-Wy<+;vax;Q79E7E=%>l{3v9IA3JGcQCl4tj$
z3Khy=(Q1fvC<bF3v#n8GhNhz#44%Ajl5h#SMPWQR`a8YS5EdhLCwzgiix^H8@uztU
zr#THfjrN@&B=D9haGPF^IjFUh*Al~2UWKSGNh0w6-4So0Kj<>pzi+JRbP&^Nn4hqM
zS_GKRQsjeBc{<&OVvAAPwAUbsk<UBNFKZlIfZkF*A$bR{_}D_|3OkMI(*(qzL+3yY
z#5e;&5Bg|T3Fv*WLnV0-$oVPHYP?T@Jq<zU^9u(+Hd>tzPjv~E>ai2xJXbS|q0hXe
zTADDjC}SDNCJi_l6M<Jar6#;ttFgD$0gupV#zu!uRjO4Vj*HQRb@40b?(1qs4D6Me
z{~c4ShrK2I@8e;e*WnFf(YO%;yg@Re;qw(tb^)!Gv9piro;$eqfUK`)X~fVEM}22U
z<e-~z{XVar{N=Z-$sf0a)?tKMV}t>Aut9XIxwgpz)^WKDj{OWM0S(2(L43_EAjDI&
z<-B-;HB`8l9^+is9!FUgUXb-2-An-uTOP)6($x*xEEP*I486eK`|uKCGp50HMD7q~
z(7QYquqFsj^J0|Du0(v0ba;0D^_ZgqT7!9s>~NpMp9NfQ+K6j3>)iyekH26cvwRrO
zw-gkI1r-GFOJVq_`yLI_ss*Xg64fm%S&rG$Jqb=I<{98Nj#|rr&;oKn<6om``=gh-
zYje3q4J_+K`TOjDXph!3qClGIRAqrEAUhMzX42F-1=OJJ-^<6mLLiLpcke!M?&}hs
zb`)@t4FHI7RR*%`q$`MMOe9;QQJ+4`fOx_;;5LDBkhsS~KcgQ>G+=^78yghW{s?9K
zJ}0bxF1)sn^-T_0reBVTGA6L*bp9;?yv47Vtu-|DQ>H<z@kJdnwe*maQQqa2df%Y|
z-t6qNX`^wL1!PMeBtBTe*^-mOo)94e(7!pA4A{aILe6ckWh@L9b0t*iyjf5z)nluT
zu`{A~C~-;l!0vj^bYa+S=evh!b#1WhI5K^gIVOcb59YTA4n3%|-95dN^hSv;<R!EF
zhA%p<I18k~rdmsPky_U8->!6JL*a=`G6A;ani>JtDa<w!{WOnM9&<ASRLMOcCszoP
z#*-k0eh3?Ba7H4MkNA%+3HeM(ncaYI@N$EQ5k7>Wq+7nY<97>mE$x?YB(~&S=(*W#
zM2itUQ|{BjSK}Px<tHKD8~16m%T?-^taxsewl`|{kPI5$(`US;$`!5!+FO9QeTWPZ
z*Gknbh_J02dvxz`t<3`n{e17kdL_#Ep|VfL-&`-Qf0Buwr3Hx9L2D$pB$FYFt@KBv
zp8{%WT_)07Sjii&wrgyq(U=PJCm(S4F(|@rA%cBcpwk9!S-qy(ITo!h^6M@?HoGmP
ze2jIICmOCA5i_{6dv<9eKSBzzj-mYW*S!Hw1Vd+joh&@BDTLZ2U|ZwIUEAo;G~wkE
z^c^%!m^;VBpS7BX{>1f?j`JNqB1_c2vL6NS(=XQ^<v6#4taK=igr<m3R^$8<Q57(m
z)!V|KBp_y~&DgT$_vSSTLQ+f*74+2^1gSvGJ+?#rxg^5&=Uwq5;0#lT26thmUyFVo
zjheEPQV&6?;}ce#da-QJG_G7paHM!=ZRMVmxD3%dMza~A{WG^h>(_{M3iQ+emFR7n
z`(hP!9+_@z&=Dcphg0e#*m=0Td+V`-{KmX~U^*^9kDk#JIB;}<gE6`7#lg|{?Ej)U
zum%W8I{^jy;Ds1(2=U-{a<{N@xp19zBjrE03|!kz%%U7So_Urpa%ug+X8{@<oJ?}o
z?xrH1_p4PmMxgSZvF*idB?rMj?<tG-L?*rIhAsPrx}hO~wW@i)+1a#A;heIA6X4bx
z?;X{TqRfu%$D=va|82HU?0jSjCwjFDXQ}g!lZq9j1DLUfYmO0>Ug_EU_NdT?ITJ^D
zO=f5`eB7*$iTCq3SQ`dmoO=$=-JF{3(7#?kcOKCSLSfV5=_gd;=N?r7e+z~&b~_(!
zM4%J)?Zw*$&Lh!zERipv`x!(Dx6~+b#XwO>Kq~|P9nC21f;Jk6Qc5;=`@y-54h9A0
z0O&Eel)ic8w}^9Il~u_%hfktaTw*52!3d~)eX7xcJ$>2(x)s0H%3S#^tO9!c??~Xx
z6SVJsZ>=gD>&feQ@hNpnMR)x}IVIc<xX=wEWL=<x_0g*|DcAk8h4g`--<QK({;e3w
zgGk!Y!xxK2Bq>Jp?O_eGmf@hyx_`7#iL+|_LY`<eT;0z~wjmRjn<nrd_{;LK8V%B`
zMB>8G183XN+VPS$=YC=__rZ0qJpar;hl5~C&booKQ2fv1;=!NKe5)ZfN8W=yhqw7@
zOdn}2lCi{<<;OE+SQc~mZNe@>i%5FR^PHTr6$Jr+^f7G^gWinBO;Bh9>)&S!(3VcQ
z?b&IEA7QSI-}{gU4B(#nKIFC3J~oH)iI^oIRiNmXw9mAjQzgJOXb|>E^WY#5UI8t-
z#ypwj1;Q&&$yb5!w95(i3&s-78SA{8XX9g<dzRlOGj;#mt2#6bzdT^9XisuajY7hJ
zOqVQ@PSnSF^j`B=fG6;ed;bEQ9XC`{7A_MKvgFg^ku%YS7+9z4fNDKaZTs!l0i&bG
zUJcE=TY(r9O%4-`4wtg!irAp;SUS|F?2hga5o^zTV}qG=0>`C$&8Yu}`j|h_ex$64
z9azv$yGtXLa=-~uN)z~Pn)_A00yjdbv3hg)P#dWPA)5G}$NSgFnmsZ?6!PfH$=%DS
z-(f5_SlaeY6YkU_K|ChD{9ud$itAwYo&q|u_*`jPiebj_F`ftTgHG2z;d)Yr|NVA&
zEUd@GY4eIKQxHf4c3b${CTTPXfP4aKoLvs2mXx)|#JZuYLBpRFKH=U@&gBQ&?vg`;
z?YtWMneNN9$FW!PlIftO3GJ39Q~0e9bXTpkj@iyR)VQ_+0kaxM{BA$_{VAR+h&&7l
zBL<R*+9Fg1XM=|waGHe;^#1;7v$7iCE0m}H;{zuk?tb)EZuex)!XS#rR}&p>%MQSX
zzZD=Eo5vY02bU^i&vRr+7|JeNs8;Zvaw$0NbkZ2h(SN~tTElbO=Ta_&3JKzqELJg%
zD^8!eA=Q!;&Tnb%s=o~|V65?4?!}{Q7~tOxX33wR00{+D37m&mtAzbZp6`+7vg7w`
zO-`Km@8eJHrlHTQw<tzRC2?XkI73%tIFehQoT46#eN|yMTo-zs&w3}JI)|I-w}IZ4
zr-91*<z?p8?3<CIn3Rw|ENlGddkd=2TfcknKsM)W<BkDBg%0oMmLY#8k#E=#i~1Bg
zay`DhMNkd`>zArL@zhajat8sa*~-dt_E=10>eBtUTQ+A8J#2`1_)0zI_}?9mqUkz1
zl9EN5%Z+)S9~W4+>Qq4T%86zfQ+JuT?e4yBwyxM%C_(O}Xc#@uEKu?TeIeJj)%;<W
z%OoYWKfL+YcGsgjPebO>s_2X#K0gM}YwbVp@!2E4ZR4B0JwJ>3+C8qFU-AQP#7&mT
ziakLC=VfVAfYe%AJ_eOt3Jsma%|&N0SqJ|bN8>7$jrB9u+ix{x_-B2YIXEt)T-L3I
zx|tv!ogtf3uq`nr(lx;U1OomZ@}a-~6rW+NH%{7bUQ^py@=@gH#cTjbi+glrzt^Dh
z+`i$wL6qNlI>Y=BOLHo8^Yg#k^_;m_Ir<a4@Cg;?+D=lR;KU}OK+jggJTc7t!?`s)
zN~3YfYd50jc+IxluZ4W*<M3Q8NWB1_FY2IxPIhy*!SGYb#uPaFd^?A27y7jwy!}-&
zF{XF!b}v83y3d8N@cH+~PmkZTBp_6li?a-lg2p<sQg>5%hCeldVGel>GR)pfbkS~G
zq*E0a5B|mG<=<MJQkR@~{%qK^g-(^Qaq{awug*Hi+Q=2bC@oA~_j@Vua|@Rk+UU?b
zc6l&8C+o7_xN)>o%-x_?5s6U$)u5Tr*-?8#zeIl$=Rq<Ru^nX@&h3AG!!o2hbm;8L
zc%rJ&eq)*8Kd+cP&BnH;%+^m^u+>;!Gi%POoY+M2xx*j8Udt;h5wUiY4d-XRBE3@Z
zxI9`H-hCHshkq#~8#)kr7d<`TuBH~uN?1<4fUyQmpPo6tMo<6GZz|&ozyd*&kBy*S
z10s8jl}Zs7I&p^O!{znkOJ1??ruh9CYxomJe|;DazW#*83hL|U(ffOD<s0Uq@_rX~
zzxwx9u4DEw3sHeYZkkg1c-{Kpe>{$s$;8X^W$ZZd=Qpe%UNW)`G0v*PWX|MTMVw|k
z?%+jD0Vt>lR-8ldj{3NQs@XAq>PVXA;Ve|GNgF5d=Qo)Op;j~`H_C5*Ro`-|DJ<b>
zUJ(o54IZXPsal_+iPPwV6ofzC7Tv8w9i+?x2qHm8HLG%akN1;g6YQxYufv6DUP$LK
zki`7VzS2Ny`(s<VBN$hF0w;$4`Aw1uo1mSTYrgGkT~*b(OE^EJf&noG2wqG}luFr%
z=#n05I_h%PzZ>gsrd2#1l{B)Ul}W6g^Y&;I0Tc3@sQ;WDG<CtPb?NU9j{AFUQ(efO
zF;C9Mlk<e`)ux{OZ~>sOb=Q$ZuTH}kThbl0veHz3A=Ps|nKg?#=3G}+vv@9+&aT`|
z`GX;wMsTuAl+XJVjGM4qg2l9<q1V0*H>Jc!^D@Gqc)DRQU7B(LNt0l`t*2L9qSn7(
zoZX%-0}OLBS}q2kl}(~El04bwWwP$eS8n@PtCFlCXT_4SVrMOCyQ#-<WR#@ZHS_z;
zTYXTOXTcGcyV4~T+KGU7kc(Lj^%V@}jTu;}GWL5@%FvX#ti3;HA~jAl&G3-;v(zHB
zD57t$*p{Gk9y!ZxK<4sVTc!;k3og3p8LZeT!=$&AAGxaven9*UOa((8C^M6n&L}y_
zLJCJFOGfPOxfCNh%4?oA#k#H9XB9pCF2>{8U~$dqz60s1nyRV<K+UwD0#oPMvh;hw
zq>NE>wjSfL{<~c<1Q{WugvvK~lT*Y_v!_6addXU$-GKZ^8`zg4%*`A#x8#$RX!iOw
zYYnyE!k(x`e0UnOY3oiL5w-lEPnzb>n8FMdXM<0(6aGW)P5XSkF{|GKZ*}%*^{G*-
z->|uU^U;|vZ#ml>ndP?XAnP4Z(P4UvUEO5Q(T8p8Vr=^5jJ=N3@=xPR_n4aDAyyU|
zsLaFI`VYH(!ahK`M5+74)jQ)yywiBNhXnzTR#EQ3k`Bdu(xGpd{BMWmZwL8Ppw{K#
zOsxB+l6|NS`}Vzh({(C;2lbF#u1c|fp=~YUn!}&kI7D|4*Ctp$YCJ=~4_wJb76dOm
zUT#Z_@a(ZaKW5)nV&9J0T1AfRTYF773ZykC!p)m?(SXJ!x{2Ah{5eQb^|etROQ;ne
zM>D_`VJ;aYLRm`o6i>8yta}RYuqhBnquGMl=i-AoL4aT*+up7`PaSjS3Y=E;vx>>o
z*^8V#2P#J3O~uc5uibo<T(KpYEZ7Umeo;pVq<GnUl$ER^RuWNDOf5mel#&Ls6FEVc
zS?AdDt9r^fnxP!ycvv@1ZTDA93XcE3{o{J@t1xxU<j<uzh3`A$Tf0LCZj8A}E;GJr
zMH*ggJxguYJX}Tc|8}TGN(fo01{DB*2k-ix%z00{ilM<=9DL~WS?tWLTI}9_`|XkG
zUv9q(&S_6C!*|sz7|3x_=gz_$rpf^n*6kYFT3gp1=Qx6ctSLN|#ac_weX3BVE^(&u
zJa+Gj*JuJ+3WQoG1m1E{<H5i6t8;?SS~TVl_2nax`+%yf3%K@W?F!X57NZM*5dWd-
zru5B0rdih(v`-*3hIJU2a~1Pl_l-8et&-#TrY+=8DWi5pUw#e<^`=?@!Aty&*A856
z-+}l!G1|2hV*<%+U$uDD?Qm}^KqkZzXAP=0KQFI-l4&4+quOHo-j0SPd(Rwbm0bwO
zL3yeLci>|#h2LQgpVioZXlXZAwJzMHwk7{s^h@sSDFJ7-Qw-hE%>4uvs(s-hIy$Cd
z81!=|*V)adnAgGJFs3dR&sLmE9YeM}<A#Wi(CRn|>Pzkd&b%GZu+dtR#>}Uodi5q&
z`!L4;-R)(lHlWK>wbqu%oTL2-Qx?AVX!#CeY^+Gk4F98O9U8=-+C>=jE1Vw?S<eH%
z_YsM_V*E_$xcc`hBi=?6K}EAmb@OVDw<(EaV5Z*2{`7abJeWB7H%h|J?H8gAbHVR}
zBm9axJNt9itAAhdb;7ZMROE``v5rwU1EZcUdUv9B=0&ug9?#b`xXYDQYj-^2>-Ot;
zk;LgyXYKM7fNn49?#!cB&W)CK%c)h-`@-CXysontZ6Y&aI=bD5YsG%Xcr^S(L%5Ce
z08{T)g}wrQ-q*=0#8q{E=8hei2OiITxSTEVO7`g`R?nrG{&CRQ{OBU+I-4@F5)wY%
z{a}zqeYqDGa^Ih5DN7JZ|1oXdC8i1uKnc;p$}kY9s;U}j63HjJ;)n_()dQ>6Z~6>1
zPIC;=N#BBGAm{n%o%|n7m|5#cO1bVsUE$_2&eIak^Fg=U!corC;B1kF+PNL+tEPpO
zJ4*3H@9%n~lKWNXjyc?*{m;IPXG=9Q73Z5H83e6pquWaikXu{XG$({sN4qGXBeuo(
zDfQ^}aInq0N@tDQsfJZ2g506GuyOy9cl~t2P^xnn>V2ps=Gn-ui#$^FYDb1)`>p6D
z-;nY>6ARoU5C8h&Sh(u&%HvS!(|^WjoS_ng!4=G$DA%4zQn8x6khfR>vXFxhE}t)`
zHY%tF0#$N&vj~x0+|J01>w(`SLkA9S(dymz$maa;Lz1h_Sf}K5#H!Pj7gh|lDv|X*
zn%1lNa33S!N$47l&o@ilpplZ>dgbc+MemE%jC$b~*Y`R#m`l)b9rYZxk!V?Um~yeP
zzi^jbaW)DH`16}2bvB8*=f0v#m^7>1ZVodb$P^)<C7fDs@z$hBxU}xagRb?2WtbK4
z4H5Cd6^2`Z%I%xsA5HmLF4tM*k}I<HPf_1u>%zwI2A(UIc^#vN<b<0|V^Jud3<KD(
zdJsnT%vvH#4=B&REZyi8Y{Cus;DS5MzxiDTQ9IR~w4B<gakPE>Uj=on|JA2fh5d`0
z)X+fWoR@dmFI3h^{j}!3F0H}NkQ7$WtF}6Km%Ci(M&dW~i`er#V|KC?b$k&Mr{V2I
z4J!6hp65M($<jmR2-(4YiIAC`w?3=@ybHiOxjXdIigkL7qixtb&KUDZ<bRXhL!Cow
zd`Q>4bzQTDKe$pQH1FtCjZdoEKZtvFFRL;Fb;#ac+`rw^{@OYFVD;IrB%fYq0hU?U
zRmW>L=7=vJH*~Zv;IuLo+R7egW%-2;c)3nn``nS64u+;9xP58M^i-q5jTRFhT_^0~
z!<{bvt^J`^zEqex<p@~Ph;vN4+Dn3l_u$43cKec|_Pk=p{1?;AWElDXK#lJkaKQ{I
z2saFn%nZhdn|SUspZ+~dWFJwdmTsB?D=#(*Dxwx?qCiQiLzvaucx`9*wL(1g$Ov%i
zUf__JEx3%)UjGIytd@0hJ8i>mf2i5l`K%+r!HBxCj32=$PHAemN)0cT;-*GK469|X
zK_f~|xH6ysyf!hAx;%Srkul?=e+23s07X>{1A%peChbuQ9XL(mT3{-VyLWTKIHvJi
znv%SxRtqPeQM4i>BwG9sBP}UDBtIoUqu3>Xk-Q!{9lDy6go(-?4xTTDiGwGY-D_)a
z|C!9jK<H@IFRhd<T;lWi-7Px+UUiX{lMTB#CMZPGes;K|^zil<xYAdjQ&oz2jdLo<
zcV)FL0A-x<?SD_mZ18HcLUfJxZRf_HMr9{i&mU?A29S_uDJP_Hl|iYh-tdecWjsRy
z=3t&ckRmMj^{Ul^nKLXs-}-mD9r~r6IXDT$_RC1@x4G4Q3|))7h_l`HBtDFcZ=)`^
zQYJ&8wrFi{vc0t5a+9)qsMM0Gm4$H9ue2>%EhJrfO{Gf=+3VQheFNRpF7@iu>da0w
z>F6p%?@uuB-m{B471PY%JG;)k>HmQI^VBQtmHW5R>ehHY7iTBrJ=xA#9j-{50&9Ai
z^M{$kzhVdSm-vj!hP4A<yL!5ERz(R7+2@R<`#pQ&z?8Xw>PP$KK15$dMm%W`U3;pw
z9=jlgXapSr5W08lk#=d|*7OhkF)c!ia;*Oi3j?{+rTyBiikrU8_-GT0UOpsmDIPO`
z>;?B=E|h}PKO&a9?}7Jv#;EUg2VkGRj*zKy?J!o`Fz8S&pMzF=ofd7McvL^N2VC;T
zDCr0-l3Ob|d`uUZHsUfdNW8NmTGnZPt?l87Pt`w0BQ;9JkeZEbsMdKRx;devlVuXU
zo!x$PN#6=)&EYUA)+X66?D0cx36=GhGugn?E`I~{<dmKE6W$`P6^Pb;-ka<bDl`=m
z)`&&Ym6|MvA1+%gqIu-u)*%$1_feNh|Jmbt96H4Q?T}af^u*}OT6V3_JIFyCx6pCl
zPk2cPY+_*A8<>5YpAGVtjIqWvSfh}}LV*!$WP1OY<GsP9uER~wjIU_$ky`!kWmZ;J
z?h#vJ_O9;{ncWw*$OiXFj{1|D`bShU5UPQzdF&xBX;|{O!K}Y#TAYPY?*p~FE`>ZY
z4lBg_lpem|fzL96&+2IxXGc$3@se|RP7w&_QHGQnWm@x)9!<{h5Q^$lRjE0`GU1UJ
zKTwlDn5lZi`u(NHKD8OXTvztj9$7&df-y;|U6b0vhYh2AC9N*!XkAWVklVhfN4E-a
zrs|Qdv`-}gM*GD@iFgz5o)DJNQK_B7-}x6?U)WeDsk!dt^x@-XJ8k$w$~@0SZC@4(
z%o-&EANLf(?nm(e+oZ`me1>KmAGrkOEwN37GAq3(OGw~Rw7fkhuRi6B?@Q@_?hVm*
z+-bS2D~wu^f|JW>jp0+{@%PC&9J7AU*E{sM^nw&8`;L@tEGsdH?<~~fPXBS1Zx$Fm
z|Au?8_kH`2AmK1Xj~@*wJrFR~*{ZmzMq>C3M`&Yn%e`hho<11Z;fgwfuDUz8SV_g3
zzi|E8jrxhp3rBWC1-tQc&#YzCZU}G>??&%<71OK&C|iPw-1DH+1>rua4Lz^AgTfQS
zD-YV-47rck-O+GmxjD{SkaWvpYB@+&BqlErv6&{CgA(bANumorAqw0c!S_&#4&b50
z8uVNva-aMT3jWwsi2cd=oYnCscU&*mD%661KjV!B?Dp69;Rq7)9&Q&>M2G!p6nRLl
zg{u5O>o)Y#<bINbW{?d|w8^FW^*n#JEqi(Z$Nh52+hE(eKCV1MJ*PD9AL~-?oM@f2
zgI!q7o%YTJLYg*^@wpz>4%8Ndk)wpI-X8pWk}NiI#qdV{zj@Rf$g@P!iP{P1qqs6^
zQq9wG!}Y@*7=KqtLruW9ZPOczG(N8lx;zeBRQp!M#)U=v!&!;FbgaZfO;PSYzcIc<
z1ABFq6zZtY)+v@G7l8SGdR~aqo;6FlE75MW`ohvB3svXb1}zZ+?k#cQzCX)s5U#u#
z(v{EF{9{<I>Qh&Z!~j_bhy0V;f1aE<{FS!Jv=)OlmulYnt=6rXdFJEU^Pg#c^M2E3
zo%IgI@}0Sz4<PsAXeyZg*&vXNF9RL=YxJI{a*HM0w>3CMSp|yD{9roll3I4h8*)te
ze*%PMIJLV`zE7LdF|VQVeW`FjUNAI5+%W@h%Zq5p^!`q1Z$8&cyLAs#%jg5ivUia|
zsp7r^i&HT3;n|hTj!se{0SUZA|I?Vqlqb!el0=gEFJX}*7N^a;>)HU=GoGLC)Ah6m
zqAkXuHf$Wi4*=0alqLoqVeq{{;PjMW2#>w{Q?gz<!4Cw-ghPv3WM}|lje$=??jw0@
zL`Y=;CZLG#Dgz^K)%xA{$%w+xez@a2r2HS<m=&BvDrrBSY^<3R%5(d_SFYpYy?pM3
zMR_l!1+eQ*QAS><0$XdKWes7L;%3j+uTJxPj(Zt%F%z%5@^@TYz?PkfE}%=HRY1ao
z%@Xv>jF)l@+e``}$CWese-AqbCzDvT2`XIAbrs#`zAwB|76%4~NV?=}O-SFF!W_R^
z^?hF(??Gxd_SZe$Pi9@4n(87*wZRdgEy5luG~}ZHPv#y^tNYA|VCd_=`5tpEy;L53
z!rj2M``(EU-F~aUUwnjj!b~?~#hnYFcP6BuwP2v+`;8{rh5cQt+!wxVw>c=3pX5Oy
z4Wt}$P@qu2b^&!VU6CbCa&6B!7oL)kJ%h0Q22gEY3K=g_vB8~Ht@>)m6x~C`Z^+C`
z=L2n<^84Sb40VS15<>O4omq8FAGapc*l{7*0y;vElR4;s{6<^d^T$$~J2?@w@pf9>
z^nmU!mp5%^RxNHo=SxTDvkOy64beW;e*Wj%=VqMSTqO3gb4or>cckzYq4<9vo*FoP
z{bq%J=W&w^!SPgh#Wbd2y9JnShp21HkJgrwVRXkU#WUPaTZpfyYx(wpZ;62qw{A5M
zKpYDF$*lMnH>r1~Gqws62!{94aZb25n`l{AdZg?PCt!e?9HL~BX_=|(9i~DUM$!y&
z%Fgheq5f+9a;Ep>xqaQuLfMf&P=e~D)rIWqd1upk%wQObXR?@4aNF6@svKVs#_AEX
z>uIoCtPM<mAw1}Zx)9%mLC?)OXFz@hJgDmjo!OKJ72+>Zb|5TsuxobcrZz1B=C_=g
z6Kc})?n3zYgjJO7ht9_XMjM`XqYPA+mKNg*U+whRXPG*|hmu6u?OW>C(aq*x8&B05
zlFac}fia?-QikZPa<oYxgNt96aZBVkrrpXojr1D)_zLp~hJZiyg{;_P1j?mXJ5t|B
zAVNJd@zJ0r4b?Z<YpkBXk(Pc#N}6+XppD9=Sg{}QhfcdnCk|)gNlV#=s&%s_wh~B|
z)D7gPl=u`NzZd|tXIl5y#G_QK9$SIGl;5e>ylnU>%y+SsvT7ysWW0|8qQN^(0t%c9
zMoHCNSx$9;k>r54)IUEnn@Vs*XT|!JGiS^2m*oG>5gR}HI_M4G^w`UbB(hsPw}RUJ
z1(r;2nND<7B8Aky|E6s<4Ouz>i%XBf<cEEF>|#9G6??Iu;d7a^js4|y4jB)@@8CS|
zH^#hAgrEtt4bMjCoxC>l_phi;ejtQs!*b0ADwK@Sj!jM7g|@XdGZiFe@s49V33%dS
zBg-+>WJi4T>lJxtk8D%qcqJiegc~L=B&Jc}1~0FJ+Ilk`bM58MQpR;Snmu~ccUB@d
zFMYbfo%3f^r1CW}W9vWEbF_!e#)Ldc294{=8fM?~r#{VBo#XXJYIW*Zm{apb=21~m
z%IyM#l-=JyL|J>afTsDo(spH8BPpK!{m7@8oFC3TdkGQ8g2l;vltU_ImP6wOY(J63
z=YbM@d!{vv|A+H{uOf5ha7z~&C&_`}e8dZqsl{AVpsZP6rW_4U?e6qCXN4$bo+bRu
zat9hftU|*5jdDf}1Kj$vDGDN-+k*L31!uu`V*q7wBw1DXiE7J}ZgXVBtdp!P<*D6;
zNut6rjT}Qh$XbUWVS)dZ?IRB#swCm?2bLCKme(Q}-rIA(p?2W581=Le=ah=|`|7)B
z2IVul6hiZ^_5$KFJTp%^-~J>J2VJbFE^`uPLR|Xz6JyaWdDNerwX_X4RBb__p>?+R
z{oiN2<Wd3dEsQUWm<MS4^E%WgsU<3p$Hy6#wlz^3TESCURkaf`gvt4+X(a-qXQd*0
z_-MAlYA?z&WusAmP${g;48urM>S4+lwea&Vrs3wY<<uG)P$#)$efs~p)F5#Koucvc
zQB+#^N@CW;actyb26Q;c)Cto2hR-M|4+1u{J3EW=VOv94K)5$-+L`I*)RSdTz@1PP
zHYyelYC3@KYQB%;OdfdD+-I?zq;p6BCGw4+8CIkT=)^G?CD&O!hjY=QMBw_LGmmq^
z7A{vA7;cVSI2E<nZligB$=>4i^Wsr4AYCQ|tq)zn-odjOsP?mW3=aIv4E=8i)J=B~
zQU3`p5Xp$Vj4Dv-HzK^xp%ST(c!vz}uVX@ZH?lQb+DfUdnMViMVSjj{>dKj0^7Erf
zy$Hnv%)rKY6h((k8#a8D^UGZwtx4^M895mX*KK1`A20-PU=0C`-|HM?5gh9+WCJ05
zt^Vg^{rP7sG1PIo7H#JM&x(ezXkKi2j|xf?0gp&+yi0lxCQST4DK~oZZU&yD>_&9J
z5vrb&;IR^5em%u)rxLE#;3;MG0df4w(H6T6Zc+bTipcP$gPcH!)&p5x;Bo8M+`=`n
zRg>K=MTZ3A7qs`h@~m;Ez#*9ZM3RLu!ow+J{fQfW)6bJD#LmpM*Sp&lr8JHv5qfRN
zQEQ}>%tenN)n$ju#gzAsFtP=AnrO;i%tx;0=;5GbCEz8=YE+;X3FxJTn%pS>$UfSP
zjOXEag_*Ps<GO@j0swa#X2$CeI7Tg76-1~Z8<~A+pDPl|CTX-AKdLJox!o*CMSkzU
zLC632t)#uAHcjB#Ef)@~Z<BnmhPi<TZtI$#V4BBnoIlhNsX0OkNtp8WS$W0AwqwIi
z&sd+0=Ac`MG5y8RaaxAB<Z47-+ws5H*CoG#dKNNML)Ge@yxuV#juSAp!_x~KZVRS9
zLXwTHkfLCEa>kp%o?CLlxZ&QyU$t9ozm{31oMY-)SV3JpjHLDWW2cyn@2J(=|2gSW
zus-h^=zyY_b%v05Oj8KZNC*lM4S-tCsyUw@&F#^b!$H;Gd4RfX_2tX+eTNoSn@jCg
z2l0diyh-M4aD9AHDU-|Fp)_Z7^(sOSySLz}7j?xT@$S)C1dMv!!&~*PWvKtXAe@*|
z!IJfTq_p=$^;OEeZxKS2*JUZet5g;$PM7-g8{-_>3h@JOqWuI8MUI$AisVAFbCBcV
zH)_(61N1Qk$p7V%fJWNZW;5i~x~T8-2V&B(u3UpCWju}1xD~nvjRlj?{aX6SkP6wx
z(H0Vz^KkgCl3^R%$>SihR$SfHVgJMItOw3Fxm=}plalTJ3I9yJfm*|5@=TOjSKykA
zXs`^Seh%ZEfua+HNaWpsx7uJUZ^plFJMkBLJRg~>0_fg#@$0b%KpA`J`lu~x<N$!J
z_-~C)HdMXfpKRjp?p=1Upo?y~x5}*LJ4Deu;p^5APe2)(BR`l)d^tvx^N|)J1-rGr
zNY^mz>NgL{Sb|1!xeJ5e6C)l&4GP88|4UlZDsH<K??LUEyH_3A%I{OaAj#psZwvkX
z9qjo009>Z)2!wMA_ber>L}gd|bdBMiTt=&#z<(a%ESP{>S>8DD6swqzbfJAX=WMji
zXQ=NDr1=kwuD1RUMJu@gG?qIa9p~?SXOe3oiHi-A=>b&K{Z~U@*u7LP(Nt%>>&S%~
z>(}T(U9QGcXy}%H))0{kXJxm}@9flUZ2>|v$^V3UvC%eCibA*R$}#KsABGp!5rW{a
zAQt+GsC-9Rxo=#RQ!@>9envo81M12rtsg6EM2;p8lu8=kLi0kpTMEVS`iot`N2+!C
zzpAeD4j?t_0dMyU)c#stl;hN|Vh)DvC{wz%9~1{ZrL><v5m{ZZ7NQ~w1m@QdCNfIf
zu6s~Uc@;MSWI@0dzSG}yTtzd$c<q50{-iZzvUjo%<2otarFJOp<QxyKQmIBczpmSD
zRF%qDnT^7(4+s(d4fa--Pz&AiWMPfshIzlFjo;Ba>4oT9A3%?W$P5s{yYU^>sAw)a
z{+0^l=%&GVCiyMO4pe3$>t6t2taD<^dx66cp^>%UuS}+KwN4=_23N$3+O&Q?T2nc>
zbx(yJXM<G7t&xZbvVKUJjw?^m-BPvJZtE`oApRRAd#K&Z|9+@48mT~Px*Ic430A@d
zE+p+|F?e8(JGCa<yq>%lX)?O{)Q4RbT7>gXLWEEDXgb+o>7x6h=VL1vQ!=~ZS}y<6
z&i%sg_bc=!<@e|j8w=La9nb&untIDpNkYS$JY6-Ve6MD}SDagSG`lID`4qw0$xHA}
zG3}LCry{Q=5|P`hAA96FeXF^bL>b{Jg+`mCQyOaXC~v(WrU2#BH%PGxyWeU1_5RxP
zqWe7p3`J$e!d!3)Hu(0x6Td31m>Tat><U=1?X>ecvVDTXY%g2|V(fP)Bm>=5{66!C
zvc2{x)GE?7lRuqe1@SC@z3Wnt>RI$RZLZJr>i*UDW+>Iq*gl{82t}AEIf{Mjn!6^Q
zD&Q@+QvfPgEyD6bM|;Cl;)7q01~vF2=r_Jz0O%^Rb%IMF+*qoQu}Xv<<#5>P#OWR2
zr9Olq8^YS*vl}Q?;@zQNHTezbcxo|5@H&MDWjtx8@hwfnor`(5pB#Pj$m^5J!@`+i
zoXf+s5r-eL(?RI@w)~Wd#4s(|<fYWb_~cI~SWoEt;WY9wX@c*5P{=C-mAspv7gj$0
zG%<o&F6I_2S4o}UD$19Z2hdQ9$@vs_yn-93Rn(UqzZ1D2S;i*WhkB(ghY<Lt{L-#q
zaLmm#mYnR}YvZXjN;Zboeu$(#wWXQmju3Q#7ULUmhASLnWoVhR%E4D%A+|4mPSdFk
z)<5B#114K3jhm`K-4Lv&Tqx;e@PsQ5|0dRj+iprtdwM$oYeRg#UwZJlunJaA@BiYF
zZ_s^sD{OmD<>dZdBad{OJktI{@f7$EV?$lX7ybN#URSPO^@if72RK4&-BKpg7D|!3
z(opU7cUZXc2~SLeCm<`NAUDnbYy2>46EU*^jIY`C<tz$~+-ga@`(>i)vmGihpU`Gb
z`k~*szjfkrvTRJrc&guhIQBNKq{_)o{Tm9uQpY@-fndLQ!K$#~O(?%f5T2(vRi}S>
zBKl0d-4y7CYq(K?mrPg4ANcJ2LU+>QPq7fF`OToC+&7A^{wL(!on}!ULN=KME~spV
z_m~BDtjGs&42LbdO<=K<aS#ps%-2p=ui0JdDr{W8(eYG34FV*m8CU0W4BGS@*6qqp
zy&wvZxM4@G1n639ZBMoI>=Uus?^3Z!=2~q+E&{Za0_iBru+|CB2QLi$ywEyfqxf9T
z9!m4UyM%37v5WF9Yx$zxUv2}wx3{&Mw7DbT`5ouH7YbMsPT;s_n5Fae0UZ4xuu=WY
zJMZ((exp_lF|RqIrlQ6>lxHx#Nuo_Sl5=P$&kl7&pQ1I$B(-uE<H93nu=RzYx@%qe
zPsJ}=h>cFwaNnWmWSac5bU0+i0<zMuc7eJe^Zkh{oi^HH>4_j~cmWy~f)}?Y0o+Uw
zM&L40q_^&w4-C6A6*%^~f!McRPdSF1#D!yFTBumUrh<`;G@R~cIGHflJ57NKP1Re-
zm0iwYqipN?F~-CC_U_GZ$;ht<Qgk$g>?-k&q0k<LUOEH-a*&LtD5ptGGEA=rAsVn@
zrDj33CGV=Y)c<uYRxm*)Hl5qf@P=n&Be*$o@pV8y5IA^pT}+oT9Kk@k>E<O#VuFU$
zMu>`-JDxK0`H}M`VWy3}Os!P#B0Cg%7CyD><q}oijZh&Gp1$p))=>5@?fc!)RPxF+
zR;N8tx!riE^pW03Bq$_#1BL7-XqU#W$v1z4e8O^`&6p;Z6#6}0J2s3P!Zh8N#0O*@
zr<P5dh|n%IRXLMRMPVc(gh$(VMs6R4JLoGh79Iv1my2A$Klt56Kg@@)Mf9e!SL~M*
zkZp9N`2v|Gq7Hl;AI@LWDPaK;)}Dp*N7N_K8NK2_K!2K1X2M9aVXk-12kc+*<$9Vz
zJGhv_w^p=iZGy|O>9Y`Qn1tNLREr*lYK_G}!S#^<&4=Wc8XogxhI5Yv*#(iCcI^Xs
zJ8>6jl|DRaG2x{o>*74EZ$z1TKRo(6gNP}?AOsN;rfzQ3K>4P<AI%_2CLHKV;NF?P
zTGH)>xXcpefZtbucH?B<-ur0iKfiz!1^}U_-f6u;zJuemJ&<Z=pPR=+9bSDUGR&^$
z%j-V0!LOb<1TI>LZqRRTJ+vyx7<p9*6=NdK4dCk7N6!+Drvyb3cHp|?Kp6Fo5jxDf
zdP^XdgShMRQ{0@lVYCZD+^xJ#OimS2C0MH<8(dqh6BZ6`+T2T>tXBB7q~LuEc&QNM
z1us$yg<Z!kFKt@Ct$t3>tIK$c*hubRtQ}B~S?noBi#CI@LSCEmZvQ1l>C1o3>o}mY
z>%?Vcg3QlwEs;R|Z>18AmMORU9=Ax5{o-xyAbvtmYUSwI7%!H`81rJLwx~rCG71>X
zQC>@0;=_@X-@!Hkyhr(EyJ<~Dl7@SGZy}$G8+eCfj;RHUN22D142A8mlDUC%3T{&H
zMZA;z4&{-q6Z4;v$T(AK4g_JAz;6>J3kl$U{VKjc4Q^HjA3Vu9p`g?IpBP1IXO^F)
zb#}Mn%D;sIa!Z_6|J%t#HkCY^6fN!NJwi?g5AOhC&Lr{5lJ@Ddsk_k~{=#WXY7lj!
zmoK8>M#_WyPit!{Z`3CATj0F+AjVXCT4m*rVaZFOIyX4bhi;oC(%uk)1Ws~n^bY(N
z>17(PW#gcKLQrdD<patlVI`)xL|AZfT`^~>a@QetA@q?24;JNum^vm$s-oBOU39u!
z!J;faL73zR&ZuoyZU+Zn+`B!EaB#5JrWaV6?%xOl17yk7%Z|U+Zc@uf$!7Gyk98s-
z^uO{G{U~SaL^|~lDK$)9^4+-A6Dj-g`v@!4$p?lL6N__o2h5zPUHpYf3Jg&G9>b7b
z$OG?q_TLwr0)?~?a*<;6Kalo-$~H0x0{5>kOrD(Gvs^}yG21afdPRO>GG&U7vEjNM
zBRKVc(ay+awjh!Wl-UfXg8?7^<Am6fUqFyfYk2H(;wNA}qPHQET-k%W^8_A#W}C<F
zJBMRa;PfTkRbMDc*%Y!&i6}5iZsRsjt|1LY!pF(zH+)=4&hHEPwN5V36e7$IxUg%f
z@=wZ3pg;!ZY+=QLTux8};#ZLB%TMsyk9Wmu0waTv#A4xKo1Swq+ObXCWHuj+MJ=HY
z)e(%aQ>t#)hv5xB+&EelSdGTl=5M#89yg442PR&0e${rDdJoE1HzW;Bb!A%?d`jWr
zvio)zov4X?Ao34<P|H_G-FMw+va%3DWYf-OQBmq>$J1ce$F4adn~d_x1NS*t{lioR
zQE7(u7cQUft2Q%nc<S_PLnW3q@thmOPT*P(h-`ClJ)#uq!^GGz?l0vc=b#<a%xh=u
zjm>E)g~%&#7+BU^%6Z=_Q2kKfqy<G;Bb32&n^o3~yrzv6i|6gt>33*UcvFzig+Spt
ze4~`=IYYy=wssgOnMzVA$u#~+BbggXICq>UM8gaXNe3H&kxfWi&%V>{;dEth=mUz$
z$27Jh@=wm_>8wk?2d_kC^pZh*MQ1(*7j*F!4N1nhJeZbdYtj=WV&(Q6sCCp%{!t&v
z4M4==w4*&0mPBYb4l7MV*5uT$FRyNO*bH_>0b7X|9>><7LfWt}#z}no?8G9wq2{Zp
zcjR*}xq7V{@x%Y^yI~ruvf=*S@nD44vUWV7eVXz^OEV0PKz@ZS+#QK~5Wt;Z>x6bL
z1ol&N4keNRiaH%<fp#V(x2T#^hdG%WWt9YoY<(k0?MzI!8hGYAxCf2_CxazTCsYL0
z&(LRg)1RI7{Wh2mlhvC;vB;d~gzH9kmFl|(RK6ygSpeQ5TC;wDkenF20da1vy}CXh
zvEJO@UCLl*$gDmcT7ls=d;23F*jbM8Fypz;RyCfUc!qOuB4`lZZb52rmo&H;fe$$C
z)^>!OXYtylvR`q+i}YsJ{km7GaDB$J(j@T+m$hzGiogVhgi6~y{tD`eO*4qbFu{0y
z`)>If9F9iDu*1R|VS@^UB!Lq~Sd;62nWpr7Bqf!5L57a|lnK)t;}y=LoR(CVthSq4
z$W`2EfBH@o`n*0}tg}`@hX8Te;GvhO2J%F#ZbUfu6CuS?JbvXnQqrm&zyIpBtTER7
zpC-A#*kQayO<kQq9dJ@wui5};BZ5x^1Nh(R#aw4CYylFy8TjJi`|HJpe>DP69k<hX
z51kD1KPhj|<+QQDy);v^?*UcbmJKO@G&+eY2rKCesrTdMcH&eS!Lb<1FgER<JY&&U
zR8-5{=b}-PoZ2U6e@2?E*QLnUOyf4%-~r3F{BaLBeL8ssCM#xk7yLaLvj8OaE`i^u
zwf(t8tFI1lDr$77Zq-Au{ngG}R#`sfFna>aK`^Uzg5kI?R3`=q)WRSSrzWnU_?+w+
zd_`*afMGI*o&;|OI{piUctSbeap^Sc61`}4L)u{vO{dVjmq@FG^Igv$d~_CKt<+x6
zf>AD)pnhWDWWtNWhn?<S@0s-u)l-ZY$i+2Wr#F{<U3LxxyAGog1vXBJ^%jJ<$6hZ&
zByn=5;E0_wL%9?kB1Sb>CJ2)TeuV$i-tam@o~zXxVs~lxx={Yg#64Ci%lvA_g}KCo
zNy36vlkY;!N(=2AtrKjxtGzi7dc_(r&ZrJKMhUwB-f-F*1|(+nRcoLdtfXrF1)PQ~
zlhwA_TuZ?m%D{EDA^(?Ij}?w-Hk{!lztAyHdgC^(nx@$noTE_6B$M=J;CzC2xHYT8
zJaOr+K|ODLWGtITg!cGX%$SQF+4i5Ewc3i-f|I2kMxP{lr2>d4UK<y1f~W`!(s77;
zzZ|Z1-G&WOzldYH3lgUo@MLsJtwb54Oc3+a?;*AB4RYRpDaq#s$R*_YTPKupxpDmY
zU+hCjeEVl}w$3!N0BEgirOAroV0zt6JM~NEm9mklr7)&Sg|EPp(6arB)Ku@@L2HDS
zOR*5xHv91|oH+JiK?RtpH{8Upk9uHR=t>^|$PFp&uR9(>o@@{Z4<SE#&X6dRR2Ddm
zJ`!L@cj4$nswQI$R3Cz_E8>DK>Xc(gm)>e5oQmNkJRB=~&wu;dQw>tUK>PHToRIEd
zBn<L7QVxbi+-%wD%TG2YL|K;}B9e77xSEC$wH-NBp8%XG`v@ujE`Ovx>O)Us^A(-G
zqx>hlH^0W(yI8f1P#aBh?!Oc~4e2vnLLTUWQ30%{aO2VJbM+8;f#E&E8gbT<vGq10
zUC~|d!6^8164-0`+&;hTW={~i{o(9(^u%5GtRV`OzI~6RqcGP}<&|=(Hk(kSF>5^!
z*hzZ#VW?}zpob4-Z}6Na4(d;(+E)&;qzH(~Ki7&P&az^Yxll07yDhb!U_E;o{(hu!
zI%|!rlZQDU4f;9glbKK2X5;Sd%B)%s?$>M7c>-U~<m?hwI2V|hx%J$S9K(yywgp*;
z^d_VI6NH<{&p!3HUC_bV_tfL=d{Ciw<lMf)jSe+cxGY(wP}OAgoOz%(CupX|`hnOx
zzT-|%#VS^;0|6+fWd?IBLsXm4hx<r!bG2I0{07rIyLO>kRz+UpAwIbEh@B!E&J;j|
z)^79F<H~oof_1LDtPKMg`|VP&P`aN-itC8*HpA%%UH&!-dHk)ILb|I`+-xH5$zM$o
zc-49yo*afN=k;i_5WOB1Zz6pN&i}dvk+t*f;TcKRZIVN=JIQ5uHx4^_S2Fh2=v9G7
zoe4{us8(MN!t#zEDDCz?5AhK^*GA@9>W0WUteNuBs1XItI%FZFjuiJxyD~5?iBV>r
zEZ)4!NSb3Yp?m~i8&xf58s`&-shr&$M)E7%WwMLy(Y!(gU_#KWZ2m6klk`z_0nbkt
zj}E@m8hi)ODO;MIn)>C&{Yg2HECK>){XJ~-JBS(pMM|4xMZsIF*c~)jp^qsaW;0a-
z?-7ztgMjhF3cWXd4L^wf<M*)pydr4nnV<~P1yoU;IyHg)e^f}l9fxS;c>W0R3?tG*
zF=XrV)lpPwks+JN+7a}%V`#90aIKqGr6A2rtAB#K3Q_3QVw<Y9+emMjjo@T5a;SL9
zD*(B^oyTq3;Tit>*&_ICx3wZ1M}W_+KM=gv*)GwqIQ&!fwO4Y9(Orkd`)x_%XTO<}
zqX@E$F9^qcNCrYWxirt<4peJpJ)E)U!~}iehtH`}1YoCQNUTOvow!Qnb`K5wcvw?Q
zIH1MhTWeBZqZA7DPcpbzW$;R5Ivr^21cKZV8X{eIP3(#gu{6;t>~*r|jYYx2l*OZw
z5@|wte6BJT1Tc+_*rczQm*;p#(Wa7ZfQv}=4MEU_+S9k9H&(m9N$am9asAuFo?M$<
zj)5`Bbk7b53t5sTGUl<l5MD{Ms)w1aKw;FmQ31}^m3Rvg(W`35NOI!`lmRjg)c{z$
z)|A%@(nyOT>=<LIjz|l6+C4igD>egy0suOl1Gy|IdiejTSIidiglKfrt1sjbAdV0g
zU%mg)60sGF00#B_=y1-AE<PZd_8x<SXFzm@^!ekRs=Vu<W>Wk@f>ft5Hf((dkLV~D
zDN9<WA0_Rj;B%(24;xg?2Z#F7+6PhEHFZ|qo|2}w|Jwix)g;nrU84OikcYH}qF`(w
zI2rwjL0;F(;eI@YXZ0l2!t&adXb+c%e@Ds*I6Sg>*<ze;EmfR@NRYq`oyLWvE)Z{;
z1sreLB=u&brvXBHBi({~9C`Oc6bQO@qD%MpB0wjc2(M>00LRxXZiy!mF__cc=VK|2
z39@3YATUK3Sm%%2BO(ajzXM%g07wa%A1ZR^0DVHhdX5thDlQfk>MxJ~cG>XB8n*UV
z4@;X(vI{?qcwoM82sG6Xz79M7*C_9t#DIh2JwxBdxnm*>Z9$DmTcMMZ_6?|@iu5tK
zBb+Ul{2T;VJ-ydNoIq#)(ctZDI?3+iF<5z(x6uN`^q132D{-d5OTQDpXB2^wY0QIT
zb!rZuD<^hKqO%De;4##HOy#+X^hLrqh|Kry|F#AE7W2|GnP1no%kv)>T+;jOLPE|p
z;B{%Lrkd{M;Kx~n&gD2`kS_@%jXQiRowGneatyu>g7c^C_f>=oP*EAEAfSDvfIAvk
zUv4ZNLip_qnHFBh!vueqn>@p>0m()^I0v2w-@SS>=nm#b)Rz{s5rgBB*N3BvyS!{3
z=WB#!F+3g+TQ-l}(p+*9YTTCp{+rh6x<4Njt6L#M4<7KCW!)V@g|_oC`i_yyqW{Kb
zaS6bQ!I+NyGV&AB-fE-&JIW`01Lqil=&%~wQ5r3>aioS2S-_`~i;I`Aj&Ms(e3Swd
zhXnbR0!{g85_?QV;8^QI)z+7&n=B<k1d-$MEMf9az$Ms@CMMA^pyIEawjHiL=!F7u
za=N7cvLhWcMqbZsHWtxwiqj&_+Zie_45kYSf89s!g1-Er9CnQEfGtXUFH!uAi)IW3
zU8r4t(KV1a&3)Ytdj|L6@86KW7dz=roMEMCCdhrb@Q_iOnd{$g-Y*#Wn*U+@gI6K{
z6-zHz%7Z4~UbNswDU=~P*^7!9WUtx=zsYcea<YC$&5Y~HHC3XXWOxMOw?k4kukYCf
zL1MFrS$8(-@+KBf84p%&ux?<SD`UHwnUT6fw%%KZQQsb^9P2kGeq@+uv3C5!c<(p#
zTCVX7093`kB?e2SS|^z32IO(s`F`kie!B(0)Hbs>Yocd(4Dn%x*5jXVcJGQ<aCxX>
zO^UWDJt~`hFD(4m(Dmg{F5P&sOtD<B`>=V+ww90HRH2&~aW0UU<^$v3Q|T4c*oEej
zzf^T{Kp+{RAmS}kB&sbGWp`g7P*g7^*R-}$B5*=y((?Na9~DQvz?J1Vg!;29=5Qh%
z`Sr}m$@<>^2p)nqqN#0G=ci?xFC4kKc^Qg?3Uk)TaSjp6w*R3hu`u_YP9{L&<enIy
zSM1{QXR6Ah8_s0i{A=LZrw!@6JRnibJKj*K;}_>#{0|%MAghYE$Y?wn=JE0lb9zz3
z_-z*s?_fwD{gt(HYZV~xq1SqEZM@}+3YT|gAhb8T7Rw&>ROLzV3_+zWb|c@wc{%bQ
zu@+os&B|COI4!gz`M@7QVJoqh2bFG0qt<EkCm1ip7nsdI#UU_rT+r};h?4%UxeKj0
z$ku<p!IX%tRIN^6bNu7j|6o5LLCR{w8cvY%pKtgt4UYrg4F5;$Togq_*S*Lb{n)>6
znLlg7szNGZ`ovU}{@=eGe$xMc2^jw$EJ3WUz4hbKdr%Tm0O2GsQZq?xg19euh=BF*
zy`K>rf}Zjk7X~OW%1a~2tY!3YkmMx;6;!3cE2Dz7{}rzR-(ne4pBt3E{Fck-^g_h)
z;@4pm%M)~u^dxExY{RYaLLFdlHIms*<oH1C21N3Eb)7^;9X;b~G~szys?RJ`9y$5N
zG=kg%hs=l6!_0lnC!st58l63|iqB{rdAmd$VHt#y`kjY*oRPTia;k%(L7WfcC?W^>
zyEpAeK8ov#N$f5FAXST$?MI#LMjIMY)`WcjGoPLeNLM9PdlohVCkt^c6yCJsWZ;Cf
zr02l+mG69tL#3?vOj_!5BpS7drO+|cjS`GPuG)H*1$sp)_KuJkK@q$PGs}FM;xghK
zdd1nYJVo~>nDDHW?0kF=*$5dd`tXo$^4#fQy`qlaqoA351A&Oz_Ud+mbH4RMqOYQA
zRB~Q%gue|RGL~Nn`z#m|A>H;*F?j^y=(B@?6D?BJ0rYdgDOs%-v?BtGlIqc?Do>})
z0JqT$J^GT?ALlJ$`u?V<Ajc%x*LCX}D=~QZNBtgDv=J16?^BF`HSO^)?XSg6x+JjQ
zl<8!){O1#ApNynp3C^GXFYFIOpc3O5qv*wWzyv$jeyY#@wruheH`JCcM>>vA>+B-b
zW4E&cI{X2nvN1D706OyLEtgR>eFn6hfGd#c4>(!oqQ08$jB^sCR+OB;zDJ@|ELUQ<
zy~z7o19hS##PM87yGRlKN53f5fp(Ck#QTq!=Aunn^h4-MR#Pz63Q!{nKXA2;I^XGp
z8#nUT|3%LkNfK-#Yy!4d-MELr^t8FIS&rFFP*{!sMzZ?RElrFhqO^A9g5BSo%nW~x
zi5r9Q;YOvCJV|h+6p^7+HvKx1zZNBZ`Y0wVD^{44?>0vUaboD$*<8>FU4klWEnOlg
z3y=cj+Ex&$+cI7*@bY%OwE}Ma3O&Z_8+jW>GB{APaco={yNooBF^v^>BlpWgWudMb
z;rddcztTUK30+5Y>i=!0-wn%MMkB}0n44TYbB-MuapjHMaZPptJ38x6(m%ph+qp4q
zk=QQT+~@Lo3gbUnjogA=3SgbT$qwge4E)9R<*{^ou1gSsn$DWL!sjI2K;3?oZs#H>
z!|QCSV?LMFF=DCz!#(Gs_lK}jY}|k_ZBYjo#~TF|hfV7jDFamN1OSKRO<_O3GiTA!
zj1)ps>N<UMq{kNeKfI)>US}2YE&+}v2OXnzd)6-!)AZa8pG6qULU#|TjY57x*Rk)w
z6y7h`yN!c#AQWK`6_P;JwWM=!5~kpdbjx7@dgZZjHoDJ7@Kkz)jV*MV2p0C?Ir4mh
zcOons3ycwzsb?(8#7I(PzrBA{MEWLn?5eonnMI@C&XyY|W&K}IHjNy6)6jw@X+KEL
zT|d+n&O6FJvr!%>`CPe*lF@<#sky5Pk!I3lxH1lPC0(V_{(*UehIddn_(5r8L*;cF
zF#}CSKLLsY#1>30jWS&>ddfk-yS%ac_&8t7*5`+`hO3D@+8fNES%e<WQ7g;B^Y>pb
z{_kIy_sjXAdn4=ZBQzww3jv)EDb90n>(IGbH02*92&^Etd#&Xwj8qHNdm<Qv)|Yps
z2%UHJ7RPYNDM<Yr2w7w4;Vz2Ez(9*qu>rj-Gw?a+??ZqE78>scvCBF)XZ+ooZ6mKr
zZ#E$$l<0-(BbyFq4#6X#jVak5&gv=|sQ05w{W;h=s4een*)CqZhX)rcnVTKuDyK4X
z2QZBfU?r>667@#1`DFxf*!O&E)K$N)C>i(TB%Xtt1A5SukEC+X^C~;-OF+2htzR5y
zC?YBARW-)fLYT|=^|$SAoaZ%DE(bgi$L|jJl$5|e+NoHGS|6!FI6ixW#UxTd#TWLL
zoR=E0ZgQ!#`;RWwj-$O7MB!iNJirj#Day_Qc)Q@3AV`}&Ph4}M_KoKmXPaEV8@+pY
z%5qmOev8`l0O0kC&0H8GR3bJS<4ux-2RwUr*WqFPIwvb4Z693h<PKy}kBzJE0dDg|
z5E}tGnrm(WGrVC(#*?NVk4u)CH6q7yz5>Q<L#JeZsch5nA*70*m2=u%t2b^DS~4v(
zuf6G<3=#FHn-ZzugB$huQxx}JAGerMt-#=05+ttWQ|QM7@TQ~t*##)`R(XF|D|)H@
zJG43+FL3W8Xc%Im+n$wy5Zn{dSgY9b+2tu}S)_oAt~`fT9|be}=pyV9V>3r@{$`gc
zlrgYd1*bw(A9tAO4>bVy_s!EkT>Ymjf3Pi|@JCRKUnE=(@l=>NAtKtnK8G~o-&fm}
zthXhcYMOQ^>AK)gLG>cB;%)5MBsIZPGoUBe_g%~P2-=k4-8z8K={l>-*o7vh=dY1d
zzPJi+vpz6N&9nLBigf2=)2mIUm1YGT=Yp{1*P~kf5?K)deALoaiklx?>%TD2zUAl)
z*OgfZ<XkF!;<~;Aa4Pm*1IbyNp>Hy+6Izsz9g42Wn=f!G7vfG5S(yP{_nW3$#4E%X
ztXO0+LdA+ywpH;<=<}o!L1Sr}i^{J6so(&OmuWYjZph9b0-yH_a5Z?R4)_M~4aXl&
zGE#pZfheMT`kP+8mBmt%aPj%ns}DTX8Z?pKt+3#P>_3_)IGOxSf|O-DknAhpc?{uD
zLI3EaX+^5V{udUhe_@79S0i(GFi~o^pnb)r%SJIo@XrRN;biIQHzD-#DJ|@_PLJRF
zz)Ok|OME*YLjmcXy^uJ&<3NYEiB>@0Abd{MWS_5xdt-|in|5k@V_h$8Rohw~+iSU?
zsP*jj)MLN2qo?HybWUB?FY3@X;kj781lGN;DJpPHllFv}NP|7Moe|;_e_K2*lM@*z
z8rq%|r(#UDlSn|fSw-Ma!XV<EtoFq)QE0E-`#5w76(6W@%k2HK$0ElUQE(rc!Cx3G
z6zl)!JGiCCrrD(Mm`1UI+Ocy_UE0+9Qe}m5flYVcz4|H3RWpL|BereHo|Yh!f|8!8
zo!3EYNXSo!v<W&oxk9BM^030kY0^Qr6UDg?`elOw@xbr$hPH`K>S5@25N;5t3i#d-
z7u)cx1MDyBMM}Ip-PW25f6--lIYoEY-`tVC&M)G8uPo+VCg|XXO*U<6-FBege3NhS
zcvsX)+ToV`^W5zR2J%rQOuK1>&e}T`DX{_}Ou<PI<(#T8Ga<@JGoa(j7H&Haw&Rzn
z{0fGI$MI;XHNzu;D8%(T_pJ5rMG^MIxys}vyY$}dzCW?$#@l-+Mw+>$Co%$A6>njU
z2SB_nUN?K@Vz&V4!b7_dp%`Az+WOoMRR^MacTxIo@6d}2gA4L3oOh!LRD532KPobv
z?*~DZh7zOl3ehXbH${YJd175upq=<zNybr*Z=A@Vr0{FP)@2RvIB*d=mtOSbN$<Qq
z9L@6r!Ka+;)>XUJ7BGcp2QK`<*Orz7j*H8h{iyG3?9X%gb_YNXCWnMXE%x{_L*c8#
z_J59Fu;Ej3(O7iXO3cP&&EY&~)!F9+xZRFuM5RgTCG9OzspC3No1DRX*%Vh0`tseh
zlec%~S!!4p{dj@!`k9+QUGg)6fRcayHRSCC!$x#(Y%TXisHdi>-$RR2ey{p-f&8KA
zJ8o!e@(Z}MYYS|WeyG2q`}{Nmi=B`zF!5CIT3F3>e9iXutNh-k{U{VEwny^w2}3vM
z+U^gey#?-3Wbne5mm$8rY)T7Y`}Lrif`O@_lMeVz8EI-p{T30Fr6niyYu00mi3E6;
zLKf$7^<*@!&Gb%H?wCqoEohYrg7w;9ej}*?oKQi7Px0sC2?@K2@13G^joI!&1mzLe
z#ESy#%Ie1lUqwYkcgs?|vaXW<ee+-~;j~pM#d>^z`giArD8}zezlXUiq!gj`O28%M
z%q1UV2Z5s|Qv+sBLQbb)z5Lj&Ro#ad?YWj&-HP(oTV9tz#GE<3RlG{1-@;_T7EK=~
zvV)~d?39u{@5<;ePqW|iwi%@=<{@+0cB1PEbwu9o!@cz}y9;>??%WOe8s;zp>?#5}
z*?eZX5f%71ud>C>kv7j~|6KJiT@QCYjH`UN>W~nUi(SfzJFV=WM8|bp5!suD%jt3?
z>bsi73#S*URcF|TJeg&eSihreZBzT@QvVkZf@kK$f3tsksrRY3*TSBh;KFC%WnHj{
z6fF9sN`TMJ^BdaL6>>kA^+;BU*nqTXB6i%R_{o55iS9Ak>h{3bH`-=ybJvTU=Rx{q
zyu5dK@n01ymg3Xpjz?ioL+n7n3%kF+dCG~656ldy$Z6vD^DLK`Da^gJaCB#_DDcAa
z-3nW%s7v00@Pg}bSJ=M~2dyqAP17qyPS8dIoQtZARo4p)-+~nKtW&7P?WCh}%lW4#
zY(XBVyZ`dUaeaDPPQhB$G9^ii+`o*!TXnbpG;t1izsv4iU}dk(P`6EI-&gUWPVqV!
zlZvbSMF(0h1uu3&>yiJ*Ss?G_woei!@|`^nF7%@7-@*5gBn24WXFH27(_PkRd8=vh
z390>hu{nkJRMxCjLHzyQ*L{P6()pA+9hRk73#}9O_iI<@e|X(6#ZfR^;C<lf+wIX0
zrpVOooFCZmNy?I|iPN`?(;UHwJYM!_*aNUjg#reGNI!&*Pnzv>Smtz%>z5~5@qM0X
zqazbEU+IE?Qd+uU%JneoD{29K@yU5N<a8FN++_9?4#4NQLDX>nM~yez;_B_B{Nh&3
zjl3E-#Ut>z!108*K#TOFou3|IMsFuF;`s|71byhO-d`DMH#fTWu*8@34_^+r$oxXR
zK6u62rn2^Y@DS{zqiYvR+<oBlTvw#{<s$!&EA14Hqxx~Vtyy_%+@n0C^i7j|_TMeM
zVk<CRXu0u%vSSJrD_yu+ww>Q$@5Ln{Ce;gL8M@|tK%vh`^$Vu0BduNlg)+1{KUAjs
zN5!==Haf0R2)x=WWw~JOh#NGND36m<=4gzLqa}#+BL1EECM0;`1|$J(=JJVOl|J<P
zxRp9^tLg+Oxs&0yZvF{7s?R<>X0>VEjIzwc+pdRt9~{`PLazJZ1fQmjK`Ryeo(w#A
z@{L4yhH~jOmZ<`TV8n+Xa4rHzYMY6{^UaD8g3lz05NhX2ZLa`{U5iETYS_5B&UnzV
zO3}MJ{6haMciZ6mx5a1jZMMSwu3)^=qQ5D@d0WM!iO*{3v&Qd5dqHv5TmR>}(z6di
zAtA5fp70WPMhI`Qy=>k5OI;#nY#jqrZo7*&+OKv$NISX*@Rps{k;1Fc^ZJ5i$=jl5
zZ|36{wZQx{=Tb0~fMEz;Nu`nn=;qb;2MvU0%zQY|iQqIo5@^|7LcF~pq21G0hIyiI
z;HM%7QxT>Y8CJL%>F#_mJ*V+O+>nWSfJVD+-#<rJmu9;hjx#%9XO(>OwcE1AcK}ZX
z8f_}|TG;Z~Rvaqg^(TTmAJ_{M?VPpKq-M``uMV0o72w}}sa>#d;DJ)h_Xaxn9n*8T
zwrrBuQhw4r2=&W=H(zTFn?coGGu!0manDs3QsQP7*O=P}eSe^{8?+ITL!PT=`2<>9
zRcx+2dU^F3f$M}B&acvPbL;iXAz#-XY2maMvoC`<x^?lU4O9*gv5{!blr|{Xq;}-5
zXe9uP`@YEAHM#M9@!O&?RDzI57oX@mC#SrDaa=>Uj_K3n5@?>5*s!4K^1i&!HhF1D
zevCIPXU~n&JuMizdMiV{NM4V0xn?=axzx7=t6cd%+WYcos`t0;5}6ttQ;{-d9x5}5
zZ3xLcMus9Yk+BSkXdt4@^E}&zjR+-?d7ft(Glq~P-uwG|j?MY~o@cH1pZA}4J!h@6
zR;zRN-rw(MyzlG2?(3>ke&q^kj|In8pzdKTGjZASxN5WcL*|=vsTG$(?d>O%HTvV-
zVX4hQK0y6YREFe1<?VZQX7s(G@@6tB>lHT_b5#1xFPcvIaiHtt%0h$$luDr<Pt39d
zf$ytXr7%Z(CwG`!k%P%iMH)E+Fb4Z+u?fkGydg6;{ZZ0tC?FTw+SZeDKzO6K$yMO!
zwwWLDV3FZ6C{kz8R6uh~)8f0Aagr|rNQ_#xt-m+%<L%{DUVS}eIk+eOm@iz$A|#}f
zjy#oBkv%jqh1J*Yg`W!hdg>{b80JVR9zF{mji`E6QhA5-#^x8yRrn@<iY?XI#%}cY
zeb{swarx;q0)JT8O|FAP<Jz^p53Qed;w!~C?0pTmH13=>7&6;r<IqS5tpD<1rR-hv
zQe#-FrZ)TW<a0<5%F<A;5$T}D=H>EEwgs_RsTjrUPhy;y6`Fv8*&A7tpqb;ra5Sw_
zy%d(+x@BgL@JWCb^8g&B`WB!6hwhNYZ6V?VEimC=fpsM*z}aM3q0%RiXgb@k<H)y%
zh;3f6?W2&)0OzJ=slt3OvF+X>Lf@0*qU}FN9>nR;K4CE*87I>rka&9>&|e#I?=blb
zLUsxIWLfqlF3V7&zG#<9tiH*@m}hbVN_f1ATL`yDA(ah0I(tJjGYl%<Tg^OYj><4#
zSr;7^>kSDtZrnjby*$rH!!$gbq$WOp;jezIX%LAV9ynKIwi6E@+jj;D0ulE|wx_>w
zXe&4KNpC8GsG_39n4;q6XD2sK77x7A$`3bCiJtNyfx;}G?Wm~lNzm#=zq;13+=A|t
zLyMi$wq#~X?I8C7<K=!gOqTUC*H|)kmmstE(e+IhrFu=ekbJZ+f{kn+utz8NR7Pmd
zu1n54qA~T)DY*s2ah<w=5618n;+INry5hF{<Ib!QaYwtvvZrxWLx-3G(O!gVIC|SW
zTEq-++1N^?&Mea3p_?U!62BXte&@{oyF5XafSq{#CinPptgnX8ze-z5f#kD`RPu?*
zQhLep(6{vBQ8)~eA;vUdL?K>(C|c+CC~#U}%jF5<Lus$0x??rwUqyKm-;;dFNswT#
z(%3WD=+XOp2{>0SCi&Xgo~bncstbA`d+QUpz){p*;kPi9`_$@`qz-;if?E)(y-uSu
zH;_d9t5Vqz8D{tafoR~5_oN)f%#o9S=#0<qG;@H8vFW2%aJ6@t5qT@U*@ib_F*)=`
z^911V6Z>IY3HwGl<U+x|B#;Tuh=|6Kcx&1Ax(v)vg=AFfmD!u}W)r;^R1|8DUv4+x
z9Gu<J{oEUA2aO<40o(aq)LgCKsAYReCYQI4tQJ%r{hvJp?8jm`1?7dKTl)^$8`78q
zh%<-Fimm#3?}bC^-?RsO4v?~L$2rc9Wa#O!&QaOweE8p>w(yQ)55kLIueXK*t~E8<
z@IPF+V-`L^2^EAj>lux-HRb-<l7R!$x*%`xJW<P6FJ;1R7-HN-<?+yJxtFgd#OFZj
zKW0NWR!SZm;xMKy5zouRP9cOs;|rvoPhmGIk)2xqiPS7WKlIA!?)u+=pJox_Z#(4P
zJ@=!Z(;|V8YE$-`NV+nC#30~8WpPUJXb#`XvZE%qJ9nM}J3TV0h!Z!-fx_1n&>4`}
z=E#jnJnEY65c&*i+ZX&6LjZHQX`}E5eq%WG<{3<6Sv8Q)`8w)^wak9R%e?g#|Jr<w
zUzYyAtXIIq>0UWl+gN|cxqoEe6RO_T3kDIuwdNSXWMX>r;9CwkMAZJTS&qHt@KH%u
zI{*73i&`)K2(@^?><Mr!b*?_%NF9<`Ogka`=-FUDJ<Tj4UNlnt&3o<AT-O@-pje*J
z-y59mqd7I*pen<d`_!f{-B6e{Hk&rs8#Uhx2TbOf?n?+M8+fI2^Xg0BQq~PrgS)Qs
z9Nw&_$lET}(~#<H0lt7Js;G~FxtjVWXUd>*{}1NOgL8qv2i^qbNG6E&1-z#)E=?)<
zqPeOIT-YMQ`vyO5&bL^O!ETI@tsS@nD#2o0{1Mq#9U1U7vouyMBAfMt>3x95#XfVA
zRYDPvbcDPFa&1hWyE|8g(j}fIINk@G5h>ck2KE37l&+}G2zi-L;d?2abl~IO-JXU%
zFO4r;Zg_YY4k<i5-R2rPuEyLPd9ml|<FJ_SA6`ro$YQE^T6zt;n0Rr}p(otczDdV<
z00&IBxflJK#V4U+@A+w30a=S%hQy~MTFBT~ZGt(3?zTF^HERgkD`0ec6+DC#P1wmD
z+t9*P)SDa4iz%<WLK*hDCq2P$GuQCr2er$o)r=QsFL?Kq>GtW0GsPQrLCTzvNyyw0
zrmGYuszWBPM52C^b;ybl@#EWNQ7Dej-y11B@X@5J1(URthxiZNu#Q2d$lB+9PvX6h
zy#l7yR>Bhzf>%6u-OB^sd^GEkU6)@2k;x0cr_w~-Hmw6bTD{P^VXAal^_8$$7`axS
z!ICb4#LEZ-CgT)-Boce^MeM>{o^K73Ln0<ZfH}D32~ap)=8xnyXPbcoRQS17TqE#z
z#yzt%eiQy4<8s^gq&$kNrbj31H%9voYsp6!q<S>BuH5~Xy+g>eKHd;z!%W@g0(NsA
z3DM*q=-~x8Li3gF8K%L$O8TlhSc5g5O1bp2_>MiO@Dwu?{5m2!NZ*k8PD-eiqF6x6
zp8I@GHg16fqwhQGfzCpcn`Gjw+Gs=)blueS#t!OInO4W;GXRI@Cb^n+VS+5h@3uv!
zknV3P6dhtw@7<|B%(8#dljJ=)PO_m`WU=(trM9SmxYe>C4RK6Qqsh$&J95MixROX1
z$|3{Sad#4duMH6Ws(1qwANh>3Gf(TjFG1Mh7jD}5l4NP;Ek_ED?Il*{3%-c8`spvJ
zg}B`S;5gY_-4UI4ofB~Kx4!g6;1&yaT#X)|DJ5euGQ(w$z?E_TFpwap^C77*n2f`Y
zs`$}9NdhX}-dn%adh-lz<MrR3T~usagM+w5H_tL5!S~tPwcDmiYT4&mB@xaTD5urv
zvD|_FeK>FJj0h2jK)~#ScODVGPqL!-IpHM1{KDH!Zd{}<&7j;lH;izuH#ZvN{35Kp
z-H#4r-c<E!?mII~gKZ5oLk-4A{5TEzD)$Ah5wmTPZB{6=W1kGyn*ltqSg+N3Fzi46
zDp7xYm6e1yXe`vaO^xlQ`QC<Z3d=zbvL?JUdUphpDFC)G`QG`we{<$#I9!pL(pTU~
zyIy3yd2jw_ibmLI$0f&z5GvW)JIP=r>BqEnu~~R^$RSpOHnE&ivi_SiNgrXC8}F^y
zg>~O&bH!_3y>-rN$JtClKS<8xbW;{njpA!7j~U1|qFMlKu2l#Oi4wp2gMfVh;ya3q
z!(yklCf-C8^OxW&IDY>%_PBB=TImIK6J8b+AREv#xH`56F*20m*t<SJp#d?Rj*i3A
z&%W32rqN_-kxL_^eRFBZSn14)sOC<nJc&4vjG^)1y?x16jKl(yp4$%Ma4y!O<lYAv
z7?5&)_Rp$LNl81OG@<p<Sk@cagsT@UBQgl_**P3k`_THKR&rLZzeQDNoY_28!NhB0
zE#fo($ZTN~H*a*DZ^DkxHQPA({bTBg7diQh)eSA8(JSwK_FsZ-oG~F-kxX0#^Xl>5
z2Zvdu-&3ovN-h~x^w*qrTm)`u__r-DbP|OX64@#6Y_I<YfGof))eDR<!L3ojfrsxB
z;!to!im^oMS>q-r$@z!C`6LssC%MD!vJ9Q;e7~w*(q@=xB3f{ye~Ab61EXPa>k}|r
z?LkHLOUQ{vWucj;?bwS9MRy&n-%{jc?auA5aa1osDNLj}!Ck1LZhCa*=yZ*)b1KJ*
z^q||;IAt~)7@MfudL$dS@9Q^&0J}P+Mlsk%j~}}Fx&T*n+nAruYI~yy-#{uedoi=O
zjOc6;#CTM_aAf>0G-)z2<?gJWf#$k!)6{;Bn7SM5kgg4Jfnnx+gwCsOF+1_Qzp2<5
z_ZKWdN&;M2<cs9@3HQ&5`idui2rIsHP3oK&_z+>A-X9HjVZI<d)|Gexo2X22&;l?z
zl@;|x=T>{WB_MnNzGl$K!!Kmvurywt^xS=z#JFp=GMw#8vM5Yt`L=^KaMt3pUL-?o
zIqVB4oY9Cl$}QMctaskjO{FvZ;C|)z^~~qwgevzrYR?%7l$^`aD^v6L8scxPHnElA
zz+Y#fS$dEmiyhQ@XNbNBM)8KR^B!aZahD5RS2K<X0}-5}n@vUW`s&NoinnNs?Y>I3
z9Jy?BA4GoDTf-zl?FOk|DAu-XApL&Y*;-tKHZ^aP*#-;5Hk=7$Tc!p|9~j4PF8UeL
zox7@?_jZQXJn^sE(?vNh2<xxHMlo3R=6TIB`)Q%m526C7`5W)&nf0+xI4c&}w|iE1
z#QD1(tkOkNI-el@H9hPd2B8g^9)W~l2y|FF+}1rSZ2HE<BhY3L`aa)-IEE>&8?o;5
zO$ekQ`^yXC)^`Q^wXB$&y(EF(Cusu_H<Qls`5Vi6({%44?=1E!{AS?3dOdWKuUmcw
z$S8@_;Qy||K~?i8B+3H)8Cu{>-f8bt?b-~M>zS16BzRV%XN}1AXBtIU4y*HCxM<(?
z{lP6X7{@&$*?y7&jKGI`fhKI-80QOi<%n^ZX}DM^w_3`J0fk^mkZ}o-F@?zoa=i-s
zS(aa~mN5_;>O6kV=@T@a0(ZC4tffd3K-%c5WzhUF8Sk0EMjS;lIqWGTWreV6Ag-|#
zFP}m6M~h9b!%Zs6S$=Rty`c-@AnY3_-QP4va2oys!GyT7AWF#$?Yy{wR~VH{HKwjk
z{udVWigy9ogYd*z0NToUa!swBO>TTe*t&M(sIVA$7Gc;rPM@ipV|h&c-c**ZRo-jR
z4B?LVC7B30AwLTR+i(CIA}(Y>;O`1i)(>S9Ccj^=fDLmSSbs2{31PZ)DG-ldIbCq>
z8%438W}vuFkec}2KJ|QToJU6@4U1A@k$;(Le{@$!+9f|U4jDV&Gor^mpPeBEjrg%u
zPYN@?T^fukh*|~PnTNCX@|{UNCM~I$EIa5gSsX6o&2BY~p9PM3fgmPJ)L};9w}ziu
zK8g29qwuGlWl&WXb6$*C*0XxdN|d@K0Z!wg@mX-rgG9X7ZVv-K+AbuC5!jwUP>uaj
z0SzGjEc18lz_wV#td$0nDG7$Qq2&$eRkrkQ2nEFsg@PpH$P3J|En}C*3R!zy2O^p~
z!T|T`rfG}`-U+_>rDjtV-3wgkcIaKMM>s=6kkV#&Ia?(5L5IcQn3!xyn{)tVS}Hpz
z!Q@Y43d3*q=%Ih6K>L}>BpjiGIwm`(Pxp)_I4(miaf2+sTt>WTscx^R)k{4u<rb`%
z58gz&#8V)6{|gFL;2@bfO%_wRtv?KY-kVfLVd0!QYmj0NwE#=k!gf1ceU@d>Gky`f
zgh>bxtIULT@}Jn**-=Iv3(-y6fKq5bct5J4IXXTb=M%IUT>nT3PE!c}538UrLsg1a
zae-qFsTacT1I?6-kuR8>*J+B9DTn~;t%7QbfOY~4P#kmqKAQWv*ueXlPUxG_9~vxb
znO@yw!oz$0oH{pkUdDxB3?luz@u<nPX|5YLZ94T_1@x|`80?9ifOI4P003T}U%su7
z#dP;~z2`<-D_@nqf?e@>aF?m|qX6peB!o$bR>K$hJTLYQGfM>t@fhHLO*g!)>#k9U
z?%2d^ogJN5F~`)WSQcH01P+1{9Co`mKbCHxzQ+w=oieN05}rsvGRU*NNPSFGjxH{_
zXmJH3EMvoTrx3wIBpnqveL-{!ia{r-`Ra9&5}>FFL{{tW3{}zDX7gxZJj_4ib=44E
z3hf@Hxc_oCqh0`M83Ez~<3e)JFr-cpZoMcG)Y!-zE1N%s=d(U&HR!dg8FraBQF;&G
z5(*jMBryibvT`2V{q#ixU9W-kfLSy&WP(6sv}w30g}L|5B6p*-$m}2@*j!u%*y*Dq
zQ30D@+@i@+o4-$1g5oD><mPg!!d;>4vUs^MF>soCtmT9(P?vU-fCc-8%T)beC$|@r
z{ma>#{0iIZG6kF$Z^hQ>bZPREuCErK>J!{(%uDj0zxY1pcnut!cSle$@Sd0&lPisH
z*a%!>5Z7??c9d}J#XP)=G2ALTj)5QTzg^u;Qa_ej-oPpeq3@s8(26wb3|h@nC<ZI;
zprYr;;1D-gdM&93ZCg%E=k4Dse3Zu<1^WqCGhV4ZOOwCKJ?a9_^HPhqf6ZJWD~b2!
z*^k7ZpzOd(nq>IHWb6J^desoI5?AmfvCh6au13MGLL<qhQ&gBuN!rfR4;hu>;<WHp
zoB{TH$2qm@(d20LFZq(bM2|-OO(DPVThn2SjOaoIUp%=D^D4B97o>02YuT(ztWQ{|
zk(d9L^j;{dYIP5DtK#G82a+9FH)d27nBIDyK6U6U-bHvE;FQMk&<@rq9P-nRwR~Em
ze)$CEHuqZiUz*{1LgoTBn^5Kj)}pYE+S3gD{C#Gt6SZK9t5T};Lzm&)y^Lq^cF#)9
zvGFF|I9!b5ICrs!4Ql@eI=BROIf9NI$6lrT$LgW~kxmVu!xTdPgInT1{#3<@JNn;z
z68mE-U_I;|PUQ3-AH)VoJ_hh3-wcl)Ie4&t|I-TcE~%2|v6G1nO!mJe@eHIj+uBEt
z{Limb0ax+EOqGM&asRXXkB0gN$*<2Y#S!eo9~z3k|Lzi;;Rfv82iOWWvHJU4NL=Cn
zNc6-z@}HmGf3&Kbumb*fNDfc({|?FF(fHR|``;n?e>Eia)0#K_ag?NBssJwEL#TPs
zJ_hjd(_Nt7q5ybq0)z&zHE8x^a9JJ>KK&8U1ps}E0qciME7&bl;?P;)RRq{dP}oSL
z)S=^HoIXJXu~$M$?!dBhT@*9W1h;pD;s!BjL$qL?1qONpLH8U8qsHL+g@^$vTmz%=
zg9Aj)_9!V;s>|M*<Ni?v=iiR@ovt)Rgc=Nz?a;wC??)TLs+xpA68>jgQoy<eQOE)Y
z)Ey|1q@@9%??-F?1LRB{*neS@CFD1%>Fhc|h@Rrv>=f1h)d(j#L3$NJ{)Gq;kz)X=
z6faKH-FCp6eaWDDcGnwU-m(a=3P|ws^vkulLi^088$$tAuo^4^#mAem#-T^&;KPEm
ziT^lf=ePjUsr((DE`=RAh&3c}qTCbYo~X9RJOx#RpNj+Pg#HSt`u*iDw$R!Z8%YGz
z#MWTOItWX@^Ffu$n^GzNoH{oAg0b>ir-Z>D{nONSF!(;&<47Oo<vUe?vs(}V2r0tl
zqXJb<^+_Ee$8RbSDge=Oc`#dt7?P#;*>4p8G3v%gdnM&2Tc<k?rb3cEA^#IXgoK{>
zs4WS!UfDB{Z~!}rh}C1kY>zqsg9cyX$qB3B(t_+~|NJ^zxESU}K8hYJJ=8Z`zUL^+
zN$9*5{xOv*AG!^F3BW$Gu{5GATCq=8iq&fY4|vp}!WfEf2~X}`$fkbtF)SuV;3GE<
zaz%jV>*e3^dSt42wcV&5a&=z*2b6#;K`6u3bt{|IipvCZ<LKeLx|*(v)93O!PumOZ
zeM&;<1rpl>tH-Wwf`Je3U*hdvh?_Lj@4SqOyb}`+%_Zs~T&O<jd5_evqvpkT<}Bo^
zAd_z`o}wk)US033_^c69UB0DgJ`1D9Ll_x&%U=YvRI;-}lo@-iqoRm8Z<pji8}X)8
zZ??8FK-K(IT-q5bpJ*frKB=f2YzY*dD15rQt51~e;M0RdJD~9h#d`oWX9#~xHwnVb
z0FD_Qr9cjx*b}oE_S20}1z;mH8z;d&G_ux<kg|dct>A0yKs>4P#FGi6DZBgm_|OEX
zm0Ui0q+CpDn&G;i(2@=Tz0`faL#vO@1_7ZIwb$qlHYh#C7O*6qouD!jclW}<yWqYe
zQO@jOqIlo#)J1>QnIpnP2UFpd0QEou*VFcU2cN%t=B(($6-W!L@wog{;GvOE<FUjX
zBc#7~Dk%A2O_C72N7AvMbUzQCp2YI#^d(o2tg`V2R_)L4V`JzB$AL&+6<B0Y;xya7
zNOLgfFsGj=>M4dH8$Y!FQRBcPE!I}0lmid~F$Urcp6kihh8+8<<F_Ctv&oAG&q_c?
zBINZJ@RGc$5xxV=6N!xiQNBbJdc<tN@I(XOZR?-$5(_tq;A_oE;bYiIJc2X8xB>%b
z0Je6U#XqJRen9i<8E(CPbjm?&i>yL04J;<%nf;=A=rQhnM0^AE1v@GpIZWVllO4+1
zV7+z0is4bL^E&(i4C)?1A|Uz$?#g->4kqL%Zl@eXrPNN$uf4HLf`r)%jv=(+$I*=%
z&|nJww+;T^|3G>b*gxopx5bO^!FVoZ-TP-iUXQFig{}v5hX;Ty-=vAMZyNSpB!@P)
zHdFMWC#HLXO2gX{KO31gi8{1|3)SiDDJ21ynEaRs7=}`e&?*R4;}KIfhhK^}y*vd<
z-e7Uca(?gk<klFp7@P+3C1h~1W#D4H8XpsO_!ojMz(92-(>@}?{(x6Iq4YBno{-uM
zfN5ww)>UIHA*I!wnw}61Gd=uS5<R%8@ZNndbvXE}3h@+V)f+odgmLWIvW6nVM?ivL
zS**2rq&y;on+cKma9673#(oB7LQWewF9$KAhl<Qm#56*7RU1B3MD$3%G{6Ld&|eLt
zpN;RpXVo)PmrMpvC)TcDN_~}2#o_-re+ix*dHJfky-Ac2JJcw70&}6qP+D8q+ZzM4
zoCwz0dq;~@LQt>LkvVft2#Cx2%9_w&s;w9k9p!Ha>_hPK1r+hx49Q3*aL~;J*bXmf
z6akV!<aa%~hkX;Td&KSt7g;926)4+9FCG+GapblGNxv7K90!j`+RC_3_8AGoEK~o|
zpQdvH*7=k8mcwZMt9w%Y{@A4l3*O=qT%1eg5%_;sL%o8CasY$^Tpz%74H0TjZ>VEh
zK~e~s5BKW(Y1tjz%$g%sp~LGBXc`MVx9XULvQiv~hyjvk0IFR;B-w<$WtG><aRD-M
zP)XSaB-;=K)6+bPB&Q}CK#}@p(U&U=fQ+kaN5e8Mx9ndhM<S%|*HO0%VI=_I5Pr$f
zKtEYr@ye<-t!FBZ64IXmIzbz^wZgI>#n%A$Z79O7#?$LI)x-)9Qe+;$%mf**_MbS@
zybIWW&n-gMyGn1F75Ng}0mU)!+{pv!0rWAzzluP40JLqaUb#UvXApt8q2kfk^QuLG
znDufyR3M?9SjbF;GaagG&^uqcQcbKs3)kT%==NB^B`p&B1ia&1)@tBG&RNb3!pJQ{
zXs`qEM6_#wQMA1bv2&1Aoh6~k$y;XY7w=l-p+<l_1a`$`a39y3^-*`jp6Q!<Q?Gx;
z%2BEE7_`Rs$x}ER`!a1qsRLIXGzi}<OkQGHZeATb#&vLv;#{<QBPK%JzGR<$uuWD8
zkT!7WBS{@mGvr>txcp(e5vWzC*-Ond@orCD<>l6fusd4M-N*=I@*$_*YXV<M17(J^
z79e72(8M5nbrJ-G>}n~rS;vAcRv^s!Ip5P{?MexDk^C;pdYGGL?MDg*npGq^(jbWI
zpy`Bu5s>YQQX~T5e2@2__44I4?H)jZmfFD#*Ad0<0;B7pJujgw>m}*P&hdp8ND{@4
z=a(_Rfai~>vVr=8<DmJV7iFr-js--<QCw3RaUdq}UK%0|k5;lWjt6Fe6_kgd0t-f1
zur}}C;A3OtU6us*ZHPvHk2&YO`QRUa&CMCst1<sQPirqm*+6rP3214LNJDfMAcR05
zCkWKqGaO;_^rH0_7Bz)rrM?8KQ!W~VHB3suWA)^tR*Kge;DpP3iAl(tqnOG+_Yr2E
zy1vAU^_&;P9YwZdqc2O%Ul>Z@Twh$=1uWH;e{u>|z00W%z4(Vb>X`)$_)+EKobrjy
z*Il7G#lqM=80g?uQMD7eEbJN?`2fdkIWRFTH$U^mIXF0UiKgV5(kBx4Z(*+>;?g&r
zTjo>Ekw1`DNAMF~Jw}_l<-+Ex0RimN)+ZnjB?TM0xeE^t_)&0dLZj2?#cKeWL0Mu)
zTW%RDPY^!0(92t<l$|o=-K_>Xi+%t_vt3XA90BVe?Wf02l8eLN)ghuArY+Q53={K)
z(@)g;$3xVW9-LB@yRkr3gX<2sKk&DgYEBo0c;isGhAZr#o-c?(zm8q&k1l0#wl39X
z2a%JcE;B;5PY#{ipD6B7O4|Jnyo%kN6?l*TY-k7Qi~Q28xJ{GSb0lN!yVaB1Pea98
z`YAqE0XZ_w_RG>%z>GH>h%CFjti>Uw$96zt?#lBT1+#VFam+t;=`sLji=kqoq_2Wv
zv$xuWp7Vf-GjHwN5buTy6(v8PmE~AVBDP8p<nVo!D$C-V_@10N+UHJ}{s~r{I7I02
zUtlc+P`+!A?3pR*Qc~BtV#48PoyQs+X5}1az^a)Q(ivR4m<A`y+GYlJosr0i1Du&+
z7{4a{_ogFpKI*|_vxafgWi?%aAloWtJNMlv!{hNAAur`m?po4~2GSxTNK}5b-;$_L
z^`?;go+qN;l6?vyA<Vws&Gc}nYB1CEQCFDS-jlGpyaTLUmmVlxT%M^aS_ryq%`Tl{
zK|62WA{+q~JkK2zovvX`zPY>?n=GZ{3~0@Lc}Z)I+)<*F_=BnK*Wu2I<%NXrb4sP3
zXuA5_QeN3|hK++0!h#w0PK!uL9rm4eH-X7?lOAmUhThp~D>+YJR+ZU+%kG=Q2EfE}
zgttPrcA*KZ<(qC!{WdXc!>bQ8f;LcB-vz`uvSo`{cc{vs86~x|>D>@FO9gt05D7LT
z^j33GS3nojm~|z-%Ru}9nEjMxd&&(m?UT`@Dvq*qPc<IGSq!CsvImnge+olv*{PMF
zG}p|Xd;3mfFy8eC&D~_z+4SB@>Ma$szRv`bin780Ca9Fx%^>hlHF=$^KJ@YH8ZvBl
zMt{am%qI^@O)^PsBD-hogD#4GDC|Va9W#<Zyg4s;Kc-hofPeK&k%;e?7oqN7Xp*X4
z@8N&fRo9Lk&C=xHJ#h_+H|Klu48@r`*;Epq2#2z<(zgL!K0D)$8cU|y(@FWbF5nTR
zBfz}EO`^aj_V{df671mu;r}S@>$&?Q9taeEl-Z@gLAe&ncdV-0D$*+UQ%>NWdCiBw
z`vF0HYs-+^%^}8=<X=`SV3L08h3FriBuJn57E~_a{~b71b0-wviv1qT-yW$%i3mgj
zwHTIJR_$@Mr83-Q`CIFt70IgXk@g(}Nb+XEm5+Sd#?zG@QT@*8R_1O%f79*L?qcM_
zH-RV{QECCkUNKI-{W3>gD{!R}x4g6z+pS+h$~^}B>ekt%_YS#rcg9fpjfAe6(k@#o
zcU4c&2(>q?IYQ)brT%%n!+kLlKZ+tn?zLGHsP;E^bNF^cLxmM^Ep*eWyRZ5)L?~+0
zKQMFFe{m2idGS%!H$PXkX#SLpoh*qnmx<~7lN&yHWI%Lj`C=39QC0g3To<$PZxq*_
zbJPYhhvV_^^RpmNR{u->TGt|)q<U?-iNBZD;ej{sEG)RaO~qk9G#C{@yoYedekg^s
z8~)7Hyi$aNp@sla*2?$<6;BEvD{)Ke6+9hzjgY7?B@9b5H@$uKTNC3r*J;iJ*REhW
z!h{j#xNfC`X!52-MiLlVG7xW`pvaBAFi@3wV>mKdKlo6E_}@_()I+=&UlKl%MR$ht
zDlZ`RXH)#hfKN^4l6xWtSCMT)h1%iWb4MpzxbR;K4}aeya6aWx9P&iQFO;TXk9DcX
z2?TOvn7YE_m~EQluHD9{#Uv9^a(<iYQI%NW(*v#+;+p{w$dUv5*4=dDLYGu)VT(M!
zw>)w`cok+vO2qycJtu)C2~lM!x1=ZU5yS4d7KK!AP2JWS>!|1}{AsXxqHi8bb|G}7
z|3{uqPiiMgaK_+aqS^5s^5<JX&nFjXkN?o*Y`-o$n*Xefkpq)$3XvgPszrGQdt1Qi
zVKSM{UZu)%)@OhUtnOPKPoD1c%YZlmQ}wzZc}JSjarwIMyvvIsnvM6F_cT>+DN#o)
zj7LCQ5iJmWJ~l$tYDJ@SF%mMMDE&B$H?$(&G0Rj_sWZ4SL}PX41H_*CgWVao6$iL4
zE;ny#vK}xXd?X;1I_nx~wja4dVIx8K1<(#H#-OaF9sCsNfwhZSAyK}U;mi+c9r~Fw
z?Z4F09Lh*G9NmCRC6XwKmwsWdsa-X5Sm>M~g$aY2MpfC-T+zgyOCfv;9n`4-YvgXg
zsZsPOw(LQUY>WjF*pC*mWhcxrZl;+HgbLmmSs37aV76I76MAR|bHXtja5Y@^Z!ikd
z9CL-PUk}GN8ZeX*nxxU@$pHh^DVH)g{uEaqC%PcSW;9WD(nC64gyRpW-KrD+HS{jG
z*x#5rsbEN!1HyQt$1xX{G%Qr&8S)wt`3MN8X+<CAL#J@yJj9!#U^xuLM`;GDxj@K<
zB3w<%{eY{`kcp6%1~dF2p0@!ClO8>8Um;bi>dUJRMde8cX(?R*!p?gzyJo(Bt%)4X
zrAaXBy_5PT+dOg%v=<bqyoyfbyosl|>Z~yN9q6%I3<;x^Zf4q1jCi~d!l&ndf(tv;
zm?iE)XZkvfmRevq#9DglBq06#SDWe$^JF%HR-(2UUBuZtY>HqT#3EaSV<jQv)91JN
z7-^Z;NM#mN08}+F3I<&#t%ij8-rV2N^QQqAMIj@f_xC9wduITR7PGR^e%6qp@wrjl
z=W5M~kiN&O9dJIfox6=ha-wzdJ=LM~yqXSX;zVUL@875v>J~0+QjoiF59qLRcL4Q+
z*0&4C%>LMnebA4*wJzv1ui>$4SYT9}QpWnU0-fzYzhK>$?^fi-s#j_Uz2~qW(g%P{
z_p+C>e4c;E7^<V_Y@`hs9o1}O88j${80bPJ7-Q908q=892;{K@imX4Y6yog)<Zs^q
z1E-Z*{BgxBJ^H7tZMhk5xjl#mLqc9)6=KjZ>evWVGgW*z;&AhN8<V+gu@@tXP340&
zakeu)O1~uC^c=Bhl>lsap*u5Kj{`1UU?`sseOgQPsn4w2PcR0AvItEEN^GLNskXb;
zUnI-E{^F1eH`!36UBXN9R-BSmn$R7AsL`*w{W?@(dY{W9%Q-fZIi^EC@a@3{<#j1Z
zpb5Q@*9ce_L}r^%3Kw3Cx|NWx1K~BJn?&1f04hN_q-#;Pz{uBhC_=H#yLAliP|VGb
zE+$<((Mtbu2Lutofq(IQi}C(^Y!ibnWmU~-KXSG;P%gm$M&~?8>oKzfYFNNu)S+xF
z$@BH(XQBIx8%VyfGD;ii51butik(r@N`)Rl4Io9QJ6;jvQa1~aYjxsX&GB6jsc~=s
zz!<&F*J?!mP5U|-GhnC5@o$WdyY_?%S_6|J?fZRjiiZ3?+bgqxSUpXlU=knp4QN89
zjZ|TBYLymUTLcRQW9=teX&!M|?nmPYru}cri4~XflEI>=83ftIqfgA@7)%DsfUE7i
z*q<1-ous3R-&ftl!o2~>Er+4I=xm3F$lYTjPuDnW9pvVVUY?*wS_*F7zr^E{@aUqB
z{fx%d8+L}FaJ;f+1~+I;dH498M<_0yMhuFCg|F2JxZ*VVU>CQYVSHd#VjFR&3jOb(
zshq>n=^K4yKr*V8r%8DobuC_tgUisqfl1e<XCNET3nS{2knb~_8Y`c=VcVS%OeUP6
zmg=Y6q%pd&v?~7l6khfflXB*+Y3TQ^amwMjT4vd4%v+l0XL<ny`32)D+u^{57`wqC
zJhOY}nD(Uw*}I!E97v-zTv3~&00|66vqRd7CN4Nc2e<&(%hx%pIP$+En@+tupiB<=
zQOCq*=e4AIGj0gIGn?NG|CzC72dc7FNY|of;1*~Y6he6fY)%R@Pnb*0x#+o|HJR3C
z+rDQt*QD7`B6^XVUI;3+cKNl(?jk$>wkDhY`*lb}cOk+2##VBFWE*JnRxMM|nynZC
zrdWlKpMYO$(&<Hz(dh1Qe)<r2S|T)J?)@~W<}tm*SN@x|y8^8U1j>d#QvN<AgJy6V
zBWRU7cc8)bEWxxLl2}kO|JB*Q4V`$+zm}?jCc{sS#=BEY^gvYj_M)+RF6ie|_Hp&(
z-tN<UTh;>cq4Ev;MdMJ~8&7MU6<tG~?q-?i+W(3Cu!We5|0}$Y2cDWdu@Ogj@W29l
zX_}#VrAaQsvL#uWFv68q`w_s9to8CL#>1lV;QC1<t^X@4efsVJT(I6zB+)L|z0j8@
z7|mjEtMJ~9qPNzg6~CZyD+>tTlnRLBuGmCxPcpyK>UKSljmOLyt|`Zf4mia*witz&
z=E*~BUZnL180@jjT}CreB;)h)m@@MY26Ziow}cafS!I;MD|%M2;s3cVGY~L8tNhWe
zedKD~w()}z$Kuetwo%bf_8_d(6v8&?Nvi0f!YMJP0+@UvP?IHIF##@!1Y~>NtBXCj
zr?e><J(@=WZFm@XQZIVlO%tFri@fKFzA{NpepcsaK=MGw%3|IHG7bA_c}!*yBK+2p
z#s)FN4C$frKs7UpAvu3TsBIqu%Va>z#DHhD4MEGUU`oZjt|W-EURbfEoE?S!gU1lC
zI~$u&778lo=iJuxLB-z$<47b<EqKd(J?;@gR}S0@y=1sMIO!>mPu@1$ADk|4AHdNU
zgub75Ie0EwmHH1Hyo1kwF7zo6#4kn_v87LE+e7XXRg5Y!FMEs^{2lXn4V@`2zeIsS
zm}}XoZxk_L0>X-8CJ_6t8;#@;QkzIB#8&#FPO3*)l=E-kASLpNmVqYDdVR2YTU|HZ
z8hCJAb1^AW`@G%Qf1fs$m&H6tx9!`jq+I~)N@T^CvdbGAA9Mpv9#nt0v%@IuV#Tus
zzzg9g1bl;LU7zE|c{)U-;1=`zHGRfm@D0;V;5EGlQpjcgIbFGEZSbk*2EpUYBPj#y
z!A)-w-)blZ@!LQZWCNPczbsClyI(57ArPHR{shk1b1iA=a>J4OXATn>HE=HS5NW#n
z$lu>g>^R#egj=_~mP_GlI+kz$>P37?pw4s+Qoqe_4n>=2jX3~!<q>*X0x!%oFoHk|
zHDM?lgc6F%c{VJC-M8#=IT9}A{N7wzc*{LT^M=b83M-{(H}BPBu}fe`@?hf*+_%Po
zoaOf!6QV7CWPQUcDM$UVUQ|ra@p@do7UR(Ra-hktswI-QjwUtU2EOROSwxboa0~dv
zwle-5i=sA(S-&ebKaiT`4kbthU@y+40+2+&7%jf(+)#oN%_j(I9tvPMdOAm~p@df%
znNX+|)nmQ|V$J|`J_dN@%iABaQUw5FsdGxu&TvGw0h&YVXd8KT2{_6|v}Yr8zKotz
z*5OcNbvAlOu3Pl@U399BaE}2+?Gf;$%+tYH6|zbgGb1!{IL<`IYeF#*X@_~ZYLx4H
zC$N+6eB0gz%s@-(AynB%1?n7<{d1!Gm&m=3C^)(SldFV(Bi<#QE=u<JDhBLyw<}X{
z`2LDlyqWu6>vrXQ7RyPk?ITA>mTyQ)Xk_o4ZWyebrQPe|!RYecioAu0WCMwP+i)U1
z{;}F`M0pchb5?C5n=7g3N-m{$C(OS9?fjuoDt0Rmi^|h8BuwufE#$s~)7;jwT1(DX
z`#OCL%;*g<AQ<cRpx8=w-0lLd(O7@)ooXSETNj&EhjqD{|K*|RxM?^L1~)*f*1ysM
zLMKca?m>Oqi-ZV{>$~Yc>i1U|DGoRT1kAd*$GgP}JET<#&dfGbhn5UkCC)%0z8ZZ)
zNoR60_w}AdzrJxYb&Ng~J@;_*S5IJ=XJ=xw0m`m&681Q9fy*b~X+g`Q84cZ7eiRPp
zr>!9hS_8eCd>6Lw>QhXM$LZP3CbN@h!^y-+Hei<_;A#Yf-rsKypaHrjA5cf1Fug+N
zK9`=zf4%t;=-ua-f0vnPwMHu)lu_w?kPUM9@lWAEY~+;Mia;5EXsw~6-uWqFAOZLl
z#_Ek;Z^Z9^VFF?NJ$F{dKLL8`?|ke^GGoUn1r-lC7WLJo)(L|9I?onSI+3hFk5|sH
zERat>L8dfoxe0)^%VPg3lzY>5fk6pE6|V<U=)9&rvV(7F$uy(($NZmJutnVHO?u1|
z3{^$6d2QZz8pJ|#cd*T36fh9(k5&wH4?}Q}w#W{r?mzE+_M@-&GEVmVz;5>^C=?#w
zIs(OxRG9MeS#(N2@+!(GY{OJQQ~2gqao{!xy?pHwi!5&pzqHpbbsA_6+HmY3l4CbR
zSiin^&GIFQ$(vk?juG>CUzvH-<ho-Bp{&0M>&ZZ8wy6Y)^Xqf~kYa9wl|+{T-cQ6q
z3(6%+$3twqQYzU&{iz`D_s@@eBcrc4IXc3uR;Qrp;ktl&zz>X*v&G%1`gbEvA<U=<
zY9>?{v9W-@Z%yog6aAsYLXrv20F9BB&)7hg06<!GvcJ|v0Yc9goB9X`pGVvh$hq2a
zmE@QN@KY~1$dxQ+siy~OJ=dCvr{Xo$gS?@2yj<*vHf4f!8Ew4(duE~vm|bV?G{nvh
z^pZsM790Cg^r;QCQe?9Y`<s{%K^u_exUAg~;tQI`bvxrDMW-F;JM#Rf0=0RPKucm+
zTr53MMSE~0u?!Cas{$*9fmXBWs3xPq2ycUPmAa<tR;F7O7m(d@^icl}J-JZB?ne=N
z_s18oL}&u^lMjYk+>FzVh0hbO(aKUetzWMg=mFUUC`&|3sHuB0;3+;30!6u(G4yM<
zPL?W1@?4+OK@`X9D+df5fEbPFs?G=$$v6TQSe&V8!=a7NnGSw_f~xbKtzj2%6yT0W
z$B4ETzKaq{iX>w8`~$kZg1}nlF=>q0oYN(!$?gQZ#1_tuoM4F87))i&>8GLjto>ec
z=S7OjvmI%z=F_L~-o`js)>`iw21q_-7J2|edM&Y6`({l^T}<Qs3yHDyS0FjuQW8E~
z>?)#YpEwG6T_;I8GnKiTWYKIlDM90@dz+wVsWrK#b56Xyfj<t-<lCZIn7;z<<oCV~
z6s~AAUQYI!gi}srRzvg-Y&Aa+WDl6YCn8*59vS+e7s{?luB$>P5g*2pVSFiZI75~3
z<gMTCckaite49p9u+NKa)Wgy52^S~gsu*EaFb##&=q)c@%e$}5SC1{_0)E-~LU_q3
zXE$a*AUM2zF==(O?&^l_%}H>acwuea_a9)0;uC3^y9*EE_BlA6vg1kq+?}71a9w_v
zY$+4OM26tzMq1%c0-qT&Y0f|Aay@3|Ls!KmR*#Lm)}n76)@Rzz03-pLYMS+~bFg1k
zzg9Hk#U5J6%!8i1_#sa-JM6qT(t^G2^hj<3U*ZuGhh-{tlOMgh8Zu|4uJS~J>K(bo
zWqx^1D94IpsB~1S5`g}!RB7tNF-|@p!%xKT^sW6<BoDeYWRxUTW6j!voy^75f$$$o
z5jdCCO0&U#$FFpXxECIdAN7dNdEVi&9Thn4%6?Mp14tb~K?gI#DpX#ik++sDWSort
z!ctlW+#xd2r)^)wXdQpkgvRJY)h@rzAAEV*fp!Ie(A+$fcl|2x1B9|WZ`@0fgt(R&
z%G$71%*PUC<LJNf!&{+0Y+JH8r&?<`zh@uXgI#6ap706Us$%7*FR97R0U%yH^e}A6
zMaJJU!Z5?40%A_4+D<*n9`j1v+#u*@Z8nHUA~y$BXDrr9kn#@Tw)JJhxKsPtk50nF
z)rr=JsmIeo31@)}a;xHD0p<6e=cDol8s*M5C++qCB{}8wtfW|Q<~fvr<cCZ1j}NH3
zKYHSMey_{><f#$pAq-)WH(^S15mVCUlpWXu-?J+wSJcuN(<9Xe%u}Zk0urbbay#Ii
z4H^ET87UAAuSb5m-H2kISy8$a)VPbqO0Km-lso3K(5o0(gY;H9282~>ooy?wcP=tO
z*aD_!h1`k*y<0y*`eLU3sNmVD^I7#iWZr6nR}{}d0`a&=y=)$)vCsDqrHJ75ktVk1
zhj=`4Qb(^7@Qr5-v%y6cu=x-c<6}@CXLz2KT-+w@0k|UnXw=JN#ut(<if_w_n_{eO
z6M|QPHgui4GQ0UY^nSVc;*yU<iasi2_LUypgmN%&=(1L)0iWWsx25jJUC``UK|E{b
zUd>pgz3@i)wNbFD+yOdP+9mRauB0>b9(zn=3U28M+LFYHS38({%xy^jC~agR^qy8*
z9e!xpiGMsuC6cQYTFP5~!Rrus9fTx*H)MZA>_-!=H~=0k486<S?IFiKDI_vrkDv^Z
z_|TVn1CGbG_|~>~>bA6j<MKk#M)MuW`kyZtIWH^4zg_qa$XIupOe%;K{6P1-mVcJa
zPho_PU3erEs>7XN*ahwC<aKH*RXUMXWj>G1x2Fur2!M;e4dA}X@_YCJXu9%L<81#m
zJ#Tiq($#5N1yqebdf$k=nj0uT9Lei(^AHOMDEN{DN0)@|pW8nM?TF3Y)stg%g?~Y-
z2GU~_y2m1Ws!;2F=$ri9z<vPtpg}N}DUJUxi3M$>X#;A5_XmqI?<YW%Pb45c+5muG
zo5%vE$8|Ai9z-Z00tw9L@c6`0FVnj*QO*gcpvRx-YU7*Vs;_~RVesiGVBQ?UYFe>c
zKf_9S%^IFMknX30N4&tnPm(}d@WI=I1lH?}3YsN#wZoX4TkjUf^h|SP*W(mGMG)6@
zU3m#Yhmh8cTLng&28WVf_%y17VCAlENxp;X!MC>)S`U@9JTje*1~2BUKReqydj6WM
zj}sa}=ZsW5|G^<V8NNUOV#azIcWx|mfR2R03n~*)=}QBcStvXdXV$2L@Te%%%!^)l
zW8T3^PNO;qI*}o!bHgiS_AuKNJEn~-d&1vt;B9^%v%RmzwgFUvT*X{RkCkx=^}NRS
zP76>zKX`ZKJvNB|C3E{zCUQ_2zJ=>Y$wmDu!+8hghJTWpeBc>1_9g%ItN+mvMnVu?
z1(533y((a?eFOo9KSTNZm&1P{T5@S{Al<Wc;KF{B*9uX0wsza>S7`ejM&O<FqlX#_
z;8R{xz$laZAgq0;cRsla?(9CerQ2BYbTBhO4D&y8X}K)?pi1{YJz#7Vl%G0)`}w=T
zpPY@yx<4PzgfQrDK&Stcl$GLLb;A=1$Lz!SKM|7F-{csS94^Fn7zo{9jv!SmR!fx^
zwqMavmUE4(J9t7A-9um=dH0zOU%`I#;7k7g(+!ew|MZJ=4usgj*xxT$cAz&4#C#7u
zzMMa(K^$i3AYT-O3%l(GYUWS_-2TJv|01q3lr2`aUtK*!^uuyi|LeEt2;q6(JZWY+
zKpy}7PXuu=cjFsWF8>^;!_R{v1?=VG(a4*J|8oB+{_F2ZMB#BiE$}X%{g0R3|3R;d
zhyn)=PdIiM_Wuj^BmZ~93p85Xd)!|Ce1OA~xdM-io7Yxq{6EjY8w7tPtk=WA{`S9(
z+!3#FLO4zg;vdxg`?w#B!bPNTL>g3v`#-OSPh=2-RESLm9$;wy{@QZhFa!U)VE$uS
g{C~M%#`lhPF%R?M{`^pR1pc{kT}c`(dB^+z0I_1IuK)l5

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/documenter.js b/previews/PR3525/assets/documenter.js
new file mode 100644
index 00000000000..f93caf595e2
--- /dev/null
+++ b/previews/PR3525/assets/documenter.js
@@ -0,0 +1,883 @@
+// Generated by Documenter.jl
+requirejs.config({
+  paths: {
+    'highlight-julia': 'https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.8.0/languages/julia.min',
+    'headroom': 'https://cdnjs.cloudflare.com/ajax/libs/headroom/0.12.0/headroom.min',
+    'jqueryui': 'https://cdnjs.cloudflare.com/ajax/libs/jqueryui/1.13.2/jquery-ui.min',
+    'minisearch': 'https://cdn.jsdelivr.net/npm/minisearch@6.1.0/dist/umd/index.min',
+    'jquery': 'https://cdnjs.cloudflare.com/ajax/libs/jquery/3.7.0/jquery.min',
+    'mathjax': 'https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.9/MathJax.js?config=TeX-AMS_HTML',
+    'headroom-jquery': 'https://cdnjs.cloudflare.com/ajax/libs/headroom/0.12.0/jQuery.headroom.min',
+    'highlight': 'https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.8.0/highlight.min',
+    'highlight-julia-repl': 'https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.8.0/languages/julia-repl.min',
+  },
+  shim: {
+  "highlight-julia": {
+    "deps": [
+      "highlight"
+    ]
+  },
+  "mathjax": {
+    "exports": "MathJax"
+  },
+  "headroom-jquery": {
+    "deps": [
+      "jquery",
+      "headroom"
+    ]
+  },
+  "highlight-julia-repl": {
+    "deps": [
+      "highlight"
+    ]
+  }
+}
+});
+////////////////////////////////////////////////////////////////////////////////
+require(['mathjax'], function(MathJax) {
+MathJax.Hub.Config({
+  "jax": [
+    "input/TeX",
+    "output/HTML-CSS",
+    "output/NativeMML"
+  ],
+  "TeX": {
+    "equationNumbers": {
+      "autoNumber": "AMS"
+    }
+  },
+  "tex2jax": {
+    "inlineMath": [
+      [
+        "$",
+        "$"
+      ],
+      [
+        "\\(",
+        "\\)"
+      ]
+    ],
+    "processEscapes": true
+  },
+  "config": [
+    "MMLorHTML.js"
+  ],
+  "extensions": [
+    "MathMenu.js",
+    "MathZoom.js",
+    "TeX/AMSmath.js",
+    "TeX/AMSsymbols.js",
+    "TeX/autobold.js",
+    "TeX/autoload-all.js"
+  ]
+}
+);
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery', 'highlight', 'highlight-julia', 'highlight-julia-repl'], function($) {
+$(document).ready(function() {
+    hljs.highlightAll();
+})
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery'], function($) {
+
+var isExpanded = true;
+
+$(document).on("click", ".docstring header", function () {
+  let articleToggleTitle = "Expand docstring";
+
+  if ($(this).siblings("section").is(":visible")) {
+    $(this)
+      .find(".docstring-article-toggle-button")
+      .removeClass("fa-chevron-down")
+      .addClass("fa-chevron-right");
+  } else {
+    $(this)
+      .find(".docstring-article-toggle-button")
+      .removeClass("fa-chevron-right")
+      .addClass("fa-chevron-down");
+
+    articleToggleTitle = "Collapse docstring";
+  }
+
+  $(this)
+    .find(".docstring-article-toggle-button")
+    .prop("title", articleToggleTitle);
+  $(this).siblings("section").slideToggle();
+});
+
+$(document).on("click", ".docs-article-toggle-button", function () {
+  let articleToggleTitle = "Expand docstring";
+  let navArticleToggleTitle = "Expand all docstrings";
+
+  if (isExpanded) {
+    $(this).removeClass("fa-chevron-up").addClass("fa-chevron-down");
+    $(".docstring-article-toggle-button")
+      .removeClass("fa-chevron-down")
+      .addClass("fa-chevron-right");
+
+    isExpanded = false;
+
+    $(".docstring section").slideUp();
+  } else {
+    $(this).removeClass("fa-chevron-down").addClass("fa-chevron-up");
+    $(".docstring-article-toggle-button")
+      .removeClass("fa-chevron-right")
+      .addClass("fa-chevron-down");
+
+    isExpanded = true;
+    articleToggleTitle = "Collapse docstring";
+    navArticleToggleTitle = "Collapse all docstrings";
+
+    $(".docstring section").slideDown();
+  }
+
+  $(this).prop("title", navArticleToggleTitle);
+  $(".docstring-article-toggle-button").prop("title", articleToggleTitle);
+});
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require([], function() {
+function addCopyButtonCallbacks() {
+  for (const el of document.getElementsByTagName("pre")) {
+    const button = document.createElement("button");
+    button.classList.add("copy-button", "fa-solid", "fa-copy");
+    button.setAttribute("aria-label", "Copy this code block");
+    button.setAttribute("title", "Copy");
+
+    el.appendChild(button);
+
+    const success = function () {
+      button.classList.add("success", "fa-check");
+      button.classList.remove("fa-copy");
+    };
+
+    const failure = function () {
+      button.classList.add("error", "fa-xmark");
+      button.classList.remove("fa-copy");
+    };
+
+    button.addEventListener("click", function () {
+      copyToClipboard(el.innerText).then(success, failure);
+
+      setTimeout(function () {
+        button.classList.add("fa-copy");
+        button.classList.remove("success", "fa-check", "fa-xmark");
+      }, 5000);
+    });
+  }
+}
+
+function copyToClipboard(text) {
+  // clipboard API is only available in secure contexts
+  if (window.navigator && window.navigator.clipboard) {
+    return window.navigator.clipboard.writeText(text);
+  } else {
+    return new Promise(function (resolve, reject) {
+      try {
+        const el = document.createElement("textarea");
+        el.textContent = text;
+        el.style.position = "fixed";
+        el.style.opacity = 0;
+        document.body.appendChild(el);
+        el.select();
+        document.execCommand("copy");
+
+        resolve();
+      } catch (err) {
+        reject(err);
+      } finally {
+        document.body.removeChild(el);
+      }
+    });
+  }
+}
+
+if (document.readyState === "loading") {
+  document.addEventListener("DOMContentLoaded", addCopyButtonCallbacks);
+} else {
+  addCopyButtonCallbacks();
+}
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery', 'headroom', 'headroom-jquery'], function($, Headroom) {
+
+// Manages the top navigation bar (hides it when the user starts scrolling down on the
+// mobile).
+window.Headroom = Headroom; // work around buggy module loading?
+$(document).ready(function () {
+  $("#documenter .docs-navbar").headroom({
+    tolerance: { up: 10, down: 10 },
+  });
+});
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery', 'minisearch'], function($, minisearch) {
+
+// In general, most search related things will have "search" as a prefix.
+// To get an in-depth about the thought process you can refer: https://hetarth02.hashnode.dev/series/gsoc
+
+let results = [];
+let timer = undefined;
+
+let data = documenterSearchIndex["docs"].map((x, key) => {
+  x["id"] = key; // minisearch requires a unique for each object
+  return x;
+});
+
+// list below is the lunr 2.1.3 list minus the intersect with names(Base)
+// (all, any, get, in, is, only, which) and (do, else, for, let, where, while, with)
+// ideally we'd just filter the original list but it's not available as a variable
+const stopWords = new Set([
+  "a",
+  "able",
+  "about",
+  "across",
+  "after",
+  "almost",
+  "also",
+  "am",
+  "among",
+  "an",
+  "and",
+  "are",
+  "as",
+  "at",
+  "be",
+  "because",
+  "been",
+  "but",
+  "by",
+  "can",
+  "cannot",
+  "could",
+  "dear",
+  "did",
+  "does",
+  "either",
+  "ever",
+  "every",
+  "from",
+  "got",
+  "had",
+  "has",
+  "have",
+  "he",
+  "her",
+  "hers",
+  "him",
+  "his",
+  "how",
+  "however",
+  "i",
+  "if",
+  "into",
+  "it",
+  "its",
+  "just",
+  "least",
+  "like",
+  "likely",
+  "may",
+  "me",
+  "might",
+  "most",
+  "must",
+  "my",
+  "neither",
+  "no",
+  "nor",
+  "not",
+  "of",
+  "off",
+  "often",
+  "on",
+  "or",
+  "other",
+  "our",
+  "own",
+  "rather",
+  "said",
+  "say",
+  "says",
+  "she",
+  "should",
+  "since",
+  "so",
+  "some",
+  "than",
+  "that",
+  "the",
+  "their",
+  "them",
+  "then",
+  "there",
+  "these",
+  "they",
+  "this",
+  "tis",
+  "to",
+  "too",
+  "twas",
+  "us",
+  "wants",
+  "was",
+  "we",
+  "were",
+  "what",
+  "when",
+  "who",
+  "whom",
+  "why",
+  "will",
+  "would",
+  "yet",
+  "you",
+  "your",
+]);
+
+let index = new minisearch({
+  fields: ["title", "text"], // fields to index for full-text search
+  storeFields: ["location", "title", "text", "category", "page"], // fields to return with search results
+  processTerm: (term) => {
+    let word = stopWords.has(term) ? null : term;
+    if (word) {
+      // custom trimmer that doesn't strip @ and !, which are used in julia macro and function names
+      word = word
+        .replace(/^[^a-zA-Z0-9@!]+/, "")
+        .replace(/[^a-zA-Z0-9@!]+$/, "");
+    }
+
+    return word ?? null;
+  },
+  // add . as a separator, because otherwise "title": "Documenter.Anchors.add!", would not find anything if searching for "add!", only for the entire qualification
+  tokenize: (string) => string.split(/[\s\-\.]+/),
+  // options which will be applied during the search
+  searchOptions: {
+    boost: { title: 100 },
+    fuzzy: 2,
+    processTerm: (term) => {
+      let word = stopWords.has(term) ? null : term;
+      if (word) {
+        word = word
+          .replace(/^[^a-zA-Z0-9@!]+/, "")
+          .replace(/[^a-zA-Z0-9@!]+$/, "");
+      }
+
+      return word ?? null;
+    },
+    tokenize: (string) => string.split(/[\s\-\.]+/),
+  },
+});
+
+index.addAll(data);
+
+let filters = [...new Set(data.map((x) => x.category))];
+var modal_filters = make_modal_body_filters(filters);
+var filter_results = [];
+
+$(document).on("keyup", ".documenter-search-input", function (event) {
+  // Adding a debounce to prevent disruptions from super-speed typing!
+  debounce(() => update_search(filter_results), 300);
+});
+
+$(document).on("click", ".search-filter", function () {
+  if ($(this).hasClass("search-filter-selected")) {
+    $(this).removeClass("search-filter-selected");
+  } else {
+    $(this).addClass("search-filter-selected");
+  }
+
+  // Adding a debounce to prevent disruptions from crazy clicking!
+  debounce(() => get_filters(), 300);
+});
+
+/**
+ * A debounce function, takes a function and an optional timeout in milliseconds
+ *
+ * @function callback
+ * @param {number} timeout
+ */
+function debounce(callback, timeout = 300) {
+  clearTimeout(timer);
+  timer = setTimeout(callback, timeout);
+}
+
+/**
+ * Make/Update the search component
+ *
+ * @param {string[]} selected_filters
+ */
+function update_search(selected_filters = []) {
+  let initial_search_body = `
+      <div class="has-text-centered my-5 py-5">Type something to get started!</div>
+    `;
+
+  let querystring = $(".documenter-search-input").val();
+
+  if (querystring.trim()) {
+    results = index.search(querystring, {
+      filter: (result) => {
+        // Filtering results
+        if (selected_filters.length === 0) {
+          return result.score >= 1;
+        } else {
+          return (
+            result.score >= 1 && selected_filters.includes(result.category)
+          );
+        }
+      },
+    });
+
+    let search_result_container = ``;
+    let search_divider = `<div class="search-divider w-100"></div>`;
+
+    if (results.length) {
+      let links = [];
+      let count = 0;
+      let search_results = "";
+
+      results.forEach(function (result) {
+        if (result.location) {
+          // Checking for duplication of results for the same page
+          if (!links.includes(result.location)) {
+            search_results += make_search_result(result, querystring);
+            count++;
+          }
+
+          links.push(result.location);
+        }
+      });
+
+      let result_count = `<div class="is-size-6">${count} result(s)</div>`;
+
+      search_result_container = `
+            <div class="is-flex is-flex-direction-column gap-2 is-align-items-flex-start">
+                ${modal_filters}
+                ${search_divider}
+                ${result_count}
+                <div class="is-clipped w-100 is-flex is-flex-direction-column gap-2 is-align-items-flex-start has-text-justified mt-1">
+                  ${search_results}
+                </div>
+            </div>
+        `;
+    } else {
+      search_result_container = `
+           <div class="is-flex is-flex-direction-column gap-2 is-align-items-flex-start">
+               ${modal_filters}
+               ${search_divider}
+               <div class="is-size-6">0 result(s)</div>
+            </div>
+            <div class="has-text-centered my-5 py-5">No result found!</div>
+       `;
+    }
+
+    if ($(".search-modal-card-body").hasClass("is-justify-content-center")) {
+      $(".search-modal-card-body").removeClass("is-justify-content-center");
+    }
+
+    $(".search-modal-card-body").html(search_result_container);
+  } else {
+    filter_results = [];
+    modal_filters = make_modal_body_filters(filters, filter_results);
+
+    if (!$(".search-modal-card-body").hasClass("is-justify-content-center")) {
+      $(".search-modal-card-body").addClass("is-justify-content-center");
+    }
+
+    $(".search-modal-card-body").html(initial_search_body);
+  }
+}
+
+/**
+ * Make the modal filter html
+ *
+ * @param {string[]} filters
+ * @param {string[]} selected_filters
+ * @returns string
+ */
+function make_modal_body_filters(filters, selected_filters = []) {
+  let str = ``;
+
+  filters.forEach((val) => {
+    if (selected_filters.includes(val)) {
+      str += `<a href="javascript:;" class="search-filter search-filter-selected"><span>${val}</span></a>`;
+    } else {
+      str += `<a href="javascript:;" class="search-filter"><span>${val}</span></a>`;
+    }
+  });
+
+  let filter_html = `
+        <div class="is-flex gap-2 is-flex-wrap-wrap is-justify-content-flex-start is-align-items-center search-filters">
+            <span class="is-size-6">Filters:</span>
+            ${str}
+        </div>
+    `;
+
+  return filter_html;
+}
+
+/**
+ * Make the result component given a minisearch result data object and the value of the search input as queryString.
+ * To view the result object structure, refer: https://lucaong.github.io/minisearch/modules/_minisearch_.html#searchresult
+ *
+ * @param {object} result
+ * @param {string} querystring
+ * @returns string
+ */
+function make_search_result(result, querystring) {
+  let search_divider = `<div class="search-divider w-100"></div>`;
+  let display_link =
+    result.location.slice(Math.max(0), Math.min(50, result.location.length)) +
+    (result.location.length > 30 ? "..." : ""); // To cut-off the link because it messes with the overflow of the whole div
+
+  if (result.page !== "") {
+    display_link += ` (${result.page})`;
+  }
+
+  let textindex = new RegExp(`\\b${querystring}\\b`, "i").exec(result.text);
+  let text =
+    textindex !== null
+      ? result.text.slice(
+          Math.max(textindex.index - 100, 0),
+          Math.min(
+            textindex.index + querystring.length + 100,
+            result.text.length
+          )
+        )
+      : ""; // cut-off text before and after from the match
+
+  let display_result = text.length
+    ? "..." +
+      text.replace(
+        new RegExp(`\\b${querystring}\\b`, "i"), // For first occurrence
+        '<span class="search-result-highlight p-1">$&</span>'
+      ) +
+      "..."
+    : ""; // highlights the match
+
+  let in_code = false;
+  if (!["page", "section"].includes(result.category.toLowerCase())) {
+    in_code = true;
+  }
+
+  // We encode the full url to escape some special characters which can lead to broken links
+  let result_div = `
+      <a href="${encodeURI(
+        documenterBaseURL + "/" + result.location
+      )}" class="search-result-link w-100 is-flex is-flex-direction-column gap-2 px-4 py-2">
+        <div class="w-100 is-flex is-flex-wrap-wrap is-justify-content-space-between is-align-items-flex-start">
+          <div class="search-result-title has-text-weight-bold ${
+            in_code ? "search-result-code-title" : ""
+          }">${result.title}</div>
+          <div class="property-search-result-badge">${result.category}</div>
+        </div>
+        <p>
+          ${display_result}
+        </p>
+        <div
+          class="has-text-left"
+          style="font-size: smaller;"
+          title="${result.location}"
+        >
+          <i class="fas fa-link"></i> ${display_link}
+        </div>
+      </a>
+      ${search_divider}
+    `;
+
+  return result_div;
+}
+
+/**
+ * Get selected filters, remake the filter html and lastly update the search modal
+ */
+function get_filters() {
+  let ele = $(".search-filters .search-filter-selected").get();
+  filter_results = ele.map((x) => $(x).text().toLowerCase());
+  modal_filters = make_modal_body_filters(filters, filter_results);
+  update_search(filter_results);
+}
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery'], function($) {
+
+// Modal settings dialog
+$(document).ready(function () {
+  var settings = $("#documenter-settings");
+  $("#documenter-settings-button").click(function () {
+    settings.toggleClass("is-active");
+  });
+  // Close the dialog if X is clicked
+  $("#documenter-settings button.delete").click(function () {
+    settings.removeClass("is-active");
+  });
+  // Close dialog if ESC is pressed
+  $(document).keyup(function (e) {
+    if (e.keyCode == 27) settings.removeClass("is-active");
+  });
+});
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery'], function($) {
+
+let search_modal_header = `
+  <header class="modal-card-head gap-2 is-align-items-center is-justify-content-space-between w-100 px-3">
+    <div class="field mb-0 w-100">
+      <p class="control has-icons-right">
+        <input class="input documenter-search-input" type="text" placeholder="Search" />
+        <span class="icon is-small is-right has-text-primary-dark">
+          <i class="fas fa-magnifying-glass"></i>
+        </span>
+      </p>
+    </div>
+    <div class="icon is-size-4 is-clickable close-search-modal">
+      <i class="fas fa-times"></i>
+    </div>
+  </header>
+`;
+
+let initial_search_body = `
+  <div class="has-text-centered my-5 py-5">Type something to get started!</div>
+`;
+
+let search_modal_footer = `
+  <footer class="modal-card-foot">
+    <span>
+      <kbd class="search-modal-key-hints">Ctrl</kbd> +
+      <kbd class="search-modal-key-hints">/</kbd> to search
+    </span>
+    <span class="ml-3"> <kbd class="search-modal-key-hints">esc</kbd> to close </span>
+  </footer>
+`;
+
+$(document.body).append(
+  `
+    <div class="modal" id="search-modal">
+      <div class="modal-background"></div>
+      <div class="modal-card search-min-width-50 search-min-height-100 is-justify-content-center">
+        ${search_modal_header}
+        <section class="modal-card-body is-flex is-flex-direction-column is-justify-content-center gap-4 search-modal-card-body">
+          ${initial_search_body}
+        </section>
+        ${search_modal_footer}
+      </div>
+    </div>
+  `
+);
+
+document.querySelector(".docs-search-query").addEventListener("click", () => {
+  openModal();
+});
+
+document.querySelector(".close-search-modal").addEventListener("click", () => {
+  closeModal();
+});
+
+$(document).on("click", ".search-result-link", function () {
+  closeModal();
+});
+
+document.addEventListener("keydown", (event) => {
+  if ((event.ctrlKey || event.metaKey) && event.key === "/") {
+    openModal();
+  } else if (event.key === "Escape") {
+    closeModal();
+  }
+
+  return false;
+});
+
+// Functions to open and close a modal
+function openModal() {
+  let searchModal = document.querySelector("#search-modal");
+
+  searchModal.classList.add("is-active");
+  document.querySelector(".documenter-search-input").focus();
+}
+
+function closeModal() {
+  let searchModal = document.querySelector("#search-modal");
+  let initial_search_body = `
+    <div class="has-text-centered my-5 py-5">Type something to get started!</div>
+  `;
+
+  searchModal.classList.remove("is-active");
+  document.querySelector(".documenter-search-input").blur();
+
+  if (!$(".search-modal-card-body").hasClass("is-justify-content-center")) {
+    $(".search-modal-card-body").addClass("is-justify-content-center");
+  }
+
+  $(".documenter-search-input").val("");
+  $(".search-modal-card-body").html(initial_search_body);
+}
+
+document
+  .querySelector("#search-modal .modal-background")
+  .addEventListener("click", () => {
+    closeModal();
+  });
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery'], function($) {
+
+// Manages the showing and hiding of the sidebar.
+$(document).ready(function () {
+  var sidebar = $("#documenter > .docs-sidebar");
+  var sidebar_button = $("#documenter-sidebar-button");
+  sidebar_button.click(function (ev) {
+    ev.preventDefault();
+    sidebar.toggleClass("visible");
+    if (sidebar.hasClass("visible")) {
+      // Makes sure that the current menu item is visible in the sidebar.
+      $("#documenter .docs-menu a.is-active").focus();
+    }
+  });
+  $("#documenter > .docs-main").bind("click", function (ev) {
+    if ($(ev.target).is(sidebar_button)) {
+      return;
+    }
+    if (sidebar.hasClass("visible")) {
+      sidebar.removeClass("visible");
+    }
+  });
+});
+
+// Resizes the package name / sitename in the sidebar if it is too wide.
+// Inspired by: https://github.com/davatron5000/FitText.js
+$(document).ready(function () {
+  e = $("#documenter .docs-autofit");
+  function resize() {
+    var L = parseInt(e.css("max-width"), 10);
+    var L0 = e.width();
+    if (L0 > L) {
+      var h0 = parseInt(e.css("font-size"), 10);
+      e.css("font-size", (L * h0) / L0);
+      // TODO: make sure it survives resizes?
+    }
+  }
+  // call once and then register events
+  resize();
+  $(window).resize(resize);
+  $(window).on("orientationchange", resize);
+});
+
+// Scroll the navigation bar to the currently selected menu item
+$(document).ready(function () {
+  var sidebar = $("#documenter .docs-menu").get(0);
+  var active = $("#documenter .docs-menu .is-active").get(0);
+  if (typeof active !== "undefined") {
+    sidebar.scrollTop = active.offsetTop - sidebar.offsetTop - 15;
+  }
+});
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery'], function($) {
+
+// Theme picker setup
+$(document).ready(function () {
+  // onchange callback
+  $("#documenter-themepicker").change(function themepick_callback(ev) {
+    var themename = $("#documenter-themepicker option:selected").attr("value");
+    if (themename === "auto") {
+      // set_theme(window.matchMedia('(prefers-color-scheme: dark)').matches ? 'dark' : 'light');
+      window.localStorage.removeItem("documenter-theme");
+    } else {
+      // set_theme(themename);
+      window.localStorage.setItem("documenter-theme", themename);
+    }
+    // We re-use the global function from themeswap.js to actually do the swapping.
+    set_theme_from_local_storage();
+  });
+
+  // Make sure that the themepicker displays the correct theme when the theme is retrieved
+  // from localStorage
+  if (typeof window.localStorage !== "undefined") {
+    var theme = window.localStorage.getItem("documenter-theme");
+    if (theme !== null) {
+      $("#documenter-themepicker option").each(function (i, e) {
+        e.selected = e.value === theme;
+      });
+    }
+  }
+});
+
+})
+////////////////////////////////////////////////////////////////////////////////
+require(['jquery'], function($) {
+
+// update the version selector with info from the siteinfo.js and ../versions.js files
+$(document).ready(function () {
+  // If the version selector is disabled with DOCUMENTER_VERSION_SELECTOR_DISABLED in the
+  // siteinfo.js file, we just return immediately and not display the version selector.
+  if (
+    typeof DOCUMENTER_VERSION_SELECTOR_DISABLED === "boolean" &&
+    DOCUMENTER_VERSION_SELECTOR_DISABLED
+  ) {
+    return;
+  }
+
+  var version_selector = $("#documenter .docs-version-selector");
+  var version_selector_select = $("#documenter .docs-version-selector select");
+
+  version_selector_select.change(function (x) {
+    target_href = version_selector_select
+      .children("option:selected")
+      .get(0).value;
+    window.location.href = target_href;
+  });
+
+  // add the current version to the selector based on siteinfo.js, but only if the selector is empty
+  if (
+    typeof DOCUMENTER_CURRENT_VERSION !== "undefined" &&
+    $("#version-selector > option").length == 0
+  ) {
+    var option = $(
+      "<option value='#' selected='selected'>" +
+        DOCUMENTER_CURRENT_VERSION +
+        "</option>"
+    );
+    version_selector_select.append(option);
+  }
+
+  if (typeof DOC_VERSIONS !== "undefined") {
+    var existing_versions = version_selector_select.children("option");
+    var existing_versions_texts = existing_versions.map(function (i, x) {
+      return x.text;
+    });
+    DOC_VERSIONS.forEach(function (each) {
+      var version_url = documenterBaseURL + "/../" + each + "/";
+      var existing_id = $.inArray(each, existing_versions_texts);
+      // if not already in the version selector, add it as a new option,
+      // otherwise update the old option with the URL and enable it
+      if (existing_id == -1) {
+        var option = $(
+          "<option value='" + version_url + "'>" + each + "</option>"
+        );
+        version_selector_select.append(option);
+      } else {
+        var option = existing_versions[existing_id];
+        option.value = version_url;
+        option.disabled = false;
+      }
+    });
+  }
+
+  // only show the version selector if the selector has been populated
+  if (version_selector_select.children("option").length > 0) {
+    version_selector.toggleClass("visible");
+  }
+});
+
+})
diff --git a/previews/PR3525/assets/extra_styles.css b/previews/PR3525/assets/extra_styles.css
new file mode 100644
index 00000000000..1f7df51cf0f
--- /dev/null
+++ b/previews/PR3525/assets/extra_styles.css
@@ -0,0 +1,4 @@
+.display-light-only {display: block;}
+.display-dark-only {display: none;}
+.theme--documenter-dark .display-light-only {display: none;}
+.theme--documenter-dark .display-dark-only {display: block;}
diff --git a/previews/PR3525/assets/full_sudoku.png b/previews/PR3525/assets/full_sudoku.png
new file mode 100644
index 0000000000000000000000000000000000000000..f3f84a9c00620b3b238d5ce2dba27c295693ddd2
GIT binary patch
literal 119423
zcmeEtWl$aMvMwPB?(Q0bySoMr?z+(6?g19=p5RV!2=4Bl;10oEgWCe`<lFn~{q1va
z)%|h)o>et9vu0lH*SeqXCtOKE3K@X_0SXEVSw>ny1qup!8@P7iV1PFWHvV={P)OF6
z;^Im&;^JgVPWGQHZOovcq{9<6VYO6;ak6y3#KlcO=Ssu3!%vkZpy|AMRT2;#_Xa~Y
z3SP7?CIX>CU!wY38AU|J+qe%Oj?I*heX4V6YM5}aKJ3U#i_H7%c<nvzj`@yf+$%P-
z->$wAsZFMj-~OfpwH<z0aE?y?qq@ZMqY7*e6qb@dw!_|qDTaog9)};>!i)au+MBR%
zoxAz#4|{jM*u)(o1Fxa*Q5|7G1EYaW_)y-j6)Aqg`V~*;D*Q~Q`Arn&TEHR_rb+LR
zgmI>_jzncgP$d&(F7Evnijyvx{wK7^me344qK9^z)E0XmlW~VMlHaaF*lSNQCQ{=M
z$Gw*q@97=h_PI?MK(j`{-kEw2re1SDN0Eb>xUJ#nFA31<mEu2s8O^-$P1^O9GpO*M
z2t~G0%jIA*h!_68ags5tjPNnINiG8&v#gl2ubrZ!<`gOiD_4L`#xj=R($q#vxXW)L
zcpN!8U;%H_!O`wxooseIEyA7%k|v@K&B)$6F&n6J3V#L$QM1|)e7pubh~=XRHQIDf
zycI58l|A#~o(dRp{V$R<>&17h(G*6;2z^*#<p>hPtZ3FE#|)z~VJeGJ3>Y;gFt+&@
z9v>hYXoHA-=;X5C?_{b3B$YMHKffaHM%toB@UYT(lYvt3O1`7IMGO*$k?scQqlanB
z1PVYeF6?*f-33grn4K;l`t#6=_D|CfqlRma_0xV?;Yk#<BOi)Z+GixYiNk(Zyt_Gk
z@8C#xfBsPhtoMygp^YY;pH2ZEwS!3FH5c65k7S{YuRp>pip(sjuf07>pxp37%qK&H
z^ZP2x&O>QtD1lT7p92ShkB1rx*$qaDLJ)a1<BLB<SE)eeY*=S#D2d_$<46()-Td1?
zh(%6dn-O0im|#-4&Dm(H+r!yNHR{FdDd#3V<SpkUU9|TErK<D97d%NMDoKpq@S7%d
zlCa0E$=?J>>4c|&k0`t6KeK*EO-OotY!bDgUP*mNm-LLbV=;kp->Y6fbk~8?guH##
zC;l612qo1K#_mmq1L}TgFs$f!meQz3e|a&&CG2UZ+3#P|N{jLOLp0jiL?0U-9ScJO
zL-x=jzs;@9OW|7&D77z&Qwl$r`)c|K9p{1T`f!Y3t7{kFj|Jv}W?#Q|BeL#dGLpjf
zuy7S5*3u7L-Gtp7_80i(d$(2UO-D!TT;h1(a{rxZ6c=lko{Pk*#xzYWf$E2O&HB7(
z*LIlaSZ}HwlU)TAO$X=5{Jic!7xx<@y}LqsFsukvC-6H+XCgSg7Drsa@x75Z_d@D;
zfAxw-bKt8SVgjF@BY9V2(EAn3kS;7OX#SK~!r#!(WOzaT%49h20@PU0HMHU8o3TE=
zW$)5cf*$E|s)G3ro8HY;g?RiHRz%POnz{?$0bTDEcb8z(XTC1E3FOs4p*Nz^aLBoo
zAz}E&(L5A(VW=2kLFALt0$(EXDYe7!;wjnVJf-n#FtlTBK9t6T;wXpUw*{s`Rzyk2
zGlvMbF(Y!-)NnaPZ70x+QAfn`CL~;-(}ER5BPZ_nRqCM2#cp$4_u;Q7KEJk=0(}$2
z>zZMO`4q!zB!ov8**{hF@ngi#9@%QA?+9i61RKsBA@wHu3sM<j(~j>?-)K6#%L=;u
zO?$`S4c|es7CgUE^xOK5vW@UXHU~WlKIV-ubSO(w+=mod8uA28H#8JXh2HiVQ4hs;
z82A#p-?&#IOZsd!T{bZ`B{zkP*{h@Pf)>On;*n%x@}X21!sSM&6XWLMZ{lLUR(#=;
zZnrwL`ebEiwP9sxWn6Du?^OTGYH+5aR8aGX(mK9;7`&~;mMsw77S)!yBHi)vl<Zcy
zj<z8IN|u5uoh?~Kw4@-TaHXhCT~o7B=~kUIk4Ysv&mgx0^px>Hc^#`pT>3guo>(za
zHCe@7CB4MG*xr&O4Vf_|NpiG^u|R!VVTxqxGv_|R7y%J-$AC?=Y&2hV89^n-aN1lN
z!_V-v4{1ATu$;`+ZkF-&HgyX+dga<Wi8@ksc9z!`)AjoGCiRzf$rfjGLX}Tqq4_3x
zn1#Lbsg<?zd--FSe3Fg%wFd$Z0uOEvmY+kJl;<^)^U}+83tKfC)f;scrN+gXwJVh>
zRnCeH_+OhWu<%6b4X_Qo8jv5L_QCW?>Jag*2A4mpKcT|(A_XA7fhmA_M6$#4W-}!S
z!QaKJVENANZ5Twzix<tBW>q)5U#C3J^2_MC*P~Z9lo-!#9*@(FC5Ai8uI=aD7suG}
z5}BazKGV!$nXgQ$A(mB^4F^`!CWV=wTxvzw;}@-Jw?6r@T&FO6VyX@Bj99d_vmvo$
z;$E{_wOTbdvW_=Zwy7T9^rz~d8f2SxsyB-?YfMa?wJ5{a;M8z0AD$DM6FaJ(GcOlj
zSXjzmV5lo;^fa<I9Ia-oDX(s~m#RgqNjKIsbN`iPab}Wj(mDAY@%nK6fXt=XrRg|~
z?=x=%5jS50GBwgIF%yXaCnO1THhf2IU2W@XD0PGAMDJMd3}O{D8&-!-FPTAnKHEP{
zHE-5BDyT_3uUyez5nW-wLbT#xmuNTt%kp&NwDuR_3doD+Gn@dI=gLjKSImvx{^I6+
z|LC&Zq59G6f$dJ_X?yo$Qf!4{#yCc{bH|%Cw6(<!`3};jHps;j%H7k>*>?6;=kiOx
z^m2Sa{IZk~d==a}j195{swqq%tR1`!DjVu1jQU$alm?_-*fTirSdF<^?M;<Ia8r=9
zI7+Cu;bXOVO}cpXMA*dC#2iazResgA5o#}EPi>ENPb;MWWuX*@6b3~eg{H*nW{QJH
zhLyrC&6zU0RA$6CNh*o**v~{KXpGTzMA0!MoS$d+A6e&EH&Y(8&_=`cu^-9Cxa?$C
zDow&nHV4#VJaNtlgejkGpaqRodQO9G!^y)(G0vzKsfNGmeBt~$NUcfjD-AENojN7t
z*0+#q!dysD5TQaiC2cuu$v^dZAUIl=y~3nz@9xVvH}!%7O@;&b+hpHLxD>Ue!lnXW
zih{JF%v$DiD&c7P=8dTX8<(le<w9G1TQbD(*$P}lCQ_yHTY#O+*=#rRyWe+p?aUrB
zqvk5PKAdvg_OX~Bs~Jz^)LiD~3chx-ryT<IZC7oS#>9>>cQdBb!@*yd-^F~vka@)F
zq_r+|k;}{HHx081_)ZSOeVze*Ssun6BE^1#qlR;(AfNTQN~u*z&%-jUx2}IW3pW_|
zNq{(eEPa)&*Lz?EKlW42C%eqxA0sKVY|BnJkVRr=irwQhxzu#_+eYK$&BXq*-GU$1
z8TM?xV-RXY3Y?h)iv(#oJH{(elP7wIbX~D)_2Z^%WM|Ta9CJc*LSnLyL65G}2K9V4
z9rzHG)?DYBI$t9{8a~=TYF`1TQ*B*vGx9A_Q6t(wy)mu1#pSw{^Xps&^DWI<TT+ci
zM?*VB%QvIU*_NFnp-S9}PGx+RX8XYNkbXiTn{E5H7FrM0lTTlNnxzZdqT4pNa~fOg
zC(5heb8R*R*p4?)+Y3L|j*rjsN4X`O+|{VI)%s{cDpY4zXJ=cZ-I>cot+zQdy#~&j
zYq(wP#miMs=6s@j%pSYJ_s_y$nU|ei|Kit-us$erBsD^H8!S%_mc%OD<rS?4T5GFu
zOA~IUGx151F9W@ud@n7v!)+xO#jBHMgEWJ?q({OAE^S5HQ!B0%M}1$DWEgK1Vgx;(
zW>Qqki}Dn*`7yz}+uC!BTYJ!BoSC=$zS{xg!4HuJxJOK$27Hk8J^N7IcAX<8FOXf+
zrdL6=t?6=gnW(0s9%8%OF5limKOrmklHL1XZ9dkJ@|$JIGV|&7s;5umy4_kk?=AeT
z9I1sMjjsiy>d^gh?Y7>lp412HVSj2XXz(`ccUGG>o!3LN$us*>?Zw`w@iI_TyTu{W
zS=+^6lQ5rwtig?IBBbPLFGSO=66FaSGGB07{foG=pz?8z!PENAZI3_Jol($H5Or;@
z{p?X`$$FvnC~thQHd{qV@GSiy^{RO@@+o<sQW<o1yLWvn*Q9dVRS5+{n3f9Pm7WPT
zq4FxH*O|~P<AQ;QEbIcp=hx-iGjh4<=kZDjk-y(wpcUWmscP|u|4KI$4R`}H@YUg|
z(2N?zsEfO*lk$d4*u?VM%iJ3%6-p>j{?n5W;p1bcO=$P(s~pdW3#3vs3ITQ-c(oit
z7)*q2)NcaW;N-1XK_!XxSK#B9xBd>XCU1F4|5zUpW?C|z<mI90fM+<UH?Ihw-U3gr
zfJ^8V;lG|GU(rCl{_8z76jZ1s)SG{%Q2_3LesRF{C(l3bufK#q!2<s<fXgEX`oB^`
zZ|A)J*Yhs$4N6p1Tt)`CtC~2Onb|p8*t=Z6MkoYcz&l85J3~R?Q2n`H$*8>h4U|7;
zsix(kCI6An#NL+4$kg80jLE~+;ZHeG0v>$8qpg{X5t)arjh!=}hakmYDfocrKd+f7
z$o@*=Vl7CaC9gy#ZtrA9#>K?K#6ls2Kt@I;;AHxVPenrVpV@(bf)o}mE)IOm%<k^)
zOz!MV_D<%^th~Iu%q(ooY;24`3Pxv7I~OAlMmuN9e-!dx<w%%0n>blIxLDfTk^L#x
z$k^W1MUaBxPeuRw{iB^`9+v;E$<Fzo-2ytu{O1cZD-#Rzzsd%(3jBG?r)242W}_`(
zX$!~<s6&XGgH7PC^#9kFf7kf8oLYa+$@2G{fBW*^bE-R=If>ib0#&*Q{d>6nnfdP@
z|Cv#M`A^UPCW(KD`LDNtoP`hsnEy3qLI_UBNg=>65?M+pssVRklKuI;0;T{I%|Gsc
zo;SuJ$ZCX8P$Ez=5~6Azua5QKwqf_*3ZDrYoRdUjxRD)>cfQ4=A;+OJiZq^Z%SkZ|
z@Tb6|F*Y(GBZD!f&&3PNq?!8N@_6bywkdQvcf9KCes+49b?<%pE0f3Eyn}u3$Kql*
z)(0NLK{yuZC@9LVf8YGc;(A~_$(k<3@n1>$p&P!z`>z}JE3Tnt^8cRx6-5_}r{DTz
z-hWRHBu4%pCN{K=E6UG*lfnq*_<Iw92LE0Y>#}Qf<QjU>z5Y>P^zTLjO@^a|wuuX^
zGv?9#&z=qYLlJHdE062R{zt$3<NowOnm(kq|KCyoa>fzCRf%R10mYmC4~sW6j|&W|
za`{FSacnNr=g;>BMd;3uI7GEbVYa_kDyD0p+Qjt^nM6QKp{qI?ZpajOYk`-t>*6RX
zp5g6iB@(eG(P$+RDR(tPr=3!bV(*1@Y_uwpI@s3h6P<Wb)O%F{`k*_jbZls1;Uo~B
z8v)g5WyqV8kqfzA<Pt6L!GdK1Lni@79sMXXGI##B5oW?|mmkEm+T=W9@8r7>L=j>+
zzLcWnfBeeuWrw>1!Ti2pYw$vDP$AFsel{*gUmI-MM2@WvUL-3rIh7tp=tifogw}vl
z@yL%?3nnv5GmU5HJXhx#W&VHEUVJeM!%OTeC944!gC>JnB^K918J$n}hIaa0r0IQh
zWsf%a;I0}nRXt6k4d#4jp_v3Cu#^6f2;v$#Na0th5pxKVPV>VKz#_4Hr%o=N`d<6p
zsJ%D#7|^`fo#dEI$>EDKa?#cr$drRPT`8nSRAF!>{tI>hrNq>m?`ZkIr724gL`OJ=
zDYU^`We$LFzzm%(#~iXrAY}NZ#GT|2LKWqmWMn74-I#>$Xhh^+-Wvv!1xpdgsKcog
z%HRxN5Q2hhAyelj_XQ?J$~6!x>4^%4PL7qh2)a(2xM^1ReAD1{xOk3N^lxYdrld>7
z;@ISCq?6X}=rVxjkpeQ|kN&cwDIVZvc3-d&qNa9l7MKy!Oi4xmMk*{HUW;6MBVJMk
zJ0QXG%ZD1sfaV6!JjnkJ+5dOQxOYItqLuEnw6qUUf%Zr2$o|K(<P+t(+*;MfA50cz
zN>oKTE+l;^?5jGsY!;x>x$Q&R+5{GA%pyNCGBJhRoGgnMDP+2I?*2@tX7frKZ+SQ@
zGg_*%jCKr!MGd*X+~a<@A0Y8Br0A42qvC)(!AeM^M{V2tON$ESo0^MTJ)ZR2x0cTn
zyN4u1AwBjyU}3($zb*;us6j4(ND@d8(xaW&*4~BRglO70^YZF0FS!wux$;lm^kFX>
zDKXVu=z>s@b-6@MF9*sy<q1X6jR-zj@*l%l@}I%18w$UJ)&!=I5=uIRu7FB+c@>Ae
z_HK0f-9zw4b&bY`PTw6aYVM8zFA@2~H6?;=b#?WLB1KYE!^ztHwNBv`$3fx;5mFyf
zx*+Ju0|rn>+xhKIGDu8WIlgSl*o5zTM%Ab<4BY{8J}9oElSa>1i5V<U&XZW_v@5RE
zdWAV^nmU*Tx40Kw6Lm=$SBeH1ii|H(QAMuL2E%clb->q;3cW2ZV}Cz+ugX?S_OwnC
zw_3|>O0-Hi<+d!2dzMTMckgpCOucbFNP1f)9qo<Ew@;SB>-;HiEjFEH*xq#Ob1~Qi
zMQuE@oM7E<Bz<xZZ-~zo&h+Z|ZCsibF&@{w?LrkwX_haB6;2z9tZO1xM_s0S@DKkY
z(Yc)InPehMV^Ye%I8f21A^-Qsd)VTa9jWW0gv2!IG?0MCKJVl%w43N?rHbzBYx)ih
zpFVS?);mfIt=c7QiUU}e!0)wVfj_{5uk(vqzn$};tAoWY<22ROMC>YV?RodIyqE-D
zAeROv_M^-+!49L0ja~UNUmf_q*siWfx|}RwzgNsc;^E7*z^xMtD`<0FEBE`NEDlA1
zF2|;Z);#~q3V%<EOr88LKG`CRle7&Hd-^nBs1SLwoAp;?lQV(o4JjTbGL<9R1;68_
zgvkB%^7ZL2LZK(MHl+Lg9<okZOIEcX9#EyAY~+av5jxwYEPe9ZrF8Q4N27P25tOyI
zHA}wxjZhbg&mNl-fij|ZKB7}5@@g7XcFeqCH6{syNSgS8&=2#X?Av0o_)#P3o19B0
z>;u1PzjY!FgCO9S>vyOj%`x~b#jd@E1jxFU5c7$SQE7_rhWEJ1$QRU*PyV$TVyv8o
z1*pI!>p8*yNMY^Of3IIi!2T&}gq_nLnIA299V2$%^zJ>#`$oh<>q*n4>QrbE*{L8O
zZ{&S+=u#}sfkUo`&I;>WHbuTvF#=3^w2O3_K%!3X_Q%s!)(v#)&77S^qbD62e0GkC
z<j#GzJv~QaQxMC0zM)1htI$)k5CV<LfT^+pk=`ZWP2+S&7hem<<@UHQ>y~k%LzOUU
zlfzP99pyd+wq3Q1=G=(eU#o=N)hrobTm>}<|2YCGWa3cgbFczPVR(C-?{|`W#Zd*f
zU#rY91fEWmJDZ`Dd)x^Xf0lf{7fwaK2=RRmnHrUXB9J!V;TqbKLNDr%A`#>ROJbmi
zjxcvX;eZ^{R%kF(FCDPB>(#AUk@Iw-3~=~jzBM*I84M9Ql;+Q@N+|QY*k>^bUD>X_
z>EFv_$X=x^JkM{ig^f$1oAce4sg@J2;;-{I#mP6NvNHfHJu%OTb}Rpa+7nRKQz6{e
zjH=1Jk?B^+>!uZ9B=vyK(Pn2Hnx|I>rzePZ08=u88ANMJ28&GaMz!!G`ShSHUV<14
z!8WgPw@NID;QGt+*-MV|-dGDr;$yW|?I+~#`mHz`l?KT69ry1cSJR3OAAAl=ie{R3
z8sG(@;A+I(b&()Q+5*QemgYfiUuMZP`EL*v)Nv=YjnbxSI601{)FC$^b?rAAl{{z6
zHAaFQHhQo;do|CoYdNROl~7jJ>NqcEfslA95`S1y?+FAtl?9OR6V)fH{UWOzp{}>9
zjJgf)&RyiNQXml}p9@Tw)%G%VcChCuO_7_S*gg}%EIborc9lHef^v3G?u|R{%aXE3
z9nrOqaCb$QNW3_|+&gZ~jC-;YPjampZ&(lS=PET`SUY1F1(_rByk(2who3?{?D*M*
zb0l)u@uHFW$!|LFcUkVFS*G{<?N#q&)&*`~K*l&>$WA)-NlW<K-xRy*;eWa<V@|pK
z64#eXcz>EJ%E0z4wa;_z+G4lEt#td-3C-D+ku=NahGht!yPIw7NTd|PMH|?r0OwQt
zrhUUhSyYrZNRGK3tK)7H(N*+4IPG+?wy4T+h1t&I@uYEMguXIJy-Xw3uuB2bxCRc4
zA?8opY(s(j4oDkqk|c6W5TD=~QXUU26>#|Spc)l}#T~LCuR^fBX;!%{Hd!4R%qtlh
zqBPd8DLQvx-m1eA8kudi>P1+(=1p8Pj)myvV30G;M00;#s^$B9PkmfJZ%|nOt~UfZ
zP`mb%=QW<LkmxXTg(^+!Gn;dV#M<O$!*HBufb()U?(mc~lM3!CK7Ik|%xfJ%+THxe
z{bJ%4T<y}&8<wt_q4Bt<-ZPu0uzk=+;!(<czU>pCX2yrF!1&X=tSNI3!S@q3XD<f<
zPZw`!JKbP-!O9j5<KRPP*T)zo@chrh%}cm@YF$?JleeG;2Y-iMR_c~CCzZxcIqd0>
z&Yum46($y(TfEX|j!p2+m9PJ(gU$1t=>5R%s-&v=IyE&_nK%5&ZMsn2;cg?eEN@rx
z&~a;^+voO|InvDvsaoi116fm<?9P_5sz=|Px9z7+cq+=!6r8hnd7=h_FX}Ye9$(?d
zK0OjcF`jv0sWhG$MmS@WHE7HA6%FQW&tt<dVMLP(lbp7U3Q*iHnrDkv8no*<gfYt~
zUFZt9u*|$zJ&z=_-$)J#x#u5=ZE$5YalKQ~GW;eQNQCU%NmMT$%Pk!LO^k+M1vMeU
ziI{!`y5>64USx@>^{D4WAWZu_?v?R~To#jzWOiQe&P-|xOQ!&^ZxC$L8oRUoyql{i
zA?D}C!NqOB&6;VQto-$^UG#aDZ3yEv)1tVP*wU#nmVjhvD^#MM2h+^UU1oSq>K(%J
zLjkwXTGH1)8fJRwk5)f2>;T+?E?dl4o=v!E>N8@ug;#Usx_kYU`2qol69uUc+`!@l
z^%IS%y6~fN;8J${l+uR@mb#Vx9H>Xj{911zh}gTXmzZ8hn4Nk;Fe~qpveka%B63>I
zZP5%@s}tN_5yoTud{o&vz<I4WsI&<$26oI8o6O1k*xW;_Tn^GfyR<{R`%nmJIFg4z
zo+0ul6Y}`ctT`8xB}(@*dGAXzMh8W2sv<vb)IhD!=Gj1IAKn=!)ch<%lnF`7imzps
zzVG*`$%G+VAmg4lx^B~)VI<81>HR3l5J!9YR;y5#F}QYko=n!B4xx{|#N4JGx}e|v
zrp<TGlM!a}So`jRrmC%N<)YbmBtlA28Y%syN5;swReZDs4HoGh#n?g-p{!CfA_PU_
z0%bMtT+?F!K0-ye%vfcj(#xbNsnF)awqpzX2`cs2IAg71bF$<;(nZTj$;?yL)R~mm
zKqu{JWs2J;+D7*^pGNaL9##Y~u(gnmshzu+LfLNHc7pzq$K>Aj?a8N2ZGi`(qbrhN
z7>Z8hYuRqc(DV@<hvQt?{1&>KFCQpmH&Ut7?h=3b$Q`%AVHqW_g+w38FBtkNxBo^K
zvuvjZ<l4^gVdqj&dsA2Fr!TQ|B4{})bz}ljWRdI^`h&ekn7%p#|4rx1^Kv?@bW@Nz
zx!PD-F7DN-K2~*3N%gy7yEv&4*+j@H>>(=#vHZbxGPm&zmR-+k?kEcutXJx`#`jL;
zK4S&^)FMaGA!Y*hB{_$y`qTOBAv5}pJ8#VKZ<mTxm(JpN5-*W_3<T^a9gtkFU#5#}
z^0Q^kDIUAGywc$_YPyBDY%Irr7HV?6?JhPgt;WUYV>2sXs^nM}>KmS*Khx4n4}J2u
zrvmNUrIkA~j&CZTn~lw;Y7=p|`G%;=zCvo{%cCOIOO=nSm;w0ggs;Ywdx#$7p%WUr
zRK2x-N7{MlfY4f0GQtdQ9~l~A89wRqe~pRYiFTp;)Gg+`jL95$U8Wg7YQW|e6>ffB
zOl}IcgDW9!AzXE{qywdy(NT}soiu_L<9R}s@iXtJdhK!D?(2I)q=wq9@q?*JZLOyJ
zO+r53LJ=zbvA0%Wl$Zm0E6`RFl!+>B4;T73L^h$yWx2N<3*f(rCx2?n6{RI`Yf<Pz
zR5&h>pc`>D1q6+oH3<qucRTFI%`ECIru9fGg{=C%JiSltM=@S)^Mnr;gu5?)a!M^Z
zF?uX}e8P2Flw%?WW(eDj7@IJ6&<x5iWcyZ*o!d1_I%;F{8GE-Yw+#(d@Q7Xe7*n(P
zhEXzL;R=>n=}Kz&C=c&Y+D?06izb2$W@zIsN8dIV_m#0l6L2CpoSj>^cfEq`8cC$j
z3-oSAHC9gfPNvTMGa%yeY=+A6e*J5lqbW#Vi;elPk8wO=oKGec_Os5%HeqsI>@*T0
zDG}AgwWpiqahD{KP!!^sS{UQSXAYEowJEY~+(tr1o(Cnz8u>t`$L;o`{;YxEr&u)D
zHJi!vX&i?mXu+#E+~EHB(2G`k|3hBp1KzL`r)A&sNFfHv?Aek;KzS5fc2DXXqHWj)
zx??bduj`|tf=l8YW@++jg$?(SQkYm9e6k8B@rhANyUu~1kGaPar5TrIU|gICE1$!D
zb!7(Sw0SS<F12+Xtq(3;zu|EHIdA)RMUgu|?enyH((qT$pcRfzgQ%(vfevZQa4HQW
z6r@MEbZob5-{)o-17&lN*Gav`un#h#EI5+6Lxj5M5I1eo{OgF$<uL4fgC-vB3{H{F
zy(H-)^OFj@`@SzIxLK4MEo<cZCPqf74Gs>%g5$o~{*-;-ma0+a*tkdS$B3ya%RsZ%
zG7saHPE?$glWh&e#!->41mvJP3-oo8meC}}z-U5lF#`jVI@@KO7Pk|thmA0$38rKF
zeKzmXe&wmBhwImq-A7++8-Ka|9whY@Dbpx-*#LV4;2G62gvK$i%&Q5$Jzjf5yV-QY
z<&0x1$hRbt@Pl%@AUW1=$t?Q-t=nf%SNQRvfu)*?H0!&@%OJDT%kmtkw{-0lt8If6
z?lqRgbxUj0nx(lWEUc}`=`VfNgf&)^fi6tNFv@Spr+eAHAwD;YgJ_49+%)&)k52dr
zdfpu+4%Me;-_<mVALF>c>*j4GhPBrXQ`)p-4D5D`Tm4Ru&&2eyY<zKA?ly1XNhVnd
z!D!6m;9H3$s=gWx)&~P4QJJ<j(r9(VuEc8g<$%0A3vHQ%kI2r9B7)VAu2X)5EcPv?
zKqdxi75egxf665hVu`sWu<DKx9v;s+dxLQe;Ih3h!fA&QncJYmLs2+`AJNg&LhTsj
zt}}3RA*Z^o-HdjfxNNW)rHIo4!!~icJ`_LrOlY=O9a<4kvvM@JHmpVibvYp4D*sr@
zi<HdRgRY3*gB|0f=V!JXzkP$|b3a8YC@A1dQNm@?Z&AZdf>VsU&~<y-QM&9F(`Qvp
z+Y=c*04!F&K-Uco5&^yZ;|Ka2x8cDSi)+6OV6(0x?vMNy8?%%<__ZLy(p`KO@!=qR
zsPg1-lo6IhPf#vnPIux6N<<)<X>a6+{1GxaU^ou>uZ7G`>FA+1c`zsr7%eW29!oaM
z?l_OD$x1u7!q#!t^%L!0?r4aZqTU@-0T^z`>0)TA{vSgjYU=7Gp^!fc``mZ!*LT>P
zPq%BKp9PR7*Za)vn{HRVkaM3s1CJ-_RdS~o__p43O|tzgApQKV?z3sqi6v;*DvrK@
z%8z;mUw@2qi~KQT!;jis<6{SL!V-ZAhLkW#WAx=S4Rs?gTcZpb*M0r)CzIdo;{vaH
zLs9Lothp71?%zcb@v2<DKPJeBZ0cQXIxDiq{)k9!`LP8na=C%4+d#!ZRJPZ_fHYl1
zUgMbuUxw4TMs3j7?{6r?Zxu?S*x>NYGe*hXk~B>c8nyZq)m;I(?tJ+MHmj1&5Zs0y
zYQFhYMx)r%!}RCOQ2Ckku{6sInvNvHCc+xL$V2{EP3&N1yh#DfpBK+SqUzq@kH(K3
zLD9!n*VtUaeuHMr+8B>wY3;YGx}1Ie4VF5S7*5DF;+lvwg{8%+EJC>UGcPR(430rp
zWoIHwDEyY9h1N9$+sI|le&5&?SvLGt-x54OBhqDO*hNHAkrjA4pHo~ha4f&%XblZn
z_P*~_czf`xm7Bq!t-Rd$VP!=R%_~+s5{EHkNH7*v2TU1!ujIHUQS5nU>%7<(9#6}t
zBYFsc902+=lebCfsoOwuZO4iizm$u)z3Rbldi4#za2IJv98j+_$^_=dJO*&lywP);
zM)0SDR)mdZ(mbbz{UU9joc@|LPbK#uvt>CR`@@BmY&1u<yI`n=*t?VBi{*Tzi{oRP
zF5Vw^)@vPtGcz+?h4Se$AV<e)EE7i>7zVv2v>4u9JWD=i%HBdWW7eT_s@?pr3eYQN
z1->uO#tkBA$rDYR*e;<4?S?1`fHlHY$!lUCbnLv77Y$6A`Ijn;h|tW6UPM9sr<gw$
zduorWZsQ^!s7+4O@rR2PfR+$ABqxP*KBIx&(a9gMLJ9LWWIerq!fc_^JnE_%u5@K;
z4B6yh_jX)u$+FNLL8p{Vag$lXIXcsyxF}1kcupqGJ>1y~GJ3`lu`S0rK3w`fx*A6%
z>BO#x2S=N#U18BCEpalj^|=AhPsv?^^+&4W@K6w!z4x{UR@?bv#NC4qM=`Sm0%DYM
zsqkeAW={WhDC1Te98xP_-k*M8Jo2RBoK<3jAi3Ip5t2N<avc|s^E?h8Fa$x=?vd+7
zV<b7od?k9(d|w{dUTnI14{*i>Z;1dbL!21BMm9_m0;qaLW3_b=w|opX2;?|{0`&BV
zFi3b%e*mBihaPw`6oPXC5-G5#+dhv@NM1V+vbgiln?+O|VQL)e!V!?dmpQVRXI&bc
zk%{44@>R%%HMw<_G`%EUG)kh?>@<6hinm)GTh2t6y=V4yaeBO_z%wAV?+n~44zE9E
z**r@`g8E_rZ63eG-cfX<vRWygF+KM%aqpF|8DXOYN8v2wXUi!fujVhPgI}naY}pI0
z(laRfqx{!8*FQcMdn^bLDBx2iN5_~vv&Hhm>#RJHahAon{57qJ%UBxTYzEiZ^l&O&
zo<d`0uG#*)1FYGQDY8Lgq0`ngAw0p)94^NTlUf=jq<a!;PNR%0wo4q--j?tBEO8NW
zQs_U9B;et51*1iZlIU;$u{NrD2Ooxh{D_LVxE2~hX}>x|%)OaLZA?`3CbxKkcw5ZP
zdKN|Ga}N1F;zx|**CDe<w$^33`D?I)pfaG>w+}lWB|qX5lqgWM6!!Of;)mX{!FU~_
zrbCvnQRcml+b+ir$5baR9r}peH60I$zFD@lp{juEor?3e#)VKgNh6p!LfVJxITSSY
zx>9{srH6<-bb2I*GY@~<QT3G|*=;38H~|*gc%3_Y2T$Jlv0G3z08})bx~OWJKkhU5
zo1fb5CQIuE;8Fgy5e;hPq72iMHj~ozi^NP8oYbV_Nv9*yHixG-Nc{z<N{Y}v1BWOH
zQ`0*1G_~G#Fd}X&|0xf`a*9PyA=+=mo)*_57c&|<I_T=PriVx1c+1V{dbp&(Aof`I
zKOFs`z`r=2J;xe5jj0xDaCw@p#0CD&bi3Hg6hy>YdNe>)E^5baR)EgVid=e)xLNRN
zydqjuftwR?O-V{H>LBgO-dBB?ptt=hR33aXa<iLm-;*P{i%U+#+IiY?bUfsZiYg1P
zHsskUF+R*x5<J5`a*DvU$#@QeoyoVGe}z@OE?8rhr!~|p)9A;ahBqI*t&l<CX35H|
z7o)J(FIX>rySdnO*W$e?|49S7c2^9Qi`0jg=?B}ON!WsV!Qqa1?nQh_A#a;xz>nuZ
zsQKi(oqz1#nwL%dXr~yaE+w{0Vr4}JA$9t0+>-Z)MI&dvb7WF=l`V&K9=}B|sjG#2
z9<=NZ_FR#J^~tpoZ`18N{6k4T!x?z)V7tb2up*yo)Oz1Lbsb26TgmubGS0NTC4IxN
za5-^MrCfD8FB*;}t4%qFW|gaz&AB%}wG9h-;cZ8y8zQ+~B}WlKyV7@Z6<9z0AhNje
zg-z>eN3fu~@B$F$UPAk}Tzak%>eCFYBV0-1uRMYPUYKTPgq38#ieZ(9V`BY5#&VIM
zW}Hv)_<<csMMo#Q!0&NiQ}DXjEL~i@?)v*YR8i&q0kD5ld)8?EAo<B%e;E@ux#e(P
z?*Nis_5w+x+u9y^LJOCfbFR-7PNGCpe8><)6MRD5a%@b;aTi3(oSA+ER7t|^RW<Go
zm*h3t2kc3^JbhV(iU-a)d+lg&p_^jj+Dn~i%(Uk3TE;PE?iK(x>OO6Ke4N8_3pVU~
zxAZc~+*y2i<2bXjkHL9?BN1xg){Cul`OLo@!5}o-@cH<@{8G9xSoREn?izJxN&=`P
z=a)LJwSOS?#x=MNQ!9Pmp~2h*;02MfSsY5w!h|}%LbA@LV&bB*8T>++^?St2z40RK
z70Ob;2B?jptZQnGnEVSITH*h79bhQ%t_YH~{83dO!1G`0bk;#RROf*Jt;j2p)YWh^
zV4AE(^yLGm#&a0D0yBN<`A9kcf&}XjfeYXE>Ssl@u*Sx}wxO*``)kBQ#?8@s5BONO
zCPJOj1I5=P-e@tvQEwSYM8CNaN*Y36(bS_x9V_%eqvw6zU+BR68n7udMk4sRi9YY=
z8m@H1G%AUUx9hjMDKKnCC;|uC<GGg--zkQER;_~2YbSdyUxaBYHV;4&gYm2aQQad2
zq#<67!Y0@|MDKT>aY)>oEDh*mCUT9oldTahBCTz|7Z$9CHP3S3bXL`Drd!<i3#@B|
zeh7@^lIQO}cwY|k7PoO-n4&9dFj5~Ec#p%Yz}nB<c|~9$)YJJKxRjgDMj!>n6vWti
zK-a|#!b)DMXn%kXQuV&@ia^Fk;$P`wLyo(E{51g*bbd*yUT(A=MrLU-G%=!`^}%43
z_pRP9?7i73zIe>$9^$rNMatxJW&24r_?3p#>zEaEM1PFXEx+No$~)H2-QsoH3|rCp
z{Mn?c=vl&T-y=?X@_>3@(otpE7S?vR&HB-DxuWWi3TRfKG%^1GRpEi0GG)!oX`30A
zXMd9PWYMe|ssF1FV-;)LJ=^ET9{_39H5q2y@6=h)r$<~+xjZmvJD2h%JtQfY_Gc>f
zjMF~+P>5>V(8#(L_Nk<lx11$dV$$3EOpqS45FEOc+iFOxi*e4T_q3~5oN=GS*SX5*
zI57Ftf6&$#UWtD4%Kbat@!GUR8vlmE5Z?}iYLn)xzox<gXEY&Gf`L127$(y7$G!x%
z6@Pl=y|??M08Rz0g*MR`v<)V54<5-n{+$F>!_$@KNztseP<`?yT9%64L4_9pP@Se4
zjK5^=xD&(Ot*4GBZkZKR&}&O@R6CCJm1818u*i_)&tOjV=?<P0J@nBTRPf7eosTI+
z+x#gd+IgCjfR(;b%+C>KJ62q=>F<?ucuy_AeS+hAwk(PYz%&JN$vyE>X;S{bKj4E>
z0^PzBK=t3nG_kX0MooDNNml^alr<em>SQxE=c<(T{R7U|BJo|;C;1J7H*csq-8=`U
zU9E9_k*?P&p)%4mUPcvLFx}5b#XmQSlW)ak#eppprI%e}B{k~hgw^-1xM3h((+`XD
zxr<;}yg<RrQ2jrib#>R1FR`y+EuuhreMYnsD#x5L%bPHOHkghD<C)B)zdkPfGBZ1A
zS(FLmy!&G^*z*LwRm?%x;CTJ4{nkL=Y2~>}<=8ataO=4({CL9gY^vs5^JpXmPVw|k
zCeHN`ZJhLjI^6lEeA6FFu13l6ilel#UkvOxgxqv>>UAGpuX~FM>n-fKZx6-)23!V5
z5i!>>N5X)0TZ)L<^A*7FE;zqGNeM0tXw79<+xs7r?-2(6dhDk@59okRMJW@@8d^rh
zO@Op8Mcz62A+9BU6JyOGQxL!gVU~|!iGVdgj7p!DeQJ8ZjX3l8)rxp-e{h+O(_>?s
zXh%mb=}4K|GQ{8$3ru{4#5F*Xa5N9_yxhIlkuuVRvqKd2d2nfhI91Okc5kd)Xo%rI
z?qDqD){36WGwQf@3@vg8dWW+N^}~07n^;Een->aEiwf5{mo~!gH~S;=0-;+K^C8CM
zM@h`VU*vvFAL$4_MB&wQ`pSg8P1$sYNmF<k-T#HE6VY4><^lXqQU;ql+Ah1{6F@8-
zF=2+v#KkM$s%a9ePrzmYIG1J<TFV{^%rdkZjkF^DLfz_{T-nw))Qud~?&T2z+W^b9
z%4`%zUw<QjdP~Eo<qe4ku&1rypyPHKd@i|@Fm>P&_60l!+(DoB<V{1zV4AtFHE#Pc
zXqsJ8!c+M^JN6i=HZA)_l(>t|Mm|g$=J)e%gNZFkKHdLu;uy}*T#m8x7z=sdIW$&1
zq%p-`M82&~*O@vz$DKXEI>&0bgQ&`14owb8nXI8oolHCp$q5J?`M$i37$>nX_J@Vu
z=oSIDu#7S%r6W7t*2n6mB<cIa+odJQGIqd~TB9onO$*k($<GIl^i{GKKnXIO*LUYs
z?@nPgG2H!;`p+$TCF%P4@hlWT>QldLk&E*PuQ4YNp=2UE*;^^2=ixRLtCxJAV6_oC
zWU~ODtG85X2^45X1uPY0?@K$ZmOd=%;vQyF|G4n#Lzn8RF&lLW?*A3*Q2x=*u7XK`
zMv-C%cT@{7*=br=Ayk#KQF~<iKWt+LH{k28H{UH;B!dmzuP5U7St?HfOi*+DHUjk&
zRbTMdjJP#nmv_Q#lCAoNQVRF99;hMG447v6(Ob!tQ`NWuDyrqwA|{ZBWwN8F+=t(2
zcmLRj&)O~(Qdo=vN+KLWu~#hn6>;5y*Bg#`)>fOKaNHfb#&>v<SN?R6crSa-r-a*~
zc&F53-EP`@`sy@Pw6Oo%z*QjXtCjf$YbkX&gY7&(^Y(k}m2ACwjS`}v*hxUV{b0a(
z5zG7Lx7L|OXylN?)ig7%#(h8UCg2dQELi}s_dZO;l%7BYX<(k>pWKv<w_8$}X1Z|^
z!&Fa8cPkrLJ9(zYNJ);bPFxsWY7**Ge501Wr~0%g5wiRq2l>>TVD`%S)W_w04_&7Y
z$s*Z&Tvg|s&r>i@pmE(!NhR-m6fUfH@RoH>^&8WB>{bgczH=yno1q(X8LXNv2*mrt
zaiT3<Ae0<B<zC00Mut_1*9{Or;p1C4k9r4q1rnm7ko0-0^sn|Odyk}2@1lwMHE#7B
z7s6@f@m}{SP|B`$d+meGzS_%KXqeSRI)Yo4n3MKLDvT#>6$_rvgk@sO!R8&Q*<FoJ
zb4VD%jS@zQp!0fJYFr)tNOb4M4|38i5~KD{E%iw!`P(%cj){#ok^M>DqeUMhA}gb`
z%2Zp+qPQ3w%`olRw%qntWqU(h;fIa2b~Nd|dk3yRc?TJk8Qg^=y_>N9wllA_uw<V^
ze=l`x9<IE@&K^B1zSH<smvM!uwn4B~(!#yo1h4Q91~Of<b-32)D^`OK4E7r_aq&Q>
z<LW^jV;-{+Eb5D<tw0U98R&Pdq7@PVtWP_bhGlsmZ(<T!muo;P@Z^D6*+$|;i=58Y
zH^vMT!7>6w3zk#N^`)4q*vJ!z4`Y1gfV*cDxVLcDMQ09l6$VQ&xd1Tsr>q`_Qtb=)
z)FRAW{m!yr{z+J3&~2|QRpq0>`<0Q?W=<n*#5{;3ZWi)|1;^EGJ$<M0MeNGj4A{Y;
zXmeQVDqMVjdVA*7DtTuDe;ed;owz?v>2~i*EMz(LPPHg_eSP2Ko8*@OCAe(xLkaaA
z`Glr{X8~1+?Q(;99}0cNQ9v3?7)oZ^yhTes_t0BBb!QM4y{Zx3Ecn`!LoMbuaa`DP
zmomZWAgGMka!Vd92*S6ceT-yA`*pKVfqE)v5SG02Znp2v72zq^JgZouw9I%Q2e4Wh
z@3Au$<dhGOOm_E@x+*B=dwyNOQ|SL<7N&bUNMhB)*lEvQkXy`Rq>b>e-ZlK1e`-=0
zKJ&>bc`6>bi-K*SLORcUU*x9Q`aN`v40qX$*6Xajkpc&U&ePM=?Y;pS)aq95^l(t5
z>!t%YR%O&{W7!@9Cq{k@xJwXXk~XgJGHS#fc{2BAR2A#{l^8+_s-jchhoK1We$+v$
z5@ixWr)+*pea%Iy-UowSL0F`q&YNErE(IQc!nj!Ehr4&}F<bBQf8YF3QBjl2c4=E8
z?e`toHn)mFAr_HySjkUAa{~a|r&7^AAP7CY#ijE!jtEZl2VFnd%zM+opYS;`X7sS*
zq}A}-Ej(NhjjpG7h?ZnrrvzMBhA*~h8V`%DSoYF%P6itX3J^q>-@0}_=WpK%F4R0F
z2orUP!aBn@gjhIPx!30*{`xj%(Ehpm3zeJ%@9!UeF8k9(s6_57j#X>m#}dR;w^dJb
z$7hiUg370jG3e1`=FrB~&xoJKvyg#6Af1F$An3ETHv*g9`xig|gTE*P9q+o|>-ER8
zMv_a0U9k9Chfgl36a2gL<w>z0)jfliWS6qAuo#xw=jnR>_hDIOPwQC+eZ#}|XeQY$
z!PB)EM1g46iIlG1bO~mJOS+?`Q8<z`^plz~K2)d5R!+X5d9R`ebcF8tJxXW2g>Xax
zd4;;s-MyHgW94KWJ8eALq1OVbwVyViGfwO%6laXH$hPOz7K))}oLi4;B6%2Fgy^uy
z(2G!xNcN{N6Cg_s{PAZqD499;Rfy!OHeZt{mU0uzUJks9y^UM6-aS;CJz6)sh)z7v
zRy&9BMD;P_)<JNHW|0%7Fr-iF0(n9pG($Y!aXF4!Q1ji6ztgj2jp*K<%x~UO&46#v
zIlgO`N6ri2`^2h`Q&jN_NXzZ*a7y1tW8sMG^xn!u=BcTR{d_!2naX#^M0|2zSi~YB
zj`O_0j{}P|?vst51;(xQn4rGx8{1cL(iJJxq;~jZR?=FFm%UMtP1l*rHB=5IM%nIq
zG&fNwPdICVmA9{?oV#ZHGXkaYjkq`{icI&zBL+^8SQwBJA-FZq;x4>I)GzcgAt51!
z&y#*B5QPt}F9A=40doUb#RdtH$QQOWNo$UjYMtjx4SvP^E5ZV<Mfzm&kl*>8Lj%{F
zb^WtsI6nRsamuufze-^vKKgvb6E9hY_Gv*B1DkySyenSo1BCV*RGybVxsAH~UKRe%
zc^#0jR-;=PnAW(vB<4r`q0nvh)fC@ZK0&IiCoRW!5F+~Ijap#W<_EOYf;t=Ia!?90
zQX#Rerx&{D$wa#c*Tn>q6Q=Mv#@D6N`h^-6T#C&Eg@W&W7X?&lD2*95ysm}zupS?T
z=^uoo@9X(Fx%hm~SE!9iZ-LeHg67y$g-3QZec!Inw3`XI4aUfd)Djjf-|miQqk2T>
zQ|O~1bPtmiC6Puu67k+%l98FL0+JXIP=eURN#(vzu)Y^;OuM0ok==N}Y2@vjeuAua
zYK<zPU6ctfd-O9C0JdVV@bjHG?rgKOc~>pY?+^SoOw#x2R@OOJH5)*rM{-to0COB8
zDVXk5JY>Z*46ti9PTohCi=Kw?&>@5Ie>5LTE?1%}%LsNNtNf*zQxk4;tyKi7z)h-Z
zSg5x|(=js23)b6h_JvPWn+)2b!vnG=_XsEL2?9|J#FC;W>${~^dd29f%fu14sDoER
zszvQ2OFwrL6*t*`JLn>FmW@O^SO0$hJ2Lkk&e5i(NihK!ECA^N;uA`yZ==q6jKP4`
zg&!!s0f5ksFAt?f<{~@nn~}2m`g$qrSAHw!U&(A^nH<j@<%rZ?%66Cy0tIYCGs*S_
z_{p|@8xV{JfdUf%|DxsWPZ-Z3B&X^#&bF85$KQa--2*tJkc^Fte{e=hz9aeKIQaqh
z)fge_$lF{r2O*xZx4wWiVL18i!{N&3tqfIafhu;9UIPZdVD>Q{%Brs~0jCq04mFYH
zfhz>5L%Y#RyqDyzc04hLe0;2oRMry*ae|u|YejHcJ5+G@^&_w$xeP*&$F(gM3Ddyd
zh1}vI6BS?V)r6?1was1Q`SnhWov?JT9XdV$?*WrlN+3@YG=!7ab`VPBzm}_?u$2Bz
zlyLh`*kL_sCv1JBFk5+Od?FeuZ?AaD@8mBVYq^d!0u}&%3cxKojD2oqfdZy5o-Wr$
za~kD3Z-ELK!+xn60uDPr=YjLk+a?P~&1z|M-=+nb)-}M6nrX5y5M6;SL94siv`n+e
z*@}!=2!F+8x3b{z!5h7CpOh)>gD+mp8j4(Kh0(!H1f{dFaRHO3Y+8y+-0m6$%dg%b
z>V1@7DcKgnSxw=C<nvEguO%+H#oyUS^1Q9q<7}CBTScC7>t3rcw!5z|9<wX6m<n1A
z6$RHrZ&6t5U80{<#XJenazAJ%JWdQ(@;KIt?spD7a2|S}hqxM#@vKJHBbI#2EQq+@
zw&ah<#``6Y9eMa$h>AK{<*Lb2jI#6F&)Lcc;zM7E0?{>7)#w@p*wZ-6r-RRxq5kP6
zZVAkfU3*0m`sj4Xg8f5eNAs#&HuHJ_ZstN_x?mlPlN1+wt2Q7Gr1_e2QY7wF)jM14
zLlVd+L7Pump^nDJH%Wr1Se)#5ndyLyP)reRxiA<FMkUD<(C>~j&<*ch2c(k%b~Hm`
z1Y&kHV`9^_l0e)vi$zBF^@9VnA#|M$g6xU8s*+!6v#`ay>whBn|38AE|2Gu-f1~IB
zH+uemM0hBK$T93+Gf`1}Ang+Ri`4Q3&wpk{t&5c39ZrG*y(@dSN>N5mZo17okxU5$
zVh-P-L4aoHtj-;{AS5)HEXmw<FZ)R%Uwxg0?8N?_sOLWjFkNsSF#uyGOMC?BzjJ$T
z0cy-C3BAJb|CPaS55`kpKXWe?QprlhZAT<<w;mYXPrzY;*7p2x6b<f(r!?%d2Xt)~
z_ftJy^ubcN{GuYcc*Fa|o;*nd=bxOvC&W=X-}l|z0?H`Y!@Xoo7jST)&}&xMeB|)F
zFm2k5peK0pHi;E+b8GO)4>S5JV-1AQ<ei-3N&S4!)xk93S0p?ZC<7omL{fNdfpY!b
zpU}m#8j85{(Q~?;yErq*{vEtJx&DFtUWP!z%JXKK_3BQ*e{N9_i@&p8keB|jUIrBZ
z#+6wH0Ynsl-SQuto4d)s6Hy53{t!{3vj0Y3Y3u~JApqgzKZq#jd4K1Ia9jN0hTsYR
zjZtFf3ovW|+RlG)L$;&<e$T%cHg*$#7&a{6zwvVZbj$%@()<U*W+3+Oq?eT(;y<Jp
z$ls_k%}@QB8vwcGKS(c8q<^OlHH-59p$^^2{F}Yv^cY#$0|1%-K^?*u{yUq;DbVc?
zo9F!Nzd1nLH@E=724L;{oz24r{yV{D8|LH>!Dfs4Z?v2fW`L#y5NZB{VDrBN_IH-j
z|7gJ05`4qowyzfC=kLEp>L?%QcmuP}$jlrH#9iS78dC8yU?IN&7@`v;szn29Bwoj_
zfJk2W@n5r=AY>qx*V_AZy*tq88FK&Ea!2EW0Opy+LJd5HE|>k&j%`3f)JN%G6@hoQ
zFK3mn^S8H05R?2C0J06dUjQVms$Q5uT3?^|JE)Unu0ntLuD5C|lfSI<m)kP(&5|{<
zL#Zzt4J(@4(HyeIcdbcjR@Xx{v`p{3Z2Mz$htkfV+s>YWux(p^X(J(t2n11DhK0|J
z^uY;Swq*6~6TT;n)@_BRX3MxtqO3*<Yj3cBeH{$K=#+jffgrIwis{4BAk~Jm9Vx_Z
zZoO=M9Qi>wgIJF+f5iSP$52^msbP1(TNbYy8?7b>@|n3gSwh2x;*`~PFY%Y>r;WR<
z*iazwN-Ax&Y>14&<r<=y`}uU#0J}R$v9{t-3_#)=&J85q_=8+J5|;>yLFd>rOpi!b
zxV~?Ytu>^imRE~NCy@Ub)@+Z&sBu|#>0{$?rDDI?6^9-H!TVQsJ`;4@?-E`Ed==Br
z+kN4f>b#sE-tmy_-{9)6%$8~PCeW%H0)a$I+S>TUW-Vt^1&Bz5Tu33Pq!lQ#DH!i)
z$N(P3V<_OA^)Qe<7Cvp-LK(l;VMWpbCw?(5el()`6M}qK9@LWE0ZSC;u+wK~5HjJT
z9jFi*mDLWVGOZWynMw2U;;7Ijk)8ApoU8Tzr@)QGkdS?XSq8}8sE4`6QBbXuRce8&
zR{iygd@B0`(QEA@6!?5(m7T_6{B6K<;cs2h3<I=rt5AVqy6tL8&H*rcbO$`~@gxE(
zyq);e#yq-4jkPXhg_~gOJ{vcfWoY_~uEQcDE46yl&a|~cSqXztv0%5PH0A)5fNbwe
z8Tn;j_cvA9M5=26z~j^Q11@G$l5ou<J3IUAE({msjXuV^M>OaHriROZ!beco(15uM
zxL6NF<wFs`lg>38O>zKio*n?;hQ?<_9U-eMF7{i|JL^=*1pIHmA(!KXMmPPW05JQa
zd_O};wp%-(-vBtiLo?_r0;4(x`+m}Ez!=4W8a*o>?cOFt5kRAbYkgUVQap8S)ET(P
z#)ARozHu?yN%e!Pk#$%{>gLqVF+%hFDiRRckr(&X0*Lo9MY@1KPle`BYIffH0u}N8
zGY}QI4n##nqFn3<h5aA)&a$blZrj#LaCdhL4k0+f-3jgxENF0tEJ$#7ch}$qhv1qJ
zEVx5(cfW&o?|tgjy>+Y34>(^}Rl=HUjydM&t+%Ij*bF*9eo)r^cD!XWw4X?Z{bIP|
zk|c5vbY!J{ypcLq%UzeuPok2_edV7;^ja-oC9ollgGcrbi1*DhR<hl&HyYBAhYL-7
z9(!^EDU_KB#dj!~9)Af8!_^K5G&F${@!kOsH~&6y`7@*1sK=PTq<&#n%+LZ+?{ok(
zsyy_VPouwB5^_fmZk7N#p~h1IQq(1I-!D}`nB&6<5MvaGa&yR+*Tal1VN3?jtI@%&
zmcicus@%J0C(ZG+;Ux+eZq*F12Qu%~HDMk4m#hA&i9Dy{fDh2|?4_F@4Lre?`~1aY
z1;O{@&o!g<)WlMBJ)s{IM)Ti4bz1R-LZuM?gysDKJQvwsSBF&|XCHM3#>dA$fF=rD
z>)Kq(?XBm0ofN4DHIq&qd@Aj6K{Sfv{H}dCyb9#u%|P{!S+QS1pl9X4)S`*-bz6we
zqw*jA=VM_!$1JjA`dfAHPRrj_mt<agJ(F~Ll?ijDoODxD==3bk#1eK6cWu$UD{5Yu
zq`mJqxzAyFUFCt3fzD>;!$YD!2`H+MDj!jKR$g4B#>Bkj*~0YUF~0zQkf7c$JTV1@
z2*fL&S@KHA1x#`3L%7sW*}5Ya>t1t?hayTnrZLyCt#(!eWYL^C!VO~<_qks!qP2^c
z{&o6SzIBVo4Mi`CNeLfs<VY5ab#G86Ki4i?s@h%7c9WYgK#%T9_63)soAR54nE3JI
z)}31t^%Tvvf4=Jis^K|CS5V*8%~lqgd4>z_`R-KN*pEIJ4zPvrj-+v;ef*V<FqP}~
zwplq(vJ+TFle(arst7i~YB$$Yb%RAA%>T8n1nmuMNn_+vpO<ZxA9A#lDs=HG8m4ip
z-O!s^UZD_i3!9!vTFg`#l-J*YHSu7Cq<qNiCiZ8A-DbTl96t9cnDgmE=;=jox1t*d
zb=(j+o@#+?zd>=ud`%Zt(eTHAgFy7mx^{Au=N})AI)s1DpY?j5X)+h3t-VGDniHI*
z*M5D(ub_7|_iHSS8j2ejw9&JV<rb}|SR`?jxF$M(MVzmPz*!H6jbUTA+PF^90|@x>
z-Kg*6G0tA-kM*P#m)H@hb$qOko&KambF?Dh+yPXtFg!xdbCDHI$D4sqW;kZ7q@3a<
zp6i@6D$1)2qxH*7nutp8Y>A>Q_MtoGSd*I8u<R)R#SJK=aV0u3N;>15)XSxSZV^M2
z#tk;r2E=?6=ZSod?BhOI3k^Mr`@{v_K%|cJz=J{$ow4lg5p%{2)so+nK2rah{apb2
zon6XM<bnLY<huHp_m&9_-k1^OWA^@`ZhWp2Ed3_ha?vEDO%LNo?|EZtFByG+Uu4bX
zmto4|xc_7U+_a&-Cp>NF>4{3^p}TR(Sf0}<I3B@r;lYXGksCakSHKH=u$$eO*Q>o?
zFnWKj-qG<}4Kf7_#BP;`U*4L%)-!d=^n`pfu}UWc6cz40adbKw-wqGl=PySz<a4|C
z@jVGxB%PRW%JA7Vdj@!Kl#Mw;<cW}s-lOm(@(+z`$Y&YPW?O4|(^H)>vB2M}`Q2B2
zjt8pu`(4ck5eB*ZQWbi>CpEeK(7^P3x)}HnJ0|h58Y>qvBvx4+j)8`G+jkQC@;TvT
zLIxPtLfD`W)Q|X};Q??yZ=lqKYK2>NN!V<b2J>$6-J6oKO-by`ROf|6;cq6yco6o3
z)3zdGQ!r|tlWSr4<RlNRqx2U^w@zQdM7}ETxkXgK;=zsi^u(Tp_LEMoQ{86Vc08#e
z-nvW&(WZz*vikLj2%l>TKAAEdGoW~+C44nOt=`ydK*3@z)3|blLHH(Opr=>vy4}}x
zzx?!&^_7ZTp%Yk`q}gnWH4$&GI5_`ysYc2Ki-dR}z3M1!Vr4`jnwHCPz?!Y{h{NCE
zneiuMz;&4%M2X`W$KUMRj`FK#RN-bF0T@qCPZFMqe$LZjiwU>!ac|!MlGLEj?GEjk
z-)DFuzjV~n>ZNc9-&T)u=qId%)<drpXYxzN5POkxGTUMGt2<WNr5vxlsuK4#qXd3H
zE#A6VJ11KNdW>lslgo`H2T!RpCz}jb-nWNT*M!oyDH}vU-hIsB95{*lUX4DJ0?ajd
zu4V}O?#Vb`v9vhDgudGhB-DQ7iSfi;jz1AYXUOsy!t53gM<+1r-@Yo3UHi}j+eFHr
z4hfo233r*w5reC~`&b}Yjbte99n$#{nW{sMrKtzT<z8k6WCB=}v*%^~;;oLSF=4Hn
z{DJ~=({4^(_Yf5AwC)_*#jf~1-lt=MwpPU8Y>aaFQiXGiM#l?d;#+y)3!EsypuSO>
zgnzy-R}RO3gL5meLls7DjIfW^F-p;Oi<P14WvjnE#H{^rJ|p*lyzzA*j#`002%20)
zrF;Wnq^`mHK*#Y0Ug<c7jg8?&o}pw~G2l2aq{AdqUS|YdxPqp}9#-Fq*G7z#Jil&(
z?YH1=C*u`Sxgl6yP|&;`DjQSUO4k}1njK9&C)@43y|Df~Dy?;gMm#xw`|4t@EllO=
zp&`PZ8hm{~@d$9`ej}H3=@{?W#WD1552010znCD{oZ4>ia-~CYbm*y(p;+DKVbQRX
zWhqDOL=YoxdjnAv%b{zss>3QNI#Iysus}yt5}eCN4(&&dP+S{sQ+|p6(9X{xS!ffz
z_tI<S3!J<s+MdEbdVvzY-zcg{`ovL#38-|co_`~-Kvd=w2*62t_A1=J`soMn+D*_w
z*&f!=QJa2{1Yk+Yv{<*x4clacl$kPPt4*BU^O*h8IInlYPp!`91>VKO-aI{C390rw
zpN~n+Cs5R_8mYnC{-{Z_Qa(HJvK<hRn@~FmgUvn;e|&A`izM~nDhc!9{O-Y<`~D!j
z*HnMf(&k7fBOx)d`O5?s%As_@$#?~T5Nj2V)S*FjN?q?+(=yIWn^?1ZG5h)<iB{5|
zYU<G^hizyA)B{+$?&}jh#M`+>MSI6|Jk?1vopK$oSyE6O7XsuTtd5Tc8eOBD4ofOF
z#*JQQ@+1r?nbS7tG%nVPnMbeFH74$CTjzdfsDlc{B;^ZOsX^C}H0`Vo>T7*S81A2a
zZC!dTfmgh;P5RpGJ>Tl?Hg)@g)Ntff8!K8w3_a0?=MR?VZOXVqEOUl!EDUesr%zbR
zByI~aKj^tIxdu$S8TslHMkIqIicro34mAo0)aTR}Eg4>>pz4Bi-`d1$Ci`!RV>GoQ
zr1bPPdhmJ=fB3275ATgh6rL$C`)S<Y%_f*$r54)t%?<1Le161kSZ{E%0h@_<NVPPn
zH-r6^=-!?st&k_LauaP~J#b-yLjj}1;sqhsr>dTswUC{$LGLE+q3E%rVKEcaA@KB7
z4YRb2<ll`?Ok9M(3o`l<&qJ*p6+JJ7ihnN!WiA`7B7n<`=0h=>3nWI%`<FZw8{S1B
zgpH?F)bvn*9~L?v;HvVU3f<3V6Sa6@{&=5$3Bk+!rt&nMrI_!~OZ2>QO)|e?P*x9y
z@Xi-jZNDI@*{06-a<G7J1#Lt=9sUy4{fcq?Kq>L3%Xz!wMYrWnfFeyTM+rT`AI#N5
zu`R-Y&xI~Rue|mt6xFbc=3!P-0iGwRH$)o|wChoc@0as4U1E19z7U=aDwPs4X1@|r
zf}{Fc0<PrhTc$fec?Em056y(%pAyLtJVOxAUBM7>dF?Mwzj*txGq-)k8-&<ID>7>c
z^cC6gU#<zKZD%JRDZ@QpOdGy#=3a7#IzE@_(3?@*)dPf%H>f76Y-ZoiG&)diY;L~H
zq?@m|)|^n<w^_$ys<`HYxk<ngFZuQ*774^}9d%66$bFi?of~;V2mQS5u7_i49muC)
za(Sj6QGwo@WpTjhxb3T|&VOf*!697~BZe}}M^Y+`<J)W3gQdaCs$>-M)LG!gd3)C~
ztto2;gPkStKAo#wvY#<;bnLg|uDpKoG-=KKoc1oIU*TBl_M3Oj?%gMEVz#$dKA&^x
z3q9S}8WVf46-HeieV?}-QdJVb8KD1U<SFK+%8F54PnemHEATe*s<(U8UhL2#wJ<OX
z^+olf%IXT_HPgvi3Xg$C(j=Nv9kSNq-N}fS$0>XhX*H;?`tt9p4v(YBO1mc_LGp#|
zAOJ~@3Tj*=%qJ%rwS+K!9p)=C@>U`30fgBz8D|a8zjinYmsC7+L9iBbVTfqUXudD?
z`{*)XqG2?)07>pi`C<;~H4|Ww7rFKZ1w|u&05@L>LhU`(m8O+u^E*p?v@OOa<M@<*
zj63@4`YyYo!IH(P!<(Gc8Wk-qi&+QxG%kgoLYWIG2?=H?eHAM@Jg%rAeCT3O@0bup
zhNC8?Y$d#>S~WAIKZn>^T}$}Azu8|0`wxjq?*w~p+>L8xy;n<_9a`iUHt--<eH=+P
zDw)S+(#i%~q>{Ht-A2d(&;%=jzo_oZqydZ8pu*MRVuD?}gE@3hWVArQrFTo2K!Fci
zOei9aibm<(YDl^Za5%0!jwP`ERI!)SEc<4ld`06qXM%r=#&3#>GPs?Dl^#p+?%VU;
z^tMr&o_IzxC@nN+MiPJe!)!)wL?F+vDOR3UX?Esz^!A;nLlwF&u8w5^^D>sjh2`*m
z+8MT#Z@)QRBj^$np<vM}P6qVv4*vWZ0PX`Cs?hzhmW<4f*xBHS*`b_d28zovq(ClX
zyp^|3|C~3UV{|NlwrgVYc;bzvRvThyP$n`6!NAQoP;Hneof~-&{xGYDcNi3OKR6OK
zGIP8^&*=HV9&zal9BDBccKg|=&@;uZt=gF83vAZ}uNB@ZrHRg+{d$L%>2=7<Z6j^j
z#bp+t1}{SW_oTaAyY|EO?;JH;wut~`h@2$qm*xb+e!avxeK%0w+W7*c$G9*ei2V3b
zY{u*#K5Z0_8yAGv5b1Q>qUIk|shT98yq+y8w{RCnJj+>Y4|#vKG_Ujfds(td#&|By
zhpbf4hI-F1L=oEnLkZ*Y!dSL;tiW{}x`}%Um7ooC0=DFoh{2C|D@2!IxG2+;A^5XX
z1iska<$*esLAoTI@$?IzsR*qEkwIQ??Dv|>sd^%#)2@~{{Q{OyuVHsqC@4*MHtfRH
ztd5-Ry7it$g7$?^PTGE>z6-jCOSsWLS_=KMV8cLo2~GE?@9i>V>?>g<^Pva)+bDtR
zv7;>%5Y>@FOJxTR48bcuWP#1{!&IqyGyI7<`s4kX(u`3MiSw#rW@sql3?R~tzyUQ;
zW|U_g+5KWl^B*21O(ZiT1I1W2|F*k5udomn$YS}AJx0#sl<Rtr02}CH>nEmC*itve
z-{gMnr?-;eu%R67GL#Yxs&-+kwP^>KMISZRWT?%C<!o!re?dA@yItX=&);bB>6A*X
z?LPg@$nvY=Z&Oly*QqYEL^XkWFH&fc{o_ZNf+m+F;x?ykxAjaF`1Wn3@RS<0Y^XjL
zCer4}<D|ELl@inC<^Sl}<&Ply@na#4kkN{6j0Y)hpwdTMB=yJNZHZa#R+^B5#DJ}x
zR7{A~s)QF)n`JpoWC8c}JAohY?XMYKHBD~G@A=Jo!*4eR<Ct9%@oBheUExJYEHt0+
zA%C>Zc>zn$w@uI3_)ha^sST$`jp;TC7j{1P`c;|p;leB~`@Y+sN<-Gob@lx6r8g|_
zn*3`_a*MbEBw{W7-FEh*_CNcNxJePRw)x>^eO5Z9Za$;^Syksy+xJ^qX@vwAl_%@U
zcK=n7y%D?8whbw+=UuPg-v1(da9Z%{tfvwxXQ=X<sVYHgjx$+}aWA5-S@9-Z!!ah6
zbwx6Zc8$2@Y)t`;QgvJyBk2h>q6$S%yj}IcZ>!2*U8+$Dzg)t02>g2#U^{$<HP}9f
zD!#Nv%sq}bne<44h996#ZJp5{3>Tm_^;<yxC?gYMTwkRh-bh#3dK#MY<F)wy>)aL6
zzpfWis}M3&?069ab@hVh<ektNnO<0uyAQ-81l+a&yZBqvV+(p6mp!C?1G^~eU+t^$
zH&PF~two+PL-AMDC^<PnGDY-5(2lb-+f@@_we=ISfa&AS)Wf}>is!>78^PMxC$u9Z
zIFf>ut~qgH$%4&vV4DCaJYD*(I2;r(b#ty@<20adWo3a9zHmdp^W2xU<9_`4rTRN-
zWFfKOvNkH)c^_Z(^jWgnM;ciz^cRHTz`3uwR%hjh!I>CQ^|DW^)KF?ewrtry#FG7E
zGnh3}K6`^Xq^ePF^ql2m1gw9B(&=0d;56OM+_3Pf80lO3wG}*?z&Sdo<VgJ;EEDqC
zkQ)-kAGVW_tcF7IHZkDFI$`kid=o1}NNls*7aWW|v!YnSPj%1!@PqTKjHXpV9W*7;
zwBL*qG2x#|n1UY2?z2iNJZu>S^=nJPAF1@I*yBopjz~yDvuEE3PxRO3^lsNaKfjqo
zs{`VqI@^WfFmtl{LXXCul7rs+6I7C3{!`TKVEUmLW+DJ9#$|u2^2gme*iuY?1%iKS
z!>`^(Gj*g&*tLd^Xh))3toEH7Tsx>qt=qY@oenX|MDd;uP%+;?9&fg&A+{l3y~W(p
zq~wG82MCuL918d=veCvlaqdh1wwJFMK}jMzAjx=lqjhNyDmiPRd%P8llNyK!wz*hW
zYXuwDl8U(z31D?9hJrDu;=%5v)miBaoLZVddns9w$pW5;s1Q`Ch50&56GXDNrok?-
z{D>%4{tJ2gj5EvP!{M&WAke9j)M!}BD{*lu@y!ani}+chgY1&gUGh1_Za>fpXiVod
z{V9SU^K!@y*&^-NFB_`Ee-Z9rG}H_>(6oN^`=P{~pKNUGvLLV)z1)zd!nXdh8cD92
znUoW_&}};Y5;wbv-kQE^n(8`uYj<}bM^)D{#}-aYjAGidwd(x?c$i{lIsiJJ=w01;
zv-8UOHybu$#9^ygb80*-;f_d4p)c8CS=w}47=jRgPK9f4(if0)E~VrCfgsHD4bg`3
z2Yg$HBL|Mw`rBLshqM{Mnu`BZ%1<A!cB3d={F$FbaZGQo0dJ->`|f6RB)c}4;U7Os
z+EE6@wtC~xk>D=bb|g4T8M0DRqVHkHe2OO+#R;m@ca@MQCm}fVKMZS#HWF&HW&^VV
z^BQrZh;vtvqI5MuaXDPV5>P}!Td1VoVkRd0Omq=!xNk@4i}cWnV!n~4R=liBPxAOJ
z0vF@aOo1#)(Kx8-OxW^Jb22cStv_>8_V^*^BD2NVS=z=0htAv8F+Tt<waJGJIJH}G
z?O)_-RvLWgQ!bNhDdYL1qTT9V-F&@h^`#~Gk#bZxy;@<ByniihM=0@H=yp5LA(x7_
zp-h!oPmv^2i=UjXV^rkzI^xwPF0Jxy=Aif$StFiT9>rrNg|=%y3R@EUa~{!G{)zU3
zVERKDQhMePg5e=;c6pbt_Y9ecMfZV2<g5$~HENJz|3e@TCo@u|zHrkRi%)Uw9iwW=
z(8;wdxeC#MkM~~dQpm%UnY4^myg^}yTzAFN3i=RKcUPs33OX~S9-ZhAHMHDjzQjZO
z;I}|PO)UQ-F0l7|#p?$%%?5nPoMjQcv7FGw%SQHK^*P`dTHu9p$iHUA-w38)Ddt>r
zTMc-Gh;RswZb#+mV9M%%Qm8f(8&-L<yx)6Nit-MPf1q$>rrX&}qQk~2faC>0xy#;_
z+~2*$T5R|OX&OOg@n}qbRB>LSZ}t>vzmakYvm}hUr}(Rt*Gl;CG4vZn&f5VtIZ#}%
z(+e?;02s<f6IOGz(QyeL@pgD<NN-FPlSYN&{b`Ucp(MBG0cnFUvL6y71gIJ{nN;+D
z7pJ|iZSk{fZT7|vgpgCf_WhZ8<rf8LJt+5xa>7KpaA#^QG`m=DZp0~RY$ZIM4zu$0
z0SkDfnHTz&;P$x3!OiofhD>P$7o|IfPTTKXOAMDzwXrxP3m6NCuhRb&$;v%@SoY+J
zd62s3f+=jr-r`&T#ZdiN-p&xe<X%8SCe*<}jVV-*5cXZgz6;|uahMcJNLj%8i<FP^
z;)Z3*hWL1f1l7;g=ImIYRT55&5zlUMp74lAk8Vb{-xq(($$;LqPJKAAfv{j9kY<i8
z6vbir>1=*t$X5`)A>kIiAgbIk=|#}`WMCM7tS@q&GP8?B1*w-441C6a$ljBqwp+06
zKnD0BzUs=Dq>=`k89&lbSs0zG9JmvJ-+N$q*zfzN`LD@TsID%M^$6}Rfvbf)FZ^Bp
z6Nz!x23&pP+#>p;8ccmt{&EDgkD;z>NYKG}`P<}5M+dtqUcuYH>*s$PGLkmqm1cg{
z%TlVAdw*9T6bj|1a<NK!i#h89`ZU7(<&O11O>X@RZuit6;4QeNhpkYS8j@TeSpOsp
zBD~TZQ=AT%qXjm+RU7ww17Yd5y`r0)<Ty}H#g`yFfLHxR-muG${-y)jk$;Sxwm-{c
z6>Blu#8RvA+xyj%pcF#O+b@e#pKgs;H#O&~RyAABRq06Fssq?E5t(f>kW-+xS;GxX
zq=6%UORO&n#gTw-sdtsDdhG)-N%u=A*dO3@T!M5qcLscNz0>f?+;*AM(VVA%5}Qxf
z4ochA%nUm90=8wVtDT1E*!eE(Lv>-Wwn0Q#WqtFeT+T);l6*KINs#l&JoD-K6Mg?7
zNk)HdFDfktvf~QcPN-~hJHy@p6EmFgFso{lq5CQfKA&?j{lX_-Wgbqxg)+P-waadT
zEME8H56MD4?PO7*DdrdV)7jo@5wDvUbdSE2ixj1G`N6|uL5yQw$LoK_psVNfr0c;8
zSU_+S{WEC3+YNQ8j3Y`}`xCb5$-?Jq(aj#{N2#P7)_<7L;h<8w+K{g+@i&Yem~3Wi
zB;bB#i}<ZXa4-E1@Yu0LJl*@=I#t;Qzj?J_CA)O_uo*q~J%fIs%vb!9%qF_u(7HaX
z*Mm+of{o4U;yNsP>EOMpJJZYFtt`ZU=r9o~vWU+@#2`P<JbHR^(sza%uV=xQM3H;h
zn>~19+S#buU6F5x8r_NBh1z(j;C5C6Z##!J29e^AO;R`y)^59Q4wal`Ce$aq&#3f*
z3aeuRBLTEb*Q;q>J2OZkB_)5h2qRAR?EK2vw^>6y1D9@!5j$Imd7O;Fj+N)G8r&O1
zj%&9&Q)e;usz+4mM_JZk$A{=J7u2e~ik8$5Ag&HMxNR8_S}+_iH!ivwEn*XZWkW2P
zce5XU{nt5UUkEg$nmpy9Imwv(YxY>V0~%6Ki1W$DC$LH7gim<=df;zH`7njdp+MWm
z8J3qakk*<|DH;nY<-^W+0iW=q=kEdFFg(bqt5cI2Z?H4G28D`?k%1>>3A*v1TlaEV
z_O|nxCOq-x&$e0qQ9;}#gF+N2h>M}m%L2_v*n;LEwYqfcKCw&QZiQ7N_-qD$@xOqd
z@fnlM^8&28|E|?T*ak>#tjPSkb2Rbt5ec?wW)bPNJjnUot?nm<@}4~z=6t>~S3M`y
zyD`eojt=9-Ei^*Hp~-TGNphx7U(T|CNby%jnSr!%o<Bh?99{RDlC01*y7c@fG^I>y
z(^%>I<5{q0V0=EHGvOk?Dh!EsnVT)NpMHtK7Ia^HxW=LMKyLRBbFx~bUmHS00m57~
zyh4N4OEKoNNHXwNk2;>(cbam2LmjSOG>l?S=*!<%**u3R>Eh-)BN!I!zfIblIQ*{X
z!X!DO?<g%_lYLV5@Ubb@rw5X#MYC&<#4!GLwzu&O&yrd1vJRlq>tv41Oik&^G%GK!
zI-eZvc3aAWC2A8}AV!A%pQe?ju{Wj_ZuX|X)i03ZHsqeogW~D^a8mg^Y3Zk4<Kgbu
z;Xu_KwH>PtEt}b(B&8H=u+6glJ0w^8^<v-Z(h6_~bGp3VN_UFX#f{jpCtC((-E<F<
z(jzvNbHvVlBSL{_aT(TK0jZT$w!5eLj9&EhRK<PYsqd==t`b0CbipRj3xUoUJv-M+
zo9Ou$>mqM#Et3-yKaP7P0lW)tVX@j`7xpH_JkzoZZL<`lF&w@KrczgpHtb&u5y%c>
zZTc}eosIg<`%y;f)Ex_Y)~6@#3ZTvywOw&hTn%^*bC!!y%P|zapRn>8-c24_Cwhzi
z-4U1m2TRw|;;-w{q9abZ-p-3kuln7ql=8N>2ks`+lC9Yp;@(w^ZY2mAt#Qy0yz(bQ
z@9AZyyWfTAm26xI{>G$T(EU!&ledrmU_=K+wDCCE1s>GA75a=mKD>D<2IDkwA5yDM
zPI$ljvdZ?;jj9HjxYsGiV*J%;I^V<Hr9)sOT}Aqu*>4ZF*HHQ^zrOn6K3jagd8JS{
znRD~GH;2irAd_FLkGl9#ERmMm0lf0(Gs?By|AyrId6yb^EY+akrq3Wn*}JKd<a0=R
zaNZsy2mT?)J8MtScb)Cc$DVJnt7I}{i4Q~yQdI|%rviTuX)D5w4UuHPd&emYCn$mZ
zvaeEq0)N7MF<h(^q6cLnmkG6`%Me*Rs}1_>-~Qqk;Sk*}Xvr!OTeQo87^<nxAT34A
z3M|8p9>_Z4F1N|&yr&Ra%)4<Ce-Q}i1u#6rC5}%IS8)b{z++OD4!PY)yru{=HYdE3
zJ$utv`k33z?;RhhKBi3zB(1vBfTF?vd2SoM%J}e>9+jZAZPTf}T9?uDVJz{dRSE6b
zS@Nj(Bzl;BV&hp?**=3}aRx?xv2eI2t<1s<UoQNA8cYH3o4Jb0FQ!(z6!lgtoq=V>
z{wL>jaoDHw?@ZYJB;6p?uZ#s52ZlSgyUucArL?T#ZGaa^^tJ?s;!PAVv~<OJhLBtn
zIm@s|@}cS|dZ^*y!YZL(<B)pMfvgPq-u?$i_PZ3ku2jrsgn=OyH*<%$_*BX84H=nl
z8=|u4Ce*fAZrv!{(j~|5D>My1PsJB1WeHXRoj0v356Zt2QHt0>+&OydN~BKc8@l@p
z8tDQREX|86<Y%YI<N=Z!08f}hJeld}d}-tof0$=|E<9v>1P8?JHBPj{v<P!vo_Zwa
zHoR$aT$jt;DY;&#GgerW7xS3LK&`p^lH3d*oQ2_sLd4~ZzwzZi%*vXY+VGl5nqIXr
zOr6!#(A%Y9VH47b)`!^TH43O~5Np>UsJId||I;j(gk{`yS@phISf_|&B<Lns!l|*i
z`f-aa`|pEWw(G3<3m2C@rm{!i$*?`JCO(g&aR~n+GGrI7Jw?OH-wYRw#p7b6ZJq%K
zVp>a)&gQYOAs}(Vw+3Na3zxNCjMhSMUvvPQ==}^ceBS{IV^%T?ZU}_QAm`_$VeJFj
z*FQ>aOmhyYo9wOdzzkFjq;=SLOGR1V4G_Zmcn&Fjh4@Pp!y8|^N*j;G{)YR5`0^e1
zf)LkKKUP;AB1)-<^(ePNS*8av18Jr)a$q7FKajs}Q@Z?6;a_#!gPO`ftF@1!Ktj&q
zDg-htGn%#fd-Y~2$9nTe9LkBmSce|v<rr%2gQdGAY}4QA3LgBi3jKF3!=n5JSP`y8
z&HSL&P|w(!erchz6h57e;%EZ-9VN+U>oyNCl;UsS)Rp*p@x*9@bj8=gMnRwYD1I`|
z9pv{Mb-Tbe;jmY~lP&#iNYC6W7sA7=Aull7>=j@X-5*+~J^eXnDz3Kod-Ef4kGc29
z=u_(vIi^RD=7v%BAA+PSF*X6EB*V?O?$<jJtNvfPD2bxoxAo|=RknM+sR0y`BPu%I
z1OKp&1l5ta)@>e0K!Y=bT{P&=k>gUPT#gs`2F??+XjT2VdT^_64UnY>hShQ2t^Qs0
zu|84YBDHz(^7rLI?+ekf=ZBvR1S{0N0*A2fv0vm-343Nq`mn|%DA^QQ7I>0=>0e*A
zzgYcZYs`lnkS;EE@LL8gjM9-~AaA_SF|5xWgg<!;zz-y6k>8@3kkUcG!SBa+9^cA;
zb0OI9%$6!|YkRnYlzKl8MAJBlk^l2etv7oxsu?R=$DKb!<FdAdertawYW&k|n-Pu$
zm09Ewb8LH)-oL#iRvoJiGKA($ieL{2q9BLAp8+q`zhGZL>vspGuE7b|wRJK*i4A`(
zsB*jb>FujL=P(6@V9}NueppQ6e~h0*6<DJEcfv(vk^+@LWIi{U2ZCvaN9F;xUtbUD
zGBD$~D?@V)G~^Zsvudd;jrz)-m@V5P#_P6Dpl!o8t<!=8)yL1kH-QA=W5Y!v18|UR
zS3qyU54@fRoo6IC@aF^+<p|h}{jGwZdk#AGT@^+u@QGQzJ#~RxPRl_cYgqyRbu~ZT
z^Y@P`2;85j5wVSmNW;I61$hPhE(WzTX!9r`sUJb2fHGQnr<iq3k4^x&K053TkfLDL
zh>@<3yV0h^mW?|VyAg|O!$FGObLXwU9+@y=S`#cfH^Cn7<J&AB=xR6iNoK4LGR;20
z0Ht%r9XOYu=agp<Aqwky>V*CvwQ!9D3%a7p;WkUqQ%>e$AUTp;WyJt*jrb(o;4_8p
zNQaWp20WXgM0<to)8w_c+RY~w*d|zv%*LBeI5OXvHAH(6)W<i-5r=J;b7dz%PFl)M
zu$!U<=soG4)maHIL*xCUbnxGjU%-9=IMCVhP1HMe3U_ta^yzTgjoVFomN$_fZg;*~
z`1G9y>uaololXspWX-pyFM?=l86<~KmAHZYhFAAFhkMhNWZ5F5xOn~`GRM`%4@ACw
zwd;7$AQ|Zl=;Jve^>L3flyh5j(Yy!>Gq1nPaGj&iF;vMrD_?|T?$8m8ACd(z)^HYD
z4U(LJlsCG<7?K#a=;W_)g-f+IQs$ILHV-DX+mx`*JlL2(h*gBdhv_#1tOdtBZ^LIK
zTjJ|JB4lY0^4;5?RQ>jexv-wRSwq?DNq1;nZX{LdO-fc{%i{kebuuq9!nkZ?egO~?
zZo)qQ#kfsPI%YXq$4KTbfC4{HB}m=<F4rsyy=`>Ay)yjoAp{4b1XS9g<lhD2H;d?d
ztMB^^bwS*}(ZKcCN#0Q9r%`z`Y1S^Es8}LrZ<itP;lDhnj&~DaasaRc17Guh$2|Vm
z=={G%=jVwL|LaWuAEN~SFXHh$k0toUCB3DkrTFd!Wm#oq5^f{#8jQO~LI;ae?kaVE
zozDvNZ(&^DLZvNJQ#@^%`pB&c_VGS{vHzOo^6v>KqZmK~bdpY?_`k-htdPFs8dx;x
zrT@Rdk5P={1@g+WvNQg;C(gk2z6d`M0$x?E<p+!Mb=uIUZETd=5gK2JSFmR0{M=ca
zqv~FERhg>kffwKzMm=@_K|P+ct{jx+IW&Z>sjV&00vU)h>%5pN;M0Jei6U;%&Vxz}
zL{0ASEXHIeEZ)J&43S?g<?yEWj;*r1f~EvQd=ODNX-&@wURpF7KvV3dp5*--I@q`@
z0>eQ5hxh<MK>js&r0KyDO!fGm`Q@Mg`oH!wTc78ND*rPCqc7?|qyo4vU@*)73=;X@
z%lg+0mVaQ6|GiuP%!v7ivibik!n9M?>g?&3_;~EkzMTyhDS^V5yXcMo8Ej(oUxQ8j
z8cWK`{4VyUyXWeyiG1+?08;t_7&BWs0Wbxlc###{8idgi7zjJl>cMU^Q<ZblU^(&H
z0#LzJfrN>G`UvqIJS9-PNV31d;bVC2><8SS(=$L9RDIa1FUMAytNSu}IUly{8ZjMT
zZZqsi^Y%^n0c*@=)RneTw-P<`mq6w;@*1<VKZZ-Mr9tlF<N;rBmW9vmKO{(>?s?;m
z7;T3`65R=M*Pq97bh`h}om8=QLh``aq=*iI%NeM*?CjqLKBHJ8Uf&=J6OWXf&Tk-}
z0xbUh)fw9Rr{SnwB@B}0G0qNmCNQ657(d<KCz<fs4{bCQPyYspq{Zds@y25Xa(MV$
zHu%V7f_mzR<2{0sfe0NSWFIaQa%%|&mlRq>x6pI7`$iD3!$t}{x=YK+If<=l69O`W
z0ZLYe4djSpK0~<o8lsCz%TR>k73uiM1KV!=rYL+QZ#@xt6&09Vbx6A>AM5xBEB=d|
z82oSKgvff)<R9UO(8}jJ_CNDV)Rmu)^v@M*N-m11XEP&VB^C^QhpfWIQ>`1#F_-$X
zPuo@Nm#gS%5h#UJxVJczA(Y~>sRg$Ww=uBvW5-3Ym^e5$G7qt^G?r>IG=s^12jvIS
zwoaDLmiHVpz5Y5)r`Rsm3uWE$nxk1~v~_q@)=w`|VK4Da9#Dn|HZe}#U45!J<21&1
zzGENK^zh&XF{$_lpD)y(5BNEyr9*U<)UCFQ&A)-cYJ0K8EuWS5qz6l3q6LQl8iEtH
z<Qp1L*`y9FCg`G*1+BeK`ZnB+P$Xfwk!zsa=O8eZZk=ek8|EfaYs$x+A^+;oJmuFO
zv<QcVOG;>1m=S>Mk3o=!He!3~cM(6>;~A4EiE!C6*up}WZZfp%&CShD5S2^D)?;mD
zCEfcx(}wEW?Pw`OT#2Hll$BL#U^vqYbLcNWl(7IY%^qTEd|qVWh;F~IPQw`v$YNg0
zd6;E3zi`{jvA;(5wbx35Mj0t*L)-52z`=)y7{iuXO(7QFTd=7y-{Dh-b;8xw(2{x6
znJziP<r_{K-qXu-$?baTMV~Sv_^~ydl~g1lIFt*WAWYv7vLOS?#O$NbHRvxO0L6sX
zX1RTlvcTc|{^o2In3UcquBgbPUne}c!G3sJtq1lqRKj17soGJXGuxt^-oF$euQMS+
zc*8Yt)p-fRce;A<43UYrZK+r3uuup&;0?S_BnE~~-59vx1IUvIvVgw}mLoTVZqx!2
zN2?SFVxYz0JT4X0;cZQeO~;|$pZsOCtt#D&VY7!fjmTIjGs1ShD`Rf$MT&&R({o_+
z|8C1ApF#)gF_w(+zW$s_jF=9?(%dcJHqKmFrOgxD5mO=hd4yNVkbH=vlM@w`E)pIK
zhk`lnDackO)BAie<r>Q;i{<}#HcZ}+gN^cW#Kk}AS#@3iLq2-2^yd+~0g-@f;HB%~
z0s^2B5_e7!^(4nyJ<bLk5%pR=a=lsrn~zqFDZ=DgwJM@0#ef0=jzxbYiNe!J_U~0D
z(YV(Nl*7J`j*jwd;!D^4f59OkL1=w8t`vP?AQ}&r_ix!@v^Duz#ogjSE;FgfcOhMZ
zg{)%d+x7R4YfP%K{%iwwB({w<{n+GT1JZDh8#cZGk5_5NBFTSFl2s%b(t|ubxHTSk
zAyGl`J0{10tOv#4^HOP?k46KH8;)Dy$UN_#=Yv5BFT$giiuFeXEjGLORkq*Csv>iA
zRW|)4Y`i}m&XG%EsfXpNFwt#tq-h0B^O#Kn92^|Nxz+BV?)}+Z)zd+k05X=XZBQn*
z0uxv>P49&_P{6EB&s#x=R=wph+D{;y!vWvM+mRG@IQ#C%IiQ69NaeI%iC0xrmTKB9
zVu0p;o5M8_*F!kKN1Flb5^3FLx2~t^fwkCatSyHx+T!+-MG5KPR<-;>A2CP42md40
z-lI(Pm`{~aYZ`_~eTUYN@9;HZMGLm@Kx>MeyB|L%j!L8rf{P9b5<PR&mw1iqcN!}i
zP&uE{sA`d>i2jING^+>7ecLqv(|ICE?O<!IpQCDHL^2(Fl*wK1;)av|)WexB2`0*u
zF&gm?u%d>$Z2zWkeBo>GSpCFvP&Yx@F`*<Ra4Ih+2eoh(AhghO4j$R=)6L<Wnwqy=
zYgsnYnyrAiGW@KJ#-O5Qv4vV4u2e!)Eo9YdT2PeMFnG!sj;2p5%bN1w9HNIjvhzoB
z$kr+RZ;}_ptHgfeP09IeiONmd8Rv<#>dERmmcRbL7>Z5&GTqk@7Mro&H57ojz^)o$
z(U;O|`8X(OV*~lX)1W=xl0Z&(|7;$le>Ws*WmO7t==pVfv5ID-K}^!rCe<VS?Aj`I
z%ro?nRN+4Upz`<|qB3~iUr<_KMqSd(S2(K9@+3dBH?6Pv-Kja9CBX5r@0rCYzBGT=
zp{}m7WQIp?&w%gR6hHCp8zr)f@yQ0<r&<?ABOug9wyp}#&XY3YAd^uhbkqtXCMC`>
zXiPPOf0d1ddK`S{2i;LRWjMdK&Y>KAO2ooA!Q7#yx?ZQ%J`Okbn2SUQ+egjjo;ME%
zFQ%8uW)kQ-Nm_Vwb2ARs3=Nw#W5~;_?NRqjQVxhBl1A<4vUp}wjZadzLT~!$t>rvR
z?+GtQ)B+QE>r-B!oGy?a<ac3A%N30BQ>a6}iz6O9s)|HFIA=m^qmLF_G%<(iIAn#1
z3#dalHxchg_}46icA~{l$Br?m`ahah^uZLMgK=qj^xQ+;!0<w<debu&9IJ(yKfDJz
z_TeF#<sp9Y%l2kw<+D@yQhDzhj+&ZJ_1}JqecgulQlws}#f(1VCLS{@l2(>emrJ&3
zE298ei32#~-t(`LmVBHbbIno9M@mYPxtT_u{1E(E?lD?078{ju+-ns<4WzQT70CA5
z)KnbeQT>g@=cV-#8~GS#X+jb1_F%4_I8LDIY}RP?V8y7?f1-91q3g$gK20k3eujKn
zl*4v`N1x`6rOLkK2oUvhR|E1kgdWc0F1tznp>xom4Se+Mh)KyB@%BAb8nlrqlbhUD
zlf_!vz44i)=*G9f{UN>8Au5;J{nqhTwnw!BkEeo{KrnBCDSLdtqi^X(K(vuHSBi?J
ztr0P3Kj!@TBSK$3Z^G!wghP%f<sw9Jztaoss8k!j80)gLg{}USr$}CfVKq=!$B*5>
zd9R?4-j3E>=MW;?Azb)=TB6zNzaZOf<Q!mriH9&lu9@pUVc5td@ofYR43KoV)l7L?
zjbf-34xRH#`K)C#QP<}EJ0(niQS_D-&duJ_b!Up+RSQ@}rBXI*>wEfrn~CC%qRaV@
z*!Epr^LBse)YWQo7KlfkaK-Ox)#)g<ZZt17S7N!ZMP=rM91{m%o0#uyUNd-kM7}a<
za=J1Jlw_{7>i9_+9a}A;_*FrK19I;k@vFbH988*$jA6+Xa437S@zrSZPN~>W{Q3D8
zZ2cLvm52IWMEeieCXBZ%{&$Pny@m1-m_GvTL9;qe?hIfa7Vf0ZzeS9fm3UK%^hG;P
zy1jRPB9aFJMXe$2lR`5vv>N(RDC&CdOd<*&T%FF_p0apd(CGK-oNb|U3*`+n245DX
zVW^Y4GSoQWdt|z#t35n4)R)YVR|a&A$u(9957`!F<a7$i8@#J;EqyS6WT7~kZ?A2;
z18#wn@2O^(#i?(`eAj)-$Mpjp@#ioVsWk(1is797j{WF&ig}5e^~>8(Hqk^B>!OXz
zD7S;OS(9`xZQzZAxSBYOlNwl*h-EJ4ZX(DMrSulEb(ggMESEWSx?EX(>G>&Wz;|&q
z31RIxcztx*YiEVS7#V$ZVNeY__E=uk&AG-)qNMUc<#gtWegNFsAFBg#+}S~|tnZ=J
z>?aZ;-RCGYhRf#cg5Wq&h&I$VF2@uy_|)@`9Gr>K<gVW}xlj7I9(T{GUN{QI&v(t+
z)`=a;5<Xv|B$<)l0z3g;*Msfv-J`*X?Be19Wy|wKJw2d6{`21^N$E-01&t`Dg5*=H
zc=|QENOtP8gQ>Elz@llhn7yTK8J6%-|4qBLQnuil-eCL2wJcsS6U54y?r5LI0z`gC
z)?G?Pbma;ekgkZ`n@o8Id!?~LvNO{w!(#A`swttn?Ou>*<JWJ2n~gZm+jgla(qGB2
z!X)BPdy;GsiK*%=cQYuL-`xist<E`&yZ&o!JY7`!SYm(SnV58ti4?fhLRs!c6Ix(1
zWQ2vR2i;)cy^>igcUAD{ovqZEor>spx%S+<rIk<4PFwmWOeH}c)N(;md9z|k5MxHb
zv)TBB$Gg+DhlcbBp3Yk1gtk*rX4tp5jT$Nk;K!yCy;_$seEa9L*5aPTIgpFqjF;1M
z&Qwgc50)3<J3L0@*S)C}Ln)kbr;37b?uVkEIgsF`tnGzN;}DDR{-a%|2E5W~*xc^$
zM$mEe4@X*saBqVYe2e&(nHdL@I|LgujrP1@Nx$swH)_uJXI1^yK@=CGRR6j6R)!l#
ztNraMz;h!*l*&W$0t}A?G$11RXa{rjo9UlJ`^Xv6<mNvRSrdG<^6|VWpRHE#lNf7j
z*r8F_HbHZFNdX&pxj$Yfz%e9eK1%Lbr&LH=J7#pr+nX7Sayr8nM}N#wyY1O^#pkjs
z?H+l({IfUiu{Mn4ups7%M@mR8(|eKyNprkp-gnOKJcc?PJ%m#cA0-u_A5Z}O$m|HZ
z+W*=rsL#1v@^q{weYB{Ab?D~1%pDMdaXP@9Hnxd9^rEz^EYHsvBJ?-e#JTbTr|;v2
z<Zv><9Q84g5U;NBEMjm(a!tb}J(qXa?;J7f2MyNV-d_8^Ddt~!WR+vE?Osakw>ta4
z-EcB?sHr0aa2vvKF8h>8CoUATap=@x$JXpJcC{TtPd##&VK5L@2QN@I3B4^@75TmW
zlO}qhG32#}L6j>x%Yddo?T|u7sOY=%m546|s-)xgmONg9`QI%o5!~3xrEfT!OKMly
zSw$~yr`|s-dctq*2iQYMsl!~89Vf@urezm=Jy%i|8cWjbqupUBI`9nKDZBv|6?(s>
zkg?{RF!!c)>;`l$!xna1jk;lf{KLE`q;kBhUZ8OdV;4LL`pmjiLE?|}N~ma)$-{zJ
zt@WWf5M@`kHDj7pD$$TiPTy4TOo^3mB>2~;z2N2hkjo6Wr4n6p_d&F&)!F%r-KmMq
zJ5bmt&(7r{os!)GM2gd_!my*g?%B%h!HUBTb-6Pk)eXgJG1usj$4sh-37lw4Em%ti
z6Mt+LnQ#_()ZFs(*8C(4eA*Lzq7N#$u+|E^JNZ#?agN$?GQ>}Q$!TEQ|K>OLd{_D8
zLAc<=x=nrKU{a6?tY>40W>G^H{`{UjW4moDx1BM=_dS~^mx8+@AQoKnX#Zr}<g3HS
z$?dhK?c0Q9-FhpUpDFB8<==x3zMF8bAiol#z3Foe<K``!fBjveBSX(eQYr8LI2MP>
zR1f~DMMUUo!HEzGou~&S@D3JWJT19+P%i7*wIz&9qvkAMOKc%b$C7<MA4{!SKVMhZ
z)HieI#Uxw<(GhX>5p9`o+CebDSk}qD{?YuHQRe3ve>MX{QWh4K`(w#pyIJi~&P!%z
zTe9x?=}XE^#ikFLuE;eHgKyTd#T2(DuruU4J1RNb7QO=0m%wf9e!&(_@^FUt4$|mY
zN}n(RbD8H6TU97;2@z~lxaOleRQ|3ZB%Bw`7X6yJ17uA$^O4;_IlDWVOG*aX61MS0
zqKZ*1RU+M(?mzS?J+fR_I#X)i%-g-!j@`-KQ#@J>W@FXHu<+Qt&pgGT{qlprJUxg0
z7w5DdR2TZB|2#tDQTyFWPEg&%cWEzx4fxIMTLhudM=W2Lu+x9U6<#PtsO+f7c-2i`
zA{jy035WyOAYw=OmD70ORq4zW@N!OqCvoV(INq-v6$ROiQ&}fHN||rZsJZWwahKje
zmvWa!+#vDbWb*EkO7D3ZjMOQ@f3@{7u7eJhMD;687`6_JiC4x$^<0ONwsa8xlzbbI
zR9=h`@{???83D0^_=So@)7Ljm+D;aA$Ewn$no1al0U%tKQ>R9hrAUiQ`>As<8e-)0
z76~zD`IHg)?|>tgi-AK)9uF7ZYsK{9$k4&8gajBepL;gv-AVe_d{)LCCwYHW3mJ(a
zFls(6l9!U1bY&4q<=Ai2O5m^N#GWTR3zQRd5Y=5z!}OE5rGNZcbFGPLoB|E3T$F|5
zp<@Q<c|BEbM+W+Q)#Xr2?&~OvJlwAy`bOzS_x^nI0}64sb!8iDLPCP%pmVp<y8{k%
zc7&MvEpWJ>QdpM&(^uVN@KVg>VVkylCivWf!sS@}>S}uhV@#i>m0=ej8F0Y}!6&yU
zWW$XVk2|emuPKz!H}yd%j+~U24=RA+G7!SIg4{`nsXdJqQM7GIq}sE-`WkOY<;PDi
z02SLqin`Dr<Nclrox59<hg<jFQJ8_{ZIDR0pDoV=jP!whqz50K$wQ){9&Lp=6B;oO
zdQTfkYAh$1uwncoY`#ilq__+GS}y?w1eI-MsG^c96ZLYfr(z)W40IIt`dSFKxgm<D
zq;ub4PT};){~kEJTJs&=7Q(MN%yK6GQN&$yPl%Ec=AEYa4S(#0;FtA4y@gUHzjnh?
z4_y;iwM6jJMaS5|8gR8XtgJBcWfqQr(wlzKX;h$Dic-=9^MQrSS@t;|R!ef{iEn7z
zd%Nvj83Cb7XVtJv(#wH~`7x!|`$CHm_!7ri+Ir2diFVl{#F~fIUMb$iIlAMa4wdw7
zveIfMbXM(Ze`%pFJh^q-OY`6X`&6?{YL~QC!k+*9I@&$Ay|qp%4^Bft3Pg(;GDxSH
zvbkS8$C|-g-(hcR<J(ya;xTSHUl(QN&4UadfyTkN`76n3qIM)P{u*zXsGcb&Gh=3s
z%eI}Jo#uQ0FO^b%J(HKOdFh1*C!%(~x0Iq-XH?Nll<w&ra2t<U4@5*eQ1A!C>T7Ay
z?R;Xwc$%TEKJRiZX}>>}Ll}Ld$(?V}F`Wbj^CgQgyWV<Q)_t>jX?9~wE4eK!<1|*Q
zf-)!fVmM_J?q7UH4F&ejfL9iCLdT^Ly`J`t^>;rTpMZwt{QVF94r&i&ZFz4&XOlAy
z)EV)~M?zPEN3myAccd12mTJIo(pV#kKRrRSu#8dGjnpDrs1OSVP8i_H3E9b=>q4e0
z4ejO(jWgCTx`64a0cH;vacDR#ehVjU0ZzaoEGSMKvM%+@S|P$e?OWa*viB&_%!G3#
z%8eHne95Wx4Ew_`HKPAy0jLtp4U}2Sxclpo$u98cYWg%%I4m?e#A><bDA9>JaNu$0
z;AfO${?sIXX)(|r5ex=NYGJV~y01qx^H9PD7ii(JgwV@c{>~P`%+!ahh0w~nW6E(z
zjvY11U?-H5vD~(J9zixl9wI{B@MH^|GS%;^&iZ2P7w!`#+P+DaNHWgeGv3O&bJZEL
z6klqdtQw=B^1L#g26>lwB~Uv<29K4nX9HOx4s7p#{!xI#ygmQts3Hi4$EJi{EhW~>
zla_^*%wUQ9B|-wbON-;Ctk`0dOAwTvyTf}XTQFv?nIGVrct}JlIvpoeW-93FkZQGt
zjMWjv_bfUZ7fTVRz5hBq!cAp6FvuS@yEYls+03dCa@izP|B;YF?g2EpOzp>?vB!l<
z91NSzr&v-8y|9uB*U=PWdcuRr67R29voNXd3At^V7QuKw%)?@CM;^~HAWESa1k*B@
z5xZQ?4Z4`0pWoi8Jx&Qdk>e~1=?opZ=T^jm9dtCwWb4h7NL|N!+sMDhWc&~`DbfQz
zu7Lx}dRG0Lu4(58DSlCuc@y+s(zct&<fqts+;XH@#f;rB+y)qoj4GZTn|S64o3~?L
zp-``+H{a~a9i#a^9~z6z=vyB?etL(+9vV+TYDGHf?0G>(QmDv!H%p!rbG$-e>epA>
zKXx2x3mxyqU$_IRq@*4hN<xF@4jf!D$p->#h?2lXd|`R8KA_t2wSGBg0VVrT1@C@B
z)xdW72cK{719w^T_fh)fc+NC#Sr5FdGW9xRhVjanDDpYOu|=u`Gu*szb!+22_gV6W
z<&P(N09uJuHD;N^UqaWdhtz+TWGT9D8D{PJ;zQ}gx|`dHYfftAI;Z=66<VWXbK1tA
z*j&wl470Zlgrqb-U?h&pZE-{5YqG~O7TfC!BXVL$^YfmQJ4yfO{6-<fc|jPX5z$Te
z2!eSM6Il&eR)brXVB<g(@x0jiSArbue%xJ_cb%KB<M*O2DpVv}$spH8O4aP8s=Nym
zQ7t{htnfp;SMm`xy-rQN3h^wTdy^xUPjzo*F%`7KXctV<Mf~u&22gyJVKlvOa~@N!
zj9vRG&2UPYraYfd{(oMFpC#338jjalatxf*Fji(PpN1tp;!jNjhDn&EqH$K$%@CiI
zaHZgkNJv&Y{1B=5{$3UF-Lysc+!tcoa|*71?Lef29Ej{o=Vs2z5eveTN_uz^ee}%$
z^HtRxd!8%Kl5ILd34eb~jl+q%ZP@Hw;vlJDY=47Paiw~OJx%S^JgE90P`^w)?0^%n
z_bYEYX86d=$=owjNod}%@r~io!K|{f#(5ADk}(SqD*r?q1%2bLNnBioE;*G;E!Xv#
zf*?K@fS0^5X>a4xzO@h+U$35InvW>m^meWIMl@%dc1*m)=hv3_OC#iPIk%mU%l+d5
zTu`RNdf@`?fKrF;1m|$2o|L2#mCnDfD3g$bCB_|u2h%uwFLnlWwY4qsyX|bheBhxM
z|ML$%5jyU3+zCf$=8K|`-cl;P^}XNj{(aNVgTu~TeOX@a^Vsj^Yi8i^!Z2hvey-Ef
z#`@ILi{o15>SVc}2#=y3Pir4`K5a04J{%kvWN>emWfE%B4#=&}bM|oOomioTRM*W8
zZD#EpDWmTxGA2Q=kr+qEhd}XozTZvy=FdCC>f>KNUdJ>v!yt$SwQhSVrLkKyr3wFY
zo>bJI&9ERe_sy?ASQBejVghS0NFSIytQ?ta_+`reDc}T$@SqUlTOCiltJ>cPQqa=E
zfv&WOjav8um!)&^!%Vfw>X;On`VVSFqnB)yFO1`IDHjx?makVgW!$S{=L8<kC*C-2
zNV`;h>PRjKUf3;Z-+$>rX~&wbtUt7ji6{bzxHI3Gh!VMx8)ZDBpVObf5eh}}i%nQ%
zxxKgXNg!hmTyt=1O!o3BujvzjCV&c>ks~%?5IU(%t1qog6dBWhdWoJiN0;8{Uiv7e
zB<#uRV>_<yt=oU56ubPdZb4dZyrqDrm|y=#x~_#w=&{FUw<A>J(S$-ecedX&NqV^m
z4MgOMQhp#KNj#32^{Bgj1_=)uQXQXX+ND+OZvG_}+U_^R8X=$9rS9#5Ihi`p*RHB^
z*V;tpnfaEeFV~~Gu0tH%VOsHTA`l&(rtI!}Xavb4StqffD;fVFO;d9qpzl3Ua+gcc
zHeIiU8J(>4j)_#Xsp5RKCa{lD?%iV*8#H+;TAkSlv!kFdYk$4)?S;@hu=3eXrLtv}
z;us2#ZTsxuuMfU|dg)N_W!T(ldatfCA|wNGK{>S*$mzry9s5x&vsrXMq&-g5+Qn0e
z>`1&ZICB11=bHRD&Y{jT^dd8Qx{PXt{r|A{R$*Or?b^2r(k0!EbO;DYcS@(EbV+wB
z-5n}OcZzfg(jg$--Q8W^`0IT?@4cR9ZLFR5`!>F{>v4E6|6`6h=eVxx{GI1?PO2R?
zd&-HXqXOx{@9fvM?a&PvtC~rm^d?HN>%{xK3Zl}V+P#Q=KHLAO`3T5>FCIl0RcTef
z!*kpjmn9;MYxcT1|7y~Y{}dVj%52wT34jM-ja+MQ$``R{+)f`OCwHb5`>uA%vM*Ub
zy_~wABy}koKfMZ;{jA{G)!goMkkAWlAj)WY+ZO#|k=59><-1Sp?fTj7qoqder<vv(
z=kH;ZO=-c6x}<HkpI1x^qi`^`*;7uTMgeC1#w7ydI8JYj-0h|aM@zKS&+D&Md@6v9
z{3|#dt2YZ7tCgEQj(<&9L^kNYs0fCP1<z*>9zjCy_w{8v9Q>-ab|KVv`?{={z1<aZ
za75~RdFC$*mmx3bNUhgKG2tZ!JWe+=;G?M0fuGUgUd;lXWc$@OQn_+P|GKpdI_R43
z2hY=v-5k;#rzw&9I!Q|3udc&dw$33OR@{{f?W>iTU#XO7J{GH4KvFUeU`3BOY9R3s
z|KTVi)$@yQK|svAt>MAH@%x6w`ya+iwc5Qh8{n?3p(X8qsI7r&pSFHQbKD3DqPKua
z4h69wUe~9O%ade;$j`mR5+mrI=$t}-NSoWmL1{oLY}P81OjTjVz-N8~FGb9_r0Dsx
z^>_Ke_}&SfXb^c?uI|?+w)&m0Sim=m6Z<^}$M3hDs93~w{8bnmAyD@(GPv!A5rX^q
z6*M&R-wrjn*q|lBfrUR-!1C2vaEU`Ly(+nW7ZPhMjlh1j4J&yaw5#9pvU;qjSuHVU
zlitfd)|u|2I5OJ%L$y_2wK*znNhFzfPs&f5tFb=WvuT@jv(X4r)p*NHwgW;SUDPdu
zj|zJ46MCE2?}pu*`lav#`_6nw^ovad+1g~#@@_6Ftw(?)pouBA3n6?3CBE249aQ|Q
z{$$Re%uMp9@B`D!yCv#OG`GdQ7}MY2kD12?gcljq$GxVV&ed)kgQ@$<hOFgdq-M{I
zFWE9r*mx^%<mYs%u}HqJm~tRBV;wZaBeQ)M5I)#;Z8WzBwUSzPY5PqCM5yRA)4-BK
zlpyiiaQnDvuX={?=p_Y5oKE;nL+$4;Je0hlOzt>*-#7HULN>$>8}8~<N%_xhNn%xZ
z)TCZ*q-2`&OnTqeYlmv)JwRV|V0hZs*NiPQGc~k+pRoj?Ki+^4#`j`IVwduA|1@_T
zw#J0-J;{F<iO8&@aQP;oz-XG%DjA)XQT%6#q*qv8Z<#(24Q^_MbOZr~EQ%HHFblA+
z&4Rl>Nqca;P{i6G#q{Me7E?A+$t_YJvDs@g#5h}pkBU3(-SPs>l%j+BUiFkE-NOS*
zrv#g+z6Yl>W%r*zr0ISY{4yNbUaD~Vt6VE$_NdS5FTLvpQ-grfi_@WRg%|rCfp8?F
zv)uGwC2kQVwB;AEc|Zz$Sl33i^_*tb35Z9=cOW|?j){}wmTE(^xu5+$H!~yW*E)!q
zy=`HqCjuLxy@8ym*JyOjXOInPq2QH+_9vvzy7EnEhGqg7)KNC(u+=|aYTzPPMDL=U
zUR;!mB|T)Tf1{kjE>hoZ+8K_APx!2lB8pe35WBa7KChIgYN6<0IX}C$q!sI&DoTEo
zMtAp@Lw!|3Ylm2=XIr=w+TI|8fb$Ca9BJ9TonQRO8PZ&}pJ~GbQHrMSn0HyP!rjP4
zX*cuz*3b=GJfnEm9exjzlK0UtuhW;MmH{Bq$5tMb{lqHP1~3uh7cY)7C{fS_vToIw
zZZv9i67FoOpGlD}ZP%;cM6mvSl%7hrsa*;TKJ2oO!@&<;ZnTq1u`DctHW@sx8G`+g
zrm?{oa^_yYhSTtq%I;_UUg8IwhedRfDwa$iVBb=GrGUQe#l{`pI)Wm5zve|rSVEJ0
z@#M<gBjt(?bk|FlJLJCKmZ<L;b=Z9-olaYMjR+G027wRPahGl5;x&8muYG2bUA>l`
zLE1FB#+h#?*U;y^HO`L-Y2(vyPE3V8oMpre#W$CuzG>aIeEjI91P_Yy@_j9Cz)-T+
zH=B>-4i~B#{Avws%b_bMD8MJ46u2pW$1ZsrzDV1+CvG)kv7p&Xpuf_y)$`8rqoVGL
zvobYtoC?RCog$1ObwZNR8g=iNu8-VRC6oE1%aVq|&*J#svSg>ukF!UNbdz$5EgB2B
z+C?cKyf+P8dFJ7YU5)!m6_cKm6X*N&HvJ_iyeT$9AC9_6LS3gQtdGJ|axALe2+19E
z0vyLy1jYL2kOSZ>pl1Pj!~#;BB<1BNWYAyDqq&R%%fy?@AoaLhpl^v8QEiVUl}*ul
ztag^2#LW3!hb5bN^!8|nU#*}(7@J2>vq3J}{-<5EedQ(5w6Mn~c`nR4uI%*<67pm1
zv$Lc}V`KGGQf((;dAx|Ea@~j1Psi?466wn%c+hrSu->sUHRFHl<V}Y=nZ|(jc9+=I
ze|6M&eR@(!@Iw;U$VgPqlx%345<%J<(FXSkm<8|m8GqrvS*<ZNC2Q^y`H~eCuquKw
z(yhzo0WPZcabY5}UvJd>SgJ`y^G{mF$oGEh%;%^3@6BBO%zTCxIJYIfF4O|jTmSa*
zs9pklf2slkx0ZHyD)8~~fzAlpO;o19T93!C%L0%B#!x1Y^T<w8xKd`v`^R^FGH#Yd
zm%`i55~0XEl*NE2x^=~{)FPc7evJksO9V!KiOBX&j}`G2Lr7eW$mWMb#2Zq7NsKJT
z_&}guQyer=ANp%P0K0{s0S!DYH&3WbBT)&k{rpv^TGXAE;_Lfl@|&^DY7J;PiM%e?
z1vC&rhBAxel!z*vTtx;;_u0n~0kWPm0zoCez(>o~8s(?Z+30qnAEkS?W5&(B(AjSf
zik3BsUN@OdwE4dCkwb{<hH;a-8Lys>#qf+1y|Fs3JMMl-eS2!=7poT_RGRccKfCXk
zeth;rd>W!s-8nSDOWNLdZ8*9?++APNBaeTQpdQnk6fS|p>|cx&sU(qkAD0fm==B4b
zPd`QpwwkL~4VTzNm0VoJKkw)vo>yQn<sd$cJMkSn#xWJ%I;9;Fin)N1=SMZy85r@0
z#S{F~a;qXGsEiRJwzQK7$9-%-l)PkSTYPIEMMCNAI!Zw-a|v`N?XkuqmHF?F%*67N
zkUn`E`B18oO)gm_b`FzrQgDQb7&iV0uS?a1&c64I-jn~5qyaKH=p}v`{1}j%_6Jq(
zqipH8(L+)@_+PJEv@q`+k4T#hl%EItp|QKTDcH7nL)kc=U*Ds?vN?L6mV)!5_30Iq
zhp^er_%x}fl=V0veseQnOWTx52dtHQ{v9X{IUeUyB7PI%dD#sjBF$Vu@!@1cl=W69
z10G%ACv)n<bvH%Sdku9|kb4tPK02HN9HU6bBig`-PMctT4N;Yht3PfHPAi<GCFazf
zbM^^dlFr`Nrxo`pg!;R!wH`TKf>i*jnNz!)YbPwMV%k|Qljghh?M2UKf12OmMGm9B
z>PRk;h``UOW9deDWI%!*PQ-(~590P(T$;6cm(-iQUjn(5it=gw#2q55MC7<XERo<I
zaQwXZy6*e#*QNY5IAFmvqVQ8wcRAC_^B%Q#Q*pUIlMB?K$IC`UZ{D;&n5|<ocpWBN
z0&G%!#3Vj_U%|I)G=r0nv_)=Ci2lqP_F{NjORX>7;CA-IssW#C|Db^I7-wI`wdE0p
zrOY0K#H$v;#*nk60@Ku|TLw~ZuCH*OI-h7ZRth}j;NX21T_I$h^=b1t{9PSXVe=K8
zy%d&iVfK7jx=c1gyTFjEluT!iNQZ7Ig-ZFo+ls8W%kJc3Fjyx2nI@p;FqKM7SIb~L
z`Bnd!2w{@=_1-ePKsg>R-GWxWzu+lG1h)4RT8V0vz^)Hb(+`lRQyEjKszOn!!d(j8
za!K&k?_#(^#1A{KxCLSsW?;&hE<6cR&{u;3spPzfK)z=T3Bs-0#ur$#=&rCZ*xStV
z76MK%+aqN{Q@XX?)5{W%Am3@VGa<X40N}D2_IF4{F@uL`Dkj{|EFfsM;S9I2F32TO
zT-KF0nNER{{K2TahwUmF+n{1YKjs3n+#IQGdFgt^)zZg#M?CcBIt?cVbq<`@#z`e;
z^-HNHIP2R-GxeNIeXfnx8-mT69lt$i^S{^)WGTfN91P<<`W28p$}Y)|WPEYx^^JZ$
z4wZ*c358!Jr8s)~k)Derc(pm(M{-&NaxgEZ|DnNo@5!w`?)eXZiFoeG&xSnPQr|s1
zC70D+3cA6{{#y8PzAt0jueu3Duu&M-%X#9xtY4B(3WmahYhNiB3PtLy3p&bDj|RdM
zLLN)#+PgP~Mv1;hY$9DCVPk@jgr|J=emC7HCGd!?vy%=aYy?6QHdMtiXj2I)E2d5U
z63E8U*4_0;{!0BVXEwOXF{D<uzln=U=liD>yOxr@U$uWXFjY52h%yIH`A1gSuaee2
z19T~Wi;i<w&#5VdJpQSruE}17v4W2QF8G6(#3Pc0!Ysz0-vT9niYnKN#2*9q*FAgk
z;=bob|FB}Wmg+&CMkgNuF8H*(xiKmMZ+qE)kOMt<k?r5I=ufdTY2na&J^z$N$Ho#=
zQi<jQ7kp*c*AxW^KjO`O#q4h!)Rw|Y{y6TlQq>6H!B5Bi6Dgsh;xv;9ehpmka}Pg?
zD`EN4jzEX4o^KSl{;ioclD76?MX#j(r*K<F+WJ=_+9%+G|9a^~L^d6?axOe}&)(ph
z05<PGj{7|zBfyB#O#YwV`HAtf>5OEozy<$Pt0BIWDqQYs3<mH^DiBQnyL*udX?8jO
zWTdSV(h@cTF8FK#UF@-iXs<s#*PyHsFvS}93!eG^%MFvAmS@;ZV4|g!!uu<_;-72+
zDaF@8lZn_0q0)~3gcc?s(1O!{LJMz60JO08-#`ohZ~O-q*6u3<(huS4rJ%G9g<8&&
z`?5WjcOT<<W$?tk3IRFO%)>(9i^H-4WEW2qKWttWtf}thImi)$Ejh6L88mnWe#<vf
zrbs!3I|FRh;06SZ0tCC{OCF6vtUog#8@cq&!_bj3(1ifoQZBG7Q^sEI04CwL%!UiR
z?jc%)VziVG-5_x%P4jX7o#Q}=2xKqoRNQB+<N-MguxhoVfjobf<35hJqsN8k1a>BW
z3Yj-#Tu?DsPr!9NQMlPk!3ctE=X#Sni7t7d{Q1v|=Uo2zr~h6dYzsq94-Ky4Y5F+;
zGf=>@Y3kH}9R;od)G7aQ@iy(iSr0NQHVs%r5><l(*8$W$|G5_baq$bmYy_*~-L+@~
z2tV+Gi_yR$V!i)w*5W@dzWJHypN?h<mM<IH8?s+V%Jh%^o3;4Qi>Da)qyFKdhc+es
zCY-*Krv^;u|My1z-i7~P+{lgJq7}%C0|T4rtswRn;Gv*E1MPEnhO-Mg&gV6zgT(oI
z&HRMUllq~0!1<2a?Q35pI6Z&$zPp}Y1=ZElp>E5&TZ70J-~js=hz^QmN1J_oG%C#$
zO9Y~mKI6mf%|R_SJ0INaO(Dm^$ykqT-=YM@V+V=!#*{o_Snb+r>?*o-*Rkfj>N37j
z#`>C_4Gnz5O#d8j>e0^TGlGaUZ+L({hLn#_M{KJp2|V0}iF(}5tj3L10tNZDdJE+j
z%y28r9;lQ#ITF_qUUyyJ`6+4HWSE502VIAK%xp6IgR1qMojG>r8aR&`0CS%!n>@bF
z{pJ7!JuoZmg6P_s7u_{hGce4JhfjWfeo3!H>-<X`YCkvG+84w}3Q0?c5}~V)K<A)k
z_ae|V+T|A+3z?XIsTXeYSy<=RqX<08MiXvu7QX6x2<Yvw`Jkd`9ls{&3f$Ak?v{n3
zhiuR&IjyWej~9Utz3u0;P%#_9j@F5kZ{^u#BiEGO_!rT}+|=;P1$SaeS&fZ^*ucpi
z$fD^yFPDN8R8_HvO5Dn{YM31Nrm@^F_KAR}WvXlD^9UgR+!Cnx=I`<V52k64UN+z2
zUP*O`eO^7g_iFg%q{F`C4FfW@<$5(ZcFglNkLsAhPE&D&=8A%LAliJ#o47){43j(j
z{OfC(e9e12|7x2m)zQ(B8R5<~b&+c~5L2OSHRKvYBYBsUoQw!K+`_vO-6s6wlM^c}
z3}aJB><~a&THXnOin2*4H7N=9V!_9!9j)vcMGU43h*)j~F04-h&X5-!XYQ*67rXh@
zfhSFRpyrwWwb)(8eLHn{rVc8q5?OaZOs{mtyK^KOpEI%d+ZL}T$1x<hStaI>K@b0u
z-vCJPs}W3M^LLxs(s%2qWtD0%5~g#B*i-Hbd43|*wu{^-o9^{<F8K2tG-6cSER;rm
z+e;Jh;n%k7z^Wcm#(>FrSiN0lP?L>6ia1z$vLn}Y-=<8~<Rb7%mUJ+EHA!=F`(RvT
z$u?u7RbL0&?5#<Jn;_>vQH$Vo+Q%PdnpIEe1T>LU`}koBio>-VSN%`}eroMM+I1m&
z_33AjG;aL8h{plgrmbtEqs0@$Rn~L(&*f8*xVU!mvU+W_3047E5Q#e^$DOt*-4cq&
zX7buGI`7~E4YVRT-~tQAe~rs3IL^qTO_%pSYpWz~%DDH75{iu)F#`b<`RTrs%zlIt
zECfG4?yxuHo(efhx_@eYMv=0oRco7}>-bv#;b!4s5#&2Xh6KiP(cA^O^^^SaW2gbf
zLuxZz=-Rd&9ca__Q_(DnNXMBkDAFyqBm%h<$cG2^5zeRw533$pJi+={*R5f@f~*W0
z1-#+(ebLk&{=4#w^JY8CXhYc1azhDV0-QzGt+5F*4Fmy5kpB7>qra!;agpLFBo?H!
z>0DDJ)mj{Sj0|+P8?gbE^1-(VR*;Lc>*E`H1WHcpC(*G;ODXRjz2G)jRT|_%#OGzh
ziDli!r?PvQQizsf(#q_|3+eOQ^{mNNSW1(q1Z5A50#_Uqh?H#cQ2+tcGo<6$eXGY4
zyGn!vtK}JL+;?^~K6Z(;imkY-H7K0g)rJVsVn#-fmV2fiIc;lXJ2G?A+-(F1MdpVS
zzCwopNEeS-mqhn`R)$Ht3PQpA>j#)v<Y^gcqq^;s)QF1@0yIQB87ATgG}Tdby|0Wv
zvYDbR22xk`b$7piX*u3Po0pOqX+K9|1LM~FOyZR$cg!Ev{<u?0I4(>M8Yv;V<zbos
zHnL^a8w_z*@dX*uJzxH4_w_fhNbky1a;|S|C`RJ3Y?=_bEV7;PpqTuMZ-^CPJ(dBG
zkoT#;)o}j%8lN2AULNY+ELrZ41RuxP_RpK0GHSjP$!Sk4eH176X6rDx9XHXYeqmy1
z--KQtRF24n-O2&cbvGL$BQof|;$JPzCxdk(27_OegoK-EN)Y>axb^+3P$^=VH7OGm
zUG2Wo$PYNm*RFor#ZSAM#QrO^@EiqxvicQkgF=%&kxH2sRc<YnPVvqkZlR1|L+q`>
zwY*jI%JVblz3InHd}<`K&#+~(ZJ%~T67i5<?*mio0HBEJK|`rTw6m(4h@5>uSW%w{
zK{702x!ppq!>oP#<H_7XZf>qL(CAd0VJQm5vIIL_Ic^PiV58=R_;NebYO6gru6ZSX
zf|2)ui!rwBREVfS1@tVDor6vHpMJ?MbSh$ak$=CUP79<WRlL*0^r{cXSyW8~@IL`&
ze~(gXa-&A8^)}bKuLCvwi0TF?<m+sgMQNpw_&2v%^n+#COn#z~Tgm5k1k#v;9QN3c
zPVw#U18(8v)4=<bK}D{H<FhkH%+kq>OaMiYRL-DWLhnLKS#s`jn+tD82g*rvz^$69
zlv|QccFR}LQhjn$9m>+)<a+Gq9J*%#8nJpy-2^D{I-8}wOCs6FF4&Sw!%AAlZTD0*
zzn;Rnie9Yv>WyueSsmF?D1ybrT}2<C1C!C5L4gNt^yG0vx=7sag!Ue#dUxgG{OI-&
zl@~`@l(~O8{gOS6pT!v^?RANLZta|ul$*JpOIXY|)>(IWUackNR$7Y|_K`*MIr^1Q
z0edx2QN2d%?Mtw)PbpOl2Ga-fv3GMul6jGEYwMDAd^{GHnDgFo+y&I$Z%o2dftb}e
zH-|0P?67U12?+j^nzIL)@=>G)`gf-)oSc$yUnI46*QGG5n-&vk|0J1|M~lC28_W_)
zoyq#5!nN#PaQcZlyLT{AGGoq%F4WiiERF9JlpODSekpUG&{BC5$;Vq%zl*mY<dMNk
zHG`>Z$#8AuQ|iwZKq_YogZmerV8!MmNg{<1)ZeWWhA91@s-nWh1cg7xs8u~WmIj8<
z&OaAIKPWcS2N4W)eG&Q&Xi6Qlj>=+x3i0xKkomKHOz_e|rMTEY4!lCzVcdSIs>;0o
zQlm?)!*NkJ&F-xC)o|hvJ&RC9%<7`yqa3zpmMKHy6Dfki7<-Sjk{_t1#P?f7=*x@)
z6A6GA$}Nz1VMQO3<N85Gs+a^>`az#x#AmPkClkSHA{yTF{J$}UD~2z8e6jp7$j8S4
zW#;#nc)n)nAND5967wud&KOGIEoT&k@K&qY%wi6~Ofjb2ZUUcZI#Uc0QF>!Y{H_$J
z9s+9Fx;to};6Exy<C>GU<HGkJX<$V=t^hF$n9eVTe!p0Wm0uD`hKr|bBxZ7CI(Rro
zw_>fq4o`VuP~dIT>a6PUZh1Elc5kgZTc1+0oriq2zhaf;BXy1!5FcfLog#|>IU!jU
z?e(JYJfE5&Ydh5eq<RgK1;Zi6N)TM3+jM>rk*d6!G?pa>*&Yy0jp8!YGP$z$?GsFP
z(bt34#3jQaD=07BF>gfeenevxWVvQ}tD|rl5WB!!K>!fM^xC80e5AxBAjtk6NG*5c
zzUR9g4RS5UiYMa>HjRwh8dtZv%#fa{UJ;zI^EwRhDI=(JrzNuW#WP%r%d<_p2GfJV
z229PabC*oUI~y!>$0@(<Xj?}h8V<z&;xuKPBxcIIzu-6Q<NN^#^O=--9WV1V#v@`j
z-gHUj^B8vl8LEGUDIgVuzvQ)OB1tb2!wG0BTo3kt3sF{O2N&KBHd5^s9x^lJ$|ulv
zQG<E>w6#4m5Enrtn1u>VpRJuiSRl1^;xTCRgO8b9(bTsqhnKym3HoC-mf+G86!W4i
zHJcp2EfVlG-Bv5l*WT;$5k;OUxXwh#z3CDK+s@xZcJwSTYh}4I{^K!;<8u0G+MGQ>
z&+{@ePI0t};0;G!sSIiv;v0ddj|o;0xCq^jVNMm^M!0`n%%vwfs+1v&zpqFZ;ModK
zWO%sBvlj><(k;(c+_9-z2v(oPTLpmJn4`D`u3buoA|6AV$JutyQiw91ZbLdZB?u~h
zsde1mhQAhsz@okVuD^7@0E##bGZRv0A(Q#!hv!<*CwcA4!<_q2<)dr?<IeHTB4R%G
zOHF4^er(7ecKNNj68JLNmH+`1h(38J;zW2G`SFTt%f@aX@0Yaju(FWJ*&nq%&3Tw*
zm;uJ;?DZ<zD3nYJj1DJk_`2KoYJ2}$S-9*5G`*BBmMSnKwsvwkj++@l7Iem7WS#41
zEei8zg`1Wo_D5(02n(FHkFV46fN$Z_E4iu9!$qJRwUH*Y&DN#%-?_}$w$KA~O^|=q
z;&#1oGXM(e;cdaKwZ@wytqB|b?r2=rm_B4!J0bEfC_%`~U*aLOb3w=y7s?pznmKTH
zO4=Q`cA1?P)Rz6gE{PL$=jH{Pau=|wABTWI*g}vOh^-@f-{Rj7@En-AE}b5L*YZ2*
z4RIp{B6Nb)%IUTW`z$QI)g~`X{0bU6K0iXYETP4OFl_H8m$L6W@nuWg_tYd#z1$|(
zq}<uS_|5!O75Lgv@=Dwutowp(YX?jhnUrExDh!k39nNVH?lc?^tm=^fP8C44-tsF|
z2cdE8dx7J5_(wNIINx7B@t*iY!F@y7?q}QPr(0Q3?h8qqXLUpR;%<?B$=n^^u+Ruz
zh}|d1y}R_!5DB7#lJ)#2Ab?==7%Q-z{3oJ(<J{=f;U}C7_dARHd+Vm!Z7*j3lM~Xq
zC86lQ6#S&lzZLvA(e3Rc2g$9{#C52LiLwDQQXg`Qe3YR(wtEgHh=hGY38+;|e~i9~
zcD6=U&F9fQmet?)O#PX&nqjAKbiRXsEpEdUGHaa|a>&wha2_O92YAq>tSEk^cm$hm
z%)Fw}z@><GV1}%<s<sUL%O2f(-yTVr4QIm#uQA1F1s{b7*a^V|XEj@x-+97=cNC+c
z>mXD2&WLJUip{o)5V@QpW8+n!SB>O;lr;JsOs|ps%JahZM{a!6LbarN&r)ePo#+tu
zektE?hbeHcks&7qR&0n^B%gn_t$_mTX`LXHVvGX75#nO)I;9^<gj9$=w@0wSvL~Uo
zHGHf06WvNUc%!!!!6<B*4)mUfi$6Xe0G+9xX1PH?iF|x$a``p54NGLD`UxPs5ok<D
zM1>;I@FJr79=`7JB9@X6*1Z-Xyx+%!VQpG?_$8CepUFf1AsMMhZZVJeP3#e~$Cewb
z7JaMPn=y}?OjL5>nUeXO*OS<Ow%#=Y3|aG2E3OV-c5O}MOVm%d)ln?R3KJ+x!?26I
zZguwku2eTh%EE?>ii=Zug{=dAeZM?@IvtluH}BOK{ZXPDIis9}Y*p391S@LoY^QUN
z^fb-ueYkSy4+QwAszj%^rz88a2c5Hf9il$UKx;3f&GhES&t=VHaCbDJ=;Ri3oyJs$
z<pu7twVt!L=gB5NC-S~r!q<de&q>l9&N(4x$w%oy_S5FTKPkgsoN+qX=04OSi)VHf
zn;<Q_H7MY=tiBs-zANs-IAV1B+2DobE=|h))A_OIYmM)hvO2}R3*K#A+Ju+!3wzAW
z%<}4{8T=waoGvCtMw_`IrCLocqhjLRCxRh!UXX7S<5Qr@lOz*=^+jH}+xRzykP?P$
zZ>e3PwD>k+78c+)dck+T+(kBc9XBNp;m2e_VOXd!h$Dv@8dpJ__N8ws#9MDHQ$3HR
zq<*qUn^A0UwKVM4%b2WPDIeT`2M8~?kBXjr<ry`K4Xj<(^TzvnX9R*gX#AHJ&_P7b
zxP<oIWbgZw!7=x#$NBE#7x<JF`Ci2Tpa(458ol8yde!KdgJuWv>eU;$ozn09;>hk=
z@Ar=?fAk#CR(xLY`QS^>G``BvbU>x>@IX8PAO;!_Yy9}T3%t4PEiV|sJ9uTsl@yrR
zyq_uFlaIZY4(ZdcXI>(pNS$p9tx(aE73k)Y&!R>x?8)Q)j9GiW{cTB!&u*9gbqbdv
z5muZ4Jv}1-{iWgNXo2!;$I*6hfNi*1qoT)_49TGrqX-fj>s_mYk7M0#6|(Vs1g6RP
z$jZ%tEO|*@ekG#CDfA-&O0bYd9ql=%B?=fq(K+h#iPfqTH=m`czN&ZJdZOobmY~Wi
zdkasU@+gY$99P@3yxL}h(wX)>7*kaVF&S1^&vAd`_u_5@VWJ=^lNNTyJ=J;I+!j($
zji*kOXiZ>8XoY}7KseHI_9>F2#VBs<kqboeUnKM|#XobkM$okFHDdJi65)v>Q|EOl
znka0%Ksv0yv$c9WIV~d*e$xpaC8vPKGwZb<?%f9y$gOU5^-?2)&1l>5sig!P?9rfx
z^iGl{`CID$)z)_9A5#A=&gL1CQj7_<L|7C;I0I-&<zvtS5=Xj(#Kg3RzcsR2MHRSx
zWHy-;A8$5n_@2v-?%O&b>(}d#tYJWPyqov|+1?@hZ{C4iaedD)V_%h%rgap=ON<3i
zUdSY1DY)Lf&A-~iK~}tzNDBn57W397ghQ=|HwhHjfDnI)1!lsy{GME=`*SlFUhLpq
zW45XNK8t6U8JuO^u)aHN$ZdgOks&^G#eukM+P~QX7^0uto9VglAP9igcZ7c+0F@TD
z7U(WE8|TBiM9lJ^mQ;+)4g_v;McbvBY)&gS?}u_Px5N_pc~p~<e%5TJYMpg!!z=gs
zP9OA2XdC&~Wc1dG3KVe^I~~!NmmhodLM#$E#_H^e5dLrjj$;n4o%@$ZD>=Y?NXNUi
zWrD(ikAddSw1%|Nr~{XF^aO%27_mM`+qp>OcYZYULHY}1h~CKbuu*@@v?I)XgK+m#
zGIrs5kcPzGbz|U*)7htDoUF9k%7bAU<^*g;P0A1<?N0$VgPV<;O$5f2))Ucv;3(r<
z;;$|q7Wd?xdLaHIyW#hgiAAp7xQ!{MS9sujHSsNJAQeK3_r58=LsE~Ip&EHEHDzjE
zwkx)_xM39{MB{HORi5o~Zs|N@m=hYcP!nM|c0E>KBioCBoz_9Di;|uAxN5;!=vGog
znPSN#$L!r9pUqqOk*Y#n)Bc0=8t*%I86p>ufhY|D3sfCnG|~U&3i7ymK$%J%0{N1k
zfPVc8B?!t8FY%&Q8G<tSf@wqN7(=dOCnnXpFs%6LBGrb>(lR-llKk@RxQ|*ua0hAB
z{()wt8CeJxb}W}$$2LZ3hlG#b!Q7^ZU4zlnJZ-yP(RW7ZYeU@SwgRkop~_XYk&XHJ
zaP2hit3f+GXDO4E2+O`?TpNTwi8}2!@g>CK_8Mz9Vd<A28rWsvScs$KU++!6k=veP
z%cre?G#Y!_J*w8w`<%CRfJI<~9p(Fv=p#-`Eb-0f=<$4PsueH_U-&^Mtc7B{(RP)V
zjhWT(N5yvEVEX2HGJ6554H1-6*;t?@x;y5g%MF|6H*y?i&}e9X^mGv&s{v&qEll?C
z^C~cy_XDj|plw77cDd`4Q9HG4IL>m(aGl)??XDL-Fpk8$`+yAo{oVU)X(fIelz6!j
zrvA3I9OR8g>{p*P#FpzOYyfReS~>TtGt7C}8wPD7Lc$veNgvzcN;+09--WXcyazd;
zQ2W%U(~36>wd~y;Dh%3=Lt<ZzdeD)rBM3MH0W#WRRWpY@HCiA=`s9#s0g@Hr0`<t*
z!oULeYsKMohewc+sy8nb)}XLz!)Wzh#?r~j*pJd1C}lYA<WHH1NuHBS(c`5Uqd<s~
zonnhl86pr=yI!_llSW~B7FbO)*d&odH!k}@MFqDXX4vB_;5v%80M4Ngh+C|^1hDzV
zTy7ByzhA0Ik>B64-yhOTS`WX*+s_&V`^odlesfyYgQ;5EXm*iF4n4Z5I!)Y?`6zrI
zv`wbhF)UkMCw$&TTUugXG+-j_mvBk&t_5c;rcW%u&iWa30SE*(2g>|#J|H%z%MzwJ
zmbD^%sn|_nqg@$=^8BdZ=59lL$hLV2ctjArf)llA&iGj62XKwYD4XtE-FIa}zv{Y+
z_f_%;twlG)A@97L8>^vm$>mk7nKXzi&%RO;hCHw0ZVi1+f6!-^`Mh@}y#amd08P&1
ziBPK;nXjxH<NUeom%`ds(dBTHTBPNCHjKdLBi9v|4XhlpXH$t@HinWk&J<rpE*k*`
zfZJf7`~_1{FLq5v;R^<UU*I$g8?L$BV8DDUop0lA&uX+KLH>g?c$RzW)`MT)Xe6d}
zBjysN(wnJKs9t4~!;v8c$Nc8SV+kzPqaWnuyr*1L>jnh~^Ew~;i2WSBcLX^5LN8t}
z5+S+YEr39sx!ZUq=B5Mjpe^)}mRc<lrbx#4!zbyRKUa8CrR3`PH4o@4R;k5*MX=a0
z^3L;6@;qe(?0dvF!4-JsFreuwK@;}bi$So)&D8R)dR(wExHlzKc<n4}MC9n)h4IqU
zmLf}Cj#0p(S$!cQp=`*lAtB8NpPqhV<Yw;vD~94k+-~!tFJku^2B*iiE|-7ue3Ro*
z=>x!{$Gz7~W!Sk3LARq<l*1EZbuC3%aCs_1t5w`TaY~d~h9)69ZMdvY2R^$qVm}al
zQMflrBjm(xSgE^y82fOr((PDf<vj}n9fBiDtR`Uw9qERYr}7%lcBo$RcDx)(wK)#I
zK<UX>6b%Zpub`U!m9K{1ff^Mhx8CESO#^LLD0r30nLP`T&9~{Dl+$`(u_@b8h*CAs
zh^U*Bm_7S3W}^#Ik;H=B&X;iURyt+zP_MQ&tQ<|I&KmBP-itUAi8b>O3AV(kW?n7v
zq>8?ER0_jma$Z^1c_<h1PJiAOaP|2OZ^{@V4mP2r(Mbe2q@~-o2-%ph00#%9wIp_}
zuIr-m9PlOVNaXP-=%{yym=~G502Z#(hJvqi#;YuvpUu>2iZHRl16h5x$r68Q(d}{=
zv*;%?AP|}#;luO>N#UfCof$8xgs=`QX{<2GVkyfmrqBlFizHLW@1B^kti2KYIFstC
z+4fvjRe0>!SmET(J^xA{YV`Ivj;rgg<<q4UP@P%KphS<~zq3=jnTj4h{;0`SsOAau
z*A|2Oe49Bn=>xQR6S)p20gZwE{jW6+3-7IoFi{$`<LDBRjqnX{pnx#ewYqk`ld{oo
z>A0^-V~Vk~HU}m5^^mS`0Kwu>D=D*f<Bou?h<Ez<8HfGyx)U(ssc%g#8=bZMaAUl!
z>roM?+DjHBtkmWQ1n<!8mN!M$7I>I*XC#m8P=i}Tyx5T2->jH#oGC%WuE7gG)aj^I
z(srO5qh+ljj1zG?KAIIhNXP<BdE;oC+1h83Jo{Y8;<b`KtCDu#o(_f3&UC{loEhVF
zW4L*&Evu5XYtDBYzv8K;?H+bN4TOp6SMaaf3$Y9!<sc41JQnI5#2}Jx^GZhT*eiE0
zb-cy|-Sk{?1e-?7<16=-5lzej!wUs0;a$B<<@ndQh*$><3al*zt_zyz-33p2!v$C4
zsqL)1tZYGAyHJXy*3QLp4#<=~1X;<J8Y>-wN!s=y&u6V$YVB4QV<d=&D8Y9lz@2#R
z7U$qP3i%kSKjS#7W7#Ny2L<}_H1Xv=SvM|Mdt6zE;SC#mXPk0605n87u05W>d@$F(
zXnqmDvEOhIIXj9cOTKUR$W9N6lA2lqw&7e3^z%U^a?+*XU@G6s^hKvZ_odJN&mS8!
z8LbA^Z&JD@&=LEsB!o!>2+|gn3eA^Lv%4<8Lg1NUf4~fg_L&kxn?7yGoCW$mKcMp~
zgL4>xyw2BugePe*!?-nZKjF{P*~T*!i@XX&LEPpGKIe4a)j%65_)P9$9KAv+21cUm
z9^hU!+#CAc3Nls*c|Srdn<r<>$$BUu!ov~Qzk_`);;wg1V$FO^2&|nl2|@e$7tt?>
z^LL;DOFyE-?;pUasrd~=EpCpNBvZ^z*Vv$1ju$H)3POS*%Cs9@B%S-dOjP)-Utd%@
zcnJCxRP25u&AFkb@&CKRh!@xaaIxJHscfy*V=y)vzbLs>VbCNUa3?8}N<3@HiXIwl
zFgE&(`WYHPEea%{XY}P;DyI#xw%c05^SYN&N=agJ5m};u{7q;o<zM;pXxC4X3cL@%
zu`!5yjv2P1la=&J{5ypSbVa8C5}00`ELTLj7CHrzu=z)_z7u3%3h;lpm0)<L^vceF
z1XuxSV{Jz!fjS{UDAKaNqa+1i)ONKI_*Z;B)OgYkLvabhF0m+=*Qr}7e46Z!=E<gL
zU0g5Hq9-o2WtCD)+PB1+=LCev7O%cxA?hi&#Q(+@uJ8}H60+!^%Z{_wwYYX!^Y7&t
zrb6s*Y~Rq@7T{pv&sw7Gp$-mMjThtPGjjsu?Kd7_!~%+TC=$=Z55Br7f{$Fw1J#-4
z&j2+8Ymo}JqDJkVT9ZXU?d7gXG#!0>@epqFjZio4>xZloqs>xgxVsH&_;WiCz;wO^
z;a&LSu07GU9m*K{;*Y8xB0!ob57I8zs2re9I2VfN&@$yrS8=FI-gBuwML<lzsQ|})
z0{+`W5en9>rQg5u?No|s#_pRvtJpUea{l2$>XZ6Hawjh8a7#@-Lf#K2>{;YaLE2EC
zT;}9EJ5b0Vy3sv$WZjf_p8{u0T#1lx`y-_e&ewc&UJ84tHskFC99eTyrfJCZYQ@1Q
z-|iyC`6Hp1tU8BKOo{Byhv@%$-d>WRd-xZs`*@MI{LtEpi>s6w8rX*0!ZEu!ALZC1
z%NhHCe2BtdL>3WruHJM`K!fj;wO`Wz$tBXSUZiYdl}GmpIP_X4`2fK6z07?kkuH*d
z-<d6A9S#`r;1e~Ivle)0heHk+l%8O+Q@?;2Sb<R;`w?b*jo$jkv3(+xYd-4kTW|AZ
zCC!?nP0_1FlqjKginc~)p9{mJQ0n=j4mm<gtZA`^MJnt@2uIl2+P7B;4)*4D&rD~+
zo;lPAtOe^S>>b3(^ABy(XDsIEd0z#n5SP!2wV}>DogZ4v2?J-(E@F4(8_(S(e(j?1
zl34qJS7PcD2~Avm1cy4S`0@7+<#6cJ@ypK65%+CbUz;5NG~f4Xd4jZ4zGXf}<;fZ=
z+Nx0tF~oSu`waI5xs87qJ>;&v!BTRtOlJ9fSiA|Gp0r}>(H~sY;Kl^rp>cjSXZE7;
zK7FkIwcLOl@%zw7=zPT`pd9AyU}dQjID@^9BGlcG3-6Yu7E82h#y5*WpRQ9|r4e&8
zJ5Fx+ur*nr#^&O@dhYiLuxmEu-LtPkOj4=@<5HmYF|zq;S3u_BCY+JbPYx4Yf6=Tf
zEV6EX?gVTo#oM{-a3+!_o4?LHY~2yo<GjGWPYUNMn5eoQi+q45sW8B9F$XbIr0{<0
zrn^5{t}j4Z<a@r@5oXqkcZAWs4wmp4v%og~7f{McXWa7UsTzjyYhIqsTfI0URb2ZP
z<us!@^<d(0_|lXCfZL!M+9+<b)YAT&hRFE3GF0SK&hY#>e|)>pX0yg|4~!dw*_3?p
zCXCxz6hU@Msp;l+-S9?gC8|*it=})7<?nbCO!Qzv6JNS-Ia^y!DW`FSsl)e$fXL+Z
z2r>cTwvj4!8GWj>dz{c-Pm=Y`thM90;@lPjw8=QbL0-7;W;8;e4%5<JkTLTh8J<rD
zUqcR<X*(vs<FA8h-_YW&svsJI5EG&<o=%mC@jD;ub~B@?*-uIXQ`aN$Yc|bH6&JU7
zjJNIwsirN9Pysb_`BqwcS`+gY7J}H4wm^wpe3KZ-y-gEAjEMs)1tbsO0$`a`rDSEt
z4$(FuWxzD1s-sORlk`-N;mX*L7rSydPu9q<@p>XHJH^_nl3VWd3)Kio4XOcQHj?lI
z@Fq8pt?ich%<V*X<$eQXGg9uH7=3r?B-LrRj9Gb#XPslYaxquIZOF2<<x+zXbBgz{
zO{gUNv$jI~#5Tdy=#9TY29dVEKnA7uehuAvQjP@7KYkT+>I~&}WcD6LT>}7NI8xZy
zG$gx<ZhNGHGr7o^@>0M$cvduov)bI*?*NuIr{kIJuiK93c+~}&NeegFx&q1?53T4>
z1hT7N*wN!TJlRg0NKSHG<|N)#uP`5W5fDKys%p6gzwrAllJHE8cn7hoY@5T(k%=<a
zIjmCcjrhFEvNZ?h?AI@AXIoE5Io?R{2v~bX+^5c6kj{{-4@sgfi)TYL8J$riWgbRL
zpu_&iW?j^s?=C2bDCCaKS)ViyQTU<f6Y|EI@YHFSaoN5y9l#&t<EaPVwvJXCWa)=d
zIGL}uoIds=J3FoZHgRUk+gmm#ynbeeQ^(H8?FCV)jDLj9^yBqG&Ai(pdKUh)gP(1#
zgn)~Ul{~M)75OdN^}*?gd=MCk^DD8JpYzzyo<W4t)M=iq!8tC!0Itw)%<pOjyWH!>
z_~;Rti<GF4iY{H9^PVO^`#BduZxlT#U2UtC?QP5ZZWwv`t6{20vR|Risk-9lcIMHB
zouv)?+^$<WNr$QvOMCVjuS$Skr2H28F!}<Fnnb<9nP_Y;I6CHIX|M_*NJR5hfJ5-p
zE_>f9DUUtTNd3|b&Y(ZC)|?e5``ErIK9q$wwsZ-xOytXooPxi{7Cf+VTm!?nDZLv^
zY4SJPXDKSoh^@<sas?VGvsov3WgJbeQdU0Bt3S^vqXs-8?-Or<ScbI;N4~KH(}q!%
zr;!w%Svr&x$LJ=m9H2-9ivWd)ZVIJNCGiIB4RLRXK}v8886+B3Fx=ZuBQj5t-sAdw
z4-rCq<SNIl%z46O0QJmD0hw#YCv~QwgJ&#lFTtQQ<-Fu26pD$4;&QC$15w@DCt*S;
zSukUnCY|V`OO5pHs^YL1g=^PM4xS6j+gP&NRYPG8!jv&c_x3t~#Vok9nZ;iPzK-v^
z(8=FhzuIj%{7p(s^~<9X#pVcEbDk(}#ri*<e!KmLe-NU9uJa6psY_{D$B+nQwIM*1
zYK^`aQN;2!b;I=0Jcd0V{Ygp*X=5(vp)4M~!^+VH4wkQ)C&70C+d#ICP}DnfzAZWA
zZZn!CPH0mAkC3C=nDW*KL>#9)c5Hw_2QCx_A)Q>2DIQMGqPg`(?}3G=Jp*ZK+6^Eh
zL`!2iKKm*_>8}caz$3w8qR`h~xH)z_yum0vs(!q$#AZKp!EMjk^H6e{uXFGw-uki!
zr{J&FHc~t6LQC#YCYnJn0`|42SglMCIe-&o>}#3FT{EeJ|M@VCZ0n;ZfOXkk9{9q8
z)Bc>QKR{^P(cL@cnuy1lVHQYK-XD0^fmXgY$KO{x0Q0J9%xP=*DTI=us|!St819N7
zfTa)*XVG3RCbK8|t2?{m%So5<g3sgpUCWN#9=A3{Bo4B^aSHI`HMp{l)lpI(emX=J
za6kKaY{3(&ZuK!U6`E;Cmo~QcfNxTzkv6wVaSkuJiF*wk_%(&!H4uOT3VUZ;APiol
zarFuF>sCFb&zwmplRqX3O9)EzE83D{#oxgM_Flv%u`Rk^iz5nrh^Np}(r}l<FuEC`
zp9QUsMUW-oe|53}H9)Fax4LQsyn;2_(Jl-_Qn>}dO?id-F1k<h^1;Y21=-%BsO@aA
z7F&Ck2udcA;c@wgxpDjW%>(^~pg=w*4`N<d79(S0F^QXFPy#$)gTXJUS|S%%LH&bJ
zlF~2jKEjh*guoRkR1dhR8`&?U2{}u~w1yf#$@nsp0eU)QjixmkCk*3ZVNQdW?XvPx
z1qOVl`i!J&zuG90ag^IVj`e{?@kdZ-z+l1mm7M#x+PI<yUo`u37dmn$4$k#IR^YV0
zntIYqSRToX5u40`(`*tWz*Mlu?dH6bf=@s%ph~83MAfz8{gk$tT>jXLHi`}^cfQul
zhx=7K0Vh!GD7W<Q0MH@ymIdi-0WfZQHyAG5qNkZCh<d(H#(ab?q{glZ8fIxcV>S8^
z3}C=P&gGauL@1{_kb6xZL5lVXb6Qf))1mw9!d~3y%;#qkHl2R=Z8Pjcf)dH7oJAl)
zxZym+PkNb>t^@S~!$1Q~+||)4c1;~qdIEl(Mln!8ha5$P6z!nPo0gRzD3vr)CsYJi
z@(n9sSq)7i9@Pj~hW}LhdjFr$#97)44z(95S1)%z6(E(cewI7}BIz*>h-4XJ=l(Yp
zup@wPE!lA_(<S#8|M2ZT3nW-TNb;1(Iw~r9Xb4^KRhZ@9wfjKnOwz|p#Pc`Ga2*v$
zpri<Kj^$oYj4wuGfPexBKmMJYfJnFh<L`?>fUT{MT1wx9zd|4ldgy;gARyv#h<E+p
zut)KOSuGhBgxk+E1iS^p4+w8jjP~y|14IGd^}j#vzfu?fd*}YIZ_BuI3`F-F8EI)H
zn6FXK-a`N8JV?pkdr3=6BgAl*cUxFkyhhZ{lOq}d2^Mex!d#0Qh+_8ktcy`!_@G;t
zSQh1-8Jbzq$H7S^!a8Ya79<mi`z`!?GDQ>`ph76#eq#K$D2unS5Go|S-Ta??3FIoG
z&};!$U0QlhQ}jSs4Rl(8AcMZJTJY}r%n!IT7mW`xuftW^ED(POmd?|n%DJyUC}xR|
zU0-$Jei$89S&@9!1_DFnDPz|r@F^-VEp8!a5JYSS7K)|XJ2>FqFo79%d~6T-QMDkJ
z@-5)Ms}yW+iUL~45m!tRpU;V^9a|bjpU~1j!7C#?P;w4rIEc+#p|14blCyy@6Gm4~
zY7_qx-U*mc{sZJn)cYUOL<Fn<z@#95?0*DT{qG<9`vv_MMfbnU`1kkm4}JH4dxQLg
z97{jmyUC@CZbE<t8HFexIjl6TG(dA#AB0s@c{l52XA1vjIhuU-Zj28;WLG*$T04Ke
ze7W$zg9`>yw*#N|FQ%)l5P{opSwu3Zc;ToNccOxD3W|}0weH{BI&=6toCBLb*x2py
zfYsb5gSq|Ze5e9Siz(}t&MP1uE(7iMY(np=<wWW90;m+zfkNAMQa^~+8-zQ4H^{x$
zXzVHD_Bq;*qQ3pw+{zJT#jMr6J>TR(F$lPGv*}7&LIA>s>NUBaJqh5{UE@?xFt~N)
zt!kbwo*3QT@q3s?*1!D?q0CG?5F^ntq6C5C>0^yay{cPY&k&hgUWT&d;)3IwDAsP!
zDipc|_F0a0O}pRvAYH$DJKd)+bLlyS()n+iVG6(BqZcOnz+~)cSz)~dm1(ry)mXH*
zF7x|Q6)kpM%20Jb?ZV>sh0KY<YZ3<t@0uUwmvS(AE4jk}OI6qc5DHK!rvrL!FLJl$
zf&nDS`MspeaHiS{Xr-82fS`_CJVqDa3HSa((^hVBI*3V+j*ZQXJf9_rzcsj_0|3g-
zT3faH!9q>lz(laV!lk}92o%sCD^#s&26I>uZ}0DlYFJA&JW#MG{y;kUbo|^mcHuAV
z1k=@Zso*c{1gF^24QA@zg01|4s?w4=zWG)!U;7qY5^HVSEI?qbJ2Be(h_ri?lZG`t
zEBJH>L;kX-qtn|6%zM9rw2E$d;Fl|xf*Ygt>3J}<s`tYSm@B|uBkR;Vez4^#4tofG
zM)4MOUVpFMaC!%v(O4l?tA_S(l!EX_PW@jb#rxB+YtYYEo{HXD9s*LLQwo|d=fSvB
zlVVKjcR_hV?gwE#r_OKyTtzw17+(E9c@^!FXdPeNw+kTg7D5;MbF`~Qp9NRdSO3!F
z#GY7*2HR`<EMtfdl?K5EkW|+M5xc`B8FDiz15X$I<JF$`J}7sNLOnK9KBj_G4)X50
zDun4~TsNxn%@`^-VNCHVY6!*^rh->R%$g{(q9aQ@11hGG^#rUiDDXz@yk@t~yOS9#
z--pl!dwLYxvX7dNd+Avm^KYQ=KcjwvR#>|r(T`SFQE34ow(!mO*F|KaAqLSn5qOVP
zWdvB;KcL4v7kA!~Pkk)KFTc_mC&Yw>EqEldlD=I*g$aknen$QJ!zn)$roOcYC?5Mg
z-_19^G7z-^AsATIX1E|jr4vin`QvYCuyvku_{gk%ALHSy!yq0tY5U@9<m1=p=5NVq
zuFm>M<#$G(;{W_8lMjDGDb!Pz()NFl6Im2;z7a4Vx<=(+_jn+n^ae(vgzsi+WBhRl
zRw*ZEDe5;;J;<|gO$aeMNt>OxR7x&qkr0%$ML+PrB1o4bLgx#h(M3A`Rn4ftZjbAK
zKPyBV82J1dlD#(TFTjl`=Y)z&fC?-fXdxAzo7mf)wpn1_0fPj8;5@44R{$U;C&_ow
zJVmn1NcKO6K-}!oR$iKO;;ZD*2giRQAsj>LREW~vFKU*W&jo*sL6LieA~~d`Jn4x@
zew0lab|*3Q$iR7~`?q72s|gdO@#stGWp~%;%a_)(ljrDcPCurAPxW#ZN!|-!<!m$+
zL)|FEH14}HUCCV%84#b?B|82hFW3bxME3a%hltsZN&fs-l){M;bcgk8<tC@HS`=1^
zsEZQkl^>*0G~<@Ap|Z&yPJLfS7fJISv@MT#9xFu(kR7)>T`cMN2%BBV!5!RVSlnMC
z7dGoaDX3yC9m#Wc+Z6~ld*3ZWCYl|^;3fB4BPBklc+`vXUy+me`1o$cs=@GZB#jCa
z<P`f}+A4kBCUBea@K$Lh2-nQExl1fjyBocV)osl$VFJ>oNxS-gfM*g)`#O7`D_n-q
z4hA-_tHyc`0n9&`ZbAF&^SAEDlS~jGgNeqkaXG^6CQ-dsrRA60i{NK!_@8UHnvP-<
zQb~x>RBqaivFBZLHsvQQ6fysUVL>f`Ff4tq>F3?CY#)-Qjge}RzU_O6677BxV7V0y
zK_4T1&?52B07@`;s<3gxy)xhs7MuT5vzd?@$zGrOa>tj8Bj0uvS;;xs@ca(28~-J~
z!Si474bg5(%7;aocU6r1)|3dhlWht#zIu$QiW;kqmh=hD_unW3EZ$+g)K;NMXnv*F
z+w2^za&%9w#V&{5vp2uF<zIuf=lc|q{W$1KQdTwr*fyDp5r^Qw#8TczZyi7L5PNrI
z`<Vvrs)IMEuh(%+KBaT9QW6A11f;QaQ;iZCc10%eKioPL6c*Cu+J_0m4!`Ik%$hlG
z509(_{k)Htis%GAC4_QITb#1}=Or=xA|(Q0B)BVp)UogZo#~;3g;_*ysmW%|x@`gP
zgkz>xdVM0N&122CW@7uAp5T4`7|0B>>3x7<RT5L<p|CkHbAcpJIE@59v!7r|XN_GO
z-865%u(663IBbCmMUax0Px8K6!;vC1%z=|oK<_0zaLrWtE8ju<%o?AVi8=1L3@=|5
zu9dW(WLQE2NnhE$hTM;%yKli;hXm?_>}DC?sezjMh`E&#KJAlNv(E{XSe=@B7E)M^
zrC;B+zi>jQd;5w18gL|fo$X<<T=^&=$oA4!MCs2wLQ!0UZOj~NfE{SfUdTv&u=ZY@
zSdCMY%T0=}%o}dQ(9-*2&F1N(*TDKaqfwhSuW#$KJ!q7Ls#G||?b!zb=^R8N;)!6>
z+{1yI2k+M+Kfs2++W+}w8~_>_fqF+y6Toa>8W593A%D8l0|RvNA8(I>3#@1v8D$m8
zFrL2-_T>G8S4aWAgu4`)oeOw74beao?&WJ*-Z3xO!9YV*=~EAox%ibZ5(Ci{ZH)wv
z?9~ZV0}ZRMd6?AukW)-m$B*i7W-86&grBe3Ky-?+Tz<UkgP`%Ip0tp!m}RhzMKAPQ
z=v_V@&$-onwN4;KSW{r`!>Cz#xbCN@ZCv-lU7RXnQ2m9&$C;aVB_Qj(*&BHinyeXQ
ze`L_IoqsZRNDHe=rQ`RM=#vSLRfe55=odu0>~RCxJbAYkmq12uVIE8-%`d7F_DuZD
zv06i!ZV4UR5%@MWz&8(#YV}k~S$lUzt4G7QZB-~cCFq}_5+I!r$UhS&sw3Q2)wQD(
zxfegh$ityUNg~)}6awEdf&iV@?7!clT`IDv&#}9RwyRA6pOLo^kx10!A}_H2<fM14
zukt$1?4we}O2h4P8d^DnSd%DqjVUaxwj!~|i}u&-D31!Ru;l9noOPiQ7LLMRL;-#<
zM2O#fm@*(1Q$kUSUcI!qM2@4|GQ;*1orxj(+YC{eA{-#3UM*WDanS%ovCE-8HG|d#
z_?_;V@Etn3e{luNjn}@_x_Quju-vJ7Oby0Ip2&kJMBLF=Q&dQWB0vjojqH{tmr|H%
z?lGAQ_Sf=O7i5jyukCO?;zWKI%MvJOhBx-Ok$BF%<K0&VC=$ksOAMvsQ^TRD(5<H1
z2sixHR~xCc0rX#(`@Z!CFI>J^lOP&|aU;ZCtrgc!sZ3XhUFM<_(<uieQd<nQjUZ}-
zFiq{2v;^-SI;?SYQoo0gvdQeP&}6H7rh$Blq_!ryE`aD1M9W@zEJr>%rtSz~#7|w$
z#%2|0@c+HYGbLmPBY2FTu21AQ`kwULGxtRdjf8V-a%n%L0Q=Ag6h>f;rB|_D1KjJ<
zifA2-LN;nFyVp9`^0uGZpQ?LT`~Mta0>152{Yfd2$j9%I?X~^rBDqu?2O#Pf6%`fA
z3B}bOn7St@5~+<pnqDG2mWQww#?d<ob&GoHH;P6R-}Ogh`j$$)Bs<KuhI9TO_TDNg
z%kW(vq@_bZx?3cqJEbH==|)Ovq~oJOy1TnUK)SnIx+JB$^Z)SMd(XkFnUgu3wU1EN
z0-pE%p1QC53gp|^wm>(b!g6p=BuuZ5Z8yhzkv=uda++s6zo;`TF-(nm8g8$WFqH@G
zzFtuOx(MqwEBXq$dmclukbdi1CYSZKXVYV|uh~4=5iD32#d9;+LGQE3lr`UXFL2?8
z<mI`-$JT1Ya!h1%e8U$192t9qTc*<C^IPIuh&;-JS<eNm;e#l3z6|G+nWd#P$Bp+J
zBC)ql#5OiRovc<<b?n^O-JF{0_+<y=?dH?C*{w}hMpLe>u1DNrOQ{FXmp@5KpVN}I
zxX0O%rhj*yQe=qgL~iIncG1xxso_l*aEW#eH}FrT4EheWufK5;&Nh7$%fTT=L1*3J
z!7Uv9Z>?Af$r5Hd4ttYQKnbuN@iBk8V3z0QN#K6Gz-D@<{>y9K?G&o+ZLTO6+edPL
z+%s}A%)swJl$$@ZQe`}3f2^_jxuN6;7BLHU1_jMfipl1Lc}vh@<$=G-la7i1L{_wL
zpjIfR)10#c71y%;s(BaBMnnh~jL{pJH{cWQ`o;!L%2xw}tTC#_1T@A?xhi^u#{2ui
zFfaOo94QRO-n@KoEM#6@G)0C@vwP`>pU)Z5spNQTuA?-JjIx~txi4qEicW1^-RF8%
zvtXOkbqN*lg#Q_<9gU()TO&UG&uCsFCN&k6-Pv8wL<ym|rad93IFkN_PoCMrBkqC2
zDV|crPE%DIyELs`%BiLdn;#^<t;vt9>N`)3t*WaV8~87g`=>V3Q0d|}bJL;LG4-d}
zV=!)(Zj}(znq%XAuz}rbnP3W4>JO~N`JD3hbbwg~KeNZ-vPxu0nOA9iZgqKwSI4?K
z(SL1Fd5r9h;l{tl32ghn<VX-Sww312#@S314V<*JwAZ@0_iB4&@rl`$FuyTV>_`Xb
zy{QW#6;tme$gQUP3D6hKH9Ff-hT!dbzwWfu!S`i7!q~cego&sxt8<Q|(<hUhVENnb
zQ~1}OkKtUpDm=*AsA-uLE?~U)rvQ}`8d?*iOzO%i=SneKpwYE_Z?*{4WM$vlXcgV{
z=!T8%kI|40Fs@Htd1_$)@&~ebo}aPW!HV34+#rhVA}$U~^EV|q89!o&l)f}_D4O;L
z=ApU&&QxKE?=%?S%uj5TZQ|6CT>@@Me}ZotL8e8cPkt&sg@R`>{6|jGjn@`j10vr3
zu0xIbJDjm(IodTKj1+a?4URIduC8Vv#z)X0*Bf%C&Vh?g6&uymvtxCu+*;{i!EIfw
z$u9eJf#UG6yuADeg*?jICyaw=55N0qvpj2kHb+pSt~8U!h>0-DxkC7?Ah9Q6GqPmQ
z?}7hhDMqgO>quDf1j1A#_oEL%j8ocfVWd<f66MrqBw>`#tkzy~?8k4Dm#(j_VD9h(
zqoSf@+;3!v`S|!mtU_<zz#<?Z(C)_&h5kXhyu5s!xRM0o!bF;`2dqxW34hnr$Zr>%
zg;j6She9;v(cVLbRDZHS0)6(OMfA?Sj?QR5ebpy#mp5dqxC~<zfSbXK|7>O&XB}9d
zZHRBpkDfE_CX!e$KaB+UQvm`a=Vngvb~j2yg_9v+0}ZQT<-xpqUcV<8cYJD!%5k?r
zQc6l(R1{iY{fD%SOjK+XDrvlLZ?9xwSs6Skx@a6XH#Z0oBBHM+h03{6v*GJVZv7I&
zAW`tne|}LIrs~O=E@y|1TQRT;dIw)bIn=sSaQ^^Y>qLyuq8NcY3{1ihO_3BEXeip8
z{1rtmf7NdGq{3csiY@H*_4?J9sd;{k_F(}O_*tTMx5qHgXC@x>gY0|<JsqlHXQ3FM
zSFkoZt#DVJ+9uQWK4GjShIG|(DCE|63?%Gpvw(N}dE`X)9S$z;C!2Jrng{+oT)S+J
z#1WG*T87uyhy7V4B`DF+(H}4!z~aY|!GT`lF86*M5y0$Ci*}5{7!C9z<EMbhfR5<a
zqHxk{g7LLZEGkO6B7wQSB*nkqV^ztA6WQVVC(Zt$AfMq|esTE%ny88?TiAFmDr+|F
z@7h{r&i)%#0k(JA^_Z`*u?0Qr@9*w1+uF!jPw-Sc0X*ZSH%LxL7x;DxML$+#Qkmj<
z9N`%*$x%(b)TKo?*KnOP7&6>JZUgWEo9$tO215;JJG&|=tg#VeM#(BPar#8$hZ6f6
zaT<S{e-U~#BmL>z6&u8Zp&GAwB2bSn=?Vn;wwsaj1<6X1&*uI)&^v{*F7}%I^WN<u
zG)a0RvU2g6Z~K?nk;MC!)n8HN<FDRXEc_lQQS(YpCIo?_3GMFeVE0L&?D*%uxWW3x
z0$S~wrZ}o-kNxL}tG~}a{NJ+3rbcG3$1?ZteC%wq?cKZR$fUg7(AlIA)NC>xn5c8?
zj%7DkZMCM3<sb0lEQBf<P?The7(d$hp`*@^OykYrQ$NByKHQ$CS9}I}hp?gfC1%0E
zuTn`<bI)Z13lFcL(6TTt?6tHVvdf19xu4RBuZ)%@)>RS{y_#mUZit`^jOL#0bF%dn
z9%nJnhd(}Fa=Zht%<i*Nc4S|7H&n8msEiC!sZP_I>_~{f)R*;D{q4i2eLmFICpN;G
zkHYn1+ScClP3UmGV&{BIvt*+ltj7{I&&!NSvfZKT(q_V8Z0i(A0!je~u)h1MT{<ao
z8k|=&?1{yf_j_Y$kYSBM*n%9f@P{r1X@o(@6j85L9!V^mZacSG^!rSofbF4juQWf)
ze%R=rI?xzn=+<r--zp^08P61&E&GojZux&4wK{sw3>U!tQ!MlGx-g`9P3!V|Ez)Lx
zzAx!U8i&6nhP$Hb=<;Y4PdGUP5u|;{CHyQs#ZuSQOcp`^8GQpQNu6UsPiNVGDw5ce
zD}rx+H~Hh|92wfrKKS*={>A|GWu|hyDy!vZxmD?)rycRZ$*C!tvkz9wSUgBV2^kqI
znL$}zo|LdzGtEn@bJFu)WT^@B>PMR^u<;=57KM^0$8+3zth=nu<|T%F&PIso;%E5<
z*DWh-^N4p06qEM9DW375pq`)YwLT^OPaMO6&a2ev?(>YM4MHcWSMU-N7W(%Tl=5&N
zFhxb6;cqpRzC7KDL*sU@NZ^RV<DgJ-eF;XSd;^EYNSRIfQyp`6*3E<EIW3S^x2h;*
zlcsM3BIGm>pL76WujaRAZ$gPyfPxy>Q4Ace+4P2e^)_7(G)#7ZVBlAG8vrVhH7)uQ
z-+z5K(WqetDzji$R6>q^`K$8tx9Pkpg9166*9kvPU6kv8wO{^AdM~gM%8qjbjH*UL
zyFWp;D%xx;)i$EoceELPJJsd=`(mNZV(h@L6JvU-mZ7yyq1i9(^qxi40CYwsfw1l^
z4K+3WN}k4J_Bmh|F$scb0A(2&nbKksmvFs1B2d-^>!nioS6wYyS=X0-d7CVxF(_~j
zC*Lqc=9Ffco(%_CD^Odl9uv}%Mu^~9NBjmlBgz^YNu=hpQ#o6D`zx2P_zculXpFCS
zhX=Gq|E_@OKmN6YpEDS@D0sLZ=Uez})QxI|72e)EH*8&fu~4Tob`%Kuhs`O3x57Qa
zCLHAh`D<E`4gwwO@`c(?dx;$9oRw8&O>?F$NC^W8O3Cro*VpTR@4%5%?ND}JpZ^;R
z1WFA09qYD7M`u7W!TIJI1+-RTs;cpnDFmDCd@2&jvr$@B4e<Gj8M=@7CXrpyG_u7T
zqKyn59_ToV8N(_Jtp-$u!vC10dG6Gi$M*fM4i{g}hGMcVR6G5&4-QL*Da9GwVk`Yw
zcNW;#SsnYbnWMIJdkLSmVk?|60O+9-hphB(b1Ji=B&5N<?)AeTJqz(qpRh6?+JL8e
zvb>@2uPmVLve}H15(|05l$n_c1%h1&ZW-Ub`;*IiHTQLGT20TtT&L;#c2k;q%!S(W
z)!CM8=1D8k&6I2!{H19NOi<eB_T;hxA|ZsHpZ@Cv1+c6IzB8n^_dr8M&d<-!r)B$`
z@%->Ao>>PO(15<=x?!Dac^>{QT0d7&VtsNAr#{>cigd-z7{$Gy?_jxQ)H*}Q5wxdU
zE%t*e<hSGa({e-$#T9QBV1xg0HTe}yb-wo()nbhg3UWvMm~5m;v{E{4cm5gE0S9xH
zu;IYbvz&*$pZ7kBansi4*VCTHUTJG`o6x4cAYZ}m{ncU2ANHk}fS|JZ_nd+=NhbZ5
zt8%;Rb<_{=73pZb^UH%H!u!~+so47jXZ@ug{-F91Mr6~Gz6K;l&MqE{D5ovOnbw=1
z0$*l6x0E0%mT+V#N%MgW2G08k5VRx%QdlcuV`6fKtiW*}y<%aatQ_rEeN>4knnA(V
z1+}HZ)P+CRbi~H4tafHj38B{l?(BK7si~z5ZF6)gs;Z4X5m?xQEpby}W?$11gdrR(
zrX&iOcbyw1%N{YpS1RPso>VS!Pb6AZEC<ZELaZ`xJ@!?Q+Ok!x7U<)clUZ<uTq9!H
zb1|d>*^e{?NB>Tu9O3WNY7?`po@{*2PP3sOjogPc#pCK77o$a;e|cyod6c`771yl!
z9>>&c6sZ+ha^OpIaDH`~T4n0MD<K9uJ%(xSON?g31WOwo06W~;B4!g?p;I<cw=*E+
zccVcOJVR0D#whS%NB|@jAd|#rtd?nUBYE0rp;eoX&Dc8t<K(a}nwsC^!fYBcF{Dk<
z#INHHAj=%a72p4%v)lFEauT!w=tmY%#U^7n7(VCo^!A2&m$JL6vEQk0<!GsFYtY%T
z1xPlob|~JW7@NwNIq!$lAz%wnaA>40e8k;vi>q>TT6e}>@oXx<ybsEV3jlVsoaR>W
ztWCziJ!;s&3pjpO4F#ka^eKdy^t<2>sg_5wC1nvTLw`8puZeowsyl?;%|kPT;4+=j
zcCP-gXuG}a9Ot|*KHR6=Rl#~)1-&jHPk)JE$GhyO&Ep2hB+8PK2nBt^%O4Dnplfdx
z(gl(>{4Jd~v5q0b{c)&wUevQwJDO~s>G^UA%+AX{hvS-)!&hg@wDN#skj+`(*@!l!
zPKR_vP-%gynqZ@PLBcSoyjurB$nbu#&DgotlMM`=^&cB~_xks0T_R&g%b$vG)|!5j
zL1>9$ra2_W1V(G-N8s8{+09W1Hc1ILhn*Dp9!|P&)5v<mge>=4Sd%Gc#kckO91}I!
ziRvSBAI=pq-dvnAgGa0<&)C1%RQTmQS{)wQ-*zY&&OHiouGx{XCQq0m;q3Ktp3Ieb
z`53Qv3k=!(k63`#;;%E1chAKWrHIBkRQm{#@2U#CrHV}}+6`B|>0kbo*W4h8bMh7P
zFZCcxuKH4L#p{YiO~d-4JY8rB;fHoT!*$ddE}rv8i>!iG3e2-U+OHoInyHNS#^btV
zdS&;Mh1Wb^?=c6AKY1ko-ub>~j0guoKoh`0J8OTp_eZmTz+HX~C}2!HqFctRO%?lx
zhCXAWkqHso7!>mTg(2apFNx6!dgbuxLl0g_mVW0m)Z4eQheja^dxBS})M;+5Y|91e
zrFag8?A2p}*tza`2l}qPn*|gNik{85$UTaJ2zeDP-4k58_t@RjWkY%V@|e0`*8g^t
z9{c5l^95`1IF0iDdr3_8<9#vGX4P+p3V+eyfYwF_rFM9yu9(v0tS%#z-JPSi%{ByH
zLwQU{$$GCkn4^v=5C%d&H7tQVXVCzY4k%xp+^g<6(m&YO1D_VkJhviN#h<{Q!lvCH
zgu?fTE*dRdlui70raYiHgA2-vRH<eQubqRp+xU-f_Z11nZYgu*TqO!M@s#GMPcX}@
zNXJM{@y40FuS>{_T4(6fY=vm2Vu|hV!l^C-yQSg)Z_@KNU_P?)mpie#l>DjlonF(&
z$-4*t#oRcL`e6-Uo&4}iR9;o#<3_CHYOJjiA-g|(Elf0Bc`FuMrunkrggN5PmWy2S
zj~*an(Vo7PW`nX3s9(dc<++!Q!o2m1$oDFwKiiZjT)rb7ZFeC-$9(|NZT$nJ%y-=t
z+`90>7qeDcHU`<6<u*UOiVr>KjJITNMTW>?NG{kb?jE@rVh3iX@hvbkE{3s*N9%iK
ziwjTSxRbtRo_>uJz%lKItuBl~wbMXrZi@FjO$vVxY5r1cQR#&7X7eCWw0!z@EpsDD
z^k`yb0fM@d;eLi>TCXqoA^;af;*svHP2F&=);RZt7pO<M^D?R#NJhA2dNz&Iu3yUX
zAv+Q+&r>{LEM?^~@zWd&QDD!x<`dS|erKHK@z#`nJzQB7DKPk9OWO`Q@W<2KzCa-F
z&dtlN;(J)j5h;4amHiRojEQ#l&V0lyw)MWY%0|Qx0zkIJxUQcBoApFtF<$(|W~V)A
ztj&lMzI8J#cw^eN7D@SQRj!xc`y!^!s6Z5Q-Pcx0RIr)|rUm&}?whN5=7-2UJ?ya*
zk1%i_Oh%IP7-k(o{vgV)_)*D%`K^DBx06n*YevZTb(7a`>L2M1Ky~_{Vzpcvc0TFf
zS+7om$_Fuuj1jd`iKp7~p&Z9ub*C_P5K2<BxN@nyAUSr@`US7~#XuPK%*KT`<(4<B
ztN3Pq_Y-HeXNu<(eYmOWwc{0mM($72=9(j>s!EjukSaOBp{-hO-?{+|{SJ9n>eU(&
zDCqC@bHlQh`FR_}yNXWMno@t&d~l=x8;Az1J>Jr}6@EU5q>$zLi8DDlzr%r*ZZ)~7
z<AFq{HfOy)?4s0}MebQxuosjH%Y)~F{$u=<|G$>`14!`pKJ0G4BmTEmyJ!`Xii1Cb
zgy)J*)XO5$0dHDw))#{5CxI2W-{$R8Ir}aBF3xdZCz5adJ4^BrvZC-C2g9CIBY6P>
zc0THK2*MxF+q50(q3hvU47*W&l-zYncr@nqqUTN`uqY`Z?~w07J3G_U6SBs*s7Nds
zLCUnmv0Xx>;hdI{QSo=8&RS1D5RqYWDHKlBGp)A3XtY}C?44(_-d!4o=wRefY87{`
zFwRLA0y+U3++2KfaP7KGwKE~pIgMIw6Yi<FXW5M4#VlDQf2`h3hTTM7BxEc{b>3OG
zJby^f`B|u+#J=3{l+Guf@8P3#ij8$yU&I>+S03@B?J>cHs}rbX+*MCTV+TS1d;hSu
zdRg65tM?^Z$CD8RcBTA~PfBASKfGalylhgMq6}ntKRhiUyDdc5J%)_6U&d*AxoOWR
za1ehBa=SP)22f3#N^!O(Ug?xOT{wXJwqQFeDIXzmqnQ;O-SXJ!YC$?D`XL}2815(e
zH*5cZ2^ITIfP_o)=!8_Kuo-plU(trH&|;|5#Kl0fT;)^QSs$87?)#C7-7~eD@ic)^
z3oXNkJ=_pX`v>$Mf|t$O{&RajxvAt<?=hzi%Q*cz$hY_u8@Jve;Mu3v{7-xMxI)Y3
zpYy$?@%{KVs$j;k`swCaOVi<bwR9F~`?PI{g#{k})6E9&hX2Tw%>GPJ#@%46p{+or
zpoQd$<KqD)oQDXTxopQF_o-dqg|O%Fv`KFHCK$L4UYwA}o!KC1W2G~}pS$94w_6MG
z6^lM9fnSn>4PH{MubBQA-IhGKQ1OeV2>-4)sk}Et0Fw_yFMgk4GEt^D$y~6CR0B6C
z<Fs?K@~Ke#jWIzQuR>w!R)Wpa2In0g*YS>|p5WuwT6oC;AXRox{Pe}<nyL0?j;Zuz
z<V_Ey@w5CI=b-hD%hzuTjMS15uojQCI}b5iP$8(&acvJq_3+8lSL+d03Xr6wEY!a5
zv{M)lbCBzUYx}vkasgYt;UpytZefI+(alQ+@QisgY7R@k)=5p`E1uLpmw(jtnCL19
zuDlX|>i%AwTf9^yb1tQ*8gZpe&n=}f*QNgQFsZ0^TI<JwB5F@}A+ggNs3Qv1X1&<8
zd&f=&!f19&8=Otgyu{%L0H^dkQAfg3i>p@HbpA_PrR(YV1>^5teH{2P1IUUp$hG4@
z3l{Z7kQqLnVbgLviO6q)s_hv%huuWJ%WA8)Bu<B@pDhq)b#VU%8f8{nRJtEkwz|%%
z9~xZJ^&Z<_jQC>alF7e;AotF`#SIZ*=M#N-g7CP5qDvK$>xZpf0dmhJhh~dPAq627
z<-p+tBjHU&E3-A<7<9Rnxm(+Q1#YG!eyz?^(V5vD^HL}Ab8KM(shs^Rv&dH_pXl~&
zv{68X>0es<4!-UE*IuYaMb)Tnhwcao8r~LuF&?@Ub4_hm1nMRS7}Jl)xvLQ2Yt%aI
z2hBeWb+CAj>oAJU)HI^YLsFu9HP62Fn>ft;scA^gY^JxtdBgVIwmJlcm5DQB5(XA1
zpX(pahT_7WxC7`m)`J>mb=}I2d|rZA(MbU*M!nUYxABjyJD<}LD(GSR!#P%!CUjst
zj*xK@zvV6$9)DMB!)oWMS5Qrnb%~_(Z*GJ9>EsQRt!=Q$21BY!3YVk;%)4hB0_iQ|
zV0YeVn-8*PC)>*Ht>kYyFkXBN?QcN$pK|1D>*_c_&Ut_t(;Bd-3aGVQ>;x@X*wfR~
z;O&iUgQYL}pnPt3MD_u;9g!r4rF=a%dch&gjr3LRQ*v!M7??j3PiH4ly!+-TMR|vI
zSJAxIa7hep_slQOOIPkGJnfdN+-hGHk&~^bkU`5VyNW6IRcDRLSYqr{X#dnX{jxJi
zHdEc^pmcis7V&V7G|ugWt4Qh?mTte;{#hUC`uL&S8c?HHL3gZ65*H6o>Ow`Q5OjuZ
zK*3uagL?|7jiR&VI=J7gBtKjxpxSY-`UEGC!K69N1Zf8r6~}MTlb$c$RHKF8CcZye
zB=5V!6?A?T+fpY|T?`_z>jVg2&*`~u<YTT>19^<aE3SJqC6zpxm_LWwh1`~0tR*7s
zt7RW1&^yZhMF4M0fP)qcl><T!)<z*v{kJbEdNn|`>v=xn-gb?ukNwp@g2C!S+HFRM
z>{N7;&weLopWH}#>X|OV7&d*CyJQjhQMTse+sXC4{7N04l>olJj9lq!w)Iz#34{jK
zMH{Es7t>ErfqOU9e%M5JIxd==m^CA{*q7I5ejgdN_J~2)cFEnieCDysSe@nE`-NB2
zqwM+*%J7ar9Uo~d(aS_{0>u8)J;wVee7!d4?rHfz8!L2ae~W+gVKRYxX9u*WrOXz?
z&ZO(gpL1da+pe*NAH0WWs9a`xfu$cz_$OWXPmeUdOC{W$H+K}{Mv+YqV)GVC6JM`I
z=hy3q*?Z>|h?~cG&iH;hB3ezCl$N5}Xo+#?FCiCMh;)|W+Rm_p>GjCT*$+=h+J_4D
zfR%WY!46;8BlFNcHcXCZiAK1=^J8?W@`J`4Kw&%^2HA4nJK0!UA#k^TV>kQl&h%?c
zg*a#x3DTdjWTNA%&8&Z#%yxh*P2}{d`@Oi(Oo!q_Ke<}VVY>rj@g0GLdG!!7!KP$1
zMY>w8Jq3>NDW)_{&s7pNt=MvjwO8ZpQhPrC>%%D&C2-ab4Gq0y;`#THy2;JLw|}2H
zIj1})il`@5&$HDWjevfUmk2>CBcOXXa)(E_ci098>bMG@ytIAYPkZnjONzBfO(%~V
z_nAc!q$;k6XI-5!FSr|Z<A`QNpB$*NV<@kSsGcz<ts28*F~oex&gkkEi$ueZjK4Vv
zAj+gAg+wa;_vkM&!RYr0BnQaV?tp<IO{7MOwrcho((xiJ&6qIuDO$qY!JO*f$4Ocr
z2ivnET}Q$qZMvftj0HZTML$(xqfzx<gs`+C9@N_oF-ely3YZ~=Wr3WUhz5l|jeLik
zOfpzyvsq!K+Yvmah9kyy@s>H=_1YAlqj%2IvV-GTaz~mA`A#A2cn@=UOc(H`ZnIrd
zWFBy<koz(kr`R2?oYX9Xh>YJX4V$l}6%`{!!v|WDO7|P=T~v5$@$m;&)kwqr%n1mY
zpWt4H^YxbW@vZL%J`CWNtDiG`DSb*T+gCT14aC;02`8TO4|MFIarv<mt1VSsk^wHc
z<*xjq`<P#**+bWdQyaU)Rz)8|j8DWfnFxZzMSq42(ETRAi8L4Wr?XOp+|zt{I%D^%
zzJg|UG+Ye)5PIa7qbb@sapQz~&Lm8`zF*SQwf~!ZkA*qCYifGVfH7FOR#RwwDg?4t
zk56}30sN$O#q)+%F&<UE1T-M|IUacxm!(iNJ7Jvj8b7TMALCTi+we0--R9Ktz`RZ-
zBhV0tZ$H}5B1S?A<cye4z?bk)!CF22df74cBK1gGY;s}&eW-%F;IG|aB9+~AU7{v#
z2V#+;>HWRHn`M+<+j(1cKNkvH9H<$NAo?^X248r$!Ckv^*9Xg&T9Xk*kR<0>O<5jt
zDxWOs9+MwMHX2022TQmqiUe5?Yz%TtRilXVD7v!y3}x_Wj5dpkinwf<!5ZPi3nRM9
z)!z*N6o2BdDDIhZbhDe=7ETc>?U`mEeG{}Hb;lT<CMR=7M@8XC^U!T0{`S-3bWN!Q
zK`{5)cn-C<<CVeyIbL7Hg&HxCdR(YUiXD8s2?7d{+_v?(fmmt~En%qEsffy*`+8cB
zt-t3zcOJDjmw*2+D1yy^6MU0kN5lX*a}C<vifTNC|Chf<<Eu^l<MYSYW}kneqC|>X
zAx0SFn+59k|Bg3~e&Tam^UL6Y{O_Op-<Jud=KsH^O4xlu;sm3(wstC;C~nE;g7CNg
z+28*o2gxADobvO7<Nc_H4JVd``J`XUj7vs_aU<11><#E|U1GUy0Vf~WWIRXSv%Srl
z(aWidl|saX@e{EWC0{5Hva0?iN;=|qgaj?-g-H&F_<wzA$A@O?4?E4${=a|!Gh9FU
zHIA;K!pm=gq7eC?FZfh{8%1>iYWnbD&>sn#wjJ={N}}okCE5T$Pt&%I?<F$<oA8(I
zM(>{3i+%*qbQ<^+xbF~i^YiyitB5D{#GNUVa@)4?3F4IYkAWqg0YJwksbjBMV1j~z
z0@S~#suFI0rhRdVLI5_0_MA8`{Ah`gH!6L*z4*|bWB6n_+U$?h?ET*oa05csyj2<a
z#p4|{u=*k#-UKkV^zhD-ew@@ZJx&IDmhhyWEzr76>JcDSGOp19iCWMe3&G}kVv2IK
zNQ#(|x)*}IxHBrd1V5a3e$Bg>i`jV;x<}mqhy5=5P8OkoXFETZdORPK;7um<GA3!4
z9Jt*@eq$P<ess9%l=192r2;>zkWc*>7lqmM8e&*a5m)#WEy^6;z(9lao8dzMyqO5M
zUmxq|-{P!BnnU38N^7wJLwu0->;c*~8Bb&C?sC5jd@E_1;x?nmkgrxBWne6kqW}|e
z9Bu9WFwP++YSK)V9C$pI)c<|F|NZp-`!N3hJ~?|IWrug-D43ZQk}`uyob%sCWh2AA
z#0&txNLPmikVB?F5KZ%6>-r@o80ZxC#6W-u?2OEpzOSxb_U}>~$U1p7Fh@tA8R+iY
zYkonk6LR_8(2mUMP3@U$fBY_2QUFoczpN`p7Q4R|b6!n{^>n-T>YiLrVS7j>ML?1Q
zcP6PC7}j0SKZlo;q#TbUBP0H4{7#*qwO9PCtgM^~)_!msu-0{f4zx`7p43~cxlSPe
zbw`9Q%9C>Gk#vuK5$HTJ>Qn#wH>?NQlz~RVhVtVftZ#nX)5qAyf?2qS;NW0=V0ZG%
zbS%Ap6=+c?!Uhb%|4W`V>#WcQ1_rL6ZA2)F+q~KP^;~ypoIwyFv4}y~Z=EK`lZW#$
zEyj21Z-2s&fHLnb_cN-B7MH{mwTV)(_%czixbi=%z0JhFN8<6k2Z2HW<b1Kbi+Z&5
z8`Oiq6p^nIA`B<W@m;%16Y$_fH?k6MV3$rbsDJz|Zc~YjtO~%&tc#k2O!r;N$ia`e
zE_>3QpZY}3yeLrHD56Al)5uEdSMRh`RS_~gpLkz#MoymYkGnvb{f@H*)b*?;!}$0>
zVNFd#qmg}@9hq^*IGG;(FT*>SMR<x@{%9FPxuRZ`eJrFqbC)HY$oQDaud3&q3ZXuZ
zi_Sq9Lm2_%x(wq&)-&&nt-sC65BH}<oXnm(`kxkqR3<_?|Hsa;($T!UykGXa%84MS
zqg1^T=Iz_Js4~j)7+oNF@YRd>55V|BzrRiO{%R&l%DyAw@qGXh8$Lh}KZ_=Z>hc#w
zv)UxgfS@1=g2FK>rYb?O>P$8|#A&J8OEpE_a4a%qr;B%DF2dq+#6X|Kj*=*xut7^p
z?C~ylSpX~AHT7R&0trGkiBFoTIs;=Ij`%<bdzSqZBUD_Rs&w&CcV3?O!rUU2ZzCD|
zHe|;d_g?dt-PXsK?DjPfaR8qR{TdClb^QQi0H3+-!R}{71fDUK+`ojU;4k$6GmP{+
zwDGc0j}b7Q!;KU=6&ev>c@3M>2JfOYeAiVaaSkm;k(EzN>_nNXSpVwv*a7timXDi(
zPlh@)CafkI9>r!g4!hC!Xg;BV>3>F?_`6shLKiA+*2L;BXS56e#YM6Q_57Fhi#7>$
zh-U<6dR5ara)X5?)#Z%shGDz`@6u;n*b{}bqLye=BIek)%HJB%pg_uxU1HZspx<QZ
z>sS0BM?}|4JQ)njXISK39hP^2XY4Vqm;RymWP$aK4qLg2+K=O_z^SG)TOvf;l57+e
zgw8_&-iOYXd2eYx{-&}7@1_nHl-T0)A4_FRj(+i`C+~;-XOy*;CI{Qf7;kLsA~hZB
z!0t~P!WVdqatIAMv-vWqBTKpEZBMT1&CZo3t!dqAdLFpt_3Oy|)U`vw;cYcD<WhBs
zK;1dYHSUP`u!s38d|q%-%c)X{GZ@wG8qLm+$KfU}lB?ssLipz01UmG$WyCcPH;Uun
zkWgW3cD3{y$uQ{G>lCSR-*{jXK#8BVQrdtiyXr1{Q3rCrPk{MA{V6hzfDMe{e3fxz
zVqzhHIg~%6o~;uoitIA>9);A5kV_@BG?xV6KPQ@$yWHS1wnAXH15|rkTetD|SO+zG
zVI#!P{C!(+n!c|<c36Z`QP1|&{tk`+wgUp5whi>dZNU?U2|os<?-Bx-pd&s78IT18
zf7?&)K@`&kT=sXXEpFV-EjI%d=+IaczDw%Np?1^koAo?J3X4;jqo;JF>c9SM{_UiR
z?8IEC=C2ZY|Jx1kOX(wynCH&@UH&RRfii-Q;f1r>nLB6yd}qbO{u4?Z(d@aUfu!@G
zKc2X9##7|n(XcN9*ALU8huv3&>W{XY=RBjD#o~h!31WAG5ydfrA@^Brt_Zw$K4if9
z8tYl=!aMePyvG<|3MYWR(3zkILCWyBA~;IT3mO2)6l+BV>663fl$Esjn>)P;+;)XU
z6dEO{cC^Cq8&cJ+7IFFN)aet@$^m-tG9-MW+ARpLhJM-B?$t3VwYnpnb9=tZjgIzp
zd0y$Ep%M3;UG?{yp|f$ZqPxjzn?n%v70B9}2!B&Tmo|km*}J;NCD-`y(8<y$R{ygF
z=rFL{nD}&Bt^X8ar2@WwR2#6k6twGUwF9Jvs5a1?Bm$9<bm~oxOtA@M_@8|&BFQEf
zNvkB3UbD;D*egGT0eFl%;Z)$J!BR;zf<1SbUGs<M{J%7styTN5FAoPCMfPhc+Cr1D
zWzJ(iEJ_91D=epw=~fxH!tv>GN1Vlq)t}3#Ra7;LtQ?ZgfYW8Fw;*Tb$ev>(&ftQc
z&P)kW&2Y(vt{~oRd^49+_2Ik+pDgYwyMVmdR;^xMYRK4z%Wq)?09_CGEwosGX{gr*
znfAcF(y=HDxd8FTvi|=5@r1oIJ<`MK728;kD^aUp!h8>kAo72h)EjcBD2%jYJ}ii=
zZZ;2)(GXl<ocyKuF4eBhxYi0M6k@G;Z6zQcAvI6OSbrte8(pmsJ1C@{EtM~gakxyi
z_2d*uo6-EK3)i?M_~l*4+sFi&G|gKn>vB$Gw?fwFxC7*KCGUipZcPK~Ieu)*Arx`S
z&7-SK0yrU0A&pO_qU9?20dcHc(;-}8%I$>0GIAAI=On{oFE)r^V5m(!;+S#tSGzex
zp7m)n;f}c~zty{p!da-6oe&B&CpfHGCdWufV6=H1lsS>+Vc;b19+I#4#^75Bt>$gb
zk>&g9Pdu?POluM7HQJ!{BT>BY6lkoFb?}a%$%!sp@#$ut>D)$h9OF^KH`Q{22DAx2
z-T`fjU$PQtT2^Y%jI>6KP|B>?sQ;|c$<8F&?5WOB9aJmeQl8dplb!#UUL;pxjKHfP
zS$K)JK~vFUBnu~&bWfcMHMy=*X^_UGOX>eOu2)J&ut{khQs4rxG|qe0gT=XFUPb-)
z{)dBxmATIE%ieSI$w!=1yzH(8RN*{6Zdr7xHe$mF)NOol?(<n5ZkH=A4txZg(@M^E
zUA6MDY-0l6M3GhZj;EV06LL9mdiz3+r771p8;a&o%t`mDhkbuB#KZAYXe#c$rdaAy
z!huahYmu^Ytdw$EW&E$#^e+SP@@H`keJiHmDqRUMpVgjfB?b?iqs3YQ4-CsYn~P{u
zh2M#uHR_VZN!^pH(9a^|<L;z?@<VXCb3YeL)5~w1i9)q&vJM6YAqjMxmJ6}_Z3=>y
zUxE$=da&W%Nth3ZDNL0dfrnzzYfW`aLhRohSQ%3o)W4sm=pW;5iWapBh&+FIIL+aF
zMV$(L#rA=*ZKj283|&QLC!P{x{fx88NeZO&e1DFzpjA6m_7a%LqkluOb7vmV0nFKo
z=3lKiE?U0ysb0;Ho5pZ{cM49GWsc1CG#&k3EgG(;Y}^von(daOq<;j_WEOFm4Z}1v
z4eV!mL3l^CCYjx?mz{oi;59CJfa7)n4qA4a@CRU<6}$BlZXPJ19{JA{eU>r|;l2E2
zBe-0F-tk?$pS&(O*0~}2_MGEYeX-8-PxcxSg0ctA>o@9i%$;wHzt`66B=uRIM<ab)
zbxGSunr=zM0mQlxXGgTgr({k6U?tfpr1<zW9}fSYVNa9yRd7!9nrtWa$<jC36^<Y@
zY^yO`QPrQRy}xGY<~R;ujB{p(G^nA+uU#GNC4fTg=pb(Wz!@|p8X)_?+1jC(C*aSX
zkQ&Ks$sonI$m7L2MMFl*20T7Mr4S?*khv}q|FwY6)?Y?RqpNTnV`=s*WSjqnewxIK
z&&u2XkdnrFbPGY3!~HGfW|WI1Nkmc7mUe6()Bdq$5>m^KGK+C_Fh7M{w819I(LW(2
zQ1V&JKlP2^lUH_G%>ubzQA_63=p6=p;ya&xAsk^Z^^8cTvTGFn$4LOBO~2M>5duNs
z<zt=mwzXQpf|5e<su6X2^}5sOGQQf!pPBe?6);{YPCo8omU2ms%@rMn@RhBb=4tO{
z-!AF<vD=9;6RVQmSQ9wB$_!uBH*PSSh>19{H1@TwpGz9cn4vrFyr<PHPcZAq+^>jp
zOmjaQSo#})BJ~vu#x#<hY-cG-9G5M5xq^r;qh_AGLj5`phHO4kGQ({m^0_UFSkZp#
zobM(N6si1f)tVlv05MH{ANIb+wOOU_viOHI;ie|b#jh^(ugpFx73E<F3l-=Tjp&ls
zMv+8W8w$FhqgLNR@5G};zGbWA1hrR>7s29wL0Vo`Z!`L>pTM5EkNe)I{3TxGm5!%8
zQ9H63Q)nY-d6h--F=KQ1HgAR{@bbRZYf3gZwdCsWx@|~CnHXBZOC2`c7x_IC*N<Yx
zd!i$EBI^mjOV1!4lycG%R??6%<R0^p&);h(_`(W!BJKJKHbMKPW*a4q)xd`oAw3T)
zEp<ywf=$(^*Nk`zEuRPX?2t|2Vu#b1ddCq|exI2yzr0BLqv>~Jg(`aUD?JLn@uJ}m
zpCB!$qO*F{wa0Un;?*qfI~2!G_3e%D1LPHV|Ev}8P>zTTkPN^$u_Bv^%GWP=jvI$K
zD%oV;JzLQ9F=9J?ygPSdSMIxo?7lz~73kc{N)c-H=m^(X6*8jB2oBm6Oty&~$UnSq
z4{W|oAn_SZNVqouq4K`>q27R?8>nQWrVi^-F|d7XCgVPxce*)=D?I<GrzD7w^WAKP
zZMaaU`{Nh$bivvYLwhczhpuvPF|dQV`U{Ig`;4JG&-4Rb6XZDX>5ZVP;%2P8#Mjra
zGY1;Wf~lLLmb?Wie(#fom#+N&M^WU)!n|DO2d0Y#|7rIIKzM9>=GH{p(TUezNS2{b
zRTStBm1p<XB`NrQ;cnLTCsaSJ|J3uzw+`%1dy~LsFiKI)F4!xZZZ%erKL;@49`^lQ
zwiSKL5YRk?0Mf0|;aUD@{L!iK)w?TkQZ{|ms1Vr-a8SD4KtMD-d}&2>v-8RUvgy?<
z^6PM3=P>I(-R916xA)#;S~lopkejS*D`#c8MV&l8d7-o$5K8R_B#V+yT$T%P;2#Lv
zB+vz^b{{u`uYtj102~)W58u!5w!m1?QS@#vU~w-ZD(VgM%&gP+Av{i%x|e)CP-OAj
z8-h*8@hV-ei)%fli~eO~Sdsiu=NlR6R{cog%Y(NI6g)*rrxycCcX!0$5;=;k(;V*g
zKAK2ttVT}!B)L_}%-og-hmo#Z4%&@cNMMfSII60uiUALSH4wt6hGP!2;=hu9=vONQ
z+p|!LzV9D;(+Ul0fzKT8H@y2OZaX`IIe3*9(d@djPOefi-M5elqWd|^tqA|5@Z!Bz
z6c&#v#})Np0@mGyH-<Ho{LGxVgv~X?ufqjCMjc6Xv%H{WhrhAjE3~X@a;@2rinuqE
zD9;YwdFz2&FF8B==ZOF!>t-y#>A{s*tTIMnP2dOV=f2sYy16ec%338q5E#Wq>yfUF
zF>!hw$=0vuvV~iJw^w2;S6so;a`F0C%yIEXW#g%M=Hc%!R6BmN_foaduf1%fMXBQw
zM4eZ4uj*cj9fXK)&6;Per?5bY4cbZj^->W=RU(42E~YV#$vuWZ=-Texp1db?^eAtF
zY+ba{<zjsQyA;*=KWl2D3x00DtF&cM8lMO`&w}XEy@Igm>?BA%E<V+7g!9+iv%L1v
zkNirO;5Gm+=M++sk~nv2Q7CNA>@}*U&(^P@rxy9jB|&f4Gs&<zY~Y_BqC-V1^y_fJ
za>CEI2xzBOJ=PMC#IJ3I=mZ!79S}sFQiUt(qlBkMgUBJTKi}Nm=B{3{{^8v<;$<+I
zRp01aQqBZTW9@rDiYC3C40K!qgflo}V`FWCYPfBoSHl>BY%&I_Ghy!k^|nsgK|HHD
zu5*4t0qmk1F&kU61D2A&#j12Eyk-fy<hPg2u<PGe1R?rMbwEJNaKpf6c|dKD^T)5n
zq}D7jb70soJN+~imKD95GHbJ8SLtlhdmf9^!|pkRe50di&uDfXOsbXuuoqN#_1Yjk
zd4(e1T-tyd!OO$&ozF1>?++D*`dz|8ec|XW%!E!8v(;Ou*WqRibhc`2{TL#zM+bb?
z=}3LfA`*V!uSF!)hu&eX{wC<={)Vh<i{;0jL?p?TvRqocKG;V#Lp*ux3>DoN^vL=T
zUsD?Ft;-z!R3k=u)oe!@uVYwBCVur2wgJ0Wnsuw!I<J~NfiWZ0JxA@8M4B;U_~Qy#
zkf;FGsg<BpR=0l;G(tTCT|+blzFlhV+~W4*Wb<-w-XGialZLoKH6n|acm$hQnAOIS
z)e0QD2kDV^r+JZo{BQSr>E_%LyC%Wp$?0QT9%?DX`^Qe05Q~boss(>Kf?nN2`l=S}
zTI8uQWikowEAWx;jRQ>5iGsjuD&vZ08y=Uy@L-%?JFHtpaI6pDo3E~4MR1v)^jErQ
zr<S+}r44jZLWFW|5tUISBUaHu6}XGI#u02E!+!Zf&n|E15H5_QJeN?qiGm=aZ5J%_
z&8Zq3*C1eOk*se9vxnOZu|EE%<^7vmXzbBZm?EWJlOc+ZR=GbJr)<=Y2ZP2gahT>0
z<BHlt>L!5#o$Eaz4JKTGo?zIX0Sq!;sP3Hr-T4bZ6$Flhkl^6@vqUduw>j=?^Uws*
z|9WXa_|Or{@R=tRu;fv87C0Tnh;_WxQS2|O%a#De4nEJ<_yB*?%r83y;Vy*E;SfDC
z(K+%H>d^;&xB?Z%v}+g~;d7ooNu&5BGe=^w$fW{dBRNR2Tl{~vCkK98Hke1oRrhmw
zhc~E1T&OFq`%KKsLvVnmlUg>?65CLWv#G{isgUydg|;u}NgN}pi>7fY57=Y$9`~H|
z9&wh5nRbgM%(;z4KMb{`iMWvfaIzUf_FL~x%QBLkiqNCO))3+OwI*4G*p5C<pJ8nc
z^<URlu{vd50Rnfc8)`x7eqf<F_J_#PCf1;#+13)tcB5&Sr*_F8*^h|VwE%mma1z3s
zkT;c(_O48oik-e9y6~-L@lNtSl{zUB2OLlrNm;3F2D5M1Hw6>$AXDN|RlYH6Qlax*
zA~keGJu@5SVb^xrd=@?H=^_W2%6voBZtO|3`<SV!h{8Jm(D}Hle7$_|?z_W!f7l&E
zf{j|Fw=h;o=61cA6#k$vzy^r$-N){$AT~Y~FM7=CI7uK-I^4ZP?|@YY>quO~H#AkB
zf#A}?l#^vRQt)cAFEj_Cg2|hYu9i13|A!R9t*Aa_^~Qme<81+%23I$RGFg2Wd}@(e
zsE<*8MP(|@wXCOYAk)KIbm!guO|Sg44Suh!yz;;gMbTX%i!k3_46@YC4o==m#LC`d
zfNVisTyi15xPg>v)*x%v{&C`s*=>6`lhqH*v`ul>e+tHB%mij6c7TMNy>U!o5Xqud
zi;950S-0Co@T1IMzC5R*i**L3>@(W35nI1Q*2FNqDLa1F<DmaX9#}kpdCsM8M8PX5
zL(*xPXtUpYr7_%nb|nF`iF;3_c9VqObw?o`mpyndoEXiaqDkgC1bR^=@1o2_iOoED
zd-UyuoUOUSQt(I#xD|(4ar39)TQwcl?&S6h^5fH2m2Y=rD{h`OE1V3GvjD;o&I+gQ
zMDrI~r*GWyO1Zld{p7;Gyjq+WsFbdb78h9G>x5f(KJ;Pmr;f~#BE0^Ag?PM*t?`$%
z#fvK$|M6;yEP2o*Y*C8N%~22NLD0l*&uEmK$zs&vP?+Pp4zaQM)5>UV1xd{xF0@rQ
z186D-N-y_kVH*q8(RMt;_0XmPl7=27gY3|#W&5&$v-Yzz?oc`(w_M|>uh;4VDFp^G
z$l40NP?t1h8h2cjL}<|I9v&h)sY`Xzr-%(`qbZNKK6!f{_!TxeK@?!Tc?uSopIGuk
zwux5PrKR`AlO9FBp&-%K`19nKsA$#wY8#0vqZIdi#L{{2YNJ`Kp<Q@{LiuZjK@cI7
zN55}Ey>rJ{hdZW&FX+lq2gwJVgs6FHj?hQM7e?<2`Ehh>utztxCMg=)imsK5W9<yl
z=7E$gAYbEz7#QCMmCQmZ2kJl<QAvtv3m&pC<LY^!=lw{Z)`c3=z9t;jf!Yz;bSPxF
zD+v@o{S)|tt7<zM-dATnp)_j|s*M|nm7aE?@rI5{;RB0Pfy>y(r&geS@+u$N^I~R8
z<9qM)7Q<)fMr#A@ti|nGa!JJosE=9A#>v&IjAZPUhmXI4cSI!p@{URqq;jhC59qnf
zTVtp`Hn~vd>N*G}9GUo-1^!hfp$wQ%5hovRCFVie35jqc!&`ZO)sBw-X&CTn{4efF
zLq6JGbM4%Pl=TphFMZ`64lUeK0=6mr&_q1=x3~|i{U~(Hs9PECHopdjRf?YuX3=ln
zlgV*5mmDJSLb5uSRsK*^>b&ZAcF*5phaY@15}4h0^iP}RwPsBKf;F@8k=wwYR6St@
z(w+GqeS5tDlBHKRsj7hivT8&}IJwL$uOm1ilI)v29gQCMjy&r>sVum+d@+Q)o*wUi
z*dI`<C320Vu#|Rgu$<FOqn2LP>jWKT&_-bdMF>CDn~=<hHel3z5}K^Yaf_=Nk64@7
zn3hVd&gtK}ufl!$8d!rqIk9+0zxVHp<ufQZR>}$02v$q06=UXPyzcy%H}Q9ww|gjo
zwkqyS67Amt%Z9$=$bX40_(A$=-mrotj}TB@9DQT=DQ`o;V~#yh|907=pr8;hbhksL
zSXtRAFmtEXx<{oTiA<Jilvf<g20)jA^CY{r?9f@POMaEU7FVE-`rP>>Q<lOD>ngqG
z<GvWdlypzrhvI@a<}&jFR)E8k@sJ@&SP{Ez621b+W53YO+*DpdwDo9*JTinl{1!WB
z@IMYaL~JaV>+l<-oA}waT$a0DG*VTQQt|3U7?e)Ucgmxi)z*Y{n+$)|!98*TTXH!J
zEs{}W3e=_;mlkiud{+E$+98aS$D6A3J)KJiolztenLOE*s4AA7;viqwzX=<+*iUp>
zwq6v!Ju_%6fiOzDyfj)T@)D1Z80c*qS|e^XK3d3!RZ$4_7=DWkJpYk*u#$#j=W}WU
zNk5#KHSulUNutBSAG}kqJYrnrkwg@uV7Lwr;Z4^NT$Ej_QhuNQrpTwX?M({iE-X7Z
z&)3v|hD22ZF(HH#a-dIME*#TW0n8sh)RRc}JF2cZ7v$#tYDoV@)#}>)pU1W5Trh@L
zn@IQIYO7xZu?c~Fq~}MUxnSp&e_qm&$;Ji0{lLulN#EHa#O((LFPWI<bCI)H+GmA8
zg!AS_SDBlbr0n=337dc3!8IjWdj)7pm*<j&3uh>cf!BZIzt|F7l2>|2N%NJUESlH7
zh%FlL)^PR8gKmDS(mlBO-cW?=`e*JPpKT2p^hZ)G#p~W^;9V2;S+@9K8tztQ*F<9D
zVbEP3d?gBgn0Vg0^6$MwE>a8IcBUa-Azt6({yUjQdz0N(S~0l0%2dV5-uM(`YVUQ&
zG8&;!<)ByTh>H`tp>x5OKkT;&hPz8ycxR3xpvqHVjc3L9A$H&Ob4Jb3psI`TNf#3u
z5ZX7muE&?vJCv9uvKqk1rSZj4vAlbCv6;#7>9K4A;TG7q`~@(fs#mCp#e5a~!<>r9
zC#*N1jF@&{4F=a3zi>B~dKT`Zu`)!99_JVqZkJl&3UDDV%t^r7J|A9k6d4>bKtSU^
z#i)DrwXp;@*i_DcQq6?1>ZH4;H^Xv#wM^oZ=DXJ)KmpW?Jx93)49`dcFvyr^)G-PT
z0i<TkU$V3=${4D#Ub)TSyWF>g@rd?#?zE}T;t(=LyM$8IRq=&RE20{HeqM0FJC~35
zjSHo9g&5|gT5efX*UVf~5V}muotWVE*s<{mY#T7b>jY-|y__f*2#r9!*kE6R<f_&A
z1`iLmyv~;Y=bm*KCx`{26>yO8pu`?c-1xps4krw^pwKixUgsO{JkREsDLJu$VS`Fo
zaSOdZ$zE^I{rcsMO+Vy{SdzUL-HBMmu{cFMa#bvXY|?+^iGY*$v^~})j{Os;)7{tC
z1&nK`Lip3_76j2>Xal=}PcIy}%~fYywcWe1D$6UrGrjBv9|R;9DufS?%E8R*m+1?l
zQ>#0*iw(0J{caZlu8KmYfnkSLkiSmgHeu$<?Q9&G^OzVPZ$d=G4g)^5N`%`~-XDJ#
z%qzBRItb27Dq53euc<mPylbZv*nrz8jc%<)2a!dc_dO3nn9%LaSvWj#r3~2O^*gz?
z=T+J)y+N68Gu8v^eWFj!A|82>z%&WBv{F-ADmP)|KE~2~;$xy$NWUZQhiYd#4p3jg
z+2}i#W-MDeY-==;l{nw5y@U*$cjNt1Usp>o0pnt_H+6*s+NmrUU1r%$wB_5F6kP-4
z_!R{7OWwcOXKQPn?5}!0-|I0ifA?u*V$(so{~nW^{8mb*DfmAh2#}5-zho!Ev;$|4
zxfZuNIw8^ER~n&({^W>S?<LjJz-5u6xryb!6|aJs6TvLdil-NpfWWHzrw3^)Tk0qP
zl#mSSfGg`ON#+6Xe8~nU;i*=xO~8AD8sQg?eiLpwnu^4|;{4U#PXFiyjKT$t%G*H@
zLwGN*owqN1{g%3_hD1SZ<)?Sl4}@$N>ZgBlg~C!rixamB$G+KZ3JczWqg&c@X<xmq
zly{i7z-$wKxC;zS%CdH{uLq`YFJ|4LV2MU7lCEC`PGkXgeEXYdNxB9{mf|+fNpdN&
zQ&3Xa{ap*?C5^_4{riGy@V}rMuUWKM*F2Rf_o!%Kw*k*NBx7E;S!U2a!4MQzvJp=H
zFPHlNmV~ffEnI&ocgMVNy4;(#0sV29(aplHi`^*$Al}&9Z14X1Ei&i%#}?lfTo^rO
zkD~ncLmKwj0o;JcO-m^D7!Qh#o@Jdc*`3yDkLkey+*8Sq|5mJYr}T$=jUBnr@{fG}
zW<=w_`)Y4QsT>;H6H$=>?s3c9m?h7T<ky$8cWr958%~B<gpjpBI`j2g<cTa0u<VuM
z{1cAjEQQk?0UjP6`3xi^LtRzq!WleHFGBoPl-cg4azZ5d_$f83tt@&4-bB<rgsNK;
z88sL_zsa@4IHWXbbDtjd%pDG=vt3kOQ`Of+pjz%}PR2wgYmDhat3*9qA&A+#FKYui
zKT(OEIVO7C#)Gkm`hPbmDwnRtWdx^x*ElS8N$cS{koq@o89XJHOk>UmtGQRl&wjm1
z%`U=NT>o14YxT~&u66IsSry(2gP|>Ok8u$JtIrDHTsIzP19aLMeW=<&^4-XsNexXl
z4>qW&0xngU!i*oXH!6<$-wlp`G){{xDlCjd=RaUY{|n$O9)o?PGCdf}n)g5!EAjCc
z@uDyHaX`rpdK{W0=%G;e{~^=>(m%~ZnxeCRVanS{bT)IiLXH_KZ^DOei`p!>qz##!
zwS@zZFV_NFIE{#5uUTLLbu$W30e&#=_3RQK`RX+6O0D&7>Xdl!sNdpcAf5jf+bMX`
zEXvSG@>#TAUxg>)Pkj<$QzXen{~s>^xAN3uJV-q^&mv94{=H-<K84!I87>CA;R+5h
zi-3|PXo$z(&N0!0SvgAX3AzWoHkGG}0;hgD!vZYF2_jmS_yn5*%0Od8#7hu8R+*uE
z(Jh`#@?90fkzUfk+arLB9_)xj5htZ)Tk-M-kqP0%jz6{Nx%;aB6%*X&wY_kVzqCO-
zMj%HkZpSKW?7wWmL+ORRpovu`qj`^h=w}R|MpcNg&F5~?<CZb(jeZeDls#<Gsr5#z
zCzyO&e{*Yl@qgHR>#(ZUw(VDG0qJf8q(M<SCMhK$DN+Iw3KEl)n5cAvbeBkXcekXZ
zba!{}d#<&f_g(9G_P3Ay{q^l*AMgKiIOZ7lxW;{7*ZDioB@|AB+|?aa*kS9WxP8H0
ztJWdtg>_4@=}yirQBaStuM!#g>6L|M-H!2jxHf&hPtoLLI0IBAQpHDs{YqadhMBm!
zK1D!%PNJdS$6<1M?po(gE^XX*73AfDYCI%>g&u%RtPsG#fDiLwD}-}yAFZqn5uYc$
zjT>DdP}r86bDPTF)LLBfxwV`q68kVwL_6NOaP?06t5%MqGJ2R8$L-_m;W_J9sS$3o
zT5*?v#V#Y@U@o4{DpArftJz8wqY4lK6S=-xlbDESRlHXS0@7xOG=CB?fsb!}wHMs*
z*_ZapQmD_n?M+~F!$6VTcj5u(Zy4F~knGfl1o~oj1UWte4@-261+KJ9?Q|mhMJ$$W
zhWl5T&A(lq%+J<UDH#ub)em@tgF5{<`&|dVcc@GTgYvB(bmb`?eT6ATWH4V(D9=Tj
zJk7!Jskd>0NeA>*jLvx70<rjLg+8_<K0bB=)nw{3`Z5tJykKMZzL0tBBZGh_>sMN%
ztI2du8m=e&qQ)6NRcsVmZj25*xf3O<3A`vJJ&7avoJ^j-cNjROAzYRaYGuvVR+2H)
zP8j9SW^bdOeN;mtbMZ;}qSg7d5{pGdA32qBitqwTuTj-K(3(qq**ksk7C<3GqrTIn
zq_*RV^KBkatR0C8L3P@C5f62?_oaA4*vI;(B&1=TUTN=UWbTk~Pi$eP@<RpZZPBW-
z<Fc7d!@9~0@<vs?0bG)@7%;kRO8;A}M|rKxVep=#rk|3sNfpJBg!_^7L!X|YE$lMS
z!~}AFx2j*07T0|DHz{1EzuYQ~p7}gGlNkLG6c!$M(598GQ(c~oDc^$2p4uEQFDzIq
z4hwcYL4;gPI<<QxRyC<oehUv@7}K`$0*P7Lw!Pv!{1{GLjKj|Y-Y8%dzs7{u`yLv1
z0Jq%Z4c7EA?v0Xic%+<pMR}&0a-yhfixK{M8*!)ebHS6Y@+QOD^tMu?*`>`<(?;is
z_u*d1s1Dd~GlycmK%>IzlC$s3wN@n02YkK9P;xIy#QGymYkS&C(>s<ne@Mh64l@(3
znq&PQva4Dl+pQ*FYlamM4ZNBIk_C$wKJnJd<Gt^vdv_YrwpM*^UqJMa8-T{emptCj
z>+tM^v3mde^f{5VsCYmWE}be1%jdxYcbYnvwJ*Y@+?ce><eG)|pF6odh~MYOOHV?9
z9BUecnwJNYZXBM|7R!<|^ODn#I{bL^uBzSqCVK7<5GHa*Hjer{n>@{)tX2o*3>yNx
zksrB(?g$^8-z);G@Y~Wk3z3&D;+T$k&fq|Q4_oz!3$*{aJ7_xf1%3`e5S*Te0{=kY
zI<(dA)QJ?nE)e86Oz2bGC){1_69{BWCno(T+cY%x<)8*}Qx8=Yh*^fy_P_iQR4d`g
z)BjjdyB+LUxVk>pKT<*u7wd=7X+}si$K<9SUfCBIP(&xRRPZM(orluq5*|vpD?Ehe
zMuxBucRqfysSy9L%F2b!16=-;&n(W})!f}9pXBVL?GBsx)(rV*ge1)+dmKu4^FXm%
zklH5!40QP^Y&x#NJG(OM5jvkK3LI%4-?wzk;tSVv90>}AZV*1@W&wiMrz1`*ncxVV
zq)@vgCbKymWJ;Az06C7vJS3^t@#qW%A$SzoX$W%qFf2K}*!-=pruWhxVa$*p(2(2R
z1gP&zxa(3vV&s*2nrM~JIu3B=s@~_l&4mS1pQ}T3?4LRA>cATVj&DPAVoaX<xkX9|
z9`p<DA=QOQ{du_}&}EqzUaKSod5dFUF|csHWR$t=j~*j=O<=~=rJxtA$8&^Q_`B}G
z#B8<l1>u!ZgWKC?N~z}HZezpRXSA`DJ4vp3kkzDDnqiayzVKGT49l1+I$<uv1QwRd
z=dP1(I#0{$g!bSTNLI)RPlN`r;lt^p@eh~xxVj>7yv}xvGgtb3N*w3BtsSDcelc}m
z_I$S#ripZ0krT?zUE4AlQ3p`$(8tWfc+N*YYU%}d9U60t<J62cDls82fV=!G&&)`m
zILf6yC*?i_BD^nGdyUkypgYRM`s=hb<cD-+gn2>rV?*D{52Gvsy+<xX2bCQp%NBPw
zt0m+b_xyrhpx&jR=p)BfpUXrNCW+zuaERx%f=yyGs=6T~y~f82QgJ~?NtAT!F<9tC
ziYJE6gwIS@TKjoxET{1iQP>qyg&8g}*QJ4{bR)(P7Hyml>)A`w+}raY_BWTfEiPyG
z7K!_N-obuo;p$%k%_U8P?60n-aNht^VZIaPGD?7wMSgivCd{RqciBg?UN~(csF^_e
zVkQ{N=0RSg(U!C!AwqiN(8G@QEteRqz2CS8#MnFUM3$Tef*nIlU%V4KVpQV$cJq{u
zE9oQp=QEr48$%R^&R|x^Le|ysmV!5@!ct`LEVYqOn-||ibJ~uUBP>5t4063HO+_L&
z;pQ4ZNBf060l6ZoZEs!>CU9@wFCI8ldcrwH>vZhv*5}V_l=v1rHz5gG#)29f)(1RV
zuiZ~h?F~lTsbtTTLV|~}AI+~-h@qOlvfcZzS@B*>bQ<{bz_FG;734HG+~S>lA;;q+
z)U4It8#R=NIVH|FjEth_avDMRfk~p&f9!`Vo|&mtJc`{vW!B#tfjpaVqma*mi(#WM
z&~qYkAHVfs_8%mm%})hp*%Sp@E;dob8d@RmSg<9g{q-hF_Ht|6EBw~~U)ang>))`M
zf3OLIpZ^160SL=^*sZ@2qw2TY9vNTV2au=Jp>k^Q5q0Nu#5bhh1dmz~o~Mwv7{xIE
zrX}IZHwH5Kemy>LiGSfQ-$5kA|AMx8xc(os4Iq1{|L5oZ|Lp1U?&=%9D8#g}vFSnb
z$NdMI0YHt<PY_To9FhQUO@%K<`!~)5PJQS|f5wE4(=vGfo#%l6kv(8NDuW3l>HiP>
zf#?8DrpFf&E<I!cl2+0ZyC+}xV0oxdg%U&)Y<E)WFKkZ>9zg!<#l3y^pL_&LHJq!+
zl6dlWuHrVK-nvEUU--v2$o~!g0Rlt-SFwqM`8TfO8w79_`}qIDKbDpL3wQf3g+1t4
z!m$wnq`zY$%L4!#xo-Ox{&5lhpZEvq@*mg;R`cJn5%BX`{^#fY&(r%4HtBzNagv{L
zK26v3uBzfQ{>tJoADh$mS~53>rp%qw^K6HIztRy~hK(Iy(x;9j^b|Jq^c17$P2X>e
zvpxWYE<idSj)uhXd?b59TNsSG@FOb=jmK&h)8%yA?|NsZmUwu0I0ZPdjk?2yGt0}%
zF`8#l+Lm87er0Mx96{1-(Yx+%n|w3tbU3eXMhl_yPy`0|ybx^DT_0;jB2Qae(VlJ*
zYj(EJRo)M28y;-k#XZk&=*akf8El%ee?<Avg%B<6c6D{N1vq>g03LtTotc?=Fii@h
z9i9@h%{`VrEEdtI%!CB4)nlgCQS)vgCO#}IwF&2GyUoM^j^y=qTWf!*)Cqj+^-rYY
zVX-XVCW#Ey6C^N8^21d2_7eVynFHt7&_Vmi3uN52!snxDX)QQjAm4}&xS_*`rPg4l
zEfLT25wo{8lC%vAgwPSIS>BT7R9N_5%l@QM@u}er^!eUDGt&JuC;-6hYAPTP^Lpz=
zRkK_(UG@>b@|m}d|8q_{SWj|9GHN8m$}>j+QqmZte9@}}4z1ee9PxoNfC?GAd+7C@
z<0aM~RESOw_H|EG9inl-9JHE2s1|+>JFcZ#eWoLKpz6NzI3hB#83+TYNa><D^|642
zD7+N25FjdEz-XdrzS)D8nYp_tFFF?B?|Lq`w^vq9z$AVF92r7LmYrA9e%T;w0+mQe
zVD*Cs#p4S!3&cSIvu@o-1ETHG$?r`V&vsF>HFa`h*QC<2Zc%1wqUHQkm)m2N1?~tx
z_sD5awEqcP&>HVs>zn-aFbm}bKBt!TCmKL|?!(TOjsbkw<E<x+qoo)>LH8VeKI<Fg
zqy<{lVXv*kRd6qby$_vsOnJ}Gi2<koM`R5l=OqIoqYpLEsk(1a#5yFDScdBY&l%hq
z2859rY)$STG1^Vnd8UVWmBllPtArd;9F_DoVPDJX#Xay%9;ZDB#8SR%zep19;JnE<
z)Q<-}*9Q}wjaR$1pY@fcBzmLR7Z!Y|y?Z$uDS+{_)Fj%k3tIQhp4H{$CXnqw{<IBO
zJIs1BJF^8y)gOZJMFLvl9}f^<8m#(2Z`RcqHTh|w=&@{15i5<`qFdZ25Eip(mGba#
z@D!5rcKQSd==!8%7)~}ZNY5ZxHuhTri$>9WGAN%M92}O%Pl2vuhsT5UZZo&Ag-}zc
zIiogl-qa`5<rSwv^&f61j5W&7FGQHM&%N=9DA*O{zRw-d#IuJa8UVbcqlc!G^DQ`U
zu2?&;irI-eUS9u2cqEj(8eEDVMaCc!b#|_KG_&PPeMmg>19Er$%urG=_S%0Hv?mE-
z_-qXAzGa@lmQ2fS{sw?B(0b?)lbMtL#bVGTBxWcziL`g{e!vXH!CuIFd2cm2gBb75
z_VGwOH?iS{HYtI}b>$q)1B0XLUXh12-3W?@P_3!U76$^0ayJd5hYkZ?MVcPT^6M`^
zs3i`BS|3N-Wep%G$FE73Hyj}d%*5EX2qCoj`Y4}3!({=?OrPY%%RH_T-~b8)FtL;B
z@-n*)g;b~U-3MA>#&d8|U!mXi-y^x%A?LNkWME(jjf_M#DP1qDJ4Dy9A62K|<ixvf
z8&GudHx;^{Q7B|3BaTu{EA%*a0q02UhEvT_XMp(K|B|qH8o?fieV!{zslri1FueCk
zEpb~*2n~gwqh$ju;dl4_G$cUeF;Zrt!<%Jz7;T8o(4G|>V1OLET>-z7w;;PiSr2wk
zm8+{fR44`v|0~3j1`o}o3yfsfj=z>2I8q?}UJ`6Xp)VJs>KwBM@-eaz{@XUhfbn<=
zje1|o?@v5Ww}aWNv3F*X3C$s1gVdJ|4I5{Q-VtKaPJz?+Phs<#r~O-~)8x7z49wOb
z+s~??X{dA3=^+g$c7De!NPRS&Pr)fVBqWczg<JBFnk*E|<GtVxA`L+BalTzTV7<np
z@EREQU?>T%)C-1PTSJzylY@4s`K)Q+5yen-0!^I|dmz}BHW{$Fjzl!P0Ex0KAmata
zcJ=3w9dV0P*__)MB|Oc9E7nxAaliBzjU8PdIsb??7tzHifrn7&6crqz_|sl>CbLxS
z;HI+bX`n&Y^l5Gcuv?{YGr3y&-oqY``F`f-vCEW1MVhu@Pe?u8S*HNPd*DB0;St9o
zuVBKuoiSNzD!;AS*d~A}|6CE{qIyRKupdqr23)bAt*LUWFfP+RpcpPk+>ovH!=l|q
zF@85jZ!1wclOOi^z6YQr9&@~$z|8s@y%K#*sMw&{L-|yhCGNmetvKn#F*35Ei5a-c
z^^i`e9#Osn#;)kbJC*H9Wkveueq)*nn&n2do<`ozY%FzFX*F3}2WGmXHI6r5?zpzc
z3-wQ4H^T@dJFqKAUa_#Tl^u6+n~(j;e&{+b1N*_`H;YRzXd9P9s@KdalHDjn_1tgl
z<rS_#d8X${&Fszu&n3!<kA0aK7PY1Oigrckw^C~Ir*X{i66Z3_@2R&$iq&&m+Pvs5
zdh5sPXP*JZwGXs}f#Uf{;#-Yz)$hQPZoS+?9vgT{PSy+BxE^-u4@91@G+TLGgG`<g
zH8UO~j^bF`1(Q-;EzkEmIG2;ODPl=8xgHUa)eqTDW8o?nIU(N%lXU$zjda*ENf_Y{
zOkODyfu6m+A653-ht~TWifVtj5-BAA#CHl^%)G4`N?43l<vz%rDC&QB>Mc};a<Kkf
z!VopqEVPu}G6RSnWr3(D04H;C(eOcM%=%@|88E3M!1D1r|IKhHe><|&FnyY>n)&+!
z!W7x74`X?;kAp4F$hGSv(XW63Y7Nl<3=9tvf6J)5-*6ZwyZqov#DwOC9<B_Y4{<@|
zRz*6Ad1S>}eV5uzgmP{tOvE=7E8aYv6j|PVQqjb~$*J@KB!^}Gv}mtr)9BZppLfTF
z2zm}S)c?2_Dpv3iw=dLu32#fS;$vFGr_~2K)G;%a@-K71?-H<ln<XcnFtUUS`(#X%
zVRUMO`HoI;@52kuxI2*ICG%6^RYYAJ7NaW}k40AyWP)mSOhvc97r5c<0sY3H#WP!s
z$o+uwnJ1vzr2zYR)5*a)wQ|!m@PUEVed-bL;FV2HZs+Xxg}d7kN|iz<9&l9>*fYk=
z)TR?~uyUub9%*%E-+up6cMxvP_B9#h0&t0a%v><8>(zKG!y!A(B-~}J`*4-asrh3M
z9dcrb%28(sh%?$oJitgBhvOO}w*YD4w!x1j;c^bvOG={U$lMx2X1?iWOkVDw$-4ub
zJW=?EOImje;%9w4>9v_MO@%}2dMu5O7r%i!6Rtyr88FaML5|%r+I2@|K7O|L4TFAT
zjuS}nwZ(_(aI_<`?`iS;3`qJko$xwSazn7tR;aY5GQ4HQSb$)c;mSC4=?`RrpP}J+
zSboM_2a^xELgKt~=Uu2d#(mP|-&N;fz*3ZI{LY~9P5Bn}KrX)B4L1S}n0I`d7D%A+
z;yNP@tBztM*K^h!r&^|^t3a((14ho46o!JZPv;{Zc{ibrU_$C)&e<U0V0hN-KUoLT
z$dCAv*726@!MKz4V+(bA$w!jvq)&C873maRG|^$!PVAp-3-HW%tZC*6A+C%+wFb<@
zliz}w8VY|+2k_=ipfs(`aGH6id4XCk5lWuTsJ$$qD6(7<c<#ooTccd6^l;~Fwg|K4
zcC()R+BxNxB6vypwrcBGMn_5WY;Ln^3;WoYb}H73JYeEJFO{%DH*jJ$Lfy03X?N^&
zNPVTz{~~Z_qUE5ij|><#&6|!rKE@r`9MKrcBm{iN*Fa1tGR0z88YA_qd+c}U=){6Y
zpT>0T$vq+uVWi9ORP)Rxu(!o?e=70{rC7q(`zw6Zj5YfU^Foa6x`9>qdGem>WZ<_@
zjFi3(BwhEDxb(_i&5O;D6ZrL!14}gc^`(xz_Jji);3GOM_2Q=BNq@yN0!yx`*2E=#
zpu;kNX*>W+waD@(D#USFS;{?XrZE0{1$H+LujG)k$wK-WQ&TX>@=u$&e-am02BAxA
zRT@4xx%ub{V#Tj$tSP_d?7j=ly?4PYhf>YM0sBrQ!soYj1@$=#kyQu6BWIdro#E(K
z%<-p_Gcc8q#z?z-5g~y-^bneeq2I6P0YVodI|{DSfhIH}G_)zGd3zfxm1imVIk4AV
zI~N8Se&W~LI2T`%U){*fI>r8y*V#%1I)+?a+`uKhu){p3laoAr<Z4R#UtC>b5HdAA
z6yO!v9nNUm6999q3UN}v0LDNgaEqW9%c7vvebg~bB>Ris7|)FQNvQqSr1BQ1nDS58
z3yhpMpXM(nWk`|Fy*Wz~o{;L;?sBqZYpD=A9+cn^kxGu#KOC=hv?`bZXN~c~EU|t-
zGj|?{pZx+?ORiiUzCsLpQw}FP3OHG-Y!ftw$m(|tC@OrJ9}rK-uMMZ++jXS{Of+yh
zA18G&$D_A+BAJy~=057Sq<4I)gtlYISdc9a)Cv+RXL>nmXKAp6k#3@&O<-7&WrmbA
z((Ry`tr)JP*{a$MylqJ?KoFAZGMfIW-%}_$iQYk>{nl6$>_GEZ$Km63MlY|kUGIe3
zHS5qp*~H(B_G8H>=xUa)!I8uAGiWKJt*wf>-k^k99~W;I64F!H?vu@BJ@7Hl!~e?R
z{=#Wqw*4ZV(}%^u*~ZS-Y66Xn<AMNaPtwALs|)?k358vWBkHese56~wsKvx{$8>5m
zAHV_|Eh|^R7tkAye^^W|h<(mfcc#4l23f5@u%LZBE>tm(4`#p38CwiRtuZO!e9}~*
zqDIiHDjJ(wAanF+kr<|<9pmTkrnPc5r(xP&Y=dae)$nR*^ZCW&Oifg@wUwTsu1Lkr
zk?S49gxAk^URm-ma)kreT#i-8@7(g9uJWt(Z*^#Ws$oAv&D|c=0ju*|?XQh_qbG)V
zy*KF5zkpvfL$Q(T<oFwxOk@qnV6#FfEY+Ob!=$TR*<kAVR>FwZ&1k?z(xlt$XtX4j
zkzxGV2M9e%wpcVo7%ivbj(J6wR~&z-1UqUytU5{W*S<vcs6Ygk@zXJ}UH6lM;)0Nw
zwFL(^xZ;+#AmvIbuLBRVviD7%W8rsFSQxDvICMp&w6#WFp6v}B$jHlw!|rYz<%JT6
z4)Gh;Yoa;k9eY@{E0}707n5`}2&{dj<%sK2`F$e_FVte*uZ;D@2@TOOIZtat*vc2S
zPMSrl`XV$?fr;cX8q5i5DktY}LdssV{@fy9z58PtCnpo-SG}8Vn(-GM4?T2vCTuwy
zW#Ar+S7KC|C<?bN7NJ<gL`=8)lKsGt!2jB*SvTj<s^&${yQ!n@b1D4*!)*uL=hr{m
zq^zXlfQjo<I(uPA&E6kdA&zIR*9V>4ow2~+Yh;CSqx4Aib};&S!YE|riRjs@<J56-
z`5~OngUc?ukY%%1Sa5&4xwp}_*KJCc=)uKWrpZ6n2M#T(diJdg;w(Z@dyF#xY#i!1
zdGE1fPpIocd+0$Eevf28@wc@f65aBp&*d4<8qv6@5^`eKUKr--YQ%h#o>myK?^H6j
zm|}R_;Ihf4eE?j3&%q8j6<EwvtV?2`u7<}dB6b-|k&|SrO&LqomU)c`*wZZCGKac0
z(_nbbu;M+x=>606uyRJ!4q1*|@BzR~#(DDbzosmhjzwLYb9(epH>eT><Z2U>xgB!@
zDUa%R7s~8z4pzRg`3X%Gm9)><z=5L3eoLdthMuU|>SC(Win<7&8$bk~h?!g)x9L?y
zUt&S7twqlwK*?wYQh+aRV)OIwCrcQFKu`c&F9ZFg$8a7*{X{eJYka{M|H5$%-lrT4
z0M<FWPfsyaBB0==8m_B>0KT;tU4Ur-eq+jSDgGO4Q2ib+HTa~9asm;g*~dwR57gcR
zO1#0JwscR&03z;ZPTK2%ijCs%m1O+;V}&?L%yU&GVZsSc=tRXw@@wE>@!8uASnF$2
z;ZHDj`VM@88)+WRygaU#c=8?-e8TT-kS^PFGvjo+#w?LPM>;k(Vayfd6AZsG{-X9e
zrSGMp@fOMEfD<X6phKjpP$O7b7NkIH{Bn3wn094Ro;DlfXgnz7Mj#a2Q~D>Vjdx$C
zY279$=oMdr2v-k|FUKy7)GvL9bQQR}<R@dMz*w`3B-I=Jp{fI4B499_l$ZCwWX*^4
znMsqjRXy^`%1Xs=m*4L*oBMd@C=P^`6coy(w#tg;&Li1%(WJ0nsywIrcug7Ve8D6Y
zC!iAk5IS~ehB_wGr}P+F!xMQl@=Y^I(=`d4Es3Vh@2X}NO*7vg=B)B>n5+@<%;T6x
zM;10*_UtEi?*)@C;?)$+N6EeU(q|##IOkRgcf9WoAw7Tdt{@Y6Aw2d~I(&I0euySq
z6$o?Ov3_7*vw1dGhH)3~|C*?gPqjy+v9zFnF|+G6g&CdZbF;%}$G#`J7Rw?E4T}q;
z$oQQyE4h$t5tTQVnt26ctpiyzPGn_)y2s^)OU`s$1Bc5EZ}v7PJt&`a9C)(78KHOv
zG89%nmy)~>UA9{5PXwOl;m<22M_f8B#4^i(yG2D{N3#8MVET6B3XV$z0<l;66=%*_
zJ!&K$N_yqyfrm2Cg@B{LqLidS*J-D$F*PTz8!7aQ|HG~4O4<Rt78t@J6ii^%Pz~f}
zPD2Ga%9?kw@;o^k&c$;y3zHrzRfGo+5(W31Ptv~5<^yYk$*Z?F#zt5|{LZBIK6W1S
zAJG=7Q`iDgnG9ed0N7;cM#o`}LuEs|P;TK6T3guq>LO(!b1Ei>g}D~{PU4BFwUvav
zx-N|4UDr+d?5LM?RsSNG6pv7_D)N1~c{0#zd-KTz(}Ttxgc3(jXx`WMuDVPtP3)&%
zEqJsbsm11ze7+LWsP`VRq&u-kKmX&&>+pw%Z>#g3$hIl}i6HbNhOOUim)g&SrKj%N
z^WLX_{9=uw4%;X7PY{7j#yA7@7x(VBD=ehxwEkQif@s2OH;~`aB!DJLOqhHBW?=FW
zL4lJG7UkQS8o=tsKY+P&0T27?yck;gBz<6zKrqYu#g<kyxXR7YFtfBhR{J|<+;wE*
z9f#-Xp8AL$OP9}nl5W!}MTo@?%gP$_VLF|?q6Ol~kH&ikzz|UJbX&yie1p5}uV4ud
z!T%1Hph^8vohNj$9Bw`JvC>ASNVgH*X|P`OV+2Km@toOx{m^f~R6mV7zMIJL#WZ9{
zo#3Vm10DTyUz+soV6qNtkyKg}7RlEFU8ALnHDcHbHHSw_Z3iojjIo3MZv_jXA{_Y2
zpm_*AUQ!i2l5X??+6!?NpMK@Fwu(?|6>VRm7Tt1tjH@Y%$y~1EXy{zOZ{r6hBt9o8
zpXDP?&p_#9e?{nt2v{->t}m`?7ZJiUlS<}hfs^U|c=olT3(!?eRn1g{6OJ%@G>t7K
zlxp?wZ9RlaZX|MRjV-v06%dTg#{I|9yAI-mD@wWba0i{o^gR$gtR~mvQDErGDB+<w
z>HLi(fn8a3pS5b`F2%4<jzJAG`2}mkXTZQjZanoLrT3r+>K{s%UK)2m7TDfm0lW)K
zNt0?AZo02Nur_c^0j>@EKow!b{t@_41d={`Tmli~*c?H33I<;&p)~&0#qVUc3PBHk
zz~+6nZVov%awnwWwibi)J1d{fL-Rc0U#Wg$VD5u&=n`V8KcRau+0t^`K9v07fXa=T
zbqttapY66Cu+-f-T-YS(810m_V7}{9@_y;dFm3tZW<s(Bh_AUX78%(&J(81u=9LBL
zXym2P>!HkOmR#8{9TT0@e8}XkWIf1~R~cLx$oi=6e#^0fIk*cM%7hT5KA!inRFB<$
zZgx@?27%eJ(G0txoKC)Qukn1^?{srAo^gKeR8Ro+fAG}PBa-nK7kr=vLFwDJ@MZ6t
zf`zQb8lf@6L&N_Hw|J<{tN%fCLotf_oNyWzu0kA}3vGAQpZcJz;Gp#&Pb}Dqrp&4^
zg7-t!A@byto>CXb2mrEkETYCtW<TsiB|dZ_HT)R5lnMbEk#LcJe2?O3L>U%EixeZ+
zfD0fA?f`2PtT%?n@~Z>j#|1G8v9$pB{o<K9gbF8Ow4Tz3dUvZ=vE&(@V}d^ZerxN(
zNC#g>FGPEz>PC@d{bUJRgFkCI#GzC5glmXyT()7+<er7YM1CjQ(SgM6M1>a~)jfJ6
zT74U(!)cu>0p3rC?}CXB-#uH0TgnG=!$2;;qFwQRv#dt0wr()y;Rhz6Dudh)@_$7*
zxM;nh%ua{`84fB;g3nZFb&6RtrS;{B8&|}<uuB5V#e&f}0ePOQirkT?J|c7wN?V`!
z1}lvQM!bd9Y^eZ**#R*(hgq$$7E$zXtPUFX!)kY)mCEAUU6c+dlz?7kk;Jv!YJY|J
zZq`eMCTwEM9AgZ4mIPkQK+3N9s10Pxd%#4xQXZJvO~*9puzUxCtP9=xuA>`pT$7ao
z$F*A!WZ^@;h{bAAPt;ioHq=R+hk*>Rg=?Mufcy1iEf=T6h~prOG)83k<=0WM944QC
zCj&aq=aZ8DP6p&w{XquANwr6tr_7tc_Igl`h{OP7P{^@%s*vZ_ReW1j$tTDVXpX-7
zrKcikaXA0+%a|wke<&vQ5OyLn=}{if`(Trgi+Ug(((>NxY|l$*(4|SJ)^C2AZAzWR
zd!i*3Hx-uBo1AgV)i^s%yC$)|2O(@Po@>1w=mmc3EbiA-hlLAQl7A*gV2Yj6nsNj_
zdDn$wsq|4N%hx4nshU^}tvCvD%6czYU^A<$<>&h8QcVHqut_pb)_?$`GXuD$vJV6U
zzg-_pg;$m0HYIafM$6eRH+3_ikjuvDYN9H=Cnq^W6uF+Q@e(>WC^|L2kbp!TB%9xr
zI#Rw>A?R)7pyp?Odyi0R#1us=jOkGp5+Pw*qs;TKxKBUhx=KnRBYu-SHb83gy^kn{
z`+&v!3<s%Or{B?~Qt+<8i{Ie~<LP(dg3GJZt=P?ixZOE5`|(q|`Fn;Kj1#7sq0kc2
zloKqD27^rf_g-nFzfzXVH3?*$7FE-&c-HsE(z@6k@I57F%gr^<L7S%vVX0XZm?&{e
zSIWT^Q$<E0)!W-y`frbr`D@6LAmm74hu<VPzQ(yFts58s*ANob+7r(jRTCC7%el7N
zTdjPnAq>@&cg?vy+ZHwxuX1@y5k2rC?TGglDH~fdsl11=XVu3@l$2fF`RNk}#bwVf
z6w4&eYGR!Go$$1gdRN0+Ki@~J<M?WP0o$4b?P6qLinZKx22ZgE90p&FIzl`bneeEp
z+iU6`$w?58alm?IK{}<iB2t)>##(tg@gNiybS1$0p;3LDcHApQsjK=H9nI<{`4MgT
z5`*#?zdB<^*TVH<G@onu%momtHeP2vw4E=~G}x_zBTq}k=I@n?()dgx$Z)(Hvg@j~
z^QjrA+x}|jhI>%pe#xNuBch=m2R&0Lo4%y>#hj(5=Wlzy8f{Ofju5v!pKU5)gpOj`
zm1!Pjt6lBl@1g}#f_9u3@Rs}wRQr)kg%YX6C+@edB+t!4qeCj$r>t&_e9lFIPOcgX
z1B1LcJC5}GU-HKn6(Q}fj=)PuYRxZyh!25wVK?N`&C(X0&ryfEAl_AHp5lko);mH>
z$;?|QJtcxW=0N+b;2oc{4Ir7JcDYF9Ha)J}MF{(wGWsXaz+?ae%n&1~DV<(oRYMAC
zYFz}Ic8UxoKCD?PpG|5N(~lP}>E7jOR2|Rordkg#>707&=Xtex2mSfe3hg#(xAOvg
zs&mF*u6W<i4lK#;orzM4*YPh`dCjS7>LLT{HGe!Ken2RwlZXIY0V-@>*jDoF$hX4A
zTe2FvjYYI?K-u3?W4}efHF(i^JDWSt2wM)Nb`6lOlh8Xq=3~;$qd1Vs(!wC)&_^V3
zA^mFIWpj*xZKO=ivA>=k(gMy13_3G_@W&ImSjI<o2fip@*y+^=r|S+%#c-H;`IjGM
zO)fP{Se3Zgug2^@qk~x<bN#U(UAoC_Q_U_VIb2$mG$jm|9#hP)iwiA9^By}h-Nh<s
z1wKH%^4imCobMy3i;<4i>j;+!rfUR;*!vZuNomusJ7S-iwop1wr>1oF2F+di*w##D
zC+M8uHiAusjh!G;g%}bDw|~_<3?^Uesq*M)0lct9X{?K8Vn<2A<uJr6+RQx$|Muk9
zJ7kC`V)>|4wy*5VrH~gwjb!)#NZa5k136#7l!Q{g#5G6}w$u@np6S=lwa^H=KHF2@
z2R|<x@z0;foOFdp(AbvzL;EW1Aar&Q#ihk!HQ4c{^iAS5>aORY{|b2P`mNH{8^sag
z<#*M6HLv&Qs}jsF3Vk;;b+OB-SH`=Yyh6gJoou>_-S$RviRYu0?cwf;$Dhf$h4prC
z#vl|$Enise-(LAiV=Jos%AYMw)cCxUgYfB7kH<qWWq-gI)CGTb1{E)|t~<F4)H2qq
z{jSx@c!<S+8(-JT@;9>6C)zn7;%K5bd7@1f&fZGCd|WxI$fgj&nY50)M7>Z^)??-v
z*A^Umnh^uDa$(Zk1Sa3}vy`eh8=#FHS~}r>u<tYB$f>_Xl?}e#mAZJ@s2^RJzQ);8
zO|uG?nw!shPxAQs){oQ9o(#y{usZBS6!ocE?L^!D8r7JfT=CPqM$c4Q-WPS9@U2e2
ztm$T}EOsuA9~cgAcR&f7xTNW;a6eq`cDAsQIMcoZtMZ2*p=|=CLCcQ}vz~fDsZext
z8=aCvN)KNv6!uI}b%4f(hW=-bhC`skC;h0nLN@zVHpJZc%#$cWBG|<P$SM7wErQIy
z-vx|R=yn~Xrq(6<%LDGW2d8CiPF6Tww3Yda6(m)vQ!&3hgH>m{Xe*X|O~6xomQOs@
z%6E1eU(JK_lueJXGh*hwcU4<qUBZszbZ{$uZSr~Zdqtb!#c|s<;~Jun_~)nG>PMf+
zU2i=f;7_|on3e!fzKUM-am~fI6X4b{=_plve8h-9he#eq9QEGoDcW^jL@ygQ|FQ|%
zMS|XKr%B%n-cosU=7jJa?e|}yUBzE5QXiVOO*{c^i7gs802xT2u)v%sBhiLRa&@|R
zh+^^IY4o~ZF*tK^BpZgPJYSRQfA1fBSTbbO_=3x0uyOFRykOqPj98Jf(2uslBkZfI
z3#TNFBhGm&Vd?hP0QXa7m|F#DZsA7BQa-VZR!q71xEz3Tep#GRyKFz@q~6Kdra4yH
zd)HUmFCf28L(L<te<WOV)1Lu7T^sJjL4sM1_)tK3mDReN=u{=P8Y;*6i-&ZCzHpS@
z#Kr|#c7Ll>S2{T5k?~3ByNXieTR`fMnQRur<Ehd>Py}E=P|#msfIph_r^R|P8lMFB
z`f7<Qd8bQ*gO+M>cUbGysoaIpix0m!1(uORLKdhuaT(BVdN<rNJk!hDt$yN*%u=0j
zSshc>kVMrrcG0y(7=omEEreT+y(C;Aj*r^vNt-`yktBxTPe?Q4H}E4FNfARVm}^=q
zVYjb$&SQjn3_^GYv<<iNpC5oRo~%Xu+g^L)07(X+3pLnU@c~HFX6wc6BSC_t83(Q8
z_F#RQ=GgqX`J3$fwUCU^Dxb{G1fj~5{paOh{h!-@baK@TQGBn{VVXO~zq&z7-O~^~
zPu}stH!4uEnKKlcR9f3liODO_MJdlxEx5<A-G-dB85$LZ0t{Rrf#kf3w!JOD81;VA
zPrvJzL!;aC<Azia;>avkzo~7)ftZ}fth*WC44JOPtB){EWZw4?`BRMRglW2kJB-?g
zLKV%qC(B(Hq!_T|LwPOkMOa;Kq!2hj^^cFIF$6`2yEDb10jgNP3wuxw2>Si|;?o!p
z7X3b|#6r!vt7T+6xjioLv{pA}Cl5=XV2R&BQIEg+v%6EVy^!k>yXL6!T^Psv_NIv1
z9J{*J^^rl7)GNZqJF?Nv!MMG9J-tS`DxY>t1;K5KBpj_074B^)p}VI|PfUKzcfx)8
zNu$7{WDYcwn!X_RsM-|QS+q@;<MSQR@J}bUZ{bATZk*J<o&(M?&0w%kGV>Vk${_#q
z6!~JXnG;&L#+}QM6tR?}W^-#y9j4gATAAYH42Lfj87V7c{hflc%g?1^dZj|C9o}6^
zC>D1GtLP)a&F85mAAO5d1u_sS4qvyv4_uLdKP6IQY9k&*R{U<D(8B?|TtKSi0i;Op
z#|VMQe<eaStr4VFwtxV$w?OgH^rBJAy<jGF$^Qyl<oxpml3yIx3_Z#uO^SyH>aY`G
z-7KnpnV9t(eJ_auhCu!eKl@E_@fqi)hU$`Vhf-0MkC{K(y|sr&<Ii~LD}@wTC-L9l
zTxME#m49d#Fd&6A{qn~+6k}nVPk6~bwya?)TF)OKG`Fh`bHF+89vJgKC^Hc%*P`72
z%qPUektD8s+S}lZkc7EmyzgAAQ@t&=IR{iaaF=Z8t16Pahy<raKkH#x#{2OwgzD<*
zeXyRS-OqhwdU2&`ACmKUCc`e7dpaN3$Kue#6m!I}s_~EQoK|q5iSK&fU7M<HCHV-?
z%*q!&IhYSMT5?`XbbNtrHOQK&E?{Zv%jhwhy)?4{D>i;m0$x#}Q-#-!Joa7r0XaWo
zn{GdkyPXm?sS?~^imfMSOK_mw6Iil0?LRvPe8bS!CM8|(9v2?7a9ah(U5pQlLqVwv
zovinTs6OVg`Lts|yVyaSl3nJe6)^(jcT}hbi9Oh1Shz=KneTg3L`^GmoiGwY0cBj@
z*cY(xIyiF3_euUb5-`Eo^I2SoZy7pzV73JKm3YA?GK#$ZlR`Xik5CuFq^a)Low6cL
z(u7KIoJQ%|cj1RmWc%HutWutd8>XkXqc4AEMZ2*jglITk<Gr$-pbUlGC^TTpA$>`n
z-Qt+rr&|sp>0%fsaU~xi<Bm&5RRFuq3v#qKcl+1D00EtYXTYF&i3hXT0SiIIcwVf#
z%kf5v?&4O8=P-vJviN9<!`IJsYOvg{2t6Ba6Z%+noBL3?oRKUKgHgTbIce2hf}?O7
z<KG`FhZ@y`7}5s-Ttm8EgWN3yi;m^+D$0`z*qM#sb}E^}d}zp#pk3{=4o+u^Om_zA
z$Duxl$*>~h+C%&yzcz_C!-bSxtBbth-3^^z)~VyM^$xB~&~A$Gdd7>W<FyNujx5lT
z;?t9%dgL-&GuXo%^^c0e84;V6P$}RE8X|A<UK;apzrIGM=Z3tJ4e=0!b|p4?It?{~
zQHxaK1`t)W0%rntP0TglB{2y)>0nQJ&iaw#BGTTk>?VQ4*K9<UT;%-Ak;+ci&8xB+
z7B|$|$0yGuPj|dzR_~woewJ}CvEz8R_$r9@AS*qc(h;2lb~r!6F!ZK9UV>ws4g#Ux
z=hN2_Dgj=~>)7wt?cl~Wk&uc^mLkG4uJ}Ehl!pjcZS0YP<fBgz*uL~-pzt7UtE1&e
zXpZ<8kj$Q7t6iU+c~k@(wNKWBpas|rS%{-+V?(!XEK(N_hT5lOw*?L+mT0rPpIR7O
z_=~II@P65xOGPiOA5K{e(NE`lQNoqJlkE9aJEe<gK+;Cu!k>i71xKP{oszhvv-M!_
z;In?l?7TBFv&tuz{YTAbg(x`)>$51vjeMgmTc7QJX%DWimM+QNy>8!cQ*dxVwV0|<
zIA|@`o)|E^n}TZ&3&9rDQru8pqi(^pd|&0*+!OlD-;;i9?vjsPVFG4U8Q@oM64^<%
zacg)qT+V|aX6cfpo0cP$pEJeLj@a{}BH%s2Ph6UtXPsS4s0*f^!8T-*g-=B^RAle+
zQobkq?EmWQV3Qhw*}Is$`eK5dueh-cEcB*EXndEXPeB&PPuWY5#X*D*H;*TtgwqL(
z8_9v5uV2DbH+0fY<+0!{^a;yw8FI`QGPn#GUy2+_B6XB4?D$_Y<SXDpze;}tWXJ)3
zPzLfNux?s7We`pQ1tA;Hls(>pkc~7mkGD)^aSY#X`Tq>rcn8n)NcaxV^mqz4%ZK{?
z=H!Q)<wL(oz}3<yAK+@~mjoYiFugQNC5Qi|mfnVkgM=RgwKUsfc!mhat1AW{c&3LU
zv<pBKKSdvY1u}fUw1;0)2R}r~T&`_<;llt7aRXoo+i$nsGQRC_)B7%l4~@%k)B7G-
zEV!tfG)oLF>V8!w_JrU~AD-;pU!v|3xTnA3FkI9f17|j>C3m7Jpn%!n@22JT>Ux66
zjRs(d$NoHM-r~7hk2aG3iXQp*+Z&yA1J2#Ro1c*%F3vVdL4k|2C5KR&mCd;2>2LlL
zXLrGXH$T7Z9nhW|!jn#x@yhkV3;vXMh2CcxFdLX7G$er8VERo%f}U^U_uU;Z8w-rV
z)c^DF{=fO~<~${S2I`aI;o(Ifjp{<7CH~@-f23^OE@aa)K7NdF%k)};2G8%_?eb@V
z*oX-FITee~qEb>`Vl4M23u%Rhz{0l;420sqUaXU3hTV`Y88N8myK6WxVbji|A{4_6
zwvRk14`0*TDf}xn<G(UAI?zHTECaDK@cxb5NO~lSRh_y2&HCSSI7se~QyzZ%hPTA1
z!lGk0Bvk}bi{<DaD<xq%&e&_N2<wv2iJp@#8AZHFfAKK=9nBTDCu5w<rD<^3t6)?r
zr+X&-30koknuU7%yR~~7v%kDV5Z*OjPP|qCvo_41F;COE;IY952aD~IXLbgGqCi({
z08Y!1v0z_?0_I_k;J;X>3;p&jExAkOm;(CS*o$^aJw;a|fmR;G`S|cy9pgvqxbeYX
zPV?UxViXX%DdCCafZhTQP)Pb`pO?EYp8t0l5&xrV?LYhXl>zky@WkJ$R`$(5-fP*^
zKg!5C>A$~c{F47HBO|x}sFL5|Ssf-c*Z)zqC?45d|NXuGaQBb0r2>y4dWHl0k1~Q4
z#s~X*l`MA$W%|GC2s1q6gA@(&A4S<mh63{Uvh^Fze*Je<CV;bFn4Zc1UM0gN=>J(q
zcK%r=(c#%3C@lm3Ubebv@j65PR+WF8{<Cg*frO?fhzq9w{t;Nl0f|t5E0d~M|NIL4
zKRv#tz@UzQc>(;-qx^sWC?EL5kT!DB)6+9O$f@-hP{S2&h&()^aeYfIiM0(m@0~z>
zko`W?$qxV2+;x2%3Gc(`X%+og8*sn^n#Bp_*o#TiWaKd!wOLds?gI&WR*Xlan5}_`
zs7rY-y|vh7gwUTx6u;ve%fgkxUGeE)fI$})g@UE8f^5607Q=3BLMI+ggZKOiXwJzc
zHMm{13}?u*?s-%(adP^=)eHd2V0D)fJe@E~ZwyxycmVgPeo{(Z*lT%uO>?QAFCgFE
zj&tjRu_et#^R4?5e_Q@_BkqoTu)yV8S6ACHrOXe-E^+~^H%20F${PFnJ!KSmWG3^l
zJE~}?@yMPlFSc9D%V1aRwG&J1Bv(26|F+T*Tvo}#WZ}v40LN6CGAOd=-K6D%)Ck)*
z63%z=G&|p@?lCh*In{L$D4|nM){@IXoUrAHIsJL!G&bXZU#Q45n%w(dv4<8K^UkY%
zR7h-Iv*wu6dq}%rOhW!OH1}Bw=x4u?Ef|y}N7AI?U|NQoiYkd*NQPg5L8cEFFq;4P
zB!ww{_j9wvs1pmGbO}$f9MV@*#OmaNT|2E0d6c`|_i=*yR#?qpfhZIYfRRW%wS5#M
z4$hzk`}584aR=XOj8R~Ta}?3QDQje;`1({~%62KX*4#Dwy(<YL0=(g2+LdbznhZBa
znffez(B3O~<uE?{bp*o9N!;T;gG3nbZqlfrd^^SY8PPuqfgDr8xIx}I*&K1pr8q&X
z+#RUn2diyO<&`J9M?m(t6F?cZi_q2x%jeE*yIKv<@^$+H-AKV%hGL((%1p)riaAoV
zBqCFr#%arhFV&8Agq$&0j2v}no4UTTmU4U}bq`ZJEYwbpHS4A+J#=_g9M{1hiDsWQ
z=^4I1sh%F{cFJa-$zEGJ&2JhfH44>Y;E3`C9J|Ho@rHg1!a?t+M&2fInr$@4oE02=
z{^VP(OZFxvCKg0}TOe8-9|emdD|VIfsf>?IfvK3-{dwvLILjiQ8v3$=U_4)&_{Ymw
zgvx1Xz#iS|!G}g;eHe<=ue*m~ZZ2#FS&`ZOzkbo?n_A`5YrYrTpgyH;x7A1Dp`NgD
z`=)4izkHVrF<Y0zo~^gwbIR|=w6sTAx>Gk=Sa)accc{P<tqj9M8TRTbNYF>f{ih6?
z+es`An)A!-?c`61-Op%KO%&?ck_J6g%&NH-W&~;XqR1(^lS<eEiw&}Dj@aRC3Xi4%
zP9LF@v|we&Ro8)i{Es+7tf61#hlf@8fKGLO0CG3dr#?dgrO3De$qaS}PvmHPu4x3a
zlz=ZDQOB!}D)?j?_$GZ?r0*I~?mRGQ<XhB!V!-C9!KO((dg_*#`9th?q~_#mX)jKM
zgdX344U^lRLTx+u&k+3_bI8vZj&2mMG1Km(K;Ls**(-(2^Sahs!t8{Xi~;(>UqE!A
z4@E}N;c*uVa0dyJYD0N`whF_Ut(ItP+x|t;U#XO*Nlm{>AO$T|kRI9+;zu5K)=lU9
z)gD2!f7DA`D7^b#rm?0mz>HZ6*|@Ps<#nygdaBP`(9ZRa&5f6s1zz`qalwXY!>N+c
zuKNx<zmY%wF8gwGb+#oQ6%p}=wQ}Iu&J-#kAz{rYLA&*Pz0sULaUu^l+rn?W9j_l=
zq<bS^?N{M;(Gp7(cYGF8-ulwn7K2>%b=|!<W;xWnE-!@)Nf)#mk(EBc9?>V;q0=Wi
zLG7sC;MICtBYQepc<i5kRBm`L98svBdzGC9gl?&0>o#QNC*A}R;!Bw&*Dh8;{k7vc
zq>9lp)Yi|w>siy%kL0`aVZ;xs95g~wwwE$28ZHFmetB8re!St0g16PER*%+M(5`{e
zJfa*?{py>bmfF{h9Yn(&C_2X@))!|<93ogfFr%&aD`c4l@O_teKTZ#=&op`*>XDkG
zkI>KG#lR`LDMvk*Z*cH}>bV)t@!lLnKuXS`$VQ+Sv~sZJqu9P^m73g#mA~>?X!U8`
z22rS;J~*7G<dcmd54Uyjbwjh%KW;<5X;&CF{-W6!`IYvxV9I}c@*xOkxi`+CZgIgG
zVWop=(XXr~X>Z$AT(2auw9TYA)W;YsP-lZ6gX{f^%6EGX!@yh1Y(*1iwnhqLlPUky
z2%olo&%S2E@-foTo#Dj^#i6PqfBVo#G0+f7j)ztGg(a~WT(#?Jlp(m_D1IKkd4EtP
z)=SR)vWM?(*`skRQ>MBVRfd7Oj#bLvpUn?DUmvI*YPX9pwTL=i>aBNjM=e){F=u`B
zkZU|wtG)fGEt`?dR<2gQ-%zj8ALPcpvW!LkR@VF1OL*6Bn#;W%GL<f%nLXfRS3H04
z6iPH;NH^79=7a`ZlZUlT6En<=CS<<1z3?G?&c<om<;DpOy&MPaUM$46FG1^aAmya`
zx#cw9Mx>_R(9c#1=v?uz^IPSqe&M5^&*G|zkmH5{qLizyhP(s*Y1o7!YXat2x17b(
z{6~eFyfy}n3)W?K>XrIV$&(r7YKN!dvgn$xztVh5W{j*mqh?-Pg{f?Gj$P8oe|^Y>
zoZ*g)PaO_LISoTMeHkeZ$&dVg2=!1olr?76l$ms^`EXhvOF0nM5{v5e7@G+cVUP=w
zu#+lWG!rV6mBIT+YDcrul!$P>%=wkib<3}eubEWeSSl=W@rc0(3yPwfo;}EqSP=1D
zMpC9g_eTl=898G9i!!GL3T<^X)9vUIj*q#_tYr(*+9i@)s9wo3TIEjuCW;uCwJF=0
z7^uR(0jpDB=JacL_&enGTmlTw(+TYI<n8UcLPRdj{B|&Df+z(w-hkAeHiy@*ME%a*
z>{Sa0E~8>mgiaW~#EJ&WSXm3jY6M6+InSsR9H?OQ2`V8iu8#U~qG$BJK)6d(8MObo
zr#QE@Td&SYt>fq~mR}>Ugi23{di?sV=%`WHA9qfyn7po5$8ei+2@b*BdP;c*=DOxZ
zRpKmK*HHwswNu4x7Eicp6AH5jhurgnlizaCpm`NcoP?Mp+Fq7I9$B=tS~6*!CjF>u
z$zIy=FG+-;U0V>VgnS92Fg~#uFo<?L$!NHy9VwZ79j8}(lLnM711cXr<g_dpXsbw3
zD`8dB7xNGPDEvwRrFwLIgqOQ^c+%!~nn-eOM0eEtH2B*ca9+Ij(EZUh@GC$UM#0FT
z0^I)WYd>2x7Bq_c6A;^8EQNt8yGul5F(t{I-j_Ws*aT}2!BM*dot2KMQFF9+H7N4a
zVgAJZD-op9!6D1;F~+P>=3}G~VsvDlxG%|!qjbB~bSUV4rL6>N9&a-Poj>(z!1@fl
zNr(pAwL;p}Yb-LIi*vC{nK2>4B^Q3^rfWZ+4MbR@W1v90^6$%GRpT*MFsZh@Xa10A
zHT?a-l<;^1es{(E@(;DtZNc3tiHo<`yCe3>b1wTbUR`O?Nh?#FJS+H#Z8>;3H4?#m
zYcnpacT}$-c>O)c*W5^>ci$um4nB-N{CuhB!LTb6R;R<fgm6o3GMkm!5W$V+#c^|t
zsDQ|Er|PQ4cy=n5@bcPX!yXL*Qn?wydx?ehY(y#nWm#HzD)(|Wglk<GQJU4LH#{bW
zsTkQla<ZGY(6eP9+-(M-EQw-g27L%n>-Plm?CQB5NGY0toe_hBlix$;)6yQ}@qiuq
zvi>jWOR<UlE58e`QKhwr@lg3gkbLT#h7fjz=bIGUcN3H7DE_?Hxqsbjoo^+Dn;pz{
zyX?G|Dq}PRHOKNh$ccngmFVOAd$Zb3n3IOug%UXD)vp*S>buMw$$YeCh3TeotYp36
zH=2Q#>Rk)};>N}XI-}_bld&{dBi!@dw3>^@)3yW5$8uvgk*u05&wf|^Y9(1ab|GCZ
z*=pAlsXY(+q%-A3{wTt#SOj&h_fRX+jh=7WS@lILY8~~&gPc3i!jfn`Aw!5QXvT~F
zzP;rP4Z`4{IH5yI9jponbas9)8-2Hjy}+_I!ygmwmpojDB{skIBJuc#T%_AgHTfcY
zcj%59X9a!R%1aAD0l$wnS=MoJyjC15Yiq1Yl33LUSTp3pro}ZJn`(k@OLlWl<cXCN
zk@9ujQLw4<!lx)pbR6m@5;M)@hc=(tTGh%5+ES>FUJh1T&GB*MI}FLq>bITLlYV?7
z7p7v>IQaJc*G>0s+VMf;dT-Gm=0$A`vgQVIw^diJ%;q@gl-<2UI%q1FLwLmkrJo|S
zqIKK2(WOdFeY;DW36Wqt{+aRWxe~pbv;C*1nl;!(p)^z3dKO}yTUx;CzQDSp*aGgV
zct!#3;+J{;U^$f4pn9ibitnbyt5p$k>MkbS1!%<u3C+6-O)I%(LMO!@1%EYvxPD%)
zp6Zcin<xEMBDl6MdrsOeldqs+7yr*ob&4)o*lH(Zvyn#Mn`&i##^N}%9SfE$lUINJ
zjY+|5HJv3+j_o9`KD1>2$hf_t(Q|py03977XsouzerHS<Lj0?`W$kfxsk8GR|Lrdy
zb2QV=Ci!g~A}=G`)`!0ZKpzq1GZ{U!D>7w%ET#n3YXb!IT;WZLQ4+z2xbB5tbzt=q
zxoeAQw&jSv=K9vu2RTIG9iy5pmG7}|e|#q5zt*_en~Ronm*0=|386vD<X8ts<h^e=
z0mXeK9kw{0FRY^VlXqVtj;5e?l@G)c?bL6V$hvCXa5NJIlpq-Ig|DpbguklkAXW1o
zd+!&x6KJUjI(pO$^s5ZkRXsq5FxoRGJRy>cuIuwN+n+sL-zd@fD(&f1qHkfy0W!4V
z@*x)UsjOFYa5^?!I$9GCTes+@k`_cah!t2e+Ox0}+8%)oXcI6XZ^|fPW_@BkFB-hd
zyU=j8i+HO3{jk`g=N_AGP1)h>4hD}nm9X(M4NEALt4P<)P`08FoO|BECE}fDz(w+2
z((RfQS`Q3(mU)4r7lFE*$Pq<W<NQtQa|b+@q(I%`73{3@y=Hx<tvi}jD#7xj<)|Um
zgmfDWhnYNnQNv*aJ?g^iF$iz&W3L{Xy|FtFe|;X7w#PhXXb+=c$Pdc5nAY{CLA}~c
z9I1A_fS!Yn)Z4x|?jpxMy{@jVdpwDl;1kt;Gtn=nCu*qmqv&SyX4Ft=^#!@%duLZ0
z2~+J{<Fk6tD@VUyob(mKQPi|q?lG?`ub<LY*O)r+X`_EEaTuncDu{3#x7O=uH<cq4
ztS2Y+cjFXWB-X~t9j}6LCVl@wWza(jO;pD%SP3Xs&o>{ZdM7prCM0d;z&Y_Wj)O5M
zbW1#4CIuG&T*Kg=4{VTbiclTWky+1pT=oWAD1l5{Ttf36Mu-F_4<SvAn7Kb=k`OcM
zHCK+~enQPm>lH8^&iIzGcRse(v(i?T6CA2(G52_(qUDX#8{zkYM5ses;L}-;3l-KL
zOat$-h;t0tZz~+J+?U1A>$%!zWD}i{%jG%s_H%xS6Zg4H7BDSbi~V2hy>(PpT^sJJ
zA}t{yDlH%lN=So%64D@D(%n2X57J6E($d}C-Cfe%-OU4M@_ygm<Nacsz5h95oO8zh
zo8fS+XRS5oyzl$^T^BXWi<cW=1{eD+I!(NM%LGpKa?-=k8KE|b5s<5PZbp-DVtl7|
z_U=kWo^6a31PX;`3T@GJT7+SPq<=2U3t%+poPm(-F&XwRprkkOUE5BI0=r##r6Aag
zGQq(~IoY6O9vH)m#|(GOfzBU`UdnFmeoe0rxwG;bST4`glzR#1fFb{pu`uQh+%rI;
zta}?+Zf%3~rp>mUGmbxB4EPj*bSX6D4RU2f0Bq@J{8}UI&C(TV`+W@KQdFOcDl1tA
zRLc2T&OM}+G<pyGCm%+x&;B5g?wS1J^}TheR|-D)5nqBe-+*gvQHPQ?g5}e>1lk@t
zhDDcRK*V(NZn!n~H}6%Kyq~Ut!0_0S^!ZgM)0Sm?$RlLP_G9fZhp}N2TwDS2l}DdS
zm>$v4pl@*5MozB2%z_<;hfE(~G4#_43|&s;E6yL{|8`m7gvZ}jX<b`jPRLFF?d(c+
z+uhghQ7nY`5ER)Y6=?k}HSi&4Pr&~Ly)G8sXap`gmc$bYWto;1NjGnsK^{d>U3*Ui
zG$2!aVQEY21t1xmx8jSLJZ}yjGC_bus|NRd#mQ4}0?k(?*|_elE?NHCdAKn1+hIVQ
zp&$M6Op6mE!}rzF!P3^Z@oe)Md8=snkJjR#CF)6O*s34%Yl%I^B=64ZmAW2Y7i!uH
zcdoXkKcP*xS0_sy_!c<0T5(ci&|r37H5yv)I?{>H8x`3ydCXhzB0Es;*xe~tX_Wi|
z>2~6{<kkF%7@fQ4AF0wLKh#N?JHAfCsCqTRaDawvwB_SCCkcjvd1JJ7D;8Z__OI~S
zS|69UK9msBH^o1Fk$KgBen2vRfq~x)q*yecB;!7yO^!W*mv9BI&g<uU&R}V+S|o0Y
zqE~GJ9&fnFrVGyB{QZ_dIE&m+zxmMnCcYqW03GUcnuAO<1FSPOW<BDi7fj){2H6@>
zvlo}=EQ|<K#p+V3Jf5@|>gjpXu3F*I)*(T9J7|}6xzmU_UjaWkoMrvtbG*Y>ovg-9
zK@rnQ5&nzij4xcPrh9_Mz6BxBv$K%t204<GL(gqDQ1HBFHGXuCwu@@aqEE@y3VK>8
zjkB|Ag19Z^<PSr2HK(m~pR3s~I2D@p$u<h_=A<CMUPQaQA`H5%YV=DPUtu_2ILcu7
zgO|-PlCsV%>FZ_7*M9ov<-G6T&~Nz6%Np((i=r43DR3`kdK=N@A9_>AI^j*9%9(yx
zDDbRU0H2r`8fa~mk><G@^s@C$SPVtrnivbnDLBLfJw@6O#}xV`6-UyP7Za##B$P|~
zz`!<61XC(1DH+dS?9T?>y`!f-e#Sm=34;-KLPN`Xf)Pq9Pl>;VlR{6`p{<0YgI_T*
z>BB37zE}9u_qAqSXEuKodfx61BRMtBF&xAGEB>>_;|SdKPhtt2kT5VUpfj7Vi=ow|
z1FzlDN~c&Sk#k-_zjmVsER2|c&|I%Z<5f`B^W$Z_B===U#!^mN!4YIie%J)wO$ign
zc{e|-HGp2w*4CvtSNa?mf9Z#pPB@(=CE)^r@hr=;5vPk2NCY?0Vd+z>83@UZUl3;Z
z*PyheOrcqI_{md}M$bb&+h`O&F5{x=+Q<=0FKxhB#LA)2Fca2*b(Zzb^aYE+z)B1H
z<0lmoTDFKjlV8hxE9ae1K?v!!Qpr9m*Bmov><gyLlrD5cKPbus?lK7}(oO#5!9fVu
zki*2UG=0jC^~x5HB(+#j*Z2y3Z)I9jMoGsVcLsjhTdpc5EH1l75|IWslNMWHXK`G9
z5UH)<BTvidNON$0_4QsAn85&i!TULx&_~eCGm|e#OM6P>yj#4vFKNn8**ODP(9ex+
zdX_&)j2ndm1T~I3)Tm+yq|lnoJhuB?QNtvMl1jHnxE9-<rVlA$-s&oy?vBQ)9)3oR
zjc8jBD!p2fh%IqkmH0@bYtx0X>)mzd8c_Ko7Zock$ONIhQ_2%`Hcc{WC7xmNaY2w>
zE1w)6)O+h>LndBkSusm(Fgbn1G{c;w>H!h2Rc04NE{FWzQlq2V!UfHd-;(&M;b(%>
zoKIXf5SlycPdshtafC0^B!lKl+dDueJ7W{Dwn$bonD*}bp(?igB3#3ZWLMFm=ZhX}
zQf>Cyg3fzYPDMVfkD-Sb_6VipNNz0b9%5%3(igBG>I9Z^Jz{$ILw8p{8qq69Owelu
zSs<A6_1J0Luxz~Ud`aJ$QO)vG1Cb-l=D3`Zdt-Tvr50~XU@x)q6`kSL1dF=elpevH
zga_Ek(Gp5qw$NcZD%{>-?G~aAN`cp-80P$ncQ=LZxU5-HyzK3tcB*bRGm9AV(^I08
zbZU!Mx0huru2n@ZPE=|~U0PY*-P(o@OLo;>C`evdrgInk#-Mqs*I3ma@6>}#l-ClE
z-z%MURm|Y9>AwJEjKLXsh9*z<7e|4JckXTn<cvg2P9qp~1iTsnU7NHA*&LwY!Vc30
zfdfp!ijEs*^_F5*I2^*WA%^d$V<fw5pb~E&^q+#HTW&if$8McbIQ)o>e)#N^vT3hr
zuYv>rB<y?Ej(_zvi_8C}7qtP7GwY_W_HqE5sGk=!3c3LkrJx&7_4Wh&VnaMQs9U2U
zCt7-=iHDG8q1QVMn-CCf25g|?lX#ah(F@8R&%3c~#2Z}Hiu*-Gszi~+zNfm~jlfIT
z52(mzuNxM{U;=k*`sI7db6qm{5nx+n%Fhofohr~1>_onQnX+Q0kEp+8%AeJSUKTc)
zX(Rb3gjDbpb@5$$o*H_)Ny$zE69M67%VUHJbW!q%{M%R)Vj2A<W-0$4rG0X^gx1y8
zRSR7EOGNFw>r&6>@ftsf4Xhif;a+Q!J+j2i0-~$<xfHWv?*$=nZ=d!m!Sb7~3C*cD
zpi~vdx`w9fNGG5m2_&~FzXI4s+6Ud)aJAJ-o&MAa<y2{w3yRFcU&n~<Nk++EU!9)U
zGTy9<$xFIf0O0~<x>}vX`-w{+dFrbW0R=L@5PJ1{N#M;M#%Z_PwUriQ*0r@;#`m+I
zB6vS0DJrU?TWeG^0CkZ>p21UN_hT6YEvNf7RIp*NennrMcht+$I}NgkY5saqFRKTg
zW7z}osh%BguruXkqp-D2ofiX5EzN18K1tD*0k|QFB`t^s0#9w^meXKC2|5QSw+dPK
zE7LEZr|dPi#wQe=K<P9m`uQyQHU9Weuu0qYKt|*0tS##Tm8}{5lCcvbs7t5vrOsLB
zI81N3(G&{n9!eS~mz2}IJUVln=fb!k3&G4E2nAm20?EdTSM7CQ0+%A^-Q@}^mp;$U
z=0ssqTg6v@-beQOLp>M3Fen!9<2&C!C)!3H+GtajO+t1}<He(oU(PCjrpLC&8lK5+
zFZ%rgI~Ozp#_3>&$bcVF;=afAr}|-#ivlZD-W&2cqBYCgTRh9Jbh?38g>j$NYgFg1
zrVI?_6q4S%b6xS2PYJZwm=;>ivau3*9QePPeB!f9J7rF6s-<r#!g2p3i_KYvuKN$f
zkV3W00i%(<cwjXq#Lh@cwPaWKoRE*pTgTA~&%G`g#!>Lv)52RNm1uDjiUoX`kzU9G
zPJ<z_76inZiGk=d+N_I6y%R_(>hV}hEHG&geC=48X;rsFOuzlr#Pg14fA;XGleiW9
zX%XP>-@z;JXk=9yczAD)q@&vI{GqR@?J>$scuV>Dg0W#4<?a?GVicsp_qdpAT$(jn
zPfbPn#G<0^h`SJtFx1Kp=yNeRnPWxJ#kU@@s$BUSAfC0)q(~MVD<I`b4Q@@qwQyBX
zTT>=EqK|1FHh8R`OVm}mDeU-fU_3-dg%aLSoPSp2w>c4-gr0Fj6xjDQKLdsJpo!B|
z5AHN7(R4gbB=kIaBDj)b;lS+|?!D>69b3Jjd%m?z%oLTJzHce}l($^sYDtagAwFNN
z9;3rCDLS^Anp*MP%$1Qs&nuH$l?esi$?jIWxOb+EWj%`sXm~4*xXbGIhF)s7{iXE7
z<nabUH~2uf%Sr!PYFZG3r)}#4mz9DxV)$k;dq&)p&kEuuQWxR-5R1{tA9O@}`{X$N
z@fp-?vf8f2m2(@!$YiTZ8!a1@CzHiS)f!0{SO<Ga@~%x1@MGs*noV`jwl}g5&Jxwr
zGx8$xuZ22aSx`qptCR|Z+=pQbn?o464z2B5G<@tyQb{&~jFNVs%fj9M$KrPP>s#5c
zX;+l+_BcO$f15uVyD*OqC9tl1&KfO@Ud<t7zRC`|vDN7HKxSb6<4<IAh2mK~Lu%{O
zdg|arj@1u#nkA~`WbLE|nWi?bjFPH9aaL!%+Z65`_TBcre(EJ{+|J3}{ZW*lXjZuG
z%;PTKLCUClqJDZxbH>d=5K5<fT%iGbl75pv#BQJ1ibx35bd~|0f4GllXDsaKCPcn3
zT)v9@oS;y$N+^*h#_OP%z#8HVf&1NHUn{*DQX806&28P)|K2)Q2W;>0%s~|_fELM2
zLAx41&uL)OwF~(XM{0ORS``-){+%nwu=~uKaOC+)NaKmmTVUrUj_qiFG<!dtEd^u|
z9QgaAzGlQ*S`Xbaje}wN7Shhpg+Bkgc3KL~6>;Si*30~%bWN8Q9WV5&>G2@a7Tlw4
zgz=r>Z;eN#aI*U$JM7#kxeHTWj@(cGafJc@a)m3vS8s{)gcGtYvinK;@7MTJYt*vL
zZi~jrs0J=O3_yD?7c@!;Kh-@G_7l{ovqbkLJa12#kydHvPQ1~~g-kZMav3xliWyF5
zAe7hVfTvD);m_IYcK!JHfatNT5pr%g(^!og-PlX^Poe>-T#b02y%bzqi(U_qO8=9C
z2r`esL>5#~teDI!`q&&c-GIILyh{Tf+=xo>;esmZh6;4a@>+gEa8LN&Og<p<>R|<l
z+6I-62cWCJgs66Za5^=l(xvPKLzT^`M!v@H72urJgl_Y*e|s8Ix1FTMuQEnWwj|MK
zr}+51TDABRJBh`3ENjy!ai06^_t(Zfwh9J^lDCOPukd9DIEc6VohhSquT)UHNp40e
zNN}mmIMX3qSr<8MlnL5PK|ekx<PlZ5e>QBlrkFQ{^B2m6dOuCTHi(RA-(RkWmCELV
z(C*hAf%u(zu4|Es-}lN|hUeY@oFXxE043vUP#|OHs9N-|d54Be#(JY*<A+XrS>ncf
zErs3b?4R-gGLX!9p5R$PMA&)LVubxak%q1<?@>zQWwpB#1P-yqV$hScT7+kzdSvEo
z_()P5D=?(<`%?4TkLP4nyzf|9<y&Of3Oy6&dTbe!3aeir<K^DNqK9-+eJ_gd1m=7f
zU*T!^kZE~wPiL0q$B$1g;m8&W)}1#h>?Fj*P!|(cvE%Cxlz5qCM)Oz5)hiuLS#jx6
zlo{2pXEJ~7OW|PL9;YnW<fpPdm!D%_X?Hwdi*tL?@z(zGk;YqRjH<3b4+KEQ88u^c
zNloR-VbiGC3iT%IQzo9oo*V{Lj#cANWe1>%*wKzgGczLn<;$04&J$3zjiMc8FzIrQ
z`%A#HcVqq(nNmF2pk3v&E_BV_mHxeJ;%vF6hmd$%Z4vD1_<=pAPH@2PNef)e*Z}+d
z3d@T^LACS74>b9_tSS}hAS!1%<V}=3G)g0nl9hq*__#kVtxs~pm_~cJcW`T*Vjo<7
z*?f2G)PyFfZHQF&3@Yn_;2&JVDH(0tL~Wv{P<>pQn`;{hZa>W=vyN?(8Vx~`QCIqa
ze&0|*bF_Eh1hJ&cHgSEnL&2RcSAoj`G8B>{LPNC_<E5cjF;2xH()i*O{tL@993P6X
zRW9n$sAb2j#;mDMI_C_;f<)Fy2&`{&4*qBfd&_m69DL_kiBX?Sb|E?_@^C;LzAsR%
z9I!-{&2@UG)$()zIvizb8=vs`n|y#J?_Z|~kCg#Z#`_bt49g7@8ttWkY^SpyVh^VX
zYPyhz0nbm(`TEK)&+%u(BcVsxUX)43tXUKXjK>5gB!|qoIOb<vOT!hBI0MI-sUK*Q
z;u-K8H0U<`JAV;an^<@gSi8mpdw{q34{XVX#y|s9B9=w8g{Px@SO47f{?68fhc7Ba
z(<WLLL|g6jNUEuJqN8hTs<J6NiP+|#nBb4^>FM7vrvp?yN|(JG$4fAA^!^KSW7Lbr
z_fnS*H;40cVSl<an1Z0E^*@|3VHN<0C%2=DJ_Yj&B-qtlwN0k&Uikz-JEjMLmOUXW
zLT_G4b-d;eHjceNXK8Jvr-<E)Tu&zQu>v?F^Mt0DMD%iE7lErs82cV7IPR#{mRdmy
ze`yhPDMd4i;eR-x3eZ4r;Qb5+pK?Esgz2obktewljV0Z@_^6E)z5bSz>h^RSvzr~0
z2N}0T>vPmtK{WrI5hJxB{{j*t4RX!m?iFafuoPXn`Go}#aA4{8{Yw7r=!%_oN62pZ
zXwsd<N1l#`#>0Mo32z#*Kl@6h%uuVVmq4KkWHh0juuD9%mLbIN_lb`u050U%-}hdj
zgW}!yWPl3;H0leoKvwSyfSIF!W-8a^VJn6f5P106Uis{)67Wn7_7$w&O6rtu6rs;;
zk%?bHZk|_!6s_N*dA@Ev@bW6Nm6Rl>g*((8D3OXbIwyvjbzpkkt)Xwi<Eq|q{}BCx
zve+^{hRWcP7OpBFi+awr&i=>ZEzNtKxVK*R```T*+`LN|dEG%HcJlquS`WAHzHF^6
z-{b^2xnV}F=(7~q4k+!$h>D58l*zQ0p>~()o5ACzR|fIV%Qef*7OCS+CJMa4U=ZoB
zZ8HoD3WpitYxX8_NVQOdr|yg;t2Yb!D-qU_AecB+fty801ksD1ryR>h4Q!ZN{2D;*
zhKWLtmjed2(&wRf!$rtjtRz5|O|mtqA2NBWRN?uB*USFEt1D3*@g8%F<q{dPBM2tz
z0wY$Py}`$H=vaLDUn1+yQE>Wybx@4E?#qkpIUoyTgaP%)Q)T_Dsnrm;#R67<p+=Be
z@#x<e!r{|ftDEOFkr2GOx+)Q5Sm?D<QF&eDt<hA$@bm+We7kH162v40G*62U(R-jx
z+)X~X;!wQsPyBT}kCAibXTRik@^dj$Tb57f`}#p^cHdGL9^vA~a!*SB0l>pN^?2W2
zz)P<?Y69#0<<dDBteE>y>1Yn`P^6-4N<@1`IkaQL;<uHYaNGULPmx(+!>uIO3rfCf
zsBpF;z?lLA^mM*s&eS8Cy+nN5_hg#u?uZw-yY!dTgfTqmmTR9hR)mn?s4)D*q4B&8
z75wc23ufxJ3P7`*du5Y)e@%#5N<t!XTeV!wN5wH=>wCyIgDd(|n9awBxbW#WG&64a
zVFXg5Z2*XItV*6q50X~W;x+b(gNEBs=mm3IAdE@47+|E*HNRY+Lta{zy+mZZwWeAb
zA;oLMOgO4x-VoGo74pE^ErlPxkL6fBH+ekFw+IMT_z#3aUg|%D|B`@v%)FKZi;*hR
z?0vcJe6Mt$klh)td;2<DjU-1V>lhIi>Gf7WRZ^T0{|cUJ^f4OWLoiiw+9o8Bw30tR
zG?l|?=qb3&P3x7DZc<9$*=(Y}Adhp<L4V5Qbhv@WP9KMR=TN#ZqIbX#nzSUialt~?
zXF92|tiH48tiJO>Czw^bg-HDDtB$P*AFnc!)N-+c(k2M(5Ysf>!uN6HCWtbya6aAd
zWPm+H)Pq#er|I%&B^jt*bEX}g0fM|5;(V#$A#Iqy{ptKbhm64gw{7<fX1DA>|9Fq4
zsE$pW9G9K@YnX*cM1^TXHpa;xnxuY)Wq3Zi@O2y#A7RgS&5yDt*zd*E{`$FT`2X>9
zTTYeeJ=hp;*2$hZJtdoR=)xg->CtFg@rNo4Gx$NXf#+vXkC}iUr1?z(E{--~enTbe
z{VBR;2768~V7<{zc^%ph;z&<hzfftmLgUz`l015^BO^|xKdKx+L+rM2cl`w<Amc98
z7%BI2VB+EeIFw<*^@h7^6Lb2|NWT@^s0Wn@F!R%|iR$p<4&_cc$G1nXrVzyA7ikV(
z>}~7^@0N_j85#r8R4v|{+WVgaco+xTHXTQ?{V*}z`!?(;m*cdRsRQ;8g|}7f!#iFo
z<>Rt0wZXKvlW*`Vk|J~V?>0lXlcUT*(Jj`(Gw@YG!KapwIx#Fh{@LXT^V4iMCAYTb
z-0wE;t)6!GV*U_J%aQi#ADydnw6Pt=n{j`9UUH7%c5`tMJV3?=)NPO4C=+b4)U3NH
z5)-a-Xri{vxMvs>tCSn;SR?MwX+JStz%kfGPj=T-6;NegiH6udxs%kj>6PMUSDY@Z
znAiP*2n+Y~PX<&6Y28lJ<~JqsaIJQPc1-4qWn8SOH-l{%rt}l<JB13vC!HA5;Mcr5
z8-9|^@X;YwxIfsQf=Kr#(+mu@og6EdttNINcn>_)Ffqe?R3XR%dm-VU@4qn7a^6dm
z{|XZ$HR&z(b6ljS#bzNXkK}KC4zb<b|Ao{L#+(^=6`4ts9tz`hQw751cWLLqK3?W#
zd@trRyr1Gy*9y6}-MQq0S^M90!$7wTaEJe&D|`Qsn)iQS&HGP&<^N5>?Ef!6zPAM0
z`nNqHAtArXhX_eO{u@t$L}oc;umLnG8}CU`{_U@eoKL`>j82XG@2M7fWB|2*<WW)n
z=RYXmS_Rf)1$f6I*kCHII$33<qgVbn5aNHZ7O`Pr+7Zum8+{HxVG9Qnp@9Lsi(Us1
z*fjwcGfhV+ybdHxXvgKt^f1`#WF7jz)4|u0PSsiW<ANRW=K76J=p4X0DCCG{h(^z#
z3k#4?#cx|J(<`f1w_8PPS=gN?ypsMEGdO$5s`<ai3^;B78C~&r+5es5zu-#$MfNZ2
zyZ@gs*4Vw@-&wQ&Mp0}JCHx~{<L_(VwdVhlY9OrnPZ&$?p7vi@v;RUFWDDB=BPrwW
zi|@8u``=l!{|0S9HvbEZ1uOg)*6hE0t7+(S{|xl_JCkyc`*+stzcDFe$z1<Q(*2)@
z_y5Gh`wwiQzD^Gkpi2M%*0-OmOZa;RIA7we3JQ24gn+^EIadvnq)6eCo)KL)06bYu
zK~sR~pmdNIH7_))E)!MfDH5NKOMvXk0908k{gxYvcoGrVj9}a`o%0JAza}S|^Md_?
z3Ass-BrBMu=3GVD0iX2d)2&h3q=`xkO{E%JqcvAdyd}$*=LZWIz|$l9c5kK<V|y$c
z?qSxtteC5iPvWO^)#uPwSYqkRkhZLibYJg%M*j-4TNfs#@RocmOJ*wg(@o?`95auE
zJdiXI2cPd4dGlau;GPb88rQyMBDz^PKDGuqwQRey)uqS2EY*8lTWR5oV{%gP%r{nO
z<b?eCgA*r7D}P9%kFAq$CnThcuMEy-o&qHX*4_|gyO6w;4?8>h^!q9Agn;o)(4(Iq
z?lozDStdC!M}vC@9x4A$h*?bKg7$ZZL@VoP-5Uh&q^}Anj${Zm8u?@Xoc#SW2Uh9@
zQ6Gm9_{E~ABv2yx+kGIIDkp?q3I*RmJwyNV3#cN}E*g<XexV6|@uMJOWNMf5Z2%kv
z2!bQ(i)E$V4pGf!ymPNwi&Fg!md=ofRX6;XQkU(^-sgnZ=d@HeyMh^U3r|2zO+}uT
zF4%L0_wmb(<o6Hpwj<p_Q|hArh&SUoBDx3WRQE8&tV<o6Q_AAU?;%-^H}%gb{~RC?
zcnmdyfoKzmC*F4K9|XNCL)FDKParre-!20I?!Q3HNna8lvENjZVd_wcp4dxnhj-6j
zF+C09Sbv$Cmze$&5Gb6h?Tmbg05Z@OG-f|OpaZT-PmJeS5oBxCw#7pp*>h8uDIfK^
z%|CK%MSS$>O|0@Te-6*ORX<DN5B?>}$cT>$zNKn~3e&i*$pe=LYSFnA(q;TZVS4@@
zLY9Iiip&NEe~}rMPi0xiKX4!z&sh<Zn1}qLq7nMV_yb*CeR6l#)B)r;rX*MZ?2vP6
z1O}AS@L-qWw%wZ^mxWd`wZlQTYa(2)Ug1dGvjKQphElnq@Z{81WM^u|8X6klK4l@U
zMRs!`@HyZla3XXz{ujY?G$aukwKOWsH7sn5dCdmBz09h2Qty{c;B2m3DB%ky^7(ze
zFiC@pgY!t+Z%>g5$4`Q(9`w5y#ABE!IH|Psr7~Jf_1`&bM7}0vjK5@$Xt>CbKnf_$
z)MS2S%y%16Na)N>Rra0+QO=5-`NC%&bG|`>jf-QoW=Kgg3mx6@?pG-0^)_?wFQx8Z
zJwk7H#1oN`Sxr>B_f--}I8P^AEM_X`w?EwjIa{&r(hjh9ZUs<@*VP^qLZGIp<B`8f
zvsW>uN-=j7ovr!?K%(r{<OS*!Vm@)%<{tMSuc?irwQ$-0(i;`0^u2vAp*BmttXMe6
zYg3sZAI)fZzXpCjFJR5u3JkE)8DjHqps?%ahvb8DlN&Z%*)iyfwbv}FFTujMz{j4y
z|G>@T`{&W~)0U~BmsS~ee5G<Ui_-dF-G#PP9HB%{a9r!NBk5f7G)(gC?Ng15>ZwQF
z-JBe=l!#l>`g(fJ-gp{@ldrT=gHt~SzGn4<cJ%l4JuQ{7-Th_WaHW6IJQDJf99rCt
zd~k_>4Agdq`w0)XyuBD=05LOj^~*vLweG6FcyFoF2inG8yK0^B)9y=5CKwDyGXj@R
zx#N8%s;pKvZMoHib;iyin?sKuB4JovLEc0Pp|zTBxJ$=W$ZW}Yv$v*jw|`S?krDlK
zq=PxQOGo-T1Gc!_Yh15N^WpKn>0k-3(}szyjM^B?Gy&CoM5s?6p$q+^IOA_PU8IO2
z#&%DN#@`6NdzF!Hw~bJ~?>hM|SoHzp9^Kg3;aD|gZ~&z==~bE>$yQr<>WO9eB&!!K
zNu7@B>5lkS#+h^jG=KkV`D?akJ}Rl59P}xAcfSd~+{RC3N)cZI11gqYzoguQnw^5p
zO-EqtL3{Ge_Y^$+))sG6t^L_*X(c74n75i~nVGG?<7)9iLoJ%v&cB4GnhZ1)A<ieP
z!)TASw~iqKbJxRAit06^5Zi5|-RrW=8gHZ>u@-0h#{)41lgFA=3svZoY$sH}_JGN0
z+6rRVr3uRK%~KRJ-CemO+E)y#L1<M<_;`?&Br8Ti={wXvm7s4@1R@;^oHo5MUb7$p
zGiMA!F1tM2ztJ5(B@bxn>DL^5F^JJi4NCb5i9IaudtbZtH}%lHg0aM?y?G(-A`Sv5
zglXd6%cH=`6vBFX%|sbgs|9-YN96%-p}QYFbwofN@7NT@J=|y2XeLcXuBYkd{UmiZ
z8J6D1kdMpKR{ZT4JLMg~wY)Xcs($72X_Lr*MVe^05CRkhOQFDgmX>d^iK|}(m&^CB
z$IZ2}X<1z(WP&>rU#4ITTy$iAd}8Xr+?uZ2icC&=CZitY<HVX-01@md(2yzRyq@v?
zs7;uYLGUM3Ic$kVPWn<lii;p<J4v6fX}T(`67P%bWUPen@kv_P#U0t&?5q?G%)w}l
ztgZ&=MAxo$?Km_?5lR`%Xx9Q?)8N*hn&M#4n~L$lKEA*<B&fRM$b$i482S37)rY<p
zbes{Ufjn7v7cmBjOWOhGO)m_;n9kQ3ExBE-;r#^DnC_2sAbZ22=RgSIpLKVQOpJ(f
zjL!Q_TZ6Ohys0c4z#b?y;6CQ3E{iPDN=yMRzsojDTe9s7ZlcQ3Wilh(q`}KT6j81J
znb0~z7+#8pT;6m}d*F8{nXD|`+Ory;QU2Xq8w53K(a;|V(Kl(oIc*`eq?PTa(T`Ol
zbyE$oZqj-o;#3E~prn*SjNC^0vRjWH%pWVz-_M~OEKnMh63-Bjl-xf$at-*LnBMM*
z?x;|SnpPJon3jT<6xyRe7Y$xJ+LC1+j#U&9wR*?ZP~j(~>UlqE2(@QE)T;2&YyE{|
zbnz)#oXdaIw&^KXVlgr`QOX7lu6Tm;H5ul{nNQXIfBer9@uJbIHMI6drYe5Q3dPA)
z3z6j-IM=3l9U~?WDWIy{HS(ne*kw`|QOEL-^+=*~1cs=NlR+AQy2gsIk~28c+4k3|
z&{m3Oe;$(6FG?g|>zzOOYLp&<*MJwLFP5sVFlnnm9*wXEN!&A0deNjvx2sG-D95Sj
z>C;+mj2@Tpbn~7X!VIJJ42_X|bfE2a@nZ6nc?&4}R!s4Pp=7*hB2SkZ0|>dIa+$Yp
z0fT`kvjCD&C13o5!I-bF5)9J=Bcn4EOW@h2v{mP9A<v#*4M#Z8GuGP13T7f^*@u0T
zHs-szO%yjZy%awsw3`3weny>kt%$dyZqV3%tx#`lu!LWdX3(hnwezV+#tlsFyoU5`
zXemAn#)o&W8=`ar3gao%`pWW)cOGNp`?zB-PZoL7O~@(}&qFh{+TLOba5L+sFBa;5
z=+LNqX@ue3Vtb6r>Jmn;05x9GoXy_-&kjBY(_Tf%d0<ILit<yBw|dv9IPo_y2V~SJ
z`)=`TD)Yc^j_d55Rp5T5fGWMi(QS0(uQDU*$orDwuml2Y>n_?z(k6cqje-cP$M@c-
z8%5X8JuIopg&;*T29$RwidQ3IjBA|f0G>ncE$so0OG+OI2nTm<3rYfY%%FFksd+DW
z4qcoGtHm+uzUMvq>&GsrUVlBHA08e~5%06EcsS#x<<C^jncDCGV6@2V`<rsaob5~)
z2nR+W^_I0z0lj`6<;oBHF0rWkPS!qlZ7kj*O$_XbO(<<yq{j-uQ6K7{^fy`dHBoq4
zj#Z%G$9dBfx0_o-ZHsHhb`gc|N-cwlzV;Ax03n&onu+1dMnK$<$l*e$kn`t^+qvLh
z^o^u~>JVizzhdWXXul@-I?-H2Ee&fJyZF%V`0y)8eT%fbvoSWNF(@^aGPfYgA{)23
zU#s3oT-gqAY6RBR30Se=Q^=jMOh;mA(fkj?{(1xAEA*=y|I3b;KqTQW^)O$r?_kdF
z#8@3Ru>O1fS%ZSR@Wbd2{iS!L@&kop-Gz(td{N>Y!h?J5htBaAj<F;SJsVeyB}AGT
zkgn?4#e6QQztG_7y{4@9><J%TT-3noyj}RXcIRha2x37)u&+Rv;C{dA(2Sm!+15rs
z93<^}uQ6)|Re9s&<OK4?Fffl*Z*V8U_yH4fqTpL5w4U8dwniu2={w@Ak@F;a!!QC<
zFf2{MYwQjthy*$+9Wvi|0&7X?CvO%dF^q5eiS3gZ(`D>2q3p%GA4z`hmG^QCJ?Md@
z6Re2R|J3zmCOUl)NhHi5Ic$}&EN@t>CsxIsZtkEfeE8b>FJU5}uZ|yae^%&bbsl_@
zWySL&sTur`6*J~5+ku)W@jd|P^ZuW}1>?hm?{op5*b+}|THm>WHJa^_OEJ8x6UjD^
zbr{xLlhkW(priZW3hxi=l0ICjMEhaP%lS&*Qgzz`KmEbebaH<72&Llvxmp`S5#TxQ
zy?qOC0Lv1~O!o!1d-Z=@>&l<bMUrQ#zTa<tD5n3O2<<&}$N~f7{YLnD$1x1<!|yj3
z%%Pi=_e=l=E~p4Q(0KHSR7&&{|JVGMQoMKsdS8M)q|8zW+=nUhSz_FG<ba>3dBf|5
z#ZL(<Ch6QY#FCl1YR~R}FGmOZS(E=8GXNRERm3~@d+xtU{lgavdwbj$;eg&3cxhJc
z0$;K3BIZFJl9Z4qW6npcRKN@>;=b<jn;1lruyC}W&pOkvRf-lwcc1pOT4}eY;#^+2
z%8l>g9nmPCzQ&P=YvKn>u^Y+SZ3V%cnSeB<M2Qh?(+h3uaV=C{t;HsH0`*{NY3WD6
zUyG<H9|u?Y5J!4Fi`Ec;4DOnDS;?UKNl4-BHb{g!!>e=q)gk%hnZ-hf<l-$H&7Jch
z-uyLR(}(w0t5G4g$JvN0FLrB`ZN9gqwbt>f%fsIU?#{qXDcUV#W)Dz8tiMf?&sz7=
z$7YI_@C{jBZ)MIztkcyl+@a!=1}B!hQZ=p!!R_ho)%hp4(S$8E{&?o2DQZ1Z%DM9=
z&ebJCH^HUR6&&=eXHE&ogCD2YT-Q{kbQ*D03YpRm>XE&%i51Bw9Ia8(&T54C!!5c<
z8Qt^H4Z1Cn+Nn3<F<8xCqh}|?=tTaaB5y?wH}qUwjVU-T(OI-9*n7N`z=flY9c=sz
zjZYc5DM<R9^XZaw52WVqddEFsl6paC1!z{943b<CK-sG}RBk@`LF+ry<hhZbF1KA=
zzcAu%+5;XH?P@f$NxU{$ldfq`HU{17rlP*RP@sBgPyF@vO4anm+)mm`knii(eJ@*6
zTGt_tY7+To;9L~F5XEH$@(T!nDz|A}oc3u-eW6CZ3jfJUhsC|#>E>|L`k;-xj=!2a
zp=HN#KIUWuPm>+?j3DF}lO9SDBZhGa?M2GDYC3^+#Vd-B<av+8fo2|@A$KeBCMiKB
z38nCYD#14i@sA~U{C+OIvyV8gJGc)%o_xP=sWCQlYb3aM@^3TCcAA8oH{dp1-~kC-
zEy&=R;L@Qi?Y#R`gm+JqsXJP!jQ=p*@&%FX@wF+H6O<{sbiHd>+NR5lU=CFP6Fprm
z7}RxDiyo*|p@anQU%&hIi2L@i)e^W{c1_Xh>^f+;o{bxJR)E3nml~JiZ_Q60ENxwO
zjCR?evQQ#jh*|%5_8*u|at8=x2c8C0vA5)l6LPhW_bNe^Ns`@8ZKv#gtO)<$y|g*L
z%~XFrZSr76z~DL%&+*Q5@?FqAD_BUp6UrX}W|m7aA9&$Dp3(A2RV7_qYz6CdZp>Lh
z?Y;*9^EH_Jsb$8Xxe<bi|8OPj{-D4f`TNft2ETEpnCS_p)}d;0bxBD!&9GF}u9P1E
z)`-bRm_4t1+DdFUIm27uSsq_*6b74=ZSMC|m5j`n;=>5;az2lcmiAOhFFJBR($LFr
z^Biq<41GI_4v{X+@Yo*@TtH8+A!280lN|E`*Qc6*<Jt&Du0MjZipf$QXJ!`h_-TQ4
zJvPY!%y$qN>rb&xx5xd%*KnIHZd+6<KMtHl+l)6;<195+@*l;=sKpXK*;cQ|!~1wV
zwK1iI-J^=T3tXa-6?1#rD&D|Xv?sl(2m7&1ZHJ6Tqvyh%f8t3191?p72^Nv-sm!l;
zS{`qA9!L(;;yV3Uxs)szQ8hY;jfX6kt@mkeq3_taX`<?*lSkksDaVRraK=DMauMM#
zfE=_+uiOXRn}rJ&?1fN*4#pEz&+j{!SUYUr$jnH8=;{?|z7>J*84N%fI`VPdFJRLl
zN-L}VC3fub;H)6?f{W?MUvWNUC&k8j+y(Ns{*lpKp>#M;?Zr&;SZ)_kwP*na^Wh&=
z3x4I8`r$dy!NRvBN7_i7$dVB}c%C=}>nL2FkO<+rR4HJ1X!QX9S^`2dc23PYJz5dd
zS)7)8Q(wQvJJco6-p2cUhs}o4L@jrQ=bBZ0?&JI&PC&j0V7-KO-*9*C?=JY*_sT{s
z!_hwX3?&Oblw>HenrzS;Mx_gIthYAN7#SU1deD2V1aS>0j9Z~d16T$unQ*Z#6YBPa
zbwEcEYb5y-tWIs5tH8u$__<X)W1^Fl!S@K9%$GCk7ploRs4s!KgOI*uu*zD57bc!{
ze|?kTCfP;Gw;IAs`Fg`+g_!G>q8MJ&=`mO4NO080ty&8qJQw3NLB!v8s>E<Xr%r^x
z$fK}Vk1C0^3w~waw`4HNY6{3fv9`re*H7MD6bU|UB>Vz)L08}i{my1x{1+#LCN8j3
zs&jO?9X{b#Blz4*UwPi|yu9tez&qS1;RE5LdqwM6h$Quly}9qrFz#`p`kY_+O_fo9
zbE_GDOn6%(r};Hpy}l6T!Hmth@HW|($m1c6>WN~$3Ot_X?jT9f^%DOj<B!MM$>?uV
zSA|kns2bhw?}qu&&q`;j@X8m+{*VU4v;G^7@Yd0Z&`Y2fx2N$y0m=0y(gmL|Ils}8
zn!J(|_D*6WXBq%6LDU0Uk_OjdlbeQw5oO(zF?fx)1QB)J>zlcMxJVEs<8ruQ@@D#Y
z*`P@G|6D|$bJ=e2`xl46_IIwyF*ny@MlP~;-WdyDrt_&|)}Wx3<y7ERuuwwAaMJUh
zVeug&^1;BQAt>x7P_y05&3D$lqxLO1E7#-gekVArI0CZPRQ%}sx<rX9d<0n;(&$Wv
zn`vCwaZh@cVAfwh2iHN32mLpT`8tl>xmrb1OyNE&dt7F9>37u**k1P@s47z*yhA{E
zZx*ulcTFU#<eJFua&DHHG!^!>=60qS8-oV0GQ{THgzQGIuFqMr=#H=O-JcOpQbzN<
zUO}=1eZOzcL4v?pG33}9cdM9fGa!XoH~oFk8g@OpKyq=OUXm*kyv~6<PHo4cYVr$w
z+!CyuT0wJovQlmkgyQYd`D#8#Dx}MIhB8BKvRT|K)fTwD4GR*!i>`Z*=a(V1O1^md
z+sS6Z&z!!S!b$vaCOIHqQcH_;@x)+IKeg=jV?6s3IJfVFmNM1Ma16_B!~>KQD3@0+
zHpH)4m)F^#PI3w~#oshVcN&`!C?t*KMVV@;ZcaTn74abrt8NK|u<)#mT8QRfOOgHj
z`9tVeHH;|eMer-;js#yM5#%1X=FfR#LII0k0?g)t)&(h;=lo?6oM)9S`C<VLW8L8s
zP;n{P7(^<)jqmVP?RI+ZwHNKU9qW#ZMajH%#}c#rT?sH1vUfVE=>e!x6Wm|P4ZacE
zeUbELq&Rcdan1Y6PAex%yc<4ufkiE=qFp^;4CM|;2VI7!tGS2D#5|I44F{MHXL|zt
z*7D?iY)ODA-sm!Q1RYeY$IunUS6;u<hP-U!w*V&M%s2I^2prL!DrXl9hi5E1A><XP
z{tR>gvdSJy7DVi?z?W)x@M*VooD)C9ZyfR=<{<RrDaYQEA%8%Zs4H)y1K%vu9estr
zN%wT8tN$`8?(n4O+Ld`jPMd5!qfSE~khkrl-e=Q%ZCig49J5Y8^K(NdBZ!PZM#=~<
z`X%Sy3f~(a&~iK-78Wyo%q?tCEANPqbnOZ>uS9t>%(?Jc2WL+^Q0uj4G-d4$24c~S
zj!a5jsP!>X67CnN^b%rHA4Gup*TCTrm0_;$l9Fo2FR16jof5G`v%SF0_G(t-FEl>T
zpo%p0G&~*D;!y-b4lN9vO8f|5WvMPOrDZ5MX}Xvw))D@>SNFz%IdJNOwydly79}cI
zei+t{+mHltS~%FHhWLXZO}G=gWdnzIuACICZ=>^XRpiV8gMrV?DI)&>;JX`6D0P$C
zqf>B#n4t7)SV}mNEmDhH=uW^M0q4B2+U|TQeDVlz)FCTQ$YNJE^%uReD(KW-8*^TK
zz7wM3O<Q3R&cv0}$(j*EI1n`{8sYS-N(ij8y$W<2+bV}+f)zUe+%xO4edUKozc`P~
zi9&JEklec><N?lm9i{m_RIuaG&Ej<wfRq#!6{;p5D5~JDDD#3F%|n{0kk(d>w@N^a
zGk>$TI%!e*Rx-Y;TVT%N+UOu&^vW9qT%osE*4S=hSnd<D^|qG~`_iW6nd4cO*qCpc
zX-0hx%J~NLJE|@*2vd{Sm=ZU-nIf+(ufkU=?^6nJ<DhtiUhhib>rX*hE@H}1?M1e4
zWMo7zV0!I;9a=EA6DgXp!?XDV*Z_~D6Em#8mNhN&-)Kzn{b^#$hszljO{NEFl$KGf
z3tnBDE&vYQ6{MqjNCtX_lATlzPVo%E_{a}zhB<fZ&wvt(h6twVgp8qyEarAYy#?2`
zPh}=`P_76i%Zk=V6IP2{x7)Pu#VxN<xxW1fp91`KMX6@7iINfWF@B>6l&Db3m*6<1
zn$ZY%n28=(S;y~!>N-Tp!a~_6A;8Iwc66`nm<Cv94utBVvT@Alnh28Hsnyvmp;KIL
zHE!ZY(mO;>Rs>_I(zY&<&v+=z7HT$bE+X{kaeXy9E`b|6CEEE_AfKTWI}evLmLjP@
zdgD%?)4SCfv^ZZZ(H!Tp&LQpgc;?Aa>*7WVQoeUQt8jn^kVZ}wXwB!dS>G6Eo^8Uj
znm0s-M~<Do6v<LdO0DHT#O7n`$)zL%3C>MsR61}C(KVwQeiYSC^mSQn<FbaRsQK5g
zkBO_o7i5FXn3bdVSc7JI+651tW!{*irPk7CNlmuF2{ypaRkcKTuQ$p1!}sT1FVw5f
zRiF#zONiMoVI3I6g9z9xc;_ooF;$QZL#ltm7{zP=#-K!$Z!d-QkPM|r0^3dp;8@=D
z$H}s<Dn!l%^%FyCr<cRFo0yo`L$AgMWPvnc@ZCPb|6uwktveYs>W9+B@HPQoYHN=>
zn=FKGDuJeE1@8uh?=hMvNQmzrR{|ugi*J0h+Gz_8mGbXt<DmQz9^*6p3d&0Ro0IG8
zp(=Zu=E)m1c^cZ5L#4}869lDhW3J$&%PEF<G@tm~8h(cjba-Lo%0y}V9&z~oetfVk
z^fMWDk3c%6n$;n1x#2E588ti>&?$(Z?ua^BaQm;{8IMu1P95`@=lWziEf3HaKh{ez
zE%aEp)Z#fduYTS$QMi`l<p%wXK&i{`q-y;^1xCBL)oJeQP`)tf6edXI?*rS>YV~QZ
zb3DNbxSYSgT){9L@41%zi1L_WzRl*|zcyP}Utg$FSQmBfu@vo}(ipH?ynFyiKW4n;
zf<#L^9-GVm9ICu1R7h7sMc^`wGTuVU=;}<0Bk35H>y~Q4=2TNzoP;8Q&T-+i;`4he
z#C{7MlC*R&YnM9&jt8oP6JgjYF#6Tth6LhQ25x>k+`O-VHB^54i<kWtz96)-(d(fl
za0UYwEUa5<LU2!ru*!@^-+m5y7^9Tv-I9v?iKCl?W0SuMi_Yr^p!ZZX9wa!H?<2if
zd^+Y(rpcPM3PL=Kp}}rJRx`3^NxB5@J9)}hNWn+opu7M$7t{|H`ksaFlL0vLeMZ6w
z7jR`x?5n${Z7rg~Hj`|(+WDEnBsbYZt06XpZ01N)b2I$pTw*ebV*_))_e)E0SlBGU
z>5r`!lh&Vp#LMHx9PTG!O&XHvr>yg;Oi~z)2_i3*15d*dI6TLx9G^{S3iEYxW}82l
zTml9a<@x4qYAJ|=z~{>ozQV{QIT&bGi%ikfoXqY|N#2}tD{XCvNR6S&9ggh}35jB$
zQHW-o-&M^04P2{|wF5iWKerw(wN`^(+koBu;wC@l?Wty6&H@-TJp+*k43tC|Pg^|Q
z>rg=HNl##1FDU#PTJ!o4N2<`{;i*(BS`5diahK85DjG;4NuD%s@oE4~Vzd8oM4>v`
z)x|uDgJ1TW-^_Q5*dfrlKosk`*~>lY<#_Y_@J-8Af(;=~WK;JS4>1{}Ejz>PjYq9L
zqPx1h+>_!o46<Y8>xZGzC+E&MEL?0?_GrFC+ZH%<0F)O15j9kX=5xTDYkT}Bfi;hA
z-<AVA-hgiqnyEd-WT^qhSb;iM6v1S9(w)CzWZl^xRd@RNuRnz2rv#*$CUXG;3tqEM
zc->^&-pI*Pc-_E5o791yQaydz_D~v@0U3sS8KXt*bp0gPq_{55GHrP0FL9`>c%Reb
z7hZ%(aw>;T<m2-l+2r;abASP>$8vO{e0cIX*$Qd>0vbpWdDw^koak|^@xdNl*#|$$
zF8lOWzN37f7?l;Fqtj4YFIk)6){cq~1eR1Ce}OKM&PWjy={!d>&BXmyoAXgaBT?Yf
zU?R_BJZ?wa+YYG$;pVEF$5N94Xb&xpv-Nsx5C}pDIc79Nl?}B<hkn_ylRg+{=h_;y
z&W|@B_f>l)7*7U$$eRF9wZNW<F?tG7K>omOc#9Xf&Exp<J^xTRF|f$&b)#9or}#^}
zKVMt7hDft+7(`}_&1%qZWaRQN8grZClgo-upG!YwO+>t?8Qh(Yonb(eN^hfYt$S9I
zunp{IRHF!08Wno#F2x(}s`j#^d~Le*p8tf+SAZ#~S*#g@P5fs+&F<F!k)J_x;$XGJ
z2DK&>;LBTv1Wl=dkw>?(X*=*lKsQTzSv%+mdeUay#o@{iG&iBL##*`=MtcY$k5gF&
zPswwe_wGbi^78keeRSG>JSP}%kUcuzn^~hvY=5{KCk)<_3sM8;!}WP6^8$OSV3B1I
z(c6OI+#G>$!Tlf<%ghX-?XeRfbgPZW2(qWZ5y!k~`-`mMs1q{Aq95ej>J9);YL1^3
zdV{7=iMm3Sq*q{t2nf%B8dJ}F2o3+~)FiaZ<NXflHU(GP+0#ICE@8ADv?I$kHSMUR
z&&zq(m0vj=Z#3|nuPLwo<(Wg(F>1ORGULmJgCgqiE%3U2d8_*9MNZDW2YU%U8mkm_
z1E?)s-IZa`@e3#KA6&i!!sD}?rpbWRZ6MW?l+lm=Bb!F@SNmIDJ~_8kPy-k#7Hkr8
zCX~SC{Fo9^zZ%3>8-35izm4$GwXkyoq2A@>Ep_Q=b?h0V(-j9B(UN}v3Bi~8aMmfw
z#{!5~eeoROJGU};fC|{ZS|a0Lq@)$<UaoY2=~Uo>(_?(QYycb^$_>wHCa2r1Yt-du
z=IlLrS4Nr%30~{2z0_*mZY&Ascuq);s}_;XSNVHA;F}cd<}`W!Lsm%0r!$KXr|^N9
z(;!_QARPor84+^_OxeJosHJm8M=%*u5<6Q2vNA+4Fci1`aX*P%WsMQ8687itdIdNG
z-4*#5ORmo8zzT(x%WACAqqB1|in`@5RPPN_{$a%Z75+`DbAvt%)u2r?Q|D_YxH*uZ
zVLV$^te-gt1S~(z!u7xEetxY=PP0U*a#q`U1H9{~kDaWrK{~|QH?mM=(55`aocVCu
zYDn=ZGpEd~y}QfQEdyu1wAT-5iCg7>F6u4a<YyspyPn+@uh(|pQ&6NIe-*vKs$W4G
zano`x6Qfng`$6PvQSReVLp@`mrtKVH^D=d>=;})0Gsghl10Csv@uT*0!u`VhRJ%f?
z{9}hXXH_R^C7+ioWkW{ULFC83^@YfOA_$0x1W1xFr-mb+yz^xAth;=3nbz!O=!2N*
z^GSz=dr^vooTUwq_}Mc_@+B=SnlBFDe*cmzhWzU5Dl047hfMR-H{Qe-*q9Ht-z2Qs
zE!doE+$6&G%Fi`;4DM&0mvWihR+0*y?QibCNMJABKjBJ0{Yjb5vdjxp-DW#cc6e!4
z$K<w`(|30zb-WnpZw||?PUkkhD;PaXg7|)_Y!S>27KX@r7051j@F^})yRA@6T_zmJ
ztDBsO65Y=8oO<-tJL+W{s;i4#U$irFz-}UJ16^eOf-%Ong7#ZD_|7x99bk+pTp54z
z-<CYfDbtfCy6HC*n`pRzXeWMMW`5@r809dQhl0=dhDUj?Ntr=7xAdjO`(h=HaIC?2
z4l+axB2$MXPft&4&=@uBE=x>I^f9t^Y+dOL7Ur6%_Gx732cE8}<w0WR+Rn7Wb(dF6
zTWyjFi^}#@KX2<1F}klUpt(z7-l(JZpoi+V`PQd~bsGOr)xW#TajMWDy|--se#_*p
zOsnd$8J5+UAGizckvLsA{<OE2bCL)f6uU4;%&Cbw6P>?SR3%d9*E2K}u}v-LSpkh9
z>+{{I9x<~cp5VI)JT<nP;Xvnz=z>=TVrAPJDEs}1Qdr)jqodYaBNWvRN7`itmq6v6
zvkdD71{q`>R(5NnnP@A}Oqtre#;JL9M>3iMZ0ee|3>4S6PqXF{=9@q18txN+sEU&8
zZ%*GCO`Y~G)Z_=TwNdnw^GbP3$~_`D*7L1<{%Ir48D?{Et9O)Oq<s4u3LmN>GhQvx
zcwKMHQ(cm7VHQg_n2#nf?fmQ7V@8X?oj&6#L4koy(NWXUZnH*ce`_?8)*^GecAtff
zo0|?S*Z!Q@&KLXLARat*5b#51{RZ+*ZYhLwz;^5nhg-h9R^qKo%sk(ndJM+}qMHit
z_VrSGGbweo)KMxiwtC9x(xvKhBGmsRHfQ8I9HKeDa*kn%k5O*vYXHly+ntD(Q`{Rt
z>sNu8-rfbL-Ig1uM!H5=aBXVpJ%&_@%(0f751|>4ZJAb*N!_<|t(J?Ql6n=qOc#4m
zvqTv>I-i$+Qc|P4TWzzU+v-z(baQH-e}1#cv6^^HBM9lZfu8R~nLqvNEapgKyH}sL
zPT<SE7*dF^YCYYO(U~GycT*~bpJw*v4?^>~39T&qDlOfJOI%FMbM>f0pwloa$qrLL
zu!|T=$F)D0SBKQgQ<^PdR=c4?<f>vy8^zScT~%i;Ry6av;S#{D2)2-3ke48j8_hPj
z){gNu6TW;I0>u2H(auAo=E8}0V$8lS<pDti9B(M45+n32uodGpM@|)P%(#5(jJrOL
zj3AC|g`ds~`ljEWMXS`TJEW?1$tB~@NsXAWufF0~Z4Zk`a#V1_aN}}%5*^|8Vk-|k
z$Au13DKX2>P&+#X$#{0sYu*0*fg&ue%MAQbprUy&`SmQ5NCTRlHPxh7u+-a&a^f+w
zml0XI=9{_)y@g*6XbXi}sj$1##3Q0%*E1b|-2z0U9gyhTOk8%mzp8%DLMCRG1i<yH
z_`2xoegcvz3Xl)i3!gp?of#m-V<U38oBqXpIkUrb;d4_}T%8Mr-|z6XciT7GvMOwI
zI)}FN3~%<Q%N~E-WP#t;1vz?+9jSYjkZ_8*u?#ypSFD0-so&=Cn$Z2vY*E(leu}Ic
zJ=__7Gq|%NjXUr?hcf5K`G`wO{-HFq{@nRvM@aBG!VoyS7Z=LJ>FPAhN}^Z4(AKgR
z*uTG}9G*OEM=U*Bvu>5`8RVwi(d5|Obue+cekQl7*Vur>zMA;ygH|f3p-LyvW8hVe
z8GYajvtn?(W=u^@T~<g7J=qxS9UJ?)99F$XmMSr3OIBVIXA&`?Rlaa;(`ACR9Ta+U
zUqyp)FdOB)MXmhHdWDSD7JF+p0H3-B2zWXxtGW~6hA+b6Y%?d0ty^d<_@ulG>Vn6i
z6A4=pW@|E3)r!Uio|?U6_RxLS)a|+3Gn10Tb1%A0ha>wtY;Vr3?4f!|Ke!xc)o9y6
zS4MQx$dS&QiCs~Z((9rRNd3oRPCTKh;Vh*YTPP;S6}fW+U*@b%z;S2!17^<kChY<E
zPK8%m&bp=7G7Ad}soNEV%kcR$HRh~r_K!97_74=*-I!jTQJ7!Fc5N~66ia*&ynbb;
z)9V|(PUG0j`s}1+0mFZDx}m+=)<TcBFSj|`1AW-i0#ntoEc|W@zW|<XBma1aQ)4F?
zC$f?Fg6HCHV(vE+K0bapFvjh@ysQ!z7ym@;`Yw5;>*nSrHFzK$WVQrn3C^SYPC@4;
zGkkn$Bb;<4_|Uwols=OT>UrjvuRiB-PpV%QsK~V<IC^TWUcq>GKFte%yDRSno1StM
zJsk`$@;+_+U{0e~zL3FF$|~sm{O+DyX44h|-17K#Ck6(@Mv|SeD*mqt)B)ABV(5vN
zDYE_?;VEJz7HGiur>UZ8BJf<U8JX>_mliHxUGU@P&&2&@WpP4(Kda;@{$1ZY+bi;|
zc>A4mHeo`oeXb%t4Eq)4B`n$M$dPs{abn4fms&D)Wpy{!&z(En<-ZMQw8#JK$IF7k
zejiobZrnfBWnr=)tMCE+4t0slT!E&UWf3Z}RlB&_=g*b@mYx1w^`gE1A+byUZ|zt)
z<!W8wiMMBxRRdo{xEy5p!2O8RElRp2VVSI$XZyT$haRRsZy)u_kcgH0hc<1Pyxxwh
z)#)qiIRyqr4hIG%fd&Q^Ak(-}b^hKsnf5*NIKz=;CRj4cd^$66k|IbEO#LL`1B(9V
zx~4$&f`w1aVc@HL=5bO0W+F)RB<F((>gldtFpVHHQv4X&ZHkwqaKMdIU}${8QgCAO
zW+hc*Ta45<WVPFTwpMUP*6y&N&*5_EiPu3zjL0?#(8B%(=XsVvMlIj~g1Et<i0MCb
XsF!=Wa>*tY1|aZs^>bP0l+XkKNc}A>

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/logo-dark-with-text-background.svg b/previews/PR3525/assets/logo-dark-with-text-background.svg
new file mode 100644
index 00000000000..eb4cf0b3249
--- /dev/null
+++ b/previews/PR3525/assets/logo-dark-with-text-background.svg
@@ -0,0 +1,48 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="415pt" height="150pt" viewBox="0 0 415 150" version="1.1">
+<g id="surface186">
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(20%,20%,20%);fill-opacity:1;" d="M 207.5 149.5 L 10.5 149.5 C 4.976562 149.5 0.5 145.023438 0.5 139.5 L 0.5 10.5 C 0.5 4.976562 4.976562 0.5 10.5 0.5 L 404.5 0.5 C 410.023438 0.5 414.5 4.976562 414.5 10.5 L 414.5 139.5 C 414.5 145.023438 410.023438 149.5 404.5 149.5 Z M 207.5 149.5 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 30 81 L 130 41 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 14 133 L 130 17 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 93 117 C 93 126.390625 85.390625 134 76 134 C 66.609375 134 59 126.390625 59 117 C 59 107.609375 66.609375 100 76 100 C 85.390625 100 93 107.609375 93 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 133 77 C 133 86.390625 125.390625 94 116 94 C 106.609375 94 99 86.390625 99 77 C 99 67.609375 106.609375 60 116 60 C 125.390625 60 133 67.609375 133 77 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 133 117 C 133 126.390625 125.390625 134 116 134 C 106.609375 134 99 126.390625 99 117 C 99 107.609375 106.609375 100 116 100 C 125.390625 100 133 107.609375 133 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(100%,100%,93.333333%);fill-opacity:1;" d="M 177.417969 60.46875 L 191.445312 60.46875 L 191.445312 107.582031 C 191.445312 111.828125 191.765625 116.0625 190.730469 120.230469 C 189.523438 125.285156 186.5 129.722656 182.238281 132.695312 C 179.277344 134.757812 175.847656 136.050781 172.261719 136.457031 C 160.707031 137.757812 149.894531 130.078125 147.324219 118.734375 L 160.941406 115.257812 C 161.59375 118.300781 163.734375 120.820312 166.628906 121.957031 C 167.917969 122.4375 169.300781 122.605469 170.664062 122.453125 C 171.925781 122.320312 173.132812 121.882812 174.1875 121.175781 C 177.234375 119.054688 177.417969 115.304688 177.417969 112.003906 Z M 206.527344 60.46875 L 220.296875 60.46875 L 220.296875 86.660156 C 220.296875 91.730469 220.640625 95.28125 221.34375 97.25 C 221.953125 99.101562 223.132812 100.714844 224.707031 101.859375 C 226.390625 103.003906 228.394531 103.585938 230.429688 103.523438 C 232.46875 103.582031 234.480469 103.015625 236.191406 101.90625 C 237.84375 100.726562 239.078125 99.054688 239.710938 97.132812 C 240.269531 95.570312 240.546875 92.226562 240.546875 87.097656 L 240.546875 60.5 L 254.222656 60.5 L 254.222656 83.511719 C 254.222656 92.984375 253.460938 99.484375 252.023438 102.972656 C 250.398438 106.984375 247.578125 110.402344 243.949219 112.761719 C 240.382812 115.042969 235.859375 116.183594 230.375 116.183594 C 224.410156 116.183594 219.597656 114.855469 215.929688 112.199219 C 212.191406 109.441406 209.464844 105.53125 208.164062 101.070312 C 207.128906 97.769531 206.570312 91.785156 206.570312 83.117188 Z M 332.804688 133.804688 L 346.65625 133.804688 L 346.65625 60.457031 L 330.816406 60.457031 L 307.835938 91.992188 L 284.855469 60.457031 L 268.90625 60.457031 L 268.90625 133.804688 L 282.753906 133.804688 L 282.753906 80.464844 L 306.152344 112.433594 L 308.878906 112.433594 L 332.804688 80.566406 Z M 359.28125 60.46875 L 374.089844 60.46875 C 382.15625 60.46875 387.933594 61.203125 391.425781 62.667969 C 394.921875 64.101562 397.867188 66.625 399.816406 69.863281 C 401.933594 73.457031 403 77.574219 402.886719 81.742188 C 403.050781 86.332031 401.632812 90.84375 398.871094 94.511719 C 396.105469 98.011719 392.28125 100.523438 387.972656 101.675781 C 385.285156 102.40625 380.382812 102.773438 373.265625 102.773438 L 373.265625 133.859375 L 359.28125 133.859375 Z M 373.265625 89.15625 L 377.664062 89.15625 C 380.117188 89.269531 382.578125 89.019531 384.957031 88.40625 C 386.265625 87.953125 387.394531 87.085938 388.167969 85.933594 C 388.980469 84.699219 389.390625 83.242188 389.347656 81.765625 C 389.496094 79.25 388.21875 76.855469 386.046875 75.582031 C 384.484375 74.625 381.523438 74.175781 377.191406 74.175781 L 373.265625 74.175781 Z M 373.265625 89.15625 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo-dark-with-text.svg b/previews/PR3525/assets/logo-dark-with-text.svg
new file mode 100644
index 00000000000..1031633f5d4
--- /dev/null
+++ b/previews/PR3525/assets/logo-dark-with-text.svg
@@ -0,0 +1,47 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="415pt" height="150pt" viewBox="0 0 415 150" version="1.1">
+<g id="surface181">
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 30 81 L 130 41 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 14 133 L 130 17 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 93 117 C 93 126.390625 85.390625 134 76 134 C 66.609375 134 59 126.390625 59 117 C 59 107.609375 66.609375 100 76 100 C 85.390625 100 93 107.609375 93 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 133 77 C 133 86.390625 125.390625 94 116 94 C 106.609375 94 99 86.390625 99 77 C 99 67.609375 106.609375 60 116 60 C 125.390625 60 133 67.609375 133 77 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 133 117 C 133 126.390625 125.390625 134 116 134 C 106.609375 134 99 126.390625 99 117 C 99 107.609375 106.609375 100 116 100 C 125.390625 100 133 107.609375 133 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(100%,100%,93.333333%);fill-opacity:1;" d="M 177.417969 60.46875 L 191.445312 60.46875 L 191.445312 107.582031 C 191.445312 111.828125 191.765625 116.0625 190.730469 120.230469 C 189.523438 125.285156 186.5 129.722656 182.238281 132.695312 C 179.277344 134.757812 175.847656 136.050781 172.261719 136.457031 C 160.707031 137.757812 149.894531 130.078125 147.324219 118.734375 L 160.941406 115.257812 C 161.59375 118.300781 163.734375 120.820312 166.628906 121.957031 C 167.917969 122.4375 169.300781 122.605469 170.664062 122.453125 C 171.925781 122.320312 173.132812 121.882812 174.1875 121.175781 C 177.234375 119.054688 177.417969 115.304688 177.417969 112.003906 Z M 206.527344 60.46875 L 220.296875 60.46875 L 220.296875 86.660156 C 220.296875 91.730469 220.640625 95.28125 221.34375 97.25 C 221.953125 99.101562 223.132812 100.714844 224.707031 101.859375 C 226.390625 103.003906 228.394531 103.585938 230.429688 103.523438 C 232.46875 103.582031 234.480469 103.015625 236.191406 101.90625 C 237.84375 100.726562 239.078125 99.054688 239.710938 97.132812 C 240.269531 95.570312 240.546875 92.226562 240.546875 87.097656 L 240.546875 60.5 L 254.222656 60.5 L 254.222656 83.511719 C 254.222656 92.984375 253.460938 99.484375 252.023438 102.972656 C 250.398438 106.984375 247.578125 110.402344 243.949219 112.761719 C 240.382812 115.042969 235.859375 116.183594 230.375 116.183594 C 224.410156 116.183594 219.597656 114.855469 215.929688 112.199219 C 212.191406 109.441406 209.464844 105.53125 208.164062 101.070312 C 207.128906 97.769531 206.570312 91.785156 206.570312 83.117188 Z M 332.804688 133.804688 L 346.65625 133.804688 L 346.65625 60.457031 L 330.816406 60.457031 L 307.835938 91.992188 L 284.855469 60.457031 L 268.90625 60.457031 L 268.90625 133.804688 L 282.753906 133.804688 L 282.753906 80.464844 L 306.152344 112.433594 L 308.878906 112.433594 L 332.804688 80.566406 Z M 359.28125 60.46875 L 374.089844 60.46875 C 382.15625 60.46875 387.933594 61.203125 391.425781 62.667969 C 394.921875 64.101562 397.867188 66.625 399.816406 69.863281 C 401.933594 73.457031 403 77.574219 402.886719 81.742188 C 403.050781 86.332031 401.632812 90.84375 398.871094 94.511719 C 396.105469 98.011719 392.28125 100.523438 387.972656 101.675781 C 385.285156 102.40625 380.382812 102.773438 373.265625 102.773438 L 373.265625 133.859375 L 359.28125 133.859375 Z M 373.265625 89.15625 L 377.664062 89.15625 C 380.117188 89.269531 382.578125 89.019531 384.957031 88.40625 C 386.265625 87.953125 387.394531 87.085938 388.167969 85.933594 C 388.980469 84.699219 389.390625 83.242188 389.347656 81.765625 C 389.496094 79.25 388.21875 76.855469 386.046875 75.582031 C 384.484375 74.625 381.523438 74.175781 377.191406 74.175781 L 373.265625 74.175781 Z M 373.265625 89.15625 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo-dark-without-text.svg b/previews/PR3525/assets/logo-dark-without-text.svg
new file mode 100644
index 00000000000..a6221d61d43
--- /dev/null
+++ b/previews/PR3525/assets/logo-dark-without-text.svg
@@ -0,0 +1,46 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="150pt" height="150pt" viewBox="0 0 150 150" version="1.1">
+<g id="surface196">
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 30 78 L 130 38 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 14 130 L 130 14 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 93 114 C 93 123.390625 85.390625 131 76 131 C 66.609375 131 59 123.390625 59 114 C 59 104.609375 66.609375 97 76 97 C 85.390625 97 93 104.609375 93 114 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 133 74 C 133 83.390625 125.390625 91 116 91 C 106.609375 91 99 83.390625 99 74 C 99 64.609375 106.609375 57 116 57 C 125.390625 57 133 64.609375 133 74 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 133 114 C 133 123.390625 125.390625 131 116 131 C 106.609375 131 99 123.390625 99 114 C 99 104.609375 106.609375 97 116 97 C 125.390625 97 133 104.609375 133 114 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo-dark.svg b/previews/PR3525/assets/logo-dark.svg
new file mode 100644
index 00000000000..4014762ca62
--- /dev/null
+++ b/previews/PR3525/assets/logo-dark.svg
@@ -0,0 +1,47 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="200pt" height="200pt" viewBox="0 0 200 200" version="1.1">
+<g id="surface191">
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 63 71.199219 L 153 35.199219 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(100%,100%,93.333333%);stroke-opacity:1;stroke-miterlimit:10;" d="M 48.601562 118 L 153 13.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 119.699219 103.601562 C 119.699219 112.050781 112.851562 118.898438 104.398438 118.898438 C 95.949219 118.898438 89.101562 112.050781 89.101562 103.601562 C 89.101562 95.148438 95.949219 88.300781 104.398438 88.300781 C 112.851562 88.300781 119.699219 95.148438 119.699219 103.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 155.699219 67.601562 C 155.699219 76.050781 148.851562 82.898438 140.398438 82.898438 C 131.949219 82.898438 125.101562 76.050781 125.101562 67.601562 C 125.101562 59.148438 131.949219 52.300781 140.398438 52.300781 C 148.851562 52.300781 155.699219 59.148438 155.699219 67.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 155.699219 103.601562 C 155.699219 112.050781 148.851562 118.898438 140.398438 118.898438 C 131.949219 118.898438 125.101562 112.050781 125.101562 103.601562 C 125.101562 95.148438 131.949219 88.300781 140.398438 88.300781 C 148.851562 88.300781 155.699219 95.148438 155.699219 103.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(100%,100%,93.333333%);fill-opacity:1;" d="M 27.554688 137 L 37.882812 137 L 37.882812 171.691406 C 37.882812 174.816406 38.117188 177.9375 37.355469 181.007812 C 36.46875 184.726562 34.242188 187.996094 31.101562 190.183594 C 28.921875 191.703125 26.398438 192.65625 23.753906 192.953125 C 15.246094 193.914062 7.285156 188.257812 5.394531 179.90625 L 15.421875 177.34375 C 15.902344 179.585938 17.476562 181.441406 19.609375 182.277344 C 20.558594 182.628906 21.574219 182.753906 22.582031 182.644531 C 23.507812 182.542969 24.398438 182.222656 25.171875 181.703125 C 27.417969 180.140625 27.554688 177.378906 27.554688 174.949219 Z M 48.988281 137 L 59.128906 137 L 59.128906 156.285156 C 59.128906 160.019531 59.378906 162.636719 59.898438 164.085938 C 60.347656 165.449219 61.214844 166.636719 62.375 167.480469 C 63.617188 168.320312 65.089844 168.75 66.589844 168.703125 C 68.089844 168.746094 69.570312 168.328125 70.832031 167.511719 C 72.046875 166.644531 72.957031 165.414062 73.425781 163.996094 C 73.835938 162.847656 74.039062 160.382812 74.039062 156.609375 L 74.039062 137.023438 L 84.109375 137.023438 L 84.109375 153.96875 C 84.109375 160.941406 83.550781 165.730469 82.488281 168.296875 C 81.292969 171.253906 79.21875 173.769531 76.542969 175.507812 C 73.917969 177.1875 70.585938 178.027344 66.546875 178.027344 C 62.15625 178.027344 58.613281 177.046875 55.914062 175.09375 C 53.160156 173.0625 51.152344 170.179688 50.195312 166.894531 C 49.433594 164.464844 49.019531 160.058594 49.019531 153.675781 Z M 141.976562 191 L 152.171875 191 L 152.171875 136.992188 L 140.507812 136.992188 L 123.589844 160.214844 L 106.667969 136.992188 L 94.921875 136.992188 L 94.921875 191 L 105.121094 191 L 105.121094 151.726562 L 122.347656 175.261719 L 124.359375 175.261719 L 141.976562 151.796875 Z M 161.472656 137 L 172.375 137 C 178.3125 137 182.570312 137.539062 185.140625 138.617188 C 187.714844 139.675781 189.882812 141.535156 191.320312 143.917969 C 192.878906 146.5625 193.664062 149.59375 193.582031 152.664062 C 193.699219 156.042969 192.65625 159.367188 190.625 162.070312 C 188.585938 164.644531 185.769531 166.496094 182.597656 167.339844 C 180.621094 167.882812 177.011719 168.152344 171.765625 168.152344 L 171.765625 191.042969 L 161.472656 191.042969 Z M 171.765625 158.125 L 175.007812 158.125 C 176.816406 158.207031 178.625 158.023438 180.378906 157.574219 C 181.34375 157.238281 182.171875 156.597656 182.742188 155.75 C 183.339844 154.839844 183.644531 153.769531 183.609375 152.679688 C 183.71875 150.828125 182.777344 149.066406 181.179688 148.128906 C 180.027344 147.421875 177.851562 147.089844 174.660156 147.089844 L 171.765625 147.089844 Z M 171.765625 158.125 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo-with-text-background.svg b/previews/PR3525/assets/logo-with-text-background.svg
new file mode 100644
index 00000000000..807e338e91c
--- /dev/null
+++ b/previews/PR3525/assets/logo-with-text-background.svg
@@ -0,0 +1,48 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="415pt" height="150pt" viewBox="0 0 415 150" version="1.1">
+<g id="surface166">
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(100%,100%,93.333333%);fill-opacity:1;" d="M 207.5 149.5 L 10.5 149.5 C 4.976562 149.5 0.5 145.023438 0.5 139.5 L 0.5 10.5 C 0.5 4.976562 4.976562 0.5 10.5 0.5 L 404.5 0.5 C 410.023438 0.5 414.5 4.976562 414.5 10.5 L 414.5 139.5 C 414.5 145.023438 410.023438 149.5 404.5 149.5 Z M 207.5 149.5 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 30 81 L 130 41 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 14 133 L 130 17 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 93 117 C 93 126.390625 85.390625 134 76 134 C 66.609375 134 59 126.390625 59 117 C 59 107.609375 66.609375 100 76 100 C 85.390625 100 93 107.609375 93 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 133 77 C 133 86.390625 125.390625 94 116 94 C 106.609375 94 99 86.390625 99 77 C 99 67.609375 106.609375 60 116 60 C 125.390625 60 133 67.609375 133 77 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 133 117 C 133 126.390625 125.390625 134 116 134 C 106.609375 134 99 126.390625 99 117 C 99 107.609375 106.609375 100 116 100 C 125.390625 100 133 107.609375 133 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(20%,20%,20%);fill-opacity:1;" d="M 177.417969 60.46875 L 191.445312 60.46875 L 191.445312 107.582031 C 191.445312 111.828125 191.765625 116.0625 190.730469 120.230469 C 189.523438 125.285156 186.5 129.722656 182.238281 132.695312 C 179.277344 134.757812 175.847656 136.050781 172.261719 136.457031 C 160.707031 137.757812 149.894531 130.078125 147.324219 118.734375 L 160.941406 115.257812 C 161.59375 118.300781 163.734375 120.820312 166.628906 121.957031 C 167.917969 122.4375 169.300781 122.605469 170.664062 122.453125 C 171.925781 122.320312 173.132812 121.882812 174.1875 121.175781 C 177.234375 119.054688 177.417969 115.304688 177.417969 112.003906 Z M 206.527344 60.46875 L 220.296875 60.46875 L 220.296875 86.660156 C 220.296875 91.730469 220.640625 95.28125 221.34375 97.25 C 221.953125 99.101562 223.132812 100.714844 224.707031 101.859375 C 226.390625 103.003906 228.394531 103.585938 230.429688 103.523438 C 232.46875 103.582031 234.480469 103.015625 236.191406 101.90625 C 237.84375 100.726562 239.078125 99.054688 239.710938 97.132812 C 240.269531 95.570312 240.546875 92.226562 240.546875 87.097656 L 240.546875 60.5 L 254.222656 60.5 L 254.222656 83.511719 C 254.222656 92.984375 253.460938 99.484375 252.023438 102.972656 C 250.398438 106.984375 247.578125 110.402344 243.949219 112.761719 C 240.382812 115.042969 235.859375 116.183594 230.375 116.183594 C 224.410156 116.183594 219.597656 114.855469 215.929688 112.199219 C 212.191406 109.441406 209.464844 105.53125 208.164062 101.070312 C 207.128906 97.769531 206.570312 91.785156 206.570312 83.117188 Z M 332.804688 133.804688 L 346.65625 133.804688 L 346.65625 60.457031 L 330.816406 60.457031 L 307.835938 91.992188 L 284.855469 60.457031 L 268.90625 60.457031 L 268.90625 133.804688 L 282.753906 133.804688 L 282.753906 80.464844 L 306.152344 112.433594 L 308.878906 112.433594 L 332.804688 80.566406 Z M 359.28125 60.46875 L 374.089844 60.46875 C 382.15625 60.46875 387.933594 61.203125 391.425781 62.667969 C 394.921875 64.101562 397.867188 66.625 399.816406 69.863281 C 401.933594 73.457031 403 77.574219 402.886719 81.742188 C 403.050781 86.332031 401.632812 90.84375 398.871094 94.511719 C 396.105469 98.011719 392.28125 100.523438 387.972656 101.675781 C 385.285156 102.40625 380.382812 102.773438 373.265625 102.773438 L 373.265625 133.859375 L 359.28125 133.859375 Z M 373.265625 89.15625 L 377.664062 89.15625 C 380.117188 89.269531 382.578125 89.019531 384.957031 88.40625 C 386.265625 87.953125 387.394531 87.085938 388.167969 85.933594 C 388.980469 84.699219 389.390625 83.242188 389.347656 81.765625 C 389.496094 79.25 388.21875 76.855469 386.046875 75.582031 C 384.484375 74.625 381.523438 74.175781 377.191406 74.175781 L 373.265625 74.175781 Z M 373.265625 89.15625 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo-with-text.pdf b/previews/PR3525/assets/logo-with-text.pdf
new file mode 100644
index 0000000000000000000000000000000000000000..aacf7bac495360cf663abca6046564c6f8ec6698
GIT binary patch
literal 2157
zcmZWr3piAH8&9qJnoqTTu}PNGWnzfAFgK(y#w~KU6q?4FF=7s9#$}~tUE1P9tK2U|
zlrSpym1NfLShtjt%cyTfdE5%Mno@k{DAv>dpYxn^{=a$8`@X;Tcb?xFtUZINL)0Z>
zu+>xL^B8@A06-`FFh)iI{xHbr3Vi`G0y$y;0KhYO{z4E&uKp|`$N<@p4~Q`~#t4Kk
z$O^!O<piYJCV^yyxKVd=%AvE{Mj?4iONFRiC!*BW#+Gc@w63O$tMT^fwe{8Y{C*{k
zIl8gkUe3a4?1%I8;Rz?1Dy%j0ezBQ4GvN(?`1lH-KQD7~{h(p_fifcF%Eq|gJeO%l
z{`Y^scR57Ve!L(Ww-hm2Q#>BaaMt+ii953Q=FaR{P`<lmapQlYXMF;Wh-zw!O`Rc)
z{nu`VOKeiAQ;YJZyzg54oae`PxF~bfS(vTLYy)!hB***>Wi@pkZ3g_;N?h=lSgBq~
zq@4`Ro${`|qUBA0;9zZ2^tSR;-Bhn&VP<eFcT?m3Kb0c&n6a>EAS6|%ezsh>&`&4m
z*3+SZt3I<XbI<`>@%p|8@3e4<(!KfCZRX9D1&{P!hlkK=)Tiq^tr`QohJx&qyre(1
zM?4wzjyyQ#`QTOSD@W-gEeW?LEj@Lhw<GvY?|aG1=k!Uh9<|wWKPcpq-wh?VYHmz>
z!_eIW+fA<hX?DD>D@?O7rB1UEZU~VT#GB3N7+`dS^Q!Ik*`Oo$x;77&K0F*EyL$7@
z3kODjn)IO!)`StwhVy>>sx6OaQt-=U>}XTjAH4dJix6&)sNvt|U8BtvZLx00^b$*o
z;UalY(yXzWN<QBX27uA85=eb;1wW~<Q&GIbHUIv`+d=yn)a0)-j&E@ZZ#vnDrE5Iw
zZYuO?PBn>kUYnFTUiGWB%4T2IZYwE6X=_7sz)qaG7H)jucFB!1Ci%vKbBS3mFC<Kb
zi!%1OkE?Pm@8;y%aeU7=yy3KyT&FR~>x^4~k!1Tav47s+#o6pllApSV2}9NHPY)FE
zOho;q^{#FuxsN10ekYC{<;TbwlpRRjv8!*S&wsa=l?=w~ZLYk(?Ypku3LIu`Zl1Ix
z{#7xk7DX0E2&wY~lk`+-PR$xxTM6@cpP_Q2!X1OsODs+Ie-%y)&~I9~>xo>9T8uiq
zjw+v-sG^h#6XgzPhNNYl`T3CZl{<Vxe2TY5lhJyP`XRSn0X@@2F<)a#=UznlV>8=&
zRZ?@pw{~zuK%Da~^+K)Fd(sr|IMMrDt#Y51R^_{`eYUYDC$@gb`bghzFOsYi=Cw0a
z5}hm_-wu~D&5Z+sT@{`<Yi_gbWcKaTtsgJGwT|3#ZcMq`CcP@FKW?P+-l^_25na0F
zHC8FnYGa2d+B^c{-4o=RgJ^d9C?`GTvtUWFcoC;{;a7{+nTSEP=A&hyN_A(-S{xkS
zXNj%$+pS@@6fTw((=Lx3Od8FIH@!WWH~q5Wo8l)&D*C%si*eG2o)O8T<*hHAwoA{N
zZq+|M?dwdJ%r%fP9L<>!O+Vtr^Z7iJ(F*Za$s3J!ZLMb+30yfN1@YjsoSmhT{*GK)
z?c2DRbIqmseR16|nQ6NBEOBD)%?6j5cVPRN=yAmHjPm^rvYiD6QStX`{Pzi34Vkt%
zW;u&;iw(Oo$WLCCV6R=SKGx8)fd(00`T-2BcYOE9LSa8O3Z~0GhE4Vk%Qf8F{%CV~
z@`kb-E|k1Vs!}N9^y>&2tLLm@ki2G-Un##<D=(Y;<FvuIKU~4x=R~@KH)0H1<_9(8
zZBP7lUBzd(L$8Qjm2j{l11tXa<fIjAtS;b6Qy@1#PNA?}8sr9^8puwyQ1Ecdg6>z!
ziMOcB8-%jU>n~i53u7p^msWnf?7>Ta*A%L%J`^c_uM?rZAn&Efbi;srpU*ys+@TJN
z+9rBMjFU*BQqcf(tG8mXOO|RLCbV!8B6bQG8KD5O@GJtd0q|xf01?2mS;!eNS>y-g
zj2xfykXG^lc<1i}K>%;h;(`DcATH&#XTcy}h&;ON-L{|)k7Wvl0Y?#t0Fa4vfJmbO
zo=eeY5MPLd3joyqKY6_c3<U?SYU)#sC3Lw@C!{jW;tK-NTI}zaVJkog2ZPHyGbD-u
zhVa;+qlGC7e&T|61O-qq%mxJjIuXZJ%+Q|ENue|No0$N87H`Az5g-gu@{TLL2ctw+
zC7`TIU=Kq+!EEHCrO9UTU<e@U5-GX_K+{(!3=|mPk?|rZ7iI<e^4J1h2<GB&ONGpk
zPnwVi@foNznhXOHfkYzENpyl9kwBtr6G$2a0&Z1J>Z+Qbnl@tzS^f}rDH|${;IkNi
zOBe=&91H=VVF)V=ppeOWWPk&F#)uRe!WCTre+AR0AToV{>CyiYM<5W9cks(N3jH4#
zRUhg8i#Qquk!2YZ!YrOY2rnt(#0v+JmQgVrAxH?I#6P_;R(uWwptD~RQy^r)!X<`8
OIuZGIfi*wE#QYydHf{p|

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/logo-with-text.svg b/previews/PR3525/assets/logo-with-text.svg
new file mode 100644
index 00000000000..60ebbc89883
--- /dev/null
+++ b/previews/PR3525/assets/logo-with-text.svg
@@ -0,0 +1,47 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="415pt" height="150pt" viewBox="0 0 415 150" version="1.1">
+<g id="surface161">
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 30 81 L 130 41 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 14 133 L 130 17 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 93 117 C 93 126.390625 85.390625 134 76 134 C 66.609375 134 59 126.390625 59 117 C 59 107.609375 66.609375 100 76 100 C 85.390625 100 93 107.609375 93 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 133 77 C 133 86.390625 125.390625 94 116 94 C 106.609375 94 99 86.390625 99 77 C 99 67.609375 106.609375 60 116 60 C 125.390625 60 133 67.609375 133 77 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 133 117 C 133 126.390625 125.390625 134 116 134 C 106.609375 134 99 126.390625 99 117 C 99 107.609375 106.609375 100 116 100 C 125.390625 100 133 107.609375 133 117 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(20%,20%,20%);fill-opacity:1;" d="M 177.417969 60.46875 L 191.445312 60.46875 L 191.445312 107.582031 C 191.445312 111.828125 191.765625 116.0625 190.730469 120.230469 C 189.523438 125.285156 186.5 129.722656 182.238281 132.695312 C 179.277344 134.757812 175.847656 136.050781 172.261719 136.457031 C 160.707031 137.757812 149.894531 130.078125 147.324219 118.734375 L 160.941406 115.257812 C 161.59375 118.300781 163.734375 120.820312 166.628906 121.957031 C 167.917969 122.4375 169.300781 122.605469 170.664062 122.453125 C 171.925781 122.320312 173.132812 121.882812 174.1875 121.175781 C 177.234375 119.054688 177.417969 115.304688 177.417969 112.003906 Z M 206.527344 60.46875 L 220.296875 60.46875 L 220.296875 86.660156 C 220.296875 91.730469 220.640625 95.28125 221.34375 97.25 C 221.953125 99.101562 223.132812 100.714844 224.707031 101.859375 C 226.390625 103.003906 228.394531 103.585938 230.429688 103.523438 C 232.46875 103.582031 234.480469 103.015625 236.191406 101.90625 C 237.84375 100.726562 239.078125 99.054688 239.710938 97.132812 C 240.269531 95.570312 240.546875 92.226562 240.546875 87.097656 L 240.546875 60.5 L 254.222656 60.5 L 254.222656 83.511719 C 254.222656 92.984375 253.460938 99.484375 252.023438 102.972656 C 250.398438 106.984375 247.578125 110.402344 243.949219 112.761719 C 240.382812 115.042969 235.859375 116.183594 230.375 116.183594 C 224.410156 116.183594 219.597656 114.855469 215.929688 112.199219 C 212.191406 109.441406 209.464844 105.53125 208.164062 101.070312 C 207.128906 97.769531 206.570312 91.785156 206.570312 83.117188 Z M 332.804688 133.804688 L 346.65625 133.804688 L 346.65625 60.457031 L 330.816406 60.457031 L 307.835938 91.992188 L 284.855469 60.457031 L 268.90625 60.457031 L 268.90625 133.804688 L 282.753906 133.804688 L 282.753906 80.464844 L 306.152344 112.433594 L 308.878906 112.433594 L 332.804688 80.566406 Z M 359.28125 60.46875 L 374.089844 60.46875 C 382.15625 60.46875 387.933594 61.203125 391.425781 62.667969 C 394.921875 64.101562 397.867188 66.625 399.816406 69.863281 C 401.933594 73.457031 403 77.574219 402.886719 81.742188 C 403.050781 86.332031 401.632812 90.84375 398.871094 94.511719 C 396.105469 98.011719 392.28125 100.523438 387.972656 101.675781 C 385.285156 102.40625 380.382812 102.773438 373.265625 102.773438 L 373.265625 133.859375 L 359.28125 133.859375 Z M 373.265625 89.15625 L 377.664062 89.15625 C 380.117188 89.269531 382.578125 89.019531 384.957031 88.40625 C 386.265625 87.953125 387.394531 87.085938 388.167969 85.933594 C 388.980469 84.699219 389.390625 83.242188 389.347656 81.765625 C 389.496094 79.25 388.21875 76.855469 386.046875 75.582031 C 384.484375 74.625 381.523438 74.175781 377.191406 74.175781 L 373.265625 74.175781 Z M 373.265625 89.15625 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo-without-text.svg b/previews/PR3525/assets/logo-without-text.svg
new file mode 100644
index 00000000000..8bf39aa46ce
--- /dev/null
+++ b/previews/PR3525/assets/logo-without-text.svg
@@ -0,0 +1,46 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="150pt" height="150pt" viewBox="0 0 150 150" version="1.1">
+<g id="surface176">
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 30 78 L 130 38 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 14 130 L 130 14 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 93 114 C 93 123.390625 85.390625 131 76 131 C 66.609375 131 59 123.390625 59 114 C 59 104.609375 66.609375 97 76 97 C 85.390625 97 93 104.609375 93 114 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 133 74 C 133 83.390625 125.390625 91 116 91 C 106.609375 91 99 83.390625 99 74 C 99 64.609375 106.609375 57 116 57 C 125.390625 57 133 64.609375 133 74 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 133 114 C 133 123.390625 125.390625 131 116 131 C 106.609375 131 99 123.390625 99 114 C 99 104.609375 106.609375 97 116 97 C 125.390625 97 133 104.609375 133 114 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo.svg b/previews/PR3525/assets/logo.svg
new file mode 100644
index 00000000000..31fa5edeeab
--- /dev/null
+++ b/previews/PR3525/assets/logo.svg
@@ -0,0 +1,47 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<svg xmlns:svg="http://www.w3.org/2000/svg"
+  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+  xmlns:cc="http://creativecommons.org/ns#"
+  xmlns:dc="http://purl.org/dc/elements/1.1/"
+xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="200pt" height="200pt" viewBox="0 0 200 200" version="1.1">
+<g id="surface171">
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 63 71.199219 L 153 35.199219 "/>
+<path style="fill:none;stroke-width:8;stroke-linecap:round;stroke-linejoin:miter;stroke:rgb(20%,20%,20%);stroke-opacity:1;stroke-miterlimit:10;" d="M 48.601562 118 L 153 13.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(79.6%,23.5%,20%);fill-opacity:1;" d="M 119.699219 103.601562 C 119.699219 112.050781 112.851562 118.898438 104.398438 118.898438 C 95.949219 118.898438 89.101562 112.050781 89.101562 103.601562 C 89.101562 95.148438 95.949219 88.300781 104.398438 88.300781 C 112.851562 88.300781 119.699219 95.148438 119.699219 103.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(22%,59.6%,14.9%);fill-opacity:1;" d="M 155.699219 67.601562 C 155.699219 76.050781 148.851562 82.898438 140.398438 82.898438 C 131.949219 82.898438 125.101562 76.050781 125.101562 67.601562 C 125.101562 59.148438 131.949219 52.300781 140.398438 52.300781 C 148.851562 52.300781 155.699219 59.148438 155.699219 67.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(58.4%,34.5%,69.8%);fill-opacity:1;" d="M 155.699219 103.601562 C 155.699219 112.050781 148.851562 118.898438 140.398438 118.898438 C 131.949219 118.898438 125.101562 112.050781 125.101562 103.601562 C 125.101562 95.148438 131.949219 88.300781 140.398438 88.300781 C 148.851562 88.300781 155.699219 95.148438 155.699219 103.601562 "/>
+<path style=" stroke:none;fill-rule:nonzero;fill:rgb(20%,20%,20%);fill-opacity:1;" d="M 27.554688 137 L 37.882812 137 L 37.882812 171.691406 C 37.882812 174.816406 38.117188 177.9375 37.355469 181.007812 C 36.46875 184.726562 34.242188 187.996094 31.101562 190.183594 C 28.921875 191.703125 26.398438 192.65625 23.753906 192.953125 C 15.246094 193.914062 7.285156 188.257812 5.394531 179.90625 L 15.421875 177.34375 C 15.902344 179.585938 17.476562 181.441406 19.609375 182.277344 C 20.558594 182.628906 21.574219 182.753906 22.582031 182.644531 C 23.507812 182.542969 24.398438 182.222656 25.171875 181.703125 C 27.417969 180.140625 27.554688 177.378906 27.554688 174.949219 Z M 48.988281 137 L 59.128906 137 L 59.128906 156.285156 C 59.128906 160.019531 59.378906 162.636719 59.898438 164.085938 C 60.347656 165.449219 61.214844 166.636719 62.375 167.480469 C 63.617188 168.320312 65.089844 168.75 66.589844 168.703125 C 68.089844 168.746094 69.570312 168.328125 70.832031 167.511719 C 72.046875 166.644531 72.957031 165.414062 73.425781 163.996094 C 73.835938 162.847656 74.039062 160.382812 74.039062 156.609375 L 74.039062 137.023438 L 84.109375 137.023438 L 84.109375 153.96875 C 84.109375 160.941406 83.550781 165.730469 82.488281 168.296875 C 81.292969 171.253906 79.21875 173.769531 76.542969 175.507812 C 73.917969 177.1875 70.585938 178.027344 66.546875 178.027344 C 62.15625 178.027344 58.613281 177.046875 55.914062 175.09375 C 53.160156 173.0625 51.152344 170.179688 50.195312 166.894531 C 49.433594 164.464844 49.019531 160.058594 49.019531 153.675781 Z M 141.976562 191 L 152.171875 191 L 152.171875 136.992188 L 140.507812 136.992188 L 123.589844 160.214844 L 106.667969 136.992188 L 94.921875 136.992188 L 94.921875 191 L 105.121094 191 L 105.121094 151.726562 L 122.347656 175.261719 L 124.359375 175.261719 L 141.976562 151.796875 Z M 161.472656 137 L 172.375 137 C 178.3125 137 182.570312 137.539062 185.140625 138.617188 C 187.714844 139.675781 189.882812 141.535156 191.320312 143.917969 C 192.878906 146.5625 193.664062 149.59375 193.582031 152.664062 C 193.699219 156.042969 192.65625 159.367188 190.625 162.070312 C 188.585938 164.644531 185.769531 166.496094 182.597656 167.339844 C 180.621094 167.882812 177.011719 168.152344 171.765625 168.152344 L 171.765625 191.042969 L 161.472656 191.042969 Z M 171.765625 158.125 L 175.007812 158.125 C 176.816406 158.207031 178.625 158.023438 180.378906 157.574219 C 181.34375 157.238281 182.171875 156.597656 182.742188 155.75 C 183.339844 154.839844 183.644531 153.769531 183.609375 152.679688 C 183.71875 150.828125 182.777344 149.066406 181.179688 148.128906 C 180.027344 147.421875 177.851562 147.089844 174.660156 147.089844 L 171.765625 147.089844 Z M 171.765625 158.125 "/>
+</g>
+<metadata
+    id="metadata21">
+    <rdf:RDF>
+        <cc:Work
+            rdf:about="">
+        <dc:creator>
+            <cc:Agent>
+            <dc:title>JuMP project</dc:title>
+            </cc:Agent>
+        </dc:creator>
+        <dc:title>Logo for JuMP</dc:title>
+        <cc:license
+            rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+        <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+        <dc:date>August 2018</dc:date>
+        </cc:Work>
+        <cc:License
+            rdf:about="http://creativecommons.org/licenses/by/4.0/">
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Reproduction" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#Distribution" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Notice" />
+        <cc:requires
+            rdf:resource="http://creativecommons.org/ns#Attribution" />
+        <cc:permits
+            rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+        </cc:License>
+    </rdf:RDF>
+</metadata>
+</svg>
+
diff --git a/previews/PR3525/assets/logo_constructor.jl b/previews/PR3525/assets/logo_constructor.jl
new file mode 100644
index 00000000000..50ec1a82fbf
--- /dev/null
+++ b/previews/PR3525/assets/logo_constructor.jl
@@ -0,0 +1,395 @@
+# This script rebuilds the JuMP logos. Install Luxor using
+#
+#   import Pkg; Pkg.pkg"add Luxor@3"
+
+using Luxor
+
+"""
+    _logo_no_text(; color::String)
+
+Creates the three Julia dots with the constraints above.
+"""
+function _logo_no_text(; color::String)
+    setcolor(color)
+    setline(8)
+    setlinecap("round")
+    line(Point(20, 68), Point(120, 28), :stroke)
+    line(Point(4, 120), Point(120, 4), :stroke)
+    pts = box(Point(86, 84), 40, 40; vertices = true)
+    @layer begin
+        setcolor(Luxor.julia_red)
+        circle(pts[1], 17, :fill)
+        setcolor(Luxor.julia_green)
+        circle(pts[3], 17, :fill)
+        setcolor(Luxor.julia_purple)
+        circle(pts[4], 17, :fill)
+    end
+    return
+end
+
+"""
+    _jump_text_outlines(; color::String)
+
+Creates the text part of the JuMP logo with the top-aligned "u".
+"""
+function _jump_text_outlines(; color::String)
+    setcolor(color)
+    @layer begin
+        # "J"
+        newsubpath()
+        move(Point(34.018, 6.789))
+        line(Point(46.768, 6.789))
+        line(Point(46.768, 49.619))
+        curve(
+            Point(46.768, 53.479),
+            Point(47.058, 57.329),
+            Point(46.118, 61.119),
+        )
+        curve(
+            Point(45.022, 65.714),
+            Point(42.273, 69.748),
+            Point(38.398, 72.449),
+        )
+        curve(
+            Point(35.706, 74.325),
+            Point(32.589, 75.501),
+            Point(29.328, 75.869),
+        )
+        curve(Point(18.823, 77.054), Point(8.995, 70.07), Point(6.658, 59.759))
+        line(Point(19.038, 56.599))
+        curve(
+            Point(19.632, 59.365),
+            Point(21.576, 61.655),
+            Point(24.208, 62.689),
+        )
+        curve(
+            Point(25.38, 63.124),
+            Point(26.637, 63.278),
+            Point(27.878, 63.139),
+        )
+        curve(
+            Point(29.023, 63.018),
+            Point(30.122, 62.62),
+            Point(31.078, 61.979),
+        )
+        curve(
+            Point(33.848, 60.049),
+            Point(34.018, 56.639),
+            Point(34.018, 53.639),
+        )
+        closepath()
+        # "U"
+        move(Point(60.478, 6.789))
+        line(Point(72.998, 6.789))
+        line(Point(72.998, 30.599))
+        curve(
+            Point(72.998, 35.209),
+            Point(73.308, 38.439),
+            Point(73.948, 40.229),
+        )
+        curve(
+            Point(74.504, 41.912),
+            Point(75.575, 43.378),
+            Point(77.008, 44.419),
+        )
+        curve(
+            Point(78.538, 45.459),
+            Point(80.359, 45.988),
+            Point(82.208, 45.929),
+        )
+        curve(
+            Point(84.063, 45.982),
+            Point(85.891, 45.469),
+            Point(87.448, 44.459),
+        )
+        curve(
+            Point(88.948, 43.389),
+            Point(90.07, 41.869),
+            Point(90.648, 40.119),
+        )
+        curve(
+            Point(91.155, 38.699),
+            Point(91.408, 35.659),
+            Point(91.408, 30.999),
+        )
+        line(Point(91.408, 6.819))
+        line(Point(103.838, 6.819))
+        line(Point(103.838, 27.739))
+        curve(
+            Point(103.838, 36.349),
+            Point(103.148, 42.259),
+            Point(101.838, 45.429),
+        )
+        curve(
+            Point(100.362, 49.077),
+            Point(97.799, 52.185),
+            Point(94.498, 54.329),
+        )
+        curve(
+            Point(91.258, 56.403),
+            Point(87.145, 57.439),
+            Point(82.158, 57.439),
+        )
+        curve(
+            Point(76.738, 57.439),
+            Point(72.362, 56.233),
+            Point(69.028, 53.819),
+        )
+        curve(
+            Point(65.63, 51.311),
+            Point(63.149, 47.755),
+            Point(61.968, 43.699),
+        )
+        curve(
+            Point(61.028, 40.699),
+            Point(60.518, 35.259),
+            Point(60.518, 27.379),
+        )
+        closepath()
+        # "M"
+        move(Point(175.278, 73.459))
+        line(Point(187.868, 73.459))
+        line(Point(187.868, 6.779))
+        line(Point(173.468, 6.779))
+        line(Point(152.578, 35.449))
+        line(Point(131.688, 6.779))
+        line(Point(117.188, 6.779))
+        line(Point(117.188, 73.459))
+        line(Point(129.778, 73.459))
+        line(Point(129.778, 24.969))
+        line(Point(151.048, 54.029))
+        line(Point(153.528, 54.029))
+        line(Point(175.278, 25.059))
+        closepath()
+        # "P"
+        move(Point(199.348, 6.789))
+        line(Point(212.808, 6.789))
+        curve(
+            Point(220.141, 6.789),
+            Point(225.395, 7.456),
+            Point(228.568, 8.789),
+        )
+        curve(
+            Point(231.747, 10.093),
+            Point(234.423, 12.387),
+            Point(236.198, 15.329),
+        )
+        curve(
+            Point(238.123, 18.596),
+            Point(239.09, 22.339),
+            Point(238.988, 26.129),
+        )
+        curve(
+            Point(239.137, 30.302),
+            Point(237.848, 34.402),
+            Point(235.338, 37.739),
+        )
+        curve(
+            Point(232.823, 40.92),
+            Point(229.346, 43.204),
+            Point(225.428, 44.249),
+        )
+        curve(
+            Point(222.988, 44.916),
+            Point(218.531, 45.249),
+            Point(212.058, 45.249),
+        )
+        line(Point(212.058, 73.509))
+        line(Point(199.348, 73.509))
+        closepath()
+        newsubpath()
+        move(Point(212.058, 32.869))
+        line(Point(216.058, 32.869))
+        curve(
+            Point(218.29, 32.974),
+            Point(220.524, 32.745),
+            Point(222.688, 32.189),
+        )
+        curve(
+            Point(223.879, 31.776),
+            Point(224.904, 30.986),
+            Point(225.608, 29.939),
+        )
+        curve(
+            Point(226.346, 28.816),
+            Point(226.72, 27.493),
+            Point(226.678, 26.149),
+        )
+        curve(
+            Point(226.815, 23.864),
+            Point(225.653, 21.687),
+            Point(223.678, 20.529),
+        )
+        curve(
+            Point(222.258, 19.659),
+            Point(219.568, 19.249),
+            Point(215.628, 19.249),
+        )
+        line(Point(212.058, 19.249))
+        closepath()
+    end
+    fillpath()
+    return
+end
+
+"""
+    logo_with_text(;
+        filename::String,
+        color::String,
+        background::String = "",
+        verbose::Bool = false,
+    )
+
+Create the JuMP logo with the text on the right-hand side.
+
+If `verbose`, print the margin lines.
+"""
+function logo_with_text(;
+    filename::String,
+    color::String,
+    background::String = "",
+    verbose::Bool = false,
+)
+    width, height, margin = 415, 150, 10
+    Drawing(width, height, filename)
+    if !isempty(background)
+        @layer begin
+            translate(O)
+            translate(width / 2, height / 2)
+            setcolor(background)
+            box(O, width - 1, height - 1, 10, :fill)
+        end
+    end
+    @layer begin
+        translate(O)
+        if verbose
+            line(Point(0, margin), Point(width, margin), :stroke)
+            line(Point(margin, 0), Point(margin, height), :stroke)
+            w = width - margin
+            line(Point(w, 0), Point(w, height), :stroke)
+            h = height - margin
+            line(Point(0, h), Point(width, h), :stroke)
+        end
+        translate(10, 13)
+        _logo_no_text(; color = color)
+        translate(130, 40)
+        scale(1.1)
+        _jump_text_outlines(; color = color)
+    end
+    finish()
+    add_license(filename)
+    return
+end
+
+"""
+    logo_square(; filename::String, color::String, verbose::Bool = false)
+
+Create the JuMP logo with the text vertically beneath the pictogram.
+
+If `verbose`, print the margin lines.
+"""
+function logo_square(; filename::String, color::String, verbose::Bool = false)
+    N, margin, w = 200, 10, 110
+    Drawing(N, N, filename)
+    @layer begin
+        translate(O)
+        if verbose
+            line(Point(0, margin), Point(N, margin), :stroke)
+            line(Point(margin, 0), Point(margin, N), :stroke)
+            line(Point(N - margin, 0), Point(N - margin, N), :stroke)
+            line(Point(0, N - margin), Point(N, N - margin), :stroke)
+            lhs = (N - w) / 2
+            line(Point(lhs, 0), Point(lhs, N), :stroke)
+            line(Point(N - lhs, 0), Point(N - lhs, N), :stroke)
+        end
+        translate((N - w) / 2, margin)
+        scale(0.9)
+        _logo_no_text(; color = color)
+        translate(-50, 135)
+        scale(0.9)
+        _jump_text_outlines(; color = color)
+    end
+    finish()
+    add_license(filename)
+    return
+end
+
+"""
+    logo_no_text(; filename::String, color::String)
+
+Create the JuMP logo without the "JuMP" text.
+"""
+function logo_no_text(; filename::String, color::String)
+    Drawing(150, 150, filename)
+    @layer begin
+        translate(O)
+        translate(10, 10)
+        _logo_no_text(; color = color)
+    end
+    finish()
+    add_license(filename)
+    return
+end
+
+function add_license(filename)
+    header = """
+    <svg xmlns:svg="http://www.w3.org/2000/svg"
+      xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+      xmlns:cc="http://creativecommons.org/ns#"
+      xmlns:dc="http://purl.org/dc/elements/1.1/"
+    """
+    footer = """
+    <metadata
+        id="metadata21">
+        <rdf:RDF>
+            <cc:Work
+                rdf:about="">
+            <dc:creator>
+                <cc:Agent>
+                <dc:title>JuMP project</dc:title>
+                </cc:Agent>
+            </dc:creator>
+            <dc:title>Logo for JuMP</dc:title>
+            <cc:license
+                rdf:resource="http://creativecommons.org/licenses/by/4.0/" />
+            <dc:description>Logo for JuMP, an algebraic modeling language and a collection of supporting packages for mathematical optimization embedded in the Julia programming language</dc:description>
+            <dc:date>August 2018</dc:date>
+            </cc:Work>
+            <cc:License
+                rdf:about="http://creativecommons.org/licenses/by/4.0/">
+            <cc:permits
+                rdf:resource="http://creativecommons.org/ns#Reproduction" />
+            <cc:permits
+                rdf:resource="http://creativecommons.org/ns#Distribution" />
+            <cc:requires
+                rdf:resource="http://creativecommons.org/ns#Notice" />
+            <cc:requires
+                rdf:resource="http://creativecommons.org/ns#Attribution" />
+            <cc:permits
+                rdf:resource="http://creativecommons.org/ns#DerivativeWorks" />
+            </cc:License>
+        </rdf:RDF>
+    </metadata>
+    </svg>
+    """
+    file = read(filename, String)
+    file = replace(file, "<svg " => header)
+    file = replace(file, "</svg>" => footer)
+    write(filename, file)
+    return
+end
+
+color(prefix) = isempty(prefix) ? "#333" : "#ffe"
+dark_color(prefix) = isempty(prefix) ? "#ffe" : "#333"
+
+for prefix in ("", "-dark")
+    path = joinpath(@__DIR__, "logo$(prefix)")
+    logo_with_text(; filename = "$(path)-with-text.svg", color = color(prefix))
+    logo_with_text(;
+        filename = "$(path)-with-text-background.svg",
+        color = color(prefix),
+        background = dark_color(prefix),
+    )
+    logo_square(; filename = "$(path).svg", color = color(prefix))
+    logo_no_text(; filename = "$(path)-without-text.svg", color = color(prefix))
+end
diff --git a/previews/PR3525/assets/n_queens4.png b/previews/PR3525/assets/n_queens4.png
new file mode 100644
index 0000000000000000000000000000000000000000..f86590cf5ed8998758e64ada7c106e42867b628d
GIT binary patch
literal 8151
zcmZ`;cT`i)*1a@AK)^!pBGN%XAoL<g?@dYsDT?$ap*JZ~MWpu*0Rqwjp(!E=Qk5c*
z&`YF+PUw8`_ugCYkGH;+buufHbLZZ3XJ+qx&cx{HsNN!FAO!&6mb#j<0j@p!Yb5x%
zZ{oxuM_fbrN>N)80P5b8UA!Q|{bsXOGtdSA|Azn&8u2gx&_4k1<`DpFSOb7m1_02y
z=d|j};Cye`XsRj$*MC0+9pxVZfK^XjS<%RMW_#8wkl8esYxi2RX|^(8vt~x1fB(Mb
zT>3W;H6}I%HnPH-5lxI)qgrE6MgpT=$PFI04|Z!|2MJ^q<R8r*ghk;g6=Wv?H>bN6
z_;q$`rqb;DlbdGw);BOtF6hB-<C-Cw2TnXz6W8a-8qkun`Q`jX9pl`}`#D|;eT?5`
zLNaVc2i)5-IE%x*d2H42i2$<#U?Dy7%=AtEimP8_rLmP@CIyf|K>!TAK;9?ys@}Zn
zuW9MBIM*g2<Yw24llI!`Vhx<;r3Jq5t^ozf8sicIz28waGyvOZ9kjE$cLq;^G~$6a
z>(rh{%$?>c3x+rk^r>pKC!R$>5BS`oydNLn#!n?|ZRSwBc^fa(EXr6=i^*?%+hAVH
z=Mc3chz!86TMVRCBfzu2s2XG*OL9=iZ2Pc>=q<y2TAiOaw14`<xds90;g4_cQ<=Ys
z>UJa}^QgTvc9BU9M^gJj_n%guO1#)focvtBK2Afk6ukY@fS@q41j*@J@%O)G`EVo=
zTcgh-Oy8477>CO$s~wAX=eIoH*h1C(zi!`=G4kV>ys-ELQDq7D_-j_p8fCV^IhZp>
z<z_MkzL6If(*x7KT_q8y$x^HY)Ya6x_)CjKJD=IIXX^|@l1~%A)|`Lv<KHFf*yE?9
zQ3pqc4YDtKxD0>{9`m>To}*kCr!hj$C%Vp-oS=-?&O+KN@8efnmG<2$&%WuE!9|SW
zj;AJ^j#KT9qCiVd&DEAcS&xu5I>9X`=U@#iONZ_Cd0BSRdG!aLah3;L>~GiCJPtT7
zN;gPF3%$BtMP0yh!#?Q}hKh=oN{equTQr`Z7l|*6s(2yJT@`Rw!-Y$7^7$Ak7_(}X
zjq4_%(f3|J;>0cL_oKYfWqK|Hlc6%nHCYt-eGC+6{h_$!XF|biPa&=~IWQ@U;!b#|
zx0~C^l+oT|!vS@6kb`P<=dDuZ+YXuaXIahOM4iW*nbS7OR41c}8TUskhlaWMD3$}v
zn`U<UThTF;h&&nL6#5|}`>=}>n~gy&^@0`mW)}TOcxXz?%fMs_ChelFtlO<aEy^)!
z{lxI^Gj9;ib#8fi9DsJ#$x5kKmpkgp&IxaMPA#_Vu22RRtKv3I)#9XilG-16E^v`N
zS~zeQQ6=#kN%5OH%&US}E>9x+G1OTYOam;SEX~p*3AV3>Zd-kPn48TPu(&{t8^|c;
z2Yw7%JNlsTR3kv4239*OkPDMK!bM-~h6=Xw;&f6r`f#TL=0vIX(T_v_(Lt2x2T^fC
zd!d}nZ*>72w@DBIPA8_zr`|o2_KO^CS&o};R30(Aisn^5)6OCUsN-D7`C|2&=+cSs
zftaF+8qFhBd+*2;OwxXFm$%++R}6`F^u1<zl=+vrj;_YdeN~9qQIApgi(y7DLT|H2
z&XXlEPYx!WB*f`!T?|EXV|zf+6|JLPoTM**dW-+r=Q<YOaX9{AZyF!)WQOVJb+R67
z+xY`2D|nmIkLhUx7wvveqYpRi3XjQz)5PvqCRIr>pM5<+m&PbT)YeJ+S?Io6MtZoZ
z&9Y~D<iTd_<M0%o-PTMS+elv<@NqGx!z<Dlf6;Cy2|UWEc8)oz6pw@62jbD>6iNKk
zNLlgR8_RP^gi06n1@8{MiwJ=t1+8de>O!--%YJtO{$3)b!27FAbkR94ns2Ps_k5KL
z^YZnOVKr9K2wsLlo`-t48J*8u9B?{;jQi;mwbZ9=>J^$LoM_Q}Orc&i8sl6Bz3hff
zwplCb+yF@m`LHQ}bT$QY9s@alQY{uAP`p9WPhK!B#GK7fM^SaFtGOqzfpfqUw5$Md
zhqb@ZU5(d%(X)mUG8|jvTIxZzdS`poCfEVXb1j5Qt>K<V)LdRqTW!cv+!q9bp*D5S
z6L7k8!lM2Z54C%+yiYb$D50ChnT`z)Pwm;a?nW}6FP2r|mfjb5@A!0bQ<Vy%j<N9v
z69GYkrS#WD4zVdD5qfMy-8`8(dbiO7Y4?+t$9iUAGY(~~l1pA~fwa~r#%B;UowVvu
zeyvq)r2=ayV3gTlc*K~gH%wuaYw#W{lwS{@s5^5oGuX@>R11Fk=3iH3D5!OLcb+-g
zJODEf7+GcVYpuj5>C&ki+l7JRrQ^9`%XCY=m&f6NWMnvHiRd7NCHIFh^GJd%0C3%Y
zEY(}+Ak}6*j>D(^!b9EYU%@eJw_b_3<-JSikG8*s{^Q%}D6FCON)JQ+gQ$D&i`}Pr
z%H_FTyRbk<D^0!IH7oJBI4qB_TC3DNlH>G$+zeDSIz9O@KK(UaEw#k;3WOSP<v<G=
zKcLkY<YPrCBMma+-I`_08#)o{^=vw?rET!2cmB;4>vyZSV{tlOcXh~n`uSk*R+g!m
zFBPuqRQaqtAnVBw8?Jz$>F(&tzk@`S=BQxT=&=mAfmnz$`sjt<=nKK%^b*X1eoiwx
z#yA)&8wlt9g@Xo*i{=dRdE^=b`dJLueY^;+!yc)moq<{W-u>G7+9Cmer%Tvg-{`<{
zDWEzR@eic`2ipIq&C{ze*LxpOVEYky#MZd+UlLHw(-UE?_zq5JY6DX$RZ#>$R&JJc
z7qCEhs65+PB202g0HdS_+Ra!L0ieeN00Iv-Es_o+kSbS-GAPCRn<NrbYVme=Wl(t2
zHv;Y}yoHMc&+1=Fk^>~@izUm&L+`{4`{pWhkz{Rd-H@2a)~@jZoqdb~5)#2q2M&;+
z%=IAvsH6abz%EGn3qRie`+ypLDtDdT(6(c`AWEm+!^Q+vV;fWN!QzU(^?N}g(CK4l
z;h9Ls%jsCzKBhPg5thoLqY>UUN3wf>9p01oGI(!s45`7ZZ&3i`1*G%u^KZ?5h_*)k
zwV!9E@$V;SxPd>NDzY*(g-@e9r1z7W;@a%9QVo7?kHpEiB;)~>gZ4U?8$`wC>l@oU
z^C5A_d<MVq%0tXLB6ECeB1SpwwcPPh$e+*aquW?H_p9GRV30@kRjHD=th8UG=$el-
zo9?s4xm?lXZ_U^G$IzUH>Ha<W^RFRyIeD^kjy4O^Zp{Lp?0;YNe~6QI4JHF*4*KNN
z<A`LLw6)?pcMS8%B$`@g860vg7UwR1eNi8tnq@sdlx4k+*FDIWI+Fn&zwwSO^1Yzd
zrKHSh_g*`hoFxX>WV2fWuLjTM4|aCWcDVx1d{|ERC{eDYjaM*tR8#0M9tfRN=gBU_
zCkL=apsq*Jf;ph$;_LI&?73r%fHW2#AjV83+DZdom{ccZ&fsi~X(rZ>y6Fc>(VS*`
z#81E$=Ul2$d6!+yq?suUqgdu(RToALBK_nCX2+)lO=5MEiE|yqESKLur{7c2Z^8y1
zvIC#AiZcdoa0n}YYyMM^+Ke)y_L`3Vygwf#2Q%tB!XVaXSx?7ZAlK?PfrrzqfpG^_
z#05FR$a<2}LB!dQHEWQ#$Lmj3&DV4CmlA;c)yaU6{E@Hjb<MYvKa1yb&(1kQM@h|1
zQ1o8GdsEos*<dsbGV8`Lq1t%{L5v4-Pp<ra>kks~cs*K?LTyX{^wQhvu-#K1;cSH;
zPjp`HpNdw>kN*033&=TW5Mn+3Vinv8?GR1Rx}JtyzdlHmjqp89ZFX1{aN5!dRg1m<
z?hZhubd$Y_1)M^y<j{h!-`FpmNz3sVMzFohIhbK6Hdh~dcyDNaX2Q_wqc$fVTYl!N
z_!pBuV=YW}bhtZ~JBjnJ=6;tyjbi$ak$11~5z*g+)-NIJZRdA4%i&uy#K6lYd|<JR
zrr5GFXlHy?b7(uah&#$g8&DvTJRClMel~5(6>`c1v~PClw#<q+Fq>mCGZnZm3P*xG
zurP>GY|mMYmj=|uVipM>;VN@<qO(oWGt+m(y}WoZ-4uN1r3R4zRW(fgVOeA7T=UL?
z>eXz>>&;%SJH@HZB3$<+c2E{Pw5reL!HiW@8h8rtm3c~gEN=j@2y$u_ZOX~doTBo`
zP7hJwT{0msZRF|i49ZvE2Ki<qUBV|}sWH-}Unjxmo>0WWcdHuMZ58!<tL|Q#<9{9j
zAUN!dz#n8TNCOmP4DiYsCsV6f_?~<lVGVk=2c1RwjO$ErG%}cMi5SjbJqeaQ+@#H`
zgzUykf-TNeNAigPB%FjjoI90F5_o4t|9Wl|ERg>AGX>D_(bS;9LzZdQZ`Y@pkr4KZ
zj|Q8Vkd0T+1_^+@HNCvD5kw;Fktd_4j);`J1ma@t4{IhGM7=q6UB9LM;U(xhT@K_x
z;`LHyH&(_v^W}ClTcd^vKPJM!nDVLl%+ry4`@zC+2P2IxwLv%xv3?T?SJA&mSeT@Z
z9)lvFF%Fu@!If0&#@C*9eQ*K&v|`wHlcKy*l^y4D)4>|@hAz`b>iNj!+KMds^ekHC
zIwz;=zRhy<T|-AUIug|iwn6O#r&hu9@vVy-Tja>~F7oo+&F^xe7(~eegQvy^1dOa5
zqO_`my5n@ODlGX2N3;kO<~!dSrbqvX3};DvS2ReMO^0ZCgT+%)dnhR_E5)y_!98io
zL(MS+b8Wt1jy4g&q>x?`y9!rLlqC+Y|6H7b`OkJB2TGOLvmZuqSCZ!77K((+{1DT3
zk$C996St6SRrKAl0JpZW?J|YUdD8c(&090qUnxI#(`7R-$G%TkF*t}wvIb@4=95L-
z2dvG?mJ9{PT6{N}EP2fYX`TmR%0~jPv%wvgt`bD%*CxJsu<DbCs)8cD$zyj*m0TC5
zYR$#H`6|=mbvt*vcQj&?^Rx%!-*@i*fgEJsv#m}Va`1m7EwAYjMQ$F%ClAXZT`FwD
zNFUam3+79EZTulyP8XblmsZUDiEIGANWMxeUhjgnQ8IO6!@$uPwO+1hVx87mioUxh
ztAeB&4%JN@lQSLz(mR*MjAZz8{b7?NdFZDs^$fCnLbPPi=@Xk)mtj9g@i59c1t;6p
z<zXiD#a_Cq5IMhRz%as|!L^xSKf+kjsN-}Pi@N-&E9Cj~3vM;JUChq6{Bsw395(O=
zAIEX`yGr2i#~MrWeWhy<a1y2*S;(QEV8r%BI^>$AwtU9L_g?0z@4mzu6A5Q%dgkbc
z#bQhI8*JcWNJy^0o>m-ptUvdXk72=)Tja7}zs7ag6W0*#h+z5l!3Y_`<;4HjBJdRo
z9TDIgN=|qIkj5(_DVn_e#D7ZxCA<PVsevm=5;0(F_UBo7?~D_UB(;cH2}5Q6@m3Gh
zb5SJ3<-;_~-}u|W6UiT(!=t6y7$I#*rHhLc$=V7nv)f8(YrSqREZa3gq8;>r!UUee
zFK=Eoy|{DuyCs?HzR?wOK%o=>8k5@3*vIa=hQ!KW?0seKm9<*AmBF~WF7$yBB=Onw
zuPz|(%SsYHv#`4jWZ0UmhJ$t+k;VdG@!{@m%gMhS|8@HTBiGAf+3-R|3r2+^<vZNt
z-!Ue_XdiVFn3Rl>N*<dPvdz@sX7i?^!vro7mgy2A3-wUV`<u<yQ(<ia2e&nFiStw&
zhwf<59USGC<0wC&ry?cchokoV*Op>N65C<ihRuu-9(ZiJjJ8F6Z~~!!U)AcvWG+^(
z^##)79ar~iLz*o4+xvTe6I(6Dq4oyyFEl;<!@cI(&}Em9_Pan7u}-hK^w}{pVdw|1
za@=~ubZl<}?)&K@2^16{_-s?@&T#<ivtgY&h^ipSW5w({oaLhG`fyJEx|u8Z)LBH1
z2pD1p-|c=oeTuLV%+ARp$L*v?k|K{iuLcS7Pbpb1wnCVs0s<GNey(oO;`CU!W9hy4
zZzQFa@@#yaHgaRApVKply^j2gs{G(hC()0lbJ$TUw`l||fI~ww6J>S$P&Gp*o_z3d
zurQ{ND8Myn^Yc~y=h}O5nWtvc<;REHl`ywWjQ-=0(q~gGE~+j=#{|(jOn%3u<t(CA
zspHN|nD^H`p2u+>s!m-MDfy>Y5>=@poGZIQtj_Z-r-j#N&9$=E4TC5xQ#X0P<&L;S
zdpxxJP#CnxF*iE^M<BD49qWakPc0PheZOHPf3Sahf|W{TeVAAdo`KDrT%2NtMTT6?
zq|dZU1f6(Up#qS|{EK0Ek6|K3XU9zk$DMZ1y}%&RVsmQF{_@i}gjSHfcq-v=02cM&
zU6DP5D$Dhi#ZJgB=6z5rtPS16#R@i{r-L+ZI1gxWy%Kkue?qdpG5cI_#btmH5G!zM
z9nLTmpx$IO<RvNkHt%ctT29tVfI7jk6cI!K7-^`CRh$th0=v;)BQ2wZR$*nJ48<y)
zloA{z?EW%*zb)g7{1X|L+l&(ICk0L#0qSo!0cU92Z2{RP4z&?{!ZU&2x;)ji!`$&`
zfihBiOb<60$y1v;cw%`cHsMw_{CTJQ@oonzoA*ty_%X&o0Cji_`FzkN-zMTM_c2G-
z7cJYAnp||RA+rV+bT?gaeDcv6cjX#olqd9bCacS^Rt(~OdQ|XWph>Kf#(Gfy@inph
zc_e0TTa%YyIdM|$(H&&S_qNK^Up<Dklau5Aduxd}Y1KcIr2$nVQapw>v6I^BuxBU2
z8=b<eUEYp6qbkw_#!x|<%(1#)$6L4;=aVHZWea1b)nVaV^#T;2)<!?N)T+&SClj?b
zPperHd(Dw9$Qz+{fxI)no|42gy%`VwTK6iSN*QYZB%RH21jK}U4kIUYFC3%I6>`Lc
z69{$;#=N6A_Mcji1gzr;0O|-!M+%_qY1G;AtWD;GUNt>X_WkLv29E~w>45Xx-_4MC
zX>-GD2&RFYLsY_WbAd?F+d1GMu>&1^dFK^Beg62N^*|J62Jg?-;YpOX#X+WbV`+j{
z@L>$AD-zN`Ln7Yf($d*si~5=6i20*6)Z0A-d-YuIYsNB2+$#q&clGUL)qtH6h^@p+
zGLSPBpmrG}kKpLBoMa`pLih?&kvfIxVHs&HYYBRpVNH3R;8X(-XLmM9S?Jl3l3=lF
zQJA%_&%OpCiI$V+7iwwB$Hs=ADwawqUSqm$+fX%s-f#E#B=rNV?HG)$5ym*cwp&+s
z18p3;%jQh;Lz0EL@^x;8PfZ}@#r&Ze1Ai7Pxcd`rtz<4|%C_5>2SgG!N*lN;*mx^V
zaKc8`cebs!rMPLby-;%xx-Dz5zn)@?T(w~V{2{Cvx$R3J@zU_E1LN4pZzm2G*fP&L
zJpo#)knKNn$T>M-nW>P4QBR))4^=^jb(HT?pXd;CZYCmg<?sQ}=!P;Fw2wnnZ(z@j
zLJ&#VO?GvY<QS->V`o#tD_8#ul#|e;VCQp*9OY0;w*DVDeb(YN0N)DSUt0eym@hwu
zD0Pz)HWs8FX%AGUDz&ruQ5$`=C?Dc;@yl8`f2cIM_OpMABn5Tagdz@GHA(z_N6Yc>
zb>3CYqnA0Ffffr<DS=+iUs$~bzBp%l_0l;xB6suN%Xy#7QN}+91+47t2zlkxK~}OH
zG|w>>JH7(}bGWG@Lu~E1Dv(6j3^kqFedKsMFwfD;j95#^mimIGJh=!Ho8n-py#M=I
zSzX`v*Y3m98toL+IMk$A1E~gIyfw`a@ce*8|2h~n!)(st@Q~Sv%C%$jDwo);mh$x+
zD#l6<i`(V`UgdvpZ2f>CR!qRtR{MAQXH}*)j^Ce3Q7?bc%RQzP<6i@f$n1~Q%A9{x
zan0BG#*BQvrt=qHVPM&vtG;}pb-TFF68A_{NxG^kD}H*rVuF6Z&40E5xmlz}<7a(N
zj3Xqnd&W-N8)oO1ONG4qRPoaPuSF<Lg!`iin6pte#D&Ix)71ZXM<lN+$YLC;XSe<T
z&2^8MZydE~h9&)z*ee4C;SXN7cA!~+Pz^l5f!%CAyo3uSxKFwCFGBBnYV#m_VUeP*
zcUI8uHA7<)-j`&tffs-R2u+J7Eid=_j}lPwJJT$|^Rh)9sLVMa5PKHcs%@JCT5dN0
zy*I$0O-JO_;)tO-4W7cC3gsA1Cs)%p??s>Fgu4w(KFpda2?y@dZY?)HFIr>|l24?s
ziw&4;DXu*;O5W3>hJ~02o8r=QHldx&X%aC~aw+p`UZ}E57=EY~&EJ5VLjN1E`E~4G
z-3+drKg)%{6)677oE3K5RI$@x<-<L=vxb+RXGFSspAarc;3}Fqc|50tSvAj^A;bY+
zW0BN__FXMP6*F73wut^5zk=Bi_wKLKa#q>AsAhpXNs*aFM^n}tUc^Ws)B_}v)|M>h
z|MiEocw`#(i$`Ob`y&FJ#tg(d*UPen1VYB1O@gHIte#OONiYF@#<iUfSIeKgR$92c
z{24nGER*8X=F><>2LVFc7Y9?TFSmoiw~w38-)qx2Ztl!51qVL5yw(}%0DTB?S0x9S
zddv6M7BSXZcjw{4Agg+lrCVkW3Dzh)KuNBUucTeXOwD);SD%8oK5t1CE{nm{@thm9
zUs%Oi;R9jGCZE0S_|y3<RKFkODn>3@mwEU~V7xdtp^Jlt-I@J4cmV7C>;r@SZBZ|Z
zW#SovqajeJa>#Om!p;<v|C`F)%irj3xjhW*1wf}`9OA=s@WZaq8TKV^@aM*!1FFIJ
zoQrnM;(q5`#@Se*R=ag<3Tk{2Lia4*O4?w=*adt$c5UMK$B}!36}+yatH`!DN(+Zt
z>w{YBF0x(21s5G*4-#Em?ViWC^_-+u*RgYB#h@MfR&Ik5Sy22?g?IPwXcuv;I|kYI
zDpEsVSZVL+Rn`kvkH6B+k4^Xgs-ZyB`b1;eCNnmHYu)V`ZWLv8Q$6|-hP`k}!$!f<
z%3Q_LtI~J6GNOi+Zq6XBmY0vZ@kFT0)jwYRQX1QK8B6>gC#teJ0mWyhGFFih^XuNX
zLE*8L#q8|7rx;Ajr_*pt<UZO|ja7T2@{YFVxoQ)=+hj>H;t}_V$#$p~AC1#5nzf0e
zUL%zP`PL--(9i8q>V{yB0|yjd0XThUj)8(#Y0iD4`29<5(mo{xnma^Qot*5}?R&xK
zn4F(A14LstdCA^pnm!~2KD8OQH5p=K=3GXi7LG#l;GtSK>9qu9w3G_Ifaa%*Zk6)_
z3+_5MZ}Nnn^se{v(Q{PB9@z6$#sW#AIdKQ81^2hy4%#mDM_JKFEjx}Azo?xYIt?}#
zEG`TtUP@<8tcTwOzJ%RLB1*2Bp+HkCLt~kbW1@zrKj43fTRzQ}!mm_*L$rf+0iM4P
z(4rPoub6SRh7$sEAF?Ri$okT{tHi5jv`W5l8K@Ufyq!*7y$kI|D1_ZnOWmHa>rY_M
zk(#7yXMq5J63)C;eBabN2(uVP6s`B+>vKwFir$Lm!#`5T<rC2fB~?fpIpkun`XlD6
zgE~u7Jp{;VdUD0<{7HdIS4%hi;lw91-DmV#CV}^1Gvq*P_>V|C#4&kb%$;7pIrp;t
zwcbz3l~XXh*ZhXssRH2JMU@o`UQh<YD``b7YCiEz@*i*e>xWua-GQll`cY#O)5+BP
zH|rI6*!{Zo3GeL1Ojp9xzh-9s!tGoG>P@{@*3+B?V&NXdz@O1oXgiOANT#UM_i9)$
z)IkbbFF+P?hl&KZL*clpbTThee3>G8xWH1x%#?~4bBkb^0{GNC5%RL^-dIWTG9tO(
zxtB--cO3ByUP)rarkv_fZ-+bg_|jz)2r^#3$w$zz`i`=;Q2b7N_Dz6o<?d6_i4We!
z>3@->m<-zaut~b4UB+yR#_*ubR#FCB7rn+O&pOm04yGcZA7HpExm-&LM}Vcx@#D{p
zZyN_{mIsUI2l|7~r>xCyt4-UO)=xx_JQV2Fj<(>}1d1ws4PRNC#vSK2c)6y>r`=c%
zHddfvnPLgLKE^#4IH#6Q{JaCX<#OP3M~wL@x53pkK8<Ag3Q3|6XH_}dcKDExsa}Fj
z#^~}^Ts|}rTOodbYnIkM)UN;Rc)f?>JALT7^%5g_?<H?c@{m3&7?s&mzq-KF7xmV|
zN&)y<c`}*MrQ+%)di62xBq{U8TU+dA#B#>dh>mk(m9f(!Z;uZgDV0OcGUhIi>99oE
z_dB(Ep8q;2QUAE>+R^)WKB#z353gum#dL_Vm-vveI^H{}ZCZH$+*D~3xDt5!?_<gT
zsa^lm|7+I+{nNEsb60uj7TB-A*3jhzJEoPgmZ#{bb};}FK|_hll1mnshYw%=+X=zo
zR4c&$4g7KM^@YQw)5#pO(d;&9yS^8PsUQ)}^S7N;kU%BLef--1A%3z#^S6s8>_BR*
z-xKdbM=<vNbAq($qc=@vb;K|m3IIn>mzq5%8ORAdORrwqIXXp+u%wc8g8GL)Zv?oj
z{>^UHvUE6!a8++6fDSFOFB)ER39IVR@Ajqxo>3ra^fSDkT<cs3@~WkaL_N?0l88uw
wE2qR<%g;CJhs_y*iV{7=2!GV={qSqb4(5%UAOc7w?zA0HSJ6?fQL+yIKL^4Jc>n+a

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/numfocus-logo.png b/previews/PR3525/assets/numfocus-logo.png
new file mode 100644
index 0000000000000000000000000000000000000000..cbc35030695811640f4fa4dfc55ce8e2b7ea26a5
GIT binary patch
literal 4883
zcmV+u6YT7XP)<h;3K|Lk000e1NJLTq004pj001lq1^@s6sTE*j000u&Nkl<Zc-rk;
z30PFu6~1q0h8cF*_eDTtR}l~p1w=svx8Q<t!F>h91x1^vtxLMNXKAyTq_(LstuZ1r
ziHT-4sj1o|YBaV<nr3U8q*0?r+y4yjkzr(*5s>)xoB#XH$GrRQ^4`7wJ@?#m?t?kt
zfCCOV;D7@TIA9+*S9ceoe|EVve%elH>b!%HS0OJzek~T3F5*KYd$Rvmcst3?4TF=L
zMh1HiW$hrssPjg?BeQx2mTGB6NuaU^rbPN9%{u(`iWl-793A3aprwTwbgoEioD}Xi
zjM<3^VZKEu(|~pj=IKIwN})V`(Ghz@96n~-4Fqua^pM6)y+^6v_?5D1(|_BM@8mhP
z2l?29zKr9TO(=>eq@;}iv38*CCdeUa%G^iA!NX>99$s$D4191{5bi%l!UY6bB^=+g
zSI?Gwo;^aCIb=l;lKU0oUAD<{mprf3{rEL`c1;u7-wJ|{<(v&VW(@&)#r4BcSyzr1
z_kU9_3Ouud-fo5<<ooYki4n83G&fnKvfnN59lC(EgHVGm0Qs(iM3c3&SXXB+s_)HV
zX>YnnBF}8VN#OYt!~4HQ^`PBWorss5L<41yNRwy3*s5&bi~?o(x{uZS*W~I|FNwp-
zmJ11~g?z7sbV!acU{IwrZT{mR$~PJzq)`*sGbf#Bb-;qdL*<1l-n9BGE7pH5<`k5D
zKftl^+C4rjY5=nWA+b-UmXGuGPiiS26dEiqsC%uwcLy=v0$KQWi9mxgEi#~_9U}PN
z3L=n9j_@C5k2avZHi&Rdo>~1HxpKiH)<i5GUB^_2`Nfk_<|>ty=Py6R$Hu3$yL5PY
zdq`vN+^DSDc-d5DM%7--GUbMR2BSf)z3Y@V0P^A=okrWgg*+zDTlz8)NxgSfNbR5B
zSt8&W^l0OxmX5Zmz1>p9%=_i(3xBVyT>lSB$MQDErbd8VdZn=M>k4LGcM}2a9+r6P
zx)#sG9Ua#1upLCW%DK3CFxwzaP^gmNx-@e9T2iabPC)(UU`Bt|tYPN-iiuLIX9wA8
z=^oXM-&IT$h3{FyQTcSz2$V;zS#^R50^b=TV5ynW(%!I<3$^w%aC&E1$jBKA4JOzm
z!ZU-C@5A4#QzJ1uyO9WIN;6)jzjuWCWHBp%{gO-NDQ{O00c&dyrS2!DSt6B<ze|=R
z9=4K>#9bDpHp$5IEqM%*AVPbjqWss?`NDv~6U+x1D*7nZ4SS-AwLPzxSTw3i8$3QR
zq-SS{0K4N!Q+rkQuTXm(f<}ZzvB3)(^HQF~%sZmP4D2o<Z0#3adt!L{U!KcPZD4Ic
z=6UY$kUov?mSnuMCMlw-D~JGNYAfZJ=Pqqzk|fk8gfB(C-;kUu%RKEWi6XPW1L?ii
zYk7_pr5z(;h=YetQP<uNp`M)?fLW(Q+J7h=Gqsvo0S&%1zVZ){r!b&Vogsp}Xr;*(
z!Ls!LE;~nr1u;QYMFC!YXkf>N_~dsN5!PaGt73yI@cY`RVDCQ60z(75`q90)(Loi|
zPG^Zg+5q`3K)&N6V`3PvJ|$`^>iuF_@6h?PqXWmZwMGr`_e`SuPv<5zBJV$)ML}Yc
z%*$0PO$M5nTUcqT-#5S&*H<VFJBB-$o#+%1<}W*JBKEB6eO*EX0c!1!c_|HV3`=`~
z>qa8DLpPrsp8npx+@!tC3}iMO7@WN4)QF5TXiM%2A_!T7D=Du$vuZyB%!u$l$7}=I
zV_gdo03kDXJk>|rwPx!*#7#3m`Cf@*x`PO?aki*wSFqbj?=%q@FgnC1i#mH;sBdm}
z5rF}9y~C!HjgsTz-kYh&_w$US`sc?4O=Cb;5J8@?XgB2*vhpg;h_J1rM1T{DG{7au
zx7aP{e+VM*F})LUeVNj*#k~Y#tb=9>WE|IVB4EGHy-?8S5Z2}AXnpQ3A{Zr6Ia`|f
zx4i=scTuAxzs-(+1lr)7mm(?l5`mA73&Z;BW+txY1{C)8$0#o!-ZQ={h`<DaDNF7?
zPIb~!)BQxyjL(sCw02dw3@z0l^czr)U4!nXi9o0b_bZ?d9~0s;pu31b!0)+8`?jB#
zOGG~}*D(CPyfQv)E(7c#f<^|&6*Ku99ia^xrz~6h0p;WT0)4w;X&}INaiDqN6S99O
zA`scJEabpud4+Z&tsF^9*PavlW|rPG5nwBjrT!W;Lj&#ayNd{LMLWZ4|LBp-xa|*T
z#_c#=n(+~~YkGT$U`75-M9?}*slDrODiQp;kO&Pb5wg0U2$mT{9(AQkV<-IrzKT!G
z`r#Vf>(zD>LBq_bz!I|7v4>AriQs*!i9kU8)gO$^x=5PgM~UI;J-dnU9e$5QXXTqo
zE)&I;G1lC^d_>RWE+GQ643>tsn+Gtm=MPK-OAn_`99lY0S-IgWO<N*Xk!p(wwdUz~
zQQ@uP{(o-DOWDVBCL(xr2N44GPR94k27LNnY38Rv&Q3-xy{n8xL=tyVUNNs|c9$57
zo;^HNL+t`O-syHCf>}_G%3M)Bp%2<3!V+eIInjaTq!se~JW{%g2!stOQFV}I4Crbi
zAb>PU_4+-_IIhz~K;USqx`#i=fZLG>1O(fjpfU)soxN!ycqo$e=i>hFBJ}%v5W%y%
zh~TGF+=88*47Y*^oSVBFwpd@1K?UO~)of|_az%XDM)I(VZNxoFgL-7s{r!WIUTq<Q
z;v~YYYuYi01BXnworwUKcC|``uk9s*W#><!&WHDJkToK}vae<Bps%M(uu6nGZ4ki%
zT~7ppG-lFz$`6m+OK=TK%X^2-L%pAnfcURtkVR&J`})V$YkB@ul=eKim|=V^Y_`hR
zyvOPEiiv10t_JccP&%~F<aQ3hcp&mvLBbhbJ+bwS(5!u&cgSmp2!&=uh_;&uLrNB?
zMEI&hM8N*u30}YR$A|ZSwhbaEG?R8`#qVI&K~Ab1Dq9&9P;xsEfewWzOV*sHGE&8i
zExai5(0&35NWfW%c`z-H9(fQ!%vBvp*RWSF`JQExfX@Di>c*Lc%~l6AEo<!e8Xz)-
zGrH)YsV>tsB8W?w6(E|qTrKx4$mGoI1VYjesYH0ELqs6lofI*L2EdXC9OO_z>T~au
zWV{bIvCOOh0h`U_HB9w%@wpv{KoFApW+4AXDu=+&y%-25W)C=y+|(nL`VC*3%HA<~
zlcfg_2H7KY59a{>35QnDjjU1xdG_K)%0m`CXLL0(0B03mG^H(DcOGXSJllwqcq3|R
zA{io_FSfMH@k|_AHXWy&LRlMtlPwPI{HOU{r)}#n5j+*i30}XCEr~!tkY*gELv?O!
zYc|0-m5ckA5a!-XH*Xgr5QL0@#TprKGU+(ZC=6jd-oXm^knj*`;&dA1Yg9&>Q2Bsm
z?4JPDo~qdi<iHY#j9A3Sr=;^S@kv73fKfOs(x7PrY2368W`^u2EnTizeF}ER43y~u
zTca-`nR9UX<g9jINY5#=IR9`R<op85)f#B4Jd{hsGLnsZ^~#E$e4%+8@eZ-Jh_IBk
zgLTP~HI_usT+;goB|oknFnN_i#i(m!ZCuDyxQHJkO@OW7`FJ-&0K3hIpblDJbnuxL
z$4l+qr=-ypKgS?c@V%3Yp~6SesU{3wlXkKQD*PghuNsiRDrv;Ban%}OtO<55&ya0l
zrfoQB;*4#!kK4`JO8yJ8Wo8U!{6!n(uEI{3VauU(8H2ZiwSz#t;sUSVr<Oz@V9gX@
zFwT2-M8>-`@Y+eD^Y91Un-;qx+C}fpfHsJ59Yk1klSH^?g$UOnS8kRF1kUKHM<*?X
zC+<TVop}+1@F-MplwBI9Px=t3`y(_H%?AX;I{`n&EG8==vxFq`;=qE5@PNKWdCki~
zYaJ6Z2Tfo+e**#GAd7{zdJu&8#KOgV5iy4*J}xPXwFf$yhKOG{wrlRp29iCDevwAK
z3$p?;qLSQ=KBzAy+s8e`&B|#X&0&Rn5m83H8?zI1@|ntrJjYvX361E82oeUvk9827
zvr*yW5(h#9tfqAyvJdhEG{jFqwsEl1y>Gr!j&VMIo_Mz2ka#{I*w>^<Y(bPb*cZN3
zhJ^DE^dT)^vpK!92;(yhWfCB9ur*wn0}eRgfCCOV;J~c_3=JO`DJk&8d8qMx9@jCD
zSkBwmhYTV3+a<atfiV@P9^ees^21OHXO__EMC6Z$L^DA(oRAgQk&r|(+El~Gpw=DD
zxw<(w=MxxRtLZWJ7*OizV!c*|bMr80+cjMGY)RMYjr6-k+&z6-UwimrSK?}R?agFa
z(#pfrXRy+b{<xov_i<`T>)~VSTh!s!{CrLd&PYTu0PpJA(uTq4!DK~`v>vJp+a2wA
zYK|%7iKBezN50u>>KUk&#rrt3&H_$}0f@7ohQ+)YZprs?v@04FyO3KI-n6Z7c)Sh|
z+c-WvDu~kH$k-wlj;LV(eD5Ff5z#T|d=~^|PPDFApo1Jw;S;hw@UA^4mW*A&0CB_}
z%W%5pWog3n2l4kPI!%H@q)1u*<G<sW_6|J5X+A181tbyB-j_&w2yNa&q_WRn5TA{$
z)#r#aY{&DRL|i_FNs@@VzQFPJU!tx}d|+?_{%*$c^ow}MhlD;EW$N`O(w(#S;#6|E
zSTOufJl}hij_cw0)GUm)XgCP687F~G!f}*i;lUh;t>AHZFYl+*LrhV4K0Y}Uj;1qG
z`IIg2BA&sd@}{5@p6@IJaE4$p&P^PrHu%WsTs|}+m&(#$(|40_K`&F!Up@?Z8ed!;
zK!0pv0G|6vgxa5_ve=<SvCe`z40~NugLvTkaoDp)CBiqHe~=MGDWwl9R3iMFW<LW|
zBK$*|H0xJw5rLR(s>9hJBhb=cuCD(R13<cWa0s&>W+|0Fi_<>IL<H=ekEq_WwIItq
z8WB|OV>gHhAk%RM;M1>VIDWoFBLcot^k-%O$v?x43N^bvib!Q11BeLNum6szP>Jvv
z23QdDi|a4|TbUV<wo)uDU#)#l0S)N^cks`UwnmKs?87m#VoM?*?{=nwh(JC8^!Xr-
z2z2%Vj;#yOEN*;AL^Q~HjwTrq0p80>)>Qxzc46;a%>X>~77$^jN`wmtvZTRhy#*}%
zsY-;eF{trSxn7z^fOF;zd`8~Ok_g!2h7b{;;subw9kZi=0m%PX%+f@a6mP=;l5I`|
zAv<pch_FK=0=a{+&kkf}fDa%Rvu{5M@-DU{0{ryn;4LhKljn~(tdXk`0q)vdEN21s
z5&>T#H$vmZ(O{@V_!mgrSDsbVC=MFB&?3B?gqXv}a37D=(m;f5m}GUTBPaud8E}I}
zm^ObmwZSA=A|gO@yp3q-(;5-b*Y4z~Li|C`8C;y9or*9SO+>&!)>SAQjY)M&iEs&O
zatk`72J59kTN<D>HlZ_eR3dyJjh<KmU;Mr85kX7q2odlM_hGh`Gc!O-c#640#*kpZ
z(=3TV1A;QEpq7tfb`@wufM(l417$Z67Oy^q4&Mk8_^U*?fcJk?hBpvvy~HelNs<C>
zbPmtI)RG7qZXiO<s$<j!f?=*i1WddqaQNjZ$WBWlFf$MlK*)pmTTk{=R}kT0OCp#r
z4NQfVrQrjpD+4;_B#7{rRxcbB7NijYYBCP0agP-uWVI$j{z`P@_V$(r&d1*y?%Am%
z&<;7)mIe&^1gxvOyK-s3z!(wq`Hb`OakmImMS=XWNcY2geFQrvS|!2?b(#7Z-gU68
zr2*GZ2ni`!H0e|#xN1bevj)QMI>-dE>z4+sb@hmh?S%SMV2~VuVK}KBB7DhYS!SXr
zYD9nm<Rli9%moqNwzS^qOJK~BaEhQ|K$;i;!|@OTJ>{^XJ)pur#j+AeM8M2=35KJB
zS^FDA@#bqpFd0Ikg!Z_bvk_qBgHWr<P}7KjcL+As#c}o#fyhhw1%WrW5y%L}Ahgm_
zm_XCurv4plBOMID*M-E?Obm1#j*dMH>tB%<AuS;_!;%Ql06R@>@PJAL+U4Lo$5)65
z=tmyQ#ou8|B?(D=a#7D4xNjl?`gLaW`|zFD2()FCD;GYFK|O-Q8#@`s4_ie&EYkg0
z({xM)Na2mNCdeMll%F#I`FCUMC4)o;boll*&JB)1=e{5>tZT#!Nn-$BC>3j91KK-`
z-N!rz&^`5E*gS>Z3k*er4$r~k`VWx3kY6KBW_AFbnSlI_R1Oio2k{Q%^B2^5o`mei
zEaMq~LD)lEX9lR(+ey7M0QOM<@;1=>lIG#O{d}oB1&e5x0iv|BO~{S_adzUpUcw-4
zpnO<T-LY%hMSV+4Ds2NnsBjQwm#K~P?6Iix8EOMspp^3qND}jhuTpm*g$SRQtNDu1
zkI!SWJ&Q?`qkba48s+v;S=b+m%mD`+aKHfv9B`lm_#Yq06=$wKOE&-j002ovPDHLk
FV1if`9;E;P

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/partial_sudoku.png b/previews/PR3525/assets/partial_sudoku.png
new file mode 100644
index 0000000000000000000000000000000000000000..4866e46def6302f2e09725429b5e588f42b52e53
GIT binary patch
literal 71711
zcmeFZWmJ`2*EWoZBGMqB(yep|NK1pHbZ@1kyJMq*0wP_~qLid`NJ@i9cS?811~%}n
z4eI4}-_Q8Q`{Ntmk8eB-7;MhH&vUK0)?71=V@?8;6r?b3-o1&0goG(0Ev}4&gggg+
z7H(VzS8iIqv_?Y0urL!7Q<4!AqfoN9F)_0=MnaMfh|)mSRB0nj(g_a@?M6<KMyp2a
z&5p#;M!E9&RZu7j-s5*@B8|asZ|3QV7ktd7dYgAGR6=6S_{o}A!H2>^7Q$N+^YYS<
zhP>uI7Ecy_cy}cnDOPYEPF{Ib6vG%c_fZ>ZE?_rp8<+BX!Dll$Wz-Kyx0GHI*e>oE
z;i>EDK6t?~e6BY+h2sCQb|H28coF7JKw9&t`6|*KY`g1_<_@3oJ4l{a6{$K=Uu1Uc
zD0If$+amFIN@IKEufb>=jkl>hjX`UD_lwNC=VG4MkhmCP7(0<4%?kB%qPuE^O3iXM
zvKZ7zW4u_f^}p&S%0g}+vDUD=L&s=)t<kw#ADR6fD$K~UC3c_Z>>VYPg~#Fs<L+IY
z5~VP?@Q%a-@8|_@S^YfEZa+**)szPu`eDL3Ki3o5p4^msT`rq|i=Umz)mTkcQ@DZj
z;Z}+Ohm2Xs-CZL~P2sv1!>_wAgI*2ap0TyFmMeao9CrWa;v)<VbnScXi!`E^NZV8|
znV3Y3izN7X_2<!ZJ0c6U7*2Wf9P9GG4T-ra;K??fOWvE#gs}%vJu|r3c*{Tcrg$4W
zj>V%jrVbf@<&k$xc!h@7tx|VfCE)5fE$EH7l#ih~6e@Sg@(WozBQO^Nt#IEyWM}ZC
zfV{u`IS1SMZ40SC`NAtbTz@SY9|7c%;iY<A*sE?f<BehTmk;lYH1#pIVFzgZXu2Oh
z{xC|=nzA)WX^ENQAe4Y6b77|K$kvYGXj@JOs{4^cq3T{dKZC*@>>3jBtK2uP$x-+*
zUzNK)@~D4QZR*-)B=yf1=&2OwH(o?M=6pzP{8StxAML{pI9eD|>zjqw81HUAy3+4`
zr<WmLfx_Ww@zbZ0SgzOY$o!Ejw)BPAWCiA!?+CoU+ap}%@NBl;)!~`SyK|4z4+rsY
zu5tA+1bKE-t2lIr-<HInmBecZI4H*@^FNvG*}97nPrTuCg0-;SnY8t;ThdjoT*Q=q
zJeGzb`V42@v>WTFL9L_}Rzp~hIk(p+wuRJ+6>E3h8YRINd&%!Lsz_IoQipm|ZsyHh
z)Qwu>t%*LRkubg1ds@jPa-}DB>2G}AEaC)y9Gn`Gx?|C-R6Qa_EqwgkTf<9e?Gv=P
zk?<L6LD4YUn!up%z*RbD5{o*PXHo>NrcQ#SntDEyGpI9dF9kk&!)CKQ85r(6e70M5
zJjx;IAiA}{$W7)^XcVU!N&EeI;q;J5-JHMMkA_%l7VA8udo^6`Lqj^vbv!7~bYba?
zP}E0Awcu+xp5^Ywz8L!Sf%k#@a}SIfy1gq8HJT%2(IfeE?I`QYeCfu`-qhXFMCN}V
zLcE21MseHs<r4}*npdi9xawLrhAM8!UE{3NRYGpBv;T5E2Q|K)`wRNoHPlCfrpWYl
zcWiNWukh3fmb>%S$#!E-`Us(jNZ-Irp?>3k#~|n-m9;-Mp1&_;kF-Ge+f-^T|Jz~I
zoS|;gcM9>eLM$b+!XTm4t!Q%sy>G@v$S4zAiRbX&rl_hCafw)U<7Q&Fi+<`BcSMeR
ztsoNEeYB)pjGQZa_`zujZI8<Rs+APvqu}kje)j7o!Mx9eZW9MK^?s3)d)xW#ae;l#
z&FrSTKOJh`lo;v_OC|XC+0kvFXxP#u`R;DrhcS7g)sRiS9{QQFWdWnEB0hio0rwqR
zFp4m;A6s;&#QVqhC?oNmaj@_e8mjw6Toq~X?uhGr<QWhA+-N!DID<DMIU{7iSr7#C
z9TuYs!;lG1MN(!8kZq@r3LOkP2n~+N3+I!rHeWF}F}F7VX>Mk2P-0MGUov6d(w~<l
zsIf|I5tiEqozvt<76__(SCu#}T_d+aarmV8erY7qV=CHsju>T;&uIzi;~7<I8X9Fv
zhic@XSd^1L>8I2{P7{u)_d`@kv#v(TlPX53#3;Kc$A5Oov@v@ShspdtTCyX9IZdrk
zp_i=JoongtkGmw8HO-bmkAwJvvhU_UXp0++W9kfulZcy-L*-(%a5f7ou`C|e*3H$@
zj?$JYwl>>0?JLnMF)Z0FjxpUF6v{sp^-DGUgrD9p6q{cpznJ<1pHH$ZwP;!3Sm4<C
z*v#FJ<;jqG%%}KVo%F97Wol(Qic(!-tXla>`O2G_`utZ7huI#!({1KxzS1n;Oz(y7
z6<zbly8xPdrgn;by#eDDCd&1+>n9l2w>>$G?!LLRa66AJhsX1&FEQ`!Aoe)(;<lyY
zCqrx#&(0cL8y@?S-gX|k&E?D%%#&nY)d>r?3kmov;~UUul-MQ{!J_iU?2B3HvU#6j
zdZLMA(Id{V5%Z#16K}Tt_e>@%MX%i6j#yb+l9{paOqoxbPd<NU5oYwnvY>0`C2do0
z3rC-QiE*HDSyb$RY4#m;E_IjOwn3pm(bbZ{=efee!=tIgOvRte+@4uH?I>U_%q^(4
zkt#wjj5p9Qc9}>r-84)#tnE2_dv#@cnZhyCv3$*+&z<)z2@hW>COyU>DGQlC7d#q&
zAYfi~T6K1>HTEaTy6&3pCfwY2z`q!mQ8Iyad!VV0cF6c^hoA=O(38BTyr4XraguRY
z>nQ7?3A2r#8$}bu;}DOB?l%Ou-Np}6J%SH(mquoenmWd0SJYMqmaXO!H>&GTqC@f&
z6T0w{9coafaHd9T<ZH-JtKd7QSg_Oi&AH^++OhK{>9Md^VPjH4XkTtjU$?|G#WuR0
zj%tl&iOqpMb6xG4AXX{H0_x@s=#RqZ#zhBT^k0|zN{eCnc|JWUcwQJUR?zL=-P=9L
zmiQ(0%l<R$2Ig-?-z>g;r52!0mwF(DNA-zHLws`Py{&74xx(SS%_p2viElqj(u(JX
zxRb2oFb7$a1O=0ExeqLzun)4&yg$~&=?KsxIHCB#ZT)yW-_YN1rdc)Ejd1g>F!h-w
zvfwl2ZyUac0h9q9c$>5%v~3aE;am|d^cwWu(rEHpvAsghjl;2qtm${t-YOIKN}IKr
z@%OqnzYfyj%rmT7goSVO&<`uzOR$B0>}ebikfJwJm{H(+uOO``GnIH2OWct=b6{l4
z!ENNYJ6u&#6$5{IW)97uc=SbiOMsKY!FVAs=S7a1R^m5`XBA&$8wqoXs(%E3pG-KV
zr00JAT*2G=@kWh6N!4BzwE?MJFs$FGukCfj7)@|Ep3KRu+WQvij<TOp`HlQ7U*%9j
zh|c;U;bUz?t>gqKgsOyl3i3(rUzENo>AIT5H551XB@w+QIu#&I{t-XP(csxUeka7F
z(8N0N_4oGo0~}-a2k;S62dag&IN8{E&ciZ;n3<@i&4sk@76~>S-ap{<=v0LLk*1N-
zvewLdka9QN8tLLpr-G9ir@-3i9a+}MipZ!KFa2*i_CM)|k{O^YkhqFsr`VxF`Hp~&
zrVg9D8`=dHX$S2eqZHMH^wrAZDk>fKzj8$kCa@mfo2rT~RIe$mrmFn-EODT6epM)+
zD6jU(9pwrepY1nI#6p&HHdU4PT~*diB07!Zg{^R{Dyq2*EcBw})sDDlN?%!ZmD1Y?
zpA>a<4e-Bnj$DTos#FzuX~6ST1||mvz6QClW{X(NaV2^*Z&noYINFHis;m!sz4J0Y
zsfQk&2}5Ph=NDdPUS&r0!Wtwi6e|A7c4})zs?3v{_f=nWYBF@R+u5i;EL!rsxuKTt
zyt1gR>hn(KWRG#ny_N;?RbhR{stm2(aVM(P#)xPc=0k;GLATTX_bRy=pA?e$@u3TI
zT7x6Ai^xB?5)b*k=U#QaJ`U6;T4iz5=Yz*D+W6^IYp=3+K&;DWJkkoRjK&JGMKlz3
z(W{*o_!f_wh}n5Yt?4GJye#0k2V?Lt){VJIH?Oj3>#1tqL$pI#a#O*3-lp&`D=sHf
zhb0~*<X*RqmwIP?TMm=9lBzrzJdP_2&p1bmb{0*#vLWTwrYjhmRXZ)^!hHIV^$(n)
z;Ga(y-)K1JW1SMfhtf6*CP?$s@=vCi+$><ui~J!j%z}1;*i(zun<q-67Q<gxKXtVf
zB`XUFZpI(S?p4eLp2jriKY?r>F76-7mMd@6<s)4uj*CUBi%&%AR=)C~!GYK~VTb7<
zh5rtm??s*WxAxtc7p_;7(D|Ee1is#Rar$NC`^)iqq<hex_&!%wHiY_BsX828RP3Mp
zl=1H#n>rsvK~knhf~20FdJ&(T)LQ!0PhR=p_I3v&3x`U8(-KYf1MzkIoAuZq1(KmL
zvmt^?;?q~4Yv<Qq+J+cjdzgjDKHeH@%9zN@BQb#2H;_=S+(o(uUR?n{LRX0Yyq3Ij
z59#W~bz~$YKQknhU!PF`zY$-d;0JNf<?pNEZ;(*IUwGih^#k(nPb1HLxccYy0(cKe
zL`6(S2K-hrv^O@kb}+SZ+`oD=9b7=OmDX}VLL#I^{9KVyrr83|A2d_dbkvlW<1@6e
zVtHm{V_?kUYGsRf4w8T?A9!hH?D&kr)ymS^fzMTt>f#f8;5FhhD;34XM;t8#sWj!4
zD8y{+jVZWU*jU)8gl<w$Pzcx?neZu#Oa8hY{3S?b>gZ_8$I9yB;=<y>$zo&woRyuI
zmzR}|gO!7W8GM4-!Ohz7nJcrk1NG%Yem_Ut*ul`=%+}G&#+m~0+-C+hPL6_9REUB8
zd|l4d*wyTxk*pnli3J2?MZCkx&cep}=h@&^0mM~4B{NrJOD%CTE3jr@3?ZHe90C`g
z|F3uc8S$?>HUGYo?e9DPdgnjyRC6%47qhVfLplomvt7S#{`<{eHwv&KB>!tAE|>Y@
zDp+Tsn*ywV_Dtxe{aW-Ju#F^U;)<%^H&8Of*A<`wNcS#(BVNB@xZYFy4hiWIl8m^B
zs_T`tM6_W-wUL&^Yg*TCN#Lm7!imytZ|w?dQjWwEvHSd8_lZLqHl=)19{R2_hRChX
zi3iO1)Vh7T_Ok4u#kb};F-5d)sfdW9@J@|~B*nP>NN_MoI{AJeu$#Q<o*!E@ewc7H
z(UIViq!s5|>#F$(3FYr^d&+KKG&h#<0u>yjEB}0F;?(()1S26+{PoR?SLZ7r1smqN
z@dD|e@28?z<bU6Z^pXVasb~_l=i1$W4fGCwCq136`^sN~fZL?~Uo6kRc^^uTa@+jx
zX+5fwM8dZxCM|yX*HAAWwWF|6D&Ar1_xbzkiy$PrIc`4FtA9WGWh7!298WV*|Gs*K
z_=;Ssm1Fnye+hOU9|&e#r-c2ltK7)xXcdb*rbLj~FQ;a&m9!+L<EP*z1PyHw^cp3f
zp3-mCo}bjJ!6o0_en#}jg7b>bnlf>;DYE~RsbndcKZKangLr1>2fYI03=bBq19@Y9
zth_2Z4>ahxN#3{2m(ia|(ds)#rEhpb{UJj9_XA*2G_DqntRciQl+dHKRGJx$@abIK
zS^;aCrXYx1>(O^=+Atg{TrGYIO3iIe;tI6-tYV`ee+Z8cHS`FMQ|sxhjJM4wvo+`s
zA>#?3X=bIH^IQlEm%b6^pF<joljE8bH17LUO}kLCF+aW@xAwoswVNfvw-?%#mZ9;L
z79YftzM(EKC}}PFg@IX%dS<9|8q8C;ku^CnQXKvmuZn9XrX?t~d^-@0H`y(f22YfR
z$^w-Jg^|pHnuk&|Dmn*ud)34ie4czJrc2>56_gfj^6qxWtfcZdtjRZ_OF$V9jS?@c
z<SO1Pf{>E1c$}QavV;CGC(ei9^Vr)R&*^D#Yr|1&f?)G@)c3972BJTnlid&S#R+H~
zy@FQ%wq>+c{)T!Qt%W$e@(#&%$D$u>z9=HxA0qRrOcK6R9(Icd%1*h+84@Wj>Mw4z
z{=)ZZlcCid9ZGg@I2RRk0wWkUzbgjc8Ih4`%%`c>p@zDROTQ(qeIL3TEA^O)Qk<&L
z`H|?CnF4Nm{8wsbZFg~Nt*=YLcaHvdd;X{G;r60k>(F48mzN(Pc)Z-<#SnE*9d~J}
zwwA|bV|<|!92$?B2%eataS9J+{i<P|Qp?!Gvy)Aa&FVva<I^4~ayPw7$L&g&P5u3$
z8m|*3zl@2Y_w0sM_f^s#lm*>p{}#{dVB&Cx*aX=h!Wm!AEb(@#+>k{XDCE@L|2#l;
zvt^<9I|JDo9zLkp3w><TSfqUY@odjtrEnds2;hH1CHI`J-ui~#>Fs@fGF6Q&3}+h@
z3FizV;qB$$Zp5$l!QyMR7%4Pnsz~Lhq}&Ry$4B-K38Igcf=@d3$xmYPjEC}lpKzI)
z_<DA{sxpB+6B9yop|HFDkcv%7O>iL9japdsD2JHOlyYJwoH%3#yms2?2fd_%I{TsD
z*;{D18a3T_j@vBZu<dUy{^GYc%4dVW)iRK-x<+jHCVz*C%d=Cg9>sM~gF%mpFz^C&
z%%1nTcYhH1OpLePA!S(@pR;Je5@JNQJtpl44zor3L#|x;ICM8`Ub;Ctho)@d_2ND}
z;m^tT?e}K_S!?`xUii8LSw!OSUjDk1elLe^?u-?3`9!AK{^EsyJuz-ZeK7ukp{O-g
z#3L_#G|C_}Ufy1<=ZR5BQ80#3oS7nJkM=sA+<~)uerR*JwOaD6r(*`+P!nu1XJWAX
zZ<Q4$<yQ!VyEWQS_ej8Za_@ryxXoDR1n08Vb9eY0m!`>O_X|%C#-t{hAk0tn?h1L9
z^O*IfuC(8hb(?5#K3>l7s6E@yj}Nh&4dOXC7&gSV8Papvci&rjvR?*0b+`HX?e*kG
ziNLCB=!C9^Q{J_QE7|$n9a+RYQ=Sg<llCP-K^z;;W>t)*8k(n!8)O!6Yln0lf_R)(
z`WrQC_6K#6&O6kjgxXQ}Ubti^BxbEduuwpc#KB@ZD|3XvqE4)RfsM0~aE8a3tblzi
z6MR1xp2;&&Sm?YyO4I*5ojU)#`xLF>gV#BHQJ6?ka6b|T-|e#CTR<gqeYCgSCuQM%
zPQwz-q@^4oWc?iO_k*^@dmZmWlQiIT5<5XWlifnH_VDc%`V#bPvc`O12id!J^PQeL
z^=oc!$Mcv2<whQd<!0-svr&dDQY3pFXjKJA>pxhYwEntY9I;bi*rH7nVIe9yT?7o<
z85yx+;iTfyVzPG)H#8Y~ye#+bHEiAxSi=#SHun&S8;Kk6D4sjN4RGNk{<j>hBH#1&
zhHe?J?A+3~0tYmF6qH-{e`&l{*Y;QI2|f|KtmAqt{xu5tSU(3lAy9X%l6dV4Z4qp-
z!Oii7&FhE+X_Q#@zC)`BkVD=N3dMnkZ2Ky#%;}z?zQLm~(s=3nbkWd44gZC9Y1_wj
zS>%0^T5KWM2X1TD>b>6Y-veRupVVj7n0L8iqzaQ}%WkNP0+spDoUP4m*6cejPWSRy
zQu>C5o>SC8<5}`Koc@ZaGlBFCXQ5Jct}B_w@l2yNG)i6DtKP$mx5|jEio>?Z7ruY5
z0*1kyOMid^;^0QJ$_ovAN()x>rg>GCafFFntGp3jrj4(!?P_k>8-=ww-OZX4Yh>Rw
zpxbMnmbvE0N*&x<#a=QW^k-wB)#ekGPWP;_dFMjq<iTub|IAi@w(K4um@RY4f6Vr5
zu|$mifRI_+Us3q<!T&Vj{B4gW0u5h-=(Y0<;nVvneN@n+@&Q-14$sc{u$&<n^V`?J
zughj`XGP1$u~YJd$$Y56#rq~hpLmO}!uxa;qWD6g_$H07sbo4CV!Cv^{WcUz`P*N2
zfAISml<JgObJd<XUPYS(TLye}|NrxoC-GmMr%?0&<J;Z#v#^6j_J4YK$o2`)ale@;
zOJVEra!l0d9`I~zZLs;@hjhu>+}}-6X`ya+l)br|Z$6S*kw-TdBCz9MKK%;4v8?vg
zs-RDSKcJImN_TI4Oap_MTQ0%j2k$E(RC)4mT9uBUmt8q?^A~-;xbK@698P-pDQxIi
z7QBWneG1|D>F-<Dl0%2|XPgK{A>La>Mkj8&3&z$D8(GOKz>u*Ble2%jR-zCK99>MX
z_k}VJdh)FUK7~<-BhSg%4Vy83dUKCDu6-UFvJibQ>|HZ#Q+rxWf55*{PVHoEk1Bm5
zTzuU^$5IIq%k2}^?!7iV4$m|9Xn*a-1k}lC;LWc=2x}#ZFu)+V7lRaSf6rk6JMdbj
zzb`;yg1ii#YvUzb1)B+!!N@itm5{x`sF9bWJCV*fz7tZ@%Ik5k)-KGdq3@Ok+i<C6
zT-rC~o76gA3-VPe>fl0OS|iG_*zm`se)+K}-|hp1>;0rZ@H{*c0qMOUY%#=a&a#fb
z5cSV2^uAwTm5U>WiYnk}mzdKQRuwWnBVwk+xXt#6zXyx&t#!OLb|uj8xo5_ulA_a}
z21nR6TM%3JvB~Dv(aWo-Z)U!V`C%C}y(M_^@pCb7VVXV#Sz{N6yt>Jzhsg&Va&e-f
zdcKH-%DB)niy?)*wGoxQ%_#_pW4B1u7=vkH`Eq*v8wK%tpA4(Yl@qXQaSv0KA;rE*
z9li-?0la(NVog`E1t^v{$K9Af;JcD_o^4#Sup>>#47HkD<}ON4Pgf!n03O)N;BzSm
zvt0k2g$x+;M>TL3vM<j9x*%nS%-#nPpT@=j{3gcyDg(YZk59*JYMEJTU<N&(q(Zu7
zWWFu+rpeNg7l&PC))vh)7M(1$(y?pAmGY0}eEycpVz5=L-7yF#6|>jmrRVFJ8In93
zcjy{esD7U8HAok2P7F-l&~Hz+f2&ra_$JtKf;$=MB{B|_^mBV}LeTK(ZdBG*489i9
z<aA!s>?t&sT}NZ+$R8kJrtIgS#OCW&*Ri6_t2s8`9(kTTV2+h>21NH1Iaw}6xOc5$
z7*`nNt8?zsWVOe=sh)hl)uDWzfHeIO*gW-MljWD-i*3fMl_<FZhK_xjvr-D&JTD~=
zW@+>>kXY{opQS!!h~A{&7Bl6z^O)64xGc-(<4@lDOOGE4!XM{{_49`0oHX;)P!h7}
zzQ0!4E)<Fb%S`K(v-q|2@^i~aSc!?tF--zQAY8BckyOuzcxzA=PB=j252-PJ1r&kX
z%z7A&mvn2X=TixNR=p2uyeo}1Bn0i*j|ASCVZ#sttI^_~!cT@D8SS+Yo7-MVgOf9h
z2q>x>H~n#`MGGF6o!CL2$@Nx(W7LU|<3fJcZ5`D>;CG{YfUrMOD^lp^98YcvqTEHH
zESvJ&mB){0Tv_F%?Di9P-(N{Hjv0vJp+Tu(d@Rh9dvQPZwrqs8JCaTbr+n|OK2Q<#
z%mdX+1L_B3F607NRxg!?*j-csRsK*Yn=R9Gy<{lvJ&ihK6Yb^X_y-Rk`IcbPrk(i?
z{8=MqU*D%x_<H%fM^E*NukU60!K&dq9VJKWlOCI$6EDDepdr@d()%^qgyp^j2Se)e
zV7y5VW!0-x^<X{voRMU15U%M(O@mTNB)lYTe}b^6?;>Py@m>YUIaU`(n33jQ5nQ2!
zF4xNsH1EB5u3i>3QiZZ@@yXCe#VRX$z1Nf$c^{`%qVyKnchXBs)}YwBT<4T8DzNsY
zK~4@Y8+flrHfF<iEu>=fr-Fy$oa_p864E#5IC}OG{-Eb>_ot<8+588p>yM4yuNmMH
zBHSO}*Xye=gzD|Vm@8=!D|XFLt<0)d_buAd!I%v>pB*U__LGh-V1VbbC~Tw)N;o5h
z?FHR#A_^$zy)P=-{UksmLl(Kd(p9^CA{#uP{1k08vUGX&QGV^2J0%9l!{S9svM=L<
zp4qac=tsaire1hQmUZ8R8qwH;-g{{Ej^t-j<X({sqBJN%`Uo$6p^v;!jXd4^AUwp0
z3+*er{<aXI8%n=JgVt>wY<2Ls++UAt@~yHM%1gl?+WOfbk64gMU)xTuAzxUS53^1w
z6T3PeSPnjh?UjC%57Dz`_g(U-S~lf*CQ^gibgV1V*d##$D35U57uy*&TSk`i0&J(u
zMKCkIg}(!0OH)PNj0F3)0Wt>rLE$)b@^Pugupj)JG|i7QZlH-s(dsA0A@$n}FwbwO
z&A}#s<q&0;u@U~rxG13o7Wcw?UufzaEEkm#NTaR{q=|pP1NVi;)WVII8YOJ=zHaa#
z6rGrs@*N2;M`Ngvzz(n$wEx2p{jf9}l+1(Foy;enZW^<J4JF|_`L&__uv{W^#CqKX
z{$cFmEbZl4PApC@*JqSXLB|?t3J_tGP4#}XDF|mwk-xN%dC)YsrQSfR7pJS#k3o6m
zgwUDHZGPvqVI_fZt7MOj3eh{5+Z{YoZ0b&9h`5Bw96lI$4I$aG0k`J!-tlhyoln7h
zk3DAY@;PP(NPXPjPr_MZW)ORW(Zi8CrrE0x?2WHaeg5T<!JR$a?Gh#e!=pDD)k05i
zKEDlQps9S1Ik=4o{`|Lf6!#Db%`c@aM>|>@DY{mDGU~)WktM3pQdAipDJAN<@;L+V
zrxaW&gp8DLJHo-Lbi{;eohPUJEDXmV*`|l~<x^5ci;C18|EYW5ypQ{mi`kE|A*T$F
zAM84VSi$X&GzRcNd@nEqZNv;>9}roKD&((!%<s~(Ya*CAJ)GhR8q=&b?V&<=4pT*K
zlsVXJ2v0lRM5vVLuvOUviXL;kfE&lj?tvqJ$VQFbveFR5+p9OT_&5}&Skpl5u}<t@
zL1bi=3fVUCYj~&_7TSENdagFwF_m2p;xtmFmq+biEbE_%a45)IL@bCZw9bok2Q0j}
ziiHycfmXOKrTC##vrD+_!$Iarz-jUwXD{U)$V6T%XevWCC#yLtfX819!!Z?Q35nS_
z722x1dd+L$4CL5EHZ@1-2ZlncgSpwBa&y7_8#sG>(i4KS%JkK>de5)AeV54#!+||W
ztG@IJAW`D3j9grLW<HRv<>@rc&7uO3EQErT6lezzz95qboN_NhxFhyUvjlARiilDy
z-7+Ft8-C_T+6ep*J@=&RY&Y6tdZ~OAkE8h*c@pK%UDT8)^mBYzj-s2rX@N^)W%)~e
z8H)2M8WS@5^s+5QAo@_%T#lroT*_?B?i*p%BvO>E@R$elnTgSN|I9tqBRnRd3v@3V
z1(>YPU*Su27=UG%(*sAICdxQW$hZ2S61c=q7VmuL?0SDYRXtfs)ITn?#W?+7%<FJM
z<s2u}{00q5lr+51T)2b;qT=QO1g0C^%f5ArpF3IrG|FIow1loecNM51JS*`seF(f&
zyDv=&BHOsj$QH;9EKE*Syh~d+ZThTYiBO2|qr>U!W|#W_@X+ceR_lc>mjA%HG(R6L
z$$m*fH?#Mp^hdhz`B9fVRmBj9v*@}<_%1^+swNvjWgtr~;XAd7QW=MzbaRf$oL<F?
zt<)ho<Ju|w?=$)0#Dj@>^6a~cm+Oi$!FqO|P8XSU;^EWD;hb!I%|A%$y9eE)ixcs=
zj&_XUk_0&u6P@-jiwr`TAb=&rEs?p5+ohoKHZ7l@PMyb!+vkG;<$=)y*qOvz2yeQ(
zHki}kOtTICdM^<cGqezU;2I-_K|RZlv0`lRR~x6Z<mX%PvJ6*%+$1g5x~1(%jh!|4
zC3%y(&D)Gwm03+rn#5Z%SpB$atJ+=b?QOF$UeT!y=Sg<Gz<-8ea}kPFybcSlXQ)tt
zb9s!Ua|@B(I2az!(NNLzu7!1g)fll0tO_;1yE7{v%aK;m1)_&45VLW|6g;C-@`Z7Q
z<OylN#i<RCJU_GRs<o0JWV9%>i|4k=T6wFA@D0Er|63S?mVDv+eJp26q8di8oU|!b
z)p6^*m((7bvrG%`#f5R9BeZE^9G6%UfJ@YInwb7mn(vDf9odO<`fjEazD1fTyD8Oq
zt<k2lHSQY1`CD&fTHe}-CNTDN(i0c8<s`>LPEKY)L%Hy)0WhZ`0339&-p;PHLijWQ
zHd+D{bGPpz!nuQQj~|P&5{5HQK}ETBbj=5Hp~wJ)j7KEr(BGC255T7Y4^D<4Och<{
z`_TAS8l@25>Fc;?FS`^GrW#>30QflqP}M~upfTKtG#8+hr6A!yx-`ze@^R@K);QpV
z!(aWf!ZZNR1vs&gFd}O}!~*}C^Y7ePO4QZ|WL=90XMYh9{uTuNclrU)XMmd?!4WIP
z`72R(r$OQI2LT|oetO6Xk<b4YB$>s;=fCD$D#HozDh{-MYRYmEanb4BK7{4|TM&dF
zK%De{El}()TuKn{6|p#oA8Ma&tzN{FzXef51Z00NK4Q*essFL~%i)hM1MJ^|{EqDY
zwLo&ea&I!UdQ2*?(-1b;r7Qfs_!+?T{MVeVeuE*u7vE8a;^LV95u{#o;J+8>KX;n-
zHvGS}=prfjd(OX6l0S?8JNo`DNbT={`(F$6pF2&20sU_+LS%{mn)7c^@b}^i>*3wF
zaNfUn`a)&@)=tC%CI075qhGJROk#fvg3uyF#_+E>%g4oM))U=e2EcL(pDKcRdZdL`
zY})hflQ7fvjT}hr<R{^ZZr(KGj_5-FlGk;WPOD70rZYH?;DeHw+d9PgPd77RTykFp
z<yw7i6=430;<=9th#%T7O6dQ(YIozQD48GL@!u@hi${-9@RCFG^}cearEmw20^rSE
zB}2hV=)=zjKM!yCEO~s#@E3fgBU1ZT%E@<kIM+l?b8$=uvZA6;W`^Eo^t%4wU33Nb
zm70UpPVLvtny(e_F!^6rA5M;WET_f9@yjPX?BPFID*}&U!#5quORVUz5rJ!Os{-1g
z=bfcsa}e>Od^~q<MaP$C4OPp189WGUh(l1Dj>bD#I1?dbQLoZ&ZB!r9)ul)PJj`|o
zB5&Lh=FypDCh~-q4ySQ5^`PKfJgcrtvSqS42x4gDKrrg3m8br&IKCR(c}cqb^hS^_
zXuSU>qUuBbCVNQu5X22k>mMEpjO1e|`r~gB=DP6TJe2Jcf>WA){tpN9?R*>~7?4MQ
z^Fwggzc`q0Db#@OdXfAOJ#^#Tf%75Ywf?5ewr2mLhtT|j0o8*9{foJRcXdt!PUmmh
z=6{9x|KHBo^l5yn@!aq2?^o4JiL4pVP>vPb1{mS?LTCIUvsPj1PMPDfvJHYt03_2S
zd?-VaJRh8>*bwXQ@7X6K@2kWe@*6bXuvz#X+i7e$jP@cqF(kJ{e`d%T!>x><=X@|v
zZ?zhPSP04??mROO#EDV0`ql1k{A*u)h>Xvfh7UHUrsDZstatP16qDAD=#LgSxWA)(
zT+@i$%|E)aV{_jCzMx2ZcB76}h_Qs82fAu6S_6<y0LpFbLIDxw;eEE7T-_ynt~>uk
z%^KJM9@B1$%@00UJ8!T^SM~uglqHypOfnk81lvBcLvT{tG*i$MD;og8OahKa^XO!&
z0bA@95yaDuR$UD9j>fIFb3!J~Y5BMZZC?T5Q4N6T{>L9}rfMd)TgW#PZNH&K1>z0<
zxX-=7MVYbNFkSD1FU-N`BKMw+(yp5CwL}nPLNicXBNfKXKs{D|X*Yr)m7;k%@Chx}
zud7Cdnejqil}jT<8BUf@49_Pvp)pNL*PuIg4H!^Q=o9w3Argo^;Asp%G&l1KodxI2
zNYZh+msWjz(rqzu+a=tfiwZXNmVmLb+-_DJBq)Xj=%dxZvd&bl=HUa06~n@{ob=0b
zV6}0B-x*pvi@gAQ_8y!2Ct=@hBd>2?Eu@_!{HNTnqeDGc+b$@u)deyD<irGqQA^Q<
zn*wmx89zjP^=G&BLxgdRykR5=aS&)*FPlmt&3+&1-3BWe(#K~ZNy`N^plyNue7ox#
z*)K%E$h_}$PUk{IB6erCMO<LXV=oA?p(bC?Gx*U*phRo<RlXEMATk?I(cA|ikB2k_
zflAHFIB$%{lIFh;MgKWP%>q1<iSaPN1%C=NQkPD;TNLz25^J4c0c18uyM!TW8*k+`
z8ZUY3eQRSdwSR`Y6bD_HG6O&2ye@p(sA$9_J^&DAiV0o^BQp!}HiRHxCS<LJ7nTFC
zys7`Q%4=`($cNJ}uKJ!UJukun49A924hRWq1myxA4eoNaL^S#ekkRYqJFO0)#XXYM
zxi<6w?Qj#52qF3aibrQ0*GwCI@-2X$5?YOw4zvf~<$7H9wSVO;H)RHd7=E%k;=P#Q
zpkSV(+9VRcT6#hKO~Y|blw~nt1ESfL`G^p2pqeN0rXb}xpdEeB+0qZ@MDe%H`qI((
z+?urDek}}#DjQ(QXApD%3GSkhttpT1bj#we&_h6`?*h`9`8su)R}MH@#_ddEoss-d
z2*}4r6Qnd6lHFH6OJ9ZWCZFFri&q!rgQEt4VXRU;l}W)g52qv}`cu;Kq=<;>EBjHq
zk@)M58an+E16<}*AaCQDbTwQ`6-6yuO!9c^GU?XhC>OSWrlUo5Lep~vYh4l-4uf9?
zYb1){baNo7m6(5~;GP*lwy`($MVVG^e)|2Kb1eNFA_My1$4-l0aR8ZT(Dy#wXMmP|
zD>UwCjMlNf)63}zvyIpMXa&5F-Q%zEgNTTz5zJ-~%;x!99W9$)-C?vNfPty<rgU5l
zUDb1$Q0rFN+y7FRGLeG<&JTw^A|ZDTw>x{*>?`m+{LHOqWUW`071Jd+GsW9dRnIXB
zKA-^!3YAL-fvu3#fb}Ie@zA?`<Sq~bLHN*hzD^A7Qyg6p>25O`%+X3Wyss-rGTKQ4
zCXvw`41g~M1w8y6^JvEt(uwGgqi+#H3hGbrN(Sh=NE5NCdSxU--7(A#S%4UKvba?)
zu+`sUWL%<pbG`|A!8V0$bSe04PozkV`GQ3U8&W2|BLg82K%{49n~Z^|e&^27Vlwu;
zK;c-$Vv^28!JA&LAzl353P2m&JfG#|GX3cUDntHas{U)3@$*Ledz3u`cJ7*=lL$fF
zQc@Svr6`XC<mTm-k!#J?(s!GWN~w}^_E?C0p2a_?#KE5KaonfRM!AUpo3g`O#OBfQ
zAx}=e%;Z6-K%&beucN`J_xY)Oi_)+=c&^#s&+RiR$ia5XI%BbtgseJhEd{A6$LNxK
z!?#%&4BHGu$<QN8BLJYAOCV}YaD^)gR0KU42!54_1RFUT)|R(pT$L|;9Ut1KX3M)Y
z&6lk~5ckM#<!3jEIk8cl&DB8dvR4J4#`4r4cTd?V^V5$blHoLwC)#D!dJ_RpdcD>o
zgy<p<UV%XA%kKp0_p<y_X#f-KabMX-T4HqEBinY@neMhGyecdt%=M-pFrW6}??W|;
z1a`l3idE#RWvdS)-lm=KO@P0IWmG&B{gILV3;m?|L2g}o9dX+QX$Bq&=pf9*$ffGp
z`qx5RTkVYbH;&Cb#>?#tT0<$L%JA0!qCImmbzV~rST4p*!6adC=o&WawSD>$Oh0F8
zwEx`ueicQ`d7;YNKP&e(Ciqih;4f=T5*Cz;(T=D?1gMru!l%v-%9<4T*J-!`@ZYY%
z?}tfDx5SvhGj1z3-yka+!y=C_%QOMl+j-5$k0;T<E}|m+J>NZcT&J!2A;->m9>vT6
z{dIDt!<7NqanIdOs*wER`V@I<qI@#=j9u^k&E^-@I`k1vq;zuwjb_T3MFd`ReqXzX
zg2&2i@&^!TmBT=DxXSJW!JOnK(W%efb~$O*fPMe0+fNB~V|rrPX(+c}s+b@=&U77!
zISym)_l4J<pXPIiDFDbq^i9q92w<D6lJ6lo+PALw-0pp6YYOU;E<)nBZ&7O&=4ilP
zn;fk|&b2Md5Wr2PA!ZuDxMZ>;;InH*U3nrZ>w2r8p=@?hXX-fG6`=1H!md(i72bWL
zPrU9o0-wam`Nt;_nTslvrL<l_9dfj2hi%W6Sk*(HjFIv=;&cffCIm>#>;o&CiZFr%
z=)ccm?*lg>P*jl(&PcD<Q~BfO5Rk^)eK0Utf=f5o_;zEqb~c#b9Cez4(#lAiKa(Zs
z9#D{Ugo3Dyx4Q#?QodE5y!ke-eX29ZWST5d0SH9`ihl6P$|XE}@(wtT?$d<_+tH|b
zd1d4FvOjn5zo4b&kzwV2!mGuCBZ^0|p!v}QXKs-QrB{Fqt^p+Qo=Z>utIeui1*gT^
zw3SYu0Jcx*)k_pCu+w^58>R>ByA2475a9-*vgqN`2Oow-tjd@Kp-7t8_k)hC=bu7^
z+Wnq<N;M-XtdVp|TMORxiyr}gU$ObK8DKav(RlT$xk(>=NurwpEK_2gBf6JKTpmV=
zS)XKf2pvp|39M85g|i?+B*KWZ<EPd8`dVW3Lya~ojTSTVJtWUK*-3o+IC1Q|4>Iiw
z-ax%dnCAlbI9`wmx|<&1v+`|Lf&Jv>3Mg6jq|l5R4!*023s$cAnlL~qF4;RS%4B|*
zJoY>7J|8t)nRUy%9R<|(yM#`1mij+Os4q5P3+oQ>1R)|l2C&6#uIwPFh(P%6?>n<0
z!i1HmW6QEl-mnJpZ^&=rfO*#L=Kn5#ka1irj1@WL>$$%&VDQZs%>V$C*ZCIAE8qVn
zCYisAVevLHB|Hu6wC;aH8;A6R9Fbf8J0CQ&;hHF<Y7wQseQc}5j6QKfKku_Dcrg5l
zkSnLg)|9Z+gmTM5^2x4b_mI;V7o(gADAW{GJ>FM*&RIDlo4>A8WD=y`ou5|rs#|O;
zohKV*pUi8oM^aGc!5kN*qQM#<BEEj`e)Xzqj=NfjeO+RsqgcV?Jm>>CIM0l)fK>)@
zP_>Lh3cx<qq_8y0z_J!EKg`*`dL5&CCCwYDR~OhfN5q~I2_qmL#7P<|O4Zq3B3t}W
z{b~$Akckfghuj1_BlD6*9g)7Ph%Ru=uWI&2@*%aMBzkt6;}s5>XQ;xkmrx|Mfx%th
z1f1?8x>h}FY%+m|l%ghF_K@@vEj;FMus|Rg#)N_*Oyf(3orzYlev?;^bVguU9QAD4
z2iyEgSVLL`@YqX;xnp?t#a()$PNwC0hd9^xHKMX4vKxr%uWN$e^V$<Y%)M}hrdxRv
z-@el8H5sy6#gijY5Fz&>RO_{May?%WjW*4d1YsIbEQv^l84w`i#bI0iBwSgqGVv0X
z%#`m<>oEx7%r-m50isQ9tjAZEqPK*01(>4rhv9%uKzLj0=Q2bT*Re=UzX$Y2r#qEd
zPc(9N38=Wi^bTK~0RL-2{=(3r8W4(G`IRdS%;y&Q)q`QddgDeaS~f9~jJSxn2^--B
ze;=JAKXPcMJk(JsM%O-I`L@Bps(zxhMu?n^c|h8OkN6~f$V%dzFbSl>2l1fBtjynJ
z5L6aF0P6P$a(B{wH8-f@$>@V}A^29qYdDHP^&OgFy0T@JZ8Mf2HcI+8l^$)Nf~x;P
zkJ&oyx}*gFD{RmFO^jBrK)Lx2N-qvhZ*%M5eYC;YTa|-=(O4KYas>5Yyr0wQ7ky@9
z#z1|AW3L7N8?mt^@v@hY2h=Lq%Ts`-XX$X4Yn|;&=kWv7zTgZfIKkc1a@h}T+GAN4
z!xk*Iha3UX)ze@peq_B1HCja`YWuR4-8b;<4L*7h%m6kc`avFp7lNlzOMKFeS9c^9
zNb5`sYmy{bxd&?6u+Ti!o|L2UdM!!0-~d{NjZ`e>7*Jf0Y?py7wSqa`#Oc3Wy(9ar
zoa{?izo?Q{_PN_@VGp|A6iB4wz040RpJ+wM<TnEk9dvIaVj5y&Kt_#XHe*@!vErDq
zteo}R$DLT;ngH>g7T!PcU3}=O0+}P5nSARqXv|{<Cae$<++EblBqY$#&9xO2n~fUB
z=!xRDKWlh1B93QfF^9%`dN*w0BXIr-4pTuNiNWcUMQHG4Ok|61@7|Kl=gQD`(x^u`
zP%$U8By(I$=#)$?`A*tG1M@Qx9A*-AW~IH$MNc1q=4Q_%AE)oWOlu`n4f0HRD!?7L
zU&Qg0ZyNlt72icDgdvW#!gZ$su&<d5Pb|>j=Cq|LweM8=+{L<9a=*#93ZJZT_5ff_
z37JF~>bc))!Q{2+F@?UIdcxB8`^^8!-w<}0zNI&HIZ{OdZC<cFpHV%BEr3RS>c$0Q
z>kE5=fjjho2-Imf{cFMG5V2SsIuB*W%d;gZ;hVG~?x6u+yJ>1c#z~pzSnD2#4va1f
z!svp^#$QI4H=0j=Z&ZB$r{azu!oLj(ZdXda4?W+Bsl`a-@&D;Gq!;rr|AQ!Si^oqP
zmaAyf@RUWoGD7%F_|3xzJ9BM6+M1mEbG~p;<cbMK+>bmyK<X3uyzlrm5bHZ|`dQBM
z#Tb6h4Ty+|&l?*+slV3LDNaG?XijO;VxqF-Lw1c8)4mWTq{g$NyLvzHhs4}hGZqAd
zoB>KsdxC(YqrEVS7Zo~yJzSoc1~pBFOGfs4JyP+(4VsFZ+oz*LHH5tqkTczK+mt6P
z0oNZE6H9E(?p{K%7ZG%X;b-%Y8h1-AM>rx6e>|MUD6ue~cTWx_cj|va3B4r?f<pPA
zSk)W^qPyl-5(`YE3Adeg9V-^CS2E(5?kk@W2?b$7XL|03JJD67t}K5du!q>oYtSXR
z&IH6cvx6GXP_kYZkjM9@`t{wx>P#sWxen+4lIbMd{LV}-yXL{Mp=i=|HA;QYZ3z*`
zvZ_@rBJl+kp2CTV=EV!_l*xh~rTYY-anH=N<;t?Ud6$4W>uaac`aG(@Oka^NIq({+
zhB0C_E_i&9W${*1y8y?4nW$RnxJ;$*`tw@fw(3A%E<>(%$<5^tE?XwsMa(3X?Ic`8
zb@?Vnoydr+rIXZQyv}zV<2*G&AxqaYb6)lv2uKKG^qgY>LaTRgG3?Z)YFly_OU1O!
zP*mt?9nwX48)IJitNpGP022Yqm?ULh3L`}L40cQUwOSS^<ZHmoe1gyrywAB9(30(&
z2Z*c$_2hb4eCD(Z7V}xu3njfT-zW&zW~#gv6J6ZTqSiu_xUEKA&xls5&{?axcNK+F
zoX;IvCIMn15Iz*-&>nezQV~$xp}M^UhRyG`O(FK%5RawK1YQ-wf?x_HDDDfUPPZ0B
zsrBN~)U7Y?x_99r_HHgyUM`tk0P?8@l}COnJ_KjJ17qN#k6M7u;<ESIfaN9#ME2l;
zcuD6!Y;8cyKe_RBb7LQ{@hcC7VbfPPL50G867<+?6L8Fjfb0#*a@N|KieEpPvfug$
z7d=oknE1{AW&K9P?C%m^Ma9b5qL?^qx*g)VTz3_F5AIUnah4lM0A;v(-taidJFLZH
z`Xy@B)cM(A(z%-IdUq@RvMcw95x$j2c+|>mwTa5zxY3NYd4bt7++iQ`skdve{i`}(
zVPfq1v>!@yqVW|4dLCj`;uWvveZ6`7Zrdy)k#1?=jn~P+fIdM?ix_$1hx~Elf@xv7
zi&&5TbOO|W1=BApVg%V)+*M)T*Q%1@S3F-GuPBO6_GbUMbKk|i`Hi*6Hhyg>By=7~
z%i0S^ifg0ZsT3s>aCVA^vQHQUGB+YBWJ|}a`3tt2iKTX^N%yi^0dqeBiXw)xXVp6b
zHosgUSh!Og^a39aQ&+tx0wI1rn1x(osr=qi@DwB4BPNe8>7lWE=QG01sVI!;?%%-&
z0RgsW=Qap*LR#ATJ04Jm6lRY?L4teJecggg7MFgGAsO7(*A5(f#QkrXcE)_~`6ljw
zsPhdX54468LPc4^#_c<pN3pL<kUlKw%dUZL2SRRCvI=%w1BeGPa3&vTUgh&90dKMW
zFxI_70TgpHHUb5(R8E%hLWBv-T%RO^@MNQz+I%%^>D^jvY?<~e5oTu|1PLL;wN9Lr
z^bhvpL@QrQ92TIHz5s=<s(33pHq}k^s}yK{&My`N5&ngR3_)D}mSq(bRpkmilEyAT
zT_-G=$cG5q%|~h6JQp&JSoyxP`#!YxXdyl-<bcF&F6^Vab2#%U=bW$)jkJb-AW5KE
z;!ydDSuMgFq5}bb8TV7A3}M&xnJMQP5VRU5oxf!aA!|($&RQMFj#-={9)>SYDQ+CK
zsby0KJ{Gw3Q}PbDbr?Hp%SDTJnVzKUTp}xrr43o|%Oa#+PT5#m0~g-q-SPGiq`#_U
ze0x(!Ge;U0jNTp5%)9WDxay6GuNr6f4!sunr?%l}Zc7h$vRY?TeoPmHV;@JDZ<3$G
z{G`^FL=4YMLZXC654?Hz;U}$(-g{d)<CiqKh@O_sB5AD_xy{`ZMp1aWovzk0im;H`
z@2YbR6FK|LsYgiX&#a%;k01+Um9QFv0tnFTb4CxEecW^@7lVpjmHi<u@f6UB+hB>w
z0kXUjO(2xyNdr*NabN~R;&EUemI5-c<al||67zT0iw`;AdHyv1Jaj}f@oW|S)AaJw
z+xmJ7=+^nGA?G3-{?n86vm4Q0RM+)SKaTHt#T^m=NB!L^1l#_rAIDcr3iQ?>3IEk9
zbhc<>3woUXZeyBm`MXsJixkoJ1OKaw2?$gFzry_Q&iuclW>&|Khhy^f<Kp777?qHJ
z^$gNo;matt90|X#$5H_DEiG)<PeI&;rJs6s2lFA2qX?#H?{M)dE?CoS>M8+2E#${Y
z0|GFJ{c!zXt&e#0FVNg(q*HMIb#(^ObtuHRhw^uW;>$)f8W7`sb+(%Ev3&5_uqAju
zT_JJQQxindhi9-BVMGxOJrWc&J`92~vXMAJ&^4l%hI>ZT**YjT&to~vW^=r^%#B;@
zCGkP;tFt;`wnly_V#dD-<U7~k`FA98Ft-9<%Z%M;fXd_BR~a|aM-0Boh@5CeSm<Qh
z97Xqw!VPdLEkJa5*(3)`22FDRH|7ErxSWeNLEm#3&?fkQbM*gr5YScmzsvRirFNAq
zGd_dVxSN@qXXPn9hHR*O7B6^P_q*e8=ugKXZ%9yy#gIE7Y4gfVdp<6JKmzm^%;#(5
z>raA!Z(3qRIIw^yYzmOBKS4p@cbzI%yWJ_!c~Bip%FoRi$pqM@Std5KLtJCw#(Y8@
zHDlb<eDfFASYMD&Jpv_I{P(r6&kh0W=DNGkxr=BPJNo2+mdNujl6vP0htAN|+V4;M
z%c-5=y}r6-u#>VMJ5C&#*>p;&=WOSI4fE&x@ldSTW@^%0qtIij#;b`oMwbdM^L+n`
zu`UB%EhuF64`iv7c3d4~4o`kZdQQ%^4dAP{z^)m~1)r0<FI;8b!U~-8zSe9+au;@|
zN1VHr#6Vk$;SuT+)$^0RKE)=$Pth*Hy8)lsx-0Y5dA(5tm6Qeq_Dt|L50hwShT88)
z57$~jyToMb-Ljnbpj-`t;-_C=@cK57wQf#b%nz!eL_Q}x0NNyK)<V6IADC-AL4lHc
zZi&p3#|Cka-s7-)Y`x+R-X~BL*cM+$BBAML&oo9INW*1*bw0ppxizbU$mlLQji`>N
zK#QG3SCZh!BHoVA)aiyx?Ck=7I0x?*iZcS%;9f_*thC#8UK-N(khiQoJy57u40eS5
z^b3q))S!Em%1z+5>#^I(lNj-w4ToMKL6aECZ?gV$rIb`PM)HBpdA3XUbW7QX-?R%;
z_&GDAMi*3Q_a5oW3E7^kmpwOVq}uM7dx1YLgDBHW*L!x^E}4Y2K1EID<{<&%#Mj>W
zal-)rlISOxzH;@tKj2_!AJNvl6*zi=9CVzfHCOlhe*rviu2qz?1T-3gTsG)nE_-Jr
z{qhKOozn>+`6VGrx*U}B?)C_r=HpTz@}A!XWQF(6c)ZWx?YyGTr4#_C(RBFp-q*Nb
z0L&sP4R;^kaO~oc|KOw`60wD^@K8hmKyYgKYoG<jpFY`B`5l>{cCpo%IwQY%e<n(1
zqed8!aHYia;V;j$uQxh^meZLej}2T$ep%aX@?Zjoh3}u)*1A)C)RTtK0DW=uZCYV&
zk=GZNs6_7fy;bo1Cv30Cr^5eC?Z%lDQ8U~RUb2W>xC#nG=|GEtu-(DP^T2qU8m;e1
zf;!+o40X(aJb(azFa&2lQ8S2==QFXyPChnD)*9?pM3p%o@@2bT7bB%EAc8yVi5-zE
z@16llNrG$y)Y2;ctTm%e^~t+Y3%IDT9T}(AK=O%2qB*)U9ncJgXs*@tlL{#JK0nhI
z1o`K?w1%_@k|MM9k6)Pz-Hsriwyp`+&?1<+i!#}=^CO$NEZwRwZuC4eDk0F>GK>xP
z26_{<SaC3F!MZ&*b;gZPz0k=T;Ui~8FQT(rD*v_VvQ1Hd<QP<S;qS4fub`Y|M>ekJ
z*OHxN!Zl9RtQ(Z}G024kO0kB5-L^ild+)|iVQf=lMbOD-Er%{Ka{+%nu;LE-UM+gn
zES>{0w?f^^cr7{Roa-iV!#c{pRln+VvzgeYf8h*te5h1`UO!7Q&$FXN%Mopq-35&f
z&A>qo1nzMc{09I;W#!_}ZX#XVm)vN&?#pRh<UN4wywfY8(FXC_ABc0ZL)n>@!nV(<
z7n`4O9+U4T2+@uKNt;@Bh&AS8haOU~FD#v2Ja0Dyz|Ic$*a52*_j_+S6weV&UX6r0
zmbmSwd(F(}6SGdwd<BjMu2FPQI3a360{{+<XdfQLl0OXYv$`+R(8pcPA3y1(y()UB
ztb+VxKM}O4mSh?1cg2xo+Jh>Ql-ab|j3Hrr!pb-iVZ|(r<(m2jL{8r3y#F~1L>DKb
zz;j{Q8$kTe?k)B;fnHXVfuJipa~eFM7+2LmhE3dva=8f4*SkSsjpseIQ0fHG)3Elb
zVj<2#l#ge^QKbOX;FSI-f2Fni$!-}+K=H<5@kbE<nv$^lJt<7-;}MQdiHYP(l$0O)
z+~StABMGny>A$-T-HyohNHfZ8K6CT0WPVz?3Dt+M+zfyAC2M`4Py_=U!S^&PMb1m0
zNsS{Cr4={HZQ;(r&ni&Hp`GYR%ezi%0d*AEJGjN=nyK4yddNUjxv4)OV`DDk@Z>Nf
zx^LX&)6$QI^SiG}Nl{_jxu1WLwMP19QjmVtbqS4@Wn;Ac;<={$q}#F0S)&kU;`w)&
zXEfq!cuetgO0_Q^4PxbPD2@i!3@#TgSywzfgU#xxHixT)_z7_(9PEH@28+`qeTgQ$
zrJ|NzLi@yng4WabgX4zxBXr5mhNL~~Lu}*`gx|$d&wB%m+Tq^RX6^Zz6RHp^)Cu)5
z=Tk&GoMfu<YaMkt&{5kpn&Q0^k?e~n0?K(3L&OhXk{rKR8kHS{FS#yF%<SbkbA^^~
zLeoyRoDTODL8XzJ=L&4VK8xQ5XWk74@%fNVa{X^f$<CxS>eY0f@B-qT&TK*eM$3C9
z;Dq4Tp5G<@v$@Zc+)VD>>Hb=zM+C_2gh0T4IJCtj3q(_jz43jBMJAxhJye{IEr4=0
ztaN#?12l~eq-G5MSP>eq&pmD*7G?z-3+gh;QvYl#jJ&#2u7tqrTKpunzKtvyd?hH%
ztSSpT)A&<D!4=@&x*PL=tbnvowc_v>d@3e&{Ge(@Nj1IABn}8WS0Di3htb(IR3Qtk
zCG`{`uVY_F<82m~#Oha0B_vW>Ckyl5LEixK5{>b@oZ1aG)tw?HcBJq1BDrP>j({BW
zDcG#XABqsYSkma%a~_fK1a)FWhyt&9E6DC*SIFtl4-azUP-`CqQ)4Gyw6HY$GF}{e
zpz(-WoqKme@w=s<<>=KOUjYK9T@4^9UJWazikzcss7`ZXQg4DQK$<aQ<9(v@CEm^h
za27KlF2Gv;fc@YUu>0KYv;#PxEZAB3#LP(0?HEFB3v49;;&fkh_mQ1tcpSQJan<j2
zWG**2h@7;qgQB<qgmLet5n83ihH8-;YM(?LaOTRN{IZcbq{l<GXQ2N`8q03V>#!hX
z0ifW$D>T)hd5T-dhNK?MtwsDWd=+}<Lo#^?@51fbmM~K)A>;O2B8S7PueJfohQ2{p
zVl`3Y<MesG{N*hGSO&cSbvx>R)>7+a3jA{ih}Dmb-7Cy{^FK<h#>`KaEJ<=c$#|-h
z&qa`ha|u^WNot~6WsGr?QvXq;^tRoC7P1bZyA!BbsXnJ5<uX|fqJGaM>?8*&5OU$Z
z_d<Be(@*M_*E#n%sjHtSK%>WjTSJ*BB448~`@9a&K@FEMZLN)=Vg;-J!`^!aRk<bq
z{)#AK02C1wkR%{k0Y#7?pyVI|lEDCy1SAW{1_cB}a?T=2vg8aVvg9Ng$vLMD8}{u7
z^o-}s{OaCY^$%~R>eN(?XU1o*)T>u_e?MJ)GX4hQMg$^m#P9m9oOQwClXrR;3KP{S
zs(mOkjgPN;_RPbL`h31<f`juxCx+wZ#}n3PbfwyeuI!1v1&rJhx$P$BRTHswgHSgY
z?&B`7)SG9d{Fo-Tm$y|Tdt8q<7$g<$kg=D|HE64<`EtgoWG+7f5#*|9OJ4J0kS#Od
z%NFqV&fZ{rgY8JA_~c{C1Lt<Hz72j*n;Tlxn>?A{6pHiiG?PX&@y@VZH5<Nt6<mTk
zR|+JLDZ3F~()UWub1et%m>+Kjx^x6pO)m)kb;q3f(K`A0+5}V4-1Lp9M7+DL8p}g^
zycf6J#^H);D?fd;hNlpzpFD1ND{%g6lO4XdNI}%v^Jz23JIjKLsbV;otH@Qqr%6s5
z3CYmHGuG)z^tjPqE=<~NSr2J$rpZ2|;sdvdF)YDpXk9aD)w!=a_V63<{`w72latlV
zuOeKag8514=<c8h1&_J!Y(k~*TGQKhZ`{3zQM_eo6)9kwX3t&Z?UCV)$}5BmowH0b
zgG}tBEWd2vSKK#4jz8r)<W6ob8I$}LRphK!6j@BDHHGMi*|{!$aXtR{{7;B}lYl$h
zI*pW~kxRKrC8eQ@-sb6N@1H`?eG-B#57*8H1y^}rBHn5C!XccjCk+=yT3R|Z?2cBh
zJ|Ri=?1cKSTzR?ECvBJFvNAP_?nS;>u2u84oc-oMe|!*P_LV=~D79>@lt3$oTjoJm
zbGcB4(Bqze!z288<$|cnbJGr#a4O4k;+7DX@V9z|!wAa>Y5X}+(!$BFLU}A8`f*!-
zkqvsk_V=$lxQ$S0LT44ld-`&(pCtR`CYtMSF-mU~Sx=GwnL^hLqbi3oyy2~xOw!c^
zxp2ig6y$RHZraR~r(i!5qI?g~<-94~Y!RfRN3Bv#+#%JxH0fd2CalbU<it6m;X8bc
zjBQFAxuga++j!@8xL@spi+Jis%T;hB{C{iTzV~1!m30j$`+@Dz!m^oYfxmOecijJ4
z1(y$<Q9tiC-i-9f=U?s>MfgR`I}e*?{2RTiw<e*$moXm#+rAm#*is*qd_rQo6v1m{
zIH}^%|LNr+BfEFwP)OP~FfEK-P17ptc`<~ad%K(Wr6yPD8^l-rg!rm^uduK@Y<Grl
z8+%G5DRXlc-{z2W&&nwZZN!wf@&%_^SpXbq`3!4NfrtXmI|-dYi_o4Ds~LK@&co`y
zG8E%UH-a~IW@^mB)l19jL&=<r3HQ$zyDA%uy2B2&bRqF|jgVN6nQkPzZ9A}}zHkoi
zRiQr);MSy1trl=YsK_wzzRGQKZIL>J9PE(sES>b;QM~K{u8aFFF#9VJg)40#<<gWX
zegy9`rpoohBU_gtHDgHJfRJgB_2AoViBMELtB3-&NW-DL+~zrU%Ziv33wZ0utx?N?
zJJrph!{u;?3jCQ%n}2#Us@s&BALZvR)n|}Cu*{T=*UppyU*t=j>Wb8(?{`|L=S<Wh
zsyz1h_8oKm;yZhf1Gs3se;~?N?GwGgDX4^0x>`{t2Fg5BdAHmy+KET#KWPY8>R_J|
z?!;U!n%Xrx&!z9%7TbdN(()yb!<HpO+4YD0yRWSFfE!dsSR<}^_*MG9Rhr+wVkd<?
z8?|iutwM-5AxCVu)>@C(s<u27(h2jUJaOnVZ`<>IV_9B9Afs$DC19~t{A0M%GaYL%
zlX0O<UFE5%k;DP*aYENJNYC`d+jJ-?^H4PQ<QcinKPFria?cG$G(>GZrnFP6=9tby
zdf;+P{8kZ-|8!i9jd>R}x(?co$WJkjyJ)pC;MkZ$#X+NUj4|I@J3TTr-iP-O6Io^K
z)XHMY<CA39SrGDv>WV#NVMVGXs&_#1TPe_dE`4ybid(oC0g9aSFJ((9Yv(}ps@dxm
zqH$2n?1|(G_Z}l|*eNzGzqwYBLwDu2{Wau051(+~@^QZssaS25iuJg})0JK@5PpEV
z<V_kn%!C!GFEgane|x{*VHEVZ9@Mon?~$^AHZ`{7w4KZrs970&|Cb~E(-C2HtiO<0
zAhuC$f9$t=OM6Y)NPdIo(Kq*8>yBr32@cXmZ3ll5foN~6GA2cymb|ih)}Ms={-s+r
z88k*HB&#^7!ual@{p{84TaR;FiI)tYwQlRSskJ*KkL9lSiraxP6@4UD4Yi?v4iZY5
z^?C{{y(k3-@G%?Z8)eOC=jd%oi&|&x@aljB6QnB9fMs6=r@cTRlPMF=P^6?MxiT<8
zwGM+F*pNhi_M{AV#M5jE_YrWUG~3m5l_HXLQ?FcL#ziLfE5V=N48iUoH2`CFc78ZE
zu@Wd?015L>q1L!UN;ea1rB$bp%xpx}%l*y1Hs4$<SKSsuPs*a=dmu_5SyBMU2ut|j
z)LIQoZ@f!AC@~hZyPDKO6v}V8eZ2E(N3bLdvXcw{wv+3zmQ6I*u3f`lo6|L86)Gs1
zeBo4X<~Qz73f7L1xu4hmFjzZCJ0r{+J8ONrgohAA5b+p#%h@RIB+W-+yEp??ij$#?
zVY@pEjy~M_GVFH7kFJ~>yIv;Ck$z|XYp$$Og;zW`A@)#(D9tCqvwwN-Bo(V$eH1o*
zR|8jk1hyu_f3{_&YH#;+2^UY+#~deROH<Wm#j#yV)qQ(sjv1jfdHuQ?W;$ZEf92w%
z400N-6NCSc!qFK{u5pf;st+%jFJ_(*5ok-4Ic|0-gWrDr%qs7<+_N3lhCJyz*Zl|h
zFoc*39tW^0zX3GgIZ;z_wkZvhUm77-D<5zWMD=G{MeM-S3k_d=#ye5#!#O}x`ypSs
zJ~8Xp3o@F(p;dxG=~YEAyKM{Z64ag&gG#rJ{w0}d>D7ZwOG=}k7m(sO<Pm*QkY87;
z-Y9YMtDvQ)&xN;&<-qA0aelDcKF6t2eOHy&4_>o=NC@T6w``oL<x_B)V*3W6nsMpN
zarSQ<IQ=QDDN3on1Y1YM&lY%%*Rr2|hhx)<=s3A7e?=tPV&N{-z}JyHW*JfYJ&dyw
zpd1vuDIh^m?t%9#t!06cIK%tQ=baoygmdol5L{{FxBvDi!Z`nwWk<xt+^ejX_Q%VT
zqB7jYQJ>c!4#GN_RKpcAv}@UGFkd2XIWguq)O8tmvpV*-%5cS7yUjKpv@5R!Rv+8x
z{`9AlVp9GjGvNcQG@VD^O3}LJzKnC1kjk+(lYMNEfSg+}v3bY&%je;SjWhQl>wop=
z*w(u~@K&|kf_~yQD=CiFE<l*1O*WB^v{+JTW1Zr~;^*PqHd$gp@TDrX?B|q<I;yJz
zL)*T_8L*p2CR+7uo8s=+Zc%Pt!#TKnH59}Vd-SHGOZLMa3>TVcSEac6kJ*A8T&a<I
z)lj7Pa$*S++ng+^-6V6bD$AXrkV><}EJ6eJww_$MzvdFs5dqGj(@lIU(lZ@HxL3|A
zjn<kX1&xbCpR05Xhjmkva3jBoy?*-%=VZRw9;64npVi4XgNHDi*8#%>q$`?D%KjrM
z+reUi4pwuk;a^$GvGSKt<}7RZoggU(pL8UQ=W<%_b;jE}M{sQEal7zsK|fy!$?nph
z56pkm1yEA3^#_}`+#x-`aHy-&HQ59nDI}St%aTvd$c*@|g`JG|wh1i2c3ds4YyR^d
zhv+HPx_YiV3(*6TVTvbS8#7G$YF?_zobhx8i_c9Dhr@I5+L$}^%j2?~1piK%916)_
zU8HH@aY@e5u0RdLv+s<h>x~6EIn?LJ+PV6EhzugnH2SJpyq%r#D3C4=m$~uHY%ds(
zyhlWRZ-yOPXjz{SE{^mb-lNcYjywL&!}CX#spslbi+WuMs%7=|SAC|*h5DRD;-IeA
z^dQOrBhx-)bAol(;bUy^16LuI9|^0yyUza6*aM=Rcvx17XzVjx3m(stjx>{K<tO~u
zT9c+26}#O-v_Tahx3s0WX}yX7`_M>geIsh&MfD29sOkykPJuU{*l*m$4)8b0r9?Xt
zKJSRZeoULd()>Inf(>y^5yT!h5+VsRrbYG3lQv-cd@ie>gn~7YNF%cCH9>4X@dCWT
zR-QCP&jBm2a9PaCau>l0gkv$=wagPFwtT9Yo^|X-1d%&gBVAky8+Z@PvJL%3U^zDl
zEe;)TJiBT3!GRHoMer9c3jQv<-rF6J1pikat826Onld$(=z`p7g1ILO{~-;v4|~MO
z0gLnhL^TMif_*vT7wqo}&cBe7|AU+nip~D}k^fHDuie$V8uXD}yNv7FH6`VDxPQ|n
zQ+=Xe9E;><@y53e#s51km&yzW3L?GF925GJ?n@;AwA+y!VYz>CeG%ktB;ayAXkCT>
z_NVT02Z)z9)0&C@mZSc^%Up-jk8WP^m7`A|LqNG5$A=l@q9Z#qF5F`mkp9G-8Y7hQ
zLT-=ncN$47kM2u1_Sk$-Ws)ztJF=p%R}{7HVE&~p|BVI$z3@{n45C`v$MswSWp8JV
zBPfji;Q9T<Eciu=Luy{q{jmF5C`5n-)hXQ<(7#Iq|G`H<C_#wRjgPg+xwgDt4ZDYw
zLxj|{*d=uGA1J=xIDi<y|NPr1{>q>K8fNxC)$a;`LPVfLe;Wl7Wc_oP|0e6-7gPOx
z6i7PkZ^Qg2LSSDX^-B)^dldg1X7@kY2+8=j{<~PepcMbFE*5eL6YTK~_Ag<&pgH#P
z%`XTcx8V8G@OVXnREQmN*Y1-~;q{ut@w!lQdt+gtA?jv9(e`L<0RLd^9)RE+=aaYG
zx^;o$vQVhK_@ptiTchfA*e&^yF23R6ymn_SB19M{wOuKtlL}Vcy-a#<U<S%RSF~dU
z97>_eREOXk+Q$cI%0OS<0|98pW5iZu!dF`Q7666>GSh{!x-C%IEJGb_Yvhd1fYD$%
z|H3<B|3}sAGBsA7f^6X*$gq|i2A!-bA8%<`w%mCS7|N=v$s=nF&xSwaQJ#UUfRu&4
zX#XZ*zdQZ?L%SkAP>3SobvQ2#k<Yz+zcq-2`sRyg*o4sOke&8Je*~X?7-Z0=cLfW+
z#FQ#iJPtn$$4V2Pgnym8MouI*TgqV*$3r=m(HM18Y@Fvj?XHdrIk3!%8h1ujq>Tb}
zk2%)Y0sNN=scS}5xRux4k3IaAZ+7tzG6WH?iT;Z8Q4`uGZ8fihd6C$oo`SFbNHrJv
zVuoxgwl(P&!KQ@%#uFkog*Ow!jlw(^3NY-L;2DEgmi%T#O3BLy;$mJM@!7!00xE4A
zdXyNK6T+&*Zi)y1=Hqw}QJ<*hiK%ykGx?rZ>mK>M7l&-eTL{I5)ksW3rnCu{Z<&%Z
z=XVDB*N1F+43g;lCtfW9+&Aec)>@C7>Y6y-Mtbk{WTsh{NZDKP1S#vdvZ~!qAMupN
zZeyMmiIJ%&w;(CUP?bwsfRTC4mFY5cefp933@}hK&c(5;Uj|1KQVN-)h3|GW@~={}
z=MT0_zMoac(AGG*+`txhQ)Oa8s$3$6r_yp7%@5sk=Hf;Hh!9-XvUAs;hF)bnd4WqD
z(OA4P(Zgp$Dy1P;d<xMLgAlE`t^iWZnEixtz88G|`mgU>5DQ-w)qX2~<}(&bMlWI9
zI`WF%J}v#<Yn`zVg8$Syd;ZkUm#0D*TMp;_#E|&}Br9v7<)VDI36qu0$z?tOXZD^M
z1U5Yoy>02%bZSr$IX}bmeyc7I$e_y2LXf(_R+dyGkTyrfa+{zttm<wJK2JZxdGgXe
z6X1Bj+vuir^*tpy1`aHvQF=5GGw-j-q&m3_iT)l*edyKny;Y<!#m2>`Obt6XzUDd=
zbKO{N7OztflJ0d9pfp)OU9mG~%6ZTSQ(c{wW4q6vLGG7{Agrw5j-^okHoD1?E&9c>
z#1hD5yDX^)`4*iP2U!vtB9QWlU9x1u*6JzpJvp-;HN&VORqdgzj!cv;FM)9xHk`Z#
z5(DWEk;3eBX{eVr+sx#bzUT0DrjD7K3rFQVXQCA?`Qx5y-vVqu)NJ?k<^7UDyE$y;
zMGqC;6y6>2|H{hz!kp~VPe(G9XZg!@j-Y)!uUDwV@ryi+wwV2qrR30dIkd4Pdmy(7
z)EiSmvY>9YE9kd=+#l!&krBhS7GqWeb~sTkP=R0)$M)LtinV2L<vHCfF~1}%dZCL1
zQC5TGk1P&j<3mc;ryM7;I}-6%%ni8+qx5zy$j~9J2;vH=Yfg+cr|0)D;ji7NK$6fA
zE+FNIx%yc2C=~@!KkYZwNkgKY(p2GCP>cX|uJ34<z3{V}cH9OXgB{rs>*nH+Cfdst
z>$Ql?TKXr497g+?GXBGi6nVXH=5k0C@}gkcraafv%;?J6FW_vIsCm{z7zGvih6lH_
zOSvdSbolw(&$tNnuI8Bb90v^1Pax}=;BL02RO~}dA-xBwhJlL-J+mL;yi=Gr!Bw$H
zhLFZ@F9a;c#Pw;zsnm>Ora&>qV1Peh6>X#o_t1uO8HfbO{tZZQDj;}~_}D#-QsBjN
zaRz>tp!HX<=?VcPNp`S2m+kdF2<m%1c2YIqUAlfmhBs0Ul=ka`>;LvagErQBC}%RU
zs&<h=T&K4AW8Bp62CWuhpyfSh7es0kE0>vaYclC=E-A2_z5vX*@91KJlH>m01Rt`@
zoJX|MGX!9q26cYN5BJtf8P>}xZ*{Fpn=*VmPrpU@1=KTiP!O~2c}^wFeJ+s$tQjSz
zVz%cs?2>FZ=KDoN!+^|=&=q0+E|XDH7oHlDO8w=0a1dhSQioE~KSXw<9t+521D`Ch
z5pRFQd&bZQ=5YAqDuP6=OVY+}b$z8E&cH`lPq`(qD-GU3*b`rfobUO^sdH&-t1nNG
z9Nod(G#%he)i~`1A3XE>g94eq!>uiD?278fxh<lvki_qC7~&%&>-*a`ixPt&n&qv&
zI(+>0_WGO%dKF?(sq>-2d(g-pT{w|G{Oc-5X()*J7gg)CA5|WkJ3c%Wz07hJX0s09
zh<)AzGHRe%@0PLWiC<Z)h!t^rZ633AfY-rqGe}u0dGxE<NR<-XgXT=9lG$v}&E1U(
zc7$syWSSbjudz;eC=VB)(%*J2Ew;$(x%29DM`|V9(iC+Gt4ykP`T}!zkVtWlBm69`
zA0V{mz|HxU?v0pINtsqoSEC~k*c^fo*NGWWFfXK(8<4<)*K}II(J{9%K)mk_ey;M@
zd>FZnd9y9^D2>Sw{%e^Z@=pfauT9fUTC{L^G}*~ldl1G|7TjunT|BoLUe`+uF8(&(
za!jv#MPTCbQ>$u5ntk7PT1#eb>>1ZpQfpz^4S!hvo47%!a{lqq$KO&j%TxI2!`<b=
zIPN?J85y-IjyV`?gR#~Rz*vh@%f$aY_aVTK0I*Rv!^*hwgVS~(4hfoTJ$ZpR*I$Zj
z$|x^qt3Xd_9!^$f8thW-5lHX;<z}l+RD5X3&?w3_w~EMkRxh|kGXZKA=8B^TG+>W7
z@s8&`+mp$zZ)s=ycj3&Noq?ZuNUPY=V2t6%(90U*V!}v)O%cJiR{N?cy?%gBisl-2
zNE)3mEWz$>BZNbEe|%NTbs4he>Xxj(aWy@MLdId9)d(Bcd)IRPz4}@elJIR?<L<DU
zY!>3%=(7kre+WFT9vu<Yrr&;iPNSoI9L1Ff2bl5ZApkcFd>GGrG5RH-mFvlc=UmO$
zyNT~%w7_w^(UK?=5q>he*QCTF!lamSt*fsl>qwXQQK1WYra*sgN8zF+LP3q$WU8~4
zGXWEZ+$iG(z(z5gDFGf1+F%_n^V7r(vS&cuk)jd$&Z8hVr#R8Rp>0<(D)-8A%q974
zHGsWmXg)??8gewb-o%T&<CPx_Lh4ZdUqENydJC>SLz2}A_=Y1`fTOx2cj~s}?X?HG
zwj2F6q^nMoP_*!ncUN%+{GP?Ltki`RGWo3>N2F(-iYdg{*Fy$yI<Hpivt77*+Z^+%
zq;kLT!VjcJgBR%){Z}!SuQ1g~<hcd=z%b;t)QiHt7gYzaw8`AfVRzbZNaQvAn$dej
zhJg%3n`YqV8spp6VYa1{JKsP0`B1L$1`zo*?C0{rwCTizuM}yPZoQ{-c^=0m%C640
z7f6JX0T2e$GK?bblS<*UhpR(77yl;9M$)_taPs*7Ro`sYiDowp%mcdEQclNxbr0CZ
zYA+y{Azvb%#$V{zpKl;m#hD{30c_(d$#%srv4B79jF#3Xw;U1ZCzZ|2=K;rnIdgZq
zD>LRDcNT8-`MyNA+f$c<IuI*gtPh)BKRbwEtIXtPZ%g~cKsF@nS(#f=>FNb%SAV`!
zJ8PHf6iAq{0SIQb_K#aM*sPCAF0b-i_sMc_3TM(*wA=NkT**PUE739dBkM7<@|TLW
zbkjW!d&G<m=EVrvpu59WT0<83IrA8h%$ECtIJxrnMv*=54LjeJ_nvI0S4DB!mL~+U
zs-e3P?*lGqS~Wcqf<vkCYpWxQ-zICWwW0#pV^c{n`O$YpeJ(cO7}UA065a|2BN~$d
z+c*3-lnhp{Pjowht6|O?Y~{$D-6*kYDNZL+J8GHc8tE$+#j9lh=)#R>4`c{0LEWiU
zkGkfhByZUbfw{x@pS_;()8iJ<%F3!8DL07fi@u<&(7CHuA}m{RwAF$uatxjs?c)l(
z<y@r5GvRmsWAA>l0^@C*$K{auvivjE2wh<!-cXy?tL--9h;?lc@s>^Bwduw0H~E5o
zkvp%@j_Z!v61Rte(rPrpLe%spSOLUuqTH>j%Qp|(B5_GSZfG48*)>PXp-n>U_7D;O
zYV<g=SIKhA{<c^9M9G2^32}{!rxIf$C0FVpJ>YiublM*DIyV*ZpJbRmzpS6DrgZGH
zJke=ayyH3_5@<IXKF|HW;8c7O6?I`t`UcpQRg>1%CK)5{xiQ~HI5e-EYqc^+iI7w+
z_o>i<c7GRFc$u7$LG5=BK>``DLz|3q?WSN?Yep!rm&|gM-4f12D6?%Q$AJ43;5>gk
z!uiNa4llv`{0>Cocuzn<kTqHq<x5mJS{`WR0{hU3J$+?kpK-XdDA?vKv6qQ`Znw#(
z?^q;5*p!q<1;$o#J9H0n^T*B1)xZd)2i14y;|L`Zi;yHi#Jc7Od^BWjPT`2*J%cQ4
zSQTYzw8?On1B<i32fuM-+ZX(DZuaE#NE4AwfJ@9JHM|6U^)nnuzD^jeWvJ!VNo2~L
zLnAc1Cx|*9QLQ^+i_k(BVZh()rebzx^<w+PD8r;zwz-h;?uffy22cm0!+jZdXscd(
z_t6E%Sv$KiuVWK{=JjB0ngZS>$WasJPk(t$TW2D*W|bG&FFtJ}XS*3Q6<n4}KN<sm
zx#1D;k>im1!rx(@nZ2uSt#y^a{2J;5@`b6xr$2PeLAs&m4xkJY)unc!<|(|2vP8><
z675SZ@~GZ|mxlV;2V-A@K$nj{-Lrd-As1VTU;9Gz@?wS9Mcdo<%kMTkN?4||#AC!)
zcrJ_zmRH6@2KKijl0N?S6s>3><9CUzaHEWt@pg#GCtRmNwtJiR@oUr-P;8Sogd}3y
zHXnDjksZEtq9}1=C2?;r83)2Nzhq!_=*t~9U=2}(Dm(3fwlZBqXh<WwSg}#I;JQp%
z_L8arTx7Yw5cqq~6hAn+X(7>E2KO#+$dj9C)20qeS*642E-v19eGeJz##k#N8@HS5
zg7@cXj=H^EtZ3bSPbK|8mk0bcffr-|NtKk_`}$tQ??h}0<bd58g6Yxl%sB?K?~XPN
zU+ku(Ze;kscGaHG{4sa_q{6BuUY(}_D^f#}xO3J}c(+sC)K&I^<$(-hy<$AgS4T4t
z>~nbGrF+P0H%lJe<-^gI%}rLleCcEUbx0nz6~e3@P#f;0d!V@g&cvZruh;t~1F6yb
zhnsE?uF%@m?VKgjYkAxv+_P8^oGKd_ELROQ%y>Q2Qaiceh%K2qM>8{(=fOn8BnX$I
zUp5<b-qM)FZvnuz*8y#{iG_VADZC95m8~d%i*fC1k~Ut&dR`_>w<fh3TsjR~hNo`r
zrn=s@8bsvOsrGU4v<I+l@*dhD%DbL|!uE1@p2r;S%8&|Ll1YVoBFu=ka(D-EJE&Sf
zC_~qwMz1H=K+PHH&d@wV1anB#H`%%W5Sl_t9Qt!ZM0n+}B|l%23EO(INh1vxu6^e2
z7qtB?$wvE#%t@X9&4%5xAs5jLfV7vKeXI))dV;aLZ(<v+_p)?wrxC5~u^ECH`&=-8
z<zI>T$JFbMiXs169K(FH8D0aX0K-Qjx^;Y~U-i0s(xxK-`Qf*@g|xM8Sin-z=0o-I
z-HWq;g?|3imGqHNoVpV9u^?%MekAr;4?e$A;nm=!#x70$g@Z)`yDK%0NSWIt+?}EO
z6mH<X<u~_pvyEK#EQ;w1F)+@S{t1h8w~gP&@>%cG9Itn&1NmgjQ|rM%LclM6AAcwt
zf?CJ77`gW^76|GZf2;KCL_2)UxIxQ0FTT7ksBkgfN3GcnC7sElk*(wLj(I2%g9O>6
z-$5|-D&B8C?XMriMIEv$OR6vxOuY4-zwGojz&)VlliRpCLdVw~xB*bGkvl4__Q3O|
zw>~@|s(jJez}TM^5L!p<!BSo$6ScW(8RRzah14KbBHCrt=2yx$W$QgspxtPAPD8`L
zYH_%K5_}>!!aT6KuMZE4&pe`6QI7OGYjf$&Vg(|A4#k54L59iu<zESIruiS5>Uh0+
z!h1fYFqdERgZRpED<LBq>|ulUeAYQ#wXVeU$TEn*TQ|{S4+EcjFiB5)Qdzjp!eMLm
z6jX8+z2nu?&hyrCD1*lg_vNbz9S*6w0pVX{Yd{6VdW7&I`SD!4kBJ;npZqs#eDM?2
zd|cjvzCBV*_=B-a<(%=GgR0tnex!OjcNPWh8pi<IAaRMX)rcU6m)eg6uB-Cj&8PN)
z;kKg{Kv`kD_<q`eBYND;NwHOaM1H&8{2{zxZWXhRfjg4Di+jpyDBIyzUiJ;&ymTi#
z=0aECKWN#!b64%#iuYFU&EGS!NrQi7WQBG|+6a2&_R+$>mTommG^)a4hSz4EW%9Fi
zFKeSo;3|B)Lkdmke1Qzpd-JKJjJZKOyIQWe2JF(HUiqhw@AazXKV_VZrm-XBDU)QG
zLEMw)zjo36$3m>6dhwMM5-k`XEd}v{kD#;Y439`i|DBZWX0=P9^TR5Fh@|y(qSXtT
zu{q?pIb1~h%;UveGYSrcxA8mA4`!xf_*MA)H$^gw6Ftu%s1xT;fdt-S1;+P!Z=Wo`
zrN2bTq)3`}#r<%}rSiGK?MXBPT~h^K>2#B0O6Lleco~{!jIDuKdAbQJtT{a5p865A
z5Ch=uMwD#J5$v^Q17YEnC*RLCIUYza3cv<a{32n7;-9&SYKUWfA|=1mR5AC>em1MR
zv^z)7-jkF)10zo{RzYF?BNp9M(Uk2^3cezyngM#3{7$^G3Gc29rc5-1aVBvLszZyP
z3Bix^wbvSicDUu0cF>tQ8G$z~CqLbsbcz45Bt>lAKkQR)M<UALTx<xr+_G@7^aQ?L
zkxVNnN%7gVDmZ~@@}dT3^3NnIdLK=zPFj$*m^I4m;WAW335e`JMbAhA=nv6!jpP5O
zt0LTU&Pbx<kHoAW_D?kJe`pVgD5b?9hxLbA68?%mWY7Ph4I#LsC%hr0wcomMFH;R^
zus?XV|DoL>I9KP%A%XQLtrzn*t@OXQRovs|{^x1?pW8RW+u<4b&wpw;m;a^~{|{{<
z5!f!0lG`h(#K#6eOUu!}VSxWbJ4!fwpZ~i+05<;rcY*!{=)OGttHJ9P{C%{x^pv>=
zyf_=}J|WH3@+iEbH!4ZBn-5n%Im4=!yEsWG#fJm5){RLyHfX>}V@E+!Ayjd~@T&`-
z_@UlEo5)i6o`$5f0?zTgzx^qZ8ziO8F2@r7tFbKLx8rY9E#ae~L{s5{;m8L_fu$kM
z`^FC;tpYcy1$VRdAlg`T40dUE$;Xk)CY;N_pBvI-d*EGgA9PJt)rl(Z^96~1Gv%UH
zCkz_=MbGZN-Mmin61r!osv>2<J>H^CWZWi^+;iCOn{9U^%DKI@H8S3Y>EVwBGxlH7
zwR`yXZerlFFr$K`i}ct@8fDCBU2o+}j~PYoy?2i{x@>M{CPT-e-HGL^c8pr*(VNR_
zvR(h!DKJqolfSNI#NVX0ofEB>J~5x;hTWBlb=pa|JDPh4#q;pq{pDrt0w;1)RS92w
zqpu2eyEVg%wgMI3SYjPop~KhnOFt`is}Ppw&>r()9heXI_K4RRSQ@Gc&W%79A$ywE
z`?qdoRK~H6${tlqxO4v~c8z-o3Tp7Mh#CF}^)Df>_73h2=|P^DqNGbFdWik<ad6C8
zFB((Z2sbmaW(#j-Rvbix+nY##SC<_SZlBp>anDp7h1RqDbPt=O-E&EZe2td$j~5ZY
zx^Lh|^2KrI<PE4^`Rs_atDS`snS?!Ng9i4_twz0hMttY>+wtj6P8yzK>NQbBxssk~
zM7uMlp8!ZfxQiqh9|CEd{qUZ7*kZ>q7qF0%IG4*nV|*BQRE@7kc;4lmlA4a|qPF89
zsM|y>B3fDeZ7?Z}dy?`l$5YU$)n4+;JN{D!%v4le!!2jc8F15EX_#HGTV&6mn8z85
zba(v$1wT^z<rTZ0PLWN03a5|Ed5EhYJXmzrC*SubwSLf9RccJvpc7KPgF%|re%ZZ*
zuXPN;zE!Si<3b<ZX#Xqva<_Yq9j(cGp6l9&FO-^Fa}1W|2>E}P=KrCkQD4A*!AGyA
zgL-J{{CP2K&-a>d4==ogM{SVe{ric_iPB;B%%|Iv+pk)$;xaKYNzV5b1*GJHrpf=?
zB|qLO@0$nJcQ%@01nGcp&1$=|Wi<`tgZy}qdeKGQw2|F5>d8&s#uvSdgI&Yxrbp#!
z`8D{>l{sR|NBc``YF7b2Ne9ql)+yj9_h^GC6BAV`XP~~*HDPdv$s}Osj@|IVYe4(-
zeTq$R(#WGLe($xc#F4mi{pk-XaDme>y~S2Wqy1kULw9{gj`76A81JfUYO(?;f*rb9
zy^B(g+>yl`Vs4BPWS^d%-lTdV{>9PZlazb-=mzu<h@bOq@FL@ug>W*HUWpV-;v;y(
zi5fRe;iWh@(khU*($8fQ9qq00>GXqOtvmF#^@K8brpIMaw|lNSb6Hg{S{?8BcyJI{
zVb86#nG)Gz@XxFt6RDMZ9iix{ef{p-K_1GKHnh_Aiw3j=3bQHC0&Z+zaB%TfVv1v9
z>F&<9C-{}4UC$2W!N#Tv9|mF|uaHn)+v~z@<Q?FleQgKQ+4gH8vDuTirMtCbovt@5
z0J&}$>JJ_PI{FcW4rrF7?#l`|%>KAKBJyc%@YT_p3P&tkU44C?v#V2Iq1Sl-LEKSW
zS}FWz3@(AM4p>JEI50s8VaA$NCW^Pd;A!{iTbbs{5AB5N%95-*xhhRRzdHP~J2JNd
zy)xnXZSKC`%iy`<1V5`wRVTx;E6&kM%@j+Z%Ir6uY+u1y&VBfH_rX|QP_(m~6W^BE
zQZEe{6`=xGjq#Oh6Z8@VC!WisSd!$xj(V@x6#X93AUv9P1T81;not~l7hHRo(P^Gc
z`QaOFm??Av6xGr^X4+|I^-80?YEl!pQ+ZBRtxIOPe&TG7G^Nn3=<9D=F7_2$F(<i_
zAGFOKIeps*Eae$ahUD*$%b%Fm!nof^1ke}l2Jk`o{HTMI<xNkE?Vas{LWk{q6>x<0
zJJZ7Ml|!ay7^Yq}t9f09O*}SA(dCWIQvHb<VnRGm`8ZL!L(@>_>jCr+Pgu@3DJd!O
ztSuJDpnK#gP#>Iz&9BrGwHmXty&)!iH-92nwACY&`%&HQBX8CYhX4$bZ_?|ZgWkQj
zsa7~s$apa5m85suQo1phDIj5wv;b4PI7Umw&2hW@0@&>8iB<d#0qvZE(WQ~<9E{Ia
z^T&rGX3%ce^svVTcJ1;r9tE?P_aw{=6e#8}4PVT9{}EdS4D=PetUvz_uOdUMj0v@~
z6-4R{5?T*9$c)CBcRbGy1CZBtYDDoqI*rpOF6W01vU#3euWe$bf}>i7JaJTjc(NR$
zpCnh~ph3;8^K$h*3D42M;GiH6H$qA#(_OJ!Ci=cFxWIf9HcgSHI3P3p-7H|vv^T#&
z{f9j5QaG;=@y9@$MBuVnw4c!H8hv%9@%)|Cjx!aMf~ZXa`D4ZQtCO+0We|7W?p#Rt
zWH6KBz-}8D;ZQF}y~sUQ@y#1&8QcLn%JFzN-l?CB6HPHm4-*~-nNm^7eLd6g8TQz1
z$dI%i(HW`os;yFz8=;}{d7|;^DIMoWsZuvS@wLP0QrF~pR=)7s+WEG1d-qX70%s<+
z5IRsZA(`-pru3xoADdW@?U+s2tc=UFK=(?P?HK??Om4;x7AC~nygN`H_TbQg1BVAB
zZVM@1a4*>jyNDBjF*G#flcPLnJIs_DNyTU2s`7IFS_6is1nMqcnLLa=dC=B<kyg<Q
zW=4H0S-q|+EhL^9INehhH5)|>$Uc||#o535Iv~v^{xC-U%+~Rjlx4@Yb(=g0Pq@Ch
z+mXo0-d=V5wk`FKnA+HnQIt65QaI*zSE_k<TDGfA`wIQCHE~B7&twe<-D;#1NOJtJ
zjt*b*U$ux$Df>pMy!I-*6<`>8Q{)^v?xHlIG`*zp4;nf6SY1DvPBccT=8LWGK!R=Y
z6CGFTi<&S}!;!=9ONnNiM(_S;Bs^Gj_#jU|t*oiqGg;ZN%sAm&Caqr1Mr7ckB(w=C
zf|;_D*m@J5QU;9{rVd|A2w;$FYz1L7zI~2^(6MdI&NMqN%A`QL%x<lY{bG7)om-@>
z+QT#S_-9`2gpo`3qN%C4$$i`0&O`Hr@cbKXZejwCyOB={j@Gn2`?h}IAdll@q-TL)
zOr|QkEqZy33ma(M#!1=dCE&oeRugDKcl6Ki*l)}?Wb<#`SqJF_YDs@AA*v+d-#?~&
z|Navh?^S~C@31QLe$Ow`Qcd_o^a``B?9G~qS%>p8F>msg<HdokHydtRO$Ux`EFUx>
z&AOFstnG`#Q%e+d2B!ipEZyz_P_K!^ugm4FDim<75}dhp$n}9Ez1wGN)X|!gM@HqM
z`A$Js2Z;!Buw?1E%DAJFELrupVUalY?VY`p_jykG2D#)0$o1JBXEKZwaNw09b@;gg
ztvlcJe16hhPIcHm|Fe<j%?h}V(yqBZh3__e{#w_38H`@T2XB$XFAIBPh68CDvXOlu
za*#)pF0O#-2163mi5)Fq=s)EvHeC>V?X^>+Pv;c#{+R@cYyvvvso>zo*Vc6X6vEtQ
z{xSKd`RwBIR~FkuBhe1nUH+MM@$AQT+{;OFF_;In68E>>s)q;o(`25}+s=77`E<g~
zsqgNsa+vY7!@CeU&baRB)JGzdc8th(#jW0$It2Zqp-7@J^eSz=UIe!(8`R842fPBZ
z?;8(dh9^_+n%=c#>i+AS8fbnL=Ee+;?`nib1=oxlS#8~V-@S@0Pa%Pfb-nr@dpq%^
zQ{=lg;rYnWNd~iG!})w_DjQnj%on($m?iTqPuL?o+!Dnhr7+h9D5;&xtgL&bO-f2i
zl6TIyx8=}4ff`&GXfV^vV_#}=VSkQ9I=35;Re;sJlxfmNVuBXa9xb(+xxXys1icl-
zQNa;U6~{OMTYB`g&;5`T2fgU~=UgRl^1HSBURNTEOiNJ1INA8(D^9<dtoEUH<D)fJ
zmct4#*@E{QE7kLwx*|(-jE&PnAx6Ye7z=u2XTen=cDtq<Wy+;carYj=n{``?*6Xv~
zePo3166cQJJCJlP<yh|~NYbPY)S0l)#)CpaJUR6mUFw(=igFw}P){7l^@u+<>}K2R
z+j!5x<snAyOL^lNl@U&VzHvkn2`$ye?K7U>1Easjt9^^I1xG?^pxi<Id(=(kz@Q*q
zbWV<3$_|1!fqThyH=L(K6?R$j)>DOOW6Uys%@pO=WQC+R$lXv}TwIft^5SzCsjb_i
zAMcFUNx5C}S43`Ie0WQiQVaSfDfQrJ$;<n-jk-JWnPg<n!~5VGBrv1k#0rv92z4eW
zCnGoDz4FuY9fEaapeB5WcYA945dHOD(o4kB%LZ-NQE1%J)ukfnmE;Lz5sw%_$8n3E
zm+e+7_+gadw1+dkUElU3GePWx&lRdB3@>GWmkksoL_WQ6&{j4yEz<?&`nLH`YbkX2
zN<e(-2u#&?W}>E9&N2hni?amVk6FU>o@~Dg?AGq}HXaF)yEUKtnN73U9X5jpTspU3
z|L86^6wLPiR!H4D%WiVnctY9zAR2q%zyT~apTutaH}z#U3U8C*P6w7dzJ)sX08V%_
z$=!hu;?=^ZaO8>Z-WGbJ;<Q+G`|hD~!Z$<%zIRE4iE-s$9;La%c<(UoA)2)WK1&?O
zH$=im8CM2ZoYA@qv)^aS23d19=51moHfHC)N9N?jSTue|&-I^XtUi6mZ+VC3+cC!7
zIa9|7-gQ5*i{1J2LHE@d=^P;-ozJSR_!YamfHj$_1{Bhpd;LfGv1{7cjxVM_8TGoc
zKqK2qT5?Z$ox`Y;1z7ud1wYuklc3jNGBk2fFa{{IA=JLyMe^r8&%U1Q;HYOA2|N#o
z_+sE))X&>hpsGv-kLkymVAwzlZ3W@(JPauQv>e8mwdG?lqYNEym@_6<`Iq|TUJMgY
zgT&fZ#gZxTElIC+(pLD7a+BRI<Le)8Rsijx>Ov-hL3q}lA1V`E#5)-o7VstCIOI#|
z7}5}hgLvAiQ&p@C$`FiGua6~QrE7Ytlp|hC*O)2%cywL$wNYcR#>WRQd!Ti0y7SmE
z=CpU-fEic=mGey!O(+sDOij<vt|L6pal0Q%US+#!n5PA(No+H8C#OYG!|H>WaqbQ<
zPU*W0k6WRfw(e8Md(QKz`E$WleOad`MX@>2;_to`uK>ul&`>MbXBMQetcQL&9X!+y
z-DlfdeunzmxTNNLh53#oX2}fKa%foOkAj@C_4L=RLcp)&`8yW&YBquj>e{oRPl0Sj
z{sMZ9k=7lb=HO?1X1MS7fP2LrfHWP@8}VK_LpdKnbzAYLL(S5q7H$!ezl+$(B-os3
z^uf2Sg*Vkyo5CR=s$Tq&Wy(Rla88x=aH#Ld6Em67V)CxpkVqbj#I=0lqW)1Y?t#VQ
z>V^oewmp<;f~w$RU|G9^FcY+VpI)#=!P17Rq0nOry9@<NWl%zCyYzg_Dzdm!>*4+N
zQ<K)%<Hl9UkhLBO@AepFCMzjx=IJg5I1I}5vbpIB&(pMNj==)h7txz*dK;}vT?C{s
z>(w+L8)(Y3b(kX0e7inLU@bXNwrH>*p!Wy)bXlcUjBjaFrjRcCVhPlOD3mh@IDElp
z2$WQ*_A=6({%n8SmVHxOT%kRRl6d(vv3aj^q<Q8PFwwh~V&+n+lq<HBQlmLCE!+EJ
zMUL3ttbQUqKtN|t)^$6he(oMxl9Wy3L#N7PFGU;>HTI6)djZ+#1j{;G1e8;>`UIa7
zm2(t5WB)Q{a}_p5%yI?%uf97j^B5t6=IP)TE+xgV$Hz`1^9-73zVkB|UevJ&&ZOl?
zuAqx=S{g8D`97i_y-|}E`uSQ39ncBPj7}i^YfNIV8;(*J1NDM8Vf*7(0wa+Nre8KI
zyFGL=7q@|25W387JE?#wV>H4)RiPUtb-Ao;*l|>k;gd@weo;Rsc>*~R`GX!77!Lhq
zsT?BWEy%FNNNU7WI_}J9v+~cTiEHubMYlm);hr#m0(DN*P24A33nZwdC~LL4jc=q#
zf#5kkjlPVjnQ@ev*tDh${Qa>kiZ&?enlN9RGh|jalisrQw;scfx3s|F-Y{!0k?2aN
zbKLG7>=QQb{4q}(ViOWlbOp82GDm{ai!^(Eje23!ODM{pYow!ZK~bBycjFD#HCwYE
zeU(~IctHLsEWu6Qumi}BZ*IWv>#K$d!7jQ9st|{enG}9zZGX~`7-16L%DTYBZt)N^
z9@oo^0k&A=GQUUCM2w>VC6!O^`ONWhu|OqCqDd0$aoy(Pl^N*8<aF*P!Fv1Y3hYg#
zV3ud@bP<-hJKu=qZZAcg<MlVY=3P;KwhXKfYX{VlX`T`M(S48p=F%ZOd5qSGqV1`}
z;Ldv@h{a5%-FteLl_bim3l?AH9<+auZvm<I7FnhC^xI*f_l_}SCOAsfnX99aYxl73
z)(ufi8s6{_b0aqwPbuwxwW-yA8I>xz3SF7<-Dja{OCED7HaMHc=3E6@fBjg+rc|~(
zMsncnxuuXN+vA~|TFb%9@2!QunaVq4dPZw4Rovy9X=L9uG+Bk^zv0S&x_Gq*1oijE
zl+J!yn%P=lo-bWXabmNn?TC%c<WF-ZS>AZ5Tx0f{S60(VBS(E1lC#;~=go#^&l<*w
z&I=VpRRepv<u%tp{JgUJB4?^-HtD8^Tz%I=8Ek<1hdPb#BVpUR7fX~aH(p^~bMN${
z*M>=!YjmGk9JsQfX>N5{Pgo0&d2|7kijlD}?E)Fy>_t$MpXeX5Y4~yrVy{0*SKI51
zh#Pk6Q7`TcfFdwup7`D_CyTO3!Ad~E2WU`EaDOnMSEWwl>25R&^U9C1_a(hM?70+s
zr)4<BR$p(4A_l+6M|PwADs{DSP-LlbZ{d@Q_?h}}MAExT9n&T!*dI=mp4X`+spY0J
zes2Rm`MU+Xu+32vqO$b2aU;!oOL4Pv+MYS>qj0jRieGsznD+bM&}+rsM7v#5(ue`%
zAzt<#PM;^}Z>h!A|L#KW7X9l(Ixk`h)O3TC((vzK;4Kt0x5zBsUgzfcwBUFRIuN~c
zp#N}HCjHBFk;Pt}02xJ$wC|f3?w5k5?8G|&TQFJsnew9PY)&o5cbW$OV?Q0sZWqpD
z?4XxPgGkiU1BD;i%XQ3ROQbi{S1If!W92ad-ND(}Hq*OzZ+B0`C<{_ff5ZPm7&PQI
z8xrVdxDG9ycXvjsU~iqL0Mu~{koC?KU34#j((Lz-efX~zgS8zCE{=abxDGq)BIL@A
zTi}JDV0Gs#at2R;_L&m<d=;T-Tk-AcvH8291viSC?oe+3sDAH*T8(QF6Ik}s&#F9Y
z#B<eFUqy@Q88ojFU5a@19w6ClcJN=+dM2aT9j2rl2%kTx3^DP^$XV?j#^Jx8S?9V~
zq4nsz{~#c&PE&x*){2C_WAmQMIW559A@c$p9*ar0AoQD#V=d^*PDrbvxPj8dwp2!0
zXi@tX^Rk!fCNy>$RG5|u4+t>l=uU3-LzcdDq}pGKEqNGMqS91X9>Mo@bd|4PTMKJ?
zA$pQhK%>$*OyRx>sW0AKipn&~^~13vw%(^+#b{7(pivm>GjuD?d@~<`-IBUF$1hp;
zf}>bgs|*b#AY($U`Dp@)M*aEkk8q!Wb#b|QNw7;fMsumz01efT+9}D|TO+)MoH?Q-
z#RQ7Fus@g#LTp&T()Bhk-PCI&C(%yUXKtyTPR-DM<Uyx+it@Z+h+z$Wa=$=lk>*j9
zDt^&=f^3)n+Ys;^+q_kGID0JwWvE=R3HB{zbktz)xG{+lc)%TdeYM0g1fR#GkoVgH
zwERp(Jd6p4Ayo(I&T2vS6cC~i;o*qCj`Ve!@07iSD~MK6Y&BgrAp{4#mQ3keNUBEm
zmZq+D@tO)XCR1ItdyB%KlGyw!O+OzMpf28aCzQZ#`n7`*1zp;Cu9*#gEAN%h-U>?d
zSvaMPltuV&_>!)Tzm;`|Pw-CZY4*Mz0!-iw=vQEqym=m%4*M<7{-$K>!e}s~ZO1rp
zg^~c$L4s?VEto$?6*Zmh{Ccs1O=NV=L>wg}Q)9mz+XxVx#EVA)*i&n@gQRNQXj1r^
zvb`(ac+IjGMd&uEIqAlJ@}P^=UJyR7*T&?&IHT?H7MfwTY+Ui?^Nqs>1HekUO!wLY
z%s8z&9%{MriLHvl3lnI#*doPV)?VSA<5$kUQ`@z>s*T}oD4#>F$HCf(z>LS3alVAQ
z%9$a@pcu=ScqeM*9w{srW(UJlHe~DG0c5w)R8o}+?i7uxl)K-m!CoO<dv9Ox+?Bk5
zVuVyLiSB$wIBq_@;M)lCt0*&THUHfti3di@Ukg3XT&Unj-(~R5Qjy@b6o`iTL`^z5
zQD$viI$wc)Dv7!wSjFBbE7gkS!hb4*I>+rNJOV4)$fw{>5am(R>x5rK>Qiu<f~ZuI
z?0)TLgWj{{ujqgO1pKw*Vg8nr(zgeWNY|7-FAbV>8dB={C|PAH;2c2N=(FUR)8@Y_
zD1)OOKv^Xd-@K_S?9JH}7%!qnvwZ!5@O<S+`fWVi%+}XR9H^D^k(g(wx7v|W!SPsw
z(?7y7S4?PJtJuL(+a_M3;3f7eI7n>He$kYUm+E=Y^)&ybh^C=Yx0T13;o00vJ8}V^
zI!E2ih#9$1a20>&q?>9Eb1yv^>TAFBoVn%T=hWPXL2n0N9AOZQq}-dkDaYS)7rRvN
zLarChDB;$4d(;>Z6`s}V9t20V5>BS-V8%Ibpnqb{RouAADuHdg^zj<{Vc+dzUfOW3
zR-|Ea>1j+4(T{5J4M<wvT4XX-p&Q|j`CLmlQ4Ma7fy7n}SlU{`5&C=j7!K;{Fg?lE
zSk6<nWLKy+hL`TYKlDo3;42d>$qCv2E=dyI|L03WuzBl7u&E{xoto7i%!1CV=JL-B
zi3iZw(+@LsCzOEv#@z+;C?P1*I=oEy*fF*8tU8OoAX_)?cvH-R@wd&j;VIZ6PdJzJ
zu8eeb1zTOZg>$L-3{Gt+y=~(&;b_G##;2(c{`rsKLHbdYKpNejfB5nqJ^(>U*n<KN
z;QRjdk5kO8pZI8a!(ha8jJY0uahU{E?E=G9$O2D7tHim%iPtfo_U=qjtn7Fpq!9{`
z1nzA2JtI~D``#C1c5<Z4Gbt0!NkdK7m$!qp0Ssc08(8!BIDZHS!nf<$q1>j?1<^tB
zT#z=IgzBX!u`H;#zXK#aTdV*#2DhfEsaiRLq$0y5(p1%>HZX6IaqNW-I+!F<n@{C=
z%xcv&ZZYH-cEYLMKUDuRNdGug`2GJlR3oInj>w)f@*j@KGvR+6k;cu+Up8nd^<QT0
zPa723{0|$%i89-_C^{I!|7=khEdOCq{yUUC!}Z@)`F)Qb5+L#(QxAdgU+&p+_6wA4
zmq$&Gzl8kO9f-cov6FE=k^T?QE1573x;J<pyACFW_*YQ<PkYjn)iLFuac+SYT9k+Y
zWXBO&8$Ztl2e<&k3VNuy=ts$OaL;J%-OQb=D65KM-Tsr90jLn41Xi9m^r1>9uL$GR
z6P04kQ_Sr#OkKmKL-}M2#A}!<y{k;;PA^+D@-T#hFmr10dcS4cBxqom)^79<H7DTI
zBu5hS5)e@kEz+(>>r8G!Fu^7cA~M&C;ra6(P{U0vQ06*tb`*SF4qhnojRW7E{{n}O
zxH^X}*CWiZ;(6o_JkN`=5-hgTR7Qaa)&sCmf_D)8h2>mcmnhOUmUQW~6w4`Nh&!}E
z$ICNwATMosO=dD_tbLn~yu#eIFaKNAvX*s0?HPG%D7drpiRj3*o_TzTq(w86V0Uz+
zGC;L9rj+v(51CHoO9fw&9bDmW3lNA><g5K)3W5$p=qlfM?Q8~MZl|H|^4;TSStbRO
zn7Co$jQ1w)wY#>%td>fb-yqU~s#DD<U1J1+fcCaFIKvdpQ)e<!`(PEA`Y&5qPPWPw
ze>(I`xccGyLpZ2!0|6&aRX`b_SVnEtla0+fU(ns$A3FkLuQ9fk4*nS)L3^GCW$$4U
zwTVQDn06~OeX|gt5=*~ytxBG832*Zc7rt`Vs>=9}%e*|64;hqVX=EpFJng0OW%&xW
zlH)dJ9`v}b(!3Jt_4yi3?v6d_?a$w8lNT5WT2OZNk=GI}^+Gl57X%ZdKr80W@>C<u
zLRhQBzNYEM$XmojdsN3f3Qmf)V|wG5Z7fgEy@P<E<+}+5JJv3yojrGL;3xlwLbD<J
zAB!UEkvi_1{m>rmy-`LHedWFR!`h&3*q(O`w5M7dgq&&I^Dl3Z9)^>kxN<isM?Wd5
z@%ImT1^IVT;Ady&WxN%xVwVltbK<Cnl_^U|JFSKo2(wJ{yPr6q!@Z^;yhvmWSZ@bM
z=m%51-N|uZU}v?>!4aVxA#b+Zpqf?ZvYGD_ATblZSXqGE(@t{#NOk6nU~!?|09IVX
z+=|JE^3}boq3yP_fZh$V%wfYmyl<XONA9Bh<JpHYqKeIKQe@HtlxuWn{q;0Kf>oMt
z%2PBJtgaZDmQ=;Jgx7+E+0GtHI?4KW@H!Pw_!Shz1K!I7rz$8cGuLHARs!_2buc-V
z!jbN&@bK2GB)!x#e6Au^G71QX)0{Erw_6RSsJ39~XwiHDhI)-+Bb)j+sd3_HXdC)C
z_38*WDb+M@F#cRKg-ziToXBP~=S$9m`CdjQ-Y~mCE_|pvA-wQJ?0}YgxSyk;f!CW%
zCe`ZlO(uURdZ&&rP@t|dP5O2=ub}d2_EUK&lcJ+8?%K-Q#aVn@6*)};y}Yk$f1K6k
zar3)xHlxLH-?%rUapS2l8y5QRiBL+Ow*<jp>QmukbIojVHc~tcmVNi~g{UCTyZ?PX
z_2C#*3z1-aBfy#`FUh^2Do>{?*tI@$1=Sm?V;b;QsW)HqO@agt^)R<B#$=G7b^(sM
zj~eI#6-K&%ww*R?<I{C9o%Kn8ZU*sWw+m)EHB<Jyf)fI7Vms>^;$}JdMULJ{7`E}M
z&SG!ASwn;0?D{7ERD3|%;~{9t%6ye(gtYppXSdGzj3OTY{)-i{tR((}-#myF9*6Xq
z+rXpDwW_sqDIm+R*P3yv+of;)O(YUo-+U-b>iMN!gxB&X1Mw|t221i{V<PdrQ`SAV
zv<-nt=JPB5)**B3(9$XBoZph?<Lkyx=11TZ!oMf$xfi=x+pz>2f$f?JSRbuezg&kG
zoLUVIA@AY6qm4rAk_Z|^a7;q5aC-Qzr0r!lwlN)?%VRFyBG6xV|F~(n4{kQFB%3Yf
z#Oyq!6V<^Chc-OF22nv6r0eEMgY>%?D9`b~kTW{{9FrgDdxMuZ9|_h;smjrhX2cl6
zt`{u65qR%nU91#O*a1nNSO@|h`~*QjZn13A;f=?X+bV|U0eE7mDqx^=kzc1_VR6Zx
z8YQB}d`7T0bPXi&0+-%+`xsZH7A@TS5rALR@D9>#nV7HVV##}SXHA$VT)fr4lOOp0
z`~PW7C^`qP?q4$kQORE)<Brc0UJOVbO3bWBkz_Kl_~*@$`|D;rUenl<cdMP`efR~J
z)VZ}z<OT$AiEmwn8Mouugy7WpxVW`FMqqn3r!#9&PjFjxe<qBtM(kOw?A{UX+ywZ$
zLrg9KKw*sMoUPAuZq{+0(CrM1sUjB)V3lRhD?UK*`)o%Ng4~`>w#w2wJ%TM`iG<yL
zZ5gkMIs=O1V<SRJEZ}2h?*-rf+AyyM;kF2bc`h5N)dh~nG)UbPArF$}e>@$@x~LiF
zb+n6;8ON^=*qhRkk|)Z=WDK*PD*#M%{4PyCvS^yslgTa`19C;szd7+lF}Ui5jTP<7
zL<E}C_)2NW<=6Ln7E5?gV<V&x&XzS~3G|Sn&t*CbhOYaq^8T(c@gC$=VZ;^eWwjpM
z9@yA2t8rDb(75`Twn*^Nx1&}-GroU)e9D`C`$2E#+;=NEx={$^LO<}+=MY|~?Lg2h
zOo60q+I#2Fmke^+76v{}$j;77O6X26=0PDnAZAoHJ+(3KIpR1_X6?w`vA3-omelc>
zl^W!<+6-6LAf>qe)0f=_6w<?;*RL1pguT5p8RBzF{gRJyul3Yyvdasg5y+l5f*bL^
zY1vxv3`UAIvnN)HKkPsqvJs7)Z)KlAE)^56K}2@NS~TjPob_z2Ls=QH11aB~;Nr1D
zEY98rFv4_J3M;Jy^xl11&Q_2}4);PQue!Z**yt<ZO%F1Eo|$%Lv$1-wciOy2TYNw%
zSxzOB@W#-AuA&evtvrzslrRaU6HKzkc7@TNb>yyE0An2|u;d{ZUnR8UJP2*?f-XQv
zV8psIoo#Q^l8qF|b|#0%%(#;0`|wGUr{%Uv8bX9e!HvQ?6sdXTg-*zUeWq+41>VFK
ztOc{RHimUt-$%l?Ge$O~g;7BHR`WvIRg^TAtuCif6t5DeY#sm`Jv4sc!M>$1UMhL^
zQ#cRQt}4|+wm*`mU6(wGxqqLUUML0&h996&c73NWpWUOdubgHi%Gl2nQ4T9YOUS(G
z0Nje%%FYFyu$fJMKi_&0$>kWiWfxt0=abi;U1&HQUxR$Nro2mt6m0XcCoG}cd()|o
z*#+Am>-F-{1~d+ptDX3ess$#%?6(OhVMWb62}?mYgkJs1i=uPTzNi?|Rg|AoQ>oG`
zC2)HX3|g=Eu|k>bCP@AbM~l~B!PQbd$td{PMI*&}(sx3wDhn1_@n4Jl26^DASKqQV
z`V1VK3*Yo%dF}sNUIk2N3zO1{MSu4D9Y^SSC@DbRUW=drRjVuI9HtebK1>y+&*vD;
z-TOg|kA@uSz2-69Wq^zL&y+GV!nR`ONj^sS1)sug?!r+^o;_24x&qLcIRb*T%5L+}
z8{)wc2*peyDWts_kiv9!&?v2x!0;7pHCnON-W11U!D&rTvm2vy$`014fV*zQ+Ci92
zVgtH0Dtw}oAS{EXh_OutlOq<w)gC=ZUjDjMh$_qvBQahnFURFzlCU9RHn1S5zs4ql
z>FmCsw6dZ}9dbMkx*_nn%pkXzm(OVhbCfRYULw`IcMZR2#-T)6%Qj&ql9?G9(ao$k
z1mEx}@S;wjdaolRvho?MXDdlt$Z1V<7<-slff-}}e6n)0QszNV`@Wf-UsAV`Qf!bD
z*m}HGG%yJ4z6Tn>mQl_pBtRsv9~)VO`c1of=mz0ye1!(lmO<Ka7B7QfkMwkfn52z>
zEpK>>zi(!Et1~s8eca)=yrHUP)uEJk<@5+CxASkiXyrX<@=>@0gY=(NWJ}uF1y8!(
z_51qtZFG(Aj0|fL!P(<8sFF^$9nC1;dWGD4;$UT}&WFi8u`|ya@2tNh=9O(bONu05
zEiTPmsBi=dM|$R}e~i`tYVWI~qFlSaZNvZ-5sOj~0R<816jY><lvX698>9^oL8U{a
zL`0;9W~41z8YG5L7(zxsh8!4}Z{LJ^&Uv2a{r-8^x4yNWv;5;Mg}Lv!@9Wyv-uo9D
zU}Svv;YU-CqfY>^Yq|dKy$1L85!H2UFPkgO7xyo7&=deyPJ1?DHsONy=Tmh$EUoS2
zh3+#fE+IK*>B)A@szWbsr!VhPQ~eU`H(ZqM&DH&;S+(HEJP4CWRFSG6<H^T`iu#*y
zv;`vh3hf{|derwT3;iOOhwdEEWa4(m;=;3_>E4jk!E7PqYZ#@{pt)c2D%&}<uAN$V
zofmHb=lL3{d`C?Qeh1x#(np?<tFu0zGF!VAdZ84CU8HMN4ka|~=6L(U>ETV{z$vbc
z&22S>vBXOLM5b4yyO?N_6A+=WzvgG;tRNfjX2&AYy4?n>PvUSo?3|R}n!@7^{m8K1
z0Uvob5(L4DNoXr4W8<V+1MrVuxaXq+b`R2^{Y_2Q-(iI{pEBipMY=0|@Rf!vm0b=^
z>&$Rr)8dTIXqB6f3mSKv!XkDswKjEUa-Xo_;uvVe^$WjmbG$>Hm|5<ahOCHBJo`rN
zR^P*c@(z8kQY~oTC=bRm4Fdjn#d0qIsPc(}ZTU#IC5K}QlkrLNryEpn;aJ<({qzv0
z3(p}{1#4^_XSq8^tmw^L5@aX^>{NCgI9m$AQHBne9jx%}E-K``dSE9}_5S(RyX!Ax
zU>>kZWxRrIt7ncFND#D+J%DvlWjwjI@f;uZTfL_B?}woVvu$@+Xl^uL4sj=3)jIEK
zs6C;<Y0edBG*KLY35T8K*M*pDpr-dAxzK|c3vx0f(faiqo2J}g?9ZE|2rV|A;74a!
z<e#g$NuKkxrHWdAnNU|2$2y$$gopfn5^w$$96h!$hGTT)JaSeoTo$|c=g@L$4!1Nr
zk`sWrko=I6K5k11Xg!M>xUNHTN8$8gJN+{_fG!*%uB~dx%1N-Ry1D#a-GTtPa0&a9
zcd!}_H!@YmK{U;2-vO4%9r`duT}7Y|5};Ntb;9W;DY3=oFb<c?Kal)9yUb~O2$H*g
z|6b%xM2do4sF<G!m*npkZsXm_RA&RJ;~wgAu!}}(SrrbY5B6|X9dG5?af3MUTBb79
z@)&X;D@|e8y(RI{P^g;`L({N!yY*S>fNiYe$`pq{W(sW%Jzo$EF7g9Pwi=4Ij)OM{
z2EAYfp*wOoUHFwGnb7wwl%F9OM{R^I#sOzA<uya904@cKg+%)0H)&s?{bk>^$_iPn
z7RH&27ztplrr`m?ONFNo*fa7%LUtb-KVz5XKW<)$muo<fN#xW)LMunb)>kJET|GPH
zbuuN#@SfzEu2(<DAtmGTe|oS&OfqkD1qMYQub51+U`_kj@YZ(l<-+^471Sr_6r$b^
z_(SJof&9r)9yWifu$3z+ujTwU&}PRc`p(ukO{x=~gP5t85RjR-GnE<8Rm6KKgrD3#
z<$Vwto%86n)cGg<0XF#;jEMzb?*gLC+1C7gm(Q^CR0WOHRQQjBv0yKAMQq_c<xAkl
zU75lpE6+|<blv{oO&(m0!K>EP1n>_$3eySRZbOuju+koANLmq+j_5M1<p_-i@<tOz
zL)THQ%lF}oFOiaJ%%<M^j?Z4cM*({~j;<Br?DM=u7V4@lQJG6Ju4IFJ(uuj_XSAJP
zXGCR4ezN8aMr{?tbaXmeRHxm5QE0cQ^@G$9v@g7*+Xh}(z8-bQd5{#VLxbXUuiWAt
zRDRl4UHdx!o>9)MdMX{5QPd7?4V@91yb=lu+JoW^58%w({M^=HHMT*?8x_P?Rqk$L
zh7<;&M!ieW2(jkO&fF?iGhPNpTgiA)Zo2NS$d2P>aKeT^JDnzmQgD!$;K<yR8*3Ol
znU?-QvUpGbUWA##-ethNQ8F4y^c^SRV@;7)EUNjXRFUBlaqhs9;8ZAQ7Dn{2;p4rz
z7LXr|C5tVmge<kBX?+%jJl_*OxeUAtk`CY(L}_QeNb_DO``vnaEWQ*JTSs9a*QRc=
z6B|j_zrXooVlMItJs)zQXOV1okz3Tz3r=+ve0DW-G*azyI~POG*gL*4ryG0HF9!XB
zj87wA6MSrBXa}i?BZ1!k4SS^1Qibn_Zsu?+GBMQ<_4dw?eN)H0)KCt6YEPq|==X#{
zVh|evZJfo<(kv8v`9eYHD72p*LCzj`Rd0sTOPLo~>$jn(#*rRd+Xf8qsc+}ne!Ltd
zCnCCg(u<KpAcFIR5*h*8(uE2TppEJ`&}K|D=(*&-+75Z5{~Oa9f3B@=bKx7)_a;Io
zyhB35S}zHpqg>YicDYw^$^SO&{}r0~zlJ^h32VI*`4fMF|9|#wNx)4jhTP>jx$M=u
z+kXFfVFSbHc1y|lASl+aKTV<lS&io!$o}PHPRD?)?uTGy#y=3NP2_osFc5Yry))#G
zPrbWs7tp&(Y+k8e{o_-`w~SNWKxh*MOn-sGK9IjXVCQ-G^eL4uA1H5OPu5XVeI*X;
z|FmSg1xb*;dl3Mx{`cTDxeEfv0(AQuj$P*bk8tdH#zqA5-TeMHJoSn+^dB+bQ&<F8
z=4ZI`8(5wx|Bt|Ojkhp@UXyM618Z7|{Ez5$R1gA@f4%$lH$>h-!v9BzoC=TN@W=O+
z{KnxI_5ULdZzvImfc;zV{|5V|Np=4b?3b%X;A<Q4pTFU2cq#u8zW(n^`M-QAv_J3J
zoaryxbNJkg{CAyhq-R$cowau&In;CTg7oSNV^w+CS+g{c%bEfURVhj_q76((;hbHr
zj}hS3GD}H%2}WHRBx~q~27ow3%f-3s=`kAlSiyYK+gp`SHqx>UKq`ThFI(FeK>VT8
z&6;uRe*mXq48B{Aje0>g^GFY<vZJ7q{y_>AaE>yHFui$N&WMhpuq{~sDqZaHV*&co
z)t0jEHr5f=p*J<z?-l7WdO?xZHt9JpM9uQn46+M7p~v)RvOgU#tD<#QwY5thz?RN3
z$@UC-jq>aKjhK=uNo<JXmRcU-^tvW>+}2wQ(qZnUS4$<(`AwDOD7*v-Ee)B{UBg%+
zV$h{tb=>|5=&0{tjvINF0ZP*jK*#PE&4vDgQ@E1(o6HVp-hM&#O+`@CL~W_XJqywY
zRLu;ycxG-&SpsbcP_$*#FzxHzp!(|e55{Q2DA|Ca$2*%yhqA>BsworZ%H&r24#wqJ
zNZ;C!NkC;&Pp#h|oSqKhY^X<5p%fLb(C}F%JDXBKu`-V^hQUH-$qXiro~r{Y@^wJw
z<e9J8rDbdtbDkX2y#*CXRNA~;wXWf=)0rv~E6}`1g(`rS?qEoR1DM7f5@*rgPUdZK
z1$9W;t*b!S?u`DWyDR8N+jA~^?#!PzvRG}@E-@osX%fSa4E4cBmj$Mvo*~_neY1Dm
zOHq_yIPq9tIsn9<Z1+O6jc$8R#I={7GUdY*;PyNVDy}}Lq!;N;`~^C_Fp|)_Yaf>W
z0jX$YMBNqNjChANJ^`|g>}m<~BeDlrRlB2ELY5=6>0p?b2<iQ9)ZXu2cAG6?9luQF
zJelNHhk8$^zo%p@HLbvrdKsCE_dcv=OagB|pDJV)Qf7^DiiM#<t8XDFNDLeU**wKR
z37kvM`mk%Q2T1SvRuKNW9p+CZuFPh7r2tP)IToye**vo(J-+SMSDKQHT^4*6MDz7H
zb6VEkEj2kIl>Jm=0*pi4CxNqrRBe?!OU$~^iZP$b4)l;)rQ;)W$wDxP5awDLVE{~5
zl#yN^;;9wR6jiVy(*;g)%VeBydBaF&b>1pHr4#9{xd~HEu0`rPMTKBALmu~VH~`c+
zXKUBXaZv118Zxeir9uWnQIVNPm8z7@zQds~jeYHPQbKR0)a?MVn>@r^`V&lTn<=0_
z>=&P<0Zjkbi=RzI6FtGY_~0KeSQ`MFfA@j`9boD>?-k?R=VIA!XNOeOFd3^!aO6=L
zP)nQp?`<)cB*Y6Rhv$X&(SbE+h?icsCRRd=7{Qws2LqIfXsHV^-UE@vgFH$vEGM*z
zi>keQF2BBuZz3mvCc$#=C=kKrDl#+gH)f^A(+Pkxe@?fE0RLCOT#$nSAaMfot#&}a
z=453gILaT=-T8WN42ELi`JCDsVB~Gh@mn~=z`&MSBPi{<1R-8v(wqVWZWE(<gBmAO
z)zmdBYNvqAIDVC;h5w+~t4znxW?ttE-DQMDuOeP}<k^&4_3ozbIDZmxM9gd`Tti`d
zu5dPOxm$?{^F3`_@ppT_ZnPia%mwOOOniU}RBXWZ=e2z9Iviy}eA@jce7ESwhkIq~
z#q31+`rYl19mVsJfAhjKB-b0m5GC@Nf+zst`%^RhF<ufs>#S#8Z)Xw2b3V7rp2BjR
z-yY^GS3MM0_G)W&0nzXd;-YP1-2Ue4-MjX@me#A)*~YV2YhI*<3IiIZY#$U;h2UY|
zFYP2ni+U8z#tb<r8V2hp4$UdIuHbt#x@32H8pW%qEy}w^K}B_r(K(p)g|Mmwhep-V
zcgt3CAbn<TX-E~MKEBW9%j@wqsAwbOS?e1+hJe{p(QG{<V)wIprD7<hRHjj#94iKV
z?hSb5?X4?zvP%E<Qs~NO6bu?w^zhz(UuyYZT`E7MfYZGeb3OIC#6o<DNJU=w53$A3
z#GkB_K#paN#b;Umx`?G{&>8#L2jqp1<u(r`9I$f{QEOSqSq5mU>@Ihz`gXEyF=$h-
z)rOfgWF)`^Y-n7?T68jAS#pFR=ooU1%aw4Ax-{?R%YEOh#*T9v?@2{j7VaZ<Pw3Pi
zdp>u5Uboj|b(y$4^43A!kFJbdPSsPpzNx?yp4j3whmxDq_Y5_wb{#t{4r|rn*IGrX
zx6B#b1HJ<~M8W%~m<xIVVHn$=TO6lwIav=4>4+9*pdv&xoZD93qL}(}D<Y&B9^0`N
zoFtjB;9cm>c86^5(N|cJf!+ce@poq<Hu^3bs}KQD-wuy4fuVF)n-0?mf+(iL_3-86
z=zdYhKHycIb!fofD|^*zq-`bU0U8%ZkJz0)10B-zxb81`QH7IZe7~+JG@hrzcDJ5P
z9Q*hNrRc(o5Hh7xa?f5EZVD!S;OoA9G9SFX=J!6s=Pdw<`pcZ}gI({!mEUh>^m#SD
z1%cc!>TO>i!mU*T+cAEvqbyhhGX({BXlh)ls~*5hCG<n+xFKR4Wi6!H=ZZU<56dsv
zp+;m#?M%Xgnyr-){YQ95{f_SId*W)frPzMYNw<ES*fBz5_m9Z$XKI$gr|%6IqiaYX
z(4YOVc{58~UczdJ>(^^WnpisaJR$g^GSDrtgv-_X#HS921g)Tf0sgwd1JZ|U;G-y|
zy(1`)eH;vjS!sdX66Ok%mbzPl(bukex-8LYi6V_^4bs|4Wp%2|$m{S`J#HiJ6|gi|
zYPeF?QwY028`FsmF|y+rjl?L09p9&9$JSbj01=QjcRdGYX-#$q1TNGSJcuar0xp&>
zRB)z#7k3W}T49+F2|O-vzCzrJKHP0o*FJ@5;JkijcI8o4&8Y`SV|hQ3O{gY>0%;va
z;j-2bYC%T0P{@@I3|^%Rd=V`(S}nUZ0WfJ9gVWw^vthT&emil?Oy60{aHGb>xe`&4
zMDQ)y4hIupfb=fd{rT7s+1JC-?Hc;J;4-k=w<W>wHYr-b-njql3tsJ7kFWUm^o8=O
zecT-AaD?w-2BV@_txj+{B8@BGZFnBadQSbxk1`mGIM{vxYuita1q-J`GS4$)V4XQZ
zhCSmI_Obq-c-rEQ9nvMC4WT^|@y7urANrl{oG|6a6#PBid9t>$?U-ReEA6bGqWcFy
zFCyaHJpc9KT+u>506%Gvjf(Mr-vE2$E8Y5WdETx|u|PzUd8=O9=A+*}qWft|P6#l&
zTWcY&WDK^iUi3rsxa<iIbY?YsxW(gbBp?jLZhgeGpmqkrUoCGj%W5kXYO!3NUuStF
z3+r?Zv}td}KEN7nR^xQUmjm*uhj915Xq@Q&!EOX*!6_kRmtdBi5>HXtV7s6GzD)uY
zV>u}t&MC%g!2`iD;xvx+<VTQMVjWe-&!GD~i?u-ck|CH06W^Rvv1T7PZ>}22DPILC
zL!eLtV)1mV?U>f!^SV$X81h)evcn-JG8<1KCYNj#y3p~SgdJOeQDW`aPFQv8^yT5f
zf-?u7hIRZpa?W!6vHUUk_E7I4XPgUM9~VJAz6T7vui?MFP>6WUQB*Bh?K!3rKXz$v
zTg28a-WiHlF+HpO{*9~KuR^AgCk0&ahB#qH>*d<%t_9&zB<0E^0#<Hebb>d`S7@m0
zC)o1oj6){pkxADFQpNBRzss0TojS>vcERbX%uMB}4cX~~&kDE!FJC)!;f4s18;tES
zwtWQD9|@Cus3J6HYsEGhalor6+UV;2_0W5a9JpHjRc4<>YctdAeE8cA)RM4=s4ycF
zK7qp3h~?sGmy^%Uk-75W<_+PG&D6N|Ry0;5Mc39}2Pw|ThqcTjmCCN4+Kv#YGjU5Y
z0z8Jhp9jgliztK``+WnZV1g1JIMgF1t2h#ey0Yg*U5o4r<>?0{MwdDE`RG6_1BqeC
z0OK6S6_+98%J?D5B+j(7{W9uRBlxp9<RvVGdm2z5e}@FL_U&r{O#?DBDKEK?xMi?;
z+bo(0MjY%IGGA;$=y@87o&36%&Z~{!Uy+O&4WdGw7%cFqzBuE2CPMY<Fqe=P_(5j|
zC6%3|Pz~B{E1_VFJObDSG-6=@IlAFH)Hga+MS|2j-)FXpe#tKl*xu3IZvE>HWZxAZ
zXRzLc(ThDzf5fgdC25oUMX?`Bmq?OhHiUZ|lwD{m0bZ6@?hz+85*?i<Sg=wz^)M;u
zQNA|2l%s2M@B;}e&&C^#3_U<uxz-KM(YWZ<wG|Hdvq!81ne6n`JFKf}+FjS<9J_1n
zfJsxRPO3Zb=Jk{j37PqU;yx<oS)yhgMehnpPW0#Dt!t>^dg@VbnBEPv9|ALyl4<Mv
zv7;{|6(E69Dd=J9vLc5<S6NmeA~O^`Ng%5l-WsXrYRXZn`9W>EZcLG{9rGjnjD76x
zUq>50A^z33**VZ|+Q$at!qxF9{W*u=fouND2Ogu?68?k|XxO@THOz!CnlrljBoP?(
z{-tofO+~TZ(HOM-Dg}YFHDt@RI;q1N!-c9D+Epp$<71S+u84~NTq{$J80~Jk+NEYA
zo@H;E+y;;v-pS!Y>0Ve|WI^-|ittNM%`2HR3x|*5!fjpNOar^{9d#s=W<E(5@S!i>
z#7wVc^9k-ek7WnQ3F|XDl3$p4mX$eZ$Ff^F7UJC3PA9q(jFG;kdCj5bE=NTh4R$1d
z(H&;-sK14{7BsS6kWt4}<y#%x^V9IV7sFf1Pt}woBmUZ;8m(j+lC;5p4iYWXqlEf@
zGOh917rUg?Ik9zhqFwx(aZx;yOQ7}Q@23(7i!GLpf}%R+9fb8n@{+^~=(n`u@h}22
z(LS~rsB$q!p8%Jepd*;;Ex=&7->##0#}?g*2*anR#D7T{WLCfUgpM+b$N?33im3E?
z&&nev4yddXeyPiolwJ#x{=uD-&_u8HsKj64Zl%HM4jFN%O|wtyq<*pNuI+K2^<6uA
z7!TNepF^-h)t^?-*Z+A$I)szn<y@=`V1@SI?%r+Xu{8P@`Lj@H35@d&D^ckYF8cun
z+4<Z?)M!*UNv9tdH19W}iOv6r)RMJ3kgSL<KPm@S)(<5bo~#PloX?JI>xEW$5N5(C
zH0=nCdfn-uZdWYZ5;ap$ax(w3&(<vp8kp*T5l+#ZF6*FVt>B#E^9v~+m+xt*L%l3A
z8y2qC_epc&VuzC=DiCla<r?++%(l>!>n<${kcv33zic+f$PUg77-W&c&<CnUK~Z+X
zBoVN;+cJ?Q1BlFf{JK8NmgD;b6WY7KAjPznwteH3yDKJ%EuKdR`Hi&}Q3yk?t88>x
zXyWZhS6t%E5SWp>H2bn&*566ehF*t?eUi?eC8()3$0wHXHV<#b2U$iTl|-Ckc#-9t
zOhn=0dMH&QFLdw#ccm1Dh<g`3@Bc825_YkZ3FQ9*KKmHIqi@cVZQH3Gue3q=+Q^F#
zdc?Qqxy;Du4BXiRcEno&0;jYu%4<Qc+!4uxU|XY-qA<It$Z<v2JR?7>gd+Ph^S1Ky
zjma{-tF+gR8wesNfbnL#FAUaHJ(rocizdQp!waBAMrKEk)1g8&BP5BTwNw1I3N`xc
zSn@M`U9ebZDs2Q?&Q)SUuD;zbNK=3DCp^|_PMzHDeJk@IdswR_75ff+sWHDl^id2y
zq!PUh)7{fsOntv<EJ~|6Y>KW2`kU3BApf^F&HOT7zpf40PLXXGT4^1Kpps;7Jt+`W
z(tu^MpVMt{;4x>w0DD);$33)y7Zzdu_dfX;elGv`IK}ox=&n6KcT<81&nr|#Ielqr
z*_^fT<?c%8t<_}D9H6}2Os%1l;9NMj?*`HJYo6w1k~vja;t_(tyS!`1**wAMhprSv
zlPPqSl+;(C>)@7pob?gs=%uM}{Y!Z#qm))go;xQ^14q}`0SHZ3$Y>pun5MHV0kK-s
zNcz(s$ZseIyIL0qkosPps(NNI;1ljMpwUUbO?6vUCKI-RS(FCnWn9#A>SK=j7RJmV
zT-<S+cbgT2YSYrO$$6Y!gt{>xb)Z5tj@63T6r>GwPpm_eO=GgU(_#+CGYcYDSz8-t
z?OITplda9YxOmtoqs~(oCHfJvc8H=b0=1c+@fKe=ihs25&3lfU>#M{QKAWf#DTDn(
z1(JUto&(r#&1?}t#8m$B^MS$CLa1BQ>;1N%pgIrZ{*^g&?B=*3{u)=8`ftENSKAad
zr)nd9J1{gut;GsFIrh|PkvEaFTpWs|JYx@NxAj3Ck#qU%<6Z*^y*-Tr_QP*1(Y*yh
zE*cnJju+HZ!3j={K~orx6bB(o?he7TQ=a5&(H|ESglpxWESZi`tS^qa5?SYAA-4Qo
zF}x>IZwTs@<P6IZNa}t!Hwp*TBL|-^hsaegy9=g-PdZmiW}%ke<)6i9S_9`8VI}^+
z{w%}*YtLJg*U@x6rHnmbCqTQ3Y#J<W%c;M2NZW#ox*w<BrD%A{j$-Gj;JJCrIk|}Z
z%#^~j$f>I6$-jQ8dP*aDpsV9hL?sW5<S$xDWp}lx9qK=c0wpXT^CV-W#u1HDAW95y
zM331;8jH3ap<**9Usg)Sl?-4MUM|>|1q%z~VfD@}OLu{3!b<>yGfA9f@xuP`Z=F+E
zOJhKc`{<jm48jLT-mWxcQZ#IygK0%~k}+Xn_Vyx8ov-~1p10SN223(vF1);cuzzbG
ze*QCe<4k?-c~Zb}ft(@E6$vNp5v?nUMN{m9$Lm_it1{VD0)@s(Nj1z>_{-bIoP6A`
zH5T1^OiCDZ0>1oT1bKgJ=z`J^o&FUBdjEwu{);>P5}E!Nclwu7Rx)>n<8^3`tvmMI
zK6`F%$KIocVKyC1SvTHqA^0Yj-+?UaF*Z=vZ5vqC9j=Mj`@}*<ar_N(6g?C3Dof_|
zKU}US>=8awZgHjT5OySSF!L61k&vfizNjyB-a>MV+P<s9J~e~y$z7Dm(b9i{!FRV2
zw>=A5B4kDU1<RFT1}s;#L{;H0Sgz+59~La=fu`t#^P7KZ5}E$cB>tpg5EG-<9|v*J
zx3qvi?~3Y_+#?tdH?wjIj-&t%=$`;{Xak}mfN8|xSf_5&<$8n#bsGkFfWYx5)dG-V
zX#pbf0O6xz-0`QhLItZv7NBZ$(x~M5tyetDPN)un?B1Yn-WMo{tH&E5Vn<9or&E7?
zqiwHzFw$YTPz9pkqvd}}39%!8176AZ|0!p-C$5RC|NASy1mORDV!y=ft4RI-%LVhV
zHH+Z8c^L-HjB-YwH<CF>d0=%bt?tXSgB?=IrGkt4ZZSF;NMehC#^tV`pK&3%ZF4!j
z!7UJFH1Fmt*#~Hx{j!&G0YIzs%TEt;ozMd9!OQdo*5XOcsu5ze<vsWqr%YVy{Xu%G
z-G3G1S2u33ESNP`8c9S&hkBll>w!5is^s!uRUrCHxY~6ok{h5nFM;C8w+N+B*Ze%F
zlwK*25M4?h%&L4{o5t}*MPdfj&{BXb1vH3}h@-L2eN!*5Lx@O(VXk6}U(jXLFi+8}
zufRyn=@v1fVoSR=3rVt(@s-Y-DeHz90mR@#dm&{K{A`+`SEl_|FIs}%b7)tgr61)$
z=?^Oex8A&hZLLjg6~W`eI$zH99x8`EwXYU5o?>n_0OVCPie>BGHSUGX(RvPW&ti?D
z=gWbbOBtYHP!N@N(|4b$px%$Ce@%hjQvyku84wc@=-V<|)=meyjqT8*N`bIFD%O0!
zEOa@`GSa3xPDPW9gfo7VP+5*NB?qtPS3J@zfWl1Om}0tfxGe2tiSH_2r$nJE-eI0_
zpJ>DVY?aK*p?ALQ#7o}J8(SwCdh)#85fm1GZak{TL}H~3rrK7YWTTX-cR)<cK>NYg
zHzjg(He1&`@N}9e!#i)`cy{mSUQt2XKZXHaE4=J|dB;5^3QLDXbLSdLzSvcSO{WBk
z6S}A<ciyv%^t=|=Pf<l$Mg_4`L#hf?+zBIoo3G;3{UI;GcWN)s?Is3m47>rE`Z$r|
zJ0s4w9lF#bg+5)6K`yp=Yv&#w(s@SoVOg>=bIbiVgD*0KNgnGR?bgMAlwYenIMnfR
zZWA?f4_`GxE{@)3R$N>3otywzemGN;^eI>Eke>^EHjW6<lBpAn_tXIA%RLH4>RRyF
zGjfm`D!XB_`SlV`%k7fp>}KFF5p)>ah&Y+*p$^mnfns5jOl;wzC<E=)V#xW6Ro4jt
zk^&B!bu)c-;Xl^D*xC<00~!`q=_Xi#Jt4%aC-Oh+V$Tb!5je-_CB~=^9y@I`<qMZr
z9ZHs>E$1v<B|KZCS(1oS{59=AwXF)7TkiqVwwb!zlsW07r;=<1Wvg*5P4heB4nxiT
zRiloya@EtC{Hu30h-MrTO4yaODwA{cS;x|j7u`j5NC0MFjKJG~q}YZa^^0~$9FSkx
z3MV_}ei~3w;h`f~IC)dS<&6s6#FPH4mM!{op}g%Fd4+0ymmz9q%3EfKTua-OWm6FB
z+b7<7WGGT!9maE-$uwFOj}KF^rz^~3@;QEpi=;mycM%<Y_&+X&Cs=6nWVS_f4rnOP
zvYJSs+5l0FCx1@vax|DYZ7@v$EwN}Xf8wQ;Y%a$dHM}pK>uJgJUimIM_F|;{uKHW4
z@b0`6tE-^19+71|-uZaX7Erv6Ug?+!p|=uT!#h2t3E#Oqt1M-6a$%}5YoX8OLqJC3
z0)Mbt=?U6MjIx}5q_5DRb0C`6>XxFO^X)!m<|1UsQ$AyCrk@MNu&3yFeUPz4Tw_DR
zFj#gRDvpBfV*R%zx7Jv0;AGWc4_ywP4+NwmcG*8FXg=|$)j3A%ij$&YGwh{jt*0(V
zayRgwOtL2oMQY$v8_v`8I(HO(zBTV}1<UKq5wjH|39IaAbB3m&L;DXM*(M>7*ZdUK
z%PiL8uxe*zFDhLq)#LuS0%Rj!UWSpo?4}r@pu($P0JW_ho68|m`FH_F>zGs4h{2ip
znaIMourecEe&{vV!wzMZtcL{UtW)c~NiFS5*A4gW9BR74SL75)v}s){TWxE$?>_Vf
z`u`=&rXYLH4|A92Dm=%i*kWT{98&RGw9WL?jqVP4k{wlkM`zz+E=+dB9HUW9U7FCM
zad;M|erxd~=zd%m-5-k<Doum$%425Tt@k+4kFu&6_SGglwQ?BeAbfkh287twfIh8&
zX|M$iO?$zlAaa*oe)+)uV~?109}!|DA;NL<lQ$aY@)Zfsl%gI(kJKfqYJ$2KS@P%G
zh;`mQ27I&&sQkYN8~8y1)R9nkKmgmoNU1C2oCmZRa@lT4SXG3u8A|l|U($P;vaFp+
zMoHJxy|Xiy8T(RyAu@k-rTB{xEjn@r7=0%7Yxm!)^Kb88C^8JKW&!%-vo$T?&G-hZ
z+gJd*?x{@C6ZjILk`w~3Dylxs1L)fZ7da+iXDE|7S_W;L&6eYS$csTEYQASdBe`C0
z2p!=;B!r)Y*J^F_oXa%>ZJ5$|-|9#qu=vfuvd4f-3b*vL1Ni&Xv3EANAd8`cap5Kd
zLVo!p;;d5}>}&J^S4i<5hE{66%6=+4mgf<>PGQgTFP<_Q@2|<n(U+^6F2xDY-&n~>
zbd0|<;OCUEMP`;u2e*Zyd72IGiQ9Jdo{A0LtCxWea>heWet_?j0Zs)t4lL_<uHERC
zBE8+zSl?*IHNk{QNXP|_+|TFFk`n;W%fzM@n%@M(s#UAOyCl3S({ncMkXWlH3E_q3
z-z0OJ@AoQKOCA*k(Wg*To2Ck~6}T<yEs(X^3v2Z<6IRellu)CXib#(L;+t9);cd$@
zuearY$JzC7?IA5)giO?n@V<iL6(wLp^_US$p+$CNAC6*dm>UsqN^RI3eBcOmh}1rH
z5{DQ9(B2+PH^uW8j5p-*<nlmRaiZiQ9f}H){ZVbtM5F6)P4qCW4C0?QtX#DQIsI{X
zu1mS~@t~iPW#&e#m{s9~qPnIJo<12dSo(KB@Ddub$M-D(&{&pld(185aC%NH8>@$4
z+#JjV=gqqzKf|SCE@z(m3Rw)!s}G(N8tF26q3p{cyG8IiAw9Aj^ka4vejJe|(kN&a
zyNe43eW`NH&`n)$BiAPE9(6sd8boikqeJ{foo)TNbQw5AAZm`YTZi6UOAK@G6Na}D
zcX@Q#^%T}rfhER9;&G#?mlP^w2^H|T4?tL<-h20D9LjZZ_bFOi&5%W4{f3~yf(^N3
z4Q32xEw7}J?$u1(lP5l6OI?j^Kr*v>>g9JbO2beu)xZ|;Md#Uj|2TG<AleK*$AtDk
zzfEh$l_NR`ZK}XKDhTb9GMVqjyTm*DQCwmif{rS%v+N#Ek|%;u#v4TIIezb{b}!&@
zGJQ9#2Gi<liTpvL56ye${4Ts@gx6FFuL&Xc(SqT^7_jHY7Pj>P!psA+=D`|N-24aF
zQlMyYm$vEHrq$^RMy;~cKwI?dNkQ=7K?!K%(dEmQX6J+<#mU_+cKf||iHK^6gw==x
z%qmSLDh`ORu!eQK29jV0$Kl9Y?IYN&>L|W6ra@C+YS@m|Xa@@iSPEsXoT}!AxG`iE
z=fzp_^Fx`}2g~e}nTluJvT6A=hXqB~R&8-t!fyWM9=2*qsb{MMkS<Qv%xBzGW`rz%
zox<DBTwutGZ;nijELs9Xvtp@Ex>E|3@4%sMzvH{jWN%s(gyeexh8hE$M%^UVF(QN$
zO5JY{Qw;S`Z==!rer3omo1-|W^!!JrW$|6fnG)Fd1Ua4gDNfvv$^hDEe<QyncoV#`
z7o1C1CHq^~A$1IG(WY&Muz~_OS9de?&S4_E*^zBS>h}N>ui-RoJI}~>m)|KWNa)?C
z3VdJ!Qy*{t<SRM4ouvfeAf%?Dp0~j4RFAsedoetk9-mXn_76a)jw-~A=oju}qP*8_
zz>1UCCsecIxI&k0zOOTJ&_WKIciN>eh#l8o#TAUv*i~27W^iR4J!I1cw|AEdj?!b|
zPjp>!YXLQn#Y<}och*AAh=Saspv-xg5<2cl$)T96Cn%$dZjgY~slMCc4|xd7s1E^w
z`Q$4yJ_Wm{qdqeuxIjnt+FIz_O`*GK$Mx-m_oL|#l<aQ2<H0cNgTrl`+NHwO1+Wjm
zY!^<Fs}`|2pe9#Mb?$gHUOgV+s$>AMBZ@2d3FN|^lHV0=>Y|T?IJ#<!w(EX*8Td^r
z(y>Rcr4E7^kK4_a@WT=MruF=S*dR&#`<fu$K$D5$Pye{Q@>Zqdty^3#flDi}Pv!cf
z)v)7ivBa)EcJ_H(w?he(O1C4qhsGwnfIThW*)QJLV*^H;Dc~W)N=Ps0@eg@IK2^Bt
zQ8*ti1`B!FF2@!sqrAL!JE&<iWQ$KYLy*Cp2y9O)LZXaCZ&kTdW5XQVC#1DM?k_pM
zx=Cc#?~END%&x232^TvTy&b+5e|c+M{HnyP(J}pe-O$aFoQT!c*@MUPQWrMr^jQ;;
z?s0F#K#Nn`HK`<b?Y<8KZrYD!HHwEcxmr<ir_f;b+qmdRPwEyCa!I*+L=nIHdembV
zVIT)ttQ{uv{*p9zw<urq+*csVxF-2`g2uKJhf&0NM$XWQ6TL89>6)1Yy@>0~%N@)X
z2ybq++yP>IytJFqN<K3dZw?;{RHZUp=Nh6nN$tzAY5S}DL<S2Fi8T_!I5g(Rb!yK5
zJ@r)YykkE&U!@r87<?yAH*(H?H7<F+ltfn|6^88OYP~2g4NaqP)jf5-IIsm>H3(R5
z*JtuOY(C9EjU~H@qdLE;^t6aHgpi?Z%u0nn|Gg{|++bIn%+Bi@M;Wyt@Wj}A#9%;H
z<W3;?hsHGqm7XIZMkgr?;nzdna@Zq^Q)gPYyQ-g_s>l)z%~H>3e%&X_Yf7Ah(EaiV
zxABbqv%go|mhDqZh?^_GYl)e%mK_!Gq5+80PsAL;L`{T%UYsQ71V$^-r&pfIrj|Uq
zU$oQ@dc<R^+PzpaV&k=gUYF0S={a5`ZZbOPHJd}q`o-~HF0%e)t>HHiUz~^t5tGt@
z|8xCdxk5%#`e<zGAbNDA<Dx!#N&ogk<gI?Qhhz6LkYMF_M5aG;Coagdr{G`i`nup~
zG-g|sPSPB;i#%ROvwzeFWKps5v)Wi+?v6D!OeEHv5_9V_J9gYwLYvn8=lc`^8$?`=
zDhe-M&0al-dQFY&SVDT|_;hB;J#+A(m8-CcIm3QLs7}WLq~gDzq2k_z-Qq#AQ0pup
zSz8rPBxeo8@=7qc9`w!&(KrfbZ0C5Q5HA9$HperFAa>&V&8^`0Ji_gYZY&?xNOU|i
z4h22oT^?gb;E-0a>OW#*<C@Tx0LuvO8a`biwg71|q+r=&1=VL<eW=xyNRy=Ha)tJb
z<Fj14=2uNWdLtLO^N27H_2^XD{k&f8MIbiD`Rn^J7YYIoid3r0V<V)|1LzD`BCpa%
zZ6LMMJ%*%K5>VhiHRsR|A$M_m0=rXy79q@n=63A1T|Aht34ZgY3M;wQbBGVn<ASBT
z+dDb~;?J-z*odSAKmg2r_{<C4ZN{Gee#fylp$dZJizD*@3##Z7YBkcQ*r$*be&j*R
zYll_6CXf3Y?$>HxT6R-j4y$-gTZ-*KJfRm74ip#3$b!sr=yVw=bLbhAVz(W-O_TG(
z!b_=INXUrz(<=;5=MTaZ0NQ0}xL(cg0`Msj<}19e4c%`0;ix4*Uo|g3pe04YlA}HV
zA-I(J>J2(SlmbPyM;?IpSD@TbuX8B_vQIN`6zGt1zZq<@sS&Qm*;!W>Pixx?Fl^^E
zJIZ0H1eZSeiORJPjyLzKB;5A5jFsgGIhm8rryq;}P9CXq>b&WDdgYM#Omm=T9;au0
z=OeKn)V;vqYzD^F$S9Nregm0!t*N?ar#ZW>BQo3B%ZHwOTyPW`><CY(+7&}Gx=I^q
zv4wHv<L3#r`8)=(9`)Yi<s}j*)tJMsRsCm{jN2oZ=`5#uf;3M1?mcEa#0Ts(=J*yp
zV0wi2+&`cg^TkQp6S<fg@e951;mtMJDND@V+Dbp~nR30mv<55T>{wo}4kOzT)WB|D
zGtd*N&=abqpvFEsWSKdHRZAklSY$Tt$D%0UqCDcxbz*H9;_w0DK7+R;r=aV+buhh^
zByBRJFoo_AD&eSgjbdkXu6U3pykc~vB2(luCN^M5ce#0pmfhk7h^^eOP;!xsnX%m@
zA<epIb>!@HKPMR0FcvRMyPi`+SnEYSh2DkyJ&K(R`yYoIN;G&2XYvinR}aRc+d;@4
zi9VL)&R50h-m-l+)JbF)J`uGIB{;055u=SZS<cch%54l!5zoy{PQ@n}j~DfpIZ6uQ
zKW2FU<pXtL&3I7Sbh+2i9GjAQ-B!aApFXgpH0{I~S&|v3T*Vv>H;h0IjKlp<+bUPk
zk6&|}C`=x^;N<wZT5J${{z@;|+g>gWXcaGl&FU=!qqB^3{-<#|5~IvwF@i)Ta?k1y
zOfYJN{fN}#GH?sJjMo@C&_cozeIzv>3!kIbzsM%*5#J{$tctez_F)vkW!*;EJ-138
z)IW<`a)SaraiOf`Hc;7cOUJkjF3l7+dZ6+$%ba!<MGvA|?a!up&=ND#?b9TztEUtj
z6f58HrvyM+R(91ugU%sPU#NSY1v4d8^m0KIUnWK{$nMj#ujg4BOTB>=+=0kNb?+>!
z=pa0OhEHD49tZR-LsY~x!3+W9%P(_A2WM{2R9YNez#T4!9z<IUsXGB{^>M1P{#4qI
zr}`tGpRj!%yB#_Pn$h&X3fk(ucY<5!s*H8b26gbC&Z!Jye-o0)J^G7t>VJRb?_RV2
zePX|Ta}nj?|7Q#4Pi5R)x8LH_Rh8<wd-L{ldHhuP3se!m{Zqk6Rs0)3m^1-lyZ(Y~
zzVHB<E5@WE;x8Yg#DDJt6&^+fQTNs8sVsa~^b#zB#`~h8IUZZn43eOw`ND53;+~$O
zW!+%Uyp(#fGLuG8e+U#Gi7E;LN#)-lA|KKPV&uO;RrpCo)rOHK6`AaeBiiZ)=Z(&~
z%LveGf=Z!6^5ML+H6m{%L_A#8wESUEJWPE24^QLYFv{LEtasesb)+#~T>mgmCcsFG
z4Hx8?rhiy1Pjmjmo%t_2hRXHtI#T<UidC2Ce+PitNi_alN7_*SPoHVxA7WL7(ZB0R
z>j^afa83@2|1DOnAs6|(xL1G#8vVc7GvlX!i&X^@M*c4D9nlB>?7!O8ihKT{AZD5Q
zySVqiFXeA`*8g`erGGR1&MJIH9ZZYS+Wn+-KG5`_9opyJMv)Zx+cR~6JSaXsFf(Eb
zm)NS;i+GD5ZfJaTzr0p`JFjQmxHyBa<CNkoTQDH?LvJ<FC_?j;78q-2uS9?vcAxGK
zF=JZ7tTSr(*pq(Zj+dmNTi2@y&VK_kUssiZiscemsp*ALixH{~Wld=N!^NP6G~SJV
zlPLKStnz}0HlyqI>9`9CVo)SC14KXrAe*KKxr<QV5d9-WINA(8nuqru=g*=ImcX?6
z=_jzL-#cGv?ODx8oQ$z=@Ov#E9Z^1g<S1gzT9(a_sEHaAtS<*qpu+_hJa!bS*g&KO
zHtCW9I)vtxrAc}ZoOZvzy0#^cH71Ac+X5t=1<Vs~=5vbW9~xzBTL@KThcOe?nJNa4
zySY7K+!}i9XXw0nX6frPFyl#syve2NR(vT!;dE@3=Sw(zqfqeY10nkBO(u#dgCU5y
zKhl1)i81aH)<lND7;dfZdPy-o2}N%bA~<@u=ST}z2T<PC5w}b;@WmF8cHoH13%A51
zH6Q8XWd@5H3B)mj%vY~1dIEDotjBo6&F+YSgkp;azJ{e8z=-y#ONP1VFc|H|no>41
z-b9oUdw+oPYfb_O=TwtOqM$cPEj*~I+%?=PxoLV5%zs{f(KLze8SJNX`;h?d{c3`l
z*v&}2SW{4Ac?v6ibi1{$)98K0WIIQ!?@8yq+kl*p*r!qzy<UROrm@16sm7OPP`4m1
zs;UUorLED743WV}hLus&WX^=L6E;B~9=EC(m0ZwQBX@Dvmix}TOMp~4|0oS6m%4>c
zE$|AjLz+)NXIsPMBxUfb;)snBy^B!3txQl+LD83X^-z9q+8%f1oEp|Qv17Gg3O=z0
z5;R44D%@yuVr*zIQO5)NJ*<VX%!QvQ`BDJe)&R<i`dJ{ms1ra*t+Tw;Pg3v?FQE0;
zH)T;xFo1FrzpJ<<-xmgOAk16;$(tN7h*-hWzBGXTi@G&<wSIRZhX!iAn7;Iuj7Gk~
zLv&N6-^QS2YA&4xd2<@rxek|QD^|XJ^`uV}l6x7G*OV^;ziZ)~hM$ucO22xy3~lhQ
ztlQKsd-8j(HJ!n@h$~@tFH`|-f(d)*gs<y5G+(wnB91y&L20BpYXL=$<@ajL>4W7Z
z%cut>fC?8;c%-}07xj|pU;+?x`TZ!-ePUVH$rqKTC#6|p!vsxkGH856h?Ae~y<kq0
zD)DuA7{yH}pI5k5${y`_UH{pmgX^C?@D0<*Jv<4~o|%?3<wh?~E=;BEZ)ZzO4|)FV
zQa@CS-N78MY)^Y$)yJJm1SM-nM!Dk;Mf6fX7&Hyu$Y!6rP?fNdU!N3B0BrOc{@m;D
z^Z2xP;EUHiH#Mm7ajv)0WULE1soh_KlM?`x>`#hb=dc6hYSnQC+7PeLZJ2<;suHd?
z?UtG2>oTOGE`LyRYhlX#L;51NAR;GnAF)<6Idul(&Y-ETCRRk?@BQIrhpF@xok>vr
zQ5)W$+IRGT!OAkQ140B%rj`H~ExV;rf~Yo&uxtcyjcv<DDXcq8-C9!!?eyh%fyuvi
z(eo|q1~VE~HgKO>RYD(i%Io#}oI$LXwQl-QD+{7ARzjC_rOq=hIJUh%%BKJV?cz$^
zO!<zYE)og^@T#Bv=^6Z3&4N-#^a(7NtYo>CbrF<Uq;JZ!T8sl<tD(|AUp8MF;ONM9
zgRZs5`TP1^v;)T=(mV+`2KH+l1H?m$yw^!h%7Vsgie?ANhtr*rht(2#R9WDWCHd1F
zZS81{V5$1^D;)`5@Ej1bGn?^{_BOj?T57WG)04o{h(ewMZ14l3L%+dP;EvsMEI4@^
z@j7aGzpf0TExOCmAp>t>_|G@7CPj6&BfT<ZEVOQE)2#5lEb)`Ou~d5=(be|4>Z#2Y
zYcDHKcolf;Gy$bX#sU4I`Ji}rd`*h23~>dR9}gOZRz75_B@+*54jUks6y#m2pre$A
zA6x}{kzWgtE1@5n9bd??`lLmgJ$j-i47$C56!Yiz1z}C58wuA`ZZ#{`p?w8$eN^|)
zA0u?+Mb5bpv`peEoa*$ttW%pkK<7O2@gLvKW<7RE+uwh4!Y|zUF{V$%9Yp^UUx!3x
zFdk9(&7r>*8IIpEBW+BIk3Ly@c@puK9izWp7^oFTmSUIUy0YBhTox*5O-N#xG+63R
zc9(nexdK#6JsVwGjq#@K4u$G~Uhw%+xUbn<DN!u~*cbo&SphgT?<B%ziE3fCK6>YG
z@>KYk64WwErM#$_q%(g_@!7cqyF9hVa-^4JmL6iwiPg0rFZ>kT)1dt2C3NPTXd~CQ
zlN#Uyx~Oo3Pw{O-xSeon1}0-M@^Ja}V(&F793rGRgJXrl_tY`nl7t1v4{uPGh>4`*
zxO41c`_mO?7{`}9pyvL<-|K}uMool7XIIUbVjoGSIn4pVUG6*eYrE{1>1$tAXRQVI
zufA#!)grX;qMl!h&@?{c2*Rv9qq|+XCTTFdg*i!5G;0dcg(jV}P7EUA=S1>K5X5?P
z#qr}*aXkq-vz9n1+k#Gm#|&@N;=H_mKA?Y9$v)-mQcbOk#q0$iidSlrek^g1QtQJI
zfV}wNdh^i?_Htc5w8q~GthF`41sv<mF&~0K{ivo+gq~Ih!FIRwr(}iLR4kpeB@Vj?
z!lO5z?=lk9oI;psPlG1YN!oXuHNSV*3l(?=5msLOp*%}mO4DDi6*LAmh}Zo_=mSoo
z$HJZo0vda>RP+9-<t26v4ghmjnx-Kwz-Zur{EA+6nmytw5f9Vxy5S_spz`k_<FKae
zyq)+oti(j8kG%{`EZ7Q1T8ynziov~sjYsdcXLy{&d$QcFbDHWuzqu;wq6z&7l;htE
zz@65RN8L`d9tbZQcp_M7Q;BRx7%>M`Xr=lWD*0}}?JOP!<Tvi$uL#GeCJ1PH`JgUH
z0b+|F3yr0QeQC|9=+lKC9=d<$$V|e0;GA^v4&ocXJ>fg5f~L)lCa!v<;}r``N48S%
zGaR0S8*n)pHXap;9H=5B;xw&&koH}5zyJ=a)L*ySea5`G$|mEt>hX7Ydkd<l0K}j&
z_LM_ZP=MgrX&n3heEIrWXw;@+H(x6ZJpW0V#gnl?tdUDz`}vZK1K;I>Xrf}-71QqP
zb1?I1?#A#gE>fJ$2Xb|aYBjt5cH5im<)RV!ofc7N>^HUobkjCYma99;*NgxK<wors
z-2tf^JH4N8{H#W-C!tPX-7~y|f>_k9jmSpV;~CK!($+85+by!*?~p+q_0^L2Tn<^l
zYc0gcTe^_w`V%sUGRuSLKG(Ea^W++Sw&_9fn!(NjU6yrq5cnXAqs<c`oU3$h7^^a}
z%xLBsvv#mIeG0(aB=^uj(h!}7;9@n!b$mTGlFZfy(%k}uXDR*ET`JBpm0QiXyy~^U
z^m=-#&6H~s73s9{y6yn9taihW?qnO$4KP=2m+EKpq47k-m!+f2gutUjXXae0$z0>|
z2jGfHf(dD>ZUO&bIN38v%8#B;=@fP5oC_rr2!z>EtJ5p@+Cq>>1x{J5U`kEoBsSHJ
zTK2T3T8wsA)m7xo`!P6wf5mD_jIbrk^p><JV(aq4cqb33HggMVN|E0Kdy@W|o%OP5
za_kr9d*A**U~4avhbHORBAm8VB5W!%Q}=qr@@^~lnn6PtY@b=JH?V#7Y)D~`m`R8Y
zg{dfq`gSVzeQ2r?azPyLaTy%vP3sZCspP`}jYU5-n@zxCTHaOEH0wLMgB1D-GiS^s
z-6F}!IHG@KglUR*xvbU0g#j6@w?@Y5K$p#CwCppPZFRh{7RSJSeM=F&4xE9f#L2c1
zAC<B|Vj1}tr70s!Uz|$uyXrm&2C=y78p&HA6X>ogDlj1qqviM(;@x8^jdAD8Pjqo~
zHBDy<S6jJJzj_*(VX^S@+ZY^%xMIrwxrb~nK?Ug*U*bXE{&rnZ2d71hgc6__gNhr=
z6%wHm=?ddaFdNywg%e%w4w_||?>e>>)UsaH+}p%<D)05N05gcRn~~-nNmk&sV8d;y
z(WM9G-C@F`M_f}r5`~aKUEsziDCS_C)4GJ$d8*SBG~`30ca<x`33xK>d$T8e!8i5X
z+KG(&3U_hk9Nb0XuOvT}#i7u-WzMG_Qjb$K4g`jv0yjGkaKRpKiDuT}DNkQ<Q1Jkn
z-nf-uq>@KQF{x(j7hcQ?%w*i-lRos;@wwN~p<?j)m3w-OPr(u>^fFeLrE|2Qb<-YM
zHj`AgR150-;E3tB5C^ro%?hCyOH0xX7DA#J?)uu($qbdGu@MYU1K_(5;d>!Q1L`6P
zFkp1>i{Bt3P2H}GDyX3$<&DoS!pOk7BtF@<W}ceM-(S00p6n=!ZeUi`$4z+QZ@e_s
zcPZ^h9GiD(*4g5ONwu6_cJYJ<l%j9++@~&2;utN<1#7jhyIY%^Zr=RknG?9zvaLsU
z6JL(6Gcp<%J^2kbQkHaA31FDuRH*Vg#}UD*)oA>}Cl=)*z>zz8Ts?NB4v*ku(d`HV
z>x?0Ik6OnrE9#k>EwO0!1*k?QX$<#2<?TMrGY7UeMXH|>NEP*9Z6lOu9532*FlZP4
zLDnoeLZb^bUm=ahDpphU)4Yr4W7>U5`QXb}VLles64}@M&t#nTk~sta8D<<j#Cw8Q
zo%j0b-N8>^R!bZ@gg0T#X1pvj9Xq|^Iqwla-_Rp8_&K)FxsM=M+8gCIDPTWh-%vPG
z*2vxX*@mfDVHBk;ODE>pbJ(@Z1B?HG=S8Ck`7UR5aRr0zq7$yd9>V&yE6&S{vmZ`e
zvv7yOmxfA>4~a`-?P<yleAayy;kmS#bJ;lT8E{zI($=`cm*Vclj4hsK4oRU+3$I>u
z2p_LOQ7w*R(6ZxmKfisfY=U3XY3iLzw$+_B-^0qdeK1?AAihZ{lvbt|%pxvgoWhgY
zYgk57EOHr1bU-={n+~KG)2gt*d=YhujoA*7?e1dN*QU+q-l)A(e2O|p-X6Pe#w5lg
zM3-e|X?`-YTRx7Z&oGpQ`ix%5+ZQkqyjQ+Wz1FEd(@tDu>_%0ZM^B=xA8(uL@??E9
zpU?G!OLPKuOh?mlsOCjTwYR9o>AiN16nF`pJTrbLKd|nOVy(b2{5_0JWuUHIn<CX{
zv81BDbd1Qt_rky`v#v6|>D7lg*;mRQYah)KGekFYo}S-vOf{|Y$8-HOn$~RmdAFw>
zj>0i|aU5zSA&*4|Q(;k$#S5$r9*yNdVLNLR!AD*0MJjV_K%T_p(M|Ed#@Y5}vRIN#
zyy@z8;><$rse4m`rNiH@R$d*&b!2if?RUX{0AuX-uS3~>g9Us`c>XM2^kHImNgRjc
zv15g`991nH_DQ2&UMs{JtPf9t?$CH)ils+GIsWQL(DxnWkLS02N%gtFWBgVJ!}Ybx
z<w}yhC04S1O_sW?xM?mM`Chk}96F+T_$<3$xdu0evuo^<mUh26JJz02?#{HMylWuV
zsieCqcSsOZ5^!*NvLRL|Y3dGiF?g@NahT^JmED<oITIh3BpYe0U+$@vy`aG;q&I#m
zQuGi2xtqt9#+o*Bh&<Z0@SUrh>aIsRvQQ3_FT)$`!VlHShR&6FEakxX?JvA{OBSTc
zUgleJR%vAFWn@TUxORw*Usnlb&+WRSYc?ovE6lnWFKp(}6ObbYCwU55=Jh8o;}sG2
z1)I6i=981t6_)DwG9P$gCio||(b>oKVetAP)A6@Ih=YzOs$2%=U~(n;)`YEl#ERf{
z<j>I&jh0;NKL$SuZ?f(!QhToSDFCL1*J4#B`^&wA_A;W;)%H$9>XGL$)|i5`q4cf=
zr4<7mbB>M)<)5tv9d(~1OpaWwCt;~4fINNn6Lw{JDfAL8H;F?t$1tQ7=H8qytjA9D
z7I%Sqbyp00Pu{Lm)<4e__cnfdQc|JP=iSq8+<;@#9lUXI;sn(T(btk>hdmvrf3`-y
z<Jterb#*1kjRHmU43pI@A=V+47Z7ZX-R6GO1kj7c(ZcX`OpzlFtC5fLd0So6QGg$8
z<|ueZEP7kU+sI{w%UR}a6w!7_hLji6j9uQ!YQ;63sTqVR6NB|)F;pAg^>FX8QzPg5
zbxaJM$Gc}0`_DYg@)1C%ImA9#+kVd|Xm2bu!}hb@(sAj=*z9xPlSjA{NQ<-VnCx4=
z&eyCgG|@09`{JPiW=Tg8hEwH5x8(Bm!{t?Ion;1PYDn1Dt{g=hBU@Iq)~mM_pUY9(
zUDR|s$5|BPW;~-i4>`LXhlo<L7E!7kMmK#7L~ZeOrRv;RC!OTFR<Z)l6EV}GHkj9C
zgAiRzGWE+R+U{f%99^^TDGui*<z=f;U9BO-3A-pSHqOFer&LZR<-lhI^!NMjy+w{)
zpk_J|JSc!|ez&A>yfwjNa7^|XaVFwK0<P;wjp~6#a8S>IhE#;8z&jHYmmeK{rEc6V
z6fTkTimxjhpCl_LHKNW`<228QFN8No^xi*()-Lf_D(I`YG)m1)>MM5U6av({z@jUs
z#mi7<iZL*-#GPPws{dSNkIo?XEGk?6s$^eE)=-1;n`4<akDfXoA+AK6)xN^1Uluc9
zGc{0gO;lQrHg}L^6C&+mhhD8p=2=og(`ZYQtV-;3a?BGvJ&ioAbLsT_WPfv2AiWep
z<SX|Io?hrFUNP*7c?XJiD1J}<f_C2RYaJhB1oh2`uXlWTKs$gg3|7XM6A6Qf#Q1?D
zG|apWlYVj6B3mx*B4|7t>i_XHPQ;D3R_xhfj}@W|ezGo-Gtp-^O@Mm(buI#neLw7S
zDZ+hWa__MP$Bee=H2jLmKuDtJbMeolbIDT8{SX)}rHEtj48ufka^D(LAr~vnf21!W
zbkdJ><UmmfF%3Y6?ghubC$H7slzVrv_(}5RvJz*At3M+x!N=slOmqMz;Dz>F?vRCe
z#sv17azU#TNl*0;VO@vAL--w-Pgs*J0OWEkpX$EG_jgC7N9*@ch>3^|Xf!rl9q@YS
zjbb8=z4Z&@_0$Z)MQO+kG=)DZq5TS)Gs`vw7(BVvoGF2in}6e)f4MKCizpL4RzFFN
z6L_b)2pDori1AK1MS8Gz-{-GCTc=NY&i;79v|@5~R0f|J5Pj8jFmiigoJ$fDg)W$V
zZbp!eE92f@)daJy9B}1+@KLcb@PGe{s7`i@Z`!owh4iJ1s_PH(|Nk%b|LTkM`;m5!
z;JML$etuEZ$1iT$vhVJuZ48?>?-bv(#cT^nJC~+HGl(WiY3u6Wto`G?uiLH%kxBVW
zAV0M6^KC5e?w5$NyGs22D{%(?&4;k>gJ16Z^;7T*v*cL}nT`H&8^4Hr#!iL1+g)s4
zo%a59?R|H}J+?r^$$iPPaXZK_+^1R?+3xal|BJtT<i6AUY(#xannHK|c}M>%m%nk5
x?l|i%#<d5u`W2MoH|@qRrjL^U^$D$PoobZYNq${rViWu)eOciWO5(=-{{zigF+czS

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/power_systems.png b/previews/PR3525/assets/power_systems.png
new file mode 100644
index 0000000000000000000000000000000000000000..b84d9249552dc33336fe2bbc9ec7c245127cbca5
GIT binary patch
literal 23020
zcmdqJcT`ka^EL`7q9EWXNf260AW<cVq$Y@{C?L?}AR>~3<k&4DpeUdsL6M+j5SpA6
z8bM_Q$(bff4h>CcLdS0(aOVB3@4olH@7}xCU2A4~fOB@JUG-GeQ?>h<uC^vCGcPk8
z9Ube9>sR&Y=orX!bo8kQ4}jmC?l>?5KIlF5G}Y(|+fK}bANJd+-d3ffD~@3yTQh>6
z54m4A@uZ_;YlQx#Z*t4IPe*4Ga^tG1fuALA<WPqgBAABalonyO)Q~O|zWDK(>D994
ztD*4Qk8>}@T@__0Sut1fxr$~MUSxc8(%5`6m}yY8G3zABC-mUslftBPBp*q4Dy{ut
zolaeQb#`FD<}z{=o#)mn)!nkO@jjc>tz;Kq$q5!hSH=|r10VJm8Q|cf<KBKG_(=PH
z5bTFe<@sOR^mKHOxK88ufgi=MjWdA{(--yZ;G>M^|J4`MdeUQ*d1qRaxD74+r!P45
zeZuD%)r$0aw1{&Xn%Ks<HyjfB_V7Y!F6*ziQg(U4X#=(i7F-Y7Q}_+dAAZjzWOmUK
z`zFXc>k2mavN~z6DsqP^h!xgY(=;pRIC>7fjx9k<j&xh~Yzp=*^N4-zP=M1uMq_?D
zqHKu9i2rQz#U2vDz<}BB#C0P(dXviAnfq!tXu{(B!VHwV`G-$xDR5zb+6!frYB7sl
zLoHs!gsRDS&bGOHm)B}vAT@X~ds4?uXjbyNYfHZ{QRzO99hX#YR^GN{5)WXd?vJb}
zt;4xz!Jwh;{-yj4aV7o!S6FY^Sv*PNyrYovr{A=>@d1V;*U#RAo!iVP;+ccJWrFxc
zTK3#Y9jq~F%yG@>+f)(az=f!`OGD9_KVdQbCT0wnFm}>Zwm8~pK<q`{Tq>N`<sf;)
zm`{(s_G2$y;typ*!yO65(2k&u+>jM&I>!52{@j&Msh|11=2X5(CS{?!roYzQ*pB(;
z{lK)Fi}0#qw)2wx3DnFG6SMf|ADuSyWI{mq4r#g_A@RvxVDrR$RYlMh=6GJ6ma_||
z?k-xzuudv)&j)IbTfM1e7b#D-?mRwLrj&d8jrUfOY@V&&=T6czO-Kgg%5<hrfRe57
z4xM%Tv+!HXp^wzQfaX?s{{x&1C&HD_8k;`s`KADFHa1ehl76o}lOaGOpZ-D=PPfdH
z?*LAv#iyA!p5-`~K>2C$1))L_Y-jxK>@GGqccP3AtB!g|oUn98Om=4f%&c3;hn%kd
zkA&BONGoRTYQii$cXfpRlZUFk?tZGqj(3;%Eu#;c_i?2;k!)hlRSfy;M29%Wt&(|z
z>U#5XaUz|Cfx~B<41O`~16JA;LviK4_A^CjM4Ju)$^QFFA}8~K%tv<>jBvk#$2+6@
zecm}-;DWs|;O&3>w4s8%z51o~SGD!NFH6p4DVmMW)gB5u*kY1xzn4fL>*i3|`WF3-
z$!*(y{2Q3%)%8_7)$(5fzrq~^DuQ?=r^mFSvP-6WO0ZwTbDJ06IRu&U9$@(%&>ozm
zOWIttLS;KA!kV(SXveBIGgg;q2=vcX7o)&6S`3w@npdIGs=nS=GC<3>H-{B@y7c#z
zx;~$)b{2S-iee2=M%917=u>OU9+ruTv<AK@`&8jv?ZVu_YGzBP&#~r`{d&2KWG$Fx
z&y>lZiBs&}cNk__=Be$whBQ^uX%62>u9;Q8UQ=4ZAK3XNzsTQyP#^KQ7?v%3Yx_vr
z%c-4yBiR`1ytSGKyiO6es%0W9pWbD&e@m%l;uW%8i4t*<)a#pZ#=dlF|1lIqgiF1s
zn45h?O0AfcFl(2MX0z|YJ?pG!P7$(blC8}f!>M7&p4_!tD#t6D>9V%@krH^(GXwat
zu9Io}d90$?uh?j@kr1No`JpRLysSFDYg|t}KQxu@WHa$q+<I3xSN-`s3&ytbMe@iB
zmzeVxFaGUE8MAy5B50?qV#2FB_>WPHdFI!?278kv{7N7ZWf#NE{|HMTCPJb0(<%id
zNmK2kv<;f8u}iVDz$V<REQRy63$dhsyh$t7jt3it+$JQDrs#7X7hHODUb!e!A?Q2{
zS)_d>BhR2FeX~XH+m9RUvg9&Eo_&)ghd`dN8C3+`8lDh#toVpASBb~Uw6KA{g;Z2Q
z%*P4?uI=YTw!JsUd;=Ae(RPy=OnjSuy=&pUTrL9D(<__s#e@%7p~F7zu>!{4fzfw5
zLc*w(gfElI9V(yq^pV38wlm8SrU*xi#HDzDIx6t|$7+mGuJrImiB11vkI7}0L)ACE
z16`0K^DCDSeW*(R!&-4$&RNA38F#JX3e*d9+Gh}%9Bkc%Fj=u=w2^eQl4WMsK0Gi{
zh5|MZf~ojiR-75tl~!6?*80&OS_B>&hpbVf4~Mo-om|R;@pI#@Q?%<+OZ5;w9U7AZ
zlMeWtyoKnI8V>J#gUbFsEtR|y>gQG)<Ge3+Qc=_x=VSw1pv$CFa6q5L^O`9yeQ7JS
z{fR)uepFpjp39xaNVPij@R~0v6_@G4)tZm>Je=gFfY7O@(-;uhm-r*~PTo=7NS}<*
z18pJ-a~~fD3|7w?3>K4SJEE^Bkqj#ikHqURU6=jhn2@q@cr9%a0}CzOk}X+@&B+tE
zirqF%9Zd@<wXd`6ltW~R>t%`)3q>7Un5*WWJub1Gt8gWgbj!7Lh)C8NYM|;sQ&BO5
zxhvUYIF|H}$>^w+9UYdMu)g;ncLFpfxBENVo5m47OCvwd?;dl5PUTDc;QH>>gJi-4
zY2L^oMz*jo+fGo|B^{=*PKf+b+^=h!^fIrDEB}Q}qi4$)j&^5pBukQ|1Nk{Fx*vBX
z{Cfn&l}1xR0GB1!;Ny(&k+T(SZ+mlIY-6Y(+qo_az;pPQ)JJ-@O|Fw11+cLZq0D_J
z`3yJCM~=;Nhh2rIN}oX-(rq-HR!VwC4*XCwD!zfOawsC$<&v&Em3Rl8aPaGl1;WXi
zn`Rr6Q&gd#yY1o_3dX@q{~McWbA_fXtX5}U4pG4pdTPnW07%^UiH}#EQ;g3`dF-_B
zH=<e0pteEWn!>|YNgOSIg869m5reae@llfz4noVb=>BZmPBt0ein8YDk(ETsTb0&+
zv;v|T&5g~)UOHZ2C7$%7vmm9B;c0j#T5>v79TlSv>0=kkZt+^?mQt{l*I1=!LHnEX
z+k7dc)E>EN**mgU`qI%umUjX<D5&|9SFxcRcQbC90i2XGMFv_pm>EsSkHy%~V&3(^
zQ{i`SR%)6eSx9-q^oTfI!%4DgoD26OYwKMP?g>+_L3FZ-u*z%LD)>$Mj%(CJ>D%sb
z>H^JOGPPt;sb2d47g7?|&1}zB(c3&Ba(gl~A@^v>0o%n~vw4~E_LXe7iiBuE>BdRp
z4a?afZPRL+CSN?uuPo;i;yRU^IVJTj0EuCK-CgTLH7(+1oirQy!m$S&aq*|Kn|-X0
z%zL~+EI(osRv%~5y{Vvwqw99hPDTrKzb%sflr3#e|1E+a_g<!UCOOG+y?ixtK(4yO
zmE%I0=Z1B?WT1o?_HCW30wT}?Ir{Eed#-y1Nu$M$$w0Qi_4(r$eXrR4d5zj*4~Vv>
zSjDkKH@>WR_EE%?Nh)siaxDtDh)hDR>_odFcc)?(0WGW7C-JFZy3LZ0FWX+gvC=Vn
zUNu5IP@-e2i;M6DIclmIpsa+pOcD?^MYdvp6gzk>>a`@g#%%a@uJI*Rx@vyy3y$OM
zy_8n_|4AH)rv*;O88x&>ViY&K)|*`_N4VPL7eBcVIb?@96;_{9&YEafCOMNyx!Ebq
zOt2vLZ8VLD#;?$*D-#iLvbEp#M8xC;|A{c}2Ji&0?1)!4o`6OvE!B~hC$f@UnYTs*
z-Ku^^iZ(=D9_6tNn9pn*KxL6c%%a0sxP8L!>MiuJp=&pXwiCm*hS==dQ*JBlcYIRM
ze)n3ELA~D3$BZluG02T{`~*8n?#luK9Z87x{c`$KX`B2+z7@rBdlp5V#gD8YFCP%&
zb0uU-2A_8>{?29Hi!h2pNmdNp>kdcZLt%$SKNYDNd;C1x7&t~6#k<iq-K@)d4Oy_R
zpK(2^VPWE;w|-<`KdAW(v5oDt`9^=?A5vw+9pMpYfIm7e&i&}eosTcU^JceMG|~R8
zPFW*cgaspvykX@-Y4ghJ2}8Gl?xY5Z&r3%WROz)^?%VlaL^_aojGhhE$D6rcDF}X1
z%D-H*SXich_Zv9n2N^d~?jU})BR5Nnwtqiw8n^lSH`5tfBKqTM&|np_g>vE!DPVcM
z9S)ax9wz>R8;3_sl%s>i!Gb{Q@#whiQF<-GYwbzlZwo3%15AaR8MCb8P;4~wg>^|&
z*<{=Aem%MTL)o_+usj+&v-tM%!3Rm-Z9nicq^piIT}aOmv^UQ@l3tXFos3qH2>%4h
zcU3G~O()=hSK{<+>VbmteoKWyZk9Bjn6ls(`3pmA1Pl6qUO#R8GV%V;i3pg$=L_3M
z?q^CWn9}k3IE*Uy%okUFnHHb;c1-8R;yI<o=bdT%65vQ%-)Si9zXlSaG{QOll24*x
zIg5|OU@VF?<t5gQVz8N|D|PrV$tNt5UoJ?*#<w%4>u3Aq6q@%bpBrSuY=4*RVhg!=
z62H&B{6>VBsCb0qdh6vFRm~91uR^S9T+!D|S|mO@C&E!kb9hA%vq{Z&`<?>+p%5Y5
zk;LOb)}{q)s#+~4nJ>EqVzF6aDJ72xo0XVg$ALx5tr|ZIFUp9Muzr8kR_#jggd}ZE
zax&(uh5vNiL;o^opKFMgzu)$D=ISf-#8`>fZgh1IT{lTGeel*ct6ZI;{e+su&(I1)
z+F|N2{`USe;(5_+W8Jk(sRAE|t8vJjxwRHac1>8Ym$6NhOv6v1)DsfbQ?IufwfLHD
zb}vN4Y>uI5+mkmy>cg5!?8$EUJVsf@Uq-E8ekkBO>TgHp?6phrKPRE(dK|%1Hyn^+
zaKl8w07fd~U*PsA;7{>rK0opMxuk@1L`@2+ke`YhqXG>{JP9<UubJ`LjW@}~J;!e|
z=1B|Zdd0oAiBXab!Z}$zPgwjyTK+sKoeTaexm4DzUCd6^R9<_9r%}gfED-_$_M3X)
zI&g|t5cN&GX3T7xyv=r(w_<LIPY#_`4_m*5??^Mwy@}s9gP(yF+WbxLIki`P;P{R}
z8DzZiy=Q*11{NpcO&xOwQi(b!qQ3buBMa9zvtpHu5=_(&4A8o)tZw&3HtX(f@4GQr
zzfC_=o<jMG#@SSqGM0@%c`bs|MgQzz3wZ#YJ3r&Ik_#7eOKd-H4msIu9q=xwqHJO(
zCqEuJ!`shv8qOa0s}ye_9mQHnBAVIV>s~eB5h$?3j4tX`I0J*5fyPLdR=ici;cTFz
zxO!F0`(cMk-)u^t1eLN>+WupD^Q$HbOOd|1v+A5E$u~o*SX*!A3>T_I3YkFfpMu^u
zf0Sr&en8j&2TWsi;f!0*Bwu6F=tdTP<}&8W=t5mx$M$+E?F&~+^u*7Cu^Vya>cOw#
zeGdLo1ROy@EX$EtksW}F{QD5gTCWXLNyD?hO^klF+L?##JwEK!eS3n{%}>38V$#VL
z;sedenf&t?<I@{)=3bo|7jMls3)gIXpjK#|T+oWGLARV(cpdgSslAH1@lKOZj#XJE
zKZDe<5SWTnkNExEDe`)IDzi}y;{4s6v~u!VO9W6Y--qPkpdSbxbN!(YuH8O4TFwi3
z3cEwpy94f`?3zVr-+XqeGH21JAiQ4ioH~rTa@a%b1pIR9L(T*J8$N!E1(iQmJ=c#L
z(ep?^TJ|?%#EpbGBaiO8)B6$}2W=Cgf+=TrpH47C$4UUikm81xH1)VKLu;cYv)X-n
zF?c1=3bNKe!|pzq+r<V@c?%di7qsy(d7pIMpVkrlT#W{I{??n<dj^!w;?E1+8$R2G
zzO|y}C<<Y3Lv-QK0k&{1oG61BkH$BQ1?J<BjW+ZeS@AcuLqh)AQ_U~nHy+YGWOYa3
z%gj5c-`mwYFY@UQL(3EGxnhs297A@i#XUCUtr;cuRfa?EFgg_yZ>U3n!h0DfL&$~w
zU*0NhSAM>~U}GgS-c-aKxVAqMzLkuf7!6!e>Rk@PuA{pDoM)IkQvom@ZxQ32wGgLg
zXX6iYTLnG*vn_Rn6WV)HHO%qs7<$;m&wa>np3;w@5+)wANIadKoslYRxN9HLc%H6N
z16UoiH=F4gSZc{cb3Z@<0n{Oc2RIq)3uTQH9FHdU<j7P2{JD>a?s<F8umaDvD?)$v
zl<TQ5{2<9CYls&Bh6zAPUUZ?L*4ZSNia~?KlC(G$GrRW>uk9_3sHt$ma&Kur@&h}D
z+#6Z5)#69$W@ONBHxn0n^mU?k7pF#B7-hRWa5wZO4i8Y}`R*?nE|>!Yh)Q5N?0w!O
z5LeE92tjqrKc=I2ac?D*mV)-0HwG>dLdjlwd-LtB{fqhC9Ic;u`GqytbM#CGO0W=q
zLGQ@jw9Dzl49gd29Rh2whSo00&AXtt7KaY9HLWjkk2_FsI$G}3?z53;GKtWmH+p^i
zre108Dz7H6uL(fX*?%#&2U8?q$kPsjEzJ?)y<hy-cTc$Hph;*;LI3W$nE1VPp>^Hi
z2F{liDE*HX#t2tJmzJdXT4KYs<yFg;c6I8s6&1|AHXYhWC9S}kw$^zy|0)STUFrH6
z!H=DwXF73Y_jh3CPQl%o{c`k_kBFOB6-iYdp&2?<sXa|SCmv<9UTkKALQ&UKQ@IKa
zxQe`-jBXCvr3U}rwyd0mUHhvH8Zy79YG~r7Xn}MAt0}3}%n1Zx6YP`QDJiHPmsdUq
zU$6m{o}6w_kJC%+DqxICa;v-Y9=M9tI(BP~%y~-H%<|_IHWIGV3X;fgN_)_ZRJU4u
z9QbE>1u%LKgId=CDzb4~c-{Gus=RXH`AZqaYJyp&2TnJ^p1+KL)VtrJm|vMVm%fVZ
z^^x;k9+p$a9kFNMlg(4tO!xbO>Raq^GQMc3q$Rw(&s0=7%UU)iwXHd;zrcU)L&WdF
zP7E<0!@LWkwOgZnFsp~(IZNqVs1;KApI7Pvs|bB&jZ1^2laoGlF9WI89D+_&9*b3G
zoaO2kYWm@wY%AcYZ9m`*`O18%ngjxGIYs<=5}cCK?n=U*y4ng(ZerbH1Zm#rK#Ri)
zd~RPJ2ANM9C*I61zr<|3u=NE1qfS_^N$G<y(h3I7MjL2SmL*g*tHraM=W!Yu3x+Fw
zcC!fbXjgjC;6B~=uu{Y^BFVYr1{h?ci;bXkWQ+~6K(=AFg6BHR)e(2-uEhW(TJ;e8
z`047%$!H)W{8x6QyU0LJpP5^&oZ{u=IDhhL(<J9@<<)08w{D|LGnjSU{9DYj9>=ra
zRW={25>fc<I62qHfa)9^QH?CkYfLa}w>{g(2FHi6->RkMA%`ZK;MCRJc~2a^5k|%L
zxI6jE7AnO%MR099M5uTI9G6ZBV`4mtlI>xL9N{GtTS2jh75Tf>tuQ_o`Mx+c)n7U*
zp+EI?lsHlE=VDXI_l-fooYnA({f>O8qs952@`e;4BiY*lsFm8S=x(Ru;U+w+W*A4w
zz6VQn7@6A+5$VK?L-pH7z)Y6BXE<Je@)}cnP3;@(nFgV*NXNF+I}y|a8CH8JYml3G
za;#yDMj6vdy0%{AKRD*|0MYU?r50znzu>{meou>`iGA#f7jzX)AQA-VCP6OO;ir7n
zSkAQQ?$@%?52)=%SOZ8#q`YMjc2_-BwI-itU6%ZUSCd=+CDq6c!bbd1f1PZ!T}ZQ9
zCj^3SxH=+U)V>x3)!jsOqw`5k!VDNIAsz`#ie@icyWHf9;>yN5`;kL}7c3!MnpJ<&
zcAr?tUV|?E;3L}`aXgM8Htd{I9f;L`%wO$ZD0yN<>`KUDyaX!=+G*;I`~8&(_QK9)
zj7HtkxowsyUF#abzkSFgeY)SXkY4~rUJWi+52N0wXtZX6=0NUYALkl4vy0AE6X6pe
z2A~)`>O78seG=m_d0CR?z51Oyf<|tqy-1MRm>jH9HrOHY8bKkdk|G)3IPuP@@Y#nJ
z0kocMDj4&HVW)hhkG&m%K=pBo=*!f^;o;Ot_yNZVqS8~<p^2h>@JFi<wkfh4^_kH{
zrFwPUm{5liJ5JI6P`lNMgOeNK-z0CDCL9t?>WfK@S1QDwO)$S?*<>ElGIOWgS)lPS
z`8|Ta*2`k?Fbadim!?vdbg>fv?UZC<@@&_69*k%6GsHqspP%xr983v3e@x4y{;k)b
z$6-<Yv63H+=oQ)WMr%fB_gA3Z+l8oYES4!3^_*2C=z-Cnd|EB`mT+;5Xgo*;!Bq)_
z-|cE)04XH&L(bybhu`o%7;oC%_8!ghkpnrypC=+BcRlAP<DDbjPfBdh;V(b8ngyfm
zN2cxU4iUl3Vk(ksQjSS7Xk;eyNMc5@J#0*@WV&;^CK*(OS8YiJe{N|=fL-1)US274
zPQ005`76rCb{*fn-NbZ=yzy25Zno70=DiN>lAr?Zl1t(nlQ`@i*#;rPs)T4ML>Dg1
z`XgjVPKLlHXZwSbh9<!OsVG6#kiyRG&{5JwAXI{^vIsy}gm~rfC1;Q}K7SKlP)A(Q
zhrfKIo&PX0YXtb-2FLsYu2FpW<!=s4hnJyK3<p7ovO?5M`FAd}Z?hKOy@+gzN}?-?
zm7I7OLrofDGXkX>jUj6cm@s&6LLLoOt%Mt&euwk99@G*qH+X-Jb#%lW#2D6npG>;5
z%8PoUkcS8^Zh>Z-{=rZ4^!^Nl28+=c5)6UaD??bhPM<DpkE_7CKCB{gkaT}*al~&d
zW*~=@HV00#bxUSTm{(m;V6R?l7R^?B*J~Mw<BZ6h%$J6bG*2g3)m^{7D1`5}3-DM{
zdZ~(?0P*qX>k97`d|gk-F8d7V?21xtSCp4B%sTXROCH6>&)p)KjjqM#7NxI+SWQl6
zB>+F4=h0D8RB`v~V>xrx#R9irr}yFYW4AU1{uUnk5G>zOz(2!413?u(nA^zVKL@WC
z&cRqP;yb3qw(@{V-g*As1njz%&~F7(f0tU}Xv7>`a=|`9;8r{e7CH2RU(qMlDFX3%
zV`^vuN_*QnJc#$LJ0z6o79t?%4(bqQHZ8SlkG8vw1`(g%0cj&%<hQ>xpIpZ5w036!
za^v>og{X9?v}%9*?GF(GydVX*{M|uJ%%yH;=HgXIERrLRL+Wf-GK$(O*!6A0ey**k
zC;bmus{&b{SFBx6UOBwt*QC&Z9iwg*{kqg5rpH@IZSgs;E*d73WSIF7*LQQPm$N}<
zF||L&$+qQ#gSTOCRx#F+0d66Du6>lnL0@6%EFQ#`0Z^!?$inyffOr_!XK8E6xbPIq
zo9%XSe3l0Do`$Tz<N}Je)Q5W5@(%#B)V10(OLY%bt+;T3Kt&_o<<QqCG5$(3??4IK
z2@zV*e(Hnb-4_m#Rn~Vs3M1QjW>adht8KS&QFD~OLFg?f+wQ-%lfI9#gLRz2?`zfE
zJz=>stfq#<xGoTN%2M-kD9sM0yeXux?)Kni&sA=17e$h{o)s}npJG}_X~<<P^fI;D
z)XL^9w2}*+<QtQdDR2dA><LA|7y8%;TAwKujd_r=@2-LUgobL1M9ZM~`+>8p;!8P8
zOfebGrj(nx6#LG5&&9diS|sLkr9R5SEza7`XC_2<Zyepel5YyzO!+$vgf0&J4;0EO
z-gFo68y*74lQz1CJP}hhdUJh+DA%u7D;=ImSlIVps6@7Mm683B@`1mY#6?+5=eG^7
zskUSl@NQE!5)}ijGF)*TYJu`~ah;jU-97{2c+xz550rXMn&4l4l*$h<m$n>~e82F^
zF@WEL#gtdGd%;ML)1%9QB8fA-E(*3U+Nwjm(x}ss#|=iy#p#A%`eCCV0p#pSQ85FU
zBNU4HUA{c(H+jm=u%$-buaC7+2ZTK&pYZtZ;}#Zhi*5r}at=4GPP&|bp{cS9TFJf~
z_^v^J1xhyG@(jfOW>vr~-x~>7EC3*s>NDi9$`sAvBZi;Tdc(U3bWfHH64+;zd0s=Y
zI?7kX06+1NH)%6zjQrrlOw(<x#uzwtJNKpGS$sN@94Hm(lOx&dI8cSPHUZ~-c-MtC
zYVwwem3`SPM@Jb++?f|SvssBg8*hBd^7!|*$OV=dqFFZ}TdZhAxdPX22SQa=@45po
zsU#z-RVlT#Wf@zBeES?y>V==9Fb7A49!eQM7vw)WPV6cw{Z$V;L-8tu^I>PA&}sKS
z{0yH*R{A>xC27hB_m<3_1fHN44nX_*IMcgbs26*5{Z2;-KK6j16|YgF*!^6x7qub>
zoTswG&d$I(ak|&#&bX%1&Q|tqT#+!a9kx9TJi}e~No?Rv9HRg&F|}?X2+-R><n<{A
z_(FaG?gBRecouNRqQ(aY9kP?}U3XC5X}OR`+7A|vj>5SE3NsS@%;%sqlu_pIrt^t5
z(T6~6>ih{5<K;(T-sgFy1y;(*9*Z}{xQgWRhs0t*Hs=^Bf*G!PcLYSxvY@Nra7aU+
zu<+eMb}@R}YrO};z<_+vtb>DpKa6WITwX=q#02C`SDH!p%V8%jVACRf4wf;2=7R9M
zZ|sahMC?Q9B8$A~9toaww(j~r@`H51;sv|5_f;WgAVyb+dC?Cw5u~C<zxXF>2BfF9
zR{G~aEd}_w^cbIT2K%E8T<4->52pRCTgk>;G&KI}olL%ydwEgeJCO6wb}F*AM|}p4
zL`O;a(B^Q!NbMSWtRFNV^c}aE>2Rvpmx}B=tIOnwsXx<sfS%*{LLUeqwG){O0bAn=
z-bOw6j6Lq5@KSeY-?JO36P30LG8qdp17!fQQvz<A+DXkH##lN9jH{1p>gQ4zVz2V;
z*S()(79ssNSM_?ELiP{+Tv4s#3bAOceDU-Oc=tucZ7i-ks$$KK08pmW;NO{veX4fY
z^3FSnqQBvXibQ_1pCaN!;UX30wc!TIKPw&VPSJ?H+ChOQ@#(n_7HConq0pu~z^0wK
zM!n9#HSK6J!&_e^7Z|1*tw-9zu^t3oJ&OFhXB$L<5<7~nf2C&s!o4giJY5+&ro+LD
z<CweoQk7Tw;?%7zzv`g!C4Hj_t8Kye(&}@p-)T07oA?TP3wDkL{F*U3YYsM@%_Tx{
zLAMCx^XlPLvU7sy`f572snrB|0&gT+^tZn3QKmR}&RUDZVnd2la5C>QM&(oJW$~C+
z%mT5cGh24vmvhPg5ZPX&-0JRYXbms!FWwCqZK#+ft%z=z95fvGReU6GRgXDB7X5}*
ze_XugRCL}_AiXX92s@awWhT1>KjNZ@==2<NZ*6I)jzbmroJ{kcWM}Tuo7-Y~;1E}U
zu^b>D{NHCl2wt?Ml#7pJgAqfFf5T2pwK*CcA{z!ME7%n55KAdOKStNP7YPH$9a<Yx
z4;cZmwZxm1IWu>eN?^3RX0_X2KW8m{<jVBdyO>lX-ZjE|S@Z6JNm<y)P{10KV_u2Q
ziM$ejLXAX7jV_7t+%xD{u2--zKjIqiLT+)D6|lrKq(>zka6?^q$x><ib&y^Ex^{Md
zFiUB*&M9Z(&g0H7KM>Tn2ybA^s?%H=4Y)QvgU0)vZtV1Dd%voSK~ko(R^pU^ncvRb
zWVv<J<%r6wSI+!FPj&Xa54#F8#C`JS{UYu)!%+Q((GD)=TXfO~|0<m5`oXl>(&n;l
zi-pGE`cToweVjHSR*TBn&GJS5UV()oOVt4u2`Uyj|0wN!vj}sZ4XZhV;4pDu@os(T
z3Ij)St^`XPLajV1Hhfg1J2I-<>3V@Y?UZf1ne4AcuD;;${;?|~Go37fg_iwib-Dl5
z<-xwMTl*^A-C{E7J0r9f<v&uVHWOC@m*W^`MLyKnzyD{fOQ-^=0?7hVlEImR;q`dd
z^_nnm26zYGzY5~KINm;EmT+WIEtbWb%h4r~|HPC;lY?}jl4hQB;(6Ze=CXT}1DA|?
zU8u|h8iHd+%l<9l{`--4t9p<jH$Jj52w&<d2=)G1EIDWTw3YF?0#DJhjL%qB>v@*8
z5w&s-P(d^;ei*=-YrEe6HrPzZ6h-`9vO+zxyF-xHH2z28uKY_{mCS|I7bxuj4FQ%~
z4J)3wl5um{UmQ{Cvt>=s8w%q)f6U8d#EYB7fm8H^Wc0s}5aeDVIl~pEK4ymzI}&I+
ziFm8oDX08vw(v;CgpaKt0xekXJQ&o`7VP^KCO4f@>6{4sbGR0cJMqlW$w#M7kH~M2
zzu)X^FP2a-tS-U<ah6zv@+FEtAS%266f=fvGyhB0`;u7#lQMG`7twn$TmQd|GA3^<
zTqeRCb1#*_nA0M<`_0}k?>EGdrly8Tb^R}gTn4)c<62C|eeS*qwB$o$FWtU`C;9v4
zySmWju#_A9G2fuYang)vPW+uCWu-GUt=<l+OE;ByFC%1sd3`?mQVsLCls2WBDsN0z
zF(1g9S)tt=e|vI!ADeh+yhh7?EBiJjbOFY-<^H-bk>k5Be~DIRRt@EX>0b=+q3bx#
zy)8TQ)7?FI&@53J+W-ImemHk`q9JVJtzF9e@V<@%PsDHW0_*;(g6$ao@6$0)K*>>*
zN@*5Fe`Dxy{?8w6?oU{HH=oabV&+<XR+$QFKF7!*;iv!gzdWa53+U>ZcSU1JQW#)-
zFOCX<c)2gD=zB%Z+iQad9FD1!{<Yn7o$pq1EBDB9Q6t*~s1!cm0>O8w%vt=HXS)(a
zCMO$i4gjv?NJ`*xbz7qSyF2Bo*vm?!rP2vdifgM7j1w?Y+L?--$V{+|?54|MVxT4N
z=%R@8*%2#%atGyE%Q;1cwA^<`w-c<8JNRchLHA69$&;{{a^Q|5SR|^_q|Uv2&=W#?
z@7Z=Jzo5c&`)^I<lIwgW#6=8saeKa;?J!*rQwZu8`E5}0?IfkZ&9YK5-MifLAyzHb
zfN#;s2WhKVACodZ1gaf-|G3)yLLLwE9ey~`ol5npU^Q_Vcz?ySA$V(qJDXmYeLwYL
zX}ud?hlS7J*`+V*ZL508bno(SgkW+;*HgBh0Dh_*Ea=+GgfPG7uZz}UfRD=<$;e#J
z11Z1dL9$K!0jd+aRu*|Rppo_RHxk{u8>-ody4+`=0t3u$o9`C%X5dos1faO4PT|gh
zjB?S;Yo6XQydG{HqqJ!}021&9b|gNu?KJ8E)BYTx;FVKGH4CptWC>bMqJ>rWh8-4d
z{0*|b`}Qj*X7)a*M=e*+{;*AkY5^dYP>4}0v%Hq;INo0|xrLzx+lkP5tBk`=vCLBs
zBRMtsy$y^3-#y9w@mto4w=xebxOlaJMe;D9RXa}+;Iwu4#6murZEFmLvRnH37LKC!
zXd9!9tS1*)-rtsS8~ws_kRT5V$|Y^->6x>R41p+5nP>HwW|91Fvwbd{pk|nCI{L7-
zvN0|6Sh<!$nqh9#LZEqL0ROvLVS-0EtwWt1cZ)HN?I2k_!`(GuWa&G<@9<xJ=ANxZ
zt3UW6J}=kguiLLMvGJt()-GCHJW)%dtXO5ln>=`npO!>jdG9~-Hn~_xQ?<dT0imuD
z+m07Cm{<93=_!d0)D0*+hhIEbaAQ9*5R)fxS9bL<?U2y>3*L052E2B<@7jeCtmGQg
zc93)Qzhze8Ke8S1iy=T#l>&-O(G~%Hpc?94PtSDAn^6$oZ{#X8yJ!O=8IvDlGYH4F
zlbTa|w=~96B`n}|?~ZII{XLY~HiH#%i2-D4nU1JIs6f5FK)Qhp9yKGCdWcED@9w{X
zpSwKbr9Ul2;hF<gzHddKi{Z&_*}*f4+vB0D7a35&(>@+`Uh(AZ$(WTFN>`-L;upIQ
z1|~B{$Ifkj01`HntUiU039n!x-an<uL|;Meb!GbC|3a^~9`p48A7tH`Qk;+EPvT=9
zwT%x4t+yrz(uBB2aThHboPi3y_<-Vp;qH}rC|t4HPhHhBHg~O_&a#|5V_BJ!f5^FZ
zDzYAqtSexvVUv!zvoNjonV5W>j-U!OM&UG?bwrhkpQC*ZYh$P(y&QugPx?J;18GCe
zy6}QXW$b$IRK3;a*_(V>%Tgah^%vLCA_}qd4)RF}aShrSrq2#W2N_To!3G@F#wAIo
zYKAfeDl780Mk}TGISj8nFWk)N$Xlh9$@9UNR#>+yhT`kdk~k0kT|-jOP;s;HXS*`>
z(!Eut{~Jbpj++q~Xl~?t2aaSwQ6QncJF(rukjSdfQ-bkCaOCy4RXfYasr9B2{YOH(
zJY_-(63k@;@d>q+eQ{drYa?H}L;@xFz2bwv8aC!xuI4e*0*1OW0lm`09)fxrG)`-E
zLmtCC&2J)4o`>JE8bwqJ{F!J-V7tI;v!~6^kxF$&E#Jx^81)V6F^cljJ^fK$l*KE(
zy|G@y1epW0SFQ0%y}+`EJwM2U)Zy$K%xl$>1Ch$rC7T5gKjGM#VTlGiUY}mfT)YVo
z1JsF5uYsaygwocqlsTZ*R_hpPmulo=3Z30S8N^56ZCiWe7;%S)xhP+q{2u-G$hN25
zZ_2t!CCtX+B1e2=jfU~vZn@C8S#S26?Y;XeEIoR&>d~~X(r*^)7y}kJCkN?6)cm4T
zm0iXt`1Nj8e42FX9l*@AYK+4KE&T0flYQA(xWt}rd0Hz5-|BbmZ@b2H_zD+3c&CZr
z5TX>``4cx4RIY?yBybkf#inFdO1-U4vn8*wb-8>@?N<_3GCDR8@Dq^)Dg(#;SP1j$
zAOo)Wgd1FoFQ3dKuWgbjoI2~r&w_Bxd_VQC0I;WqYkV#og}~7L;N0JW?_!{ux$0x2
z_tsRbPBi9AeyWgFmAkc~y)g3{*V-Y{WGwdQj$`0>JFAqec0hcO0NONkKLKgEm9}Ai
z`|jddBNt|}hV%Q4wBGj0n+S{1Q`DUU$qAibu05+~2T3pL2}ch^*YA>;iog*iYjS65
z`7MJAY4P)K!14tesen8`VF-9lKMSfGay`lVGp>RKzVlP30nms7NJ75VcPEMyHSxu;
zb{iP5^WoB}NT$xA$Qp`4XZxZ+4^;EQ$@CNh5+i$Q(opI=K6`R|$1wM!D;$yILMl|$
zNWHd^C*ksPPt(3}bo2wbTLO&jNtJ{Zs9xM?>Qun#D+kdchsg`=sL80y;Uhj+akg#~
zl*QLsKvD#J@tOtRm#>b3bK7;C#mJo&s~<5RXj>@SWku@tWU^g-&J-|;gXHfasF!I5
z2sk=0tBCMgl!2BCgBjAbcYY}k?pPPzVDIvF8J26@70$OFZq#$}ii1h6H4arYDoG(G
z9G?xRoZnpO|5^`l^0K%0fv_MLNAGP-L8>O4cxM0<*(T}hp^hRgBl;4k9&GcnEBg?!
z1P}VE_?r4deyFo6^}OW^KMUYNV_cGrY}?JI_3v6x$hyX^sUpcgs(+ud5TmeM<@bOJ
znBu1~`J&N2^72k-tG-@@zjVr6p7K!6ruuqsrc(I_3e=`h_+_oRY7z+Iv%Jk+<2^{z
zXe!q75Se>1B-dc3KN#?{_ZAy@;ekyI`y{jCdDteRar^ah)nD5opkE>;<Z)dC1uBKE
zh(68di)o;ufczq$mJ4Nx1keL^Dq(uV>f5wdHQ^=Dx>I;b`Do?uCAP;k($)Z|NQD7J
zX`QN<<oY{e5N$bhWHW3}$H0;H!mRk}A!1q)3VldpG~`kY26VFM1i$TnOa@U74=S_2
z)Ox<)=SXD<iKlQ^moPyAagP&d_X??oN4n%fE!T>GN$CT7>b=EIMvf9;+NgtCjlJF_
z{mXJL2T6dfe)qf{evhl7d{TFem>w-JVi~L%j8FCZ{Y>tc!pfKR-@Sg{B7AzZfb+_a
z4#9jPAKxmWGd%k~E-E%<UMz1oCLnHAu+U7VU~Hpky6;)<nSCge9oM_*hshsRTpL^R
z77!VERCd!SVmDg#$garPUom=F&4A$U*@$n&ut$sE0~!>E9)7$0LaOblRh0}<ASVWF
z#>j#JgCSd)?7Xyw-+H6aVk!G;K+g8VNZS;4J-@6?+68W>a){i!J1GOEEpayfdjdN-
zfVQ-{QEQ-}-J~-pXNz&pr-7s{%BDTjKd;r%KR-NpwZ$wG5Te&Y>f!etI{FjrB>@xY
zJ|4mp4x?>1Swp2jg~Jw6v!HeyZIy$#3`?&^>oDlLL0DO{A9*s)32hh)QhmUsK>#Fp
zvym!^Tz!sDUI`Mi)U-+%g^F2|><@4tytO9u8pR|`^;{s#_nYur#9CE?61Ptk=o6gG
zRJfx9hLi-Pn@GQ9!9g<A-U*$>Rb_mBdtCz=!ZbMV=-p7(-w8e%{WF-kkIx&R`nv2`
z(0UQn=JwElXqH(BC#s4kufk{Fdp5L#dOCi<Cg%J6MgRXgHUpeA*#6sAj&p`xVX0_9
zhz9jrz@JRd50MsH&=N|3(6*pFUOpY8cr;>O4D+}SspR|hF=Mx~YW>t666EkvWFxc{
zr-t%5LtwTC$xt0O;+E0qATaI5b4kueecuy57E09mQwaS{qU+a(2u%u51=9!X!cC&i
z&YcYDfd8|$@cJ0lJu6~=>l05@Er6OZeZKG3%_nQzLdMhrC8|X6ZlKBW`Nmk~>LJn}
z8%lRQNS;f;`Tm*yx>Xk<HJDPe{9Lk*XBk7Iw!J)32<kqd8IpXGYQZ(y4hpMKEZ8KW
zbZ>I%$M=8t!T=}Ql2UD=uY?1u;|Rg%a1FSqZ-X`i!2{mOKigOB06Bt31IID}0DVdE
zv3xUdU4L!gxs2F{k>H@4dJoKGFXqQ>0}n`0ABR0+L=v!LAYO0aG%~-4dx%b`N9U8a
zb(nPFkOt`lFX%s<J=pUdatlQh;d~FU`MpQA#Qrigb;vBD5*M3V4(Pv$tI0zyd#18L
zjeI38O@M*PfickN#Cl-C8=b2|I}bYQ(NUrym{AC?<Qvh{LSkKbVPS@ZC5KGzQE!=^
zm5`|+Q4N9~8u*&h7&9Pcq9C6Adhoi=mFYJcg(>MIOSIiK4D<=Loz6)I#?}v@xN|$y
z5ku@nH-N;77jg!AJ>PDoDhv2ey_#sTqIy)+>h)Xz4M<RML$_du>bau{P{{AY`t@QY
z15H~2_huwfkAQ|cgo2iAU@TVT-`r5?nlbX7J*68?(O(LrK>bWu%kz=E+Vh%HA@Zm2
z>?-}u{V_I(ug9rAt9;vMM>m)8KTK8<HwU2Ve-S9He?xW?OQvpL9=UqwSUzYjkRfkm
z!&$8;3r5-H`jNaEty5Cu0E!BK=cR}<xQ5tM_)4vsP%SMxeM0_E6$=0`;;SWJt2Z4p
z>jvHY8u@+;89=f>@0b^F?(uU#mdnh!*crPD8jThr;s=EY{gp3MNv~Za>&GPK*(Agz
zTAtN(3f&$l>`K?&f;v;7ZeMGtCom_sYvu2tX;G`(9&u2<?O_-9Ja~{Sr2{5^ULR9a
z<f0$9h{u3rgB?UfmM37>lLHoV%_A@@waBz+`?oe~<8BFFObOtkop+-o?w)Q<&jn1G
znKl}hbJ}z!@yIsz?8w(ljd9J9Bik2+8Bo4^du)30XJG(1_6+EA_af86)A%dr_EUXs
z__{kl`nex@7;GTf60WOR_UmFU?>3O*63_MuWE|kY|4ges-joKee$5m1d0%Zgaf|`=
zIYXhU>>xRZj{$x_V|>LG6qndy6xXiL3h2T>7CT?A#r{X>_9J^(Ltvq&@V=E$>kE`*
zd9vrQgH+r-9AmNTM*n?6b6nE`r2QHF_3a5hi{iUp{NG)HfjC1@x8xefB>q6wpIx2C
zrx$|c_xC;wr6-szcsoh(zrL7v0NIAWtPb20By>T%4Y;zk47LD*ltf|t3TV&Dw6;1(
zZr$^~5OVMULstOO16j&FE<Dtb4YB@s&{Xk*q|qifd6)sU?;o(+lS4EZ8MykN2qO#5
z3czjPK_{5Ez)$__<{<AqM3za|_3W4XLD=>tSjSEcoB$~3@`Hl~A#tphWUaJw^>qAm
zGK1C7_t}@~Y=;&<m9m9kX7-|unJ!ZQ0|zq#IO^lznD+$oBq$Lx{+Oi!v<5YTz9Fpt
z{W~n&5{IE?CSy?a{?DWdqZnB&Y##6<%}2IB{}Gx;+m+cTpe!#r45F#kIyg@e#>#IZ
ztO!Io+KXMWNQOrNfA1n$lGch<Ql?8UKo|h#BVgf1Af{fkrho*39qU|r8_NGDgITo}
zL6MFH@G*7+0LdFcK*rIet^{RFK(v`#=+>@^1t9-R|H}U0&IQ^psrDxkFr=JiCa4YV
zOebXjhy@YRdBx2En}=VZHrXa;o`t=j&P}2g<q}_k02SLef_UFOjnVwOp@tntR1gmw
zW}`d99~}TC4xkkK-`31~YO3FAR|&?<1875HF8N?(z@@J2o!!3R|NH$S)7W=459K*e
zaVjk<AfCdfAO1&0*f;jn?!PWDM0rn7{-*>W&!DQLufAdje69P>vk6|y^dR?TBJL^R
z<^OSBps1O;^UJOQqzt2P9UttFXW=^In&JSg|LeFkz_-DCK<jCnffT6E0o~dgl=w|Z
z`7hsSdx8Im5Ewj-U`3M&npGH^za>+81Ugr+Pt=w#id&g7uVbfWjRO6>Y*0KPZ<Bt5
z!+1!H)vj(7w9ZE8YhTHfI)}I0#9nAR;d#p#5K_Pg^|L7~9buE7OOtE81!_*fpy})H
z+;xzutuq2Xqr3U>yEc3(0J!9fA@I_Av=n)vFwsffkhcP~K!-ynpe@zCw+uRWFxwKo
zk)0i)jw-a3cc$F8MtpsL?5F->pe6f`9tYa{-W{|e+{eZ&0CtD<^02$QAK>>UrKJGq
zKpUsa#)Z&8o4MGh?34fzG2<dY>0L-3`XSY>80(^72=zYuQFmq)^r5a~05CAR_Z3V?
zx^!se#diDXpbiM2YdxHSvUKhs8YBYa```aimQpBo5nL$dnex~r4*l0>ynncIiC5NE
zfIj?B>%s><c$)};NPoF(b3to`B9i~<Xp;l8_#&{`Y>Li;UW{W4;BxHk#15NbD?$U5
zJZFBrt+|!cOog4R7C*BKZ}v(qWJ+{FPT8dW&GC*lZ3TM_c|Uccqvlg=wdbsMVCwUm
zf;DX$n7E3;vUeNs0cn>ZwiIxuIg?3JBueVg-rfpkeZ3rnQ$W>G@$~)1MS*wUc8Mhz
zTS1HbY6v{Dbeej2HbYID)3MU@2dI%tNMwiIEH|U^zQe-@u7VMVHg#HZZMn;ZoNNr`
zFJ5x}GptcXx)hgO+38%V;3qC<4Lhx19<czIxu`zd<Ts4SI~~On%r4%UFVWL2$x4P7
zO8a#;<S!=5x-fZn7;o^BX)R}Jgdsq2Ht7sN_;`(|AuNQaNXZs){o$iM1*`@RNqvZG
zmA9uqlRDxf(9IPe&ttz{qIxe7UVu6<IAqO5#9-%+)a?1-`_7G#%BT7UwOA$PrLtyw
z2XB*vO)G1x$`m!aX3=Xy9eiM?xknq8emP6vNrloHJu+&O9KmX7o49?|Gno$IgSL<B
z;iiZR&h$|>NBMHTa#&KrsPFB|VpOf=#`kWBKO63M(ae@OP_eI>F-1h7xkUGY!q<<y
zSL1Be-3|ongcq2NZ>fs)hka<K4-LZ&(}(gOo)Wo?!~~5u;k~hjF9H;Dd>n-oJ!C%N
z63S0285NQSa{k%Bt%3;d37Ji8OFwjJuB@>*`91t}UX%ApU9WSW@<dv~M>AF?GUCF7
zlUvJPW|fkwA4y#w-@<7z9g{_s+q7Qh$5_2XE5%qcg`=I#E-v8G2FL^|aR<xY{(QzH
zZpNnl!@F;T!8=_EbsTMQ8o5m=O?jc*Zr`=~m7K_%A9C#@r2dX>(@A2n3C$~Up1VyH
z(0=D;rjePZ-_v^UWxZQ>wEPp2?Rx`VCk0G_aNZpkrtbnq7hvHVk&1?njn(_&ADY-D
zdpG>0d7`tL2{7ZZ0wY^ZyST-p7ypjdRsfAPmseB~CbApng4u;yD+@}x-VU1c3A0#d
z3FAq+GI>Wb_gRxTZiO66pQz*<aU8&^7H?I}1)i!EVJ7F5>`S<_=e6u>OZej&!rQ>n
z9z+C$KkVR7Zi^ne5PZIH?v^t=PmC`hHjrH?_mhiBKQD`Z^)gTBeq`%dyP2TVK&ru@
zXTE)7x_(6gPm@mvO0X)Gh0815XNAUX5W8LW6!`DF{uEZ85D|sElF}EU%X1v>g653Q
zPaU?gGq4@}qdx<|3LU;0UsM?|_M^&Ib2m;j+Pf+5kN$n^0y$L#8+pf-NoHk(Wjr9-
z8g^A=8|&18yD)3P2y_*JR^q$BfbcQ+-53GalAnui6CNj(zRi<<Psm$vQuj7Z@Yc(8
zdB$%bGUXR~0C~q=@4~v>BSh$lGDp*-_B0yPT;W2+y&kOU?W6;!@)7F;o^V=8zdSQg
zOGNFabNxYs!8^~!Om<fDZdu8^dDgbQc-n;z^-?McVa1o1za|i^-&--wey-MfKT-$x
z3w1BfM=<Yk5gduA<eVI06B#G1M)X!%MOrq>K?qD3Dy2WIYpZz{DRePX%IrMLZG@WP
zt2{JoF98gt-dkb&A(+E%Tc6oUR;Lxit-Qoe`Xv#G7%_Yis4xe=!@~t{f9S@2h@6uj
zfE#@DS58T~-dxc#v%?r7`Bmk7`RN-D+4jNL5C{TqiAUZS*^R7(E%y_7PlU@=`Gk~g
zmiPHAzkZ7JlXU6Knm>y_wkecdi2aIWJK8)aGvJZIgqc(UP~7X(e&B^lC4Vvn|2!&|
zMT4si(Z{H%z=bI-S(7&J7MU}y>MWSCp6t%dw(Qz+?`mdvHS~HYo?VXlfhF&wPste;
zEd~zb{lWYy+q6TsXGzydrS-Pb>`CE3DzU?C3p-wP9~^>6J)XwcPeYlS>m42ovcdfC
zQu&K=4h&XoLZKvx0tn>jLNkY^xY(51walh+n_`lstLB`RLZPKsQYt!6^>DYtNbil_
zEpyN`)AcRYL`Tb|qef+l0VNOaBEKCyH8L};g=TD&`H8z*Ionq@AL95I?SyTw7WTxs
z-S{)6uRA)!UMN#<yz_XAtHKG~68ZR@<QrYlckHU9Z^dM)YBZLG-fWCs7JWhh(Q1IH
zulL6zsdbfgJ@KFT<3Z2PadV1`ONW|4Ks<tWj9D8gU;Z#EA;kiBjw1cxQy1;@Vb|AK
z)jMY0UdZDc?-4_;H(hjamqrSQoD!!hVs`hq?@{_V9UTSqb`%bj(m#UQk|9bWQi+dF
z4m2fEfE4~u{b2~ILV@H73t`AVez&|^Tj4vyLiqm!^Zg%O`hOitN)Qhx^yp6jN_Tm9
z-vzNfCqnmLmWgU)>3?=K1>8cJ2w@U<io)9brXA?m%Q@KT+HM^rgLubE?)THvzk}o2
zHjizFeBF(_k#z14)W%VuPr4Cm|8)Jb`onMRG5iCJ=kPyaPix5778XGKdLy{b$;BmM
z&a-lCJLD_B$KQh1p{%lV!}}IO!S}FeW83j&IjG;ygc`ubRNg|6<oKo5vme&BnC=5u
z4mk)A^^BKY!9Qo;gzoJvA?rB<A5B1d0R7?D6DcytjWASV5x7DW4)O+xQKkz#yIKqF
z0B{c2739`uYmnMTZ_ytHIU6vTeUG|(>fHqEjG!wu#=(SM?kg=T?1$+L{s|ZzJjS;W
zy`V60Oco>;yJ^>WDEp75gG3e7Zq5d$le^7ErZSZU9Kr$CQ^y$8FB-vRm8{#Ec&32=
z3<E;~4iRAi`m`=)IG1%^o0v&*i;f}4v(tf70Lcv9`>d@Ah%DVrRcOmUvi!m2-A0h>
z|J4VWad+@TQxl+%R{ot;xlt0j2`C3qo@2^o++n_fWWeA(SGC<-8e9cmdh?%0SGrC^
zfiZM-V{#0!j|Q5V1!RZa8;v)oE$C_m|A<if3LeYuI0EPg>#^XiD3oXiNn{PAMBkj&
zAi08YT<0H+Ikz=YsfQl}M0g|kCYs3$C_>5`7{vXDK`q8>zI1fA_A(;+dWQfQavhK{
zmwqKHRC=PTv|<K|08bKQ4X%o^vn%3NzNBiWlMK&MuwzUQc>kQ2wiyGBvNZ;i9lRf#
zTT0eY9fngug4dK!Fn9mTn5>yqs*)NHdoU9D=q=<bcMmeYUXtmAyUIY|O0AZ972~E*
zYS5)&zsV?TaC7PqbcJn~ZG;k>Z*>%qn!bp$QH6`&JqMY}u4<jfTY0w16zkO0Y;pK6
z<YqU3p#E6hYLweIt6?QjU1)?DY)B1?hzt2m`yTDqNh0`t4uVj_ko>9;B6uc1a6SKv
z{?R`&Ejy22Tq0LuC^()`7^rx+rPF3XRut1{1ZpisAXSYJO7S~&!n^0r1OS@nc4Qrp
zToi$jH(aLA__ToB9Njon`4DT>Pu&l`4@hVU+=YW~Zbzmbne#}KtlPdfd;HPcuzzHb
zRWHemW8Q!+Y|Xxc>n9|G6S{oB-y(Ybm8QV6nD(U0lEXkgt8Sf>#Qk>@EPM`uI*-{G
zVhoiEF2fwD1KdxO8NZml0Fgt5=}66)oSLH!WI$5NAurP<9@^mJkykkYNF2yq8E7jg
z8mXxB8sB%(P2(|QKDtgmCCuEdmhk&H8PqdQD^tPMgzu#UWxP?QKL+2xva0_Wur9lo
z16Pkd5k-Hwv|F<&e&o}=WvNWtY4_brJn6DsLQsIjwa!lViNvJkx47ngj%xdRaf$ka
z&+UXOxT?{XD2Tl13$=Al9;clh{wO)grNs72`XB9-Wkj<4hVDoiyOj4ffc|rSs8~M`
zWe4!)DJoE2PS=+|SmY$vW^h`27-<;q+D+$j^-qTApgyjnzLiSd$o9)jhHiQPN>Nhq
zncn#>iTo?VU!FvRSo2{(DXc21^3VO7+|7qr30m}z812*}q3>^!9L|Gdh<&qjE#=_8
zflJlSOwwd9)UNf9e#h>^Fx0TmGgC06k(Gw?+3%0~0OkV^x-q0b0b5($WJhhQFa1up
znksi#O7Ow;`_pgy56D7yEVnN#ni8QJ`{C`HQ|;gm6F-uK9Sa9UhqQ8*(?Y%^4Z>06
zT5Rf*P&Rqt6G8Z|6A(ODO$9^W3V`sXGx*m;iN}$52F3tBH-IZG$9L)apXZXk<*&^-
zV}!R;k;Hjxz~jVVit>>ZW@=3@yCdM*CHiysg#!(id>_0yI5Fgej$-1Vi~Pqeyc><6
zK^JnYI(lPh<n1On8_<5us*NtK$%Kal(W+6<I>Vg;-ZUnfSuG<5`q^FPBbA8d78Et~
z0^k=?`YaCGGgEXb%OB<0i_?uu2MYxN34I4;G0B;{B}YZ*c2`zmoxWGrC=tK-0vXK!
z+|KRT@F`hNUCshVy7%W+t`b<P7gEtkA~Dk}d5}A7LmEH(_%kClR$D%I{qIzh4xy#=
zXI(Fcvl#L|meT=OkUR(RM4#7k#wdUQ;6T^g*sY1<^H=Id+QX25CZC7zdO-O5qL{3k
z>K;kVIgxz~blUnpADPFw8aPLN0d5w75@J(P>SQ~fySN@5dO&b0ofVoxjWmDRV8gN<
z^xUprN9}IOLC7(L48AQPl0qP8nEhlV<b_xn5w46s+9w}F*E+k;1M7IA4fviV&wLIw
z;Gw6hC{U-Yg8tpanYLtw@;W9!)6m&<L@iVq=U@P6H2kkxt~9EtBnv-`5LqNWtwJM4
z+?U8=iy}l>mBv<(H6V)wMI<0ZgMt#shK<-N(k+4r3IR#jiR==#APP1JMz#PN69xiA
z1W|&)5W-Zn`<(edXMWeIQ?KgXy7$(--}fnlFe_&>@gVPH9zeeYSs2ZXYp<g+%O3Z(
zu7w+z*QK;RSMrGimaFmeZzmhyi+-^&;n^|5+W1SmSXrj><>X3QJj1Nt{I2TpMO01K
zv<YAtyN?u?x1%0r|D*;$(6=~r{XPO3G?7R26eyy)TqZh6H!?J=Dr?O?DvXu9b_nFw
zd;)cY9=$lwEBSk+o9wB1(*<Q+*xma?Xkg-@K28C*39xpLCmzdEfFKUztk7b+B6S7f
zZN^RBz|!xxKXubTw~na(K%v=aAN(~kfR49O2yL*6zU|9_<MIwJ7WYQJXkCSLRNl~c
zypi{AxuV|s#8Ui7)}dKm><wI6syEiII>h4(V{}usR=WQymgq>Gd5R?~m*r5yX0@ry
zq20@_;yv1)!xbpfl)5~v;QKy!oRDmctp!6k*>a(Db+0@6hnbcsi|+1=O#aR8cBIDu
z!=wWxB|f{BCkQ%H^dW#bF*iY2@=^QyZ-jzU(?9!U(F9o=rLp%65+6CkZrQOA)gk6z
zXjsOEFt{38Gn4E?_w9VeUK2u$@f%?p7EktPCGsTXTKOJ*{j(zLkQw;)bjs0@!3+GM
z$!cc}c`W8CZai}(B-=ajOBIM(X$9TfUgp{ae^?VcXSCHTtv|4q6aFE<cH^5_8n)Sd
zYSW_td<(|s+Fc4VYJ*EWLR5=CYUYfY?onT=7e~B7yE5PwcQO|Z4j+()LEJps{e_^8
z(dHO*svvQnc7vKC=hKv7y&e8v8J-(?vo!0=sV6@KWKU3^zS^u?+A&ZcnNc3CLkhOd
zZkUL!GXio$W{Vq!YJ*`^D<j}>zoij1Bl*($aZ7joSwb)1iS<P@Zi0xL7_kHoFCa|p
zo6()KsA(%q_1ao|FFryqQph8sP6LIgT^Qz6<f5*g^Enl&nv$3qxO=?=A2^MN!APU7
znLuW_(xk!0^WsM&h7FyMjxvJl+t$qYZFPVf+G87$djSH$<9;fLRr$rC!ll;zjXwj{
zijDq>Z}!Kc^~fIN4G)aD_HNC4nGd7Lo(pn78UckS{<rOE%$nhcyaek_G;8dCLC%N!
zAd7KdyG`a&JKv0Xe_`KfE!b0XZScxfr2%VkkKJ1j0GZGJG2EzBzve+>r;y?dM<S@t
zfcaCOa)g$Job!`Kxcl)YQ_CRKh?ezPl;*MUVnRNku!zhr4)fqV$>+QLTf1h)jf)Ck
zx(80^MF~GvAFn)U=C^MJxPdQ0#`x01YIjvhYqsIcu3XeznMv(fDa}Nj<QRfp+DRQ_
z;BC^qhLE^knjjThW9Hbw45|z=POn)B$Y2Ws1b2|U^tSrl;w&N#oQ^3(GcZm0cIoFi
zSWgXRP`20V_av%Op~Nx9`^0`wvsmI^^-GCtC?m+!<##flOBt{qN=)mlzkiI?OcMYp
zeloK+AXom0a0xFW?9gn{$bFy8GX1OwBk3K8dFa2!;gnC7_UN((M1$-=Tj>&3f>fcH
zS?)(zvN{hjV5^Lhm_3ub(rb{M+^nNl0K{cOVE1@Gu@7AySt!PQKV3R!NlA%Ogww27
zDjp0w=zCue4OGFr_JQb(Sh|<6NzBYZOFx4?_ai?S{IS1Bx;Ye}pc6GMvIITdOJdu}
z(YOfT?{3>js?1L~lx9`kOY8rN0YNb^3j_Y&7TvpV+-xwFIuOnxvTu;QFx3LO{e#hG
z&<k6H0JEc^A71G?|CclQ$B!^*hz5?(NDA63@j&hcy$;Y!mDwqW0<znLm1YA}<uG`o
zQd4^vd`ZyTs|YLecT&>Aq5vd-m=otps+era{3(d4RI;$?umpIxVIVbvH~;)%^9V`x
zYvrgiTZF4A3(xfk2nPoSP?`2r{xn{5#9IM!$J6i2Di3>!U+Z2r0Bc_{fVkS=6^ygc
zl!S0<eE+hXPJlt6NsoXwyoLR5GME97QurhdKrFBH75T4G2Bxk@BNvtF&UpVGNgyEj
zol3!Y{yyshsFe|fOo;kLsqTT)WC1|r)WJLBEo{5<`*m(DK2f@F(YU^<Wu7I4C7+Zk
zi`AD~aW&vsw|D*9dD!}t>5E(#-3esnJ`TYJYu+lAFS|BJM@9KzL|V0p2SE3$Yjd)7
zi}ev9Z<2e-hm3r3-9KrOe^z#;W$0Je&P>ZXQ#0rrJ7@C5ADEpS9as9Siu1<D*R7MW
zc{T^ZwEMA6ULrs)DXWnAz!f%5aGJs$(ItW5V1RMd?@BHDaiH6@KImkGd~gP>w`FE>
z)gO3PUCE`#UJ~}=c572c%qnV>lFhqSGt8d9=hB!U0nJa1R4KpUS5q+zE=>l0o0<Y2
z#+Ow1tI+WScT{j&C(E-d<}IRhyllyUqT@b%96RbL{uu~>J>#&cIIB$d3Nc~9!v?O0
zhRQxg7A@3{OKvcGMjF(}6ngS5U(eT1h)(DY<>Q_q55=pJazxdUtdM=TQ~Q{6!7uAd
z3>q#INs8!7#gy<EV;nVkqt7IxNdrp1ZKOx_{LJ4ILF3xx#}y%(YfawCjP2I$4{GT)
z6*UT)Qri;FVu|%T4xStAU{+ixDa;8Kel*K0#MMr%T2<m}&{g8tH}~kPI9h(r>D+*z
z=<!z>Y&C>KfrIk=OxtHUYurx}ENkTRXg2NI+N=F9eu!>g)^wECZ?(~dIu|^C@l_GQ
zaQS4vT{yY~!wO?xs(r4#cl8h{?Rb`Ni@VNz_~6{o-iPf5u_<5=6h%Pu7=wErelq>t
zdK2P#*)U=zf`x$6!K;H_Ww;x)g{bjRXJWOTee-eY#go(z0rVa6!FsS@T*!2?n!#=N
z#-cwF{B4#Cf?XNLy7P5qqAt~NedbYK1223rH*3$R_6>P+0&;~`yUEhyF?A^UFsLB;
z==>X#sTyM7x#)e;;2>$@rD{~!a1hO!hdC0P@0=1J>MmJZ^jHl1pA{}oZhH&qBJ%LI
z=J(UG9H!~{USx7DObfzoQko7*<7&U|+pIFL-BkpV?uG?pn`y23GI9GditA5vBe}x^
zAef8>HP>aJW=-4x&b8VxWv(aOA2a5f(%(cn-8~hc3Z(;$Eqb3J=j(S~$WzA0Yo`~b
z9bI2SpM-KbS+@g|ocGL}yAQPTVmwOSoL0&e`*cX@Fvv-{m(VCP0sFTdFTC@QiP*e>
zW*X`K-Zo~4udpVPSK{WRC65ZoL&dOE=)WuPS0{Tkzt_ISj$~{jyQ(s`Q!4FXUmh?A
zfI@PkMvGYQWBGWvk8t}37!Fi}sGhS}HYTm2qThbw5cbdQQqjA085Hy@)Tzp_b(5#v
zbRIa&G5C1fIwP7OfwZ_%`K*F-i8>F~Ox%!t)Yqh@Gv74RUYpl`D*WwrIUzRf6<9FK
zKc5)XlcK7~`Qby|LM{Not1m!jHI!#rAq+c(#cWymy3lXnOIST1*)NHI)3ynF!qy$s
z`+{N>Tz9&sDYw#Ez#G4Re+KU2S+H$6nX=n{0M3m9Ua(~RP<El=b(G@*VNHHr839m-
zwqAx#Go~FBF=`Fv;s_cRuCEB#WMm+72hPUKDWyA458|EyQ%&RMI=#M{_FFX8-RN3=
zIk0&E+k}n%$wYRSljJTbD%B__-<SBT?w-2&jHp1tx9#-ZE{@PQO6npn2w5WT=Yfdv
zOkQf0uS;n;ncdkz(Qw<Fv~DS#jj7?OK#_o|@_*AGuCkk7f|6&JpLi2etW;=*eziY-
JxYRBn?mubCf#(1K

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/space_shuttle.png b/previews/PR3525/assets/space_shuttle.png
new file mode 100644
index 0000000000000000000000000000000000000000..dca29fb2ee895cb81597ad303ace5027522d75f5
GIT binary patch
literal 292247
zcmeEt^;cZ~wj}NvTsja)fCP7U4-!bQ4#9%EI|O$E1b26LYuudx!D-yxUFPI_@0~Ym
z-unw?%}>w`-K)>({i$8Ms&<HyqBJHNDH;q645qA%#Ag^7L>d?vI2#l=;2AOzIu{ri
zN*GxQQ56^1!`2`#{OKE^=ci9jda$_YA3ktdaO_EnSQM%tdcymDWd58d?(I*d@Rbrr
zj1)CDx|+Xr{4mf676D&bWaY!qiR1CSJWWkX!1#&&Bjs!DhjmBi>#;Qc@e!--rg7f!
zJ1_1%iZ>!CIB@<j|Lq3{JlsnKNnh?8R9G=@97+U~|NF-(jj~lU>VJFspa1nS67{BY
zuIc#b_v(LoZyY#@roGqy!+ZXHaWW!o_R=UEJ?6*%`GetX<-7lnAIKZWgb*RvjvKv{
zgYLgw&cCiPHs$#LzKnmrRVEmRatO*$IhpxC+>5_21}+x)|6eT1|NmzH&(ivTx{Q|w
zQI;BQH~J~O9*-d>9x<^JwmL>M<7aJ0^a;!v?B+8ec(g5xpVIg^kD)hxtvBh;8>1=Q
zHP<Iz2S2%|)GE#_GX(8t3fH<zN{U++y&664Vg@N%g`G+b!9>j34#IX8#<OE-_QvB(
zBg{O`$ElE~$EJ&VORp+V_nGG>ZmT7mhLslAi$6tQdL<=88Qi&-fk%rMJLBns4;SOY
z5A{=;|AkDqWWf1-;`Wvg`ZuhYI3X~7Ys@rXn6EKzmA>DI6z*$(=?rdI36atle!866
zqA2&cdD;piVm34~3WV#VeeHfYT+z;Zv6WEpDTCMc+~IJu=w{iI86V{b%e7~JQB5k}
zVnlMie&RJXc0D$t?_#~>EXEg3c~y2!=)gh-0UCm4gT?!z<WiH+;=$q%^!w5=LTHG-
zHw~5zPDjn5LmzJOweK&|($NupvqUP@BVR{n4}Unc+mykz!a(%xUil%d<{&$5sG7lT
z-3Af#)Cx4xXg1x87p&Cra)uO>IWcTY9LNvvKR%*2dY*P9e9E}nAFgP7-Y>B>Gw(i_
zDepcqH8oZL@wm5tb<qBJGb{XjyK!H4;D?0a0el~=0xRgCy8puRRb~L8Z*<_OF~$A`
zv%eWo9tWSFo*zzT%e7wCf+*S^jv>{P=dyT0sjQawMeN4=is|7qX|;!ZQ+C=E_hxiB
zkGDCW@KqD=wW3v(Kf8dNX#ae`M>$@ola9q#_WhZiYjDiZlBx6Jy!;~t&p;c0{-v67
zaMAhjTdm32n3lVX^0s6MUM>o=1cL4O@`zwkQk*KMDyN?4XGGaG#3q^%0`m9i9JZ6*
zKc05R?qc!Yz1Fs~^e-taHG@_g(%~#M8$#&lv$H=ItEqm0=uOe|^_|4MRZzkgRpU_}
z8XZkx`LSNfz5MNb{#Caf!!Q?P!Kuu|zN&d32I=G8(p_(FKLf3iWR4RDpYe5S!UHz(
z<9TMDTzqkIIZ3|WS6oGt#bQS~9CB`_Z*Ee7cx5#?tivIBdA)bnm9r~zP3{ebvtY$P
zwK+M`5mr#Rd?GXibF=P=W;|NOG(N}ukr~FtYhy!6eaDqlLFenGJL~pVo3qKhllFFx
zt$vCM?N+y^9i1OfSF<Z#FRkEmj{j<QOL-_bTWxa83~7Y_AdNE%T>H!OLvI*S`~7|i
zdIFQD{c&nOiKKJ{c}4MidU|1(Qzco`q~a7@MVA)wATCC6-ljW<^U-{jzOk!e-P`vE
zGx<Sqoe`kdCB@~?ggD35jyDkrJlbCyZ6*?x_1Px5$t%hk=_^W6rKwu*xBZt|d~^Cg
zno)cB6H&Jepi<fBi*6&ZcA@KGY2zyTiz56YGSH5a8~$(&e7oAJm`}Du+6#|&GftMr
zd%vIcWRad3*crdhgk$%$QlKG7|0Fj3TCc~4RJb0YH=i+<#%;aaGFfdm90I3TpA)!M
zn9L*-N>`E!V{fmMO8LInN~3<O=g<UXXEfzxu~?N>2>Lm%IAWI_Xjszx=6y2dU?Vea
zaCYucX!$;6ON!#3kQv^pMEECt!@Mo?nOw`I+QAT5RPLify|2Yq;jej%#UFx<#uCA*
zcxn#aESAHz7~NI`7Px2#*gM;L+^XCT!+cSWx&y<0G~pSmJ&F<siV_<yvxC)E{8cdp
zifKXs7U=xA8f^kF%F}&qy*DhO?cw=+i(dUli`&)0xbVyQqILVr-A;z{>8Q*X>i=-0
zWTL+eCXD%~;@_jyp%v0$-Tt)AUr{|dInHfaTvpWd*d2#X#FQ@R{ydUYA-}k{&@_B8
z?(jkAYca9W-B~dv8CRK3MirB#V2`8l@~?zV6lS6OQ|c@cn@8l071xI}CCc09Ia}Y-
z_}KUyj1<S!-+nhx;C44y%+my)k7+@S<Ws5~4<eYIj!_YPEj8;RU|b1rj-l<wCvBno
z`>1#{0N~9AwuOp0oK;O$SBIi7w_a^m{%C!$diFrKJ3(Pqo$<#u+94N89Y>9(9E#5U
z$cEsvGFxv6oV&ALD`+i4e^s_fqP_h^hI{^0$1}WvDxM*tp$&M-yz7^Z|CGXTB?ww;
zpt{q`L*YciS?Q)7OEAj9xER7aQE1sOuT*0>9I@c^ry3NlJByjw!xm*0K=ckg&C&fQ
z>DY2!dCE?Hc*a3a?@t<$A$US%^pC<~EofI)9J#xeqgzEqYuYg}@#f1e%*7vvU3hz&
z?08UJYGL*sIuS$i`C#m1HJbA_uJ*U8?*L4A948qw@*{0#Wa%YdpEj(RLfJUBHj47M
zI;3s1X;V$0%dHP+Xkn`cjQ#aFPM7|8VZDl}0<Qa)ds8~;ycfU$O?|p&2;L=P(r$U|
z##XP;eqauByn!lT#{Y*DH}V6m$J6<nG#~mu{pk5S!jHSHgr|pV%Z19bor77R=3So_
zt}5E^?;jghtnNJI7S+n!9ZzuiZKLselh}<sIZWC5QN@ukfqwBtD2eLh{ywL96k|fH
zO06kL^GSL=7#Zu#*Po1a1frACX5D^8$UHP%hr1-Ce?JmOul|#apUd*dO*DSCR2Jwi
zpPW3zqGCS@JsYgwh6LGP@G83<j5bh0-IgALF3^@iPsel4{b<ot!RJqwCx~C$92cYW
zhP7X_*ev`p)l;0Se+pBgoGTv{T)Y2$P&;kj?Rz@dpW$$LO845i=X5Q6bfMTv@y=lX
zVMAzlChs8eSI&(gNMsv0WXlOn!tI$j#5}>%_1Po?B>;ZNh31-s=9@Pckx>%$Z#4T@
zie=Z~{bcoAj56T-_0!&$!CWN#7du&f4oQCLKAj&VcY(z+m>1dZwV_xuW{_X2cdaH@
z;X$)$N%ET^Gk+qjv5~WEI&(CKK-@WH;sY*kJF@VR6blWhS<#GH2e}yx$pCgZIBW14
zi@KV25y?+Ne4$jHqG9y3opBw?kHy7gbQ3B0es2ztv9=4Fv-W7HuNA?rcg1lf?Cdv9
z;UH{FUgh7L2m7gWA<ciGJRN{?j)83*FZ=7q27t5NuMNeS&sMx#FIvBid+A-VA&vM~
z5v-nw0o`;ZfCCit_X!_<7$4!6$BTtJ^J9r0_j?5?+}2Dw%@(_>Z<>OzU(ScU(9zLo
zym(A|&j~_{si+wl7~)m0dW|yxS|OLfv{<RPfj%yHb6wdWq5PhSsZ(fgkF4HezTR%S
z`0!9Xm6B1T5vRd=u^H5)BZ0#!O3G$pLx*z+%~7B0M(wV9+*Qt5CVR4)n0OV>4=M<I
zsHqlC<uSc3ohPR~%N_`$WU%+Gb#q(@!lN}buy6ps<LY}ceuU^R^T)>QfIMEF{rr$I
z-OkS<juw`tKFm>#oE2M<<_kIw{gHlc@uCZPSblvyQ*AhGsDNBHSDEh7q7P7k$BS`B
zTBDh;=70d(Aa?C2d<7yB`-s!^Nuad$7aY$w2Sxl8q^IRYKX;{8x?ap&?U4b;u9e*&
zYlAe(V7UIM&`Kn@(MY31C8@{qYlI$ev)=n~E_=A#0Qi^+<Bk03;GYmb`xd(Q;g_WK
z!yWS`wy0@)J7=DfBu2ZQw!Uxalk-!;C;3@OhNYZFH3VhO<ol88BfHJ=^NLwqDNl*d
zA}n-1e<(2n>R~UfVVPI$GvPLZi2O9i)Ynl$rF8dgaN3zuMBdOvbhX*^{->QWeh;2;
zBR*d7Ni`0TBDosEw<iI@pEJGtiaq;^efx^3e4g7WH9Jh&_s2t#!q4Z!%wCUYLGE&2
z{QfsAO#Ow00xjHR?0*`$7c3mWX>YevtSf_6z@~c<mW!X#1)8i+mg~#D-W))80Bnel
zi*GO`0Z<?`pBgWn%P^S^;9VoY<0rS9Ekx5!-+&Ue0#y|?^z6qBTZ+qrDd+8Qa0{g!
zIf9+EOBM||(1Npo-Y}EWO-w;Z9X>WKbqi)UX8cyUJMim!ybxyFceqN1f!J%aO6_B!
zZG?e*!gnfVyviRB!5NW^lHR|>KFKO4X*QOdP!=i@3!mL7jvM7#|5jKGY%vL5DHqU&
z?iCg2kZ^2jm^R%6<BEQMGn0bo=;u^sY-nh`(sr@Sni+Onl3D3;=&ld6{=r?=N^?ne
z9?|>Fn4&UYMr}_+y8Xrc5K_BZ^BpFjo848%722i5Q@>xB&^Lra9LP9U4Ohq`;^KF<
zh-Rb^l>97%vAMxrT;<FxOdz7d4YT^aXk%41$5H>*D2)Djiwy=_k?E~XTLrRh)aM;y
zo3S4IscN62lIF-5_#<^ksCOJ3#y4lb9ctgVgP|jvp78On$YUhy#m9L0(NQ=frvYwJ
zdDzvJ-5VyXS?`eKuHZ_!A=NJADekkn)RaL=vl^)k!>MOAvg>R-n(En7tj4V;3r7T}
zt1Pb8ok}TFT%1&iv)Xe=oZBXz?s_Nz4T&ij_izO0-bvfb^ZvAX$R4&=r+Qf6|5iGl
zg8*{(OaUha{fCJ+cH`Uvg!}Jkv^|<au(dUnFF#+ax8c0v8RXEMU(fZ`8P5C*qh{Op
zsr*r9p@+wTHEKlJ_wO0q&+drsZ;mu=TbgTYUxhc4ykq+CVQ^&R0~6EQJ46oScpA`J
zZyIDDuYlKv65ZaGkxx*+yPv`n7@*YNE*@X==r}$Nwp7;#jk0ZHsHAR?be5l-Olk!$
zA$_rRLCq#FKe5TZoeL!>A!OF}HJuDa15d_-KzuGXK;tqVPv3}G<)ZS<g~bhWv4&PV
zK)nTwhq;Jz#Kp|N>e>Z1{a)233^+6aPitlGMg;y$W>MF|Q%kfZq0t`>t#{Zvj}*Rt
zYp)aEVDF^J^x-A@t^Q_{rQh`Yx@1_U#hjd=XP_CqdbxJXZOvioH)4d~3aiRWr+i0E
zD*bd^6jJryFBpw(p6T_azOrP&BUFjiqA1)_x$Y)q{M<AW@I`%<_M6L5znEmQOFtxM
zk!9L#+`28bV;-)!h(@pV{CXvgiRa)hP|n}DxN76**957Wv+H*vUkusW+A1ip=CR9d
z6p8~SdYOU|zpH-1s|Qtj#1JUNPHRD~<&#hK7Lx0#i}&IY`|L&%u+(7<2x@IrxgF7^
zBEFoNXmL2LZx*qyo?IUV$gOxfU+r_UX|;F%=r-Z|M^jqb`pfKxnOV|a1AwrkfbthK
z0YWU`etif`$q$!@j!*YjzfLfKE>L4UTAenxG@k(wwYJv#;Ofbo2f2ziHV*EWV9KEq
zwK8zMh3gN|H#K|lJlY0Xv*GL)RFz;<l(OM+=}7Xc9(~YkZr)imhA+FpM|#7pxK{T|
za&|KS$w@_{d8ND&WeIv+kbQ9V`xf8Skdm;4QN$xg<rvO<A$%tpWf=ZSdLo5j2JX2}
zRGBGQSXwXj{hTfa7VKQ|U{_@pWqr2j{QjlGDWlinnzYL3>G@ht|B#s(uYp?&9jg%`
zBDB$#I|1#xAD)4NF<)6+sxEpo!+TO57r(5p-&!0K6B4lo$D=&cVlAZ*Q0Z7NxV~_-
z=9c~%dA;rLcrdBsX}Ft<O`A2q>~oN4*}vb+TIA7t9v2T%b%gKTnLHm-N5UjwzlTnq
ze_VO*ZGv$~+E`UWEQiB-Oi(B)dp(ZFVT0*lp)w3+YwCH`1y<onjMvv~;x9e<P02r8
zTuc(!Im5iZz2D)UWz>kLnyg)e(hq;E{5F#Qu@3jg>~U*}TE&_K7;XVuVxcyTw`Lo-
z|2|L1-p4=7t>v?IAWqPQT1SHZl0&-SeL}Yu=}tT>H8k&?esVR=pLx~&HS7A0wBwD@
zw{_<2G(2$23c?@<rhf~HFaMl<VT24Ug@48ZMw~=0!;iB>+9N_PEX>R=Ysibj&@)ti
z(h?qR5E{Z_Ed~JveHx$P*O#`Z1O~0Xkk85!6P{pjo*@Bwi8nz7=g)A`JC@HTedSqh
zYCk$3T`z>w_?Z>?B|{Pn6r*&Gg-@UBZ^km(VH%_p7rCPJ)C}l<S;2}MQ8aNjw0PP`
zpTxhl)P2Tzcl^bzY6qb3uPvNcfQjVIuQ>+;NpyQLQN6QT(<z{3`G4SM9vW#~)P3LS
zn0esqi}{Gh*w4Z?&mt<J>bQDDN$TfoCMVCT$VQ*v2mcP?J<<0RPN+PeI52CV*~1KT
zOjYk=jxSYOHZ&$c(3ywKJRS>;66th;c=e{X$9I@DTv+(~k4V;fQ{9TbDLYa=4wxS8
z8iyq~xa}v|>i0LLRRJhB0|}VW#A&spHRHq4jL2FYL4;lg#JW58aM5~WHsK~Dym2G{
z7{ITLVxrD*(j~p#7XFpG3c21Qu~HFMZU4+1pZhCvgA)OCA@2Ut?zE&^#E-hS9{@Kt
z_iJ))9vy@%s6^1+4CB|2AN?G^Ng7W7?yRV_nzsY^Zax8LS)tX}=xkeG4wv7nK@=A4
zn>B_EW}sJpIJ=9>QH59b=8v_SRa9qxTwQS!F?-%jO_#JkU8fj=>Ejs}Tg_(~+e$83
z|DziIMgX+$6qrCRyuZTuDL0i(ZKKO?Zy$<qkn`O+_J;87r}1N`J=V^CAZO)@x6@6n
zR-^7j+Y=bXFTj_n<0Tr)tNcMojWn_m$tQWQTOgNPyw?;rR0WcXJgjq;?a|U)PeY#I
zMWPU=R{j+zBMQ%4aZh;{0r#dBU@QviF{R-b|HF+C!=1}bTc1_VIhW$o4bJ)GDsx(1
zF&rSfN`=5!LvGQzv8}4A1+jb}O5w8ffx6;`l$*{24chbZ7uLZ{9ye*c0!>#~rdRbO
zxVo6u$)@2nI~N}8PiL4gL!~9nt2K?G73B7nFzlxOKs?hC?}J@P^(1~?tO`g{ca#N>
zq>dl9;SD{Xzbk)uOpS~6DOpgV#p<GLB-J3PqC9)4ex4!Z#-KSzPyfD_<fh*0qOYXF
zyi|<@1<IfaaZf16u7@xS3U<qFn|FQ7o_kG!16oXsw)P7@<f-Btj%da@YTLM97E@p!
zkTF)?rF>qr)wCl#E3f)yn_&z=L4`8Vr)ZI?W9u|9dG?&*+V6PmzZQ$dht~25UOCX=
zd~0U{4M*#}dYH_U+wj?~OU>SURjQ|fTvKDa0Z37;g|F#LTntZFX(`-2ri-WZw{%$q
zXl1qt`WNe(3LL}Y^9bTbjN2l5Yx&L9{GNU!5NY4ur_`)?zC2wW2*+qHuG~(^1X2B!
z3RbN|fT00uL$F8w_fdv#$Cdf)&Pejrwt0n4qmAgR-f4Pu>w!4~-ig}UI)Fd9Ik?M4
z#qFf;{b`mQo-Ub^P%baV6mR6zuTd&Y!aYRUR_^EA;lYbIg~MLf5BI5QF_B7q7d~}7
z(<FC5$mo2y)Qk3!UOpB{)dL`s11rBN$n1-?qBpn%GHBi`<7#(cX}8-tuiPVzD17En
zKI1p4@iw#Bd3Inz=yN_=lAOW}m`;ETV@6jo!7uqW>|=H$aYRfv*gROQA+FRg#GQX1
zB~VJX`!1r(<w-u8fxNO@1V2u*C&Q1MhrUe9c;QUqJA-|hNg(Ar+7XSc1%w2<VTwO#
z76jt41bAXO2lDi0{EjA$NHgDme){dq7e&B4fbuIpWxzrQ#CJK^&u_KdA~}Hxt!AWo
zp9A{<Os-`-+N0;Kte=*Eo~4$xkmNOiVAs{juhpPC3b*+ke)IcgRVs`CS@@Rj8horl
z#yY0yu3HNNDUFsp>S?EV5X5=!dUFGW*f)zAuVu|!1&rYU)@S(zX*m6XB(m9d1yM_f
zKdy&qx!emthKlozDfm=eatBK6D8@!J1WeD}Pulv|mkHpjZ#x}4m)!L=8%%ih9uZ14
zA~?q_k;x}IEykVZzQTve7@H%j*8LKM+#D`uAi6&1kA6%}smBv^J-_5$NrgC|dMx}d
zIL$bxm$-QQF9GQ`1^2I%4X+pkC`vwF=y1_=R?CfNE{0nt+*aS(>?Efyf!Wk7aK2vk
zK#8)^{breD%9v%P^mQubtbbGT0t(ym<0CaOHu;*Qs5gU&d|}|BNj&(c>#Sb>>m0>l
zTU+~+=<4&U1GaCXJYDc{TatHq<xoh=Z`X?*c@Zo5i^S(t|A2d9=u2cnDy3-WzW>{u
zWR?bxyHHkY=^X;{Uj&#MR0c}s57k)V<g9jgp>ZMo(W)p%T=KMzr!|xxkaFZht>6YZ
zysOv=LlMg<i&Cl_WLl-T*`T!^&F1S9yrN=fGnEz&BZ`L`Rz7^+^S5{$b@N9bTfJk%
z_n2GXW2VB(W`uOd=a?vbt$QQr6y~z1)Hh;Fj-wwJi`LH1ZF0K4w)E)pgye9rL||ZC
zh>6tIrpHh5>A6w0viGzhh<kq=l_0$k`BY9^h7qNvA5NF<Tz{_KTIXajOFU5MQTO5M
zU{bUyDA&A^jv%gI6H!y(T-QD0^v9-*?e=qNcFsE<xQj@?WIdYTwfa&qJJF>O5XLeC
z{Neh%YTrkrRdsCq1(v5_@~MK+-I=C~CEg6SaeJ}^R0*&rz{<5rINs9IVn@;iTjY5&
zer}W_zQxrj>n}T`<bqXCmH^kK?j|v}H`|=8e^RfA7GG@N3Q8Y~`A0NZ{0j$`-V$2>
z*o-nE9e}jspa5~b*ajRJTR=7*PZ##u|Fg%M)8`M@sZ1@MpiX}B`*}@?QbWlhmMCvT
zUZ!DNa!=`c4bZU-K?3*{>4<1t#BuM;!nK!ukAh*U=E+++1O(tzAss##avqKcOG&Qj
zqLNIN@_w6^C_Ki@b5tzj><r?)NP1a&IGJPj{b?g2oLlc?s`};}4Nh243<AX`m|^=6
zsBQB*=zUAQ`-yM8V8f&&@ed?P!Vk>&f+v%$1fO?vfYA<sMujwnFk<B7vJ@w?j|^<{
zB%Zx^KIc}rl1cBdxv2_UatA`ORNIJSdU$JcZEfHHfnC+wV0;_&+fsU6?^I;3Rr$aQ
zxdC_W%>!}xRL<&NZ*-M@vR-p`)gpV<n*p3ItE%MX5NX)gD6H4E&xhOU;>A3}AGnik
z0`0Lc_m*6t%??kuIgz3hM#ymHwzT&yLDCd%tDPWzr)I$~98@id>Axl*(iO_oQps4k
zdW&2Pm?nYK70wx0*uvEoduS)eb+~QCG$T|X1BLEPhqu#bdC1x-9h_;??A_k3quvB~
z+OfZ0;3*mA7x_yoFzFfOv$-|i^Qar`NTMFie<&_qgidkcqr9#t_Q8SIV1<Rt$k3@Y
zSzCt(BZF=2Qnot%i1?^4qP(oW>h>AUDUU2V@9*pCNIj=%K#Qhyet2FV_EyV^wfrXw
zAeZo?*)1;iG*&qK^3A_|0d{~$oNSlK?MMN1ax}E#Z;{-cZ?!BvDp9&vA8<vN8XXTO
zb2;8OUmdv0E;O0X4iS{tVeOn{PwovCAYUFO#K2qL*isa-sdT_KqepWNCvSs~-c_Oz
z5rn@A3d}wA-5hN!N43A$VRqT5o4r7b@OOpXbMPm+J`v^tkdA}EyoYa{X=^mws3ueY
z%e2CHQ7?8a^Fhah{)(5^5^{C6W_^yP?Q?3q6SUrvKa+zbX6X$R=!3;UJV8wwUpGtV
z!-9Gnx=9OWy(ZG1mB2YWTBn0xTd*h|r?rh8wShpgmjIlA4;MV-iw`XsA%e^XdcH9%
z?JNeK{Hox_TT962CM%AnV9HPi8OX8<MpFBhgyY4`ezUuh&P+4i8-jiU92Q2o+=Ipn
z6q1ebNzM8mf&H#jCA{tXtK)|`>PQ2MAy^1!KKns_-9}j2(FsHt9|g-@U&GY(oqvM(
z5)95W&|+CUy4e9?S{`$0p`(UzPt}hYTk1-=bS5bl)26bHOsZIPUh&K~0DHP)e|O{?
zG?>4Z6<tvNHi#H`6vsd{#CWya11D+$B_l;=6{Hgzd(U>|>2Wh&70p&j=ADNM6Q6g+
zZ6L4v8Ggxy&||ZlWU!d<XWBn}L8t9;nEAzWMu(O(_1_xf*k8vk&7RSO24Idh($qE^
zeLwRu7)j==J0j8`@_2ArqGp6!)zw2nt!|NI{N=^PPD^uDSt6&?Z^Y;Q_?vBpOKAvY
zexKb>`l4{@XVS`@l^)UrO9)7Rjf#wmr{GK7Y3Ar&Z8<%K#}<>VT=pFL5aI4B20$X}
z2CaOC5PR3K_J>>GWL31@F95P0b<1j7S1QZ*+u(*kO!Lj~QB>Zbi)GNo@-6oB28eeA
zm5kf+d2{3a!K@)milrPoLu^rl^-7cBK|SCvwLVG8E?4S(0R;M6A7pdtmK$novTTPW
z$M|9SmW*ag<@lu3-AG1cFP^^4OkW}#N41#wkogHdnQJb}OZCsS`AEjn$ZX6q#qOo~
zq}g=hQXIJ++|_1yZ+m~<IH5Kbr{Hjn-`Y=J1?Nf^eR^~v?G2>{1@LK$oMH?s*CB0V
z7-$C}*w2m$?hP#e)JPhJGj>!mOfyM432#ZOO4HFmIN-m!F_V%)8tMY>-|sB(ksyr5
zEn13Cu7<y(R7=z#f-zig!LT|v8X)=@?E+1wFK1ZGP`!mp1{lLWM|e0ikwtF#+<Atp
zj_<Npw%6!!bDZ1d#(oN4L&ae{k%+iaD%CmRN3alNV`HfUB<a;kml{oDQ0)XuD>X_t
z=j=ZTUjQ;|gXR1-=8AW!QzMqrEeKZGk<PqepT=J~5$n%~uuhcSDJ%ZJhGYBV{(&{~
zH^7l1ZK{S&c)^(1{WEbsGXHhN+hHP|-T^cMLgWKD=jtk|o_AN=iltZx+zNQ_EOpyq
z0j%!3+(VQO)!CnE)mp5dW`HFtkw^DUZ4x#=N_LbEjjZmPwUxxWWehZ%HP70LQw5x>
zWI=mHOMi59%q}P(5@~xJ2k2R2T}+0LW>|}EY|XP9>!a`uLCt4Mdyjt3R|k`Bd;w*g
zV!hh-*Q;nJK|>IB5gyAv7@anFtGHNya(eyZF+2mvmL07}OF!Z0TJQbOHUJ4Oc&(c?
ze8!Q}eLsZOmJQ5+kO20YcXdT<teW@^OIr$VbprHsd!uny#{Lx2R-mpf!!Z#9dXGhU
zMgL&R^j({V1N{S5FlpZ9!~Q^6cq&<`m3`WvU71h8G|`OW$KhVLv=&_Ca^v)4ge4?R
zQf~}B=do8`tlw}X#}yZI(^eD&5XV||a(p%}h6JJ5!}(98L~QXC=htMK2m8(_&_Ry0
zjL(jJI~UJ>J4`8Z*llELllUuk7sKjbF;P+nM*621FFnw(Oa^{5Em1EU<cm5Y6vMU`
zi?m`6fa^rOAj%vjP5(tX^rb+4bYMV)E_%DxPpYh!)3V`1qVsd<F^>wTpA9PfWYZfN
z34ba*i!P;Bfx7N*ElMO(i7EB@B$O@eNu+7xN&JFytG00TLP)k7Q9&o`-WNqbPsM{U
zj^@@T;KB2BUFN0Z0Nd>kC#`Wgdv(}{N|jti8y^rPgHV`&m4u@4g5j*pnRDDzWWUq?
zbk?6iNA`I9bRlN^x2e4F0dcCA118=y7NF(4r8@l^1v%}Gt7sP@E8{m!yk4j~5ckpT
z$BMM`C12}ZU)Kk!1uZ%{K}SPdn*=v^sCdC$tPB{d<P{Z#`o(z<mRcSO=+&BVp-RQc
zd7o}h7At^)lZ$7#iVpJa`koMm7S)BxEE-G1^mQrUu+Vv2dwI-!N|!B5gvHY{(Aixf
zF@@dz`0);1Ziv1`KIc|raHee0`6!U@vr<}ED}e0|wQsrGGjmJ3_)!u~K}-W<gJlTT
zmq_P2+Tc3(PuuCONw}g)4j`1bVH`q<o^*gF3b=nQY2B$TCsVFGgGFERn<+p;gGYd^
zg`So!9-<i~l3$;?TU#_G=Cexg7s%qie_v|7=vrhx3lMLMJvMk5+mY&==<K*Sb)oA4
z>T&1&O+M4LTjU!1cxKHiV{01!E1ED<=K>#0?bs;!j~_pBrVvSzbj766X#^}{$@~2>
z5I4&P;7u1-5ffwFda*COm(#DEJG55hjyOi%j5wm813?>S(|T}(HIaOt5~cW%Vd)(V
z1>15%|HmPt@#Lqm%7*UL4z|1`)F!<$jkfbRQSo?Yog@yLC4Mng1)c*DC$KBxR5pk~
zQO)J;k8XAKpjN%30O?SytxY%c*)^r`9|uSm{>JLC@vSW-?90^m3-ZXus7LvX^vMVW
zb6H~iUwcGDAq%>LOfq&Y4&?0R7noa~=>8eDj!*=1=h}JXB&})HQG^ToVQ>v;@v)=Z
z{(FA3qE&gv2FI!L;v!Lc7dj;Dce|8LTFO^a%%dDRx?ivz$C8xWJ5|mAiybi5max^!
zAFmfrLWIB72K>9zQTV_pAA9u}X3GNc!4949jym(%F@XH8_Y{?B)+-PTv(nNg5E0<Y
z_j=!NZxdkbE?S@GMn=r~uecrn-Z;;7Yn4npKv;l3w43jY0hqPSqni8cuZeQkSVVkJ
z+|Fh)Q1eN!)ohwn-MtjmzUdATum97xCcC<n5%O(HOo)~yV21)uD2fVV7_IfEG`6ce
zE9h*+#l?}fpxfpC#K!4bV<+3BpXH+SrCU+Ixwy_%bpQ+PNIIvn?hEV%5yPeBUk;=%
z^^-TzbD#mVuL6$R-d9{avMJ5Ac=iq0=!@7uNP6F3%T#MVn_db(Z^hlvc{obm9Uhb!
z#JlF=c=U_9H$YMvDl`(N`B5{5%l>}k35?vbcyW%-U+Yz9Suz7eW$4l{XDlK%dH^8{
z1GQVwP0Wc~!}x%1F@l6MUp+FtM7<y@fgfEwqPRIJCT3@b>?Y#Gl6`|)rcFWqqcfz5
zbu&>OOAda5SmpGUVZ)~{nBXZApPQ2WkSAB%!fe!Bbe=@KFwt83Au7FNN(J6=Q6(-3
zFg!>S8b{RmVFjuj^(S;_qtU^Rr$Je*W)Yb*imj}7Y5@?HRH!4FkwqSz{E&{1Rh5<N
zggm<s!S*&`lvSzb{Ct-Ce}M03b4#>To@8V2{m<X)Ybzr$0uu{b-ldN5mROS<r<RQs
z8h->0m{{Ud)`zIVjJUQd-Lm%s=}K02kMn25ciRHCrm|!jVeycVv~@Jwva8J=uG5a2
z$@t}heNI*ibRg~vOxl-8=qTi2L{R-ZF?JS$C4+=Ph!WNNueHZL^Y%Lx-|l{KoVMS7
zg2R;lE60l1Q;Ebvcqj6z0B2@(Ok@>EcRT{#3=q08?+-EMi^V5=7iT&=6gDmd1TBKv
z1qBW|vV6O;35M0z^jZ7mpb#g(dIu!3gIUPQdaoRUD4}6q<*o@Gjs-N^@QjsC$KTB9
zY+{nPW(ULFUPG!+>g3B+;>3y5$s{AAWtlT|1P&U4t^d7fPXM})IG{v~3u%5BEG{*X
zCt(k45#Yf`QOw|ikK^!5axqW#NwdE=$DHc={k_|c%W{Dn4I#tjk`-rY4&7csI)WGo
zxV&6+dQ}#)*i+VUKs_p<o<NuHC*oIydh@BOd78dMFCB9@{orR+Y($sLM4wDYKObaw
z`desz-aP_W`;P<5`nfHtHH8KJhuE<h2lg}S6~X|5c{L9BhLd4F7U!{o5Q1JehoFw_
zt>Jj_KTI(Z3iMye)@;UH?*f+fu*pZXTU-Xy1(O=6K^qZhw^g(STez9&d?v-V?Myaq
z`fBs?$wrffH@gaTtWBWO1w5X5x*}a1BmKjMGBc#lo`gP;M%$QjWC}kc7Yh{9t_}KW
zw?~pO-GnZ(#r12phSkY=5|8rMW&3>$e)KJK_Eq^f^dA^Osmw{OT@XLI@{_S~wVmA`
z02WH%0yMlcc|^HI`X&JgII!63s7;D*NKw{Cacdd-gn$hER>q2vOZt(v-DQfw0Rn|}
zSl>Vuq=_Z{7n@0QSoL=dsX}qKt#8Gl)`SJR?*d07NjUHKXw#SKEs?{Z!Qgs*@Pv(j
z2A|=6J>#OrX_I9l)&dS7LI$UDGeW&lf^Ug%hKfb^A%^{#v?<YeW&Do!BS7r|PBCq@
zcsu*^A0Tx1H;*h60pQ(c<8L+K_giV|(}$mAK+@=VbB{gHlD_!O;Lvg_*iYxRgU(`o
z2BUV@nVml#FIMd4<s~K{E1@9zjtku3S66p81*B2}j_<SQee3C={lV0fmLqF><?eKr
zc!t;Y{Qd$8^`dRYe$f|^QnKF-8M%cWimF-@7z*6?mxll;OZZWbU)qvNJPEb&_c9!y
zw{GPvEq&xsDUVlW2ync$d6te5{#w(FE1ZB|S6iB$c`74o+7$T%NRO43qvn(b{VLyO
z_5|eTV!qe!u>KHXq-EFX9q(3EOC&i1k%FY82Uk&!k+;vPoOMgvqnge~V?DO=i@1h$
zTwPv#&dWZ2^Q)RjTpKu=!(&7r1f=zK$$LZnMV+b|wh1>f^0W7%SVv0RTUq<sYPtbg
zDyz?BI9*RsgspU6O42Zn+k<ihRXQ@R?QG$|)9t#vur%~EyRfi+V{T%1U9xgyqUJ%a
z8TE1+Q!D*^@DCuJSk5<N7QB(kyxZ8LO)Ie0mx!993f>6R^TTT*tnmIBkV){##Xb=a
z0c=Q6kh*AV+ugm^j8rYE_8KE?T7%CSPW`42-+L{QAWgvZGlc6#o>;h3*wv^9#5{vZ
z0=0Dn_1Mqh$}ha-HM9{LR`Qmy03^ohM9Wil>0`H&Z1esS^Tm}H5jva^X1YoX=_#L{
z(CBX-mviTa!`MJ$jv4ff2`aRU4;$;*c^hL8e|q+6w;8qXuu=xD=^zR%{VTk4H?po)
zrjVbvs^^!^X6&x9HJZ+alT*gvzDolg`fm<%EhGY@uNxb*Cf&)}s7+4Oqhzrz(_L?U
zRNHpMN}65I|0JOkF>AH|c=`)-(l6dqoBuMThyP|4610@~Kw_&pp&iUGPbYWiL_i|S
zbYZo`VRbc2m0q1<h!-PdCYvF2sQ!pQN*M?jb~m|_0XS3-kI`*scL5;K-EKE%+mr94
zOt0HWJ_GDg>Q9tsobdyxXEKlecl@_l0?FyfxY!L!)jGXWHN?B}{IsMd>#p073{S!a
zg-~`XwPQK%(%iSj+BKF29!KZMh!}+_vB8qcT$XeRTb`><H-@4MZ1L!dSr4++Lz{;5
zh@&Ew9&=>Y`NYU<N$+|s#BLS1N&-#AqquBds|lfLLH6UMrjXo?D;{c{3j?{B_ec8q
z-&>pR_Q_Dba_P1@wct<mG@0n?1&I3DHNHXJL6}yeN?Z8+m)X6*JQFuV)zPU>0`mpN
z))_4)Xk9pHV`xeR-hbRPS(uYYbfDGXO71CjYj9>tuY+HK`WJNy>`n*$;A$qu_g>W-
zcx9UY2&jZ`OLE|^+F&g*Qx1wxY{FK^(Sz|8lwEKt-@;|M@wRSyY+Y~zbWrqQ;B6eF
z*?nU7^S8>a-W~^QhK5D@UXGL=B7)-#9QTanfXOIqTyO`;PXAMdYl=08Dz^wpL|^s<
zU$@<o{f(pj%FIp_t=_8bxQm8_t{hqE9I@u9@507=t9G|+ET*kxx-&ritvi@u{@R#`
zSQ~~@ge}>7h--~~^7t4AgeZ#wi^y<H;pPvO=|YU>SGyzWi5k-lt0I3pz<45ylmw-G
za9jVgXWhmNgr|qwpRX6&FDu~Exc;R{Vf@vk66Hkk-T))#3o6`f`HRQ#5^pyP_G`Wt
z`&p~MF=1TrR++POr0916&p!FWVMHz?JHXg{;}-~sns*mO7{+_^KxTQ<1r>205P1um
z^@s@(EUSmc^`#<RT|0(>D3#B>c)Xu$S5#n{x{|UoqHW8_D9<lmmJjH;RKJ|xeHOX_
zQ(VY&Du%xww7ucvvKRd<0Y)6FAgPBZt<Ej92z&7&+(*w>$R6q1Mnt_BY?bbGFjUXE
z_@VMyEkejRd{Uh2Fkc}>8iS(Bbc)cLyJ7UBrFe2J<ejNs0zM%|1_F+bm#mK}jA~?1
zguC#TWbiknthw+6Sijg~6A=9#0%ynP(@(M?{kU(p^W3q0ETbY6YMN<UgbzA?t?Msi
z@$$OfK<%yRaQri2jQSDX>I|9hEKC8y^0#L#kLf9%zCSuHh%U`kaQo6j;Qxi?1c{Z9
zrvMRle@>>B#5t%}N>zdQqg216(P%BX0yjVgO+?yZ0Ro#&ZUK4kWu0JT?iMZo<Hm-H
z-U74c{Lf04Br-htwzaZ&3L}OnDrYTPzk$o;jaBg<u%hBJFiic%CjYh!CWq)Pw7Oq=
z0JHYwOu7F2MBv~I^W5hIVaPHhA+iCogNSM5_uBcJzFYMcN7M7x@qmM=1|RH=UF7Yd
zQ%CQ6x=m17O=1XX2_X`Z`UIsl`N<lfw$}q5>ejic_F8LxJdl@{?maWx+%dV!Px<i$
zMl`g)p}X#4L{zN~kM`5weC=O@4CYz*j$hpKA2hl(|EplkuFB;6L(r&Chnd?4st<XT
zwYcAbx84QdZRN>{-0ojqvZvBzU69bqlqa`3pR@saR@IW3KN5}W`u*#0j;dG7tw|fg
z#O;Q|T-veN1{BFRdyk_jLN`Jr{TGgS1G%{(h^Elsntl}K?*?_*d!#Pk`ZWZ#=@<;Y
zo>iLR<YX#ZgZxuSPhBvJCir^P>5H*!%P6;a^D)o$spO2$CNr;GuUJo1z))&~;-W%k
zlQ>0{Xo3jC+^}jIKfUU#U8H<7Y>+^%!myH)Q~jO@ZoV5PCW-(*XyhPp#6S&}&q*%E
zM)WhuN{vU2f=5Z=S2Wp`MPW{NW99o~I5}E6jJ^&R2DC4TirP9^qx_kR+nsjg=cd;0
z{lD7?{dO6im`nRBD=U5yx&q>YnGEd7F{~$N$6Ao(mFs1IE9p|QP8xwHip{!?ZNG49
z+?60z(^f7_B+^;w3bJ6Aq1ku+dH)5_s;{A}`g6_l@OzGqw>{mkO~!S(2PdEV<)&Xf
z&|2wnkjX>e5%NdK!COgqe)l>d9O*~0Dk8yc*yhrtX)H6Xyf*{RtjX+lI!MO@Hi&EE
zqus<r^xl+_#c~U$q^OJ9tI_&Yx+NA9Q!BCXaOc2fF6O?YQAcXo5Y7hL*st&nSBXbJ
zj%vw&M2|@}mfr)CMIXGl%<R&8wjVlFeCyp(Xu7@D?Fwjd=}IMmM^|r9Z?8mAK^T<y
zS~Cuk@Dz?Q&=(xntQ54zZQ|<ul0dNYtNo~_OFTnw?i(FO(u{xn6`sFaXKMe%Gx2@M
z04uH0VXvoXDp2n%`Vo+x#C&v(jN&PMDtN}_0PZM7DCj&UQ22ZJbC{v3*BP&kM2et0
z7Ol(;EB5KpVmpeoKOi?HJyZk17rLA79y;dUpOv+GVXC_!>QAh1Xp{DomCpMsJ4R(D
zFu?&92KZ5LiswUVFzEvhzd7OGfR`3^IZMc)wXOE2P9h5#zRkhWld&|15q@h2vIRTW
zovzn=^CeSXK0%Czj|;5e6G>E#%tpwUB$20oW6;GRR8=UBgf@eh!b{ktZT~DoVLR0s
zwm$nVUR=*ZQYO;yxkaN56@osGT=^_XwQc|{sWT0p>$lyvvId(A&wclYg^~)v>9dd{
z{x%rHbVQYucrBaULcqNSQFvky$T<@w5ZH?eyr-N97(NQe#3&>LPgv6u+GkhA(W?pG
zx%3al&^)+tT@iX9Bz{VF&hp8#>-TKugkJvGjDXiR4X10SfjgGAqJ~3Sr~3oH)om*x
zh8&6wyI*82ry=JnW>!uMZV7L~IOn=xzfcJ*LA7V{-u>KWJCzTJfWM!xU5YFc@yw_N
zD0B=@XM5s1zQbS@>`Uz<k_t$JCm4_k|Hvb6>nEBlzs@KKgQs!Ry%*1BIy&n65Z{Tw
z(Y-D~^7Boloy*xoG<Kio>GpxOH!7S9d7nJG$qS%Y5i&oIa3cD@LkVt^jZTOlSyf7v
zH;*03CYu3lx%+4LP~mp*7p%$JzmELzKmNQeXpaHVPZ4}hm|vpe<7aX-Zz5^D-pZh{
z6C@NDn-_dad5wnPezO~w)Fez5><g$~LXK#$&RB@PYr(o-0hzS_yE<|WHt0Pc8sSp1
zkkw%$Hm%~&;NV7LP8?oX64qPQYPJF8Tx~@Tq<5vHM6Si$ujRfQ_nuKndD9wFXOl;6
zUkS-WEMHHi#{{M6Q{>}|37R<*=}?bt;^DhkbN3m=xPMb6uh4F?yXrZ^17Gs+SZ_oa
z67;uHH(I_;k|il!ME3J1QI*N7)%U`PRPUe!A=sCKdzpf|sS4kie~rb*@-4A0TZY(o
zc;DeSHVOI+4_1Di7@z@B&Ay3;2z@7~@iWX`pRrXd4qFG2uLu1aqLE`7SW1{Kv)?J;
z6M`j#X7+V2_*=~Z7VPaW<i$_VHj_tDFU85HWEHK?fR25Rj{XKOYZ)D6O4|<@^Mv7)
z$6CscCyBP^-1Ny}V+uT=ZK<){JiM}tUm7|)xz@I>lH8Q<JJdDkO}$Q}u!=o=NFaKm
zLcEL6W<&2MpotDARCE-(WS5FEG~oU0`)-ZxesE`;(fb?zMA^eqzQc&g8?(8Aa**K|
zMp2Yq-ug3YDYouuyX!$y)#(qE*LuHbDv|Kp4>VZe7=ZhjqRY9~ihMp6gSROp=nhD;
zre6wNnOUzxP+o1*!1!2`W5U`AUKqi{bA7rDD%}S{;cYf&lK^eDOL&ybn))Z&3FrlX
ztr=U8388;SJGM^-koi_}y3P&=8%q~F_>-oDi9)oJk`!Wgt=YZ&{mpOPl%BKiKqLbY
zPo111O2Xe46R!C~Avd~xhsm7gCzjCd`GWMo-&QQHVsQkoYK}}sDoGh2r>-hdDlFn$
zk9y<bihjRi3t8<L{J3rD$d!yHCNyG+F%~IdHGLXwZ7ya+%k@Rmn_=s`9HImnK~zaT
z{w(jA_jzy<#<J!0!a7T{L_q;uL%F)?{&f(Csd8rOo>0_1oC+*Ot3v}Y0SDp?1H#}9
zLmM|cJw8f3TKxbs!?{q~!@?_cgj3Y-Z9)uX_~A>J2*h~q*_w+-m7L1)%B?${2@3k!
zzZujr=0y((Ro3DOFsBSf*$|If2MN$l<TJ&1HbYn3v|w(;`lx)JiRhz$if%BZ3lt5K
zn6)<u|4799-LI*91w=zPCxZCHU&dYH^TS&HL=%)9gr&K+{;|JFm%ov)QP>ICNq67u
zyl%^Ou%ghRh6|IZ2bI7(UUCh6fk65b$Sh77ow>80!wI@Drp$&~Q}gb{i<)xjznDx$
z;4gSj&iu@#MUXXuMTscp1XF{e2RF7if5eCRlZ;Gv-eyj;vO1j7TYe3kDV5cy2iY=x
zRmB?{m~&ysqVpahv|~x=Ve!7Eom6LLuu4&t_shV&A?p!U^N!FqK?=L<3wWhepTw-S
zFOMl>s<+U_Px;y$*lPpw*M`UA>m7a_)oF=(iK?8O;>g%Z!q%|&<5h2Qf3JZv`R4|z
z7WAXu0s<F#g(LR0g+>$xeGExHsD$74__1Mj;N@}UW;SUo5$Iyic7G-Lt*A+Fz~hNR
z@F8noxdrx*fDMy7K&#SebSVO)e0bz)<s2ra_Zx^wc$939W!!;4S^w&m9vH!qQ}$6Z
z4Q$fu&i4c(n{}m*(Q(YF^%98n*Xt|RmKPRo4n}+mC*#LbNR$Kz2G1iTDSxK4<oSg!
z`hmGdHb#TfLY14tPx<?z9&b5=ELoy}$rq#?NYrc8PbK-IrE#x*4zY9)6s!q_J&;o2
z*?C4zEDNPlcA`kdob8nBxqR_cQ4#rENC_J$>C54{5Dt9Kwa(;;A0Dy&o0cwg&M@OH
z+q}cP3HS<8_{@%xDAA}w_6<%F8<IvwKgbvoA~TA-aC^c67XuTu`9D+{xJ2bEpyUVO
z8Q>xm;x(uJI-=EfXz;7>i7t&qxW8Z3EZhO&e`;OY$;&c#^3CX067yVVIBh)}fDrOS
z`u=i%OTvd`5zBz}a5-ka%sv4qJ)Yh->&V|H%_rlqWT(Me^j;pEOUfoZvmBb&poICK
zRYRIsGh&g=U62an(1z;b8heLBlaY*erHqEu@6M`IovEp!0GSci);e&k|HARK+9?if
zdO8lrr^38Gr~A%r=o5I2wrjh(adCXfBdWrQS95$8=EUy+GGT{kYwNo@%VQ`7Wk}Eu
zdBX*vfWEVK6nF410-0FWLT&n)d87Yi0obG6huL3HWZG|F5d)8+$l*|#mjhmD$L3V=
zWv|FSS5DCUvxa1a4Q!z`1mRr)VKbi>M7S6l`IZL9l)DVyeX1c+=D`eWpX6FV7hJps
z4SZlEgUf2LfYy7=hi2#WCn8vWhs#BLixX>>O9yZvfQ%^qwP`|SZgq$3zK|hf*u$ya
zROL9z!v^k{UN@2>lFWyiuZGMR;%Xx>E_^(u=|yF$gyjFxqWQp;7c?8zoyqB%QQ&DA
z{&*=Q`@jctV(1%Z4SC>Vw6rXs&2YQ;$&Iln%Ziog-W$}62o1o2so%{sYtHYEmdL5J
z!JFYx8SVLY^XKg1_${t6`x}hTqg6|k@QxK4S3<+yz^(zpx4Ucsl|pI)vh|D`$|4)q
zBw{CHw69T=bTa0lW9!Y1_@=g<l15q3<D`3+2#Ee<9}&M3a3+9dN$r)=fYMZ#ogE#K
zC@3$pPL<vBBO-5*{XD6EhSEFG3Y2(FazMVXBE*zCPYd$Yib;SlBc*l$*vnLloox`?
zp+IM8tTsbJ|JNL96RfJ8S8uIDJJ4U{GSOZw$EbE!g0|C^=^W?X9e6?!$ZK+QEzW~t
zq}tehiQWkmj^zt_b#?)4V4)83rl+HUUyce!bQ!54$i+e|^a_rUX>4TY&GV9V;W7&k
z_Nz8{)y`~XH3E6Ht6Y}zC}#bt#|tT88epE(Fv;|hB)bOIevmK8DuJ^d2yY8l@`5=W
z3|CrO6rDYUhCTxO_+62qugg-yA_CTkFwigqSZ?eMODNaM*<X1H6ArDV2PXK(zw*x$
zGi24F`L7^;3M@?%65?Atc?_Slzx)99SAjuoFQ4-odK0C^{H;<6Hil77?Xrtn64-T1
z;rZc9C;rHej*TQZ@8BIAI+BB4V(IGoK8mpWUUvx@KviJ(1{K|0l=mG7uDI%ZnV~qx
zd&191ed(pm*klE=d_VV+WnFO*cE8qJaTf6&@TuH=ytfIjpB|Ad8H8%xROIHhK#>hg
z^EQXa9fW%WvrRLs=NIw~ck9%3E<86q-TB2kOI0d|KMK;5$KMV1cHg>Idc&dH>LP}d
z_UeDj+9wcb(^O<oEz5Va#3MBH9xvz`6)d_RE5g5(X&GSK{GBag$-j!4LHbz$!E9Th
z7r$}JhAiAv{1qFzBC=b^{D3UtB-EV^P%_B35mfjk!9_(|fbn8?snK#wgr!NgW`BXs
zc4N%R7ix8N44c$?A(VD9x>El06ykK!tBc&wzStp;W-ZqEfNM58qDT?m>9KlBReGBl
zk^qtWf+^!`#O-Z07nL)fdO<0(7$dXvr>#*dO0`w!7uk$wCKmVu(-8>M<-L(#4T@8#
z0TRvHPyMl0i0%WB-TS+)GH>tEU;Ohl6zSJ;rj~mYY!WH?3MEoO5e4jt@~XLEXEO{^
zs{X0sBv@3EYviDOeDNm2&oXy~aIT&6w{_*2x{5-OOjJ5l2iP?-pPG^GY=tXD19rFs
z0$zQ)xa%3r!Of{4ZJ`_l0tY+ot=_ouqjwDfeKh~XSTlb9b?dILa^~^ioU!0memoqN
z?y)QXLYL;z_BWdhExtLLE&vRx;e;}gcVKQXB~y$QrAvySlzWB=K_zBjmkh9DGdyFd
zzvB4`W|(zAqK*=e^D9^jZ<_}^9aZRl(<?pRc%<fQg(NS<tw5g~t4L28oElQq#zVs<
z>+uTF`O>An{zHk9rQDQ4f03W>a&t&hl*tO_-VB3fO5AR-Uyzafh~l-djd1=@7fN$L
zT@U5TXZC%(F3jPODqnG3hMhcmApAd-&lrTgQzfzMvTf;T*&{u@)bGGh1pPDE425w;
z$4G@TDsZw)KE%T}zZ-37@lC*|M0A5^G+t*>g2S*s{neVkZ8O{ma}p8t5oYiaa}z<B
zKyuq0Nw$mOGsw)A#1~=kPKa#-#Auo_vF##xlMBjxW-si_oJy`|2O>1U;Io|QY;8=^
z1;VkTqvR&0!go4asL}1r2E4)3_jGxIrhIwt3=><uE;WiEL>H3CmKR)Ay&ywX`&8nV
zY{v%sGHAD+62}6eNH1((Mv}zI`1<wndcAGgD19Jgy0;~K_9$b{=4nxvL?-YYY@=@;
z$gM?XTLGjUfjsxl=4Jq|=nTuTpxdL1<HG|5vB0Sz15pAH*hd_hnM2DYqK`_OH^ZZe
zfy|hc!en8ZQ~aD-fKDtSm7w)Gm|{Q08>~(@R`$Z#$Op68Jg&hvxgwgxqJZx$fNgUh
zactqa=by9MH(%RwhGD{3A`2iCf@<-<s#pd87fs(BSLfsYosH#Mb}cTqY}>YNYuUD2
zu4UWiTDEQbx%c@#zyEvH>74sqc<V9(Uq$(|qJlrO+G24F@C&k9?XpQpOW}!F^xALx
zST(xviwhf^ewqJCo`Cw4OUj@0`uT=B^G_H{26lY9h)nQw(+dvW_4Z=bA2AluWjxBH
zTo3r5HPN=OaTe>cGgcV?hDI-V^k*tK-|%W%h6KC(-aVKv*1<hkX|e<a^6b`2N4&O%
z1=QSE<cyZanbFpzrzVPj8*9Xqsc4SOpooVjTBvm6)8@hZ(PgfVm;&xJr6H49gp!O2
z$4zL#g1XRumQV>oxkMoKkkd-BTVKMytYF`tt5kL1lK}qtKQqKa!1oi-M3Dv>WP*xJ
zA&fKj<RIQo@4EE$I{zRjs;);fZtPX@D`$g`>c&?|#|pSyN1{B0fc=Hsk!V*Q&1+15
zi%ur8;M{bwJ&<nS8U9t&LCf{d75yoMAAJ0QUY&&%5#d+Qkxy|Xu0uP6s6(e)wHt&s
zv*n4Bom_J?>dpzPg%EUiLQWUxAOG~9a4YUcfXW+7A%yIiLJ9Ze@{N&l?*QYUl9zw;
zV@t(W-L-PcfQF0Nbi(_(_0UW@_c$iKAb|Of&Mtgn6tSyDUbQV=DxC|R{-UMmZ2f!#
zs?FyY$C<9Pe(S`T0Q-B$>W0sCMjf0hQLC3(7{x<ZgYjmUe(03<>|M}+DFUYe`6gEd
zAj0H*nd!%0;R`x|vGsQMHBeAQoq~<aw-GZ+L9{s8z<HtA?Gg#)o~ug?5e$<D!S5sA
zf?_0&8x~J@LDpcDN0|PUQqjSsrlJK@0lh-V+%t_%SW&XESB8LlFUO~CE(V`#+pEMs
zx*PxZYP*{r)t8R$^UkMN?a%Jp?Ewj3>KWkb-%hRn*cUdcH=Am6KW^9adp{?9K8&XL
zpF2nIUs3D+s_-@TOA9W_({V}e+FYCS@hgg5ci@@lNs>if@%Pw$a|4Ch;Y8k*j%%k?
zP$N1C1szp?MbtQYv~OcSmDr8Yiy^1;*o3hH^}b8iUGr;W=J1$3tVKmdABBYs)_PAc
z_XU0%<xVvGk`><Hmq<7RGUtj%S^TxQnuJB3S{EXek=f(nhAzF(6pUI@U^T5J``ecc
z36*C1bV2CizZOg8FwNG_5$t+?M*rAvrS|H4L6P!3k7HXoQ#nM~ta8)8C6<dNUG;VS
zh-E7YKXp&3zu1S0C|wN5t276$`VT^y9-P@q%x*#8Xh1LsHrmw88C?%i2>%YuPm=2q
z23t^8wciOyGhH%{0eJ{jM}o92&G&5w&zTlQb4zl2TY>ppHj@Ixl$m<=L6G8<loc2B
zC^_+!p5NCTiIA4x?_9s?8xBV;bY>4n<1e<Ei~};lJk(RiS75=W(`(vMm(bAGYQkgK
zzKhwiB5f9qiGZ9=WD?!f?P~5I0!?uJ8}l>3^(1q%nISTNh7n}7ZX3FHzdxKO(H4|C
z^SNzRmgOE_%e!&_A9S~W)BXG&aPP=wGhv`LIh4^HJK{2Yo6sXFUlX<NFio3PN07l&
zSr-Qt_YUu^@$muDEdX~@PG?;FRNZF&QCq<==>`FS;%pK=b{Q<sHyX@=39+g$6Zjl$
z1GyGJgaVK@b2cQCdCdYJi|)hnybCDY>d8<>`g+a*sm~u~HsgzN!{F<h1@N~tv%@_@
zXchJ;@8@}BKJC`)VZsYS6iWcW2HZY&LI6zAeI7CQj0*M%4q?02QtO4)!tQc*brcS~
zDRG#@Q_rm>xxgQ^QmqD!)W;1{;?gX^xpw5GqOQmY;LI#W{Ade-ZB3;Y5}KV*0c*DT
zEC4V<zSqCZv<5A`>u?wrbKxtsKN8Dk^uvQR!mj*bZ(?FKh_j&t8ewk|Ge8YZ4vGtG
zX7#y)>(Nu-oB~DGLmiuY;S)f~Q{S#9Y@xkrp$MyALNzbN3}iXHIqZM{0TB@~8D@!i
zW<9egLR(9!X6A8)?tAkIS~yFS-C(1#m2WNWLLL4_nnzNH*c5wz>ho5X&f_9p{6AL_
z9T%KOdR;HcA8Ao>Nj!ym>yZ9N&&fzBxDXOwp9iNd;XYoRZh2)6c+eBjBlR|8-<<D0
z7tnTJo!tUJ?_BGx&o9>d6nse4u(fs_LsUk9PRoc#N2CSI#pb>wC@QDWR#e^YLcBE$
z+N~wh)78yUYXOK~$#2_iXXCIf#ik21Rd~o%>s!C$@~S-_``7~UZE+tV)_IvKj{|<L
z{Ov@qX0hdxsLnw6F!u#r!^gwD((3d5-l6BMUox>m=ki?j^k(8Qbe9f@Iwr+8L0mpg
zD3t(Hl<B=D1K%TlMzRoiw-3T<z?(;)M|XWACE+==mpE=pisFDXU<OlN6rGtt8ncD}
z1DL=7^%(eLQ{-kGRAp!gr@?l?U^TLj%U~Tw*Jed|hP_PM>3Y3dW=j71D{|U>SF(X@
zX;pN~^<i8yh{3JI0xRP!-VQwXeGh4e`N|Z#*|bil>-j(4mhcU~pz42G&)a{Nk?HOb
zXDC1~722~8V3qdIOfdjQM|d!q9cuY7{to>&b=rZJQdQT1JLk-7IQVP8d~_qeyb;2j
zsxl%WxzA{Ude%^dPj5_!imIxET<Ls;kS<D|Iu_ZrVU?-kFQ&htKC$7~VrJ_s=VUj^
z5*i-9z45y$^SYQQ2SLCl@lPQP7#|u9jT3<G-@k$V@KRCD&W*=ha<MO!+Wqtf4``mO
zmZtjLUIG2}3^;OO(eHU|gRm>ZAgsukl!orlfQLsu`#W`NTY+D7fc1gr`e+wuTsyeN
zZ+=6MC$mnI87%eJBz-_=%}^J6)0Mao;yESZB{m#{p7%#Epgg@sR8$&o)tZ5rk|JJ<
z?LbTs@Px0$e>I}35*FRVM_iS6Uvt9D6Rf{dUI()>@y4NL!Yhq&YNy!#iD13ybxTID
z(hiAw%rkf8J0R3+7ljIqLwKRyBp9b&vfiFr9j)roEX)C@3}Vp5ASC7xO3w!*t1p+^
zAHUQ~+(sXJjXRrCjrwwl#cMdT&U_nh%2Ge+bL;}&*1{ar_wF(|oxrIF&|(1=s6M!|
z1Dcd-YYgk}d7J(N_{3xbKVMeMC}Ck3DIwK;Djw;;6*HM```kFb{Ba2@Gd;MyGFiMM
zqg`)KeNCw~=fIz2T((1zesxW<uaukezWn+!9<`XE+JD>v-4{Ys@%nHM_=g{ZG(E->
zSWkv9DbkpbBNMR86P0o88n7tE^-B+?k9^Ct9)Wu4+vOqEd7MowDQRIbSW%6D|C#%#
z-x6ns=M8Uh2pfsCZ~f;QNj_^rqnr1=_w|C1)v%DLBBFTL4y+;(hP??tvbPO4lHd=5
zY6m}fxAbth!33XE*UZc(iK^R8-=I|Y$IGeS<6)%e|Lks4BL6^AFB@~ypq|Z;p6E?G
zJZf5D*G}`^zNK)bt#KRm*GEFU=c%wFsniuVG7>OSlbr3eRBSomH_(*?)C<fLs&b9J
zzWp};u}TgVqSDq1IF{O7&xaxwsz4%(d1?I=n|tuu1x|;|c<sd>&`E$?q7*>GACv5#
z$UKn3tl_?RF#%w_?F0OA#tl55`fIPiATV+C{d&!mA`O_SuEhuuA%6l}B^*-*Y(%ty
zCV&H=z_}N);d8-d$t(Z0P%3V=oxMr!73zVOjpGpgZ6f36G7?8gaM|LxZ+Tg74RBsS
zjM!WFFaZNy#hs}Fy+++#J~RRLOfw-vSQgPODs%eWJus~T&;wvM`<79~uVODgl%kAi
zmXPZ<5CpyWV>P_E=r=yg)jqHg>>0(@zwp=>&3L=-agk~^n?G+4AnPbQNE(TDbD93d
zZs;4LL(xSSezv#u)$K*OY(QHDTMAjnr0IGGq%r==w=YNL^5ZJ_*p-=BJus~Umi2*P
z$lp9~(ESHsjbtp?{Yh#A?cfy|lF)j&_1M=e%<nz(+)`KBBLMR+Fp)J+>HM3Jk#2bh
zQ7^REsj*Hbv^F@cGDZiaKbw2F(Dd2~O5N9>;q5}Cks6LAx!c;g#>&fn4Xgdud{<n&
z0NiuN*#WBg8vOvlSnSPoFYkk1%Vl~3RY<}`i<}-WdKq4_@txpQ_wu`H?k!&PjbtZ7
z8bi$*X5u*#ZN@yC6s=sgAzm?a;C;rizoYMX`v3Rc0h&3_z)^CM)dq*#V#Lgfa;r?u
z{vC1{-sTKO)6G_^$EC+1-?v%Hmbh)<-S|U>N5ZA&V63@CGcq1X(4nBbK;^ju5tbDY
zO7V5!To3gW!*JN~6nel%m2Y6zaZEFFB)RDj^xNF9h~?_Jo(N*-ag`vJyZdIB5Z&XW
z`F*qm90ic1ZvGQ&b>ln!AR@MhCb~Mz^~>S(fueN0KbZ$2+1lQH9!cA^x@~!2Vk;47
zy#&F;NZcJ`hZvjCgwz#;T?pr{9dZ>6nR8a)Ko7qtR#`?{J_DdDJvz+8>H6@aJa+h1
zQ4vOKt3J}1k4!GxuSF>yEPC}3*aRx-y7IvIMhl~+?T>e8Yg|d!!0mK{i<M4FbeN6B
z2VMJl#ohqC^q~ZDsnE`kgB_9FR?<LGW6tAa96*atNBiT%7Eqw8xSZ$rw~`m_Q?pFF
zlQorK$GSmS|FS>P@f1G;FWhh>=}<>I0HRU=&Uk~D%k|7Vf6fVz>Br?8+k(H2O^s%=
zIxW0;ME50N$mD+a5d|{*%%(FYmx5tbTyeAkyDl0W#N+umlTB&({s7U8=X8#Ah>DiT
zyoiPi;X4+?%UNT7H9ep`u}DB<GXfNYfNZOEdSSz(57E&p#P}zKuVdB7F+N!>o%lD1
zqo0?Z2H0^F5vcT1^R{V>7b8{TiKEZqfw#1FZ>~|BM9E(Tzgzdvsg?v5rSmCpJ9z8=
zEwILZ_=U`~)^1iExXdZ>`oty~Val~-JzY6#J9%X97UJhf=zIB8lLO)#KoniD-JMCg
zb)M2TS)j8+d;x^0ZEd^){krM*YG(yEps}^XaKM`}8Q<Vq=D`$gRG$#yy`}*d{9#OA
zSl$8I#;_Km04}M^?ya#yZ$4o;)j7Jp+lCl|h({7n^RtzwpYGUad}1+PLa@PAiJ^$M
z3qW*q72CK!4XM6l@mR?!y?3qx&98>x7wPCX20_KCvE&MX6ap~O8&@GpPFjXg*PSl8
zPP=p=;sG~CcsIxsNc{t($fq7c4{>Szg^P<z?|)2bM?3e}O*8&_O|JLZH6Z`PpWxl&
zgP=N;=p)=(#9ENU-CwltFz1@R#VXxf!oLdRSNg|-cYg%{Vn5ab76M{>Gtw_rzt2}I
z*pqJHg8iLEiE&oLhZcV8bUQ6xl%HrysM>I0nuv5DT-^khNIVi_6DH3@OPjD-;N|Un
z*`OX{;<alHj_-`sqq5FOsRtaXK!Q}aogP`+NRsrb#Pmo{p#uW9sGHtucqf;qG+_F1
zoABD0HCD;wGP|fuce=TO?_9OAY#kf3q+Qrw-<UBCZmca(>=&I5WGcQXmJd*eY11pc
zMuMb{Z2ndWsZkv_n_d2-3aA61S3z?%>YUTvy!BpPM?L|kgWKikp8IIcSL^0FB;#0Y
zUc<j53K#xaM6h)~Oz(~W;qk?KEp5Pp%ybi>7X%3y6BAntJ2+)na<WPyrA!9!nb-um
z4NI)>)3g*)4~M@KYrV3S+r!t56<GDk5arC~9aHaVbH^2AHajQW%EB7i`t%E$*p-f<
zS6>$4rVv#C1{A>e6hZ@*yuD*tKQ7l{^s?#@`9vnkix&zRl!oa&-eIh0*o!Y?!tsi%
z;FoBG4w6<RV9iuXwkF$3OEjS`cT^$=gmwW{R0q%_ne@c9+N`#90@OSM5a>o0hhxh+
zV8qV){nAwfvY8Ja;X&Y2P5_Aa@K`Ok6#g}l$z#F^a!GHiWqjfP)SuP>_H%V^(SOOe
zXHGO`EF!)iEHt%lhhD4M_S1f+5FEOm94G5M?Wzi$fI8!DyO2(+xplL+{A48tf6FU1
zG0`fGzg2@G%&DW?^fiXz>;Cqlo}Qbebx{u!UAIIgOa;6?3kaV;Tfp+J`rY_Cdu=af
zc7k4HU%@XqzH4KuFO}tGgBdmjl|O&wBc<bCd0hz8;aZLlMiQMA7vR*Dhtm#)dg#8?
z&ykA&!r1X~K5gRT*D2q;yybr-_8ib-zq-a3_y+ax;RopXy`BSxd|Wnbepaug%>I~^
zr2N`1>M?<vj`)_OX#$p-4(+CKu`K-#C$v|2hH@5_aZHILU&|yiK#P)f#<O%8?g5>w
z5EODPo_G#~O9oFsK7Xg(r|)XOlaP1LMyksefFgSVihY9=j&o05AYzHtygO==O(ZSd
ze!XzPZM@K=zd;FCpdLv=JpR>v?B`Nu7Km&AjT-OAL5x45^In3A&vn5xunLl4r1Y8~
zb|hsLa-MI18bB6?Un$$Mr?kM2{u!>Y1TbRnF~2l4(GF%qz$gIdNBG&4iJ`SrqIJB8
zD)ZKPY|=jvQe3{FD=QVtWqzzNXY*#f#oIaL=hd|AxHn$5Txi)>bRkyIHFpn<<+R(a
zLp`OIK^b{oE883(<@IjKr2T1qZ9}|k==K>(&Z*4@o*U%#PY}_ZOw7H|_Hnd}nl32j
z_O3ym-Bpn!73uGcOR3i>p*`ywL#%s5Pohqnb)K_7zUZh6Ft6Kvr_lE*>jJ>(Xh7+r
zqj$#bzBR7zeKc_Ax$N;=hXJtg;vJ3NyEUjsUua@hJrqPEN;g9SIB(&D(y$n4zsOHw
ziYe3CULSE4ykeeM)OKJ$dgGo=o*eGF%C3ANhZHY18UI@O729--$`)C!svI`fICz=x
z^oq*-!@94{;NJ7N2=pE5Yz5JGw+kynzIe(sbbM?<m2TLfz=gV!aJc|hIHx~BV3!f}
z(qVne8|S7$gWP_JTvOKpulv=$6GX`DqU09pfw0EYCBw~4B0!HWcNp5WXYN`<>gOGP
zboS>%pWo-}rr%4heg@w^^5Z|-y<OLf^MxBTnBT|qVI>Z8JNKucipo8(bmek*jFoY1
z51f=+_Af7oBB&B$?$)OJD7+n~B)@{~K*(*-)-J@bF=dBX3f-qzTm=5DTK586^{oJZ
zoi2sBx)_rY6S&5U4%5anE9<=*BD30j&TlgDs<3v!wLF(ocvAV`NE$-8$zEh1o$w1b
zI7WF!7acL7h*ifVJ>sh|x>ggu$!Nru5@gX}%>ug8C%?5>L$4$Mgxv`cGl~=m5Ax~i
z3}Y0ybxV`r5h#LIQv_XOo1mhjs4SogDRBUC9GTrd6op48PUKNUvA1<OndAJ)_t4cZ
z6|-~lA-62@OPC1?LL4kUu@?ddflUBRB8?0OK9C=VI0<AwiyU2_oNkOWRL3SOeagY~
z`-eExt63o8Qf@6wx6@{3+3#Z(($y>2Bc})P`E`}~`FWo8`;GR?jq-fHf`&*Uz<ZK4
z?r7@rRN4V8aj=hf7GwP}Qd^u{p{vti9<E(Nn)!KO3Q%<wkGrwFwwy7$JrTRk@<_{&
zCxo1?-p7B6fa#;uQ^XsSslrmEBy~SK;O|I4C31TC9ufZteL_^Ud97?O!hu;OX=eLJ
zh83ku79P=<pgc+yl_5i@E`H9oX4omX(ijUgQurYx5^;;<U+0w5;NEFG;}n@0{@W3m
zNyc8!U@z!zujSK)M}-JMupi1Q>mA-(Md@uXz@RRI6he1i&Wkj)Xs@l;<z^D(bnot-
zGBDapCTG3DZLKqWDq>C_W(A8LbWP}Q^Ggo883-gCw{hv{;-61pjGm72NYP6JjEaz)
z<f8_+aTL<-4l51EunmXuSS3BIULMYXu)8BKQ}u~G*f_<q0G1o%J0*zlR*}-!*_W=u
zJ&*Rw94}^<`{@`D)!3k05O&bkx~oD3f_A&WG#$@~Fwwnix6RkRL^;L2+soS(^3En_
zsIC&`xQw;UIowR$1!ONM-uJV5<j<#))9S8+3EfUC1`ROpVm19ag}Hc*$y!xi&hIUM
zY0#yI_pxCT6EmA=uA(&}xV2*3&gU1i)|)RwqNCA+1eOLI--m`|I=zfv9^2E@($%wC
zYwlF5wBLcWO2cCb#&H<<8i?p7nwysv3_`(^G9WZiCWTnlB2Iu}UpH0%{0g5VY%3i>
z<FAYetMVj7HLeDSA%dd-JIWk9gM<h^?FIF$<{2zHhkq6x$9`O+H37tr<B<bo0^UFE
zVj=pgH2kH-N=R&w@mn86{0#0j-eDMC!Ss7eN@SCDJZ;8OMaFiZtD;ZMP~3E}zcD?H
zxj`C($CgQZU|iD$va4SLh=uIBu8V<xKsw!2Ge+;sat%WDklE|;yR??BakX}<=hz-K
z;wgIsP6auSuCl3=dJp-)2Y-N4qX9;u+yH~K$k%U<6-_Z`btW{J^qQ3?wCNN$Q2FwO
zYjL1p(jGlzHYvLl+y!e*g{XuU&BED8!cguFk-DJVX7LX;`pSlA=6l!{&K)fLN08oL
zK*Ums`i;{E@Xk&1u-hI^t)nYI&o=J9th!x%b(i=w5F(5catf;rh7U1$9i;DuA%6}c
zbZjBMG;l^<N)D0D`9CedI&~3emFC_D?H3&J(Q9~-0GO~!oF%}T82}DlBAw%S=FVJZ
zY$WN@Z>he?0q{)x(rK%XeQycmvxVlgG*o?0QOd|G6$;a^U{6(8rb<G8AZifZhc%mq
zg`tRU=L{3)Vevuyir!vLTrN6ak^aC)^`7re(mif^ZB{Nj{iyda-RDQ*JlgE}9*{*u
z68IMZUgFYYzTup)#7B0vZR2e{_wy|^k+S>A!0%+PLe7jRJK%Bxp^)Oy?Lf5%T*i#Z
z7V4&TZ3al1{><_ExS18Z?0VF#J(UTGB;?x=t`n$f*1(fumd4%?4l-@6g5=L*n1c-W
zUurqbxpeyoj^%x$B6A2w!Yb!3_L<*LMgGcYzkZ77$Yw2N^-7A&=ee<bc(z>6kd=#I
z?m*St%3MN(wcz?j(lWCg<eFE1z8W*`y`DpeYaxwU+2=WXI>m=f>k+Rat8KOYZis9W
zI4o#C2**^@g+AHJU)H1|xNxS#DDCxj%RsBuB>2Purn9f|n*zZ@*1<=QROBupw9f<?
zG6!^vQ^y`ZU<2>5F^$`o036)Js!W1{{ssvwFm^3StE3Xh27BD7osRK2X_j-Cq_y9M
z5#{49NK&y~_9uY(b5r8``WHxDM{8NYKqRw)-CF92gK6J}me#}I{9DtHlG{K-H8;ff
zDOg<f@#h14=Rz&dPu9XUEQH4!W+s-q>^mA0n8o(~R*ryuh*DkpHk4yNJW(zTx*mk4
zM3@?yGs&?Ushju`=+5n9WR4y}7QH;PBXg!&7{sbZgZb#P&#SJ(F;T+S3EM%l?WV^O
z;~h+&ANm~Biw^I=_9-e)I~g?H;_(#S<J;A)K$l?3`nOp{h?N@(Jcq(qqL`m-SwAPX
zn{!J^;cW0V4$8GV%wgS<-x4h@Cv29Mp4OHQWAM0;1bIWLhE;sVoAA+e-3<<_)?w-E
z75Up83kJl6K`+21GpkQJ{L6F6aEU$R;A<AfkBJ8i!@Ws-M^|H-jj5S3r+98FM@rhH
z-3j8Kr^&w34p7N1hO_H#2LctYoBE8~)+g@bzejD(2ud;W^{xf^nCcJCQ?uz-MRnK9
zUHvqW&)2o6qNWWiLPNN9m{mZ!ngU^mb0{&f%~tJ*F7k^5SNc~rm)qyt6jycYjb#kS
z_nbA?4y>VW``*|o-1{f-+7^R;5dA@Uwn_s|zSfJ7+o=FH*2PftdIOD`(tInGx<_$)
ztvMM}keFR?vzODOGz`a9gYRR%62>vpPj3iXrDywK?!PL8s#jJ}@!`U>lMN1O)ULT2
zo}Mkc>G#2-)p<-pu60WKC-rR4Bj)2tc)%*jXe+ejY0dn?e7%i;yRlr62=i>y7+|(;
z&p8y!LlxFi&?mU5K_vY)0vUr4+F$61Pj9#O1Q`37**a@ACKPoQRMw4!ds67M=p4bX
zRqQtEY=G>mL<-p|Pq#g^ZCeTw{Jk-<sWVUC#c(v6enPw^2O}dov?=#v2tnRwbnPkX
za%dl~zUw~EGU_1QP!ib3YDE<$hj{xByQ`E$ucDtjF?F#9h#?Mr0ia*(f9=HEo{l)u
zCRv?8iI3R69@7mRBfj{27)^@{{%9+u_Fw)2@@t6}&o2nGY`^!Ef|DWC238U2p{3l7
zyIPULDV<gSwCQ|2<u?^GgTFFEiTQLaE}zMmGG|F=(&w2<-v0jhwxce1IGFi7tI>p#
zoY~2VUF$JLfDP;-mRDQr@MtFKF9fhhni7BG+LL5OPjASF_5nzK4wxr(nasoyx^}I5
zf7of<1@@_IcnJ|-!)G5}yNfqq9!XB_Pxs8u9^4TeFdIQ4-@;D0ooc`Ic{w(tK6~5L
zlqrn<gClOXl?hBBdmpBU?B_h$0iPELEVfHeJ?}zDXRDv7-QIUVJn*YYnuC&3y^KW8
zSM-|p4nZu)6nrd&NyP@&l{=hRo*Q$cU5YJP6k!PY)}kW(JGIVNJPzleR8|nXc3mQG
zF?7jwn0fHfm>LB{a}$)TLm<<!MA%l$B=QGE)Wo03*#ofx&fiq%%~1ku2S?^$rOAx7
z7!}?gh{F>GDv70ZSUvar{XM3r8b36UULOecBOTV;LiksEiY<OH^}27BF74q3jKsy4
z_CvzwO@7-w%xKb40E&Rhq_I|Z{@}hpC&j>RWMdY-DS}0~2$#~BX)-pm(`>L8@S(D?
zSf{bOea!Ao7C6c(q|)jN#vcasB^WF1C94%j`x@S5u$~H=Ai01wqFlD4B(HKgNu|(h
z-Je*3MG3bk+`z@e4a1H0)NKYN4M}zE7xeFey*QE*C@(RGjK*75B!F_EiX?br;w58$
zLH|8KPvJg=nh{vYsIi12-~}lik<<zJ-|>hrGGR-$gn+r~;Q-f8v(poCuM9Rbal9fl
zmcY+<?P(Er0a{nyq0_+;6y__;KFa$6ERR<KyFHEnDZD;3*LO56mU#Hv1BQ{m^N+9B
zkfxL6H||)VJ(rPl`{>CUl;B-c@6gC{xx7XIQL`_6k-+TS7lZ{Qf77g1ag)V|#rSvV
z!PF#;_G6lGJLnSAJ0nkW8g|v_5GD~O8^#r&F(pth^5mR;&s?N$+v=0nd2diF9!KPS
z4+4FQ(KD8WNvWc}b};l83G{X2ge=4k%KfwX^8V?aRhRwghDB&$cZ^WTe3d)NZ|V4H
zcAMHHYqq786(Gm^?)((k9~O6jGfY<+GFZnm;Gudo2Dc<LNoUtldwjpBYvvSuo#I<-
zb@jX1<m4om@X^~$HXgL1{qe%C#9MWd2I9&3He$$QYr^Qv-B(KM6~7tzVx4quZS5;t
z=vat@@Lq~QW;`g42)N}BTh>akLE`+`-^bq^@}EWINitKUW|@UIsJ=j~;}1hd<OI*C
zyQtExxWit!^9oOet-z%{l=4AK?COv!vW)T4xd=wO(+dT3EI|%*MV@K+^HF!MHQ>)V
zA{eIyGE6qqEcra;hXV6S0{>d{mJQ-Up2fUzqFWpTseu6PLQ9lmbrFRKO8c~IEb?8O
zcvsr)UYza1KjQGDsjXQM0lW)}8j0aS3^Q!6BhOq^$!UBj+OASV%Ruu`WjAIO8Yl<D
ze4k(<8+gTN(-lKUo7~SgQq@76Z(wFP?*m7`ClbA=gSC`f11sv#J?>&aEMM+w4OMv*
z6tv%=SxGbt(F(;VynXd28YYESfmuEer<AO?yn#~%(Ro?^BT;=W?%^_kgJ?!;UPffL
zhnIg39(19$+(rhv{C@p1J*&=eU$Q4*`bT^DE>u-NRXL|Nz5?+;r@dquAgT`|GBO$t
zFmXF&3ac>xF+29tUISIRp6B!Ex;S^8*M017s^%&e^&)=h{n!38>x;S=PyTMrtnb<!
zo*iXo?>N9dB9qzdUld!q5>878h6C}}-*`3D3(X5%`m5mGK6j+;TX(6FtF|b7c(~=$
zm!Lb8+wzB#^7&Yv7jr2q-0sBmwTr9ST_#NbZuAuV?IFWh^{w^f$%^1A_Rr1F*V|EG
zRvSci@ho?S`d8?P41os9WvKW^=NREFce;n~<KyvswJF-qb`49<`+8Q4u={rEb+Bp^
zBSHrn32&|4E_c5rnB>WnKYiD&;+Rp?`QeNcc+CRf(4N9lq?Jb@A&Aj-)~d=n2P_9v
zap#^t5Zob()7o)8^@)4vK&`R^QW07n%B|RJy@e_ZU?dS6)2@DCHwB?(Z^EB{kBHAm
zF<O?lH|Zx3r6q5~jL>P?)k>-8H)d6toG5Ek_X9`GGEY-MG(sI#SL}0@1r<_cN<u$d
z)~y7c+er61cM+~1rIt6Saz04PDxfB7fJb$KkKppH4HIGHgn!ay+{#UyMno&-V(_!7
z9SzWVgo7;<+d)b}T!hgJ2PT@ctuL9dkZV4IUuiO6^Md+)zhD!97PDMHyjroHs3*6T
zgVqCj?|m9o+tpfl?!7y4v)h&0VpG#lUeB`@gA`QyA9c?Fz&4q|+3tWAG$XE3QR6dQ
z<>lvbd5U<7Mz4fth~S}%XbNU_3&#j{w+lv?46z{3Qh&ls@qz^jf9H-7OaEqW@@(7h
z*w;n(P%>cR6qeHu8MGkrkVT%_kFoR61BvNLdBw3EMHQ(Qnt7WY!4i&)_k+o1<qj52
zmlkcdzrU#TC9NbEhxKL=m%Xbu3>?Sm%Z1*V0?R-mWG-aSpYJJ6<|J_LIEr7=qWgyl
ze9nMXQ7?<}d!8J{$vC88r7*qcRxm6(oG=>624#`DTY<Nl7~AIBpC<k70!mRH(69hV
zQ{N^*c57?xPEVV(sW`DghKCwn!nv#Ob{@x}DN52E$BBQOo|^kFwyws}me-@;h(+ZV
z8N$@sxJ4i$94O>6{?va&5kDfr;*IU2JU3^<6eD2#3D<v0JjA5l{5Rq8Nhc3|OfRw>
zg8&m<@6^8<sc+}wxm|LP(DDATYK?lf+&+E%g^OHpg~Z7`K0B&r+lH_6Ud${T2nceG
z=ru%4outufzQlW88qr`r92rL+7l<lgF8N{`A5<&;I#y%7OlE{(qA?d7h`>ZRpWiPg
zGL4>HQBw-au_)EryCqFe=@`fbS1=i4n&vOt_T?DmT^U`!q!&`YRtat(Af3C)hE0#{
zcB16~f0=vykW5*BmD-kW1hID;9YYjtoq)sHc&2CEb0N3ZY+ARLxo$?{Y}lhwk|W7X
zYfsnFuwG!`V1j#s^VS_(dy@e(z-b|BuVTb%Ypu$%<)&skqb!;om(81G@VncHv<0JR
zJfebepcwN+%^=mZmI{SOD--r#p)c=fzt3sZI)4BIQ*QzgxM{$JjO0P2q~>=z$#)Nj
z;Z)}_-UKu_=(c2bSQkQ2uGG_!eQVubx&Bb(J_tR{8Oc<dNTyeR<1QfwDtW4SS+r)p
z8jKP_iea7ah*~}MN@JkK&W6``P_mP<LE?d&@`?O^!#L|}hw%aem_ss!FL3M~K7ktB
z@qCL`Cq;r=r=Kab2g%>ng3@5=fat+@N&$$oT(`*I^>SV$9rt?_>_?XMr8&sqe#izO
ziDwX71tv`!D<#UMsznnEg&3cl08U&KY<~B|C#Lo{L$$Sc3>jy%je@<``E)g&?xR$C
z#*8e&F1nwP!JEqUrqit-E2lXh=W{SoVxWFMG%TVVw_LyzFcJDfH&e`Iy!X^SiuY0f
zjzIN??RGtW5Pn?8`eiN9_&eMG*LFKI|4nuq>woTkc3ClB0ry|ZBNE0Bx&S77n)me;
z8u<CiW?#G34H^cy)h?^%)0Z4QZqK`?D4hC(X}z!(xa)5ym~HlhdE>8-?dF3KDjf)h
zGnE}n^D^*52rNZDbB-8HC(;!^AH^FG$(S$K8V`p~?Kny<(>X=Z#yKbw59Co3*Zx?}
zvr22tBUB9ul$%#jqwGZ`Sqr-x(#e_1U|DtX2Q1I-jUy|MD9*5YD9C0^s900VHlaJA
za=_cwd3g6XF0Q*xcXGL1#8e7x=bZ%Hp>DdcT9WfZLVEn%o&HIvyC?%MeK|lNP3(LW
zt<hA0qgppQxtPFhw*!{G`>j=|%#pT{Lz^9wrAI;tSMz$NIi$IWq>ZL6g4X}N>!9X&
zZa_?=Z?=9`OLkT!PhY3nem`#<K&R$PrsJmnY@}OVPN+<7q-G|~jR??VELB^!*?1+9
zT?YTgr=WyiB@2)R%M?D-%b>@|S9vlt<~iZ_*)Xo%*SsJ)v~DL#eEiapltxP86|)-7
zjvDOBDK=EM{<S*E_omC>hgd8-gHBr(EOk$<T<s1}N|C?1FZXwQk}+Lp;-|)wSiYfx
z+-s))(2eT&P1^20S_d71U}J0H#0`UXlb_OrL~neX0h#XqjW+5Gh+5OrkNM&lTFeh&
ztdqqfFVC9jqY{5ms_n5{htVI3sMqGq`Krfpx4sbXbABJv6hmtcMx~)M)j-+Dxs`*e
zJK`OYRgm}xLKF!dTwoZ7hrXMP6_rnJB**{VI?-HXzaet91WhTtpj1uUb2+iw1C_~$
zq9ei^a|kI2UvVPZF7)q>^PKPD`Xle_+2v*<xzLY|e=YY(_FqZ!qqNidadbI{Fp=?{
zR7vf#eiaI9^9^4MD{4^11wz4vDKy8=mm7L>>5RBFNl=$HLu-X!Rta@d^n|uvOC#n`
zz`CiAgII%tFYbIHesVRLXdW-Iq8F9f*L$k9dEN1}w_i-A(t0961fRhSF<YfM6H!a|
zmOL=))>5&uvAis+8q?nN$&G2P;hSQ%NV3gvYPA$Svk409LrUpLKbdGjSKs!&J1c2b
zIwTnr+pNx=>K{$yZTEe+g33hUw52^kPC{IJpC!|Mbnt2r*iubtqLAYS#YAsJ&+}Oq
z!Gl08T#qI=X*-gEpP&|LvP<;eimHdf$-x$<Mn3Ox*>wFT8{skO;RxR+(YkTGSGN#x
zaq|_d6R~=3-7YtrTsOPewY@rFl0$LFyd;qgSYteG1y@$|O@xAK<Y@%5&BrglJf8(H
zz2WS6A9uCxAxZVbLU?}3gbPnRRMFX<FleK$`%P{M1^0&bh0SbGA;cz;Nt+A;q--T7
z>3}X@UzpG7VB$3K*kckyjnft~)>9XpTGP0>;5XG#+~7^Hy_=QT1JF$p%PMl2&8kY#
z)VADOvbqsr*!`C!Txp$bzEE%<rEa4sRPp3^KX)V*1>U`Cni4>D@7u1o(E&Ah+vD)o
zP56RNzF2~4fZT_woZfLtW=vMH=?@MedU^ugl$C%z_!a#N#m@u&wdzJIJ-!OJT&HPn
zsmOfBt=W)n=O=m5+6AkS=LfD!-<#u-%@P+8Pr84#%f9X3%8+a8MutEA=nd8N2HYg+
zdfGXy?*4b3gzx2Ock}(j&((VK`&j_AT9x|#OYlNdjNY2dqX?Gv(#pz^w+>#Y^;&~R
zA3Q^q-I9MC)1ZoaJfvX>ZykvUL;|%dc#f~(MH$E#nA_Obbz$(x0lwkMZ<eyNlCTp=
zww_Po!SwV@;xv=)wStl`lz$?)T^)c{<}fZ?D)%fdru-LGBg1wjM51+5Hk>+|YR<qX
zK3Tj?l1zX$Y_dZTc2jo+yl6jyuN<U~hD<gUDfvgCvSpYcmu@eEvj)<RCee6NgXMaS
z^18BeXJhB>aY;_sRtI5oqnoh0fARwebipqL5JedxFRSqZid*Uh6ru~1iW`payeO?-
z&5WR=ceDv4fvoNyWV*pwe$aBiOx0JY72*?vRezaG;f!pT+POE(5Hl!gffHy?)&WI>
z%lQdA_y_5me89)P%+2N2VgqpSE8zWAZ^h4lzh#2A$6^}*Pdv;AglG{55tqeOcgda5
zW|CDzB@L7UJaC-|9;2wQ<}iXBNRYYmD=ncWaJk=Lx`*Gf`*|1*4BYu%K?S;G&$PF_
z0`w{NTZNktFAEuI2WmO=ctA%{&Akd5b@ZhX<_^-X=-AhtpNj%h=yeE6@mfbAyza8x
z*|x?$em=u5R%)pI$_;)!o1k_dMepfSc)fEqzAfaJ)zq#;y?kLp;otL@@|aRk<tb&6
z;LA#G1|nL1>~#_Ib?+jv2?vYa2wQm>PqPrnt1ZQpXz4qMk#q`66Ezz;j0L?B88WG<
z$4SpRPX;AVN|V%Kg}_RJ+4&cv{YE^+<)rpknXB_~8H&<?$D~cNpnmSh@&}2to2513
zj>pURB1Vhez5x`sZl4>8Z_E!odZ9(BsZ%|lga44<&%w%)!?*mY9vX(uzulW602l3Z
z>bLRmX6Pk0yiF>WwoOs}HRjYyce|s0n)C9^32^EsNF+PY4LL?5@U7lUSi|x=c!c-A
zNrMW-aG3feJbklx%V-j65XE4-%JG4|=zi=1#nQ3XM2V9f!Uzq70wc#&M-EpYK9Qw8
zYBRtYh!=&YjS6zn(=IGC0ySGMa^{*OBUNRt(Y%3geIjDiBHGd32CG(0Jjd&IEeA`M
zL1V_e(A$_>h(F!D+_|0NHQOJXZ|x{A%_zrdv79q?<XD3leta)bGhQV);E`5Ez4&un
zek*)2c`E*VF-{4W#M)maA|B-&Qd@w?61!%cD@E60*tDHKfW-M`$_`Gb_-G@>OzZIh
zNY%_8x5{qIaj|z$)p4~6*w!RdCbMkvXqv4ZoB_V*F!%LWD4}OdIVkB1C0@XO5ZQ8n
zk}_ZWxvPC)tk;IC4;s25x!nk$GX#JwIEcEndTm)z@30I!O=AKk@d`AZ4h~JD0M=13
zH;4kInu}gS6bpG`?MqkE0xM3eLqiVVUIuj1SLur&I^B=5@{G}>UC?1eZDiU4iW~S9
zE+?RINaadT1Dm!#$8YQJO$|-ZKRfNLHyqzt7w_7)1Sf>Z|F%|+tB_Dea<RIT+YCF<
zP+?1<u3f<mn$CnFSzM9lFha>LiWm3r)O3!GqWGU4qR%=R2VA}a))2o*6}9d*Pno<J
zB<8PjkR@M<kx3$s>1HIErBi)<^56Fa9Z|T=^!Ob~c*=-}L|3@#JW_DHo=pnp4pm-6
z;x3bi&b;n^!+Oh0mTeJ-+lesye)8FsG(94_6s738{Q%z_`S<9tGyS$GIYxjm-5Ljg
z;HPRoAN!FRptf4Q+~Ihr_&ubX0dowHow<!o8~WQZ!nL3%y7T7i0y|MJN$nbHJ1Xt%
z*cIfju~vy3QsSITf%<YPrSG?nl(ss)1x4q;uEcQ8r~ewGbJJIvgs;=OR;V6Yg_EL1
z^@A4--&!zn1EF9)YdlqdD?VQ<AR<X_OxdQcjDnEdnSEg`Ap3>_mkyTaO0lK9R;G18
zCCX-sVv>3VAl_(E$YO<)K)=i2F)vOW^j22j6hk?!69*!mYDV-VCSQ<D*s9m;Sslmy
z?I@RTj})l0^wd}R76SzrCP~gT6#2p_LZ<_UVuC`~ijP$iOc>h#dQJUaU`Ef#%lYgY
zNwJ0G)@dS@nhUUeX(c%^{lyV4+WeRvHdw3ilkBXGER~z?Pe<ROEYOGstu1V5&`2&)
zx$G=#u&x!!tV5+eTOEzlJ7GOk>c!z!W01MWCgXn+ObbvgLJ0AaH)OIMA=^~<X;Cgp
zQz@x>jP52*S|v3S2rT1Y1P~KDELfyFcHkb6KI*6OABgcz%TQ_tS3_tV<#c;LkMXAG
zxNUl1J?0&WdWkj36(7pkwQ;6%$|-F5!wh&?p%a_2aC{L>?i$Q0ApXvq<U3(BUXjE`
z=%hRWqT?Izp3l$Mnltl)>CzXR(iD?v*HN~J!vpj_la;EpZHP(^0xe`gpY4r3o;HJ;
zgaZEKaWRqfY;;h<P!35&;~!M8V+3ZBK|d?E4rH`M`M$p_T~jOs9kNecA7FA`Os@GF
z1aYr!7^t)+oWD1-T;QwE7H0m}g2(=`t|i<i7J<%Cmg%5#jEB?NkMnAJ@Ar0oR=f^|
z4_TZk=XgIHrRNd#2l3rz`$U@G?^37}OKuFoM2+)d5>y}{@Ur*JgymzDE{)|4t14?C
z%NQ3IiJgOBpnk`FW87QohSVsAp;3$DW3^qGkoxR^bNZ?*g&r$X6B!p&3|V)bRlww`
zLy*})Ac)>6O-**mpi{g{MtfJ0F{TpOrZ4OGzNars7jLs(_uCz<qO8mWptBl)?{ePa
z_9@+ZeB7F53x$clns9%gi+e0LY&Iw4fH{==Y+IQpEV0~@&f>kLBshh@y{1grO0>;G
z?0~>BS3VYmA;_MXymT^lR|M>WyS^-pn9jcIxQrtldEDWAQ%t4&9@I?I)fhrja^Uua
zGZ9#!qZ+Kq2M^RYJb>TdiA?lm-N|o4nrUDL&EI_n`YJCl^Mn$<sPtWu%wX6tnfQ-T
z!TwcXiAJHvB$;1dvEy_!8)$@cCb5G<%Z7}@r%o5C!lXij7ovvcvkzmh=L{zQMLUeR
zG_;i#_K6tcMOuydri9lPE4&}170x5a$4UVVV$%$O`8p3{?{RgTbQlT0O1ui?o)rEn
z0kgfseY;{rHL4%l1yB4ef1jWSf1vSXgxWoBEkP&Ras&g~BePP7n4L0$DXw62%cd2~
z2mtMVSYxsHMw!>(+g&FV30>gU86Vg2n|LJRBi>9LqGKFGCJ!8pB0<g)hps^)1Ux#4
zV$`k)fvk}xxXudG*I@8?{7I_!eEbK^+z_XHL5?a{4q%zbZx-ammn*aAe%<iO^}ex)
z;H)-zPh61r$D>{W70u0E_74vL&6O};0T-Gbg{_8TbUZ{(*GYl#bhFI`x7r8AlEQ&{
zy+&KOu)U#<pT?xh)A0_2Y$K_2l$dsyE-hx)f9o3zDn)~bJtZ`P2?RxEq#IIKe=zO0
zfUI>eDfJ5BIH;ec*l6*ymjV7!>SJ+BjzKEZ7}i`X2Zz7D-FO9ZuNIgMi43FEhZFP(
z?l>Z|iZ=Cvm_e5)T$KDK<=WbjVsg1h(+lFT@5_~>l~GZf4$7+{ApHiOnK(oN1>T7V
zi;z26QEr+5oHcPD_}K4jHCm?sUN$JX8-dhnZ#l$<;tc5-wrXr(-vq!Gh{~>2p%4~W
z@K!Pm3XkCj1JpW7nr9n6(1bpF@rOe(fegh=DO8^DSU2BD9F2p7cAF^}KqBK#B~F~T
z(PX?#UQEEDfSBH2P4Ge^j@mJoJnw(gfFzsCO+Uu9Abe<6#IiSU60J=B^~M3tedYvj
zol1|P5FY2uaXZG@6qPwSi@2x}|6=9*FQ(2|Jb>hOFmN$U@b0*=@XpB&cxtEq&g(ht
zs}wk0WudfnZ34F-{#+Kgoa$nQ_61LMd^NU$er|%luK&vMcGS4}+o0;FItn%AXqAu$
zHj&01ryBP$-8c4t<#1$qfhSK6cY?>DD0h2zfadMZM^^4E;S{-rgPdmI@Flh%>Y=#W
zqi15+!D~@%`FZt!S^(=$B}`Oj_MI@OyabSMA27my_9PM+T5C1|%#RqsI|x0uoU|I(
z+M7sHzlm+{2a4FQ(`Nxw2odIw!*jQP3XxCe<hZ8)m{Y=~fBh2qETkm$b#kHyxPtq9
zzo_ST6^gv@TjTYlT#atk$8>_3My*m&W&Qq*QY_Zjp>LpHEH>fe8}A98dGfu5@c@UB
zutXD9Iq=FcXfT?h2Pv?G3pF>8GeM)OrQac$7{?!2`Zrp=<f@rHv*z2U7i?DdUq&(U
ze;!^_)SkD&;)JA}lGEm;|7Gu`CzrOfpc`!48MrPEAl6{+X*W{O(ff;_G#}N%7xPAy
z5wGULGF_qZr#{{c<V|b7!1AmQ`|}O{S1vaGR!U+!SbvI^cuT3G3DX}`qND;j*<gS{
z3NEL<D4-i9N{u>4)oDGB>q6OHl8}jjz1`2UndQ#U>wQlQdP`kRVI6HJt-Vobtz_y)
zHTG?{4i&Or=R<5Q*3MzEP%N(UoheF%EhBz5Z;^hG3yKfdb^)7N*}|^Kvp;n9huE{(
z)4*Fqiy!J)s^e^wD>f^x>Z)}}1l+h22N*>WGoDY{AR%gs%giP|Od4N7NLk^Vb_-2#
zlh0F5ko}hNi2%#9Y%{J1J84pdut{rA{ws$l>=-D;mK&8ohy@t_^k(~IkDP6b!%I|p
z0Ioyj@j!OqB}cp6=I$h=u|RckaSG-RHZu?@{wK$j>FeLg$z23lGDkEhVPB{{_(SwZ
z5GODCYE=9Ty?M5Ss=OPSgW7X<Sl%)~<+9RP&E8d;PNW9&yP|D&BRvy1d2_+@8Wh@B
z^X=&I=zhayemt<6DzT#}I*Ymk`l^jIMzr5~4?28A!27_r!d%xk1s(Cs7Bo%8%dCp*
zJYG%5J>5rzRTKfY%W~ebNIF;HN0}OMBd0{49>o8RhI!ki-}ABa@$)+N^WAg--J<$G
zIL_xkIIe4b_Rh0wABWM|=m{Ko!*81B@^d>d98GuP@BF6yxeL)E=*Bc@j2U#@Qq{CL
zUolCd)2KS);DdMSND9o}S7GIw3|%Bur!gAUfDn;A3Hn3j)tK0-HyE$#q_PV0decc>
zXF!@}JR7~FtO9|T7<CTQbZD(RX$5syz$}ZJW)`lH#hhqy4_*SI4HaXhA+EqCt;H@Q
z!>)p;$iQa8IaWdMv6Li&<51*?kNrns7P%Ekhu8i$r-~YDfhfE5dV>raqkG_9CGlB*
z>#4sD>F=*eH)PIHoNnBaWO*{<fNDpj+7Qd&w|3?omQl&BDSsE7apyuD{giNTBmYjT
zlO+2fQz}a*X9XT{i#V$tH(g_NK!`#jMQe+wC&|y04!BXqR9oQyG4g|mcl1d*8wMGl
zbVRaQ72^QE%*MbfjqtC}L<GK&J00=ak=Z?(YQLr|0Uo!~F6+nresM8oT5$QrvyN`Q
z)%6Z}=hTq{a4R@$BLTh`gXTNYSH&7M&b=TEu^;pT2(rZL-`WZ)Z&0U#4PY8;D&u3p
z1w>}$T6&Q4&mkHc#WAnZd*ItFEJ9e_E;<@db^C9%wRGD?UJkfXR&OagIi4#we8hM^
z+}%q@z0$xY>#n&Xm<vh0W@O^x(%GbE5{k}6OUlM4IprBhI|mgABzn9jQw-)=GfuVG
z>?Q%K_Q$W`*<UgRha+Gc^AI~>ujz~j<E}jLRP=nyD6n6^IT-T3Vn3@HdNK@Fg&XR;
z1W#o)GXjksH~`()=oXTB8ZsL6H9#ouoM_r*3srj6;}e<i?e=!Md%sc4v+Y0h&GsLH
z;@5nk_G4wv5UT6tgoZpfm)-0Fn$X*R?`j(Wwdd8e%WxJR@Y^Xmu6A;GUPse{f>68G
zO!BC&L5MY)g<%k~%+!Lhau;?2_ECtdq*|AYSlEVz#&V&9TFY^74s#)zzxfZhKb&Q=
zI52_s{p#(9V2kP5Rlu~R<_I~nU5D7`aBj^bgJ{(syBmTF7f*JEV~Vto7V+>8Adg%s
ztdaYZK`^9Zn3i<)Fsnv3)d$;V$Xo{!$ObW+YDKTz;6Bp>+z-}Lb@7?7jpt9J%3MTW
zpBXWyEN~Cw_VD+2jeo=bo<izE_{}_^iI$CiQ_n-+kyB3mucoDI!kAn1sl^Utse9I5
zq9e{3=ntk(P;xrwWdiePuvime8iZYRX(WmkxJR^&dbiD9Rx3Rt9nXiGG}R<~w|&WX
z<>gr*;mj|<eG3P4eCS&p-&t6qBctk6B~mP3M<_Xz`{z$c`0DP16|c+I;lej{Lv%4^
z`#%GtO2yLaoo2b_Nahj+C^VL1kmIi7=pZdYJ>b-fN#iY5!jytt(mj|xeS2TTs|~eB
zDR-E3kw8F%>}k9kPI9*UMFEBYdeakkHLDR`<Cd;TAFU1O#W?oY<Z`^vT2S+miDS|_
zh9y)rlS3b>0&Nt0f(549z2Bsq$OCKpC%1iXKynY(F8O~txA2w#)^Mn#Nte@^?O$km
zS$9XIvHVc6cT24mKxY$skceT6UauT+%_pN+4X3a>)Gpsk#SpyH%@5iriL!FfBFBTP
z^9n=PA&~Pjvz=UVnKPT9Me9rLdj$eOFtkIGCXMTNLe`o{OO6WJ4mn1=4d~=P78#s<
zo`(6dxjk+g)QbxP0nRl$V9xrVZi%Ix_}E_d66fh%tK0ANcQb(HMMh3_14Is=-v`+0
zb<^dVh<8obE*1-T?brUFxw>03*2SVi@_*-C;E|q~e#Ug~ON(O?b$pj@mOpajI9v$8
z3B-mcp6{13z$C9y$1%PIc(AkiU(o$Z;Gx)f=3G?;I1r_x1k_=}`M*vBBk=CTMfV)z
zEL{5VZwL~x$9H2=%-g=9cslNvKTqpN|74O8R?Rso!SuA(y-;IS@?(9=*EPFycf7L`
zIot*l6HDx8S{rj-^Oqu@(K3@A4N#g8@XX(>p~Zs_S&ds4bizSM@4uBFyzeO~<*Aa*
z=Z2W5kVSgLJN(+?0&*%4;4s#p?2;(aZ8@y3Hpvx7X%C1oe7qjo_WC0G1hu8Dq-~KP
z`rAQCSqHNAmEu0W?6}19Mf1zJlg7<NHh6T~!!_8i$uXhjbExSuKe;rxUn}}H+voB6
z$72FxzHl{m6tmj_`cAUZ^FGX8_t}@IGqw)2zYamz*W>6W+dp8vIAV34x<QJF9fytk
zqqu@Jlhu}LY902BNKGuDjM1Go8ga5fzoY&?uFkS8t1jBYba!_O58d4<APv&p^+-!d
zcOxO)C?MV44bmXp-Q9h*@43#G^9w{+d#yR=xW|~9n)v&>ujRW0^oPF9o*Ku?wf8ae
zt5seOu=GAMJQHHk7hW09o53N$ywHON&{put862d+T3XRO`WPrjY|o_Rh|Ap9L{-xS
zKBoTp%GtIYPOmxoe!c!ttP!6@CiZY??}Fri;81NBO6$WKQnT-epyh6h!$lA6V^J`?
zlnp3ye6;YwH(01#|2|<<>qctk*cC1wjg_fjza*{%|3m^GvL)aWX>{^*o%7yvXjyb0
zNlR)V0DXw!?e!lYyry4d^;9v@RhMD}RIj33U@5~{tMSZw^WlWpai-~7<N^;;kPmx*
zl!*Wf+SlPB;Z%yGWE6`vD$S1xMF;^{h;~y*hxaO}R6R(6C$&;PXp-q$Y?dnSyC(BE
zt;Q|mK)ucU5Y|hpME&`QFHcpNre-pkVvO@CBZa3Q90hB^&OJv+IDD4~ZfNWpF%jKy
z6`3ZU!TNiSuCusBCW4_B=gWS^Glk`wN_}Lq&&dHOrdCZ5I+eB$ZXE++CtNGI=fgHF
zT69N+ibqWnMW|PMEhez~SUWVu5p(iN7qNo%3x1!HfLYOCP`P|z(+HD&Lr~j+tOIN<
z@n7pZ^#vUVyYr^HcYmP&M$nT*;np9bPd5^ag@`X`tzYTG(()T?KpMR1LmdMS3LlK@
zhqx;u0>jJ>vE1|puy&`@(T`#vx<9fo2sO)kp9j3)vBtl@_H}<kx;`9=hHvQncesFs
zlz)kM1`?&z2&Jn+-!WYtKY2};q+#GJn*l$k2TdDc6Q#*eD`$HzC?5%S<cx>*JDJ13
zN*d5S#C;8WE5Eb9G#p3#LJxJEUvN8aAk{MRoaOhK;QC??XK`Hur8y*V+5{&(`vrZ;
z_{EK?SxPhjK~e!Kd@?r!*Szv?Zuopm9oDTXol`~%M>Ydm(mCI28HH9TmxFh~2J(;E
zd@8;AdMm%n>(=MHllHIAXx$y-ZmkwFan*@*yw0nMIgoq0jS=x^fj<myqEh@!5tW1$
z!!l?gK7r76-5#n;L<NpB?_y(zP_h!@v~iHB;g_LB;=zWePpf+5@3)oo!;kJhrzx?b
z03-M3Uq9hzK{TV(vCruUY=Z|_%|RfZcEz<<8IM)BN0uktFIg^vtM>-k06asKT;8)$
z0g<I!0h*I6<LiMiBZh3W*9#9<lN_N%pnzA`=6CKfuMPjKhe7-0d`O^A(!O=dQL#5#
z<D$ozVQnFF$xTQJ_i6T$Q8!PYxpE?w%wes?2Ov$oJT*WP^#J@(4XiWFYP}muS{A5@
z!eh?nb@+$0UA?V(-A|-*dtK%i9KamO6zl+@4-FOC&9-_410FIwoj&_#KV`l*x?Aol
z?Oer@wl>-?S7antA?^GOs5#`TFAJrI=AVnAyibcBFaND0Nyg`qK@U$r3lowq9)!Xm
zC^N-yh1s;90FsakL=R`jh|yP!H?%ixHstu~!h@b*lkXFVdV<@swWVr`k`F=p(^<_v
zCeJ|dk1Es@hjk3KwdG<(v_LVbBE|Pl1k0vH`nf`w8PMm28WK*+N(sM`SM+~8-s8p!
zvuIaU0T)6i5QZP&-k#xV3p_@|JUni$dnA<@hbI`qGI#`;3XALr8)5zJRC<_L9}cHe
z{g5}TiX?I(T$eUvBTsmot(qC%D+K+}5>6@6r8MQ&BGzN!XFnDu%2qIu3$I1rc*O}W
z7zaAG_n3Vbasy_-;~9cR70D`=e^Gj4TQ&(22s0<~4B4v8Hs=1mA@|&uaB!#BcQb*O
z233*ib=6(uL1WMX6#)ZA4d8wN)qYyCLv`A=N^KlvV6b(*AaRhT5|Vt9{Bg*5R9i;d
zKk(DYI@0o??~-iD5!3UTDp1&AEl>^{G;#v+cy7kPk5eAB&}>5%GrlT8OSo26Pq;pR
zlwfafbA?N=qTp^$+y@CiL1zVsz8)%%Bs2M3kMX`gHFmZ?wl-RC!X%hCpp)X4rjc}G
zzDi6vw|qhW{Z}@W(aJ#=jiATc6KR?rj{?!VTuzBc7OuBMDZdcPnL*{vBP1W%gbZm6
zQ}!9~vOJ%+p92%%@7x*$1^LVXQbg?gFCSPtEtxSBc4c|)RB(SxugVe?*lfB2IMN0y
zWnQ1g1<(F%oZ++KsU&j{aI&mi1Iq6y*L;mx#nn3384e(qik=l9`2gDjt@W@9PRQ`=
zd6N&@O(U(6dAa2vGw>7GEHbq;_`O3eeOJjiejG_*nE+-GfJ6l^=Yk~!{jHPjrP~DH
zuRu8jsHU6*nPjHVhMRzTnN?iscX|tFsjBvsx{~mC-huCav=%URcue=vw|~27IxT9v
zdMHxPxg2TrW4GDDapKoc6?Fd>79LAhDuFL`0Y+$>#VV>N2`+>NMXo5o5e+H&M@HWV
zVTkF@EBrDMGz{nqjwi88bxX=6(VkNVOwyWdbmAVKzw=2TD7xr?zxX9tSf*xA6|nV1
zWHSU&wAIHkhQProatRd2zQfkNsVIrSUX_671BVnZQYe;nQ3zD+$!ds>6+7YNyc6`@
z-dB!|V93a)!fiwMv3lUY2ponGM$mb6JSUbbf-`n<kDER#xR>)(A}Nf)TzUEn>jUh8
z1kHDbtKUlTqe-Qj-#x!6sLw>G*NR?#D&^u9ABfG7)1Xe)#tr-c+u3xC5Bw?E?e@xs
z2r7l%WFNi}&XWpXJ6|&?bk_`gaPo)7`$b7zJRrJS^D2>|MK^afx1~j5hx+H%1t}DB
zPn2B1+!`3%0G5;8Kk?9JmsQj5keG+|m38_#CKhhdM*enc;FWDq;_yy5WS*&#!Y7!h
z=db@d*@yIKUo{c#ln9FUE)zdK)QDLqa*jXKS5eZvZd|K4!gf;5Z=LTi`{=F)Nrr9$
z(4r@VvWjQqcdyEoB9?G)>&|w-KijDMHg|XNxTArr+TJzS|3n;RwlJ~Pde#IHm8|{4
ze;n=$A9EJ&WpTqQMRs}LvXtjiJ`o4>mkCDa1(F2N&>Z!Q0LxOD_0ujiE09D`6M9Q5
zEvheTOpVWFS!*@nWAcIwn8K5n&lc)ifm=qjbs`OE>bqnW(T&X-V9Xf#JNtfhW)!#k
zeZ^t#xZ{n8NfAiXhg8Swla-d{L<5H3{T4wL%|}NNdE=3|m1yKVC(YH{JDh-Y2^7rU
zPsK&^6(L9!Y(&7#%&L_7Kova4pzYakzHmD&&pMID$vZg=oWmTFSG{;SYaJTYTFhRV
z6##hAYdnoR?p&*St9gU2XnDE%x0$K)U`wZV?>7b+r%Vh<T5&1)+Xmb6z*YS4uy7~e
zMZz0!ZgxjBm(b#K7AZuRo9)~Ifr~uHi=}j`>W?c=Wrbn8+vtqh-h>`qPBnFggT2kR
z2NT48z4XOz%UcUx(95zR3=<U_+*(~+9l7q*9!^(hHz^90k4IE^EMu$^eT~)Cl@v@M
z72^zg9<97c5rkBY)w!?r>g{2x)x=r}<HK|kRI%y|jPvVIRp+NqCOGM_;mft2^qLFt
zT10C${Zb7kye*EiA<T}kEn14P2$@mb$c61fr@*-)Cgz1Sk*Ba-t#dnF6x?~ERHs#r
zKl~}>GKkpG)R*TNkais*iB|TlNPu`}4n{J-dgyTnF9JVUdC*H&cNzG!tL^R$t;v66
z`3L+lW^$2bR)`vsKNr*V`#Ro?h0D`4!zjSs-KJvmTR!v*fZ)D!XXQ!-snK2oPpJLd
zup+?Yug}Hz_nu3Qd>ToCfM%tAOf#r99oW&75+_G7g=fGpCvq!xy!neYm@GI<&rmG^
zg@bYL1t>^HZPqruPahwCg@;sd7S*}$%~kjq_#VY|h)oEAKKELsTYq+W3f&GEQYXWj
z;K4>EqJ6R0@rofeltF6@rMz(#1@<U=*`K)RHGUa<j<c2$dnpW1{$WjYF)C)z%NQ(o
zcp#{aAXq{&|8DlF4zKp<lRGdR_xX?4?^s*BX?e>cd^jqte0xdxj%47#qF?24)kn`<
zeyV^qZlpQzu&eQFKSE!p%+x1Xts3dxKl4}LA<`}b91<J~v9?83m*S_3fu1ENpBYzh
zSvwU9G%Z|83Y5Mj&97Ln?|rt9>E`sx^xVV1-Yx1e^sf$JLFc*ORR4F8t5i#EUq~k}
zONvgQ-UQ|d!2V1O+n*sJr{4tKPCE_)ky$!jClS?haf=Mst9-A1(F1X-*j`Ole(&7?
zs@GWemBP}FwjZoGh`Sak25OVTIi6zoR5c2#CSq63hx+@AUYqtDF(sm})m5?|2KbZ)
zGxXffSN%37G40iA@?HL{G+!HLvN`RKBi;8UmoHZkx%T_ZKVb`N$5TCCs6_j?F|~rk
z6ydW$+J~VQMgW5I@BVW?6Y>-S`0{`5i!V#A4a@y`OL>PRM!3h8Q<CNZHxnfaxz-4s
z==`ZwD_l&|{9*LtEh+)By3G2Q`Bu4(s|cjvW{F&|U`GjYt>yS$>r9&23p^df=41YD
z{mL)5PPoRBta*bE5suli8@w#vDQm#Z8omh5jMLN?!fFB-=j6>25^3RB3D$R(W9b>i
zrJYSiq9bmt%)F65vX7*z``8(g_dc&WD?{!mkZ4`z8(ADW%>7BIyid+>7zS>kZ-41L
ztt8S`P-#93Me)ODZw5DPVO;t_8YRB)VBhZcFjgx3doDc}bNzy}z9;68CKN^~&L*Sy
zIvGyuGD=%`7MToiBs#%1Tw`xtE96kVgqkaDGvix%M0K;7gOBIF0N01_zERwWT`VyS
zLiisP*Y7VJkpcz31&!6oIMJkH0|*JRI!JHun1f)giLOh;p$=gm76T40l9_efE@AOw
zecZ~t-3mP$e>UN!wXhl|y6^okoj(_Ode{c!X@*_hXLn7oO$K1*RR6D8D_%}QyCY5w
zh`qZ5eT@C~j_&6H0Ax6dMQtkHI@R9y$k|#L^jQ)Wcv2m?bQ06WV7((+zL}i~q2*aq
z_9Cb-)CkrI)^t2I+g)@w5^;9tP{!-u?&quTz?h?<Az;_y=x>Veb)N&^1K{)^US9xa
zqvwf-I-yq7m!oW_)e|CKKz}3hc3=is3Vv$~aoHbVu04mSRUQFsO<A5$H3Ks;hrdE;
zcoBb4x{>axV9fH>5fEha{ehhWj4>!+4!sO-Rw!IU2krNlrb~aqFtfJ2Sq4f{hCtnb
zH4pWLf~|vb1FvKzgS(|;+`}k-X)o)CaJ$M5%fsnierf2nYz!E9ba6)(b*-EcyGe$k
zjkbM!tMSyYx4VD`Ut0wR2={3R@`HAEdY8(jJgDgCsNvG|bFJOkym9P%+$y(wXw9)~
zR-+3K5|<Mejj)|lXWr~8(HTkps>apwd||sD*^go5JlgrIG8tTl&#tV&+2xEPPOID|
z&{Oo>=-hC5yqyE*5|PN!h7aGmn}`Lgl~6r4yR=lR@Utr;H-N>>)AusP{dMbTq%u=f
z6k!Iz2=kmW%&cEf<?kqc$!U`cII<@ao2xxCMbWkYQrR-dO5m8^Wl#9(;wJ}svp#=C
zS(@=Q2b~k$MQn^6)m>0^*(gPv)>o)!S1W=DTPN7%OQ4DQ)@~v!b2|l?>?UG-?B~iB
zA*X@cfJTj%!>-q6A>dDZM(8dMf^^bFC<<5K=}#C-`@meM{Bbwt26v^YyAxrylWr4$
z{!zNYLsEwy5c^oa>4E8FR%p&!u7;73gtk-Z(!`i>dPB|m4W_a~i!N0)Hv*AWfe%n)
zH__3eyi;rWVO=a_9!M3I|M@8N96Bs!OG|ZV$Ev^+tZ#3pjX+A?`Fa-==#(QXz!gp@
z<zI0!v#a!L`$04<0D#d<H~Ld3JRkby;HDQM8;<GI;aVJYc9GK-NdAQCGdm?|K)8Q3
zWla=4@|3=Qp$h4kQsFHJ8^-izr6*J8!#IPS$!^QiYR^f$2WuyyznqWWQb*JjCl(1*
zOl%v~m={|v@M~Bl<wC=wq53&K0JaZ$bzYY$;qf~RZEH`6{LB?RCuUUd?R?)R6%EJ?
zg7Q~$JDK40-q$E$u(zAck;I(~(Bj#z9s}PsFv&uIDU}B08!svCq$_;hfU;4aH)rem
zRvG*&vLco{E0x!w`$sXW2|B&jwMMkp+2fiphx__h-m41AyYU*mHRXy{N?=m=^^g3o
zKtrf7-o`pbZ>Q-tPh_YC7!hR2C_8<tTGuK9LuX3l5l$+R;dHAqRppXZG~)ag9hzDF
z%!=lM(+;n*)2|=VH?^?Bu#jZa9T~}n+K`Z=W|;CaNI7~Sb1J&usdDNVwn-xhC|~ST
zAYV>i5(JBN_Xb5a<HzZlT%_H69}i6tTx;$bylr=&<r5Br9NOEpzLNpb4PpxQnKCoe
zOJsYrAHvKOMIq4~)D|#q=KTZbjqQddS8Kj50}Zi6UqG2SpB%KRRI^+zSO8N3-JBoV
zD67kTtSw2i?E@bOWw5`UpPJ%~X65h7QlnKR@U$Jx+Cf9_DM|;8>+D<%!D5(1{dw)u
z1W7e6$W3h@fj9F-o_~=9HX`6EIC0frG_=Q`ph64wP+9PW+VBIQy7B1D#0j9t)+kG;
z5xj)dOH%kXBliC?3)<IDG}+v_2}l=vSgRIhU9nv`@^BgHc!dw2>~qaCu>@)fVD%{_
z!a{PT_9NiN%OQRy@mU{nj>z^CrU{!!6V@8{H}+35;rDyk==FO&HhKq*<lX%GPiKnx
z54-(2li}w3G~NZcVBP_#Q(@Vl+Pgm#l{lo5h=`}|Z3HVElM%WxBOvrDjWIcCbbVya
zU8+jXW#~34Ajy6upOAd}&+Z-|WBNJw#-yMUbYbXZP+7)mdAv0{0Y|F!?4kJ>7pR3r
zz8;N$l~|2BOm|DP&`Cv-!JMF~j~b|Pw7o`Y3__0WSNWif%5f^oMLF=m3UL~*hQwIC
zRC(Hq+X8g2W|@ZunN@Byoy})&Y86zr{!iADxq%D8x7p#1_}t|xC++Q=_+lCE=GV%x
zZ>OwhEv*4i7M}HeA#N9wm}p5>E}k<`p6Ki8H1f%!T;6?uQw;H}<mk*7H@=5)!Ym0W
z-8+16q-z^qx`ObYto@Q`s%gy6*r?{Z-H@~9yqhbIs6XiJ;|8Vy5^w<fAOFWsU5RI4
zOjXLf3y0w1P4U|9Q}ifc$x}7@2*ZsA*>e{k>P4<Qzb~auB-3Q1GZoeJVMe;1?w?|s
zMImciU}lyIyy`nWzF^QrIm(UZNTEfLDEVL9of(u2f%q&TuC2W6TdOuUUx^)6yqdr>
zSDbs1X0*~yTYnhJ=wOT1o~jR4!A9H41V%8VTX$u6aJ?vH;HgVLC(6^T=Fr<1vuv`N
z>AAca1M|Cp@@04XrMl(;KB2fzYJBN}R`peUR^)WDR}2=4gK$Uuo3h!nU^4oH%^fQz
z{(!)e4BzSl{Fah89&~d4InWHJx>l#Be6Sd`*ir%&(Z27XCwm`3xUfiSfO3F6mDbN2
z-w`+L@JU4djgOWm=L^+dXDgn^yj$*00PyQ>iMg9C^5rAYo3)fy7T|zhXdjsyz0Q4=
zP6j3O0~;tHJ|B+KhnUXGkV889KC$FY#}{zMjd#)iJC!+G!3DFOwf_yxzUVcewV!tY
zV~qFq*YVbh@8NM(Tj9sJT|CHf&O7CT_0Rvu0zB~hit|7-ZDHfBJ5UUgOt88J@UOV>
z*JJ5?vDhj4J&b3=cRyX1m3NPR3Q&zS@+pMLWBjuX`{aUanZ>1|?=t(O^e(22^;SR^
zCIbv5ZGv2kZdMF_2RFE#7eP31oNCs@*y@Tav>pg$s9%-rVNS$YVxTu8s*;RSR+q6u
zHzTt+{bCnnaeQD38jVGa<^AS61**ABXbkkhOL`BK(kcNmX?b1k)Zr>mpM0xFsGOZw
zn5c9?ro{5ROp3g?V=7#d#^U`U!N?lohbzI<?^i_s#=VpuyhDm;D`@!~?5eH_qlpk1
z`JBsa_M#Vo5|uTl$LW<_Ip$yKw{_OaznR8<vDhZyIu!TMS+`qae-4eaCxL0)%-ams
zc1tCZt)NI*w2tr%t$l1m`pkhH$`28`fOJU<*u#Rtz7N^lIp;X1NjTz)>OESUFS0Qn
z9OEcj4HCkspdG;x)fjsjTU}XZo{$8$$W08pF>nn}5-+qVZGOCfNbS|nHc`8l0Oy3K
z|5SAU5ZyO~^wDtMu9A1fXW!r-)y8i5716X7$hhP!Tc^wSj{o8ba@h`Jr0a7E@%W0e
z(#4jAZ<Ojdo6D$8#I_RD2H!k)6L>UA|09<enJr!q_D^nqsI37zKQ-;-C$O3^AxWa_
zXBKFTzNr#eX=q*B9kX2Lw)>R(&*iLI#wU&Y_;%g}KbAOV`8|3r>ezlzoVhtBl8k;8
zGLFAUQ?oYMN@gtZ16dM~ia3J-qBj$^7g0ad0+{iU+zCa>TcSk0%;JSf1sGv|aKMa;
z$W?ZOA<Q^{`2;58qxP7X+EzT-)d#Lgz=!dW;{HBO84v{g_P|SR%n)4LtUnGu;e=DO
z9h@6~t?qmmhRyV|n$J4ynjws{1jcd&!?IE9$Ak|+bcr^#?O}q&#yEYDF?VlZ-)Eho
zlyZ(Vq*Sf2f~3+YA&m)m(c}m1nzaPkTAXK07=JUp3{i*8f8+yjCvri<B^~1nlMl}}
zuDH&82M}EVhgb;?pKKEoasgx`U^Q8$pt~W-))HjW=WwTV0Isye-A2U8wSv=d48pFW
zA4`zifT*|*h?HT9?R^Nfv1=(f_^321IDFE0!a!p48^f&bLUyR8-2V;k;}CKO1hjq=
znx{a*la2Pe$8&jOL;Y6-F-t(>Vq{=X)XOxCeF%jv*XkVlb+bd^D31Q|hjSdwJPKf6
zZKzH(z%`A_OeX`H?5|B-18U6N!++2BT+nW?7jVkXfnF=45xfjIW^#MF<hwF~cxPu1
zN$*r(U&HyT`$hut^Q!l@BEU&P-}ckmjLh!HY|u^~MwmyUQNHhwWp%!Hi9<~um7ujE
zQkiJ?(-xS^s8@yw;BU(`6#d!<-Hn8Rvr{+1YqJ*jl*YaV^qVduJ-9W$i}6=dmM`2Q
z`R3)y&<MdFjbywu6dytKcyukD5*MzfssElW=BI!7^$Tsm=RRVsr}-(8QPvk+BUB}}
z=vLasE9E0(JSlr$i&M#yg+(FTU;=P<P@U7v;b=HY8hhzxds|v^8IxdqFRbyA>{kCY
zMeo?um*7xu1Rm{01%nRn{<!DR(RA*N8`@S>P8~>=NlMV{^|cD^hrvNQ6=r*EKK%K{
z19$wU;3nEkR<X>Zf=9^U4{-KfS#joYRRoAWHR~TLw*<L?4;_F;w;5g|0vd3l;Fy1(
zcwp6?Zt0&OeIMP&A}w*8!rL)V1Y~Ps!273Vm_6XB#?pjBIeJ-&i5>$qz}utJE-O$&
zSqA_$)NVb6wTLU7Ngu*i2K`0~U-8~nvJJ9K=@{y-;1zwi_zqIeele_^nvRmU+VB#A
zE*^dr?(9C?^?ALW&jF&N1kXECv~FRM@y-Mt{U7Ue`bKVB<9!ugndj7kXYSf1@Tcfg
zP{53JhC4gjZ!N$BbF=XC`>OA8`51G#eV**peBd^o<wQ1s@x31^@|K2L$d$eT4yHFD
zPKye@`UgwkdG93SF9Dd=e;wOSmOkxI-X?96C$8SQcGGz3{Q_&Y8aac6ks1Y<{MM9_
zfb7Q}YWQC6h#Kw)<*4M_>Q@)gkC)`SH?xw4U=v5t*yVNzRDA@8h6Z;-XOqs7WQfEu
z+Gm6&$rY*Pv0aUjHr(W0U6zgTg2{H|4*kgyvv!<#JOX~%0RCtphV#0T`1Ytj3ZslX
z+epB>3w-`{4OY9O`@jRdpXG9<<MSl(^gC0)UUwX*+E4GduVsWqkkDw=%A<}o*S^bo
zY?bI%c)z#b1EX#kAwK;z`jm(;vY)0Yah9nHl@#-GETAG|sfJy1m<N9t!F23AapuM`
zdjDumAOBri3J3R=0{Dnsw234W0mN5|s1A}ul$+$L#WZ!kH#r3d`z{eV>)(V8FIb6o
zN;x9i-IVCywVIgE1Q>UtrAt7^nB!|ZnqFTyToUPA8hizeWh&c`(wkl4ZjYL10xyG~
z0wk-VK72wG`P4v1p){OQrq{sSUJ5M?9g5oga8SiRlLxw=*umG7Rg3z$HQb3mo$mH8
zun07BeWzz{XKRny`*Ux|2*Ybf!b)jVArgilsv_ts+eT}^@<{nC_s_%RupH3JAH<1r
zAvR!hcgxGuEM+F<XtytrVjvC;$70mMk$|gz7MFKbp>-j&&O+S+jq8ypBv?JYt{=8S
zPOw%xA@pxBa+<6MZNOf|cxCE-vM|Pe0i+xSXg(Zh^Y}Z9wbw$xObY8fHN`a>ia(`}
z(HnGn9Xt3ve=wN;ugY+D`bQhIddyF=YyEGRw!YS`98GfD*y%HUyLV2P3SDD_eq!B`
zFD~#YHtVR*qxYNf&F=N9gcijN7LK*rXhWV!@2`l&^@Te66s?F-LZYm6le?pM;2{Qt
zHtDw}o-FGxcE{mAJ}GIxa#FkibVUZw*Ztb+>t$#CPVJM1f^PemGJ~>2Isr^=wutV}
zs4bzQ*ZIuw><Q%vO7Mz20}7=rY*_Z{+!&V3ua3Isz+WkaS&!04Hn;3|%jt*y;bz|$
zLx{-VuD#`(FjKCpT&T;{`fN7UPD<XL;ic4s92`cu91_UtojjwD^da48Ma<3+CMf#=
z7I_GqMshFk2iBD|Tm|0KQkKrGKO?M0wRiO!eYTdfvNbDGyr&lO$?qq3-Fgl7iT8s^
zn9xPC6sC=6nvcT=(1A&sesw!DH0lS?gyZ5KPmeB=_EgjsJ80W~5amd!$eocY$4Tjs
zlZE^XMv74dDRSm7+x(|O8p*q>hLvIma)Lf{tZ9JmSyKSfJqegkJ5JF!vC=d<=e``r
z{{+xk^!FpoYVwcNhm@eAVrmitG|Z@m0I21h3Mu44T!l~ah@X`f{?Z?3NEh*y)8;up
zsPI&3|B6+DvcYV=*`#f=0))RHIKyxK1u3c=Y!G}q&82M9TT8%v5GU}g!DiSg=v-C;
zyD&ED8lA({sH?m=r-n)^Axcw(`U852$paAH@e*ZBc&eDeBYJmW_AeaEH$13(c>_e&
zjr0XmRh^~EkVv!uc#j4}1n=fD9Bga_Oa$4Ma23x->EY2xbz0Br`5Yl3_JhgeTTh4f
zjJ(P>i`PXo(#erM*gp_$SnOVLK$DU5<iitf@u((icBRgv9>0ouUwHnh(X^mRB?F=_
zZqflmu1Nctj6gadB6~k{!z*;u*c(7Ik#y2wlxt(jxmrN@Bxpw3@au47_wrbYRGzNP
zGWRGXHw%LgpyL8a`9IN$IO%s*JI&<v4(_0XYHL5Ha;YLNz@w8dR@*h2G{c%l^zv+@
z*8=%VQhPZ00}-ecATzt(W3XptAz?<t*sRoA>&hn<>_xBXK;@|t)<x9XZ&IHjwRzpJ
z`r@`sRq~;bT=A*SlnR>Vm!`xS3-WWEeL7a-!=MCl6;sBh#@kAcXn(#I=$&I865$*|
zl~#y3nAW-_@aX?ANc>3TUXe1>;5p1VzOa_ox>Y0Y!y(M;u=XkVXsCksixrT9lX7?N
zH=iYn&y9_w(xGVHhF+)WMh1zY<SN%UqC&+`n~7AZliNEqcCk!RlfICq{<(5t_epwC
z7&e=k!@{m!?m>UCZ4Vy^BC}~JGXJBSR3XWJB^JQl^`&a7`(-ZP+3*z~{~$Q%4tRM0
zb!Yd6(9##+p{#NUpSOY|yzZsd)yy&r>(9IL04AY#^@?nVLH)PR1H~sHjUti1ZIK>L
z$L|@4%PM+~-o0=*KunVQZG%cQzSsfQvZq46%8n3#_>=e(&6{;|uGYzjh}fRgn%$p3
zr@1lv(J(fqSCbw@X4z-*DY9QM_)3PVaT4q2mSRPvE-OYD-68x94))ucR|W_#Km!F4
ziJ`{fAl^@S9kfoBKM@+l+_Wn!Z*tYi8htV3T<Ig{^D>dr<d=^P(Cl(^#<+)w-3Z7-
zM<M0di5bIP^Tbm^Z|Y5LruK(2E|Sc9*pnfD=eE#!t}G*-Q38y&h5vNo&O<__=XXsj
z2%#Nv?*sfuVD;Hrvt3H2(U<0I6UZF7W|MF;EsvX@Ly=7LdaZZ2Ky&(XP*w}s=_(YN
z1>xH0ynAFbQy4Y6Ap2yRyd0sAA(%bL#ehZ=CRky$Fuv6<KYr-7hdOVF-;5KEZNk7z
ztH4Of%&arU$5ezv0I}r|F&w6`+mz^8``}tHtHz$_78ug@zOhMpkDkWi`H5lytBGoj
z3vxPEDJbQbn#Rz1W+)8GqLs&+6&6$~Pvf$G;i}S=BdxDUO!zP=J=oIJVuG0eb15i}
zhy&J+JO5zrbMeX?WJ2e3D&?{nlWyJ4iHHv?xd;8UHR76_?hTL)F}TA1U<Dk5FULb2
zFVl>PAZ6S>q=3ihWd%2~5Iq}O2&lqgB$E!C02)Dk#Wl%^kFZ8T-dr`Kgp5~t;B3b>
z`);@V_|jmlN{(vhFnwAIAL);XSZm@`sXOXbOpVFp(i^Tg-)MZK?*HQa1%v3VTHC4J
zmDV`{+f}J%g{6_2o{SsNH+SXav{(hplA{0-2$2Ro<B0mQJ+yfeEtRp_O(pQs&603L
zFpFHyNIH45i{}b65jSKKPN)iR)pOK(`c>DA$6SqdS+lLmfupHH=*g$UU4*&qamebg
zNkds;SF5c4j$G^}dLq<CmTaWD^!LlxQ0b)3K7S<KN!6HA?nibT#!n(~83A);sdWn&
zF-%+_A2%dcet7K}oh+N<wm=TMXa@&N%Uzx<vEFWuC&xOeoj<>s4&&m3Ky8Ficps}a
zikH0JvLJk%z+q`+1zvjuRKd?yHyoxMinETAv=b{Qdfl^cZ>{I;?-PE$Z{;iv(~1AD
z+gI)+oc!uDP<QzLu>S7zoaJsp)LjsrisHBL!QT&+T^OU%5O>Sqvni{rJr&KaEb0Ov
zyKPTLM9}3vgYYzmZ_w}#ZEvAQF&+ni!}9}}HK8MQ-XyZb<qX7mk5L)DFVPG#rJPMu
zzdIFm1ZMiN#Jq-kI-BCOFmWbm{bBHlvsIp%h<IfR>`63p>un|b+>)~>V{F{L910m(
z_=Tu8P6HZPgXndJ^n1{xHlw0^UO(T#^Y>l0fJq*!<dO;APP|2ZNd7|xxq=WzE{oIm
zpaCYKE2{w^pq%a8TsrQ*qN!1lZ}0^8IKIb=K>8I>;I#$R@`bAB_d~wn5k6I&AYz&=
zRZLu|0P{5@k`M6Pyi?Sir$z)h6VW>L`Z%n|2CUx{>Bm{{xq8v0(t=51f8PNXc_8%%
z2rDt!i%ydofB!jmH-0z1aseME%zM`iAuCn4B~f!tr_wxXN07zQHQ|pOvKgr&y1c(x
zfpwQFhpI&Q+hnogxF@iHFEuAK5L4tQ+GYy_?Ki`aSp1N`p9q&iZAOwcnwE!z{kOlZ
ze#X~_4~5|(O)z~Zgr*7_64}iH<mXfl&p*|L*x_daXoC#Ek^UDWP~{k}Uu&DjQst#f
z_rd7*Zlk2{sJ-LY4XGcy@;d9Pb+FO32&Jv?0w!bo7lP}B;9RlY3)--CXt^l`3amF9
zmutze?;TfHbwKg&eDEW`W7+M*)_S4q0cUz^;F!u=V6As*BY6IYO*cyNI;DbcGMIRz
zs24rpfe1U4a(+;+R82MW^H^>-EqYJ}L7`86^KO}l&-vr7Fp#A%@;YPztg{!kvj;p0
zfRK;w?=Ce4)0%#KQLo2itnv`-xq3!ip`r#=nt1^j0<LfwfmzTfUswF8nIb57J0VFX
z;t!;!<BJa$-Ao4i5hU7tEO?C-N8P4+#vvj-mM$w``kzJViO##xYUAGFHM?2P5cJhF
z{PR@84^dy=yg=Q|m_|T0-LDX?<ShdQVVC8h?6=<T=>c^Qpk||0;+q_Xc)vq#lEXSl
zriS?Dg*zr%>@(EHM|>^wGN9`?kd0$>F%3b?zx7CC;Yr=z`U*J2KViYmr<5FrT8P1|
zP*QTOUqUJrp!4r0sj9JG`yLj;CjW+#_!!-oBE$G6cAsYO$LBu@_y7+Ys$)=2l&Nt^
zxj3FVou*Zd!i=*dcs*n2D@`5DlNd+cpQ7NC@?b*}=PbE2<{Yrd!z5!xcBR5oe@c64
z7G;<B(jK^=K=(V{&<sa_keiebcBg{@-z5NPkwiE{o90=BYdP{4hGl3<NkC{(uqx>G
zHhnLTI%B=_eFI3_^OW}5*_3fVU!0F0BNuW_^&1)+D=$1na5243{sJVLEVfzb0cV%(
zN)3gr4+^(}_KuAb@8Kvu25~);?}v)<R&OxUrsT`ga48XeaP=2`;UlG{73+2J2OLv$
z>SVe632fKbEo=P%{2L7`iXteAV+rPBbK~n}>Wbyy*M{x0h3e;Qqcs<x$N=O}^p4wx
z9T{C$T(*XOUzV;x@e@RoUo6)gpSe?;$uDHF-MQ|2KkwAn`niC-;q$Zuz%SY^ypD*b
z*mU3KYkh{0KDTq{<Kp8;6W{Wg4}`Aq?)<=@Ve(`KJE&<K6k;J{hVTkrmni6x{gynr
z6t~OqUa6<8W-K|wkMylh6+*O|>u{6njDpjjCf7kp7&KoG0M0+fb=1^qiw&uew>c0^
z05ElYFL6UbAVYzjuo1yWVKZGv@`FOyu4eo;fo2f`&>uDWD_<79Siz%9ReS@-&C#@4
z@Jmfs%*47o7Rq-=mZ}9WV}{RTVw2C;TnDeKhAV|X-{1ZN|KHCJ(GMa;-X1Dr_uuy-
z)|qr#P~tTRVyeixId^e)XlN0+Q=mdwJB9uQIRF$3ghXPph}RuC-xSiW-6ML`&{d$?
z$-HYKPP07D&rq9dcgzP{#s|4Vl?*&fQbCW0p%K>hgxuat2m#oYSeW0aBZU|O&0P__
zgobX**OAiDH3Kx#jO@-VaTjWJmlL?!b^Mc`*2>8I$LKZW$}zza-zTmQ`Fx4B)rP1y
zYizXxHf)>FT;4yhy17IaCa5HCe}f>WQZX_Wlth9E(Lq5rv6Y@7Jjyw7r9%y}`pLss
zfp^4r#=~DN$_5>;{>%TIcv4ey)6|*^<D_qd^nlW2#;KIgXs2P+26PVO?k+FYX7GH;
z;&ljE$kB>_dhdBaG!I%OYjX0sBgF`Q&yS${tWkx{-3-CWC*^})*AUd|u%z_Ba=j`#
zTV^0n^#MIvoR`VkGz4ns#IN`X8{q~{36eMUeIhsMf?jGlSf4-@R0HwRiTns<b4btp
zH5Qy*-vQ-#54d%??EsTmYpjGfZ8YzvIf6R~42)+q60CL_;^q8wI>%%%Jx-de33GZd
zvcDSjE^#yB=Fed-?W9jrC`c1jpGfwZV@}}3Ol$7LD@e`@E;tDn<J>EB+TDhH?kBRj
z!;DNdR=)&PkgVTF8gK+y=Sp}aW{WurTu?55O!m6+j7VAg{;YTyikO95E`{GSdK6ed
zBjkTKP6DdC4jZV;%=<5rD978h&XEd6S3I}7I8b;qj=Nh?Y>vPan;Dr1ucnwaCp-s+
z5I2d2)^nawi2W#n)9W@{(8}=FNCc_3$-J2NWb5(I@1jba?S5i*TQ_RT=l=<%&u_ne
zmSzS3o1NMYA1(m|;OPDJk+tn0Psn`YQr-y~>D0|H2?)*O0H8jvdr}G$0vL0JnYGsY
zu2uKd+ZC7XJ}r!;R>}d=)f)amCiFlTgk(1I8f==EpVqn`9<g<zF`HHlTJN11%p`l^
z*ilqfxO%UnP)1qgl^?_ZBviw53X=7I;%~vU^eeu$ZODNN;-pX2M7C?xfCe?ITUugi
zk+P(^;iBAaNz22pyS2*EKW*+^WtS|i>Zb)j*QfHPn-EJi7ahQ9zGF;vL+?=(;5E0N
zl&NCmRE|%MTU83%khm(^s|?9>7&eb0UJOUAo<04r%G=7XdV)DiZ<>@^Xl|wDFQS0Z
zB+l<J@!1ycd?>I_-J~(gRrUm6gH-?q<FlIskH457a?fAGqqfxnO6K)G^AAS#3#8Pf
zuJN5QqL~K7@K{H@{pp-(4gUZpm}X@H4J$*7ZzT{)pgdr2g(*yQ9Zbuy`L$=&O>4NS
z9>#%XQP4@@4@S@rIlby0zd`=$A*5JHBVZHqD}ZOVov(R?2q#ba$?*I`X+DabthJ{^
zT$bb?Qwy3OI*W<OzO;Rm<#bQd!tDC7j<T@@=S{nGkN_X2PKl4-L7iZ(#U(A|ZrU$7
z8Np26hIn<j!{JSb)!)u1wLo${xk-(0<hVIBC`0Q}34S^Y4^{9_F+PY)*Uxp(>At3P
z#plCmL{xxm+ouU&r~QROeJ$<sBml3`)UV2-k}PF%gROiO)F+Z9Mp`Y`^=zCVL~g;(
zInb>-?l<kZc!;rI)TP_cDmGG!mG5`saUC(|hA|Ysx|XJnM$SK=LcuQasi*Q`>XH)e
z+y3Y9npCt>vklU>Gp}=CdfCZ<uYRccpG{!-Kl0=*x@i9yc*S4u*U#6_^MSSph*`@~
zr+SqT@9Q9V1|>uJ;|Q$f2BcU5HJGO-kzZWT2dl0pO*q|aB4c}qFfTTzeu3h)52zI}
z3)vB_ql1UI-Dvkw_peuL!-<7))kel2vh5cm;owr5X1$$lovu_9h@JXeBfdB4JLb=Y
zIKz&V_beG8!xE{i?@orTH0zI-PO+dgk9R<aji1c60%K6kifH59c)Nk(rJ=9JI6AYG
zXGK3Y${~=E-ld_4n=Q37e~^u5prR;)v0OZE^>r#71nf@>VdHizAQ_wlR4M!g{tl2<
zZY1=eZe(FutN%s82WSgR@|jHj>mC{EdkRKMk&F$%him9EI|~FD=7Z6i_!S0M+wA~z
zzdrX#fJr=QsJj}AY~LSlef5*6k?1EFWLBLGr#A$9-2Ly^J|FRR#DsO8gV)(j$YkdB
zvl@=YmvXU6)W2eL$o^om)y1V-*$MrjN}_Ow#+4_efi#>Plt{YxQlg~=RJE0@Pw6m$
zrDCGMN1w^<!ow#1;oHO?Lm(uH=<(q<5D|p_8O>p9l>irymJOb&8nN~dj58wckn=zy
zMr>BiSqk`f*x-N{pNaMW!UaBP#mXE#lbGgxns6|i1Q94ur+@)15c+Hppzri;1Tqcx
zo+c2DsyPq)X`hSHbL}(Mdc|dAD$Lq-Yl$zDhXlbSOX>m-7)<i`FtnGfIe7Xp2#|_|
z>{lL$E85A{@p*@J_eAI`V&#lwBzK1Rdq_fih8v52peoUU-nKF{8%_Bo=e8ZO`8E7Z
zUV;J2f`HvScm-rkA+RFmIW`2{699_)@rikk0l95h(B)I!@mS~Et;72r6cdmqC?){?
zzWJY#fG@!2+Q9}+Df?v$NPYGgQ51c>{jCa&oJ=smm*oca&a(i#O6<M-1wqiG<M`?2
z(4hHf78r_tIn_bp*YsHACufyWm5o%Ua){J47U!~*lc)cKQ*9thu|6irp%5p#ss07C
zlS{lKf@5{SbSm0Xf)<*BM=S+=rdw5JnzFR9ff#bRd)}zSoY%sDN6PriR!A=zD+S-=
zia@FvZ_T0aK*$e}(qdvVq<S48b>}X_Aymm_MQdXEvzr1f=54z}1O6cT*oGEka3`Nq
zZ4*U-(4CxDgT+~hus#SP22t47oGHZ=Z!<}#Z=+n$gAvn52%l-JTMX3FlVl`j^iuEs
z)a{NW{+*IeR7W03^pqQNGXsb$fZQhFx=^Evlbog(1xMjC1B;sODuliI=?Ru~ZdGw^
z(;}EnA`!@-w3~KaTMMvWoSY{DIJ(d-No)}O)z5zKe)fNqn`_;|#XsP2PF(+dl~<$t
zwY$!2?0ovGoO0ocq&}gnapwVktM(;jY&k<s^XN7m!;05LI{-Jxko)pcmZz`tBNkHM
zCyz0f@zQ!J|D}WCce)#bk`a%9$!lTra7Bm7xuXCRK8t(@4+(yqxk8ej_-;HL3!Zh&
z{|YRC&V21lby=<3o5<*NGMo0&aMiHHIsX_`%O}0U8%vzvZ^TQWcryS?8<yDKD^$v@
z6kbByVv^-Gw+Q?IX~D6GH>^A(b2Xe6NAkcb9}(x1CcZvc*<DQSUbrkP+;x)c_K`}v
zDZSY3JCAA^_=)~QswoWMVQB^AEF%KZHTt00>tJRRT--L#@D!;`B^(S8dcfn?-ltx8
z@@&ALr>EOeQ1oB)!qvaN#*=pX9ca`a_C??SdE?)YjX>+Z_rl{Fw&$b#6^$<Q_JxF!
zCxHjKb7h7zjaKK~ji`kXw*5u%7n`e$s)mHEAuNNp8jF!1IjmsrlsWTNR}LC7$KikF
zpDuj(c^bGz+oUTnwe&WcY6WZ@C$hSyh5wzX==5uqmdd7^k2FUb<-3PPrjd7*Jgfu}
z2;P?NW?p!>n{0@#1lFkhkUP5)f$0BO0G*-wg&h>k%KpL{Qe{Ic8MN#eH8h<;4UA{I
z!7%8v>;t4t5goEoRyTenTpnlk(47pH5~lzah}BS^4QpfdnbM*$D;K08*a*lWW(a=5
zW-Qd3(TrplTc@Qi58WjJ2C-DGgk}3?ZUA@kUGeCcf|8y7`xlEp^kQ5+o$AgV=U^6w
zW-VP&T}hyPHxHK@rV=K?3^q_VUqT;}o_|>M1uQcCf>(<_?_%sX6EUYzS+RGIVnoEq
zEkMSRUuwwo{E@mFM^TX(IMAz2m#8oXl$@$;Os(a@&j7f{q}}wppd1>?>lY&#f!Y<V
zUb$~|>Cpn!vp5;nB@`dp@599P%m6ZHPh2PYtyS~W0bQZceC!#f>M*EUGl>^|4mQRk
zcqF*^?>volx2KDY0ObKtIbtEfV_>CAB-l<B1+nI^3~F~>=fBRzLRW?^dWl!=2_jZf
z#)XG_0RP3TQ}=m>Q9pST($)ML5Wk{WngGwSsMEB0h1g^?f+m#Z_mn)CO*HY#jUK&E
z&ej4R0wV~Xs{Gx2@E!i&5L}w}5ILo<;-nRM0WRhr+jFO<dZ$l$;%f@*GUNpa={CTl
zBJy}MGvlzD@~@{pKmB(B-rvpn7O|exPKac1*f@7yPt2pLGwJZv2>&nC4ph*O{rL|q
zbq(%UVV}HbtM(@2^U4!sk)?CQ#1x~djBmP}srhnxxAe03LkN=e(W_j%Kg~K5#7xNl
z;K3{$`6AjM>Gx1Y!4|EzcnNuAuHN=DogKUY*y1%LXB6|@Tqw4rmkfS{gyRVQc1h|j
z`EVL#k-r;wG*AOs#mk%$KiXn*99m42T$sUd7)+m@XH6@(=Msr`KR7fBHXJDy;GWVL
z`~1~9`Y^PZm?v!MMPjN-$1=L0$Gb`{X|RxZj#F85X0j@quRu7ohl4TOWU<Aj_ybw8
zI@bas_lLymgx>>vY)-W@D>{#!l4)wYcKXHhO8zk6;tJ8ymel(TO_DzxwzRNGOlWop
z5{-hgvE9|YE}EMQzT9B{)M7HC5e$N$tJa6<g_P(vBP7u4v4FEp9+13~W3FwtJZ-t0
z=&W+uCKedB2q+l7&dAUM(psH#-XS&2hi&C~<79pl7eX{$`68uknka?8&8EIl<j`T1
zPEZOHD@1<Gryp?|BFe#VY8lVmdN#;jw^&?F)pp{lsIoY2@D&w#2TVfu<R_LQiHt$v
z%aAqHU=k<>rbV-V;$zEKRNx{r6#iD{p8R<Wn;{InTBjc!*9Yo?w#=!5?uGy+4}u4*
zWhH3V<98{Y2p_uQkqb&F%vuYcLZB~)TP6++v_N+TR$l<Hcj$_j=zak5gtK;w8VVNK
ze$6pPo*9h6Pk8F|p^&x`uaJr*>T4l39y@e$;eU6GL(AsJnL4Sb#ua{M6o#F_-Jkkv
z?e6cKX4%5u+7?+|A485RlFi+?gy8CgAQtrsY$LpF@%ZIEUZ-P%p5#6-_W)fsFt?dE
za*0_1>7DS{u)y3wm&vTVu1(ko#_G}GtiNpMyxjnKs=h^9(T5u|K-A^vKeIi^eg@q>
z?%yJ&XCqf9AWj3&2N0{HE79_)8aQt{DF3Y*dAcR;$Cp;+PRmHy?G_@>-UD|mD^P7%
zutCqo1<vM6{;B`i6}<Ds$6ExEWrUz#7GKaj%}fD*#0^L}yuAoF3URUv9x4wXj$)5<
z)|PH!&CtSC4EvRaTGet)`AdaL>)KoDGs#31>1v(Hm_#~H(?g;lDA!}UBz%imPMEDz
zo<uT%Pm6;ryOfY^?`QX2evS;8FBg!@nRx7g<3)uWjSD33c>-w<LHwmmdt_DwP$mte
z^Y?U5w%QGa8J-&N?<egMDLB#%qHjf?#UZz#!k3$(n^?C41&+e=!<(q{1D~N-C@2C?
zT2G=VYFJUZxKL474K#F`Sv5{ao7#NVMjyw7_W1=wIy+s?%OfKr!=p#tEJ^+f6roFf
z{3!L2C5iBUZQ|8?i>dL3^7tk}w`pa{!29hlL@pV++J+uhEomuzGSkM_Qp(wgA%AK8
znaPdnyb|+scCfZLC&PCiZ>Ex;M%3q9b`<5klpYAbQ*1_6JMfbrGbV|5$`ukjZhxN6
z8~B`hN={H_Ho!l@USP23H$6P=8)~<@`CtN}*rJdorq&L09O@626p}HNI&MhbOFZjp
zMt42h)9|Jw6!y%6b{ImE!QlrdljPwmDQz<#Bc`bhsrJIdGmp33-cVtF`<KlIH$~2+
z5CKOq*^lO0?Jb>P#}T*rawE~*a(aTpTa`Hi2tL3Rq1WME5DG%+Jv@3@^X_(3{h)A%
zJ>`UWPj-)H;zGR!Nr_J%s~>Y~3R`0tGQ_VtvgIE0brYGp{<KAeu*H?yX8}(?F_ly3
zEW%04(Cu-B|9T-cHCVrKR!0~&d~rG9M&!YAC&q%cAP`c><*Dr;^sMk}O{tyO`bwAY
z>n-6&7@-K{XKM9tL;1&{ke(+F%2~e47kViyF=HXcH9u#k`$|S_dmbf*h%)D2>Owe-
zcLF)y6B;S$AvGy(BtK&7w*A`<*s*`+mF;^(2p6S-qKEw<$Q>=3FyMS32&#4OZ>e!=
zj_)1+CX%1?$k-ktZ)Mm0SJ-sm#{XK7F(~H_+wt^t4D146GGYRLY2GY}?SLP?>#G4L
zzY2M)eMY>Gg&4*1eY{)~OoSt9TTPRx<Bl;cYqy5c6TIr@!|Kl@(p5~FX(4?@4Bn^K
zi3@*|TDxw^G^M4)QvW!EyI?f0J?JBq?C3|JgiBarIU_o5{+JX`!}imzmI(aziP(ng
z_Wm4C7we#)m0cM>B1VRXona9}@AwBGKa#f_>gz?{7Ph~f4sI{SwRsYe&=<RnZ2PPK
zG?xsYzBORI>7iq|k5liJL?c_i1+on*W2`Ya>UJP^i7u9Y<(}rnCU5sWI_<=hiY7(!
zQS3y~Z1l?TIRAmG1E(<a?#2ha%FCOwx;hDUo?1UQfd&p8@Q(W>v04HUhOXHT<8Lyx
zIW==1P?i<cFZ{`jUZXbR)1+K^zx{n}R&ZOv-iBFl$)EZyHF;13-)5Z~1p8O!X_8Ac
z_S8RH$ml04w9w!!UB4Eg%&aRbfxqZHLE>Y6z+c5K@A2>3Pa65YimpE#fSwB@wE_CD
z<G^f|B>ih%;I!YTR;18>4un09MZlV^aS)=3N=JnBF?=POG(;-ZW`>P&mG|{SF}`?I
zgdC#<LRJpCT1be=XG=<JUk6h{e<Oc~5^K|{D~kJ=wno344j_uAA?{LDxnjT_#LXID
z8%#3jG#^8Ct$-d)9+N~WO2In`un)?a2cOWxAqL@>GLvU=1qMDA_OuV-T)fmy96H3Q
z*7W=aF=_3bLx~y>!)12v1@`hKWmm0W&%^O8m|o$8+o0Dlv|OFH^EN^wPNY#8V<e5G
z8~H0DE<TT3B^2sEK799kQE%zjt<|_j;;IxouMnHrIa|H8ZFrN>2WC^xCwF=d&rL=9
zjXB<-P%L@m!C~l)t8OFAm2Fo`!z^9!0=_{a4K{+1`K_KH0W|?`rZU8GkoSBP4f0m+
zZ>r;m4%J$v?c}R2TSLjQYF0d44=nlyyWs>2hfd)bowvt8$jy&is>pxDwb4d|$+DG(
zKf`F<swyh{islaOHjYSRP@HhpZ`VdHtE#6~c`%SgF#|kJHo@tXJ!My>RpJP=Ao=ub
z)Zdw`w49mh@@_^u;txb4PLDEC>#)0lCaFB(iC!Xfbux~6a+0yZD?diu5%%Qe$(_uh
zeDC3MV@;*2wGmCc`^ZYE8+TeHTpAT7rT)-rWJj?veJRA$-4{$y!!)8!LGDk0l{feQ
z>TiQzuq#`XM=`TJ8<M(1O(x>(dYMlHeBy-|*wSGNgTl=7=Pxefloj-O_!eIflAaXT
z4hM8KcRj@^ivx&axDS%B1dsb$@x^6O50W=9zP;nrQ2yeUI@EqfCn4g!LtJ+2`8^t&
z-C^9^0hWZi#qs~<+O0iA`RVG6*|6qjDvWHg!3SRD%X~uScM{<Sv%&EjWj;TahHV+u
z{-lHA2OcMjNbr6X&BNO4R+>}H?zhn134H(CvFjNq6MtIN+N|N|4;iB4#5ClMyT*2d
z5iBYt?hqozsdsP5(6yGGGJhGIz4GAD!}^|M)(G<mI}ahmijjmD8s9e-1_#kPFD(%?
zOEGvgKD6=tQdq|s`XY|V@;YKN1D|n!Dk}SC5!s$|a8hwOBsR`7QK*FF`h+380re7U
zcyWg&IbX_{yc|1WJ^dgrZs9#RP<YVU>1+YqXB#WmEiK3rMnm-KuOIc%>E?n58jg9V
z_TIzm?OjH59&oMGSN?~mbB>Ptect}YcG}pDZ6_PsPGhGr8r!y=hHcWgv28bQY}@#~
zKi_kn=l`5N*?Vuy%=MaUgs<Q>^ZlRwpT-QO>27ADnW>=0zeBAvxC-RgB!argRo0}j
znVkd-ALkP{(m{Rbs$eSWN3`cCDjVBT4!SkE+_eDheMkkEz6RUHyS@^+v^}<Eb~x12
z(F0bDjyT(JP*EBa(vPgG?Go7IL&SVZKlVU!{?Je9R@BMcO#aF(dTeArs1|QHp{;gp
zLaqxJD3T4b#rG>}<koLr!6ayfOlhf`_5~J4o{Q4fl6gvhN{v$FL)^snSsrzis|ivL
zWYdZw=}uBxL;5^Bne;Cmto7>jF-L+AqtKAdYx?w+8kK}z9akTzhsuCGcaZ#8a~0b&
zv!jWC6Y#q!4a>eQ@A^Py!5QBY#UDHN#^fVVtZ9wIXILk=H87K8c4ox8&aW@5Y2C-B
zFHGT;C(<0n7it6NE#7fYVYS3veQo5r<B_dUnxg}(4}Vd4sR{i#23HRxEMzD&v+TRR
zEasAVx;s)$0mTJ#StR~ISVspo_2V$lL{^_dTf*!4${Go+$g8f^i7T1vd9x3!9DTSx
z@uH-T(l8YlTnG`w8+3)r{SVnfW}C<$rEcn>Zz(Vqwn*L;jZIt#p{RvEHwp0Vm&owp
zpd@A5JYmn-T|FxAzvd%2=xqFu9=Ow|A!C%yy`g<q*$I~<l8lUS&Y7pn!LVlzyxkY6
zW8$35!rjHHR3O1$skC`+V^^P<4cHEE*{KBBMO>;83JR^kygJ?Vc=5#KD{d(JB>J$9
zC=j$*lHJHAE5N9D&(~UDuEisUFa!My`M-eP_%Nf`Fj9f?s^_h)_meJ8V4J;aIJ>|K
zr_oPywD>as>60h1q5s<=@Z{PaZjyD!xB5_u^YRQ>jNOe!oc>w+h5tJz9_1upu$*C7
z3dR^)!$`+=(cIQtk>Xk1yVJ|*r`Zy$mm_lo9qt!2iXpfL&rWTP(FvSDz1O_9i$e6w
z8tmyM+!V?xouES%m(L;Bt~Q+HxLEsr4)|2fl#lo=IbqZSJ*U*i&fFrFC(1Ufz)Aos
zE`gLjJhy4HX)si+t3B2~Hl}Oh*p}DQeY^)L*PrT+z==CveV|m=ApuX{XI`z^4`V(B
zN8>$OfTj9G+03<6JNqc2ZQ80gTEhE>7C&VaBJqGiY8CEw0{G}p7u9AIhe*~T%TE~6
z7HQQOqogq5LYyq-PDeTK=QV{K!b%I0G><AznO8~&Y|CvI_=(q5JxJsuWR!$3SVv+;
z0a8c$5a&$X7MzzB(Kx7P6WS1%uDvwSk4h^CQFkQ;Zw*bB@|n+ETS-e*blzOEJY#5|
zLeV*^h5QILVeNjA!qb!3_@q?U>Rq-cthH4vR<=FlB`OKAPKqVbg&oBf)h)8Oii|dq
zjHO80?$!;-W)Sb!YDq9X={pT_ZNRcokjCLZ#^OVS5XreD%6N%o6rXuGlcvPI%ru@U
z#Swe};}(o-I`U8_rlM4PLRl;7%yefMxFSwM8(6fj-et$!)ZQ0{mEh7Le*>^s4hapy
zk`;c62#d>pn;0C5-(ydB2nQmUh9>Ej3<MoaC)Bvg4{rzyJLJE31aRB0zm(YbyeT9Y
zq+ROoeR_TyUqD25|71rG`~hO6xE`SS6|xs)RF&>6^uCjyWpzWcPu|9RqIQJHp>VEt
zWdF*sfBj#6^m5j;;d3)_@Nr!|j%5q^kEZe_iW1%(8bdkgh6HW70X9SfBriep8>PaZ
zd^bg_;R_N2GtYK^lWeJiwKFyW+!~H{B$su<pzj4|g^!?Crcjz(kL0^5M+z&oo)G7P
zWWoHeMVuI)T6G4U?#~3Gis?Ef=<bQqghqdd5e`;SA4PsR#|PI;E%}q()de*QIu`1n
z#)vU&hargXNfgtg0eyJ1Y%IjyyEEwqJ~H_*#}E@q@Y+Vy!-!ODKk6rzL=mw;cM9<K
zE-9PvuC^XArtx{qtp<X41bN@&B?~U=;-yVjuQFsd;mC~xEzCN)BoD1bjzz>&tERzA
zx53x_2vdM>8^~Xw*!c5GoGxB8I&|GX%9s&{(u41129J;s#3CMKA>9$BPv4bsq|v9n
zDj%PeGmknhfO)h3GV9YB4Iy6TsR+~95|~>qdkOrV`u5S9>N*?76>^Ym3<c5SXw})y
zm=@J4stdhU4uY{@ixo%@t3tC9zvE?&Z_}nPlOCD(MYy9l7hZMV-tX5@T%VqUvZZna
zUb1Q|oJB5=M^WI_InAawZK_+pok)>t>%WN|m2xl>ev6m4{wqy^e~d!b|6F71^Akjc
z{!A<IDc5@&USr_2hVYgc1{`Pwtr#34zFn&g<yrtfyc->3^51&G;cvU~OUOB78nY+i
zarFSWyP^w{SXvOGKV{8BNx&yk+cxAcvBoL0{9@fNe2zpp{3x%MMAU84J0Slo4dObq
zCg*tBo!`(Y=tBASK1L*g9<5q@w7~&#B(JnTVo@IGL9f_t7(V|!%)s5Q1ErK+f(NdK
zwh|wxM9B-vL)r6q_kTNgSFpdyALQ%sO#4Xq@Ams3OaH@bwf1SL{U|t^>2dk_`$^`N
zavkZ_)Qt*4pBqg~H*+4m@c^^NRS~?|zbX>84|_tMoJWUc%V5OUYaH!l84Z5W`bGI#
zA4%@_Vr1z?nNY0Ue*l?e;VwjSAQB#6%=k&+Fj{qLlQZGur)1m&c~?4&WTKp&jWIOB
z8HS3FF&}FNo;;X7#*1)@n~0?hWsFOn-H_v{KAHk^r%W8b-X~V?o^q701;s&Cm7}yx
zdlW~Z$Si0Vx301+;SW*{7#s_wzB;TP&u1Kyojb|&>vpNiVa0x?;fZ}aKAV`_o_|DV
zkE3p|wwDAyG)Qnxli_oyMZaZQdBV8YGVU2PDb3*bk#73TO*9ib?H=eqba#9)e~e+C
zJWbnDLd6S&$`Car8;eQp==iI2ya!Gc+0}y>u3>^TI~q>+^FO}}76?eSuuX`4Z~py8
zKF=oLQ;dRkT<AT#4unr(DA|gxZex9;oRJxh3oz((j}_eV1d+?hv4jihpI$xQS$3a;
zv^Yr0*XrFr1l>2-9o}D?Lxl)*bR0XTMZ4hJ^VPKPY|#&K#UenGrEub|-@PN*{kpB`
z_bTg*pL&7t4XDKYvu^R@AYF4p&sK-Se2R;5?t!N$Fp9~p|AEA9sf-7zs>8%ocz}#V
z!F!QTlKeSl4|-~SO|BMZdKQElT>+;atrK|-ywC&h0t)X<r2!Gf=E9>cJ+|D*Lh$Tr
z!Q7OUptCx=M${Mjd@zTJ#JlIN9DSM3dQiC@?6^vh7PldT-FiD^`Uqg|{+*3PC~#ax
z2+P==4_<?lh3P7LJ`pM6coupo@l~%M|5Vt~s+S$>qQmk_Gbs@cgdK)+rKS?fnP*}j
zp-)5_W1CGh<;jiz(_(e!JD+{*Vx@yY;QptGHb{kJ{_|0elaF_;%~gjk-*0QlP%n4>
zXMbgAhVa8wG%~`4DsbD%t8Is2X>g4ww{;P}6hWnq8RA)B%>JhBz?ud3IV&^0bY^6H
zD{Qa7hN{*Zb$7$5>Uhr5qk8FeVW#muI;w-ex6W|WU?sqa)Ue*GA=IEVfdA>6c7|8l
zSCO08@he_G;Xss^UNaMyR^r!j2s#|JX>v?!ztP23V-YZW5!=?8OQ{efS^X^*wf~|3
zF>LLkoCeh$FG#>#7*xp+>5&~FS6M$s)dRgW(-pqS#jmL+qQt4h;9Jlf7v<HF^v{{l
z(4tyH(ey6s`zqkqr;)`W5PPT#dCU~^1YMwk)qizdq*D?7G>jz&3U;(W-UKEF*v8Q%
z=x@+hfUZa8T2xxEC8B6WM#JD1;&xyhI(T71&#RITnCF`fXpHznVMi61W2^*YsZ8kH
z(Uy53`_*}QG`7#6X8LR61Fbfg9Xb@|v|GmNKxE9PqAty+R-d{USAm47q?=YIqNv@i
z`X9Pf95K|`9NgF(8aUVcx_G2Way?xrP?TlfSXPF{#7$?qo-{~7b!9zaq!O`wKN|K&
z7S^ey8QrHhSLq78lC2T<j%l}!g+99qt!m0#F5$}k8;6aSuZny|W%4laM8RnqJeHGd
zDM@g<hDON82Iz;vGT41u&XF^OdR<jJv9G9zGah~`Ckqa^emfKH@7Jgn1)@yyG5Miu
zh_Yz31j$anhMaZ^tr_{SwcVr6ZX5-m8-kJAA|PP5we<n@bH<dY66c9zPYdENl2Flk
z(9F%117Ni{<^N%B;?lqW>G8@Suyo{Yr=zKCDf8{&;=_o4edd1WY$z`HjgZsPquHL+
zOi)DA81K3^vojRJ7=v|W+L<uZHHGDXlsc+{G9tVWbAf5=A;;HLJ22tEHkmoW4BN18
zv~)+&)8_cCNOHu35xzwVOph2PB{ZyHobVaGT#wnYp`f|nsAxS&>f2}1Lz9~nq?AM^
zAmcb<Ya)O&^l{gu+Olb+#nG(R@m`;H0qE$#{=c}1hr<-ZoU|ySpv2ZdjbP+?s3VAC
zR9)0BC`mH+N@(_KD@{J7qgj{&9)BfVxLIBxPTW9f#=9nKF@~3vs>@g58MiCNe#=5#
z%}gRhxfek)c(qEVomQCNv8GAA{bm}@<_Cp|D6bxRmBL0h<+f*28L?|0g6qy6oeJE+
zPxr(H^wT@l+O^-t4_T|*Uw=gwn+w6_wq6e$<oVhk&sziXY!zmt%;%pzh@!E}!~4?S
zGL-PAml;igk^{=P`5Pff#Nxm0&H$?rXTcijr;aS=9vaGbmbRWYL><xBS(4don+W%b
z$7?20L~x#4>lbW4=&>v4v9q-7Fd!0@saDvI#p^_(_eSrFkF_KZ6kxFcqA-Xhv1BIe
zoU>SN4qg{H6j6y*9^J7YwG&gZcuk`h(KsPaytE03TOJ2!f(JQ0V!(n@$F%qiya0m`
zfjr9ycBr-h;h!}Yd|wYcWEka#pBh)`C|y|lzPOn2y)$7>&%MvoBCkD7{~4??t_ICz
zr|lCW@3}%ampqmL1}pE9RhkqPZ061iscaRtRI;QL0xRp+qmh-4rV_)(Dc~jE_+u0A
z7@0!2j~om#j{nt%9kgqJ{toQ2OCSo&GdY>qcg@n;RBs1vx)=3!df_bI<uszunU9Ad
z;k(yoZV)6hgl`jiqGUBagAgL_C~Sr+-j{DrQvsN!w3p|*BBV&STt&_>#a+emc%;-e
z&+D(7RtmJ;rdcHdXLM<#;MiNO0RONHWpqB&@tz1NvhDB6gmI<CWf!v0e@a>N!2#mm
ziIzBLF1Z$0!40ae{xuzcAuXh$Q)90Nt9+bJ7UC7h?!5Pp5T3{FhSb*7JROyJlyqxU
ztNZ^7a~2Ja*Vta>d650oC>+GJ>Hv?oe|A<v-oL3=z?&^&ifoD@(IG*wWMut99Dr%7
zeh&#c8Pv(Hl+)^2>hPW_!O?I3_BYF`#8qsWDV93=&lZv^ga|frc`bR6$g$ScXDXb_
z;4f)!$_Lbt2sZF?P^#U|zjmW#8NqLX5%b`UuzO<RnZI2Q8+uf2o;Tql^KIqcob+T8
zt+ushSj^zRDFYj&?}DWAdC=qfXz&!F`i^6ejtrv?`U`CE-Xg!1I?xw&<LF508NP7o
zW(8dq_UNGvU*c?kzgBkCo22N&&p92f{@P@`@g8jC$HS6h-4?F9YT$K6NkvpP18ccc
zBg*-;jf!2e63^Bl3$ngL?#o-rDAP<#u)3Z3HBy<6UEK@~rd-ot1;%52hx3->*O!bL
z)36-gLo+RIG#J1nNrox;zv7Yd2}TRf-{vYNmo}@6ph`ZvV0BKI?ZMrXV04yU$YkD)
zfK=Lv*ax^7CRN`Z->=>e=VRt|VsL27{+sA0OgI8YJQaL~391H(pSPYg?s;1H|FT3n
zuYA#PxzEXNJP6&E`cOVeM3|B%^qZUb5<wa-U+j+C1U87#g@peF8&Q<XaIo?ICfxb7
zpG`8c^2d{Wg8nkT`M=Fb;-ryn+WSen-O;$qHh+uz-}w4I<OUDAJ47{)a=U6ocKb!8
z;HQ-Kw(^Rjwv3Y#STf(^?TiH0CmoC^k!8j|{~{K!L-wO6$3#0dS-h_Y^xNFv_Sn8g
zBdPuEYY{P)i<+&+z?`s7L08LxA?R`P5j@tI+SY7a5iz;PM^xjC$gkzCft)F;RWJ|a
z<-GmDe5A0PQfmDi!3yss(@C(`63{ddjO{Ai|LTHnfM|!a1|2j5A7*vzVV(7La34z3
z6>~VBQ`11B4?11tum5h(RFuNHGui1zDl=6O#vyvgHEjUz?$c{f>IBx0+l<?<D8Y+_
zgih@Nl|_01FL}?;%I@$zaHQSAJ~OqoUhmpZKQK5{g|EH$u1Bg~0nf;dJU&4~P^)#4
z?ES0bR*=M^f$szWe=@ima~hxz2Hxlpv;?5SBqVAXIJUf9VdYK<KfE$>qiC^FBc0js
z8GP@8)jncfUKBMnPR87%jfXv|3xJceDq53o@E*x<e+v*?=TPjlf=xnAyGE~czCV5F
zuEXcxK@}a){f3A-YUIJW`gKlqUE>ZR>ob+Ut|R6D_XVh+IB;gx4CaCH7-JoWSxqh7
zgtDP+KMA47TTFk_uzB^FNurd4%k>)^)u`BMt?}YhTjvA^Tls`SyUaDGb!-2=>Py`H
zK5F?Fg4YJ#i$X4x{JF!@aLhDb$dCPJgq#7Lo9eAc_?|4b#%&cP(yQ)|-(&GRPsZBk
zTh(Y?={Yb?B<>t5TU$DS&<cEQ596`&PK2pc<Zq<n0hiY!ICA+u&&%olPjW6WYweH|
z^acN2d|y3vvQqs|KK<w0b$@Qx;<Hp!BA1SZA3GZk;zcOV?goZ!qnuD@b}TgL-SZX?
zl%rYQ5D%fX5$rUWRgdPc^C<qQ@|i)y4_XKR*njo|;q!vE;vnST)jIt825%;by?R;>
zRN_cJSc~d{EBBoR|9*jPvQz5*i?jp10{*<$iKri<Fm!`TX!Wn&y~~J==$qYYyZOm|
zZkx48pV!%1%Rw+&Bl=RR#(wlZ5O@QGKIWw}HDBB6N<7I4kIopo5-fLCBZU#(lFDNc
ziXSxVuR-ntwq5u|6PRm*&cVeh-D5`9p|(yulCiT<94x4P2`K3c@aN7oDBITz!+>jN
z$_cXBgxT72>nS89XyOap8dib?M8F1(-Htjc;CI3>45q0_#zNt}8A9+Jm4z3z4O@Mu
z#2o5aD$6eqPD6s#5?U*5<kXHs?tw0dZV1a1>7VrH2ZV5$D#Sr_Dv%0JEe0O%1HiKn
zh!8jlTWd<2M)ccLc1=y^P#MA&PuIyJm%JcE4v6E4uZ{N?bMA))J^b!bxaxj1h4L4>
zHF`25*c0wjsGfiBB4k!TEGjD==;IeC&>aMOMKnX7guw`T;(kf$Li1j`xp)`y`CL%!
z<tJFHEiUnwrKvBP7w`uG!Rai+SP`8db3;gfgf1yrJ*694bnZ45Hs@S%r!c$?+#KaK
ze(1BmX8Kpj`kvc?mK>*A`fi>~$-Zve^{<=Hb=ED$c&>bR6H5ly&ex4J0y$TSp0Z78
z=~%Yl?a-^>;X!1KW%9kRs+rR(|6oktgF+edlxP{pXC=9<L>-BR=TH(I$=NNd;cs7t
z_{tJi3AJr^@!W7?CeL31)h`?sp8=~zthfWiTYAHb1YLF8YEM>8{nSBUl>6m_rW(*k
z-c{BY{P$)4^A*kaeP?b^KRabg;Kp836N|tm%T(?bLc3ibYF`d(;<`Nl@h^H!D>3qV
z#p&K9=dj@<^qLQoB_R(j4J{L$xLC0MZ}!R1phrBBu+!uw8CG5<jdu3s*32LyR`6<j
zk6Ru6#)4nY0S6c3f%Z`3(VT_6r%lvlW)AC4?pJ&K_iZF*#JH!Es#Xi5u6ID=sWw|F
zc!c3u9_<nP-6GGGK9YBOT1sd!bfeGg;eFfyR9<(O?Jp^e0xbQ3`Cc}_itgNT>nNaW
z$fhX-pRR5|X21+lgU*6Nqk$^y0FJ?KGlb-NrD*NEKFgqx(ESf2zd@<v#<QjG|Nbr-
zu%6b<6rFDXiiE!;BivWZaSzM&H5hdnx%(=g+8wiWzpE^&<2Zc9(ggZpl3eDmaTI#0
ze$0l7@unA1fy)Qcr3mfH0e9ts$=x;yv%#PPjUYWUKg}$*L+YXA-5Vl)MOUHc%c8fJ
z+kbU?Vx7b;@PV3z_{ib;^10XOb~DeV(<<w(f(aYRCp+JEecmCzjfyoeN@B+Bo^Qjd
zFO;3C)`4fH(E?0KUW4g4>YRNvk6tt^Xf^wAP@9N=lP&x=;hQ~r+7zC^?rY&erG8VO
zLuM&MfWsaWHoLfho8V1o!Y(ixQgE$wC+@Q(MW($)1Nxm2zVmgEimWxe7&&GXeuA&G
zAzY6H`8b{Z++E<O_>3)8*jZ~DHm^{h5|a_v_RUj~;3*>K`)3q_9aLGXh-QSMGZY8Y
zdppcQ5oVb;E(GtRKj{55rduZxTYFn@f!v<uMzxX--!=Z`0OF&#UKuo;Z}ayTN*n{%
zd!1lKx9_$0fYju7Iy2;cx^fndSQN);3iD{Yb%Lrf9k9T3j#<l8sZ6PaLT~MZ*z9V3
z={CbdIfvL=w|NwyHf0fB+6y>40@;0T+DVhk0n>7jdF-5Wr6)QuNCI;~j;7flxG}RP
zef#4d43zj-&luFhGY~6K4+^oKCgpn@=p%()SheJpUBC0&cX$4#EFOJ;g4ch2xbWYQ
zzj?Q08p<9L8u(8<6dJG}MswqXLnF)p#u$A(<$nwaSqvl*g@D`ie@WD%1Xy$V96gmI
z_1ChBROubDp1~p!Iv>1(?<v*f13tbe;}w+CtVGI^fd_Z6Mi)Von0q*Uh8#PGiB_zO
zsmzT{i${XXBCO5-(5o!lKujkd;qiq3{&+JRE~he20#?qq<u4na2l{yqrA8SC$h`dN
zBUQQ`VIi!2@IFuc?x&OHb-<2-X@jK93iMgMG4#7C(WBzuo@`SlEbrHJ=xW`W1RieJ
zo?eLrUNqX<8aDoUFSGt?$u&U}XeHH^ROTwZ7Ozv&_^#(sn_gtY{MSJWG{JTtfzXH-
z3NztTbdAFk+<#Xo3VChNs90Fz^Lgw;6YAdw7YSZZ_gSLh-@dv`;mD?`)XAoaVT+qJ
zIjq<HmB@30JhxwJrcEC~7208F*c&wld=!Vu@_gvhM}W_h%-B-f$;W?r+yP#JofF_-
zoG-Oxhhi)-al>n~1ZvPBf!5{HF*7A`(@>3<t?fK?XOi%fRUmg^aKC5h$`@L19o+@<
z1!;Nm`IyTNgT5b~G9h<emC7--TlEd1J&L64*&FS`YEyWiyng!;UPoEcFMnYG-E>6R
zo&8`-(-}~WK`IFQooR9BO~<FcnGH{+S0c`4{1~B7j4N9;HQ|pLj5MC~8I5sA>&<}S
zKhJ|r_Ni|U!m!Vd_MA(rv(A@&zgKDhWpP@B@|1nCA;sovLrOafQ{nw}d)aFwTHmao
zB-f-ES=vv@6w_T{htzhPjS(cZYn9!J(_4*wdqhR*At8j}NpnkD(IN3Wd{y!v|1of|
zTCY$T+CPnC4hyAR5Bs3_?s%*cbQ2LC2Rc+TP*@Qr6$S_s5>1X;9FGPC7B6f3qeW+w
z%cTGrv=sDMaU-7qx!*Ynrz}Y6Dx6725j_ouN!xXis7tpInbi4o>TaH^#aLIC2x}n`
z(TWa=B|B(jRJwXPlnc)Ze|yvQv(TQf&AJK^HD6gfulb__H$53_?~L(xY)N_8)H6y2
zHLr5z@~`1sQa#WEI4Fi)g(p%XU+4z4h}NK?yQ%FS*2Cf%a;ZKnXqG`<aOfuL7@c{g
zXRky4@0My4FK}%Q4vMcAp8ruG8#mwH&4&2=PH~q)d|!xp9N`1gRd6*Q%>&2%B^#Q$
z)pw=*;Pra`!HBgWOL!R%%<35pv39o{2RnR9IEXU(1G8zzDXxQ;MFDDwHFmuJvx6k+
z_yULyc-(G;^`J;4^Sx1x7WsQ%Ib}T7&_o(X(Td6jgYlj=GvMSj%#VJRUq|p?wC<H|
z*(_EaB+de>U)mw)yKwsONcnxF0STt)_Z!bx+|I$^>WHXd7edlMnTjTktiLdQCW`ic
zK82^xTX`Lfz+`}x`fBsl`&?*oOFN(mmB;*eRjo`B3e0JFtC+{}v{F&+2~@=ql$8Oa
z``!CVL2oOjWc>~=MrWrYTy5O5l|R(!BllzXWVGtYBSHD*z~PJ+xV_}R7&i+qf<&XC
zDRix#2Dj*dJ8AZqT2y(6S|BQV&<%r?qRzIFz#^!$gXYTSXrQsIW}mVKPd{|u<$jW}
zcc}1{`26Xn;*QJc3UeJb6pMs_R!sb@A6>})h2LBxJb9oi=nWu82h|bOpSVEUhbD<I
zRZS&xJ7>#IJC76*)R1Vu0>DV}`?eekf$CZ15Ff-K2np^C0c&SK)wX+dG=`?+51W=;
z((ADJx0i0vY6@qTHO=$`(R9CKI&b4tLx_ExFydBa83G3Ku~0P)W&n{C3QqUk%!n(q
z?)d;hus3SPc2O=k&2ov5N3ED4Q{d?iPBPY@sdEt?#wS19xCOvIZe8Nq>=Q%+@`;I2
zh!?f19w2KbI9N4Wkp3q45{Uo33z)TIeDJ*tWCecYEmpLd@PJ36JquBVPw49zu9BX1
zviL6Q^ES$g7_es9#G_#_#ya;e^<0(UE%(4g;=n!f&ym<O*8{UZ2IL;jg}Ws;ks<=&
z451u*L(MbWYo23+V@{U1>3&N_mF6<z=6#t(eJE%34$TMm(B%3q6wotu#5h6E8uT?#
z7;RvaLBZB7BhY|UZ{DA){oh8IlCr7S+aW(LKHj=YNB;RBzc0{V0&w-kP}lne0E7a)
z76xqDp~{Uo(8v~H&kcUK8hV%hi%w35k6JZ%Y=v^~1j=k@=*rU&h#o=OR-AP3#QIki
zrBh6CeXNuDUH`qk>mEC?fGzZO`1vRfwE+87kjtd+`SL-8REq5pDMLQ+F9O8^w+X&6
zr=Tp%BdOKKji<3ylI9NxKlHx3)i~nSMqjBD3{K5C7|r76fvp$SWt$k{VU*dW5o`G#
z4iV9kV6hBLa+o9qDi(9KkoVz$wOW^luvij8dz*v!KfvsO7s+-0Z7quOS&_D}h8)X0
zFRJtrA(rEJ(B*2v&gOIB*<-pP-)Rg)*+edTJQ@y(-~#K(zh%I!QN5LYHaog~BP74L
z?z9zAII)g$1O}-V`|)VGKOb!d7M^pJNkSA;V=ttrpG+jgq9HPK>1Pt1h+6km5K9!@
zr$w~g!&qk@f#vnYkS#i43)x{M$#TRor_4J@Slo0~0&c8Ap$TUA7|LC3MEHD2JrnW}
zDO}BUdE^f!gNEI63<?9!r%z^CP;(LTxq#@8gk<jNSh&AWU_WkN$K3Ilq#UoAzxBK+
zmbmzKn~dGtat-f{LY~yEOA6C=OHnXEIEL4g)BvM|t^iv5vw~B}qLji3Tf$l#BkdLi
zRM+O5+gBYr7}H09mN%R6+(KJgEnTkZLp4EK3HmNqXQ{fsFuMS#<)Hf!X)MNDH6>O+
z+nYIfog?hulk2{=6;ui;=#IhMtexArU;MpMWA_rr!VtchSp9A`>j6TN=z|JDBCU#i
zPRWg~OJGjZ_O68x_mYqgJb~*)lseUj9*5E+hcv8Nul7ClH~@5w!;P|wx;^?7F|u7o
zNc(Zz7~&I{Du}<B@T8*U2at;0D+6X7+JVz9w#kSuUs2aQhlqlwc<PneZXQ#~sO0Xi
zq2UWTs7g|hNPha%-mU`VfLyZSnS;J7kS!?xFIy;FZ&5Q|2e5*>YT>68-<FDHS3Mt0
z@!u+|2&pO*V|OS8O&IA6vXlnMhUzyd$O*+k9`MuI8g!%Pc(?~i-UI?`jnw{VhhKw2
zayD@dx{P-K;MddhY>3!1rF#k}{*dLc*R6!4od|Ni|8xn<c^jX@3SpkR-uutg!2ecT
zszAT(C4+<V);#+Q(<s*-B-1;lt(I_%e&lRGSf=gr4`1wTYRMgR50Sht6}&Uc9gbTj
zD4hX+LO25tkdVt?4K%sDDWC?u{!4^!dn}d%*?@2<?Co_HnBHEvOJJt_A|n0F)FS_x
zO@L^Mg;CdFJip0k{1C2G-ZvYif|f<Pt5o5;mV8!I9fib~l(yZ`^B{~g`Z%OX+901`
z+>B$(S*N;_%ra_$*J%oCp0lOEfSKGPWs^H67?C3s@@XN*QH)fKuYI-WPQ<JWU?ye9
zuT@$L0dUiIMTm|oN9hBi?&u;3nwg>m#>snS0I=6^KdWkBrId{>5s%@h*;pp)*y-V?
z<?aXG&|^yeS7V%9%DSpS6J&qic*d_;Z7@c+gNQx|H-)-XRfXfzv_uQfy<l?WN(v7j
z%+wS&iIU8WiGMjqwI$iwg<Ej`XABBOtC`2M54r^G7<nILg(9AjIRCr8t;S-)D*=xl
zw+gQA8%&?T+JcaOPwx&j-xF^=+p)(_!^ezPjX#V!-ko5-gyx?nEL@k{oD_*yDzDo0
zS7Zm2RLgDjhPKwcJT-<NKyFU2S!tqQ9`J;N65vt@VTbtWId;Y6u<pFjJqHZGWF_7!
zJVi3Z@~JPd1gtM1gQ8aOW*kIz(q;M+)JENVSY~K{4kK8y3okf1kA79F>e_E54;VOR
zh0;S75Av>{I{bYztR_SR0S*5^_9~EQE_gqwOG8K19;@W)X4(j|t2dogU33-sy`yyn
zkDn91<WWNqUlWM0-}eCu@M<7mDOAGN1r?|b`u<NV@ijOj8L8F1zvw!8HuUi8u1?D+
zK<dzDT&0|+?;^L{V86eH$>JCRPvFQIw}i()FCuBP2q^BrapoVB&~foF@;ovRrTr#R
z$mcSBx=s~85UX-uy@SnnvH+rhE_*0x!9Xg<lHYHS0KI$O{1zug;uV4d58J+jm=E@=
z-t)-nFk*Z;9>78|X(RS;c})Cy>Vc4@0g6@BEf5P}rR{~9j40+RJv;iH(q|`C-MRfs
zkBZ=+-I6-OUDn-Pd#r)Y=qK5BToDWYSJcK%p{H0tzQ`lODT3}{R}w3KeSbNZqQ`)T
zgpcys0eE+y<O@*AMfVMZe9M^*yh~Zle1JyBy^?&q3-9+!U2@Ov&m6Q5(|3Ntlr$Vh
z-JCJADFQG1lPxY!nwBs>J!fA%g{*rJWc5)X1S-4oF6&^dlB9~P1rJw5fYua_g7cZ@
zuW{^SiOA;Huc5J}wEZ##c$bO+&9(YtqZYx_pLW<8mY5U*y=HFXh)+BK9}vlP!Mchz
z=?v0QLttyqoi4kC_~?2BQorr)?F30-N&I`OIc*+swC4)id5oO#8w+bNFxCbD$pU%;
zmRiu(ZO!be=WI>yVr+t1!B<T&_5q??`|w#Dy)#vo{mIPqF*9a?4@hJ57_8%2R~>J*
zj}WDJs=qnkiy=FzJxt?qYeFnW6)r)~0UZYq<Z$8KzR&#G!fT_8H|C<oG>OLvtW~n^
zUKbrLz%A9Xlioz{UE5=~xI@t)jr&s((o+91$e7o>03-5`^wZ#GBYb~1n{GSeDDM2H
zxssr{;bCDoLMDI+`jk|D5<Vzy6T67Eb=X~9`V<Gu7hNVpR`(iOYKAWxN=a~B8c2_6
z_bz4;3mP4(mSVfEs2a>l=Zc~82ig$)zKm_}TLWxs)PI<RnqIL5);Xfj(_T7Gp1;V4
zkgw%LmV$%j*!>|^L_u2}PKfmIOoC>Ac;zyGZse$HWn_@P$UJF45B?CXDrBQt)XE=X
z-d3ZnAJ%GRwdY~ht5py4$vcS=ClCeYdl93+1}V`p@cVu+DVVkp6G8)r(`Om1CQ`ja
z+5ZsI83`+60%an*lS|r5_KPcWcN0BpOO^&U<!j$qAw!yp3>mU{#y!y8&3S#E2zme2
zDvj-1?qQ8lA31}th8YElZC)6qnZ-+OoE<PlV1-haQs4WI(=vz4ap0vQp{jxos?qYI
z6DacNR^_9kEOl6J>`#zjez?e*IoqKkY{Iy~c%X81vQ^2;oI31W!1#w1o_S?i{%n%|
zKIXJzg`J`n<Q<DfmsA@paSTf)r>Np~q~Vw?3DR6w?X!`^#e*N2A*RaVf?&P1r!t0?
zuAd*pyF9z%gx4lsMALg{v(oumHMu7>M2A}<MFI;yRqlPk@Atz6(D{KIYbFI2R`{#O
zAU5kaaNfzMu4r6hoUmC3@&-7Xx4!Zp4nNnTnWw^P=n(2LfVMTH3Nu4)O$r=BX7Ibf
zmCkjLa<akL;L~NFohb<k|5j#XfOeIHPEW=`2uR!1H!ugggPeB<jsuxJhHl1P7H6ZQ
z4CY#A8fhdxQ=5yltl-;xi(i2yC;!ayzgcJWV#&nn=p1nppPJA(QG7xtss0dV42>Rc
zF8O3?9a2RH&+k3%#^%@)u=~=~u!|GAs!P0+m=?9$=jYon2YHKKN*gUM`?a0U5H+x|
zDGi*WcgJ%PjZJa4s0rf6Iz~_;TV4_4YLXd*2ulFV0mxbc1rF!KiKf?;ERi5Tr_Cgg
zj|t~+d;|Uu=Dxe&jgKuf1rrzh)PD@`R{skYX4Y5y`y+jy_6&c>PDMRv*)Kn1{G1Ql
zLrj2L+*WCpwsZ8r%2B=a!Y_q#It4v8DOBM9G3}PyLQ%YC6s2_f!Bf?B6g&Zt<rY;7
z0<y4Y%P54pK*fVmC(fMzw4%l0EG(1Va<rF`#BTNdQVfk=#`UQAGBo}pg{nVIWJ|}g
z|Hs77ILX-=k;u^*P9w+L^PoiLG{i6+hD^y3x~xXy70NNHB<k^t@A6}`8XrauXH$@0
zTA=ET2a+@xGQq-OyH!s?JsU#9N^xD-r(A`CrU(mEt7C|MxOHuPbbFBfA8HI>T4!?i
zbQ@AgLgwn;COEM8PFC2)SUP_jQv0g;iPY*v^VvN?br(dQOt)2U>l#{7Bv#&pc^y!%
z)*7zSZnS+``EvqP+iy%&TX0m(o=hCfYXybVZq`686a_?7{L?IU&Tp4f!2(|ksaQXh
zt<a3_GQ9O!0_c#n3wRX%r6`iW(}?E|EPhEt8c@aUX;V}lS#TGh5t+f9b1&p%f_^pr
zjTi&fifcWAB#EK&Hf(zVC_MSz_fHbh>A=j*FF;jQ_%`Q$hbZ@7D8R&7Ph{Q6zC$jg
zBL-BYi+hD~X^gOTZNZ&~s^ruHP5oS%-MG7CMff?>r1n^sFuxLL|G;t4qWM4vjX)SS
zB|!#$a>xE9Me*A<gDYfvuESs#B9QP)1mWV#Qz(w`V>d3iaPi#P1fmX|WfK`K-c0JK
z?MCGLwyl!&P5$ZVhVVCa;09Z)`ie$QMB0ltP`JFa54yn9lP-QDtPyEN8hlSn11^N4
zf%F`npOX9agX#g}XJG8oQURB}LL^B9#zAcrO#B52Cih;XVkhP#%5gvOFTaM|1!k7t
z?&qzw^qux1v2r6x^qjvoLr9k*9E@=hsq-3Vi$d)0NZszJLhk8J`&bVTAD%MVt|LDu
zvdQOS2C+Q<xFl3@)qjl8j(aftxNG}(xnK3~@BPDTaf6scu2o>|oKMcZJ)b-H@dg-P
z-7Gpt+hwmOg;_s|V5fSADAf^0Sd1DJtcjfxiD0LF8d(BmR`tX_#zP=6<e!c^taiHG
zq|qtIa#%19b-d0R1N|P8IlNykqeq(F0lMSA#)SU@WE4B32A5AHlK^RAiAtC;W_G}=
zJxr37%8E}y!1<RwQzxyWU+FXQ9C+rG6mqdSoVMQ&6x~G%-TMiii@e8H{rsv><nR$}
zYTDAd*Ic@n#?WT9BC!j<P)U2l5cFl`Q<M&Zp@<udcF!F+1n>7V=E`xLHGcx*1>|hc
z9b^Laz}FZ$g*{NLd=gq<JrdY@;R<qi-5nfSuKlZ2re16~6ns)+bf#tzgGH216*2}}
zJP^|U`GBCN&KQy^)9xGM%|4^YX`+;+Jlwe}%b=YsY5XLf0I>HuyG$tKF9hdT8japX
zmu0ZIJ7eb&oV=f&&BxN32>BQ?-h7~1VkViDO{%GU`-`f7I#dh5Ov5;@sO{lQKPCiX
zLbR#s(KY)ZIw~cXGCm}U9ismlaID&a9ZlsMRPlCVa#fyx&}=Wd-REw9Hp#)hZmGeS
zmGTF40+}r1)j0ir11ITho$1)8^+;w!H|$;`{ydUYdRqTsA4AvQt+E8531paDo)gpp
zcRO-48yOg{nj`&Zq8Leh=;n!pHhR$a@paFy-K<*dnShSK={mg$;p;-(c7t#Ai64fv
zLF>(pRHW!LuVw6>sGf2zVW?M`(n#G!hxcv?j^XvYER!8UM(y|3ZClhvC}wb@%sER>
ztEu;1Z(hgCIq*l{wG5${!q&|s(V_8+pH+PbuR6BcFi{k^b4tmT2wZ=a3|{YUneK+F
zlqB)EAVI5>jco#A9z0@h$~DR{^5tqwN;IL{A<s0wQ$0r8Z`hll=heIHfiQ_#LZjFY
zfJ;_e_c>XrZJmU{y)ph$eI5nU!+CO-Eq!zNcqjP};{Dl1YCl=&3G*KIKcX@$`T^@u
zmjT<pb8E2WVuCd=9u6E2>Sz?l={$?THV?B_QHF=^)~iOkS{g03G56UQw`?+R@v#D2
zCJX^xzt5o^VK?iI&&XkPzKObnD^lC@jfLw*;YQ~!$ZFOm*>Wnc&tJH_G97iFtY{!f
z0%R?HDK?hLLFe1w9~8WN$GFNWt68Z17mA{x*hg72naZMqxB8eCw>16sOlsV%(miK_
z+G*hRd6)Z~h?i~R+elF5-Ln^mAxji&x!cz`OYcr3CfESQ5EaTwu%<9ZOt6SaTbT8T
z99TQ9-}HT%)(D&YN&ctr_EVj)xADNh*VJU{4x#(2+AnPB%xSWRMJqMNVob?I&igu$
zXUo7Hs@2|Aw`WA3EtWM8weSeYOULPFE<?-?mYk8HJXw^R+~tzxq-SXZENT}2>UV+|
zbbL6H0EI($<f@TmuM9mUARecd3AuXmTM`m&#+fvN0GNr@>zAOg{LxQitcbqH8raJ2
z><w{)3AYjknm<Fp6Vdgsz6Q_^;tw?a(lVcHXSPe~*K-vjH<)s?-Ym249YMdV_r=kF
zE!_O|Lo^2O350TGFhe~xFd$EUq1$Z=S<|Wtls9rMrO_{O_U@FFTuH{8>N<J6uZ+o|
zGFcfOnS3op<6=jkQ5E|7=Nrlmznc5w7Bzx4n1$r6@%19KRm~%9LR=?+IWeUs77L1i
z*^osyp3pqXsuw<FL6a|3<cq-jYRF(ADx5uUfj>ChRLWgtqJrKqhuZ<-!wcI2X4-Y^
z9%{1>YFhjoop@9@4#ovkTJik>nD_oPm1Z2meDxQ78b`5}Yfp9SYZP#fF2;SD+z?Ul
zuw3D%&~D!izpS5#{0jMhz5u~CNd51dWZ&Qoq2~CtA8V}-iti@!C_$*3JHT|X^}sjS
zUAE=l5kc6KT0d`LGZ^B~$qYHb+HAjk-kCq17fXLEDRuWh{>(K21kj)VuGVueVt`Uk
z=4la{u;_gZGCd5M%|wf$E{`8P#AVQ~L)cVyiYw!gB_OvExs8-SJ?{ffG*OAO!}s7s
z=ey%S6c@x#_YKz9q}@Rl8SP~`{)@qsgu$Z=J93fZl|~t6a(zi9_(5@`2?P)rt|N7+
zSF_?23vQShvjJs{k`g(SOjqb3SIHe?uLCm6gDeyM*nhsCR19tb6U#!y;InPt80V3)
z<I;M6w;5r}8I|y7&`DO9iNlFw%QYp+XLs?_<dnjbAr6HOA#ixQO8YI8XocC2Xo?Xp
zS^4xh`SvvclajbSuTOy21pkaJp=|@L7pT4h>d{U<q0$L&nvX~fzb(Wl+Hkt?Qo|FV
z!^-0cNLBST_#q;@2Vml66wpZbX<EEOhd4hMdmWcQsnjRlNI(Hbj)sWF<re{0jXBUk
zU^fO$8g<Y1L=46|J|U*&i7<`N2_0xIT^;J5Pk}BYG$EFKtCy==QH4+wYd~cHsJH)e
z(@gHsk$#QA#7uV_WViysJsy`7vf^z?%>{l`8f8qu=DNUA)qBGfhE0d)7oSV9Kq-xv
zR!L8R#IPvIO$Agx`4&BvO2GM8AbmtJ@0RcEuKOkDI9KxQM#4yjbUpygPDP?vC9(+y
z%$H7yR^6#qo<!YSA#Xd`<im&YL%6P~NUPqYAD_XKkZj#!XFvCc6N`Y~_W9wg%?ay~
zA*9P>Bu)SCuw$OQG9*)85Qez9uyloL9PTx2{b!n<)W_&Astw8dPb0~OU5~v;Xia}4
z_K3IILlk6l;=JF2<hUU>W!BKpo#_)qVn4#J!&&6BuMs^r7J3%?=E}N~nMF0*XBP|l
zn^voEwx-hZ?K-ueZ|kSJ-k(p_f9qub&eqz*?);Be!tUh#nE&~oLf`eaTYtYpt@&G5
z>_SIdps<_D8)y!8d`<WxR%k0$5);eE{R2XcCQY+i@8+;C9?%d8MwG!${Zin&7O}}r
z=cPT!Wm-u0zbAeu9gS?~|NLJ2fjvfR-p~fpj5!0rA*>)I7{qxxrkWtIX)TXj!*tBh
zgg8FMsVjzf(w9xdClA}@VbW9yWAKq5OHhY2T4`ME6R6#v{+pUoqk*DuWn>~d#pD7q
z20OS@)xzo;guxbLi9N)h*_e7Hz5B#p%2**-7RHcFT(Wft=n|ME%K>-R%s-0%u#|uT
zlY~;Ne-zK`4BaKE$jsIA!%_B_SrZ2)3<<B7Ex@Yv=Nhn)HO!|nm&z=|teHB@_omj^
zNy~-ZF}$`WGInkG;c3!+1HJ#c=kh~k<?HTJUmn&M7P-v}mD%xoe>Hn#gKi?wkA+ES
zhpga1_0JP(o**eV0#)W#ACVtj=vuWJYfbDnA79f0Z3f9ZgY*xB8um+%J;jSucI$Pi
zLC&`)RyM-#fKl{OuFdAppPr^Wuw`C9@=Vn62QX_Ej@oD*Ae&n)BFWC!p^sA;_hdV+
zd0`L!QhlCm#KD999AZ&{U1(Jv4I$}iUs+n;#sX|ee)~4xhde<*{D^VICaX->1BWp>
zMjicwb{+x%#rcjR;6CbbkB_PUnbM(6xn!-PG+GTh3B}NN_Wxp|=DrsVbSncd=@Gz?
zQiMwi>PC~f6A}MXt_uPO29;)RT2_s_Fk$lP<H2rHE17rKgeOQ*U0{AQQ{`Cbz7b1<
zKt1{Vns>guh*_p>z|qsy{k5ykttqE$0#%9zQ=}nY^lkT~kcCrN=p>ripnxmmdLe~~
z-@V?;`PMt*>$!^-f$}rBVPR8r6bCuw`u#Y+mC&BJ9f%f${tJmee5-ciEwD#_JiU>;
zcT<!imOQJ%ngk?_5d|gL6&V`-j8Q;X2Nbx4Dj54u&n~`3oc_*)m*MfP4?>zs(bhUz
zw2IrfM$)Vme0r&x$#D@~h+p`654!_{)AjjI>*V~_|LG@)UAem9?`H!OqmP9${YbZ=
zmX|Tal}Q-rNF}2$$kVaTb~)1QdT2_X?x|P&u(1@e;AL;cr!z}upvux{#fm_1(seh{
zlyklF2*Wv2U1+3x(?i%zIoHAJ?V6l=<>#4O4y->3Z#c8q1U@(kf<I>|#ruEj&vpOv
zhhz$;r}3U8-EZ=g7to@P>8pT?{I<tY)ZQkJ>9jY%pKcY(UmQ#ou8bkvVIFOi)}WPC
z8a@phT8ZB)+cRd-md64iZJn5q);ofj?bVLVDf!A9k-&~)Oc)r-(p-bEVjUaSF<sE6
z?d($ng#KiUXj<(b?|_n@Gy7N66f;Vku*--+zY)0e^1eUgITbyXFEYByuQA|S%+FzQ
z=)y;KltUPNhS`A?aRW+@?g&t#Y;wrRvR09XaLP+=0zOuWIPB+dqz`*G-BytSzbR%+
zhaY~h`NRP&Mq@$hyDc~SKu@$C5VI!`AaU+)Zyg{{KY)o6u%Cm(Chh=N1OKUrwu<%7
zvP%no=*y(g@FrwxP*dc-5yT4opiwf{c6Vlo>9#03<Cg15{);N@T3G|58ksuCa#+Vk
z){ZskaD1B>k$)jyz75^cYAT7ISYs+haOG6tW>$}mX|SCEr^0SHGdN|$i9@EK1Aff7
z*n6F(Bsw=mnY5a(SwW+TpcIEDV-`TOfx^+<*!|H)8M9mjc$&|#SS&#D)7b-$n|zj}
zd(Y#YgWrGh@Vk}f$^_Pv_5Z57bL|&V_3r?Sw?Cfe+xKS?;XaZCtQ;|JtM2b)-dmHj
zH{2qIeA;TfW3h>Lv5GU0B*GeGP_){3q1|K;rNS`;Q_GsI`{tWU(J_pC)*^xnsi6e(
z+7mZq)^SwO=6)gQve_L@>~p9+B(qf{icOp4KnCwS!0Z-XqV32}B%N2o0s*S~rV?a*
z`rqEn+GJ6oLSo<Ljwe2`vgVB#-;GLT+F&yuCTcN*ocLs63J}f7%{5%=;9o7U)h*Kj
zc3BBBRW%dc7EF+Nt4>fGaL)X~3{rc<U>WUN9(XQ>V?XZx@5@Xr2~?mluRzwHzCeTJ
zRe7xYDr1vL5QsGa>SF#f+rIPi6)A<BgK~8bg^NpFuTblFTVq4bu2!ODdY7n1KfVd>
zv5!?QwR%3?{IR^wm(d&VOI(mNHFWuDTuxA$vzDHDr*>5LcF-622_&gQRPuG1Ns2<G
zD@Lik_|%aGQBz~J(mpek|D<ss<Y4Z+#LV!QZUbA^+CD`QwYkyb>{UAG^|;~>^MhW|
z)8^#A{shyKz@Wo{kMyQ|C8S2&k+Fm4cq^6t{`fE`py=eiiYcnYTd#Xxl53RM3BOMF
z0Y|P@iumt~$)&McS2St&VR#jl@1iqDq4g-rLVKJ5yu~jx6ljM|)ENcZw{guf8@Y^V
zj;>sv8BpC{lb>(eLx4;+Y+fh7q3)7xm2S)@0<B(ZjT+LQ6Zm?`ZrZlxD1+;~utc-0
z0wcs-DM<ss(!zr8KSj#&Yw!B(WYJJ_BoyR61@H11!F*Yg^KnLht=5OeiLjn0G2|GN
z5UESfeR8*)>;w|5!@I-84p^`eF%2;YXTjmF-rnINqswS(FgE-Bwd<ooyNjy=0OZ3R
z|Dz+&wwQdGU0zS;l%{YUS9pV`s32GhrfUe1pZ|Vum+iiw?X-JJU7r6rC%NXi_(LKZ
z3VpuEhk8-i?Z*uuWqLWUJ;qHk?P1a}MZ~I*4!F0r8HuY5+=&W<6i>6P^-Us_x3c@W
zWLmyWb$$l0v0DuN&)S=9HEibVbqXyBxLKNh-yfX-F*q{ed&>O6z}8WOBj43zXqC@Z
zA6nDfYj1RQ*S&5YR$e|MeHlWFn(5n}B~)n(>O7BN8bNVJjM)&x)hIlJVM6AnM>j88
zCdTDK0aIcUW@K)twzSCHfV&Ld797Y^>;2al(*+<=sw<Y}g|+T9uKtJX)|HG2x>6Q|
z<+OF}ZC(;M_?;#Wr3lgbp;qQ(!1d^=Lii}HxPs;jkyw8~`q<*LM6<B1o^tB&QC=3+
zU_qFHP|ly)t@~juVJPn?RtdCfh*TO!`?X8fUnT#A7~tLmjZAvYj;u<%nrw`hHvP~=
zkIR1HJi@OWgb$VnxqvDF%&!=TPt&Fn2o2yg>OhYNC1ABc4A|~@EokY};4;bD_qQy~
zcYH%<1Rp!cp(Q;kT6^>7G3{lSbMF|4Hy$845c}nQZS-p}Jpk|p*jQF{3;yNQjT&RD
zKea?kruA1l*gefr@U3f3KW<XB47}f44(5C`t<dMOZ4*LW^rm?>(Fz&*%%Pn%XO8fs
zN+w}vU^+0uy8VaO4O_m=`xf!FUDK*uZPJpKawH7@%Jx4!9EDT?73f)W*8R4XOs(tv
zY$G4W<V}40W8-Xy&+08$*U+(Nb(Nn5`$vrhXZLqLzQpxF;;0?Z;e;Uz379Zz9Kh)R
zMkFO<w0LwdA=s#P$HoodS$quPk1BYSMPOmr3-fXx0eP|GD2xDbm#*@89l-H_H&_^$
z2<E8n-T<sM6wbR2{=RO--3c4|Y<axbnU4X4fy*0d#Hg`4MjG3u;lbk+fRCLo`v>?c
z$R!kg{^ve544`RSnFz0ADWkF!WK0nXg`npU(mRZD%NfkIyVvIUe2d)J_r;}vFbL!4
zCA4qk^EwW(^mdb?wPqRJ@1a|2!gLSKejsE!a7T0+K@^!5Fl&qE*DbWl-ycPrxWDt3
zyU=|Z(>q9~UkkgPVU~Q!SYT`&6NJ&i#YEU{4WV&+C8jAPU4z*gZplFK(FBk0PmMeN
z4R$X;qhY)-*JqcG?&ZQ)IV@+X=2D^q9-J@@Aqi1}EX*#3Gy;Gs=`HTWb0MD(Sg5nc
z$8u}6N--M21dT4DjaHFlHPU}|Dn(I0vksGJF_Lrhwtjv#Zt#6k92_QX+TGfN>04{C
z#xe?n%ZqafOMvNs)2!6y^j(ShO=5Q-Mf#held3jnwzhY$R<GuB!nZ)xp5%#19_Q~|
zrFsDz5UEq4kei2s*dwhN`w58P9#pAD%2oBxW;X=r;GEd9Q%ZIFwFb}WBg-72O%C#I
zocpBzqv;xh<7~rqHg?i9wr$(C8oRO4*tVJmjg7{(ZJUj4H#qNi&YVeRCqJ??yYD>D
ztqZqjjLsR-l8Wmn<6z7FrE{!RYw<ycfa@B-HgLK&Do1gMp#|%Da9rPdhuLRT5B?}q
zPh~M=-vGP)G7{n3*=8@Xu`}kigIN}7{<Sxom9z9*Zui%izSlb5&VU|rgfrxy!hrg4
zbGAh===M|;;XJLi8j#;eKV#QG9O%mOFaX)zeXGM0h|vINWh8UH`REA^g_Fkc_Q7gw
z;Kl$<P~Ahf_yqj-7G<dM8M2p<v<JVV445ay)l{Z{%V9;_SERZv;c5ngO>oq%zMy0(
zE55Ga*HfBJUhD=YHWnA*zUAGyz8x>1hUKQ$B*eYj0UVeZbel52!9z(4M19t_gGM|7
zKLNjQp<jrrNj`!gRiM&-e7jnF{7z8vMCAB@p$T_`A`3wx4eoh&GWF_L<FcGhH5_Z$
zY0i)#rhW+!JYROh+fVJM-l#nO3(HE^#ax-JvxR@%tK<^_$yUf!NDQ7~m(KxkYyccE
zPz#k=PLnKz0WjcLYt$S`jw-;Sv}E)SL=!89xTiiK4mxPWho4v0eenyH3J3$J)$-Zw
zo(E%#6C6E6abr~_1!{BWP^>QNMLz1xX|oX$^(pF&z3H*SbM(T9H8xtDA?LU>tMtRk
zc^YrzEQXp$l@^oo%BPB2O*(GiQeH@x8IBdSMH-`PjW#G5U<^w#w)l-*_&dZDLg1dL
z%HKCo#w_>4VHJaO6-FU9Ac-P&hWE!$diTM;q-w~$DOc##o&UPFQeG{i5U;s(f-*6S
zUY9ekM~YnPu^|u_Ks9K>@9i%-k28ug%}9DJmN7YsHz85)qBG=ozQ+!3o5^a#cL%r;
zK$`+s8r4!w-n^NNZq>TD>xnW6ar0{6$9|Wyl+Tbwk%K$F`#Z;z%~WxO-eICc1TT13
zbZ=?5GX@UV<E*6^76oeoNo(23?k7J^;S3Tm(&9#SQqcEaT&z~|6Z60rX9(oM=hzAu
z8-JK<_ghhZ#ctetQQOM=Ic_TTW(|2J3S}~lNjq<ALdwn!IiF#=X#7Qvv?I_;6Jwb0
z6Vnx6KOjA)$<}*(*VyM@3+tc2U;VNY9wU2e2oD*8E6&>jPg}mrun~k$x4q?u&uNbc
z)7>7src^DsctW~;wIK?^=LPhcR(#teh6EyFc|@E$QK4c3c3ytikB2@0%p=5i^#&+z
z*UFPD<zc#)o%Z(dAe>M3at~8BiP;P!GF+_wrYC|L19^@jH8aY5MhjFzTI<^#fg1YX
z(@DF@3;cVkL@tFC!6Y;|&m7^hkE#knWz}?)=ccfNtP6yR&!}+&Fwo-j_xvYu^f^%m
z)&KI<^DS|Bira1nNSwcIVvr7GsK=TAyqze0&exRM*#j}U%<I3iTjpi)(R=p_loVH;
z|FQv*Qn<=*Gd?;jrOHY>wyobX)Wi+(V<PH28%u64HO9bzCM-{h1ij$3iuxp!CL%US
z>Kk{}-YXPDP|?^UdUH&@=P|}~cb)bNFzh5%5+WM=?yXIqwKx+fTWr@Z`0PMJg|EiQ
zayEAZF&+x3qXSy68cs^R4OZr1NoHrk53f^&%l4bD_P^NYQMc$Dl;2%T%WcUvYKCAG
zkf@;S3>-7bhU%x+ZZh*mK6Wl>;u(5jS&}vO*Ww<7-0M4mMKta+@SBf?v%R$|t&q7(
z!L#Z35dF>rjAMBgbgEpzzL*P=PgNx|$JMMl)&1wmG~tK&IAli~s|)B$0a26J9RoZo
zdQ=hf=}jQn=GAud6(Enm;XvLFWwOaM^I_94pE#|ctO7&)Vi_1LZyyDdMkF;o)zfTh
z_WV0n492X71b*KOgz{<3@FHqr0Johs230UH%&9Q^ka(%(5D=IH<tsMEmF<K6Cy$H7
zF-FX?ivAz(9<NTWSsK_(8lU@z0@ud#`GAE8Wp44v#j9~XpRlL$K4#fbK){lUR$RZp
z4H@~VhsyWe+{0MU0sfL*sC|_1Y!0?R>RoKr+y^BAu`u%AJB<vfs!NT`Gj`~sqA322
zI_&{CP^k<otv39S2WhdQOP1LrTg*C4T`jpD<Mo)+gkuBNj)k$nyTN7OGO-^a+IEBo
zeJmB$b7#|)kIDJhN~twb)XCGq*vBB~7n^G)=_ZN{LX~411N4z*LyDh{KSLU=RxU}8
z^Xsdd;;kdjBlKTLpT$&~RD!2+AD?BbL(H;JdCjk_dIB`G!b<dCf>hyszsAR-+CIHG
z0)b(Qz-!98{*e0SGo_&BZ*a}qu+YZ|QHQlX@^%L}Ac_Lqkbu{>z(;@Q^REE%1<>7B
z5pX-6V{1jmoe6+J*xDHoTh5XwH5E(}*1rS}FsfG6=9LmN=Zjb;9sL8TD#`q%fi2EU
ziD!nj6AdVJvak64O&3&;t1*HVlcc}t9*^DIk9#+WHggTG&H+CzV5%aPf<DIWzN}bj
zNlDFiwmg&2UHI-Xeu$aDNk_gG!Yqv%ol+eK5ty}aCZ2!u`7!n_L+F`Vfq#ZMi#>0T
zy>0ggeBBYLp{dD9AW;|y`v@!<Wn6%BO#7?=QdSe8{alo!6A8s=<e4Is+4dEz-5A_T
z13UkTN?5XIO^r4NU>y-!+qIuY(fxPzy<7<bsJQlQ&}5!phR`d_zr=`D+5!4=_tj~p
z3+`1Y)^Ve+LANt~q0#=HXLf)quq2TN6BzgpdahCki;x03t{1#0KzJ(eQ7&^Cnkhfp
zjgVJBS+)r6qlm*=g*U(8T|_>N96pn$#ZMJt%5&EUgzXitmOb&3Js2l|AyB6$yu~_e
z3w=c!(H1%4vZQ~KW9H2f`V+}iQF4@buOX0Y+wtqC^U{6aOlOcxh${&TxtSsdn5k+9
z#HzA7G2SQ&U1un^vU`n+4U<JpvDbIDV8@121y#{rs@@dL;!3GJ0Ni9t9R=*=;+rmZ
z_}tvE%qKjKqKy0D2Y-GXSoL10xu~@ppg#RB{;ndN{5#?ea>ff%)Vb?ta$>bE);%Q%
ze5p$(;OaZIxa=TjL=Z_c!9^cpwb92L6yU=_xf(LYSTXlwgSfF$5t1m#0+k0D<H&w3
zL8>xd<tsc5t^avA934n{7h!x_(Z?Q8b(;`!T$ks$;To`r6h~*#VBj+)uw`gT@+K}0
zc7|Saa_j)2lmpEQt<ch_nWn7rs+QYHE`YwUgL;SE7tpAO=11x1RJbjT{VhH031xx&
zB~rc}m~)BS|LC{+xOD&YJ-$vnncvRS#D?Q<Xu!_|)aXu1F%SgETC>T40xd=pnLIWC
z%++MK9SAQ#>W+)>=h{IGQZ2LwQc9U69zTxlf<1*Tfo(8%?M+#PDJD`sZZG_}VTj&L
z$ay2zw0+_;if&CSpz7jeA)7>{QnHCYlf#Y@JOgRmTr%1jGXFVlUE(295d@FRjG44D
ztGcoeR8*S)zvDmmlgO#d6(`4wAfvq%gz;CMxPs3rH96S$LR~tDQ0?!WexTcbVB2wl
zd?vUs#R3a58ZEIa8>gY{gt55xJOUiN@4PDoT0(ku)Opa1Rs|U0ZcO3=iJUkS{N*X6
zTl&=Rw@bQ!`-N%xAn9>|wajzq?;i)zGz6%C{tE!ArQ%w`{esU;Q-RQKyazLkH&6Le
zu(Q_Yo>F7h>V)7+T;@iCiUb+E!7;F63+@q;gK^L?l8-2g*Q70qY(Q)hflg<&B(E)O
zWYKr}+qF>r?rGbQ=g0L9R<?KGVhPi+e)tXx4K_K$l&)73-WkLq(OS6#eY4%R6Ohs@
zkoXBs8coD3Y}t#zDb~lage_PC<ej}vwJy1~DHT>LMje&2?QpqB@s=P+y+Vf}Nm`J0
z?J5xHAoEDYqArTS=fa~`%$<V<v@2~x0D@@ObD!LETJLzmDGI=w@}AiVe1By!sQcBp
zfUecYgA-CHk#EIxUf26#;1BoDxO6zW_R@JKBW9%q1Kbs|ocr)kKA3gz7is;f*4tB{
z8~TW(TFpxiB`sVPb7#&_L;YGOqDSNdf`U~m@9(Rih>WV;5YSM9Dh#SNnWxs4vFvdG
z?obuBn>J%omY^^6BAITBs8Bq$)k~M??}*jg>&Te>9B3Z)WFe5a|B4?N)6nB1vgm`u
zRtIqP^=>}KfKQ7exHr*MFdWiPqGB@8-Cq_6l|4o6kT$}trr>`aTx#XPRq>pGS2FQm
zXfBzkd^VuGO|9;NAIJD`g>EW6*pDuXZ?cD=sTTC*<83z!RxX=?sF_&n12;}V9SFNu
zpV-@t2S0A{kmHyt)j*4%tr`4@0dB=ZoIs;`@fv96O{EA92R|)9CJ#}my_s&%4J}bj
zw`n3UOX(y)Y@A{-XblfdY_?ilm+KC%j5Vw<l?csa{f#x*ai-tyBGusbF1cI-;i1nQ
zcY@~-c_3`+8KgqX=gELPXHQ%3izq6YVA}$V$kzGS2Wgl62%FTP{#Sow)N+T0zpKQ~
zB)~yt85$=WSS*B0QhOqfz!wttMwyv%%7b>RFhI(NSMBvgdThq$_GyU?wG=)lhMgv0
zWltSt%<PVs3}gzBHb6`Zbr?#03FR}arpLn!u<3V>FUarKaLbdWh?1Jm_8}$5G9}Jm
zMk(O#1`|@uk0ahrO9sl0%6aE^18>)#ws`fN&VTtrw&f_rZ3?2&(PRW64IzGms}h<6
zj7sZZO*np&uT=ber&}ugGHBCoLnvulVUE187V0NUXzTURkm(7P+$Q-cNQg@;Ea=lZ
z>DZ@O*=~-e_pB%Bx7Q$%!)W&BbFEjPU6wi++jNfwMqF&>>YdlGjBf_>ZuzqD9Y#Cr
z1IY+sI4L5l<W(Yiz8uF$tj=(_4{xvucti{SVMdHY8SMZooS*Mz6`0hCVJ#|<91*)v
z)<JlJh!0oHZ9xK?`<!M>qbox%qy%v=u+JdofDwIWe?d-Jr#d<z!0F^W0ab7f+BdKL
zOCO|~qM_r!L@lru#ofK@=TT)Pw2sa=yl-BYZFf@u{AjN7l#&uGm&z!`F1|_DUo4I(
z_^Iy_@Po5XU%AR`vUDeWmgsXKB;B9MP2!HY>_0x4qHmes2$H~Z*6Y%*&*$mSeD!?o
zdh___#R!RLOsR;9CjIJmqs;ROr5w&Zz<i)rp-TNYP(lXsUc$6AO=A&0wg6W2(0cGl
zB$!bWnRZ418lSssl|%;vpqsb)l}>POy!~B@_ILYxu3aCR%Ct_Of<7+Tgf*p}95;pA
zB0}tMvTD#!Jn$sg%@x~fM#89bOksPP1-`%?`;b9NI*G^teUC;dmJyj6mwp3gbiiG~
z8rvO!TQZ%c*xe)+P-VA@>@zE7VjmunFu{dPM*_DsG$l$MK<_pIFKH>$U?A%Z*-EiU
z{jT?t51NR8<qX_bjP=7OMw788_z+H;WHgC?ffm(%5d(%x{&%r7D1z5J08>QfA*<yH
z;7s}d^~B=m_L5I8qT*AoT|mGVdLX@xIE96RP-6;=%6UOr-Kh{Z6}4qE>xGh(3Dre%
zfh2|#t&NZ}ul}`HZ+?Hr56rBZzGIb>2NRKYrnQeR*R$>fn+SJQ#OeuqvU$WpoZ$As
zpi}j=>aW1kw@g~^!fP5jQH;fO%0;$-un=O;@++_uXPiLJ$&|m~@ttM-i0V2+$shmZ
zZH<>l8LzX>D$^;E*P^w~h^e+qe<GZoGyZVHf`#5#)qfszT?XJ_Ai$moIrf=)<T?0E
zlSA6RT=Ju5YK?h62=fYd#4`M>IX&PQMo);EGdQpq5lhS1{(mk&s#Dq5u5RbWai(tP
z=Mfblb#c9yNA}v5brRTzMf6s52Sh)4+JanpiZ5j5*p!r_%i1U_%3z~al`gPQkCzjx
z^Ci;}a#t_{H9FpJ%(Ej8tKeCAwA%>;iA68AuSRTR6dryGplsoXv648;_4eCA8ntcv
zoe3uF|5Yu8<Z@MgmAG9drsW!4^c!g!kGH!7AlqR#se`*0z}6Oe-x;WE0RyBRy&v!Y
zhq^1p_y%=;EjPxw9F%RxAGYw6f;EA}<BOr~vku0kGu{$!DSCT^MO4~E-Evhyf{pCz
z%r&L0_z|o5a5GFeZB_>WQBgun8clji^-WFmnh`uTaq2S`;su=bQ_cF!vgi+JVFz3f
zOJw=W@^ivdSaMGOx$ESfJ}(6eic~yJV3Z~b3Q~lDi-o3GNwwRyka@Y?$Q4drE?h-Y
z0g{R_??G52kHnM6;>xtgW=|p0LK)<A+*RO2Q;+bW;s@4CekXGWdqaYWg5mgn@hbbF
zxot3>%WmHwVvwk@e(8WYvuVuJ={x`c=2WI<u%gi3t4$TfAO1%+y(U0dN~cNl<u;<a
ztomgoj{_?99<1YC7Lz4B@iTD<grJ5YHHQzCGYZ*_=gNh+Pb+q}ek$;rCoEDGN`mF3
z>D=p^C7kp{@#DXRX?6p-c*kYiH!rM)PNaF`V>9Da=Q~qPzs)uzz{`Z%iB$OUWYham
z;B(_H65W?Q*Xk_dCi1u5Ni78{?LoyZEzLHtzIpuuu!o3^%>KgM)|;T_X{{yKN<3f%
zI?4=^w;x}M_jB5O0G6o!g&PuFYkD`;Bm|EXEX@UASMMjH75|+CALW{K+r}BL&ZGxJ
zByj>de!)`gKev(`yiPkW54iN@K3`&b#OZ!1LW(m2Bn^7r4|wY7U)3FZsqV0d+;0ZJ
zsbK0pz3$LDVDoaVY0lxtrQlBz8WG(TH^b{fdvQ9ojY!!WGmHZG1vyy=nHNmm_pIBR
z)WIUrRk}r<F~!`dt-}`rfHT3L@v7t0i+YC~QKp0uK$3Gb1uAO03(Iz?*duo?313{s
zGFf}Ek^35^RNoK>rE&LZRhK?KBvOC)?|;b!|GIX!k8b>)+P5duUoIo~{2$H-A^|%q
ziMbtB&0YYPB0fGF8<DwP&DimW-ZWpo8MW0gvFy!OAAC&^QHV`M9<PBXZ?qeRapdb{
z;c#*gp!&V{0<cK|X#rdg{MrKcDXh3=ZrTbrG(Fy;g-C!W%;F!0btn!+CP#q?9wSJU
zg|uMQO!NppnB%&5@a1eGZo%9p+vR$Z<Vgrwl5tMi6l0xS3JrDy2{J=^76+&(hZE)t
zVTHBQ_(gr@HUwtrUgl_udP^;ddC5O%BMk)ekG5DieTk@0d*sXl!zczmm!ZjYnk01c
zIBBDzBe9TH_QL5ekYJqh8PsvfhO%xCr2*z?q^&BJ(9K#IRMTrEt+)%^ELZ`taAn<p
zhSNB95UQ9&3EL2kzg-7}5CLvH_sQ~3X0Yi*u}n7HZO`WVFS~$CBcSNSnc;f+y9C{8
z3z!Yl@Sz*6VH&Z|!iG0EK8_S|5wDW$->fp4fe2%MF0SeuVSk4&<Tm(k9trrC(l{s{
z@QLbi>3jKiYQi@KdF}bsmK$-$p-lZEK3|N&ZvaaIFGPY7)dm0d%aI4KkQI;9_HAb=
zBOI%RE118y4Y25?6UHtJmmK$1^MXu(gv=&2pD9-AIYg1oA$&x2JBS7bX=H$#9U%Dd
zVKd{Ix}f8dY-{HoCPa-}_*u9=l#VMu9O~2+Y>DFqgWE;F3DplZ_q@M=*Kqeet3!If
zijwDS3~N|`g%OZ&w>kVe>_8m%dz&2wC>s9}*nQZ>vJP#2nK?_2)PSR~8lnk8rXYOC
z^Bg0seNW@A<R^7D4N$k(FrfI#M^SX<3(1~)@JhwYYv^}!fCTj2U(@#D)<ZK*1<e@-
zoBnRVCmb?Va_z`5{9TM>6YWyIRUzF$a~PrCUq4^#d;%h%jg3RR96i;PPi{6;@3qzf
zZ)cm?^p_x~d>dfxoTmFm>t|13iyo&}L&(7ZKs@>nwtq|78&8hfy@*-`@h@afjk|#%
zjOKg?Fyadbq4(YB;<%?Lt04AxivKWT27^SR`NyLi|1GBV62RA-{+nd+2j=s11Rn#W
z_Ic^du~WR<P+Wbn7N=F0kyZ~!w2c@0`fgV=E;4OrWRZ4g(sq{r%dOcfc1`+akP5wa
z@x=<V^fyqsCnPl6+Vq#U&%lR)27&&%ll5G#Lt_F|$k+IMP`W8xslL3ZM!>+z#4Hpe
zZs3~%Z7RL)&E{^`6L7%fn1bro%!G)FVUpM~`4fk+iEp5mV@Kpp+>&l^z6m}O7sDU6
zS+zrRqoYqWI7ZKF)7Q1q<<Km8{4~a2o|6yaxhCM4#@mkJXRR4)&<BGAujEzsWNBi{
zV=>3$q(={d?kgq|y*z7k#S8H|pd_xvD-atqSB4YPViFJ4T1Gg4imm%jIwN*j)I7cI
z?QjVgX#dCLcjOHqD>#Rl?G+0b$#$Fwj+FQN(ZbH^-8T?}=0U7QaVQytQ5f&>G*&Je
zj3sCER12XJh+X12!nmFL{v4SX3{7_Lg=v_hHI(c1UTE}n(?3!$^vcmdydFzJpe`|i
zTZy86GjFNMee3rNkfld7Fj_gHe_X6RAUeq+4-DKKK*t@Wjwk<u^~`HOi|$5t&SotM
zp6Tj}^n~ot&Fezg9=fMZsNa@ZBYRhHO1KQMQd5Dq*Zp@BihzY6^^2=SG0P2@v}AEv
z2!5Bp_R7IViAjziZK5A(NsdR;4Bp*($ZfVA8FUq2kq`qHLMm!qqb@x75+qNE8L|5<
zh5s}=B%|5(Kg$sSy-(m%Re(c9yw)grbeatD8siW?&M8>dq)Q<lA$k&{MBzh1Ht;(M
zCBw%$H<0gL0kln1S>DvjOg?}cK`z;g*fNjuXwNAfeG*%=#pDlPSWejRVujvo=mrF_
zJcyvqsCc8>hF?eE^sljJ-+D}P7wF=Yy9E!%0FnLO=SSMo%cRQ-NdG^lKZrJMs}h~K
zu^hkK(eC$u{y(r{0K1`PLLS*G))zj7H;X~EV@J_@h@`gFr(}N;QLqxsh{H3pbuttL
znVj9WY#38TA8!EVz-{7+o506%wEm8kcZ-T?6buU$b0BKk(r$!Jfe`S`U?;Iy9XpZX
zBA_8sv+mqy7PakPUeH8Rqs&*wq-vBiMJRb~rGSxBmJ9Z{Ew4PzxN=1njAC_W+3%9^
zJPTgprAJ&h8Mxl7XAHBwhTHpg{{VOUVk;2k{6k;qMsNa5g8MD%swJPnXv3;WBGS)8
z`7)>1l-%LsTkx9TI?-y<zeZPsd|!kGFil&^TX#z>lH6pX#93*o{4avOvVrqO4+IGd
z!?vy_v?I6=y9K<8U<bY|hrG5Zj2xtlH0<K434Hm=sMGNS*o+Z+lyw#UMfn5wo6rIv
zp?qqNNRT)vm-k0sM&Z~okRO6Zhv+s4Sx;+7NJfDtKb^w`@<g;N`L?+gf*&e_UW*4%
z4T;LV7k_A(Qt%pJ3=f38BEMNLEFYm?i7fEK&X?AW_u6rOZ3VQuQG;_Q?=XnxTt@RS
zy8B8Q8xq<OoGurPJ=H3|%?<y3xgo=r{yDJ>Yqk>zJAOWN;|-!l|J2#h6&-^JbJI4O
z`KLJR@vPqKUF;4w!T!~`Lv%<}y;k3Udh30DD|>5;(8d`4j_S8C@weD9ryRdlKvMZ@
z+_;rg%hw0O6(?|bM5&m0XhHv0qV8!$6^84pX`AFWij`OM_M@Dqyy_3EtrTIjs;Ix^
zZR;dp#U`_BK3Jx~5ZAcn<cThSkJEIHPwYIf^NjR-UoZfMWU*4Yb4I<fcHmr%8SO*d
z*R1NU+qLN_zV3p01Rv1N&X?h*y$)yb%95B#blIg81|no0n;D=!p950gUGKHbZxG$g
z8jLZ|@ST~V()4im%k4PRJY!kL1;B^6`tculv3(+%&D!tN;Cco;w71>1vlrANV{ix@
zZ>~V4Wv71B#3a2g_R;d|$<hjOmq5s`agp}wH7ojvkXEX0AO=c?rBTAAv|^{Qd6#Ly
zWu|qD-4NRIO3SqX%G*BL?Qhu0;{tIAn1Z^gcO~Y0Z434~pEnnwtof68jAb)m4Fl>`
zPF*-2IO<p-jWPv4ljmM##oEH)jz%?zARE4sSJ|$qOLvG@bXF?~!t8BulAtia7O+Z8
z!9m&a`HeOl*uXl?E@0|!KnxWR!+U*A=O_>*br;yqh9idf`x_#vKpGRaIP3bpiRf&G
zOz_#bEE9HMr=1+wloc%D04mHRwsWL6O(goA3@1Ou`4s^iy=YJTC@!mAn`s<8ezMxW
z8<?sBB%Ww>l%1?XLn6F9pb-t!?y?LJMk2q{9JCq$eJjS*6Nq>NY;6u<1{S_@r(E-2
z+wHR+439mG;SR9$I?xLkeZz)fjWqFd?`eUqkg;w*<1@NNX>9s$o6R>(y~x~4-iUO=
zP=xtCE}6^hYU2a(_4Qz8Bk;qB*u%D+g1pCH^<!=if57tQWUK8C0dHiWJiSF){_Jxi
zE{%}aEW5>8*@!TS)wT!R;Y0x{!QoiBHiU1WyR~}lcGpzmWIXg`*cww&N=;UR&Y=4+
zKJ{`?KoEHGRP(R#DwH|sNX1&fcd6)LC&Yr1QNODX<;)Xa|Cx36O?}kB?2^zJJV=&@
z2YeS7!Iwk?k&OD*BGJJrPbp$k!w<_>VB6Py$xNvG9vA@#(8m#?m>&Zb5LNpvR<`>G
zP6?78U{L7liq%6GDg<fhrqA+t36kUIuM!=|ct{7VsDNDxZ;J%(V{vIdy~OC9jyELX
zhG-Tq|FOwZyM0{8=(jSyzT6o87j3_{4-dx?2w7aOx*Sb>ucqoQR)&X_Q5ALe<;g-B
zofud)0Wqf_^|Hg71e1zEdARE<c^nMR;#$qssKm-3v4Cg92In7;oS;kdkCgh|#d7((
zU3%&^`xVBJA74=W*6DpIkUIw-gX(NWro)?;VV8YiBd0NXAtRB8PU|oM)2T2^x8y$z
zY@cYKejN+__4xCk1^3C#S4#oAW7g!$+`<`|QGJlHs7c|`o@yER8k@*IX?dmjYKYdE
zDHa3)oq(i(N`zvX4{0**%?eHm;Pf*KtYX+`AUIH!!N%N^>m*h$!SLMO8z2fz<RHfJ
zSiCBhTCC#$_RbU|k^-`H8T<bhAOOqr>{f3vq^0zb$zYkA8i|PhJ+JV%91bPk)HyN#
zHGIc3u?qpuIi5O~8I~I%70G6};h>9|aA*RAr*0QsA_a-uJ@}e2a(4tInFoPlvZeAH
zg;@N5v_P>MV^pz)I2_#^A1#*iDKtHc<vW~0u$@cKC$nar&KpTwdj9WkPzWzWqByJL
zUuz_!(I|&HZ(K7TV|$4j0Q!@FfbqgVONOrPXJx7zZv;#zfBLQtz+VB-7KlyZlg|hg
z;twQ0zqwy*{R2z}ij}-!7HV+cpWmD!wZ!8XRq^r}=c!!lU89*R&Y%r`&c^k0LRVvQ
z^94~)!o;aI9N)#t6>t4Miz9+OxCIE^fLwv=rcrE4+?WOW*mrR{1(yoDN7@yX?}WN$
zr_~CVTI)uJPY5sW33FR;07AwGN2HGz!;bZf5$-1Xg*30zZCnuWQ|yp!^P$<euZNg=
zo2=;XrdNyIQ?@S@vb7>If%A)TgzjIg;EJjtOVGZD|Lut;;KJ~|9y0yNMfShK)mL^a
z<%+!w*g0CAoNWm7sq_}mgXJV3w9ebb#?-kr^)TUHzad-v^#?0p!Asb#?I8D$oUqw7
z5C;K^o^pt5>=oG=c3+Ix`c24ybDs{8y$zPZJLLEIB%{H}oG{ieri>`#);>ai@oc?a
zZ!DIgnF=F&b*iSl<6bT+GNBn_*st3xeTGM?SPF(a9uCo8#QuHOKM@#wBiHTrkCH}%
zAQc&a4T=gPrG@UvjHHE+A{G((LS7aCUrJpn9T%QCMFRLILeB89xi7#?ZT#lr{SLkL
zhn6}@vE-)H^~&d}f6)1&s_IXb+NQp~_to}?oev2;?0qVDcMe!)lOtD4m`OZ+4@5t>
za?I~k!e5~KYRU3hkss`k(qs-`atCF91W$i|{_fR;2}9#4DTuEK+qXJ-gMtc8DK@t+
z*U!*h6OJM3@q%E`ET4z4;NH_z;)=lu=%*5Hn#g071LMie)GrWogc~3Ue~K^!QHC=+
zyuS9}!)c2$nmt`&q)oE^w%`fpcmml=*z-D(boY=Crd6P9w(1DXXj(admf?9;-A?N@
z^;A!dY#iA9L0{S8(*u}cj%V{h$)Um+T1@aH$7+=Lxvs=G<mI>lydtlNHsNof%!%AP
zW`zpq{&eV`<3Jj55N_*-*&$F>RlN?DdW1%V$M^H7!&Q*<bJ#>hnC7gcJR_-*S1@8s
zD3bgxZ#D2vy%3Q6l}Hgjm^F4P;<ij(g|?BC?G9Suvp1`!_gJQ#Q;<Tl)PuB%vYh4{
z@4K=^s%dl1XVwOPNTpcjejx|satWMuC9;H9(t+r0QvbrFa61_uIL@ur4Zq?})JCG^
z%!tym`WHLA?m#dkh@{LcSU8Rhv$|0iGG7lhLq{J>ZnOs8U*Tw2&T!1+8H1Po<N*Qq
zn`-hX6~4%0zjz^bJk5S9ne^bwHZQKPyL1j03==P7y`G)mNq44DUS2MW%7G?ehAdPE
z;QAd=M!q5k)nb4p_Pny+dUPbZhzi|z<mK^jaw^VkMte)LyiG8X<*W5B8f~#oj%!v6
z`hUEA)4lpF$;L+f1KM2OK9B98x$+5a?a1%*iO53^lE?f$EYJ95A%I20xtONa`Y$Vd
zQXuG-yi&1RT_JLPUzVPqI%7jM0Rg9m56#K9L5;UR0enUl)F=XT+;|U<5eoi?#TdF0
zPSwgI=FGV$-L)k26|}d{@s3->%XXT;p1uP=_v6MlDIi|mg4_@BRVB9nE~YKr`xTl|
zFM<VlOkT+@E;hPqayFf%Wo>Ou9r<-Fh4-_2yWyb{PY8sE+nuQy8UCLi!FXKttZY>!
zB^LSJ3pQ4xqck#_Rg~o9L&N$pv5~Q@s7YxF*e)c8KKm~b*Y$rb`ld4l#r`4ixxVg4
zCK;FXDTZVtBEusilYaySi3N!%g=KTQvY!7~v0AtW-ojQg9g+hV4=aS{$;f)@b02z~
zwvx88veMGB@*I7W$~Z(tt|u!k-I1A|p7P_s+Sysw+FD0a@-P;U%cag?vnwkj)&1(@
z9R>v<noxkzPe(_?Kznm(ZEIy^V`FY<i{|nK4@?9OpQa^NkZqsrkGf6B`Gj(k)i+7=
zbUt_IT9Mce#-Aea-T|9Xh<Ed`>*)4s!po!_x`)!16kt?5>=nTxAa*K<N=m^H*@Bx%
z(kc`?JK_sW?YUKo(eZZo2kYAwinzJJXRJy_Zsrk}U@7xMq$P^P^T2P=w7d@du=b%#
zs+L|eceeLVzUjm#7rLh{<mU?~*;}ABnnRKi7wF3Jgl#9>XWa#iwQSaQ3ORJz1g${K
z2Z2pyzYP+r0O1W(@T_S`B>%<-Bs&8Ak=ZD2S?e?LzZRspxOk!1m^ZMsV6Yq{Y0_{O
z!q0qhm@dWIIDR9)0tVAX#A2DE@?BzStM(#$i~k(V=v%#!1yxh1m69A1Y0V^bhDq5B
zR@Nl^`en;GKH!$8ptKE^)B~lthemmCv6w4ADyu%Ui2CKh7!N*#1LvhFziNL*Rd@?)
zf9K$HF#tX;U^B7cCoR}ta>laKKRj;xS~}M0!b@DCXtp7!(;c&XaaG{3<do;!fZo+5
zc`xxE2@DH5V9A9^TZdq0g9YC33f<k|{+d0*?6#p0@~PW?Qm0;<vbNT#1rppq_*JGZ
zgi@`|$$`yhgbJcYg=6X1=7+Ujbaw7_fQsv0ZU6*z=~f1zI@5tP9x^tI@fP^o8{!x`
zli~aI!FX!-g}uMI(dWE{k>tTgjxP+~+IlJ<&*E}4bfNRF6*aig*tBB>jSsfWa!m38
zXg=4TT*}ZhRJ%xVpNtAchR~=i?b!Q<1fFIZ0pT}i2_4aUbnmfay3{1EfB#;Bf;@cl
z(qv;_=9S^Pmq6-F22T;F6oL80Ats8WJ5Y1MWv1{&ewNtUO5_5*7g@W9DV&L=wRLsH
zEiL=K!F?SaZ??ITTQYKT-RZxaPfSGB??WT5BW4n2_J+Zc3HTfQuLS)KKlRt$FYi~o
zJd;w7UXd<smRj2?3>tk7<|q`SFV1bT*0iRxEX?eUS=JY9Q${C<(gS(S4-qO3*BBeL
zO|EMCtWHjor_MH(Y;RYN>&-H*ru4m^&Ut*Um3waXNP0<16tj5=+<0&GOl{>i4hHDT
z%B;BR>s(x}Qc_Y<f2C-CTa3cZNTi&8DD2J3|55?Dib=`@X-7eoy?-;gO|0GrTIM3h
zBWq4;@h3?`(DUR+_cSS&O1T?hXN~Y{(IWZeKbu0l7I!Qst-w;X23sXOqcQ)ham8R>
zFJD!pH;)%~<2AP~q@}yvQz4&nUC2ycK5L3Ja0_<LL%dy7UePf2j7qV0g6_VDMgneK
z6qLBlzL{3is>3D#%j;oo1Pz}(A;zC7Mnlp<TT0yZ>sw{0ZL}@pHe3##W)2px&<cob
z8&))$P}$B29x)&i5G60N8~a8wKKXWr2}ECsG^nhdU5>@k!#TYB=VWb#av5Ly0AJG!
zI+hT|2jkA6^bATMBEiJAT)8CQ`2|sjdo?lJT^r`E{WYIN`H;W#i7;wi0?zlxf+g#4
zU<nO+0CYWWC1hVbC@c<q;!ac(1vV#}-GX==b~*+Lag16(Swwf8?`o4SHx>ltnq=K1
z4<cQG`oP5Ar#ki<MW@xlbl>?BXeB#J(#!ML#T`qCA#2A7;)xCjg$ADji;In>RMN#i
z#2xwekU~Y4r#%o8ttDf)x**}tmz7+{qj59hi4a;`sEBk)9ewxVs1#gBy1+`$e?aS)
z=F3;T-IU}=NlJZMoI$e{eCkg1#EiW+cQEupbMW%Exz;y_mRa2{ns)4tCj)D}om-8~
zUz10Z!;BN?brcjpyQIuKtC*vOQbGLV@M<vGE|pVY&OP8iC@$`pG_5Sx5jIGaP#+Te
zr6$aO<8Sz*Y8J!2!L#}6l!3}(4vG4+P{z<vVZj0}ttZ8&>=F!Nh##q>n?TfZ7T8Oc
z4ZiQel@|oAjrntPb$xyL8y()-+N)(&R+@~4it^I@?wq#^qcyobJpje^;%2*Z-Xgp?
zbs75Up^C9wlTo*8V5`^dZ=Yw1%Y=gYXZK=ApdgQXcIPB1RrnxXDx-^w(`t=<R_f7E
zWk#p_>+>_<@Hsr~zkMkH^K^m2@AhUYZRxRc62WF+lFy_U6CE8ND<{*Io}KQ+@S>yW
z=C+l^x2moE(%R~NcYAxek16Q$ybt6>DkWMvTI)aM<+OA*Hr9XW%5lZ>mr|C1OB46_
zI2DKdw!E@J{zcBKt^MD>SLj!m!r<Ut##(t=3IBIr-*@1rvvU(IvZcJdwG58S7L}u+
zsezM=gCjz`agpO&a`NK0Z@_Ev0S1QIfmu;e5e0=(baXT!CME_Z=0<x?PKGxf-_hX%
z9*&jO@=vv&jD~TMadDARf>=(Rz{<IV+}u52m&lsGQGCv4)x|oOxBdMO!|oxohq0n@
z{EKrehMH^};}?H_go+6U+e2Nxwbg0>5u3E;ntmYE?QDL2d4<O3Bg97sct)h0l#Dbd
zrw<but)4HT67geWbf*#u7g;9apv75=vZ}H@<J(~+8q65dMNoSlrX@mzn=aG~l#@$A
z-;dtYgv9ZPr9h97Au&xOPL4p5UXO-(=hgxZfll9xnJ-ap-+yUr&&*f?NoRH{J31D}
z>EUlQJQ`T^T;UZ0mq;siV&qpWpl`Lj)_HPNP&F!I6niEMc_F9!{+m5!SCj;tl3%S{
z1d_7+B&3nZ2;IoCrh?;3GB0#O@SE^PHt!Ola-N2T<<+^YtgNDpOtQ!mCA|CvUIHg)
z^DR5%0?ClP5VJ2qbP%NTPo?Kw4@Qw})lv~>2aE|>g%pHTcl&K)&hh<`UNXZkbfCZU
zd(*Uqo_J`m(jxuyr(x2NKuxdqhuubRleBNTZpc4TNVcr_n&9{K7&4qbq|rUAS3atx
zzfQjfEr&I{Y4!dZSi&B;*)IXP^5yv^E>nVZd@L*r&T-T7CrtG?lloS58|gakfq9;?
z@`42Os;UuVBTdd+vaBdL(((tSv4Fk_yL%B^Q+N!cS45$&iRmedw+}z{Rd@;R;u5eu
zgj0qt(3X4J#oS=2Ndjd=BkKxk87Y~4Abs2||K78P>L8HZ@7G6FAT~6At5Wa&I9Q&=
z8PrW<WmS{gL7XVv0H->=(z;kg5n5V`j*24Odx)5uO``buW)ZOg9<W;K#EQ6q3RmI9
zvDPqa9VGT#c18pVPjl-LkfzG9h;riI5+v%WZumib+QIh%G%(tm%BXq+&6Udvs3vUl
z_BIdD?=v-9%!oc$yq%(zCn@)w9L*`3>Vow2ZN7Y5WMpg%90GzLCi|@;HEJcn!9nqa
zf_20G&C@A%*=GWJ`Wj5`I&CY+1G+lD{X2h}TUnb6IBc2d@*N$}Z7f$jY&BT@TdMSa
ze@D3xLUp{?k~W%Yy^d%}ODh~tWt*R`wlLqC9H;ZS&V9Z~qKYTviV)8tAb@y(XDyR2
zU;M8!AAH*Wyx)swGVJjB_joZJPb4KJwcX_#owGBYBa}2#*6i>(P_p&GLlYkplbaT*
zlp_fI5fdBh`}Pdy8^}C*)b76b!ccp?Z0Mi2#U+0r^Z#4`M@`eu<Ar}HMAsO@Lqn31
zk|Q@ABw?>dIeVRXuS7zMKchcx4v|*jX!ckVLXSm{!)L~T|J&;PK8nT4%lrDbKRmVC
zpwsXDbubzW_$bR(jqt4<)y>WI1qEfbH&#wgH+xi1w^!!d9i9w42dX|k`>XX-Vxqv~
z>sViQ$aum;aerDY&&<NdWXk07*GU79UM;>m($Y>(pJZv)I%_kJPHoPq-u^2Vt`w3T
z!H@OZk%mM-fQ{?vGKosb<_xbU3fW~-p&uvimnTo*y~K`1wgBetFsVB&y-lCGW|(Rv
zWzrIMWjt#q#3d2K%&w6jxn#xRf}9T&8;?9iPd+pjR;ul4hgoF?UJybV4a3^RCX<Ip
z_fDW}5s83`s2@U&pXiqU#~6faK=|@$*y}40kvB6OX#z3i2eu<|_v2VuHDxitP?BOS
zd#vb<LMG(u{b+KVF^SV!D5FxYGglcj;nXFDPnKFvYf@0aYXtt4h*T+$9{g#=OsOi@
z`}0T3=g0h<4v6oUM!X?)_6*gM)2^tvCUmxNuN8umZsn9RlQqyy#2~uuET*MTA}^c5
z<7Q<)lm#Cjkfd{p+El3ri^n(gb%Cr@X1~+9L<1>i1bwW*T6x+6rjaD}mq47a$rojt
z#)?C4_CUR95?qMFwd@}n5Vz7sabK*dRlxN1qt(?2V-w#xf!=<-$u?Hd9|U?&i=x)X
z8Mo98MM=zEtWP61N<+*E#%+|~zt$n(lfiozU6WCA6%(7!gMg{8cSt;pJWP26N;;ro
zs;dVR;#nptg31sZaV)mzG@GHAdbtse;-ZVk9m^lKj#2x=@z-LiLbo2%JSjP7gV)P(
z81x!aOHVw_qY$WRRqFr#$mi0}tj5X)VaNr+M*Flpx$fV(SfZ5kBaDWKwq<SZDKQ}l
zg7Ox`1K1sFCe+4s@$Ai&CnL^qNBPa-g@I|vj@BVT`8gmPLMjonE1_~sH4R!zCesbk
zv{TL72k$tdPr<9INqhY}6-OGRDDw1tv9iK}OGP7{mdHjFa5}%9?fmd|yIiSt3JxD1
zOGa*byrVyoU#^84HNxWNp*QFpTkYTUJu|<0_kM(ig3LsSkBEwqSCnl_^c1+7%KG<o
znVj-#VR50Uslg4{1mo-L)%nrx(EabFB*&o3E0kR~_qo$QzVeUscK<X^neoq8l1kjN
zYQN`e&{G)iFhGY!`Thd2FnU9xkqz9mk5?PLf{(q)mZ1-G(H`6fzO)VJA3xAf`VDuw
zYinzPi(XLP?Jp$J3%#jGefpeVC;{dgq2x<^@8XOm=u3D-v^>y<*B23hkSIKba)Iti
za=SofFPq$~#sF?g$N$~*Z#W8(-{(d1(RB97kG8;%z^0|8(?5Um`})>-npM@R28V~T
z1u-xk?^o;%g{~AbmRc&@oV7Z7tg8|%=1iJS`d$z+G5Okf+dVw(Z`TjJfyV~~$nW1x
zlGLiqjyC+XwA59;&97|*HjC>;t_F-EzA}I?_1S(urxQs8wGZ|+WZqFXBri51z{P@<
zidW(+2i{QpvQ7X;C4&F*%et}Wodw?dn^mcuE*Sj#ec6p^h-ek7eRf)L65Fj_;VOO7
zgUxao=rdnAjZR)l-0<ZE2F~}~tN)s9sgucHmqca+rjXSrWKp5G1Da$~CvObV4ve^Q
zKSUEh=zHL3nnw78^7jcu6@9g4&w|v+Hz-;TiFCpc?q==k6)xgMys#f&#FJL;JJ!^p
zdC&XWJ3jN^U!k=}XvO}pO<6l*IA+h!)VB~Bkyw-xJ=3Ep_kAhuCpxIxowWr@Xy?1z
zynK$)`!Irdovdsd^?fj~>kJR*8JaS^!6+h1JXz4MbYwh9j9jbWE(FmMWn`gjO^BBq
zt8N_#tyOLZ5DzWS58r^Bvp@#1x%uUp8F3AxWR3-oPdyFj4fw?H`go!?nHmt((!yPa
zwo0?G@shF>G7)En+3rOnjv)fv0me5fF{KfVow%9e&<eO#wD1wESpH~J1}R>a`$Cl-
z%>k2*z#J6E>ff4An~mePDKgHg@I^&TKWUK$V<x2N5G!+zKA#`sg}4>o_Nn|Nvwwp=
zaMDy-UyhEhTpuShxPW5M|MRKaDvd6|!`$SI2aFUNB$r@pnO647`n`ihm+aqE@*K{-
za(@`Cb<JUZB?^iP=W-W*-QG+pf5(n<qbN38gG88iE<3L9IUN6sUF8MvkwZfzxa>FN
zrCad7_;Kz$-)#DQY#A8X=x?U*F*NF1emAox5!V;V9FoK0t1@gyKAGqZ3_R*1Gs!QQ
zS5tFw*!EOcS2xh0M2j5I*AIpW)`^-syZ_G~#JkkagyDy+ZOno*kT_-B8Os=dS~Iv&
z7D!zn{q3OWY$PKiGbKb&b$v^k<Pex$FLcQ`zvvG~DI9*D%xDJWYcJ0?dw^cu4=|1t
z`gq<0Hn(~{oGx;GK1<4}KzEABS-I+DrQ{c?p6v7p8#*}X+UROpSXf9&xu~g0d3yt|
zp4y&v3JSpQ=zPgnQ8=_QS-JA^KR|&#CO0lCCmTg1<fpE6e?OY^vqWiJZXArYqobp^
z4fk7BWmy?DQ+0Kvjjg4nm9?ax4<wU=gU#9bF{*py7e;;3OO25O+(w=1Wg-8^b0E%m
zzuX>73%%b1;Njcr8DO^kvJ8|9R`zzqP4U}~x6pq@$=`VXwv{Bu$HmRP4r7ey#vmm2
z{E014+qWC6?DF}z{dd*<xkvOlW7P-*F9wB4iO^7>w*YGRd{Q!qc!KS{qpMD<r1JVZ
zpuX96-0?g=O8d>XvR3Kjy53=x)Ls2qTQIB?o)s4ztE3<|J=v+Nr@A(?AWC+oS|$>w
zvk!d);@w?-q5CDAO<`82D}4_6Vvk6ykZZigyl3^oFzdRAYp0>{2sCP!Sp1N`Z6Cyl
z&(*4I&yL~wTzR>k80V2;jiWMc2wq&IwTmJ7kyVZgSoRpD*TF`%o>w|BWMhR5aXcAv
zHx)`@sLMD`UMX$nOluw$=i|UkvXR7I^G?j{rIwRN(LFL762YE0_964^;#4@&Vly{e
z70zJx1=1r@G^s)G!WP(VKUhdODK#QU#UClv0w7abAb+Oa))xFt7bb&hZCT+EMt{zo
zhxnsXC$k-JZ)`kUmSo%)_=*b;x&4!EC^;h9jl`!^UG#$BkYh%9KEhv8U<4gL%WuSU
z@J|+>+X5H+u6-!%SI<@vKXK-BIJ%hGrkXCob%cG|<E+ES;az0yMoa96NzamBv^!E}
zm+i_qmS)69{>rq|6CHxX6K~J8U_SbcAk{9M5asfG!rbqkKc55Uz~k7Lra-@O_D)D`
zNbOtW6mOEpWi~tC_i55&O4nXt%p@e@sZv=vcHsLsXeyDVqIX2;iN$)zlOZ}b_t!z5
zZYNjZ*2w4eY8VbhI{G>~lqV<i0sI#9r@W@w8JQFdD9;$ovEUN>I$VeXd#ge_R}V4F
zY!Lix%mvPO@>APq3&63fvtD%uhGXu|ch|S<aDySH9T8079O%mMthP$_u#PC~v~nX<
zc~TkVIO~^DP5FQQJfze1iR&51O@`%<q2%FwAz~F629;zVFIK&8b_Wq=u|$K!9$&9k
zf#Y>Xi{%ZH4Q(Agmq4gU<zsjNnns&{j}ZSJ!?XBtyh$i|XlR7Eni~oWTg%ECizEs(
z{QpfhR~vSJTs+C_pw98%#8QdO7E(RVmna!_fmLw4hIjUU|Ng!3Fzko3x`nCfK15(N
zC59`S;QPTi*7-dl4$jNV(ei&mB>JkGh!FSm;so#b3d=omT}SuK+6HK@f%{8@B|zKa
zF7(fQY4VIqwTjX(Djr41_w~L^jYLFb7@26b+oJlDahqV>r-RE*w05j&YHEs3qsmd|
zv|+`mwlWJyJ^vLM2zL_&Q||&w+7q0Otj$`??%9<OeS=MIiqQmFsnuHTYm5Dx1mMrj
zFBcn)0FOZo1{o3tIt=*Uo{zSA@)OAe0}VFIb@m%0?JiVLk69&ygFur$sh}`&E^5U-
zog?sHY!K@H{Ma<~PqYk*3^zgCmeLD?yk*JFBF$l0qqz+vQQ#Sw38w4z{P#F5^yzc6
z8!B0e>F)YXL!F<Vx+Oc&lZBCyoY`{wauaXaL91<Lc5Wi($ozN4MAPqXql3Bwpc#66
zzCz|E#0J`O5|W$OhYX;uY;=84#gTT7p6g8DlZKok?`OSv?IEl2qVhNm&Z4^?@9l9F
zxF4FIcbZM_r<~UacHM}CaK75l08k<aGY)3%&{~<YeeyrKNRj-74Qg5#qiQ3oaURQ-
zUlCI5d!t}3)e4N2pkR^$PmFZ*F=6=jV>2Eh!RHiYqa(@!M9bM~F1>D6q(RT6P=ytE
zzeXF|Xe`|tC8;vuqdQ2iG5VDX3pzW2Qer$(bEN)_HDcqi1<M&qx?s=@N907}nQz!a
zL=?Oeq#!3F!=uyuB@3hKySgyq5w5i=0>@g2hDfn8PFPqsGXM6<Mw?f1^q}p0(~_G`
zvncWxuy9CD1gp@~QkpE|<JBK_lCnMJV9$5xv>a#1Q9NKF8^tdY)X3$sl9I(q!QZ7F
zVdLlcriukYgpePmy;og21qER4l;d0RoX)ww6c0M3zJS<ad5t*cu}>=1=(UYFmV1|~
z8Y+UwovErv<P#OLCBiT&x0uVkl^TB%cRAI;rOZ-drqSbEBO#f}PLH>1>Z^sHiE++5
zLA$DRQzxJOfR2?MM|Ai9m~1X=2lw~12>FeJ$_mun%?aACx(lA4AMe?l?P6l0iOIG8
z%Fw$iVPSzUZYB)Sy7&*znBjWL>b9XhcNJ{JuL_k@x%opPg?g($J+WJRC6oFC;G6Z5
zH2%gh%6xOZD3DUJ-|kBK>CC`O%X{pRyUn<bbhgCD!+q<bG0DPac-msR_lg9x$70~Y
zz=*35klu1JRaZ1LG`1G$lD=H7UjaEc-v9F-%8&j8%9zIEcp{;nEd7N<KNpf@-e1_>
z!X1Rl0+pO_ziute&d$1DZFT-TV=APj+jlNfrTHR<i-$`n<aJ+@yW=)hIdOka7a56_
zpHHTk0|y7s>wKMi6_=&U<?Y>8?e6MkaN07k<dx@$B@llZqV(U%W^Jy?=9@O}YOH(^
z_0Et08UjV+IJ(S~whEm_^Tzep#!JGpWdIwve_ZRB^geSVBHGq(f4toCcKqxWgPGaR
zD+rH=#1{f)RX(iu1Hb409+8ptPlxn(M~t{6&Z=OUFb*((qJ2dZ5v@*Zd2Wd&&d?M!
zG1*N9CirMytI5g94ca{5OtGS4f$>`Rj~g_5LIT&nZ!e|5yTAOkAnPNeWQ1hI`a(!Z
z=yoW69J!O55U>Xtj>G5saXk~*`R@q`oEKMD&ExLNj|76c&33Ej7YD7QF2TcX95x%7
z`Wg&|ts7QlNx=o3m7AnCXHXh0l@kmfhO-O6mp1@nP(fDbT~spO5~yOp-mn{pv{{I7
zu~9J{nF(pXvlHEqruOosfH+ghP>EKn1;ksxc1`H>em=+lBa*`{+kpqkLFx2ve_l?M
zWN2te+6I{hs60N-3_o8#az52*z5$aCB|?Fz=}85dHZMm0-RnRI-A+zECd1x%tLAJz
zl+%5@H@8=4LX?QOC`Fl8x2Ma_-QAySuh20B6q+Q(F)cp75BMnnK=GS5=%fWqp*e64
z+!lE?Sd<5ena3LTHcII^aI~ssV{Oax{dF=#4vzXOhV&B{c1L(Y0b@uOufD2koXst)
zQZ|3`{o|dHPnuocDFV2KWuU}F`UKjsAX*_O9Y`6T*^tyb%39RU{BcyRPhtXkr%U{3
zf)MOC1>O;UJ*76dz<OBTUkgLyFfea>QHv@pj=`Y;V2CG*;SNi44a_2cdBk*3_Zau~
zDC>}%k)(ZlN41igIfnd=Vc8j|)Qvm<6$1G`G@WHwR$bS%>5}e{R6@E#y1S8<?(S}+
z8$`OhTN(-J?(XiE5ae6;^BwOWe!>Opwdb5;oMSH7!-I{YEDo>mi+5i7T|v7sA7|m(
zP(IePdMT?37m=WKGJdKOLg&nU*Vd^zwu#vBE2+mN{2;4%;pQ%P;UxYB&PmC+?@Y@Z
zKYm1bLapz+_n>wXy_npAj1aEubq2N9Gu-F4j|R8J^V{k80tMR!s_=;O(c;+!TjpZS
zcI-Y7D0=di^MqerUC-S}bH<6drpJ69;e+_}*62)Rt-TF|w(2Prkdo6!598YMg^C$*
zQ<~Cef?P>v6mP^x<JoOIY=$IJ>aS{JYE{?HYljwQ=FSGj>DOQowQEV?+O1|7`$SFz
zX#O5XrQ`@s?qT-n>=Fnc{ip1Sn4kR8xx8okg(|hr9LW4=Q}7Z$<{_wlpht*_Sd797
z@ilAao(=hymZSfK0{?s9IlOI%NLYFqV|sR1pqH|$;dBA;e5;EJ#j^R`UyfLE7*;bi
zs?+UFLz0Dx4eT3+e8*STyZnIcgvF@WmYDho+U$d3sx-C8wP|*ArZUUi_F#za?i^(3
zfd7g%L9wRK!v*n-l@!wse0nyr@=hsf{mSX$uMQw{JvErKmNk}@m4Ua@RZno{*_rk5
zr)gzn^IoB=a}vR~@ob{SYPYZ8CE~TwQKP<hkeUXBt<8-EyR@X)|6H!c>eZbl&%X<{
zenS4d4rfQ#z)J%Y^M~F^iLAKTm<ZXH@A2PVIR2U`m<`-#p6Pf#xsw7-{Xd^eeQ|Cc
zI4C;WtGv9FI~RJbBK_s1Q+A~)eSQy?^VnohJiKBtl@ZAECWtA(Cu(K3Z=?c$GoLd{
z*Qq?+GKyEYDS1e;5Qrq?r#3@O4f(f#$_CjPL4=kl>r*kDFC2=pqmd;1t@>AFTug$z
zr|RGXeE&Z8`!^l1ITH>kCNpEXJ$L5n`XnDN*AJUV7tKe1%99W+wpM8YuMab<KdSOq
z=*fG_Gubm@-K;GBMf<Duy4#^VQA29Q0Wp0kU|A$(eSO>#_O#TJbsiOZ3(%|ahb36Q
ziWC!V|4*pQq1eQj*<k17955B6SU<S@_dJ@rNS8bUObgJ;0E+|h8c#Gh$lZ2x`(Ve2
z-<{-^L4p#((x}tnp4Qx+spi%DA&bwEm620`ZP!a!*wEFLL52a<4juk6<UW+*gxHDg
z_ZO6qKg!Jd2^-zLY`<sGVOHd*H}Ox9VB7zoNTo11LLCoQ7ZHAeGcw+#my-vT?-9O_
zFDj<oDfbFrID7X1-RgH&!0u-`VOkBs&`{yADu^()<@Zz*%uYPrMDUmqioz3$DI%o$
zYDL9Um|;iPG<re;cCZcv-L3o)UX+T4T19!7OqAMbI-nlJz7q~;%SOu<TKH%hBT)jI
zKo~SeY9jiPZXR2ExzGuMGi^b(5RetxVM1&H1l!LoNg658OYE!sA!iFOerniJvVTHN
zyZK6HLFv(KSV5WA-B#7y3UY?d$YMEy8EWmna?7_*LLY;jn;#T(i~l&;4ZGY|&=2MC
zBj^f#!e*m|>Qm|5-0NSoAD<be6%-C~=DQ$2b-;M%cBu=7$U%mpLn#V+GBCDQR`#x$
z{Q1nqmH01MOKbQ5KD70yrzpW0L87WbOe?9Ad=rf}Pz$QKBMv=&vNJmF!(?C@&c4Qn
zK#i#&=KVa-F1L7bESt`k|N95&RP#>j870^$Bz8O!CZE2<E4k3AA5HL|*y{Hkrpx~(
ziOT;bi3Wwy%_8&!p10i$d*Kn8PT}JB4;##g##%|&5ZP$NOIQ|-WNkXoi6YQpY=wtu
zL1AK1#CW?H6(oS4V63V1yE~GeNv<#~QMg>~Ty%a7rTi1cAivB~0fgl$N3HXAW9q5W
zU8VyQ^AtYcm5qgB<qH3o?LnZWC9n(!_*<8Uc><|`;5bscV+tkM!`{Hl=WQEW8>vn&
z(6QBi<KNT4cn034e~ofQ24Tjj^(hGs&c2B$rEZ4{BL&6PlH=RguK&8@R+lu;F;s8%
zmRoA3r)z0xPUILjIb3vv+~O{c>DLaecqQD6tcJX^#amfo&^9q|81mbuXb*`xv#~j|
z#a(8ochxj$URtoL%jzuX1l73@9I;3HW1~S+*-UkAhOezQA*1DAswWtMKC)1y)!yUV
zdFX}4G${`x6QJbYZ|)ynp8oQ`?CJ|c?W>&rra(fJ5+C2<n-N#hAgl1_d~K~AjT9G0
zkfExyybO<{>8JaaQbk0&py{QoABMD6)ilzwKibHiMDx2+9UQVKBTvHND2lWh!XC6;
zDzeTFW=cpS{yP~Q;F!7!5tTGT{FC`y;Lt2Hb&icjD<F=<?+cxkl~`SU(e@m=Q&I#F
zosBkEYT%?CkIVyJrUBznf$UEgzqk9N&1T26b=GP+&gc)FG0rS&LrX(icCEIew6ub(
z*U!xyyuLoM!~#xK^k+u8x?WZ%RcAH<e=RJNM@_V=E}DU0>=^_ZTNu7OJuu9bP&}Dr
z;sG+;uUFfBAeA*o>*}qe16&m?F1z1);fQ>0KJV_5ZUfYDwU}I7-IlvtSAn+x+A->n
zdR=MBU)An3=cdBfdH25OBM@pMzLEcYM3&m8nP*dMgIU8F7m675gkjBXD{6Z6I*`YP
zi)IzGpz$>opZ`;}x?{9EtN?t@uO_y>rwhdOQ>A0Eu3_QS#O}q6cR0)|JXKKgvSDHu
zqJ;!e>ZGJjzLH>(9RxP@-5>dVn4&McJSH*n5H*|z7>7CwGzRkMjgPvmnT=z^pI+p+
zoo;(O-m79mXKzm{ki%kD@ex|dIB--^c>=FO*jE(mlu7)b^FJIk-sWKeX@Zb?XHRU5
zmh{TW?}lh(-dN|@g|mk`oc-&aIvpM3@@5+9>a$l*B1`MJcz6ySNo}}<!ODWnieAhu
zHnp2ZO%m2~mU73)dxIinkjuX;+4@}Zv51W=t$waf>XVs3LuxY5C>7;dpgZ1(OsNf)
zL8CzovO>%VdMH_ZOjv@M(J3mvvI*ETv>~4jKUnDV^)=eTskU(7XF7(iUvWs&HFXLx
zJR&F)NfOj5FnuWc@LM2X?t{0)5yZ+~9~#ovTMg?{aBslsrnq4$!QiGJV9_pq9Wy-N
z?90l`%m@E{m^~Y*oZCUd5x@NmrSd-BZ%T;naJkiKo2W=d+)_*2ZJ`RVSh?M~!~&~d
zHCn+@>n?X={MmGSv*vMjm5_S$j*FE|He2A_!hZ4Cg-O4&7fYU^*>QVm311>$xAzym
za(y8yGbiVtr(3+45-wKOo1Ohp^S(YHhV2*a12JrojdfC7>>C{H<)3?Jcbx{8x&mdt
zPYav({;C#stgF4)->@c|G8=$I4#s-U+p6t1AqMkToyP7Jv#$GI&s#jr@hoee!I8Dz
z-nqMqtV6zF_~FvZ$}`**di}Ps`Q?5<_JPL<%+Uf~k3$htSmqxl`uhj;JJX{vCGl|h
z?JR8UZ0yAKWNwlqlJbZ$VHqSU;3Beke!Vu1=pwH+T7}c6la7UENa4^fy!Lp62oo5j
zgw#1`L+lgd;_F9_y^Ot*b}`BaP%%>$JDy;>bFnvl*H8xv#>D6^@7;m%y}j>X3IRV9
z7Cz*?fP+Qc=EXQs=YJk&oqH-*7#dHI{VP;Ot1NMf!Rw2z$M2#jyvwz=ui7d>V|tsu
z55%ZVWXdG$o7|Rv@;iq<CJ6j=c>TNi=EuoLe}Sp-Lu0Mgm3MomsM{D4P%M?&KQm9|
z8<{!2+L$-{Rkef3aI1SO^&QKi{Tj_DhN>$2-0QaK;sidJ<3S!)+7FG)uugk&(<ta>
z8E0bQa+WdhYSdJE|H<C<Gg?)UO15X1_I8$7PKat@at}kZNKfB$^$9H!r%q^xYQ#^I
z!VKPr$`DT3=QTlPQ-n$;{AU2<sa3FJNB#)<%#af*@dsl4_)Ltvh0wra@fe*Hlkpp)
zQ64S2UC7vX=HRUC8N%P|osfgjBz!bto~;fXjF1y(uWmvrEFUILZVw9p-C=*bF*46a
z6BaR&I$eUGF^05h`Uj4<^)qHdg_68<`nZh<Q=p!;i9PXmx{#XHCRYlv%OEKVq2^cU
zUJd1~z3-n4ki4$+g1@c4r-ShAwmI&B$l`{Ema?V^^YLcyM;1?a5cln)l2AvkLanV_
zj7l%JK0!y$o`bVA&LtMvSTk=FSrBc7f;hY0QV#M$>b^XF<mkp{>ekGiK0y)nN$K)`
z>v_+13v9x4{8N&8_&mgtySpJWrx&t{_s|Mx_MbIbrnFuPhC;6j<R+s+$LgKXtL>M%
zY;{yrSS_z$Iq~4IVSA7Om-hQG%q}K$W?U(K*Y|I9Hycn5xF+{rxD-+D&{8wT+!`Bd
z@Z_oD4df$(b&&!}GiRVQIIm!6X||?{i7tB<zt;tIyUN6Np~s9@0b)a?rYIx_adY>V
zcUHz~KJ0?WVr&TT6DxkyEH(K`E=42_OG~@o-tzK(&KC6l_wUu)N%Egz?>zF{aRNr-
zZA7KS1nV+Pa#B|d8mWNK-}7dpx|ZtVKZe4>W?o)$YIobg@c%W9U1sM-I_)(DkZfEm
zY1WpO*0#GxwI2O9`_}61o_1GjJNp3K8MG68gZ_*5p4W+<o8}`e>+{<3P-*EqmO21;
ze(iFi<e{Ww_yA5x1`kiOlk@Eb4!;+GIG-M+EgccJwizo*t1HW^%FAl1akhV<I=X{*
zvw#WC=XUN>wfQ*>_4yxP)HD{^)fOX31V1s*d-?p@GPL{GnYr4^Rqw>gYOg0Nuh^EJ
z81F$T@J~f$W`5nTJs2t&G@^0DDBLY?Tj0ZG3Gn{(@bqr6*E~Hr=rEdB8KeY|UDe?P
z)`dQxiH!hfXKmX6m^x57H!?W?K0Ooj`0SY+MSuqs%t&Uym1FWyQrDZFo00MBTk7Xl
z|34PMB1edk4y}q1#$o26Z0vLCx8$Uprg{}h1o60)L{Z{+sBlWLSO_E*1SNRBwntjN
z;VTDXdW(Dd2?C!VRJB2Ml<S@O17?}h0t^hhGySqaoPdUg0>lGg`c|J^FTj7(rM4-<
z+)_+1$DGG}8IwBB^}_q*MKSc7US!<l@Q9b1+T6-oz8a05&)e%{js7kkF7Ko=;**N;
zIdK*R^mN^-kBiIQ{plhh$5p%_1LH)S<L=tBvm_enpPRLorG?n}ZWDkQf>~=ly{W6+
z4Hg#eYj~`*;v@hoY8BY~xlrZsRr^{kXT#A<WpIk15n(@7#ad?lp4Wuk(=y`kr?0Fm
z|3pLcZAn3<0`~5bmc!LxGZ2V32#7E0){T|5?+!wSFsn$ELT(9?sb17f=(h9bgaYRX
zY1X-8c<>8rEvcEp&Q!;~vKmDj(?Xc&Y=2ZK#c?SKK3gV3xz|AXLaPx^;LFC!Hg1+i
zUz=_{hlJQmLGz(?7Kd!fJJK2(rQKgiU01i>CSn$J<wH4>l(PE)wJ1+~trRD%D?3hQ
zBCuL&VMEC%IXDynK9-?|fn-qHVc~?nac>ute^ArXxGc<{(BKQ5`li?zi&9D$WY8pz
z2iT-;=RPWu$I*D*Bt$J8MMKKM5P1<CJewI(1%LV-HX<{MM*4XyU6dpS{-JF81SXxx
zgf-rTgPUe)WG%ng4ygeDiD<jGs0U=YKHWv<X=s+^58-`NZ}_R;^7<^Fs$erSr*M47
z9?Ra?n1~OlEvN$V2KNW5NqzM9I`j5^3(xx#ZglACr`^~<$+#%s6zQa=e^Hy?8j%Xv
z##Xf(2~aC0B{ZwI#1G9$a5I0&%fv7<gO~*1Urfo4{YSvC@Vy<Z?tQl&XGp5dX18+u
zKD@wuCszVbU@Z;Dp~Bxi3h!`);qpCy;Gq5C0+a$C*PFQsVqPF>4Ufs#vaGu7&zN@c
zx=CaTIP-tFb(}5$7B{K?BL=Lj@yJC&0)~^UeA^InwKwRwcLx=hj5%3ZMe03w%iive
zlk{pxk1lJf%34$WxhRjVejl^7|Gd0%aBptb`sV6aEMzxD{v~xEIN<NkiFMVhZ`QWz
z+#j+(kjXo0nwZ@6KYrW$K6lDh4>8qt<tVLXDqkf<e{4GM=V<7zDELDw<j0?hfdlmg
z)fT0-Sq)WqHPdWCNB5fP&$hN-C2GEWrDOekaH6f8Raz#y7Bi@-t@q`wh5OA;*lpsx
z@otPRrDovwKn_2V=Z)A>SyIWL%hg}bg9<e<8D!{QX`M0WFFfjdV{DgE%k)?Pso}iE
zt8~zA^UL}Rq#fWJ0MY@&nf%W|OeJrtXTPE>5n;MVM@q2D2Ql?W#KU@;ZKOgWWgiFC
z`r<1)LpV$RM|^0PHibN+VeL5Y>awa|ey>!g6_KY?Fz<iM(j9&_IoDlVr?ax4|Mck>
zkPoO2)ko&{j&79$2ccfvp6oVS5I>gcn~d)JOFXy-&g%jy>Zg~jfYZ^`^3~P%iebPl
z0w*+`FvIopI<OsH_SvK+V)d=Gr4ZCxm0_KTADt=C_;eWs{BGm?k?$TfG`_5ZCgS<U
zGhE}r$w_nUc?j@AJPdAYN86rzQ`<mJ!NSA@xILI{_T$9GRKpU(0y9yWUvv;oAx#|F
zc62E(fuJ=AEXHSexH`M#S>9TQ4If}xiN}%5x@DU5SuSS`g?@x0@OXm&^FQ$q8>qFs
zlVP8vMpeEr=wR01Z34Vi9>qRkf^NTCH6d?bvq<_hk)D%@{Yxtf7Xq6FNKzys?6ea2
z7<Eb2IZA$CEiE2Em`ow>GSN#^=5D|hvThd(6=oXebcj?5_kbVqaWVBi)UeYKU(iW+
zUURf33Bt3A9UB4x|I3GA1qec)pjIoswe%(Lc#bsIna!zvRAO!uQ#&oCs{qNr`S-aC
z(QJ)*)M0czNL4tW#nKt>Zckw)#1meR!oo^iQ<TG5S!wGIo~d{*iQ-9koyGu!O@)T0
z&BMC4ZH|@$i);=RKJpi?kmPhN>;3ynJbbPrh#0SvcNN<X(L}!d`!Y=%%!S2`Jk53u
zL1JWY?}>+E{D}hw`-GUQsj(7Dx1S4h1J~{wAM)FZ89#l}Dl<E5Q?-ULAmPvCb`i<T
zzjBU!z9jwbdwVV@?WB6c%l%<;?0;mhH{8ZAf5ir+L;^b@nJ-y-70`jpau2X_cOsiT
zcXue^gz5I16H$?E4i0a^4;-gc>TzoE2U-(ahbL|pZ-GniSel)kwA0gbKud5uOP#Cm
z%RB5ae2h@us6){AU$jQr^t3W-9LsU>zgxw4h2zl_Cc&30QAKxm;DKJ8^WyF9K*f^W
z0yHn86Z+JQFm84YhFb$~LT77l>*%Co#d^1x|DX3nW^9WG3poQn+o9V<aIa*MX|iTu
zWi`(H9EVif-yzI&-NmW3Y05mD#4rw5ly!a|eo)&#^R}N7r4OsiwYD!xQO`F~M6mRC
zwW-fdQ?htfrMJPcT-HeRSzB0Ho?^3w$<UQm*`HuPOd(wxevyAoDlz>p&9b*PJUXf<
z)!`(-`kDQ1FE3cq?R=?q*th@~ppxG3es3ox>g6aTe(gTcv7~}Tju%cl$-K?Zva+L|
zJKops*h+HmR`rC4%@Ut*AG&Fgr2-El?y(z`s5E9uqk#*IRFt0I{jOpjCGbw1d8)3k
zaLp|6{ym#>v3%At18tgmGz${e{z;@(EGxoDGA&Y~fYRsA!uIC+@`IjWk@yDzjJ-}l
z)On{#6cW0Ws#-0B@!RrrV3p`tG%U>90=pHoY|<%A=H{lrISg<AO*XK%Pp3joeT;9x
z2*sK~XgT!v?s(RH)0c~zSyT0E7BCS)N=pF{Y`3;$*gOSSgHy=tKcSnFOy~R4hS%-R
z18|q12mmdVuG}iX9Fuj2l^Uizh|Cz5?n|OXiLp_5YSvcb#y??=lz9N5DP$q{h4@0N
zNW0trTldB?aWa6Hw!8<V7Z`uRn*l>f{ZTc>`(RQgrem<u43<h9;UY;=C>~80Qmpn=
zm;zb1mvZaHP#4Ey9|>nDqbi7{&SO!wqR&5kUP>254C@Gzi<i+5=iBaW0@Z&HOKT7G
zHs0L#ADe}}Uj087{s^b7F(U_QY0}|;`i9Y3imLP&;}|V}Y>SSMV3Ce2!u?J-2;#yW
zt`}9yS!lvaukoh}TlL}YHG6Yo@Lnp2p1Ec2RuxqR<ERD*^|z4Yny_wR6fMhKY0)!B
z^eD69t?+F<7eZhS#qHMS=F4sdCnTrU8$R;Hnd2{BUYsi`D73oQ1kA`6r^5Y8Pr;Bc
zo1Xb@vPO>3n;#I?`cq-<(}l!^*MhesXCPr0;la_&)h#M5hf56hLvWbysarLu7mHsQ
zD<80738!d&D=TqzG4YjiKs+h#Ge{m|PBEL1d6T~KNS4fL6FX%m)d^}>3G+gd^a#Qb
z3|VCV5iWb#uOKx`ov+<t{;>5QFm$c<k`mHvzxnT~5AE&woNFbqM5%0Eo45!z!thEU
z^oQrW&bRPn?U`8(6_xBXNB$SNZ#gwN0K98-Q9Yn)$kM3fYkU*hDf;I9aPi>c{sz8H
zK|w*G-OD0OKUXfx-@CN2uVQuy!&Ib*P*;;V!b!mW<0m4qv$h>?bG(%?GO|#ro%~S+
zkoU|?6RHMfi@H7v-jw7Cz3dFY7m)h?6H!r-vHSN*tvFhn9u~qVt!V4}L7))PHNRK^
zv7}n_E1i{%m96z<eiOLS?v$wA!@IN^=Q<AZo{FWoQPg*}MS=0+qH6y>$=q~(W2;Ng
zMDDGo8urC<yy|=}&NP{PGTpK1J@<UfGB$0w6$pK8c0MrE(NX_1>P?%)7>t8`{QKl>
zd!v1RR=+e67-l<a^s{VTga|yxK*Hk(4qn=<IVL8Px#>POKP@f?yUe~!U_#{V4FNco
z+v~LRXVKPs%>={HDrvf9D|^SDPa`-Ca#$n3P(ohbi8$WVSYuS=54}gERd$-{4?agK
znmvZR(3bW%GnGk}k-xt~4JXWql9d5jn&RRjfE#9gCU86ICeN`II9;f8)jk^%8#mxs
z0-PVFcbGg$cGw?MZY<m%+3DJk;4%s00vDy{lX*Iujt?xXh1+p^YVeGAT=rKhuYb;U
zg7r4%m!E*sGIkfuW@z}=?b6NJ(a^UM%*d;&^xDm4Re%&4#-ax25xnIY=xUC>A$niA
zKt3ZQnY^i6HpbhT6Fxw>QfQRk++Ku+enWCp%SPq31sJ~jA!Dkbw1pHKNtMcYS`Egc
za1e67&ZjItCE75B-5TX{_O)#1G${1)hESgm<`}koBB*)BQvsM_jzvs6Ct+n0i5#Z!
zuNjv85`j}Mk8brjfSwgq9R&(EXq14h(rK9Lws2&qcq#Ol7XGBJ&e60hq>s%;|Gsgu
zkv=zyqLDMEH-M9*`ezHF)N^-24M}0OSrI$+4jviP9lG<^5I@-6A4(zGSeV#G4c&yx
zg;%BQpNW4_zZXBR-_TDJ=`Z*WMTrv7eQfCUIKy2Fl>$+u2K!y55;w6#N~>CFN4BDA
z#j<rFufnagtd)cJ6t^NJwId#Ct8);K(b3&rG6e=-bJ+qfpeGjPPT^1J)vZj^k0$tf
zDlcCPDu(zY?;zuv`f@l*6Y^td3;x_KNo?`_d4->pAhC)a_pfF~$;v|B%+D$>CVE{A
z<YLWC?JzLkBc)PN84zMkxjf(I&YxN;=IZ7>-!J-aBj-MT;P~)jpB+e${MX*`Aq-j6
zj*$^O8vVrr-H%YmG~c6b<SKRGV<f1P>A;|BaXqGffJ?Mc^^V(({sTi*Te(ch{J+!Y
z&Dy7PP=DExjQEICArR36e!T?{t4>|M`qi(jHUs?T*UQc5R=0tlo*&rEULvAsbGxxq
zJA`{#E8XMsBcdp+^?vOMP%!JjJ!>|%ml}$o)2QBPf2++e|1tL*&?ak;J*+Bquy`#d
zU-iPc_G2No_y?z1b4y`fSy5;3BqXEQWM!585o@S+O_e=+qz#@^RehwD&S?mE!|XCq
zZRz*WHa59O5XN+G1!`~1Os~!KtUzmN>MW9(ynylz5n;d6ZfB{r-}x8Ker;%CtOWzd
zm7Ipzq~=X8H>t+yXM@&n99ugudrnV~pxZq?hroml^bc&dI5r$jV<{W6wc-iSz3#cE
zOR87EVl)(xum`bW<}rlJRTqY|j3z@wM%?Ml<ia`LpAkVV^i<5vMwOUS`Wl=QP~LDL
z3(gV<dDKp?n`{O&cB<+se|bH+#>!0^{GIjL@nkA6c(vJ1=kVabz<Sb^3uwE5#Pe!t
z2ht=j9!WyN5A~NvX<6snc*CQIO>noEH23GSv;qj+j?ab}WRXM~G}tjHD6kI~`khXc
z++-HOmISh1ZQCiHmQIbc^X+;o9S}hTE{xRG6&$54QwlW-=JU?bdSeJ1-$zzD%hFYr
zRpD{4Q{1q{;9C53D)I`O{4;A<A|5l|II>56S`t{<O$)7UFH<v%CNEnjW*Mf%TZogd
ztDcT{;PCsqM8$_ar7}Xr*but7(+#Y4LtO=|09u+<X<1YseXQSS-)d49N~Zh%#3~eO
zLPfQrdbnTW@zFCfvaHm$H~WJq5TG$Yn0$)1qDfn5QF2ORiZ+)_^Pf8wWQmr;FPjNl
z?YBoIT8$a}n<f}Myy}4u5ZW^5Eh^=Cnwsq`E1_ARZTa2xBnDjBJNY_xD)}M4!{i~e
zwYF&Lkn5n052K0Jlzhz1Zwk>3RZZke!qFsHfvDrjppxShL$)1i1E<CaL}S7x-LzpB
zL_<_wkEVBlF%uLE&zh`QkJ!7^U?e?MX=(7WT)NGv9zGqJd*c(al=8$Ms{9&qDV~8C
zu0R=%N3k2e4TOL4?d%Q**7r`S@8T^_g_9wVKkv>g^E4CVE<f3<OE?F|WI(vMq(lGq
zkEOyjmA{0?mSr2WdJN|K5ItVH<{uR%RHio3REx~*rwi9*(#<vn|I5Fz30m7qi@t}O
z_@Vz%HiOF!iPw{rnyw-*&(OjGp<<*ENxm&wpvT_=7lP_#^4C@L7Ro7zF}FWl0zg($
zLPD?0XL(~GJ151Rhn!{Why_SBAh4MzbKhda>RlaD6r8ngWqxR@^Aq+xT89VA^+jh`
z^1SPOQ+{+b(eSY4`C6%&vZ%8+e2#On{w8NR-A{KZu^5KJ4XGrsIw1OU;M1VeS2|3N
zpDr$h$}6Yh^2A@0rso<Bw?qpT%d2sk8k<0V8{i_l>XjlXE$s>>-+HS+1HD0xoBxAI
z{9u>G1wxzq!<Dz1)=VGCl-?IreJ~s!s@u)Xe=ohA?h($DL?yo7Xk0cex}`uZWka}j
zb-2_0Ww@E`YpxKcuh5a0k)p5sMN~Fk!h2dsMgPK=p^DXdh`kC`0OnKFw|hRCB<m))
zEJ(5$Sk|K8M@7S|f@?YJz3>Aor*<|PRZALd^0B5kWtw^X)L?UncfjfbIwlySK_!!y
zk%{HQ^L)|lwiEP!XgdADO<!#2e1*ie*y}W3zpvB52@ebN@wfh1yK_l>hYD5le~pnv
z)dx>X;8|hSm@)VCnEN~q5*g5m3GZ(D-ByXwfAV|z$n)tK(1ib;^;Q>xakmJ#|0Y(@
zsIB_?K*hrU9_(N5axQ5rR;NGi5kz`7<nft-f`1M-RB9@Q9z?M%_`nqgmENWh@wm0B
zgMEEa4A0mehv+;A+iAa(>vGk~TJ1hp_hAbSJzTL>`GpUf57~ssTye!V1<4`anVdyX
zoFYURP@fuWRm`4-6`O|&*$!KntgYtCZFjj<0dx-h&0FF>4eiOG6jhX$Q(Ed;%%h+4
z%nsKmDK~cGYKyLrlSySjj1aoo&93Z-aum%IAxmVfwg|{8m2!k`8=l=3X6=7IJ4?)k
zeJ`NU?YRmDs{To5a66<^(Fim9*rk@~Ipo$V^-uGpDDunEAns~uhd`Ao!~5o4x?W8W
z)OF1Cz^kGhKF^bOXALW>a`Y0f%>Z_`-8R7D#cwICL3h!pLZI8lpV;s<1j`8=CzX7G
z+g9zsKxb})tVONsDm1aTlF?CCZtWj!KT>9)4MTg`>%wy=uAPS1RiLRN@pl)Rz_1!|
zLh?C#KNn#V%4T8$a+6{xF`E7ZDMCT-&3@(L(W-^!nX@zaT0HU`ul45S5p4Uf{s*_b
zz4HD$7}x2Wa<KL;v7vw}C;iBgRxg*^B2XNN?#MfgO*RTSHQ&^>efZX`^aFU%hH{Tb
zM>Kd;@44$FC8orLem%5c{RPD4{oNf%m6#Qmz{A6}RXdsWn`@`0YfonU8nY}*IV}DV
z7WaEl-|OlM10nw)TuwoDvXQkpRdDb`+7WSKPzn^ns-ooCR;0d$r=^MLlSqlYE!*=Q
zIK7ayBmCT`7BFxmy?m8hrm632Q~x*xzp=19KlR&OOh^MyWD+iawx3^Q&WU`N^j9^+
z3AdUm!YNU?bnt14#==_r0~wf(4@7h=q>eV$y*oMWw3}>TlUloNejqv&G1h<mweX)p
zsY=}^wB})_=V{th*zT@vg(+{G=j(fI=q&3Eym~0Q0(F`0H$KWJ9AvrKAQ#hTa*p6v
z53gr7$r~s>gW}wIc{{IEb4XA6p;Uw1#@z_S`ITc>KDP4+hN^Rr;sM$OFmO@95{IpS
z!C=b!3U2%~87Gmp1Foe_1lGd9t+O-T$cVIUCNe6ud_3qlEXKtwC*#YmH{bs4g6tY*
zT}x%-otMNxTJ3DR?<>#`dqx<zHb<=PX%B(o4iyJ4A^8=78@I(`yLYDE=|Jo{rIh`9
zrZMl*%!(fy9T9t&R*Q0ni<P#QgVNA!Nf9?ClmFYE7Qs@R)ANy%7c~qowNBm?vU*dF
zdW;*w!`dz>$I%O<BJ1Ktip{(F0(ny#ipd!hl?ONpx&ia=>N%%Ml48Ua9NO{CmLk_%
zRzV{;v>d7g5+$sxtj%8bWu=u8wXkM_d=bm3$;wg;(E<<^zZd(rp+ENLBGcq}2^{jA
z&MT9TETUv>%yKG1$WrL;2$lY{<5SBx)ZY8lsuE+k1g@AAafu%HIR$Q;2{9vqF{FW7
ztzp|`EYdfcU~EU-*M=W`_b6AF2>PY_NR$l3ZRzPB!;JM{#W+N%9ZFg<aEXJL+@?>J
zRJ0+{gg-XK;}Ov2kPQb-S6W(_r9v0Dc-`noC@0afG&q5?$o>%<v)_|~m+7>kM9qn{
zv_L?(3l4f6BSt~KHg~#b`-mP2mnDS?pED0MXdYJ(d)eLoCb%wJ3Sn9c0fPP~Co6(9
ze=l-2{r~ZWXx{y*@f(ZY66F0a9_H`XMWFF_dobs!A_`^!K_%HsM&tv;&76hnvtgB8
zxb*nVBp%b1E67BLJ3j$EIT1!&HFN25<MnP<5wosvYva|~oWjlKR_mk%#`{ODfu)rd
z+osq24UT}0&qOBbWJkhRJ>}p#(i{JO=ttVhyV}hhAoNX{VbI!Cl?3mN+}qGv*75Bq
z64P$IrYo;3uS}f&*!J|u%KFj@d<9!0TwLAVT-~L;+CFxI;X>NIrm*B!jXv%1@#)2R
z(p}IS2#mf4^LM@`l63vr@UwK+7i>5;LF^C{EY!l3_3BT|W_<i~euUlZ9=^KLo({>`
z)tOh;Bs=p#Y!i2fR#JIiLTffqbse9TbaVDX?}N|fE=OQm=wxaTFATFcYA~u6{BYLJ
zMP&KP+Kcz_@FIvX^Lg!+g4}9F)>m`{Qo&WJfKoH>$$KKyvvpFCy1Cb0=?fD_k!NU3
z!4?`VbV`59)PhfD=WJ%w*)ic9{Rk2m8pPT>4h^?A$a&AfQ)AT}$ONH~ORGme5SkK)
zAqHaI#Ir@We_^}lvq6Nx$#jX^{REr?GY#}VC~fk|yuy;2y1JqxWk+yS)aF6}s><u@
zR_W^6pkvX8gq@Qgqu=p8CqqYF8|+T4wSVt9gfRfm5FJi}c^R)1+s9?U84ElCB0omr
z_w)>(hPym(Us|03Io#UYf2gflt=nc3mGS_D-3iHTfU|jM0-9q=Y=Czw=Lx6w3bZSb
z>OZllEfsLqs8h)J;#OKh(Y}h3fo+=SDIM(L%wcznCgh3J&b@0}u{k)ImoW<s3kFj}
zOUsuQ9F}3htdytTC}S|%UQYJftCoQk^MnjVTx)Gv)|OTW=ll1rWFd(CL?*3RJI#eC
znq}&{=!oCm4%_4#LwA?V;H2;+^rkvt$9`@J8#${%K}kY-kvV5i=EP}p+xz+WbWCp{
z6lwSUhEl>>vl+`vqzrF)my(MQL;jA#3hMfn7NyAiS-~Mk4wCY>O2C$buuPX+USwQU
zvz7rd5z!(SnXT>RH#1iLiX9~9UPZG*Gd*HH1HvkmD&bN~@+h!ZDFO{%!3_11>q^|Y
z5+p2CgesKP^VsCRd7#rW@<<hh@<Wi24=a!H;1$D7BTYFN2JlAAh!i06;+wy1_ySE0
zL@<U8?@*-fMZ0N*>E@KYR^Cx2)A_#~o1fwI5#57wH~hbHmqk>iM!WCy)Zp(U>Mw36
z#aV?e9+W4g2Ez-V7C3LUV2e})>rr_5?}`7PA>gzJ5j#P?nYGK^=T?F$d1GXx6(8^L
z*XUx;4-zk1J67+mWE>+sey_))<ib77Amq%!cO!9#mq*`$IBdCGUYGt`@pXM^X=b5k
z&~0ITJ&bP;pthydqLi&AHSHz8h^&6SLaWpjmZ+|+Z&2EuqBl3Xfk%0ftI+_Jv+f)5
zOthrtrU75~DvcGf7>7H)yG#_Jj*jQ@I&H-X>Y1RXmR1PI-pcd9bLfxa1F$(EcHw6=
zrv>W~rh8QT-+)jwSC8D8hr0;xXK=WQcP{=Yu%lJc7`6y~Q6I9$?!}Lz#wApZhCT~v
ze^s7LwCnn9V~R8H`Sg&asxq0y>;GaAu&J;)KfwsxnW;Mm;6H#62zQO<al<$@4Koek
z7?}ir3D1c|4HkF)d~^_^U2XQZmvSiWeja1ux`_NVWif*>Ctm@d7N+H%BhwV>x^D0%
zTE>9NVfSn{X&l~EZjnOD9|eIdkYr*lZ{{o_>8qnn1B5hv^E=|DIjd7JOr=f@#9iJk
zLk1b+E#Fw|u68tMBA`lle0O-e=)9R>EmC=cB4lJahm9nW>~j1b_@1E<R8Fa{{vd-6
ze45JmQLtEIMs?OWKOYT1`oLcgYQ0U3OG<9-cPc3Y7#fHqMu5!rk8EXasQUf7@*Q0)
zRx-eCMV?ja6q(`(k<UToCaDx*xIcvEATrT<jm@t%PT@YD$Os7+hCH5IE?IB6`G^0k
z>Gw^9a)K~Q`LJ@IhyydiZpRsdd6ju<b1`f;kL9OPA5Yci0^x#_A{QlIG3g)PR}U!J
zC+qrbN*pyznhEVRkdedJP$o0V)DM*Oytg3U5ydv1EZ-FlHRl&hd=D9+mBQ32(^pjE
z<|Z9)jdwK3JPyMi^{}gKc*x0^hZv8Xa~syzctY<IDn)tlCXVjFB5oL3H$GsCW$_K0
z*`rL$`RA-3)aU5Pw%TqhyWeS7-2p;Gry-RIOeTH8_>)78ORUmcC3zkVlsgnmktPhQ
z?vVr1)$^yONMD>~byVN|<;`CRLe?u4rAo~F(lqcPPzlPx8a*-7=M26u;53{5m?$2s
zw>mtj|IGBp82TZP=a}?+UUC>`Ol^E@f^vD;rvDq)M*bwio4472D}<QVPQl{}66R6o
z?=7Y!$M}R)T2;_k?G>Gp3z9`d>w-Cl$Db=m{eJ&CKVNPBTvc74PjP%>`i4G97x8~A
zz}GI>_qelefd>up04>mO^65v-A>Q6LzrDG<_7J07Yb|(U+5WoTsoC1!hKu|A{oQVp
zD+(Bbh_lw$JE4uFnVgv1AMT)wFs!1-tFk~jN!);3!>Q<obs0>y6f)I~4H+n2P&d}R
zu4dd#m|>nyNw7Gfg6!Gz<9!ADS#0h%f+UA%12PQ<Xx5ZOQyW`bTZ^iyh+7`oKk_Hq
z<>rWHi05d|tS-;5FHFPwM0e0zlyb%2t>SO;mM>4O^(qp_`QstP9m=1CS}TlVDKs2>
z|M6v`Ey?}KY38_=?eCw9&AmFuv;r))wxyXl8xULqbCigvNR6&SoBmb6F<J_p@LYwE
zDNztw$=TS|M8FqjnIr+StniivA~FH>_}lmTvQ=qtVedaCLn@%GWy%V97B+qkhB9!4
z2_zb>9{LhXf0&?*RT-oI%kS1`t$%7O{_AQW=fpmdz5$1OR>f+M?_Z|OxVG#BWo?z2
zxuv^tu1%2E`~^|A7lvjLm~|DeI1rN*jSo&6h?^#97vav&A8vFx|66<NgGU9?*9k9@
zC|XpC`4s3;1qXvGFyb$3HD~|$;VNgUBq^v0nM1dblbMoZ4-|^({r6OckT911C~Q3E
ztKXJ~wBQ5Y!51Q#eZv+QrKFi7OkMLMWP-YoDOTDY=VBCtGO;j&49RO?7tdPwagW{-
zVi?2{qcFKIZEPF<PMLcEob2P{wD18~EEhwHFS=<|LZwLU6UF%unAw)Az5Ectople`
zWriMnNw$!9S+$ZDN)pH2^}(wm`{snr$p>uk6Sh6NI5IfN>FFo6L!6v&+}`t6RYDC>
z@2{ZbhMnP5;Od|Dt}mgGh(2!iz<s#E{`RHe=mb3vqC3*WG@U-UF}(%dmn^kgLKpHK
zSTU9#N+szX29y^W*-;0cIA)rpq~jKvtU$Vkbme{>dB{O-ncV(+ft>_0RfIn|s@O`k
z*IFO%&_<wQiF52tU5|GtS3dbj9p90?;~&Hd`eLn{gM9~wOUh9W;cTwZDmnB$LLJ!x
zC5=@(EeN(pcy<GanO;PYU9MN*+Razu@y+8uJn^k}+?VskrZs8S!+_*w4`OlZu>`?E
zEnsXFu88m(a>6N8qGda&8iKly(K)XB11`T;z7ox3h@?kL`|J+dxcOwZJ;x%D?ul33
z#KbZ?-QUQ|O0SZPZ)vJ+&$Cc$(lsqT=aWY@TGvWYV!*tU#orm6c;T4&v#_LDc5QZp
z+|uf}H29lNuKC86m7guaQRV>Hq?d|6wI?LQ`9Z5Gn(8`8y8rPE^+=n+mN;hI$L6GT
zWtW;VgUc2ypc8KQ-OCfYWRcrvK$znvK+#(yv$C?aJU;?Owx2(d1fmt*;IG}0;N{81
z#rff3DbZUUI*DRznk*b#7Uu++e?1ObGK3b&@VM>yID^E6jVr*3i~kvJ?{=)U*VSDE
zi^FEU-gUknN6sx6WQ!Kt+RpDcu|TKE%iCUA36)CILj3a)TW*A-mA`*qH%OvlclisO
zP9DB1dVKM$^}Q$|(jK>gdLJd!lZS(Eh;}f_TT8sYkV{W^SkRzj3@<@^P^zL2jTp0x
zY7Gn@K3s^(m#;cNwz|}oAv8F6hrChmIXYrGn)ZN)0SVSOY_u)^6?2dRBx!FL8sN!j
zBE23KxgS14740g%BM^<XTm`BVSct$Cqs>rWGdVdP6Bol@YoGHsy{U3|-sXVG!Bwl%
z`tQURS#IRvW~YX>-f(&32O3Mz&AQCNrv4ELI}Re@_fq+)N@p@N^f@3k=A7-@iB!)b
zQX!VPKqPIGTiYb`2^LJes5aL3xyaD5kAo6PTC*I}CuHYQKT5TFV!X{csbz%%+BO29
zhC<@7?z{Unive$2WV+ydNX+l^ZBz~3+332nW9lC)D<>Nc$qb9aKU_ww0!Rn1z0=*r
zy;d(FD$bZ6+2Rou0#)z#q=*R3#Q)A(&7gu-Go5RY_H6GHKlHV1#W3GDcU*i6;7zDP
zn=A}c5@wOGeUrwaB|pU(Qtg-cJ^p*LphBoW11&k`#)#%idSc>@f+N>lJPJTFk&zKV
z9KBZ*lyEn@KJ*Ei0d07kdfJUjb(|RWuwA^C&u66Hg++;oGh$AGtF%1UJTWn1$jGr(
zv~U#fiD1T^kq^B~qJ#|>a}Z<Fv8eRfYzoo}73wGT`{Pm=Ezv?Jbo-q6cn^@*D#M#$
z{xk&%hqAYxd6)S?jFxp}=dwTg>h1b}-R5z|-(5w)2egqPm?|>ra)lZKjfVN7&Sbjp
zsV)gVNm~4T@$R-)em#8v2?p_y)=J$7%&Y%sd{NEa_o~kDI=pfBB*K9`Ty_WD9$)`l
zY5N&5Gd2;xD7#B*K9%PDP_K0TYHY{KOZ^wuA&nwg4mq=^TNSB6?zW<xIO~kxcs*he
zDZpV!V%CKKh11TOEfzY(g-Ee-yuUxje-!3Riz5ZQA06}wdcaGo<@U9B%z<$q$I~Tt
zHQd*`-8OJ%tEjQ6j<uwDN(&J=_D3v)ke{*DsE*C(dUt^Qv4(FKhS$e_*WEsmp@+H%
z)3E*t*9cEYo@yMWi9$`~i|SkM)pXO`9U$pJui7%*H#=>}5_7h)X7V#Bc=yM+P7ri3
z7FNv~@LCWlU>|KSx;dR-g;-BNcUOkWpbt}=YFYFy%ENqByihBif8(Cf5xU|b=?hSk
zSbfe_MdBEWkpj%^l7nx&eP+sAW@UtFhzfs|mN?lS(2!?3XXj?L?@IBx2X>i&5n%=P
zpn@ReAI>OE!;_y2@A%i&A1;@@w|(ZS&N5fh35!IocjJ-*J|qeU<4%VuodOG2@b%HS
zSxiM;J3TEML>E`w2PcwcQvk>_ohuFI4crO+t)DA66+1u(AEhgA^QP`;^q{>H?W6m_
z$d+O1wC{w;$WDted4saGz??&H^King!gZ^UVaS{28hx_yI#XjPGpn>UyeL2DE)$D#
zG!{>bh$n*jM(!u84y+7oZ}`^pf$_TdG|xV_yqM-ivx914O`&K{T}MGFp$p$K-Rt=y
z_tFX_uRdnFr3^Qo19C#PseLMhnKV}qnW+XCI~GmcVyrz@vh(Bf%aaqhC;b34=*IoX
zp9ZJ~=${2?L_*9KwrQ2E8$m44l7+sz1p+#Vl5UQ$vf247)P9%I1tz#Tr{!q?B^&&*
zK*HU^REO+9RpDvV5*bjW-H=OR@<f!8MYExrq`0M%Q5bMt|D)h&|F)HjTzKE(VNYd2
ziV{(QRhknAJ#Fa#$4+8X@G&gRM8;eQ9hzOXw6;Jts8<wVwbSVVIdb$Ev?x-}rH~7O
zpW^D&s$ZV&nHnd1{I%KAS~vVn{#U_<*!!aXKW(S;qCkQ7&F_d{wz^nFbuSx@1M(gb
zO|hiqFE~c;rdhn*FSY7iKDE9zuyg1%8SQ7Udozg*>bCmaTT1hk8DBxz1xkMZV2Cr+
zxi)?bi{t*l$sX8k3?VL2tsrghozemvxi*Q$aqo(U;&CtzH5%rM`pws=uJr|rb0MW2
zJc)ZSOMEfSO(qz%GgY%ZWO3d~AR7%)E_Vndu?)1ZjV~J0q+(+i34H!3gvb9_hm{9|
z5(8SpuUXw+1nj;8U3P`{cg(mgja=uukN9j*{gT1hiSopX9x1=McOW!youVqN#go6d
z5-qfXeEIe5JuimWh<iy*#pnsFx(oI-!1=(&F5cQpA)rpYbhJ|_7hH92H5xg>dq$L9
zJAH$x9i2OICG(BdL-4_9N?-60x`7QLlgCGeqcX~E6!F<=+`Pe&$Cg2nm<ggse<qJ4
zVxtFuKxuiaFRy*;EGsY3YSyZ|NA-i|Ny|65Pa<Bd-KFhSJT3Nbtqq5Ehil~(WnXBM
z!NLM6LBVh);^g?U38B~l_sd^kGsjMURM*A^*v$hx_3&Bq7q<&1y{_Jw=;}zF)O1Ei
z&9l~?-R>@E0N&23uti46TAhfcFjCsA);D!2i1jG>@6iK6s`;n;R52>CA8}NZT;AMi
z*rTs}Wr}Lha85mzsk-reXR_6Zg{YZEH(GSd6uc=J+mp);)D7FWzTb;U(j~*P)VeaM
z#%5jilLGbeUoMV`g<6-%Cp-~s*FZ|qaobFbm;7CVuow-M4=rt&TqE2}v@^O<LP}qK
z`pitnZwN*K4tuO>IV?@p7Hb;e@q`WWQzg-3z)=Ii*2qe-F&+DASaAmdVv(3@A$UQ(
zSm}dKgv$6uSMfYG8vOo9-u7W5!3R2KQxw?nDWY_3z(ggdrVcgPMYVe~4N%WT5Vo?y
z(j?!_nbQ2okaX`4RxW1`-8SEg-d1IY9Ked<{W10Fx2A-8@b<M&0>;q?kz*8sv_-I%
zSM+B;Yh>MEIuD~{kV@48L8!EEY9(`ZV*^3L7b}@b=Hl<))xpw8rD~&wjJ`rU8tuRi
z7&eXk^qJ`!i+KGof~U8*|6rFltcSY#{QS0hq`co4($NyokK&&=EokVZ6pNKU&8OGj
zL{?}df!$|2Cucz~OBRA$i@h`RruQ5CfAc3K5BJu<YFgNi&G9NtomA`=@`Xl+g<-0u
z{O3^VFCATZtxcUEu1Oady_U6;-Sw~Mv+@Tm(orc&z?@4Qw>UnvbXN>O6DOzq9VxX}
zpIe$4INim`Z@)qB-_Gf?=Mb9zHn$yVHi!rI*$MI<q6(VxmaomtY^)i*4_-%5er@q*
zgft?J{(BYDe$+G0F`yO(KRdg+xiSq~cZ$%mvARKNX=hmzf{v(*^3lxNmMdMJcoEM@
z(!rF(2fjKu1TdF;Yh3Xz!N3CSmF%}XPBWhrCt32J0$~;-9o`u^IQ*3Z%_?W-|4!Ce
z2)N+c=~OIc507womJ%(e`Zg1;I$;C{;XVezAZ_<J!>^vD7MA)?J-x5O`b8ff3t2Em
z?iul7Aa!%E?s!S^F^a2*@~(%OSq^%%V4-1mNzaQ?8(3Y-dh@f{zqYQmsr1NELc(nB
z>Bu-N?#5s)f!5F9vg_Spft`sSyq;sx4cLPxZe#nnd$-*uNk99t2Uh%ef=h_}_4TzZ
zEuB<mBDT;0+vfJpHkHv)R<)rhR<?tWi}<s{5l92fHQv?NN6KC!;x@=XTx`hxv)Z!F
zMsT4RjR9@}_A8R_*|+Y|CNErz&y)ML7gffhdpGRAOV|U-Trd{`<Cnf~ctlzT$P~c9
zvVC|XijW-dOllqi8q+vEA97xAZe>-$EU4h)P}U2{fhDB5+!zC1AxKxlh|{rTCEtT^
zjH8huU((6(%(mZZA4v36-fR55%Bg=RJAoPgQ4E&=20?l+Q&`F*;yo;;@E*O>66NK^
zm_qh(c3O7kEyc25mfmtbLuEA1mb=}&oxHrOyK5Z%)BQu%{Pt3Fg?&i%aaxu4rE^YH
z3X;W%?uWt5-IUtM-3u8B3{>NCE!ES7YU|bV`wkvwoa^Bw3{iW$&N?MVor-iqF>6_k
zka32l<D+ZoYDlDGu902FM%EAZUWe=Dkt1Lu5c7Ft3b*fPT*Zx6=U>$u;j@YJm9xd~
zFK5E?!^RKNNzzCQircpwSh6+7qp)Krv9JGlWJ=K!MD(*P9{+nKoeG|H!p2c3&^Zbd
zNB5n-6v1{N5dX~o>^qR$c5E0C4Q_+;&zo8Hp+Hf45rcoX)X>fD_cClRWvYLZu}Sgr
zW(Kb<ccK4U(Bc1}_eFgFcD+r~7%|;M9HVWRedVlrwq&AGzu9o-Q3IjL@{73X>OvJf
z3O+D<@~c*V=J38@7*8RgV>V(y2OxyPRg^-tG_At6RvFbne)+NY?it!JZqL1K4aE%i
zioWbn>b4<Vxz&<XD?@jSh$p>Ht6~1w;d;b~h>$-0XT{*@2a~NEoE_Y8vQ`D}<V*ch
zw$#`hzV#yG61%MSjq=utpZIGNOpOxHCx#ZY5o3eKNS{l0Ym5DyQ*t;A>Yy&uS5H%3
zgmqI<$Pi_|LG6{*b^Etln2H8#v9Sam1rs3z3X?1lN*AO%Jouegot$Lxx-;pt*?`~d
zxofIj-qd<U<z%I9wlqbOQMVS|u}~j-ExX=e#^CL{@fuW1@f?Do0(62VMU2WcGr3Z0
z{{RG{7S#qmal}#*5;l!UhxM{<HcZj5N<yoqUKbatb|MBhRQUu#va9N`F6&pf>iYhJ
zmba_zgn~hK`Rj=-|G?76pA0&aSv=O2T!o2clW*4`K|gsUh42X?;w{?u9!IIgFauos
zC@^|>i`(!tBZ>ELw$|j))9Y9uL05m>Y`y;cbir>np4n=4f5PAXp!;%^IK9-Z<n8q5
z+gg5HNg>9&VeCw99OG%1PP<~fH_WHt8$=VK@@0+44+h?%*SL1kzY~``oB~;L*_jS`
zL*T!C3lTA{hVU#`=owGbzE<ahbghb}8m%vq9gHPR9>=@OjlEVXYm}j7J>cAZMpap+
zCvQ^G#h;tCD~316PPD{zout{4wHaV$l3<6tjLH|74eza$M~7e~a}GCZgAqws(H8xh
z1$o=Yf`*W`s-44B{~P;^T}-CrB&tVPgF_MC7CsSnS6`5yq(Y906e0A_nJ-$)xr;G+
z(Du<|o$=~VyDI8sm?T)%SZFcm0Y8-`F?04|Q4GjdCK(lvAqH)sE}TCP@1uRh9+aH_
zUYrp$^j^USYDkc0j~N(<v%*f<dHI}K@4mS8_vlVQ5@rU4?_!|8%j=~kiT%A3u{X5M
zMEd!W9OX!=aOMV&h}I+J2E(lWN$~Z_U%D<@e_hNnZA-8Pnt|$}rfb>%@s(6R<7UM+
zl3j(R2{Q-!rRNF|P=7AH{Wmf_p`ZU0Ol~sqJ-k#T>3?l7JD)C~zThuaZE)q+`^ERH
zVa}XJkJ^N!!<aJosYSfD8ah$%==%9RBSv5AlX_bEy3^B$s~@go@bbuE)<Lj`Ji_m4
zv*9@cInLa_2mKx~@q1_c(E$!ygpOY!R;^i_3NZJ}1AOWsnzBA3xx&q!uRc8AoxEP~
zW*kejo?)+hB_?6K=pe#RXBI_EE(DJarl*s8)vIMexGKQG5JV!#dGZttTe(Nt#VOvv
zlDX>MV5zPduzSy|ejEzW3`lp-&4sCo+whD@iClOX8A;ysR&sy7YqiyJr`K!ze>8n_
zRHg0rceZWYw(Y42lWVfA$+ny9nryqK$?jxM)?~Zh`}w}>_iw9p);ae%`?~i2XnXx>
zzh)KMN{I@6Jo{@3RFYOt$G_SBfws%P>XU0jW#HvZ<$pHEiTJYaQqg#$v)b8{6u|pT
zVH5&Nn)uf;%a<5wU2m(bcRlpobLt@B2I!o=o4Lk~x#8f`o<&8ALMTXR)&E=cjw*Wm
z9VD4)2OiO02!=h5t~t-zN~yB9ON~9bOp1NERB>CLZe++6@DXpn&XObe&FEXYd>N1%
z7yB9e-RcF1q>HLQ-V*V7IFskB5fJ(9_kPc7<a;g6ht-%dRrV!!?dGDiqOM1zz)^Ha
z>Kb?rWI`dg!#jLitvX?^?W=o2-UA}2C+EQ8_~>E=5lROrA=(zxazjt}@t^bQy{Syd
zWspW#vk*Ktd$o88<D$9qO@o*x`n{R^IYoRW?subO7Eat|2}k2GM^z2E-)CK@(pH)t
zdnUfsj-?0H@cj^+hoDi;V>SszB;d@-QS4SzCb#vP(Uh-WfDum}3;e~L{S7+1{C-Gh
z>O_eJx)1E2N5yM6D@(AhN1&wikjE0?5XS|FD|z2Tpd%ghixu6Z<F__wOarrzw?^KK
zkELiA<yRPvD!2O*h4kX|icBYNQUbB?60p#Hty6%TPXIJ1fk!c;?B4yH5X<P#C_4>e
zCGQRMV<D=C%;QZ`F#Smfe$I(F@S*?aOqYxTGXRp<8zI7J(-x|TbofmL5vim6%&I7m
zosxZhG-G6Ev{<Sb5r6Z?BC{HTTJXVyl2&ovlQvJO4GZISajNO_3F^NJ>N*H`o6pb8
z4|n1kerk5odT6cDGs@-a@Hwg-oV=KA-8@bzPG>UgyzF^&+^D4IuYUbGHnSiR{JVA2
zE_jJ>%A;)i`lWC>ZT}JWPc*qRn?7+7fg?|Z`T2G~s$i>a-{E=7U9k14-6BI>&5fHE
zTFjJmUKmN*C`$S|eY-tM{|-hrqWY_l>g9NXA40Df8yb?2T8V=^0))B%#VU@lNIx@t
z(K5A{Z&2bU{YoxN_JDPZATv+IkO{Az9ZgeBzn%0!T?>C_EKSvR+P^pp$&ewdMuI7x
zlp%*kfl1*14x85ut~&Uj1K4)aq{@yxI)sB|F?vMW_D8CM4+`CZYcYV0sl8`35K6(R
zS+`EzX+7wA-MUfC-8t9UBE2An6X`(OXLQ&FO1=%fO0>6T1UV>ue~0F6exg;In53>R
zYIF)0O;{vbQq@fjTxs*7nK!`pSbYDcUy_u7pgIeBUrttp6GE+uGXn@)@_O4p>G9dp
zYc~b~kJa0a{zF7VQ=_%chI79}+|>fob^!2mj~RJR@uA)pYp_IXH^`AA$YRx>Oyo2N
zTI+kBl|09$&iz23nn_l&G@*OTmBt2TaHDbSMp7r|NaEE*nMWw%z0E0VVy6Q82nsJ{
zOauF-0#CjMYtjgSv?T(}FdlDU2Uya*UZ*!vMTI)8=MSY3ag}q}c7}T2v0R6`JTEvr
z_REJ5asBE8K7rU$zhsTkqyKUz9MPf#z7GzCO&uFQEz9i~efhW_5q;~YmFIQb;PbuO
zjy9h#3~*Oq)wzIVQAIVMr0MDOUA-|7^vvgXoi=MdUbSh&){dn88YW9RkNpb|dYYP+
z3c6rdNh|(*u-Lp&bR<WzWC$^ASl0VZ;jYovVGHnshLU+7%MfYKV^+WzC2M`uZ&L5>
zh5<NqnWa3{#H6F_wGC3nw^*pTkgBmlx$4=cW6PCwuisW&@zM144}EKHZh{Z2VBFrE
z9Rhye)D62V?17$Y5WddJW*qw_PWcJ?nYjh$r-rs1)%<8n;qn&j)4s)JUcZ%Yqm#2x
zs!7q^U2GIqLFNgLuF9|MGhe%c-NXc7^;<&X^!`nP#)LDdX%MY8`@>J}7w0~GB#jEe
zGE(djq3Gd}8yEt8D$s{th2IlOffR>f|27`h?74PRtLwbT(r1og)FFHjt3rtvEZ05#
zEfPR^Ueg_jN+~Na%wnpIH_)BDEuG8s-5XId-2sjKA~>kCt!j?_AYBc@NqJpfJIwcg
zIRVl3k#GHADln<s4s-YM!)4LeZSp<`+NC-4@T1Ph3?b*awU*O_e4+CPg}WWX38)T+
zz7jm4YThj9g#BAs>VR41XJVBk83k906qIy5<bsKW3l`n>a)Z-KpcR|=rftclW7jS4
z#RaJqWY)4mgpy9z-a}IN<p%aGh;26SH8^?yQO@{BCsf@gy{>&cOw%HG*tqtcU=KzL
zv`n$ye&epnoQ=0hUpP#C+lgUxBgw$rp<BhX$T9mSNj}JoyFb2zNCiagwMkN5bAC;3
z1R2aG*BdxPdQ(*X-z!$5aJTZrOqI-UFKn=M%Z=uI`yyFDD)m7y>xsX*hd5iu<(l5R
z|Ky#BJIWZO4)dleR>I<q&G;7tY4T$)+;dH%HSgs`<C%E~2S_F`(AA%|a(cIz3JQtX
z^0%Ej{qK4aRWy6|Z6i^q>Fa!{S(Hs6sjJ~WE;LIy%3$x)ousDit1ZWmr<q%|Ly4ZE
znn$pVFQUOdqmE0|s>u_OTd)%MUvOYZDsk{Hn>zr%Y-r$eeAsw)J(^H7Oc^Q=b0E*9
zhv8I9Y@+H)pmT-de&IC&i)^U{63CBdcbKzzS@jz47Nam<A5RoNiTnNa$_cmXK4UB9
zm+EwU&PGA6o|n*XL0>)}=j(c5-uO|%1D+SZQQ>`Hy+^FB<q7*dtZscgygzPPa_Xed
z|F}3?J=bev&Jc7Qtlt4Z8#4dzOwAQ`i>DoVrZRC^f<DtP?u01B0=|30%v}Q|4{lr;
zM4ED6Wt=CvA}G3}&13x;5DzMx!H{9+@t||l!SMPWP4R3N%o~-lo1QlPf@T%pjjAil
z2i*&?Q`Ka=#X-~ZlI|wwp2N6sx%{pUf6qEHObAe=Z}}OgC#<2sx-XO*Er>WA5qe)c
zJ)&a8=2OGkoek!~;NTe5V#Ri!DhWcf@BTnwwMaBWlNxlp?OAYZIFIAZd*YivQ&(iA
zjhh{0eD4t13?8M&s%!@dxbQ=YGJ%ddJ?3@%$iT<MsI`P=re&{>fTzu@*Ewm{FlITD
z#?Moz_7PUw6HWW|Npa9~L~8Hau2xQ3Wkx0BVZNi(+qfAb$=2q=?Q)|<z3Ty|+n=FG
z#jSp50=~x5MU`0zJuR%)YWz#w-x8Uj&Rr-u3CMX9=kof<Z_>Sj4PMF4KSTe3L_bbu
znE7svCNy4W;@3@FK5t&n-T0>Sg(rONAqwTFaQb`?#yeY1hhqt(DX{Gs^pMX)jUTG{
z*V`_;eGdeH0PIX6x*(nYXvrTS<zw@hDSB^Tpx3+)4r2D|w$Yd|9ILaaYTqDfndPMl
zt<p<Ip~adNaet{UYWVKiS^ZGzGLLMLJ;l-vg-XGXD%V@EkN6Bh-nM*%A?#6R+$z43
zMly1N_N0|KUlLT&S|BiYm*J9`U|q+VLFU2C;EQmlD$}Gq3of_BrB%08-YUqQ1LtMf
z7T@feMU`V{-vhF1T{}nVKL}>)8aWN=J{F6_TJ?xBd>b1#!kd!)jcHwak#bK)C&D_G
zZ<Q7%w0tQJ41{ci5cLgFB&ZZ>k6!P<DFPC{{`>cRQ^1qo<Bh6^je9lvP1eTq;2ap6
zVOzrIEoU6pJ5BrFq)lZL`eoTo(A`gFozKeI0h6$h)Anqx;3NJZabiO%zAp=}f1aGA
zfE8cXs9ARV2tKO_KMZ=|2K}9x2`nXhPFS9j<<$?h^t}QdkbnCBX#p09(Y}=Kj&T`m
zB~|O_z{vum-VTWNV7TpG2wl*WFH5duljO2w;+x0iTF<{J&G}LlViB_plN*l+9{UJn
z5+U1{SOY<i98n)Fr;fSINg%i%&K@;BdJm3&8tD_mJZew$Ki%tpf;`Fo*-b%Dz}(vI
zd(t>or*!u2`{iZ{y-@Fng)$sDo7?md_#L2-@cEv<AZKr8%JuU!p2|vZyD;D{XydCQ
z!JDH6`8Jy%Rz7As`u=*}#qQz?xG*l~&5t_iQjNt;8@Fhf6UfxxIh?W)R<h=!6xg&Y
z{t)~dq%D-mDwcUSmECH$aTRZYZsFYz@`@gOMElL1uXIb2ctI#BgS{DyA|wxWzwAU5
zx`^0aGB5TvjWnT;p%-R7lAoJ%8l<IY=@RW9XC;(6f%^{``E<FaZ8g^omx7TdhjtF%
z03RoEC<`f7;ppqeFYltLFgKpQfUQ?p6rzPawIKs18LrYI)whQvdMQv$-vp8a1iO{l
zeg+vEy1N+?b>NM`>wKt5;x*MCG>t!UquiZ8!)ql(VdMaHo}i})hT<V&cQvU}kc{m1
z*i(fD56T^8<(<FkcfB-)kS((ow--4k|Ch;3aKG1Feszfk6n6x~gwX|UIrG(pYcGoX
z=Obl!slIN|9LH{e&Hr!`ud<LJv92Xe;<eTRY!Tm&*~O~m<+}B&@%C#xFfC2B)zZer
z7{Aw<OyS3?9m?N$I$6S17CCezOUTGL>BsQl6v?c|E=?`f8&lF58T6_}jT~IW)g0~_
zEM`>*m2#%En!77jjcY_K>BrDeU7}>qWCG0IeklF+E&O<T1bm&b7XHlUU5fj-P9E?C
z2VEy7pc~|V<}93e61_FE23L<Lr5gAK|Hn;c;`P?ot5NnjY2{h`lf?lTSnT=^nO<~X
zY85hO$i^4A;U3fWAc8?LPWOBv++fsW<%lKg^JRP`kB9<Tn!xn_Ab()sw8-8(c-S_v
z5L?HA<H@MuZggb-L7@w#9aJ-RAl1e=elOERU=&g$vV+*|zqbTnQPuj_hri%4jp(>T
z1)8k|O;+lGuI_xZtL|`GI6ga{ADrxBH<=>L1^S1gKzhnLzc^grfHh5a%Fy-rk*K8t
z?f5d8BfRt;(2*H$p0v_sAtX@#A$bn0WKMivH)A84wG-b-%B~UQkfXyW!}pXEG%E1f
zxv?1ZbwRkuTK5EkDtk<UaP^6Ay&?xK<gCd5HhL-9V;g2m8#VLudP<eD-X5~6;)z7`
zzv&Ui43h+W@S{Y{bDaRuJfR;|&qKJ00w@jFj=l1Om>yN<#kWUo4r@X_!?EJ5D8@Dm
z-!wG<P~G||@jGYfly;ZJIV>!SVHb=C-@Ds2)bu0Ej^ucne77u2=vPZRnK&S}QiLt_
z?hi~j5xN2+8L>74W~mnELT2yHgcDa1tI(Y3@6p#?UkfC^M`($sm_En+zXWHQ7JOiw
z9AYimyonvsEc3qYA_zKFe#>7^s!Iu}VJ;;xb%T@;TJcTliuAA!lz3<pL?NdoCR722
zNgC)fBfk$E4D3?O-FR$3gz90<a~G${m0)jGnRQ0(#(>IL*-bK=f6eK<Uvh?eAST@T
zCUhr%0<-!tx5FTxjkE`EBhnLv7?Z&e|DPw^uboRdazSHNax*fAL)fDS0VWWIa8bj>
zo=`X*mmiRGNjN5BU!VL=Y}76~U6cqv=zh6E)uEu>d4b3?#(rw-4|CS)^t9SnEydt6
zQ6Q4RW!a&OipvSYSy4~UV-t|F4>29*@8Uf71^ln5o;Y7#MMd&?JueCe05|u>?b=Uz
zt|_li5kJeBoF1R$zuAkP|Bz@Bz9#BSl~PBz|0FrPe}OO@L&b?#PEF4?4R^M~9@5;Q
z?CmbVn!dmHdN<4$4b+~32HARc?)QsDkh$>n1Y+riDG+}-N8_UC2`(p?9HG=E9s=ts
z8v5fYxy?euDF(ODx#!Wec;VFXEMUcgtkM$EGQ1NK;c>EZ@-#O8268)8bNvhJ?K@TD
z@bxe`F_RgkY@WH3nl%wRl+4xe2C>M@m9Cr6%%deepKdnT-*w@{<<=GX_<K+1_kU=n
zkC6Q&6tYe`1Qv!+3h8~HX#BW6Idxz?;TCCLJ%#>P^nB#``S?&qT3fxI$FoU`j&**1
z<EQVhcO~+m^_E%!0k6?N(1?#0o1Rsj!>5Qp-^R+S`>lj3^4$<ni#WB3B_|d6HFD%X
zuhG8*DP*PMS=gM?o{w@dyG(``J7f7rbSo4|BI0D5PW@}qsZk}|gr3MDmb5@}%A0^>
zx;6_}A^wS4wF2R6O1(r!G%R0r<<6fYF?HVc`WR5}F}1eJX_-L4-uNxy2(PmjK&>5e
z1rqT2U3R*mFM%ZT_ciW>_1kOO{KS1H2-T4QS?h_nvA%-()ki>nH%6P!cNX7NHUp4t
z?$eVxAw~prBfmX?YxsDZe7;OY511idF9Vzo#@%D4f--x)p2S*|t2^s0C~d)AQCW+v
zlGkgYyc%<)D3`f@tJ^ddn~p|Qx1}myGeejIWzM(ivYw$#94g>VoT+>tdB|k}K}y$W
z?;S-`7SxHsUuoBdkH^yc%GTpb*4C{AN9NNv^J`-f+|KP4bVTp(o1dfBpMsye&7aZB
z%`)rC-Z3?wVF+jQItu&mnuSxWM_@R=B#MKepq@`q)#ja$Ln+`^r`Kv}nhu7i2&B47
z!0_ct8?*R76U8^*Fk#VkHw8(f2gyPNvPC^Vf(NMQUMbVz$M8F($MceKn0|5c|M;B)
z8aqT7Q<Ax0`MyvD;)B#tpi-aQCiQb`P&Tgu9y2E&D~AR{3N%8<PYy_l;0eh5Dc<pc
z1vU=jHATZfF*+CE>o;*Vb9pukf7gLLi$zk$RHTYxh5Ya;^lXQ={6NgM|LouS57D=4
z68#)GTCOuPVsjKhZa7^9mg%m#dY-MgznAiS;_t3{upb=)Llot+Ur?g#&3X<s(*{m9
z11m3W@OTHcJVK!(0BangZ5WpK#U&2=gP3)Qhbl)W(wXx)ZR3etlt>2Yude>8Rx1)$
z7+SJf{~3aA`vw{%N1RPToLa>ZQh}_WwqBWYse{siM0cujajBqbhq~N+ZLyp-P*wkZ
zo;Bg{HHB+sxq+ZQmoYWnbV=25L63LBu4#hvh;ULXTODm98JC=lOoau!BvX&rP@EM^
z4c>~6DYEx?>1qf~*+gQVIof3^_qDC^(&bGn!@GvQPZ9yHXqoc)U=(GjpFGi_r3<R3
z@WUnUxeK6!I=!BA<1SSKOkWh=4ei!DwbQEoG_!Xk!<lv4t+sq#tEwDRs++tRrUtJ^
z<)xL4k-)gQ=knx!nMP5CR&(-`iB6cE6f>w5he-*jYi;ThrR;+#lfT4g%tXa+JN(fm
z5%9{$(^Xgb#Tm^cs3bed0K3?f*q<|%_T4<GkfYINXCDV$LM%RY;*bEoOAjJwKMDd}
zC>hwvx~5J$1MFFHzy2eQInu`gV35Hx6hYS>nwlXn-CV9~cX6k@IRef_{l=5#O<TV2
zQ*B%RvL6MzSA4xxI=fFdw=<-l4{d$!{JH5Oo$coHr5#NcsQ7&42AbITq=>+5MkHdk
zGI_$eRAUfucifQOw%x3dHjcX+#YHNHc&H2^(e7r4P<EQa$r6C@QoDG8$L}lkTFKMW
za)00*j|L-1&b#k4+*Mk|n8);05~NsUe#NnmDCl|SYdU_K&T2&62A3*ey@N;}dN*8K
zlpk?Bg5CNRA3E9cV`1Nc<r^KBzY@y?#_pKLUU9;N#AD)hlH)}ZnfG)q<M&ERglGsB
zX`J|B%&6u=S$P?-+M&aO>#zhCkJE*YFEfV0U8ArvTv1^6lv!PgGH9wCuJVXCB_`z^
z-MQBNKw9VKYK1|>MQ+zdW$kyY{dAzcO9c-oA~?)<3*z%=q73Asc$1J%<tu=4q!BG%
zw8G~gOf=PSo|{OoLm{}D%h7S=6P@4(Gpx8Vf?zG4<G@h;%^K(XbzJ9ow11>CZSI8V
zj{{fuAGP76BqhuJL^S=Az>g9Lk=uZel(&^uxKE@1fg|b9zs`6Bqfp&!mcIjN`028J
zx7SIf$$_Ob=h(?j2J`P%O-C+4{i1zDFW<!28jFaP^GauC6&?;+9=c~qoNUa3?MO?+
zFzOy0!iA*9)u0wVkOZI0z865X&5+C(^r_X9|8Cpm3L4CEat1M9$KeZi%;wnlqlS$q
zTPN1)RHKNb3Lu$t?R8E?`$!W&s|`d2axcRD8{^a7CXA5M!K4r5?w2moBtW*kVVq3*
zksaN+nw_KEo<*#BAnwO1<_c7dji*UIO1?rMNz=5ki_Zj`TMwPtR4f^g>RkL%;?Igg
z%5evnSZ}(bnV()AS_PmJpirry^@AC{1s81riB22GEF&4RSJ4=)9X_c4fXtPa@9)Fb
zj!J~Hb)v|S7qg}36Os`%+6_0e{ivi1OQ(~BZ}`tU6-rwEa57J~z`{?tTZSY~oc_He
z9A)xU%?#<9Z%ioQ5&FH(6QxrCuR8yH=EYSGp{Wvx_4p@a#*OEjmQk)M#0z`MiPB#2
zm;EvU-H0UM3<H^Ucpv;L1OOMq`=w^yQWQdtW?(#u4bunO@-oXh-cqI2c8B*9V9Fv9
zO;RW&2|#HXl&|4<J9=2Zj85aXa0VEH)dbPEr|d8Oln8$ofw}E`g*{+;gB%ie{M_kx
z2`aSyOV%P(`t(h6>Z><dCo7l-4VgiAH-oxYoo~?9Xu46}%NpSm7_lU}O0J~%iJb$p
z)nH?(DjWgb&xxqvFL`_wl|?SMhg1DdSb9>Eur;?FS#sPw{$Fbe<B{?t!Qbp^WA&Zq
z(%%TElJQZzvXAE_ApT^4S@bV<Pq?h?@CA@z0h>P5fnE<-$b6i{ZO(Q?1};u6Gj$S^
z0gYI+zFsxry54rK8Z$dbIf!m4Xbp}gxU4N5=(I#OxVeYM=%8?jD}4z-GvO#iKNV~R
zCHp(zpWz71V9k=~HcR0IR41IlN}Vs3p<1hE<uGhUpR})_B~a2MqnjI&j<hMHyUB^_
zy*Ukgo|MfBa2m{XFZc$U%D(DwK))9gD^iUl;B$K1WxS%=oIqRteq!xq=F`oi6pEh$
zYnCPb3`cW%<49AQpzFIqY0lXAynqq4u@k!bfaQsBUD+-K?iEp&yTJk7VqIdp7NPUf
zO6d<U%=$1hyJ8*x?L)KmN?nu^Ib&lT$*}gG-hy>XdEg9`3e*9@OO2)L9N#<R((Ck7
zSQ~*~a8zY!7zqi7GG$_lNBr`hTa-=rsab!PW)``}`9gi$JyrB==HD@IsDwe(HB;r>
z@Mab6+{GcXx}a<bvAp0t3<?2lv8bgr>_6zfx_1ixO8S@g&>JLU7g!5KEaqZ^aTW;|
z6qW#b=V-KOu|_ygvedt<e@hxc48~5ZkWoBM?G;PIc->`<5`*pbqCaXyWJ$Et;nVl9
z(trsZ4mmF`gsczD4Zi`@Ti_&B9{3ZV=SBAb_3RNO1>~F_y^T25zJ!cXJPV4Ph3m66
z(RS92^Aj_`)4~PM$4G%p$J|&@e`>h&1$MvAXcPX1rY~op(2d~O^G%^4V$Oc4JsDkz
z6${6-sa3;li^^wD{c7wP!E~J^I1-<?a(7V*eS4Td%-soea{C@$kJARjv3MQ6V{-s?
z*=JyI{`mTNb5gc{$x^B`l5RxF;O=r7LF)f>9YOke<oOBIL<gs(Cr}fP4@8j>Z5Dew
zn_X_i+$Q_MKW^LXH$Ol_F`k#gS91Q(Wdn|5I0T#>LZsm}d=ebc);UVe2(}U()+|!p
zv4Dej%ItPu6Mh5>VlR54n2^<|!=`mDN&Jr8zyA);<qpm`pG*;HIYz6=>6k?-elP_I
z3aoHS)74yF7Ah}c{Ck@kR*z+haW*nWXP3kk=AXWyMm<1ML7CP<lYKQj^=}!d^aelN
zoh*x=nG@RGc=baXJ8zfVo5Zv^BN$^|9MoSV3c2y|Z&F;t^u;d|3?Z;v%3zZyiy#Z(
z9)utgvrmr;+XN)Bd5T9Y&Gx|<3TjNm@0)3OGOz<8tWSA!)klj<NBS3zfzQh4ze}~5
zbS#BA&fVD{L5DF@xP?&k1@S~FZ~NErgBTrz5^{+gRN1z_`P)-#jqX;DPre$8tZBxd
zzzYbfqY@pJLb<a5(f0{p=g!!SMa<P!_kRp?-~Gwk!Tc9?>8l;ogk)D{9(8*Z914_*
zN)vR}e@;<`MqpxlABhwU>!<1#SY<ln$MVi_G1e8F83eQ4{l-!^up7x+^pu8&+U8(m
zt(94ls;D|+l!43JE+N?DMyL+p3d|0MTzh_r3B`vAWLa5}Zo$_yB~65KCw8{#Cw38v
zZa#iIUB91?FanPW$369kEn<kH9kcCaFn?as<X2ntdglU)Q_ybD<+*wj`Izg8<-bGU
z#yhL5JjCh^<JLJ=Hem=#SAB_CnkOlyN>;p@NpJ$Of9-HN7UkYV8M_~1sxv%1=E^>I
z7Rr0OPC!3c4ush*m(noq>MZ<(=3VH#73&BJ964=URcw^o?{V{QS0+{zcDMoUGbf9B
zJ^me_l#TCuU-~DHNh7jc1KUsvVDH`)M*qnK&sLk-N)J(Ju2p_<K($TgkSwq~bDQq>
zI0@_dk_1C35`)Y8W~87FIjSM1P;^_G67<t#qT@|uc2s%5KUv<F{qCpe(7e6<+KcXJ
zz@CuH1YhEN#Q6{0gGMviO#@2`qz5PqK*b}xj%ysQ!$=y6-K-U};)dkgbENz>&&ef#
zeWu62fEXcM<n@g62Ls>pt&-#tAnZ88OZR`?EdgAXlbQUsVq5-D=HX>l>y_>REI|_=
zVE){A;PcSy`Fx{^LMKcV2ZCLdLu!<i*HE}-m#4HhPLz=mYgNyCCMv!*v5Ny88OdOd
z4a?@U6<>GFdY*U+rc9CYySU8;raxSM1}>?kU~j^f69CJRpx{*}U=t`r325PB#m*{^
zA?4O0+-%Lm@WuyIzbNR=WeV-Vk+n%Ml-pGY)*~g}Ufzq>hu%>@K-rlOq3p;i`L(nM
zGqJFKDW+l*(c#>L`GO4o&rOTgEvK3%56a3LDe@HCpE{BNr>H@SDT(O$Ci4ANl+2KD
ztxxRK($($-#t(NQ(~iVck}IDbLvf$ZL*Qypl|@mApy&%@c=5qJv_#S}&6kSGa{W%T
z%w2<=&$=f{fajucz6vS7);EEJ-y;|`yGhc9z`ZrKsIKje>tdThsZF%YGe(xHa|>rq
zNmH`yd&3wPZhd<+>iz$P10Z?lZ4+rMz`}#6yP!pCul*#=(A5VxH0|*6#s+><t+-c(
zb0Ox+edZLn%ArHBH3rh3n;090oJFulRKTY2A#_^I2+#2vsZW`QAV)UdjOsT#P40HD
zkkxQTAH|Qv<A+XR5v9L0+_<?i_njdmM!!jsVo2oO6oc1Ku{Kfq&z+(@U<<CNRQuN*
zIP|z73oWYDF9gVEBB8T4E952Zxpbtkx_ETXEV79uc2($d&)qtpK_MhkYh8~$z+mm)
z^9tv{DinsQP@$N3gyfG8I+aA-b#9jmATw)i#9qkItavusw-KAIx!@&#iCnKfW}`TF
z=|8zS-iP1gOUU?M*r`XK1YRs=lzfRoLvE$=6ArE@;+izGAT7H#0%9W?<wmFxN#p2?
z;~)M%zZrcIWbU|=sJxJ&z+$=h>wjzhV&x~TgB5~)sfSlS<;}v*h5lBzwb5gH8T=d*
zj!L}LU@vGQ;%>9-|F+`n*mwT`Fo=BN$o%Zae`wc63mq%QcJOk}GrLp?aj1<MWtIm(
z8PjwuBui!u%o%NBs4ZXVFcb?LE%Dkiw{@o{v&N`@k}i(~quiJ8u<y*KsTE364lp1j
z8mgfGWB4aV&z(^+np-t2R=Sewpp<6O^a@evEja+p2g1uWK0p&x01j#RZzSz!U1j;%
zT*F>F$W)8P$bb7Cwbq~qCq}04X*UKSZvR8i-SCc1>-xw8TKgp2huDq0Ig-4dw)-0>
zWtY3(ANIe|a6sP%BGkI#Vj&sT+n=`1TTU-668Z(Ha-?dm<4@++K8X+sd;4y#-Dfpx
z-sO>*9BJ?vMHZ>1PABByga6*2+W2;Ygax0@Sl$v!*`Smrjb7*Pqr9$Fv!&mV4AVF+
z``fXgJGWWN<T@wXJWSc7sDc4(ARvgm_s_z7rGWzE+HdS{y7q_&O%t((Lm&)}G!Y1?
zZws_>Agj08e#uP%cmKy+&zLx4SQ}riOnveAmtNlkn=QkhFS!EB;YHB<#HP-mW(aR#
z1uVW<=TgeVzek9X&<3Ul?Ry=!xN9x7fUGTL6TYC#j%8-u=5B|<Thqy&xe;$tumFD%
z*08}29K2pJs<+}>u)3sR?$l<$lhbY}Jolm}*6rAs_c4^C9bSrFdvchrClV}}Ilj8F
zP=yj=1PKDqmRYM<@Yma_?6LIi$?FUyUy&q*&x-*6&jD1?#I0eRoj#}k@V)%cLf`A5
z^cy5gwz7&@hmcj6hBMU_YH#`9DLE5S_F}sA1u!zHxbGZ#{lC-R#Pdauu{H;jX(4Y?
zO0ZFv+Uo`?i2Q+&s+D^GVnwG3(g-C6FR$lCF7}o&laZD|se~Cu=-&%>u)&3JgP+18
zpmX=m4IY4>5<o^C3gq8fgB|7;4}noikTz=-udumMjSkinp2$5WB$y;hhgEa!+-S0x
z>~uMYu~x%U5aB3ZT{sJIf**Bzd=X0CV(cf`O=M-+0y5@!A|~uwIP(c23i9$c&00?i
zzsqVEc56U*E2%L(m(I4~t=f263y8?2C%>yf_ydwg54p;#XY|z0<0QK&>wPmcnfA|m
zl5y|T*5?UsXKZW4Ozv&P?9CJCLA#and%c#lxULL)k1`YlKMrEEiBq^&(Q57-A$N;`
zm$<sVhZlfUz_{(IENXPD`wP6`jTYC(DLwAckT{W2sXgb8qdcv|iEFdy_~%h<0IJ2>
z9S<IgoVwIb#uZCTAbSpZ5^r_fBJZ299y{FGPhV8vHj|-)R8u0RNK%GyTe3M4vcU1J
z!0h?*B2jXz#A0NnL1R|5B-uF&K^#)F{@#BDNJMZajg0NsEE}IS7)<Tha{I18)8~R|
zta=~+evPr-Uw*lL-bqI|i0ErQ$@QEDWQ|R_qQJh9b;myHfYy^0ln}+wk5|%<kH@}<
z;C>)>$!@*d{Gxi5_5O4QYCy2np4V&pcSYS7zt@_!$IH*{e~YIcu-W)1<Re^X{OrlB
zhMoV7afQl)k|CM?V!|8@^f_M&35K#fpgGAFg42HF7NOnE%SnRO^4)W}7mv*uH%dsg
z@%ODI*eE5gs^e?cL4n96N*69P_o``LT*y<WEzme!g_10ooH`o2sWZpP6q$6=L?IGF
z=F0|>GDxl<GQA&bHv!H~=ly};tZQfRRvwqajN^yE&rYY0KRl^5Jk?ubSo;uOygwCW
zWK9sR4L69FH<D4Mb;-U2{Sl&_#}_X7bNVliRwYltyfIAH&fJ=VgBc2)B*-`6iZ{jr
znsPi5SGgvl852C?At^*=z5*e2fS$%l8WznXG^rmSoYsqWe%z7rMFuy3D5Ez6Gt_!8
z-6nv;hIFaL)}Y00jN`6NBseu_rh9c+MkZz`rNL{?!V?2yN{;Co;Ob@kN%TTLTZygM
zI|2@AZ`%4kiwpU`Dyvkj1hQw^)A0%(OL<fb9Hhbapz~;FO6g9aJbJ$aY%gZ}z<V+)
zS!a)zwq%GjbI;POF}73`Qv&(|k`J3sZl!yw#Z-o-rW(KJw7vzG3q6I;x>cj@IZHjH
zaBnFaCCbVRghSno1MUJuq0O~DNALQn>6wfRBH~-HnfA!uVx(9W6G~V7o4s-q?C7u?
z9-BkLFWQA_>o2=yw=R+eGEGx9o&r$l`h=?6%)qbKkW>tEgCt)OY%k}-?QXVy?}s~0
zsHAH8>CgV+ozE8j@pAL204Hzln3};@UM?^m;^oB0m%#csU&}U!&A*Rff4SQ(L7V(8
zAss@!&eO}PA~CDF64|I~-uuqC1)=EAj}ZL=D@eG{^}mM9+>hz$zuciu3I60SukC0G
z_=HtVXCmY_F)Q;Rj)eZ(VBf`FEYs8SuV#-{C9B4R?}b)9+LFm7HW3kkGQM=uY_eyE
zQLL=$7YoHpRe@Z%@`U_%_g~ykCnyw65s1mD$fTl=lkoAw;@yXmj^iWKYkz&?Lb;Eg
z*9kK*9g5dCvr|*HkS=TF_vILtUf@6W^iFgkMIKp1?G7lAU3MT`?!$zW_We+`ui4q?
zxCH<>-A1diC-yyW4^|7V?Z--hYk_?2;{zDwnrxnYQJ<yno^OY|ea|})q*k}t<aPtt
z6HaVGXmCP+ZVJf81H58{>~?%{%fI&?kJsEM%hh6VSk^jBmxmP7?H_0;i4M5IAVj}e
z0o;=LRBqQ@(`k|#g<C^4aZ%IKsj&P}mI!)F)Z9M{UlmeQ2onAz1)~n{7N7Lmsx4h+
z)YFQo>1Ii1FF9I&jat2+*c3)C27)xHAj)5jsv-Z5++~ww%Cl9M9OJhi1oS1i@&J^u
z8ZKtL8b>Fu!;L0?3Kyy^z4YhN-5)gbMmhg(+=LXdS?A}B6FU|3<0O?ME6>1gN`%`(
z_9`8YRC!C+596m{Wn`^eCoY(dUDXZ6kW@&ffRe2_jm|;XU&FpCtb9q+rE@wc`9u#*
zcG`W{?~gI6KR6_KV68uiXi;iM<?b31$GCSges^k7%t+GHQqw<zyc#=SJhpIOVo5o*
zJ~rh`3o`ICp>Y-PI9)W@Hq0{j^j7tQ9u)Ih?3e#f3-F;tSyr-MR}b`|o<Fi*_E1Ir
z-9Fw(KdxYD)SLV5ZY7ax$S{$loKy+@B!(IZE|ZqOnIL3M^2B1GhEjM$K!o73%rldc
zmN@D07wEJ=1N7@QrP#(v^uWb>TY%W-2Y_>D)S#42oIgmob<!97yM7KKui6~5L%|;@
zhn~XTFl_1dLK_u0EGoD+L$|#tah?kwYhvRev@U}-bFAtViYzThU3-sNSvUJCyhN|<
zMlyDOFP_Dvp!xd@G8nGgd{VJUrsj(U6-1()tEf~Fu=CvA!h1P2rOHmz@|9Md5=N@4
z-|Pa^Vw;Xa%xciR>lB?e(Yn?w%zcNIQ%7DCZmt|H0x$;jUV(?3u&HBG34^{eH|!@k
zj1)`-dQXi6y*;hHJChxVm)%X)0BrJde<3-7{d#}FY59=z_;AtnxlPW`dllwF*$kGz
z#LZUZ96Ek@=ucaKrkg~b)<@-=j`{<dj0K^C^ZC3%Dp`c{BohuYnZFm3jwfwm%#=n=
z&h!BP)24}0W_u5fTZ3*|2_}mAYmPJqcSk)eH2jvgt0s-&qJ3#M7z@btR%;5dw*V(N
zpU>eL`hd~J-wsCSQ{zGd&Ba;3hsms6|Jn0+vC(^e?IrRFn^>~9wtbiXe3>Ui4v2}L
z7ry{or2i8l?s*t>Ux5I@u*FOUsP7BBO>t$>&9lSiNF08b+0O1}*MokS$AC@4(T%y6
z>lJ0nZJRGF24D2wSW0s&wsV~Mau56{@IJDJf`WKA`X${O%1vMk$t$1;$>KJp*;6ZW
ze{!LM;bE0|u-R45C?xYK!cRqZ1_4{|(eX&ZTNUpES~#C6(1@|EZTFkZ<{2Ygw&*+|
zA@N!C+xh(dHD|4NzHz9H7EG7AX{?l54pzH7|DhaWWh$0uQ9HRTwX+{eWCg^(1kbD)
z!-qCkc6oosQpzeB7>x372;{HQJO|^yJ^lC@W{;g}9h3F!8zo|sURYPLg$CsZ>n}Ck
zW#);HyVe5q-a;N{Q23eh6>=p%?$0qQXh^~T#Qx;NLi^Iop-vWh2TK3h4h&dEA!7t_
zyqY4EG5rvwnuwtk0GTIdR;%9A3nqz9fF%2QH3B@<`7irzwbD<HD}e5>EB4FI$$S~&
z;dB)+BDT8m*B%~VbSIO}G4=R|MK3wDLAc?FXJPNI!)S;^udI;N&8$u4|FW=k4qlOI
zio>iiLGpPqR#uN+B%I>uLL3kO@t&x0h6ok>-Pnjz55ScU_r`*J6|d0DI|_90Cu)jG
z@ZeR(M@E>lHWm^atDpy4)|yX)@Cle4YM|Co6Nu*VFExE&Ge#Cn<p7;TR(4<n_xjme
zqa`4M7(xKQ1%sd{!ny-c3)d);s?CyP;p-=bXbtw$GS<TGoX*pTB@IR>u~RgB?6`U9
z6_}r|4l|jYxGK3KF<r}mo;e~A|0799FS*6RwR1m{#>`NLY{5^dYprBcN2ZySHiA$Y
z4f9a}EZP~KWGEp-2^I<K#p}!om+}TC`GROiYijfTl2p6Z?K#wW?-<N0q1FKS64zR5
z)*UvghoS5y`-4^VQn$X^?&OOU!-L^-ut#vEa+mQi&7<h-3dn^!v5pFaQpwA4Ed~VB
zasqiEOsO`(xy8v72mI()@Od<&)V2lC1?J<t9OlDwFL;|L&h@dPtot=MAc(+Drras{
zkL|2QTYy2E=;CS%b=+uozxuu7Zo52{!DG6TSeUzpyMw)g640ynIEHDc)8cYafl13`
zaZ8);d$ZXl?L!#+SA*O+tNR?_vS=})))yCjJT9_?O=Y$WS=n&>f%8TdDd5M>TUeLJ
z1;5XszI{!Qr74(2*Z$k)8ys~*g(_9rSOXe7-J6~THkc5fh6k`K)zXTQ1(H|BRK98m
z27JYiXI|+GwKtUtt_x$z6#K13Bfqeid1O}QPbI}8?GJtqVswm(pC*5xTZ5k0BXuxF
zRi(jq&aoq?1)@E+&oY~;7#NNwbCehMB|^qt9xB`BN;EN28mnD?$j(7vAMVP>`dU2X
z9Md5L8q#y!<i}iO$b2`POO~_5dj1C}FhxvOuBc;hTxG)1JYDgn<+vR^>Jt00p`YNV
zz`Vq-SsVrrw)XqK8^_!EN2po2QzAQ*!B$ssR(rc*n{oa6PGJH$FYttlY1!t=Vc{|-
zoM+QmE02gyK<ZZwN26fr3N(K7V!knSds5dL0zOrp;lCSq>02LsgfhL~+Dm*r@wNWe
z)vTGgq&ExzVlrW;?d=S;VtbTGYIGkl(6k!TL6Cs~f^|4sOC$pHbI9Ou=}euCGn<NU
zwn|!Q#Gl5E4GxGLEb}%PNkBUI@RC!sZfB0@?4$T?I(8COmGTlSp-cho$K7EfRkgX}
zIuS+08FHQ0bxINGuYNnvcC1;bW|Ydl8t^4<ys50As&p0|Kqb}TzbWJGB>kM;(#DU1
z>+PZJlTGJZ@PTa6wMiZAFa}w;Y|Mrg)TTHaZxB-?wen|FDjLE9{aZJ+qg%TMC(Zu2
zK-9+0`rgA+$af<$DsQSLfdSFXKUR%4ou-i-RNj%)t-^kaYFaEF#$1AGsZe9Mzxi2z
z)g51hkg$FkDA%un?G$-@Yam{x_0hHjvN0zC?HuiwoQ+~dcGcAizld+c%D}MsAJPyV
z`HE|qmW&xrLLOR4cLEv|AW87NsOIk#OLX=6O2nL}M4uOjl`3z_gSv0YM^w{C!-m;K
zR(uUD_n`92<yq$4A3U;zb^=iyz@6<##$!)nwb*T8%DcTg4>LKBfWsCTd}xn?!|kbH
ztr!vyP{ZSBCV)}P^L(SnZLBd0ljiQ^9>Tz;7nuA0W2-|Vn6g9zW@<OT)#<}Oy$9H;
z_`QWIpqEWu(gNZqU_)yRYj5O;iKoNX-yYj#V51srA6FuNhh%o^f16E&Z6AQV&6zym
zFLT|d*S}~4(pT|E$Dq@=olwycsAVDo<YRwp>whcd70xp85MxTJuH$tBbQ-I3<~ArC
zR~b&2_+WH}hDOoVnTAM|wCbzX<AMYDnMldg^=~#$IEMu^HqjC=7@OE<!e6=LPJj=S
zkk9wyvwuK?rz|qw1L1<iV2Ae#P{k=3MM9PZs)+wJv;+TIiy?{<E>ne{G8BKz7&3wO
zi(|D>3@wvrU)+kO({veW!A%JZ(E(NhUE0IhZk53Sy~lz~U91^8YYPsGh$VcBoK=39
zr~-JwGT=^$OCSMGTx>9se8&FJy)Hg*C_(Z4VbP_r{jAxiCyduMHYYJ>N6Dm>Wa*gK
zhkt$-Z_pdiH%%Ccmn@#Xv(=c|9skvf#oxMWpk7WlxAXHvl*^iU%Bp^eBqHQ2GSY|M
z)Z42%a!HD`VqHDGd89{9s8@h&D^}Ob^;291+xx4>(=;5pULnQq&cp5uw)?}2eVf9-
zq~;sX+<d)z9}D75wRVijjz;4*5u#z8H;ys7eP0z)Y!IU(cts=Eq}IG#Vsgh26H!Y>
z&=yS{mzQ`aGr7wEvSv{sE2Sbp$44)YJYOjSS~ww!`<WF+nCX@*942OwJd#3t4^MLE
zK@O!H-(uPNRKnU*&f>7c<runPS&4iiWQn}7Xd-sQ94fsllnZ69qtF5$!mEpV7}KnA
z#bD#cYjQi<7+)r$cY&!lF7-d9;&zg6zZhH5^V<C0y1~KF`l6FCU@Z%a`Njhv-{*1G
zJr#5&fD=>}CE&-@zr`|@e0Kx~U#r=-Evsfbg;o%yjHaeWZ_htweN-62a+?UM#y|~g
zx#;Fw=tCzY&LVwv8ZCVexZNja>%vLJ77zXnVlPh@=qb668ASDn;%he=RCkBlu*T_&
z-`CT~f1`^YXLnEm%sn`14vL%<p<lM<;pxVsZokj<5u8Pr$8zbcFa(AXW}pmE_)Uk$
z;ZEmb>eq_CpL{V2P^Xgqyv_ei&%x;Z{P3X&c--`i^6IW_eHf680MJ6<Xm%>a(LF*T
zmH|G>-;ItHI7c>}g9&w8c<G|wl4*=MyIS)<+azx!DxBebz4#ZuYTv<KYv?iyZa+n(
z%($QPI817{psFyBDkYG@N&kt2iw^P$+^PK8GI2MQ4&wXyleX0yt0y!%D>|;d691!=
za@Id;r*b8XtCT}0t4^^22~VvQ+8AQy&|#~`b9bMfC{3x%_0rDQ63{YHDafCi;*Ucy
z;Qx*KS?P&ndNOu6Is%2!t0dnm&yc;?@tjy0tZz3tLYrSw0sS7i*{Rv|a?V|-8hW|$
z3w-}BFiAV-VQR<ZtU0(SN6vCCGv_TLT}$QH?*{QUWBl#}xWUlL*Q@Px7!p1wxocB}
zUti5Q4@8Rj_*8zr(?rUnFg<8fjQQ@FjB(+oRND{TVB71ph<Mz!7xpVf@&F*yN5AEV
zU65puw>?`$ztP7F{`0vF`bV8d!O;Fv%Pr8&snZda%;YL#`}H7T`yF@y4%jQXATcsP
zVigX<DU9;#wKH|zZ_x{oXgD{u(96ZJ?F1huQDGUl`S_xY-Z05ym8AzITb#QgCgC#c
z-`&g>a(iB+;K(>RLoVdN4(_#wVz`Fk6D1k@@IUzGf)~YRnRnv(d=Ww19)phiojYW*
z{Hq3rEQF+&s|_ehqdpx<Blyjt4!i1V{2o@t^X2-sMMI01?+s@%_@iPB7eP=i+wLiE
zEaSXVoq?Y6r<9cAHjDR>5IBHq7|<d5?P;~-M8|@t(lb~S^%)RoVj!{MpeXfkV4p8J
zg@Z?`crS?`=lW)1MIfRj;UyE6N8lF_9pqIA<ykY5?oB@c3H;v{*O{DES6H>{0e}od
zB^EL}xJd;7qCX{8qv)YH{xH;fd<1=dM;F<Q-`Ks3$?nD_44aD#l9;D3%9aagOxNYl
z9a5l^8O<80-IPIeM6LWiSdP*K@mYW1F&HJ8ph=K}&y^nYaZAICc1wm5-Pitx8l~&N
zd~)D?;SqW9GXBJsjl=s<r5jQdg&z%%n?1Fpm|ol6=D9VF&1jU%*Wp-)Yc&k6QCmG9
z+OVfp=Y2ZZu;=dXp6~m1{huu2Y?V{s<$cqW8t)z$g`PiNFA<;r48Wp3T>R~^>JJ7~
z66{G=fK?mKq!kgD<-3HqMFkkIT<R00Q;$h5{FpVEdsUip@cu`RK$nXlW%&g(0Saa)
zHK!Ou{I~e)TBR|HXGwNuyt6oA`WM9?pb)TVzU9%QtkQdY{xmrpgarTLHl+DJ%UiM8
z(b$R~{O@trfa47Dka;^hc#<z8CZ6k%&X~vAEL0V$A!}j<yx?cL!uLN)Y3U`gE`=3R
zo>-<w-PU*Bf4$D{0ip#!8cPTl&7}0d-BpMe1z}9UkTs_ku$XC!rT=O?(T`8*j!T`-
z6neLv`SqD2^rI8FeStWa(H$N$mhwDgqRa8V6>B>FT751N04w#0;kc4C*oXk+dFL;2
z@TMb=r<)@u!9+wa`9-OD)BE+FQC1c%&p5i=1~glBr3)r5wsz-^0k7*L38b$f-&fV<
zDy<k4drO7vVgERJQ<Wdg8#fB1?&=abwQo3#AgU6Lo4La|j2)rAK&alWU!d*7gp~dG
zJ{baLl%-D9;Ba^WbSiCiHuGIah|9HGCJsA_v_BEG;)rb4)_p5^BJ*vbf8YN+F=6^D
zhJ!Z+To@oTdp13_8b{c}EDoJBo6l+<S`QXH0;Zeup{^hdSgdZg$tCN?M$>nrmA_g6
z;Kv8=1P`s#u;NOGLf%3@b2iMPU*E_B*!G<|`0Gf2u#lD8s?v~=320{sS~>u{%!-<6
zhDVEH`FdFbLhZD1<7$lgU+}niI$#Ry)BWSC^Uy?)nYlKOzS<5^0kzb&q94i0dW;c+
z!dizl`&r+f%%9&Y-piA|&8>G16#V7-V?Q8$wJt6J6q5aruq=9=cPk)Ft*#Iv%5eqv
z;hrMrMLcwT9OFdQJ3@rO9Zn;ivi;I1Gjt?x-;SJdGuP)~B_1O!GTelag79%z!LTSY
zFPLf#w9>;o{R2N|#nQ|<eKqj)ztIn$Gp73zR!4wvGZqlLDw{KwDK2zK-_-Z*21~a+
ze!NGNspJ5TzD+`P!+;7>z1sf-gy;j|02uNF+$R5QY+XBu3}PeL`{dc)#!_fOR{yQn
zgnj7ibNc%R@RIEOJHnkF#G)1((FcrxRs0ixxKYv*r*KcN=M<rP-5h@D9;NKY`h~-z
zX4aV#CN@ZYMH4m<aMNwDgPw*h#I2SM&zcOAwb6<ghNF`t0gW<ZwJ#kb9!!xB<Hbt|
zvvpubZxNS0(V#I3VjAze-91YJ-sVhxzu|kr#T$r|ocJf4T~)erX=UuLJr@(asG3{K
z>BE~?(weC6;$wY*l>dBxc~YcaLrVY#jk*rMZ|G6XM*KAHF=TihZ!DV-S|u3TFtf0O
zO@ZBzLla+3u@l0H#KOwIlGlpS-PE^!ygkPd?KV@f9f`w4tZ5muLAwHB4D#u_lVgxY
z(Ff9J-UdAlvgx%cG)h<-7#_(S*vOeTn~Bz`jBIlE#(Xcp^xqMtr95!;y6DiBvC&;K
zWx?^Eg}xk}l5~mSWWLa(vAM7x11_A}NcGoTH~*0e%5!E`4|1YLKctsFoDnQF{jrM=
zo82N*fz=|F+3?}7eCV$Qi}IBl4ktj*nO?_$8{^@EO!E7i>t2saC`Bo*q!|ELK~azN
z1RuM@8-kF#R}JAAbjjmc=(d=jA$!F}V=Ci2ghkr)Fr>$*y@`UYJQ1`aCTv_YYCOhp
z*|-AZYHVnEwYh_OX<|6^Ne`ei;%KVI%((T=3<JOr_|~B#)ZIFTSo#^;sX=>?bB<#P
zH<^(CsWW(MNWm5pO~A+kco)cJPdz&TLozAc2|EpcKGW)?K!|)9=+$;*-A4NI_~5%v
z-!lUBd<Da5d2=ctZFTMpCFAj<zp59oO)xzG98SQVpiQk2x7i66TeEFZ^%+003W~(*
zoqK5g?cv5WbH{W&_->{fp3yMp&XB_L<+6Xo1#Qq1%>m;6-x`MY)UIMTIv%e=P*GTP
zn#}+i*17IC^q7N^56Cnoecy*k{w_%}ODtgtid;3MY^N6rtvrasOHjZcm6FB-6{tJ4
z4t+WdN$WYMZW<xUoP(XO!ryhIVz!`Z9m`a$DFjA~R~gZw69<Y)IqZERis5sKbSLXa
zrN1f_b7yCxt>Ntv#1FF?^?HB2NyF5Gm8?x1*`bCPFD|y_van~4ZvcoOpqV1%?dEiS
zi7kYL^|YFg@Ae1Ie4q^>;Xvb?+uCZoe7Qf3lG@F#1CRgM^xR5P9|Dv#0Z%)yE`WW}
z2r&9DM|r(XpRWtHU1kW{di4GEJ7#AWd|}+9sn+kkd4CiwukAfgdYg18SOYPov=S*z
z-+k?t!mbA}eHmtivXvNzZjW;rK!}O#?HstFQoM+DU?w2Rhpsx*MWLmymhTWt`xPlg
zi=m2_8Q(@V{i_E#VV&Hq%9p`Wmn$cxyvYN^tgU%((*eLGi-OMwXJwVN7*S~@Io_an
z=QOuB2Ej&8TuY=va1XbmT{Qa?Db9K;XmKlDM@nJB1mDKtSF9FU&iKJLZhwaN1{3HS
zaR$jr&>JdcUj9@d#54~LGUKX2s~1v`I_Aq2cTVzZhBG0U%|VCW!F<|}5}cbsxyvc3
z!|BnG<B**pi5@g}Br`i?WMR#N!f_)zn>4!=t)}dA3En;d0u+Lu5%r5ACWA>af2$|j
zgF?L^bX(@;u%k4xV9pXr%W-y={{oO)cH00?RSR(Jwb(At!%B4qYxIDSj_r+9(R!$Q
z4T|j4hI?Ih*JGoC6)qvR=PzL=X?N_3EqBDG)WDsfdE7U0^C6}zrS@RUPlM=c>HpeV
z)o0YG3dy4hJ0$Fjr!|S9J&5yL^~v7EkZiYm@JWk7Wt^|@gs*gUvKCAmbLc6av3N^H
zPTJ8}(*_szGuco>@@Axp|38|}F}Tk5>*9^ACJh?fwvEQNt;V)(8;u&<ww*M#ZR@?C
zcjo^kGtKnFIXUOLuf6wL>$hQzk!jpo7HiJniyzL|=(4)y<yw*-TQOx+BNr`P!{qnZ
zY27<&jFzvE;l+9kp&WG_VDQ?N7_@M2Z@!Q&erGFzg}14;I@hg*(dh<uA<lKksFt8I
z`IG96P(y$#?5cOTUDs7;f5bw*n+ahx{C?)PY6KH6*hdP(!}mQnPz7aYQ5U79^R8P{
zzkZFpu}fSz3<-wdG+7@1Q##-M;(E1&$Md|yq%51hKiO1j&FZwholL#S<9L`w;W4@s
zh(h9PH@aK&07;B$slTwFTI&{oSO^Zd2q32dys(KA7x{w6j}pK-GH1450<u8DoR(;8
zrC2QdWuXV0M}P^#P*nayZ4_)mGA?59*zXlDzAs0+j#BFs{2=&{0@R^sU_h(=a7g9R
zYq`GE5Tm`>{pGEX>h@xS_C+&;y$dHWc?}SLtT!7>fKV-vBs>EmQ@<ISf!CR5+YM}3
zJ6~xne?6?UqmVxG|9akw1tv(OP%*Y$FX-d3`1~$hI7`*K8(q$86BhGt_vfOqD@`hA
zKHjfgj0X=0XY4sZ<s=auBc#3@+)K$%<||MPp@Jr*-^6y9a!C5vA@RB0zv_`ko;P!t
z!B+Uk-ygaMwq*g`%!N$G@1ID8(O6Mh&CgIugdON-lj~&?cP3{v?9Ee~>Qye`VV4VF
z#F<fm_5(+%P?b(=wps1UNIWvQbfiW-E^%(qANJ1(^3XAJs$01|Y{MUg;bV54kb+hr
zLOr)aFEw=Pn8C#RX$<8OrO;`Txpcr4DTnhlmfeoDzc+uy?4fvppCu1I13T)J@>!V@
zn<Y{B!vORfnQch*{K?8K3i2!Ee>M}{^R=t@m)K(x$3IzP0w8t{qtLB=#6z28ut0~%
zI1gfPP`3zW2~Y)Qn)Lv;hCf^J{ocsot;Wh3h{#RR)sd~w6Mutn+&fDnbob_err0vi
zxBXGn)Z<nlyzBkiu$+ZhatsQ_9cHPjkr`7cd5Hd}TU7$dhCvI0G>bF90}>=Y#+qvO
ztbRE|8cx|jDxGZ1d>6ukldWWOO3L4*)g5h@<Eeeqd}y(?7fqxwEYG7J?Op<e5?P2v
z3KpF}Z27hlx?M+KqU2Z(q%0=0A|q=wLKfKr=Nis#@0!DCi<Ih$hd1d^8|Y9hlvc_B
zFSS|~r7+}(IoC-w2O0h|2MuDTl3m2500lhj=DT1L2CkiyPw$0ui{C~z;@g+raP!B#
z;fU%8gfgN|Tr+jk$mk0Kf@+DBQ-K>tU4LY9Eyo)re%4)M0yXo#r2|BpXtec&v3=gI
z@wP#}|1o;z{H1bm^9->O76+WNkJtWK3mVaMD!6eRo?D|lBm!P9!po)%e^i9g820$8
zTnAOjxc75^!8jmnvbmW9!-?ADxhm&H0D1ARTFdukOoc3>4xm;6wLfrYGM1oznh+%$
zd_Oh!4igpr>5>!9aE1h$IBRD~-Kvb{*Gi$6mtN}wS<Q`?UN7VX)MjRck>OwKV{0s#
zPAu^g2e2?UwfIcL-eKg?4m*9pQg^1)o$lpi^Ao#|ck9Uh^83C0Z5SSpfIKA11V<-P
zW%>wM$pApgdad`-u4v?NG~QaJBgU$J72sW(uU`U!_)WGOF2Ic2Z|f`8_gz?*D(l;<
zLff-i?$@4UE{RTuoxy{aty~5}zEz3reVRZtpZPVg;Tte@*q_|0_!5NW&6d5a?F(Xp
zq{ftB=Oe2*8b{frjtA*U`cP*`gb>XF-3yIQ-NcTfKwgrvH5KcR5p&D?ZG&y%88+5}
z<H%7=%-e6Q=`kqsmUNk^nWKv8Cdnr3{R=D8pc3^)qn=McCdLIQ>WpvZTq?_B@VQIP
zk;o-c-%g5%jJc<A2CdDUFa~%^I{p2TuZ2jIG_1#N6;|$_Vf0bKruYqe=Ul^SUILB8
zqk*>zj49zQNR0KJ;8K)MDt76=SEf&57z(KLtRYwAPVJj($L_{Ti}8YRBS!!ea23Es
zwN$AgPVbj8d({eRNt`WeK^(T6>(2kn^{+wrF&jW$>R){|<_h%nLPjwiop&ew`hY?*
zD$0LQC&H2|WJt<(FF+QSfLRfzO8nq{6;)&n(zJR1!#!i@Ky7;jglUGi8*u4nkB|g|
zJyt9fQP!o>CerRmQRBq`D{YUbH;6AoVH(vaOgP}^34ne3i7#ti+gjmrAivX8$$hz1
z-48EEMgT7?nnMzF{0L<iZ4<h!TEH%TS9aSV=%iV`%B+d4^(SgmOvJy+0MbfR-lL{5
zW7(~PiVISrIxz)T*8btNv9OQXJ2K_unQdcggln%h!j_CTY1Ft;nB_;`Vb}^BpGoyT
zM!W{y^!3uTT$N*RTsQ4WV{@fTiOfM!5kBmA2K_hHb`9^N=NlzHU+8xcXbC`^jnETj
z3Z9yVK06~+BPPa<GL>Uy3u2*yg&YaFh73UQB{+9#ec+}s7wFm@aWZPXv5Q<pffXmw
zn;~^`qJ!aOx;hHbMN086?qZSjJ=*N|e&b8yx9tV8QzSg<fbN&eVz3K<bOFuh4V18C
zp4D@R1`z_0^9DKp#I4YwgUs@|Fida?x>U=7>d)eW@%A=dQJU)fz(t&KPAG{ZDDM8*
zyenZ+33x}L^9T*6vMi1slY)I~mm%EnRButuLJw$|OVxQcxCrS$J3-*hZ89|w;OBJP
zuG#HF!>%RoqwaUui^aSPWiC<Yk!8vz#F;|R+}kZ0_jI|^Xfs{`d=WWU|FYTA0O7`y
z_i6pCT6U$);trtbl2$*(`o5xa2ERYg%kh6g^Zb79u0~P(00!sJ{a3!ez<LMbP$YVd
z775wt%>t_>`-0)Fhr_ade7sf_oq`e%7ouzxI5Mz7%`S-~(k+UCnIWd@S<9uW4Yz24
z#(@8M0jji~j@qB%uzzF`F!O?@D3ucZ!X}~}Pl50oMl<9%!uNYBTGo7Vl9uAyP*#91
zT1dH}3L*OTeW(U)>P=40%$(uvJzpIs0&WJ=xh;sbQf~Yv$-%;kOI7-eSRzskf@!E5
zIYQtU=u62cd}KEn#VWkI`HN051a}j17_Kh4Osp<jNt7S7_#R6Zce<hpQvt>|C&<&N
z-GNUvuT$wNc_l}2JI{J5i)A;W5bf;R3nS<;S^DsizlKI0|Dw_X{kcsG7F?nzS@<{&
zB<MwPobUi(w)e1ku^<j{CJDbiC;BQhVrKaIXnuaBZv9slZp;HIfgQkNS<Bgj19U)m
z;Onef*{bd1mTEO%KcwVlw;f+cmi0mBShwjOg|$cgBZ3Sd8W|{s_^6dK0FxmH8&VmK
z4*ZV{l-G^u^C~bX#Q!QKuMzpojnTOBJ!ehpj%ALafw|a;k@qY|3GZf(V*<Q++`cJI
z%3VLjlm?`mj6?HmzPHf&Y2%ysAooAT^O6f#-xum(FmtcukrDx96nf;wjuRXn%x(3%
z;4}FPghCysw&lbcmzp|W&U*ek1WCI%oCjYLSK!G|Wd_FE!_{)rM(Z8^?8aHNX}hqU
zr`HJTjw{UItNUtobgxLrTyAzu{I%C57!wUTj!^sqvak+vyJycOg-_!Nol8mHuXDMx
zRBR;it|mi7xU{q5P+4UpkAf{>ge|O2lKVC;EHjD%ySslE+3k-LS$qZT5HlCEjcX01
zN<hR{P0fs8_-9w#WU>MmG}4dgzyqA<VP=kdnP<-yF;AlO*->yYz|#VXjnq4J2XBQ~
zIW(d$ushCTsVnv;yl*(Su%1)YDsdinH=XNGaf|NHPMC~P+s0oft409Ptg`HH7N2v5
zre6jg2|Q=<su<{gC|FQSJY<aJKiOvPU|@Y{H!u+dMigYSX)^%OB?@J;-ZS>U@YMTd
zJ1_@Dc>Vw?KOZ0})&2QU{hz^h!8(ZF9W=GYW929o|EO-IJhd7q8?A;KEdI+cLxIzH
zT=6OCX4e5hVpMXu#YOD&S22&Ur`n?cM9A-f5u5C+E7UJmm+KHI{?dq<!q81RAv=;{
zAwAYzjCyX$muI2k&>;oU_`ZrMQjyXK4G_U%h$KP(3Qc?&I^5?EPfvgH3aYgcZ>OLL
zy>)iJZLYoL@FeCckfrMLu$iG4y)rsT$w(BU$^XljNUDYZ<+^RMO{PXH<iDFLQF)Ak
z6$hO!L9_L2&yv&Pvqu&v3&ekV=-|0d5vK~~PPf#V3kFqkzOAtqq-_hxMM$tjUR7ol
zBD<>-qza<U`%}x4<$t={BYV6%u?;VeK3DaV!k>Tb0)e+o8aB;bFNA|GKY(Xpb#$5f
z+U@oNq(r3Hria$@``q&jsl+*cex28M*B;7`KhSsn=NlT0;cIdAeU8f~i`2P=GGDLD
z=&UqfOrt?u4Jl>M(1}ZiiPEs7OtQ3RR<@U1**pyVb3g`_gy8tCSoIbHh`>R88%kKD
zV<_!DqSHS8@TyKEpJsy6@Up10T<*7w0b9ovas<5!`YmLoob{9a&rAa(X-keo#cIgB
zlA~oA3M+3@Q8L^%xiDpy)IJ@`eFcutu&z$Nkqp3VT-v*#9HPaZV9OLnTTf_)>B1ZE
zVnEMSQpkl5Loekas|++Dz^$R==-AJQ+9~-f@Uz@@3;T914`?A4{;|i2vG!U!zJd|v
z(cXi^vgewAd-G;-^_y7KuY={f#R%(?x^~bf`>Wj=6!7SC&I|5M#|(;v1*h7sH0cCD
z9!ZA{fB`<Yw#_IMCRTz1=r69*s90ll%^GfS(QfieVIa||z18*m4G8@`9M2^$lX&YL
zcj|uz06CG>bGlb;p>h7bxNM0c^8yt7sD!^Ev_{P{s73VxDPqmX@joK1OY)#A9nhd-
ze~BADOawSHT+pU4vd7BKv!qk4CHqAnV{oWcJddX=;Q@PfFQ+!3<deBFr}uA{P1`-I
z!+>;{Q2Caw<IyGWVf~(5&=nCfHCpLh!NhTk$aJSC65sQ74=~S{$>9he-TJTGTWX)K
zIAI2c=8g|ziy|E-Ebkkr6jJ}emH0)!yuWyn>%AkFR{`n^eXqdw!f4`E#TJA8d(dUk
z^P}v=DiS`EHNcqUcD*bsZ7ndxW{4a}J%LYx5|`L3oI0JnwO7=4*oW-m(0|Y>AJ4!=
z5d>l8h%-8Z&7whiDGamZ7URp42I}~o>`C^b;9=Ag>G}!|l=&7Hgci>bc$^?BtK=V9
zp>94*8|~!JCO}OJsQld`l$)f*XnOeAK?C$D(m!A>7Py40RZp_%31LdIqI%fTk7}z?
z`j|847Xk4ch%f*+=yEo0n_=U_<@ZAl!y0$#wIgHo)iZF>mc{ug`&47DL}@emom!$;
zVk`t6``P_C!9X&(EM;5sytSRkL}~Jb6K!2qMv#IhDLnAy=R3%PUqJ#~Rt$SL`9vg*
z=Pa+@T1`#>jP`&Vo1MujZ0P#{F_=v@Cit_G-xvLhxkK;}`LTHM0G(y)^TB^=6y4<k
z|JTbjo=%&K<<bsNCf&En%x<buj|NJzN3w<b-PP&@V5~KS<*o}ONg4m53JH-Qo>Frv
z&zJVkl4gxNy9?F+vlNM{>J2h`(YUJfd>?ym6MXN$Loen8n6zic3F>C83pS7ov!T7-
zP#La^?BwTA;Gn6?J48j4n1?9ZGTE}6;^A=C{3xtsgQVMuO3+Ru;DFo{Lwco0&5Xx7
zxQFRLeLL@W{GZW_X*f_G2T?R0_bGjUo|-nzn*bOLNXtO{ILz}`7LojA>W{4^PB-nX
z3B?zBgYdf9^OX1IE4~>cTWTS^2{!KMpU4+Hf3WCP#|xhFjTNaNWmi!s#2&PMErjLO
z4{^*s5e4v03=|Jv1LKuQ1U};IFZs{b)9H9G<Qiv?@A$iW8fh^B$Fucrb5`^1T72)<
z6tcOcm9`v?mwg$mR)Dc>tyag=+6U}vHLtgPQ`;F(&;tesom<oIA`Xj8^O>@=#|Rc3
znL@B9!|!W?`+w0)r^t+%I`0UjLx>oTc`58rAlumA`qXnCNUBXd-%8;)^%c#IJBS^Z
z@}fr!oFRk&{TVRiH!SY_F|Gz+3S+Q1CA+>)qW{gwme@Hp0jOH{eh#^SadflA-s<b~
zX-m7+VJ=9m&GkYoOGF)Wrg;lETU%@OPlwT&z+kYx`@f5--nI(@vB~!XF6tV8)|Z6m
zbCkThR-+PX^GhEHZS_MU6t8^#S1Iy6y$`^ks#dAH1O6q?Cw%j|>;sg<lqs%>`$Qrx
zZ;Bs2<!!}UB~#g}&J^WY>Qvd(jLY(x;0TQRhkxi5Hr!GMVR@jr;xKwXNtJ|e9=~t1
z^Ww$_#;5HgO{B2}4v;ID+WODbqsoYabYq*Ae_W{8x)%{JzWqo}7sCk-@mQ!EWX_l%
zrTVE38AA{+jdI8^ndhM-)RJJ!2>dIyaPm^KMc?47>*L-UO)_(O-jw7{WS{9LgfdTS
z5JQ(@mn&J~(RAk2p@Vx$e-yk`)2etZ&SF<tbgwyCH0FVz^`|)vof#|fnEp$F90C=0
zQr|6cElDvK`IyMaeouiQ|Mmg(Qb(7cA&~%AYUDt|^5zinh#&%xP@M06PbTxAS(Tuj
z>;C_H${y|uz~un4|DuI0vzr6O+}`6WD1drS70;C)YIJZ?n`{UqXyW3^5D@{cjMVMk
zcXhncD>R-ifiN5{tlk;@Lk9mF;^h#?dIGB9wHq4aUlWDT-%--;5D&g#sL4qV2bx2=
z=?K+^)a){`*k84z;t&(yN4taU!w$#ho-TBKl%tZ5+cMJrk^GjfFmEvkV-qe&$FR38
z#dn6c0Wk}HN@ZDTwsk#%SG{HG3bV`@G&cf)M6h^3-OTNO60M6Q{?FFdsw?iE<~5nN
zN9GVNga|K*?a%Oy0bnVQAP?jh)r+=&hjy@QU!-$WKo3{C9|@&es$*qQSLO|*5^IlS
z%Bf_7q7j3=p0sUVqSvGPz+${jp<T<{RE#~k<gYNZ;jLiB;dJm?nLA*)(ga{%|2ab4
z?^2+U@R%sH9a<DnkNv_k5G{alcd0d<n)w9#X%T4g8x<aUI4E)6t!OmIXjW^{5mWqB
ziaNLQeuPu#o^?!PC+D!_oC2M|t;>l5rdIulwUU$OCPXuVv>Ob0T!twDT4_}mFOVrV
z5{sES@ORWj=4CFK)*oAQXiL;`$Q6QzsFM(5pEU+ormTeSGP#^fwW)U68>-X+IM})^
z=I4KBH@Gq4f3}FMgyr|E(&)InAJ+me`2DiE+w*<e_dkAad3Coc%OJ2$I6;~Ou*Y^f
z-yNRVN?)uLK>eNr1jzq@qChsq3rH);W%DT6niFh+qDfiTsn^&S%>xuo|AY}yD{&`J
zdY!g_^XZNpKjz;OGu(kR={)$qL4M6jvf<&8r;piHux4~;q+^(%`@sw|kbP%GT*HcS
zLxoFpA{0V{*9dk6ON9q!n>}h*NGRT#VFQwYu97c7m8gOfeHm;bWS%;8F#c>Oj52~f
zdWKLIm|KOGT4j#j$H|4Zb5}&Fglsny89KLqdcVDIfci_~TDC9(ov&2|t#4La_9c;&
zk4eHxZIlxrmIB#<-79EB4eA6Avc;DPpg^5Gz+U8I^~<9cJa&wZJ+{Si&%)fT1AIsO
z?NM;TV_)|GUg@xLV*hr_M~lWqoh4DT(AW0A1XS_I+Yx_vT&O(S-WtUw#J_^U!-1wM
zCA%f8Ny|KQ*`daa&S*ow3jc`so#Q8p^P3j{UDQ+)Pf_i0+#XZ5I+!I$7@LIO3u$Uw
z)mdr6;^Ecd$1q&@-*egh_B>QPradTuVH<Gg0FaMv_j&YNnluq_;PG87m)+&Kgf|xx
z3`AvtC+VxIGs1??;d<~avA7H-zrRh=yFY+-lg4ECnapT(@&sbN06%r5{t(C=g_ypV
zwL1Fl4Y{)0ZjPbZ>UFyq0Yf!_b=Cvlr+rIZi(=9x4&*f555s@%=I(jti6|Z2_N3-t
z5)u^{_#=6+ubs=z?%fFCH3Edq*8u$p?8ym&Td8cwuaD4&xa#~6{CJJ<tbN3npP0Fb
z-^#L)1(Ll4XZI4{5Gz~+qX*^fuj{D%iPeL|eEfUOT|fi51-b}LuT=g<RRV6}Feyw2
zwEbq02TB^%P>fj?Lk6Cc@)IlK43Qf3i#@Se6UF;c9isKFH~lTotLbd!X5TAAwsQb&
zKb_5c)Q=RqnhonUzAq36;ecke%9$!}z=}X-2ns4HzZ_B^ZIdij8^SI0tKJ$9s%Ttt
z2r`ighqT@b7BKUSp28B2iAR#UiSIy1AW~+i{~hyyVN>L5>45lK_8qxLuBd+54Y1xl
zRi6HAE$#@G3xgnd)Yhf7y)#R)j2+59W-R-@U|zR8mD}Co{T34phaQ8;mQ{+nS*r5}
zVBM=VYj5vMm4-a}@qNDDfQba~S(pKa^B?b5YS-PE88Kf^)xIC)`wuz1-lmHY$-qHB
zpDCvO9rPbg35ak|r&Mz|U0*kzv14uuQUq*danfuNt^d)W;z<tCKVMkdTz=a0A+t#D
z50jNkm7rFi>Xkr|mZcs)a03n7H5Z`2(2F~pr=8s2mtk?>kRny%`U#94ut>+*vcax-
zfG#s$;vGQt0tivhk0F^uRjO<g<USF?<VC-EuuPBomk7_&+N5~K@VH5^GzhHDF*~b2
zzrg=e=ZuG{2u+G0z->>$2QO9Xh<~9m<7*_EcJ&-dh<_qdWEa8Q#-<P*<%cxUAYfx(
zPgtIe8&m0+zjGi*09tDs?)Wo%kPFPqdp-b@Yy7SAFgqx<Ggrg2p@S$$Zc}Q^RdDP9
zKz`|waJ#II@UD#scsN~#83?^QNM76Hs96Ap=GWTy%g-li>n7L8Q3USW=No?K)wRFH
zL_u}wNkWYoD*Do8N%?;!*&M(J0+x>0Gb#`<YLI^M`R#p2H<87O#ivoJU#Ze=XEff<
zvs$ucw92D*R|$ZR`~3&gudLFx1sJ^*M5<xl9WLJ_!X45BR9u8~(CXi}g`<FazglBJ
zJVor>vL0|;yPExs1F-4o{1UwdsA*pULXo_~zHh!tekbxo3hGGM+Pj&|;;RP`AtwnY
zpnpmoVRTcs?b~<&h|gbOZt8v-wLzKSTOykc^zi2kUAz`B;KYcjEylA6K+%LXU)(NI
zefCxRME0qIcJlK&H$h<8UKkz41O-_cTohRYO_U7;A{loDEX=yj)&R!;8zD$tR+je(
z-&+^`FQ>z)bx*U)#qzzV0D`k8&+-Jx?9<az#JkC-jq=yMP+K)M6}2Z$@4rvpxlir@
z4dWt__u32~c$ttEL3Qy!7kYf=VW6W{R95y=JB{YWLh}XEZ$Cf)&m%om#n;nu)T{*T
zr92=u6*zl4bY{!L?KV;zt@RUuEFq=aqthOkdrR?aYASg~^Q4Fco=9FajyYmf3<l^r
zL9M!s;gM$ZZGfzOQ%8oBA3^^0PYhvuYBvrxy+-O`4gM%=@iJ*cm;O2eYA4!m^nUVG
zS(RJM-RHVTu-)Y<Kb540P`CK@(?Q&#?Sm{*`1Tt*ZfR*fKk+)}uL9XLu#@Um%sEtS
zUA-Q|dSXhZ_(L`<lO~r|OfBD$y#5bFNMRHQ(QGj_eC2{(lGe^tI=vU`56I^qr)yQC
z3-wx0H%7yi+O=h`yC%MIGIYEd0vJopA{`<GY+|;3$JwfYZoZ~Czk!L3HW+!wAFs}n
zYCC)1iBA~w?IqEPEJs=trCmLBIiFbr!7OkcL-r{?^HkBDHV2&CE6!Yr{HSTg@N5ME
z_-t~k$(DA1GAcif@(!{Lp^kdxC-3jtg(PmyGJQaV#&e)CqvA2m&Ev!CwQt%1yk8Y+
zwdpiI?*Yettovsp@3&4q``y!I>T-H*=;x2G=V5-EJRFWg{r;+F{y5>!`DkolTgzKn
zWo7GVjqB4T(jh(O%lcgdRLtU3v24!mO<nBM<%*h;*3ko0ugPS+VYiz*#SF0nfqipE
z#%j(++tc~KjTvjACx~CBBncv#17bYsQ?Az*<cb<(dgOSKr<(pq^xvbI&-F6E{HsCM
z-~JJc$MVlsfBAz!J-`!s_uR;WT;djRO0;L}B&}a{KluPU{kjd(hWPZ_m<baqL@Y9o
z;B_V#yKdC4(L>D^PIZscWk?z7xPti~mg^nvQX0Qb;!2<ZHZ3$k8O9z#U$^Qdwd#O7
zB&g0iJ#FWLGFPZ_S3M{OZ>p8r-5;o?dOQM@0k}nsxBKJP_xXb{U%ou8KlNR!7|@-M
zr-eKHIkwi8HolTn6Z?M`@C%pyV&h;udrOA+{O>TzfQR2&wJUz)q^u9;O|xt|wp=l@
zdRIKpku+UwxtjXc&C{N^2WYo*>f7z0)Qw1=`6W%?Ksy&#_ix7*<eS(nYiB#k#%~4G
zHxY4*)KwdJ_%fA2$HB%)%uuBajRUDYTPem}TamWn40i94d*hOlskA!9P1z=xv!~dZ
zXI1|ar)07A+xb77UM8^WAHCiE8n7pu#?@B66@;1P!-#FMSpJTUHzy}2#bB{&Ch)qX
z;y{N)nYQ<u=kWc0We$X_)~!#Go-=1)Wx+a{)?#@2c)=*6pT9VoQnOucFge0#2eHww
zMO8j+y;~%g&FZqNrO1qS?x<eSDYvyQCNH(M?o#x;klb)ymv|p)D}Pu_wZyz?p4I2l
zuz!_4;GDycEi3J8^7$BIwfYyIpsy`na~HV*6rt~K8fL~IPr27~)UgjI?e=@7>}a5F
zI(1N-Jz^c*7Hk<chFKW)GRd~VW~`?#DCk!4eJ>@sUayZ68tpF5Z83-#BTXvV-O8?&
z{$1ij|N5U|LR(1aFXox<W<rUqMeXO*ePvn}=}>ZLUp1o;{3%DVVFMYGF>*?@Ug+k9
zO*>c%Dhi!up(2mPShf7*b`VbzN}zI6A%qz6tPmP61(yoC`ql>-s%e{%HbzEC?pT^f
zk=vD2Oyceh;Cri)=j%UeR-*KNMNWqQWeo73xNrMu{*A_uUZS4@_*9T^I6964MGHb=
zNfM0Yct4-dHLlJ#>ZeUzL5q}lN};gEtq12$veJ056~$UXg*lQY>)N&^8rFs79cg8C
zXE)+56{J)Z>~X_M%_Z?X!j=I~7W@KRj&XWtEAz2EW)fK9i8Zx6jw@~V?^;f>j(b49
zEJ^cjUBc%9c2i`nTp6%AG+HFT2pse!(f*~wRX<2Bxo;qeL`{1xEvuZtfL{Ek^L0B8
z2Gj;=z;__bV0uR$)IKK{_jaP)w!6O5YYkX8YfR*RcPK*x;{#`3v%|YD#}C4STQuyO
zq^+oO(E4DtB$LUMd`Lh;O?k1{X?-82;Aq<2>@4+iAeCB#$S?2v2Xo6usowr*N+*vq
z>U(nRqjyiWvDMmAMyYJ0S<J85!RnILWeCTkr;}Mn;6GX7f?ouq^HHh)h9=9E5<o8Y
z0~PR)q7uNR(j&MDQzxo?fh#5(tNghnuvzhpHAiJmf&7g~C4k?i+V>&@fTRAlvG|C1
z#9RzvAX*#&ku8Y>lST?hFR43G8UXs9F$J6t!NwU;T{;3pE>JO`2ljmpFDdS3&3@@6
zO!W(Lq|22yad-h{iw=@DuSrZ;vQuJ);X-F>+Ug=ifP{5|1uq^=c6MV#eiy8Ia3w~S
zJ!p?V$B!1y$Q(*kyEVV@On#I5^~QeTpjXzVuByo8wd^>c(Wk{*zCqgBNnwPQ$L_9;
z!||*}B|2ppZqmGFCj4E5;Kv~XM;4QPH^iGLYf7V7;3(lJsGM-ee5dZ#{cNaII<4)-
z;_L10MY(E~sg1!)3AIO(OXl-fdEGFQ-3oz#ulMK7_+F2#Ub9Q~#tG;>O#sM}j+NV9
z<0!zCZg*=<$@vm`DEfFqQpw_jqgeNYw%ask2v>Rn2m>I90@KwEIvp7*#hKbT;09B>
z!tKI}hJoohpm97K^Op1(r<&gLIh}f!%e8kNjIJBNwwmHDMw88TcQ*hr%D8s!Qo^!R
zq&6lgxm2YAmG0s@gCF$_N*KC7q*Xp&GCjM*<#W662MHYnS}*rpqo8-+$>O=n*e83y
zHg4;mnlbP~G%jQzMTO`}SdyxvMo;66k)}f16=Yb^nVB(ZRYR@%LuAH&CDDdu3t~Lt
z#j4N=bCO%w6ZvN1d<O%Yoi&4XH@#JKUG%+l@U2`zE_Vb$(EY~JOuACOU`_7Y%gRCL
zZO;*c1u>lq_1U~a3Qxj_fm)*mc#JLf&}}4=($hK7=+AeHsIl3tbJ(0$QE|~`cqQAT
zvAJC)PgxfiQ(Ic$>Rb)rB@A65%A!w6wBUa0(7lKSsg%JunyPH?AKj^kunQ_+F>VZm
z8nRQSgTM0x8Vy*aSYImO%&OA}j^Q$|@H7)8wQmRCp(iCLb@4i)KhbPJ&mY3!)9W<d
z7Zj<h%PPS{;mNMS1f@f%;B5-$c~bqR3qeoZA%1QvYY$eQ+*xVs{5(CCT5I~P*RmZ5
zL*IQ%BDxU^*C&pE&uBZf-r4?y3EZ@rTDI>_iPqFt8{NK7ba8+7zI?iz&$>Tf_3B7e
zvN&FD=88~(I?w*lKjC7uq_~66ePs?vE3K<|E2Q#pc#rp-XN@1h%FA{bzC<`4(ao3*
zx`4w{Z2KsX(hfHw!<Q)KaahrYsZbt`{?Ode<)`@L-%ptUza9OXo@09dZU9TwWv{8w
z(S6&08sl&})A#%t#s=TUsA=Ypa`f)wK(zX1);GItXVlLakY*nGXn41i-2u1IO_g@(
z27QE<SK@d{Gs+knL0<|8ahoR0-|;~<qMMZIiD=e)YlIit#5rTz5=FPc&XP8XgDKLc
zc8u{395D`5_49>h@I1Ff{<Y>r!39Xrcq&d2QGcBA^L)8;K<Bm#k=QUL+=-H-2fmLQ
zhqyDPTYnzP5Lmr`bo(B{ycg?SiZF2+F5eBy<@9ATBo5+*4|TtuBFW}BA-GD>js(Uh
zQRb}2a7=jn>l0uqIIe+fFd^AKt?<>)n<%0=C;l$dmsNKC`M$q*Fu$4A<a*^3*TJUg
zXQ270-M#7eMaimRZBxBP>JB(U&6|F&v>|}5T}}2Dos;0%*X`biqn1>o1^$7s8X7lx
zcds)wL|$jLFn||n8#*c?=EarO&|npqGVG^pUnD91B^N2vk5tjF$$!+z>G3=|sald_
zgXrA-1RQIlVQRZWe321ksORLivqZg29-E0s40;_4OSa5Pmv-S4<LRagZUPlPFE<XH
zht~j`*e%LdTDBwZ`b{jh`<2D=NPUQC$oZ8q$5gr{rm%0fIOhF(i{-aPi3<qtzcT|V
z4~nWsmNN*k=e92GUcdx%k*pqG2I+X`FpdMdClQ9?rI;ZDkZNjv79(*stdfrGKm{ZW
zm+YI6Y=rqIp{f6bc1W=b^Ce0WHeN9cVZcz<8_St$@v9@Z5%fCz8P953F>8Ia7ZTCw
z6sq+25@=()ZEdo+=lup$jh(w_;J5c*v%XgdGDE`RF|@k(JCOy$2@^Yh!?|v*kLUE=
zJb9&_0|&DPO?+d*9T&ojv9!{=LM!PF6znSdNj(z8AI22*Xy{2y%zi;>HqGh(wqi(?
zFndR&|GJMX;`KAA-4I8R&h^S0-xbrcq{nSLeCoT+D1iYFGbDUIN&S?7&vm7^cY2?k
z6L^lhfC`KpDF5h}Op8~cU-DE&tU-Q~xED1iIRk837eL>&(O_}6&gnQD?Op4OPti)`
z`}%m!dXo%O4)ooguTk?m&pWoh{clnU+#V=$|MLR8LTWIHB*v7^CWa3HSM6`DakJ@+
zL5b{#^MBsUGx_sdURH}^DxrwDHr6cAIbK{#>DV2Lu0N`eBsWy*to{8l$0HJo((+}l
zkw>h<jWyG`Mf)4G&HA{+O#Ata(P2YauQ86X$A>aou5CRWtmhd6q*ASbx96GU((X~I
z^eW3)SP!vKqDjCNw?yZ2i4(kk>=?|Z$SIw>yP77J={4<tO`~y-HCwRS=~P7?3yS4*
zB+Uk#?VFXW3AT(1CvGF))=O!sz88^m2)oaOj~z-J(@Ow1)Fbvzf=uvT>e<+L=%KTz
zZ?AcDbABjtW0FajxKS2!*Kx?FA;JzV^p2oq8Z!mC67cvA#=T!_90B86%(M~_F^^C8
z0?oX*^SNzgRtNmL@VJ%>5?A+ycjIH4W6kZ{cIn{r^6e^9<{z0vL&$2^>uV?gkJ;`G
z)M~bwPgpfR8>&PTeUa`GYnaA&Xfam4#Qp<k=)WrkVmPdZlIKXs`&S5`^byRWq`AXu
zI1OWH!po|75H?Dd04=8JmP4@9W6{#XU<n((T&;Sg#sD>%kX)`{Rv)yAqeQxyp{X!g
zJ!M05#>3RDpx4+{GNY=_Zvj`%ypfyrASTaLJrZ@zylE}?Y0aEL#>3eV=sy7eY|@iZ
z^KnvN6C5$6u~2;_s<PZqti8U$44w9;9F9#qO(L_BF2+W=SY1G3PgvKDT|yf1B|yA3
zV7MrqH-mZp1C-I;hwG=TR<0MDQXybSDPv0s_q`$ej}~Gqc5+VKbfzt<`JdW!&0d{>
zbRZG^_DncR-wA9uqd5>eNhU{7HZ&3*`W|j5hRA_KdjBifxH)ydXY?JiAlT>NS9{OQ
z7yJ#0WS3J#E5hQ5E`Or?kOB|m-5-bF=$Ye}Nl6z0=!t`D?sI2a%_-$Kk`%)~qlmIa
zO?=mP1=np;tt{DXRzXXzd_HPzXMny}P1nO>W7RqCoNDOT+?&y1(l!Gqv?E(6RycV*
zgu*Dh!;gj!mILR=1ZIgr&Dg6-GQ#4lTni9}-;rhOi1X=u%0xRS^({P1dgi3<>JDs)
zlaV)7h2+0S@U}kbkpPAkiTQnlVK_~PPH}tb3h0+e*8NbpJy7OeP_VBFMKJslQe075
z`3^0bkr6E(o2}mLva$ZOiO=dKGJPrkPxccYm*sh*d26%7!to-R`b7ElU!@kit&3@&
zy&%V?%;(-Pem9T??4DM0KRk~(7~gWca2AU6GdlyM*%n5lqp9?E_xTin0%kbI_}4)#
z?~+<J`}4TJs8l+IuvV=q7m|v4+SCC8G)z86lt<Uzgxf~Y7Bw_x?c%|9y@oMf7EYw^
zMr*o-YhP&8siZ*kXL$T%J*@9O+OCut$PNbmVue;aB6aROZ!(*(d1yDAw1fmFY_oti
zar)4{nWcBqMcn5km<85NXR@~hoT4Jl>x<B_t1P<5*giXOu#cv*#fO;0XySV=tvJRg
zZB4m|?L<7A0bBYV0A5_QJDha`r6!{~nM}hXsvU?rnGZ$4ef?f`{QiEIEe(bIaoSVF
zT=GqNb55*%pb$)v&uD%LQ#lum$DQ030xzJt%Wc+>2@L&=uAxrl62_;k{SIS0_escC
ziTPbj=Mp%4jSW5zlg(d_SDU=I?pvP+806I?nUOp}){2b0Et5Na5TT#?m{%L=-q?Yp
z#nVhh8~2pjG&0S|YvB7+-xH%d<h@^7^ZNn($M2@)^}k|h?-Y79UU13~vDkNp1-;;X
z-d&F!^P>^-KT4*Vsv29E*7X~p2ZrLBE8^9~UbLuNQHl&|Ie)?GxT7~SY=+fg^WcXL
zZEjdPhN0#N93eKouJSGK5Q~#{aw$+siqK1!u%r&2%9kINm@$XDAC@}hs^M}<tEs48
zC<-gxoj03xv)wd~v&tjtH*w%3oIgIh<|)5~9ff#)d)T&r^V_MO)z@q_d|2naJ7)6n
z2o^v&3bFJpj`uXL<6_zSX-t6OEHw3f&-&m6aKu{xI_f84D*U?_NM5gF%lbM-y3=JD
zUgQ2hDe+MHeD_$FZ9_OWFu~CaW-Wh@tO}Vmb2cm34aP*b&Y$l`#~ZHh9~Xc0dd%po
zx;gGu&Iw>H3$<8&^8|TCgT!Bgw0S}yRRozar~jr)AMO(DWEwl8^0oV~0mb8a@9h&M
zrl>0Il|#y6cijZuzsU+*UN0T&zG1h*3Co;k0mCj(u){ZUDwVa)34=FTs%js5k~3@g
zj}x}w*kAdtb3;+;Wz+pR<%%@Id7dkkW=pkAHYUjMU<m{Hw@{S)w-%|%a<B|#8eyyb
zN~)JvY)>AY;8;MGGI29wV$qn!D355Ose<n*1PiWV0+$f4n%?fjVKVjY@xKkLb3x%?
zIc^CpTbPl2rVD>X;m|L)nyYK`ZveyCQ08B5*Q?ETSNyBu%dPIO@oaA3Vg?vRz1@If
zKMC32`vHYMmC0|qKoNl4>xk%K*|RXJl;YFiG=1r<c3S1okJXbNDzOTzzLD;5qS#br
zaARZ_MeN1BXyINDYe8BeVLqA01YkV`OkVyofEgG#K|qqY&uI_u@U{Y{d5YePZusVL
zkip_TdeCjV*p#>)ud#O4Eg){1QFw$<NhMw~a2uVInt_Nhj!w!<<z{WHw$ur|n>*dw
z-{<W(=TpE<>xHd;a7?td<i}}9T(heF>IqD3wylMN&{n$5*zWH*UwYYX@<go0r7e8u
z=}hiwaf@c)bMi8s*<R&BWCL+SaHFwLr+HT{o~y{i&@gCwM%(Y7<wur;bsD|+up1Rm
zSQ9Q=znSO>NSCHAP<AR3Ac1|j?N)O|ovxJ{vYk306QDd~+-P+acW`tGwWOqprPpU!
zGj2xO_wDdeTp4+owkIw8nW7$t+*BU;C))I^0*m?W;k47`f{OqP67(W&4ZCJF%iH7e
z=i|i8nB(Go)y?U{Q27!zHbQX`KJCR)`4e6cNeMTymYu@7c-F7_Rh(U!`o$eZ)Bce~
z!eE=)NTuVy&^#hY3acLP#;iK*9!<rj^%a3znWe;eW&}-BrB21Y&K_4Du!lFWgFc_>
z^(p|=dsyV`0wvuP*tc;`v8NYH38`?9P|%hCGRSCDB9rZM;ZY##3*51AUL?2hwfU%W
zK0Ah!+aPkO$w@w~U&zd0PsrE%gZ!W}lcw$S5FH_5M(aDLg*r2On6cHQXLKgRAHU?I
z+LeoISF9a>*|eiBa=r!1re=y!Fk~y-YN-4YT)}Flz?E&xY1uHW9k;<30rwKj*09p7
z*5cvhJWJ>RrHNXB?Xberiwmh;K9}ulk~o}YVd=*Qre)quHXr*-WKU;{F$)(u|MsGn
z^tRfHv|B$&Gf9f#C0!?zWlS7fy{x!W=LhPjs)I=fk}j+x`<|u_SWY=tu;SFBv{hxP
zARSvve=F8H4z@_IzoZWV&0ioSk4~3~m8#TACRP~IBF=-m<I#UL;_xGos)p(@yS2+%
zFW(l+x|a>8V?TO<hD=amEOZY+s;x>}y&XX|B$%OxN{dkTs&Dm*+_>cNK0H5^(K_rx
zQor|+_zZmB`S5tiVcA)fOQGY|QP_EK>wddz_kD-nHGbx&k#zwxnDu=HtdCmFefjdh
zoe2Q?Jj@Ph;5W#X9!_=HEC<4{=1b_moL>AtpsE+yzqf%W4z8AaYHa0nT3S^b`11};
z6E+u^lFVCV*~k=f=sPqj6!IpH6Vq1hw>9QYrhnv|KPD(;h#}sj?A$_4PB}sC)|Pj2
zJDC}>V@vb4JLi5LYSqC<E(n`rM3+d=L`{HI<J|&l5Dm>7eg-U$DyMu#IlKfA_)8ZL
zisc_r^E!jXJwkyxh2`jHbBeuY=CEa}IIG(8GGAtmq92>gGb3ZyJy8^=s4B<A@qr$9
zO<vwZ2z}~H&=&h8jYf{WSvW%$zh-FH?$twbv6DC`B&|;w-)dX)qo<n%%_J3GhAVS-
z5Y3}>jO@emD>X#o+P+_XVetBIrHt!%9*il)^5VgWr-?fpUFuXNr7F%rDz1D2#FB$z
zxy8b8h$>v=pXCCRFccn3fSWkS(`S7dBU`cE8nf6Xlhd0_-m6-({~V)|Ei^;Uv$D%#
z|E@pRXLoe0j3_i~BJYGD(sffTFAg1KN=7wWQb$C*B;HTab84SXAV_;$4zi{;1`Q|B
z<hTPo8QIZlG+|G{atGT(Han)lXV<<F6WLoqn=+5Ne?mDems{`S;4?58==4!_|8_;W
zDom1<(<1|7Y{HsVeS6dIwVS>X1;?#iIoBbtTLK$7UPXV}Pe^nGYo;#VyNh5rr<+8u
z9BuIwOwJvq_cOe3j}9MReeelT3fVLa+BN+-)LueiX>{2GW}jI3A^|^`$9(RCAdJx<
z$v`cQf^#v=3YU^eU2q0Q4IG;pO{B69J_^D!W45B%8RPXXqmuK7Gx+0g!-c5(!``29
zQ?XInXuS{Z@_2e%`pfOEUr_io9+&B6;q;8RFFW}3Z!Kk~p!S5v-rsAPPq_>VW9NE^
zv?${HdzGo*_EIp-8%|Z`f9Ha~Y*zmED%Ng<+i{U6UK)!s=(7uv;Os4L`k@DxtMhqs
zi&wPquR<5@%|Ya(uwwOKssJmTLE{@K2--|5=;INJa0ynt&{^LMRdpoC&T4YceV*M3
z)*L4yDxW#@foP2-57^S|A98UY47=|5x@7^1x#b0~{gGN3Y_`cPM#txxqxAI$K#G{k
zVYOQOcbk!sA;RNxca(btAp5p^!QP)P*1EU3?3B@c&HyfE<C^ska#^~!;{upHJ;VLp
zH=vuenMiAOJgoY%?gu&wUH6j%>n?e7%H1$iRcgM8<_4j$7%jGmK&hE}6DD`J!q;oK
zdx0<%W2TpsCK1WTob(epd~{8a4uR&d3*eDbARgu6zWz)S23L%QN7K6u{vPI}OOcH7
zLx0bGL8PPp#l@84)dAC%KBO*?1ka;yxfTuZ5S7fgF6n09KSpUZtm1G=VZR8$?Hzwd
z<2nghDh=<luzHI)8%S_ZIx)Fdb>Z8QsNl_)>1Q9g)(<_#VE<13_ao%B0CBj*k<FR2
zO~^`%D{-yGfLKQIyIkGGQLkM?<Av4h28{Tg???65r%rq34FyDKH(~1A<wpD7aM!c(
z*YI2^%ZCVn6d0DFFsN}wr8wx(oxFh4Pho0w3e|#e4wn>u;Un4iM7jB~xw**BHtln?
zaYG-&@7Of9lK|RVvxM|#UOWab`1tBN-WY=dc?{^?K_4mG=bbBi<rVBhq(!ka5bIYv
z_kMxzU*O_a$P(tDj;2}p8S!$xYsm~$<RRwz5M-<@yg~+I2FaPu?@^q#dh_Jx2nZC|
zOi!AtDzqv+_ppUZf5YAX*Ag;n&07L}=Jg6=9Y8LVb}aaMy8NS=yAZIwMga-^7+mgF
zrP=6pwd6|yYp@`lcW_(j)z%O{sQCGA_4jJ5-jetRt&u?`Cf74M4_lVBHds}5E$4i&
zo;?T}X^BVh`#iSe_J4PaMQ>}k37AKQOlx^WyvPnN<_K=yg7%s@Io)0P?x*{dJ<HvX
z%xyaJDSqdZu(^%HS9hWJ18D?V`NRyv+M19}?i9rwo=ui7x5@Oep5f4B{14B2S5rCO
z`d@<Q^|Nn;eeqK4ma9&WV6H>(JBCkn<*iOU%->dLWf8GoIhT`{rEa`Q6;HH5kXui*
zWsh<)(;2cy0Zz-9<5BvfF&UgjwxJ|*8<HAF_GMm!QTW0Uch~Vs<r~k#uHpej8)@>?
z>_dwBnzmV4Ry%G9ITX`ho={r75*(IvSW-<bhVQI6GBfM5x_WuV#V$*iyue`K_~BO~
zB;N}f44VOGEfB$=N#pmuy5t4+&CsF2s>9<f`R}oC6q~)BkjFtyAHgBE8+~<o<I`#|
zxh%j%@$);)Y?mq(0~^Q=xBZ2InU{8_s+{Uy1A<D6btn)J4!(SS$X8D4_|pJ?iG<F2
zj{uUYLE#Y%ZF9eDfmbOZF1u&}3d2V3x)mN`#OaF(7P5IP5|zK~RsQwIg(DCYKYk3p
z+>;VXAje><3DFCoY!m&IgZ{o`tl7Nuw0q1543~A<yn<QEuo%pSVdqa|H~EVB$8s?i
zQ(RrXeR6UXX>{Pi1DhhPsS-QgO384dsTtRL?G_B10BD5S=q@QsE>^2*evwS{1<oA-
zL$~Ru^qVc2tX`sf7=XS!_Qhog`v8`q1G+m<FkIM;7Lvc7x^$6YT7oZbr2nZ4(9S>j
zcd^!>_RwDW2<{_H_tYPX^xsu)%=C*t((~!^1<b&--<Ux&F3d?nb%w8$(zz6cshm?G
zGzRDGe*Qqu3#4rfxF?hx0uNOQ!p@3aylYd*&MR%rt89<alF*+!<xMJeovrJ_m?lC@
znFGsA@FGnQW#Bn%!GqXxf5&K#oLx<ju>t{a8C75%wpj2aTj;)>V1k~5C+W%_@S|DI
z<;^%&?(uo!ICl8V)f4O+<1OTXuTXs2@L6i7&&pmR4<FzzSl3P6w;#m=aF<T4xU!tc
zQOoCak&^~>T+*i4aTh=8gMn%5MXOn;3N=g5=moOyb$mRZ1b)li<B7}AzhCY}N?mYu
zA6#|q>g`hJlJU+=wIp+Z5CO=TZBDH*4FYLH7oG<DaKK-&-Q{9v_X02qo&b7Isi)0x
z&97=%5}{)ceNZ>kY{DPwt>vFhLr?4a0pG3cC<F0E&O*|v5n5ei^0fvLVjGVsdh@Cc
z>$qxC&+JD1l5U_Nb~t(D$t44{2Ps=E(AX{1m>JG0*pcUwUJK09fDLzW#A0)LB(F3}
zk8WU+P5hAv@=|!!xp1s?tJxmw4u=ks)_~Q(sp!(<<==&^jAvw6^ndl2Mc;r7Haj~D
zCIdT7=j^P-yp5lO1fa5GXeM*g*(Dd0RVa+P+exv+8{SSA=CBcq*qANUrdlrbU9<{W
z0oFLvqJaXt8#Y&AIY=Q&+bU5Q6O0~_QS~bO{h8U?0$p_=aY~$|2zNV?uC43$7sJmd
zIr1F2%jU_fZEQwN4mx}n9pPKY<RzD5)AyQV$4&gVzO*i0&skZwt2#XTSnQ|KIg{a7
zJV(y#Ew8)DY`~ljNEAm?xPQUE4pJYx@p%BZQ)(`-Z}DgNKkc>#i~bO-`Ay))1BR4m
zYxPSFl(s%VRXv*cgDf&ym#P-)kYPP$EKQ@!l?fah3JC{~-$k2c$X!qI1P5%`WT+`J
zMZw)@9pr5~Bbi`1x#(yhT#wEleIxo?)EYJle{28A>wV%ul<;6wU#NsV!Vi73Hot&g
ztPK$^sdmTp7keY36}k%1SB4*Zg}4l$5mWaX?{IraA=6ov3!|x!qJSq}N-Fa?6f||1
z8Tm<_sQ3LlY((=7m#fdG==9|VzwRSUr?sJ)ijJG<b{>XU6fAguVQuzDmoOfUntY`-
zRP;?umJ62_T0s=K(t0(;C**GT&5zE{Z6H#ULZka<fM^g{T*Q_uQEai5^Rfslz6box
zXfL!rY#eiYY{w5w@p=J^RbW2FCIeO!x-Y@P>f@s8miJe4_eY*F7^Hy#{uZ%eUei6|
z#2MUzZY96hK8FQ3l-8ry{f;>BeO4luGiYg-Rn4j@?Hu}1ILHjmBx}gtuMa+)LnCYv
z#@wG#x7dJh&Y+3x2N+R}Q`e|HKY)cIsD(z><WrlTULeiG^zEIpD@LxP{`$+-^_es-
zMUIhbaFqWH8Fi<0o*NIsSR9w)@Ks+b7-8yXtpNe^UF2&RpsDHDzgEm2q!?GP?SWyy
zDg1N)AjE0B$p`ju&=Yw2IZx~)RJxG6`c}_3&z^Kj*;Lb!)FXoP`HbwkA$$n>jiu^f
zoN-Xu*v9rP=veh=gBh0*i_6}h%+|YVAgx?Hyb9;bD?517st*4L>+9e#IEYvTCWZ?O
zx`A~8hL*x@U`7STEs!+<@CpfJ2};^v;=I!Gi|3+A)*v>5uy|I-f4ZX=79x7q)>-4k
z>K8{VOM$w;Cg~yB82=;t<aCnlr*v4XPW302F6M_fTgHWrZO<7}1g=glzw*}d1Os2f
z+WyZBdNrl`6;gNaPh=PkE|T;R0{apbHcO&w?2IVDDXYujyK^^<NGv-tawP6mcb*88
zV4y!pJ*XkI+~toHM|&u)Yi~rppLU{DaaY@YOG|GJh0fPZ@-}QUh?Px`MuCDz`!zz?
zL(e@YKNDudR!~%_qUE}AhJ=ow%meEHV?B^SDFH_GNSyF&_-c-~`+bXm!Y_yb-k-EZ
z39Fjl?f4qSH_sefW*SmU7Z+a~hc>~!A>rh{AU~<TRVaT(>?-Gk)MAyyAVziZ`z%KL
z4OD-0bcmyh3idx+FF<MmpluIyXM$Z{z?Vok2k$?UG_XV#(74m4kGkJJ-$Ie_cQfqX
z+5dqLxky=+&+l+}?2lw5mk$A%fNC{ck1`NB9?qL&H4N_>^S8(8X}!NG8@ffx*xs4(
z<~(}lCV2Fh`i<PLPXN+Yyc(FT&Iy}yCr_1ihW|2{<#2IZsDVc&mXD7@%!h#E>vaR=
zq4kLGwR1WVmjYZc;;|n77G{X?ecI%&xG*`+cud*p{WN8l|L&5>z?jp$n6gFxrpval
za*ktK=ss{fVL+aE^-dRPLPJLW4HVC&H#dYLn49%aJc>Zubp{SI`|ssvw=JLKqp8^z
z@7GJ?=H{+^O8+dabwo!0UCAuWg~f8dB5TC7oz_}qWb%SAwozv-&9#;rlmi}tl+D#4
zX<!>PX2!IT)Y^rfAN$SI*FsJYAXs#LZ41)m+(Y$M{?_w-`gnS}YEJwO1Qi0?=My5I
z3XJPXksWVe1AqK51JwO9h1JQ+xb;mxT%d?$&7{w&-mEL?*kutHRjuK^hpr#61fySC
zxO_KkZfVju59T6`og`erO^~`B(d8%E2f>8*HEQO>jNr(!Dc>+*S!?CwT_q%sjbUQ}
z>CX_|Mn;|W|38|}GN{U~jlwiYgM@U4bcb|Emx6SIba!{Ch;(;%cXxM5cXy}6x8EP%
z3^SZLKhG0;?{!~mEneJ*X=8zCNwH4k;fS%j;hJx?W;l2X1OX$QDWxVUJ$td*Oo}X6
z&YsZ3LX%E)l2l%hO5(1bX-o}sXT7YXbrFf}K#|RrG$BNrdZ&aNiNTTp$7<(|s5qE2
zBP_tdfmgyjw_%3+sN42$yan7fj%(g%(}S;>vUeM}cdLu%Z6I)RqX6Eg(vGZ#u1<^-
z9kD@(L0tenhcKI3^N)Ck0PPSVsh3rx`Mm`{$`hs%!|5dUd}mH<-`tH=FWkfoEK~ro
zp73=VRZb<D=X34Ac<L@OBGeDiN_-5x=9YX^T2kog^+ldLQ6pvQzvP%#VX1J8z{3bM
z18W}?1xXYA-6kZ<zg}GS-6d_!4J{;r>G+lUYtJz>^o_G)#w9St-+nQ2@|KwB!!sw?
z9$K(kB+!lK*xMO)Y3Z++FrFtcXbHFd!N@``#XwtWu=NE}2hOM-5Q1#|&}NqMV`P$(
z;)m`gO(O_-H38&!5^359`^>S4Cm=H?p|0ABw2w1fjLP71#ZwKD_yXk{Sk8r}Mn|~I
zr8V~~!G=v2!F+||<*#IYTG!@z$M|Aagc<f{W`l%Dx7GFD#>nw{UeDM2?e%4u_Mi9h
zU+YQ7HWc4=3E%%RILa6qo$KoKyj~yq*iBK0I1=wVJzb7ppC?UC#}k`8=90xiBRDud
zG?Kutq;+48?`r)a<W*HxmbDQdlgd{g&^m0m7|5<f+P?knu$1hgKB9?tr9?pA)JCsX
z`pIr1Gcz`=#pI@CYWQ~XfG=BT-B^>7RX#!#%R9+SHcX*j2*0=d=am^7$34?=mHwI+
zTqR21G8PB^&ph*-gPcp1tXum2Q{yvO-B5lch@m~db^pR26gb>puqYCkbEg-j7`k4v
z^BvupkaRvJL{RxHXJ}vw!uSsvx)S;7e&EiMyWJjjiRm({X%W_h`ZiA!T`ElUQ+uOi
z2QgTR!GHU^h2Xwz>x<f%15J!{=-Bwn@AK6b$rY>VO)%nF2``;F+}mp#Q>7u>)~NE@
z9og1fYBeuW4ta5+#d01#3!$C~GwL7r@x1GkRU7n?ja0-rLDh+-wUwp`8w|!r$p^Bh
zWR3+bCtd;2FwvCK*d~!Yp*PN9MFmgJrFC0$9HeY6|0=&XEjcV4v|+X{Jm`DBtj{qc
z<vFcDpfFT6lt*d)(?77I_)24DUi91X5{!z+NGl)A@_{U^bp^8-+p?8H0@C-{wn9wE
z1oW0RNa;IX?Z^bsL6WQv!QZsx8q!p&AkpB0Q~T#g+XQnr%Oo-w8B(wO!?d1v(H?+g
zbD`Tgq*F#SIwm)xFC!iIOVk+j>P*iQoc3K>=!fhe1P{hj*;%2+^NPsbM;&pHL1RFs
zvID#8Fp*`QEb^z2<!}y7qH?BN8aW59miNkkW_*~F;<*~lESjM$T|Z}!TDa;0utE9q
zgnu?hC0*2}lItWGPfBc(H7YdG9|{-V{vA(JQj%E8Oz>M%(^psTk@%~kn1Y-0jR2og
z#q}h16A)Gp;DaPa74jqu$2*pn<p%Al>{Myw`jBR?bEe(W!90p8V=lcxY|TUtgu_e|
zO6BH*_V)<vieipn3@}sSnh^+(S~6Dpmsl|$!450X=%?MECG?f@q?~q-)2G&^YaC|H
z(pW@AQ{Euct&yl2Mr-TjV$iy3EzFRU_=FV7!a-!ha4{nPBkGa8DL_+|D{1t{z?O8W
z%Ro~x((-Ua8}~n*@A%J`J}+x--dFt-f7Q~<<a(FClnT8)vp23c*hF>M<|I!8JA2pp
zb((<VUQL~)2yv^T-qO2sFddZL?q}A%=o+8sXjJmXnDJ;>m!gcPukFPMiDzMt@toy|
zE^^|6_eYj5`)E>(s)*@FsnhmFi0A(-fT)(f6znCmc}GGMj5k6hZV>q;eD?>9K;<(h
zlK%b*xF&ROoW!?Z9cmAy)Oozw1TZvZME(_ga9BS@e*bQ*zcDmxxmtNN4duVSYP`Iu
zj>;6qh~p-Xma+(XcTod6^^!Zxm0DN*l{S3HAI#~AbXqZC^oPpTs|-(&j4aA$r(A=q
zksULa&0&-fe=I&Sd2=nNPn)(I4XLO;=Io<|=n>yf1Ms5Ju06S3wY@WhCK@j@AcE#A
zCB)4w$*U9+%X_qA*X{WehlaX(!I$0`Z;zYjW4*WH{g~I?1U~JKaSLX0Ua{uCxOZo(
z5z4-aL}UAC?d->=)zQ;U=M(RHqG5ghWU4QnIlZH!gk6&b#M@AeZpJ6~U+>8wA(nzx
zNH<cw6>yade!Bh|GfBbkVObOW^d05Y(2g(ch00{z+*L+9lXF#mrleshEnq*ez=7wx
z%}P)^`ZTobB%`)gS+Cz&X;EDW9}?x_;Ub6sOC!5U6SsyHS17W&t8mkHt^4FDN~5tb
zy}lt0Yq`I+rV!^R4~*I#bT<<bj~H!y0;UE@)3e&6=R8mmYQ0)r<{e_9#JJvOi6+X3
zYDmt8ZAw?6S1~VLp5HM_tf$G?F}8v)$njyUHod;f%wC2#1I_KdG-@ibnd2Z%%AHy)
z@}Y(AB*C^lArxc^ny7=q&u^qNZ|2cjIBzGBeMZ|GT|E)Er}t|o_kM0GR3K5YVW0p{
z4<K=dW@wJsUWIqMiKVEy+&D`8HF7xc_avxi&$&_bwk~a5m6rEwi?$QQlOn=%C7z4N
zG1S)d>n3bkl@7lllw@pwoAISf6+QApt!vHX<uZ#bOCumUWlwgY3^g1L848F-xqpfo
z!PdND{OnHY!D^^Ab}OIH-qiZzJ0reY-%GI$D~f~7Mns(L$N8OadkUB$Pu&k#3e1+p
zcv0@rqKDO1f)knQL&9>S<=0vX<2xT>IbB4+&wc8K$#wyR3=>x(J{PauU*7=Patwf=
zJK9qT&v7w$H21gF`}#OrOB=S_nWTUB=QCTzQFe(ZPoNUW7X*y@a^Qu-x36^A2@VF%
zDBseaayeZ@HsmX4H5tx<vW5v2RWdk#RnCsjtv;P_`$%#UU-*Mk!)d-Umg%wL**d7|
zL)7Nx#U0-xN&u;<Oluf`pWoceWe)A_;U%^|=G#H@s}q|gAx~gyHnI|>>QSv$Vu+1e
z$mLwI#4w9z+bbQ>#cQeE-8?~!Q6q^%OmQRX3X^FF;F=Qu>OXuya!EAj)^%Tf3d!f$
z@ctD^B*&c@$`JAgdR|kK^=_Hn6d>0&%<JR7$wLRLLa!iGKhvBxB50#{zyH~<wcMV~
zXQYJ(dja1K0vJx`TNK5-DQ1L{4C(KBrrcyLr)n>alYK~rMji-&5sT>_)Jt%@jBzuo
zsvST6J_kbRnBm8Foq!6_QvJHr;Yy<7xS%1cq46zTq%t72tdl!_?tNw9@A_*|<CzeY
zlixsmVdN^H^dzMZb{k*S?daLPgBQI;8Jj;!N&NCKd3c53Qtnj5=*q*>+`VOyGLyHG
zHLa|{-~#m?ycx96b30dmQ<pS8yf5sio$a>IS0U6Zv}I9I{bpG%xxY)Ab2dFbdVFre
zaSZDWqF+*=+ZlUl;B`MmpmoE<8mAGTaK9qlc<y;{%ncQsJ6%pu<w9HA8MFby+_&DR
z2b|sP6(6$QynKAkt}J8z1Kb&2Bh#~{^tV232F%Cf{fXZ6VKojU?tWOl`Dg{*4y}qT
zYKfX~;$$wYkcfA}r12cW_5u<)5-%a|xs%I;pNJ5kkhC%_%`9>>5<U>LRmRb)f_taE
zep{@!0DYChuF?_Ls(Rs83SwGqsz`r*7I`Z`(3fvLg@|pgSZI9od1rEa4q}s=0}qXe
zx8ANR+dynNt{-gAJrDcrGs?8q7Z98*pWYB?-4CTdtn*>Y(k}KLyjpOnGHj=R7Buzz
zD!b^Pg8J1(xnAGTDPA%(|3t3g2X`+K>@^`nCk3tT_Hjt|<xyC|e?*qcxlJA4L7&-@
z#pm2Z?evjIamB*=xBP$lRhD0i!ximc$W3CeSmLE}NRi-;OHM|u-+eX=j^|PWwiQII
zx2To1Vz7h(M3l$2D=0zF{pfPQ9PPQP{e@aFxz?j^cm^9AwdMNXc_#kzW=|-06eUkR
ztK<`q|3D!Fne^48<Nn_Yo%=lk<@AI34zprGZ$09->27zVJK*ima*gw+in*mlv4iCs
zyoeIYOR$FgJma7qf+DuU>A!L_2M$>b=|M`ts1^_&8-RfWS8l#SOA)BXv`HEBJ<2}{
zx?@dhBk<*O&CIwR`&2%a+<cZM*$X9)p`WgxUX7KIwbY`EspF#x#JG8$eP!XfR*xpB
z_#H^h^L8!ZSbsJ28qvn-tjV(^9bWFzq+K2iqw#@wSh-*VMhxS-YEpV}ojmq2YLn&X
zmoMcIgV%EL5eka=VVT>DUR`Ig=q&~0D*hLw##9etpbtpSZFl1BGUJ|vzteX8+*io!
z;R38#;GYsr{pRb-KY!8@Eg%M3FIJznxvnq!ZdA^liDz1p&KET~IBy001gO06Li&Y^
zOr#9<4m(lwZ^sB@MG0iTTbVR`Be+36+3Mxc>VzzP9`YgGu(7_?-I3}-YCN{)SaG~2
za@O7<dL=B!UzCksYDH`DzT1RrCveTuZb_5;Kx*WSE|I7S(d3EDHpi&v#+REZkoP^R
z0-k36&*kHsN*>>ORYD~fPcY-twiQJ2%buDcH3XP>rT56-xNF774Sje@#Guanm9V5b
zo@vFH+-L1zo&+4FaF5Gf>so3TuLi68*1t`o5D@F>1A?4d(Zd@!IvBY^PwpO{mAI@}
z5}$4kIIkWtK7S|03KL|HxKg5#<tHvYv}M0u_V=9FbqJ(w95|jgj&MC-V(r8w_u3nE
zh7D~OMre1lK|qxH#5O@)6MjsF#AM(M@uc}YZ7|1L!TjU$S~W%Nc_z1Yg+gqyEz)8B
zA|q+!&=>Zi6P)tBuP(0=2HM1Dqa`9f)Pb{!<$T?4ROW$N;vAVLwV_`S=aq|8m=8t5
zylP&5udVtx{=_qa@QOZ<ru$i|tOL6}WHPFJrB!^oy!Tr;O_#e`>e!59#ELBRU(&d!
z@8^Jf71Vw<_>_>NY`Q|0%c?c`Kv>oC`B{Jsy>YSSAjAL#1Vnwj3mc?rlwf!#h&>ri
z#Y3ulhCYpe5P+%%X<)|ZjX4cZKwY!DaQWVHzBG;1`^ix(gNhm9F5^Fo#LE4iSJn1)
z$kUT()&)rD00q9d=l|?6c<OD}S|0(8=nK;2<t4U@vFY~#QZ~4sUNE!K)|?RT<<nb=
zSM#@A!xLcAzqt)mvvEM<l^Z+srGf6{mi|}nUHD?c@y5&{6|S5uM27)Yy;g=1-A>s8
z_F+^D#w@ee$?4(m-;@Y2!wXo}bPp5lLyfOs;d(JXmzoWaA8dxPeoXt$Ny4N#U7>ak
z7Q49XNyZPRsIu657c;z>SNZU8J;3Ap2S~^D*8fmf(#OqUV7sTp+b09k3{2>}66H)A
zSztzke;$`NK}QX%(Z6$<+_By?lPkL*P#yp!WcW`CwOt2@*DCnJCrQY?dKV;-Xbi7i
z;NcQdHmF!{+<u{VZMyYp2J8`t-;fI!(XVxT=>jfeQ97$3Op}RRI6?8eRfT+ZL9Cq^
z4BF&CtU7g~=;hEKKqb{viWWf5oHcYts9$xln>$(Q+t?Dbmrx7je@Cc;F;SC#WGUa+
zNGc_ACR?LK*gF&?+o${v@x(s#NR~4Y9+~j<#8JHFSE)Z2>3Rt^lo>~2?;tE2S!6%i
zu)uUT=M7(5wHEZg;@-Z`5ZRWIt4uX+;768TZuDPf<ad{~Peev7I$y6Zh7bYX@&qaa
zGMMh&9=<|3DH<ZRFJKpp|JNE1Di(kIA=`nhfBZ3J#mKzQ>iPG?df6pXx)H8Xw;LbP
zW0vA!S?04-&D=W+<JZWYJ2pG)f|D#rq?^2^di)ID26h{jZUL6_2He?><=~>jO`H<3
z_u-=VV8TZ#%Z!Y~_F|9iP5Qyb^LZ5d%N->ngvdFfD-p3&;4<rv)qONJp{X6r-i>zE
z9foxXL#{%p4!bIo+hR`UZ~{f`FyK$qdp&fpX<Mg2ity)PTKevSndNL%FxdXf5(GUY
zeE~K@HM60etKJU7PY9^4r%U0a_Z3CA($1NUaYT`lYPs|!Cq35hBn^ZAR+tG`*l^nz
zQIF8{{vz|ZHY2Yyxm~)({bfOY$^YSEC{1}0;a#&A(q)?APY&0Da>7EZA>(Te%ISRm
zyhYKT+b`~-A~nx(qd%WngA-6c)&-cH+VJ3)l(dBGc7bsynC514{5K+~Xn!6)!ltL6
zNR9jc7~1iIV*1^uX=c323(N?%V%{d+yl}(CnN+@&Y1BQ<CFmK5kHhg_?(9fc_yQHR
z?O%IoSqt%(NZFX>E(yA_Be*c$&I>+KZEBMxJC3Z7QUz0o01gKh!d)EweZO|TZ_3I5
zZ%DM93Vm11Bb35C`Hj+vyL@$eX*g)8adnA+On<g~D?j9lw1-TkmjX>Te}?0+Z=cUa
zaPjRDY*S}mDU*V@9D;G&NJSG;4yvLiXyb+?wx#PGV0MF~NuKBy;D(3fRG(2I#J7#~
zKFA-?B`P5m+PRW4BtAx|sHcxGXX3m<OroRJ`kpS`?uQ6!xu5YjiIYds)z(IdSs8z9
zgbzUiEyZsK_5(h*-af8PB>B^efuVm|SsZ3B3?@EjUkW{MuD{;(pxvLZE9OqaqY?_b
zU9iNAsa{_u%d2vlYl!4k4H-7%;2S0Ebp3RSVBU!3Pljz!A4_TZ%;((fFgMJpP0X`G
z!q(Hxm~VoTOFgSvv!v0aXL{Cm_;u_#COtVOv$nWiDBqM=`c6f!b+pnYDh^$2##z#Q
zhkOSLZ{!19PPxsK@&JNdmTZr-FZrKFbn37%vZhTwWGHho?&-hzRRmDQP<#E1Ty0Kz
z%4>-M7n@c}zmp;fK`Y%cL9)H?R%*xf9UssH&&PZwPEjUlx7s}KGX%7Ps}5`;&fkoD
zQ&zB8Pt>gmFuYxd*8dKfDloz5@iLAJ?f8g^uAM@HHYuIyk3^vrU)S~h3jIAMBXyCh
z%)i+YK6g`H<_1o)Bw0s5h}=H}YU^V#K|>A-z{+>yaykXG{wbZ&ucbR)W7a9}`RWed
z#C0?`>3C0<)8FTCnYW41?i>@XBh$rfI5}~OFx}+AE6>+3To;(U!uyr0U=BiEi=66<
zi=T=HNGb|>#3u4<d%mC(<cJg>I`Mg_Eo!U;92G7TIgbkF3Hj28gn+kf$$a_t#3f+$
zJ@8n`UgOV(7Lw%JC#Uld8Ch}$$<l@tO<N+v2lS2D(5!PEvmNYLl&?j;U%}Qvpg=%k
z{rWyeJ#x8CBlORzhtIZy-FTyuc=cCtR|cqt07-nH9H>K6En&>~^)<qGgGi_{XZ;m+
zc@U+;^Zcx>VrdEV95<WJpYhhL6J6a1Kf7;scqZI_5HOr=sWAm4^5soZjm)r^muu#>
z&PhMzjK&f{E(xx+q@i>s9&>w(v1W<{Ys@Y@9}Mh*zrOK9r2+6}dY&BEd!Y30Rqj$y
z!s<W?f0}`7{Y`Xu&%n$p=bC?N#=n@2nvw4*YA;SFD~s*mJ4wj&5&aisu$Osn*4)P%
zBE{TlUOP+JbpG6&unvgRtz{x}gWqvVDVqmT+u&*l+&;Wl7#h3C*bGGvqnYywb$pC3
zMdS@;TP0GYWEIqY6!f`U3;~c1wS1x!F`o&#Fj=3%>mPj{{BUS0s56{Dy<0<0S#{w-
z4L-N!bxSC(#oA|JsLiPRAo6XOx2~n{blv-YHA3k1$7#p&ZbgxkYP`UIbQ#Mlm7MG2
z#b(dn`ii0aP_QAZfVB{#j>XNT5rNH9diPVjJ^+Va2$!*6rQ{=;Uc19>QWt4-mc*Cx
z!z^A(ja1U<n}Ht$66>7Smg9ah6}(Bl9+-6t@T-!JR0rl}j`)^1dD~$}mj}X}OB7HD
zwH11%+=0@Htq{W_{;-j6@;Moa#J0a~=C9r4aav|}QG<k?hu|L-JI2V0n&`vfV)_hm
z#lr^ks;bPl<8dYa6t`(k)H<R$3)1q(x7PumEj$_#*@B4*`lhqz3(xK4?kn^g!Lzl)
zNHj9uxk-)u-GdjT+lC>ssM_Ob3tc>1s;Pp8!@tUZjIY*Ru2moi?Z1(>Ywk?siteXg
zu?Q$0h*?yaM72&s=O0~yH<Z2k9R#^w%oEELDoA7?z$+$ai0%-ggCzz8&#|I*<lilY
zprp{qr_L5gDz^Uf1yj$6crT~CdWDW&?nQoc85bJmL2f3vul+H83DhwM+-#3f+n;l1
zI)n4$OUT%_hj+-c&CB?`2w79<T_cSZs-on+H+A#lm3^`ho<jN<d*KZFg{O>Yz;HBL
z&g3fiPc3oo$S;?wt<Te@%NPak(5+^?(r<+8V7K81y>xGc(J7ftk_^YUdMFFk9gES-
ze1G#r+Kd`s8*?iTW*?wh)X(EKIyIT>k2A+JYTZ5M&L?8pq{8XdSlkY6(**cKC<Lyz
zy1zfi2Ey_E2s4q-!)lA~%ZMIj!tukD5ecL2)m5XtI5Thz*8)=4KrD1TRnTbcsBYj7
zu_;~X%p&34SHHka`p+a^BK<6Uo#ckvmR`18B(w;6LgqriCTQYs8Og(rvQ0$7V<W4N
zmNluUWR3RUTt`u|FXcbs28`YfQ$Thbr9qgmvCj|gTt5(u2*`4qTxXU!zj|>#3iwyp
z$1_fce0q0VqIHT|vzW|WYvKs;K677f2~f<f^HFB0s0dr*m$7uN9$<Fxo`z0Svjekr
zwu!%_TmEX9XSGZlQ|X;ktO3P&<Mlm%l8NKF>9{HvCxN|Gz}+)sRFr83!CzI<)$Ge1
zu*L)~i2QTTc=^_}vOlgG9~Y~T!gjm1>S;l;{CvO8O#iXX?v;Oavnvq3_nq&<#isA`
zzFuM(nqcpt+1bZvbfa-DJjQn64}n7ewKyj1eq(<B?fUfc8H%y$sekYB8fVn~+}ng|
zN9p8@ys~+4eoj2laYhf&B|}G*OFF0pppB7gk*1}sdSL(VNkF#)mv%iM@fL`N2Ahbv
zN<`gM#|t3=S~u|M9bGK?DQ($8S$4=z|MX)g;9rSHLtgtsCP<*{^MyYUWv6HU!HI11
z8QKQmNWqCcm}~Dut?0(|8E|O9wr4Z>5=;j!X}Ip(A03CF7wq(WV~=I<vk+K4;_w+o
z7a&99hnYN)Kw@7(6xY!e&qAY{WP0~Cc|pk3qnKz5HZr$XlY6E7F2q6d@KD>e%m6?-
zF*I#9X7vk1YzP|gqSea{p7ZHnC?ptCC+6Yad|?&Gb=YlsU+b&IJfT)djJxr4cG*K0
zvfG(iGy;2enH-!_#+Vkv(y{lorp#D~cpb7$8_vMURo!`$JV|_Jl9G~wy-~aHO8f3p
zszXBU%mwl%Jx2##!qaXcagRoVL1A(-2~8LYAfGa74X?X-JAiQUs&j#dLw{Ku=Az=}
z%z=CAf0}1Da*K&FruIqYiSSyyk+lejxLK3<JQ>$3R*Gty$7yQfMY(cJ8Q3P-Qoe??
zV$eHSTy~JWixM$4thGH<p-<uHhS&<<s{yOc=dklZhrfvTXEh7)jL1*c5;x6AL`7)$
z>Lrp=*fD3i=%L@~vbA~$b|`7mhvID42wzgL+|Gj=ztXFFzub6#OE2lR9iCA{R&Thg
z0c(|~!NXuWp#3DC??Jo$ny2X<ZiX1A8(mU&bu~CoJUs4exk0;w98|tS!RPTbJ`m$(
zVZHiyETup`X{kd!{^Y063k~qW1IzL0QeDg05vUK*G&6q>-Eo*V?Lr<0+~LVKQ${(s
zt(>R|e8*V!p`I@yN+Gh8(r;$k;%*YuA|3_|@~&L-Lv)fJh7`YOhns9a!PpD`EiZ(F
z@f;kb92O(&`XUM)D-gI@xcX6G>byz>8cw&}3mATmI8U3jk@6dJ_M5;sVzO)Ex3|IX
zlS{HXcI!JB?ZOIsM%@cmLi2(5eV891-@|}yq49XlFGa%c<h~#8y!gz%F{V~fWp430
ztOv(e#5$ie5fZSE?P(?^zX?5GD&C*17wA;3|9L5KXgXhQYp|V2_jSq)^l5izkCS7~
z*v&^FDCl_1pPC%4jp=!NJmX^$o;8Myh-x{^BGRv!LtQa}q=gCoYULauwej@Ju~<7J
zo&niZzp-~j>EHkZqZseEYC&_kL(LGKXRAWO@<&g~i}<LW4_}FTD7A=5p5=@b=gy)*
z2mR0%%ZD;!>qZ`B#`m9_&8huQHN_9WGL_GR7vN{HbIq(>%Ioz&TQa<CWy7`1GT*%V
zkyi)oC1HBLG%#;mLOVd3r@(vr6n46LzC<tYJ2hFm3|Upp=wUl_$cUK={@tGX{0D`k
zM9_pYfzq(DVE*HA3kSdPP}$?p?IaY5l9Rgt3e@OrGPJLp%Ft3^N+$9Us9UT+ZNVmP
ziu(RO0c$E~H>3PN!qBopAt1MWUhxAWusJ(lOl~sym-q{p)5nckymB>hT7mc=?qZ00
zi1iO|+~IfuXkYrOGdyL=FmfwtQN8jB5QY8zokk1n%-lRc?P@hxs@hE|J8NJQdNW$P
z;AT@I9Nh$=jG4R`P*?M}<VcAVDUX>8x+Lr{{!&KR9T=)hWm1gkYP4SdQv@>5=37?p
zi8=H)K)Cxvn(}QMNm43*3bS#wyMVj7M`$t04e?JZ$6(bH7}a#X^+{u;ZnnduIjiqW
zh_qV}!)r;SUkGF`@+^p?_{0$%V5_?&O*zjqq3Rb17ohny?0;#0w(5;B_diEm*+bN(
zPhyGLR7;6L?d@d6z~2}PK97F2fGecO)tn=QsQcpNNu45z=|s-KyqZyOblD56wm_X$
z@Ngz~XCT(gd?L^7N?_1sGw&9s`uLP$E2q(7a+-G(28vSlQ!Yr>KCQJaCqchG{w)I=
ztCvkQUl#mi*Rusy-?Np?3Dg_`cXT*k`{(XQr2CUaE!W#*ABh{cYyoPh%vID#E^lQh
zzVu=JOZM^>st<{B(3Eb#g3hd>hvk%`2)XM9pNBJKH%;_`FKK7cy9nDhH0PqufbHTY
z=+gLC>k=(YCdT*3@8{DI@u)-`Jn2%Fn0-dx4xc3Kn)YPC^c`fONWgQ)n4@aBO_5L0
zBw|T48bYX^YqzpdrH}33p9@u*Z=2YtFvZYgxs|i2Tv1HBdk8E@RHpGa7{keoee_Z$
zkNeY$2_eH=Tl!`NU6zBPhy^S2)jZR{DD}>m?i%Fa^qxMQe(0b}NO@hzj*`kC79crZ
zK39O}hVxb(d%XmsF#>UP(2C2bl#Yww?yb~8=k7K7SAIj!oxhk*Ek^^U6gwsoDZlSR
zA@ACV<v##k*&zpA2ig61=DU+ySWsn8>3O2CXjLP|og&tXG<y?7HeX!VH-Ez=Z$hA>
zC|jRWYv61aSO^M>{k+Pj^qW1}E`u>Bz^<OhB|9Kp+nn#l+<WsmvCFJV{nRD&BwC<)
zAe8%sHYoxTrPW`Ub}M#8wM7=WE6&(EU~bGr<7Gr*JxspvMd98_nn45CJHFl;wM`*M
zIBMLguEiK_I23-h9b#j)4bn#YcRNQ_OKkkfSJQ;i2ql1E6@K#l$<S#9a!FrJ;ZW)5
zKdv^{iE(L<{CvWFTsHouKtq`U9*zVUg^Gb=TwH~mA1QwV5RTf7HM9g$N(&ZYK96il
z)s{$|GX+as>{4=%ob5%E-j*?RQjb7evkTmk&Yl7RRQuAJzd#JC746>S*7RiP3M4``
zC9=K(1CsT!kgwx#O7A@}&&J_r5-%n9@?UEiQ_IS`sB)8EYlB7Cs2_ARg%{19ztj%=
zArko8QaUiWAMy{{pewquptp#U;dpGdJv7JOqhlm3bs6Dk*jmqCxohRs$%r8;vqo+X
z*Agg)WFaRG_gyy&$Sk$aG8R(>W_*wfx8um{;CxpttQa>Qfap5KEut5aBo|FzYpCEX
zKs9C@>o9g{OqYP$WDror!}~7yFCvdU3n7RCprKdz&m1(0#~zBpI1P6WUQh>76u8XK
z8ym0CA^G`l5BI*d`$or|8SBm71T_7?=k}2C_L!hgs${>lGfhSE$IFMO=ZD}{)Bbuo
zuVv3uIPjd;Ej%3AClUU8q-G~)WqgV^zh%?LQ{W4CQp`)EuTAez3WvO&CrFMZn$^5|
z6wL~bA^LWlG45QQZ%MvyG!)C`8@5{!YvV{=(uIfm@kQlRki($I!Um=iE9&9BI9r-j
zzD9NJp9WGjvt2kImOlKVpBt~Qkp;0s{%C|h0~6%%t2WU>LQ&oGOLK47(?_+q^dQhh
zB%4$4!UjjfeBcqXA-fS<zo-jBBt%9|>VBJpRgyoO3B7A(Er3S-6L}mEOO&UF{;T)(
zn%UR;f*=O{UtHsg@9Rv)>(hM4^q-qC0&YF;!(D5w`q!6p--2b>yRzj5yjiNnNpz07
zti==YCd&$DghTnye`I-sYSG1%7*Mh+j;MSMGvbdQvSb|aMo3{r4O9e(AuQ+@t!wzq
z3lv%%gEIpuG25Z&EVBd|CqEA1Tj(xHQxe^5dl)y*8Pe{sTW};Y={B>u+#CRQp-=?Q
z&Q#$0ex&m}{DO%-LWoIC-%)Q~rBeC^5*eajBpR$oDpNz>9u{B5C;Jl|znANwyF*hb
z9@qliQ>GI6^HfsPGPLqmld)#C{tmBO4<kM@r~qs!HiA=f%e%uAoO|4D$-&fxgJ_e+
ztVJUu5z7Cw0NT3kFKHU*sRCNxt9e$WW|vNHGk7#uL0HB4s7Gkhv)-vKALVPPnX8N-
z%wPR%O({98?cMJ-B4^o7IE)K6)jUc9Qo@8-fn?2``-4|2qViT@eW=B5?GQQ;wjy3I
z#&BtH^5#q*4C`d2Hd}SxU16LdV0<kM5G>cV1@@c$+M=1K+oM2bYd}R^0Q+cM+&7Re
z@H2*zkCIa?vl8}Zq|fR54W*l>Ov6EKN=i7uGWdO-jj6RV=lVE83)p#gU$uA<P0J66
z7#(I8QOX@MtAl4-y6*(^CiT1#M-`J8+Ni$Mgf_wnb2M13HQTQH=#h%M9u%*Ktv<Ms
z2a_W)U?!$(=o`s2iL<TFRU7w~tHdHAh2$)HiTEAnCHN;c^VI;$G#oR+ur`J&FPZQT
zHM>TrGQN<@$`l3RRizRo5CiBZ(wNsZ1%20^K@i*R`OK@rCin4wqol`vf)p;RR<Fyi
z{KRkn)|Vd_d|n?-w=|!jG2X2<+EZZuNi8l^Z}GT2eou3ETnKOOD+Z73WqHI=D*e~0
zniO8*mK|!`H<(=_{U?4F(bu?belZWym(@ZZ=cgIP_<aXP?XS=tSoIh!T8JH}9%~}}
zh))f*w0M<q*W2+<T#{ZwUIi9eFtwO-naDcj0YrVmEMM)9YL}etJ%M<UJ2U42m8Zd4
ztM%^P&%ltDLOFpbs7%OvOp_Ytv(%<6yV#$X7;kt-wO!`>YKa5yl_-1T8vL4_h_%4V
zUO=M8Lb&?)h<m*^N~MHRJ#NLM2#-FsJkZbT_=qE#MRhUm!&&e;(X!(EIHD)`@~{o)
zi-`u?H&em57T5EkR?3j9MMC7y5Qj%dJlZw^KH}IO3N?sX<()-1G-|LZJ~4=8v0Y}o
zj`a>*&yRvI1F&i8dcx5Hs-a<6lcG#v33)UGBJyP>@)1(+QJ5MzTag8n?8c;>WC9M0
zByT;S)hf8zyl!`X8uaL^<P{ZoE~R1O3W(fy3A<7u`yU3r>;Irncd^E2EPdCUMJEWO
z1lzleg(z9(f|5aqH8RR&Qy=bVo}xU1)SdI&By&q`ywiFrNvpisu1f8mIO6_P&;F>?
zfjB2gCdAV}SN`9!=q{yzjz}?InvK0eqnZ1d31$tw1$QEs&C+y@=lz$72)`wh3X-?<
z1)Zy^m;!)B;|zEVs(iZK$?cfQ3wyU><-Ao9L+J@~sK>@=?be)~psMU?z4u!~l|Tp#
z51P>-eWK8QSIp&2G#LFTwv?$LvBFvv3-T4DhUmJ>ZU4LzJ74inl&EKnOOl6hPOn;|
zJa0y}Lv{V;@egx;8*iEKIqJG+=YLrjl~0B!i<IX07{V4f-_#sJ^^0lnJyYRUs$Ec3
zd9`r9NNjF>^}@-sS%Taje$Ay)J78P;&?=m()xT;PoivA>JePX&vw+Q`5`-`!f(ooR
zR+v!G%fQ?&e(<H{l1DOuxt49&u-c)jYs06irR18zq<<P0>ZWD#1%i$baYH7ej|(-E
z#YUGDO9O-@B+r#_{=96kSyWa1iWvDzA0szFD+%-$<KrWPAQIg=hdC|f0u&<tAh^hT
z;6AXqNd$_I+ruQ8Ms)kj-7<j`G~u~Q)%nZCrdTi4jk0W(l4FqxFuX0+n4q{lzw=1<
zR}|y~SrVX(X}RM~*eJ&zjwNP!Vh?dxYukrxC{mv%Ie?#ceNmbIYiB$CA4`^qAET*b
z*m;UyU?D@<BW`Hm7<ghN#kGsc;HHy4@KU}hfE^>2k@)t<uRFgLMKn+<mJE;oUV%Fs
z&sN^l1k^(i>uq?5y{D!bptBhkltUU;37nz#Tii!{{&nxh7T!ZxWc4S+g*_qj=RjNp
z6uejx*^W{;I`LdkjXmCvijTbwmnsJ;gv(h{$Ubirto=fv8~DN1Lpjm6=b$z^!+!l0
zE&Qe9)S&@OGyUXkVYDHCBI(6O%=?8x=pKOI;M`LkXwS2&&i^>z`)_6)bfGBvzPbV?
zaBzOh6ALGd&Tvtm!x8cbfJV;Tqzf-((s%j-#-!c14{1LV4rEquMHB^eRKGTR-V1sJ
z*D|`8Hy+x1H1a@}7pSV*Sw4t*WNsvnOiBweK|IH_xEuyrw*+fix^YEr{0*29gBhUy
zszHM!=CpYFj&*vF^8jtts*W$!W^w!e!G9O&Xtd1-m$r?M%)cmDu=ARNS#wJ_%v}7+
z#3Lr#Lr|5!W9)3D&2zmgFf?hz9!)?hX>l_^lvi7ItoP%=mro^Im*@zY2TWBG(TC*J
zt$>fDpU@!<v-f<~^|gCmum=)<hF5E9Z3&*veZ#6pw$P4)L=BM-$tV~1MFzZ1bOI`0
zv`y_)-p1#gE)JKQL~zMIlQ3k{d|8qoyVyQN3ZmrRn48P~FzlcE`C9GA4Q0p-RSKg^
zNl$RoW$zD6HkmiIWMuPf%3k36F<xqe#AK&hRfm4nLHBN1cpYp5kDAlN<W$32sszh^
zPYCJ@Fn%O1{5<K2mEb4a!ufh9X(yk#G2T#H^gW|7M=IyT#qY{I1_5SbdSa<PHGGX)
z*wI~g4pAbvFT1p*3Y1b1h7TNK6hZr4g2=84Hb$#+$3RlKgzed2EU3e%S&>#cqnG60
zF$P2ifkj%9OC1DczJ;=E(Wdll|8#;>0c5p8gVXQt8fS6dc1YvjFlRdx^oVKcKN0~!
z5H+EcE2vhl-7BxUnl1^hh;s!YQC+t`0+}pC_AZsrHOtHLF*}&#H_1P8ii4NYPLcEW
zmvP(-U!UFKIH2{PtwLw7llVR#0=7d4EMe!bFCdH~kmq65vy9O?90euAb{%Ae{R@Qe
zgo}7S1gs)}Uo~6JF2~ba0Q!f~fg3{_?R;eVJ`-d0L1Mb86-S-arb;~&6~Ef@p7;}t
zU;C(+EpPg8p|tvD81;<CQ%X)oHh;XI1GYc6tRRt9?k9(B3xi-FwK(S4fD3P_T4rD;
z%+=xuJ_lwHo3TMZWqOM;N?UhEQYeVC1zGlDzU$Xg$Pe?ALRQh@eJAzw{aULb78Gjj
zF3QEnob{^($?*jY(@ReMHmiLVg)5w2q@~MeuU*u!M-&?u9^PDs)fX83oXv`f|F4JS
zO}v<ecrotZy3oa2{YabZeLA?2mgv=9hv0S0%u#~K0R`rLJ=P!&zXg6DFau~?B9^u8
zF=f%zjxMetlcJ7csOsB`HD=zKdl88cB5i8pY8fE_E)MIJN{vbb?DID{927-GZIW_$
z%u?w9luW0x--Qg6K0SQ00`0ZL&SoX6eX+ZauH|dOKb94^@qa7P%Qo0q`xo-f=$(cQ
zI2l14jcPnio?fPuDHNQ*ZFq!opT!CU-6Og753An@b80EYl@XG^I!wQa@Q;>{Y!0ya
zl(#@x+*L0h!(YKJp7z`yxcqBOJ1lv>Y?CnB{9=9M@?%Y#jgiX=)$Rgt$#W%t<<j_f
zC$n<IEtU@*GCJ%Y!-s6W09^3|2Yt5w-B-(H#?5^1XjqKt8M11{|KC_}o)hQw%B+>3
z%X9DOA^S1G^6q5Zh`uJWMA(0}J2I{;p|?_YY&kvI=(f>qw&bq4`=`KwL?oi4Y}5A(
zg$n6CXOOq2YkmW`#_|3I9zjPQ{E^QJx@~Lf?au$eWcK~f9H4lDB%AUfeHEM5=|jTB
zG5QHa&C2_f@gq!#o@?!*7ER|fsG5H&EifT=??W6(Ww-OnW<zF9f9Isj1=1!j+;{fW
zAa=2PXfDB->;7W>aF<SJF5!VSudolnl`$5$c6+u8B#704%(12%I||2wMvQlrH9<Y2
zls{g;85Z}lq>lE>)&h+&H5MUtOp;X=1>dl2!n^+Z3{i6HAr<NbX7KwI)TZPL;P9#p
zIo0Q#(kGnoIz1qqY8?@#k2G-MB?CY98yF2%D0*3UzE{MYz@yXmymz<mE&=fe{H}M}
zY*7y&gX$dQl3Y(ToMpcGJnkj{a`ElveWl#XW`wWP$Q4k_aTwKs7e=ZXs}+Ph<w&-%
znBIA&aM^tFyh}Z~+`VRWzgqdUB02`Ue{~9__Q`ubEyEYm{JDxbdH^%boSm~U?i>40
z`|2etLFOOEB86cNvz*c+gSEgbop)==s!a~zz(qDXtq6zzd+)zx)qBY)<ObCc<uQ9j
z5O12PJ-io6IjQujEz?t!*CTXV9n%;pW>iDlKO4?$VowXC!Gg?;Klp;F5^zz{EO0@D
zY`ne*p_hZBQG+gIY0%^xY&gSWm=_jPj0-;3;YAZD-gjGh3ePH(IHnfoVo~arS{U~}
zbZKqw*ZaQHDgX2YpTPqN7XBVP+Gl2uUT&K;6(b}HEu38@nGF^8wa>sHwHIl;vkxPR
z(J)}@Dy5C4*BXXhU}M}sOaI-BD~vtWB>}FWL3WG&C6WdjM)N4PFC6au)g*^m63ms=
zj;KefocwTAP2B!S!>OIw<J+g1!wq})I!C1o7$@VA+st<tGm}A#k)i!wG}3%r$;o4+
zyHtU(U{8Gc6j(iQZfD^|#bJp)_^tI5G3uEWhlaE0dP{E<D&ebO2j#!YkQmu%{oI<R
zv`_SFm(?vh-ixQ0@ApX25DAutg}}VS`>2YBUsniq)PO1ElXvwF(@$s8q{k$&@{X|^
z)(>N(ji+b^s$u2LbqG+M_MAy2lc8XmL5cgKy+9N_K@uDF%()s<-zO9eaFUVkXj{4P
zJtP?OpYZxl0PuilkI#Q_8HmVTKSqoCN=?g*8*Ld$kcC33v)|^VW>Hb5*J0l+TY(;x
z3l1~Zi>-53b=!<867ZbK4%@+BTRu0*M0$_o_70T~4;R0BKk<W6Dr!P*s`4bJEG1%@
zwe}WC7p)iKR6Q<Z9CDXyRy@>JyhwDF1)Ry#*dCIhtEe@Pa^^8w2)vL$B}l{F4v`Na
zd%(gvxN3|V6b@3d-t?eg_5KcPTr9_=2!%Km#1!|DE&fB5aq;Am#2(q{ClvkMk-eHn
z98TuXENEZwK@zDo$UQrLulL@`L;)k%anxJ*s-bhhrP>h0?3;`wa)Mk`FnS%1zNMoF
zFeR>#Zu1FP4wtO!ZD0STfVFGY_hJKX%cWoVCc|c>(f93@8vw(%ZVGCtn$0^s_D218
zbo3PIUJ=taJ)tNdIHSKGB}E~%uyX0q*lN_`(tsl!i2mk+>^Wf9%30@(kd{89PfmOH
z>=;aLhsX^fUXC4NhSY(qrGbl5A<-VpZ_W13Xh%tkJS2BKNqI%gZf}NHocY|+f-9O9
zdWIM7B=lrLiU&XE<*VHK*sax&Rg_sPDb_kFxejzcWWSKN!jORL_`ed>()^N@q0xoT
z!!x;fWe@1lJ9C9YCAMH%0yk9N?ZQy<DX_meTP*YZ{vF$KRl8j|gCSiml!A2<x}Y0z
zQ|hFn<8e=KXXeE=@$Rj4<b~PehwIszvy9h6d1nqf4~s4duq#(P4W3>ureb!PLcun8
zJfV6mnXtE6XIry=OqkomR7?Gp<}M##A2wEm6W1B0QX*7fE8@sqDyh757Txrd^huOd
zlX_I~@2oBXm$4bRMp<H#hyF5J(@_rd=VX?GkgX7Dt3a9@=P30FReBU)O|n)<7&t}I
z|I86Fe{Lt@UNiogJm9+yqUkqWroY-vbQ@U4f<!H1n7%ZJuL=pFk4qe!DI&%MPzD=)
zQ^Cgo{tMTQN{<02EFo)`Q}V|UYNNix8qYhgpPUybsJO=3&|TaGt&bqt;cuyiBvToO
zcC{p$JOmy>6^?V8@N_t+@FZ33j1DpoWP`!#zOg@yA)~I5_rvQ--WT*OL|_wxtRLb}
z4jCY!?3hSS10K@y*-1Q}9Gco?s+XT0{7Md<fnAla+yUC<i&imnO}zFEVIb(cLh~A!
z@p-~ScF#~xgf?PVM)#zz(P>@Si;;~h7a40ExPz+)*bNsPJ?~FuT483$;>Ha~Sv#YP
z97OU{uFx;?jtpCKec6(RcC`*$-^agE5j7hdUp?A!q>t}37r#esIYsScZ`zu(oX2H)
z=sdh4H4RaW3+nUlRnamQrEMC;K4^Y1&VRX__P$vFObK|1Sk^p3CcUQ*niEa6(6%5V
zZsB`~_5A!zb{rU2DhR!K4QZ6kxANv^{{j$pDSm8MyO9r16dCn}czBwfc2}Q=Ze(rP
z@tanIRSJms0!dZ~YcMU76R7Scccl#=zkPLf;idqbKG1$3KNNr`3LN%95I}>?dOI-I
zgcERs4@B=mEB?+|FV>n5AII3&3E6jt@(b84&2p=O%A<pesC*MtKUikIsjSjJGDTW=
zu76hW=YdnF`cafT$)(Y_rKj(-hXXU*^@*?lCJPPYqjwZH{b2w0(438b5tD-Hl}9^v
zl$qC#<HUdA?f`2!4E}hEyaBsZGgAeF9{cehWl*fS`U_dwSd?w11Dy7#sE?6jaCT^*
z@^OGVTxDj__0T@zXFWf42LHf7i-&6%h`wdp_XD`)I<^qW?_!q=a3iAjX_MB|iQx+s
z>{GMrPLErp&R+F))?eOuPS2UR_cz7NU1>#K=xWJmFY~SmxKu=su0#D62+c*_MV7#P
zRnS9N1}NT!Alas+UzWogX(?OB65&0rr(|cEZmPUtdG?UO+Je>J*tw~jCywij&euGq
zB!$1Fq5Z9Anms@<Gnq)1m#{}^Q8s+$u;I>289vRQVVzOt05q`(m%8LH;_8-GV_e_h
z?rb=ZAI#_+$)yUoEU29c>m;!&f=iZEX`GvIjr)Bgb>On?J!VaSXOzMSHl%NFOZN|-
zc!6Kil$9ht?4~fwmfg@N0alS4`KAxC{O19JToej-APikm;O=33!WD8Xu_A5vA5h8V
z2^xTa82sX>nIhZ-k>6+Rd@z8uQ8oGk^+5P*T5Lt`1ccC*fHJ`183Zqw<{VHI`yPSW
z3sNK~w(0mSwC&ukCVOz}McE{gjHb`e&yVyA7UXB)FR_HMSkNNEk*VxZ&|;`;EQfo9
zPvCD2Z*pv^vK)F8z1MxcD7mOeYgK18K}XH|{PFTMy}bO>HzFIwG)JF(*Fmzk9;xQ3
zfGMj6zN}Y%J8P`!G|+>vxOsDt<|pUtbztXNrTp&OIMQVlhIn9~1@myI!`*Y=eSC$R
z&4p5{{rOtcNbScNR6-ug6mqHa!pc;syzf1^)9Ib+@s^Z)DyibdM5OXqP_-@SH$Sd^
zU8f#!5Qn*~8c(K(vr`C1%2*8=J6;-xO84J<5iqXru5x2bu+7a8kipl7UvIDbb`)5-
z{-`3s|MXY|GP9$#*PT?~`|{;*#f5O#5@H)GnDh4<VSIM!64V{B!-(3CO0S+Z+hM@4
za@h~M++KG-w${jzMdI{Og)~5lH9BPAq8kjSx;7+ZEZY10DJaLDRd1t?te`SdD$mm5
z=niE5cpX+>ZtEH%1-hQI+CPsF2)*2Z^5!A&?4qKgro3HnWZ@xbt*gUMpz;Ki1rWuL
zEstHSmhbjJRmNHi8f(XhtPIKVoN4#NrR=&+v**NG$+x-23t|-vdSrbWd2N%T2lO^8
z>z{C0QdWyCBOsQx&UzVjaa_)bJZ{YRv$=ZWp#+I2zRRbaDnC88B!d%os%nnW0#X8Z
z)FAS2^Tjh7tc2keU@xzi(T*D0tVo6lyTmd#LFy1$bvRjj`CD}Bv<W*{3erKVmwyK%
zQcOaMoYqq(?T<uv0h@+b)qzgj{Up_l`3VDp5jls2OFt<Md$$awi5E4ESn8q@!pDJL
zhZ}hQbh*`LSVH)>gqQne_9oa0KVBN2cA6;cYvTe@B%Bm6u0C8x3NXM?;T;`QpO>(C
zq%tXyQ5^Va?0mP9S5zTiHiLKY`g>II>07w?<1-y^;f-#+t}R<~sR<TbEuR2W!^YiH
z;ASEOfs&LAZX{&LFCXXx?EXurdkS<*cz!d}-`DJ`0qfhhBGPsh6Q>v;D`<Ip`}x|<
zTbN<rpUj5Y*v;dIsNN)!a?kHBB`iGuPQnma^=P!=2+v?YwF@!(asAoND^Ms|Ld3wG
zgI65I?mA~Sg>|IT)}~%&k<Nd%H>rA00C8mTrIagpX>b!l=v>(yW$<N_a_AVDnt(D)
ztel5h#kCP}@a-q21drQ=6o|B0OTR>&Ve<Sz9RAnXHY}u&Eh#ue{rHA@M-j<Stcy`E
zLpz56_E@sZv1|lxfxou|jOE2M9ka$%2Zodjz+KT2AE$K!XhLaUs!gGW2D~wenlbv$
z)9$YQo>1T}TeR$)nNZQG`0n9=ngx@^rHA<is%V^c4fP^hSZdpD96rg+1SV{(tCM7s
z8RO#4q$KT74mF8wDL0L|5#uB2%)mQ-WD(THo;@dE55?I16xg?Y`z@8f4HT)rDhh<V
zo&pPC>g{;xub6_^6O>Bn&RV1UC}XXBO5@&CO#AX_^yzW+%3LC(MoiACqr;jAE9v|i
zLpnx_(VKizm>gNmoPMGu`R_@IAI>_-p7IMwuP(AC;u2!Z-QJ*s>KT_-s7i)+34Sqp
z$@jD263FIqtZ_h%QYVPP*yZowXWP3glYRI@%x_oNn(G7}jYd%3B87yP`7l1os4k%>
zGUZ415|eatrsD~`V9r~;AEyghM@W(H5|JhStn#>Hh8!L)rXmENA8w0gK=-8zgg7%j
z39r>dHb@{4a+sw-(DQw{1z8;S))(<DuB^CEr^~LIT3Q5GC?+xH6XiNJrsIS+$ik0<
z9X7jZlBm9~X?V63dT)n^?E5P={_Uy85<5I)<d>9U%x^=<b#2Y}3mvv3W#X#+v}j&*
z6gXW_iNZhOmO5A#?+Bg0u^`PpF(Eg2c>5n#D4s9YEyvNTU4E_F`~`Yls^>rBZ~pZU
z6u#X5?b)A+;3Dbwi6-GDO3u_+Y*#%n(H4fIv6^Q57)_AdZLkKgD$U44b3;z?5@B`^
zn^iOcy1D~%dF;~}Xt^J9zms22xee=cjHO<pO2~dT{Q1F7{Ri@^P}Dej?H3NzTco#h
z`jr;~_NH(+vHfWUlGeA^2~*p*FaFu7FyY}QKRo<z50H3Ph|{_ZIW(}@`Lo7dME1xQ
zCo{$Mbcu7m!Zg;eT|7@3>{8{Cct5a~g~va_)qHg$WH4p<sMDp?dwL0JH3$FKp(Pw2
zE6DFZCBh4Qdt?V@<$yXQxZnPu@0Mm#qRiQyL$#9IM?Kr^*zbREjBFA(WK#8s5Vb)K
zLX@vzLvr>~e{|b7fMoHMIAeyfQ9y7m$TZcVl<Lk5gu_D7y!^;xJB5i#r(%q>>4aT~
zNiS}7%`MQ<nzJvogY!456mvr84o7ItDo$FakSc@95OONMDvj6FfQ$Lel_{eth1A(=
zGLQcaZjHyrUmyh<Vy!VbcF2l;c#TUVO#K}lMM2!#$!HUjEGPtG@rCrO6431&^7Mv_
z`3~4?pco3coKSj;{Xm{tB4=FAwo-|ib2;4F9q_o<s4-zoQu)TBm9~_D`<EI?1Rkal
zt??<&$5m61XVk(P6?dRq;^TT}wXUD@JYE=7c?=$G2I9)<?&Ukd_@%Mk%b_Q;ADS8>
z$@D%4lT%6Oqh76TUQ;oJogdbEqP#z_of=<3AEJKMFl8b7<OKt7p){2yF@6_?_h(Ec
z|H~cvSBpMC%C~1RmB4UsJqfEzR5{@lvn*~-xyg*=<xOYD3Tg9+fqLfQx#C53Q#_RX
z=kOE`;&dk^Z1Sprqs7MLhSF=%V!CvtQ3m0zBalEYZ~22tnUDbbzPH|<e+YlgO!snd
z%80N6Jqd|?Ep{^;v*SCahP1C{BA=;^yY%7J_Ype^6{HE51qCNX4Lr${s!Orx{_mdC
z;R+4+D!YgBDfOcne73XK6dxpRx}f34(?E`1d>LHY>wU++7c?P`y88O9VdD>BAWj{e
zqX@$Kt{$Uu&KY&?x(}u}lG|hps>z_?<$b-Mj^-@?kO&#DXxjbqpKZ@}A34AV?Nh!i
z$S@1Tt$ttSc<Rs0lq3I28;kd#CSs|4R9a;ZfDic}P3IVu>GyZ>Y)*Dfwq~+z+ml^W
zO?FMTZQJ%_+qP{!*YCg9^RhR0t-I@7=bZiB`?K?Jf(G%l6v*srS9sh?bqIg;*Zcw?
zdc+l&#^kQOKojQs-|&K17y)lT1@LKE9A?2+v-%$0Ay_pdyQm*X@;?0za4SD-qnUgr
zJAhgwyhzL(fntZ^5ik%0#wCWsTzt|GJQ87hK2FXE@^8!VK{@#6`EDMA5w~%(>Xcpn
z;H$3a=y<gEzlo#Cd?|+U;;hX3qYAR8j^FexKO<Au6Dj2%E;mV0qrYUHjsbpPJZ_BI
za4G^sfMWC#BRyd_IBs1#X2L|dgylr@=ri6E5CH0W8FVU?EF?A#7_n?@&#+}IT0;!%
z<2Xma=-}lGp0lYHtfi`$K6sR}1LI8=UkUYX=hPvz@Jv|*8JL}K@zJp*yun)>?3Qkc
z&heWbHgvp+-&|8E75zjV?fe%qOM(?NvE$U%vEApJw=(<}73Kw;#D{7B_Vt_!q|m87
z0=H#-&Gs+bGdl{jc|9Am0w(RzSM|mL1tW4dWng$KCIM3zH!op7H8#Rf!uPgux|IHK
z3l%OA&11sCdYN`43^P+)&R~Cj<Jwi;ez0b$L$B>XnDqRRXST_tqw&cM!|S7zzwppq
zYSmh#kmo5Aq6tbvju<9(|Hw&)x#ZAlICcE__Z8WEZ_qK@<CQleTa7(K3^qG5a3$0Q
zkFg!TPH=VWxJ>W}F29e~YZJ)n9LN6aqUMIToxeo#%^+s1>zrWg#oAZ!qmniI!;yH3
zyEuR1a=y;{XLLSGcP-2V8HeLIADO?Za$)9;tPWWjqiSiM7vQP_n?bd1brCn^tb%TY
z-9|1=mlh&f9uKbk0j}+|wyJ|vc+L+L6BCc;<M03P1-K*53Gqp3MCN4O^<|p;R831@
z#m^3-P)b37y$V1p9S5o%C$o5~xng$wi{1h%!v*S{NJ&{MB*ry8my5)ejJd(d*BgaC
zYN!6DKxiE%GEl^iQO<z~9g0jA=KJp)|AkK;SL5DAGM`V8>^%-0L~Dh5_<lD|<(V=n
zgs)o64!{=wBhT#(b>3@KgCV$OXYuN1a#(CoT6MQLkQ0XUyI$v%!bV*GX9#=V6Rr#Y
zd~CmR$a@}0q&-`z@-XS^*V|b2UWHh`?MtZz8t?#G%2#9A8Q?n&HOSC3qkvcNS57Pj
zDpUC&bqrM$l1~^b<3wGKbE@s2Ig3%saW=%&29YWtRN!NBmy(cr8(U&tmJC=lIW$7U
z!s3CQBwjyL#%vwMG1fc?{kUQ@nI_o|#gV2FWcwSHc+h>luTuX3B}-ams1pSvsCT@m
zn4x5lN=6Gv2)PUw8c8PV-iQheYHqC9c~P#<fIoi=&~cfT=6*5}duzLoi+jxILl^t{
zgv=ka^uSL4J}7g!xVTsiw5pi3?KJ~4l{A|N5PV3DhHYd{UYiVI{JAJJ!2=QZmjnw>
zD`iVXJix0phbWfGk1g^Sv7-y5UrX0Ed|}eM^|@9dUi7agJjV1c&Nqs1S?0s7Eaxgq
zC}D*4ija|hZqZNOJ}(u{68Y@T>(4hLCT6OY%#xzWx}f_%FxR-o&&_|J>Q_!{+SeHZ
z89_iiiP7Oi6?HbC8*>z2s8$+pvZ-<MSW{jg9tw(`k=QPHbG@^gz_t-`HUi;1cJ}Q{
z&If$|JW!l4`s=`u3iR6B4yRq}(`7eJQSRy2FDoGMfLvmvwPiz<&&AV>AQh+Ifk}u{
zyU|W&b+>qXX-BcMfRViH-X@j{LnCIfdQ3i~xtG!uV#;=hJZAHN;E&&K+l%0VF53m~
z&uCoOKK@pQGr=-#;tKmt2H>GT4XCigXj(gs&HNDD7!#n?xyt7+;UQ$zxMRMopn`H<
zVOXk@X3+}>mcC)3TSmm%Mr=EIG<H}lm7TF!uixKF@3kj2){;l#u{zHcO$^LNK>v%y
z8tSUCJfLxJ7c)jrn>2A?iM=ysxic3-8tI)3X|15jPaU6x389lU86ylQk16*3St(lU
zH>qO$!x1U}ch0Zm1C+yvzG5W{96X2JXIS8wx}x{V=GU25c%+LT8)KkTAMWhWhEW&&
z&TOM;-&h7jG7eEaUk1hLzM_Oe5K}YM8o=raU0Xt688a~$bvisR*CC6*kC<qMzWt1{
zA2rQ#95mbwsL((uS;CtTk0R6tM(AgAMTodez6_L$l^u7>iuuWOI!^o3zp97cKDs~u
zmiY&uMt!}ZCQ`lK0CfZ%Zg-=Jv_Rk|oz-!1T5yU*N{o-+5V-OKwMAtmB`PK&E*u*`
z4c}b5M>J5n6qikSlt;`;9-rth$Z=SbViyxRy{n1c{sW%8WuAZwhyW&!mEZ2`SC$cS
zTaIhnbx1m~%+08W)sYw%4VobRmRBA9alT?~=`~X?0NvW+&XYRmLY->y&wH#-pw3V`
zRiM<2EPi6-wV-cCV=pp&9CuYJL=nl~fCCP)l~vPlB%UPH7`gy|=*}qg$DPf~o&G1_
zi{WqVF6&<S`npJ6=by{;zTOt3`X-EtD~&vvyKs<`b|I%%kSv}`VP<n0iCIDw1acAc
zBPzu&gQ3^Zah6*&V5X0RKe&XOT`v5w>z3!jRl5jyzB{DeQ~1^LAW|r~`RSSMhs)K4
z#nb7K_?i_K9#(S$ppuglY&YsM#~EnN3#pDa_3|`H|D^t?QmtlBC6PP+?ZI5fyAKZm
zo*S}uR?n#tR%=nA5d{7~_A0(eM~fXE8b15(U$<(%SC(j1rPnz)q<Ji>mVifi|029&
zpYSyq-x`LbY1Jc+a)8@OsFn`ja^Meaa1|Q{^_&|SGP-q8-#L6or75=3oeTmAxSyQq
zX?@DJ(D;0&O|M`&;=#6MeD2Aq`(Qfsl<n#`%qNuowhS25x#e8{0k%>ud_vW3y!<0n
zp<SqWNDqG}DER#hY2TG`^yg{XXB_n+YxXu&$XWOlQ~<}YjV$n!<sc1R%B0hLqK@U`
z!S~c8h8$hN>wrhd_cP#od8$Q+Q=#R<sk|OGFVxjod!kDgo5^bIzig{q0qFwPS`j!j
zj@XOj3SXFV?&G4lypk}%#D)0D8g#>22<(xvMeg04B*~A&Oh`yK2eJYR#B-`h!_|}X
zITf+-unsCNRy8=;Ap$v-j0_HMA=QPSQB8T68AaGNDzq`o;RVR`hE=(Wrh7nvWg??a
z%e_iB<zS?j;59U>p`K;sNYwX8nORo~brHSVA8sD!C5piElVTuap~l3%Ls&33Lv0;L
zaTw;wr%Tu{B;W-QA_t&i_YY_MUS@-9Q^*91osL*>1G;)jVSdN$E|;@QXI-zS<=Q2<
zICcNlQp$nW0H3SP|Aygzg>&=8qbgjlIss(t^v{t|X{rCf+9TC+KzT<r5+RpoXIl{S
zK7@}w&~^;?*&>F}`WeX8dr|v!2!(kPIHrzkWge_OY49VZFT{+W1sOWy0`}bhAYS15
zn%FaA`6=`tp!UrKra&hNx~nSTd>=6s%$E2uiT#NEk=wvi!MZd(JY^U{sob!t!6JI}
zJ23x-)ohB7-)T9qTy<7dY_HT)IKLvC3D+!ld$=#yGt8QS@lCd<bupwi6Jw`JukuuY
z?N^RB6MiHmvFYYyi=KUTzvsVBj~=z_wU4W>^!4j=@7I%Jd$PfEvTG8`5-S&KOsYe1
z_r&Rx7A1BeuK;?#@QFMxGi?=3yN2uLN}2=<O9h)QK_s?x+|;Dm!dQ<p72(Q!m|`h<
zE8-;OJ%*r^GHK$wQMks3pu#JLOg59WbIn=vIT&vBc214lQ+ms@q6znAjM?G<tHDDD
z_f?L6ex*nUNE#P%KEyIrA*2FiI(X39ii%c|>Yd&nL6e1<P{_Ci<UV+!U#r&G`v|wL
ze&xtDqq7`lyb%9XH)&s+%ypOucTs;}@chdp;l%Gwb&uaTczhA^XYA?KB=!w}+YQij
zt9^C^@!(!pU9yZKu#>)K1K+_hF7D>KN%Z^3Ua9$wR@trL;c+y{EB|-~)VJX02l^b_
zyI(LEKq`x7q4BZ>r(YS8@`eHx2nD%+`frd{uB4(Uvl)9_U{OCrrg6qIqU$$}A=3~-
zTgwY8uu}bDxDMLPmmi^g`~tD#K7XRBYC-fIa)1u%yS=XG{guwLpwv2BD!%H2gfAS~
zG!9N4IW%rIF9@iJkuk%w^&Ek<9Es|SbV6!s(55KIz85<fGCs#C{Z>StI?2Z7{Q{E=
zTdN;JT}$WW8+N3mw+9(hNncNR$dqMG%38jJW538iMrH-R#Zne`;#8GOk{Ao!$dXwY
zd?NzKN?AB(SPS3cdgJgksmu3zZKOocNqU%qlIkS3{!M<=Y)SZ7qtaFfT~UBJO+JI;
zr}I2^=cLH)qYSj?a49&@ari>EDYa~k5PaBey<@&^z;d-&|Fl6aJrwOx3IyC6AUBK6
zrl_ldk-o!1YrWaO+-ra`>Vh>>1KhDbr{@FfdjV?s-1<5`3<RToM~lVY;K^JO@5~dk
zmgJP}2B4_?^f;`TV)J(X>=TU`l{4#CAWM?L*v>P#LfF8L8pOr5ji~^AO{mC_FoE^v
z4vLUlv29_Fos>x3QS0c=18!>h;?XnUy^|{Sg~Btg6!4c>##}~q%h;G)023NtHk{KM
z#F2*DfojhF!v_5_so7z~sRhZIE-Rr8J(<qC8IXKl8`vexzLP1RnDda}(F}QsjJC;{
z*7b8lQ)+PWuJebw&djEQg-o|PVzoKZq;u{OJZrCIQOArNrrN1fXm8(!490g`#Jt$`
z>{qqwc>(#2N7ffM+x2%PXFJdxHOM(6#i9{brkl|Go53;TCQe=dZFr#fu$o9<8#|&B
z9ki-+q5Ea<zN`#)<<G7dR|Uz1Y2EE<q#tLc>E6wQ6>MHv-lFQ>wT*fi$@?%UaZGvu
zX}sr|Ts3{V7W^oI8yZ}njxl+}_6#5zlAF5pli`rxlXT$$Y|=v~j}AT}_ZyIn(}y(z
zNb-rFgL%2*DM7`)OF6HHF@;&Y9>1=80PY~Km=*`v7L6@y--nHuD{;Zdo^LTN9LvS+
zvoRv}N+O_|G=pLI1CP7jAP9D9Ak&m3BU_g){wKcbrjm?D7Vs)>zMb@Vyu5o6-j_bU
zg*jr>iR<LDFp|yuTrq<Jb~AOzxM?E-TSmmbp=$_tL?<&k?hKp4`hwyFtq9YHHqE9c
z2y3F+TGfxZ*%Kz0wg%nRJa#P<3zHs$IuJ>27L&pa4tm^5r_LGwi1#Rv_D3RW-gY{y
z*UyjOO;00tK>oZIdvEO(Uaa`+H)q-NN?fBq-|Re!E$w5z#x}c-yZE_M`x}BHuVh{s
z_dY3!NEvsmA-;!WGJnyQ(;$PDa{TX}K^+?~2&Ou1g7`Nh@#Aw++s=iWR{LqBEs(4}
zt(B@!SjP$xH646Zt2EuZy}=zvL7dj?eO7Ardk<z!x3BFGAAG5-=u_;ZboPKBcVTgo
zjG29*p(%`4%1C2u9;@FLnM#k^PP?4hL1`Zvbt3L<4{cJixbSdQD#*-M2H(dDD7XcJ
z8=!1B1_aZkbsH>2WwMo5=H~7^{;$WT|Io$JaV+!j5P+v{w*5In_IVc8)!nGdY~R#v
za9xfe`BCW(h-?9`{H!*k4S*DR44}QaI#I<r)GKvNT-r|7$P;VT<vfts7Ky_UHB07-
zM#L!&xdcu=4W1j-DV}_GM}Yn^X{`830WTH{?stDf0vSJG25?)T{>Fb}JTL-A<6&w*
zg*(OmLrU@_#a8L`Cq)dB8pD!!(r8HQJE9mpACK!XfWI-o5aQ`T{D>;E8RMSm@$#n{
zD%_OUgb<{Q9c9O)%Rkk6s38%i=;ZNp9kqw-!bWFdOhG@Mpw`!;AqD!?eO@B{$9*m+
zTx9?J0ltfj)Ov3UqxQzehirWx*((xrP3i;eySZszi`0=zfV!I^p>`eY-&^w!l<dgK
z=euJ~X$qd>=stA;E9tfahm6q>3Hr+6U(XAMrY6lHjRc-jOQqYDZKgl*(sAV@<9QAk
z=HyyUL0YdR4jns<ZXG;oZtKtz1Fa&OEqJ@lX5~@LE;({$4H&CEvq&wX#q1V<3nQi$
z?cl#Ny}cmy&zeCJwo4}uPG(`GcF>tn$m+x8x)}uP7f)mu#VTF;&{2QcGfbwBu74M*
zU@O2iW_Z3~+i7G8Hvi3Oc%UFt+d4{s7^qdwWbgy$xm2iP#SG_P>LYXMEb3}g3pQg;
zvCg%>e-4EcAVZoD6scLxNM`V<wRw|s&-7d=RDeW`X3;+PN2Y<FeHnZwshh*lPM<KS
zgZ(t%z*lv*0j^9x5GDM#P0ogq8o1X>;tX?2ApFE@W<l34;s5q2?kdOVMqB5Pk=iG@
zRJPdyCXPLQ{H?-@ijIN?;FTK6*xg}rx42WXJ~%^~J$4Dk83_WCl=ADhic{F2jMUFS
zQs&4<JYG!y&DIVv;Y`Ai11BvqZgmH?Ow^ZHXM|#7wqD~F6$ukjT&{EpenaoK!wGGG
zzwO#~i@?uDs|#W~g0bov16@vnzQv~&%R=LKTJu{6xeg;SMbxy8(-b<Id)igq6vmM%
z3^Br@AeN`XCQrxv6>n}U^~Sk#($}Ie1#WrV>He44_>8HZ*gPCUC;%74luF7S!1xx`
z51>zhYtZ6fc+4)2$G?E2fN!?!wb*P{y`<;%P%qD})9S<-E3w|?g_pX83~<q0j*wEe
zWM!K#0e#{hGJl|dFZ|Jz&wVd+wM@z|I@HxdCpUx2VzF6uSeTR!M_8979qY_wMM`{-
zWnq3zK3S4P7^mnl2P@poHMlQlvYnF~ME=UC6LB#1`}n$;K(2BWjSq3h{YIhg=qQjE
zk%5x>Ys`WsO%^|D%7Y`snn&$5RQCeB09oMRYq1j$6-F;9GFdMO!>ID!C@=OM$=>h~
zME4NcyApU4M~zQB{PjKFyMH(goxy9^M5$<z$tS(56cXvJ3wa}fu!xi2c=N4YA5Fho
zvA4>K8GTV^jUV0d4-E*vK!~JLv}bRhcO>>%0mB{S#9(lLBB>b;d_s30pnII-S}{1@
z?Eb$!0jQZ_nyYqdRHz7-W*cn9%xE7ot8LP_AUmhXB(S4Zy|Yk_{2|gmKx$e(hr{Xi
zVEjvqP9}bVf?iVe2a2F{L%d2fX}&RJy(++RiDuE{f|)XTh!Laz<;k>t!N6r85hM+n
ztuwD>#-vHW(7beY2b!6#4#Tmoa4*57(**gwXVk6i9wjq3ght;sT8$RH*91e|%8`Q?
zM(_gl-G&LxW(CysU_dDi<A8!mn;<X<#Y}R$I_z7f25{mFC_N^o`MkVXY4;f~VP@=P
zLfYbrJ9Ri;9vRcu$<L-9t#F>EFRNSAg`mXo#_;OR5$p40X&<N>9+rKTfP4=Jw{F6!
zn`m@hp;O~NN@vUjYhKp1z3p5$ktNI_Qg80ewak|#bS?a$m^lg55$ott8CleEJad4F
zg*r3b!zWi&xPArXPJvlLRblWSr<%!SatYg=tC%2xp+(Em+UE`sn=7HPDrUmF2#nhx
z&alPn1>J|_i&IDABKL0FYqEpd*I<dca@5Bt*NSi@+wFr0O3IUyB-`CX|KnIs7-=%e
zZRhc~OyiAe9xD^A{SZ^KlU#-V)(wBBRp{lvV@{X)?GAF^(QGo~EAO-|?uDcrBG#jM
zM+^1C0d=9Hrq8+xG`I*yS<RWNZRP?K3K?5NPf1JDP0Uhanm5t7<|W?ZA__Z?GIeqr
zh03Ta4yg{TrDRpM#r}tFEyv>(f6ML%#$1sM-nEktuJC4Kh39$646stR>aCjp*?s|8
zEKh(l4A@23HroC(_5t*Mz!3ffTG!M4dQ00W*K9R3=~jP)fX5OTRp`QS@#8ubv|2W+
z1Nc<H`(9Q)_acnC5{wrmfW&5W!;&np^{XdUc}B5_(hEDR|CZg>@}506#cHh`NCfV)
zbUuiIJWy%&g)G&4#QR@1RqRMrqdbq_g-uewv;N{Em~oAB#voUsBf+U8QZ=*I=(LQS
zG{{@4jlf>g(1C(lE1uPK$7I1Qqiph!I8n#i`B*`AAWIq!Lu`0LQJ_i1mpi27n#+^e
z3S72BG5!P=leV^YdV0DKg~!_vk&jB)heJx0B;6_kx*Q*s_62<Fy8jGVJMHSA=_6H!
zh<OfgzD9V8pICUN5z*qAtQ;0&Se|8i>r7I3Y68B{4^?{VRr-=Zf7PX&0sBVIjIu>i
z^h`C=$ErCTfH#xI?NOx!i|?>tP-X!?<)HK<luhR?44(#8p<P0ouu;sZ+<Cth(-TyW
ztxK;Y;NspD5+*c1Za#)@0%Fv95^T6_R#mutv)>2yq_%`2Rn3;^)~fI))kp~shd_I0
zm+!`jDMK)h$)A$=(>Y-kA225~{Oe+E^A+m>x?{oMwQyeCf!!&RU`-N}psbLj#P^70
zEi}*GBPiq(9jzwki(=P=iNjsbM)L^~XScAL<{PSJtl79o3R;!zJB+}tP%W{V6;&Dw
z)(>VO@yD|ZCqWDBLiy&2Lw5+g<%1+-f;$_#;C6Ez-ou#WE-1F(HR$_4Jo>OlCNZ15
zvktfWC%{`fc62iU3PGdM0b?^gPByV3Mxsa(TcioxBBDQ@vqXO(=&S$UdcVRM6yZT8
z=@R*8C;W=bNqvKi-Rp_77*sr8a<QzxF->`xs%b)uFq71;=(U_1u~<K;N3-vVx?N5y
ztn$2_9#eX=si=Tx(dm6i$BC{8%is_K8P_q<KuaL>QsOC0hPwrzx|iy1fyh4GXO5Ct
zp|j4gY`zJ(D&UgD!?i{C2nJI9{MeuG$2Y+JyKJE|UM?*5%f{%Sf_nu3$bAbx!n2K8
zCbm$=$h#>x7igbT{e3&XD!(}4>2Wx`LyFDUR7`lqZZkR$WEOS0Tsc!E)@@qX=I5Qh
zdIOS@lGXs}8;0tBZGxU2aG__;op5;^j<?D4x&4`?1v>T$=wGsQ+atl0?pM<J)|?*C
z-6$Aw_1c`ZtF~#rGp|IP(LoY09!K8>pPf2URd-DUs?pfc6TrxeL}6!~;DEP)KD$UT
z@yAa{c0z?Z1q}tD_>EfLkoorGe2kN5DnU#yL+UfFD^s5kqlc=hWic>PU(DqfH^6XG
z;1$!Gr3nynF2T7@4wuz^_uMU5qWzEv0=4ilTA4NZLClXSJ*~f?L5C_|d$8!OEpyln
z8vkZi*9@$m{2Q_r#A$!J7aXSV0!8FCf4SEQI|ETti29T0iA5+5(rq~41i4tjwqENv
z-Sj}{Tj!KwoI5v-nKUCz+XoICIM;y@@5y5uum8~59b8XAq;v%dR7=bhGNh}*NO=x&
zq8{42Akler01PMWVWc<=gui0+_R;%Y-e(Sei3{ip!uV9dY$N|MJj92e0x@>)oj+|q
zQ9my)*fdys2Vd|QtL->EASnTCRUm{56|Xu{;b0GYG(tMv@-UzBNca5n#^YuSvySZo
zs$-#Q(m85@D<jO++By{^<{v3Q&ha5WM-IH01j~We#6vBsT8O~A>l;x{EcnZ)VS_|J
zBU;gaLGQmXet?;hU@ddttFHnjU|Y)>R!8~5gmxGgUKjiaDt;UJHQ-nD(vS;VtR~3m
z`K(qdezcO;l+_vH-5~?^=X#BDqS@o+n$K>(<T-SQ=fNr9$!sKpZ|M{*=-*RyBh%Sx
zIrL{lQ5yA!GVs?b1b4`<0comAdzvtHc>@z%eK@gqZRjpB7Hj#=X~;SYDup1myJ_3F
zdU$XDea)JIuVv%_)3;mqHd*c!%kOv}#+R-}7^SO7t^3belsZM1e|h+y6BV{!?*QuN
z$rYr9DDgAiAwv31LOU+iT6hm}1tg(?buLb(I8kOx<lF9u6vCiDE}0d=ur5|Ay41xX
z9<fo!FuHjKXmKN&4GIBG!~ThK1Ok&%s~6`bz)bA}CSFgME0pkoa)tGEe(t7l4k?(9
zf&7v`L(?J*110D#o!X%JShtsJ(Z;FigQ-h85IMjo2un=pBm)Qiyet+$#3gwWNo;cz
z&=sCyHFi}FEYdinEI7#Qn%iT(n^!zOL2T|jk2@x5OG<Njh(TCZxzhYQa=3&3x6t_T
z_C%!9W&Hy9va-*gAZ&PCuGUQeiE%8Mf!F=HX71$ucC0!L*Z|%El|9XNL-^}sdj1{&
zH*&kAecdK^r^psfu~}>U@aZktnTnqc0G^{vY3>K$Vg;b7Kdu9ik?}cHl(nyrgIHI<
z4Ba$oxE&NBxa72IYJ9$WFLm3z+q}Uy2%*1&{y`C8TS5sHT;<_alU&4;0qh>n&-sZ!
zWiZ+_rXvfp=eBM!oegufP%5UpugRn<!pUFqIRZOU3oOSM0jj#+|B76aJEY~sxeQ3+
zCrFoLA`Y@gMy7eclS#`hG6yU7hI%t^xi*%y0jGS1%+V@O#h!e*xPK!X6kvo|R^}aF
zc%Z3Y-Bu1hFAG;DPtYyniAbtPV)v$k?B^za%-9i@2O2-1e~!xI9j=YP*S||tY#OVs
zsw|wziq-30L*p~lBr%&IvnUJu6MYB|+JL0aIR_6}I;PSi<&TV>!>9(tmw>gxtt(<U
zMuiN`Pbe0yd%OzTtu*`rhA9n*Ac`#Y7=i+s3Yb6z@=NSe50?28dSDOtHZ&<<w*vOq
zjgzKT-7WzMcuTFZ1-~oOoKVfkfARAA6g4X&paZFVxba}uRQ0Nb$GD2NOc>~d%^-v#
zAeNT`!+Q@aC9cmAC#-;$Tij73p(Nncr$p~u_>_7XWO5yD!H|A#U!uNHwv=v-luT@3
zClnGtFiI!37e9~>f}$vMTD@{w03PRzo_7xiTFB!m9Q)eXo$3lu>90<e)E1jvuF9-_
zItP1<9aaMWcws2sv30b==>Y!)8Pc!m^1*L*O<XxC-%X;xXz{SXNHbBHvq3Ml_W0v5
zl)8Wp7{MSbd>DQWAB7Qcs&a_6O+ST(<?ymO90Iv$u_bW(O48~QIlJAvsxgK-IkWW!
zH(D7=Nv}3w@dp!mzOQfRx<+pni%)UZA2I!m955U;jNQQ71t6cL_UOcQ_~mhg{w6wS
z8mn^6Is4a8P}Oq7v&UJuCa5$#<ccsVXh&#@w2_1H%907}a!7?>28L^HaRp|-!$*u!
zJwa6<T4JHxEK67E%h7a5*lisukX956P$}(#`#_kA;)Nukqk8(EQ06lrTWAocLMntI
z{4%&xniHV>p&?ePWT4{*0MR|U_>~tR=p}KA5}DSqJYv9Qlu^={6<RG?*Rl!@H!Q1-
zr0S(t;71L5`fcB!VPt!^{d%m>{1=RLp|+XI)&~d>`}Y>rb{fr}yMN(h?OPx}mR46^
zfdeLE_j?^64ggfCX4@@W3k%rifMf<8otGVPpuEUba;DuGkJoTNW2N53ddux`eL#S!
z&lqW}N>=OA<0g-sx80fB={cl<zuS}D%SVCS2GePRibwHx9pGKe&muyF5lc}X|HYG9
z6z(Cq?@Pu^Wm|cS9s(jHz8`#+>WMPwKd>cFJ|)LREA_TlE)e?wW|d+5YuY73$AJuU
z=pKJWC7p5?0OyS*(<h(Gn++J#$kyMO!Mh<EhR!3)5?~zqbV*$Jl)NmNJE-e+0u+PK
z2mTJG&uhD`m!y?-Y*{sZ;slJlP~+Lln4ts*j!sTRy8M+vKdiP50-cueB$JI*d0HCR
z8Gg+{7OF&Rm)4#zq`-fWnQM>#?*%~76QgjN4hR$F%&-lg2$e;pvYFWMU2Qoz@Rbhy
zwo9|JlmWN@?Cz%`L%*Rets@%#YtIRG+|lgP4R0bH%89+Sy^39HT#c%J-@#+UUmt}y
zehbStY{l18_$S_lg8Is2tIR@s;1X=&U?oovWh*|t3%L0B4~H5H!Oy*wTzo8c)m&b}
z%Uf43wRWS;^%ph|@mrc!2r*w{yNRZfh1}3-BzeE@M_T3fC$eha1R=SsRFcLI6v-6#
zJaJOsMdWxZQz2a|R>sHPfh!o52odqHS&3j}5?Nn`)LyuV*uvsMG{DBq;(fCp>+*sw
z?qD<7TzA`cya|?Y`q<vdSFdhd(7#64cO26mM;ZS{gdln@YIwrDMcDcy(jTvPFZU8_
zy5r^Nt5;a(lq<7q!lXgpHPIkZS%ZQ64T$$R94f+JRcfYG$*qh}Q$UKUD;!I4e!afi
z!>Bi%q`;RJhC<Tnv}~L)PO|M7){A7Ee(`RhwBP=I{HGTD<-<<`t5nq3VnB1uCe3fP
zl2idRte*uYY$O{iTee_+Z=7QyeY0%0wv5h}1zv#8DZA?HzLI@&f8IMDXPIzc9Mr94
zne006kIqjs7*2B8_dybU@_tYp396i9%<tf>X<BWJN4#VtR(Ua~#`KW}9G3OmSoAtS
ztdA8|yJk5sT>_r#Sql8?O>34Djp}FLWYowH@t*n8Tv<f@Uu?o<isDkw6B)~#FV^b;
z2ZK6g=WKONp2pJcEzsP=pjBh=wb_Z({xRkAde{B+@ehE=JDojVWP87so?s?G+X3oh
zF|b5`?*$E5^VQgW0c%p6hdqKq^lG;B9FQ@KWxOvwp484a+PmADvM72f&W6C|C&C0T
zc&!`t+BX$((@t=r1`9Z^W01eMdMx`NaK{PdO32#qUWYXrVWylM3uj0pLtu^)qdSn?
z3)D2PmD2XOA9byrAY3O=cjgJGXJJNC)UT1oP=Bu_Gc_Bg@sSI)qt$-KlQ$F5)QL#3
z%bD48Wl#*H`^I9&gg?Q)Zh()duoK<JZ`gHX=kvzBb^kUn^?52|8@#busJzm`Y~;-c
z_k4J08V|W#?@!_y=JWQ(f4SZokloM!RXVIUS=$`?AZkY%GS18|qn7^d`TBatXk(On
z&j3{s(jp!x5jyz|?jz5H4F{h8t*-C_DSE`RJl|T~+rClHZWoWK>}QXv<8WjDkV558
z47@n9irFW>*v92q0JlheFK)H>56)hmek@2qiNuGHg0^lo1L+=u<PAM{3p}@E8@`Rl
ze_>KuE!c8@^|}XlZ;hiyOus!2VH&y7$%{qJW45lHNx_0?CBv3#fDR(ZT~km%I{-%>
z)1Z!nLyln<5nfbuGS|FSoB7bfrKJ;P2f~+`&AWjg<~(THJAS5J1>y@Qs!K>rliAd)
z)|q&8OYOJbHbN#4MRcO3U8^_CR;OwkY*dKojLfbp?u<#F*tvZPD0{$SdKKQAiuBGe
zXlzt8#=bQ#fU#kzS$>OvmJ-L2!PI$)-qWBP)fQ9(Jd`kfV}_0|?<BiJBwogUpOU)y
z&3k%Sj~Irt3h}Hg|7bT^r*n9N4JH6Xf^43X1;A`v#^O4&#-gd_I7{*^($RlYN?V>q
zN3v&Jtjr|4;2$Vrc@3qUNsbhz85(0`yVOHGCXDhC?uhqOY&8%m>E^QdW^7vd8c4N7
zC4xq6h_FraTLgBnRczhnXmW#+MsC)_%3aL*-hl<Cfh>a-^f%{1{mD-eL)e}$EG-x*
z;0M3wji44&k~mI?euuDMGFrkk|5yS=8zMyyH3S7%u*EzV5)g&u15M5kW4w$7oC{+t
zCw+S=*~Co_QW3LlJpmWqQ841wOd%}+luw2F6|lVOeto>&9ZiO>oU?Cvhtu14h(hD6
z>`Y8d+=%S1r`7`zqtxfw=S^O?x?YbzSA5R%DKCzPHme;6fJsmxvT)-nz{>sSW<v35
zQTs2>#K}09=*iclN11}UaXh)#X(S364+alpNgo{s+AReI)o{M>1FA5yc<0qkib%5d
z^oD7ns$4%kJTEv$lny5M^BglMofcQf*;Z0!-0Deck|ZfJ6IK%mjr1(4WILGxRx%I&
z3@Nf)!mN#xTj;(S*KgcXDMH|6Iz&<Z_Q81U-T$3`eSXhrS%7wP#dD`h0CLU`%t)Hb
zXMVeGuOqp!jdms8i3xfsX6^KUh3FOAtl5$-q6Fy(f4|F%;fs|6Y%JEtp56dH{bADQ
zOu6a-%VK&<q$q+2-sa%?v)wt4J&fCjd@_4c;Jhw4)};o|kU2Qvyv_@EYVl(&(L96u
zm3I9i3aJGw6^e;Xj@JN63K@|i9W#zCpG={YHDyAr_#b3cB3nHVNK~Q$hhcrY%FQ^g
zbgYG&|9)OW4uZYeCcHdv9apB}&j~zHuL&edrUGbE#p>;1)DvTy2#_9&2{rPV9*a9e
zgLn-iOkO%kyIayt=w2c+#owl-_rE$f3tq$X-_F9vhl(;Gp@0+i^Iwqu3{HL6i|B8R
zGU1((9^Q(oK)`Q3Hz0Uf!5L~~m(S;B^d(&Vcr0NQ_XX+Ludd{efe!d`D-T-B$Zicc
z`dH%Js72}(e9tj*Hot_?<k>?^m^|&w_kUCjjeNdM?}Et%V4k&8=!}Cq*TY=2nGAp*
zBJtDwTe6iZcZT3t;lfFTdU=&ko$1E5UjQ%%D;;|4Y<~Y}<04oIqlPW3I5Z~4Zcs^}
z_!h}}gEp(Inlqu^|4=!i1Q(PaTHr2NjeRJctif_$P)1w89nuO_#Kqv1#5gpA#DNlq
z&*?sXXkS?Wo$@GNu~6hZVuA+=@u8`54G0VXjtEI&nt}=nvBCM@{lx;4n2GjTK?uu8
zC7hYQvgi&FA{$Fd;o+5#a%Lyg!?%dy#oc@#?tdcZ-(wz&wL!8%4T}dR(?<@6b+o=A
zUKiYmgi$XIsf6Y)i5V+0U=e?lrcgS;O+ftIynWL5I9LFN_PGF*)`<qlA`Sz44)P_x
z!B1^mJlx+uE`it!Jv}}D=ks1B_XN<^&@#9F{<ihCX;}*ZukCx+U;YL2iPKUuQhViH
z8O7^mwE!er-u5TpaO1<2F-o(zLSaP#mhzVxr#{As7%V<IQuEu0=6ABW6%*;?sch~o
z(Sewe6^%T4SpD%4W6T&^4HQSj8CRjj%|U|Q(ufL1-@vn62@N5r0@7zNcr3bOEj8~-
zxEhc^zYIE%cI06a)SO9Q2SfOei^yUYpyxU=tCER^5J4O~cenr+Z72$N@76Wcq5o2`
zMvFPQnDlhS>G8;DBS|YtwDJA)e=b5IqIWQcFktpaW8?M*iq&jx{MX`+@8E-&G!weu
zg1+z5*cp6_ViYa&Iiyd_B)cga?fWqf(WqWAtNZ--r1uZ}*_=2{8@4qW%O9dzma?U4
zkit+*)%&hCoxg7K1}MV0hJF%N3K04w+0b@7a4Qzpp5qu1f`jz#Ks0$0b@1>7I@!BJ
zluww{w<p-6-4~#2sd+$4_*^_TiDd#_yN2Sw`H>#Mj^f2bx!1l^ME5WtsYQ`unSxf8
z4hUr~($?P`7L;<S=3vV56?rrOj}mElr+;=+%P^gmIgEr^=S-f8rQdxnWJE=gb_fFe
z-K&>hOnK%ze)jBb92)r%qh2lglstpNjZJ3n{>^g|z3ygf!yMk+CUF{0Zi%?>kULx~
zdW6uq<Zr>9P+~z<4Qr%G5*;l(B2|&6iaDH2!fVnMe1^$vWK$SU2<+BFqUsgHCrIBF
z!u5(YtGC$JQztR+8=H6TTNnX9Ny^FN9_pN1cK@uu!`tWfO`P0F1YEC!XLCR?UJ!MH
zp7-x};K-REQ!VoAk4_!`9t_I;BxsQ${JgSQv{0*E#6p5$s|BKk;=W4`cfPI%_#u(l
zvUQ~zZzV*;y9#dbZf;t06EV$9%*d!cDX0h!qD4NC_vo@Yb0&xTtw=|8BkD9OM3;m@
z?q}E&7@Uu$pJk76@qF@G$5LjyAn|=uRCS61u6~85|9C&HM3N{9MYiBJl6?f1gSW7l
z>26@p>y<_kRw{U!K9arL6T;LXp#}coNOiC3UZ>aoM<xa{4mNbrBtqyQsEr_nKg+E{
zQr6rBuh;AO>z{0Z*xSgOCvViZUEO(mS{BV$#iM%S+zR+t)Bnq?tabhuU5Qk4bi70U
z`Xs@O%;Ir#m>5-^EAISsBB|DnX3z~rcDt-Rv9kfL82hDjYU%NsG>$0}X0w{i(laT{
z1joWTnBuLA86_>^*T{q*dILZTX4)v0l|hy}Iz@qP4u^<B8Anwe7ma9gAC3uW##ISc
zL$R7YkhB)aoQ}#Ng1*Qkl|o9W09&)~4eM9jMS-u=GQReM6H@CMsZ8p-R<fK4F!?A1
zc5_Ln@GQF)L7%~YGtbRiUg`<V@7*WwRzF(5*wrV;<1p^_tNBRnWsoKZ6MP5h&5a;4
zxP}{?Tl|`XqKOaE`!iy=^*DWmi_-cgs!@PNq;)RmU4elh*XU%fs-0HLni?#Gr0UWJ
z<TL1jcZ$g@t#L?&Z_O^v54%;)5e`Nu-u6LT!fGi*whFXDZI1}xhzN@6K$Q3Un`h70
zPzNjS%W`k?uIfH+4kH98?dtTr47Z~O;1WWMepqX;fS;>!8h&DE6zwB0=zV_ARiG*(
zP^#KDjMoi4wGxYVXyJL6n!!x}7k=u;y<tDcp^vRapVaHECY@L;ZXsDonWjrfv_5E`
z@Q5|NT5!8azvkg|(q?g|c0w%&Z(|p4o9$U=dmj=@RPj6<+=2)D?d|*}T+5eJ1Q|69
zQgIExN`UGXrpZuI9)%GZtzOXLw~qv=pm647$m|$~GT!iCR-w+#eQ|woFeEcBDgt-%
zjB6?0-$eo2$XT3@RdQzA`}+1jmho+=i%w1((^Sj(+`NyOIxf5`T(SqkG1YZxhkHDY
zq0Uvf!a=7p=0)7g#f_0D@SprLmOa5?N#gv>wPELNox;U9--!B4SXc0k_VWWR37s=5
zq<#i*zPBFP5Axv%1?GX>wg3@#N(Y8(IMjG*Y*0kcdjr7mHgSL_xt^j}<sMRv#dH@J
zlX<EEh+X@`GeDDccq`QEu@XUu9!-n70z#_4ojCw(!rHsDuo53z1}0~A9jK95f%vM>
z6mz}>(55XyR9)viKsu7h`#N{Pkl*3?7@C$Zh2^CY^zW~0AZU?0g_H%<`T?9lODn5x
zKpiCd+_?V_48OSepFY^>fqWXszF}!=`*~VkJ*lE*yV3EoAsQ}@=#vG|*MaXNSs%X}
z7zzUtQj7f!%f-X6F_$;@ijE!|F`cewRc<UY_zlb_;EPI<Ku;KxD}q0_$xA}J9Q3q!
z{@3?DIATvnHfMnlH)ZTtmz`6woOC>i(65)+kfQ8PPm-C-JCOM@GW6}1`d~QXD=rnr
zfUy)|kF!lrIe>&s#*77(?YWQ|!{>P~+q(<FWH4~v_4!NIwng#3yHCO+Xa)g)Jiksg
zyQWPK%wM-|l8|36Gf_S)e^(5T0>8gx7;^=LLV$({X>AvO&xf$qlUX*Xi!n#8Od5_+
ztbkfqMby8mMx81gpm?&Z5k)58Vlkes2P{Dp$;7ngu&z>$^(LyNh7i|OkrdgOOs!&n
z1wIaZEyZlkZM%Hu4i0=#&ym=-oXC>xAZhp#$@JFUR_84MRqd*~+dtv}1{etk5OJEX
zg0w;9LrMkZfBQ{VnU-2eg;!cKsZgzrxFyIqc!bQUt4C~t0;c2<a0%<c%y}Uiu!DL_
z#Ek_Ka{2HAMiRUxy$7@Wa2$cAhr0zRR0P~E((=@e!>envlh?Hl9uH#5|JiA>-2Usz
zoaM_%5{C(0%dq+3ZCXEZ6fo+qrGs3S&3*Dc7fFW%i<l+>W&a)fH;&EckUkfx#)%mG
z1m&)8+9k=Sh{hD@dJnyJ(;q-hKDAN~yCvN|(IF)IT{o;Zd^AM6>Gqb%cOjW>Bo?*V
zWYfMzdhM8%s*4M_hA^iU0y>kersLuecXl6Fr5l7aaTfe9Ype?cW$N+^B1cs47SU53
zSq1_NgeAgLlz-gj!?ag5C=VUd&g7HLpCJJ8i63cnD5YZSDlKV0w)Pcfp<w?Zvey~@
zezz3Clj{+SSqd_&6fBZ*V_Sk+rWoY}+pT2;AtqDOtTYFMH3Yllp%DkF94$U+PQVHR
z^*sH5XbF>H7U%yh;Y(UvlUshX*9Fl;i^>X5a(sajfz=DFBouQxqZa8s_@Do49q0>+
z6P4D(=dMz3CUnBaSEaz~H2?GmictV^2hij4U!EE0C9$(;eVg*}-f+PBZ2xoi__XyE
zGZCh*^J)08`p2&8^R?%5YFdgb*%|lR=?kE2dcMyLK|d0ewW(de2VRpCgZU)3qs#YI
zV~nCuDX%lLL)<3LX%>1u1C#^2OQ7VzDF++1>JORtw6Xywdgy`OIvKx`MX;+sBJtdn
zJy}+Vcc<xUf(3w{;9?^cs&ug6c0Oae7*T=Bzbp(F<S%4%!7DuqW}>TbP_aO;<V86^
z&=Vg2x&HQZ0r^ceWqpK9Vb8+p+dIH)?s62THnpm6O^`vN!V1g}@y>Ct<IHf%`^isH
z1X%pq)ap&f*!YYS>?ZyE1er@D*%``ysj70vbb|;;mHQF?u^hKSkx(~MhKAC0<)X3w
zt&%rB;CH;I-S4B<g1m2<-&Qx<F7=-ofqu;?a*^8*P9eI}N(7=|Adq}zj4NQpbZ9d#
zXB*iqhT68ieV$&xxZqcu=PQN1aq(FDVSHl-(tbixpKbx`5$x`+9tXnZ%`1qdw#=io
zBC~w`Lhen`PNK9=v}rg2%_9l7&g=KmPo5FPUjO9un=vOf_waj)a_F8T7ReLfn*t(%
z9IhH=Ql+B#ow7jyYLxTj0QMu>bCeVULf_XINx!<3(yKs7QSTeiI?~^Y<yUxvYN26W
zz5~30&G5<oqT0FH#P$tehh(mr3bx>d!??$zW)SjRFmc|V6R(W3)8WHydBJReOooQ&
zX7AypM?&lsc|xLr=_&ZB&%m;;^=ISt4<1kg_V;P1v;8TB-Ru%bxQ--A=jSy1><?L~
zSpG%E3a2Qd3q>>HWQ1gNV+_1YS*ho;Imgk}`UOj)eft-|3@B=e#VEZ(m>*?eX#J%$
z<~ACJQ>5uKM^2sG17`9}U^dJiPJaT3jBs+L;eN|$Y(+BMZz_=1GS3B__gZwReaBvT
zi0~^q-~s6-T!3Q`8w^c)?dGgNBrf=E=o{>+WnQ9)VJJ8>65q@69`G^wYoT=Hg`~xC
z)Kt%a$}Yr`8vB4>IG1OA%2-K4dzcc#;jZaAx3klL2CY5Ak(jJjMP^q2BEn@<&#4Do
zeoj~2+tTIicpUDR%M#0ry7|18r*k&|)Z4`oAFoFj%kSB;XmvGw{+-|LUyo1SU(-(I
zg^(3j+Q;318>6XV{bP>k%esCQpjN)y4?v?0b<1>kXs807O@MTl$^23I(`u2B_qKJ!
zTG5?0?gZgOxv#O}WMOo4#Z#CVJ&WIrkmx`zxIO;{BG)<pPner%vJOqAsOA<;|4G51
z21~pN;e7;qZq3>JQqe4{A<j@<;-+(2Ef0QZtO6TKv{{{!4R*q0#y7{8)6-QTza{y_
zK_f1R#5|q($;7t){7rf5YbkO_T;X{veSORO?E+9J?oSs#vcH~(NOgLAIQD`2oz#wO
zLf=KN&k2bqy})4pHhj+aHOSxKb$C7g;dofP%TRMg0wDGquZR&s4#zuM^<~rMT#JHE
zvtVX6E$K=)*JWpozz)>q)H-TNONHkQ_<;akgKmq8K<K`c2f(lF`ZMNiI_LR1{!*#k
zbhY6UQ#&HBqC&{y8VIX=7MTS6;bZDmKyL3u+Q3-#P2WSUe%3>1s^k^M-Yjz5a+S4H
zHJeV847yS9m3Ryv=7NOg5)A40c*UiEnw}iRVQByvSfzw|>0e2CsWHrQUGCMSY)Qd#
z$nse%CarTq<Usq*7{tjtbEb^b2%5ub^SV`hQEbt&DIqf^(x9>!fEkc@@;%;2wbN+)
zwNd0JP+N(W*jw)}UX(+|kuUpLqv4UhJz`OvI!FHumTOThOEB-#6NH#8NjdwO7F(-O
zmwwj$uOO7lbuH>}Iz8;coY*W+Svrq{A^;@IaUZgVJ7<(wvr??r`p@#b#5Qrb-t`=a
zgTXEBta8B8o`}iRRHj)k3>4{g3Ab*HJj2@+YL<HDq-)vj`K(KF+^ivKaP)GUN&4D`
z;#AOPAah;PxibEzN!fRkyRaQYQ@BHvGUxn~s1({NISx3K+FF7<d_1PQpznsJa_AUl
zw(_`BIsJBQ{P8knG{!8guHPSS!}oyL$!3FFxTVkRueV16VDp#fy&(%dYf9`^D<_Kj
zNwuI7yxf@P0rs1Pd_)Qxbj!@0NZv)D--rfuEh2@KO{{mhL8`F1F}vbyv-PY!e5+9W
z7ia27S<nq@YDR@>QCbB<SlG666dEP{Eysv)CdaTE<1lTK9yI}l*H<S7@cV$c2*nyA
zJAi>WTlmS{bfr`0I?}LIgh0p<^Gn&|F%?vS(DS3%j7C0KDA&j5>jR)or>CW*^&)jI
zt*@`oJCz??Zft#Ax<GY$UtMloF>3SK-eV8O0XUMgl-PXJPFAalQ~4;umyRa2V9k|+
zU)uY#ry>1P>RBWH@ZkEZfD70mEF;$5<wsV6SANZHpw8bKV;bqQ6(gTuC-MTwikA$T
z12-S04j{@FzcSjPzC(o$G4wBS6Qxgs;SlZ$%7Dwb%K;G)sI-6Yy}r&B280u!@sm3X
zeu>No-Cs1@Cj607=|q9-G1|xbgGej<D{-g0W8~w%DbxX${%X6&lh1*Bm}=apabY@0
z`ag`Q{=C|jZTP%?Q?$8YG$FxxlQ7k44zkUkjjv0^U3|vd^?(qNa&n!!aJUWq=S7A`
zO5-w2*{sQC%kBR7cW^y1mdoe*OSQb$$8jy1F*95HnTY7YoIKCNzR`yhRDG}UPr;Gj
z!3&o<uMGRr-CQRaAcAgQw$;dXdqK!JT6#)tZgZIZ9f*kz{<-G5orE5U4`Qr2K5Jh9
zcCi+4@ZaB&3hiK5QAUJKKv6{aw00ES=OgpoyM7P?fR3A1UVna+K`9#f?E2*litr8d
zrjy7dpj`OB5-A{+4Yo%!jq0m)H?oRjHyKleLd4~H-{VKBuhS{VpsY6SPoPTf@Gq^A
zwcH{VeQE5V8+$+O7*7b-1XAyZs$Jc(uD~YLJw&bonW=jDR#fq{EjBnov(_z`X44$G
zIc)yUr)_oiZ=CH>8wTS6SU#BjIHVpbF~yu%923aS+K{@Pl)e}lvho-U^mjpEOcx$g
zx~dqB;zhs6CtmRCyZQ7Amhl7Um4>i)^RT0~S|(#2@ZBv{@n-cYRF-YU=m8a<gw@~D
zBv)~Ur*!muh|yuhjHeuMZIF`PZ6F637Sh3bL-<h^pYbT`t;9>WNyO?A6KMjg*%kA2
zugyVrv@dM0H5Ix7+&fQDv#@SqwQ9H<i`1T01}7LJaKfACJKG+B$o=KPn0`zoDJk)q
zBrm3T^k|ryb<3zN+)0*{1IxcO)sQer<`}tq3Fd$M=%fYpoIz`_-{mg24T|bde*vMt
zKoH4B);Z9CWkB#Ml1+!^S?De*Qe#UPjpz$bwYL52fOOAK<qa=Oo{JT30+;GZVw<Hm
zFal|=kI<a}-NZ5v>38@xr#oKGFP<k7=<C+Dr(kU$6ta&_53)RQC_@Gh_dC4R6ZK}N
zZ{Hp_L|VRo*hbNuW>0Hu3i{z>n30&vkB+|dIAz%QPU;GlimqYE`D)ei@SGD$NQR;h
z01FXtdiIy2QSL}t6ZYz>{;^qWBJdUU`L%A;`8NH(n;t*{m`LTcIy<afSgIE9i@o}M
zs0PH4bZsqr--g12OTgnu#N!$hLV0?2`gp$DXx{5TGZo!~ff!o$pGS+3+2OHb<!s5z
zJ#W{zb5mgZ82gIYBQo{T!S_?(83=y!IqaKr<?jZ7xW9`L^5Jg<;wLHs1y%)o`7&&j
zHvL9Mumbv2I8!a_yk1GA-?AsEENwe}D?c_)g+bn*&$zPS%1dQ`0!Ab7H-01QFR7Sh
zDJ^!zads3&0Ve#9Qx~5e$<Kk+(650{V1;k2uDSy#O#OW^pw-MLKxzrIogZb=*wN>G
zF8k}@;|b`^F}Jb#c;o*pZCKkT4d!&No<SJsNu0->h}rfM?R`|H&zJQ8{V!g7GpY(%
zYtnhLTpI=;Tt3cw1XHHgFP+-g^`e*9T^`qgJ94Vs6hdGh)c|{7yg~NmlQA$z1=KTq
zhU$r(x>a4dW_!Tz*VXCH`|`XeW9{}YXo?~t8%XC$9iDD#cNs2@uK*C~feTCs$gjLI
z_mSN~Kyfh%8^Y(ieTf^8W|wwN;nQoiJmh8tjHbc~@Sg^MDs2HD31g&3Z0U2`pRq@8
zuh&PP(-X!jy#V6qjRX6(OxvAE3%i8m1O~m9<s~&lZXnb7;b(t!83XBbz6_hZc4Y5f
z;aszcO!V>|Iz&js<qDG(D?$o@&z^e2ke|*s<kf`Wai5~_Y5f<)C<^fkS%{BXe`~Rb
zv=2ykOB=UT2s||iIMc+~!G7ipoi(p=we-tpb3zX+{7kEMkH-`M)r+3knK!=z?M<H(
zrU4t)L$VdIhTIfjI!6|ws82FE3)5sZ^kdj&J)(okj_cAAi(d$^XGh2nRPy;m-!zPQ
z1a?E6mm5o{eQRiJg<;*{IXU^)t;XisL%@GxgCx}n%PH%BG^LNv<vMI}NxAU<UH~9~
zc3k>WH+D96t2QkK1J0xJf&}riEf}96`q3p(=IF73UcNZkzF`~QT$4V=NlM^Wq&Ut%
zNF^H(ny_Fj-ph2#tJ<9kN7ZEto$O`5_p|~J#c2ZOKMVGBzqo`G1wg%eZ@nYBn!~d5
z5l14bHMvuxIQ~DHt}&{!H(KYKn#{?zZF6d}?V4=6sV3XDYqD+Iwr$_@zjxgat@={y
ztbXTx_ukKbfEP)TAQ(OFlgZWJV3+gg;F-rf=kJ(;4#zxrk_9ARgxorxh!f|}%j#;9
zpGf@?rH)*wQ`AscY$WC50%NjOOtn%7ZgO%wM-09ZHS|<uwf3L3)Q>AUQ4$qz)b<|<
zlaq=s89^#ViV9efG4z%+GwA-+TQyGEBV(ivZtcLhk$`ZcBA5r7=vOwcv6>|to>~Ej
zs<EJie#;d>ZW<W7%^f79@YjaSxWTd*fr9W?J_JPwVml5+E(j+D`<g^B4HA6}CO*z;
zxS!qkhlbIc9N2a;Wmeh8jQ@o|F>K(Xk<EIw+B0)BopLuAS8qor9drZ`$oyad0o%un
z%r#f-7KcKj>@MFge)cYaIn2cHiumgnu%`g>ab6$J-X7027;QHJvY6|MvUmdUO1e<d
z;4mMrEEUVX{iiZL7f{yGk*1h3nq*m7o6PMINn9Yz!VRZTF`!Lhh9^rJd*hMZPt_HN
z@CGcR;bDs3>N6+8YD7<W2gYWHz0i_|UODn_(88ufCag;vvHZr@Hkr8Lp>bsr&Y*qO
z`*F3DaUIBDKO$syJ;cYifH9%VTo|+bRL#Hvv+ejx^~=}$>tmyh*|n=stVXbDbJz%T
zj|5P)A%@VMYdY&JJe)~H<D0Sn2*;%RFRTb?1c6O8unBjESdM=QY%UuO4O2>~T-|wu
z2n4!!fvdUvqnuB0EV8x63w6Ltnn;TMu*QfYI-66f!tkUhdnSLh3|<#}gg;av4|^me
zr%F@iVYw!ruhr>d>4^!O!4B|r16ISC{Ek^!Gkb8f8t<3q8jh)t-fswxXK!v_#aZkc
zxVqU}C%3}uo0us}hgojW-=+1s_Z|QsD1C!%`{XnF;j#`;&uG3B(x{ZHdDps+v16l_
zGbNQU&EXJDu>}>JBQzytwkCl&>|?)Nu^e93u5O@H$U4(p0UJ-6Onf1@>XgUcG4XI6
zIB`sdWkc7WcSa)(jKZK5I<^6;+n*q#)k)DMo?-r8<+rIU;kh5!<5XUM=>)R{_8(v2
z$-~975fly-X!6B87EdB5WHU}zDz$+31)V}}?n0D2JlFR;+Rzqr53;)s?sI!IvHZVP
z7mEWc)|a*_9Iq&e5OxdqIB@0cHLl_Y5WrT>@cmvnzzBm^+RMN^Q91s8$WGZDfH*d1
zQ#c1FpIBHplM$oTOVj3kaQXnyI--1hAYNE<x#3RqYnOSqbLUR~5rS8`G%XaRd<PS{
znre)|c@x%}2HHM=MW)^A1^L%5-kPko;EmS@F2ZUHlK8KK(EcBRq?GCmo0PST6)JKz
z<qK8dem5UyGyi#K3SjPkTv9T?yh@=sy-6`BBxRYCkZTQX2Gd!{qq6<7VYqN2$0!`*
z5bFJYsrEVyN#zwuFflXMAK>^~mVJyc<0&Jejwp-dd;ODjre!7>kht3h=4U5WFN67G
zfThT8vs}Q9xdA$y!$J(TJA0wIul=cz0uxu#s4DBPDDAT8mKW&oeU@qeJBw*f%2>5#
z%i&(MD#{r*N+7omF!v>_`eWtOlSw1~A?6M&-Q6#Bwp^O<Fk1x7FI!zNf8tE(jEzZr
z$6<fVO!WlZF3+4sL~nt7cfeZK2e8rmfJOw2P1b<QS={mhpjI{N@3RBp+1AT|J<mGe
zCox^F<*F$Me3szZKBlAP)pWqy2{=vzMk}^8E2%WSA(ToyYWBBi6Dk<tWGSA#Op`TP
zKw^&Vv7PVmq>B6m4T_P^Sgd-0h?%<ksS{8m0axr<)+=KSs0cx@=byUlGG4>9APXci
zECf<+3BZPS61Pt)O9tEk)O*M<)L3b*@+%2lqay38_?OK}`UBvfU=PxfVHdN}vp2DO
zm3~9t(zyB0<>D3CW@dIR$G%&!t^r~WkZ-adS%N0n0sn2{A+&7_4%b-PTu->o9<|AL
z>m3eG0Z;%XKyN`=tFsuy#mAz5+q7JSt%O*f(YDq%53;|gs@ZIS&lxJ*A~QWKY_tbF
zJMq_TL9tSYMSDk4ony3*k0CIQ1FYJ-hxwC;y}|bBuTTm^TwbdhE<Ad~e~WPbfT$g?
zXak5JPq3iLOJ=UE8-L@T_7cknBd`Htl4nF*cyed=$i?5=U-oiR(YsyB1;rzVbL!mE
zZ6=p?77aUFK4;bmZgt?*(9ap`U@@tX0lGa?Xp4(B^t-%f4v}6x#xtVcA&Dj3s}JOD
zeUT^xgnTNt^=M*)MICnBx)I|4z+1e02_XiuK8T{#g1&wrQc-K*`5;5x0a<2x)#DsO
zs*1%d!jzksFu}ywB1Og(Wr5N6q+l=Y+CZ%9b7(xLUN_Bi$7jxrVK$%RByeOI^jdzt
z2D*-@$^T5_FC@kJrAo%m%-e5l|NJf#{Cn$Sq2Fs#rc853#<m^?jPujHZe8nL3|Qa0
zglK)csS2FG^i!K;sPp~G1$fYUKEHV{){U!#ACXG7DXkC~AKwBry#nZ)a$ZBArv6(O
zwYR{%Hb8esL}N7=y~#E$Nw5cCLl!!0oygFS9T6w7=)J_7GIQaosE$e<A`Y2Joui??
zqs%Eg)9$ZOuVv40LhKg?!>Xps6Yd(4KR^lt{l1p2qDO5WGgcIoaO~5T*3D&{Yxsd=
zpM_WrQM%eVhiOp~KKa~=G54Ffi!JlUAU;7TnF>t_lTshf0{-9dLDUGLUMEw8y^DQ2
z0HXD9=@#thslXc(r0^k83-q!}Nkt|wT{Vh`*`^B_PsR(X88W12NJpt)EwCPxkyRcB
zi$S=D`7D-Sv6f*@p5Cqs8%fL!&hR=teebdBn6x`z4*l=c0(n;Z*T@~Mm{`D}?S{$b
z@SFrzFlE&`h>(E*r3R?4z?bVK>sG6CTj`0alF}Sdb>UyyWsBF^%%*lB;fA`Ni4AfH
zi`;;)OQ2O@+N=CQ-~-Zg4Xk>Wz4aM2U+<4ss^zxAle=xPr2jp#cjDmY-MTRc<t~-h
zQ^HwNJfTwgHwlND=!}mfAW0Xoh}S&~)G)LI{#e{dzvhW8xxxiR_s>KpnRhJLqizr&
z!i2d)@|jj|hYYUGdF^+42Y{xy0N&@l?q|WO`ik%jC;YNd>Bd(A8;I}apx)p<;If;w
zjqY=4>B;)RVdF#9CJi#C#FbPC9cA4#EBL$c%-K>P^L90XExUqG#eMN47^`7z@9s5_
zFGk?}`3AOF3s8yzc(7GtW*IM!^KC+l;~~?lrG-CN1OYm0iktZ(cjnBg-R%>1*<pPW
z^}6PBNBwzS1Eb{2d+w<wo?~^22&GIV5zL_bu4t|<=&JYUZX$rMRs%+g>vQIfJKN(G
zsdWH}+N_-m{GbUM83pIL28tJ+jhfwd^=DsMS4r8Y)!rchuFd2+RLGC5+Lf%F93}>l
zVCHQwbL8XtFvpN=y9Q2~Zwyk>Lj=U8V45gd_Bc2)bEZ6?nPeCx(Tal~$zO-y%0AD-
zeS#dp`LO|U!m<{jP8ikNhb{Fuzi*`Y<y|P2+jV;wEK8ab?6*ZJ=m55C+aazeZ_IKe
zR^Y{B8Odpb7xb|TQw98ZDR<`0ZXtu+c-^it85hlSfY`xoJ{oEI;}}O;MF1i*Erjn3
zYN;-A3oI@0SE?BR>(B7pPb9LG#|S`iMIH_>e}5rKhGqraTwV!M%!=5us~7Z5wB7i;
zFrS1+Mv$8jLEZcclP~}f0+`OpV0|(BfC8^j3!r8IkRxNnAxz~lZyKUL0#H<0)i~RB
zszbDIAyOQrevkRgF3PIoEEQd|H6>uFyq5`=xKRP2lIdhIxzq+q#UYXhEM}9?!m-Na
zTK2oj?oz)2Eoc2o=B!N_ogo5Bu2>DsH0VsKn9&VUqi13TB=aAbZhScd?N}lRL6BL0
zlHtvAb>>1y_?~}@Mpkr(A!@%)F$SQ1X9M`ypn%!L_^REBEgf0Uu^uS)LWv2Qt`vqz
z%TYpVQz|5|1V$8io<6xf#0eYz1m<SE)b1cM@E;$?e3*M9P92+efW@;4joO5p<shJB
z^1u-P#{(>AUeC9G%x!7cIj9WWpt_`gAo{)=0kvt(^|Jn>&AaUcV|l0R?eqzuE&o>e
z09Xxg`P@|kqczj-9UbmOc(>NCKQ|0O)!0Y%CgAaT7uD&Aa=%Ft`)RAjnw2+U<pztk
zn*AdR5yy6YpO-vXhc`{qAV5lcR5t94T<K4KP4Hm)ehkKLobY;coFQ#2IiX`~S4s#b
z6&k1?oJ}^Pj~7r^ff41zrnDl3CgOqx4a)Z&a2tC1G0_z<(&2N_3XH3(X=xI5VZNOT
zTOVhhh2B`j`dG^yuxQmG2)_J)G{{IITBuEhW7R*%>k}!MQZALZ5b(c19g-u&32Lxh
z@_k%M3Qy+AWAZXmL0=zhlJUdQwnn5AQ;1tI6mrfstL0Qi9w=*e>KqHCm4bVU7T2|p
zWHMfm@LLk6K+x_`bv*fruEBy)B%8{8`hsjWnHBhY8O82X3zW%SC&J<}BQbicb@#8&
zK%nQ|?DWRW7xgH)p{{@a_3t*1R(-)Wkaqb1gFA!J1C`WLm=rfY%(WnB?-QE=)^+jO
zZ<U+oJ!XXach8X#9Mz8q4DHF)y$6;nF9k+7$sgClfy*b5)3s%AP%Tg(gn{t*$y(cQ
ztL6us=Hm%*rreT<e8PgST#)-*)Li6nYBLCqgmb^$f5>K3XxHrat$G9ksdn!|e~4HK
zfXTc{$FpWH&RhuvAj|5kl$NpnyoUa6n!ZGEZ?2C?@{^O1W>$h4u)oM(K~Sw>?$qHv
zMb3EubbyEhJlw-ng%*jL+W6VjN_aVv??i)A1qW1PvSBkaiN#YBnJi7f|C_VoezqC&
z>53;OK`r9RS|iN;@8*cq0VAwAs_8~vnG4IRBpM}d5Vn4q*Sv8vH3a5pp!6;EjhzNh
zjCm3)r%%-HZv2*U&cUZDy{A&OYuokf75|SAW2Wxe6?-@q1~lbj^N|Gc9hbr2{>U12
zFp$>RWs-;i5kjYNNdA}*8Xicr8`Y|lfg#htnJmj^{Q+on82;|#qw#Qs*I7H_MG%HY
zVP~h3?THR`m|^KG7?3q|#p)BbIk+TviMh-gAiPsIoBIOrs=%X<P60wjeuivg$RHOZ
zm-%tKKC1#QS==OxXHukIkIM>1DH;C}4FC`$ir{J2Q#u)MxvBcl!<iQQ<1yVfEzRa*
z`VX))ZEI;cc)Z}*0^;dPjTP0Zd4LFnSu$zzj&}zj>r@g87w*6Fh0Ulx`$y{Th;F%g
zrw-`XJTm<m$pv3!s6}osj6xGCq=S^|zkjjDmC{D`Cv9@S`%PO}*)y!1t_;&9hLLa6
zB|caaaR;*>d;0q#xuKd6-pF%jFc<R$_>Sa}1qcXsjpRzw0^HO#|7zm+31oekef+;M
zd0%%L35pY4ej^`yfdoersen!XKS0<hjd>5(xZdTfN5z(q9TM?$4{-80OTPU2B=}u9
z`E>Q!13(T>=Jc5#PG;gm6|?vqKLN0KG+ND>{(^J>E4~tZ%dke>zbM@)3ASwc_kv%X
z1#}_y-o4@RKe`QBV={K<Sx^xI6*`Ez1tJ`Q+y`?RHK<Kmcmumg9HB^0^U?(;V45_M
z-qLE;M{a?LUpakzKAKEZ(p7w&jBthU?FJ$faBbRSy}eJ*Sf|=g;RNR`u1Y)nVg}Q}
z6XX3FqMpbB8)ngIbchTG$~->RidV3B9*|ry0ZZXIvmYALFsp{H#4sMiC$|n^q)ci4
zCli&SCz0PFj%$OL67NIM5(!hb-VtFgeQ#zlxdVgZOJwF}0YPK@>Y5X@G83Hsa@J$n
z()$!13kHFO@6fQF6%BSo(#c=N8vwbAQTiDGD3Sc9e{Q6>96HaYS7=&YNi-9mlv~z;
zR;Qn)1sR%8T`#CCoyuT@^qmmVny->ft%lo@$U2}J%442B6iKvEDE{kyNCO}U_6RI~
zuq(Z6uk#zA+av0W#*1RUK{_c?Y9vb>IV7M9<;H~<+fyVbwL&gMi7uY))+cQvEn><V
zWh`y4ckgpLB=}#S)AU+Tk555mWnS@f-(-nyw~dXq#}b4;qh>SB6~AGhsJD>o6I39L
z#C{NhoJ6Qk?4Na06=ucNs7n@#QRQMTSFPTr&S^A~HxAY`G;n`E;Piat)rdp6_IK(Q
zCWT79kg;M(l4LrBWr2nEF|n?5Qr$;_H+A4DTtZLPrQJ~lNI%Ao?`WZAmJ1!Nt^|Az
zXtJJB11K}<hO3rJ)!LPD^y$Ax4scLU5Tp~iX!t~fwK0mrrKh0!6v8KbK6?iMxVNkA
zc|(mt7GN0o1}jxM{w_8U32uKFOh4n-(6WDTp0H?hzZ=tPvi;A!5-?fc>;PuO4a?uZ
zTY#vMt<Tpdb5@Uo*#jLO7h0Wm<K5>QK%Q%SeqCj9pdZr3Yv1!8^dI)Se|_Z90f_F6
z3E*`Zk<z7j$?qki8~shHnKD6{!B&(2ejd#sU~lpG=?yIj9hVL-MR^g5zE6um2nxCm
zftT=EkooN@B>!eqsko4D?|Qb;9rT(3!B_YsboUz}%-{Z1YRrDrQmro$rF%@qmKueR
z368DzfW?OM`5E<q6_!|$I7;<mmMWdMlN1G`!wDVz2Cj{gR)(;F7w89q>3r#>39ZIh
z1*9^EZ$C-*IjuaeuN4NdDf(VUb~q8lxuqe~HwKy%UTz#qg8Jwf5n9qP+AJE>AO4~L
zgD)?zyR(0D=H6SdWCM_J#RZa$KT=mXlR^#4Xe|dLNIsx0v%jRb@yvJXnKE+i>aXro
z9RV1K)9Klf8E$_it%9%Bg@%EO&nK_<2oX{knWL=0``m2+Sb=pipP;w0&{=XgW6=bt
z3Xee};Jdib)6w(xgGl|*9MkYzrHX<l$N78!#abYnqN$ph^H0v*IPvqvUZsQYH={rP
zex3COMmahDkg$ShLzl$ZvO~{NkPt7T;y-gGxa}y@L5`=T;r<vp^g5Bvq301cmj@}R
z%cyM&8${d*l5-EYni{4N6cUYOUZ}UN)AlIqYxD}-v`lBDrc33O9NMkgeu6d2+M1nE
z<F#B{d}M1lu!fj&*desDAS^sw{fKId*$*%-z(yb&;e6}MLWON6dK-~lvf3OvW3Jfz
z1-Ym%n4>Hod4AISFU11jDE2cXo1ogw@>MIv>^~EhC2~1_pLd9;TH>)vk{Ge81!WU$
zOKJ?cF!vWj<)7Go0u|oHlxq}pVLuxU8&eDw)ZJ-8k#*FI*?Ht!NwHZn&>d@^<xWcU
zIZqM)?8R-bEtD?-cQ8)mK%!xps7iaxx6pc2Pyv38Z~${rK#?v;tu=rV?S{IS^ev6?
z_nci6$nvvGt4Ay*i?<B)!W2Vry?hTM3r{j$+u6M-@QP9TtxGAqh;J%a#8Uet&yBHh
z4Mxl1cvc`1GbOe-e8ilWuV_%qq-d;q0Qg)Ao81BSitrLZet;>H8{uy>w!LNl7#S9K
zUG8`|F+;T~RXkoNw~tomIx}I$3@JDUKykiM`IZt(ax|XN@)_T7G?iJWY2Ed;aRrd;
zJg&Dv{})uf+GyKaUw;n~c+c%@d<4W4KwOqKpuq)HJp`6RSJR8ndmV2PKe%22XLf)%
z+~_>kxHlyCPVm_I_A4yU@=b0fh%7WG4$YD92mS+7K4D%SKQ&51MUFn0GILzwcQ|QH
zQ;J8Yr$1R4d-OTPSjb|6kV5xuvP(ABfJ2e<1T6p6TZi`2r(J97sj~C0DXR^Nb+e}|
zZ)0R=W{&jKuFAK*A7T%hRL60@4@)FzYw``$qN_Gq=UrA?963ls)B+q4vKVk+?epa*
zGOZch0h<e`<I=Y_mn)`TpL;mJ65>^XmC=|&b>anw9KKOL47pFccTQAv-TT2W?ETJT
zxmyl&ANPGqcYgDFE_U0M&+|E>%dRJ%Yyp_i)z2KLAL%nx@x}8C^P9d`-X#K)d}|8s
zNWbg3wCO`(>B)+>-%1{SinY&C5Xdg4WxUsm6ie(97~UUS!Y#i%3AN1#h}a+}(d|t4
zA@Z*`D4C53T1u0ErNw>+8?wxS2pN;cNGZJBm2N8d+fBhoqz&T&)yFG${)pG6S_zir
z-ZU?3{`C2F*XDA{n<J-}S^u3__yhD;+`<>{*Y3p97N^ncS<<HDJ!b&&s^gLObc51S
zu{eJahd+g~lCHF>+wv?)`7!864-&P(M0yiSy9NqGTp6hb`ismB(&5b3UZO!mZi$*J
z*6t*}6oz}Xov+3XR|dY<Z>napg?X99edYNajHoiv!G}LLs5z~h($(q~gLq}XkeA>M
zE^Hgd^392c1&-E`xj}YlLrOlkgD13FfhTpEj_KegfKZ+#Kb71w+lPB6Fyl??-=t*2
z-a<k1TdSCW8Tb)^1q&wM-M+wT^b&3{uiEAKw9w)4aK301>o0)fYy$`LggtcY+Ys@g
z5Z-svqoDkjp5ZgBhDklWZ71cHlj>kxlF4nd`fF*+Z?ber&vT=K!!(Fk!ArjUXHIWo
ztmEu>wur^z4*Z+8v`b)|WUFQ9y|`k$;s;l9jcBVpZ6dd*z@V`bbUp}&DJNQJr1wjr
z1in{FXxg=64`Ow8G>*%K5PITGrB^2(B^pO9Bwki7t<pS^8f~0V@wr0cLC@?Dkj;_|
z=EP2VW?V1tLWde<jQ*;GNpoHeemf+!`n=0qhWs1jF@F)_t1Jb%PI=}c2%I(rI(ZbC
z;)d$eAPDMz>@nGsvkhmZRMUl%OS(Hy5@?s>lhVm_dd+sRGyc)sLnu=DKWIvQhf-GU
zNE#)M&yFB}Z4)8IZhHCUQ8>A`|0nRhe7n>A!E-qiLeTOOoK6{a-<q}xIGrQn@IJpV
z8EvL8?Zr2oVLx5i{<fh{{0Y#yU*7Mg083E2)A^JCgdRun#hn1>R*<k6DBy^P6Q0()
zmU6u8?}KMhdEL+`AxjJMk7sjFlevB3{7i=HG<%yFY@RQ-(XJP(7j|HYP`Zp5aU-ms
z5q2amRG4Gv2W@uQc<^nT84sp4>)Na%W~JMbOI;qrs8wr1Sf6+5#0PeqrDiiEgw$p3
z`H-}__4})xA4o^I79)pnq`!=E*Ja2oz1;VYvwc5bIQ$C^uE%0<8YxktI{A$b!&EF+
zsd1=2^_LfoC9*HsiR)Tt4okLLaCoX-Bxnep82b^M4lN%wYkW2E?Vsxcp|jfDkX>t#
z%%8Lfa%ts4;M<-y9G5nlHl5%M-`RPLxHd3k9Lx!Qk7^vUx6<OrO`EF1gVAo4E^i73
zhaJ7b(X3oBP3%_Ms!Q`^+YmS<_?hxNb&|jB4U&9b>r+;|vL;J(o&@6r!~Y02&#uv|
zrv9nGZ_r!5an?$!U5^%h0Fy*OkAHTn>@VicKXH<z+@(vQ9qwjAFjSFKqGJ53jc>Ee
zD=Uc+5qU3Ai~;wX`Nh5Ra}QSlshYyecd!<N8)4&T_BQEiCp<U_R05yom^!G_q$xG$
zR7;i+JIf9xRX-=__cdYy(fo-T{hmB6+g6idyugQOlJS~UMZ_qE34CZ*Mp)Ts+|Yh+
zmD&R~u65!;AqVd34-3HGsON<{=0j@bXo+99=5!Ced{{jHd*`bl3`@pZS_W}OvN#`o
z^teoxt0y`gnseGvXfs)m6smp(SQD*g`*~~#{)z5VVO7y%Sk*?B_p3&%o3dB){M6r~
zs2i_bWky8PNv&A+9gTk=RI%L3?2b%3S*Bo8X?P1pS#(BAHPk-4uI#pHG3^F@hx84Z
zPzcSFJ5C1HxVc|!-cXNS7zAmhZf=u;UU1&@p{#b5dxI81NlD=E1tw9#qO(v$h=vAV
z1!ab%fGev4x_xJC0B(?Rz1od2_x6%a=S@)%945|NN&}&vCanr*)dY9^#f@m$`nqDJ
zfF7bSz0jjG83|tL3MHn~DvQPgoX0IoDleofDw2*BwnsOS^wg2G@_4)CDd%*G$=!;V
zzb*7Pi9Lj5vyazEIDSn|``&)!=;-E$qI5tE@N;Jzc>Jm}j_p`z0uSrn{vZOnbp<^F
z*dT$o_d8SyAAa}S`IEU~U*P-x2%x0Tdi`8R1b|iiCeMj9wyTc2%V#xR_h*0Ts*APg
zZv$=fE~l;!Y=Jgidp*7a*W<@#MKW33u9xlT1l518{RIUTJ}40=)gX%bUUCckz9KqO
z<qiL-v)8E?GIPE69;*zl)fplA>QbVtb(n4S28Cx6+^su*yzY4)&%K?hlrL_xVFk)>
zx_v|fPo3p^>Ap?@IPlC~48SOebBIZ)+i=-^z7r(<d%1FYJ|cWl)AN3C#qg*Z)NQug
zS*|fEXoSP0Uw>lQZcD#kHLgCV*b~E~RX4VD*!CSyAROaev?OUj!SA%~%JlubbM!Bn
zxZEY6qTONjb@7{nIyGu{BOT@+KFG%|m6o1HsJEaGmi{wOf?{)VOX*S^vY$c1rmwN@
zHoE0r@$*`l?k@#oL#Wy{3h&-(rcJb35$yZ9`EsBnj|=NLLU3M>{5<A&s?P2>vN7ll
z{^Qh{sqJ6-u-wXU^3Ic~e9pBWE?Y$!1oN8^C!0m+pd~f<3?z&=z2aU-hDFIsHi)9h
z#^J0EiY*;!Ci0V7iik=!VElG&;w5<)VGs?B0wx;=vE^W2K(Gm~A3-J3%h82wlb%^$
zBd)IXU!F^?+sgmv0$9bWQ#S0Y?5NC_TYu}^+$V>d3<x=2oui}y4UrD8dY4yzgv9qf
zze(2h`n^|V-HfR8@BKGhl!b%Gd}A6ILb1eLUQ$FN(&R-6cbjhXoD0zzKCx#0cgMko
z9gK0_Xf0aEkTD!siN4qr6FDR34)Y%jgRp+yjK=SB6y-jqXdU*jWnzk6z^xYPX`<a_
zStSJB=6Ts}381tI`H8&UL>f<gTAr<nsoEpmr=kN|*oo;4erKgHDhZ5c)f#d}o{jST
z5W@R9$cBzRg7-32eo75DXaf#+5^xx>aq-f%ujedNx*KO@=Co4wss=&A!n_*P9GKE{
z;}VVYejV;HxtUF$2~t6%lM)B(R3+jrMaH}g7SOOvOyW*xH{lK5_d5YqnLJyJ*B(XS
zsIDSlVo6EUcuUFzbz{b@9k&_^Q^Ra$zpW&8)=fLjjeO^C2O~IMb>@bsOgn`{!U}KW
z3Z;+6eRvW_HG}c}wl-s2Vi^MP`kP5J_a^QiBuyNiJ+T8Z6^B8TDcO#kzZkdeiY@zs
zCkTB0>E9Mm$wgv2A59y0SGQvC$<f}QFM0!4qj)41*FnRGGzMMOmTvX117HMtvnP<n
z|8drGcO;j|@BMfVWUNf$3;4Wm`-||aY}8pk0bofxzxS)-6mR(KB-O4{MSE3IBTxm-
zULU$XlC&<{jy5TR1wMOIjws~Pcr6brp0?84%x~g;;c!^5H@gE;>0b!R5^)c-t(?qE
z{Yt4`x$FCOBW2mE9rCRB(xleE5ZX36OfOo=q?0C6KPLa!?*V0IH%iO4BhOOow^KWo
z(ZV-3&$UhxN-496Rp(sEh|kqiFDql-C|$Q|K2P5Gv1|d~EYZK1YM&v5Pp(@&Yh7~L
z_6*S-&E?gq=|dNn3wvQbO&hWZw24}qS$CK7RCAO~_}RqiQx6lR?GG~&;Wk|diw$Rs
zN?*4(g>!VN;y$O7Qv5M<cjfLf%3-Ijt=(H^H`KZV#U-HGdUw$E02ON9@0JcXG^m$}
z4oytTsYf1S!2}*cxYMubl}%6Z?k!Mv;VAJ#cYtD~OhT1R>cI^Z+>c{+4?f6<zN!D?
z4k~~s`I6<{%|uF-R3i~BTkT<_Oz#m=S+5&&IzV%nVEVl$Mh_190iAgMPjYdPQLH%_
zMk#duD>%_YF_nd2oVSgXOKS%uuS&n`#Vt%OgU>bCC+-UY>Lq!~T?OVDMi?(9#Pz#i
z@M~rp(*THZ^cxruS#Yel_WO97NRtzLo(7S`_k_l1DAMDW>U<7^LJjrlL`bGo#l)>W
zJCCgS;dV7wk`*ly&cQX1&EFEP&r(oW#l6LJbOWV78`@cduXO7q>Dh&&mPOcn_Zc?Q
zq_dko`uiU{O+fk-?+xn7Ms%sNqJ6z5INqY`9=><fpBYQ|8QBev&L$XqX1i4bhbPMa
z(tq!&<(p9^)qcUvA&`X$XA_+qPk3SHK0qh0PT@8H&%h`e7((KevI@!r#%S2wPRXLj
z;Vd?=1wVi|h|q4;xgBdvhbwmW1W6l*Jg!{&NvONI>HW*L>ti!UC~)S+MsUk#uK1Tk
zhRnPH^0WV(hW!^~-Y0dE7E(M5WO>Bsgk}yU>kivQa)qUzfeIwwE2bWHkan(l1c?}A
z5>=eto_ZM>{E$@+arhY&*lE>p;S@mxg%VPhb%RwAd+n?nE9gkyZccKG>F_lB&0#Fu
zwA3nfK=#ar+i2_GrWN&$X8TRaaq&qY9!!eX-{xYo&J{SBEncA_MGJ=@v~jC38KC&i
zN{pee@&^?=d?B=YY0(oU+EcgIh=~52)^~154$W!`)n`jxS1L=a=c^({WuxaH4$71$
zIBs+ahTh(T-)l2Tj>0ZvY{emD;f<3Brv8B)I<^pRZDdYhPPdJ^-+6y-11amt^g8T!
zks4p_pFy2!I?j9D{^H4QI=`%wS7nzTFI>L>jxKO;58K&155myld0Z`e^rfZ8q6I$P
z4#zbpjULCU&8D(G9-6lPtK>kuPvpQZz&6rqbGZN(AVsoS+=nM-;@=5;j<$1wzh?#&
z$b?TG1l)OgEBQmfD5JMg?%Fh|)9O-owsH5<X{F9H+9vZX<<mtGoT`iQ@Z}Yc$C)1F
zAD7}=v@QjoRq^hc**>o=zujNf%x=t(X%{J)kPnd<&pO>R^M_Ft(#4K~+tb8yUH6xR
zq(cmcBrkf7{YRd`Xm7}n^RMSQInzzekg+kYzt^%vF(@O4;A?BkP8)upIu<-ee+_vd
zt~2`v?|)N1pQ{)vukc`n4WbAMjv;R-Kv{M3`^u|+#1(L_ciT-C*Bm(3<u7P@q>`3v
zZUrC+Zjv`0?1p6&qoa5G&5=kkTP2*ry#^5dF~08zxR@1=m1Mm{G!Mg0XfK<ii@ia3
zHbNPnNLmQ9`^KAr0wKgNf1ok&LNZZ?2;)<#k&;XrjaCP4(=XHx-}Y#;-j3cx;*bx7
zVYE~gxlCrb256FEhv(PbR`EqLd>##G*iIH{AG=i))GJo(VMGz$V(09zw<*=+?e?Lb
z_*^|+m68Ixd`_&ev=)mG)m#E~7B5{DsBd>&50Z5oR4?F~jA#npLUnpr4)E$WWrHNG
zmU4T~_IqNBBsBbVxI$TG8V^R(0Y?XLy96|gfqG)|#Bc8&%)AWmLhpP-b(9u7AbpW*
zr#Pk^T%KD#ZF6*%xh1C|46CCHmdSLM2VM2=D*a{uwjHFBhBe}d>M|o3j<LMC^*!&Z
zgKZYT%4uHmn<_hZvf3_J8oi-)|1|@rHd@($*vg?9NWU-%Yv_92UHex?ZpNC$`00qI
zVMmB&Q2SA>-eB5~N;3UYE0J84FTocSm@%Wk5YG^^(mQyUtr0%?u+bYLXKHi@$z6s1
zFpd#6Okv7CEos<)k~)GR8Ek#MkBQ|Lkq^3Y3t12J;x_z(_;z#>I|#<PrN&Tk?F1g%
zJwS;XtG)~lE_QUEW%S2ZK(`#jGG)HTXMlo(v=L$+lQ-*VSLK=Px9ACo2c<Wu9iYSv
zGbu_8=`t;VCQE{5<wT83t7{3+IA%D7s_FTY7=jxPwmpJDYgWrOY1zkWL0D@!R;^Mq
zUt4D|^-0c?R<2GBC(iWz?r)am!HpJ~hvMp1iYvsi$O?x&j6y|7-oMH@e$m?2)<J<E
z&<sAT93Vq)`=Uy#)7ma=U98LJ;bF2sGNRRJqt|Xe0_@BOJ6T<x^MM3v{M9ZwfiN~a
zkriFFdt9e8Ym#6A{>y_J+sTXuwD$!Z>$j(|+PD$pE?$TI=NkgKjOUl5cnTmvunC9}
zyZ}!18xSQAkdK8mmu;FjfClsW<TW~(?hQz!(x~pbS^cA?hjq8EM3U`wpBqB(`ULI@
zkH>zzxMtgOQ>10neiD2Jg@9{!0XXCQrH9^PI<6uTxoSJo>m%?!Z!sHce}0D}A@JRu
zBoe7U6Y$eK%~`Wae;K5WCRG9j+gII#rv)tA_1^Pz&uk?6=#BYAU%uuogA<N!#ZXvu
zon8n6)`2Hgqs8tMeA?w7tT7rwz~>Y^!F71Ub7pIJSZh?e*gtr=^JT0XnEv`L>p>R<
zZu}{66xe<7z*8)nI*2RzD)q?0m1%R0QnGylNtzQxk_(eSrRt1Wr=@rfm1a2RXEz33
zQpmB4?Utg2d#$8(m?uKTKvR?fpWdjBzm1ejFCd6n8Fn&Pw~{Ff;GWc-Il4-GC{ZBQ
z%t$9aUv$4z-)*Fd+_FUDmK9=(3GHg%=ch@U`E-SrvN^S|9JQ{Gmdzcls<&E<Pi{3>
zw&K>nx-xBk&lS-PH&VOr>_=9DST1_q?u5Xdxn5#K|E+>eVG11`UWj8pg6q_K@()xf
zg7;Tzh_I0MQjm<(E>IBvwA-(l??slbVopG|p6|0-0Y1U2jjyuuQak}0-a7C)1qE;<
zcl;<#FiBqR>%_s2ou-*~q<SqdEUN;kt8QBA=hL9$Ex<dUn&okSSOKqR%_9T(#m}o{
zqB}R#*Y%%pu@=U=IUt!$IGhFFxaxZ91fUcJ5^)2*fPji;&P9vp3UhLw2Lvx&bE)U8
zBvKlo5;^))(jo8(d}}T5zCSnB!~yHTuNVY|_|)s?VnkttS<Mo9-LjzOJSyV&I+Lqy
z6Jq2;Nz)(86eW5_U<2hbg3Z5#e{d`zi#2OPlQah#8jPhyHtD~2%r9RUIv&>-Pf4Uo
z5;JYJwX##|s1+zJ9GA9+2tAQ6hN*$&3d5|S7T1ChZnU9`c1fIEj40A6kQYiDPRVQ+
ze3PjV?>Ixf=ZOi=gB{FK!gMvF4vXU&A8+>q-!|QAIDUwUDy0wh^Ulx9G|NKeDSarw
zbIE3K1+`6I1N`B=ut-=1I4-21*KqZK`>EE9F?2e6hAhmN!Mg#J2AyVbjM^4r6G4Na
zs<UiPe6epOyw3p54U=BG|3y@IXZe%-?ezgiK)2Cw@|<f=_s3WNp4|J{rKe5jPEpo|
zr#^%Y8i3tmPZwX|HLbf$WxDdcPWLfHxcj`$|GpY`_xxOMf#ZL-dl+5N;;`Lls5AKL
z6mQ$OwBt0^fQJGr%PR}Zqb?(+39D$%p!Zaos|pUjSUX#)(CygHhXhVP7jOvx7?u}c
z=kHx@^XsUm?&-vP!3a?R;N<Z{MzTM58WX7vJiDLN|3zPqe^KPh8pkpsAn$n8d}^h?
z>UiU4Cof`l%KU3s`o4l0A{FB{bKr~Mpm$$L5@e}bSyf)@4TxtcKSo4%r*lQ#Vf?i_
z%^#!Lu~_NaX7}&y5ctAI9a-W^dw~frP9<9}(9Xv0r-%y00zphWe_On_;dWaV$R9_<
z@T8IP_W_PoT#rXKPir>{Iz*<VmSe`)5Q)c4b4RaMpSi>h&j@@r>?up~XUO-y{UgI7
ziKp>&s3;)uHd!JGJ^M675**ij8}Sbs+jf%itoNUyZ9y0Z_<&FAQ8Q_a9AoYNbPEn$
zYQuq%qsg}%Y3rl6%N9(_<(Yp+A5Xx*!;RN)>}1m^$Or^{F#tFp|KqSxLagD?<c1q9
z)!%uS!4En;-u3t|g}9BF@A``3?%M`ZhcC)F)(CQJBRc^*_-<0$k*x2mE9dkBh!yIc
zAis15&!Zcdf;8$jTOE%X>#ia8koX)sGh>CK|3Y=zN6oFxMj2&@*-6ItU4-1D{`I^9
z@j;bg>K^m_>TpNoyWXxP8w1x|mg{HRF8ID4kFu^S7Bduudz=V5TIO-_-Cp_Rks6<A
zlGlNL<g9K5m*MM(Y|M;Ef^8MIOmENY{vvbd7s5**kyZ3wdf6Iu;{Zo4MoFyczk%g*
zd7H?+1qt5L-5;NCk84}rwwIYAA-SjqUzV>jsk{RRH;QI@V`Hr4kBY>Q@487$Q-jx%
zbtiG$%7d_UG%=|spzo;+C+J?4$r4m69ZfZ6O_u=RjQ1^C=|ol6vaIf=)|?VG^7Y)d
zb}IWzh&uitg<8_sWIjSG?xY@hDRm-TA~qn)c(aowXjE9#SdQg~M46gpv|!JgARR(W
z>ACQ(X}qk5wIcBL)opo&hn=yX$u;I@N~m=9YaC3A8?+ZB$5#7E?W3tr?o(&%mkA74
z=tRGrRhLsSni|f?&}UB`7X8YkAE7RV9@8(k*i()^ZpIXVyzD~@o1hB_#cBmDWNS<R
zf;G8~u)B977Gvr=2qWXyAxNS4Ie|8xtn!uR{kCMgTyK0++jP{<XL6eWgkQ2-?8kp~
zK79g_UWn%cIZx<(R7WhqL%vGKGlgf}pG%uuPRn(20z5{8XJ`B`^OyRJw1;qFMH@&l
zj-6~!6m#VA1yHf-h6AH29o5HNmWQ;=!MsU1aHu}M=d(%GpL*aeQ`rJt`JJB+Cv(wi
zi*j(kS4)sdRB82Og>1&gjD~y|-xwyLlQR1BQ02kYMCV%8byqN9AVf4dfnb81;C}+!
zvCHUnFHAK_QE2=pQ5#e@<r`R-w(A7a)AWjEX?6Z?;zpB1ok4%>t0T(GOEbAE(dQLh
zZwiN*CJ*97Gx`Fu45E1EFYP>y#crUTl(R&8zWBQiRj{1&j*$uR1Zf2p-7fpn1z}@Q
zFVS}^pt4)mdfNk7r&2#;r!F@penQ9A`Cj(~)G$P6J{(%<PeGj{%?cwm+yAQG?5I8$
zGtnBZ&9)ULW`h<fVrQ<!iQ~$St*M)tG=VNM-xhv?6hXVSZ#^i>K?^aSX-4TRZffvF
z{j+Gf^H%`Nfs-#u0DGSHm<~lk!$yh;o(L2cgpx~K#j5HwUU&@k1w#oDWxDs3U$pev
zv<%1n<PMrj8p%DGYRzf=uPs+b7GPcrAgjQ{!}nS`W4g9n(`yu?E6Zs+GoNIzjKZ0*
z22bo(>eX}UWV8yFtp{h0oIem#N`&<N+hc!UCjPfwATNELm|m~~dMS^Z#y<yx$)|$7
zZ9uS&twQ5&#Bqw>RjldquZ;Qpu!2_OBFohJpm<%1-8||*HS?ZMz8!tVuq-PzN24fe
zsbNc@JnfB9a3r`r+&4=io~GHlur{V3#TNIARyp1{!}f5Q?@So*Dk@TYr7{Ry>}u5-
zIAQ(z!~w#9`AZ$77HG^_%1X!<mRv7C+^$!CnK|!@*LIA?a$}9a3Q8n}^aS0qrNrj$
z7;vT2;a*_%aDqH2Izp#>DaMrPKw-GXM`yK|@?r7UIj!85dp=ZEhw$ot5}Zq!xLf0?
zGTr*bX1x%2sjDfieWmdm_gFa5WQ`g-I`NB<r?aptO82J6Lx-#Iz;P@@wvX?J9~^=k
z!zqbRx_i-N$?V1aScge{Y0b2{OR5yFBg7}Z?JahEkztqXTtoV~xWC-k@i`~QQK4Js
zGO^kKI9a8!0E~Fs@icbl*L|t@Vwp7F%fuTaJ#3!G!kX%KUKc=YC7H#!voI1*E|<;z
zF`;2Q@wMFtfI`+QtrvR-GyoQ>BKk+RN+8qPY)|B(IR|sjDWHOgQZ#5@iBhGXcm?m0
zbe<t+@8#wQJ(LU}%v*lL|B?-nhHGIkBz*yvDek~~3G|I|Yj>{rSdo0Lm$*m9<2x3H
z!73!&2kf}$@a6^!&{4mvx+K23Ud9#QjS9y|?sAh0rH7ag7g6dogiSWSY(bAWJ;UX=
zlBNRM6W}mtdqq+L)E+`JzG!zkq7c`{v{{aXBZ)CW#MAp|STIcM00EoI19XpIVC_wV
z9VPCA8@g#3QOozSXeigl?@JPV1Kx*`^6EUz9X=FU0VIqYKieKzHyw{QysJA+#}+R>
z2@}c?Ao{`t{uv&ewuNjw09DEnEl(=|&8eggS=jNuod4|A0AC8;Tre}|wBQ@A^*r5S
z^tXkOcC;7Wn2$hBQ3cJ8En+>09uy}x*rp*G^G^Hb4p8`?IYGjCz|&xSB9dH+!_JGl
zqi18}#R{wtI9z?C4P(USTe(o&t+9GOuBibFWT$B)0XSa|6+LL!Xeu8i3vSBGN{W^k
zOOe|1a>0{9gmH{wt-K0S`5}aH4&+XSX3aO`G(6b2v;DvP&nIwNpNl(eJWh`6o%=8V
zI~`8Kp1*x;z)f({&I>N%h`Sf)kKYzTs}YXZH0w<Pgo<)CgZDzAwi%~Ro;Kd5ylXF?
z9*yV*$0527<45njg);%NH=|hRFUo_=-4;qncnFV+DpTs)MsO*>9AOct;hU`29t+j6
zZ>C<vveM*w<~CfzC*3VU4soEZ$z|-}+g_%BeC_nX%(%Va1{;ynP3|DikNfaA1F|-z
zVafgVc@tMU%ou)xZbt|=+3hw-+SC5js`*a4KB%l!hU5GrlHv-RyIvyM23S%S+ob9g
zF>CB+%7RfA8nAFtV};3Nu^u$2QN&Nm^+%*kHlk|8iy+Z~J)G_$kP2GE#hG~}+WC?u
zj9%;H3@UnGn^eimnACz~!DE1~1CF*IRh@@vn;jh1s{uIqP|0qc`a391tQ0*A-`QaX
z;|H&DH1;KlQ-83#!HgRA(e#61VXk({pIAk)&h6h>u^2jz3GIc%O&%cYkj+LQwar|{
zGYHyWIkA5|&5-H!k^Nb(U3$9)NAz)TJ{@+Gi2vAu#?N4Phi!0@i04C&)<Yu^rOq7q
zlA!Vr)XlK@apP%81Otf(L1c>M3C7hk@*lJY1X!1b4ogd^zCiQy)7InKxTiFV_9iJE
z({sYf<(o8fb}KBHxu3?NLm87jDNQA~gl~hA!xSj9oo0!&0t8KDERRflrD%c~4TcF5
zgA_r>So)vEV>IPH{R<J)lBJ(H5@4(dCWQ<?OwtaN!aXBWpZ}B+oTk2k#Z&yMEH4#9
zkMs6{S2CVTf7*d2pkG$q@*6z0zr`=YH!@PuSR6Mr63;9#>9X;dwk+_sTq5{>*AdRI
z=Zq!il|AE8!swX*S{0NZhMgnhAIUxhpoJ;}*L`c3kE&v+RV@1<{+TmJ5@U>OQni_2
zQ~Pu*sVoqRoqY<~lV0KoPtav=v`EO{M;z`?x__tVLP^(q^s!!iC?b5F2ZqWcAaM78
zrqoW!9l<fS0acm~x2OHSjA+b{E6gf`#}x>plV)Tkvo%=RgM15O(hsY<aH~+TYbJ#q
z<XIP7u=$Y=GIMJ<!@XZnrZ*=Xi|}1k;A&p+>7uDiLLyzxSvtbpaW*_)D#42=K7{oO
zfu;V($n5O^O2}B4b_AR0B8gQbf_q7P#^m4Pc5%aRfBhL6asq2%K(Sh6Nw1n#B>1>a
zf$bcIJqmV+1R#VedXUeWJXq@qOyoc>{b}N`e`o5jJ!~>*XvG>=C}cp*WaE;uq%S^G
z?s!u{WDfsB#3c|TN>^Ze)gQd7?)%EHS}XiaKxFiwIT09A38$m@(dfh!=1=@;0P9HB
zxG`gaKS%->hcI8%c-m}IjhWoPeP6cvSTMORPp?Tkf+LH{^%q^`&R`FEb-Ni+E@}xn
zLPr!cWHD{`vF`d5qWdF0uq$=QAjq4@<auFF=-O#6B(FCA3kr*L&zYDgod(O7-XSP&
zjt<_}sTmt4=5HEq_&l~iKPe3eQDrhT$7S8*KH}D>8*?NTe(g&@k(jDiJBnQL=wUQ&
z$9`c21NuHcAn=U@hjuA1A2V@8b10(k`nu$y-t&;~_69&vxoD5DQ}-JBu9rui+bu?_
zHeTjQ5y!K{1J(}K8iAaa4(cI+MPq~}cAM43@*#A-HrF9S{|scZ#GATFPE`>3g?(0T
zvSF%qd_rmpEy|UhTSH!v4dLW&^1|oTCcFGXOuQM+y-yj?+4h65VK1X5@(`{43fr3B
z+)zp0-=7UKVucw75lRM_Pw}1WSbcMBl7yg3aO}=Ud?4d6dnpw+2&Pj8#vzY_V8Ch=
zNA!)65sn^57vvL~>;r8#HM>kFD2}Cw`nC=pUK#+5rU6(_M=qJIq-D@2|LIkEyl6@v
zxY^yoA8AG4y*#&bFCX<Y8T{8wCMHK6s%VPLXR13dLufb-2ou2wvN^c;KK+XNScavu
zId>Y`f;XtL+AEiQYl~KklF$ehGiT_MrG+73CU#x9{dOTjmNQYxs7l(@dHk9-i2Eb=
z5v(>kQOSDgbRL;jY^b2`759@ih7sb<!$naSz{Xzg1kM58(p8q)rTMamh0!QlU(0ZG
zr9$)=iDL>fVGbV*JNBNS#>3s0!H<v}3^90a?HK*SAY$SXQFTOuood}WoBhABDQoYg
z3i*spExKpwXn;190OeMU@?4pArb@po{RTz5h|J|;%PLEODSbXh#-!i>SMG8I=+VBE
z((BWS<#%JU#PpF7<JOh~LNP+C#g%1*2qkVw>7!dLQmi&6Ee8)m5O<n%2FBQLE6^jR
zf6P{dWA}XaqanX6FvjTyPwb@W^;k)c#r=oG^$75_^zA)>bu8Re$MjoRR;M6jJQsct
z1j>)NCgyD;h8Vhp3`r)et6Wm`^yDlc4>xWFo}1c+Z92v~#ep8Wtst4E&V|_s{>Heh
z<v_o$itq0Rc2oK{QVs6xF1jRrUz*6avf@A2ixtozfoGlaj3#jsa8OWb>R@84O%#P{
znym725H=%2;|J2lyB@@e2bdb>zqz_Yi)IA=KI#ddmJ*fn_p!6GHsR`Tl@%oP4I;pl
znRTcQIIyst94q9XSy&XyWQAqF_J?LOV2rcQIhGPW0r+6zgfZY|W5|93{Oo+W&ecyN
z+jFbu`|@y<&9D%k&G*LhD81)f5`%&<t$1R^n)cn?;9Ew?y|0H48s$+$+wZ%^w=d@_
z;NN}Yl+YcW;pB;KM2%hAbDJ=7Gf2qiW5b;h)kE7WM0A-Me(LOII?5MJ5E9LTT!F&1
z8QLv9T!EpFeyf38oIl{G3-az1kPcGiBUUnq6@#A2FtzhY4Z9zCrv1d!h4JX7BJJBe
zRm>FKvaiItihC#UTw?DVU&zk-lh+@&@VmUDLUP?vJImn;Q@Lto9gmK`J|_~`X-}nA
z)6@U0MP05o<Q{Bw0?WgM7k?o%5vKAg(i<j##<8H|ZfAImjVG#|`5yX)mQb<VZf;gO
z>SS-K0bOZktBK1uliKmIUrZ7&F9vJBWJSNkYc1}4(NpXGriWc$=U_ZS_kq#{&{aMs
zrGmIma_2T_K%2Iy-S}Sb1QbgMSi3eW;Bvku-O%BIMn-a_Ds2zyuTf?r{v7?rpH?}-
zv>(+&;8TxuhK!tHV|S{FZ5&xfT1}0D3c>2O12Yn1xDUfhZPqA12&Wa#Hh?`#^7aSm
z6X48A+T8JkeAlWq*5KB07Ukdv-Mh1)jipc0qTR4DC;VX>5gnHRLRCg<Yr(E2S@1_u
zPAVwKkPGWO{PA3qZ+2&ZU=VH<&%QqhZ7Kvm3xalmc&@Y&gb=&hwdCf1%k_&@xl|E5
z5@(2^i<l(B4WIw#0-#x_1gzEey5}mh1HSP3q+AS1j_xk-cF48$#$k8|QK*JDv<2HL
ze!{eRv@A#nJNe`xF%CVX3wnsU`uS9{QjFUqGd09qkZ}qvoJKq=!WtM3w7X2T0!iV8
z4xHFsdy8wW&Q<9GfOc#Wd3QvnX5Q+MvIy!BR*czp>-ll_Q&|qs$gx|5p8Z~Xu1b}+
zvKi>BtbnGS&m29pUwM(CpOSH8WJbvQp<oA{XknxM9pud|?pvt6ANT9=%(4?XK{&+?
zU9d3UIdPVZk~rny;9z{ABHW_`P%>XL;lbbZ+AIz%3fQw`^!k;}S@u1@5pl2`JZ<^v
z(I=M<hJ1c>4-aKa7Aq>h6}CkT#3Ll5&G)8Seb!BNJ_C%Gu8&8Y)nkTeT$9go@t`73
zaeMwEu}*sy%naJMm*$`=*;o|;0zwTi2Wg1%=QT^r^b17aSkslU1a@hk+9lo0`|7r<
zmG%KJ9LYFp;Wo0}B-fS>;?U$InX5_(4zg}J0>O=L40)+u32U^@`gc2psiL2wtb8h{
zle?6b*xKUO->q!`X5;3;nA*ef>;ph-yR^3I4QE16%Sft*#<wik8*GXI`{CAQJ~M0i
z_R!U6E6`D^){i+I1ApA{bSRt7ZnRO1gKm)Whu3ZJWL&fCr$GacR&19<oPvqlrx{(G
zj4XYHrWZX`C*q-Rz2QQdc)G}K*|rNpFxMdqpMx%mS$#q|Y_6+&&Ua_lr0umIk@2uT
zn8W@^a+`;uLG2)vLs&aqUysz*Oy=Cmtr-~rscky*eFnKNyV-sl9_T#jotw&#P#h|5
zp4LsF#}cBqp(x%+G#((Yqy28GG~nhvZCPI2(Fs|Ql>^!*Of6C`EClZ?yTzuye1~wu
ztujr4-80g>jN|Y}kGx9t?&e^T>!!Ew@^<dufv46_aDb{;q^SY5;$~N%xhyaLbgEy?
z%m$%DK8J2UInLkAlnPqp^xgiB>?}>(jg|~;Yyz0J#U2zch?CI>P>4~6%6HBGrtNhD
z(=C9+cL1D-$Z;9TTo_k05Mjwcvu-gGa8$=kbBs|>IX(tQLL1S2gXg#p!~db_ETiJu
zf^8j~;KAM9CAbIo#$7^i2<{HS-QC?axVuBJ;O+!#+<E)l`|fWBj2_M2YgNtq=B&DH
zj0N>^ocbGWRBww1HsfF$QNQ`!@P)bJiC!550hdl^iz;#Rgi8CF2sMRBD*r1af;95_
z!3)8cnx6Z1;9qGpB}ryzg$&8PX?($2piW`_9oZ%tS-ahNr59qUeMlCalp@staf+{k
zq(75P4$pOr5fg?b>Kdu=kA<0)XZ^yY_^31%`7#ho2ejveT<)~m;-6-Mq2Ho^%9uMY
z$jP}*9FQhv*^HSW?&CD~>dSY7`qV{F!w1nBG;6xwSN}Etgi{>15C}Ik;&?%%3^748
z)rb;QANkUSm2MOBo3=>;qIU$ma%8B-w;j&;0gQ3qXmbAo8Y>`lyYGHP*ze2;&!k<=
zX7f6QIO)Nsg|2e(u4iR@M=1^cwa5h3{8U4nxiR*6QtU3`aRjF2vzuq)05XI-aoHjM
zA98@LsPvUUrtS7WMSXtFV99&>mG5s5OOrkJ&6VweB}aaso+Yo`R{;!TvUsWPolxbO
zrxkPN=tZ>xC-B`hbtT>MUo(-W^SP_~3<J0Arqy3KN5<;@vstaZzJ)R;VGy%!A!3K>
zZs#YhYLud2l~$PVrS4Y0!@dcYvzJK}_zyI(O2kIxLVk|KVd*H*w4+wc^X&h=1nM7I
zhHG+HC;4!pf@3GcKHzYcyY9K{>H%iEyN}^&HWQmZCBu*<W{y@WH3++_XCgF;C{hgr
z2Y1$SI3n7AIvZkd3QZOYA0c-M;EVhXh8TDB0>U0?CFsrO9m*XH6e9~X6rf`6NrJuM
z#FQJc5)xf|d(II<SU0KFPk5i=z}B~Y{chr>-roE6g_BB{epdl-9j+IhSA`Nqp@yMh
z*Z~s5;fzwedU4c0C%>JhZE1ZRiCu<FT_C{@2>Hde2`n5Vs=*o8(=%Z%o=j1Y_q4bt
zNpDBxFhHmu=;a`|K2-Q3jN+*q8f~;AD+%0-w@kYLib|KK9(a7&L7{4#dm$Xd$fO?f
zH;Bkt*knx01v}!~QAF0S24z$$QHcsUNSiSm<m6UvV5Wz`vI6CD_B{m%A1i(Oau=?~
z?}4lO5rUivsIX})Nu#R#z9ynb;nQytfov!kIczkhT;$Staz*}p9noJRRuWgj1VYE0
zOQ3$ZREqK@Vjw5e;M6&vi~t|!*Cj3{N5HL#xeCyG9222A9PR~}$kdhQdbR`gFrYN-
zbjXuU?-K0kQn9yXfqg{A={`ku_N8}9C_kIvL+;1JKGJ{o>oy%ND*IG3X)ZpyN82BK
zZKwK{coLCe^~l^-q)Z{TxVRAG<H6W%jd_dyi|N$@XAV0~zBC#&VZ81EFW*9NIOpf5
zjN8>BU>G)AD9ORb71xE=O2DzACy@R<WXjyJ`V&8X=$YMZ9oS_2*bL~OK19@>{;RXV
zWDfwo*^qt)P?IvCY<bEM;p^jwSYc{{52-Q0e)F0tGyY1bFXT=^<WQtYdGk>y$7o=!
z@)>I~s^3jn$BqX-RxC}!iP?}bB_mT8g;?MgcmoOFuSaKVyWbDjF1v;A2f!qL&nEzm
zS*n=(ez0UXlf`@Ua@mtB{4o+N{P9}r#|)6iM<jl)4i~RaTR<S)WO`%I`~4cQBFXV|
z!BQpx|A7;butV*--!6_d4M~K&_y3q1{swlE(YNd)4moqdX%jw?(}T{`y#az&oHSvQ
zc~0`8+82bucZcxJ@XI;_`amr9=L?HU>G!{GjEuLK@-}DoG7YHoLQsVe{X6Z#0%gGx
z=t4szaJ+9(q@mB^SH)ZOSdb5T@8IzjyD9B%Z!A*z*d<h(t1le`8(lRMKe@NsJ*)*U
z!l6(3B{qtgv6}e87_L#4vu%ar1SFvP;-q(>4A8fo>NO#e9okLfoeeWLU@o>m;u?7k
zd^w8ZSM=u)nB;FRe@3N+AvW*OaNpBpA6;b*(Mr}<mR0Wqz7PhJ@nq$o1S!2JG{ld!
zHUG2|SwTzCT>J*g&-=cV19+9oxal;MVyH~t{V*D6<S}D-Hs%6@Bcb#zxp^a!Hi%gh
z5}1<C;!R3ccI{fr*Zf9qt!~KjlSl+Tvvde98$R=lDesD?3&RvCCY1Z${C)oDyxLV4
ztgWgnKVNOTNu-b@hl>FhRd&=JqjyZ&ma%0r#4YAW5q4(P?=$Uxh}Zpbcc%7IY0y+o
z=WRE_9Mx5AqMCOmB9#>;*My=r7gH!#`}_4~z6=_VuAwbZQyJ)iv&Rmj<P|5CO@7F;
z2qur=uy&JQMvQ`Hs}4xFCag-}<~ky&N^btzz}qUFWT}|&Ky?q{*ti~*#VZLtoH_9<
zr)}l7uQ}#Xj6Dsh@TM{h?~V;crm7%JF$Yu&!lm!;=De!3r6aM1psOB&@H&7cglGAG
z<HlJxi@n3AN~1@1T@re+woDN`sm|Hwdlc?UcVVI~C#6W3;3dd01^G45<vv++zhW2X
z+2GllBMh<ISESEkAIHjY$i}_c3<H6L2dDle$+{@4J&fgF1u2GS<gF1K;(<7fS5mde
z=@AVc@5d`3%z>E>vVAnJ!*}gI6BY&dHh-?#brk*UZXNG=KDE9!q^7&cUM6&Ue~#9D
zq_NT7adbP`X|?@~(yl{`FISXDDpKyfy=@LGOL%(Dw2OVSe+613rH|{0^hR@F%(b^`
zJNlut`VQC$ZjNShd7Tb#PYe|S#v*iYr03&(Ilj=)=l%(2iU|&JNQE8>3qEf&k>wan
zPfM%Mimh{{MSUEMfdqn65}MGK?QW6Q06-mk*$R#=M{!pa3F8VDcv0k?Fc#e1K?#EL
zd>O%_sw3WcVvYtVx(Qh<k<|3X|N4DyeLu083qfyRqc1lwU-zL5TWo>%)y}7#xV41o
zW8I*b9)$)>1Y5w<5)>$u_uGZmcuac=tVuv(vUuvkkALsk7B|L2pBL=HO(&kCN(MDN
zjaI_^BHkYVBb@|$1%h@&O!?E~Plz3&{x$WR;wMOe44n2@9Fs6^`(dGhC}HRcrW#y~
zRo6yP@6zX!>Ov+w%jRRzy=R+e9!pnFP*V(7u*ze}gZSR@?%}dM@GU^jOh&}EwqQiG
zZcOxx1Y#3$yqs<#2Vppj*muQ+<aFXgzo3wfmlDNnA^YAY+SONp9eQ_vlBGZ7O&)U`
z4ZV_yM^8A>D__^MEs1Sb<8GXAZwOqXhy60dMBe-I_HcNAI_q>at#6j09IaQvO>C>a
z7q>JK-ej~r#bP(DJ(EPI3NYro9gjQ%ji7Y8C5STw2>7SJlpfjmQ;=gIYf_9e7dZA<
z?P&Z=226)e2m0)Kq&(8lLL0yRv+4>;el@8js8Sac$jCTkM@d24?oV1XL9el=@&?$M
zy3mc~v$Q_kMl?GNaV?v3$p5(it9Cf8Fqw<ZD1Ar%L7>=UAfcvDCWD$mw7PR&k4OAy
zzxxX-2&5|e8?UzSW#_(sQl?BEu1d=fN~3ZRX*=+b=grka7tqf%Tn+nZd<u*=BE+JY
z+^0VcNaS<Cvdu=!+l84(9mg8R6cnMrS4z>Q`h-Z87blYOMbB?r2tMM~{v9qA1wvnb
z)F-dF7@(<utD0>m*#Q{a&oMpkS_ygGGmh{NIuIROWD(JKexFat+3<|=_AXixHp?6!
zT8JzMT)CJFW3(i|H4;x@==1Pa4?z9z01pWGX4oUTTXUVT{^$7z-Uomp{tq0yJ$Kvk
z1MDk(V!2pyQ$T<Wz8v)LI!%b!0&+=r=F!LtXEDm26-2RS4kZ^8I7)uO1k%I&$KTNv
z_K{Y!bwLQ!G6+e&092QUor2Z#z6jzqWT-;>{Az^O&^XVe+LybpFPq}ix&I{iVq@}Q
zM0>rEs03WbLcF^&a%d`8WLX#I<~(*|D>8Urs^gi;Y9JcA@%9|EiDd|pYhFwSIJ-Qz
zoZpbu(1bXNTr9<%BM=yACKG<}ZE8So2?{8ye_8}S-?rCRviF$$aOk%+cl9zteAhZJ
z(=vN%bfQ0^hmEYAG`ZgG;Db~HU#Gt|!0<4nKkZ{O3T^Q5B<`R2xYzy4yU*oY%a-RF
z(28deUKr6bj((<L5K^qbbYeF77BPgeYe)C@hg<uT8r|^&4O)S9h+FHUYvTN`^J>MG
zqTs?n;fa)hx3aS3Kcc_BTy<M+-(7aT5X({r85{kCn~xYMR+neFPNPyeITeYo;Yyxn
z_i2`<Ne4?a(cwlQwcj;wYToKSgu@8q<^A)HkJx3#K4}yx&h-fQZ%H?g;tjIW`#1^7
zwCl74Y*XSk2n$*Xj0!X~l|09)+2s@8OusXUYzG&FjiAS;$7?r8I|Yo{o4Ko}A$jl`
zsUFow7ScuKX~t&QS}r$*e&iECku>oB%h<`cmk(7keXVQLi+^%d@rq>5pLDPJ^xG=<
zbP}&P1^bM`)M_|-t$At7^=4{-!~?0CRc@Kxo<x=v`)>qQg1kG4Qpso-maLM%ih(-|
zU799|QZpfs_z9bVnPZz8AOA}=uP7@WO?u6xe6ta3#TWU-bLY-lC><v*!`{afz0W}&
z|6X5#VsPDi`zlrHzhk<E!yH6nVn`tuxmV9OCHnVKf7;|a=^S0ur=pu^E)n9|51vPA
z)Ne~@ji_alT4s~}&$ncK7V7=?>#Lm+V&6x<RKJ(np7#hEl}8c*Dpm{@2ND$g|8!8G
zo&(yvx&GN8gS(6E{FzwaySsXlm|Q-W=ZCEs^(Qv{i+5oMI5c))!1x0E!0+ykJJ-jz
zb$vZy406br6!-5xi4GW!E(!Xwd>o~(ocLDZH|+Fa>$AY3w+JLjz;UHfrPug<z)YwO
z^TN)pJ=y;dH3-P5C55INCq!eoc&Ql2y8!%D&+7@h_FkzhhMeDDD@nXtKaEQ<2#G6g
zgXy_ZI(PEnNOXD}5Isd2lWblelhBZK7Ai}hbw}*}Emr@^2R2PRdUZCaLkO#U011D!
zXPbLKRh$s*o&EtK+E;BvTPB*vJ_mJ%WIRkEP}}o-WfsIicDGszfrMQqWF(q3jquaH
zwR6(`Q|;-}uih0bWZrd>$N1>LbUNg@)qt&u*@sghrC8sR?@I;(uENmYG#^~^>vd@7
z>bywG5ghz*Z_>efAd)_u!N**iqY1#$_^2s&XAR+0!!8>;L8Xp0$pWycm;v+&4dmBC
zpm|Zx$jrN+R*tY=5fh=kCHqMR8x9zNO&U7@3kHn)4CXr?qAcgB@J0(N-wH>^3gcE+
z8Pleb!?!d@NvbA8c4t0Ck59RldM(-VO{X-z>*a^lA#*$?a>RA2)3+4O6(m=V1@H`W
zVunMsi%ol1j*Q@^S^t%@*Y%Q0{}F#^Tt4Xx1|R+Gs)TEI^3G9PXZ#x;1R0;LxG`7R
zkSVm(<@O<0TlqB@KbToHoiOnyO8gv+isd1#_T84>WtUy2P-1%|c{>|?>Ba7ZJhjqp
z%{Q4_hPc0Qfo-x6V?Tn;0<mLc8#3g9AqpJrN8L)C348R$2GA5R!V(~xqxVy3yVjHU
zd5&h`cey-yk6kZ)`htwo_5;Lh{+!R@F-j+Ef7MH)GC7ooe|R9<9Hs>scW<8S2L{gI
zYyppP+_K26^K=e#m-6$k4zJ>c*D4dDW(n{zt!RMQg{se&%6+d?)VRM2x|#)=gl_S(
z0~`#;kGmzqREy+P6g!`rsg~i7j<R4w3Y|NsdhCS>WvCr##>6S^7Q$DcqW}mM9+&6a
zr<)^+ovV-bjaKKES3iv#pmPBK1Kys;JtadZ>JLapr}8N57|&#c3f7pX5N~0a#y@@4
z^lAM}A(JgjmxKTU2T)#Gb3=uT><p)ET@@tsDWW>9IHy+->N)Nju7y)%G0YkM_))3W
zvKLl~^8-QZWi!vnnVzoy3^Y~$IeNYJ)lE<oHr^dFt#R2{!R!=q$=*Nx;z7u0AL_*P
z#`vSPjrssbwusliSFK&{9E834+k?Q#%It?av6?nGnB0L06HDG)gv0gIXAZGHT@6Y<
zW`E^t)6(-C4yZ&krw-0&+5YA=*;rG86hLC3`}rz20ig8*Fu-7Zw#Ux!ET(xS`(4Ck
z<<0{6wGGXK|I?l<`veL!{4uk|Xbgb*?ZA?A#}0C7<xM9*?`JHC#LID_A&F8#D)Fku
zQocM6+OXRrNS6H+6?#RvI<4MnJy<Cnxjl7|&{sIe((mX3AbB64;QdjoEjP!AMs#J8
z{yDjxzHkd`iya}@INc;`*GPI#t3d+wQh@YWjS}wLtsS`pAFC>!bVK2Dn!ZZN=XK=b
zT9#LfS%TUb%75rPD73xFhR6J=+)RDQCYH%$gGFYP6l&qKFc+V4!~)=guGLIe`*7v{
z<WFcZ(H-jvV+riE*Vb<2%eoF<4@YOhZTODlbVcJ1Z|izq3B)$=+x9baSyTU+_;}fT
zNx!R1NM|gS<;CNT)ikD59BHY`t8LzZj$=uYUXSU(gBkoBD&00^KwyPQk3+NjJ2cFB
zH(REJ-HbX)zN9M;hm+_D8RmmGKkl@A$?aB<*LeS|#Y`8r1t+ZT-a9;CW<Fgn1WP1e
zwFGN^<#mTbst4$3|LbKZ(>YSM;XD4e0JuPn;RjtJwQ8hPJ;$<);)a9<F$G*E8-f^W
zEiGOlQpCX_P^ud`?>okmJxCaCQK>%yt*_Iu{a7Q*!pR+sGuP~#c`nKOMaLyT2l`yB
zo0{KIy?MWXX!-(sJXK4TNcisufV#ekkO%sU;cU<@BI|(ZPuCWzPa;w>4t|JF2^_Kc
zW#IX{Oxy1F;iF-Jk&|C`pQ&r0Ug!I135hzJGz?_>ih-pNj;s0pKPRRE6mO|UH})&Q
z)$2k=@qd3E1a6p6VsCiiR^>QTKXn694&hQ|4y*j$C<Pk-KmE#!O9t-|gcITwCUCMf
z%{>DrqP1lMC!Z}q>>KPGT@~oY)k@Xck1qPef(L6`2Hy=1aA3`!T3kByVQl7+NoGMs
z*$0uU@{Mv1N_T+80GNl&v*i#XjP%EA{pX?GF%~Uo%I>&7MKO=u82I;_pDLSc29<0c
zQo!Dw6q{5!Q}<T%PV9uGFaVYwr@czd?n+#XbY<zycj~$_8m^+7^hOiMtv7JsJlMj?
z?cen}5hwC-$n0oLh#<yr&nj>)^w$0@zqyhMwcBi1boVaOA?;a;xM&5@bBKjK4W*Zd
zUln7=>_<?&gNfX@5Wtw?Zw2>y`DqNSF=xZdWk8h7?hW?}-0P%9>(}Hlwh?l{9gEkJ
ze|BX6+$La&UXC{wraZQ6YeSbLi}_WysNvBg`mG#$-(u){3lIRxdPaC6b9B8N2jR6S
znWr$--zjhIw44G#yKq|EFN?;{V$jIrWyodtr3`6v`WragZA+~FKueg4Rn8tO^1Q88
z%@4E7&deRig(|EEbxaub<Sir+sidFn#UhiuAKSO=7q=^?im;`yMEJQOP>uR2mf)@1
z6q%XSlIcdlV{8OlEgv#lK+jG|_sF~*i#9w(Bp8{%!PCq>u*Q!kjWYoOF9oUJg>FWI
z5f^g~$D-V1G{oYy>h*^EVCuyqZZ#7%U>ya4nH@EK=_=TidR~^tPAy^jFfn|P#(GJk
z!(H*?V?K^QeSG_)P|H35LMUMh3x)9Ytg^P}>0~gxvFBrYrEj8z`LefoIpDu`74Y;e
zqmBU1fZqqbeeztnQ#hRwTgN~8YMId~wZ8;;?D_*d3*bFsPzlyW{=G+N^?iY=NKCW6
z4G6bAMTHeprZrX5zLZ_>iAHHy(94qR9*M=5cFnmvJoCC}KUxe=bVlI{14~NXgZHSr
zKQ=o<^u6?nB`A$al$3vdW#9S@v92tFf-@8awYd9Kahlr{{Q60t@Er}?6#d;LFGx$B
zcg{7GKQCA;zLhkipTW(OQl<vdt}wtN6w9bHa+lSsDv>U)73z>lu@J>H!1m9@^?#=(
zz>h{^GacO8{pS5^x9t402Le#J_S}AKMi1_a*5%Y`kEe-&R+#$g>{I0eY2UaCGpV%m
zR-)wjob$90r5=awNo{-PitV^yY^AWpfQw<WsShgNKcaYVxa-A+!@N6u+YmUIH{3QY
z_caI12fgbB53sm1K7*Pnk+)94i@#MDY)Dt$!4c{}r&SMV6nX*rxlv~}N%MG;0;MRO
z)s=!`Df~EaW==3>DD<)vBMS<{>iO8Qls)XOSuALIo#T|2=_^_dMCVL8^we?`{95Cq
zE~va?mN~V{@=Ezy%$S)^UZz}JhK!<hzj^672-2p%J=IGf;;dy5n)#%ihj9pezFqzs
zejxpeIQh@}Jw8N&139#8z5XW`((jE9(lwy3KiKlj)Vu=#rWI(LY7DxD%ivmooc-7Q
zt4i}dGQK;ba?0UwcqEhbzjNRP1esvVvelu-c4bWKVs<UIz<hY#lq8Oo&iE=ee5FH#
zA+cPacOV@o=kaO3Epms@)+~7y*tiG(jUe|o$i#l)1X}*=6gG`NTAm?K)&m%&iMDJg
zYpQnGjLk*5jP}6Cy=2;Tw`o;A(cF2zY&@Wpj*!4psivnu!Sy^pXJQC{cq?xv`(4v9
z&)oU_gabi7DbXl;10^*as!z%kRzHvm)%?u?>~B@h-ath19u-({Ej>E=9`}XGoB3#I
zQ_PA=Fo}Z@W}NR#gWUY<I3+#TX0cAi=^0IDNfNPlo4Fyyt|U+mWV)O1n~Wi3(*M!h
zee|&VF@Ic|vpL^th$un1W1+i<Q-u&D62`xj8}2?DsH0^r1b6i}0Npr1Wc2vUQ05G#
zw@~%iP6+Oq5q!zt%5^rcUCG|Heu7{ryh2v*cO%dqnR?ef50P^;R>o%&k1HZ%F9T19
zGr%%)9KSCT7fO)pcM%L6`fM{XY^v6OW46AEgzIP_i%2FP<m#&Mp<&0Q?NoY#uVf#~
zzw${Cc1#ocid4yhuHd#8RTfCGY>au=cM@2|2o^myJ5fV{AO>GCz+*7nb~!VEG<Li#
zR}4xh{T`DwhJ#pa#$hpz7iKGXKS}{4G<g3uB5R*K(8XwEf-4C8KHO_5z>12S695f@
z+xv`*S_DmMS-GI9u53}nlsHDw;_;RE^3!_|pL)33MBlFzo?IP)Zbr-rURp}qcCyw|
zWsx{B!i9KzG&=qA^?sQVs+S+U_)1(W+4}r07c12-fEUrBqZln>&=J(#Y*q3Wr8qns
zI<;_?ZiCW8AXz5qBDn&s#vNv7%G)azzdXXAaq|1;SsAhsXKfLZ^l8x9u*_JBGj%La
z*W;$80A<C?Q90*chZd{nQ1tA=6^ax_%ewo`t>t{Q!~2*sKH27B%CAtNmt8>hj`?`p
z115-OcvPDKs?R`~JIp`>LtB^4BA?AZg}aD)cLjeDuErHx1=vF}p5{{ns?ahPqR9&$
z+`9lXmGABGEQ$c6)96<4`hS5OK&9Hr;f135^05z?R<=J9g|jED|Mj9|3BJ69V+jA(
z?g7bC<0dzRR;C>T3La_Yi6q_5tzCg?T0oG+pK_R(<2Z}g>iPNUi$h15_fJW82+#w|
zG$E-7z3QK^_TJU=^=i*B!ikofO2L3bq31Cd$8_=1`Cq4TepvzRrl<8Zkf0;rHO=qt
zsuG7qcHl0p{(fbi>rn%YHoCno#Rl|=sc<3JCmaSmlTUYd?#{O8^Y-4UOR-}@$k`MT
z1bwv_AEC}HJl6YY%eqO2)<RPdq+6L-)9??191qlAT6mG1SSj*+^6tow0t1iu$lzBj
z==j$SV(3P8ZlPT+E-dHGI`3EatJhpfgB28Gs~!OhnZDec8XR?u)7x0k|FHmz*0wk9
zoj@)9M?_K~^iS=Y>X$Z!yYB`O17`51G=cj=eqm)EP2=wFEt;Z}Kh?PRigi^Ok@+Ts
z4W0l((d8Wa)e3+i&7Oas7{EL(MhVRAe@~()olF0kB}E34;rQ$znXmiwjV>>^*=9xZ
zZt_rJW1`i(p7E<bkTJt9r~M6H01?}92>m)k17_^Xo-!#@7Zz}*Z9QzXc-c|7OkAT^
z^QKR`8bK7X!8mklDTeaL&9h;2Iw;Bh$-X()27fNgnZJ-J;ZsemhgM5Nl-dIvG4B+p
z!sI-x|Hb5ZoN*N9N`|Cm85AeaO}0V1wtSmS5S}c{SF@@?*fGDZPC5xYo+yj6qEl(=
zz8kkU8>aqYUd5DHO>6LJ^CYY3bBUHM9i|+4bB<njSm%XVz{TAi|BIfBOD(bH@|j{M
z{RVpcB#tbgrOgXNWykp5afauDWsSr47wA+0SmzSol}6pXKKVAEX$9I`KCeDWn1maj
zNzdeJXlcuPEufIc{J+N)QTVwGm^-=+Fi#Vv+UaiwOz>ZAtp)%Q7C_id?MTd<{o~aa
z)&Lq6G&{nE8tuh8>&pTOIo{->^7$h7d={{p6G2@^`rm)9t(2JjL<{}Iwe%j9v~q{w
zXK})uV>}|}TGI;NHUdfM*TE~AQ!A@n;<u<7LkP%+^H69XltSoo>@h$V#{U^ZbjSbU
zvLv|XcE(<DP4By{UpeV?KPs|mdp2qLWd<Z7J4(Onzx4-yRrV(bjS7@H&itWN^<`+M
z8xw9s-1eV?d)Hp>iN3Y{#^d@yGFp|NZa|JpiYUas7A5~i18L?!(;M6|UWF=YYwt<C
zx@2#UvhboY9@K7V#+oyAb3p{PgzEh48obFFnOtR%JGB5W5vDt;lvp5PB5Z9p9>}e=
z$f!(kDl_ZL*eJTWjd^p`R-NUoYGLj1YBlgujS~w}E2devwSs3=_44$7mdiJ56?imG
z{?^7mobn&(AZm>FlqQsJ&K!H<>O~69)>f2g3^4IxjHhoQtfJ3t^Wlk0;q;<>y5PS&
zTP?4tFtSAG_!e$?Ro&oL;v<<|vdM$}qEnrHPVJRUhTd6za%C!zZv3}uyeVH{eV+Ds
zf4^F$?$MSywPi50a;XY{V7qqO1BhiLp?>haw1-i_V?l`(R?vSg889=d!Fm<f$(5dw
zANIg|WSa^y;X87aE0m16mUiq0?fSXWg_W23?0NEu@s|^1gy9h@)j$yNCs;P~r=CU0
z#;59{*mJEPfPq*)%Y66|iO!H_6@!&rrJPqd#YRJ)H)Pgg0wj?k$Ho~F?_DqrHlvnB
z^FXm9;?1uTT)5qGdB*sNrsQ}Ux6gi5GdW$OmP39x|05;nus))Ce@ZQF?Sx8o+a|y0
z1!gC)Bp+Zvsf*hZ(|A=Iz4-OvTW4K54RBjw9)v~sCwmP$aS<<a4OVU%tuI$hg`|q_
zRWD%32;sb7|1a5>Wk6dZN+VpLr=gbbXl?N(?=;-~c8^v0w@Ukgj@%~E2Mw$E5mNKv
z)a9YBL+H)95WJGkWUxrn?G?=xx{D)=;16#u{Tc(Df4RR`TIs<v@X;<_os;k|YE}fe
zOGy4($5mX+zpUbY)Ok*^IB^#eH(7`_sZ);<l0cLfmF&sA2j%2yUh{`P_eK=fJ59Yn
zkwJxQ`$jlB#uoZCAOaoT1CV*>L}jq96Snj{`*l2%X-wuS1%nRH_@ov@e;|rOb725i
z2Sb@e1ws=kVotgsrtA$@$^}L_ozX&Db@%uF$8aA{Lv_HF0gN_*lcs$9C4(X5sx4hW
zm0D#b!K0jPtWxbjZRQZ6XkI?fK2r}hCGz03Gu}PVAhImmy-LgKFm;<8?v#h}*FY>q
zz4+TStPfmJL*ZU|M19er-fThl*Tipl*3tmFc5Jvn_FahO6vSF9mJ$+Z@YQVi<}_%A
z$xKm4h}R>3y{asLw{eS0lA6M|_F>a!i=q!v^1YbjoYLNeKc(E*s7$|z5*PIFC1%&S
zUVD%r%AgczB*ywk3}6l2zfax((|F*wP~rY0;7hyta2c|$+1G9FugDxq4rtUb1X>P{
zw$)`ZU<N4m4DP9kWvj+Ppp4yP&i4Tgwtj?><Mn?v;Z`AMoQCjtd+ON1a`*$-r0waG
z6Gvk&^6YrB$nb0&`)1jU&_|XYdID98xu)2!di}ZsauM6UbY{1X|9Rs4u2J+JZ)OaA
zpYA#?V}0&daO$;o-@PsgoY#l|u|BF)qylh1(Y7kO^roWyVb*&LbwU_Hs5h!gfyPwj
zivplVbfaHh^Ys64E2}Hm@tm8<=czdXr`TP{znXum<&7hVi=nIw>l#~s0DW@>W6V~j
ze4K=tPTl?5<(*?2$*KK7gcg0FUp`K(T6TRQ5SDM_+i#Pw03-<zL$bK%0%TgMV{!mE
z3~GH(-DHRaNg@p!;!s@v2BDG0_Ba1}n#}Z0ppG{AU9mudo_dx<;S?JXoyZ1cv21!L
zjR9#Zx`5vM{rQ&r$~X=3dYHowOfXO{9;YrADO~ZT!c{l1tWe(3u{IAS&k7%MAbnxy
zTE*6nGcR0?0Xol?)_pi#8rh<|dEp@s`^#&55$3o0uQD4=r4Hl!<xtP})>K@6`}It^
z|G=|<<Y|vhq|MwY>Y>T8OQuV7TNWQWo-=@H{0@@ICghA&dDfq|*h!R>=o6cNczmdY
z+XXyu<-uWk5f%y$I(ryg=J)c<BPLuS3mO}6nbO5LEIPZ07LrtW*t~>`Q&lKjz6KS-
zF6k3&vhl?YNGg0*r0x?gnoI|%lLwLHbr|EGc>Tza0y`EEEZtYeKnDi(@Rq3;YXBZ0
zx9ZX|-urIG##%%u8;Iwf@4kt-FlTLwGZ2mDwT0~)ML<WEY)OOEa?7z-pKpH#_=+x;
zXN_)+_Jor1Ffy?mJrHYo5WMqL<R$|}0UjhhoiveU_p*by$yU@5Cz?hVg;1x*cW=|@
zWh@y-ZmP#n*7#&k_&uz2;<)dMFVYK}n>--BG}=Jt(XrlF47Bb}o?K!RFL@Y7;>)9$
zy*!g6==J7r{h?^`U^=Ss^W<P+g5TZ)&Ik@rM`JOB-<$zeDOWLz$MLpOrHvO`$rE|n
zzGdVeW{Erp<d0MDaF)vPacl@*z(*VVEmr);=e{%N>{q?7M|$hUQMq5pqJORt*hc~p
z5c#9asUxlc4?8(Ov+IgUJn6QPnEi0ks6LTu+X(dqvm81?@rCr7&?!gNO`<uQ5Yj@m
zoVlMD4aqvbY=#6gl%aQ=`t@7Efv8~J`0G#YV7Q~q(8+?xF{u)hP`QF$;#2>Ee^Y_O
zf(_*F!{ou%e|=X03>Ij%%gS)A2yf2PYO?uyZg(fLjv7`r0--Gz3U_FoL+1;FCG<x;
zEgi^&xCFAR0<{MjUU-!P_z7UTN^tl!4HG<`mKAAdm%jvECi~2475vDuZ(#9fa`<0&
zM&o9d&kZtY2>~AeqRnk8L#d2|w)<-!m3JgM$b%9|VWf+_8%RgMn>Pn-h%*NSRB6#`
z?ckME=yfMatBWsUM@~t00=tQ5lggpKUWAMmvrhJ&YmMT_Wp$LTl}zaZh184T{K!el
zm55Oab2Aif!v^gLIvsr#X>LnS>VhgG=iW>hl;{|lmSWLr;_u3-qnSg*Ww}Xb*j??q
zZF`R#LN>>zhFInNL;J5N)81@4JhbT=;Wvou#zq!%dO0=ourB?rx0k`jc=bs;rauC9
z|9m??u?f(2L5-dBc|7NWx#DVpF-!D`a?JlE?-=-BY+RJVFXQyoW+O_u`9vb580U=w
z6FqcWy5vl%T1SP;%9=BWA)4Hpk@Hcbhxu@f)Z<nFJfInC11{h6q3<!zz;^qcSQIG_
zZoo-s=OvI{LbzAWw+F<agi60jE1Dzne5H<;H8fh}vS#KMzF9a1lo6M+J{GZynDey*
zW#DrDI7R_d5qE&Vwk_b~4fqwBaAG46k1n9VUY-_Zz6(H|ZbuOslob)oxC4||2e4u~
zv=tiEnu-6)$4!yx{w$ITcA`5<Quf@R7G>7->@+P=!?@TrPV6<U=ttmdwOV|he^Id1
zYO>n!7kbmE(voO$Vs=qsBfbvUMd?E`5%MxAf)itm(Ya2^>Cz)I29FDM7K12s5MO?M
zj|hs&g0z8_eL*`-S)oEp7!zvjXn-t7ko+M91ONiEL4qMj!#fUXy?LznAM*>3xSQET
zO6MuV55p%ws#gmI{@vVm)o=x(hOA|MW)~~wy1-td(?feto`ugz^w$Y^aRtwi5VqiI
z>a>{V>o}_^Al)s|gyhDFjUaGUV%%p9i2TJYHbz+Aq%U|L<IE#<!jy!-8iIzu0bc<d
z65hd7bd%Zqepf`Gy#H$MtMgyjECxDX1__j~BhID#_Qn!c3Q8Baa6b8!00jwDBWh3l
zV8H0!)TPqq0a`Vd?AdO+dn-Ij6piIzWC({$a-A8Ybi*?M)x`?EvYvpb#0(siy5r<q
zc4hPm!vOFGai#j;P(>RvM|q&J%@}YnEYiAbGq&JXiiJn%H@Kg?J#6C0QA+Ub34yBQ
z+>|4{TP-lJkA7@hr7KjwM9=TTMcnSg3$q%aqCZ~P?y1DXd97T>nl37p_ooIMf3D_D
zoZ#qcXr;jh9xW$CD7IY`tx>GV<*K;pgQ;%>*TLvSWCfvbds_a0%FuE0{7<ZE_d6h8
zI?@GB`aK<JzG8Z2KrQ1;c301rDAocc>p@ATSbSQ1*7{=`uxNe*diXIuibgk490dRc
zR0zTRkZnp?JZuL|GFt{V)Gcyg_f!{y|N0gCa<^?ILGm|Uq97U-yb@J4N?VL<x}UDx
zp^51NZscJ@BSt2EimRoX_m(=i=={kVw=n7rtw9DwW~Chw`&W{gh4h^jH(sGXUowFP
zrNgC0w+FE4H7OJGxmy622vX(acgH+w6PAlwuk!@;U@{=OHbp>7OBkI}##C}vY?asl
zgZ1Riny5f6S}{OJeiDqL<^<x$CoEcQe^3ti+86$(k@j5#6j}Ce1b5ZU_(H@?1}Tua
z(Ph6=04O?Zw%v;tNmmm}KG?5v-E_!o5qLz8D5*<xpZc;i?_i|2fh=X7Eo+ruTd15}
zqu4Q>J!2=RA<D23@F4Sw2L6o&>#cozzz+?tA2_NsSTHnk(RE}1-DyYYE~tiyDB+J$
zX{6vFPi$VF1vj_noNoi+;NYiX1Y2+J=<y}C1uQRbB{{n><k%65vjvk<C>u-LJ%pKX
z!dM{?<-*<pCg5n(%hxLS=KUmA*72WZ%ShyotpmtNbA}Yi2~Fh19&y2$Fy%NNA}ph6
zACSY6E1F`BUFjWuQ`q2Jx&tVi58t&p(+Pm<VgoYlVyRS^;84ED7S}^HN<2v5R9^2N
zMM#{jP2yBTb#iFKMz*fJfDC&Yt&gUGoT#PNp0QS=@#_LHPgs;9SaM26t#3%(v&D{l
zVyO&;&eVxt?{vEVSRslCitVJK_~1=U;K%_>EV^2cCIQf77pvs+NRu|ga}-{!-TLpg
z1@BL><iW<}^4-#pQcz{g30U!*$Eo=s+xPRy^vD-+X<aGOhu>qWz$1{#(q%E!$1?Ow
z|F4G!9H52!S;O^W_@3VW*Ep{`lIBQ%riLy2QQ5k*dXgfJQY&3esc}#_d+7heffgTO
z<W=04^F}%v|1M!c{n74C{o0Jo?{RoaS4x=E^Ei(zK@&NI{Jojpa3Zx<=ePUHW^>ZZ
zn}VWl;@%w|_EY=*UKruU*t`3iYv~#Z5}eo<YOY~1G|#?N^1zA0Fz<#65{ns-e?4Vo
zAP)vh@6IXwD3ZipU|D!t*pc+e2Cb;ZD{~7s8uV6o>z$M`6wg<GmC;VU29p(^kM@VL
zcnTmSaWqF&mZuM2O=(|K%2v095`Zh$qXH(G%0)3EIDbcpl>I7S*s)hblo)y-5pJ4i
z%FJNR$m%m%zO_x}$_C;w1)r}9%L=qUgx(G(bM|iF#`a2z&k!y0?W?3IegA=)v&R3(
z)R|R0rtzxE)JW3LT4ssQ7AjJ#2W~icF3?Ur$<uQu0#k5%e1gY<sb|%5P-%tMICcba
zt=m2Z1wO$~HMr?wq2@aeAVH-ODPhHanIRBTuv8O^>z{m+s<<_uJ9c%rLVTiFR-0nE
zghF5EB+djNR{%z_7AX}^!`Vh400!6}#Nnk9T`!KVO*Pm1YOqnN$H<xZ+a$^C&{Pfe
zrIy|%UoB9rLt`!4gIWTFsD#V?p2L61RU(Z`bY}6h&Aa1sZ>x@@8sD0%o!OaWa}h_I
z$_)g%Q-Yfe9zjt9Z=M!~-FI!JPx($29But=PDjkU0XG2EB*G!NCv)oa>&ou0k}7$!
zqIutFXsRrNX9@%(fAZK4C`_6$etJKe&<1#uaKPHq>O=wS6$JvoEgRX<v!B5{0bH-w
zD6f9aP;q$gKsQ3nt3#Xe?J)F5_XrY+xoBAk%+$_p7zy|sg*$zPLB<LMWptwMJsGuI
zu{+ZhVYt3%b^^K^@7{?O(9l%n_gt8_aH<%ZQMYNG{)Rg3r%_jFU>7w7t+DtH_@9_l
zDc`^6m!brH+mWuK=XM;CHIIi^RsN+zYjqOr>KTN)N5A=Vv;Z~EWt&Wt5A)y%9f)*@
z*;gM^PEu@?Ytig;Sn%zEo!wTL1~M?(eFvx;$^IB1SMvVdvzLj-t!vuDb|3?3BdqbD
zMVaD5jkE7<Mh`?Y;g<1lfW{TqQ^>|D!rs0eX;#u&)H*F`CywEz0)M2$JQr^sSPcN2
zi$D-T4!Z|E{ec28UwQ*z<?QzP<CpV+dShE}60d-KkU9y5EuV+)%s)(IbswkzsW@{P
z!o+?hj)WHfn?0tpgjEU)8R<g*>pGF5eh`S{CjGXbq8qi5#<Z`s0O7$Y;XNI$mXa=Q
zBRMI_{3$0DMtN=}&{Ra6d4yW$*Z3Ul11T?E5QJt!29tK@*)-X}+zI1n_cF5M1V-2p
z9R#X>Cm^M>UK5QSW5p$|I*C{_98L}h_lZNNVB`GJVfTH#e0wAblWp^nf0P~H@s6?R
z-D+LpLnO_nx&A^$%A3b0(`SEq9yf8k>e}u}1e=(>cRne_;Fh2M?b2fR*&Uq0c^i%R
zMxRE2rBo8-K+c0iEJ27{=}MC*LTe=7K3fT^<{d_(<jShP=7_-u*~XM?Z~^oSz)5vL
zw0+qNupB)Azzwi<H1pbvaIadseMkNhV!$eGb>*ag+EZAeM%T`WNki|X3-Kr+f1Wxj
zaDSwh@{NOy{ybm3U3fX=x_b|HdXS7;kzy+WYwx`tRt)@p91bh2jSdl*D)`f75W;+!
z&=@O2DW&;D_FNrwv{;ah62#p>g-0VS*MN>Bk&bD)|4T%s?X<iu?eMeSR%ZAJpT@Cn
z`QY0AxS#>a{K)%UB+GwvJU(y486&xz!Qw%F;U5503Qwz7-fdAvMY!Z{)@>gBx$i3E
z$BKr;_ESrqO5JBx%@A$xrUEuWl>m->*G<kVZnu{Ck4VZ6x2a#^{7wR2$~sCDI%Ga|
zsMkdEGcZ@*YScyTN_xP;7>;)H0C2!H+L>(rTA=nC+@k<(m(`0=rqF)6GP??;sTct=
zpF~jQ6(s|~;}y!2$$7(iHSm3l2lwrLS<?$l5mo$#v~2l|oCx}6j=?4sP5mIs6nF{a
zn<KwEv#IukPo&T^1Z*h_wMXgPKUgI<tuj+_un3h(cvm)_Z|3qF_c4cz8&~@M2ejSx
z6W;2x<k`{l!TOCN){n;={%U}xY=jLZM%5urBg|RBNxW?I_GR#tM;;2#Ichwe32cJz
zh*Tj{Ek!k=lFpd)xyf?}aR~ismQ|ETB>^B%Z{f-T#4k?mHRVxn>qw72^Jo%Cx9FdJ
z%oYm)zGcO90V;eY19JZidGicilnmZA<v!lmy5MgTby4#t?FYk|Ek8>q`1TPbs6%K}
z)qu+F<Lue(!XRYb&-{Tow(EZslHIdjz;d9NBjB=stO=;<WfG4y+Z9%#noGQtIx(%Y
zhwxH+pIF4cPpI-FZ3h-emVDLU*CrXO=!r17R<B&L)Tu=7H*>4eZbTtfJ(>nm4e>b(
zvHj2#pUm8%&xvIY;o?-qfE>0$a1eHc<LFo!nQLjsYJ(DAi76VWgGw7wBHhG@C~b&+
z{nKd+H}6{j+AgKo3{3LxyLG!$wx`qx;a<O|S^LFWm?n7UIU)0D{<qqbmvH+4w<2-#
z%t9~#2-Q0+Z32@C0l{a>GA-hQo~Y5QK|h2;{l(fHm_iih7EMH{=(Jq?O^61*!G7Fm
zs-b^$afz@4om&uoA~#PfgLdzEbVqpQTeJ+_Mm*>^Psh<e%9m5l`CYqS=NL5gS1JJ^
z7|8j4zv++U>cGfVkb$7`_3JcFo~00X6?4p2qh>&TBbwB+vF&4~+Y^7qmOiu@2mqVF
zhW1}eH3>oztYFLiRlbF>O<|#L!H%YBvR`OL|M5olgCs1RYIoQHF8<6oEH*f96=5O3
zhKi6fwls_?K?2Iwgd9Tv8nNx0#15EAw`Gg-Rxx_0F*uYIg6d-?lRMYseFB>icYDfW
z7})CFvw(Tbp00RP$ETK1_`sCPpfmhRlW0kJ88>z?h|FDln1W;YI8)8Nq1OXt)pPT)
zJBFmwc`))tuBaSNrl@i9I79%{)S??R3IkN=0)vd!_;;vQJ9r!hO2rMON*|66G5)Yj
zL&kUp)?tl#d&}SqM9&0QfpqGh7(l|&^jQl)RpUGV`}lFmzg|V#aQ~5nhbN+j>7?iM
zH(7tFwS)Q4znwm9NnK43ir!d6f{io*F3sNQ!mGm$#fN{LAL_8xT0>j{##hN@Dyv*q
zU4O#|ru)s;P<Te@C96irrQ&&=_^f4`_mVf@IThLIa^N%Kav$V7!{hD*OM2H?0|XFU
z!mjxo;*I)9ysS<PWOa|EFm3tWf@#buvsseH(2OO=+gPei#MIB=7^2)Ek4|hxxcX;J
zG?+=#Qg@u`r21~ELQPi1bsTfay!Vj(a@j5C#6CQsiT9(qH8;KIr&0}FJt%W}e;@ID
zc<l5`0JbkYLKK;V{wY*A#3UmsFQ+)BVo8I9oLH)P$8K6vEQnF5sx#uJ%7J<l@`Jic
z3#7a5>8KVwF@-vTnq64CL{ws=5l2U%RF?RI&qaYE)9(3?{hd$^dbHRR0U$N|&VGG1
zs<!hI?1;Re^Q5E>=f9>8?@l>Cm<8Q<zhDpPnR2Zh=DgE?KM}ErYDeO}bv1N0a)^q&
z;$n=Sl;+_>rog-nc90@VkH7NJ;K{I8tro%mEk}7`D9q|Mz5WX|Ex;0orRVk<8A@!Q
zrev`p4WonFMQeq1ExM#;jr+j36h!ff`0K!LhR&jrUaALOydix3xU#->qiQh`d7DC4
z(4b96g|}_lsy^VOT%{{zHtdFsQhi!voyPj=6je^JbtkJ?l8Ny)+19<w90B+5*j(bI
z$z)I|548n!_M20sSy~rG@OXQy%S*8BimsIb?V{u1y*p=<Ly@1T&^9&8JQ;sOqG}V=
zrj5Hl=NTzB9^Jpd6}-FaZ=WsNsI>+5FI*&^!kN%C@RcxBy2OYTmbn0t#@cM!EJ4L)
zm^Kfq2-9<_+gpLE0a!@NWlCp67`}p@T$x4l-<Drb$JPBb5X$X&r2p#7`Hv^UQ>9gn
zvzoL2>Bjpo{U`)dw843Zbng0Gv0=3^(+v)S-y7G<e?8ax`m_8E%obEPoHy#3464n>
z^8iimd&eSi^<RAEfohRk3E%wHu^igg>bTmU1RSqzwV1Lhwo#NTs)A08IYIK49UJ|R
zy+&72s!IG8!vR3xD>`i3H}+&faO>WSr{`t&jqQVbcjRNI#Mrb_II83>d<?y_He<op
z7E!%*^S<bDK~v<QYJaPIwJ*KC7dozSA(c^ZUj(%vIippt!(8<8WHzrOTEKc)qD0Z9
zSc8;D_MadoOexb5cm|mGccKp}sUI#Dm*)SKe-JiHimC!8V!B&QP$wF^q$9HxgO+tZ
zd}*jFU#641sgz$E?^-)L0PS8F1{+g`<go)KA;>uJgXb3!!r3|F_5e{whr9Pc@0GsW
z+g{x-)S+ZZzqZF2>%aF$Nr4q>hr9zVRoO<($}DVBaRb@YyCPGp`R5#N2*kfz=Z`HY
z%k92|-(aC>bAy`{rVC7vjz;w~to5hi1UbJje_S8++}_<z``N$q)MJP4;3b_ySHf_Q
zXSJxJMHM2)Mm6|i%ahrqGc5!~5jxN<Et*-I%oEr#`W>De=_!Vb8~b%X?)t6!pq@_b
zgTH5%MdDU;XmBF0CPyw+FPxeXwys*Jz$a$coWfTAa9P_O(a&rr`_yjIe1Cnw@Mq$1
zw`nk^ypk}(*>wP0P-~ubE%0YAV*=Z3g*^8*Hhe1!kzSzx)&1Y<jypvAB#Z2yZ+O=q
zgh++avK>FQ<|C5vg<{3xuOvrmho|PB&Td#-|B1B)uM+^4Mz_<=0Wq--7w&F{|9ym7
z!M&)$ua76CAGWbJ^hxx3&=)xc>*JzRaRJU?O&eBpw_HX+OmVtw$L^#lI)5tY`*;VI
zr=xH+Paa&?bY<_$t_MT!ouS(9`;a8{q<$@_-|CewI0Enc&(UJ*`oQX|+P<SIjHyKs
zsQ_VAQDfPG3wioU;G6(%OfiRTKBI4q2yH?=NU(rh6=64yzqJ6dlPG%Ld7XBm3xp20
z->7Zp0rzVl|NRaMRF6cELP<4J6gZ&ID1DEdOQ&$nj;~a!*JO5))kbTZ#nR3KafScV
z^vRrzX)Fg_%&4;79cgq`^#53ZgmW1#houUh?f3e^@z|))zfeXTXi=lV#yCFx%E-;z
zu|wo#jW0{2_zJCC=e;0%U|>Y(v#;D*xGefYa&<Vaesy;yLkh-yZn#8>{45_nHs+E~
z$MCcD*Vd>lnF2>S^q#Cw)ojfMrfa7T$*+!$_=2AIX9CA=3(h}>zo1<utcp~uw}nh%
zlvS<BzutaZSs)f=l(e=|9l4t{Nd1evU?;=9hFSOBO*$G=7Axwet46D40eJJz(J{)@
zw^g`xG<-v0qx-h84a*irt<Fnq!#wstURT&hll>_Ax|W;GL-+f+nLJl0nY9b23JaOB
zjc|V;AZE$GW(<wv3JXN323opaQL9-+SL^!O-(=tGI&TukIu4kP7N84gRaahhe<ohB
zDFhbf-oyT_>c4}<gVoG@wncWhTyL9a+V6f0M0`kRzXOP~HzwlSqi5nP!EG-iH+TPU
z+c~u3zv5eQQTkdqha<FEz*_W7T&7KKlIJ7t0wx;oZE%z)$?KUea+E2akMUB#8G|ak
z54vWnk4msHwG-`;pC)i7L<qVWYEYpy)4NI6)?X1YIoUH*FeauX43C`O=;&RmcUJ-R
zMEm*N2hskUwS*GuL!9Zkx>W_*MCy_9?RAy=tSmRfRESz}#=iV*$xg$kx#Cnk&y%VN
z0}mHG+C)fgLy?$BQ^FEOvYP&a(v4x&n$-!IZ*07Vlpm)*!haJ;!n&^YNoQ64o2G|(
zkX|l!qU2RNkkyJhwshs#jqls}qrklprHb|J={b}AG;(04xG$V&SfJOYSMFeKtQ^f9
zs~7!MMXW9VRYg}t!7jhb#c{Ad0H0s_1eBVw((lrIS2BH0BYxt<!eq?on`u}h#2Y3v
zt87sts8i0#g@p8Z`D8Ratp_2eCI6I>gQHl<KnJ;tnx{osj3pYQowC$_eIbhe3#{5=
zy?_%44c<i4b=`h~%D0&>H=3x#2=WS8YtHNVzLSk>)}dBjo^kPx2Ns{QaZrtiL0w(|
zJ+3knq-QUGMH?WAN}P4Uh0fQpJKTR=)YMmS!P`#Zu;9z3L)NWapk_6h4^c(_JM^t-
z6Blno9=RbES$)7yb@4f;ca7ie!u#VF>dH@%S{Cn#<EjCvx0aD`;?Xu_1;cLdp{SFW
zp4+*J{YTt?IlQi40k>gCeFTBuLwoQ#VC8jcQ`;ok8t+#4|JRnzf7jODNUpb4^D5&L
z#+|yAa*|iC9dl0tnEnK*rx?3CbJ8UF8y&ZzuWfLr9#;JBNuT{>d`70XmKx(!Bod47
z`;eTuzAKPNt9p{9elMWm@U~E3PBF`4*{7^1Q&ZL>%}>dp@ijvPY)%HDjzf1Wrcut?
zMoSg@{ZbDupt+O;pCD@R_>rp#t$Q%|4J~r|P~gY*5q3|j+t&V`pWAu`=i54!5t^Gr
z=&7~#Nr|8fIj4ptl9}i-4=xXIY5E9uDKYCo{~~qX5nG^Q{yRG?K&U_;T#%^}sm6yB
z<FLGKBwR&$aJ`b1AiE|+${!3Y{&}k2;7G{AqAo*)qgeP(Xqm!bUPLH#HesTFFkv4}
z8@vl}JzRe812oz8<T36aI}aR&hRs>eNTw|1!_!rPme<--4oAA^zI9IJUkW)PpA|N)
z@Q0zTs}(ZG(6FOAzI>e;43f{d2$&<(Es>tRdD`6xCG<R-9jVejl20ur7m-laJu#b@
z>I*Pi{(5T_&GY;xhH%=ZiO!gQ&I>I5-0Adz89g*6%2=r|0FHwp8ayU~W7FpKn^cyX
zS+;wYV_`A<nLwdNiHDPzh~U~_``cm1KLi9G1P5gcMG>6G{~axxC=4=gvx=W1YZtl@
z#Y!9MkLF%qUYhq>9V_kSx$|eWrDFdhZ~15olWgJkh+(Sd7N|0Zk^*+{Uy)pVJkxh9
zQY(*+>TbFV-5}l?HvB~b@Rn_SqqKHIvJ^+eCiZe>8SVAkV44!eQ+~HYl3o<L->ZcR
z8_OGHwy4S%`Y)nqx;Ma?)ceqz1Ma;i;_rUCUwhbm$;-E5l)MW<5GWoU265Kveu|To
zJCNmMINDItuGQJ_+v<c9z!LU4{eM)wWmuHq7Bws-CEY2~4FW^Aba!`34PDYmBS<$$
zNp}y8G)Q-M!w^c>_jt~EzxR57b1@7v_p|S{*IIk89{hKM9J=>`&qc~hcGEyq$*zA}
z{nz27zOe;MkZIreeV)uaPvU}(E%=-s4u+vBpFEfwC}k&EB;9JN+Rhn5Z>?JCv5tv?
zFlx%4?@=V(550OYoKLh!@ZxN;8ss>V6C+cus6KEJGQjihoycAHC%CUY7C^Xtp2R9s
zVc!4_eb#v7Mm~WKgG!H)iwH9rzqdApH8n78I&z<{eV~KoKCWe2DIu5er(~G#o5=^l
zSZ~`186~zPMwiY^SDdDuTi`7V?iD6}ym)!mweGrE=C~qBN>V+YCF8AKj&$!PRFu9S
zC<O6vu&&uRN<%&}@!<c)7Z+HJr2;nm*4>x4v}CC57}b%QL~86`>86s^ocr_9Ok~4;
zg*9`{(qMig88%F3J~&YBuK8xGWlCcaFaKB5u<Gm1?{wg4sIT7Pac4^@j~E->UvMlT
zj`sS~f7S8Ye!dB3RmiEk_=d?zcMtP6)llr5_aiS+j5L-~=FP%s)TWRo4+Ub7W`UGS
z)XO0)hi9N#Am9|ua2;J;NM-3wCJrXGNi%C&rY_cl=*{p?upuex^D2I+%RAb3OV%s6
z@BJ<UO`mAW7Vz{3^qvc(ib-;!%K7`(n%qw4gM<rWd&g)>NBX#hu33Qn6JVrY>GXSq
z`tvudLf1MlHagE?{uQK~Ktz9F(EwoIFq_$LmfN$f)%YM{-Oo{D6g_ug=y<)ZB}<nR
z5tTJ_xbm)Kp=58~k@Oc<ne=-@(KT#?Fa2>xNp8=lC(44*yPhqIkq2@*661RCr`Iqs
zk31bpvTH(<hTfGzXZQw}s4O<K9m6;qcbKPL-;dZaOmYl685Y9CigCIlyHrTKt6l~|
zq>3Z=ls@Er*(U>zrxFTKycpVD-(kPE?)+4)OcXkg?7^j7;55JIQo0}8FYRxYrGeZv
z7N!sLwNA`y&CE;$?XXt%F<B7|lcYTCO`5}7)Uf3-m*snOyjzqX(HlwVqY)0@IX1qb
zuQkaL@H3$&jHNwQV#@KlDpgk&Mu0o7pWwASnqDZWUSZw-ZB-ugfj?S=VJBEVq`BO^
zFV*IcDNBz0dZmsByjQYD7>1h*CIU-khLOOV$YcZ=hIj>nQO&h&pJHZ#aRfgNi*Zyv
zl=)Cdze{5&5jAd4^BaYPjEV{lzT;$DA?%njh$ZXNUUasOozo4of9WKQ4kW5ojT0uI
zFAS9P_IFjNsQ(K7?uxhUjH)}&0x@|ZtzPv{*JNSnvLUl}xqI?R0DgsL#jF$$U$M72
zYJ~dOuarI8?0ojz6qU@Z<99cj(iS${-@xvT_dDc$D@ZN@*PymfUaVv6XfrqTJ^Xf`
zETS6qzCLIE^q(|?l<6+Pv@R{*{F!LGW||FOZJX=(FG7AP7rJ=9)fag>+yNjLmOEuA
zQVGPk;WRh70-6WE3_jX2p{Du}`*F5F0Jk62f%M(e8Cz=BW4J3~_@~q9Rjg`?n9*wT
zrlEm@RSxUrT?TDKZ#;+!0gMq8yHHV!HD$!5^N=B76>lB7srS7<(c`(pG<<7Z;xi!u
zYGAW$qXc8Yt~@YyK}x*`{am}G8_}un99X*cu|(gnE!lNkxVXCb7MQXOzD@B(smh8M
z?0z(%27hcQpm6n6HExRpdU&mYGz7>a0I+@<tlPnznec|a#;yJER4K0)G2_Z{Vb1d2
z47wkDg)a8wQCP;b`r1x$0xL@R->zF{dGHrr=QlOCFU#CrG~&qNyI@ig^%z{Wb*b5z
z+&L6)d;}{iflWnJ=?VAVtc19lM#>q|c;}d|QjTr^`S}$ce*m@p<hxLls20jj-+tA;
z8ET0*Nw8N$#VFBByiX0~4ZUZFLFJO<lZ>AlI0kN6&(M7$D}0<7(`HHM;VbxBjdpB#
zIN2FK;&BBgqH$($XO2Jrw8%{-Ns=+wsd-G<k*De?tIFJ@UQN3c#Z|2Py%jJ_U`Ntw
z>Fe<NrL2;xG}1EbWs`7y&{}G@9!ZTvMbgbW!TZyk|HDzS&=U=?*K)%&yyg|Yr&>1I
zL;dCBO2H}>qkWS`sU+n`K?H30V_Vl|j@rjKAB--^7ngti_EYMY%~Zg__4J2L#V$%M
z8GaG}ByHtJz9gnT)k6%}i{Wc8(UP5<>|3fM5q}@^73R2L392^kX(A&4BAl?m69eL}
z^c_XECwsx!K5YqXG&8!Z6->$Ft%I+hf@(<QTo*C$Wm0L?b?w|Er}J1l$hmU$9@-sP
z{FvimVm)1e$SLD7`QsW60>?yKHT?dY57nIky_?3bR|_T)N~*NQi^GK{^w}zewGqw*
zCq_hTHG3+&Z@IkKIvW}Sp5H)|!t|4>H+Z>BzCO27slG=EiQbE%<NtIKxO4Ks5)XVP
z<#~PaDaeU2F{UrEPVHjyoC1>^C8RRTX@yE~^%>m=&#OC6)ayFK{ZVnn`bVh`o8Dh$
zJ}{j~<KYsgvIrd9`HT|ZbqvpXp|Gi;5v)hQ_gb)-$$T!EtXJg}yFPky!`zooq$i&&
z>=m>1*BFmExpGq8wK;HsuxE8a!#I%@xwM!$Mt|J^DrHKVJY7Aig<UiEcZX%ua@ibV
zjE}ZpgwjVT9>q<MeHerpyBMi9Ay}>{SjA&}Fb|FBmn2~WR^6-Vqz+ocfM>KEB`0R6
zTR9_nw%jBV>9@q-bs}h(3`6Os5hqGf^nA~Xp7TgUb*yCGp9)3^m8ab3Z*Q_&Z{J&w
zf1^|J4_SM4^40%cDm4xQQe9eCve*Kx3QBw8qn2lzfoRSHk>w`vGt7B$Dar0*k?mVu
zRpU=%a{`QuTMTaXKQo&`SpcD~&wr*!sOwJh%5jN5sx07oZlUeS+i$2Nlk_C>8h|OA
z0kw!8r;D|gqbau{caydmYW0sYmV-h+outjyQ37e0@c0NLt@qj!fUVKJP#CEiVFP}Y
z;cxJbEwFn*J+-;Il<BOvqdh9S-A9i=11<?Ji*C#1YW^gRTpq)jO0f)%+RQVU*Pxq8
zP$NWb3a7eEa)(l|kFLA)t+3NzihsLYGVtH$E3tB!r>WmekrxAs4XSLk%zKxidX!?=
z^x1<E`en!S7zHiJpBsUHfdB&Mt(h6?pomaoqW2L_rgV+?oo>n_aT3>yIF{>i%VQuW
z!rxYS9PBw`3<kp9^WwbN2c26<R`jIm6*rr^ElHzn26unA`{jqkr0>(Ma|9m0)w5p+
zKPSylY@@xeGYdkK)Wb5de#@4GP{!RKTIZhh^Ebw^n#J0-BDee3CxvFD0Rg$m{+Y|Q
zB2zuTG{fdh%0PAI`1buWV6Y7A;nN_m=9Z=L8x0}=UO$$p;LadZa>`0(OL1&}8Ng%G
z%KNd%WteiCXf-xpQWmo(p15On4~#I19A8jXwVb+Oe~v8PY@7R10OC0ILiv=zsBPEM
z4$-J=+l?xI$Omb}^o18XGiN%C56Ab{&0m^@Q%$8?|8RXek5xyZnMWf3@MoB*v32zC
zwZ}@ba)sNs_#&%QsDWMEL?3WodcJDgG749UA`&a8!<{&VfTP~Vw2-tHKA1rs`gTx@
z6?0t1G_~vB0_NF&Em3pr1_$&LbYoj6pz0q$d8HGq0bWH}<hgmt)_M=!$nN+t`AkBR
zAud&`o(Y4cCxIan0dLAnPb;cU&7^c$O)ceVgf4Y|<WfRTAxUyVJanG^?gnU&ZZ{ks
z04gJHY_A`GnrLYz#p#BvqYCeV3Naq%istY=6x%uh-(cIp1%;ItSA?zfyXCD{W>%Z;
zeN>mxQ|jkc7LxZlCvzC|8D+SCevYvB-bsSPm<qs)vf&DGc2%8V;C5@?@IXxGJ9kg4
z;nxQR>J{Wo!(si@eRJLMC*?LsjV~YYIy=&&b-!Ij<s<pUDPceUw~A_P0H=!g!c>kT
zWP0J_rPu-@QIHB36G4s-G_-y~kVI_^D7cz;Pk2aLgXXQH3%r6Q$jA21KWj<yOM!L+
z!Bb#6#lvDPxKeNKS4jtvBs*)4_}W}d%97>~fzUZGQJYm$%lqe9y<M97_Oc@Hs6;!f
z=)yIQl`5rAQG3U$e9Z|J%PlQyH1CS|e@AY>U&*RwSAs{pHBR!lJ9@5xf5A2k+$6ZC
zYckK%mPlsnc!yoqc&YnVJDgNP!cyXsPe`-3^PPApb|j50T8tn$tN`_Tn3(jo^Qvy8
zPL(*A`?n>6A`jq3N_{P5Yd>3QPSmecJCq(;^}1qxnHy@3pJmo2S$zryT6hqsm;JsJ
z(Xs0fl=Xo9_QUP#?8$ER$-j=~RSx6n;iMXvsP8{48UrF{;Nwb@cp^?-W&$q%uo;Ow
zN;Uo9^tCmIY4l^ob&ds7#$V2|4@Wge*Pt}w!}g9y5Z`pGT+A)3ck>^;D-%&J!h%_G
zUl=#yU$2Fr%DLWJ?06kHap_jgazJW@txt0xR3$9w(s7a7D?~}!7-U_m7Km=)&3nH>
zek4Zdw@GLTIsF*2bB7~TW%juBQcFOs((1vftXiuw%F(oYaApjlKB#vwVn_n0jcR?7
zHzZC}IyFn&Q^cCAyI0uN(%SUci3aN1@d9G>vy6Vy_OQs|i3l$F?p7N4QYE7nawC@+
z&3O`T3r|>uW1{qWWg5e-XjfG_v3=MUOg+{)hQ9Mh)@GZ2^F-Fnev<I*nu^u)l3W^E
z)3jaeD}ICKcjNP1H~~^5EVDm|2857Hmg|K^@{TAer#;?|-7L1I?Z+?%G-iely4M%T
zkcS`5SDXR+>B9T$E$|yQ1=;ZI8uiI=AnKmw?)pks=woJk)4KK8rSJF_Ron;JIhuP(
z%Ubrg6sCEF@$>u*X13=N>o|VO8YX-@OeXFs#|I%8Yn>=+&ZthesNaA<eijxuv*qf|
zys7vk?}?>DoP39*$8FsQ1|VJ}C*_J`v0er|`h}eC9W=bV(=qep-76wb7d!gE82~mH
zxO5?ub}<xy%oe;!L<Hn&staskS-Y*2iOSsZ>N%IQYtqi*4Nu_TySN5?4w$?R{8uf;
z$rn4#2N#S3fU<qAWgDYGOZSpRbKZ&ptu*L>0_~UrbZyj9jP}?*OJ-C~tjUQDQj$%I
z)UTEhK1L-Wy(wC?1khbTnV4Ldc|0sYh3%064ihORJ}s^jL1>7n#%2ba+>XSFfu$fI
zFWurAd&$w0n!EZr;}lg0EuW)1x;~TE@LM6zyS#eD=j7Q^vd8O4xg6^KXgFV1B(}Ef
z%g&Jzdb;R+JQF%wwzH~7S&bJRVzF1E#v&jijS4N01<&z&;<>OBMyk7Q@bR5rog)~@
zOjBsb(x{=t_kb5_0tE=(8hg7h$<k%Dtn>d8t?ZhD2`VqZwEvJY<m3vU<Nv@{aw!bl
zE-hz$f~NZbBqa%iP><1KAq^E|a;pn!I*qCZDmyV{6Jiw#D8V44p;94zYK61$xc+@G
z{`3g|AIYi4Fx<L9h6HZxS9BloR(}PXu#S*TDhep`Yk^2^D!clTU&EI0vpg3aTIMIm
zv+u)E$3;X)Bh@A%*s#h|5ITn1z`xdzN_c2vW4IQuT-H7(mtPax5{*?e(R#uH)X8ZM
zmYeA3<AF8LHNQuPW&(N5OU9NYI9fB9vpBx=vbno0<6irf9+QJ=$^~NmW^jKvZkg57
z1zhnnp>D^93h1{GYmq{E`50}_oHSn$V;{xq-ba0^P&FFDqK%TNF-M3$g|*~+7lFs2
zg?}t<_ibRoKdkX@8wU_{>{H?+ry$|T>Z9P$;SQ|~=(?o6tHDPNQc3EAtL3#7P1ZzF
zDV0~okZSoLrWYe~;1PMKRc$EZxDHpoP$ez7aIdDs5LOVWZDi;H<R)uX>(bST;39dd
zUX9U+d}8R4lN}jZwXo^K<55CjTgO;MlQgp+#2gUo9_!##lO7J#Ztw6|3j=;z%G5*$
zHbh1+4>u}O14!6P_j{8O3rn!FqZ^#-$(Zeq<6^BHuDI^LAc77MP>tHgCt(`QxJMB^
zbyiDQCkLX>>n$b;bH!4W&iY^4-MHywP|sPvl7SeRaFfL&A6wOY%-Oz++}r_jUZ!IP
zr!x~_Jf3#^eNqz6LBC+*p{@z&^&d~td5U(I^;6iw)POlRMf(h^j<ZJRN2}ddckPvS
z0Z5#ZMlR8HnChVllAA1E6yjG_!su*k(_rnRLmt`3nhmluF^S8Rc~|I`q<B0H>*qGx
zVBoHQh?=!)l0s4t)<&UI*(Ff-{PCN09-Bf>u%Tdj$fwyP3x3xJhGIqN+Q}FmQ>i}h
zFB{J<l<1!2g1NEF2`iI`-P0O52(i7rh5|ii-=ueqVsB~@7?|FBTzmki7Ql$sP%j&6
zUc@DsPgInx?3BRgt!CaNee-7h<>m3*ecLNvtuo+#PJ31xKZ`ku3B-YCESj#xti3`-
z6%3D}C=KtT<%ePj%Xv_wZ>6yD<0%!mr~H8pE3;Mpqdzeq0nxWlkblk29r>nLV8{k8
z?XV0ESr={RH)K(Hv@e$SBTE|Lv5llmNhLb(1K-;k0=ekV9YE3;{Gbjy$3Y4o@E+-Y
zIOS(Q=ZUsNCpZq8*{?okywkQSBn=To$o$}lfGh=e$zmGBw)qy`v3&+Q;m-0}y+oYg
zby~W@GZ;OkP}=|Ahi#;OS<Ap<O>aLbkUdtr3wyd=62D~<@+I`&onq8R4%!$QxjQ2x
zMXz1jpQ_5JMML;CoP--w?BbF$rkHypWk{=LGf3}DnI4~v><_XJLi?tkf{-eg7ZN?)
zVU;|%Jq#?aU2}fSuDGm?X!fi1)ACrkIW3i($nv-<<(>20Upi>N9LR+Bx>1vN|LzW3
zsw&Tr7q@ZY4__p;JZ@!_$4^Ww&&WICvIzaT7MH){_L)O8yx8@|{zoTzW&6ddOXG({
zJbMiIHG7y2B1(<sW%ORdT9f$Rg8FZR<oXwVILg6fm(0G}oC+oKJPNpX?wqJRQely?
zV%KJn47L<w@9Rdr0<kGYlA;m}()D*dt7}{eO+lRYMwl@?D}Tk{f?dnNsT_e0TDI<2
z(jWC5dX8&iB;;BTpgznuL2uuogxGeJ9<gZhkjB6<dD&UGJ>HyUIc6C0H&~9|X0Cq;
zU%N+q?=Mls$)UtLl+(324W5TBOhO%RI%eE9x5j89Z+Y!u@KJ2N{c#(-q~(*eI^me=
zH}t&Xcc}lric*P)`SgM>*)=-%MJk~m#_8B~>ErlPwm)wMFe$?a0XD5Lpl-_#3p-~B
zVBK<4Id-)FZ9K8w@R|@gSIjdq#vv;+GI}6}a;3=0H^F#kennOu-kQ=8A5K-NbfH!P
z|5im0Gw+_<FL{Tei+aQmn&1O}eQf7Q%CVFC=f#bjy%$IX4m3$ti%ZT=Izy^KE!xx%
zzT@^}MgKCY<j?_eh0FUmA_O%QB$O}w0RTzAVx_k2D<(0tJwc!0_^?PD2I(B{pPWc@
zZIK@e-uVA!F(y;u=|+__E465)I)B5xLmwA;2#_ZXxCPEDv}c8l62xRoaa0DnT2Ya>
zhSAXd?$Ej>z+F9~^p7-iwSIBlD+67`AFI_6+bI@qa%VJAOI3)$)I!X09GNE4W5KoZ
zhgKw*6szuh1Sk41XQ~v8IGCeBesbNB+pggBKY#PB7spgavgC19-)n-tBD|7XzI|fz
z-r57I7tzeEvOFKp0}j7dvJ1g5YD+(tc%;OI>o4Z|Pc4>d4hD&*90mr=+NDlB{A4Bx
zRXxJ^jPA)X*(l`daZ@VgprJIDCgV@p?Q}1cC3BTjWyJKeSRb=wh>FTHrik;F+P7F4
zG)`h}*I!LAHqXwR?1_lhzUd$@M@&&94=1dA{tgblI`Z7`drwBNwD3f77K@YO(kN!7
ziG<P>b3>mR??0B|ZZXjX{9^5M&B<{X2Bgg;Q)KwRXk=X|aO-yEyj`DBn~W;NNZ;2x
zCVWTvtRaaikL}t)(df%hQRz#C%s2rXZenH0{7c}>@p}yf0TZT{r3<o_T2M-z7!^M*
ztj6Pxe%<XF()`q0Buf@@w)|m8a`z9{AdNj=8=z`<uvD1JlZh9XR~G5gNOyKr>CV@=
ze87za5<Cxot<@4`CSoN~ZN_%k|I>7N{a!_k-98Rf0lP?N#PIBCGfzH`A?(yr=eH2-
z@~Q}RnB?mqg_-xO{fxftv+X1szB*Qf?TN_8-M)db)d-DGCVTz(!80fEj}!gowF<^*
zNI~Sm4`Y{UWI6zoPTdn!oovYX^P}#5Q9r;|%&V}cqnMg~2Xla9)S#seIh3DK$Wmuq
zBw76xhupa-G=@M$Z?-CxokDv1PF>~$cBEpa;Lqw}&*6mBVdqqiSUz=UlT=p{MZbBP
z0MszbSe+!$qV0EXjWKl{a*|RW%nEm3#6>)&Mk#7yp_6Z%-1ej<=gTphPCTh<<6sos
zMksApK`F^>xlws}FdS|aEn1Um>HJcOeYp^$?A7wib*>b0Oi_^WUD3w@CMlq!EXZ76
z-h*E|rJMHDZ)%973VgXHx!JJZ0E(6<v~X&+8NAjC?6m~!Hr1sh7HuwSXrmGL0qsvY
za~kf6XG>aQ6*lzEtE|{yyxwMMZU?K|R}yL8p=T1ZL*?85KoL=%|G1*GjRQJt{;U9-
zqsr+04~@SYc&#JAeS*2`cE?=})Ss4#&2x+GQ-k<^pgVy4IFGal`-Y!TbZT+Sa8ljT
z`P43w?YV0jZjILf1B0JG59lhZYihQ;mVykbCFItDs9)p?bo|DHB)=KJjxLdQsR-i6
z4kJB`ICT6UFF?f=TB|2<1uyAt0mzON>+1@F`*0egjw@~nR9&kHb|e&E`JFTlb$_6a
zH$CP$J9|5@OFmDO*O(s8+6N%^2P1PSMiJT&2l=;T42B7o=Ivr}Op)beWap;>lfj!s
zTx3Pz1;}N0@4==$BLl_JH-O~xAD2Y?L#uYDpS`+PRH$00-q4;Ub9`8i(|6(4%@*_6
zvy-Fjx`jgBfajG>lfY+YT4~Z5Gf@3c+qZlO7RuBPTC+}+lB<W*$cUL{FyUh{^D+cO
zIQrbbQ>@YqyJfO+g?YII5BB+5w;=IpvX~J486hrmx;z=fISO6v({eM2++IlL8Mjr4
zN0tuIbB-a8bIwZsc30y!avL8ju{f%jnQ*JeKTJErN-gwcN3gDM>sj%VgXn#wIWF5z
zbIUz`oBZ3jmJO}tDvAl<EUu~pgkV?T#OAoJ0!!7j-M=QneZS`Ey1oI#;thv9of%^<
z=V5=Jb5}tA+efur<g4Ppe~M-Uh;;Mcovb$=jqCQdMQc-P9eHYe2)o`{-B<IKB5r?7
z>9Te{f%ANxGXIM~7N3EJ#_y59*q4u^Rd<{~yGpNZVz%So8+byMH}Hok=3GDS?SCja
z7AU6=#lCO>VtcBE!rP0O!l|4oT7G2*xw>p!jk0!awxi~Fdwg8LIK=ICKW1yflrKcM
zZ@w(F5OSB^cc8$%mSYPaNpST(CJgn1M?Fs6EUoOc%h8vjx^XGifW@6)n?x&bD}PA&
z^F;Qxw}g(nPpLFKM#@2Uh+gqnZ6r3d)I9QU9gp^D^KAnYYuhiyr;X>6@JZ&{Kdf~X
zmpQtzm-ehbJf@H+h?OCO2Yf%@Sbs_#Z+&O}Zo~%6$`bb--LKAN{rPy&g;CLyjf50^
z?8ocHlr(8%;BBES;4ibJ!2KYSD^?UyWRivVZC<8eDF}P%i;@SvK>m@HI?tKRFI_*6
z!a3s7az*aI0oyvNUBr%2VIz}}zK+k6a=4zlWQtV+R+DC6SI+Dc>lsokmFOKBMvxM9
zla?LfKH(^tATiTteoAbrK|9;T%m?}7T--}rG9`F?R)Yrr-Zz{2a)}-AY{+trOS<^K
z-hW(ixDAJ$KSJjBc0SDyPi4|Pxxg(s)<5_miZ?oEo#PE2@Spji-=4N?{b>%;MZgpF
zGQZN+2?KhZ`nQ%9#5G<8R4Syf0_t~R)=LdXEOAq{sm}}w9aAC3I7NYhPd2e+JfCT^
z<0gcLd>*qsm!$8`sDyW4&V>K6-2K;ilIpz)G`WVIPh+Vm?WY{53wE!S5{zfUiaaSY
z1W|k9wcAJ(!x(<Ifrt|*VfTh3v3$pez-5Z?U1%J<2s;6mcerGX*OAN0Mg1N~*^B@m
z2Dg5+iuLVMDa3!krkG8?iIjU&O>_`O$t#nK#u88CD_N|2bTgTr0@bkqLn>8<Apkf5
zT}Dug;+85!mB~Ke_~Gdu<GgF*Z@rSp%%${SJ=37(50-l_5XSX{HACInm=f!{x=o&6
zvtSRavw?buuT;X<twWlCj=~LdZP%yhZX04F%3@p!TZd+P@`-)>M3uMUpKjNDUI86{
zeqp8!a3(A*7!pL%Y?d#`?YL>Gqo@zfg2kqo4_y)L;r}!^<H3I=Gzbz`=S0{<oE0@u
zl~_8yVHQRAVqNE4{1B>Vn4J6s@Y@EB8O9=PwA|Gou5R+eB{N7X5b5Xh;{5&1o|bLC
z*4(v}dK_DfTFYR4K&d+xDGlt^tRYV0`;)-!Lh;HV+l*l>=7N2ro6*`SFZi{SvTSRa
z$Dxj$&B`k7B@0Kj^d!x(&s^?eS?L-Bb?dgE*`l;CpuM0?u*srAJJ&2I6{`Z-Y0dDv
zRBm|Rb{O8AnK^zQz1`r6!N+qby0J<l%gqR#WE4NeVObqI$_TpB)e{GM&!*Mr6<<Bo
zVntPS8dN&2G>yI3w07-$f<pZcCjS0c@&{jcH#w{bzue(8mag=ClVQeEc9XE|H~zi#
zc^c;hIV3-glNmE=Z#TFCx1_?}143F&ovis43*OE_&KFSXCItViF2BU|%1CQnl?dM_
zYp0H7dcNkSRy+b#A6s699gf?C37$%>lkz}5_Pc`KVgBXfkMU=9fGx00HA+!O2?nq1
zGPXQ4M_)WTu?H`!?8$pL;x?ifd|`Y*lg61(m!!?X{)rdptTnV3ZLfC+nB%J1d}ib*
zRSGXP0-`L&X8_4WL6`>@iB1ihKbay|_;3*!jOouovxoZ-IlY8#%Omi3+A;?(%x1aA
zdiRt-G+Z@goPxR?K$_I<dRrINvk8z1u0uhHH`jlmoR-I9JZpk)kwc7ocNaKmv`=ED
zrp7dolBR3hBtEFP3O29WzT00{8Z-!QlXq++bNyQRqTD3g?%v2?5+dHWUt?`%=r!Q|
z2YP;TBy?5NTWVmyMAjTsi%9Q^(3|_-^f@f%QEAP`<=iOvmx+B@Db=A&I$pEfz{pSA
zY!VUccsxr|8kavv4aAD72_8S^*_WwqCwu}cp{((alEvX@t`ZkvItX(+)iZYz;MKG*
zA+$Y2(7MpH{>!3ct5HAM_T^b(e0f+*OqG?uQ8@?Z<435^`_p+0Li?u`$F5J|iY&tB
zM5ib`(6Vx%f7V(O0@>mZObW%_Q>?~;m_N4K;4Gn|4MgwRqFEMyC)@D7n;L8zW)a?9
zSB$+L`=4=5`^)BIrT&J``B%9{bkH0UhPB|NI->Z;YQKa!6uNlcY_Y@COv!H#Ha<a2
zfHbEa5xep&+wBRN8Ml@MIn+EZ)~@(djQ*<Q&II{43s*TMR$lyc>$-K0r_1~v6|Woj
z1@m7wz|kR=A3?${E@g>NjTm<zbxqhYOCJ$WrDi%j+rB+!UE@z5*>@@7{edvC?bjMI
zc3H^9&hEDP_qiJK(254v`zEu0V5uz#bL>yJ`sLare9J1M5m$-~!w$n6UC1VbD}0l~
zUdX{F2h}i?@3>x9(!kvjFXqqEvS0@s$=fi;w(A?X;`0rlc5F2GLGK5J(km4FH(Qf}
z7TeXzo0r#7%2d(l;Lue|n&x_3ey<^WvVPvCZy@6?pF3JsO_M>%0N06$r1tHzb>2-U
zq#AMSlik&<f3BO5uaX(DjMgUIlrFF_aJ`<J5PT*rHK#MTb&e<{mI6>A1N-{yI}X;f
zh-TCLDci*KX+;7BuakL)wAHo+VE%Ew;sjA3nO{yPTwbeW{vP>ElPexawK*`q=OxKR
zcj3s&rmPJ0@1+D&pomAqR_A%g@+{s@r=I(R6-|%sDuA@_di}Fh^(0zbdp*{?p||>I
z-j5L5b+>(x1Dw~6{p0c_9GZ4-M#REWk_%#U1PPRG)Ucf&6Wa-*t2U>v6;FS(D)XS7
zO}oOoP2P%6rvLS%J!yY28H^<Zc+;`{=5?*iRv+JJ*nXpgI8__xP9)<SgAX|}454aP
z?XEfiHwFJQiLBb*w8zQ*`q-VvAfQ+1wVj7*Z`YTB-)OwUe+Sc&)Hwg^TLPNK*;Uf>
zCDoD*VN4cTUw``%4tItNQN592x%hq={OjiMf)r<w+^=x9?xe5QOdYbI!G%DLw?v?&
zDB25};EYgggJ5*+3ytK9q2BL`Kbj_LrQatiSbl(xf$VrbUr+A0SkhwxJNge1>TNpb
zoaxO!O^+LLk;<nqX{N4Iun7L0?0tP$pt4V8lhBI|<auT279SUWBy$^p*Vr%$tg_cq
z@q4-(9!<QdexJs9&;)n0Rt*>mx}R@<cy2rcYEqZmnG|uEQh%bVqWcPjov?sHY|Wl5
zbi}@3GQy;Ne8QNfS+BIz!V7If98!%>#8Ju+r<kAWMbNk3IH1#QFo%X6kD^8Gm}jJE
zCC_unZ>u(_)Rb!%eH{px#?1R-^CxdaVCWOY8)FYQqW8+ctqAu3z16zQmf9^~ZrSCD
z2_JVOyJYW4%~xQKyMGrpJ19-Ki<C0u4d48n=Dyo=yc2JmJwCm_mH41<axhOardQxS
zc;*Qhw{x6ModJkF7ZkrnFhnDg(Rp74=ePsNLWc(=geY0SPg9@ntGAzbzShfCsvhLo
z;l3+I*TO_xTf_=$S=q2Q_7xNURh0|~wsy})|G%&7PdAV~@^Jp@TXaOrv#;PU>w_|w
zccY4c*se$nIOPEb*7)!IsoqwP-ZV2!!L+4Oag~Brti<-C!kW+i`GhXy3mM;N#`YUY
z89!!J6Lzf#DIAwh2}<wsrum@8Ov9lV#9*x>kf4|6S<l+ths0!8j%@Wcvh{J9tR~GM
zASEQK)Gg&2?wgmNd^qeD;EAs8eE34kWStzGc1xlK^MO>6zgR@)$#%90E>DL*_u0$v
zDHTt`_lV~9DFEm(OkB<jk7jOkJ3NF=e_&4&rTdgk8}jgKd^M?U^uA_1WlT>ni=54H
zC#mYNx)D~tg#y)oJMPnaKR2<?#U7=1?18u$VKxL(U@kNuWM{O+PMs%%nCt?Vt|&pE
zuTFD9AjSUF#c)Ih5*1g-Y7w4aq>_53n0X+={#wRVEmz1f%Q7fmKR2)Mz$5*xqEyRg
z*1l1Q?4t>lmf0Vq%n!d=dt$$R>+F}*t)S)0Ct>kx_v7JXjZ-acO72}%0fA?evdc+W
zrb$|tiePJIaSJzcsjBE_#P+gKRXX|@)EmRzg421<UwE`DFya4*JlrJ;UBm=ma^EV?
zjRYMCQZb%7ey4armBn|P8<>cVjfO{bN=DB2(%4^vQY6r8W0k|~eLCSPZq!d-0lsuZ
zBf$XV|Cv=iFnqh*Puk8qjH_lBTS|%xRigP<;MS$$r@deY`!H9$mEu*;f8zRjyCPI?
zj-W8^LyE#+w}225Qt!nr#4`~pIs(d_p1BW9A6CIpYb}}625G?JnPl-{#c@6bB9xGY
zs+Eh~#OU@H`dSHM_up#m&s<t_p1=L-51qJ5ccWW?0F*|jyhh%o583e$U-Pw|$Io11
z+gon51d#*Zg|A8=u8xiJzTI7++D6Ux9u<Lt*nj3P-uZt?ddNA?&Hbcs9!hCp8%WF4
zw{s+t#W@q(@ncrm+C{exfkLk#lki%Ff>gOM!A7=;7k7webx5Q60@z?>z<*L#n)m?t
zoxck`q*>=+?XhPTxk8(Znp+|}jfRKOqT@&dd|jhsmd;%~BslfYrW8+GMzQKaW!&yO
zmWvf)^`Hc{6lH#LkWANORX`1(^5{@{uT(aRWi17}R!t+;fTAVs6*flZniYvmv;i<0
zL8uP7tl$6z2jNuB6oDhzGQ)JWLBPw%6p=kH#urIR@wD}=c;WLa)-X<dXi(BKx1=v0
zXg-c9;>}PD-MFmC7)cMUM_$Aex&OP-_N;GlFzyKaY5V?+vnX`taoC5(yaZYe!z4$u
zCTUwVjm~jls&AVFQgd!Mr(fk%2Gu)XUmn+vNrCw-t=oqIKY;zd5UME_3e=BnTrS{U
zNXn#Qx%(`IFwA=F(<9D~eK4@7PSp1}){(rE9Pw4GYV++&2gD4Ra`*<Bw?K#wS@T;E
z!VvmfL--|r)ce<pmzmrMu;Duk9!jh4<?uQHj0%|^7Yg%SM6QyI&^2@H{$yrk_S8it
zjp}QSP(w$g0<vsQW|uQo$y4fCTMJCb{v(s5ji5GS-oD6vo66(-8}wbU&g8hT^OF@R
zy;Hm-9yI2X+)<zy2tJBhICfL;Osy|HMGDD}QVW&<$7)+1<hagOXz-^d{>hVRKl$v-
znEiEzsi9YVYV1U4|EF*KqvO!YXo|5{PpJA$#_WGwyL4=Z55&HtUq$;1OtP4C0=+2>
z=E<xE-fvXK$MO<-^0kTm&a-TiKBN5Ofm3<60zW?|z?6bef58$+U#FG^Tp;~wL@Es?
zRn)`MWd0HWRNeswq{>QLOK6F>%CFyMH|kd%e~Bi^4As>!iJVsv0ecedU(_rn;Zb^w
zzYd;_eQ0}hOf{%GpY0%E_R3l&yj7b~Eel<q4b+J)A3%=<UrF6r>-Y^BhRu*<KbBIp
z3tiHQIt;Kc&$aTCB5hp?sdqhJFKlSo(R+T4*-~GjdR<&9J2ivoIppT7C<{a4<YH7o
zWUoaLWhyc}pBt7}1G(3cAwfc->wx=0=oxfceeh2@I}Z}t5BdRk#(KXCSIrZHTXKP?
zTZ8dL4JSaKikz*Oa$1Ly>OdazbFxd$eS@GSWG7}=&e?=M;zZ_P%qI|*dUcdlOSY(f
zONO)}uS~l?h#m002LTgi&&mN$IFgNMQZ2BX4lN{6xw{@5-Kfo?NK)HI$bTkRP2hRY
zFyjf!tUW@NT-QIeCU+Pxr!dx7--=Tv*RCu^Y1-xJ{Z*jAt#4Vfp=@=N^7u|Z{7gNA
zZS*lmf#E*k&2(Zt@J^km(a%NP?Y21xx42J^S??S*kBVL7rqB<RD@K_ntX975Y<v7v
zaeKP$`RHdpEMNUw(Avsa*%A1*e%DXv*1CXvBh+h91`vYr#sI!Cga7-29y=D;bh%KD
zY|g5Fid|&~#WH`74GP+O!770)`j_f?ehOxxkX%qe{HtHZ4Zuzj$A?G7RTte=`N+qb
zeiuy@MExec08~<XnPq8fv>VWcfwXScy7}CiN1NnYJ^w}&z%{b_Eep%1c2cGVJH}hW
zdSIDXVSU>^M#Es~a|+^5P0MbBx1C-Wr<1r2W9(k>f`-MwJu0Ms8Jk=11#ailAdK%o
z$XJm^X>xglzQJmv-R}w7eC!hL43C&G@|FO9;Ggq8ctJ**8=2E9SaKer7-Dy9l&*xD
zxxtF?&exQ{l(}C6Hov8{7a#=<dQ;=L@n1Ik_0NCAb%2YqX_A5`WF}T#F(quna8rK&
zcT1;Ehq-`OA+ID`%wzQubsdZX{pfLTnR@F;_<J~X^{(}V2OUO;_xBlPXSR&IeXkQR
zT2&C;L5Dw*`mliU?%i9HKGm9wCKh8o#_UF&dD=J;ZbB%EV?iM2%KjKJE{>afJS0$+
z6zY~NXT8L&2Rl(Xtt)U)Uth;ZlRonFQc-{2_1dGrPdDykUAz<DAaVOwT=ovZ`X%@?
z^YGYpjXr%iW64ioH0%8Spu?~oh{C2!kE9uh5&%{!&C_;|uYFDX%z`URe~-NpP9M4E
zu+_^s!76;VtN!r<82yQU`}S@B6yu$Q6~ZciM5gN~qdFA`S1QvaA6(K3Wj%}{Ms*)Y
z7*!nGr0w_`3~K%`S$jde_8p~2`$RF}mSjMTUYD=qAnVyDC8WKGJNu{(tDGnMuRgQ(
z+C>OYMiz?qDz9GPfU02qx`{ygj&o1}rj8`u)RPkrQ7`<#W|jNf<2wh8Uay85#WA;U
zyXs7>d_oNQzsx%3RfJ%CPPj}pRBF&BR)F<rmt6v^SkdP})p5W9Xax=cBTl^a+Gk|`
z@*TzeR@)PJm>UI@ONrxC&hgggrWMmkb~<2CY|`G9@gS%oPwwxVuM89=8J9nVer^0K
z0vFg1^8e+VHoJ!4^aJj;)^67Un>+PiG-)Bvs+%ap1O{Yjf=^h{5BkpFHp#h87c5v1
zJ>ZR}Ohk1(_${>C4d>W#?U7bOPlngT&WgD$i#!ePx0TC7+ANM1{X(%@hAn^@H6jN?
z3<_6m>WV1&)^;FDzpVQ9cb^DOX~S=McIdWs-37NHE?HzAw_a?aS;Fq*X_Q~<Ah}OI
zAQNnK6jz>Y46VD9#=Ce64>d%LCI5nV6}Du|k-#;Z&E44>nLCN)+on%?c<d9Gw$%F2
zn}wbWpW-&JI02hlUD8jXL;VVaEZLfuq+hccx#RC?4?G)5vVQ*<M_%mEbCcBIxgXxl
zyOxdZ`@XR&>2<m2h?O=)$kzi#2f>{P$Gihrd0*5K>zF335qj<GZGl86i)g%_p;5Lj
ze#_}XzhWp^B#p@=4nynuF*Tgs)L2zFGgB_*nJwACn)C~(9!^96|74qnagM(Yo}jxo
z<suSUSU2#3p_3^vz@m$38aD+}tEK7r<{1CeOCLrVQITPLszL3#3Q3ukq%~)gG^b9N
zMk$bOcHp!5xKmz@xjrz+M5gb~zt=<aY>*GKmu5)1>aVlPnB%u?RwAaeEQ8TOrh!p_
z?{(%tEybwQJl<Gsbw7dLBw#>1o5TGc|I0bwv0ZmpcPU}!!0DNHZ9=GEEk@^n#X|N-
zLYPq&d%mm_Y&gkFPE2sHrH!lbOH~~Y%YH>rzXdW+5ABm!xroH|X&8XJQ`2J5F7v2~
z`GV5Yz@N1=&@2{%i&>!GxhOeUAxgjPf^Tb4iJME;PO|Rq60Q>~h2v3cob|)ebHNY7
z+jzL#_>3XQE`l$eXES*E*#*pM?8dw<B~2dTQAHIg)c>`!&{Dvgm+<+IB!Mvi*1je6
zB0PCyoVK*=c*>0|*JkL)jI9lnT($zf!12_uuUOUbzAh`58v9$}?}Y&WF@WrVaX$BP
z7FYFjW?!^ugcGVZu_G?fCmvEKq}bRf`D%pz5>^ts<Z=-A_>l{%KRxPBj5tG`Rg(<8
zyb!NDq$*(NS7`LRdc)HBbI8HEtzzR&T|VtE%l;9UX4tYOUc@AZymb}#`q1ND0rfTp
z&`q5|rRh_JZ7azIGCtQ#-xgRox`<;sKAf!c^VG~DI$qJ+ok9~y@KG-=%r9x>*=>QE
zQs=aLCU?z1k;=a>u@4SCIr7QN|8yh7PLWR1!Ijl9ou4qi7;z7=tKonhr&w?n*=s_G
z-!vd%+UvDOuJPrXXX8}M_&(03xpjdD`ro_!m5luErtO^D!Jj0`YuM3D%j1)k)VO%o
z{_s0K@lmmm+g)F*y3qQavrZdDpV>3=e4u{!O^y5J74FJ2Ff@NTdw{s^vj*T`)kMA$
z$QKIsvn)80@}K<WJOE6*0F-`LaM9_iTiu9=yOAR3$&0FQ^E++RFbhRV*ovLhwfY@S
zB2j6yYsf!pNLl?5*FHfUh>UZeoQ-zP8HOY&;Gu{7`bcB^O#)}ZP{8M?e3ahuYT9y^
zplEm37g#idS_c6?H0&rHF);A3<HBdEX?a*EbdP{zvw)l4O{(bgoXSeQl6LQLthg2{
zZq%B_4we7osg4LFsHdB(d4tcq>j-dXa2A~dZk}hCPMK{|EpD%+uH2$b7MRRhFIK{y
zEp^d$uhk2kD~HA|^3j=r=a8OG`5s&9^<yx&)}Z6=S2S>P99v$#q>ifEp2aK1LJntf
zeRWGgQc~QI$l>}KX(^uQ9y|dAO;3Gmc(ZmnMc*$|oHMk_eEUNM7C*bUKgtoY3V1Cu
zfp`RA-`Wn$I}3&!E=3r;VUI<)sd@Wd|7jPeM_(|)Y)66fp5s^C1~lmWJFqq#-TEl{
zyS9v5O!L^o7`p&}zB%aWb<eh-%2taFd<JcUHa*_vsa2jx>0h9$AY;u|c`)^6VL)!X
zR}R^>{`b{`$nR`cnlOZ)O4KS>LM>d$?<ZjMbVl-*J)S%UQHX%2Rbx$*W7>oOGdqjp
zeEsp?r0VkPo2K8H1~kO-14L2wYC*}b{yU1mXK0=6PiwPu%&KM4`@+6fwgJU<{E8Jl
zBVg{V7=1U<+Iv%yo~7AB5=#knhS}nV`~bf2_%wRXBCs=?`EWymwDS>kdSOb~A>+5m
z&EgL5g6ecYkn^f339d@Gvvi(LRf_^qu0lU{Q#clp?D(!kb%=aB8dK3vQf^>k2LTr&
zcLFb$NQZ-Fseb~36Ao*7^TrZDm4HC-J=5#G_afFmd@Fz1Pg|YvW1m7_W4FM5nDsgy
z?ZHD_(LwyaWC3EL&ZAFNkhwKoK56@%P)dJQY-D^Nm95=tT(>LrPpEcEC-+H$Zmq<$
zQX(<pI>)K3vY1QKy*<cDTyptlxO+D0ZrVPt)qr&CfN>E`ONGZFapim&ELUrAX5979
z3_+F2<MTz7>Fztn>zXU(Y(+gzJI)w8S*kkdD?hjRbF(IYcPyHgEq&!1@_K{HfX6dW
zpUwK6Xtpj!P1|d!dJqRA>%w3u&Bbkr?b`)7{K{vEA|XMn3)EGQ69?eq>JmAzZwnwT
zb8jF@8$yX=nclp{@B!viJrg&4|F6mZpK}(BVZYJ&Tt6Ya;Fc)SMI=j$%6leUWJ=n2
zioFd#3I8EE#cAD<pQ@m$pY^bgM)tCS3D@W(>uP@l)I8qd03}VDEyIk7kj)~ErNv<|
zt;iwxkm_<z;@C1=rJ9~JT2I!6xjz17luo)~@rKKg{;(3a#rYK48UeE&W;_;kvJa7u
z;~!39IUp?^7%8Fi?v&_^sz4VoYy8!8CMU#B*aDoh>NXD#?&3lLvUSN6Pz8F>Wy9<e
zW#;Xd?ux-5I%zW2P77Rw7NKrZR`Cvj6zSH*5z<NGr6-9960(a+D(ozK1}P<hsf%YF
z5SOuymxpGCL>0E$L?vDz8qoNsa0Y~-APL<dzi8&HcGX^)VbBByHA1c8WPgRXYkB+c
z{;zkoELYf|9`cE?0l*lt*!}nF&fuNWxWKJ4@G%28fQB9^L=LS-TP?53HrQXj(CRA#
zZ*f$EtfgELIw*r5O@P6#rGh6z)WYHU@JYdH5;0Gs(!55`@i<5Pq~!-h{ztb<<xR=D
zt$HbrQ!2LH9Dhix>H?S~pEqr;4%m;;H~<(8PJxG<t@8=HE2gYyj#!b$2Dap5o0SHM
zB8;>UCG9Mo3|ghp7aUBzgx%=zu8Y#p=+hU>i)se2=;UW&_i{L*mRMY%qWb_Qr?%A~
zfau?2y+->l(ZBi1gW>zgA}7F>%w*KY1o^EIJ^|I?NX`oy`afO(ILID*TD(QlTT=OM
zB!<H2B^7F^Sl*E%g>mPoB+bGs-@CkULtw>$BeOAH6m=gNq$Rrb-ETdzuBb(S3fyG=
zt@jHYAo}i=LMzI6)hn7!I@C~el?25yzLDUOV89&xN3zOCJ-uS#7V$jf+lJL%@Tk&G
z1ll+%nW)^z*249aG=3ompER8l-lDN!@;D!OXSuu@umW^khGn*JFZH<5B6bT9MhfJ(
z)!<*q@Kf!GFflL{cNFHoL-+-o%ese)<zviF+4i}&^KbG3Fr6-~VDmDV&bd&btIe#9
z#m~WOvAlc0p<M;ys-Lq0MvRh|#AeClrSE1Wj(VQhsRU5JCzYesHMQ$$f3O1cAPkrq
z5{khult~d&JH~Se@EFr?JWf0U{^wj?wI4|&WmXG*Bl)2xGI6wYC{eo*YU^18Rx2iL
z6<I;R1g#g7d$`fhz7lJ5rpd&PKF{uOSMXi1$!w39a2KUf56)}nETJ2MbFHQCD(0^=
z4$^mk!X@xikW}8>w*5f1KN!yrY(BBYc{(k60%p{Ce;8pgzahrz2p{ajePGrVk~Is@
zUjnUoHVSdbc`#A16nH=@ZvaEtc9SU}4!@p_E(VbQkF*oUc)FWWKJR!HtJTQM2ZWk0
z)wp)?UX-lN_bGKK;@_qsxsD_lckudYrp^0~R)rG3va$J{<vwF9wmG_R<@`mFVOH3V
z`8L8>H<ru@s#e~`cI++K%Kglpdd`}yGrlqqV;ZRfk0+nlWhF&izKlSY3LDvMD#xFg
zm>e89SJ8I-H&65~S=(5e<}D|>L1XLt6jvGwYMaMsy6J>B*6UBVnHVB>aX+iup|*LK
zpS?M<lmE~N<nT$WLkx08(1!PM4I)@`qFYPbKEF*#M77s~rR1K7n%*e|0yOs_>Pq|&
z0Q>{`Tk2TrJ5R5C;n+2U{IKb=`B??VFZ)*D-VD?X12wJ|dE{%UM0x)zBHCn51w~TF
zU`6e-;0*BcO`X5XmU^jMSfO2*F2$kLD7uDS1~LEylu?p1&Prj%k?E2dbY~PJBomuS
zaWLVxz+(+8QB}Ec=qsGWeNC2@U5Nt7VFkdQrI^lj1duEggR+n!C5hQf-b{F<<FED^
z<{cZOa~XlYB!6oI(S6@(SA!d?phWG@_XOsbXDD*13&Uu3t;k4Oq=-bXl(B>Kye)>>
zpt85*r<>!u)b?Q8w|Ea)pqNbJNtH&IMpofpc3ju*{hyo({1pK_N5DVBv-iJCor(PJ
zKGj$l;CpQi8>s^DsbkfQ=k7Q{BD!gyEVh_5<<@vaQ#qDvh1rQSILJ!4*&4aKL3x{i
zp4B6izUPZ&J_|6l6$Xtd(;=gP@g30cU4tkOoe=tnz<F2-ODwFb2&hJv+&Q1!u@bR|
zzh$w7QMoBQvMV8L1-G)46((I;a+h1U-m^mbRD|0<Lq3xxd)GfoGa!?jH+eM!E2*Bt
zdUONY+yM5!HhRwW{Y-5aaSopHPs`AKk--3l@_8?byC8OOXZCLy++kAB@Y&o=Ugp$M
zRBV3OytI}?gF*56kMK|N1KaC*NncCD-~0bfD~JV%`D||QV-}u`gcO%u>*j*hSv2fV
z8I@SvyjIpB8K|~t)9+Xz!DRwme>qkPJXCq5SAkv|viS<k)ci9U-9Ngv=u;v(|GgU-
ztp+L^4STFh&KX~aC?M6>LAJr7xXVp=jkI~HKj-6-M61L{<|X(IR|yeo3u$RGOHg^p
z&g@HC*X%BSMa#4Do(mctm|uTM6si4vqA3LpQDkNR9SgQC<QM{$NANAqmxj<U8{2UK
zy>qzG1@hJXf$`>9a=AN{C}j?@t$-1RJe>KIKsv9Ir+wYlvF?#1NC+3@=hDL1ID<Fz
zqk83ht{nB`ik*VJvP#Jn)aU$MJmBH~Jpr!&F$e_+Kem?=P2EYnZe>OHsov!PlZ$Y9
z5!xQm&$yv|!0c<VRF$YXSI&(DcPx+8PwzqXBlGY4RwIS7w*X7$3(t&!9{%Dx^#cHX
z=Pl1lH&rIom{Drf)4)8WV7%HD&2(*2DC?yMv#c58Y?Yu$%DmFjdCO^&2`|$Dj!n`e
z*pq*zcEk&bAX}hiXA<zP%9-Z>CIMB*LHOJ9=3@qCaI5`!v9(dmb@5<j04O2Ycoju6
z?c!fi+W6){ZKZ&9hA91a7!ZY^>j~YS-%=mIYEI9xPDlMURhY@nYsrEcMIUNLTaL7D
zVT<CPK0MN+^0%t(#Tn3wKaOoZAs(ooC2eKqV81h3W0sPRU7Thovjw~NrVx;gxd$1R
zH@=PTb7NDiA<0mp{e64uc()cWT@Bb2?}L0GU#^z{*JZ|PoA&FM!t=$SksM~4#NSO#
zz)HNp9ujU4oV2vv^~ZeBRh6ie?$S=4O)5o~PL&pkVjZ>E92w=k7hLO9kqO_NROVVS
zxM}DQnF<9?jw$8=hm$JPd@0Y7l1+-=!(DgOEqB1`&kwq7%IUyOk8)9Fns^n%?f8)(
zPT6+Oa*~b~P176)%>@5tM%0FOr;B%$G890w3jh%f{VkI4wa|cedKPld=Z*OWr~Dr~
z=V1wkU_j1{7z0-SOU`v73VQuV!S%Zaz0_b@JgL2(`4q$s{oEgmDpeB48*B|}ufRRk
zna66=;TkT2prnX;kvMm`bCnXKgdo2C%Co_Ba7guCuNnNc0=E<oFokFvyN}gMlTnF&
zKF!zqrjg{3ImQ6Q^%MYDU)lkSVLOXFJb0G2(!gyQFA4pYgYIC<Qgo3}HJ49)(mfSF
zClR$y?Hdh$hPkU}Kl*@r)8NM-6$=v+CtysDz{u$Xlto(UoYm!V;7YScrb44swREk}
zGRS@*%V5%jL?5-KyddmnrPYpkm$sngXc0Pq+-Zv?D*qI)=A=^F=r;pGzMb_sr==}G
z*Z^<x`DvSA>oJWf;-du~mhdH2I<7g+0xx-k*>@@pig9$~rKJy?>m6p=_A{SjW{(-A
z(C^Kbz#lw(=Z-<IEsyW5><U43ksT*i5<IQcfOu>Rv<8rP*~?ZWF<@5&yR6`peO?LB
z{I6z(nW&e@dz7>6CCM-cEq_^Bc-uP-3L9v*oJl27u0<KLVb){To}=BShO4K@y>gBI
z|EPKkt~$1@3lMjQ;O?$Lf<xdya0~A4?iSqL-60%Yf_rcX7Tn!E1h=mH?svOKQ@`L0
zYVVqB+R}mm1mwG2MG>!uTDO*pe64Fc-Rp~?cT*`LdEU`ExFx|cr!AWI<U@wENf<JL
zUxVTb47zUnUzyZvu#$fp1@o2WGz3JqfX{)A8Hejnks?#!YrZ`X#s4xjS3=WG@ZHb$
z<3K8o)vBfIzz|+sSpbS<?m>IfVjt?(j4aNZR>Z{%!?0<ERYq$+F`TRez*TVJP{6B!
z{Ohb>1i#J$FCRrJS*9?8s(yX@Tw1U%j1}CarE5>qfJZpyX%R<SuC+F<o2O7TWQ`@m
z-2!`6==e)@8b58w^{6)pZZ}SxGZ$9oQ_UaM=)8ZJl{ld%kCb~D(OesNbRdvCCdP%<
zdGvw|yidxNgM@C5R2k7GbB<Ac^AdI~UbnwT>N}lk5+nNzB_{D)WCBqt6xT?mWH4e9
z+WrEiN6lVLt+=h!b^!KtyAH(zA|uH6NZMWAEYFY=s9>jgwTXao5_-?UUjVq0ae?)!
zFwAGdN4SsT3~y@$Xw6vHmXCndYwl|wBute!BE5}<6zmV<AWLi9Zd*cxzKSebgd1nN
zA_PWS3Y(RM_8Ypi%IZXU`;o))W8RIc`(QM_tx!B#&MOAkyWh6)+!M~w@*L;HT2Qf?
zo2NPYTAmar-t|>Hw?I18?Ul^t9vkpw9A0IZ`wUnv-%Z1DyE+=Cy6ey^H^-Yu<_VyY
zjHJNSnR^h<PrLet?P$gs{aDE|LSE-id1g-&Gi(N}`3#kop||Prk7@i{<N^fRrbr*&
zo~yGry`_3}CSOYq{C(q5;i!_3fi=7XE{~KKog`;b={mZSzb>xsE%KIu>teYyy`cq$
zD$o$7{5!1mTy>5vJr2`cJydDGeo@)Z7j#TRhRVIYPK|CR&BbeKnmidMl$JKPlmu9+
zi_nx5Dh`ShUwiEO@3!rGZh%vORi_)O2@uU5m>KqA=D?v_f8M-q=<Y9W?}_LhMycvU
zQavx_IxmGWg{VeN<t0AmxsRqoJi|~h;W-+(UiZNqB|~}i*JODjA0Gh}4W7Q|@n2QN
z;a}G2lUW~-hPP(dtm*BZjZm#i3R;Ga2CCAOn+qMBk(g|}$1|habj|2?8oomBQ~O*P
zcNy<z0bc9^y2i51AUQpEtGV$k%aM}I>b}D5Rr@YfH_IQ(j4_MZ{L8fMh3+M-!Gs#q
z@^ijK_K>2Xm;!|=HW1e0uuv@M;W!w4w*wsd%o=2!-`?iF_yJk<*7cg!-PkgYozi+;
z)IpyapGs?fUV&U~NiLnT<qe<{Bk#Tc%ugJCgA{#*_v3n8Nc@_$-uzmQ07O_VuUm&L
zR9b$mN_BFGuda;Xl3XP>rNY-beXD9Uu-$t5Y2}XJ$oXs6jm5u2I-saY0oZ&u9#|%y
zr$V0y$=_Uz^;kcd3VW1WNft{NU0N~jDWG{|Y_hGiHz{!8y*-#dOQ7Yh(Q?L_fY$L-
zoI0&{X|&=aYkQSW+|aD)R#t<R2LZyM^J9_<z4xHDZo!})4Sr;C*eYU57jZNl*xLF6
z$6X1?q=|BjA&cVNzuY;urnfF8BX@cisv%14e$VqA15jV`#OiXFGF7>;V5--3Y3bYp
zK;}gmZrRDoZFxrZt)x)Xu#AFJEQ-X~CEr*yCH~Ub!}CqU_tm#}fs)QOFQ&!f{YYV$
zIdBj}{)_|_ql+{E06m`p=%H=nPENH!#qkJ@x4nmoBX-Ax8k8+C|4FF5h%L1Guofp&
zEa)<xvU^~l|1{REdM^w(E*hOlQ?7$sM{~<Ymg+-+k+%*o<*?zs6QDT5PGle;CLvWf
zyIdt_OH3=dr5!~50Ky+Ov9yEnLZoB|)P|}qkPf>}O~=i<yc_4eTkNd7<cdh;H0835
z+$SSokUT?L<v+Lk<6aeD9R9XZIymJ1{`GyzBERoMjZvbMXJMWghrUn^yq)6QfzJ)D
zwJzswZ*^Xcl2nBGRcIr$#GpY`DD8TYqXR}l{yKZ@u&Ct58vCgkNSXjB99ux+CC!P<
z`5pK2(dAzi9S9(1!~3r^pTD9D0|KT#!-U>%;l%u1>atiGmGa^wWMh_EqUYyW-ih`-
zhrM>(OKr6@<Y*x3eAFMGBXF;N!rG1kB3|WgYy>t3hWTobEzcX8AlFMUUUdEZgp(<K
zg%(oOpyIf!_Nx;5n)U{W3Kb{-u9(@0nvsqLhroyVB|?XeP7iB-pGb8S&_(ur@7_ii
zTAtGuV8X0I^PtX53h2(R3<%&g4rWprh6z;SbwJh^9~&!w*3iJYSj)#6+(V|hRt|l(
z-~RMTk;Z@=O3-5h`3f2h1@5wQw}#+Z{vVs<LjXa%P<z1BT$qDXeSoaRD2`LTuvbf=
zTb-r`2E&S`J_mRx^<2hX$F5x>CzmvGFrII(?t8v}TGv;NY-U1=`LmIwWUFJqDYH^%
z1o5Ph+*rp<n^p;yk7#dn8Jw=12eGH-h7u<6KBDGF?N{tOD%(TB;?(<b3T1atwv_Z7
z7gf50Rk(&e_Z=w}&{$*0vsPQFH+0!29YdEujkM2TCHr)}Z}?0xdwbt;FTS0#ee3dF
zD3vwv2AgFGxSw6!k)7oVyuaoncFOP5_L{?p**1L1&aPPp5y2MqacY!v4@obcITOF}
z-Jk=!P-ATeT3LFv@U5d(r3bNdAD{n3v%}`aysZ6S;It3PcE|l&tprq*8l%4e>tV=z
zAC&8LX?8c<`CyvbCevX&+FG1DFZ#egs0$zujZtcx1Ru^Cr*_d__Z@1C=MN^k+yH2P
z_<Agu#f4J;Z55ZR(fvxA4~s5yh@pwS0_h-!VE~(<R`4AEp_GfW3Fd;BmB|$>2tJq(
z^t7kU?9=?1+CaY+g!#KnbP}<{B^i6>F<M2sY`9@zbc+RiK%ZQAfM=lnG(v?840TlA
zmz(Sv$8vsnY$3a&iTF`dIx$B%o8q>Ro2jYBhHIJ2IXk3d+80=EmY*#&<@&igEt&JM
zI?HqHboO^JnFuHn)7q(c3|4!tFL^gU+WkoZjON(!LjK200R1`>*HGgTP*i5l=YJn8
zKZln!Tv5uy-7G_FZY0LbzfgLyKu9d?GCfnJZRrHpYTw*Sp~{*lpg9C3Uef8U$H72?
z55)Tz1y-PnC=Fu{d9nykU&>Tl97Am7vkZwX&XBJ}PVh=^^aci3J;V2{tCi^AAB&>H
zgDE`qF{Vv8)i#(9366pbj<GxKQ?-L(Dhq-Eq3#{%5^#V2@&@4O%l6$#0RNwy9@D&g
z43?%Fsk?Bf(@g+{Sa?d`3y#ayts>m2rfb+OXwI%>uxX^^2Fk&Qguxdd4$JmSt-FMb
zeJVaEmasj{dd+A;oi%Q)F^X(=#%}uFup9^K(F5f<|KG3<q|@_+0_yJJ$v((Zo~^?s
zP!(bxGKpYJntsqaK9X67tx){ZCUDt$x?fuBl#Y)S0%Sd&$12Bx4`&RuN!ZP{q3OiR
zt^1SvZsCZga9b`^OtJ99ZaH7th%2Na#;PtzO^GCqf|7T^8Bb(Zi9AO}cH0u`4#*};
z88h(wmSr?t25OkU7m!hYPZ|4q!7?lcc?)6x%FT|`xqBz@%!?8P1_*$v9zE}qFW+tP
z@x#Q@#mRA)aK?36Kl0+X4kLXk-OA`9iy6>s4pl6nX4EP^rM@%dQ)gTxZiF}^B_?4F
zDXQ(M-+6JF>Sy15xtSau?Y&yIqpO20(wwz)A}2N*uxBkN5G3Za|280>)?l7t%fmei
zi)!Npo3kp%I`9+yLPAheW2UdqB%Y;r{*Rh>vmxaeG@W)qY9;D0G{SE}f(srL4U)yE
zee0&0r21Ex>d7Jx=MT({V$RVKdBMYm7k~$R9f<9B6vq{*FpkX=rep<kA_Q5or%SEd
z#?~>fncGMNw0K;sdA&B3TP=X;@_u;lS9n*uSIy79{g!0~N-7M1MW)V=RkF3+J4n3m
zxo>|r-}@|Z%`2;|n{Vf@Tq`P)KW(r5fvn^gHQBU?=IRU<lYc3jw~o~=wrYW~MI>Xf
z`h|!<uT<H2JK13`IGY;>!~t)zMaKV|(cJ6*=~}bMaZwoT>T-VsQCtVDCh%ELm*O-y
z7b17YX4FtMr?%+Y0t<Q<ZfpO!ShtsU7puv(PU?a$u6cd|X8KT-Qe^vPN9qes+^F3l
zCA!)xXCy-rcX639l`nGzkwjstTEmX%(RAB2qP^Jv5U&!@yi9L5`S-WCe@We^DWHSJ
zI>YhOIy-?x1@NIx%7&*{I#eRJg?(3Ty>4A}_U^NvK~HB}nJ1JmW=Lm)uob4+XUA(#
z`6}pU0FPXb-8yJW8cB!4mU6HJbz2A2vaCf8lK?Jj0XLP+*u~8ZIoYew38Xn*-cs_5
zEf!Iw5)$0MNR;J!5(yMrfYnal;4aK_iai>qz^{d_GnS=FYgNRo_RH+|b8HkT>1<ZB
zep?&<!?|83e$aDVIAIkDH0%<QzJUKJZC0o#Et}j#&7bL)aDDIN4}XXYRVoG2ig{6!
zO=*YCV-+|*IkQw8bOPFmg3mvD-2g6QO7Q&0W%u(%alA19vFG1#$Hj9G!xSJ;b@v&a
z-||WLncqCaq%+pzw-UFDsp_})A-7YW1RoPKCcb$5tyaGtQ8LL4f#(}1orq&vI_J1p
zh{GVNuiYEgiSUqh@EGJa#m$vOPO1jd!h2=JI|l!9?*FJRb@0u0pO-q6Do_AL4+n})
zWVX#2uU@?pUPX5j?|db`9+^f3Yx3F50TE<c`T!)kCi!=NsVWG5VNF@vDtG({c{~j@
zJ_jUvVxhwT%|?Jo)q2j%qYGG0VB4m$kCmZ(2bPUyS<5UF$R1Z+uU;bqComb8e~BAN
zPRhxb?xJ-_Y@ou%r?1s)0{q}LZP!mksV|yPp(^ykZv#K)Z59IKWC1(nXn%w#J?)`h
z7khbRyNaVOtdLr_v}7taSPst&!L!;JeA2LzzfN=;Oq-Nh)iH*A%D{JM1Jn^eUv#~1
z5x@g4)MQ3CpgIYruuZR5kWgh_DeG$-qitLqVu60tFG#7!GvRx%N&HMSi4(N$v6|Hd
zT{4XCTl=}r(%Pf_Lzo{#?}w9mKNQ`t?KRsMr7Vw!Nvgrg&uPN;Nh9BGHdkNZ%FX>%
z$Aib7_g0niD<QUy1iLv!I#<u&d=|*~TJ1p{OH_^a$6X2fAJOG9HpU)n_Z%og1lHX9
zlzVLuSS(b`Z<2Va0UW_+bioM7I}HPW?)w8BK*MFrcLP84T<7~fIcRp(@YuwfD=-lI
z4J1Ba2%Qh&83gIL{~S5e3BqyBfTdHv1<kUq5+@dMO^kv~s5-{bIuq=R%wcx&;heD5
zP7Qn?OP5%Bu4{nufMV#^<NrqUx(>m*i5=~6ys<Oh_H~37|5XKo_`^NH$P+*URiQ_E
z=mKnYZPLseBpYG5TOUp4sO1<8oev&w$n@K+S?CQ5qvLky`T65o`m*L{J52GG2CUT=
z?GejbKMX<Y^JOL6(A)>ekA0fXQznuwb@{~PrKv8#Jv7JMDsha$NfrUxWuP%=5a2`z
zX=mtc!5`V(mqG2JW76;mUQ=P6T2~|w8JUN<%p24>D7*qpIOHn#_%dfx(j??$x`&_u
z>O|e{qc;-KG-WRKw@b=`kHb1hVI&a~z;^<u+ZpmJlLMGkp}V8aE}-`Is?7k*O|AqN
zY%DfA95KK`OY2aM!d!f@d>#j*-MUY2+VBT)n3VT0t?jG+ajU0yY1s~hZPT`B<<N>)
zge1vnKx|b1xsB6un`vUF{SP1%Y&vi0eY%eNnyGBeqiXoe9t)(D?KB}uo^NRg=K&jJ
z6o+r_GqK{&IHk)zXj@$3bWeu~Agn?YOTzc`hlZYsAgqhy=d%vy!6-ajw(dU-<M+Ak
zyyCkSdAIZF28eIj@;m3&^00MJd&$V?KmoqAX_08wvnwhS>^i%Strc&jgJvx@b04GT
znkS<cNw3*Y)%iGoH$(V1y$ZMR_OO;sFU757J?C+tCvhOQ8}nby<$-i5mh|EJd;W&k
zz0mNI9?v3V-nwY0tvaMXgUqJ&y`+m2o9D|YU@QIyM5Z+Cmh#<o*L>>-8r$S^=k9+d
zrT5|N$YL+31=BLrd1iNJPZR>7Dop5ZqG+Gfl~@Z@uFg~XB<90eE%>?gM(Li>*tVo#
zObJ_kdant6&=PZHM7pT_Vn&>Cd4-m9#t63!>=0%WN{I24P>1JA3RpwmiCZI?#Fg5l
z5qd*!FJyN<cw4hgcM!(>%zuE&Hs=#YR72h#q0{WG5FK9YAZ%iWPEs=fM%%A~kb!|8
zsI2$7Ae+?dF*51mfTZ3!T`L8I^CXDgedemLj5O{pU9i_VaxA0aR~(5{YOku!X1j5c
zL3?nawc5KWNvXa7?*e05C9?wou&*GVS|yYF^l7<i1e1rZQ`fE)GE>)@DZQ)XcJk+C
zSj8t=pIxEFdD<N2k3ZrXO#c*Wwr$MW024la>*fzDi(DsAh~c?$>S$u3e0JN{i4wFl
z-mcqbe23(jOMr<HdS7!1XcyxHzMJ}<Z;vcG%?NL*ucLJNKTiMvXk0zT5L$GVQLr$*
z)uob+O~-AR{NdLf{jGBYf1PEq)lzO@!EVyzMS%PSH3LclW)LV(w^Uo>+s5npgx?5@
z=k<%uf5-c3b3d?J<fZPhx&{2v<b^wblyPTB@=WR*fB)s7oOO}tkC><~4o;w3=@IaM
zc>S#({z4)}U>x9{WwT({x_|Hq_$6g2sV>e&eE2}5s)N{}|73<QYEP<qiPkzIS9&Jr
zB#@nc6_MWFc2f}GU5!CZ?(HPA+?8MAV8oTnyN7arQ0_j{&1({vA+_2khECgdhs;*@
z&zRydyvn9^I*)y!Lr06BcSeW(2{@_V0d4Q$bnqKwcw=@5=JgCpIS}AfuYQ(Tyq0Kq
z&T-U5jwK8haC@iZ4Y)TP#tZN-?krE&WWH!7&?Lt%Jo<~pFsj2WgyC<09Zi?->HI&N
zTryC^>c^Q&(bJ|k)>^OLwi0lZIUb4Y&Mp1NabWzMKTvBvhA~-x)shfrXv==6h5#Ql
zxyHb5nSo8b!X6RIRNLx225c%Bf6t-+{r;W;^Ya`QAu1x`u*RpgGMer*mag(NHYxzV
zrltj%v>aiKyBFn`Dd#7F+)5XNTvxhG$USu>OdQG3PY>M{Hf3chdV1@lQn0TJr-SH<
z8X~u9x3fk`9zs(${-8Kt?>=#3Vm(gXk_zVDd@lPf>h$rmxP4HSbD0IPbL8K;1@T{k
zWRRVNENAFi4HkJK!4niBXJn{vr7o~S)x5t0f8Hs2)Lq}bP2GRi?K=A=>#}TXz(&p`
z;{W2wlYD;SYG-S8Z0WGs;*fvr@z@t^P+ZFO3HOG$DTrT#T(N<(jWU~+)L+^ec3EcZ
zOH;>pO~PNzqR#`NDC!M~6L{d*g>tg-;9qq;;7w*m#}T_3W!+exfd4&PIDdM9_KD5)
zMpOz)4q`W-iOL|SUh=5e#Qe4{alph%r((0vyyY%uR-`bmwt4&qCTQD~p9mq0^qcKl
z5}v+<uI^de*#nPRP1*u00hg6>8H$)w`YQa*HUYnOaHXMO;<mTs&3|5ipY={IA9v^2
zn`A3N<~I(!cn@xFUddn-@rN^ufy-tO*IX{Mihgs0tGU?+&sE&R_FE#O2aTz}aUbR5
z%@b&`$gkptuspYU+ifJtac|*+aHk|h(>)s-Q4B;=uu?jmswLYj&3vSXcTuGY!_cKl
zK~~6#i5mJ|b|-s{MuGP?Q$h|xsG25fHQSFv&fq|hF(_v>)w+dLmI^0P+DyIrNrGdj
z3oKWoV%Bu2UJ;f-P@|&41Y(DqktuxL`U#{X1<ZCz?l*Me&?%R3K!l+kH<HN>I!clT
z#avygGdL?M;Hn~j4ND9pv(SwL)ad~5*CFH}zZFLte~JvYtnYo;I$v$V5U&K>8?f>`
zE?2lOH++<diI6PhG8|VYS~}l1oJ)0wiZ{<LFa7fU(1iE2zq&AH^A7gXNWMA_O4W9~
z!TG+(O(yV9MraU!Km&XY586(OCuXH8By0M#-+Q7#1rAy)3nwBXtIglW?tV?VrGLK2
zA5UIGyNCGS)BEMBw`J2`=XAq5_YURl0e3*D*y=}ll{N-dL7ut|RgC`_5u*H+XGfR+
z5_U&Zwck%oCBJ*XsZO*kwTrrIhbs6SOwmJ*)K`}xttHav*!GIqIOp_(<NG+Yb9`wD
zKBQITL#rf#sRvhjv`i#B>N9Ls)P~LARj}XBCiM-}xt3?BooBKt@++R}jH{9@fU#=7
zZYR>Hd4buGj64cyNFBFizJRq`C7f*WK%3t_)j))yQVzzlDL+FGD}UVZR#ix+qX%7u
zL^xJ?8Q(}SuC9S&XZHi85bbnw-m^SJn6L8O`e1PdJc+gyQzJ7ePz-mcZ1YID;s<sX
zbzkFWs6gHCec=cSN};~><?8g;nox_%hvGr@LJ>_Wj%bZ*vN9wob`hF!2p+*ojq|P^
zsq_%df9lvv^RJ*I_PCf@8eub|HgV>!_<~W%oysG5sM6`H8{IEfCLR>dk1sOINUrq!
zM@z@wT7J0Ama*6Y;jE{nU&n`37{ZUg8ZX*zYQB<H3;ITU?LocfS(B?TS!i!5iX}i)
z5kyq11xAjZW7|BE33aBuOsLe~Ye(M*Ca(;N%zQDj&o<htSxa@{XFfC`dBNNaCf%~`
z$9Nd+<A$%PcMiC<!vFgySOHFV6k@@LCSCRFfhS+c`kAUZ)+urnQV@>bh9xuRd{S>X
ze0Wf=W5|)7X-QbTht~_Ax9-zj2k)k9bmJ_EVj0W~#xr+h%V~!@@D5;HqlI0+9+|Z)
zZXQ9xf5cKO6UPM4OxNEA-6>Ir5#bF)q)OPoLWlk2*inU{pS4QJ%OFDFNjMig9QuTo
z4FoV+p~@_>)3UH_3dc2xQz&|6S-1ox_w_zMNwe>hLbJ^FE1wJvD5Y4I#{J@ikq$T4
z22a)nynPO(tOqU6M4Bcms@L|o-!)-m>Iys)T3Fy+SCaI6?EBRQMdh%JgcD+QNK+b@
z#F8iXzI&<h`WUZ)P*h6nDGB@Qm*py`P_jf6fm22mo~3~vB5!(_t%=xHN#_Dje4%l_
z5EkcRD00DZ7t{b}<I}6hoVz4VK*nZb*UreTeq-m0Ybem4v6Nz+ih3d%q6=o#ZhZ*F
z)30hYQBsKEH>BEM_ql86ajktFwhf<nyAZfMnw`8q+lW=-#hDbg*%?a1tq=n~x1Eu4
ztcoA6#XEk6##Y;Xdp}J*W|nU19R@pt1$8wk(otm_F!6_CF&;X}(_^r3RWfsQ)%!4e
zG0KO|w>S-TnbOcmgyn8W**Hch?%B4#+9&^aXYY9y^z?jte}i7u?JKWY-9>LuO^h^n
zsT9`oncsZd*e8|X9?(k=K!u87;J~4=AKXpZkPJa{LJaT=krc#@8vtU6fN#kAa7{ot
z;K!%ViHr}CLj<tx>`;nlYk#q(^x<-@hETR!85~3{e@K0dSpup)pm=u837n~|H8;qj
zhA25cGjem_m|;YoYS>;Lc|RRXzUw@>Q)XJl1(Sr>VW&)GC>F=@KE3qkdisi;%0VWU
zhtMSu)q_v-l|_KJ$i#`CBuk%wY^DKX1P75Jk2))7TQvbi69R?u?)`bQL5s_R-}cU_
z7$ov=32(H7`>t1W?HMTCU-oXQMKNQ(yX3|vtD8*}CxeR|ynSbggZCFy<KS)E?p306
zx*-#LL<t*)fjB&w<G4c7vQ8QY71t8>c5Ay{b_r`X7*wQQ$$gW+Y=pG+V(l>xr8GcC
zx#LO&6mbB%0AN2S0`(I8MvXn!LpoQusdm{JU%^3PUw!WV%%4#CU)QX@$LC+oDvBI)
zdz`Pi#i@&4KOW{gY&M@)^KRgu;&YD)8&BDD+ID+vT0FpAoxw$amz>v*S4I@FCs2Yo
zh&C%tVjHpDc`^pa;0Y)(%r>+^ao5zOI--*bj_z|#=84+UrjDMk4a%%6`rN+r{RBEn
zX99is|GT}xM}fP;VlEDpKj!yQkLo#|tZMN1`PaE7voGqF$-dckzp-*q2I>C{c2$@E
z%jnYPclOun$J*neg3HpGqTm~K6B)jW2?8XCDIp{p0%f1p{vFbS6Ja+n0?zOmA}A*L
z70#?vsI*`R8!Mwp?FJf3%al}qANxEWAhmvOQcD?TEHQN3`Fv2VwP?)IEQh}saVk!D
z?7@hEQmxfuG*4%PNK!28?%%|Q!O1v^8&YS*8tgw)o-!B56&8HrO|lT7?uq^Zf@Ed-
z9cLk>Fei~Ue)9+gA*qFHBWsnZE}0YfDad*He8(}3#5~U_k;(f0kEy2o$lr1gqp9H=
z-ZTU_iqvhcH2irC$T3^Fak8fHD%XX|QhyAzA8~3VFwJYb<eVfZb&&@PYmF30im`&v
zb#n0J@|`y3T#9B|2ep(?hoqbA=MVZ+);Q0O%|ZiF@?p)UuRDG+1}`cG(hLi&HY+<Q
z!{_Ge-EX)kYRQUeFy{xKz3<2}ZGxwmwfE|l**cDPnim#;n%QqoTfG2UgMWFGRAdu0
z`3VDD1SjLQq5@rAT^jG_muT63Hj3!>nZ2vtF?uy>E}@jSGh5{R&GZb`-Sk`^x2tu<
zE-pAkD8=6W8XlqYI>eS1zdK_nOn#6B`T972IX=A{?z)y8Cw6}SO?q(G_TL3x2Jio{
zJ#4q?5mKDwCKufw0Bqpc0xqB-silVTb6E>#5JI12&G<Pgk(5%TR<+$ZGA^_3V#61i
z62M{z?SO&3U5bIS>uat?@^NYZdU*Fsg7g_WkE?bdb6p~ZC1aZO@#p0rgs5MMn5R%1
z7esfj4VrH87M9tpcnBc!g06egUOm^*s%ZJ|M;}6z{7!t*@8ec4@wt9Gz4wq#Gwo8b
z?h@MSqUMYsY-{@QIw$Zr#2)9e@4I?ojs8fH<~Z8?tGX32F5ma7Akl}@ZmtMD^Up(2
z*{8jgN2t(}33N0(=G+xqYnvXI3q2E1ob%FIx`Fd(R?}$|T&{m4+3VA|T#eG{vUU9<
zHizGNJ8CGUbk^4tM&(L5OBJ?jxRriL7aOH~`b0ZR?>0_p5!%7z?A6@RB;!yS5@9IF
zn>iU&tN>Lq46*?-TvI0vYjyHXa@9qQCt$@KDv4q*Is2Acqq++kZw_QBypA$F=L^H%
zKZnognlIggzDLu5;(*q%iD|WWC@1{9n_5|rYf9<#%Du>EdskWeLPX6KzUy;$(XT<L
z57Q3Y&^nIPJ^)L}(xa4O;C)>xMd}YjqP;yfeRK6OT)8o}EWjET*+TK5*~x23p{AhR
z^ih;Q8;HnWvOGN=s_Z5_AO4|k=@;t#-{UCbvo(@<$vA%z1_4dPN~sorJX2V+=#N38
zprbWO`OMpmga8HqdAfBN-dli0xAmy@+uLbnZI@U%W_l35J^ed@!u?~2r{Ex~E*)-A
z`hsd1l`;b*%@#8Sj+Jq;LVEa$^V<Gh)<U`J?$sRKPx#lMrDNy!$0(+SD!ur4oy303
z2IIBkKZ37+ix{x!dR{g)@X3f6y##zDd-cm8FD;ojbU%_#ow~RAk-zp48L~xPJo#4b
zA$(((AVw;PWRL8pG<R<oEOgX^@}+0y_HMnJ8B?rzIZXx5NQ&P;b<3J@S0Cq-pU51F
z9I&t<7F3a+@m>N<0e@bwEKxs5?+y_-x9L*!TGwbcWY8$kNVf3lRTKUR-yNKK%KeUt
zb$Rnxblkk(P&7p_d0^ILtOA=*2+1zwljGmQ7p%dUJdSHx)SJN%Yk2^7Ln0dJS}MRu
z0~KUh;=2{ihHp>=4Gq28Y`@;*vZ!I;7kbjtVYBW4By+EN$U1E<)&ziX#6O%4T!_hV
z9ECt_*B@X$={X+y#5c0m=#j5NP=Wz{yUMNZAE>S`#&~qoZgRc%1AG`Xb(MryjvjG1
zpI$!?9&^-q;%n9?@b?~NUAv`|hcR2X$iI|&dz^<sYx~BnaTEIgBDijNAPd<2@~6^Q
zj0Y8sSFDz#^cD{2>r|shC=9I}w<%r4jr;tg#D){W_x7uXfbHG$#Va1=EM=MK&JnF<
zZKuDe6<YOUXo4w$z1HxmWn~b)LYcEVlIsi(wNkcnVF!5*sp3ihG4M|>7Zz2OgMC1}
zeiDQ0J|0E{HkK0HC2aoQJ~?vztXZiDf@p&?O5%oS`IVrWsf(Teqwrug_&ddjaD+0~
zs{LpAYMoZEoKHAue_`AT>*fYVqjAI0wZWL8Q{o8Yo6lD-R)?Q#?`DT%4LXl9B{+-U
zQyo4Q^^(ed7P*JaZ`*e~>hjp+0;Z|7wAfasYQGl57#I%yU2{Y0(lNC62FUL!)f|it
zU?<5R@5bPlUpp2KZI9-Dc&;!KiJTPdKhJE7xIz6a>cbADKO@A6X6iIh;)tdjO>cl8
z?IjRb$>g;9Q}H#QoP^r<dS4@kZ3F=0Y?<81l8xULvIK##Yq3hd%kSBJs0y1-!50u_
z9(B<cm%A+3bbrvN?<X0<E?Co{^;<BRm5K`~Uc(06Tx3fL+5Y~fwx^Pr_%Tia5~nF`
zFyb_!!?92ojO9OcDNeB8t6?K_)C=V?&c@<0@+#|-v3f|l^}mPjgV6NqY9-qpa13<#
zbpd|H5-KdThy#Q$vx_bh`Gb7nI?%)2tN3%F-j0XJE*1_1G7kRDg3h?GR^#L$`Vh8N
zgth?x`iO#3%tyk%&`ZJ?bKwMHkam>SWKuStIy#dejppknl}cHKvEl~SHUfs>Nee07
zxK1At5;dV}Gh2i)H|ZcwZs*L%G0CkWS=P!2vg$5fy&$d>O33{FUJD#<v~`z$J@o>4
zT|PxVVY>R`M9P$Ttvt9w{E1y&F<P}NRwcEjCgW8Y3DA<qe?iAeO+w8z-YUZybzS@E
z1R)GNnvhu2)emg9Kv%@)nlx%O;M)yKWfEC<4_b^@nl)=|6E8(t>b!k=kS<+FmNpmc
zh7gX}lSxgrw?oR@B2D1d743y1qXu?v1N9GZ9M<NRA5%wZR5Hg|yCrN?dmvge+0ynJ
zqT-eI#g!L=d7D(Y+|0$wT!&b3%!*;GTD@u07e#1oi^}4UicrT(+exh~iR(f?KoUR*
zCyniqemlGVXYji_v3c+WF4>bUv7yDi4!~Pbw>-=bTIcq-2)-9_d}e7;$oz=dcX?Bm
zcLPkO0O?+e<QUSj4Oh`*A8qRhumV+TH?Uf);JSX?H-r#5j4&d<U(T?ZGQ>*iCkWg7
zNnZ+EL5l{zeX^HQwa+K<WId4}`VU2(S%4$%p6(pH2L12TLxWB%^z?aZ;hg<#`Qoho
z=~6>Jn{%E&`ol{-E2s!}v|T(gBSjj@5$%en6yn3{(|067xSIZRHEu(}JYQhZg@s&x
zw_%@nJy^}ofPCO!IcjL<sTW)cK6Gvu-1csEL|yt`9*#Uf#XIw&8za$l7995aKAPoL
z+w&+w1!pU5FvaIp=p2I`A8g?6+CDYw%uV*~ERt5ddm6`yAA;*@*#TGK$C9rOMxO=Y
zCXa3r;;$->(NbXZp?GC67fug`_0c<?F5$9MyC6Ue>H16<Qq==-otg7%?4GS_S)L1)
zG&RsfoiId%P8jEhd#vTjC#?r*ux=kmr1Bnae$coX1`AtXqTFN&I#(7wjmi|WKhJmN
zul;?L5q4-RV%8{-my?o0bL?5s9`;OQDH5(H5+8ExQLw|zZX_9L2M*1hVTDSVM`&`~
zfqk2Ht59Vq*QMVZAA=1Vv&@0!n4W&~d>%7v#e4S0<=$S=YIU8qo#KwejZeohT*L9C
zpG#Tkx+6W#Nn;%PUkl_et?l@NVjaG+rOzXYDVHq<Qo$0q5-OH(o^e9H9@dekM`A^)
z(WW!uV-->y>Ue2|qka4Qu&2?nm+-9N!fg=o-$0gf*IW2GylbOZ>wB*MOlYI_>dH0Q
zrmYr=Bq7ph(1NirMd~QGjP${6bk$M=tG(}~(0yz}eCvK$cJ&t14neH(aY&ZBBT?E;
zFr>axdd+8pCx$1?VqKdA3|1Qa<vDLbFmDZ6;4{uZUC3?ILM0{@6W|3pi%cGhCunh=
zx8A|^M;u*RZ@0F<u^vj5-(P6Hqs#XlJBDK3^v*szX63yi{rbIeoXCP_+4{$$J0I5P
z(d`K#hBdm9*LXTrP@l`muH_fV6s{_1wIX$F3=qEF*ru<|@~>kq0Emaw@8)@12c}44
z;!{CeCw=j!6^GZi_B!aVn$+RvZ_06$btw#OP!TKCuKw8^*p|>!p?x}PdPB&^F{MG5
z2PH9XL#|*s040Bfot~<#NPrQK>Jy;E-X21IqK-z3fJ`IFl9rNGGnT7F#d<^j7bz!s
z0{YH<r)H*9ySi}6L7B%&5jIxs2SYeoz4m7^z)CavctgVT{2Bda_2{$_7=wXUDUip^
zL(RI+)dW>1D*SlVW92B`UXbpXrhQ1>#0S!+`{0%}-$tuifxpg;k44H;=dLvcwexgu
zr}fEf0C^MKd3}lu0PQE?Lndc*$NxP!?W9M?7r-CTEHM!8Tr5k0Z%=6?fJ<h}$sAUM
zP)o=HIot(Vh`6`5((E;{f{Vcs_dVDcChI5mBsdxlre67{W4vNO3?R0gi6uy`nJqmM
zPohdOVBXU1tSY1rpWlX4^^-J12hr0oe=u=aS>w2QAppt<$)26I^kNrSxe=0BTw+dH
zcFjho&C`e9-{6^6n)z<W9wLT`dR33Eb)*5XRo+NK=-MtQk6kTk5^2gGoZY~R4G$A&
zIN-e;I!uH>K;7F<Y8&&Jqt?Kb%)d#X;~&PBU{`Fnlwfx9pDZ7yD`aN93^YqqgI9t@
z-q>Cm+xylJr`?0~dp6UtA_%MbKZDNTwchsD>jQ7C&&bZ;DmCkv1`A&badoD7tdTFQ
zf-lolz|N%B@$Ra*_Um$P9fBU@CkGGcaNeidmBtT>NOhbf<gSjiDwC+C&~+AO&Y=$R
z#75un%96gMh`#N;Jd(CNUs0$_uZ<8AIrRJC?qLO63nQzHh2QL9N}CIr@OxZWMFVdj
zQ|%nhZ`JSRjC_rTdsqEhe3;sWB?TsEYqU{5yTMV~=kIPEL7108pY4BAyd?eKnGEoO
z@Z1_E`?l78K5G~;s19q{Pvp>6a`@r&4;;02Ei;UUV0QsKOvX$Bm^?}u$||JzcQ{#%
zpvTFa6q(mA%Xab*f|zAcqptloK0*)crG`Rlqtm8|1W1Lb?-0Tc&AMy5s7fXW39zw=
z*caEZ!G)Hm2WV92O2g}<-zYQfixdTZg#9jY+GUM@UjqXZ!%Y?M#tWHn+<13Dgo5Uk
zQw2%~xKQva?5T7cPS4&z4l1@|_u92s!fz`eAS|<ZruCR^QwFM2>X|L<5_c>JzuvR`
zM~&HREbp&Cq0E|GKY5FEqQ79q)a(3P-2J2?RxkTe&EfJzeV_Fm99{Y1&l>XBdDjfm
zY2!$yu9b6y7czS4aGz?x;5kdgiU8-I`MyvXjRuZyF%n}~Y$TI_UzLe|L{TpkUyUjF
z@pnO}Ce@4^0d>(hxil;9o&R{0_9@S$0IJM@OQ=H+ci$?VLe5f$$L0Hmugz})3(75_
zG2tDyZ=J@YjZ-w(h|vb~`IjOAJ|OVISDD?Gzg8Pp_B;l2@Gr|%yKUkO2F#L0zmpEr
z)hZ4?JOH274j(2){az2r5q3I}<ndi8h!8ZflKhDe{x8>b(HBplb9{Smp~T}8|DDYj
z1wqH7X`{BP6Blp4z=V8|N!hzre0*J{m)KX^oS0Vc*PzrrP2Ud@3gG!3puy=OBQXxB
zs*!<;?LUU(n1eBD#pO-ALJgtf0Nv1)jFu5dS%+X?h=b+y-$4l6uccFDrqIeQ8U>s!
zfggQrNb({jg@hp;e|o)KoAz66gJ>pzvASi``<Y-@@$}@6N+WcTaY^K7?0gaWK~;dW
z<giBNLO(=n?eagBxUl_&!rsY|tN$7C%9tK23gw$pzw)Se3)h2rdiU!^RVEj3l|68n
z<ooV0%*JRzA@+_ZArX2%LW^0)#PP$Trlle=^3*_p{=)tF5ovDNCjKQK9F0pEGKS$R
zxvRa>Od}C><<6KKJ(zs|ej1qm_H4kkMNs07{RL)O&>qze#D+0{`)I7ZnXe@xYjeZ1
zW-q_1)|l00paTD#41Ds4_|0zysieA1e8otYdU~kSpPrr)CIQDXetO9wu-f<Tt46#N
zA8-O+q{u#d5d56f>u3to2`2WJ8N_g)a=?%~%hc_oPc1HY$}5hP9@&#VX?1S<X+9Zl
zbR!6~T6HVesDTS@e?W@f@%-gi$z1EJ?J;jXT5KDnxUb!{=tlaF*K$~e_VWWgGk$4u
zs8Cq>ik48pT2)SN{RJ(!3vkL><|Vnko>wXe^$(+qRd$btgEA~&&sGO@sqZ=nVYjj1
z^1ExXmSG2+1+;*JJq9Y{Pj|=A-7u^GxxI}a;Epk@9bcwx@#60DszQm1&aCeHw;lq+
z7GpD5l7ReW+K^L(e(N6Yql4y&rCk;m`Z3&#wz{DH9dVQjzrW5h`yM)Ue)G`sftV7B
zkIIi6fZN)yK@T-i*DpHg`uJ7?2ojWl9~#k;C<G%J%U>*p=7%dH2{cq;RnOUxE3lgv
z*&@%qz298q+lm*NZ!6S$p#$UY(pTs%bxaZng?3A$W^yV@*1O-JiC&1<Mgn`@G0Mq3
zBv1<+ugSD8(T6myJCq~@l%QXL#$!n9Ky+TAa$4x#BEG6LSD_DyqnCXPri|nIBl&rw
z$+r7p{eF|H8x8qUFv@DvYnX(rT2r}l{;FGDsz-znEmTd?2tTmJ{cJ6iw7R4mln+@F
z9FkOoRyJ%QU;YPRA`?;bCJ(?2_+D+)=^suQKIkXeg_aa(pF(ymPYAb20wMCa+zvUC
zTPAwPkyj?ZOt!xN<0r13|F`74)ot}og0uRaWs(%w`QSvw!H_t%r!>UoA5Fo#g3e5w
z$YX0O;4$_IIFgzjjnPJ0-^O>bW3G5q7JeVhN<7(PZ$JnYi9KCAreGzR4`Lq-96<2!
zYC>%}```&SIsT1EWb6LR>5Y3DL>>ZmJuM(5!YF)<m1|q;4l8fwhU-w0lhOZnep}mf
zZC;#@+ltCU=bK?EA0r`u+B@?Wn(#$Q4`onv1`>4*WD;v+?nqfc(i$La=Y$ognqGh=
z6R9ScX+NoPRbFn^U-oT-FT=vW|8~)*Qy500T$~*TO^SvQM{S-^fF21|<?yGDb~3ek
zZzcyX>`{SPkR0+i7*%O>t+6WxB{*ukD8w(>RS=}N6FMuqqaW=HOUi9UOWlM|hrW|%
zb{T2${8CuVw<OP&)ZXQA)!?wyEA#NFA^-O6Z~4sP_V(xSnw|r~kQOH+z@K|_(dDrp
z6)y~IcrHF3;EQX4RL<9C3Ce=G!RPd?As(Tr-7>@%s@3n!SF0w>qom<#JFRKEJIFpk
zm^YQ2Cv(|?i`-8+*$PMc=-)p!j>zq)1-{K$!?ORY!nX&i*@F7N^9OMs@bw?z2pBH<
ze8rj{D!U>h097IU>o7Zils2v^1q>mK=jx)PwEJk%dyG>|yX671H7h@-0s^T+fZl1`
zDfGp)CnpYEK-;HIJGHQLS1-{Jo;*$&_-W^OSYIynJ$H*mgZG#8FBUm%HD=Hs1i4~h
z`Y~(cn<z&Nzu#WF`<#@o<I7^SZ1`OU1Cv@@L%ipE@P`4GSy|u~c8-@cMzlU7<i(F0
zVljV^DuwBL$5Z?mnMpO%>K63bw?%Y8n9-4)JOUJRvB!MgDY_|bd;HtOe6qFIs4<_%
zeyu1p8taR<pY!)|zdo;73xij5Dp*YtyE!L9wWTcXz6IuK!T_S-OvH#HL%ZIsYS$V=
z4g!JX3GB!<SeBA6%wk<Q{fXdYS_45B^$yT>O<;c(X|VDk<1NW}b;Izfdu5YOF|_B8
z8j!Bkj8Tqyz(=>Hph>hpl8;B|BKUs8#Gjn^;?wm2e%_(=j=yKh<AL3fbF<!Lc!k^M
zBA*w%?=qH!1%!Kbt>}R=+(NMt#yVfo#Gn{Rij7y%L^_L3`dfo>tGfO+)bMBr*#a0E
z9<KUR9gPb$Ic4}Xtg6TN8iKy@K<s9L*y5W5k(Yh10mw73?u=!I)}~l=^Z5J!ojSeH
zi3QwWW=T5ABVb>x9pD^%UNLehsmNn_2;;5)sy&diBFk^v;eRmb@bB-q>_*ot#_nG}
zXmt|5dk6+$q)$_`CTHpvE9-tA-@Vn@fsP)+DzwzY7bH$(HWoL|=U}M!igu|Z%W&5h
zB?pO<>47NKmCF|mfaXqc;|!KJMV=g+LHADNl`Fx2UI3se2a;tNRebD83qZtd5#kB+
z$3nP@gt>eJ%t{lBEJ&3NO|HSTZ>plIE6b>s6rg?WS{(&Ejm50M#Opih7(wI$LZ`OZ
znI%U`h~HKo?toeB1)!Temd+YkTMg%jE3e63mHZFaJ*49Cz3y{88b9?%(R@`Ts8(E?
zUzMVA*6L#c=pySVw%>FxOK>P?K^5BLEDc*JNh*QLQWC#GOW*b3J||`$u6;cI#Wm=d
zk|n5^8w#hK*qqw@%Mw?#Q07CR5J>JrK2$wZmKkh$u(=7W7FZ2=@He{L!{Mzb@YU?O
z=!hpwU2pW^l~{+ryX1XccyS$8@Le@)3hJLzqy4Vl!L=+{Q=E?J*_;=*u|8{(Rf`!Z
zBb&~u%47(gshg<xJ0Aae>+a8E(MC7sz7AvZbJlXe#9XXMEmZiLNUca4_XXttp1^lS
zA;+UDqqdSGCRuRF?9n~{SBSICk9X#Pa1OH@`J=mC6QloZgwc&fIYxzQwb1g{{C=kJ
zd8ZVqm#vEr9mgUOhMZSPB*PV!7ZrI}Nyz*!{pv4#$hcTGp_CjM{=|_q%w-f9xJOm%
zdK>uC#un$)3lCAGBA$iD^u{5!kzJVSgb&1LpTg}!i=C985&Rmwg$U=6jYm|vyDG<t
z6w4ybJ>^Zz<3%z?53+sO^fn;;3kqh|wa&Lt73@$fF(E!@a~wkmUK^3I14`0%`|_f=
zlI!V2%R?ynPxdzBFRYcPIy?oK%}QW2m&ik8PNHNb`M!Tj=uNZSWpnk_8D&bd^_b;z
z5tx2-wAk4M;ZE)%o8AoORj<Y}Hx5@e*)#pn3Bjzk(20$h!sdS(e^Fv%Iu-3APmd=N
zn0&K#?Pw?86LvqtFKh!y-pZQJ-+%ykH&!{o^1AO1<-i1N98^?etn>zcGVbLdEO<mN
z@~MF6rXZ_V<-%JF+#n0FWgb;pnDk~#Dxx5+6B_iUcec@}aha3+kAnBbRoEcwKce9N
zAaG-21CWm?8|=TxoH(KRz)S}l=@p}<c64$uL~ex-`3{YlmaTC7TvZ}4&5n1K%k_0n
z@2I(2Iz-owxyOn==_vMshU#_J$1To)ke-wLnUXrL#!p96wZmaq*9D!@(0H>!F;MCq
zfHvn=ul{;4==c%Aa$LH>yNi|s3aKaDIy+KFA>CqPB@MV~^?9GQtsNR2f7CySP*a45
z%h=)_djx?*4?1x}&=HJ_1?}D-gz72jLFC1x+LCNAEFYn7o+KHUvs2p~2NyRzZJvJI
z0PlGwwt~)|mm~>BXv7=@PeboaHr#oc7G0I=fQn;gU7$ATSPY@2H(?x!e9hjTpE;a0
zRV<Xp1o4CgU;Qi7*OW+LogX@`j7*^>W7-&sbXh(usKKjATh&JWX32>P5aozIKy6tA
zgMy}t|L;P}SKiI*-Qn-~0uunw==l~yGbP-z6O6u4t9SoLmTjW@#Cj>~0!R^h`}o{q
zUnc@anmG6$vQcB?{^JK`*7ZWS^M}!9L&u7xK|p!IO@$T-jGV>Dc&YIhL@An|CIe<Q
z<Nz`j2+SXeFIVeu`uS;1l1}y@;n~zHuXg-Dz~gh*+aeV=*WSV{199Q(gBi8Z4|ft_
zfIyc-;C{(;BEf-Ec!$`Z3ZPZ5n`$%G7B*6{Up)7);%Dq^!X)WnLpVlPuNUpLyr9b9
zqA$uH0kAq8ZrSoG?pc@GLhyHeZo&VNCY{C88qz+TLRJM!anzpwq%!q0I;r&ol=0dh
zjzDEpo|T(y2bOIA$sZ!R-kIImdi|faKX2S18}nz=W%SQP+4hl@7dv_Yx{GDZ3ay^Q
za7u~10aIMeN7q)s{C{QGdTZ>~s_PSex`{6M(f^kwpkIc|vaQ$Er}m=bvVCZ(P`8F{
zGgT(jP@NM)(E@*5kd%}CL~37XI}jaK8#>OX25r{$6(!8fxs{e8_+T~FsHjO--~T3-
z7bsUJ>+W@IfZ%}DOB?SO>iXe)O6nTTVi|c99E}_MX}%mpUXqd;b&Q-e&PdgrJGF1$
zde`pI{RroeM=I>=xvh$9h|aFo`Oi#w!t}ndup%SRC-tvRBeD)a$~|X-Jdcqv!&%M~
zssb^}g4iC35@^_78iis|F%PN96694Y&qJ26%Aem_c<Vy<tW;Uht*)gDc1UeTE8o;*
zg+<)gsJ}xzo@-zOgXO&L1m4d7(ew}RlhCiPufQB)K4(ZE21Vbx){RiDnF26VuoP!E
z@R4x_A#rb<_&md_Pg*LRACRQjIv$b$rcFEFu4a*p7&7GR!HhqK5GmB1W6^?e#rAa9
z?EXx#!9V6S2EP@N?nC&CZvZNJzg7HpeB`>jYx@aT5pGagaE3it=p#`y-ECLXtJuwk
zcDJZBh8Dxcw0}pV4L~m8mHR3bwg;DaL>teEv;T^yQ570G!X=e0S@QjZ%h8wyp~zEe
zQ$wIBn0NO8!?2;52IC7fT~7AA98cA?`vAZhP%kV9hpFmNizaz4f0hO}l*8by{K7Sz
z=lnb0PaTUlffYfs5Ly{*uMZd7Jxyf+a6NKmlm@T&mB8tKy}6LGR)bV=$83>XHU7@o
z=37s{s8OM__kjw3rQwk>c($WNEsn^f2ly8r{Jq8o#p55}KVH=WJ~pyc<D|+)m;NW*
zKC%Eo>((HBQhfZSW|A$t2exb`wT>2T#mic?b|YKecSup5MA1%5ze2v>YK&h`NF5V4
z_pkNXUtbVShMJr0#IO+|y_UxAWX$$gsigk0@*h(XJLmt^=;i%aBJCBrdcN<y-~46(
zcIOJ(s9|MG&54ZCpEeZ5h%7mEq&CmmAV!gt-(9stq=tc+VT%BY#!rqgsiyEn3hkn?
z*NqW!^)^7Gzl-W6_z2M@O=1+ZV?GdknyRzT8Y_dlBZw9O(8V1ZUVrNwuBmbS;3?&X
zH}?rbh3)r^4+C|W`OUb0>@w0f_TiRc?Bg$ijq&vlPgk*zt|4CA<stAPPG;h0Qq=B`
z0nNPL`dxG6U$S}03#dq=zv~UyZ<}%Ea-YX3<UVXXlf{t<T<mD!vFQI@oBNUskOOn0
z7HoQVJdo5ethcS=td+Q#zYzFB(v+D0QgCG<%Xno*?wo<ftZ1X%AWLvnSF!=s<S<V!
zNv@3GuR~06h+K_y?67z^&R)ZeSb#QoA4zhLSDbxkxMq531gB|i1ey#VUz!#RFl)_(
zlf8-fzgu}etNK>5cgIqWKCn;`lw^1~aJS3GP%*Qlt1*^?o*Jq!YCsPJ$1tB0<JD_^
ze>*D-(yg(fPzR%07?C5l*1h-Qy-R(26QBL`|0iu6(Gy6IE`jpco7Yjbu$NLaq(%nQ
zk&+c9XEx${8~lul-CKBIT1v%}M;XlOUG=Akse7%6*LXh^VFOU^k!%5rRq7C`Nc!x4
zBxrZLe2!jyac7k}5q|{*t9KHcJS-B_>m{}O^<G=!s-1KtDFp@Q^9Zv}9lkYMSW)8B
z6o{VCq+g@h&4{$`yk#9g<7^#!SXA66tr#fSx$}wI%y|;|s2?Mc$&)Vcwy8i3Y@%WJ
z*+ahy>ry67V{wh+gH2Kfw&1_OcpYFi3;W~odo4G3Y!`xTbe`7)4&q2Z0f@nX%cpt#
zd~h3*A^5j(zv-EEH9Xra%>k!pb|LK~Na2^6iFtXI12o*Q3*gd)b!&q<>5p{dhGWab
zpEFRmNjb0pA)w}?MI1T9`bvy<@PHE*@lQv}1;|OK#$$rf+e&u!PpwNAR#&C`Aq^Mj
ztv{5;vFW!)<bI5spc)Je46I%~z2Yg~oJ7wfq_<9W3Igc2>OevmYb)NGFjFn{zk2)A
zdb%SY!d4s8=a7ma^z$1N*i!z6zV1gE{d@u3#|dcsW;LJogw6C}Hxd3LX5M*8`Mm#z
zYu~H)3?aeoRO0@r0SHF`uH#|I&*a{niU1+|X<d#rHD4l(j2V*(a+6PaNHCvy^}p@f
z)OJPU208wiX?Nulic0i{55i<f{QX^j7g>xzoaD82CSv?N9m|SsWb~f5MRD4xqML%?
z3<Is*nq$=-GOQU6d*-tdV$Fxts-GG=ee3!nd-ytw4aW;Q_I}%;c$lMPh2S0(Nx3*H
z(I%1ViB<-@Tz?j&Es`|2=N8^08q@0W-Egl<6~TGSY;Vyh{V8d9VQoZC1Cr`cWSBz!
z_FLen=?k7xhEE6^RT({B$i>YY;Tva&*M}?$QW=_biKuH)L{5C9{vwt!!pDlyX$jjV
z7puIhnvgSrckQezS<R_{G0jYmBWo?j%A)@w$DB9E7>J5a8xaZH7Gv@MkoA^9c`U#Z
zDDF;h5AN>nt|7R)f4BsP;2tb^aDuzLOK^90Cs@$8x%ZxP-m8}%`I9PAwd~IH^mO-B
z(A#`Q?FG>Ex9~r<5{5}u0z%wb^xgQEat&=&aUo60nLOxGgG>VcFk<V86Fk2wTU<0N
zA4@dKv=3h)JiT>^Hkit=cEBq*WTY=8`I~FhrQ>1u!RW|gb0^q!Oubt4q%6T7_4QK5
z<T`1>KGpyWsu8T{37(SjuaP?TPkTF>kH<X#!?tDC3H9%U4>kbk93Y=jeN;8nZ)ex?
zemI5d#uV}C6SICu-WgF@qG(BNZ5<3z8d{Zk2W*ONOJ}B1d<*5PV6YW8m5W2?Sxl&$
z^u)K4ruUA{oXJ$xuE+cjYreo8S@y`yJlHuzgRX!a^LPYO1v!vqh5dYcZ5BrfFJ0nD
z4S)A1%Ror{9FcJ_9C6m!-Ov)VhUxT4YP2<wTI~qN15=!E)<{k8p%kh-2juKG{PGmD
zXUIIj6_g`9%u}#5w=nVj)YS`+hW$PihDU5?&gQi}U%BI%u`0<;)u|wtgpBFO&xM7p
zE?qzlj*;B`VC&=EoC!V)PFIJUKX?f7g{vWjK80?+2N}%+FEuL1YXSIDIev?g<63aw
zIF<fHW<s#(MI?n|9s`5LU#1@PIA5_RdOTmRJ4}A>6gYe$nBYIugX$vjIwBiw3UMgu
zs>*SclwI^ExKJE3VG$@OAtzWY?NKC3CVCpy4m%3T)2mBqzIcfGM4j`xhp9}*QOdP<
zu2=I_*sqfGG^6a;$DIw9XHO_?<IvXS|8`A|sF^%YJJ0#j^z#k^^UH8bB9C7&&DlAw
zbxmmK;Br0L%s*5OCMPgxQ6c%i-f@%o?$!K&3z{MJujVzQ8AwE;vtvxo!l;LuJ{Yt<
z)bMTp^76eV|BO79(Hp1UJa!hH%^8~M5h5>H6RZx&Q2<rx;|Y05vvy;&&77%PUT)D*
z5Uz>G3vm4VH;MO7y1!oC$u4I<G5&e&fvXo3E5u5!hnH#0p+on);2WtO9ZJ+BBUNEv
zOOPl#i)wVJ8!dc)U!=+0hdt>$Ctvpwa#Z0j_u80?4M^GC=oZ&w!1dW*onIRJEJy@R
z_5k}Vd1yXZ8y(fp2{;Q<GR=5nk%lEZa@Ljtjn>m2{A>=AC3kiW-;|X&Kc8NR8h<W@
zaAP4%KG+cQ*q)8Lg+EdEHL3PK)yOOPuJXg+0KvJTZK$4-GhG*ugb{)Iw78#O)&T2e
z`h3~px;IGImL8u2<LUYN63|_{P~{Mzi8;DSvX0=Fpk&qkScixNL2Bu$Aqyui`=AfC
z>Uc0ELMG|u!VyIO6xVxtB!=_q8(aY7PJrJB!$i+NZ362<158^_PI{8{<LPcZ!2?X{
zKdQWDSfzAWfI7dwSRV>6;MFIN6JC-KuIB6HKt>Gth!G1_;;a(w`9gJV|0)s34ba5b
zW}~~9#9oin$_wSQrz+A*dL~wpW5~5W=x4__FumB>(2CMQhG>WD?3Zbl+2Z6%xanMD
z@TgGaJdwoNl~GH5y77!_IB4MCkfaswej+?c+-iL^HzzQmu-T_H_;O7Nv~vK}hgWU5
z5;d&u#uc+8CyV5}^?-}M(qAExMJd2;TV?9+0|2rF3Kk%+F=9S>-AxU~gVh~Pi;9d5
zHr7!q@%^r!P3?Fko-syb?z&mwwNwU5!Y3*vowtqWh2LfHbJOc}F+bL+KH#RxVHIe7
z8Y{r7+hY6-2_jw7Z%FzYO(2ogH)qpioz5!6%A9J%Vn`0p9lg8*sC*H-B7V8MrjiBJ
z8jl&AEVp*}8+AWDgbkhhoyKsNpOTLQ%}i%<A&l7P5fC=gSz+Dm%MFYcmz1kt??M*p
zn5%2xL?*4=0ZKp(8gn>=YK7t~$UobWk(o$pLk(--9AfBdc;S1rO`%%L(-m?R&q;*}
z;#Ft;5txL|yBk|<|5n8Ak%w;XO&|Ky54Rb_uRb2$(D~(>1^GBZFt1GmGSTRt9V(-6
zJNue8zQ9^!R2kG6bkUpif83C&BUSd~q2t^}W5gifioEYjcZ52gj#L2~xQZbW=h4p&
z02EiXNTXj&$Xt(|pQFKwVGd@aUnNYP+zmPE<=W~_kBf2aH)hBvXNW;ygNM`Wp-_e*
zUyTS-M7Y)Eh!EtZ;<ZerEvLV0pLK2&CLE%Vd{E?o#Qe(gt0Ep_WdE}2(Q2O>+3Tl$
z)bC3EOFZ_O?doOAcT@M&p)k^~lHdmxw&zs@gB?5%-}RVw!-rkne3;;-%MR?je<)V`
z;R=_{^Vi|Ma^!d5sKgODcMwzx4qumh;QWpfV5eadtT;9up729Gbfg9tm{Gog*6)Ab
z;hzB#sk4<9+YP~o^c0`dt{6a}*4pGC?mdAO*>F*jW=9-T9TKD$RHa){eZ|EgW>H}I
zJJRN{6TulTwDuagpeLy{6@{<f8p*90pJ~?)x<3qAQ~aiw-~FV+6(QDf*cYMmLnr0`
zQCZ*@ftW*aB=5JgSv2ZL*D&b;sibN#sxGIhfvAodiMd?|TH_sl;O7g+H}&>T6FKjD
ziY^Y3va;^3nwwI(UCOWFi_E=*78;72j5iFS`BKd&6%bo_Nfh{oVWOmSajxZvI`TI0
z%wUiONbw3wP??q%$5Z;mT^dKkqz7E!<hjs>q)8J8`uZEdr)gUv$;&eaCmFtG2j-M`
zdtulgYHA_5OZwiGZ9}YWbxTeWo&el`C_ZI}i`V`{mNinxAIZ+ppZN1ZSj!1e$pY}*
zdUsIx;p{g&V{=$FlsM^dhunT7@j0$y)dCHf3H`PMx!SG5guq(XBu-3#Z=oC+#Sk|t
zO;0FI^YyhE9lOg@bc@HYnEstdo^KJvD^~3)Y6^9baHzJ?`);?OVI8`})wg35u8p_j
z;v5tbzOSA0Koz=Zf4vWeMa@Hx5_|6!@t9=UfJxidDQ)y?dc23@O&s?Ctxtr5FrS8-
zRk8VI7Sl`S!R?af0$hK@kEs4=9;B{9+y2s^W=cq-Ub>EKj;(e3n>lyz|7$=9@ZLJS
z?gst<R~(W7N>3&STg2kgk)C0&%AZk}AD+yYW=#Q=F8+GrtPm>Q&l+nH*ve>lOXI94
zxKzZw2H~Rh4$JV+rOgEJ>Ss$v<se6*NB!xW@>G!PK8!RLII+^FhcK!r6TX`gan?^|
z1DxLz?4`Af%rX%iqtPNEo9gC+KNgw0`E240zx!RZ;7ojP@%e)gD&$8`m>dm2Q3)=_
zt~=zm4hREd6w~8JuJSP;&L&~b?JS>Y7LAL$$-`#g<VV7HY%Xq)Vwmm}8x9QZEr2lM
zgt^}9Yj%N|cR$sj(R$VbA`>kYu;jq4qR<%?SoME!h2QxF)6j;$VjR+@C<~ZzZ2DO(
z<YqJ&=%u7#=&)_1P$8f#Ef=9@Cv#Zb^`yh1nhUS>aemTH1*enJXFK15U9tt0CF^hp
zf(fs;L|r!-r0n}|IFeKN_07i;@&*~|ByhItFY*z@^`(xTWD<b%uBYWX=)=(^p{Jh_
zD(>)eBbbuqu)skaRurU{WIZy8p3otAne_2ld!)<Y*{E^lSbH}BH&)Snp+2=8|E%tb
z=!5#t0$>f2ikRoA4;8Q@SbJi>*24bX=uiTljf9<;?J!X>yM*N1DE9_;{V|&7aG>w?
z*}0sGAZml<st0b6^L<FkyJIZB*7jKAS~RZP1VW-8e@OCXjjQ5yDHcncz~|37s9a?#
z5E!fbvnV8SWK=@y4#{4t=W9O=*9zl`;oUJ3aOu14ZvqmmIxl%xH6u8Uo7O(`i1f=T
zt7!btSzwtnb{=-f6CH!M9)#SnX5cT7Ky~vgrl|Xxdqa=!Pq{3KZ#{x05dDiSOy*Ga
zLs&?+!{QZ)W9HThzn%(4?bFbL(;@Bo@PIuQsmj7@TM@{`xu#Du>N#s&Pa-zF7YP+^
zvPEg)U=t2UM6#&ccUi;+IOE?&sBLexH(N|bMngp{kDq+pD!wjJsCH?op-a91%WgdX
zW<$SY%3<-hL*bX}ZCDZk&#Nthst)(<Xhf=Pzrp3d@ZhY@_t(aH^Zwi0n~pNk1{Df9
z;3G5xSQk^G{3S^QbA}u}aQ!w$)*4NlbxCd7Q7ZeHsnwrzHuDac++ya#eyE9XxK#@N
zHi;uxvBZLrZATwcfiY~r5N(tAzg|@h3jmpivvo@Zb|}`fin#Kft8N)=a@BA!NE0|x
znx;t^OcGf5jXNLBM}4t3&$|^~?uiz&g#oqp5ioL6RZEReuzlz@B6lr^**RQynV)mY
zDI3kcM1UE)((KstLDQ)aWTU&R8vtt>%xVGLSdQx_qn*}4OJXj%cPfn`W>NRxi)fhI
z%|NAHEhTUv#L;2NC*3D>+H8rCTc)PCr`p@PTXLbCcp39@$6?h26s*iXUD0?;p|wtf
zs<=2k=jnlb>1e%bBQha4-P$3{y)p{dD{<tVl-Ygl+CAaKNr5q7S0{Yl4p0hWU@!>K
z%NWrL=Famvs0mrRNn;Dw-}LBGSt1!)3QnA8Yq($3D(2#ah7l&PuG;Kc%GF6paD5OO
z+5EUO=t+GH7?Kmv^}^_E;V*;lMo5EynBYDQ7WZMVwLP2>z0H;<8u(+?E45#};3YUd
z0;u3zw=vUqK_YD2@2UAZ)kbjzsM>(sVRk$p39Maz(^$R&Wic&3#<$*i{@bCU3ccIU
z^&<zmM0(9qiSUilFoc9XfK-W!X`a{dkgdQ)FWuIK@ZF{R%t*doEk}5QiM-b(<obQ*
z=O@o_@rzH7`&SiC|H6J;&#kL%5W)Zk78Y!f@oA6aZJY3iSafvF3R=x{`%x0Oyk;6U
zb2?Yo)<OxnvKylME~i<GSzud6jmyT|7wZ1i2$wo6IB>I(QHj26i|<KnQ%k@)grk$}
z!^t)DuDJDf^sSwb=@-2CCTZBy#4ILdo{!p9&Xy9p+hRIzTr#1C%~yE6F^Wkr0dr3h
zD>!h+yB$x5d9hGI=u3YT=;KHxN{Z<)gL@pm>I5(ji2C!3r8R#LpGIcPczfQ|jBs*2
zM#Q}-gS_`#&xJ{c3yI0#0wO+(RPWIN_<64j!s%z}Cvestw_clr(qnC97~ZdC3pbXF
zv|;@op&J}A6i}O?Asu6TvTrA<LC<M3-E5GCJKgOPlPwUZT7_P>WnQSGTlZxgxNa}s
zF@{^d@CEFo?E4R)wchNGXQVf-;O(lew<n-n7b&K-zW;umx4FA$@c?XN*ZD*39j?cK
zso7(vKZ5#j64(gR3q1xHLv5T~UT(IyaCg>QZ`#>lH0OQuBv^qOf*st4k#Od*JBbBJ
zF_ZrSz5Xa(sGA=hBKRzI_I&2@sVmC{0U#S~KwmGz{uP&Xm%{q-x!avKD!+=sy*lFb
zqhTqLLYI<;=1Hw5>pA3nP*y#y_XV%b!JOw|g&Q6$>=DyoWI1(fd&L}#?EpNDKtV(w
zeWav`y#GdF3^1c8a3+Lvr$u*7<<Tr`^~>WmcOvVf&O-hyStA(eAbQL!rnp7nVN9hX
z>B0i*AU*?pMFnamL$7^K;LW}7mnEnt$4Ffz7%-sB{ndGkm@C5(>yNSY^tkRwA(b=)
z3pH(GB#dF|wB_Xkkzf7MMvp20G<kf}Zvr1A#bn`CIlKRNQ87E2bx^%bo#nMa0BSE(
zSI}WPHAW@;IoB$B?nLfnv8G;OnW37_S1FZT)^H)U8X(FapjD>!nZAMQhhI&3ruY4a
zu<_2lkaraia}KyC(BhY+%)LjwPs57~D+mxPyVTvUf6X%Q_Ydc*jak6V8<sC^=;;S<
zh0bj~13>Y<H{H$6O;q-J^xF!psJ_&OshSVUz4jUI=;HU+Lj@>7Yu-9hARblK)9|SZ
zsKITzAMkqGJ(}F#JT5ILajIXvFfN^keKE2oQxE=+(cdC|#1mj)VX5|i50qddhYRhf
zf|Cg_WV_<3#Euv=)}-JmLH~{{?yHwbX6wl>O2gp&+yUsn+1+NAqbar>=WM#@&3sHx
zQ(`J~>#a&@D9aB<`i@CUY;%hQv8lDkflo^O+P+@b{$okr1FX23&!Y^4Cd)UVBn!Kf
z&9}B+HybyKiH|oef~-2u7gj;d7cjKhw~t@r_s#4~<V|QoYpKN&a2vWaz6UYo5B47<
zXk})(pvO^HB-uwL&}<O>SchMF!#!lS0yTq?|8Z=QI^@&OW+bZgdKf2QcTNI~C<i3(
zj@~a)a9>#sD%@})3qymNhVh}q0MMed9y3P+k&+EGLF`Wnxmi9nT6C!Qjxa7^5O+zR
zUC}{Hoc6$k`*%qoJGOTD!*X6_uyf&hm4%+*eEat+Z28N)-+RZtr4rBii}kESHhXc|
z?EK;)^&_P6M2@gMg4Kv0Ag5>oZB8^oUb)*he{3Z~sZ*!IQ7&T6WOaGlRZhZ^*Kg}#
zg>~l(W{$}wf>+1)(pIHC<PZ2cF;(Wt7rc>>2lbB@fc7)^uvpas(Bqc<_&NXn7Hv9l
zI_dwL2|@)O0m;6cc|$G>trf?*%n=WbRDlE=%?$i6$Kl*m(d|m`4sGdT@BPyGzXPrO
z*7Ed}3<Y*^(bQ?G2s!Yrjkj*~(lfn_pB{+bGv^3fp?kKq;v;T++(K>($rt3HKvQrA
zcG)^LsOMq5eOJjfd1J`VEvM<3ZeE2Frc1=j4~V&10?UMZMN~EN5e3ONZnL-GnyK8`
z3aamjXv9*JCKLgoW*glnoTru(&0BYi8ysq;aj@dk>QqAu#%t?yEAAtbsofA;l*}*G
zX%ioRnbK{=|5Q6|B<5J)hNMsa9EC0uknYNZH<zMoMTl%uZT_u#<zo%kTlEw>U>a?T
zk7Q0Ag_`E`9(aZAG<wnBXar7X0@kF1)2dHy8$PdY-wob&5Sg_iL8hJrKgM-DY9x+*
z?Qdn5>aBkRPD&*L?uq$*))}(F{TZWiVNT906&SDG!tLgI2KwkH=vTxbqKR+=^U$v^
zEAa9)1%#L9g-`c$B!J9)&R^Om>^}^N?76j6v%A{3Xr4jvS_G<mZzTN+teCSE&__C?
z%DRE%hcMWY<<Z<~3Mg^FySKa}IT=dsq?wvpQPuT)Vhf7KJO1u_XV*@KC)14*V4<Tf
z!&R@`mF#Ewgta^3I*WHEWra3p%1%C#pk><QA192fT)@}Xj=%WexZU*TgiG!j%|}Hk
z^yWEE)HELQ^B`LL3OiU0#r?`cBDZ2F@->h&n|2F=7jc}B>CvYbVETb__U^SyI4eC%
z-@In<{bJ&6_?Q}nbvsbd!?l&f+x6{hjVaLZbE+Occ}r6V`rM|b>4DGB=C%C^vZGwf
z8pryOtTuzxw?2&j*$6EIQm!%giqp-ZMUNH2uCP{zp{BX^r=t>i4{^B~cHh>P_<L1B
z$y_2_OS#(eLFeNQeZFx2rr+~9X51tR`wTek=<}CnU@Z;l=RXD{Tj=WCpF@Uh8WQrT
zRA>Ob%$A(os=cZ!2dpEUkSTHO)UfhxdP|0au0rn*tMRA|UaaNskc$epxg#WBb+DXN
z`Tmg*KaIkf7|1gWPKwG7suRTVgDBybM{?(BkkVHCGxx*jW9WYz)i1#*VV0|9%N%wa
zb3m>v{v8h@H~_czs0ffnj>}mS+c=#$Ka;CQATUP57y_;-)%yg?=$z`G73(rXR)i^0
z%)Za-u}a*ZhjtOwRMBKoP;|{J2O0YqY;pH33Wjz+!I&djO@-4E9fsv=5j%IjY0xlG
z&MJl)+_9dVR(5EJ2WuYTe3Xh-!gg#3z?~r(=g{OJj+N<A>EFfdO`t->Y4__nL=^h{
z%cbb<>XBSzc^7$$;iyxD2EVQq+1_FOJrw}0ufIv&p-c+^Q>y|lwbCIttcN73=T`VA
zk_Vs);ZH}#9YE%j#FD2YaAp18q!2OA-U5Ne9rUBbX@<>b4-iNDjuwoNR6d4ZMf93|
z3S)4vlKX22DhrnKckkkH+=?7y4jXUcE(b+??aSbD$Yebifa?NY4?Ik&$94lig6C;y
zNCL6s2>Dqb(TR>mvMoeRn`nYsQ@yO=8V#2`#krU2u6Cs(*_SHFf{dywNKdN%{)maG
zn+=@f+r&YaeLIlux0L8TLiR9nCv2GYFPPX$z3a~H3;PY721g9j2Nrqc1mP_Krgub9
z$ljk8?X1^bK=iW@_3ZQ1x}trC;p3#dUv6NmIg5Z5`l{zEM7HfYGMst}Pdy2F;CDEa
zQuw|V>ovIW7Xq`DT^p5HN;RJEO%R56QbrX#XOCD2@;p=p0IXwW6jMTKbhrR6<oI`V
zoC#{&wKc4Gn(^YxlO~-4ybC|<G%;7EoFzP+Up;~@T0fdi49v|uK`tbV4GeZ7M$uWQ
zADNwcLZ{>Q4m))w@SQPt9DM^~4&dTRMaRZ1pudV9JBw<l%h^}ZT7;IFIeS5%+2Hm5
zn8aLx5ubXPDl)Myg)Y?TBj28*aDVLwQi5&%XHngP+Es{;8L({=;KP)mZ223uOQ`>m
z-4}3Y0nm<`*t@-{=^+Xd_F%)EUx3i-B!Zf&9Hd>pWeuR9{hasS>Hr7TD|g>7yeq^M
zGUBOU!5&9yde79UU<qu!I~df!u=?(yjwsLa$Lu;uQ?a5(e=49j9+zR+V1U=cLrdbl
zsjyD)$Db-W1p;9G1^W8@Kd}CS1cYx5k=U*aqFF@g9L2?O-+G9O+fm7(N>d6h+<8KN
zWarfBLL(%T8GZ>g5`d?fI+*mYo#(lz2$`)I#Ml=5!S4Q%2uTvXlWhgVO@K5Uj&KJ7
z-y56$Av-*1$0cS@l;sk#<ywmG%xTKJ;*B!>eoYES@w5cMeKb;c)_=AN`(-8s%obz<
z>_>vOOVHmq(!&Q;A=<e~S!M*|g_yp&791(1N~4KPdg#T_1zQ-kCejwb-yFsm8qzS3
znecKz<1AEeI^Jzo0k&#jw?piAgf1bA+{#3+!B1w(q3$-n9>WC*j5RI2M(YhaIV=ib
z)UW3Fh}yal3p?1(AsBn6V^|d!dVm0c({^!SNA^QY1#W<r4;e*j(kS<)-`FE&yZ0i~
z(%ks$zQN6zi)Ged*>l%DNu9^Vk{5u)$)#g|0pgsn@r$z~BS44@H2a7jf~&$IPc}C7
zmic~0`o!-A?31lPSZIL+0xh}d&xM4TVC5v6nU73}8$(8NA$cg|W@ew5MWnviTn8Cw
zm(??$uRo;yJXP|A2GowP9{nD}|0mD^<N&`te^gb09<qIF22X)puCIX~9c4(rehes7
zDC@j`3NfOI+-#sB2yry@+@=K9fY;IN=(PB3kN5nS`d_9giL#hDO!U!$wvA{olnd~Y
z!(b%8gjv>e^lQ2mnCo;lsVXybOv_9KSdE600ct?PI0?#ErA#K}V(eaXGeMEtF|OFn
zm_ZDW@s86XUhWFp<-KcYJUp~Vh7Ll6p`eu^S^qItiqbM?;N=lTs-tf7Xs&F#Zyh4>
z&J<mw9{AK>1MM(t!d|U{OVP25DQe4V%H}TSGJ?pgB=nm>#7EF~@PyNrERo%Src9do
zDM2q6s&Bo>H~Y>7mcEdF9k?d7hGMn0E~qdRv&oihE2zPq2LDN}P~)^wrzCT9kFBNG
zxK79*2Z;PP$I*{=9;OD?V$vjjo-dErK$vA^euwV@^j~L!pvMLP`G(%eHx?I*HEcZJ
z8M!t><BBdz^(&{IL6G=e6P1XTdX5uGw0>kVo>CQaB{?K9<T2UY2_sl=d|%R}0yLXc
zxP*D2)uCJ$*YmJeftN?w<ILpwyvZ3zCfXd$*m>Uj52`N<Tmh>`%319lkM`+8+{kKj
z&{G630VVvZ{@Ha~w*{APKmU}n(00*oc|qg?p22}#u|Bh%&gL`&P*wv`Ih<^^9I+<3
zN~y3C?Z}Unce9c@r0P|`8cRUO2s4)_o;SKdtyn1+#b$e_hlYMDFbeV`M`@8C3l-%u
zgR4=Z4ly`u)7b7JH8UDt|Ng=K1TYBz9?=|sKiOGUn|6428isBCw($MiR6wlXy)E`l
zGW+4t^f&OFe&ONR9UEn>d$xnqr4qQQbP4i6a0d%vNewAT&2rnV`0U<s1LX@)+zH)o
z1t(~&Ap}gH-rB8Nuu|&@<8W<GU@;UX?aq?P>T)S7I}LH(u~FziX4=4f-hZNXl%RjP
zfVn%NH17!tQKdC5fFELS>22C1uDRiUxZ3Umyy~ql>kpYkM8dDPvk{ooLRxP^P}{&4
zo~>P9Qng43qEGIzdsoAAtT?TEz-yBn*pLf}R>!RCRki+|axd2q>M~3Z<LNz&DhZ(z
ze9d5$qlCMNVqzx6>=E=heyfnHb9zjITYZC|<V*Db|H`5~|0WaceGBrt{p7#KJ!Ern
z7bV7Wb`R#q85hU!Uc>|iTLFpMv`C^{WPH|9F%Iywu!uNJ7FtCc2+|b3W&|J2V9Hu%
z_pS4kn-`hoB3x-OO#5YiW_%5v`|uio+xU&me!UD_v_Fwl;LXUz4LoC&KgZiW!VnD*
zwcfda-#%hgSP5hsQJ&<tM3WPg?Bev7NRY9Fm3zPYZEa%XJ@>=G_2#=3nz1jxk!lNK
zTNb2~KdWyyEN%#B@~;Ejas6i!|GY4fX!JRxa+M@Ts;7^(__R-?I}p>(_)lV^Uq$x5
zYT5`&%(Ldk8J5~0(?VWcq!-Fq(<RCuB3LCN<{1191<7_(1tPIycAdEl1)TTz2Xof-
z`8^sM8ifFEP(IuD;!e}h-q!wgr`xg1p=!wGcP4L~@XZkrPCm*40lK@p%kAh9RJU0-
z#|H^?ii>H?*@!PAu+fNv6@nMh5P723TaviD@nvX6YoD@dnY#NV&WS?}X3Lv5ES{D0
z&NUUs0&3vmiMB2N?SDTd9(dx%<E0uv09#V7l<QcYTe*&g^$78ic-6I~!^yV);94Rw
zk{&l0D7mA_Ine~1PbnE7dk=S3j=QGYz18MefNO{U5_VXUWR~-3%iU3XcPf@^x_Xp3
zfE;P<0~7Gwn2J48ZDry{gphy3j^Wn(uo+-5s8128I*hl`IPi0EE{=tzZWYlW(hq?U
z9R4aLit3{p#3ZNH$M2{vH{hSMnb8{BnQFW$bag+b$Z%bgAe#xwL;$u`U58FLAQJ-k
zp1m(ys4<cqGfgQVFIXF5mk3ot<}n@cCX-d6nP+OkGmYJt(xIVS)Kh9uCj0eWwSTI9
z%bAzcSLIV14AB!E)LtW3JzmHFX}3S(<qxg{!uYn!PVOgN!YHZt+~nDOGiRD;x9Bex
z(|LViS`xNjGCDf$JPCG<xRD`tCrdPJLcRrYj}Bg{?d#R3pI<CD*UX!AdCb+U$tm&F
zc)`dLH^2!PGW7U>cMSRE15&hiDMP|duK)Pj$>TtP4;J)(*q+-lp(u(2!weAhKZbQR
zA$x5GjJQ$M*CnL6eLHPj7amw}QU+f8cO#aHttX>h&k!Gt6P@5sEEvtqgWXN;fEj0E
zCVC%eEg#C}u*4Ed%qyK$SczFCKZFlr1o05MLT6Y*1z}{MhM1@cDEcQ4Rzc$D=9yF(
ze)n1E2P$RMfS;v(Uyq8qVI5fbHKWIN&@)t+s18%dBsGV=<Ha?M$+bkilON9#C2_^+
zT~u`h_w~nUrjEc?!LI9r_X%EaAnro=GSfw~f(_mSXAc7_1Dwlx7seE=5o^EJDW{q(
z*6xm_%1mKsz={m!91RVH+aP}Ii5#pTp4(GPWW#`Bl1MF0@BjQQn;Ocg-qDa8o1NFK
zr<?vDFS@i}=5+uMn`J&`q*)zC;xlDc=?6eOoJr<kqg>Bs|9EhRZ9=GlNO0aXrx(0G
zPzaGa`kS@JYk*hA&?Rdi5x-EaLc6kCx~-Rdub+6UP;b+iDUsgiy<?kxs_mMa70_RU
zb~vvU|M3nwqQ1xxKLUvVr$TJHll|=}NenC$I>;<!(1{E`3e!{8mfZrkwf}JECa4&v
zQhyerZUI;qRcf~xkl!k4A7gH)VAap=U;hCsYZ*JY{b+*ppkQBB*l^5d++Cz*x$D<$
zE>j6}{rBHcU;(aX813Qv*m)wxaj23(w8`99N(VpSLkHSzO7;_qP|)hGW&g?b(AU}z
z${!=a4Mof{4~`D=5~#C1PXM8l?Nb9n56ezjobj{hJzfy0)a|=A2l>Ge{anjm2jT`B
z?nUR}g?EfdlAZ?IG7r|fx~-r(RfRg0CY>g5()#IG2Zqg;55nk4+p0kCryBWWG%jr>
z;-aK`JZY-plW;K`uN7e<-z3g{zVM6DZE<5&(`6CCA;8czQyjbbJslSxA1Y<ASd?4u
zb7iisu0~nLS-RPoUC~BT&L2fX(y6=RW;h813P84n|AFWMM96<gm^yyd*AgiKTy11+
z?*^(4gmbno#YmrbFA3jE%tlAo9XgPOa0-gQ%=vHhIehq=KMB$~Q0#^mh0~2=56feP
zgKL0T3FVg&{L>E+Q|a|Zx;SGA2CkU;TiIfxp8vTUaHfsbYjNMYwNHaD5pq0@v41c^
za+zMr1fJaB0f?iKfZ^xCmnq6P8iQ`C5RSWg406(O;=KOO#F!9>+t1A1&;Sgx`duRM
zhfff>R4gv3+@zGVCNws$!C{w{fO!bm00QPt-MecWSvf8pQ<8=2L+g5*6uMT<`$w=&
zJc{5*7>o2TJouJoxm|w_W?YH`93Oztmr7tzkM#$GDvdu+hllp0O&u+)m0MGzPx)B+
zQ$a~X{bOV-zAHM{UFvdBG!B<-rezY-&Qk>1)_!Dg#aQ`mGd|b1iHUUH-xW=74}EdM
zY;WKFoVEh`EZ$cQ-{j`u!^AKLqXj`HymM{+KSG<rFrVwZRM3RmbG|%X*!eZ1Wlc@*
z+L!#$u+iSV;@N1>jii1Rn;IRHlo2Xx*r9S4>rkhkH<7K8afOw+qnn2`^pSl$d2j-T
z(aT-Io8tdXD*5~O;~&8*;Gk5l!v-9NN{Gl*fj+jI1`P%fE;Xln)`Lv2%Jkju&HK{d
zs=$5d+y#I?lI)h~2#QHj;)pcbE6m~Q&dh-J!J0J)Kk=Icb`$Z_#|%4gY(MXpp3r-E
zd$#a}Ad{~NEyyd;x<}eE?!>V{OxouYm2Qj~{qcpryf|=hdZ#^U7y#jnkd<o6&QI9)
z*C|+%%CArO_Sy4xh}~e9_o_GufJC>~dkTpmo1-u(N)Dc`LyW<pR{#h#(c$*^d&9Q6
z^%;nPiI9LPYrBR~R}c7#igA4+l@Vxuk`6*;2W`u<m}zNMGV@p}1U7-0Xo5><GLWDZ
zu(iyWEgHB_4*=$7u2+IDcgg-hbQ<?$8s>)}?}szq+q>Mqt=(b8HSFLD$fxsgYgLH3
zy}x&OcZL0u;?rqWm_$Toz>z@ok~(&a0Y(TS|Frd^pj_n>(hrmxz5@TD;!|@;q5|bf
z?a(72EQ<YZ1Eda&nwv*ayQ}{%&N^8Dh>LkXj?IE1?eCHL8~Q>982teoL0%w-C{nJ*
zMzc0KZyuK9r=3-;7$-m^ew}9ttOD+|1S04%3kyDrgfLg^cO*hB8cqIEp{SqI3jC6)
zDOX^m<2fr0DN%M6fSXtAvWpsS=2{*gbt}e^GUbmU5}i&%KC}_C0Y`Xzb|ds~=29F-
z@ZSGY9PPTaixH|{=vhnV;FCWWtRjfY>jo+1l2D6S&!_we^CJUsPxpp%KY=k}_>aPb
zfUq`UotDS|g)adTMBp9GatH(~s?@!r4Qdr!Gq42Qg(aV7@s0FIQoS1;n$#d-iayAX
zq7MGqtH|KWZCh=!Z`)ouh1=oq{F6%mmJ~ju_)bG2ynzV0auxF-(z!+g<P*~v28Gr(
zf!b(Rw?l|VTxBW5-Mi`-;nS(5His(o2v@Y6?-nBj7LA%*u70HnX7$Sr{%Tn0L*wO=
z^-0HOB><MX);)$iCjJ{0PrZKl8yRhOPTtNp=HZ59wN8P2>BUlqP_@ewYMH^lSk>sd
z1j6MevBaQ#%#aBmNE~r(1YDwk7Dpsl1Bxr5VK7K4&&g_G|DLlyG=vN250=inQU@$E
zbBGBtnK~Xsu`*`{hD=S1PYw;CGkveEBGD@#1Hbak3SjMW(Q-%~O~eh%T|YgRgTK5S
zU)MbTO5m*aH(hc0QA9f>92u}Re$o>I#J0T~a)`N=xlB8dPxJhswldS_&_wju6jHF*
z@Vs=WNYFxs)J89X@A`!mX+9Mp*Qj(9QiQ`P#F&p@Um%6G<sfgKh+)N{$y5d;W^g~r
zKaI-21DY-Xiqce)4iw-3gD~KlxoEy?^uB06eSJttl{eSxkrDEH;n`?xomn3rrk3XK
zhnA7wP?-TAJ+a)iZyv2}5^WgZth3BNaR8z>?f}<M=S>&fBLvO*3f>UZy5K5|A2Va~
zPf(7yL#gf~>A(00NTXecH1u<0NgM6Xx$9Ax+r^wx_K6Zp<rZNn$&DcPXUSGLb~5#r
zUcA|-FMuN2%sW;G5BW**xcCYv8y!0KHA>U{&?w}kO`e=gB#<P@@AQ@(s)S9}8lt`%
zc<`#Doa248D{C7=B7#J54_3FdqR`mHM$~rq76n<kTB}Z-q<liB-M)WfR3e)*YWD!#
zP>S~FCOr5Ex{4@^T02PcMdI7IdEnF<KF{a-bye9!NYW8bCxwP^=yz8uww$?bvl-to
zXydrI?jMbn!VG#``wvmgYRBgc15{d+wcg$kwAOv!>VE_IX<@tEjYzEQbF^hxQVTU3
zO1WCnk(C-1?xeGeMrAfJ<IAB{vAcbCoh2FyC*AUL`;SiI?Y;n85>cGPYCcoY;e9{#
zj?cv=w`Ctw)TU0O##gzmT##DUK-`cZbrkTiQl~Dc35e$sL92S~BYF4gBV6Jw*WqMF
zprNIE3j!2aRVCE-ab2XY@)dIc;X-h|4^sX+d2jIV^8Z(Gc#er54^~0)=JTtbp0m4{
zV#b&Ohxd<w9v>ka<Z#)3!|X3;f7T52>CF#0Xb4PAIOrZ-k2&uXJfNqYkSNts-rQ)z
zASwki`6Z+Ypfp$KEuB`KZhT!e8zY_V>~#6gj8=>LiBr)nRKD-4<x3LO2frvquFaGk
z*;gmXXS{9w;quDGq#fEg=H`>#c_1O{sm&3JE1#=1YVc0t^pv|kTgc%Bm<ezUeeiy!
zA*gnXqnJE)1}se-f;TZpnT)7olhiOhzlh2(;6(J_(Z1U8`;kGFjq0whfU;fc{#Se|
z{_Y3FS)#45V%)~`g9TNjL9)aw0p=g`US}X<h79J8{<GX~K&dO?icTF3V55pYY-l5P
zx}`CjhNh;}>!#alY2pys-O@akx$-p#PXf~ddRjU;bY=73>ZGOqgX&aaY$)Z*f%RVd
zav_0VuNw*^=rE$_lXb?ORYgRJ&&sozhoPf;P`IPa@_rdGm7O&i4RjQDq5z}Y8~+af
ze=G=(q@d&N5vqA%FG}T3N%tvEIp!zv1A9XeyXETRpyNYBAYhn4w`prZ0f^;QcKqVz
z+Pcc$N*aO{yh{hx8pRDMff-RRJ5wuKu(HdJ#0fh92R;x1yI8D+iO^3e)2$RFXQ~>_
zm4zA0jGv6LunygtI-kR^wd&H+jsq;N)_#6UW%7k339B8#;fORkP4*Qcqe3p0gflS%
zdQ7(Oy;ld7cflBeOi|r<LLNULA14g?AxK4uUnnr8Q;X`6;nDKCAgpgfBggPCT2_@K
z=#&QpGI6>bgeq3RK^y}P**!!Gi%=W6hR86EI~T0^_@y-tRl;rA4(*BzoQ_v5Ag=Gg
zTq?9tuPELEczPfZ$u8>YVUB<7qTOW6TWFkS<;n4du>xzW+u5TNUf=3mK%Gv$nH8oV
zDMlTm`?&^Pa51iqL>5IJb&3d(Hxvg1As2I*IZHOI*oVM69I3!Kjsa`ZoLD0Eo7GF}
zOyE<PUEQ1P{kIu7VhCK#)|dji#-jb9jeserk(9ErjFAIf?k0je7p`7;hpKKSqtNc>
zV`#*xj)N5}Eq>PA<sCCJo+T<qI%JndJbP3U=>!Fz_;~UJo_!~gh)b4IilS=fLbFjm
zKXL{1dV_eRNjciiMdoVEVc-n~)>D}Pl4!ECBgaM#&z0(H2Nsxw$@S?r8m5dUWIGt{
zv^b1YrRabFK-smVkGCC2YUBXxW0c4XP5x5^OE|3+q!ZO?JbQ4;abLwsPNe+H|C`ZB
zkmFdrVbWlYWCV3S@giKKDv;1L4J#%(-5i3}TCRSpn2YGVT#x*jOH+lukL>BMF#&`N
zdUw3F;!9pxnwe>}#(aZy1scWaLS+_<1w(AdWSY-CkdtQAarFpn0G+i(D^<e4Snx%I
z%BxA4I0w2vVA0=cL*i}0KDCLjPGL;L`X)1{$VIlg9<#kLeV*6@o6c69WqVEXsVOh{
zEQRCxBKd-128;u+%n8E(1rY$ET(C2rT~s94eaDwA#BjweN@}b83Ra~+hgI5BY_N7z
z$tz^s5&z~}XMpkF?D+jd3}EHF7NzQLcY(X1r&aW}ae<*tWPsZh15Mm7@Pb1OQ-lXi
znYey={~B{(@QZIgy2@h(Cob$Sn+2PUTv{%`i@?kl3-%&f_T6i|6*-;`bs$B&m1;>U
zu81UcXGDU{UJwVeNoBDVS!I;lNLsGor@H^N7NH<cCpJ?ZL4=#vOSBrq?17`af#-Gk
ztP&tL0rsB92kz7;VdcM_dxQ+`T*6oxg(_3){+K}1V@&MbcrrV52nl@~$8ywzH7XSq
zAqG-ts}JouSJs2+!d`_o6}cw>SueQgU6O2K^xHRw)mFE=A(BZT#0=17v)PPS<RD(T
z!}qN&eL8J|?YlEClv|sTIYV-CHpu@;Ueqn$AQzWeR|h}$q{w7Yk9#ZGc5HD1j;u$M
zIF>pZ>^rbLI4uM)hzXaSv)*S9e87Z!c{hG@_>T#4ix?n}0H@X_5QYfu@hV1R1n#pF
z4|xwp-dl~2t`nJnhQb)I&rlrF#NUR^=u>1ap7B9h_WA=SS2!q|E-c^aL_&*SQN(Et
zNFL3+)IX)f{y?7eqe*muuQSg*Qq(za0E3nu??C6bz00>c6^f$G47l(2H{UUB{?P*9
zQ*kY-{hY1wygB#SAkXf-b$$(9AgFBX;dVvYLH*`xRfK)D=T*P}XG||)b!K4%xXG`V
zrS1WPiZ{~?lB6kkAcMb=h!e;3@lCnyvg>@ZKkz%*m5u|Lpfu4+21F+ds5B7`7Me+z
zh*3L*aOIwTxkCO7<{%!r6xSGfmOZ%!7>LC|*)u7Zrpu*}tNV<qE|i3O!#B?X_9*NM
z-L}Tr<AS$ofVZ&nF*4fP>IpvjMn?GTf94CDN;oioue_Wr1(9HHXV-0(Ssl}}%|0^7
zW0YwCd-l-4DoVv?6f`1B0WT|~W9`AgFE~3p+d_R<EeernR}Bs=0%7i7-0E&?TUu6T
zi;cE2FBrD#($;#SamQXqF6?{hdvURDnwTNu&`5z(h}g*5IBjCRPxp!sA<R!SkD51)
zw|n>wYkDk4=R1*R`FhXYh0NDOUMLsyv;Gi!Z+0wd|D@4=6?$sCOIM`jaMKb&843;z
zA^?mO49=$?ju6;?&TAU>@ol7Aj7mGP(tBFW*I8z|KMDGGrBcDyD|)OJ4ei2*0NMV#
z38{^#$}}Y!Tv;PRx5K8B<zj&vY40-qJa#uUYQY)DrIJNyG%lW8&YPu0r$y(ox9a5H
z=zU{#J_MU$dt8QSUxS|QJI@ag_Vk0Xz8xwkQL?oATyJ<^#vvnRzZX^^*@q)qur*Ea
z?v^igZ^czT!3$oH^44dYK|L|mVCK`dqGU#6_Ki3RsrW21Yow0ebKA#xaeuu}b`Zl>
zrg<!ryvNE(Bh1E?2LT7i@?!^=EEJ=tByaAQ`U`xb)16i?V%hITenIu@KkiY9zH)hN
z?-h+d7xSwMdI~?OWx3zbCcJiJf7uxU{_|2xjr+T@R_}+4R+3FA`x%wEYuSs0UYMFe
z{+X|*_T@xeB#W(yZLS#LcxKTsYWUL7^Lt&SprIy)p)o8upO5+nmNfONrtD9SbJ$DU
z1R#fF#9v-4cr*_Jk7(QIVBjKPXkbuaf#M$m0Xu!2dS$|U6tlG+5Ex98e7%|)b+~Gf
z4PTsJWd0gel23S`7q|xoBK$ld^ctNX%X3pi0754*qPo-%$~47u$3XIE7f?EnT?<Ov
zlDAO|iTXs7;C-<7OW}S-dBPIbTRD3IS8pZsWK~=a1_~vSR$^B$F2|5jOU7aQ`*~j+
z!}A&Mq2I2pVXAOnNuJPScb}%uL*zD(+F1U(d|tVos=uA^can7^1=TvyFVRpNPsh!n
zJT49=l;LI=;I9lzT#Sp=gsrfij!H|zH&|tzH-%OGFfeiY&U1#KcLHO>Z&DQ-CETmy
z?$ovEd`H#u7bY51p9aPmhU6*eyRTOiGjfU%tCipOWThE}b^!f#<jXN335j8?)itkW
zwk$JF)e7f$)Ad4D&9xsb_cuI{$KLOkYblwOb}D)XLz>TYoa&{5ZZf3{L~&5>sbn()
z&YxBAHmn+DzUH{!t#rMUOMwhFphpFng*7~B{`aYai)^8Z8I3Jegyh+oZuis7EIp0S
zcMY`c8><IKuU>q(=3=B?n#xeX#8r5^x*9&3*L*+N9m-;<_Y2m-WMVL3Sdy2uT-Q@o
z_yaYjXdqixr0Whcz4`$=r%P{xqFtzq!su+F?L)-OW@zNu^3kS|s)N&sv$O<-(oU$B
zwZ&o`MnPp{44o~FRjo8T#^wr}D(t{RF;NY?J7c9$I+CisT-NSDW=Foym64iywk`Hl
zdK|y?Sf(h6otRCVK9SS!q0Kf{$4A1>=xPU~pE-?`P6~e}Ri=1S5>?W#)0hGrhx^fg
z;GE0@YIr_+9;XoY7#xHK|J**{GWi}%^$Qk-kmoh&Q!0bb%S;rr)SC*>XYM{x`|a!k
z!BM=J%1aHl++N_@i8HiXAKi}O@P+Uk9%K5KGVQ;hED~?lr)k*kPUONS{8(Ay`P8MT
zY;>kmr(ZdD(e*BBXwF#QOIU@Vf!uyG=`Y*MGg$<_pKVD(BY*;wn*XUdj^Q7xeV=s;
z<s|y~mon0g;fzy5Oa|)n0H^DiY;-~k=UG#z#!PzZA&%!{?Dh}2bsy3BXrq86UXiOA
zcLC9zCTnknUpbyCb9QFTu54+zm6|G?M3TFS)@(~*fi@GCH)RP|8Lom|JR0T4Db`+N
zn&n??{;>D_wrZu+9cD>45z<?Cw3xObj!M&R<mTTNoRj%g{aNurY;PfGQYMbHr|b9^
z%Iv+_hv#AEKKJ1?67Lt<%t-&8YZC9P+%UX6tadxMq0wY~_KjnyIzuYxbWa*h+w&SY
zghOf_mxK6H=bad??048OO=j=gX`qS>#qqxZ*C5yGER~ny4x*Ub9|tDMLI!MsUtu-x
z?PDddRV?~)Rez_Eu~w)&yp;SUv1t4T-8r++uyWy9?hf>T_x1fc&ad6GM~pJdB6hrk
zA#d&RXWUr`PQD9*70Y2Fa+0@@Z~7I({}glJ33unlb_3UHK2>lw2g*_O*h^F2V)Tk2
z4UHJ<9PzZUtjJQ$!eDL|fu<Wt(wX$!kDL4-o^8>vQirFCQfIjs(~~(F{~RVOkOCDY
z(f9lTHx65=7**DeS<^&@wuO?*htusoUxDIN_E%?~*tW>mA<R+kuD9`gY!at)lFB}I
zX)Y@N``H}9HVc`|?`W=NHm&D-fT}-;9h$r{gPXCb>XP`{qKzH7!*%vzEaF_r_jd$$
z&3E3kP@9i>_vWL(w>IhbBpijq6pf7Zb_FY3Z<i{crxMNA9^qQC0cYbd>(>ue%g?2X
zz~a@ex8~EPQt2#+Xtfc!KNiQ2pIl4QwpViL_sowf{HVglny&g;aojF?=(710l&bGD
zvI7klBmDCAAmOTbg(&~~+__;QJW^3uQt?-q?VUw@%J0(>Z`*>jp;U%%mkJ)}iOUee
zy{);(9_wDITG=sg{jZ*ndk2SKFrJ}VCdU4D()}R>ROo*?Dd7y<*-}mZh8_CO>lOJJ
zu+@i|HIc<1N9#Q!#c5b?Cr;M<g>`|l`Y~H2+vh2b+%(DrGWcP`qf3!e05OH7i7WJW
zs!_#t#e#}G^;;FpRjCBVU2QM#y>?%UDxzs)Un)g(&P2G5KlEY|8%?SJPpq{1c=dYP
z5jT5GY#UqC_p#0QH`E~#-(TXW!jWB)n_WB4HGw_MzS~j1I7Y@yTIdJ}E-R28X0eMR
zYq(>J+VyIxFnl4{c&VMgrIz}kR#GZ4P!F|xku_D5l|(eZzlABk&W!*+;ghiM&CUKK
z%i2hssx)=H#}IXt`o?1truq{>N8_}O@iyxh+q8pK;CSeIhS}SWd{wwVK-{j_%2#!5
zpRrFA@PlUWH7A~DfY_^krK|Oal1HvD{G>_7mPf2!GeGI6;2x&5=TKdAcw>~>%Phxe
z$Ln!v?I4jZTd$CD)gw0(I-?y14`*d<kv+%?L7H%JMNzrned_6{0sp6m^2T*fcx>7?
zv?&X<>^z#r>T@)2ZHJ|@_LJVIseAx<MRZUBBGKqmFPHC^fzi#kdw+dqW-CFW#s%g=
zsoc5hcPPbn=wEcdbOmy0Y`6@|<Lixker;7qun~PwWoMKYDTu&iv>n~!(w$HUcB*aM
zq_}So@4No`jgzs@d7)DOp^!Nn$l}G{QQYW;dyZ*)biOdIoU%=_zGi2W?)1;tS}h|;
z*^-QYf08Ny!%?P@m2_*SbLxP9-&?9@ttayv@Tk(_Xg0bI3h8>ApEu~_8OR#_{3mbI
zqU;l>5$;%Qd17W~wa__vL0#PLQxL*Vhs_<82Alt%WdG53iO!o*dSFuX-OCNT>;g_g
zTy{LvNE45gocO=Jf@D#JHnY{<D>>79G-04%O`uj?Id$w=cPls_k>{c2n&#onr`Hxq
zj^<(JWim5UaMx>ykKLY3Kf2)N^9wnB|9$@c_RuBLLiwI#g#S+g2OuL2+?l84p$F6J
z{pH&w{?q&I*b>B`SPDN=nOw3=k>f|!G;-?#bw%#tpr%%1ZNKv8YwGM+y~+YMHoAof
zvz0L{>u=G|@zPI+O5Kr1NBzEXsbuSCSFVtAPxM49W>IAqgMtPExj%~b)w9}3FYphH
zZPq~L`(9p@73==*ecwvGJsF?aZaaNg8K}$TM~Y+UY;YU>3Uub2V$j2Og^gr>=~aWK
z;?56_CT%&A6ROOcU+?*q1-5sAm)T{4&)e!XRILzoB3f3nMy*fHQfn#G*d!q}>%y~8
z1LQW;t2W*z@L_CzaWmc~>oxEtuj66E`}{(y57l;2+^n@lTJrj3a84KAkfCBCOA0k5
zUO``(lfN(Ucx0)PQ^(buhuAE0s%&&#hkkD#E&??Bd$kv}WSs~&;I;`H^gGiXBFS6H
z-^bDOA0%QVMC!QJ9jM>t?lD9qgIYoAWhC`9+92iIUPp!STBz=#vA=U19Vj?8q}ArW
zj*e(*_|)rFrcsqYWX+#7%}Hd>_=n_#GKHS%+i@qyAl@MoPs*-rsd4^L^v6sZ{LDjF
zD)E@{XB?eR?Ak-)Ew3E;MI>Bl>8d9FHh)Ti8uS-xV@;Ol9e!Pw!P|5V@^G?gdel}c
zCC2-`$T7Pm`&2ml;u}mY>U)43)qVtA=jBX1FIytm=PzL6u>^P1;4avpicc&zV80!&
z@@MB_@toWF%hGbCTaB#lCT8IGcK__2@9!^Sd7{?3J1MIos?Qk!9w=&4FWqMwQ_Mu-
z6Fn=0iNQ57X0}$-3WuhuMFfg+MocrNBs#yy9wgPcxfJR3MTn>0W}gnw5ent<=YP(5
z>-g;Qj=OtYZgjw&*YPBALyY>;`13h({qsz>uEdNUfQe~0aU@CZO<8ymE)nlgi7I%*
zuFb{VHCXp3>RB@9I&HRB#*K=;B71#oq%KsSW~A{mIcj)}Q7JqA!>1p|Xnb{}i{p(m
zvnz`|ZNFrsxR27^--<d=2Ryq|dG<Hz*IAHS%68!Q7gpbVHhod22%Y~#e#Xi4@4?IA
zRV?xm%qPHmXH@w99_{c|^7$D}Qklk}Qu^9pwh-g;_*267me|&O<p#52>fxn7CJ}>j
z7Vv4m54&FOn)Bm-EiE#?=<p>XvoZC^ltrb37FDcWG+AQ2qZ|cfO(L=ge>F~CcJdE2
zsS=!{<c$NZ?U**_={DU>e2sbs#huU86jN|{WzG7RQ5*kY3@`iak$iNqk2M%F{~rdc
zuzG#_+mdo};te~*f-Tf4|0U9z7su!JZLfdpsW?gCdfP*^vK@kLtzwiPrUsgI0h`W3
zwMwZw)8zi7F>!5ZMS-M%<^D;-_1B$dJEvM>KDIUHW@a1Y^Ey$(nf9CsOELs|GsLyu
z(IpEB*PddmS<4)o_q&Io8JAU1>1P>AlY1xRCd*?=BlSO=b`_f#@bx?aj76}5W_cmh
z<}Ro-jq~;AzU0c9UZ9t<@yZ2*#IzG10R3JoDV0;{llJ(_5<f3h6H-KGtGSIWT3B90
zGTQ^Low;Jo@|TyE9C7`ISYgqk5L_zLXdg`q`+Qo$@rQUO9ZLnh!|27g!#i%2l}m#0
zCLj>pnNySI^~F-ykfRx_siKI4`WF>cGbYnKv0l_FpgkSpsoCH1VPoU{Lnysasre^U
z;eP)}MWbsE_AshGZhOB$%69Q${qI|N%N<1Xko%6;1OM-H>_ZO}^uAY6aIl~mWhfX^
z;j1jNtWP_;_)2n<eB5ofq=ftX5G{E+>2RMrnxgV$I}W&&<MU16Tn-W7p44N7Ts*(9
z*^g1@<%79_0-F+dDI#Yl`s^s`5B58J$k<rPjsofHQzmd@qe)*%oG;QWH(mZ8s@^gz
z$}a30Hb_B`1}UYxdkE=nkOq<N4oPWg$ze$8?i^AY96~^vp;NjWzKiF5pV#~Q=?{*>
zYtP>2UguitTzk}4w9%9*mpFO9Tu<cWEF#JD`#ecxJ8|8FRrI(0l~>oxB#^kzR>{`x
z?ajhBUmcs$Zu=x0pEX`~yWQ6ycKY1lx4D~|oPay7Hh@Y4)XS{9tNs{pAynu0=1JYh
z!pMoLpwmGV_#gxwUoJ^b5to5AxzvHA=Vl*J2*BH6oNi+_Q?E65hd+V@N|s0uQSEtk
z_usM#r6FajlwA3EN6aSMyoN+qA*VAgTE`}pj!CYWP7Erh4J5>Ovc0s()7l*dio>Df
zoiyVeB)~b$<fX^|cFI7{mMisXG*c)cEmJ(MZ%ji)Dy;z7*qz44iI^r^+p_%~iiF<8
zetWl|OaJ56R>!y3=zceO%IEoOAqJ3-G6;BWlyr0}I!h6$8_Coz%6#?UwH1XmovmO-
ztKquBz#}W}`RVyLXL&0ewlbsXUXopcl2yzkoB{23*Q1YH+mx|)@WaBMYjfstbABKQ
zOTzmA)|dariW}Ct_uZ*WV{<nYPe)2=!I>JVzD#yHEMYLJ^ZxUK$D??b@J=4k2|x-n
z+THiPSk6~JzDR)N%e75p6f@GS%@lpy9d-{R%43Bpv};$iL(O2Kx8-6!-nyb0*~9Ba
z!YD#fsS905IDP#=8@l3UyM|?dW^jEqJV+yGXA^}j%4cb6>`Y%w|FoR`tM+zpMm&FN
z=K7!o*pVrW#6a`Kn1!2;^Taeffh9g4tf%m95kpVe595RlA6Qg%J^9po1Fa1!-k_U-
z4BM9+DHb(&_l@pPrl*ZOUVap%qz&h?<)s}6iJ`?^=w;xe&E0{NVeN?Uy=PBlNGh%U
zfJcjgU~-tuEx8utJw~bYCxJr9C2wg2>;3@DX>@vP^F#xT6WdD^ziwMyKa1<(IVct`
zDbhz`G)zNr&2#Z_ayw@5;f-7GFc<rL2>rGQUr=%|RcFE?xOh!aX~4_nvm*)aLqd9L
zR{QgLfp^qW%dcm9Ii4?u|M;ZcHddm@=s6)w)H1n`-4~hLuGeS}vJ{B{-szTCTY%nW
z3EO$tQb5$QzL?_)5l%@Fk77BVt5Ki%247K`JZG+pknl{u*ny=;`<DT4hxd<UJT05}
z)3S+^<CEb1zhx`?pJgi;_Uq9T12bl-pv&_xli=}hMNr}S;19h@Ad;F^95!{bwm0ob
z8(F4q{bo~?sw90|8>;9Lu*$N6k@KL<Vy$!_;_V;4(TrtSaDcoXRLM3#l#;?swDp!q
zv7Z!9=6zQTSm2_UECv3;j)qDsa3WG+V2a)usD%sz16&(cMS4v@;z#k7z%@<DhVX8g
znIjNkk~!vw0OV7O{fulp{qTBtenR~W8xw8O`>-h`T2{8x9ax$}`#wN~g#(9SKafbV
zzC>T%dWe}gny1`!(Kf{WyZIG^9EUVyo(|fqQEWHge*fd13A;VIrlp@Y5ZOe^Ib@tb
z_!0`>goCCvC1-haV-MSO`eqg_;!Z=O9*b;^9oIfs(p#ZGyHx(m1-#JyAW={K4~uKr
zjVP`9#cj1F{i3%_n^RIiuQauVScv$4w&61b3=agV@J2JrS=v-8iMbu1qM<=&BLYIO
z&dcoKt`oysnQkL+|E+1V%U-c#b;1IZxe6=qTDqJ*uC#1Z9e7T7{xin1Lh6(^02led
zuZ1!K$vA_TXxhYxY|c>P_azHGX#BQVN3n_t0R_-XscRpzIA@a&ZMiDSQ>jWHSG{3(
zN`oYKQc^~(G(oFy(G__SiqCsSi<Z=4^goLazctoKhlH!q;wByZ-mHwq$N1@&Y1sZ3
zE$d!zTmy7)g-q_nSEio$H81X4MN^|oO4#wBqC?yHh3Kv5LbZ#9w{4EDLo!3XyQjUN
z%N#Pf=XTGUGNG=Bpg|~U211Nbovy+77x|j%FS(N?++GNiV%Ha3dYeki7^~_bZ~zrK
zbexVnE!)D=vMop>AtU|ovONoUJeY0qDCtWhSg(W6o_zyOC99E)1v>Av#m6ri<0sSv
z2H3=ecK~C6y)Ia5k}X3Chk*=*3G>nIJEKri9gy9oA@znL*B7B1-}^Lxrz1EjV0)4*
zII`!hqy8(g@(xj^P*YU)8uK5gvwiE~{aHr<s#E{1NXk0REdB0dWm?0J0^QbR64rcc
z*O3=w1O&|q^*vR#vEPnzeM6FPP4VpbFeo0I@!pRpvvyt#^sTe{9W#~{Urdlc_C$|s
zm~7Lvr!U-*WYCRd;E*tp>&G;jY$>HH8`{+M)AAsWQW92^RH~FjDJ#^N%}liN+S=t$
z*)}xSOEIds#CU!Pw4S)135>Jmh~{aAqYC6GY57h}d?U4;H7NOmvDp#}y*eaUIe=aj
zcS8R64hE{;`{l^e>XOY`mhOUX-&A{9H3<nCYA;s!mo0c{mVu2r6_MukO)*@Q28Uma
z;;*)E#+Vxc6DC6rX97-p2??+0CwvFwyD1|{W*b6t5%@^aG+h%B_sGyRx_r9LuwZ0}
zKvp#b14icmH%eU@1iTFnL}&dRDr7Z#HR5Da+hZAF_zl#DD1QLqKg-|Wly!TCEva0v
z)539K$*{J0GGe>Ll?_&eQec^bXX7$!eR}1}irp`EEGm}1c0o@bS2o9soEn3Bc%Aer
zrZCMhPGd+xjNExA2Bpeo9+&etEowGg^BL^|0V`p4b#wRb`+c&9z8O`s2g?MnI!hb^
z`R!Z<*79&Wv@7hF9`NsisHC3JH4>6Hq5Usw@$5yIz>B5~!V2=uUrZBcUu|ypcnMm~
zQYx&Rtepmj@^Uzc=GH$J1$c$yal<{`hrg$shkQYy$tl!OEO@(6W18kr3Sqhz_<3kC
zB8!U&dM{LAU0?;RO<<{2+4Kp$2pz~yzf1Wr<Kt<1wj!PU{{HB@7MX8&otc|_VMN8P
zt~?ntWSpdut;K=WeRU(U0h}fIjreq*p7y@(4W;}lsu}$EQSVYS9ne^uVM#B#?`Bzg
z;hR&M%gEXhz$UBm%=)@Tr+OBvO^^>tCM_={(l)MLq8XeV>gH^wFz(MUI@mFL+ic&_
zL>0AD;m{~Go8ZKMil?rZs;%7it(`2MO6zrY+ieKalJ_rqRLAub{;{w!^ON*{`-o?6
zQa~tsv$b2$Eh7{Tt}g4YC{@Y&U}J08pxJ9slVH9zm6L#{7F%Gv=j$zi>kwQ?F?j^g
z6RLSlqm=r>TPs1u2}Q=%Vz)dBffjJ3NMRcAt5fQ1lAe8aPAWW2S7aC!<VzqUntRi6
zT%d;l?@B7lG(BmZc5<=e2Gy|h3vxt^eC_#Z0jP<?cnQUs&6YReId&RpDj!f6cWlp0
z8PMKks(k<9Rq;)We&^xl`wAC=m$V`;b;qPkK&2MKLufuPneTc$fIfi2kj||jwiPIq
z|8Xm{em;ZUDgH}65Oyz=wFHni<x0FWBh8C@5^`nJsdfU)f?9MLKz{@fRPx{lb`&(B
zz_*s_&Xy+|2YD9Sg&>^B0wI_EgR8teCVm~@@M0Exyj!?#{N|$eV}V5F6UzTyoX-$8
z>0SVUgN!(9c~f$tHE>3bff7<WgsW7V${s0Ty<iBi*;AMfcgLC{{F?*RHd7T~CRHUW
zO-RK^^6<_dJoG3WgdQn@Tp+hsEUh;I($bJ%B8-wn;;dUUhBp2Bb(Z%96hNtZ_RwfD
z%(;SnM7J*UF?3t5B<OjvNQU1p3o_5`F`W*5nF0+~BQ`46t{_aYjz210CWoCN+K)_E
znZ4()&`+{$zObe_;&2>ba}d%LJ@nuxXSww06mj3v3Sa+GXG9Uojn>W8bZFSdnv=3#
zC(V;0yQ|c7IGC%y{c+FLb~>WrqxaJQHmTbJ6S2$Ic#k@NR*PP~bksZQ+gV5I^}Tu?
zeyj%lg+7%=S2ct_4sI{<j1|<&oy6+s`U$fTTiviOBrflX$zN`;o%Q-H90_g50;<>p
zYS;C@D)v=!xy_(uE~9!Nb;eZ}hwc-WyD0h)p}I9ssUE#1<=TnV$3xXqT@V+EGMx;w
zprAl>MxjRZh#~&rw^7;E{Mya=RQt~cEYwNI&SP)kXT^f7bxy8wlZan7DN#3S<XPH|
z?$APcdaT0qi8V9Tp}&jh!Bua*?owr{usn`<n7j-Ux#>(Y#_3txiBf8pVVSg3YJIh&
z$t^^v>E70=&@Kg-T~1l=ctHim5nfTn5L-v1o1e-11)KsxVz$*kZjNMDrL`U`$Cd+o
z(eNJ*9UfV==*H?)YKUggQPVdUfiCLuAoi;R?Av#ztvZm9vGrR7TaSJ(SJ(i8)M!)`
zfn1VxQWojgoBX8GeI?ALJ~ZFg4H)~u)j=xA(kav-Y8#5n^Lrwo9Gf2t6t2r;|G26l
z1B6B2OJG152VeXnw?P3WPBu|bl>sZ5b&!ps27x$qhhax0$9~p+s><Kx+#$=y8nO0w
zY!UI$r}cbt4`%G_`hn}A;2-u474saVY1K!dHD*t%xr&0?`=UE4e}`@vikbxeeYRZU
zg8h<IELRz2oEe?&_T7pOOuKI|*T>Iy+o1`WN}2tQD$;g(RL?&MUXhMmL)LeqSln}V
z-3h7t9#qFYGapweZDw}R{f(S~x4~m%Nn>eoaZsUQe$j&`9rybcII$w{dDyNtSTlJ`
z^4KkKf+K_SDkjNF{}co^pPPFFJrVML<2Im~cUrCMvyci}XKPr+p4au-BfqtA(W;*_
zpm|05=cNU~iEbbPBR7S{$jD_2xUsnAi<cN-3%<xznqC7*-RNgOTZBY!KcAQ*-2W95
zp1u1i{i#1Dquwil?YXUGGrc%}iso^!SbTnxM&B?2#mk+ZAO+Cw?SVHNP@qk1vEW~I
zr}s5jb18*;kt$TRYD01G#4vBFyWMIHDu1#mc=2QF?v@|sA9C<~yrZH2O|OX=Im>|-
z1P<Ntyv{+7yGF?RUoQY2dPV(0Cc;&j6^vZ`sJa;{rox)BnBfE77VWzb`@G4c9F?IH
z$?7&SRU=f%ld`o@k%PR0yd#h63$PMf{?~Uv(L?Pg@do(NkZMrsa}IatoTW$q+x?MD
zA@6JdMe?p~&|xblPx#zd$=anI6hUG4+_4JYCmNfRCMhjiYx3=I7`fLvz<_#Of7VfO
zwQ{r+t#|g^PBX-IWqFY3D(PSH!MSo@G6MZr@D)cw``4#_8hQV<`=dHlPn882|Df5n
z!(MY@XX-PTYPQ{AyIJ?~n9ZXjyqCaW&&KI}zWtYICohnQFl2I=Xpk3nu}=9@A~Nz=
zmPHzGa@eV>Ft<2rM@G^kywp=n)*=c3lXM9ga-RzwWalopEMgKh(FE2R+iqDg+TA;+
zdsf}!DD14LaDS?HffW#b7EZbb$zp9~LABJQ{nSM6{p4^emg;9>|2BS!^fB~O|NMBt
z`{-?Is|pB-C5)BMHx3K%otQ&-<Kd6{Wgtrov>{inQLLySEbUQR04@E=qP2^YmWiW)
zviJ*6t}w&L8u+_{G<%Fiyr@rd`o)uJcB+ef``P08H2{^lqo?08_E*EJw14SYyHM#N
zh01|WCsO2pmI;HR>b|`3Z}bI|bJrTG<oE38lW*O{1)tt$Q;SmOTlcDkNPG>#L7*%#
zcg};~ajgwlfD1JtWs(-cw{E@3%NoXwepf-{Ky&zXaszcOjpetGKjk;?dp03tWcJku
z*$mc!4sUK(5z%6`3B3Uo(DXYYpRh7ZZ|K?vq;KcmkGhJ6-~Jg3U+ooP!>C8EOV`1T
z!RpCk9gEit*TEGag`W<Us4LxQ0H2x8Iif2<vQ+~*Sf1%AX!FK2^tml>Tw>9*4(@4~
zGl-uZ5i)dq<I#;T9UD}d=(_of{O--=OEh3*>Ody@EQNl0bqM*pyrx$??rL^i#LT$O
ztceT`t;d@rcatJiG3bf6->(Rl-~2mxj=$f`)VSGnq)Ug|rX&;Niezz%cG3QEb+POo
z0O*-~pa)5WSp)_(8{@3hEgMj0y!LkQ+0Aq53b*JVE=>(ET@lM`7#Lz(o9?(dm!<VP
z88cI+inn4l`q{%_%fU&8#2VAhnXm(s;J|}8_%GdK-XB~}BW(Qk+41z-OI-|XP{ni<
zSr95Iin|?0v(n?u@X1z)9hmntqE14!kz{e?E*X;9&`X}yk_?R>`8?a<Cf-f3(g&Aj
zfdcymA^rQpT;27xFEV@k{gmD}2g=YL+Aw9T5wC>#p(qmPJ~}N~4;GC%X84a%8D`tq
z=%k{)Pn;v(oHK=^Qzg8*lU{AJ2N4z=9<(&Pk{3L=;2%BbBZNo$m#3__DhHQAWM%)a
zlYABcjri&RaN3+ls7)xnQG+`_Y<6D!$-y*~XZmG=+udo_y)y9oaTcaC<kFcV-Z(oo
z{(hY{sHs$-Az(UefX8FWZ^&Yl$ZAt~zf3bTT4A7!#0^ihst1G&%Gq49aa4WP+M%@s
zI8;h1O_f7FH}fwi(e1*uaS-q#*Z0?04X<KN4kC1aoG<Z^(4VSnBH1gIgb*(B(nSDr
zN4&%9x-!U^m+50ghu@B>Gzk&Vn;mR`^^)aJ1^WCRLI^N~m+ZMfomKAD`n}aQo55#h
zkV3;L%X8QDf-RgDZDa4K>;}VgB*+`{-}~+INxISkQ_;IarIQceZg$@+`T0Noe!k)7
zf7NOIAH-VwKZc&HM=#AZ^8wKj&X%97t3!tDU{*xZUs0d%<wox&?EImpBQ0=!QT@OJ
zFWX}NeJaaaj5u32hTK84P<Fp`t88vX%yK>R<YjsGh6-9QZEVnnR{?UhHSfK^yWG*5
z?B+B^!G^u#G>7J-3swgUH~Go4y43nEn>SspR$iDM+_uvD`8CF9z^OjHG^0V-?S`OU
zseNx<-k_*^sW`t@GAq=)7e1jRnGY+>a{gOuGwnpb5y{%t>hSmwbkR}+lm)pY-BO`Y
z&0kRIzcI04N$y&^oW;{abI#@O-(X7<?J+BTZPs;IJnK)Zd^mSF>E&2^W0Zrv`&J%C
zSvw18iNGTinMLA0xA*>d0nQ!%=$$YBg9(fgE)JI-yCPW;PjhyBbXne`5-SvFOt?Fx
zx*GM)#Kq0$<xoJ*#{L}11kMUE^`=7l^kW#T+f`O^{laXV#>fJVeddP|<MO~gIKS?J
zHx#gyC0>=kSM)TL;3Wi^-66Ab>Yz!F1Z9i{CUi^L_JeloCY}(~{h)I9W^lBgQcOk8
zWoynIL3f=&0j^-CN<97K@J4q7<r==j{vQL4d6E2Lsm+N{Gw(CVj^4}^dL)%An@2`^
z(T7i#;&&%77QigUqT|}b%)8W8^$Tq1XZF~+K0lAORN&T|a{3)6)uO&XiOrbEa4uCk
zmGakA%6wc(`+KVz8uj1gwT!#L1OV0pC&tK1YsTKK*|-u3;Se!mgQ4*+D88uUnC6dz
z|Fg&D=Nk{tJ|GbMD@bh9STlfzwYil^>%>o;#LgV{9-9M9_e`FetZ?G3)&gjop}44k
zn)<TbYO~Vsde>!CU<=6F?(qx!!@HdY`g|pn%neVG>~fhY$c<pww)Z_%Y-G`?yV83+
zd~_PcA>pzOQxf+SNr!4z#E<ee#d#v2`HD2`4YK+kx1M)XiefNnkKmz7L-&&zwL=`j
z>kmk23xj#TyEunb6`YwGO*li<TvmMtPXN_cV-Jia{Eas8iTs+OmhMGDVtttE)#(=V
zsnPe{+3-IQaa{g!pJID@?M}5%)k?t4yc-n<LmJmI$sJy2v2*RG-G0YU@%Q<K(x`Ir
z9P7gtK_|+WQexG;;pj;)1NY6$nN*0H<aZv-n|1O(T3B?K$RSi$&;OA$fF2EijJ`fv
zP9090At@hF?ha9TAu<OItqvT!5T{ThgXVAWcrcMs$I&j(z1aBqzkU^US@|Ut&LNa+
zM@w5wlnGRy(l6X53{^vdUFD5;J4KW+t@{lkoNIwF`nChpIk!{a@X|41c1#V<S)EpW
z6saIzey-w6%Fhown-bhdoxVG?T@CRgbFJ;W95&U+i2`G{c(0Oq5H9njZFxn`Szack
zO7|}g&e3M+`i&4-080cIunIp`_%y(Y{4e2`7;~`$pu;^n**{jC2uTv<-E$u~PY2Ba
zP5$PNMAHU_<wn&lg<9g)oISX23KlC?vuUt$t5eH=n|E;L(-C*&S*&k{6yV_h)kyNk
z0{Od?<yZgd317ZA+@G7Yb)%pR4(9(P4j@L)0rpH4+xaz)FN(Z`N%Vc3mnRM$3rU2E
zcoBF3!u%fZ$=3UdR5^TDD7N_XtwV`OGc)FTVTI#w`H?<)=vANSa8Nj`zcM?viB!>&
z32&nfniIvv!mE~kpjA5eXbp^?I%vKntH_d2lS?Pbw@|mjn$i5swbA73sKu)V^oBLb
z*}oUo*n?t#e{ih6@HR*F4F`Q66$^0iq&f~dG=f?>b^>~cyu+0VRClS_ZC>;WYqr78
zyI3u4p$>u}%$ct4$Pcwr*U{BYzbfWVHNn9#wpuZLYw}67{=k!b_$Na>62300*7e5f
z5I%q~XAg=o{as+mK}Yl?!g|j(gnfi<Jk|cy@a>n+aS=-YSQ@^sQiS0Vw8z(H0?_BT
ziZbhaRa!KnAQ_9;0uFn!{5O@J=6Ia6duk*vfL@I0+rG!$!PpOs`M{{$cO^#g!^4#{
z6B0<h5ID(aXwG1p#9R96TR1d@x=5v!w+^Q8jTV<Eu3`SsT^hI{VOzdmEXC6=J_wU~
zu?=0c_z(;-W}3<+jM-YWnsJd@z~z%^*<fp^)HXBMt4!RF8j7Q-?j0*mKA~mCkNXZD
zWd3LDe1rPD`#?e3_!$N<Y<HV?myiR=G?$GX{Gtn<AgKolhx3>CRJ~<&h>g~=cKhPy
zRI@X)UO;Z-Hu?B)@i-up7vL2et<0mBF#s_Y116DU@Z|88=Y8)$?|Jd)8cCGc(EjZr
z40(hGTX<iOOpt!ByGtF7yW<Paz|st9ygy)hL#asj95clN1P`uaz}7~ad3ntWwDPaK
zXiQ*X%8K*3kww7O)M$R{{;Btug;s@b0;|Thu;x%P0=V<QEFv{2>RBn|o2xJEn2>N#
zmJ6P!is-)@UMS*mU;$JT$+O63@DxR`GrWd#+l1Sq+#EcZ8X^|1;&H;nAtg1z4ewbu
zG8Yz-<|T<=GM*#5mkI~qsGY@M<>kS`tWMd@Lcg1Ji|Ml6809B!srzbhw%Jave$l#a
zF4+2`>XKEAGauvwvFnCQ{i1Dy8NZ!{PvF&#8V?aVCs0!+AbYgRcTe6w;2E{oX?;<E
zX69APY*Y{f*7`fltk>vASVg`m9N<j8Y4ps~=ijFKCbeAe|419f-VA-vb-`L)*@)1W
zwC8v~#HIkI@${UVfNA-z09l;wgJZki-Q{%&V;X%mW-eW0U1N!0#?|dM=IP_KMV3qz
zGGE~RX<@(Btr30;OYwz57hl(4fQWu>fnR0jgOqhCWf2>zguJqo5E<O=ox6P;%mJ7l
z?JfB-rW~^z2dB2r+<SOPd9~=`VeRO7!CQs$+cY@`_vPFM0+U^J@H*E`#FZh_eplUu
zAvTXTvVg|dl;MX4$3kVMe)Hx2CBS5V+r9f4xrGR4gFSp|y&iU9(#4sWl3l3|-mnH+
z*HFJ@(j>$bje#G;fT`3b2F0k8+8>&Uspwtnh4Cr@UjdxG_|*|mvY50ml|K55WX%Xd
z&=S2feZC(N{=4puvMHWy8Ng-^L5=Xb*1@JylPgban8keEK&mQnA5NR8^e$6SeHJ7h
z{D+L*jf2c8Z$h_P17fguws=b>$kDJ3bX0F=Q+kN=Igzvu7(Jt8M$%;{tj?CZJjJR>
zV*l{C(6R*;Y*<ckX{wc_9e-u5rn_m*Ha;*oeuo_<U$BPS>)JXXWfR3Xy}J`1y-0bM
zT*-O5BQGV1xg=MdIxIg>NYIB9@k9G3Y&h>1XX%mI%+km(FWhS=P9C`5!Twv$^T&q1
z=k=I>54`}VYk5V{wR<55ODYJKl==>Pku(eUPq0V73We8Ehj}UGfR(geT}yi7vkb%I
z(;xG?wZlzz9(@#yqw+~}4g2#VbVZ$*LC;j|k=`^R{J)Ft`G<G70l?#}c5`=v7<Uo9
zMNlr22-|h-oP#Dc>y=DeVpU}`uKvm&?SPpKIBydc34ayTDZAa9<{Se1HMqrLsmg*-
zLB4+X)14?f1iRdV-|AkUIycU~+0PDjrp(X_>3q5!cM4%6qkFO{Ptz4gxvyLk_?$Ct
zUFKu=`+yTQq&U6^XR359f@Ug;LI{Q3XLNloeP5NNCByC$af<kis4yD$jtfWl&g%d=
z0Ptn_-iNLXDvs{YH?p2Ns2%hrm+R%HIy8E%x3g!LPwmD@j_r?V=YO<7aU`^(nYjDo
zs-)DiG{vaFT0Sc}JP5$dWAP}d(&2J8OyYhtU}^Lme1D=MI+MdzIMC=WS59QRU0(bj
zU;UF66FKDP0lbBlS41Q-RMW46QF!(Pb4+1v6KZ7cYj#kM1@>*j`~wLwVe46re0&71
zFEv_qC%#{7L90)2Cjaq|w1%bv53VFs^?B?brEveb5A?g&#cYRV{G44&FpJkMmG*Y(
zAE-6Qy={l*&bNTHk(^FMkHuasjzVQJWjaqI^EXBUi>{~TAAq-n+Eo|0lg7GC7k!<a
z{bHW^rj!{Y>-9%7sD0i8EF97>@8;s{m`nFjls3}pFrFUPw)Gld#m0bkZY>5xJbW>d
zaX4~wcnQ>(e%h;SbFq)%sufdnaDY0lmx?@qN){mg4@EkX_>F*3$S?5pSf*HiKk8hf
z7B(p`MU1JZMyFXa*~iOrHwTQkhi6V$yCFJn*LwvClj&$zOfC6sW$2iop_#pT_KP_P
zd*~mQdj8?7<P!t6VV&_kSmvc$x--Q<XpVVMFpid}^bMxc)$P(D==O}$@2Kri{-P&G
zh$^m`xbO&+fjY?<OPjcIwLQW3QP*Bc4WETDH?ubl(Z?2Eg_i&79_7THBR}rUFRH<Z
z!dE5T2XH-H%pXel*5)+j=a?(Pu+lm6{~D*yZdIgX(rrh##=dJzJR3R%YzmZdd$Q`v
zh1}h^8T{M*8A=>laWZ=3q=%JHJziuiU|co4?q&qmjM$-l@BvFszXNI}uWiknogw;?
zKs9&4VTWuEzyKP>pG8lc%g1b8x~4m$k5K5xD?aYNMui-^i4(_&+@$oPect6oH^Brp
zr1nR_S=_j5!|W!x@^+Jc3{+g>U$<h<=q_&qG4Wsi+i%7(Ka<PnNsN=3%-?-m*H%M7
zt+yFx9-m`~QwXWwd*w`Q-ZP9BwwAja8hWo}fK^|AHs>m3QB%Lu(xfd=G*)kKc~xq}
z{|H#YYbw*Nsm3>vM`wR{$ykE<QQH-Y-)>g8hFKpBEx&6#g_t&h)YA2TZ|P4c4*Koc
z$0TLIei$D3TIQL#MM@k_mSNrP1y->Eo?qnxj#T?W6ey<+P|?^7`iH}D$OE5jQ-B(j
zaZ{WCG)7sVt%aVVlRGrlgQX6m2VM(Ua(O~pCS`gptI20)^?dSn#dKu_`^~<DL4=C2
zofuwSyM=;CPyms>2uTOPHdN*4wSZ{oXbK5(^e?U2&g9CE{hX24MG#X^6N%ej{;TQF
zP+E-;41M>PW{-|CQU_lhI76favu9!~v}5>Qy=|BRH+*L7t8F&ZUCf1LdY;fL`EolY
zB{AM9QN?BQ!%6pnqAv^xe8T<*7Ft3rF`fxrHKCllgxsrH6RQ?7z*<n&q7D?J!%)`(
z;#TZsotzapwJP-Q_vKmjYVBXnPD{LSAaW+MvP`u~QLnURf}zPxNqc=D+Joj;L5)X=
zcRGhRHUi9AXYDaU?%^byEH1^wQ@rgD(f>nbl3i}+#x{aXTHN>JuOs0#X7clxf8G0N
z<u(|G1u5;A&J4Da+#{!_(lX@A>FH&ZTCBPB_DA?~eWa98YJnO=JJ=W=R((8Z-3Hi0
z4hMgtxGaji_yAoJ30;2v4^=%W%u5}Q*Z!@9>tdZ`a~xchnJOwpvvt6<!%kRQFV3&N
zu|ymSZJ`?W>uVR0M$@s?-O5MGOTy&+3wcDR4`lsC<i2Oe=iS86OvE4=OGRmRJzNw;
zn?OjXUO}%FBy^Ea0#9she_IPlbgPIM+BqFfIXJ$cIkYUfboUVeHFvC8vHe1ou~9Eq
zUxCY74E^vaUd)8=Qj9FHcB$Lkn|rnsKEGuMOJ^H;cXM~8_}yo`{ZTP<T)64x?-G+?
zlOy1woxXcWb<S^a(+h6B01Et$ic2jVQ2K&hB3{RD8axe*@DS-$HOxYa3l13R0jBBG
zdA(d45&fhT`H$V{0GP*%ZX28EDeuGQOTwOyw$C?&J+-@){%s?+L#!FV&+lOy_=X?u
z97rV?1^K;J*T=_jgpvLG4BM7gz2rx*Cs0r@r}ZzZ*J6yI)^6Xg`JkD?gQ8o;l!g^-
zEo5uLh4oh0MR^=t;%M_&u^?#RSAT|`GSbXb){e*}kMc08%iF5mFkYaPkPBU}J9j%1
zIyoP|<nU<o8DmTTkx9fE?JawebeN`i_r2b;$Jr%*$+`kPfZU2;XL8NJhdY35KM06D
zx-<bsIsnpXiuW<gehzrfW%MO7hNHh4fSmBr3BaYV3*sT06+L}a$@;6IlU%5U4rA`#
zwo&2vln+d*WAKBsvu%LlY<idOK2pD<K4zN*f}r!H?i;jtxHtuMVxm3=nmg>OW&D2|
z2*U`$z5Uj{eyWO82amnWuFCl6TeCf1sERcjV&4Q{A|!I?lqdN%V#=JFKB~0q4Y2jv
zIE)arpe~2batIbk!g9E7K_-Ja6qzG)5D?7Otyw$8#g@iC!+T!xI5vUeG^lSQIAOpw
zHtBw9=*M%&ts?t)w);|Ql&J}evkOlW95sHc_JSnZmpg$wrw&yLkDDoBq41u<D#*2O
zxdgQg7;b}FWDx_4GBj9NMm)#H!w$6=HQWp{-<{ew7W)hc!nV!v5EYfriN7W<v@LB|
zoT&I^{#5Hx$&A=kD)Q{dC^}AbSqUi8Oh|d8PkwUX(u<3`GjD0J7$rkuZ@R0(diLwy
zm=gWrpLn;-Ci!W;VSn|Z8=bog+>k&ida{hr&zMK_f}fdlQ{I}WlX?|ve%PX>Z|Shz
z0)B&aJhTiqx?9!iXW_S=T~m(NYqr`o=i2H&?XCVyqPHoAH9G@IM<(`L_lW?U{2XVI
zNah7sYLii7)!0KAZz{Vy)Jv4Uvn%lCLOd3nE**U|?91_e=0D4uj`ZV10Jo>jaY1vh
z)ibFqX;}-|=lnu3cJ>Hq3gS12Qo(#*X5<sftU`PT;}{-fR$IVE0k>grx9KIDa~k-X
zHSif+tQ&Li>dYdJO!bN>3O$z)vqI_dwJKEOwJRB8+g6p5VD>w?Hu^V$4jz3M2WCqu
zBn2HK$d}bXN21?q%=@n);Tw8$HQiyD-`j>lyYiMJU{zJp%oV2M<=%u&=t0TBS5l-b
zczy<e8d~4Go<eXX#adinbN0%FlJEvFjOF<F5_2rU3F^RsH-@UuQK>iadb}O&aMh;c
za-Ddh1<e+Ap#Y0s4=^Tkg)3ROl3VNL{+a9AoLDnuCTiWvM8;>;_8if%sPuDZCR%^1
zJB1FT(&E%P-hHXq*psuJTdXjn_g+@iZNhz)Dlj;Z&!vF3_R~%8o71S|HFjrveC*`N
zs}q3#1H5fL_uYQL%*zJahL+LorT^ot5m9^etZnY3`kVp4^_N4E=k#-8kZsPTYT?<m
z%kH1gP^J%Dbuw8u=VxSFIQ#16SdvS%^V3_<>@}n=@n|Oim%{DEK4{3#7c;3J@t1Fr
z54I!5zjaHgMA@5fbeP!Oq70YN{vB&ff2B5|KtXN?$}}an?MuorcZ2EF+IL@8TrGk}
zZ)=DqdD0CJHeV?VwXF^oe*7KO^N~|$euIkY_IlE{QOKq31q;~)ka#mj5}A&yYKDX8
zMqMo2_NKIdg50P}%f@0e*ua^8y<~EiMes2YpyT8}+X+s%_u{OgBG$vn6cIfd?Q~RK
z`(d0*<P#SZm((Bz>*TXK@z?8O5r_jmeKHoV<J<G6CnvH9m_jErg(6?np6CW6LDr<f
zhc$9mEH8Q!4Qus$_-i8*eRI^zAm?7+F%A(cK`tzz?W(c#)oa(nie)bIys=bS0#;4y
zeWkM)VdktK_fd)p7nAyS?GL#A8l#Abx4B|u7yr=(y3qnI-S@4z=FyklfopRElh-!T
zgnk+J80TeM(UAJ}%Z|i=3D(cBv*I0)ZTr=^afxUi^EIN67&+)n@-&!@irBDJAt9bI
z;**nl$<iX8c?ng2UQRf7ua5Ru%;*Xl!+%%hn0Mb)fX44QZK?u1j@#Egdw$o@gg;)k
z;De5R_JszqooZNl?nus8nf@ST8niqB=t=|-O2_+^_1o`=QcdONa1zoWY&Gj$)#0*4
z>^yyX(7A2b{&~3NoC0gv)b23PxY(hMG`3{cOdZkHh4gLgYn6joeE%8J{}fzSgociD
zd~s&%L07_^&ndD;*k&oQ1q=Fd;c7@09F#J0LU}ZT5V7!GMy<*xaR#Dk6Bc?uane0K
zWa<NvDUo19kMprhX1DH_q&{d^TiLTeFf<P8oF>YBh3$=&rW_k-#PUvLMb37${)|P5
zB}<{$)vEDMcaJt;&R$t44h;6<x_RbzAqvcO+`z%^@vI+HcmXsL5Z$2>la{NFig3E4
zw-fcSYh=*o+1UlS^X#D;A5)8^=CrZOIeD;X8u2%u)8=>xtlf{hb!^O}@NuTMSk*74
zrYsD#Q4{V(DGFZRiFCQbF)NK0W6!Jtoc;5Tx;Ot_$Bx<*Kr+kY$X^+KYpx{r{aFo~
z+<2>qy?oUwjbv1qOtF#spcz23K5+<Q4uFsGU@eyaBGdbTlJX?9V&1Tw3xIILFXOlb
z&X=()wk5+Mwz@f?dH_j%&Ct^@OYr=}W-yK;^&tb+?hBYvG3*HA9_a?1ZJP`i(`Xil
z$_M1{{`SY&Ka&EHXll$4zOQf4Kq5Xfs4o8@5<q``qA(SqQBv4OR1gFjy1!;^B=`6;
z_Vm9oI`&PxRxOcukPiGAqK0_>JNaM^ZduPbkw85Bv3x9I0ZqAxJ2iS!qnEGEd;LVW
zP3O{M$3`<-;|ri01vrqudh7y-LtJ4*zW+!YB<?7C0DOaPJs6*>3o8zUjym#5z^h2(
z$rNZfLrv$7+Af71`>YC=+_vDsv{{cEl>lZsn%k|gAo-;);OASJ#&|~BhrTjYK$<}k
z(=+HxM#uO56~^i3uDYL1pZK965@UsFp69^+ekw4<v{``LuyxJB*7^CYr0u7;BR~Wy
zAS&||_T|l(I@`f_Wf(7W*>3p^B7H+k`{ng&$Z>Ga4pKsTmP|<=X{sl`0TW>0R2Dow
zvntGd3*5+ivs*Or^v`2A7{0$~Ankqz45-UWh-zq-p3C%}h!qE(4KZ-feaWA)^yu5h
zsxWY<FRe%NEuGkCM9$s9l}qxf6({t|*1+Q8H^FZ|J7EliSN_eu`S}5gIjM{^^8W_)
z0L$C+7XXaL$Mtlf4>Q6botq#;3pJ^(UUnAHCuxO6McFF!;mC31e}M)X^8+Ak5wHpl
zkwD}HDuIcc@V|QNb9p(haN$+X-YGbm6O>EB-|Itn#wpfy;Hd_Ib!(eLA9~XEu&wgO
z=Kw+`!fcU(mJ`RdJ3I2|wxoDd@X%6@1C%ifPXAS>L+|VyNf^2x?5j#Ss_b9<ts9V0
zm+t0x0R{(n4!Eg5S!hH>#Yd-Ns++Jzxp&rlx-Oxr!ax+t;RS32tLvnhp?7Z&T^ZX1
z(;?jkGZ^2pI7?sU|LOpG9JZ_Lbk10mMDGzF4M&y#-{oOD-)_KB10HjwVUW2(vZJ$3
z9&BIaOLbm9t)}m+^%OgXP+3>H-R^r-7;;KD;2MYZ6p^wj2W}aYy|K7(IqarccIruf
zIib>CT@p^f;ejJ7!2zHUCR(t9SxQShScZkgMy4Q_Q+UT0XV5Qx3nyY22Nkf7F1$x4
z|M1S%p5IWrQ^XL~1A)-WnQ@7Z1+Q&D#ddqbt<GB|-mM)6cWum@c1?Cj<-;Qhm|73!
zc8OSYOL_lJ6vKy@or(HNDboU?qfc|EiiEVGa(A3k<zufvYK0*D29<PFKuc;BHu7iK
z+I9g<+_~f=B*gzJ<tE+>VD#A&2n$!c9~li6WRsezy8Rp<y0dI+tiitMk92#p63x}7
z)BgE<Oq%RV7kj_((Q@}tXA_vN64a^yAML@=ydJ;(GA_ZPXu&aWC>*EY5{t{PO{ZBC
zO3EJ%dM@w2rl*Y&4T=U0y#709=y(3N-v4e*ae>-Y<xLEJ9lQ@lS&BYx%@f5F`Gxf&
zHgDj|m)oeocVF=PWD2<_{?h;W&J+)Le_NFKHfMmIFCVOLJ<DPaEke>GRHHF%nmqWp
zD<r!NXIDM*Ha8~NiAaEn6_20KhmTQ;wrCVLz9`;_!?v09Vw{n01ttlOH<Q1d!}wUq
zkmxR3uLxJ(|Hpb^_^t|=o~j0HCMI<%X=X_)R2*Pn@Gvssc!Tj{hSOgwUNs+w`=h?B
zhxM&2{_8(wWaR=thyceI_=2Co>WtM4rG+BMMw>&BZRq7lbN|Xg75{>h2qeqb_xqUZ
z#H-;)TEDvo?LA%R%dSd4Zos<)h$ymBmM<@-a-^FY(ij~P_cuz>XCS%7?Gkgg{gkNd
zsEKy(OHaedkmD{W{3=|VV=-KmE49B|Kr8@*$Ee``SzMe%DQxN!axtrrS}AX8xl*<U
zy{2`By=yU#j$UETR9t{lO$0qFI8pX=K6E&}T%*Ds-r#F^_hzIq?Ab4&C1F)hvi}@|
z-KSyQTJ4JQv%%vuvI)xg26OG5sER!9aZx$vmDq|_Y(!Mn=0u2yAjjKFP6*=UNb^db
zn80`T#R6~j;X3>+B*~ysK1x4q&_QY+bUuw?R6&MHb6zrVeqE$v%l#}3v17^rPovQL
zI`EyZ07R|#g+r|H%@q+a1!6n9CQ3V)XlZovZf0ORR7vtMPD1DKl;boAT(G}7%2iwx
zxE``BCYWO#YTPWRn`lkefDh)`PR_Ls{iJj>1#D8Vp^p`Sp{{aczg)#hqjn^5$!%DB
z_|(k)l^da*019CXhy`bfM!qTl;%<Z{OY{Y*=DoY$xOJnv<Rc`aa&Ub!*yZy4A91uG
z1-x6b^pneKis?_CK{JyMudOWg-^3ED<}YwJUIZ(2BGL1{>n=y1vx=6(_d(yHLHza#
zNtrlx+}#|^g-e$&dN4D3GAEhgdzAbO1UH*Q(xkmE>c^TjSSADC>UKhVUtht>qG@rN
zK7+Y!b(fhlcCvpD6*qyK{JjOU^Cp=TJD>^BAendXv=Zs7wpgZX%-gj$7b8fx?56jU
znj-bZXcl}<nWdScm<_9r<~~^Qc6&nPaVzcXs9uho&8g*2krIP3of7JS-pzSpv+Bl<
zgV_=oy8>*%TMG|uY14(szl17*sTaQgYycR_9v}nvf(^5+INxK(g(j@^YYA1#)c8P^
zG=>vZExT<hfoWX7-aWG`=P3td$Aw84fraQaznG05suLRriO*u0aUtjIG9M8;X~Px|
zG+$3$F?&!ZWu=i~*ZA<nCVW)YOirCSWAGO(*7uE>fi3x2zqen<V6P_>^SPUCGus&(
z6wyMOR9oj@m$z6TL%%gkYliF(&DVlTKYnJ+1A?WY{+#c1BP)1s4sJbc8BTd$yV&<h
zbX;DFq28-Dnq&Mv)KCp?+0(gL&F+?HZ`fVBW|M@qj9V876kODQ$Vhv%06zOs#KcqV
zKc78;vNu;>I%OG7|Jntd*uk-OH9pO`^=Bf+R)IoyGIr`x=yK72uDI~w;L(paa>31+
z=g)wbz`o98RQpMC93OOa;~Vd1iO?3P*pewf`*#$bm=Ktg%R;<&4TEJ$LuBnSsmH|&
z2Vf%kai+ST*}%)>xBohd-(q4(7<~?^;P27M_#tt&Ew4|D52kL62}yEu086_ppFZ&C
zrBA4-9B1*vx-zr>X4Ggp8Gr{C__r8}xr_(&bA|exfrJZSRvenJ(`lWNL=Lvv)}+(?
zHHq~8ZxKNw`KLmmu-WL37Qob}%gyD!BcGtkLHCM!sKr0EA^?!(0XoPmC0;uw(p&<V
zfC%1NiR%HJ(aw{x6@S4}eK=3~1AP(<U{-;9iFbptK;ww)ItSQfNFw@Sl)ZyaVCMtz
zbH+Mz5NC8jkl=U0m>Ct^`agQW{P*a>`AG<1hdD3ml}>lNA!;hhguH9Y8{I<-zm$L@
zqWLbul%!t&^s4Xa7W{o1&tLRL=<j@>+DozEDV0l(_gh-EQf%rbK<mFye~UxwV|yK8
z(IsQ+kxrgkF}1teY@hdSBzNOm1zOU$YF3q0B)@Y0lz11U<*EiwG+*m+nEq%5LMH<l
zVdJCs-IV{4AX+~FBGyYFb|U_(?K>j2LTRL^?<5rgL{=P0I?ZQ&z%2tnuV#146p%Vw
zQC@8m9>i!Mdb|oO^Z>YD#HxZ9PSR9QfmsYIMW$ownDqqaRaELyJt4_G-&giHT7!w3
z>#vkpG(M#dt-lR7ecLGIvCukhvFo4_E(HI0A14hKYS@!y<;~^qhsU7jD1Y%{CW@lP
z&G{yU$<grPlVE%!8XxCkFDy)0Om0#EW{e?jWX_e*%9aDf4I%E-g_K_2pAdBDw|-iQ
zmyCQe)+2o>({3))u5cGi^{a1&JA0mdsh_gEy4mU^B`(L9_#QPxpH!N0zy~#r0;o<v
zni`HIZ@7MZ0}RqUTA(YV{F{0>MFnPDJ$Us<L(O||Ugfmk4}@Z2?kF^<R6GDV3id!A
zn5oM6?;7(0EorZAnbmdW%ri)WvhP)#*=O1~#cZsw%ae48%LjhW&9HD&FZaOtE73%a
zzAF9x-88mbW$EE@5{?XZQ)m2b-=I%_le@o?r>{i@hg~qJkL!rO75OzB^U4e`0^MTB
zsMdyl=UmH8Z6Fv@9o?3aiX0J?u4H?WeKOD!<|%ePk@fI@B|VYbol)R1j)I!ce)1FW
zA3t<0Y;%D7%O%h~ZrZjC9&T^I!y}A!#sl4f-rru17G~97-!lOpez}FFjQpPu#|hY<
zU&^^g`b<Cjpi5^wk^qeCv(45BUBA+6ovPuxZUbES9ank;<hAQX+2=W5p{;ySBWhuF
zImO5XWTFDk9`}>@q@SW9fD_a6lRz$7inU{Tfu6oS6QyLRp>?p9aISz0D-hK@U8#M|
z<X}|n5CTI>u^C^Y)0~h^eJLDNP|Nx`(1Mb3$I#h+1_Ruka2+Mu(G<su?|D{DE9^Ti
z3v@?l?I)Uwyo~iyrEqCZw&odcCaMF{^}!IekYx75sAy?yx4mFP;QVJk_M-uqE+wnO
z=Jc%itMljtCO+AJOahyr|ABl!R6>DI-#CH%`#ecmv+dtD6p;A$0iMvc4aE3m#*f1C
zFJPu1TdH_^P@WVRB5S1EP-q=$YQpUGzWOyx^$(LU-p8LbjTEmZE&_{}i?-Z3NTE{=
zLIInt_I$3FtbJ)aju&Q|3m!{Bkw^qSdoyd9p3eNfk+)*OnIGm0Hh$xsOMFVtq^!Ds
z6;kZuJ%J^}I*rQW_`z%QJ;&{Wrs?mxg}SAmQBebgrPJuXJ7`-^ZXBWO#OQDY@Ow0*
z``{yUwpC{BKnLRol8@|U4XqOpHCOMQ4Xw>v%y~f{U}oj?NIXRP#sAY~1G<r)29I6)
z`))VBKSf~?8{Hx`iX<24)&1cFw+&3I+B65Wl>^MzNaX>2sS=(3<AUhKg>#4wkGs3K
zWLGUshZc{AkwDBZg@M0VeofC!`%p{1Y)*!uUE~V^&CvsJC|i$834YfP=T+kisRN96
zD&V?}1uvUL{0v{)TsN9UxW{Ek^1BT)rp=A`yMb0)R6-aN`*2l~^?#G0#4&}ir97Z(
z`Oftqc3D`IOQ91oEw2MUuAd7S=;X{xPygCH3j6rYZcZe;*vM4Qq7g*<=innLK(X~N
zr5zO<u-J|Rnk1-YXIE~l;7PVL*;6cn2B!~K??%|!+z;O^QV&uf6Nb2=!^wU2u&Btp
znI3EfrRoGVVR;0nDP(?VYZ2dQ`Ia@HZ$@R?M>*uW4EDq!<WXD6{QE{Jr#<j`s5~xO
zR^+SEwue!lGN5T*p1(d`-i%DO|2>qofK8mopf|nm*YSR5C^070h=N`H>$^p)rTTjU
z_Gv?pdn`19C&Y5IYmbuKG6dvH^oMWf%ase^mpJy7F(fkdI@j-x#wM~`OyNHTPI#`d
z-Mgg$Bu}FEfq&=`!at7V8Oq&d$5ZnH{OWw4O#)<U_C5Wty4^4Zuc}g#5#~Kcn`3U~
zMD#uWV9B#QUU>9?Lg8k7oH5p+iUs{RgBseo^n$}|aJCrUk|Msp@*wnDMVpc>p3JXb
zPbmA0PoZlvLqo&S1so^09Ddlc^(tj_)(uPepMi`MnF6P-4&!Fmc>Y?%%lJNP4b|Qh
zPBI^-6QTfK>T;d&l3M^5d-o1f8pVBw?;Mx`^rtT^iHj`&Pj9t+mWJ^Nn_>D>e*ue%
zBBnFm$TsQ>i<b5N2q97BByqH%$U44BB1661d_#Ba!h$4lfi2wcEwBIg!>dsC7TXFt
zCpc*p#5oKgL%eZiZ?VngEnMMMgPumAyBk0VIz)4{!ev3f1WXxz|9iMP1yZmPTx}+(
zE^A>HKZ6y38C)#8mpH58$e;R)&o8>Ugb48A4dE{QvXpAqj-$uI6nJl;RDAP<1a_#w
zukF4e7dlC+l_;6ypNZd25Gk4Jo!Ep*?0Bv^4$0#mK<s$?6(yvvU#JSr1ZG_&D2uoN
zc<HH2x$Pq0Oda`d=sYVS)Rwv)PY6Fo89T9VBih6HSTSXVt<*Xul|SXVl|j0<#IHX}
zj{p>`;#f3{{|w`O{fh8HuhDKzO2;fk;rk+0{Ek?0*W?0*5=IPL$dVi|GN|(6XrRy3
zkwXZ3Ht&_D=vL$z_GiF=yT5NJW$*MFXRvnXD%n`Tw~gQjDSm`z0Lfb+PRq3RpFp<F
zc60g8m_dyt<{3n*l1SiuUH!=j+Pu$tmd*yUnOUf6P%v(=pEVB0^Q-e=(9CS9H*0xe
zPtsRZ%_ty6>Y6zwZnmiUVR7TK!U&kl?z|=d&N5(XO#mWrlur$98=Fn?k2dFvxIOC0
z^nu!JJ?tJW{bxtR<)KSz7w)l?a1|LwAI1e}j+i^K$Qb{*lD_`jdjn(;{zPMK52hbB
zfn@@3^rF@2xM7LA{gDaMophb3GspfKBuL8_PpHVhnE0LBWx5JpnSPm{b$@2jO^LdR
z-=658wm}<q`9=dah$_(qo$tkYW!;nn15HS-8BZXrY3Bh0+RQB?KMH!9>F-@cgwik#
zj-Dx1^63j^m+6T(cc1%97*uL9^JS4a0UD5J%#1kIM>zpvnqFEQq|uyBYoRP#Ad)8j
zoa9jm?Owm=x%WF!p^T{>1{+G^1TBLvn!tWfp_YFNK}$t~aRILGxq2kxmkte}B;6o)
zygKNwEg-Pu@<6Ee_B2d04EJdjieq15y!>}D1AdY4pFkiO5Y|X5d{GTRv-)X6WfyGO
zv}~zNtwht>!9??5^)uQ`{fBCkSR()w`Cf|P3d$P2uDBCe#~nkWmHKVB{te7HXCE_=
zA~{k>%O?1H7;nR^10wNkb;>DsiInX&zJ3Pa8br~YoXQBjWz4uHq~tb-{5fn237s>2
zJ$~Y)o7}C*Qzumy{cz>Tcb?#F0%ZFT;k<&h_;Adpn`B@nM=}WQ!Eu<JO=weX^_>gg
zr<%0R0B&WJmuq#%q2oY2)W!b%rRW3x*UDIoCw-9gqKflh>3;SC88P5oTpWlO5GQ2D
zigGFbBsWPKf#?J-O{Hh4i=j;f&az(fnN;^@V)fCSPCVftW6<@iJ63LFA09XH_180X
z93XLuEmH*`T_C0hTl|r(4VuG0O3QYb*SVY9^o_40CSwJgqnHY_!04ewO%Q%)<jM8z
zUYXJ&se~}=YzPmVi;_i@P7!JonRS-^tAv@IwklsBz{T&b2?sb3OZOui(dLnxeQ`X0
z$v0u~z$gm^?F5c)M)99BAgXY_!~aqA+U~0&2GY}Si1e!p`M)1W&ivx`@^Go{n*WOd
zOV-G-ydgVnHJqgJ8DFM~0xW#UqCi%O0zzZftsk$fj8S`kWx*1c_)EY%ovv7c?cL$#
zD}6T>b7o4xz6HrWt{=lvn|Yi%^cdkckm`U{4XFv{uqJF`V8<EVEf6F#39t4MJ4nID
zzKk+n07PnX(D<J?x{R~-I%P#(rhc3)w~13CkP)3^dyjGQcbHGLAR2WbkdPvHdd69R
zIEkK044X+v<AEudZ2lqsKeo+)m*o4Sh<gv66weZGgL-jV|B(dG-Z^elRGIXT_+(2I
z$)sc#sp&2a;!5w%AD^7`WD0*Z<}cN%ygt~_DkFYd&~Ii-^RRPyAjPoD@o45ssm&He
zKtULM#$lstd`P-oZbw^);5z3vVg6QsOg^x|ti`l_y}d3WU*19RpvZ*Gd(VkkUpFdq
zw4S*ijd5gaq!8Jyd%PxjEkf+(%xA6vepi%Z-8ydX^>(^1{E{A8$+iDJV@!Q#D5U+z
z+*1_J&1*_@A&eHL<kAc$+$D48-flqnd-OW57r)$ksl=oARC)3Fwf}!LXhC8*>+uf!
zA&yv8#c+uxHTrjhJ-?1g3GGKVLithL8JX!{&g~bIXn<p7sp)9<$<ffR&C7`0XLpaM
zA9)fzv8QVe<Q8g~fiOB`@oGtoa#F2J)El+yqITbBs;HHigZPI;x;Ih!rZHZpHSBJ{
zh^5s7t&AIeR0Z(N^8UZ(zWbl*_l@6MltM_HkWq?c7BX{?QD%~PY$5B|dxh*(<k)1(
z&R)l+I7aq3IJWGSy}q~iNZ-%n`v-jc;XKacp~vgo_w~B&Ydo*#b=}fgQIF)zt}(M~
zuPkHzin?aEI!}h+L{EVRoe*?cr-l@%d+(+TAQfD})IA0@q~o4%EH1}aV`*&6GY;&l
zImsFHyI8<5mI82j&)BiJr=lbs)+YgMx(dQSEzI#Voo*WUBY2bjU8c7mCiY8f?|%(d
zu_UAD$)IvR9~tAleqwpBWi;CtW}aMMG3R*iu}m5L<x~R6R&R;Hl4xagLu*%qQUgiz
z#0kscBx@T9kIE<LlbHxSRDj+07^VI2k^aq>!!ZJ!yw7b6U59K1V+uwj#MVcGoihpr
zE2wE+e4*s|ODn^=!Cr8F(f#ZQdR~9<_xTDvJ18Jf%3x<MHB24IkUSssH1cJT{wC^g
z{-H%@9frd3ZZIGq?lb&5xb~lElF@ey4N%JJvG{ux*Je1ptnM+JS5u7RE#uu+H65jq
z#2zou3VnTyxLW?uI3y0Ade++O*<s30#i#WoL@SDd!H#nG`VL$Jbk}F|hlQz!G7d7~
zBjM9%MWIU}rLQPf7w5V)PFGQY5Z7|l+Ou6$0|{taRDNSs8b)ShE|{DZD-up$aiw1S
z;aZ-KJhx#ig6zv)fGKTs6n^veYB|PfB|H^al)A&Y0ze&X%5gIp&-UfQ<PTDcJd_iB
zy&OU<U{8}#C*t=_|I(b0ppcQH)CE$5h2w~~Uu4pKlsxoP>)U(xAHLZXu|j=<BPg;M
zygN^0zK&~XebA2wpj$~48`26StR13%Md=$(OMgrFl98D3cJ~u4JG6TL$qiO%d8P04
zC7ErUdnFAGpRymx_JH<HgmlDiC37ZfNTvJiypa#|VZa(zzx#)`N52*lr1&QBN&&Zb
zO-h}#&PyZ!b0=()d7|6(OSIc&dg$xJq6Qe^J#Qs^1Ao!`FQ?Hn`Y^Vm{(vMfb4Sa1
zv!yJKKECp0Ht|<0=MQBHjZO~1G!6EtBkmSMY!gWg!JV|{$1kb-Mk3AyIFIEya;MN{
zp)3A<E~P9q!#Iq3MK8oQztM@(>TZO;`yOWfe5xNQ?a@9MG5fvoO(2FTR%Vh?in8=n
zDo5=g*RB<0eu}SR)Wj<7#+A7f?ENRO#Uek>9MqcK(kATcOunvK0&5-Q;>=YT0VDzo
z?cLQ<(;+|VXU`HQ?mAisCVz0Xb2cSp!rthXk#EW?ju~Rhn3H-qkt$!Tf{{DE;WrU`
zzz}>*oA!ebvq~P{LT>Rna}ByzyeLE7H{Uhv_*H|j*Vr%(U=wa?+>a+fA0?R7Vee5~
zXmb4y)+d+iZmO2A|9rbbVmN>(pQO7I^z!;>q0jVeXcW;l1LmbO=<#W%)p7z%S7O8a
z-E%)v^<X^l8nxfs^rL<}DR&{A)Z}4b_ai4tl}|U}C#}v{*F!^CQ8?e){NcR1&BG<>
zQW+p-UA91ls|(mYN2lK(F%Vn)exRVjho(ohs9}CRH0z@@e#~z5mMAURnLn?GjQm&p
z=cgriPD2Vj=mA#0D4$qjKkf^v1ER?W1dl!R`V8OJp4>T&r<TeGBo3*h%kj8>zPPZP
zV0xW{ifIU4Ctj$(qUGkXtvWsfjo+d=ZaE!2@K&|q%01x`<9G3mz5U<Qr!eO^hdi}h
zc2Vj#9!)Qw9G&*E(IyDiUhN3MvMIkMYv4sOny;dh@@=ftk+_LU{sFIF(ov1q97ska
z(vR^7yGhCL-}5z;XwOTx?0i5$jIFHK;tLz}zV>VDx{6YaMAs}yLe+q>H2BekCy#W*
zuMO`*wOp^pT6m;E?XK9_d>*?(1|KH)a|>YN*;&nL6&-U0hTjB@iV&QV8N7zy=KJXG
z<27fWT;t^oq;u!XI%Kss{{Z}YfFJ2XvV^c>a7>$QPOHsJC@WtwH)^XlW}@g#qAE`@
zx6Psxr~FM<EyKrqTeU9h(9>vrTE|I|wejx_VvjW+s?cXzF@A2oQotcF8sGj|>N09-
zaYKBLJ^D0T4ev_m!zE5;6)IoFobp=|)J_N89!Te(XL$8jp~B$Yc)Zi`V*XQ)=C5Ld
z8P&_S-@J8u-$+G%0WT{ud{Co6r)`dp({yu=!0)!K=5bz-5Qx;ljWa@PGKL%<3FeS$
z<Ke`e^LSy8U_UdcUBlktbg%rp#cMT(N}k(D2TqISDrJ=}MTR+_qPrJqGd;LM8KkrN
z+SQYyn_O#}zgjGNYV#oqS~MDK3wH9l0~SN@s^>IUBTai+7G|+*IWUE{!^h+p-c#7#
zbyn}gHJK$i*^RTPdnE5pKlvyPaHLoX$z(3TSuov;<Jmb$%J!=jKOm@5w?1p0k;~UX
zYw`Ed-(k@5T7Qf^16rq<Yh7CYra4<V%v#5ytQco%)+Xb897R9gB%=3As}4=RtRDNx
z@s`a)9+iGuW|@uPnA<fSZ>x0q8}4mSc1-$bFRC?Ft=C=|%7v}ZZ#7)?BFUJ6^j?ZO
z{al>I?9yE>T4d@Vn2ba97GN6p_cG{J_Y-<-)m@+~PTb>fZuCe-+7^!+4BKrKX?!vF
zn@9VN$Ukt`_&nw}_w72-pRV@ZZR{hW+km!VM}qSGrF2c7;TP|h9%^xGOmaH7Aj%9h
zfiZKO0!rsmUYS@(XHmxe7smuaN^}u7Y4op~-67e^eVRjHw9;Op*|&?McG400>pttR
zkluMoML;buzpPjDQT=DLt|jY|e}<vKwRu*aD48$KWRLpI#D?H?Kb`^y-rfeu+PhCS
zaAWWwm0cn*)v(-(3p8DMam59{Uhbq8NIobV3+t+h!~Ntk;g+@`{Abeb`U9_~SYI+(
zl_B`~#AGiOdN;bBVJMlv=m~3-^h;_+Mhp$;Ta+M+{2AXVheZGt&3{>B6gMQh6@TKZ
zbkY*zZ8XSQ#!3O1briL-#$|(cQ`$>#o&o^ZhgHlJH+x6HV55PT6G;*=E{qAJ?<Qe~
z*Jyl<tNEOzINsxOyZ1h3s!Nl-8#k~5w0peQ?=r~5VExMl=#wZO7%4T~yK*9V^>m&9
z#NZ1ZlDJncct1ds=*9j9y=#9Jxf-$DT#_;R8pu6$uUU9%P2>5e-UVnGGb)Dkg<X^H
zM|RgW{4h2b>?4Q^gApEOSzMB)$CjgLveI0QkavfW-<zHf!JPP9-3kc1n`aRH9Y53j
z5HLz>dl_>(H0<)!-8TFN*qdg%?RW>;sS*TBvxH^PcVVtgVi_gjmWlB3#!K2=lP|tY
zlqft<l8U74EU(rlbxA(hmE1?!;h7PBIwIaDp2Yi8@Es&LVE!Fa?~ABSFbBGzCnU?L
zOtf)4{FkrUNpaHYc`I{pw`=PH@}$KtfRqI*BkKWYJb@JD^Nn1dJ0liR)SLUmm7kA!
zb6~OXt6lH*jrxt7aN1tp;xk!|kY|N%mvH7E_9)3T^L~8oy^3k<`a<<Iaup0^k()3)
z=IxT%s*L4+)1DV_r7q~`(`WS$o)N|x-nNRF+8LHckG;zx3i>Id?JPjM2Nx!hzRsL>
z`VpYx_)mJ7{$Cd9KK9x1P!3aR`4_T;*tnNwurhtc$LYN<8c%wu8tvh#&NK93Ldp-M
zkk{#Q{6F^mViAYmr|)pA%!m_d*;;(F<A1#H{G<7&O`2CH`NkP(KVpqt(&VXxILW2y
z!`L~%yqGbLaA8-wLsEI86#P62zf)7(Uw2d475Jyco>|*6Xy(en#VP%#1o%E-VdSj@
z?Blrea^X8WURdn?1UVCC$o*dC?X7NFnrv-gQ1^mvbzJ!kot{t0+g}G=8-Q67wc$up
z7mN^~Ac}RbK!<tk_NWHgKlfU?-pkSV6ZWypxbnSxc$@lQht&6C$ue*IaY2<k1o&`i
zh9v=-xJ2_2xOP}f``Zj|GNh!Jk#P?dDK@rw7R75^^JsXsykkIw-q*3Am^RQ3)J7Iq
zhIt=M^Qi>s8Za0Ur9meDAlpz?PHyZ4TL&7TA~;aecUWL@6OSwDMG^n}{s6;(`nt1i
z{yvTuxdCLY@}M{ay7Tm2aEMq2kzAF}QA2IQYYzsMa)z!dPr-Djb%75R9vw0Z;9mVf
z*z+@Gfo>kbZHv;q+>#j6_CP3D>o1x6F{H_*lUxmW*@s<Z3Rrb#j_mf!j!x*#lgT4M
zW=AR)lI)R_JI09*!SBh1LUGTIH5PQe1Kz~mtKY+~5IWII!D#(T4Lk&c_c>8@4^*Gf
z0$cx-7OLdP@TZPkCqMUE=4Ar@pLZ28D9d}OEz9FRYwPd?r2iGRbvFR(lCx{<Z_da$
znvF!fG5ql%SvO<VI`keZmM>4bc-N_ii^KH2HYHleWBHFi{`!tb?6(>cLb%RPelLHA
zCX{2>`B5vC&ai3_F6$GVTX;CVu}2?>!z)<s;PC&Du~VLNGg4OC3zoP+8S;aVKG}vV
zDl|r6&o=5wQPGjzE~?nXBqOJ)j^tVkN@X3{ar|aS?D!$(2zf6&?3<cpzHNA@Y(Aey
zrm^hU@)d=0%eb5(wiKZdm$OA$sLLte!fB7d=?0I$e6!?$&zrBp3=I`CzPO205C;Fh
zB11{UsB+t1eGFnt`Ak(-(xl-|Ytf_n@O#W~grst@1Rb>qPeLkac*%nbb#YQ*zaJ~!
zjIWTwAr{8^3w-Fr;$ElbA6pJP2<2beEgGqO-AO<tkw&FoY1Ut&Z*=70HtnH(nD2m=
z`;Z5jo1%HGX{D<9(vJDN#w*F<8lCv78}7sc@X5IeN0lT^*X$((q_u^wURqSSr#vs_
zbhO#5N!>%2jq>$P-#A2``fB{L>q;u^+3`Z%=eUrJ3<!f`pIy#XFF3!C%@NXa=Spm#
zsa)Ad_NvCIE*owYttX19PjZr${TSuqqvz`s^UW1;@G5zH5+d%t!@|a+10OJ19VaKJ
z^-Dy`Vj@1Yk;r1gCiqV5)V&vunvJW}dAl1sMKxt(g|{;s_{|r+QsH%}6s1?Y)3ipO
zeY>i`ew&_Jh5xIDnPp_cn?&lmTr<_gPX~fARtX0r$+<I@Pi?*qaG2#EjO#j147{J}
zW%4aJ%$cC7)>8eL(rv4ed?zHHC7gXM#%xt3V0P7^t&g_$17_ouCS{Etg=j%tVY8tQ
zY*o9oZH(y7-kaMLxaR`@Tk7XPoPg`Bzp6Eh4|A{Xs+etOVY8^l7mFLHk(RIV{g0p8
zj{UH%5m+}_9sUw8hh50KHUw{1a{P-`oUmypp>;+JdZbsE8<e#d5mF>$6`k>nh_kq~
znD&K1fnbmN?QUjjO_QXjEHW&NlxaatEy=I5cnjXDJ&_g+t{Q+m8QXd4w7~9r!>5SW
zO3d4qL`;YR*0kcLMf{>WB|QrakUwkJsjtJryLpY+8dCH!;9pVW(Q#zdZ*6T|Ix^96
zpvutXHR7d$q|b4zy)e#?;I;gfHX<c?^z-caXVPh#$9|i9j<Ul7yKsM6P-}W*+jQTO
zsC$0V0)&-B+l4fansp7)x=<Pyqx4Hwe#$vnmv5vmz`7}<lmO-4UCmy->j%9-4GSC<
zY=F66Dy~q8EMjv(47Nt%O2JI`ImP&&?<U_9aOXKCF^Z<~pkAGlq@miD_kp9Iz5C$e
zB51UB<3CaJR%82x-DZueU9Rw~Eo<BM#mjy$X~vXos4h}bg{aWI9ZhS$pt11m?8GD9
zqpMQ;xG`!gNqLB%X@$^Ahq!rQd!pr?lq5b60~RT5I!q#{$@LDOh2|jJx2u7fsul#2
z+ulL3Q)tZtJF9-)wUcS6$HB2pmm4|`W4)l6he%PbJ0xf9Bqu=0);$bRDqdV0SzDC9
z6RYdw-YK?cR!|au$~AeITw>u=oBleLqK)4sJr!KN&S|tq6ja=qh;?Ug<tb<Wt^YO?
zebsYnc@6V|Gt2w3jc@t}!_(lI1NaQ@OG=S@(no{kwQ*@iFzI^?+{}Wv8KfiVdl4y1
zRZ7(lVVrr#yE?sU?F=zFzT+G(V3M3^Swd<*h4j429nE8%VVvG6SJAx&MnnZQTc()9
z`SOU7E3s;(;j0hYrM~^#oj1)BbNTXPwVNlN>yDin)6#KcexUfw!EQ=x$r!z4XjilL
zT}Wjmuq|=R-6JIb=s1=qE1Tu)Pla!=@yMzt3&s9MX`)pd=)&9G?EdcQt-Z;*l?QWk
z3gsAO!Y>A=Kdc3dt!u5O$PE^GI~|&=k<VktpT<|M%ok*)Fs$wN?Q|n%Aa#3eF=BMB
zzf$vOa1or1cb$58UTn<oBM&r3V&@Ie+k_~J_2nr)E5-x!CeFN`=5XpdmlxM~zU@fH
z+CN&kQDc3)Pkz=nx65!(`1{j4&wWMyK8)ud4@gwi)JFCupOxCAq3&{xj}dczLk<iV
z<+_zk5Q>(*xlwI=Bmi5LcReOGfXQ5GQe_iR*B#l!&#=GJ<6KnIsP!s6yI>)Zs<hx%
z`=(cg0H$ZjLyB0l2$$VogPjF-{eyW7dr(OAbVX)KX|?^sgC_>ux~)<Z#MNs?UMuo6
z+3;#p=-Q_Q>3(U_)AaS|x5}zO>QM-URPD-ZdOCb!Yt-G_^#9pBy!fl0^patrkB)|z
z)QwC9+)rEVg!rY3f=WwWky!^nv^{>FD<Q^Uos2u>B|-+q^__e)mcoGiPS^)5GyUfj
zd#y==NwvjFcG4?#3CPCLqUR-p${{PretV!c=t4rYc{ye?RC>aGE>eYig_%2Ng2oJK
z)g$V_^g+v_rl-=f(WA)6@g>x6I`PQ8x~RyS;Z*pRO%5;@76XhZw>$qWRFSc;rpEPb
z!p*?*Wo_+rLT2ya!}&Wz<um!CyL?ZF%pJyd8tV2ZxuF%xJmyxro7Uls=yxs1sa8WJ
zDuJNlS$rwUplL1fYn!eR1x^{kGWWU>jj<G2<KzcDvHtbaF6km>3S%xi1KM>z8;8y|
z1<p8R!9`MDy*(&BmnN2RegWz{XO-PR_SPjXYn!JS$K9>@QSJlub2%Zqvy6^LToUW<
zM#X8$3x&P6(U{yMOHRe(q_T2Z*B#eltIC?S$?3`G><129%(v#rh6qs9PdDxOi{ShU
zV^kJ;W$@upNZL%+p|kI9peZ@6v2TRDAUTdzL|*^H8+lIduM;BOqE+*GU+fbVRG<1%
zdA$1QE*OKeR*=5|;R$CgQA%Y&CCpU};IBj^L8TaYabm`MiZf9NeEQvcd<UG}@Z>sC
z_}t1>V(X-*rd{uUTjCGb;`!~CCXUv+Tlq$cdVA7uj*ZBtTT61LtQ`g!peS^`4UDtg
zJsRv!v`?4pztpxRFN-fqC`w6^j?~53%1l%riU=9ot1;LZ%M9V8UU^U;6p>|>-~u~c
zfRx;RO>{KKs2=1(H6Nb|VP7?69aq@pIIEh#k6mkDJK74n?(@}l*EZ#LE7B~(t+O!f
z_I*CSGv^Q{aH<$?kN@lAz_F53oUB|RTB>BKja49h5Lod5hL^f#slv+F-d^^F4Fo!=
z6m=DIDi+R7RM6d?u+`y_e0b6c7xXWoU}wU<XmKm{POe014pzqX7-{Rebi$OZ5)06@
z9J`<3&LwkQCFyORB@T4Oo{N(@bN9^HuvB}@dinVRS(C)3b6L%UH6wc6o{U@_{(R&6
zb@*H(QyP@_@TPxV<=1jN-Vf+I&hB#)04K|0A+sC#`)tLr6joG}fcbM`CHmDOLU@oQ
zWr#CUbH?C0X`FcJ&7D|Nk6++qfO{%B>yR^D)B~BMNfJ#V8O_NLR1qoyZ8i2=HTf_D
z0kZ`RR*<5!&l`lDWViAY4@}%{VAb;1lPot_;9=8pzQ+tMM9cl)j3U*7A77Yfb@MPg
zmO>2qxGDEV?MlUI{<zC?x!SffiH_67tCdD)6xLV%4tKyueK5Wta1m|T^I1c03RYa-
z_P5`|3aZ+E<*4>UD5B1}!qCx8%UN2%!w#JQJzcE3tWa*xYAea@Rrv6mUHD9%fVSgb
zrEcN{t`93hm2I?oRBqlLk{^OHM(ob)h?%B{m^WBmLhZJq7gvukb@Smt7}t?%HfrN-
z4yzaDox#=Hvex||)YKWRUEMcu8Wk&UY;@aO*Q6jYJhlizVO;+bUF-C?tnt@B;o13m
z+aeejAodUmp@B&1&Mj}k<q-341#Zar&@XXtap|U#r*&K@@Y=V6{?1FS#>1UQfi6Ui
z1AN=QtD~XpOK}kXY?8hC?SMhb(9bc-w_0Y9Y=bh(U6_@cxk2r)&5Y3^LDUNdzw!uL
z-YUX^C+3LR53mUKqt#T)`pUB1{nJawN_RO(z^Vc`9^+js3J2%9;O6Xe_9|AiG-O%I
zGorW0^chPxQ4Xhdrhe=A>4MPS0T{aY$mLp{nvi8rOQ8nIvf+$y&*v8+JmYW=mDTFK
zqU3Nf(+o}%lXBY<UCbNW(nsXt0#p$E*&$?<rlqvPiAGV$=(mRwsD&icikmg5G2OI+
zyq*@E&*(UA(#_5X{bwaXTqTZOP-9)AXTwDD^8Ti(<_&hJKdLwEeozEj-M*w~OWwo2
z&!<3Oop3>W!`rk#8Z9>76YiLf>^+#0-K*}Baf3j}?H174BzE6o$P;CI=lwJ{*Hj3!
z6u$4u#<f_9&}=o5G|Fso_RzJh<->R;_7xQrRsE}5jJ`l!9w*hQG6hTxO}Ic3`-_lK
zTFy>^r!*CxMj&*vi6=(isn4VAe?o0(9gbUq$KUk_D!)**BC`#p&;pF>#?pf-kX9X5
zapYEe&gpW}354B0;@9~7uNrFF3I(>m9;{u_J-K1kBYrc@0oC&$Xd<Gv62}5D$E`;B
zbOlyJKECrzNq)j5>V)SwAi2Uq^f0nO({jxz>a1z5WgJDCm$`a!FTv5O{a4_4V_z*|
zh5`X|oOrOt&c+7MgEFicZsUTC;Eci&_rLRjh3(|2|4oZ!Oj$h;im%W3<?+&P-@#Hg
z+$j<S{-Gn!l**6rt0mey8LEgJgkYMMaMjT0H|HtJ>+#)V!k__IpvnEkAAvKiD=)JR
z5AQE0ffTEmYOB>ipzNm704wx?X8Xje*W<QkJRh9X{m2$%sg5T{mnU}T%TEWcoS7_W
zNK%Rn$P{eV7`vc0J2YlB%6<s*`{ct8BnuZ3kCc^_w=mC4h<Rm!PWIuX30&x8%nem7
zt%;+0?-56)fN$-Ju?8?HgTCEVc!A|qd4JKyk6qU$>u|yx-!F9@#rkZt%;feOpo4N|
zceJT|F5Ti?@yzR$*g+b&N2EDRWW?tZHkL3c1`Al1i7znoiyUnpG;LNcH}(b-%G0xW
z&ZY-Jdhz1wfbT-y8Pgpk!9bxV^k&rS>L0=-AU^Kx@$QI(f|qTyB;8Sg)&g-g+5;Es
z-_Y(aoDA5c&ojSXVtAS(3zhbL{JPY9Lsf+Wh0n}8gfoQIQKd>V^t7knkJwz+*)5Q=
zCNa?5@8skX<3%Gqi@dttS7#6I=^A6`SQ4CBHPgxw_?rEGZFilX-5w+_BHi(>ifa#J
z@IOvl$6Ju4U$DJYScJL0m5CZmL(L>P?N%<H;!iFa!oZr!l4w=J{_~1vS3J|x(~(EJ
zN>{Q*UP?VQ(oR2o@m{SIPi-v@wto^wZL?nat7yBMZx5w?7}(G263t@RwbS|JEmQW#
z+%Ig2ahhlqg1Li_hN{zr=~<ipbu70K%xv_gk<`Ui@wK`%J2iSTJR=(}r@#r^Gl_+K
zDtl6AD!;uug7BTY?Xuajx&di^3Qqx1JHxb%$$z;XmL~~LqWfXYl5O9}+lQ>o)BrRC
zQ@oU!GTqSGcha>;XVyVhypyA%CTD!3Xw>J2UKUqgEUy`pi6oWNC6i9e-kpgiQ-qPT
z(G_-PMs)VhJZh~xH!4fMYR`c?MxL$8U-U4T?aU9w&bwE`oWCEOW%23<2vgW>i6rIe
zX~o7P7`3UVjwBUvT5r(J(gcObU7VD-JC@VUZSw^lc1>(D^@vt$1I3~~d0CWkc%J|*
z2V-UDZI6EO8HJkI$xdKqH{-<OtmDcCS6O-QuXoI~R?5s|%<;lsma(hYIe;Pi+t)qb
ze&9qZ?QzSXH^cXiF-H^J7EXOFNj1lc1FOYJqoGNiPuI8;r7^Gi$6eMmfwBh|a4d}3
zgTw*7_R*&4zlGJy;1gW-mZZrW5zS4{%7-jbMIa1POG{Svb*F;V*!`8<nqIHEV$5W3
z?s>SOskP5q%hdz<H97=Gy=HA6^zF_}G(KPG)bgf!pRFs6>MACD*set!uzgtsOxwrq
zCMwNU8b8$<K|O!+vr8u6tcA5s%CKdqvsGRaFBKl^BP8&7b-ya+O=ZsMda@sF%p&wJ
zHuo|BJQyuIS(W~`Ml9@r`X2pn&090BuJ<2X&q2dYMq^{<F*kNLe=2?gPCf~`VFZak
zs_J&%wb6@Lz_%*E^aLMs?K8~0abR+Y|GeSwzQ}58&y+9Uz{Oy-ct@eUprq7HR=rD0
z{<N?339_-aDXUt0kU6ga4^aH`qZ5B9hh!!>A)t3~x-x){jr7lJTr7r^6xfd1^>UZ(
zuWJ2ySBJMWd!w54SN2(btL9^uPxli%Y?}z7<lLS3O(ToT_=S>EwNGVw888-)8>$<2
zXWFz3zK-!ASL_dbCvx_bqHcE;IKs4scw$(H?Hx)+hfaP<1bl7N+X*$ix%L3A+68vb
zuzR9UhZ74zQy!ZXvI{Hr`kLU;*tVXv^^FG-noQ449e*BJl|#G8^_rSz+VtkB1tJkC
z*UL{Sb@eJgvQS1vAgZ;};&zM&l*Z+XGsq_N?bfU`%ZuXF!?;8xO+R^bm7E+O?doDA
zmXT*$?5A5P_}U&IEeS9s4QdsaHNJDet$U&p%;gJp+M0ds?o*k?$|n3B*+R<!nsokM
zl$nz!I0LJl>yKOI1T(VA7N`&rY2)y+54=>12QfZBjFEE{WHxcdZV}=$_wl<TEcyM5
zL(+?8d`p)R-dQRYeWf#(OTonebD2NUp%zc#?1A}bN{*c)$C>JN3ut04I35*;)Pi1I
zO~T3uAIymOj2(WNUtD@#Z4<_A@98~h^w!yRnN`9Y*7B>VQ;iq#aqKmq<|ju)pYjOF
z6*P23q?);1WzEUUI#dv{*ed*^=H=s?NfUdXT=!LL^`#Y1hk>nR0xptP+KWF^48$9(
z1(`&<BzwrlVS)U?-74CZVG+0U^U@Vh?k_AwWF(zVBVgy~;P|Y25P8VZ^&56a_oPh^
zRm7?}>4m?$8ER0F{Rb=vPv+v^7TDEm8baND9pr)=(ME`}JAPTh#Bvi(gfKstsyjfR
z+s&s@T8KOuRGVCX2qb>NC|i=gHqgQ}W7x8K!+l%j!N6PJnL~vg8M`k*wiSiTYn5O1
z=HGa_l|x$(j}a;1EUUR^SFIbdf#buznSCJ!!`GAjUljMoj$lj^%z|d7hQtWAW4Q^A
zy+FmF(wj;=@uShM0HcYB@>rUWv-s$a`$INUuY99Xn%f*Yo(q%BuQVxh(%ma5LThiJ
z*2`0Xs3KH4b$P}r_^Yad_Dk|8MLx0Q>^8xtSR8sDYv1aU)<$9!Jl=Ac@3la*<9iAo
zMIj@>`7ZD6o7K8l)(y|hMCU60D$@!g1g*U|{0Ry1b%%)jZgYmwa;0bYE=M_hg%{HC
zQA;k~8b1{kqKe^_wrh|g*Qsv!DPke5Ad9jZyipT!dyk=Q%GJs|w9l5%Nm{{gs7AM-
zh=XM~2u6^|+B|;H(jtC(i#Yl+5Q?$mHH}Lb2kAYo7N5J|gEp--zpvTcntA5YWy=Ld
z!!&kxO8hVR6{jwEY@)P-Vn1G6ytX|$hw;%j%*<Gr8E#}t>BZN#)9)>!3(oQAA_f;?
z&Z!4xQ-sd6Q2;$vcmU&CbJjJ#QF^q70lE9*sddt(I~@7IqEuM_diEO;iH3VdMn<|3
zoyv(>8SD1%m9UD-GR8pG-mwyD*+;Lu8&5cEAadXL-Da<t4($HK&79LcudwHp#nJ{|
z#qqCwUzn@XJxA$q1Png4C}z-7aH$DdDxLX@F{YInk`sr}k4%$EGBv$&U65}aBb5(F
z14Cj_E!jo*I~7Aqfg7<Qy-)Ka(hQo!Tkf)$X}#nw9@5Tot5fIwY?#0H?xfSgeSy`Y
zYX8!@%UM^YoG!{WJv-(Aceri0pHS;25hN#y<rr={pqtm*lat5Qv~@h)U`mQgFONB6
zrPJhMdyJgp77NLS9Z>gkH3`}N+$cMQ+Z1!&0eLKGSU`~2Z(-cho*Zat0hH6edopAq
znq`v}wU1oC<WR_|#mAT@u?o0<xO{9p*KJWX7gbk<oF=p^Ti-C)@@?suR!mP7k8JA*
z4>rAD;MGE>Q0}<EN?+^<<LZKhYfOKQ*JX?0(TZBpBp|ap8sVC6ZS}#VsOgFro9fQ5
zXc1FGt@Mry@Sc+J>H{BX6pZlv532wXn8bNFzV&lv1gn7K14>1z`i06$q~?!uSd~i|
z1R$49O%+4RA;-}pu|?5N6=PLBdONQCJeY#Y!WG<m5%n(Hq63@H#&6yEsuA8Q=r7P@
zw^K{Iz&Y?@p`Et12{~xLI#FCxBgl>6X2bC$1V2y33%cF@vlzpT=eR00)1t%3iYatO
z>P5@!TgwVZSpHaal3)z&nxBgMoT-4i`4@w!G}AA#&Oe<$-kd{d+UV}RY2xf7yV^Dr
zQjmbEdcNcSAuSaj<)QsyRN4<Qvo{YB8)I`*ciA64z2x1XgYRKnS*muPitI3c-}$W?
zc)@gqB_$=m#Ec5(r&1n>_m60`tt^LA^vCKRG&qNEo3Vd6o?~~L12`D+7qs@Mt6F_g
zvePCmAepOg%a?DU<@)e+s;`gjoc)f>PImIBx`lf~N#H`girwq}dtWuihU`rHk?jD0
z?g=(Za)ZA&l2^oeat<zY_VY1b9MOzxg1h_f>3Naj-WjWLcKFGCGV?u_dN%gA!C-%w
znsV*VkVDNGa{cVPA2h}G7b1lp?#2Kxrg>kHg4NvhcbdbwxvzL4l87YSo+gGOW)-MT
zYd@P9+Vhka*%&O_HQ6HvFbpPr(MAFp)%JD2dQowQ_orv^xIiEWZx*A!2r*e1EPws{
z+K8j?-Y=^Hk$$8#qFbeJe`tJbfE|Q&BYZ=ZrkT^cBIR{2s~urW%-8!8T2To5AVVKq
z9#zbux_X7yun^3++EbS=#c;p9it&05em=yw(g#CWZn?Y*A%{ej4MW1wBiTR?U^{Kr
z+82=rH@392bhZ6@n}lyCM8U{K4Y5$s;OM5Zce2)%Q#I$;f`%J(N3wrIIw|l4>$6&^
z!oIUll#L7X45j`IBx$UOs*R5t<#K}2w6wJK`o(#o)PAII?g<r!wV=5%W<rS!=a$i#
zQUm`&>}rr<kIAd)sm#>$S!4Fn*N%()vf(}o>`PDF8sb)&S&y&JtYJ=8G2>uX-N+Mz
zxyH6(Ut=CMNW#ji#5l5~#OIuOMFva8-lHX}20f8wQY=_lq*dp00rr>7jyvYHe9Dz~
zi!>L1%7YP~YG+=SqBwHW?>-!r?Eeh@FQ%X?#Kj3k?(3D8zr1xipsc=j*tlLX+hQ?7
z*W0{;NO9jxs&fT}yV?_0FkG^W+5p+>jM{1^w-{O*O?bn!=M}ga@lNpJDXz$2$Y(R+
zp`h2A6C7sIi>xfm;du_n!b|yaT_5Qs%j=^;V8ih9_;@wIrtU)MxY*K}axWq}ib-_T
z`24GSvCY2W6QEd4lk(WnO3UY-k`}rxKM)s_9#meOI)}1yaQn8hl35D#3Ti!Inm)kb
ziOsk+l-y>H=wTfxMSt?%aAjABu+^5<gPbP}^5lX~vWcUef#|<@h-uJWh<@aTqwASx
zpv?STN*VkEf37_FCx_Wuk>i6V5cix|wK8cn`^_@j7rnYg`en7c#t!bAW&Ps(9bKED
zDbp?JS5YgvMMLTzSYC!z(eIfvWoN^*h}SC@YUk$Wig(B+1@UB5!69q>Oz=B+qvPnq
zmJFrIb8m`&?-fh&`pQ6{I{Akj--Na7NJ3dudoP_&P&w)_W5;(cy&yKCvwWFc0dhUk
zOw+KM&?aC(GJ5&vHsKS|ZU-5NJGAVqUdAJ#!Vx(8gX_P64LrI+5glV=V|QH#UE<+a
zuhE;bDCQ$B814Hl>G`RWZHa^W7T%0G4NwO4I1<3*sMfGjLdf}J0WM`;Zn9~w`>o^^
zXYPYFCC1mx%lS(=b?ndUs6Ul<51RV!G$uczq9M3~*YO!RSOOs~Hj4i&sPH7gKUt2O
zYkr=AcReD-kWpQlmW+lHv7Fmd)#Kkb(EC?G)zX>fe%1DbYY*-5e7^D&lL<4DZjitw
z_%lf2L#Z+ZESRa?`h(Z<A%=TY((~yHf+)5Ld@!L_E`G|dV*@Xq5WK_U<PN@zxW-Un
za$}0#xIX5e154kgd%AHtOj-;JbUxXEo<?~bNUXlzU3nqS*M3idihI7B(b1HqoKe8S
z!Pbhx_)gncM$9XN(+xPeMfbEe{*cC9;qRK*&$xf<2o`n$a(8K3ub}S7n@xKmTN%^v
zwgLCsc1N9<&A?<(h>*~<;(gUoU>kfap`A#ZE#Wr0QCX(EOC@q^Z;Sn`g^a-`&BEhZ
zNJ`){)Mjy=p8sJ9$Iu{+iz^f(fKwI@<f!Z%JNAb;n8odVG5^YPc!NWra&+fzo=63=
zL|Objix^%R{3%QO!vwpC4rX$9DuiiuEtlD@Wrd6Vr@W8tZK$P?fFI(Ql*%>hvO&Pc
z-S=>eOoMX;a-BNycw1i#=ZO-~Ma2U?zTaD`55@ztV;YKh6>;KU-60Gsj%<4Sq+{??
z<hFo&5CZUXhXmN2jv%!g;tKR+$64qv@pa=p1kmj*LG<{}U1Y;LHgFTBafU%_p;T`m
z7D0!uB8%T4lYX05fiS;gu%r0SxA$CwOX9-O)?$APqqV508`6L=!*(?`6)0n`GulPT
zb-<aHc<Oc%ppGY07^^n)dWDUoP%Nu}MXf=9T-3uE$J&Dx2PIBq2mb1DqlK~MB1dIs
zsVr*P$>r!&!Q5MbadzS|Q4H)N$cg@NUfl&_Y$8%57Yo~RR<h72BZxGECKZCVx%b^c
z6BYcp4U{po0`mp3ADIM}W})RFhRoDRy}#^Fv3q0R6s%_rNx5rh0h5;NBCSx3JorXq
zw6J>BO@PAiA%N*Jx4v2`T-Y#0mcp}oQUeQbD!C51UCh>;@v#2ND68IlqYsAcYySTC
zJ$*D`E81S*AVj$QXwD@%sQ4ziZIe**^eWS7qKz>r2?z`D=)Et_u^Ib$U5pYDj|Sw0
z*(!KI%#u&F37JIptMmmwt8$O!0*k06DjH{zS!V@zUHgc&wsm~Na4_H_#EHS!iQoFx
zj$EQ(C9;-lY_$wPv=fN1gt&`0!E5fR0)bSUqLL#!+GiOWvfoCqt<vbfa^MGb%_Cnb
zch?>ti~ZTTq#I37SelUo0dpn{`xe70hi(n{DGV;~OmB?C*#rox``ob!-qjP}5nFfH
zKbyYsM@Xj#Lb?@~4Tmb@YOT&LRRgQskG}{b#^%lQa%I(!Cd2L@;`d@r^Q&h&Ee`si
z2YaP)egVm&*&oL@tJAZZej*tGiuoy&oSbzcf5Jw9`sF_^%vMM?|C;A{SnI~{nu7GV
z0n(wmdY-CNcV_u3poF$gRAJ~z=L*wurNPJANq;}`%z2xYXWZ3v%=A-hM+ul-y`^7v
zv(*0mRM~`xE(TPWoi`KF#x8yKVbm^GRBoIVN;q0BAs^x_F5rJG>`dm3UKBeTZcL|s
z>vQ0X@oB}T66%5;IxCI%4M8>K@l};N4x(wT9U>gve%6kRiC6t}{-!*P7C90xOBNjf
zjiRE$vFgV!Nu9xS3}!XJ3i*c)F%{Y_^*82Miyvz!JooCYaJ=U5*ht{v=>uwxkxo-{
zf#o7aNB1<!-EMdLi2l#^-z<72c0@ZrXjcx^VnN0u%IBzz64zZek{fh5BD+I^L?=5y
z6*EaqTSD^U1PWn$H%tPGL8nCGfptq7yoYzGE6BP`M>TZW2L#t!11cUTjYA20i}Koc
zC<!3$L+ksjqg05QTx7Yt{V|)f)w#Rq&8kkN=mzrMJ-BX}iT51%^c#477dyd=N7Qd$
zAwqQ4<|p*C>J8Ba3&b*AhXpR-TAg;<l!=sq+hR4h2YED)7tU6@9lGq)10S*;ca`OW
z%48E!b=gc@Sq-(ic)^M!7)0cBFc&2+9npo?Q332pFjSb0_TsG{;()wT!>l5~_j8+}
zPN#KTz=wenx38M^?o;J&m6i7oWXs2_7DBTGpu3@%MPKyhsKi?X#)mnWQkU*8ikTVf
zQ<ma*@k-}@rM<K5!mpGCoZAlo?8!FuC@|Uu(LIUjq25a$uW8OK^_gm1*8=l3YBkY9
zXq}1;7dv;C5x{^rGIPq$dkT^m^sI<Y9NueCk5f6?VES!vv9M?GL5TeqJDy!%n?zbC
z3NZr<g)S{`+PXGvnGMOK+=;(b*_{wVjU53}D(KOaviVt1vWhFiaG1Tu@~!J*=evca
z=c)bR_ALBjsTnepKa!g|hjfgZhB($r2L@JaBig3SS&xkE&ZDVU8_Yr{9@-ug2T4a`
zkZb{lpw^SD4KHfwYr^U(7h*g4cwRmO`FAK09nSOWgJP1+MP>p(XyAOGOIPPy8hSIe
z*D71)`_a;^8U3n%DrM8_{Q_pcwVd63Mx6HSNd9$Fm2*#pZnd71lY6|2N40~2fq?*+
zGeP1x57eVG@9hQk=3RX>cBUNb`Y4i-He=qDppOlGm8IHB<s27M8qfD%^*AlyDZk16
zaO6f!Dupb02&h+kTmx*9&Dr&DX~|UBFm>R}q@h~y>ld03i$ep-g+ps^dhF(ASIg8X
zZ@IW<sqB&|IJ;YcRvpA@z~uoSF6Owm`lD+A@yiH^F%{7X1mev6scXq}p}R1x$&rP`
z^|v|9CF1?D<SA>%8pXJ1E0sikn!VUv*={m1BqBx2!!7(JYjBlbPlfFZGC$+7iDh;c
zP{ldOY$d$QQ3mvM*w*&K8wu;&2D8$gN9>ht0x}<&XC9B+sr{g`Ka!HP0LVC}S9s;L
zfm`2U-N7;ua%VbdJVfQyxcK3(9!XkSg3B|+KdyM11C$GCPQ%R;_&aFl_&Zrao{Ca7
z@aJoPjF1VmF7^whJ=E=Y6?H5FF+f4>;W7qXTtv<DVJYr*cI?55O}9KDBgCuz7qAQR
z>%~*;7g|(SSg*q|4C*zyGbj1mA1!b0v7%|&4R^g%`vCQ)jJuC$`goFzTZ<T!#x_}x
z-&SzrFgyU*f{2W-=x^&Hti$EWn0S2hoC==#(b69BHPmolxBRYXm&Z-%pjNbczbwL@
z=U_b%s0YZ#M`e2{J<VCb#zJU?%0n+UhKub-MB5W<Kq>!BZ`?Lbvw1!|T1_rX`4%)M
zmg0TOddW7^NE*qfQ1{6I#!ReWX<;42xiwuhw#xhCJ-(XU4<OB_mAV6;(rsbW157z;
z9xm2$%i)s8fi4D2eX?kM#zNME@tKUS67a%e$PHb8=b(S#EaA!}@N?HKuLU|*q-T{1
zcpm7E^>j3fq?8gY_#_2NLH)Prf6Y6vvat<dF0m2|(=iHvPoPi6{;iN$o}I+E2reJd
zxIP!F8xehcN6@512W{=vWL@MI8A=mc6coK`c54cN^16(Lo%{TjU92GZM2gegEBE{@
zib8H2_p|;!V7Ts@(bz#AX_m*A8iRXkM%R3w*0&Ur;dWY@@w~KKKJaEYBUD<$_Ur`N
z^aP7jSJ3ZAyc7Of53;YF()T91#aE|Z9p~DXZ~#Nh4wcq0QM140aP%!K=|QcK35oFM
zuUK~+^dD0GCM4p~S(f3*#q&}~Abp*_Z|mBGHwmr{G28(Oi0K;UsJ^|ueWXgI%e!8g
z9Q<s%e3Z){n46|2NQ8W<-kbP5vMCB09&O&rLbFSYY7GOXggo@5V<p9E4o@Zyh`!`0
zN&b~T1gGTLpi>CuMfiNRQZpcKj@31Q5;j)kQ3C%|v>*v+UDpmlomm7J?W=l^Z~kFl
zBhGW_0AGvj8Lc{wc#xOFV7Ik~KmWr61onKkAd!S*c;Y+-Jvy-2yHJ*5I?Iho{DFWU
zubXxoD_G*;OxbWm2Af4rYeb!g3HhxG2^>RtU~_g8DOCPQn+b8XoEHuH?D4%#x{2)i
z*p45-MidC4{zcrZ;-%x~<C8{x6UQE78%ctUdxh7A{0{|_Q<)r4(I!mEzX{k`BsbDJ
zOG`P{0!ZxFEB63^oav(WOA^b|8tl=UO6Y>yC9KnqgNTZ9Se=My?VujOvmvR#%VfT0
zWudkB^N2|Bkpa)QIgri+OI|+tY!30i=Npj1La7j)oCSDgePJwU0sG*#S=hcnW1j3s
z@E&eHC1AVYbz*t?VqZygEYhX4VZzoPw%hCNebrR_gW#MB1GdBU*W$&IeOEt^C+M-E
z(?nR_m*oPy;NkF0oK0Rlndj<5pwya4f_ovqHIm~9`&%e7D=zdGP4&i)!uTe?u#Qt|
z8<1B<gGB^C<zR#Sp$xh80i=w#*;xpR#SIGTHb5>`1~I@P<`n>M(db)D@kPW;XPJS6
zKp>_kYVZov8=nzDAKxPA>aU~#Pdks;nVt5D|9Gw%F<{A6x^yF773NXwHbAa-HqP0E
z;F+lbk+UAXt$g(RU3r?{{{W|9NLkB5pBHAzZcG}IM2mv|#^m}y<HjHAr+;k)y3wow
zLjy6(FwiP~faA%$23nI6PvHA|_wVWr@i(kpij%MwBy}{wU&p=ve8%58NeTk?L^_6N
zzsYC+YHHzmIS4Lt^<o=;fAhS0ZqJD$k}qEOKi^aV*r-+h`GRWo5B;GY2Z)2wzwPD4
z0r=-zE(lo)U@!kL=zp7b9WNkj(&jf0ev9Pa1q_1c46#1=-(~*4j^^J#uneBFO#bPA
zml=JIC*VAyxc0yMvI9c4!<YKsWy-t+V)Ke7@VOA`-{0!1z;mi_@Bi;I{~wO(pMCv*
cC;np(5<C7)wRqe%Eb#AzxSUwNsE*hF0WGimTL1t6

literal 0
HcmV?d00001

diff --git a/previews/PR3525/assets/themes/documenter-dark.css b/previews/PR3525/assets/themes/documenter-dark.css
new file mode 100644
index 00000000000..691b83abda7
--- /dev/null
+++ b/previews/PR3525/assets/themes/documenter-dark.css
@@ -0,0 +1,7 @@
+html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .pagination-next,html.theme--documenter-dark .pagination-link,html.theme--documenter-dark .pagination-ellipsis,html.theme--documenter-dark .file-cta,html.theme--documenter-dark .file-name,html.theme--documenter-dark .select select,html.theme--documenter-dark .textarea,html.theme--documenter-dark .input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark .button{-moz-appearance:none;-webkit-appearance:none;align-items:center;border:1px solid transparent;border-radius:.4em;box-shadow:none;display:inline-flex;font-size:1rem;height:2.5em;justify-content:flex-start;line-height:1.5;padding-bottom:calc(0.5em - 1px);padding-left:calc(0.75em - 1px);padding-right:calc(0.75em - 1px);padding-top:calc(0.5em - 1px);position:relative;vertical-align:top}html.theme--documenter-dark .pagination-previous:focus,html.theme--documenter-dark .pagination-next:focus,html.theme--documenter-dark .pagination-link:focus,html.theme--documenter-dark .pagination-ellipsis:focus,html.theme--documenter-dark .file-cta:focus,html.theme--documenter-dark .file-name:focus,html.theme--documenter-dark .select select:focus,html.theme--documenter-dark .textarea:focus,html.theme--documenter-dark .input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:focus,html.theme--documenter-dark .button:focus,html.theme--documenter-dark .is-focused.pagination-previous,html.theme--documenter-dark .is-focused.pagination-next,html.theme--documenter-dark .is-focused.pagination-link,html.theme--documenter-dark .is-focused.pagination-ellipsis,html.theme--documenter-dark .is-focused.file-cta,html.theme--documenter-dark .is-focused.file-name,html.theme--documenter-dark .select select.is-focused,html.theme--documenter-dark .is-focused.textarea,html.theme--documenter-dark .is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-focused.button,html.theme--documenter-dark .pagination-previous:active,html.theme--documenter-dark .pagination-next:active,html.theme--documenter-dark .pagination-link:active,html.theme--documenter-dark .pagination-ellipsis:active,html.theme--documenter-dark .file-cta:active,html.theme--documenter-dark .file-name:active,html.theme--documenter-dark .select select:active,html.theme--documenter-dark .textarea:active,html.theme--documenter-dark .input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:active,html.theme--documenter-dark .button:active,html.theme--documenter-dark .is-active.pagination-previous,html.theme--documenter-dark .is-active.pagination-next,html.theme--documenter-dark .is-active.pagination-link,html.theme--documenter-dark .is-active.pagination-ellipsis,html.theme--documenter-dark .is-active.file-cta,html.theme--documenter-dark .is-active.file-name,html.theme--documenter-dark .select select.is-active,html.theme--documenter-dark .is-active.textarea,html.theme--documenter-dark .is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active,html.theme--documenter-dark .is-active.button{outline:none}html.theme--documenter-dark .pagination-previous[disabled],html.theme--documenter-dark .pagination-next[disabled],html.theme--documenter-dark .pagination-link[disabled],html.theme--documenter-dark .pagination-ellipsis[disabled],html.theme--documenter-dark .file-cta[disabled],html.theme--documenter-dark .file-name[disabled],html.theme--documenter-dark .select select[disabled],html.theme--documenter-dark .textarea[disabled],html.theme--documenter-dark .input[disabled],html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input[disabled],html.theme--documenter-dark .button[disabled],fieldset[disabled] html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark fieldset[disabled] .pagination-previous,fieldset[disabled] html.theme--documenter-dark .pagination-next,html.theme--documenter-dark fieldset[disabled] .pagination-next,fieldset[disabled] html.theme--documenter-dark .pagination-link,html.theme--documenter-dark fieldset[disabled] .pagination-link,fieldset[disabled] html.theme--documenter-dark .pagination-ellipsis,html.theme--documenter-dark fieldset[disabled] .pagination-ellipsis,fieldset[disabled] html.theme--documenter-dark .file-cta,html.theme--documenter-dark fieldset[disabled] .file-cta,fieldset[disabled] html.theme--documenter-dark .file-name,html.theme--documenter-dark fieldset[disabled] .file-name,fieldset[disabled] html.theme--documenter-dark .select select,fieldset[disabled] html.theme--documenter-dark .textarea,fieldset[disabled] html.theme--documenter-dark .input,fieldset[disabled] html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark fieldset[disabled] .select select,html.theme--documenter-dark .select fieldset[disabled] select,html.theme--documenter-dark fieldset[disabled] .textarea,html.theme--documenter-dark fieldset[disabled] .input,html.theme--documenter-dark fieldset[disabled] #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark #documenter .docs-sidebar fieldset[disabled] form.docs-search>input,fieldset[disabled] html.theme--documenter-dark .button,html.theme--documenter-dark fieldset[disabled] .button{cursor:not-allowed}html.theme--documenter-dark .tabs,html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .pagination-next,html.theme--documenter-dark .pagination-link,html.theme--documenter-dark .pagination-ellipsis,html.theme--documenter-dark .breadcrumb,html.theme--documenter-dark .file,html.theme--documenter-dark .button,.is-unselectable{-webkit-touch-callout:none;-webkit-user-select:none;-moz-user-select:none;-ms-user-select:none;user-select:none}html.theme--documenter-dark .navbar-link:not(.is-arrowless)::after,html.theme--documenter-dark .select:not(.is-multiple):not(.is-loading)::after{border:3px solid rgba(0,0,0,0);border-radius:2px;border-right:0;border-top:0;content:" ";display:block;height:0.625em;margin-top:-0.4375em;pointer-events:none;position:absolute;top:50%;transform:rotate(-45deg);transform-origin:center;width:0.625em}html.theme--documenter-dark .admonition:not(:last-child),html.theme--documenter-dark .tabs:not(:last-child),html.theme--documenter-dark .pagination:not(:last-child),html.theme--documenter-dark .message:not(:last-child),html.theme--documenter-dark .level:not(:last-child),html.theme--documenter-dark .breadcrumb:not(:last-child),html.theme--documenter-dark .block:not(:last-child),html.theme--documenter-dark .title:not(:last-child),html.theme--documenter-dark .subtitle:not(:last-child),html.theme--documenter-dark .table-container:not(:last-child),html.theme--documenter-dark .table:not(:last-child),html.theme--documenter-dark .progress:not(:last-child),html.theme--documenter-dark .notification:not(:last-child),html.theme--documenter-dark .content:not(:last-child),html.theme--documenter-dark .box:not(:last-child){margin-bottom:1.5rem}html.theme--documenter-dark .modal-close,html.theme--documenter-dark .delete{-webkit-touch-callout:none;-webkit-user-select:none;-moz-user-select:none;-ms-user-select:none;user-select:none;-moz-appearance:none;-webkit-appearance:none;background-color:rgba(10,10,10,0.2);border:none;border-radius:9999px;cursor:pointer;pointer-events:auto;display:inline-block;flex-grow:0;flex-shrink:0;font-size:0;height:20px;max-height:20px;max-width:20px;min-height:20px;min-width:20px;outline:none;position:relative;vertical-align:top;width:20px}html.theme--documenter-dark .modal-close::before,html.theme--documenter-dark .delete::before,html.theme--documenter-dark .modal-close::after,html.theme--documenter-dark .delete::after{background-color:#fff;content:"";display:block;left:50%;position:absolute;top:50%;transform:translateX(-50%) translateY(-50%) rotate(45deg);transform-origin:center center}html.theme--documenter-dark .modal-close::before,html.theme--documenter-dark .delete::before{height:2px;width:50%}html.theme--documenter-dark .modal-close::after,html.theme--documenter-dark .delete::after{height:50%;width:2px}html.theme--documenter-dark .modal-close:hover,html.theme--documenter-dark .delete:hover,html.theme--documenter-dark .modal-close:focus,html.theme--documenter-dark .delete:focus{background-color:rgba(10,10,10,0.3)}html.theme--documenter-dark .modal-close:active,html.theme--documenter-dark .delete:active{background-color:rgba(10,10,10,0.4)}html.theme--documenter-dark .is-small.modal-close,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.modal-close,html.theme--documenter-dark .is-small.delete,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.delete{height:16px;max-height:16px;max-width:16px;min-height:16px;min-width:16px;width:16px}html.theme--documenter-dark .is-medium.modal-close,html.theme--documenter-dark .is-medium.delete{height:24px;max-height:24px;max-width:24px;min-height:24px;min-width:24px;width:24px}html.theme--documenter-dark .is-large.modal-close,html.theme--documenter-dark .is-large.delete{height:32px;max-height:32px;max-width:32px;min-height:32px;min-width:32px;width:32px}html.theme--documenter-dark .control.is-loading::after,html.theme--documenter-dark .select.is-loading::after,html.theme--documenter-dark .loader,html.theme--documenter-dark .button.is-loading::after{animation:spinAround 500ms infinite linear;border:2px solid #dbdee0;border-radius:9999px;border-right-color:transparent;border-top-color:transparent;content:"";display:block;height:1em;position:relative;width:1em}html.theme--documenter-dark .hero-video,html.theme--documenter-dark .modal-background,html.theme--documenter-dark .modal,html.theme--documenter-dark .image.is-square img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-square img,html.theme--documenter-dark .image.is-square .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-square .has-ratio,html.theme--documenter-dark .image.is-1by1 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by1 img,html.theme--documenter-dark .image.is-1by1 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by1 .has-ratio,html.theme--documenter-dark .image.is-5by4 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by4 img,html.theme--documenter-dark .image.is-5by4 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by4 .has-ratio,html.theme--documenter-dark .image.is-4by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by3 img,html.theme--documenter-dark .image.is-4by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by3 .has-ratio,html.theme--documenter-dark .image.is-3by2 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by2 img,html.theme--documenter-dark .image.is-3by2 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by2 .has-ratio,html.theme--documenter-dark .image.is-5by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by3 img,html.theme--documenter-dark .image.is-5by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by3 .has-ratio,html.theme--documenter-dark .image.is-16by9 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-16by9 img,html.theme--documenter-dark .image.is-16by9 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-16by9 .has-ratio,html.theme--documenter-dark .image.is-2by1 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by1 img,html.theme--documenter-dark .image.is-2by1 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by1 .has-ratio,html.theme--documenter-dark .image.is-3by1 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by1 img,html.theme--documenter-dark .image.is-3by1 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by1 .has-ratio,html.theme--documenter-dark .image.is-4by5 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by5 img,html.theme--documenter-dark .image.is-4by5 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by5 .has-ratio,html.theme--documenter-dark .image.is-3by4 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by4 img,html.theme--documenter-dark .image.is-3by4 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by4 .has-ratio,html.theme--documenter-dark .image.is-2by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by3 img,html.theme--documenter-dark .image.is-2by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by3 .has-ratio,html.theme--documenter-dark .image.is-3by5 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by5 img,html.theme--documenter-dark .image.is-3by5 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by5 .has-ratio,html.theme--documenter-dark .image.is-9by16 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-9by16 img,html.theme--documenter-dark .image.is-9by16 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-9by16 .has-ratio,html.theme--documenter-dark .image.is-1by2 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by2 img,html.theme--documenter-dark .image.is-1by2 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by2 .has-ratio,html.theme--documenter-dark .image.is-1by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by3 img,html.theme--documenter-dark .image.is-1by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by3 .has-ratio,.is-overlay{bottom:0;left:0;position:absolute;right:0;top:0}html.theme--documenter-dark .navbar-burger{-moz-appearance:none;-webkit-appearance:none;appearance:none;background:none;border:none;color:currentColor;font-family:inherit;font-size:1em;margin:0;padding:0}/*! minireset.css v0.0.6 | MIT License | github.com/jgthms/minireset.css */html,body,p,ol,ul,li,dl,dt,dd,blockquote,figure,fieldset,legend,textarea,pre,iframe,hr,h1,h2,h3,h4,h5,h6{margin:0;padding:0}h1,h2,h3,h4,h5,h6{font-size:100%;font-weight:normal}ul{list-style:none}button,input,select,textarea{margin:0}html{box-sizing:border-box}*,*::before,*::after{box-sizing:inherit}img,video{height:auto;max-width:100%}iframe{border:0}table{border-collapse:collapse;border-spacing:0}td,th{padding:0}td:not([align]),th:not([align]){text-align:inherit}.has-text-white{color:#fff !important}a.has-text-white:hover,a.has-text-white:focus{color:#e6e6e6 !important}.has-background-white{background-color:#fff !important}.has-text-black{color:#0a0a0a !important}a.has-text-black:hover,a.has-text-black:focus{color:#000 !important}.has-background-black{background-color:#0a0a0a !important}.has-text-light{color:#ecf0f1 !important}a.has-text-light:hover,a.has-text-light:focus{color:#cfd9db !important}.has-background-light{background-color:#ecf0f1 !important}.has-text-dark{color:#282f2f !important}a.has-text-dark:hover,a.has-text-dark:focus{color:#111414 !important}.has-background-dark{background-color:#282f2f !important}.has-text-primary{color:#375a7f !important}a.has-text-primary:hover,a.has-text-primary:focus{color:#28415b !important}.has-background-primary{background-color:#375a7f !important}.has-text-primary-light{color:#f1f5f9 !important}a.has-text-primary-light:hover,a.has-text-primary-light:focus{color:#cddbe9 !important}.has-background-primary-light{background-color:#f1f5f9 !important}.has-text-primary-dark{color:#4d7eb2 !important}a.has-text-primary-dark:hover,a.has-text-primary-dark:focus{color:#7198c1 !important}.has-background-primary-dark{background-color:#4d7eb2 !important}.has-text-link{color:#1abc9c !important}a.has-text-link:hover,a.has-text-link:focus{color:#148f77 !important}.has-background-link{background-color:#1abc9c !important}.has-text-link-light{color:#edfdf9 !important}a.has-text-link-light:hover,a.has-text-link-light:focus{color:#c0f6ec !important}.has-background-link-light{background-color:#edfdf9 !important}.has-text-link-dark{color:#15987e !important}a.has-text-link-dark:hover,a.has-text-link-dark:focus{color:#1bc5a4 !important}.has-background-link-dark{background-color:#15987e !important}.has-text-info{color:#024c7d !important}a.has-text-info:hover,a.has-text-info:focus{color:#012d4b !important}.has-background-info{background-color:#024c7d !important}.has-text-info-light{color:#ebf7ff !important}a.has-text-info-light:hover,a.has-text-info-light:focus{color:#b9e2fe !important}.has-background-info-light{background-color:#ebf7ff !important}.has-text-info-dark{color:#0e9dfb !important}a.has-text-info-dark:hover,a.has-text-info-dark:focus{color:#40b1fc !important}.has-background-info-dark{background-color:#0e9dfb !important}.has-text-success{color:#008438 !important}a.has-text-success:hover,a.has-text-success:focus{color:#005122 !important}.has-background-success{background-color:#008438 !important}.has-text-success-light{color:#ebfff3 !important}a.has-text-success-light:hover,a.has-text-success-light:focus{color:#b8ffd6 !important}.has-background-success-light{background-color:#ebfff3 !important}.has-text-success-dark{color:#00eb64 !important}a.has-text-success-dark:hover,a.has-text-success-dark:focus{color:#1fff7e !important}.has-background-success-dark{background-color:#00eb64 !important}.has-text-warning{color:#ad8100 !important}a.has-text-warning:hover,a.has-text-warning:focus{color:#7a5b00 !important}.has-background-warning{background-color:#ad8100 !important}.has-text-warning-light{color:#fffaeb !important}a.has-text-warning-light:hover,a.has-text-warning-light:focus{color:#ffedb8 !important}.has-background-warning-light{background-color:#fffaeb !important}.has-text-warning-dark{color:#d19c00 !important}a.has-text-warning-dark:hover,a.has-text-warning-dark:focus{color:#ffbf05 !important}.has-background-warning-dark{background-color:#d19c00 !important}.has-text-danger{color:#9e1b0d !important}a.has-text-danger:hover,a.has-text-danger:focus{color:#6f1309 !important}.has-background-danger{background-color:#9e1b0d !important}.has-text-danger-light{color:#fdeeec !important}a.has-text-danger-light:hover,a.has-text-danger-light:focus{color:#fac3bd !important}.has-background-danger-light{background-color:#fdeeec !important}.has-text-danger-dark{color:#ec311d !important}a.has-text-danger-dark:hover,a.has-text-danger-dark:focus{color:#f05c4c !important}.has-background-danger-dark{background-color:#ec311d !important}.has-text-black-bis{color:#121212 !important}.has-background-black-bis{background-color:#121212 !important}.has-text-black-ter{color:#242424 !important}.has-background-black-ter{background-color:#242424 !important}.has-text-grey-darker{color:#282f2f !important}.has-background-grey-darker{background-color:#282f2f !important}.has-text-grey-dark{color:#343c3d !important}.has-background-grey-dark{background-color:#343c3d !important}.has-text-grey{color:#5e6d6f !important}.has-background-grey{background-color:#5e6d6f !important}.has-text-grey-light{color:#8c9b9d !important}.has-background-grey-light{background-color:#8c9b9d !important}.has-text-grey-lighter{color:#dbdee0 !important}.has-background-grey-lighter{background-color:#dbdee0 !important}.has-text-white-ter{color:#ecf0f1 !important}.has-background-white-ter{background-color:#ecf0f1 !important}.has-text-white-bis{color:#fafafa !important}.has-background-white-bis{background-color:#fafafa !important}.is-flex-direction-row{flex-direction:row !important}.is-flex-direction-row-reverse{flex-direction:row-reverse !important}.is-flex-direction-column{flex-direction:column !important}.is-flex-direction-column-reverse{flex-direction:column-reverse !important}.is-flex-wrap-nowrap{flex-wrap:nowrap !important}.is-flex-wrap-wrap{flex-wrap:wrap !important}.is-flex-wrap-wrap-reverse{flex-wrap:wrap-reverse !important}.is-justify-content-flex-start{justify-content:flex-start !important}.is-justify-content-flex-end{justify-content:flex-end !important}.is-justify-content-center{justify-content:center !important}.is-justify-content-space-between{justify-content:space-between !important}.is-justify-content-space-around{justify-content:space-around !important}.is-justify-content-space-evenly{justify-content:space-evenly !important}.is-justify-content-start{justify-content:start !important}.is-justify-content-end{justify-content:end !important}.is-justify-content-left{justify-content:left !important}.is-justify-content-right{justify-content:right !important}.is-align-content-flex-start{align-content:flex-start !important}.is-align-content-flex-end{align-content:flex-end !important}.is-align-content-center{align-content:center !important}.is-align-content-space-between{align-content:space-between !important}.is-align-content-space-around{align-content:space-around !important}.is-align-content-space-evenly{align-content:space-evenly !important}.is-align-content-stretch{align-content:stretch !important}.is-align-content-start{align-content:start !important}.is-align-content-end{align-content:end !important}.is-align-content-baseline{align-content:baseline !important}.is-align-items-stretch{align-items:stretch !important}.is-align-items-flex-start{align-items:flex-start !important}.is-align-items-flex-end{align-items:flex-end !important}.is-align-items-center{align-items:center !important}.is-align-items-baseline{align-items:baseline !important}.is-align-items-start{align-items:start !important}.is-align-items-end{align-items:end !important}.is-align-items-self-start{align-items:self-start !important}.is-align-items-self-end{align-items:self-end !important}.is-align-self-auto{align-self:auto !important}.is-align-self-flex-start{align-self:flex-start !important}.is-align-self-flex-end{align-self:flex-end !important}.is-align-self-center{align-self:center !important}.is-align-self-baseline{align-self:baseline !important}.is-align-self-stretch{align-self:stretch !important}.is-flex-grow-0{flex-grow:0 !important}.is-flex-grow-1{flex-grow:1 !important}.is-flex-grow-2{flex-grow:2 !important}.is-flex-grow-3{flex-grow:3 !important}.is-flex-grow-4{flex-grow:4 !important}.is-flex-grow-5{flex-grow:5 !important}.is-flex-shrink-0{flex-shrink:0 !important}.is-flex-shrink-1{flex-shrink:1 !important}.is-flex-shrink-2{flex-shrink:2 !important}.is-flex-shrink-3{flex-shrink:3 !important}.is-flex-shrink-4{flex-shrink:4 !important}.is-flex-shrink-5{flex-shrink:5 !important}.is-clearfix::after{clear:both;content:" ";display:table}.is-pulled-left{float:left !important}.is-pulled-right{float:right !important}.is-radiusless{border-radius:0 !important}.is-shadowless{box-shadow:none !important}.is-clickable{cursor:pointer !important;pointer-events:all !important}.is-clipped{overflow:hidden !important}.is-relative{position:relative !important}.is-marginless{margin:0 !important}.is-paddingless{padding:0 !important}.m-0{margin:0 !important}.mt-0{margin-top:0 !important}.mr-0{margin-right:0 !important}.mb-0{margin-bottom:0 !important}.ml-0{margin-left:0 !important}.mx-0{margin-left:0 !important;margin-right:0 !important}.my-0{margin-top:0 !important;margin-bottom:0 !important}.m-1{margin:.25rem !important}.mt-1{margin-top:.25rem !important}.mr-1{margin-right:.25rem !important}.mb-1{margin-bottom:.25rem !important}.ml-1{margin-left:.25rem !important}.mx-1{margin-left:.25rem !important;margin-right:.25rem !important}.my-1{margin-top:.25rem !important;margin-bottom:.25rem !important}.m-2{margin:.5rem !important}.mt-2{margin-top:.5rem !important}.mr-2{margin-right:.5rem !important}.mb-2{margin-bottom:.5rem !important}.ml-2{margin-left:.5rem !important}.mx-2{margin-left:.5rem !important;margin-right:.5rem !important}.my-2{margin-top:.5rem !important;margin-bottom:.5rem !important}.m-3{margin:.75rem !important}.mt-3{margin-top:.75rem !important}.mr-3{margin-right:.75rem !important}.mb-3{margin-bottom:.75rem !important}.ml-3{margin-left:.75rem !important}.mx-3{margin-left:.75rem !important;margin-right:.75rem !important}.my-3{margin-top:.75rem !important;margin-bottom:.75rem !important}.m-4{margin:1rem !important}.mt-4{margin-top:1rem !important}.mr-4{margin-right:1rem !important}.mb-4{margin-bottom:1rem !important}.ml-4{margin-left:1rem !important}.mx-4{margin-left:1rem !important;margin-right:1rem !important}.my-4{margin-top:1rem !important;margin-bottom:1rem !important}.m-5{margin:1.5rem !important}.mt-5{margin-top:1.5rem !important}.mr-5{margin-right:1.5rem !important}.mb-5{margin-bottom:1.5rem !important}.ml-5{margin-left:1.5rem !important}.mx-5{margin-left:1.5rem !important;margin-right:1.5rem !important}.my-5{margin-top:1.5rem !important;margin-bottom:1.5rem !important}.m-6{margin:3rem !important}.mt-6{margin-top:3rem !important}.mr-6{margin-right:3rem !important}.mb-6{margin-bottom:3rem !important}.ml-6{margin-left:3rem !important}.mx-6{margin-left:3rem !important;margin-right:3rem !important}.my-6{margin-top:3rem !important;margin-bottom:3rem !important}.m-auto{margin:auto !important}.mt-auto{margin-top:auto !important}.mr-auto{margin-right:auto !important}.mb-auto{margin-bottom:auto !important}.ml-auto{margin-left:auto !important}.mx-auto{margin-left:auto !important;margin-right:auto !important}.my-auto{margin-top:auto !important;margin-bottom:auto !important}.p-0{padding:0 !important}.pt-0{padding-top:0 !important}.pr-0{padding-right:0 !important}.pb-0{padding-bottom:0 !important}.pl-0{padding-left:0 !important}.px-0{padding-left:0 !important;padding-right:0 !important}.py-0{padding-top:0 !important;padding-bottom:0 !important}.p-1{padding:.25rem !important}.pt-1{padding-top:.25rem !important}.pr-1{padding-right:.25rem !important}.pb-1{padding-bottom:.25rem !important}.pl-1{padding-left:.25rem !important}.px-1{padding-left:.25rem !important;padding-right:.25rem !important}.py-1{padding-top:.25rem !important;padding-bottom:.25rem !important}.p-2{padding:.5rem !important}.pt-2{padding-top:.5rem !important}.pr-2{padding-right:.5rem !important}.pb-2{padding-bottom:.5rem !important}.pl-2{padding-left:.5rem !important}.px-2{padding-left:.5rem !important;padding-right:.5rem !important}.py-2{padding-top:.5rem !important;padding-bottom:.5rem !important}.p-3{padding:.75rem !important}.pt-3{padding-top:.75rem !important}.pr-3{padding-right:.75rem !important}.pb-3{padding-bottom:.75rem !important}.pl-3{padding-left:.75rem !important}.px-3{padding-left:.75rem !important;padding-right:.75rem !important}.py-3{padding-top:.75rem !important;padding-bottom:.75rem !important}.p-4{padding:1rem !important}.pt-4{padding-top:1rem !important}.pr-4{padding-right:1rem !important}.pb-4{padding-bottom:1rem !important}.pl-4{padding-left:1rem !important}.px-4{padding-left:1rem !important;padding-right:1rem !important}.py-4{padding-top:1rem !important;padding-bottom:1rem !important}.p-5{padding:1.5rem !important}.pt-5{padding-top:1.5rem !important}.pr-5{padding-right:1.5rem !important}.pb-5{padding-bottom:1.5rem !important}.pl-5{padding-left:1.5rem !important}.px-5{padding-left:1.5rem !important;padding-right:1.5rem !important}.py-5{padding-top:1.5rem !important;padding-bottom:1.5rem !important}.p-6{padding:3rem !important}.pt-6{padding-top:3rem !important}.pr-6{padding-right:3rem !important}.pb-6{padding-bottom:3rem !important}.pl-6{padding-left:3rem !important}.px-6{padding-left:3rem !important;padding-right:3rem !important}.py-6{padding-top:3rem !important;padding-bottom:3rem !important}.p-auto{padding:auto !important}.pt-auto{padding-top:auto !important}.pr-auto{padding-right:auto !important}.pb-auto{padding-bottom:auto !important}.pl-auto{padding-left:auto !important}.px-auto{padding-left:auto !important;padding-right:auto !important}.py-auto{padding-top:auto !important;padding-bottom:auto !important}.is-size-1{font-size:3rem !important}.is-size-2{font-size:2.5rem !important}.is-size-3{font-size:2rem !important}.is-size-4{font-size:1.5rem !important}.is-size-5{font-size:1.25rem !important}.is-size-6{font-size:1rem !important}.is-size-7,html.theme--documenter-dark .docstring>section>a.docs-sourcelink{font-size:.75rem !important}@media screen and (max-width: 768px){.is-size-1-mobile{font-size:3rem !important}.is-size-2-mobile{font-size:2.5rem !important}.is-size-3-mobile{font-size:2rem !important}.is-size-4-mobile{font-size:1.5rem !important}.is-size-5-mobile{font-size:1.25rem !important}.is-size-6-mobile{font-size:1rem !important}.is-size-7-mobile{font-size:.75rem !important}}@media screen and (min-width: 769px),print{.is-size-1-tablet{font-size:3rem !important}.is-size-2-tablet{font-size:2.5rem !important}.is-size-3-tablet{font-size:2rem !important}.is-size-4-tablet{font-size:1.5rem !important}.is-size-5-tablet{font-size:1.25rem !important}.is-size-6-tablet{font-size:1rem !important}.is-size-7-tablet{font-size:.75rem !important}}@media screen and (max-width: 1055px){.is-size-1-touch{font-size:3rem !important}.is-size-2-touch{font-size:2.5rem !important}.is-size-3-touch{font-size:2rem !important}.is-size-4-touch{font-size:1.5rem !important}.is-size-5-touch{font-size:1.25rem !important}.is-size-6-touch{font-size:1rem !important}.is-size-7-touch{font-size:.75rem !important}}@media screen and (min-width: 1056px){.is-size-1-desktop{font-size:3rem !important}.is-size-2-desktop{font-size:2.5rem !important}.is-size-3-desktop{font-size:2rem !important}.is-size-4-desktop{font-size:1.5rem !important}.is-size-5-desktop{font-size:1.25rem !important}.is-size-6-desktop{font-size:1rem !important}.is-size-7-desktop{font-size:.75rem !important}}@media screen and (min-width: 1216px){.is-size-1-widescreen{font-size:3rem !important}.is-size-2-widescreen{font-size:2.5rem !important}.is-size-3-widescreen{font-size:2rem !important}.is-size-4-widescreen{font-size:1.5rem !important}.is-size-5-widescreen{font-size:1.25rem !important}.is-size-6-widescreen{font-size:1rem !important}.is-size-7-widescreen{font-size:.75rem !important}}@media screen and (min-width: 1408px){.is-size-1-fullhd{font-size:3rem !important}.is-size-2-fullhd{font-size:2.5rem !important}.is-size-3-fullhd{font-size:2rem !important}.is-size-4-fullhd{font-size:1.5rem !important}.is-size-5-fullhd{font-size:1.25rem !important}.is-size-6-fullhd{font-size:1rem !important}.is-size-7-fullhd{font-size:.75rem !important}}.has-text-centered{text-align:center !important}.has-text-justified{text-align:justify !important}.has-text-left{text-align:left !important}.has-text-right{text-align:right !important}@media screen and (max-width: 768px){.has-text-centered-mobile{text-align:center !important}}@media screen and (min-width: 769px),print{.has-text-centered-tablet{text-align:center !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-centered-tablet-only{text-align:center !important}}@media screen and (max-width: 1055px){.has-text-centered-touch{text-align:center !important}}@media screen and (min-width: 1056px){.has-text-centered-desktop{text-align:center !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-centered-desktop-only{text-align:center !important}}@media screen and (min-width: 1216px){.has-text-centered-widescreen{text-align:center !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-centered-widescreen-only{text-align:center !important}}@media screen and (min-width: 1408px){.has-text-centered-fullhd{text-align:center !important}}@media screen and (max-width: 768px){.has-text-justified-mobile{text-align:justify !important}}@media screen and (min-width: 769px),print{.has-text-justified-tablet{text-align:justify !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-justified-tablet-only{text-align:justify !important}}@media screen and (max-width: 1055px){.has-text-justified-touch{text-align:justify !important}}@media screen and (min-width: 1056px){.has-text-justified-desktop{text-align:justify !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-justified-desktop-only{text-align:justify !important}}@media screen and (min-width: 1216px){.has-text-justified-widescreen{text-align:justify !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-justified-widescreen-only{text-align:justify !important}}@media screen and (min-width: 1408px){.has-text-justified-fullhd{text-align:justify !important}}@media screen and (max-width: 768px){.has-text-left-mobile{text-align:left !important}}@media screen and (min-width: 769px),print{.has-text-left-tablet{text-align:left !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-left-tablet-only{text-align:left !important}}@media screen and (max-width: 1055px){.has-text-left-touch{text-align:left !important}}@media screen and (min-width: 1056px){.has-text-left-desktop{text-align:left !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-left-desktop-only{text-align:left !important}}@media screen and (min-width: 1216px){.has-text-left-widescreen{text-align:left !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-left-widescreen-only{text-align:left !important}}@media screen and (min-width: 1408px){.has-text-left-fullhd{text-align:left !important}}@media screen and (max-width: 768px){.has-text-right-mobile{text-align:right !important}}@media screen and (min-width: 769px),print{.has-text-right-tablet{text-align:right !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-right-tablet-only{text-align:right !important}}@media screen and (max-width: 1055px){.has-text-right-touch{text-align:right !important}}@media screen and (min-width: 1056px){.has-text-right-desktop{text-align:right !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-right-desktop-only{text-align:right !important}}@media screen and (min-width: 1216px){.has-text-right-widescreen{text-align:right !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-right-widescreen-only{text-align:right !important}}@media screen and (min-width: 1408px){.has-text-right-fullhd{text-align:right !important}}.is-capitalized{text-transform:capitalize !important}.is-lowercase{text-transform:lowercase !important}.is-uppercase{text-transform:uppercase !important}.is-italic{font-style:italic !important}.is-underlined{text-decoration:underline !important}.has-text-weight-light{font-weight:300 !important}.has-text-weight-normal{font-weight:400 !important}.has-text-weight-medium{font-weight:500 !important}.has-text-weight-semibold{font-weight:600 !important}.has-text-weight-bold{font-weight:700 !important}.is-family-primary{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif !important}.is-family-secondary{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif !important}.is-family-sans-serif{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif !important}.is-family-monospace{font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace !important}.is-family-code{font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace !important}.is-block{display:block !important}@media screen and (max-width: 768px){.is-block-mobile{display:block !important}}@media screen and (min-width: 769px),print{.is-block-tablet{display:block !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-block-tablet-only{display:block !important}}@media screen and (max-width: 1055px){.is-block-touch{display:block !important}}@media screen and (min-width: 1056px){.is-block-desktop{display:block !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-block-desktop-only{display:block !important}}@media screen and (min-width: 1216px){.is-block-widescreen{display:block !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-block-widescreen-only{display:block !important}}@media screen and (min-width: 1408px){.is-block-fullhd{display:block !important}}.is-flex{display:flex !important}@media screen and (max-width: 768px){.is-flex-mobile{display:flex !important}}@media screen and (min-width: 769px),print{.is-flex-tablet{display:flex !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-flex-tablet-only{display:flex !important}}@media screen and (max-width: 1055px){.is-flex-touch{display:flex !important}}@media screen and (min-width: 1056px){.is-flex-desktop{display:flex !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-flex-desktop-only{display:flex !important}}@media screen and (min-width: 1216px){.is-flex-widescreen{display:flex !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-flex-widescreen-only{display:flex !important}}@media screen and (min-width: 1408px){.is-flex-fullhd{display:flex !important}}.is-inline{display:inline !important}@media screen and (max-width: 768px){.is-inline-mobile{display:inline !important}}@media screen and (min-width: 769px),print{.is-inline-tablet{display:inline !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-inline-tablet-only{display:inline !important}}@media screen and (max-width: 1055px){.is-inline-touch{display:inline !important}}@media screen and (min-width: 1056px){.is-inline-desktop{display:inline !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-inline-desktop-only{display:inline !important}}@media screen and (min-width: 1216px){.is-inline-widescreen{display:inline !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-inline-widescreen-only{display:inline !important}}@media screen and (min-width: 1408px){.is-inline-fullhd{display:inline !important}}.is-inline-block{display:inline-block !important}@media screen and (max-width: 768px){.is-inline-block-mobile{display:inline-block !important}}@media screen and (min-width: 769px),print{.is-inline-block-tablet{display:inline-block !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-inline-block-tablet-only{display:inline-block !important}}@media screen and (max-width: 1055px){.is-inline-block-touch{display:inline-block !important}}@media screen and (min-width: 1056px){.is-inline-block-desktop{display:inline-block !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-inline-block-desktop-only{display:inline-block !important}}@media screen and (min-width: 1216px){.is-inline-block-widescreen{display:inline-block !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-inline-block-widescreen-only{display:inline-block !important}}@media screen and (min-width: 1408px){.is-inline-block-fullhd{display:inline-block !important}}.is-inline-flex{display:inline-flex !important}@media screen and (max-width: 768px){.is-inline-flex-mobile{display:inline-flex !important}}@media screen and (min-width: 769px),print{.is-inline-flex-tablet{display:inline-flex !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-inline-flex-tablet-only{display:inline-flex !important}}@media screen and (max-width: 1055px){.is-inline-flex-touch{display:inline-flex !important}}@media screen and (min-width: 1056px){.is-inline-flex-desktop{display:inline-flex !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-inline-flex-desktop-only{display:inline-flex !important}}@media screen and (min-width: 1216px){.is-inline-flex-widescreen{display:inline-flex !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-inline-flex-widescreen-only{display:inline-flex !important}}@media screen and (min-width: 1408px){.is-inline-flex-fullhd{display:inline-flex !important}}.is-hidden{display:none !important}.is-sr-only{border:none !important;clip:rect(0, 0, 0, 0) !important;height:0.01em !important;overflow:hidden !important;padding:0 !important;position:absolute !important;white-space:nowrap !important;width:0.01em !important}@media screen and (max-width: 768px){.is-hidden-mobile{display:none !important}}@media screen and (min-width: 769px),print{.is-hidden-tablet{display:none !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-hidden-tablet-only{display:none !important}}@media screen and (max-width: 1055px){.is-hidden-touch{display:none !important}}@media screen and (min-width: 1056px){.is-hidden-desktop{display:none !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-hidden-desktop-only{display:none !important}}@media screen and (min-width: 1216px){.is-hidden-widescreen{display:none !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-hidden-widescreen-only{display:none !important}}@media screen and (min-width: 1408px){.is-hidden-fullhd{display:none !important}}.is-invisible{visibility:hidden !important}@media screen and (max-width: 768px){.is-invisible-mobile{visibility:hidden !important}}@media screen and (min-width: 769px),print{.is-invisible-tablet{visibility:hidden !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-invisible-tablet-only{visibility:hidden !important}}@media screen and (max-width: 1055px){.is-invisible-touch{visibility:hidden !important}}@media screen and (min-width: 1056px){.is-invisible-desktop{visibility:hidden !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-invisible-desktop-only{visibility:hidden !important}}@media screen and (min-width: 1216px){.is-invisible-widescreen{visibility:hidden !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-invisible-widescreen-only{visibility:hidden !important}}@media screen and (min-width: 1408px){.is-invisible-fullhd{visibility:hidden !important}}html.theme--documenter-dark{/*!
+  Theme: a11y-dark
+  Author: @ericwbailey
+  Maintainer: @ericwbailey
+
+  Based on the Tomorrow Night Eighties theme: https://github.com/isagalaev/highlight.js/blob/master/src/styles/tomorrow-night-eighties.css
+*/}html.theme--documenter-dark html{background-color:#1f2424;font-size:16px;-moz-osx-font-smoothing:grayscale;-webkit-font-smoothing:antialiased;min-width:300px;overflow-x:auto;overflow-y:scroll;text-rendering:optimizeLegibility;text-size-adjust:100%}html.theme--documenter-dark article,html.theme--documenter-dark aside,html.theme--documenter-dark figure,html.theme--documenter-dark footer,html.theme--documenter-dark header,html.theme--documenter-dark hgroup,html.theme--documenter-dark section{display:block}html.theme--documenter-dark body,html.theme--documenter-dark button,html.theme--documenter-dark input,html.theme--documenter-dark optgroup,html.theme--documenter-dark select,html.theme--documenter-dark textarea{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif}html.theme--documenter-dark code,html.theme--documenter-dark pre{-moz-osx-font-smoothing:auto;-webkit-font-smoothing:auto;font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace}html.theme--documenter-dark body{color:#fff;font-size:1em;font-weight:400;line-height:1.5}html.theme--documenter-dark a{color:#1abc9c;cursor:pointer;text-decoration:none}html.theme--documenter-dark a strong{color:currentColor}html.theme--documenter-dark a:hover{color:#1dd2af}html.theme--documenter-dark code{background-color:rgba(255,255,255,0.05);color:#ececec;font-size:.875em;font-weight:normal;padding:.1em}html.theme--documenter-dark hr{background-color:#282f2f;border:none;display:block;height:2px;margin:1.5rem 0}html.theme--documenter-dark img{height:auto;max-width:100%}html.theme--documenter-dark input[type="checkbox"],html.theme--documenter-dark input[type="radio"]{vertical-align:baseline}html.theme--documenter-dark small{font-size:.875em}html.theme--documenter-dark span{font-style:inherit;font-weight:inherit}html.theme--documenter-dark strong{color:#f2f2f2;font-weight:700}html.theme--documenter-dark fieldset{border:none}html.theme--documenter-dark pre{-webkit-overflow-scrolling:touch;background-color:#282f2f;color:#fff;font-size:.875em;overflow-x:auto;padding:1.25rem 1.5rem;white-space:pre;word-wrap:normal}html.theme--documenter-dark pre code{background-color:transparent;color:currentColor;font-size:1em;padding:0}html.theme--documenter-dark table td,html.theme--documenter-dark table th{vertical-align:top}html.theme--documenter-dark table td:not([align]),html.theme--documenter-dark table th:not([align]){text-align:inherit}html.theme--documenter-dark table th{color:#f2f2f2}html.theme--documenter-dark .box{background-color:#343c3d;border-radius:8px;box-shadow:none;color:#fff;display:block;padding:1.25rem}html.theme--documenter-dark a.box:hover,html.theme--documenter-dark a.box:focus{box-shadow:0 0.5em 1em -0.125em rgba(10,10,10,0.1),0 0 0 1px #1abc9c}html.theme--documenter-dark a.box:active{box-shadow:inset 0 1px 2px rgba(10,10,10,0.2),0 0 0 1px #1abc9c}html.theme--documenter-dark .button{background-color:#282f2f;border-color:#4c5759;border-width:1px;color:#375a7f;cursor:pointer;justify-content:center;padding-bottom:calc(0.5em - 1px);padding-left:1em;padding-right:1em;padding-top:calc(0.5em - 1px);text-align:center;white-space:nowrap}html.theme--documenter-dark .button strong{color:inherit}html.theme--documenter-dark .button .icon,html.theme--documenter-dark .button .icon.is-small,html.theme--documenter-dark .button #documenter .docs-sidebar form.docs-search>input.icon,html.theme--documenter-dark #documenter .docs-sidebar .button form.docs-search>input.icon,html.theme--documenter-dark .button .icon.is-medium,html.theme--documenter-dark .button .icon.is-large{height:1.5em;width:1.5em}html.theme--documenter-dark .button .icon:first-child:not(:last-child){margin-left:calc(-0.5em - 1px);margin-right:.25em}html.theme--documenter-dark .button .icon:last-child:not(:first-child){margin-left:.25em;margin-right:calc(-0.5em - 1px)}html.theme--documenter-dark .button .icon:first-child:last-child{margin-left:calc(-0.5em - 1px);margin-right:calc(-0.5em - 1px)}html.theme--documenter-dark .button:hover,html.theme--documenter-dark .button.is-hovered{border-color:#8c9b9d;color:#f2f2f2}html.theme--documenter-dark .button:focus,html.theme--documenter-dark .button.is-focused{border-color:#8c9b9d;color:#17a689}html.theme--documenter-dark .button:focus:not(:active),html.theme--documenter-dark .button.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(26,188,156,0.25)}html.theme--documenter-dark .button:active,html.theme--documenter-dark .button.is-active{border-color:#343c3d;color:#f2f2f2}html.theme--documenter-dark .button.is-text{background-color:transparent;border-color:transparent;color:#fff;text-decoration:underline}html.theme--documenter-dark .button.is-text:hover,html.theme--documenter-dark .button.is-text.is-hovered,html.theme--documenter-dark .button.is-text:focus,html.theme--documenter-dark .button.is-text.is-focused{background-color:#282f2f;color:#f2f2f2}html.theme--documenter-dark .button.is-text:active,html.theme--documenter-dark .button.is-text.is-active{background-color:#1d2122;color:#f2f2f2}html.theme--documenter-dark .button.is-text[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-text{background-color:transparent;border-color:transparent;box-shadow:none}html.theme--documenter-dark .button.is-ghost{background:none;border-color:rgba(0,0,0,0);color:#1abc9c;text-decoration:none}html.theme--documenter-dark .button.is-ghost:hover,html.theme--documenter-dark .button.is-ghost.is-hovered{color:#1abc9c;text-decoration:underline}html.theme--documenter-dark .button.is-white{background-color:#fff;border-color:transparent;color:#0a0a0a}html.theme--documenter-dark .button.is-white:hover,html.theme--documenter-dark .button.is-white.is-hovered{background-color:#f9f9f9;border-color:transparent;color:#0a0a0a}html.theme--documenter-dark .button.is-white:focus,html.theme--documenter-dark .button.is-white.is-focused{border-color:transparent;color:#0a0a0a}html.theme--documenter-dark .button.is-white:focus:not(:active),html.theme--documenter-dark .button.is-white.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(255,255,255,0.25)}html.theme--documenter-dark .button.is-white:active,html.theme--documenter-dark .button.is-white.is-active{background-color:#f2f2f2;border-color:transparent;color:#0a0a0a}html.theme--documenter-dark .button.is-white[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-white{background-color:#fff;border-color:#fff;box-shadow:none}html.theme--documenter-dark .button.is-white.is-inverted{background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .button.is-white.is-inverted:hover,html.theme--documenter-dark .button.is-white.is-inverted.is-hovered{background-color:#000}html.theme--documenter-dark .button.is-white.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-white.is-inverted{background-color:#0a0a0a;border-color:transparent;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-white.is-loading::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}html.theme--documenter-dark .button.is-white.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-white.is-outlined:hover,html.theme--documenter-dark .button.is-white.is-outlined.is-hovered,html.theme--documenter-dark .button.is-white.is-outlined:focus,html.theme--documenter-dark .button.is-white.is-outlined.is-focused{background-color:#fff;border-color:#fff;color:#0a0a0a}html.theme--documenter-dark .button.is-white.is-outlined.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-white.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-white.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-white.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-white.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}html.theme--documenter-dark .button.is-white.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-white.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-white.is-inverted.is-outlined{background-color:transparent;border-color:#0a0a0a;color:#0a0a0a}html.theme--documenter-dark .button.is-white.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-white.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-white.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-white.is-inverted.is-outlined.is-focused{background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .button.is-white.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-white.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-white.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-white.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-white.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-white.is-inverted.is-outlined{background-color:transparent;border-color:#0a0a0a;box-shadow:none;color:#0a0a0a}html.theme--documenter-dark .button.is-black{background-color:#0a0a0a;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-black:hover,html.theme--documenter-dark .button.is-black.is-hovered{background-color:#040404;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-black:focus,html.theme--documenter-dark .button.is-black.is-focused{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-black:focus:not(:active),html.theme--documenter-dark .button.is-black.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(10,10,10,0.25)}html.theme--documenter-dark .button.is-black:active,html.theme--documenter-dark .button.is-black.is-active{background-color:#000;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-black[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-black{background-color:#0a0a0a;border-color:#0a0a0a;box-shadow:none}html.theme--documenter-dark .button.is-black.is-inverted{background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .button.is-black.is-inverted:hover,html.theme--documenter-dark .button.is-black.is-inverted.is-hovered{background-color:#f2f2f2}html.theme--documenter-dark .button.is-black.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-black.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#0a0a0a}html.theme--documenter-dark .button.is-black.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-black.is-outlined{background-color:transparent;border-color:#0a0a0a;color:#0a0a0a}html.theme--documenter-dark .button.is-black.is-outlined:hover,html.theme--documenter-dark .button.is-black.is-outlined.is-hovered,html.theme--documenter-dark .button.is-black.is-outlined:focus,html.theme--documenter-dark .button.is-black.is-outlined.is-focused{background-color:#0a0a0a;border-color:#0a0a0a;color:#fff}html.theme--documenter-dark .button.is-black.is-outlined.is-loading::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}html.theme--documenter-dark .button.is-black.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-black.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-black.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-black.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-black.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-black.is-outlined{background-color:transparent;border-color:#0a0a0a;box-shadow:none;color:#0a0a0a}html.theme--documenter-dark .button.is-black.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-black.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-black.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-black.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-black.is-inverted.is-outlined.is-focused{background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .button.is-black.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-black.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-black.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-black.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}html.theme--documenter-dark .button.is-black.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-black.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-light{background-color:#ecf0f1;border-color:transparent;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-light:hover,html.theme--documenter-dark .button.is-light.is-hovered{background-color:#e5eaec;border-color:transparent;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-light:focus,html.theme--documenter-dark .button.is-light.is-focused{border-color:transparent;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-light:focus:not(:active),html.theme--documenter-dark .button.is-light.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(236,240,241,0.25)}html.theme--documenter-dark .button.is-light:active,html.theme--documenter-dark .button.is-light.is-active{background-color:#dde4e6;border-color:transparent;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-light[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-light{background-color:#ecf0f1;border-color:#ecf0f1;box-shadow:none}html.theme--documenter-dark .button.is-light.is-inverted{background-color:rgba(0,0,0,0.7);color:#ecf0f1}html.theme--documenter-dark .button.is-light.is-inverted:hover,html.theme--documenter-dark .button.is-light.is-inverted.is-hovered{background-color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-light.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-light.is-inverted{background-color:rgba(0,0,0,0.7);border-color:transparent;box-shadow:none;color:#ecf0f1}html.theme--documenter-dark .button.is-light.is-loading::after{border-color:transparent transparent rgba(0,0,0,0.7) rgba(0,0,0,0.7) !important}html.theme--documenter-dark .button.is-light.is-outlined{background-color:transparent;border-color:#ecf0f1;color:#ecf0f1}html.theme--documenter-dark .button.is-light.is-outlined:hover,html.theme--documenter-dark .button.is-light.is-outlined.is-hovered,html.theme--documenter-dark .button.is-light.is-outlined:focus,html.theme--documenter-dark .button.is-light.is-outlined.is-focused{background-color:#ecf0f1;border-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-light.is-outlined.is-loading::after{border-color:transparent transparent #ecf0f1 #ecf0f1 !important}html.theme--documenter-dark .button.is-light.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-light.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-light.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-light.is-outlined.is-loading.is-focused::after{border-color:transparent transparent rgba(0,0,0,0.7) rgba(0,0,0,0.7) !important}html.theme--documenter-dark .button.is-light.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-light.is-outlined{background-color:transparent;border-color:#ecf0f1;box-shadow:none;color:#ecf0f1}html.theme--documenter-dark .button.is-light.is-inverted.is-outlined{background-color:transparent;border-color:rgba(0,0,0,0.7);color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-light.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-light.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-light.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-light.is-inverted.is-outlined.is-focused{background-color:rgba(0,0,0,0.7);color:#ecf0f1}html.theme--documenter-dark .button.is-light.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-light.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-light.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-light.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #ecf0f1 #ecf0f1 !important}html.theme--documenter-dark .button.is-light.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-light.is-inverted.is-outlined{background-color:transparent;border-color:rgba(0,0,0,0.7);box-shadow:none;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .button.is-dark,html.theme--documenter-dark .content kbd.button{background-color:#282f2f;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-dark:hover,html.theme--documenter-dark .content kbd.button:hover,html.theme--documenter-dark .button.is-dark.is-hovered,html.theme--documenter-dark .content kbd.button.is-hovered{background-color:#232829;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-dark:focus,html.theme--documenter-dark .content kbd.button:focus,html.theme--documenter-dark .button.is-dark.is-focused,html.theme--documenter-dark .content kbd.button.is-focused{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-dark:focus:not(:active),html.theme--documenter-dark .content kbd.button:focus:not(:active),html.theme--documenter-dark .button.is-dark.is-focused:not(:active),html.theme--documenter-dark .content kbd.button.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(40,47,47,0.25)}html.theme--documenter-dark .button.is-dark:active,html.theme--documenter-dark .content kbd.button:active,html.theme--documenter-dark .button.is-dark.is-active,html.theme--documenter-dark .content kbd.button.is-active{background-color:#1d2122;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-dark[disabled],html.theme--documenter-dark .content kbd.button[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-dark,fieldset[disabled] html.theme--documenter-dark .content kbd.button{background-color:#282f2f;border-color:#282f2f;box-shadow:none}html.theme--documenter-dark .button.is-dark.is-inverted,html.theme--documenter-dark .content kbd.button.is-inverted{background-color:#fff;color:#282f2f}html.theme--documenter-dark .button.is-dark.is-inverted:hover,html.theme--documenter-dark .content kbd.button.is-inverted:hover,html.theme--documenter-dark .button.is-dark.is-inverted.is-hovered,html.theme--documenter-dark .content kbd.button.is-inverted.is-hovered{background-color:#f2f2f2}html.theme--documenter-dark .button.is-dark.is-inverted[disabled],html.theme--documenter-dark .content kbd.button.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-dark.is-inverted,fieldset[disabled] html.theme--documenter-dark .content kbd.button.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#282f2f}html.theme--documenter-dark .button.is-dark.is-loading::after,html.theme--documenter-dark .content kbd.button.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-dark.is-outlined,html.theme--documenter-dark .content kbd.button.is-outlined{background-color:transparent;border-color:#282f2f;color:#282f2f}html.theme--documenter-dark .button.is-dark.is-outlined:hover,html.theme--documenter-dark .content kbd.button.is-outlined:hover,html.theme--documenter-dark .button.is-dark.is-outlined.is-hovered,html.theme--documenter-dark .content kbd.button.is-outlined.is-hovered,html.theme--documenter-dark .button.is-dark.is-outlined:focus,html.theme--documenter-dark .content kbd.button.is-outlined:focus,html.theme--documenter-dark .button.is-dark.is-outlined.is-focused,html.theme--documenter-dark .content kbd.button.is-outlined.is-focused{background-color:#282f2f;border-color:#282f2f;color:#fff}html.theme--documenter-dark .button.is-dark.is-outlined.is-loading::after,html.theme--documenter-dark .content kbd.button.is-outlined.is-loading::after{border-color:transparent transparent #282f2f #282f2f !important}html.theme--documenter-dark .button.is-dark.is-outlined.is-loading:hover::after,html.theme--documenter-dark .content kbd.button.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-dark.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .content kbd.button.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-dark.is-outlined.is-loading:focus::after,html.theme--documenter-dark .content kbd.button.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-dark.is-outlined.is-loading.is-focused::after,html.theme--documenter-dark .content kbd.button.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-dark.is-outlined[disabled],html.theme--documenter-dark .content kbd.button.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-dark.is-outlined,fieldset[disabled] html.theme--documenter-dark .content kbd.button.is-outlined{background-color:transparent;border-color:#282f2f;box-shadow:none;color:#282f2f}html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined:hover,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined:focus,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined.is-focused,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined.is-focused{background-color:#fff;color:#282f2f}html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined.is-loading.is-focused::after,html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #282f2f #282f2f !important}html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined[disabled],html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-dark.is-inverted.is-outlined,fieldset[disabled] html.theme--documenter-dark .content kbd.button.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-primary,html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink{background-color:#375a7f;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-primary:hover,html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink:hover,html.theme--documenter-dark .button.is-primary.is-hovered,html.theme--documenter-dark .docstring>section>a.button.is-hovered.docs-sourcelink{background-color:#335476;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-primary:focus,html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink:focus,html.theme--documenter-dark .button.is-primary.is-focused,html.theme--documenter-dark .docstring>section>a.button.is-focused.docs-sourcelink{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-primary:focus:not(:active),html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink:focus:not(:active),html.theme--documenter-dark .button.is-primary.is-focused:not(:active),html.theme--documenter-dark .docstring>section>a.button.is-focused.docs-sourcelink:not(:active){box-shadow:0 0 0 0.125em rgba(55,90,127,0.25)}html.theme--documenter-dark .button.is-primary:active,html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink:active,html.theme--documenter-dark .button.is-primary.is-active,html.theme--documenter-dark .docstring>section>a.button.is-active.docs-sourcelink{background-color:#2f4d6d;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-primary[disabled],html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-primary,fieldset[disabled] html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink{background-color:#375a7f;border-color:#375a7f;box-shadow:none}html.theme--documenter-dark .button.is-primary.is-inverted,html.theme--documenter-dark .docstring>section>a.button.is-inverted.docs-sourcelink{background-color:#fff;color:#375a7f}html.theme--documenter-dark .button.is-primary.is-inverted:hover,html.theme--documenter-dark .docstring>section>a.button.is-inverted.docs-sourcelink:hover,html.theme--documenter-dark .button.is-primary.is-inverted.is-hovered,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-hovered.docs-sourcelink{background-color:#f2f2f2}html.theme--documenter-dark .button.is-primary.is-inverted[disabled],html.theme--documenter-dark .docstring>section>a.button.is-inverted.docs-sourcelink[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-primary.is-inverted,fieldset[disabled] html.theme--documenter-dark .docstring>section>a.button.is-inverted.docs-sourcelink{background-color:#fff;border-color:transparent;box-shadow:none;color:#375a7f}html.theme--documenter-dark .button.is-primary.is-loading::after,html.theme--documenter-dark .docstring>section>a.button.is-loading.docs-sourcelink::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-primary.is-outlined,html.theme--documenter-dark .docstring>section>a.button.is-outlined.docs-sourcelink{background-color:transparent;border-color:#375a7f;color:#375a7f}html.theme--documenter-dark .button.is-primary.is-outlined:hover,html.theme--documenter-dark .docstring>section>a.button.is-outlined.docs-sourcelink:hover,html.theme--documenter-dark .button.is-primary.is-outlined.is-hovered,html.theme--documenter-dark .docstring>section>a.button.is-outlined.is-hovered.docs-sourcelink,html.theme--documenter-dark .button.is-primary.is-outlined:focus,html.theme--documenter-dark .docstring>section>a.button.is-outlined.docs-sourcelink:focus,html.theme--documenter-dark .button.is-primary.is-outlined.is-focused,html.theme--documenter-dark .docstring>section>a.button.is-outlined.is-focused.docs-sourcelink{background-color:#375a7f;border-color:#375a7f;color:#fff}html.theme--documenter-dark .button.is-primary.is-outlined.is-loading::after,html.theme--documenter-dark .docstring>section>a.button.is-outlined.is-loading.docs-sourcelink::after{border-color:transparent transparent #375a7f #375a7f !important}html.theme--documenter-dark .button.is-primary.is-outlined.is-loading:hover::after,html.theme--documenter-dark .docstring>section>a.button.is-outlined.is-loading.docs-sourcelink:hover::after,html.theme--documenter-dark .button.is-primary.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .docstring>section>a.button.is-outlined.is-loading.is-hovered.docs-sourcelink::after,html.theme--documenter-dark .button.is-primary.is-outlined.is-loading:focus::after,html.theme--documenter-dark .docstring>section>a.button.is-outlined.is-loading.docs-sourcelink:focus::after,html.theme--documenter-dark .button.is-primary.is-outlined.is-loading.is-focused::after,html.theme--documenter-dark .docstring>section>a.button.is-outlined.is-loading.is-focused.docs-sourcelink::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-primary.is-outlined[disabled],html.theme--documenter-dark .docstring>section>a.button.is-outlined.docs-sourcelink[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-primary.is-outlined,fieldset[disabled] html.theme--documenter-dark .docstring>section>a.button.is-outlined.docs-sourcelink{background-color:transparent;border-color:#375a7f;box-shadow:none;color:#375a7f}html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined:hover,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink:hover,html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.is-hovered.docs-sourcelink,html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined:focus,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink:focus,html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined.is-focused,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.is-focused.docs-sourcelink{background-color:#fff;color:#375a7f}html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.is-loading.docs-sourcelink:hover::after,html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.is-loading.is-hovered.docs-sourcelink::after,html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.is-loading.docs-sourcelink:focus::after,html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined.is-loading.is-focused::after,html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.is-loading.is-focused.docs-sourcelink::after{border-color:transparent transparent #375a7f #375a7f !important}html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined[disabled],html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-primary.is-inverted.is-outlined,fieldset[disabled] html.theme--documenter-dark .docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-primary.is-light,html.theme--documenter-dark .docstring>section>a.button.is-light.docs-sourcelink{background-color:#f1f5f9;color:#4d7eb2}html.theme--documenter-dark .button.is-primary.is-light:hover,html.theme--documenter-dark .docstring>section>a.button.is-light.docs-sourcelink:hover,html.theme--documenter-dark .button.is-primary.is-light.is-hovered,html.theme--documenter-dark .docstring>section>a.button.is-light.is-hovered.docs-sourcelink{background-color:#e8eef5;border-color:transparent;color:#4d7eb2}html.theme--documenter-dark .button.is-primary.is-light:active,html.theme--documenter-dark .docstring>section>a.button.is-light.docs-sourcelink:active,html.theme--documenter-dark .button.is-primary.is-light.is-active,html.theme--documenter-dark .docstring>section>a.button.is-light.is-active.docs-sourcelink{background-color:#dfe8f1;border-color:transparent;color:#4d7eb2}html.theme--documenter-dark .button.is-link{background-color:#1abc9c;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-link:hover,html.theme--documenter-dark .button.is-link.is-hovered{background-color:#18b193;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-link:focus,html.theme--documenter-dark .button.is-link.is-focused{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-link:focus:not(:active),html.theme--documenter-dark .button.is-link.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(26,188,156,0.25)}html.theme--documenter-dark .button.is-link:active,html.theme--documenter-dark .button.is-link.is-active{background-color:#17a689;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-link[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-link{background-color:#1abc9c;border-color:#1abc9c;box-shadow:none}html.theme--documenter-dark .button.is-link.is-inverted{background-color:#fff;color:#1abc9c}html.theme--documenter-dark .button.is-link.is-inverted:hover,html.theme--documenter-dark .button.is-link.is-inverted.is-hovered{background-color:#f2f2f2}html.theme--documenter-dark .button.is-link.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-link.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#1abc9c}html.theme--documenter-dark .button.is-link.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-link.is-outlined{background-color:transparent;border-color:#1abc9c;color:#1abc9c}html.theme--documenter-dark .button.is-link.is-outlined:hover,html.theme--documenter-dark .button.is-link.is-outlined.is-hovered,html.theme--documenter-dark .button.is-link.is-outlined:focus,html.theme--documenter-dark .button.is-link.is-outlined.is-focused{background-color:#1abc9c;border-color:#1abc9c;color:#fff}html.theme--documenter-dark .button.is-link.is-outlined.is-loading::after{border-color:transparent transparent #1abc9c #1abc9c !important}html.theme--documenter-dark .button.is-link.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-link.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-link.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-link.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-link.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-link.is-outlined{background-color:transparent;border-color:#1abc9c;box-shadow:none;color:#1abc9c}html.theme--documenter-dark .button.is-link.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-link.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-link.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-link.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-link.is-inverted.is-outlined.is-focused{background-color:#fff;color:#1abc9c}html.theme--documenter-dark .button.is-link.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-link.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-link.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-link.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #1abc9c #1abc9c !important}html.theme--documenter-dark .button.is-link.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-link.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-link.is-light{background-color:#edfdf9;color:#15987e}html.theme--documenter-dark .button.is-link.is-light:hover,html.theme--documenter-dark .button.is-link.is-light.is-hovered{background-color:#e2fbf6;border-color:transparent;color:#15987e}html.theme--documenter-dark .button.is-link.is-light:active,html.theme--documenter-dark .button.is-link.is-light.is-active{background-color:#d7f9f3;border-color:transparent;color:#15987e}html.theme--documenter-dark .button.is-info{background-color:#024c7d;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-info:hover,html.theme--documenter-dark .button.is-info.is-hovered{background-color:#024470;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-info:focus,html.theme--documenter-dark .button.is-info.is-focused{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-info:focus:not(:active),html.theme--documenter-dark .button.is-info.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(2,76,125,0.25)}html.theme--documenter-dark .button.is-info:active,html.theme--documenter-dark .button.is-info.is-active{background-color:#023d64;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-info[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-info{background-color:#024c7d;border-color:#024c7d;box-shadow:none}html.theme--documenter-dark .button.is-info.is-inverted{background-color:#fff;color:#024c7d}html.theme--documenter-dark .button.is-info.is-inverted:hover,html.theme--documenter-dark .button.is-info.is-inverted.is-hovered{background-color:#f2f2f2}html.theme--documenter-dark .button.is-info.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-info.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#024c7d}html.theme--documenter-dark .button.is-info.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-info.is-outlined{background-color:transparent;border-color:#024c7d;color:#024c7d}html.theme--documenter-dark .button.is-info.is-outlined:hover,html.theme--documenter-dark .button.is-info.is-outlined.is-hovered,html.theme--documenter-dark .button.is-info.is-outlined:focus,html.theme--documenter-dark .button.is-info.is-outlined.is-focused{background-color:#024c7d;border-color:#024c7d;color:#fff}html.theme--documenter-dark .button.is-info.is-outlined.is-loading::after{border-color:transparent transparent #024c7d #024c7d !important}html.theme--documenter-dark .button.is-info.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-info.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-info.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-info.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-info.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-info.is-outlined{background-color:transparent;border-color:#024c7d;box-shadow:none;color:#024c7d}html.theme--documenter-dark .button.is-info.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-info.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-info.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-info.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-info.is-inverted.is-outlined.is-focused{background-color:#fff;color:#024c7d}html.theme--documenter-dark .button.is-info.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-info.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-info.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-info.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #024c7d #024c7d !important}html.theme--documenter-dark .button.is-info.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-info.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-info.is-light{background-color:#ebf7ff;color:#0e9dfb}html.theme--documenter-dark .button.is-info.is-light:hover,html.theme--documenter-dark .button.is-info.is-light.is-hovered{background-color:#def2fe;border-color:transparent;color:#0e9dfb}html.theme--documenter-dark .button.is-info.is-light:active,html.theme--documenter-dark .button.is-info.is-light.is-active{background-color:#d2edfe;border-color:transparent;color:#0e9dfb}html.theme--documenter-dark .button.is-success{background-color:#008438;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-success:hover,html.theme--documenter-dark .button.is-success.is-hovered{background-color:#073;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-success:focus,html.theme--documenter-dark .button.is-success.is-focused{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-success:focus:not(:active),html.theme--documenter-dark .button.is-success.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(0,132,56,0.25)}html.theme--documenter-dark .button.is-success:active,html.theme--documenter-dark .button.is-success.is-active{background-color:#006b2d;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-success[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-success{background-color:#008438;border-color:#008438;box-shadow:none}html.theme--documenter-dark .button.is-success.is-inverted{background-color:#fff;color:#008438}html.theme--documenter-dark .button.is-success.is-inverted:hover,html.theme--documenter-dark .button.is-success.is-inverted.is-hovered{background-color:#f2f2f2}html.theme--documenter-dark .button.is-success.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-success.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#008438}html.theme--documenter-dark .button.is-success.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-success.is-outlined{background-color:transparent;border-color:#008438;color:#008438}html.theme--documenter-dark .button.is-success.is-outlined:hover,html.theme--documenter-dark .button.is-success.is-outlined.is-hovered,html.theme--documenter-dark .button.is-success.is-outlined:focus,html.theme--documenter-dark .button.is-success.is-outlined.is-focused{background-color:#008438;border-color:#008438;color:#fff}html.theme--documenter-dark .button.is-success.is-outlined.is-loading::after{border-color:transparent transparent #008438 #008438 !important}html.theme--documenter-dark .button.is-success.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-success.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-success.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-success.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-success.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-success.is-outlined{background-color:transparent;border-color:#008438;box-shadow:none;color:#008438}html.theme--documenter-dark .button.is-success.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-success.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-success.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-success.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-success.is-inverted.is-outlined.is-focused{background-color:#fff;color:#008438}html.theme--documenter-dark .button.is-success.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-success.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-success.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-success.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #008438 #008438 !important}html.theme--documenter-dark .button.is-success.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-success.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-success.is-light{background-color:#ebfff3;color:#00eb64}html.theme--documenter-dark .button.is-success.is-light:hover,html.theme--documenter-dark .button.is-success.is-light.is-hovered{background-color:#deffec;border-color:transparent;color:#00eb64}html.theme--documenter-dark .button.is-success.is-light:active,html.theme--documenter-dark .button.is-success.is-light.is-active{background-color:#d1ffe5;border-color:transparent;color:#00eb64}html.theme--documenter-dark .button.is-warning{background-color:#ad8100;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-warning:hover,html.theme--documenter-dark .button.is-warning.is-hovered{background-color:#a07700;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-warning:focus,html.theme--documenter-dark .button.is-warning.is-focused{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-warning:focus:not(:active),html.theme--documenter-dark .button.is-warning.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(173,129,0,0.25)}html.theme--documenter-dark .button.is-warning:active,html.theme--documenter-dark .button.is-warning.is-active{background-color:#946e00;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-warning[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-warning{background-color:#ad8100;border-color:#ad8100;box-shadow:none}html.theme--documenter-dark .button.is-warning.is-inverted{background-color:#fff;color:#ad8100}html.theme--documenter-dark .button.is-warning.is-inverted:hover,html.theme--documenter-dark .button.is-warning.is-inverted.is-hovered{background-color:#f2f2f2}html.theme--documenter-dark .button.is-warning.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-warning.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#ad8100}html.theme--documenter-dark .button.is-warning.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-warning.is-outlined{background-color:transparent;border-color:#ad8100;color:#ad8100}html.theme--documenter-dark .button.is-warning.is-outlined:hover,html.theme--documenter-dark .button.is-warning.is-outlined.is-hovered,html.theme--documenter-dark .button.is-warning.is-outlined:focus,html.theme--documenter-dark .button.is-warning.is-outlined.is-focused{background-color:#ad8100;border-color:#ad8100;color:#fff}html.theme--documenter-dark .button.is-warning.is-outlined.is-loading::after{border-color:transparent transparent #ad8100 #ad8100 !important}html.theme--documenter-dark .button.is-warning.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-warning.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-warning.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-warning.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-warning.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-warning.is-outlined{background-color:transparent;border-color:#ad8100;box-shadow:none;color:#ad8100}html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined.is-focused{background-color:#fff;color:#ad8100}html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #ad8100 #ad8100 !important}html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-warning.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-warning.is-light{background-color:#fffaeb;color:#d19c00}html.theme--documenter-dark .button.is-warning.is-light:hover,html.theme--documenter-dark .button.is-warning.is-light.is-hovered{background-color:#fff7de;border-color:transparent;color:#d19c00}html.theme--documenter-dark .button.is-warning.is-light:active,html.theme--documenter-dark .button.is-warning.is-light.is-active{background-color:#fff3d1;border-color:transparent;color:#d19c00}html.theme--documenter-dark .button.is-danger{background-color:#9e1b0d;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-danger:hover,html.theme--documenter-dark .button.is-danger.is-hovered{background-color:#92190c;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-danger:focus,html.theme--documenter-dark .button.is-danger.is-focused{border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-danger:focus:not(:active),html.theme--documenter-dark .button.is-danger.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(158,27,13,0.25)}html.theme--documenter-dark .button.is-danger:active,html.theme--documenter-dark .button.is-danger.is-active{background-color:#86170b;border-color:transparent;color:#fff}html.theme--documenter-dark .button.is-danger[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-danger{background-color:#9e1b0d;border-color:#9e1b0d;box-shadow:none}html.theme--documenter-dark .button.is-danger.is-inverted{background-color:#fff;color:#9e1b0d}html.theme--documenter-dark .button.is-danger.is-inverted:hover,html.theme--documenter-dark .button.is-danger.is-inverted.is-hovered{background-color:#f2f2f2}html.theme--documenter-dark .button.is-danger.is-inverted[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-danger.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#9e1b0d}html.theme--documenter-dark .button.is-danger.is-loading::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-danger.is-outlined{background-color:transparent;border-color:#9e1b0d;color:#9e1b0d}html.theme--documenter-dark .button.is-danger.is-outlined:hover,html.theme--documenter-dark .button.is-danger.is-outlined.is-hovered,html.theme--documenter-dark .button.is-danger.is-outlined:focus,html.theme--documenter-dark .button.is-danger.is-outlined.is-focused{background-color:#9e1b0d;border-color:#9e1b0d;color:#fff}html.theme--documenter-dark .button.is-danger.is-outlined.is-loading::after{border-color:transparent transparent #9e1b0d #9e1b0d !important}html.theme--documenter-dark .button.is-danger.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-danger.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-danger.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-danger.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}html.theme--documenter-dark .button.is-danger.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-danger.is-outlined{background-color:transparent;border-color:#9e1b0d;box-shadow:none;color:#9e1b0d}html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined:hover,html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined.is-hovered,html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined:focus,html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined.is-focused{background-color:#fff;color:#9e1b0d}html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined.is-loading:hover::after,html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined.is-loading.is-hovered::after,html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined.is-loading:focus::after,html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #9e1b0d #9e1b0d !important}html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined[disabled],fieldset[disabled] html.theme--documenter-dark .button.is-danger.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}html.theme--documenter-dark .button.is-danger.is-light{background-color:#fdeeec;color:#ec311d}html.theme--documenter-dark .button.is-danger.is-light:hover,html.theme--documenter-dark .button.is-danger.is-light.is-hovered{background-color:#fce3e0;border-color:transparent;color:#ec311d}html.theme--documenter-dark .button.is-danger.is-light:active,html.theme--documenter-dark .button.is-danger.is-light.is-active{background-color:#fcd8d5;border-color:transparent;color:#ec311d}html.theme--documenter-dark .button.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.button{font-size:.75rem}html.theme--documenter-dark .button.is-small:not(.is-rounded),html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.button:not(.is-rounded){border-radius:3px}html.theme--documenter-dark .button.is-normal{font-size:1rem}html.theme--documenter-dark .button.is-medium{font-size:1.25rem}html.theme--documenter-dark .button.is-large{font-size:1.5rem}html.theme--documenter-dark .button[disabled],fieldset[disabled] html.theme--documenter-dark .button{background-color:#8c9b9d;border-color:#5e6d6f;box-shadow:none;opacity:.5}html.theme--documenter-dark .button.is-fullwidth{display:flex;width:100%}html.theme--documenter-dark .button.is-loading{color:transparent !important;pointer-events:none}html.theme--documenter-dark .button.is-loading::after{position:absolute;left:calc(50% - (1em * 0.5));top:calc(50% - (1em * 0.5));position:absolute !important}html.theme--documenter-dark .button.is-static{background-color:#282f2f;border-color:#5e6d6f;color:#dbdee0;box-shadow:none;pointer-events:none}html.theme--documenter-dark .button.is-rounded,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.button{border-radius:9999px;padding-left:calc(1em + 0.25em);padding-right:calc(1em + 0.25em)}html.theme--documenter-dark .buttons{align-items:center;display:flex;flex-wrap:wrap;justify-content:flex-start}html.theme--documenter-dark .buttons .button{margin-bottom:0.5rem}html.theme--documenter-dark .buttons .button:not(:last-child):not(.is-fullwidth){margin-right:.5rem}html.theme--documenter-dark .buttons:last-child{margin-bottom:-0.5rem}html.theme--documenter-dark .buttons:not(:last-child){margin-bottom:1rem}html.theme--documenter-dark .buttons.are-small .button:not(.is-normal):not(.is-medium):not(.is-large){font-size:.75rem}html.theme--documenter-dark .buttons.are-small .button:not(.is-normal):not(.is-medium):not(.is-large):not(.is-rounded){border-radius:3px}html.theme--documenter-dark .buttons.are-medium .button:not(.is-small):not(.is-normal):not(.is-large){font-size:1.25rem}html.theme--documenter-dark .buttons.are-large .button:not(.is-small):not(.is-normal):not(.is-medium){font-size:1.5rem}html.theme--documenter-dark .buttons.has-addons .button:not(:first-child){border-bottom-left-radius:0;border-top-left-radius:0}html.theme--documenter-dark .buttons.has-addons .button:not(:last-child){border-bottom-right-radius:0;border-top-right-radius:0;margin-right:-1px}html.theme--documenter-dark .buttons.has-addons .button:last-child{margin-right:0}html.theme--documenter-dark .buttons.has-addons .button:hover,html.theme--documenter-dark .buttons.has-addons .button.is-hovered{z-index:2}html.theme--documenter-dark .buttons.has-addons .button:focus,html.theme--documenter-dark .buttons.has-addons .button.is-focused,html.theme--documenter-dark .buttons.has-addons .button:active,html.theme--documenter-dark .buttons.has-addons .button.is-active,html.theme--documenter-dark .buttons.has-addons .button.is-selected{z-index:3}html.theme--documenter-dark .buttons.has-addons .button:focus:hover,html.theme--documenter-dark .buttons.has-addons .button.is-focused:hover,html.theme--documenter-dark .buttons.has-addons .button:active:hover,html.theme--documenter-dark .buttons.has-addons .button.is-active:hover,html.theme--documenter-dark .buttons.has-addons .button.is-selected:hover{z-index:4}html.theme--documenter-dark .buttons.has-addons .button.is-expanded{flex-grow:1;flex-shrink:1}html.theme--documenter-dark .buttons.is-centered{justify-content:center}html.theme--documenter-dark .buttons.is-centered:not(.has-addons) .button:not(.is-fullwidth){margin-left:0.25rem;margin-right:0.25rem}html.theme--documenter-dark .buttons.is-right{justify-content:flex-end}html.theme--documenter-dark .buttons.is-right:not(.has-addons) .button:not(.is-fullwidth){margin-left:0.25rem;margin-right:0.25rem}@media screen and (max-width: 768px){html.theme--documenter-dark .button.is-responsive.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-responsive{font-size:.5625rem}html.theme--documenter-dark .button.is-responsive,html.theme--documenter-dark .button.is-responsive.is-normal{font-size:.65625rem}html.theme--documenter-dark .button.is-responsive.is-medium{font-size:.75rem}html.theme--documenter-dark .button.is-responsive.is-large{font-size:1rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .button.is-responsive.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-responsive{font-size:.65625rem}html.theme--documenter-dark .button.is-responsive,html.theme--documenter-dark .button.is-responsive.is-normal{font-size:.75rem}html.theme--documenter-dark .button.is-responsive.is-medium{font-size:1rem}html.theme--documenter-dark .button.is-responsive.is-large{font-size:1.25rem}}html.theme--documenter-dark .container{flex-grow:1;margin:0 auto;position:relative;width:auto}html.theme--documenter-dark .container.is-fluid{max-width:none !important;padding-left:32px;padding-right:32px;width:100%}@media screen and (min-width: 1056px){html.theme--documenter-dark .container{max-width:992px}}@media screen and (max-width: 1215px){html.theme--documenter-dark .container.is-widescreen:not(.is-max-desktop){max-width:1152px}}@media screen and (max-width: 1407px){html.theme--documenter-dark .container.is-fullhd:not(.is-max-desktop):not(.is-max-widescreen){max-width:1344px}}@media screen and (min-width: 1216px){html.theme--documenter-dark .container:not(.is-max-desktop){max-width:1152px}}@media screen and (min-width: 1408px){html.theme--documenter-dark .container:not(.is-max-desktop):not(.is-max-widescreen){max-width:1344px}}html.theme--documenter-dark .content li+li{margin-top:0.25em}html.theme--documenter-dark .content p:not(:last-child),html.theme--documenter-dark .content dl:not(:last-child),html.theme--documenter-dark .content ol:not(:last-child),html.theme--documenter-dark .content ul:not(:last-child),html.theme--documenter-dark .content blockquote:not(:last-child),html.theme--documenter-dark .content pre:not(:last-child),html.theme--documenter-dark .content table:not(:last-child){margin-bottom:1em}html.theme--documenter-dark .content h1,html.theme--documenter-dark .content h2,html.theme--documenter-dark .content h3,html.theme--documenter-dark .content h4,html.theme--documenter-dark .content h5,html.theme--documenter-dark .content h6{color:#f2f2f2;font-weight:600;line-height:1.125}html.theme--documenter-dark .content h1{font-size:2em;margin-bottom:0.5em}html.theme--documenter-dark .content h1:not(:first-child){margin-top:1em}html.theme--documenter-dark .content h2{font-size:1.75em;margin-bottom:0.5714em}html.theme--documenter-dark .content h2:not(:first-child){margin-top:1.1428em}html.theme--documenter-dark .content h3{font-size:1.5em;margin-bottom:0.6666em}html.theme--documenter-dark .content h3:not(:first-child){margin-top:1.3333em}html.theme--documenter-dark .content h4{font-size:1.25em;margin-bottom:0.8em}html.theme--documenter-dark .content h5{font-size:1.125em;margin-bottom:0.8888em}html.theme--documenter-dark .content h6{font-size:1em;margin-bottom:1em}html.theme--documenter-dark .content blockquote{background-color:#282f2f;border-left:5px solid #5e6d6f;padding:1.25em 1.5em}html.theme--documenter-dark .content ol{list-style-position:outside;margin-left:2em;margin-top:1em}html.theme--documenter-dark .content ol:not([type]){list-style-type:decimal}html.theme--documenter-dark .content ol.is-lower-alpha:not([type]){list-style-type:lower-alpha}html.theme--documenter-dark .content ol.is-lower-roman:not([type]){list-style-type:lower-roman}html.theme--documenter-dark .content ol.is-upper-alpha:not([type]){list-style-type:upper-alpha}html.theme--documenter-dark .content ol.is-upper-roman:not([type]){list-style-type:upper-roman}html.theme--documenter-dark .content ul{list-style:disc outside;margin-left:2em;margin-top:1em}html.theme--documenter-dark .content ul ul{list-style-type:circle;margin-top:0.5em}html.theme--documenter-dark .content ul ul ul{list-style-type:square}html.theme--documenter-dark .content dd{margin-left:2em}html.theme--documenter-dark .content figure{margin-left:2em;margin-right:2em;text-align:center}html.theme--documenter-dark .content figure:not(:first-child){margin-top:2em}html.theme--documenter-dark .content figure:not(:last-child){margin-bottom:2em}html.theme--documenter-dark .content figure img{display:inline-block}html.theme--documenter-dark .content figure figcaption{font-style:italic}html.theme--documenter-dark .content pre{-webkit-overflow-scrolling:touch;overflow-x:auto;padding:0;white-space:pre;word-wrap:normal}html.theme--documenter-dark .content sup,html.theme--documenter-dark .content sub{font-size:75%}html.theme--documenter-dark .content table{width:100%}html.theme--documenter-dark .content table td,html.theme--documenter-dark .content table th{border:1px solid #5e6d6f;border-width:0 0 1px;padding:0.5em 0.75em;vertical-align:top}html.theme--documenter-dark .content table th{color:#f2f2f2}html.theme--documenter-dark .content table th:not([align]){text-align:inherit}html.theme--documenter-dark .content table thead td,html.theme--documenter-dark .content table thead th{border-width:0 0 2px;color:#f2f2f2}html.theme--documenter-dark .content table tfoot td,html.theme--documenter-dark .content table tfoot th{border-width:2px 0 0;color:#f2f2f2}html.theme--documenter-dark .content table tbody tr:last-child td,html.theme--documenter-dark .content table tbody tr:last-child th{border-bottom-width:0}html.theme--documenter-dark .content .tabs li+li{margin-top:0}html.theme--documenter-dark .content.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.content{font-size:.75rem}html.theme--documenter-dark .content.is-normal{font-size:1rem}html.theme--documenter-dark .content.is-medium{font-size:1.25rem}html.theme--documenter-dark .content.is-large{font-size:1.5rem}html.theme--documenter-dark .icon{align-items:center;display:inline-flex;justify-content:center;height:1.5rem;width:1.5rem}html.theme--documenter-dark .icon.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.icon{height:1rem;width:1rem}html.theme--documenter-dark .icon.is-medium{height:2rem;width:2rem}html.theme--documenter-dark .icon.is-large{height:3rem;width:3rem}html.theme--documenter-dark .icon-text{align-items:flex-start;color:inherit;display:inline-flex;flex-wrap:wrap;line-height:1.5rem;vertical-align:top}html.theme--documenter-dark .icon-text .icon{flex-grow:0;flex-shrink:0}html.theme--documenter-dark .icon-text .icon:not(:last-child){margin-right:.25em}html.theme--documenter-dark .icon-text .icon:not(:first-child){margin-left:.25em}html.theme--documenter-dark div.icon-text{display:flex}html.theme--documenter-dark .image,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img{display:block;position:relative}html.theme--documenter-dark .image img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img img{display:block;height:auto;width:100%}html.theme--documenter-dark .image img.is-rounded,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img img.is-rounded{border-radius:9999px}html.theme--documenter-dark .image.is-fullwidth,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-fullwidth{width:100%}html.theme--documenter-dark .image.is-square img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-square img,html.theme--documenter-dark .image.is-square .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-square .has-ratio,html.theme--documenter-dark .image.is-1by1 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by1 img,html.theme--documenter-dark .image.is-1by1 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by1 .has-ratio,html.theme--documenter-dark .image.is-5by4 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by4 img,html.theme--documenter-dark .image.is-5by4 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by4 .has-ratio,html.theme--documenter-dark .image.is-4by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by3 img,html.theme--documenter-dark .image.is-4by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by3 .has-ratio,html.theme--documenter-dark .image.is-3by2 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by2 img,html.theme--documenter-dark .image.is-3by2 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by2 .has-ratio,html.theme--documenter-dark .image.is-5by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by3 img,html.theme--documenter-dark .image.is-5by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by3 .has-ratio,html.theme--documenter-dark .image.is-16by9 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-16by9 img,html.theme--documenter-dark .image.is-16by9 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-16by9 .has-ratio,html.theme--documenter-dark .image.is-2by1 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by1 img,html.theme--documenter-dark .image.is-2by1 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by1 .has-ratio,html.theme--documenter-dark .image.is-3by1 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by1 img,html.theme--documenter-dark .image.is-3by1 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by1 .has-ratio,html.theme--documenter-dark .image.is-4by5 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by5 img,html.theme--documenter-dark .image.is-4by5 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by5 .has-ratio,html.theme--documenter-dark .image.is-3by4 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by4 img,html.theme--documenter-dark .image.is-3by4 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by4 .has-ratio,html.theme--documenter-dark .image.is-2by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by3 img,html.theme--documenter-dark .image.is-2by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by3 .has-ratio,html.theme--documenter-dark .image.is-3by5 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by5 img,html.theme--documenter-dark .image.is-3by5 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by5 .has-ratio,html.theme--documenter-dark .image.is-9by16 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-9by16 img,html.theme--documenter-dark .image.is-9by16 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-9by16 .has-ratio,html.theme--documenter-dark .image.is-1by2 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by2 img,html.theme--documenter-dark .image.is-1by2 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by2 .has-ratio,html.theme--documenter-dark .image.is-1by3 img,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by3 img,html.theme--documenter-dark .image.is-1by3 .has-ratio,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by3 .has-ratio{height:100%;width:100%}html.theme--documenter-dark .image.is-square,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-square,html.theme--documenter-dark .image.is-1by1,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by1{padding-top:100%}html.theme--documenter-dark .image.is-5by4,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by4{padding-top:80%}html.theme--documenter-dark .image.is-4by3,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by3{padding-top:75%}html.theme--documenter-dark .image.is-3by2,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by2{padding-top:66.6666%}html.theme--documenter-dark .image.is-5by3,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-5by3{padding-top:60%}html.theme--documenter-dark .image.is-16by9,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-16by9{padding-top:56.25%}html.theme--documenter-dark .image.is-2by1,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by1{padding-top:50%}html.theme--documenter-dark .image.is-3by1,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by1{padding-top:33.3333%}html.theme--documenter-dark .image.is-4by5,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-4by5{padding-top:125%}html.theme--documenter-dark .image.is-3by4,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by4{padding-top:133.3333%}html.theme--documenter-dark .image.is-2by3,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-2by3{padding-top:150%}html.theme--documenter-dark .image.is-3by5,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-3by5{padding-top:166.6666%}html.theme--documenter-dark .image.is-9by16,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-9by16{padding-top:177.7777%}html.theme--documenter-dark .image.is-1by2,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by2{padding-top:200%}html.theme--documenter-dark .image.is-1by3,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-1by3{padding-top:300%}html.theme--documenter-dark .image.is-16x16,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-16x16{height:16px;width:16px}html.theme--documenter-dark .image.is-24x24,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-24x24{height:24px;width:24px}html.theme--documenter-dark .image.is-32x32,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-32x32{height:32px;width:32px}html.theme--documenter-dark .image.is-48x48,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-48x48{height:48px;width:48px}html.theme--documenter-dark .image.is-64x64,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-64x64{height:64px;width:64px}html.theme--documenter-dark .image.is-96x96,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-96x96{height:96px;width:96px}html.theme--documenter-dark .image.is-128x128,html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img.is-128x128{height:128px;width:128px}html.theme--documenter-dark .notification{background-color:#282f2f;border-radius:.4em;position:relative;padding:1.25rem 2.5rem 1.25rem 1.5rem}html.theme--documenter-dark .notification a:not(.button):not(.dropdown-item){color:currentColor;text-decoration:underline}html.theme--documenter-dark .notification strong{color:currentColor}html.theme--documenter-dark .notification code,html.theme--documenter-dark .notification pre{background:#fff}html.theme--documenter-dark .notification pre code{background:transparent}html.theme--documenter-dark .notification>.delete{right:.5rem;position:absolute;top:0.5rem}html.theme--documenter-dark .notification .title,html.theme--documenter-dark .notification .subtitle,html.theme--documenter-dark .notification .content{color:currentColor}html.theme--documenter-dark .notification.is-white{background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .notification.is-black{background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .notification.is-light{background-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .notification.is-dark,html.theme--documenter-dark .content kbd.notification{background-color:#282f2f;color:#fff}html.theme--documenter-dark .notification.is-primary,html.theme--documenter-dark .docstring>section>a.notification.docs-sourcelink{background-color:#375a7f;color:#fff}html.theme--documenter-dark .notification.is-primary.is-light,html.theme--documenter-dark .docstring>section>a.notification.is-light.docs-sourcelink{background-color:#f1f5f9;color:#4d7eb2}html.theme--documenter-dark .notification.is-link{background-color:#1abc9c;color:#fff}html.theme--documenter-dark .notification.is-link.is-light{background-color:#edfdf9;color:#15987e}html.theme--documenter-dark .notification.is-info{background-color:#024c7d;color:#fff}html.theme--documenter-dark .notification.is-info.is-light{background-color:#ebf7ff;color:#0e9dfb}html.theme--documenter-dark .notification.is-success{background-color:#008438;color:#fff}html.theme--documenter-dark .notification.is-success.is-light{background-color:#ebfff3;color:#00eb64}html.theme--documenter-dark .notification.is-warning{background-color:#ad8100;color:#fff}html.theme--documenter-dark .notification.is-warning.is-light{background-color:#fffaeb;color:#d19c00}html.theme--documenter-dark .notification.is-danger{background-color:#9e1b0d;color:#fff}html.theme--documenter-dark .notification.is-danger.is-light{background-color:#fdeeec;color:#ec311d}html.theme--documenter-dark .progress{-moz-appearance:none;-webkit-appearance:none;border:none;border-radius:9999px;display:block;height:1rem;overflow:hidden;padding:0;width:100%}html.theme--documenter-dark .progress::-webkit-progress-bar{background-color:#343c3d}html.theme--documenter-dark .progress::-webkit-progress-value{background-color:#dbdee0}html.theme--documenter-dark .progress::-moz-progress-bar{background-color:#dbdee0}html.theme--documenter-dark .progress::-ms-fill{background-color:#dbdee0;border:none}html.theme--documenter-dark .progress.is-white::-webkit-progress-value{background-color:#fff}html.theme--documenter-dark .progress.is-white::-moz-progress-bar{background-color:#fff}html.theme--documenter-dark .progress.is-white::-ms-fill{background-color:#fff}html.theme--documenter-dark .progress.is-white:indeterminate{background-image:linear-gradient(to right, #fff 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-black::-webkit-progress-value{background-color:#0a0a0a}html.theme--documenter-dark .progress.is-black::-moz-progress-bar{background-color:#0a0a0a}html.theme--documenter-dark .progress.is-black::-ms-fill{background-color:#0a0a0a}html.theme--documenter-dark .progress.is-black:indeterminate{background-image:linear-gradient(to right, #0a0a0a 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-light::-webkit-progress-value{background-color:#ecf0f1}html.theme--documenter-dark .progress.is-light::-moz-progress-bar{background-color:#ecf0f1}html.theme--documenter-dark .progress.is-light::-ms-fill{background-color:#ecf0f1}html.theme--documenter-dark .progress.is-light:indeterminate{background-image:linear-gradient(to right, #ecf0f1 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-dark::-webkit-progress-value,html.theme--documenter-dark .content kbd.progress::-webkit-progress-value{background-color:#282f2f}html.theme--documenter-dark .progress.is-dark::-moz-progress-bar,html.theme--documenter-dark .content kbd.progress::-moz-progress-bar{background-color:#282f2f}html.theme--documenter-dark .progress.is-dark::-ms-fill,html.theme--documenter-dark .content kbd.progress::-ms-fill{background-color:#282f2f}html.theme--documenter-dark .progress.is-dark:indeterminate,html.theme--documenter-dark .content kbd.progress:indeterminate{background-image:linear-gradient(to right, #282f2f 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-primary::-webkit-progress-value,html.theme--documenter-dark .docstring>section>a.progress.docs-sourcelink::-webkit-progress-value{background-color:#375a7f}html.theme--documenter-dark .progress.is-primary::-moz-progress-bar,html.theme--documenter-dark .docstring>section>a.progress.docs-sourcelink::-moz-progress-bar{background-color:#375a7f}html.theme--documenter-dark .progress.is-primary::-ms-fill,html.theme--documenter-dark .docstring>section>a.progress.docs-sourcelink::-ms-fill{background-color:#375a7f}html.theme--documenter-dark .progress.is-primary:indeterminate,html.theme--documenter-dark .docstring>section>a.progress.docs-sourcelink:indeterminate{background-image:linear-gradient(to right, #375a7f 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-link::-webkit-progress-value{background-color:#1abc9c}html.theme--documenter-dark .progress.is-link::-moz-progress-bar{background-color:#1abc9c}html.theme--documenter-dark .progress.is-link::-ms-fill{background-color:#1abc9c}html.theme--documenter-dark .progress.is-link:indeterminate{background-image:linear-gradient(to right, #1abc9c 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-info::-webkit-progress-value{background-color:#024c7d}html.theme--documenter-dark .progress.is-info::-moz-progress-bar{background-color:#024c7d}html.theme--documenter-dark .progress.is-info::-ms-fill{background-color:#024c7d}html.theme--documenter-dark .progress.is-info:indeterminate{background-image:linear-gradient(to right, #024c7d 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-success::-webkit-progress-value{background-color:#008438}html.theme--documenter-dark .progress.is-success::-moz-progress-bar{background-color:#008438}html.theme--documenter-dark .progress.is-success::-ms-fill{background-color:#008438}html.theme--documenter-dark .progress.is-success:indeterminate{background-image:linear-gradient(to right, #008438 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-warning::-webkit-progress-value{background-color:#ad8100}html.theme--documenter-dark .progress.is-warning::-moz-progress-bar{background-color:#ad8100}html.theme--documenter-dark .progress.is-warning::-ms-fill{background-color:#ad8100}html.theme--documenter-dark .progress.is-warning:indeterminate{background-image:linear-gradient(to right, #ad8100 30%, #343c3d 30%)}html.theme--documenter-dark .progress.is-danger::-webkit-progress-value{background-color:#9e1b0d}html.theme--documenter-dark .progress.is-danger::-moz-progress-bar{background-color:#9e1b0d}html.theme--documenter-dark .progress.is-danger::-ms-fill{background-color:#9e1b0d}html.theme--documenter-dark .progress.is-danger:indeterminate{background-image:linear-gradient(to right, #9e1b0d 30%, #343c3d 30%)}html.theme--documenter-dark .progress:indeterminate{animation-duration:1.5s;animation-iteration-count:infinite;animation-name:moveIndeterminate;animation-timing-function:linear;background-color:#343c3d;background-image:linear-gradient(to right, #fff 30%, #343c3d 30%);background-position:top left;background-repeat:no-repeat;background-size:150% 150%}html.theme--documenter-dark .progress:indeterminate::-webkit-progress-bar{background-color:transparent}html.theme--documenter-dark .progress:indeterminate::-moz-progress-bar{background-color:transparent}html.theme--documenter-dark .progress:indeterminate::-ms-fill{animation-name:none}html.theme--documenter-dark .progress.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.progress{height:.75rem}html.theme--documenter-dark .progress.is-medium{height:1.25rem}html.theme--documenter-dark .progress.is-large{height:1.5rem}@keyframes moveIndeterminate{from{background-position:200% 0}to{background-position:-200% 0}}html.theme--documenter-dark .table{background-color:#343c3d;color:#fff}html.theme--documenter-dark .table td,html.theme--documenter-dark .table th{border:1px solid #5e6d6f;border-width:0 0 1px;padding:0.5em 0.75em;vertical-align:top}html.theme--documenter-dark .table td.is-white,html.theme--documenter-dark .table th.is-white{background-color:#fff;border-color:#fff;color:#0a0a0a}html.theme--documenter-dark .table td.is-black,html.theme--documenter-dark .table th.is-black{background-color:#0a0a0a;border-color:#0a0a0a;color:#fff}html.theme--documenter-dark .table td.is-light,html.theme--documenter-dark .table th.is-light{background-color:#ecf0f1;border-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .table td.is-dark,html.theme--documenter-dark .table th.is-dark{background-color:#282f2f;border-color:#282f2f;color:#fff}html.theme--documenter-dark .table td.is-primary,html.theme--documenter-dark .table th.is-primary{background-color:#375a7f;border-color:#375a7f;color:#fff}html.theme--documenter-dark .table td.is-link,html.theme--documenter-dark .table th.is-link{background-color:#1abc9c;border-color:#1abc9c;color:#fff}html.theme--documenter-dark .table td.is-info,html.theme--documenter-dark .table th.is-info{background-color:#024c7d;border-color:#024c7d;color:#fff}html.theme--documenter-dark .table td.is-success,html.theme--documenter-dark .table th.is-success{background-color:#008438;border-color:#008438;color:#fff}html.theme--documenter-dark .table td.is-warning,html.theme--documenter-dark .table th.is-warning{background-color:#ad8100;border-color:#ad8100;color:#fff}html.theme--documenter-dark .table td.is-danger,html.theme--documenter-dark .table th.is-danger{background-color:#9e1b0d;border-color:#9e1b0d;color:#fff}html.theme--documenter-dark .table td.is-narrow,html.theme--documenter-dark .table th.is-narrow{white-space:nowrap;width:1%}html.theme--documenter-dark .table td.is-selected,html.theme--documenter-dark .table th.is-selected{background-color:#375a7f;color:#fff}html.theme--documenter-dark .table td.is-selected a,html.theme--documenter-dark .table td.is-selected strong,html.theme--documenter-dark .table th.is-selected a,html.theme--documenter-dark .table th.is-selected strong{color:currentColor}html.theme--documenter-dark .table td.is-vcentered,html.theme--documenter-dark .table th.is-vcentered{vertical-align:middle}html.theme--documenter-dark .table th{color:#f2f2f2}html.theme--documenter-dark .table th:not([align]){text-align:left}html.theme--documenter-dark .table tr.is-selected{background-color:#375a7f;color:#fff}html.theme--documenter-dark .table tr.is-selected a,html.theme--documenter-dark .table tr.is-selected strong{color:currentColor}html.theme--documenter-dark .table tr.is-selected td,html.theme--documenter-dark .table tr.is-selected th{border-color:#fff;color:currentColor}html.theme--documenter-dark .table thead{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .table thead td,html.theme--documenter-dark .table thead th{border-width:0 0 2px;color:#f2f2f2}html.theme--documenter-dark .table tfoot{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .table tfoot td,html.theme--documenter-dark .table tfoot th{border-width:2px 0 0;color:#f2f2f2}html.theme--documenter-dark .table tbody{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .table tbody tr:last-child td,html.theme--documenter-dark .table tbody tr:last-child th{border-bottom-width:0}html.theme--documenter-dark .table.is-bordered td,html.theme--documenter-dark .table.is-bordered th{border-width:1px}html.theme--documenter-dark .table.is-bordered tr:last-child td,html.theme--documenter-dark .table.is-bordered tr:last-child th{border-bottom-width:1px}html.theme--documenter-dark .table.is-fullwidth{width:100%}html.theme--documenter-dark .table.is-hoverable tbody tr:not(.is-selected):hover{background-color:#282f2f}html.theme--documenter-dark .table.is-hoverable.is-striped tbody tr:not(.is-selected):hover{background-color:#282f2f}html.theme--documenter-dark .table.is-hoverable.is-striped tbody tr:not(.is-selected):hover:nth-child(even){background-color:#2d3435}html.theme--documenter-dark .table.is-narrow td,html.theme--documenter-dark .table.is-narrow th{padding:0.25em 0.5em}html.theme--documenter-dark .table.is-striped tbody tr:not(.is-selected):nth-child(even){background-color:#282f2f}html.theme--documenter-dark .table-container{-webkit-overflow-scrolling:touch;overflow:auto;overflow-y:hidden;max-width:100%}html.theme--documenter-dark .tags{align-items:center;display:flex;flex-wrap:wrap;justify-content:flex-start}html.theme--documenter-dark .tags .tag,html.theme--documenter-dark .tags .content kbd,html.theme--documenter-dark .content .tags kbd,html.theme--documenter-dark .tags .docstring>section>a.docs-sourcelink{margin-bottom:0.5rem}html.theme--documenter-dark .tags .tag:not(:last-child),html.theme--documenter-dark .tags .content kbd:not(:last-child),html.theme--documenter-dark .content .tags kbd:not(:last-child),html.theme--documenter-dark .tags .docstring>section>a.docs-sourcelink:not(:last-child){margin-right:.5rem}html.theme--documenter-dark .tags:last-child{margin-bottom:-0.5rem}html.theme--documenter-dark .tags:not(:last-child){margin-bottom:1rem}html.theme--documenter-dark .tags.are-medium .tag:not(.is-normal):not(.is-large),html.theme--documenter-dark .tags.are-medium .content kbd:not(.is-normal):not(.is-large),html.theme--documenter-dark .content .tags.are-medium kbd:not(.is-normal):not(.is-large),html.theme--documenter-dark .tags.are-medium .docstring>section>a.docs-sourcelink:not(.is-normal):not(.is-large){font-size:1rem}html.theme--documenter-dark .tags.are-large .tag:not(.is-normal):not(.is-medium),html.theme--documenter-dark .tags.are-large .content kbd:not(.is-normal):not(.is-medium),html.theme--documenter-dark .content .tags.are-large kbd:not(.is-normal):not(.is-medium),html.theme--documenter-dark .tags.are-large .docstring>section>a.docs-sourcelink:not(.is-normal):not(.is-medium){font-size:1.25rem}html.theme--documenter-dark .tags.is-centered{justify-content:center}html.theme--documenter-dark .tags.is-centered .tag,html.theme--documenter-dark .tags.is-centered .content kbd,html.theme--documenter-dark .content .tags.is-centered kbd,html.theme--documenter-dark .tags.is-centered .docstring>section>a.docs-sourcelink{margin-right:0.25rem;margin-left:0.25rem}html.theme--documenter-dark .tags.is-right{justify-content:flex-end}html.theme--documenter-dark .tags.is-right .tag:not(:first-child),html.theme--documenter-dark .tags.is-right .content kbd:not(:first-child),html.theme--documenter-dark .content .tags.is-right kbd:not(:first-child),html.theme--documenter-dark .tags.is-right .docstring>section>a.docs-sourcelink:not(:first-child){margin-left:0.5rem}html.theme--documenter-dark .tags.is-right .tag:not(:last-child),html.theme--documenter-dark .tags.is-right .content kbd:not(:last-child),html.theme--documenter-dark .content .tags.is-right kbd:not(:last-child),html.theme--documenter-dark .tags.is-right .docstring>section>a.docs-sourcelink:not(:last-child){margin-right:0}html.theme--documenter-dark .tags.has-addons .tag,html.theme--documenter-dark .tags.has-addons .content kbd,html.theme--documenter-dark .content .tags.has-addons kbd,html.theme--documenter-dark .tags.has-addons .docstring>section>a.docs-sourcelink{margin-right:0}html.theme--documenter-dark .tags.has-addons .tag:not(:first-child),html.theme--documenter-dark .tags.has-addons .content kbd:not(:first-child),html.theme--documenter-dark .content .tags.has-addons kbd:not(:first-child),html.theme--documenter-dark .tags.has-addons .docstring>section>a.docs-sourcelink:not(:first-child){margin-left:0;border-top-left-radius:0;border-bottom-left-radius:0}html.theme--documenter-dark .tags.has-addons .tag:not(:last-child),html.theme--documenter-dark .tags.has-addons .content kbd:not(:last-child),html.theme--documenter-dark .content .tags.has-addons kbd:not(:last-child),html.theme--documenter-dark .tags.has-addons .docstring>section>a.docs-sourcelink:not(:last-child){border-top-right-radius:0;border-bottom-right-radius:0}html.theme--documenter-dark .tag:not(body),html.theme--documenter-dark .content kbd:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink:not(body){align-items:center;background-color:#282f2f;border-radius:.4em;color:#fff;display:inline-flex;font-size:.75rem;height:2em;justify-content:center;line-height:1.5;padding-left:0.75em;padding-right:0.75em;white-space:nowrap}html.theme--documenter-dark .tag:not(body) .delete,html.theme--documenter-dark .content kbd:not(body) .delete,html.theme--documenter-dark .docstring>section>a.docs-sourcelink:not(body) .delete{margin-left:.25rem;margin-right:-.375rem}html.theme--documenter-dark .tag.is-white:not(body),html.theme--documenter-dark .content kbd.is-white:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-white:not(body){background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .tag.is-black:not(body),html.theme--documenter-dark .content kbd.is-black:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-black:not(body){background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .tag.is-light:not(body),html.theme--documenter-dark .content kbd.is-light:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-light:not(body){background-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .tag.is-dark:not(body),html.theme--documenter-dark .content kbd:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-dark:not(body),html.theme--documenter-dark .content .docstring>section>kbd:not(body){background-color:#282f2f;color:#fff}html.theme--documenter-dark .tag.is-primary:not(body),html.theme--documenter-dark .content kbd.is-primary:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink:not(body){background-color:#375a7f;color:#fff}html.theme--documenter-dark .tag.is-primary.is-light:not(body),html.theme--documenter-dark .content kbd.is-primary.is-light:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-light:not(body){background-color:#f1f5f9;color:#4d7eb2}html.theme--documenter-dark .tag.is-link:not(body),html.theme--documenter-dark .content kbd.is-link:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-link:not(body){background-color:#1abc9c;color:#fff}html.theme--documenter-dark .tag.is-link.is-light:not(body),html.theme--documenter-dark .content kbd.is-link.is-light:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-link.is-light:not(body){background-color:#edfdf9;color:#15987e}html.theme--documenter-dark .tag.is-info:not(body),html.theme--documenter-dark .content kbd.is-info:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-info:not(body){background-color:#024c7d;color:#fff}html.theme--documenter-dark .tag.is-info.is-light:not(body),html.theme--documenter-dark .content kbd.is-info.is-light:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-info.is-light:not(body){background-color:#ebf7ff;color:#0e9dfb}html.theme--documenter-dark .tag.is-success:not(body),html.theme--documenter-dark .content kbd.is-success:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-success:not(body){background-color:#008438;color:#fff}html.theme--documenter-dark .tag.is-success.is-light:not(body),html.theme--documenter-dark .content kbd.is-success.is-light:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-success.is-light:not(body){background-color:#ebfff3;color:#00eb64}html.theme--documenter-dark .tag.is-warning:not(body),html.theme--documenter-dark .content kbd.is-warning:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-warning:not(body){background-color:#ad8100;color:#fff}html.theme--documenter-dark .tag.is-warning.is-light:not(body),html.theme--documenter-dark .content kbd.is-warning.is-light:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-warning.is-light:not(body){background-color:#fffaeb;color:#d19c00}html.theme--documenter-dark .tag.is-danger:not(body),html.theme--documenter-dark .content kbd.is-danger:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-danger:not(body){background-color:#9e1b0d;color:#fff}html.theme--documenter-dark .tag.is-danger.is-light:not(body),html.theme--documenter-dark .content kbd.is-danger.is-light:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-danger.is-light:not(body){background-color:#fdeeec;color:#ec311d}html.theme--documenter-dark .tag.is-normal:not(body),html.theme--documenter-dark .content kbd.is-normal:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-normal:not(body){font-size:.75rem}html.theme--documenter-dark .tag.is-medium:not(body),html.theme--documenter-dark .content kbd.is-medium:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-medium:not(body){font-size:1rem}html.theme--documenter-dark .tag.is-large:not(body),html.theme--documenter-dark .content kbd.is-large:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-large:not(body){font-size:1.25rem}html.theme--documenter-dark .tag:not(body) .icon:first-child:not(:last-child),html.theme--documenter-dark .content kbd:not(body) .icon:first-child:not(:last-child),html.theme--documenter-dark .docstring>section>a.docs-sourcelink:not(body) .icon:first-child:not(:last-child){margin-left:-.375em;margin-right:.1875em}html.theme--documenter-dark .tag:not(body) .icon:last-child:not(:first-child),html.theme--documenter-dark .content kbd:not(body) .icon:last-child:not(:first-child),html.theme--documenter-dark .docstring>section>a.docs-sourcelink:not(body) .icon:last-child:not(:first-child){margin-left:.1875em;margin-right:-.375em}html.theme--documenter-dark .tag:not(body) .icon:first-child:last-child,html.theme--documenter-dark .content kbd:not(body) .icon:first-child:last-child,html.theme--documenter-dark .docstring>section>a.docs-sourcelink:not(body) .icon:first-child:last-child{margin-left:-.375em;margin-right:-.375em}html.theme--documenter-dark .tag.is-delete:not(body),html.theme--documenter-dark .content kbd.is-delete:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body){margin-left:1px;padding:0;position:relative;width:2em}html.theme--documenter-dark .tag.is-delete:not(body)::before,html.theme--documenter-dark .content kbd.is-delete:not(body)::before,html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body)::before,html.theme--documenter-dark .tag.is-delete:not(body)::after,html.theme--documenter-dark .content kbd.is-delete:not(body)::after,html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body)::after{background-color:currentColor;content:"";display:block;left:50%;position:absolute;top:50%;transform:translateX(-50%) translateY(-50%) rotate(45deg);transform-origin:center center}html.theme--documenter-dark .tag.is-delete:not(body)::before,html.theme--documenter-dark .content kbd.is-delete:not(body)::before,html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body)::before{height:1px;width:50%}html.theme--documenter-dark .tag.is-delete:not(body)::after,html.theme--documenter-dark .content kbd.is-delete:not(body)::after,html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body)::after{height:50%;width:1px}html.theme--documenter-dark .tag.is-delete:not(body):hover,html.theme--documenter-dark .content kbd.is-delete:not(body):hover,html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body):hover,html.theme--documenter-dark .tag.is-delete:not(body):focus,html.theme--documenter-dark .content kbd.is-delete:not(body):focus,html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body):focus{background-color:#1d2122}html.theme--documenter-dark .tag.is-delete:not(body):active,html.theme--documenter-dark .content kbd.is-delete:not(body):active,html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-delete:not(body):active{background-color:#111414}html.theme--documenter-dark .tag.is-rounded:not(body),html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:not(body),html.theme--documenter-dark .content kbd.is-rounded:not(body),html.theme--documenter-dark #documenter .docs-sidebar .content form.docs-search>input:not(body),html.theme--documenter-dark .docstring>section>a.docs-sourcelink.is-rounded:not(body){border-radius:9999px}html.theme--documenter-dark a.tag:hover,html.theme--documenter-dark .docstring>section>a.docs-sourcelink:hover{text-decoration:underline}html.theme--documenter-dark .title,html.theme--documenter-dark .subtitle{word-break:break-word}html.theme--documenter-dark .title em,html.theme--documenter-dark .title span,html.theme--documenter-dark .subtitle em,html.theme--documenter-dark .subtitle span{font-weight:inherit}html.theme--documenter-dark .title sub,html.theme--documenter-dark .subtitle sub{font-size:.75em}html.theme--documenter-dark .title sup,html.theme--documenter-dark .subtitle sup{font-size:.75em}html.theme--documenter-dark .title .tag,html.theme--documenter-dark .title .content kbd,html.theme--documenter-dark .content .title kbd,html.theme--documenter-dark .title .docstring>section>a.docs-sourcelink,html.theme--documenter-dark .subtitle .tag,html.theme--documenter-dark .subtitle .content kbd,html.theme--documenter-dark .content .subtitle kbd,html.theme--documenter-dark .subtitle .docstring>section>a.docs-sourcelink{vertical-align:middle}html.theme--documenter-dark .title{color:#fff;font-size:2rem;font-weight:500;line-height:1.125}html.theme--documenter-dark .title strong{color:inherit;font-weight:inherit}html.theme--documenter-dark .title:not(.is-spaced)+.subtitle{margin-top:-1.25rem}html.theme--documenter-dark .title.is-1{font-size:3rem}html.theme--documenter-dark .title.is-2{font-size:2.5rem}html.theme--documenter-dark .title.is-3{font-size:2rem}html.theme--documenter-dark .title.is-4{font-size:1.5rem}html.theme--documenter-dark .title.is-5{font-size:1.25rem}html.theme--documenter-dark .title.is-6{font-size:1rem}html.theme--documenter-dark .title.is-7{font-size:.75rem}html.theme--documenter-dark .subtitle{color:#8c9b9d;font-size:1.25rem;font-weight:400;line-height:1.25}html.theme--documenter-dark .subtitle strong{color:#8c9b9d;font-weight:600}html.theme--documenter-dark .subtitle:not(.is-spaced)+.title{margin-top:-1.25rem}html.theme--documenter-dark .subtitle.is-1{font-size:3rem}html.theme--documenter-dark .subtitle.is-2{font-size:2.5rem}html.theme--documenter-dark .subtitle.is-3{font-size:2rem}html.theme--documenter-dark .subtitle.is-4{font-size:1.5rem}html.theme--documenter-dark .subtitle.is-5{font-size:1.25rem}html.theme--documenter-dark .subtitle.is-6{font-size:1rem}html.theme--documenter-dark .subtitle.is-7{font-size:.75rem}html.theme--documenter-dark .heading{display:block;font-size:11px;letter-spacing:1px;margin-bottom:5px;text-transform:uppercase}html.theme--documenter-dark .number{align-items:center;background-color:#282f2f;border-radius:9999px;display:inline-flex;font-size:1.25rem;height:2em;justify-content:center;margin-right:1.5rem;min-width:2.5em;padding:0.25rem 0.5rem;text-align:center;vertical-align:top}html.theme--documenter-dark .select select,html.theme--documenter-dark .textarea,html.theme--documenter-dark .input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input{background-color:#1f2424;border-color:#5e6d6f;border-radius:.4em;color:#dbdee0}html.theme--documenter-dark .select select::-moz-placeholder,html.theme--documenter-dark .textarea::-moz-placeholder,html.theme--documenter-dark .input::-moz-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input::-moz-placeholder{color:#868c98}html.theme--documenter-dark .select select::-webkit-input-placeholder,html.theme--documenter-dark .textarea::-webkit-input-placeholder,html.theme--documenter-dark .input::-webkit-input-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input::-webkit-input-placeholder{color:#868c98}html.theme--documenter-dark .select select:-moz-placeholder,html.theme--documenter-dark .textarea:-moz-placeholder,html.theme--documenter-dark .input:-moz-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:-moz-placeholder{color:#868c98}html.theme--documenter-dark .select select:-ms-input-placeholder,html.theme--documenter-dark .textarea:-ms-input-placeholder,html.theme--documenter-dark .input:-ms-input-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:-ms-input-placeholder{color:#868c98}html.theme--documenter-dark .select select:hover,html.theme--documenter-dark .textarea:hover,html.theme--documenter-dark .input:hover,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:hover,html.theme--documenter-dark .select select.is-hovered,html.theme--documenter-dark .is-hovered.textarea,html.theme--documenter-dark .is-hovered.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-hovered{border-color:#8c9b9d}html.theme--documenter-dark .select select:focus,html.theme--documenter-dark .textarea:focus,html.theme--documenter-dark .input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:focus,html.theme--documenter-dark .select select.is-focused,html.theme--documenter-dark .is-focused.textarea,html.theme--documenter-dark .is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .select select:active,html.theme--documenter-dark .textarea:active,html.theme--documenter-dark .input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:active,html.theme--documenter-dark .select select.is-active,html.theme--documenter-dark .is-active.textarea,html.theme--documenter-dark .is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{border-color:#1abc9c;box-shadow:0 0 0 0.125em rgba(26,188,156,0.25)}html.theme--documenter-dark .select select[disabled],html.theme--documenter-dark .textarea[disabled],html.theme--documenter-dark .input[disabled],html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input[disabled],fieldset[disabled] html.theme--documenter-dark .select select,fieldset[disabled] html.theme--documenter-dark .textarea,fieldset[disabled] html.theme--documenter-dark .input,fieldset[disabled] html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input{background-color:#8c9b9d;border-color:#282f2f;box-shadow:none;color:#fff}html.theme--documenter-dark .select select[disabled]::-moz-placeholder,html.theme--documenter-dark .textarea[disabled]::-moz-placeholder,html.theme--documenter-dark .input[disabled]::-moz-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input[disabled]::-moz-placeholder,fieldset[disabled] html.theme--documenter-dark .select select::-moz-placeholder,fieldset[disabled] html.theme--documenter-dark .textarea::-moz-placeholder,fieldset[disabled] html.theme--documenter-dark .input::-moz-placeholder,fieldset[disabled] html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input::-moz-placeholder{color:rgba(255,255,255,0.3)}html.theme--documenter-dark .select select[disabled]::-webkit-input-placeholder,html.theme--documenter-dark .textarea[disabled]::-webkit-input-placeholder,html.theme--documenter-dark .input[disabled]::-webkit-input-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input[disabled]::-webkit-input-placeholder,fieldset[disabled] html.theme--documenter-dark .select select::-webkit-input-placeholder,fieldset[disabled] html.theme--documenter-dark .textarea::-webkit-input-placeholder,fieldset[disabled] html.theme--documenter-dark .input::-webkit-input-placeholder,fieldset[disabled] html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input::-webkit-input-placeholder{color:rgba(255,255,255,0.3)}html.theme--documenter-dark .select select[disabled]:-moz-placeholder,html.theme--documenter-dark .textarea[disabled]:-moz-placeholder,html.theme--documenter-dark .input[disabled]:-moz-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input[disabled]:-moz-placeholder,fieldset[disabled] html.theme--documenter-dark .select select:-moz-placeholder,fieldset[disabled] html.theme--documenter-dark .textarea:-moz-placeholder,fieldset[disabled] html.theme--documenter-dark .input:-moz-placeholder,fieldset[disabled] html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:-moz-placeholder{color:rgba(255,255,255,0.3)}html.theme--documenter-dark .select select[disabled]:-ms-input-placeholder,html.theme--documenter-dark .textarea[disabled]:-ms-input-placeholder,html.theme--documenter-dark .input[disabled]:-ms-input-placeholder,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input[disabled]:-ms-input-placeholder,fieldset[disabled] html.theme--documenter-dark .select select:-ms-input-placeholder,fieldset[disabled] html.theme--documenter-dark .textarea:-ms-input-placeholder,fieldset[disabled] html.theme--documenter-dark .input:-ms-input-placeholder,fieldset[disabled] html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input:-ms-input-placeholder{color:rgba(255,255,255,0.3)}html.theme--documenter-dark .textarea,html.theme--documenter-dark .input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input{box-shadow:inset 0 0.0625em 0.125em rgba(10,10,10,0.05);max-width:100%;width:100%}html.theme--documenter-dark .textarea[readonly],html.theme--documenter-dark .input[readonly],html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input[readonly]{box-shadow:none}html.theme--documenter-dark .is-white.textarea,html.theme--documenter-dark .is-white.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-white{border-color:#fff}html.theme--documenter-dark .is-white.textarea:focus,html.theme--documenter-dark .is-white.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-white:focus,html.theme--documenter-dark .is-white.is-focused.textarea,html.theme--documenter-dark .is-white.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-white.textarea:active,html.theme--documenter-dark .is-white.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-white:active,html.theme--documenter-dark .is-white.is-active.textarea,html.theme--documenter-dark .is-white.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(255,255,255,0.25)}html.theme--documenter-dark .is-black.textarea,html.theme--documenter-dark .is-black.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-black{border-color:#0a0a0a}html.theme--documenter-dark .is-black.textarea:focus,html.theme--documenter-dark .is-black.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-black:focus,html.theme--documenter-dark .is-black.is-focused.textarea,html.theme--documenter-dark .is-black.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-black.textarea:active,html.theme--documenter-dark .is-black.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-black:active,html.theme--documenter-dark .is-black.is-active.textarea,html.theme--documenter-dark .is-black.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(10,10,10,0.25)}html.theme--documenter-dark .is-light.textarea,html.theme--documenter-dark .is-light.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-light{border-color:#ecf0f1}html.theme--documenter-dark .is-light.textarea:focus,html.theme--documenter-dark .is-light.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-light:focus,html.theme--documenter-dark .is-light.is-focused.textarea,html.theme--documenter-dark .is-light.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-light.textarea:active,html.theme--documenter-dark .is-light.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-light:active,html.theme--documenter-dark .is-light.is-active.textarea,html.theme--documenter-dark .is-light.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(236,240,241,0.25)}html.theme--documenter-dark .is-dark.textarea,html.theme--documenter-dark .content kbd.textarea,html.theme--documenter-dark .is-dark.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-dark,html.theme--documenter-dark .content kbd.input{border-color:#282f2f}html.theme--documenter-dark .is-dark.textarea:focus,html.theme--documenter-dark .content kbd.textarea:focus,html.theme--documenter-dark .is-dark.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-dark:focus,html.theme--documenter-dark .content kbd.input:focus,html.theme--documenter-dark .is-dark.is-focused.textarea,html.theme--documenter-dark .content kbd.is-focused.textarea,html.theme--documenter-dark .is-dark.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .content kbd.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar .content form.docs-search>input.is-focused,html.theme--documenter-dark .is-dark.textarea:active,html.theme--documenter-dark .content kbd.textarea:active,html.theme--documenter-dark .is-dark.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-dark:active,html.theme--documenter-dark .content kbd.input:active,html.theme--documenter-dark .is-dark.is-active.textarea,html.theme--documenter-dark .content kbd.is-active.textarea,html.theme--documenter-dark .is-dark.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active,html.theme--documenter-dark .content kbd.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar .content form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(40,47,47,0.25)}html.theme--documenter-dark .is-primary.textarea,html.theme--documenter-dark .docstring>section>a.textarea.docs-sourcelink,html.theme--documenter-dark .is-primary.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-primary,html.theme--documenter-dark .docstring>section>a.input.docs-sourcelink{border-color:#375a7f}html.theme--documenter-dark .is-primary.textarea:focus,html.theme--documenter-dark .docstring>section>a.textarea.docs-sourcelink:focus,html.theme--documenter-dark .is-primary.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-primary:focus,html.theme--documenter-dark .docstring>section>a.input.docs-sourcelink:focus,html.theme--documenter-dark .is-primary.is-focused.textarea,html.theme--documenter-dark .docstring>section>a.is-focused.textarea.docs-sourcelink,html.theme--documenter-dark .is-primary.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .docstring>section>a.is-focused.input.docs-sourcelink,html.theme--documenter-dark .is-primary.textarea:active,html.theme--documenter-dark .docstring>section>a.textarea.docs-sourcelink:active,html.theme--documenter-dark .is-primary.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-primary:active,html.theme--documenter-dark .docstring>section>a.input.docs-sourcelink:active,html.theme--documenter-dark .is-primary.is-active.textarea,html.theme--documenter-dark .docstring>section>a.is-active.textarea.docs-sourcelink,html.theme--documenter-dark .is-primary.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active,html.theme--documenter-dark .docstring>section>a.is-active.input.docs-sourcelink{box-shadow:0 0 0 0.125em rgba(55,90,127,0.25)}html.theme--documenter-dark .is-link.textarea,html.theme--documenter-dark .is-link.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-link{border-color:#1abc9c}html.theme--documenter-dark .is-link.textarea:focus,html.theme--documenter-dark .is-link.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-link:focus,html.theme--documenter-dark .is-link.is-focused.textarea,html.theme--documenter-dark .is-link.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-link.textarea:active,html.theme--documenter-dark .is-link.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-link:active,html.theme--documenter-dark .is-link.is-active.textarea,html.theme--documenter-dark .is-link.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(26,188,156,0.25)}html.theme--documenter-dark .is-info.textarea,html.theme--documenter-dark .is-info.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-info{border-color:#024c7d}html.theme--documenter-dark .is-info.textarea:focus,html.theme--documenter-dark .is-info.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-info:focus,html.theme--documenter-dark .is-info.is-focused.textarea,html.theme--documenter-dark .is-info.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-info.textarea:active,html.theme--documenter-dark .is-info.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-info:active,html.theme--documenter-dark .is-info.is-active.textarea,html.theme--documenter-dark .is-info.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(2,76,125,0.25)}html.theme--documenter-dark .is-success.textarea,html.theme--documenter-dark .is-success.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-success{border-color:#008438}html.theme--documenter-dark .is-success.textarea:focus,html.theme--documenter-dark .is-success.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-success:focus,html.theme--documenter-dark .is-success.is-focused.textarea,html.theme--documenter-dark .is-success.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-success.textarea:active,html.theme--documenter-dark .is-success.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-success:active,html.theme--documenter-dark .is-success.is-active.textarea,html.theme--documenter-dark .is-success.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(0,132,56,0.25)}html.theme--documenter-dark .is-warning.textarea,html.theme--documenter-dark .is-warning.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-warning{border-color:#ad8100}html.theme--documenter-dark .is-warning.textarea:focus,html.theme--documenter-dark .is-warning.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-warning:focus,html.theme--documenter-dark .is-warning.is-focused.textarea,html.theme--documenter-dark .is-warning.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-warning.textarea:active,html.theme--documenter-dark .is-warning.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-warning:active,html.theme--documenter-dark .is-warning.is-active.textarea,html.theme--documenter-dark .is-warning.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(173,129,0,0.25)}html.theme--documenter-dark .is-danger.textarea,html.theme--documenter-dark .is-danger.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-danger{border-color:#9e1b0d}html.theme--documenter-dark .is-danger.textarea:focus,html.theme--documenter-dark .is-danger.input:focus,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-danger:focus,html.theme--documenter-dark .is-danger.is-focused.textarea,html.theme--documenter-dark .is-danger.is-focused.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-focused,html.theme--documenter-dark .is-danger.textarea:active,html.theme--documenter-dark .is-danger.input:active,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-danger:active,html.theme--documenter-dark .is-danger.is-active.textarea,html.theme--documenter-dark .is-danger.is-active.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(158,27,13,0.25)}html.theme--documenter-dark .is-small.textarea,html.theme--documenter-dark .is-small.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input{border-radius:3px;font-size:.75rem}html.theme--documenter-dark .is-medium.textarea,html.theme--documenter-dark .is-medium.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-medium{font-size:1.25rem}html.theme--documenter-dark .is-large.textarea,html.theme--documenter-dark .is-large.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-large{font-size:1.5rem}html.theme--documenter-dark .is-fullwidth.textarea,html.theme--documenter-dark .is-fullwidth.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-fullwidth{display:block;width:100%}html.theme--documenter-dark .is-inline.textarea,html.theme--documenter-dark .is-inline.input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-inline{display:inline;width:auto}html.theme--documenter-dark .input.is-rounded,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input{border-radius:9999px;padding-left:calc(calc(0.75em - 1px) + 0.375em);padding-right:calc(calc(0.75em - 1px) + 0.375em)}html.theme--documenter-dark .input.is-static,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-static{background-color:transparent;border-color:transparent;box-shadow:none;padding-left:0;padding-right:0}html.theme--documenter-dark .textarea{display:block;max-width:100%;min-width:100%;padding:calc(0.75em - 1px);resize:vertical}html.theme--documenter-dark .textarea:not([rows]){max-height:40em;min-height:8em}html.theme--documenter-dark .textarea[rows]{height:initial}html.theme--documenter-dark .textarea.has-fixed-size{resize:none}html.theme--documenter-dark .radio,html.theme--documenter-dark .checkbox{cursor:pointer;display:inline-block;line-height:1.25;position:relative}html.theme--documenter-dark .radio input,html.theme--documenter-dark .checkbox input{cursor:pointer}html.theme--documenter-dark .radio:hover,html.theme--documenter-dark .checkbox:hover{color:#8c9b9d}html.theme--documenter-dark .radio[disabled],html.theme--documenter-dark .checkbox[disabled],fieldset[disabled] html.theme--documenter-dark .radio,fieldset[disabled] html.theme--documenter-dark .checkbox,html.theme--documenter-dark .radio input[disabled],html.theme--documenter-dark .checkbox input[disabled]{color:#fff;cursor:not-allowed}html.theme--documenter-dark .radio+.radio{margin-left:.5em}html.theme--documenter-dark .select{display:inline-block;max-width:100%;position:relative;vertical-align:top}html.theme--documenter-dark .select:not(.is-multiple){height:2.5em}html.theme--documenter-dark .select:not(.is-multiple):not(.is-loading)::after{border-color:#1abc9c;right:1.125em;z-index:4}html.theme--documenter-dark .select.is-rounded select,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.select select{border-radius:9999px;padding-left:1em}html.theme--documenter-dark .select select{cursor:pointer;display:block;font-size:1em;max-width:100%;outline:none}html.theme--documenter-dark .select select::-ms-expand{display:none}html.theme--documenter-dark .select select[disabled]:hover,fieldset[disabled] html.theme--documenter-dark .select select:hover{border-color:#282f2f}html.theme--documenter-dark .select select:not([multiple]){padding-right:2.5em}html.theme--documenter-dark .select select[multiple]{height:auto;padding:0}html.theme--documenter-dark .select select[multiple] option{padding:0.5em 1em}html.theme--documenter-dark .select:not(.is-multiple):not(.is-loading):hover::after{border-color:#8c9b9d}html.theme--documenter-dark .select.is-white:not(:hover)::after{border-color:#fff}html.theme--documenter-dark .select.is-white select{border-color:#fff}html.theme--documenter-dark .select.is-white select:hover,html.theme--documenter-dark .select.is-white select.is-hovered{border-color:#f2f2f2}html.theme--documenter-dark .select.is-white select:focus,html.theme--documenter-dark .select.is-white select.is-focused,html.theme--documenter-dark .select.is-white select:active,html.theme--documenter-dark .select.is-white select.is-active{box-shadow:0 0 0 0.125em rgba(255,255,255,0.25)}html.theme--documenter-dark .select.is-black:not(:hover)::after{border-color:#0a0a0a}html.theme--documenter-dark .select.is-black select{border-color:#0a0a0a}html.theme--documenter-dark .select.is-black select:hover,html.theme--documenter-dark .select.is-black select.is-hovered{border-color:#000}html.theme--documenter-dark .select.is-black select:focus,html.theme--documenter-dark .select.is-black select.is-focused,html.theme--documenter-dark .select.is-black select:active,html.theme--documenter-dark .select.is-black select.is-active{box-shadow:0 0 0 0.125em rgba(10,10,10,0.25)}html.theme--documenter-dark .select.is-light:not(:hover)::after{border-color:#ecf0f1}html.theme--documenter-dark .select.is-light select{border-color:#ecf0f1}html.theme--documenter-dark .select.is-light select:hover,html.theme--documenter-dark .select.is-light select.is-hovered{border-color:#dde4e6}html.theme--documenter-dark .select.is-light select:focus,html.theme--documenter-dark .select.is-light select.is-focused,html.theme--documenter-dark .select.is-light select:active,html.theme--documenter-dark .select.is-light select.is-active{box-shadow:0 0 0 0.125em rgba(236,240,241,0.25)}html.theme--documenter-dark .select.is-dark:not(:hover)::after,html.theme--documenter-dark .content kbd.select:not(:hover)::after{border-color:#282f2f}html.theme--documenter-dark .select.is-dark select,html.theme--documenter-dark .content kbd.select select{border-color:#282f2f}html.theme--documenter-dark .select.is-dark select:hover,html.theme--documenter-dark .content kbd.select select:hover,html.theme--documenter-dark .select.is-dark select.is-hovered,html.theme--documenter-dark .content kbd.select select.is-hovered{border-color:#1d2122}html.theme--documenter-dark .select.is-dark select:focus,html.theme--documenter-dark .content kbd.select select:focus,html.theme--documenter-dark .select.is-dark select.is-focused,html.theme--documenter-dark .content kbd.select select.is-focused,html.theme--documenter-dark .select.is-dark select:active,html.theme--documenter-dark .content kbd.select select:active,html.theme--documenter-dark .select.is-dark select.is-active,html.theme--documenter-dark .content kbd.select select.is-active{box-shadow:0 0 0 0.125em rgba(40,47,47,0.25)}html.theme--documenter-dark .select.is-primary:not(:hover)::after,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink:not(:hover)::after{border-color:#375a7f}html.theme--documenter-dark .select.is-primary select,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink select{border-color:#375a7f}html.theme--documenter-dark .select.is-primary select:hover,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink select:hover,html.theme--documenter-dark .select.is-primary select.is-hovered,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink select.is-hovered{border-color:#2f4d6d}html.theme--documenter-dark .select.is-primary select:focus,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink select:focus,html.theme--documenter-dark .select.is-primary select.is-focused,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink select.is-focused,html.theme--documenter-dark .select.is-primary select:active,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink select:active,html.theme--documenter-dark .select.is-primary select.is-active,html.theme--documenter-dark .docstring>section>a.select.docs-sourcelink select.is-active{box-shadow:0 0 0 0.125em rgba(55,90,127,0.25)}html.theme--documenter-dark .select.is-link:not(:hover)::after{border-color:#1abc9c}html.theme--documenter-dark .select.is-link select{border-color:#1abc9c}html.theme--documenter-dark .select.is-link select:hover,html.theme--documenter-dark .select.is-link select.is-hovered{border-color:#17a689}html.theme--documenter-dark .select.is-link select:focus,html.theme--documenter-dark .select.is-link select.is-focused,html.theme--documenter-dark .select.is-link select:active,html.theme--documenter-dark .select.is-link select.is-active{box-shadow:0 0 0 0.125em rgba(26,188,156,0.25)}html.theme--documenter-dark .select.is-info:not(:hover)::after{border-color:#024c7d}html.theme--documenter-dark .select.is-info select{border-color:#024c7d}html.theme--documenter-dark .select.is-info select:hover,html.theme--documenter-dark .select.is-info select.is-hovered{border-color:#023d64}html.theme--documenter-dark .select.is-info select:focus,html.theme--documenter-dark .select.is-info select.is-focused,html.theme--documenter-dark .select.is-info select:active,html.theme--documenter-dark .select.is-info select.is-active{box-shadow:0 0 0 0.125em rgba(2,76,125,0.25)}html.theme--documenter-dark .select.is-success:not(:hover)::after{border-color:#008438}html.theme--documenter-dark .select.is-success select{border-color:#008438}html.theme--documenter-dark .select.is-success select:hover,html.theme--documenter-dark .select.is-success select.is-hovered{border-color:#006b2d}html.theme--documenter-dark .select.is-success select:focus,html.theme--documenter-dark .select.is-success select.is-focused,html.theme--documenter-dark .select.is-success select:active,html.theme--documenter-dark .select.is-success select.is-active{box-shadow:0 0 0 0.125em rgba(0,132,56,0.25)}html.theme--documenter-dark .select.is-warning:not(:hover)::after{border-color:#ad8100}html.theme--documenter-dark .select.is-warning select{border-color:#ad8100}html.theme--documenter-dark .select.is-warning select:hover,html.theme--documenter-dark .select.is-warning select.is-hovered{border-color:#946e00}html.theme--documenter-dark .select.is-warning select:focus,html.theme--documenter-dark .select.is-warning select.is-focused,html.theme--documenter-dark .select.is-warning select:active,html.theme--documenter-dark .select.is-warning select.is-active{box-shadow:0 0 0 0.125em rgba(173,129,0,0.25)}html.theme--documenter-dark .select.is-danger:not(:hover)::after{border-color:#9e1b0d}html.theme--documenter-dark .select.is-danger select{border-color:#9e1b0d}html.theme--documenter-dark .select.is-danger select:hover,html.theme--documenter-dark .select.is-danger select.is-hovered{border-color:#86170b}html.theme--documenter-dark .select.is-danger select:focus,html.theme--documenter-dark .select.is-danger select.is-focused,html.theme--documenter-dark .select.is-danger select:active,html.theme--documenter-dark .select.is-danger select.is-active{box-shadow:0 0 0 0.125em rgba(158,27,13,0.25)}html.theme--documenter-dark .select.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.select{border-radius:3px;font-size:.75rem}html.theme--documenter-dark .select.is-medium{font-size:1.25rem}html.theme--documenter-dark .select.is-large{font-size:1.5rem}html.theme--documenter-dark .select.is-disabled::after{border-color:#fff !important;opacity:0.5}html.theme--documenter-dark .select.is-fullwidth{width:100%}html.theme--documenter-dark .select.is-fullwidth select{width:100%}html.theme--documenter-dark .select.is-loading::after{margin-top:0;position:absolute;right:.625em;top:0.625em;transform:none}html.theme--documenter-dark .select.is-loading.is-small:after,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-loading:after{font-size:.75rem}html.theme--documenter-dark .select.is-loading.is-medium:after{font-size:1.25rem}html.theme--documenter-dark .select.is-loading.is-large:after{font-size:1.5rem}html.theme--documenter-dark .file{align-items:stretch;display:flex;justify-content:flex-start;position:relative}html.theme--documenter-dark .file.is-white .file-cta{background-color:#fff;border-color:transparent;color:#0a0a0a}html.theme--documenter-dark .file.is-white:hover .file-cta,html.theme--documenter-dark .file.is-white.is-hovered .file-cta{background-color:#f9f9f9;border-color:transparent;color:#0a0a0a}html.theme--documenter-dark .file.is-white:focus .file-cta,html.theme--documenter-dark .file.is-white.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(255,255,255,0.25);color:#0a0a0a}html.theme--documenter-dark .file.is-white:active .file-cta,html.theme--documenter-dark .file.is-white.is-active .file-cta{background-color:#f2f2f2;border-color:transparent;color:#0a0a0a}html.theme--documenter-dark .file.is-black .file-cta{background-color:#0a0a0a;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-black:hover .file-cta,html.theme--documenter-dark .file.is-black.is-hovered .file-cta{background-color:#040404;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-black:focus .file-cta,html.theme--documenter-dark .file.is-black.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(10,10,10,0.25);color:#fff}html.theme--documenter-dark .file.is-black:active .file-cta,html.theme--documenter-dark .file.is-black.is-active .file-cta{background-color:#000;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-light .file-cta{background-color:#ecf0f1;border-color:transparent;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .file.is-light:hover .file-cta,html.theme--documenter-dark .file.is-light.is-hovered .file-cta{background-color:#e5eaec;border-color:transparent;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .file.is-light:focus .file-cta,html.theme--documenter-dark .file.is-light.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(236,240,241,0.25);color:rgba(0,0,0,0.7)}html.theme--documenter-dark .file.is-light:active .file-cta,html.theme--documenter-dark .file.is-light.is-active .file-cta{background-color:#dde4e6;border-color:transparent;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .file.is-dark .file-cta,html.theme--documenter-dark .content kbd.file .file-cta{background-color:#282f2f;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-dark:hover .file-cta,html.theme--documenter-dark .content kbd.file:hover .file-cta,html.theme--documenter-dark .file.is-dark.is-hovered .file-cta,html.theme--documenter-dark .content kbd.file.is-hovered .file-cta{background-color:#232829;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-dark:focus .file-cta,html.theme--documenter-dark .content kbd.file:focus .file-cta,html.theme--documenter-dark .file.is-dark.is-focused .file-cta,html.theme--documenter-dark .content kbd.file.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(40,47,47,0.25);color:#fff}html.theme--documenter-dark .file.is-dark:active .file-cta,html.theme--documenter-dark .content kbd.file:active .file-cta,html.theme--documenter-dark .file.is-dark.is-active .file-cta,html.theme--documenter-dark .content kbd.file.is-active .file-cta{background-color:#1d2122;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-primary .file-cta,html.theme--documenter-dark .docstring>section>a.file.docs-sourcelink .file-cta{background-color:#375a7f;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-primary:hover .file-cta,html.theme--documenter-dark .docstring>section>a.file.docs-sourcelink:hover .file-cta,html.theme--documenter-dark .file.is-primary.is-hovered .file-cta,html.theme--documenter-dark .docstring>section>a.file.is-hovered.docs-sourcelink .file-cta{background-color:#335476;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-primary:focus .file-cta,html.theme--documenter-dark .docstring>section>a.file.docs-sourcelink:focus .file-cta,html.theme--documenter-dark .file.is-primary.is-focused .file-cta,html.theme--documenter-dark .docstring>section>a.file.is-focused.docs-sourcelink .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(55,90,127,0.25);color:#fff}html.theme--documenter-dark .file.is-primary:active .file-cta,html.theme--documenter-dark .docstring>section>a.file.docs-sourcelink:active .file-cta,html.theme--documenter-dark .file.is-primary.is-active .file-cta,html.theme--documenter-dark .docstring>section>a.file.is-active.docs-sourcelink .file-cta{background-color:#2f4d6d;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-link .file-cta{background-color:#1abc9c;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-link:hover .file-cta,html.theme--documenter-dark .file.is-link.is-hovered .file-cta{background-color:#18b193;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-link:focus .file-cta,html.theme--documenter-dark .file.is-link.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(26,188,156,0.25);color:#fff}html.theme--documenter-dark .file.is-link:active .file-cta,html.theme--documenter-dark .file.is-link.is-active .file-cta{background-color:#17a689;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-info .file-cta{background-color:#024c7d;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-info:hover .file-cta,html.theme--documenter-dark .file.is-info.is-hovered .file-cta{background-color:#024470;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-info:focus .file-cta,html.theme--documenter-dark .file.is-info.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(2,76,125,0.25);color:#fff}html.theme--documenter-dark .file.is-info:active .file-cta,html.theme--documenter-dark .file.is-info.is-active .file-cta{background-color:#023d64;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-success .file-cta{background-color:#008438;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-success:hover .file-cta,html.theme--documenter-dark .file.is-success.is-hovered .file-cta{background-color:#073;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-success:focus .file-cta,html.theme--documenter-dark .file.is-success.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(0,132,56,0.25);color:#fff}html.theme--documenter-dark .file.is-success:active .file-cta,html.theme--documenter-dark .file.is-success.is-active .file-cta{background-color:#006b2d;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-warning .file-cta{background-color:#ad8100;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-warning:hover .file-cta,html.theme--documenter-dark .file.is-warning.is-hovered .file-cta{background-color:#a07700;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-warning:focus .file-cta,html.theme--documenter-dark .file.is-warning.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(173,129,0,0.25);color:#fff}html.theme--documenter-dark .file.is-warning:active .file-cta,html.theme--documenter-dark .file.is-warning.is-active .file-cta{background-color:#946e00;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-danger .file-cta{background-color:#9e1b0d;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-danger:hover .file-cta,html.theme--documenter-dark .file.is-danger.is-hovered .file-cta{background-color:#92190c;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-danger:focus .file-cta,html.theme--documenter-dark .file.is-danger.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(158,27,13,0.25);color:#fff}html.theme--documenter-dark .file.is-danger:active .file-cta,html.theme--documenter-dark .file.is-danger.is-active .file-cta{background-color:#86170b;border-color:transparent;color:#fff}html.theme--documenter-dark .file.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.file{font-size:.75rem}html.theme--documenter-dark .file.is-normal{font-size:1rem}html.theme--documenter-dark .file.is-medium{font-size:1.25rem}html.theme--documenter-dark .file.is-medium .file-icon .fa{font-size:21px}html.theme--documenter-dark .file.is-large{font-size:1.5rem}html.theme--documenter-dark .file.is-large .file-icon .fa{font-size:28px}html.theme--documenter-dark .file.has-name .file-cta{border-bottom-right-radius:0;border-top-right-radius:0}html.theme--documenter-dark .file.has-name .file-name{border-bottom-left-radius:0;border-top-left-radius:0}html.theme--documenter-dark .file.has-name.is-empty .file-cta{border-radius:.4em}html.theme--documenter-dark .file.has-name.is-empty .file-name{display:none}html.theme--documenter-dark .file.is-boxed .file-label{flex-direction:column}html.theme--documenter-dark .file.is-boxed .file-cta{flex-direction:column;height:auto;padding:1em 3em}html.theme--documenter-dark .file.is-boxed .file-name{border-width:0 1px 1px}html.theme--documenter-dark .file.is-boxed .file-icon{height:1.5em;width:1.5em}html.theme--documenter-dark .file.is-boxed .file-icon .fa{font-size:21px}html.theme--documenter-dark .file.is-boxed.is-small .file-icon .fa,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-boxed .file-icon .fa{font-size:14px}html.theme--documenter-dark .file.is-boxed.is-medium .file-icon .fa{font-size:28px}html.theme--documenter-dark .file.is-boxed.is-large .file-icon .fa{font-size:35px}html.theme--documenter-dark .file.is-boxed.has-name .file-cta{border-radius:.4em .4em 0 0}html.theme--documenter-dark .file.is-boxed.has-name .file-name{border-radius:0 0 .4em .4em;border-width:0 1px 1px}html.theme--documenter-dark .file.is-centered{justify-content:center}html.theme--documenter-dark .file.is-fullwidth .file-label{width:100%}html.theme--documenter-dark .file.is-fullwidth .file-name{flex-grow:1;max-width:none}html.theme--documenter-dark .file.is-right{justify-content:flex-end}html.theme--documenter-dark .file.is-right .file-cta{border-radius:0 .4em .4em 0}html.theme--documenter-dark .file.is-right .file-name{border-radius:.4em 0 0 .4em;border-width:1px 0 1px 1px;order:-1}html.theme--documenter-dark .file-label{align-items:stretch;display:flex;cursor:pointer;justify-content:flex-start;overflow:hidden;position:relative}html.theme--documenter-dark .file-label:hover .file-cta{background-color:#232829;color:#f2f2f2}html.theme--documenter-dark .file-label:hover .file-name{border-color:#596668}html.theme--documenter-dark .file-label:active .file-cta{background-color:#1d2122;color:#f2f2f2}html.theme--documenter-dark .file-label:active .file-name{border-color:#535f61}html.theme--documenter-dark .file-input{height:100%;left:0;opacity:0;outline:none;position:absolute;top:0;width:100%}html.theme--documenter-dark .file-cta,html.theme--documenter-dark .file-name{border-color:#5e6d6f;border-radius:.4em;font-size:1em;padding-left:1em;padding-right:1em;white-space:nowrap}html.theme--documenter-dark .file-cta{background-color:#282f2f;color:#fff}html.theme--documenter-dark .file-name{border-color:#5e6d6f;border-style:solid;border-width:1px 1px 1px 0;display:block;max-width:16em;overflow:hidden;text-align:inherit;text-overflow:ellipsis}html.theme--documenter-dark .file-icon{align-items:center;display:flex;height:1em;justify-content:center;margin-right:.5em;width:1em}html.theme--documenter-dark .file-icon .fa{font-size:14px}html.theme--documenter-dark .label{color:#f2f2f2;display:block;font-size:1rem;font-weight:700}html.theme--documenter-dark .label:not(:last-child){margin-bottom:0.5em}html.theme--documenter-dark .label.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.label{font-size:.75rem}html.theme--documenter-dark .label.is-medium{font-size:1.25rem}html.theme--documenter-dark .label.is-large{font-size:1.5rem}html.theme--documenter-dark .help{display:block;font-size:.75rem;margin-top:0.25rem}html.theme--documenter-dark .help.is-white{color:#fff}html.theme--documenter-dark .help.is-black{color:#0a0a0a}html.theme--documenter-dark .help.is-light{color:#ecf0f1}html.theme--documenter-dark .help.is-dark,html.theme--documenter-dark .content kbd.help{color:#282f2f}html.theme--documenter-dark .help.is-primary,html.theme--documenter-dark .docstring>section>a.help.docs-sourcelink{color:#375a7f}html.theme--documenter-dark .help.is-link{color:#1abc9c}html.theme--documenter-dark .help.is-info{color:#024c7d}html.theme--documenter-dark .help.is-success{color:#008438}html.theme--documenter-dark .help.is-warning{color:#ad8100}html.theme--documenter-dark .help.is-danger{color:#9e1b0d}html.theme--documenter-dark .field:not(:last-child){margin-bottom:0.75rem}html.theme--documenter-dark .field.has-addons{display:flex;justify-content:flex-start}html.theme--documenter-dark .field.has-addons .control:not(:last-child){margin-right:-1px}html.theme--documenter-dark .field.has-addons .control:not(:first-child):not(:last-child) .button,html.theme--documenter-dark .field.has-addons .control:not(:first-child):not(:last-child) .input,html.theme--documenter-dark .field.has-addons .control:not(:first-child):not(:last-child) #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control:not(:first-child):not(:last-child) form.docs-search>input,html.theme--documenter-dark .field.has-addons .control:not(:first-child):not(:last-child) .select select{border-radius:0}html.theme--documenter-dark .field.has-addons .control:first-child:not(:only-child) .button,html.theme--documenter-dark .field.has-addons .control:first-child:not(:only-child) .input,html.theme--documenter-dark .field.has-addons .control:first-child:not(:only-child) #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control:first-child:not(:only-child) form.docs-search>input,html.theme--documenter-dark .field.has-addons .control:first-child:not(:only-child) .select select{border-bottom-right-radius:0;border-top-right-radius:0}html.theme--documenter-dark .field.has-addons .control:last-child:not(:only-child) .button,html.theme--documenter-dark .field.has-addons .control:last-child:not(:only-child) .input,html.theme--documenter-dark .field.has-addons .control:last-child:not(:only-child) #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control:last-child:not(:only-child) form.docs-search>input,html.theme--documenter-dark .field.has-addons .control:last-child:not(:only-child) .select select{border-bottom-left-radius:0;border-top-left-radius:0}html.theme--documenter-dark .field.has-addons .control .button:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .button.is-hovered:not([disabled]),html.theme--documenter-dark .field.has-addons .control .input:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):hover,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .input.is-hovered:not([disabled]),html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-hovered:not([disabled]),html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-hovered:not([disabled]),html.theme--documenter-dark .field.has-addons .control .select select:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .select select.is-hovered:not([disabled]){z-index:2}html.theme--documenter-dark .field.has-addons .control .button:not([disabled]):focus,html.theme--documenter-dark .field.has-addons .control .button.is-focused:not([disabled]),html.theme--documenter-dark .field.has-addons .control .button:not([disabled]):active,html.theme--documenter-dark .field.has-addons .control .button.is-active:not([disabled]),html.theme--documenter-dark .field.has-addons .control .input:not([disabled]):focus,html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):focus,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):focus,html.theme--documenter-dark .field.has-addons .control .input.is-focused:not([disabled]),html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-focused:not([disabled]),html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-focused:not([disabled]),html.theme--documenter-dark .field.has-addons .control .input:not([disabled]):active,html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):active,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):active,html.theme--documenter-dark .field.has-addons .control .input.is-active:not([disabled]),html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-active:not([disabled]),html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-active:not([disabled]),html.theme--documenter-dark .field.has-addons .control .select select:not([disabled]):focus,html.theme--documenter-dark .field.has-addons .control .select select.is-focused:not([disabled]),html.theme--documenter-dark .field.has-addons .control .select select:not([disabled]):active,html.theme--documenter-dark .field.has-addons .control .select select.is-active:not([disabled]){z-index:3}html.theme--documenter-dark .field.has-addons .control .button:not([disabled]):focus:hover,html.theme--documenter-dark .field.has-addons .control .button.is-focused:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .button:not([disabled]):active:hover,html.theme--documenter-dark .field.has-addons .control .button.is-active:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .input:not([disabled]):focus:hover,html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):focus:hover,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):focus:hover,html.theme--documenter-dark .field.has-addons .control .input.is-focused:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-focused:not([disabled]):hover,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-focused:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .input:not([disabled]):active:hover,html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):active:hover,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):active:hover,html.theme--documenter-dark .field.has-addons .control .input.is-active:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-active:not([disabled]):hover,html.theme--documenter-dark #documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-active:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .select select:not([disabled]):focus:hover,html.theme--documenter-dark .field.has-addons .control .select select.is-focused:not([disabled]):hover,html.theme--documenter-dark .field.has-addons .control .select select:not([disabled]):active:hover,html.theme--documenter-dark .field.has-addons .control .select select.is-active:not([disabled]):hover{z-index:4}html.theme--documenter-dark .field.has-addons .control.is-expanded{flex-grow:1;flex-shrink:1}html.theme--documenter-dark .field.has-addons.has-addons-centered{justify-content:center}html.theme--documenter-dark .field.has-addons.has-addons-right{justify-content:flex-end}html.theme--documenter-dark .field.has-addons.has-addons-fullwidth .control{flex-grow:1;flex-shrink:0}html.theme--documenter-dark .field.is-grouped{display:flex;justify-content:flex-start}html.theme--documenter-dark .field.is-grouped>.control{flex-shrink:0}html.theme--documenter-dark .field.is-grouped>.control:not(:last-child){margin-bottom:0;margin-right:.75rem}html.theme--documenter-dark .field.is-grouped>.control.is-expanded{flex-grow:1;flex-shrink:1}html.theme--documenter-dark .field.is-grouped.is-grouped-centered{justify-content:center}html.theme--documenter-dark .field.is-grouped.is-grouped-right{justify-content:flex-end}html.theme--documenter-dark .field.is-grouped.is-grouped-multiline{flex-wrap:wrap}html.theme--documenter-dark .field.is-grouped.is-grouped-multiline>.control:last-child,html.theme--documenter-dark .field.is-grouped.is-grouped-multiline>.control:not(:last-child){margin-bottom:0.75rem}html.theme--documenter-dark .field.is-grouped.is-grouped-multiline:last-child{margin-bottom:-0.75rem}html.theme--documenter-dark .field.is-grouped.is-grouped-multiline:not(:last-child){margin-bottom:0}@media screen and (min-width: 769px),print{html.theme--documenter-dark .field.is-horizontal{display:flex}}html.theme--documenter-dark .field-label .label{font-size:inherit}@media screen and (max-width: 768px){html.theme--documenter-dark .field-label{margin-bottom:0.5rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .field-label{flex-basis:0;flex-grow:1;flex-shrink:0;margin-right:1.5rem;text-align:right}html.theme--documenter-dark .field-label.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.field-label{font-size:.75rem;padding-top:0.375em}html.theme--documenter-dark .field-label.is-normal{padding-top:0.375em}html.theme--documenter-dark .field-label.is-medium{font-size:1.25rem;padding-top:0.375em}html.theme--documenter-dark .field-label.is-large{font-size:1.5rem;padding-top:0.375em}}html.theme--documenter-dark .field-body .field .field{margin-bottom:0}@media screen and (min-width: 769px),print{html.theme--documenter-dark .field-body{display:flex;flex-basis:0;flex-grow:5;flex-shrink:1}html.theme--documenter-dark .field-body .field{margin-bottom:0}html.theme--documenter-dark .field-body>.field{flex-shrink:1}html.theme--documenter-dark .field-body>.field:not(.is-narrow){flex-grow:1}html.theme--documenter-dark .field-body>.field:not(:last-child){margin-right:.75rem}}html.theme--documenter-dark .control{box-sizing:border-box;clear:both;font-size:1rem;position:relative;text-align:inherit}html.theme--documenter-dark .control.has-icons-left .input:focus~.icon,html.theme--documenter-dark .control.has-icons-left #documenter .docs-sidebar form.docs-search>input:focus~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-left form.docs-search>input:focus~.icon,html.theme--documenter-dark .control.has-icons-left .select:focus~.icon,html.theme--documenter-dark .control.has-icons-right .input:focus~.icon,html.theme--documenter-dark .control.has-icons-right #documenter .docs-sidebar form.docs-search>input:focus~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-right form.docs-search>input:focus~.icon,html.theme--documenter-dark .control.has-icons-right .select:focus~.icon{color:#282f2f}html.theme--documenter-dark .control.has-icons-left .input.is-small~.icon,html.theme--documenter-dark .control.has-icons-left #documenter .docs-sidebar form.docs-search>input~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-left form.docs-search>input~.icon,html.theme--documenter-dark .control.has-icons-left .select.is-small~.icon,html.theme--documenter-dark .control.has-icons-right .input.is-small~.icon,html.theme--documenter-dark .control.has-icons-right #documenter .docs-sidebar form.docs-search>input~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-right form.docs-search>input~.icon,html.theme--documenter-dark .control.has-icons-right .select.is-small~.icon{font-size:.75rem}html.theme--documenter-dark .control.has-icons-left .input.is-medium~.icon,html.theme--documenter-dark .control.has-icons-left #documenter .docs-sidebar form.docs-search>input.is-medium~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-left form.docs-search>input.is-medium~.icon,html.theme--documenter-dark .control.has-icons-left .select.is-medium~.icon,html.theme--documenter-dark .control.has-icons-right .input.is-medium~.icon,html.theme--documenter-dark .control.has-icons-right #documenter .docs-sidebar form.docs-search>input.is-medium~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-right form.docs-search>input.is-medium~.icon,html.theme--documenter-dark .control.has-icons-right .select.is-medium~.icon{font-size:1.25rem}html.theme--documenter-dark .control.has-icons-left .input.is-large~.icon,html.theme--documenter-dark .control.has-icons-left #documenter .docs-sidebar form.docs-search>input.is-large~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-left form.docs-search>input.is-large~.icon,html.theme--documenter-dark .control.has-icons-left .select.is-large~.icon,html.theme--documenter-dark .control.has-icons-right .input.is-large~.icon,html.theme--documenter-dark .control.has-icons-right #documenter .docs-sidebar form.docs-search>input.is-large~.icon,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-right form.docs-search>input.is-large~.icon,html.theme--documenter-dark .control.has-icons-right .select.is-large~.icon{font-size:1.5rem}html.theme--documenter-dark .control.has-icons-left .icon,html.theme--documenter-dark .control.has-icons-right .icon{color:#5e6d6f;height:2.5em;pointer-events:none;position:absolute;top:0;width:2.5em;z-index:4}html.theme--documenter-dark .control.has-icons-left .input,html.theme--documenter-dark .control.has-icons-left #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-left form.docs-search>input,html.theme--documenter-dark .control.has-icons-left .select select{padding-left:2.5em}html.theme--documenter-dark .control.has-icons-left .icon.is-left{left:0}html.theme--documenter-dark .control.has-icons-right .input,html.theme--documenter-dark .control.has-icons-right #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark #documenter .docs-sidebar .control.has-icons-right form.docs-search>input,html.theme--documenter-dark .control.has-icons-right .select select{padding-right:2.5em}html.theme--documenter-dark .control.has-icons-right .icon.is-right{right:0}html.theme--documenter-dark .control.is-loading::after{position:absolute !important;right:.625em;top:0.625em;z-index:4}html.theme--documenter-dark .control.is-loading.is-small:after,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.is-loading:after{font-size:.75rem}html.theme--documenter-dark .control.is-loading.is-medium:after{font-size:1.25rem}html.theme--documenter-dark .control.is-loading.is-large:after{font-size:1.5rem}html.theme--documenter-dark .breadcrumb{font-size:1rem;white-space:nowrap}html.theme--documenter-dark .breadcrumb a{align-items:center;color:#1abc9c;display:flex;justify-content:center;padding:0 .75em}html.theme--documenter-dark .breadcrumb a:hover{color:#1dd2af}html.theme--documenter-dark .breadcrumb li{align-items:center;display:flex}html.theme--documenter-dark .breadcrumb li:first-child a{padding-left:0}html.theme--documenter-dark .breadcrumb li.is-active a{color:#f2f2f2;cursor:default;pointer-events:none}html.theme--documenter-dark .breadcrumb li+li::before{color:#8c9b9d;content:"\0002f"}html.theme--documenter-dark .breadcrumb ul,html.theme--documenter-dark .breadcrumb ol{align-items:flex-start;display:flex;flex-wrap:wrap;justify-content:flex-start}html.theme--documenter-dark .breadcrumb .icon:first-child{margin-right:.5em}html.theme--documenter-dark .breadcrumb .icon:last-child{margin-left:.5em}html.theme--documenter-dark .breadcrumb.is-centered ol,html.theme--documenter-dark .breadcrumb.is-centered ul{justify-content:center}html.theme--documenter-dark .breadcrumb.is-right ol,html.theme--documenter-dark .breadcrumb.is-right ul{justify-content:flex-end}html.theme--documenter-dark .breadcrumb.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.breadcrumb{font-size:.75rem}html.theme--documenter-dark .breadcrumb.is-medium{font-size:1.25rem}html.theme--documenter-dark .breadcrumb.is-large{font-size:1.5rem}html.theme--documenter-dark .breadcrumb.has-arrow-separator li+li::before{content:"\02192"}html.theme--documenter-dark .breadcrumb.has-bullet-separator li+li::before{content:"\02022"}html.theme--documenter-dark .breadcrumb.has-dot-separator li+li::before{content:"\000b7"}html.theme--documenter-dark .breadcrumb.has-succeeds-separator li+li::before{content:"\0227B"}html.theme--documenter-dark .card{background-color:#fff;border-radius:.25rem;box-shadow:#171717;color:#fff;max-width:100%;position:relative}html.theme--documenter-dark .card-footer:first-child,html.theme--documenter-dark .card-content:first-child,html.theme--documenter-dark .card-header:first-child{border-top-left-radius:.25rem;border-top-right-radius:.25rem}html.theme--documenter-dark .card-footer:last-child,html.theme--documenter-dark .card-content:last-child,html.theme--documenter-dark .card-header:last-child{border-bottom-left-radius:.25rem;border-bottom-right-radius:.25rem}html.theme--documenter-dark .card-header{background-color:rgba(0,0,0,0);align-items:stretch;box-shadow:0 0.125em 0.25em rgba(10,10,10,0.1);display:flex}html.theme--documenter-dark .card-header-title{align-items:center;color:#f2f2f2;display:flex;flex-grow:1;font-weight:700;padding:0.75rem 1rem}html.theme--documenter-dark .card-header-title.is-centered{justify-content:center}html.theme--documenter-dark .card-header-icon{-moz-appearance:none;-webkit-appearance:none;appearance:none;background:none;border:none;color:currentColor;font-family:inherit;font-size:1em;margin:0;padding:0;align-items:center;cursor:pointer;display:flex;justify-content:center;padding:0.75rem 1rem}html.theme--documenter-dark .card-image{display:block;position:relative}html.theme--documenter-dark .card-image:first-child img{border-top-left-radius:.25rem;border-top-right-radius:.25rem}html.theme--documenter-dark .card-image:last-child img{border-bottom-left-radius:.25rem;border-bottom-right-radius:.25rem}html.theme--documenter-dark .card-content{background-color:rgba(0,0,0,0);padding:1.5rem}html.theme--documenter-dark .card-footer{background-color:rgba(0,0,0,0);border-top:1px solid #ededed;align-items:stretch;display:flex}html.theme--documenter-dark .card-footer-item{align-items:center;display:flex;flex-basis:0;flex-grow:1;flex-shrink:0;justify-content:center;padding:.75rem}html.theme--documenter-dark .card-footer-item:not(:last-child){border-right:1px solid #ededed}html.theme--documenter-dark .card .media:not(:last-child){margin-bottom:1.5rem}html.theme--documenter-dark .dropdown{display:inline-flex;position:relative;vertical-align:top}html.theme--documenter-dark .dropdown.is-active .dropdown-menu,html.theme--documenter-dark .dropdown.is-hoverable:hover .dropdown-menu{display:block}html.theme--documenter-dark .dropdown.is-right .dropdown-menu{left:auto;right:0}html.theme--documenter-dark .dropdown.is-up .dropdown-menu{bottom:100%;padding-bottom:4px;padding-top:initial;top:auto}html.theme--documenter-dark .dropdown-menu{display:none;left:0;min-width:12rem;padding-top:4px;position:absolute;top:100%;z-index:20}html.theme--documenter-dark .dropdown-content{background-color:#282f2f;border-radius:.4em;box-shadow:#171717;padding-bottom:.5rem;padding-top:.5rem}html.theme--documenter-dark .dropdown-item{color:#fff;display:block;font-size:0.875rem;line-height:1.5;padding:0.375rem 1rem;position:relative}html.theme--documenter-dark a.dropdown-item,html.theme--documenter-dark button.dropdown-item{padding-right:3rem;text-align:inherit;white-space:nowrap;width:100%}html.theme--documenter-dark a.dropdown-item:hover,html.theme--documenter-dark button.dropdown-item:hover{background-color:#282f2f;color:#0a0a0a}html.theme--documenter-dark a.dropdown-item.is-active,html.theme--documenter-dark button.dropdown-item.is-active{background-color:#1abc9c;color:#fff}html.theme--documenter-dark .dropdown-divider{background-color:#ededed;border:none;display:block;height:1px;margin:0.5rem 0}html.theme--documenter-dark .level{align-items:center;justify-content:space-between}html.theme--documenter-dark .level code{border-radius:.4em}html.theme--documenter-dark .level img{display:inline-block;vertical-align:top}html.theme--documenter-dark .level.is-mobile{display:flex}html.theme--documenter-dark .level.is-mobile .level-left,html.theme--documenter-dark .level.is-mobile .level-right{display:flex}html.theme--documenter-dark .level.is-mobile .level-left+.level-right{margin-top:0}html.theme--documenter-dark .level.is-mobile .level-item:not(:last-child){margin-bottom:0;margin-right:.75rem}html.theme--documenter-dark .level.is-mobile .level-item:not(.is-narrow){flex-grow:1}@media screen and (min-width: 769px),print{html.theme--documenter-dark .level{display:flex}html.theme--documenter-dark .level>.level-item:not(.is-narrow){flex-grow:1}}html.theme--documenter-dark .level-item{align-items:center;display:flex;flex-basis:auto;flex-grow:0;flex-shrink:0;justify-content:center}html.theme--documenter-dark .level-item .title,html.theme--documenter-dark .level-item .subtitle{margin-bottom:0}@media screen and (max-width: 768px){html.theme--documenter-dark .level-item:not(:last-child){margin-bottom:.75rem}}html.theme--documenter-dark .level-left,html.theme--documenter-dark .level-right{flex-basis:auto;flex-grow:0;flex-shrink:0}html.theme--documenter-dark .level-left .level-item.is-flexible,html.theme--documenter-dark .level-right .level-item.is-flexible{flex-grow:1}@media screen and (min-width: 769px),print{html.theme--documenter-dark .level-left .level-item:not(:last-child),html.theme--documenter-dark .level-right .level-item:not(:last-child){margin-right:.75rem}}html.theme--documenter-dark .level-left{align-items:center;justify-content:flex-start}@media screen and (max-width: 768px){html.theme--documenter-dark .level-left+.level-right{margin-top:1.5rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .level-left{display:flex}}html.theme--documenter-dark .level-right{align-items:center;justify-content:flex-end}@media screen and (min-width: 769px),print{html.theme--documenter-dark .level-right{display:flex}}html.theme--documenter-dark .media{align-items:flex-start;display:flex;text-align:inherit}html.theme--documenter-dark .media .content:not(:last-child){margin-bottom:.75rem}html.theme--documenter-dark .media .media{border-top:1px solid rgba(94,109,111,0.5);display:flex;padding-top:.75rem}html.theme--documenter-dark .media .media .content:not(:last-child),html.theme--documenter-dark .media .media .control:not(:last-child){margin-bottom:.5rem}html.theme--documenter-dark .media .media .media{padding-top:.5rem}html.theme--documenter-dark .media .media .media+.media{margin-top:.5rem}html.theme--documenter-dark .media+.media{border-top:1px solid rgba(94,109,111,0.5);margin-top:1rem;padding-top:1rem}html.theme--documenter-dark .media.is-large+.media{margin-top:1.5rem;padding-top:1.5rem}html.theme--documenter-dark .media-left,html.theme--documenter-dark .media-right{flex-basis:auto;flex-grow:0;flex-shrink:0}html.theme--documenter-dark .media-left{margin-right:1rem}html.theme--documenter-dark .media-right{margin-left:1rem}html.theme--documenter-dark .media-content{flex-basis:auto;flex-grow:1;flex-shrink:1;text-align:inherit}@media screen and (max-width: 768px){html.theme--documenter-dark .media-content{overflow-x:auto}}html.theme--documenter-dark .menu{font-size:1rem}html.theme--documenter-dark .menu.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.menu{font-size:.75rem}html.theme--documenter-dark .menu.is-medium{font-size:1.25rem}html.theme--documenter-dark .menu.is-large{font-size:1.5rem}html.theme--documenter-dark .menu-list{line-height:1.25}html.theme--documenter-dark .menu-list a{border-radius:3px;color:#fff;display:block;padding:0.5em 0.75em}html.theme--documenter-dark .menu-list a:hover{background-color:#282f2f;color:#f2f2f2}html.theme--documenter-dark .menu-list a.is-active{background-color:#1abc9c;color:#fff}html.theme--documenter-dark .menu-list li ul{border-left:1px solid #5e6d6f;margin:.75em;padding-left:.75em}html.theme--documenter-dark .menu-label{color:#fff;font-size:.75em;letter-spacing:.1em;text-transform:uppercase}html.theme--documenter-dark .menu-label:not(:first-child){margin-top:1em}html.theme--documenter-dark .menu-label:not(:last-child){margin-bottom:1em}html.theme--documenter-dark .message{background-color:#282f2f;border-radius:.4em;font-size:1rem}html.theme--documenter-dark .message strong{color:currentColor}html.theme--documenter-dark .message a:not(.button):not(.tag):not(.dropdown-item){color:currentColor;text-decoration:underline}html.theme--documenter-dark .message.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.message{font-size:.75rem}html.theme--documenter-dark .message.is-medium{font-size:1.25rem}html.theme--documenter-dark .message.is-large{font-size:1.5rem}html.theme--documenter-dark .message.is-white{background-color:#fff}html.theme--documenter-dark .message.is-white .message-header{background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .message.is-white .message-body{border-color:#fff}html.theme--documenter-dark .message.is-black{background-color:#fafafa}html.theme--documenter-dark .message.is-black .message-header{background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .message.is-black .message-body{border-color:#0a0a0a}html.theme--documenter-dark .message.is-light{background-color:#f9fafb}html.theme--documenter-dark .message.is-light .message-header{background-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .message.is-light .message-body{border-color:#ecf0f1}html.theme--documenter-dark .message.is-dark,html.theme--documenter-dark .content kbd.message{background-color:#f9fafa}html.theme--documenter-dark .message.is-dark .message-header,html.theme--documenter-dark .content kbd.message .message-header{background-color:#282f2f;color:#fff}html.theme--documenter-dark .message.is-dark .message-body,html.theme--documenter-dark .content kbd.message .message-body{border-color:#282f2f}html.theme--documenter-dark .message.is-primary,html.theme--documenter-dark .docstring>section>a.message.docs-sourcelink{background-color:#f1f5f9}html.theme--documenter-dark .message.is-primary .message-header,html.theme--documenter-dark .docstring>section>a.message.docs-sourcelink .message-header{background-color:#375a7f;color:#fff}html.theme--documenter-dark .message.is-primary .message-body,html.theme--documenter-dark .docstring>section>a.message.docs-sourcelink .message-body{border-color:#375a7f;color:#4d7eb2}html.theme--documenter-dark .message.is-link{background-color:#edfdf9}html.theme--documenter-dark .message.is-link .message-header{background-color:#1abc9c;color:#fff}html.theme--documenter-dark .message.is-link .message-body{border-color:#1abc9c;color:#15987e}html.theme--documenter-dark .message.is-info{background-color:#ebf7ff}html.theme--documenter-dark .message.is-info .message-header{background-color:#024c7d;color:#fff}html.theme--documenter-dark .message.is-info .message-body{border-color:#024c7d;color:#0e9dfb}html.theme--documenter-dark .message.is-success{background-color:#ebfff3}html.theme--documenter-dark .message.is-success .message-header{background-color:#008438;color:#fff}html.theme--documenter-dark .message.is-success .message-body{border-color:#008438;color:#00eb64}html.theme--documenter-dark .message.is-warning{background-color:#fffaeb}html.theme--documenter-dark .message.is-warning .message-header{background-color:#ad8100;color:#fff}html.theme--documenter-dark .message.is-warning .message-body{border-color:#ad8100;color:#d19c00}html.theme--documenter-dark .message.is-danger{background-color:#fdeeec}html.theme--documenter-dark .message.is-danger .message-header{background-color:#9e1b0d;color:#fff}html.theme--documenter-dark .message.is-danger .message-body{border-color:#9e1b0d;color:#ec311d}html.theme--documenter-dark .message-header{align-items:center;background-color:#fff;border-radius:.4em .4em 0 0;color:rgba(0,0,0,0.7);display:flex;font-weight:700;justify-content:space-between;line-height:1.25;padding:0.75em 1em;position:relative}html.theme--documenter-dark .message-header .delete{flex-grow:0;flex-shrink:0;margin-left:.75em}html.theme--documenter-dark .message-header+.message-body{border-width:0;border-top-left-radius:0;border-top-right-radius:0}html.theme--documenter-dark .message-body{border-color:#5e6d6f;border-radius:.4em;border-style:solid;border-width:0 0 0 4px;color:#fff;padding:1.25em 1.5em}html.theme--documenter-dark .message-body code,html.theme--documenter-dark .message-body pre{background-color:#fff}html.theme--documenter-dark .message-body pre code{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .modal{align-items:center;display:none;flex-direction:column;justify-content:center;overflow:hidden;position:fixed;z-index:40}html.theme--documenter-dark .modal.is-active{display:flex}html.theme--documenter-dark .modal-background{background-color:rgba(10,10,10,0.86)}html.theme--documenter-dark .modal-content,html.theme--documenter-dark .modal-card{margin:0 20px;max-height:calc(100vh - 160px);overflow:auto;position:relative;width:100%}@media screen and (min-width: 769px){html.theme--documenter-dark .modal-content,html.theme--documenter-dark .modal-card{margin:0 auto;max-height:calc(100vh - 40px);width:640px}}html.theme--documenter-dark .modal-close{background:none;height:40px;position:fixed;right:20px;top:20px;width:40px}html.theme--documenter-dark .modal-card{display:flex;flex-direction:column;max-height:calc(100vh - 40px);overflow:hidden;-ms-overflow-y:visible}html.theme--documenter-dark .modal-card-head,html.theme--documenter-dark .modal-card-foot{align-items:center;background-color:#282f2f;display:flex;flex-shrink:0;justify-content:flex-start;padding:20px;position:relative}html.theme--documenter-dark .modal-card-head{border-bottom:1px solid #5e6d6f;border-top-left-radius:8px;border-top-right-radius:8px}html.theme--documenter-dark .modal-card-title{color:#f2f2f2;flex-grow:1;flex-shrink:0;font-size:1.5rem;line-height:1}html.theme--documenter-dark .modal-card-foot{border-bottom-left-radius:8px;border-bottom-right-radius:8px;border-top:1px solid #5e6d6f}html.theme--documenter-dark .modal-card-foot .button:not(:last-child){margin-right:.5em}html.theme--documenter-dark .modal-card-body{-webkit-overflow-scrolling:touch;background-color:#fff;flex-grow:1;flex-shrink:1;overflow:auto;padding:20px}html.theme--documenter-dark .navbar{background-color:#375a7f;min-height:4rem;position:relative;z-index:30}html.theme--documenter-dark .navbar.is-white{background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-white .navbar-brand .navbar-link{color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-white .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-white .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-white .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-white .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-white .navbar-brand .navbar-link.is-active{background-color:#f2f2f2;color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-brand .navbar-link::after{border-color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-burger{color:#0a0a0a}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-white .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-white .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-white .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-white .navbar-end .navbar-link{color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-white .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-white .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-white .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-white .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-white .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-white .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-white .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-white .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-white .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-white .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-white .navbar-end .navbar-link.is-active{background-color:#f2f2f2;color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-white .navbar-end .navbar-link::after{border-color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-white .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-white .navbar-item.has-dropdown.is-active .navbar-link{background-color:#f2f2f2;color:#0a0a0a}html.theme--documenter-dark .navbar.is-white .navbar-dropdown a.navbar-item.is-active{background-color:#fff;color:#0a0a0a}}html.theme--documenter-dark .navbar.is-black{background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-black .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-black .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-black .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-black .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-black .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-black .navbar-brand .navbar-link.is-active{background-color:#000;color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-black .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-black .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-black .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-black .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-black .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-black .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-black .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-black .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-black .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-black .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-black .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-black .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-black .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-black .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-black .navbar-end .navbar-link.is-active{background-color:#000;color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-black .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-black .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-black .navbar-item.has-dropdown.is-active .navbar-link{background-color:#000;color:#fff}html.theme--documenter-dark .navbar.is-black .navbar-dropdown a.navbar-item.is-active{background-color:#0a0a0a;color:#fff}}html.theme--documenter-dark .navbar.is-light{background-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-light .navbar-brand .navbar-link{color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-light .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-light .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-light .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-light .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-light .navbar-brand .navbar-link.is-active{background-color:#dde4e6;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-brand .navbar-link::after{border-color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-burger{color:rgba(0,0,0,0.7)}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-light .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-light .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-light .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-light .navbar-end .navbar-link{color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-light .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-light .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-light .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-light .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-light .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-light .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-light .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-light .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-light .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-light .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-light .navbar-end .navbar-link.is-active{background-color:#dde4e6;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-light .navbar-end .navbar-link::after{border-color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-light .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-light .navbar-item.has-dropdown.is-active .navbar-link{background-color:#dde4e6;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .navbar.is-light .navbar-dropdown a.navbar-item.is-active{background-color:#ecf0f1;color:rgba(0,0,0,0.7)}}html.theme--documenter-dark .navbar.is-dark,html.theme--documenter-dark .content kbd.navbar{background-color:#282f2f;color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-brand>.navbar-item,html.theme--documenter-dark .content kbd.navbar .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-dark .navbar-brand .navbar-link,html.theme--documenter-dark .content kbd.navbar .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .content kbd.navbar .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-dark .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .content kbd.navbar .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-dark .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .content kbd.navbar .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-dark .navbar-brand .navbar-link:focus,html.theme--documenter-dark .content kbd.navbar .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-dark .navbar-brand .navbar-link:hover,html.theme--documenter-dark .content kbd.navbar .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-dark .navbar-brand .navbar-link.is-active,html.theme--documenter-dark .content kbd.navbar .navbar-brand .navbar-link.is-active{background-color:#1d2122;color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-brand .navbar-link::after,html.theme--documenter-dark .content kbd.navbar .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-burger,html.theme--documenter-dark .content kbd.navbar .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-dark .navbar-start>.navbar-item,html.theme--documenter-dark .content kbd.navbar .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-dark .navbar-start .navbar-link,html.theme--documenter-dark .content kbd.navbar .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-dark .navbar-end>.navbar-item,html.theme--documenter-dark .content kbd.navbar .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-dark .navbar-end .navbar-link,html.theme--documenter-dark .content kbd.navbar .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .content kbd.navbar .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-dark .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .content kbd.navbar .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-dark .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .content kbd.navbar .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-dark .navbar-start .navbar-link:focus,html.theme--documenter-dark .content kbd.navbar .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-dark .navbar-start .navbar-link:hover,html.theme--documenter-dark .content kbd.navbar .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-dark .navbar-start .navbar-link.is-active,html.theme--documenter-dark .content kbd.navbar .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-dark .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .content kbd.navbar .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-dark .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .content kbd.navbar .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-dark .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .content kbd.navbar .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-dark .navbar-end .navbar-link:focus,html.theme--documenter-dark .content kbd.navbar .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-dark .navbar-end .navbar-link:hover,html.theme--documenter-dark .content kbd.navbar .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-dark .navbar-end .navbar-link.is-active,html.theme--documenter-dark .content kbd.navbar .navbar-end .navbar-link.is-active{background-color:#1d2122;color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-start .navbar-link::after,html.theme--documenter-dark .content kbd.navbar .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-dark .navbar-end .navbar-link::after,html.theme--documenter-dark .content kbd.navbar .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .content kbd.navbar .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-dark .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .content kbd.navbar .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-dark .navbar-item.has-dropdown.is-active .navbar-link,html.theme--documenter-dark .content kbd.navbar .navbar-item.has-dropdown.is-active .navbar-link{background-color:#1d2122;color:#fff}html.theme--documenter-dark .navbar.is-dark .navbar-dropdown a.navbar-item.is-active,html.theme--documenter-dark .content kbd.navbar .navbar-dropdown a.navbar-item.is-active{background-color:#282f2f;color:#fff}}html.theme--documenter-dark .navbar.is-primary,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink{background-color:#375a7f;color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-brand>.navbar-item,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-primary .navbar-brand .navbar-link,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-primary .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-primary .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-primary .navbar-brand .navbar-link:focus,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-primary .navbar-brand .navbar-link:hover,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-primary .navbar-brand .navbar-link.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link.is-active{background-color:#2f4d6d;color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-brand .navbar-link::after,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-burger,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-primary .navbar-start>.navbar-item,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-primary .navbar-start .navbar-link,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-primary .navbar-end>.navbar-item,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-primary .navbar-end .navbar-link,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-primary .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-primary .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-primary .navbar-start .navbar-link:focus,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-primary .navbar-start .navbar-link:hover,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-primary .navbar-start .navbar-link.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-primary .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-primary .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-primary .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-primary .navbar-end .navbar-link:focus,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-primary .navbar-end .navbar-link:hover,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-primary .navbar-end .navbar-link.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link.is-active{background-color:#2f4d6d;color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-start .navbar-link::after,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-primary .navbar-end .navbar-link::after,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-primary .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-primary .navbar-item.has-dropdown.is-active .navbar-link,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-item.has-dropdown.is-active .navbar-link{background-color:#2f4d6d;color:#fff}html.theme--documenter-dark .navbar.is-primary .navbar-dropdown a.navbar-item.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-dropdown a.navbar-item.is-active{background-color:#375a7f;color:#fff}}html.theme--documenter-dark .navbar.is-link{background-color:#1abc9c;color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-link .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-link .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-link .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-link .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-link .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-link .navbar-brand .navbar-link.is-active{background-color:#17a689;color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-link .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-link .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-link .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-link .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-link .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-link .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-link .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-link .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-link .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-link .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-link .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-link .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-link .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-link .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-link .navbar-end .navbar-link.is-active{background-color:#17a689;color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-link .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-link .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-link .navbar-item.has-dropdown.is-active .navbar-link{background-color:#17a689;color:#fff}html.theme--documenter-dark .navbar.is-link .navbar-dropdown a.navbar-item.is-active{background-color:#1abc9c;color:#fff}}html.theme--documenter-dark .navbar.is-info{background-color:#024c7d;color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-info .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-info .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-info .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-info .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-info .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-info .navbar-brand .navbar-link.is-active{background-color:#023d64;color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-info .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-info .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-info .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-info .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-info .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-info .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-info .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-info .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-info .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-info .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-info .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-info .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-info .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-info .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-info .navbar-end .navbar-link.is-active{background-color:#023d64;color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-info .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-info .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-info .navbar-item.has-dropdown.is-active .navbar-link{background-color:#023d64;color:#fff}html.theme--documenter-dark .navbar.is-info .navbar-dropdown a.navbar-item.is-active{background-color:#024c7d;color:#fff}}html.theme--documenter-dark .navbar.is-success{background-color:#008438;color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-success .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-success .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-success .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-success .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-success .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-success .navbar-brand .navbar-link.is-active{background-color:#006b2d;color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-success .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-success .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-success .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-success .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-success .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-success .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-success .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-success .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-success .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-success .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-success .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-success .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-success .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-success .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-success .navbar-end .navbar-link.is-active{background-color:#006b2d;color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-success .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-success .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-success .navbar-item.has-dropdown.is-active .navbar-link{background-color:#006b2d;color:#fff}html.theme--documenter-dark .navbar.is-success .navbar-dropdown a.navbar-item.is-active{background-color:#008438;color:#fff}}html.theme--documenter-dark .navbar.is-warning{background-color:#ad8100;color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-warning .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-warning .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-warning .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-warning .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-warning .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-warning .navbar-brand .navbar-link.is-active{background-color:#946e00;color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-warning .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-warning .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-warning .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-warning .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-warning .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-warning .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-warning .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-warning .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-warning .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-warning .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-warning .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-warning .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-warning .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-warning .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-warning .navbar-end .navbar-link.is-active{background-color:#946e00;color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-warning .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-warning .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-warning .navbar-item.has-dropdown.is-active .navbar-link{background-color:#946e00;color:#fff}html.theme--documenter-dark .navbar.is-warning .navbar-dropdown a.navbar-item.is-active{background-color:#ad8100;color:#fff}}html.theme--documenter-dark .navbar.is-danger{background-color:#9e1b0d;color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-brand>.navbar-item,html.theme--documenter-dark .navbar.is-danger .navbar-brand .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-brand>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-danger .navbar-brand>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-danger .navbar-brand>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-danger .navbar-brand .navbar-link:focus,html.theme--documenter-dark .navbar.is-danger .navbar-brand .navbar-link:hover,html.theme--documenter-dark .navbar.is-danger .navbar-brand .navbar-link.is-active{background-color:#86170b;color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-brand .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-burger{color:#fff}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar.is-danger .navbar-start>.navbar-item,html.theme--documenter-dark .navbar.is-danger .navbar-start .navbar-link,html.theme--documenter-dark .navbar.is-danger .navbar-end>.navbar-item,html.theme--documenter-dark .navbar.is-danger .navbar-end .navbar-link{color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-start>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-danger .navbar-start>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-danger .navbar-start>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-danger .navbar-start .navbar-link:focus,html.theme--documenter-dark .navbar.is-danger .navbar-start .navbar-link:hover,html.theme--documenter-dark .navbar.is-danger .navbar-start .navbar-link.is-active,html.theme--documenter-dark .navbar.is-danger .navbar-end>a.navbar-item:focus,html.theme--documenter-dark .navbar.is-danger .navbar-end>a.navbar-item:hover,html.theme--documenter-dark .navbar.is-danger .navbar-end>a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-danger .navbar-end .navbar-link:focus,html.theme--documenter-dark .navbar.is-danger .navbar-end .navbar-link:hover,html.theme--documenter-dark .navbar.is-danger .navbar-end .navbar-link.is-active{background-color:#86170b;color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-start .navbar-link::after,html.theme--documenter-dark .navbar.is-danger .navbar-end .navbar-link::after{border-color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar.is-danger .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar.is-danger .navbar-item.has-dropdown.is-active .navbar-link{background-color:#86170b;color:#fff}html.theme--documenter-dark .navbar.is-danger .navbar-dropdown a.navbar-item.is-active{background-color:#9e1b0d;color:#fff}}html.theme--documenter-dark .navbar>.container{align-items:stretch;display:flex;min-height:4rem;width:100%}html.theme--documenter-dark .navbar.has-shadow{box-shadow:0 2px 0 0 #282f2f}html.theme--documenter-dark .navbar.is-fixed-bottom,html.theme--documenter-dark .navbar.is-fixed-top{left:0;position:fixed;right:0;z-index:30}html.theme--documenter-dark .navbar.is-fixed-bottom{bottom:0}html.theme--documenter-dark .navbar.is-fixed-bottom.has-shadow{box-shadow:0 -2px 0 0 #282f2f}html.theme--documenter-dark .navbar.is-fixed-top{top:0}html.theme--documenter-dark html.has-navbar-fixed-top,html.theme--documenter-dark body.has-navbar-fixed-top{padding-top:4rem}html.theme--documenter-dark html.has-navbar-fixed-bottom,html.theme--documenter-dark body.has-navbar-fixed-bottom{padding-bottom:4rem}html.theme--documenter-dark .navbar-brand,html.theme--documenter-dark .navbar-tabs{align-items:stretch;display:flex;flex-shrink:0;min-height:4rem}html.theme--documenter-dark .navbar-brand a.navbar-item:focus,html.theme--documenter-dark .navbar-brand a.navbar-item:hover{background-color:transparent}html.theme--documenter-dark .navbar-tabs{-webkit-overflow-scrolling:touch;max-width:100vw;overflow-x:auto;overflow-y:hidden}html.theme--documenter-dark .navbar-burger{color:#fff;-moz-appearance:none;-webkit-appearance:none;appearance:none;background:none;border:none;cursor:pointer;display:block;height:4rem;position:relative;width:4rem;margin-left:auto}html.theme--documenter-dark .navbar-burger span{background-color:currentColor;display:block;height:1px;left:calc(50% - 8px);position:absolute;transform-origin:center;transition-duration:86ms;transition-property:background-color, opacity, transform;transition-timing-function:ease-out;width:16px}html.theme--documenter-dark .navbar-burger span:nth-child(1){top:calc(50% - 6px)}html.theme--documenter-dark .navbar-burger span:nth-child(2){top:calc(50% - 1px)}html.theme--documenter-dark .navbar-burger span:nth-child(3){top:calc(50% + 4px)}html.theme--documenter-dark .navbar-burger:hover{background-color:rgba(0,0,0,0.05)}html.theme--documenter-dark .navbar-burger.is-active span:nth-child(1){transform:translateY(5px) rotate(45deg)}html.theme--documenter-dark .navbar-burger.is-active span:nth-child(2){opacity:0}html.theme--documenter-dark .navbar-burger.is-active span:nth-child(3){transform:translateY(-5px) rotate(-45deg)}html.theme--documenter-dark .navbar-menu{display:none}html.theme--documenter-dark .navbar-item,html.theme--documenter-dark .navbar-link{color:#fff;display:block;line-height:1.5;padding:0.5rem 0.75rem;position:relative}html.theme--documenter-dark .navbar-item .icon:only-child,html.theme--documenter-dark .navbar-link .icon:only-child{margin-left:-0.25rem;margin-right:-0.25rem}html.theme--documenter-dark a.navbar-item,html.theme--documenter-dark .navbar-link{cursor:pointer}html.theme--documenter-dark a.navbar-item:focus,html.theme--documenter-dark a.navbar-item:focus-within,html.theme--documenter-dark a.navbar-item:hover,html.theme--documenter-dark a.navbar-item.is-active,html.theme--documenter-dark .navbar-link:focus,html.theme--documenter-dark .navbar-link:focus-within,html.theme--documenter-dark .navbar-link:hover,html.theme--documenter-dark .navbar-link.is-active{background-color:rgba(0,0,0,0);color:#1abc9c}html.theme--documenter-dark .navbar-item{flex-grow:0;flex-shrink:0}html.theme--documenter-dark .navbar-item img{max-height:1.75rem}html.theme--documenter-dark .navbar-item.has-dropdown{padding:0}html.theme--documenter-dark .navbar-item.is-expanded{flex-grow:1;flex-shrink:1}html.theme--documenter-dark .navbar-item.is-tab{border-bottom:1px solid transparent;min-height:4rem;padding-bottom:calc(0.5rem - 1px)}html.theme--documenter-dark .navbar-item.is-tab:focus,html.theme--documenter-dark .navbar-item.is-tab:hover{background-color:rgba(0,0,0,0);border-bottom-color:#1abc9c}html.theme--documenter-dark .navbar-item.is-tab.is-active{background-color:rgba(0,0,0,0);border-bottom-color:#1abc9c;border-bottom-style:solid;border-bottom-width:3px;color:#1abc9c;padding-bottom:calc(0.5rem - 3px)}html.theme--documenter-dark .navbar-content{flex-grow:1;flex-shrink:1}html.theme--documenter-dark .navbar-link:not(.is-arrowless){padding-right:2.5em}html.theme--documenter-dark .navbar-link:not(.is-arrowless)::after{border-color:#fff;margin-top:-0.375em;right:1.125em}html.theme--documenter-dark .navbar-dropdown{font-size:0.875rem;padding-bottom:0.5rem;padding-top:0.5rem}html.theme--documenter-dark .navbar-dropdown .navbar-item{padding-left:1.5rem;padding-right:1.5rem}html.theme--documenter-dark .navbar-divider{background-color:rgba(0,0,0,0.2);border:none;display:none;height:2px;margin:0.5rem 0}@media screen and (max-width: 1055px){html.theme--documenter-dark .navbar>.container{display:block}html.theme--documenter-dark .navbar-brand .navbar-item,html.theme--documenter-dark .navbar-tabs .navbar-item{align-items:center;display:flex}html.theme--documenter-dark .navbar-link::after{display:none}html.theme--documenter-dark .navbar-menu{background-color:#375a7f;box-shadow:0 8px 16px rgba(10,10,10,0.1);padding:0.5rem 0}html.theme--documenter-dark .navbar-menu.is-active{display:block}html.theme--documenter-dark .navbar.is-fixed-bottom-touch,html.theme--documenter-dark .navbar.is-fixed-top-touch{left:0;position:fixed;right:0;z-index:30}html.theme--documenter-dark .navbar.is-fixed-bottom-touch{bottom:0}html.theme--documenter-dark .navbar.is-fixed-bottom-touch.has-shadow{box-shadow:0 -2px 3px rgba(10,10,10,0.1)}html.theme--documenter-dark .navbar.is-fixed-top-touch{top:0}html.theme--documenter-dark .navbar.is-fixed-top .navbar-menu,html.theme--documenter-dark .navbar.is-fixed-top-touch .navbar-menu{-webkit-overflow-scrolling:touch;max-height:calc(100vh - 4rem);overflow:auto}html.theme--documenter-dark html.has-navbar-fixed-top-touch,html.theme--documenter-dark body.has-navbar-fixed-top-touch{padding-top:4rem}html.theme--documenter-dark html.has-navbar-fixed-bottom-touch,html.theme--documenter-dark body.has-navbar-fixed-bottom-touch{padding-bottom:4rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .navbar,html.theme--documenter-dark .navbar-menu,html.theme--documenter-dark .navbar-start,html.theme--documenter-dark .navbar-end{align-items:stretch;display:flex}html.theme--documenter-dark .navbar{min-height:4rem}html.theme--documenter-dark .navbar.is-spaced{padding:1rem 2rem}html.theme--documenter-dark .navbar.is-spaced .navbar-start,html.theme--documenter-dark .navbar.is-spaced .navbar-end{align-items:center}html.theme--documenter-dark .navbar.is-spaced a.navbar-item,html.theme--documenter-dark .navbar.is-spaced .navbar-link{border-radius:.4em}html.theme--documenter-dark .navbar.is-transparent a.navbar-item:focus,html.theme--documenter-dark .navbar.is-transparent a.navbar-item:hover,html.theme--documenter-dark .navbar.is-transparent a.navbar-item.is-active,html.theme--documenter-dark .navbar.is-transparent .navbar-link:focus,html.theme--documenter-dark .navbar.is-transparent .navbar-link:hover,html.theme--documenter-dark .navbar.is-transparent .navbar-link.is-active{background-color:transparent !important}html.theme--documenter-dark .navbar.is-transparent .navbar-item.has-dropdown.is-active .navbar-link,html.theme--documenter-dark .navbar.is-transparent .navbar-item.has-dropdown.is-hoverable:focus .navbar-link,html.theme--documenter-dark .navbar.is-transparent .navbar-item.has-dropdown.is-hoverable:focus-within .navbar-link,html.theme--documenter-dark .navbar.is-transparent .navbar-item.has-dropdown.is-hoverable:hover .navbar-link{background-color:transparent !important}html.theme--documenter-dark .navbar.is-transparent .navbar-dropdown a.navbar-item:focus,html.theme--documenter-dark .navbar.is-transparent .navbar-dropdown a.navbar-item:hover{background-color:rgba(0,0,0,0);color:#dbdee0}html.theme--documenter-dark .navbar.is-transparent .navbar-dropdown a.navbar-item.is-active{background-color:rgba(0,0,0,0);color:#1abc9c}html.theme--documenter-dark .navbar-burger{display:none}html.theme--documenter-dark .navbar-item,html.theme--documenter-dark .navbar-link{align-items:center;display:flex}html.theme--documenter-dark .navbar-item.has-dropdown{align-items:stretch}html.theme--documenter-dark .navbar-item.has-dropdown-up .navbar-link::after{transform:rotate(135deg) translate(0.25em, -0.25em)}html.theme--documenter-dark .navbar-item.has-dropdown-up .navbar-dropdown{border-bottom:1px solid rgba(0,0,0,0.2);border-radius:8px 8px 0 0;border-top:none;bottom:100%;box-shadow:0 -8px 8px rgba(10,10,10,0.1);top:auto}html.theme--documenter-dark .navbar-item.is-active .navbar-dropdown,html.theme--documenter-dark .navbar-item.is-hoverable:focus .navbar-dropdown,html.theme--documenter-dark .navbar-item.is-hoverable:focus-within .navbar-dropdown,html.theme--documenter-dark .navbar-item.is-hoverable:hover .navbar-dropdown{display:block}.navbar.is-spaced html.theme--documenter-dark .navbar-item.is-active .navbar-dropdown,html.theme--documenter-dark .navbar-item.is-active .navbar-dropdown.is-boxed,.navbar.is-spaced html.theme--documenter-dark .navbar-item.is-hoverable:focus .navbar-dropdown,html.theme--documenter-dark .navbar-item.is-hoverable:focus .navbar-dropdown.is-boxed,.navbar.is-spaced html.theme--documenter-dark .navbar-item.is-hoverable:focus-within .navbar-dropdown,html.theme--documenter-dark .navbar-item.is-hoverable:focus-within .navbar-dropdown.is-boxed,.navbar.is-spaced html.theme--documenter-dark .navbar-item.is-hoverable:hover .navbar-dropdown,html.theme--documenter-dark .navbar-item.is-hoverable:hover .navbar-dropdown.is-boxed{opacity:1;pointer-events:auto;transform:translateY(0)}html.theme--documenter-dark .navbar-menu{flex-grow:1;flex-shrink:0}html.theme--documenter-dark .navbar-start{justify-content:flex-start;margin-right:auto}html.theme--documenter-dark .navbar-end{justify-content:flex-end;margin-left:auto}html.theme--documenter-dark .navbar-dropdown{background-color:#375a7f;border-bottom-left-radius:8px;border-bottom-right-radius:8px;border-top:1px solid rgba(0,0,0,0.2);box-shadow:0 8px 8px rgba(10,10,10,0.1);display:none;font-size:0.875rem;left:0;min-width:100%;position:absolute;top:100%;z-index:20}html.theme--documenter-dark .navbar-dropdown .navbar-item{padding:0.375rem 1rem;white-space:nowrap}html.theme--documenter-dark .navbar-dropdown a.navbar-item{padding-right:3rem}html.theme--documenter-dark .navbar-dropdown a.navbar-item:focus,html.theme--documenter-dark .navbar-dropdown a.navbar-item:hover{background-color:rgba(0,0,0,0);color:#dbdee0}html.theme--documenter-dark .navbar-dropdown a.navbar-item.is-active{background-color:rgba(0,0,0,0);color:#1abc9c}.navbar.is-spaced html.theme--documenter-dark .navbar-dropdown,html.theme--documenter-dark .navbar-dropdown.is-boxed{border-radius:8px;border-top:none;box-shadow:0 8px 8px rgba(10,10,10,0.1), 0 0 0 1px rgba(10,10,10,0.1);display:block;opacity:0;pointer-events:none;top:calc(100% + (-4px));transform:translateY(-5px);transition-duration:86ms;transition-property:opacity, transform}html.theme--documenter-dark .navbar-dropdown.is-right{left:auto;right:0}html.theme--documenter-dark .navbar-divider{display:block}html.theme--documenter-dark .navbar>.container .navbar-brand,html.theme--documenter-dark .container>.navbar .navbar-brand{margin-left:-.75rem}html.theme--documenter-dark .navbar>.container .navbar-menu,html.theme--documenter-dark .container>.navbar .navbar-menu{margin-right:-.75rem}html.theme--documenter-dark .navbar.is-fixed-bottom-desktop,html.theme--documenter-dark .navbar.is-fixed-top-desktop{left:0;position:fixed;right:0;z-index:30}html.theme--documenter-dark .navbar.is-fixed-bottom-desktop{bottom:0}html.theme--documenter-dark .navbar.is-fixed-bottom-desktop.has-shadow{box-shadow:0 -2px 3px rgba(10,10,10,0.1)}html.theme--documenter-dark .navbar.is-fixed-top-desktop{top:0}html.theme--documenter-dark html.has-navbar-fixed-top-desktop,html.theme--documenter-dark body.has-navbar-fixed-top-desktop{padding-top:4rem}html.theme--documenter-dark html.has-navbar-fixed-bottom-desktop,html.theme--documenter-dark body.has-navbar-fixed-bottom-desktop{padding-bottom:4rem}html.theme--documenter-dark html.has-spaced-navbar-fixed-top,html.theme--documenter-dark body.has-spaced-navbar-fixed-top{padding-top:6rem}html.theme--documenter-dark html.has-spaced-navbar-fixed-bottom,html.theme--documenter-dark body.has-spaced-navbar-fixed-bottom{padding-bottom:6rem}html.theme--documenter-dark a.navbar-item.is-active,html.theme--documenter-dark .navbar-link.is-active{color:#1abc9c}html.theme--documenter-dark a.navbar-item.is-active:not(:focus):not(:hover),html.theme--documenter-dark .navbar-link.is-active:not(:focus):not(:hover){background-color:rgba(0,0,0,0)}html.theme--documenter-dark .navbar-item.has-dropdown:focus .navbar-link,html.theme--documenter-dark .navbar-item.has-dropdown:hover .navbar-link,html.theme--documenter-dark .navbar-item.has-dropdown.is-active .navbar-link{background-color:rgba(0,0,0,0)}}html.theme--documenter-dark .hero.is-fullheight-with-navbar{min-height:calc(100vh - 4rem)}html.theme--documenter-dark .pagination{font-size:1rem;margin:-.25rem}html.theme--documenter-dark .pagination.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.pagination{font-size:.75rem}html.theme--documenter-dark .pagination.is-medium{font-size:1.25rem}html.theme--documenter-dark .pagination.is-large{font-size:1.5rem}html.theme--documenter-dark .pagination.is-rounded .pagination-previous,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.pagination .pagination-previous,html.theme--documenter-dark .pagination.is-rounded .pagination-next,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.pagination .pagination-next{padding-left:1em;padding-right:1em;border-radius:9999px}html.theme--documenter-dark .pagination.is-rounded .pagination-link,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.pagination .pagination-link{border-radius:9999px}html.theme--documenter-dark .pagination,html.theme--documenter-dark .pagination-list{align-items:center;display:flex;justify-content:center;text-align:center}html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .pagination-next,html.theme--documenter-dark .pagination-link,html.theme--documenter-dark .pagination-ellipsis{font-size:1em;justify-content:center;margin:.25rem;padding-left:.5em;padding-right:.5em;text-align:center}html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .pagination-next,html.theme--documenter-dark .pagination-link{border-color:#5e6d6f;color:#1abc9c;min-width:2.5em}html.theme--documenter-dark .pagination-previous:hover,html.theme--documenter-dark .pagination-next:hover,html.theme--documenter-dark .pagination-link:hover{border-color:#8c9b9d;color:#1dd2af}html.theme--documenter-dark .pagination-previous:focus,html.theme--documenter-dark .pagination-next:focus,html.theme--documenter-dark .pagination-link:focus{border-color:#8c9b9d}html.theme--documenter-dark .pagination-previous:active,html.theme--documenter-dark .pagination-next:active,html.theme--documenter-dark .pagination-link:active{box-shadow:inset 0 1px 2px rgba(10,10,10,0.2)}html.theme--documenter-dark .pagination-previous[disabled],html.theme--documenter-dark .pagination-previous.is-disabled,html.theme--documenter-dark .pagination-next[disabled],html.theme--documenter-dark .pagination-next.is-disabled,html.theme--documenter-dark .pagination-link[disabled],html.theme--documenter-dark .pagination-link.is-disabled{background-color:#5e6d6f;border-color:#5e6d6f;box-shadow:none;color:#fff;opacity:0.5}html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .pagination-next{padding-left:.75em;padding-right:.75em;white-space:nowrap}html.theme--documenter-dark .pagination-link.is-current{background-color:#1abc9c;border-color:#1abc9c;color:#fff}html.theme--documenter-dark .pagination-ellipsis{color:#8c9b9d;pointer-events:none}html.theme--documenter-dark .pagination-list{flex-wrap:wrap}html.theme--documenter-dark .pagination-list li{list-style:none}@media screen and (max-width: 768px){html.theme--documenter-dark .pagination{flex-wrap:wrap}html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .pagination-next{flex-grow:1;flex-shrink:1}html.theme--documenter-dark .pagination-list li{flex-grow:1;flex-shrink:1}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .pagination-list{flex-grow:1;flex-shrink:1;justify-content:flex-start;order:1}html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .pagination-next,html.theme--documenter-dark .pagination-link,html.theme--documenter-dark .pagination-ellipsis{margin-bottom:0;margin-top:0}html.theme--documenter-dark .pagination-previous{order:2}html.theme--documenter-dark .pagination-next{order:3}html.theme--documenter-dark .pagination{justify-content:space-between;margin-bottom:0;margin-top:0}html.theme--documenter-dark .pagination.is-centered .pagination-previous{order:1}html.theme--documenter-dark .pagination.is-centered .pagination-list{justify-content:center;order:2}html.theme--documenter-dark .pagination.is-centered .pagination-next{order:3}html.theme--documenter-dark .pagination.is-right .pagination-previous{order:1}html.theme--documenter-dark .pagination.is-right .pagination-next{order:2}html.theme--documenter-dark .pagination.is-right .pagination-list{justify-content:flex-end;order:3}}html.theme--documenter-dark .panel{border-radius:8px;box-shadow:#171717;font-size:1rem}html.theme--documenter-dark .panel:not(:last-child){margin-bottom:1.5rem}html.theme--documenter-dark .panel.is-white .panel-heading{background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .panel.is-white .panel-tabs a.is-active{border-bottom-color:#fff}html.theme--documenter-dark .panel.is-white .panel-block.is-active .panel-icon{color:#fff}html.theme--documenter-dark .panel.is-black .panel-heading{background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .panel.is-black .panel-tabs a.is-active{border-bottom-color:#0a0a0a}html.theme--documenter-dark .panel.is-black .panel-block.is-active .panel-icon{color:#0a0a0a}html.theme--documenter-dark .panel.is-light .panel-heading{background-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .panel.is-light .panel-tabs a.is-active{border-bottom-color:#ecf0f1}html.theme--documenter-dark .panel.is-light .panel-block.is-active .panel-icon{color:#ecf0f1}html.theme--documenter-dark .panel.is-dark .panel-heading,html.theme--documenter-dark .content kbd.panel .panel-heading{background-color:#282f2f;color:#fff}html.theme--documenter-dark .panel.is-dark .panel-tabs a.is-active,html.theme--documenter-dark .content kbd.panel .panel-tabs a.is-active{border-bottom-color:#282f2f}html.theme--documenter-dark .panel.is-dark .panel-block.is-active .panel-icon,html.theme--documenter-dark .content kbd.panel .panel-block.is-active .panel-icon{color:#282f2f}html.theme--documenter-dark .panel.is-primary .panel-heading,html.theme--documenter-dark .docstring>section>a.panel.docs-sourcelink .panel-heading{background-color:#375a7f;color:#fff}html.theme--documenter-dark .panel.is-primary .panel-tabs a.is-active,html.theme--documenter-dark .docstring>section>a.panel.docs-sourcelink .panel-tabs a.is-active{border-bottom-color:#375a7f}html.theme--documenter-dark .panel.is-primary .panel-block.is-active .panel-icon,html.theme--documenter-dark .docstring>section>a.panel.docs-sourcelink .panel-block.is-active .panel-icon{color:#375a7f}html.theme--documenter-dark .panel.is-link .panel-heading{background-color:#1abc9c;color:#fff}html.theme--documenter-dark .panel.is-link .panel-tabs a.is-active{border-bottom-color:#1abc9c}html.theme--documenter-dark .panel.is-link .panel-block.is-active .panel-icon{color:#1abc9c}html.theme--documenter-dark .panel.is-info .panel-heading{background-color:#024c7d;color:#fff}html.theme--documenter-dark .panel.is-info .panel-tabs a.is-active{border-bottom-color:#024c7d}html.theme--documenter-dark .panel.is-info .panel-block.is-active .panel-icon{color:#024c7d}html.theme--documenter-dark .panel.is-success .panel-heading{background-color:#008438;color:#fff}html.theme--documenter-dark .panel.is-success .panel-tabs a.is-active{border-bottom-color:#008438}html.theme--documenter-dark .panel.is-success .panel-block.is-active .panel-icon{color:#008438}html.theme--documenter-dark .panel.is-warning .panel-heading{background-color:#ad8100;color:#fff}html.theme--documenter-dark .panel.is-warning .panel-tabs a.is-active{border-bottom-color:#ad8100}html.theme--documenter-dark .panel.is-warning .panel-block.is-active .panel-icon{color:#ad8100}html.theme--documenter-dark .panel.is-danger .panel-heading{background-color:#9e1b0d;color:#fff}html.theme--documenter-dark .panel.is-danger .panel-tabs a.is-active{border-bottom-color:#9e1b0d}html.theme--documenter-dark .panel.is-danger .panel-block.is-active .panel-icon{color:#9e1b0d}html.theme--documenter-dark .panel-tabs:not(:last-child),html.theme--documenter-dark .panel-block:not(:last-child){border-bottom:1px solid #ededed}html.theme--documenter-dark .panel-heading{background-color:#343c3d;border-radius:8px 8px 0 0;color:#f2f2f2;font-size:1.25em;font-weight:700;line-height:1.25;padding:0.75em 1em}html.theme--documenter-dark .panel-tabs{align-items:flex-end;display:flex;font-size:.875em;justify-content:center}html.theme--documenter-dark .panel-tabs a{border-bottom:1px solid #5e6d6f;margin-bottom:-1px;padding:0.5em}html.theme--documenter-dark .panel-tabs a.is-active{border-bottom-color:#343c3d;color:#17a689}html.theme--documenter-dark .panel-list a{color:#fff}html.theme--documenter-dark .panel-list a:hover{color:#1abc9c}html.theme--documenter-dark .panel-block{align-items:center;color:#f2f2f2;display:flex;justify-content:flex-start;padding:0.5em 0.75em}html.theme--documenter-dark .panel-block input[type="checkbox"]{margin-right:.75em}html.theme--documenter-dark .panel-block>.control{flex-grow:1;flex-shrink:1;width:100%}html.theme--documenter-dark .panel-block.is-wrapped{flex-wrap:wrap}html.theme--documenter-dark .panel-block.is-active{border-left-color:#1abc9c;color:#17a689}html.theme--documenter-dark .panel-block.is-active .panel-icon{color:#1abc9c}html.theme--documenter-dark .panel-block:last-child{border-bottom-left-radius:8px;border-bottom-right-radius:8px}html.theme--documenter-dark a.panel-block,html.theme--documenter-dark label.panel-block{cursor:pointer}html.theme--documenter-dark a.panel-block:hover,html.theme--documenter-dark label.panel-block:hover{background-color:#282f2f}html.theme--documenter-dark .panel-icon{display:inline-block;font-size:14px;height:1em;line-height:1em;text-align:center;vertical-align:top;width:1em;color:#fff;margin-right:.75em}html.theme--documenter-dark .panel-icon .fa{font-size:inherit;line-height:inherit}html.theme--documenter-dark .tabs{-webkit-overflow-scrolling:touch;align-items:stretch;display:flex;font-size:1rem;justify-content:space-between;overflow:hidden;overflow-x:auto;white-space:nowrap}html.theme--documenter-dark .tabs a{align-items:center;border-bottom-color:#5e6d6f;border-bottom-style:solid;border-bottom-width:1px;color:#fff;display:flex;justify-content:center;margin-bottom:-1px;padding:0.5em 1em;vertical-align:top}html.theme--documenter-dark .tabs a:hover{border-bottom-color:#f2f2f2;color:#f2f2f2}html.theme--documenter-dark .tabs li{display:block}html.theme--documenter-dark .tabs li.is-active a{border-bottom-color:#1abc9c;color:#1abc9c}html.theme--documenter-dark .tabs ul{align-items:center;border-bottom-color:#5e6d6f;border-bottom-style:solid;border-bottom-width:1px;display:flex;flex-grow:1;flex-shrink:0;justify-content:flex-start}html.theme--documenter-dark .tabs ul.is-left{padding-right:0.75em}html.theme--documenter-dark .tabs ul.is-center{flex:none;justify-content:center;padding-left:0.75em;padding-right:0.75em}html.theme--documenter-dark .tabs ul.is-right{justify-content:flex-end;padding-left:0.75em}html.theme--documenter-dark .tabs .icon:first-child{margin-right:.5em}html.theme--documenter-dark .tabs .icon:last-child{margin-left:.5em}html.theme--documenter-dark .tabs.is-centered ul{justify-content:center}html.theme--documenter-dark .tabs.is-right ul{justify-content:flex-end}html.theme--documenter-dark .tabs.is-boxed a{border:1px solid transparent;border-radius:.4em .4em 0 0}html.theme--documenter-dark .tabs.is-boxed a:hover{background-color:#282f2f;border-bottom-color:#5e6d6f}html.theme--documenter-dark .tabs.is-boxed li.is-active a{background-color:#fff;border-color:#5e6d6f;border-bottom-color:rgba(0,0,0,0) !important}html.theme--documenter-dark .tabs.is-fullwidth li{flex-grow:1;flex-shrink:0}html.theme--documenter-dark .tabs.is-toggle a{border-color:#5e6d6f;border-style:solid;border-width:1px;margin-bottom:0;position:relative}html.theme--documenter-dark .tabs.is-toggle a:hover{background-color:#282f2f;border-color:#8c9b9d;z-index:2}html.theme--documenter-dark .tabs.is-toggle li+li{margin-left:-1px}html.theme--documenter-dark .tabs.is-toggle li:first-child a{border-top-left-radius:.4em;border-bottom-left-radius:.4em}html.theme--documenter-dark .tabs.is-toggle li:last-child a{border-top-right-radius:.4em;border-bottom-right-radius:.4em}html.theme--documenter-dark .tabs.is-toggle li.is-active a{background-color:#1abc9c;border-color:#1abc9c;color:#fff;z-index:1}html.theme--documenter-dark .tabs.is-toggle ul{border-bottom:none}html.theme--documenter-dark .tabs.is-toggle.is-toggle-rounded li:first-child a{border-bottom-left-radius:9999px;border-top-left-radius:9999px;padding-left:1.25em}html.theme--documenter-dark .tabs.is-toggle.is-toggle-rounded li:last-child a{border-bottom-right-radius:9999px;border-top-right-radius:9999px;padding-right:1.25em}html.theme--documenter-dark .tabs.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.tabs{font-size:.75rem}html.theme--documenter-dark .tabs.is-medium{font-size:1.25rem}html.theme--documenter-dark .tabs.is-large{font-size:1.5rem}html.theme--documenter-dark .column{display:block;flex-basis:0;flex-grow:1;flex-shrink:1;padding:.75rem}.columns.is-mobile>html.theme--documenter-dark .column.is-narrow{flex:none;width:unset}.columns.is-mobile>html.theme--documenter-dark .column.is-full{flex:none;width:100%}.columns.is-mobile>html.theme--documenter-dark .column.is-three-quarters{flex:none;width:75%}.columns.is-mobile>html.theme--documenter-dark .column.is-two-thirds{flex:none;width:66.6666%}.columns.is-mobile>html.theme--documenter-dark .column.is-half{flex:none;width:50%}.columns.is-mobile>html.theme--documenter-dark .column.is-one-third{flex:none;width:33.3333%}.columns.is-mobile>html.theme--documenter-dark .column.is-one-quarter{flex:none;width:25%}.columns.is-mobile>html.theme--documenter-dark .column.is-one-fifth{flex:none;width:20%}.columns.is-mobile>html.theme--documenter-dark .column.is-two-fifths{flex:none;width:40%}.columns.is-mobile>html.theme--documenter-dark .column.is-three-fifths{flex:none;width:60%}.columns.is-mobile>html.theme--documenter-dark .column.is-four-fifths{flex:none;width:80%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-three-quarters{margin-left:75%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-two-thirds{margin-left:66.6666%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-half{margin-left:50%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-one-third{margin-left:33.3333%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-one-quarter{margin-left:25%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-one-fifth{margin-left:20%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-two-fifths{margin-left:40%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-three-fifths{margin-left:60%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-four-fifths{margin-left:80%}.columns.is-mobile>html.theme--documenter-dark .column.is-0{flex:none;width:0%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-0{margin-left:0%}.columns.is-mobile>html.theme--documenter-dark .column.is-1{flex:none;width:8.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-1{margin-left:8.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-2{flex:none;width:16.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-2{margin-left:16.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-3{flex:none;width:25%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-3{margin-left:25%}.columns.is-mobile>html.theme--documenter-dark .column.is-4{flex:none;width:33.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-4{margin-left:33.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-5{flex:none;width:41.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-5{margin-left:41.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-6{flex:none;width:50%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-6{margin-left:50%}.columns.is-mobile>html.theme--documenter-dark .column.is-7{flex:none;width:58.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-7{margin-left:58.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-8{flex:none;width:66.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-8{margin-left:66.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-9{flex:none;width:75%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-9{margin-left:75%}.columns.is-mobile>html.theme--documenter-dark .column.is-10{flex:none;width:83.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-10{margin-left:83.33333337%}.columns.is-mobile>html.theme--documenter-dark .column.is-11{flex:none;width:91.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-11{margin-left:91.66666674%}.columns.is-mobile>html.theme--documenter-dark .column.is-12{flex:none;width:100%}.columns.is-mobile>html.theme--documenter-dark .column.is-offset-12{margin-left:100%}@media screen and (max-width: 768px){html.theme--documenter-dark .column.is-narrow-mobile{flex:none;width:unset}html.theme--documenter-dark .column.is-full-mobile{flex:none;width:100%}html.theme--documenter-dark .column.is-three-quarters-mobile{flex:none;width:75%}html.theme--documenter-dark .column.is-two-thirds-mobile{flex:none;width:66.6666%}html.theme--documenter-dark .column.is-half-mobile{flex:none;width:50%}html.theme--documenter-dark .column.is-one-third-mobile{flex:none;width:33.3333%}html.theme--documenter-dark .column.is-one-quarter-mobile{flex:none;width:25%}html.theme--documenter-dark .column.is-one-fifth-mobile{flex:none;width:20%}html.theme--documenter-dark .column.is-two-fifths-mobile{flex:none;width:40%}html.theme--documenter-dark .column.is-three-fifths-mobile{flex:none;width:60%}html.theme--documenter-dark .column.is-four-fifths-mobile{flex:none;width:80%}html.theme--documenter-dark .column.is-offset-three-quarters-mobile{margin-left:75%}html.theme--documenter-dark .column.is-offset-two-thirds-mobile{margin-left:66.6666%}html.theme--documenter-dark .column.is-offset-half-mobile{margin-left:50%}html.theme--documenter-dark .column.is-offset-one-third-mobile{margin-left:33.3333%}html.theme--documenter-dark .column.is-offset-one-quarter-mobile{margin-left:25%}html.theme--documenter-dark .column.is-offset-one-fifth-mobile{margin-left:20%}html.theme--documenter-dark .column.is-offset-two-fifths-mobile{margin-left:40%}html.theme--documenter-dark .column.is-offset-three-fifths-mobile{margin-left:60%}html.theme--documenter-dark .column.is-offset-four-fifths-mobile{margin-left:80%}html.theme--documenter-dark .column.is-0-mobile{flex:none;width:0%}html.theme--documenter-dark .column.is-offset-0-mobile{margin-left:0%}html.theme--documenter-dark .column.is-1-mobile{flex:none;width:8.33333337%}html.theme--documenter-dark .column.is-offset-1-mobile{margin-left:8.33333337%}html.theme--documenter-dark .column.is-2-mobile{flex:none;width:16.66666674%}html.theme--documenter-dark .column.is-offset-2-mobile{margin-left:16.66666674%}html.theme--documenter-dark .column.is-3-mobile{flex:none;width:25%}html.theme--documenter-dark .column.is-offset-3-mobile{margin-left:25%}html.theme--documenter-dark .column.is-4-mobile{flex:none;width:33.33333337%}html.theme--documenter-dark .column.is-offset-4-mobile{margin-left:33.33333337%}html.theme--documenter-dark .column.is-5-mobile{flex:none;width:41.66666674%}html.theme--documenter-dark .column.is-offset-5-mobile{margin-left:41.66666674%}html.theme--documenter-dark .column.is-6-mobile{flex:none;width:50%}html.theme--documenter-dark .column.is-offset-6-mobile{margin-left:50%}html.theme--documenter-dark .column.is-7-mobile{flex:none;width:58.33333337%}html.theme--documenter-dark .column.is-offset-7-mobile{margin-left:58.33333337%}html.theme--documenter-dark .column.is-8-mobile{flex:none;width:66.66666674%}html.theme--documenter-dark .column.is-offset-8-mobile{margin-left:66.66666674%}html.theme--documenter-dark .column.is-9-mobile{flex:none;width:75%}html.theme--documenter-dark .column.is-offset-9-mobile{margin-left:75%}html.theme--documenter-dark .column.is-10-mobile{flex:none;width:83.33333337%}html.theme--documenter-dark .column.is-offset-10-mobile{margin-left:83.33333337%}html.theme--documenter-dark .column.is-11-mobile{flex:none;width:91.66666674%}html.theme--documenter-dark .column.is-offset-11-mobile{margin-left:91.66666674%}html.theme--documenter-dark .column.is-12-mobile{flex:none;width:100%}html.theme--documenter-dark .column.is-offset-12-mobile{margin-left:100%}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .column.is-narrow,html.theme--documenter-dark .column.is-narrow-tablet{flex:none;width:unset}html.theme--documenter-dark .column.is-full,html.theme--documenter-dark .column.is-full-tablet{flex:none;width:100%}html.theme--documenter-dark .column.is-three-quarters,html.theme--documenter-dark .column.is-three-quarters-tablet{flex:none;width:75%}html.theme--documenter-dark .column.is-two-thirds,html.theme--documenter-dark .column.is-two-thirds-tablet{flex:none;width:66.6666%}html.theme--documenter-dark .column.is-half,html.theme--documenter-dark .column.is-half-tablet{flex:none;width:50%}html.theme--documenter-dark .column.is-one-third,html.theme--documenter-dark .column.is-one-third-tablet{flex:none;width:33.3333%}html.theme--documenter-dark .column.is-one-quarter,html.theme--documenter-dark .column.is-one-quarter-tablet{flex:none;width:25%}html.theme--documenter-dark .column.is-one-fifth,html.theme--documenter-dark .column.is-one-fifth-tablet{flex:none;width:20%}html.theme--documenter-dark .column.is-two-fifths,html.theme--documenter-dark .column.is-two-fifths-tablet{flex:none;width:40%}html.theme--documenter-dark .column.is-three-fifths,html.theme--documenter-dark .column.is-three-fifths-tablet{flex:none;width:60%}html.theme--documenter-dark .column.is-four-fifths,html.theme--documenter-dark .column.is-four-fifths-tablet{flex:none;width:80%}html.theme--documenter-dark .column.is-offset-three-quarters,html.theme--documenter-dark .column.is-offset-three-quarters-tablet{margin-left:75%}html.theme--documenter-dark .column.is-offset-two-thirds,html.theme--documenter-dark .column.is-offset-two-thirds-tablet{margin-left:66.6666%}html.theme--documenter-dark .column.is-offset-half,html.theme--documenter-dark .column.is-offset-half-tablet{margin-left:50%}html.theme--documenter-dark .column.is-offset-one-third,html.theme--documenter-dark .column.is-offset-one-third-tablet{margin-left:33.3333%}html.theme--documenter-dark .column.is-offset-one-quarter,html.theme--documenter-dark .column.is-offset-one-quarter-tablet{margin-left:25%}html.theme--documenter-dark .column.is-offset-one-fifth,html.theme--documenter-dark .column.is-offset-one-fifth-tablet{margin-left:20%}html.theme--documenter-dark .column.is-offset-two-fifths,html.theme--documenter-dark .column.is-offset-two-fifths-tablet{margin-left:40%}html.theme--documenter-dark .column.is-offset-three-fifths,html.theme--documenter-dark .column.is-offset-three-fifths-tablet{margin-left:60%}html.theme--documenter-dark .column.is-offset-four-fifths,html.theme--documenter-dark .column.is-offset-four-fifths-tablet{margin-left:80%}html.theme--documenter-dark .column.is-0,html.theme--documenter-dark .column.is-0-tablet{flex:none;width:0%}html.theme--documenter-dark .column.is-offset-0,html.theme--documenter-dark .column.is-offset-0-tablet{margin-left:0%}html.theme--documenter-dark .column.is-1,html.theme--documenter-dark .column.is-1-tablet{flex:none;width:8.33333337%}html.theme--documenter-dark .column.is-offset-1,html.theme--documenter-dark .column.is-offset-1-tablet{margin-left:8.33333337%}html.theme--documenter-dark .column.is-2,html.theme--documenter-dark .column.is-2-tablet{flex:none;width:16.66666674%}html.theme--documenter-dark .column.is-offset-2,html.theme--documenter-dark .column.is-offset-2-tablet{margin-left:16.66666674%}html.theme--documenter-dark .column.is-3,html.theme--documenter-dark .column.is-3-tablet{flex:none;width:25%}html.theme--documenter-dark .column.is-offset-3,html.theme--documenter-dark .column.is-offset-3-tablet{margin-left:25%}html.theme--documenter-dark .column.is-4,html.theme--documenter-dark .column.is-4-tablet{flex:none;width:33.33333337%}html.theme--documenter-dark .column.is-offset-4,html.theme--documenter-dark .column.is-offset-4-tablet{margin-left:33.33333337%}html.theme--documenter-dark .column.is-5,html.theme--documenter-dark .column.is-5-tablet{flex:none;width:41.66666674%}html.theme--documenter-dark .column.is-offset-5,html.theme--documenter-dark .column.is-offset-5-tablet{margin-left:41.66666674%}html.theme--documenter-dark .column.is-6,html.theme--documenter-dark .column.is-6-tablet{flex:none;width:50%}html.theme--documenter-dark .column.is-offset-6,html.theme--documenter-dark .column.is-offset-6-tablet{margin-left:50%}html.theme--documenter-dark .column.is-7,html.theme--documenter-dark .column.is-7-tablet{flex:none;width:58.33333337%}html.theme--documenter-dark .column.is-offset-7,html.theme--documenter-dark .column.is-offset-7-tablet{margin-left:58.33333337%}html.theme--documenter-dark .column.is-8,html.theme--documenter-dark .column.is-8-tablet{flex:none;width:66.66666674%}html.theme--documenter-dark .column.is-offset-8,html.theme--documenter-dark .column.is-offset-8-tablet{margin-left:66.66666674%}html.theme--documenter-dark .column.is-9,html.theme--documenter-dark .column.is-9-tablet{flex:none;width:75%}html.theme--documenter-dark .column.is-offset-9,html.theme--documenter-dark .column.is-offset-9-tablet{margin-left:75%}html.theme--documenter-dark .column.is-10,html.theme--documenter-dark .column.is-10-tablet{flex:none;width:83.33333337%}html.theme--documenter-dark .column.is-offset-10,html.theme--documenter-dark .column.is-offset-10-tablet{margin-left:83.33333337%}html.theme--documenter-dark .column.is-11,html.theme--documenter-dark .column.is-11-tablet{flex:none;width:91.66666674%}html.theme--documenter-dark .column.is-offset-11,html.theme--documenter-dark .column.is-offset-11-tablet{margin-left:91.66666674%}html.theme--documenter-dark .column.is-12,html.theme--documenter-dark .column.is-12-tablet{flex:none;width:100%}html.theme--documenter-dark .column.is-offset-12,html.theme--documenter-dark .column.is-offset-12-tablet{margin-left:100%}}@media screen and (max-width: 1055px){html.theme--documenter-dark .column.is-narrow-touch{flex:none;width:unset}html.theme--documenter-dark .column.is-full-touch{flex:none;width:100%}html.theme--documenter-dark .column.is-three-quarters-touch{flex:none;width:75%}html.theme--documenter-dark .column.is-two-thirds-touch{flex:none;width:66.6666%}html.theme--documenter-dark .column.is-half-touch{flex:none;width:50%}html.theme--documenter-dark .column.is-one-third-touch{flex:none;width:33.3333%}html.theme--documenter-dark .column.is-one-quarter-touch{flex:none;width:25%}html.theme--documenter-dark .column.is-one-fifth-touch{flex:none;width:20%}html.theme--documenter-dark .column.is-two-fifths-touch{flex:none;width:40%}html.theme--documenter-dark .column.is-three-fifths-touch{flex:none;width:60%}html.theme--documenter-dark .column.is-four-fifths-touch{flex:none;width:80%}html.theme--documenter-dark .column.is-offset-three-quarters-touch{margin-left:75%}html.theme--documenter-dark .column.is-offset-two-thirds-touch{margin-left:66.6666%}html.theme--documenter-dark .column.is-offset-half-touch{margin-left:50%}html.theme--documenter-dark .column.is-offset-one-third-touch{margin-left:33.3333%}html.theme--documenter-dark .column.is-offset-one-quarter-touch{margin-left:25%}html.theme--documenter-dark .column.is-offset-one-fifth-touch{margin-left:20%}html.theme--documenter-dark .column.is-offset-two-fifths-touch{margin-left:40%}html.theme--documenter-dark .column.is-offset-three-fifths-touch{margin-left:60%}html.theme--documenter-dark .column.is-offset-four-fifths-touch{margin-left:80%}html.theme--documenter-dark .column.is-0-touch{flex:none;width:0%}html.theme--documenter-dark .column.is-offset-0-touch{margin-left:0%}html.theme--documenter-dark .column.is-1-touch{flex:none;width:8.33333337%}html.theme--documenter-dark .column.is-offset-1-touch{margin-left:8.33333337%}html.theme--documenter-dark .column.is-2-touch{flex:none;width:16.66666674%}html.theme--documenter-dark .column.is-offset-2-touch{margin-left:16.66666674%}html.theme--documenter-dark .column.is-3-touch{flex:none;width:25%}html.theme--documenter-dark .column.is-offset-3-touch{margin-left:25%}html.theme--documenter-dark .column.is-4-touch{flex:none;width:33.33333337%}html.theme--documenter-dark .column.is-offset-4-touch{margin-left:33.33333337%}html.theme--documenter-dark .column.is-5-touch{flex:none;width:41.66666674%}html.theme--documenter-dark .column.is-offset-5-touch{margin-left:41.66666674%}html.theme--documenter-dark .column.is-6-touch{flex:none;width:50%}html.theme--documenter-dark .column.is-offset-6-touch{margin-left:50%}html.theme--documenter-dark .column.is-7-touch{flex:none;width:58.33333337%}html.theme--documenter-dark .column.is-offset-7-touch{margin-left:58.33333337%}html.theme--documenter-dark .column.is-8-touch{flex:none;width:66.66666674%}html.theme--documenter-dark .column.is-offset-8-touch{margin-left:66.66666674%}html.theme--documenter-dark .column.is-9-touch{flex:none;width:75%}html.theme--documenter-dark .column.is-offset-9-touch{margin-left:75%}html.theme--documenter-dark .column.is-10-touch{flex:none;width:83.33333337%}html.theme--documenter-dark .column.is-offset-10-touch{margin-left:83.33333337%}html.theme--documenter-dark .column.is-11-touch{flex:none;width:91.66666674%}html.theme--documenter-dark .column.is-offset-11-touch{margin-left:91.66666674%}html.theme--documenter-dark .column.is-12-touch{flex:none;width:100%}html.theme--documenter-dark .column.is-offset-12-touch{margin-left:100%}}@media screen and (min-width: 1056px){html.theme--documenter-dark .column.is-narrow-desktop{flex:none;width:unset}html.theme--documenter-dark .column.is-full-desktop{flex:none;width:100%}html.theme--documenter-dark .column.is-three-quarters-desktop{flex:none;width:75%}html.theme--documenter-dark .column.is-two-thirds-desktop{flex:none;width:66.6666%}html.theme--documenter-dark .column.is-half-desktop{flex:none;width:50%}html.theme--documenter-dark .column.is-one-third-desktop{flex:none;width:33.3333%}html.theme--documenter-dark .column.is-one-quarter-desktop{flex:none;width:25%}html.theme--documenter-dark .column.is-one-fifth-desktop{flex:none;width:20%}html.theme--documenter-dark .column.is-two-fifths-desktop{flex:none;width:40%}html.theme--documenter-dark .column.is-three-fifths-desktop{flex:none;width:60%}html.theme--documenter-dark .column.is-four-fifths-desktop{flex:none;width:80%}html.theme--documenter-dark .column.is-offset-three-quarters-desktop{margin-left:75%}html.theme--documenter-dark .column.is-offset-two-thirds-desktop{margin-left:66.6666%}html.theme--documenter-dark .column.is-offset-half-desktop{margin-left:50%}html.theme--documenter-dark .column.is-offset-one-third-desktop{margin-left:33.3333%}html.theme--documenter-dark .column.is-offset-one-quarter-desktop{margin-left:25%}html.theme--documenter-dark .column.is-offset-one-fifth-desktop{margin-left:20%}html.theme--documenter-dark .column.is-offset-two-fifths-desktop{margin-left:40%}html.theme--documenter-dark .column.is-offset-three-fifths-desktop{margin-left:60%}html.theme--documenter-dark .column.is-offset-four-fifths-desktop{margin-left:80%}html.theme--documenter-dark .column.is-0-desktop{flex:none;width:0%}html.theme--documenter-dark .column.is-offset-0-desktop{margin-left:0%}html.theme--documenter-dark .column.is-1-desktop{flex:none;width:8.33333337%}html.theme--documenter-dark .column.is-offset-1-desktop{margin-left:8.33333337%}html.theme--documenter-dark .column.is-2-desktop{flex:none;width:16.66666674%}html.theme--documenter-dark .column.is-offset-2-desktop{margin-left:16.66666674%}html.theme--documenter-dark .column.is-3-desktop{flex:none;width:25%}html.theme--documenter-dark .column.is-offset-3-desktop{margin-left:25%}html.theme--documenter-dark .column.is-4-desktop{flex:none;width:33.33333337%}html.theme--documenter-dark .column.is-offset-4-desktop{margin-left:33.33333337%}html.theme--documenter-dark .column.is-5-desktop{flex:none;width:41.66666674%}html.theme--documenter-dark .column.is-offset-5-desktop{margin-left:41.66666674%}html.theme--documenter-dark .column.is-6-desktop{flex:none;width:50%}html.theme--documenter-dark .column.is-offset-6-desktop{margin-left:50%}html.theme--documenter-dark .column.is-7-desktop{flex:none;width:58.33333337%}html.theme--documenter-dark .column.is-offset-7-desktop{margin-left:58.33333337%}html.theme--documenter-dark .column.is-8-desktop{flex:none;width:66.66666674%}html.theme--documenter-dark .column.is-offset-8-desktop{margin-left:66.66666674%}html.theme--documenter-dark .column.is-9-desktop{flex:none;width:75%}html.theme--documenter-dark .column.is-offset-9-desktop{margin-left:75%}html.theme--documenter-dark .column.is-10-desktop{flex:none;width:83.33333337%}html.theme--documenter-dark .column.is-offset-10-desktop{margin-left:83.33333337%}html.theme--documenter-dark .column.is-11-desktop{flex:none;width:91.66666674%}html.theme--documenter-dark .column.is-offset-11-desktop{margin-left:91.66666674%}html.theme--documenter-dark .column.is-12-desktop{flex:none;width:100%}html.theme--documenter-dark .column.is-offset-12-desktop{margin-left:100%}}@media screen and (min-width: 1216px){html.theme--documenter-dark .column.is-narrow-widescreen{flex:none;width:unset}html.theme--documenter-dark .column.is-full-widescreen{flex:none;width:100%}html.theme--documenter-dark .column.is-three-quarters-widescreen{flex:none;width:75%}html.theme--documenter-dark .column.is-two-thirds-widescreen{flex:none;width:66.6666%}html.theme--documenter-dark .column.is-half-widescreen{flex:none;width:50%}html.theme--documenter-dark .column.is-one-third-widescreen{flex:none;width:33.3333%}html.theme--documenter-dark .column.is-one-quarter-widescreen{flex:none;width:25%}html.theme--documenter-dark .column.is-one-fifth-widescreen{flex:none;width:20%}html.theme--documenter-dark .column.is-two-fifths-widescreen{flex:none;width:40%}html.theme--documenter-dark .column.is-three-fifths-widescreen{flex:none;width:60%}html.theme--documenter-dark .column.is-four-fifths-widescreen{flex:none;width:80%}html.theme--documenter-dark .column.is-offset-three-quarters-widescreen{margin-left:75%}html.theme--documenter-dark .column.is-offset-two-thirds-widescreen{margin-left:66.6666%}html.theme--documenter-dark .column.is-offset-half-widescreen{margin-left:50%}html.theme--documenter-dark .column.is-offset-one-third-widescreen{margin-left:33.3333%}html.theme--documenter-dark .column.is-offset-one-quarter-widescreen{margin-left:25%}html.theme--documenter-dark .column.is-offset-one-fifth-widescreen{margin-left:20%}html.theme--documenter-dark .column.is-offset-two-fifths-widescreen{margin-left:40%}html.theme--documenter-dark .column.is-offset-three-fifths-widescreen{margin-left:60%}html.theme--documenter-dark .column.is-offset-four-fifths-widescreen{margin-left:80%}html.theme--documenter-dark .column.is-0-widescreen{flex:none;width:0%}html.theme--documenter-dark .column.is-offset-0-widescreen{margin-left:0%}html.theme--documenter-dark .column.is-1-widescreen{flex:none;width:8.33333337%}html.theme--documenter-dark .column.is-offset-1-widescreen{margin-left:8.33333337%}html.theme--documenter-dark .column.is-2-widescreen{flex:none;width:16.66666674%}html.theme--documenter-dark .column.is-offset-2-widescreen{margin-left:16.66666674%}html.theme--documenter-dark .column.is-3-widescreen{flex:none;width:25%}html.theme--documenter-dark .column.is-offset-3-widescreen{margin-left:25%}html.theme--documenter-dark .column.is-4-widescreen{flex:none;width:33.33333337%}html.theme--documenter-dark .column.is-offset-4-widescreen{margin-left:33.33333337%}html.theme--documenter-dark .column.is-5-widescreen{flex:none;width:41.66666674%}html.theme--documenter-dark .column.is-offset-5-widescreen{margin-left:41.66666674%}html.theme--documenter-dark .column.is-6-widescreen{flex:none;width:50%}html.theme--documenter-dark .column.is-offset-6-widescreen{margin-left:50%}html.theme--documenter-dark .column.is-7-widescreen{flex:none;width:58.33333337%}html.theme--documenter-dark .column.is-offset-7-widescreen{margin-left:58.33333337%}html.theme--documenter-dark .column.is-8-widescreen{flex:none;width:66.66666674%}html.theme--documenter-dark .column.is-offset-8-widescreen{margin-left:66.66666674%}html.theme--documenter-dark .column.is-9-widescreen{flex:none;width:75%}html.theme--documenter-dark .column.is-offset-9-widescreen{margin-left:75%}html.theme--documenter-dark .column.is-10-widescreen{flex:none;width:83.33333337%}html.theme--documenter-dark .column.is-offset-10-widescreen{margin-left:83.33333337%}html.theme--documenter-dark .column.is-11-widescreen{flex:none;width:91.66666674%}html.theme--documenter-dark .column.is-offset-11-widescreen{margin-left:91.66666674%}html.theme--documenter-dark .column.is-12-widescreen{flex:none;width:100%}html.theme--documenter-dark .column.is-offset-12-widescreen{margin-left:100%}}@media screen and (min-width: 1408px){html.theme--documenter-dark .column.is-narrow-fullhd{flex:none;width:unset}html.theme--documenter-dark .column.is-full-fullhd{flex:none;width:100%}html.theme--documenter-dark .column.is-three-quarters-fullhd{flex:none;width:75%}html.theme--documenter-dark .column.is-two-thirds-fullhd{flex:none;width:66.6666%}html.theme--documenter-dark .column.is-half-fullhd{flex:none;width:50%}html.theme--documenter-dark .column.is-one-third-fullhd{flex:none;width:33.3333%}html.theme--documenter-dark .column.is-one-quarter-fullhd{flex:none;width:25%}html.theme--documenter-dark .column.is-one-fifth-fullhd{flex:none;width:20%}html.theme--documenter-dark .column.is-two-fifths-fullhd{flex:none;width:40%}html.theme--documenter-dark .column.is-three-fifths-fullhd{flex:none;width:60%}html.theme--documenter-dark .column.is-four-fifths-fullhd{flex:none;width:80%}html.theme--documenter-dark .column.is-offset-three-quarters-fullhd{margin-left:75%}html.theme--documenter-dark .column.is-offset-two-thirds-fullhd{margin-left:66.6666%}html.theme--documenter-dark .column.is-offset-half-fullhd{margin-left:50%}html.theme--documenter-dark .column.is-offset-one-third-fullhd{margin-left:33.3333%}html.theme--documenter-dark .column.is-offset-one-quarter-fullhd{margin-left:25%}html.theme--documenter-dark .column.is-offset-one-fifth-fullhd{margin-left:20%}html.theme--documenter-dark .column.is-offset-two-fifths-fullhd{margin-left:40%}html.theme--documenter-dark .column.is-offset-three-fifths-fullhd{margin-left:60%}html.theme--documenter-dark .column.is-offset-four-fifths-fullhd{margin-left:80%}html.theme--documenter-dark .column.is-0-fullhd{flex:none;width:0%}html.theme--documenter-dark .column.is-offset-0-fullhd{margin-left:0%}html.theme--documenter-dark .column.is-1-fullhd{flex:none;width:8.33333337%}html.theme--documenter-dark .column.is-offset-1-fullhd{margin-left:8.33333337%}html.theme--documenter-dark .column.is-2-fullhd{flex:none;width:16.66666674%}html.theme--documenter-dark .column.is-offset-2-fullhd{margin-left:16.66666674%}html.theme--documenter-dark .column.is-3-fullhd{flex:none;width:25%}html.theme--documenter-dark .column.is-offset-3-fullhd{margin-left:25%}html.theme--documenter-dark .column.is-4-fullhd{flex:none;width:33.33333337%}html.theme--documenter-dark .column.is-offset-4-fullhd{margin-left:33.33333337%}html.theme--documenter-dark .column.is-5-fullhd{flex:none;width:41.66666674%}html.theme--documenter-dark .column.is-offset-5-fullhd{margin-left:41.66666674%}html.theme--documenter-dark .column.is-6-fullhd{flex:none;width:50%}html.theme--documenter-dark .column.is-offset-6-fullhd{margin-left:50%}html.theme--documenter-dark .column.is-7-fullhd{flex:none;width:58.33333337%}html.theme--documenter-dark .column.is-offset-7-fullhd{margin-left:58.33333337%}html.theme--documenter-dark .column.is-8-fullhd{flex:none;width:66.66666674%}html.theme--documenter-dark .column.is-offset-8-fullhd{margin-left:66.66666674%}html.theme--documenter-dark .column.is-9-fullhd{flex:none;width:75%}html.theme--documenter-dark .column.is-offset-9-fullhd{margin-left:75%}html.theme--documenter-dark .column.is-10-fullhd{flex:none;width:83.33333337%}html.theme--documenter-dark .column.is-offset-10-fullhd{margin-left:83.33333337%}html.theme--documenter-dark .column.is-11-fullhd{flex:none;width:91.66666674%}html.theme--documenter-dark .column.is-offset-11-fullhd{margin-left:91.66666674%}html.theme--documenter-dark .column.is-12-fullhd{flex:none;width:100%}html.theme--documenter-dark .column.is-offset-12-fullhd{margin-left:100%}}html.theme--documenter-dark .columns{margin-left:-.75rem;margin-right:-.75rem;margin-top:-.75rem}html.theme--documenter-dark .columns:last-child{margin-bottom:-.75rem}html.theme--documenter-dark .columns:not(:last-child){margin-bottom:calc(1.5rem - .75rem)}html.theme--documenter-dark .columns.is-centered{justify-content:center}html.theme--documenter-dark .columns.is-gapless{margin-left:0;margin-right:0;margin-top:0}html.theme--documenter-dark .columns.is-gapless>.column{margin:0;padding:0 !important}html.theme--documenter-dark .columns.is-gapless:not(:last-child){margin-bottom:1.5rem}html.theme--documenter-dark .columns.is-gapless:last-child{margin-bottom:0}html.theme--documenter-dark .columns.is-mobile{display:flex}html.theme--documenter-dark .columns.is-multiline{flex-wrap:wrap}html.theme--documenter-dark .columns.is-vcentered{align-items:center}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns:not(.is-desktop){display:flex}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-desktop{display:flex}}html.theme--documenter-dark .columns.is-variable{--columnGap: 0.75rem;margin-left:calc(-1 * var(--columnGap));margin-right:calc(-1 * var(--columnGap))}html.theme--documenter-dark .columns.is-variable>.column{padding-left:var(--columnGap);padding-right:var(--columnGap)}html.theme--documenter-dark .columns.is-variable.is-0{--columnGap: 0rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-0-mobile{--columnGap: 0rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-0-tablet{--columnGap: 0rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-0-tablet-only{--columnGap: 0rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-0-touch{--columnGap: 0rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-0-desktop{--columnGap: 0rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-0-desktop-only{--columnGap: 0rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-0-widescreen{--columnGap: 0rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-0-widescreen-only{--columnGap: 0rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-0-fullhd{--columnGap: 0rem}}html.theme--documenter-dark .columns.is-variable.is-1{--columnGap: .25rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-1-mobile{--columnGap: .25rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-1-tablet{--columnGap: .25rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-1-tablet-only{--columnGap: .25rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-1-touch{--columnGap: .25rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-1-desktop{--columnGap: .25rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-1-desktop-only{--columnGap: .25rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-1-widescreen{--columnGap: .25rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-1-widescreen-only{--columnGap: .25rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-1-fullhd{--columnGap: .25rem}}html.theme--documenter-dark .columns.is-variable.is-2{--columnGap: .5rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-2-mobile{--columnGap: .5rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-2-tablet{--columnGap: .5rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-2-tablet-only{--columnGap: .5rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-2-touch{--columnGap: .5rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-2-desktop{--columnGap: .5rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-2-desktop-only{--columnGap: .5rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-2-widescreen{--columnGap: .5rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-2-widescreen-only{--columnGap: .5rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-2-fullhd{--columnGap: .5rem}}html.theme--documenter-dark .columns.is-variable.is-3{--columnGap: .75rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-3-mobile{--columnGap: .75rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-3-tablet{--columnGap: .75rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-3-tablet-only{--columnGap: .75rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-3-touch{--columnGap: .75rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-3-desktop{--columnGap: .75rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-3-desktop-only{--columnGap: .75rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-3-widescreen{--columnGap: .75rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-3-widescreen-only{--columnGap: .75rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-3-fullhd{--columnGap: .75rem}}html.theme--documenter-dark .columns.is-variable.is-4{--columnGap: 1rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-4-mobile{--columnGap: 1rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-4-tablet{--columnGap: 1rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-4-tablet-only{--columnGap: 1rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-4-touch{--columnGap: 1rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-4-desktop{--columnGap: 1rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-4-desktop-only{--columnGap: 1rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-4-widescreen{--columnGap: 1rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-4-widescreen-only{--columnGap: 1rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-4-fullhd{--columnGap: 1rem}}html.theme--documenter-dark .columns.is-variable.is-5{--columnGap: 1.25rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-5-mobile{--columnGap: 1.25rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-5-tablet{--columnGap: 1.25rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-5-tablet-only{--columnGap: 1.25rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-5-touch{--columnGap: 1.25rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-5-desktop{--columnGap: 1.25rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-5-desktop-only{--columnGap: 1.25rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-5-widescreen{--columnGap: 1.25rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-5-widescreen-only{--columnGap: 1.25rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-5-fullhd{--columnGap: 1.25rem}}html.theme--documenter-dark .columns.is-variable.is-6{--columnGap: 1.5rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-6-mobile{--columnGap: 1.5rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-6-tablet{--columnGap: 1.5rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-6-tablet-only{--columnGap: 1.5rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-6-touch{--columnGap: 1.5rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-6-desktop{--columnGap: 1.5rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-6-desktop-only{--columnGap: 1.5rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-6-widescreen{--columnGap: 1.5rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-6-widescreen-only{--columnGap: 1.5rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-6-fullhd{--columnGap: 1.5rem}}html.theme--documenter-dark .columns.is-variable.is-7{--columnGap: 1.75rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-7-mobile{--columnGap: 1.75rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-7-tablet{--columnGap: 1.75rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-7-tablet-only{--columnGap: 1.75rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-7-touch{--columnGap: 1.75rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-7-desktop{--columnGap: 1.75rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-7-desktop-only{--columnGap: 1.75rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-7-widescreen{--columnGap: 1.75rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-7-widescreen-only{--columnGap: 1.75rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-7-fullhd{--columnGap: 1.75rem}}html.theme--documenter-dark .columns.is-variable.is-8{--columnGap: 2rem}@media screen and (max-width: 768px){html.theme--documenter-dark .columns.is-variable.is-8-mobile{--columnGap: 2rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .columns.is-variable.is-8-tablet{--columnGap: 2rem}}@media screen and (min-width: 769px) and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-8-tablet-only{--columnGap: 2rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark .columns.is-variable.is-8-touch{--columnGap: 2rem}}@media screen and (min-width: 1056px){html.theme--documenter-dark .columns.is-variable.is-8-desktop{--columnGap: 2rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){html.theme--documenter-dark .columns.is-variable.is-8-desktop-only{--columnGap: 2rem}}@media screen and (min-width: 1216px){html.theme--documenter-dark .columns.is-variable.is-8-widescreen{--columnGap: 2rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){html.theme--documenter-dark .columns.is-variable.is-8-widescreen-only{--columnGap: 2rem}}@media screen and (min-width: 1408px){html.theme--documenter-dark .columns.is-variable.is-8-fullhd{--columnGap: 2rem}}html.theme--documenter-dark .tile{align-items:stretch;display:block;flex-basis:0;flex-grow:1;flex-shrink:1;min-height:min-content}html.theme--documenter-dark .tile.is-ancestor{margin-left:-.75rem;margin-right:-.75rem;margin-top:-.75rem}html.theme--documenter-dark .tile.is-ancestor:last-child{margin-bottom:-.75rem}html.theme--documenter-dark .tile.is-ancestor:not(:last-child){margin-bottom:.75rem}html.theme--documenter-dark .tile.is-child{margin:0 !important}html.theme--documenter-dark .tile.is-parent{padding:.75rem}html.theme--documenter-dark .tile.is-vertical{flex-direction:column}html.theme--documenter-dark .tile.is-vertical>.tile.is-child:not(:last-child){margin-bottom:1.5rem !important}@media screen and (min-width: 769px),print{html.theme--documenter-dark .tile:not(.is-child){display:flex}html.theme--documenter-dark .tile.is-1{flex:none;width:8.33333337%}html.theme--documenter-dark .tile.is-2{flex:none;width:16.66666674%}html.theme--documenter-dark .tile.is-3{flex:none;width:25%}html.theme--documenter-dark .tile.is-4{flex:none;width:33.33333337%}html.theme--documenter-dark .tile.is-5{flex:none;width:41.66666674%}html.theme--documenter-dark .tile.is-6{flex:none;width:50%}html.theme--documenter-dark .tile.is-7{flex:none;width:58.33333337%}html.theme--documenter-dark .tile.is-8{flex:none;width:66.66666674%}html.theme--documenter-dark .tile.is-9{flex:none;width:75%}html.theme--documenter-dark .tile.is-10{flex:none;width:83.33333337%}html.theme--documenter-dark .tile.is-11{flex:none;width:91.66666674%}html.theme--documenter-dark .tile.is-12{flex:none;width:100%}}html.theme--documenter-dark .hero{align-items:stretch;display:flex;flex-direction:column;justify-content:space-between}html.theme--documenter-dark .hero .navbar{background:none}html.theme--documenter-dark .hero .tabs ul{border-bottom:none}html.theme--documenter-dark .hero.is-white{background-color:#fff;color:#0a0a0a}html.theme--documenter-dark .hero.is-white a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-white strong{color:inherit}html.theme--documenter-dark .hero.is-white .title{color:#0a0a0a}html.theme--documenter-dark .hero.is-white .subtitle{color:rgba(10,10,10,0.9)}html.theme--documenter-dark .hero.is-white .subtitle a:not(.button),html.theme--documenter-dark .hero.is-white .subtitle strong{color:#0a0a0a}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-white .navbar-menu{background-color:#fff}}html.theme--documenter-dark .hero.is-white .navbar-item,html.theme--documenter-dark .hero.is-white .navbar-link{color:rgba(10,10,10,0.7)}html.theme--documenter-dark .hero.is-white a.navbar-item:hover,html.theme--documenter-dark .hero.is-white a.navbar-item.is-active,html.theme--documenter-dark .hero.is-white .navbar-link:hover,html.theme--documenter-dark .hero.is-white .navbar-link.is-active{background-color:#f2f2f2;color:#0a0a0a}html.theme--documenter-dark .hero.is-white .tabs a{color:#0a0a0a;opacity:0.9}html.theme--documenter-dark .hero.is-white .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-white .tabs li.is-active a{color:#fff !important;opacity:1}html.theme--documenter-dark .hero.is-white .tabs.is-boxed a,html.theme--documenter-dark .hero.is-white .tabs.is-toggle a{color:#0a0a0a}html.theme--documenter-dark .hero.is-white .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-white .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-white .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-white .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-white .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-white .tabs.is-toggle li.is-active a:hover{background-color:#0a0a0a;border-color:#0a0a0a;color:#fff}html.theme--documenter-dark .hero.is-white.is-bold{background-image:linear-gradient(141deg, #e8e3e4 0%, #fff 71%, #fff 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-white.is-bold .navbar-menu{background-image:linear-gradient(141deg, #e8e3e4 0%, #fff 71%, #fff 100%)}}html.theme--documenter-dark .hero.is-black{background-color:#0a0a0a;color:#fff}html.theme--documenter-dark .hero.is-black a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-black strong{color:inherit}html.theme--documenter-dark .hero.is-black .title{color:#fff}html.theme--documenter-dark .hero.is-black .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-black .subtitle a:not(.button),html.theme--documenter-dark .hero.is-black .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-black .navbar-menu{background-color:#0a0a0a}}html.theme--documenter-dark .hero.is-black .navbar-item,html.theme--documenter-dark .hero.is-black .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-black a.navbar-item:hover,html.theme--documenter-dark .hero.is-black a.navbar-item.is-active,html.theme--documenter-dark .hero.is-black .navbar-link:hover,html.theme--documenter-dark .hero.is-black .navbar-link.is-active{background-color:#000;color:#fff}html.theme--documenter-dark .hero.is-black .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-black .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-black .tabs li.is-active a{color:#0a0a0a !important;opacity:1}html.theme--documenter-dark .hero.is-black .tabs.is-boxed a,html.theme--documenter-dark .hero.is-black .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-black .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-black .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-black .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-black .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-black .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-black .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#0a0a0a}html.theme--documenter-dark .hero.is-black.is-bold{background-image:linear-gradient(141deg, #000 0%, #0a0a0a 71%, #181616 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-black.is-bold .navbar-menu{background-image:linear-gradient(141deg, #000 0%, #0a0a0a 71%, #181616 100%)}}html.theme--documenter-dark .hero.is-light{background-color:#ecf0f1;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .hero.is-light a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-light strong{color:inherit}html.theme--documenter-dark .hero.is-light .title{color:rgba(0,0,0,0.7)}html.theme--documenter-dark .hero.is-light .subtitle{color:rgba(0,0,0,0.9)}html.theme--documenter-dark .hero.is-light .subtitle a:not(.button),html.theme--documenter-dark .hero.is-light .subtitle strong{color:rgba(0,0,0,0.7)}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-light .navbar-menu{background-color:#ecf0f1}}html.theme--documenter-dark .hero.is-light .navbar-item,html.theme--documenter-dark .hero.is-light .navbar-link{color:rgba(0,0,0,0.7)}html.theme--documenter-dark .hero.is-light a.navbar-item:hover,html.theme--documenter-dark .hero.is-light a.navbar-item.is-active,html.theme--documenter-dark .hero.is-light .navbar-link:hover,html.theme--documenter-dark .hero.is-light .navbar-link.is-active{background-color:#dde4e6;color:rgba(0,0,0,0.7)}html.theme--documenter-dark .hero.is-light .tabs a{color:rgba(0,0,0,0.7);opacity:0.9}html.theme--documenter-dark .hero.is-light .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-light .tabs li.is-active a{color:#ecf0f1 !important;opacity:1}html.theme--documenter-dark .hero.is-light .tabs.is-boxed a,html.theme--documenter-dark .hero.is-light .tabs.is-toggle a{color:rgba(0,0,0,0.7)}html.theme--documenter-dark .hero.is-light .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-light .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-light .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-light .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-light .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-light .tabs.is-toggle li.is-active a:hover{background-color:rgba(0,0,0,0.7);border-color:rgba(0,0,0,0.7);color:#ecf0f1}html.theme--documenter-dark .hero.is-light.is-bold{background-image:linear-gradient(141deg, #cadfe0 0%, #ecf0f1 71%, #fafbfc 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-light.is-bold .navbar-menu{background-image:linear-gradient(141deg, #cadfe0 0%, #ecf0f1 71%, #fafbfc 100%)}}html.theme--documenter-dark .hero.is-dark,html.theme--documenter-dark .content kbd.hero{background-color:#282f2f;color:#fff}html.theme--documenter-dark .hero.is-dark a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .content kbd.hero a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-dark strong,html.theme--documenter-dark .content kbd.hero strong{color:inherit}html.theme--documenter-dark .hero.is-dark .title,html.theme--documenter-dark .content kbd.hero .title{color:#fff}html.theme--documenter-dark .hero.is-dark .subtitle,html.theme--documenter-dark .content kbd.hero .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-dark .subtitle a:not(.button),html.theme--documenter-dark .content kbd.hero .subtitle a:not(.button),html.theme--documenter-dark .hero.is-dark .subtitle strong,html.theme--documenter-dark .content kbd.hero .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-dark .navbar-menu,html.theme--documenter-dark .content kbd.hero .navbar-menu{background-color:#282f2f}}html.theme--documenter-dark .hero.is-dark .navbar-item,html.theme--documenter-dark .content kbd.hero .navbar-item,html.theme--documenter-dark .hero.is-dark .navbar-link,html.theme--documenter-dark .content kbd.hero .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-dark a.navbar-item:hover,html.theme--documenter-dark .content kbd.hero a.navbar-item:hover,html.theme--documenter-dark .hero.is-dark a.navbar-item.is-active,html.theme--documenter-dark .content kbd.hero a.navbar-item.is-active,html.theme--documenter-dark .hero.is-dark .navbar-link:hover,html.theme--documenter-dark .content kbd.hero .navbar-link:hover,html.theme--documenter-dark .hero.is-dark .navbar-link.is-active,html.theme--documenter-dark .content kbd.hero .navbar-link.is-active{background-color:#1d2122;color:#fff}html.theme--documenter-dark .hero.is-dark .tabs a,html.theme--documenter-dark .content kbd.hero .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-dark .tabs a:hover,html.theme--documenter-dark .content kbd.hero .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-dark .tabs li.is-active a,html.theme--documenter-dark .content kbd.hero .tabs li.is-active a{color:#282f2f !important;opacity:1}html.theme--documenter-dark .hero.is-dark .tabs.is-boxed a,html.theme--documenter-dark .content kbd.hero .tabs.is-boxed a,html.theme--documenter-dark .hero.is-dark .tabs.is-toggle a,html.theme--documenter-dark .content kbd.hero .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-dark .tabs.is-boxed a:hover,html.theme--documenter-dark .content kbd.hero .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-dark .tabs.is-toggle a:hover,html.theme--documenter-dark .content kbd.hero .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-dark .tabs.is-boxed li.is-active a,html.theme--documenter-dark .content kbd.hero .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-dark .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-dark .tabs.is-toggle li.is-active a,html.theme--documenter-dark .content kbd.hero .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-dark .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#282f2f}html.theme--documenter-dark .hero.is-dark.is-bold,html.theme--documenter-dark .content kbd.hero.is-bold{background-image:linear-gradient(141deg, #0f1615 0%, #282f2f 71%, #313c40 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-dark.is-bold .navbar-menu,html.theme--documenter-dark .content kbd.hero.is-bold .navbar-menu{background-image:linear-gradient(141deg, #0f1615 0%, #282f2f 71%, #313c40 100%)}}html.theme--documenter-dark .hero.is-primary,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink{background-color:#375a7f;color:#fff}html.theme--documenter-dark .hero.is-primary a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-primary strong,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink strong{color:inherit}html.theme--documenter-dark .hero.is-primary .title,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .title{color:#fff}html.theme--documenter-dark .hero.is-primary .subtitle,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-primary .subtitle a:not(.button),html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .subtitle a:not(.button),html.theme--documenter-dark .hero.is-primary .subtitle strong,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-primary .navbar-menu,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .navbar-menu{background-color:#375a7f}}html.theme--documenter-dark .hero.is-primary .navbar-item,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .navbar-item,html.theme--documenter-dark .hero.is-primary .navbar-link,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-primary a.navbar-item:hover,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink a.navbar-item:hover,html.theme--documenter-dark .hero.is-primary a.navbar-item.is-active,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink a.navbar-item.is-active,html.theme--documenter-dark .hero.is-primary .navbar-link:hover,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .navbar-link:hover,html.theme--documenter-dark .hero.is-primary .navbar-link.is-active,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .navbar-link.is-active{background-color:#2f4d6d;color:#fff}html.theme--documenter-dark .hero.is-primary .tabs a,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-primary .tabs a:hover,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-primary .tabs li.is-active a,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs li.is-active a{color:#375a7f !important;opacity:1}html.theme--documenter-dark .hero.is-primary .tabs.is-boxed a,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs.is-boxed a,html.theme--documenter-dark .hero.is-primary .tabs.is-toggle a,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-primary .tabs.is-boxed a:hover,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-primary .tabs.is-toggle a:hover,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-primary .tabs.is-boxed li.is-active a,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-primary .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-primary .tabs.is-toggle li.is-active a,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-primary .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#375a7f}html.theme--documenter-dark .hero.is-primary.is-bold,html.theme--documenter-dark .docstring>section>a.hero.is-bold.docs-sourcelink{background-image:linear-gradient(141deg, #214b62 0%, #375a7f 71%, #3a5796 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-primary.is-bold .navbar-menu,html.theme--documenter-dark .docstring>section>a.hero.is-bold.docs-sourcelink .navbar-menu{background-image:linear-gradient(141deg, #214b62 0%, #375a7f 71%, #3a5796 100%)}}html.theme--documenter-dark .hero.is-link{background-color:#1abc9c;color:#fff}html.theme--documenter-dark .hero.is-link a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-link strong{color:inherit}html.theme--documenter-dark .hero.is-link .title{color:#fff}html.theme--documenter-dark .hero.is-link .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-link .subtitle a:not(.button),html.theme--documenter-dark .hero.is-link .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-link .navbar-menu{background-color:#1abc9c}}html.theme--documenter-dark .hero.is-link .navbar-item,html.theme--documenter-dark .hero.is-link .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-link a.navbar-item:hover,html.theme--documenter-dark .hero.is-link a.navbar-item.is-active,html.theme--documenter-dark .hero.is-link .navbar-link:hover,html.theme--documenter-dark .hero.is-link .navbar-link.is-active{background-color:#17a689;color:#fff}html.theme--documenter-dark .hero.is-link .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-link .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-link .tabs li.is-active a{color:#1abc9c !important;opacity:1}html.theme--documenter-dark .hero.is-link .tabs.is-boxed a,html.theme--documenter-dark .hero.is-link .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-link .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-link .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-link .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-link .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-link .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-link .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#1abc9c}html.theme--documenter-dark .hero.is-link.is-bold{background-image:linear-gradient(141deg, #0c9764 0%, #1abc9c 71%, #17d8d2 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-link.is-bold .navbar-menu{background-image:linear-gradient(141deg, #0c9764 0%, #1abc9c 71%, #17d8d2 100%)}}html.theme--documenter-dark .hero.is-info{background-color:#024c7d;color:#fff}html.theme--documenter-dark .hero.is-info a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-info strong{color:inherit}html.theme--documenter-dark .hero.is-info .title{color:#fff}html.theme--documenter-dark .hero.is-info .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-info .subtitle a:not(.button),html.theme--documenter-dark .hero.is-info .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-info .navbar-menu{background-color:#024c7d}}html.theme--documenter-dark .hero.is-info .navbar-item,html.theme--documenter-dark .hero.is-info .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-info a.navbar-item:hover,html.theme--documenter-dark .hero.is-info a.navbar-item.is-active,html.theme--documenter-dark .hero.is-info .navbar-link:hover,html.theme--documenter-dark .hero.is-info .navbar-link.is-active{background-color:#023d64;color:#fff}html.theme--documenter-dark .hero.is-info .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-info .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-info .tabs li.is-active a{color:#024c7d !important;opacity:1}html.theme--documenter-dark .hero.is-info .tabs.is-boxed a,html.theme--documenter-dark .hero.is-info .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-info .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-info .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-info .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-info .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-info .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-info .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#024c7d}html.theme--documenter-dark .hero.is-info.is-bold{background-image:linear-gradient(141deg, #003a4c 0%, #024c7d 71%, #004299 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-info.is-bold .navbar-menu{background-image:linear-gradient(141deg, #003a4c 0%, #024c7d 71%, #004299 100%)}}html.theme--documenter-dark .hero.is-success{background-color:#008438;color:#fff}html.theme--documenter-dark .hero.is-success a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-success strong{color:inherit}html.theme--documenter-dark .hero.is-success .title{color:#fff}html.theme--documenter-dark .hero.is-success .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-success .subtitle a:not(.button),html.theme--documenter-dark .hero.is-success .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-success .navbar-menu{background-color:#008438}}html.theme--documenter-dark .hero.is-success .navbar-item,html.theme--documenter-dark .hero.is-success .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-success a.navbar-item:hover,html.theme--documenter-dark .hero.is-success a.navbar-item.is-active,html.theme--documenter-dark .hero.is-success .navbar-link:hover,html.theme--documenter-dark .hero.is-success .navbar-link.is-active{background-color:#006b2d;color:#fff}html.theme--documenter-dark .hero.is-success .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-success .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-success .tabs li.is-active a{color:#008438 !important;opacity:1}html.theme--documenter-dark .hero.is-success .tabs.is-boxed a,html.theme--documenter-dark .hero.is-success .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-success .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-success .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-success .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-success .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-success .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-success .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#008438}html.theme--documenter-dark .hero.is-success.is-bold{background-image:linear-gradient(141deg, #005115 0%, #008438 71%, #009e5d 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-success.is-bold .navbar-menu{background-image:linear-gradient(141deg, #005115 0%, #008438 71%, #009e5d 100%)}}html.theme--documenter-dark .hero.is-warning{background-color:#ad8100;color:#fff}html.theme--documenter-dark .hero.is-warning a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-warning strong{color:inherit}html.theme--documenter-dark .hero.is-warning .title{color:#fff}html.theme--documenter-dark .hero.is-warning .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-warning .subtitle a:not(.button),html.theme--documenter-dark .hero.is-warning .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-warning .navbar-menu{background-color:#ad8100}}html.theme--documenter-dark .hero.is-warning .navbar-item,html.theme--documenter-dark .hero.is-warning .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-warning a.navbar-item:hover,html.theme--documenter-dark .hero.is-warning a.navbar-item.is-active,html.theme--documenter-dark .hero.is-warning .navbar-link:hover,html.theme--documenter-dark .hero.is-warning .navbar-link.is-active{background-color:#946e00;color:#fff}html.theme--documenter-dark .hero.is-warning .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-warning .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-warning .tabs li.is-active a{color:#ad8100 !important;opacity:1}html.theme--documenter-dark .hero.is-warning .tabs.is-boxed a,html.theme--documenter-dark .hero.is-warning .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-warning .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-warning .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-warning .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-warning .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-warning .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-warning .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#ad8100}html.theme--documenter-dark .hero.is-warning.is-bold{background-image:linear-gradient(141deg, #7a4700 0%, #ad8100 71%, #c7b500 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-warning.is-bold .navbar-menu{background-image:linear-gradient(141deg, #7a4700 0%, #ad8100 71%, #c7b500 100%)}}html.theme--documenter-dark .hero.is-danger{background-color:#9e1b0d;color:#fff}html.theme--documenter-dark .hero.is-danger a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),html.theme--documenter-dark .hero.is-danger strong{color:inherit}html.theme--documenter-dark .hero.is-danger .title{color:#fff}html.theme--documenter-dark .hero.is-danger .subtitle{color:rgba(255,255,255,0.9)}html.theme--documenter-dark .hero.is-danger .subtitle a:not(.button),html.theme--documenter-dark .hero.is-danger .subtitle strong{color:#fff}@media screen and (max-width: 1055px){html.theme--documenter-dark .hero.is-danger .navbar-menu{background-color:#9e1b0d}}html.theme--documenter-dark .hero.is-danger .navbar-item,html.theme--documenter-dark .hero.is-danger .navbar-link{color:rgba(255,255,255,0.7)}html.theme--documenter-dark .hero.is-danger a.navbar-item:hover,html.theme--documenter-dark .hero.is-danger a.navbar-item.is-active,html.theme--documenter-dark .hero.is-danger .navbar-link:hover,html.theme--documenter-dark .hero.is-danger .navbar-link.is-active{background-color:#86170b;color:#fff}html.theme--documenter-dark .hero.is-danger .tabs a{color:#fff;opacity:0.9}html.theme--documenter-dark .hero.is-danger .tabs a:hover{opacity:1}html.theme--documenter-dark .hero.is-danger .tabs li.is-active a{color:#9e1b0d !important;opacity:1}html.theme--documenter-dark .hero.is-danger .tabs.is-boxed a,html.theme--documenter-dark .hero.is-danger .tabs.is-toggle a{color:#fff}html.theme--documenter-dark .hero.is-danger .tabs.is-boxed a:hover,html.theme--documenter-dark .hero.is-danger .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}html.theme--documenter-dark .hero.is-danger .tabs.is-boxed li.is-active a,html.theme--documenter-dark .hero.is-danger .tabs.is-boxed li.is-active a:hover,html.theme--documenter-dark .hero.is-danger .tabs.is-toggle li.is-active a,html.theme--documenter-dark .hero.is-danger .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#9e1b0d}html.theme--documenter-dark .hero.is-danger.is-bold{background-image:linear-gradient(141deg, #75030b 0%, #9e1b0d 71%, #ba380a 100%)}@media screen and (max-width: 768px){html.theme--documenter-dark .hero.is-danger.is-bold .navbar-menu{background-image:linear-gradient(141deg, #75030b 0%, #9e1b0d 71%, #ba380a 100%)}}html.theme--documenter-dark .hero.is-small .hero-body,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.hero .hero-body{padding:1.5rem}@media screen and (min-width: 769px),print{html.theme--documenter-dark .hero.is-medium .hero-body{padding:9rem 4.5rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .hero.is-large .hero-body{padding:18rem 6rem}}html.theme--documenter-dark .hero.is-halfheight .hero-body,html.theme--documenter-dark .hero.is-fullheight .hero-body,html.theme--documenter-dark .hero.is-fullheight-with-navbar .hero-body{align-items:center;display:flex}html.theme--documenter-dark .hero.is-halfheight .hero-body>.container,html.theme--documenter-dark .hero.is-fullheight .hero-body>.container,html.theme--documenter-dark .hero.is-fullheight-with-navbar .hero-body>.container{flex-grow:1;flex-shrink:1}html.theme--documenter-dark .hero.is-halfheight{min-height:50vh}html.theme--documenter-dark .hero.is-fullheight{min-height:100vh}html.theme--documenter-dark .hero-video{overflow:hidden}html.theme--documenter-dark .hero-video video{left:50%;min-height:100%;min-width:100%;position:absolute;top:50%;transform:translate3d(-50%, -50%, 0)}html.theme--documenter-dark .hero-video.is-transparent{opacity:0.3}@media screen and (max-width: 768px){html.theme--documenter-dark .hero-video{display:none}}html.theme--documenter-dark .hero-buttons{margin-top:1.5rem}@media screen and (max-width: 768px){html.theme--documenter-dark .hero-buttons .button{display:flex}html.theme--documenter-dark .hero-buttons .button:not(:last-child){margin-bottom:0.75rem}}@media screen and (min-width: 769px),print{html.theme--documenter-dark .hero-buttons{display:flex;justify-content:center}html.theme--documenter-dark .hero-buttons .button:not(:last-child){margin-right:1.5rem}}html.theme--documenter-dark .hero-head,html.theme--documenter-dark .hero-foot{flex-grow:0;flex-shrink:0}html.theme--documenter-dark .hero-body{flex-grow:1;flex-shrink:0;padding:3rem 1.5rem}@media screen and (min-width: 769px),print{html.theme--documenter-dark .hero-body{padding:3rem 3rem}}html.theme--documenter-dark .section{padding:3rem 1.5rem}@media screen and (min-width: 1056px){html.theme--documenter-dark .section{padding:3rem 3rem}html.theme--documenter-dark .section.is-medium{padding:9rem 4.5rem}html.theme--documenter-dark .section.is-large{padding:18rem 6rem}}html.theme--documenter-dark .footer{background-color:#282f2f;padding:3rem 1.5rem 6rem}html.theme--documenter-dark hr{height:1px}html.theme--documenter-dark h6{text-transform:uppercase;letter-spacing:0.5px}html.theme--documenter-dark .hero{background-color:#343c3d}html.theme--documenter-dark a{transition:all 200ms ease}html.theme--documenter-dark .button{transition:all 200ms ease;border-width:1px;color:#fff}html.theme--documenter-dark .button.is-active,html.theme--documenter-dark .button.is-focused,html.theme--documenter-dark .button:active,html.theme--documenter-dark .button:focus{box-shadow:0 0 0 2px rgba(140,155,157,0.5)}html.theme--documenter-dark .button.is-white.is-hovered,html.theme--documenter-dark .button.is-white:hover{background-color:#fff}html.theme--documenter-dark .button.is-white.is-active,html.theme--documenter-dark .button.is-white.is-focused,html.theme--documenter-dark .button.is-white:active,html.theme--documenter-dark .button.is-white:focus{border-color:#fff;box-shadow:0 0 0 2px rgba(255,255,255,0.5)}html.theme--documenter-dark .button.is-black.is-hovered,html.theme--documenter-dark .button.is-black:hover{background-color:#1d1d1d}html.theme--documenter-dark .button.is-black.is-active,html.theme--documenter-dark .button.is-black.is-focused,html.theme--documenter-dark .button.is-black:active,html.theme--documenter-dark .button.is-black:focus{border-color:#0a0a0a;box-shadow:0 0 0 2px rgba(10,10,10,0.5)}html.theme--documenter-dark .button.is-light.is-hovered,html.theme--documenter-dark .button.is-light:hover{background-color:#fff}html.theme--documenter-dark .button.is-light.is-active,html.theme--documenter-dark .button.is-light.is-focused,html.theme--documenter-dark .button.is-light:active,html.theme--documenter-dark .button.is-light:focus{border-color:#ecf0f1;box-shadow:0 0 0 2px rgba(236,240,241,0.5)}html.theme--documenter-dark .button.is-dark.is-hovered,html.theme--documenter-dark .content kbd.button.is-hovered,html.theme--documenter-dark .button.is-dark:hover,html.theme--documenter-dark .content kbd.button:hover{background-color:#3a4344}html.theme--documenter-dark .button.is-dark.is-active,html.theme--documenter-dark .content kbd.button.is-active,html.theme--documenter-dark .button.is-dark.is-focused,html.theme--documenter-dark .content kbd.button.is-focused,html.theme--documenter-dark .button.is-dark:active,html.theme--documenter-dark .content kbd.button:active,html.theme--documenter-dark .button.is-dark:focus,html.theme--documenter-dark .content kbd.button:focus{border-color:#282f2f;box-shadow:0 0 0 2px rgba(40,47,47,0.5)}html.theme--documenter-dark .button.is-primary.is-hovered,html.theme--documenter-dark .docstring>section>a.button.is-hovered.docs-sourcelink,html.theme--documenter-dark .button.is-primary:hover,html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink:hover{background-color:#436d9a}html.theme--documenter-dark .button.is-primary.is-active,html.theme--documenter-dark .docstring>section>a.button.is-active.docs-sourcelink,html.theme--documenter-dark .button.is-primary.is-focused,html.theme--documenter-dark .docstring>section>a.button.is-focused.docs-sourcelink,html.theme--documenter-dark .button.is-primary:active,html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink:active,html.theme--documenter-dark .button.is-primary:focus,html.theme--documenter-dark .docstring>section>a.button.docs-sourcelink:focus{border-color:#375a7f;box-shadow:0 0 0 2px rgba(55,90,127,0.5)}html.theme--documenter-dark .button.is-link.is-hovered,html.theme--documenter-dark .button.is-link:hover{background-color:#1fdeb8}html.theme--documenter-dark .button.is-link.is-active,html.theme--documenter-dark .button.is-link.is-focused,html.theme--documenter-dark .button.is-link:active,html.theme--documenter-dark .button.is-link:focus{border-color:#1abc9c;box-shadow:0 0 0 2px rgba(26,188,156,0.5)}html.theme--documenter-dark .button.is-info.is-hovered,html.theme--documenter-dark .button.is-info:hover{background-color:#0363a3}html.theme--documenter-dark .button.is-info.is-active,html.theme--documenter-dark .button.is-info.is-focused,html.theme--documenter-dark .button.is-info:active,html.theme--documenter-dark .button.is-info:focus{border-color:#024c7d;box-shadow:0 0 0 2px rgba(2,76,125,0.5)}html.theme--documenter-dark .button.is-success.is-hovered,html.theme--documenter-dark .button.is-success:hover{background-color:#00aa48}html.theme--documenter-dark .button.is-success.is-active,html.theme--documenter-dark .button.is-success.is-focused,html.theme--documenter-dark .button.is-success:active,html.theme--documenter-dark .button.is-success:focus{border-color:#008438;box-shadow:0 0 0 2px rgba(0,132,56,0.5)}html.theme--documenter-dark .button.is-warning.is-hovered,html.theme--documenter-dark .button.is-warning:hover{background-color:#d39e00}html.theme--documenter-dark .button.is-warning.is-active,html.theme--documenter-dark .button.is-warning.is-focused,html.theme--documenter-dark .button.is-warning:active,html.theme--documenter-dark .button.is-warning:focus{border-color:#ad8100;box-shadow:0 0 0 2px rgba(173,129,0,0.5)}html.theme--documenter-dark .button.is-danger.is-hovered,html.theme--documenter-dark .button.is-danger:hover{background-color:#c12110}html.theme--documenter-dark .button.is-danger.is-active,html.theme--documenter-dark .button.is-danger.is-focused,html.theme--documenter-dark .button.is-danger:active,html.theme--documenter-dark .button.is-danger:focus{border-color:#9e1b0d;box-shadow:0 0 0 2px rgba(158,27,13,0.5)}html.theme--documenter-dark .label{color:#dbdee0}html.theme--documenter-dark .button,html.theme--documenter-dark .control.has-icons-left .icon,html.theme--documenter-dark .control.has-icons-right .icon,html.theme--documenter-dark .input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark .pagination-ellipsis,html.theme--documenter-dark .pagination-link,html.theme--documenter-dark .pagination-next,html.theme--documenter-dark .pagination-previous,html.theme--documenter-dark .select,html.theme--documenter-dark .select select,html.theme--documenter-dark .textarea{height:2.5em}html.theme--documenter-dark .input,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark .textarea{transition:all 200ms ease;box-shadow:none;border-width:1px;padding-left:1em;padding-right:1em}html.theme--documenter-dark .select:after,html.theme--documenter-dark .select select{border-width:1px}html.theme--documenter-dark .control.has-addons .button,html.theme--documenter-dark .control.has-addons .input,html.theme--documenter-dark .control.has-addons #documenter .docs-sidebar form.docs-search>input,html.theme--documenter-dark #documenter .docs-sidebar .control.has-addons form.docs-search>input,html.theme--documenter-dark .control.has-addons .select{margin-right:-1px}html.theme--documenter-dark .notification{background-color:#343c3d}html.theme--documenter-dark .card{box-shadow:none;border:1px solid #343c3d;background-color:#282f2f;border-radius:.4em}html.theme--documenter-dark .card .card-image img{border-radius:.4em .4em 0 0}html.theme--documenter-dark .card .card-header{box-shadow:none;background-color:rgba(18,18,18,0.2);border-radius:.4em .4em 0 0}html.theme--documenter-dark .card .card-footer{background-color:rgba(18,18,18,0.2)}html.theme--documenter-dark .card .card-footer,html.theme--documenter-dark .card .card-footer-item{border-width:1px;border-color:#343c3d}html.theme--documenter-dark .notification.is-white a:not(.button){color:#0a0a0a;text-decoration:underline}html.theme--documenter-dark .notification.is-black a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .notification.is-light a:not(.button){color:rgba(0,0,0,0.7);text-decoration:underline}html.theme--documenter-dark .notification.is-dark a:not(.button),html.theme--documenter-dark .content kbd.notification a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .notification.is-primary a:not(.button),html.theme--documenter-dark .docstring>section>a.notification.docs-sourcelink a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .notification.is-link a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .notification.is-info a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .notification.is-success a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .notification.is-warning a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .notification.is-danger a:not(.button){color:#fff;text-decoration:underline}html.theme--documenter-dark .tag,html.theme--documenter-dark .content kbd,html.theme--documenter-dark .docstring>section>a.docs-sourcelink{border-radius:.4em}html.theme--documenter-dark .menu-list a{transition:all 300ms ease}html.theme--documenter-dark .modal-card-body{background-color:#282f2f}html.theme--documenter-dark .modal-card-foot,html.theme--documenter-dark .modal-card-head{border-color:#343c3d}html.theme--documenter-dark .message-header{font-weight:700;background-color:#343c3d;color:#fff}html.theme--documenter-dark .message-body{border-width:1px;border-color:#343c3d}html.theme--documenter-dark .navbar{border-radius:.4em}html.theme--documenter-dark .navbar.is-transparent{background:none}html.theme--documenter-dark .navbar.is-primary .navbar-dropdown a.navbar-item.is-active,html.theme--documenter-dark .docstring>section>a.navbar.docs-sourcelink .navbar-dropdown a.navbar-item.is-active{background-color:#1abc9c}@media screen and (max-width: 1055px){html.theme--documenter-dark .navbar .navbar-menu{background-color:#375a7f;border-radius:0 0 .4em .4em}}html.theme--documenter-dark .hero .navbar,html.theme--documenter-dark body>.navbar{border-radius:0}html.theme--documenter-dark .pagination-link,html.theme--documenter-dark .pagination-next,html.theme--documenter-dark .pagination-previous{border-width:1px}html.theme--documenter-dark .panel-block,html.theme--documenter-dark .panel-heading,html.theme--documenter-dark .panel-tabs{border-width:1px}html.theme--documenter-dark .panel-block:first-child,html.theme--documenter-dark .panel-heading:first-child,html.theme--documenter-dark .panel-tabs:first-child{border-top-width:1px}html.theme--documenter-dark .panel-heading{font-weight:700}html.theme--documenter-dark .panel-tabs a{border-width:1px;margin-bottom:-1px}html.theme--documenter-dark .panel-tabs a.is-active{border-bottom-color:#17a689}html.theme--documenter-dark .panel-block:hover{color:#1dd2af}html.theme--documenter-dark .panel-block:hover .panel-icon{color:#1dd2af}html.theme--documenter-dark .panel-block.is-active .panel-icon{color:#17a689}html.theme--documenter-dark .tabs a{border-bottom-width:1px;margin-bottom:-1px}html.theme--documenter-dark .tabs ul{border-bottom-width:1px}html.theme--documenter-dark .tabs.is-boxed a{border-width:1px}html.theme--documenter-dark .tabs.is-boxed li.is-active a{background-color:#1f2424}html.theme--documenter-dark .tabs.is-toggle li a{border-width:1px;margin-bottom:0}html.theme--documenter-dark .tabs.is-toggle li+li{margin-left:-1px}html.theme--documenter-dark .hero.is-white .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-black .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-light .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-dark .navbar .navbar-dropdown .navbar-item:hover,html.theme--documenter-dark .content kbd.hero .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-primary .navbar .navbar-dropdown .navbar-item:hover,html.theme--documenter-dark .docstring>section>a.hero.docs-sourcelink .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-link .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-info .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-success .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-warning .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark .hero.is-danger .navbar .navbar-dropdown .navbar-item:hover{background-color:rgba(0,0,0,0)}html.theme--documenter-dark h1 .docs-heading-anchor,html.theme--documenter-dark h1 .docs-heading-anchor:hover,html.theme--documenter-dark h1 .docs-heading-anchor:visited,html.theme--documenter-dark h2 .docs-heading-anchor,html.theme--documenter-dark h2 .docs-heading-anchor:hover,html.theme--documenter-dark h2 .docs-heading-anchor:visited,html.theme--documenter-dark h3 .docs-heading-anchor,html.theme--documenter-dark h3 .docs-heading-anchor:hover,html.theme--documenter-dark h3 .docs-heading-anchor:visited,html.theme--documenter-dark h4 .docs-heading-anchor,html.theme--documenter-dark h4 .docs-heading-anchor:hover,html.theme--documenter-dark h4 .docs-heading-anchor:visited,html.theme--documenter-dark h5 .docs-heading-anchor,html.theme--documenter-dark h5 .docs-heading-anchor:hover,html.theme--documenter-dark h5 .docs-heading-anchor:visited,html.theme--documenter-dark h6 .docs-heading-anchor,html.theme--documenter-dark h6 .docs-heading-anchor:hover,html.theme--documenter-dark h6 .docs-heading-anchor:visited{color:#f2f2f2}html.theme--documenter-dark h1 .docs-heading-anchor-permalink,html.theme--documenter-dark h2 .docs-heading-anchor-permalink,html.theme--documenter-dark h3 .docs-heading-anchor-permalink,html.theme--documenter-dark h4 .docs-heading-anchor-permalink,html.theme--documenter-dark h5 .docs-heading-anchor-permalink,html.theme--documenter-dark h6 .docs-heading-anchor-permalink{visibility:hidden;vertical-align:middle;margin-left:0.5em;font-size:0.7rem}html.theme--documenter-dark h1 .docs-heading-anchor-permalink::before,html.theme--documenter-dark h2 .docs-heading-anchor-permalink::before,html.theme--documenter-dark h3 .docs-heading-anchor-permalink::before,html.theme--documenter-dark h4 .docs-heading-anchor-permalink::before,html.theme--documenter-dark h5 .docs-heading-anchor-permalink::before,html.theme--documenter-dark h6 .docs-heading-anchor-permalink::before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f0c1"}html.theme--documenter-dark h1:hover .docs-heading-anchor-permalink,html.theme--documenter-dark h2:hover .docs-heading-anchor-permalink,html.theme--documenter-dark h3:hover .docs-heading-anchor-permalink,html.theme--documenter-dark h4:hover .docs-heading-anchor-permalink,html.theme--documenter-dark h5:hover .docs-heading-anchor-permalink,html.theme--documenter-dark h6:hover .docs-heading-anchor-permalink{visibility:visible}html.theme--documenter-dark .docs-light-only{display:none !important}html.theme--documenter-dark pre{position:relative;overflow:hidden}html.theme--documenter-dark pre code,html.theme--documenter-dark pre code.hljs{padding:0 .75rem !important;overflow:auto;display:block}html.theme--documenter-dark pre code:first-of-type,html.theme--documenter-dark pre code.hljs:first-of-type{padding-top:0.5rem !important}html.theme--documenter-dark pre code:last-of-type,html.theme--documenter-dark pre code.hljs:last-of-type{padding-bottom:0.5rem !important}html.theme--documenter-dark pre .copy-button{opacity:0.2;transition:opacity 0.2s;position:absolute;right:0em;top:0em;padding:0.5em;width:2.5em;height:2.5em;background:transparent;border:none;font-family:"Font Awesome 6 Free";color:#fff;cursor:pointer;text-align:center}html.theme--documenter-dark pre .copy-button:focus,html.theme--documenter-dark pre .copy-button:hover{opacity:1;background:rgba(255,255,255,0.1);color:#1abc9c}html.theme--documenter-dark pre .copy-button.success{color:#259a12;opacity:1}html.theme--documenter-dark pre .copy-button.error{color:#cb3c33;opacity:1}html.theme--documenter-dark pre:hover .copy-button{opacity:1}html.theme--documenter-dark .admonition{background-color:#282f2f;border-style:solid;border-width:1px;border-color:#5e6d6f;border-radius:.4em;font-size:1rem}html.theme--documenter-dark .admonition strong{color:currentColor}html.theme--documenter-dark .admonition.is-small,html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input.admonition{font-size:.75rem}html.theme--documenter-dark .admonition.is-medium{font-size:1.25rem}html.theme--documenter-dark .admonition.is-large{font-size:1.5rem}html.theme--documenter-dark .admonition.is-default{background-color:#282f2f;border-color:#5e6d6f}html.theme--documenter-dark .admonition.is-default>.admonition-header{background-color:#5e6d6f;color:#fff}html.theme--documenter-dark .admonition.is-default>.admonition-body{color:#fff}html.theme--documenter-dark .admonition.is-info{background-color:#282f2f;border-color:#024c7d}html.theme--documenter-dark .admonition.is-info>.admonition-header{background-color:#024c7d;color:#fff}html.theme--documenter-dark .admonition.is-info>.admonition-body{color:#fff}html.theme--documenter-dark .admonition.is-success{background-color:#282f2f;border-color:#008438}html.theme--documenter-dark .admonition.is-success>.admonition-header{background-color:#008438;color:#fff}html.theme--documenter-dark .admonition.is-success>.admonition-body{color:#fff}html.theme--documenter-dark .admonition.is-warning{background-color:#282f2f;border-color:#ad8100}html.theme--documenter-dark .admonition.is-warning>.admonition-header{background-color:#ad8100;color:#fff}html.theme--documenter-dark .admonition.is-warning>.admonition-body{color:#fff}html.theme--documenter-dark .admonition.is-danger{background-color:#282f2f;border-color:#9e1b0d}html.theme--documenter-dark .admonition.is-danger>.admonition-header{background-color:#9e1b0d;color:#fff}html.theme--documenter-dark .admonition.is-danger>.admonition-body{color:#fff}html.theme--documenter-dark .admonition.is-compat{background-color:#282f2f;border-color:#137886}html.theme--documenter-dark .admonition.is-compat>.admonition-header{background-color:#137886;color:#fff}html.theme--documenter-dark .admonition.is-compat>.admonition-body{color:#fff}html.theme--documenter-dark .admonition-header{color:#fff;background-color:#5e6d6f;align-items:center;font-weight:700;justify-content:space-between;line-height:1.25;padding:0.5rem .75rem;position:relative}html.theme--documenter-dark .admonition-header:before{font-family:"Font Awesome 6 Free";font-weight:900;margin-right:.75rem;content:"\f06a"}html.theme--documenter-dark details.admonition.is-details>.admonition-header{list-style:none}html.theme--documenter-dark details.admonition.is-details>.admonition-header:before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f055"}html.theme--documenter-dark details.admonition.is-details[open]>.admonition-header:before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f056"}html.theme--documenter-dark .admonition-body{color:#fff;padding:0.5rem .75rem}html.theme--documenter-dark .admonition-body pre{background-color:#282f2f}html.theme--documenter-dark .admonition-body code{background-color:rgba(255,255,255,0.05)}html.theme--documenter-dark .docstring{margin-bottom:1em;background-color:rgba(0,0,0,0);border:1px solid #5e6d6f;box-shadow:none;max-width:100%}html.theme--documenter-dark .docstring>header{cursor:pointer;display:flex;flex-grow:1;align-items:stretch;padding:0.5rem .75rem;background-color:#282f2f;box-shadow:0 0.125em 0.25em rgba(10,10,10,0.1);box-shadow:none;border-bottom:1px solid #5e6d6f}html.theme--documenter-dark .docstring>header code{background-color:transparent}html.theme--documenter-dark .docstring>header .docstring-article-toggle-button{min-width:1.1rem;padding:0.2rem 0.2rem 0.2rem 0}html.theme--documenter-dark .docstring>header .docstring-binding{margin-right:0.3em}html.theme--documenter-dark .docstring>header .docstring-category{margin-left:0.3em}html.theme--documenter-dark .docstring>section{position:relative;padding:.75rem .75rem;border-bottom:1px solid #5e6d6f}html.theme--documenter-dark .docstring>section:last-child{border-bottom:none}html.theme--documenter-dark .docstring>section>a.docs-sourcelink{transition:opacity 0.3s;opacity:0;position:absolute;right:.375rem;bottom:.375rem}html.theme--documenter-dark .docstring>section>a.docs-sourcelink:focus{opacity:1 !important}html.theme--documenter-dark .docstring:hover>section>a.docs-sourcelink{opacity:0.2}html.theme--documenter-dark .docstring:focus-within>section>a.docs-sourcelink{opacity:0.2}html.theme--documenter-dark .docstring>section:hover a.docs-sourcelink{opacity:1}html.theme--documenter-dark .documenter-example-output{background-color:#1f2424}html.theme--documenter-dark .outdated-warning-overlay{position:fixed;top:0;left:0;right:0;box-shadow:0 0 10px rgba(0,0,0,0.3);z-index:999;background-color:#282f2f;color:#fff;border-bottom:3px solid #9e1b0d;padding:10px 35px;text-align:center;font-size:15px}html.theme--documenter-dark .outdated-warning-overlay .outdated-warning-closer{position:absolute;top:calc(50% - 10px);right:18px;cursor:pointer;width:12px}html.theme--documenter-dark .outdated-warning-overlay a{color:#1abc9c}html.theme--documenter-dark .outdated-warning-overlay a:hover{color:#1dd2af}html.theme--documenter-dark .content pre{border:1px solid #5e6d6f}html.theme--documenter-dark .content code{font-weight:inherit}html.theme--documenter-dark .content a code{color:#1abc9c}html.theme--documenter-dark .content h1 code,html.theme--documenter-dark .content h2 code,html.theme--documenter-dark .content h3 code,html.theme--documenter-dark .content h4 code,html.theme--documenter-dark .content h5 code,html.theme--documenter-dark .content h6 code{color:#f2f2f2}html.theme--documenter-dark .content table{display:block;width:initial;max-width:100%;overflow-x:auto}html.theme--documenter-dark .content blockquote>ul:first-child,html.theme--documenter-dark .content blockquote>ol:first-child,html.theme--documenter-dark .content .admonition-body>ul:first-child,html.theme--documenter-dark .content .admonition-body>ol:first-child{margin-top:0}html.theme--documenter-dark pre,html.theme--documenter-dark code{font-variant-ligatures:no-contextual}html.theme--documenter-dark .breadcrumb a.is-disabled{cursor:default;pointer-events:none}html.theme--documenter-dark .breadcrumb a.is-disabled,html.theme--documenter-dark .breadcrumb a.is-disabled:hover{color:#f2f2f2}html.theme--documenter-dark .hljs{background:initial !important}html.theme--documenter-dark .katex .katex-mathml{top:0;right:0}html.theme--documenter-dark .katex-display,html.theme--documenter-dark mjx-container,html.theme--documenter-dark .MathJax_Display{margin:0.5em 0 !important}html.theme--documenter-dark html{-moz-osx-font-smoothing:auto;-webkit-font-smoothing:auto}html.theme--documenter-dark li.no-marker{list-style:none}html.theme--documenter-dark #documenter .docs-main>article{overflow-wrap:break-word}html.theme--documenter-dark #documenter .docs-main>article .math-container{overflow-x:auto;overflow-y:hidden}@media screen and (min-width: 1056px){html.theme--documenter-dark #documenter .docs-main{max-width:52rem;margin-left:20rem;padding-right:1rem}}@media screen and (max-width: 1055px){html.theme--documenter-dark #documenter .docs-main{width:100%}html.theme--documenter-dark #documenter .docs-main>article{max-width:52rem;margin-left:auto;margin-right:auto;margin-bottom:1rem;padding:0 1rem}html.theme--documenter-dark #documenter .docs-main>header,html.theme--documenter-dark #documenter .docs-main>nav{max-width:100%;width:100%;margin:0}}html.theme--documenter-dark #documenter .docs-main header.docs-navbar{background-color:#1f2424;border-bottom:1px solid #5e6d6f;z-index:2;min-height:4rem;margin-bottom:1rem;display:flex}html.theme--documenter-dark #documenter .docs-main header.docs-navbar .breadcrumb{flex-grow:1}html.theme--documenter-dark #documenter .docs-main header.docs-navbar .docs-sidebar-button{display:block;font-size:1.5rem;padding-bottom:0.1rem;margin-right:1rem}html.theme--documenter-dark #documenter .docs-main header.docs-navbar .docs-right{display:flex;white-space:nowrap;gap:1rem;align-items:center}html.theme--documenter-dark #documenter .docs-main header.docs-navbar .docs-right .docs-icon,html.theme--documenter-dark #documenter .docs-main header.docs-navbar .docs-right .docs-label{display:inline-block}html.theme--documenter-dark #documenter .docs-main header.docs-navbar .docs-right .docs-label{padding:0;margin-left:0.3em}@media screen and (max-width: 1055px){html.theme--documenter-dark #documenter .docs-main header.docs-navbar .docs-right .docs-navbar-link{margin-left:0.4rem;margin-right:0.4rem}}html.theme--documenter-dark #documenter .docs-main header.docs-navbar>*{margin:auto 0}@media screen and (max-width: 1055px){html.theme--documenter-dark #documenter .docs-main header.docs-navbar{position:sticky;top:0;padding:0 1rem;transition-property:top, box-shadow;-webkit-transition-property:top, box-shadow;transition-duration:0.3s;-webkit-transition-duration:0.3s}html.theme--documenter-dark #documenter .docs-main header.docs-navbar.headroom--not-top{box-shadow:.2rem 0rem .4rem #171717;transition-duration:0.7s;-webkit-transition-duration:0.7s}html.theme--documenter-dark #documenter .docs-main header.docs-navbar.headroom--unpinned.headroom--not-top.headroom--not-bottom{top:-4.5rem;transition-duration:0.7s;-webkit-transition-duration:0.7s}}html.theme--documenter-dark #documenter .docs-main section.footnotes{border-top:1px solid #5e6d6f}html.theme--documenter-dark #documenter .docs-main section.footnotes li .tag:first-child,html.theme--documenter-dark #documenter .docs-main section.footnotes li .docstring>section>a.docs-sourcelink:first-child,html.theme--documenter-dark #documenter .docs-main section.footnotes li .content kbd:first-child,html.theme--documenter-dark .content #documenter .docs-main section.footnotes li kbd:first-child{margin-right:1em;margin-bottom:0.4em}html.theme--documenter-dark #documenter .docs-main .docs-footer{display:flex;flex-wrap:wrap;margin-left:0;margin-right:0;border-top:1px solid #5e6d6f;padding-top:1rem;padding-bottom:1rem}@media screen and (max-width: 1055px){html.theme--documenter-dark #documenter .docs-main .docs-footer{padding-left:1rem;padding-right:1rem}}html.theme--documenter-dark #documenter .docs-main .docs-footer .docs-footer-nextpage,html.theme--documenter-dark #documenter .docs-main .docs-footer .docs-footer-prevpage{flex-grow:1}html.theme--documenter-dark #documenter .docs-main .docs-footer .docs-footer-nextpage{text-align:right}html.theme--documenter-dark #documenter .docs-main .docs-footer .flexbox-break{flex-basis:100%;height:0}html.theme--documenter-dark #documenter .docs-main .docs-footer .footer-message{font-size:0.8em;margin:0.5em auto 0 auto;text-align:center}html.theme--documenter-dark #documenter .docs-sidebar{display:flex;flex-direction:column;color:#fff;background-color:#282f2f;border-right:1px solid #5e6d6f;padding:0;flex:0 0 18rem;z-index:5;font-size:1rem;position:fixed;left:-18rem;width:18rem;height:100%;transition:left 0.3s}html.theme--documenter-dark #documenter .docs-sidebar.visible{left:0;box-shadow:.4rem 0rem .8rem #171717}@media screen and (min-width: 1056px){html.theme--documenter-dark #documenter .docs-sidebar.visible{box-shadow:none}}@media screen and (min-width: 1056px){html.theme--documenter-dark #documenter .docs-sidebar{left:0;top:0}}html.theme--documenter-dark #documenter .docs-sidebar .docs-logo{margin-top:1rem;padding:0 1rem}html.theme--documenter-dark #documenter .docs-sidebar .docs-logo>img{max-height:6rem;margin:auto}html.theme--documenter-dark #documenter .docs-sidebar .docs-package-name{flex-shrink:0;font-size:1.5rem;font-weight:700;text-align:center;white-space:nowrap;overflow:hidden;padding:0.5rem 0}html.theme--documenter-dark #documenter .docs-sidebar .docs-package-name .docs-autofit{max-width:16.2rem}html.theme--documenter-dark #documenter .docs-sidebar .docs-package-name a,html.theme--documenter-dark #documenter .docs-sidebar .docs-package-name a:hover{color:#fff}html.theme--documenter-dark #documenter .docs-sidebar .docs-version-selector{border-top:1px solid #5e6d6f;display:none;padding:0.5rem}html.theme--documenter-dark #documenter .docs-sidebar .docs-version-selector.visible{display:flex}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu{flex-grow:1;user-select:none;border-top:1px solid #5e6d6f;padding-bottom:1.5rem}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu>li>.tocitem{font-weight:bold}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu>li li{font-size:.95rem;margin-left:1em;border-left:1px solid #5e6d6f}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu input.collapse-toggle{display:none}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu ul.collapsed{display:none}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu input:checked~ul.collapsed{display:block}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu label.tocitem{display:flex}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu label.tocitem .docs-label{flex-grow:2}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu label.tocitem .docs-chevron{display:inline-block;font-style:normal;font-variant:normal;text-rendering:auto;line-height:1;font-size:.75rem;margin-left:1rem;margin-top:auto;margin-bottom:auto}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu label.tocitem .docs-chevron::before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f054"}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu input:checked~label.tocitem .docs-chevron::before{content:"\f078"}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu .tocitem{display:block;padding:0.5rem 0.5rem}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu .tocitem,html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu .tocitem:hover{color:#fff;background:#282f2f}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu a.tocitem:hover,html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu label.tocitem:hover{color:#fff;background-color:#32393a}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu li.is-active{border-top:1px solid #5e6d6f;border-bottom:1px solid #5e6d6f;background-color:#1f2424}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu li.is-active .tocitem,html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu li.is-active .tocitem:hover{background-color:#1f2424;color:#fff}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu li.is-active ul.internal .tocitem:hover{background-color:#32393a;color:#fff}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu>li.is-active:first-child{border-top:none}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu ul.internal{margin:0 0.5rem 0.5rem;border-top:1px solid #5e6d6f}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu ul.internal li{font-size:.85rem;border-left:none;margin-left:0;margin-top:0.5rem}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu ul.internal .tocitem{width:100%;padding:0}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu ul.internal .tocitem::before{content:"⚬";margin-right:0.4em}html.theme--documenter-dark #documenter .docs-sidebar form.docs-search{margin:auto;margin-top:0.5rem;margin-bottom:0.5rem}html.theme--documenter-dark #documenter .docs-sidebar form.docs-search>input{width:14.4rem}html.theme--documenter-dark #documenter .docs-sidebar #documenter-search-query{color:#868c98;width:14.4rem;box-shadow:inset 0 1px 2px rgba(10,10,10,0.1)}@media screen and (min-width: 1056px){html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu{overflow-y:auto;-webkit-overflow-scroll:touch}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu::-webkit-scrollbar{width:.3rem;background:none}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu::-webkit-scrollbar-thumb{border-radius:5px 0px 0px 5px;background:#3b4445}html.theme--documenter-dark #documenter .docs-sidebar ul.docs-menu::-webkit-scrollbar-thumb:hover{background:#4e5a5c}}@media screen and (max-width: 1055px){html.theme--documenter-dark #documenter .docs-sidebar{overflow-y:auto;-webkit-overflow-scroll:touch}html.theme--documenter-dark #documenter .docs-sidebar::-webkit-scrollbar{width:.3rem;background:none}html.theme--documenter-dark #documenter .docs-sidebar::-webkit-scrollbar-thumb{border-radius:5px 0px 0px 5px;background:#3b4445}html.theme--documenter-dark #documenter .docs-sidebar::-webkit-scrollbar-thumb:hover{background:#4e5a5c}}html.theme--documenter-dark kbd.search-modal-key-hints{border-radius:0.25rem;border:1px solid rgba(245,245,245,0.6);box-shadow:0 2px 0 1px rgba(245,245,245,0.6);cursor:default;font-size:0.9rem;line-height:1.5;min-width:0.75rem;text-align:center;padding:0.1rem 0.3rem;position:relative;top:-1px}html.theme--documenter-dark .search-min-width-50{min-width:50%}html.theme--documenter-dark .search-min-height-100{min-height:100%}html.theme--documenter-dark .search-modal-card-body{max-height:calc(100vh - 15rem)}html.theme--documenter-dark .search-result-link{border-radius:0.7em;transition:all 300ms}html.theme--documenter-dark .search-result-link:hover,html.theme--documenter-dark .search-result-link:focus{background-color:rgba(0,128,128,0.1)}html.theme--documenter-dark .search-result-link .property-search-result-badge,html.theme--documenter-dark .search-result-link .search-filter{transition:all 300ms}html.theme--documenter-dark .property-search-result-badge,html.theme--documenter-dark .search-filter{padding:0.15em 0.5em;font-size:0.8em;font-style:italic;text-transform:none !important;line-height:1.5;color:#f5f5f5;background-color:rgba(51,65,85,0.501961);border-radius:0.6rem}html.theme--documenter-dark .search-result-link:hover .property-search-result-badge,html.theme--documenter-dark .search-result-link:hover .search-filter,html.theme--documenter-dark .search-result-link:focus .property-search-result-badge,html.theme--documenter-dark .search-result-link:focus .search-filter{color:#333;background-color:#f1f5f9}html.theme--documenter-dark .search-filter{color:#333;background-color:#f5f5f5;transition:all 300ms}html.theme--documenter-dark .search-filter:hover,html.theme--documenter-dark .search-filter:focus{color:#333}html.theme--documenter-dark .search-filter-selected{color:#f5f5f5;background-color:rgba(139,0,139,0.5)}html.theme--documenter-dark .search-filter-selected:hover,html.theme--documenter-dark .search-filter-selected:focus{color:#f5f5f5}html.theme--documenter-dark .search-result-highlight{background-color:#ffdd57;color:black}html.theme--documenter-dark .search-divider{border-bottom:1px solid #5e6d6f}html.theme--documenter-dark .search-result-title{width:85%;color:#f5f5f5}html.theme--documenter-dark .search-result-code-title{font-size:0.875rem;font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace}html.theme--documenter-dark #search-modal .modal-card-body::-webkit-scrollbar,html.theme--documenter-dark #search-modal .filter-tabs::-webkit-scrollbar{height:10px;width:10px;background-color:transparent}html.theme--documenter-dark #search-modal .modal-card-body::-webkit-scrollbar-thumb,html.theme--documenter-dark #search-modal .filter-tabs::-webkit-scrollbar-thumb{background-color:gray;border-radius:1rem}html.theme--documenter-dark #search-modal .modal-card-body::-webkit-scrollbar-track,html.theme--documenter-dark #search-modal .filter-tabs::-webkit-scrollbar-track{-webkit-box-shadow:inset 0 0 6px rgba(0,0,0,0.6);background-color:transparent}html.theme--documenter-dark .w-100{width:100%}html.theme--documenter-dark .gap-2{gap:0.5rem}html.theme--documenter-dark .gap-4{gap:1rem}html.theme--documenter-dark .gap-8{gap:2rem}html.theme--documenter-dark{background-color:#1f2424;font-size:16px;min-width:300px;overflow-x:auto;overflow-y:scroll;text-rendering:optimizeLegibility;text-size-adjust:100%}html.theme--documenter-dark .ansi span.sgr1{font-weight:bolder}html.theme--documenter-dark .ansi span.sgr2{font-weight:lighter}html.theme--documenter-dark .ansi span.sgr3{font-style:italic}html.theme--documenter-dark .ansi span.sgr4{text-decoration:underline}html.theme--documenter-dark .ansi span.sgr7{color:#1f2424;background-color:#fff}html.theme--documenter-dark .ansi span.sgr8{color:transparent}html.theme--documenter-dark .ansi span.sgr8 span{color:transparent}html.theme--documenter-dark .ansi span.sgr9{text-decoration:line-through}html.theme--documenter-dark .ansi span.sgr30{color:#242424}html.theme--documenter-dark .ansi span.sgr31{color:#f6705f}html.theme--documenter-dark .ansi span.sgr32{color:#4fb43a}html.theme--documenter-dark .ansi span.sgr33{color:#f4c72f}html.theme--documenter-dark .ansi span.sgr34{color:#7587f0}html.theme--documenter-dark .ansi span.sgr35{color:#bc89d3}html.theme--documenter-dark .ansi span.sgr36{color:#49b6ca}html.theme--documenter-dark .ansi span.sgr37{color:#b3bdbe}html.theme--documenter-dark .ansi span.sgr40{background-color:#242424}html.theme--documenter-dark .ansi span.sgr41{background-color:#f6705f}html.theme--documenter-dark .ansi span.sgr42{background-color:#4fb43a}html.theme--documenter-dark .ansi span.sgr43{background-color:#f4c72f}html.theme--documenter-dark .ansi span.sgr44{background-color:#7587f0}html.theme--documenter-dark .ansi span.sgr45{background-color:#bc89d3}html.theme--documenter-dark .ansi span.sgr46{background-color:#49b6ca}html.theme--documenter-dark .ansi span.sgr47{background-color:#b3bdbe}html.theme--documenter-dark .ansi span.sgr90{color:#92a0a2}html.theme--documenter-dark .ansi span.sgr91{color:#ff8674}html.theme--documenter-dark .ansi span.sgr92{color:#79d462}html.theme--documenter-dark .ansi span.sgr93{color:#ffe76b}html.theme--documenter-dark .ansi span.sgr94{color:#8a98ff}html.theme--documenter-dark .ansi span.sgr95{color:#d2a4e6}html.theme--documenter-dark .ansi span.sgr96{color:#6bc8db}html.theme--documenter-dark .ansi span.sgr97{color:#ecf0f1}html.theme--documenter-dark .ansi span.sgr100{background-color:#92a0a2}html.theme--documenter-dark .ansi span.sgr101{background-color:#ff8674}html.theme--documenter-dark .ansi span.sgr102{background-color:#79d462}html.theme--documenter-dark .ansi span.sgr103{background-color:#ffe76b}html.theme--documenter-dark .ansi span.sgr104{background-color:#8a98ff}html.theme--documenter-dark .ansi span.sgr105{background-color:#d2a4e6}html.theme--documenter-dark .ansi span.sgr106{background-color:#6bc8db}html.theme--documenter-dark .ansi span.sgr107{background-color:#ecf0f1}html.theme--documenter-dark code.language-julia-repl>span.hljs-meta{color:#4fb43a;font-weight:bolder}html.theme--documenter-dark .hljs{background:#2b2b2b;color:#f8f8f2}html.theme--documenter-dark .hljs-comment,html.theme--documenter-dark .hljs-quote{color:#d4d0ab}html.theme--documenter-dark .hljs-variable,html.theme--documenter-dark .hljs-template-variable,html.theme--documenter-dark .hljs-tag,html.theme--documenter-dark .hljs-name,html.theme--documenter-dark .hljs-selector-id,html.theme--documenter-dark .hljs-selector-class,html.theme--documenter-dark .hljs-regexp,html.theme--documenter-dark .hljs-deletion{color:#ffa07a}html.theme--documenter-dark .hljs-number,html.theme--documenter-dark .hljs-built_in,html.theme--documenter-dark .hljs-literal,html.theme--documenter-dark .hljs-type,html.theme--documenter-dark .hljs-params,html.theme--documenter-dark .hljs-meta,html.theme--documenter-dark .hljs-link{color:#f5ab35}html.theme--documenter-dark .hljs-attribute{color:#ffd700}html.theme--documenter-dark .hljs-string,html.theme--documenter-dark .hljs-symbol,html.theme--documenter-dark .hljs-bullet,html.theme--documenter-dark .hljs-addition{color:#abe338}html.theme--documenter-dark .hljs-title,html.theme--documenter-dark .hljs-section{color:#00e0e0}html.theme--documenter-dark .hljs-keyword,html.theme--documenter-dark .hljs-selector-tag{color:#dcc6e0}html.theme--documenter-dark .hljs-emphasis{font-style:italic}html.theme--documenter-dark .hljs-strong{font-weight:bold}@media screen and (-ms-high-contrast: active){html.theme--documenter-dark .hljs-addition,html.theme--documenter-dark .hljs-attribute,html.theme--documenter-dark .hljs-built_in,html.theme--documenter-dark .hljs-bullet,html.theme--documenter-dark .hljs-comment,html.theme--documenter-dark .hljs-link,html.theme--documenter-dark .hljs-literal,html.theme--documenter-dark .hljs-meta,html.theme--documenter-dark .hljs-number,html.theme--documenter-dark .hljs-params,html.theme--documenter-dark .hljs-string,html.theme--documenter-dark .hljs-symbol,html.theme--documenter-dark .hljs-type,html.theme--documenter-dark .hljs-quote{color:highlight}html.theme--documenter-dark .hljs-keyword,html.theme--documenter-dark .hljs-selector-tag{font-weight:bold}}html.theme--documenter-dark .hljs-subst{color:#f8f8f2}
diff --git a/previews/PR3525/assets/themes/documenter-light.css b/previews/PR3525/assets/themes/documenter-light.css
new file mode 100644
index 00000000000..60a317a4c53
--- /dev/null
+++ b/previews/PR3525/assets/themes/documenter-light.css
@@ -0,0 +1,9 @@
+.pagination-previous,.pagination-next,.pagination-link,.pagination-ellipsis,.file-cta,.file-name,.select select,.textarea,.input,#documenter .docs-sidebar form.docs-search>input,.button{-moz-appearance:none;-webkit-appearance:none;align-items:center;border:1px solid transparent;border-radius:4px;box-shadow:none;display:inline-flex;font-size:1rem;height:2.5em;justify-content:flex-start;line-height:1.5;padding-bottom:calc(0.5em - 1px);padding-left:calc(0.75em - 1px);padding-right:calc(0.75em - 1px);padding-top:calc(0.5em - 1px);position:relative;vertical-align:top}.pagination-previous:focus,.pagination-next:focus,.pagination-link:focus,.pagination-ellipsis:focus,.file-cta:focus,.file-name:focus,.select select:focus,.textarea:focus,.input:focus,#documenter .docs-sidebar form.docs-search>input:focus,.button:focus,.is-focused.pagination-previous,.is-focused.pagination-next,.is-focused.pagination-link,.is-focused.pagination-ellipsis,.is-focused.file-cta,.is-focused.file-name,.select select.is-focused,.is-focused.textarea,.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-focused.button,.pagination-previous:active,.pagination-next:active,.pagination-link:active,.pagination-ellipsis:active,.file-cta:active,.file-name:active,.select select:active,.textarea:active,.input:active,#documenter .docs-sidebar form.docs-search>input:active,.button:active,.is-active.pagination-previous,.is-active.pagination-next,.is-active.pagination-link,.is-active.pagination-ellipsis,.is-active.file-cta,.is-active.file-name,.select select.is-active,.is-active.textarea,.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active,.is-active.button{outline:none}.pagination-previous[disabled],.pagination-next[disabled],.pagination-link[disabled],.pagination-ellipsis[disabled],.file-cta[disabled],.file-name[disabled],.select select[disabled],.textarea[disabled],.input[disabled],#documenter .docs-sidebar form.docs-search>input[disabled],.button[disabled],fieldset[disabled] .pagination-previous,fieldset[disabled] .pagination-next,fieldset[disabled] .pagination-link,fieldset[disabled] .pagination-ellipsis,fieldset[disabled] .file-cta,fieldset[disabled] .file-name,fieldset[disabled] .select select,.select fieldset[disabled] select,fieldset[disabled] .textarea,fieldset[disabled] .input,fieldset[disabled] #documenter .docs-sidebar form.docs-search>input,#documenter .docs-sidebar fieldset[disabled] form.docs-search>input,fieldset[disabled] .button{cursor:not-allowed}.tabs,.pagination-previous,.pagination-next,.pagination-link,.pagination-ellipsis,.breadcrumb,.file,.button,.is-unselectable{-webkit-touch-callout:none;-webkit-user-select:none;-moz-user-select:none;-ms-user-select:none;user-select:none}.navbar-link:not(.is-arrowless)::after,.select:not(.is-multiple):not(.is-loading)::after{border:3px solid rgba(0,0,0,0);border-radius:2px;border-right:0;border-top:0;content:" ";display:block;height:0.625em;margin-top:-0.4375em;pointer-events:none;position:absolute;top:50%;transform:rotate(-45deg);transform-origin:center;width:0.625em}.admonition:not(:last-child),.tabs:not(:last-child),.pagination:not(:last-child),.message:not(:last-child),.level:not(:last-child),.breadcrumb:not(:last-child),.block:not(:last-child),.title:not(:last-child),.subtitle:not(:last-child),.table-container:not(:last-child),.table:not(:last-child),.progress:not(:last-child),.notification:not(:last-child),.content:not(:last-child),.box:not(:last-child){margin-bottom:1.5rem}.modal-close,.delete{-webkit-touch-callout:none;-webkit-user-select:none;-moz-user-select:none;-ms-user-select:none;user-select:none;-moz-appearance:none;-webkit-appearance:none;background-color:rgba(10,10,10,0.2);border:none;border-radius:9999px;cursor:pointer;pointer-events:auto;display:inline-block;flex-grow:0;flex-shrink:0;font-size:0;height:20px;max-height:20px;max-width:20px;min-height:20px;min-width:20px;outline:none;position:relative;vertical-align:top;width:20px}.modal-close::before,.delete::before,.modal-close::after,.delete::after{background-color:#fff;content:"";display:block;left:50%;position:absolute;top:50%;transform:translateX(-50%) translateY(-50%) rotate(45deg);transform-origin:center center}.modal-close::before,.delete::before{height:2px;width:50%}.modal-close::after,.delete::after{height:50%;width:2px}.modal-close:hover,.delete:hover,.modal-close:focus,.delete:focus{background-color:rgba(10,10,10,0.3)}.modal-close:active,.delete:active{background-color:rgba(10,10,10,0.4)}.is-small.modal-close,#documenter .docs-sidebar form.docs-search>input.modal-close,.is-small.delete,#documenter .docs-sidebar form.docs-search>input.delete{height:16px;max-height:16px;max-width:16px;min-height:16px;min-width:16px;width:16px}.is-medium.modal-close,.is-medium.delete{height:24px;max-height:24px;max-width:24px;min-height:24px;min-width:24px;width:24px}.is-large.modal-close,.is-large.delete{height:32px;max-height:32px;max-width:32px;min-height:32px;min-width:32px;width:32px}.control.is-loading::after,.select.is-loading::after,.loader,.button.is-loading::after{animation:spinAround 500ms infinite linear;border:2px solid #dbdbdb;border-radius:9999px;border-right-color:transparent;border-top-color:transparent;content:"";display:block;height:1em;position:relative;width:1em}.hero-video,.modal-background,.modal,.image.is-square img,#documenter .docs-sidebar .docs-logo>img.is-square img,.image.is-square .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-square .has-ratio,.image.is-1by1 img,#documenter .docs-sidebar .docs-logo>img.is-1by1 img,.image.is-1by1 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-1by1 .has-ratio,.image.is-5by4 img,#documenter .docs-sidebar .docs-logo>img.is-5by4 img,.image.is-5by4 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-5by4 .has-ratio,.image.is-4by3 img,#documenter .docs-sidebar .docs-logo>img.is-4by3 img,.image.is-4by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-4by3 .has-ratio,.image.is-3by2 img,#documenter .docs-sidebar .docs-logo>img.is-3by2 img,.image.is-3by2 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by2 .has-ratio,.image.is-5by3 img,#documenter .docs-sidebar .docs-logo>img.is-5by3 img,.image.is-5by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-5by3 .has-ratio,.image.is-16by9 img,#documenter .docs-sidebar .docs-logo>img.is-16by9 img,.image.is-16by9 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-16by9 .has-ratio,.image.is-2by1 img,#documenter .docs-sidebar .docs-logo>img.is-2by1 img,.image.is-2by1 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-2by1 .has-ratio,.image.is-3by1 img,#documenter .docs-sidebar .docs-logo>img.is-3by1 img,.image.is-3by1 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by1 .has-ratio,.image.is-4by5 img,#documenter .docs-sidebar .docs-logo>img.is-4by5 img,.image.is-4by5 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-4by5 .has-ratio,.image.is-3by4 img,#documenter .docs-sidebar .docs-logo>img.is-3by4 img,.image.is-3by4 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by4 .has-ratio,.image.is-2by3 img,#documenter .docs-sidebar .docs-logo>img.is-2by3 img,.image.is-2by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-2by3 .has-ratio,.image.is-3by5 img,#documenter .docs-sidebar .docs-logo>img.is-3by5 img,.image.is-3by5 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by5 .has-ratio,.image.is-9by16 img,#documenter .docs-sidebar .docs-logo>img.is-9by16 img,.image.is-9by16 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-9by16 .has-ratio,.image.is-1by2 img,#documenter .docs-sidebar .docs-logo>img.is-1by2 img,.image.is-1by2 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-1by2 .has-ratio,.image.is-1by3 img,#documenter .docs-sidebar .docs-logo>img.is-1by3 img,.image.is-1by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-1by3 .has-ratio,.is-overlay{bottom:0;left:0;position:absolute;right:0;top:0}.navbar-burger{-moz-appearance:none;-webkit-appearance:none;appearance:none;background:none;border:none;color:currentColor;font-family:inherit;font-size:1em;margin:0;padding:0}.has-text-white{color:#fff !important}a.has-text-white:hover,a.has-text-white:focus{color:#e6e6e6 !important}.has-background-white{background-color:#fff !important}.has-text-black{color:#0a0a0a !important}a.has-text-black:hover,a.has-text-black:focus{color:#000 !important}.has-background-black{background-color:#0a0a0a !important}.has-text-light{color:#f5f5f5 !important}a.has-text-light:hover,a.has-text-light:focus{color:#dbdbdb !important}.has-background-light{background-color:#f5f5f5 !important}.has-text-dark{color:#363636 !important}a.has-text-dark:hover,a.has-text-dark:focus{color:#1c1c1c !important}.has-background-dark{background-color:#363636 !important}.has-text-primary{color:#4eb5de !important}a.has-text-primary:hover,a.has-text-primary:focus{color:#27a1d2 !important}.has-background-primary{background-color:#4eb5de !important}.has-text-primary-light{color:#eef8fc !important}a.has-text-primary-light:hover,a.has-text-primary-light:focus{color:#c3e6f4 !important}.has-background-primary-light{background-color:#eef8fc !important}.has-text-primary-dark{color:#1a6d8e !important}a.has-text-primary-dark:hover,a.has-text-primary-dark:focus{color:#228eb9 !important}.has-background-primary-dark{background-color:#1a6d8e !important}.has-text-link{color:#2e63b8 !important}a.has-text-link:hover,a.has-text-link:focus{color:#244d8f !important}.has-background-link{background-color:#2e63b8 !important}.has-text-link-light{color:#eff3fb !important}a.has-text-link-light:hover,a.has-text-link-light:focus{color:#c6d6f1 !important}.has-background-link-light{background-color:#eff3fb !important}.has-text-link-dark{color:#3169c4 !important}a.has-text-link-dark:hover,a.has-text-link-dark:focus{color:#5485d4 !important}.has-background-link-dark{background-color:#3169c4 !important}.has-text-info{color:#209cee !important}a.has-text-info:hover,a.has-text-info:focus{color:#1081cb !important}.has-background-info{background-color:#209cee !important}.has-text-info-light{color:#ecf7fe !important}a.has-text-info-light:hover,a.has-text-info-light:focus{color:#bde2fa !important}.has-background-info-light{background-color:#ecf7fe !important}.has-text-info-dark{color:#0e72b4 !important}a.has-text-info-dark:hover,a.has-text-info-dark:focus{color:#1190e3 !important}.has-background-info-dark{background-color:#0e72b4 !important}.has-text-success{color:#22c35b !important}a.has-text-success:hover,a.has-text-success:focus{color:#1a9847 !important}.has-background-success{background-color:#22c35b !important}.has-text-success-light{color:#eefcf3 !important}a.has-text-success-light:hover,a.has-text-success-light:focus{color:#c2f4d4 !important}.has-background-success-light{background-color:#eefcf3 !important}.has-text-success-dark{color:#198f43 !important}a.has-text-success-dark:hover,a.has-text-success-dark:focus{color:#21bb57 !important}.has-background-success-dark{background-color:#198f43 !important}.has-text-warning{color:#ffdd57 !important}a.has-text-warning:hover,a.has-text-warning:focus{color:#ffd324 !important}.has-background-warning{background-color:#ffdd57 !important}.has-text-warning-light{color:#fffbeb !important}a.has-text-warning-light:hover,a.has-text-warning-light:focus{color:#fff1b8 !important}.has-background-warning-light{background-color:#fffbeb !important}.has-text-warning-dark{color:#947600 !important}a.has-text-warning-dark:hover,a.has-text-warning-dark:focus{color:#c79f00 !important}.has-background-warning-dark{background-color:#947600 !important}.has-text-danger{color:#da0b00 !important}a.has-text-danger:hover,a.has-text-danger:focus{color:#a70800 !important}.has-background-danger{background-color:#da0b00 !important}.has-text-danger-light{color:#ffeceb !important}a.has-text-danger-light:hover,a.has-text-danger-light:focus{color:#ffbbb8 !important}.has-background-danger-light{background-color:#ffeceb !important}.has-text-danger-dark{color:#f50c00 !important}a.has-text-danger-dark:hover,a.has-text-danger-dark:focus{color:#ff3429 !important}.has-background-danger-dark{background-color:#f50c00 !important}.has-text-black-bis{color:#121212 !important}.has-background-black-bis{background-color:#121212 !important}.has-text-black-ter{color:#242424 !important}.has-background-black-ter{background-color:#242424 !important}.has-text-grey-darker{color:#363636 !important}.has-background-grey-darker{background-color:#363636 !important}.has-text-grey-dark{color:#4a4a4a !important}.has-background-grey-dark{background-color:#4a4a4a !important}.has-text-grey{color:#6b6b6b !important}.has-background-grey{background-color:#6b6b6b !important}.has-text-grey-light{color:#b5b5b5 !important}.has-background-grey-light{background-color:#b5b5b5 !important}.has-text-grey-lighter{color:#dbdbdb !important}.has-background-grey-lighter{background-color:#dbdbdb !important}.has-text-white-ter{color:#f5f5f5 !important}.has-background-white-ter{background-color:#f5f5f5 !important}.has-text-white-bis{color:#fafafa !important}.has-background-white-bis{background-color:#fafafa !important}.is-flex-direction-row{flex-direction:row !important}.is-flex-direction-row-reverse{flex-direction:row-reverse !important}.is-flex-direction-column{flex-direction:column !important}.is-flex-direction-column-reverse{flex-direction:column-reverse !important}.is-flex-wrap-nowrap{flex-wrap:nowrap !important}.is-flex-wrap-wrap{flex-wrap:wrap !important}.is-flex-wrap-wrap-reverse{flex-wrap:wrap-reverse !important}.is-justify-content-flex-start{justify-content:flex-start !important}.is-justify-content-flex-end{justify-content:flex-end !important}.is-justify-content-center{justify-content:center !important}.is-justify-content-space-between{justify-content:space-between !important}.is-justify-content-space-around{justify-content:space-around !important}.is-justify-content-space-evenly{justify-content:space-evenly !important}.is-justify-content-start{justify-content:start !important}.is-justify-content-end{justify-content:end !important}.is-justify-content-left{justify-content:left !important}.is-justify-content-right{justify-content:right !important}.is-align-content-flex-start{align-content:flex-start !important}.is-align-content-flex-end{align-content:flex-end !important}.is-align-content-center{align-content:center !important}.is-align-content-space-between{align-content:space-between !important}.is-align-content-space-around{align-content:space-around !important}.is-align-content-space-evenly{align-content:space-evenly !important}.is-align-content-stretch{align-content:stretch !important}.is-align-content-start{align-content:start !important}.is-align-content-end{align-content:end !important}.is-align-content-baseline{align-content:baseline !important}.is-align-items-stretch{align-items:stretch !important}.is-align-items-flex-start{align-items:flex-start !important}.is-align-items-flex-end{align-items:flex-end !important}.is-align-items-center{align-items:center !important}.is-align-items-baseline{align-items:baseline !important}.is-align-items-start{align-items:start !important}.is-align-items-end{align-items:end !important}.is-align-items-self-start{align-items:self-start !important}.is-align-items-self-end{align-items:self-end !important}.is-align-self-auto{align-self:auto !important}.is-align-self-flex-start{align-self:flex-start !important}.is-align-self-flex-end{align-self:flex-end !important}.is-align-self-center{align-self:center !important}.is-align-self-baseline{align-self:baseline !important}.is-align-self-stretch{align-self:stretch !important}.is-flex-grow-0{flex-grow:0 !important}.is-flex-grow-1{flex-grow:1 !important}.is-flex-grow-2{flex-grow:2 !important}.is-flex-grow-3{flex-grow:3 !important}.is-flex-grow-4{flex-grow:4 !important}.is-flex-grow-5{flex-grow:5 !important}.is-flex-shrink-0{flex-shrink:0 !important}.is-flex-shrink-1{flex-shrink:1 !important}.is-flex-shrink-2{flex-shrink:2 !important}.is-flex-shrink-3{flex-shrink:3 !important}.is-flex-shrink-4{flex-shrink:4 !important}.is-flex-shrink-5{flex-shrink:5 !important}.is-clearfix::after{clear:both;content:" ";display:table}.is-pulled-left{float:left !important}.is-pulled-right{float:right !important}.is-radiusless{border-radius:0 !important}.is-shadowless{box-shadow:none !important}.is-clickable{cursor:pointer !important;pointer-events:all !important}.is-clipped{overflow:hidden !important}.is-relative{position:relative !important}.is-marginless{margin:0 !important}.is-paddingless{padding:0 !important}.m-0{margin:0 !important}.mt-0{margin-top:0 !important}.mr-0{margin-right:0 !important}.mb-0{margin-bottom:0 !important}.ml-0{margin-left:0 !important}.mx-0{margin-left:0 !important;margin-right:0 !important}.my-0{margin-top:0 !important;margin-bottom:0 !important}.m-1{margin:.25rem !important}.mt-1{margin-top:.25rem !important}.mr-1{margin-right:.25rem !important}.mb-1{margin-bottom:.25rem !important}.ml-1{margin-left:.25rem !important}.mx-1{margin-left:.25rem !important;margin-right:.25rem !important}.my-1{margin-top:.25rem !important;margin-bottom:.25rem !important}.m-2{margin:.5rem !important}.mt-2{margin-top:.5rem !important}.mr-2{margin-right:.5rem !important}.mb-2{margin-bottom:.5rem !important}.ml-2{margin-left:.5rem !important}.mx-2{margin-left:.5rem !important;margin-right:.5rem !important}.my-2{margin-top:.5rem !important;margin-bottom:.5rem !important}.m-3{margin:.75rem !important}.mt-3{margin-top:.75rem !important}.mr-3{margin-right:.75rem !important}.mb-3{margin-bottom:.75rem !important}.ml-3{margin-left:.75rem !important}.mx-3{margin-left:.75rem !important;margin-right:.75rem !important}.my-3{margin-top:.75rem !important;margin-bottom:.75rem !important}.m-4{margin:1rem !important}.mt-4{margin-top:1rem !important}.mr-4{margin-right:1rem !important}.mb-4{margin-bottom:1rem !important}.ml-4{margin-left:1rem !important}.mx-4{margin-left:1rem !important;margin-right:1rem !important}.my-4{margin-top:1rem !important;margin-bottom:1rem !important}.m-5{margin:1.5rem !important}.mt-5{margin-top:1.5rem !important}.mr-5{margin-right:1.5rem !important}.mb-5{margin-bottom:1.5rem !important}.ml-5{margin-left:1.5rem !important}.mx-5{margin-left:1.5rem !important;margin-right:1.5rem !important}.my-5{margin-top:1.5rem !important;margin-bottom:1.5rem !important}.m-6{margin:3rem !important}.mt-6{margin-top:3rem !important}.mr-6{margin-right:3rem !important}.mb-6{margin-bottom:3rem !important}.ml-6{margin-left:3rem !important}.mx-6{margin-left:3rem !important;margin-right:3rem !important}.my-6{margin-top:3rem !important;margin-bottom:3rem !important}.m-auto{margin:auto !important}.mt-auto{margin-top:auto !important}.mr-auto{margin-right:auto !important}.mb-auto{margin-bottom:auto !important}.ml-auto{margin-left:auto !important}.mx-auto{margin-left:auto !important;margin-right:auto !important}.my-auto{margin-top:auto !important;margin-bottom:auto !important}.p-0{padding:0 !important}.pt-0{padding-top:0 !important}.pr-0{padding-right:0 !important}.pb-0{padding-bottom:0 !important}.pl-0{padding-left:0 !important}.px-0{padding-left:0 !important;padding-right:0 !important}.py-0{padding-top:0 !important;padding-bottom:0 !important}.p-1{padding:.25rem !important}.pt-1{padding-top:.25rem !important}.pr-1{padding-right:.25rem !important}.pb-1{padding-bottom:.25rem !important}.pl-1{padding-left:.25rem !important}.px-1{padding-left:.25rem !important;padding-right:.25rem !important}.py-1{padding-top:.25rem !important;padding-bottom:.25rem !important}.p-2{padding:.5rem !important}.pt-2{padding-top:.5rem !important}.pr-2{padding-right:.5rem !important}.pb-2{padding-bottom:.5rem !important}.pl-2{padding-left:.5rem !important}.px-2{padding-left:.5rem !important;padding-right:.5rem !important}.py-2{padding-top:.5rem !important;padding-bottom:.5rem !important}.p-3{padding:.75rem !important}.pt-3{padding-top:.75rem !important}.pr-3{padding-right:.75rem !important}.pb-3{padding-bottom:.75rem !important}.pl-3{padding-left:.75rem !important}.px-3{padding-left:.75rem !important;padding-right:.75rem !important}.py-3{padding-top:.75rem !important;padding-bottom:.75rem !important}.p-4{padding:1rem !important}.pt-4{padding-top:1rem !important}.pr-4{padding-right:1rem !important}.pb-4{padding-bottom:1rem !important}.pl-4{padding-left:1rem !important}.px-4{padding-left:1rem !important;padding-right:1rem !important}.py-4{padding-top:1rem !important;padding-bottom:1rem !important}.p-5{padding:1.5rem !important}.pt-5{padding-top:1.5rem !important}.pr-5{padding-right:1.5rem !important}.pb-5{padding-bottom:1.5rem !important}.pl-5{padding-left:1.5rem !important}.px-5{padding-left:1.5rem !important;padding-right:1.5rem !important}.py-5{padding-top:1.5rem !important;padding-bottom:1.5rem !important}.p-6{padding:3rem !important}.pt-6{padding-top:3rem !important}.pr-6{padding-right:3rem !important}.pb-6{padding-bottom:3rem !important}.pl-6{padding-left:3rem !important}.px-6{padding-left:3rem !important;padding-right:3rem !important}.py-6{padding-top:3rem !important;padding-bottom:3rem !important}.p-auto{padding:auto !important}.pt-auto{padding-top:auto !important}.pr-auto{padding-right:auto !important}.pb-auto{padding-bottom:auto !important}.pl-auto{padding-left:auto !important}.px-auto{padding-left:auto !important;padding-right:auto !important}.py-auto{padding-top:auto !important;padding-bottom:auto !important}.is-size-1{font-size:3rem !important}.is-size-2{font-size:2.5rem !important}.is-size-3{font-size:2rem !important}.is-size-4{font-size:1.5rem !important}.is-size-5{font-size:1.25rem !important}.is-size-6{font-size:1rem !important}.is-size-7,.docstring>section>a.docs-sourcelink{font-size:.75rem !important}@media screen and (max-width: 768px){.is-size-1-mobile{font-size:3rem !important}.is-size-2-mobile{font-size:2.5rem !important}.is-size-3-mobile{font-size:2rem !important}.is-size-4-mobile{font-size:1.5rem !important}.is-size-5-mobile{font-size:1.25rem !important}.is-size-6-mobile{font-size:1rem !important}.is-size-7-mobile{font-size:.75rem !important}}@media screen and (min-width: 769px),print{.is-size-1-tablet{font-size:3rem !important}.is-size-2-tablet{font-size:2.5rem !important}.is-size-3-tablet{font-size:2rem !important}.is-size-4-tablet{font-size:1.5rem !important}.is-size-5-tablet{font-size:1.25rem !important}.is-size-6-tablet{font-size:1rem !important}.is-size-7-tablet{font-size:.75rem !important}}@media screen and (max-width: 1055px){.is-size-1-touch{font-size:3rem !important}.is-size-2-touch{font-size:2.5rem !important}.is-size-3-touch{font-size:2rem !important}.is-size-4-touch{font-size:1.5rem !important}.is-size-5-touch{font-size:1.25rem !important}.is-size-6-touch{font-size:1rem !important}.is-size-7-touch{font-size:.75rem !important}}@media screen and (min-width: 1056px){.is-size-1-desktop{font-size:3rem !important}.is-size-2-desktop{font-size:2.5rem !important}.is-size-3-desktop{font-size:2rem !important}.is-size-4-desktop{font-size:1.5rem !important}.is-size-5-desktop{font-size:1.25rem !important}.is-size-6-desktop{font-size:1rem !important}.is-size-7-desktop{font-size:.75rem !important}}@media screen and (min-width: 1216px){.is-size-1-widescreen{font-size:3rem !important}.is-size-2-widescreen{font-size:2.5rem !important}.is-size-3-widescreen{font-size:2rem !important}.is-size-4-widescreen{font-size:1.5rem !important}.is-size-5-widescreen{font-size:1.25rem !important}.is-size-6-widescreen{font-size:1rem !important}.is-size-7-widescreen{font-size:.75rem !important}}@media screen and (min-width: 1408px){.is-size-1-fullhd{font-size:3rem !important}.is-size-2-fullhd{font-size:2.5rem !important}.is-size-3-fullhd{font-size:2rem !important}.is-size-4-fullhd{font-size:1.5rem !important}.is-size-5-fullhd{font-size:1.25rem !important}.is-size-6-fullhd{font-size:1rem !important}.is-size-7-fullhd{font-size:.75rem !important}}.has-text-centered{text-align:center !important}.has-text-justified{text-align:justify !important}.has-text-left{text-align:left !important}.has-text-right{text-align:right !important}@media screen and (max-width: 768px){.has-text-centered-mobile{text-align:center !important}}@media screen and (min-width: 769px),print{.has-text-centered-tablet{text-align:center !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-centered-tablet-only{text-align:center !important}}@media screen and (max-width: 1055px){.has-text-centered-touch{text-align:center !important}}@media screen and (min-width: 1056px){.has-text-centered-desktop{text-align:center !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-centered-desktop-only{text-align:center !important}}@media screen and (min-width: 1216px){.has-text-centered-widescreen{text-align:center !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-centered-widescreen-only{text-align:center !important}}@media screen and (min-width: 1408px){.has-text-centered-fullhd{text-align:center !important}}@media screen and (max-width: 768px){.has-text-justified-mobile{text-align:justify !important}}@media screen and (min-width: 769px),print{.has-text-justified-tablet{text-align:justify !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-justified-tablet-only{text-align:justify !important}}@media screen and (max-width: 1055px){.has-text-justified-touch{text-align:justify !important}}@media screen and (min-width: 1056px){.has-text-justified-desktop{text-align:justify !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-justified-desktop-only{text-align:justify !important}}@media screen and (min-width: 1216px){.has-text-justified-widescreen{text-align:justify !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-justified-widescreen-only{text-align:justify !important}}@media screen and (min-width: 1408px){.has-text-justified-fullhd{text-align:justify !important}}@media screen and (max-width: 768px){.has-text-left-mobile{text-align:left !important}}@media screen and (min-width: 769px),print{.has-text-left-tablet{text-align:left !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-left-tablet-only{text-align:left !important}}@media screen and (max-width: 1055px){.has-text-left-touch{text-align:left !important}}@media screen and (min-width: 1056px){.has-text-left-desktop{text-align:left !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-left-desktop-only{text-align:left !important}}@media screen and (min-width: 1216px){.has-text-left-widescreen{text-align:left !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-left-widescreen-only{text-align:left !important}}@media screen and (min-width: 1408px){.has-text-left-fullhd{text-align:left !important}}@media screen and (max-width: 768px){.has-text-right-mobile{text-align:right !important}}@media screen and (min-width: 769px),print{.has-text-right-tablet{text-align:right !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.has-text-right-tablet-only{text-align:right !important}}@media screen and (max-width: 1055px){.has-text-right-touch{text-align:right !important}}@media screen and (min-width: 1056px){.has-text-right-desktop{text-align:right !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.has-text-right-desktop-only{text-align:right !important}}@media screen and (min-width: 1216px){.has-text-right-widescreen{text-align:right !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.has-text-right-widescreen-only{text-align:right !important}}@media screen and (min-width: 1408px){.has-text-right-fullhd{text-align:right !important}}.is-capitalized{text-transform:capitalize !important}.is-lowercase{text-transform:lowercase !important}.is-uppercase{text-transform:uppercase !important}.is-italic{font-style:italic !important}.is-underlined{text-decoration:underline !important}.has-text-weight-light{font-weight:300 !important}.has-text-weight-normal{font-weight:400 !important}.has-text-weight-medium{font-weight:500 !important}.has-text-weight-semibold{font-weight:600 !important}.has-text-weight-bold{font-weight:700 !important}.is-family-primary{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif !important}.is-family-secondary{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif !important}.is-family-sans-serif{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif !important}.is-family-monospace{font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace !important}.is-family-code{font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace !important}.is-block{display:block !important}@media screen and (max-width: 768px){.is-block-mobile{display:block !important}}@media screen and (min-width: 769px),print{.is-block-tablet{display:block !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-block-tablet-only{display:block !important}}@media screen and (max-width: 1055px){.is-block-touch{display:block !important}}@media screen and (min-width: 1056px){.is-block-desktop{display:block !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-block-desktop-only{display:block !important}}@media screen and (min-width: 1216px){.is-block-widescreen{display:block !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-block-widescreen-only{display:block !important}}@media screen and (min-width: 1408px){.is-block-fullhd{display:block !important}}.is-flex{display:flex !important}@media screen and (max-width: 768px){.is-flex-mobile{display:flex !important}}@media screen and (min-width: 769px),print{.is-flex-tablet{display:flex !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-flex-tablet-only{display:flex !important}}@media screen and (max-width: 1055px){.is-flex-touch{display:flex !important}}@media screen and (min-width: 1056px){.is-flex-desktop{display:flex !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-flex-desktop-only{display:flex !important}}@media screen and (min-width: 1216px){.is-flex-widescreen{display:flex !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-flex-widescreen-only{display:flex !important}}@media screen and (min-width: 1408px){.is-flex-fullhd{display:flex !important}}.is-inline{display:inline !important}@media screen and (max-width: 768px){.is-inline-mobile{display:inline !important}}@media screen and (min-width: 769px),print{.is-inline-tablet{display:inline !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-inline-tablet-only{display:inline !important}}@media screen and (max-width: 1055px){.is-inline-touch{display:inline !important}}@media screen and (min-width: 1056px){.is-inline-desktop{display:inline !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-inline-desktop-only{display:inline !important}}@media screen and (min-width: 1216px){.is-inline-widescreen{display:inline !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-inline-widescreen-only{display:inline !important}}@media screen and (min-width: 1408px){.is-inline-fullhd{display:inline !important}}.is-inline-block{display:inline-block !important}@media screen and (max-width: 768px){.is-inline-block-mobile{display:inline-block !important}}@media screen and (min-width: 769px),print{.is-inline-block-tablet{display:inline-block !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-inline-block-tablet-only{display:inline-block !important}}@media screen and (max-width: 1055px){.is-inline-block-touch{display:inline-block !important}}@media screen and (min-width: 1056px){.is-inline-block-desktop{display:inline-block !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-inline-block-desktop-only{display:inline-block !important}}@media screen and (min-width: 1216px){.is-inline-block-widescreen{display:inline-block !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-inline-block-widescreen-only{display:inline-block !important}}@media screen and (min-width: 1408px){.is-inline-block-fullhd{display:inline-block !important}}.is-inline-flex{display:inline-flex !important}@media screen and (max-width: 768px){.is-inline-flex-mobile{display:inline-flex !important}}@media screen and (min-width: 769px),print{.is-inline-flex-tablet{display:inline-flex !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-inline-flex-tablet-only{display:inline-flex !important}}@media screen and (max-width: 1055px){.is-inline-flex-touch{display:inline-flex !important}}@media screen and (min-width: 1056px){.is-inline-flex-desktop{display:inline-flex !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-inline-flex-desktop-only{display:inline-flex !important}}@media screen and (min-width: 1216px){.is-inline-flex-widescreen{display:inline-flex !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-inline-flex-widescreen-only{display:inline-flex !important}}@media screen and (min-width: 1408px){.is-inline-flex-fullhd{display:inline-flex !important}}.is-hidden{display:none !important}.is-sr-only{border:none !important;clip:rect(0, 0, 0, 0) !important;height:0.01em !important;overflow:hidden !important;padding:0 !important;position:absolute !important;white-space:nowrap !important;width:0.01em !important}@media screen and (max-width: 768px){.is-hidden-mobile{display:none !important}}@media screen and (min-width: 769px),print{.is-hidden-tablet{display:none !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-hidden-tablet-only{display:none !important}}@media screen and (max-width: 1055px){.is-hidden-touch{display:none !important}}@media screen and (min-width: 1056px){.is-hidden-desktop{display:none !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-hidden-desktop-only{display:none !important}}@media screen and (min-width: 1216px){.is-hidden-widescreen{display:none !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-hidden-widescreen-only{display:none !important}}@media screen and (min-width: 1408px){.is-hidden-fullhd{display:none !important}}.is-invisible{visibility:hidden !important}@media screen and (max-width: 768px){.is-invisible-mobile{visibility:hidden !important}}@media screen and (min-width: 769px),print{.is-invisible-tablet{visibility:hidden !important}}@media screen and (min-width: 769px) and (max-width: 1055px){.is-invisible-tablet-only{visibility:hidden !important}}@media screen and (max-width: 1055px){.is-invisible-touch{visibility:hidden !important}}@media screen and (min-width: 1056px){.is-invisible-desktop{visibility:hidden !important}}@media screen and (min-width: 1056px) and (max-width: 1215px){.is-invisible-desktop-only{visibility:hidden !important}}@media screen and (min-width: 1216px){.is-invisible-widescreen{visibility:hidden !important}}@media screen and (min-width: 1216px) and (max-width: 1407px){.is-invisible-widescreen-only{visibility:hidden !important}}@media screen and (min-width: 1408px){.is-invisible-fullhd{visibility:hidden !important}}/*! minireset.css v0.0.6 | MIT License | github.com/jgthms/minireset.css */html,body,p,ol,ul,li,dl,dt,dd,blockquote,figure,fieldset,legend,textarea,pre,iframe,hr,h1,h2,h3,h4,h5,h6{margin:0;padding:0}h1,h2,h3,h4,h5,h6{font-size:100%;font-weight:normal}ul{list-style:none}button,input,select,textarea{margin:0}html{box-sizing:border-box}*,*::before,*::after{box-sizing:inherit}img,video{height:auto;max-width:100%}iframe{border:0}table{border-collapse:collapse;border-spacing:0}td,th{padding:0}td:not([align]),th:not([align]){text-align:inherit}html{background-color:#fff;font-size:16px;-moz-osx-font-smoothing:grayscale;-webkit-font-smoothing:antialiased;min-width:300px;overflow-x:auto;overflow-y:scroll;text-rendering:optimizeLegibility;text-size-adjust:100%}article,aside,figure,footer,header,hgroup,section{display:block}body,button,input,optgroup,select,textarea{font-family:"Lato Medium",-apple-system,BlinkMacSystemFont,"Segoe UI","Helvetica Neue","Helvetica","Arial",sans-serif}code,pre{-moz-osx-font-smoothing:auto;-webkit-font-smoothing:auto;font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace}body{color:#222;font-size:1em;font-weight:400;line-height:1.5}a{color:#2e63b8;cursor:pointer;text-decoration:none}a strong{color:currentColor}a:hover{color:#363636}code{background-color:rgba(0,0,0,0.05);color:#000;font-size:.875em;font-weight:normal;padding:.1em}hr{background-color:#f5f5f5;border:none;display:block;height:2px;margin:1.5rem 0}img{height:auto;max-width:100%}input[type="checkbox"],input[type="radio"]{vertical-align:baseline}small{font-size:.875em}span{font-style:inherit;font-weight:inherit}strong{color:#222;font-weight:700}fieldset{border:none}pre{-webkit-overflow-scrolling:touch;background-color:#f5f5f5;color:#222;font-size:.875em;overflow-x:auto;padding:1.25rem 1.5rem;white-space:pre;word-wrap:normal}pre code{background-color:transparent;color:currentColor;font-size:1em;padding:0}table td,table th{vertical-align:top}table td:not([align]),table th:not([align]){text-align:inherit}table th{color:#222}@keyframes spinAround{from{transform:rotate(0deg)}to{transform:rotate(359deg)}}.box{background-color:#fff;border-radius:6px;box-shadow:#bbb;color:#222;display:block;padding:1.25rem}a.box:hover,a.box:focus{box-shadow:0 0.5em 1em -0.125em rgba(10,10,10,0.1),0 0 0 1px #2e63b8}a.box:active{box-shadow:inset 0 1px 2px rgba(10,10,10,0.2),0 0 0 1px #2e63b8}.button{background-color:#fff;border-color:#dbdbdb;border-width:1px;color:#222;cursor:pointer;justify-content:center;padding-bottom:calc(0.5em - 1px);padding-left:1em;padding-right:1em;padding-top:calc(0.5em - 1px);text-align:center;white-space:nowrap}.button strong{color:inherit}.button .icon,.button .icon.is-small,.button #documenter .docs-sidebar form.docs-search>input.icon,#documenter .docs-sidebar .button form.docs-search>input.icon,.button .icon.is-medium,.button .icon.is-large{height:1.5em;width:1.5em}.button .icon:first-child:not(:last-child){margin-left:calc(-0.5em - 1px);margin-right:.25em}.button .icon:last-child:not(:first-child){margin-left:.25em;margin-right:calc(-0.5em - 1px)}.button .icon:first-child:last-child{margin-left:calc(-0.5em - 1px);margin-right:calc(-0.5em - 1px)}.button:hover,.button.is-hovered{border-color:#b5b5b5;color:#363636}.button:focus,.button.is-focused{border-color:#3c5dcd;color:#363636}.button:focus:not(:active),.button.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(46,99,184,0.25)}.button:active,.button.is-active{border-color:#4a4a4a;color:#363636}.button.is-text{background-color:transparent;border-color:transparent;color:#222;text-decoration:underline}.button.is-text:hover,.button.is-text.is-hovered,.button.is-text:focus,.button.is-text.is-focused{background-color:#f5f5f5;color:#222}.button.is-text:active,.button.is-text.is-active{background-color:#e8e8e8;color:#222}.button.is-text[disabled],fieldset[disabled] .button.is-text{background-color:transparent;border-color:transparent;box-shadow:none}.button.is-ghost{background:none;border-color:rgba(0,0,0,0);color:#2e63b8;text-decoration:none}.button.is-ghost:hover,.button.is-ghost.is-hovered{color:#2e63b8;text-decoration:underline}.button.is-white{background-color:#fff;border-color:transparent;color:#0a0a0a}.button.is-white:hover,.button.is-white.is-hovered{background-color:#f9f9f9;border-color:transparent;color:#0a0a0a}.button.is-white:focus,.button.is-white.is-focused{border-color:transparent;color:#0a0a0a}.button.is-white:focus:not(:active),.button.is-white.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(255,255,255,0.25)}.button.is-white:active,.button.is-white.is-active{background-color:#f2f2f2;border-color:transparent;color:#0a0a0a}.button.is-white[disabled],fieldset[disabled] .button.is-white{background-color:#fff;border-color:#fff;box-shadow:none}.button.is-white.is-inverted{background-color:#0a0a0a;color:#fff}.button.is-white.is-inverted:hover,.button.is-white.is-inverted.is-hovered{background-color:#000}.button.is-white.is-inverted[disabled],fieldset[disabled] .button.is-white.is-inverted{background-color:#0a0a0a;border-color:transparent;box-shadow:none;color:#fff}.button.is-white.is-loading::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}.button.is-white.is-outlined{background-color:transparent;border-color:#fff;color:#fff}.button.is-white.is-outlined:hover,.button.is-white.is-outlined.is-hovered,.button.is-white.is-outlined:focus,.button.is-white.is-outlined.is-focused{background-color:#fff;border-color:#fff;color:#0a0a0a}.button.is-white.is-outlined.is-loading::after{border-color:transparent transparent #fff #fff !important}.button.is-white.is-outlined.is-loading:hover::after,.button.is-white.is-outlined.is-loading.is-hovered::after,.button.is-white.is-outlined.is-loading:focus::after,.button.is-white.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}.button.is-white.is-outlined[disabled],fieldset[disabled] .button.is-white.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-white.is-inverted.is-outlined{background-color:transparent;border-color:#0a0a0a;color:#0a0a0a}.button.is-white.is-inverted.is-outlined:hover,.button.is-white.is-inverted.is-outlined.is-hovered,.button.is-white.is-inverted.is-outlined:focus,.button.is-white.is-inverted.is-outlined.is-focused{background-color:#0a0a0a;color:#fff}.button.is-white.is-inverted.is-outlined.is-loading:hover::after,.button.is-white.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-white.is-inverted.is-outlined.is-loading:focus::after,.button.is-white.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}.button.is-white.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-white.is-inverted.is-outlined{background-color:transparent;border-color:#0a0a0a;box-shadow:none;color:#0a0a0a}.button.is-black{background-color:#0a0a0a;border-color:transparent;color:#fff}.button.is-black:hover,.button.is-black.is-hovered{background-color:#040404;border-color:transparent;color:#fff}.button.is-black:focus,.button.is-black.is-focused{border-color:transparent;color:#fff}.button.is-black:focus:not(:active),.button.is-black.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(10,10,10,0.25)}.button.is-black:active,.button.is-black.is-active{background-color:#000;border-color:transparent;color:#fff}.button.is-black[disabled],fieldset[disabled] .button.is-black{background-color:#0a0a0a;border-color:#0a0a0a;box-shadow:none}.button.is-black.is-inverted{background-color:#fff;color:#0a0a0a}.button.is-black.is-inverted:hover,.button.is-black.is-inverted.is-hovered{background-color:#f2f2f2}.button.is-black.is-inverted[disabled],fieldset[disabled] .button.is-black.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#0a0a0a}.button.is-black.is-loading::after{border-color:transparent transparent #fff #fff !important}.button.is-black.is-outlined{background-color:transparent;border-color:#0a0a0a;color:#0a0a0a}.button.is-black.is-outlined:hover,.button.is-black.is-outlined.is-hovered,.button.is-black.is-outlined:focus,.button.is-black.is-outlined.is-focused{background-color:#0a0a0a;border-color:#0a0a0a;color:#fff}.button.is-black.is-outlined.is-loading::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}.button.is-black.is-outlined.is-loading:hover::after,.button.is-black.is-outlined.is-loading.is-hovered::after,.button.is-black.is-outlined.is-loading:focus::after,.button.is-black.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}.button.is-black.is-outlined[disabled],fieldset[disabled] .button.is-black.is-outlined{background-color:transparent;border-color:#0a0a0a;box-shadow:none;color:#0a0a0a}.button.is-black.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}.button.is-black.is-inverted.is-outlined:hover,.button.is-black.is-inverted.is-outlined.is-hovered,.button.is-black.is-inverted.is-outlined:focus,.button.is-black.is-inverted.is-outlined.is-focused{background-color:#fff;color:#0a0a0a}.button.is-black.is-inverted.is-outlined.is-loading:hover::after,.button.is-black.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-black.is-inverted.is-outlined.is-loading:focus::after,.button.is-black.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #0a0a0a #0a0a0a !important}.button.is-black.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-black.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-light{background-color:#f5f5f5;border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-light:hover,.button.is-light.is-hovered{background-color:#eee;border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-light:focus,.button.is-light.is-focused{border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-light:focus:not(:active),.button.is-light.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(245,245,245,0.25)}.button.is-light:active,.button.is-light.is-active{background-color:#e8e8e8;border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-light[disabled],fieldset[disabled] .button.is-light{background-color:#f5f5f5;border-color:#f5f5f5;box-shadow:none}.button.is-light.is-inverted{background-color:rgba(0,0,0,0.7);color:#f5f5f5}.button.is-light.is-inverted:hover,.button.is-light.is-inverted.is-hovered{background-color:rgba(0,0,0,0.7)}.button.is-light.is-inverted[disabled],fieldset[disabled] .button.is-light.is-inverted{background-color:rgba(0,0,0,0.7);border-color:transparent;box-shadow:none;color:#f5f5f5}.button.is-light.is-loading::after{border-color:transparent transparent rgba(0,0,0,0.7) rgba(0,0,0,0.7) !important}.button.is-light.is-outlined{background-color:transparent;border-color:#f5f5f5;color:#f5f5f5}.button.is-light.is-outlined:hover,.button.is-light.is-outlined.is-hovered,.button.is-light.is-outlined:focus,.button.is-light.is-outlined.is-focused{background-color:#f5f5f5;border-color:#f5f5f5;color:rgba(0,0,0,0.7)}.button.is-light.is-outlined.is-loading::after{border-color:transparent transparent #f5f5f5 #f5f5f5 !important}.button.is-light.is-outlined.is-loading:hover::after,.button.is-light.is-outlined.is-loading.is-hovered::after,.button.is-light.is-outlined.is-loading:focus::after,.button.is-light.is-outlined.is-loading.is-focused::after{border-color:transparent transparent rgba(0,0,0,0.7) rgba(0,0,0,0.7) !important}.button.is-light.is-outlined[disabled],fieldset[disabled] .button.is-light.is-outlined{background-color:transparent;border-color:#f5f5f5;box-shadow:none;color:#f5f5f5}.button.is-light.is-inverted.is-outlined{background-color:transparent;border-color:rgba(0,0,0,0.7);color:rgba(0,0,0,0.7)}.button.is-light.is-inverted.is-outlined:hover,.button.is-light.is-inverted.is-outlined.is-hovered,.button.is-light.is-inverted.is-outlined:focus,.button.is-light.is-inverted.is-outlined.is-focused{background-color:rgba(0,0,0,0.7);color:#f5f5f5}.button.is-light.is-inverted.is-outlined.is-loading:hover::after,.button.is-light.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-light.is-inverted.is-outlined.is-loading:focus::after,.button.is-light.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #f5f5f5 #f5f5f5 !important}.button.is-light.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-light.is-inverted.is-outlined{background-color:transparent;border-color:rgba(0,0,0,0.7);box-shadow:none;color:rgba(0,0,0,0.7)}.button.is-dark,.content kbd.button{background-color:#363636;border-color:transparent;color:#fff}.button.is-dark:hover,.content kbd.button:hover,.button.is-dark.is-hovered,.content kbd.button.is-hovered{background-color:#2f2f2f;border-color:transparent;color:#fff}.button.is-dark:focus,.content kbd.button:focus,.button.is-dark.is-focused,.content kbd.button.is-focused{border-color:transparent;color:#fff}.button.is-dark:focus:not(:active),.content kbd.button:focus:not(:active),.button.is-dark.is-focused:not(:active),.content kbd.button.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(54,54,54,0.25)}.button.is-dark:active,.content kbd.button:active,.button.is-dark.is-active,.content kbd.button.is-active{background-color:#292929;border-color:transparent;color:#fff}.button.is-dark[disabled],.content kbd.button[disabled],fieldset[disabled] .button.is-dark,fieldset[disabled] .content kbd.button,.content fieldset[disabled] kbd.button{background-color:#363636;border-color:#363636;box-shadow:none}.button.is-dark.is-inverted,.content kbd.button.is-inverted{background-color:#fff;color:#363636}.button.is-dark.is-inverted:hover,.content kbd.button.is-inverted:hover,.button.is-dark.is-inverted.is-hovered,.content kbd.button.is-inverted.is-hovered{background-color:#f2f2f2}.button.is-dark.is-inverted[disabled],.content kbd.button.is-inverted[disabled],fieldset[disabled] .button.is-dark.is-inverted,fieldset[disabled] .content kbd.button.is-inverted,.content fieldset[disabled] kbd.button.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#363636}.button.is-dark.is-loading::after,.content kbd.button.is-loading::after{border-color:transparent transparent #fff #fff !important}.button.is-dark.is-outlined,.content kbd.button.is-outlined{background-color:transparent;border-color:#363636;color:#363636}.button.is-dark.is-outlined:hover,.content kbd.button.is-outlined:hover,.button.is-dark.is-outlined.is-hovered,.content kbd.button.is-outlined.is-hovered,.button.is-dark.is-outlined:focus,.content kbd.button.is-outlined:focus,.button.is-dark.is-outlined.is-focused,.content kbd.button.is-outlined.is-focused{background-color:#363636;border-color:#363636;color:#fff}.button.is-dark.is-outlined.is-loading::after,.content kbd.button.is-outlined.is-loading::after{border-color:transparent transparent #363636 #363636 !important}.button.is-dark.is-outlined.is-loading:hover::after,.content kbd.button.is-outlined.is-loading:hover::after,.button.is-dark.is-outlined.is-loading.is-hovered::after,.content kbd.button.is-outlined.is-loading.is-hovered::after,.button.is-dark.is-outlined.is-loading:focus::after,.content kbd.button.is-outlined.is-loading:focus::after,.button.is-dark.is-outlined.is-loading.is-focused::after,.content kbd.button.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}.button.is-dark.is-outlined[disabled],.content kbd.button.is-outlined[disabled],fieldset[disabled] .button.is-dark.is-outlined,fieldset[disabled] .content kbd.button.is-outlined,.content fieldset[disabled] kbd.button.is-outlined{background-color:transparent;border-color:#363636;box-shadow:none;color:#363636}.button.is-dark.is-inverted.is-outlined,.content kbd.button.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}.button.is-dark.is-inverted.is-outlined:hover,.content kbd.button.is-inverted.is-outlined:hover,.button.is-dark.is-inverted.is-outlined.is-hovered,.content kbd.button.is-inverted.is-outlined.is-hovered,.button.is-dark.is-inverted.is-outlined:focus,.content kbd.button.is-inverted.is-outlined:focus,.button.is-dark.is-inverted.is-outlined.is-focused,.content kbd.button.is-inverted.is-outlined.is-focused{background-color:#fff;color:#363636}.button.is-dark.is-inverted.is-outlined.is-loading:hover::after,.content kbd.button.is-inverted.is-outlined.is-loading:hover::after,.button.is-dark.is-inverted.is-outlined.is-loading.is-hovered::after,.content kbd.button.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-dark.is-inverted.is-outlined.is-loading:focus::after,.content kbd.button.is-inverted.is-outlined.is-loading:focus::after,.button.is-dark.is-inverted.is-outlined.is-loading.is-focused::after,.content kbd.button.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #363636 #363636 !important}.button.is-dark.is-inverted.is-outlined[disabled],.content kbd.button.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-dark.is-inverted.is-outlined,fieldset[disabled] .content kbd.button.is-inverted.is-outlined,.content fieldset[disabled] kbd.button.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-primary,.docstring>section>a.button.docs-sourcelink{background-color:#4eb5de;border-color:transparent;color:#fff}.button.is-primary:hover,.docstring>section>a.button.docs-sourcelink:hover,.button.is-primary.is-hovered,.docstring>section>a.button.is-hovered.docs-sourcelink{background-color:#43b1dc;border-color:transparent;color:#fff}.button.is-primary:focus,.docstring>section>a.button.docs-sourcelink:focus,.button.is-primary.is-focused,.docstring>section>a.button.is-focused.docs-sourcelink{border-color:transparent;color:#fff}.button.is-primary:focus:not(:active),.docstring>section>a.button.docs-sourcelink:focus:not(:active),.button.is-primary.is-focused:not(:active),.docstring>section>a.button.is-focused.docs-sourcelink:not(:active){box-shadow:0 0 0 0.125em rgba(78,181,222,0.25)}.button.is-primary:active,.docstring>section>a.button.docs-sourcelink:active,.button.is-primary.is-active,.docstring>section>a.button.is-active.docs-sourcelink{background-color:#39acda;border-color:transparent;color:#fff}.button.is-primary[disabled],.docstring>section>a.button.docs-sourcelink[disabled],fieldset[disabled] .button.is-primary,fieldset[disabled] .docstring>section>a.button.docs-sourcelink{background-color:#4eb5de;border-color:#4eb5de;box-shadow:none}.button.is-primary.is-inverted,.docstring>section>a.button.is-inverted.docs-sourcelink{background-color:#fff;color:#4eb5de}.button.is-primary.is-inverted:hover,.docstring>section>a.button.is-inverted.docs-sourcelink:hover,.button.is-primary.is-inverted.is-hovered,.docstring>section>a.button.is-inverted.is-hovered.docs-sourcelink{background-color:#f2f2f2}.button.is-primary.is-inverted[disabled],.docstring>section>a.button.is-inverted.docs-sourcelink[disabled],fieldset[disabled] .button.is-primary.is-inverted,fieldset[disabled] .docstring>section>a.button.is-inverted.docs-sourcelink{background-color:#fff;border-color:transparent;box-shadow:none;color:#4eb5de}.button.is-primary.is-loading::after,.docstring>section>a.button.is-loading.docs-sourcelink::after{border-color:transparent transparent #fff #fff !important}.button.is-primary.is-outlined,.docstring>section>a.button.is-outlined.docs-sourcelink{background-color:transparent;border-color:#4eb5de;color:#4eb5de}.button.is-primary.is-outlined:hover,.docstring>section>a.button.is-outlined.docs-sourcelink:hover,.button.is-primary.is-outlined.is-hovered,.docstring>section>a.button.is-outlined.is-hovered.docs-sourcelink,.button.is-primary.is-outlined:focus,.docstring>section>a.button.is-outlined.docs-sourcelink:focus,.button.is-primary.is-outlined.is-focused,.docstring>section>a.button.is-outlined.is-focused.docs-sourcelink{background-color:#4eb5de;border-color:#4eb5de;color:#fff}.button.is-primary.is-outlined.is-loading::after,.docstring>section>a.button.is-outlined.is-loading.docs-sourcelink::after{border-color:transparent transparent #4eb5de #4eb5de !important}.button.is-primary.is-outlined.is-loading:hover::after,.docstring>section>a.button.is-outlined.is-loading.docs-sourcelink:hover::after,.button.is-primary.is-outlined.is-loading.is-hovered::after,.docstring>section>a.button.is-outlined.is-loading.is-hovered.docs-sourcelink::after,.button.is-primary.is-outlined.is-loading:focus::after,.docstring>section>a.button.is-outlined.is-loading.docs-sourcelink:focus::after,.button.is-primary.is-outlined.is-loading.is-focused::after,.docstring>section>a.button.is-outlined.is-loading.is-focused.docs-sourcelink::after{border-color:transparent transparent #fff #fff !important}.button.is-primary.is-outlined[disabled],.docstring>section>a.button.is-outlined.docs-sourcelink[disabled],fieldset[disabled] .button.is-primary.is-outlined,fieldset[disabled] .docstring>section>a.button.is-outlined.docs-sourcelink{background-color:transparent;border-color:#4eb5de;box-shadow:none;color:#4eb5de}.button.is-primary.is-inverted.is-outlined,.docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink{background-color:transparent;border-color:#fff;color:#fff}.button.is-primary.is-inverted.is-outlined:hover,.docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink:hover,.button.is-primary.is-inverted.is-outlined.is-hovered,.docstring>section>a.button.is-inverted.is-outlined.is-hovered.docs-sourcelink,.button.is-primary.is-inverted.is-outlined:focus,.docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink:focus,.button.is-primary.is-inverted.is-outlined.is-focused,.docstring>section>a.button.is-inverted.is-outlined.is-focused.docs-sourcelink{background-color:#fff;color:#4eb5de}.button.is-primary.is-inverted.is-outlined.is-loading:hover::after,.docstring>section>a.button.is-inverted.is-outlined.is-loading.docs-sourcelink:hover::after,.button.is-primary.is-inverted.is-outlined.is-loading.is-hovered::after,.docstring>section>a.button.is-inverted.is-outlined.is-loading.is-hovered.docs-sourcelink::after,.button.is-primary.is-inverted.is-outlined.is-loading:focus::after,.docstring>section>a.button.is-inverted.is-outlined.is-loading.docs-sourcelink:focus::after,.button.is-primary.is-inverted.is-outlined.is-loading.is-focused::after,.docstring>section>a.button.is-inverted.is-outlined.is-loading.is-focused.docs-sourcelink::after{border-color:transparent transparent #4eb5de #4eb5de !important}.button.is-primary.is-inverted.is-outlined[disabled],.docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink[disabled],fieldset[disabled] .button.is-primary.is-inverted.is-outlined,fieldset[disabled] .docstring>section>a.button.is-inverted.is-outlined.docs-sourcelink{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-primary.is-light,.docstring>section>a.button.is-light.docs-sourcelink{background-color:#eef8fc;color:#1a6d8e}.button.is-primary.is-light:hover,.docstring>section>a.button.is-light.docs-sourcelink:hover,.button.is-primary.is-light.is-hovered,.docstring>section>a.button.is-light.is-hovered.docs-sourcelink{background-color:#e3f3fa;border-color:transparent;color:#1a6d8e}.button.is-primary.is-light:active,.docstring>section>a.button.is-light.docs-sourcelink:active,.button.is-primary.is-light.is-active,.docstring>section>a.button.is-light.is-active.docs-sourcelink{background-color:#d8eff8;border-color:transparent;color:#1a6d8e}.button.is-link{background-color:#2e63b8;border-color:transparent;color:#fff}.button.is-link:hover,.button.is-link.is-hovered{background-color:#2b5eae;border-color:transparent;color:#fff}.button.is-link:focus,.button.is-link.is-focused{border-color:transparent;color:#fff}.button.is-link:focus:not(:active),.button.is-link.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(46,99,184,0.25)}.button.is-link:active,.button.is-link.is-active{background-color:#2958a4;border-color:transparent;color:#fff}.button.is-link[disabled],fieldset[disabled] .button.is-link{background-color:#2e63b8;border-color:#2e63b8;box-shadow:none}.button.is-link.is-inverted{background-color:#fff;color:#2e63b8}.button.is-link.is-inverted:hover,.button.is-link.is-inverted.is-hovered{background-color:#f2f2f2}.button.is-link.is-inverted[disabled],fieldset[disabled] .button.is-link.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#2e63b8}.button.is-link.is-loading::after{border-color:transparent transparent #fff #fff !important}.button.is-link.is-outlined{background-color:transparent;border-color:#2e63b8;color:#2e63b8}.button.is-link.is-outlined:hover,.button.is-link.is-outlined.is-hovered,.button.is-link.is-outlined:focus,.button.is-link.is-outlined.is-focused{background-color:#2e63b8;border-color:#2e63b8;color:#fff}.button.is-link.is-outlined.is-loading::after{border-color:transparent transparent #2e63b8 #2e63b8 !important}.button.is-link.is-outlined.is-loading:hover::after,.button.is-link.is-outlined.is-loading.is-hovered::after,.button.is-link.is-outlined.is-loading:focus::after,.button.is-link.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}.button.is-link.is-outlined[disabled],fieldset[disabled] .button.is-link.is-outlined{background-color:transparent;border-color:#2e63b8;box-shadow:none;color:#2e63b8}.button.is-link.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}.button.is-link.is-inverted.is-outlined:hover,.button.is-link.is-inverted.is-outlined.is-hovered,.button.is-link.is-inverted.is-outlined:focus,.button.is-link.is-inverted.is-outlined.is-focused{background-color:#fff;color:#2e63b8}.button.is-link.is-inverted.is-outlined.is-loading:hover::after,.button.is-link.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-link.is-inverted.is-outlined.is-loading:focus::after,.button.is-link.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #2e63b8 #2e63b8 !important}.button.is-link.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-link.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-link.is-light{background-color:#eff3fb;color:#3169c4}.button.is-link.is-light:hover,.button.is-link.is-light.is-hovered{background-color:#e4ecf8;border-color:transparent;color:#3169c4}.button.is-link.is-light:active,.button.is-link.is-light.is-active{background-color:#dae5f6;border-color:transparent;color:#3169c4}.button.is-info{background-color:#209cee;border-color:transparent;color:#fff}.button.is-info:hover,.button.is-info.is-hovered{background-color:#1497ed;border-color:transparent;color:#fff}.button.is-info:focus,.button.is-info.is-focused{border-color:transparent;color:#fff}.button.is-info:focus:not(:active),.button.is-info.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(32,156,238,0.25)}.button.is-info:active,.button.is-info.is-active{background-color:#1190e3;border-color:transparent;color:#fff}.button.is-info[disabled],fieldset[disabled] .button.is-info{background-color:#209cee;border-color:#209cee;box-shadow:none}.button.is-info.is-inverted{background-color:#fff;color:#209cee}.button.is-info.is-inverted:hover,.button.is-info.is-inverted.is-hovered{background-color:#f2f2f2}.button.is-info.is-inverted[disabled],fieldset[disabled] .button.is-info.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#209cee}.button.is-info.is-loading::after{border-color:transparent transparent #fff #fff !important}.button.is-info.is-outlined{background-color:transparent;border-color:#209cee;color:#209cee}.button.is-info.is-outlined:hover,.button.is-info.is-outlined.is-hovered,.button.is-info.is-outlined:focus,.button.is-info.is-outlined.is-focused{background-color:#209cee;border-color:#209cee;color:#fff}.button.is-info.is-outlined.is-loading::after{border-color:transparent transparent #209cee #209cee !important}.button.is-info.is-outlined.is-loading:hover::after,.button.is-info.is-outlined.is-loading.is-hovered::after,.button.is-info.is-outlined.is-loading:focus::after,.button.is-info.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}.button.is-info.is-outlined[disabled],fieldset[disabled] .button.is-info.is-outlined{background-color:transparent;border-color:#209cee;box-shadow:none;color:#209cee}.button.is-info.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}.button.is-info.is-inverted.is-outlined:hover,.button.is-info.is-inverted.is-outlined.is-hovered,.button.is-info.is-inverted.is-outlined:focus,.button.is-info.is-inverted.is-outlined.is-focused{background-color:#fff;color:#209cee}.button.is-info.is-inverted.is-outlined.is-loading:hover::after,.button.is-info.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-info.is-inverted.is-outlined.is-loading:focus::after,.button.is-info.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #209cee #209cee !important}.button.is-info.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-info.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-info.is-light{background-color:#ecf7fe;color:#0e72b4}.button.is-info.is-light:hover,.button.is-info.is-light.is-hovered{background-color:#e0f1fd;border-color:transparent;color:#0e72b4}.button.is-info.is-light:active,.button.is-info.is-light.is-active{background-color:#d4ecfc;border-color:transparent;color:#0e72b4}.button.is-success{background-color:#22c35b;border-color:transparent;color:#fff}.button.is-success:hover,.button.is-success.is-hovered{background-color:#20b856;border-color:transparent;color:#fff}.button.is-success:focus,.button.is-success.is-focused{border-color:transparent;color:#fff}.button.is-success:focus:not(:active),.button.is-success.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(34,195,91,0.25)}.button.is-success:active,.button.is-success.is-active{background-color:#1ead51;border-color:transparent;color:#fff}.button.is-success[disabled],fieldset[disabled] .button.is-success{background-color:#22c35b;border-color:#22c35b;box-shadow:none}.button.is-success.is-inverted{background-color:#fff;color:#22c35b}.button.is-success.is-inverted:hover,.button.is-success.is-inverted.is-hovered{background-color:#f2f2f2}.button.is-success.is-inverted[disabled],fieldset[disabled] .button.is-success.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#22c35b}.button.is-success.is-loading::after{border-color:transparent transparent #fff #fff !important}.button.is-success.is-outlined{background-color:transparent;border-color:#22c35b;color:#22c35b}.button.is-success.is-outlined:hover,.button.is-success.is-outlined.is-hovered,.button.is-success.is-outlined:focus,.button.is-success.is-outlined.is-focused{background-color:#22c35b;border-color:#22c35b;color:#fff}.button.is-success.is-outlined.is-loading::after{border-color:transparent transparent #22c35b #22c35b !important}.button.is-success.is-outlined.is-loading:hover::after,.button.is-success.is-outlined.is-loading.is-hovered::after,.button.is-success.is-outlined.is-loading:focus::after,.button.is-success.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}.button.is-success.is-outlined[disabled],fieldset[disabled] .button.is-success.is-outlined{background-color:transparent;border-color:#22c35b;box-shadow:none;color:#22c35b}.button.is-success.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}.button.is-success.is-inverted.is-outlined:hover,.button.is-success.is-inverted.is-outlined.is-hovered,.button.is-success.is-inverted.is-outlined:focus,.button.is-success.is-inverted.is-outlined.is-focused{background-color:#fff;color:#22c35b}.button.is-success.is-inverted.is-outlined.is-loading:hover::after,.button.is-success.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-success.is-inverted.is-outlined.is-loading:focus::after,.button.is-success.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #22c35b #22c35b !important}.button.is-success.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-success.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-success.is-light{background-color:#eefcf3;color:#198f43}.button.is-success.is-light:hover,.button.is-success.is-light.is-hovered{background-color:#e3faeb;border-color:transparent;color:#198f43}.button.is-success.is-light:active,.button.is-success.is-light.is-active{background-color:#d8f8e3;border-color:transparent;color:#198f43}.button.is-warning{background-color:#ffdd57;border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-warning:hover,.button.is-warning.is-hovered{background-color:#ffda4a;border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-warning:focus,.button.is-warning.is-focused{border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-warning:focus:not(:active),.button.is-warning.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(255,221,87,0.25)}.button.is-warning:active,.button.is-warning.is-active{background-color:#ffd83e;border-color:transparent;color:rgba(0,0,0,0.7)}.button.is-warning[disabled],fieldset[disabled] .button.is-warning{background-color:#ffdd57;border-color:#ffdd57;box-shadow:none}.button.is-warning.is-inverted{background-color:rgba(0,0,0,0.7);color:#ffdd57}.button.is-warning.is-inverted:hover,.button.is-warning.is-inverted.is-hovered{background-color:rgba(0,0,0,0.7)}.button.is-warning.is-inverted[disabled],fieldset[disabled] .button.is-warning.is-inverted{background-color:rgba(0,0,0,0.7);border-color:transparent;box-shadow:none;color:#ffdd57}.button.is-warning.is-loading::after{border-color:transparent transparent rgba(0,0,0,0.7) rgba(0,0,0,0.7) !important}.button.is-warning.is-outlined{background-color:transparent;border-color:#ffdd57;color:#ffdd57}.button.is-warning.is-outlined:hover,.button.is-warning.is-outlined.is-hovered,.button.is-warning.is-outlined:focus,.button.is-warning.is-outlined.is-focused{background-color:#ffdd57;border-color:#ffdd57;color:rgba(0,0,0,0.7)}.button.is-warning.is-outlined.is-loading::after{border-color:transparent transparent #ffdd57 #ffdd57 !important}.button.is-warning.is-outlined.is-loading:hover::after,.button.is-warning.is-outlined.is-loading.is-hovered::after,.button.is-warning.is-outlined.is-loading:focus::after,.button.is-warning.is-outlined.is-loading.is-focused::after{border-color:transparent transparent rgba(0,0,0,0.7) rgba(0,0,0,0.7) !important}.button.is-warning.is-outlined[disabled],fieldset[disabled] .button.is-warning.is-outlined{background-color:transparent;border-color:#ffdd57;box-shadow:none;color:#ffdd57}.button.is-warning.is-inverted.is-outlined{background-color:transparent;border-color:rgba(0,0,0,0.7);color:rgba(0,0,0,0.7)}.button.is-warning.is-inverted.is-outlined:hover,.button.is-warning.is-inverted.is-outlined.is-hovered,.button.is-warning.is-inverted.is-outlined:focus,.button.is-warning.is-inverted.is-outlined.is-focused{background-color:rgba(0,0,0,0.7);color:#ffdd57}.button.is-warning.is-inverted.is-outlined.is-loading:hover::after,.button.is-warning.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-warning.is-inverted.is-outlined.is-loading:focus::after,.button.is-warning.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #ffdd57 #ffdd57 !important}.button.is-warning.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-warning.is-inverted.is-outlined{background-color:transparent;border-color:rgba(0,0,0,0.7);box-shadow:none;color:rgba(0,0,0,0.7)}.button.is-warning.is-light{background-color:#fffbeb;color:#947600}.button.is-warning.is-light:hover,.button.is-warning.is-light.is-hovered{background-color:#fff8de;border-color:transparent;color:#947600}.button.is-warning.is-light:active,.button.is-warning.is-light.is-active{background-color:#fff6d1;border-color:transparent;color:#947600}.button.is-danger{background-color:#da0b00;border-color:transparent;color:#fff}.button.is-danger:hover,.button.is-danger.is-hovered{background-color:#cd0a00;border-color:transparent;color:#fff}.button.is-danger:focus,.button.is-danger.is-focused{border-color:transparent;color:#fff}.button.is-danger:focus:not(:active),.button.is-danger.is-focused:not(:active){box-shadow:0 0 0 0.125em rgba(218,11,0,0.25)}.button.is-danger:active,.button.is-danger.is-active{background-color:#c10a00;border-color:transparent;color:#fff}.button.is-danger[disabled],fieldset[disabled] .button.is-danger{background-color:#da0b00;border-color:#da0b00;box-shadow:none}.button.is-danger.is-inverted{background-color:#fff;color:#da0b00}.button.is-danger.is-inverted:hover,.button.is-danger.is-inverted.is-hovered{background-color:#f2f2f2}.button.is-danger.is-inverted[disabled],fieldset[disabled] .button.is-danger.is-inverted{background-color:#fff;border-color:transparent;box-shadow:none;color:#da0b00}.button.is-danger.is-loading::after{border-color:transparent transparent #fff #fff !important}.button.is-danger.is-outlined{background-color:transparent;border-color:#da0b00;color:#da0b00}.button.is-danger.is-outlined:hover,.button.is-danger.is-outlined.is-hovered,.button.is-danger.is-outlined:focus,.button.is-danger.is-outlined.is-focused{background-color:#da0b00;border-color:#da0b00;color:#fff}.button.is-danger.is-outlined.is-loading::after{border-color:transparent transparent #da0b00 #da0b00 !important}.button.is-danger.is-outlined.is-loading:hover::after,.button.is-danger.is-outlined.is-loading.is-hovered::after,.button.is-danger.is-outlined.is-loading:focus::after,.button.is-danger.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #fff #fff !important}.button.is-danger.is-outlined[disabled],fieldset[disabled] .button.is-danger.is-outlined{background-color:transparent;border-color:#da0b00;box-shadow:none;color:#da0b00}.button.is-danger.is-inverted.is-outlined{background-color:transparent;border-color:#fff;color:#fff}.button.is-danger.is-inverted.is-outlined:hover,.button.is-danger.is-inverted.is-outlined.is-hovered,.button.is-danger.is-inverted.is-outlined:focus,.button.is-danger.is-inverted.is-outlined.is-focused{background-color:#fff;color:#da0b00}.button.is-danger.is-inverted.is-outlined.is-loading:hover::after,.button.is-danger.is-inverted.is-outlined.is-loading.is-hovered::after,.button.is-danger.is-inverted.is-outlined.is-loading:focus::after,.button.is-danger.is-inverted.is-outlined.is-loading.is-focused::after{border-color:transparent transparent #da0b00 #da0b00 !important}.button.is-danger.is-inverted.is-outlined[disabled],fieldset[disabled] .button.is-danger.is-inverted.is-outlined{background-color:transparent;border-color:#fff;box-shadow:none;color:#fff}.button.is-danger.is-light{background-color:#ffeceb;color:#f50c00}.button.is-danger.is-light:hover,.button.is-danger.is-light.is-hovered{background-color:#ffe0de;border-color:transparent;color:#f50c00}.button.is-danger.is-light:active,.button.is-danger.is-light.is-active{background-color:#ffd3d1;border-color:transparent;color:#f50c00}.button.is-small,#documenter .docs-sidebar form.docs-search>input.button{font-size:.75rem}.button.is-small:not(.is-rounded),#documenter .docs-sidebar form.docs-search>input.button:not(.is-rounded){border-radius:2px}.button.is-normal{font-size:1rem}.button.is-medium{font-size:1.25rem}.button.is-large{font-size:1.5rem}.button[disabled],fieldset[disabled] .button{background-color:#fff;border-color:#dbdbdb;box-shadow:none;opacity:.5}.button.is-fullwidth{display:flex;width:100%}.button.is-loading{color:transparent !important;pointer-events:none}.button.is-loading::after{position:absolute;left:calc(50% - (1em * 0.5));top:calc(50% - (1em * 0.5));position:absolute !important}.button.is-static{background-color:#f5f5f5;border-color:#dbdbdb;color:#6b6b6b;box-shadow:none;pointer-events:none}.button.is-rounded,#documenter .docs-sidebar form.docs-search>input.button{border-radius:9999px;padding-left:calc(1em + 0.25em);padding-right:calc(1em + 0.25em)}.buttons{align-items:center;display:flex;flex-wrap:wrap;justify-content:flex-start}.buttons .button{margin-bottom:0.5rem}.buttons .button:not(:last-child):not(.is-fullwidth){margin-right:.5rem}.buttons:last-child{margin-bottom:-0.5rem}.buttons:not(:last-child){margin-bottom:1rem}.buttons.are-small .button:not(.is-normal):not(.is-medium):not(.is-large){font-size:.75rem}.buttons.are-small .button:not(.is-normal):not(.is-medium):not(.is-large):not(.is-rounded){border-radius:2px}.buttons.are-medium .button:not(.is-small):not(.is-normal):not(.is-large){font-size:1.25rem}.buttons.are-large .button:not(.is-small):not(.is-normal):not(.is-medium){font-size:1.5rem}.buttons.has-addons .button:not(:first-child){border-bottom-left-radius:0;border-top-left-radius:0}.buttons.has-addons .button:not(:last-child){border-bottom-right-radius:0;border-top-right-radius:0;margin-right:-1px}.buttons.has-addons .button:last-child{margin-right:0}.buttons.has-addons .button:hover,.buttons.has-addons .button.is-hovered{z-index:2}.buttons.has-addons .button:focus,.buttons.has-addons .button.is-focused,.buttons.has-addons .button:active,.buttons.has-addons .button.is-active,.buttons.has-addons .button.is-selected{z-index:3}.buttons.has-addons .button:focus:hover,.buttons.has-addons .button.is-focused:hover,.buttons.has-addons .button:active:hover,.buttons.has-addons .button.is-active:hover,.buttons.has-addons .button.is-selected:hover{z-index:4}.buttons.has-addons .button.is-expanded{flex-grow:1;flex-shrink:1}.buttons.is-centered{justify-content:center}.buttons.is-centered:not(.has-addons) .button:not(.is-fullwidth){margin-left:0.25rem;margin-right:0.25rem}.buttons.is-right{justify-content:flex-end}.buttons.is-right:not(.has-addons) .button:not(.is-fullwidth){margin-left:0.25rem;margin-right:0.25rem}@media screen and (max-width: 768px){.button.is-responsive.is-small,#documenter .docs-sidebar form.docs-search>input.is-responsive{font-size:.5625rem}.button.is-responsive,.button.is-responsive.is-normal{font-size:.65625rem}.button.is-responsive.is-medium{font-size:.75rem}.button.is-responsive.is-large{font-size:1rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.button.is-responsive.is-small,#documenter .docs-sidebar form.docs-search>input.is-responsive{font-size:.65625rem}.button.is-responsive,.button.is-responsive.is-normal{font-size:.75rem}.button.is-responsive.is-medium{font-size:1rem}.button.is-responsive.is-large{font-size:1.25rem}}.container{flex-grow:1;margin:0 auto;position:relative;width:auto}.container.is-fluid{max-width:none !important;padding-left:32px;padding-right:32px;width:100%}@media screen and (min-width: 1056px){.container{max-width:992px}}@media screen and (max-width: 1215px){.container.is-widescreen:not(.is-max-desktop){max-width:1152px}}@media screen and (max-width: 1407px){.container.is-fullhd:not(.is-max-desktop):not(.is-max-widescreen){max-width:1344px}}@media screen and (min-width: 1216px){.container:not(.is-max-desktop){max-width:1152px}}@media screen and (min-width: 1408px){.container:not(.is-max-desktop):not(.is-max-widescreen){max-width:1344px}}.content li+li{margin-top:0.25em}.content p:not(:last-child),.content dl:not(:last-child),.content ol:not(:last-child),.content ul:not(:last-child),.content blockquote:not(:last-child),.content pre:not(:last-child),.content table:not(:last-child){margin-bottom:1em}.content h1,.content h2,.content h3,.content h4,.content h5,.content h6{color:#222;font-weight:600;line-height:1.125}.content h1{font-size:2em;margin-bottom:0.5em}.content h1:not(:first-child){margin-top:1em}.content h2{font-size:1.75em;margin-bottom:0.5714em}.content h2:not(:first-child){margin-top:1.1428em}.content h3{font-size:1.5em;margin-bottom:0.6666em}.content h3:not(:first-child){margin-top:1.3333em}.content h4{font-size:1.25em;margin-bottom:0.8em}.content h5{font-size:1.125em;margin-bottom:0.8888em}.content h6{font-size:1em;margin-bottom:1em}.content blockquote{background-color:#f5f5f5;border-left:5px solid #dbdbdb;padding:1.25em 1.5em}.content ol{list-style-position:outside;margin-left:2em;margin-top:1em}.content ol:not([type]){list-style-type:decimal}.content ol.is-lower-alpha:not([type]){list-style-type:lower-alpha}.content ol.is-lower-roman:not([type]){list-style-type:lower-roman}.content ol.is-upper-alpha:not([type]){list-style-type:upper-alpha}.content ol.is-upper-roman:not([type]){list-style-type:upper-roman}.content ul{list-style:disc outside;margin-left:2em;margin-top:1em}.content ul ul{list-style-type:circle;margin-top:0.5em}.content ul ul ul{list-style-type:square}.content dd{margin-left:2em}.content figure{margin-left:2em;margin-right:2em;text-align:center}.content figure:not(:first-child){margin-top:2em}.content figure:not(:last-child){margin-bottom:2em}.content figure img{display:inline-block}.content figure figcaption{font-style:italic}.content pre{-webkit-overflow-scrolling:touch;overflow-x:auto;padding:0;white-space:pre;word-wrap:normal}.content sup,.content sub{font-size:75%}.content table{width:100%}.content table td,.content table th{border:1px solid #dbdbdb;border-width:0 0 1px;padding:0.5em 0.75em;vertical-align:top}.content table th{color:#222}.content table th:not([align]){text-align:inherit}.content table thead td,.content table thead th{border-width:0 0 2px;color:#222}.content table tfoot td,.content table tfoot th{border-width:2px 0 0;color:#222}.content table tbody tr:last-child td,.content table tbody tr:last-child th{border-bottom-width:0}.content .tabs li+li{margin-top:0}.content.is-small,#documenter .docs-sidebar form.docs-search>input.content{font-size:.75rem}.content.is-normal{font-size:1rem}.content.is-medium{font-size:1.25rem}.content.is-large{font-size:1.5rem}.icon{align-items:center;display:inline-flex;justify-content:center;height:1.5rem;width:1.5rem}.icon.is-small,#documenter .docs-sidebar form.docs-search>input.icon{height:1rem;width:1rem}.icon.is-medium{height:2rem;width:2rem}.icon.is-large{height:3rem;width:3rem}.icon-text{align-items:flex-start;color:inherit;display:inline-flex;flex-wrap:wrap;line-height:1.5rem;vertical-align:top}.icon-text .icon{flex-grow:0;flex-shrink:0}.icon-text .icon:not(:last-child){margin-right:.25em}.icon-text .icon:not(:first-child){margin-left:.25em}div.icon-text{display:flex}.image,#documenter .docs-sidebar .docs-logo>img{display:block;position:relative}.image img,#documenter .docs-sidebar .docs-logo>img img{display:block;height:auto;width:100%}.image img.is-rounded,#documenter .docs-sidebar .docs-logo>img img.is-rounded{border-radius:9999px}.image.is-fullwidth,#documenter .docs-sidebar .docs-logo>img.is-fullwidth{width:100%}.image.is-square img,#documenter .docs-sidebar .docs-logo>img.is-square img,.image.is-square .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-square .has-ratio,.image.is-1by1 img,#documenter .docs-sidebar .docs-logo>img.is-1by1 img,.image.is-1by1 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-1by1 .has-ratio,.image.is-5by4 img,#documenter .docs-sidebar .docs-logo>img.is-5by4 img,.image.is-5by4 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-5by4 .has-ratio,.image.is-4by3 img,#documenter .docs-sidebar .docs-logo>img.is-4by3 img,.image.is-4by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-4by3 .has-ratio,.image.is-3by2 img,#documenter .docs-sidebar .docs-logo>img.is-3by2 img,.image.is-3by2 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by2 .has-ratio,.image.is-5by3 img,#documenter .docs-sidebar .docs-logo>img.is-5by3 img,.image.is-5by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-5by3 .has-ratio,.image.is-16by9 img,#documenter .docs-sidebar .docs-logo>img.is-16by9 img,.image.is-16by9 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-16by9 .has-ratio,.image.is-2by1 img,#documenter .docs-sidebar .docs-logo>img.is-2by1 img,.image.is-2by1 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-2by1 .has-ratio,.image.is-3by1 img,#documenter .docs-sidebar .docs-logo>img.is-3by1 img,.image.is-3by1 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by1 .has-ratio,.image.is-4by5 img,#documenter .docs-sidebar .docs-logo>img.is-4by5 img,.image.is-4by5 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-4by5 .has-ratio,.image.is-3by4 img,#documenter .docs-sidebar .docs-logo>img.is-3by4 img,.image.is-3by4 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by4 .has-ratio,.image.is-2by3 img,#documenter .docs-sidebar .docs-logo>img.is-2by3 img,.image.is-2by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-2by3 .has-ratio,.image.is-3by5 img,#documenter .docs-sidebar .docs-logo>img.is-3by5 img,.image.is-3by5 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-3by5 .has-ratio,.image.is-9by16 img,#documenter .docs-sidebar .docs-logo>img.is-9by16 img,.image.is-9by16 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-9by16 .has-ratio,.image.is-1by2 img,#documenter .docs-sidebar .docs-logo>img.is-1by2 img,.image.is-1by2 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-1by2 .has-ratio,.image.is-1by3 img,#documenter .docs-sidebar .docs-logo>img.is-1by3 img,.image.is-1by3 .has-ratio,#documenter .docs-sidebar .docs-logo>img.is-1by3 .has-ratio{height:100%;width:100%}.image.is-square,#documenter .docs-sidebar .docs-logo>img.is-square,.image.is-1by1,#documenter .docs-sidebar .docs-logo>img.is-1by1{padding-top:100%}.image.is-5by4,#documenter .docs-sidebar .docs-logo>img.is-5by4{padding-top:80%}.image.is-4by3,#documenter .docs-sidebar .docs-logo>img.is-4by3{padding-top:75%}.image.is-3by2,#documenter .docs-sidebar .docs-logo>img.is-3by2{padding-top:66.6666%}.image.is-5by3,#documenter .docs-sidebar .docs-logo>img.is-5by3{padding-top:60%}.image.is-16by9,#documenter .docs-sidebar .docs-logo>img.is-16by9{padding-top:56.25%}.image.is-2by1,#documenter .docs-sidebar .docs-logo>img.is-2by1{padding-top:50%}.image.is-3by1,#documenter .docs-sidebar .docs-logo>img.is-3by1{padding-top:33.3333%}.image.is-4by5,#documenter .docs-sidebar .docs-logo>img.is-4by5{padding-top:125%}.image.is-3by4,#documenter .docs-sidebar .docs-logo>img.is-3by4{padding-top:133.3333%}.image.is-2by3,#documenter .docs-sidebar .docs-logo>img.is-2by3{padding-top:150%}.image.is-3by5,#documenter .docs-sidebar .docs-logo>img.is-3by5{padding-top:166.6666%}.image.is-9by16,#documenter .docs-sidebar .docs-logo>img.is-9by16{padding-top:177.7777%}.image.is-1by2,#documenter .docs-sidebar .docs-logo>img.is-1by2{padding-top:200%}.image.is-1by3,#documenter .docs-sidebar .docs-logo>img.is-1by3{padding-top:300%}.image.is-16x16,#documenter .docs-sidebar .docs-logo>img.is-16x16{height:16px;width:16px}.image.is-24x24,#documenter .docs-sidebar .docs-logo>img.is-24x24{height:24px;width:24px}.image.is-32x32,#documenter .docs-sidebar .docs-logo>img.is-32x32{height:32px;width:32px}.image.is-48x48,#documenter .docs-sidebar .docs-logo>img.is-48x48{height:48px;width:48px}.image.is-64x64,#documenter .docs-sidebar .docs-logo>img.is-64x64{height:64px;width:64px}.image.is-96x96,#documenter .docs-sidebar .docs-logo>img.is-96x96{height:96px;width:96px}.image.is-128x128,#documenter .docs-sidebar .docs-logo>img.is-128x128{height:128px;width:128px}.notification{background-color:#f5f5f5;border-radius:4px;position:relative;padding:1.25rem 2.5rem 1.25rem 1.5rem}.notification a:not(.button):not(.dropdown-item){color:currentColor;text-decoration:underline}.notification strong{color:currentColor}.notification code,.notification pre{background:#fff}.notification pre code{background:transparent}.notification>.delete{right:.5rem;position:absolute;top:0.5rem}.notification .title,.notification .subtitle,.notification .content{color:currentColor}.notification.is-white{background-color:#fff;color:#0a0a0a}.notification.is-black{background-color:#0a0a0a;color:#fff}.notification.is-light{background-color:#f5f5f5;color:rgba(0,0,0,0.7)}.notification.is-dark,.content kbd.notification{background-color:#363636;color:#fff}.notification.is-primary,.docstring>section>a.notification.docs-sourcelink{background-color:#4eb5de;color:#fff}.notification.is-primary.is-light,.docstring>section>a.notification.is-light.docs-sourcelink{background-color:#eef8fc;color:#1a6d8e}.notification.is-link{background-color:#2e63b8;color:#fff}.notification.is-link.is-light{background-color:#eff3fb;color:#3169c4}.notification.is-info{background-color:#209cee;color:#fff}.notification.is-info.is-light{background-color:#ecf7fe;color:#0e72b4}.notification.is-success{background-color:#22c35b;color:#fff}.notification.is-success.is-light{background-color:#eefcf3;color:#198f43}.notification.is-warning{background-color:#ffdd57;color:rgba(0,0,0,0.7)}.notification.is-warning.is-light{background-color:#fffbeb;color:#947600}.notification.is-danger{background-color:#da0b00;color:#fff}.notification.is-danger.is-light{background-color:#ffeceb;color:#f50c00}.progress{-moz-appearance:none;-webkit-appearance:none;border:none;border-radius:9999px;display:block;height:1rem;overflow:hidden;padding:0;width:100%}.progress::-webkit-progress-bar{background-color:#ededed}.progress::-webkit-progress-value{background-color:#222}.progress::-moz-progress-bar{background-color:#222}.progress::-ms-fill{background-color:#222;border:none}.progress.is-white::-webkit-progress-value{background-color:#fff}.progress.is-white::-moz-progress-bar{background-color:#fff}.progress.is-white::-ms-fill{background-color:#fff}.progress.is-white:indeterminate{background-image:linear-gradient(to right, #fff 30%, #ededed 30%)}.progress.is-black::-webkit-progress-value{background-color:#0a0a0a}.progress.is-black::-moz-progress-bar{background-color:#0a0a0a}.progress.is-black::-ms-fill{background-color:#0a0a0a}.progress.is-black:indeterminate{background-image:linear-gradient(to right, #0a0a0a 30%, #ededed 30%)}.progress.is-light::-webkit-progress-value{background-color:#f5f5f5}.progress.is-light::-moz-progress-bar{background-color:#f5f5f5}.progress.is-light::-ms-fill{background-color:#f5f5f5}.progress.is-light:indeterminate{background-image:linear-gradient(to right, #f5f5f5 30%, #ededed 30%)}.progress.is-dark::-webkit-progress-value,.content kbd.progress::-webkit-progress-value{background-color:#363636}.progress.is-dark::-moz-progress-bar,.content kbd.progress::-moz-progress-bar{background-color:#363636}.progress.is-dark::-ms-fill,.content kbd.progress::-ms-fill{background-color:#363636}.progress.is-dark:indeterminate,.content kbd.progress:indeterminate{background-image:linear-gradient(to right, #363636 30%, #ededed 30%)}.progress.is-primary::-webkit-progress-value,.docstring>section>a.progress.docs-sourcelink::-webkit-progress-value{background-color:#4eb5de}.progress.is-primary::-moz-progress-bar,.docstring>section>a.progress.docs-sourcelink::-moz-progress-bar{background-color:#4eb5de}.progress.is-primary::-ms-fill,.docstring>section>a.progress.docs-sourcelink::-ms-fill{background-color:#4eb5de}.progress.is-primary:indeterminate,.docstring>section>a.progress.docs-sourcelink:indeterminate{background-image:linear-gradient(to right, #4eb5de 30%, #ededed 30%)}.progress.is-link::-webkit-progress-value{background-color:#2e63b8}.progress.is-link::-moz-progress-bar{background-color:#2e63b8}.progress.is-link::-ms-fill{background-color:#2e63b8}.progress.is-link:indeterminate{background-image:linear-gradient(to right, #2e63b8 30%, #ededed 30%)}.progress.is-info::-webkit-progress-value{background-color:#209cee}.progress.is-info::-moz-progress-bar{background-color:#209cee}.progress.is-info::-ms-fill{background-color:#209cee}.progress.is-info:indeterminate{background-image:linear-gradient(to right, #209cee 30%, #ededed 30%)}.progress.is-success::-webkit-progress-value{background-color:#22c35b}.progress.is-success::-moz-progress-bar{background-color:#22c35b}.progress.is-success::-ms-fill{background-color:#22c35b}.progress.is-success:indeterminate{background-image:linear-gradient(to right, #22c35b 30%, #ededed 30%)}.progress.is-warning::-webkit-progress-value{background-color:#ffdd57}.progress.is-warning::-moz-progress-bar{background-color:#ffdd57}.progress.is-warning::-ms-fill{background-color:#ffdd57}.progress.is-warning:indeterminate{background-image:linear-gradient(to right, #ffdd57 30%, #ededed 30%)}.progress.is-danger::-webkit-progress-value{background-color:#da0b00}.progress.is-danger::-moz-progress-bar{background-color:#da0b00}.progress.is-danger::-ms-fill{background-color:#da0b00}.progress.is-danger:indeterminate{background-image:linear-gradient(to right, #da0b00 30%, #ededed 30%)}.progress:indeterminate{animation-duration:1.5s;animation-iteration-count:infinite;animation-name:moveIndeterminate;animation-timing-function:linear;background-color:#ededed;background-image:linear-gradient(to right, #222 30%, #ededed 30%);background-position:top left;background-repeat:no-repeat;background-size:150% 150%}.progress:indeterminate::-webkit-progress-bar{background-color:transparent}.progress:indeterminate::-moz-progress-bar{background-color:transparent}.progress:indeterminate::-ms-fill{animation-name:none}.progress.is-small,#documenter .docs-sidebar form.docs-search>input.progress{height:.75rem}.progress.is-medium{height:1.25rem}.progress.is-large{height:1.5rem}@keyframes moveIndeterminate{from{background-position:200% 0}to{background-position:-200% 0}}.table{background-color:#fff;color:#222}.table td,.table th{border:1px solid #dbdbdb;border-width:0 0 1px;padding:0.5em 0.75em;vertical-align:top}.table td.is-white,.table th.is-white{background-color:#fff;border-color:#fff;color:#0a0a0a}.table td.is-black,.table th.is-black{background-color:#0a0a0a;border-color:#0a0a0a;color:#fff}.table td.is-light,.table th.is-light{background-color:#f5f5f5;border-color:#f5f5f5;color:rgba(0,0,0,0.7)}.table td.is-dark,.table th.is-dark{background-color:#363636;border-color:#363636;color:#fff}.table td.is-primary,.table th.is-primary{background-color:#4eb5de;border-color:#4eb5de;color:#fff}.table td.is-link,.table th.is-link{background-color:#2e63b8;border-color:#2e63b8;color:#fff}.table td.is-info,.table th.is-info{background-color:#209cee;border-color:#209cee;color:#fff}.table td.is-success,.table th.is-success{background-color:#22c35b;border-color:#22c35b;color:#fff}.table td.is-warning,.table th.is-warning{background-color:#ffdd57;border-color:#ffdd57;color:rgba(0,0,0,0.7)}.table td.is-danger,.table th.is-danger{background-color:#da0b00;border-color:#da0b00;color:#fff}.table td.is-narrow,.table th.is-narrow{white-space:nowrap;width:1%}.table td.is-selected,.table th.is-selected{background-color:#4eb5de;color:#fff}.table td.is-selected a,.table td.is-selected strong,.table th.is-selected a,.table th.is-selected strong{color:currentColor}.table td.is-vcentered,.table th.is-vcentered{vertical-align:middle}.table th{color:#222}.table th:not([align]){text-align:left}.table tr.is-selected{background-color:#4eb5de;color:#fff}.table tr.is-selected a,.table tr.is-selected strong{color:currentColor}.table tr.is-selected td,.table tr.is-selected th{border-color:#fff;color:currentColor}.table thead{background-color:rgba(0,0,0,0)}.table thead td,.table thead th{border-width:0 0 2px;color:#222}.table tfoot{background-color:rgba(0,0,0,0)}.table tfoot td,.table tfoot th{border-width:2px 0 0;color:#222}.table tbody{background-color:rgba(0,0,0,0)}.table tbody tr:last-child td,.table tbody tr:last-child th{border-bottom-width:0}.table.is-bordered td,.table.is-bordered th{border-width:1px}.table.is-bordered tr:last-child td,.table.is-bordered tr:last-child th{border-bottom-width:1px}.table.is-fullwidth{width:100%}.table.is-hoverable tbody tr:not(.is-selected):hover{background-color:#fafafa}.table.is-hoverable.is-striped tbody tr:not(.is-selected):hover{background-color:#fafafa}.table.is-hoverable.is-striped tbody tr:not(.is-selected):hover:nth-child(even){background-color:#f5f5f5}.table.is-narrow td,.table.is-narrow th{padding:0.25em 0.5em}.table.is-striped tbody tr:not(.is-selected):nth-child(even){background-color:#fafafa}.table-container{-webkit-overflow-scrolling:touch;overflow:auto;overflow-y:hidden;max-width:100%}.tags{align-items:center;display:flex;flex-wrap:wrap;justify-content:flex-start}.tags .tag,.tags .content kbd,.content .tags kbd,.tags .docstring>section>a.docs-sourcelink{margin-bottom:0.5rem}.tags .tag:not(:last-child),.tags .content kbd:not(:last-child),.content .tags kbd:not(:last-child),.tags .docstring>section>a.docs-sourcelink:not(:last-child){margin-right:.5rem}.tags:last-child{margin-bottom:-0.5rem}.tags:not(:last-child){margin-bottom:1rem}.tags.are-medium .tag:not(.is-normal):not(.is-large),.tags.are-medium .content kbd:not(.is-normal):not(.is-large),.content .tags.are-medium kbd:not(.is-normal):not(.is-large),.tags.are-medium .docstring>section>a.docs-sourcelink:not(.is-normal):not(.is-large){font-size:1rem}.tags.are-large .tag:not(.is-normal):not(.is-medium),.tags.are-large .content kbd:not(.is-normal):not(.is-medium),.content .tags.are-large kbd:not(.is-normal):not(.is-medium),.tags.are-large .docstring>section>a.docs-sourcelink:not(.is-normal):not(.is-medium){font-size:1.25rem}.tags.is-centered{justify-content:center}.tags.is-centered .tag,.tags.is-centered .content kbd,.content .tags.is-centered kbd,.tags.is-centered .docstring>section>a.docs-sourcelink{margin-right:0.25rem;margin-left:0.25rem}.tags.is-right{justify-content:flex-end}.tags.is-right .tag:not(:first-child),.tags.is-right .content kbd:not(:first-child),.content .tags.is-right kbd:not(:first-child),.tags.is-right .docstring>section>a.docs-sourcelink:not(:first-child){margin-left:0.5rem}.tags.is-right .tag:not(:last-child),.tags.is-right .content kbd:not(:last-child),.content .tags.is-right kbd:not(:last-child),.tags.is-right .docstring>section>a.docs-sourcelink:not(:last-child){margin-right:0}.tags.has-addons .tag,.tags.has-addons .content kbd,.content .tags.has-addons kbd,.tags.has-addons .docstring>section>a.docs-sourcelink{margin-right:0}.tags.has-addons .tag:not(:first-child),.tags.has-addons .content kbd:not(:first-child),.content .tags.has-addons kbd:not(:first-child),.tags.has-addons .docstring>section>a.docs-sourcelink:not(:first-child){margin-left:0;border-top-left-radius:0;border-bottom-left-radius:0}.tags.has-addons .tag:not(:last-child),.tags.has-addons .content kbd:not(:last-child),.content .tags.has-addons kbd:not(:last-child),.tags.has-addons .docstring>section>a.docs-sourcelink:not(:last-child){border-top-right-radius:0;border-bottom-right-radius:0}.tag:not(body),.content kbd:not(body),.docstring>section>a.docs-sourcelink:not(body){align-items:center;background-color:#f5f5f5;border-radius:4px;color:#222;display:inline-flex;font-size:.75rem;height:2em;justify-content:center;line-height:1.5;padding-left:0.75em;padding-right:0.75em;white-space:nowrap}.tag:not(body) .delete,.content kbd:not(body) .delete,.docstring>section>a.docs-sourcelink:not(body) .delete{margin-left:.25rem;margin-right:-.375rem}.tag.is-white:not(body),.content kbd.is-white:not(body),.docstring>section>a.docs-sourcelink.is-white:not(body){background-color:#fff;color:#0a0a0a}.tag.is-black:not(body),.content kbd.is-black:not(body),.docstring>section>a.docs-sourcelink.is-black:not(body){background-color:#0a0a0a;color:#fff}.tag.is-light:not(body),.content kbd.is-light:not(body),.docstring>section>a.docs-sourcelink.is-light:not(body){background-color:#f5f5f5;color:rgba(0,0,0,0.7)}.tag.is-dark:not(body),.content kbd:not(body),.docstring>section>a.docs-sourcelink.is-dark:not(body),.content .docstring>section>kbd:not(body){background-color:#363636;color:#fff}.tag.is-primary:not(body),.content kbd.is-primary:not(body),.docstring>section>a.docs-sourcelink:not(body){background-color:#4eb5de;color:#fff}.tag.is-primary.is-light:not(body),.content kbd.is-primary.is-light:not(body),.docstring>section>a.docs-sourcelink.is-light:not(body){background-color:#eef8fc;color:#1a6d8e}.tag.is-link:not(body),.content kbd.is-link:not(body),.docstring>section>a.docs-sourcelink.is-link:not(body){background-color:#2e63b8;color:#fff}.tag.is-link.is-light:not(body),.content kbd.is-link.is-light:not(body),.docstring>section>a.docs-sourcelink.is-link.is-light:not(body){background-color:#eff3fb;color:#3169c4}.tag.is-info:not(body),.content kbd.is-info:not(body),.docstring>section>a.docs-sourcelink.is-info:not(body){background-color:#209cee;color:#fff}.tag.is-info.is-light:not(body),.content kbd.is-info.is-light:not(body),.docstring>section>a.docs-sourcelink.is-info.is-light:not(body){background-color:#ecf7fe;color:#0e72b4}.tag.is-success:not(body),.content kbd.is-success:not(body),.docstring>section>a.docs-sourcelink.is-success:not(body){background-color:#22c35b;color:#fff}.tag.is-success.is-light:not(body),.content kbd.is-success.is-light:not(body),.docstring>section>a.docs-sourcelink.is-success.is-light:not(body){background-color:#eefcf3;color:#198f43}.tag.is-warning:not(body),.content kbd.is-warning:not(body),.docstring>section>a.docs-sourcelink.is-warning:not(body){background-color:#ffdd57;color:rgba(0,0,0,0.7)}.tag.is-warning.is-light:not(body),.content kbd.is-warning.is-light:not(body),.docstring>section>a.docs-sourcelink.is-warning.is-light:not(body){background-color:#fffbeb;color:#947600}.tag.is-danger:not(body),.content kbd.is-danger:not(body),.docstring>section>a.docs-sourcelink.is-danger:not(body){background-color:#da0b00;color:#fff}.tag.is-danger.is-light:not(body),.content kbd.is-danger.is-light:not(body),.docstring>section>a.docs-sourcelink.is-danger.is-light:not(body){background-color:#ffeceb;color:#f50c00}.tag.is-normal:not(body),.content kbd.is-normal:not(body),.docstring>section>a.docs-sourcelink.is-normal:not(body){font-size:.75rem}.tag.is-medium:not(body),.content kbd.is-medium:not(body),.docstring>section>a.docs-sourcelink.is-medium:not(body){font-size:1rem}.tag.is-large:not(body),.content kbd.is-large:not(body),.docstring>section>a.docs-sourcelink.is-large:not(body){font-size:1.25rem}.tag:not(body) .icon:first-child:not(:last-child),.content kbd:not(body) .icon:first-child:not(:last-child),.docstring>section>a.docs-sourcelink:not(body) .icon:first-child:not(:last-child){margin-left:-.375em;margin-right:.1875em}.tag:not(body) .icon:last-child:not(:first-child),.content kbd:not(body) .icon:last-child:not(:first-child),.docstring>section>a.docs-sourcelink:not(body) .icon:last-child:not(:first-child){margin-left:.1875em;margin-right:-.375em}.tag:not(body) .icon:first-child:last-child,.content kbd:not(body) .icon:first-child:last-child,.docstring>section>a.docs-sourcelink:not(body) .icon:first-child:last-child{margin-left:-.375em;margin-right:-.375em}.tag.is-delete:not(body),.content kbd.is-delete:not(body),.docstring>section>a.docs-sourcelink.is-delete:not(body){margin-left:1px;padding:0;position:relative;width:2em}.tag.is-delete:not(body)::before,.content kbd.is-delete:not(body)::before,.docstring>section>a.docs-sourcelink.is-delete:not(body)::before,.tag.is-delete:not(body)::after,.content kbd.is-delete:not(body)::after,.docstring>section>a.docs-sourcelink.is-delete:not(body)::after{background-color:currentColor;content:"";display:block;left:50%;position:absolute;top:50%;transform:translateX(-50%) translateY(-50%) rotate(45deg);transform-origin:center center}.tag.is-delete:not(body)::before,.content kbd.is-delete:not(body)::before,.docstring>section>a.docs-sourcelink.is-delete:not(body)::before{height:1px;width:50%}.tag.is-delete:not(body)::after,.content kbd.is-delete:not(body)::after,.docstring>section>a.docs-sourcelink.is-delete:not(body)::after{height:50%;width:1px}.tag.is-delete:not(body):hover,.content kbd.is-delete:not(body):hover,.docstring>section>a.docs-sourcelink.is-delete:not(body):hover,.tag.is-delete:not(body):focus,.content kbd.is-delete:not(body):focus,.docstring>section>a.docs-sourcelink.is-delete:not(body):focus{background-color:#e8e8e8}.tag.is-delete:not(body):active,.content kbd.is-delete:not(body):active,.docstring>section>a.docs-sourcelink.is-delete:not(body):active{background-color:#dbdbdb}.tag.is-rounded:not(body),#documenter .docs-sidebar form.docs-search>input:not(body),.content kbd.is-rounded:not(body),#documenter .docs-sidebar .content form.docs-search>input:not(body),.docstring>section>a.docs-sourcelink.is-rounded:not(body){border-radius:9999px}a.tag:hover,.docstring>section>a.docs-sourcelink:hover{text-decoration:underline}.title,.subtitle{word-break:break-word}.title em,.title span,.subtitle em,.subtitle span{font-weight:inherit}.title sub,.subtitle sub{font-size:.75em}.title sup,.subtitle sup{font-size:.75em}.title .tag,.title .content kbd,.content .title kbd,.title .docstring>section>a.docs-sourcelink,.subtitle .tag,.subtitle .content kbd,.content .subtitle kbd,.subtitle .docstring>section>a.docs-sourcelink{vertical-align:middle}.title{color:#222;font-size:2rem;font-weight:600;line-height:1.125}.title strong{color:inherit;font-weight:inherit}.title:not(.is-spaced)+.subtitle{margin-top:-1.25rem}.title.is-1{font-size:3rem}.title.is-2{font-size:2.5rem}.title.is-3{font-size:2rem}.title.is-4{font-size:1.5rem}.title.is-5{font-size:1.25rem}.title.is-6{font-size:1rem}.title.is-7{font-size:.75rem}.subtitle{color:#222;font-size:1.25rem;font-weight:400;line-height:1.25}.subtitle strong{color:#222;font-weight:600}.subtitle:not(.is-spaced)+.title{margin-top:-1.25rem}.subtitle.is-1{font-size:3rem}.subtitle.is-2{font-size:2.5rem}.subtitle.is-3{font-size:2rem}.subtitle.is-4{font-size:1.5rem}.subtitle.is-5{font-size:1.25rem}.subtitle.is-6{font-size:1rem}.subtitle.is-7{font-size:.75rem}.heading{display:block;font-size:11px;letter-spacing:1px;margin-bottom:5px;text-transform:uppercase}.number{align-items:center;background-color:#f5f5f5;border-radius:9999px;display:inline-flex;font-size:1.25rem;height:2em;justify-content:center;margin-right:1.5rem;min-width:2.5em;padding:0.25rem 0.5rem;text-align:center;vertical-align:top}.select select,.textarea,.input,#documenter .docs-sidebar form.docs-search>input{background-color:#fff;border-color:#dbdbdb;border-radius:4px;color:#222}.select select::-moz-placeholder,.textarea::-moz-placeholder,.input::-moz-placeholder,#documenter .docs-sidebar form.docs-search>input::-moz-placeholder{color:#707070}.select select::-webkit-input-placeholder,.textarea::-webkit-input-placeholder,.input::-webkit-input-placeholder,#documenter .docs-sidebar form.docs-search>input::-webkit-input-placeholder{color:#707070}.select select:-moz-placeholder,.textarea:-moz-placeholder,.input:-moz-placeholder,#documenter .docs-sidebar form.docs-search>input:-moz-placeholder{color:#707070}.select select:-ms-input-placeholder,.textarea:-ms-input-placeholder,.input:-ms-input-placeholder,#documenter .docs-sidebar form.docs-search>input:-ms-input-placeholder{color:#707070}.select select:hover,.textarea:hover,.input:hover,#documenter .docs-sidebar form.docs-search>input:hover,.select select.is-hovered,.is-hovered.textarea,.is-hovered.input,#documenter .docs-sidebar form.docs-search>input.is-hovered{border-color:#b5b5b5}.select select:focus,.textarea:focus,.input:focus,#documenter .docs-sidebar form.docs-search>input:focus,.select select.is-focused,.is-focused.textarea,.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.select select:active,.textarea:active,.input:active,#documenter .docs-sidebar form.docs-search>input:active,.select select.is-active,.is-active.textarea,.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{border-color:#2e63b8;box-shadow:0 0 0 0.125em rgba(46,99,184,0.25)}.select select[disabled],.textarea[disabled],.input[disabled],#documenter .docs-sidebar form.docs-search>input[disabled],fieldset[disabled] .select select,.select fieldset[disabled] select,fieldset[disabled] .textarea,fieldset[disabled] .input,fieldset[disabled] #documenter .docs-sidebar form.docs-search>input,#documenter .docs-sidebar fieldset[disabled] form.docs-search>input{background-color:#f5f5f5;border-color:#f5f5f5;box-shadow:none;color:#6b6b6b}.select select[disabled]::-moz-placeholder,.textarea[disabled]::-moz-placeholder,.input[disabled]::-moz-placeholder,#documenter .docs-sidebar form.docs-search>input[disabled]::-moz-placeholder,fieldset[disabled] .select select::-moz-placeholder,.select fieldset[disabled] select::-moz-placeholder,fieldset[disabled] .textarea::-moz-placeholder,fieldset[disabled] .input::-moz-placeholder,fieldset[disabled] #documenter .docs-sidebar form.docs-search>input::-moz-placeholder,#documenter .docs-sidebar fieldset[disabled] form.docs-search>input::-moz-placeholder{color:rgba(107,107,107,0.3)}.select select[disabled]::-webkit-input-placeholder,.textarea[disabled]::-webkit-input-placeholder,.input[disabled]::-webkit-input-placeholder,#documenter .docs-sidebar form.docs-search>input[disabled]::-webkit-input-placeholder,fieldset[disabled] .select select::-webkit-input-placeholder,.select fieldset[disabled] select::-webkit-input-placeholder,fieldset[disabled] .textarea::-webkit-input-placeholder,fieldset[disabled] .input::-webkit-input-placeholder,fieldset[disabled] #documenter .docs-sidebar form.docs-search>input::-webkit-input-placeholder,#documenter .docs-sidebar fieldset[disabled] form.docs-search>input::-webkit-input-placeholder{color:rgba(107,107,107,0.3)}.select select[disabled]:-moz-placeholder,.textarea[disabled]:-moz-placeholder,.input[disabled]:-moz-placeholder,#documenter .docs-sidebar form.docs-search>input[disabled]:-moz-placeholder,fieldset[disabled] .select select:-moz-placeholder,.select fieldset[disabled] select:-moz-placeholder,fieldset[disabled] .textarea:-moz-placeholder,fieldset[disabled] .input:-moz-placeholder,fieldset[disabled] #documenter .docs-sidebar form.docs-search>input:-moz-placeholder,#documenter .docs-sidebar fieldset[disabled] form.docs-search>input:-moz-placeholder{color:rgba(107,107,107,0.3)}.select select[disabled]:-ms-input-placeholder,.textarea[disabled]:-ms-input-placeholder,.input[disabled]:-ms-input-placeholder,#documenter .docs-sidebar form.docs-search>input[disabled]:-ms-input-placeholder,fieldset[disabled] .select select:-ms-input-placeholder,.select fieldset[disabled] select:-ms-input-placeholder,fieldset[disabled] .textarea:-ms-input-placeholder,fieldset[disabled] .input:-ms-input-placeholder,fieldset[disabled] #documenter .docs-sidebar form.docs-search>input:-ms-input-placeholder,#documenter .docs-sidebar fieldset[disabled] form.docs-search>input:-ms-input-placeholder{color:rgba(107,107,107,0.3)}.textarea,.input,#documenter .docs-sidebar form.docs-search>input{box-shadow:inset 0 0.0625em 0.125em rgba(10,10,10,0.05);max-width:100%;width:100%}.textarea[readonly],.input[readonly],#documenter .docs-sidebar form.docs-search>input[readonly]{box-shadow:none}.is-white.textarea,.is-white.input,#documenter .docs-sidebar form.docs-search>input.is-white{border-color:#fff}.is-white.textarea:focus,.is-white.input:focus,#documenter .docs-sidebar form.docs-search>input.is-white:focus,.is-white.is-focused.textarea,.is-white.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-white.textarea:active,.is-white.input:active,#documenter .docs-sidebar form.docs-search>input.is-white:active,.is-white.is-active.textarea,.is-white.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(255,255,255,0.25)}.is-black.textarea,.is-black.input,#documenter .docs-sidebar form.docs-search>input.is-black{border-color:#0a0a0a}.is-black.textarea:focus,.is-black.input:focus,#documenter .docs-sidebar form.docs-search>input.is-black:focus,.is-black.is-focused.textarea,.is-black.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-black.textarea:active,.is-black.input:active,#documenter .docs-sidebar form.docs-search>input.is-black:active,.is-black.is-active.textarea,.is-black.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(10,10,10,0.25)}.is-light.textarea,.is-light.input,#documenter .docs-sidebar form.docs-search>input.is-light{border-color:#f5f5f5}.is-light.textarea:focus,.is-light.input:focus,#documenter .docs-sidebar form.docs-search>input.is-light:focus,.is-light.is-focused.textarea,.is-light.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-light.textarea:active,.is-light.input:active,#documenter .docs-sidebar form.docs-search>input.is-light:active,.is-light.is-active.textarea,.is-light.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(245,245,245,0.25)}.is-dark.textarea,.content kbd.textarea,.is-dark.input,#documenter .docs-sidebar form.docs-search>input.is-dark,.content kbd.input{border-color:#363636}.is-dark.textarea:focus,.content kbd.textarea:focus,.is-dark.input:focus,#documenter .docs-sidebar form.docs-search>input.is-dark:focus,.content kbd.input:focus,.is-dark.is-focused.textarea,.content kbd.is-focused.textarea,.is-dark.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.content kbd.is-focused.input,#documenter .docs-sidebar .content form.docs-search>input.is-focused,.is-dark.textarea:active,.content kbd.textarea:active,.is-dark.input:active,#documenter .docs-sidebar form.docs-search>input.is-dark:active,.content kbd.input:active,.is-dark.is-active.textarea,.content kbd.is-active.textarea,.is-dark.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active,.content kbd.is-active.input,#documenter .docs-sidebar .content form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(54,54,54,0.25)}.is-primary.textarea,.docstring>section>a.textarea.docs-sourcelink,.is-primary.input,#documenter .docs-sidebar form.docs-search>input.is-primary,.docstring>section>a.input.docs-sourcelink{border-color:#4eb5de}.is-primary.textarea:focus,.docstring>section>a.textarea.docs-sourcelink:focus,.is-primary.input:focus,#documenter .docs-sidebar form.docs-search>input.is-primary:focus,.docstring>section>a.input.docs-sourcelink:focus,.is-primary.is-focused.textarea,.docstring>section>a.is-focused.textarea.docs-sourcelink,.is-primary.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.docstring>section>a.is-focused.input.docs-sourcelink,.is-primary.textarea:active,.docstring>section>a.textarea.docs-sourcelink:active,.is-primary.input:active,#documenter .docs-sidebar form.docs-search>input.is-primary:active,.docstring>section>a.input.docs-sourcelink:active,.is-primary.is-active.textarea,.docstring>section>a.is-active.textarea.docs-sourcelink,.is-primary.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active,.docstring>section>a.is-active.input.docs-sourcelink{box-shadow:0 0 0 0.125em rgba(78,181,222,0.25)}.is-link.textarea,.is-link.input,#documenter .docs-sidebar form.docs-search>input.is-link{border-color:#2e63b8}.is-link.textarea:focus,.is-link.input:focus,#documenter .docs-sidebar form.docs-search>input.is-link:focus,.is-link.is-focused.textarea,.is-link.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-link.textarea:active,.is-link.input:active,#documenter .docs-sidebar form.docs-search>input.is-link:active,.is-link.is-active.textarea,.is-link.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(46,99,184,0.25)}.is-info.textarea,.is-info.input,#documenter .docs-sidebar form.docs-search>input.is-info{border-color:#209cee}.is-info.textarea:focus,.is-info.input:focus,#documenter .docs-sidebar form.docs-search>input.is-info:focus,.is-info.is-focused.textarea,.is-info.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-info.textarea:active,.is-info.input:active,#documenter .docs-sidebar form.docs-search>input.is-info:active,.is-info.is-active.textarea,.is-info.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(32,156,238,0.25)}.is-success.textarea,.is-success.input,#documenter .docs-sidebar form.docs-search>input.is-success{border-color:#22c35b}.is-success.textarea:focus,.is-success.input:focus,#documenter .docs-sidebar form.docs-search>input.is-success:focus,.is-success.is-focused.textarea,.is-success.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-success.textarea:active,.is-success.input:active,#documenter .docs-sidebar form.docs-search>input.is-success:active,.is-success.is-active.textarea,.is-success.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(34,195,91,0.25)}.is-warning.textarea,.is-warning.input,#documenter .docs-sidebar form.docs-search>input.is-warning{border-color:#ffdd57}.is-warning.textarea:focus,.is-warning.input:focus,#documenter .docs-sidebar form.docs-search>input.is-warning:focus,.is-warning.is-focused.textarea,.is-warning.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-warning.textarea:active,.is-warning.input:active,#documenter .docs-sidebar form.docs-search>input.is-warning:active,.is-warning.is-active.textarea,.is-warning.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(255,221,87,0.25)}.is-danger.textarea,.is-danger.input,#documenter .docs-sidebar form.docs-search>input.is-danger{border-color:#da0b00}.is-danger.textarea:focus,.is-danger.input:focus,#documenter .docs-sidebar form.docs-search>input.is-danger:focus,.is-danger.is-focused.textarea,.is-danger.is-focused.input,#documenter .docs-sidebar form.docs-search>input.is-focused,.is-danger.textarea:active,.is-danger.input:active,#documenter .docs-sidebar form.docs-search>input.is-danger:active,.is-danger.is-active.textarea,.is-danger.is-active.input,#documenter .docs-sidebar form.docs-search>input.is-active{box-shadow:0 0 0 0.125em rgba(218,11,0,0.25)}.is-small.textarea,.is-small.input,#documenter .docs-sidebar form.docs-search>input{border-radius:2px;font-size:.75rem}.is-medium.textarea,.is-medium.input,#documenter .docs-sidebar form.docs-search>input.is-medium{font-size:1.25rem}.is-large.textarea,.is-large.input,#documenter .docs-sidebar form.docs-search>input.is-large{font-size:1.5rem}.is-fullwidth.textarea,.is-fullwidth.input,#documenter .docs-sidebar form.docs-search>input.is-fullwidth{display:block;width:100%}.is-inline.textarea,.is-inline.input,#documenter .docs-sidebar form.docs-search>input.is-inline{display:inline;width:auto}.input.is-rounded,#documenter .docs-sidebar form.docs-search>input{border-radius:9999px;padding-left:calc(calc(0.75em - 1px) + 0.375em);padding-right:calc(calc(0.75em - 1px) + 0.375em)}.input.is-static,#documenter .docs-sidebar form.docs-search>input.is-static{background-color:transparent;border-color:transparent;box-shadow:none;padding-left:0;padding-right:0}.textarea{display:block;max-width:100%;min-width:100%;padding:calc(0.75em - 1px);resize:vertical}.textarea:not([rows]){max-height:40em;min-height:8em}.textarea[rows]{height:initial}.textarea.has-fixed-size{resize:none}.radio,.checkbox{cursor:pointer;display:inline-block;line-height:1.25;position:relative}.radio input,.checkbox input{cursor:pointer}.radio:hover,.checkbox:hover{color:#222}.radio[disabled],.checkbox[disabled],fieldset[disabled] .radio,fieldset[disabled] .checkbox,.radio input[disabled],.checkbox input[disabled]{color:#6b6b6b;cursor:not-allowed}.radio+.radio{margin-left:.5em}.select{display:inline-block;max-width:100%;position:relative;vertical-align:top}.select:not(.is-multiple){height:2.5em}.select:not(.is-multiple):not(.is-loading)::after{border-color:#2e63b8;right:1.125em;z-index:4}.select.is-rounded select,#documenter .docs-sidebar form.docs-search>input.select select{border-radius:9999px;padding-left:1em}.select select{cursor:pointer;display:block;font-size:1em;max-width:100%;outline:none}.select select::-ms-expand{display:none}.select select[disabled]:hover,fieldset[disabled] .select select:hover{border-color:#f5f5f5}.select select:not([multiple]){padding-right:2.5em}.select select[multiple]{height:auto;padding:0}.select select[multiple] option{padding:0.5em 1em}.select:not(.is-multiple):not(.is-loading):hover::after{border-color:#222}.select.is-white:not(:hover)::after{border-color:#fff}.select.is-white select{border-color:#fff}.select.is-white select:hover,.select.is-white select.is-hovered{border-color:#f2f2f2}.select.is-white select:focus,.select.is-white select.is-focused,.select.is-white select:active,.select.is-white select.is-active{box-shadow:0 0 0 0.125em rgba(255,255,255,0.25)}.select.is-black:not(:hover)::after{border-color:#0a0a0a}.select.is-black select{border-color:#0a0a0a}.select.is-black select:hover,.select.is-black select.is-hovered{border-color:#000}.select.is-black select:focus,.select.is-black select.is-focused,.select.is-black select:active,.select.is-black select.is-active{box-shadow:0 0 0 0.125em rgba(10,10,10,0.25)}.select.is-light:not(:hover)::after{border-color:#f5f5f5}.select.is-light select{border-color:#f5f5f5}.select.is-light select:hover,.select.is-light select.is-hovered{border-color:#e8e8e8}.select.is-light select:focus,.select.is-light select.is-focused,.select.is-light select:active,.select.is-light select.is-active{box-shadow:0 0 0 0.125em rgba(245,245,245,0.25)}.select.is-dark:not(:hover)::after,.content kbd.select:not(:hover)::after{border-color:#363636}.select.is-dark select,.content kbd.select select{border-color:#363636}.select.is-dark select:hover,.content kbd.select select:hover,.select.is-dark select.is-hovered,.content kbd.select select.is-hovered{border-color:#292929}.select.is-dark select:focus,.content kbd.select select:focus,.select.is-dark select.is-focused,.content kbd.select select.is-focused,.select.is-dark select:active,.content kbd.select select:active,.select.is-dark select.is-active,.content kbd.select select.is-active{box-shadow:0 0 0 0.125em rgba(54,54,54,0.25)}.select.is-primary:not(:hover)::after,.docstring>section>a.select.docs-sourcelink:not(:hover)::after{border-color:#4eb5de}.select.is-primary select,.docstring>section>a.select.docs-sourcelink select{border-color:#4eb5de}.select.is-primary select:hover,.docstring>section>a.select.docs-sourcelink select:hover,.select.is-primary select.is-hovered,.docstring>section>a.select.docs-sourcelink select.is-hovered{border-color:#39acda}.select.is-primary select:focus,.docstring>section>a.select.docs-sourcelink select:focus,.select.is-primary select.is-focused,.docstring>section>a.select.docs-sourcelink select.is-focused,.select.is-primary select:active,.docstring>section>a.select.docs-sourcelink select:active,.select.is-primary select.is-active,.docstring>section>a.select.docs-sourcelink select.is-active{box-shadow:0 0 0 0.125em rgba(78,181,222,0.25)}.select.is-link:not(:hover)::after{border-color:#2e63b8}.select.is-link select{border-color:#2e63b8}.select.is-link select:hover,.select.is-link select.is-hovered{border-color:#2958a4}.select.is-link select:focus,.select.is-link select.is-focused,.select.is-link select:active,.select.is-link select.is-active{box-shadow:0 0 0 0.125em rgba(46,99,184,0.25)}.select.is-info:not(:hover)::after{border-color:#209cee}.select.is-info select{border-color:#209cee}.select.is-info select:hover,.select.is-info select.is-hovered{border-color:#1190e3}.select.is-info select:focus,.select.is-info select.is-focused,.select.is-info select:active,.select.is-info select.is-active{box-shadow:0 0 0 0.125em rgba(32,156,238,0.25)}.select.is-success:not(:hover)::after{border-color:#22c35b}.select.is-success select{border-color:#22c35b}.select.is-success select:hover,.select.is-success select.is-hovered{border-color:#1ead51}.select.is-success select:focus,.select.is-success select.is-focused,.select.is-success select:active,.select.is-success select.is-active{box-shadow:0 0 0 0.125em rgba(34,195,91,0.25)}.select.is-warning:not(:hover)::after{border-color:#ffdd57}.select.is-warning select{border-color:#ffdd57}.select.is-warning select:hover,.select.is-warning select.is-hovered{border-color:#ffd83e}.select.is-warning select:focus,.select.is-warning select.is-focused,.select.is-warning select:active,.select.is-warning select.is-active{box-shadow:0 0 0 0.125em rgba(255,221,87,0.25)}.select.is-danger:not(:hover)::after{border-color:#da0b00}.select.is-danger select{border-color:#da0b00}.select.is-danger select:hover,.select.is-danger select.is-hovered{border-color:#c10a00}.select.is-danger select:focus,.select.is-danger select.is-focused,.select.is-danger select:active,.select.is-danger select.is-active{box-shadow:0 0 0 0.125em rgba(218,11,0,0.25)}.select.is-small,#documenter .docs-sidebar form.docs-search>input.select{border-radius:2px;font-size:.75rem}.select.is-medium{font-size:1.25rem}.select.is-large{font-size:1.5rem}.select.is-disabled::after{border-color:#6b6b6b !important;opacity:0.5}.select.is-fullwidth{width:100%}.select.is-fullwidth select{width:100%}.select.is-loading::after{margin-top:0;position:absolute;right:.625em;top:0.625em;transform:none}.select.is-loading.is-small:after,#documenter .docs-sidebar form.docs-search>input.is-loading:after{font-size:.75rem}.select.is-loading.is-medium:after{font-size:1.25rem}.select.is-loading.is-large:after{font-size:1.5rem}.file{align-items:stretch;display:flex;justify-content:flex-start;position:relative}.file.is-white .file-cta{background-color:#fff;border-color:transparent;color:#0a0a0a}.file.is-white:hover .file-cta,.file.is-white.is-hovered .file-cta{background-color:#f9f9f9;border-color:transparent;color:#0a0a0a}.file.is-white:focus .file-cta,.file.is-white.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(255,255,255,0.25);color:#0a0a0a}.file.is-white:active .file-cta,.file.is-white.is-active .file-cta{background-color:#f2f2f2;border-color:transparent;color:#0a0a0a}.file.is-black .file-cta{background-color:#0a0a0a;border-color:transparent;color:#fff}.file.is-black:hover .file-cta,.file.is-black.is-hovered .file-cta{background-color:#040404;border-color:transparent;color:#fff}.file.is-black:focus .file-cta,.file.is-black.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(10,10,10,0.25);color:#fff}.file.is-black:active .file-cta,.file.is-black.is-active .file-cta{background-color:#000;border-color:transparent;color:#fff}.file.is-light .file-cta{background-color:#f5f5f5;border-color:transparent;color:rgba(0,0,0,0.7)}.file.is-light:hover .file-cta,.file.is-light.is-hovered .file-cta{background-color:#eee;border-color:transparent;color:rgba(0,0,0,0.7)}.file.is-light:focus .file-cta,.file.is-light.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(245,245,245,0.25);color:rgba(0,0,0,0.7)}.file.is-light:active .file-cta,.file.is-light.is-active .file-cta{background-color:#e8e8e8;border-color:transparent;color:rgba(0,0,0,0.7)}.file.is-dark .file-cta,.content kbd.file .file-cta{background-color:#363636;border-color:transparent;color:#fff}.file.is-dark:hover .file-cta,.content kbd.file:hover .file-cta,.file.is-dark.is-hovered .file-cta,.content kbd.file.is-hovered .file-cta{background-color:#2f2f2f;border-color:transparent;color:#fff}.file.is-dark:focus .file-cta,.content kbd.file:focus .file-cta,.file.is-dark.is-focused .file-cta,.content kbd.file.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(54,54,54,0.25);color:#fff}.file.is-dark:active .file-cta,.content kbd.file:active .file-cta,.file.is-dark.is-active .file-cta,.content kbd.file.is-active .file-cta{background-color:#292929;border-color:transparent;color:#fff}.file.is-primary .file-cta,.docstring>section>a.file.docs-sourcelink .file-cta{background-color:#4eb5de;border-color:transparent;color:#fff}.file.is-primary:hover .file-cta,.docstring>section>a.file.docs-sourcelink:hover .file-cta,.file.is-primary.is-hovered .file-cta,.docstring>section>a.file.is-hovered.docs-sourcelink .file-cta{background-color:#43b1dc;border-color:transparent;color:#fff}.file.is-primary:focus .file-cta,.docstring>section>a.file.docs-sourcelink:focus .file-cta,.file.is-primary.is-focused .file-cta,.docstring>section>a.file.is-focused.docs-sourcelink .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(78,181,222,0.25);color:#fff}.file.is-primary:active .file-cta,.docstring>section>a.file.docs-sourcelink:active .file-cta,.file.is-primary.is-active .file-cta,.docstring>section>a.file.is-active.docs-sourcelink .file-cta{background-color:#39acda;border-color:transparent;color:#fff}.file.is-link .file-cta{background-color:#2e63b8;border-color:transparent;color:#fff}.file.is-link:hover .file-cta,.file.is-link.is-hovered .file-cta{background-color:#2b5eae;border-color:transparent;color:#fff}.file.is-link:focus .file-cta,.file.is-link.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(46,99,184,0.25);color:#fff}.file.is-link:active .file-cta,.file.is-link.is-active .file-cta{background-color:#2958a4;border-color:transparent;color:#fff}.file.is-info .file-cta{background-color:#209cee;border-color:transparent;color:#fff}.file.is-info:hover .file-cta,.file.is-info.is-hovered .file-cta{background-color:#1497ed;border-color:transparent;color:#fff}.file.is-info:focus .file-cta,.file.is-info.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(32,156,238,0.25);color:#fff}.file.is-info:active .file-cta,.file.is-info.is-active .file-cta{background-color:#1190e3;border-color:transparent;color:#fff}.file.is-success .file-cta{background-color:#22c35b;border-color:transparent;color:#fff}.file.is-success:hover .file-cta,.file.is-success.is-hovered .file-cta{background-color:#20b856;border-color:transparent;color:#fff}.file.is-success:focus .file-cta,.file.is-success.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(34,195,91,0.25);color:#fff}.file.is-success:active .file-cta,.file.is-success.is-active .file-cta{background-color:#1ead51;border-color:transparent;color:#fff}.file.is-warning .file-cta{background-color:#ffdd57;border-color:transparent;color:rgba(0,0,0,0.7)}.file.is-warning:hover .file-cta,.file.is-warning.is-hovered .file-cta{background-color:#ffda4a;border-color:transparent;color:rgba(0,0,0,0.7)}.file.is-warning:focus .file-cta,.file.is-warning.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(255,221,87,0.25);color:rgba(0,0,0,0.7)}.file.is-warning:active .file-cta,.file.is-warning.is-active .file-cta{background-color:#ffd83e;border-color:transparent;color:rgba(0,0,0,0.7)}.file.is-danger .file-cta{background-color:#da0b00;border-color:transparent;color:#fff}.file.is-danger:hover .file-cta,.file.is-danger.is-hovered .file-cta{background-color:#cd0a00;border-color:transparent;color:#fff}.file.is-danger:focus .file-cta,.file.is-danger.is-focused .file-cta{border-color:transparent;box-shadow:0 0 0.5em rgba(218,11,0,0.25);color:#fff}.file.is-danger:active .file-cta,.file.is-danger.is-active .file-cta{background-color:#c10a00;border-color:transparent;color:#fff}.file.is-small,#documenter .docs-sidebar form.docs-search>input.file{font-size:.75rem}.file.is-normal{font-size:1rem}.file.is-medium{font-size:1.25rem}.file.is-medium .file-icon .fa{font-size:21px}.file.is-large{font-size:1.5rem}.file.is-large .file-icon .fa{font-size:28px}.file.has-name .file-cta{border-bottom-right-radius:0;border-top-right-radius:0}.file.has-name .file-name{border-bottom-left-radius:0;border-top-left-radius:0}.file.has-name.is-empty .file-cta{border-radius:4px}.file.has-name.is-empty .file-name{display:none}.file.is-boxed .file-label{flex-direction:column}.file.is-boxed .file-cta{flex-direction:column;height:auto;padding:1em 3em}.file.is-boxed .file-name{border-width:0 1px 1px}.file.is-boxed .file-icon{height:1.5em;width:1.5em}.file.is-boxed .file-icon .fa{font-size:21px}.file.is-boxed.is-small .file-icon .fa,#documenter .docs-sidebar form.docs-search>input.is-boxed .file-icon .fa{font-size:14px}.file.is-boxed.is-medium .file-icon .fa{font-size:28px}.file.is-boxed.is-large .file-icon .fa{font-size:35px}.file.is-boxed.has-name .file-cta{border-radius:4px 4px 0 0}.file.is-boxed.has-name .file-name{border-radius:0 0 4px 4px;border-width:0 1px 1px}.file.is-centered{justify-content:center}.file.is-fullwidth .file-label{width:100%}.file.is-fullwidth .file-name{flex-grow:1;max-width:none}.file.is-right{justify-content:flex-end}.file.is-right .file-cta{border-radius:0 4px 4px 0}.file.is-right .file-name{border-radius:4px 0 0 4px;border-width:1px 0 1px 1px;order:-1}.file-label{align-items:stretch;display:flex;cursor:pointer;justify-content:flex-start;overflow:hidden;position:relative}.file-label:hover .file-cta{background-color:#eee;color:#222}.file-label:hover .file-name{border-color:#d5d5d5}.file-label:active .file-cta{background-color:#e8e8e8;color:#222}.file-label:active .file-name{border-color:#cfcfcf}.file-input{height:100%;left:0;opacity:0;outline:none;position:absolute;top:0;width:100%}.file-cta,.file-name{border-color:#dbdbdb;border-radius:4px;font-size:1em;padding-left:1em;padding-right:1em;white-space:nowrap}.file-cta{background-color:#f5f5f5;color:#222}.file-name{border-color:#dbdbdb;border-style:solid;border-width:1px 1px 1px 0;display:block;max-width:16em;overflow:hidden;text-align:inherit;text-overflow:ellipsis}.file-icon{align-items:center;display:flex;height:1em;justify-content:center;margin-right:.5em;width:1em}.file-icon .fa{font-size:14px}.label{color:#222;display:block;font-size:1rem;font-weight:700}.label:not(:last-child){margin-bottom:0.5em}.label.is-small,#documenter .docs-sidebar form.docs-search>input.label{font-size:.75rem}.label.is-medium{font-size:1.25rem}.label.is-large{font-size:1.5rem}.help{display:block;font-size:.75rem;margin-top:0.25rem}.help.is-white{color:#fff}.help.is-black{color:#0a0a0a}.help.is-light{color:#f5f5f5}.help.is-dark,.content kbd.help{color:#363636}.help.is-primary,.docstring>section>a.help.docs-sourcelink{color:#4eb5de}.help.is-link{color:#2e63b8}.help.is-info{color:#209cee}.help.is-success{color:#22c35b}.help.is-warning{color:#ffdd57}.help.is-danger{color:#da0b00}.field:not(:last-child){margin-bottom:0.75rem}.field.has-addons{display:flex;justify-content:flex-start}.field.has-addons .control:not(:last-child){margin-right:-1px}.field.has-addons .control:not(:first-child):not(:last-child) .button,.field.has-addons .control:not(:first-child):not(:last-child) .input,.field.has-addons .control:not(:first-child):not(:last-child) #documenter .docs-sidebar form.docs-search>input,#documenter .docs-sidebar .field.has-addons .control:not(:first-child):not(:last-child) form.docs-search>input,.field.has-addons .control:not(:first-child):not(:last-child) .select select{border-radius:0}.field.has-addons .control:first-child:not(:only-child) .button,.field.has-addons .control:first-child:not(:only-child) .input,.field.has-addons .control:first-child:not(:only-child) #documenter .docs-sidebar form.docs-search>input,#documenter .docs-sidebar .field.has-addons .control:first-child:not(:only-child) form.docs-search>input,.field.has-addons .control:first-child:not(:only-child) .select select{border-bottom-right-radius:0;border-top-right-radius:0}.field.has-addons .control:last-child:not(:only-child) .button,.field.has-addons .control:last-child:not(:only-child) .input,.field.has-addons .control:last-child:not(:only-child) #documenter .docs-sidebar form.docs-search>input,#documenter .docs-sidebar .field.has-addons .control:last-child:not(:only-child) form.docs-search>input,.field.has-addons .control:last-child:not(:only-child) .select select{border-bottom-left-radius:0;border-top-left-radius:0}.field.has-addons .control .button:not([disabled]):hover,.field.has-addons .control .button.is-hovered:not([disabled]),.field.has-addons .control .input:not([disabled]):hover,.field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):hover,#documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):hover,.field.has-addons .control .input.is-hovered:not([disabled]),.field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-hovered:not([disabled]),#documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-hovered:not([disabled]),.field.has-addons .control .select select:not([disabled]):hover,.field.has-addons .control .select select.is-hovered:not([disabled]){z-index:2}.field.has-addons .control .button:not([disabled]):focus,.field.has-addons .control .button.is-focused:not([disabled]),.field.has-addons .control .button:not([disabled]):active,.field.has-addons .control .button.is-active:not([disabled]),.field.has-addons .control .input:not([disabled]):focus,.field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):focus,#documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):focus,.field.has-addons .control .input.is-focused:not([disabled]),.field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-focused:not([disabled]),#documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-focused:not([disabled]),.field.has-addons .control .input:not([disabled]):active,.field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):active,#documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):active,.field.has-addons .control .input.is-active:not([disabled]),.field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-active:not([disabled]),#documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-active:not([disabled]),.field.has-addons .control .select select:not([disabled]):focus,.field.has-addons .control .select select.is-focused:not([disabled]),.field.has-addons .control .select select:not([disabled]):active,.field.has-addons .control .select select.is-active:not([disabled]){z-index:3}.field.has-addons .control .button:not([disabled]):focus:hover,.field.has-addons .control .button.is-focused:not([disabled]):hover,.field.has-addons .control .button:not([disabled]):active:hover,.field.has-addons .control .button.is-active:not([disabled]):hover,.field.has-addons .control .input:not([disabled]):focus:hover,.field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):focus:hover,#documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):focus:hover,.field.has-addons .control .input.is-focused:not([disabled]):hover,.field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-focused:not([disabled]):hover,#documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-focused:not([disabled]):hover,.field.has-addons .control .input:not([disabled]):active:hover,.field.has-addons .control #documenter .docs-sidebar form.docs-search>input:not([disabled]):active:hover,#documenter .docs-sidebar .field.has-addons .control form.docs-search>input:not([disabled]):active:hover,.field.has-addons .control .input.is-active:not([disabled]):hover,.field.has-addons .control #documenter .docs-sidebar form.docs-search>input.is-active:not([disabled]):hover,#documenter .docs-sidebar .field.has-addons .control form.docs-search>input.is-active:not([disabled]):hover,.field.has-addons .control .select select:not([disabled]):focus:hover,.field.has-addons .control .select select.is-focused:not([disabled]):hover,.field.has-addons .control .select select:not([disabled]):active:hover,.field.has-addons .control .select select.is-active:not([disabled]):hover{z-index:4}.field.has-addons .control.is-expanded{flex-grow:1;flex-shrink:1}.field.has-addons.has-addons-centered{justify-content:center}.field.has-addons.has-addons-right{justify-content:flex-end}.field.has-addons.has-addons-fullwidth .control{flex-grow:1;flex-shrink:0}.field.is-grouped{display:flex;justify-content:flex-start}.field.is-grouped>.control{flex-shrink:0}.field.is-grouped>.control:not(:last-child){margin-bottom:0;margin-right:.75rem}.field.is-grouped>.control.is-expanded{flex-grow:1;flex-shrink:1}.field.is-grouped.is-grouped-centered{justify-content:center}.field.is-grouped.is-grouped-right{justify-content:flex-end}.field.is-grouped.is-grouped-multiline{flex-wrap:wrap}.field.is-grouped.is-grouped-multiline>.control:last-child,.field.is-grouped.is-grouped-multiline>.control:not(:last-child){margin-bottom:0.75rem}.field.is-grouped.is-grouped-multiline:last-child{margin-bottom:-0.75rem}.field.is-grouped.is-grouped-multiline:not(:last-child){margin-bottom:0}@media screen and (min-width: 769px),print{.field.is-horizontal{display:flex}}.field-label .label{font-size:inherit}@media screen and (max-width: 768px){.field-label{margin-bottom:0.5rem}}@media screen and (min-width: 769px),print{.field-label{flex-basis:0;flex-grow:1;flex-shrink:0;margin-right:1.5rem;text-align:right}.field-label.is-small,#documenter .docs-sidebar form.docs-search>input.field-label{font-size:.75rem;padding-top:0.375em}.field-label.is-normal{padding-top:0.375em}.field-label.is-medium{font-size:1.25rem;padding-top:0.375em}.field-label.is-large{font-size:1.5rem;padding-top:0.375em}}.field-body .field .field{margin-bottom:0}@media screen and (min-width: 769px),print{.field-body{display:flex;flex-basis:0;flex-grow:5;flex-shrink:1}.field-body .field{margin-bottom:0}.field-body>.field{flex-shrink:1}.field-body>.field:not(.is-narrow){flex-grow:1}.field-body>.field:not(:last-child){margin-right:.75rem}}.control{box-sizing:border-box;clear:both;font-size:1rem;position:relative;text-align:inherit}.control.has-icons-left .input:focus~.icon,.control.has-icons-left #documenter .docs-sidebar form.docs-search>input:focus~.icon,#documenter .docs-sidebar .control.has-icons-left form.docs-search>input:focus~.icon,.control.has-icons-left .select:focus~.icon,.control.has-icons-right .input:focus~.icon,.control.has-icons-right #documenter .docs-sidebar form.docs-search>input:focus~.icon,#documenter .docs-sidebar .control.has-icons-right form.docs-search>input:focus~.icon,.control.has-icons-right .select:focus~.icon{color:#222}.control.has-icons-left .input.is-small~.icon,.control.has-icons-left #documenter .docs-sidebar form.docs-search>input~.icon,#documenter .docs-sidebar .control.has-icons-left form.docs-search>input~.icon,.control.has-icons-left .select.is-small~.icon,.control.has-icons-right .input.is-small~.icon,.control.has-icons-right #documenter .docs-sidebar form.docs-search>input~.icon,#documenter .docs-sidebar .control.has-icons-right form.docs-search>input~.icon,.control.has-icons-right .select.is-small~.icon{font-size:.75rem}.control.has-icons-left .input.is-medium~.icon,.control.has-icons-left #documenter .docs-sidebar form.docs-search>input.is-medium~.icon,#documenter .docs-sidebar .control.has-icons-left form.docs-search>input.is-medium~.icon,.control.has-icons-left .select.is-medium~.icon,.control.has-icons-right .input.is-medium~.icon,.control.has-icons-right #documenter .docs-sidebar form.docs-search>input.is-medium~.icon,#documenter .docs-sidebar .control.has-icons-right form.docs-search>input.is-medium~.icon,.control.has-icons-right .select.is-medium~.icon{font-size:1.25rem}.control.has-icons-left .input.is-large~.icon,.control.has-icons-left #documenter .docs-sidebar form.docs-search>input.is-large~.icon,#documenter .docs-sidebar .control.has-icons-left form.docs-search>input.is-large~.icon,.control.has-icons-left .select.is-large~.icon,.control.has-icons-right .input.is-large~.icon,.control.has-icons-right #documenter .docs-sidebar form.docs-search>input.is-large~.icon,#documenter .docs-sidebar .control.has-icons-right form.docs-search>input.is-large~.icon,.control.has-icons-right .select.is-large~.icon{font-size:1.5rem}.control.has-icons-left .icon,.control.has-icons-right .icon{color:#dbdbdb;height:2.5em;pointer-events:none;position:absolute;top:0;width:2.5em;z-index:4}.control.has-icons-left .input,.control.has-icons-left #documenter .docs-sidebar form.docs-search>input,#documenter .docs-sidebar .control.has-icons-left form.docs-search>input,.control.has-icons-left .select select{padding-left:2.5em}.control.has-icons-left .icon.is-left{left:0}.control.has-icons-right .input,.control.has-icons-right #documenter .docs-sidebar form.docs-search>input,#documenter .docs-sidebar .control.has-icons-right form.docs-search>input,.control.has-icons-right .select select{padding-right:2.5em}.control.has-icons-right .icon.is-right{right:0}.control.is-loading::after{position:absolute !important;right:.625em;top:0.625em;z-index:4}.control.is-loading.is-small:after,#documenter .docs-sidebar form.docs-search>input.is-loading:after{font-size:.75rem}.control.is-loading.is-medium:after{font-size:1.25rem}.control.is-loading.is-large:after{font-size:1.5rem}.breadcrumb{font-size:1rem;white-space:nowrap}.breadcrumb a{align-items:center;color:#2e63b8;display:flex;justify-content:center;padding:0 .75em}.breadcrumb a:hover{color:#363636}.breadcrumb li{align-items:center;display:flex}.breadcrumb li:first-child a{padding-left:0}.breadcrumb li.is-active a{color:#222;cursor:default;pointer-events:none}.breadcrumb li+li::before{color:#b5b5b5;content:"\0002f"}.breadcrumb ul,.breadcrumb ol{align-items:flex-start;display:flex;flex-wrap:wrap;justify-content:flex-start}.breadcrumb .icon:first-child{margin-right:.5em}.breadcrumb .icon:last-child{margin-left:.5em}.breadcrumb.is-centered ol,.breadcrumb.is-centered ul{justify-content:center}.breadcrumb.is-right ol,.breadcrumb.is-right ul{justify-content:flex-end}.breadcrumb.is-small,#documenter .docs-sidebar form.docs-search>input.breadcrumb{font-size:.75rem}.breadcrumb.is-medium{font-size:1.25rem}.breadcrumb.is-large{font-size:1.5rem}.breadcrumb.has-arrow-separator li+li::before{content:"\02192"}.breadcrumb.has-bullet-separator li+li::before{content:"\02022"}.breadcrumb.has-dot-separator li+li::before{content:"\000b7"}.breadcrumb.has-succeeds-separator li+li::before{content:"\0227B"}.card{background-color:#fff;border-radius:.25rem;box-shadow:#bbb;color:#222;max-width:100%;position:relative}.card-footer:first-child,.card-content:first-child,.card-header:first-child{border-top-left-radius:.25rem;border-top-right-radius:.25rem}.card-footer:last-child,.card-content:last-child,.card-header:last-child{border-bottom-left-radius:.25rem;border-bottom-right-radius:.25rem}.card-header{background-color:rgba(0,0,0,0);align-items:stretch;box-shadow:0 0.125em 0.25em rgba(10,10,10,0.1);display:flex}.card-header-title{align-items:center;color:#222;display:flex;flex-grow:1;font-weight:700;padding:0.75rem 1rem}.card-header-title.is-centered{justify-content:center}.card-header-icon{-moz-appearance:none;-webkit-appearance:none;appearance:none;background:none;border:none;color:currentColor;font-family:inherit;font-size:1em;margin:0;padding:0;align-items:center;cursor:pointer;display:flex;justify-content:center;padding:0.75rem 1rem}.card-image{display:block;position:relative}.card-image:first-child img{border-top-left-radius:.25rem;border-top-right-radius:.25rem}.card-image:last-child img{border-bottom-left-radius:.25rem;border-bottom-right-radius:.25rem}.card-content{background-color:rgba(0,0,0,0);padding:1.5rem}.card-footer{background-color:rgba(0,0,0,0);border-top:1px solid #ededed;align-items:stretch;display:flex}.card-footer-item{align-items:center;display:flex;flex-basis:0;flex-grow:1;flex-shrink:0;justify-content:center;padding:.75rem}.card-footer-item:not(:last-child){border-right:1px solid #ededed}.card .media:not(:last-child){margin-bottom:1.5rem}.dropdown{display:inline-flex;position:relative;vertical-align:top}.dropdown.is-active .dropdown-menu,.dropdown.is-hoverable:hover .dropdown-menu{display:block}.dropdown.is-right .dropdown-menu{left:auto;right:0}.dropdown.is-up .dropdown-menu{bottom:100%;padding-bottom:4px;padding-top:initial;top:auto}.dropdown-menu{display:none;left:0;min-width:12rem;padding-top:4px;position:absolute;top:100%;z-index:20}.dropdown-content{background-color:#fff;border-radius:4px;box-shadow:#bbb;padding-bottom:.5rem;padding-top:.5rem}.dropdown-item{color:#222;display:block;font-size:0.875rem;line-height:1.5;padding:0.375rem 1rem;position:relative}a.dropdown-item,button.dropdown-item{padding-right:3rem;text-align:inherit;white-space:nowrap;width:100%}a.dropdown-item:hover,button.dropdown-item:hover{background-color:#f5f5f5;color:#0a0a0a}a.dropdown-item.is-active,button.dropdown-item.is-active{background-color:#2e63b8;color:#fff}.dropdown-divider{background-color:#ededed;border:none;display:block;height:1px;margin:0.5rem 0}.level{align-items:center;justify-content:space-between}.level code{border-radius:4px}.level img{display:inline-block;vertical-align:top}.level.is-mobile{display:flex}.level.is-mobile .level-left,.level.is-mobile .level-right{display:flex}.level.is-mobile .level-left+.level-right{margin-top:0}.level.is-mobile .level-item:not(:last-child){margin-bottom:0;margin-right:.75rem}.level.is-mobile .level-item:not(.is-narrow){flex-grow:1}@media screen and (min-width: 769px),print{.level{display:flex}.level>.level-item:not(.is-narrow){flex-grow:1}}.level-item{align-items:center;display:flex;flex-basis:auto;flex-grow:0;flex-shrink:0;justify-content:center}.level-item .title,.level-item .subtitle{margin-bottom:0}@media screen and (max-width: 768px){.level-item:not(:last-child){margin-bottom:.75rem}}.level-left,.level-right{flex-basis:auto;flex-grow:0;flex-shrink:0}.level-left .level-item.is-flexible,.level-right .level-item.is-flexible{flex-grow:1}@media screen and (min-width: 769px),print{.level-left .level-item:not(:last-child),.level-right .level-item:not(:last-child){margin-right:.75rem}}.level-left{align-items:center;justify-content:flex-start}@media screen and (max-width: 768px){.level-left+.level-right{margin-top:1.5rem}}@media screen and (min-width: 769px),print{.level-left{display:flex}}.level-right{align-items:center;justify-content:flex-end}@media screen and (min-width: 769px),print{.level-right{display:flex}}.media{align-items:flex-start;display:flex;text-align:inherit}.media .content:not(:last-child){margin-bottom:.75rem}.media .media{border-top:1px solid rgba(219,219,219,0.5);display:flex;padding-top:.75rem}.media .media .content:not(:last-child),.media .media .control:not(:last-child){margin-bottom:.5rem}.media .media .media{padding-top:.5rem}.media .media .media+.media{margin-top:.5rem}.media+.media{border-top:1px solid rgba(219,219,219,0.5);margin-top:1rem;padding-top:1rem}.media.is-large+.media{margin-top:1.5rem;padding-top:1.5rem}.media-left,.media-right{flex-basis:auto;flex-grow:0;flex-shrink:0}.media-left{margin-right:1rem}.media-right{margin-left:1rem}.media-content{flex-basis:auto;flex-grow:1;flex-shrink:1;text-align:inherit}@media screen and (max-width: 768px){.media-content{overflow-x:auto}}.menu{font-size:1rem}.menu.is-small,#documenter .docs-sidebar form.docs-search>input.menu{font-size:.75rem}.menu.is-medium{font-size:1.25rem}.menu.is-large{font-size:1.5rem}.menu-list{line-height:1.25}.menu-list a{border-radius:2px;color:#222;display:block;padding:0.5em 0.75em}.menu-list a:hover{background-color:#f5f5f5;color:#222}.menu-list a.is-active{background-color:#2e63b8;color:#fff}.menu-list li ul{border-left:1px solid #dbdbdb;margin:.75em;padding-left:.75em}.menu-label{color:#6b6b6b;font-size:.75em;letter-spacing:.1em;text-transform:uppercase}.menu-label:not(:first-child){margin-top:1em}.menu-label:not(:last-child){margin-bottom:1em}.message{background-color:#f5f5f5;border-radius:4px;font-size:1rem}.message strong{color:currentColor}.message a:not(.button):not(.tag):not(.dropdown-item){color:currentColor;text-decoration:underline}.message.is-small,#documenter .docs-sidebar form.docs-search>input.message{font-size:.75rem}.message.is-medium{font-size:1.25rem}.message.is-large{font-size:1.5rem}.message.is-white{background-color:#fff}.message.is-white .message-header{background-color:#fff;color:#0a0a0a}.message.is-white .message-body{border-color:#fff}.message.is-black{background-color:#fafafa}.message.is-black .message-header{background-color:#0a0a0a;color:#fff}.message.is-black .message-body{border-color:#0a0a0a}.message.is-light{background-color:#fafafa}.message.is-light .message-header{background-color:#f5f5f5;color:rgba(0,0,0,0.7)}.message.is-light .message-body{border-color:#f5f5f5}.message.is-dark,.content kbd.message{background-color:#fafafa}.message.is-dark .message-header,.content kbd.message .message-header{background-color:#363636;color:#fff}.message.is-dark .message-body,.content kbd.message .message-body{border-color:#363636}.message.is-primary,.docstring>section>a.message.docs-sourcelink{background-color:#eef8fc}.message.is-primary .message-header,.docstring>section>a.message.docs-sourcelink .message-header{background-color:#4eb5de;color:#fff}.message.is-primary .message-body,.docstring>section>a.message.docs-sourcelink .message-body{border-color:#4eb5de;color:#1a6d8e}.message.is-link{background-color:#eff3fb}.message.is-link .message-header{background-color:#2e63b8;color:#fff}.message.is-link .message-body{border-color:#2e63b8;color:#3169c4}.message.is-info{background-color:#ecf7fe}.message.is-info .message-header{background-color:#209cee;color:#fff}.message.is-info .message-body{border-color:#209cee;color:#0e72b4}.message.is-success{background-color:#eefcf3}.message.is-success .message-header{background-color:#22c35b;color:#fff}.message.is-success .message-body{border-color:#22c35b;color:#198f43}.message.is-warning{background-color:#fffbeb}.message.is-warning .message-header{background-color:#ffdd57;color:rgba(0,0,0,0.7)}.message.is-warning .message-body{border-color:#ffdd57;color:#947600}.message.is-danger{background-color:#ffeceb}.message.is-danger .message-header{background-color:#da0b00;color:#fff}.message.is-danger .message-body{border-color:#da0b00;color:#f50c00}.message-header{align-items:center;background-color:#222;border-radius:4px 4px 0 0;color:#fff;display:flex;font-weight:700;justify-content:space-between;line-height:1.25;padding:0.75em 1em;position:relative}.message-header .delete{flex-grow:0;flex-shrink:0;margin-left:.75em}.message-header+.message-body{border-width:0;border-top-left-radius:0;border-top-right-radius:0}.message-body{border-color:#dbdbdb;border-radius:4px;border-style:solid;border-width:0 0 0 4px;color:#222;padding:1.25em 1.5em}.message-body code,.message-body pre{background-color:#fff}.message-body pre code{background-color:rgba(0,0,0,0)}.modal{align-items:center;display:none;flex-direction:column;justify-content:center;overflow:hidden;position:fixed;z-index:40}.modal.is-active{display:flex}.modal-background{background-color:rgba(10,10,10,0.86)}.modal-content,.modal-card{margin:0 20px;max-height:calc(100vh - 160px);overflow:auto;position:relative;width:100%}@media screen and (min-width: 769px){.modal-content,.modal-card{margin:0 auto;max-height:calc(100vh - 40px);width:640px}}.modal-close{background:none;height:40px;position:fixed;right:20px;top:20px;width:40px}.modal-card{display:flex;flex-direction:column;max-height:calc(100vh - 40px);overflow:hidden;-ms-overflow-y:visible}.modal-card-head,.modal-card-foot{align-items:center;background-color:#f5f5f5;display:flex;flex-shrink:0;justify-content:flex-start;padding:20px;position:relative}.modal-card-head{border-bottom:1px solid #dbdbdb;border-top-left-radius:6px;border-top-right-radius:6px}.modal-card-title{color:#222;flex-grow:1;flex-shrink:0;font-size:1.5rem;line-height:1}.modal-card-foot{border-bottom-left-radius:6px;border-bottom-right-radius:6px;border-top:1px solid #dbdbdb}.modal-card-foot .button:not(:last-child){margin-right:.5em}.modal-card-body{-webkit-overflow-scrolling:touch;background-color:#fff;flex-grow:1;flex-shrink:1;overflow:auto;padding:20px}.navbar{background-color:#fff;min-height:3.25rem;position:relative;z-index:30}.navbar.is-white{background-color:#fff;color:#0a0a0a}.navbar.is-white .navbar-brand>.navbar-item,.navbar.is-white .navbar-brand .navbar-link{color:#0a0a0a}.navbar.is-white .navbar-brand>a.navbar-item:focus,.navbar.is-white .navbar-brand>a.navbar-item:hover,.navbar.is-white .navbar-brand>a.navbar-item.is-active,.navbar.is-white .navbar-brand .navbar-link:focus,.navbar.is-white .navbar-brand .navbar-link:hover,.navbar.is-white .navbar-brand .navbar-link.is-active{background-color:#f2f2f2;color:#0a0a0a}.navbar.is-white .navbar-brand .navbar-link::after{border-color:#0a0a0a}.navbar.is-white .navbar-burger{color:#0a0a0a}@media screen and (min-width: 1056px){.navbar.is-white .navbar-start>.navbar-item,.navbar.is-white .navbar-start .navbar-link,.navbar.is-white .navbar-end>.navbar-item,.navbar.is-white .navbar-end .navbar-link{color:#0a0a0a}.navbar.is-white .navbar-start>a.navbar-item:focus,.navbar.is-white .navbar-start>a.navbar-item:hover,.navbar.is-white .navbar-start>a.navbar-item.is-active,.navbar.is-white .navbar-start .navbar-link:focus,.navbar.is-white .navbar-start .navbar-link:hover,.navbar.is-white .navbar-start .navbar-link.is-active,.navbar.is-white .navbar-end>a.navbar-item:focus,.navbar.is-white .navbar-end>a.navbar-item:hover,.navbar.is-white .navbar-end>a.navbar-item.is-active,.navbar.is-white .navbar-end .navbar-link:focus,.navbar.is-white .navbar-end .navbar-link:hover,.navbar.is-white .navbar-end .navbar-link.is-active{background-color:#f2f2f2;color:#0a0a0a}.navbar.is-white .navbar-start .navbar-link::after,.navbar.is-white .navbar-end .navbar-link::after{border-color:#0a0a0a}.navbar.is-white .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-white .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-white .navbar-item.has-dropdown.is-active .navbar-link{background-color:#f2f2f2;color:#0a0a0a}.navbar.is-white .navbar-dropdown a.navbar-item.is-active{background-color:#fff;color:#0a0a0a}}.navbar.is-black{background-color:#0a0a0a;color:#fff}.navbar.is-black .navbar-brand>.navbar-item,.navbar.is-black .navbar-brand .navbar-link{color:#fff}.navbar.is-black .navbar-brand>a.navbar-item:focus,.navbar.is-black .navbar-brand>a.navbar-item:hover,.navbar.is-black .navbar-brand>a.navbar-item.is-active,.navbar.is-black .navbar-brand .navbar-link:focus,.navbar.is-black .navbar-brand .navbar-link:hover,.navbar.is-black .navbar-brand .navbar-link.is-active{background-color:#000;color:#fff}.navbar.is-black .navbar-brand .navbar-link::after{border-color:#fff}.navbar.is-black .navbar-burger{color:#fff}@media screen and (min-width: 1056px){.navbar.is-black .navbar-start>.navbar-item,.navbar.is-black .navbar-start .navbar-link,.navbar.is-black .navbar-end>.navbar-item,.navbar.is-black .navbar-end .navbar-link{color:#fff}.navbar.is-black .navbar-start>a.navbar-item:focus,.navbar.is-black .navbar-start>a.navbar-item:hover,.navbar.is-black .navbar-start>a.navbar-item.is-active,.navbar.is-black .navbar-start .navbar-link:focus,.navbar.is-black .navbar-start .navbar-link:hover,.navbar.is-black .navbar-start .navbar-link.is-active,.navbar.is-black .navbar-end>a.navbar-item:focus,.navbar.is-black .navbar-end>a.navbar-item:hover,.navbar.is-black .navbar-end>a.navbar-item.is-active,.navbar.is-black .navbar-end .navbar-link:focus,.navbar.is-black .navbar-end .navbar-link:hover,.navbar.is-black .navbar-end .navbar-link.is-active{background-color:#000;color:#fff}.navbar.is-black .navbar-start .navbar-link::after,.navbar.is-black .navbar-end .navbar-link::after{border-color:#fff}.navbar.is-black .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-black .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-black .navbar-item.has-dropdown.is-active .navbar-link{background-color:#000;color:#fff}.navbar.is-black .navbar-dropdown a.navbar-item.is-active{background-color:#0a0a0a;color:#fff}}.navbar.is-light{background-color:#f5f5f5;color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-brand>.navbar-item,.navbar.is-light .navbar-brand .navbar-link{color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-brand>a.navbar-item:focus,.navbar.is-light .navbar-brand>a.navbar-item:hover,.navbar.is-light .navbar-brand>a.navbar-item.is-active,.navbar.is-light .navbar-brand .navbar-link:focus,.navbar.is-light .navbar-brand .navbar-link:hover,.navbar.is-light .navbar-brand .navbar-link.is-active{background-color:#e8e8e8;color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-brand .navbar-link::after{border-color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-burger{color:rgba(0,0,0,0.7)}@media screen and (min-width: 1056px){.navbar.is-light .navbar-start>.navbar-item,.navbar.is-light .navbar-start .navbar-link,.navbar.is-light .navbar-end>.navbar-item,.navbar.is-light .navbar-end .navbar-link{color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-start>a.navbar-item:focus,.navbar.is-light .navbar-start>a.navbar-item:hover,.navbar.is-light .navbar-start>a.navbar-item.is-active,.navbar.is-light .navbar-start .navbar-link:focus,.navbar.is-light .navbar-start .navbar-link:hover,.navbar.is-light .navbar-start .navbar-link.is-active,.navbar.is-light .navbar-end>a.navbar-item:focus,.navbar.is-light .navbar-end>a.navbar-item:hover,.navbar.is-light .navbar-end>a.navbar-item.is-active,.navbar.is-light .navbar-end .navbar-link:focus,.navbar.is-light .navbar-end .navbar-link:hover,.navbar.is-light .navbar-end .navbar-link.is-active{background-color:#e8e8e8;color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-start .navbar-link::after,.navbar.is-light .navbar-end .navbar-link::after{border-color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-light .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-light .navbar-item.has-dropdown.is-active .navbar-link{background-color:#e8e8e8;color:rgba(0,0,0,0.7)}.navbar.is-light .navbar-dropdown a.navbar-item.is-active{background-color:#f5f5f5;color:rgba(0,0,0,0.7)}}.navbar.is-dark,.content kbd.navbar{background-color:#363636;color:#fff}.navbar.is-dark .navbar-brand>.navbar-item,.content kbd.navbar .navbar-brand>.navbar-item,.navbar.is-dark .navbar-brand .navbar-link,.content kbd.navbar .navbar-brand .navbar-link{color:#fff}.navbar.is-dark .navbar-brand>a.navbar-item:focus,.content kbd.navbar .navbar-brand>a.navbar-item:focus,.navbar.is-dark .navbar-brand>a.navbar-item:hover,.content kbd.navbar .navbar-brand>a.navbar-item:hover,.navbar.is-dark .navbar-brand>a.navbar-item.is-active,.content kbd.navbar .navbar-brand>a.navbar-item.is-active,.navbar.is-dark .navbar-brand .navbar-link:focus,.content kbd.navbar .navbar-brand .navbar-link:focus,.navbar.is-dark .navbar-brand .navbar-link:hover,.content kbd.navbar .navbar-brand .navbar-link:hover,.navbar.is-dark .navbar-brand .navbar-link.is-active,.content kbd.navbar .navbar-brand .navbar-link.is-active{background-color:#292929;color:#fff}.navbar.is-dark .navbar-brand .navbar-link::after,.content kbd.navbar .navbar-brand .navbar-link::after{border-color:#fff}.navbar.is-dark .navbar-burger,.content kbd.navbar .navbar-burger{color:#fff}@media screen and (min-width: 1056px){.navbar.is-dark .navbar-start>.navbar-item,.content kbd.navbar .navbar-start>.navbar-item,.navbar.is-dark .navbar-start .navbar-link,.content kbd.navbar .navbar-start .navbar-link,.navbar.is-dark .navbar-end>.navbar-item,.content kbd.navbar .navbar-end>.navbar-item,.navbar.is-dark .navbar-end .navbar-link,.content kbd.navbar .navbar-end .navbar-link{color:#fff}.navbar.is-dark .navbar-start>a.navbar-item:focus,.content kbd.navbar .navbar-start>a.navbar-item:focus,.navbar.is-dark .navbar-start>a.navbar-item:hover,.content kbd.navbar .navbar-start>a.navbar-item:hover,.navbar.is-dark .navbar-start>a.navbar-item.is-active,.content kbd.navbar .navbar-start>a.navbar-item.is-active,.navbar.is-dark .navbar-start .navbar-link:focus,.content kbd.navbar .navbar-start .navbar-link:focus,.navbar.is-dark .navbar-start .navbar-link:hover,.content kbd.navbar .navbar-start .navbar-link:hover,.navbar.is-dark .navbar-start .navbar-link.is-active,.content kbd.navbar .navbar-start .navbar-link.is-active,.navbar.is-dark .navbar-end>a.navbar-item:focus,.content kbd.navbar .navbar-end>a.navbar-item:focus,.navbar.is-dark .navbar-end>a.navbar-item:hover,.content kbd.navbar .navbar-end>a.navbar-item:hover,.navbar.is-dark .navbar-end>a.navbar-item.is-active,.content kbd.navbar .navbar-end>a.navbar-item.is-active,.navbar.is-dark .navbar-end .navbar-link:focus,.content kbd.navbar .navbar-end .navbar-link:focus,.navbar.is-dark .navbar-end .navbar-link:hover,.content kbd.navbar .navbar-end .navbar-link:hover,.navbar.is-dark .navbar-end .navbar-link.is-active,.content kbd.navbar .navbar-end .navbar-link.is-active{background-color:#292929;color:#fff}.navbar.is-dark .navbar-start .navbar-link::after,.content kbd.navbar .navbar-start .navbar-link::after,.navbar.is-dark .navbar-end .navbar-link::after,.content kbd.navbar .navbar-end .navbar-link::after{border-color:#fff}.navbar.is-dark .navbar-item.has-dropdown:focus .navbar-link,.content kbd.navbar .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-dark .navbar-item.has-dropdown:hover .navbar-link,.content kbd.navbar .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-dark .navbar-item.has-dropdown.is-active .navbar-link,.content kbd.navbar .navbar-item.has-dropdown.is-active .navbar-link{background-color:#292929;color:#fff}.navbar.is-dark .navbar-dropdown a.navbar-item.is-active,.content kbd.navbar .navbar-dropdown a.navbar-item.is-active{background-color:#363636;color:#fff}}.navbar.is-primary,.docstring>section>a.navbar.docs-sourcelink{background-color:#4eb5de;color:#fff}.navbar.is-primary .navbar-brand>.navbar-item,.docstring>section>a.navbar.docs-sourcelink .navbar-brand>.navbar-item,.navbar.is-primary .navbar-brand .navbar-link,.docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link{color:#fff}.navbar.is-primary .navbar-brand>a.navbar-item:focus,.docstring>section>a.navbar.docs-sourcelink .navbar-brand>a.navbar-item:focus,.navbar.is-primary .navbar-brand>a.navbar-item:hover,.docstring>section>a.navbar.docs-sourcelink .navbar-brand>a.navbar-item:hover,.navbar.is-primary .navbar-brand>a.navbar-item.is-active,.docstring>section>a.navbar.docs-sourcelink .navbar-brand>a.navbar-item.is-active,.navbar.is-primary .navbar-brand .navbar-link:focus,.docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link:focus,.navbar.is-primary .navbar-brand .navbar-link:hover,.docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link:hover,.navbar.is-primary .navbar-brand .navbar-link.is-active,.docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link.is-active{background-color:#39acda;color:#fff}.navbar.is-primary .navbar-brand .navbar-link::after,.docstring>section>a.navbar.docs-sourcelink .navbar-brand .navbar-link::after{border-color:#fff}.navbar.is-primary .navbar-burger,.docstring>section>a.navbar.docs-sourcelink .navbar-burger{color:#fff}@media screen and (min-width: 1056px){.navbar.is-primary .navbar-start>.navbar-item,.docstring>section>a.navbar.docs-sourcelink .navbar-start>.navbar-item,.navbar.is-primary .navbar-start .navbar-link,.docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link,.navbar.is-primary .navbar-end>.navbar-item,.docstring>section>a.navbar.docs-sourcelink .navbar-end>.navbar-item,.navbar.is-primary .navbar-end .navbar-link,.docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link{color:#fff}.navbar.is-primary .navbar-start>a.navbar-item:focus,.docstring>section>a.navbar.docs-sourcelink .navbar-start>a.navbar-item:focus,.navbar.is-primary .navbar-start>a.navbar-item:hover,.docstring>section>a.navbar.docs-sourcelink .navbar-start>a.navbar-item:hover,.navbar.is-primary .navbar-start>a.navbar-item.is-active,.docstring>section>a.navbar.docs-sourcelink .navbar-start>a.navbar-item.is-active,.navbar.is-primary .navbar-start .navbar-link:focus,.docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link:focus,.navbar.is-primary .navbar-start .navbar-link:hover,.docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link:hover,.navbar.is-primary .navbar-start .navbar-link.is-active,.docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link.is-active,.navbar.is-primary .navbar-end>a.navbar-item:focus,.docstring>section>a.navbar.docs-sourcelink .navbar-end>a.navbar-item:focus,.navbar.is-primary .navbar-end>a.navbar-item:hover,.docstring>section>a.navbar.docs-sourcelink .navbar-end>a.navbar-item:hover,.navbar.is-primary .navbar-end>a.navbar-item.is-active,.docstring>section>a.navbar.docs-sourcelink .navbar-end>a.navbar-item.is-active,.navbar.is-primary .navbar-end .navbar-link:focus,.docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link:focus,.navbar.is-primary .navbar-end .navbar-link:hover,.docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link:hover,.navbar.is-primary .navbar-end .navbar-link.is-active,.docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link.is-active{background-color:#39acda;color:#fff}.navbar.is-primary .navbar-start .navbar-link::after,.docstring>section>a.navbar.docs-sourcelink .navbar-start .navbar-link::after,.navbar.is-primary .navbar-end .navbar-link::after,.docstring>section>a.navbar.docs-sourcelink .navbar-end .navbar-link::after{border-color:#fff}.navbar.is-primary .navbar-item.has-dropdown:focus .navbar-link,.docstring>section>a.navbar.docs-sourcelink .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-primary .navbar-item.has-dropdown:hover .navbar-link,.docstring>section>a.navbar.docs-sourcelink .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-primary .navbar-item.has-dropdown.is-active .navbar-link,.docstring>section>a.navbar.docs-sourcelink .navbar-item.has-dropdown.is-active .navbar-link{background-color:#39acda;color:#fff}.navbar.is-primary .navbar-dropdown a.navbar-item.is-active,.docstring>section>a.navbar.docs-sourcelink .navbar-dropdown a.navbar-item.is-active{background-color:#4eb5de;color:#fff}}.navbar.is-link{background-color:#2e63b8;color:#fff}.navbar.is-link .navbar-brand>.navbar-item,.navbar.is-link .navbar-brand .navbar-link{color:#fff}.navbar.is-link .navbar-brand>a.navbar-item:focus,.navbar.is-link .navbar-brand>a.navbar-item:hover,.navbar.is-link .navbar-brand>a.navbar-item.is-active,.navbar.is-link .navbar-brand .navbar-link:focus,.navbar.is-link .navbar-brand .navbar-link:hover,.navbar.is-link .navbar-brand .navbar-link.is-active{background-color:#2958a4;color:#fff}.navbar.is-link .navbar-brand .navbar-link::after{border-color:#fff}.navbar.is-link .navbar-burger{color:#fff}@media screen and (min-width: 1056px){.navbar.is-link .navbar-start>.navbar-item,.navbar.is-link .navbar-start .navbar-link,.navbar.is-link .navbar-end>.navbar-item,.navbar.is-link .navbar-end .navbar-link{color:#fff}.navbar.is-link .navbar-start>a.navbar-item:focus,.navbar.is-link .navbar-start>a.navbar-item:hover,.navbar.is-link .navbar-start>a.navbar-item.is-active,.navbar.is-link .navbar-start .navbar-link:focus,.navbar.is-link .navbar-start .navbar-link:hover,.navbar.is-link .navbar-start .navbar-link.is-active,.navbar.is-link .navbar-end>a.navbar-item:focus,.navbar.is-link .navbar-end>a.navbar-item:hover,.navbar.is-link .navbar-end>a.navbar-item.is-active,.navbar.is-link .navbar-end .navbar-link:focus,.navbar.is-link .navbar-end .navbar-link:hover,.navbar.is-link .navbar-end .navbar-link.is-active{background-color:#2958a4;color:#fff}.navbar.is-link .navbar-start .navbar-link::after,.navbar.is-link .navbar-end .navbar-link::after{border-color:#fff}.navbar.is-link .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-link .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-link .navbar-item.has-dropdown.is-active .navbar-link{background-color:#2958a4;color:#fff}.navbar.is-link .navbar-dropdown a.navbar-item.is-active{background-color:#2e63b8;color:#fff}}.navbar.is-info{background-color:#209cee;color:#fff}.navbar.is-info .navbar-brand>.navbar-item,.navbar.is-info .navbar-brand .navbar-link{color:#fff}.navbar.is-info .navbar-brand>a.navbar-item:focus,.navbar.is-info .navbar-brand>a.navbar-item:hover,.navbar.is-info .navbar-brand>a.navbar-item.is-active,.navbar.is-info .navbar-brand .navbar-link:focus,.navbar.is-info .navbar-brand .navbar-link:hover,.navbar.is-info .navbar-brand .navbar-link.is-active{background-color:#1190e3;color:#fff}.navbar.is-info .navbar-brand .navbar-link::after{border-color:#fff}.navbar.is-info .navbar-burger{color:#fff}@media screen and (min-width: 1056px){.navbar.is-info .navbar-start>.navbar-item,.navbar.is-info .navbar-start .navbar-link,.navbar.is-info .navbar-end>.navbar-item,.navbar.is-info .navbar-end .navbar-link{color:#fff}.navbar.is-info .navbar-start>a.navbar-item:focus,.navbar.is-info .navbar-start>a.navbar-item:hover,.navbar.is-info .navbar-start>a.navbar-item.is-active,.navbar.is-info .navbar-start .navbar-link:focus,.navbar.is-info .navbar-start .navbar-link:hover,.navbar.is-info .navbar-start .navbar-link.is-active,.navbar.is-info .navbar-end>a.navbar-item:focus,.navbar.is-info .navbar-end>a.navbar-item:hover,.navbar.is-info .navbar-end>a.navbar-item.is-active,.navbar.is-info .navbar-end .navbar-link:focus,.navbar.is-info .navbar-end .navbar-link:hover,.navbar.is-info .navbar-end .navbar-link.is-active{background-color:#1190e3;color:#fff}.navbar.is-info .navbar-start .navbar-link::after,.navbar.is-info .navbar-end .navbar-link::after{border-color:#fff}.navbar.is-info .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-info .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-info .navbar-item.has-dropdown.is-active .navbar-link{background-color:#1190e3;color:#fff}.navbar.is-info .navbar-dropdown a.navbar-item.is-active{background-color:#209cee;color:#fff}}.navbar.is-success{background-color:#22c35b;color:#fff}.navbar.is-success .navbar-brand>.navbar-item,.navbar.is-success .navbar-brand .navbar-link{color:#fff}.navbar.is-success .navbar-brand>a.navbar-item:focus,.navbar.is-success .navbar-brand>a.navbar-item:hover,.navbar.is-success .navbar-brand>a.navbar-item.is-active,.navbar.is-success .navbar-brand .navbar-link:focus,.navbar.is-success .navbar-brand .navbar-link:hover,.navbar.is-success .navbar-brand .navbar-link.is-active{background-color:#1ead51;color:#fff}.navbar.is-success .navbar-brand .navbar-link::after{border-color:#fff}.navbar.is-success .navbar-burger{color:#fff}@media screen and (min-width: 1056px){.navbar.is-success .navbar-start>.navbar-item,.navbar.is-success .navbar-start .navbar-link,.navbar.is-success .navbar-end>.navbar-item,.navbar.is-success .navbar-end .navbar-link{color:#fff}.navbar.is-success .navbar-start>a.navbar-item:focus,.navbar.is-success .navbar-start>a.navbar-item:hover,.navbar.is-success .navbar-start>a.navbar-item.is-active,.navbar.is-success .navbar-start .navbar-link:focus,.navbar.is-success .navbar-start .navbar-link:hover,.navbar.is-success .navbar-start .navbar-link.is-active,.navbar.is-success .navbar-end>a.navbar-item:focus,.navbar.is-success .navbar-end>a.navbar-item:hover,.navbar.is-success .navbar-end>a.navbar-item.is-active,.navbar.is-success .navbar-end .navbar-link:focus,.navbar.is-success .navbar-end .navbar-link:hover,.navbar.is-success .navbar-end .navbar-link.is-active{background-color:#1ead51;color:#fff}.navbar.is-success .navbar-start .navbar-link::after,.navbar.is-success .navbar-end .navbar-link::after{border-color:#fff}.navbar.is-success .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-success .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-success .navbar-item.has-dropdown.is-active .navbar-link{background-color:#1ead51;color:#fff}.navbar.is-success .navbar-dropdown a.navbar-item.is-active{background-color:#22c35b;color:#fff}}.navbar.is-warning{background-color:#ffdd57;color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-brand>.navbar-item,.navbar.is-warning .navbar-brand .navbar-link{color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-brand>a.navbar-item:focus,.navbar.is-warning .navbar-brand>a.navbar-item:hover,.navbar.is-warning .navbar-brand>a.navbar-item.is-active,.navbar.is-warning .navbar-brand .navbar-link:focus,.navbar.is-warning .navbar-brand .navbar-link:hover,.navbar.is-warning .navbar-brand .navbar-link.is-active{background-color:#ffd83e;color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-brand .navbar-link::after{border-color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-burger{color:rgba(0,0,0,0.7)}@media screen and (min-width: 1056px){.navbar.is-warning .navbar-start>.navbar-item,.navbar.is-warning .navbar-start .navbar-link,.navbar.is-warning .navbar-end>.navbar-item,.navbar.is-warning .navbar-end .navbar-link{color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-start>a.navbar-item:focus,.navbar.is-warning .navbar-start>a.navbar-item:hover,.navbar.is-warning .navbar-start>a.navbar-item.is-active,.navbar.is-warning .navbar-start .navbar-link:focus,.navbar.is-warning .navbar-start .navbar-link:hover,.navbar.is-warning .navbar-start .navbar-link.is-active,.navbar.is-warning .navbar-end>a.navbar-item:focus,.navbar.is-warning .navbar-end>a.navbar-item:hover,.navbar.is-warning .navbar-end>a.navbar-item.is-active,.navbar.is-warning .navbar-end .navbar-link:focus,.navbar.is-warning .navbar-end .navbar-link:hover,.navbar.is-warning .navbar-end .navbar-link.is-active{background-color:#ffd83e;color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-start .navbar-link::after,.navbar.is-warning .navbar-end .navbar-link::after{border-color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-warning .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-warning .navbar-item.has-dropdown.is-active .navbar-link{background-color:#ffd83e;color:rgba(0,0,0,0.7)}.navbar.is-warning .navbar-dropdown a.navbar-item.is-active{background-color:#ffdd57;color:rgba(0,0,0,0.7)}}.navbar.is-danger{background-color:#da0b00;color:#fff}.navbar.is-danger .navbar-brand>.navbar-item,.navbar.is-danger .navbar-brand .navbar-link{color:#fff}.navbar.is-danger .navbar-brand>a.navbar-item:focus,.navbar.is-danger .navbar-brand>a.navbar-item:hover,.navbar.is-danger .navbar-brand>a.navbar-item.is-active,.navbar.is-danger .navbar-brand .navbar-link:focus,.navbar.is-danger .navbar-brand .navbar-link:hover,.navbar.is-danger .navbar-brand .navbar-link.is-active{background-color:#c10a00;color:#fff}.navbar.is-danger .navbar-brand .navbar-link::after{border-color:#fff}.navbar.is-danger .navbar-burger{color:#fff}@media screen and (min-width: 1056px){.navbar.is-danger .navbar-start>.navbar-item,.navbar.is-danger .navbar-start .navbar-link,.navbar.is-danger .navbar-end>.navbar-item,.navbar.is-danger .navbar-end .navbar-link{color:#fff}.navbar.is-danger .navbar-start>a.navbar-item:focus,.navbar.is-danger .navbar-start>a.navbar-item:hover,.navbar.is-danger .navbar-start>a.navbar-item.is-active,.navbar.is-danger .navbar-start .navbar-link:focus,.navbar.is-danger .navbar-start .navbar-link:hover,.navbar.is-danger .navbar-start .navbar-link.is-active,.navbar.is-danger .navbar-end>a.navbar-item:focus,.navbar.is-danger .navbar-end>a.navbar-item:hover,.navbar.is-danger .navbar-end>a.navbar-item.is-active,.navbar.is-danger .navbar-end .navbar-link:focus,.navbar.is-danger .navbar-end .navbar-link:hover,.navbar.is-danger .navbar-end .navbar-link.is-active{background-color:#c10a00;color:#fff}.navbar.is-danger .navbar-start .navbar-link::after,.navbar.is-danger .navbar-end .navbar-link::after{border-color:#fff}.navbar.is-danger .navbar-item.has-dropdown:focus .navbar-link,.navbar.is-danger .navbar-item.has-dropdown:hover .navbar-link,.navbar.is-danger .navbar-item.has-dropdown.is-active .navbar-link{background-color:#c10a00;color:#fff}.navbar.is-danger .navbar-dropdown a.navbar-item.is-active{background-color:#da0b00;color:#fff}}.navbar>.container{align-items:stretch;display:flex;min-height:3.25rem;width:100%}.navbar.has-shadow{box-shadow:0 2px 0 0 #f5f5f5}.navbar.is-fixed-bottom,.navbar.is-fixed-top{left:0;position:fixed;right:0;z-index:30}.navbar.is-fixed-bottom{bottom:0}.navbar.is-fixed-bottom.has-shadow{box-shadow:0 -2px 0 0 #f5f5f5}.navbar.is-fixed-top{top:0}html.has-navbar-fixed-top,body.has-navbar-fixed-top{padding-top:3.25rem}html.has-navbar-fixed-bottom,body.has-navbar-fixed-bottom{padding-bottom:3.25rem}.navbar-brand,.navbar-tabs{align-items:stretch;display:flex;flex-shrink:0;min-height:3.25rem}.navbar-brand a.navbar-item:focus,.navbar-brand a.navbar-item:hover{background-color:transparent}.navbar-tabs{-webkit-overflow-scrolling:touch;max-width:100vw;overflow-x:auto;overflow-y:hidden}.navbar-burger{color:#222;-moz-appearance:none;-webkit-appearance:none;appearance:none;background:none;border:none;cursor:pointer;display:block;height:3.25rem;position:relative;width:3.25rem;margin-left:auto}.navbar-burger span{background-color:currentColor;display:block;height:1px;left:calc(50% - 8px);position:absolute;transform-origin:center;transition-duration:86ms;transition-property:background-color, opacity, transform;transition-timing-function:ease-out;width:16px}.navbar-burger span:nth-child(1){top:calc(50% - 6px)}.navbar-burger span:nth-child(2){top:calc(50% - 1px)}.navbar-burger span:nth-child(3){top:calc(50% + 4px)}.navbar-burger:hover{background-color:rgba(0,0,0,0.05)}.navbar-burger.is-active span:nth-child(1){transform:translateY(5px) rotate(45deg)}.navbar-burger.is-active span:nth-child(2){opacity:0}.navbar-burger.is-active span:nth-child(3){transform:translateY(-5px) rotate(-45deg)}.navbar-menu{display:none}.navbar-item,.navbar-link{color:#222;display:block;line-height:1.5;padding:0.5rem 0.75rem;position:relative}.navbar-item .icon:only-child,.navbar-link .icon:only-child{margin-left:-0.25rem;margin-right:-0.25rem}a.navbar-item,.navbar-link{cursor:pointer}a.navbar-item:focus,a.navbar-item:focus-within,a.navbar-item:hover,a.navbar-item.is-active,.navbar-link:focus,.navbar-link:focus-within,.navbar-link:hover,.navbar-link.is-active{background-color:#fafafa;color:#2e63b8}.navbar-item{flex-grow:0;flex-shrink:0}.navbar-item img{max-height:1.75rem}.navbar-item.has-dropdown{padding:0}.navbar-item.is-expanded{flex-grow:1;flex-shrink:1}.navbar-item.is-tab{border-bottom:1px solid transparent;min-height:3.25rem;padding-bottom:calc(0.5rem - 1px)}.navbar-item.is-tab:focus,.navbar-item.is-tab:hover{background-color:rgba(0,0,0,0);border-bottom-color:#2e63b8}.navbar-item.is-tab.is-active{background-color:rgba(0,0,0,0);border-bottom-color:#2e63b8;border-bottom-style:solid;border-bottom-width:3px;color:#2e63b8;padding-bottom:calc(0.5rem - 3px)}.navbar-content{flex-grow:1;flex-shrink:1}.navbar-link:not(.is-arrowless){padding-right:2.5em}.navbar-link:not(.is-arrowless)::after{border-color:#2e63b8;margin-top:-0.375em;right:1.125em}.navbar-dropdown{font-size:0.875rem;padding-bottom:0.5rem;padding-top:0.5rem}.navbar-dropdown .navbar-item{padding-left:1.5rem;padding-right:1.5rem}.navbar-divider{background-color:#f5f5f5;border:none;display:none;height:2px;margin:0.5rem 0}@media screen and (max-width: 1055px){.navbar>.container{display:block}.navbar-brand .navbar-item,.navbar-tabs .navbar-item{align-items:center;display:flex}.navbar-link::after{display:none}.navbar-menu{background-color:#fff;box-shadow:0 8px 16px rgba(10,10,10,0.1);padding:0.5rem 0}.navbar-menu.is-active{display:block}.navbar.is-fixed-bottom-touch,.navbar.is-fixed-top-touch{left:0;position:fixed;right:0;z-index:30}.navbar.is-fixed-bottom-touch{bottom:0}.navbar.is-fixed-bottom-touch.has-shadow{box-shadow:0 -2px 3px rgba(10,10,10,0.1)}.navbar.is-fixed-top-touch{top:0}.navbar.is-fixed-top .navbar-menu,.navbar.is-fixed-top-touch .navbar-menu{-webkit-overflow-scrolling:touch;max-height:calc(100vh - 3.25rem);overflow:auto}html.has-navbar-fixed-top-touch,body.has-navbar-fixed-top-touch{padding-top:3.25rem}html.has-navbar-fixed-bottom-touch,body.has-navbar-fixed-bottom-touch{padding-bottom:3.25rem}}@media screen and (min-width: 1056px){.navbar,.navbar-menu,.navbar-start,.navbar-end{align-items:stretch;display:flex}.navbar{min-height:3.25rem}.navbar.is-spaced{padding:1rem 2rem}.navbar.is-spaced .navbar-start,.navbar.is-spaced .navbar-end{align-items:center}.navbar.is-spaced a.navbar-item,.navbar.is-spaced .navbar-link{border-radius:4px}.navbar.is-transparent a.navbar-item:focus,.navbar.is-transparent a.navbar-item:hover,.navbar.is-transparent a.navbar-item.is-active,.navbar.is-transparent .navbar-link:focus,.navbar.is-transparent .navbar-link:hover,.navbar.is-transparent .navbar-link.is-active{background-color:transparent !important}.navbar.is-transparent .navbar-item.has-dropdown.is-active .navbar-link,.navbar.is-transparent .navbar-item.has-dropdown.is-hoverable:focus .navbar-link,.navbar.is-transparent .navbar-item.has-dropdown.is-hoverable:focus-within .navbar-link,.navbar.is-transparent .navbar-item.has-dropdown.is-hoverable:hover .navbar-link{background-color:transparent !important}.navbar.is-transparent .navbar-dropdown a.navbar-item:focus,.navbar.is-transparent .navbar-dropdown a.navbar-item:hover{background-color:#f5f5f5;color:#0a0a0a}.navbar.is-transparent .navbar-dropdown a.navbar-item.is-active{background-color:#f5f5f5;color:#2e63b8}.navbar-burger{display:none}.navbar-item,.navbar-link{align-items:center;display:flex}.navbar-item.has-dropdown{align-items:stretch}.navbar-item.has-dropdown-up .navbar-link::after{transform:rotate(135deg) translate(0.25em, -0.25em)}.navbar-item.has-dropdown-up .navbar-dropdown{border-bottom:2px solid #dbdbdb;border-radius:6px 6px 0 0;border-top:none;bottom:100%;box-shadow:0 -8px 8px rgba(10,10,10,0.1);top:auto}.navbar-item.is-active .navbar-dropdown,.navbar-item.is-hoverable:focus .navbar-dropdown,.navbar-item.is-hoverable:focus-within .navbar-dropdown,.navbar-item.is-hoverable:hover .navbar-dropdown{display:block}.navbar.is-spaced .navbar-item.is-active .navbar-dropdown,.navbar-item.is-active .navbar-dropdown.is-boxed,.navbar.is-spaced .navbar-item.is-hoverable:focus .navbar-dropdown,.navbar-item.is-hoverable:focus .navbar-dropdown.is-boxed,.navbar.is-spaced .navbar-item.is-hoverable:focus-within .navbar-dropdown,.navbar-item.is-hoverable:focus-within .navbar-dropdown.is-boxed,.navbar.is-spaced .navbar-item.is-hoverable:hover .navbar-dropdown,.navbar-item.is-hoverable:hover .navbar-dropdown.is-boxed{opacity:1;pointer-events:auto;transform:translateY(0)}.navbar-menu{flex-grow:1;flex-shrink:0}.navbar-start{justify-content:flex-start;margin-right:auto}.navbar-end{justify-content:flex-end;margin-left:auto}.navbar-dropdown{background-color:#fff;border-bottom-left-radius:6px;border-bottom-right-radius:6px;border-top:2px solid #dbdbdb;box-shadow:0 8px 8px rgba(10,10,10,0.1);display:none;font-size:0.875rem;left:0;min-width:100%;position:absolute;top:100%;z-index:20}.navbar-dropdown .navbar-item{padding:0.375rem 1rem;white-space:nowrap}.navbar-dropdown a.navbar-item{padding-right:3rem}.navbar-dropdown a.navbar-item:focus,.navbar-dropdown a.navbar-item:hover{background-color:#f5f5f5;color:#0a0a0a}.navbar-dropdown a.navbar-item.is-active{background-color:#f5f5f5;color:#2e63b8}.navbar.is-spaced .navbar-dropdown,.navbar-dropdown.is-boxed{border-radius:6px;border-top:none;box-shadow:0 8px 8px rgba(10,10,10,0.1), 0 0 0 1px rgba(10,10,10,0.1);display:block;opacity:0;pointer-events:none;top:calc(100% + (-4px));transform:translateY(-5px);transition-duration:86ms;transition-property:opacity, transform}.navbar-dropdown.is-right{left:auto;right:0}.navbar-divider{display:block}.navbar>.container .navbar-brand,.container>.navbar .navbar-brand{margin-left:-.75rem}.navbar>.container .navbar-menu,.container>.navbar .navbar-menu{margin-right:-.75rem}.navbar.is-fixed-bottom-desktop,.navbar.is-fixed-top-desktop{left:0;position:fixed;right:0;z-index:30}.navbar.is-fixed-bottom-desktop{bottom:0}.navbar.is-fixed-bottom-desktop.has-shadow{box-shadow:0 -2px 3px rgba(10,10,10,0.1)}.navbar.is-fixed-top-desktop{top:0}html.has-navbar-fixed-top-desktop,body.has-navbar-fixed-top-desktop{padding-top:3.25rem}html.has-navbar-fixed-bottom-desktop,body.has-navbar-fixed-bottom-desktop{padding-bottom:3.25rem}html.has-spaced-navbar-fixed-top,body.has-spaced-navbar-fixed-top{padding-top:5.25rem}html.has-spaced-navbar-fixed-bottom,body.has-spaced-navbar-fixed-bottom{padding-bottom:5.25rem}a.navbar-item.is-active,.navbar-link.is-active{color:#0a0a0a}a.navbar-item.is-active:not(:focus):not(:hover),.navbar-link.is-active:not(:focus):not(:hover){background-color:rgba(0,0,0,0)}.navbar-item.has-dropdown:focus .navbar-link,.navbar-item.has-dropdown:hover .navbar-link,.navbar-item.has-dropdown.is-active .navbar-link{background-color:#fafafa}}.hero.is-fullheight-with-navbar{min-height:calc(100vh - 3.25rem)}.pagination{font-size:1rem;margin:-.25rem}.pagination.is-small,#documenter .docs-sidebar form.docs-search>input.pagination{font-size:.75rem}.pagination.is-medium{font-size:1.25rem}.pagination.is-large{font-size:1.5rem}.pagination.is-rounded .pagination-previous,#documenter .docs-sidebar form.docs-search>input.pagination .pagination-previous,.pagination.is-rounded .pagination-next,#documenter .docs-sidebar form.docs-search>input.pagination .pagination-next{padding-left:1em;padding-right:1em;border-radius:9999px}.pagination.is-rounded .pagination-link,#documenter .docs-sidebar form.docs-search>input.pagination .pagination-link{border-radius:9999px}.pagination,.pagination-list{align-items:center;display:flex;justify-content:center;text-align:center}.pagination-previous,.pagination-next,.pagination-link,.pagination-ellipsis{font-size:1em;justify-content:center;margin:.25rem;padding-left:.5em;padding-right:.5em;text-align:center}.pagination-previous,.pagination-next,.pagination-link{border-color:#dbdbdb;color:#222;min-width:2.5em}.pagination-previous:hover,.pagination-next:hover,.pagination-link:hover{border-color:#b5b5b5;color:#363636}.pagination-previous:focus,.pagination-next:focus,.pagination-link:focus{border-color:#3c5dcd}.pagination-previous:active,.pagination-next:active,.pagination-link:active{box-shadow:inset 0 1px 2px rgba(10,10,10,0.2)}.pagination-previous[disabled],.pagination-previous.is-disabled,.pagination-next[disabled],.pagination-next.is-disabled,.pagination-link[disabled],.pagination-link.is-disabled{background-color:#dbdbdb;border-color:#dbdbdb;box-shadow:none;color:#6b6b6b;opacity:0.5}.pagination-previous,.pagination-next{padding-left:.75em;padding-right:.75em;white-space:nowrap}.pagination-link.is-current{background-color:#2e63b8;border-color:#2e63b8;color:#fff}.pagination-ellipsis{color:#b5b5b5;pointer-events:none}.pagination-list{flex-wrap:wrap}.pagination-list li{list-style:none}@media screen and (max-width: 768px){.pagination{flex-wrap:wrap}.pagination-previous,.pagination-next{flex-grow:1;flex-shrink:1}.pagination-list li{flex-grow:1;flex-shrink:1}}@media screen and (min-width: 769px),print{.pagination-list{flex-grow:1;flex-shrink:1;justify-content:flex-start;order:1}.pagination-previous,.pagination-next,.pagination-link,.pagination-ellipsis{margin-bottom:0;margin-top:0}.pagination-previous{order:2}.pagination-next{order:3}.pagination{justify-content:space-between;margin-bottom:0;margin-top:0}.pagination.is-centered .pagination-previous{order:1}.pagination.is-centered .pagination-list{justify-content:center;order:2}.pagination.is-centered .pagination-next{order:3}.pagination.is-right .pagination-previous{order:1}.pagination.is-right .pagination-next{order:2}.pagination.is-right .pagination-list{justify-content:flex-end;order:3}}.panel{border-radius:6px;box-shadow:#bbb;font-size:1rem}.panel:not(:last-child){margin-bottom:1.5rem}.panel.is-white .panel-heading{background-color:#fff;color:#0a0a0a}.panel.is-white .panel-tabs a.is-active{border-bottom-color:#fff}.panel.is-white .panel-block.is-active .panel-icon{color:#fff}.panel.is-black .panel-heading{background-color:#0a0a0a;color:#fff}.panel.is-black .panel-tabs a.is-active{border-bottom-color:#0a0a0a}.panel.is-black .panel-block.is-active .panel-icon{color:#0a0a0a}.panel.is-light .panel-heading{background-color:#f5f5f5;color:rgba(0,0,0,0.7)}.panel.is-light .panel-tabs a.is-active{border-bottom-color:#f5f5f5}.panel.is-light .panel-block.is-active .panel-icon{color:#f5f5f5}.panel.is-dark .panel-heading,.content kbd.panel .panel-heading{background-color:#363636;color:#fff}.panel.is-dark .panel-tabs a.is-active,.content kbd.panel .panel-tabs a.is-active{border-bottom-color:#363636}.panel.is-dark .panel-block.is-active .panel-icon,.content kbd.panel .panel-block.is-active .panel-icon{color:#363636}.panel.is-primary .panel-heading,.docstring>section>a.panel.docs-sourcelink .panel-heading{background-color:#4eb5de;color:#fff}.panel.is-primary .panel-tabs a.is-active,.docstring>section>a.panel.docs-sourcelink .panel-tabs a.is-active{border-bottom-color:#4eb5de}.panel.is-primary .panel-block.is-active .panel-icon,.docstring>section>a.panel.docs-sourcelink .panel-block.is-active .panel-icon{color:#4eb5de}.panel.is-link .panel-heading{background-color:#2e63b8;color:#fff}.panel.is-link .panel-tabs a.is-active{border-bottom-color:#2e63b8}.panel.is-link .panel-block.is-active .panel-icon{color:#2e63b8}.panel.is-info .panel-heading{background-color:#209cee;color:#fff}.panel.is-info .panel-tabs a.is-active{border-bottom-color:#209cee}.panel.is-info .panel-block.is-active .panel-icon{color:#209cee}.panel.is-success .panel-heading{background-color:#22c35b;color:#fff}.panel.is-success .panel-tabs a.is-active{border-bottom-color:#22c35b}.panel.is-success .panel-block.is-active .panel-icon{color:#22c35b}.panel.is-warning .panel-heading{background-color:#ffdd57;color:rgba(0,0,0,0.7)}.panel.is-warning .panel-tabs a.is-active{border-bottom-color:#ffdd57}.panel.is-warning .panel-block.is-active .panel-icon{color:#ffdd57}.panel.is-danger .panel-heading{background-color:#da0b00;color:#fff}.panel.is-danger .panel-tabs a.is-active{border-bottom-color:#da0b00}.panel.is-danger .panel-block.is-active .panel-icon{color:#da0b00}.panel-tabs:not(:last-child),.panel-block:not(:last-child){border-bottom:1px solid #ededed}.panel-heading{background-color:#ededed;border-radius:6px 6px 0 0;color:#222;font-size:1.25em;font-weight:700;line-height:1.25;padding:0.75em 1em}.panel-tabs{align-items:flex-end;display:flex;font-size:.875em;justify-content:center}.panel-tabs a{border-bottom:1px solid #dbdbdb;margin-bottom:-1px;padding:0.5em}.panel-tabs a.is-active{border-bottom-color:#4a4a4a;color:#363636}.panel-list a{color:#222}.panel-list a:hover{color:#2e63b8}.panel-block{align-items:center;color:#222;display:flex;justify-content:flex-start;padding:0.5em 0.75em}.panel-block input[type="checkbox"]{margin-right:.75em}.panel-block>.control{flex-grow:1;flex-shrink:1;width:100%}.panel-block.is-wrapped{flex-wrap:wrap}.panel-block.is-active{border-left-color:#2e63b8;color:#363636}.panel-block.is-active .panel-icon{color:#2e63b8}.panel-block:last-child{border-bottom-left-radius:6px;border-bottom-right-radius:6px}a.panel-block,label.panel-block{cursor:pointer}a.panel-block:hover,label.panel-block:hover{background-color:#f5f5f5}.panel-icon{display:inline-block;font-size:14px;height:1em;line-height:1em;text-align:center;vertical-align:top;width:1em;color:#6b6b6b;margin-right:.75em}.panel-icon .fa{font-size:inherit;line-height:inherit}.tabs{-webkit-overflow-scrolling:touch;align-items:stretch;display:flex;font-size:1rem;justify-content:space-between;overflow:hidden;overflow-x:auto;white-space:nowrap}.tabs a{align-items:center;border-bottom-color:#dbdbdb;border-bottom-style:solid;border-bottom-width:1px;color:#222;display:flex;justify-content:center;margin-bottom:-1px;padding:0.5em 1em;vertical-align:top}.tabs a:hover{border-bottom-color:#222;color:#222}.tabs li{display:block}.tabs li.is-active a{border-bottom-color:#2e63b8;color:#2e63b8}.tabs ul{align-items:center;border-bottom-color:#dbdbdb;border-bottom-style:solid;border-bottom-width:1px;display:flex;flex-grow:1;flex-shrink:0;justify-content:flex-start}.tabs ul.is-left{padding-right:0.75em}.tabs ul.is-center{flex:none;justify-content:center;padding-left:0.75em;padding-right:0.75em}.tabs ul.is-right{justify-content:flex-end;padding-left:0.75em}.tabs .icon:first-child{margin-right:.5em}.tabs .icon:last-child{margin-left:.5em}.tabs.is-centered ul{justify-content:center}.tabs.is-right ul{justify-content:flex-end}.tabs.is-boxed a{border:1px solid transparent;border-radius:4px 4px 0 0}.tabs.is-boxed a:hover{background-color:#f5f5f5;border-bottom-color:#dbdbdb}.tabs.is-boxed li.is-active a{background-color:#fff;border-color:#dbdbdb;border-bottom-color:rgba(0,0,0,0) !important}.tabs.is-fullwidth li{flex-grow:1;flex-shrink:0}.tabs.is-toggle a{border-color:#dbdbdb;border-style:solid;border-width:1px;margin-bottom:0;position:relative}.tabs.is-toggle a:hover{background-color:#f5f5f5;border-color:#b5b5b5;z-index:2}.tabs.is-toggle li+li{margin-left:-1px}.tabs.is-toggle li:first-child a{border-top-left-radius:4px;border-bottom-left-radius:4px}.tabs.is-toggle li:last-child a{border-top-right-radius:4px;border-bottom-right-radius:4px}.tabs.is-toggle li.is-active a{background-color:#2e63b8;border-color:#2e63b8;color:#fff;z-index:1}.tabs.is-toggle ul{border-bottom:none}.tabs.is-toggle.is-toggle-rounded li:first-child a{border-bottom-left-radius:9999px;border-top-left-radius:9999px;padding-left:1.25em}.tabs.is-toggle.is-toggle-rounded li:last-child a{border-bottom-right-radius:9999px;border-top-right-radius:9999px;padding-right:1.25em}.tabs.is-small,#documenter .docs-sidebar form.docs-search>input.tabs{font-size:.75rem}.tabs.is-medium{font-size:1.25rem}.tabs.is-large{font-size:1.5rem}.column{display:block;flex-basis:0;flex-grow:1;flex-shrink:1;padding:.75rem}.columns.is-mobile>.column.is-narrow{flex:none;width:unset}.columns.is-mobile>.column.is-full{flex:none;width:100%}.columns.is-mobile>.column.is-three-quarters{flex:none;width:75%}.columns.is-mobile>.column.is-two-thirds{flex:none;width:66.6666%}.columns.is-mobile>.column.is-half{flex:none;width:50%}.columns.is-mobile>.column.is-one-third{flex:none;width:33.3333%}.columns.is-mobile>.column.is-one-quarter{flex:none;width:25%}.columns.is-mobile>.column.is-one-fifth{flex:none;width:20%}.columns.is-mobile>.column.is-two-fifths{flex:none;width:40%}.columns.is-mobile>.column.is-three-fifths{flex:none;width:60%}.columns.is-mobile>.column.is-four-fifths{flex:none;width:80%}.columns.is-mobile>.column.is-offset-three-quarters{margin-left:75%}.columns.is-mobile>.column.is-offset-two-thirds{margin-left:66.6666%}.columns.is-mobile>.column.is-offset-half{margin-left:50%}.columns.is-mobile>.column.is-offset-one-third{margin-left:33.3333%}.columns.is-mobile>.column.is-offset-one-quarter{margin-left:25%}.columns.is-mobile>.column.is-offset-one-fifth{margin-left:20%}.columns.is-mobile>.column.is-offset-two-fifths{margin-left:40%}.columns.is-mobile>.column.is-offset-three-fifths{margin-left:60%}.columns.is-mobile>.column.is-offset-four-fifths{margin-left:80%}.columns.is-mobile>.column.is-0{flex:none;width:0%}.columns.is-mobile>.column.is-offset-0{margin-left:0%}.columns.is-mobile>.column.is-1{flex:none;width:8.33333337%}.columns.is-mobile>.column.is-offset-1{margin-left:8.33333337%}.columns.is-mobile>.column.is-2{flex:none;width:16.66666674%}.columns.is-mobile>.column.is-offset-2{margin-left:16.66666674%}.columns.is-mobile>.column.is-3{flex:none;width:25%}.columns.is-mobile>.column.is-offset-3{margin-left:25%}.columns.is-mobile>.column.is-4{flex:none;width:33.33333337%}.columns.is-mobile>.column.is-offset-4{margin-left:33.33333337%}.columns.is-mobile>.column.is-5{flex:none;width:41.66666674%}.columns.is-mobile>.column.is-offset-5{margin-left:41.66666674%}.columns.is-mobile>.column.is-6{flex:none;width:50%}.columns.is-mobile>.column.is-offset-6{margin-left:50%}.columns.is-mobile>.column.is-7{flex:none;width:58.33333337%}.columns.is-mobile>.column.is-offset-7{margin-left:58.33333337%}.columns.is-mobile>.column.is-8{flex:none;width:66.66666674%}.columns.is-mobile>.column.is-offset-8{margin-left:66.66666674%}.columns.is-mobile>.column.is-9{flex:none;width:75%}.columns.is-mobile>.column.is-offset-9{margin-left:75%}.columns.is-mobile>.column.is-10{flex:none;width:83.33333337%}.columns.is-mobile>.column.is-offset-10{margin-left:83.33333337%}.columns.is-mobile>.column.is-11{flex:none;width:91.66666674%}.columns.is-mobile>.column.is-offset-11{margin-left:91.66666674%}.columns.is-mobile>.column.is-12{flex:none;width:100%}.columns.is-mobile>.column.is-offset-12{margin-left:100%}@media screen and (max-width: 768px){.column.is-narrow-mobile{flex:none;width:unset}.column.is-full-mobile{flex:none;width:100%}.column.is-three-quarters-mobile{flex:none;width:75%}.column.is-two-thirds-mobile{flex:none;width:66.6666%}.column.is-half-mobile{flex:none;width:50%}.column.is-one-third-mobile{flex:none;width:33.3333%}.column.is-one-quarter-mobile{flex:none;width:25%}.column.is-one-fifth-mobile{flex:none;width:20%}.column.is-two-fifths-mobile{flex:none;width:40%}.column.is-three-fifths-mobile{flex:none;width:60%}.column.is-four-fifths-mobile{flex:none;width:80%}.column.is-offset-three-quarters-mobile{margin-left:75%}.column.is-offset-two-thirds-mobile{margin-left:66.6666%}.column.is-offset-half-mobile{margin-left:50%}.column.is-offset-one-third-mobile{margin-left:33.3333%}.column.is-offset-one-quarter-mobile{margin-left:25%}.column.is-offset-one-fifth-mobile{margin-left:20%}.column.is-offset-two-fifths-mobile{margin-left:40%}.column.is-offset-three-fifths-mobile{margin-left:60%}.column.is-offset-four-fifths-mobile{margin-left:80%}.column.is-0-mobile{flex:none;width:0%}.column.is-offset-0-mobile{margin-left:0%}.column.is-1-mobile{flex:none;width:8.33333337%}.column.is-offset-1-mobile{margin-left:8.33333337%}.column.is-2-mobile{flex:none;width:16.66666674%}.column.is-offset-2-mobile{margin-left:16.66666674%}.column.is-3-mobile{flex:none;width:25%}.column.is-offset-3-mobile{margin-left:25%}.column.is-4-mobile{flex:none;width:33.33333337%}.column.is-offset-4-mobile{margin-left:33.33333337%}.column.is-5-mobile{flex:none;width:41.66666674%}.column.is-offset-5-mobile{margin-left:41.66666674%}.column.is-6-mobile{flex:none;width:50%}.column.is-offset-6-mobile{margin-left:50%}.column.is-7-mobile{flex:none;width:58.33333337%}.column.is-offset-7-mobile{margin-left:58.33333337%}.column.is-8-mobile{flex:none;width:66.66666674%}.column.is-offset-8-mobile{margin-left:66.66666674%}.column.is-9-mobile{flex:none;width:75%}.column.is-offset-9-mobile{margin-left:75%}.column.is-10-mobile{flex:none;width:83.33333337%}.column.is-offset-10-mobile{margin-left:83.33333337%}.column.is-11-mobile{flex:none;width:91.66666674%}.column.is-offset-11-mobile{margin-left:91.66666674%}.column.is-12-mobile{flex:none;width:100%}.column.is-offset-12-mobile{margin-left:100%}}@media screen and (min-width: 769px),print{.column.is-narrow,.column.is-narrow-tablet{flex:none;width:unset}.column.is-full,.column.is-full-tablet{flex:none;width:100%}.column.is-three-quarters,.column.is-three-quarters-tablet{flex:none;width:75%}.column.is-two-thirds,.column.is-two-thirds-tablet{flex:none;width:66.6666%}.column.is-half,.column.is-half-tablet{flex:none;width:50%}.column.is-one-third,.column.is-one-third-tablet{flex:none;width:33.3333%}.column.is-one-quarter,.column.is-one-quarter-tablet{flex:none;width:25%}.column.is-one-fifth,.column.is-one-fifth-tablet{flex:none;width:20%}.column.is-two-fifths,.column.is-two-fifths-tablet{flex:none;width:40%}.column.is-three-fifths,.column.is-three-fifths-tablet{flex:none;width:60%}.column.is-four-fifths,.column.is-four-fifths-tablet{flex:none;width:80%}.column.is-offset-three-quarters,.column.is-offset-three-quarters-tablet{margin-left:75%}.column.is-offset-two-thirds,.column.is-offset-two-thirds-tablet{margin-left:66.6666%}.column.is-offset-half,.column.is-offset-half-tablet{margin-left:50%}.column.is-offset-one-third,.column.is-offset-one-third-tablet{margin-left:33.3333%}.column.is-offset-one-quarter,.column.is-offset-one-quarter-tablet{margin-left:25%}.column.is-offset-one-fifth,.column.is-offset-one-fifth-tablet{margin-left:20%}.column.is-offset-two-fifths,.column.is-offset-two-fifths-tablet{margin-left:40%}.column.is-offset-three-fifths,.column.is-offset-three-fifths-tablet{margin-left:60%}.column.is-offset-four-fifths,.column.is-offset-four-fifths-tablet{margin-left:80%}.column.is-0,.column.is-0-tablet{flex:none;width:0%}.column.is-offset-0,.column.is-offset-0-tablet{margin-left:0%}.column.is-1,.column.is-1-tablet{flex:none;width:8.33333337%}.column.is-offset-1,.column.is-offset-1-tablet{margin-left:8.33333337%}.column.is-2,.column.is-2-tablet{flex:none;width:16.66666674%}.column.is-offset-2,.column.is-offset-2-tablet{margin-left:16.66666674%}.column.is-3,.column.is-3-tablet{flex:none;width:25%}.column.is-offset-3,.column.is-offset-3-tablet{margin-left:25%}.column.is-4,.column.is-4-tablet{flex:none;width:33.33333337%}.column.is-offset-4,.column.is-offset-4-tablet{margin-left:33.33333337%}.column.is-5,.column.is-5-tablet{flex:none;width:41.66666674%}.column.is-offset-5,.column.is-offset-5-tablet{margin-left:41.66666674%}.column.is-6,.column.is-6-tablet{flex:none;width:50%}.column.is-offset-6,.column.is-offset-6-tablet{margin-left:50%}.column.is-7,.column.is-7-tablet{flex:none;width:58.33333337%}.column.is-offset-7,.column.is-offset-7-tablet{margin-left:58.33333337%}.column.is-8,.column.is-8-tablet{flex:none;width:66.66666674%}.column.is-offset-8,.column.is-offset-8-tablet{margin-left:66.66666674%}.column.is-9,.column.is-9-tablet{flex:none;width:75%}.column.is-offset-9,.column.is-offset-9-tablet{margin-left:75%}.column.is-10,.column.is-10-tablet{flex:none;width:83.33333337%}.column.is-offset-10,.column.is-offset-10-tablet{margin-left:83.33333337%}.column.is-11,.column.is-11-tablet{flex:none;width:91.66666674%}.column.is-offset-11,.column.is-offset-11-tablet{margin-left:91.66666674%}.column.is-12,.column.is-12-tablet{flex:none;width:100%}.column.is-offset-12,.column.is-offset-12-tablet{margin-left:100%}}@media screen and (max-width: 1055px){.column.is-narrow-touch{flex:none;width:unset}.column.is-full-touch{flex:none;width:100%}.column.is-three-quarters-touch{flex:none;width:75%}.column.is-two-thirds-touch{flex:none;width:66.6666%}.column.is-half-touch{flex:none;width:50%}.column.is-one-third-touch{flex:none;width:33.3333%}.column.is-one-quarter-touch{flex:none;width:25%}.column.is-one-fifth-touch{flex:none;width:20%}.column.is-two-fifths-touch{flex:none;width:40%}.column.is-three-fifths-touch{flex:none;width:60%}.column.is-four-fifths-touch{flex:none;width:80%}.column.is-offset-three-quarters-touch{margin-left:75%}.column.is-offset-two-thirds-touch{margin-left:66.6666%}.column.is-offset-half-touch{margin-left:50%}.column.is-offset-one-third-touch{margin-left:33.3333%}.column.is-offset-one-quarter-touch{margin-left:25%}.column.is-offset-one-fifth-touch{margin-left:20%}.column.is-offset-two-fifths-touch{margin-left:40%}.column.is-offset-three-fifths-touch{margin-left:60%}.column.is-offset-four-fifths-touch{margin-left:80%}.column.is-0-touch{flex:none;width:0%}.column.is-offset-0-touch{margin-left:0%}.column.is-1-touch{flex:none;width:8.33333337%}.column.is-offset-1-touch{margin-left:8.33333337%}.column.is-2-touch{flex:none;width:16.66666674%}.column.is-offset-2-touch{margin-left:16.66666674%}.column.is-3-touch{flex:none;width:25%}.column.is-offset-3-touch{margin-left:25%}.column.is-4-touch{flex:none;width:33.33333337%}.column.is-offset-4-touch{margin-left:33.33333337%}.column.is-5-touch{flex:none;width:41.66666674%}.column.is-offset-5-touch{margin-left:41.66666674%}.column.is-6-touch{flex:none;width:50%}.column.is-offset-6-touch{margin-left:50%}.column.is-7-touch{flex:none;width:58.33333337%}.column.is-offset-7-touch{margin-left:58.33333337%}.column.is-8-touch{flex:none;width:66.66666674%}.column.is-offset-8-touch{margin-left:66.66666674%}.column.is-9-touch{flex:none;width:75%}.column.is-offset-9-touch{margin-left:75%}.column.is-10-touch{flex:none;width:83.33333337%}.column.is-offset-10-touch{margin-left:83.33333337%}.column.is-11-touch{flex:none;width:91.66666674%}.column.is-offset-11-touch{margin-left:91.66666674%}.column.is-12-touch{flex:none;width:100%}.column.is-offset-12-touch{margin-left:100%}}@media screen and (min-width: 1056px){.column.is-narrow-desktop{flex:none;width:unset}.column.is-full-desktop{flex:none;width:100%}.column.is-three-quarters-desktop{flex:none;width:75%}.column.is-two-thirds-desktop{flex:none;width:66.6666%}.column.is-half-desktop{flex:none;width:50%}.column.is-one-third-desktop{flex:none;width:33.3333%}.column.is-one-quarter-desktop{flex:none;width:25%}.column.is-one-fifth-desktop{flex:none;width:20%}.column.is-two-fifths-desktop{flex:none;width:40%}.column.is-three-fifths-desktop{flex:none;width:60%}.column.is-four-fifths-desktop{flex:none;width:80%}.column.is-offset-three-quarters-desktop{margin-left:75%}.column.is-offset-two-thirds-desktop{margin-left:66.6666%}.column.is-offset-half-desktop{margin-left:50%}.column.is-offset-one-third-desktop{margin-left:33.3333%}.column.is-offset-one-quarter-desktop{margin-left:25%}.column.is-offset-one-fifth-desktop{margin-left:20%}.column.is-offset-two-fifths-desktop{margin-left:40%}.column.is-offset-three-fifths-desktop{margin-left:60%}.column.is-offset-four-fifths-desktop{margin-left:80%}.column.is-0-desktop{flex:none;width:0%}.column.is-offset-0-desktop{margin-left:0%}.column.is-1-desktop{flex:none;width:8.33333337%}.column.is-offset-1-desktop{margin-left:8.33333337%}.column.is-2-desktop{flex:none;width:16.66666674%}.column.is-offset-2-desktop{margin-left:16.66666674%}.column.is-3-desktop{flex:none;width:25%}.column.is-offset-3-desktop{margin-left:25%}.column.is-4-desktop{flex:none;width:33.33333337%}.column.is-offset-4-desktop{margin-left:33.33333337%}.column.is-5-desktop{flex:none;width:41.66666674%}.column.is-offset-5-desktop{margin-left:41.66666674%}.column.is-6-desktop{flex:none;width:50%}.column.is-offset-6-desktop{margin-left:50%}.column.is-7-desktop{flex:none;width:58.33333337%}.column.is-offset-7-desktop{margin-left:58.33333337%}.column.is-8-desktop{flex:none;width:66.66666674%}.column.is-offset-8-desktop{margin-left:66.66666674%}.column.is-9-desktop{flex:none;width:75%}.column.is-offset-9-desktop{margin-left:75%}.column.is-10-desktop{flex:none;width:83.33333337%}.column.is-offset-10-desktop{margin-left:83.33333337%}.column.is-11-desktop{flex:none;width:91.66666674%}.column.is-offset-11-desktop{margin-left:91.66666674%}.column.is-12-desktop{flex:none;width:100%}.column.is-offset-12-desktop{margin-left:100%}}@media screen and (min-width: 1216px){.column.is-narrow-widescreen{flex:none;width:unset}.column.is-full-widescreen{flex:none;width:100%}.column.is-three-quarters-widescreen{flex:none;width:75%}.column.is-two-thirds-widescreen{flex:none;width:66.6666%}.column.is-half-widescreen{flex:none;width:50%}.column.is-one-third-widescreen{flex:none;width:33.3333%}.column.is-one-quarter-widescreen{flex:none;width:25%}.column.is-one-fifth-widescreen{flex:none;width:20%}.column.is-two-fifths-widescreen{flex:none;width:40%}.column.is-three-fifths-widescreen{flex:none;width:60%}.column.is-four-fifths-widescreen{flex:none;width:80%}.column.is-offset-three-quarters-widescreen{margin-left:75%}.column.is-offset-two-thirds-widescreen{margin-left:66.6666%}.column.is-offset-half-widescreen{margin-left:50%}.column.is-offset-one-third-widescreen{margin-left:33.3333%}.column.is-offset-one-quarter-widescreen{margin-left:25%}.column.is-offset-one-fifth-widescreen{margin-left:20%}.column.is-offset-two-fifths-widescreen{margin-left:40%}.column.is-offset-three-fifths-widescreen{margin-left:60%}.column.is-offset-four-fifths-widescreen{margin-left:80%}.column.is-0-widescreen{flex:none;width:0%}.column.is-offset-0-widescreen{margin-left:0%}.column.is-1-widescreen{flex:none;width:8.33333337%}.column.is-offset-1-widescreen{margin-left:8.33333337%}.column.is-2-widescreen{flex:none;width:16.66666674%}.column.is-offset-2-widescreen{margin-left:16.66666674%}.column.is-3-widescreen{flex:none;width:25%}.column.is-offset-3-widescreen{margin-left:25%}.column.is-4-widescreen{flex:none;width:33.33333337%}.column.is-offset-4-widescreen{margin-left:33.33333337%}.column.is-5-widescreen{flex:none;width:41.66666674%}.column.is-offset-5-widescreen{margin-left:41.66666674%}.column.is-6-widescreen{flex:none;width:50%}.column.is-offset-6-widescreen{margin-left:50%}.column.is-7-widescreen{flex:none;width:58.33333337%}.column.is-offset-7-widescreen{margin-left:58.33333337%}.column.is-8-widescreen{flex:none;width:66.66666674%}.column.is-offset-8-widescreen{margin-left:66.66666674%}.column.is-9-widescreen{flex:none;width:75%}.column.is-offset-9-widescreen{margin-left:75%}.column.is-10-widescreen{flex:none;width:83.33333337%}.column.is-offset-10-widescreen{margin-left:83.33333337%}.column.is-11-widescreen{flex:none;width:91.66666674%}.column.is-offset-11-widescreen{margin-left:91.66666674%}.column.is-12-widescreen{flex:none;width:100%}.column.is-offset-12-widescreen{margin-left:100%}}@media screen and (min-width: 1408px){.column.is-narrow-fullhd{flex:none;width:unset}.column.is-full-fullhd{flex:none;width:100%}.column.is-three-quarters-fullhd{flex:none;width:75%}.column.is-two-thirds-fullhd{flex:none;width:66.6666%}.column.is-half-fullhd{flex:none;width:50%}.column.is-one-third-fullhd{flex:none;width:33.3333%}.column.is-one-quarter-fullhd{flex:none;width:25%}.column.is-one-fifth-fullhd{flex:none;width:20%}.column.is-two-fifths-fullhd{flex:none;width:40%}.column.is-three-fifths-fullhd{flex:none;width:60%}.column.is-four-fifths-fullhd{flex:none;width:80%}.column.is-offset-three-quarters-fullhd{margin-left:75%}.column.is-offset-two-thirds-fullhd{margin-left:66.6666%}.column.is-offset-half-fullhd{margin-left:50%}.column.is-offset-one-third-fullhd{margin-left:33.3333%}.column.is-offset-one-quarter-fullhd{margin-left:25%}.column.is-offset-one-fifth-fullhd{margin-left:20%}.column.is-offset-two-fifths-fullhd{margin-left:40%}.column.is-offset-three-fifths-fullhd{margin-left:60%}.column.is-offset-four-fifths-fullhd{margin-left:80%}.column.is-0-fullhd{flex:none;width:0%}.column.is-offset-0-fullhd{margin-left:0%}.column.is-1-fullhd{flex:none;width:8.33333337%}.column.is-offset-1-fullhd{margin-left:8.33333337%}.column.is-2-fullhd{flex:none;width:16.66666674%}.column.is-offset-2-fullhd{margin-left:16.66666674%}.column.is-3-fullhd{flex:none;width:25%}.column.is-offset-3-fullhd{margin-left:25%}.column.is-4-fullhd{flex:none;width:33.33333337%}.column.is-offset-4-fullhd{margin-left:33.33333337%}.column.is-5-fullhd{flex:none;width:41.66666674%}.column.is-offset-5-fullhd{margin-left:41.66666674%}.column.is-6-fullhd{flex:none;width:50%}.column.is-offset-6-fullhd{margin-left:50%}.column.is-7-fullhd{flex:none;width:58.33333337%}.column.is-offset-7-fullhd{margin-left:58.33333337%}.column.is-8-fullhd{flex:none;width:66.66666674%}.column.is-offset-8-fullhd{margin-left:66.66666674%}.column.is-9-fullhd{flex:none;width:75%}.column.is-offset-9-fullhd{margin-left:75%}.column.is-10-fullhd{flex:none;width:83.33333337%}.column.is-offset-10-fullhd{margin-left:83.33333337%}.column.is-11-fullhd{flex:none;width:91.66666674%}.column.is-offset-11-fullhd{margin-left:91.66666674%}.column.is-12-fullhd{flex:none;width:100%}.column.is-offset-12-fullhd{margin-left:100%}}.columns{margin-left:-.75rem;margin-right:-.75rem;margin-top:-.75rem}.columns:last-child{margin-bottom:-.75rem}.columns:not(:last-child){margin-bottom:calc(1.5rem - .75rem)}.columns.is-centered{justify-content:center}.columns.is-gapless{margin-left:0;margin-right:0;margin-top:0}.columns.is-gapless>.column{margin:0;padding:0 !important}.columns.is-gapless:not(:last-child){margin-bottom:1.5rem}.columns.is-gapless:last-child{margin-bottom:0}.columns.is-mobile{display:flex}.columns.is-multiline{flex-wrap:wrap}.columns.is-vcentered{align-items:center}@media screen and (min-width: 769px),print{.columns:not(.is-desktop){display:flex}}@media screen and (min-width: 1056px){.columns.is-desktop{display:flex}}.columns.is-variable{--columnGap: 0.75rem;margin-left:calc(-1 * var(--columnGap));margin-right:calc(-1 * var(--columnGap))}.columns.is-variable>.column{padding-left:var(--columnGap);padding-right:var(--columnGap)}.columns.is-variable.is-0{--columnGap: 0rem}@media screen and (max-width: 768px){.columns.is-variable.is-0-mobile{--columnGap: 0rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-0-tablet{--columnGap: 0rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-0-tablet-only{--columnGap: 0rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-0-touch{--columnGap: 0rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-0-desktop{--columnGap: 0rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-0-desktop-only{--columnGap: 0rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-0-widescreen{--columnGap: 0rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-0-widescreen-only{--columnGap: 0rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-0-fullhd{--columnGap: 0rem}}.columns.is-variable.is-1{--columnGap: .25rem}@media screen and (max-width: 768px){.columns.is-variable.is-1-mobile{--columnGap: .25rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-1-tablet{--columnGap: .25rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-1-tablet-only{--columnGap: .25rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-1-touch{--columnGap: .25rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-1-desktop{--columnGap: .25rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-1-desktop-only{--columnGap: .25rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-1-widescreen{--columnGap: .25rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-1-widescreen-only{--columnGap: .25rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-1-fullhd{--columnGap: .25rem}}.columns.is-variable.is-2{--columnGap: .5rem}@media screen and (max-width: 768px){.columns.is-variable.is-2-mobile{--columnGap: .5rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-2-tablet{--columnGap: .5rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-2-tablet-only{--columnGap: .5rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-2-touch{--columnGap: .5rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-2-desktop{--columnGap: .5rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-2-desktop-only{--columnGap: .5rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-2-widescreen{--columnGap: .5rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-2-widescreen-only{--columnGap: .5rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-2-fullhd{--columnGap: .5rem}}.columns.is-variable.is-3{--columnGap: .75rem}@media screen and (max-width: 768px){.columns.is-variable.is-3-mobile{--columnGap: .75rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-3-tablet{--columnGap: .75rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-3-tablet-only{--columnGap: .75rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-3-touch{--columnGap: .75rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-3-desktop{--columnGap: .75rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-3-desktop-only{--columnGap: .75rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-3-widescreen{--columnGap: .75rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-3-widescreen-only{--columnGap: .75rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-3-fullhd{--columnGap: .75rem}}.columns.is-variable.is-4{--columnGap: 1rem}@media screen and (max-width: 768px){.columns.is-variable.is-4-mobile{--columnGap: 1rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-4-tablet{--columnGap: 1rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-4-tablet-only{--columnGap: 1rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-4-touch{--columnGap: 1rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-4-desktop{--columnGap: 1rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-4-desktop-only{--columnGap: 1rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-4-widescreen{--columnGap: 1rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-4-widescreen-only{--columnGap: 1rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-4-fullhd{--columnGap: 1rem}}.columns.is-variable.is-5{--columnGap: 1.25rem}@media screen and (max-width: 768px){.columns.is-variable.is-5-mobile{--columnGap: 1.25rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-5-tablet{--columnGap: 1.25rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-5-tablet-only{--columnGap: 1.25rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-5-touch{--columnGap: 1.25rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-5-desktop{--columnGap: 1.25rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-5-desktop-only{--columnGap: 1.25rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-5-widescreen{--columnGap: 1.25rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-5-widescreen-only{--columnGap: 1.25rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-5-fullhd{--columnGap: 1.25rem}}.columns.is-variable.is-6{--columnGap: 1.5rem}@media screen and (max-width: 768px){.columns.is-variable.is-6-mobile{--columnGap: 1.5rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-6-tablet{--columnGap: 1.5rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-6-tablet-only{--columnGap: 1.5rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-6-touch{--columnGap: 1.5rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-6-desktop{--columnGap: 1.5rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-6-desktop-only{--columnGap: 1.5rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-6-widescreen{--columnGap: 1.5rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-6-widescreen-only{--columnGap: 1.5rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-6-fullhd{--columnGap: 1.5rem}}.columns.is-variable.is-7{--columnGap: 1.75rem}@media screen and (max-width: 768px){.columns.is-variable.is-7-mobile{--columnGap: 1.75rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-7-tablet{--columnGap: 1.75rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-7-tablet-only{--columnGap: 1.75rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-7-touch{--columnGap: 1.75rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-7-desktop{--columnGap: 1.75rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-7-desktop-only{--columnGap: 1.75rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-7-widescreen{--columnGap: 1.75rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-7-widescreen-only{--columnGap: 1.75rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-7-fullhd{--columnGap: 1.75rem}}.columns.is-variable.is-8{--columnGap: 2rem}@media screen and (max-width: 768px){.columns.is-variable.is-8-mobile{--columnGap: 2rem}}@media screen and (min-width: 769px),print{.columns.is-variable.is-8-tablet{--columnGap: 2rem}}@media screen and (min-width: 769px) and (max-width: 1055px){.columns.is-variable.is-8-tablet-only{--columnGap: 2rem}}@media screen and (max-width: 1055px){.columns.is-variable.is-8-touch{--columnGap: 2rem}}@media screen and (min-width: 1056px){.columns.is-variable.is-8-desktop{--columnGap: 2rem}}@media screen and (min-width: 1056px) and (max-width: 1215px){.columns.is-variable.is-8-desktop-only{--columnGap: 2rem}}@media screen and (min-width: 1216px){.columns.is-variable.is-8-widescreen{--columnGap: 2rem}}@media screen and (min-width: 1216px) and (max-width: 1407px){.columns.is-variable.is-8-widescreen-only{--columnGap: 2rem}}@media screen and (min-width: 1408px){.columns.is-variable.is-8-fullhd{--columnGap: 2rem}}.tile{align-items:stretch;display:block;flex-basis:0;flex-grow:1;flex-shrink:1;min-height:min-content}.tile.is-ancestor{margin-left:-.75rem;margin-right:-.75rem;margin-top:-.75rem}.tile.is-ancestor:last-child{margin-bottom:-.75rem}.tile.is-ancestor:not(:last-child){margin-bottom:.75rem}.tile.is-child{margin:0 !important}.tile.is-parent{padding:.75rem}.tile.is-vertical{flex-direction:column}.tile.is-vertical>.tile.is-child:not(:last-child){margin-bottom:1.5rem !important}@media screen and (min-width: 769px),print{.tile:not(.is-child){display:flex}.tile.is-1{flex:none;width:8.33333337%}.tile.is-2{flex:none;width:16.66666674%}.tile.is-3{flex:none;width:25%}.tile.is-4{flex:none;width:33.33333337%}.tile.is-5{flex:none;width:41.66666674%}.tile.is-6{flex:none;width:50%}.tile.is-7{flex:none;width:58.33333337%}.tile.is-8{flex:none;width:66.66666674%}.tile.is-9{flex:none;width:75%}.tile.is-10{flex:none;width:83.33333337%}.tile.is-11{flex:none;width:91.66666674%}.tile.is-12{flex:none;width:100%}}.hero{align-items:stretch;display:flex;flex-direction:column;justify-content:space-between}.hero .navbar{background:none}.hero .tabs ul{border-bottom:none}.hero.is-white{background-color:#fff;color:#0a0a0a}.hero.is-white a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-white strong{color:inherit}.hero.is-white .title{color:#0a0a0a}.hero.is-white .subtitle{color:rgba(10,10,10,0.9)}.hero.is-white .subtitle a:not(.button),.hero.is-white .subtitle strong{color:#0a0a0a}@media screen and (max-width: 1055px){.hero.is-white .navbar-menu{background-color:#fff}}.hero.is-white .navbar-item,.hero.is-white .navbar-link{color:rgba(10,10,10,0.7)}.hero.is-white a.navbar-item:hover,.hero.is-white a.navbar-item.is-active,.hero.is-white .navbar-link:hover,.hero.is-white .navbar-link.is-active{background-color:#f2f2f2;color:#0a0a0a}.hero.is-white .tabs a{color:#0a0a0a;opacity:0.9}.hero.is-white .tabs a:hover{opacity:1}.hero.is-white .tabs li.is-active a{color:#fff !important;opacity:1}.hero.is-white .tabs.is-boxed a,.hero.is-white .tabs.is-toggle a{color:#0a0a0a}.hero.is-white .tabs.is-boxed a:hover,.hero.is-white .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-white .tabs.is-boxed li.is-active a,.hero.is-white .tabs.is-boxed li.is-active a:hover,.hero.is-white .tabs.is-toggle li.is-active a,.hero.is-white .tabs.is-toggle li.is-active a:hover{background-color:#0a0a0a;border-color:#0a0a0a;color:#fff}.hero.is-white.is-bold{background-image:linear-gradient(141deg, #e8e3e4 0%, #fff 71%, #fff 100%)}@media screen and (max-width: 768px){.hero.is-white.is-bold .navbar-menu{background-image:linear-gradient(141deg, #e8e3e4 0%, #fff 71%, #fff 100%)}}.hero.is-black{background-color:#0a0a0a;color:#fff}.hero.is-black a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-black strong{color:inherit}.hero.is-black .title{color:#fff}.hero.is-black .subtitle{color:rgba(255,255,255,0.9)}.hero.is-black .subtitle a:not(.button),.hero.is-black .subtitle strong{color:#fff}@media screen and (max-width: 1055px){.hero.is-black .navbar-menu{background-color:#0a0a0a}}.hero.is-black .navbar-item,.hero.is-black .navbar-link{color:rgba(255,255,255,0.7)}.hero.is-black a.navbar-item:hover,.hero.is-black a.navbar-item.is-active,.hero.is-black .navbar-link:hover,.hero.is-black .navbar-link.is-active{background-color:#000;color:#fff}.hero.is-black .tabs a{color:#fff;opacity:0.9}.hero.is-black .tabs a:hover{opacity:1}.hero.is-black .tabs li.is-active a{color:#0a0a0a !important;opacity:1}.hero.is-black .tabs.is-boxed a,.hero.is-black .tabs.is-toggle a{color:#fff}.hero.is-black .tabs.is-boxed a:hover,.hero.is-black .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-black .tabs.is-boxed li.is-active a,.hero.is-black .tabs.is-boxed li.is-active a:hover,.hero.is-black .tabs.is-toggle li.is-active a,.hero.is-black .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#0a0a0a}.hero.is-black.is-bold{background-image:linear-gradient(141deg, #000 0%, #0a0a0a 71%, #181616 100%)}@media screen and (max-width: 768px){.hero.is-black.is-bold .navbar-menu{background-image:linear-gradient(141deg, #000 0%, #0a0a0a 71%, #181616 100%)}}.hero.is-light{background-color:#f5f5f5;color:rgba(0,0,0,0.7)}.hero.is-light a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-light strong{color:inherit}.hero.is-light .title{color:rgba(0,0,0,0.7)}.hero.is-light .subtitle{color:rgba(0,0,0,0.9)}.hero.is-light .subtitle a:not(.button),.hero.is-light .subtitle strong{color:rgba(0,0,0,0.7)}@media screen and (max-width: 1055px){.hero.is-light .navbar-menu{background-color:#f5f5f5}}.hero.is-light .navbar-item,.hero.is-light .navbar-link{color:rgba(0,0,0,0.7)}.hero.is-light a.navbar-item:hover,.hero.is-light a.navbar-item.is-active,.hero.is-light .navbar-link:hover,.hero.is-light .navbar-link.is-active{background-color:#e8e8e8;color:rgba(0,0,0,0.7)}.hero.is-light .tabs a{color:rgba(0,0,0,0.7);opacity:0.9}.hero.is-light .tabs a:hover{opacity:1}.hero.is-light .tabs li.is-active a{color:#f5f5f5 !important;opacity:1}.hero.is-light .tabs.is-boxed a,.hero.is-light .tabs.is-toggle a{color:rgba(0,0,0,0.7)}.hero.is-light .tabs.is-boxed a:hover,.hero.is-light .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-light .tabs.is-boxed li.is-active a,.hero.is-light .tabs.is-boxed li.is-active a:hover,.hero.is-light .tabs.is-toggle li.is-active a,.hero.is-light .tabs.is-toggle li.is-active a:hover{background-color:rgba(0,0,0,0.7);border-color:rgba(0,0,0,0.7);color:#f5f5f5}.hero.is-light.is-bold{background-image:linear-gradient(141deg, #dfd8d9 0%, #f5f5f5 71%, #fff 100%)}@media screen and (max-width: 768px){.hero.is-light.is-bold .navbar-menu{background-image:linear-gradient(141deg, #dfd8d9 0%, #f5f5f5 71%, #fff 100%)}}.hero.is-dark,.content kbd.hero{background-color:#363636;color:#fff}.hero.is-dark a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.content kbd.hero a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-dark strong,.content kbd.hero strong{color:inherit}.hero.is-dark .title,.content kbd.hero .title{color:#fff}.hero.is-dark .subtitle,.content kbd.hero .subtitle{color:rgba(255,255,255,0.9)}.hero.is-dark .subtitle a:not(.button),.content kbd.hero .subtitle a:not(.button),.hero.is-dark .subtitle strong,.content kbd.hero .subtitle strong{color:#fff}@media screen and (max-width: 1055px){.hero.is-dark .navbar-menu,.content kbd.hero .navbar-menu{background-color:#363636}}.hero.is-dark .navbar-item,.content kbd.hero .navbar-item,.hero.is-dark .navbar-link,.content kbd.hero .navbar-link{color:rgba(255,255,255,0.7)}.hero.is-dark a.navbar-item:hover,.content kbd.hero a.navbar-item:hover,.hero.is-dark a.navbar-item.is-active,.content kbd.hero a.navbar-item.is-active,.hero.is-dark .navbar-link:hover,.content kbd.hero .navbar-link:hover,.hero.is-dark .navbar-link.is-active,.content kbd.hero .navbar-link.is-active{background-color:#292929;color:#fff}.hero.is-dark .tabs a,.content kbd.hero .tabs a{color:#fff;opacity:0.9}.hero.is-dark .tabs a:hover,.content kbd.hero .tabs a:hover{opacity:1}.hero.is-dark .tabs li.is-active a,.content kbd.hero .tabs li.is-active a{color:#363636 !important;opacity:1}.hero.is-dark .tabs.is-boxed a,.content kbd.hero .tabs.is-boxed a,.hero.is-dark .tabs.is-toggle a,.content kbd.hero .tabs.is-toggle a{color:#fff}.hero.is-dark .tabs.is-boxed a:hover,.content kbd.hero .tabs.is-boxed a:hover,.hero.is-dark .tabs.is-toggle a:hover,.content kbd.hero .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-dark .tabs.is-boxed li.is-active a,.content kbd.hero .tabs.is-boxed li.is-active a,.hero.is-dark .tabs.is-boxed li.is-active a:hover,.hero.is-dark .tabs.is-toggle li.is-active a,.content kbd.hero .tabs.is-toggle li.is-active a,.hero.is-dark .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#363636}.hero.is-dark.is-bold,.content kbd.hero.is-bold{background-image:linear-gradient(141deg, #1f191a 0%, #363636 71%, #46403f 100%)}@media screen and (max-width: 768px){.hero.is-dark.is-bold .navbar-menu,.content kbd.hero.is-bold .navbar-menu{background-image:linear-gradient(141deg, #1f191a 0%, #363636 71%, #46403f 100%)}}.hero.is-primary,.docstring>section>a.hero.docs-sourcelink{background-color:#4eb5de;color:#fff}.hero.is-primary a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.docstring>section>a.hero.docs-sourcelink a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-primary strong,.docstring>section>a.hero.docs-sourcelink strong{color:inherit}.hero.is-primary .title,.docstring>section>a.hero.docs-sourcelink .title{color:#fff}.hero.is-primary .subtitle,.docstring>section>a.hero.docs-sourcelink .subtitle{color:rgba(255,255,255,0.9)}.hero.is-primary .subtitle a:not(.button),.docstring>section>a.hero.docs-sourcelink .subtitle a:not(.button),.hero.is-primary .subtitle strong,.docstring>section>a.hero.docs-sourcelink .subtitle strong{color:#fff}@media screen and (max-width: 1055px){.hero.is-primary .navbar-menu,.docstring>section>a.hero.docs-sourcelink .navbar-menu{background-color:#4eb5de}}.hero.is-primary .navbar-item,.docstring>section>a.hero.docs-sourcelink .navbar-item,.hero.is-primary .navbar-link,.docstring>section>a.hero.docs-sourcelink .navbar-link{color:rgba(255,255,255,0.7)}.hero.is-primary a.navbar-item:hover,.docstring>section>a.hero.docs-sourcelink a.navbar-item:hover,.hero.is-primary a.navbar-item.is-active,.docstring>section>a.hero.docs-sourcelink a.navbar-item.is-active,.hero.is-primary .navbar-link:hover,.docstring>section>a.hero.docs-sourcelink .navbar-link:hover,.hero.is-primary .navbar-link.is-active,.docstring>section>a.hero.docs-sourcelink .navbar-link.is-active{background-color:#39acda;color:#fff}.hero.is-primary .tabs a,.docstring>section>a.hero.docs-sourcelink .tabs a{color:#fff;opacity:0.9}.hero.is-primary .tabs a:hover,.docstring>section>a.hero.docs-sourcelink .tabs a:hover{opacity:1}.hero.is-primary .tabs li.is-active a,.docstring>section>a.hero.docs-sourcelink .tabs li.is-active a{color:#4eb5de !important;opacity:1}.hero.is-primary .tabs.is-boxed a,.docstring>section>a.hero.docs-sourcelink .tabs.is-boxed a,.hero.is-primary .tabs.is-toggle a,.docstring>section>a.hero.docs-sourcelink .tabs.is-toggle a{color:#fff}.hero.is-primary .tabs.is-boxed a:hover,.docstring>section>a.hero.docs-sourcelink .tabs.is-boxed a:hover,.hero.is-primary .tabs.is-toggle a:hover,.docstring>section>a.hero.docs-sourcelink .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-primary .tabs.is-boxed li.is-active a,.docstring>section>a.hero.docs-sourcelink .tabs.is-boxed li.is-active a,.hero.is-primary .tabs.is-boxed li.is-active a:hover,.hero.is-primary .tabs.is-toggle li.is-active a,.docstring>section>a.hero.docs-sourcelink .tabs.is-toggle li.is-active a,.hero.is-primary .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#4eb5de}.hero.is-primary.is-bold,.docstring>section>a.hero.is-bold.docs-sourcelink{background-image:linear-gradient(141deg, #1bc7de 0%, #4eb5de 71%, #5fa9e7 100%)}@media screen and (max-width: 768px){.hero.is-primary.is-bold .navbar-menu,.docstring>section>a.hero.is-bold.docs-sourcelink .navbar-menu{background-image:linear-gradient(141deg, #1bc7de 0%, #4eb5de 71%, #5fa9e7 100%)}}.hero.is-link{background-color:#2e63b8;color:#fff}.hero.is-link a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-link strong{color:inherit}.hero.is-link .title{color:#fff}.hero.is-link .subtitle{color:rgba(255,255,255,0.9)}.hero.is-link .subtitle a:not(.button),.hero.is-link .subtitle strong{color:#fff}@media screen and (max-width: 1055px){.hero.is-link .navbar-menu{background-color:#2e63b8}}.hero.is-link .navbar-item,.hero.is-link .navbar-link{color:rgba(255,255,255,0.7)}.hero.is-link a.navbar-item:hover,.hero.is-link a.navbar-item.is-active,.hero.is-link .navbar-link:hover,.hero.is-link .navbar-link.is-active{background-color:#2958a4;color:#fff}.hero.is-link .tabs a{color:#fff;opacity:0.9}.hero.is-link .tabs a:hover{opacity:1}.hero.is-link .tabs li.is-active a{color:#2e63b8 !important;opacity:1}.hero.is-link .tabs.is-boxed a,.hero.is-link .tabs.is-toggle a{color:#fff}.hero.is-link .tabs.is-boxed a:hover,.hero.is-link .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-link .tabs.is-boxed li.is-active a,.hero.is-link .tabs.is-boxed li.is-active a:hover,.hero.is-link .tabs.is-toggle li.is-active a,.hero.is-link .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#2e63b8}.hero.is-link.is-bold{background-image:linear-gradient(141deg, #1b6098 0%, #2e63b8 71%, #2d51d2 100%)}@media screen and (max-width: 768px){.hero.is-link.is-bold .navbar-menu{background-image:linear-gradient(141deg, #1b6098 0%, #2e63b8 71%, #2d51d2 100%)}}.hero.is-info{background-color:#209cee;color:#fff}.hero.is-info a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-info strong{color:inherit}.hero.is-info .title{color:#fff}.hero.is-info .subtitle{color:rgba(255,255,255,0.9)}.hero.is-info .subtitle a:not(.button),.hero.is-info .subtitle strong{color:#fff}@media screen and (max-width: 1055px){.hero.is-info .navbar-menu{background-color:#209cee}}.hero.is-info .navbar-item,.hero.is-info .navbar-link{color:rgba(255,255,255,0.7)}.hero.is-info a.navbar-item:hover,.hero.is-info a.navbar-item.is-active,.hero.is-info .navbar-link:hover,.hero.is-info .navbar-link.is-active{background-color:#1190e3;color:#fff}.hero.is-info .tabs a{color:#fff;opacity:0.9}.hero.is-info .tabs a:hover{opacity:1}.hero.is-info .tabs li.is-active a{color:#209cee !important;opacity:1}.hero.is-info .tabs.is-boxed a,.hero.is-info .tabs.is-toggle a{color:#fff}.hero.is-info .tabs.is-boxed a:hover,.hero.is-info .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-info .tabs.is-boxed li.is-active a,.hero.is-info .tabs.is-boxed li.is-active a:hover,.hero.is-info .tabs.is-toggle li.is-active a,.hero.is-info .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#209cee}.hero.is-info.is-bold{background-image:linear-gradient(141deg, #05a6d6 0%, #209cee 71%, #3287f5 100%)}@media screen and (max-width: 768px){.hero.is-info.is-bold .navbar-menu{background-image:linear-gradient(141deg, #05a6d6 0%, #209cee 71%, #3287f5 100%)}}.hero.is-success{background-color:#22c35b;color:#fff}.hero.is-success a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-success strong{color:inherit}.hero.is-success .title{color:#fff}.hero.is-success .subtitle{color:rgba(255,255,255,0.9)}.hero.is-success .subtitle a:not(.button),.hero.is-success .subtitle strong{color:#fff}@media screen and (max-width: 1055px){.hero.is-success .navbar-menu{background-color:#22c35b}}.hero.is-success .navbar-item,.hero.is-success .navbar-link{color:rgba(255,255,255,0.7)}.hero.is-success a.navbar-item:hover,.hero.is-success a.navbar-item.is-active,.hero.is-success .navbar-link:hover,.hero.is-success .navbar-link.is-active{background-color:#1ead51;color:#fff}.hero.is-success .tabs a{color:#fff;opacity:0.9}.hero.is-success .tabs a:hover{opacity:1}.hero.is-success .tabs li.is-active a{color:#22c35b !important;opacity:1}.hero.is-success .tabs.is-boxed a,.hero.is-success .tabs.is-toggle a{color:#fff}.hero.is-success .tabs.is-boxed a:hover,.hero.is-success .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-success .tabs.is-boxed li.is-active a,.hero.is-success .tabs.is-boxed li.is-active a:hover,.hero.is-success .tabs.is-toggle li.is-active a,.hero.is-success .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#22c35b}.hero.is-success.is-bold{background-image:linear-gradient(141deg, #12a02c 0%, #22c35b 71%, #1fdf83 100%)}@media screen and (max-width: 768px){.hero.is-success.is-bold .navbar-menu{background-image:linear-gradient(141deg, #12a02c 0%, #22c35b 71%, #1fdf83 100%)}}.hero.is-warning{background-color:#ffdd57;color:rgba(0,0,0,0.7)}.hero.is-warning a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-warning strong{color:inherit}.hero.is-warning .title{color:rgba(0,0,0,0.7)}.hero.is-warning .subtitle{color:rgba(0,0,0,0.9)}.hero.is-warning .subtitle a:not(.button),.hero.is-warning .subtitle strong{color:rgba(0,0,0,0.7)}@media screen and (max-width: 1055px){.hero.is-warning .navbar-menu{background-color:#ffdd57}}.hero.is-warning .navbar-item,.hero.is-warning .navbar-link{color:rgba(0,0,0,0.7)}.hero.is-warning a.navbar-item:hover,.hero.is-warning a.navbar-item.is-active,.hero.is-warning .navbar-link:hover,.hero.is-warning .navbar-link.is-active{background-color:#ffd83e;color:rgba(0,0,0,0.7)}.hero.is-warning .tabs a{color:rgba(0,0,0,0.7);opacity:0.9}.hero.is-warning .tabs a:hover{opacity:1}.hero.is-warning .tabs li.is-active a{color:#ffdd57 !important;opacity:1}.hero.is-warning .tabs.is-boxed a,.hero.is-warning .tabs.is-toggle a{color:rgba(0,0,0,0.7)}.hero.is-warning .tabs.is-boxed a:hover,.hero.is-warning .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-warning .tabs.is-boxed li.is-active a,.hero.is-warning .tabs.is-boxed li.is-active a:hover,.hero.is-warning .tabs.is-toggle li.is-active a,.hero.is-warning .tabs.is-toggle li.is-active a:hover{background-color:rgba(0,0,0,0.7);border-color:rgba(0,0,0,0.7);color:#ffdd57}.hero.is-warning.is-bold{background-image:linear-gradient(141deg, #ffae24 0%, #ffdd57 71%, #fffa71 100%)}@media screen and (max-width: 768px){.hero.is-warning.is-bold .navbar-menu{background-image:linear-gradient(141deg, #ffae24 0%, #ffdd57 71%, #fffa71 100%)}}.hero.is-danger{background-color:#da0b00;color:#fff}.hero.is-danger a:not(.button):not(.dropdown-item):not(.tag):not(.pagination-link.is-current),.hero.is-danger strong{color:inherit}.hero.is-danger .title{color:#fff}.hero.is-danger .subtitle{color:rgba(255,255,255,0.9)}.hero.is-danger .subtitle a:not(.button),.hero.is-danger .subtitle strong{color:#fff}@media screen and (max-width: 1055px){.hero.is-danger .navbar-menu{background-color:#da0b00}}.hero.is-danger .navbar-item,.hero.is-danger .navbar-link{color:rgba(255,255,255,0.7)}.hero.is-danger a.navbar-item:hover,.hero.is-danger a.navbar-item.is-active,.hero.is-danger .navbar-link:hover,.hero.is-danger .navbar-link.is-active{background-color:#c10a00;color:#fff}.hero.is-danger .tabs a{color:#fff;opacity:0.9}.hero.is-danger .tabs a:hover{opacity:1}.hero.is-danger .tabs li.is-active a{color:#da0b00 !important;opacity:1}.hero.is-danger .tabs.is-boxed a,.hero.is-danger .tabs.is-toggle a{color:#fff}.hero.is-danger .tabs.is-boxed a:hover,.hero.is-danger .tabs.is-toggle a:hover{background-color:rgba(10,10,10,0.1)}.hero.is-danger .tabs.is-boxed li.is-active a,.hero.is-danger .tabs.is-boxed li.is-active a:hover,.hero.is-danger .tabs.is-toggle li.is-active a,.hero.is-danger .tabs.is-toggle li.is-active a:hover{background-color:#fff;border-color:#fff;color:#da0b00}.hero.is-danger.is-bold{background-image:linear-gradient(141deg, #a70013 0%, #da0b00 71%, #f43500 100%)}@media screen and (max-width: 768px){.hero.is-danger.is-bold .navbar-menu{background-image:linear-gradient(141deg, #a70013 0%, #da0b00 71%, #f43500 100%)}}.hero.is-small .hero-body,#documenter .docs-sidebar form.docs-search>input.hero .hero-body{padding:1.5rem}@media screen and (min-width: 769px),print{.hero.is-medium .hero-body{padding:9rem 4.5rem}}@media screen and (min-width: 769px),print{.hero.is-large .hero-body{padding:18rem 6rem}}.hero.is-halfheight .hero-body,.hero.is-fullheight .hero-body,.hero.is-fullheight-with-navbar .hero-body{align-items:center;display:flex}.hero.is-halfheight .hero-body>.container,.hero.is-fullheight .hero-body>.container,.hero.is-fullheight-with-navbar .hero-body>.container{flex-grow:1;flex-shrink:1}.hero.is-halfheight{min-height:50vh}.hero.is-fullheight{min-height:100vh}.hero-video{overflow:hidden}.hero-video video{left:50%;min-height:100%;min-width:100%;position:absolute;top:50%;transform:translate3d(-50%, -50%, 0)}.hero-video.is-transparent{opacity:0.3}@media screen and (max-width: 768px){.hero-video{display:none}}.hero-buttons{margin-top:1.5rem}@media screen and (max-width: 768px){.hero-buttons .button{display:flex}.hero-buttons .button:not(:last-child){margin-bottom:0.75rem}}@media screen and (min-width: 769px),print{.hero-buttons{display:flex;justify-content:center}.hero-buttons .button:not(:last-child){margin-right:1.5rem}}.hero-head,.hero-foot{flex-grow:0;flex-shrink:0}.hero-body{flex-grow:1;flex-shrink:0;padding:3rem 1.5rem}@media screen and (min-width: 769px),print{.hero-body{padding:3rem 3rem}}.section{padding:3rem 1.5rem}@media screen and (min-width: 1056px){.section{padding:3rem 3rem}.section.is-medium{padding:9rem 4.5rem}.section.is-large{padding:18rem 6rem}}.footer{background-color:#fafafa;padding:3rem 1.5rem 6rem}h1 .docs-heading-anchor,h1 .docs-heading-anchor:hover,h1 .docs-heading-anchor:visited,h2 .docs-heading-anchor,h2 .docs-heading-anchor:hover,h2 .docs-heading-anchor:visited,h3 .docs-heading-anchor,h3 .docs-heading-anchor:hover,h3 .docs-heading-anchor:visited,h4 .docs-heading-anchor,h4 .docs-heading-anchor:hover,h4 .docs-heading-anchor:visited,h5 .docs-heading-anchor,h5 .docs-heading-anchor:hover,h5 .docs-heading-anchor:visited,h6 .docs-heading-anchor,h6 .docs-heading-anchor:hover,h6 .docs-heading-anchor:visited{color:#222}h1 .docs-heading-anchor-permalink,h2 .docs-heading-anchor-permalink,h3 .docs-heading-anchor-permalink,h4 .docs-heading-anchor-permalink,h5 .docs-heading-anchor-permalink,h6 .docs-heading-anchor-permalink{visibility:hidden;vertical-align:middle;margin-left:0.5em;font-size:0.7rem}h1 .docs-heading-anchor-permalink::before,h2 .docs-heading-anchor-permalink::before,h3 .docs-heading-anchor-permalink::before,h4 .docs-heading-anchor-permalink::before,h5 .docs-heading-anchor-permalink::before,h6 .docs-heading-anchor-permalink::before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f0c1"}h1:hover .docs-heading-anchor-permalink,h2:hover .docs-heading-anchor-permalink,h3:hover .docs-heading-anchor-permalink,h4:hover .docs-heading-anchor-permalink,h5:hover .docs-heading-anchor-permalink,h6:hover .docs-heading-anchor-permalink{visibility:visible}.docs-dark-only{display:none !important}pre{position:relative;overflow:hidden}pre code,pre code.hljs{padding:0 .75rem !important;overflow:auto;display:block}pre code:first-of-type,pre code.hljs:first-of-type{padding-top:0.5rem !important}pre code:last-of-type,pre code.hljs:last-of-type{padding-bottom:0.5rem !important}pre .copy-button{opacity:0.2;transition:opacity 0.2s;position:absolute;right:0em;top:0em;padding:0.5em;width:2.5em;height:2.5em;background:transparent;border:none;font-family:"Font Awesome 6 Free";color:#222;cursor:pointer;text-align:center}pre .copy-button:focus,pre .copy-button:hover{opacity:1;background:rgba(34,34,34,0.1);color:#2e63b8}pre .copy-button.success{color:#259a12;opacity:1}pre .copy-button.error{color:#cb3c33;opacity:1}pre:hover .copy-button{opacity:1}.admonition{background-color:#b5b5b5;border-style:solid;border-width:1px;border-color:#363636;border-radius:4px;font-size:1rem}.admonition strong{color:currentColor}.admonition.is-small,#documenter .docs-sidebar form.docs-search>input.admonition{font-size:.75rem}.admonition.is-medium{font-size:1.25rem}.admonition.is-large{font-size:1.5rem}.admonition.is-default{background-color:#b5b5b5;border-color:#363636}.admonition.is-default>.admonition-header{background-color:#363636;color:#fff}.admonition.is-default>.admonition-body{color:#fff}.admonition.is-info{background-color:#def0fc;border-color:#209cee}.admonition.is-info>.admonition-header{background-color:#209cee;color:#fff}.admonition.is-info>.admonition-body{color:rgba(0,0,0,0.7)}.admonition.is-success{background-color:#bdf4d1;border-color:#22c35b}.admonition.is-success>.admonition-header{background-color:#22c35b;color:#fff}.admonition.is-success>.admonition-body{color:rgba(0,0,0,0.7)}.admonition.is-warning{background-color:#fff3c5;border-color:#ffdd57}.admonition.is-warning>.admonition-header{background-color:#ffdd57;color:rgba(0,0,0,0.7)}.admonition.is-warning>.admonition-body{color:rgba(0,0,0,0.7)}.admonition.is-danger{background-color:#ffaba7;border-color:#da0b00}.admonition.is-danger>.admonition-header{background-color:#da0b00;color:#fff}.admonition.is-danger>.admonition-body{color:rgba(0,0,0,0.7)}.admonition.is-compat{background-color:#bdeff5;border-color:#1db5c9}.admonition.is-compat>.admonition-header{background-color:#1db5c9;color:#fff}.admonition.is-compat>.admonition-body{color:rgba(0,0,0,0.7)}.admonition-header{color:#fff;background-color:#363636;align-items:center;font-weight:700;justify-content:space-between;line-height:1.25;padding:0.5rem .75rem;position:relative}.admonition-header:before{font-family:"Font Awesome 6 Free";font-weight:900;margin-right:.75rem;content:"\f06a"}details.admonition.is-details>.admonition-header{list-style:none}details.admonition.is-details>.admonition-header:before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f055"}details.admonition.is-details[open]>.admonition-header:before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f056"}.admonition-body{color:#222;padding:0.5rem .75rem}.admonition-body pre{background-color:#f5f5f5}.admonition-body code{background-color:rgba(0,0,0,0.05)}.docstring{margin-bottom:1em;background-color:rgba(0,0,0,0);border:1px solid #dbdbdb;box-shadow:2px 2px 3px rgba(10,10,10,0.1);max-width:100%}.docstring>header{cursor:pointer;display:flex;flex-grow:1;align-items:stretch;padding:0.5rem .75rem;background-color:#f5f5f5;box-shadow:0 0.125em 0.25em rgba(10,10,10,0.1);box-shadow:none;border-bottom:1px solid #dbdbdb}.docstring>header code{background-color:transparent}.docstring>header .docstring-article-toggle-button{min-width:1.1rem;padding:0.2rem 0.2rem 0.2rem 0}.docstring>header .docstring-binding{margin-right:0.3em}.docstring>header .docstring-category{margin-left:0.3em}.docstring>section{position:relative;padding:.75rem .75rem;border-bottom:1px solid #dbdbdb}.docstring>section:last-child{border-bottom:none}.docstring>section>a.docs-sourcelink{transition:opacity 0.3s;opacity:0;position:absolute;right:.375rem;bottom:.375rem}.docstring>section>a.docs-sourcelink:focus{opacity:1 !important}.docstring:hover>section>a.docs-sourcelink{opacity:0.2}.docstring:focus-within>section>a.docs-sourcelink{opacity:0.2}.docstring>section:hover a.docs-sourcelink{opacity:1}.documenter-example-output{background-color:#fff}.outdated-warning-overlay{position:fixed;top:0;left:0;right:0;box-shadow:0 0 10px rgba(0,0,0,0.3);z-index:999;background-color:#ffaba7;color:rgba(0,0,0,0.7);border-bottom:3px solid #da0b00;padding:10px 35px;text-align:center;font-size:15px}.outdated-warning-overlay .outdated-warning-closer{position:absolute;top:calc(50% - 10px);right:18px;cursor:pointer;width:12px}.outdated-warning-overlay a{color:#2e63b8}.outdated-warning-overlay a:hover{color:#363636}.content pre{border:1px solid #dbdbdb}.content code{font-weight:inherit}.content a code{color:#2e63b8}.content h1 code,.content h2 code,.content h3 code,.content h4 code,.content h5 code,.content h6 code{color:#222}.content table{display:block;width:initial;max-width:100%;overflow-x:auto}.content blockquote>ul:first-child,.content blockquote>ol:first-child,.content .admonition-body>ul:first-child,.content .admonition-body>ol:first-child{margin-top:0}pre,code{font-variant-ligatures:no-contextual}.breadcrumb a.is-disabled{cursor:default;pointer-events:none}.breadcrumb a.is-disabled,.breadcrumb a.is-disabled:hover{color:#222}.hljs{background:initial !important}.katex .katex-mathml{top:0;right:0}.katex-display,mjx-container,.MathJax_Display{margin:0.5em 0 !important}html{-moz-osx-font-smoothing:auto;-webkit-font-smoothing:auto}li.no-marker{list-style:none}#documenter .docs-main>article{overflow-wrap:break-word}#documenter .docs-main>article .math-container{overflow-x:auto;overflow-y:hidden}@media screen and (min-width: 1056px){#documenter .docs-main{max-width:52rem;margin-left:20rem;padding-right:1rem}}@media screen and (max-width: 1055px){#documenter .docs-main{width:100%}#documenter .docs-main>article{max-width:52rem;margin-left:auto;margin-right:auto;margin-bottom:1rem;padding:0 1rem}#documenter .docs-main>header,#documenter .docs-main>nav{max-width:100%;width:100%;margin:0}}#documenter .docs-main header.docs-navbar{background-color:#fff;border-bottom:1px solid #dbdbdb;z-index:2;min-height:4rem;margin-bottom:1rem;display:flex}#documenter .docs-main header.docs-navbar .breadcrumb{flex-grow:1}#documenter .docs-main header.docs-navbar .docs-sidebar-button{display:block;font-size:1.5rem;padding-bottom:0.1rem;margin-right:1rem}#documenter .docs-main header.docs-navbar .docs-right{display:flex;white-space:nowrap;gap:1rem;align-items:center}#documenter .docs-main header.docs-navbar .docs-right .docs-icon,#documenter .docs-main header.docs-navbar .docs-right .docs-label{display:inline-block}#documenter .docs-main header.docs-navbar .docs-right .docs-label{padding:0;margin-left:0.3em}@media screen and (max-width: 1055px){#documenter .docs-main header.docs-navbar .docs-right .docs-navbar-link{margin-left:0.4rem;margin-right:0.4rem}}#documenter .docs-main header.docs-navbar>*{margin:auto 0}@media screen and (max-width: 1055px){#documenter .docs-main header.docs-navbar{position:sticky;top:0;padding:0 1rem;transition-property:top, box-shadow;-webkit-transition-property:top, box-shadow;transition-duration:0.3s;-webkit-transition-duration:0.3s}#documenter .docs-main header.docs-navbar.headroom--not-top{box-shadow:.2rem 0rem .4rem #bbb;transition-duration:0.7s;-webkit-transition-duration:0.7s}#documenter .docs-main header.docs-navbar.headroom--unpinned.headroom--not-top.headroom--not-bottom{top:-4.5rem;transition-duration:0.7s;-webkit-transition-duration:0.7s}}#documenter .docs-main section.footnotes{border-top:1px solid #dbdbdb}#documenter .docs-main section.footnotes li .tag:first-child,#documenter .docs-main section.footnotes li .docstring>section>a.docs-sourcelink:first-child,#documenter .docs-main section.footnotes li .content kbd:first-child,.content #documenter .docs-main section.footnotes li kbd:first-child{margin-right:1em;margin-bottom:0.4em}#documenter .docs-main .docs-footer{display:flex;flex-wrap:wrap;margin-left:0;margin-right:0;border-top:1px solid #dbdbdb;padding-top:1rem;padding-bottom:1rem}@media screen and (max-width: 1055px){#documenter .docs-main .docs-footer{padding-left:1rem;padding-right:1rem}}#documenter .docs-main .docs-footer .docs-footer-nextpage,#documenter .docs-main .docs-footer .docs-footer-prevpage{flex-grow:1}#documenter .docs-main .docs-footer .docs-footer-nextpage{text-align:right}#documenter .docs-main .docs-footer .flexbox-break{flex-basis:100%;height:0}#documenter .docs-main .docs-footer .footer-message{font-size:0.8em;margin:0.5em auto 0 auto;text-align:center}#documenter .docs-sidebar{display:flex;flex-direction:column;color:#0a0a0a;background-color:#f5f5f5;border-right:1px solid #dbdbdb;padding:0;flex:0 0 18rem;z-index:5;font-size:1rem;position:fixed;left:-18rem;width:18rem;height:100%;transition:left 0.3s}#documenter .docs-sidebar.visible{left:0;box-shadow:.4rem 0rem .8rem #bbb}@media screen and (min-width: 1056px){#documenter .docs-sidebar.visible{box-shadow:none}}@media screen and (min-width: 1056px){#documenter .docs-sidebar{left:0;top:0}}#documenter .docs-sidebar .docs-logo{margin-top:1rem;padding:0 1rem}#documenter .docs-sidebar .docs-logo>img{max-height:6rem;margin:auto}#documenter .docs-sidebar .docs-package-name{flex-shrink:0;font-size:1.5rem;font-weight:700;text-align:center;white-space:nowrap;overflow:hidden;padding:0.5rem 0}#documenter .docs-sidebar .docs-package-name .docs-autofit{max-width:16.2rem}#documenter .docs-sidebar .docs-package-name a,#documenter .docs-sidebar .docs-package-name a:hover{color:#0a0a0a}#documenter .docs-sidebar .docs-version-selector{border-top:1px solid #dbdbdb;display:none;padding:0.5rem}#documenter .docs-sidebar .docs-version-selector.visible{display:flex}#documenter .docs-sidebar ul.docs-menu{flex-grow:1;user-select:none;border-top:1px solid #dbdbdb;padding-bottom:1.5rem}#documenter .docs-sidebar ul.docs-menu>li>.tocitem{font-weight:bold}#documenter .docs-sidebar ul.docs-menu>li li{font-size:.95rem;margin-left:1em;border-left:1px solid #dbdbdb}#documenter .docs-sidebar ul.docs-menu input.collapse-toggle{display:none}#documenter .docs-sidebar ul.docs-menu ul.collapsed{display:none}#documenter .docs-sidebar ul.docs-menu input:checked~ul.collapsed{display:block}#documenter .docs-sidebar ul.docs-menu label.tocitem{display:flex}#documenter .docs-sidebar ul.docs-menu label.tocitem .docs-label{flex-grow:2}#documenter .docs-sidebar ul.docs-menu label.tocitem .docs-chevron{display:inline-block;font-style:normal;font-variant:normal;text-rendering:auto;line-height:1;font-size:.75rem;margin-left:1rem;margin-top:auto;margin-bottom:auto}#documenter .docs-sidebar ul.docs-menu label.tocitem .docs-chevron::before{font-family:"Font Awesome 6 Free";font-weight:900;content:"\f054"}#documenter .docs-sidebar ul.docs-menu input:checked~label.tocitem .docs-chevron::before{content:"\f078"}#documenter .docs-sidebar ul.docs-menu .tocitem{display:block;padding:0.5rem 0.5rem}#documenter .docs-sidebar ul.docs-menu .tocitem,#documenter .docs-sidebar ul.docs-menu .tocitem:hover{color:#0a0a0a;background:#f5f5f5}#documenter .docs-sidebar ul.docs-menu a.tocitem:hover,#documenter .docs-sidebar ul.docs-menu label.tocitem:hover{color:#0a0a0a;background-color:#ebebeb}#documenter .docs-sidebar ul.docs-menu li.is-active{border-top:1px solid #dbdbdb;border-bottom:1px solid #dbdbdb;background-color:#fff}#documenter .docs-sidebar ul.docs-menu li.is-active .tocitem,#documenter .docs-sidebar ul.docs-menu li.is-active .tocitem:hover{background-color:#fff;color:#0a0a0a}#documenter .docs-sidebar ul.docs-menu li.is-active ul.internal .tocitem:hover{background-color:#ebebeb;color:#0a0a0a}#documenter .docs-sidebar ul.docs-menu>li.is-active:first-child{border-top:none}#documenter .docs-sidebar ul.docs-menu ul.internal{margin:0 0.5rem 0.5rem;border-top:1px solid #dbdbdb}#documenter .docs-sidebar ul.docs-menu ul.internal li{font-size:.85rem;border-left:none;margin-left:0;margin-top:0.5rem}#documenter .docs-sidebar ul.docs-menu ul.internal .tocitem{width:100%;padding:0}#documenter .docs-sidebar ul.docs-menu ul.internal .tocitem::before{content:"⚬";margin-right:0.4em}#documenter .docs-sidebar form.docs-search{margin:auto;margin-top:0.5rem;margin-bottom:0.5rem}#documenter .docs-sidebar form.docs-search>input{width:14.4rem}#documenter .docs-sidebar #documenter-search-query{color:#707070;width:14.4rem;box-shadow:inset 0 1px 2px rgba(10,10,10,0.1)}@media screen and (min-width: 1056px){#documenter .docs-sidebar ul.docs-menu{overflow-y:auto;-webkit-overflow-scroll:touch}#documenter .docs-sidebar ul.docs-menu::-webkit-scrollbar{width:.3rem;background:none}#documenter .docs-sidebar ul.docs-menu::-webkit-scrollbar-thumb{border-radius:5px 0px 0px 5px;background:#e0e0e0}#documenter .docs-sidebar ul.docs-menu::-webkit-scrollbar-thumb:hover{background:#ccc}}@media screen and (max-width: 1055px){#documenter .docs-sidebar{overflow-y:auto;-webkit-overflow-scroll:touch}#documenter .docs-sidebar::-webkit-scrollbar{width:.3rem;background:none}#documenter .docs-sidebar::-webkit-scrollbar-thumb{border-radius:5px 0px 0px 5px;background:#e0e0e0}#documenter .docs-sidebar::-webkit-scrollbar-thumb:hover{background:#ccc}}kbd.search-modal-key-hints{border-radius:0.25rem;border:1px solid rgba(0,0,0,0.6);box-shadow:0 2px 0 1px rgba(0,0,0,0.6);cursor:default;font-size:0.9rem;line-height:1.5;min-width:0.75rem;text-align:center;padding:0.1rem 0.3rem;position:relative;top:-1px}.search-min-width-50{min-width:50%}.search-min-height-100{min-height:100%}.search-modal-card-body{max-height:calc(100vh - 15rem)}.search-result-link{border-radius:0.7em;transition:all 300ms}.search-result-link:hover,.search-result-link:focus{background-color:rgba(0,128,128,0.1)}.search-result-link .property-search-result-badge,.search-result-link .search-filter{transition:all 300ms}.property-search-result-badge,.search-filter{padding:0.15em 0.5em;font-size:0.8em;font-style:italic;text-transform:none !important;line-height:1.5;color:#f5f5f5;background-color:rgba(51,65,85,0.501961);border-radius:0.6rem}.search-result-link:hover .property-search-result-badge,.search-result-link:hover .search-filter,.search-result-link:focus .property-search-result-badge,.search-result-link:focus .search-filter{color:#f1f5f9;background-color:#333}.search-filter{color:#333;background-color:#f5f5f5;transition:all 300ms}.search-filter:hover,.search-filter:focus{color:#333}.search-filter-selected{color:#f5f5f5;background-color:rgba(139,0,139,0.5)}.search-filter-selected:hover,.search-filter-selected:focus{color:#f5f5f5}.search-result-highlight{background-color:#ffdd57;color:black}.search-divider{border-bottom:1px solid #dbdbdb}.search-result-title{width:85%;color:#333}.search-result-code-title{font-size:0.875rem;font-family:"JuliaMono","SFMono-Regular","Menlo","Consolas","Liberation Mono","DejaVu Sans Mono",monospace}#search-modal .modal-card-body::-webkit-scrollbar,#search-modal .filter-tabs::-webkit-scrollbar{height:10px;width:10px;background-color:transparent}#search-modal .modal-card-body::-webkit-scrollbar-thumb,#search-modal .filter-tabs::-webkit-scrollbar-thumb{background-color:gray;border-radius:1rem}#search-modal .modal-card-body::-webkit-scrollbar-track,#search-modal .filter-tabs::-webkit-scrollbar-track{-webkit-box-shadow:inset 0 0 6px rgba(0,0,0,0.6);background-color:transparent}.w-100{width:100%}.gap-2{gap:0.5rem}.gap-4{gap:1rem}.gap-8{gap:2rem}.ansi span.sgr1{font-weight:bolder}.ansi span.sgr2{font-weight:lighter}.ansi span.sgr3{font-style:italic}.ansi span.sgr4{text-decoration:underline}.ansi span.sgr7{color:#fff;background-color:#222}.ansi span.sgr8{color:transparent}.ansi span.sgr8 span{color:transparent}.ansi span.sgr9{text-decoration:line-through}.ansi span.sgr30{color:#242424}.ansi span.sgr31{color:#a7201f}.ansi span.sgr32{color:#066f00}.ansi span.sgr33{color:#856b00}.ansi span.sgr34{color:#2149b0}.ansi span.sgr35{color:#7d4498}.ansi span.sgr36{color:#007989}.ansi span.sgr37{color:gray}.ansi span.sgr40{background-color:#242424}.ansi span.sgr41{background-color:#a7201f}.ansi span.sgr42{background-color:#066f00}.ansi span.sgr43{background-color:#856b00}.ansi span.sgr44{background-color:#2149b0}.ansi span.sgr45{background-color:#7d4498}.ansi span.sgr46{background-color:#007989}.ansi span.sgr47{background-color:gray}.ansi span.sgr90{color:#616161}.ansi span.sgr91{color:#cb3c33}.ansi span.sgr92{color:#0e8300}.ansi span.sgr93{color:#a98800}.ansi span.sgr94{color:#3c5dcd}.ansi span.sgr95{color:#9256af}.ansi span.sgr96{color:#008fa3}.ansi span.sgr97{color:#f5f5f5}.ansi span.sgr100{background-color:#616161}.ansi span.sgr101{background-color:#cb3c33}.ansi span.sgr102{background-color:#0e8300}.ansi span.sgr103{background-color:#a98800}.ansi span.sgr104{background-color:#3c5dcd}.ansi span.sgr105{background-color:#9256af}.ansi span.sgr106{background-color:#008fa3}.ansi span.sgr107{background-color:#f5f5f5}code.language-julia-repl>span.hljs-meta{color:#066f00;font-weight:bolder}/*!
+  Theme: Default
+  Description: Original highlight.js style
+  Author: (c) Ivan Sagalaev <maniac@softwaremaniacs.org>
+  Maintainer: @highlightjs/core-team
+  Website: https://highlightjs.org/
+  License: see project LICENSE
+  Touched: 2021
+*/pre code.hljs{display:block;overflow-x:auto;padding:1em}code.hljs{padding:3px 5px}.hljs{background:#F3F3F3;color:#444}.hljs-comment{color:#697070}.hljs-tag,.hljs-punctuation{color:#444a}.hljs-tag .hljs-name,.hljs-tag .hljs-attr{color:#444}.hljs-keyword,.hljs-attribute,.hljs-selector-tag,.hljs-meta .hljs-keyword,.hljs-doctag,.hljs-name{font-weight:bold}.hljs-type,.hljs-string,.hljs-number,.hljs-selector-id,.hljs-selector-class,.hljs-quote,.hljs-template-tag,.hljs-deletion{color:#880000}.hljs-title,.hljs-section{color:#880000;font-weight:bold}.hljs-regexp,.hljs-symbol,.hljs-variable,.hljs-template-variable,.hljs-link,.hljs-selector-attr,.hljs-operator,.hljs-selector-pseudo{color:#ab5656}.hljs-literal{color:#695}.hljs-built_in,.hljs-bullet,.hljs-code,.hljs-addition{color:#397300}.hljs-meta{color:#1f7199}.hljs-meta .hljs-string{color:#38a}.hljs-emphasis{font-style:italic}.hljs-strong{font-weight:bold}
diff --git a/previews/PR3525/assets/themeswap.js b/previews/PR3525/assets/themeswap.js
new file mode 100644
index 00000000000..9f5eebe6aa2
--- /dev/null
+++ b/previews/PR3525/assets/themeswap.js
@@ -0,0 +1,84 @@
+// Small function to quickly swap out themes. Gets put into the <head> tag..
+function set_theme_from_local_storage() {
+  // Initialize the theme to null, which means default
+  var theme = null;
+  // If the browser supports the localstorage and is not disabled then try to get the
+  // documenter theme
+  if (window.localStorage != null) {
+    // Get the user-picked theme from localStorage. May be `null`, which means the default
+    // theme.
+    theme = window.localStorage.getItem("documenter-theme");
+  }
+  // Check if the users preference is for dark color scheme
+  var darkPreference =
+    window.matchMedia("(prefers-color-scheme: dark)").matches === true;
+  // Initialize a few variables for the loop:
+  //
+  //  - active: will contain the index of the theme that should be active. Note that there
+  //    is no guarantee that localStorage contains sane values. If `active` stays `null`
+  //    we either could not find the theme or it is the default (primary) theme anyway.
+  //    Either way, we then need to stick to the primary theme.
+  //
+  //  - disabled: style sheets that should be disabled (i.e. all the theme style sheets
+  //    that are not the currently active theme)
+  var active = null;
+  var disabled = [];
+  var primaryLightTheme = null;
+  var primaryDarkTheme = null;
+  for (var i = 0; i < document.styleSheets.length; i++) {
+    var ss = document.styleSheets[i];
+    // The <link> tag of each style sheet is expected to have a data-theme-name attribute
+    // which must contain the name of the theme. The names in localStorage much match this.
+    var themename = ss.ownerNode.getAttribute("data-theme-name");
+    // attribute not set => non-theme stylesheet => ignore
+    if (themename === null) continue;
+    // To distinguish the default (primary) theme, it needs to have the data-theme-primary
+    // attribute set.
+    if (ss.ownerNode.getAttribute("data-theme-primary") !== null) {
+      primaryLightTheme = themename;
+    }
+    // Check if the theme is primary dark theme so that we could store its name in darkTheme
+    if (ss.ownerNode.getAttribute("data-theme-primary-dark") !== null) {
+      primaryDarkTheme = themename;
+    }
+    // If we find a matching theme (and it's not the default), we'll set active to non-null
+    if (themename === theme) active = i;
+    // Store the style sheets of inactive themes so that we could disable them
+    if (themename !== theme) disabled.push(ss);
+  }
+  var activeTheme = null;
+  if (active !== null) {
+    // If we did find an active theme, we'll (1) add the theme--$(theme) class to <html>
+    document.getElementsByTagName("html")[0].className = "theme--" + theme;
+    activeTheme = theme;
+  } else {
+    // If we did _not_ find an active theme, then we need to fall back to the primary theme
+    // which can either be dark or light, depending on the user's OS preference.
+    var activeTheme = darkPreference ? primaryDarkTheme : primaryLightTheme;
+    // In case it somehow happens that the relevant primary theme was not found in the
+    // preceding loop, we abort without doing anything.
+    if (activeTheme === null) {
+      console.error("Unable to determine primary theme.");
+      return;
+    }
+    // When switching to the primary light theme, then we must not have a class name
+    // for the <html> tag. That's only for non-primary or the primary dark theme.
+    if (darkPreference) {
+      document.getElementsByTagName("html")[0].className =
+        "theme--" + activeTheme;
+    } else {
+      document.getElementsByTagName("html")[0].className = "";
+    }
+  }
+  for (var i = 0; i < document.styleSheets.length; i++) {
+    var ss = document.styleSheets[i];
+    // The <link> tag of each style sheet is expected to have a data-theme-name attribute
+    // which must contain the name of the theme. The names in localStorage much match this.
+    var themename = ss.ownerNode.getAttribute("data-theme-name");
+    // attribute not set => non-theme stylesheet => ignore
+    if (themename === null) continue;
+    // we'll disable all the stylesheets, except for the active one
+    ss.disabled = !(themename == activeTheme);
+  }
+}
+set_theme_from_local_storage();
diff --git a/previews/PR3525/assets/warner.js b/previews/PR3525/assets/warner.js
new file mode 100644
index 00000000000..3f6f5d0083a
--- /dev/null
+++ b/previews/PR3525/assets/warner.js
@@ -0,0 +1,52 @@
+function maybeAddWarning() {
+  // DOCUMENTER_NEWEST is defined in versions.js, DOCUMENTER_CURRENT_VERSION and DOCUMENTER_STABLE
+  // in siteinfo.js.
+  // If either of these are undefined something went horribly wrong, so we abort.
+  if (
+    window.DOCUMENTER_NEWEST === undefined ||
+    window.DOCUMENTER_CURRENT_VERSION === undefined ||
+    window.DOCUMENTER_STABLE === undefined
+  ) {
+    return;
+  }
+
+  // Current version is not a version number, so we can't tell if it's the newest version. Abort.
+  if (!/v(\d+\.)*\d+/.test(window.DOCUMENTER_CURRENT_VERSION)) {
+    return;
+  }
+
+  // Current version is newest version, so no need to add a warning.
+  if (window.DOCUMENTER_NEWEST === window.DOCUMENTER_CURRENT_VERSION) {
+    return;
+  }
+
+  // Add a noindex meta tag (unless one exists) so that search engines don't index this version of the docs.
+  if (document.body.querySelector('meta[name="robots"]') === null) {
+    const meta = document.createElement("meta");
+    meta.name = "robots";
+    meta.content = "noindex";
+
+    document.getElementsByTagName("head")[0].appendChild(meta);
+  }
+
+  const div = document.createElement("div");
+  div.classList.add("outdated-warning-overlay");
+  const closer = document.createElement("button");
+  closer.classList.add("outdated-warning-closer", "delete");
+  closer.addEventListener("click", function () {
+    document.body.removeChild(div);
+  });
+  const href = window.documenterBaseURL + "/../" + window.DOCUMENTER_STABLE;
+  div.innerHTML =
+    'This documentation is not for the latest stable release, but for either the development version or an older release.<br><a href="' +
+    href +
+    '">Click here to go to the documentation for the latest stable release.</a>';
+  div.appendChild(closer);
+  document.body.appendChild(div);
+}
+
+if (document.readyState === "loading") {
+  document.addEventListener("DOMContentLoaded", maybeAddWarning);
+} else {
+  maybeAddWarning();
+}
diff --git a/previews/PR3525/background/algebraic_modeling_languages/index.html b/previews/PR3525/background/algebraic_modeling_languages/index.html
new file mode 100644
index 00000000000..ce52312ba4c
--- /dev/null
+++ b/previews/PR3525/background/algebraic_modeling_languages/index.html
@@ -0,0 +1,129 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Algebraic modeling languages · JuMP</title><meta name="title" content="Algebraic modeling languages · JuMP"/><meta property="og:title" content="Algebraic modeling languages · JuMP"/><meta property="twitter:title" content="Algebraic modeling languages · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox" checked/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Algebraic modeling languages</a><ul class="internal"><li><a class="tocitem" href="#What-is-an-algebraic-modeling-language?"><span>What is an algebraic modeling language?</span></a></li><li><a class="tocitem" href="#From-user-to-solver"><span>From user to solver</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Background Information</a></li><li class="is-active"><a href>Algebraic modeling languages</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Algebraic modeling languages</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/background/algebraic_modeling_languages.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="algebraic-modeling-language"><a class="docs-heading-anchor" href="#algebraic-modeling-language">Algebraic modeling languages</a><a id="algebraic-modeling-language-1"></a><a class="docs-heading-anchor-permalink" href="#algebraic-modeling-language" title="Permalink"></a></h1><p>JuMP is an algebraic modeling language for mathematical optimization written in the <a href="https://julialang.org">Julia language</a>. In this page, we explain what an algebraic modeling language actually is.</p><h2 id="What-is-an-algebraic-modeling-language?"><a class="docs-heading-anchor" href="#What-is-an-algebraic-modeling-language?">What is an algebraic modeling language?</a><a id="What-is-an-algebraic-modeling-language?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-an-algebraic-modeling-language?" title="Permalink"></a></h2><p>If you have taken a class in mixed-integer linear programming, you will have seen a formulation like:</p><p class="math-container">\[\begin{aligned}
+\min \;     &amp; c^\top x \\
+\text{s.t.} &amp; A x = b  \\
+            &amp; x \ge 0  \\
+            &amp; x_i \in \mathbb{Z}, \quad \forall i \in \mathcal{I}
+\end{aligned}\]</p><p>where <code>c</code>, <code>A</code>, and <code>b</code> are appropriately sized vectors and matrices of data, and <span>$\mathcal{I}$</span> denotes the set of variables that are integer.</p><p>Solvers expect problems in a <em>standard form</em> like this because it limits the types of constraints that they need to consider. This makes writing a solver much easier.</p><div class="admonition is-info"><header class="admonition-header">What is a solver?</header><div class="admonition-body"><p>A solver is a software package that computes solutions to one or more classes of problems.</p><p>For example, <a href="https://github.com/ERGO-Code/HiGHS">HiGHS</a> is a solver for linear programming (LP) and mixed integer programming (MIP) problems. It incorporates algorithms such as the simplex method and the interior-point method.</p><p>JuMP currently supports a number of open-source and commercial solvers, which can be viewed in the <a href="../../installation/#Supported-solvers">Supported-solvers</a> table.</p></div></div><p>Despite the textbook view of a linear program, you probably formulated problems algebraically like so:</p><p class="math-container">\[\begin{aligned}
+\max \;     &amp; \sum\limits_{i = 1}^n c_i x_i                   \\
+\text{s.t.} &amp; \sum\limits_{i = 1}^n w_i x_i \le b             \\
+            &amp; x_i \ge 0 \quad \forall i = 1,\ldots,n          \\
+            &amp; x_i \in \mathbb{Z} \quad \forall i = 1,\ldots,n.
+\end{aligned}\]</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Do you recognize this formulation? It&#39;s the knapsack problem.</p></div></div><p>Users prefer to write problems in <em>algebraic form</em> because it is more convenient. For example, we used <span>$\le b$</span>, even though the standard form only supported constraints of the form <span>$Ax = b$</span>.</p><p>We could convert our knapsack problem into the standard form by adding a new slack variable <span>$x_0$</span>:</p><p class="math-container">\[\begin{aligned}
+\max \;     &amp; \sum\limits_{i = 1}^n c_i x_i            \\
+\text{s.t.} &amp; x_0 + \sum\limits_{i = 1}^n w_i x_i = b  \\
+            &amp; x_i \ge 0 \quad \forall i = 0,\ldots,n   \\
+            &amp; x_i \in \mathbb{Z} \quad \forall i = 1,\ldots,n.
+\end{aligned}\]</p><p>However, as models get more complicated, this manual conversion becomes more and more error-prone.</p><p>An algebraic modeling language is a tool that simplifies the translation between the algebraic form of the modeler, and the standard form of the solver.</p><p>Each algebraic modeling language has two main parts:</p><ol><li>A domain specific language for the user to write down problems in algebraic form.</li><li>A converter from the algebraic form into a standard form supported by the solver (and back again).</li></ol><p>Part 2 is less trivial than it might seem, because each solver has a unique application programming interface (API) and data structure for representing optimization models and obtaining results.</p><p>JuMP uses the <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface.jl</a> package to abstract these differences between solvers.</p><h3 id="What-is-MathOptInterface?"><a class="docs-heading-anchor" href="#What-is-MathOptInterface?">What is MathOptInterface?</a><a id="What-is-MathOptInterface?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-MathOptInterface?" title="Permalink"></a></h3><p>MathOptInterface (MOI) is an abstraction layer designed to provide an interface to mathematical optimization solvers so that users do not need to understand multiple solver-specific APIs. MOI can be used directly, or through a higher-level modeling interface like JuMP.</p><p>There are three main parts to MathOptInterface:</p><ol><li><p>A solver-independent API that abstracts concepts such as adding and deleting variables and constraints, setting and getting parameters, and querying results. For more information on the MathOptInterface API, read the <a href="../../moi/#moi_documentation">documentation</a>.</p></li><li><p>An automatic rewriting system based on equivalent formulations of a constraint. For more information on this rewriting system, read the <a href="../../manual/models/#LazyBridgeOptimizer">LazyBridgeOptimizer</a> section of the manual, and our <a href="https://arxiv.org/abs/2002.03447">paper on arXiv</a>.</p></li><li><p>Utilities for managing how and when models are copied to solvers. For more information on this, read the <a href="../../manual/models/#CachingOptimizer">CachingOptimizer</a> section of the manual.</p></li></ol><h2 id="From-user-to-solver"><a class="docs-heading-anchor" href="#From-user-to-solver">From user to solver</a><a id="From-user-to-solver-1"></a><a class="docs-heading-anchor-permalink" href="#From-user-to-solver" title="Permalink"></a></h2><p>This section provides a brief summary of the steps that happen in order to translate the model that the user writes into a model that the solver understands.</p><h3 id="Step-I:-writing-in-algebraic-form"><a class="docs-heading-anchor" href="#Step-I:-writing-in-algebraic-form">Step I: writing in algebraic form</a><a id="Step-I:-writing-in-algebraic-form-1"></a><a class="docs-heading-anchor-permalink" href="#Step-I:-writing-in-algebraic-form" title="Permalink"></a></h3><p>JuMP provides the first part of an algebraic modeling language using the <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>, <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a>, and <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macros.</p><p>For example, here&#39;s how we write the knapsack problem in JuMP:</p><pre><code class="language-julia-repl hljs">julia&gt; using JuMP, HiGHS
+
+julia&gt; function algebraic_knapsack(c, w, b)
+           n = length(c)
+           model = Model(HiGHS.Optimizer)
+           set_silent(model)
+           @variable(model, x[1:n] &gt;= 0, Int)
+           @objective(model, Max, sum(c[i] * x[i] for i = 1:n))
+           @constraint(model, sum(w[i] * x[i] for i = 1:n) &lt;= b)
+           optimize!(model)
+           return value.(x)
+       end
+algebraic_knapsack (generic function with 1 method)
+
+julia&gt; algebraic_knapsack([1, 2], [0.5, 0.5], 1.25)
+2-element Vector{Float64}:
+ 0.0
+ 2.0</code></pre><p>This formulation is compact, and it closely matches the algebraic formulation of the model we wrote out above.</p><h3 id="Step-II:-algebraic-to-functional"><a class="docs-heading-anchor" href="#Step-II:-algebraic-to-functional">Step II: algebraic to functional</a><a id="Step-II:-algebraic-to-functional-1"></a><a class="docs-heading-anchor-permalink" href="#Step-II:-algebraic-to-functional" title="Permalink"></a></h3><p>For the next step, JuMP&#39;s macros re-write the variables and constraints into a functional form. Here&#39;s what the JuMP code looks like after this step:</p><pre><code class="language-julia-repl hljs">julia&gt; using JuMP, HiGHS
+
+julia&gt; function nonalgebraic_knapsack(c, w, b)
+           n = length(c)
+           model = Model(HiGHS.Optimizer)
+           set_silent(model)
+           x = [VariableRef(model) for i = 1:n]
+           for i = 1:n
+               set_lower_bound(x[i], 0)
+               set_integer(x[i])
+               set_name(x[i], &quot;x[$i]&quot;)
+           end
+           obj = AffExpr(0.0)
+           for i = 1:n
+               add_to_expression!(obj, c[i], x[i])
+           end
+           set_objective(model, MAX_SENSE, obj)
+           lhs = AffExpr(0.0)
+           for i = 1:n
+               add_to_expression!(lhs, w[i], x[i])
+           end
+           con = build_constraint(error, lhs, MOI.LessThan(b))
+           add_constraint(model, con)
+           optimize!(model)
+           return value.(x)
+       end
+nonalgebraic_knapsack (generic function with 1 method)
+
+julia&gt; nonalgebraic_knapsack([1, 2], [0.5, 0.5], 1.25)
+2-element Vector{Float64}:
+ 0.0
+ 2.0</code></pre><p>Hopefully you agree that the macro version is much easier to read.</p><h3 id="Part-III:-JuMP-to-MathOptInterface"><a class="docs-heading-anchor" href="#Part-III:-JuMP-to-MathOptInterface">Part III: JuMP to MathOptInterface</a><a id="Part-III:-JuMP-to-MathOptInterface-1"></a><a class="docs-heading-anchor-permalink" href="#Part-III:-JuMP-to-MathOptInterface" title="Permalink"></a></h3><p>In the third step, JuMP converts the functional form of the problem, that is, <code>nonalgebraic_knapsack</code>, into the MathOptInterface API:</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; import HiGHS
+
+julia&gt; function mathoptinterface_knapsack(optimizer, c, w, b)
+           n = length(c)
+           model = MOI.instantiate(optimizer)
+           MOI.set(model, MOI.Silent(), true)
+           x = MOI.add_variables(model, n)
+           for i in 1:n
+               MOI.add_constraint(model, x[i], MOI.GreaterThan(0.0))
+               MOI.add_constraint(model, x[i], MOI.Integer())
+               MOI.set(model, MOI.VariableName(), x[i], &quot;x[$i]&quot;)
+           end
+           MOI.set(model, MOI.ObjectiveSense(), MOI.MAX_SENSE)
+           obj = MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(c, x), 0.0)
+           MOI.set(model, MOI.ObjectiveFunction{typeof(obj)}(), obj)
+           MOI.add_constraint(
+               model,
+               MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(w, x), 0.0),
+               MOI.LessThan(b),
+           )
+           MOI.optimize!(model)
+           return MOI.get.(model, MOI.VariablePrimal(), x)
+       end
+mathoptinterface_knapsack (generic function with 1 method)
+
+julia&gt; mathoptinterface_knapsack(HiGHS.Optimizer, [1.0, 2.0], [0.5, 0.5], 1.25)
+2-element Vector{Float64}:
+ 0.0
+ 2.0</code></pre><p>The code is becoming more verbose and looking less like the mathematical formulation that we started with.</p><h3 id="Step-IV:-MathOptInterface-to-HiGHS"><a class="docs-heading-anchor" href="#Step-IV:-MathOptInterface-to-HiGHS">Step IV: MathOptInterface to HiGHS</a><a id="Step-IV:-MathOptInterface-to-HiGHS-1"></a><a class="docs-heading-anchor-permalink" href="#Step-IV:-MathOptInterface-to-HiGHS" title="Permalink"></a></h3><p>As a final step, the <a href="https://github.com/jump-dev/HiGHS.jl">HiGHS.jl</a> package converts the MathOptInterface form, that is, <code>mathoptinterface_knapsack</code>, into a HiGHS-specific API:</p><pre><code class="language-julia-repl hljs">julia&gt; using HiGHS
+
+julia&gt; function highs_knapsack(c, w, b)
+           n = length(c)
+           model = Highs_create()
+           Highs_setBoolOptionValue(model, &quot;output_flag&quot;, false)
+           for i in 1:n
+               Highs_addCol(model, c[i], 0.0, Inf, 0, C_NULL, C_NULL)
+               Highs_changeColIntegrality(model, i-1, 1)
+           end
+           Highs_changeObjectiveSense(model, -1)
+           Highs_addRow(
+               model,
+               -Inf,
+               b,
+               Cint(length(w)),
+               collect(Cint(0):Cint(n-1)),
+               w,
+           )
+           Highs_run(model)
+           x = fill(NaN, 2)
+           Highs_getSolution(model, x, C_NULL, C_NULL, C_NULL)
+           Highs_destroy(model)
+           return x
+       end
+highs_knapsack (generic function with 1 method)
+
+julia&gt; highs_knapsack([1.0, 2.0], [0.5, 0.5], 1.25)
+2-element Vector{Float64}:
+ 0.0
+ 2.0</code></pre><p>We&#39;ve now gone from a algebraic model that looked identical to the mathematical model we started with, to a verbose function that uses HiGHS-specific functionality.</p><p>The difference between <code>algebraic_knapsack</code> and <code>highs_knapsack</code> highlights the benefit that algebraic modeling languages provide to users. Moreover, if we used a different solver, the solver-specific function would be entirely different. A key benefit of an algebraic modeling language is that you can change the solver without needing to rewrite the model.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../api/JuMP.Containers/">« JuMP.Containers</a><a class="docs-footer-nextpage" href="../../developers/contributing/">Contributing »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/changelog/index.html b/previews/PR3525/changelog/index.html
new file mode 100644
index 00000000000..3b1b94cd65d
--- /dev/null
+++ b/previews/PR3525/changelog/index.html
@@ -0,0 +1,15 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Release notes · JuMP</title><meta name="title" content="Release notes · JuMP"/><meta property="og:title" content="Release notes · JuMP"/><meta property="twitter:title" content="Release notes · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL=".."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../assets/documenter.js"></script><script src="../search_index.js"></script><script src="../siteinfo.js"></script><script src="../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../assets/themeswap.js"></script><link href="../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../"><img class="docs-light-only" src="../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../">Introduction</a></li><li><a class="tocitem" href="../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../manual/models/">Models</a></li><li><a class="tocitem" href="../manual/variables/">Variables</a></li><li><a class="tocitem" href="../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../manual/containers/">Containers</a></li><li><a class="tocitem" href="../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/">Introduction</a></li><li><a class="tocitem" href="../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li class="is-active"><a href>Release notes</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Release notes</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/changelog.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Release-notes"><a class="docs-heading-anchor" href="#Release-notes">Release notes</a><a id="Release-notes-1"></a><a class="docs-heading-anchor-permalink" href="#Release-notes" title="Permalink"></a></h1><p>The format is based on <a href="https://keepachangelog.com/en/1.0.0/">Keep a Changelog</a>, and this project adheres to <a href="https://semver.org/spec/v2.0.0.html">Semantic Versioning</a>.</p><h2 id="Version-1.15.1-(September-23,-2023)"><a class="docs-heading-anchor" href="#Version-1.15.1-(September-23,-2023)">Version 1.15.1 (September 23, 2023)</a><a id="Version-1.15.1-(September-23,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.15.1-(September-23,-2023)" title="Permalink"></a></h2><h3 id="Fixed"><a class="docs-heading-anchor" href="#Fixed">Fixed</a><a id="Fixed-1"></a><a class="docs-heading-anchor-permalink" href="#Fixed" title="Permalink"></a></h3><ul><li>Fixed support for single argument <code>min</code> and <code>max</code> operators (#3522)</li><li>Fixed error message for <a href="../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a> when called with a <a href="../api/JuMP/#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a> (#3506)</li><li>Fixed constraint tags with broadcasted constraints (#3515)</li><li>Fixed MethodError in <code>MA.scaling</code> (#3518)</li><li>Fixed support for arrays of <a href="../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a> variables (#3524)</li></ul><h3 id="Other"><a class="docs-heading-anchor" href="#Other">Other</a><a id="Other-1"></a><a class="docs-heading-anchor-permalink" href="#Other" title="Permalink"></a></h3><ul><li>Updated to Documenter@1 (#3501)</li><li>Fixed links to data in tutorials (#3512)</li><li>Fixed typo in TSP tutorial (#3516)</li><li>Improved error message for <a href="../api/JuMP/#JuMP.VariableNotOwned"><code>VariableNotOwned</code></a> errors (#3520)</li><li>Fixed various JET errors (#3519)</li></ul><h2 id="Version-1.15.0-(September-15,-2023)"><a class="docs-heading-anchor" href="#Version-1.15.0-(September-15,-2023)">Version 1.15.0 (September 15, 2023)</a><a id="Version-1.15.0-(September-15,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.15.0-(September-15,-2023)" title="Permalink"></a></h2><p>This is a large minor release because it adds an entirely new data structure and API path for working with nonlinear programs. The previous nonlinear interface remains unchanged and is documented at <a href="../manual/nlp/#Nonlinear-Modeling-(Legacy)">Nonlinear Modeling (Legacy)</a>. The new interface is a treated as a non-breaking feature addition and is documented at <a href="../manual/nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a>.</p><h3 id="Breaking"><a class="docs-heading-anchor" href="#Breaking">Breaking</a><a id="Breaking-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking" title="Permalink"></a></h3><p>Although the new nonlinear interface is a feature addition, there are two changes which might be breaking for a very small number of users.</p><ul><li>The syntax inside JuMP macros is parsed using a different code path, even for linear and quadratic expressions. We made this change to unify how we parse linear, quadratic, and nonlinear expressions. In all cases, the new code returns equivalent expressions, but because of the different order of operations, there are three changes to be aware of when updating:<ul><li>The printed form of the expression may change, for example from <code>x * y</code> to <code>y * x</code>. This can cause tests which test the <code>String</code> representation of a model to fail.</li><li>Some coefficients may change slightly due to floating point round-off error.</li><li>Particularly when working with a JuMP extension, you may encounter a <code>MethodError</code> due to a missing or ambiguous method. These errors are due to previously existing bugs that were not triggered by the previous parsing code. If you encounter such an error, please open a GitHub issue.</li></ul></li><li>The methods for <code>Base.:^(x::VariableRef, n::Integer)</code> and <code>Base.:^(x::AffExpr, n::Integer)</code> have changed. Previously, these methods supported only <code>n = 0, 1, 2</code> and they always returned a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a>, even for the case when <code>n = 0</code> or <code>n = 1</code>. Now:<ul><li><code>x^0</code> returns <code>one(T)</code>, where <code>T</code> is the <a href="../api/JuMP/#JuMP.value_type"><code>value_type</code></a> of the model (defaults to <code>Float64</code>)</li><li><code>x^1</code> returns <code>x</code></li><li><code>x^2</code> returns a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a></li><li><code>x^n</code> where <code>!(0 &lt;= n &lt;= 2)</code> returns a <a href="../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a>.</li></ul>We made this change to support nonlinear expressions and to align the mathematical definition of the operation with their return type. (Previously, users were surprised that <code>x^1</code> returned a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a>.) As a consequence of this change, the methods are now not type-stable. This means that the compiler cannot prove that <code>x^2</code> returns a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a>. If benchmarking shows that this is a performance problem, you can use the type-stable <code>x * x</code> instead of <code>x^2</code>.</li></ul><h3 id="Added"><a class="docs-heading-anchor" href="#Added">Added</a><a id="Added-1"></a><a class="docs-heading-anchor-permalink" href="#Added" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP/#JuMP.triangle_vec"><code>triangle_vec</code></a> which simplifies adding <a href="../moi/reference/standard_form/#MathOptInterface.LogDetConeTriangle"><code>MOI.LogDetConeTriangle</code></a> and <a href="../moi/reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>MOI.RootDetConeTriangle</code></a> constraints (#3456)</li><li>Added the new nonlinear interface. This is a very large change. See the documentation at <a href="../manual/nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a> and the (long) discussion in <a href="https://github.com/jump-dev/JuMP.jl/pull/3106">JuMP.jl#3106</a>. Related PRs are (#3468) (#3472) (#3475) (#3483) (#3487) (#3488) (#3489) (#3504) (#3509)</li></ul><h3 id="Fixed-2"><a class="docs-heading-anchor" href="#Fixed-2">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-2" title="Permalink"></a></h3><ul><li>Fixed uses of <code>@nospecialize</code> which cause precompilation failures in Julia v1.6.0 and v1.6.1. (#3464)</li><li>Fixed adding a container of <a href="../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a> (#3473)</li><li>Fixed return type of <code>x^0</code> and <code>x^1</code> to no longer return <code>QuadExpr</code> (see note in <code>Breaking</code> section above) (#3474)</li><li>Fixed error messages in <a href="../api/JuMP/#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, <a href="../api/JuMP/#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, <a href="../api/JuMP/#JuMP.FixRef"><code>FixRef</code></a>, <a href="../api/JuMP/#JuMP.IntegerRef"><code>IntegerRef</code></a>, <a href="../api/JuMP/#JuMP.BinaryRef"><code>BinaryRef</code></a>, <a href="../api/JuMP/#JuMP.ParameterRef"><code>ParameterRef</code></a> and related functions (#3494)</li><li>Fixed type inference of empty containers in JuMP macros (#3500)</li></ul><h3 id="Other-2"><a class="docs-heading-anchor" href="#Other-2">Other</a><a class="docs-heading-anchor-permalink" href="#Other-2" title="Permalink"></a></h3><ul><li>Added GAMS to solver documentation (#3357)</li><li>Updated various tutorials (#3459) (#3460) (#3462) (#3463) (#3465) (#3490) (#3492) (#3503)</li><li>Added <a href="../tutorials/linear/multi_commodity_network/#The-network-multi-commodity-flow-problem">The network multi-commodity flow problem</a> tutorial (#3491)</li><li>Added <a href="../tutorials/applications/two_stage_stochastic/#Two-stage-stochastic-programs">Two-stage stochastic programs</a> tutorial (#3466)</li><li>Added better error messages for unsupported operations in <code>LinearAlgebra</code> (#3476)</li><li>Updated to the latest version of Documenter (#3484) (#3495) (#3497)</li><li>Updated GitHub action versions (#3507)</li></ul><h2 id="Version-1.14.1-(September-2,-2023)"><a class="docs-heading-anchor" href="#Version-1.14.1-(September-2,-2023)">Version 1.14.1 (September 2, 2023)</a><a id="Version-1.14.1-(September-2,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.14.1-(September-2,-2023)" title="Permalink"></a></h2><h3 id="Fixed-3"><a class="docs-heading-anchor" href="#Fixed-3">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-3" title="Permalink"></a></h3><ul><li>Fix links in Documentation (#3478)</li></ul><h2 id="Version-1.14.0-(August-27,-2023)"><a class="docs-heading-anchor" href="#Version-1.14.0-(August-27,-2023)">Version 1.14.0 (August 27, 2023)</a><a id="Version-1.14.0-(August-27,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.14.0-(August-27,-2023)" title="Permalink"></a></h2><h3 id="Added-2"><a class="docs-heading-anchor" href="#Added-2">Added</a><a class="docs-heading-anchor-permalink" href="#Added-2" title="Permalink"></a></h3><ul><li>Added <a href="../extensions/DimensionalData/#DimensionalData.jl">DimensionalData.jl</a> extension (#3413)</li><li>Added syntactic sugar for the <a href="../moi/reference/standard_form/#MathOptInterface.Parameter"><code>MOI.Parameter</code></a> set (#3443)<ul><li><a href="../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a></li><li><a href="../api/JuMP/#JuMP.ParameterRef"><code>ParameterRef</code></a></li><li><a href="../api/JuMP/#JuMP.is_parameter"><code>is_parameter</code></a></li><li><a href="../api/JuMP/#JuMP.parameter_value"><code>parameter_value</code></a></li><li><a href="../api/JuMP/#JuMP.set_parameter_value"><code>set_parameter_value</code></a></li></ul></li></ul><h3 id="Fixed-4"><a class="docs-heading-anchor" href="#Fixed-4">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-4" title="Permalink"></a></h3><ul><li>Fixed <code>model_convert</code> for <code>BridgeableConstraint</code> (#3437)</li><li>Fixed printing models with integer coefficients larger than <code>typemax(Int)</code> (#3447)</li><li>Fixed support for constant left-hand side functions in a complementarity constraint (#3452)</li></ul><h3 id="Other-3"><a class="docs-heading-anchor" href="#Other-3">Other</a><a class="docs-heading-anchor-permalink" href="#Other-3" title="Permalink"></a></h3><ul><li>Updated packages used in documentation (#3444) (#3455)</li><li>Fixed docstring tests (#3445)</li><li>Fixed printing change for MathOptInterface (#3446)</li><li>Fixed typos in documentation (#3448) (#3457)</li><li>Added SCIP to callback documentation (#3449)</li></ul><h2 id="Version-1.13.0-(July-27,-2023)"><a class="docs-heading-anchor" href="#Version-1.13.0-(July-27,-2023)">Version 1.13.0 (July 27, 2023)</a><a id="Version-1.13.0-(July-27,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.13.0-(July-27,-2023)" title="Permalink"></a></h2><h3 id="Added-3"><a class="docs-heading-anchor" href="#Added-3">Added</a><a class="docs-heading-anchor-permalink" href="#Added-3" title="Permalink"></a></h3><ul><li>Added support for generic number types (#3377) (#3385)</li><li>Added fallback for <a href="../moi/reference/standard_form/#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>MOI.AbstractSymmetricMatrixSetTriangle</code></a> and <a href="../moi/reference/standard_form/#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>MOI.AbstractSymmetricMatrixSetSquare</code></a> (#3424)</li></ul><h3 id="Fixed-5"><a class="docs-heading-anchor" href="#Fixed-5">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-5" title="Permalink"></a></h3><ul><li>Fixed <a href="../api/JuMP/#JuMP.set_start_values"><code>set_start_values</code></a> with <a href="../moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.SlackBridge"><code>MOI.Bridges.Objective.SlackBridge</code></a> (#3422)</li><li>Fixed flakey doctest in <code>variables.md</code> (#3425)</li><li>Fixed names on <code>CITATION.bib</code> (#3423)</li></ul><h3 id="Other-4"><a class="docs-heading-anchor" href="#Other-4">Other</a><a class="docs-heading-anchor-permalink" href="#Other-4" title="Permalink"></a></h3><ul><li>Added Loraine.jl to the installation table (#3426)</li><li>Removed Penopt.jl from packages.toml (#3428)</li><li>Improved problem statement in cannery example of tutorial (#3430)</li><li>Minor cleanups in <a href="../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> implementation (#3429)</li><li>Changed <code>nested_problems.jl</code>: outer/inner to upper/lower (#3433)</li><li>Removed second SDP relaxation in OPF tutorial (#3432)</li></ul><h2 id="Version-1.12.0-(June-19,-2023)"><a class="docs-heading-anchor" href="#Version-1.12.0-(June-19,-2023)">Version 1.12.0 (June 19, 2023)</a><a id="Version-1.12.0-(June-19,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.12.0-(June-19,-2023)" title="Permalink"></a></h2><h3 id="Added-4"><a class="docs-heading-anchor" href="#Added-4">Added</a><a class="docs-heading-anchor-permalink" href="#Added-4" title="Permalink"></a></h3><ul><li>Added <code>coefficient_type</code> keyword argument to <a href="../api/JuMP/#JuMP.add_bridge"><code>add_bridge</code></a> and <a href="../api/JuMP/#JuMP.remove_bridge"><code>remove_bridge</code></a> (#3394)</li></ul><h3 id="Fixed-6"><a class="docs-heading-anchor" href="#Fixed-6">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-6" title="Permalink"></a></h3><ul><li>Fixed error message for matrix in <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> (#3369)</li><li>Fixed <code>EditURL</code> for custom documentation pages (#3373)</li><li>Fixed return type annotations for <a href="../moi/reference/constraints/#MathOptInterface.ConstraintPrimal"><code>MOI.ConstraintPrimal</code></a> and <a href="../moi/reference/constraints/#MathOptInterface.ConstraintDual"><code>MOI.ConstraintDual</code></a> (#3381)</li><li>Fixed printing change in Julia nightly (#3391)</li><li>Fixed printing of <code>Complex</code> coefficients (#3397)</li><li>Fixed printing of constraints in <code>text/latex</code> mode (#3405)</li><li>Fixed performance issue in <a href="../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> (#3410)</li><li>Fixed bug when variables added to set of wrong dimension (#3411)</li></ul><h3 id="Other-5"><a class="docs-heading-anchor" href="#Other-5">Other</a><a class="docs-heading-anchor-permalink" href="#Other-5" title="Permalink"></a></h3><ul><li>Added more solver READMEs to the documentation (#3358) (#3360) (#3364) (#3365) (#3366) (#3368) (#3372) (#3374) (#3376) (#3379) (#3387) (#3389)</li><li>Added StatusSwitchingQP.jl to the installation table (#3354)</li><li>Updated checklist for adding a new solver (#3370)</li><li>Updated <code>extension-tests.yml</code> action (#3371) (#3375)</li><li>Color logs in GitHub actions (#3392)</li><li>Added new tutorials<ul><li><a href="../tutorials/applications/optimal_power_flow/#Optimal-power-flow">Optimal power flow</a> (#3395) (#3412)</li><li><a href="../tutorials/conic/simple_examples/#Lovász-numbers">Lovász numbers</a> (#3399)</li><li><a href="../tutorials/conic/dualization/#Dualization">Dualization</a> (#3402)</li></ul></li><li>Updated JuMP paper citation (#3400)</li><li>Changed GitHub action to upload LaTeX logs when building documentation (#3403)</li><li>Fixed printing of SCS log in documentation (#3406)</li><li>Updated solver versions (#3407)</li><li>Updated documentation to use Julia v1.9 (#3398)</li><li>Replaced <code>_value_type</code> with <code>MOI.Utilities.value_type</code> (#3414)</li><li>Fixed a typo in docstring (#3415)</li><li>Refactored API documentation (#3386)</li><li>Updated SCIP license (#3420)</li></ul><h2 id="Version-1.11.1-(May-19,-2023)"><a class="docs-heading-anchor" href="#Version-1.11.1-(May-19,-2023)">Version 1.11.1 (May 19, 2023)</a><a id="Version-1.11.1-(May-19,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.11.1-(May-19,-2023)" title="Permalink"></a></h2><h3 id="Fixed-7"><a class="docs-heading-anchor" href="#Fixed-7">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-7" title="Permalink"></a></h3><ul><li>Fixed a poor error message when <code>sum(::DenseAxisArray; dims)</code> was called (#3338)</li><li>Fixed support for dependent sets in the <a href="../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro (#3344)</li><li>Fixed a performance bug in constraints with sparse symmetric matrices (#3349)</li></ul><h3 id="Other-6"><a class="docs-heading-anchor" href="#Other-6">Other</a><a class="docs-heading-anchor-permalink" href="#Other-6" title="Permalink"></a></h3><ul><li>Improved the printing of complex numbers (#3332)</li><li>When printing, sets which contain constants ending in <code>.0</code> now print as integers. This follows the behavior of constants in functions (#3341)</li><li>Added <code>InfiniteOpt</code> to the extensions documentation (#3343)</li><li>Added more documentation for the exponential cone (#3345) (#3347)</li><li>Added checklists for developers (#3346) (#3355)</li><li>Fixed test support upcoming Julia nightly (#3351)</li><li>Fixed <code>extension-tests.yml</code> action (#3353)</li><li>Add more solvers to the documentation (#3359) (#3361) (#3362)</li></ul><h2 id="Version-1.11.0-(May-3,-2023)"><a class="docs-heading-anchor" href="#Version-1.11.0-(May-3,-2023)">Version 1.11.0 (May 3, 2023)</a><a id="Version-1.11.0-(May-3,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.11.0-(May-3,-2023)" title="Permalink"></a></h2><h3 id="Added-5"><a class="docs-heading-anchor" href="#Added-5">Added</a><a class="docs-heading-anchor-permalink" href="#Added-5" title="Permalink"></a></h3><ul><li>Added new methods to <a href="../api/JuMP/#JuMP.print_active_bridges"><code>print_active_bridges</code></a> for printing a particular objective, constraint, or variable (#3316)</li></ul><h3 id="Fixed-8"><a class="docs-heading-anchor" href="#Fixed-8">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-8" title="Permalink"></a></h3><ul><li>Fixed tests for MOI v1.14.0 release (#3312)</li><li>Fixed indexing containers when an axis is <code>Vector{Any}</code> that contains a <code>Vector{Any}</code> element (#3280)</li><li>Fixed <code>getindex(::AbstractJuMPScalar)</code> which is called for an expression like <code>x[]</code> (#3314)</li><li>Fixed bug in <code>set_string_names_on_creation</code> with a vector of variables (#3322)</li><li>Fixed bug in <code>memoize</code> function in nonlinear documentation (#3337)</li></ul><h3 id="Other-7"><a class="docs-heading-anchor" href="#Other-7">Other</a><a class="docs-heading-anchor-permalink" href="#Other-7" title="Permalink"></a></h3><ul><li>Fixed typos in the documentation (#3317) (#3318) (#3328)</li><li>Added a test for the order of setting start values (#3315)</li><li>Added READMEs of solvers and extensions to the docs (#3309) (#3320) (#3327) (#3329) (#3333)</li><li>Style improvements to <code>src/variables.jl</code> (#3324)</li><li>Clarify that column generation does not find global optimum (#3325)</li><li>Add a GitHub actions workflow for testing extensions prior to release (#3331)</li><li>Document the release process for JuMP (#3334)</li><li>Fix links to discourse and chatroom (#3335)</li></ul><h2 id="Version-1.10.0-(April-3,-2023)"><a class="docs-heading-anchor" href="#Version-1.10.0-(April-3,-2023)">Version 1.10.0 (April 3, 2023)</a><a id="Version-1.10.0-(April-3,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.10.0-(April-3,-2023)" title="Permalink"></a></h2><h3 id="Added-6"><a class="docs-heading-anchor" href="#Added-6">Added</a><a class="docs-heading-anchor-permalink" href="#Added-6" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP/#JuMP.Nonnegatives"><code>Nonnegatives</code></a>, <a href="../api/JuMP/#JuMP.Nonpositives"><code>Nonpositives</code></a> and <a href="../api/JuMP/#JuMP.Zeros"><code>Zeros</code></a>, and support vector-valued inequality syntax in the JuMP macros (#3273)</li><li>Added special support for <code>LinearAlgebra.Symmetric</code> and <code>LinearAlgebra.Hermitian</code> matrices in <a href="../api/JuMP/#JuMP.Zeros"><code>Zeros</code></a> constraints (#3281) (#3296)</li><li>Added <a href="../api/JuMP/#JuMP.HermitianMatrixSpace"><code>HermitianMatrixSpace</code></a> and the <code>Hermitian</code> tag for generating a matrix of variables that is Hermitian (#3292) (#3293)</li><li>Added <a href="../api/JuMP/#JuMP.Semicontinuous"><code>Semicontinuous</code></a> and <a href="../api/JuMP/#JuMP.Semiinteger"><code>Semiinteger</code></a> (#3302)</li><li>Added support for keyword indexing of containers (#3237)</li></ul><h3 id="Fixed-9"><a class="docs-heading-anchor" href="#Fixed-9">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-9" title="Permalink"></a></h3><ul><li>Fixed <code>[compat]</code> bound for MathOptInterface in <code>Project.toml</code> (#3272)</li></ul><h3 id="Other-8"><a class="docs-heading-anchor" href="#Other-8">Other</a><a class="docs-heading-anchor-permalink" href="#Other-8" title="Permalink"></a></h3><ul><li>Split out the <a href="../tutorials/nonlinear/nested_problems/#Nested-optimization-problems">Nested optimization problems</a> tutorial (#3274)</li><li>Updated doctests to ensure none have hidden state (#3275) (#3276)</li><li>Clarified how lazy constraints may revisit points (#3278)</li><li>Added <a href="../tutorials/conic/tips_and_tricks/#P-Norm">P-Norm</a> example (#3282)</li><li>Clarified docs that macros create new bindings (#3284)</li><li>Fixed threading example (#3283)</li><li>Added plot to <a href="../tutorials/conic/simple_examples/#The-minimum-distortion-problem">The minimum distortion problem</a> (#3288)</li><li>Added Google style rules for Vale and fixed warnings (#3285)</li><li>Added citation for the JuMP 1.0 paper (#3294)</li><li>Updated package versions in the documentation (#3298)</li><li>Added comment for the order in which start values must be set (#3303)</li><li>Improved error message for unrecognized constraint operators (#3311)</li></ul><h2 id="Version-1.9.0-(March-7,-2023)"><a class="docs-heading-anchor" href="#Version-1.9.0-(March-7,-2023)">Version 1.9.0 (March 7, 2023)</a><a id="Version-1.9.0-(March-7,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.9.0-(March-7,-2023)" title="Permalink"></a></h2><h3 id="Added-7"><a class="docs-heading-anchor" href="#Added-7">Added</a><a class="docs-heading-anchor-permalink" href="#Added-7" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP/#JuMP.get_attribute"><code>get_attribute</code></a> and <a href="../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a>. These replace <a href="../api/JuMP/#JuMP.get_optimizer_attribute"><code>get_optimizer_attribute</code></a> and <a href="../api/JuMP/#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a>, although the <code>_optimizer_</code> functions remain for backward compatibility. (#3219)</li><li>Added <a href="../api/JuMP/#JuMP.set_start_values"><code>set_start_values</code></a> for setting all supported start values in a model (#3238)</li><li>Add <a href="../api/JuMP/#JuMP.remove_bridge"><code>remove_bridge</code></a> and <a href="../api/JuMP/#JuMP.print_active_bridges"><code>print_active_bridges</code></a> (#3259)</li></ul><h3 id="Fixed-10"><a class="docs-heading-anchor" href="#Fixed-10">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-10" title="Permalink"></a></h3><ul><li>The matrix returned by a variable in <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> is now a <code>LinearAlgebra.Hermitian</code> matrix. This is potentially breaking if you have written code to assume the return is a <code>Matrix</code>. (#3245) (#3246)</li><li>Fixed missing support for <code>Base.isreal</code> of expressions (#3252)</li></ul><h3 id="Other-9"><a class="docs-heading-anchor" href="#Other-9">Other</a><a class="docs-heading-anchor-permalink" href="#Other-9" title="Permalink"></a></h3><ul><li>Fixed a thread safety issue in the <a href="../tutorials/algorithms/parallelism/#Parallelism">Parallelism</a> tutorial (#3240) (#3243)</li><li>Improved the error message when unsupported operators are used in <code>@NL</code> macros (#3236)</li><li>Clarified the documentation to say that matrices in <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> must be <code>LinearAlgebra.Hermitian</code> (#3241)</li><li>Minor style fixes to internal macro code (#3247)</li><li>Add <a href="../tutorials/conic/quantum_discrimination/#Quantum-state-discrimination">Quantum state discrimination</a> tutorial (#3250)</li><li>Improve error message when <code>begin...end</code> not passed to plural macros (#3255)</li><li>Document how to register function with varying number of input arguments (#3258)</li><li>Tidy tests by removing unneeded <code>JuMP.</code> prefixes (#3260)</li><li>Clarified the introduction to the <a href="../manual/complex/#Complex-number-support">Complex number support</a> tutorial (#3262)</li><li>Fixed typos in the Documentation (#3263) (#3266) (#3268) (#3269)</li></ul><h2 id="Version-1.8.2-(February-27,-2023)"><a class="docs-heading-anchor" href="#Version-1.8.2-(February-27,-2023)">Version 1.8.2 (February 27, 2023)</a><a id="Version-1.8.2-(February-27,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.8.2-(February-27,-2023)" title="Permalink"></a></h2><h3 id="Fixed-11"><a class="docs-heading-anchor" href="#Fixed-11">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-11" title="Permalink"></a></h3><ul><li>Fixed dot product between complex JuMP expression and number (#3244)</li></ul><h3 id="Other-10"><a class="docs-heading-anchor" href="#Other-10">Other</a><a class="docs-heading-anchor-permalink" href="#Other-10" title="Permalink"></a></h3><ul><li>Polish simple SDP examples (#3232)</li></ul><h2 id="Version-1.8.1-(February-23,-2023)"><a class="docs-heading-anchor" href="#Version-1.8.1-(February-23,-2023)">Version 1.8.1 (February 23, 2023)</a><a id="Version-1.8.1-(February-23,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.8.1-(February-23,-2023)" title="Permalink"></a></h2><h3 id="Fixed-12"><a class="docs-heading-anchor" href="#Fixed-12">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-12" title="Permalink"></a></h3><ul><li>Fixed support for <code>init</code> in nonlinear generator expressions (#3226)</li></ul><h3 id="Other-11"><a class="docs-heading-anchor" href="#Other-11">Other</a><a class="docs-heading-anchor-permalink" href="#Other-11" title="Permalink"></a></h3><ul><li>Use and document <code>import MathOptInterface as MOI</code> (#3222)</li><li>Removed references in documentation to multiobjective optimization being unsupported (#3223)</li><li>Added tutorial on multi-objective portfolio optimization (#3227)</li><li>Refactored some of the conic tutorials (#3229)</li><li>Fixed typos in the documentation (#3230)</li><li>Added tutorial on parallelism (#3231)</li></ul><h2 id="Version-1.8.0-(February-16,-2023)"><a class="docs-heading-anchor" href="#Version-1.8.0-(February-16,-2023)">Version 1.8.0 (February 16, 2023)</a><a id="Version-1.8.0-(February-16,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.8.0-(February-16,-2023)" title="Permalink"></a></h2><h3 id="Added-8"><a class="docs-heading-anchor" href="#Added-8">Added</a><a class="docs-heading-anchor-permalink" href="#Added-8" title="Permalink"></a></h3><ul><li>Added <code>--&gt;</code> syntax support for indicator constraints. The old syntax of <code>=&gt;</code> remains supported (#3207)</li><li>Added <code>&lt;--&gt;</code> syntax for reified constraints. For now, few solvers support reified constraints (#3206)</li><li>Added <a href="../api/JuMP/#JuMP.fix_discrete_variables"><code>fix_discrete_variables</code></a>. This is most useful for computing the dual of a mixed-integer program (#3208)</li><li>Added support for vector-valued objectives. For details, see the <a href="../tutorials/linear/multi_objective_knapsack/#Multi-objective-knapsack">Multi-objective knapsack</a> tutorial (#3176)</li></ul><h3 id="Fixed-13"><a class="docs-heading-anchor" href="#Fixed-13">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-13" title="Permalink"></a></h3><ul><li>Fixed a bug in <a href="../api/JuMP/#JuMP.lp_sensitivity_report"><code>lp_sensitivity_report</code></a> by switching to an explicit LU factorization of the basis matrix (#3182)</li><li>Fixed a bug that prevented <code>[; kwarg]</code> arguments in macros (#3220)</li></ul><h3 id="Other-12"><a class="docs-heading-anchor" href="#Other-12">Other</a><a class="docs-heading-anchor-permalink" href="#Other-12" title="Permalink"></a></h3><ul><li>Minor fixes to the documentation (#3200) (#3201) (#3203) (#3210)</li><li>Added tutorial <a href="../tutorials/linear/constraint_programming/#Constraint-programming">Constraint programming</a> (#3202)</li><li>Added more examples to <a href="../tutorials/conic/tips_and_tricks/#conic_tips_and_tricks">Tips and Tricks</a></li><li>Remove <code>_distance_to_set</code> in favor of <code>MOI.Utilities.distance_to_set</code> (#3209)</li><li>Improve <a href="../tutorials/linear/diet/#The-diet-problem">The diet problem</a> tutorial by adding the variable as a column in the dataframe (#3213)</li><li>Improve <a href="../tutorials/linear/knapsack/#The-knapsack-problem-example">The knapsack problem example</a> tutorial (#3216) (#3217)</li><li>Added the <a href="../tutorials/conic/ellipse_approx/#Ellipsoid-approximation">Ellipsoid approximation</a> tutorial (#3218)</li></ul><h2 id="Version-1.7.0-(January-25,-2023)"><a class="docs-heading-anchor" href="#Version-1.7.0-(January-25,-2023)">Version 1.7.0 (January 25, 2023)</a><a id="Version-1.7.0-(January-25,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.7.0-(January-25,-2023)" title="Permalink"></a></h2><h3 id="Added-9"><a class="docs-heading-anchor" href="#Added-9">Added</a><a class="docs-heading-anchor-permalink" href="#Added-9" title="Permalink"></a></h3><ul><li>Added support for <code>view</code> of a <a href="../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> (#3152) (#3180)</li><li>Added support for containers of variables in <a href="../api/JuMP/#JuMP.ComplexPlane"><code>ComplexPlane</code></a> (#3184)</li><li>Added support for <code>minimum</code> and <code>maximum</code> generators in nonlinear expressions (#3189)</li><li>Added <code>SnoopPrecompile</code> statements that reduce the time-to-first-solve in Julia 1.9 (#3193) (#3195) (#3196) (#3197)</li></ul><h3 id="Other-13"><a class="docs-heading-anchor" href="#Other-13">Other</a><a class="docs-heading-anchor-permalink" href="#Other-13" title="Permalink"></a></h3><ul><li>Large refactoring of the tests (#3166) (#3167) (#3168) (#3169) (#3170) (#3171)</li><li>Remove unreachable code due to <code>VERSION</code> checks (#3172)</li><li>Document how to test JuMP extensions (#3174)</li><li>Fix method ambiguities in <code>Containers</code> (#3173)</li><li>Improve error message that is thrown when <code>=</code> is used instead of <code>==</code> in the <a href="../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro (#3178)</li><li>Improve the error message when <code>Bool</code> is used instead of <code>Bin</code> in the <a href="../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro (#3180)</li><li>Update versions of the documentation (#3185)</li><li>Tidy the import of packages and remove unnecessary prefixes (#3186) (#3187)</li><li>Refactor <code>src/JuMP.jl</code> by moving methods into more relevant files (#3188)</li><li>Fix docstring of <a href="../api/JuMP/#JuMP.Model"><code>Model</code></a> not appearing in the documentation (#3198)</li></ul><h2 id="Version-1.6.0-(January-1,-2023)"><a class="docs-heading-anchor" href="#Version-1.6.0-(January-1,-2023)">Version 1.6.0 (January 1, 2023)</a><a id="Version-1.6.0-(January-1,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.6.0-(January-1,-2023)" title="Permalink"></a></h2><h3 id="Added-10"><a class="docs-heading-anchor" href="#Added-10">Added</a><a class="docs-heading-anchor-permalink" href="#Added-10" title="Permalink"></a></h3><ul><li>Added a <code>result</code> keyword argument to <a href="../api/JuMP/#JuMP.solution_summary"><code>solution_summary</code></a> to allow summarizing models with multiple solutions (#3138)</li><li>Added <a href="../api/JuMP/#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a>, which is a useful tool when debugging infeasible models (#3140)</li><li>Added <a href="../api/JuMP/#JuMP.has_start_value"><code>has_start_value</code></a> (#3157)</li><li>Added support for <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> in constraints (#3154)</li></ul><h3 id="Fixed-14"><a class="docs-heading-anchor" href="#Fixed-14">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-14" title="Permalink"></a></h3><ul><li>Fixed promotion of complex expressions (#3150) (#3164)</li></ul><h3 id="Other-14"><a class="docs-heading-anchor" href="#Other-14">Other</a><a class="docs-heading-anchor-permalink" href="#Other-14" title="Permalink"></a></h3><ul><li>Added Benders tutorial with in-place resolves (#3145)</li><li>Added more <a href="../tutorials/linear/tips_and_tricks/#linear_tips_and_tricks">Tips and tricks</a> for linear programs (#3144) (#3163)</li><li>Clarified documentation that <code>start</code> can depend on the indices of a variable container (#3148)</li><li>Replace instances of <code>length</code> and <code>size</code> by the recommended <code>eachindex</code> and <code>axes</code> (#3149)</li><li>Added a warning explaining why the model is dirty when accessing solution results from a modified model (#3156)</li><li>Clarify documentation that <code>PSD</code> ensures a symmetric matrix (#3159)</li><li>Maintenance of the JuMP test suite (#3146) (#3158) (#3162)</li></ul><h2 id="Version-1.5.0-(December-8,-2022)"><a class="docs-heading-anchor" href="#Version-1.5.0-(December-8,-2022)">Version 1.5.0 (December 8, 2022)</a><a id="Version-1.5.0-(December-8,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.5.0-(December-8,-2022)" title="Permalink"></a></h2><h3 id="Added-11"><a class="docs-heading-anchor" href="#Added-11">Added</a><a class="docs-heading-anchor-permalink" href="#Added-11" title="Permalink"></a></h3><ul><li>Add support for complex-valued variables:<ul><li><a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> (#3109)</li><li><a href="../api/JuMP/#JuMP.ComplexPlane"><code>ComplexPlane</code></a> and <a href="../api/JuMP/#JuMP.ComplexVariable"><code>ComplexVariable</code></a> (#3134)</li></ul></li><li>Add support for <a href="../moi/reference/models/#MathOptInterface.OptimizerWithAttributes"><code>MOI.OptimizerWithAttributes</code></a> in <a href="../api/JuMP/#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a> and <a href="../api/JuMP/#JuMP.get_optimizer_attribute"><code>get_optimizer_attribute</code></a> (#3129)</li></ul><h3 id="Fixed-15"><a class="docs-heading-anchor" href="#Fixed-15">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-15" title="Permalink"></a></h3><ul><li>Fixed error message for vectorized interval constraints (#3123)</li><li>Fixed passing <code>AbstractString</code> to <a href="../api/JuMP/#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a> (#3127)</li></ul><h3 id="Other-15"><a class="docs-heading-anchor" href="#Other-15">Other</a><a class="docs-heading-anchor-permalink" href="#Other-15" title="Permalink"></a></h3><ul><li>Update package versions used in docs (#3119) (#3133) (#3139)</li><li>Fixed output of diet tutorial (#3120)</li><li>Explain how to use <code>Dates.period</code> in <a href="../api/JuMP/#JuMP.set_time_limit_sec"><code>set_time_limit_sec</code></a> (#3121)</li><li>Update to <code>JuliaFormatter</code> v1.0.15 (#3130)</li><li>Fixed <code>HTTP</code> server example in web_app.jl (#3131)</li><li>Update docs to build with <code>Documenter#master</code> (#3094)</li><li>Add tests for <code>LinearAlgebra</code> operations (#3132)</li><li>Tidy these release notes (#3135)</li><li>Added documentation for <a href="../manual/complex/#Complex-number-support">Complex number support</a> (#3141)</li><li>Removed the &quot;workforce scheduling&quot; and &quot;steelT3&quot; tutorials (#3143)</li></ul><h2 id="Version-1.4.0-(October-29,-2022)"><a class="docs-heading-anchor" href="#Version-1.4.0-(October-29,-2022)">Version 1.4.0 (October 29, 2022)</a><a id="Version-1.4.0-(October-29,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.4.0-(October-29,-2022)" title="Permalink"></a></h2><h3 id="Added-12"><a class="docs-heading-anchor" href="#Added-12">Added</a><a class="docs-heading-anchor-permalink" href="#Added-12" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> which converts a container into a vector of <code>NamedTuple</code>s to support the Tables.jl interface. This simplifies converting <a href="../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> and <a href="../api/JuMP.Containers/#JuMP.Containers.SparseAxisArray"><code>Containers.SparseAxisArray</code></a> objects into tabular forms such as a DataFrame (#3104)</li><li>Added a new method to <a href="../api/JuMP.Containers/#JuMP.Containers.container"><code>Containers.container</code></a> so that index names are passed to the container (#3088)</li></ul><h3 id="Fixed-16"><a class="docs-heading-anchor" href="#Fixed-16">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-16" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>copy_to(dest::Model, src::MOI.ModelLike)</code> when <code>src</code> has nonlinear components (#3101)</li><li>Fixed the printing of <code>(-1.0 + 0.0im)</code> coefficients in complex expressions (#3112)</li><li>Fixed a parsing bug in nonlinear expressions with generator statements that contain multiple <code>for</code> statements (#3116)</li></ul><h3 id="Other-16"><a class="docs-heading-anchor" href="#Other-16">Other</a><a class="docs-heading-anchor-permalink" href="#Other-16" title="Permalink"></a></h3><ul><li>Converted the multi-commodity flow tutorial to use an SQLite database (#3098)</li><li>Fixed a number of typos in the documentation (#3103) (#3107) (#3018)</li><li>Improved various style aspects of the PDF documentation (#3095) (#3098) (#3102)</li></ul><h2 id="Version-1.3.1-(September-28,-2022)"><a class="docs-heading-anchor" href="#Version-1.3.1-(September-28,-2022)">Version 1.3.1 (September 28, 2022)</a><a id="Version-1.3.1-(September-28,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.3.1-(September-28,-2022)" title="Permalink"></a></h2><h3 id="Fixed-17"><a class="docs-heading-anchor" href="#Fixed-17">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-17" title="Permalink"></a></h3><ul><li>Fixed a performance issue in <code>relax_integrality</code> (#3087)</li><li>Fixed the type stability of operators with <code>Complex</code> arguments (#3072)</li><li>Fixed a bug which added additional <code>+()</code> terms to some nonlinear expressions (#3091)</li><li>Fixed potential method ambiguities with <code>AffExpr</code> and <code>QuadExpr</code> objects (#3092)</li></ul><h3 id="Other-17"><a class="docs-heading-anchor" href="#Other-17">Other</a><a class="docs-heading-anchor-permalink" href="#Other-17" title="Permalink"></a></h3><ul><li>Added <a href="https://vale.sh">vale</a> as a linter for the documentation (#3080)</li><li>Added a tutorial on debugging JuMP models (#3043)</li><li>Fixed a number of typos in the documentation (#3079) (#3083)</li><li>Many other small tweaks to the documentation (#3068) (#3073) (#3074) (#3075) (#3076) (#3077) (#3078) (#3081) (#3082) (#3084) (#3085) (#3089)</li></ul><h2 id="Version-1.3.0-(September-5,-2022)"><a class="docs-heading-anchor" href="#Version-1.3.0-(September-5,-2022)">Version 1.3.0 (September 5, 2022)</a><a id="Version-1.3.0-(September-5,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.3.0-(September-5,-2022)" title="Permalink"></a></h2><h3 id="Added-13"><a class="docs-heading-anchor" href="#Added-13">Added</a><a class="docs-heading-anchor-permalink" href="#Added-13" title="Permalink"></a></h3><ul><li>Support slicing in <code>SparseAxisArray</code> (#3031)</li></ul><h3 id="Fixed-18"><a class="docs-heading-anchor" href="#Fixed-18">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-18" title="Permalink"></a></h3><ul><li>Fixed a bug introduced in v1.2.0 that prevented <code>DenseAxisArray</code>s with <code>Vector</code> keys (#3064)</li></ul><h3 id="Other-18"><a class="docs-heading-anchor" href="#Other-18">Other</a><a class="docs-heading-anchor-permalink" href="#Other-18" title="Permalink"></a></h3><ul><li>Released the JuMP logos under the CC BY 4.0 license (#3063)</li><li>Minor tweaks to the documentation (#3054) (#3056) (#3057) (#3060) (#3061) (#3065)</li><li>Improved code coverage of a number of files (#3048) (#3049) (#3050) (#3051) (#3052) (#3053) (#3058) (#3059)</li></ul><h2 id="Version-1.2.1-(August-22,-2022)"><a class="docs-heading-anchor" href="#Version-1.2.1-(August-22,-2022)">Version 1.2.1 (August 22, 2022)</a><a id="Version-1.2.1-(August-22,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.2.1-(August-22,-2022)" title="Permalink"></a></h2><h3 id="Fixed-19"><a class="docs-heading-anchor" href="#Fixed-19">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-19" title="Permalink"></a></h3><ul><li>Fixed a bug when parsing two-sided nonlinear constraints (#3045)</li></ul><h2 id="Version-1.2.0-(August-16,-2022)"><a class="docs-heading-anchor" href="#Version-1.2.0-(August-16,-2022)">Version 1.2.0 (August 16, 2022)</a><a id="Version-1.2.0-(August-16,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.2.0-(August-16,-2022)" title="Permalink"></a></h2><h3 id="Breaking-2"><a class="docs-heading-anchor" href="#Breaking-2">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-2" title="Permalink"></a></h3><p>This is a large minor release because it significantly refactors the internal code for handling nonlinear programs to use the <code>MathOptInterface.Nonlinear</code> submodule that was introduced in MathOptInterface v1.3.0. As a consequence, the internal datastructure in <code>model.nlp_data</code> has been removed, as has the <code>JuMP._Derivatives</code> submodule. Despite the changes, the public API for nonlinear programming has not changed, and any code that uses only the public API and that worked with v1.1.1 will continue to work with v1.2.0.</p><h3 id="Added-14"><a class="docs-heading-anchor" href="#Added-14">Added</a><a class="docs-heading-anchor-permalink" href="#Added-14" title="Permalink"></a></h3><ul><li>Added <code>all_constraints(model; include_variable_in_set_constraints)</code> which simplifies returning a list of all constraint indices in the model.</li><li>Added the ability to delete nonlinear constraints via <code>delete(::Model, ::NonlinearConstraintRef)</code>.</li><li>Added the ability to provide an explicit Hessian for a multivariate user-defined function.</li><li>Added support for querying the primal value of a nonlinear constraint via <code>value(::NonlinearConstraintRef)</code></li></ul><h3 id="Fixed-20"><a class="docs-heading-anchor" href="#Fixed-20">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-20" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>Containers.DenseAxisArray</code> so that it now supports indexing with keys that hash to the same value, even if they are different types, for example, <code>Int32</code> and <code>Int64</code>.</li><li>Fixed a bug printing the model when the solver does not support <code>MOI.Name</code>.</li></ul><h3 id="Other-19"><a class="docs-heading-anchor" href="#Other-19">Other</a><a class="docs-heading-anchor-permalink" href="#Other-19" title="Permalink"></a></h3><ul><li>Added a constraint programming formulation to the Sudoku tutorial.</li><li>Added newly supported solvers Pajarito, Clarabel, and COPT to the installation table.</li><li>Fixed a variety of other miscellaneous issues in the documentation.</li></ul><h2 id="Version-1.1.1-(June-14,-2022)"><a class="docs-heading-anchor" href="#Version-1.1.1-(June-14,-2022)">Version 1.1.1 (June 14, 2022)</a><a id="Version-1.1.1-(June-14,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.1.1-(June-14,-2022)" title="Permalink"></a></h2><h3 id="Other-20"><a class="docs-heading-anchor" href="#Other-20">Other</a><a class="docs-heading-anchor-permalink" href="#Other-20" title="Permalink"></a></h3><ul><li>Fixed problem displaying LaTeX in the documentation</li><li>Minor updates to the style guide</li><li>Updated to MOI v1.4.0 in the documentation</li></ul><h2 id="Version-1.1.0-(May-25,-2022)"><a class="docs-heading-anchor" href="#Version-1.1.0-(May-25,-2022)">Version 1.1.0 (May 25, 2022)</a><a id="Version-1.1.0-(May-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.1.0-(May-25,-2022)" title="Permalink"></a></h2><h3 id="Added-15"><a class="docs-heading-anchor" href="#Added-15">Added</a><a class="docs-heading-anchor-permalink" href="#Added-15" title="Permalink"></a></h3><ul><li>Added <code>num_constraints(::Model; count_variable_in_set_constraints)</code> to simplify the process of counting the number of constraints in a model</li><li>Added <code>VariableRef(::ConstraintRef)</code> for querying the variable associated with a bound or integrality constraint.</li><li>Added <code>set_normalized_coefficients</code> for modifying the variable coefficients of a vector-valued constraint.</li><li>Added <code>set_string_names_on_creation</code> to disable creating <code>String</code> names for variables and constraints. This can improve performance.</li></ul><h3 id="Fixed-21"><a class="docs-heading-anchor" href="#Fixed-21">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-21" title="Permalink"></a></h3><ul><li>Fixed a bug passing <code>nothing</code> to the <code>start</code> keyword of <code>@variable</code></li></ul><h3 id="Other-21"><a class="docs-heading-anchor" href="#Other-21">Other</a><a class="docs-heading-anchor-permalink" href="#Other-21" title="Permalink"></a></h3><ul><li>New tutorials:<ul><li>Sensitivity analysis of a linear program</li><li>Serving web apps</li></ul></li><li>Minimal ellipse SDP tutorial refactored and improved</li><li>Docs updated to the latest version of each package</li><li>Lots of minor fixes and improvements to the documentation</li></ul><h2 id="Version-1.0.0-(March-24,-2022)"><a class="docs-heading-anchor" href="#Version-1.0.0-(March-24,-2022)">Version 1.0.0 (March 24, 2022)</a><a id="Version-1.0.0-(March-24,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-1.0.0-(March-24,-2022)" title="Permalink"></a></h2><p><strong>Read more about this release, along with an acknowledgement of all the contributors in our <a href="https://jump.dev/blog/1.0.0-release/">JuMP 1.0.0 is released</a> blog post.</strong></p><h3 id="Breaking-3"><a class="docs-heading-anchor" href="#Breaking-3">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-3" title="Permalink"></a></h3><ul><li>The previously deprecated functions (v0.23.0, v0.23.1) have been removed. Deprecation was to improve consistency of function names:<ul><li><code>num_nl_constraints</code> (see <code>num_nonlinear_constraints</code>)</li><li><code>all_nl_constraints</code> (see <code>all_nonlinear_constraints</code>)</li><li><code>add_NL_expression</code> (see <code>add_nonlinear_expression</code>)</li><li><code>set_NL_objective</code>  (see <code>set_nonlinear_objective</code>)</li><li><code>add_NL_constraint</code> (see <code>add_nonlinear_constraint</code>)</li><li><code>nl_expr_string</code> (see <code>nonlinear_expr_string</code>)</li><li><code>nl_constraint_string</code> (see <code>nonlinear_constraint_string</code>)</li><li><code>SymMatrixSpace</code> (see <code>SymmetricMatrixSpace</code>)</li></ul></li><li>The unintentionally exported variable <code>JuMP.op_hint</code> has been renamed to the unexported <code>JuMP._OP_HINT</code></li></ul><h3 id="Fixed-22"><a class="docs-heading-anchor" href="#Fixed-22">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-22" title="Permalink"></a></h3><ul><li>Fixed a bug writing .nl files</li><li>Fixed a bug broadcasting <code>SparseAxisArray</code>s</li></ul><h2 id="Version-0.23.2-(March-14,-2022)"><a class="docs-heading-anchor" href="#Version-0.23.2-(March-14,-2022)">Version 0.23.2 (March 14, 2022)</a><a id="Version-0.23.2-(March-14,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.23.2-(March-14,-2022)" title="Permalink"></a></h2><h3 id="Added-16"><a class="docs-heading-anchor" href="#Added-16">Added</a><a class="docs-heading-anchor-permalink" href="#Added-16" title="Permalink"></a></h3><ul><li>Added <code>relative_gap</code> to <code>solution_summary</code></li><li><code>register</code> now throws an informative error if the function is not differentiable using ForwardDiff. In some cases, the check in <code>register</code> will encounter a false negative, and the informative error will be thrown at run-time. This usually happens when the function is non-differentiable in a subset of the domain.</li></ul><h3 id="Fixed-23"><a class="docs-heading-anchor" href="#Fixed-23">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-23" title="Permalink"></a></h3><ul><li>Fixed a scoping issue when extending the <code>container</code> keyword of containers</li></ul><h3 id="Other-22"><a class="docs-heading-anchor" href="#Other-22">Other</a><a class="docs-heading-anchor-permalink" href="#Other-22" title="Permalink"></a></h3><ul><li>Docs updated to the latest version of each package</li></ul><h2 id="Version-0.23.1-(March-2,-2022)"><a class="docs-heading-anchor" href="#Version-0.23.1-(March-2,-2022)">Version 0.23.1 (March 2, 2022)</a><a id="Version-0.23.1-(March-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.23.1-(March-2,-2022)" title="Permalink"></a></h2><h3 id="Deprecated"><a class="docs-heading-anchor" href="#Deprecated">Deprecated</a><a id="Deprecated-1"></a><a class="docs-heading-anchor-permalink" href="#Deprecated" title="Permalink"></a></h3><ul><li><code>nl_expr_string</code> and <code>nl_constraint_string</code> have been renamed to <code>nonlinear_expr_string</code> and <code>nonlinear_constraint_string</code>. The old methods still exist with deprecation warnings. This change should impact very few users because to call them you must rely on private internals of the nonlinear API. Users are encouraged to use <code>sprint(show, x)</code> instead, where <code>x</code> is the nonlinear expression or constraint of interest.</li></ul><h3 id="Added-17"><a class="docs-heading-anchor" href="#Added-17">Added</a><a class="docs-heading-anchor-permalink" href="#Added-17" title="Permalink"></a></h3><ul><li>Added support for <code>Base.abs2(x)</code> where <code>x</code> is a variable or affine expression. This is mainly useful for complex-valued constraints.</li></ul><h3 id="Fixed-24"><a class="docs-heading-anchor" href="#Fixed-24">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-24" title="Permalink"></a></h3><ul><li>Fixed addition of complex and real affine expressions</li><li>Fixed arithmetic for Complex-valued quadratic expressions</li><li>Fixed variable bounds passed as <code>Rational{Int}(Inf)</code></li><li>Fixed printing of the coefficient <code>(0 + 1im)</code></li><li>Fixed a bug when <code>solution_summary</code> is called prior to <code>optimize!</code></li></ul><h2 id="Version-0.23.0-(February-25,-2022)"><a class="docs-heading-anchor" href="#Version-0.23.0-(February-25,-2022)">Version 0.23.0 (February 25, 2022)</a><a id="Version-0.23.0-(February-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.23.0-(February-25,-2022)" title="Permalink"></a></h2><p><strong>JuMP v0.23.0 is a breaking release. It is also a release-candidate for JuMP v1.0.0. That is, if no issues are found with the v0.23.0 release, then it will be re-tagged as v1.0.0.</strong></p><h3 id="Breaking-4"><a class="docs-heading-anchor" href="#Breaking-4">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-4" title="Permalink"></a></h3><ul><li>Julia 1.6 is now the minimum supported version</li><li>MathOptInterface has been updated to v1.0.0</li><li>All previously deprecated functionality has been removed</li><li><code>PrintMode</code>, <code>REPLMode</code> and <code>IJuliaMode</code> have been removed in favor of the MIME types <code>MIME&quot;text/plain&quot;</code> and <code>MIME&quot;text/latex&quot;</code>. Replace instances of <code>::Type{REPLMode}</code> with <code>::MIME&quot;text/plain&quot;</code>, <code>REPLMode</code> with <code>MIME(&quot;text/plain&quot;)</code>, <code>::Type{IJuliaMode}</code> with <code>::MIME&quot;text/latex&quot;</code>, and <code>IJuliaMode</code> with <code>MIME(&quot;text/latex&quot;)</code>.</li><li>Functions containing the <code>nl_</code> acronym have been renamed to the more explicit <code>nonlinear_</code>. For example, <code>num_nl_constraints</code> is now <code>num_nonlinear_constraints</code> and <code>set_NL_objective</code> is now <code>set_nonlinear_objective</code>. Calls to the old functions throw an error explaining the new name.</li><li><code>SymMatrixSpace</code> has been renamed to <code>SymmetricMatrixSpace</code></li></ul><h3 id="Added-18"><a class="docs-heading-anchor" href="#Added-18">Added</a><a class="docs-heading-anchor-permalink" href="#Added-18" title="Permalink"></a></h3><ul><li>Added <code>nonlinear_dual_start_value</code> and <code>set_nonlinear_dual_start_value</code></li><li>Added preliminary support for <code>Complex</code> coefficient types</li></ul><h3 id="Fixed-25"><a class="docs-heading-anchor" href="#Fixed-25">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-25" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>solution_summary</code></li></ul><h3 id="Other-23"><a class="docs-heading-anchor" href="#Other-23">Other</a><a class="docs-heading-anchor-permalink" href="#Other-23" title="Permalink"></a></h3><ul><li>MILP examples have been migrated from GLPK to HiGHS</li><li>Fixed various typos</li><li>Improved section on setting constraint start values</li></ul><h3 id="Troubleshooting-problems-when-updating"><a class="docs-heading-anchor" href="#Troubleshooting-problems-when-updating">Troubleshooting problems when updating</a><a id="Troubleshooting-problems-when-updating-1"></a><a class="docs-heading-anchor-permalink" href="#Troubleshooting-problems-when-updating" title="Permalink"></a></h3><p>If you experience problems when updating, you are likely using previously deprecated functionality. (By default, Julia does not warn when you use deprecated features.)</p><p>To find the deprecated features you are using, start Julia with <code>--depwarn=yes</code>:</p><pre><code class="nohighlight hljs">$ julia --depwarn=yes</code></pre><p>Then install JuMP v0.22.3:</p><pre><code class="language-julia hljs">julia&gt; using Pkg
+julia&gt; pkg&quot;add JuMP@0.22.3&quot;</code></pre><p>And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.</p><h2 id="Version-0.22.3-(February-10,-2022)"><a class="docs-heading-anchor" href="#Version-0.22.3-(February-10,-2022)">Version 0.22.3 (February 10, 2022)</a><a id="Version-0.22.3-(February-10,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.22.3-(February-10,-2022)" title="Permalink"></a></h2><h3 id="Fixed-26"><a class="docs-heading-anchor" href="#Fixed-26">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-26" title="Permalink"></a></h3><ul><li>Fixed a reproducibility issue in the TSP tutorial</li><li>Fixed a reproducibility issue in the <code>max_cut_sdp</code> tutorial</li><li>Fixed a bug broadcasting an empty SparseAxisArray</li></ul><h3 id="Other-24"><a class="docs-heading-anchor" href="#Other-24">Other</a><a class="docs-heading-anchor-permalink" href="#Other-24" title="Permalink"></a></h3><ul><li>Added a warning and improved documentation for the modify-then-query case</li><li>Fixed a typo in the docstring of <code>RotatedSecondOrderCone</code></li><li>Added Aqua.jl as a check for code health</li><li>Added introductions to each section of the tutorials</li><li>Improved the column generation and Benders decomposition tutorials</li><li>Updated documentation to MOI v0.10.8</li><li>Updated JuliaFormatter to v0.22.2</li></ul><h2 id="Version-0.22.2-(January-10,-2022)"><a class="docs-heading-anchor" href="#Version-0.22.2-(January-10,-2022)">Version 0.22.2 (January 10, 2022)</a><a id="Version-0.22.2-(January-10,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.22.2-(January-10,-2022)" title="Permalink"></a></h2><h3 id="Added-19"><a class="docs-heading-anchor" href="#Added-19">Added</a><a class="docs-heading-anchor-permalink" href="#Added-19" title="Permalink"></a></h3><ul><li>The function <code>all_nl_constraints</code> now returns all nonlinear constraints in a model</li><li><code>start_value</code> and <code>set_start_value</code> can now be used to get and set the primal start for constraint references</li><li>Plural macros now return a tuple containing the elements that were defined instead of <code>nothing</code></li><li>Anonymous variables are now printed as <code>_[i]</code> where <code>i</code> is the index of the variable instead of <code>noname</code>. Calling <code>name(x)</code> still returns <code>&quot;&quot;</code> so this is non-breaking.</li></ul><h3 id="Fixed-27"><a class="docs-heading-anchor" href="#Fixed-27">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-27" title="Permalink"></a></h3><ul><li>Fixed handling of <code>min</code> and <code>max</code> in nonlinear expressions</li><li>CartesianIndex is no longer allowed as a key for DenseAxisArrays.</li></ul><h3 id="Other-25"><a class="docs-heading-anchor" href="#Other-25">Other</a><a class="docs-heading-anchor-permalink" href="#Other-25" title="Permalink"></a></h3><ul><li>Improved the performance of GenericAffExpr</li><li>Added a tutorial on the Travelling Salesperson Problem</li><li>Added a tutorial on querying the Hessian of a nonlinear program</li><li>Added documentation on using custom solver binaries.</li></ul><h2 id="Version-0.22.1-(November-29,-2021)"><a class="docs-heading-anchor" href="#Version-0.22.1-(November-29,-2021)">Version 0.22.1 (November 29, 2021)</a><a id="Version-0.22.1-(November-29,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.22.1-(November-29,-2021)" title="Permalink"></a></h2><h3 id="Added-20"><a class="docs-heading-anchor" href="#Added-20">Added</a><a class="docs-heading-anchor-permalink" href="#Added-20" title="Permalink"></a></h3><ul><li>Export <code>OptimizationSense</code> enum, with instances: <code>MIN_SENSE</code>, <code>MAX_SENSE</code>, and <code>FEASIBILITY_SENSE</code></li><li>Add <code>Base.isempty(::Model)</code> to match <code>Base.empty(::Model)</code></li></ul><h3 id="Fixed-28"><a class="docs-heading-anchor" href="#Fixed-28">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-28" title="Permalink"></a></h3><ul><li>Fix bug in container with tuples as indices</li><li>Fix bug in <code>set_time_limit_sec</code></li></ul><h3 id="Other-26"><a class="docs-heading-anchor" href="#Other-26">Other</a><a class="docs-heading-anchor-permalink" href="#Other-26" title="Permalink"></a></h3><ul><li>Add tutorial &quot;Design patterns for larger models&quot;</li><li>Remove release notes section from PDF</li><li>General edits of the documentation and error messages</li></ul><h2 id="Version-0.22.0-(November-10,-2021)"><a class="docs-heading-anchor" href="#Version-0.22.0-(November-10,-2021)">Version 0.22.0 (November 10, 2021)</a><a id="Version-0.22.0-(November-10,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.22.0-(November-10,-2021)" title="Permalink"></a></h2><p><strong>JuMP v0.22 is a breaking release</strong></p><h3 id="Breaking-5"><a class="docs-heading-anchor" href="#Breaking-5">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-5" title="Permalink"></a></h3><p>JuMP 0.22 contains a number of breaking changes. However, these should be invisible for the majority of users. You will mostly encounter these breaking changes if you: wrote a JuMP extension, accessed <code>backend(model)</code>, or called <code>@SDconstraint</code>.</p><p>The breaking changes are as follows:</p><ul><li>MathOptInterface has been updated to v0.10.4. For users who have interacted with the MOI backend, this contains a large number of breaking changes. Read the <a href="https://jump.dev/MathOptInterface.jl/v0.10/release_notes/#v0.10.0-(September-6,-2021)">MathOptInterface release notes</a> for more details.</li><li>The <code>bridge_constraints</code> keyword argument to <code>Model</code> and <code>set_optimizer</code> has been renamed <code>add_bridges</code> to reflect that more thing were bridged than just constraints.</li><li>The <code>backend(model)</code> field now contains a concrete instance of a <code>MOI.Utilities.CachingOptimizer</code> instead of one with an abstractly typed optimizer field. In most cases, this will lead to improved performance. However, calling <code>set_optimizer</code> after <code>backend</code> invalidates the old backend. For example:<pre><code class="language-julia hljs">model = Model()
+b = backend(model)
+set_optimizer(model, GLPK.Optimizer)
+@variable(model, x)
+# b is not updated with `x`! Get a new b by calling `backend` again.
+new_b = backend(model)</code></pre></li><li>All usages of <code>@SDconstraint</code> are deprecated. The new syntax is <code>@constraint(model, X &gt;= Y, PSDCone())</code>.</li><li>Creating a <code>DenseAxisArray</code> with a <code>Number</code> as an axis will now display a warning. This catches a common error in which users write <code>@variable(model, x[length(S)])</code> instead of <code>@variable(model, x[1:length(S)])</code>.</li><li>The <code>caching_mode</code> argument to <code>Model</code>, for example, <code>Model(caching_mode = MOIU.MANUAL)</code> mode has been removed. For more control over the optimizer, use <code>direct_model</code> instead.</li><li>The previously deprecated <code>lp_objective_perturbation_range</code> and <code>lp_rhs_perturbation_range</code> functions have been removed. Use <code>lp_sensitivity_report</code> instead.</li><li>The <code>.m</code> fields of <code>NonlinearExpression</code> and <code>NonlinearParameter</code> have been renamed to <code>.model</code>.</li><li>Infinite variable bounds are now ignored. Thus, <code>@variable(model, x &lt;= Inf)</code> will show <code>has_upper_bound(x) == false</code>. Previously, these bounds were passed through to the solvers which caused numerical issues for solvers expecting finite bounds.</li><li>The <code>variable_type</code> and <code>constraint_type</code> functions were removed. This should only affect users who previously wrote JuMP extensions. The functions can be deleted without consequence.</li><li>The internal functions <code>moi_mode</code>, <code>moi_bridge_constraints</code>, <code>moi_add_constraint</code>, and <code>moi_add_to_function_constant</code> are no longer exported.</li><li>The un-used method <code>Containers.generate_container</code> has been deleted.</li><li>The <code>Containers</code> API has been refactored, and <code>_build_ref_sets</code> is now public as <code>Containers.build_ref_sets</code>.</li><li>The <code>parse_constraint_</code> methods for extending <code>@constraint</code> at parse time have been refactored in a breaking way. Consult the Extensions documentation for more details and examples.</li></ul><h3 id="Added-21"><a class="docs-heading-anchor" href="#Added-21">Added</a><a class="docs-heading-anchor-permalink" href="#Added-21" title="Permalink"></a></h3><ul><li>The <code>TerminationStatusCode</code> and <code>ResultStatusCode</code> enums are now exported by JuMP. Prefer <code>termination_status(model) == OPTIMAL</code> instead of <code>== MOI.OPTIMAL</code>, although the <code>MOI.</code> prefix way still works.</li><li>Copy a <code>x::DenseAxisArray</code> to an <code>Array</code> by calling <code>Array(x)</code>.</li><li><code>NonlinearExpression</code> is now a subtype of <code>AbstractJuMPScalar</code></li><li>Constraints such as <code>@constraint(model, x + 1 in MOI.Integer())</code> are now supported.</li><li><code>primal_feasibility_report</code> now accepts a function as the first argument.</li><li>Scalar variables <code>@variable(model, x[1:2] in MOI.Integer())</code> creates two variables, both of which are constrained to be in the set <code>MOI.Integer</code>.</li><li>Conic constraints can now be specified as inequalities under a different partial ordering. So <code>@constraint(model, x - y in MOI.Nonnegatives())</code> can now be written as <code>@constraint(model, x &gt;= y, MOI.Nonnegatives())</code>.</li><li>Names are now set for vectorized constraints.</li></ul><h3 id="Fixed-29"><a class="docs-heading-anchor" href="#Fixed-29">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-29" title="Permalink"></a></h3><ul><li>Fixed a performance issue when <code>show</code> was called on a <code>SparseAxisArray</code> with a large number of elements.</li><li>Fixed a bug displaying barrier and simplex iterations in <code>solution_summary</code>.</li><li>Fixed a bug by implementing <code>hash</code> for <code>DenseAxisArray</code> and <code>SparseAxisArray</code>.</li><li>Names are now only set if the solver supports them. Previously, this prevented solvers such as Ipopt from being used with <code>direct_model</code>.</li><li><code>MutableArithmetics.Zero</code> is converted into a <code>0.0</code> before being returned to the user. Previously, some calls to <code>@expression</code> would return the undocumented <code>MutableArithmetics.Zero()</code> object. One example is summing over an empty set <code>@expression(model, sum(x[i] for i in 1:0))</code>. You will now get <code>0.0</code> instead.</li><li><code>AffExpr</code> and <code>QuadExpr</code> can now be used with <code>== 0</code> instead of <code>iszero</code>. This fixes a number of issues relating to Julia standard libraries such as <code>LinearAlgebra</code> and <code>SparseArrays</code>.</li><li>Fixed a bug when registering a user-defined function with splatting.</li></ul><h3 id="Other-27"><a class="docs-heading-anchor" href="#Other-27">Other</a><a class="docs-heading-anchor-permalink" href="#Other-27" title="Permalink"></a></h3><ul><li>The documentation is now available as a PDF.</li><li>The documentation now includes a full copy of the MathOptInterface documentation to make it easy to link concepts between the docs. (The MathOptInterface documentation has also been significantly improved.)</li><li>The documentation contains a large number of improvements and clarifications on a range of topics. Thanks to @sshin23, @DilumAluthge, and @jlwether.</li><li>The documentation is now built with Julia 1.6 instead of 1.0.</li><li>Various error messages have been improved to be more readable.</li></ul><h2 id="Version-0.21.10-(September-4,-2021)"><a class="docs-heading-anchor" href="#Version-0.21.10-(September-4,-2021)">Version 0.21.10 (September 4, 2021)</a><a id="Version-0.21.10-(September-4,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.10-(September-4,-2021)" title="Permalink"></a></h2><h3 id="Added-22"><a class="docs-heading-anchor" href="#Added-22">Added</a><a class="docs-heading-anchor-permalink" href="#Added-22" title="Permalink"></a></h3><ul><li>Added <code>add_NL_expression</code></li><li><code>add_NL_xxx</code> functions now support <code>AffExpr</code> and <code>QuadExpr</code> as terms</li></ul><h3 id="Fixed-30"><a class="docs-heading-anchor" href="#Fixed-30">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-30" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>solution_summary</code></li><li>Fixed a bug in <code>relax_integrality</code></li></ul><h3 id="Other-28"><a class="docs-heading-anchor" href="#Other-28">Other</a><a class="docs-heading-anchor-permalink" href="#Other-28" title="Permalink"></a></h3><ul><li>Improved error message in <code>lp_sensitivity_report</code></li></ul><h2 id="Version-0.21.9-(August-1,-2021)"><a class="docs-heading-anchor" href="#Version-0.21.9-(August-1,-2021)">Version 0.21.9 (August 1, 2021)</a><a id="Version-0.21.9-(August-1,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.9-(August-1,-2021)" title="Permalink"></a></h2><h3 id="Added-23"><a class="docs-heading-anchor" href="#Added-23">Added</a><a class="docs-heading-anchor-permalink" href="#Added-23" title="Permalink"></a></h3><ul><li>Containers now support arbitrary container types by passing the type to the <code>container</code> keyword and overloading <code>Containers.container</code>.</li><li><code>is_valid</code> now supports nonlinear constraints</li><li>Added <code>unsafe_backend</code> for querying the inner-most optimizer of a JuMP model.</li><li>Nonlinear parameters now support the plural <code>@NLparameters</code> macro.</li><li>Containers (for example, <code>DenseAxisArray</code>) can now be used in vector-valued constraints.</li></ul><h3 id="Other-29"><a class="docs-heading-anchor" href="#Other-29">Other</a><a class="docs-heading-anchor-permalink" href="#Other-29" title="Permalink"></a></h3><ul><li>Various improvements to the documentation.</li></ul><h2 id="Version-0.21.8-(May-8,-2021)"><a class="docs-heading-anchor" href="#Version-0.21.8-(May-8,-2021)">Version 0.21.8 (May 8, 2021)</a><a id="Version-0.21.8-(May-8,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.8-(May-8,-2021)" title="Permalink"></a></h2><h3 id="Added-24"><a class="docs-heading-anchor" href="#Added-24">Added</a><a class="docs-heading-anchor-permalink" href="#Added-24" title="Permalink"></a></h3><ul><li>The <code>@constraint</code> macro is now extendable in the same way as <code>@variable</code>.</li><li><code>AffExpr</code> and <code>QuadExpr</code> can now be used in nonlinear macros.</li></ul><h3 id="Fixed-31"><a class="docs-heading-anchor" href="#Fixed-31">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-31" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>lp_sensitivity_report</code>.</li><li>Fixed an inference issue when creating empty <code>SparseAxisArray</code>s.</li></ul><h2 id="Version-0.21.7-(April-12,-2021)"><a class="docs-heading-anchor" href="#Version-0.21.7-(April-12,-2021)">Version 0.21.7 (April 12, 2021)</a><a id="Version-0.21.7-(April-12,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.7-(April-12,-2021)" title="Permalink"></a></h2><h3 id="Added-25"><a class="docs-heading-anchor" href="#Added-25">Added</a><a class="docs-heading-anchor-permalink" href="#Added-25" title="Permalink"></a></h3><ul><li>Added <code>primal_feasibility_report</code>, which can be used to check whether a primal point satisfies primal feasibility.</li><li>Added <code>coefficient</code>, which returns the coefficient associated with a variable in affine and quadratic expressions.</li><li>Added <code>copy_conflict</code>, which returns the IIS of an infeasible model.</li><li>Added <code>solution_summary</code>, which returns (and prints) a struct containing a summary of the solution.</li><li>Allow <code>AbstractVector</code> in vector constraints instead of just <code>Vector</code>.</li><li>Added <code>latex_formulation(model)</code> which returns an object representing the latex formulation of a model. Use <code>print(latex_formulation(model))</code> to print the formulation as a string.</li><li>User-defined functions in nonlinear expressions are now automatically registered to aid quick model prototyping. However, a warning is printed to encourage the manual registration.</li><li>DenseAxisArray&#39;s now support broadcasting over multiple arrays.</li><li>Container indices can now be iterators of <code>Base.SizeUnknown</code>.</li></ul><h3 id="Fixed-32"><a class="docs-heading-anchor" href="#Fixed-32">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-32" title="Permalink"></a></h3><ul><li>Fixed bug in <code>rad2deg</code> and <code>deg2rad</code> in nonlinear expressions.</li><li>Fixed a MethodError bug in <code>Containers</code> when forcing container type.</li><li>Allow partial slicing of a DenseAxisArray, resolving an issue from 2014.</li><li>Fixed a bug printing variable names in IJulia.</li><li>Ending an IJulia cell with <code>model</code> now prints a summary of the model (like in the REPL) not the latex formulation. Use <code>print(model)</code> to print the latex formulation.</li><li>Fixed a bug when copying models containing nested arrays.</li></ul><h3 id="Other-30"><a class="docs-heading-anchor" href="#Other-30">Other</a><a class="docs-heading-anchor-permalink" href="#Other-30" title="Permalink"></a></h3><ul><li>Tutorials are now part of the documentation, and more refactoring has taken place.</li><li>Added JuliaFormatter added as a code formatter.</li><li>Added some precompilation statements to reduce initial latency.</li><li>Various improvements to error messages to make them more helpful.</li><li>Improved performance of <code>value(::NonlinearExpression)</code>.</li><li>Improved performance of <code>fix(::VariableRef)</code>.</li></ul><h2 id="Version-0.21.6-(January-29,-2021)"><a class="docs-heading-anchor" href="#Version-0.21.6-(January-29,-2021)">Version 0.21.6 (January 29, 2021)</a><a id="Version-0.21.6-(January-29,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.6-(January-29,-2021)" title="Permalink"></a></h2><h3 id="Added-26"><a class="docs-heading-anchor" href="#Added-26">Added</a><a class="docs-heading-anchor-permalink" href="#Added-26" title="Permalink"></a></h3><ul><li>Added support for skew symmetric variables via <code>@variable(model, X[1:2, 1:2] in SkewSymmetricMatrixSpace())</code>.</li><li><code>lp_sensitivity_report</code> has been added which significantly improves the performance of querying the sensitivity summary of an LP. <code>lp_objective_perturbation_range</code> and <code>lp_rhs_perturbation_range</code> are deprecated.</li><li>Dual warm-starts are now supported with <code>set_dual_start_value</code> and <code>dual_start_value</code>.</li><li><code>∈</code> (<code>\in&lt;tab&gt;</code>) can now be used in macros instead of <code>=</code> or <code>in</code>.</li><li>Use <code>haskey(model::Model, key::Symbol)</code> to check if a name <code>key</code> is registered in a model.</li><li>Added <code>unregister(model::Model, key::Symbol)</code> to unregister a name <code>key</code> from <code>model</code>.</li><li>Added <code>callback_node_status</code> for use in callbacks.</li><li>Added <code>print_bridge_graph</code> to visualize the bridging graph generated by MathOptInterface.</li><li>Improved error message for containers with duplicate indices.</li></ul><h3 id="Fixed-33"><a class="docs-heading-anchor" href="#Fixed-33">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-33" title="Permalink"></a></h3><ul><li>Various fixes to pass tests on Julia 1.6.</li><li>Fixed a bug in the printing of nonlinear expressions in IJulia.</li><li>Fixed a bug when nonlinear expressions are passed to user-defined functions.</li><li>Some internal functions that were previously exported are now no longer exported.</li><li>Fixed a bug when relaxing a fixed binary variable.</li><li>Fixed a <code>StackOverflowError</code> that occurred when <code>SparseAxisArray</code>s had a large number of elements.</li><li>Removed an unnecessary type assertion in <code>list_of_constraint_types</code>.</li><li>Fixed a bug when copying models with registered expressions.</li></ul><h3 id="Other-31"><a class="docs-heading-anchor" href="#Other-31">Other</a><a class="docs-heading-anchor-permalink" href="#Other-31" title="Permalink"></a></h3><ul><li>The documentation has been significantly overhauled. It now has distinct sections for the manual, API reference, and examples. The existing examples in <code>/examples</code> have now been moved to <code>/docs/src/examples</code> and rewritten using <code>Literate.jl</code>, and they are now included in the documentation.</li><li>JuliaFormatter has been applied to most of the codebase. This will continue to roll out over time, as we fix upstream issues in the formatter, and will eventually become compulsory.</li><li>The root cause of a large number of method invalidations has been resolved.</li><li>We switched continuous integration from Travis and Appveyor to GitHub Actions.</li></ul><h2 id="Version-0.21.5-(September-18,-2020)"><a class="docs-heading-anchor" href="#Version-0.21.5-(September-18,-2020)">Version 0.21.5 (September 18, 2020)</a><a id="Version-0.21.5-(September-18,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.5-(September-18,-2020)" title="Permalink"></a></h2><h3 id="Fixed-34"><a class="docs-heading-anchor" href="#Fixed-34">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-34" title="Permalink"></a></h3><ul><li>Fixed deprecation warnings</li><li>Throw <code>DimensionMismatch</code> for incompatibly sized functions and sets</li><li>Unify treatment of <code>keys(x)</code> on JuMP containers</li></ul><h2 id="Version-0.21.4-(September-14,-2020)"><a class="docs-heading-anchor" href="#Version-0.21.4-(September-14,-2020)">Version 0.21.4 (September 14, 2020)</a><a id="Version-0.21.4-(September-14,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.4-(September-14,-2020)" title="Permalink"></a></h2><h3 id="Added-27"><a class="docs-heading-anchor" href="#Added-27">Added</a><a class="docs-heading-anchor-permalink" href="#Added-27" title="Permalink"></a></h3><ul><li>Add debug info when adding unsupported constraints</li><li>Add <code>relax_integrality</code> for solving continuous relaxation</li><li>Allow querying constraint conflicts</li></ul><h3 id="Fixed-35"><a class="docs-heading-anchor" href="#Fixed-35">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-35" title="Permalink"></a></h3><ul><li>Dispatch on <code>Real</code> for <code>MOI.submit</code></li><li>Implement <code>copy</code> for <code>CustomSet</code> in tests</li><li>Don&#39;t export private macros</li><li>Fix invalid assertion in nonlinear</li><li>Error if constraint has <code>NaN</code> right-hand side</li><li>Improve speed of tests</li><li>Lots of work modularizing files in <code>/test</code></li><li>Improve line numbers in macro error messages</li><li>Print nonlinear subexpressions</li><li>Various documentation updates</li><li>Dependency updates:<ul><li>Datastructures 0.18</li><li>MathOptFormat v0.5</li><li>Prep for MathOptInterface 0.9.15</li></ul></li></ul><h2 id="Version-0.21.3-(June-18,-2020)"><a class="docs-heading-anchor" href="#Version-0.21.3-(June-18,-2020)">Version 0.21.3 (June 18, 2020)</a><a id="Version-0.21.3-(June-18,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.3-(June-18,-2020)" title="Permalink"></a></h2><ul><li>Added Special Order Sets (SOS1 and SOS2) to JuMP with default weights to ease the creation of such constraints (#2212).</li><li>Added functions <code>simplex_iterations</code>, <code>barrier_iterations</code> and <code>node_count</code> (#2201).</li><li>Added function <code>reduced_cost</code> (#2205).</li><li>Implemented <code>callback_value</code> for affine and quadratic expressions (#2231).</li><li>Support <code>MutableArithmetics.Zero</code> in objective and constraints (#2219).</li><li>Documentation improvements:<ul><li>Mention tutorials in the docs (#2223).</li><li>Update COIN-OR links (#2242).</li><li>Explicit link to the documentation of <code>MOI.FileFormats</code> (#2253).</li><li>Typo fixes (#2261).</li></ul></li><li>Containers improvements:<ul><li>Fix <code>Base.map</code> for <code>DenseAxisArray</code> (#2235).</li><li>Throw <code>BoundsError</code> if number of indices is incorrect for <code>DenseAxisArray</code> and <code>SparseAxisArray</code> (#2240).</li></ul></li><li>Extensibility improvements:<ul><li>Implement a <code>set_objective</code> method fallback that redirects to <code>set_objective_sense</code> and <code>set_objective_function</code> (#2247).</li><li>Add <code>parse_constraint</code> method with arbitrary number of arguments (#2051).</li><li>Add <code>parse_constraint_expr</code> and <code>parse_constraint_head</code> (#2228).</li></ul></li></ul><h2 id="Version-0.21.2-(April-2,-2020)"><a class="docs-heading-anchor" href="#Version-0.21.2-(April-2,-2020)">Version 0.21.2 (April 2, 2020)</a><a id="Version-0.21.2-(April-2,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.2-(April-2,-2020)" title="Permalink"></a></h2><ul><li>Added <code>relative_gap()</code> to access <code>MOI.RelativeGap()</code> attribute (#2199).</li><li>Documentation fixes:<ul><li>Added link to source for docstrings in the documentation (#2207).</li><li>Added docstring for <code>@variables</code> macro (#2216).</li><li>Typo fixes (#2177, #2184, #2182).</li></ul></li><li>Implementation of methods for Base functions:<ul><li>Implemented <code>Base.empty!</code> for <code>JuMP.Model</code> (#2198).</li><li>Implemented <code>Base.conj</code> for JuMP scalar types (#2209).</li></ul></li></ul><h3 id="Fixed-36"><a class="docs-heading-anchor" href="#Fixed-36">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-36" title="Permalink"></a></h3><ul><li>Fixed sum of expression with scalar product in macro (#2178).</li><li>Fixed writing of nonlinear models to MathOptFormat (#2181).</li><li>Fixed construction of empty SparseAxisArray (#2179).</li><li>Fixed constraint with zero function (#2188).</li></ul><h2 id="Version-0.21.1-(Feb-18,-2020)"><a class="docs-heading-anchor" href="#Version-0.21.1-(Feb-18,-2020)">Version 0.21.1 (Feb 18, 2020)</a><a id="Version-0.21.1-(Feb-18,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.1-(Feb-18,-2020)" title="Permalink"></a></h2><ul><li>Improved the clarity of the <code>with_optimizer</code> deprecation warning.</li></ul><h2 id="Version-0.21.0-(Feb-16,-2020)"><a class="docs-heading-anchor" href="#Version-0.21.0-(Feb-16,-2020)">Version 0.21.0 (Feb 16, 2020)</a><a id="Version-0.21.0-(Feb-16,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.21.0-(Feb-16,-2020)" title="Permalink"></a></h2><h3 id="Breaking-6"><a class="docs-heading-anchor" href="#Breaking-6">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-6" title="Permalink"></a></h3><ul><li><p>Deprecated <code>with_optimizer</code> (#2090, #2084, #2141). You can replace <code>with_optimizer</code> by either nothing, <code>optimizer_with_attributes</code> or a closure:</p><ul><li>replace <code>with_optimizer(Ipopt.Optimizer)</code> by <code>Ipopt.Optimizer</code>.</li><li>replace <code>with_optimizer(Ipopt.Optimizer, max_cpu_time=60.0)</code> by <code>optimizer_with_attributes(Ipopt.Optimizer, &quot;max_cpu_time&quot; =&gt; 60.0)</code>.</li><li>replace <code>with_optimizer(Gurobi.Optimizer, env)</code> by <code>() -&gt; Gurobi.Optimizer(env)</code>.</li><li>replace <code>with_optimizer(Gurobi.Optimizer, env, Presolve=0)</code> by <code>optimizer_with_attributes(() -&gt; Gurobi.Optimizer(env), &quot;Presolve&quot; =&gt; 0)</code>.</li></ul><p>alternatively to <code>optimizer_with_attributes</code>, you can also set the attributes separately with <code>set_optimizer_attribute</code>.</p></li><li><p>Renamed <code>set_parameter</code> and <code>set_parameters</code> to <code>set_optimizer_attribute</code> and <code>set_optimizer_attributes</code> (#2150).</p></li><li><p>Broadcast should now be explicit inside macros. <code>@SDconstraint(model, x &gt;= 1)</code> and <code>@constraint(model, x + 1 in SecondOrderCone())</code> now throw an error instead of broadcasting <code>1</code> along the dimension of <code>x</code> (#2107).</p></li><li><p><code>@SDconstraint(model, x &gt;= 0)</code> is now equivalent to <code>@constraint(model, x in PSDCone())</code> instead of <code>@constraint(model, (x .- 0) in PSDCone())</code> (#2107).</p></li><li><p>The macros now create the containers with <code>map</code> instead of <code>for</code> loops, as a consequence, containers created by <code>@expression</code> can now have any element type and containers of constraint references now have concrete element types when possible. This fixes a long-standing issue where <code>@expression</code> could only be used to generate a collection of linear expressions. Now it works for quadratic expressions as well (#2070).</p></li><li><p>Calling <code>deepcopy(::AbstractModel)</code> now throws an error.</p></li><li><p>The constraint name is now printed in the model string (#2108).</p></li></ul><h3 id="Added-28"><a class="docs-heading-anchor" href="#Added-28">Added</a><a class="docs-heading-anchor-permalink" href="#Added-28" title="Permalink"></a></h3><ul><li>Added support for solver-independent and solver-specific callbacks (#2101).</li><li>Added <code>write_to_file</code> and <code>read_from_file</code>, supported formats are CBF, LP, MathOptFormat, MPS and SDPA (#2114).</li><li>Added support for complementarity constraints (#2132).</li><li>Added support for indicator constraints (#2092).</li><li>Added support for querying multiple solutions with the <code>result</code> keyword (#2100).</li><li>Added support for constraining variables on creation (#2128).</li><li>Added method <code>delete</code> that deletes a vector of variables at once if it is supported by the underlying solver (#2135).</li><li>The arithmetic between JuMP expression has be refactored into the MutableArithmetics package (#2107).</li><li>Improved error on complex values in NLP (#1978).</li><li>Added an example of column generation (#2010).</li></ul><h3 id="Fixed-37"><a class="docs-heading-anchor" href="#Fixed-37">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-37" title="Permalink"></a></h3><ul><li>Incorrect coefficients generated when using Symmetric variables (#2102)</li></ul><h2 id="Version-0.20.1-(Oct-18,-2019)"><a class="docs-heading-anchor" href="#Version-0.20.1-(Oct-18,-2019)">Version 0.20.1 (Oct 18, 2019)</a><a id="Version-0.20.1-(Oct-18,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.20.1-(Oct-18,-2019)" title="Permalink"></a></h2><ul><li>Add sections on <code>@variables</code> and <code>@constraints</code> in the documentation (#2062).</li><li>Fixed product of sparse matrices for Julia v1.3 (#2063).</li><li>Added <code>set_objective_coefficient</code> to modify the coefficient of a linear term of the objective function (#2008).</li><li>Added <code>set_time_limit_sec</code>, <code>unset_time_limit_sec</code> and <code>time_limit_sec</code> to set and query the time limit for the solver in seconds (#2053).</li></ul><h2 id="Version-0.20.0-(Aug-24,-2019)"><a class="docs-heading-anchor" href="#Version-0.20.0-(Aug-24,-2019)">Version 0.20.0 (Aug 24, 2019)</a><a id="Version-0.20.0-(Aug-24,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.20.0-(Aug-24,-2019)" title="Permalink"></a></h2><ul><li>Documentation updates.</li><li>Numerous bug fixes.</li><li>Better error messages (#1977, #1978, #1997, #2017).</li><li>Performance improvements (#1947, #2032).</li><li>Added LP sensitivity summary functions <code>lp_objective_perturbation_range</code> and <code>lp_rhs_perturbation_range</code> (#1917).</li><li>Added functions <code>dual_objective_value</code>, <code>raw_status</code> and <code>set_parameter</code>.</li><li>Added function <code>set_objective_coefficient</code> to modify the coefficient of a linear term of the objective (#2008).</li><li>Added functions <code>set_normalized_rhs</code>, <code>normalized_rhs</code>, and <code>add_to_function_constant</code> to modify and get the constant part of a constraint (#1935, #1960).</li><li>Added functions <code>set_normalized_coefficient</code> and <code>normalized_coefficient</code> to modify and get the coefficient of a linear term of a constraint (#1935, #1960).</li><li>Numerous other improvements in MOI 0.9, see the <code>NEWS.md</code> file of MOI for more details.</li></ul><h2 id="Version-0.19.2-(June-8,-2019)"><a class="docs-heading-anchor" href="#Version-0.19.2-(June-8,-2019)">Version 0.19.2 (June 8, 2019)</a><a id="Version-0.19.2-(June-8,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.19.2-(June-8,-2019)" title="Permalink"></a></h2><ul><li>Fix a bug in derivatives that could arise in models with nested nonlinear subexpressions.</li></ul><h2 id="Version-0.19.1-(May-12,-2019)"><a class="docs-heading-anchor" href="#Version-0.19.1-(May-12,-2019)">Version 0.19.1 (May 12, 2019)</a><a id="Version-0.19.1-(May-12,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.19.1-(May-12,-2019)" title="Permalink"></a></h2><ul><li>Usability and performance improvements.</li><li>Bug fixes.</li></ul><h2 id="Version-0.19.0-(February-15,-2019)"><a class="docs-heading-anchor" href="#Version-0.19.0-(February-15,-2019)">Version 0.19.0 (February 15, 2019)</a><a id="Version-0.19.0-(February-15,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.19.0-(February-15,-2019)" title="Permalink"></a></h2><p><strong>JuMP 0.19 contains significant breaking changes.</strong></p><h3 id="Breaking-7"><a class="docs-heading-anchor" href="#Breaking-7">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-7" title="Permalink"></a></h3><ul><li><p>JuMP&#39;s abstraction layer for communicating with solvers changed from <a href="https://github.com/JuliaOpt/MathProgBase.jl">MathProgBase</a> (MPB) to <a href="https://github.com/JuliaOpt/MathOptInterface.jl">MathOptInterface</a> (MOI). MOI addresses many longstanding design issues. (See @mlubin&#39;s <a href="https://www.juliaopt.org/meetings/bordeaux2018/lubin.pdf">slides</a> from JuMP-dev 2018.) JuMP 0.19 is compatible only with solvers that have been updated for MOI. See the <a href="https://www.juliaopt.org/JuMP.jl/dev/installation/">installation guide</a> for a list of solvers that have and have not yet been updated.</p></li><li><p>Most solvers have been renamed to <code>PackageName.Optimizer</code>. For example, <code>GurobiSolver()</code> is now <code>Gurobi.Optimizer</code>.</p></li><li><p>Solvers are no longer added to a model via <code>Model(solver = XXX(kwargs...))</code>. Instead use <code>Model(with_optimizer(XXX, kwargs...))</code>. For example, <code>Model(with_optimizer(Gurobi.Optimizer, OutputFlag=0))</code>.</p></li><li><p>JuMP containers (for example, the objects returned by <code>@variable</code>) have been redesigned. <code>Containers.SparseAxisArray</code> replaces <code>JuMPDict</code>, <code>JuMPArray</code> was rewritten (inspired by <code>AxisArrays</code>) and renamed <code>Containers.DenseAxisArray</code>, and you can now request a container type with the <code>container=</code> keyword to the macros. See the corresponding <a href="https://www.juliaopt.org/JuMP.jl/dev/variables/#Variable-containers-1">documentation</a> for more details.</p></li><li><p>The statuses returned by solvers have changed. See the possible status values <a href="https://www.juliaopt.org/MathOptInterface.jl/stable/apireference.html#Termination-Status-1">here</a>. The MOI statuses are much richer than the MPB statuses and can be used to distinguish between previously indistinguishable cases (for example, did the solver have a feasible solution when it stopped because of the time limit?).</p></li><li><p>Starting values are separate from result values. Use <code>value</code> to query the value of a variable in a solution. Use <code>start_value</code> and <code>set_start_value</code> to get and set an initial starting point provided to the solver. The solutions from previous solves are no longer automatically set as the starting points for the next solve.</p></li><li><p>The data structures for affine and quadratic expressions <code>AffExpr</code> and <code>QuadExpr</code> have changed. Internally, terms are stored in dictionaries instead of lists. Duplicate coefficients can no longer exist. Accessors and iteration methods have changed.</p></li><li><p><code>JuMPNLPEvaluator</code> no longer includes the linear and quadratic parts of the model in the evaluation calls. These are now handled separately to allow NLP solvers that support various types of constraints.</p></li><li><p>JuMP solver-independent callbacks have been replaced by solver-specific callbacks. See your favorite solver for more details. (See the note below: No solver-specific callbacks are implemented yet.)</p></li><li><p>The <code>norm()</code> syntax is no longer recognized inside macros. Use the <code>SecondOrderCone()</code> set instead.</p></li><li><p>JuMP no longer performs automatic transformation between special quadratic forms and second-order cone constraints. Support for these constraint classes depends on the solver.</p></li><li><p>The symbols <code>:Min</code> and <code>:Max</code> are no longer used as optimization senses. Instead, JuMP uses the <code>OptimizationSense</code> enum from MathOptInterface. <code>@objective(model, Max, ...)</code>, <code>@objective(model, Min, ...)</code>, <code>@NLobjective(model, Max, ...)</code>, and <code>@objective(model, Min, ...)</code> remain valid, but <code>@objective(m, :Max, ...)</code> is no longer accepted.</p></li><li><p>The sign conventions for duals has changed in some cases for consistency with conic duality (see the <a href="https://www.juliaopt.org/MathOptInterface.jl/v0.6.2/apimanual.html#Duals-1">documentation</a>). The <code>shadow_price</code> helper method returns duals with signs that match conventional LP interpretations of dual values as sensitivities of the objective value to relaxations of constraints.</p></li><li><p><code>@constraintref</code> is no longer defined. Instead, create the appropriate container to hold constraint references manually. For example,</p><pre><code class="language-julia hljs">constraints = Dict() # Optionally, specify types for improved performance.
+for i in 1:N
+  constraints[i] = @constraint(model, ...)
+end</code></pre></li><li><p>The <code>lowerbound</code>, <code>upperbound</code>, and <code>basename</code> keyword arguments to the <code>@variable</code> macro have been renamed to <code>lower_bound</code>, <code>upper_bound</code>, and <code>base_name</code>, for consistency with JuMP&#39;s new <a href="https://www.juliaopt.org/JuMP.jl/dev/style/">style recommendations</a>.</p></li><li><p>We rely on broadcasting syntax to apply accessors to collections of variables, for example, <code>value.(x)</code> instead of <code>getvalue(x)</code> for collections. (Use <code>value(x)</code> when <code>x</code> is a scalar object.)</p></li></ul><h3 id="Added-29"><a class="docs-heading-anchor" href="#Added-29">Added</a><a class="docs-heading-anchor-permalink" href="#Added-29" title="Permalink"></a></h3><ul><li><p>Splatting (like <code>f(x...)</code>) is recognized in restricted settings in nonlinear expressions.</p></li><li><p>Support for deleting constraints and variables.</p></li><li><p>The documentation has been completely rewritten using docstrings and Documenter.</p></li><li><p>Support for modeling mixed conic and quadratic models (for example, conic models with quadratic objectives and bi-linear matrix inequalities).</p></li><li><p>Significantly improved support for modeling new types of constraints and for extending JuMP&#39;s macros.</p></li><li><p>Support for providing dual warm starts.</p></li><li><p>Improved support for accessing solver-specific attributes (for example, the irreducible inconsistent subsystem).</p></li><li><p>Explicit control of whether symmetry-enforcing constraints are added to PSD constraints.</p></li><li><p>Support for modeling exponential cones.</p></li><li><p>Significant improvements in internal code quality and testing.</p></li><li><p>Style and naming guidelines.</p></li><li><p>Direct mode and manual mode provide explicit control over when copies of a model are stored or regenerated. See the corresponding <a href="https://www.juliaopt.org/JuMP.jl/dev/solvers/">documentation</a>.</p></li></ul><h3 id="Regressions"><a class="docs-heading-anchor" href="#Regressions">Regressions</a><a id="Regressions-1"></a><a class="docs-heading-anchor-permalink" href="#Regressions" title="Permalink"></a></h3><p>There are known regressions from JuMP 0.18 that will be addressed in a future release (0.19.x or later):</p><ul><li><p>Performance regressions in model generation (<a href="https://github.com/JuliaOpt/JuMP.jl/issues/1403">issue</a>). Please file an issue anyway if you notice a significant performance regression. We have plans to address a number of performance issues, but we might not be aware of all of them.</p></li><li><p>Fast incremental NLP solves are not yet reimplemented (<a href="https://github.com/JuliaOpt/JuMP.jl/issues/1185">issue</a>).</p></li><li><p>We do not yet have an implementation of solver-specific callbacks.</p></li><li><p>The column generation syntax in <code>@variable</code> has been removed (that is, the <code>objective</code>, <code>coefficients</code>, and <code>inconstraints</code> keyword arguments). Support for column generation will be re-introduced in a future release.</p></li><li><p>The ability to solve the continuous relaxation (that is, via <code>solve(model; relaxation = true)</code>) is not yet reimplemented (<a href="https://github.com/JuliaOpt/JuMP.jl/issues/1611">issue</a>).</p></li></ul><h2 id="Version-0.18.5-(December-1,-2018)"><a class="docs-heading-anchor" href="#Version-0.18.5-(December-1,-2018)">Version 0.18.5 (December 1, 2018)</a><a id="Version-0.18.5-(December-1,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.18.5-(December-1,-2018)" title="Permalink"></a></h2><ul><li>Support views in some derivative evaluation functions.</li><li>Improved compatibility with PackageCompiler.</li></ul><h2 id="Version-0.18.4-(October-8,-2018)"><a class="docs-heading-anchor" href="#Version-0.18.4-(October-8,-2018)">Version 0.18.4 (October 8, 2018)</a><a id="Version-0.18.4-(October-8,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.18.4-(October-8,-2018)" title="Permalink"></a></h2><ul><li>Fix a bug in model printing on Julia 0.7 and 1.0.</li></ul><h2 id="Version-0.18.3-(October-1,-2018)"><a class="docs-heading-anchor" href="#Version-0.18.3-(October-1,-2018)">Version 0.18.3 (October 1, 2018)</a><a id="Version-0.18.3-(October-1,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.18.3-(October-1,-2018)" title="Permalink"></a></h2><ul><li>Add support for Julia v1.0 (Thanks @ExpandingMan)</li><li>Fix matrix expressions with quadratic functions (#1508)</li></ul><h2 id="Version-0.18.2-(June-10,-2018)"><a class="docs-heading-anchor" href="#Version-0.18.2-(June-10,-2018)">Version 0.18.2 (June 10, 2018)</a><a id="Version-0.18.2-(June-10,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.18.2-(June-10,-2018)" title="Permalink"></a></h2><ul><li>Fix a bug in second-order derivatives when expressions are present (#1319)</li><li>Fix a bug in <code>@constraintref</code> (#1330)</li></ul><h2 id="Version-0.18.1-(April-9,-2018)"><a class="docs-heading-anchor" href="#Version-0.18.1-(April-9,-2018)">Version 0.18.1 (April 9, 2018)</a><a id="Version-0.18.1-(April-9,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.18.1-(April-9,-2018)" title="Permalink"></a></h2><ul><li>Fix for nested tuple destructuring (#1193)</li><li>Preserve internal model when relaxation=true (#1209)</li><li>Minor bug fixes and updates for example</li></ul><h2 id="Version-0.18.0-(July-27,-2017)"><a class="docs-heading-anchor" href="#Version-0.18.0-(July-27,-2017)">Version 0.18.0 (July 27, 2017)</a><a id="Version-0.18.0-(July-27,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.18.0-(July-27,-2017)" title="Permalink"></a></h2><ul><li>Drop support for Julia 0.5.</li><li>Update for ForwardDiff 0.5.</li><li>Minor bug fixes.</li></ul><h2 id="Version-0.17.1-(June-9,-2017)"><a class="docs-heading-anchor" href="#Version-0.17.1-(June-9,-2017)">Version 0.17.1 (June 9, 2017)</a><a id="Version-0.17.1-(June-9,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.17.1-(June-9,-2017)" title="Permalink"></a></h2><ul><li>Use of <code>constructconstraint!</code> in <code>@SDconstraint</code>.</li><li>Minor bug fixes.</li></ul><h2 id="Version-0.17.0-(May-27,-2017)"><a class="docs-heading-anchor" href="#Version-0.17.0-(May-27,-2017)">Version 0.17.0 (May 27, 2017)</a><a id="Version-0.17.0-(May-27,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.17.0-(May-27,-2017)" title="Permalink"></a></h2><ul><li><strong>Breaking change</strong>: Mixing quadratic and conic constraints is no longer supported.</li><li><strong>Breaking change</strong>: The <code>getvariable</code> and <code>getconstraint</code> functions are replaced by indexing on the corresponding symbol. For instance, to access the variable with name <code>x</code>, one should now write <code>m[:x]</code> instead of <code>getvariable(m, :x)</code>. As a consequence, creating a variable and constraint with the same name now triggers a warning, and accessing one of them afterwards throws an error. This change is breaking only in the latter case.</li><li>Addition of the <code>getobjectivebound</code> function that mirrors the functionality of the MathProgBase <code>getobjbound</code> function except that it takes into account transformations performed by JuMP.</li><li>Minor bug fixes.</li></ul><p>The following changes are primarily of interest to developers of JuMP extensions:</p><ul><li>The new syntax <code>@constraint(model, expr in Cone)</code> creates the constraint ensuring that <code>expr</code> is inside <code>Cone</code>. The <code>Cone</code> argument is passed to <code>constructconstraint!</code> which enables the call to the dispatched to an extension.</li><li>The <code>@variable</code> macro now calls <code>constructvariable!</code> instead of directly calling the <code>Variable</code> constructor. Extra arguments and keyword arguments passed to <code>@variable</code> are passed to <code>constructvariable!</code> which enables the call to be dispatched to an extension.</li><li>Refactor the internal function <code>conicdata</code> (used build the MathProgBase conic model) into smaller sub-functions to make these parts reusable by extensions.</li></ul><h2 id="Version-0.16.2-(March-28,-2017)"><a class="docs-heading-anchor" href="#Version-0.16.2-(March-28,-2017)">Version 0.16.2 (March 28, 2017)</a><a id="Version-0.16.2-(March-28,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.16.2-(March-28,-2017)" title="Permalink"></a></h2><ul><li>Minor bug fixes and printing tweaks</li><li>Address deprecation warnings for Julia 0.6</li></ul><h2 id="Version-0.16.1-(March-7,-2017)"><a class="docs-heading-anchor" href="#Version-0.16.1-(March-7,-2017)">Version 0.16.1 (March 7, 2017)</a><a id="Version-0.16.1-(March-7,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.16.1-(March-7,-2017)" title="Permalink"></a></h2><ul><li>Better support for <code>AbstractArray</code> in JuMP (Thanks @tkoolen)</li><li>Minor bug fixes</li></ul><h2 id="Version-0.16.0-(February-23,-2017)"><a class="docs-heading-anchor" href="#Version-0.16.0-(February-23,-2017)">Version 0.16.0 (February 23, 2017)</a><a id="Version-0.16.0-(February-23,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.16.0-(February-23,-2017)" title="Permalink"></a></h2><ul><li><strong>Breaking change</strong>: JuMP no longer has a mechanism for selecting solvers by default (the previous mechanism was flawed and incompatible with Julia 0.6). Not specifying a solver before calling <code>solve()</code> will result in an error.</li><li><strong>Breaking change</strong>: User-defined functions are no longer global. The first argument to <code>JuMP.register</code> is now a JuMP <code>Model</code> object within whose scope the function will be registered. Calling <code>JuMP.register</code> without a <code>Model</code> now produces an error.</li><li><strong>Breaking change</strong>: Use the new <code>JuMP.fix</code> method to fix a variable to a value or to update the value to which a variable is fixed. Calling <code>setvalue</code> on a fixed variable now results in an error in order to avoid silent behavior changes. (Thanks @joaquimg)</li><li>Nonlinear expressions now print out similarly to linear/quadratic expressions (useful for debugging!)</li><li>New <code>category</code> keyword to <code>@variable</code>. Used for specifying categories of anonymous variables.</li><li>Compatibility with Julia 0.6-dev.</li><li>Minor fixes and improvements (Thanks @cossio, @ccoffrin, @blegat)</li></ul><h2 id="Version-0.15.1-(January-31,-2017)"><a class="docs-heading-anchor" href="#Version-0.15.1-(January-31,-2017)">Version 0.15.1 (January 31, 2017)</a><a id="Version-0.15.1-(January-31,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.15.1-(January-31,-2017)" title="Permalink"></a></h2><ul><li>Bugfix for <code>@LinearConstraints</code> and friends</li></ul><h2 id="Version-0.15.0-(December-22,-2016)"><a class="docs-heading-anchor" href="#Version-0.15.0-(December-22,-2016)">Version 0.15.0 (December 22, 2016)</a><a id="Version-0.15.0-(December-22,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.15.0-(December-22,-2016)" title="Permalink"></a></h2><ul><li>Julia 0.5.0 is the minimum required version for this release.</li><li>Document support for BARON solver</li><li>Enable info callbacks in more states than before, for example, for recording solutions. New <code>when</code> argument to <code>addinfocallback</code> (<a href="https://github.com/JuliaOpt/JuMP.jl/pull/814">#814</a>, thanks @yeesian)</li><li>Improved support for anonymous variables. This includes new warnings for potentially confusing use of the traditional non-anonymous syntax:<ul><li>When multiple variables in a model are given the same name</li><li>When non-symbols are used as names, for example, <code>@variable(m, x[1][1:N])</code></li></ul></li><li>Improvements in iterating over JuMP containers (<a href="https://github.com/JuliaOpt/JuMP.jl/pull/836">#836</a>, thanks @IssamT)</li><li>Support for writing variable names in .lp file output (Thanks @leethargo)</li><li>Support for querying duals to SDP problems (Thanks @blegat)</li><li>The comprehension syntax with curly braces <code>sum{}</code>, <code>prod{}</code>, and <code>norm2{}</code> has been deprecated in favor of Julia&#39;s native comprehension syntax <code>sum()</code>, <code>prod()</code> and <code>norm()</code> as previously announced. (For early adopters of the new syntax, <code>norm2()</code> was renamed to <code>norm()</code> without deprecation.)</li><li>Unit tests rewritten to use Base.Test instead of FactCheck</li><li>Improved support for operations with matrices of JuMP types (Thanks @ExpandingMan)</li><li>The syntax to halt a solver from inside a callback has changed from <code>throw(CallbackAbort())</code> to <code>return JuMP.StopTheSolver</code></li><li>Minor bug fixes</li></ul><h2 id="Version-0.14.2-(December-12,-2016)"><a class="docs-heading-anchor" href="#Version-0.14.2-(December-12,-2016)">Version 0.14.2 (December 12, 2016)</a><a id="Version-0.14.2-(December-12,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.14.2-(December-12,-2016)" title="Permalink"></a></h2><ul><li>Allow singleton anonymous variables (includes bugfix)</li></ul><h2 id="Version-0.14.1-(September-12,-2016)"><a class="docs-heading-anchor" href="#Version-0.14.1-(September-12,-2016)">Version 0.14.1 (September 12, 2016)</a><a id="Version-0.14.1-(September-12,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.14.1-(September-12,-2016)" title="Permalink"></a></h2><ul><li>More consistent handling of states in informational callbacks, includes a new <code>when</code> parameter to <code>addinfocallback</code> for specifying in which state an informational callback should be called.</li></ul><h2 id="Version-0.14.0-(August-7,-2016)"><a class="docs-heading-anchor" href="#Version-0.14.0-(August-7,-2016)">Version 0.14.0 (August 7, 2016)</a><a id="Version-0.14.0-(August-7,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.14.0-(August-7,-2016)" title="Permalink"></a></h2><ul><li>Compatibility with Julia 0.5 and ForwardDiff 0.2</li><li>Support for &quot;anonymous&quot; variables, constraints, expressions, and parameters, for example, <code>x = @variable(m, [1:N])</code> instead of <code>@variable(m, x[1:N])</code></li><li>Support for retrieving constraints from a model by name via <code>getconstraint</code></li><li><code>@NLconstraint</code> now returns constraint references (as expected).</li><li>Support for vectorized expressions within lazy constraints</li><li>On Julia 0.5, parse new comprehension syntax <code>sum(x[i] for i in 1:N if isodd(i))</code> instead of <code>sum{ x[i], i in 1:N; isodd(i) }</code>. The old syntax with curly braces will be deprecated in JuMP 0.15.</li><li>Now possible to provide nonlinear expressions as &quot;raw&quot; Julia <code>Expr</code> objects instead of using JuMP&#39;s nonlinear macros. This input format is useful for programmatically generated expressions.</li><li><code>s/Mathematical Programming/Mathematical Optimization/</code></li><li>Support for local cuts (Thanks to @madanim, Mehdi Madani)</li><li>Document Xpress interface developed by @joaquimg, Joaquim Dias Garcia</li><li>Minor bug and deprecation fixes (Thanks @odow, @jrevels)</li></ul><h2 id="Version-0.13.2-(May-16,-2016)"><a class="docs-heading-anchor" href="#Version-0.13.2-(May-16,-2016)">Version 0.13.2 (May 16, 2016)</a><a id="Version-0.13.2-(May-16,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.13.2-(May-16,-2016)" title="Permalink"></a></h2><ul><li>Compatibility update for MathProgBase</li></ul><h2 id="Version-0.13.1-(May-3,-2016)"><a class="docs-heading-anchor" href="#Version-0.13.1-(May-3,-2016)">Version 0.13.1 (May 3, 2016)</a><a id="Version-0.13.1-(May-3,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.13.1-(May-3,-2016)" title="Permalink"></a></h2><ul><li>Fix broken deprecation for <code>registerNLfunction</code>.</li></ul><h2 id="Version-0.13.0-(April-29,-2016)"><a class="docs-heading-anchor" href="#Version-0.13.0-(April-29,-2016)">Version 0.13.0 (April 29, 2016)</a><a id="Version-0.13.0-(April-29,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.13.0-(April-29,-2016)" title="Permalink"></a></h2><ul><li>Most exported methods and macros have been renamed to avoid camelCase. See the list of changes <a href="https://github.com/JuliaOpt/JuMP.jl/blob/e53d0db67cde2a4b80d0c1281f4b49eb0128a1f5/src/deprecated.jl#L30">here</a>. There is a 1-1 mapping from the old names to the new, and it is safe to simply replace the names to update existing models.</li><li>Specify variable lower/upper bounds in <code>@variable</code> using the <code>lowerbound</code> and <code>upperbound</code> keyword arguments.</li><li>Change name printed for variable using the <code>basename</code> keyword argument to <code>@variable</code>.</li><li>New <code>@variables</code> macro allows multi-line declaration of groups of variables.</li><li>A number of solver methods previously available only through MathProgBase are now exposed directly in JuMP. The fix was <a href="https://youtu.be/qF1lZPJ3a5A">recorded</a> live.</li><li>Compatibility fixes with Julia 0.5.</li><li>The &quot;end&quot; indexing syntax is no longer supported within JuMPArrays which do not use 1-based indexing until upstream issues are resolved, see <a href="https://github.com/JuliaOpt/JuMP.jl/issues/730">here</a>.</li></ul><h2 id="Version-0.12.2-(March-9,-2016)"><a class="docs-heading-anchor" href="#Version-0.12.2-(March-9,-2016)">Version 0.12.2 (March 9, 2016)</a><a id="Version-0.12.2-(March-9,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.12.2-(March-9,-2016)" title="Permalink"></a></h2><ul><li>Small fixes for nonlinear optimization</li></ul><h2 id="Version-0.12.1-(March-1,-2016)"><a class="docs-heading-anchor" href="#Version-0.12.1-(March-1,-2016)">Version 0.12.1 (March 1, 2016)</a><a id="Version-0.12.1-(March-1,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.12.1-(March-1,-2016)" title="Permalink"></a></h2><ul><li>Fix a regression in slicing for JuMPArrays (when not using 1-based indexing)</li></ul><h2 id="Version-0.12.0-(February-27,-2016)"><a class="docs-heading-anchor" href="#Version-0.12.0-(February-27,-2016)">Version 0.12.0 (February 27, 2016)</a><a id="Version-0.12.0-(February-27,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.12.0-(February-27,-2016)" title="Permalink"></a></h2><ul><li>The automatic differentiation functionality has been completely rewritten with a number of user-facing changes:<ul><li><code>@defExpr</code> and <code>@defNLExpr</code> now take the model as the first argument. The previous one-argument version of <code>@defExpr</code> is deprecated; all expressions should be named. For example, replace <code>@defExpr(2x+y)</code> with <code>@defExpr(jump_model, my_expr, 2x+y)</code>.</li><li>JuMP no longer uses Julia&#39;s variable binding rules for efficiently re-solving a sequence of nonlinear models. Instead, we have introduced nonlinear parameters. This is a breaking change, so we have added a warning message when we detect models that may depend on the old behavior.</li><li>Support for user-defined functions integrated within nonlinear JuMP expressions.</li></ul></li><li>Replaced iteration over <code>AffExpr</code> with <code>Number</code>-like scalar iteration; previous iteration behavior is now available via <code>linearterms(::AffExpr)</code>.</li><li>Stopping the solver via <code>throw(CallbackAbort())</code> from a callback no longer triggers an exception. Instead, <code>solve()</code> returns <code>UserLimit</code> status.</li><li><code>getDual()</code> now works for conic problems (Thanks @emreyamangil.)</li></ul><h2 id="Version-0.11.3-(February-4,-2016)"><a class="docs-heading-anchor" href="#Version-0.11.3-(February-4,-2016)">Version 0.11.3 (February 4, 2016)</a><a id="Version-0.11.3-(February-4,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.11.3-(February-4,-2016)" title="Permalink"></a></h2><ul><li>Bug-fix for problems with quadratic objectives and semidefinite constraints</li></ul><h2 id="Version-0.11.2-(January-14,-2016)"><a class="docs-heading-anchor" href="#Version-0.11.2-(January-14,-2016)">Version 0.11.2 (January 14, 2016)</a><a id="Version-0.11.2-(January-14,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.11.2-(January-14,-2016)" title="Permalink"></a></h2><ul><li>Compatibility update for Mosek</li></ul><h2 id="Version-0.11.1-(December-1,-2015)"><a class="docs-heading-anchor" href="#Version-0.11.1-(December-1,-2015)">Version 0.11.1 (December 1, 2015)</a><a id="Version-0.11.1-(December-1,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.11.1-(December-1,-2015)" title="Permalink"></a></h2><ul><li>Remove usage of <code>@compat</code> in tests.</li><li>Fix updating quadratic objectives for nonlinear models.</li></ul><h2 id="Version-0.11.0-(November-30,-2015)"><a class="docs-heading-anchor" href="#Version-0.11.0-(November-30,-2015)">Version 0.11.0 (November 30, 2015)</a><a id="Version-0.11.0-(November-30,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.11.0-(November-30,-2015)" title="Permalink"></a></h2><ul><li>Julia 0.4.0 is the minimum required version for this release.</li><li>Fix for scoping semantics of index variables in sum{}. Index variables no longer leak into the surrounding scope.</li><li>Addition of the <code>solve(m::Model, relaxation=true)</code> keyword argument to solve the standard continuous relaxation of model <code>m</code></li><li>The <code>getConstraintBounds()</code> method allows access to the lower and upper bounds of all constraints in a (nonlinear) model.</li><li>Update for breaking changes in MathProgBase</li></ul><h2 id="Version-0.10.3-(November-20,-2015)"><a class="docs-heading-anchor" href="#Version-0.10.3-(November-20,-2015)">Version 0.10.3 (November 20, 2015)</a><a id="Version-0.10.3-(November-20,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.10.3-(November-20,-2015)" title="Permalink"></a></h2><ul><li>Fix a rare error when parsing quadratic expressions</li><li>Fix <code>Variable()</code> constructor with default arguments</li><li>Detect unrecognized keywords in <code>solve()</code></li></ul><h2 id="Version-0.10.2-(September-28,-2015)"><a class="docs-heading-anchor" href="#Version-0.10.2-(September-28,-2015)">Version 0.10.2 (September 28, 2015)</a><a id="Version-0.10.2-(September-28,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.10.2-(September-28,-2015)" title="Permalink"></a></h2><ul><li>Fix for deprecation warnings</li></ul><h2 id="Version-0.10.1-(September-3,-2015)"><a class="docs-heading-anchor" href="#Version-0.10.1-(September-3,-2015)">Version 0.10.1 (September 3, 2015)</a><a id="Version-0.10.1-(September-3,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.10.1-(September-3,-2015)" title="Permalink"></a></h2><ul><li>Fixes for ambiguity warnings.</li><li>Fix for breaking change in precompilation syntax in Julia 0.4-pre</li></ul><h2 id="Version-0.10.0-(August-31,-2015)"><a class="docs-heading-anchor" href="#Version-0.10.0-(August-31,-2015)">Version 0.10.0 (August 31, 2015)</a><a id="Version-0.10.0-(August-31,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.10.0-(August-31,-2015)" title="Permalink"></a></h2><ul><li>Support (on Julia 0.4 and later) for conditions in indexing <code>@defVar</code> and <code>@addConstraint</code> constructs, for example, <code>@defVar(m, x[i=1:5,j=1:5; i+j &gt;= 3])</code></li><li>Support for vectorized operations on Variables and expressions. See the documentation for details.</li><li>New <code>getVar()</code> method to access variables in a model by name</li><li>Support for semidefinite programming.</li><li>Dual solutions are now available for general nonlinear problems. You may call <code>getDual</code> on a reference object for a nonlinear constraint, and <code>getDual</code> on a variable object for Lagrange multipliers from active bounds.</li><li>Introduce warnings for two common performance traps: too many calls to <code>getValue()</code> on a collection of variables and use of the <code>+</code> operator in a loop to sum expressions.</li><li>Second-order cone constraints can be written directly with the <code>norm()</code> and <code>norm2{}</code> syntax.</li><li>Implement MathProgBase interface for querying Hessian-vector products.</li><li>Iteration over <code>JuMPContainer</code>s is deprecated; instead, use the <code>keys</code> and <code>values</code> functions, and <code>zip(keys(d),values(d))</code> for the old behavior.</li><li><code>@defVar</code> returns <code>Array{Variable,N}</code> when each of <code>N</code> index sets are of the form <code>1:nᵢ</code>.</li><li>Module precompilation: on Julia 0.4 and later, <code>using JuMP</code> is now much faster.</li></ul><h2 id="Version-0.9.3-(August-11,-2015)"><a class="docs-heading-anchor" href="#Version-0.9.3-(August-11,-2015)">Version 0.9.3 (August 11, 2015)</a><a id="Version-0.9.3-(August-11,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.9.3-(August-11,-2015)" title="Permalink"></a></h2><ul><li>Fixes for FactCheck testing on julia v0.4.</li></ul><h2 id="Version-0.9.2-(June-27,-2015)"><a class="docs-heading-anchor" href="#Version-0.9.2-(June-27,-2015)">Version 0.9.2 (June 27, 2015)</a><a id="Version-0.9.2-(June-27,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.9.2-(June-27,-2015)" title="Permalink"></a></h2><ul><li>Fix bug in @addConstraints.</li></ul><h2 id="Version-0.9.1-(April-25,-2015)"><a class="docs-heading-anchor" href="#Version-0.9.1-(April-25,-2015)">Version 0.9.1 (April 25, 2015)</a><a id="Version-0.9.1-(April-25,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.9.1-(April-25,-2015)" title="Permalink"></a></h2><ul><li>Fix for Julia 0.4-dev.</li><li>Small infrastructure improvements for extensions.</li></ul><h2 id="Version-0.9.0-(April-18,-2015)"><a class="docs-heading-anchor" href="#Version-0.9.0-(April-18,-2015)">Version 0.9.0 (April 18, 2015)</a><a id="Version-0.9.0-(April-18,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.9.0-(April-18,-2015)" title="Permalink"></a></h2><ul><li>Comparison operators for constructing constraints (for example, <code>2x &gt;= 1</code>) have been deprecated. Instead, construct the constraints explicitly in the <code>@addConstraint</code> macro to add them to the model, or in the <code>@LinearConstraint</code> macro to create a stand-alone linear constraint instance.</li><li><code>getValue()</code> method implemented to compute the value of a nonlinear subexpression</li><li>JuMP is now released under the Mozilla Public License version 2.0 (was previously LGPL). MPL is a copyleft license which is less restrictive than LGPL, especially for embedding JuMP within other applications.</li><li>A number of performance improvements in ReverseDiffSparse for computing derivatives.</li><li><code>MathProgBase.getsolvetime(m)</code> now returns the solution time reported by the solver, if available. (Thanks @odow, Oscar Dowson)</li><li>Formatting fix for LP format output. (Thanks @sbebo, Leonardo Taccari).</li></ul><h2 id="Version-0.8.0-(February-17,-2015)"><a class="docs-heading-anchor" href="#Version-0.8.0-(February-17,-2015)">Version 0.8.0 (February 17, 2015)</a><a id="Version-0.8.0-(February-17,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.8.0-(February-17,-2015)" title="Permalink"></a></h2><ul><li>Nonlinear subexpressions now supported with the <code>@defNLExpr</code> macro.</li><li>SCS supported for solving second-order conic problems.</li><li><code>setXXXCallback</code> family deprecated in favor of <code>addXXXCallback</code>.</li><li>Multiple callbacks of the same type can be registered.</li><li>Added support for informational callbacks via <code>addInfoCallback</code>.</li><li>A <code>CallbackAbort</code> exception can be thrown from callback to safely exit optimization.</li></ul><h2 id="Version-0.7.4-(February-4,-2015)"><a class="docs-heading-anchor" href="#Version-0.7.4-(February-4,-2015)">Version 0.7.4 (February 4, 2015)</a><a id="Version-0.7.4-(February-4,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.7.4-(February-4,-2015)" title="Permalink"></a></h2><ul><li>Reduced costs and linear constraint duals are now accessible when quadratic constraints are present.</li><li>Two-sided nonlinear constraints are supported.</li><li>Methods for accessing the number of variables and constraints in a model are renamed.</li><li>New default procedure for setting initial values in nonlinear optimization: project zero onto the variable bounds.</li><li>Small bug fixes.</li></ul><h2 id="Version-0.7.3-(January-14,-2015)"><a class="docs-heading-anchor" href="#Version-0.7.3-(January-14,-2015)">Version 0.7.3 (January 14, 2015)</a><a id="Version-0.7.3-(January-14,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.7.3-(January-14,-2015)" title="Permalink"></a></h2><ul><li>Fix a method ambiguity conflict with Compose.jl (cosmetic fix)</li></ul><h2 id="Version-0.7.2-(January-9,-2015)"><a class="docs-heading-anchor" href="#Version-0.7.2-(January-9,-2015)">Version 0.7.2 (January 9, 2015)</a><a id="Version-0.7.2-(January-9,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.7.2-(January-9,-2015)" title="Permalink"></a></h2><ul><li>Fix a bug in <code>sum(::JuMPDict)</code></li><li>Added the <code>setCategory</code> function to change a variables category (for example, continuous or binary)</li></ul><p>after construction, and <code>getCategory</code> to retrieve the variable category.</p><h2 id="Version-0.7.1-(January-2,-2015)"><a class="docs-heading-anchor" href="#Version-0.7.1-(January-2,-2015)">Version 0.7.1 (January 2, 2015)</a><a id="Version-0.7.1-(January-2,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.7.1-(January-2,-2015)" title="Permalink"></a></h2><ul><li>Fix a bug in parsing linear expressions in macros. Affects only Julia 0.4 and later.</li></ul><h2 id="Version-0.7.0-(December-29,-2014)"><a class="docs-heading-anchor" href="#Version-0.7.0-(December-29,-2014)">Version 0.7.0 (December 29, 2014)</a><a id="Version-0.7.0-(December-29,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.7.0-(December-29,-2014)" title="Permalink"></a></h2><h3 id="Linear/quadratic/conic-programming"><a class="docs-heading-anchor" href="#Linear/quadratic/conic-programming">Linear/quadratic/conic programming</a><a id="Linear/quadratic/conic-programming-1"></a><a class="docs-heading-anchor-permalink" href="#Linear/quadratic/conic-programming" title="Permalink"></a></h3><ul><li><strong>Breaking change</strong>: The syntax for column-wise model generation has been changed to use keyword arguments in <code>@defVar</code>.</li><li>On Julia 0.4 and later, variables and coefficients may be multiplied in any order within macros. That is, variable*coefficient is now valid syntax.</li><li>ECOS supported for solving second-order conic problems.</li></ul><h3 id="_nonlinear_programming_release_notes"><a class="docs-heading-anchor" href="#_nonlinear_programming_release_notes">Nonlinear programming</a><a id="_nonlinear_programming_release_notes-1"></a><a class="docs-heading-anchor-permalink" href="#_nonlinear_programming_release_notes" title="Permalink"></a></h3><ul><li>Support for skipping model generation when solving a sequence of nonlinear models with changing data.</li><li>Fix a memory leak when solving a sequence of nonlinear models.</li><li>The <code>@addNLConstraint</code> macro now supports the three-argument version to define sets of nonlinear constraints.</li><li>KNITRO supported as a nonlinear solver.</li><li>Speed improvements for model generation.</li><li>The <code>@addNLConstraints</code> macro supports adding multiple (groups of) constraints at once. Syntax is similar to <code>@addConstraints</code>.</li><li>Discrete variables allowed in nonlinear problems for solvers which support them (currently only KNITRO).</li></ul><h3 id="General"><a class="docs-heading-anchor" href="#General">General</a><a id="General-1"></a><a class="docs-heading-anchor-permalink" href="#General" title="Permalink"></a></h3><ul><li>Starting values for variables may now be specified with <code>@defVar(m, x, start=value)</code>.</li><li>The <code>setSolver</code> function allows users to change the solver subsequent to model creation.</li><li>Support for &quot;fixed&quot; variables via the <code>@defVar(m, x == 1)</code> syntax.</li><li>Unit tests rewritten to use FactCheck.jl, improved testing across solvers.</li></ul><h2 id="Version-0.6.3-(October-19,-2014)"><a class="docs-heading-anchor" href="#Version-0.6.3-(October-19,-2014)">Version 0.6.3 (October 19, 2014)</a><a id="Version-0.6.3-(October-19,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.6.3-(October-19,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug in multiplying two AffExpr objects.</li></ul><h2 id="Version-0.6.2-(October-11,-2014)"><a class="docs-heading-anchor" href="#Version-0.6.2-(October-11,-2014)">Version 0.6.2 (October 11, 2014)</a><a id="Version-0.6.2-(October-11,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.6.2-(October-11,-2014)" title="Permalink"></a></h2><ul><li>Further improvements and bug fixes for printing.</li><li>Fixed a bug in <code>@defExpr</code>.</li><li>Support for accessing expression graphs through the MathProgBase NLP interface.</li></ul><h2 id="Version-0.6.1-(September-19,-2014)"><a class="docs-heading-anchor" href="#Version-0.6.1-(September-19,-2014)">Version 0.6.1 (September 19, 2014)</a><a id="Version-0.6.1-(September-19,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.6.1-(September-19,-2014)" title="Permalink"></a></h2><ul><li>Improvements and bug fixes for printing.</li></ul><h2 id="Version-0.6.0-(September-9,-2014)"><a class="docs-heading-anchor" href="#Version-0.6.0-(September-9,-2014)">Version 0.6.0 (September 9, 2014)</a><a id="Version-0.6.0-(September-9,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.6.0-(September-9,-2014)" title="Permalink"></a></h2><ul><li>Julia 0.3.0 is the minimum required version for this release.</li><li><code>buildInternalModel(m::Model)</code> added to build solver-level model in memory without optimizing.</li><li>Deprecate <code>load_model_only</code> keyword argument to <code>solve</code>.</li><li>Add groups of constraints with <code>@addConstraints</code> macro.</li><li>Unicode operators now supported, including <code>∑</code> for <code>sum</code>, <code>∏</code> for <code>prod</code>, and <code>≤</code>/<code>≥</code></li><li>Quadratic constraints supported in <code>@addConstraint</code> macro.</li><li>Quadratic objectives supported in <code>@setObjective</code> macro.</li><li>MathProgBase solver-independent interface replaces Ipopt-specific interface for nonlinear problems<ul><li><strong>Breaking change</strong>: <code>IpoptOptions</code> no longer supported to specify solver options, use <code>m = Model(solver=IpoptSolver(options...))</code> instead.</li></ul></li><li>New solver interfaces: ECOS, NLopt, and nonlinear support for MOSEK</li><li>New option to control whether the lazy constraint callback is executed at each node in the B&amp;B tree or just when feasible solutions are found</li><li>Add support for semicontinuous and semi-integer variables for those solvers that support them.</li><li>Add support for index dependencies (for example, triangular indexing) in <code>@defVar</code>, <code>@addConstraint</code>, and <code>@defExpr</code> (for example, <code>@defVar(m, x[i=1:10,j=i:10])</code>).<ul><li>This required some changes to the internal structure of JuMP containers, which may break code that explicitly stored <code>JuMPDict</code> objects.</li></ul></li></ul><h2 id="Version-0.5.8-(September-24,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.8-(September-24,-2014)">Version 0.5.8 (September 24, 2014)</a><a id="Version-0.5.8-(September-24,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.8-(September-24,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug with specifying solvers (affects Julia 0.2 only)</li></ul><h2 id="Version-0.5.7-(September-5,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.7-(September-5,-2014)">Version 0.5.7 (September 5, 2014)</a><a id="Version-0.5.7-(September-5,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.7-(September-5,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug in printing models</li></ul><h2 id="Version-0.5.6-(September-2,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.6-(September-2,-2014)">Version 0.5.6 (September 2, 2014)</a><a id="Version-0.5.6-(September-2,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.6-(September-2,-2014)" title="Permalink"></a></h2><ul><li>Add support for semicontinuous and semi-integer variables for those solvers that support them.<ul><li><strong>Breaking change</strong>: Syntax for <code>Variable()</code> constructor has changed (use of this interface remains discouraged)</li></ul></li><li>Update for breaking changes in MathProgBase</li></ul><h2 id="Version-0.5.5-(July-6,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.5-(July-6,-2014)">Version 0.5.5 (July 6, 2014)</a><a id="Version-0.5.5-(July-6,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.5-(July-6,-2014)" title="Permalink"></a></h2><ul><li>Fix bug with problem modification: adding variables that did not appear in existing constraints or objective.</li></ul><h2 id="Version-0.5.4-(June-19,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.4-(June-19,-2014)">Version 0.5.4 (June 19, 2014)</a><a id="Version-0.5.4-(June-19,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.4-(June-19,-2014)" title="Permalink"></a></h2><ul><li>Update for breaking change in MathProgBase which reduces loading times for <code>using JuMP</code></li><li>Fix error when MIPs not solved to optimality</li></ul><h2 id="Version-0.5.3-(May-21,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.3-(May-21,-2014)">Version 0.5.3 (May 21, 2014)</a><a id="Version-0.5.3-(May-21,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.3-(May-21,-2014)" title="Permalink"></a></h2><ul><li>Update for breaking change in ReverseDiffSparse</li></ul><h2 id="Version-0.5.2-(May-9,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.2-(May-9,-2014)">Version 0.5.2 (May 9, 2014)</a><a id="Version-0.5.2-(May-9,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.2-(May-9,-2014)" title="Permalink"></a></h2><ul><li>Fix compatibility with Julia 0.3 prerelease</li></ul><h2 id="Version-0.5.1-(May-5,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.1-(May-5,-2014)">Version 0.5.1 (May 5, 2014)</a><a id="Version-0.5.1-(May-5,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.1-(May-5,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug in coefficient handling inside lazy constraints and user cuts</li></ul><h2 id="Version-0.5.0-(May-2,-2014)"><a class="docs-heading-anchor" href="#Version-0.5.0-(May-2,-2014)">Version 0.5.0 (May 2, 2014)</a><a id="Version-0.5.0-(May-2,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.5.0-(May-2,-2014)" title="Permalink"></a></h2><ul><li>Support for nonlinear optimization with exact, sparse second-order derivatives automatically computed. Ipopt is currently the only solver supported.</li><li><code>getValue</code> for <code>AffExpr</code> and <code>QuadExpr</code></li><li><strong>Breaking change</strong>: <code>getSolverModel</code> replaced by <code>getInternalModel</code>, which returns the internal MathProgBase-level model</li><li>Groups of constraints can be specified with <code>@addConstraint</code> (see documentation for details). This is not a breaking change.</li><li><code>dot(::JuMPDict{Variable},::JuMPDict{Variable})</code> now returns the corresponding quadratic expression.</li></ul><h2 id="Version-0.4.1-(March-24,-2014)"><a class="docs-heading-anchor" href="#Version-0.4.1-(March-24,-2014)">Version 0.4.1 (March 24, 2014)</a><a id="Version-0.4.1-(March-24,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.4.1-(March-24,-2014)" title="Permalink"></a></h2><ul><li>Fix bug where change in objective sense was ignored when re-solving a model.</li><li>Fix issue with handling zero coefficients in AffExpr.</li></ul><h2 id="Version-0.4.0-(March-10,-2014)"><a class="docs-heading-anchor" href="#Version-0.4.0-(March-10,-2014)">Version 0.4.0 (March 10, 2014)</a><a id="Version-0.4.0-(March-10,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.4.0-(March-10,-2014)" title="Permalink"></a></h2><ul><li>Support for SOS1 and SOS2 constraints.</li><li>Solver-independent callback for user heuristics.</li><li><code>dot</code> and <code>sum</code> implemented for <code>JuMPDict</code> objects. Now you can say <code>@addConstraint(m, dot(a,x) &lt;= b)</code>.</li><li>Developers: support for extensions to JuMP. See definition of Model in <code>src/JuMP.jl</code> for more details.</li><li>Option to construct the low-level model before optimizing.</li></ul><h2 id="Version-0.3.2-(February-17,-2014)"><a class="docs-heading-anchor" href="#Version-0.3.2-(February-17,-2014)">Version 0.3.2 (February 17, 2014)</a><a id="Version-0.3.2-(February-17,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.3.2-(February-17,-2014)" title="Permalink"></a></h2><ul><li>Improved model printing<ul><li>Preliminary support for IJulia output</li></ul></li></ul><h2 id="Version-0.3.1-(January-30,-2014)"><a class="docs-heading-anchor" href="#Version-0.3.1-(January-30,-2014)">Version 0.3.1 (January 30, 2014)</a><a id="Version-0.3.1-(January-30,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.3.1-(January-30,-2014)" title="Permalink"></a></h2><ul><li>Documentation updates</li><li>Support for MOSEK</li><li>CPLEXLink renamed to CPLEX</li></ul><h2 id="Version-0.3.0-(January-21,-2014)"><a class="docs-heading-anchor" href="#Version-0.3.0-(January-21,-2014)">Version 0.3.0 (January 21, 2014)</a><a id="Version-0.3.0-(January-21,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.3.0-(January-21,-2014)" title="Permalink"></a></h2><ul><li>Unbounded/infeasibility rays: <code>getValue()</code> will return the corresponding components of an unbounded ray when a model is unbounded, if supported by the selected solver. <code>getDual()</code> will return an infeasibility ray (Farkas proof) if a model is infeasible and the selected solver supports this feature.</li><li>Solver-independent callbacks for user generated cuts.</li><li>Use new interface for solver-independent QCQP.</li><li><code>setlazycallback</code> renamed to <code>setLazyCallback</code> for consistency.</li></ul><h2 id="Version-0.2.0-(December-15,-2013)"><a class="docs-heading-anchor" href="#Version-0.2.0-(December-15,-2013)">Version 0.2.0 (December 15, 2013)</a><a id="Version-0.2.0-(December-15,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.2.0-(December-15,-2013)" title="Permalink"></a></h2><h3 id="Breaking-8"><a class="docs-heading-anchor" href="#Breaking-8">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-8" title="Permalink"></a></h3><ul><li>Objective sense is specified in <code>setObjective</code> instead of in the <code>Model</code> constructor.</li><li><code>lpsolver</code> and <code>mipsolver</code> merged into single <code>solver</code> option.</li></ul><h3 id="Added-30"><a class="docs-heading-anchor" href="#Added-30">Added</a><a class="docs-heading-anchor-permalink" href="#Added-30" title="Permalink"></a></h3><ul><li>Problem modification with efficient LP restarts and MIP warm-starts.</li><li>Relatedly, column-wise modeling now supported.</li><li>Solver-independent callbacks supported. Currently we support only a &quot;lazy constraint&quot; callback, which works with Gurobi, CPLEX, and GLPK. More callbacks coming soon.</li></ul><h2 id="Version-0.1.2-(November-16,-2013)"><a class="docs-heading-anchor" href="#Version-0.1.2-(November-16,-2013)">Version 0.1.2 (November 16, 2013)</a><a id="Version-0.1.2-(November-16,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.1.2-(November-16,-2013)" title="Permalink"></a></h2><ul><li>Bug fixes for printing, improved error messages.</li><li>Allow <code>AffExpr</code> to be used in macros; for example, <code>ex = y + z; @addConstraint(m, x + 2*ex &lt;= 3)</code></li></ul><h2 id="Version-0.1.1-(October-23,-2013)"><a class="docs-heading-anchor" href="#Version-0.1.1-(October-23,-2013)">Version 0.1.1 (October 23, 2013)</a><a id="Version-0.1.1-(October-23,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.1.1-(October-23,-2013)" title="Permalink"></a></h2><ul><li>Update for solver specification API changes in MathProgBase.</li></ul><h2 id="Version-0.1.0-(October-3,-2013)"><a class="docs-heading-anchor" href="#Version-0.1.0-(October-3,-2013)">Version 0.1.0 (October 3, 2013)</a><a id="Version-0.1.0-(October-3,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#Version-0.1.0-(October-3,-2013)" title="Permalink"></a></h2><ul><li>Initial public release.</li></ul></article><nav class="docs-footer"><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/developers/checklists/index.html b/previews/PR3525/developers/checklists/index.html
new file mode 100644
index 00000000000..d70f06e88df
--- /dev/null
+++ b/previews/PR3525/developers/checklists/index.html
@@ -0,0 +1,65 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Checklists · JuMP</title><meta name="title" content="Checklists · JuMP"/><meta property="og:title" content="Checklists · JuMP"/><meta property="twitter:title" content="Checklists · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox" checked/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../contributing/">Contributing</a></li><li><a class="tocitem" href="../extensions/">Extensions</a></li><li><a class="tocitem" href="../custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../style/">Style Guide</a></li><li><a class="tocitem" href="../roadmap/">Roadmap</a></li><li class="is-active"><a class="tocitem" href>Checklists</a><ul class="internal"><li><a class="tocitem" href="#Making-a-release"><span>Making a release</span></a></li><li><a class="tocitem" href="#Adding-a-new-solver-to-the-documentation"><span>Adding a new solver to the documentation</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Developer Docs</a></li><li class="is-active"><a href>Checklists</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Checklists</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/developers/checklists.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Checklists"><a class="docs-heading-anchor" href="#Checklists">Checklists</a><a id="Checklists-1"></a><a class="docs-heading-anchor-permalink" href="#Checklists" title="Permalink"></a></h1><p>The purpose of this page is to collate a series of checklists for commonly performed changes to the source code of JuMP.</p><p>In each case, copy the checklist into the description of the pull request.</p><h2 id="Making-a-release"><a class="docs-heading-anchor" href="#Making-a-release">Making a release</a><a id="Making-a-release-1"></a><a class="docs-heading-anchor-permalink" href="#Making-a-release" title="Permalink"></a></h2><p>In preparation for a release, use the following checklist. These steps can be done in the same commit, or separately. The last commit should have the message &quot;Prep for vX.Y.Z.&quot;</p><pre><code class="nohighlight hljs">## Pre-release
+
+ - [ ] Check that the pinned packages in `docs/Project.toml` are updated. We pin
+       the versions so that changes in the solvers (changes in printing, small
+       numeric changes) do not break the printing of the JuMP docs in arbitrary
+       commits.
+ - [ ] Check that the `rev` fields in `docs/packages.toml` are updated. We pin
+       the versions of solvers and extensions to ensure that changes to their
+       READMEs do not break the JuMP docs in arbitrary commits, and to ensure
+       that the versions are compatible with the latest JuMP and
+       MathOptInterface releases.
+ - [ ] Update `docs/src/changelog.md`
+ - [ ] Run https://github.com/jump-dev/JuMP.jl/actions/workflows/extension-tests.yml
+       using a `workflow_dispatch` trigger to check for any changes in JuMP that
+       broke extensions.
+ - [ ] Change the version number in `Project.toml`
+ - [ ] Update the links in README.md
+ - [ ] The commit messages in this PR do not contain `[ci skip]`
+
+## The release
+
+ - [ ] After merging this pull request, comment `[at]JuliaRegistrator register` in
+       the GitHub commit. This should automatically publish a new version to the
+       Julia registry, as well as create a tag, and rebuild the documentation
+       for this tag.
+
+       These steps can take quite a bit of time (1 hour or more), so don&#39;t be
+       surprised if the new documentation takes a while to appear. In addition,
+       the links in the README will be broken until JuliaHub fetches the new
+       version on their servers.
+
+## Post-release
+
+ - [ ] Once the tag is created, update the relevant `release-` branch. The latest
+       release branch at the time of writing is `release-1.0` (we haven&#39;t
+       back-ported any patches that needed to create a `release-1.Y` branch). To
+       to update the release branch with the v1.10.0 tag, do:
+       ```
+       git checkout release-1.0
+       git pull
+       git merge v1.10.0
+       git push
+       ```</code></pre><h2 id="Adding-a-new-solver-to-the-documentation"><a class="docs-heading-anchor" href="#Adding-a-new-solver-to-the-documentation">Adding a new solver to the documentation</a><a id="Adding-a-new-solver-to-the-documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Adding-a-new-solver-to-the-documentation" title="Permalink"></a></h2><p>Use the following checklist when adding a new solver to the JuMP documentation.</p><pre><code class="nohighlight hljs">## Basic
+
+ - [ ] Check that the solver is a registered Julia package
+ - [ ] Check that the solver supports the long-term support release of Julia
+ - [ ] Check that the solver has a MathOptInterface wrapper
+ - [ ] Check that the tests call `MOI.Test.runtests`. Some test excludes are
+       permissible, but the reason for skipping a particular test should be
+       documented.
+ - [ ] Check that the README and/or documentation provides an example of how to
+       use the solver with JuMP
+
+## Documentation
+
+ - [ ] Add a new row to the table in `docs/src/installation.md`
+
+## Optional
+
+ - [ ] Add package metadata to `docs/packages.toml`</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../roadmap/">« Roadmap</a><a class="docs-footer-nextpage" href="../../packages/solvers/">Introduction »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/developers/contributing/index.html b/previews/PR3525/developers/contributing/index.html
new file mode 100644
index 00000000000..edab9f21482
--- /dev/null
+++ b/previews/PR3525/developers/contributing/index.html
@@ -0,0 +1,28 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Contributing · JuMP</title><meta name="title" content="Contributing · JuMP"/><meta property="og:title" content="Contributing · JuMP"/><meta property="twitter:title" content="Contributing · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox" checked/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Contributing</a><ul class="internal"><li><a class="tocitem" href="#Code-of-Conduct"><span>Code of Conduct</span></a></li><li><a class="tocitem" href="#Join-the-community-forum"><span>Join the community forum</span></a></li><li><a class="tocitem" href="#Join-the-developer-chatroom"><span>Join the developer chatroom</span></a></li><li><a class="tocitem" href="#Improve-the-documentation"><span>Improve the documentation</span></a></li><li><a class="tocitem" href="#File-a-bug-report"><span>File a bug report</span></a></li><li><a class="tocitem" href="#Contribute-code-to-JuMP"><span>Contribute code to JuMP</span></a></li></ul></li><li><a class="tocitem" href="../extensions/">Extensions</a></li><li><a class="tocitem" href="../custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../style/">Style Guide</a></li><li><a class="tocitem" href="../roadmap/">Roadmap</a></li><li><a class="tocitem" href="../checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Developer Docs</a></li><li class="is-active"><a href>Contributing</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Contributing</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/developers/contributing.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="How-to-contribute-to-JuMP"><a class="docs-heading-anchor" href="#How-to-contribute-to-JuMP">How to contribute to JuMP</a><a id="How-to-contribute-to-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-contribute-to-JuMP" title="Permalink"></a></h1><p>Welcome, this document explains some ways you can contribute to JuMP.</p><h2 id="Code-of-Conduct"><a class="docs-heading-anchor" href="#Code-of-Conduct">Code of Conduct</a><a id="Code-of-Conduct-1"></a><a class="docs-heading-anchor-permalink" href="#Code-of-Conduct" title="Permalink"></a></h2><p>This project and everyone participating in it is governed by the <a href="https://github.com/jump-dev/JuMP.jl/blob/master/CODE_OF_CONDUCT.md">JuMP Code of Conduct</a>. By participating, you are expected to uphold this code.</p><h2 id="Join-the-community-forum"><a class="docs-heading-anchor" href="#Join-the-community-forum">Join the community forum</a><a id="Join-the-community-forum-1"></a><a class="docs-heading-anchor-permalink" href="#Join-the-community-forum" title="Permalink"></a></h2><p>First up, join the <a href="https://jump.dev/forum">community forum</a>.</p><p>The forum is a good place to ask questions about how to use JuMP. You can also use the forum to discuss possible feature requests and bugs before raising a GitHub issue (more on this below).</p><p>Aside from asking questions, the easiest way you can contribute to JuMP is to help answer questions on the forum.</p><h2 id="Join-the-developer-chatroom"><a class="docs-heading-anchor" href="#Join-the-developer-chatroom">Join the developer chatroom</a><a id="Join-the-developer-chatroom-1"></a><a class="docs-heading-anchor-permalink" href="#Join-the-developer-chatroom" title="Permalink"></a></h2><p>If you&#39;re interested in contributing code to JuMP, the next place to join is the <a href="https://jump.dev/chatroom">developer chatroom</a>. Let us know what you have in mind, and we can point you in the right direction.</p><h2 id="Improve-the-documentation"><a class="docs-heading-anchor" href="#Improve-the-documentation">Improve the documentation</a><a id="Improve-the-documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Improve-the-documentation" title="Permalink"></a></h2><p>Chances are, if you asked (or answered) a question on the community forum, then it is a sign that the <a href="https://jump.dev/JuMP.jl/dev/">documentation</a> could be improved. Moreover, since it is your question, you are probably the best-placed person to improve it.</p><p>The docs are written in Markdown and are built using <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a>. You can find the source of all the docs <a href="https://github.com/jump-dev/JuMP.jl/tree/master/docs">here</a>.</p><p>If your change is small (like fixing typos, or one or two sentence corrections), the easiest way to do this is via GitHub&#39;s online editor. (GitHub has <a href="https://help.github.com/articles/editing-files-in-another-user-s-repository/">help</a> on how to do this.)</p><p>If your change is larger, or touches multiple files, you will need to make the change locally and then use Git to submit a pull request. (See <a href="#Contribute-code-to-JuMP">Contribute code to JuMP</a> below for more on this.)</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you need any help, come join the <a href="https://jump.dev/chatroom">developer chatroom</a> and we will walk you through the process.</p></div></div><h2 id="File-a-bug-report"><a class="docs-heading-anchor" href="#File-a-bug-report">File a bug report</a><a id="File-a-bug-report-1"></a><a class="docs-heading-anchor-permalink" href="#File-a-bug-report" title="Permalink"></a></h2><p>Another way to contribute to JuMP is to file <a href="https://github.com/jump-dev/JuMP.jl/issues/new?template=bug_report.md">bug reports</a>.</p><p>Make sure you read the info in the box where you write the body of the issue before posting. You can also find a copy of that info <a href="https://github.com/jump-dev/JuMP.jl/blob/master/.github/ISSUE_TEMPLATE/bug_report.md">here</a>.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you&#39;re unsure whether you have a real bug, post on the <a href="https://jump.dev/forum">community forum</a> first. Someone will either help you fix the problem, or let you know the most appropriate place to open a bug report.</p></div></div><h2 id="Contribute-code-to-JuMP"><a class="docs-heading-anchor" href="#Contribute-code-to-JuMP">Contribute code to JuMP</a><a id="Contribute-code-to-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Contribute-code-to-JuMP" title="Permalink"></a></h2><p>Finally, you can also contribute code to JuMP.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>If you do not have experience with Git, GitHub, and Julia development, the first steps can be a little daunting. However, there are lots of tutorials available online, including these for:</p><ul><li><a href="https://guides.github.com/activities/hello-world/">GitHub</a></li><li><a href="https://try.github.io/">Git and GitHub</a></li><li><a href="https://git-scm.com/book/en/v2">Git</a></li><li><a href="https://docs.julialang.org/en/v1/stdlib/Pkg/#Developing-packages-1">Julia package development</a></li></ul><p>If you need any help, come join the <a href="https://jump.dev/chatroom">developer chatroom</a> and we will walk you through the process.</p></div></div><p>Once you are familiar with Git and GitHub, the workflow for contributing code to JuMP is similar to the following:</p><p><strong>Step 1: decide what to work on</strong></p><p>The first step is to find an <a href="https://github.com/jump-dev/JuMP.jl/issues">open issue</a> (or open a new one) for the problem you want to solve. Then, <em>before</em> spending too much time on it, discuss what you are planning to do in the issue to see if other contributors are fine with your proposed changes. Getting feedback early can improve code quality, and avoid time spent writing code that does not get merged into JuMP.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>At this point, remember to be patient and polite; you may get a <em>lot</em> of comments on your issue. However, do not be afraid. Comments mean that people are willing to help you improve the code that you are contributing to JuMP.</p></div></div><p><strong>Step 2: fork JuMP</strong></p><p>Go to <a href="https://github.com/jump-dev/JuMP.jl">https://github.com/jump-dev/JuMP.jl</a> and click the &quot;Fork&quot; button in the top-right corner. This will create a copy of JuMP under your GitHub account.</p><p><strong>Step 3: install JuMP locally</strong></p><p>Open Julia and run:</p><pre><code class="language-julia hljs">] dev JuMP</code></pre><p>This will download the JuMP Git repository to <code>~/.julia/dev/JuMP</code>. If you&#39;re on Windows, this will be <code>C:\\Users\\&lt;my_name&gt;\\.julia\\dev\\JuMP</code>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><code>] command</code> means &quot;first type <code>]</code> to enter the Julia pkg mode, then type the rest. Don&#39;t copy-paste the code directly.</p></div></div><p><strong>Step 4: checkout a new branch</strong></p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>In the following, replace any instance of <code>GITHUB_ACCOUNT</code> with your GitHub user name.</p></div></div><p>The next step is to checkout a development branch. In a terminal (or command prompt on Windows), run:</p><pre><code class="nohighlight hljs">$ cd ~/.julia/dev/JuMP
+
+$ git remote add GITHUB_ACCOUNT https://github.com/GITHUB_ACCOUNT/JuMP.jl.git
+
+$ git checkout master
+
+$ git pull
+
+$ git checkout -b my_new_branch</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Lines starting with <code>$</code> mean &quot;run these in a terminal (command prompt on Windows).&quot;</p></div></div><p><strong>Step 5: make changes</strong></p><p>Now make any changes to the source code inside the <code>~/.julia/dev/JuMP</code> directory.</p><p>Make sure you:</p><ul><li>Follow the <a href="../style/#Style-guide">Style guide</a> and run <a href="../style/#JuliaFormatter">JuliaFormatter</a></li><li>Add tests and documentation for any changes or new features</li></ul><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>When you change the source code, you&#39;ll need to restart Julia for the changes to take effect. This is a pain, so install <a href="https://github.com/timholy/Revise.jl">Revise.jl</a>.</p></div></div><p><strong>Step 6a: test your code changes</strong></p><p>To test that your changes work, run the JuMP test-suite by opening Julia and running:</p><pre><code class="language-julia hljs">cd(&quot;~/.julia/dev/JuMP&quot;)
+] activate .
+] test</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Running the tests might take a long time (~10–15 minutes).</p></div></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you&#39;re using Revise.jl, you can also run the tests by calling <code>include</code>:</p><pre><code class="language-julia hljs">include(&quot;test/runtests.jl&quot;)</code></pre><p>This can be faster if you want to re-run the tests multiple times.</p></div></div><p><strong>Step 6b: test your documentation changes</strong></p><p>Open Julia, then run:</p><pre><code class="language-julia hljs">cd(&quot;~/.julia/dev/JuMP/docs&quot;)
+] activate .
+include(&quot;src/make.jl&quot;)</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Building the documentation might take a long time (~10 minutes).</p></div></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If there&#39;s a problem with the tests that you don&#39;t know how to fix, don&#39;t worry. Continue to step 5, and one of the JuMP contributors will comment on your pull request telling you how to fix things.</p></div></div><p><strong>Step 7: make a pull request</strong></p><p>Once you&#39;ve made changes, you&#39;re ready to push the changes to GitHub. Run:</p><pre><code class="nohighlight hljs">$ cd ~/.julia/dev/JuMP
+
+$ git add .
+
+$ git commit -m &quot;A descriptive message of the changes&quot;
+
+$ git push -u GITHUB_ACCOUNT my_new_branch</code></pre><p>Then go to <a href="https://github.com/jump-dev/JuMP.jl">https://github.com/jump-dev/JuMP.jl</a> and follow the instructions that pop up to open a pull request.</p><p><strong>Step 8: respond to comments</strong></p><p>At this point, remember to be patient and polite; you may get a <em>lot</em> of comments on your pull request. However, do not be afraid. A lot of comments means that people are willing to help you improve the code that you are contributing to JuMP.</p><p>To respond to the comments, go back to step 5, make any changes, test the changes in step 6, and then make a new commit in step 7. Your PR will automatically update.</p><p><strong>Step 9: cleaning up</strong></p><p>Once the PR is merged, clean-up your Git repository ready for the next contribution.</p><pre><code class="nohighlight hljs">$ cd ~/.julia/dev/JuMP
+
+$ git checkout master
+
+$ git pull</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>If you have suggestions to improve this guide, please make a pull request. It&#39;s particularly helpful if you do this after your first pull request because you&#39;ll know all the parts that could be explained better.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../background/algebraic_modeling_languages/">« Algebraic modeling languages</a><a class="docs-footer-nextpage" href="../extensions/">Extensions »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/developers/custom_solver_binaries/index.html b/previews/PR3525/developers/custom_solver_binaries/index.html
new file mode 100644
index 00000000000..d2959e99773
--- /dev/null
+++ b/previews/PR3525/developers/custom_solver_binaries/index.html
@@ -0,0 +1,93 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Custom binaries · JuMP</title><meta name="title" content="Custom binaries · JuMP"/><meta property="og:title" content="Custom binaries · JuMP"/><meta property="twitter:title" content="Custom binaries · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox" checked/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../contributing/">Contributing</a></li><li><a class="tocitem" href="../extensions/">Extensions</a></li><li class="is-active"><a class="tocitem" href>Custom binaries</a><ul class="internal"><li><a class="tocitem" href="#Background"><span>Background</span></a></li><li><a class="tocitem" href="#jll_structure"><span>Explore the JLL you want to override</span></a></li><li><a class="tocitem" href="#compile_ecos"><span>Compile a custom binary</span></a></li><li><a class="tocitem" href="#Overriding-a-single-library"><span>Overriding a single library</span></a></li><li><a class="tocitem" href="#Overriding-an-entire-artifact"><span>Overriding an entire artifact</span></a></li><li><a class="tocitem" href="#Using-Cbc-with-a-custom-binary"><span>Using Cbc with a custom binary</span></a></li></ul></li><li><a class="tocitem" href="../style/">Style Guide</a></li><li><a class="tocitem" href="../roadmap/">Roadmap</a></li><li><a class="tocitem" href="../checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Developer Docs</a></li><li class="is-active"><a href>Custom binaries</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Custom binaries</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/developers/custom_solver_binaries.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="How-to-use-a-custom-binary"><a class="docs-heading-anchor" href="#How-to-use-a-custom-binary">How to use a custom binary</a><a id="How-to-use-a-custom-binary-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-use-a-custom-binary" title="Permalink"></a></h1><p>Many solvers are not written in Julia, but instead in languages like C or C++. JuMP interacts with these solvers through binary dependencies.</p><p>For many open-source solvers, we automatically install the appropriate binary when you run <code>Pkg.add(&quot;Solver&quot;)</code>. For example, <code>Pkg.add(&quot;ECOS&quot;)</code> will also install the ECOS binary.</p><p>This page explains how this installation works, and how you can use a custom binary.</p><div class="admonition is-compat"><header class="admonition-header">Compat</header><div class="admonition-body"><p>These instructions require Julia 1.6 or later.</p></div></div><h2 id="Background"><a class="docs-heading-anchor" href="#Background">Background</a><a id="Background-1"></a><a class="docs-heading-anchor-permalink" href="#Background" title="Permalink"></a></h2><p>Each solver that JuMP supports is structured as a Julia package. For example, the interface for the <a href="https://github.com/embotech/ecos">ECOS</a> solver is provided by the <a href="https://github.com/jump-dev/ECOS.jl">ECOS.jl</a> package.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This page uses the example of ECOS.jl because it is simple to compile. Other solvers follow similar conventions. For example, the interface to the Clp solver is provided by Clp.jl.</p></div></div><p>The ECOS.jl package provides an interface between the C API of ECOS and MathOptInterface. However, it does not handle the installation of the solver binary; that is the job for a JLL package.</p><p>A JLL is a Julia package that wraps a pre-compiled binary. Binaries are built using <a href="https://github.com/JuliaPackaging/Yggdrasil">Yggdrasil</a> (for example, <a href="https://github.com/JuliaPackaging/Yggdrasil/blob/master/E/ECOS/build_tarballs.jl">ECOS</a>) and hosted in the <a href="https://github.com/JuliaBinaryWrappers">JuliaBinaryWrappers</a> GitHub repository (for example, <a href="https://github.com/JuliaBinaryWrappers/ECOS_jll.jl">ECOS_jll.jl</a>).</p><p>JLL packages contain little code. Their only job is to <code>dlopen</code> a dynamic library, along with any dependencies.</p><p>JLL packages manage their binary dependencies using <a href="https://pkgdocs.julialang.org/v1/artifacts/">Julia&#39;s artifact system</a>. Each JLL package has an <code>Artifacts.toml</code> file which describes where to find each binary artifact for each different platform that it might be installed on. Here is the <a href="https://github.com/JuliaBinaryWrappers/ECOS_jll.jl/blob/main/Artifacts.toml">Artifacts.toml file for ECOS_jll.jl</a>.</p><p>The binaries installed by the JLL package should be sufficient for most users. In rare cases, however, you may require a custom binary. The two main reasons to use a custom binary are:</p><ul><li>You want a binary with custom compilation settings (for example, debugging)</li><li>You want a binary with a set of dependencies that are not available on Yggdrasil (for example, a commercial solver like Gurobi or CPLEX).</li></ul><p>The following sections explain how to replace the binaries provided by a JLL package with the custom ones you have compiled. As a reminder, we use ECOS as an example for simplicity, but the steps are the same for other solvers.</p><h2 id="jll_structure"><a class="docs-heading-anchor" href="#jll_structure">Explore the JLL you want to override</a><a id="jll_structure-1"></a><a class="docs-heading-anchor-permalink" href="#jll_structure" title="Permalink"></a></h2><p>The first step is to explore the structure and filenames of the JLL package we want to override.</p><p>Find the location of the files using <code>.artifact_dir</code>:</p><pre><code class="language-julia hljs">julia&gt; using ECOS_jll
+
+julia&gt; ECOS_jll.artifact_dir
+&quot;/Users/oscar/.julia/artifacts/2addb75332eff5a1657b46bb6bf30d2410bc7ecf&quot;</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This path may be different on other machines.</p></div></div><p>Here is what it contains:</p><pre><code class="language-julia hljs">julia&gt; readdir(ECOS_jll.artifact_dir)
+4-element Vector{String}:
+ &quot;include&quot;
+ &quot;lib&quot;
+ &quot;logs&quot;
+ &quot;share&quot;
+
+julia&gt; readdir(joinpath(ECOS_jll.artifact_dir, &quot;lib&quot;))
+1-element Vector{String}:
+ &quot;libecos.dylib&quot;</code></pre><p>Other solvers may have a <code>bin</code> directory containing executables. To use a custom binary of ECOS, we need to replace <code>/lib/libecos.dylib</code> with our custom binary.</p><h2 id="compile_ecos"><a class="docs-heading-anchor" href="#compile_ecos">Compile a custom binary</a><a id="compile_ecos-1"></a><a class="docs-heading-anchor-permalink" href="#compile_ecos" title="Permalink"></a></h2><p>The next step is to compile a custom binary. Because ECOS is written in C with no dependencies, this is easy to do if you have a C compiler:</p><pre><code class="language-julia hljs">oscar@Oscars-MBP jll_example % git clone https://github.com/embotech/ecos.git
+[... lines omitted ...]
+oscar@Oscars-MBP jll_example % cd ecos
+oscar@Oscars-MBP ecos % make shared
+[... many lines omitted...]
+oscar@Oscars-MBP ecos % mkdir lib
+oscar@Oscars-MBP ecos % cp libecos.dylib lib</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Compiling custom solver binaries is an advanced operation. Due to the complexities of compiling various solvers, the JuMP community is unable to help you diagnose and fix compilation issues.</p></div></div><p>After this compilation step, we now have a folder <code>/tmp/jll_example/ecos</code> that contains <code>lib</code> and <code>include</code> directories with the same files as <code>ECOS_jll</code>:</p><pre><code class="language-julia hljs">julia&gt; readdir(joinpath(&quot;ecos&quot;, &quot;lib&quot;))
+1-element Vector{String}:
+ &quot;libecos.dylib&quot;</code></pre><h2 id="Overriding-a-single-library"><a class="docs-heading-anchor" href="#Overriding-a-single-library">Overriding a single library</a><a id="Overriding-a-single-library-1"></a><a class="docs-heading-anchor-permalink" href="#Overriding-a-single-library" title="Permalink"></a></h2><p>To override the <code>libecos</code> library, we need to know what <code>ECOS_jll</code> calls it. (In most cases, it will also be <code>libecos</code>, but not always.)</p><p>There are two ways you can check.</p><ol><li>Check the bottom of the JLL&#39;s GitHub README. For example, <a href="https://github.com/JuliaBinaryWrappers/ECOS_jll.jl#products">ECOS_jll</a> has a single <code>LibraryProduct</code> called <code>libecos</code>.</li><li>Type <code>ECOS_jll.</code> and the press the <code>[TAB]</code> key twice to auto-complete available options:<pre><code class="language-julia hljs">julia&gt; ECOS_jll.
+LIBPATH           PATH_list          best_wrapper       get_libecos_path   libecos_handle
+LIBPATH_list      __init__           dev_jll            is_available       libecos_path
+PATH              artifact_dir       find_artifact_dir  libecos</code></pre>Here you can see there is <code>libecos</code>, and more usefully for us, <code>libecos_path</code>.</li></ol><p>Once you know the name of the variable to override (the one that ends in <code>_path</code>), use  <a href="https://github.com/JuliaPackaging/Preferences.jl">Preferences.jl</a> to specify a new path:</p><pre><code class="language-julia hljs">using Preferences
+set_preferences!(
+    &quot;LocalPreferences.toml&quot;,
+    &quot;ECOS_jll&quot;,
+    &quot;libecos_path&quot; =&gt; &quot;/tmp/jll_example/ecos/lib/libecos&quot;
+)</code></pre><p>This will create a file in your current directory called <code>LocalPreferences.toml</code> with the contents:</p><pre><code class="language-julia hljs">[ECOS_jll]
+libecos_path = &quot;/tmp/jll_example/ecos/lib/libecos&quot;</code></pre><p>Now if you restart Julia, you will see:</p><pre><code class="language-julia hljs">julia&gt; using ECOS_jll
+
+julia&gt; ECOS_jll.libecos
+&quot;/tmp/jll_example/ecos/lib/libecos&quot;</code></pre><p>To go back to using the default library, just delete the <code>LocalPreferences.toml</code> file.</p><h2 id="Overriding-an-entire-artifact"><a class="docs-heading-anchor" href="#Overriding-an-entire-artifact">Overriding an entire artifact</a><a id="Overriding-an-entire-artifact-1"></a><a class="docs-heading-anchor-permalink" href="#Overriding-an-entire-artifact" title="Permalink"></a></h2><p>Sometimes a solver may provide a number of libraries and executables, and specifying the path for each of the becomes tedious. In this case, we can use Julia&#39;s <code>Override.toml</code> to replace an entire artifact.</p><p>Overriding an entire artifact requires you to replicate the structure and contents of the JLL package that we <a href="#jll_structure">explored above</a>.</p><p>In most cases you need only reproduce the <code>include</code>, <code>lib</code>, and <code>bin</code> directories (if they exist). You can safely ignore any <code>logs</code> or <code>share</code> directories. Take careful note of what files each directory contains and what they are called.</p><p>For our ECOS example, we already reproduced the structure when we <a href="#compile_ecos">compiled ECOS</a>.</p><p>So, now we need to tell Julia to use our custom installation instead of the default. We can do this by making an override file at <code>~/.julia/artifacts/Overrides.toml</code>.</p><p><code>Overrides.toml</code> has the following content:</p><pre><code class="language-julia hljs"># Override for ECOS_jll
+2addb75332eff5a1657b46bb6bf30d2410bc7ecf = &quot;/tmp/jll_example/ecos&quot;</code></pre><p>where <code>2addb75332eff5a1657b46bb6bf30d2410bc7ecf</code> is the folder from the original <code>ECOS_jll.artifact_dir</code> and <code>&quot;/tmp/jll_example/ecos&quot;</code> is the location of our new installation. Replace these as appropriate for your system.</p><p>If you restart Julia after creating the override file, you will see:</p><pre><code class="language-julia hljs">julia&gt; using ECOS_jll
+
+julia&gt; ECOS_jll.artifact_dir
+&quot;/tmp/jll_example/ecos&quot;</code></pre><p>Now when we use ECOS it will use our custom binary.</p><h2 id="Using-Cbc-with-a-custom-binary"><a class="docs-heading-anchor" href="#Using-Cbc-with-a-custom-binary">Using Cbc with a custom binary</a><a id="Using-Cbc-with-a-custom-binary-1"></a><a class="docs-heading-anchor-permalink" href="#Using-Cbc-with-a-custom-binary" title="Permalink"></a></h2><p>As a second example, we demonstrate how to use <a href="https://github.com/jump-dev/Cbc.jl">Cbc.jl</a> with a custom binary.</p><h3 id="Explore-the-JLL-you-want-to-override"><a class="docs-heading-anchor" href="#Explore-the-JLL-you-want-to-override">Explore the JLL you want to override</a><a id="Explore-the-JLL-you-want-to-override-1"></a><a class="docs-heading-anchor-permalink" href="#Explore-the-JLL-you-want-to-override" title="Permalink"></a></h3><p>First, let&#39;s check where <code>Cbc_jll</code> is installed:</p><pre><code class="language-julia hljs">julia&gt; using Cbc_jll
+
+julia&gt; Cbc_jll.artifact_dir
+&quot;/Users/oscar/.julia/artifacts/e481bc81db5e229ba1f52b2b4bd57484204b1b06&quot;
+
+julia&gt; readdir(Cbc_jll.artifact_dir)
+5-element Vector{String}:
+ &quot;bin&quot;
+ &quot;include&quot;
+ &quot;lib&quot;
+ &quot;logs&quot;
+ &quot;share&quot;
+
+julia&gt; readdir(joinpath(Cbc_jll.artifact_dir, &quot;bin&quot;))
+1-element Vector{String}:
+ &quot;cbc&quot;
+
+julia&gt; readdir(joinpath(Cbc_jll.artifact_dir, &quot;lib&quot;))
+10-element Vector{String}:
+ &quot;libCbc.3.10.5.dylib&quot;
+ &quot;libCbc.3.dylib&quot;
+ &quot;libCbc.dylib&quot;
+ &quot;libCbcSolver.3.10.5.dylib&quot;
+ &quot;libCbcSolver.3.dylib&quot;
+ &quot;libCbcSolver.dylib&quot;
+ &quot;libOsiCbc.3.10.5.dylib&quot;
+ &quot;libOsiCbc.3.dylib&quot;
+ &quot;libOsiCbc.dylib&quot;
+ &quot;pkgconfig&quot;</code></pre><h3 id="Compile-a-custom-binary"><a class="docs-heading-anchor" href="#Compile-a-custom-binary">Compile a custom binary</a><a id="Compile-a-custom-binary-1"></a><a class="docs-heading-anchor-permalink" href="#Compile-a-custom-binary" title="Permalink"></a></h3><p>Next, we need to compile Cbc. Cbc can be difficult to compile (it has a lot of dependencies), but for macOS users there is a homebrew recipe:</p><pre><code class="nohighlight hljs">(base) oscar@Oscars-MBP jll_example % brew install cbc
+[ ... lines omitted ... ]
+(base) oscar@Oscars-MBP jll_example % brew list cbc
+/usr/local/Cellar/cbc/2.10.5/bin/cbc
+/usr/local/Cellar/cbc/2.10.5/include/cbc/ (76 files)
+/usr/local/Cellar/cbc/2.10.5/lib/libCbc.3.10.5.dylib
+/usr/local/Cellar/cbc/2.10.5/lib/libCbcSolver.3.10.5.dylib
+/usr/local/Cellar/cbc/2.10.5/lib/libOsiCbc.3.10.5.dylib
+/usr/local/Cellar/cbc/2.10.5/lib/pkgconfig/ (2 files)
+/usr/local/Cellar/cbc/2.10.5/lib/ (6 other files)
+/usr/local/Cellar/cbc/2.10.5/share/cbc/ (59 files)
+/usr/local/Cellar/cbc/2.10.5/share/coin/ (4 files)</code></pre><h3 id="Override-single-libraries"><a class="docs-heading-anchor" href="#Override-single-libraries">Override single libraries</a><a id="Override-single-libraries-1"></a><a class="docs-heading-anchor-permalink" href="#Override-single-libraries" title="Permalink"></a></h3><p>To use <code>Preferences.jl</code> to override specific libraries we first check the names of each library in <code>Cbc_jll</code>:</p><pre><code class="language-julia hljs">julia&gt; Cbc_jll.
+LIBPATH               cbc                    get_libcbcsolver_path  libOsiCbc_path
+LIBPATH_list          cbc_path               is_available           libcbcsolver
+PATH                  dev_jll                libCbc                 libcbcsolver_handle
+PATH_list             find_artifact_dir      libCbc_handle          libcbcsolver_path
+__init__              get_cbc_path           libCbc_path
+artifact_dir          get_libCbc_path        libOsiCbc
+best_wrapper          get_libOsiCbc_path     libOsiCbc_handle</code></pre><p>Then we add the following to <code>LocalPreferences.toml</code>:</p><pre><code class="language-julia hljs">[Cbc_jll]
+cbc_path = &quot;/usr/local/Cellar/cbc/2.10.5/bin/cbc&quot;
+libCbc_path = &quot;/usr/local/Cellar/cbc/2.10.5/lib/libCbc.3.10.5&quot;
+libOsiCbc_path = &quot;/usr/local/Cellar/cbc/2.10.5/lib/libOsiCbc.3.10.5&quot;
+libcbcsolver_path = &quot;/usr/local/Cellar/cbc/2.10.5/lib/libCbcSolver.3.10.5&quot;</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Note that capitalization matters, so <code>libcbcsolver_path</code> corresponds to <code>libCbcSolver.3.10.5</code>.</p></div></div><h3 id="Override-entire-artifact"><a class="docs-heading-anchor" href="#Override-entire-artifact">Override entire artifact</a><a id="Override-entire-artifact-1"></a><a class="docs-heading-anchor-permalink" href="#Override-entire-artifact" title="Permalink"></a></h3><p>To use the homebrew install as our custom binary we add the following to <code>~/.julia/artifacts/Overrides.toml</code>:</p><pre><code class="language-julia hljs"># Override for Cbc_jll
+e481bc81db5e229ba1f52b2b4bd57484204b1b06 = &quot;/usr/local/Cellar/cbc/2.10.5&quot;</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../extensions/">« Extensions</a><a class="docs-footer-nextpage" href="../style/">Style Guide »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/developers/extensions/index.html b/previews/PR3525/developers/extensions/index.html
new file mode 100644
index 00000000000..8d828eb5f31
--- /dev/null
+++ b/previews/PR3525/developers/extensions/index.html
@@ -0,0 +1,304 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Extensions · JuMP</title><meta name="title" content="Extensions · JuMP"/><meta property="og:title" content="Extensions · JuMP"/><meta property="twitter:title" content="Extensions · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox" checked/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../contributing/">Contributing</a></li><li class="is-active"><a class="tocitem" href>Extensions</a><ul class="internal"><li><a class="tocitem" href="#Compatibility"><span>Compatibility</span></a></li><li><a class="tocitem" href="#Define-a-new-set"><span>Define a new set</span></a></li><li><a class="tocitem" href="#extend_variable_macro"><span>Extend <code>@variable</code></span></a></li><li><a class="tocitem" href="#Extend-[@constraint](@ref)"><span>Extend <code>@constraint</code></span></a></li><li><a class="tocitem" href="#The-extension-dictionary"><span>The extension dictionary</span></a></li><li><a class="tocitem" href="#Defining-new-JuMP-models"><span>Defining new JuMP models</span></a></li><li><a class="tocitem" href="#Set-an-optimize!-hook"><span>Set an <code>optimize!</code> hook</span></a></li><li><a class="tocitem" href="#Creating-new-container-types"><span>Creating new container types</span></a></li><li><a class="tocitem" href="#Performance-tips-for-extensions"><span>Performance tips for extensions</span></a></li></ul></li><li><a class="tocitem" href="../custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../style/">Style Guide</a></li><li><a class="tocitem" href="../roadmap/">Roadmap</a></li><li><a class="tocitem" href="../checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Developer Docs</a></li><li class="is-active"><a href>Extensions</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Extensions</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/developers/extensions.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="extensions_manual"><a class="docs-heading-anchor" href="#extensions_manual">Extensions</a><a id="extensions_manual-1"></a><a class="docs-heading-anchor-permalink" href="#extensions_manual" title="Permalink"></a></h1><p>JuMP provides a variety of ways to extend the basic modeling functionality.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This documentation in this section is still a work-in-progress. The best place to look for ideas and help when writing a new JuMP extension are existing JuMP extensions. Examples include:</p><ul><li><a href="https://github.com/joaquimg/BilevelJuMP.jl">BilevelJuMP.jl</a></li><li><a href="https://github.com/atoptima/Coluna.jl">Coluna.jl</a></li><li><a href="https://github.com/infiniteopt/InfiniteOpt.jl">InfiniteOpt.jl</a></li><li><a href="https://github.com/zavalab/Plasmo.jl">Plasmo.jl</a></li><li><a href="https://github.com/jump-dev/PolyJuMP.jl">PolyJuMP.jl</a></li><li><a href="https://github.com/odow/SDDP.jl">SDDP.jl</a></li><li><a href="https://github.com/martinbiel/StochasticPrograms.jl">StochasticPrograms.jl</a></li><li><a href="https://github.com/jump-dev/SumOfSquares.jl">SumOfSquares.jl</a></li><li><a href="https://github.com/vOptSolver/vOptGeneric.jl">vOptGeneric.jl</a></li></ul></div></div><h2 id="Compatibility"><a class="docs-heading-anchor" href="#Compatibility">Compatibility</a><a id="Compatibility-1"></a><a class="docs-heading-anchor-permalink" href="#Compatibility" title="Permalink"></a></h2><p>When writing JuMP extensions, you should carefully consider the compatibility guarantees that JuMP makes. In particular:</p><ul><li>All functions, structs, and constants which do not begin with an underscore (<code>_</code>) are public. These are always safe to use, and they should all have corresponding documentation.</li><li>All identifiers beginning with an underscore (<code>_</code>) are private. These are not safe to use, because they may break in any JuMP release, including patch releases.</li><li>Unless explicitly mentioned in the documentation, all fields of a struct are private. These are not safe to use, because they may break in any JuMP release, including patch releases. An example of a field which is safe to use is the <code>model.ext</code> extension dictionary, which is documented in <a href="#The-extension-dictionary">The extension dictionary</a>.</li></ul><p>In general, we strongly encourage you to use only the public API of JuMP. If you are missing a feature, please open a GitHub issue.</p><p>However, if you <em>do</em> use the private API (for example, because your feature request has not been implemented yet), then you must carefully restrict the versions of JuMP that your package is compatible with in the <code>Project.toml</code> file. The easiest way to do this is via the <a href="https://pkgdocs.julialang.org/v1/compatibility/">hyphen specifiers</a>. For example, if your package supports all JuMP versions between v1.0.0 and v1.1.1, do:</p><pre><code class="nohighlight hljs">JuMP = &quot;1.0.0 - 1.1.1&quot;</code></pre><p>Then, whenever JuMP releases a new version, you should check if your package is still compatible and update the bound accordingly.</p><h2 id="Define-a-new-set"><a class="docs-heading-anchor" href="#Define-a-new-set">Define a new set</a><a id="Define-a-new-set-1"></a><a class="docs-heading-anchor-permalink" href="#Define-a-new-set" title="Permalink"></a></h2><p>To define a new set for JuMP, subtype <code>MOI.AbstractScalarSet</code> or <code>MOI.AbstractVectorSet</code> and implement <code>Base.copy</code> for the set.</p><pre><code class="language-julia-repl hljs">julia&gt; struct NewMOIVectorSet &lt;: MOI.AbstractVectorSet
+           dimension::Int
+       end
+
+julia&gt; Base.copy(x::NewMOIVectorSet) = x
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @constraint(model, x in NewMOIVectorSet(2))
+[x[1], x[2]] ∈ NewMOIVectorSet(2)</code></pre><p>However, for vector-sets, this requires the user to specify the dimension argument to their set, even though we could infer it from the length of <code>x</code>!</p><p>You can make a more user-friendly set by subtyping <a href="../../api/JuMP/#JuMP.AbstractVectorSet"><code>AbstractVectorSet</code></a> and implementing <a href="../../api/JuMP/#JuMP.moi_set"><code>moi_set</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; struct NewVectorSet &lt;: JuMP.AbstractVectorSet end
+
+julia&gt; JuMP.moi_set(::NewVectorSet, dim::Int) = NewMOIVectorSet(dim)
+
+julia&gt; @constraint(model, x in NewVectorSet())
+[x[1], x[2]] ∈ NewMOIVectorSet(2)</code></pre><h2 id="extend_variable_macro"><a class="docs-heading-anchor" href="#extend_variable_macro">Extend <code>@variable</code></a><a id="extend_variable_macro-1"></a><a class="docs-heading-anchor-permalink" href="#extend_variable_macro" title="Permalink"></a></h2><p>Just as <code>Bin</code> and <code>Int</code> create binary and integer variables, you can extend the <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro to create new types of variables. Here is an explanation by example, where we create a <code>AddTwice</code> type, that creates a tuple of two JuMP variables instead of a single variable.</p><p>First, create a new struct. This can be anything. Our struct holds a <a href="../../api/JuMP/#JuMP.VariableInfo"><code>VariableInfo</code></a> object that stores bound information, and whether the variable is binary or integer.</p><pre><code class="language-julia-repl hljs">julia&gt; struct AddTwice
+           info::JuMP.VariableInfo
+       end</code></pre><p>Second, implement <a href="../../api/JuMP/#JuMP.build_variable"><code>build_variable</code></a>, which takes <code>::Type{AddTwice}</code> as an argument, and returns an instance of <code>AddTwice</code>. Note that you can also receive keyword arguments.</p><pre><code class="language-julia-repl hljs">julia&gt; function JuMP.build_variable(
+           _err::Function,
+           info::JuMP.VariableInfo,
+           ::Type{AddTwice};
+           kwargs...
+       )
+           println(&quot;Can also use $kwargs here.&quot;)
+           return AddTwice(info)
+       end</code></pre><p>Third, implement <a href="../../api/JuMP/#JuMP.add_variable"><code>add_variable</code></a>, which takes the instance of <code>AddTwice</code> from the previous step, and returns something. Typically, you will want to call <a href="../../api/JuMP/#JuMP.add_variable"><code>add_variable</code></a> here. For example, our <code>AddTwice</code> call is going to add two JuMP variables.</p><pre><code class="language-julia-repl hljs">julia&gt; function JuMP.add_variable(
+           model::JuMP.Model,
+           duplicate::AddTwice,
+           name::String,
+       )
+           a = JuMP.add_variable(
+               model,
+               JuMP.ScalarVariable(duplicate.info),
+               &quot;$(name)_a&quot;,
+            )
+           b = JuMP.add_variable(
+               model,
+               JuMP.ScalarVariable(duplicate.info),
+               &quot;$(name)_b&quot;,
+            )
+           return (a, b)
+       end</code></pre><p>Now <code>AddTwice</code> can be passed to <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> similar to <code>Bin</code> or <code>Int</code>, or through the <code>variable_type</code> keyword. However, now it adds two variables instead of one.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2], variable_type = AddTwice, kw = i)
+Can also use Base.Pairs(:kw =&gt; 1) here.
+Can also use Base.Pairs(:kw =&gt; 2) here.
+2-element Vector{Tuple{VariableRef, VariableRef}}:
+ (x[1]_a, x[1]_b)
+ (x[2]_a, x[2]_b)
+
+julia&gt; num_variables(model)
+4
+
+julia&gt; first(x[1])
+x[1]_a
+
+julia&gt; last(x[2])
+x[2]_b</code></pre><h2 id="Extend-[@constraint](@ref)"><a class="docs-heading-anchor" href="#Extend-[@constraint](@ref)">Extend <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a></a><a id="Extend-[@constraint](@ref)-1"></a><a class="docs-heading-anchor-permalink" href="#Extend-[@constraint](@ref)" title="Permalink"></a></h2><p>The <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro has three steps that can be intercepted and extended: parse time, build time, and add time.</p><h3 id="Parse"><a class="docs-heading-anchor" href="#Parse">Parse</a><a id="Parse-1"></a><a class="docs-heading-anchor-permalink" href="#Parse" title="Permalink"></a></h3><p>To extend the <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro at parse time, implement one of the following methods:</p><ul><li><a href="../../api/JuMP/#JuMP.parse_constraint_head"><code>parse_constraint_head</code></a></li><li><a href="../../api/JuMP/#JuMP.parse_constraint_call"><code>parse_constraint_call</code></a></li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Extending the constraint macro at parse time is an advanced operation and has the potential to interfere with existing JuMP syntax. Please discuss with the <a href="https://gitter.im/JuliaOpt/jump-dev">developer chatroom</a> before publishing any code that implements these methods.</p></div></div><p><a href="../../api/JuMP/#JuMP.parse_constraint_head"><code>parse_constraint_head</code></a> should be implemented to intercept an expression based on the <code>.head</code> field of <code>Base.Expr</code>. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; using JuMP
+
+julia&gt; const MutableArithmetics = JuMP._MA;
+
+julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; function JuMP.parse_constraint_head(
+           _error::Function,
+           ::Val{:(:=)},
+           lhs,
+           rhs,
+       )
+           println(&quot;Rewriting := as ==&quot;)
+           new_lhs, parse_code = MutableArithmetics.rewrite(lhs)
+           build_code = :(
+               build_constraint($(_error), $(new_lhs), MOI.EqualTo($(rhs)))
+           )
+           return false, parse_code, build_code
+       end
+
+julia&gt; @constraint(model, x + x := 1.0)
+Rewriting := as ==
+2 x = 1</code></pre><p><a href="../../api/JuMP/#JuMP.parse_constraint_call"><code>parse_constraint_call</code></a> should be implemented to intercept an expression of the form <code>Expr(:call, op, args...)</code>. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; using JuMP
+
+julia&gt; const MutableArithmetics = JuMP._MA;
+
+julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; function JuMP.parse_constraint_call(
+           _error::Function,
+           is_vectorized::Bool,
+           ::Val{:my_equal_to},
+           lhs,
+           rhs,
+       )
+           println(&quot;Rewriting my_equal_to to ==&quot;)
+           new_lhs, parse_code = MutableArithmetics.rewrite(lhs)
+           build_code = if is_vectorized
+               :(build_constraint($(_error), $(new_lhs), MOI.EqualTo($(rhs)))
+           )
+           else
+               :(build_constraint.($(_error), $(new_lhs), MOI.EqualTo($(rhs))))
+           end
+           return parse_code, build_code
+       end
+
+julia&gt; @constraint(model, my_equal_to(x + x, 1.0))
+Rewriting my_equal_to to ==
+2 x = 1</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>When parsing a constraint you can recurse into sub-constraint (for example, the <code>{expr}</code> in <code>z =&gt; {x &lt;= 1}</code>) by calling <a href="../../api/JuMP/#JuMP.parse_constraint"><code>parse_constraint</code></a>.</p></div></div><h3 id="Build"><a class="docs-heading-anchor" href="#Build">Build</a><a id="Build-1"></a><a class="docs-heading-anchor-permalink" href="#Build" title="Permalink"></a></h3><p>To extend the <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro at build time, implement a new <a href="../../api/JuMP/#JuMP.build_constraint"><code>build_constraint</code></a> method.</p><p>This may mean implementing a method for a specific function or set created at parse time, or it may mean implementing a method which handles additional positional arguments.</p><p><a href="../../api/JuMP/#JuMP.build_constraint"><code>build_constraint</code></a> must return an <a href="../../api/JuMP/#JuMP.AbstractConstraint"><code>AbstractConstraint</code></a>, which can either be an <a href="../../api/JuMP/#JuMP.AbstractConstraint"><code>AbstractConstraint</code></a> already supported by JuMP, for example, <code>ScalarConstraint</code> or <code>VectorConstraint</code>, or a custom <a href="../../api/JuMP/#JuMP.AbstractConstraint"><code>AbstractConstraint</code></a> with a corresponding <a href="../../api/JuMP/#JuMP.add_constraint"><code>add_constraint</code></a> method (see <a href="#extension_add_constraint">Add</a>).</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>The easiest way to extend <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> is via an additional positional argument to <a href="../../api/JuMP/#JuMP.build_constraint"><code>build_constraint</code></a>.</p></div></div><p>Here is an example of adding extra arguments to <a href="../../api/JuMP/#JuMP.build_constraint"><code>build_constraint</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; struct MyConstrType end
+
+julia&gt; function JuMP.build_constraint(
+            _error::Function,
+            f::JuMP.GenericAffExpr,
+            set::MOI.EqualTo,
+            extra::Type{MyConstrType};
+            d = 0,
+       )
+            new_set = MOI.LessThan(set.value + d)
+            return JuMP.build_constraint(_error, f, new_set)
+       end
+
+julia&gt; @constraint(model, my_con, x == 0, MyConstrType, d = 2)
+my_con : x ≤ 2</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Only a single positional argument can be given to a particular constraint. Extensions that seek to pass multiple arguments (for example, <code>Foo</code> and <code>Bar</code>) should combine them into one argument type (for example, <code>FooBar</code>).</p></div></div><h3 id="extension_add_constraint"><a class="docs-heading-anchor" href="#extension_add_constraint">Add</a><a id="extension_add_constraint-1"></a><a class="docs-heading-anchor-permalink" href="#extension_add_constraint" title="Permalink"></a></h3><p><a href="../../api/JuMP/#JuMP.build_constraint"><code>build_constraint</code></a> returns an <a href="../../api/JuMP/#JuMP.AbstractConstraint"><code>AbstractConstraint</code></a> object. To extend <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> at add time, define a subtype of <a href="../../api/JuMP/#JuMP.AbstractConstraint"><code>AbstractConstraint</code></a>, implement <a href="../../api/JuMP/#JuMP.build_constraint"><code>build_constraint</code></a> to return an instance of the new type, and then implement <a href="../../api/JuMP/#JuMP.add_constraint"><code>add_constraint</code></a>.</p><p>Here is an example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; struct MyTag
+           name::String
+       end
+
+julia&gt; struct MyConstraint{S} &lt;: AbstractConstraint
+           name::String
+           f::AffExpr
+           s::S
+       end
+
+julia&gt; function JuMP.build_constraint(
+            _error::Function,
+            f::AffExpr,
+            set::MOI.AbstractScalarSet,
+            extra::MyTag,
+       )
+            return MyConstraint(extra.name, f, set)
+       end
+
+julia&gt; function JuMP.add_constraint(
+            model::Model,
+            con::MyConstraint,
+            name::String,
+       )
+            return add_constraint(
+                model,
+                ScalarConstraint(con.f, con.s),
+                &quot;$(con.name)[$(name)]&quot;,
+            )
+       end
+
+julia&gt; @constraint(model, my_con, 2x &lt;= 1, MyTag(&quot;my_prefix&quot;))
+my_prefix[my_con] : 2 x - 1 ≤ 0</code></pre><h2 id="The-extension-dictionary"><a class="docs-heading-anchor" href="#The-extension-dictionary">The extension dictionary</a><a id="The-extension-dictionary-1"></a><a class="docs-heading-anchor-permalink" href="#The-extension-dictionary" title="Permalink"></a></h2><p>Every JuMP model has a field <code>.ext::Dict{Symbol,Any}</code> that can be used by extensions. This is useful if your extensions to <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> and <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> need to store information between calls.</p><p>The most common way to initialize a model with information in the <code>.ext</code> dictionary is to provide a new constructor:</p><pre><code class="language-julia-repl hljs">julia&gt; function MyModel()
+           model = Model()
+           model.ext[:MyModel] = 1
+           return model
+       end
+MyModel (generic function with 1 method)
+
+julia&gt; model = MyModel()
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+
+julia&gt; model.ext
+Dict{Symbol, Any} with 1 entry:
+  :MyModel =&gt; 1</code></pre><p>If you define extension data, implement <a href="../../api/JuMP/#JuMP.copy_extension_data"><code>copy_extension_data</code></a> to support <a href="../../api/JuMP/#JuMP.copy_model"><code>copy_model</code></a>.</p><h2 id="Defining-new-JuMP-models"><a class="docs-heading-anchor" href="#Defining-new-JuMP-models">Defining new JuMP models</a><a id="Defining-new-JuMP-models-1"></a><a class="docs-heading-anchor-permalink" href="#Defining-new-JuMP-models" title="Permalink"></a></h2><p>If extending individual calls to <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> and <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> is not sufficient, it is possible to implement a new model via a subtype of <a href="../../api/JuMP/#JuMP.AbstractModel"><code>AbstractModel</code></a>. You can also define new <a href="../../api/JuMP/#JuMP.AbstractVariableRef"><code>AbstractVariableRef</code></a>s to create different types of JuMP variables.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Extending JuMP in this manner is an advanced operation. We strongly encourage you to consider how you can use the methods mentioned in the previous sections to achieve your aims instead of defining new model and variable types. Consult the <a href="https://gitter.im/JuliaOpt/jump-dev">developer chatroom</a> <em>before</em> starting work on this.</p></div></div><p>If you define new types, you will need to implement a considerable number of methods, and doing so will require a detailed understanding of the JuMP internals. Therefore, the list of methods to implement is currently undocumented.</p><p>The easiest way to extend JuMP by defining a new model type is to follow an existing example. A simple example to follow is the <a href="https://github.com/jump-dev/JuMP.jl/blob/master/test/JuMPExtension.jl">JuMPExtension module</a> in the JuMP test suite. The best example of an external JuMP extension that implements an <a href="../../api/JuMP/#JuMP.AbstractModel"><code>AbstractModel</code></a> is <a href="https://github.com/infiniteopt/InfiniteOpt.jl">InfiniteOpt.jl</a>.</p><h3 id="Testing-JuMP-extensions"><a class="docs-heading-anchor" href="#Testing-JuMP-extensions">Testing JuMP extensions</a><a id="Testing-JuMP-extensions-1"></a><a class="docs-heading-anchor-permalink" href="#Testing-JuMP-extensions" title="Permalink"></a></h3><p>The JuMP test suite contains a large number of tests for JuMP extensions. You can run these tests by copying the MIT-licensed <a href="https://github.com/jump-dev/JuMP.jl/blob/master/test/Kokako.jl"><code>Kokako.jl</code></a> file in the JuMP tests into your <code>/test</code> folder, and then adding this snippet to your <code>/test/runtests.jl</code> file:</p><pre><code class="language-julia hljs">using MyJuMPExtension
+import JuMP
+include(&quot;Kokako.jl&quot;)
+const MODULES_TO_TEST = Kokako.include_modules_to_test(JuMP)
+Kokako.run_tests(
+    MODULES_TO_TEST,
+    MyJuMPExtension.MyModel,
+    MyJuMPExtension.MyVariableRef;
+    test_prefix = &quot;test_extension_&quot;,
+)</code></pre><h2 id="Set-an-optimize!-hook"><a class="docs-heading-anchor" href="#Set-an-optimize!-hook">Set an <code>optimize!</code> hook</a><a id="Set-an-optimize!-hook-1"></a><a class="docs-heading-anchor-permalink" href="#Set-an-optimize!-hook" title="Permalink"></a></h2><p>Some extensions require modification to the problem after the user has finished constructing the problem, but before <code>optimize!</code> is called. For these situations, JuMP provides <a href="../../api/JuMP/#JuMP.set_optimize_hook"><code>set_optimize_hook</code></a>, which lets you intercept the <a href="../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a> call.</p><p>Here&#39;s a simple example of adding an optimize hook that extends <a href="../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a> to take a keyword argument <code>silent</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; using JuMP, HiGHS
+
+julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; @variable(model, x &gt;= 1.5, Int);
+
+julia&gt; @objective(model, Min, x);
+
+julia&gt; function silent_hook(model; silent::Bool)
+           if silent
+               set_silent(model)
+           else
+               unset_silent(model)
+           end
+           ## Make sure you set ignore_optimize_hook = true, or we&#39;ll
+           ## recursively enter the optimize hook!
+           return optimize!(model; ignore_optimize_hook = true)
+       end
+silent_hook (generic function with 1 method)
+
+julia&gt; set_optimize_hook(model, silent_hook)
+silent_hook (generic function with 1 method)
+
+julia&gt; optimize!(model; silent = true)
+
+julia&gt; optimize!(model; silent = false)
+Solution has               num          max          sum
+Col     infeasibilities      0            0            0
+Integer infeasibilities      0            0            0
+Row     infeasibilities      0            0            0
+Row     residuals            0            0            0
+Presolving model
+0 rows, 0 cols, 0 nonzeros
+0 rows, 0 cols, 0 nonzeros
+Presolve: Optimal
+
+Solving report
+  Status            Optimal
+  Primal bound      2
+  Dual bound        2
+  Gap               0% (tolerance: 0.01%)
+  Solution status   feasible
+                    2 (objective)
+                    0 (bound viol.)
+                    0 (int. viol.)
+                    0 (row viol.)
+  Timing            0.00 (total)
+                    0.00 (presolve)
+                    0.00 (postsolve)
+  Nodes             0
+  LP iterations     0 (total)
+                    0 (strong br.)
+                    0 (separation)
+                    0 (heuristics)</code></pre><h2 id="Creating-new-container-types"><a class="docs-heading-anchor" href="#Creating-new-container-types">Creating new container types</a><a id="Creating-new-container-types-1"></a><a class="docs-heading-anchor-permalink" href="#Creating-new-container-types" title="Permalink"></a></h2><p>JuMP macros (for example, <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>) accept a <code>container</code> keyword argument to force the type of container that is chosen. By default, JuMP supports <code>container = Array</code>, <code>container = DenseAxisArray</code>, <code>container = SparseAxisArray</code> and <code>container = Auto</code>. You can extend support to user-defined types by implementing <a href="../../api/JuMP.Containers/#JuMP.Containers.container"><code>Containers.container</code></a>.</p><p>For example, here is a container that reverses the order of the indices:</p><pre><code class="language-julia-repl hljs">julia&gt; struct Foo end
+
+julia&gt; function Containers.container(f::Function, indices, ::Type{Foo})
+           return reverse([f(i...) for i in indices])
+       end
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3], container = Foo)
+3-element Vector{VariableRef}:
+ x[3]
+ x[2]
+ x[1]
+
+julia&gt; x[1]
+x[3]
+
+julia&gt; @variable(model, y[1:3, 1:2], container = Foo)
+3×2 Matrix{VariableRef}:
+ y[3,2]  y[3,1]
+ y[2,2]  y[2,1]
+ y[1,2]  y[1,1]
+
+julia&gt; y[1, 1]
+y[3,2]
+
+julia&gt; @variable(model, z[i=1:3; isodd(i)], container = Foo)
+2-element Vector{VariableRef}:
+ z[3]
+ z[1]
+
+julia&gt; z[2]
+z[1]</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>If you are a general user, you should not need to create a new container type. Instead, consider following <a href="../../manual/variables/#User-defined-containers">User-defined containers</a> and create a new container using standard Julia syntax. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3])
+3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+
+julia&gt; y = reverse(x)
+3-element Vector{VariableRef}:
+ x[3]
+ x[2]
+ x[1]</code></pre></div></div><h2 id="Performance-tips-for-extensions"><a class="docs-heading-anchor" href="#Performance-tips-for-extensions">Performance tips for extensions</a><a id="Performance-tips-for-extensions-1"></a><a class="docs-heading-anchor-permalink" href="#Performance-tips-for-extensions" title="Permalink"></a></h2><p>The function-in-set design of MathOptInterface causes type stability issues in Julia if you try to iterate over all of the constraints in a model. The easiest way to fix this is to use a function barrier.</p><p>For example, instead of:</p><pre><code class="language-julia hljs">function all_names_slow(model)
+    names = Set{String}()
+    for ci in all_constraints(model)
+        push!(names, name(ci))
+    end
+    return names
+end</code></pre><p>use:</p><pre><code class="language-julia hljs">function _function_barrier(names, model, ::Type{F}, ::Type{S}) where {F,S}
+    for ci in all_constraints(model, F, S)
+        push!(names, name(ci))
+    end
+    return
+end
+
+function all_names_fast(model)
+    names = Set{String}()
+    for (F, S) in list_of_constraint_types(model)
+        _function_barrier(names, model, F, S)
+    end
+    return names
+end</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>It is important to explicitly type the <code>F</code> and <code>S</code> arguments. If you leave them untyped, for example, <code>function _function_barrier(names, model, F, S)</code>, Julia will not specialize the function calls and performance will not be improved.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../contributing/">« Contributing</a><a class="docs-footer-nextpage" href="../custom_solver_binaries/">Custom binaries »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/developers/roadmap/index.html b/previews/PR3525/developers/roadmap/index.html
new file mode 100644
index 00000000000..b55dad6f212
--- /dev/null
+++ b/previews/PR3525/developers/roadmap/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Roadmap · JuMP</title><meta name="title" content="Roadmap · JuMP"/><meta property="og:title" content="Roadmap · JuMP"/><meta property="twitter:title" content="Roadmap · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox" checked/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../contributing/">Contributing</a></li><li><a class="tocitem" href="../extensions/">Extensions</a></li><li><a class="tocitem" href="../custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../style/">Style Guide</a></li><li class="is-active"><a class="tocitem" href>Roadmap</a></li><li><a class="tocitem" href="../checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Developer Docs</a></li><li class="is-active"><a href>Roadmap</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Roadmap</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/developers/roadmap.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Development-roadmap"><a class="docs-heading-anchor" href="#Development-roadmap">Development roadmap</a><a id="Development-roadmap-1"></a><a class="docs-heading-anchor-permalink" href="#Development-roadmap" title="Permalink"></a></h1><p>The JuMP developers have compiled this roadmap document to share their plans and goals with the JuMP community. Contributions to roadmap issues are especially invited.</p><p>Most of these issues will require changes to both JuMP and MathOptInterface, and are non-trivial in their implementation. They are in no particular order, but represent broad themes that we see as areas in which JuMP could be improved.</p><ul><li><strong>Done <a href="https://github.com/jump-dev/JuMP.jl/pull/3106">#3106</a></strong> Make nonlinear programming a first-class citizen. There have been many issues and discussions about this: currently nonlinear constraints are handled through a <code>MOI.NLPBlock</code> and have various limitations and restrictions.<ul><li><a href="https://github.com/jump-dev/JuMP.jl/issues/1185">https://github.com/jump-dev/JuMP.jl/issues/1185</a></li><li><a href="https://github.com/jump-dev/JuMP.jl/issues/1198">https://github.com/jump-dev/JuMP.jl/issues/1198</a></li><li><a href="https://github.com/jump-dev/JuMP.jl/issues/2788">https://github.com/jump-dev/JuMP.jl/issues/2788</a></li><li><a href="https://github.com/jump-dev/MathOptInterface.jl/issues/846">https://github.com/jump-dev/MathOptInterface.jl/issues/846</a></li><li><a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1397">https://github.com/jump-dev/MathOptInterface.jl/issues/1397</a></li></ul></li><li><strong>Done <a href="https://github.com/jump-dev/JuMP.jl/pull/3385">#3385</a></strong> Add support for coefficient types other than <code>Float64</code>: <a href="https://github.com/jump-dev/JuMP.jl/issues/2025">https://github.com/jump-dev/JuMP.jl/issues/2025</a> Since the very beginning, JuMP has hard-coded the coefficient type as <code>Float64</code>. This has made it impossible to support solvers which can use other types such as <code>BigFloat</code> or <code>Rational{BigInt}</code>.</li><li>Add support for constraint programming: <a href="https://github.com/jump-dev/JuMP.jl/issues/2227">https://github.com/jump-dev/JuMP.jl/issues/2227</a> JuMP has a strong focus on linear, conic and nonlinear optimization problems. We want to add better support for constraint programming.</li><li><strong>Done <a href="https://github.com/jump-dev/JuMP.jl/pull/3176">#3176</a></strong> Add support for multiobjective problems: <a href="https://github.com/jump-dev/JuMP.jl/issues/2099">https://github.com/jump-dev/JuMP.jl/issues/2099</a> JuMP is restricted to problems with scalar-valued objectives. We want to extend this to vector-valued problems.</li><li>Refactor the internal code of JuMP&#39;s macros. The code in <code>src/macros.jl</code> is some of the oldest part of JuMP and is difficult to read, modify, and extend. We should overhaul the internals of JuMP&#39;s macros–-without making user-visible breaking changes–-to improve their long-term maintainability.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../style/">« Style Guide</a><a class="docs-footer-nextpage" href="../checklists/">Checklists »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/developers/style/index.html b/previews/PR3525/developers/style/index.html
new file mode 100644
index 00000000000..bf3cfde7f25
--- /dev/null
+++ b/previews/PR3525/developers/style/index.html
@@ -0,0 +1,185 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Style Guide · JuMP</title><meta name="title" content="Style Guide · JuMP"/><meta property="og:title" content="Style Guide · JuMP"/><meta property="twitter:title" content="Style Guide · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox" checked/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../contributing/">Contributing</a></li><li><a class="tocitem" href="../extensions/">Extensions</a></li><li><a class="tocitem" href="../custom_solver_binaries/">Custom binaries</a></li><li class="is-active"><a class="tocitem" href>Style Guide</a><ul class="internal"><li><a class="tocitem" href="#Style-guide"><span>Style guide</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Testing"><span>Testing</span></a></li></ul></li><li><a class="tocitem" href="../roadmap/">Roadmap</a></li><li><a class="tocitem" href="../checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Developer Docs</a></li><li class="is-active"><a href>Style Guide</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Style Guide</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/developers/style.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Style-guide-and-design-principles"><a class="docs-heading-anchor" href="#Style-guide-and-design-principles">Style guide and design principles</a><a id="Style-guide-and-design-principles-1"></a><a class="docs-heading-anchor-permalink" href="#Style-guide-and-design-principles" title="Permalink"></a></h1><h2 id="Style-guide"><a class="docs-heading-anchor" href="#Style-guide">Style guide</a><a id="Style-guide-1"></a><a class="docs-heading-anchor-permalink" href="#Style-guide" title="Permalink"></a></h2><p>This section describes the coding style rules that apply to JuMP code and that we recommend for JuMP models and surrounding Julia code. The motivations for a style guide include:</p><ul><li>conveying best practices for writing readable and maintainable code</li><li>reducing the amount of time spent on <a href="https://en.wikipedia.org/wiki/Law_of_triviality">bike-shedding</a> by establishing basic naming and formatting conventions</li><li>lowering the barrier for new contributors by codifying the existing practices (for example, you can be more confident your code will pass review if you follow the style guide)</li></ul><p>In some cases, the JuMP style guide diverges from the <a href="https://docs.julialang.org/en/v1.0.0/manual/style-guide/">Julia style guide</a>. All such cases will be explicitly noted and justified.</p><p>The JuMP style guide adopts many recommendations from the <a href="https://github.com/google/styleguide">Google style guides</a>.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The style guide is always a work in progress, and not all JuMP code follows the rules. When modifying JuMP, please fix the style violations of the surrounding code (that is, leave the code tidier than when you started). If large changes are needed, consider separating them into another PR.</p></div></div><h3 id="JuliaFormatter"><a class="docs-heading-anchor" href="#JuliaFormatter">JuliaFormatter</a><a id="JuliaFormatter-1"></a><a class="docs-heading-anchor-permalink" href="#JuliaFormatter" title="Permalink"></a></h3><p>JuMP uses <a href="https://github.com/domluna/JuliaFormatter.jl">JuliaFormatter.jl</a> as an auto-formatting tool.</p><p>We use the options contained in <a href="https://github.com/jump-dev/JuMP.jl/blob/master/.JuliaFormatter.toml"><code>.JuliaFormatter.toml</code></a>.</p><p>To format code, <code>cd</code> to the JuMP directory, then run:</p><pre><code class="language-julia hljs">] add JuliaFormatter@1
+using JuliaFormatter
+format(&quot;docs&quot;)
+format(&quot;src&quot;)
+format(&quot;test&quot;)</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>A continuous integration check verifies that all PRs made to JuMP have passed the formatter.</p></div></div><p>The following sections outline extra style guide points that are not fixed automatically by JuliaFormatter.</p><h3 id="Abstract-types-and-composition"><a class="docs-heading-anchor" href="#Abstract-types-and-composition">Abstract types and composition</a><a id="Abstract-types-and-composition-1"></a><a class="docs-heading-anchor-permalink" href="#Abstract-types-and-composition" title="Permalink"></a></h3><p>Specifying types for method arguments is mostly optional in Julia. The benefit of abstract method arguments is that it enables functions and types from one package to be used with functions and types from another package via multiple dispatch.</p><p>However, abstractly typed methods have two main drawbacks:</p><ol><li>It&#39;s possible to find out that you are working with unexpected types deep in the call chain, potentially leading to hard-to-diagnose <a href="https://docs.julialang.org/en/v1/manual/methods/#Defining-Methods"><code>MethodError</code>s</a>.</li><li>Untyped function arguments can lead to correctness problems if the user&#39;s choice of input type does not satisfy the assumptions made by the author of the function.</li></ol><p>As a motivating example, consider the following function:</p><pre><code class="language-julia-repl hljs">julia&gt; function my_sum(x)
+           y = 0.0
+           for i in 1:length(x)
+               y += x[i]
+           end
+           return y
+       end
+my_sum (generic function with 1 method)</code></pre><p>This function contains a number of implicit assumptions about the type of <code>x</code>:</p><ul><li><code>x</code> supports 1-based <code>getindex</code> and implements <code>length</code></li><li>The element type of <code>x</code> supports addition with <code>0.0</code>, and then with the result of <code>x + 0.0</code>.</li></ul><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>As a motivating example for the second point, <a href="../../api/JuMP/#JuMP.VariableRef"><code>VariableRef</code></a> plus <code>Float64</code> produces an <a href="../../api/JuMP/#JuMP.AffExpr"><code>AffExpr</code></a>. Do not assume that <code>+(::A, ::B)</code> produces an instance of the type <code>A</code> or <code>B</code>.</p></div></div><p><code>my_sum</code> works as expected if the user passes in <code>Vector{Float64}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; my_sum([1.0, 2.0, 3.0])
+6.0</code></pre><p>but it doesn&#39;t respect input types, for example returning a <code>Float64</code> if the user passes <code>Vector{Int}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; my_sum([1, 2, 3])
+6.0</code></pre><p>but it throws a <code>MethodError</code> if the user passes <code>String</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; my_sum(&quot;abc&quot;)
+ERROR: MethodError: no method matching +(::Float64, ::Char)
+[...]</code></pre><p>This particular <code>MethodError</code> is hard to debug, particularly for new users, because it mentions <code>+</code>, <code>Float64</code>, and <code>Char</code>, none of which were called or passed by the user.</p><h4 id="Dealing-with-MethodErrors"><a class="docs-heading-anchor" href="#Dealing-with-MethodErrors">Dealing with <code>MethodError</code>s</a><a id="Dealing-with-MethodErrors-1"></a><a class="docs-heading-anchor-permalink" href="#Dealing-with-MethodErrors" title="Permalink"></a></h4><p>This section diverges from the <a href="https://docs.julialang.org/en/v1.6/manual/style-guide/#Avoid-writing-overly-specific-types">Julia style guide</a>, as well as other common guides like <a href="https://github.com/SciML/SciMLStyle#generic-code-is-preferred-unless-code-is-known-to-be-specific">SciML</a>. The following suggestions are intended to provide a friendlier experience for novice Julia programmers, at the cost of limiting the power and flexibility of advanced Julia programmers.</p><p>Code should follow the <code>MethodError</code> principle:</p><div class="admonition is-info"><header class="admonition-header">The MethodError principle</header><div class="admonition-body"><p>A user should see a <code>MethodError</code> only for methods that they called directly.</p></div></div><p>Bad:</p><pre><code class="language-julia hljs">_internal_function(x::Integer) = x + 1
+# The user sees a MethodError for _internal_function when calling
+# public_function(&quot;a string&quot;). This is not very helpful.
+public_function(x) = _internal_function(x)</code></pre><p>Good:</p><pre><code class="language-julia hljs">_internal_function(x::Integer) = x + 1
+# The user sees a MethodError for public_function when calling
+# public_function(&quot;a string&quot;). This is easy to understand.
+public_function(x::Integer) = _internal_function(x)</code></pre><p>If it is hard to provide an error message at the top of the call chain, then the following pattern is also ok:</p><pre><code class="language-julia hljs">_internal_function(x::Integer) = x + 1
+function _internal_function(x)
+    error(
+        &quot;Internal error. This probably means that you called &quot; *
+        &quot;`public_function()`s with the wrong type.&quot;,
+    )
+end
+public_function(x) = _internal_function(x)</code></pre><h4 id="Dealing-with-correctness"><a class="docs-heading-anchor" href="#Dealing-with-correctness">Dealing with correctness</a><a id="Dealing-with-correctness-1"></a><a class="docs-heading-anchor-permalink" href="#Dealing-with-correctness" title="Permalink"></a></h4><p>Dealing with correctness is harder, because Julia has no way of formally specifying interfaces that abstract types must implement. Instead, here are two options that you can use when writing and interacting with generic code:</p><p><strong>Option 1: use concrete types and let users extend new methods.</strong></p><p>In this option, explicitly restrict input arguments to concrete types that are tested and have been validated for correctness. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; function my_sum_option_1(x::Vector{Float64})
+           y = 0.0
+           for i in 1:length(x)
+               y += x[i]
+           end
+           return y
+       end
+my_sum_option_1 (generic function with 1 method)
+
+julia&gt; my_sum_option_1([1.0, 2.0, 3.0])
+6.0</code></pre><p>Using concrete types satisfies the <code>MethodError</code> principle:</p><pre><code class="language-julia-repl hljs">julia&gt; my_sum_option_1(&quot;abc&quot;)
+ERROR: MethodError: no method matching my_sum_option_1(::String)</code></pre><p>and it allows other types to be supported in future by defining new methods:</p><pre><code class="language-julia-repl hljs">julia&gt; function my_sum_option_1(x::Array{T,N}) where {T&lt;:Number,N}
+           y = zero(T)
+           for i in eachindex(x)
+               y += x[i]
+           end
+           return y
+       end
+my_sum_option_1 (generic function with 2 methods)</code></pre><p>Importantly, these methods do not have to be defined in the original package.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Some usage of abstract types is okay. For example, in <code>my_sum_option_1</code>, we allowed the element type, <code>T</code>, to be a subtype of <code>Number</code>. This is fairly safe, but it still has an implicit assumption that <code>T</code> supports <code>zero(T)</code> and <code>+(::T, ::T)</code>.</p></div></div><p><strong>Option 2: program defensively, and validate all assumptions.</strong></p><p>An alternative is to program defensively, and to rigorously document and validate all assumptions that the code makes. In particular:</p><ol><li>All assumptions on abstract types that aren&#39;t guaranteed by the definition of the abstract type (for example, optional methods without a fallback) should be documented.</li><li>If practical, the assumptions should be checked in code, and informative error messages should be provided to the user if the assumptions are not met. In general, these checks may be expensive, so you should prefer to do this once, at the highest level of the call-chain.</li><li>Tests should cover for a range of corner cases and argument types.</li></ol><p>For example:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    test_my_sum_defensive_assumptions(x::AbstractArray{T}) where {T}
+
+Test the assumptions made by `my_sum_defensive`.
+&quot;&quot;&quot;
+function test_my_sum_defensive_assumptions(x::AbstractArray{T}) where {T}
+    try
+        # Some types may not define zero.
+        @assert zero(T) isa T
+        # Check iteration supported
+        @assert iterate(x) isa Union{Nothing,Tuple{T,Int}}
+        # Check that + is defined
+        @assert +(zero(T), zero(T)) isa Any
+    catch err
+        error(
+            &quot;Unable to call my_sum_defensive(::$(typeof(x))) because &quot; *
+            &quot;it failed an internal assumption&quot;,
+        )
+    end
+    return
+end
+
+&quot;&quot;&quot;
+    my_sum_defensive(x::AbstractArray{T}) where {T}
+
+Return the sum of the elements in the abstract array `x`.
+
+## Assumptions
+
+This function makes the following assumptions:
+
+ * That `zero(T)` is defined
+ * That `x` supports the iteration interface
+ * That  `+(::T, ::T)` is defined
+&quot;&quot;&quot;
+function my_sum_defensive(x::AbstractArray{T}) where {T}
+    test_my_sum_defensive_assumptions(x)
+    y = zero(T)
+    for xi in x
+        y += xi
+    end
+    return y
+end
+
+# output
+
+my_sum_defensive</code></pre><p>This function works on <code>Vector{Float64}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; my_sum_defensive([1.0, 2.0, 3.0])
+6.0</code></pre><p>as well as <code>Matrix{Rational{Int}}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; my_sum_defensive([(1//2) + (4//3)im; (6//5) + (7//11)im])
+17//10 + 65//33*im</code></pre><p>and it throws an error when the assumptions aren&#39;t met:</p><pre><code class="language-julia-repl hljs">julia&gt; my_sum_defensive([&#39;a&#39;, &#39;b&#39;, &#39;c&#39;])
+ERROR: Unable to call my_sum_defensive(::Vector{Char}) because it failed an internal assumption
+[...]</code></pre><p>As an alternative, you may choose not to call <code>test_my_sum_defensive_assumptions</code> within <code>my_sum_defensive</code>, and instead ask users of <code>my_sum_defensive</code> to call it in their tests.</p><h3 id="Juxtaposed-multiplication"><a class="docs-heading-anchor" href="#Juxtaposed-multiplication">Juxtaposed multiplication</a><a id="Juxtaposed-multiplication-1"></a><a class="docs-heading-anchor-permalink" href="#Juxtaposed-multiplication" title="Permalink"></a></h3><p>Only use juxtaposed multiplication when the right-hand side is a symbol.</p><p>Good:</p><pre><code class="language-julia hljs">2x  # Acceptable if there are space constraints.
+2 * x  # This is preferred if space is not an issue.
+2 * (x + 1)</code></pre><p>Bad:</p><pre><code class="language-julia hljs">2(x + 1)</code></pre><h3 id="Empty-vectors"><a class="docs-heading-anchor" href="#Empty-vectors">Empty vectors</a><a id="Empty-vectors-1"></a><a class="docs-heading-anchor-permalink" href="#Empty-vectors" title="Permalink"></a></h3><p>For a type <code>T</code>, <code>T[]</code> and <code>Vector{T}()</code> are equivalent ways to create an empty vector with element type <code>T</code>. Prefer <code>T[]</code> because it is more concise.</p><h3 id="Comments"><a class="docs-heading-anchor" href="#Comments">Comments</a><a id="Comments-1"></a><a class="docs-heading-anchor-permalink" href="#Comments" title="Permalink"></a></h3><p>For non-native speakers and for general clarity, comments in code must be proper English sentences with appropriate punctuation.</p><p>Good:</p><pre><code class="language-julia hljs"># This is a comment demonstrating a good comment.</code></pre><p>Bad:</p><pre><code class="language-julia hljs"># a bad comment</code></pre><h3 id="JuMP-macro-syntax"><a class="docs-heading-anchor" href="#JuMP-macro-syntax">JuMP macro syntax</a><a id="JuMP-macro-syntax-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-macro-syntax" title="Permalink"></a></h3><p>For consistency, always use parentheses.</p><p>Good:</p><pre><code class="language-julia hljs">@variable(model, x &gt;= 0)</code></pre><p>Bad:</p><pre><code class="language-julia hljs">@variable model x &gt;= 0</code></pre><p>For consistency, always use <code>constant * variable</code> as opposed to <code>variable * constant</code>. This makes it easier to read models in ambiguous cases like <code>a * x</code>.</p><p>Good:</p><pre><code class="language-julia hljs">a = 4
+@constraint(model, 3 * x &lt;= 1)
+@constraint(model, a * x &lt;= 1)</code></pre><p>Bad:</p><pre><code class="language-julia hljs">a = 4
+@constraint(model, x * 3 &lt;= 1)
+@constraint(model, x * a &lt;= 1)</code></pre><p>In order to reduce boilerplate code, prefer the plural form of macros over lots of repeated calls to singular forms.</p><p>Good:</p><pre><code class="language-julia hljs">@variables(model, begin
+    x &gt;= 0
+    y &gt;= 1
+    z &lt;= 2
+end)</code></pre><p>Bad:</p><pre><code class="language-julia hljs">@variable(model, x &gt;= 0)
+@variable(model, y &gt;= 1)
+@variable(model, z &lt;= 2)</code></pre><p>An exception is made for calls with many keyword arguments, since these need to be enclosed in parentheses in order to parse properly.</p><p>Acceptable:</p><pre><code class="language-julia hljs">@variable(model, x &gt;= 0, start = 0.0, base_name = &quot;my_x&quot;)
+@variable(model, y &gt;= 1, start = 2.0)
+@variable(model, z &lt;= 2, start = -1.0)</code></pre><p>Also acceptable:</p><pre><code class="language-julia hljs">@variables(model, begin
+    x &gt;= 0, (start = 0.0, base_name = &quot;my_x&quot;)
+    y &gt;= 1, (start = 2.0)
+    z &lt;= 2, (start = -1.0)
+end)</code></pre><p>While we always use <code>in</code> for <code>for</code>-loops, it is acceptable to use <code>=</code> in the container declarations of JuMP macros.</p><p>Okay:</p><pre><code class="language-julia hljs">@variable(model, x[i=1:3])</code></pre><p>Also okay:</p><pre><code class="language-julia hljs">@variable(model, x[i in 1:3])</code></pre><h3 id="Naming"><a class="docs-heading-anchor" href="#Naming">Naming</a><a id="Naming-1"></a><a class="docs-heading-anchor-permalink" href="#Naming" title="Permalink"></a></h3><pre><code class="language-julia hljs">module SomeModule end
+function some_function end
+const SOME_CONSTANT = ...
+struct SomeStruct
+  some_field::SomeType
+end
+@enum SomeEnum ENUM_VALUE_A ENUM_VALUE_B
+some_local_variable = ...
+some_file.jl # Except for ModuleName.jl.</code></pre><h3 id="Exported-and-non-exported-names"><a class="docs-heading-anchor" href="#Exported-and-non-exported-names">Exported and non-exported names</a><a id="Exported-and-non-exported-names-1"></a><a class="docs-heading-anchor-permalink" href="#Exported-and-non-exported-names" title="Permalink"></a></h3><p>Begin private module level functions and constants with an underscore. All other objects in the scope of a module should be exported. (See JuMP.jl for an example of how to do this.)</p><p>Names beginning with an underscore should only be used for distinguishing between exported (public) and non-exported (private) objects. Therefore, never begin the name of a local variable with an underscore.</p><pre><code class="language-julia hljs">module MyModule
+
+export public_function, PUBLIC_CONSTANT
+
+function _private_function()
+    local_variable = 1
+    return
+end
+
+function public_function end
+
+const _PRIVATE_CONSTANT = 3.14159
+const PUBLIC_CONSTANT = 1.41421
+
+end</code></pre><h3 id="Use-of-underscores-within-names"><a class="docs-heading-anchor" href="#Use-of-underscores-within-names">Use of underscores within names</a><a id="Use-of-underscores-within-names-1"></a><a class="docs-heading-anchor-permalink" href="#Use-of-underscores-within-names" title="Permalink"></a></h3><p>The Julia style guide recommends avoiding underscores &quot;when readable,&quot; for example, <code>haskey</code>, <code>isequal</code>, <code>remotecall</code>, and <code>remotecall_fetch</code>. This convention creates the potential for unnecessary bikeshedding and also forces the user to recall the presence/absence of an underscore, for example, &quot;was that argument named <code>basename</code> or <code>base_name</code>?&quot;. For consistency, <em>always use underscores</em> in variable names and function names to separate words.</p><h3 id="Use-of-!"><a class="docs-heading-anchor" href="#Use-of-!">Use of <code>!</code></a><a id="Use-of-!-1"></a><a class="docs-heading-anchor-permalink" href="#Use-of-!" title="Permalink"></a></h3><p>Julia has a convention of appending <code>!</code> to a function name if the function modifies its arguments. We recommend to:</p><ul><li>Omit <code>!</code> when the name itself makes it clear that modification is taking place, for example, <code>add_constraint</code> and <code>set_name</code>. We depart from the Julia style guide because <code>!</code> does not provide a reader with any additional information in this case, and adherence to this convention is not uniform even in base Julia itself (consider <code>Base.println</code> and <code>Base.finalize</code>).</li><li>Use <code>!</code> in all other cases. In particular it can be used to distinguish between modifying and non-modifying variants of the same function like <code>scale</code> and <code>scale!</code>.</li></ul><p>Note that <code>!</code> is <em>not</em> a self-documenting feature because it is still ambiguous which arguments are modified when multiple arguments are present. Be sure to document which arguments are modified in the method&#39;s docstring.</p><p>See also the Julia style guide recommendations for <a href="https://docs.julialang.org/en/v1/manual/style-guide/#Write-functions-with-argument-ordering-similar-to-Julia-Base">ordering of function arguments</a>.</p><h3 id="Abbreviations"><a class="docs-heading-anchor" href="#Abbreviations">Abbreviations</a><a id="Abbreviations-1"></a><a class="docs-heading-anchor-permalink" href="#Abbreviations" title="Permalink"></a></h3><p>Abbreviate names to make the code more readable, not to save typing. Don&#39;t arbitrarily delete letters from a word to abbreviate it (for example, <code>indx</code>). Use abbreviations consistently within a body of code (for example, do not mix <code>con</code> and <code>constr</code>, <code>idx</code> and <code>indx</code>).</p><p>Common abbreviations:</p><ul><li><code>num</code> for <code>number</code></li><li><code>con</code> for <code>constraint</code></li></ul><h3 id="No-one-letter-variable-names"><a class="docs-heading-anchor" href="#No-one-letter-variable-names">No one-letter variable names</a><a id="No-one-letter-variable-names-1"></a><a class="docs-heading-anchor-permalink" href="#No-one-letter-variable-names" title="Permalink"></a></h3><p>Where possible, avoid one-letter variable names.</p><p>Use <code>model = Model()</code> instead of <code>m = Model()</code></p><p>Exceptions are made for indices in loops.</p><h3 id="@enum-vs.-Symbol"><a class="docs-heading-anchor" href="#@enum-vs.-Symbol"><code>@enum</code> vs. <code>Symbol</code></a><a id="@enum-vs.-Symbol-1"></a><a class="docs-heading-anchor-permalink" href="#@enum-vs.-Symbol" title="Permalink"></a></h3><p>The <code>@enum</code> macro lets you define types with a finite number of values that are explicitly enumerated (like <code>enum</code> in C/C++). <code>Symbol</code>s are lightweight strings that are used to represent identifiers in Julia (for example, <code>:x</code>).</p><p><code>@enum</code> provides type safety and can have docstrings attached to explain the possible values. Use <code>@enum</code>s when applicable, for example, for reporting statuses. Use strings to provide long-form additional information like error messages.</p><p>Use of <code>Symbol</code> should typically be reserved for identifiers, for example, for lookup in the JuMP model (<code>model[:my_variable]</code>).</p><h3 id="using-vs.-import"><a class="docs-heading-anchor" href="#using-vs.-import"><code>using</code> vs. <code>import</code></a><a id="using-vs.-import-1"></a><a class="docs-heading-anchor-permalink" href="#using-vs.-import" title="Permalink"></a></h3><p><code>using ModuleName</code> brings all symbols exported by the module <code>ModuleName</code> into scope, while <code>import ModuleName</code> brings only the module itself into scope. (See the Julia <a href="https://docs.julialang.org/en/v1/manual/modules/#modules-1">manual</a>) for examples and more details.</p><p>For the same reason that <code>from &lt;module&gt; import *</code> is not recommended in python (<a href="https://www.python.org/dev/peps/pep-0008/#imports">PEP 8</a>), avoid <code>using ModuleName</code> except in throw-away scripts or at the REPL. The <code>using</code> statement makes it harder to track where symbols come from and exposes the code to ambiguities when two modules export the same symbol.</p><p>Prefer <code>using ModuleName: x, p</code> to <code>import ModuleName.x, ModuleName.p</code> and <code>import MyModule: x, p</code> because the <code>import</code> versions allow method extension without qualifying with the module name.</p><p>Similarly, <code>using ModuleName: ModuleName</code> is an acceptable substitute for <code>import ModuleName</code>, because it does not bring all symbols exported by <code>ModuleName</code> into scope. However, we prefer <code>import ModuleName</code> for consistency.</p><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>This section describes the writing style that should be used when writing documentation for JuMP (and supporting packages).</p><p>We can recommend the documentation style guides by <a href="https://www.divio.com/blog/documentation/">Divio</a>, <a href="https://developers.google.com/style/">Google</a>, and <a href="https://www.writethedocs.org/guide/">Write the Docs</a> as general reading for those writing documentation. This guide delegates a thorough handling of the topic to those guides and instead elaborates on the points more specific to Julia and documentation that use <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter</a>.</p><ul><li>Be concise</li><li>Use lists instead of long sentences</li><li>Use numbered lists when describing a sequence, for example, (1) do X, (2) then Y</li><li>Use bullet points when the items are not ordered</li><li>Example code should be covered by doctests</li><li>When a word is a Julia symbol and not an English word, enclose it with backticks. In addition, if it has a docstring in this doc add a link using <code>@ref</code>. If it is a plural, add the &quot;s&quot; after the closing backtick. For example,<pre><code class="nohighlight hljs">[`VariableRef`](@ref)s</code></pre></li><li>Use <a href="https://juliadocs.github.io/Documenter.jl/v0.21/man/syntax/#@meta-block-1"><code>@meta</code></a> blocks for TODOs and other comments that shouldn&#39;t be visible to readers. For example,<pre><code class="language- hljs">```@meta
+# TODO: Mention also X, Y, and Z.
+```</code></pre></li></ul><h3 id="Docstrings"><a class="docs-heading-anchor" href="#Docstrings">Docstrings</a><a id="Docstrings-1"></a><a class="docs-heading-anchor-permalink" href="#Docstrings" title="Permalink"></a></h3><ul><li>Every exported object needs a docstring</li><li>All examples in docstrings should be <a href="https://juliadocs.github.io/Documenter.jl/stable/man/doctests/"><code>jldoctests</code></a></li><li>Always use complete English sentences with proper punctuation</li><li>Do not terminate lists with punctuation (for example, as in this doc)</li></ul><p>Here is an example:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    signature(args; kwargs...)
+
+Short sentence describing the function.
+
+Optional: add a slightly longer paragraph describing the function.
+
+## Notes
+
+ - List any notes that the user should be aware of
+
+## Examples
+
+```jldoctest
+julia&gt; 1 + 1
+2
+```
+&quot;&quot;&quot;</code></pre><h2 id="Testing"><a class="docs-heading-anchor" href="#Testing">Testing</a><a id="Testing-1"></a><a class="docs-heading-anchor-permalink" href="#Testing" title="Permalink"></a></h2><p>Use a module to encapsulate tests, and structure all tests as functions. This avoids leaking local variables between tests.</p><p>Here is a basic skeleton:</p><pre><code class="language-julia hljs">module TestPkg
+
+using Test
+
+function runtests()
+    for name in names(@__MODULE__; all = true)
+        if startswith(&quot;$(name)&quot;, &quot;test_&quot;)
+            @testset &quot;$(name)&quot; begin
+                getfield(@__MODULE__, name)()
+            end
+        end
+    end
+    return
+end
+
+_helper_function() = 2
+
+function test_addition()
+    @test 1 + 1 == _helper_function()
+    return
+end
+
+end # module TestPkg
+
+TestPkg.runtests()</code></pre><p>Break the tests into multiple files, with one module per file, so that subsets of the codebase can be tested by calling <code>include</code> with the relevant file.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../custom_solver_binaries/">« Custom binaries</a><a class="docs-footer-nextpage" href="../roadmap/">Roadmap »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/extensions/DimensionalData/index.html b/previews/PR3525/extensions/DimensionalData/index.html
new file mode 100644
index 00000000000..66707916a24
--- /dev/null
+++ b/previews/PR3525/extensions/DimensionalData/index.html
@@ -0,0 +1,42 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>rafaqz/DimensionalData.jl · JuMP</title><meta name="title" content="rafaqz/DimensionalData.jl · JuMP"/><meta property="og:title" content="rafaqz/DimensionalData.jl · JuMP"/><meta property="twitter:title" content="rafaqz/DimensionalData.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li class="is-active"><a class="tocitem" href>rafaqz/DimensionalData.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>rafaqz/DimensionalData.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>rafaqz/DimensionalData.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/extensions/DimensionalData.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="DimensionalData.jl"><a class="docs-heading-anchor" href="#DimensionalData.jl">DimensionalData.jl</a><a id="DimensionalData.jl-1"></a><a class="docs-heading-anchor-permalink" href="#DimensionalData.jl" title="Permalink"></a></h1><p><a href="https://github.com/rafaqz/DimensionalData.jl">DimensionalData.jl</a> provides tools and abstractions for working with rectangular arrays that have named dimensions.</p><div class="admonition is-compat"><header class="admonition-header">Compat</header><div class="admonition-body"><p>Using the DimensionalData extension with JuMP requires Julia v1.9 or later.</p></div></div><p>The DimensionalData extension in JuMP lets you construct a <code>DimensionalData.DimArray</code> as an alternative to <a href="../../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> in the JuMP macros.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>DimensionalData.jl is licensed under the <a href="https://github.com/rafaqz/DimensionalData.jl/blob/main/LICENSE">MIT license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install DimensionalData using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;DimensionalData&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Activate the extension by loading both JuMP and DimensionalData:</p><pre><code class="language-julia-repl hljs">julia&gt; using JuMP, DimensionalData</code></pre><p>Then, pass <code>container = DimensionalData.DimArray</code> in the <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>, <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>, or <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macros:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(
+           model,
+           x[i = 2:4, j = [&quot;a&quot;, &quot;b&quot;]] &gt;= i,
+           container = DimensionalData.DimArray,
+       )
+3×2 DimArray{VariableRef,2} with dimensions:
+  Dim{:i} Sampled{Int64} 2:4 ForwardOrdered Regular Points,
+  Dim{:j} Categorical{String} String[&quot;a&quot;, &quot;b&quot;] ForwardOrdered
+    &quot;a&quot;     &quot;b&quot;
+ 2  x[2,a]  x[2,b]
+ 3  x[3,a]  x[3,b]
+ 4  x[4,a]  x[4,b]</code></pre><p>Here <code>x</code> is a <code>DimensionalData.Dim</code> array object, so indexing uses the DimensionalData syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; x[At(2), At(&quot;a&quot;)]
+x[2,a]
+
+julia&gt; x[2, 2]
+x[3,b]</code></pre><p>You can use <code>container = DimensionalData.DimArray</code> in the <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macro:</p><pre><code class="language-julia-repl hljs">julia&gt; @expression(
+           model,
+           expr[j = [&quot;a&quot;, &quot;b&quot;]],
+           sum(x[At(i), At(j)] for i in 2:4),
+           container = DimensionalData.DimArray,
+       )
+2-element DimArray{AffExpr,1} with dimensions:
+  Dim{:j} Categorical{String} String[&quot;a&quot;, &quot;b&quot;] ForwardOrdered
+ &quot;a&quot;  x[2,a] + x[3,a] + x[4,a]
+ &quot;b&quot;  x[2,b] + x[3,b] + x[4,b]</code></pre><p>and in <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(
+           model,
+           [j = [&quot;a&quot;, &quot;b&quot;]],
+           expr[At(j)] &lt;= 1,
+           container = DimensionalData.DimArray,
+       )
+2-element DimArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},1} with dimensions:
+  Dim{:j} Categorical{String} String[&quot;a&quot;, &quot;b&quot;] ForwardOrdered
+ &quot;a&quot;  x[2,a] + x[3,a] + x[4,a] ≤ 1
+ &quot;b&quot;  x[2,b] + x[3,b] + x[4,b] ≤ 1</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>See the <a href="https://rafaqz.github.io/DimensionalData.jl/stable/">DimensionalData.jl documentation</a> for more details on the syntax and features of <code>DimensionalData.DimArray</code>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../packages/Plasmo/">« plasmo-dev/Plasmo.jl</a><a class="docs-footer-nextpage" href="../../moi/">Introduction »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/extensions/introduction/index.html b/previews/PR3525/extensions/introduction/index.html
new file mode 100644
index 00000000000..db358fe1f3d
--- /dev/null
+++ b/previews/PR3525/extensions/introduction/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#Adding-new-extensions"><span>Adding new extensions</span></a></li><li><a class="tocitem" href="#Weak-dependencies"><span>Weak dependencies</span></a></li></ul></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/extensions/introduction.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Introduction"><a class="docs-heading-anchor" href="#Introduction">Introduction</a><a id="Introduction-1"></a><a class="docs-heading-anchor-permalink" href="#Introduction" title="Permalink"></a></h1><p>This section of the documentation contains brief documentation for some popular JuMP extensions. The list of extensions is not exhaustive, but instead is intended to help you discover popular JuMP extensions, and to give you an overview of the types of extensions that are possible to write with JuMP.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>Packages beginning with <code>jump-dev/</code> are developed and maintained by the JuMP developers.</p><p>Packages that do not begin with <code>jump-dev/</code> are developed independently. The developers of these packages requested or consented to the inclusion of their README contents in the JuMP documentation for the benefit of users.</p><h2 id="Adding-new-extensions"><a class="docs-heading-anchor" href="#Adding-new-extensions">Adding new extensions</a><a id="Adding-new-extensions-1"></a><a class="docs-heading-anchor-permalink" href="#Adding-new-extensions" title="Permalink"></a></h2><p>Written an extension? Add it to this section of the JuMP documentation by making a pull request to the <a href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/packages.toml"><code>docs/packages.toml</code></a> file.</p><h2 id="Weak-dependencies"><a class="docs-heading-anchor" href="#Weak-dependencies">Weak dependencies</a><a id="Weak-dependencies-1"></a><a class="docs-heading-anchor-permalink" href="#Weak-dependencies" title="Permalink"></a></h2><p>Some extensions listed in this section are implemented using the <a href="https://pkgdocs.julialang.org/v1/creating-packages/#Weak-dependencies">weak dependency</a> feature added to Julia in v1.9. These extensions are activated if and only if you have <code>JuMP</code> and the other package loaded into your current scope with <code>using</code> or <code>import</code>.</p><div class="admonition is-compat"><header class="admonition-header">Compat</header><div class="admonition-body"><p>Using a weak dependency requires Julia v1.9 or later.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../packages/SCIP/">« scipopt/SCIP.jl</a><a class="docs-footer-nextpage" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/index.html b/previews/PR3525/index.html
new file mode 100644
index 00000000000..5d589443352
--- /dev/null
+++ b/previews/PR3525/index.html
@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="assets/documenter.js"></script><script src="search_index.js"></script><script src="siteinfo.js"></script><script src="../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="assets/themeswap.js"></script><link href="assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href><img class="docs-light-only" src="assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox" checked/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#What-is-JuMP?"><span>What is JuMP?</span></a></li><li><a class="tocitem" href="#Resources-for-getting-started"><span>Resources for getting started</span></a></li><li><a class="tocitem" href="#How-the-documentation-is-structured"><span>How the documentation is structured</span></a></li><li><a class="tocitem" href="#Citing-JuMP"><span>Citing JuMP</span></a></li><li><a class="tocitem" href="#NumFOCUS"><span>NumFOCUS</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li></ul></li><li><a class="tocitem" href="should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="manual/models/">Models</a></li><li><a class="tocitem" href="manual/variables/">Variables</a></li><li><a class="tocitem" href="manual/constraints/">Constraints</a></li><li><a class="tocitem" href="manual/expressions/">Expressions</a></li><li><a class="tocitem" href="manual/objective/">Objectives</a></li><li><a class="tocitem" href="manual/containers/">Containers</a></li><li><a class="tocitem" href="manual/solutions/">Solutions</a></li><li><a class="tocitem" href="manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="manual/complex/">Complex number support</a></li><li><a class="tocitem" href="manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="api/JuMP/">JuMP</a></li><li><a class="tocitem" href="api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="developers/contributing/">Contributing</a></li><li><a class="tocitem" href="developers/extensions/">Extensions</a></li><li><a class="tocitem" href="developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="developers/style/">Style Guide</a></li><li><a class="tocitem" href="developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="packages/solvers/">Introduction</a></li><li><a class="tocitem" href="packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/">Introduction</a></li><li><a class="tocitem" href="moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="moi/manual/models/">Models</a></li><li><a class="tocitem" href="moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/background/duality/">Duality</a></li><li><a class="tocitem" href="moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="moi/reference/models/">Models</a></li><li><a class="tocitem" href="moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Introduction</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/index.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><img class="display-light-only" src="assets/logo-with-text.svg" alt="JuMP logo"/>
+<img class="display-dark-only" src="assets/logo-dark-with-text.svg" alt="JuMP logo"/><h1 id="Introduction"><a class="docs-heading-anchor" href="#Introduction">Introduction</a><a id="Introduction-1"></a><a class="docs-heading-anchor-permalink" href="#Introduction" title="Permalink"></a></h1><p>Welcome to the documentation for JuMP.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>This documentation is also available in PDF format: <a href="JuMP.pdf">JuMP.pdf</a>.</p></div></div><h2 id="What-is-JuMP?"><a class="docs-heading-anchor" href="#What-is-JuMP?">What is JuMP?</a><a id="What-is-JuMP?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-JuMP?" title="Permalink"></a></h2><p><a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> is a domain-specific modeling language for <a href="https://en.wikipedia.org/wiki/Mathematical_optimization">mathematical optimization</a> embedded in <a href="https://julialang.org/">Julia</a>. It currently supports a number of open-source and commercial solvers for a variety of problem classes, including linear, mixed-integer, second-order conic, semidefinite, and nonlinear programming.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you aren&#39;t sure if you should use JuMP, read <a href="should_i_use/#Should-you-use-JuMP?">Should you use JuMP?</a>.</p></div></div><h2 id="Resources-for-getting-started"><a class="docs-heading-anchor" href="#Resources-for-getting-started">Resources for getting started</a><a id="Resources-for-getting-started-1"></a><a class="docs-heading-anchor-permalink" href="#Resources-for-getting-started" title="Permalink"></a></h2><p>There are few ways to get started with JuMP:</p><ul><li>Read the <a href="installation/#Installation-Guide">Installation Guide</a>.</li><li>Read the introductory tutorials <a href="tutorials/getting_started/getting_started_with_julia/#Getting-started-with-Julia">Getting started with Julia</a> and <a href="tutorials/getting_started/getting_started_with_JuMP/#Getting-started-with-JuMP">Getting started with JuMP</a>.</li><li>Browse some of our modeling tutorials, including classics such as <a href="tutorials/linear/diet/#The-diet-problem">The diet problem</a>, or the <a href="tutorials/nonlinear/simple_examples/#Maximum-likelihood-estimation">Maximum likelihood estimation</a> problem using nonlinear programming.</li></ul><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Need help? Join the <a href="https://jump.dev/forum">community forum</a> to search for answers to commonly asked questions.</p><p>Before asking a question, make sure to read the post <a href="https://discourse.julialang.org/t/psa-make-it-easier-to-help-you/14757">make it easier to help you</a>, which contains a number of tips on how to ask a good question.</p></div></div><h2 id="How-the-documentation-is-structured"><a class="docs-heading-anchor" href="#How-the-documentation-is-structured">How the documentation is structured</a><a id="How-the-documentation-is-structured-1"></a><a class="docs-heading-anchor-permalink" href="#How-the-documentation-is-structured" title="Permalink"></a></h2><p>Having a high-level overview of how this documentation is structured will help you know where to look for certain things.</p><ul><li><p><strong>Tutorials</strong> contain worked examples of solving problems with JuMP. Start here if you are new to JuMP, or you have a particular problem class you want to model.</p></li><li><p>The <strong>Manual</strong> contains short code-snippets that explain how to achieve specific tasks in JuMP. Look here if you want to know how to achieve a particular task, such as how to <a href="manual/variables/#delete_a_variable">Delete a variable</a> or how to <a href="manual/objective/#Modify-an-objective-coefficient">Modify an objective coefficient</a>.</p></li><li><p>The <strong>API Reference</strong> contains a complete list of the functions you can use in JuMP. Look here if you want to know how to use a particular function.</p></li><li><p>The <strong>Background information</strong> section contains background reading material to provide context to JuMP. Look here if you want an understanding of what JuMP is and why we created it, rather than how to use it.</p></li><li><p>The <strong>Developer docs</strong> section contains information for people contributing to JuMP development or writing JuMP extensions. Don&#39;t worry about this section if you are using JuMP to formulate and solve problems as a user.</p></li><li><p>The <strong>MathOptInterface</strong> section is a self-contained copy of the documentation for MathOptInterface. Look here for functions and constants beginning with <code>MOI.</code>, as well as for general information on how MathOptInterface works.</p></li></ul><h2 id="Citing-JuMP"><a class="docs-heading-anchor" href="#Citing-JuMP">Citing JuMP</a><a id="Citing-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-JuMP" title="Permalink"></a></h2><p>If you find JuMP useful in your work, we kindly request that you cite the following paper (<a href="https://arxiv.org/abs/2206.03866">preprint</a>):</p><pre><code class="language-bibtex hljs">@article{Lubin2023,
+    author = {Miles Lubin and Oscar Dowson and Joaquim {Dias Garcia} and Joey Huchette and Beno{\^i}t Legat and Juan Pablo Vielma},
+    title = {{JuMP} 1.0: {R}ecent improvements to a modeling language for mathematical optimization},
+    journal = {Mathematical Programming Computation},
+    year = {2023},
+    doi = {10.1007/s12532-023-00239-3}
+}</code></pre><h2 id="NumFOCUS"><a class="docs-heading-anchor" href="#NumFOCUS">NumFOCUS</a><a id="NumFOCUS-1"></a><a class="docs-heading-anchor-permalink" href="#NumFOCUS" title="Permalink"></a></h2><p><img src="assets/numfocus-logo.png" alt="NumFOCUS logo"/></p><p>JuMP is a Sponsored Project of NumFOCUS, a 501(c)(3) nonprofit charity in the United States. NumFOCUS provides JuMP with fiscal, legal, and administrative support to help ensure the health and sustainability of the project. Visit <a href="https://numfocus.org">numfocus.org</a> for more information.</p><p>You can support JuMP by <a href="https://numfocus.salsalabs.org/donate-to-jump/index.html">donating</a>.</p><p>Donations to JuMP are managed by NumFOCUS. For donors in the United States, your gift is tax-deductible to the extent provided by law. As with any donation, you should consult with your tax adviser about your particular tax situation.</p><p>JuMP&#39;s largest expense is the annual JuMP-dev workshop. Donations will help us provide travel support for JuMP-dev attendees and take advantage of other opportunities that arise to support JuMP development.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>JuMP is licensed under the <a href="https://mozilla.org/MPL/2.0/">MPL-2.0 software license</a>. Consult the <a href="https://github.com/jump-dev/JuMP.jl/blob/master/LICENSE.md">license</a> and the <a href="https://www.mozilla.org/en-US/MPL/2.0/FAQ/">Mozilla FAQ</a> for more information. In addition, JuMP is typically used in conjunction with solver packages and extensions which have their own licences. Consult their package repositories for the specific licenses that apply.</p></article><nav class="docs-footer"><a class="docs-footer-nextpage" href="should_i_use/">Should you use JuMP? »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/installation/index.html b/previews/PR3525/installation/index.html
new file mode 100644
index 00000000000..04d9ebd5fdc
--- /dev/null
+++ b/previews/PR3525/installation/index.html
@@ -0,0 +1,24 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Installation Guide · JuMP</title><meta name="title" content="Installation Guide · JuMP"/><meta property="og:title" content="Installation Guide · JuMP"/><meta property="twitter:title" content="Installation Guide · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL=".."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../assets/documenter.js"></script><script src="../search_index.js"></script><script src="../siteinfo.js"></script><script src="../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../assets/themeswap.js"></script><link href="../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../"><img class="docs-light-only" src="../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox" checked/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../">Introduction</a></li><li><a class="tocitem" href="../should_i_use/">Should you use JuMP?</a></li><li class="is-active"><a class="tocitem" href>Installation Guide</a><ul class="internal"><li><a class="tocitem" href="#Install-Julia"><span>Install Julia</span></a></li><li><a class="tocitem" href="#Install-JuMP"><span>Install JuMP</span></a></li><li><a class="tocitem" href="#Install-a-solver"><span>Install a solver</span></a></li><li><a class="tocitem" href="#Supported-solvers"><span>Supported solvers</span></a></li><li><a class="tocitem" href="#AMPL-based-solvers"><span>AMPL-based solvers</span></a></li><li><a class="tocitem" href="#GAMS-based-solvers"><span>GAMS-based solvers</span></a></li><li><a class="tocitem" href="#NEOS-based-solvers"><span>NEOS-based solvers</span></a></li><li><a class="tocitem" href="#Common-installation-issues"><span>Common installation issues</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../manual/models/">Models</a></li><li><a class="tocitem" href="../manual/variables/">Variables</a></li><li><a class="tocitem" href="../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../manual/containers/">Containers</a></li><li><a class="tocitem" href="../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/">Introduction</a></li><li><a class="tocitem" href="../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Introduction</a></li><li class="is-active"><a href>Installation Guide</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Installation Guide</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/installation.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Installation-Guide"><a class="docs-heading-anchor" href="#Installation-Guide">Installation Guide</a><a id="Installation-Guide-1"></a><a class="docs-heading-anchor-permalink" href="#Installation-Guide" title="Permalink"></a></h1><p>This guide explains how to install Julia and  JuMP. If you have installation troubles, read the <a href="#Common-installation-issues">Common installation issues</a> section below.</p><h2 id="Install-Julia"><a class="docs-heading-anchor" href="#Install-Julia">Install Julia</a><a id="Install-Julia-1"></a><a class="docs-heading-anchor-permalink" href="#Install-Julia" title="Permalink"></a></h2><p>JuMP is a package for <a href="https://julialang.org">Julia</a>. To use JuMP, first <a href="https://julialang.org/downloads/">download and install</a> Julia.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you are new to Julia, read our <a href="../tutorials/getting_started/getting_started_with_julia/#Getting-started-with-Julia">Getting started with Julia</a> tutorial.</p></div></div><h3 id="Choosing-a-version"><a class="docs-heading-anchor" href="#Choosing-a-version">Choosing a version</a><a id="Choosing-a-version-1"></a><a class="docs-heading-anchor-permalink" href="#Choosing-a-version" title="Permalink"></a></h3><p>You can install the &quot;Current stable release&quot; or the &quot;Long-term support (LTS) release.&quot;</p><ul><li>The &quot;Current stable release&quot; is the latest release of Julia. It has access to newer features, and is likely faster.</li><li>The &quot;Long-term support release&quot; is an older version of Julia that has continued to receive bug and security fixes. However, it may not have the latest features or performance improvements.</li></ul><p>For most users, you should install the &quot;Current stable release,&quot; and whenever Julia releases a new version of the current stable release, you should update your version of Julia. Note that any code you write on one version of the current stable release will continue to work on all subsequent releases.</p><p>For users in restricted software environments (for example, your enterprise IT controls what software you can install), you may be better off installing the long-term support release because you will not have to update Julia as frequently.</p><h2 id="Install-JuMP"><a class="docs-heading-anchor" href="#Install-JuMP">Install JuMP</a><a id="Install-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Install-JuMP" title="Permalink"></a></h2><p>From Julia, JuMP is installed using the built-in package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;JuMP&quot;)</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>We recommend you create a Pkg <em>environment</em> for each project you use JuMP for, instead of adding lots of packages to the global environment. The <a href="https://julialang.github.io/Pkg.jl/v1/environments/">Pkg manager documentation</a> has more information on this topic.</p></div></div><p>When we release a new version of JuMP, you can update with:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.update(&quot;JuMP&quot;)</code></pre><h2 id="Install-a-solver"><a class="docs-heading-anchor" href="#Install-a-solver">Install a solver</a><a id="Install-a-solver-1"></a><a class="docs-heading-anchor-permalink" href="#Install-a-solver" title="Permalink"></a></h2><p>JuMP depends on solvers to solve optimization problems. Therefore, you will need to install one before you can solve problems with JuMP.</p><p>Install a solver using the Julia package manager, replacing <code>&quot;Clp&quot;</code> by the Julia package name as appropriate.</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Clp&quot;)</code></pre><p>Once installed, you can use Clp as a solver with JuMP as follows, using <a href="../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a> to set solver-specific options:</p><pre><code class="language-julia hljs">using JuMP
+using Clp
+model = Model(Clp.Optimizer)
+set_attribute(model, &quot;LogLevel&quot; =&gt; 1)
+set_attribute(model, &quot;PrimalTolerance&quot; =&gt; 1e-7)</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Most packages follow the <code>ModuleName.Optimizer</code> naming convention, but exceptions may exist. See the README of the Julia package&#39;s GitHub repository for more details on how to use a particular solver, including any solver-specific options.</p></div></div><h2 id="Supported-solvers"><a class="docs-heading-anchor" href="#Supported-solvers">Supported solvers</a><a id="Supported-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Supported-solvers" title="Permalink"></a></h2><p>Most solvers are not written in Julia, and some require commercial licenses to use, so installation is often more complex.</p><ul><li>If a solver has <code>Manual</code> in the <code>Installation</code> column, the solver requires a manual installation step, such as downloading and installing a binary, or obtaining a commercial license. Consult the README of the relevant Julia package for more information.</li><li>If the solver has <code>Manualᴹ</code> in the <code>Installation</code> column, the solver requires an installation of <a href="https://www.mathworks.com/products/matlab.html">MATLAB</a>.</li><li>If the <code>Installation</code> column is missing an entry, installing the Julia package will download and install any relevant solver binaries automatically, and you shouldn&#39;t need to do anything other than <code>Pkg.add</code>.</li></ul><p>Solvers with a missing entry in the <code>Julia Package</code> column are written in Julia. The link in the <code>Solver</code> column is the corresponding Julia package.</p><table><tr><th style="text-align: right">Solver</th><th style="text-align: right">Julia Package</th><th style="text-align: right">Installation</th><th style="text-align: right">License</th><th style="text-align: right">Supports</th></tr><tr><td style="text-align: right"><a href="https://github.com/lanl-ansi/Alpine.jl">Alpine.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">Triad NS</td><td style="text-align: right">(MI)NLP</td></tr><tr><td style="text-align: right"><a href="https://www.artelys.com/knitro">Artelys Knitro</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/KNITRO.jl">KNITRO.jl</a></td><td style="text-align: right">Manual</td><td style="text-align: right">Comm.</td><td style="text-align: right">(MI)LP, (MI)SOCP, (MI)NLP</td></tr><tr><td style="text-align: right"><a href="http://minlp.com/baron">BARON</a></td><td style="text-align: right"><a href="https://github.com/joehuchette/BARON.jl">BARON.jl</a></td><td style="text-align: right">Manual</td><td style="text-align: right">Comm.</td><td style="text-align: right">(MI)NLP</td></tr><tr><td style="text-align: right"><a href="http://github.com/coin-or/Bonmin">Bonmin</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/AmplNLWriter.jl">AmplNLWriter.jl</a></td><td style="text-align: right"></td><td style="text-align: right">EPL</td><td style="text-align: right">(MI)NLP</td></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/Cbc">Cbc</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Cbc.jl">Cbc.jl</a></td><td style="text-align: right"></td><td style="text-align: right">EPL</td><td style="text-align: right">(MI)LP</td></tr><tr><td style="text-align: right"><a href="https://github.com/oxfordcontrol/CDCS">CDCS</a></td><td style="text-align: right"><a href="https://github.com/oxfordcontrol/CDCS.jl">CDCS.jl</a></td><td style="text-align: right">Manualᴹ</td><td style="text-align: right">GPL</td><td style="text-align: right">LP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/cddlib/cddlib">CDD</a></td><td style="text-align: right"><a href="https://github.com/JuliaPolyhedra/CDDLib.jl">CDDLib.jl</a></td><td style="text-align: right"></td><td style="text-align: right">GPL</td><td style="text-align: right">LP</td></tr><tr><td style="text-align: right"><a href="https://github.com/oxfordcontrol/Clarabel.jl">Clarabel.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">Apache</td><td style="text-align: right">LP, QP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/Clp">Clp</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Clp.jl">Clp.jl</a></td><td style="text-align: right"></td><td style="text-align: right">EPL</td><td style="text-align: right">LP</td></tr><tr><td style="text-align: right"><a href="https://www.shanshu.ai/copt">COPT</a></td><td style="text-align: right"><a href="https://github.com/COPT-Public/COPT.jl">COPT.jl</a></td><td style="text-align: right"></td><td style="text-align: right">Comm.</td><td style="text-align: right">(MI)LP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/oxfordcontrol/COSMO.jl">COSMO.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">Apache</td><td style="text-align: right">LP, QP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="http://github.com/coin-or/Couenne">Couenne</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/AmplNLWriter.jl">AmplNLWriter.jl</a></td><td style="text-align: right"></td><td style="text-align: right">EPL</td><td style="text-align: right">(MI)NLP</td></tr><tr><td style="text-align: right"><a href="https://www.ibm.com/analytics/cplex-optimizer/">CPLEX</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/CPLEX.jl">CPLEX.jl</a></td><td style="text-align: right">Manual</td><td style="text-align: right">Comm.</td><td style="text-align: right">(MI)LP, (MI)SOCP</td></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/Csdp">CSDP</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/CSDP.jl">CSDP.jl</a></td><td style="text-align: right"></td><td style="text-align: right">EPL</td><td style="text-align: right">LP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/darnstrom/daqp">DAQP</a></td><td style="text-align: right"><a href="https://github.com/darnstrom/DAQP.jl">DAQP.jl</a></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">(Mixed-binary) QP</td></tr><tr><td style="text-align: right"><a href="https://github.com/psorlab/EAGO.jl">EAGO.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">NLP</td></tr><tr><td style="text-align: right"><a href="https://github.com/ifa-ethz/ecos">ECOS</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/ECOS.jl">ECOS.jl</a></td><td style="text-align: right"></td><td style="text-align: right">GPL</td><td style="text-align: right">LP, SOCP</td></tr><tr><td style="text-align: right"><a href="https://www.fico.com/en/products/fico-xpress-optimization-suite">FICO Xpress</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Xpress.jl">Xpress.jl</a></td><td style="text-align: right">Manual</td><td style="text-align: right">Comm.</td><td style="text-align: right">(MI)LP, (MI)SOCP</td></tr><tr><td style="text-align: right"><a href="http://www.gnu.org/software/glpk/">GLPK</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/GLPK.jl">GLPK.jl</a></td><td style="text-align: right"></td><td style="text-align: right">GPL</td><td style="text-align: right">(MI)LP</td></tr><tr><td style="text-align: right"><a href="https://gurobi.com">Gurobi</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Gurobi.jl">Gurobi.jl</a></td><td style="text-align: right">Manual</td><td style="text-align: right">Comm.</td><td style="text-align: right">(MI)LP, (MI)SOCP</td></tr><tr><td style="text-align: right"><a href="https://github.com/ERGO-Code/HiGHS">HiGHS</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/HiGHS.jl">HiGHS.jl</a></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">(MI)LP, QP</td></tr><tr><td style="text-align: right"><a href="https://github.com/chriscoey/Hypatia.jl">Hypatia.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">LP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/Ipopt">Ipopt</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Ipopt.jl">Ipopt.jl</a></td><td style="text-align: right"></td><td style="text-align: right">EPL</td><td style="text-align: right">LP, QP, NLP</td></tr><tr><td style="text-align: right"><a href="https://github.com/lanl-ansi/Juniper.jl">Juniper.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">(MI)SOCP, (MI)NLP</td></tr><tr><td style="text-align: right"><a href="https://github.com/kocvara/Loraine.jl">Loraine.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">LP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/sshin23/MadNLP.jl">MadNLP.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">LP, QP, NLP</td></tr><tr><td style="text-align: right"><a href="https://www.mosek.com/">MOSEK</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/MosekTools.jl">MosekTools.jl</a></td><td style="text-align: right">Manual</td><td style="text-align: right">Comm.</td><td style="text-align: right">(MI)LP, (MI)SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/stevengj/nlopt">NLopt</a></td><td style="text-align: right"><a href="https://github.com/JuliaOpt/NLopt.jl">NLopt.jl</a></td><td style="text-align: right"></td><td style="text-align: right">GPL</td><td style="text-align: right">LP, QP, NLP</td></tr><tr><td style="text-align: right"><a href="https://osqp.org/">OSQP</a></td><td style="text-align: right"><a href="https://github.com/oxfordcontrol/OSQP.jl">OSQP.jl</a></td><td style="text-align: right"></td><td style="text-align: right">Apache</td><td style="text-align: right">LP, QP</td></tr><tr><td style="text-align: right"><a href="http://pages.cs.wisc.edu/~ferris/path.html">PATH</a></td><td style="text-align: right"><a href="https://github.com/chkwon/PATHSolver.jl">PATHSolver.jl</a></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">MCP</td></tr><tr><td style="text-align: right"><a href="https://github.com/jump-dev/Pajarito.jl">Pajarito.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MPL-2</td><td style="text-align: right">(MI)NLP, (MI)SOCP, (MI)SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/jump-dev/Pavito.jl">Pavito.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MPL-2</td><td style="text-align: right">(MI)NLP</td></tr><tr><td style="text-align: right"><a href="http://www.penopt.com/penbmi.html">Penbmi</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Penopt.jl/">Penopt.jl</a></td><td style="text-align: right"></td><td style="text-align: right">Comm.</td><td style="text-align: right">Bilinear SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/mariohsouto/ProxSDP.jl">ProxSDP.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">LP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="https://raposa.usc.es/">RAPOSa</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/AmplNLWriter.jl">AmplNLWriter.jl</a></td><td style="text-align: right">Manual</td><td style="text-align: right">RAPOSa</td><td style="text-align: right">(MI)NLP</td></tr><tr><td style="text-align: right"><a href="https://scipopt.org/">SCIP</a></td><td style="text-align: right"><a href="https://github.com/scipopt/SCIP.jl">SCIP.jl</a></td><td style="text-align: right"></td><td style="text-align: right">Apache</td><td style="text-align: right">(MI)LP, (MI)NLP</td></tr><tr><td style="text-align: right"><a href="https://github.com/cvxgrp/scs">SCS</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/SCS.jl">SCS.jl</a></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">LP, QP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="http://sdpa.sourceforge.net/">SDPA</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/SDPA.jl">SDPA.jl</a>, <a href="https://github.com/ericphanson/SDPAFamily.jl">SDPAFamily.jl</a></td><td style="text-align: right"></td><td style="text-align: right">GPL</td><td style="text-align: right">LP, SDP</td></tr><tr><td style="text-align: right"><a href="https://blog.nus.edu.sg/mattohkc/softwares/sdpnalplus/">SDPNAL</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/SDPNAL.jl">SDPNAL.jl</a></td><td style="text-align: right">Manualᴹ</td><td style="text-align: right">CC BY-SA</td><td style="text-align: right">LP, SDP</td></tr><tr><td style="text-align: right"><a href="https://blog.nus.edu.sg/mattohkc/softwares/sdpt3/">SDPT3</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/SDPT3.jl">SDPT3.jl</a></td><td style="text-align: right">Manualᴹ</td><td style="text-align: right">GPL</td><td style="text-align: right">LP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="http://sedumi.ie.lehigh.edu/">SeDuMi</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/SeDuMi.jl">SeDuMi.jl</a></td><td style="text-align: right">Manualᴹ</td><td style="text-align: right">GPL</td><td style="text-align: right">LP, SOCP, SDP</td></tr><tr><td style="text-align: right"><a href="https://github.com/PharosAbad/StatusSwitchingQP.jl">StatusSwitchingQP.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MIT</td><td style="text-align: right">LP, QP</td></tr><tr><td style="text-align: right"><a href="https://github.com/ds4dm/Tulip.jl">Tulip.jl</a></td><td style="text-align: right"></td><td style="text-align: right"></td><td style="text-align: right">MPL-2</td><td style="text-align: right">LP</td></tr></table><p>Where:</p><ul><li>LP = Linear programming</li><li>QP = Quadratic programming</li><li>SOCP = Second-order conic programming (including problems with convex quadratic constraints or objective)</li><li>MCP = Mixed-complementarity programming</li><li>NLP = Nonlinear programming</li><li>SDP = Semidefinite programming</li><li>(MI)XXX = Mixed-integer equivalent of problem type <code>XXX</code></li></ul><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Developed a solver or solver wrapper? This table is open for new contributions. Edit the <a href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/installation.md">installation.md</a> file, and use the checklist <a href="../developers/checklists/#Adding-a-new-solver-to-the-documentation">Adding a new solver to the documentation</a> when opening the pull request.</p></div></div><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Developing a solver or solver wrapper? See <a href="../manual/models/#jump_models">Models</a> and the <a href="https://jump.dev/MathOptInterface.jl/stable/">MathOptInterface docs</a> for more details on how JuMP interacts with solvers. Please get in touch via the <a href="https://jump.dev/pages/governance/#developer-chatroom">Developer Chatroom</a> with any questions about connecting new solvers with JuMP.</p></div></div><h2 id="AMPL-based-solvers"><a class="docs-heading-anchor" href="#AMPL-based-solvers">AMPL-based solvers</a><a id="AMPL-based-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#AMPL-based-solvers" title="Permalink"></a></h2><p>Use <a href="https://github.com/jump-dev/AmplNLWriter.jl">AmplNLWriter</a> to access solvers that support the <a href="https://en.wikipedia.org/wiki/Nl_(format)">NL format</a>.</p><p>Some solvers, such as <a href="https://github.com/coin-or/Bonmin">Bonmin</a> and <a href="https://github.com/coin-or/Couenne">Couenne</a> can be installed via the Julia package manager. Others need to be manually installed.</p><p>Consult the AMPL documentation for a <a href="https://ampl.com/products/solvers/all-solvers-for-ampl/">complete list of supported solvers</a>.</p><h2 id="GAMS-based-solvers"><a class="docs-heading-anchor" href="#GAMS-based-solvers">GAMS-based solvers</a><a id="GAMS-based-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#GAMS-based-solvers" title="Permalink"></a></h2><p>Use <a href="https://github.com/GAMS-dev/gams.jl">GAMS.jl</a> to access solvers available through <a href="https://www.gams.com">GAMS</a>. Such solvers include: <a href="https://www.gams.com/latest/docs/S_ALPHAECP.html">AlphaECP</a>, <a href="https://www.gams.com/latest/docs/S_ANTIGONE.html">Antigone</a>, <a href="https://www.gams.com/latest/docs/S_BARON.html">BARON</a>, <a href="https://www.gams.com/latest/docs/S_CONOPT.html">CONOPT</a>, <a href="https://www.gams.com/latest/docs/S_COUENNE.html">Couenne</a>, <a href="https://www.gams.com/latest/docs/S_LOCALSOLVER.html">LocalSolver</a>, <a href="https://www.gams.com/latest/docs/S_PATHNLP.html">PATHNLP</a>, <a href="https://www.gams.com/latest/docs/S_SHOT.html">SHOT</a>, <a href="https://www.gams.com/latest/docs/S_SNOPT.html">SNOPT</a>, <a href="https://www.gams.com/latest/docs/S_SOPLEX.html">SoPlex</a>. See a complete list <a href="https://www.gams.com/latest/docs/S_MAIN.html">here</a>.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p><a href="https://github.com/GAMS-dev/gams.jl">GAMS.jl</a> requires an installation of the commercial software <a href="https://www.gams.com">GAMS</a> for which a <a href="https://www.gams.com/latest/docs/UG_License.html#GAMS_Community_Licenses">free community license</a> exists.</p></div></div><h2 id="NEOS-based-solvers"><a class="docs-heading-anchor" href="#NEOS-based-solvers">NEOS-based solvers</a><a id="NEOS-based-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#NEOS-based-solvers" title="Permalink"></a></h2><p>Use <a href="https://github.com/odow/NEOSServer.jl">NEOSServer.jl</a> to access solvers available through the <a href="https://neos-server.org">NEOS Server</a>.</p><h2 id="Common-installation-issues"><a class="docs-heading-anchor" href="#Common-installation-issues">Common installation issues</a><a id="Common-installation-issues-1"></a><a class="docs-heading-anchor-permalink" href="#Common-installation-issues" title="Permalink"></a></h2><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>When in doubt, run <code>import Pkg; Pkg.update()</code> to see if updating your packages fixes the issue. Remember you will need to exit Julia and start a new session for the changes to take effect.</p></div></div><h3 id="Check-the-version-of-your-packages"><a class="docs-heading-anchor" href="#Check-the-version-of-your-packages">Check the version of your packages</a><a id="Check-the-version-of-your-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Check-the-version-of-your-packages" title="Permalink"></a></h3><p>Each package is versioned with a <a href="https://semver.org">three-part number</a> of the form <code>vX.Y.Z</code>. You can check which versions you have installed with <code>import Pkg; Pkg.status()</code>.</p><p>This should almost always be the most-recent release. You can check the releases of a package by going to the relevant GitHub page, and navigating to the &quot;releases&quot; page. For example, the list of JuMP releases is available at: <a href="https://github.com/jump-dev/JuMP.jl/releases">https://github.com/jump-dev/JuMP.jl/releases</a>.</p><p>If you post on the <a href="https://jump.dev/forum">community forum</a>, please include the output of <code>Pkg.status()</code>.</p><h3 id="Unsatisfiable-requirements-detected"><a class="docs-heading-anchor" href="#Unsatisfiable-requirements-detected">Unsatisfiable requirements detected</a><a id="Unsatisfiable-requirements-detected-1"></a><a class="docs-heading-anchor-permalink" href="#Unsatisfiable-requirements-detected" title="Permalink"></a></h3><p>Did you get an error like <code>Unsatisfiable requirements detected for package JuMP</code>? The Pkg documentation has a <a href="https://julialang.github.io/Pkg.jl/v1/managing-packages/#conflicts">section on how to understand and manage these conflicts</a>.</p><h3 id="Installing-new-packages-can-make-JuMP-downgrade-to-an-earlier-version"><a class="docs-heading-anchor" href="#Installing-new-packages-can-make-JuMP-downgrade-to-an-earlier-version">Installing new packages can make JuMP downgrade to an earlier version</a><a id="Installing-new-packages-can-make-JuMP-downgrade-to-an-earlier-version-1"></a><a class="docs-heading-anchor-permalink" href="#Installing-new-packages-can-make-JuMP-downgrade-to-an-earlier-version" title="Permalink"></a></h3><p>Another common complaint is that after adding a new package, code that previously worked no longer works.</p><p>This usually happens because the new package is not compatible with the latest version of JuMP. Therefore, the package manager rolls-back JuMP to an earlier version. Here&#39;s an example.</p><p>First, we add JuMP:</p><pre><code class="language-julia hljs">(jump_example) pkg&gt; add JuMP
+  Resolving package versions...
+Updating `~/jump_example/Project.toml`
+  [4076af6c] + JuMP v0.21.5
+Updating `~/jump_example/Manifest.toml`
+  ... lines omitted ...</code></pre><p>The <code>+ JuMP v0.21.5</code> line indicates that JuMP has been added at version <code>0.21.5</code>. However, watch what happens when we add <a href="https://github.com/iainnz/JuMPeR.jl">JuMPeR</a>:</p><pre><code class="language-julia hljs">(jump_example) pkg&gt; add JuMPeR
+  Resolving package versions...
+Updating `~/jump_example/Project.toml`
+  [4076af6c] ↓ JuMP v0.21.5 ⇒ v0.18.6
+  [707a9f91] + JuMPeR v0.6.0
+Updating `~/jump_example/Manifest.toml`
+  ... lines omitted ...</code></pre><p>JuMPeR gets added at version <code>0.6.0</code> (<code>+ JuMPeR v0.6.0</code>), but JuMP gets downgraded from <code>0.21.5</code> to <code>0.18.6</code> (<code>↓ JuMP v0.21.5 ⇒ v0.18.6</code>)! The reason for this is that JuMPeR doesn&#39;t support a version of JuMP newer than <code>0.18.6</code>.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Pay careful attention to the output of the package manager when adding new packages, especially when you see a package being downgraded.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../should_i_use/">« Should you use JuMP?</a><a class="docs-footer-nextpage" href="../tutorials/getting_started/introduction/">Introduction »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/callbacks/index.html b/previews/PR3525/manual/callbacks/index.html
new file mode 100644
index 00000000000..39816634eef
--- /dev/null
+++ b/previews/PR3525/manual/callbacks/index.html
@@ -0,0 +1,87 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Solver-independent Callbacks · JuMP</title><meta name="title" content="Solver-independent Callbacks · JuMP"/><meta property="og:title" content="Solver-independent Callbacks · JuMP"/><meta property="twitter:title" content="Solver-independent Callbacks · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li class="is-active"><a class="tocitem" href>Solver-independent Callbacks</a><ul class="internal"><li><a class="tocitem" href="#Available-solvers"><span>Available solvers</span></a></li><li><a class="tocitem" href="#Things-you-can-and-cannot-do-during-solver-independent-callbacks"><span>Things you can and cannot do during solver-independent callbacks</span></a></li><li><a class="tocitem" href="#Lazy-constraints"><span>Lazy constraints</span></a></li><li><a class="tocitem" href="#User-cuts"><span>User cuts</span></a></li><li><a class="tocitem" href="#Heuristic-solutions"><span>Heuristic solutions</span></a></li></ul></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Solver-independent Callbacks</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Solver-independent Callbacks</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/callbacks.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="callbacks_manual"><a class="docs-heading-anchor" href="#callbacks_manual">Solver-independent Callbacks</a><a id="callbacks_manual-1"></a><a class="docs-heading-anchor-permalink" href="#callbacks_manual" title="Permalink"></a></h1><p>Many mixed-integer (linear, conic, and nonlinear) programming solvers offer the ability to modify the solve process. Examples include changing branching decisions in branch-and-bound, adding custom cutting planes, providing custom heuristics to find feasible solutions, or implementing on-demand separators to add new constraints only when they are violated by the current solution (also known as lazy constraints).</p><p>While historically this functionality has been limited to solver-specific interfaces, JuMP provides solver-independent support for three types of callbacks:</p><ol><li>lazy constraints</li><li>user-cuts</li><li>heuristic solutions</li></ol><h2 id="Available-solvers"><a class="docs-heading-anchor" href="#Available-solvers">Available solvers</a><a id="Available-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Available-solvers" title="Permalink"></a></h2><p>Solver-independent callback support is limited to a few solvers. This includes <a href="https://github.com/jump-dev/CPLEX.jl">CPLEX</a>, <a href="https://github.com/jump-dev/GLPK.jl">GLPK</a>, <a href="https://github.com/jump-dev/Gurobi.jl">Gurobi</a>, <a href="https://github.com/jump-dev/Xpress.jl">Xpress</a>, and <a href="https://github.com/scipopt/SCIP.jl">SCIP</a>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>While JuMP provides a solver-independent way of accessing callbacks, you should not assume that you will see identical behavior when running the same code on different solvers. For example, some solvers may ignore user-cuts for various reasons, while other solvers may add every user-cut. Read the underlying solver&#39;s callback documentation to understand details specific to each solver.</p></div></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This page discusses solver-<em>independent</em> callbacks. However, each solver listed above also provides a solver-<em>dependent</em> callback to provide access to the full range of solver-specific features. Consult the solver&#39;s README for an example of how to use the solver-dependent callback. This will require you to understand the C interface of the solver.</p></div></div><h2 id="Things-you-can-and-cannot-do-during-solver-independent-callbacks"><a class="docs-heading-anchor" href="#Things-you-can-and-cannot-do-during-solver-independent-callbacks">Things you can and cannot do during solver-independent callbacks</a><a id="Things-you-can-and-cannot-do-during-solver-independent-callbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Things-you-can-and-cannot-do-during-solver-independent-callbacks" title="Permalink"></a></h2><p>There is a limited range of things you can do during a callback. Only use the functions and macros explicitly stated in this page of the documentation, or in the <a href="../../tutorials/linear/callbacks/#callbacks_tutorial">Callbacks tutorial</a>.</p><p>Using any other part of the JuMP API (for example, adding a constraint with <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> or modifying a variable bound with <a href="../../api/JuMP/#JuMP.set_lower_bound"><code>set_lower_bound</code></a>) is undefined behavior, and your solver may throw an error, return an incorrect solution, or result in a segfault that aborts Julia.</p><p>In each of the three solver-independent callbacks, there are two things you may query:</p><ul><li><a href="../../api/JuMP/#JuMP.callback_node_status"><code>callback_node_status</code></a> returns an <a href="../../moi/reference/callbacks/#MathOptInterface.CallbackNodeStatusCode"><code>MOI.CallbackNodeStatusCode</code></a> enum indicating if the current primal solution is integer feasible.</li><li><a href="../../api/JuMP/#JuMP.callback_value"><code>callback_value</code></a> returns the current primal solution of a variable.</li></ul><p>If you need to query any other information, use a solver-dependent callback instead. Each solver supporting a solver-dependent callback has information on how to use it in the README of their GitHub repository.</p><p>If you want to modify the problem in a callback, you <em>must</em> use a lazy constraint.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>You can only set each callback once. Calling <code>set</code> twice will over-write the earlier callback. In addition, if you use a solver-independent callback, you cannot set a solver-dependent callback.</p></div></div><h2 id="Lazy-constraints"><a class="docs-heading-anchor" href="#Lazy-constraints">Lazy constraints</a><a id="Lazy-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Lazy-constraints" title="Permalink"></a></h2><p>Lazy constraints are useful when the full set of constraints is too large to explicitly include in the initial formulation. When a MIP solver reaches a new solution, for example with a heuristic or by solving a problem at a node in the branch-and-bound tree, it will give the user the chance to provide constraints that would make the current solution infeasible. For some more information about lazy constraints, see this <a href="https://orinanobworld.blogspot.com/2012/08/user-cuts-versus-lazy-constraints.html">blog post by Paul Rubin</a>.</p><p>A lazy constraint callback can be set using the following syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; import GLPK
+
+julia&gt; model = Model(GLPK.Optimizer);
+
+julia&gt; @variable(model, x &lt;= 10, Int)
+x
+
+julia&gt; @objective(model, Max, x)
+x
+
+julia&gt; function my_callback_function(cb_data)
+           status = callback_node_status(cb_data, model)
+           if status == MOI.CALLBACK_NODE_STATUS_FRACTIONAL
+               # `callback_value(cb_data, x)` is not integer (to some tolerance).
+               # If, for example, your lazy constraint generator requires an
+               # integer-feasible primal solution, you can add a `return` here.
+               return
+           elseif status == MOI.CALLBACK_NODE_STATUS_INTEGER
+               # `callback_value(cb_data, x)` is integer (to some tolerance).
+           else
+               @assert status == MOI.CALLBACK_NODE_STATUS_UNKNOWN
+               # `callback_value(cb_data, x)` might be fractional or integer.
+           end
+           x_val = callback_value(cb_data, x)
+           if x_val &gt; 2 + 1e-6
+               con = @build_constraint(x &lt;= 2)
+               MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+           end
+       end
+my_callback_function (generic function with 1 method)
+
+julia&gt; set_attribute(model, MOI.LazyConstraintCallback(), my_callback_function)</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The lazy constraint callback <em>may</em> be called at fractional or integer nodes in the branch-and-bound tree. There is no guarantee that the callback is called at <em>every</em> primal solution.</p></div></div><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Only add a lazy constraint if your primal solution violates the constraint. Adding the lazy constraint irrespective of feasibility may result in the solver returning an incorrect solution, or lead to many constraints being added, slowing down the solution process.</p><pre><code class="language-julia hljs">model = Model(GLPK.Optimizer)
+@variable(model, x &lt;= 10, Int)
+@objective(model, Max, x)
+function bad_callback_function(cb_data)
+    # Don&#39;t do this!
+    con = @build_constraint(x &lt;= 2)
+    MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+end
+function good_callback_function(cb_data)
+    if callback_value(x) &gt; 2
+        con = @build_constraint(x &lt;= 2)
+        MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+    end
+end
+set_attribute(model, MOI.LazyConstraintCallback(), good_callback_function)</code></pre></div></div><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>During the solve, a solver may visit a point that was cut off by a previous lazy constraint, for example, because the earlier lazy constraint was removed during presolve. If this happens, you must re-add the lazy constraint.</p></div></div><h2 id="User-cuts"><a class="docs-heading-anchor" href="#User-cuts">User cuts</a><a id="User-cuts-1"></a><a class="docs-heading-anchor-permalink" href="#User-cuts" title="Permalink"></a></h2><p>User cuts, or simply cuts, provide a way for the user to tighten the LP relaxation using problem-specific knowledge that the solver cannot or is unable to infer from the model. Just like with lazy constraints, when a MIP solver reaches a new node in the branch-and-bound tree, it will give the user the chance to provide cuts to make the current relaxed (fractional) solution infeasible in the hopes of obtaining an integer solution. For more details about the difference between user cuts and lazy constraints see the aforementioned <a href="https://orinanobworld.blogspot.com/2012/08/user-cuts-versus-lazy-constraints.html">blog post</a>.</p><p>A user-cut callback can be set using the following syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; import GLPK
+
+julia&gt; model = Model(GLPK.Optimizer);
+
+julia&gt; @variable(model, x &lt;= 10.5, Int)
+x
+
+julia&gt; @objective(model, Max, x)
+x
+
+julia&gt; function my_callback_function(cb_data)
+           x_val = callback_value(cb_data, x)
+           con = @build_constraint(x &lt;= floor(x_val))
+           MOI.submit(model, MOI.UserCut(cb_data), con)
+       end
+my_callback_function (generic function with 1 method)
+
+julia&gt; set_attribute(model, MOI.UserCutCallback(), my_callback_function)</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>User cuts must not change the set of integer feasible solutions. Equivalently, user cuts can only remove fractional solutions. If you add a cut that removes an integer solution (even one that is not optimal), the solver may return an incorrect solution.</p></div></div><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The user-cut callback <em>may</em> be called at fractional nodes in the branch-and-bound tree. There is no guarantee that the callback is called at <em>every</em> fractional primal solution.</p></div></div><h2 id="Heuristic-solutions"><a class="docs-heading-anchor" href="#Heuristic-solutions">Heuristic solutions</a><a id="Heuristic-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Heuristic-solutions" title="Permalink"></a></h2><p>Integer programming solvers frequently include heuristics that run at the nodes of the branch-and-bound tree. They aim to find integer solutions quicker than plain branch-and-bound would to tighten the bound, allowing us to fathom nodes quicker and to tighten the integrality gap.</p><p>Some heuristics take integer solutions and explore their &quot;local neighborhood&quot; (for example, flipping binary variables, fix some variables and solve a smaller MILP) and others take fractional solutions and attempt to round them in an intelligent way.</p><p>You may want to add a heuristic of your own if you have some special insight into the problem structure that the solver is not aware of, for example, you can consistently take fractional solutions and intelligently guess integer solutions from them.</p><p>A heuristic solution callback can be set using the following syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; import GLPK
+
+julia&gt; model = Model(GLPK.Optimizer);
+
+julia&gt; @variable(model, x &lt;= 10.5, Int)
+x
+
+julia&gt; @objective(model, Max, x)
+x
+
+julia&gt; function my_callback_function(cb_data)
+           x_val = callback_value(cb_data, x)
+           status = MOI.submit(
+               model, MOI.HeuristicSolution(cb_data), [x], [floor(Int, x_val)]
+           )
+           println(&quot;I submitted a heuristic solution, and the status was: &quot;, status)
+       end
+my_callback_function (generic function with 1 method)
+
+julia&gt; set_attribute(model, MOI.HeuristicCallback(), my_callback_function)</code></pre><p>The third argument to <code>submit</code> is a vector of JuMP variables, and the fourth argument is a vector of values corresponding to each variable.</p><p><code>MOI.submit</code> returns an enum that depends on whether the solver accepted the solution. The possible return codes are:</p><ul><li><code>MOI.HEURISTIC_SOLUTION_ACCEPTED</code></li><li><code>MOI.HEURISTIC_SOLUTION_REJECTED</code></li><li><code>MOI.HEURISTIC_SOLUTION_UNKNOWN</code></li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Some solvers may accept partial solutions. Others require a feasible integer solution for every variable. If in doubt, provide a complete solution.</p></div></div><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The heuristic solution callback <em>may</em> be called at fractional nodes in the branch-and-bound tree. There is no guarantee that the callback is called at <em>every</em> fractional primal solution.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../solutions/">« Solutions</a><a class="docs-footer-nextpage" href="../complex/">Complex number support »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/complex/index.html b/previews/PR3525/manual/complex/index.html
new file mode 100644
index 00000000000..72e836ff1a1
--- /dev/null
+++ b/previews/PR3525/manual/complex/index.html
@@ -0,0 +1,151 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Complex number support · JuMP</title><meta name="title" content="Complex number support · JuMP"/><meta property="og:title" content="Complex number support · JuMP"/><meta property="twitter:title" content="Complex number support · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li class="is-active"><a class="tocitem" href>Complex number support</a><ul class="internal"><li><a class="tocitem" href="#Complex-valued-variables"><span>Complex-valued variables</span></a></li><li><a class="tocitem" href="#Complex-valued-variable-bounds"><span>Complex-valued variable bounds</span></a></li><li><a class="tocitem" href="#Complex-valued-equality-constraints"><span>Complex-valued equality constraints</span></a></li><li><a class="tocitem" href="#Hermitian-PSD-Cones"><span>Hermitian PSD Cones</span></a></li><li><a class="tocitem" href="#Hermitian-PSD-constraints"><span>Hermitian PSD constraints</span></a></li></ul></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Complex number support</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Complex number support</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/complex.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Complex-number-support"><a class="docs-heading-anchor" href="#Complex-number-support">Complex number support</a><a id="Complex-number-support-1"></a><a class="docs-heading-anchor-permalink" href="#Complex-number-support" title="Permalink"></a></h1><p>This page explains the complex-valued variables and constraints that JuMP supports. For a worked-example using these features, read the <a href="../../tutorials/conic/quantum_discrimination/#Quantum-state-discrimination">Quantum state discrimination</a> tutorial.</p><h2 id="Complex-valued-variables"><a class="docs-heading-anchor" href="#Complex-valued-variables">Complex-valued variables</a><a id="Complex-valued-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Complex-valued-variables" title="Permalink"></a></h2><p>Create a complex-valued variable using <a href="../../api/JuMP/#JuMP.ComplexPlane"><code>ComplexPlane</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x in ComplexPlane())
+real(x) + imag(x) im</code></pre><p>Note that <code>x</code> is not a <a href="../../api/JuMP/#JuMP.VariableRef"><code>VariableRef</code></a>; instead, it is an affine expression with <code>Complex{Float64}</code>-valued coefficients:</p><pre><code class="language-julia-repl hljs">julia&gt; typeof(x)
+GenericAffExpr{ComplexF64, VariableRef}</code></pre><p>Behind the scenes, JuMP has created two real-valued variables, with names <code>&quot;real(x)&quot;</code> and <code>&quot;imag(x)&quot;</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; all_variables(model)
+2-element Vector{VariableRef}:
+ real(x)
+ imag(x)
+
+julia&gt; name.(all_variables(model))
+2-element Vector{String}:
+ &quot;real(x)&quot;
+ &quot;imag(x)&quot;</code></pre><p>Use the <code>real</code> and <code>imag</code> functions on <code>x</code> to return a real-valued affine expression representing each variable:</p><pre><code class="language-julia-repl hljs">julia&gt; typeof(real(x))
+AffExpr (alias for GenericAffExpr{Float64, GenericVariableRef{Float64}})
+
+julia&gt; typeof(imag(x))
+AffExpr (alias for GenericAffExpr{Float64, GenericVariableRef{Float64}})</code></pre><p>To create an anonymous variable, use the <code>set</code> keyword argument:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x = @variable(model, set = ComplexPlane())
+_[1] + _[2] im</code></pre><h2 id="Complex-valued-variable-bounds"><a class="docs-heading-anchor" href="#Complex-valued-variable-bounds">Complex-valued variable bounds</a><a id="Complex-valued-variable-bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Complex-valued-variable-bounds" title="Permalink"></a></h2><p>Because complex-valued variables lack a total ordering, the definition of a variable bound for a complex-valued variable is ambiguous. If you pass a real- or complex-valued argument to keywords such as <code>lower_bound</code>, <code>upper_bound</code>, and <code>start_value</code>, JuMP will apply the real and imaginary parts to the associated real-valued variables.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(
+           model,
+           x in ComplexPlane(),
+           lower_bound = 1.0,
+           upper_bound = 2.0 + 3.0im,
+           start = 4im,
+       )
+real(x) + imag(x) im
+
+julia&gt; vars = all_variables(model)
+2-element Vector{VariableRef}:
+ real(x)
+ imag(x)
+
+julia&gt; lower_bound.(vars)
+2-element Vector{Float64}:
+ 1.0
+ 0.0
+
+julia&gt; upper_bound.(vars)
+2-element Vector{Float64}:
+ 2.0
+ 3.0
+
+julia&gt; start_value.(vars)
+2-element Vector{Float64}:
+ 0.0
+ 4.0</code></pre><h2 id="Complex-valued-equality-constraints"><a class="docs-heading-anchor" href="#Complex-valued-equality-constraints">Complex-valued equality constraints</a><a id="Complex-valued-equality-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Complex-valued-equality-constraints" title="Permalink"></a></h2><p>JuMP reformulates complex-valued equality constraints into two real-valued constraints: one representing the real part, and one representing the imaginary part. Thus, complex-valued equality constraints can be solved any solver that supports the real-valued constraint type.</p><p>For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @constraint(model, (1 + 2im) * x[1] + 3 * x[2] == 4 + 5im)
+(1 + 2im) x[1] + 3 x[2] = (4 + 5im)
+
+julia&gt; optimize!(model)
+
+julia&gt; value.(x)
+2-element Vector{Float64}:
+ 2.5
+ 0.5</code></pre><p>is equivalent to</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @constraint(model, 1 * x[1] + 3 * x[2] == 4)  # real component
+x[1] + 3 x[2] = 4
+
+julia&gt; @constraint(model, 2 * x[1] == 5)  # imag component
+2 x[1] = 5
+
+julia&gt; optimize!(model)
+
+julia&gt; value.(x)
+2-element Vector{Float64}:
+ 2.5
+ 0.5</code></pre><p>This also applies if the variables are complex-valued:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x in ComplexPlane());
+
+julia&gt; @constraint(model, (1 + 2im) * x + 3 * x == 4 + 5im)
+(4 + 2im) real(x) + (-2 + 4im) imag(x) = (4 + 5im)
+
+julia&gt; optimize!(model)
+
+julia&gt; value(x)
+1.3 + 0.6000000000000001im</code></pre><p>which is equivalent to</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x_real);
+
+julia&gt; @variable(model, x_imag);
+
+julia&gt; @constraint(model, x_real - 2 * x_imag + 3 * x_real == 4)
+4 x_real - 2 x_imag = 4
+
+julia&gt; @constraint(model, x_imag + 2 * x_real + 3 * x_imag == 5)
+2 x_real + 4 x_imag = 5
+
+julia&gt; optimize!(model)
+
+julia&gt; value(x_real) + value(x_imag) * im
+1.3 + 0.6000000000000001im</code></pre><h2 id="Hermitian-PSD-Cones"><a class="docs-heading-anchor" href="#Hermitian-PSD-Cones">Hermitian PSD Cones</a><a id="Hermitian-PSD-Cones-1"></a><a class="docs-heading-anchor-permalink" href="#Hermitian-PSD-Cones" title="Permalink"></a></h2><p>JuMP supports creating matrices where are Hermitian.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, H[1:3, 1:3] in HermitianPSDCone())
+3×3 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(H[1,1])                    …  real(H[1,3]) + imag(H[1,3]) im
+ real(H[1,2]) - imag(H[1,2]) im     real(H[2,3]) + imag(H[2,3]) im
+ real(H[1,3]) - imag(H[1,3]) im     real(H[3,3])</code></pre><p>Behind the scenes, JuMP has created nine real-valued decision variables:</p><pre><code class="language-julia-repl hljs">julia&gt; all_variables(model)
+9-element Vector{VariableRef}:
+ real(H[1,1])
+ real(H[1,2])
+ real(H[2,2])
+ real(H[1,3])
+ real(H[2,3])
+ real(H[3,3])
+ imag(H[1,2])
+ imag(H[1,3])
+ imag(H[2,3])</code></pre><p>and a <code>Vector{VariableRef}-in-MOI.HermitianPositiveSemidefiniteConeTriangle</code> constraint:</p><pre><code class="language-julia-repl hljs">julia&gt; num_constraints(model, Vector{VariableRef}, MOI.HermitianPositiveSemidefiniteConeTriangle)
+1</code></pre><p>The <a href="../../moi/reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>MOI.HermitianPositiveSemidefiniteConeTriangle</code></a> set can be efficiently bridged to <a href="../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a>, so it can be solved by any solver that supports PSD constraints.</p><p>Each element of <code>H</code> is an affine expression with <code>Complex{Float64}</code>-valued coefficients:</p><pre><code class="language-julia-repl hljs">julia&gt; typeof(H[1, 1])
+GenericAffExpr{ComplexF64, VariableRef}
+
+julia&gt; typeof(H[2, 1])
+GenericAffExpr{ComplexF64, VariableRef}</code></pre><h2 id="Hermitian-PSD-constraints"><a class="docs-heading-anchor" href="#Hermitian-PSD-constraints">Hermitian PSD constraints</a><a id="Hermitian-PSD-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Hermitian-PSD-constraints" title="Permalink"></a></h2><p>The <a href="../../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> can also be used in the <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; import LinearAlgebra
+
+julia&gt; H = LinearAlgebra.Hermitian([x[1] 1im; -1im -x[2]])
+2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ x[1]  im
+ -im   -x[2]
+
+julia&gt; @constraint(model, H in HermitianPSDCone())
+[x[1]  im;
+ -im   -x[2]] ∈ HermitianPSDCone()</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The matrix <code>H</code> in <code>H in HermitianPSDCone()</code> must be a <code>LinearAlgebra.Hermitian</code> matrix type. A <code>build_constraint</code> error will be thrown if the matrix is a different matrix type.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../callbacks/">« Solver-independent Callbacks</a><a class="docs-footer-nextpage" href="../nonlinear/">Nonlinear Modeling »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/constraints/index.html b/previews/PR3525/manual/constraints/index.html
new file mode 100644
index 00000000000..b3af07e1fa6
--- /dev/null
+++ b/previews/PR3525/manual/constraints/index.html
@@ -0,0 +1,737 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Constraints · JuMP</title><meta name="title" content="Constraints · JuMP"/><meta property="og:title" content="Constraints · JuMP"/><meta property="twitter:title" content="Constraints · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li class="is-active"><a class="tocitem" href>Constraints</a><ul class="internal"><li><a class="tocitem" href="#Add-a-constraint"><span>Add a constraint</span></a></li><li><a class="tocitem" href="#Vectorized-constraints"><span>Vectorized constraints</span></a></li><li><a class="tocitem" href="#Containers-of-constraints"><span>Containers of constraints</span></a></li><li><a class="tocitem" href="#Registered-constraints"><span>Registered constraints</span></a></li><li><a class="tocitem" href="#Anonymous-constraints"><span>Anonymous constraints</span></a></li><li><a class="tocitem" href="#Constraint-names"><span>Constraint names</span></a></li><li><a class="tocitem" href="#String-names,-symbolic-names,-and-bindings"><span>String names, symbolic names, and bindings</span></a></li><li><a class="tocitem" href="#The-@constraints-macro"><span>The <code>@constraints</code> macro</span></a></li><li><a class="tocitem" href="#constraint_duality"><span>Duality</span></a></li><li><a class="tocitem" href="#Modify-a-constant-term"><span>Modify a constant term</span></a></li><li><a class="tocitem" href="#Modify-a-variable-coefficient"><span>Modify a variable coefficient</span></a></li><li><a class="tocitem" href="#Delete-a-constraint"><span>Delete a constraint</span></a></li><li><a class="tocitem" href="#Start-values"><span>Start values</span></a></li><li><a class="tocitem" href="#Constraint-containers"><span>Constraint containers</span></a></li><li><a class="tocitem" href="#Accessing-constraints-from-a-model"><span>Accessing constraints from a model</span></a></li><li><a class="tocitem" href="#MathOptInterface-constraints"><span>MathOptInterface constraints</span></a></li><li><a class="tocitem" href="#Set-inequality-syntax"><span>Set inequality syntax</span></a></li><li><a class="tocitem" href="#Second-order-cone-constraints"><span>Second-order cone constraints</span></a></li><li><a class="tocitem" href="#Rotated-second-order-cone-constraints"><span>Rotated second-order cone constraints</span></a></li><li><a class="tocitem" href="#Semi-integer-and-semi-continuous-variables"><span>Semi-integer and semi-continuous variables</span></a></li><li><a class="tocitem" href="#Special-Ordered-Sets-of-Type-1"><span>Special Ordered Sets of Type 1</span></a></li><li><a class="tocitem" href="#Special-Ordered-Sets-of-Type-2"><span>Special Ordered Sets of Type 2</span></a></li><li><a class="tocitem" href="#Indicator-constraints"><span>Indicator constraints</span></a></li><li><a class="tocitem" href="#Semidefinite-constraints"><span>Semidefinite constraints</span></a></li><li><a class="tocitem" href="#Complementarity-constraints"><span>Complementarity constraints</span></a></li></ul></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Constraints</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Constraints</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/constraints.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="jump_constraints"><a class="docs-heading-anchor" href="#jump_constraints">Constraints</a><a id="jump_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#jump_constraints" title="Permalink"></a></h1><p>JuMP is based on the <a href="../../moi/#moi_documentation">MathOptInterface (MOI) API</a>. Because of this, JuMP uses the following standard form to represent problems:</p><p class="math-container">\[\begin{align}
+    &amp; \min_{x \in \mathbb{R}^n} &amp; f_0(x)
+    \\
+    &amp; \;\;\text{s.t.} &amp; f_i(x) &amp; \in \mathcal{S}_i &amp; i = 1 \ldots m
+\end{align}\]</p><p>Each constraint, <span>$f_i(x) \in \mathcal{S}_i$</span>, is composed of a function and a set. For example, instead of calling <span>$a^\top x \le b$</span> a <em>less-than-or-equal-to</em> constraint, we say that it is a <em>scalar-affine-in-less-than</em> constraint, where the function <span>$a^\top x$</span> belongs to the <em>less-than</em> set <span>$(-\infty, b]$</span>. We use the shorthand <em>function-in-set</em> to refer to constraints composed of different types of functions and sets.</p><p>This page explains how to write various types of constraints in JuMP. For nonlinear constraints, see <a href="../nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a> instead.</p><h2 id="Add-a-constraint"><a class="docs-heading-anchor" href="#Add-a-constraint">Add a constraint</a><a id="Add-a-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-constraint" title="Permalink"></a></h2><p>Add a constraint to a JuMP model using the <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro. The syntax to use depends on the type of constraint you wish to add.</p><h3 id="Add-a-linear-constraint"><a class="docs-heading-anchor" href="#Add-a-linear-constraint">Add a linear constraint</a><a id="Add-a-linear-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-linear-constraint" title="Permalink"></a></h3><p>Create linear constraints using the <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3]);
+
+julia&gt; @constraint(model, c1, sum(x) &lt;= 1)
+c1 : x[1] + x[2] + x[3] ≤ 1
+
+julia&gt; @constraint(model, c2, x[1] + 2 * x[3] &gt;= 2)
+c2 : x[1] + 2 x[3] ≥ 2
+
+julia&gt; @constraint(model, c3, sum(i * x[i] for i in 1:3) == 3)
+c3 : x[1] + 2 x[2] + 3 x[3] = 3
+
+julia&gt; @constraint(model, c4, 4 &lt;= 2 * x[2] &lt;= 5)
+c4 : 2 x[2] ∈ [4, 5]</code></pre><h3 id="Normalization"><a class="docs-heading-anchor" href="#Normalization">Normalization</a><a id="Normalization-1"></a><a class="docs-heading-anchor-permalink" href="#Normalization" title="Permalink"></a></h3><p>JuMP normalizes constraints by moving all of the terms containing variables to the left-hand side and all of the constant terms to the right-hand side. Thus, we get:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, c, 2x + 1 &lt;= 4x + 4)
+c : -2 x ≤ 3</code></pre><h3 id="quad_constraints"><a class="docs-heading-anchor" href="#quad_constraints">Add a quadratic constraint</a><a id="quad_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#quad_constraints" title="Permalink"></a></h3><p>In addition to affine functions, JuMP also supports constraints with quadratic terms. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; @variable(model, t &gt;= 0)
+t
+
+julia&gt; @constraint(model, my_q, x[1]^2 + x[2]^2 &lt;= t^2)
+my_q : x[1]² + x[2]² - t² ≤ 0</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Because solvers can take advantage of the knowledge that a constraint is quadratic, prefer adding quadratic constraints using <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>, rather than <a href="../../api/JuMP/#JuMP.@NLconstraint"><code>@NLconstraint</code></a>.</p></div></div><h2 id="Vectorized-constraints"><a class="docs-heading-anchor" href="#Vectorized-constraints">Vectorized constraints</a><a id="Vectorized-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Vectorized-constraints" title="Permalink"></a></h2><p>You can also add constraints to JuMP using vectorized linear algebra. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; A = [1 2; 3 4]
+2×2 Matrix{Int64}:
+ 1  2
+ 3  4
+
+julia&gt; b = [5, 6]
+2-element Vector{Int64}:
+ 5
+ 6
+
+julia&gt; @constraint(model, con_vector, A * x == b)
+con_vector : [x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Zeros(2)
+
+julia&gt; @constraint(model, con_scalar, A * x .== b)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}, ScalarShape}}:
+ con_scalar : x[1] + 2 x[2] = 5
+ con_scalar : 3 x[1] + 4 x[2] = 6</code></pre><p>The two constraints, <code>==</code> and <code>.==</code> are similar, but subtly different. The first creates a single constraint that is a <a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a> constraint. The second creates a vector of <a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a> constraints.</p><p>Which formulation to choose depends on the solver, and what you want to do with the constraint object <code>con_vector</code> or <code>con_scalar</code>.</p><ul><li>If you are using a conic solver, expect the dual of <code>con_vector</code> to be a <code>Vector{Float64}</code>, and do not intend to delete a row in the constraint, choose the <code>==</code> formulation.</li><li>If you are using a solver that expects a list of scalar constraints, for example HiGHS, or you wish to delete part of the constraint or access a single row of the constraint, for example, <code>dual(con_scalar[2])</code>, then use the broadcast <code>.==</code>.</li></ul><p>JuMP reformulates both constraints into the other form if needed by the solver, but choosing the right format for a particular solver is more efficient.</p><p>You can also use <code>&lt;=</code>, <code>.&lt;=</code> , <code>&gt;=</code>, and <code>.&gt;=</code> as comparison operators in the constraint.</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, A * x &lt;= b)
+[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonpositives(2)
+
+julia&gt; @constraint(model, A * x .&lt;= b)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ x[1] + 2 x[2] ≤ 5
+ 3 x[1] + 4 x[2] ≤ 6
+
+julia&gt; @constraint(model, A * x &gt;= b)
+[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonnegatives(2)
+
+julia&gt; @constraint(model, A * x .&gt;= b)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:
+ x[1] + 2 x[2] ≥ 5
+ 3 x[1] + 4 x[2] ≥ 6</code></pre><h3 id="Vectorized-matrix-constraints"><a class="docs-heading-anchor" href="#Vectorized-matrix-constraints">Vectorized matrix constraints</a><a id="Vectorized-matrix-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Vectorized-matrix-constraints" title="Permalink"></a></h3><p>In most cases, you cannot use the non-broadcasting syntax for general matrices. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, X[1:2, 1:2])
+2×2 Matrix{VariableRef}:
+ X[1,1]  X[1,2]
+ X[2,1]  X[2,2]
+
+julia&gt; @constraint(model, X &gt;= 0)
+ERROR: At none:1: `@constraint(model, X &gt;= 0)`: Unsupported matrix in vector-valued set. Did you mean to use the broadcasting syntax `.&gt;=` instead? Alternatively, perhaps you are missing a set argument like `@constraint(model, X &gt;= 0, PSDCone())` or `@constraint(model, X &gt;= 0, HermmitianPSDCone())`.
+Stacktrace:
+[...]</code></pre><p>Instead, to represent matrix inequalities you must always use the element-wise broadcasting <code>.==</code>, <code>.&gt;=</code>, or <code>.&lt;=</code>, or use the <a href="#Set-inequality-syntax">Set inequality syntax</a>.</p><p>There are two exceptions: if the result of the left-hand side minus the right-hand side is a <code>LinearAlgebra.Symmetric</code> matrix or a <code>LinearAlgebra.Hermitian</code> matrix, you may use the non-broadcasting equality syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; using LinearAlgebra
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, X[1:2, 1:2], Symmetric)
+2×2 Symmetric{VariableRef, Matrix{VariableRef}}:
+ X[1,1]  X[1,2]
+ X[1,2]  X[2,2]
+
+julia&gt; @constraint(model, X == LinearAlgebra.I)
+[X[1,1] - 1  X[1,2];
+ X[1,2]      X[2,2] - 1] ∈ Zeros()</code></pre><p>Despite the model showing the matrix in <a href="../../api/JuMP/#JuMP.Zeros"><code>Zeros</code></a>, this will add only three rows to the constraint matrix because the symmetric constraints are redundant. In contrast, the broadcasting syntax adds four linear constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, X .== LinearAlgebra.I)
+2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}, ScalarShape}}:
+ X[1,1] = 1  X[1,2] = 0
+ X[1,2] = 0  X[2,2] = 1</code></pre><p>The same holds for <code>LinearAlgebra.Hermitian</code> matrices:</p><pre><code class="language-julia-repl hljs">julia&gt; using LinearAlgebra
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, X[1:2, 1:2] in HermitianPSDCone())
+2×2 Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(X[1,1])                    real(X[1,2]) + imag(X[1,2]) im
+ real(X[1,2]) - imag(X[1,2]) im  real(X[2,2])
+
+julia&gt; @constraint(model, X == LinearAlgebra.I)
+[real(X[1,1]) - 1                real(X[1,2]) + imag(X[1,2]) im;
+ real(X[1,2]) - imag(X[1,2]) im  real(X[2,2]) - 1] ∈ Zeros()
+
+julia&gt; @constraint(model, X .== LinearAlgebra.I)
+2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{ComplexF64}, MathOptInterface.EqualTo{ComplexF64}}, ScalarShape}}:
+ real(X[1,1]) = 1                    real(X[1,2]) + imag(X[1,2]) im = 0
+ real(X[1,2]) - imag(X[1,2]) im = 0  real(X[2,2]) = 1</code></pre><h2 id="Containers-of-constraints"><a class="docs-heading-anchor" href="#Containers-of-constraints">Containers of constraints</a><a id="Containers-of-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Containers-of-constraints" title="Permalink"></a></h2><p>The <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro supports creating collections of constraints. We&#39;ll cover some brief syntax here; read the <a href="#Constraint-containers">Constraint containers</a> section for more details:</p><p>Create arrays of constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3]);
+
+julia&gt; @constraint(model, c[i=1:3], x[i] &lt;= i^2)
+3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ c[1] : x[1] ≤ 1
+ c[2] : x[2] ≤ 4
+ c[3] : x[3] ≤ 9
+
+julia&gt; c[2]
+c[2] : x[2] ≤ 4</code></pre><p>Sets can be any Julia type that supports iteration:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3]);
+
+julia&gt; @constraint(model, c[i=2:3, [&quot;red&quot;, &quot;blue&quot;]], x[i] &lt;= i^2)
+2-dimensional DenseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},2,...} with index sets:
+    Dimension 1, 2:3
+    Dimension 2, [&quot;red&quot;, &quot;blue&quot;]
+And data, a 2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ c[2,red] : x[2] ≤ 4  c[2,blue] : x[2] ≤ 4
+ c[3,red] : x[3] ≤ 9  c[3,blue] : x[3] ≤ 9
+
+julia&gt; c[2, &quot;red&quot;]
+c[2,red] : x[2] ≤ 4</code></pre><p>Sets can depend upon previous indices:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3]);
+
+julia&gt; @constraint(model, c[i=1:3, j=i:3], x[i] &lt;= j)
+JuMP.Containers.SparseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, 2, Tuple{Int64, Int64}} with 6 entries:
+  [1, 1]  =  c[1,1] : x[1] ≤ 1
+  [1, 2]  =  c[1,2] : x[1] ≤ 2
+  [1, 3]  =  c[1,3] : x[1] ≤ 3
+  [2, 2]  =  c[2,2] : x[2] ≤ 2
+  [2, 3]  =  c[2,3] : x[2] ≤ 3
+  [3, 3]  =  c[3,3] : x[3] ≤ 3</code></pre><p>and you can filter elements in the sets using the <code>;</code> syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:9]);
+
+julia&gt; @constraint(model, c[i=1:9; mod(i, 3) == 0], x[i] &lt;= i)
+JuMP.Containers.SparseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, 1, Tuple{Int64}} with 3 entries:
+  [3]  =  c[3] : x[3] ≤ 3
+  [6]  =  c[6] : x[6] ≤ 6
+  [9]  =  c[9] : x[9] ≤ 9</code></pre><h2 id="Registered-constraints"><a class="docs-heading-anchor" href="#Registered-constraints">Registered constraints</a><a id="Registered-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Registered-constraints" title="Permalink"></a></h2><p>When you create constraints, JuMP registers them inside the model using their corresponding symbol. Get a registered name using <code>model[:key]</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model()
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, my_c, 2x &lt;= 1)
+my_c : 2 x ≤ 1
+
+julia&gt; model
+A JuMP Model
+Feasibility problem with:
+Variable: 1
+`AffExpr`-in-`MathOptInterface.LessThan{Float64}`: 1 constraint
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+Names registered in the model: my_c, x
+
+julia&gt; model[:my_c] === my_c
+true</code></pre><h2 id="Anonymous-constraints"><a class="docs-heading-anchor" href="#Anonymous-constraints">Anonymous constraints</a><a id="Anonymous-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Anonymous-constraints" title="Permalink"></a></h2><p>To reduce the likelihood of accidental bugs, and because JuMP registers constraints inside a model, creating two constraints with the same name is an error:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, c, 2x &lt;= 1)
+c : 2 x ≤ 1
+
+julia&gt; @constraint(model, c, 2x &lt;= 1)
+ERROR: An object of name c is already attached to this model. If this
+    is intended, consider using the anonymous construction syntax, e.g.,
+    `x = @variable(model, [1:N], ...)` where the name of the object does
+    not appear inside the macro.
+
+    Alternatively, use `unregister(model, :c)` to first unregister
+    the existing name from the model. Note that this will not delete the
+    object; it will just remove the reference at `model[:c]`.
+[...]</code></pre><p>A common reason for encountering this error is adding constraints in a loop.</p><p>As a work-around, JuMP provides <em>anonymous</em> constraints. Create an anonymous constraint by omitting the name argument:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; c = @constraint(model, 2x &lt;= 1)
+2 x ≤ 1</code></pre><p>Create a container of anonymous constraints by dropping the name in front of the <code>[</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3]);
+
+julia&gt; c = @constraint(model, [i = 1:3], x[i] &lt;= i)
+3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ x[1] ≤ 1
+ x[2] ≤ 2
+ x[3] ≤ 3</code></pre><h2 id="Constraint-names"><a class="docs-heading-anchor" href="#Constraint-names">Constraint names</a><a id="Constraint-names-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-names" title="Permalink"></a></h2><p>In addition to the symbol that constraints are registered with, constraints have a <code>String</code> name that is used for printing and writing to file formats.</p><p>Get and set the name of a constraint using <a href="../../api/JuMP/#JuMP.name"><code>name(::JuMP.ConstraintRef)</code></a> and <a href="../../api/JuMP/#JuMP.set_name"><code>set_name(::JuMP.ConstraintRef, ::String)</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; @constraint(model, con, x &lt;= 1)
+con : x ≤ 1
+
+julia&gt; name(con)
+&quot;con&quot;
+
+julia&gt; set_name(con, &quot;my_con_name&quot;)
+
+julia&gt; con
+my_con_name : x ≤ 1</code></pre><p>Override the default choice of name using the <code>base_name</code> keyword:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; con = @constraint(model, [i=1:2], x &lt;= i, base_name = &quot;my_con&quot;)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ my_con[1] : x ≤ 1
+ my_con[2] : x ≤ 2</code></pre><p>Note that names apply to each element of the container, not to the container of constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; name(con[1])
+&quot;my_con[1]&quot;
+
+julia&gt; set_name(con[1], &quot;c&quot;)
+
+julia&gt; con
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ c : x ≤ 1
+ my_con[2] : x ≤ 2</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>For some models, setting the string name of each constraint can take a non-trivial portion of the total time required to build the model. Turn off <code>String</code> names by passing <code>set_string_name = false</code> to <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con, x &lt;= 2, set_string_name = false)
+x ≤ 2</code></pre><p>See <a href="../../tutorials/getting_started/performance_tips/#Disable-string-names">Disable string names</a> for more information.</p></div></div><h3 id="Retrieve-a-constraint-by-name"><a class="docs-heading-anchor" href="#Retrieve-a-constraint-by-name">Retrieve a constraint by name</a><a id="Retrieve-a-constraint-by-name-1"></a><a class="docs-heading-anchor-permalink" href="#Retrieve-a-constraint-by-name" title="Permalink"></a></h3><p>Retrieve a constraint from a model using <a href="../../api/JuMP/#JuMP.constraint_by_name"><code>constraint_by_name</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; constraint_by_name(model, &quot;c&quot;)
+c : x ≤ 1</code></pre><p>If the name is not present, <code>nothing</code> will be returned:</p><pre><code class="language-julia-repl hljs">julia&gt; constraint_by_name(model, &quot;bad_name&quot;)</code></pre><p>You can only look up individual constraints using <a href="../../api/JuMP/#JuMP.constraint_by_name"><code>constraint_by_name</code></a>. Something like this will not work:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; con = @constraint(model, [i=1:2], x &lt;= i, base_name = &quot;my_con&quot;)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ my_con[1] : x ≤ 1
+ my_con[2] : x ≤ 2
+
+julia&gt; constraint_by_name(model, &quot;my_con&quot;)</code></pre><p>To look up a collection of constraints, do not use <a href="../../api/JuMP/#JuMP.constraint_by_name"><code>constraint_by_name</code></a>. Instead, register them using the <code>model[:key] = value</code> syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(); @variable(model, x);
+
+julia&gt; model[:con] = @constraint(model, [i=1:2], x &lt;= i, base_name = &quot;my_con&quot;)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ my_con[1] : x ≤ 1
+ my_con[2] : x ≤ 2
+
+julia&gt; model[:con]
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ my_con[1] : x ≤ 1
+ my_con[2] : x ≤ 2</code></pre><h2 id="String-names,-symbolic-names,-and-bindings"><a class="docs-heading-anchor" href="#String-names,-symbolic-names,-and-bindings">String names, symbolic names, and bindings</a><a id="String-names,-symbolic-names,-and-bindings-1"></a><a class="docs-heading-anchor-permalink" href="#String-names,-symbolic-names,-and-bindings" title="Permalink"></a></h2><p>It&#39;s common for new users to experience confusion relating to constraints. Part of the problem is the difference between the name that a constraint is registered under and the <code>String</code> name used for printing.</p><p>Here&#39;s a summary of the differences:</p><ul><li>Constraints are created using <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>.</li><li>Constraints can be named or anonymous.</li><li>Named constraints have the form <code>@constraint(model, c, expr)</code>. For named constraints:<ul><li>The <code>String</code> name of the constraint is set to <code>&quot;c&quot;</code>.</li><li>A Julia variable <code>c</code> is created that binds <code>c</code> to  the JuMP constraint.</li><li>The name <code>:c</code> is registered as a key in the model with the value <code>c</code>.</li></ul></li><li>Anonymous constraints have the form <code>c = @constraint(model, expr)</code>. For anonymous constraints:<ul><li>The <code>String</code> name of the constraint is set to <code>&quot;&quot;</code>.</li><li>You control the name of the Julia variable used as the binding.</li><li>No name is registered as a key in the model.</li></ul></li><li>The <code>base_name</code> keyword can override the <code>String</code> name of the constraint.</li><li>You can manually register names in the model via <code>model[:key] = value</code>.</li></ul><p>Here&#39;s an example of the differences:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; c_binding = @constraint(model, 2x &lt;= 1, base_name = &quot;c&quot;)
+c : 2 x ≤ 1
+
+julia&gt; model
+A JuMP Model
+Feasibility problem with:
+Variable: 1
+`AffExpr`-in-`MathOptInterface.LessThan{Float64}`: 1 constraint
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+Names registered in the model: x
+
+julia&gt; c
+ERROR: UndefVarError: `c` not defined
+
+julia&gt; c_binding
+c : 2 x ≤ 1
+
+julia&gt; name(c_binding)
+&quot;c&quot;
+
+julia&gt; model[:c_register] = c_binding
+c : 2 x ≤ 1
+
+julia&gt; model
+A JuMP Model
+Feasibility problem with:
+Variable: 1
+`AffExpr`-in-`MathOptInterface.LessThan{Float64}`: 1 constraint
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+Names registered in the model: c_register, x
+
+julia&gt; model[:c_register]
+c : 2 x ≤ 1
+
+julia&gt; model[:c_register] === c_binding
+true
+
+julia&gt; c
+ERROR: UndefVarError: `c` not defined</code></pre><h2 id="The-@constraints-macro"><a class="docs-heading-anchor" href="#The-@constraints-macro">The <code>@constraints</code> macro</a><a id="The-@constraints-macro-1"></a><a class="docs-heading-anchor-permalink" href="#The-@constraints-macro" title="Permalink"></a></h2><p>If you have many <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> calls, use the <a href="../../api/JuMP/#JuMP.@constraints"><code>@constraints</code></a> macro to improve readability:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraints(model, begin
+           2x &lt;= 1
+           c, x &gt;= -1
+       end)
+(2 x ≤ 1, c : x ≥ -1)
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ c : x ≥ -1
+ 2 x ≤ 1</code></pre><p>The <a href="../../api/JuMP/#JuMP.@constraints"><code>@constraints</code></a> macro returns a tuple of the constraints that were defined.</p><h2 id="constraint_duality"><a class="docs-heading-anchor" href="#constraint_duality">Duality</a><a id="constraint_duality-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_duality" title="Permalink"></a></h2><p>JuMP adopts the notion of <a href="../../moi/background/duality/#Duality">conic duality from MathOptInterface</a>. For linear programs, a feasible dual on a <code>&gt;=</code> constraint is nonnegative and a feasible dual on a <code>&lt;=</code> constraint is nonpositive. If the constraint is an equality constraint, it depends on which direction is binding.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>JuMP&#39;s definition of duality is independent of the objective sense. That is, the sign of feasible duals associated with a constraint depends on the direction of the constraint and not whether the problem is maximization or minimization. <strong>This is a different convention from linear programming duality in some common textbooks.</strong> If you have a linear program, and you want the textbook definition, you probably want to use <a href="../../api/JuMP/#JuMP.shadow_price"><code>shadow_price</code></a> and <a href="../../api/JuMP/#JuMP.reduced_cost"><code>reduced_cost</code></a> instead.</p></div></div><p>The dual value associated with a constraint in the most recent solution can be accessed using the <a href="../../api/JuMP/#JuMP.dual"><code>dual</code></a> function. Use <a href="../../api/JuMP/#JuMP.has_duals"><code>has_duals</code></a> to check if the model has a dual solution available to query. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, con, x &lt;= 1)
+con : x ≤ 1
+
+julia&gt; @objective(model, Min, -2x)
+-2 x
+
+julia&gt; has_duals(model)
+false
+
+julia&gt; optimize!(model)
+
+julia&gt; has_duals(model)
+true
+
+julia&gt; dual(con)
+-2.0
+
+julia&gt; @objective(model, Max, 2x)
+2 x
+
+julia&gt; optimize!(model)
+
+julia&gt; dual(con)
+-2.0</code></pre><p>To help users who may be less familiar with conic duality, JuMP provides <a href="../../api/JuMP/#JuMP.shadow_price"><code>shadow_price</code></a>, which returns a value that can be interpreted as the improvement in the objective in response to an infinitesimal relaxation (on the scale of one unit) in the right-hand side of the constraint. <a href="../../api/JuMP/#JuMP.shadow_price"><code>shadow_price</code></a> can be used only on linear constraints with a <code>&lt;=</code>, <code>&gt;=</code>, or <code>==</code> comparison operator.</p><p>In the example above, <code>dual(con)</code> returned <code>-2.0</code> regardless of the optimization sense. However, in the second case when the optimization sense is <code>Max</code>, <a href="../../api/JuMP/#JuMP.shadow_price"><code>shadow_price</code></a> returns:</p><pre><code class="language-julia-repl hljs">julia&gt; shadow_price(con)
+2.0</code></pre><h3 id="Duals-of-variable-bounds"><a class="docs-heading-anchor" href="#Duals-of-variable-bounds">Duals of variable bounds</a><a id="Duals-of-variable-bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Duals-of-variable-bounds" title="Permalink"></a></h3><p>To query the dual variables associated with a variable bound, first obtain a constraint reference using one of <a href="../../api/JuMP/#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, <a href="../../api/JuMP/#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, or <a href="../../api/JuMP/#JuMP.FixRef"><code>FixRef</code></a>, and then call <a href="../../api/JuMP/#JuMP.dual"><code>dual</code></a> on the returned constraint reference. The <a href="../../api/JuMP/#JuMP.reduced_cost"><code>reduced_cost</code></a> function may simplify this process as it returns the shadow price of an active bound of a variable (or zero, if no active bound exists).</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &lt;= 1)
+x
+
+julia&gt; @objective(model, Min, -2x)
+-2 x
+
+julia&gt; optimize!(model)
+
+julia&gt; dual(UpperBoundRef(x))
+-2.0
+
+julia&gt; reduced_cost(x)
+-2.0</code></pre><h2 id="Modify-a-constant-term"><a class="docs-heading-anchor" href="#Modify-a-constant-term">Modify a constant term</a><a id="Modify-a-constant-term-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-a-constant-term" title="Permalink"></a></h2><p>This section explains how to modify the constant term in a constraint. There are multiple ways to achieve this goal; we explain three options.</p><h3 id="Option-1:-change-the-right-hand-side"><a class="docs-heading-anchor" href="#Option-1:-change-the-right-hand-side">Option 1: change the right-hand side</a><a id="Option-1:-change-the-right-hand-side-1"></a><a class="docs-heading-anchor-permalink" href="#Option-1:-change-the-right-hand-side" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.set_normalized_rhs"><code>set_normalized_rhs</code></a> to modify the right-hand side (constant) term of a linear or quadratic  constraint. Use <a href="../../api/JuMP/#JuMP.normalized_rhs"><code>normalized_rhs</code></a> to query the right-hand side term.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con, 2x &lt;= 1)
+con : 2 x ≤ 1
+
+julia&gt; set_normalized_rhs(con, 3)
+
+julia&gt; con
+con : 2 x ≤ 3
+
+julia&gt; normalized_rhs(con)
+3.0</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><a href="../../api/JuMP/#JuMP.set_normalized_rhs"><code>set_normalized_rhs</code></a> sets the right-hand side term of the normalized constraint. See <a href="#Normalization">Normalization</a> for more details.</p></div></div><h3 id="Option-2:-use-fixed-variables"><a class="docs-heading-anchor" href="#Option-2:-use-fixed-variables">Option 2: use fixed variables</a><a id="Option-2:-use-fixed-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Option-2:-use-fixed-variables" title="Permalink"></a></h3><p>If constraints are complicated, for example, they are composed of a number of components, each of which has a constant term, then it may be difficult to calculate what the right-hand side term is in the standard form.</p><p>For this situation, JuMP includes the ability to <em>fix</em> variables to a value using the <a href="../../api/JuMP/#JuMP.fix"><code>fix</code></a> function. Fixing a variable sets its lower and upper bound to the same value. Thus, changes in a constant term can be simulated by adding a new variable and fixing it to different values. Here is an example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, const_term)
+const_term
+
+julia&gt; @constraint(model, con, 2x &lt;= const_term + 1)
+con : 2 x - const_term ≤ 1
+
+julia&gt; fix(const_term, 1.0)</code></pre><p>The constraint <code>con</code> is now equivalent to <code>2x &lt;= 2</code>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Fixed variables are not replaced with constants when communicating the problem to a solver. Therefore, even though <code>const_term</code> is fixed, it is still a decision variable, and so <code>const_term * x</code> is bilinear.</p></div></div><h3 id="Option-3:-modify-the-function&#39;s-constant-term"><a class="docs-heading-anchor" href="#Option-3:-modify-the-function&#39;s-constant-term">Option 3: modify the function&#39;s constant term</a><a id="Option-3:-modify-the-function&#39;s-constant-term-1"></a><a class="docs-heading-anchor-permalink" href="#Option-3:-modify-the-function&#39;s-constant-term" title="Permalink"></a></h3><p>The third option is to use <a href="../../api/JuMP/#JuMP.add_to_function_constant"><code>add_to_function_constant</code></a>. The constant given is added to the function of a <code>func</code>-in-<code>set</code> constraint. In the following example, adding <code>2</code> to the function has the effect of removing <code>2</code> to the right-hand side:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con, 2x &lt;= 1)
+con : 2 x ≤ 1
+
+julia&gt; add_to_function_constant(con, 2)
+
+julia&gt; con
+con : 2 x ≤ -1
+
+julia&gt; normalized_rhs(con)
+-1.0</code></pre><p>In the case of interval constraints, the constant is removed from each bound:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con, 0 &lt;= 2x + 1 &lt;= 2)
+con : 2 x ∈ [-1, 1]
+
+julia&gt; add_to_function_constant(con, 3)
+
+julia&gt; con
+con : 2 x ∈ [-4, -2]</code></pre><h2 id="Modify-a-variable-coefficient"><a class="docs-heading-anchor" href="#Modify-a-variable-coefficient">Modify a variable coefficient</a><a id="Modify-a-variable-coefficient-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-a-variable-coefficient" title="Permalink"></a></h2><h3 id="Scalar-constraints"><a class="docs-heading-anchor" href="#Scalar-constraints">Scalar constraints</a><a id="Scalar-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Scalar-constraints" title="Permalink"></a></h3><p>To modify the coefficients for a linear term (modifying the coefficient of a quadratic term is not supported) in a constraint, use <a href="../../api/JuMP/#JuMP.set_normalized_coefficient"><code>set_normalized_coefficient</code></a>. To query the current coefficient, use <a href="../../api/JuMP/#JuMP.normalized_coefficient"><code>normalized_coefficient</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @constraint(model, con, 2x[1] + x[2] &lt;= 1)
+con : 2 x[1] + x[2] ≤ 1
+
+julia&gt; set_normalized_coefficient(con, x[2], 0)
+
+julia&gt; con
+con : 2 x[1] ≤ 1
+
+julia&gt; normalized_coefficient(con, x[2])
+0.0</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><a href="../../api/JuMP/#JuMP.set_normalized_coefficient"><code>set_normalized_coefficient</code></a> sets the coefficient of the normalized constraint. See <a href="#Normalization">Normalization</a> for more details.</p></div></div><h3 id="Vector-constraints"><a class="docs-heading-anchor" href="#Vector-constraints">Vector constraints</a><a id="Vector-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Vector-constraints" title="Permalink"></a></h3><p>To modify the coefficients of a vector-valued constraint, use <a href="../../api/JuMP/#JuMP.set_normalized_coefficients"><code>set_normalized_coefficients</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, con, [2x + 3x, 4x] in MOI.Nonnegatives(2))
+con : [5 x, 4 x] ∈ MathOptInterface.Nonnegatives(2)
+
+julia&gt; set_normalized_coefficients(con, x, [(1, 3.0)])
+
+julia&gt; con
+con : [3 x, 4 x] ∈ MathOptInterface.Nonnegatives(2)
+
+julia&gt; set_normalized_coefficients(con, x, [(1, 2.0), (2, 5.0)])
+
+julia&gt; con
+con : [2 x, 5 x] ∈ MathOptInterface.Nonnegatives(2)</code></pre><h2 id="Delete-a-constraint"><a class="docs-heading-anchor" href="#Delete-a-constraint">Delete a constraint</a><a id="Delete-a-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Delete-a-constraint" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a> to delete a constraint from a model. Use <a href="../../api/JuMP/#JuMP.is_valid"><code>is_valid</code></a> to check if a constraint belongs to a model and has not been deleted.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con, 2x &lt;= 1)
+con : 2 x ≤ 1
+
+julia&gt; is_valid(model, con)
+true
+
+julia&gt; delete(model, con)
+
+julia&gt; is_valid(model, con)
+false</code></pre><p>Deleting a constraint does not unregister the symbolic reference from the model. Therefore, creating a new constraint of the same name will throw an error:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, con, 2x &lt;= 1)
+ERROR: An object of name con is already attached to this model. If this
+    is intended, consider using the anonymous construction syntax, e.g.,
+    `x = @variable(model, [1:N], ...)` where the name of the object does
+    not appear inside the macro.
+
+    Alternatively, use `unregister(model, :con)` to first unregister
+    the existing name from the model. Note that this will not delete the
+    object; it will just remove the reference at `model[:con]`.
+[...]</code></pre><p>After calling <a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a>, call <a href="../../api/JuMP/#JuMP.unregister"><code>unregister</code></a> to remove the symbolic reference:</p><pre><code class="language-julia-repl hljs">julia&gt; unregister(model, :con)
+
+julia&gt; @constraint(model, con, 2x &lt;= 1)
+con : 2 x ≤ 1</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p><a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a> does not automatically <a href="../../api/JuMP/#JuMP.unregister"><code>unregister</code></a> because we do not distinguish between names that are automatically registered by JuMP macros, and names that are manually registered by the user by setting values in <a href="../../api/JuMP/#JuMP.object_dictionary"><code>object_dictionary</code></a>. In addition, deleting a constraint and then adding a new constraint of the same name is an easy way to introduce bugs into your code.</p></div></div><h2 id="Start-values"><a class="docs-heading-anchor" href="#Start-values">Start values</a><a id="Start-values-1"></a><a class="docs-heading-anchor-permalink" href="#Start-values" title="Permalink"></a></h2><p>Provide a starting value (also called warmstart) for a constraint&#39;s primal and dual solutions using <a href="../../api/JuMP/#JuMP.set_start_value"><code>set_start_value</code></a> and <a href="../../api/JuMP/#JuMP.set_dual_start_value"><code>set_dual_start_value</code></a>.</p><p>Query the starting value for a constraint&#39;s primal and dual solution using <a href="../../api/JuMP/#JuMP.start_value"><code>start_value</code></a> and <a href="../../api/JuMP/#JuMP.dual_start_value"><code>dual_start_value</code></a>. If no start value has been set, the methods will return <code>nothing</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, con, x &gt;= 10)
+con : x ≥ 10
+
+julia&gt; start_value(con)
+
+julia&gt; set_start_value(con, 10.0)
+
+julia&gt; start_value(con)
+10.0
+
+julia&gt; dual_start_value(con)
+
+julia&gt; set_dual_start_value(con, 2)
+
+julia&gt; dual_start_value(con)
+2.0</code></pre><p>Vector-valued constraints require a vector:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3])
+3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+
+julia&gt; @constraint(model, con, x in SecondOrderCone())
+con : [x[1], x[2], x[3]] in MathOptInterface.SecondOrderCone(3)
+
+julia&gt; dual_start_value(con)
+
+julia&gt; set_dual_start_value(con, [1.0, 2.0, 3.0])
+
+julia&gt; dual_start_value(con)
+3-element Vector{Float64}:
+ 1.0
+ 2.0
+ 3.0</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>To simplify setting start values for all variables and constraints in a model, see <a href="../../api/JuMP/#JuMP.set_start_values"><code>set_start_values</code></a>. The <a href="../../tutorials/conic/start_values/#Primal-and-dual-warm-starts">Primal and dual warm-starts</a> tutorial also gives a detailed description of how to iterate over constraints in the model to set custom start values.</p></div></div><h2 id="Constraint-containers"><a class="docs-heading-anchor" href="#Constraint-containers">Constraint containers</a><a id="Constraint-containers-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-containers" title="Permalink"></a></h2><p>Like <a href="../variables/#Variable-containers">Variable containers</a>, JuMP provides a mechanism for building groups of constraints compactly. References to these groups of constraints are returned in <em>containers</em>. Three types of constraint containers are supported: <code>Array</code>s, <code>DenseAxisArray</code>s, and <code>SparseAxisArray</code>s. We explain each of these in the following.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>You can read more about containers in the <a href="../containers/#Containers">Containers</a> section.</p></div></div><h3 id="constraint_arrays"><a class="docs-heading-anchor" href="#constraint_arrays">Arrays</a><a id="constraint_arrays-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_arrays" title="Permalink"></a></h3><p>One way of adding a group of constraints compactly is the following:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con[i = 1:3], i * x &lt;= i + 1)
+3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ con[1] : x ≤ 2
+ con[2] : 2 x ≤ 3
+ con[3] : 3 x ≤ 4</code></pre><p>JuMP returns references to the three constraints in an <code>Array</code> that is bound to the Julia variable <code>con</code>. This array can be accessed and sliced as you would with any Julia array:</p><pre><code class="language-julia-repl hljs">julia&gt; con[1]
+con[1] : x ≤ 2
+
+julia&gt; con[2:3]
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ con[2] : 2 x ≤ 3
+ con[3] : 3 x ≤ 4</code></pre><p>Anonymous containers can also be constructed by dropping the name (for example, <code>con</code>) before the square brackets:</p><pre><code class="language-julia-repl hljs">julia&gt; con = @constraint(model, [i = 1:2], i * x &lt;= i + 1)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ x ≤ 2
+ 2 x ≤ 3</code></pre><p>Just like <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>, JuMP will form an <code>Array</code> of constraints when it can determine at parse time that the indices are one-based integer ranges. Therefore <code>con[1:b]</code> will create an <code>Array</code>, but <code>con[a:b]</code> will not. A special case is <code>con[Base.OneTo(n)]</code> which will produce an <code>Array</code>. If JuMP cannot determine that the indices are one-based integer ranges (for example, in the case of <code>con[a:b]</code>), JuMP will create a <code>DenseAxisArray</code> instead.</p><h3 id="DenseAxisArrays"><a class="docs-heading-anchor" href="#DenseAxisArrays">DenseAxisArrays</a><a id="DenseAxisArrays-1"></a><a class="docs-heading-anchor-permalink" href="#DenseAxisArrays" title="Permalink"></a></h3><p>The syntax for constructing a <a href="../../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>DenseAxisArray</code></a> of constraints is very similar to the <a href="../variables/#variable_jump_arrays">syntax for constructing</a> a <code>DenseAxisArray</code> of variables.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con[i = 1:2, j = 2:3], i * x &lt;= j + 1)
+2-dimensional DenseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},2,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+    Dimension 2, 2:3
+And data, a 2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ con[1,2] : x ≤ 3    con[1,3] : x ≤ 4
+ con[2,2] : 2 x ≤ 3  con[2,3] : 2 x ≤ 4</code></pre><h3 id="SparseAxisArrays"><a class="docs-heading-anchor" href="#SparseAxisArrays">SparseAxisArrays</a><a id="SparseAxisArrays-1"></a><a class="docs-heading-anchor-permalink" href="#SparseAxisArrays" title="Permalink"></a></h3><p>The syntax for constructing a <a href="../../api/JuMP.Containers/#JuMP.Containers.SparseAxisArray"><code>SparseAxisArray</code></a> of constraints is very similar to the <a href="../variables/#variable_sparseaxisarrays">syntax for constructing</a> a <code>SparseAxisArray</code> of variables.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, con[i = 1:2, j = 1:2; i != j], i * x &lt;= j + 1)
+JuMP.Containers.SparseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, 2, Tuple{Int64, Int64}} with 2 entries:
+  [1, 2]  =  con[1,2] : x ≤ 3
+  [2, 1]  =  con[2,1] : 2 x ≤ 2</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>If you have many index dimensions and a large amount of sparsity, read <a href="../variables/#Performance-considerations">Performance considerations</a>.</p></div></div><h3 id="Forcing-the-container-type"><a class="docs-heading-anchor" href="#Forcing-the-container-type">Forcing the container type</a><a id="Forcing-the-container-type-1"></a><a class="docs-heading-anchor-permalink" href="#Forcing-the-container-type" title="Permalink"></a></h3><p>When creating a container of constraints, JuMP will attempt to choose the tightest container type that can store the constraints. However, because this happens at parse time, it does not always make the best choice. Just like in <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>, you can force the type of container using the <code>container</code> keyword. For syntax and the reason behind this, take a look at the <a href="../variables/#variable_forcing">variable docs</a>.</p><h3 id="Constraints-with-similar-indices"><a class="docs-heading-anchor" href="#Constraints-with-similar-indices">Constraints with similar indices</a><a id="Constraints-with-similar-indices-1"></a><a class="docs-heading-anchor-permalink" href="#Constraints-with-similar-indices" title="Permalink"></a></h3><p>Containers are often used to create constraints over a set of indices. However, you&#39;ll often have cases in which you are repeating the indices:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @variable(model, y[1:2]);
+
+julia&gt; @constraints(model, begin
+           [i=1:2, j=1:2, k=1:2], i * x[j] &lt;= k
+           [i=1:2, j=1:2, k=1:2], i * y[j] &lt;= k
+       end);</code></pre><p>This is hard to read and leads to a lot of copy-paste. A more readable way is to use a for-loop:</p><pre><code class="language-julia-repl hljs">julia&gt; for i=1:2, j=1:2, k=1:2
+           @constraints(model, begin
+               i * x[j] &lt;= k
+               i * y[j] &lt;= k
+           end)
+       end</code></pre><h2 id="Accessing-constraints-from-a-model"><a class="docs-heading-anchor" href="#Accessing-constraints-from-a-model">Accessing constraints from a model</a><a id="Accessing-constraints-from-a-model-1"></a><a class="docs-heading-anchor-permalink" href="#Accessing-constraints-from-a-model" title="Permalink"></a></h2><p>Query the types of function-in-set constraints in a model using <a href="../../api/JuMP/#JuMP.list_of_constraint_types"><code>list_of_constraint_types</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2] &gt;= i, Int);
+
+julia&gt; @constraint(model, x[1] + x[2] &lt;= 1);
+
+julia&gt; list_of_constraint_types(model)
+3-element Vector{Tuple{Type, Type}}:
+ (AffExpr, MathOptInterface.LessThan{Float64})
+ (VariableRef, MathOptInterface.GreaterThan{Float64})
+ (VariableRef, MathOptInterface.Integer)</code></pre><p>For a given combination of function and set type, use <a href="../../api/JuMP/#JuMP.num_constraints"><code>num_constraints</code></a> to access the number of constraints and <a href="../../api/JuMP/#JuMP.all_constraints"><code>all_constraints</code></a> to access a list of their references:</p><pre><code class="language-julia-repl hljs">julia&gt; num_constraints(model, VariableRef, MOI.Integer)
+2
+
+julia&gt; cons = all_constraints(model, VariableRef, MOI.Integer)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}, ScalarShape}}:
+ x[1] integer
+ x[2] integer</code></pre><p>You can also count the total number of constraints in the model, but you must explicitly choose whether to count <code>VariableRef</code> constraints such as bound and integrality constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; num_constraints(model; count_variable_in_set_constraints = true)
+5
+
+julia&gt; num_constraints(model; count_variable_in_set_constraints = false)
+1</code></pre><p>The same also applies for <a href="../../api/JuMP/#JuMP.all_constraints"><code>all_constraints</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; all_constraints(model; include_variable_in_set_constraints = true)
+5-element Vector{ConstraintRef}:
+ x[1] + x[2] ≤ 1
+ x[1] ≥ 1
+ x[2] ≥ 2
+ x[1] integer
+ x[2] integer
+
+julia&gt; all_constraints(model; include_variable_in_set_constraints = false)
+1-element Vector{ConstraintRef}:
+ x[1] + x[2] ≤ 1</code></pre><p>If you need finer-grained control on which constraints to include, use a variant of:</p><pre><code class="language-julia-repl hljs">julia&gt; sum(
+           num_constraints(model, F, S) for
+           (F, S) in list_of_constraint_types(model) if F != VariableRef
+       )
+1</code></pre><p>Use <a href="../../api/JuMP/#JuMP.constraint_object"><code>constraint_object</code></a> to get an instance of an <a href="../../api/JuMP/#JuMP.AbstractConstraint"><code>AbstractConstraint</code></a> object that stores the constraint data:</p><pre><code class="language-julia-repl hljs">julia&gt; con = constraint_object(cons[1])
+ScalarConstraint{VariableRef, MathOptInterface.Integer}(x[1], MathOptInterface.Integer())
+
+julia&gt; con.func
+x[1]
+
+julia&gt; con.set
+MathOptInterface.Integer()</code></pre><h2 id="MathOptInterface-constraints"><a class="docs-heading-anchor" href="#MathOptInterface-constraints">MathOptInterface constraints</a><a id="MathOptInterface-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-constraints" title="Permalink"></a></h2><p>Because JuMP is based on MathOptInterface, you can add any constraints supported by MathOptInterface using the function-in-set syntax. For a list of supported functions and sets, read <a href="../../moi/manual/standard_form/#Standard-form-problem">Standard form problem</a>.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>We use <code>MOI</code> as an alias for the <code>MathOptInterface</code> module. This alias is defined by <code>using JuMP</code>. You may also define it in your code as follows:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre></div></div><p>For example, the following two constraints are equivalent:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3]);
+
+julia&gt; @constraint(model, 2 * x[1] &lt;= 1)
+2 x[1] ≤ 1
+
+julia&gt; @constraint(model, 2 * x[1] in MOI.LessThan(1.0))
+2 x[1] ≤ 1</code></pre><p>You can also use any set defined by MathOptInterface:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, x - [1; 2; 3] in MOI.Nonnegatives(3))
+[x[1] - 1, x[2] - 2, x[3] - 3] ∈ MathOptInterface.Nonnegatives(3)
+
+julia&gt; @constraint(model, x in MOI.ExponentialCone())
+[x[1], x[2], x[3]] ∈ MathOptInterface.ExponentialCone()</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Similar to how JuMP defines the <code>&lt;=</code> and <code>&gt;=</code> syntax as a convenience way to specify <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan</code></a> and <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a> constraints, the remaining sections in this page describe functions and syntax that have been added for the convenience of common modeling situations.</p></div></div><h2 id="Set-inequality-syntax"><a class="docs-heading-anchor" href="#Set-inequality-syntax">Set inequality syntax</a><a id="Set-inequality-syntax-1"></a><a class="docs-heading-anchor-permalink" href="#Set-inequality-syntax" title="Permalink"></a></h2><p>For modeling convenience, the syntax <code>@constraint(model, x &gt;= y, Set())</code> is short-hand for <code>@constraint(model, x - y in Set())</code>. Therefore, the following calls are equivalent:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; y = [0.5, 0.75];
+
+julia&gt; @constraint(model, x &gt;= y, MOI.Nonnegatives(2))
+[x[1] - 0.5, x[2] - 0.75] ∈ MathOptInterface.Nonnegatives(2)
+
+julia&gt; @constraint(model, y &lt;= x, MOI.Nonnegatives(2))
+[x[1] - 0.5, x[2] - 0.75] ∈ MathOptInterface.Nonnegatives(2)
+
+julia&gt; @constraint(model, x - y in MOI.Nonnegatives(2))
+[x[1] - 0.5, x[2] - 0.75] ∈ MathOptInterface.Nonnegatives(2)</code></pre><p>Non-zero constants are not supported in this syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, x &gt;= 1, MOI.Nonnegatives(2))
+ERROR: Operation `sub_mul` between `Vector{VariableRef}` and `Int64` is not allowed. This most often happens when you write a constraint like `x &gt;= y` where `x` is an array and `y` is a constant. Use the broadcast syntax `x .- y &gt;= 0` instead.
+Stacktrace:
+[...]</code></pre><p>Use instead:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, x .- 1 &gt;= 0, MOI.Nonnegatives(2))
+[x[1] - 1, x[2] - 1] ∈ MathOptInterface.Nonnegatives(2)</code></pre><h2 id="Second-order-cone-constraints"><a class="docs-heading-anchor" href="#Second-order-cone-constraints">Second-order cone constraints</a><a id="Second-order-cone-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Second-order-cone-constraints" title="Permalink"></a></h2><p>A <a href="../../api/JuMP/#JuMP.SecondOrderCone"><code>SecondOrderCone</code></a> constrains the variables <code>t</code> and <code>x</code> to the set:</p><p class="math-container">\[||x||_2 \le t,\]</p><p>and <span>$t \ge 0$</span>. It can be added as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, t)
+t
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; @constraint(model, [t; x] in SecondOrderCone())
+[t, x[1], x[2]] ∈ MathOptInterface.SecondOrderCone(3)</code></pre><h2 id="Rotated-second-order-cone-constraints"><a class="docs-heading-anchor" href="#Rotated-second-order-cone-constraints">Rotated second-order cone constraints</a><a id="Rotated-second-order-cone-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Rotated-second-order-cone-constraints" title="Permalink"></a></h2><p>A <a href="../../api/JuMP/#JuMP.RotatedSecondOrderCone"><code>RotatedSecondOrderCone</code></a> constrains the variables <code>t</code>, <code>u</code>, and <code>x</code> to the set:</p><p class="math-container">\[||x||_2^2 \le 2 t \cdot u\]</p><p>and <span>$t, u \ge 0$</span>. It can be added as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, t)
+t
+
+julia&gt; @variable(model, u)
+u
+
+julia&gt; @variable(model, x[1:2])
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; @constraint(model, [t; u; x] in RotatedSecondOrderCone())
+[t, u, x[1], x[2]] ∈ MathOptInterface.RotatedSecondOrderCone(4)</code></pre><h2 id="Semi-integer-and-semi-continuous-variables"><a class="docs-heading-anchor" href="#Semi-integer-and-semi-continuous-variables">Semi-integer and semi-continuous variables</a><a id="Semi-integer-and-semi-continuous-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Semi-integer-and-semi-continuous-variables" title="Permalink"></a></h2><p>Semi-continuous variables are constrained to the set <span>$x \in \{0\} \cup [l, u]$</span>.</p><p>Create a semi-continuous variable using the <a href="../../api/JuMP/#JuMP.Semicontinuous"><code>Semicontinuous</code></a> set:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, x in Semicontinuous(1.5, 3.5))
+x in MathOptInterface.Semicontinuous{Float64}(1.5, 3.5)</code></pre><p>Semi-integer variables  are constrained to the set <span>$x \in \{0\} \cup \{l, l+1, \dots, u\}$</span>.</p><p>Create a semi-integer variable using the <a href="../../api/JuMP/#JuMP.Semiinteger"><code>Semiinteger</code></a> set:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @constraint(model, x in Semiinteger(1.0, 3.0))
+x in MathOptInterface.Semiinteger{Float64}(1.0, 3.0)</code></pre><h2 id="Special-Ordered-Sets-of-Type-1"><a class="docs-heading-anchor" href="#Special-Ordered-Sets-of-Type-1">Special Ordered Sets of Type 1</a><a id="Special-Ordered-Sets-of-Type-1-1"></a><a class="docs-heading-anchor-permalink" href="#Special-Ordered-Sets-of-Type-1" title="Permalink"></a></h2><p>In a Special Ordered Set of Type 1 (often denoted SOS-I or SOS1), at most one element can take a non-zero value.</p><p>Construct SOS-I constraints using the <a href="../../api/JuMP/#JuMP.SOS1"><code>SOS1</code></a> set:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3])
+3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+
+julia&gt; @constraint(model, x in SOS1())
+[x[1], x[2], x[3]] in MathOptInterface.SOS1{Float64}([1.0, 2.0, 3.0])</code></pre><p>Although not required for feasibility, solvers can benefit from an ordering of the variables (for example, the variables represent different factories to build, at most one factory can be built, and the factories can be ordered according to cost). To induce an ordering, a vector of weights can be provided, and the variables are ordered according to their corresponding weight.</p><p>For example, in the constraint:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, x in SOS1([3.1, 1.2, 2.3]))
+[x[1], x[2], x[3]] in MathOptInterface.SOS1{Float64}([3.1, 1.2, 2.3])</code></pre><p>the variables <code>x</code> have precedence <code>x[2]</code>, <code>x[3]</code>, <code>x[1]</code>.</p><h2 id="Special-Ordered-Sets-of-Type-2"><a class="docs-heading-anchor" href="#Special-Ordered-Sets-of-Type-2">Special Ordered Sets of Type 2</a><a id="Special-Ordered-Sets-of-Type-2-1"></a><a class="docs-heading-anchor-permalink" href="#Special-Ordered-Sets-of-Type-2" title="Permalink"></a></h2><p>In a Special Ordered Set of Type 2 (SOS-II), at most two elements can be non-zero, and if there are two non-zeros, they must be consecutive according to the ordering induced by a weight vector.</p><p>Construct SOS-II constraints using the <a href="../../api/JuMP/#JuMP.SOS2"><code>SOS2</code></a> set:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, x in SOS2([3.0, 1.0, 2.0]))
+[x[1], x[2], x[3]] in MathOptInterface.SOS2{Float64}([3.0, 1.0, 2.0])</code></pre><p>The possible non-zero pairs are (<code>x[1]</code>, <code>x[3]</code>) and (<code>x[2]</code>, <code>x[3]</code>):</p><p>If the weight vector is omitted, JuMP induces an ordering from <code>1:length(x)</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, x in SOS2())
+[x[1], x[2], x[3]] in MathOptInterface.SOS2{Float64}([1.0, 2.0, 3.0])</code></pre><h2 id="Indicator-constraints"><a class="docs-heading-anchor" href="#Indicator-constraints">Indicator constraints</a><a id="Indicator-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Indicator-constraints" title="Permalink"></a></h2><p>Indicator constraints consist of a binary variable and a linear constraint. The constraint holds when the binary variable takes the value <code>1</code>. The constraint may or may not hold when the binary variable takes the value <code>0</code>.</p><p>To enforce the constraint <code>x + y &lt;= 1</code> when the binary variable <code>a</code> is <code>1</code>, use:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; @variable(model, a, Bin)
+a
+
+julia&gt; @constraint(model, a =&gt; {x + y &lt;= 1})
+a =&gt; {x + y ≤ 1}</code></pre><p>If the constraint must hold when <code>a</code> is zero, add <code>!</code> or <code>¬</code> before the binary variable;</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, !a =&gt; {x + y &lt;= 1})
+!a =&gt; {x + y ≤ 1}</code></pre><h2 id="Semidefinite-constraints"><a class="docs-heading-anchor" href="#Semidefinite-constraints">Semidefinite constraints</a><a id="Semidefinite-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Semidefinite-constraints" title="Permalink"></a></h2><p>To constrain a matrix to be positive semidefinite (PSD), use <a href="../../api/JuMP/#JuMP.PSDCone"><code>PSDCone</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, X[1:2, 1:2])
+2×2 Matrix{VariableRef}:
+ X[1,1]  X[1,2]
+ X[2,1]  X[2,2]
+
+julia&gt; @constraint(model, X &gt;= 0, PSDCone())
+[X[1,1]  X[1,2];
+ X[2,1]  X[2,2]] ∈ PSDCone()</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Where possible, prefer constructing a matrix of <a href="../variables/#Semidefinite-variables">Semidefinite variables</a> using the <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro, rather than adding a constraint like <code>@constraint(model, X &gt;= 0, PSDCone())</code>. In some solvers, adding the constraint via <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> is less efficient, and can result in additional intermediate variables and constraints being added to the model.</p></div></div><p>The inequality <code>X &gt;= Y</code> between two square matrices <code>X</code> and <code>Y</code> is understood as constraining <code>X - Y</code> to be positive semidefinite.</p><pre><code class="language-julia-repl hljs">julia&gt; Y = [1 2; 2 1]
+2×2 Matrix{Int64}:
+ 1  2
+ 2  1
+
+julia&gt; @constraint(model, X &gt;= Y, PSDCone())
+[X[1,1] - 1  X[1,2] - 2;
+ X[2,1] - 2  X[2,2] - 1] ∈ PSDCone()</code></pre><h3 id="Symmetry"><a class="docs-heading-anchor" href="#Symmetry">Symmetry</a><a id="Symmetry-1"></a><a class="docs-heading-anchor-permalink" href="#Symmetry" title="Permalink"></a></h3><p>Solvers supporting PSD constraints usually expect to be given a matrix that is <em>symbolically</em> symmetric, that is, for which the expression in corresponding off-diagonal entries are the same. In our example, the expressions of entries <code>(1, 2)</code> and <code>(2, 1)</code> are respectively <code>X[1,2] - 2</code> and <code>X[2,1] - 2</code> which are different.</p><p>To bridge the gap between the constraint modeled and what the solver expects, solvers may add an equality constraint <code>X[1,2] - 2 == X[2,1] - 2</code> to force symmetry. Use <code>LinearAlgebra.Symmetric</code> to explicitly tell the solver that the matrix is symmetric:</p><pre><code class="language-julia-repl hljs">julia&gt; import LinearAlgebra
+
+julia&gt; Z = [X[1, 1] X[1, 2]; X[1, 2] X[2, 2]]
+2×2 Matrix{VariableRef}:
+ X[1,1]  X[1,2]
+ X[1,2]  X[2,2]
+
+julia&gt; @constraint(model, LinearAlgebra.Symmetric(Z) &gt;= 0, PSDCone())
+[X[1,1]  X[1,2];
+ X[1,2]  X[2,2]] ∈ PSDCone()</code></pre><p>Note that the lower triangular entries are ignored even if they are different so use it with caution:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, LinearAlgebra.Symmetric(X) &gt;= 0, PSDCone())
+[X[1,1]  X[1,2];
+ X[1,2]  X[2,2]] ∈ PSDCone()</code></pre><p>(Note the <code>(2, 1)</code> element of the constraint is <code>X[1,2]</code>, not <code>X[2,1]</code>.)</p><h2 id="Complementarity-constraints"><a class="docs-heading-anchor" href="#Complementarity-constraints">Complementarity constraints</a><a id="Complementarity-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Complementarity-constraints" title="Permalink"></a></h2><p>A mixed complementarity constraint <code>F(x) ⟂ x</code> consists of finding <code>x</code> in the interval <code>[lb, ub]</code>, such that the following holds:</p><ul><li><code>F(x) == 0</code> if <code>lb &lt; x &lt; ub</code></li><li><code>F(x) &gt;= 0</code> if <code>lb == x</code></li><li><code>F(x) &lt;= 0</code> if <code>x == ub</code></li></ul><p>JuMP supports mixed complementarity constraints via <code>complements(F(x), x)</code> or <code>F(x) ⟂ x</code> in the <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro. The interval set <code>[lb, ub]</code> is obtained from the variable bounds on <code>x</code>.</p><p>For example, to define the problem <code>2x - 1 ⟂ x</code> with <code>x ∈ [0, ∞)</code>, do:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 0)
+x
+
+julia&gt; @constraint(model, 2x - 1 ⟂ x)
+[2 x - 1, x] ∈ MathOptInterface.Complements(2)</code></pre><p>This problem has a unique solution at <code>x = 0.5</code>.</p><p>The perp operator <code>⟂</code> can be entered in most editors (and the Julia REPL) by typing <code>\perp&lt;tab&gt;</code>.</p><p>An alternative approach that does not require the <code>⟂</code> symbol uses the <code>complements</code> function as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; @constraint(model, complements(2x - 1, x))
+[2 x - 1, x] ∈ MathOptInterface.Complements(2)</code></pre><p>In both cases, the mapping <code>F(x)</code> is supplied as the first argument, and the matching variable <code>x</code> is supplied as the second.</p><p>Vector-valued complementarity constraints are also supported:</p><pre><code class="language-julia-repl hljs">julia&gt; @variable(model, -2 &lt;= y[1:2] &lt;= 2)
+2-element Vector{VariableRef}:
+ y[1]
+ y[2]
+
+julia&gt; M = [1 2; 3 4]
+2×2 Matrix{Int64}:
+ 1  2
+ 3  4
+
+julia&gt; q = [5, 6]
+2-element Vector{Int64}:
+ 5
+ 6
+
+julia&gt; @constraint(model, M * y + q ⟂ y)
+[y[1] + 2 y[2] + 5, 3 y[1] + 4 y[2] + 6, y[1], y[2]] ∈ MathOptInterface.Complements(4)</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../variables/">« Variables</a><a class="docs-footer-nextpage" href="../expressions/">Expressions »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/containers/index.html b/previews/PR3525/manual/containers/index.html
new file mode 100644
index 00000000000..d075e8fedf5
--- /dev/null
+++ b/previews/PR3525/manual/containers/index.html
@@ -0,0 +1,227 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Containers · JuMP</title><meta name="title" content="Containers · JuMP"/><meta property="og:title" content="Containers · JuMP"/><meta property="twitter:title" content="Containers · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li class="is-active"><a class="tocitem" href>Containers</a><ul class="internal"><li><a class="tocitem" href="#Array"><span>Array</span></a></li><li><a class="tocitem" href="#DenseAxisArray"><span>DenseAxisArray</span></a></li><li><a class="tocitem" href="#SparseAxisArray"><span>SparseAxisArray</span></a></li><li><a class="tocitem" href="#Forcing-the-container-type"><span>Forcing the container type</span></a></li><li><a class="tocitem" href="#How-different-container-types-are-chosen"><span>How different container types are chosen</span></a></li></ul></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Containers</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Containers</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/containers.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Containers"><a class="docs-heading-anchor" href="#Containers">Containers</a><a id="Containers-1"></a><a class="docs-heading-anchor-permalink" href="#Containers" title="Permalink"></a></h1><p>JuMP provides specialized containers similar to <a href="https://github.com/JuliaArrays/AxisArrays.jl"><code>AxisArrays</code></a> that enable multi-dimensional arrays with non-integer indices.</p><p>These containers are created automatically by JuMP&#39;s macros. Each macro has the same basic syntax:</p><pre><code class="language-julia hljs">@macroname(model, name[key1=index1, index2; optional_condition], other stuff)</code></pre><p>The containers are generated by the <code>name[key1=index1, index2; optional_condition]</code> syntax. Everything else is specific to the particular macro.</p><p>Containers can be named, for example, <code>name[key=index]</code>, or unnamed, for example, <code>[key=index]</code>. We call unnamed containers <em>anonymous</em>.</p><p>We call the bits inside the square brackets and before the <code>;</code> the <em>index sets</em>. The index sets can be named, for example, <code>[i = 1:4]</code>, or they can be unnamed, for example, <code>[1:4]</code>.</p><p>We call the bit inside the square brackets and after the <code>;</code> the <em>condition</em>. Conditions are optional.</p><p>In addition to the standard JuMP macros like <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> and <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>, which construct containers of variables and constraints respectively, you can use <a href="../../api/JuMP.Containers/#JuMP.Containers.@container"><code>Containers.@container</code></a> to construct containers with arbitrary elements.</p><p>We will use this macro to explain the three types of containers that are natively supported by JuMP: <code>Array</code>, <a href="../../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a>, and <a href="../../api/JuMP.Containers/#JuMP.Containers.SparseAxisArray"><code>Containers.SparseAxisArray</code></a>.</p><h2 id="Array"><a class="docs-heading-anchor" href="#Array">Array</a><a id="Array-1"></a><a class="docs-heading-anchor-permalink" href="#Array" title="Permalink"></a></h2><p>An <code>Array</code> is created when the index sets are rectangular and the index sets are of the form <code>1:n</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; Containers.@container(x[i = 1:2, j = 1:3], (i, j))
+2×3 Matrix{Tuple{Int64, Int64}}:
+ (1, 1)  (1, 2)  (1, 3)
+ (2, 1)  (2, 2)  (2, 3)</code></pre><p>The result is a normal Julia <code>Array</code>, so you can do all the usual things.</p><h3 id="Slicing"><a class="docs-heading-anchor" href="#Slicing">Slicing</a><a id="Slicing-1"></a><a class="docs-heading-anchor-permalink" href="#Slicing" title="Permalink"></a></h3><p>Arrays can be sliced</p><pre><code class="language-julia-repl hljs">julia&gt; x[:, 1]
+2-element Vector{Tuple{Int64, Int64}}:
+ (1, 1)
+ (2, 1)
+
+julia&gt; x[2, :]
+3-element Vector{Tuple{Int64, Int64}}:
+ (2, 1)
+ (2, 2)
+ (2, 3)</code></pre><h3 id="Looping"><a class="docs-heading-anchor" href="#Looping">Looping</a><a id="Looping-1"></a><a class="docs-heading-anchor-permalink" href="#Looping" title="Permalink"></a></h3><p>Use <code>eachindex</code> to loop over the elements:</p><pre><code class="language-julia-repl hljs">julia&gt; for key in eachindex(x)
+           println(x[key])
+       end
+(1, 1)
+(2, 1)
+(1, 2)
+(2, 2)
+(1, 3)
+(2, 3)</code></pre><h3 id="Get-the-index-sets"><a class="docs-heading-anchor" href="#Get-the-index-sets">Get the index sets</a><a id="Get-the-index-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Get-the-index-sets" title="Permalink"></a></h3><p>Use <code>axes</code> to obtain the index sets:</p><pre><code class="language-julia-repl hljs">julia&gt; axes(x)
+(Base.OneTo(2), Base.OneTo(3))</code></pre><h3 id="Broadcasting"><a class="docs-heading-anchor" href="#Broadcasting">Broadcasting</a><a id="Broadcasting-1"></a><a class="docs-heading-anchor-permalink" href="#Broadcasting" title="Permalink"></a></h3><p>Broadcasting over an Array returns an Array</p><pre><code class="language-julia-repl hljs">julia&gt; swap(x::Tuple) = (last(x), first(x))
+swap (generic function with 1 method)
+
+julia&gt; swap.(x)
+2×3 Matrix{Tuple{Int64, Int64}}:
+ (1, 1)  (2, 1)  (3, 1)
+ (1, 2)  (2, 2)  (3, 2)</code></pre><h3 id="Tables"><a class="docs-heading-anchor" href="#Tables">Tables</a><a id="Tables-1"></a><a class="docs-heading-anchor-permalink" href="#Tables" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> to convert the <code>Array</code> into a <a href="https://github.com/JuliaData/Tables.jl">Tables.jl</a> compatible <code>Vector{&lt;:NamedTuple}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; table = Containers.rowtable(x; header = [:I, :J, :value])
+6-element Vector{NamedTuple{(:I, :J, :value), Tuple{Int64, Int64, Tuple{Int64, Int64}}}}:
+ (I = 1, J = 1, value = (1, 1))
+ (I = 2, J = 1, value = (2, 1))
+ (I = 1, J = 2, value = (1, 2))
+ (I = 2, J = 2, value = (2, 2))
+ (I = 1, J = 3, value = (1, 3))
+ (I = 2, J = 3, value = (2, 3))</code></pre><p>Because it supports the <a href="https://github.com/JuliaData/Tables.jl">Tables.jl</a> interface, you can pass it to any function which accepts a table as input:</p><pre><code class="language-julia-repl hljs">julia&gt; import DataFrames;
+
+julia&gt; DataFrames.DataFrame(table)
+6×3 DataFrame
+ Row │ I      J      value
+     │ Int64  Int64  Tuple…
+─────┼──────────────────────
+   1 │     1      1  (1, 1)
+   2 │     2      1  (2, 1)
+   3 │     1      2  (1, 2)
+   4 │     2      2  (2, 2)
+   5 │     1      3  (1, 3)
+   6 │     2      3  (2, 3)</code></pre><h2 id="DenseAxisArray"><a class="docs-heading-anchor" href="#DenseAxisArray">DenseAxisArray</a><a id="DenseAxisArray-1"></a><a class="docs-heading-anchor-permalink" href="#DenseAxisArray" title="Permalink"></a></h2><p>A <a href="../../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> is created when the index sets are rectangular, but not of the form <code>1:n</code>. The index sets can be of any type.</p><pre><code class="language-julia-repl hljs">julia&gt; x = Containers.@container([i = 1:2, j = [:A, :B]], (i, j))
+2-dimensional DenseAxisArray{Tuple{Int64, Symbol},2,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+    Dimension 2, [:A, :B]
+And data, a 2×2 Matrix{Tuple{Int64, Symbol}}:
+ (1, :A)  (1, :B)
+ (2, :A)  (2, :B)</code></pre><h3 id="Slicing-2"><a class="docs-heading-anchor" href="#Slicing-2">Slicing</a><a class="docs-heading-anchor-permalink" href="#Slicing-2" title="Permalink"></a></h3><p>DenseAxisArrays can be sliced</p><pre><code class="language-julia-repl hljs">julia&gt; x[:, :A]
+1-dimensional DenseAxisArray{Tuple{Int64, Symbol},1,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+And data, a 2-element Vector{Tuple{Int64, Symbol}}:
+ (1, :A)
+ (2, :A)
+
+julia&gt; x[1, :]
+1-dimensional DenseAxisArray{Tuple{Int64, Symbol},1,...} with index sets:
+    Dimension 1, [:A, :B]
+And data, a 2-element Vector{Tuple{Int64, Symbol}}:
+ (1, :A)
+ (1, :B)</code></pre><h3 id="Looping-2"><a class="docs-heading-anchor" href="#Looping-2">Looping</a><a class="docs-heading-anchor-permalink" href="#Looping-2" title="Permalink"></a></h3><p>Use <code>eachindex</code> to loop over the elements:</p><pre><code class="language-julia-repl hljs">julia&gt; for key in eachindex(x)
+           println(x[key])
+       end
+(1, :A)
+(2, :A)
+(1, :B)
+(2, :B)</code></pre><h3 id="Get-the-index-sets-2"><a class="docs-heading-anchor" href="#Get-the-index-sets-2">Get the index sets</a><a class="docs-heading-anchor-permalink" href="#Get-the-index-sets-2" title="Permalink"></a></h3><p>Use <code>axes</code> to obtain the index sets:</p><pre><code class="language-julia-repl hljs">julia&gt; axes(x)
+(Base.OneTo(2), [:A, :B])</code></pre><h3 id="Broadcasting-2"><a class="docs-heading-anchor" href="#Broadcasting-2">Broadcasting</a><a class="docs-heading-anchor-permalink" href="#Broadcasting-2" title="Permalink"></a></h3><p>Broadcasting over a DenseAxisArray returns a DenseAxisArray</p><pre><code class="language-julia-repl hljs">julia&gt; swap(x::Tuple) = (last(x), first(x))
+swap (generic function with 1 method)
+
+julia&gt; swap.(x)
+2-dimensional DenseAxisArray{Tuple{Symbol, Int64},2,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+    Dimension 2, [:A, :B]
+And data, a 2×2 Matrix{Tuple{Symbol, Int64}}:
+ (:A, 1)  (:B, 1)
+ (:A, 2)  (:B, 2)</code></pre><h3 id="Access-internal-data"><a class="docs-heading-anchor" href="#Access-internal-data">Access internal data</a><a id="Access-internal-data-1"></a><a class="docs-heading-anchor-permalink" href="#Access-internal-data" title="Permalink"></a></h3><p>Use <code>Array(x)</code> to copy the internal data array into a new <code>Array</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; Array(x)
+2×2 Matrix{Tuple{Int64, Symbol}}:
+ (1, :A)  (1, :B)
+ (2, :A)  (2, :B)</code></pre><p>To access the internal data without a copy, use <code>x.data</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; x.data
+2×2 Matrix{Tuple{Int64, Symbol}}:
+ (1, :A)  (1, :B)
+ (2, :A)  (2, :B)</code></pre><h3 id="Tables-2"><a class="docs-heading-anchor" href="#Tables-2">Tables</a><a class="docs-heading-anchor-permalink" href="#Tables-2" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> to convert the <code>DenseAxisArray</code> into a <a href="https://github.com/JuliaData/Tables.jl">Tables.jl</a> compatible <code>Vector{&lt;:NamedTuple}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; table = Containers.rowtable(x; header = [:I, :J, :value])
+4-element Vector{NamedTuple{(:I, :J, :value), Tuple{Int64, Symbol, Tuple{Int64, Symbol}}}}:
+ (I = 1, J = :A, value = (1, :A))
+ (I = 2, J = :A, value = (2, :A))
+ (I = 1, J = :B, value = (1, :B))
+ (I = 2, J = :B, value = (2, :B))</code></pre><p>Because it supports the <a href="https://github.com/JuliaData/Tables.jl">Tables.jl</a> interface, you can pass it to any function which accepts a table as input:</p><pre><code class="language-julia-repl hljs">julia&gt; import DataFrames;
+
+julia&gt; DataFrames.DataFrame(table)
+4×3 DataFrame
+ Row │ I      J       value
+     │ Int64  Symbol  Tuple…
+─────┼────────────────────────
+   1 │     1  A       (1, :A)
+   2 │     2  A       (2, :A)
+   3 │     1  B       (1, :B)
+   4 │     2  B       (2, :B)</code></pre><h3 id="Keyword-indexing"><a class="docs-heading-anchor" href="#Keyword-indexing">Keyword indexing</a><a id="Keyword-indexing-1"></a><a class="docs-heading-anchor-permalink" href="#Keyword-indexing" title="Permalink"></a></h3><p>If all axes are named, you can use keyword indexing:</p><pre><code class="language-julia-repl hljs">julia&gt; x[i = 2, j = :A]
+(2, :A)
+
+julia&gt; x[i = :, j = :B]
+1-dimensional DenseAxisArray{Tuple{Int64, Symbol},1,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+And data, a 2-element Vector{Tuple{Int64, Symbol}}:
+ (1, :B)
+ (2, :B)</code></pre><h2 id="SparseAxisArray"><a class="docs-heading-anchor" href="#SparseAxisArray">SparseAxisArray</a><a id="SparseAxisArray-1"></a><a class="docs-heading-anchor-permalink" href="#SparseAxisArray" title="Permalink"></a></h2><p>A <a href="../../api/JuMP.Containers/#JuMP.Containers.SparseAxisArray"><code>Containers.SparseAxisArray</code></a> is created when the index sets are non-rectangular. This occurs in two circumstances:</p><p>An index depends on a prior index:</p><pre><code class="language-julia-repl hljs">julia&gt; Containers.@container([i = 1:2, j = i:2], (i, j))
+JuMP.Containers.SparseAxisArray{Tuple{Int64, Int64}, 2, Tuple{Int64, Int64}} with 3 entries:
+  [1, 1]  =  (1, 1)
+  [1, 2]  =  (1, 2)
+  [2, 2]  =  (2, 2)</code></pre><p>The <code>[indices; condition]</code> syntax is used:</p><pre><code class="language-julia-repl hljs">julia&gt; x = Containers.@container([i = 1:3, j = [:A, :B]; i &gt; 1], (i, j))
+JuMP.Containers.SparseAxisArray{Tuple{Int64, Symbol}, 2, Tuple{Int64, Symbol}} with 4 entries:
+  [2, A]  =  (2, :A)
+  [2, B]  =  (2, :B)
+  [3, A]  =  (3, :A)
+  [3, B]  =  (3, :B)</code></pre><p>Here we have the index sets <code>i = 1:3, j = [:A, :B]</code>, followed by <code>;</code>, and then a condition, which evaluates to <code>true</code> or <code>false</code>: <code>i &gt; 1</code>.</p><h3 id="Slicing-3"><a class="docs-heading-anchor" href="#Slicing-3">Slicing</a><a class="docs-heading-anchor-permalink" href="#Slicing-3" title="Permalink"></a></h3><p>Slicing is supported:</p><pre><code class="language-julia-repl hljs">julia&gt; y = x[:, :B]
+JuMP.Containers.SparseAxisArray{Tuple{Int64, Symbol}, 1, Tuple{Int64}} with 2 entries:
+  [2]  =  (2, :B)
+  [3]  =  (3, :B)</code></pre><h3 id="Looping-3"><a class="docs-heading-anchor" href="#Looping-3">Looping</a><a class="docs-heading-anchor-permalink" href="#Looping-3" title="Permalink"></a></h3><p>Use <code>eachindex</code> to loop over the elements:</p><pre><code class="language-julia-repl hljs">julia&gt; for key in eachindex(y)
+           println(y[key])
+       end
+(2, :B)
+(3, :B)</code></pre><h3 id="Broadcasting-3"><a class="docs-heading-anchor" href="#Broadcasting-3">Broadcasting</a><a class="docs-heading-anchor-permalink" href="#Broadcasting-3" title="Permalink"></a></h3><p>Broadcasting over a SparseAxisArray returns a SparseAxisArray</p><pre><code class="language-julia-repl hljs">julia&gt; swap(x::Tuple) = (last(x), first(x))
+swap (generic function with 1 method)
+
+julia&gt; swap.(y)
+JuMP.Containers.SparseAxisArray{Tuple{Symbol, Int64}, 1, Tuple{Int64}} with 2 entries:
+  [2]  =  (:B, 2)
+  [3]  =  (:B, 3)</code></pre><h3 id="Tables-3"><a class="docs-heading-anchor" href="#Tables-3">Tables</a><a class="docs-heading-anchor-permalink" href="#Tables-3" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> to convert the <code>SparseAxisArray</code> into a <a href="https://github.com/JuliaData/Tables.jl">Tables.jl</a> compatible <code>Vector{&lt;:NamedTuple}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; table = Containers.rowtable(x; header = [:I, :J, :value])
+4-element Vector{NamedTuple{(:I, :J, :value), Tuple{Int64, Symbol, Tuple{Int64, Symbol}}}}:
+ (I = 2, J = :A, value = (2, :A))
+ (I = 2, J = :B, value = (2, :B))
+ (I = 3, J = :A, value = (3, :A))
+ (I = 3, J = :B, value = (3, :B))</code></pre><p>Because it supports the <a href="https://github.com/JuliaData/Tables.jl">Tables.jl</a> interface, you can pass it to any function which accepts a table as input:</p><pre><code class="language-julia-repl hljs">julia&gt; import DataFrames;
+
+julia&gt; DataFrames.DataFrame(table)
+4×3 DataFrame
+ Row │ I      J       value
+     │ Int64  Symbol  Tuple…
+─────┼────────────────────────
+   1 │     2  A       (2, :A)
+   2 │     2  B       (2, :B)
+   3 │     3  A       (3, :A)
+   4 │     3  B       (3, :B)</code></pre><h3 id="Keyword-indexing-2"><a class="docs-heading-anchor" href="#Keyword-indexing-2">Keyword indexing</a><a class="docs-heading-anchor-permalink" href="#Keyword-indexing-2" title="Permalink"></a></h3><p>If all axes are named, you can use keyword indexing:</p><pre><code class="language-julia-repl hljs">julia&gt; x[i = 2, j = :A]
+(2, :A)
+
+julia&gt; x[i = :, j = :B]
+JuMP.Containers.SparseAxisArray{Tuple{Int64, Symbol}, 1, Tuple{Int64}} with 2 entries:
+  [2]  =  (2, :B)
+  [3]  =  (3, :B)</code></pre><h2 id="Forcing-the-container-type"><a class="docs-heading-anchor" href="#Forcing-the-container-type">Forcing the container type</a><a id="Forcing-the-container-type-1"></a><a class="docs-heading-anchor-permalink" href="#Forcing-the-container-type" title="Permalink"></a></h2><p>Pass <code>container = T</code> to use <code>T</code> as the container. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; Containers.@container([i = 1:2, j = 1:2], i + j, container = Array)
+2×2 Matrix{Int64}:
+ 2  3
+ 3  4
+
+julia&gt; Containers.@container([i = 1:2, j = 1:2], i + j, container = Dict)
+Dict{Tuple{Int64, Int64}, Int64} with 4 entries:
+  (1, 2) =&gt; 3
+  (1, 1) =&gt; 2
+  (2, 2) =&gt; 4
+  (2, 1) =&gt; 3</code></pre><p>You can also pass <code>DenseAxisArray</code> or <code>SparseAxisArray</code>.</p><h2 id="How-different-container-types-are-chosen"><a class="docs-heading-anchor" href="#How-different-container-types-are-chosen">How different container types are chosen</a><a id="How-different-container-types-are-chosen-1"></a><a class="docs-heading-anchor-permalink" href="#How-different-container-types-are-chosen" title="Permalink"></a></h2><p>If the compiler can prove <em>at compile time</em> that the index sets are rectangular, and indexed by a compact set of integers that start at <code>1</code>, <a href="../../api/JuMP.Containers/#JuMP.Containers.@container"><code>Containers.@container</code></a> will return an array. This is the case if your index sets are visible to the macro as <code>1:n</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; Containers.@container([i=1:3, j=1:5], i + j)
+3×5 Matrix{Int64}:
+ 2  3  4  5  6
+ 3  4  5  6  7
+ 4  5  6  7  8</code></pre><p>or an instance of <code>Base.OneTo</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; set = Base.OneTo(3)
+Base.OneTo(3)
+
+julia&gt; Containers.@container([i=set, j=1:5], i + j)
+3×5 Matrix{Int64}:
+ 2  3  4  5  6
+ 3  4  5  6  7
+ 4  5  6  7  8</code></pre><p>If the compiler can prove that the index set is rectangular, but not necessarily of the form <code>1:n</code> at compile time, then a <a href="../../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> will be constructed instead:</p><pre><code class="language-julia-repl hljs">julia&gt; set = 1:3
+1:3
+
+julia&gt; Containers.@container([i=set, j=1:5], i + j)
+2-dimensional DenseAxisArray{Int64,2,...} with index sets:
+    Dimension 1, 1:3
+    Dimension 2, Base.OneTo(5)
+And data, a 3×5 Matrix{Int64}:
+ 2  3  4  5  6
+ 3  4  5  6  7
+ 4  5  6  7  8</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>What happened here? Although we know that <code>set</code> contains <code>1:3</code>, at compile time the <code>typeof(set)</code> is a <code>UnitRange{Int}</code>. Therefore, Julia can&#39;t prove that the range starts at <code>1</code> (it only finds this out at runtime), and it defaults to a  <code>DenseAxisArray</code>. The case where we explicitly wrote <code>i = 1:3</code> worked because the macro can &quot;see&quot; the <code>1</code> at compile time.</p></div></div><p>However, if you know that the indices do form an <code>Array</code>, you can force the container type with <code>container = Array</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; set = 1:3
+1:3
+
+julia&gt; Containers.@container([i=set, j=1:5], i + j, container = Array)
+3×5 Matrix{Int64}:
+ 2  3  4  5  6
+ 3  4  5  6  7
+ 4  5  6  7  8</code></pre><p>Here&#39;s another example with something similar:</p><pre><code class="language-julia-repl hljs">julia&gt; a = 1
+1
+
+julia&gt; Containers.@container([i=a:3, j=1:5], i + j)
+2-dimensional DenseAxisArray{Int64,2,...} with index sets:
+    Dimension 1, 1:3
+    Dimension 2, Base.OneTo(5)
+And data, a 3×5 Matrix{Int64}:
+ 2  3  4  5  6
+ 3  4  5  6  7
+ 4  5  6  7  8
+
+julia&gt; Containers.@container([i=1:a, j=1:5], i + j)
+1×5 Matrix{Int64}:
+ 2  3  4  5  6</code></pre><p>Finally, if the compiler cannot prove that the index set is rectangular, a <a href="../../api/JuMP.Containers/#JuMP.Containers.SparseAxisArray"><code>Containers.SparseAxisArray</code></a> will be created.</p><p>This occurs when some indices depend on a previous one:</p><pre><code class="language-julia-repl hljs">julia&gt; Containers.@container([i=1:3, j=1:i], i + j)
+JuMP.Containers.SparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 6 entries:
+  [1, 1]  =  2
+  [2, 1]  =  3
+  [2, 2]  =  4
+  [3, 1]  =  4
+  [3, 2]  =  5
+  [3, 3]  =  6</code></pre><p>or if there is a condition on the index sets:</p><pre><code class="language-julia-repl hljs">julia&gt; Containers.@container([i = 1:5; isodd(i)], i^2)
+JuMP.Containers.SparseAxisArray{Int64, 1, Tuple{Int64}} with 3 entries:
+  [1]  =  1
+  [3]  =  9
+  [5]  =  25</code></pre><p>The condition can depend on multiple indices, the only requirement is that it is an expression that returns <code>true</code> or <code>false</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; condition(i, j) = isodd(i) &amp;&amp; iseven(j)
+condition (generic function with 1 method)
+
+julia&gt; Containers.@container([i = 1:2, j = 1:4; condition(i, j)], i + j)
+JuMP.Containers.SparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 2 entries:
+  [1, 2]  =  3
+  [1, 4]  =  5</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../objective/">« Objectives</a><a class="docs-footer-nextpage" href="../solutions/">Solutions »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/expressions/index.html b/previews/PR3525/manual/expressions/index.html
new file mode 100644
index 00000000000..4d01816366c
--- /dev/null
+++ b/previews/PR3525/manual/expressions/index.html
@@ -0,0 +1,230 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Expressions · JuMP</title><meta name="title" content="Expressions · JuMP"/><meta property="og:title" content="Expressions · JuMP"/><meta property="twitter:title" content="Expressions · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li class="is-active"><a class="tocitem" href>Expressions</a><ul class="internal"><li><a class="tocitem" href="#Affine-expressions"><span>Affine expressions</span></a></li><li><a class="tocitem" href="#Quadratic-expressions"><span>Quadratic expressions</span></a></li><li><a class="tocitem" href="#Nonlinear-expressions"><span>Nonlinear expressions</span></a></li><li><a class="tocitem" href="#Initializing-arrays"><span>Initializing arrays</span></a></li></ul></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Expressions</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Expressions</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/expressions.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Expressions"><a class="docs-heading-anchor" href="#Expressions">Expressions</a><a id="Expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Expressions" title="Permalink"></a></h1><p>JuMP has three types of expressions: affine, quadratic, and nonlinear. These expressions can be inserted into constraints or into the objective. This is particularly useful if an expression is used in multiple places in the model.</p><h2 id="Affine-expressions"><a class="docs-heading-anchor" href="#Affine-expressions">Affine expressions</a><a id="Affine-expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Affine-expressions" title="Permalink"></a></h2><p>There are four ways of constructing an affine expression in JuMP: with the <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macro, with operator overloading, with the <code>AffExpr</code> constructor, and with <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a>.</p><h3 id="Macros"><a class="docs-heading-anchor" href="#Macros">Macros</a><a id="Macros-1"></a><a class="docs-heading-anchor-permalink" href="#Macros" title="Permalink"></a></h3><p>The recommended way to create an affine expression is via the <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macro.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; ex = @expression(model, 2x + y - 1)
+2 x + y - 1</code></pre><p>This expression can be used in the objective or added to a constraint. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; @objective(model, Min, 2 * ex - 1)
+4 x + 2 y - 3
+
+julia&gt; objective_function(model)
+4 x + 2 y - 3</code></pre><p>Just like variables and constraints, named expressions can also be created. For example</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i = 1:3]);
+
+julia&gt; @expression(model, expr[i = 1:3], i * sum(x[j] for j in i:3));
+
+julia&gt; expr
+3-element Vector{AffExpr}:
+ x[1] + x[2] + x[3]
+ 2 x[2] + 2 x[3]
+ 3 x[3]</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>You can read more about containers in the <a href="../containers/#Containers">Containers</a> section.</p></div></div><h3 id="Operator-overloading"><a class="docs-heading-anchor" href="#Operator-overloading">Operator overloading</a><a id="Operator-overloading-1"></a><a class="docs-heading-anchor-permalink" href="#Operator-overloading" title="Permalink"></a></h3><p>Expressions can also be created without macros. However, note that in some cases, this can be much slower that constructing an expression using macros.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; ex = 2x + y - 1
+2 x + y - 1</code></pre><h3 id="Constructors"><a class="docs-heading-anchor" href="#Constructors">Constructors</a><a id="Constructors-1"></a><a class="docs-heading-anchor-permalink" href="#Constructors" title="Permalink"></a></h3><p>A third way to create an affine expression is by the <code>AffExpr</code> constructor. The first argument is the constant term, and the remaining arguments are variable-coefficient pairs.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; ex = AffExpr(-1.0, x =&gt; 2.0, y =&gt; 1.0)
+2 x + y - 1</code></pre><h3 id="add_to_expression!"><a class="docs-heading-anchor" href="#add_to_expression!"><code>add_to_expression!</code></a><a id="add_to_expression!-1"></a><a class="docs-heading-anchor-permalink" href="#add_to_expression!" title="Permalink"></a></h3><p>The fourth way to create an affine expression is by using <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a>. Compared to the operator overloading method, this approach is faster because it avoids constructing temporary objects. The <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macro uses <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a> behind-the-scenes.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; ex = AffExpr(-1.0)
+-1
+
+julia&gt; add_to_expression!(ex, 2.0, x)
+2 x - 1
+
+julia&gt; add_to_expression!(ex, 1.0, y)
+2 x + y - 1</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Read the section <a href="#Initializing-arrays">Initializing arrays</a> for some cases to be careful about when using <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a>.</p></div></div><h3 id="Removing-zero-terms"><a class="docs-heading-anchor" href="#Removing-zero-terms">Removing zero terms</a><a id="Removing-zero-terms-1"></a><a class="docs-heading-anchor-permalink" href="#Removing-zero-terms" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.drop_zeros!"><code>drop_zeros!</code></a> to remove terms from an affine expression with a <code>0</code> coefficient.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @expression(model, ex, x + 1 - x)
+0 x + 1
+
+julia&gt; drop_zeros!(ex)
+
+julia&gt; ex
+1</code></pre><h3 id="Coefficients"><a class="docs-heading-anchor" href="#Coefficients">Coefficients</a><a id="Coefficients-1"></a><a class="docs-heading-anchor-permalink" href="#Coefficients" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.coefficient"><code>coefficient</code></a> to return the coefficient associated with a variable in an affine expression.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; @expression(model, ex, 2x + 1)
+2 x + 1
+
+julia&gt; coefficient(ex, x)
+2.0
+
+julia&gt; coefficient(ex, y)
+0.0</code></pre><h2 id="Quadratic-expressions"><a class="docs-heading-anchor" href="#Quadratic-expressions">Quadratic expressions</a><a id="Quadratic-expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Quadratic-expressions" title="Permalink"></a></h2><p>Like affine expressions, there are four ways of constructing a quadratic expression in JuMP: macros, operator overloading, constructors, and <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a>.</p><h3 id="Macros-2"><a class="docs-heading-anchor" href="#Macros-2">Macros</a><a class="docs-heading-anchor-permalink" href="#Macros-2" title="Permalink"></a></h3><p>The <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macro can be used to create quadratic expressions by including quadratic terms.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; ex = @expression(model, x^2 + 2 * x * y + y^2 + x + y - 1)
+x² + 2 x*y + y² + x + y - 1</code></pre><h3 id="Operator-overloading-2"><a class="docs-heading-anchor" href="#Operator-overloading-2">Operator overloading</a><a class="docs-heading-anchor-permalink" href="#Operator-overloading-2" title="Permalink"></a></h3><p>Operator overloading can also be used to create quadratic expressions. The same performance warning (discussed in the affine expression section) applies.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; ex = x^2 + 2 * x * y + y^2 + x + y - 1
+x² + 2 x*y + y² + x + y - 1</code></pre><h3 id="Constructors-2"><a class="docs-heading-anchor" href="#Constructors-2">Constructors</a><a class="docs-heading-anchor-permalink" href="#Constructors-2" title="Permalink"></a></h3><p>Quadratic expressions can also be created using the <code>QuadExpr</code> constructor. The first argument is an affine expression, and the remaining arguments are pairs, where the first term is a <code>JuMP.UnorderedPair</code> and the second term is the coefficient.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; aff_expr = AffExpr(-1.0, x =&gt; 1.0, y =&gt; 1.0)
+x + y - 1
+
+julia&gt; quad_expr = QuadExpr(
+           aff_expr,
+           UnorderedPair(x, x) =&gt; 1.0,
+           UnorderedPair(x, y) =&gt; 2.0,
+           UnorderedPair(y, y) =&gt; 1.0,
+       )
+x² + 2 x*y + y² + x + y - 1</code></pre><h3 id="add_to_expression!-2"><a class="docs-heading-anchor" href="#add_to_expression!-2"><code>add_to_expression!</code></a><a class="docs-heading-anchor-permalink" href="#add_to_expression!-2" title="Permalink"></a></h3><p>Finally, <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a> can also be used to add quadratic terms.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; ex = QuadExpr(x + y - 1.0)
+x + y - 1
+
+julia&gt; add_to_expression!(ex, 1.0, x, x)
+x² + x + y - 1
+
+julia&gt; add_to_expression!(ex, 2.0, x, y)
+x² + 2 x*y + x + y - 1
+
+julia&gt; add_to_expression!(ex, 1.0, y, y)
+x² + 2 x*y + y² + x + y - 1</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Read the section <a href="#Initializing-arrays">Initializing arrays</a> for some cases to be careful about when using <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a>.</p></div></div><h3 id="Removing-zero-terms-2"><a class="docs-heading-anchor" href="#Removing-zero-terms-2">Removing zero terms</a><a class="docs-heading-anchor-permalink" href="#Removing-zero-terms-2" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.drop_zeros!"><code>drop_zeros!</code></a> to remove terms from a quadratic expression with a <code>0</code> coefficient.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @expression(model, ex, x^2 + x + 1 - x^2)
+0 x² + x + 1
+
+julia&gt; drop_zeros!(ex)
+
+julia&gt; ex
+x + 1</code></pre><h3 id="Coefficients-2"><a class="docs-heading-anchor" href="#Coefficients-2">Coefficients</a><a class="docs-heading-anchor-permalink" href="#Coefficients-2" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.coefficient"><code>coefficient</code></a> to return the coefficient associated with a pair of variables in a quadratic expression.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; @expression(model, ex, 2*x*y + 3*x)
+2 x*y + 3 x
+
+julia&gt; coefficient(ex, x, y)
+2.0
+
+julia&gt; coefficient(ex, x, x)
+0.0
+
+julia&gt; coefficient(ex, y, x)
+2.0
+
+julia&gt; coefficient(ex, x)
+3.0</code></pre><h2 id="Nonlinear-expressions"><a class="docs-heading-anchor" href="#Nonlinear-expressions">Nonlinear expressions</a><a id="Nonlinear-expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear-expressions" title="Permalink"></a></h2><p>Nonlinear expressions in JuMP are represented by a <a href="../../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a> object. See <a href="../nonlinear/#Nonlinear-expressions-in-detail">Nonlinear expressions in detail</a> for more details.</p><h2 id="Initializing-arrays"><a class="docs-heading-anchor" href="#Initializing-arrays">Initializing arrays</a><a id="Initializing-arrays-1"></a><a class="docs-heading-anchor-permalink" href="#Initializing-arrays" title="Permalink"></a></h2><p>JuMP implements <code>zero(AffExpr)</code> and <code>one(AffExpr)</code> to support various functions in <code>LinearAlgebra</code> (for example, accessing the off-diagonal of a <code>Diagonal</code> matrix).</p><pre><code class="language-julia-repl hljs">julia&gt; zero(AffExpr)
+0
+
+julia&gt; one(AffExpr)
+1</code></pre><p>However, this can result in a subtle bug if you call <a href="../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a> or the <a href="https://github.com/jump-dev/MutableArithmetics.jl">MutableArithmetics API</a> on an element created by <code>zeros</code> or <code>ones</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; x = zeros(AffExpr, 2)
+2-element Vector{AffExpr}:
+ 0
+ 0
+
+julia&gt; add_to_expression!(x[1], 1.1)
+1.1
+
+julia&gt; x
+2-element Vector{AffExpr}:
+ 1.1
+ 1.1</code></pre><p>Notice how we modified <code>x[1]</code>, but we also changed <code>x[2]</code>!</p><p>This happened because <code>zeros(AffExpr, 2)</code> calls <code>zero(AffExpr)</code> once to obtain a zero element, and then creates an appropriately sized array filled with the same element.</p><p>This also happens with broadcasting calls containing a conversion of <code>0</code> or <code>1</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; x = Vector{AffExpr}(undef, 2)
+2-element Vector{AffExpr}:
+ #undef
+ #undef
+
+julia&gt; x .= 0
+2-element Vector{AffExpr}:
+ 0
+ 0
+
+julia&gt; add_to_expression!(x[1], 1.1)
+1.1
+
+julia&gt; x
+2-element Vector{AffExpr}:
+ 1.1
+ 1.1</code></pre><p>The recommended way to create an array of empty expressions is as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; x = Vector{AffExpr}(undef, 2)
+2-element Vector{AffExpr}:
+ #undef
+ #undef
+
+julia&gt; for i in eachindex(x)
+           x[i] = AffExpr(0.0)
+       end
+
+julia&gt; add_to_expression!(x[1], 1.1)
+1.1
+
+julia&gt; x
+2-element Vector{AffExpr}:
+ 1.1
+ 0</code></pre><p>Alternatively, use non-mutating operation to avoid updating <code>x[1]</code> in-place:</p><pre><code class="language-julia-repl hljs">julia&gt; x = zeros(AffExpr, 2)
+2-element Vector{AffExpr}:
+ 0
+ 0
+
+julia&gt; x[1] += 1.1
+1.1
+
+julia&gt; x
+2-element Vector{AffExpr}:
+ 1.1
+ 0</code></pre><p>Note that for large expressions this will be slower due to the allocation of additional temporary objects.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../constraints/">« Constraints</a><a class="docs-footer-nextpage" href="../objective/">Objectives »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/models/index.html b/previews/PR3525/manual/models/index.html
new file mode 100644
index 00000000000..80599d48acc
--- /dev/null
+++ b/previews/PR3525/manual/models/index.html
@@ -0,0 +1,237 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Models · JuMP</title><meta name="title" content="Models · JuMP"/><meta property="og:title" content="Models · JuMP"/><meta property="twitter:title" content="Models · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Models</a><ul class="internal"><li><a class="tocitem" href="#Create-a-model"><span>Create a model</span></a></li><li><a class="tocitem" href="#solver_options"><span>Solver options</span></a></li><li><a class="tocitem" href="#Changing-the-number-types"><span>Changing the number types</span></a></li><li><a class="tocitem" href="#Print-the-model"><span>Print the model</span></a></li><li><a class="tocitem" href="#Turn-off-output"><span>Turn off output</span></a></li><li><a class="tocitem" href="#Set-a-time-limit"><span>Set a time limit</span></a></li><li><a class="tocitem" href="#Write-a-model-to-file"><span>Write a model to file</span></a></li><li><a class="tocitem" href="#Read-a-model-from-file"><span>Read a model from file</span></a></li><li><a class="tocitem" href="#Relax-integrality"><span>Relax integrality</span></a></li><li><a class="tocitem" href="#Backends"><span>Backends</span></a></li><li><a class="tocitem" href="#Direct-mode"><span>Direct mode</span></a></li></ul></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Models</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Models</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/models.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="jump_models"><a class="docs-heading-anchor" href="#jump_models">Models</a><a id="jump_models-1"></a><a class="docs-heading-anchor-permalink" href="#jump_models" title="Permalink"></a></h1><p>JuMP models are the fundamental building block that we use to construct optimization problems. They hold things like the variables and constraints, as well as which solver to use and even solution information.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>JuMP uses &quot;optimizer&quot; as a synonym for &quot;solver.&quot; Our convention is to use &quot;solver&quot; to refer to the underlying software, and use &quot;optimizer&quot; to refer to the Julia object that wraps the solver. For example, <code>HiGHS</code> is a solver, and <code>HiGHS.Optimizer</code> is an optimizer.</p></div></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>See <a href="../../installation/#Supported-solvers">Supported solvers</a> for a list of available solvers.</p></div></div><h2 id="Create-a-model"><a class="docs-heading-anchor" href="#Create-a-model">Create a model</a><a id="Create-a-model-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-model" title="Permalink"></a></h2><p>Create a model by passing an optimizer to <a href="../../api/JuMP/#JuMP.Model"><code>Model</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer)
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><p>If you don&#39;t know which optimizer you will be using at creation time, create a model without an optimizer, and then call <a href="../../api/JuMP/#JuMP.set_optimizer"><code>set_optimizer</code></a> at any time prior to <a href="../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model()
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+
+julia&gt; set_optimizer(model, HiGHS.Optimizer)</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Don&#39;t know what the fields <code>Model mode</code> and <code>CachingOptimizer state</code> mean? Read the <a href="#Backends">Backends</a> section.</p></div></div><h3 id="What-is-the-difference?"><a class="docs-heading-anchor" href="#What-is-the-difference?">What is the difference?</a><a id="What-is-the-difference?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-the-difference?" title="Permalink"></a></h3><p>For most models, there is no difference between passing the optimizer to <a href="../../api/JuMP/#JuMP.Model"><code>Model</code></a>, and calling <a href="../../api/JuMP/#JuMP.set_optimizer"><code>set_optimizer</code></a>.</p><p>However, if an optimizer does not support a constraint in the model, the timing of when an error will be thrown can differ:</p><ul><li>If you pass an optimizer, an error will be thrown when you try to add the constraint.</li><li>If you call <a href="../../api/JuMP/#JuMP.set_optimizer"><code>set_optimizer</code></a>, an error will be thrown when you try to solve the model via <a href="../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a>.</li></ul><p>Therefore, most users should pass an optimizer to <a href="../../api/JuMP/#JuMP.Model"><code>Model</code></a> because it provides the earliest warning that your solver is not suitable for the model you are trying to build. However, if you are modifying a problem by adding and deleting different constraint types, you may need to use <a href="../../api/JuMP/#JuMP.set_optimizer"><code>set_optimizer</code></a>. See <a href="#Switching-optimizer-for-the-relaxed-problem">Switching optimizer for the relaxed problem</a> for an example of when this is useful.</p><h3 id="Reducing-time-to-first-solve-latency"><a class="docs-heading-anchor" href="#Reducing-time-to-first-solve-latency">Reducing time-to-first-solve latency</a><a id="Reducing-time-to-first-solve-latency-1"></a><a class="docs-heading-anchor-permalink" href="#Reducing-time-to-first-solve-latency" title="Permalink"></a></h3><p>By default, JuMP uses <a href="#LazyBridgeOptimizer">bridges</a> to reformulate the model you are building into an equivalent model supported by the solver.</p><p>However, if your model is already supported by the solver, bridges add latency (read <a href="../../tutorials/getting_started/performance_tips/#The-&quot;time-to-first-solve&quot;-issue">The &quot;time-to-first-solve&quot; issue</a>). This is particularly noticeable for small models.</p><p>To reduce the &quot;time-to-first-solve,s&quot; try passing <code>add_bridges = false</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer; add_bridges = false);</code></pre><p>or</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; set_optimizer(model, HiGHS.Optimizer; add_bridges = false)</code></pre><p>However, be wary. If your model and solver combination needs bridges, an error will be thrown:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(SCS.Optimizer; add_bridges = false);
+
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @constraint(model, 2x &lt;= 1)
+ERROR: Constraints of type MathOptInterface.ScalarAffineFunction{Float64}-in-MathOptInterface.LessThan{Float64} are not supported by the solver.
+
+If you expected the solver to support your problem, you may have an error in your formulation. Otherwise, consider using a different solver.
+
+The list of available solvers, along with the problem types they support, is available at https://jump.dev/JuMP.jl/stable/installation/#Supported-solvers.
+[...]</code></pre><h3 id="Solvers-which-expect-environments"><a class="docs-heading-anchor" href="#Solvers-which-expect-environments">Solvers which expect environments</a><a id="Solvers-which-expect-environments-1"></a><a class="docs-heading-anchor-permalink" href="#Solvers-which-expect-environments" title="Permalink"></a></h3><p>Some solvers accept (or require) positional arguments such as a license environment or a path to a binary executable. For these solvers, you can pass a function to <a href="../../api/JuMP/#JuMP.Model"><code>Model</code></a> which takes zero arguments and returns an instance of the optimizer.</p><p>A common use-case for this is passing an environment or sub-solver to the optimizer:</p><pre><code class="language-julia-repl hljs">julia&gt; import HiGHS
+
+julia&gt; import MultiObjectiveAlgorithms as MOA
+
+julia&gt; model = Model(() -&gt; MOA.Optimizer(HiGHS.Optimizer))
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: MOA[algorithm=MultiObjectiveAlgorithms.Lexicographic, optimizer=HiGHS]</code></pre><h2 id="solver_options"><a class="docs-heading-anchor" href="#solver_options">Solver options</a><a id="solver_options-1"></a><a class="docs-heading-anchor-permalink" href="#solver_options" title="Permalink"></a></h2><p>JuMP uses &quot;attribute&quot; as a synonym for &quot;option.&quot; Use <a href="../../api/JuMP/#JuMP.optimizer_with_attributes"><code>optimizer_with_attributes</code></a> to create an optimizer with some attributes initialized:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(
+           optimizer_with_attributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; false),
+       )
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><p>Alternatively, use <a href="../../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a> to set an attribute after the model has been created:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_attribute(model, &quot;output_flag&quot;, false)
+
+julia&gt; get_attribute(model, &quot;output_flag&quot;)
+false</code></pre><p>You can also modify attributes within an <a href="../../api/JuMP/#JuMP.optimizer_with_attributes"><code>optimizer_with_attributes</code></a> object:</p><pre><code class="language-julia-repl hljs">julia&gt; solver = optimizer_with_attributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; true);
+
+julia&gt; get_attribute(solver, &quot;output_flag&quot;)
+true
+
+julia&gt; set_attribute(solver, &quot;output_flag&quot;, false)
+
+julia&gt; get_attribute(solver, &quot;output_flag&quot;)
+false
+
+julia&gt; model = Model(solver);</code></pre><h2 id="Changing-the-number-types"><a class="docs-heading-anchor" href="#Changing-the-number-types">Changing the number types</a><a id="Changing-the-number-types-1"></a><a class="docs-heading-anchor-permalink" href="#Changing-the-number-types" title="Permalink"></a></h2><p>By default, the coefficients of affine and quadratic expressions are numbers of type either <code>Float64</code> or <code>Complex{Float64}</code> (see <a href="../complex/#Complex-number-support">Complex number support</a>).</p><p>The type <code>Float64</code> can be changed using the <a href="../../api/JuMP/#JuMP.GenericModel"><code>GenericModel</code></a> constructor:</p><pre><code class="language-julia-repl hljs">julia&gt; model = GenericModel{Rational{BigInt}}();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @expression(model, expr, 1 // 3 * x)
+1//3 x
+
+julia&gt; typeof(expr)
+GenericAffExpr{Rational{BigInt}, GenericVariableRef{Rational{BigInt}}}</code></pre><p>Using a <code>value_type</code> other than <code>Float64</code> is an advanced operation and should be used only if the underlying solver actually solves the problem using the provided value type.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><a href="../nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a> is currently restricted to the <code>Float64</code> number type.</p></div></div><h2 id="Print-the-model"><a class="docs-heading-anchor" href="#Print-the-model">Print the model</a><a id="Print-the-model-1"></a><a class="docs-heading-anchor-permalink" href="#Print-the-model" title="Permalink"></a></h2><p>By default, <code>show(model)</code> will print a summary of the problem:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(); @variable(model, x &gt;= 0); @objective(model, Max, x);
+
+julia&gt; model
+A JuMP Model
+Maximization problem with:
+Variable: 1
+Objective function type: VariableRef
+`VariableRef`-in-`MathOptInterface.GreaterThan{Float64}`: 1 constraint
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+Names registered in the model: x</code></pre><p>Use <code>print</code> to print the formulation of the model (in IJulia, this will render as LaTeX.</p><pre><code class="language-julia-repl hljs">julia&gt; print(model)
+Max x
+Subject to
+ x ≥ 0</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This format is specific to JuMP and may change in any future release. It is not intended to be an instance format. To write the model to a file, use <a href="../../api/JuMP/#JuMP.write_to_file"><code>write_to_file</code></a> instead.</p></div></div><p>Use <a href="../../api/JuMP/#JuMP.latex_formulation"><code>latex_formulation</code></a> to display the model in LaTeX form.</p><pre><code class="language-julia-repl hljs">julia&gt; latex_formulation(model)
+$$ \begin{aligned}
+\max\quad &amp; x\\
+\text{Subject to} \quad &amp; x \geq 0\\
+\end{aligned} $$</code></pre><p>In IJulia (and Documenter), ending a cell in with <a href="../../api/JuMP/#JuMP.latex_formulation"><code>latex_formulation</code></a> will render the model in LaTeX:</p><pre><code class="language-julia hljs">latex_formulation(model)</code></pre><p class="math-container">\[ \begin{aligned}
+\max\quad &amp; x\\
+\text{Subject to} \quad &amp; x \geq 0\\
+\end{aligned} \]</p><h2 id="Turn-off-output"><a class="docs-heading-anchor" href="#Turn-off-output">Turn off output</a><a id="Turn-off-output-1"></a><a class="docs-heading-anchor-permalink" href="#Turn-off-output" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.set_silent"><code>set_silent</code></a> and <a href="../../api/JuMP/#JuMP.unset_silent"><code>unset_silent</code></a> to disable or enable printing output from the solver.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; unset_silent(model)</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Most solvers will also have a <a href="#solver_options">solver-specific option</a> to provide finer-grained control over the output. Consult their README&#39;s for details.</p></div></div><h2 id="Set-a-time-limit"><a class="docs-heading-anchor" href="#Set-a-time-limit">Set a time limit</a><a id="Set-a-time-limit-1"></a><a class="docs-heading-anchor-permalink" href="#Set-a-time-limit" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.set_time_limit_sec"><code>set_time_limit_sec</code></a>, <a href="../../api/JuMP/#JuMP.unset_time_limit_sec"><code>unset_time_limit_sec</code></a>, and <a href="../../api/JuMP/#JuMP.time_limit_sec"><code>time_limit_sec</code></a> to manage time limits.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_time_limit_sec(model, 60.0)
+
+
+julia&gt; time_limit_sec(model)
+60.0
+
+julia&gt; unset_time_limit_sec(model)
+
+julia&gt; limit = time_limit_sec(model)
+
+julia&gt; limit === nothing
+true</code></pre><p>If your time limit is encoded as a <code>Dates.Period</code> object, use the following code to convert it to <code>Float64</code> for <a href="../../api/JuMP/#JuMP.set_time_limit_sec"><code>set_time_limit_sec</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; import Dates
+
+julia&gt; seconds(x::Dates.Period) = 1e-3 * Dates.value(round(x, Dates.Millisecond))
+seconds (generic function with 1 method)
+
+julia&gt; set_time_limit_sec(model, seconds(Dates.Hour(1)))
+
+julia&gt; time_limit_sec(model)
+3600.0</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Some solvers do not support time limits. In these cases, an error will be thrown.</p></div></div><h2 id="Write-a-model-to-file"><a class="docs-heading-anchor" href="#Write-a-model-to-file">Write a model to file</a><a id="Write-a-model-to-file-1"></a><a class="docs-heading-anchor-permalink" href="#Write-a-model-to-file" title="Permalink"></a></h2><p>JuMP can write models to a variety of file-formats using <a href="../../api/JuMP/#JuMP.write_to_file"><code>write_to_file</code></a> and <a href="../../api/JuMP/#Base.write-Tuple{IO, GenericModel}"><code>Base.write</code></a>.</p><p>For most common file formats, the file type will be detected from the extension.</p><p>For example, here is how to write an <a href="https://en.wikipedia.org/wiki/MPS_(format)">MPS file</a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; write_to_file(model, &quot;model.mps&quot;)</code></pre><p>Other supported file formats include:</p><ul><li><code>.cbf</code> for the <a href="https://docs.mosek.com/latest/capi/cbf-format.html">Conic Benchmark Format</a></li><li><code>.lp</code> for the <a href="https://docs.mosek.com/latest/capi/lp-format.html">LP file format</a></li><li><code>.mof.json</code> for the <a href="https://jump.dev/MathOptFormat/">MathOptFormat</a></li><li><code>.nl</code> for <a href="https://en.wikipedia.org/wiki/Nl_(format)">AMPL&#39;s NL file format</a></li><li><code>.rew</code> for the <a href="https://www.gurobi.com/documentation/9.5/refman/rew_format.html">REW file format</a></li><li><code>.sdpa</code> and &quot;.dat-s&quot; for the <a href="http://plato.asu.edu/ftp/sdpa_format.txt">SDPA file format</a></li></ul><p>To write to a specific <code>io::IO</code>, use <a href="../../api/JuMP/#Base.write-Tuple{IO, GenericModel}"><code>Base.write</code></a>. Specify the file type by passing a <a href="../../moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.FileFormat"><code>MOI.FileFormats.FileFormat</code></a> enum.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; io = IOBuffer();
+
+julia&gt; write(io, model; format = MOI.FileFormats.FORMAT_MPS)</code></pre><h2 id="Read-a-model-from-file"><a class="docs-heading-anchor" href="#Read-a-model-from-file">Read a model from file</a><a id="Read-a-model-from-file-1"></a><a class="docs-heading-anchor-permalink" href="#Read-a-model-from-file" title="Permalink"></a></h2><p>JuMP models can be created from file formats using <a href="../../api/JuMP/#JuMP.read_from_file"><code>read_from_file</code></a> and <code>Base.read</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = read_from_file(&quot;model.mps&quot;)
+A JuMP Model
+Minimization problem with:
+Variables: 0
+Objective function type: AffExpr
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+
+julia&gt; seekstart(io);
+
+julia&gt; model2 = read(io, Model; format = MOI.FileFormats.FORMAT_MPS)
+A JuMP Model
+Minimization problem with:
+Variables: 0
+Objective function type: AffExpr
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Because file formats do not serialize the containers of JuMP variables and constraints, the names in the model will <em>not</em> be registered. Therefore, you cannot access named variables and constraints via <code>model[:x]</code>. Instead, use <a href="../../api/JuMP/#JuMP.variable_by_name"><code>variable_by_name</code></a> or <a href="../../api/JuMP/#JuMP.constraint_by_name"><code>constraint_by_name</code></a> to access specific variables or constraints.</p></div></div><h2 id="Relax-integrality"><a class="docs-heading-anchor" href="#Relax-integrality">Relax integrality</a><a id="Relax-integrality-1"></a><a class="docs-heading-anchor-permalink" href="#Relax-integrality" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.relax_integrality"><code>relax_integrality</code></a> to remove any integrality constraints from the model, such as integer and binary restrictions on variables. <a href="../../api/JuMP/#JuMP.relax_integrality"><code>relax_integrality</code></a> returns a function that can be later called with zero arguments to re-add the removed constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Int)
+x
+
+julia&gt; num_constraints(model, VariableRef, MOI.Integer)
+1
+
+julia&gt; undo = relax_integrality(model);
+
+julia&gt; num_constraints(model, VariableRef, MOI.Integer)
+0
+
+julia&gt; undo()
+
+julia&gt; num_constraints(model, VariableRef, MOI.Integer)
+1</code></pre><h3 id="Switching-optimizer-for-the-relaxed-problem"><a class="docs-heading-anchor" href="#Switching-optimizer-for-the-relaxed-problem">Switching optimizer for the relaxed problem</a><a id="Switching-optimizer-for-the-relaxed-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Switching-optimizer-for-the-relaxed-problem" title="Permalink"></a></h3><p>A common reason for relaxing integrality is to compute dual variables of the relaxed problem. However, some mixed-integer linear solvers (for example, Cbc) do not return dual solutions, even if the problem does not have integrality restrictions.</p><p>Therefore, after <a href="../../api/JuMP/#JuMP.relax_integrality"><code>relax_integrality</code></a> you should call <a href="../../api/JuMP/#JuMP.set_optimizer"><code>set_optimizer</code></a> with a solver that does support dual solutions, such as Clp.</p><p>For example, instead of:</p><pre><code class="language-julia hljs">using JuMP, Cbc
+model = Model(Cbc.Optimizer)
+@variable(model, x, Int)
+undo = relax_integrality(model)
+optimize!(model)
+reduced_cost(x)  # Errors</code></pre><p>do:</p><pre><code class="language-julia hljs">using JuMP, Cbc, Clp
+model = Model(Cbc.Optimizer)
+@variable(model, x, Int)
+undo = relax_integrality(model)
+set_optimizer(model, Clp.Optimizer)
+optimize!(model)
+reduced_cost(x)  # Works</code></pre><h2 id="Backends"><a class="docs-heading-anchor" href="#Backends">Backends</a><a id="Backends-1"></a><a class="docs-heading-anchor-permalink" href="#Backends" title="Permalink"></a></h2><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This section discusses advanced features of JuMP. For new users, you may want to skip this section. You don&#39;t need to know how JuMP manages problems behind the scenes to create and solve JuMP models.</p></div></div><p>A JuMP <a href="../../api/JuMP/#JuMP.Model"><code>Model</code></a> is a thin layer around a <em>backend</em> of type <a href="../../moi/reference/models/#MathOptInterface.ModelLike"><code>MOI.ModelLike</code></a> that stores the optimization problem and acts as the optimization solver.</p><p>However, if you construct a model like <code>Model(HiGHS.Optimizer)</code>, the backend is not a <code>HiGHS.Optimizer</code>, but a more complicated object.</p><p>From JuMP, the MOI backend can be accessed using the <a href="../../api/JuMP/#JuMP.backend"><code>backend</code></a> function. Let&#39;s see what the <a href="../../api/JuMP/#JuMP.backend"><code>backend</code></a> of a JuMP <a href="../../api/JuMP/#JuMP.Model"><code>Model</code></a> is:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; b = backend(model)
+MOIU.CachingOptimizer{MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}, MOIU.UniversalFallback{MOIU.Model{Float64}}}
+in state EMPTY_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
+  fallback for MOIU.Model{Float64}
+with optimizer MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}
+  with 0 variable bridges
+  with 0 constraint bridges
+  with 0 objective bridges
+  with inner model A HiGHS model with 0 columns and 0 rows.</code></pre><p>Uh oh. Even though we passed a <code>HiGHS.Optimizer</code>, the backend is a much more complicated object.</p><h3 id="CachingOptimizer"><a class="docs-heading-anchor" href="#CachingOptimizer">CachingOptimizer</a><a id="CachingOptimizer-1"></a><a class="docs-heading-anchor-permalink" href="#CachingOptimizer" title="Permalink"></a></h3><p>A <code>MOIU.CachingOptimizer</code> is a layer that abstracts the difference between solvers that support incremental modification (for example, they support adding variables one-by-one), and solvers that require the entire problem in a single API call (for example, they only accept the <code>A</code>, <code>b</code> and <code>c</code> matrices of a linear program).</p><p>It has two parts:</p><ol><li>A cache, where the model can be built and modified incrementally<pre><code class="language-julia-repl hljs">julia&gt; b.model_cache
+MOIU.UniversalFallback{MOIU.Model{Float64}}
+fallback for MOIU.Model{Float64}</code></pre></li><li>An optimizer, which is used to solve the problem<pre><code class="language-julia-repl hljs">julia&gt; b.optimizer
+MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}
+with 0 variable bridges
+with 0 constraint bridges
+with 0 objective bridges
+with inner model A HiGHS model with 0 columns and 0 rows.</code></pre></li></ol><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The <a href="#LazyBridgeOptimizer">LazyBridgeOptimizer</a> section explains what a <code>LazyBridgeOptimizer</code> is.</p></div></div><p>The <code>CachingOptimizer</code> has logic to decide when to copy the problem from the cache to the optimizer, and when it can efficiently update the optimizer in-place.</p><p>A <code>CachingOptimizer</code> may be in one of three possible states:</p><ul><li><code>NO_OPTIMIZER</code>: The CachingOptimizer does not have any optimizer.</li><li><code>EMPTY_OPTIMIZER</code>: The CachingOptimizer has an empty optimizer, and it is not synchronized with the cached model.</li><li><code>ATTACHED_OPTIMIZER</code>: The CachingOptimizer has an optimizer, and it is synchronized with the cached model.</li></ul><p>A <code>CachingOptimizer</code> has two modes of operation:</p><ul><li><code>AUTOMATIC</code>: The <code>CachingOptimizer</code> changes its state when necessary. For example, <a href="../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a> will automatically call <code>attach_optimizer</code> (an optimizer must have been previously set). Attempting to add a constraint or perform a modification not supported by the optimizer results in a drop to <code>EMPTY_OPTIMIZER</code> mode.</li><li><code>MANUAL</code>: The user must change the state of the <code>CachingOptimizer</code> using <a href="../../api/JuMP/#MathOptInterface.Utilities.reset_optimizer-Tuple{GenericModel}"><code>MOIU.reset_optimizer(::JuMP.Model)</code></a>, <a href="../../api/JuMP/#MathOptInterface.Utilities.drop_optimizer-Tuple{GenericModel}"><code>MOIU.drop_optimizer(::JuMP.Model)</code></a>, and <a href="../../api/JuMP/#MathOptInterface.Utilities.attach_optimizer-Tuple{GenericModel}"><code>MOIU.attach_optimizer(::JuMP.Model)</code></a>. Attempting to perform an operation in the incorrect state results in an error.</li></ul><p>By default <a href="../../api/JuMP/#JuMP.Model"><code>Model</code></a> will create a <code>CachingOptimizer</code> in <code>AUTOMATIC</code> mode.</p><h3 id="LazyBridgeOptimizer"><a class="docs-heading-anchor" href="#LazyBridgeOptimizer">LazyBridgeOptimizer</a><a id="LazyBridgeOptimizer-1"></a><a class="docs-heading-anchor-permalink" href="#LazyBridgeOptimizer" title="Permalink"></a></h3><p>The second layer that JuMP applies automatically is a <a href="../../moi/submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a>. A <a href="../../moi/submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a> is an MOI layer that attempts to transform the problem from the formulation provided by the user into an equivalent problem supported by the solver. This may involve adding new variables and constraints to the optimizer. The transformations are selected from a set of known recipes called <em>bridges</em>.</p><p>A common example of a bridge is one that splits an interval constraint like <code>@constraint(model, 1 &lt;= x + y &lt;= 2)</code> into two constraints, <code>@constraint(model, x + y &gt;= 1)</code> and <code>@constraint(model, x + y &lt;= 2)</code>.</p><p>Use the <code>add_bridges = false</code> keyword to remove the bridging layer:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer; add_bridges = false)
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS
+
+julia&gt; backend(model)
+MOIU.CachingOptimizer{HiGHS.Optimizer, MOIU.UniversalFallback{MOIU.Model{Float64}}}
+in state EMPTY_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
+  fallback for MOIU.Model{Float64}
+with optimizer A HiGHS model with 0 columns and 0 rows.</code></pre><p>Bridges can be added and removed from a <a href="../../moi/submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a> using <a href="../../api/JuMP/#JuMP.add_bridge"><code>add_bridge</code></a> and <a href="../../api/JuMP/#JuMP.remove_bridge"><code>remove_bridge</code></a>. Use <a href="../../api/JuMP/#JuMP.print_active_bridges"><code>print_active_bridges</code></a> to see which bridges are used to reformulate the model. Read the <a href="../../tutorials/conic/ellipse_approx/#Ellipsoid-approximation">Ellipsoid approximation</a> tutorial for more details.</p><h3 id="Unsafe-backend"><a class="docs-heading-anchor" href="#Unsafe-backend">Unsafe backend</a><a id="Unsafe-backend-1"></a><a class="docs-heading-anchor-permalink" href="#Unsafe-backend" title="Permalink"></a></h3><p>In some advanced use-cases, it is necessary to work with the inner optimization model directly. To access this model, use <a href="../../api/JuMP/#JuMP.unsafe_backend"><code>unsafe_backend</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; backend(model)
+MOIU.CachingOptimizer{MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}, MOIU.UniversalFallback{MOIU.Model{Float64}}}
+in state EMPTY_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.UniversalFallback{MOIU.Model{Float64}}
+  fallback for MOIU.Model{Float64}
+with optimizer MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}
+  with 0 variable bridges
+  with 0 constraint bridges
+  with 0 objective bridges
+  with inner model A HiGHS model with 0 columns and 0 rows.
+
+julia&gt; unsafe_backend(model)
+A HiGHS model with 0 columns and 0 rows.</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><a href="../../api/JuMP/#JuMP.backend"><code>backend</code></a> and <a href="../../api/JuMP/#JuMP.unsafe_backend"><code>unsafe_backend</code></a> are advanced routines. Read their docstrings to understand the caveats of their usage, and only call them if you wish to access low-level solver-specific functions.</p></div></div><h2 id="Direct-mode"><a class="docs-heading-anchor" href="#Direct-mode">Direct mode</a><a id="Direct-mode-1"></a><a class="docs-heading-anchor-permalink" href="#Direct-mode" title="Permalink"></a></h2><p>Using a <code>CachingOptimizer</code> results in an additional copy of the model being stored by JuMP in the <code>.model_cache</code> field. To avoid this overhead, create a JuMP model using <a href="../../api/JuMP/#JuMP.direct_model"><code>direct_model</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = direct_model(HiGHS.Optimizer())
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: DIRECT
+Solver name: HiGHS</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Solvers that do not support incremental modification do not support <code>direct_model</code>. An error will be thrown, telling you to use a <code>CachingOptimizer</code> instead.</p></div></div><p>The benefit of using <a href="../../api/JuMP/#JuMP.direct_model"><code>direct_model</code></a> is that there are no extra layers (for example, <code>Cachingoptimizer</code> or <code>LazyBridgeOptimizer</code>) between <code>model</code> and the provided optimizer:</p><pre><code class="language-julia-repl hljs">julia&gt; backend(model)
+A HiGHS model with 0 columns and 0 rows.</code></pre><p>A downside of direct mode is that there is no bridging layer. Therefore, only constraints which are natively supported by the solver are supported. For example, <code>HiGHS.jl</code> does not implement quadratic constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; model = direct_model(HiGHS.Optimizer());
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @constraint(model, x[1]^2 + x[2]^2 &lt;= 2)
+ERROR: Constraints of type MathOptInterface.ScalarQuadraticFunction{Float64}-in-MathOptInterface.LessThan{Float64} are not supported by the solver.
+
+If you expected the solver to support your problem, you may have an error in your formulation. Otherwise, consider using a different solver.
+
+The list of available solvers, along with the problem types they support, is available at https://jump.dev/JuMP.jl/stable/installation/#Supported-solvers.
+Stacktrace:</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Another downside of direct mode is that the behavior of querying solution information after modifying the problem is solver-specific. This can lead to errors, or the solver silently returning an incorrect value. See <a href="../solutions/#OptimizeNotCalled-errors">OptimizeNotCalled errors</a> for more information.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../tutorials/applications/two_stage_stochastic/">« Two-stage stochastic programs</a><a class="docs-footer-nextpage" href="../variables/">Variables »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/nlp/index.html b/previews/PR3525/manual/nlp/index.html
new file mode 100644
index 00000000000..a1a29a7921e
--- /dev/null
+++ b/previews/PR3525/manual/nlp/index.html
@@ -0,0 +1,347 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Nonlinear Modeling (Legacy) · JuMP</title><meta name="title" content="Nonlinear Modeling (Legacy) · JuMP"/><meta property="og:title" content="Nonlinear Modeling (Legacy) · JuMP"/><meta property="twitter:title" content="Nonlinear Modeling (Legacy) · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li class="is-active"><a class="tocitem" href>Nonlinear Modeling (Legacy)</a><ul class="internal"><li><a class="tocitem" href="#Set-a-nonlinear-objective"><span>Set a nonlinear objective</span></a></li><li><a class="tocitem" href="#Add-a-nonlinear-constraint"><span>Add a nonlinear constraint</span></a></li><li><a class="tocitem" href="#Create-a-nonlinear-expression"><span>Create a nonlinear expression</span></a></li><li><a class="tocitem" href="#Create-a-nonlinear-parameter"><span>Create a nonlinear parameter</span></a></li><li><a class="tocitem" href="#Syntax-notes"><span>Syntax notes</span></a></li><li><a class="tocitem" href="#User-defined-Functions"><span>User-defined Functions</span></a></li><li><a class="tocitem" href="#Factors-affecting-solution-time"><span>Factors affecting solution time</span></a></li><li><a class="tocitem" href="#Querying-derivatives-from-a-JuMP-model"><span>Querying derivatives from a JuMP model</span></a></li><li><a class="tocitem" href="#Raw-expression-input"><span>Raw expression input</span></a></li><li><a class="tocitem" href="#Known-performance-issues"><span>Known performance issues</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Nonlinear Modeling (Legacy)</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Nonlinear Modeling (Legacy)</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/nlp.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Nonlinear-Modeling-(Legacy)"><a class="docs-heading-anchor" href="#Nonlinear-Modeling-(Legacy)">Nonlinear Modeling (Legacy)</a><a id="Nonlinear-Modeling-(Legacy)-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear-Modeling-(Legacy)" title="Permalink"></a></h1><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This page describes the legacy nonlinear interface to JuMP. It has a number of quirks and limitations that prompted the development of a new nonlinear interface. The new interface is documented at <a href="../nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a>. This legacy interface will remain for all future <code>v1.X</code> releases of JuMP. The two nonlinear interfaces cannot be combined.</p></div></div><p>JuMP has support for general smooth nonlinear (convex and nonconvex) optimization problems. JuMP is able to provide exact, sparse second-order derivatives to solvers. This information can improve solver accuracy and performance.</p><p>There are three main changes to solve nonlinear programs in JuMP.</p><ul><li>Use <a href="../../api/JuMP/#JuMP.@NLobjective"><code>@NLobjective</code></a> instead of <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a></li><li>Use <a href="../../api/JuMP/#JuMP.@NLconstraint"><code>@NLconstraint</code></a> instead of <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a></li><li>Use <a href="../../api/JuMP/#JuMP.@NLexpression"><code>@NLexpression</code></a> instead of <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a></li></ul><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>There are some restrictions on what syntax you can use in the <code>@NLxxx</code> macros. Make sure to read the <a href="#Syntax-notes">Syntax notes</a>.</p></div></div><h2 id="Set-a-nonlinear-objective"><a class="docs-heading-anchor" href="#Set-a-nonlinear-objective">Set a nonlinear objective</a><a id="Set-a-nonlinear-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Set-a-nonlinear-objective" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.@NLobjective"><code>@NLobjective</code></a> to set a nonlinear objective.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @NLobjective(model, Min, exp(x[1]) - sqrt(x[2]))</code></pre><p>To modify a nonlinear objective, call <a href="../../api/JuMP/#JuMP.@NLobjective"><code>@NLobjective</code></a> again.</p><h2 id="Add-a-nonlinear-constraint"><a class="docs-heading-anchor" href="#Add-a-nonlinear-constraint">Add a nonlinear constraint</a><a id="Add-a-nonlinear-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-nonlinear-constraint" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.@NLconstraint"><code>@NLconstraint</code></a> to add a nonlinear constraint.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @NLconstraint(model, exp(x[1]) &lt;= 1)
+exp(x[1]) - 1.0 ≤ 0
+
+julia&gt; @NLconstraint(model, [i = 1:2], x[i]^i &gt;= i)
+2-element Vector{NonlinearConstraintRef{ScalarShape}}:
+ x[1] ^ 1.0 - 1.0 ≥ 0
+ x[2] ^ 2.0 - 2.0 ≥ 0
+
+julia&gt; @NLconstraint(model, con[i = 1:2], prod(x[j] for j = 1:i) == i)
+2-element Vector{NonlinearConstraintRef{ScalarShape}}:
+ (*)(x[1]) - 1.0 = 0
+ x[1] * x[2] - 2.0 = 0</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>You can only create nonlinear constraints with <code>&lt;=</code>, <code>&gt;=</code>, and <code>==</code>. More general <code>Nonlinear</code>-in-<code>Set</code> constraints are not supported.</p></div></div><p>Delete a nonlinear constraint using <a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; delete(model, con[1])</code></pre><h2 id="Create-a-nonlinear-expression"><a class="docs-heading-anchor" href="#Create-a-nonlinear-expression">Create a nonlinear expression</a><a id="Create-a-nonlinear-expression-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-nonlinear-expression" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.@NLexpression"><code>@NLexpression</code></a> to create nonlinear expression objects. The syntax is identical to <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a>, except that the expression can contain nonlinear terms.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; expr = @NLexpression(model, exp(x[1]) + sqrt(x[2]))
+subexpression[1]: exp(x[1]) + sqrt(x[2])
+
+julia&gt; my_anon_expr = @NLexpression(model, [i = 1:2], sin(x[i]))
+2-element Vector{NonlinearExpression}:
+ subexpression[2]: sin(x[1])
+ subexpression[3]: sin(x[2])
+
+julia&gt; @NLexpression(model, my_expr[i = 1:2], sin(x[i]))
+2-element Vector{NonlinearExpression}:
+ subexpression[4]: sin(x[1])
+ subexpression[5]: sin(x[2])</code></pre><p>Nonlinear expression can be used in <a href="../../api/JuMP/#JuMP.@NLobjective"><code>@NLobjective</code></a>, <a href="../../api/JuMP/#JuMP.@NLconstraint"><code>@NLconstraint</code></a>, and even nested in other <a href="../../api/JuMP/#JuMP.@NLexpression"><code>@NLexpression</code></a>s.</p><pre><code class="language-julia-repl hljs">julia&gt; @NLobjective(model, Min, expr^2 + 1)
+
+julia&gt; @NLconstraint(model, [i = 1:2], my_expr[i] &lt;= i)
+2-element Vector{NonlinearConstraintRef{ScalarShape}}:
+ subexpression[4] - 1.0 ≤ 0
+ subexpression[5] - 2.0 ≤ 0
+
+julia&gt; @NLexpression(model, nested[i = 1:2], sin(my_expr[i]))
+2-element Vector{NonlinearExpression}:
+ subexpression[6]: sin(subexpression[4])
+ subexpression[7]: sin(subexpression[5])</code></pre><p>Use <a href="../../api/JuMP/#JuMP.value"><code>value</code></a> to query the value of a nonlinear expression:</p><pre><code class="language-julia-repl hljs">julia&gt; set_start_value(x[1], 1.0)
+
+julia&gt; value(start_value, nested[1])
+0.7456241416655579
+
+julia&gt; sin(sin(1.0))
+0.7456241416655579</code></pre><h2 id="Create-a-nonlinear-parameter"><a class="docs-heading-anchor" href="#Create-a-nonlinear-parameter">Create a nonlinear parameter</a><a id="Create-a-nonlinear-parameter-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-nonlinear-parameter" title="Permalink"></a></h2><p>For nonlinear models only, JuMP offers a syntax for explicit &quot;parameter&quot; objects, which are constants in the model that can be efficiently updated between solves.</p><p>Nonlinear parameters are declared by using the <a href="../../api/JuMP/#JuMP.@NLparameter"><code>@NLparameter</code></a> macro and may be indexed by arbitrary sets analogously to JuMP variables and expressions.</p><p>The initial value of the parameter must be provided on the right-hand side of the <code>==</code> sign.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @NLparameter(model, p[i = 1:2] == i)
+2-element Vector{NonlinearParameter}:
+ parameter[1] == 1.0
+ parameter[2] == 2.0</code></pre><p>Create anonymous parameters using the <code>value</code> keyword:</p><pre><code class="language-julia-repl hljs">julia&gt; anon_parameter = @NLparameter(model, value = 1)
+parameter[3] == 1.0</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>A parameter is not an optimization variable. It must be fixed to a value with <code>==</code>. If you want a parameter that is <code>&lt;=</code> or <code>&gt;=</code>, create a variable instead using <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>.</p></div></div><p>Use <a href="../../api/JuMP/#JuMP.value"><code>value</code></a> and <a href="../../api/JuMP/#JuMP.set_value"><code>set_value</code></a> to query or update the value of a parameter.</p><pre><code class="language-julia-repl hljs">julia&gt; value.(p)
+2-element Vector{Float64}:
+ 1.0
+ 2.0
+
+julia&gt; set_value(p[2], 3.0)
+3.0
+
+julia&gt; value.(p)
+2-element Vector{Float64}:
+ 1.0
+ 3.0</code></pre><p>Nonlinear parameters must be used <em>within nonlinear macros</em> only.</p><h3 id="When-to-use-a-parameter"><a class="docs-heading-anchor" href="#When-to-use-a-parameter">When to use a parameter</a><a id="When-to-use-a-parameter-1"></a><a class="docs-heading-anchor-permalink" href="#When-to-use-a-parameter" title="Permalink"></a></h3><p>Nonlinear parameters are useful when solving nonlinear models in a sequence:</p><pre><code class="language-julia hljs">using JuMP, Ipopt
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, z)
+@NLparameter(model, x == 1.0)
+@NLobjective(model, Min, (z - x)^2)
+optimize!(model)
+@show value(z) # Equals 1.0.
+
+# Now, update the value of x to solve a different problem.
+set_value(x, 5.0)
+optimize!(model)
+@show value(z) # Equals 5.0</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">value(z) = 1.0
+value(z) = 5.0</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Using nonlinear parameters can be faster than creating a new model from scratch with updated data because JuMP is able to avoid repeating a number of steps in processing the model before handing it off to the solver.</p></div></div><h2 id="Syntax-notes"><a class="docs-heading-anchor" href="#Syntax-notes">Syntax notes</a><a id="Syntax-notes-1"></a><a class="docs-heading-anchor-permalink" href="#Syntax-notes" title="Permalink"></a></h2><p>The syntax accepted in nonlinear macros is more restricted than the syntax for linear and quadratic macros. We note some important points below.</p><h3 id="Scalar-operations-only"><a class="docs-heading-anchor" href="#Scalar-operations-only">Scalar operations only</a><a id="Scalar-operations-only-1"></a><a class="docs-heading-anchor-permalink" href="#Scalar-operations-only" title="Permalink"></a></h3><p>Except for the splatting syntax discussed below, all expressions must be simple scalar operations. You cannot use <code>dot</code>, matrix-vector products, vector slices, etc.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @variable(model, y);
+
+julia&gt; c = [1, 2];
+
+julia&gt; @NLobjective(model, Min, c&#39; * x + 3y)
+ERROR: Unexpected array [1 2] in nonlinear expression. Nonlinear expressions may contain only scalar expressions.
+[...]</code></pre><p>Translate vector operations into explicit <code>sum()</code> operations:</p><pre><code class="language-julia-repl hljs">julia&gt; @NLobjective(model, Min, sum(c[i] * x[i] for i = 1:2) + 3y)</code></pre><p>Or use an <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; @expression(model, expr, c&#39; * x)
+x[1] + 2 x[2]
+
+julia&gt; @NLobjective(model, Min, expr + 3y)
+</code></pre><h3 id="Splatting"><a class="docs-heading-anchor" href="#Splatting">Splatting</a><a id="Splatting-1"></a><a class="docs-heading-anchor-permalink" href="#Splatting" title="Permalink"></a></h3><p>The <a href="https://docs.julialang.org/en/v1/manual/faq/#...-splits-one-argument-into-many-different-arguments-in-function-calls-1">splatting operator</a>   <code>...</code> is recognized in a very restricted setting for expanding function   arguments. The expression splatted can be <em>only</em> a symbol. More complex   expressions are not recognized.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:3]);
+
+julia&gt; @NLconstraint(model, *(x...) &lt;= 1.0)
+x[1] * x[2] * x[3] - 1.0 ≤ 0
+
+julia&gt; @NLconstraint(model, *((x / 2)...) &lt;= 0.0)
+ERROR: Unsupported use of the splatting operator. JuMP supports splatting only symbols. For example, `x...` is ok, but `(x + 1)...`, `[x; y]...` and `g(f(y)...)` are not.</code></pre><h2 id="User-defined-Functions"><a class="docs-heading-anchor" href="#User-defined-Functions">User-defined Functions</a><a id="User-defined-Functions-1"></a><a class="docs-heading-anchor-permalink" href="#User-defined-Functions" title="Permalink"></a></h2><p>JuMP natively supports the set of univariate and multivariate functions recognized by the <code>MOI.Nonlinear</code> submodule. In addition to this list of functions, it is possible to register custom <em>user-defined</em> nonlinear functions. User-defined functions can be used anywhere in <a href="../../api/JuMP/#JuMP.@NLobjective"><code>@NLobjective</code></a>, <a href="../../api/JuMP/#JuMP.@NLconstraint"><code>@NLconstraint</code></a>, and <a href="../../api/JuMP/#JuMP.@NLexpression"><code>@NLexpression</code></a>.</p><p>JuMP will attempt to automatically register functions it detects in your nonlinear expressions, which usually means manually registering a function is not needed. Two exceptions are if you want to provide custom derivatives, or if the function is not available in the scope of the nonlinear expression.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>User-defined functions must return a scalar output. For a work-around, see <a href="../../tutorials/nonlinear/tips_and_tricks/#User-defined-operators-with-vector-outputs">User-defined operators with vector outputs</a>.</p></div></div><h3 id="Automatic-differentiation"><a class="docs-heading-anchor" href="#Automatic-differentiation">Automatic differentiation</a><a id="Automatic-differentiation-1"></a><a class="docs-heading-anchor-permalink" href="#Automatic-differentiation" title="Permalink"></a></h3><p>JuMP does not support black-box optimization, so all user-defined functions must provide derivatives in some form. Fortunately, JuMP supports <strong>automatic differentiation of user-defined functions</strong>, a feature to our knowledge not available in any comparable modeling systems.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Automatic differentiation is <em>not</em> finite differencing. JuMP&#39;s automatically computed derivatives are not subject to approximation error.</p></div></div><p>JuMP uses <a href="https://github.com/JuliaDiff/ForwardDiff.jl">ForwardDiff.jl</a> to perform automatic differentiation; see the ForwardDiff.jl <a href="https://www.juliadiff.org/ForwardDiff.jl/v0.10.2/user/limitations.html">documentation</a> for a description of how to write a function suitable for automatic differentiation.</p><h4 id="Common-mistakes-when-writing-a-user-defined-function"><a class="docs-heading-anchor" href="#Common-mistakes-when-writing-a-user-defined-function">Common mistakes when writing a user-defined function</a><a id="Common-mistakes-when-writing-a-user-defined-function-1"></a><a class="docs-heading-anchor-permalink" href="#Common-mistakes-when-writing-a-user-defined-function" title="Permalink"></a></h4><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Get an error like <code>No method matching Float64(::ForwardDiff.Dual)</code>? Read this section, and see the guidelines at <a href="https://www.juliadiff.org/ForwardDiff.jl/release-0.10/user/limitations.html">ForwardDiff.jl</a>.</p></div></div><p>The most common error is that your user-defined function is not generic with respect to the number type, that is, don&#39;t assume that the input to the function is <code>Float64</code>.</p><pre><code class="language-julia hljs">f(x::Float64) = 2 * x  # This will not work.
+f(x::Real)    = 2 * x  # This is good.
+f(x)          = 2 * x  # This is also good.</code></pre><p>Another reason you may encounter this error is if you create arrays inside your function which are <code>Float64</code>.</p><pre><code class="language-julia hljs">function bad_f(x...)
+    y = zeros(length(x))  # This constructs an array of `Float64`!
+    for i = 1:length(x)
+        y[i] = x[i]^i
+    end
+    return sum(y)
+end
+
+function good_f(x::T...) where {T&lt;:Real}
+    y = zeros(T, length(x))  # Construct an array of type `T` instead!
+    for i = 1:length(x)
+        y[i] = x[i]^i
+    end
+    return sum(y)
+end</code></pre><h3 id="Register-a-function"><a class="docs-heading-anchor" href="#Register-a-function">Register a function</a><a id="Register-a-function-1"></a><a class="docs-heading-anchor-permalink" href="#Register-a-function" title="Permalink"></a></h3><p>To register a user-defined function with derivatives computed by automatic differentiation, use the <a href="../../api/JuMP/#JuMP.register"><code>register</code></a> method as in the following example:</p><pre><code class="language-julia hljs">square(x) = x^2
+f(x, y) = (x - 1)^2 + (y - 2)^2
+
+model = Model()
+
+register(model, :square, 1, square; autodiff = true)
+register(model, :my_f, 2, f; autodiff = true)
+
+@variable(model, x[1:2] &gt;= 0.5)
+@NLobjective(model, Min, my_f(x[1], square(x[2])))</code></pre><p>The above code creates a JuMP model with the objective function <code>(x[1] - 1)^2 + (x[2]^2 - 2)^2</code>. The arguments to <a href="../../api/JuMP/#JuMP.register"><code>register</code></a> are:</p><ol><li>The model for which the functions are registered.</li><li>A Julia symbol object which serves as the name of the user-defined function in JuMP expressions.</li><li>The number of input arguments that the function takes.</li><li>The Julia method which computes the function</li><li>A flag to instruct JuMP to compute exact gradients automatically.</li></ol><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>The symbol <code>:my_f</code> doesn&#39;t have to match the name of the function <code>f</code>. However, it&#39;s more readable if it does. Make sure you use <code>my_f</code> and not <code>f</code> in the macros.</p></div></div><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>User-defined functions cannot be re-registered and will not update if you modify the underlying Julia function. If you want to change a user-defined function between solves, rebuild the model or use a different name. To use a different name programmatically, see <a href="#Raw-expression-input">Raw expression input</a>.</p></div></div><h3 id="Register-a-function-and-gradient"><a class="docs-heading-anchor" href="#Register-a-function-and-gradient">Register a function and gradient</a><a id="Register-a-function-and-gradient-1"></a><a class="docs-heading-anchor-permalink" href="#Register-a-function-and-gradient" title="Permalink"></a></h3><p>Forward-mode automatic differentiation as implemented by ForwardDiff.jl has a computational cost that scales linearly with the number of input dimensions. As such, it is not the most efficient way to compute gradients of user-defined functions if the number of input arguments is large. In this case, users may want to provide their own routines for evaluating gradients.</p><h4 id="Univariate-functions"><a class="docs-heading-anchor" href="#Univariate-functions">Univariate functions</a><a id="Univariate-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Univariate-functions" title="Permalink"></a></h4><p>For univariate functions, the gradient function <code>∇f</code> returns a number that represents the first-order derivative:</p><pre><code class="language-julia hljs">f(x) = x^2
+∇f(x) = 2x
+model = Model()
+register(model, :my_square, 1, f, ∇f; autodiff = true)
+@variable(model, x &gt;= 0)
+@NLobjective(model, Min, my_square(x))</code></pre><p>If <code>autodiff = true</code>, JuMP will use automatic differentiation to compute the hessian.</p><h4 id="Multivariate-functions"><a class="docs-heading-anchor" href="#Multivariate-functions">Multivariate functions</a><a id="Multivariate-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Multivariate-functions" title="Permalink"></a></h4><p>For multivariate functions, the gradient function <code>∇f</code> must take a gradient vector as the first argument that is filled in-place:</p><pre><code class="language-julia hljs">f(x, y) = (x - 1)^2 + (y - 2)^2
+function ∇f(g::AbstractVector{T}, x::T, y::T) where {T}
+    g[1] = 2 * (x - 1)
+    g[2] = 2 * (y - 2)
+    return
+end
+
+model = Model()
+register(model, :my_square, 2, f, ∇f)
+@variable(model, x[1:2] &gt;= 0)
+@NLobjective(model, Min, my_square(x[1], x[2]))</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Make sure the first argument to <code>∇f</code> supports an <code>AbstractVector</code>, and do not assume the input is <code>Float64</code>.</p></div></div><h3 id="Register-a-function,-gradient,-and-hessian"><a class="docs-heading-anchor" href="#Register-a-function,-gradient,-and-hessian">Register a function, gradient, and hessian</a><a id="Register-a-function,-gradient,-and-hessian-1"></a><a class="docs-heading-anchor-permalink" href="#Register-a-function,-gradient,-and-hessian" title="Permalink"></a></h3><p>You can also register a function with the second-order derivative information, which is a scalar for univariate functions, and a symmetric matrix for multivariate functions.</p><h4 id="Univariate-functions-2"><a class="docs-heading-anchor" href="#Univariate-functions-2">Univariate functions</a><a class="docs-heading-anchor-permalink" href="#Univariate-functions-2" title="Permalink"></a></h4><p>Pass a function which returns a number representing the second-order derivative:</p><pre><code class="language-julia hljs">f(x) = x^2
+∇f(x) = 2x
+∇²f(x) = 2
+model = Model()
+register(model, :my_square, 1, f, ∇f, ∇²f)
+@variable(model, x &gt;= 0)
+@NLobjective(model, Min, my_square(x))</code></pre><h4 id="Multivariate-functions-2"><a class="docs-heading-anchor" href="#Multivariate-functions-2">Multivariate functions</a><a class="docs-heading-anchor-permalink" href="#Multivariate-functions-2" title="Permalink"></a></h4><p>For multivariate functions, the hessian function <code>∇²f</code> must take an <code>AbstractMatrix</code> as the first argument, the lower-triangular of which is filled in-place:</p><pre><code class="language-julia hljs">f(x...) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2
+function ∇f(g, x...)
+    g[1] = 400 * x[1]^3 - 400 * x[1] * x[2] + 2 * x[1] - 2
+    g[2] = 200 * (x[2] - x[1]^2)
+    return
+end
+function ∇²f(H, x...)
+    H[1, 1] = 1200 * x[1]^2 - 400 * x[2] + 2
+    # H[1, 2] = -400 * x[1]  &lt;-- Not needed. Fill the lower-triangular only.
+    H[2, 1] = -400 * x[1]
+    H[2, 2] = 200.0
+    return
+end
+
+model = Model()
+register(model, :rosenbrock, 2, f, ∇f, ∇²f)
+@variable(model, x[1:2])
+@NLobjective(model, Min, rosenbrock(x[1], x[2]))</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>You may assume the Hessian matrix <code>H</code> is initialized with zeros, and because <code>H</code> is symmetric, you need only to fill in the non-zero of the lower-triangular terms. The matrix type passed in as <code>H</code> depends on the automatic differentiation system, so make sure the first argument to the Hessian function supports an <code>AbstractMatrix</code> (it may be something other than <code>Matrix{Float64}</code>). However, you may assume only that <code>H</code> supports <code>size(H)</code> and <code>setindex!</code>. Finally, the matrix is treated as dense, so the performance will be poor on functions with high-dimensional input.</p></div></div><h3 id="User-defined-functions-with-vector-inputs"><a class="docs-heading-anchor" href="#User-defined-functions-with-vector-inputs">User-defined functions with vector inputs</a><a id="User-defined-functions-with-vector-inputs-1"></a><a class="docs-heading-anchor-permalink" href="#User-defined-functions-with-vector-inputs" title="Permalink"></a></h3><p>User-defined functions which take vectors as input arguments (for example, <code>f(x::Vector)</code>) are <em>not</em> supported. Instead, use Julia&#39;s splatting syntax to create a function with scalar arguments. For example, instead of</p><pre><code class="language-julia hljs">f(x::Vector) = sum(x[i]^i for i in 1:length(x))</code></pre><p>define:</p><pre><code class="language-julia hljs">f(x...) = sum(x[i]^i for i in 1:length(x))</code></pre><p>This function <code>f</code> can be used in a JuMP model as follows:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[1:5] &gt;= 0)
+f(x...) = sum(x[i]^i for i in 1:length(x))
+register(model, :f, 5, f; autodiff = true)
+@NLobjective(model, Min, f(x...))</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Make sure to read the syntax restrictions of <a href="#Splatting">Splatting</a>.</p></div></div><h2 id="Factors-affecting-solution-time"><a class="docs-heading-anchor" href="#Factors-affecting-solution-time">Factors affecting solution time</a><a id="Factors-affecting-solution-time-1"></a><a class="docs-heading-anchor-permalink" href="#Factors-affecting-solution-time" title="Permalink"></a></h2><p>The execution time when solving a nonlinear programming problem can be divided into two parts, the time spent in the optimization algorithm (the solver) and the time spent evaluating the nonlinear functions and corresponding derivatives. Ipopt explicitly displays these two timings in its output, for example:</p><pre><code class="nohighlight hljs">Total CPU secs in IPOPT (w/o function evaluations)   =      7.412
+Total CPU secs in NLP function evaluations           =      2.083</code></pre><p>For Ipopt in particular, one can improve the performance by installing advanced sparse linear algebra packages, see <a href="../../installation/#Installation-Guide">Installation Guide</a>. For other solvers, see their respective documentation for performance tips.</p><p>The function evaluation time, on the other hand, is the responsibility of the modeling language. JuMP computes derivatives by using reverse-mode automatic differentiation with graph coloring methods for exploiting sparsity of the Hessian matrix. As a conservative bound, JuMP&#39;s performance here currently may be expected to be within a factor of 5 of AMPL&#39;s. Our <a href="https://mlubin.github.io/pdf/jump-sirev.pdf">paper in SIAM Review</a> has more details.</p><h2 id="Querying-derivatives-from-a-JuMP-model"><a class="docs-heading-anchor" href="#Querying-derivatives-from-a-JuMP-model">Querying derivatives from a JuMP model</a><a id="Querying-derivatives-from-a-JuMP-model-1"></a><a class="docs-heading-anchor-permalink" href="#Querying-derivatives-from-a-JuMP-model" title="Permalink"></a></h2><p>For some advanced use cases, one may want to directly query the derivatives of a JuMP model instead of handing the problem off to a solver. Internally, JuMP implements the <a href="../../moi/reference/nonlinear/#MathOptInterface.AbstractNLPEvaluator"><code>MOI.AbstractNLPEvaluator</code></a> interface. To obtain an NLP evaluator object from a JuMP model, use <a href="../../api/JuMP/#JuMP.NLPEvaluator"><code>NLPEvaluator</code></a>. <a href="../../api/JuMP/#JuMP.index"><code>index</code></a> returns the <a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> corresponding to a JuMP variable. <code>MOI.VariableIndex</code> itself is a type-safe wrapper for <code>Int64</code> (stored in the <code>.value</code> field.)</p><p>For example:</p><pre><code class="language-julia-repl hljs">julia&gt; raw_index(v::MOI.VariableIndex) = v.value
+raw_index (generic function with 1 method)
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; @NLobjective(model, Min, sin(x) + sin(y))
+
+julia&gt; values = zeros(2)
+2-element Vector{Float64}:
+ 0.0
+ 0.0
+
+julia&gt; x_index = raw_index(JuMP.index(x))
+1
+
+julia&gt; y_index = raw_index(JuMP.index(y))
+2
+
+julia&gt; values[x_index] = 2.0
+2.0
+
+julia&gt; values[y_index] = 3.0
+3.0
+
+julia&gt; d = NLPEvaluator(model)
+Nonlinear.Evaluator with available features:
+  * :Grad
+  * :Jac
+  * :JacVec
+  * :Hess
+  * :HessVec
+  * :ExprGraph
+
+julia&gt; MOI.initialize(d, [:Grad])
+
+julia&gt; MOI.eval_objective(d, values)
+1.0504174348855488
+
+julia&gt; sin(2.0) + sin(3.0)
+1.0504174348855488
+
+julia&gt; ∇f = zeros(2)
+2-element Vector{Float64}:
+ 0.0
+ 0.0
+
+julia&gt; MOI.eval_objective_gradient(d, ∇f, values)
+
+julia&gt; ∇f[x_index], ∇f[y_index]
+(-0.4161468365471424, -0.9899924966004454)
+
+julia&gt; cos(2.0), cos(3.0)
+(-0.4161468365471424, -0.9899924966004454)</code></pre><p>Only nonlinear constraints (those added with <a href="../../api/JuMP/#JuMP.@NLconstraint"><code>@NLconstraint</code></a>), and nonlinear objectives (added with <a href="../../api/JuMP/#JuMP.@NLobjective"><code>@NLobjective</code></a>) exist in the scope of the <a href="../../api/JuMP/#JuMP.NLPEvaluator"><code>NLPEvaluator</code></a>.</p><p>The <a href="../../api/JuMP/#JuMP.NLPEvaluator"><code>NLPEvaluator</code></a> <em>does not evaluate derivatives of linear or quadratic constraints or objectives</em>.</p><p>The <a href="../../api/JuMP/#JuMP.index"><code>index</code></a> method applied to a nonlinear constraint reference object returns its index as a <a href="../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.ConstraintIndex"><code>MOI.Nonlinear.ConstraintIndex</code></a>. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @NLconstraint(model, cons1, sin(x) &lt;= 1);
+
+julia&gt; @NLconstraint(model, cons2, x + 5 == 10);
+
+julia&gt; typeof(cons1)
+NonlinearConstraintRef{ScalarShape} (alias for ConstraintRef{GenericModel{Float64}, MathOptInterface.Nonlinear.ConstraintIndex, ScalarShape})
+
+julia&gt; index(cons1)
+MathOptInterface.Nonlinear.ConstraintIndex(1)
+
+julia&gt; index(cons2)
+MathOptInterface.Nonlinear.ConstraintIndex(2)</code></pre><p>Note that for one-sided nonlinear constraints, JuMP subtracts any values on the right-hand side when computing expressions. In other words, one-sided nonlinear constraints are always transformed to have a right-hand side of zero.</p><p>This method of querying derivatives directly from a JuMP model is convenient for interacting with the model in a structured way, for example, for accessing derivatives of specific variables. For example, in statistical maximum likelihood estimation problems, one is often interested in the Hessian matrix at the optimal solution, which can be queried using the <a href="../../api/JuMP/#JuMP.NLPEvaluator"><code>NLPEvaluator</code></a>.</p><h2 id="Raw-expression-input"><a class="docs-heading-anchor" href="#Raw-expression-input">Raw expression input</a><a id="Raw-expression-input-1"></a><a class="docs-heading-anchor-permalink" href="#Raw-expression-input" title="Permalink"></a></h2><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This section requires advanced knowledge of Julia&#39;s <code>Expr</code>. You should read the <a href="https://docs.julialang.org/en/v1/manual/metaprogramming/#Expressions-and-evaluation">Expressions and evaluation</a> section of the Julia documentation first.</p></div></div><p>In addition to the <a href="../../api/JuMP/#JuMP.@NLexpression"><code>@NLexpression</code></a>, <a href="../../api/JuMP/#JuMP.@NLobjective"><code>@NLobjective</code></a> and <a href="../../api/JuMP/#JuMP.@NLconstraint"><code>@NLconstraint</code></a> macros, it is also possible to provide Julia <code>Expr</code> objects directly by using <a href="../../api/JuMP/#JuMP.add_nonlinear_expression"><code>add_nonlinear_expression</code></a>, <a href="../../api/JuMP/#JuMP.set_nonlinear_objective"><code>set_nonlinear_objective</code></a> and <a href="../../api/JuMP/#JuMP.add_nonlinear_constraint"><code>add_nonlinear_constraint</code></a>.</p><p>This input form may be useful if the expressions are generated programmatically, or if you experience compilation issues with the macro input (see <a href="#Known-performance-issues">Known performance issues</a> for more information).</p><h3 id="Add-a-nonlinear-expression"><a class="docs-heading-anchor" href="#Add-a-nonlinear-expression">Add a nonlinear expression</a><a id="Add-a-nonlinear-expression-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-nonlinear-expression" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.add_nonlinear_expression"><code>add_nonlinear_expression</code></a> to add a nonlinear expression to the model.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; expr = :($(x) + sin($(x)^2))
+:(x + sin(x ^ 2))
+
+julia&gt; expr_ref = add_nonlinear_expression(model, expr)
+subexpression[1]: x + sin(x ^ 2.0)</code></pre><p>This is equivalent to</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr_ref = @NLexpression(model, x + sin(x^2))
+subexpression[1]: x + sin(x ^ 2.0)</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>You must interpolate the variables directly into the expression <code>expr</code>.</p></div></div><h3 id="Set-the-objective-function"><a class="docs-heading-anchor" href="#Set-the-objective-function">Set the objective function</a><a id="Set-the-objective-function-1"></a><a class="docs-heading-anchor-permalink" href="#Set-the-objective-function" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.set_nonlinear_objective"><code>set_nonlinear_objective</code></a> to set a nonlinear objective.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr = :($(x) + $(x)^2)
+:(x + x ^ 2)
+
+julia&gt; set_nonlinear_objective(model, MIN_SENSE, expr)</code></pre><p>This is equivalent to</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @NLobjective(model, Min, x + x^2)</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>You must use <code>MIN_SENSE</code> or <code>MAX_SENSE</code> instead of <code>Min</code> and <code>Max</code>.</p></div></div><h3 id="Add-a-constraint"><a class="docs-heading-anchor" href="#Add-a-constraint">Add a constraint</a><a id="Add-a-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-constraint" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.add_nonlinear_constraint"><code>add_nonlinear_constraint</code></a> to add a nonlinear constraint.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr = :($(x) + $(x)^2)
+:(x + x ^ 2)
+
+julia&gt; add_nonlinear_constraint(model, :($(expr) &lt;= 1))
+(x + x ^ 2.0) - 1.0 ≤ 0</code></pre><p>This is equivalent to</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @NLconstraint(model, Min, x + x^2 &lt;= 1)
+(x + x ^ 2.0) - 1.0 ≤ 0</code></pre><h3 id="More-complicated-examples"><a class="docs-heading-anchor" href="#More-complicated-examples">More complicated examples</a><a id="More-complicated-examples-1"></a><a class="docs-heading-anchor-permalink" href="#More-complicated-examples" title="Permalink"></a></h3><p>Raw expression input is most useful when the expressions are generated programmatically, often in conjunction with user-defined functions.</p><p>As an example, we construct a model with the nonlinear constraints <code>f(x) &lt;= 1</code>, where <code>f(x) = x^2</code> and <code>f(x) = sin(x)^2</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; function main(functions::Vector{Function})
+           model = Model()
+           @variable(model, x)
+           for (i, f) in enumerate(functions)
+               f_sym = Symbol(&quot;f_$(i)&quot;)
+               register(model, f_sym, 1, f; autodiff = true)
+               add_nonlinear_constraint(model, :($(f_sym)($(x)) &lt;= 1))
+           end
+           print(model)
+           return
+       end
+main (generic function with 1 method)
+
+julia&gt; main([x -&gt; x^2, x -&gt; sin(x)^2])
+Feasibility
+Subject to
+ f_1(x) - 1.0 ≤ 0
+ f_2(x) - 1.0 ≤ 0</code></pre><p>As another example, we construct a model with the constraint <code>x^2 + sin(x)^2 &lt;= 1</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; function main(functions::Vector{Function})
+           model = Model()
+           @variable(model, x)
+           expr = Expr(:call, :+)
+           for (i, f) in enumerate(functions)
+               f_sym = Symbol(&quot;f_$(i)&quot;)
+               register(model, f_sym, 1, f; autodiff = true)
+               push!(expr.args, :($(f_sym)($(x))))
+           end
+           add_nonlinear_constraint(model, :($(expr) &lt;= 1))
+           print(model)
+           return
+       end
+main (generic function with 1 method)
+
+julia&gt; main([x -&gt; x^2, x -&gt; sin(x)^2])
+Feasibility
+Subject to
+ (f_1(x) + f_2(x)) - 1.0 ≤ 0</code></pre><h3 id="Registered-functions-with-a-variable-number-of-arguments"><a class="docs-heading-anchor" href="#Registered-functions-with-a-variable-number-of-arguments">Registered functions with a variable number of arguments</a><a id="Registered-functions-with-a-variable-number-of-arguments-1"></a><a class="docs-heading-anchor-permalink" href="#Registered-functions-with-a-variable-number-of-arguments" title="Permalink"></a></h3><p>User defined functions require a fixed number of input arguments. However, sometimes you will want to use a registered function like:</p><pre><code class="language-julia-repl hljs">julia&gt; f(x...) = sum(exp(x[i]^2) for i in 1:length(x));</code></pre><p>with different numbers of arguments.</p><p>The solution is to register the same function <code>f</code> for each unique number of input arguments, making sure to use a unique name each time. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; A = [[1], [1, 2], [2, 3, 4], [1, 3, 4, 5]];
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:5]);
+
+julia&gt; funcs = Set{Symbol}();
+
+julia&gt; for a in A
+           key = Symbol(&quot;f$(length(a))&quot;)
+           if !(key in funcs)
+               push!(funcs, key)
+               register(model, key, length(a), f; autodiff = true)
+           end
+           add_nonlinear_constraint(model, :($key($(x[a]...)) &lt;= 1))
+       end
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ f1(x[1]) - 1.0 ≤ 0
+ f2(x[1], x[2]) - 1.0 ≤ 0
+ f3(x[2], x[3], x[4]) - 1.0 ≤ 0
+ f4(x[1], x[3], x[4], x[5]) - 1.0 ≤ 0</code></pre><h2 id="Known-performance-issues"><a class="docs-heading-anchor" href="#Known-performance-issues">Known performance issues</a><a id="Known-performance-issues-1"></a><a class="docs-heading-anchor-permalink" href="#Known-performance-issues" title="Permalink"></a></h2><p>The macro-based input to JuMP&#39;s nonlinear interface can cause a performance issue if you:</p><ol><li>write a macro with a large number (hundreds) of terms</li><li>call that macro from within a function instead of from the top-level in global scope.</li></ol><p>The first issue does not depend on the number of resulting terms in the mathematical expression, but rather the number of terms in the Julia <code>Expr</code> representation of that expression. For example, the expression <code>sum(x[i] for i in 1:1_000_000)</code> contains one million mathematical terms, but the <code>Expr</code> representation is just a single sum.</p><p>The most common cause, other than a lot of tedious typing, is if you write a program that automatically writes a JuMP model as a text file, which you later execute. One example is <a href="https://github.com/lanl-ansi/MINLPLib.jl">MINLPlib.jl</a> which automatically transpiled models in the GAMS scalar format into JuMP examples.</p><p>As a rule of thumb, if you are writing programs to automatically generate expressions for the JuMP macros, you should target the <a href="#Raw-expression-input">Raw expression input</a> instead. For more information, read <a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1997">MathOptInterface Issue#1997</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../nonlinear/">« Nonlinear Modeling</a><a class="docs-footer-nextpage" href="../../api/JuMP/">JuMP »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/nonlinear/index.html b/previews/PR3525/manual/nonlinear/index.html
new file mode 100644
index 00000000000..7f95872bd27
--- /dev/null
+++ b/previews/PR3525/manual/nonlinear/index.html
@@ -0,0 +1,199 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Nonlinear Modeling · JuMP</title><meta name="title" content="Nonlinear Modeling · JuMP"/><meta property="og:title" content="Nonlinear Modeling · JuMP"/><meta property="twitter:title" content="Nonlinear Modeling · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li class="is-active"><a class="tocitem" href>Nonlinear Modeling</a><ul class="internal"><li><a class="tocitem" href="#Set-a-nonlinear-objective"><span>Set a nonlinear objective</span></a></li><li><a class="tocitem" href="#Add-a-nonlinear-constraint"><span>Add a nonlinear constraint</span></a></li><li><a class="tocitem" href="#Create-a-nonlinear-expression"><span>Create a nonlinear expression</span></a></li><li><a class="tocitem" href="#Nonlinear-expressions-in-detail"><span>Nonlinear expressions in detail</span></a></li><li><a class="tocitem" href="#jump_user_defined_operators"><span>User-defined operators</span></a></li></ul></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Nonlinear Modeling</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Nonlinear Modeling</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/nonlinear.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Nonlinear-Modeling"><a class="docs-heading-anchor" href="#Nonlinear-Modeling">Nonlinear Modeling</a><a id="Nonlinear-Modeling-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear-Modeling" title="Permalink"></a></h1><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This page describes a new nonlinear interface to JuMP. It replaces the legacy <code>@NL</code> interface, which is documented at <a href="../nlp/#Nonlinear-Modeling-(Legacy)">Nonlinear Modeling (Legacy)</a>. The API described below is stable, and it will not break with future 1.X releases of JuMP. However, solver support may be limited, and there may be gaps in functionality compared with the legacy interface. To report a bug, or request a missing feature, please <a href="https://github.com/jump-dev/JuMP.jl/issues/new/choose">open an issue</a>.</p></div></div><p>JuMP has support for nonlinear (convex and nonconvex) optimization problems. JuMP is able to automatically provide exact, sparse second-order derivatives to solvers. This information can improve solver accuracy and performance.</p><h2 id="Set-a-nonlinear-objective"><a class="docs-heading-anchor" href="#Set-a-nonlinear-objective">Set a nonlinear objective</a><a id="Set-a-nonlinear-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Set-a-nonlinear-objective" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> to set a nonlinear objective.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @objective(model, Min, exp(x[1]) - sqrt(x[2]))
+exp(x[1]) - sqrt(x[2])</code></pre><p>To modify a nonlinear objective, call <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> again.</p><h2 id="Add-a-nonlinear-constraint"><a class="docs-heading-anchor" href="#Add-a-nonlinear-constraint">Add a nonlinear constraint</a><a id="Add-a-nonlinear-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-nonlinear-constraint" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> to add a nonlinear constraint.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @constraint(model, exp(x[1]) &lt;= 1)
+exp(x[1]) - 1.0 ≤ 0
+
+julia&gt; @constraint(model, con[i = 1:2], 2^x[i] &gt;= i)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarNonlinearFunction, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:
+ con[1] : (2.0 ^ x[1]) - 1.0 ≥ 0
+ con[2] : (2.0 ^ x[2]) - 2.0 ≥ 0</code></pre><p>Delete a nonlinear constraint using <a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; delete(model, con[1])</code></pre><h2 id="Create-a-nonlinear-expression"><a class="docs-heading-anchor" href="#Create-a-nonlinear-expression">Create a nonlinear expression</a><a id="Create-a-nonlinear-expression-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-nonlinear-expression" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> to create nonlinear expression objects:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; expr = @expression(model, exp(x[1]) + sqrt(x[2]))
+exp(x[1]) + sqrt(x[2])
+
+julia&gt; my_anon_expr = @expression(model, [i = 1:2], sin(x[i]))
+2-element Vector{NonlinearExpr}:
+ sin(x[1])
+ sin(x[2])
+
+julia&gt; @expression(model, my_expr[i = 1:2], sin(x[i]))
+2-element Vector{NonlinearExpr}:
+ sin(x[1])
+ sin(x[2])</code></pre><p>A <a href="../../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a> can be used in <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a>, <a href="../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>, and even nested in other <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a>s.</p><pre><code class="language-julia-repl hljs">julia&gt; @objective(model, Min, expr^2 + 1)
+((exp(x[1]) + sqrt(x[2])) ^ 2.0) + 1.0
+
+julia&gt; @constraint(model, [i = 1:2], my_expr[i] &lt;= i)
+2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarNonlinearFunction, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ sin(x[1]) - 1.0 ≤ 0
+ sin(x[2]) - 2.0 ≤ 0
+
+julia&gt; @expression(model, nested[i = 1:2], sin(my_expr[i]))
+2-element Vector{NonlinearExpr}:
+ sin(sin(x[1]))
+ sin(sin(x[2]))</code></pre><p>Use <a href="../../api/JuMP/#JuMP.value"><code>value</code></a> to query the value of a nonlinear expression:</p><pre><code class="language-julia-repl hljs">julia&gt; set_start_value(x[1], 1.0)
+
+julia&gt; value(start_value, nested[1])
+0.7456241416655579
+
+julia&gt; sin(sin(1.0))
+0.7456241416655579</code></pre><h2 id="Nonlinear-expressions-in-detail"><a class="docs-heading-anchor" href="#Nonlinear-expressions-in-detail">Nonlinear expressions in detail</a><a id="Nonlinear-expressions-in-detail-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear-expressions-in-detail" title="Permalink"></a></h2><p>Nonlinear expressions in JuMP are represented by a <a href="../../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a> object.</p><h3 id="Constructors"><a class="docs-heading-anchor" href="#Constructors">Constructors</a><a id="Constructors-1"></a><a class="docs-heading-anchor-permalink" href="#Constructors" title="Permalink"></a></h3><p>Nonlinear expressions can be created using the <a href="../../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a> constructors:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr = NonlinearExpr(:sin, Any[x])
+sin(x)</code></pre><p>or via operator overloading:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr = sin(x)
+sin(x)</code></pre><h3 id="Supported-arguments"><a class="docs-heading-anchor" href="#Supported-arguments">Supported arguments</a><a id="Supported-arguments-1"></a><a class="docs-heading-anchor-permalink" href="#Supported-arguments" title="Permalink"></a></h3><p>Nonlinear expressions can contain a mix of numbers, <a href="../../api/JuMP/#JuMP.AffExpr"><code>AffExpr</code></a>, <a href="../../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a>, and other <a href="../../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; aff = x + 1;
+
+julia&gt; quad = x^2 + x;
+
+julia&gt; expr = cos(x) * sin(quad) + aff
+(cos(x) * sin(x² + x)) + (x + 1)</code></pre><h3 id="Supported-operators"><a class="docs-heading-anchor" href="#Supported-operators">Supported operators</a><a id="Supported-operators-1"></a><a class="docs-heading-anchor-permalink" href="#Supported-operators" title="Permalink"></a></h3><p>The list of supported operators may vary between solvers. Given an optimizer, query the list of supported operators using <a href="../../moi/reference/models/#MathOptInterface.ListOfSupportedNonlinearOperators"><code>MOI.ListOfSupportedNonlinearOperators</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.get(Ipopt.Optimizer(), MOI.ListOfSupportedNonlinearOperators())
+85-element Vector{Symbol}:
+ :+
+ :-
+ :abs
+ :sqrt
+ :cbrt
+ :abs2
+ :inv
+ :log
+ :log10
+ :log2
+ ⋮
+ :min
+ :max
+ :&amp;&amp;
+ :||
+ :&lt;=
+ :(==)
+ :&gt;=
+ :&lt;
+ :&gt;</code></pre><p>In some univariate cases, the operator is defined in <a href="https://github.com/JuliaMath/SpecialFunctions.jl"><code>SpecialFunctions.jl</code></a>. To use these functions, you must explicitly import <code>SpecialFunctions.jl</code></p><pre><code class="language-julia-repl hljs">julia&gt; import Ipopt
+
+julia&gt; op = MOI.get(Ipopt.Optimizer(), MOI.ListOfSupportedNonlinearOperators());
+
+julia&gt; :erfcx in op
+true
+
+julia&gt; :dawson in op
+true
+
+julia&gt; import SpecialFunctions
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @expression(model, SpecialFunctions.erfcx(x))
+erfcx(x)
+
+julia&gt; @expression(model, SpecialFunctions.dawson(x))
+dawson(x)</code></pre><h3 id="Limitations"><a class="docs-heading-anchor" href="#Limitations">Limitations</a><a id="Limitations-1"></a><a class="docs-heading-anchor-permalink" href="#Limitations" title="Permalink"></a></h3><p>Some nonlinear expressions cannot be created via operator overloading. For example, to minimize the likelihood of bugs in user-code, we have not overloaded comparisons such as <code>&lt;</code> and <code>&gt;=</code> between JuMP objects:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; x &lt; 1
+ERROR: Cannot evaluate `&lt;` between a variable and a number.
+[...]</code></pre><p>Instead, wrap the expression in the <a href="../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macro:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr = @expression(model, x &lt; 1)
+x &lt; 1</code></pre><p>For technical reasons, other operators that are not overloaded include <code>||</code>, <code>&amp;&amp;</code>, and <code>ifelse</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr = @expression(model, ifelse(x &lt; -1 || x &gt;= 1, x^2, 0.0))
+ifelse((x &lt; -1) || (x &gt;= 1), x², 0.0)</code></pre><p>As an alternative, use the <code>JuMP.op_</code> functions, which fallback to the various comparison and logical operators:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; expr = op_ifelse(
+           op_or(op_strictly_less_than(x, -1), op_greater_than_or_equal_to(x, 1)),
+           x^2,
+           0.0,
+       )
+ifelse((x &lt; -1) || (x &gt;= 1), x², 0.0)</code></pre><p>The available functions are:</p><table><tr><th style="text-align: left">JuMP function</th><th style="text-align: left">Julia function</th></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_ifelse"><code>op_ifelse</code></a></td><td style="text-align: left"><code>ifelse</code></td></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_and"><code>op_and</code></a></td><td style="text-align: left"><code>&amp;&amp;</code></td></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_or"><code>op_or</code></a></td><td style="text-align: left"><code>||</code></td></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_greater_than_or_equal_to"><code>op_greater_than_or_equal_to</code></a></td><td style="text-align: left"><code>&gt;=</code></td></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_less_than_or_equal_to"><code>op_less_than_or_equal_to</code></a></td><td style="text-align: left"><code>&lt;=</code></td></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_equal_to"><code>op_equal_to</code></a></td><td style="text-align: left"><code>==</code></td></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_strictly_greater_than"><code>op_strictly_greater_than</code></a></td><td style="text-align: left"><code>&gt;</code></td></tr><tr><td style="text-align: left"><a href="../../api/JuMP/#JuMP.op_strictly_less_than"><code>op_strictly_less_than</code></a></td><td style="text-align: left"><code>&lt;</code></td></tr></table><h3 id="Fields"><a class="docs-heading-anchor" href="#Fields">Fields</a><a id="Fields-1"></a><a class="docs-heading-anchor-permalink" href="#Fields" title="Permalink"></a></h3><p>Each <a href="../../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a> has two fields.</p><p>The <code>.head</code> field is a <code>Symbol</code> that represents the operator being called:</p><pre><code class="language-julia-repl hljs">julia&gt; expr.head
+:sin</code></pre><p>The <code>.args</code> field is a <code>Vector{Any}</code> containing the arguments to the operator:</p><pre><code class="language-julia-repl hljs">julia&gt; expr.args
+1-element Vector{Any}:
+ x</code></pre><h2 id="jump_user_defined_operators"><a class="docs-heading-anchor" href="#jump_user_defined_operators">User-defined operators</a><a id="jump_user_defined_operators-1"></a><a class="docs-heading-anchor-permalink" href="#jump_user_defined_operators" title="Permalink"></a></h2><p>In addition to a standard list of univariate and multivariate operators recognized by the <code>MOI.Nonlinear</code> submodule, JuMP supports <em>user-defined</em> Julia operators.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>User-defined operators must return a scalar output. For a work-around, see <a href="../../tutorials/nonlinear/tips_and_tricks/#User-defined-operators-with-vector-outputs">User-defined operators with vector outputs</a>.</p></div></div><h3 id="Add-an-operator"><a class="docs-heading-anchor" href="#Add-an-operator">Add an operator</a><a id="Add-an-operator-1"></a><a class="docs-heading-anchor-permalink" href="#Add-an-operator" title="Permalink"></a></h3><p>Add a user-defined operator using the <a href="../../api/JuMP/#JuMP.@operator"><code>@operator</code></a> macro:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; square(x) = x^2</code><code class="nohighlight hljs ansi" style="display:block;">square (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; f(x, y) = (x - 1)^2 + (y - 2)^2</code><code class="nohighlight hljs ansi" style="display:block;">f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @operator(model, op_square, 1, square)</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(square, :op_square)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @operator(model, op_f, 2, f)</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(f, :op_f)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2]);</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, op_f(x[1], op_square(x[2])))</code><code class="nohighlight hljs ansi" style="display:block;">op_f(x[1], op_square(x[2]))</code></pre><p>The arguments to <a href="../../api/JuMP/#JuMP.@operator"><code>@operator</code></a> are:</p><ol><li>The model to which the operator is added.</li><li>A Julia symbol object which serves as the name of the user-defined operator in JuMP expressions. This name must not be the same as that of the function.</li><li>The number of scalar input arguments that the function takes.</li><li>A Julia method which computes the function.</li></ol><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>User-defined operators cannot be deleted.</p></div></div><p>You can obtain a reference to the operator using the <code>model[:key]</code> syntax:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; square(x) = x^2</code><code class="nohighlight hljs ansi" style="display:block;">square (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @operator(model, op_square, 1, square)</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(square, :op_square)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; op_square_2 = model[:op_square]</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(square, :op_square)</code></pre><h3 id="Add-an-operator-without-macros"><a class="docs-heading-anchor" href="#Add-an-operator-without-macros">Add an operator without macros</a><a id="Add-an-operator-without-macros-1"></a><a class="docs-heading-anchor-permalink" href="#Add-an-operator-without-macros" title="Permalink"></a></h3><p>The <a href="../../api/JuMP/#JuMP.@operator"><code>@operator</code></a> macro is syntactic sugar for <a href="../../api/JuMP/#JuMP.add_nonlinear_operator"><code>add_nonlinear_operator</code></a>. Thus, the non-macro version of the preceding example is:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; square(x) = x^2</code><code class="nohighlight hljs ansi" style="display:block;">square (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; f(x, y) = (x - 1)^2 + (y - 2)^2</code><code class="nohighlight hljs ansi" style="display:block;">f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; op_square = add_nonlinear_operator(model, 1, square; name = :op_square)</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(square, :op_square)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model[:op_square] = op_square</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(square, :op_square)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; op_f = add_nonlinear_operator(model, 2, f; name = :op_f)</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(f, :op_f)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model[:op_f] = op_f</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(f, :op_f)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2]);</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, op_f(x[1], op_square(x[2])))</code><code class="nohighlight hljs ansi" style="display:block;">op_f(x[1], op_square(x[2]))</code></pre><h3 id="Operators-with-the-same-name-as-an-existing-function"><a class="docs-heading-anchor" href="#Operators-with-the-same-name-as-an-existing-function">Operators with the same name as an existing function</a><a id="Operators-with-the-same-name-as-an-existing-function-1"></a><a class="docs-heading-anchor-permalink" href="#Operators-with-the-same-name-as-an-existing-function" title="Permalink"></a></h3><p>A common error encountered is the following:</p><pre><code class="language-julia-repl hljs">julia&gt; using JuMP
+
+julia&gt; model = Model();
+
+julia&gt; f(x) = x^2
+f (generic function with 1 method)
+
+julia&gt; @operator(model, f, 1, f)
+ERROR: Unable to add the nonlinear operator `:f` with the same name as
+an existing function.
+[...]</code></pre><p>This error occurs because <code>@operator(model, f, 1, f)</code> is equivalent to:</p><pre><code class="language-julia hljs">julia&gt; f = add_nonlinear_operator(model, 1, f; name = :f)</code></pre><p>but <code>f</code> already exists as a Julia function.</p><p>If you evaluate the function without adding it as an operator, JuMP will trace the function using operator overloading:</p><pre><code class="language-julia-repl hljs">julia&gt; @variable(model, x);
+
+julia&gt; f(x)
+x²</code></pre><p>To force JuMP to treat <code>f</code> as a user-defined operator and not trace it, add the operator using <a href="../../api/JuMP/#JuMP.add_nonlinear_operator"><code>add_nonlinear_operator</code></a> and define a new method which manually creates a <a href="../../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; _ = add_nonlinear_operator(model, 1, f; name = :f)
+NonlinearOperator(f, :f)
+
+julia&gt; f(x::AbstractJuMPScalar) = NonlinearExpr(:f, Any[x])
+f (generic function with 2 methods)
+
+julia&gt; @expression(model, log(f(x)))
+log(f(x))</code></pre><h3 id="Gradients-and-Hessians"><a class="docs-heading-anchor" href="#Gradients-and-Hessians">Gradients and Hessians</a><a id="Gradients-and-Hessians-1"></a><a class="docs-heading-anchor-permalink" href="#Gradients-and-Hessians" title="Permalink"></a></h3><p>By default, JuMP will use automatic differentiation to compute the gradient and Hessian of user-defined operators. If your function is not amenable to automatic differentiation, or you can compute analytic derivatives, you may pass additional arguments to <a href="../../api/JuMP/#JuMP.@operator"><code>@operator</code></a> to compute the first- and second-derivatives.</p><h4 id="Univariate-functions"><a class="docs-heading-anchor" href="#Univariate-functions">Univariate functions</a><a id="Univariate-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Univariate-functions" title="Permalink"></a></h4><p>For univariate functions, a gradient function <code>∇f</code> returns a number that represents the first-order derivative. You may, in addition, pass a third function which returns a number representing the second-order derivative:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; f(x) = x^2</code><code class="nohighlight hljs ansi" style="display:block;">f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; ∇f(x) = 2x</code><code class="nohighlight hljs ansi" style="display:block;">∇f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; ∇²f(x) = 2</code><code class="nohighlight hljs ansi" style="display:block;">∇²f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @operator(model, op_f, 1, f, ∇f, ∇²f)  # Providing ∇²f is optional</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(f, :op_f)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, op_f(x))</code><code class="nohighlight hljs ansi" style="display:block;">op_f(x)</code></pre><h4 id="Multivariate-functions"><a class="docs-heading-anchor" href="#Multivariate-functions">Multivariate functions</a><a id="Multivariate-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Multivariate-functions" title="Permalink"></a></h4><p>For multivariate functions, the gradient function <code>∇f</code> must take an <code>AbstractVector</code> as the first argument that is filled in-place. The Hessian function, <code>∇²f</code>, must take an <code>AbstractMatrix</code> as the first argument, the lower-triangular of which is filled in-place:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; f(x...) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2</code><code class="nohighlight hljs ansi" style="display:block;">f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; function ∇f(g::AbstractVector{T}, x::T...) where {T}
+           g[1] = 400 * x[1]^3 - 400 * x[1] * x[2] + 2 * x[1] - 2
+           g[2] = 200 * (x[2] - x[1]^2)
+           return
+       end</code><code class="nohighlight hljs ansi" style="display:block;">∇f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; function ∇²f(H::AbstractMatrix{T}, x::T...) where {T}
+           H[1, 1] = 1200 * x[1]^2 - 400 * x[2] + 2
+           # H[1, 2] = -400 * x[1]  &lt;-- Not needed. Fill the lower-triangular only.
+           H[2, 1] = -400 * x[1]
+           H[2, 2] = 200.0
+           return
+       end</code><code class="nohighlight hljs ansi" style="display:block;">∇²f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @operator(model, rosenbrock, 2, f, ∇f, ∇²f)  # Providing ∇²f is optional</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(f, :rosenbrock)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2])</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, rosenbrock(x[1], x[2]))</code><code class="nohighlight hljs ansi" style="display:block;">rosenbrock(x[1], x[2])</code></pre><p>You may assume the Hessian matrix <code>H</code> is initialized with zeros, and because <code>H</code> is symmetric, you need only to fill in the non-zero lower-triangular terms. The matrix type passed in as <code>H</code> depends on the automatic differentiation system, so make sure the first argument to the Hessian function supports an <code>AbstractMatrix</code> (it may be something other than <code>Matrix{Float64}</code>). Moreover, you may assume only that <code>H</code> supports <code>size(H)</code> and <code>setindex!</code>. Finally, the matrix is treated as dense, so the performance will be poor on functions with high-dimensional input.</p><h3 id="User-defined-operators-with-vector-inputs"><a class="docs-heading-anchor" href="#User-defined-operators-with-vector-inputs">User-defined operators with vector inputs</a><a id="User-defined-operators-with-vector-inputs-1"></a><a class="docs-heading-anchor-permalink" href="#User-defined-operators-with-vector-inputs" title="Permalink"></a></h3><p>User-defined operators which take vectors as input arguments (for example, <code>f(x::Vector)</code>) are <em>not</em> supported. Instead, use Julia&#39;s splatting syntax to create a function with scalar arguments. For example, instead of:</p><pre><code class="language-julia hljs">f(x::Vector) = sum(x[i]^i for i in 1:length(x))</code></pre><p>define:</p><pre><code class="language-julia hljs">f(x...) = sum(x[i]^i for i in 1:length(x))</code></pre><p>Another approach is to define the splatted function as an anonymous function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:5])</code><code class="nohighlight hljs ansi" style="display:block;">5-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+ x[4]
+ x[5]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; f(x::Vector) = sum(x[i]^i for i in 1:length(x))</code><code class="nohighlight hljs ansi" style="display:block;">f (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @operator(model, op_f, 5, (x...) -&gt; f(collect(x)))</code><code class="nohighlight hljs ansi" style="display:block;">NonlinearOperator(#6, :op_f)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, op_f(x...))</code><code class="nohighlight hljs ansi" style="display:block;">op_f(x[1], x[2], x[3], x[4], x[5])</code></pre><h3 id="Automatic-differentiation"><a class="docs-heading-anchor" href="#Automatic-differentiation">Automatic differentiation</a><a id="Automatic-differentiation-1"></a><a class="docs-heading-anchor-permalink" href="#Automatic-differentiation" title="Permalink"></a></h3><p>JuMP does not support black-box optimization, so all user-defined operators must provide derivatives in some form. Fortunately, JuMP supports automatic differentiation of user-defined operators.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Automatic differentiation is <em>not</em> finite differencing. JuMP&#39;s automatically computed derivatives are not subject to approximation error.</p></div></div><p>JuMP uses <a href="https://github.com/JuliaDiff/ForwardDiff.jl">ForwardDiff.jl</a> to perform automatic differentiation of user-defined operators; see the ForwardDiff.jl <a href="https://www.juliadiff.org/ForwardDiff.jl/v0.10.2/user/limitations.html">documentation</a> for a description of how to write a function suitable for automatic differentiation.</p><h4 id="Common-mistakes-when-writing-a-user-defined-operator"><a class="docs-heading-anchor" href="#Common-mistakes-when-writing-a-user-defined-operator">Common mistakes when writing a user-defined operator</a><a id="Common-mistakes-when-writing-a-user-defined-operator-1"></a><a class="docs-heading-anchor-permalink" href="#Common-mistakes-when-writing-a-user-defined-operator" title="Permalink"></a></h4><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Get an error like <code>No method matching Float64(::ForwardDiff.Dual)</code>? Read this section, and see the guidelines at <a href="https://www.juliadiff.org/ForwardDiff.jl/release-0.10/user/limitations.html">ForwardDiff.jl</a>.</p></div></div><p>The most common error is that your user-defined operator is not generic with respect to the number type, that is, don&#39;t assume that the input to the function is <code>Float64</code>.</p><pre><code class="language-julia hljs">f(x::Float64) = 2 * x  # This will not work.
+f(x::Real)    = 2 * x  # This is good.
+f(x)          = 2 * x  # This is also good.</code></pre><p>Another reason you may encounter this error is if you create arrays inside your function which are <code>Float64</code>.</p><pre><code class="language-julia hljs">function bad_f(x...)
+    y = zeros(length(x))  # This constructs an array of `Float64`!
+    for i in 1:length(x)
+        y[i] = x[i]^i
+    end
+    return sum(y)
+end
+
+function good_f(x::T...) where {T&lt;:Real}
+    y = zeros(T, length(x))  # Construct an array of type `T` instead!
+    for i in 1:length(x)
+        y[i] = x[i]^i
+    end
+    return sum(y)
+end</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../complex/">« Complex number support</a><a class="docs-footer-nextpage" href="../nlp/">Nonlinear Modeling (Legacy) »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/objective/index.html b/previews/PR3525/manual/objective/index.html
new file mode 100644
index 00000000000..68fd27ff172
--- /dev/null
+++ b/previews/PR3525/manual/objective/index.html
@@ -0,0 +1,168 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Objectives · JuMP</title><meta name="title" content="Objectives · JuMP"/><meta property="og:title" content="Objectives · JuMP"/><meta property="twitter:title" content="Objectives · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li class="is-active"><a class="tocitem" href>Objectives</a><ul class="internal"><li><a class="tocitem" href="#Set-a-linear-objective"><span>Set a linear objective</span></a></li><li><a class="tocitem" href="#Set-a-quadratic-objective"><span>Set a quadratic objective</span></a></li><li><a class="tocitem" href="#Set-a-nonlinear-objective"><span>Set a nonlinear objective</span></a></li><li><a class="tocitem" href="#Query-the-objective-function"><span>Query the objective function</span></a></li><li><a class="tocitem" href="#Evaluate-the-objective-function-at-a-point"><span>Evaluate the objective function at a point</span></a></li><li><a class="tocitem" href="#Query-the-objective-sense"><span>Query the objective sense</span></a></li><li><a class="tocitem" href="#Modify-an-objective"><span>Modify an objective</span></a></li><li><a class="tocitem" href="#Modify-an-objective-coefficient"><span>Modify an objective coefficient</span></a></li><li><a class="tocitem" href="#Modify-the-objective-sense"><span>Modify the objective sense</span></a></li><li><a class="tocitem" href="#Set-a-vector-valued-objective"><span>Set a vector-valued objective</span></a></li></ul></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Objectives</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Objectives</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/objective.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Objectives"><a class="docs-heading-anchor" href="#Objectives">Objectives</a><a id="Objectives-1"></a><a class="docs-heading-anchor-permalink" href="#Objectives" title="Permalink"></a></h1><p>This page describes macros and functions related to linear and quadratic objective functions only, unless otherwise indicated. For nonlinear objective functions, see <a href="../nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a>.</p><h2 id="Set-a-linear-objective"><a class="docs-heading-anchor" href="#Set-a-linear-objective">Set a linear objective</a><a id="Set-a-linear-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Set-a-linear-objective" title="Permalink"></a></h2><p>Use the <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> macro to set a linear objective function.</p><p>Use <code>Min</code> to create a minimization objective:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x + 1)
+2 x + 1</code></pre><p>Use <code>Max</code> to create a maximization objective:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Max, 2x + 1)
+2 x + 1</code></pre><h2 id="Set-a-quadratic-objective"><a class="docs-heading-anchor" href="#Set-a-quadratic-objective">Set a quadratic objective</a><a id="Set-a-quadratic-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Set-a-quadratic-objective" title="Permalink"></a></h2><p>Use the <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> macro to set a quadratic objective function.</p><p>Use <code>^2</code> to have a variable squared:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, x^2 + 2x + 1)
+x² + 2 x + 1</code></pre><p>You can also have bilinear terms between variables:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, y)
+y
+
+julia&gt; @objective(model, Max, x * y + x + y)
+x*y + x + y</code></pre><h2 id="Set-a-nonlinear-objective"><a class="docs-heading-anchor" href="#Set-a-nonlinear-objective">Set a nonlinear objective</a><a id="Set-a-nonlinear-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Set-a-nonlinear-objective" title="Permalink"></a></h2><p>Use the <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> macro to set a nonlinear objective function:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &lt;= 1);
+
+julia&gt; @objective(model, Max, log(x))
+log(x)</code></pre><h2 id="Query-the-objective-function"><a class="docs-heading-anchor" href="#Query-the-objective-function">Query the objective function</a><a id="Query-the-objective-function-1"></a><a class="docs-heading-anchor-permalink" href="#Query-the-objective-function" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.objective_function"><code>objective_function</code></a> to return the current objective function.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x + 1)
+2 x + 1
+
+julia&gt; objective_function(model)
+2 x + 1</code></pre><h2 id="Evaluate-the-objective-function-at-a-point"><a class="docs-heading-anchor" href="#Evaluate-the-objective-function-at-a-point">Evaluate the objective function at a point</a><a id="Evaluate-the-objective-function-at-a-point-1"></a><a class="docs-heading-anchor-permalink" href="#Evaluate-the-objective-function-at-a-point" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.value"><code>value</code></a> to evaluate an objective function at a point specifying values for variables.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @objective(model, Min, 2x[1]^2 + x[1] + 0.5*x[2])
+2 x[1]² + x[1] + 0.5 x[2]
+
+julia&gt; f = objective_function(model)
+2 x[1]² + x[1] + 0.5 x[2]
+
+julia&gt; point = Dict(x[1] =&gt; 2.0, x[2] =&gt; 1.0);
+
+julia&gt; value(z -&gt; point[z], f)
+10.5</code></pre><h2 id="Query-the-objective-sense"><a class="docs-heading-anchor" href="#Query-the-objective-sense">Query the objective sense</a><a id="Query-the-objective-sense-1"></a><a class="docs-heading-anchor-permalink" href="#Query-the-objective-sense" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.objective_sense"><code>objective_sense</code></a> to return the current objective sense.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x + 1)
+2 x + 1
+
+julia&gt; objective_sense(model)
+MIN_SENSE::OptimizationSense = 0</code></pre><h2 id="Modify-an-objective"><a class="docs-heading-anchor" href="#Modify-an-objective">Modify an objective</a><a id="Modify-an-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-an-objective" title="Permalink"></a></h2><p>To modify an objective, call <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> with the new objective function.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x)
+2 x
+
+julia&gt; @objective(model, Max, -2x)
+-2 x</code></pre><p>Alternatively, use <a href="../../api/JuMP/#JuMP.set_objective_function"><code>set_objective_function</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x)
+2 x
+
+julia&gt; new_objective = @expression(model, -2 * x)
+-2 x
+
+julia&gt; set_objective_function(model, new_objective)</code></pre><h2 id="Modify-an-objective-coefficient"><a class="docs-heading-anchor" href="#Modify-an-objective-coefficient">Modify an objective coefficient</a><a id="Modify-an-objective-coefficient-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-an-objective-coefficient" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.set_objective_coefficient"><code>set_objective_coefficient</code></a> to modify an objective coefficient.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x)
+2 x
+
+julia&gt; set_objective_coefficient(model, x, 3)
+
+julia&gt; objective_function(model)
+3 x</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>There is no way to modify the coefficient of a quadratic term. Set a new objective instead.</p></div></div><h2 id="Modify-the-objective-sense"><a class="docs-heading-anchor" href="#Modify-the-objective-sense">Modify the objective sense</a><a id="Modify-the-objective-sense-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-the-objective-sense" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.set_objective_sense"><code>set_objective_sense</code></a> to modify the objective sense.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x)
+2 x
+
+julia&gt; objective_sense(model)
+MIN_SENSE::OptimizationSense = 0
+
+julia&gt; set_objective_sense(model, MAX_SENSE);
+
+julia&gt; objective_sense(model)
+MAX_SENSE::OptimizationSense = 1</code></pre><p>Alternatively, call <a href="../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> and pass the existing objective function.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @objective(model, Min, 2x)
+2 x
+
+julia&gt; @objective(model, Max, objective_function(model))
+2 x</code></pre><h2 id="Set-a-vector-valued-objective"><a class="docs-heading-anchor" href="#Set-a-vector-valued-objective">Set a vector-valued objective</a><a id="Set-a-vector-valued-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Set-a-vector-valued-objective" title="Permalink"></a></h2><p>Define a multi-objective optimization problem by passing a vector of objectives:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @objective(model, Min, [1 + x[1], 2 * x[2]])
+2-element Vector{AffExpr}:
+ x[1] + 1
+ 2 x[2]
+
+julia&gt; f = objective_function(model)
+2-element Vector{AffExpr}:
+ x[1] + 1
+ 2 x[2]</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>The <a href="../../tutorials/linear/multi_objective_knapsack/#Multi-objective-knapsack">Multi-objective knapsack</a> tutorial provides an example of solving a multi-objective integer program.</p></div></div><p>In most cases, multi-objective optimization solvers will return multiple solutions, corresponding to points on the Pareto frontier. See <a href="../solutions/#Multiple-solutions">Multiple solutions</a> for information on how to query and work with multiple solutions.</p><p>Note that you must set a single objective sense, that is, you cannot have both minimization and maximization objectives. Work around this limitation by choosing <code>Min</code> and negating any objectives you want to maximize:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @expression(model, obj1, 1 + x[1])
+x[1] + 1
+
+julia&gt; @expression(model, obj2, 2 * x[1])
+2 x[1]
+
+julia&gt; @objective(model, Min, [obj1, -obj2])
+2-element Vector{AffExpr}:
+ x[1] + 1
+ -2 x[1]</code></pre><p>Defining your objectives as expressions allows flexibility in how you can solve variations of the same problem, with some objectives removed and constrained to be no worse that a fixed value.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2]);
+
+julia&gt; @expression(model, obj1, 1 + x[1])
+x[1] + 1
+
+julia&gt; @expression(model, obj2, 2 * x[1])
+2 x[1]
+
+julia&gt; @expression(model, obj3, x[1] + x[2])
+x[1] + x[2]
+
+julia&gt; @objective(model, Min, [obj1, obj2, obj3])  # Three-objective problem
+3-element Vector{AffExpr}:
+ x[1] + 1
+ 2 x[1]
+ x[1] + x[2]
+
+julia&gt; # optimize!(model), look at the solution, talk to stakeholders, then
+       # decide you want to solve a new problem where the third objective is
+       # removed and constrained to be better than 2.0.
+       nothing
+
+julia&gt; @objective(model, Min, [obj1, obj2])   # Two-objective problem
+2-element Vector{AffExpr}:
+ x[1] + 1
+ 2 x[1]
+
+julia&gt; @constraint(model, obj3 &lt;= 2.0)
+x[1] + x[2] ≤ 2</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../expressions/">« Expressions</a><a class="docs-footer-nextpage" href="../containers/">Containers »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/solutions/index.html b/previews/PR3525/manual/solutions/index.html
new file mode 100644
index 00000000000..31cd798174e
--- /dev/null
+++ b/previews/PR3525/manual/solutions/index.html
@@ -0,0 +1,374 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Solutions · JuMP</title><meta name="title" content="Solutions · JuMP"/><meta property="og:title" content="Solutions · JuMP"/><meta property="twitter:title" content="Solutions · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li class="is-active"><a class="tocitem" href>Solutions</a><ul class="internal"><li><a class="tocitem" href="#Solutions-summary"><span>Solutions summary</span></a></li><li><a class="tocitem" href="#Why-did-the-solver-stop?"><span>Why did the solver stop?</span></a></li><li><a class="tocitem" href="#Primal-solutions"><span>Primal solutions</span></a></li><li><a class="tocitem" href="#Dual-solutions"><span>Dual solutions</span></a></li><li><a class="tocitem" href="#Recommended-workflow"><span>Recommended workflow</span></a></li><li><a class="tocitem" href="#OptimizeNotCalled-errors"><span>OptimizeNotCalled errors</span></a></li><li><a class="tocitem" href="#Accessing-attributes"><span>Accessing attributes</span></a></li><li><a class="tocitem" href="#Sensitivity-analysis-for-LP"><span>Sensitivity analysis for LP</span></a></li><li><a class="tocitem" href="#Conflicts"><span>Conflicts</span></a></li><li><a class="tocitem" href="#Multiple-solutions"><span>Multiple solutions</span></a></li><li><a class="tocitem" href="#Checking-feasibility-of-solutions"><span>Checking feasibility of solutions</span></a></li></ul></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Solutions</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Solutions</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/solutions.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="jump_solutions"><a class="docs-heading-anchor" href="#jump_solutions">Solutions</a><a id="jump_solutions-1"></a><a class="docs-heading-anchor-permalink" href="#jump_solutions" title="Permalink"></a></h1><p>This section of the manual describes how to access a solved solution to a problem. It uses the following model as an example:</p><pre><code class="language-julia-repl hljs">julia&gt; begin
+           model = Model(HiGHS.Optimizer)
+           set_silent(model)
+           @variable(model, x &gt;= 0)
+           @variable(model, y[[:a, :b]] &lt;= 1)
+           @objective(model, Max, -12x - 20y[:a])
+           @expression(model, my_expr, 6x + 8y[:a])
+           @constraint(model, my_expr &gt;= 100)
+           @constraint(model, c1, 7x + 12y[:a] &gt;= 120)
+           optimize!(model)
+           print(model)
+       end
+Max -12 x - 20 y[a]
+Subject to
+ 6 x + 8 y[a] ≥ 100
+ c1 : 7 x + 12 y[a] ≥ 120
+ x ≥ 0
+ y[a] ≤ 1
+ y[b] ≤ 1</code></pre><h2 id="Solutions-summary"><a class="docs-heading-anchor" href="#Solutions-summary">Solutions summary</a><a id="Solutions-summary-1"></a><a class="docs-heading-anchor-permalink" href="#Solutions-summary" title="Permalink"></a></h2><p><a href="../../api/JuMP/#JuMP.solution_summary"><code>solution_summary</code></a> can be used for checking the summary of the optimization solutions.</p><pre><code class="language-julia-repl hljs">julia&gt; solution_summary(model)
+* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : -2.05143e+02
+  Objective bound    : -0.00000e+00
+  Relative gap       : Inf
+  Dual objective value : -2.05143e+02
+
+* Work counters
+  Solve time (sec)   : 6.70987e-04
+  Simplex iterations : 2
+  Barrier iterations : 0
+  Node count         : -1
+
+julia&gt; solution_summary(model; verbose = true)
+* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : -2.05143e+02
+  Objective bound    : -0.00000e+00
+  Relative gap       : Inf
+  Dual objective value : -2.05143e+02
+  Primal solution :
+    x : 1.54286e+01
+    y[a] : 1.00000e+00
+    y[b] : 1.00000e+00
+  Dual solution :
+    c1 : 1.71429e+00
+
+* Work counters
+  Solve time (sec)   : 6.70987e-04
+  Simplex iterations : 2
+  Barrier iterations : 0
+  Node count         : -1</code></pre><h2 id="Why-did-the-solver-stop?"><a class="docs-heading-anchor" href="#Why-did-the-solver-stop?">Why did the solver stop?</a><a id="Why-did-the-solver-stop?-1"></a><a class="docs-heading-anchor-permalink" href="#Why-did-the-solver-stop?" title="Permalink"></a></h2><p>Use<a href="../../api/JuMP/#JuMP.termination_status"><code>termination_status</code></a> to understand why the solver stopped.</p><pre><code class="language-julia-repl hljs">julia&gt; termination_status(model)
+OPTIMAL::TerminationStatusCode = 1</code></pre><p>The <a href="../../moi/reference/models/#MathOptInterface.TerminationStatusCode"><code>MOI.TerminationStatusCode</code></a> enum describes the full list of statuses that could be returned.</p><p>Common return values include <code>OPTIMAL</code>, <code>LOCALLY_SOLVED</code>, <code>INFEASIBLE</code>, <code>DUAL_INFEASIBLE</code>, and <code>TIME_LIMIT</code>.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>A return status of <code>OPTIMAL</code> means the solver found (and proved) a globally optimal solution. A return status of <code>LOCALLY_SOLVED</code> means the solver found a locally optimal solution (which may also be globally optimal, but it could not prove so).</p></div></div><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>A return status of <code>DUAL_INFEASIBLE</code> does not guarantee that the primal is unbounded. When the dual is infeasible, the primal is unbounded if there exists a feasible primal solution.</p></div></div><p>Use <a href="../../api/JuMP/#JuMP.raw_status"><code>raw_status</code></a> to get a solver-specific string explaining why the optimization stopped:</p><pre><code class="language-julia-repl hljs">julia&gt; raw_status(model)
+&quot;kHighsModelStatusOptimal&quot;</code></pre><h2 id="Primal-solutions"><a class="docs-heading-anchor" href="#Primal-solutions">Primal solutions</a><a id="Primal-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Primal-solutions" title="Permalink"></a></h2><h3 id="Primal-solution-status"><a class="docs-heading-anchor" href="#Primal-solution-status">Primal solution status</a><a id="Primal-solution-status-1"></a><a class="docs-heading-anchor-permalink" href="#Primal-solution-status" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.primal_status"><code>primal_status</code></a> to return an <a href="../../moi/reference/models/#MathOptInterface.ResultStatusCode"><code>MOI.ResultStatusCode</code></a> enum describing the status of the primal solution.</p><pre><code class="language-julia-repl hljs">julia&gt; primal_status(model)
+FEASIBLE_POINT::ResultStatusCode = 1</code></pre><p>Other common returns are <code>NO_SOLUTION</code>, and <code>INFEASIBILITY_CERTIFICATE</code>. The first means that the solver doesn&#39;t have a solution to return, and the second means that the primal solution is a certificate of dual infeasibility (a primal unbounded ray).</p><p>You can also use <a href="../../api/JuMP/#JuMP.has_values"><code>has_values</code></a>, which returns <code>true</code> if there is a solution that can be queried, and <code>false</code> otherwise.</p><pre><code class="language-julia-repl hljs">julia&gt; has_values(model)
+true</code></pre><h3 id="Objective-values"><a class="docs-heading-anchor" href="#Objective-values">Objective values</a><a id="Objective-values-1"></a><a class="docs-heading-anchor-permalink" href="#Objective-values" title="Permalink"></a></h3><p>The objective value of a solved problem can be obtained via <a href="../../api/JuMP/#JuMP.objective_value"><code>objective_value</code></a>. The best known bound on the optimal objective value can be obtained via <a href="../../api/JuMP/#JuMP.objective_bound"><code>objective_bound</code></a>. If the solver supports it, the value of the dual objective can be obtained via <a href="../../api/JuMP/#JuMP.dual_objective_value"><code>dual_objective_value</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; objective_value(model)
+-205.14285714285714
+
+julia&gt; objective_bound(model)  # HiGHS only implements objective bound for MIPs
+-0.0
+
+julia&gt; dual_objective_value(model)
+-205.1428571428571</code></pre><h3 id="Primal-solution-values"><a class="docs-heading-anchor" href="#Primal-solution-values">Primal solution values</a><a id="Primal-solution-values-1"></a><a class="docs-heading-anchor-permalink" href="#Primal-solution-values" title="Permalink"></a></h3><p>If the solver has a primal solution to return, use <a href="../../api/JuMP/#JuMP.value"><code>value</code></a> to access it:</p><pre><code class="language-julia-repl hljs">julia&gt; value(x)
+15.428571428571429</code></pre><p>Broadcast <a href="../../api/JuMP/#JuMP.value"><code>value</code></a> over containers:</p><pre><code class="language-julia-repl hljs">julia&gt; value.(y)
+1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [:a, :b]
+And data, a 2-element Vector{Float64}:
+ 1.0
+ 1.0</code></pre><p><a href="../../api/JuMP/#JuMP.value"><code>value</code></a> also works on expressions:</p><pre><code class="language-julia-repl hljs">julia&gt; value(my_expr)
+100.57142857142857</code></pre><p>and constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; value(c1)
+120.0</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Calling <a href="../../api/JuMP/#JuMP.value"><code>value</code></a> on a constraint returns the constraint function evaluated at the solution.</p></div></div><h2 id="Dual-solutions"><a class="docs-heading-anchor" href="#Dual-solutions">Dual solutions</a><a id="Dual-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Dual-solutions" title="Permalink"></a></h2><h3 id="Dual-solution-status"><a class="docs-heading-anchor" href="#Dual-solution-status">Dual solution status</a><a id="Dual-solution-status-1"></a><a class="docs-heading-anchor-permalink" href="#Dual-solution-status" title="Permalink"></a></h3><p>Use <a href="../../api/JuMP/#JuMP.dual_status"><code>dual_status</code></a> to return an <a href="../../moi/reference/models/#MathOptInterface.ResultStatusCode"><code>MOI.ResultStatusCode</code></a> enum describing the status of the dual solution.</p><pre><code class="language-julia-repl hljs">julia&gt; dual_status(model)
+FEASIBLE_POINT::ResultStatusCode = 1</code></pre><p>Other common returns are <code>NO_SOLUTION</code>, and <code>INFEASIBILITY_CERTIFICATE</code>. The first means that the solver doesn&#39;t have a solution to return, and the second means that the dual solution is a certificate of primal infeasibility (a dual unbounded ray).</p><p>You can also use <a href="../../api/JuMP/#JuMP.has_duals"><code>has_duals</code></a>, which returns <code>true</code> if there is a solution that can be queried, and <code>false</code> otherwise.</p><pre><code class="language-julia-repl hljs">julia&gt; has_duals(model)
+true</code></pre><h3 id="Dual-solution-values"><a class="docs-heading-anchor" href="#Dual-solution-values">Dual solution values</a><a id="Dual-solution-values-1"></a><a class="docs-heading-anchor-permalink" href="#Dual-solution-values" title="Permalink"></a></h3><p>If the solver has a dual solution to return, use <a href="../../api/JuMP/#JuMP.dual"><code>dual</code></a> to access it:</p><pre><code class="language-julia-repl hljs">julia&gt; dual(c1)
+1.7142857142857142</code></pre><p>Query the duals of variable bounds using <a href="../../api/JuMP/#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, <a href="../../api/JuMP/#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, and <a href="../../api/JuMP/#JuMP.FixRef"><code>FixRef</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; dual(LowerBoundRef(x))
+0.0
+
+julia&gt; dual.(UpperBoundRef.(y))
+1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [:a, :b]
+And data, a 2-element Vector{Float64}:
+ -0.5714285714285694
+  0.0</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>JuMP&#39;s definition of duality is independent of the objective sense. That is, the sign of feasible duals associated with a constraint depends on the direction of the constraint and not whether the problem is maximization or minimization. <strong>This is a different convention from linear programming duality in some common textbooks.</strong> If you have a linear program, and you want the textbook definition, you probably want to use <a href="../../api/JuMP/#JuMP.shadow_price"><code>shadow_price</code></a> and <a href="../../api/JuMP/#JuMP.reduced_cost"><code>reduced_cost</code></a> instead.</p></div></div><pre><code class="language-julia-repl hljs">julia&gt; shadow_price(c1)
+1.7142857142857142
+
+julia&gt; reduced_cost(x)
+-0.0
+
+julia&gt; reduced_cost.(y)
+1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [:a, :b]
+And data, a 2-element Vector{Float64}:
+  0.5714285714285694
+ -0.0</code></pre><h2 id="Recommended-workflow"><a class="docs-heading-anchor" href="#Recommended-workflow">Recommended workflow</a><a id="Recommended-workflow-1"></a><a class="docs-heading-anchor-permalink" href="#Recommended-workflow" title="Permalink"></a></h2><p>The recommended workflow for solving a model and querying the solution is something like the following:</p><pre><code class="language-julia-repl hljs">julia&gt; begin
+           if termination_status(model) == OPTIMAL
+               println(&quot;Solution is optimal&quot;)
+           elseif termination_status(model) == TIME_LIMIT &amp;&amp; has_values(model)
+               println(&quot;Solution is suboptimal due to a time limit, but a primal solution is available&quot;)
+           else
+               error(&quot;The model was not solved correctly.&quot;)
+           end
+           println(&quot;  objective value = &quot;, objective_value(model))
+           if primal_status(model) == FEASIBLE_POINT
+               println(&quot;  primal solution: x = &quot;, value(x))
+           end
+           if dual_status(model) == FEASIBLE_POINT
+               println(&quot;  dual solution: c1 = &quot;, dual(c1))
+           end
+       end
+Solution is optimal
+  objective value = -205.14285714285714
+  primal solution: x = 15.428571428571429
+  dual solution: c1 = 1.7142857142857142</code></pre><h2 id="OptimizeNotCalled-errors"><a class="docs-heading-anchor" href="#OptimizeNotCalled-errors">OptimizeNotCalled errors</a><a id="OptimizeNotCalled-errors-1"></a><a class="docs-heading-anchor-permalink" href="#OptimizeNotCalled-errors" title="Permalink"></a></h2><p>Due to differences in how solvers cache solutions internally, modifying a model after calling <a href="../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a> will reset the model into the <code>MOI.OPTIMIZE_NOT_CALLED</code> state. If you then attempt to query solution information, an <code>OptimizeNotCalled</code> error will be thrown.</p><p>If you are iteratively querying solution information and modifying a model, query all the results first, then modify the problem.</p><p>For example, instead of:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 0);
+
+julia&gt; optimize!(model)
+
+julia&gt; termination_status(model)
+OPTIMAL::TerminationStatusCode = 1
+
+julia&gt; set_upper_bound(x, 1)
+
+julia&gt; x_val = value(x)
+┌ Warning: The model has been modified since the last call to `optimize!` (or `optimize!` has not been called yet). If you are iteratively querying solution information and modifying a model, query all the results first, then modify the model.
+└ @ JuMP ~/work/JuMP.jl/JuMP.jl/src/optimizer_interface.jl:712
+ERROR: OptimizeNotCalled()
+Stacktrace:
+[...]
+
+julia&gt; termination_status(model)
+OPTIMIZE_NOT_CALLED::TerminationStatusCode = 0</code></pre><p>do</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 0);
+
+julia&gt; optimize!(model);
+
+julia&gt; x_val = value(x)
+0.0
+
+julia&gt; termination_status(model)
+OPTIMAL::TerminationStatusCode = 1
+
+julia&gt; set_upper_bound(x, 1)
+
+julia&gt; set_lower_bound(x, x_val)
+
+julia&gt; termination_status(model)
+OPTIMIZE_NOT_CALLED::TerminationStatusCode = 0</code></pre><p>If you know that your particular solver supports querying solution information after modifications, you can use <a href="../../api/JuMP/#JuMP.direct_model"><code>direct_model</code></a> to bypass the <code>MOI.OPTIMIZE_NOT_CALLED</code> state:</p><pre><code class="language-julia-repl hljs">julia&gt; model = direct_model(HiGHS.Optimizer());
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 0);
+
+julia&gt; optimize!(model)
+
+julia&gt; termination_status(model)
+OPTIMAL::TerminationStatusCode = 1
+
+julia&gt; set_upper_bound(x, 1)
+
+julia&gt; x_val = value(x)
+0.0
+
+julia&gt; set_lower_bound(x, x_val)
+
+julia&gt; termination_status(model)
+OPTIMAL::TerminationStatusCode = 1</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Be careful doing this. If your particular solver does not support querying solution information after modification, it may silently return incorrect solutions or throw an error.</p></div></div><h2 id="Accessing-attributes"><a class="docs-heading-anchor" href="#Accessing-attributes">Accessing attributes</a><a id="Accessing-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Accessing-attributes" title="Permalink"></a></h2><p><a href="../../moi/#moi_documentation">MathOptInterface</a> defines many model attributes that can be queried. Some attributes can be directly accessed by getter functions. These include:</p><ul><li><a href="../../api/JuMP/#JuMP.solve_time"><code>solve_time</code></a></li><li><a href="../../api/JuMP/#JuMP.relative_gap"><code>relative_gap</code></a></li><li><a href="../../api/JuMP/#JuMP.simplex_iterations"><code>simplex_iterations</code></a></li><li><a href="../../api/JuMP/#JuMP.barrier_iterations"><code>barrier_iterations</code></a></li><li><a href="../../api/JuMP/#JuMP.node_count"><code>node_count</code></a></li></ul><h2 id="Sensitivity-analysis-for-LP"><a class="docs-heading-anchor" href="#Sensitivity-analysis-for-LP">Sensitivity analysis for LP</a><a id="Sensitivity-analysis-for-LP-1"></a><a class="docs-heading-anchor-permalink" href="#Sensitivity-analysis-for-LP" title="Permalink"></a></h2><p>Given an LP problem and an optimal solution corresponding to a basis, we can question how much an objective coefficient or standard form right-hand side coefficient (c.f., <a href="../../api/JuMP/#JuMP.normalized_rhs"><code>normalized_rhs</code></a>) can change without violating primal or dual feasibility of the basic solution.</p><p>Note that not all solvers compute the basis, and for sensitivity analysis, the solver interface must implement <code>MOI.ConstraintBasisStatus</code>.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Read the <a href="../../tutorials/linear/lp_sensitivity/#Sensitivity-analysis-of-a-linear-program">Sensitivity analysis of a linear program</a> for more information on sensitivity analysis.</p></div></div><p>To give a simple example, we could analyze the sensitivity of the optimal solution to the following (non-degenerate) LP problem:</p><pre><code class="language-julia-repl hljs">julia&gt; begin
+           model = Model(HiGHS.Optimizer)
+           set_silent(model)
+           @variable(model, x[1:2])
+           set_lower_bound(x[2], -0.5)
+           set_upper_bound(x[2], 0.5)
+           @constraint(model, c1, x[1] + x[2] &lt;= 1)
+           @constraint(model, c2, x[1] - x[2] &lt;= 1)
+           @objective(model, Max, x[1])
+           print(model)
+       end
+Max x[1]
+Subject to
+ c1 : x[1] + x[2] ≤ 1
+ c2 : x[1] - x[2] ≤ 1
+ x[2] ≥ -0.5
+ x[2] ≤ 0.5</code></pre><p>To analyze the sensitivity of the problem we could check the allowed perturbation ranges of, for example, the cost coefficients and the right-hand side coefficient of the constraint <code>c1</code> as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; optimize!(model)
+
+julia&gt; value.(x)
+2-element Vector{Float64}:
+  1.0
+ -0.0
+
+julia&gt; report = lp_sensitivity_report(model);
+
+julia&gt; x1_lo, x1_hi = report[x[1]]
+(-1.0, Inf)
+
+julia&gt; println(&quot;The objective coefficient of x[1] could decrease by $(x1_lo) or increase by $(x1_hi).&quot;)
+The objective coefficient of x[1] could decrease by -1.0 or increase by Inf.
+
+julia&gt; x2_lo, x2_hi = report[x[2]]
+(-1.0, 1.0)
+
+julia&gt; println(&quot;The objective coefficient of x[2] could decrease by $(x2_lo) or increase by $(x2_hi).&quot;)
+The objective coefficient of x[2] could decrease by -1.0 or increase by 1.0.
+
+julia&gt; c_lo, c_hi = report[c1]
+(-1.0, 1.0)
+
+julia&gt; println(&quot;The RHS of c1 could decrease by $(c_lo) or increase by $(c_hi).&quot;)
+The RHS of c1 could decrease by -1.0 or increase by 1.0.</code></pre><p>The range associated with a variable is the range of the allowed perturbation of the corresponding objective coefficient. Note that the current primal solution remains optimal within this range; however the corresponding dual solution might change since a cost coefficient is perturbed. Similarly, the range associated with a constraint is the range of the allowed perturbation of the corresponding right-hand side coefficient. In this range the current dual solution remains optimal, but the optimal primal solution might change.</p><p>If the problem is degenerate, there are multiple optimal bases and hence these ranges might not be as intuitive and seem too narrow, for example, a larger cost coefficient perturbation might not invalidate the optimality of the current primal solution. Moreover, if a problem is degenerate, due to finite precision, it can happen that, for example, a perturbation seems to invalidate a basis even though it doesn&#39;t (again providing too narrow ranges). To prevent this, increase the <code>atol</code> keyword argument to <a href="../../api/JuMP/#JuMP.lp_sensitivity_report"><code>lp_sensitivity_report</code></a>. Note that this might make the ranges too wide for numerically challenging instances. Thus, do not blindly trust these ranges, especially not for highly degenerate or numerically unstable instances.</p><h2 id="Conflicts"><a class="docs-heading-anchor" href="#Conflicts">Conflicts</a><a id="Conflicts-1"></a><a class="docs-heading-anchor-permalink" href="#Conflicts" title="Permalink"></a></h2><p>When the model you input is infeasible, some solvers can help you find the cause of this infeasibility by offering a conflict, that is, a subset of the constraints that create this infeasibility. Depending on the solver, this can also be called an IIS (irreducible inconsistent subsystem).</p><p>If supported by the solver, use <a href="../../api/JuMP/#JuMP.compute_conflict!"><code>compute_conflict!</code></a> to trigger the computation of a conflict. Once this process is finished, query the <a href="../../moi/reference/models/#MathOptInterface.ConflictStatus"><code>MOI.ConflictStatus</code></a> attribute to check if a conflict was found.</p><p>If found, copy the IIS to a new model using <a href="../../api/JuMP/#JuMP.copy_conflict"><code>copy_conflict</code></a>, which you can then print or write to a file for easier debugging:</p><pre><code class="language-julia hljs">julia&gt; using JuMP
+
+julia&gt; import Gurobi
+
+julia&gt; model = Model(Gurobi.Optimizer)
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: Gurobi
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 0)
+x
+
+julia&gt; @constraint(model, c1, x &gt;= 2)
+c1 : x ≥ 2.0
+
+julia&gt; @constraint(model, c2, x &lt;= 1)
+c2 : x ≤ 1.0
+
+julia&gt; optimize!(model)
+
+julia&gt; compute_conflict!(model)
+
+julia&gt; if get_attribute(model, MOI.ConflictStatus()) == MOI.CONFLICT_FOUND
+           iis_model, _ = copy_conflict(model)
+           print(iis_model)
+       end
+Feasibility
+Subject to
+ c1 : x ≥ 2.0
+ c2 : x ≤ 1.0</code></pre><p>If you need more control over the list of constraints that appear in the conflict, iterate over the list of constraints and query the <a href="../../moi/reference/models/#MathOptInterface.ConstraintConflictStatus"><code>MOI.ConstraintConflictStatus</code></a> attribute:</p><pre><code class="language-julia hljs">julia&gt; list_of_conflicting_constraints = ConstraintRef[]
+ConstraintRef[]
+
+julia&gt; for (F, S) in list_of_constraint_types(model)
+           for con in all_constraints(model, F, S)
+               if get_attribute(con, MOI.ConstraintConflictStatus()) == MOI.IN_CONFLICT
+                   push!(list_of_conflicting_constraints, con)
+               end
+           end
+       end
+
+julia&gt; list_of_conflicting_constraints
+2-element Vector{ConstraintRef}:
+ c1 : x ≥ 2.0
+ c2 : x ≤ 1.0</code></pre><h2 id="Multiple-solutions"><a class="docs-heading-anchor" href="#Multiple-solutions">Multiple solutions</a><a id="Multiple-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Multiple-solutions" title="Permalink"></a></h2><p>Some solvers support returning multiple solutions. You can check how many solutions are available to query using <a href="../../api/JuMP/#JuMP.result_count"><code>result_count</code></a>.</p><p>Functions for querying the solutions, for example, <a href="../../api/JuMP/#JuMP.primal_status"><code>primal_status</code></a>, <a href="../../api/JuMP/#JuMP.dual_status"><code>dual_status</code></a>, <a href="../../api/JuMP/#JuMP.value"><code>value</code></a>, <a href="../../api/JuMP/#JuMP.dual"><code>dual</code></a>, and <a href="../../api/JuMP/#JuMP.solution_summary"><code>solution_summary</code></a> all take an additional keyword argument <code>result</code> which can be used to specify which result to return.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Even if <a href="../../api/JuMP/#JuMP.termination_status"><code>termination_status</code></a> is <code>OPTIMAL</code>, some of the returned solutions may be suboptimal. However, if the solver found at least one optimal solution, then <code>result = 1</code> will always return an optimal solution. Use <a href="../../api/JuMP/#JuMP.objective_value"><code>objective_value</code></a> to assess the quality of the remaining solutions.</p></div></div><pre><code class="language-julia-repl hljs">julia&gt; using JuMP
+
+julia&gt; import MultiObjectiveAlgorithms as MOA
+
+julia&gt; import HiGHS
+
+julia&gt; model = Model(() -&gt; MOA.Optimizer(HiGHS.Optimizer));
+
+julia&gt; set_attribute(model, MOA.Algorithm(), MOA.Dichotomy())
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x1 &gt;= 0)
+x1
+
+julia&gt; @variable(model, 0 &lt;= x2 &lt;= 3)
+x2
+
+julia&gt; @objective(model, Min, [3x1 + x2, -x1 - 2x2])
+2-element Vector{AffExpr}:
+ 3 x1 + x2
+ -x1 - 2 x2
+
+julia&gt; @constraint(model, 3x1 - x2 &lt;= 6)
+3 x1 - x2 ≤ 6
+
+julia&gt; optimize!(model)
+
+julia&gt; solution_summary(model; result = 1)
+* Solver : MOA[algorithm=MultiObjectiveAlgorithms.Dichotomy, optimizer=HiGHS]
+
+* Status
+  Result count       : 3
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;Solve complete. Found 3 solution(s)&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : [0.00000e+00,0.00000e+00]
+  Objective bound    : [0.00000e+00,-9.00000e+00]
+  Relative gap       : Inf
+  Dual objective value : -9.00000e+00
+
+* Work counters
+  Solve time (sec)   : 1.82720e-03
+  Simplex iterations : 1
+  Barrier iterations : 0
+  Node count         : -1
+
+julia&gt; for i in 1:result_count(model)
+           println(&quot;Solution $i&quot;)
+           println(&quot;   x = &quot;, value.([x1, x2]; result = i))
+           println(&quot; obj = &quot;, objective_value(model; result = i))
+       end
+Solution 1
+   x = [0.0, 0.0]
+ obj = [0.0, 0.0]
+Solution 2
+   x = [0.0, 3.0]
+ obj = [3.0, -6.0]
+Solution 3
+   x = [3.0, 3.0]
+ obj = [12.0, -9.0]</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>The <a href="../../tutorials/linear/multi_objective_knapsack/#Multi-objective-knapsack">Multi-objective knapsack</a> tutorial provides more examples of querying multiple solutions.</p></div></div><h2 id="Checking-feasibility-of-solutions"><a class="docs-heading-anchor" href="#Checking-feasibility-of-solutions">Checking feasibility of solutions</a><a id="Checking-feasibility-of-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Checking-feasibility-of-solutions" title="Permalink"></a></h2><p>To check the feasibility of a primal solution, use <a href="../../api/JuMP/#JuMP.primal_feasibility_report"><code>primal_feasibility_report</code></a>, which takes a <code>model</code>, a dictionary mapping each variable to a primal solution value (defaults to the last solved solution), and a tolerance <code>atol</code> (defaults to <code>0.0</code>).</p><p>The function returns a dictionary which maps the infeasible constraint references to the distance between the primal value of the constraint and the nearest point in the corresponding set. A point is classed as infeasible if the distance is greater than the supplied tolerance <code>atol</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model(HiGHS.Optimizer);
+
+julia&gt; set_silent(model)
+
+julia&gt; @variable(model, x &gt;= 1, Int);
+
+julia&gt; @variable(model, y);
+
+julia&gt; @constraint(model, c1, x + y &lt;= 1.95);
+
+julia&gt; point = Dict(x =&gt; 1.9, y =&gt; 0.06);
+
+julia&gt; primal_feasibility_report(model, point)
+Dict{Any, Float64} with 2 entries:
+  x integer         =&gt; 0.1
+  c1 : x + y ≤ 1.95 =&gt; 0.01
+
+julia&gt; primal_feasibility_report(model, point; atol = 0.02)
+Dict{Any, Float64} with 1 entry:
+  x integer =&gt; 0.1</code></pre><p>If the point is feasible, an empty dictionary is returned:</p><pre><code class="language-julia-repl hljs">julia&gt; primal_feasibility_report(model, Dict(x =&gt; 1.0, y =&gt; 0.0))
+Dict{Any, Float64}()</code></pre><p>To use the primal solution from a solve, omit the <code>point</code> argument:</p><pre><code class="language-julia-repl hljs">julia&gt; optimize!(model)
+
+julia&gt; primal_feasibility_report(model; atol = 0.0)
+Dict{Any, Float64}()</code></pre><p>Calling <a href="../../api/JuMP/#JuMP.primal_feasibility_report"><code>primal_feasibility_report</code></a> without the <code>point</code> argument is useful when <a href="../../api/JuMP/#JuMP.primal_status"><code>primal_status</code></a> is <code>FEASIBLE_POINT</code> or <code>NEARLY_FEASIBLE_POINT</code>, and you want to assess the solution quality.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>To apply <a href="../../api/JuMP/#JuMP.primal_feasibility_report"><code>primal_feasibility_report</code></a> to infeasible models, you must also provide a candidate point (solvers generally do not provide one). To diagnose the source of infeasibility, see <a href="#Conflicts">Conflicts</a>.</p></div></div><p>Pass <code>skip_mising = true</code> to skip constraints which contain variables that are not in <code>point</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; primal_feasibility_report(model, Dict(x =&gt; 2.1); skip_missing = true)
+Dict{Any, Float64} with 1 entry:
+  x integer =&gt; 0.1</code></pre><p>You can also use the functional form, where the first argument is a function that maps variables to their primal values:</p><pre><code class="language-julia-repl hljs">julia&gt; optimize!(model)
+
+julia&gt; primal_feasibility_report(v -&gt; value(v), model)
+Dict{Any, Float64}()</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../containers/">« Containers</a><a class="docs-footer-nextpage" href="../callbacks/">Solver-independent Callbacks »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/manual/variables/index.html b/previews/PR3525/manual/variables/index.html
new file mode 100644
index 00000000000..dec9fef7a63
--- /dev/null
+++ b/previews/PR3525/manual/variables/index.html
@@ -0,0 +1,637 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Variables · JuMP</title><meta name="title" content="Variables · JuMP"/><meta property="og:title" content="Variables · JuMP"/><meta property="twitter:title" content="Variables · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox" checked/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../models/">Models</a></li><li class="is-active"><a class="tocitem" href>Variables</a><ul class="internal"><li><a class="tocitem" href="#Create-a-variable"><span>Create a variable</span></a></li><li><a class="tocitem" href="#Registered-variables"><span>Registered variables</span></a></li><li><a class="tocitem" href="#anonymous_variables"><span>Anonymous variables</span></a></li><li><a class="tocitem" href="#Variable-names"><span>Variable names</span></a></li><li><a class="tocitem" href="#variable_names_and_bindings"><span>String names, symbolic names, and bindings</span></a></li><li><a class="tocitem" href="#Create,-delete,-and-modify-variable-bounds"><span>Create, delete, and modify variable bounds</span></a></li><li><a class="tocitem" href="#Binary-variables"><span>Binary variables</span></a></li><li><a class="tocitem" href="#Integer-variables"><span>Integer variables</span></a></li><li><a class="tocitem" href="#Start-values"><span>Start values</span></a></li><li><a class="tocitem" href="#delete_a_variable"><span>Delete a variable</span></a></li><li><a class="tocitem" href="#Variable-containers"><span>Variable containers</span></a></li><li><a class="tocitem" href="#Semidefinite-variables"><span>Semidefinite variables</span></a></li><li><a class="tocitem" href="#Symmetric-variables"><span>Symmetric variables</span></a></li><li><a class="tocitem" href="#variables"><span>The <code>@variables</code> macro</span></a></li><li><a class="tocitem" href="#jump_variables_on_creation"><span>Variables constrained on creation</span></a></li><li><a class="tocitem" href="#Parameters"><span>Parameters</span></a></li></ul></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../expressions/">Expressions</a></li><li><a class="tocitem" href="../objective/">Objectives</a></li><li><a class="tocitem" href="../containers/">Containers</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../complex/">Complex number support</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Variables</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Variables</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/manual/variables.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="jump_variables"><a class="docs-heading-anchor" href="#jump_variables">Variables</a><a id="jump_variables-1"></a><a class="docs-heading-anchor-permalink" href="#jump_variables" title="Permalink"></a></h1><p>The term <em>variable</em> in mathematical optimization has many meanings. For example, <em>optimization</em> variables (also called decision variables) are the unknowns <span>$x$</span> that we are solving for in the problem:</p><p class="math-container">\[\begin{align}
+    &amp; \min_{x \in \mathbb{R}^n} &amp; f_0(x)
+    \\
+    &amp; \;\;\text{s.t.} &amp; f_i(x) &amp; \in \mathcal{S}_i &amp; i = 1 \ldots m
+\end{align}\]</p><p>To complicate things, Julia uses <em>variable</em> to mean a binding between a name and a value. For example, in the statement:</p><pre><code class="language-julia-repl hljs">julia&gt; x = 1
+1</code></pre><p><code>x</code> is a variable that stores the value <code>1</code>.</p><p>JuMP uses <em>variable</em> in a third way, to mean an instance of the <a href="../../api/JuMP/#JuMP.VariableRef"><code>VariableRef</code></a> struct. JuMP variables are the link between Julia and the optimization variables inside a JuMP model.</p><p>This page explains how to create and manage JuMP variables in a variety of contexts.</p><h2 id="Create-a-variable"><a class="docs-heading-anchor" href="#Create-a-variable">Create a variable</a><a id="Create-a-variable-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-variable" title="Permalink"></a></h2><p>Create variables using the <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; typeof(x)
+VariableRef (alias for GenericVariableRef{Float64})
+
+julia&gt; num_variables(model)
+1</code></pre><p>Here <code>x</code> is a Julia variable that is bound to a <a href="../../api/JuMP/#JuMP.VariableRef"><code>VariableRef</code></a> object, and we have added 1 decision variable to our model.</p><p>To make the binding more explicit, we could have written:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x = @variable(model, x)
+x</code></pre><p>but there is no need to in general; the macro does it for us.</p><p>When creating a variable, you can also specify variable bounds:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x_free)
+x_free
+
+julia&gt; @variable(model, x_lower &gt;= 0)
+x_lower
+
+julia&gt; @variable(model, x_upper &lt;= 1)
+x_upper
+
+julia&gt; @variable(model, 2 &lt;= x_interval &lt;= 3)
+x_interval
+
+julia&gt; @variable(model, x_fixed == 4)
+x_fixed
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ x_fixed = 4
+ x_lower ≥ 0
+ x_interval ≥ 2
+ x_upper ≤ 1
+ x_interval ≤ 3</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>When creating a variable with a single lower- or upper-bound, and the value of the bound is not a numeric literal (for example, <code>1</code> or <code>1.0</code>), the name of the variable <em>must</em> appear on the left-hand side. Putting the name on the right-hand side is an error. For example, to create a variable <code>x</code>:</p><pre><code class="language-julia hljs">a = 1
+@variable(model, x &gt;= 1)      # ✓ Okay
+@variable(model, 1.0 &lt;= x)    # ✓ Okay
+@variable(model, x &gt;= a)      # ✓ Okay
+@variable(model, a &lt;= x)      # × Not okay
+@variable(model, x &gt;= 1 / 2)  # ✓ Okay
+@variable(model, 1 / 2 &lt;= x)  # × Not okay</code></pre></div></div><h3 id="Containers-of-variables"><a class="docs-heading-anchor" href="#Containers-of-variables">Containers of variables</a><a id="Containers-of-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Containers-of-variables" title="Permalink"></a></h3><p>The <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro also supports creating collections of JuMP variables. We&#39;ll cover some brief syntax here; read the <a href="#Variable-containers">Variable containers</a> section for more details.</p><p>You can create arrays of JuMP variables:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2])
+2×2 Matrix{VariableRef}:
+ x[1,1]  x[1,2]
+ x[2,1]  x[2,2]
+
+julia&gt; x[1, 2]
+x[1,2]</code></pre><p>Index sets can be named, and bounds can depend on those names:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, sqrt(i) &lt;= x[i = 1:3] &lt;= i^2)
+3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+
+julia&gt; x[2]
+x[2]</code></pre><p>Sets can be any Julia type that supports iteration:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i = 2:3, j = 1:2:3, [&quot;red&quot;, &quot;blue&quot;]] &gt;= 0)
+3-dimensional DenseAxisArray{VariableRef,3,...} with index sets:
+    Dimension 1, 2:3
+    Dimension 2, 1:2:3
+    Dimension 3, [&quot;red&quot;, &quot;blue&quot;]
+And data, a 2×2×2 Array{VariableRef, 3}:
+[:, :, &quot;red&quot;] =
+ x[2,1,red]  x[2,3,red]
+ x[3,1,red]  x[3,3,red]
+
+[:, :, &quot;blue&quot;] =
+ x[2,1,blue]  x[2,3,blue]
+ x[3,1,blue]  x[3,3,blue]
+
+julia&gt; x[2, 1, &quot;red&quot;]
+x[2,1,red]</code></pre><p>Sets can depend upon previous indices:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, u[i = 1:2, j = i:3])
+JuMP.Containers.SparseAxisArray{VariableRef, 2, Tuple{Int64, Int64}} with 5 entries:
+  [1, 1]  =  u[1,1]
+  [1, 2]  =  u[1,2]
+  [1, 3]  =  u[1,3]
+  [2, 2]  =  u[2,2]
+  [2, 3]  =  u[2,3]</code></pre><p>and we can filter elements in the sets using the <code>;</code> syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, v[i = 1:9; mod(i, 3) == 0])
+JuMP.Containers.SparseAxisArray{VariableRef, 1, Tuple{Int64}} with 3 entries:
+  [3]  =  v[3]
+  [6]  =  v[6]
+  [9]  =  v[9]</code></pre><h2 id="Registered-variables"><a class="docs-heading-anchor" href="#Registered-variables">Registered variables</a><a id="Registered-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Registered-variables" title="Permalink"></a></h2><p>When you create variables, JuMP registers them inside the model using their corresponding symbol. Get a registered name using <code>model[:key]</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model()
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; model
+A JuMP Model
+Feasibility problem with:
+Variable: 1
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+Names registered in the model: x
+
+julia&gt; model[:x] === x
+true</code></pre><p>Registered names are most useful when you start to write larger models and want to break up the model construction into functions:</p><pre><code class="language-julia-repl hljs">julia&gt; function set_objective(model::Model)
+           @objective(model, Min, 2 * model[:my_x] + 1)
+           return
+       end
+set_objective (generic function with 1 method)
+
+julia&gt; model = Model();
+
+julia&gt; @variable(model, my_x);
+
+julia&gt; set_objective(model)
+
+julia&gt; print(model)
+Min 2 my_x + 1
+Subject to</code></pre><h2 id="anonymous_variables"><a class="docs-heading-anchor" href="#anonymous_variables">Anonymous variables</a><a id="anonymous_variables-1"></a><a class="docs-heading-anchor-permalink" href="#anonymous_variables" title="Permalink"></a></h2><p>To reduce the likelihood of accidental bugs, and because JuMP registers variables inside a model, creating two variables with the same name is an error:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; @variable(model, x)
+ERROR: An object of name x is already attached to this model. If this
+    is intended, consider using the anonymous construction syntax, e.g.,
+    `x = @variable(model, [1:N], ...)` where the name of the object does
+    not appear inside the macro.
+
+    Alternatively, use `unregister(model, :x)` to first unregister
+    the existing name from the model. Note that this will not delete the
+    object; it will just remove the reference at `model[:x]`.
+[...]</code></pre><p>A common reason for encountering this error is adding variables in a loop.</p><p>As a work-around, JuMP provides <em>anonymous</em> variables. Create a scalar valued anonymous variable by omitting the name argument:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x = @variable(model)
+_[1]</code></pre><p>Anonymous variables get printed as an underscore followed by a unique index of the variable.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The index of the variable may not correspond to the column of the variable in the solver.</p></div></div><p>Create a container of anonymous JuMP variables by dropping the name in front of the <code>[</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; y = @variable(model, [1:2])
+2-element Vector{VariableRef}:
+ _[1]
+ _[2]</code></pre><p>The <code>&lt;=</code> and <code>&gt;=</code> short-hand cannot be used to set bounds on scalar-valued anonymous JuMP variables. Instead, use the <code>lower_bound</code> and <code>upper_bound</code> keywords:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x_lower = @variable(model, lower_bound = 1.0)
+_[1]
+
+julia&gt; x_upper = @variable(model, upper_bound = 2.0)
+_[2]
+
+julia&gt; x_interval = @variable(model, lower_bound = 3.0, upper_bound = 4.0)
+_[3]</code></pre><h2 id="Variable-names"><a class="docs-heading-anchor" href="#Variable-names">Variable names</a><a id="Variable-names-1"></a><a class="docs-heading-anchor-permalink" href="#Variable-names" title="Permalink"></a></h2><p>In addition to the symbol that variables are registered with, JuMP variables have a <code>String</code> name that is used for printing and writing to file formats.</p><p>Get and set the name of a variable using <a href="../../api/JuMP/#JuMP.name"><code>name</code></a> and <a href="../../api/JuMP/#JuMP.set_name"><code>set_name</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; name(x)
+&quot;x&quot;
+
+julia&gt; set_name(x, &quot;my_x_name&quot;)
+
+julia&gt; x
+my_x_name</code></pre><p>Override the default choice of name using the <code>base_name</code> keyword:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2], base_name = &quot;my_var&quot;)
+2-element Vector{VariableRef}:
+ my_var[1]
+ my_var[2]</code></pre><p>Note that names apply to each element of the container, not to the container of variables:</p><pre><code class="language-julia-repl hljs">julia&gt; name(x[1])
+&quot;my_var[1]&quot;
+
+julia&gt; set_name(x[1], &quot;my_x&quot;)
+
+julia&gt; x
+2-element Vector{VariableRef}:
+ my_x
+ my_var[2]</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>For some models, setting the string name of each variable can take a non-trivial portion of the total time required to build the model. Turn off <code>String</code> names by passing <code>set_string_name = false</code> to <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, set_string_name = false)
+_[1]</code></pre><p>See <a href="../../tutorials/getting_started/performance_tips/#Disable-string-names">Disable string names</a> for more information.</p></div></div><h3 id="Retrieve-a-variable-by-name"><a class="docs-heading-anchor" href="#Retrieve-a-variable-by-name">Retrieve a variable by name</a><a id="Retrieve-a-variable-by-name-1"></a><a class="docs-heading-anchor-permalink" href="#Retrieve-a-variable-by-name" title="Permalink"></a></h3><p>Retrieve a variable from a model using <a href="../../api/JuMP/#JuMP.variable_by_name"><code>variable_by_name</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; variable_by_name(model, &quot;my_x&quot;)
+my_x</code></pre><p>If the name is not present, <code>nothing</code> will be returned:</p><pre><code class="language-julia-repl hljs">julia&gt; variable_by_name(model, &quot;bad_name&quot;)</code></pre><p>You can only look up individual variables using <a href="../../api/JuMP/#JuMP.variable_by_name"><code>variable_by_name</code></a>. Something like this will not work:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, [i = 1:2], base_name = &quot;my_var&quot;)
+2-element Vector{VariableRef}:
+ my_var[1]
+ my_var[2]
+
+julia&gt; variable_by_name(model, &quot;my_var&quot;)</code></pre><p>To look up a collection of variables, do not use <a href="../../api/JuMP/#JuMP.variable_by_name"><code>variable_by_name</code></a>. Instead, register them using the <code>model[:key] = value</code> syntax:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; model[:x] = @variable(model, [i = 1:2], base_name = &quot;my_var&quot;)
+2-element Vector{VariableRef}:
+ my_var[1]
+ my_var[2]
+
+julia&gt; model[:x]
+2-element Vector{VariableRef}:
+ my_var[1]
+ my_var[2]</code></pre><h2 id="variable_names_and_bindings"><a class="docs-heading-anchor" href="#variable_names_and_bindings">String names, symbolic names, and bindings</a><a id="variable_names_and_bindings-1"></a><a class="docs-heading-anchor-permalink" href="#variable_names_and_bindings" title="Permalink"></a></h2><p>It&#39;s common for new users to experience confusion relating to JuMP variables. Part of the problem is the overloaded use of &quot;variable&quot; in mathematical optimization, along with the difference between the name that a variable is registered under and the <code>String</code> name used for printing.</p><p>Here&#39;s a summary of the differences:</p><ul><li>JuMP variables are created using <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>.</li><li>JuMP variables can be named or anonymous.</li><li>Named JuMP variables have the form <code>@variable(model, x)</code>. For named variables:<ul><li>The <code>String</code> name of the variable is set to <code>&quot;x&quot;</code>.</li><li>A Julia variable <code>x</code> is created that binds <code>x</code> to  the JuMP variable.</li><li>The name <code>:x</code> is registered as a key in the model with the value <code>x</code>.</li></ul></li><li>Anonymous JuMP variables have the form <code>x = @variable(model)</code>. For anonymous variables:<ul><li>The <code>String</code> name of the variable is set to <code>&quot;&quot;</code>. When printed, this is replaced with <code>&quot;_[i]&quot;</code> where <code>i</code> is the index of the variable.</li><li>You control the name of the Julia variable used as the binding.</li><li>No name is registered as a key in the model.</li></ul></li><li>The <code>base_name</code> keyword can override the <code>String</code> name of the variable.</li><li>You can manually register names in the model via <code>model[:key] = value</code></li></ul><p>Here&#39;s an example that should make things clearer:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x_binding = @variable(model, base_name = &quot;x&quot;)
+x
+
+julia&gt; model
+A JuMP Model
+Feasibility problem with:
+Variable: 1
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+
+julia&gt; x
+ERROR: UndefVarError: `x` not defined
+
+julia&gt; x_binding
+x
+
+julia&gt; name(x_binding)
+&quot;x&quot;
+
+julia&gt; model[:x_register] = x_binding
+x
+
+julia&gt; model
+A JuMP Model
+Feasibility problem with:
+Variable: 1
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.
+Names registered in the model: x_register
+
+julia&gt; model[:x_register]
+x
+
+julia&gt; model[:x_register] === x_binding
+true
+
+julia&gt; x
+ERROR: UndefVarError: `x` not defined</code></pre><h2 id="Create,-delete,-and-modify-variable-bounds"><a class="docs-heading-anchor" href="#Create,-delete,-and-modify-variable-bounds">Create, delete, and modify variable bounds</a><a id="Create,-delete,-and-modify-variable-bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Create,-delete,-and-modify-variable-bounds" title="Permalink"></a></h2><p>Query whether a variable has a bound using <a href="../../api/JuMP/#JuMP.has_lower_bound"><code>has_lower_bound</code></a>, <a href="../../api/JuMP/#JuMP.has_upper_bound"><code>has_upper_bound</code></a>, and <a href="../../api/JuMP/#JuMP.is_fixed"><code>is_fixed</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; has_lower_bound(x_free)
+false
+
+julia&gt; has_upper_bound(x_upper)
+true
+
+julia&gt; is_fixed(x_fixed)
+true</code></pre><p>If a variable has a particular bound, query the value of it using <a href="../../api/JuMP/#JuMP.lower_bound"><code>lower_bound</code></a>, <a href="../../api/JuMP/#JuMP.upper_bound"><code>upper_bound</code></a>, and <a href="../../api/JuMP/#JuMP.fix_value"><code>fix_value</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; lower_bound(x_interval)
+2.0
+
+julia&gt; upper_bound(x_interval)
+3.0
+
+julia&gt; fix_value(x_fixed)
+4.0</code></pre><p>Querying the value of a bound that does not exist will result in an error.</p><p>Delete variable bounds using <a href="../../api/JuMP/#JuMP.delete_lower_bound"><code>delete_lower_bound</code></a>, <a href="../../api/JuMP/#JuMP.delete_upper_bound"><code>delete_upper_bound</code></a>, and <a href="../../api/JuMP/#JuMP.unfix"><code>unfix</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; delete_lower_bound(x_lower)
+
+julia&gt; has_lower_bound(x_lower)
+false
+
+julia&gt; delete_upper_bound(x_upper)
+
+julia&gt; has_upper_bound(x_upper)
+false
+
+julia&gt; unfix(x_fixed)
+
+julia&gt; is_fixed(x_fixed)
+false</code></pre><p>Set or update variable bounds using <a href="../../api/JuMP/#JuMP.set_lower_bound"><code>set_lower_bound</code></a>, <a href="../../api/JuMP/#JuMP.set_upper_bound"><code>set_upper_bound</code></a>, and <a href="../../api/JuMP/#JuMP.fix"><code>fix</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; set_lower_bound(x_lower, 1.1)
+
+julia&gt; set_upper_bound(x_upper, 2.1)
+
+julia&gt; fix(x_fixed, 4.1)</code></pre><p>Fixing a variable with existing bounds will throw an error. To delete the bounds prior to fixing, use <code>fix(variable, value; force = true)</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x &gt;= 1)
+x
+
+julia&gt; fix(x, 2)
+ERROR: Unable to fix x to 2 because it has existing variable bounds. Consider calling `JuMP.fix(variable, value; force=true)` which will delete existing bounds before fixing the variable.
+
+julia&gt; fix(x, 2; force = true)
+
+
+julia&gt; fix_value(x)
+2.0</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Use <a href="../../api/JuMP/#JuMP.fix"><code>fix</code></a> instead of <code>@constraint(model, x == 2)</code>. The former modifies variable bounds, while the latter adds a new linear constraint to the problem.</p></div></div><h2 id="Binary-variables"><a class="docs-heading-anchor" href="#Binary-variables">Binary variables</a><a id="Binary-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Binary-variables" title="Permalink"></a></h2><p>Binary variables are constrained to the set <span>$x \in \{0, 1\}$</span>.</p><p>Create a binary variable by passing <code>Bin</code> as an optional positional argument:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Bin)
+x</code></pre><p>Check if a variable is binary using <a href="../../api/JuMP/#JuMP.is_binary"><code>is_binary</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; is_binary(x)
+true</code></pre><p>Delete a binary constraint using <a href="../../api/JuMP/#JuMP.unset_binary"><code>unset_binary</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; unset_binary(x)
+
+julia&gt; is_binary(x)
+false</code></pre><p>Binary variables can also be created by setting the <code>binary</code> keyword to <code>true</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, binary=true)
+x</code></pre><p>or by using <a href="../../api/JuMP/#JuMP.set_binary"><code>set_binary</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; set_binary(x)</code></pre><h2 id="Integer-variables"><a class="docs-heading-anchor" href="#Integer-variables">Integer variables</a><a id="Integer-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Integer-variables" title="Permalink"></a></h2><p>Integer variables are constrained to the set <span>$x \in \mathbb{Z}$</span>.</p><p>Create an integer variable by passing <code>Int</code> as an optional positional argument:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, Int)
+x</code></pre><p>Check if a variable is integer using <a href="../../api/JuMP/#JuMP.is_integer"><code>is_integer</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; is_integer(x)
+true</code></pre><p>Delete an integer constraint using <a href="../../api/JuMP/#JuMP.unset_integer"><code>unset_integer</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; unset_integer(x)
+
+julia&gt; is_integer(x)
+false</code></pre><p>Integer variables can also be created by setting the <code>integer</code> keyword to <code>true</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x, integer=true)
+x</code></pre><p>or by using <a href="../../api/JuMP/#JuMP.set_integer"><code>set_integer</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; set_integer(x)</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>The <a href="../../api/JuMP/#JuMP.relax_integrality"><code>relax_integrality</code></a> function relaxes all integrality constraints in the model, returning a function that can be called to undo the operation later on.</p></div></div><h2 id="Start-values"><a class="docs-heading-anchor" href="#Start-values">Start values</a><a id="Start-values-1"></a><a class="docs-heading-anchor-permalink" href="#Start-values" title="Permalink"></a></h2><p>There are two ways to provide a primal starting solution (also called MIP-start or a warmstart) for each variable:</p><ul><li>using the <code>start</code> keyword in the <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro</li><li>using <a href="../../api/JuMP/#JuMP.set_start_value"><code>set_start_value</code></a></li></ul><p>The starting value of a variable can be queried using <a href="../../api/JuMP/#JuMP.start_value"><code>start_value</code></a>. If no start value has been set, <a href="../../api/JuMP/#JuMP.start_value"><code>start_value</code></a> will return <code>nothing</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; start_value(x)
+
+julia&gt; @variable(model, y, start = 1)
+y
+
+julia&gt; start_value(y)
+1.0
+
+julia&gt; set_start_value(y, 2)
+
+julia&gt; start_value(y)
+2.0</code></pre><p>The <code>start</code> keyword argument can depend on the indices of a variable container:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, z[i = 1:2], start = i^2)
+2-element Vector{VariableRef}:
+ z[1]
+ z[2]
+
+julia&gt; start_value.(z)
+2-element Vector{Float64}:
+ 1.0
+ 4.0</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Some solvers do not support start values. If a solver does not support start values, an <code>MathOptInterface.UnsupportedAttribute{MathOptInterface.VariablePrimalStart}</code> error will be thrown.</p></div></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>To set the optimal solution from a previous solve as a new starting value, use <a href="../../api/JuMP/#JuMP.all_variables"><code>all_variables</code></a> to get a vector of all the variables in the model, then run:</p><pre><code class="language-julia hljs">x = all_variables(model)
+x_solution = value.(x)
+set_start_value.(x, x_solution)</code></pre><p>Alternatively, use <a href="../../api/JuMP/#JuMP.set_start_values"><code>set_start_values</code></a>.</p></div></div><h2 id="delete_a_variable"><a class="docs-heading-anchor" href="#delete_a_variable">Delete a variable</a><a id="delete_a_variable-1"></a><a class="docs-heading-anchor-permalink" href="#delete_a_variable" title="Permalink"></a></h2><p>Use <a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a> to delete a variable from a model. Use <a href="../../api/JuMP/#JuMP.is_valid"><code>is_valid</code></a> to check if a variable belongs to a model and has not been deleted.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; is_valid(model, x)
+true
+
+julia&gt; delete(model, x)
+
+julia&gt; is_valid(model, x)
+false</code></pre><p>Deleting a variable does not unregister the corresponding name from the model. Therefore, creating a new variable of the same name will throw an error:</p><pre><code class="language-julia-repl hljs">julia&gt; @variable(model, x)
+ERROR: An object of name x is already attached to this model. If this
+    is intended, consider using the anonymous construction syntax, e.g.,
+    `x = @variable(model, [1:N], ...)` where the name of the object does
+    not appear inside the macro.
+
+    Alternatively, use `unregister(model, :x)` to first unregister
+    the existing name from the model. Note that this will not delete the
+    object; it will just remove the reference at `model[:x]`.
+[...]</code></pre><p>After calling <a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a>, call <a href="../../api/JuMP/#JuMP.unregister"><code>unregister</code></a> to remove the symbolic reference:</p><pre><code class="language-julia-repl hljs">julia&gt; unregister(model, :x)
+
+julia&gt; @variable(model, x)
+x</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p><a href="../../api/JuMP/#JuMP.delete"><code>delete</code></a> does not automatically <a href="../../api/JuMP/#JuMP.unregister"><code>unregister</code></a> because we do not distinguish between names that are automatically registered by JuMP macros and names that are manually registered by the user by setting values in <a href="../../api/JuMP/#JuMP.object_dictionary"><code>object_dictionary</code></a>. In addition, deleting a variable and then adding a new variable of the same name is an easy way to introduce bugs into your code.</p></div></div><h2 id="Variable-containers"><a class="docs-heading-anchor" href="#Variable-containers">Variable containers</a><a id="Variable-containers-1"></a><a class="docs-heading-anchor-permalink" href="#Variable-containers" title="Permalink"></a></h2><p>JuMP provides a mechanism for creating collections of variables in three types of data structures, which we refer to as <em>containers</em>.</p><p>The three types are <code>Array</code>s, <code>DenseAxisArray</code>s, and <code>SparseAxisArray</code>s. We explain each of these in the following.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>You can read more about containers in the <a href="../containers/#Containers">Containers</a> section.</p></div></div><h3 id="Arrays"><a class="docs-heading-anchor" href="#Arrays">Arrays</a><a id="Arrays-1"></a><a class="docs-heading-anchor-permalink" href="#Arrays" title="Permalink"></a></h3><p>We have already seen the creation of an array of JuMP variables with the <code>x[1:2]</code> syntax. This can be extended to create multi-dimensional arrays of JuMP variables. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2])
+2×2 Matrix{VariableRef}:
+ x[1,1]  x[1,2]
+ x[2,1]  x[2,2]</code></pre><p>Arrays of JuMP variables can be indexed and sliced as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; x[1, 2]
+x[1,2]
+
+julia&gt; x[2, :]
+2-element Vector{VariableRef}:
+ x[2,1]
+ x[2,2]</code></pre><p>Variable bounds can depend upon the indices:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2, j=1:2] &gt;= 2i + j)
+2×2 Matrix{VariableRef}:
+ x[1,1]  x[1,2]
+ x[2,1]  x[2,2]
+
+julia&gt; lower_bound.(x)
+2×2 Matrix{Float64}:
+ 3.0  4.0
+ 5.0  6.0</code></pre><p>JuMP will form an <code>Array</code> of JuMP variables when it can determine at compile time that the indices are one-based integer ranges. Therefore <code>x[1:b]</code> will create an <code>Array</code> of JuMP variables, but <code>x[a:b]</code> will not. If JuMP cannot determine that the indices are one-based integer ranges (for example, in the case of <code>x[a:b]</code>), JuMP will create a <code>DenseAxisArray</code> instead.</p><h3 id="variable_jump_arrays"><a class="docs-heading-anchor" href="#variable_jump_arrays">DenseAxisArrays</a><a id="variable_jump_arrays-1"></a><a class="docs-heading-anchor-permalink" href="#variable_jump_arrays" title="Permalink"></a></h3><p>We often want to create arrays where the indices are not one-based integer ranges. For example, we may want to create a variable indexed by the name of a product or a location. The syntax is the same as that above, except with an arbitrary vector as an index as opposed to a one-based range. The biggest difference is that instead of returning an <code>Array</code> of JuMP variables, JuMP will return a <code>DenseAxisArray</code>. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, [:A,:B]])
+2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+    Dimension 2, [:A, :B]
+And data, a 2×2 Matrix{VariableRef}:
+ x[1,A]  x[1,B]
+ x[2,A]  x[2,B]</code></pre><p>DenseAxisArrays can be indexed and sliced as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; x[1, :A]
+x[1,A]
+
+julia&gt; x[2, :]
+1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [:A, :B]
+And data, a 2-element Vector{VariableRef}:
+ x[2,A]
+ x[2,B]</code></pre><p>Bounds can depend upon indices:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=2:3, j=1:2:3] &gt;= 0.5i + j)
+2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:
+    Dimension 1, 2:3
+    Dimension 2, 1:2:3
+And data, a 2×2 Matrix{VariableRef}:
+ x[2,1]  x[2,3]
+ x[3,1]  x[3,3]
+
+julia&gt; lower_bound.(x)
+2-dimensional DenseAxisArray{Float64,2,...} with index sets:
+    Dimension 1, 2:3
+    Dimension 2, 1:2:3
+And data, a 2×2 Matrix{Float64}:
+ 2.0  4.0
+ 2.5  4.5</code></pre><h3 id="variable_sparseaxisarrays"><a class="docs-heading-anchor" href="#variable_sparseaxisarrays">SparseAxisArrays</a><a id="variable_sparseaxisarrays-1"></a><a class="docs-heading-anchor-permalink" href="#variable_sparseaxisarrays" title="Permalink"></a></h3><p>The third container type that JuMP natively supports is <code>SparseAxisArray</code>. These arrays are created when the indices do not form a rectangular set. For example, this applies when indices have a dependence upon previous indices (called <em>triangular indexing</em>). JuMP supports this as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:2, j=i:2])
+JuMP.Containers.SparseAxisArray{VariableRef, 2, Tuple{Int64, Int64}} with 3 entries:
+  [1, 1]  =  x[1,1]
+  [1, 2]  =  x[1,2]
+  [2, 2]  =  x[2,2]</code></pre><p>We can also conditionally create variables via a JuMP-specific syntax. This syntax appends a comparison check that depends upon the named indices and is separated from the indices by a semi-colon (<code>;</code>). For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[i=1:4; mod(i, 2)==0])
+JuMP.Containers.SparseAxisArray{VariableRef, 1, Tuple{Int64}} with 2 entries:
+  [2]  =  x[2]
+  [4]  =  x[4]</code></pre><h4 id="Performance-considerations"><a class="docs-heading-anchor" href="#Performance-considerations">Performance considerations</a><a id="Performance-considerations-1"></a><a class="docs-heading-anchor-permalink" href="#Performance-considerations" title="Permalink"></a></h4><p>When using the semi-colon as a filter, JuMP iterates over <em>all</em> indices and evaluates the conditional for each combination. If there are many index dimensions and a large amount of sparsity, this can be inefficient.</p><p>For example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; N = 10
+10
+
+julia&gt; S = [(1, 1, 1), (N, N, N)]
+2-element Vector{Tuple{Int64, Int64, Int64}}:
+ (1, 1, 1)
+ (10, 10, 10)
+
+julia&gt; @time @variable(model, x1[i=1:N, j=1:N, k=1:N; (i, j, k) in S])
+  0.203861 seconds (392.22 k allocations: 23.977 MiB, 99.10% compilation time)
+JuMP.Containers.SparseAxisArray{VariableRef, 3, Tuple{Int64, Int64, Int64}} with 2 entries:
+  [1, 1, 1   ]  =  x1[1,1,1]
+  [10, 10, 10]  =  x1[10,10,10]
+
+julia&gt; @time @variable(model, x2[S])
+  0.045407 seconds (65.24 k allocations: 3.771 MiB, 99.15% compilation time)
+1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [(1, 1, 1), (10, 10, 10)]
+And data, a 2-element Vector{VariableRef}:
+ x2[(1, 1, 1)]
+ x2[(10, 10, 10)]</code></pre><p>The first option is slower because it is equivalent to:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x1 = Dict{NTuple{3,Int},VariableRef}()
+Dict{Tuple{Int64, Int64, Int64}, VariableRef}()
+
+julia&gt; for i in 1:N
+           for j in 1:N
+               for k in 1:N
+                   if (i, j, k) in S
+                       x1[i, j, k] = @variable(model, base_name = &quot;x1[$i,$j,$k]&quot;)
+                   end
+               end
+           end
+       end
+
+julia&gt; x1
+Dict{Tuple{Int64, Int64, Int64}, VariableRef} with 2 entries:
+  (1, 1, 1)    =&gt; x1[1,1,1]
+  (10, 10, 10) =&gt; x1[10,10,10]</code></pre><p>If performance is a concern, explicitly construct the set of indices instead of using the filtering syntax.</p><h3 id="variable_forcing"><a class="docs-heading-anchor" href="#variable_forcing">Forcing the container type</a><a id="variable_forcing-1"></a><a class="docs-heading-anchor-permalink" href="#variable_forcing" title="Permalink"></a></h3><p>When creating a container of JuMP variables, JuMP will attempt to choose the tightest container type that can store the JuMP variables. Thus, it will prefer to create an Array before a DenseAxisArray and a DenseAxisArray before a SparseAxisArray. However, because this happens at compile time, JuMP does not always make the best choice. To illustrate this, consider the following example:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; A = 1:2
+1:2
+
+julia&gt; @variable(model, x[A])
+1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, 1:2
+And data, a 2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code></pre><p>Since the value (and type) of <code>A</code> is unknown at parsing time, JuMP is unable to infer that <code>A</code> is a one-based integer range. Therefore, JuMP creates a <code>DenseAxisArray</code>, even though it could store these two variables in a standard one-dimensional <code>Array</code>.</p><p>We can share our knowledge that it is possible to store these JuMP variables as an array by setting the <code>container</code> keyword:</p><pre><code class="language-julia-repl hljs">julia&gt; @variable(model, y[A], container=Array)
+2-element Vector{VariableRef}:
+ y[1]
+ y[2]</code></pre><p>JuMP now creates a vector of JuMP variables instead of a DenseAxisArray. Choosing an invalid container type will throw an error.</p><h3 id="User-defined-containers"><a class="docs-heading-anchor" href="#User-defined-containers">User-defined containers</a><a id="User-defined-containers-1"></a><a class="docs-heading-anchor-permalink" href="#User-defined-containers" title="Permalink"></a></h3><p>In addition to the built-in container types, you can create your own collections of JuMP variables.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This is a point that users often overlook: you are not restricted to the built-in container types in JuMP.</p></div></div><p>For example, the following code creates a dictionary with symmetric matrices as the values:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; variables = Dict{Symbol,Array{VariableRef,2}}(
+           key =&gt; @variable(model, [1:2, 1:2], Symmetric, base_name = &quot;$(key)&quot;)
+           for key in [:A, :B]
+       )
+Dict{Symbol, Matrix{VariableRef}} with 2 entries:
+  :A =&gt; [A[1,1] A[1,2]; A[1,2] A[2,2]]
+  :B =&gt; [B[1,1] B[1,2]; B[1,2] B[2,2]]</code></pre><p>Another common scenario is a request to add variables to existing containers, for example:</p><pre><code class="language-julia hljs">using JuMP
+model = Model()
+@variable(model, x[1:2] &gt;= 0)
+# Later I want to add
+@variable(model, x[3:4] &gt;= 0)</code></pre><p>This is not possible with the built-in JuMP container types. However, you can use regular Julia types instead:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x = model[:x] = @variable(model, [1:2], lower_bound = 0, base_name = &quot;x&quot;)
+2-element Vector{VariableRef}:
+ x[1]
+ x[2]
+
+julia&gt; append!(x, @variable(model, [1:2], lower_bound = 0, base_name = &quot;y&quot;));
+
+julia&gt; model[:x]
+4-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ y[1]
+ y[2]</code></pre><h2 id="Semidefinite-variables"><a class="docs-heading-anchor" href="#Semidefinite-variables">Semidefinite variables</a><a id="Semidefinite-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Semidefinite-variables" title="Permalink"></a></h2><p>Declare a square matrix of JuMP variables to be positive semidefinite by passing <code>PSD</code> as an optional positional argument:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2], PSD)
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ x[1,1]  x[1,2]
+ x[1,2]  x[2,2]</code></pre><p>This will ensure that <code>x</code> is symmetric, and that all of its eigenvalues are nonnegative.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p><code>x</code> must be a square 2-dimensional <code>Array</code> of JuMP variables; it cannot be a DenseAxisArray or a SparseAxisArray.</p></div></div><h2 id="Symmetric-variables"><a class="docs-heading-anchor" href="#Symmetric-variables">Symmetric variables</a><a id="Symmetric-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Symmetric-variables" title="Permalink"></a></h2><p>Declare a square matrix of JuMP variables to be symmetric (but not necessarily positive semidefinite) by passing <code>Symmetric</code>  as an optional positional argument:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2], Symmetric)
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ x[1,1]  x[1,2]
+ x[1,2]  x[2,2]</code></pre><h2 id="variables"><a class="docs-heading-anchor" href="#variables">The <code>@variables</code> macro</a><a id="variables-1"></a><a class="docs-heading-anchor-permalink" href="#variables" title="Permalink"></a></h2><p>If you have many <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> calls, JuMP provides the macro <a href="../../api/JuMP/#JuMP.@variables"><code>@variables</code></a> that can improve readability:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variables(model, begin
+           x
+           y[i=1:2] &gt;= i, (start = i, base_name = &quot;Y_$i&quot;)
+           z, Bin
+       end)
+(x, VariableRef[Y_1[1], Y_2[2]], z)
+
+julia&gt; print(model)
+Feasibility
+Subject to
+ Y_1[1] ≥ 1
+ Y_2[2] ≥ 2
+ z binary</code></pre><p>The <a href="../../api/JuMP/#JuMP.@variables"><code>@variables</code></a> macro returns a tuple of the variables that were defined.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Keyword arguments must be contained within parentheses.</p></div></div><h2 id="jump_variables_on_creation"><a class="docs-heading-anchor" href="#jump_variables_on_creation">Variables constrained on creation</a><a id="jump_variables_on_creation-1"></a><a class="docs-heading-anchor-permalink" href="#jump_variables_on_creation" title="Permalink"></a></h2><p>All uses of the <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro documented so far translate into separate calls for variable creation and the adding of any bound or integrality constraints.</p><p>For example, <code>@variable(model, x &gt;= 0, Int)</code>, is equivalent to:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x)
+x
+
+julia&gt; set_lower_bound(x, 0.0)
+
+julia&gt; set_integer(x)</code></pre><p>Importantly, the bound and integrality constraints are added <em>after</em> the variable has been created.</p><p>However, some solvers require a set specifying the variable domain to be given when the variable is first created. We say that these variables are <em>constrained on creation</em>.</p><p>Use <code>in</code> within <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> to access the special syntax for constraining variables on creation.</p><p>For example, the following creates a vector of variables that belong to the <a href="../../api/JuMP/#JuMP.SecondOrderCone"><code>SecondOrderCone</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, y[1:3] in SecondOrderCone())
+3-element Vector{VariableRef}:
+ y[1]
+ y[2]
+ y[3]</code></pre><p>For contrast, the standard syntax is as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; @variable(model, x[1:3])
+3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+
+julia&gt; @constraint(model, x in SecondOrderCone())
+[x[1], x[2], x[3]] ∈ MathOptInterface.SecondOrderCone(3)</code></pre><p>An alternate syntax to <code>x in Set</code> is to use the <code>set</code> keyword of <a href="../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>. This is most useful when creating anonymous variables:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; x = @variable(model, [1:3], set = SecondOrderCone())
+3-element Vector{VariableRef}:
+ _[1]
+ _[2]
+ _[3]</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>You cannot delete the constraint associated with a variable constrained on creation.</p></div></div><h3 id="Example:-positive-semidefinite-variables"><a class="docs-heading-anchor" href="#Example:-positive-semidefinite-variables">Example: positive semidefinite variables</a><a id="Example:-positive-semidefinite-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Example:-positive-semidefinite-variables" title="Permalink"></a></h3><p>An alternative to the syntax in <a href="#Semidefinite-variables">Semidefinite variables</a>, declare a matrix of JuMP variables to be positive semidefinite using <a href="../../api/JuMP/#JuMP.PSDCone"><code>PSDCone</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2] in PSDCone())
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ x[1,1]  x[1,2]
+ x[1,2]  x[2,2]</code></pre><h3 id="Example:-symmetric-variables"><a class="docs-heading-anchor" href="#Example:-symmetric-variables">Example: symmetric variables</a><a id="Example:-symmetric-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Example:-symmetric-variables" title="Permalink"></a></h3><p>As an alternative to the syntax in <a href="#Symmetric-variables">Symmetric variables</a>, declare a matrix of JuMP variables to be symmetric using <a href="../../api/JuMP/#JuMP.SymmetricMatrixSpace"><code>SymmetricMatrixSpace</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2] in SymmetricMatrixSpace())
+2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:
+ x[1,1]  x[1,2]
+ x[1,2]  x[2,2]</code></pre><h3 id="Example:-skew-symmetric-variables"><a class="docs-heading-anchor" href="#Example:-skew-symmetric-variables">Example: skew-symmetric variables</a><a id="Example:-skew-symmetric-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Example:-skew-symmetric-variables" title="Permalink"></a></h3><p>Declare a matrix of JuMP variables to be skew-symmetric using <a href="../../api/JuMP/#JuMP.SkewSymmetricMatrixSpace"><code>SkewSymmetricMatrixSpace</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2] in SkewSymmetricMatrixSpace())
+2×2 Matrix{AffExpr}:
+ 0        x[1,2]
+ -x[1,2]  0</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Even though <code>x</code> is a 2 by 2 matrix, only one decision variable is added to <code>model</code>; the remaining elements in <code>x</code> are linear transformations of the single variable.</p></div></div><p>Because the returned matrix <code>x</code> is <code>Matrix{AffExpr}</code>, you cannot use variable-related functions on its elements:</p><pre><code class="language-julia-repl hljs">julia&gt; set_lower_bound(x[1, 2], 0.0)
+ERROR: MethodError: no method matching set_lower_bound(::AffExpr, ::Float64)
+[...]</code></pre><p>Instead, you can convert an upper-triangular elements to a variable as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; to_variable(x::AffExpr) = first(keys(x.terms))
+to_variable (generic function with 1 method)
+
+julia&gt; to_variable(x[1, 2])
+x[1,2]
+
+julia&gt; set_lower_bound(to_variable(x[1, 2]), 0.0)</code></pre><h3 id="Example:-Hermitian-positive-semidefinite-variables"><a class="docs-heading-anchor" href="#Example:-Hermitian-positive-semidefinite-variables">Example: Hermitian positive semidefinite variables</a><a id="Example:-Hermitian-positive-semidefinite-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Example:-Hermitian-positive-semidefinite-variables" title="Permalink"></a></h3><p>Declare a matrix of JuMP variables to be Hermitian positive semidefinite using <a href="../../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, H[1:2, 1:2] in HermitianPSDCone())
+2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(H[1,1])                    real(H[1,2]) + imag(H[1,2]) im
+ real(H[1,2]) - imag(H[1,2]) im  real(H[2,2])</code></pre><p>This adds 4 real variables in the <a href="../../moi/reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>MOI.HermitianPositiveSemidefiniteConeTriangle</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; first(all_constraints(model, Vector{VariableRef}, MOI.HermitianPositiveSemidefiniteConeTriangle))
+[real(H[1,1]), real(H[1,2]), real(H[2,2]), imag(H[1,2])] ∈ MathOptInterface.HermitianPositiveSemidefiniteConeTriangle(2)</code></pre><h3 id="Example:-Hermitian-variables"><a class="docs-heading-anchor" href="#Example:-Hermitian-variables">Example: Hermitian variables</a><a id="Example:-Hermitian-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Example:-Hermitian-variables" title="Permalink"></a></h3><p>Declare a matrix of JuMP variables to be Hermitian using the <code>Hermitian</code> tag:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2], Hermitian)
+2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(x[1,1])                    real(x[1,2]) + imag(x[1,2]) im
+ real(x[1,2]) - imag(x[1,2]) im  real(x[2,2])</code></pre><p>This is equivalent to declaring the variable in <a href="../../api/JuMP/#JuMP.HermitianMatrixSpace"><code>HermitianMatrixSpace</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x[1:2, 1:2] in HermitianMatrixSpace())
+2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:
+ real(x[1,1])                    real(x[1,2]) + imag(x[1,2]) im
+ real(x[1,2]) - imag(x[1,2]) im  real(x[2,2])</code></pre><h3 id="Why-use-variables-constrained-on-creation?"><a class="docs-heading-anchor" href="#Why-use-variables-constrained-on-creation?">Why use variables constrained on creation?</a><a id="Why-use-variables-constrained-on-creation?-1"></a><a class="docs-heading-anchor-permalink" href="#Why-use-variables-constrained-on-creation?" title="Permalink"></a></h3><p>For most users, it does not matter if you use the constrained on creation syntax. Therefore, use whatever syntax you find most convenient.</p><p>However, if you use <a href="../../api/JuMP/#JuMP.direct_model"><code>direct_model</code></a>, you may be forced to use the constrained on creation syntax.</p><p>The technical difference between variables constrained on creation and the standard JuMP syntax is that variables constrained on creation calls <code>MOI.add_constrained_variables</code>, while the standard JuMP syntax calls <code>MOI.add_variables</code> and then <code>MOI.add_constraint</code>.</p><p>Consult the implementation of solver package you are using to see if your solver requires <code>MOI.add_constrained_variables</code>.</p><h2 id="Parameters"><a class="docs-heading-anchor" href="#Parameters">Parameters</a><a id="Parameters-1"></a><a class="docs-heading-anchor-permalink" href="#Parameters" title="Permalink"></a></h2><p>Some solvers have explicit support for parameters, which are constants in the model that can be efficiently updated between solves.</p><p>JuMP implements parameters by a decision variable constrained on creation to the <a href="../../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a> set.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, p[i = 1:2] in Parameter(i))
+2-element Vector{VariableRef}:
+ p[1]
+ p[2]</code></pre><p>Create anonymous parameters using the <code>set</code> keyword:</p><pre><code class="language-julia-repl hljs">julia&gt; anon_parameter = @variable(model, set = Parameter(1.0))
+_[4]</code></pre><p>Use <a href="../../api/JuMP/#JuMP.parameter_value"><code>parameter_value</code></a> and <a href="../../api/JuMP/#JuMP.set_parameter_value"><code>set_parameter_value</code></a> to query or update the value of a parameter.</p><pre><code class="language-julia-repl hljs">julia&gt; parameter_value.(p)
+2-element Vector{Float64}:
+ 1.0
+ 2.0
+
+julia&gt; set_parameter_value(p[2], 3.0)
+
+julia&gt; parameter_value.(p)
+2-element Vector{Float64}:
+ 1.0
+ 3.0</code></pre><h3 id="Limitations"><a class="docs-heading-anchor" href="#Limitations">Limitations</a><a id="Limitations-1"></a><a class="docs-heading-anchor-permalink" href="#Limitations" title="Permalink"></a></h3><p>Parameters are implemented as decision variables belonging to the <a href="../../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a> set. If the solver supports the <a href="../../moi/reference/standard_form/#MathOptInterface.Parameter"><code>MOI.Parameter</code></a> set, it may decide to replace all instances of the parameter variable by the associated constant. If the solver does not support parameters, it will add the parameter as a decision variable with fixed bounds.</p><p>The most important implication of this design is that JuMP treats a parameter multiplied by a decision variable as a quadratic expression, even though it is equivalent to a linear expression.</p><pre><code class="language-julia-repl hljs">julia&gt; model = Model();
+
+julia&gt; @variable(model, x);
+
+julia&gt; @variable(model, p in Parameter(2));
+
+julia&gt; px = @expression(model, p * x)
+p*x
+
+julia&gt; typeof(px)
+QuadExpr (alias for GenericQuadExpr{Float64, GenericVariableRef{Float64}})</code></pre><h3 id="When-to-use-a-parameter"><a class="docs-heading-anchor" href="#When-to-use-a-parameter">When to use a parameter</a><a id="When-to-use-a-parameter-1"></a><a class="docs-heading-anchor-permalink" href="#When-to-use-a-parameter" title="Permalink"></a></h3><p>Parameters are most useful when solving nonlinear models in a sequence:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP, Ipopt</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(Ipopt.Optimizer);</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_silent(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, p in Parameter(1.0))</code><code class="nohighlight hljs ansi" style="display:block;">p</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, (x - p)^2)</code><code class="nohighlight hljs ansi" style="display:block;">x² - 2 p*x + p²</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; value(x)</code><code class="nohighlight hljs ansi" style="display:block;">1.0</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_parameter_value(p, 5.0)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; value(x)</code><code class="nohighlight hljs ansi" style="display:block;">5.0</code></pre><p>Using parameters can be faster than creating a new model from scratch with updated data because JuMP is able to avoid repeating a number of steps in processing the model before handing it off to the solver.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../models/">« Models</a><a class="docs-footer-nextpage" href="../constraints/">Constraints »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/assets/latency.png b/previews/PR3525/moi/assets/latency.png
new file mode 100755
index 0000000000000000000000000000000000000000..9b51f3b25fc63c559400a3fe628a33348b6e30ff
GIT binary patch
literal 25175
zcmc$`2UJsC*Di{PilPEGRH~o?QWd2)6+yay^bP`s7J4s6R0Jt6(nPA#q=VACB1Njw
zd!&ROkOT-JgxnoOv-!U7+;RT@-1Fae42F5Lv)A5x?KS6o<};tUf*z>Ioj7*!7#SJa
z3Hkdn>SSaG*~rMqnJEtg-<UHT%p@Z_APu>D_ksM~yDSf!pq3C@3o<f}2y+vYQ}?+!
zT1-t%Oj<g)IFC8GtG|31qHf~fTK~EJ;8(NbAA`O%9tt~fKvMe3asG>EeX5%3Dc|!g
zK5f@?nC%^oJr>&s?XwUzM4WQ3e;7_)uCD$n61K+7qN2ddaSa$$iR@!^o8mF)rE@I$
zX_RvuEZ&D-CZ04cI8YSHvUJeo)}`B>(i>-4^jSpMt{-agBY(7b`BG3OThY6Xr(|-4
z5kc`qSqBbezCNj-&ZQt-?Cf{jLe0`kHs~oE|1<To;^^1wE@wiHhaA6fUV7q$8XkLC
znBhXBFpXK<*74&Rm!v<ZQWo{bA3r`4F2GOecjskRvbhNj`IfuC?rk$Q*3knYaGl_u
zkWd!eW#sYWxjj1s0z+qgeg3!lUzjZdp{|TTAh=&Rco5BV_{Bju+4w&ud{+YJ#DFQ)
zve1#YR8}Ut27IO@I}ivVqX0e~0RAolf57a&enEBw`1cg>CzC?{=jcJUl!JdhlQR>4
zcw6JHygcx)hPji4g}w7*s7oewrvWfjKjfi~i;l9As5#V**Ypw8%!1d$&VhIenYf22
z@X^k~#gxUv&eq;p)I)-G_ZOnTXX4j<tSq}faj}tL)lq)Hau@1k!6L}Z&&$s$d5ndH
zMcnC;rKq}$?B40X|4Fbuc5!hK<>PaAcjt8%;DtI_@m&`Y5#i&%!FS^Z5AX{fXHR<<
zQx6_{XSUyi{252a!r9yj;@|>-+OrVHH8q2}x=66H5-;@kzu)(1;Q=9C$=-SIv498U
zBOc+q&dbmDcWmHPapJe44<H^Ewz@J9J79W%Ye?R@DImW4`+qw^x+3UI9nzV?{QJ%X
z9r@QYHJvS-?n3Q=OS(vs-pt-*!3XzF6z3y8IcO$+PxJ1#z&uMH6X*N;&Lof3IGw&q
zMkY-rFLV2$#{uM+{|Cc?M2-#Du8Pv$rdxMDa%zXi>9IUALY@~_9fZ&~)4oe}QAbe<
zHnu#M-cgU2qHuF~Npbc3^o^^6iC>DBr4Am@zh$b#;(MIq*Dv89c$@P&t2I}FPHL98
zGFwf5%z`5_?9pOU&-&PGt}|u`Gs;{kid$G1?@B7#7`Mr*=^sw6>8ka%S}Eyo&+=cj
zBqVRk8OxWUZlbrGvj~Hq%22!?#14>CP|~nSKOuuD^ZhCyAmZog;a;q+Rbw{=1R_Q8
zM$n#>V#dqB@x33Kuu-;8E3D-liTI#wmC-XqS)zXA&k(>*1S_O7PgAo69C<z1r{X7O
z#!5Q#iS!~ldSE8d;yBZt>T{4pg9Ce4CXW0}I&)Q1D-(X??%yi`x3qESZ8#IwH0?sK
z-<2rPNcH`v&lg_^!ZV2nY%19-CLY{)^XLAa)GO2Kc+lRk6?lIDrz~>v*p<KM6&^a&
z<lWV3cfXop=&2RV@G9xZGI`C9^MMUr#WS-fyvmp5Zo^3TR+>_OC!RJLV)wZL8hdni
zu84O$Dh_)t$QB?vr@T1BNOk2jaSnIqXi%Lz;3_Ls4R(fhu$ew$H2C7Il$0BnmVP4X
zWLr+y=lF=YhvWa8Uvp4o^b=G3``ABcj<%<Le)SJZGOY)>KpgxecyLj2{Des$Wb;Ay
zJmy^O(6H$7ji|rRP6k^)H%HAZaZa-5qC{e@yJnQ!-*KPR_n$g19vULqb7O9zUvzsT
z=I?3s9jE5nLxgDOxkH|Me9BHU0KL%Cr@+%kd>&s=w&7xM#ohlqp2<N<?rO^G<T$`J
zXa9M27HN}3N|&m!qC`f*V;k<)W4kl5H#HgT54?U{Ln{O4ta(h!{2l3uv_3iF(j&{k
z#26tqJNdTQzc)q(yFzi;aND}RmLl8)bW<5z55_~H4AAkkY=v~kok2&2PtUno1;$Z(
z_bgz<LyDVutUw2YrR(5ruc*41J!!sv){vKX;OfMTx@~LamWu9seFB_G5|eeN@e$px
zeHA$ebwNY)8>VR2TnI7?Oos4;JS|^#Rbv4?qP~2J*QJlY;`(hfQ{CZtF55`1zei#5
zS7_J@jVqk4xK<@=Ykn_T=@32D3R<0w6f@pvb8qpHm3yz;PX9d*_%M8~MR^cCx0pWW
zwXED`rDqq}mJFJsDA5XbIEv*IT4kpTG^=E={zEebCnS5WQD3&8{Mt{YMLOpIxxPM&
zXjkx&h#w92{oh&T<?)!Wkd`OSV@(z&i3Mu53Lzd<w!GlB1k$N61(xBEUiQ3!#fvfE
zGn7-XtdFz~ZuI9`D3M-m{Tu028)LS@@Ohm}q(zc)1!iO(($KkGeVbbUV+;@IVQ;}6
z3#bzw+GvCbIPIX#OAL!iOYTX%?3oWzFJ<s+86~vguSID|HusFY<0>n*qs0TG7w?da
zJOvt&UOJ`v0w;Rg^3lOwg|ok%-R|MRHS!9+JZt7QstA=bn-3An{=G=j+k~B9`L2q$
zrr1!AR;ioqyZ@P=9rOSaFqCHh7b`vN{MUU`zHtau6|pUIV@oM#=zR_T7zYeAb9o%!
z<K{z{WXuR+9cnaXtiiq{7bKnHOMG(4AHm&g8GW@Cr$9H_Lf%1lDc+32O^#%<WUx@a
z??|`iYgxwd>nFKL?uxSb$^*$C*PPi3%dYSkgO+X+g@USH>UeRJS>9hK27E9{q!7G!
za6_Vu(U_gLk%pyEm%`G>f4S1_O1JTf;DF|nue)QLP*Fy8eQM305o9|o!F_%Pd_nf_
z$WFn^4W#FJNVt9{=*StEU!C^398V?+M@#Pdea>K*=8gPR7n4_y^<CX0&?(%f<`6<`
z>Fjwgmu|}4HYk2h26_ZarMoO;G+9DsE7OpRL#+X%xjjJMcI+$Vk~vg%Yq)@q>gsFK
zA=o@+J7y*)q;(`hrA}!4y{{+`Z%Ky_kSoLD76kH8ZkI6`2!Xt-5w4&sX7GLgCV`DP
z644@o$+(5TGYwjeWU!3-?UK({dRInmB3MmTNUL5thrFZnljz05vIQQncX~nIQTqwi
za9`l%S<nKd$l#-~lU?Xymh5?3Tf_h|sy8eT{Aq&blz%+IZ_hSgUCL%5*H`@c4P!K@
zCszHU#+Oj?T-G^o8D6`<U{+P$Y0xE@(`j&L_x20ZDK6GPZqY~Ba#De!%ai)bqiYw<
z8QgXQo&DA^Qh+Xee&}4uh3lOy=OpQuME~9&%=L^;uVvbW!Z260Gk?DY9LW8YA(_iU
zCPEarbD}$L(SK*OUimZ!-<{Uclt3Zm0}fDN?myMy3(*uaFnZ}yX?Qo(Hcz8|{M26)
zE*Bxk_o^apU$*`}j|r*wb3XZ8oir7;Jg}W#Kg=ft^0r;}g9;oucI>_%D<@RWDZnim
zQWBm|0eXL;0*OBY?mu>Uo65pUdQ|Br()8p%EYp16pT#aJ=bW(cB;5x~;eEi9T@Dxc
z$Pv1K?HQ(wAj1$9oEJluDv+ClD!P{nqbZiXuzDp`>~IfjF62fqP87ga<`6Sqf*cj2
zTY|8&1-D;Cba~U)(vVn<Q@?|;);W&k$I%tEgW7_TPyfyu8H|%MDydku2BLgPRTpfa
zKPwO2G^w+{K@a+}@Ei=~sWl{6c0qUZMJ2^sQuqQ4dYF7{**4V%o)j1XVQUKqU80d3
z?+4}Pr+)eY3cpUf&uOsoM>Zu(0IuiQCneIf(>y-5h63VXq}EqgU<w6=3PG!!43<SX
zeU5cqb}3O@`HFEI=>DWlL}2G(zDphVjdAOU>aQeE&BA{Lh8_2b*6<Vq5KNPPv1%5>
z5ctfLOZ*zX8Jgo^)8oZH8GSS#6+m`HMV>es3H782T5^W!@kFsIf?hS{O^V;)?z7RW
zGe^N@r+4<TTT<W9@e_kx{(iZdH{(tZwvcDK9XEo;c=oR22bu`lfuRP3(gnA<9i;im
zA|_ewGREBCko7(l$50WEpXt`PK6^IC-;2LcCuPW8%kMnoFwgjXN{~TN*gqsj#$+j<
z4A<26rJ!hGqDj7fBw|qd;D%<AJSbe$hf>`H9H1a;#VYwUDYyW@w~@T2>3m=!)-!;K
z*YeaI(hSrW?-^-v5CV+tF2nBG$-X@hDR@ej^kXs@{n^KYwBJq)p7`vS`eENKS-?=N
zY3)LdDS=!orDs46c+?g)8mp!G9CZi4pBJsCEt$_CNXu*R*q)Vcl=ZA!fcbJUcz`hB
zPfMz`z;Gt2WC$Yj^nM6K*RmpuIxX)h_L{yJr<CLCas~9#e}~08sXYvbXlcPyI!yCY
zkPUG<uf<;ji21XnQ+7JvBI{qAEnh^~aPGW<5aa?>c^2yewoumYW;5%vTJdW!_uVJk
zqgykU2EGX;DNLX?YG2Kld2eLu7(=q@W4E=2WySYiWU(~*-hC#kNAq9JK*5%Up9f}N
zCKeNYn9ZC)3v}SIVv5gu%^6t}sG?piXu?8{{<hoDMm-S3c~UQW)He9=(&D4^^Poj)
zFP-9(r1`PMPWHH$4u3n?(o|DM675`&BWV6^rm_Uk5e~p=3%Iq=;*{@eeVb)+!4nr;
zuXNFW<9oaoSNjbTmXuE6{N4tY1>!yJ03__`;AlO`p92q{!TOK>%U=2-cht>j9}Cdx
zB7-3)%qTXBgxGwM1F`5mhqXHg^-)B4s;PoN-p_>7Lw`u7J3ZQ$tG%dKT94BL-miC0
z<+@alC;?m4q@s7Fu;P#k@{RnevC?&aLJ8Idu&~s`e7<n<SlyRM=sH|&>8@`u9*d8F
zh8TR|{K6zE7Q10I7(Fh5>TBgGttv;htt;ms*x}xn7{!07ml%}tEWSELwHGzCuIxqU
zK^9C#NjCB_*u~eMINEMZPhh5o;s;oTzM7hYUPHfP%8;j)+`*0Y_u41_E;v{o&-W!U
zE~${40i}Zf)PO8Fg^>u({C9rEX^C|@fhYPBF+y{k7^Iih0=+U|t^A1B`a~<Q-`EOH
zXMhXXgg6eyB?|#LnB<iu3DULKX@M~NRhVt@V8I#Eo$4E!sKrNE)@vR4NC2(5Vz%$L
ztgq<{>7UjxT33LA*8{q}*z3yJ2LAoke{kK|C|*7p(Xsq@vU)u+*Z*Gm0681XZy^&%
zAYh4pU6e~~j8yy1{LU=t#qE<3ovdQN4(YOMhOQev@FyR*Qb<Vk3nl<qk+%`c+(wLm
zQO!XD+Olmqx*VRlSM5e1yRp%@E&4QWEaB|{j9;V-r!Km^$u%Eq?8D-_6%fdkS%xuG
zT7_5mGHWjujUmRgvv018HxvrX8b##W8>rgE{SIC_dD4DDuaoZs5N<SO%t|jOJu(aw
z)=H>cEJafRjFt{K5J#l$2nhWhAXlW4iH|@9X{B|?_jhZ7P5sULJr59E(H2Y^4>)Bv
zhdqBnwn=!OwZ%wdWpdg>>&Ih?eL!#f{AmseNT^rO#aAW!jLG%X_b0xIDQ8J`Vv<7z
zpy=Sc_5x0e)LuGnP2j{Tl<O~!9x4x1>>&9bTP6TV3zZ6Z(#ZJY!h-gtQfFatH+Poa
z3FT?Mu>glkjg(k;2;iAm^MOezs(^t}Ev3@Ezc~6Dw)GkK+1rlp;pq4j{1&z=oTsd7
zlG0Wb7I4H!VHN%kSC%PEFvs?WJ%{qM`E;-C*r4YDK&o@*tY>eTag@oi1CgzvTQ@U5
zG^?-kLD(DR2K(%e<Xl$Mdb3nh$R*+8mcPXp!)9CTx6!$_98sWUac`8FLCo-D!HlVT
zK2MRMW75Ofq7!-Cr{1_M(OtQCILf^IVYLI%Aj(S)&3FgY;s1CbKbw`LVNzsbDRew!
zncY=PTqE%@UB!aSpVb9Oh7>QC66uZNp=K@8t`N%W_~g6+dTLlG7r(#V_G!C<dxFbF
z`%Y6JItnX+n<nswaxI-C*E^W@2g^T>T&@@cIi8;x$AAU5g%0d0wqKZdApF<RV0mmr
zXHnDt#CMUw+ne;VeqQ~91HP4L>_w+tI1v9ZW#If&p@0<hdxNr9F7g~{tP@uXKXklw
zlojk<Rl5LsOs(JkB4f;141@&gl}SKABmCtT(&DwH-30<xJF#Uum30IuzqOcgeB3~?
zBEC7PCLMDdanvBZ-jL6j{z{z-2Hl6srw+@YrtZo(MK`eyzxOP*n3icVLNa6XMB_tM
z<3qWt$urQ8QHCf!3!#S8^1;$2Mb6qd<1A=UxDVS>JH}7zki-_=I5T+%%N#v-c2-QP
zLc*|i>Qj-6F?M1q)k{WVQNuAOP}hiN6PbjcuSI9d9%zK#vhOgLXnhp-+|sk{^4WVt
zfFq>)bFIKXZl{N^tq*Z?bD-{?cRO**DKb2067+0GiH>b6R(IF0=})$d$dhJDNH;~w
zXN@Bn;yU^7Ied@Ye<z~=7)-1T7iuMHrF57>37pl^a1c{?&xD9m7W)A721VuAvd+xK
zkPydxM$H2Q99A4XtcC}I1$lS=s}H=oFRo|YI%l&lgI&CwasVfKo9o42L$24N<4CqT
z8#rqDH6`#+1K7<FuQ?<;P%A;rhIDM(^RUE#3S<4gf@w9o)fK?9FygNC6`Zf<QMRY8
zT7?gFzpjby?&s`5Co}73zCLr;#GH}=h|8jWTk|eUpXr>Rmug3g`gvWgAq@M7mY8p{
zLAQC$;_uUsNMOgQM8<mh+=|4J@bB+%M_jXywGd<y!w`iZjC(~|d0j()-3VRy$fmi9
z5x+k7N*wiSZGHWtay#3yk7&#cZ?)#3=GtWmCo2j0;ka7<(%No?s0$g_HP;JrH4tUp
zrTP+nE6)whDJII0BjFhs_i7Q*7v@K;xokw};t_oxKker2^h!jh2tagEfPWFv@zo6c
zJ7Hw7j=!j=Sl6w?k7VjBg|I_^QIXKE3H4i~RTM90QY&PPRQi+6K$8qh*g@gbmU|z$
z5bFA*7xa>5&dg0G`LqSxUoy3_tai`qs2JD)1tM5HFW{B{%#rnSfc4%N@JrM>^fuAT
z^`;55PajOh96$8d*9s9BVM)V9x<?91CF4Jt$t_=V5%7Y#owCDn>c_=GRq0cp#01EW
zL&uBXhTb3MP~W$V=WaC#FSU2kuYHHyCgmB@IPwlbwomb<2@wE|q>5j&#d#^P6yBy#
zNvL?O;@7Jhi>g*)N3F1A7&0^P+@5dvp{@Yng={@BN6c#RVrVMo%xYE8uSn1N?y)Xa
zr2E5Y*OioCL?<*~YEWQc85Q{ganCj?u2W_#Kb`1&J{kw`^nBR4QoW>_JCIrUa=v3M
zHrq4G-viFohx#JL%*UN?mSr@v(^9vLLH)?C&1sp*xO@~XO#NleQfU_DSk6c1Gtd?=
zb<q#HT>F}-1MPG?HhUmb%0n2B+O(RFaEE8T{#v`ntiQ9hL(Cr{(r=vEwOe|H-|qEA
zc3vxZfj}tUD$6(>bh}#!eVb|{2a4P#yY3YzD!`~?vPo(1pwce)=-#@=J$AS}K~j;t
zd+4?AT}U}#Ve}U5rdg>>-0P$eQy53F)gTxCysP-VJN>>iOYdLOELZ14pxgZCHyh$W
zWJ7L{hWkHZL$W@%oDVz;T5H3n_SlV)*D-v5ALszNIXS@TsX_J3i@-S*OQ8z+#lnKq
z#UNN_>q_J{9N#F%n|yBo{0tY)T*X9@lA+SyThvb!Bl$xZ6E3l7Hcp7OKHT-NO|=$t
zN3>*W{<eqhC$@BRk#LN0yy_9^>L_%pV`#N?oB_sc8oOhon)yMs*Oi~AH|H83xpQ1Z
z#&B%49zMo-%q;j9JDi?-pn>htrLNmPwo(|R)HklG>?a(AcWll9Ew^^O5#RMa+n{MV
za_XX+odebt$Cnz``n-?Smb9EG68?38A;3H7Jd~rn|CmwN&T<XE@JWDM@VXx~RfK8D
zUod!OUYRh`Nj1_lKH(TrsntBOfGY8uTMA`dL^#v!)_r=;07VOWL)=D4y`-{<Cux<i
z@av|)S3>Vi-=<kdid{7WIenNGg=P2Vr@--IJ<}LY&_pZ4jQj-Dy}E<TRh-X!ojKf3
ziib2T{AXdKchbgR)<kAzO>><Fy*Em3cAz{FUUOmIrmi+2Rv2_B<1#-yzQE}kR8AB>
z?(u4aGn$PabQs`Wz_{z;oqdVW?LF8P;k&Gy^IZLrJjrd4!Cvg~r-T1j`P0WI2H9DT
zg4TP=3E1j*v85k+(Rhv7XEQ=jL9a~Q3P@zunZ4b{S~X~3@jNuMK1x-R!j&Jc#V-5#
zoza?P<BnN$OqB_Jj_^}-ukqpW_3$VyYMku5kC{pK%L*<h$iWQ~OwOw>(B(uxZOewU
zD){O)CyevmdKOhM(ae&Pj`4`xtM(2DVL2t?tP5e0VFP;kaRVAVz|W+QeG;xkl<%x{
z5k4Gywbj=Uzm6l~9~5J9dSdcu(&T8+)K*K?%2!K8m7)&cSyi43e7n@4U|2n|vb)u_
zGe5;CYZjih3l&41J$gy#wtnX9@60kDV=@9l#PssNeVJ@_ZY9V`6nave3l&gj4)aAX
zWk`S+4U^QPQFNCq?7ugJx;^V9Mb7%GXNkP7x7HtCx7*yV4CDbh_Oi>fi)-GpCh!Mx
z_bot2)&I|UT`!3MCQ%;Jcr=fRp)N|b`2EcqNC|Ww7yOZ!?KM*_by9&=Z~t8)t9v%}
zKl?GPE#%v{7HlzM1;~W*<gT}}6kehj_qUT|^l|{a(bjyFTZs1jU;y9D6J$w!Y&K#l
zGd$h=h*A5<v}bfi%%h-2j@cykCc?lg4Y9)q&+-&~<JkNOi|rm~tLjFqK}1~$h54z2
z!(R$_)H2!Oiu1i;)v7sfSbdiiecy`Vw0vE&+q6%H&f*i@YX$qZwIg-kWURGYzqV}t
z?0Au3an&x#XXSZujE&{A$b3S1CDLTrZGHNK5nY+Dh(y%PVT<W0dPI%)w3WzCr_03m
zt{fM*w-xSnv-j+CBm}A_><+LkLakNYG7_b}xF0v&*R?rY;UaFGu@84KMIr4WfI&iB
z_!=J(>Jil&^cO4DpQ8a1TPlV}RitVNpbG#pTLnW@JZ%8exqq+865ZQ>D#%JnHkktD
zA=eY?iig2ShDPVrF16J%6%ukJ_lm{{Xo*Uj0QEz_$@%Zh5o<)82UkIei$*6>+WwL{
z^y~Pyp0}u?eRN{J9E+?8v$sI^fFIs#I}OC+n1qs}Yp~XBD7UnpmC&-s#^$wxY#>Uu
z{usGHpOqevd+t_Xxn2f=N_Qs5{?gJ;P(6(n&2#(i=eEht@K1rf*X{omFK4N!xS_B6
z7~oqpq$Iq#R=%7Vz~{#n5aG|bgXE3DO>)QDOaacs7BqmV7z=*UtEMZuOGzE$T4hJ`
zO?Ss}%}>h(U^zZNd#=60NTIT2i0m54oTw1|t^JsWk9{);FT>QRTez6%EI)gnoswPa
z>2DV7tJgsUWb<Fgs388U4RuO&BV})SG9zbxO$GQWu1E?b9@magL~J%q!adQ#UrS7G
zBBuDAEw_d5Nlu8*zpWT`O%q*Vof>c^{3@CHjh#$kYOw-U1!>OiJ_&9+KPT4aF#d*#
z?tvIz@$7DGzfF(f=*6=d3=Liq@3)1}8?imnw|(*#Cro$ND}CRhcU;^RVlhjqu^3;i
z^u(N<b!uX)vvv8y$MwrfnJti*aFa;NUS=7ySRh}&e5Bu=v}k|{`weLYcTRq|1d5ks
z@1+Z_H<?=bji&^bO~M|61OsJn>)#8qHC?FaJNa&lgqq855jl9s;nBBM>HVNL4cwIF
z>4?!MV>C<j_>?T>zZZdZuskq&7FBVZTBt#-gqn1U{efMsz3mUzeoojvn`BZyW9a5)
zC+vyofl?;i=47j1+e#|q-+~!X={b1p{}SD94AryaXp8)7@no=j!~_<A_-`VJsb6ub
zKF?3Uk)e2{G>{v(f{sk@q&fRrPxYPrr*s;NyH_g#;ud1Hbq8=rkFL$)ijEk!>o++v
zCx;~)ZAvhR(Lh>@B^ad=t9tY&FsNy7zBKI>->ZxO#0%cKC(^7cW!RyL^r&p*?vC<`
z8eK#SGh>)4K4qhy-<47Ye3-XHWs&$xE{mJ2?D;#52jS<b>JpY2^x+{~j^dM>(QtPW
z_s-1cQ;f}spRk$n@R4Rm`UJ567lcRia2BItxY$`lY*z23eoT^KVR-c{Z*??prfQD4
zF@nz23VtnKYgW^hY384byCT~Q9M=Z$>BZ=lJ0jYwk{y~y7Y51`>CSa8DOa7lQMj8i
z(Ho_VFxJWSnLKJ>zHyeMlqwXVXl2*4u&3DQa5O+3A}xj!e7}?Uk_nsKr=+z&X{&JH
z#Gnwuv_ymSrC^dYRy?PdyX?aCIn#etAlSQ?{wRTVHQe_ebpm;#6v{PR%}1RMG|Njx
zcZ6_EH#m={u$MNu8Y<-tsFn`%+mR@!C-n=5ys#bj=jFLd7l=;PS4*9=;#s7Bic&aH
zMR>7os3hf^+sK14WuYLc)K?-u49;H{PrX^`(4KGGtGweYogzu~U|@0OoO_c6=u?2w
zEmTjkN3SID8R^j}b9Mhf$8SsR9xHmM1(H1R6}|pTLoNet1KojvR_qzNb|&9!O6IjA
z?Fn-DwF&D}`8Sk%9S4rcu`e&)Jg%i1x~Sj=RD;$QY66rqkdB6DBRq6r@a3W8Wify2
zC`-V*<{<pM+|CF>CPr(+d&+8Y1Dpu>&&OU<SwGBtCm#8-^vJ#~D&LY_M+{y~R%Az4
zD@Y|6x6Td*8NfH~zM{Ta__-G$CwvR04Wv|vWL6`#V)TqB6TT;~=Lh0c-()8;KX3GF
zqccEwOQ4p;7S#h9!s}jdZ2xNgIUJ=J0{4s?XGqd+6cKN8nO~mm`BAV^pfxr63Yk5H
zY5M6kUhQUZ(P*V@!@>v^wk}NFDndUyV{kGjKc1OUS!&GPv0PqqL&~FEFQ2&})tDeZ
zOu&Dxz|AZx&exhQ3`%CCR?&4vw?WDiFe}qUryCn_wKW8ttyeoWWxaEMy}+N6W-yl=
zaj~cuKV&(!(ot@V9ypu~iS|Y7rQLr6;#R+BOMgP=h8<qY^V+8cdRPC&8-H7xSPd-#
z0qiIBYlq%<+{a@Ztn_+>V%YvJ{sZKvDHf<%X~%Sx4{jXAQtms)^;ifd!-8UCaXJ!A
zb(DYv3xW#zpzO#&g8bER$fnYw*XF)ho=5!p=E45AMMw}bWGPIjfA1)}&``uz_70ri
zP>#No_M}yxIV@34t2_c)^@Y|%_Z(eXO*8LO_2Qy#Nz%Wp&*ZHv@s~1Jcj-l&{$g-g
z$@wG1YgJ{x3c{O|#eDz3L#Nqaqud^q4!!qPKXh`ePJ+me0pKr_{hqP5&h$EwR8L~B
z94Hxc3ZiE6;FOd4WboQzD&Jo0O{t-Th=MbqjN(L|&kH-3VZq<(5KN_+=<G*d5vDOZ
zrZ1RNL+2LSC*eoktXi;LHSE37xs9XuF&Bxst;prb#3=rctgG0@&sUxz1YWaXc+;u6
zh-;DrG{D7YBtr6YGX}d5nSTK4JnEaE&QfQwgyHB0f`||E2A1A^a*m!_?Fm5t#K_e~
z)S^2FvYp(ShLR>l-5!@4b$@kKtO#=)e{R>i-Wv{Aat<7Coq*q0W!lucHW2nwY_n4+
zXU%atUPGR4BEUx@8h#fm%ooMjyxRFAL#&5Rwr%~nI!1WLBG`Faj(VHBKt)n}a+_~`
z>y@oqhCVWZb6t>K`sWtK<i(ri8>52~@b&0xuJq)Iah-FwFQZlf$7P%P04(N^{+$in
zxeL|nJPLXZ{r3QoN18dshO47CsfZ8b0W?)%LVe#0R$`>Ckde@X4)ZP{J1fMtIKbiN
z>!!vDa#Xf5FhC&Td3e?G5gXGLSJIj;JVXKD5b`5Kyre}nNJ)-B=c>nU|M!Sl>w3xm
z|7ieAT+Qk^WMr#|<BY;tA5^N}mV^4!Maoyz84;>AEiLFi@5M`zu`7Tw=IrXk#845%
ztedd{rti+iEIv0Qp*Cq_8bGV3-I==*FG=Yu)?7cABdU!nb6hOHo)(jTFM)Y|s>Y?Y
zq`A6g#mp<M)H9)q-9Fw&1&?|~ohCQ#o2lSy4{zpV^yl|otg4fT?##zpixhwIK;i20
zZ-k+1-gc%fxO)TMwE39$r&%jv5*@mr%bK&55#H0OYF{u(Ew!T%q~`>n(<`ZhwlYh`
zO3sT98id`${P67aJ!_FJ19EjLQXPwkVILPvw2KJB8RC%_=A8W&<@jp;acxthV|IoQ
zlW!$%!-WuD8|Q|w8UE&T@71Cg!>2(%AoggWh2?b8xO&osp<Uv!ezdcQU&65;skKe7
zSbSAYnoQV~^WTV%dNF{AN@`u-{`UTB_>A*vrvHxIJXevbr+=r<iq-t+@<XJmNS#@#
z({W;jb)9*{V_@!{*j0d~Up)0;*%@^!=dG4wm%Cw{Vs+=f)MU|z;tU86D;{%kQmqAK
zRb)Vn)$jhS0`lsGoq!;l=<m&H&T*asXJ5tFzAuUAq#p3j8NUH4gPU*^ekXUN`G;Bd
z?d@E!Yvtb+FB0qH{cE8J5`=wHPXXHsxG!T<FL^+o1{4PDD1R$oAgTBFnd@G?s}7YC
zI0(LgLn_gypSPNpmA6wRQAqXH6uVG7_v&fLzMt0+^`I38sxA!{(5ZQW3_ssErJ9n{
zIe^Ki?j4vn&<T)UR?HdBuBM4Yy3tYJLIFNl^F4rSwa@_IRcvPRxy~r%I|WtCu0C1)
zCdr2{9YCy>C#niPBJX0L#tg!jrhCJ_t;0P{eQ@{baAI%HEqMH@R~-5|_$&eH_aF<s
z+M}9jp0ietEvd!cdg)w~Fze`}p(i}vG2GW9IHgrpi&;w4>;{Am??eexqoL+n{fk&}
z7dQK;4Y>y7q;Qwvpr&enE9$3#zi9gQNO;!vPW-Iaw%bNV<2t8nU4o>tSkLB4PV_P8
zq|#LN$OcmG6Z+@qP+aXl*=UwzW=Dxo$nqC`&fM6_bIXiL0(Ab{+K}ZDfA#blhq+nL
zw&d9<^mCWhmdyEgH-$!b(%0yyt2Gh&JJ?RK)sYXXBhGLAlh7A>InDAJytsU46ARLv
zmKL*O_53g`+VpC#x}tBt70cN+XYTps^VVg&apl9Cn>R`op49X3@^8``x|z+_^<8PS
zj8mN&H&4}F(~yb=`Umtg;R571S+V;IpaQwT|5i%Fx1*GUSJ*GEy}Bo^h3|@MFTTWs
z(|cbE0FMC0IfdT%{;hH@5|LUTL-)9r^4nC1t><YnkW<eMkS-IO%PTvR{`r2yKIyOK
zpW2+JeH_S`9_-0$6`qNJ);UGL1|a)d@6`g8Zh5#LNOoPLsOWfko0{qc_}9pzE>)jG
z^L@RUWQV6>1VH2519;%*M(d;_w|mQ0-n;}e?s~L;l(f%vGeA#p^ADh(w2V(>k^Px8
z5|#;Y1<J9%iB3Tz;X^WSB2b{2{^jQ-_&V1!xVk6q^Gv~b)r@VEpm@Y$>Dy{EDYGpH
z+r_9(Bu==a?@LVf6j4;g{z=*~%g(8XF2|R?pAX^r_BB!&UKGPc(_ER*!<8s!Ix6md
zRdZ#Ju22b9h;=v4>MT*$YM*jHTNkt?vfaif`H$Ojt;ATYbI?sWv|Ha=KwwYB*y&G6
z)+SM&PMh9YH#%o~gR0Q36QQZ*2!=6yB;bcDk-vfa8Fzb3x;2*Q*uMy(@8GwJMpuo$
znOJ&sgt|niDoNqdQVI~`65;LHsbYcTBhVEbx9^qCL~*<sywTckb5%Rr6P;WZI3l+t
z&|t37olrG3wU)QClW}c!ZUDX7-n?WWiCNGvz%BHoS0#7NPX9bFxA{}pKjHnhx|Re!
zly_x&dSgNSm6MrBhVV)&_N?iuKIg#PX5x5@<I8PjfSurmZC0JZWb_pWtvD|DA2Bo5
z_FlREWatfnw~}dgWOX-<6142-1LsQ0#P)yzc*xway9T3?7|0~1juG|Ct<hy{iGI-P
zd$^{&%C}$?Y(v%l5OCsUM74f0WNi4zLnJ9b8@>34o4PnVUThy4oO1=VtjbD%Gw>FC
z)&D61C?4F<ssLB4rD@L}FIMYES?Q%;!QJ`i3dphk4-^@9BS6}I>EJ(27YhQLZ6Pv5
zhfM^{L^aBNpyMW{J1i<}tgc2+3MdXmu`V6X<e2iQT=jX|_i{_N#<&I(XW2WrmBlEq
zOWzhlxRK+Z+%sN{)F#a2?_?|ZbF8>4HLK3z^OcKyl_qXLuDyty96rf~(<(rB<f@L~
zs`|I%XXRb<sReTQR9ov92&dd#M?M!L7ruSe>tG`U95Hi5Rg`1C02QFB*R}s7`B?qZ
zf93p9q)`N0<6avVHVbcm4kd^qm3EonB>#7D#nJ0r09?lUnRBmjNTnx-KTj{-ETiG|
zcJ?*DXxlUR=yZMz`q~D*H=JPIxTuCM)9r67VSjm&$s_Cau=jMQwdDLDb(n@smnK9D
zRqOEBfmbsqknmM-g$ZkI#D(CN%t+^@8nIIq<C`X+u$us<9J4bC^RMM?_OhDtvB{71
zswE8fL>sqHt-K8a`U`j-<ipzfoE&_8AU|q`oJK~29Xq@uKD*j&C%yOU8Xj#%OumJT
z7!(YodJ=m7$k93iE<r)7;xIYl-u*bBUK|Ii7uQ!(PS+M^JN+qsx<>57;JiTJ#X}~U
zr|)}KZXL0JPu_6@ce`l)sePh)f=HA!Kqermrrp`FL21>D-%bqsX7ou`f_o_D+FfL+
zS>OEt<d834l6b8H<aU(5<zdqWLV1Ew2vB{ho)tFp>L2mWmiITw!Qg`56=v`5!>@Nj
zHbaH`F6d#e9t8>yt)rAnkx|b;1`7PXJF@?SP+nFn3;5@hxp#<?U>13hXbg}96KQmn
z|5b1>H^g;rcn1VV;F;HZ(t!#&P@{UQa+3K0;K^^48fgN$*!hNrPd3@6QEC|gA)Kt}
zDLCeO5PQ!^Ky5ZWEaSss4kl)lu4{3gwKNTnpkuxdF;*ZF7afRSA5;}o(?7t!8NjL8
zRvm_&M^-;~$3>~sUU%3P;{fhfb3_VI?}VrGjpQ|pFO)8ZJ~yXZKe^(CA4p}QuL+5S
zibty>#gCQ=Q=4b=sSbx{%x;_}e0wj4J>39TstFOsY=xg__3(x_n)Wl#I?rXfOfG+R
z%ptc~7cGNc+M)CGoMqR}z8u?QBAMkU#Kql7DA@kBy7p?5r!mT3)gl<)r}bvCoEh&`
zdu|DE>0J5F*ul;OUEyWr0H>NzG;6*H!b?X=immHS*fLXl*BI^W=k6p&1}!1q>dHqY
z4>fD77uDbm*lMNuDHk9LbdNNc*Baoh<Ke%|h6dD)a*aC&CQnOkS*{Cl`Dt_J_p})t
zgswrmPOLmTPD%5DVp4KRACB(`#&99tp08UhH|Fuz7NR#t8>Q{IU&>|j#3Z_i5DZ^>
z_seUHN^zeR;`G%wty0}~RKN;1p25to<9+}lJ}p}&3N?7GIV-pf{`lS=DVKBBGLDoh
zvV1@Drx|9EnA+N#B!XA225pP6z`*~jj2Hi05YcAG4iZZO(y85YZYkfNaxQ)ASQ7{~
z(CDa}WB~0;fEm?5uu3-aT(qF0O3HsoImoiIjkRanK3h^e1}JWQGBi(l@TgzjCvMi<
z)7xs*fWt}sgFjmO{}OwYcK}Cm`8MdTiZ6YiQ_dNBnHI_lHej;dF0&yl_H#;LrdxZO
zIR@Hwhp#W{?II_m6yC*)s7D4wOFcOlNe#s+lv&jEw~|s?JWPO~$qQAZHaB?Z8v%$W
z=c&S@Pq}jpU^nM)kEI%6#D5q@36gitqQV$GuJhg7%9Mj_B-MP<8wbGl;_4kyj}(~`
zC?1>Z;U8PiSTq%*zibo>%wEt~QdAhQQnwR+?tj=dhq=lsCb!E;MdL&3iw@SXl`KZ8
zX31+8XYYADY=bzc$W}V}Y0@_MH+8}p&dn&dweTT7KyPuRKKqJ2H>;<F5YkSp!(twG
z0_7iok4fF!5{&K1>?@_8PZlQdF>HB`7gS{HsWpELr_V~V>%S6)E2WZB_3p`^SmDi&
z4nSw=E(5B}o-AfU-!CuyeEJcaf#lD((hQ(P>Gk0Zs)1z7q#~x~cRUz4Gq&kHD_`d4
zl21%Ow6N<p%qp;A95|!8PU_M)ALUwr=Et6|9%}Cg1m;4wS$4+rW_Gf+a2MCx{LPPG
za)jP}Tzv4TVk7qbf?%q5zt2iV@0lce(~FD+263P9m@oL%-2%%HWuOux$;I0G{qAOn
zPp)SIS%ZVXZ<Wo)aHfV(s%$HXYnKM}A$zrH9>+oC<-Z6A7SBf%K3`MYt4)t?H@rVg
zS~!Ar(o*XhKs!kda<?|U9XJNMU^y1(wM}jQUw8Sb15GN2k&->u%!WrvgHrw$BY!8>
zCS^dPDl(YhUJ`;Dh}WWTMS<@>Vo!;zb`jhaP4lKwx@?58O(P)UztH2CE|fQPstRb9
zfDF*fwOQXH3OCIccLT+!uMdseRUma`d}ZWlL68fze`Z`8!#n1C<zB6tR$Od|!VXc$
z%`MrTZaT!?X&}$;G^UZS+K!taa?$R@0Fea(bXjDJn^)BOvpZ)~2R=6#_q{??a8_?P
z?n=7>52<|i^m-P`*Y}@7pvjzbK`q1K<;T_yAjs4V`^^MutEbhBb1pDg<Xenz<n-n-
ze2Rn`w7sjSJsqaT5U4n0nP*|RQk&<#CA@t77KSgR)sf3J7|ojTqarrk9lw2T-i*l}
zQ601{j8SP4kr0Qo=8ASrZs$JlD-$bQnJPv{eip&xRU=i=BABW*1<|;!sw&iVuGxjS
zvDj|J6t&iboIByNp^7#u?u0}lfc&%lR`ogB=qtZyWRDW7k&;#@&tF#lh6s=OcJw@*
z)OK4jp-MG0k&hNTB7y6{v)9JO9v;XoZ;FQ7mpJZZCLFo4Z6LXmlW4!;=_Jrbn8Xl7
z4XwAG-bc@qkug0b?gha3YSGkyYRX+Rd9o7!cnSklUX`BIZyhBPT<e(PQ8bUH6P^jB
zzJse$-%3Xf%}qE$)!=0d(9ge&c=HX7+c9*|8SLmOF(+QQ)M%N^sMq0vaP45+>`_c=
zy5z+|x_{|*0+gT8cbwuQnnr?8neWq^avuAeU}ns5fi;3LG;<hl^Xvbk3Ui=SbM^`-
zHo-WL5;L8Iznz*2h!&Ds5(s@K_K7K9{G$m1UwSlqTFOVs-aXdu&gg<XhcwVswkxDO
zy79kT*pU9TlDo~{ecWtUZoFsRuz6xH!4|sN_RK1=Xh~#-AaV-7$sxRpCaMTm79NxH
z<YkwsO(KBHA+|l)3F`KmTK4s)7DEAo|Mq6lLAA2m5^VBW?!6UCv~JeTDLX}~&T>1K
z4JBT8W0?t>!N@j1RIi#@0R`$gai!V>L;PwyWTcF_W8@_LkDCjtK%PV8C=yw57n+_k
zZ&HFBFV1-se=_d}d+Ui2j!dE$Z7vMOGeP?xOf{%rAdew=D%06-=*q^YG_JT4VGYCK
z=HIMRJEzJAx`IAfosY(cIZyX_Cue0dH|L-jl17Teg;fu)78xp;o*xOcN?>=^_P@2&
zA>$+OE*YAfOzZwjIw2?P8Zqq5J8j);6giDW0;=50^299TR^6MfzPl?v0?2B9VZN#A
zH&ulWewsnspWaclmI&mt!#!th`fgb-FX{KKzxT_(mp)Z)B)o<`GZ23x+D1jhYKkG3
zkWq%PQ{tkEvU#i!P$$p_6(Q6TCSlMtm(6LN!;TDo{?qjH-wqnn8&Wl$0yNVmV-%As
zlh(_L3j?j?blm>#@fcydht-}u(A1FER_j9PxznCQD>);FF$n><=jjCAuj|OkQa)`A
zXNhgUucJekCZC&w@!3m2gPzX76NxiQAV&^xK&e7`mTa$GKfqW?;fZGODXBfr{xiCs
zmqY{UxWg=on?>Z|E<Io&g`qr7K&^jYz{KX`pbcov4xb~o=Rsv()X0?>e%vNy?KFek
zdx?{SV)9A1q`$Ba@EXL_>C0~FG@@JA9Sw>w^)YAXGX1Vmv%+>;XlpIH4vUe@--Pkd
zo@(@eCUXMl3_jLVAPo(mDju_U=N=i-5d#VbqSzmWmvBHKATnLe+ei{n0vDZdt8sKG
z_8XulR(Er_Wlh_7656qSnH;l#E|~ghXPO$Lf{v7gej^gv%QMNwi^@T@tQV*j@iEUl
zAqtP{2HlH-w-k(cUH#Fjbx|AJbFCcqv+QM=IOpxSG41YD&k=P$iLM!^Y5I{y7ZF4G
z*p<y(y@HL^am$71fz*7+V+V+bxEXWn!%1_zOj<`zer&kb&ctcg@%7I?&WtoYWbB$)
z<(kiaxvk2u^2W&Uu0+lWp-ri1x$)+Y{(vUX?!nZOT9{oB)vHO}Z4DL7NWqFKhA<Zo
z8ITJ=<GUBTG(Ba8jh6r_4N)>_H2cz=|E-hc_2?VnGva(JxZGnSSoxg}#AXA|D64;w
zAhY>nOe(80%VqoXT9xm*P8+=n3v=ol$uVEt+}kyv$lF&kof$qmg<BTT+=gO~hxu2G
zf2~l@*U)^rjSJ682uB!r2^hx$w)FLuxL<=sHUoNVqFl=ghwDtfrZ(>+7`nKj2PE$y
z+%zl>4dk<^b=2H|Kj4I{)VFJl7HU6MQn#iz))?S-iblD{9qz?^`zf1$Rsg&6jxhMs
z&rO}SZhP_|Q7e7n9<i58!^>>K>nx5DcU}cyBrxO1d;=mHyP{^Mcg?R`(C(RLb8*Zn
zv`4i&6;5z<^=V_*OQ)@(9kh9H4z%hYtEPxQ3jtbU3phkbHU$V>Ab>O?wv?7cian`6
zbEp;Qyl)e4`ERkP2^XjeL1{M;o9)SIhabeS|25nL<gY09WD%@HTA9x75~=tA7-bJl
z3y;h~=%ASATiH?+NSgDc{v$cC&&gnXMD506+tVO0))OU4&90^Xk(wo|Tz3M=ET2eg
zfb*{$`A49n4+ube<s7Fjpi{h#Mr2=<<J^mK$)V4IqC^*=_Th74q9Y2Fd;3s7lV9ej
z2Ds3kg24Ul#2dU?XHcIHQjNrv1DeI?MwcaNPWJ~hkt;P18WV@Ff9Sb0c6b#?nvO44
z)?~MFq%sDi6QiB86+ByO0ap$vn;%XBs+pe7wdX`>p>rh?@}PM}k{b02WqvF<u(R_7
zc_2Yn^7_5R{u?J2=rTNU1}lXggp%8y5%70MnlX+JAz#pO+3n^H<~ri9EOLd=qRCD=
z<AJ|?*T=2;)yInLc<pgcZ>zR)T=r<SAV+f;Pr<|4dx+~C9AOX<_h8kmI|)b3H6-0<
zYVqtMkF}@Vuiw`W^Icl}3{YD=x4NRg&3f3D5%VvhS^AzvAh?P~&yIAC>*fs*Jb}Um
zdSS!IND!T$`LUYz(BWoY>r#nERA;e6;s<6z`MA$YBC-!E?ibzxnZP8f8x#PRbUc6J
zRZm~|?VvGlQxnnharow~!f6|oC-u@S%O_Sm--}J^Ec>zLhgt;J!4=ba{0o|wp-s!$
zio7=Qe46&trK2^56-R<yB2gUcVVm+jZT?u}3^{9B#MZ;7ts4^$m#JE7WaKrHg;rD-
z9qsrA!o8iP-Y-oK{;tK11j_D!qSLcK5QY3dL6pG$#s8?h_^;*8yF0Wbg5`|Ge82n7
zhUK)$L1qH<)XC23di6MIk&5Olg}lG3aJb>a^ZK#F*GQUVV7DnY;ubzcNnFRf5;oFf
z36>J;aM>oFmv%e4J{D2O4U@REK?ZWX_+f7#W0P3e&Da<M`;ACf92_K1$8Y_26Sg;@
zYBZ{~tG)MU9#wN@<%P3e<$efZf5vy#WxhT7o*KK9o`-+{r=*~;<H%}MCwIHMufxr;
z^`?^BMh%ro<zksx%2h^^dU-V!IQ3gnK$|NB2tC?uvu>F#{+;M{{^e0d=J3*7{Sb2W
z7oe`7nx3!L1w^DacOOB8Q=UX(EmKhhPE;e+4nIGAH-2znuuk}}Iqu<xl$Fk8j_3)-
zV8bkPkr$33toCP&u0^54k;XM;6uG{&xUbBaK#QItpa{C2&L{amB(Q^+$!s+OlBUt&
zPewg4-Y;eqhsuOA5Y<d^wURAf#DJ7$^k7Mjp-cFxDWi&`!EV*rVZ6Y(>d35`h~>In
zd|OdR5;FoV$z0bkWH!s*0I#QJQXz)3XjNzW#dtZEFpO`*SVc8Xp)%ocvy>lmD%$DG
zatz#Arm>`{*bQc86n1xd&CMVO_0Vu7)8Rz!4K$$LWP@#r87w~sirV?d`1~W~f?iO~
zX4i$5X=9Gzrjq!V7q4#f=9E_}8!%5no!2qq1UtiPa7z)cyk_`XI%;cd^+=C|YPG=i
zSA07|#+lTC*X8PVBBMH`1Y+{-ub;ba5QOVlTmB;G=<D_9y!OyoOf#Yd6Ws`LSi!Dx
zC<}BsH`HBNw=ojvDDGcr#rQEZB{5piHGj|nWQHvtFlA2L={}O%i~ehUMMbh}8Jbc1
zNh}tlXT6!?o%enFZh+Q4#W3xhTb20*8~z(jrliUu8LX5-fp1p<Sek|)ZSeI3*b`}9
z-UcW{>g0m!c$B%3Pk{XbDDpu21?U?Rw+mnZa=GXEu8Jw(S|LRS&);_8IKhwGAbhIB
z{e4BKY6Cep;(k4n281^=c=mSrxMgRGC0P{<zuKQ>;oF%a$0on#QGmnBPuW9%w14(D
z6-V|3=6fXV-`O9RK>_e*6VR=T*1U>9w%FDUC}z5G-@GgW#33Q0*4HAo$uwwn9$R%E
z9<O5I!I{nZ%$$F2CmtDzDXd`gzoAr<^VV1;Ijl5{7{!2c#qrXiIF!@Yo#2~CsCySR
z;Fr5&oWj#9sGyqj=2bTuOPf`hxpD?>g>QX**yrEu(9_k7zSazkI<XqOnIQNjb@a2(
zgw9E4dB_|4;v~A@p*_87e`>xr_RG6utGY6jn{%pN1FH2Mf2H%8QX{Hadi#JW)Y{|Z
z=Eb5-4=I;Nh0JP%sM_laeifpshnqRckKk$?w&jl6P!%apqluLiA_FMerQl+lZ2wf1
z6Yk@CXv65d!K?SZPI=fL2c;CX{kLZ(gL^wC%KP8gE3s4;*oz6bPQRNdv398Ow>)DH
zBLns+it>Kjl7|(-5=&S>9rjn1_uP_!t8z%Oqm*d5l;_n(yu1>E$zo#1)<H^Z**jZi
zU{&?g_D9I4q$}bxYR!L&S4;*xay3Bs)GgV5&P>R;x7voF4<h^=oG-GQG<v7BZl}08
zUi_;pZAr1l9x+}z{bQ!u??mX7rhn20LdPr7s5);>I9+<%p5Wc83D|LriQN<Q-ueIZ
zEm9!Hiw8;mmxZ{C71k!8%qabG-oUGUmGHb{H^sUZ9@md&mw}Dy{IQ$3&Q@x+4i+s^
z5TK-aQ!8D3Axp|1*xqle$q9^}o=W3yUw(hfB>Bl@5cdq=x8K_x6|1F*gAgUl_cn8e
z{>;wpZ4YUQ?tNC9vDZPJagU1`WFs-Be{Xj*+B69=DV9R5zv6D2#^SxbV0M$6oIhJG
z<dj8%HXeY{(-7NY^keB*6YU2cunGNyXn-!G#reC+WfmFdkLMd-B1BU8Gq#oW7B2>*
z65Ald7Z?_=M;Gc4=939OIBF|?5ewA8aO^jvW-mfaGG~TX5^@bIJDOI_mW=aX>#*+G
zG4E1oH5)3dGOE>T9B9+E5185Ejde>eFX=5ynuTJP3JSE^8U?h7g^GKtc88xK;{CVM
zUEo*^&Fe)>`KIS4xRlo`2YO<h)0s-UodgQJm^#X%QK$TEjzysf#4Q^dSN+#;bw+)j
zIvf3->X$Qw3J53LT+O^H+JJH)v0W1}&KhHWZPBso0bwlK`z2)#0na<alX<~^KuTP^
z%HZ`q#||IGng!Q<=)?Rm+H(U_PRe|c_ninoG?$c*PjT}4X893LMAWYv;mAcC4I=tR
zD_+r7e}UHZrHzh9wCF%t5aNvjQ+k5nj9T5B_dtVytd2;2kwFFuv(g$A+dWW+)d?<T
zTYb<S^zYlxs50Z9m$6~reYMg<rgDL|W$SGJ8G1j0NY`@Hf<;g?Z!)BPqBVh5Iv=3r
zTG<fXm{*v#N7#NV1~UUJ{I_;DkBR*+Z64FA!!{lgvI_11H>m-ngT{Oi1j)DtCyC0z
zFaN+p&aV}k>{I6^0a}%jx#6-|(=-WtaOU)YVheAin>+LWIjxtBbP=5g_9V-Ajq$Q{
zTB?6F56CsEwI2D3JT=F6>TPdMR>{?ho~skED>!xhOLGseQG5dC`P~3fgLaE9Lww(4
zN6b;;&Pt|BDNJpb8D&p&GfUNMn7aez7;Rq5vz%^5=X*V^w!XaY1qvIlz6_t-1a^p-
zW_1e<zA$L7+N|1@RzXEIkxC&1nXnujO*0(pZZrZks@4R|$pSfIt;enb-fnHO(cg*a
zON+HEdIBFdJAH{__{;~mHdCe4&W@zSwfs?ms3$cgb#1GdE$l(UJ+E7G{pgp)kt@2B
zMssm5jn;eHG5CQQ0B^X^MQh*i0erEpQe^%WwOOe_M^QW6M(TX)yE64kQ3WF&nQbSG
zb)WC-#B=Ya8Y{YG!>N_Vq9r<wEV-JKru6H4j_WRcvN@ZXfSs?@{9{z({`tn22JtTn
zZ(kdxg_?_&*9CvKQt!Qmm<@Uj2p-uqqi5={+Vkk;ItG7Yw~~Ypn3pQ%CN^X$aQvy@
z#k9qeHj)8$X4BeDAdBgN0^WsZzHLcVZX`W*z33l7U~oqbB+q<4n50*gHfh@{;F0~i
z9E2v(sxK2zfV-aDx3SEXKRf0OU)#4UlC1@?LnpZj9@%To4GJC7Cc)&ZtZ6$pWDag{
zPPT#-Fo-=D&Rv$ZPoABtN!kUY#B3N$823?eTML*aMD$I`gA4b=hZd0cSm}W08|#h%
zg%yK9=}$o|w#S=y=Bb%#!?*=yyHhJJ#=1j8DwVPuN7L#pzNi2Rz>OkBIYxX>33C;H
zZU(w5D*aXm-%72zox8}1nGU7=f%g^o_53=3yf692)^I`<Xma=4ZIx0iNP{D*#z)iG
z`7K~m)EO>@vlGo4`A?P>$QONmvt#Yz>MhR>0(JFX)vgS{P45<763^#1?<9?29J2|_
zoygjU*GFRs5<vFQ=j4TnD<-6_zGLS_)FC1>v{w?(RJ!O0!0%&CcM5tdDlD>$j}{C_
ze0krQ?1ze6gVeH1xHmUvje19|IZ~^`38#!^l*7w2uUTyK5^99Ez9BggxMC;hCxL$S
z8OY@L>Z2Gx0s0rlN>>bf&4H?pI-%&Z<L0klwrAnnnO=uICVT58t7197P{UVsgU!{t
zqsl*avDXn6Z$FScJW&RODXfQ-F(YoUV5Zo4v-$@i+34acU~AcbZ3hQ}V5r3yad(az
z|EgR@aj=<V-L`60pA_kUH>0C@^IDn-?O?lbGg;+dcdN7}pHgM3xb4l}F--v51P!p|
z5<E^mF>ZA>tP|&t@lUJ$EIWK77Tklb%VP920SAXO>&g_=Ci>(2@o>&EGv-5jAl&+-
zzV{HSp=HINfICBd>3SpC%S7}F!SHR|I{yt<+Y2CE%nxju+nG^^9jRl?(}v29i{Z+$
z+`XBk?^Q+6h8zNg`V4j&<h}0FksiX<GPD^Uvdvp}O+-u}S#+dKx!|S<2!~6XaC{%D
zmB5!`bVYGOv$E7|een_Y_IxO*?K*;B$EctF4wV0^_$|=jz?PUPk;&I5>j=vY@K7ch
zg<8#l+F7sUOv26rAyO3ECx7`S_zD1+pR|zM#=vJ{nF%9S1o)h3+D)_MQ1bWjdH!yQ
zm7eyEsGUW=C8fmC6#Q^B7w3gUyeuB1xiwilDpOX2?a3Xo8|fx&em*jvjK>F__WqR^
z%3Ct|?Xu*+Rx<kFw>s=LbGQHc3LZDl4EM8ANcN5CF2UZEl1jETTiP&!F5VsnZ^#vM
z_GK*nr-qY7_BeM!Vs904V@+rvKRuSPj9)9mUTIdp8`Lho{i^}r7R#p{*Bl_fJ<vwY
zh}k~k|7Mv$O>nGqhmRL~+w&F>ZZC9TvNvy-d9TqJNVHEs#d*e`AFYXGw)Mgr-KsSE
z+Sohq=9IMh{dAW=falpHIg^lL0W87kTZ3nH*0;qXyBz|Djo*&mdSRN+4#MXZx*0o*
z(e6Alz&v2=@4t<AB~0wJ;b%MQ2s>j0&r6L@7FH7R@oUR*{*j$YV%oki&9Sn^0O`>>
zgy%Y@D%XEr(RjhQ(j_Z6#6LUmiFo*t*E;zGA%&X;92cVPSj%e-x?>B~J&qbfjJoQ4
zw6(sNjp*zI`fmnai$BGLexsmkVRR_}SC;~0@&DP!)SYKur(Ww`d3^eZ{(mp}+siL$
zzWs6by!VF{=N2~{bFy8(Ddw4V*&C1g|KIEPpV)uIeABEQyMk_2)kNxjKCxk64A%_c
zF7{cUfycKzK3z1^J?`t%N#B=j5V}5R{giET@zYM*d|IVHeUEzU)E&=imTvj_18CgD
z@Y<bEY&F*$F@3JRbnWr!PmTKbygIG=-ZN2kon^dcuI)aZ>6K4H<Mk^2xwH$<{)9SH
zdHt`Vlhr{Ho!RHYy`uZ|Pp8UNUY+iJZ}K6pnBsjfC#~KW^J(6ochbf`pxAi%{@=Uz
zZjsj?yEl64tbvBtrZxK0mDIrD6?98T?DLFMweNyXfBliwe9TPsP1W<If1Z^strNWG
zBm7!*&AyoPi;uY9cwxNuM&~AF!~SsH)UK~aSEnoIT&fbez_UJoe@u;M_<Ufrcun7<
z{N~lw)nAuvKlVx1nn~~a*7&_2XSyu~u0(2nbv#F^5`OS)fi}mH+5i(v16|z<d;aTv
uzjpTbo2A7Oe;x$-M}Lae)Oqwe_1}NFt>-kaO>|3S00K`}KbLh*2~7a)f!<XB

literal 0
HcmV?d00001

diff --git a/previews/PR3525/moi/background/duality/index.html b/previews/PR3525/moi/background/duality/index.html
new file mode 100644
index 00000000000..67ddfdb209b
--- /dev/null
+++ b/previews/PR3525/moi/background/duality/index.html
@@ -0,0 +1,84 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Duality · JuMP</title><meta name="title" content="Duality · JuMP"/><meta property="og:title" content="Duality · JuMP"/><meta property="twitter:title" content="Duality · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox" checked/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Duality</a><ul class="internal"><li><a class="tocitem" href="#Duality-and-scalar-product"><span>Duality and scalar product</span></a></li><li><a class="tocitem" href="#Dual-for-problems-with-quadratic-functions"><span>Dual for problems with quadratic functions</span></a></li><li><a class="tocitem" href="#Dual-for-square-semidefinite-matrices"><span>Dual for square semidefinite matrices</span></a></li></ul></li><li><a class="tocitem" href="../infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Background</a></li><li class="is-active"><a href>Duality</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Duality</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/duality.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Duality"><a class="docs-heading-anchor" href="#Duality">Duality</a><a id="Duality-1"></a><a class="docs-heading-anchor-permalink" href="#Duality" title="Permalink"></a></h1><p>Conic duality is the starting point for MOI&#39;s duality conventions. When all functions are affine (or coordinate projections), and all constraint sets are closed convex cones, the model may be called a conic optimization problem.</p><p>For a minimization problem in geometric conic form, the primal is:</p><p class="math-container">\[\begin{align}
+&amp; \min_{x \in \mathbb{R}^n} &amp; a_0^T x + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; A_i x + b_i &amp; \in \mathcal{C}_i &amp; i = 1 \ldots m
+\end{align}\]</p><p>and the dual is a maximization problem in standard conic form:</p><p class="math-container">\[\begin{align}
+&amp; \max_{y_1, \ldots, y_m} &amp; -\sum_{i=1}^m b_i^T y_i + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; a_0 - \sum_{i=1}^m A_i^T y_i &amp; = 0
+\\
+&amp; &amp; y_i &amp; \in \mathcal{C}_i^* &amp; i = 1 \ldots m
+\end{align}\]</p><p>where each <span>$\mathcal{C}_i$</span> is a closed convex cone and <span>$\mathcal{C}_i^*$</span> is its dual cone.</p><p>For a maximization problem in geometric conic form, the primal is:</p><p class="math-container">\[\begin{align}
+&amp; \max_{x \in \mathbb{R}^n} &amp; a_0^T x + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; A_i x + b_i &amp; \in \mathcal{C}_i &amp; i = 1 \ldots m
+\end{align}\]</p><p>and the dual is a minimization problem in standard conic form:</p><p class="math-container">\[\begin{align}
+&amp; \min_{y_1, \ldots, y_m} &amp; \sum_{i=1}^m b_i^T y_i + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; a_0 + \sum_{i=1}^m A_i^T y_i &amp; = 0
+\\
+&amp; &amp; y_i &amp; \in \mathcal{C}_i^* &amp; i = 1 \ldots m
+\end{align}\]</p><p>A linear inequality constraint <span>$a^T x + b \ge c$</span> is equivalent to <span>$a^T x + b - c \in \mathbb{R}_+$</span>, and <span>$a^T x + b \le c$</span> is equivalent to <span>$a^T x + b - c \in \mathbb{R}_-$</span>. Variable-wise constraints are affine constraints with the appropriate identity mapping in place of <span>$A_i$</span>.</p><p>For the special case of minimization LPs, the MOI primal form can be stated as:</p><p class="math-container">\[\begin{align}
+&amp; \min_{x \in \mathbb{R}^n} &amp; a_0^T x &amp;+ b_0
+\\
+&amp; \;\;\text{s.t.}
+&amp;A_1 x &amp; \ge b_1\\
+&amp;&amp; A_2 x &amp; \le b_2\\
+&amp;&amp; A_3 x &amp; = b_3
+\end{align}\]</p><p>By applying the stated transformations to conic form, taking the dual, and transforming back into linear inequality form, one obtains the following dual:</p><p class="math-container">\[\begin{align}
+&amp; \max_{y_1,y_2,y_3} &amp; b_1^Ty_1 + b_2^Ty_2 + b_3^Ty_3 &amp;+ b_0
+\\
+&amp; \;\;\text{s.t.}
+&amp;A_1^Ty_1 + A_2^Ty_2 + A_3^Ty_3 &amp; = a_0\\
+&amp;&amp; y_1 &amp;\ge 0\\
+&amp;&amp; y_2 &amp;\le 0
+\end{align}\]</p><p>For maximization LPs, the MOI primal form can be stated as:</p><p class="math-container">\[\begin{align}
+&amp; \max_{x \in \mathbb{R}^n} &amp; a_0^T x &amp;+ b_0
+\\
+&amp; \;\;\text{s.t.}
+&amp;A_1 x &amp; \ge b_1\\
+&amp;&amp; A_2 x &amp; \le b_2\\
+&amp;&amp; A_3 x &amp; = b_3
+\end{align}\]</p><p>and similarly, the dual is:</p><p class="math-container">\[\begin{align}
+&amp; \min_{y_1,y_2,y_3} &amp; -b_1^Ty_1 - b_2^Ty_2 - b_3^Ty_3 &amp;+ b_0
+\\
+&amp; \;\;\text{s.t.}
+&amp;A_1^Ty_1 + A_2^Ty_2 + A_3^Ty_3 &amp; = -a_0\\
+&amp;&amp; y_1 &amp;\ge 0\\
+&amp;&amp; y_2 &amp;\le 0
+\end{align}\]</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>For the LP case, the signs of the feasible dual variables depend only on the sense of the corresponding primal inequality and not on the objective sense.</p></div></div><h2 id="Duality-and-scalar-product"><a class="docs-heading-anchor" href="#Duality-and-scalar-product">Duality and scalar product</a><a id="Duality-and-scalar-product-1"></a><a class="docs-heading-anchor-permalink" href="#Duality-and-scalar-product" title="Permalink"></a></h2><p>The scalar product is different from the canonical one for the sets <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a>, <a href="../../reference/standard_form/#MathOptInterface.LogDetConeTriangle"><code>LogDetConeTriangle</code></a>, <a href="../../reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>RootDetConeTriangle</code></a>.</p><p>If the set <span>$C_i$</span> of the section <a href="#Duality">Duality</a> is one of these three cones, then the rows of the matrix <span>$A_i$</span> corresponding to off-diagonal entries are twice the value of the <code>coefficients</code> field in the <a href="../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a> for the corresponding rows. See <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a> for details.</p><h2 id="Dual-for-problems-with-quadratic-functions"><a class="docs-heading-anchor" href="#Dual-for-problems-with-quadratic-functions">Dual for problems with quadratic functions</a><a id="Dual-for-problems-with-quadratic-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Dual-for-problems-with-quadratic-functions" title="Permalink"></a></h2><h3 id="Quadratic-Programs-(QPs)"><a class="docs-heading-anchor" href="#Quadratic-Programs-(QPs)">Quadratic Programs (QPs)</a><a id="Quadratic-Programs-(QPs)-1"></a><a class="docs-heading-anchor-permalink" href="#Quadratic-Programs-(QPs)" title="Permalink"></a></h3><p>For quadratic programs with only affine conic constraints,</p><p class="math-container">\[\begin{align*}
+&amp; \min_{x \in \mathbb{R}^n} &amp; \frac{1}{2}x^TQ_0x + a_0^T x + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; A_i x + b_i &amp; \in \mathcal{C}_i &amp; i = 1 \ldots m.
+\end{align*}\]</p><p>with cones <span>$\mathcal{C}_i \subseteq \mathbb{R}^{m_i}$</span> for <span>$i = 1, \ldots, m$</span>, consider the Lagrangian function</p><p class="math-container">\[L(x, y) = \frac{1}{2}x^TQ_0x + a_0^T x + b_0 - \sum_{i = 1}^m y_i^T (A_i x + b_i).\]</p><p>Let <span>$z(y)$</span> denote <span>$\sum_{i = 1}^m A_i^T y_i - a_0$</span>, the Lagrangian can be rewritten as</p><p class="math-container">\[L(x, y) = \frac{1}{2}{x}^TQ_0x - z(y)^T x + b_0 - \sum_{i = 1}^m y_i^T b_i.\]</p><p>The condition <span>$\nabla_x L(x, y) = 0$</span> gives</p><p class="math-container">\[0 = \nabla_x L(x, y) = Q_0x + a_0 - \sum_{i = 1}^m y_i^T b_i\]</p><p>which gives <span>$Q_0x = z(y)$</span>. This allows to obtain that</p><p class="math-container">\[\min_{x \in \mathbb{R}^n} L(x, y) = -\frac{1}{2}x^TQ_0x + b_0 - \sum_{i = 1}^m y_i^T b_i\]</p><p>so the dual problem is</p><p class="math-container">\[\max_{y_i \in \mathcal{C}_i^*} \min_{x \in \mathbb{R}^n} -\frac{1}{2}x^TQ_0x + b_0 - \sum_{i = 1}^m y_i^T b_i.\]</p><p>If <span>$Q_0$</span> is invertible, we have <span>$x = Q_0^{-1}z(y)$</span> hence</p><p class="math-container">\[\min_{x \in \mathbb{R}^n} L(x, y) = -\frac{1}{2}z(y)^TQ_0^{-1}z(y) + b_0 - \sum_{i = 1}^m y_i^T b_i\]</p><p>so the dual problem is</p><p class="math-container">\[\max_{y_i \in \mathcal{C}_i^*} -\frac{1}{2}z(y)^TQ_0^{-1}z(y) + b_0 - \sum_{i = 1}^m y_i^T b_i.\]</p><h3 id="Quadratically-Constrained-Quadratic-Programs-(QCQPs)"><a class="docs-heading-anchor" href="#Quadratically-Constrained-Quadratic-Programs-(QCQPs)">Quadratically Constrained Quadratic Programs (QCQPs)</a><a id="Quadratically-Constrained-Quadratic-Programs-(QCQPs)-1"></a><a class="docs-heading-anchor-permalink" href="#Quadratically-Constrained-Quadratic-Programs-(QCQPs)" title="Permalink"></a></h3><p>Given a problem with both quadratic function and quadratic objectives:</p><p class="math-container">\[\begin{align*}
+&amp; \min_{x \in \mathbb{R}^n} &amp; \frac{1}{2}x^TQ_0x + a_0^T x + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; \frac{1}{2}x^TQ_ix + a_i^T x + b_i &amp; \in \mathcal{C}_i &amp; i = 1 \ldots m.
+\end{align*}\]</p><p>with cones <span>$\mathcal{C}_i \subseteq \mathbb{R}$</span> for <span>$i = 1 \ldots m$</span>, consider the Lagrangian function</p><p class="math-container">\[L(x, y) = \frac{1}{2}x^TQ_0x + a_0^T x + b_0 - \sum_{i = 1}^m y_i (\frac{1}{2}x^TQ_ix + a_i^T x + b_i)\]</p><p>A pair of primal-dual variables <span>$(x^\star, y^\star)$</span> is optimal if</p><ul><li><span>$x^\star$</span> is a minimizer of<p class="math-container">\[\min_{x \in \mathbb{R}^n} L(x, y^\star).\]</p>That is,<p class="math-container">\[0 = \nabla_x L(x, y^\star) = Q_0x + a_0 - \sum_{i = 1}^m y_i^\star (Q_ix + a_i).\]</p></li><li>and <span>$y^\star$</span> is a maximizer of<p class="math-container">\[\max_{y_i \in \mathcal{C}_i^*} L(x^\star, y).\]</p>That is, for all <span>$i = 1, \ldots, m$</span>, <span>$\frac{1}{2}x^TQ_ix + a_i^T x + b_i$</span> is either zero or in the <a href="https://en.wikipedia.org/wiki/Normal_cone">normal cone</a> of <span>$\mathcal{C}_i^*$</span> at <span>$y^\star$</span>. For instance, if <span>$\mathcal{C}_i$</span> is <span>$\{ z \in \mathbb{R} : z \le 0 \}$</span>, this means that if <span>$\frac{1}{2}x^TQ_ix + a_i^T x + b_i$</span> is nonzero at <span>$x^\star$</span> then <span>$y_i^\star = 0$</span>. This is the classical complementary slackness condition.</li></ul><p>If <span>$\mathcal{C}_i$</span> is a vector set, the discussion remains valid with <span>$y_i(\frac{1}{2}x^TQ_ix + a_i^T x + b_i)$</span> replaced with the scalar product between <span>$y_i$</span> and the vector of scalar-valued quadratic functions.</p><h2 id="Dual-for-square-semidefinite-matrices"><a class="docs-heading-anchor" href="#Dual-for-square-semidefinite-matrices">Dual for square semidefinite matrices</a><a id="Dual-for-square-semidefinite-matrices-1"></a><a class="docs-heading-anchor-permalink" href="#Dual-for-square-semidefinite-matrices" title="Permalink"></a></h2><p>The set <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a> is a self-dual. That is, querying <a href="../../reference/constraints/#MathOptInterface.ConstraintDual"><code>ConstraintDual</code></a> of a <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a> constraint returns a vector that is itself a member of <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a>.</p><p>However, the dual of <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a> is not so straight forward. This section explains the duality convention we use, and how it is derived.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>If you have a <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a> constraint, the result matrix <span>$A$</span> from <a href="../../reference/constraints/#MathOptInterface.ConstraintDual"><code>ConstraintDual</code></a> is not positive semidefinite. However, <span>$A + A^\top$</span> is positive semidefinite.</p></div></div><p>Let <span>$\mathcal{S}_+$</span> be the cone of symmetric semidefinite matrices in the <span>$\frac{n(n+1)}{2}$</span> dimensional space of symmetric <span>$\mathbb{R}^{n \times n}$</span> matrices. That is, <span>$\mathcal{S}_+$</span> is the set <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a>. It is well known that <span>$\mathcal{S}_+$</span> is a self-dual proper cone.</p><p>Let <span>$\mathcal{P}_+$</span> be the cone of symmetric semidefinite matrices in the <span>$n^2$</span> dimensional space of <span>$\mathbb{R}^{n \times n}$</span> matrices. That is <span>$\mathcal{P}_+$</span> is the set <a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a>.</p><p>In addition, let <span>$\mathcal{D}_+$</span> be the cone of matrices <span>$A$</span> such that <span>$A+A^\top \in \mathcal{P}_+$</span>.</p><p><span>$\mathcal{P}_+$</span> is not proper because it is not solid (it is not <span>$n^2$</span> dimensional), so it is not necessarily true that <span>$\mathcal{P}_+^{**} = \mathcal{P}_+$</span>.</p><p>However, this <em>is</em> the case, because we will show that <span>$\mathcal{P}_+^{*} = \mathcal{D}_+$</span> and <span>$\mathcal{D}_+^{*} = \mathcal{P}_+$</span>.</p><p>First, let us see why <span>$\mathcal{P}_+^{*} = \mathcal{D}_+$</span>.</p><p>If <span>$B$</span> is symmetric, then</p><p class="math-container">\[\langle A,B \rangle = \langle A^\top, B^\top \rangle = \langle A^\top, B\rangle\]</p><p>so</p><p class="math-container">\[2\langle A, B \rangle = \langle A, B \rangle + \langle A^\top, B \rangle = \langle A + A^\top , B \rangle.\]</p><p>Therefore, <span>$\langle A,B\rangle \ge 0$</span> for all <span>$B \in \mathcal{P}_+$</span> if and only if <span>$\langle A+A^\top,B\rangle \ge 0$</span> for all <span>$B \in \mathcal{P}_+$</span>. Since <span>$A+A^\top$</span> is symmetric, and we know that <span>$\mathcal{S}_+$</span> is self-dual, we have shown that <span>$\mathcal{P}_+^{*}$</span> is the set of matrices <span>$A$</span> such that <span>$A+A^\top \in \mathcal{P}_+$</span>.</p><p>Second, let us see why <span>$\mathcal{D}_+^{*} = \mathcal{P}_+$</span>.</p><p>Since <span>$A \in \mathcal{D}_+$</span> implies that <span>$A^\top \in \mathcal{D}_+$</span>, <span>$B \in \mathcal{D}_+^{*}$</span> means that <span>$\langle A+A^\top,B\rangle \ge 0$</span> for all <span>$A \in \mathcal{D}_+$</span>, and hence <span>$B \in \\mathcal{P}_+$</span>.</p><p>To see why it should be symmetric, simply notice that if <span>$B_{i,j} &lt; B_{j,i}$</span>, then <span>$\langle A,B\rangle$</span> can be made arbitrarily small by setting <span>$A_{i,j} = A_{i,j} + s$</span> and <span>$A_{j,i} = A_{j,i} - s$</span>, with <span>$s$</span> arbitrarily large, and <span>$A$</span> stays in <span>$\mathcal{D}_+$</span> because <span>$A+A^\top$</span> does not change.</p><p>Typically, the primal/dual pair for semidefinite programs is presented as:</p><p class="math-container">\[\begin{align}
+       \min &amp; \langle C, X \rangle \\
+\text{s.t.} \;\; &amp; \langle A_k, X\rangle = b_k \forall k \\
+            &amp; X                     \in \mathcal{S}_+
+\end{align}\]</p><p>with the dual</p><p class="math-container">\[\begin{align}
+       \max &amp; \sum_k b_k y_k \\
+\text{s.t.} \;\; &amp; C - \sum A_k y_k \in \mathcal{S}_+
+\end{align}\]</p><p>If we allow <span>$A_k$</span> to be non-symmetric, we should instead use:</p><p class="math-container">\[\begin{align}
+       \min &amp; \langle C, X \rangle \\
+\text{s.t.} \;\; &amp; \langle A_k, X\rangle = b_k \forall k \\
+            &amp; X                     \in \mathcal{D}_+
+\end{align}\]</p><p>with the dual</p><p class="math-container">\[\begin{align}
+       \max &amp; \sum b_k y_k \\
+\text{s.t.} \;\; &amp; C - \sum A_k y_k \in \mathcal{P}_+
+\end{align}\]</p><p>This is implemented as:</p><p class="math-container">\[\begin{align}
+       \min &amp; \langle C,   Z \rangle + \langle C - C^\top, S \rangle \\
+\text{s.t.} \;\; &amp; \langle A_k, Z \rangle + \langle A_k - A_k^\top, S \rangle = b_k \forall k \\
+            &amp; Z \in \mathcal{S}_+
+\end{align}\]</p><p>with the dual</p><p class="math-container">\[\begin{align}
+       \max &amp; \sum b_k y_k \\
+\text{s.t.} \;\; &amp; C+C^\top - \sum (A_k+A_k^\top) y_k \in \mathcal{S}_+ \\
+            &amp; C-C^\top - \sum(A_k-A_k^\top) y_k = 0
+\end{align}\]</p><p>and we recover <span>$Z = X + X^\top$</span>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../manual/modification/">« Problem modification</a><a class="docs-footer-nextpage" href="../infeasibility_certificates/">Infeasibility certificates »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/background/infeasibility_certificates/index.html b/previews/PR3525/moi/background/infeasibility_certificates/index.html
new file mode 100644
index 00000000000..545bd2a14a7
--- /dev/null
+++ b/previews/PR3525/moi/background/infeasibility_certificates/index.html
@@ -0,0 +1,32 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Infeasibility certificates · JuMP</title><meta name="title" content="Infeasibility certificates · JuMP"/><meta property="og:title" content="Infeasibility certificates · JuMP"/><meta property="twitter:title" content="Infeasibility certificates · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox" checked/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../duality/">Duality</a></li><li class="is-active"><a class="tocitem" href>Infeasibility certificates</a><ul class="internal"><li><a class="tocitem" href="#Conic-duality"><span>Conic duality</span></a></li><li><a class="tocitem" href="#Unbounded-problems"><span>Unbounded problems</span></a></li><li><a class="tocitem" href="#Infeasible-problems"><span>Infeasible problems</span></a></li></ul></li><li><a class="tocitem" href="../naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Background</a></li><li class="is-active"><a href>Infeasibility certificates</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Infeasibility certificates</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/infeasibility_certificates.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Infeasibility-certificates"><a class="docs-heading-anchor" href="#Infeasibility-certificates">Infeasibility certificates</a><a id="Infeasibility-certificates-1"></a><a class="docs-heading-anchor-permalink" href="#Infeasibility-certificates" title="Permalink"></a></h1><p>When given a conic problem that is infeasible or unbounded, some solvers can produce a certificate of infeasibility. This page explains what a certificate of infeasibility is, and the related conventions that MathOptInterface adopts.</p><h2 id="Conic-duality"><a class="docs-heading-anchor" href="#Conic-duality">Conic duality</a><a id="Conic-duality-1"></a><a class="docs-heading-anchor-permalink" href="#Conic-duality" title="Permalink"></a></h2><p>MathOptInterface uses conic duality to define infeasibility certificates. A full explanation is given in the section <a href="../duality/#Duality">Duality</a>, but here is a brief overview.</p><h3 id="Minimization-problems"><a class="docs-heading-anchor" href="#Minimization-problems">Minimization problems</a><a id="Minimization-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Minimization-problems" title="Permalink"></a></h3><p>For a minimization problem in geometric conic form, the primal is:</p><p class="math-container">\[\begin{align}
+&amp; \min_{x \in \mathbb{R}^n} &amp; a_0^\top x + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; A_i x + b_i &amp; \in \mathcal{C}_i &amp; i = 1 \ldots m,
+\end{align}\]</p><p>and the dual is a maximization problem in standard conic form:</p><p class="math-container">\[\begin{align}
+&amp; \max_{y_1, \ldots, y_m} &amp; -\sum_{i=1}^m b_i^\top y_i + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; a_0 - \sum_{i=1}^m A_i^\top y_i &amp; = 0
+\\
+&amp; &amp; y_i &amp; \in \mathcal{C}_i^* &amp; i = 1 \ldots m,
+\end{align}\]</p><p>where each <span>$\mathcal{C}_i$</span> is a closed convex cone and <span>$\mathcal{C}_i^*$</span> is its dual cone.</p><h3 id="Maximization-problems"><a class="docs-heading-anchor" href="#Maximization-problems">Maximization problems</a><a id="Maximization-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Maximization-problems" title="Permalink"></a></h3><p>For a maximization problem in geometric conic form, the primal is:</p><p class="math-container">\[\begin{align}
+&amp; \max_{x \in \mathbb{R}^n} &amp; a_0^\top x + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; A_i x + b_i &amp; \in \mathcal{C}_i &amp; i = 1 \ldots m,
+\end{align}\]</p><p>and the dual is a minimization problem in standard conic form:</p><p class="math-container">\[\begin{align}
+&amp; \min_{y_1, \ldots, y_m} &amp; \sum_{i=1}^m b_i^\top y_i + b_0
+\\
+&amp; \;\;\text{s.t.} &amp; a_0 + \sum_{i=1}^m A_i^\top y_i &amp; = 0
+\\
+&amp; &amp; y_i &amp; \in \mathcal{C}_i^* &amp; i = 1 \ldots m.
+\end{align}\]</p><h2 id="Unbounded-problems"><a class="docs-heading-anchor" href="#Unbounded-problems">Unbounded problems</a><a id="Unbounded-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Unbounded-problems" title="Permalink"></a></h2><p>A problem is unbounded if and only if:</p><ol><li>there exists a feasible primal solution</li><li>the dual is infeasible.</li></ol><p>A feasible primal solution—if one exists—can be obtained by setting <a href="../../reference/models/#MathOptInterface.ObjectiveSense"><code>ObjectiveSense</code></a> to <code>FEASIBILITY_SENSE</code> before optimizing. Therefore, most solvers stop after they prove the dual is infeasible via a certificate of dual infeasibility, but <em>before</em> they have found a feasible primal solution. This is also the reason that MathOptInterface defines the <code>DUAL_INFEASIBLE</code> status instead of <code>UNBOUNDED</code>.</p><p>A certificate of dual infeasibility is an improving ray of the primal problem. That is, there exists some vector <span>$d$</span> such that for all <span>$\eta &gt; 0$</span>:</p><p class="math-container">\[A_i (x + \eta d) + b_i \in \mathcal{C}_i,\ \ i = 1 \ldots m,\]</p><p>and (for minimization problems):</p><p class="math-container">\[a_0^\top (x + \eta d) + b_0 &lt; a_0^\top x + b_0,\]</p><p>for any feasible point <span>$x$</span>. The latter simplifies to <span>$a_0^\top d &lt; 0$</span>. For maximization problems, the inequality is reversed, so that <span>$a_0^\top d &gt; 0$</span>.</p><p>If the solver has found a certificate of dual infeasibility:</p><ul><li><a href="../../reference/models/#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a> must be <code>DUAL_INFEASIBLE</code></li><li><a href="../../reference/models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> must be <code>INFEASIBILITY_CERTIFICATE</code></li><li><a href="../../reference/variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a> must be the corresponding value of <span>$d$</span></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>ConstraintPrimal</code></a> must be the corresponding value of <span>$A_i d$</span></li><li><a href="../../reference/models/#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a> must be the value <span>$a_0^\top d$</span>. Note that this is the value of the objective function at <code>d</code>, ignoring the constant <code>b_0</code>.</li></ul><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The choice of whether to scale the ray <span>$d$</span> to have magnitude <code>1</code> is left to the solver.</p></div></div><h2 id="Infeasible-problems"><a class="docs-heading-anchor" href="#Infeasible-problems">Infeasible problems</a><a id="Infeasible-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Infeasible-problems" title="Permalink"></a></h2><p>A certificate of primal infeasibility is an improving ray of the dual problem. However, because infeasibility is independent of the objective function, we first homogenize the primal problem by removing its objective.</p><p>For a minimization problem, a dual improving ray is some vector <span>$d$</span> such that for all <span>$\eta &gt; 0$</span>:</p><p class="math-container">\[\begin{align}
+-\sum_{i=1}^m A_i^\top (y_i + \eta d_i) &amp; = 0 \\
+(y_i + \eta d_i) &amp; \in \mathcal{C}_i^* &amp; i = 1 \ldots m,
+\end{align}\]</p><p>and:</p><p class="math-container">\[-\sum_{i=1}^m b_i^\top (y_i + \eta d_i) &gt; -\sum_{i=1}^m b_i^\top y_i,\]</p><p>for any feasible dual solution <span>$y$</span>. The latter simplifies to <span>$-\sum_{i=1}^m b_i^\top d_i &gt; 0$</span>. For a maximization problem, the inequality is <span>$\sum_{i=1}^m b_i^\top d_i &lt; 0$</span>. (Note that these are the same inequality, modulo a <code>-</code> sign.)</p><p>If the solver has found a certificate of primal infeasibility:</p><ul><li><a href="../../reference/models/#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a> must be <code>INFEASIBLE</code></li><li><a href="../../reference/models/#MathOptInterface.DualStatus"><code>DualStatus</code></a> must be <code>INFEASIBILITY_CERTIFICATE</code></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintDual"><code>ConstraintDual</code></a> must be the corresponding value of <span>$d$</span></li><li><a href="../../reference/models/#MathOptInterface.DualObjectiveValue"><code>DualObjectiveValue</code></a> must be the value <span>$-\sum_{i=1}^m b_i^\top d_i$</span> for minimization problems and <span>$\sum_{i=1}^m b_i^\top d_i$</span> for maximization problems.</li></ul><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The choice of whether to scale the ray <span>$d$</span> to have magnitude <code>1</code> is left to the solver.</p></div></div><h3 id="Infeasibility-certificates-of-variable-bounds"><a class="docs-heading-anchor" href="#Infeasibility-certificates-of-variable-bounds">Infeasibility certificates of variable bounds</a><a id="Infeasibility-certificates-of-variable-bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Infeasibility-certificates-of-variable-bounds" title="Permalink"></a></h3><p>Many linear solvers (for example, Gurobi) do not provide explicit access to the primal infeasibility certificate of a variable bound. However, given a set of linear constraints:</p><p class="math-container">\[\begin{align}
+l_A \le A x \le u_A \\
+l_x \le x \le u_x,
+\end{align}\]</p><p>the primal certificate of the variable bounds can be computed using the primal certificate associated with the affine constraints, <span>$d$</span>. (Note that <span>$d$</span> will have one element for each row of the <span>$A$</span> matrix, and that some or all of the elements in the vectors <span>$l_A$</span> and <span>$u_A$</span> may be <span>$\pm \infty$</span>. If both <span>$l_A$</span> and <span>$u_A$</span> are finite for some row, the corresponding element in `<code>d</code>  must be <code>0</code>.)</p><p>Given <span>$d$</span>, compute <span>$\bar{d} = d^\top A$</span>. If the bound is finite, a certificate for the lower variable bound of <span>$x_i$</span> is <span>$\max\{\bar{d}_i, 0\}$</span>, and a certificate for the upper variable bound is <span>$\min\{\bar{d}_i, 0\}$</span>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../duality/">« Duality</a><a class="docs-footer-nextpage" href="../naming_conventions/">Naming conventions »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/background/motivation/index.html b/previews/PR3525/moi/background/motivation/index.html
new file mode 100644
index 00000000000..35b732ad12c
--- /dev/null
+++ b/previews/PR3525/moi/background/motivation/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Motivation · JuMP</title><meta name="title" content="Motivation · JuMP"/><meta property="og:title" content="Motivation · JuMP"/><meta property="twitter:title" content="Motivation · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox" checked/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li class="is-active"><a class="tocitem" href>Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../duality/">Duality</a></li><li><a class="tocitem" href="../infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Introduction</a></li><li class="is-active"><a href>Motivation</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Motivation</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/motivation.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Motivation"><a class="docs-heading-anchor" href="#Motivation">Motivation</a><a id="Motivation-1"></a><a class="docs-heading-anchor-permalink" href="#Motivation" title="Permalink"></a></h1><p>MathOptInterface (MOI) is a replacement for <a href="https://github.com/JuliaOpt/MathProgBase.jl">MathProgBase</a>, the first-generation abstraction layer for mathematical optimization previously used by <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> and <a href="https://github.com/jump-dev/Convex.jl">Convex.jl</a>.</p><p>To address a number of limitations of MathProgBase, MOI is designed to:</p><ul><li>Be simple and extensible<ul><li>unifying linear, quadratic, and conic optimization,</li><li>seamlessly facilitating extensions to essentially arbitrary constraints and functions (for example, indicator constraints, complementarity constraints, and piecewise-linear functions)</li></ul></li><li>Be fast<ul><li>by allowing access to a solver&#39;s in-memory representation of a problem without writing intermediate files (when possible)</li><li>by using multiple dispatch and avoiding requiring containers of non-concrete types</li></ul></li><li>Allow a solver to return multiple results (for example, a pool of solutions)</li><li>Allow a solver to return extra arbitrary information via attributes (for example, variable- and constraint-wise membership in an irreducible inconsistent subset for infeasibility analysis)</li><li>Provide a greatly expanded set of status codes explaining what happened during the optimization procedure</li><li>Enable a solver to more precisely specify which problem classes it supports</li><li>Enable both primal and dual warm starts</li><li>Enable adding and removing both variables and constraints by indices that are not required to be consecutive</li><li>Enable any modification that the solver supports to an existing model</li><li>Avoid requiring the solver wrapper to store an additional copy of the problem data</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../">« Introduction</a><a class="docs-footer-nextpage" href="../../tutorials/example/">Solving a problem using MathOptInterface »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/background/naming_conventions/index.html b/previews/PR3525/moi/background/naming_conventions/index.html
new file mode 100644
index 00000000000..d83d1eb9bd7
--- /dev/null
+++ b/previews/PR3525/moi/background/naming_conventions/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Naming conventions · JuMP</title><meta name="title" content="Naming conventions · JuMP"/><meta property="og:title" content="Naming conventions · JuMP"/><meta property="twitter:title" content="Naming conventions · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox" checked/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../duality/">Duality</a></li><li><a class="tocitem" href="../infeasibility_certificates/">Infeasibility certificates</a></li><li class="is-active"><a class="tocitem" href>Naming conventions</a><ul class="internal"><li><a class="tocitem" href="#Sets"><span>Sets</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Background</a></li><li class="is-active"><a href>Naming conventions</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Naming conventions</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/naming_conventions.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Naming-conventions"><a class="docs-heading-anchor" href="#Naming-conventions">Naming conventions</a><a id="Naming-conventions-1"></a><a class="docs-heading-anchor-permalink" href="#Naming-conventions" title="Permalink"></a></h1><p>MOI follows several conventions for naming functions and structures. These should also be followed by packages extending MOI.</p><h2 id="Sets"><a class="docs-heading-anchor" href="#Sets">Sets</a><a id="Sets-1"></a><a class="docs-heading-anchor-permalink" href="#Sets" title="Permalink"></a></h2><p>Sets encode the structure of constraints. Their names should follow the following conventions:</p><ul><li>Abstract types in the set hierarchy should begin with <code>Abstract</code> and end in <code>Set</code>, for example, <a href="../../reference/standard_form/#MathOptInterface.AbstractScalarSet"><code>AbstractScalarSet</code></a>, <a href="../../reference/standard_form/#MathOptInterface.AbstractVectorSet"><code>AbstractVectorSet</code></a>.</li><li>Vector-valued conic sets should end with <code>Cone</code>, for example, <a href="../../reference/standard_form/#MathOptInterface.NormInfinityCone"><code>NormInfinityCone</code></a>, <a href="../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>SecondOrderCone</code></a>.</li><li>Vector-valued Cartesian products should be plural and not end in <code>Cone</code>, for example, <a href="../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>Nonnegatives</code></a>, not <code>NonnegativeCone</code>.</li><li>Matrix-valued conic sets should provide two representations: <code>ConeSquare</code> and <code>ConeTriangle</code>, for example, <a href="../../reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>RootDetConeTriangle</code></a> and <a href="../../reference/standard_form/#MathOptInterface.RootDetConeSquare"><code>RootDetConeSquare</code></a>. See <a href="../../manual/standard_form/#Matrix-cones">Matrix cones</a> for more details.</li><li>Scalar sets should be singular, not plural, for example, <a href="../../reference/standard_form/#MathOptInterface.Integer"><code>Integer</code></a>, not <code>Integers</code>.</li><li>As much as possible, the names should follow established conventions in the domain where this set is used: for instance, convex sets should have names close to those of <a href="http://web.cvxr.com/cvx/doc/">CVX</a>, and constraint-programming sets should follow <a href="https://www.minizinc.org/doc-latest/en/">MiniZinc</a>&#39;s constraints.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../infeasibility_certificates/">« Infeasibility certificates</a><a class="docs-footer-nextpage" href="../../reference/standard_form/">Standard form »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/changelog/index.html b/previews/PR3525/moi/changelog/index.html
new file mode 100644
index 00000000000..2e90bce877a
--- /dev/null
+++ b/previews/PR3525/moi/changelog/index.html
@@ -0,0 +1,34 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Release notes · JuMP</title><meta name="title" content="Release notes · JuMP"/><meta property="og:title" content="Release notes · JuMP"/><meta property="twitter:title" content="Release notes · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../">Introduction</a></li><li><a class="tocitem" href="../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../manual/models/">Models</a></li><li><a class="tocitem" href="../manual/variables/">Variables</a></li><li><a class="tocitem" href="../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../background/duality/">Duality</a></li><li><a class="tocitem" href="../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../reference/models/">Models</a></li><li><a class="tocitem" href="../reference/variables/">Variables</a></li><li><a class="tocitem" href="../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li class="is-active"><a href>Release notes</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Release notes</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/changelog.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Release-notes"><a class="docs-heading-anchor" href="#Release-notes">Release notes</a><a id="Release-notes-1"></a><a class="docs-heading-anchor-permalink" href="#Release-notes" title="Permalink"></a></h1><p>The format is based on <a href="https://keepachangelog.com/en/1.0.0/">Keep a Changelog</a>, and this project adheres to <a href="https://semver.org/spec/v2.0.0.html">Semantic Versioning</a>.</p><h2 id="v1.20.0-(September-7,-2023)"><a class="docs-heading-anchor" href="#v1.20.0-(September-7,-2023)">v1.20.0 (September 7, 2023)</a><a id="v1.20.0-(September-7,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.20.0-(September-7,-2023)" title="Permalink"></a></h2><h3 id="Added"><a class="docs-heading-anchor" href="#Added">Added</a><a id="Added-1"></a><a class="docs-heading-anchor-permalink" href="#Added" title="Permalink"></a></h3><ul><li>Added <a href="../reference/standard_form/#MathOptInterface.Scaled"><code>Scaled</code></a> set (#2237) (#2263) (#2264)</li><li>Added <a href="../reference/models/#MathOptInterface.ObjectiveLimit"><code>ObjectiveLimit</code></a> attribute (#2257)</li></ul><h3 id="Other"><a class="docs-heading-anchor" href="#Other">Other</a><a id="Other-1"></a><a class="docs-heading-anchor-permalink" href="#Other" title="Permalink"></a></h3><ul><li>Updated dependencies (#2258)</li><li>Improved performance of <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> utilities (#2259)</li><li>Fixed docstrings (#2261)</li></ul><h2 id="v1.19.0-(August-15,-2023)"><a class="docs-heading-anchor" href="#v1.19.0-(August-15,-2023)">v1.19.0 (August 15, 2023)</a><a id="v1.19.0-(August-15,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.19.0-(August-15,-2023)" title="Permalink"></a></h2><h3 id="Added-2"><a class="docs-heading-anchor" href="#Added-2">Added</a><a class="docs-heading-anchor-permalink" href="#Added-2" title="Permalink"></a></h3><ul><li>Added <a href="../reference/standard_form/#MathOptInterface.VectorNonlinearFunction"><code>VectorNonlinearFunction</code></a> (#2201)</li><li>Added new bridges<ul><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IntegerToZeroOneBridge"><code>Bridges.Constraint.IntegerToZeroOneBridge</code></a> (#2205)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToScalarQuadraticBridge"><code>Bridges.Constraint.ToScalarQuadraticBridge</code></a> (#2235)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToVectorQuadraticBridge"><code>Bridges.Constraint.ToVectorQuadraticBridge</code></a> (#2235)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToScalarNonlinearBridge"><code>Bridges.Constraint.ToScalarNonlinearBridge</code></a> (#2233) (#2235)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.FunctionConversionBridge"><code>Bridges.Constraint.FunctionConversionBridge</code></a> (#2235)</li></ul></li><li>Added <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.bridging_cost"><code>Bridges.bridging_cost</code></a> for setting a bridge-specific cost in the bridging graph (#2235)</li><li>Added <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.eval_variables"><code>Utilities.eval_variables</code></a> support for <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> (#2218) (#2219)</li><li>Added support for <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> in <code>FileFormats.NL</code> (#2228) (#2231)</li><li>Added support for writing non-<code>Float64</code> functions in <code>FileFormats.MOF</code></li><li>Added <code>Utilities.lazy_map</code> (#2254)</li></ul><h3 id="Fixed"><a class="docs-heading-anchor" href="#Fixed">Fixed</a><a id="Fixed-1"></a><a class="docs-heading-anchor-permalink" href="#Fixed" title="Permalink"></a></h3><ul><li>Fixed method ambiguities in <code>operate</code> (#2224)</li><li>Fixed reading LP file with a <code>-infinity &lt;= x &lt;= +infinity</code> variable (#2225)</li><li>Fixed missing <code>require</code> in <code>Test.test_nonlinear_duals</code> (#2230)</li><li>Fixed bug in <a href="../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>ConstraintPrimal</code></a> of <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.QuadtoSOCBridge"><code>Bridges.Constraint.QuadtoSOCBridge</code></a> (#2240)</li></ul><h3 id="Other-2"><a class="docs-heading-anchor" href="#Other-2">Other</a><a class="docs-heading-anchor-permalink" href="#Other-2" title="Permalink"></a></h3><ul><li>Added extensions to <code>solver-tests.yml</code> (#2229)</li><li>Refactored <code>test/Benchmarks</code> (#2234)</li><li>Fixed warnings in tests (#2241) (#2243)</li><li>Small refactoring of bridges for upcoming <code>VectorNonlinearFunction</code> (#2244) (#2245)</li><li>Fixed various typos (#2251) (#2255)</li><li>Partitioned how we run the tests on GitHub actions (#2252) (#2253)</li></ul><h2 id="v1.18.0-(June-23,-2023)"><a class="docs-heading-anchor" href="#v1.18.0-(June-23,-2023)">v1.18.0 (June 23, 2023)</a><a id="v1.18.0-(June-23,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.18.0-(June-23,-2023)" title="Permalink"></a></h2><h3 id="Added-3"><a class="docs-heading-anchor" href="#Added-3">Added</a><a class="docs-heading-anchor-permalink" href="#Added-3" title="Permalink"></a></h3><ul><li>Added <code>Bridges.Objective.SlackBridgePrimalDualStart</code> (#2194)</li><li>Added <a href="../reference/nonlinear/#MathOptInterface.constraint_gradient_structure"><code>constraint_gradient_structure</code></a> and <a href="../reference/nonlinear/#MathOptInterface.eval_constraint_gradient"><code>eval_constraint_gradient</code></a> (#2200)</li></ul><h3 id="Fixed-2"><a class="docs-heading-anchor" href="#Fixed-2">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-2" title="Permalink"></a></h3><ul><li>Fixed a missing <code>@require</code> in <code>MOI.Test</code> (#2195) (#2196)</li><li>Fixed incorrect usage of <code>Utilities.operate!</code> in bridges (#2207) (#2216)</li><li>Fixed splatting nonlinear expression with univariate operator (#2221)</li></ul><h3 id="Other-3"><a class="docs-heading-anchor" href="#Other-3">Other</a><a class="docs-heading-anchor-permalink" href="#Other-3" title="Permalink"></a></h3><ul><li>Removed unused argument names (#2199)</li><li>Reduced memory requirement for tests (#2204)</li><li>Refactored <code>Utilities.promote_operation</code> (#2206)</li><li>Improved code style in <code>Utilities/mutable_arithmetics.jl</code> (#2209)</li><li>Refactored various methods in <code>Utilities/functions.jl</code> (#2208) (#2212) (#2213) (#2214) (#2215)</li></ul><h2 id="v1.17.1-(June-6,-2023)"><a class="docs-heading-anchor" href="#v1.17.1-(June-6,-2023)">v1.17.1 (June 6, 2023)</a><a id="v1.17.1-(June-6,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.17.1-(June-6,-2023)" title="Permalink"></a></h2><h3 id="Fixed-3"><a class="docs-heading-anchor" href="#Fixed-3">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-3" title="Permalink"></a></h3><ul><li>Fixed comparison of <a href="../reference/errors/#MathOptInterface.ScalarFunctionConstantNotZero"><code>ScalarFunctionConstantNotZero</code></a> (#2190)</li></ul><h3 id="Other-4"><a class="docs-heading-anchor" href="#Other-4">Other</a><a class="docs-heading-anchor-permalink" href="#Other-4" title="Permalink"></a></h3><ul><li>Added documentation for enum instances (#2186)</li><li>Updated chatroom links in documentation (#2188)</li><li>Changed the documentation to build on Julia v1.9 (#2191)</li></ul><h2 id="v1.17.0-(June-1,-2023)"><a class="docs-heading-anchor" href="#v1.17.0-(June-1,-2023)">v1.17.0 (June 1, 2023)</a><a id="v1.17.0-(June-1,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.17.0-(June-1,-2023)" title="Permalink"></a></h2><h3 id="Added-4"><a class="docs-heading-anchor" href="#Added-4">Added</a><a class="docs-heading-anchor-permalink" href="#Added-4" title="Permalink"></a></h3><ul><li>Added <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> (#2059)</li><li>Added a variety of tests for <a href="../reference/standard_form/#MathOptInterface.NormSpectralCone"><code>NormSpectralCone</code></a>, <a href="../reference/standard_form/#MathOptInterface.NormNuclearCone"><code>NormNuclearCone</code></a>, and <a href="../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>HermitianPositiveSemidefiniteConeTriangle</code></a> (#2174)</li><li>Added <code>Utilities.value_type</code> for computing the value type of a function (#2176)</li></ul><h3 id="Fixed-4"><a class="docs-heading-anchor" href="#Fixed-4">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-4" title="Permalink"></a></h3><ul><li>Fixed support for external sets in <code>Utilities.loadfromstring!</code> (#2177)</li><li>Fixed <code>promote_operation</code> for <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> (#2179)</li><li>Fixed two issues in <code>FileFormats.LP</code> when reading files with quadratic functions (#2182) (#2184)</li></ul><h2 id="v1.16.0-(May-16,-2023)"><a class="docs-heading-anchor" href="#v1.16.0-(May-16,-2023)">v1.16.0 (May 16, 2023)</a><a id="v1.16.0-(May-16,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.16.0-(May-16,-2023)" title="Permalink"></a></h2><h3 id="Added-5"><a class="docs-heading-anchor" href="#Added-5">Added</a><a class="docs-heading-anchor-permalink" href="#Added-5" title="Permalink"></a></h3><ul><li>Added support for MathOptFormat v1.3 and v1.4 (#2158) (#2169)</li><li>Added new method to the nonlinear API (#2162) (#2164)<ul><li><a href="../reference/nonlinear/#MathOptInterface.eval_hessian_constraint"><code>eval_hessian_constraint</code></a></li><li><a href="../reference/nonlinear/#MathOptInterface.eval_hessian_objective"><code>eval_hessian_objective</code></a></li><li><a href="../reference/nonlinear/#MathOptInterface.hessian_constraint_structure"><code>hessian_constraint_structure</code></a></li><li><a href="../reference/nonlinear/#MathOptInterface.hessian_objective_structure"><code>hessian_objective_structure</code></a></li></ul></li><li>Added new sets<ul><li><a href="../reference/standard_form/#MathOptInterface.NormCone"><code>NormCone</code></a> (#2119)</li><li><code>ScaledPositiveSemidefiniteConeTriangle</code> (#2154)</li></ul></li></ul><h3 id="Fixed-5"><a class="docs-heading-anchor" href="#Fixed-5">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-5" title="Permalink"></a></h3><ul><li>Fixed support for Julia v1.9 to work around a bug in the upstream Julia compiler (#2161) (#2163)</li><li>Fixed a correctness bug in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge"><code>Bridges.Constraint.HermitianToSymmetricPSDBridge</code></a> (#2171)</li><li>Fixed <code>convert(::VariableIndex, ::ScalarAffineFunction)</code> when the function has terms with <code>0</code>coefficients (#2173)</li></ul><h3 id="Other-5"><a class="docs-heading-anchor" href="#Other-5">Other</a><a class="docs-heading-anchor-permalink" href="#Other-5" title="Permalink"></a></h3><ul><li>Fixed <code>solver-tests.yml</code> (#2157)</li><li>Updated documentation links to developer chatroom (#2160)</li><li>Added various tests for bridges (#2156)</li><li>Added checklists to the developer documentation (#2167) (#2168)</li></ul><h2 id="v1.15.1-(April-25,-2023)"><a class="docs-heading-anchor" href="#v1.15.1-(April-25,-2023)">v1.15.1 (April 25, 2023)</a><a id="v1.15.1-(April-25,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.15.1-(April-25,-2023)" title="Permalink"></a></h2><h3 id="Fixed-6"><a class="docs-heading-anchor" href="#Fixed-6">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-6" title="Permalink"></a></h3><ul><li>Fixed deleting a variable in a bridged objective (#2150)</li></ul><h2 id="v1.15.0-(April-19,-2023)"><a class="docs-heading-anchor" href="#v1.15.0-(April-19,-2023)">v1.15.0 (April 19, 2023)</a><a id="v1.15.0-(April-19,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.15.0-(April-19,-2023)" title="Permalink"></a></h2><h3 id="Added-6"><a class="docs-heading-anchor" href="#Added-6">Added</a><a class="docs-heading-anchor-permalink" href="#Added-6" title="Permalink"></a></h3><ul><li>Added <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.VectorFunctionizeBridge"><code>Bridges.Objective.VectorFunctionizeBridge</code></a> (#2139)</li></ul><h3 id="Fixed-7"><a class="docs-heading-anchor" href="#Fixed-7">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-7" title="Permalink"></a></h3><ul><li>Fixed support for <code>Rational</code> in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitIntervalBridge"><code>Bridges.Constraint.SplitIntervalBridge</code></a> (#2137)</li><li>Fixed printing of LaTeX models (#2141)</li><li>Fixed <a href="../reference/modification/#MathOptInterface.modify"><code>modify</code></a> in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.VectorSlackBridge"><code>Bridges.Objective.VectorSlackBridge</code></a> (#2144)</li><li>Fixed NAME record with spaces in <code>FileFormats.MPS</code> (#2146)</li><li>Fixed deleting a variable in a bridged objective (#2147)</li></ul><h3 id="Other-6"><a class="docs-heading-anchor" href="#Other-6">Other</a><a class="docs-heading-anchor-permalink" href="#Other-6" title="Permalink"></a></h3><ul><li>Add a test for variables in one-sided open <a href="../reference/standard_form/#MathOptInterface.Interval"><code>Interval</code></a> sets (#2133)</li><li>Minor style fixes in the source code (#2148)</li></ul><h2 id="v1.14.1-(April-6,-2023)"><a class="docs-heading-anchor" href="#v1.14.1-(April-6,-2023)">v1.14.1 (April 6, 2023)</a><a id="v1.14.1-(April-6,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.14.1-(April-6,-2023)" title="Permalink"></a></h2><h3 id="Fixed-8"><a class="docs-heading-anchor" href="#Fixed-8">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-8" title="Permalink"></a></h3><ul><li>Fixed a bug in <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.print_active_bridges"><code>Bridges.print_active_bridges</code></a> (#2135)</li></ul><h3 id="Other-7"><a class="docs-heading-anchor" href="#Other-7">Other</a><a class="docs-heading-anchor-permalink" href="#Other-7" title="Permalink"></a></h3><ul><li>Added a warning when an ambiguous string is passed to <code>exclude</code> in <a href="../submodules/Test/reference/#MathOptInterface.Test.runtests"><code>Test.runtests</code></a> (#2136)</li></ul><h2 id="v1.14.0-(April-4,-2023)"><a class="docs-heading-anchor" href="#v1.14.0-(April-4,-2023)">v1.14.0 (April 4, 2023)</a><a id="v1.14.0-(April-4,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.14.0-(April-4,-2023)" title="Permalink"></a></h2><h3 id="Added-7"><a class="docs-heading-anchor" href="#Added-7">Added</a><a class="docs-heading-anchor-permalink" href="#Added-7" title="Permalink"></a></h3><ul><li>Added support for starting values in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.QuadtoSOCBridge"><code>Bridges.Constraint.QuadtoSOCBridge</code></a> (#2115)</li><li>Added support for <code>Regex</code> in the <code>include</code> and <code>exclude</code> arguments to <a href="../submodules/Test/reference/#MathOptInterface.Test.runtests"><code>Test.runtests</code></a> (#2129)</li><li>Added <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.print_active_bridges"><code>Bridges.print_active_bridges</code></a> methods for individual objectives and constraints (#2128)</li></ul><h3 id="Fixed-9"><a class="docs-heading-anchor" href="#Fixed-9">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-9" title="Permalink"></a></h3><ul><li>Fixed <a href="../reference/models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> when parsing <code>.sol</code> files in <code>FileFormats.NL</code> (#2130)</li></ul><h2 id="v1.13.2-(March-21,-2023)"><a class="docs-heading-anchor" href="#v1.13.2-(March-21,-2023)">v1.13.2 (March 21, 2023)</a><a id="v1.13.2-(March-21,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.13.2-(March-21,-2023)" title="Permalink"></a></h2><h3 id="Fixed-10"><a class="docs-heading-anchor" href="#Fixed-10">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-10" title="Permalink"></a></h3><ul><li>Fixed splatting of containers in <code>MOI.Nonlinear</code> (#2120)</li><li>Fixed a bug reading LP files with default bounds (#2121)</li><li>Fixed a bug in which <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge"><code>Bridges.Constraint.HermitianToSymmetricPSDBridge</code></a> was not enabled by default (#2123)</li></ul><h3 id="Other-8"><a class="docs-heading-anchor" href="#Other-8">Other</a><a class="docs-heading-anchor-permalink" href="#Other-8" title="Permalink"></a></h3><ul><li>Fixed typos in the documentation (#2114)</li><li>Functions now print to the REPL in algebraic form. This is potentially breaking if you have tests which rely on a specific <code>String</code> form of MOI functions. (#2112) (#2126)</li></ul><h2 id="v1.13.1-(March-3,-2023)"><a class="docs-heading-anchor" href="#v1.13.1-(March-3,-2023)">v1.13.1 (March 3, 2023)</a><a id="v1.13.1-(March-3,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.13.1-(March-3,-2023)" title="Permalink"></a></h2><h3 id="Other-9"><a class="docs-heading-anchor" href="#Other-9">Other</a><a class="docs-heading-anchor-permalink" href="#Other-9" title="Permalink"></a></h3><ul><li>Added the Google style guide to the documentation linter Vale, and fixed the resulting warnings (#2110)</li><li>Improved the docstrings in <code>src/functions.jl</code> (#2108)</li></ul><h2 id="v1.13.0-(February-28,-2023)"><a class="docs-heading-anchor" href="#v1.13.0-(February-28,-2023)">v1.13.0 (February 28, 2023)</a><a id="v1.13.0-(February-28,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.13.0-(February-28,-2023)" title="Permalink"></a></h2><h3 id="Added-8"><a class="docs-heading-anchor" href="#Added-8">Added</a><a class="docs-heading-anchor-permalink" href="#Added-8" title="Permalink"></a></h3><ul><li>Added <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NumberConversionBridge"><code>Bridges.Constraint.NumberConversionBridge</code></a> (#2091)</li><li>Added <a href="../reference/standard_form/#MathOptInterface.Parameter"><code>Parameter</code></a> set (#2095) (#2105) (#2106) (#2109)</li><li>Added <code>with_cache_type</code> argument to <a href="../reference/models/#MathOptInterface.instantiate"><code>instantiate</code></a> (#2097)</li><li>Added support for <a href="../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>HermitianPositiveSemidefiniteConeTriangle</code></a> in <code>Utilities.Model</code> (#2100)</li></ul><h3 id="Fixed-11"><a class="docs-heading-anchor" href="#Fixed-11">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-11" title="Permalink"></a></h3><ul><li>Fixed bug when <code>Utilities.@product_of_sets</code> is empty (#2101)</li><li>Fixed <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.print_active_bridges"><code>Bridges.print_active_bridges</code></a> when variable bridge is an <a href="../reference/standard_form/#MathOptInterface.AbstractScalarSet"><code>AbstractScalarSet</code></a> (#2107)</li></ul><h3 id="Other-10"><a class="docs-heading-anchor" href="#Other-10">Other</a><a class="docs-heading-anchor-permalink" href="#Other-10" title="Permalink"></a></h3><ul><li>Added tests for vector-valued objective functions in <code>FileFormats.MOF</code> (#2093)</li><li>Used and documented preference for <code>import MathOptInterface as MOI</code> (#2096)</li><li>Fix and test links in the documentation with <code>linkcheck = true</code> (#2098)</li><li>Improved docstrings of sets in <code>src/sets.jl</code> (#2099)</li><li>Skip checking flakey links in documentation with <code>linkcheck_ignore</code>  (#2103)</li></ul><h2 id="v1.12.0-(February-10,-2023)"><a class="docs-heading-anchor" href="#v1.12.0-(February-10,-2023)">v1.12.0 (February 10, 2023)</a><a id="v1.12.0-(February-10,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.12.0-(February-10,-2023)" title="Permalink"></a></h2><h3 id="Added-9"><a class="docs-heading-anchor" href="#Added-9">Added</a><a class="docs-heading-anchor-permalink" href="#Added-9" title="Permalink"></a></h3><ul><li>Added support for vector-valued objective functions (#2070)</li><li>Added a <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.distance_to_set"><code>Utilities.distance_to_set</code></a> method for <a href="../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>SecondOrderCone</code></a> (#2060)</li></ul><h3 id="Fixed-12"><a class="docs-heading-anchor" href="#Fixed-12">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-12" title="Permalink"></a></h3><ul><li>Fixed a number of constraint bridges so that <code>Bridges.final_touch</code> can be called multiple times without forcing a rebuild of the reformulation (#2089)</li></ul><h3 id="Other-11"><a class="docs-heading-anchor" href="#Other-11">Other</a><a class="docs-heading-anchor-permalink" href="#Other-11" title="Permalink"></a></h3><ul><li>Added new tests that a set <a href="../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> appears in <a href="../reference/models/#MathOptInterface.ListOfModelAttributesSet"><code>ListOfModelAttributesSet</code></a> (#2085)</li><li>Improved the docstrings of a number of constraint-programming related sets (#2087)</li></ul><h2 id="v1.11.5-(January-24,-2023)"><a class="docs-heading-anchor" href="#v1.11.5-(January-24,-2023)">v1.11.5 (January 24, 2023)</a><a id="v1.11.5-(January-24,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.11.5-(January-24,-2023)" title="Permalink"></a></h2><h3 id="Fixed-13"><a class="docs-heading-anchor" href="#Fixed-13">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-13" title="Permalink"></a></h3><ul><li>Fixed a bug writing <code>.lp</code> files with an off-diagonal quadratic objective (#2082)</li></ul><h3 id="Other-12"><a class="docs-heading-anchor" href="#Other-12">Other</a><a class="docs-heading-anchor-permalink" href="#Other-12" title="Permalink"></a></h3><ul><li>Added <code>SnoopPrecompile</code> directives for reduced time-to-first-X in Julia v1.9 (#2080)</li></ul><h2 id="v1.11.4-(January-12,-2023)"><a class="docs-heading-anchor" href="#v1.11.4-(January-12,-2023)">v1.11.4 (January 12, 2023)</a><a id="v1.11.4-(January-12,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.11.4-(January-12,-2023)" title="Permalink"></a></h2><h3 id="Fixed-14"><a class="docs-heading-anchor" href="#Fixed-14">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-14" title="Permalink"></a></h3><ul><li>Fixed a bug reading <code>.lp</code> files with an <code>Integer</code> section (#2078)</li></ul><h2 id="v1.11.3-(January-12,-2023)"><a class="docs-heading-anchor" href="#v1.11.3-(January-12,-2023)">v1.11.3 (January 12, 2023)</a><a id="v1.11.3-(January-12,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.11.3-(January-12,-2023)" title="Permalink"></a></h2><h3 id="Fixed-15"><a class="docs-heading-anchor" href="#Fixed-15">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-15" title="Permalink"></a></h3><ul><li>Fixed a performance bug when deleting a vector of constraints (#2072)</li><li>Fixed a bug reading <code>.lp</code> files with terms like <code>x -1 y</code> (#2076)</li></ul><h3 id="Other-13"><a class="docs-heading-anchor" href="#Other-13">Other</a><a class="docs-heading-anchor-permalink" href="#Other-13" title="Permalink"></a></h3><ul><li>Documented the two-argument method of <a href="../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a> (#2074)</li></ul><h2 id="v1.11.2-(January-2,-2023)"><a class="docs-heading-anchor" href="#v1.11.2-(January-2,-2023)">v1.11.2 (January 2, 2023)</a><a id="v1.11.2-(January-2,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.11.2-(January-2,-2023)" title="Permalink"></a></h2><h3 id="Fixed-16"><a class="docs-heading-anchor" href="#Fixed-16">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-16" title="Permalink"></a></h3><ul><li>Fixed a bug reading <code>.mof.json</code> files with <code>ConstraintName</code> set for <code>VariableIndex</code> constraints (#2066)</li><li>Fixed a bug reading <code>.mof.json</code> files with nonlinear objectives and no constraints (#2068)</li></ul><h2 id="v1.11.1-(December-22,-2022)"><a class="docs-heading-anchor" href="#v1.11.1-(December-22,-2022)">v1.11.1 (December 22, 2022)</a><a id="v1.11.1-(December-22,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.11.1-(December-22,-2022)" title="Permalink"></a></h2><h3 id="Fixed-17"><a class="docs-heading-anchor" href="#Fixed-17">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-17" title="Permalink"></a></h3><ul><li>Fixed a bug reading <code>.mof.json</code> files with integer coefficients for affine and quadratic functions (#2063)</li></ul><h2 id="v1.11.0-(December-2,-2022)"><a class="docs-heading-anchor" href="#v1.11.0-(December-2,-2022)">v1.11.0 (December 2, 2022)</a><a id="v1.11.0-(December-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.11.0-(December-2,-2022)" title="Permalink"></a></h2><h3 id="Added-10"><a class="docs-heading-anchor" href="#Added-10">Added</a><a class="docs-heading-anchor-permalink" href="#Added-10" title="Permalink"></a></h3><ul><li>Added <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.PenaltyRelaxation"><code>Utilities.PenaltyRelaxation</code></a> and <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>Utilities.ScalarPenaltyRelaxation</code></a> (#1995)</li><li>Added <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.distance_to_set"><code>Utilities.distance_to_set</code></a> (#2048)</li><li>Added support for <a href="../reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>ConstraintPrimalStart</code></a> and <a href="../reference/constraints/#MathOptInterface.ConstraintDualStart"><code>ConstraintDualStart</code></a> in <code>FileFormats.MOF</code> (#2056)</li></ul><h3 id="Other-14"><a class="docs-heading-anchor" href="#Other-14">Other</a><a class="docs-heading-anchor-permalink" href="#Other-14" title="Permalink"></a></h3><ul><li>Tidied these release notes (#2055)</li></ul><h2 id="v1.10.0-(November-22,-2022)"><a class="docs-heading-anchor" href="#v1.10.0-(November-22,-2022)">v1.10.0 (November 22, 2022)</a><a id="v1.10.0-(November-22,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.10.0-(November-22,-2022)" title="Permalink"></a></h2><h3 id="Added-11"><a class="docs-heading-anchor" href="#Added-11">Added</a><a class="docs-heading-anchor-permalink" href="#Added-11" title="Permalink"></a></h3><ul><li>Added new methods <code>set(::OptimizerWithAttributes, ::RawOptimizerAttribute, value)</code> and <code>get(::OptimizerWithAttributes, ::RawOptimizerAttribute)</code> (#2049)</li><li>Added new methods <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.outer_keys"><code>Utilities.DoubleDicts.outer_keys</code></a> and <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys"><code>Utilities.DoubleDicts.nonempty_outer_keys</code></a> (#2052)</li></ul><h3 id="Fixed-18"><a class="docs-heading-anchor" href="#Fixed-18">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-18" title="Permalink"></a></h3><ul><li>Fixed <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.SlackBridge"><code>Bridges.Objective.SlackBridge</code></a> when the objective function is complex-valued (#2036) (#2038)</li><li>Fixed docstring of <a href="../submodules/Test/reference/#MathOptInterface.Test.runtests"><code>Test.runtests</code></a> to clarify the <code>warn_unsupported</code> argument (#2037)</li><li>Fixed reading of free variables in <code>FileFormats.LP</code> (#2044)</li><li>Fixed numerous edge cases reading files from QPLIB using <code>FileFormats.LP</code> (#2042) (#2044)</li><li>Fixed situations in which <code>x^y</code> returns a complex value in <code>Nonlinear</code> (#2050)</li></ul><h3 id="Other-15"><a class="docs-heading-anchor" href="#Other-15">Other</a><a class="docs-heading-anchor-permalink" href="#Other-15" title="Permalink"></a></h3><ul><li>Improved the error message thrown when a user-defined nonlinear function does not accept splatted input (#2032)</li><li>Removed specialized iterators for <code>keys</code> and <code>values</code> in <code>Utilities.CleverDicts</code> (#2051)</li></ul><h2 id="v1.9.0-(October-29,-2022)"><a class="docs-heading-anchor" href="#v1.9.0-(October-29,-2022)">v1.9.0 (October 29, 2022)</a><a id="v1.9.0-(October-29,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.9.0-(October-29,-2022)" title="Permalink"></a></h2><h3 id="Added-12"><a class="docs-heading-anchor" href="#Added-12">Added</a><a class="docs-heading-anchor-permalink" href="#Added-12" title="Permalink"></a></h3><ul><li>Added default fallback for getting <a href="../reference/models/#MathOptInterface.ListOfConstraintIndices"><code>ListOfConstraintIndices</code></a> and <a href="../reference/models/#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints</code></a> when the constraint type is unsupported by the model (#2021)</li><li>Added support for <code>min</code> and <code>max</code> in nonlinear expressions (#2023)</li><li>Added support for <code>Indicator{EqualTo{T}}</code> constraints in <code>FileFormats.MPS</code> (#2022)</li><li>Added default fallback for <a href="../reference/models/#MathOptInterface.write_to_file"><code>write_to_file</code></a> and <a href="../reference/models/#MathOptInterface.read_from_file"><code>read_from_file</code></a> (#2029)</li></ul><h3 id="Fixed-19"><a class="docs-heading-anchor" href="#Fixed-19">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-19" title="Permalink"></a></h3><ul><li>Fixed <code>Constraint.ZeroOneBridge</code> by adding new bounds as affine constraints instead of variable bounds (#1879)</li><li>Fixed reading free rows in <code>FileFormats.MPS</code> files (#2009)</li><li>Fixed parsing of <code>OBJSENSE</code> blocks in <code>FileFormats.MPS</code> files (#2016) (#2019)</li><li>Fixed the parsing of deeply nested nonlinear expressions by removing the use of recursion (#2020)</li><li>Fixed the requirements check in <code>Test.test_constrainnt_get_ConstraintIndex</code> (#2024)</li></ul><h2 id="v1.8.2-(September-20,-2022)"><a class="docs-heading-anchor" href="#v1.8.2-(September-20,-2022)">v1.8.2 (September 20, 2022)</a><a id="v1.8.2-(September-20,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.8.2-(September-20,-2022)" title="Permalink"></a></h2><h3 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h3><ul><li>Added <a href="https://vale.sh">vale</a> as a documentation linter (#2002)</li><li>Improved styling of code blocks in the PDF (#1999) (#2000)</li><li>Fixed a number of typos in the documentation (#2001) (#2003)</li></ul><h2 id="v1.8.1-(September-12,-2022)"><a class="docs-heading-anchor" href="#v1.8.1-(September-12,-2022)">v1.8.1 (September 12, 2022)</a><a id="v1.8.1-(September-12,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.8.1-(September-12,-2022)" title="Permalink"></a></h2><h3 id="Fixed-20"><a class="docs-heading-anchor" href="#Fixed-20">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-20" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>supports(::AbstractBridgeOptimizer</code> for constraint attributes (#1991) (#1992)</li></ul><h2 id="v1.8.0-(September-1,-2022)"><a class="docs-heading-anchor" href="#v1.8.0-(September-1,-2022)">v1.8.0 (September 1, 2022)</a><a id="v1.8.0-(September-1,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.8.0-(September-1,-2022)" title="Permalink"></a></h2><h3 id="Added-13"><a class="docs-heading-anchor" href="#Added-13">Added</a><a class="docs-heading-anchor-permalink" href="#Added-13" title="Permalink"></a></h3><ul><li>Added new sets<ul><li><a href="../reference/standard_form/#MathOptInterface.HyperRectangle"><code>HyperRectangle</code></a> (#1961)</li><li><a href="../reference/standard_form/#MathOptInterface.Reified"><code>Reified</code></a> (#1955)</li></ul></li><li>Added new bridges (#1955)<ul><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge"><code>Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ReifiedCountDistinctToMILPBridge"><code>Bridges.Constraint.ReifiedCountDistinctToMILPBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitHyperRectangleBridge"><code>Bridges.Constraint.SplitHyperRectangleBridge</code></a></li></ul></li><li>Added support for <code>atan(y, x)</code> in <code>Nonlinear</code> (#1987)</li></ul><h3 id="Fixed-21"><a class="docs-heading-anchor" href="#Fixed-21">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-21" title="Permalink"></a></h3><ul><li>Lazily construct expressions in <code>Nonlinear</code> so that expressions are updated when <code>Nonlinear.Parameter</code> values are updated (#1984)</li><li>Allow <code>NORM_LIMIT</code> as a <code>TerminationStatus</code> for unbounded problems in <code>Test</code> (#1990)</li></ul><h2 id="v1.7.0-(August-16,-2022)"><a class="docs-heading-anchor" href="#v1.7.0-(August-16,-2022)">v1.7.0 (August 16, 2022)</a><a id="v1.7.0-(August-16,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.7.0-(August-16,-2022)" title="Permalink"></a></h2><h3 id="Added-14"><a class="docs-heading-anchor" href="#Added-14">Added</a><a class="docs-heading-anchor-permalink" href="#Added-14" title="Permalink"></a></h3><ul><li>Added new sets<ul><li><a href="../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>HermitianPositiveSemidefiniteConeTriangle</code></a></li></ul></li><li>Added new optimizer-independent options<ul><li><a href="../reference/models/#MathOptInterface.RelativeGapTolerance"><code>RelativeGapTolerance</code></a></li><li><a href="../reference/models/#MathOptInterface.AbsoluteGapTolerance"><code>AbsoluteGapTolerance</code></a></li></ul></li><li>Added new bridges<ul><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge"><code>Bridges.Constraint.GeoMeanToPowerBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge"><code>Bridges.Constraint.HermitianToSymmetricPSDBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorGreaterToLessThanBridge"><code>Bridges.Constraint.IndicatorGreaterToLessThanBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorLessToGreaterThanBridge"><code>Bridges.Constraint.IndicatorLessToGreaterThanBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitComplexZerosBridge"><code>Bridges.Constraint.SplitComplexZerosBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitComplexEqualToBridge"><code>Bridges.Constraint.SplitComplexEqualToBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.QuadratizeBridge"><code>Bridges.Objective.QuadratizeBridge</code></a></li></ul></li><li>Added support for generic number types in <code>Utilities.loadfromstring!</code></li><li>Updated <code>FileFormats.MOF</code> to MathOptFormat v1.1, enabling support for constraint programming sets in the <code>MOF</code> file format</li><li>Added support in various <code>FileFormats</code> for<ul><li>indicator constraints in <code>FileFormats.MPS</code></li><li>quadratic constraints and an objective in <code>FileFormats.LP</code></li><li>quadratic constraints and an objective in <code>FileFormats.MPS</code></li></ul></li></ul><h3 id="Fixed-22"><a class="docs-heading-anchor" href="#Fixed-22">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-22" title="Permalink"></a></h3><ul><li>Fixed some missing promotion rules</li></ul><h3 id="Other-16"><a class="docs-heading-anchor" href="#Other-16">Other</a><a class="docs-heading-anchor-permalink" href="#Other-16" title="Permalink"></a></h3><ul><li>Improved the performance of Jacobian products in <code>Nonlinear</code></li><li>Removed an un-needed <code>copy</code> in <code>Utilities.modify_function</code></li><li>Various clean-ups in <code>Bridges/bridge_optimizer.jl</code></li></ul><h2 id="v1.6.1-(July-23,-2022)"><a class="docs-heading-anchor" href="#v1.6.1-(July-23,-2022)">v1.6.1 (July 23, 2022)</a><a id="v1.6.1-(July-23,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.6.1-(July-23,-2022)" title="Permalink"></a></h2><h3 id="Fixed-23"><a class="docs-heading-anchor" href="#Fixed-23">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-23" title="Permalink"></a></h3><ul><li>Added support for <code>ExponentialCone</code> in <code>MatrixOfConstraints</code></li><li>Fix <code>PSDSquare_3</code> test to reflect a previously fixed bug getting the <code>ConstraintDual</code> of a <code>PositiveSemidefiniteConeSquare</code> constraint</li></ul><h2 id="v1.6.0-(July-2,-2022)"><a class="docs-heading-anchor" href="#v1.6.0-(July-2,-2022)">v1.6.0 (July 2, 2022)</a><a id="v1.6.0-(July-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.6.0-(July-2,-2022)" title="Permalink"></a></h2><h3 id="Added-15"><a class="docs-heading-anchor" href="#Added-15">Added</a><a class="docs-heading-anchor-permalink" href="#Added-15" title="Permalink"></a></h3><ul><li>Added <code>Bridges.needs_final_touch</code> and <code>Bridges.final_touch</code></li><li>Added new bridges from constraint programming sets to mixed-integer linear programs:<ul><li><code>AllDifferentToCountDistinctBridge</code></li><li><code>CountAtLeastToCountBelongsBridge</code></li><li><code>CountBelongsToMILPBridge</code></li><li><code>CountDistinctToMILPBridge</code></li><li><code>CountGreaterThanToMILPBridge</code></li><li><code>CircuitToMILPBridge</code></li></ul></li></ul><h3 id="Fixed-24"><a class="docs-heading-anchor" href="#Fixed-24">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-24" title="Permalink"></a></h3><ul><li>Relax an instance of <code>::Vector</code> to <code>::AbstractVector</code> in <code>MOI.Nonlinear</code></li><li>Fix <code>BinPackingToMILPBridge</code> to respect variable bounds</li><li>Fix <code>SemiToBinaryBridge</code> to throw error if other bounds are set</li></ul><h2 id="v1.5.0-(June-27,-2022)"><a class="docs-heading-anchor" href="#v1.5.0-(June-27,-2022)">v1.5.0 (June 27, 2022)</a><a id="v1.5.0-(June-27,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.5.0-(June-27,-2022)" title="Permalink"></a></h2><h3 id="Added-16"><a class="docs-heading-anchor" href="#Added-16">Added</a><a class="docs-heading-anchor-permalink" href="#Added-16" title="Permalink"></a></h3><ul><li>Added <code>GetAttributeNotAllowed</code> for solvers to indicate when getting an attribute encounters an error</li><li>Added <code>Utilities.get_fallback</code> support for <code>ObjectiveValue</code> and <code>DualObjectiveValue</code></li><li>Added new bridges:<ul><li><code>RootDetConeSquare</code> to <code>RootDetConeTriangle</code></li><li><code>LogDetConeSquare</code> to <code>LogDetConeTriangle</code></li><li><code>BinPacking</code> to a mixed-integer linear program</li><li><code>Table</code> to a mixed-integer linear program</li></ul></li><li>Added <code>Bridges.print_active_bridges</code> to display the current optimal hyper-path in a <code>Bridges.LazyBridgeOptimizer</code></li></ul><h3 id="Fixed-25"><a class="docs-heading-anchor" href="#Fixed-25">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-25" title="Permalink"></a></h3><ul><li>Fixed <code>ZeroOne</code> tests with lower and upper bounds</li><li>Fixed error in <code>FileFormats.LP</code> when reading a malformed file</li><li>Fixed reading of nonlinear programs in <code>FileFormats.MOF</code></li><li>Fixed bug in <code>ConstraintDual</code> when using <code>SquareBridge</code></li></ul><h3 id="Other-17"><a class="docs-heading-anchor" href="#Other-17">Other</a><a class="docs-heading-anchor-permalink" href="#Other-17" title="Permalink"></a></h3><ul><li>Improved documentation of nonlinear API</li><li>Documented duality convention for <code>PositiveSemidefiniteConeSquare</code> sets</li><li>Fixed typo in <code>Bridges.Constraint.QuadToSOCBridge</code> docstring</li></ul><h2 id="v1.4.0-(June-9,-2022)"><a class="docs-heading-anchor" href="#v1.4.0-(June-9,-2022)">v1.4.0 (June 9, 2022)</a><a id="v1.4.0-(June-9,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.4.0-(June-9,-2022)" title="Permalink"></a></h2><h3 id="Added-17"><a class="docs-heading-anchor" href="#Added-17">Added</a><a class="docs-heading-anchor-permalink" href="#Added-17" title="Permalink"></a></h3><ul><li>Added a number of sets for constraint programming:<ul><li><code>AllDifferent</code></li><li><code>BinPacking</code></li><li><code>Circuit</code></li><li><code>CountAtLeast</code></li><li><code>CountBelongs</code></li><li><code>CountDistinct</code></li><li><code>CountGreaterThan</code></li><li><code>Cumulative</code></li><li><code>Path</code></li><li><code>Table</code></li></ul></li><li>Added support for user-defined hessians in <code>Nonlinear</code></li><li>Added <code>Bridges.runtests</code> to simplify the testing of bridge implementations</li></ul><h3 id="Fixed-26"><a class="docs-heading-anchor" href="#Fixed-26">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-26" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>FileFormats.NL</code> when writing univariate <code>*</code></li></ul><h3 id="Other-18"><a class="docs-heading-anchor" href="#Other-18">Other</a><a class="docs-heading-anchor-permalink" href="#Other-18" title="Permalink"></a></h3><ul><li>Began a large refactoring of the <code>Bridges</code> submodule, with greatly improved documentation.</li></ul><h2 id="v1.3.0-(May-27,-2022)"><a class="docs-heading-anchor" href="#v1.3.0-(May-27,-2022)">v1.3.0 (May 27, 2022)</a><a id="v1.3.0-(May-27,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.3.0-(May-27,-2022)" title="Permalink"></a></h2><h3 id="Added-18"><a class="docs-heading-anchor" href="#Added-18">Added</a><a class="docs-heading-anchor-permalink" href="#Added-18" title="Permalink"></a></h3><ul><li>Add <code>MOI.Nonlinear</code> submodule. This is a large new submodule that has been refactored from code that was in JuMP. For now, it should be considered experimental.</li><li>Add <code>FileFormats.NL.SolFileResults(::IO, ::Model)</code></li><li>Add <code>FileFormats.NL.read!(::IO, ::Model)</code></li><li>Add <code>MOI.modify</code> that accepts a vector of modifications</li></ul><h3 id="Fixed-27"><a class="docs-heading-anchor" href="#Fixed-27">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-27" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>Test</code> which attempted to include non-<code>.jl</code> files</li><li>Fixed a bug in <code>FileFormats</code> for models with open interval constraints</li></ul><h3 id="Other-19"><a class="docs-heading-anchor" href="#Other-19">Other</a><a class="docs-heading-anchor-permalink" href="#Other-19" title="Permalink"></a></h3><ul><li>Fixed a performance issue in <code>Utilities.DoubleDict</code></li><li>Various minor improvements to the documentation</li></ul><h2 id="v1.2.0-(April-25,-2022)"><a class="docs-heading-anchor" href="#v1.2.0-(April-25,-2022)">v1.2.0 (April 25, 2022)</a><a id="v1.2.0-(April-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.2.0-(April-25,-2022)" title="Permalink"></a></h2><h3 id="Added-19"><a class="docs-heading-anchor" href="#Added-19">Added</a><a class="docs-heading-anchor-permalink" href="#Added-19" title="Permalink"></a></h3><ul><li>Add support for the <code>FORMAT_REW</code>/<code>.rew</code> file format in <code>FileFormats</code>.</li></ul><h3 id="Fixed-28"><a class="docs-heading-anchor" href="#Fixed-28">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-28" title="Permalink"></a></h3><ul><li>Fix bug handling of default variable bounds in <code>FileFormats.LP</code></li><li>Fix <code>FileFormats.MPS</code> to not write <code>OBJSENSE</code> by default since this is only supported by some readers.</li></ul><h2 id="v1.1.2-(March-31,-2022)"><a class="docs-heading-anchor" href="#v1.1.2-(March-31,-2022)">v1.1.2 (March 31, 2022)</a><a id="v1.1.2-(March-31,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.1.2-(March-31,-2022)" title="Permalink"></a></h2><h3 id="Fixed-29"><a class="docs-heading-anchor" href="#Fixed-29">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-29" title="Permalink"></a></h3><ul><li>Fix a range of bugs in <code>FileFormats.LP</code></li><li>Fix reading of problem dimensions in <code>FileFormats.SDPA</code></li></ul><h2 id="v1.1.1-(March-23,-2022)"><a class="docs-heading-anchor" href="#v1.1.1-(March-23,-2022)">v1.1.1 (March 23, 2022)</a><a id="v1.1.1-(March-23,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.1.1-(March-23,-2022)" title="Permalink"></a></h2><h3 id="Fixed-30"><a class="docs-heading-anchor" href="#Fixed-30">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-30" title="Permalink"></a></h3><ul><li>Fix bug in <code>test_model_UpperBoundAlreadySet</code></li><li>Fix bug in <code>test_infeasible_</code> tests</li><li>Fix bug in <code>test_objective_ObjectiveFunction_blank</code></li><li>Relax restriction of <code>MOI.AbstractOptimizer</code> to <code>MOI.ModelLike</code> in <code>Utilities.CachingOptimizer</code> and <code>instantiate</code>.</li></ul><h3 id="New-tests"><a class="docs-heading-anchor" href="#New-tests">New tests</a><a id="New-tests-1"></a><a class="docs-heading-anchor-permalink" href="#New-tests" title="Permalink"></a></h3><ul><li>Add <code>test_conic_empty_matrix</code> that checks conic solvers support problems with no variables.</li></ul><h2 id="v1.1.0-(March-2,-2022)"><a class="docs-heading-anchor" href="#v1.1.0-(March-2,-2022)">v1.1.0 (March 2, 2022)</a><a id="v1.1.0-(March-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.1.0-(March-2,-2022)" title="Permalink"></a></h2><h3 id="Added-20"><a class="docs-heading-anchor" href="#Added-20">Added</a><a class="docs-heading-anchor-permalink" href="#Added-20" title="Permalink"></a></h3><ul><li>Added <code>MOI.Utilities.throw_unsupported(::UniversalFallback)</code> for simplifying solver wrappers which copy from a <code>UniversalFallback</code>.</li></ul><h2 id="v1.0.2-(March-1,-2022)"><a class="docs-heading-anchor" href="#v1.0.2-(March-1,-2022)">v1.0.2 (March 1, 2022)</a><a id="v1.0.2-(March-1,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.0.2-(March-1,-2022)" title="Permalink"></a></h2><h3 id="Fixed-31"><a class="docs-heading-anchor" href="#Fixed-31">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-31" title="Permalink"></a></h3><ul><li>Fixed a bug in the <code>test_model_ScalarFunctionConstantNotZero</code> test</li><li>Fixed the error type when an <code>AbstractFunctionConversionBridge</code> cannot get or set an attribute</li><li>Identified a correctness bug in <code>RSOCtoPSDBridge</code>. We now thrown an error instead of returning an incorrect result.</li></ul><h2 id="v1.0.1-(February-25,-2022)"><a class="docs-heading-anchor" href="#v1.0.1-(February-25,-2022)">v1.0.1 (February 25, 2022)</a><a id="v1.0.1-(February-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.0.1-(February-25,-2022)" title="Permalink"></a></h2><h3 id="Fixed-32"><a class="docs-heading-anchor" href="#Fixed-32">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-32" title="Permalink"></a></h3><ul><li>Fixed a bug in which OptimizerAttributes were not copied in CachingOptimizer</li><li>Fixed a bug in which <code>shift_constant</code> did not promote mixed types of coefficients</li><li>Fixed a bug in which deleting a constraint of a bridged variable threw <code>ErrorException</code> instead of <code>MOI.DeleteNotAllowed</code></li><li>Fixed a bug in which <code>add_constraint</code> in <code>MatrixOfConstraints</code> did not canonicalize the function</li><li>Fixed a bug when modifying scalar constants of a function containing a bridged variable</li><li>Fixed a bug in which <code>final_touch</code> was not always called with a <code>CachingOptimizer</code></li></ul><h2 id="v1.0.0-(February-17,-2022)"><a class="docs-heading-anchor" href="#v1.0.0-(February-17,-2022)">v1.0.0 (February 17, 2022)</a><a id="v1.0.0-(February-17,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v1.0.0-(February-17,-2022)" title="Permalink"></a></h2><p>Although tagged as a breaking release, v1.0.0 is v0.10.9 with deprecations removed, similar to how Julia 1.0 was Julia 0.7 with deprecations removed.</p><h3 id="Breaking"><a class="docs-heading-anchor" href="#Breaking">Breaking</a><a id="Breaking-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking" title="Permalink"></a></h3><ul><li>Julia 1.6 is now the minimum supported version</li><li>All deprecations have been removed</li></ul><h3 id="Troubleshooting-problems-when-updating"><a class="docs-heading-anchor" href="#Troubleshooting-problems-when-updating">Troubleshooting problems when updating</a><a id="Troubleshooting-problems-when-updating-1"></a><a class="docs-heading-anchor-permalink" href="#Troubleshooting-problems-when-updating" title="Permalink"></a></h3><p>If you experience problems when updating, you are likely using previously deprecated features. (By default, Julia does not warn when you use deprecated features.)</p><p>To find the deprecated features you are using, start Julia with <code>--depwarn=yes</code>:</p><pre><code class="nohighlight hljs">$ julia --depwarn=yes</code></pre><p>Then install MathOptInterface v0.10.9:</p><pre><code class="language-julia hljs">julia&gt; using Pkg
+julia&gt; pkg&quot;add MathOptInterface@0.10&quot;</code></pre><p>And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.</p><h2 id="v0.10.9-(February-16,-2022)"><a class="docs-heading-anchor" href="#v0.10.9-(February-16,-2022)">v0.10.9 (February 16, 2022)</a><a id="v0.10.9-(February-16,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.9-(February-16,-2022)" title="Permalink"></a></h2><h3 id="Added-21"><a class="docs-heading-anchor" href="#Added-21">Added</a><a class="docs-heading-anchor-permalink" href="#Added-21" title="Permalink"></a></h3><ul><li>Added <code>MOI.Utilities.FreeVariables</code> as a new <code>VariablesConstrainer</code> for conic solvers</li><li>Added <code>MOI.default_cache</code> for specifying the model used in <code>CachingOptimizer</code></li></ul><h3 id="Fixed-33"><a class="docs-heading-anchor" href="#Fixed-33">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-33" title="Permalink"></a></h3><ul><li>Fixed LaTeX printing of <code>MOI.Interval</code> sets</li></ul><h3 id="Other-20"><a class="docs-heading-anchor" href="#Other-20">Other</a><a class="docs-heading-anchor-permalink" href="#Other-20" title="Permalink"></a></h3><ul><li>Added Aqua.jl as a CI check, and fixed suggested issues</li><li>The constructors of GeoMeanBridge, StructOfConstraints, and CachingOptimizer were changed from outer to inner constructors. This change is technically breaking, but does not impact users who followed the documented API.</li></ul><h2 id="v0.10.8-(February-3,-2022)"><a class="docs-heading-anchor" href="#v0.10.8-(February-3,-2022)">v0.10.8 (February 3, 2022)</a><a id="v0.10.8-(February-3,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.8-(February-3,-2022)" title="Permalink"></a></h2><h3 id="Added-22"><a class="docs-heading-anchor" href="#Added-22">Added</a><a class="docs-heading-anchor-permalink" href="#Added-22" title="Permalink"></a></h3><ul><li>Added a <code>Base.read!</code> for <code>FileFormats.LP</code>.</li></ul><h3 id="Fixed-34"><a class="docs-heading-anchor" href="#Fixed-34">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-34" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>MutableSparseMatrix</code></li><li>Fixed a bug when calling <code>operate!(vcat, ...)</code> with <code>Number</code> arguments</li><li>Removed unintended export of deprecated symbols</li><li>Fixed a bug with <code>PowerCone</code> and <code>DualPowerCone</code> in <code>MatrixOfConstraints</code>.</li></ul><h2 id="v0.10.7-(January-5,-2022)"><a class="docs-heading-anchor" href="#v0.10.7-(January-5,-2022)">v0.10.7 (January 5, 2022)</a><a id="v0.10.7-(January-5,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.7-(January-5,-2022)" title="Permalink"></a></h2><h3 id="Added-23"><a class="docs-heading-anchor" href="#Added-23">Added</a><a class="docs-heading-anchor-permalink" href="#Added-23" title="Permalink"></a></h3><ul><li>Added test for modifying the constant vector in a <code>VectorAffineFunction-in-Zeros</code> constraint.</li></ul><h3 id="Fixed-35"><a class="docs-heading-anchor" href="#Fixed-35">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-35" title="Permalink"></a></h3><ul><li>Fixed the order in which sets are added to a <code>LazyBridgeOptimizer</code>. Compared to v0.10.6, this may result in bridged models being created with a different number (and order) of variables and constraints. However, it was necessary to fix cases which were previously rejected as unsupported, even though there was a valid bridge transformation.</li><li>Fixed an error message in <code>FileFormats.CBF</code></li><li>Fixed comparison in <code>test_linear_integration_Interval</code></li><li>Fixed errors for <code>ConstraintPrimal</code> in a <code>CachingOptimizer</code></li><li>Fixed printing of models with non-<code>Float64</code> coefficients.</li></ul><h3 id="Other-21"><a class="docs-heading-anchor" href="#Other-21">Other</a><a class="docs-heading-anchor-permalink" href="#Other-21" title="Permalink"></a></h3><ul><li>Various improvements to reduce time-to-first-solve latency</li><li>Improved error message when an optimizer does not support <code>compute_conflict!</code></li></ul><h2 id="v0.10.6-(November-30,-2021)"><a class="docs-heading-anchor" href="#v0.10.6-(November-30,-2021)">v0.10.6 (November 30, 2021)</a><a id="v0.10.6-(November-30,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.6-(November-30,-2021)" title="Permalink"></a></h2><h3 id="Added-24"><a class="docs-heading-anchor" href="#Added-24">Added</a><a class="docs-heading-anchor-permalink" href="#Added-24" title="Permalink"></a></h3><ul><li>Added new documentation and tests for infeasibility certificates</li><li>Added a version control system for the tests in <code>MOI.Test.runtests</code>. Pass <code>exclude_tests_after = v&quot;0.10.5&quot;</code> to run tests added in v0.10.5 and earlier.</li><li><code>MOI.Test.runtests</code> now supports generic number types. To specify the number type <code>T</code>, pass <code>MOI.Test.Config(T)</code>.</li><li>Added <code>infeasible_status</code> to <code>MOI.Test.Config</code> for solvers which return <code>LOCALLY_INFEASIBLE</code></li><li>CachingOptimizers now use a fallback for <code>ConstraintPrimal</code>. This should enable solvers using a CachingOptimizer to pass tests requiring <code>ConstraintPrimal</code>.</li></ul><h3 id="Fixed-36"><a class="docs-heading-anchor" href="#Fixed-36">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-36" title="Permalink"></a></h3><ul><li>Fixed a StackOverflow bug in <code>copy_to</code></li><li>Fixed error thrown when nonconvex quadratic constraints cannot be bridged</li><li>Fixed a bug in <code>copy_to</code> for <code>FileFormats.NL.Model</code></li><li>Fixed a bug in <code>FileFormats.NL</code> when printing large integers</li><li>Remove a common test failure for <code>LowerBoundAlreadySet</code> tests</li><li><code>Utilities.num_rows</code> is now exported</li><li>Remove parts of failing <code>test_model_copy_to_xxx</code> tests due to bridges</li></ul><h2 id="v0.10.5-(November-7,-2021)"><a class="docs-heading-anchor" href="#v0.10.5-(November-7,-2021)">v0.10.5 (November 7, 2021)</a><a id="v0.10.5-(November-7,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.5-(November-7,-2021)" title="Permalink"></a></h2><h3 id="Fixed-37"><a class="docs-heading-anchor" href="#Fixed-37">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-37" title="Permalink"></a></h3><ul><li>Fixed getter in <code>UniversalFallback</code></li><li>Fixed <code>test_solve_conflict_zeroone_ii</code></li></ul><h3 id="Other-22"><a class="docs-heading-anchor" href="#Other-22">Other</a><a class="docs-heading-anchor-permalink" href="#Other-22" title="Permalink"></a></h3><ul><li>Make <code>normalize_and_add_constraint</code> more flexible</li><li>Update paper BibTeX</li></ul><h2 id="v0.10.4-(October-26,-2021)"><a class="docs-heading-anchor" href="#v0.10.4-(October-26,-2021)">v0.10.4 (October 26, 2021)</a><a id="v0.10.4-(October-26,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.4-(October-26,-2021)" title="Permalink"></a></h2><h3 id="Added-25"><a class="docs-heading-anchor" href="#Added-25">Added</a><a class="docs-heading-anchor-permalink" href="#Added-25" title="Permalink"></a></h3><ul><li>Add <code>SolverVersion</code> attribute</li><li>Add new tests:<ul><li><code>test_solve_conflict_zeroone_ii</code></li><li><code>test_nonlinear_objective</code></li></ul></li><li><code>Utilities.VariablesContainer</code> now supports <code>ConstraintFunction</code> and <code>ConstraintSet</code></li><li>The documentation is now available as a PDF</li></ul><h3 id="Other-23"><a class="docs-heading-anchor" href="#Other-23">Other</a><a class="docs-heading-anchor-permalink" href="#Other-23" title="Permalink"></a></h3><ul><li>Update to MutableArithmetics 0.3</li><li>Various improvements to the documentation</li></ul><h2 id="v0.10.3-(September-18,-2021)"><a class="docs-heading-anchor" href="#v0.10.3-(September-18,-2021)">v0.10.3 (September 18, 2021)</a><a id="v0.10.3-(September-18,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.3-(September-18,-2021)" title="Permalink"></a></h2><h3 id="Fixed-38"><a class="docs-heading-anchor" href="#Fixed-38">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-38" title="Permalink"></a></h3><ul><li>Fixed bug which prevented callbacks from working through a CachingOptimizer</li><li>Fixed bug in <code>Test</code> submodule</li></ul><h2 id="v0.10.2-(September-16,-2021)"><a class="docs-heading-anchor" href="#v0.10.2-(September-16,-2021)">v0.10.2 (September 16, 2021)</a><a id="v0.10.2-(September-16,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.2-(September-16,-2021)" title="Permalink"></a></h2><ul><li>Updated MathOptFormat to v1.0</li><li>Updated JSONSchema to v1.0</li><li>Added <code>Utilities.set_with_dimension</code></li><li>Added two-argument <code>optimize!(::AbstractOptimizer, ::ModelLike)</code></li><li>The experimental feature <code>copy_to_and_optimize!</code> has been removed</li><li><code>Det</code> bridges now support getting <code>ConstraintFunction</code> and <code>ConstraintSet</code></li><li>Various minor bug fixes identified by improved testing</li></ul><h2 id="v0.10.1-(September-8,-2021)"><a class="docs-heading-anchor" href="#v0.10.1-(September-8,-2021)">v0.10.1 (September 8, 2021)</a><a id="v0.10.1-(September-8,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.1-(September-8,-2021)" title="Permalink"></a></h2><ul><li>Various fixes to <code>MOI.Test</code></li></ul><h2 id="v0.10.0-(September-6,-2021)"><a class="docs-heading-anchor" href="#v0.10.0-(September-6,-2021)">v0.10.0 (September 6, 2021)</a><a id="v0.10.0-(September-6,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.10.0-(September-6,-2021)" title="Permalink"></a></h2><p><strong>MOI v0.10 is a significant breaking release. There are a large number of user-visible breaking changes and code refactors, as well as a substantial number of new features.</strong></p><h3 id="Breaking-in-MOI"><a class="docs-heading-anchor" href="#Breaking-in-MOI">Breaking in MOI</a><a id="Breaking-in-MOI-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-MOI" title="Permalink"></a></h3><ul><li><code>SingleVariable</code> has been removed; use <code>VariableIndex</code> instead</li><li><code>SingleVariableConstraintNameError</code> has been renamed to <code>VariableIndexConstraintNameError</code></li><li><code>SettingSingleVariableFunctionNotAllowed</code> has been renamed to <code>SettingVariableIndexFunctionNotAllowed</code></li><li><code>VariableIndex</code> constraints should not support <code>ConstraintName</code></li><li><code>VariableIndex</code> constraints should not support <code>ConstraintBasisStatus</code>; implement <code>VariableBasisStatus</code> instead</li><li><code>ListOfConstraints</code> has been renamed to <code>ListOfConstraintTypesPresent</code></li><li><code>ListOfConstraintTypesPresent</code> should now return <code>Tuple{Type,Type}</code> instead of <code>Tuple{DataType,DataType}</code></li><li><code>SolveTime</code> has been renamed to <code>SolveTimeSec</code></li><li><code>IndicatorSet</code> has been renamed to <code>Indicator</code></li><li><code>RawParameter</code> has been renamed to <code>RawOptimizerAttribute</code> and now takes <code>String</code> instead of <code>Any</code> as the only argument</li><li>The <code>.N</code> field in result attributes has been renamed to <code>.result_index</code></li><li>The <code>.variable_index</code> field in <code>ScalarAffineTerm</code> has been renamed to <code>.variable</code></li><li>The <code>.variable_index_1</code> field in <code>ScalarQuadraticTerm</code> has been renamed to <code>.variable_1</code></li><li>The <code>.variable_index_2</code> field in <code>ScalarQuadraticTerm</code> has been renamed to <code>.variable_2</code></li><li>The order of <code>affine_terms</code> and <code>quadratic_terms</code> in <code>ScalarQuadraticFunction</code> and <code>VectorQuadraticFunction</code> have been reversed. Both functions now accept quadratic, affine, and constant terms in that order.</li><li>The <code>index_value</code> function has been removed. Use <code>.value</code> instead.</li><li><code>isapprox</code> has been removed for <code>SOS1</code> and <code>SOS2</code>.</li><li>The <code>dimension</code> argument to <code>Complements(dimension::Int)</code> should now be the length of the corresponding function, instead of half the length. An <code>ArgumentError</code> is thrown if <code>dimension</code> is not even.</li><li><code>copy_to</code> no longer takes keyword arguments:<ul><li><code>copy_names</code>: now copy names if they are supported by the destination solver</li><li><code>filter_constraints</code>: use <code>Utilities.ModelFilter</code> instead</li><li><code>warn_attributes</code>: never warn about optimizer attributes</li></ul></li></ul><h3 id="Breaking-in-Bridges"><a class="docs-heading-anchor" href="#Breaking-in-Bridges">Breaking in <code>Bridges</code></a><a id="Breaking-in-Bridges-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-Bridges" title="Permalink"></a></h3><ul><li><code>Constraint.RSOCBridge</code> has been renamed to <code>Constraint.RSOCtoSOCBridge</code></li><li><code>Constraint.SOCRBridge</code> has been renamed to <code>Constraint.SOCtoRSOCBridge</code></li><li>Bridges now return vectors that can be modified by the user. Previously, some bridges returned views instead of copies.</li><li><code>Bridges.IndexInVector</code> has been unified into a single type. Previously, there was a different type for each submodule within <code>Bridges</code></li><li>The signature of indicator bridges has been fixed. Use <code>MOI.Bridges.Constraint.IndicatortoSOS1{Float64}(model)</code>.</li></ul><h3 id="Breaking-in-FileFormats"><a class="docs-heading-anchor" href="#Breaking-in-FileFormats">Breaking in <code>FileFormats</code></a><a id="Breaking-in-FileFormats-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-FileFormats" title="Permalink"></a></h3><ul><li><code>FileFormats.MOF.Model</code> no longer accepts <code>validate</code> argument. Use the JSONSchema package to validate the MOF file. See the documentation for more information.</li></ul><h3 id="Breaking-in-Utilities"><a class="docs-heading-anchor" href="#Breaking-in-Utilities">Breaking in <code>Utilities</code></a><a id="Breaking-in-Utilities-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-Utilities" title="Permalink"></a></h3><ul><li>The datastructure of <code>Utilities.Model</code> (and models created with <code>Utilities.@model</code>) has been significantly refactored in a breaking way. This includes the way that objective functions and variable-related information is stored.</li><li><code>Utilities.supports_default_copy</code> has been renamed to <code>supports_incremental_interface</code></li><li><code>Utilities.automatic_copy_to</code> has been renamed to <code>Utilities.default_copy_to</code></li><li>The allocate-load API has been removed</li><li><code>CachingOptimizer</code>s are now initialized as <code>EMPTY_OPTIMIZER</code> instead of <code>ATTACHED_OPTIMIZER</code>. If your code relies on the optimizer being attached, call <code>MOIU.attach_optimizer(model)</code> after creation.</li><li>The field names of <code>Utilities.IndexMap</code> have been renamed to <code>var_map</code> and <code>con_map</code>. Accessing these fields directly is considered a private detail that may change. Use the public <code>getindex</code> and <code>setindex!</code> API instead.</li><li>The size argument to <code>Utilities.CleverDicts.CleverDict(::Integer)</code> has been removed.</li><li>The size argument to <code>Utilities.IndexMap(::Integer)</code> has been removed.</li><li><code>Utilities.DoubleDicts</code> have been significantly refactored. Consult the source code for details.</li><li><code>Utilities.test_models_equal</code> has been moved to <code>MOI.Test</code></li></ul><h3 id="Breaking-in-Test"><a class="docs-heading-anchor" href="#Breaking-in-Test">Breaking in <code>Test</code></a><a id="Breaking-in-Test-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-Test" title="Permalink"></a></h3><ul><li><code>MOI.Test</code> has been renamed to <code>MOI.DeprecatedTest</code></li><li>An entirely new <code>MOI.Test</code> submodule has been written. See the documentation for details. The new <code>MOI.Test</code> submodule may find many bugs in the implementations of existing solvers that were previously untested.</li></ul><h3 id="Other-changes:"><a class="docs-heading-anchor" href="#Other-changes:">Other changes:</a><a id="Other-changes:-1"></a><a class="docs-heading-anchor-permalink" href="#Other-changes:" title="Permalink"></a></h3><ul><li><code>attribute_value_type</code> has been added</li><li><code>copy_to_and_optimize!</code> has been added</li><li><code>VariableBasisStatus</code> has been added</li><li><code>print(model)</code> now prints a human-readable description of the model</li><li>Various improvements to the <code>FileFormats</code> submodule<ul><li><code>FileFormats.CBF</code> was refactored and received bugfixes</li><li>Support for MathOptFormat v0.6 was added in <code>FileFormats.MOF</code></li><li><code>FileFormats.MPS</code> has had bugfixes and support for more features such as <code>OBJSENSE</code> and objective constants.</li><li><code>FileFormats.NL</code> has been added to support nonlinear files</li></ul></li><li>Improved type inference throughout to reduce latency</li></ul><h3 id="Updating"><a class="docs-heading-anchor" href="#Updating">Updating</a><a id="Updating-1"></a><a class="docs-heading-anchor-permalink" href="#Updating" title="Permalink"></a></h3><p>A helpful script when updating is:</p><pre><code class="language-julia hljs">for (root, dirs, files) in walkdir(&quot;.&quot;)
+    for file in files
+        if !endswith(file, &quot;.jl&quot;)
+            continue
+        end
+        path = joinpath(root, file)
+        s = read(path, String)
+        for pair in [
+            &quot;.variable_index&quot; =&gt; &quot;.variable&quot;,
+            &quot;RawParameter&quot; =&gt; &quot;RawOptimizerAttribute&quot;,
+            &quot;ListOfConstraints&quot; =&gt; &quot;ListOfConstraintTypesPresent&quot;,
+            &quot;TestConfig&quot; =&gt; &quot;Config&quot;,
+            &quot;attr.N&quot; =&gt; &quot;attr.result_index&quot;,
+            &quot;SolveTime&quot; =&gt; &quot;SolveTimeSec&quot;,
+            &quot;DataType&quot; =&gt; &quot;Type&quot;,
+            &quot;Utilities.supports_default_copy_to&quot; =&gt;
+                &quot;supports_incremental_interface&quot;,
+            &quot;SingleVariableConstraintNameError&quot; =&gt;
+                &quot;VariableIndexConstraintNameError&quot;,
+            &quot;SettingSingleVariableFunctionNotAllowed&quot; =&gt;
+                &quot;SettingVariableIndexFunctionNotAllowed&quot;,
+            &quot;automatic_copy_to&quot; =&gt; &quot;default_copy_to&quot;,
+        ]
+            s = replace(s, pair)
+        end
+        write(path, s)
+    end
+end</code></pre><h2 id="v0.9.22-(May-22,-2021)"><a class="docs-heading-anchor" href="#v0.9.22-(May-22,-2021)">v0.9.22 (May 22, 2021)</a><a id="v0.9.22-(May-22,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.22-(May-22,-2021)" title="Permalink"></a></h2><p>This release contains backports from the ongoing development of the v0.10 release.</p><ul><li>Improved type inference in <code>Utilities</code>, <code>Bridges</code> and <code>FileFormats</code> submodules to reduce latency.</li><li>Improved performance of <code>Utilities.is_canonical</code>.</li><li>Fixed <code>Utilities.pass_nonvariable_constraints</code> with bridged variables.</li><li>Fixed performance regression of <code>Utilities.Model</code>.</li><li>Fixed ordering of objective setting in parser.</li></ul><h2 id="v0.9.21-(April-23,-2021)"><a class="docs-heading-anchor" href="#v0.9.21-(April-23,-2021)">v0.9.21 (April 23, 2021)</a><a id="v0.9.21-(April-23,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.21-(April-23,-2021)" title="Permalink"></a></h2><ul><li>Added <code>supports_shift_constant</code>.</li><li>Improve performance of bridging quadratic constraints.</li><li>Add precompilation statements.</li><li>Large improvements to the documentation.</li><li>Fix a variety of inference issues, benefiting precompilation and reducing initial latency.</li><li><code>RawParameter</code>s are now ignored when resetting a <code>CachingOptimizer</code>. Previously, changing the underlying optimizer after <code>RawParameter</code>s were set would throw an error.</li><li><code>Utilities.AbstractModel</code> is being refactored. This may break users interacting with private fields of a model generated using <code>@model</code>.</li></ul><h2 id="v0.9.20-(February-20,-2021)"><a class="docs-heading-anchor" href="#v0.9.20-(February-20,-2021)">v0.9.20 (February 20, 2021)</a><a id="v0.9.20-(February-20,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.20-(February-20,-2021)" title="Permalink"></a></h2><ul><li>Improved performance of <code>Utilities.ScalarFunctionIterator</code></li><li>Added support for <code>compute_conflict</code> to MOI layers</li><li>Added test with zero off-diagonal quadratic term in objective</li><li>Fixed double deletion of nested bridged <code>SingleVariable</code>/<code>VectorOfVariables</code> constraints</li><li>Fixed modification of un-set objective</li><li>Fixed function modification with duplicate terms</li><li>Made unit tests abort without failing if the problem class is not supported</li><li>Formatted code with JuliaFormatter</li><li>Clarified <code>BasisStatusCode</code>&#39;s docstring</li></ul><h2 id="v0.9.19-(December-1,-2020)"><a class="docs-heading-anchor" href="#v0.9.19-(December-1,-2020)">v0.9.19 (December 1, 2020)</a><a id="v0.9.19-(December-1,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.19-(December-1,-2020)" title="Permalink"></a></h2><ul><li>Added <code>CallbackNodeStatus</code> attribute</li><li>Added bridge from <code>GreaterThan</code> or <code>LessThan</code> to <code>Interval</code></li><li>Added tests for infeasibility certificates and double optimize</li><li>Fixed support for Julia v1.6</li><li>Re-organized MOI docs and added documentation for adding a test</li></ul><h2 id="v0.9.18-(November-3,-2020)"><a class="docs-heading-anchor" href="#v0.9.18-(November-3,-2020)">v0.9.18 (November 3, 2020)</a><a id="v0.9.18-(November-3,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.18-(November-3,-2020)" title="Permalink"></a></h2><ul><li>Various improvements for working with complex numbers</li><li>Added <code>GeoMeantoRelEntrBridge</code> to bridge a <code>GeometricMeanCone</code> constraint to a relative entropy constraint</li></ul><h2 id="v0.9.17-(September-21,-2020)"><a class="docs-heading-anchor" href="#v0.9.17-(September-21,-2020)">v0.9.17 (September 21, 2020)</a><a id="v0.9.17-(September-21,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.17-(September-21,-2020)" title="Permalink"></a></h2><ul><li>Fixed <code>CleverDict</code> with variable of negative index value</li><li>Implement <code>supports_add_constrained_variable</code> for <code>MockOptimizer</code></li></ul><h2 id="v0.9.16-(September-17,-2020)"><a class="docs-heading-anchor" href="#v0.9.16-(September-17,-2020)">v0.9.16 (September 17, 2020)</a><a id="v0.9.16-(September-17,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.16-(September-17,-2020)" title="Permalink"></a></h2><ul><li>Various fixes:<ul><li>32-bit support</li><li><code>CleverDict</code> with abstract value type</li><li>Checks in test suite</li></ul></li></ul><h2 id="v0.9.15-(September-14,-2020)"><a class="docs-heading-anchor" href="#v0.9.15-(September-14,-2020)">v0.9.15 (September 14, 2020)</a><a id="v0.9.15-(September-14,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.15-(September-14,-2020)" title="Permalink"></a></h2><ul><li>Bridges improvements:<ul><li>(R)SOCtoNonConvexQuad bridge</li><li>ZeroOne bridge</li><li>Use <code>supports_add_constrained_variable</code> in <code>LazyBridgeOptimizer</code></li><li>Exposed <code>VariableBridgeCost</code> and <code>ConstraintBridgeCost</code> attributes</li><li>Prioritize constraining variables on creation according to these costs</li><li>Refactor bridge debugging</li></ul></li><li>Large performance improvements across all submodules</li><li>Lots of documentation improvements</li><li>FileFormats improvements:<ul><li>Update MathOptFormat to v0.5</li><li>Fix supported objectives in <code>FileFormats</code></li></ul></li><li>Testing improvements:<ul><li>Add name option for <code>basic_constraint_test</code></li></ul></li><li>Bug fixes and missing methods<ul><li>Add <code>length</code> for iterators</li><li>Fix bug with duplicate terms</li><li>Fix order of <code>LinearOfConstraintIndices</code></li></ul></li></ul><h2 id="v0.9.14-(May-30,-2020)"><a class="docs-heading-anchor" href="#v0.9.14-(May-30,-2020)">v0.9.14 (May 30, 2020)</a><a id="v0.9.14-(May-30,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.14-(May-30,-2020)" title="Permalink"></a></h2><ul><li>Add a solver-independent interface for accessing the set of conflicting constraints an Irreducible Inconsistent Subsystem (#1056).</li><li>Bump JSONSchema dependency from v0.2 to v0.3 (#1090).</li><li>Documentation improvements:<ul><li>Fix typos (#1054, #1060, #1061, #1064, #1069, #1070).</li><li>Remove the outdated recommendation for a package implementing MOI for a solver <code>XXX</code> to be called <code>MathOptInterfaceXXX</code> (#1087).</li></ul></li><li>Utilities improvements:<ul><li>Fix <code>is_canonical</code> for quadratic functions (#1081, #1089).</li><li>Implement <code>add_constrained_variable[s]</code> for <code>CachingOptimizer</code> so that it is added as constrained variables to the underlying optimizer (#1084).</li><li>Add support for custom objective functions for <code>UniversalFallback</code> (#1086).</li><li>Deterministic ordering of constraints in <code>UniversalFallback</code> (#1088).</li></ul></li><li>Testing improvements:<ul><li>Add <code>NormOneCone</code>/<code>NormInfinityCone</code> tests (#1045).</li></ul></li><li>Bridges improvements:<ul><li>Add bridges from <code>Semiinteger</code> and <code>Semicontinuous</code> (#1059).</li><li>Implement getting <code>ConstraintSet</code> for <code>Variable.FlipSignBridge</code> (#1066).</li><li>Fix setting <code>ConstraintFunction</code> for <code>Constraint.ScalarizeBridge</code> (#1093).</li><li>Fix <code>NormOne</code>/<code>NormInf</code> bridges with nonzero constants (#1045).</li><li>Fix StackOverflow in <code>debug</code> (#1063).</li></ul></li><li>FileFormats improvements:<ul><li>[SDPA] Implement the extension for integer variables (#1079).</li><li>[SDPA] Ignore comments after <code>m</code> and <code>nblocks</code> and detect <code>dat-s</code> extension (#1077).</li><li>[SDPA] No scaling of off-diagonal coefficient (#1076).</li><li>[SDPA] Add missing negation of constant (#1075).</li></ul></li></ul><h2 id="v0.9.13-(March-24,-2020)"><a class="docs-heading-anchor" href="#v0.9.13-(March-24,-2020)">v0.9.13 (March 24, 2020)</a><a id="v0.9.13-(March-24,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.13-(March-24,-2020)" title="Permalink"></a></h2><ul><li>Added tests for <code>Semicontinuous</code> and <code>Semiinteger</code> variables (#1033).</li><li>Added tests for using <code>ExprGraph</code>s from NLP evaluators (#1043).</li><li>Update version compatibilities of dependencies (#1034, #1051, #1052).</li><li>Fixed typos in documentation (#1044).</li></ul><h2 id="v0.9.12-(February-28,-2020)"><a class="docs-heading-anchor" href="#v0.9.12-(February-28,-2020)">v0.9.12 (February 28, 2020)</a><a id="v0.9.12-(February-28,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.12-(February-28,-2020)" title="Permalink"></a></h2><ul><li>Fixed writing <code>NLPBlock</code> in MathOptFormat (#1037).</li><li>Fixed <code>MockOptimizer</code> for result attributes with non-one result index (#1039).</li><li>Updated test template with <code>instantiate</code> (#1032).</li></ul><h2 id="v0.9.11-(February-21,-2020)"><a class="docs-heading-anchor" href="#v0.9.11-(February-21,-2020)">v0.9.11 (February 21, 2020)</a><a id="v0.9.11-(February-21,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.11-(February-21,-2020)" title="Permalink"></a></h2><ul><li>Add an option for the model created by <code>Utilities.@model</code> to be a subtype of <code>AbstractOptimizer</code> (#1031).</li><li>Described dual cone in docstrings of <code>GeoMeanCone</code> and <code>RelativeEntropyCone</code> (#1018, #1028).</li><li>Fixed typos in documentation (#1022, #1024).</li><li>Fixed warning of unsupported attribute (#1027).</li><li>Added more rootdet/logdet conic tests (#1026).</li><li>Implemented <code>ConstraintDual</code> for <code>Constraint.GeoMeanBridge</code>, <code>Constraint.RootDetBridge</code> and <code>Constraint.LogDetBridge</code> and test duals in tests with <code>GeoMeanCone</code> and <code>RootDetConeTriangle</code> and <code>LogDetConeTriangle</code> cones (#1025, #1026).</li></ul><h2 id="v0.9.10-(January-31,-2020)"><a class="docs-heading-anchor" href="#v0.9.10-(January-31,-2020)">v0.9.10 (January 31, 2020)</a><a id="v0.9.10-(January-31,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.10-(January-31,-2020)" title="Permalink"></a></h2><ul><li>Added <code>OptimizerWithAttributes</code> grouping an optimizer constructor and a list of optimizer attributes (#1008).</li><li>Added <code>RelativeEntropyCone</code> with corresponding bridge into exponential cone constraints (#993).</li><li>Added <code>NormSpectralCone</code> and <code>NormNuclearCone</code> with corresponding bridges into positive semidefinite constraints (#976).</li><li>Added <code>supports_constrained_variable(s)</code> (#1004).</li><li>Added <code>dual_set_type</code> (#1002).</li><li>Added tests for vector specialized version of <code>delete</code> (#989, #1011).</li><li>Added PSD3 test (#1007).</li><li>Clarified dual solution of <code>Tests.pow1v</code> and <code>Tests.pow1f</code> (#1013).</li><li>Added support for <code>EqualTo</code> and <code>Zero</code> in <code>Bridges.Constraint.SplitIntervalBridge</code> (#1005).</li><li>Fixed <code>Utilities.vectorize</code> for empty vector (#1003).</li><li>Fixed free variables in LP writer (#1006).</li></ul><h2 id="v0.9.9-(December-29,-2019)"><a class="docs-heading-anchor" href="#v0.9.9-(December-29,-2019)">v0.9.9 (December 29, 2019)</a><a id="v0.9.9-(December-29,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.9-(December-29,-2019)" title="Permalink"></a></h2><ul><li>Incorporated MathOptFormat.jl as the FileFormats submodule. FileFormats provides readers and writers for a number of standard file formats and MOF, a file format specialized for MOI (#969).</li><li>Improved performance of deletion of vector of variables in <code>MOI.Utilities.Model</code> (#983).</li><li>Updated to MutableArithmetics v0.2 (#981).</li><li>Added <code>MutableArithmetics.promote_operation</code> allocation tests (#975).</li><li>Fixed inference issue on Julia v1.1 (#982).</li></ul><h2 id="v0.9.8-(December-19,-2019)"><a class="docs-heading-anchor" href="#v0.9.8-(December-19,-2019)">v0.9.8 (December 19, 2019)</a><a id="v0.9.8-(December-19,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.8-(December-19,-2019)" title="Permalink"></a></h2><ul><li>Implemented MutableArithmetics API (#924).</li><li>Fixed callbacks with <code>CachingOptimizer</code> (#959).</li><li>Fixed <code>MOI.dimension</code> for <code>MOI.Complements</code> (#948).</li><li>Added fallback for <code>add_variables</code> (#972).</li><li>Added <code>is_diagonal_vectorized_index</code> utility (#965).</li><li>Improved linear constraints display in manual (#963, #964).</li><li>Bridges improvements:<ul><li>Added <code>IndicatorSet</code> to <code>SOS1</code> bridge (#877).</li><li>Added support for starting values for <code>Variable.VectorizeBridge</code> (#944).</li><li>Fixed <code>MOI.add_constraints</code> with non-bridged variable constraint on bridged variable (#951).</li><li>Fixed corner cases and docstring of <code>GeoMeanBridge</code> (#961, #962, #966).</li><li>Fixed choice between variable or constraint bridges for constrained variables (#973).</li><li>Improve performance of bridge shortest path (#945, #946, #956).</li><li>Added docstring for <code>test_delete_bridge</code> (#954).</li><li>Added Variable bridge tests (#952).</li></ul></li></ul><h2 id="v0.9.7-(October-30,-2019)"><a class="docs-heading-anchor" href="#v0.9.7-(October-30,-2019)">v0.9.7 (October 30, 2019)</a><a id="v0.9.7-(October-30,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.7-(October-30,-2019)" title="Permalink"></a></h2><ul><li>Implemented <code>_result_index_field</code> for <code>NLPBlockDual</code> (#934).</li><li>Fixed copy of model with starting values for vector constraints (#941).</li><li>Bridges improvements:<ul><li>Improved performance of <code>add_bridge</code> and added <code>has_bridge</code> (#935).</li><li>Added <code>AbstractSetMapBridge</code> for bridges between sets <code>S1</code>, <code>S2</code> such that there is a linear map <code>A</code> such that <code>A*S1 = S2</code> (#933).</li><li>Added support for starting values for <code>FlipSignBridge</code>, <code>VectorizeBridge</code>, <code>ScalarizeBridge</code>, <code>SlackBridge</code>, <code>SplitIntervalBridge</code>, <code>RSOCBridge</code>, <code>SOCRBridge</code> <code>NormInfinityBridge</code>, <code>SOCtoPSDBridge</code> and <code>RSOCtoPSDBridge</code> (#933, #936, #937, #938, #939).</li></ul></li></ul><h2 id="v0.9.6-(October-25,-2019)"><a class="docs-heading-anchor" href="#v0.9.6-(October-25,-2019)">v0.9.6 (October 25, 2019)</a><a id="v0.9.6-(October-25,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.6-(October-25,-2019)" title="Permalink"></a></h2><ul><li>Added complementarity constraints (#913).</li><li>Allowed <code>ModelLike</code> objects as value of attributes (#928).</li><li>Testing improvements:<ul><li>Added <code>dual_objective_value</code> option to <code>MOI.Test.TestConfig</code> (#922).</li><li>Added <code>InvalidIndex</code> tests in <code>basic_constraint_tests</code> (#921).</li><li>Added tests for the constant term in indicator constraint (#929).</li></ul></li><li>Bridges improvements:<ul><li>Added support for starting values for <code>Functionize</code> bridges (#923).</li><li>Added variable indices context to variable bridges (#920).</li><li>Fixed a typo in printing o <code>debug_supports</code> (#927).</li></ul></li></ul><h2 id="v0.9.5-(October-9,-2019)"><a class="docs-heading-anchor" href="#v0.9.5-(October-9,-2019)">v0.9.5 (October 9, 2019)</a><a id="v0.9.5-(October-9,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.5-(October-9,-2019)" title="Permalink"></a></h2><ul><li>Clarified <code>PrimalStatus</code>/<code>DualStatus</code> to be <code>NO_SOLUTION</code> if <code>result_index</code> is out of bounds (#912).</li><li>Added tolerance for checks  and use <code>ResultCount</code> + 1 for the <code>result_index</code> in <code>MOI.Test.solve_result_status</code> (#910, #917).</li><li>Use <code>0.5</code> instead of <code>2.0</code> for power in PowerCone in <code>basic_constraint_test</code> (#916).</li><li>Bridges improvements:<ul><li>Added debug utilities for unsupported variable/constraint/objective (#861).</li><li>Fixed deletion of variables in bridged <code>VectorOfVariables</code> constraints (#909).</li><li>Fixed <code>result_index</code> with objective bridges (#911).</li></ul></li></ul><h2 id="v0.9.4-(October-2,-2019)"><a class="docs-heading-anchor" href="#v0.9.4-(October-2,-2019)">v0.9.4 (October 2, 2019)</a><a id="v0.9.4-(October-2,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.4-(October-2,-2019)" title="Permalink"></a></h2><ul><li>Added solver-independent MIP callbacks (#782).</li><li>Implements submit for <code>Utilities.CachingOptimizer</code> and <code>Bridges.AbstractBridgeOptimizer</code> (#906).</li><li>Added tests for result count of solution attributes (#901, #904).</li><li>Added <code>NumberOfThreads</code> attribute (#892).</li><li>Added <code>Utilities.get_bounds</code> to get the bounds on a variable (#890).</li><li>Added a note on duplicate coefficients in documentation (#581).</li><li>Added result index in <code>ConstraintBasisStatus</code> (#898).</li><li>Added extension dictionary to <code>Utilities.Model</code> (#884, #895).</li><li>Fixed deletion of constrained variables for CachingOptimizer (#905).</li><li>Implemented <code>Utilities.shift_constraint</code> for <code>Test.UnknownScalarSet</code> (#896).</li><li>Bridges improvements:<ul><li>Added Variable.RSOCtoSOCBridge (#907).</li><li>Implemented <code>MOI.get</code> for <code>ConstraintFunction</code>/<code>ConstraintSet</code> for <code>Bridges.Constraint.SquareBridge</code> (#899).</li></ul></li></ul><h2 id="v0.9.3-(September-20,-2019)"><a class="docs-heading-anchor" href="#v0.9.3-(September-20,-2019)">v0.9.3 (September 20, 2019)</a><a id="v0.9.3-(September-20,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.3-(September-20,-2019)" title="Permalink"></a></h2><ul><li>Fixed ambiguity detected in Julia v1.3 (#891, #893).</li><li>Fixed missing sets from <code>ListOfSupportedConstraints</code> (#880).</li><li>Fixed copy of <code>VectorOfVariables</code> constraints with duplicate indices (#886).</li><li>Added extension dictionary to MOIU.Model (#884).</li><li>Implemented <code>MOI.get</code> for function and set for <code>GeoMeanBridge</code> (#888).</li><li>Updated documentation for SingleVariable indices and bridges (#885).</li><li>Testing improvements:<ul><li>Added more comprehensive tests for names (#882).</li><li>Added tests for <code>SingleVariable</code> duals (#883).</li><li>Added tests for <code>DualExponentialCone</code> and <code>DualPowerCone</code> (#873).</li></ul></li><li>Improvements for arbitrary coefficient type:<ul><li>Fixed <code>==</code> for sets with mutable fields (#887).</li><li>Removed some <code>Float64</code> assumptions in bridges (#878).</li><li>Automatic selection of <code>Constraint.[Scalar|Vector]FunctionizeBridge</code> (#889).</li></ul></li></ul><h2 id="v0.9.2-(September-5,-2019)"><a class="docs-heading-anchor" href="#v0.9.2-(September-5,-2019)">v0.9.2 (September 5, 2019)</a><a id="v0.9.2-(September-5,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.2-(September-5,-2019)" title="Permalink"></a></h2><ul><li>Implemented model printing for <code>MOI.ModelLike</code> and specialized it for models defined in MOI (864).</li><li>Generalized <code>contlinear</code> tests for arbitrary coefficient type (#855).</li><li>Fixed <code>supports_constraint</code> for <code>Semiinteger</code> and <code>Semicontinuous</code> and <code>supports</code> for <code>ObjectiveFunction</code> (#859).</li><li>Fixed Allocate-Load copy for single variable constraints (#856).</li><li>Bridges improvements:<ul><li>Add objective bridges (#789).</li><li>Fixed <code>Variable.RSOCtoPSDBridge</code> for dimension 2 (#869).</li><li>Added <code>Variable.SOCtoRSOCBridge</code> (#865).</li><li>Added <code>Constraint.SOCRBridge</code> and disable <code>MOI.Bridges.Constraint.SOCtoPSDBridge</code> (#751).</li><li>Fixed <code>added_constraint_types</code> for <code>Contraint.LogDetBridge</code> and <code>Constraint.RootDetBridge</code> (#870).</li></ul></li></ul><h2 id="v0.9.1-(August-22,-2019)"><a class="docs-heading-anchor" href="#v0.9.1-(August-22,-2019)">v0.9.1 (August 22, 2019)</a><a id="v0.9.1-(August-22,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.1-(August-22,-2019)" title="Permalink"></a></h2><ul><li>Fix support for Julia v1.2 (#834).</li><li>L<em>1 and L</em>∞ norm epigraph cones and corresponding bridges to LP were added (#818).</li><li>Added tests to <code>MOI.Test.nametest</code> (#833).</li><li>Fix <code>MOI.Test.soc3test</code> for solvers not supporting infeasibility certificates (#839).</li><li>Implements <code>operate</code> for operators <code>*</code> and <code>/</code> between vector function and constant (#837).</li><li>Implements <code>show</code> for <code>MOI.Utilities.IndexMap</code> (#847).</li><li>Fix corner cases for mapping of variables in <code>MOI.Utilities.CachingOptimizer</code> and substitution of variables in <code>MOI.Bridges.AbstractBridgeOptimizer</code> (#848).</li><li>Fix transformation of constant terms for <code>MOI.Bridges.Constraint.SOCtoPSDBridge</code> and <code>MOI.Bridges.Constraint.RSOCtoPSDBridge</code> (#840).</li></ul><h2 id="v0.9.0-(August-13,-2019)"><a class="docs-heading-anchor" href="#v0.9.0-(August-13,-2019)">v0.9.0 (August 13, 2019)</a><a id="v0.9.0-(August-13,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.9.0-(August-13,-2019)" title="Permalink"></a></h2><ul><li>Support for Julia v0.6 and v0.7 was dropped (#714, #717).</li><li>A <code>MOI.Utilities.Model</code> implementation of <code>ModelLike</code>, this should replace most use cases of <code>MOI.Utilities.@model</code> (#781).</li><li><code>add_constrained_variable</code> and <code>add_constrained_variables</code> were added (#759).</li><li>Support for indicator constraints was added (#709, #712).</li><li><code>DualObjectiveValue</code> attribute was added (#473).</li><li><code>RawParameter</code> attribute was added (#733).</li><li>A <code>dual_set</code> function was added (#804).</li><li>A <code>Benchmarks</code> submodule was added to facilitate solver benchmarking (#769).</li><li>A <code>submit</code> function was added, this may for instance allow the user to submit solutions or cuts to the solver from a callback (#775).</li><li>The field of <code>ObjectiveValue</code> was renamed to <code>result_index</code> (#729).</li><li>The <code>_constant</code> and <code>Utilities.getconstant</code> function were renamed to <code>constant</code></li><li><code>REDUCTION_CERTIFICATE</code> result status was added (#734).</li><li>Abstract matrix sets were added (#731).</li><li>Testing improvements:<ul><li>The testing guideline was updated (#728).</li><li>Quadratic tests were added (#697).</li><li>Unit tests for <code>RawStatusString</code>, <code>SolveTime</code>, <code>Silent</code> and <code>SolverName</code> were added (#726, #741).</li><li>A rotated second-order cone test was added (#759).</li><li>A power cone test was added (#768).</li><li>Tests for <code>ZeroOne</code> variables with variable bounds were added (#772).</li><li>An unbounded test was added (#773).</li><li>Existing tests had a few updates (#702, #703, #763).</li></ul></li><li>Documentation improvements:<ul><li>Added a section on <code>CachingOptimizer</code> (#777).</li><li>Added a section on <code>UniversalFallback</code>, <code>Model</code> and <code>@model</code> (#762).</li><li>Transition the knapsack example to a doctest with MockOptimizer (#786).</li></ul></li><li>Utilities improvements:<ul><li>A <code>CleverDict</code> utility was added for a vector that automatically transform into a dictionary once a first index is removed (#767).</li><li>The <code>Utilities.constant</code> function was renamed to <code>Utilities.constant_vector</code> (#740).</li><li>Implement optimizer attributes for CachingOptimizer (#745).</li><li>Rename <code>Utilities.add_scalar_constraint</code> to <code>Utilities.normalize_and_add_constraint</code> (#801).</li><li><code>operate</code> with <code>vcat</code>, <code>SingleVariable</code> and <code>VectorOfVariables</code> now returns a <code>VectorOfVariables</code> (#616).</li><li>Fix a type piracy of <code>operate</code> (#784).</li><li>The <code>load_constraint</code> fallback signature was fixed (#760).</li><li>The <code>set_dot</code> function was extended to work with sparse arrays (#805).</li></ul></li><li>Bridges improvements:<ul><li>The bridges no longer store the constraint function and set before it is bridged, the bridges now have to implement <code>ConstraintFunction</code> and <code>ConstraintSet</code> if the user wants to recover them. As a consequence, the <code>@bridge</code> macro was removed (#722).</li><li>Bridge are now instantiated with a <code>bridge_constraint</code> function instead of using a constructor (#730).</li><li>Fix constraint attributes for bridges (#699).</li><li>Constraint bridges were moved to the <code>Bridges/Constraint</code> submodule so they should now inherit from <code>MOI.Bridges.Constraint.Abstract</code> and should implement <code>MOI.Bridges.Constraint.concrete_bridge_type</code> instead of <code>MOI.Bridges.concrete_bridge_type</code> (#756).</li><li>Variable bridges were added in (#759).</li><li>Various improvements (#746, #747).</li></ul></li></ul><h2 id="v0.8.4-(March-13,-2019)"><a class="docs-heading-anchor" href="#v0.8.4-(March-13,-2019)">v0.8.4 (March 13, 2019)</a><a id="v0.8.4-(March-13,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.8.4-(March-13,-2019)" title="Permalink"></a></h2><ul><li>Performance improvement in <code>default_copy_to</code> and bridge optimizer (#696).</li><li>Add <code>Silent</code> and implement setting optimizer attributes in caching and mock optimizers (#695).</li><li>Add <code>Functionize</code> bridges (SingleVariable and VectorOfVariables) (#659).</li><li>Minor typo fixes (#694).</li></ul><h2 id="v0.8.3-(March-6,-2019)"><a class="docs-heading-anchor" href="#v0.8.3-(March-6,-2019)">v0.8.3 (March 6, 2019)</a><a id="v0.8.3-(March-6,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.8.3-(March-6,-2019)" title="Permalink"></a></h2><ul><li>Use zero constant in scalar constraint function of <code>MOI.Test.copytest</code> (#691).</li><li>Fix variable deletion with <code>SingleVariable</code> objective function (#690).</li><li>Fix <code>LazyBridgeOptimizer</code> with bridges that add no constraints (#689).</li><li>Error message improvements (#673, #685, #686, #688).</li><li>Documentation improvements (#682, #683, #687).</li><li>Basis status:<ul><li>Remove <code>VariableBasisStatus</code> (#679).</li><li>Test <code>ConstraintBasisStatus</code> and implement it in bridges (#678).</li></ul></li><li>Fix inference of <code>NumberOfVariables</code> and <code>NumberOfConstraints</code> (#677).</li><li>Implement division between a quadratic function and a number (#675).</li></ul><h2 id="v0.8.2-(February-7,-2019)"><a class="docs-heading-anchor" href="#v0.8.2-(February-7,-2019)">v0.8.2 (February 7, 2019)</a><a id="v0.8.2-(February-7,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.8.2-(February-7,-2019)" title="Permalink"></a></h2><ul><li>Add <code>RawStatusString</code> attribute (#629).</li><li>Do not set names to the optimizer but only to the cache in <code>CachingOptimizer</code> (#638).</li><li>Make scalar MOI functions act as scalars in broadcast (#646).</li><li>Add function utilities:<ul><li>Implement <code>Base.zero</code> (#634), <code>Base.iszero</code> (#643), add missing arithmetic operations (#644, #645) and fix division (#648).</li><li>Add a <code>vectorize</code> function that turns a vector of <code>ScalarAffineFunction</code> into a <code>VectorAffineFunction</code> (#642).</li></ul></li><li>Improve support for starting values:<ul><li>Show a warning in copy when starting values are not supported instead of throwing an error (#630).</li><li>Fix <code>UniversalFallback</code> for getting an variable or constraint attribute set to no indices (#623).</li><li>Add a test in <code>contlineartest</code> with partially set <code>VariablePrimalStart</code>.</li></ul></li><li>Bridges improvements:<ul><li>Fix <code>StackOverFlow</code> in <code>LazyBridgeOptimizer</code> when there is a cycle in the graph of bridges.</li><li>Add <code>Slack</code> bridges (#610, #650).</li><li>Add <code>FlipSign</code> bridges (#658).</li></ul></li><li>Add tests with duplicate coefficients in <code>ScalarAffineFunction</code> and <code>VectorAffineFunction</code> (#639).</li><li>Use tolerance to compare <code>VariablePrimal</code> in <code>rotatedsoc1</code> test (#632).</li><li>Use a zero constant in <code>ScalarAffineFunction</code> of constraints in <code>psdt2</code> (#622).</li></ul><h2 id="v0.8.1-(January-7,-2019)"><a class="docs-heading-anchor" href="#v0.8.1-(January-7,-2019)">v0.8.1 (January 7, 2019)</a><a id="v0.8.1-(January-7,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.8.1-(January-7,-2019)" title="Permalink"></a></h2><ul><li>Adding an NLP objective now overrides any objective set using the <code>ObjectiveFunction</code> attribute (#619).</li><li>Rename <code>fullbridgeoptimizer</code> into <code>full_bridge_optimizer</code> (#621).</li><li>Allow custom constraint types with <code>full_bridge_optimizer</code> (#617).</li><li>Add <code>Vectorize</code> bridge which transforms scalar linear constraints into vector linear constraints (#615).</li></ul><h2 id="v0.8.0-(December-18,-2018)"><a class="docs-heading-anchor" href="#v0.8.0-(December-18,-2018)">v0.8.0 (December 18, 2018)</a><a id="v0.8.0-(December-18,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.8.0-(December-18,-2018)" title="Permalink"></a></h2><ul><li>Rename all enum values to follow the JuMP naming guidelines for constants, for example, <code>Optimal</code> becomes <code>OPTIMAL</code>, and <code>DualInfeasible</code> becomes <code>DUAL_INFEASIBLE</code>.</li><li>Rename CachingOptimizer methods for style compliance.</li><li>Add an <code>MOI.TerminationStatusCode</code> called <code>ALMOST_DUAL_INFEASIBLE</code>.</li></ul><h2 id="v0.7.0-(December-13,-2018)"><a class="docs-heading-anchor" href="#v0.7.0-(December-13,-2018)">v0.7.0 (December 13, 2018)</a><a id="v0.7.0-(December-13,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.7.0-(December-13,-2018)" title="Permalink"></a></h2><ul><li>Test that <code>MOI.TerminationStatus</code> is <code>MOI.OptimizeNotCalled</code> before <code>MOI.optimize!</code> is called.</li><li>Check <code>supports_default_copy_to</code> in tests (#594).</li><li>Key pieces of information like optimality, infeasibility, etc., are now reported through <code>TerminationStatusCode</code>. It is typically no longer necessary to check the result statuses in addition to the termination status.</li><li>Add perspective dimension to log-det cone (#593).</li></ul><h2 id="v0.6.4-(November-27,-2018)"><a class="docs-heading-anchor" href="#v0.6.4-(November-27,-2018)">v0.6.4 (November 27, 2018)</a><a id="v0.6.4-(November-27,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.6.4-(November-27,-2018)" title="Permalink"></a></h2><ul><li>Add <code>OptimizeNotCalled</code> termination status (#577) and improve documentation of other statuses (#575).</li><li>Add a solver naming guideline (#578).</li><li>Make <code>FeasibilitySense</code> the default <code>ObjectiveSense</code> (#579).</li><li>Fix <code>Utilities.@model</code> and <code>Bridges.@bridge</code> macros for functions and sets defined outside MOI (#582).</li><li>Document solver-specific attributes (#580) and implement them in <code>Utilities.CachingOptimizer</code> (#565).</li></ul><h2 id="v0.6.3-(November-16,-2018)"><a class="docs-heading-anchor" href="#v0.6.3-(November-16,-2018)">v0.6.3 (November 16, 2018)</a><a id="v0.6.3-(November-16,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.6.3-(November-16,-2018)" title="Permalink"></a></h2><ul><li>Variables and constraints are now allowed to have duplicate names. An error is thrown only on lookup. This change breaks some existing tests. (#549)</li><li>Attributes may now be partially set (some values could be <code>nothing</code>). (#563)</li><li>Performance improvements in Utilities.Model (#549, #567, #568)</li><li>Fix bug in QuadtoSOC (#558).</li><li>New <code>supports_default_copy_to</code> method that optimizers should implement to control caching behavior.</li><li>Documentation improvements.</li></ul><h2 id="v0.6.2-(October-26,-2018)"><a class="docs-heading-anchor" href="#v0.6.2-(October-26,-2018)">v0.6.2 (October 26, 2018)</a><a id="v0.6.2-(October-26,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.6.2-(October-26,-2018)" title="Permalink"></a></h2><ul><li>Improve hygiene of <code>@model</code> macro (#544).</li><li>Fix bug in copy tests (#543).</li><li>Fix bug in UniversalFallback attribute getter (#540).</li><li>Allow all correct solutions for <code>solve_blank_obj</code> unit test (#537).</li><li>Add errors for Allocate-Load and bad constraints (#534).</li><li>[performance] Add specialized implementation of <code>hash</code> for <code>VariableIndex</code> (#533).</li><li>[performance] Construct the name to object dictionaries lazily in model (#535).</li><li>Add the <code>QuadtoSOC</code> bridge which transforms <code>ScalarQuadraticFunction</code> constraints into <code>RotatedSecondOrderCone</code> (#483).</li></ul><h2 id="v0.6.1-(September-22,-2018)"><a class="docs-heading-anchor" href="#v0.6.1-(September-22,-2018)">v0.6.1 (September 22, 2018)</a><a id="v0.6.1-(September-22,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.6.1-(September-22,-2018)" title="Permalink"></a></h2><ul><li>Enable <code>PositiveSemidefiniteConeSquare</code> set and quadratic functions in <code>MOIB.fullbridgeoptimizer</code> (#524).</li><li>Add warning in the bridge between <code>PositiveSemidefiniteConeSquare</code> and <code>PositiveSemidefiniteConeTriangle</code> when the matrix is almost symmetric (#522).</li><li>Modify <code>MOIT.copytest</code> to not add multiples constraints on the same variable (#521).</li><li>Add missing keyword argument in one of <code>MOIU.add_scalar_constraint</code> methods (#520).</li></ul><h2 id="v0.6.0-(August-30,-2018)"><a class="docs-heading-anchor" href="#v0.6.0-(August-30,-2018)">v0.6.0 (August 30, 2018)</a><a id="v0.6.0-(August-30,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.6.0-(August-30,-2018)" title="Permalink"></a></h2><ul><li>The <code>MOIU.@model</code> and <code>MOIB.@bridge</code> macros now support functions and sets defined in external modules. As a consequence, function and set names in the macro arguments need to be prefixed by module name.</li><li>Rename functions according to the <a href="https://jump.dev/JuMP.jl/stable/developers/style/">JuMP style guide</a>:<ul><li><code>copy!</code> with keyword arguments <code>copynames</code> and <code>warnattributes</code> -&gt; <code>copy_to</code> with keyword arguments <code>copy_names</code> and <code>warn_attributes</code>;</li><li><code>set!</code> -&gt; <code>set</code>;</li><li><code>addvariable[s]!</code> -&gt; <code>add_variable[s]</code>;</li><li><code>supportsconstraint</code> -&gt; <code>supports_constraint</code>;</li><li><code>addconstraint[s]!</code> -&gt; <code>add_constraint[s]</code>;</li><li><code>isvalid</code> -&gt; <code>is_valid</code>;</li><li><code>isempty</code> -&gt; <code>is_empty</code>;</li><li><code>Base.delete!</code> -&gt; <code>delete</code>;</li><li><code>modify!</code> -&gt; <code>modify</code>;</li><li><code>transform!</code> -&gt; <code>transform</code>;</li><li><code>initialize!</code> -&gt; <code>initialize</code>;</li><li><code>write</code> -&gt; <code>write_to_file</code>; and</li><li><code>read!</code> -&gt; <code>read_from_file</code>.</li></ul></li><li>Remove <code>free!</code> (use <code>Base.finalize</code> instead).</li><li>Add the <code>SquarePSD</code> bridge which transforms <code>PositiveSemidefiniteConeTriangle</code> constraints into <code>PositiveSemidefiniteConeTriangle</code>.</li><li>Add result fallback for <code>ConstraintDual</code> of variable-wise constraint, <code>ConstraintPrimal</code> and <code>ObjectiveValue</code>.</li><li>Add tests for <code>ObjectiveBound</code>.</li><li>Add test for empty rows in vector linear constraint.</li><li>Rework errors: <code>CannotError</code> has been renamed <code>NotAllowedError</code> and the distinction between <code>UnsupportedError</code> and <code>NotAllowedError</code> is now about whether the element is not supported (for example, it cannot be copied a model containing this element) or the operation is not allowed (either because it is not implemented, because it cannot be performed in the current state of the model, or because it cannot be performed for a specific index)</li><li><code>canget</code> is removed. <code>NoSolution</code> is added as a result status to indicate that the solver does not have either a primal or dual solution available (See #479).</li></ul><h2 id="v0.5.0-(August-5,-2018)"><a class="docs-heading-anchor" href="#v0.5.0-(August-5,-2018)">v0.5.0 (August 5, 2018)</a><a id="v0.5.0-(August-5,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.5.0-(August-5,-2018)" title="Permalink"></a></h2><ul><li>Fix names with CachingOptimizer.</li><li>Cleanup thanks to @mohamed82008.</li><li>Added a universal fallback for constraints.</li><li>Fast utilities for function canonicalization thanks to @rdeits.</li><li>Renamed <code>dimension</code> field to <code>side_dimension</code> in the context of matrix-like sets.</li><li>New and improved tests for cases like duplicate terms and <code>ObjectiveBound</code>.</li><li>Removed <code>cantransform</code>, <code>canaddconstraint</code>, <code>canaddvariable</code>, <code>canset</code>, <code>canmodify</code>, and <code>candelete</code> functions from the API. They are replaced by a new set of errors that are thrown: Subtypes of <code>UnsupportedError</code> indicate unsupported operations, while subtypes of <code>CannotError</code> indicate operations that cannot be performed in the current state.</li><li>The API for <code>copy!</code> is updated to remove the CopyResult type.</li><li>Updates for the new JuMP style guide.</li></ul><h2 id="v0.4.1-(June-28,-2018)"><a class="docs-heading-anchor" href="#v0.4.1-(June-28,-2018)">v0.4.1 (June 28, 2018)</a><a id="v0.4.1-(June-28,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.4.1-(June-28,-2018)" title="Permalink"></a></h2><ul><li>Fixes vector function modification on 32 bits.</li><li>Fixes Bellman-Ford algorithm for bridges.</li><li>Added an NLP test with <code>FeasibilitySense</code>.</li><li>Update modification documentation.</li></ul><h2 id="v0.4.0-(June-23,-2018)"><a class="docs-heading-anchor" href="#v0.4.0-(June-23,-2018)">v0.4.0 (June 23, 2018)</a><a id="v0.4.0-(June-23,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.4.0-(June-23,-2018)" title="Permalink"></a></h2><ul><li>Helper constructors for <code>VectorAffineTerm</code> and <code>VectorQuadraticTerm</code>.</li><li>Added <code>modify_lhs</code> to <code>TestConfig</code>.</li><li>Additional unit tests for optimizers.</li><li>Added a type parameter to <code>CachingOptimizer</code> for the <code>optimizer</code> field.</li><li>New API for problem modification (#388)</li><li>Tests pass without deprecation warnings on Julia 0.7.</li><li>Small fixes and documentation updates.</li></ul><h2 id="v0.3.0-(May-25,-2018)"><a class="docs-heading-anchor" href="#v0.3.0-(May-25,-2018)">v0.3.0 (May 25, 2018)</a><a id="v0.3.0-(May-25,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.3.0-(May-25,-2018)" title="Permalink"></a></h2><ul><li>Functions have been redefined to use arrays-of-structs instead of structs-of-arrays.</li><li>Improvements to <code>MockOptimizer</code>.</li><li>Significant changes to <code>Bridges</code>.</li><li>New and improved unit tests.</li><li>Fixes for Julia 0.7.</li></ul><h2 id="v0.2.0-(April-24,-2018)"><a class="docs-heading-anchor" href="#v0.2.0-(April-24,-2018)">v0.2.0 (April 24, 2018)</a><a id="v0.2.0-(April-24,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.2.0-(April-24,-2018)" title="Permalink"></a></h2><ul><li>Improvements to and better coverage of <code>Tests</code>.</li><li>Documentation fixes.</li><li><code>SolverName</code> attribute.</li><li>Changes to the NLP interface (new definition of variable order and arrays of structs for bound pairs and sparsity patterns).</li><li>Addition of NLP tests.</li><li>Introduction of <code>UniversalFallback</code>.</li><li><code>copynames</code> keyword argument to <code>MOI.copy!</code>.</li><li>Add Bridges submodule.</li></ul><h2 id="v0.1.0-(February-28,-2018)"><a class="docs-heading-anchor" href="#v0.1.0-(February-28,-2018)">v0.1.0 (February 28, 2018)</a><a id="v0.1.0-(February-28,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#v0.1.0-(February-28,-2018)" title="Permalink"></a></h2><ul><li>Initial public release.</li><li>The framework for MOI was developed at the JuMP-dev workshop at MIT in June 2017 as a sorely needed replacement for MathProgBase.</li></ul></article><nav class="docs-footer"><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/developer/checklists/index.html b/previews/PR3525/moi/developer/checklists/index.html
new file mode 100644
index 00000000000..e5174165510
--- /dev/null
+++ b/previews/PR3525/moi/developer/checklists/index.html
@@ -0,0 +1,116 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Checklists · JuMP</title><meta name="title" content="Checklists · JuMP"/><meta property="og:title" content="Checklists · JuMP"/><meta property="twitter:title" content="Checklists · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox" checked/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Checklists</a><ul class="internal"><li><a class="tocitem" href="#Making-a-release"><span>Making a release</span></a></li><li><a class="tocitem" href="#Adding-a-new-set"><span>Adding a new set</span></a></li><li><a class="tocitem" href="#Adding-a-new-bridge"><span>Adding a new bridge</span></a></li><li><a class="tocitem" href="#Updating-MathOptFormat"><span>Updating MathOptFormat</span></a></li></ul></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Developer Docs</a></li><li class="is-active"><a href>Checklists</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Checklists</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/developer/checklists.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Checklists"><a class="docs-heading-anchor" href="#Checklists">Checklists</a><a id="Checklists-1"></a><a class="docs-heading-anchor-permalink" href="#Checklists" title="Permalink"></a></h1><p>The purpose of this page is to collate a series of checklists for commonly performed changes to the source code of MathOptInterface.</p><p>In each case, copy the checklist into the description of the pull request.</p><h2 id="Making-a-release"><a class="docs-heading-anchor" href="#Making-a-release">Making a release</a><a id="Making-a-release-1"></a><a class="docs-heading-anchor-permalink" href="#Making-a-release" title="Permalink"></a></h2><p>Use this checklist when making a release of the MathOptInterface repository.</p><pre><code class="nohighlight hljs">## Basic
+
+ - [ ] `version` field of `Project.toml` has been updated
+       - If a breaking change, increment the MAJOR field and reset others to 0
+       - If adding new features, increment the MINOR field and reset PATCH to 0
+       - If adding bug fixes or documentation changes, increment the PATCH field
+
+## Documentation
+
+ - [ ] Add a new entry to `docs/src/changelog.md`, following existing style
+
+## Tests
+
+ - [ ] The `solver-tests.yml` GitHub action does not have unexpected failures.
+       To run the action, go to:
+       https://github.com/jump-dev/MathOptInterface.jl/actions/workflows/solver-tests.yml
+       and click &quot;Run workflow&quot;</code></pre><h2 id="Adding-a-new-set"><a class="docs-heading-anchor" href="#Adding-a-new-set">Adding a new set</a><a id="Adding-a-new-set-1"></a><a class="docs-heading-anchor-permalink" href="#Adding-a-new-set" title="Permalink"></a></h2><p>Use this checklist when adding a new set to the MathOptInterface repository.</p><pre><code class="nohighlight hljs">## Basic
+
+ - [ ] Add a new `AbstractScalarSet` or `AbstractVectorSet` to `src/sets.jl`
+ - [ ] If `isbitstype(S) == false`, implement `Base.copy(set::S)`
+ - [ ] If `isbitstype(S) == false`, implement `Base.:(==)(x::S, y::S)`
+ - [ ] If an `AbstractVectorSet`, implement `dimension(set::S)`, unless the
+       dimension is given by `set.dimension`.
+
+## Utilities
+
+ - [ ] If an `AbstractVectorSet`, implement `Utilities.set_dot`,
+       unless the dot product between two vectors in the set is equivalent to
+       `LinearAlgebra.dot`
+ - [ ] If an `AbstractVectorSet`, implement `Utilities.set_with_dimension` in
+       `src/Utilities/matrix_of_constraints.jl`
+ - [ ] Add the set to the `@model` macro at the bottom of `src/Utilities.model.jl`
+
+## Documentation
+
+ - [ ] Add a docstring, which gives the mathematical definition of the set,
+       along with an `## Example` block containing a `jldoctest`
+ - [ ] Add the docstring to `docs/src/reference/standard_form.md`
+ - [ ] Add the set to the relevant table in `docs/src/manual/standard_form.md`
+
+## Tests
+
+ - [ ] Define a new `_set(::Type{S})` method in `src/Test/test_basic_constraint.jl`
+       and add the name of the set to the list at the bottom of that files
+ - [ ] If the set has any checks in its constructor, add tests to `test/sets.jl`
+
+## MathOptFormat
+
+ - [ ] Open an issue at `https://github.com/jump-dev/MathOptFormat` to add
+       support for the new set {{ replace with link to the issue }}
+
+## Optional
+
+ - [ ] Implement `dual_set(::S)` and `dual_set_type(::Type{S})`
+ - [ ] Add new tests to the `Test` submodule exercising your new set
+ - [ ] Add new bridges to convert your set into more commonly used sets</code></pre><h2 id="Adding-a-new-bridge"><a class="docs-heading-anchor" href="#Adding-a-new-bridge">Adding a new bridge</a><a id="Adding-a-new-bridge-1"></a><a class="docs-heading-anchor-permalink" href="#Adding-a-new-bridge" title="Permalink"></a></h2><p>Use this checklist when adding a new bridge to the MathOptInterface repository.</p><p>The steps are mostly the same, but locations depend on whether the bridge is a <code>Constraint</code>, <code>Objective</code>, or <code>Variable</code> bridge. In each case below, replace <code>XXX</code> with the appropriate type of bridge.</p><pre><code class="nohighlight hljs">## Basic
+
+ - [ ] Create a new file in `src/Bridges/XXX/bridges`
+ - [ ] Define the bridge, following existing examples. The name of the bridge
+       struct must end in `Bridge`
+ - [ ] Check if your bridge can be a subtype of [`MOI.Bridges.Constraint.SetMapBridge`](@ref)
+ - [ ] Define a new `const` that is a `SingleBridgeOptimizer` wrapping the
+       new bridge. The name of the const must be the name of the bridge, less
+       the `Bridge` suffix
+ - [ ] `include` the file in `src/Bridges/XXX/bridges/XXX.jl`
+ - [ ] If the bridge should be enabled by default, add the bridge to
+       `add_all_bridges` at the bottom of `src/Bridges/XXX/XXX.jl`
+
+## Tests
+
+ - [ ] Create a new file in the appropriate subdirectory of `tests/Bridges/XXX`
+ - [ ] Use `MOI.Bridges.runtests` to test various inputs and outputs of the
+       bridge
+ - [ ] If, after opening the pull request to add the bridge, some lines are not
+       covered by the tests, add additional bridge-specific tests to cover the
+       untested lines.
+
+## Documentation
+
+ - [ ] Add a docstring which uses the same template as existing bridges.
+ - [ ] Add the docstring to `docs/src/submodules/Bridges/list_of_bridges.md`
+
+## Final touch
+
+If the bridge depends on run-time values of other variables and constraints in
+the model:
+
+ - [ ] Implement `MOI.Utilities.needs_final_touch(::Bridge)`
+ - [ ] Implement `MOI.Utilities.final_touch(::Bridge, ::MOI.ModelLike)`
+ - [ ] Ensure that `final_touch` can be called multiple times in a row</code></pre><h2 id="Updating-MathOptFormat"><a class="docs-heading-anchor" href="#Updating-MathOptFormat">Updating MathOptFormat</a><a id="Updating-MathOptFormat-1"></a><a class="docs-heading-anchor-permalink" href="#Updating-MathOptFormat" title="Permalink"></a></h2><p>Use this checklist when updating the version of MathOptFormat.</p><pre><code class="nohighlight hljs">## Basic
+
+ - [ ] The file at `src/FileFormats/MOF/mof.X.Y.schema.json` is updated
+ - [ ] The constants `SCHEMA_PATH`, `VERSION`, and `SUPPORTED_VERSIONS` are
+       updated in `src/FileFormats/MOF/MOF.jl`
+
+## New sets
+
+ - [ ] New sets are added to the `@model` in `src/FileFormats/MOF/MOF.jl`
+ - [ ] New sets are added to the `@enum` in `src/FileFormats/MOF/read.jl`
+ - [ ] `set_to_moi` is defined for each set in `src/FileFormats/MOF/read.jl`
+ - [ ] `head_name` is defined for each set in `src/FileFormats/MOF/write.jl`
+ - [ ] A new unit test calling `_test_model_equality` is aded to
+       `test/FileFormats/MOF/MOF.jl`
+
+## Tests
+
+ - [ ] The version field in `test/FileFormats/MOF/nlp.mof.json` is updated
+
+## Documentation
+
+ - [ ] The version fields are updated in `docs/src/submodules/FileFormats/overview.md`</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../submodules/Test/reference/">« API Reference</a><a class="docs-footer-nextpage" href="../../release_notes/">Release notes »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/index.html b/previews/PR3525/moi/index.html
new file mode 100644
index 00000000000..24fff695079
--- /dev/null
+++ b/previews/PR3525/moi/index.html
@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL=".."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../assets/documenter.js"></script><script src="../search_index.js"></script><script src="../siteinfo.js"></script><script src="../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../assets/themeswap.js"></script><link href="../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../"><img class="docs-light-only" src="../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../">Introduction</a></li><li><a class="tocitem" href="../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../manual/models/">Models</a></li><li><a class="tocitem" href="../manual/variables/">Variables</a></li><li><a class="tocitem" href="../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../manual/containers/">Containers</a></li><li><a class="tocitem" href="../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox" checked/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#What-is-MathOptInterface?"><span>What is MathOptInterface?</span></a></li><li><a class="tocitem" href="#How-the-documentation-is-structured"><span>How the documentation is structured</span></a></li><li><a class="tocitem" href="#Citing-MathOptInterface"><span>Citing MathOptInterface</span></a></li></ul></li><li><a class="tocitem" href="background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="manual/models/">Models</a></li><li><a class="tocitem" href="manual/variables/">Variables</a></li><li><a class="tocitem" href="manual/constraints/">Constraints</a></li><li><a class="tocitem" href="manual/solutions/">Solutions</a></li><li><a class="tocitem" href="manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="background/duality/">Duality</a></li><li><a class="tocitem" href="background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="reference/models/">Models</a></li><li><a class="tocitem" href="reference/variables/">Variables</a></li><li><a class="tocitem" href="reference/constraints/">Constraints</a></li><li><a class="tocitem" href="reference/modification/">Modifications</a></li><li><a class="tocitem" href="reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Introduction</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/index.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="moi_documentation"><a class="docs-heading-anchor" href="#moi_documentation">Introduction</a><a id="moi_documentation-1"></a><a class="docs-heading-anchor-permalink" href="#moi_documentation" title="Permalink"></a></h1><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This documentation in this section is a copy of the official MathOptInterface documentation available at <a href="https://jump.dev/MathOptInterface.jl/v1.20.0">https://jump.dev/MathOptInterface.jl/v1.20.0</a>. It is included here to make it easier to link concepts between JuMP and MathOptInterface.</p></div></div><h2 id="What-is-MathOptInterface?"><a class="docs-heading-anchor" href="#What-is-MathOptInterface?">What is MathOptInterface?</a><a id="What-is-MathOptInterface?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-MathOptInterface?" title="Permalink"></a></h2><p><a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface.jl</a> (MOI) is an abstraction layer designed to provide a unified interface to mathematical optimization solvers so that users do not need to understand multiple solver-specific APIs.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This documentation is aimed at developers writing software interfaces to solvers and modeling languages using the MathOptInterface API. If you are a user interested in solving optimization problems, we encourage you instead to use MOI through a higher-level modeling interface like <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> or <a href="https://github.com/jump-dev/Convex.jl">Convex.jl</a>.</p></div></div><h2 id="How-the-documentation-is-structured"><a class="docs-heading-anchor" href="#How-the-documentation-is-structured">How the documentation is structured</a><a id="How-the-documentation-is-structured-1"></a><a class="docs-heading-anchor-permalink" href="#How-the-documentation-is-structured" title="Permalink"></a></h2><p>Having a high-level overview of how this documentation is structured will help you know where to look for certain things.</p><ul><li>The <strong>Tutorials</strong> section contains articles on how to use and implement the MathOptInteraface API. Look here if you want to write a model in MOI, or write an interface to a new solver.</li><li>The <strong>Manual</strong> contains short code-snippets that explain how to use the MOI API. Look here for more details on particular areas of MOI.</li><li>The <strong>Background</strong> section contains articles on the theory behind MathOptInterface. Look here if you want to understand <em>why</em>, rather than <em>how</em>.</li><li>The <strong>API Reference</strong> contains a complete list of functions and types that comprise the MOI API. Look here is you want to know how to use (or implement) a particular function.</li><li>The <strong>Submodules</strong> section contains stand-alone documentation for each of the submodules within MOI. These submodules are not required to interface a solver with MOI, but they make the job much easier.</li></ul><h2 id="Citing-MathOptInterface"><a class="docs-heading-anchor" href="#Citing-MathOptInterface">Citing MathOptInterface</a><a id="Citing-MathOptInterface-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-MathOptInterface" title="Permalink"></a></h2><p>A <a href="https://arxiv.org/abs/2002.03447">paper describing the design and features of MathOptInterface</a> is available on <a href="https://arxiv.org">arXiv</a>.</p><p>If you find MathOptInterface useful in your work, we kindly request that you cite the following paper:</p><pre><code class="language-bibtex hljs">@article{legat2021mathoptinterface,
+    title={{MathOptInterface}: a data structure for mathematical optimization problems},
+    author={Legat, Beno{\^\i}t and Dowson, Oscar and Garcia, Joaquim Dias and Lubin, Miles},
+    journal={INFORMS Journal on Computing},
+    year={2021},
+    doi={10.1287/ijoc.2021.1067},
+    publisher={INFORMS}
+}</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../extensions/DimensionalData/">« rafaqz/DimensionalData.jl</a><a class="docs-footer-nextpage" href="background/motivation/">Motivation »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/manual/constraints/index.html b/previews/PR3525/moi/manual/constraints/index.html
new file mode 100644
index 00000000000..2b04b0cf318
--- /dev/null
+++ b/previews/PR3525/moi/manual/constraints/index.html
@@ -0,0 +1,26 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Constraints · JuMP</title><meta name="title" content="Constraints · JuMP"/><meta property="og:title" content="Constraints · JuMP"/><meta property="twitter:title" content="Constraints · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li class="is-active"><a class="tocitem" href>Constraints</a><ul class="internal"><li><a class="tocitem" href="#Add-a-constraint"><span>Add a constraint</span></a></li><li><a class="tocitem" href="#Delete-a-constraint"><span>Delete a constraint</span></a></li><li><a class="tocitem" href="#Constraint-attributes"><span>Constraint attributes</span></a></li><li><a class="tocitem" href="#Constraints-by-function-set-pairs"><span>Constraints by function-set pairs</span></a></li><li><a class="tocitem" href="#JuMP-mapping"><span>JuMP mapping</span></a></li></ul></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Constraints</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Constraints</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/constraints.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Constraints"><a class="docs-heading-anchor" href="#Constraints">Constraints</a><a id="Constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Constraints" title="Permalink"></a></h1><h2 id="Add-a-constraint"><a class="docs-heading-anchor" href="#Add-a-constraint">Add a constraint</a><a id="Add-a-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-constraint" title="Permalink"></a></h2><p>Use <a href="../../reference/constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a> to add a single constraint.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Nonnegatives(2))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonnegatives}(1)</code></pre><p><a href="../../reference/constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a> returns a <a href="../../reference/constraints/#MathOptInterface.ConstraintIndex"><code>ConstraintIndex</code></a> type, which is used to refer to the added constraint in other calls.</p><p>Check if a <a href="../../reference/constraints/#MathOptInterface.ConstraintIndex"><code>ConstraintIndex</code></a> is valid using <a href="../../reference/constraints/#MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.ConstraintIndex}"><code>is_valid</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.is_valid(model, c)
+true</code></pre><p>Use <a href="../../reference/constraints/#MathOptInterface.add_constraints"><code>add_constraints</code></a> to add a number of constraints of the same type.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraints(
+           model,
+           [x[1], x[2]],
+           [MOI.GreaterThan(0.0), MOI.GreaterThan(1.0)]
+       )
+2-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}}:
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(1)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(2)</code></pre><p>This time, a vector of <a href="../../reference/constraints/#MathOptInterface.ConstraintIndex"><code>ConstraintIndex</code></a> are returned.</p><p>Use <a href="../../reference/constraints/#MathOptInterface.supports_constraint"><code>supports_constraint</code></a> to check if the model supports adding a constraint type.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.supports_constraint(
+           model,
+           MOI.VariableIndex,
+           MOI.GreaterThan{Float64},
+       )
+true</code></pre><h2 id="Delete-a-constraint"><a class="docs-heading-anchor" href="#Delete-a-constraint">Delete a constraint</a><a id="Delete-a-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Delete-a-constraint" title="Permalink"></a></h2><p>Use <a href="../../reference/variables/#MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}"><code>delete</code></a> to delete a constraint.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.delete(model, c)
+
+julia&gt; MOI.is_valid(model, c)
+false</code></pre><h2 id="Constraint-attributes"><a class="docs-heading-anchor" href="#Constraint-attributes">Constraint attributes</a><a id="Constraint-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-attributes" title="Permalink"></a></h2><p>The following attributes are available for constraints:</p><ul><li><a href="../../reference/constraints/#MathOptInterface.ConstraintName"><code>ConstraintName</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>ConstraintPrimalStart</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintDualStart"><code>ConstraintDualStart</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>ConstraintPrimal</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintDual"><code>ConstraintDual</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintBasisStatus"><code>ConstraintBasisStatus</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.CanonicalConstraintFunction"><code>CanonicalConstraintFunction</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintSet"><code>ConstraintSet</code></a></li></ul><p>Get and set these attributes using <a href="../../reference/models/#MathOptInterface.get"><code>get</code></a> and <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.set(model, MOI.ConstraintName(), c, &quot;con_c&quot;)
+
+julia&gt; MOI.get(model, MOI.ConstraintName(), c)
+&quot;con_c&quot;</code></pre><h2 id="Constraints-by-function-set-pairs"><a class="docs-heading-anchor" href="#Constraints-by-function-set-pairs">Constraints by function-set pairs</a><a id="Constraints-by-function-set-pairs-1"></a><a class="docs-heading-anchor-permalink" href="#Constraints-by-function-set-pairs" title="Permalink"></a></h2><p>Below is a list of common constraint types and how they are represented as function-set pairs in MOI. In the notation below, <span>$x$</span> is a vector of decision variables, <span>$x_i$</span> is a scalar decision variable, <span>$\alpha, \beta$</span> are scalar constants, <span>$a, b$</span> are constant vectors, <code>A</code> is a constant matrix and <span>$\mathbb{R}_+$</span> (resp. <span>$\mathbb{R}_-$</span>) is the set of non-negative (resp. non-positive) real numbers.</p><h3 id="Linear-constraints"><a class="docs-heading-anchor" href="#Linear-constraints">Linear constraints</a><a id="Linear-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Linear-constraints" title="Permalink"></a></h3><table><tr><th style="text-align: right">Mathematical Constraint</th><th style="text-align: right">MOI Function</th><th style="text-align: right">MOI Set</th></tr><tr><td style="text-align: right"><span>$a^Tx \le \beta$</span></td><td style="text-align: right"><code>ScalarAffineFunction</code></td><td style="text-align: right"><code>LessThan</code></td></tr><tr><td style="text-align: right"><span>$a^Tx \ge \alpha$</span></td><td style="text-align: right"><code>ScalarAffineFunction</code></td><td style="text-align: right"><code>GreaterThan</code></td></tr><tr><td style="text-align: right"><span>$a^Tx = \beta$</span></td><td style="text-align: right"><code>ScalarAffineFunction</code></td><td style="text-align: right"><code>EqualTo</code></td></tr><tr><td style="text-align: right"><span>$\alpha \le a^Tx \le \beta$</span></td><td style="text-align: right"><code>ScalarAffineFunction</code></td><td style="text-align: right"><code>Interval</code></td></tr><tr><td style="text-align: right"><span>$x_i \le \beta$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>LessThan</code></td></tr><tr><td style="text-align: right"><span>$x_i \ge \alpha$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>GreaterThan</code></td></tr><tr><td style="text-align: right"><span>$x_i = \beta$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>EqualTo</code></td></tr><tr><td style="text-align: right"><span>$\alpha \le x_i \le \beta$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>Interval</code></td></tr><tr><td style="text-align: right"><span>$Ax + b \in \mathbb{R}_+^n$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>Nonnegatives</code></td></tr><tr><td style="text-align: right"><span>$Ax + b \in \mathbb{R}_-^n$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>Nonpositives</code></td></tr><tr><td style="text-align: right"><span>$Ax + b = 0$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>Zeros</code></td></tr></table><p>By convention, solvers are not expected to support nonzero constant terms in the <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>s the first four rows of the preceding table because they are redundant with the parameters of the sets. For example, encode <span>$2x + 1 \le 2$</span> as <span>$2x \le 1$</span>.</p><p>Constraints with <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> in <a href="../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a>, <a href="../../reference/standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a>, <a href="../../reference/standard_form/#MathOptInterface.EqualTo"><code>EqualTo</code></a>, or <a href="../../reference/standard_form/#MathOptInterface.Interval"><code>Interval</code></a> sets have a natural interpretation as variable bounds. As such, it is typically not natural to impose multiple lower- or upper-bounds on the same variable, and the solver interfaces will throw respectively <a href="../../reference/errors/#MathOptInterface.LowerBoundAlreadySet"><code>LowerBoundAlreadySet</code></a> or <a href="../../reference/errors/#MathOptInterface.UpperBoundAlreadySet"><code>UpperBoundAlreadySet</code></a>.</p><p>Moreover, adding two <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraints on the same variable with the same set is impossible because they share the same index as it is the index of the variable, see <a href="../../reference/constraints/#MathOptInterface.ConstraintIndex"><code>ConstraintIndex</code></a>.</p><p>It is natural, however, to impose upper- and lower-bounds separately as two different constraints on a single variable. The difference between imposing bounds by using a single <a href="../../reference/standard_form/#MathOptInterface.Interval"><code>Interval</code></a> constraint and by using separate <a href="../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a> and <a href="../../reference/standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a> constraints is that the latter will allow the solver to return separate dual multipliers for the two bounds, while the former will allow the solver to return only a single dual for the interval constraint.</p><h3 id="Conic-constraints"><a class="docs-heading-anchor" href="#Conic-constraints">Conic constraints</a><a id="Conic-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Conic-constraints" title="Permalink"></a></h3><table><tr><th style="text-align: right">Mathematical Constraint</th><th style="text-align: right">MOI Function</th><th style="text-align: right">MOI Set</th></tr><tr><td style="text-align: right"><span>$\lVert Ax + b\rVert_2 \le c^Tx + d$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>SecondOrderCone</code></td></tr><tr><td style="text-align: right"><span>$y \ge \lVert x \rVert_2$</span></td><td style="text-align: right"><code>VectorOfVariables</code></td><td style="text-align: right"><code>SecondOrderCone</code></td></tr><tr><td style="text-align: right"><span>$2yz \ge \lVert x \rVert_2^2, y,z \ge 0$</span></td><td style="text-align: right"><code>VectorOfVariables</code></td><td style="text-align: right"><code>RotatedSecondOrderCone</code></td></tr><tr><td style="text-align: right"><span>$(a_1^Tx + b_1,a_2^Tx + b_2,a_3^Tx + b_3) \in \mathcal{E}$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>ExponentialCone</code></td></tr><tr><td style="text-align: right"><span>$A(x) \in \mathcal{S}_+$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>PositiveSemidefiniteConeTriangle</code></td></tr><tr><td style="text-align: right"><span>$B(x) \in \mathcal{S}_+$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>PositiveSemidefiniteConeSquare</code></td></tr><tr><td style="text-align: right"><span>$x \in \mathcal{S}_+$</span></td><td style="text-align: right"><code>VectorOfVariables</code></td><td style="text-align: right"><code>PositiveSemidefiniteConeTriangle</code></td></tr><tr><td style="text-align: right"><span>$x \in \mathcal{S}_+$</span></td><td style="text-align: right"><code>VectorOfVariables</code></td><td style="text-align: right"><code>PositiveSemidefiniteConeSquare</code></td></tr></table><p>where <span>$\mathcal{E}$</span> is the exponential cone (see <a href="../../reference/standard_form/#MathOptInterface.ExponentialCone"><code>ExponentialCone</code></a>), <span>$\mathcal{S}_+$</span> is the set of positive semidefinite symmetric matrices, <span>$A$</span> is an affine map that outputs symmetric matrices and <span>$B$</span> is an affine map that outputs square matrices.</p><h3 id="Quadratic-constraints"><a class="docs-heading-anchor" href="#Quadratic-constraints">Quadratic constraints</a><a id="Quadratic-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Quadratic-constraints" title="Permalink"></a></h3><table><tr><th style="text-align: right">Mathematical Constraint</th><th style="text-align: right">MOI Function</th><th style="text-align: right">MOI Set</th></tr><tr><td style="text-align: right"><span>$\frac{1}{2}x^TQx + a^Tx + b \ge 0$</span></td><td style="text-align: right"><code>ScalarQuadraticFunction</code></td><td style="text-align: right"><code>GreaterThan</code></td></tr><tr><td style="text-align: right"><span>$\frac{1}{2}x^TQx + a^Tx + b \le 0$</span></td><td style="text-align: right"><code>ScalarQuadraticFunction</code></td><td style="text-align: right"><code>LessThan</code></td></tr><tr><td style="text-align: right"><span>$\frac{1}{2}x^TQx + a^Tx + b = 0$</span></td><td style="text-align: right"><code>ScalarQuadraticFunction</code></td><td style="text-align: right"><code>EqualTo</code></td></tr><tr><td style="text-align: right">Bilinear matrix inequality</td><td style="text-align: right"><code>VectorQuadraticFunction</code></td><td style="text-align: right"><code>PositiveSemidefiniteCone...</code></td></tr></table><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>For more details on the internal format of the quadratic functions see <a href="../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a> or <a href="../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p></div></div><h3 id="Discrete-and-logical-constraints"><a class="docs-heading-anchor" href="#Discrete-and-logical-constraints">Discrete and logical constraints</a><a id="Discrete-and-logical-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Discrete-and-logical-constraints" title="Permalink"></a></h3><table><tr><th style="text-align: right">Mathematical Constraint</th><th style="text-align: right">MOI Function</th><th style="text-align: right">MOI Set</th></tr><tr><td style="text-align: right"><span>$x_i \in \mathbb{Z}$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>Integer</code></td></tr><tr><td style="text-align: right"><span>$x_i \in \{0,1\}$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>ZeroOne</code></td></tr><tr><td style="text-align: right"><span>$x_i \in \{0\} \cup [l,u]$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>Semicontinuous</code></td></tr><tr><td style="text-align: right"><span>$x_i \in \{0\} \cup \{l,l+1,\ldots,u-1,u\}$</span></td><td style="text-align: right"><code>VariableIndex</code></td><td style="text-align: right"><code>Semiinteger</code></td></tr><tr><td style="text-align: right">At most one component of <span>$x$</span> can be nonzero</td><td style="text-align: right"><code>VectorOfVariables</code></td><td style="text-align: right"><code>SOS1</code></td></tr><tr><td style="text-align: right">At most two components of <span>$x$</span> can be nonzero, and if so they must be adjacent components</td><td style="text-align: right"><code>VectorOfVariables</code></td><td style="text-align: right"><code>SOS2</code></td></tr><tr><td style="text-align: right"><span>$y = 1 \implies a^T x \in S$</span></td><td style="text-align: right"><code>VectorAffineFunction</code></td><td style="text-align: right"><code>Indicator</code></td></tr></table><h2 id="JuMP-mapping"><a class="docs-heading-anchor" href="#JuMP-mapping">JuMP mapping</a><a id="JuMP-mapping-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-mapping" title="Permalink"></a></h2><p>The following bullet points show examples of how JuMP constraints are translated into MOI function-set pairs:</p><ul><li><code>@constraint(m, 2x + y &lt;= 10)</code> becomes <code>ScalarAffineFunction</code>-in-<code>LessThan</code></li><li><code>@constraint(m, 2x + y &gt;= 10)</code> becomes <code>ScalarAffineFunction</code>-in-<code>GreaterThan</code></li><li><code>@constraint(m, 2x + y == 10)</code> becomes <code>ScalarAffineFunction</code>-in-<code>EqualTo</code></li><li><code>@constraint(m, 0 &lt;= 2x + y &lt;= 10)</code> becomes <code>ScalarAffineFunction</code>-in-<code>Interval</code></li><li><code>@constraint(m, 2x + y in ArbitrarySet())</code> becomes <code>ScalarAffineFunction</code>-in-<code>ArbitrarySet</code>.</li></ul><p>Variable bounds are handled in a similar fashion:</p><ul><li><code>@variable(m, x &lt;= 1)</code> becomes <code>VariableIndex</code>-in-<code>LessThan</code></li><li><code>@variable(m, x &gt;= 1)</code> becomes <code>VariableIndex</code>-in-<code>GreaterThan</code></li></ul><p>One notable difference is that a variable with an upper and lower bound is translated into two constraints, rather than an interval, that is:</p><ul><li><code>@variable(m, 0 &lt;= x &lt;= 1)</code> becomes <code>VariableIndex</code>-in-<code>LessThan</code> <em>and</em>  <code>VariableIndex</code>-in-<code>GreaterThan</code>.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../variables/">« Variables</a><a class="docs-footer-nextpage" href="../solutions/">Solutions »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/manual/models/index.html b/previews/PR3525/moi/manual/models/index.html
new file mode 100644
index 00000000000..a570d76b4e9
--- /dev/null
+++ b/previews/PR3525/moi/manual/models/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Models · JuMP</title><meta name="title" content="Models · JuMP"/><meta property="og:title" content="Models · JuMP"/><meta property="twitter:title" content="Models · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form problem</a></li><li class="is-active"><a class="tocitem" href>Models</a><ul class="internal"><li><a class="tocitem" href="#Attributes"><span>Attributes</span></a></li><li><a class="tocitem" href="#ModelLike-API"><span>ModelLike API</span></a></li><li><a class="tocitem" href="#AbstractOptimizer-API"><span>AbstractOptimizer API</span></a></li></ul></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Models</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Models</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/models.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Models"><a class="docs-heading-anchor" href="#Models">Models</a><a id="Models-1"></a><a class="docs-heading-anchor-permalink" href="#Models" title="Permalink"></a></h1><p>The most significant part of MOI is the definition of the <strong>model API</strong> that is used to specify an instance of an optimization problem (for example, by adding variables and constraints). Objects that implement the model API must inherit from the <a href="../../reference/models/#MathOptInterface.ModelLike"><code>ModelLike</code></a> abstract type.</p><p>Notably missing from the model API is the method to solve an optimization problem. <code>ModelLike</code> objects may store an instance (for example, in memory or backed by a file format) without being linked to a particular solver. In addition to the model API, MOI defines <a href="../../reference/models/#MathOptInterface.AbstractOptimizer"><code>AbstractOptimizer</code></a> and provides methods to solve the model and interact with solutions. See the <a href="../solutions/#manual_solutions">Solutions</a> section for more details.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Throughout the rest of the manual, <code>model</code> is used as a generic <code>ModelLike</code>, and <code>optimizer</code> is used as a generic <code>AbstractOptimizer</code>.</p></div></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>MOI does not export functions, but for brevity we often omit qualifying names with the MOI module. Best practice is to have</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre><p>and prefix all MOI methods with <code>MOI.</code> in user code. If a name is also available in base Julia, we always explicitly use the module prefix, for example, with <code>MOI.get</code>.</p></div></div><h2 id="Attributes"><a class="docs-heading-anchor" href="#Attributes">Attributes</a><a id="Attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Attributes" title="Permalink"></a></h2><p>Attributes are properties of the model that can be queried and modified. These include constants such as the number of variables in a model <a href="../../reference/models/#MathOptInterface.NumberOfVariables"><code>NumberOfVariables</code></a>), and properties of variables and constraints such as the name of a variable (<a href="../../reference/variables/#MathOptInterface.VariableName"><code>VariableName</code></a>).</p><p>There are four types of attributes:</p><ul><li>Model attributes (subtypes of <a href="../../reference/models/#MathOptInterface.AbstractModelAttribute"><code>AbstractModelAttribute</code></a>) refer to properties of a model.</li><li>Optimizer attributes (subtypes of <a href="../../reference/models/#MathOptInterface.AbstractOptimizerAttribute"><code>AbstractOptimizerAttribute</code></a>) refer to properties of an optimizer.</li><li>Constraint attributes (subtypes of <a href="../../reference/constraints/#MathOptInterface.AbstractConstraintAttribute"><code>AbstractConstraintAttribute</code></a>) refer to properties of an individual constraint.</li><li>Variable attributes (subtypes of <a href="../../reference/variables/#MathOptInterface.AbstractVariableAttribute"><code>AbstractVariableAttribute</code></a>) refer to properties of an individual variable.</li></ul><p>Some attributes are values that can be queried by the user but not modified, while other attributes can be modified by the user.</p><p>All interactions with attributes occur through the <a href="../../reference/models/#MathOptInterface.get"><code>get</code></a> and <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a> functions.</p><p>Consult the docstrings of each attribute for information on what it represents.</p><h2 id="ModelLike-API"><a class="docs-heading-anchor" href="#ModelLike-API">ModelLike API</a><a id="ModelLike-API-1"></a><a class="docs-heading-anchor-permalink" href="#ModelLike-API" title="Permalink"></a></h2><p>The following attributes are available:</p><ul><li><a href="../../reference/models/#MathOptInterface.ListOfConstraintAttributesSet"><code>ListOfConstraintAttributesSet</code></a></li><li><a href="../../reference/models/#MathOptInterface.ListOfConstraintIndices"><code>ListOfConstraintIndices</code></a></li><li><a href="../../reference/models/#MathOptInterface.ListOfConstraintTypesPresent"><code>ListOfConstraintTypesPresent</code></a></li><li><a href="../../reference/models/#MathOptInterface.ListOfModelAttributesSet"><code>ListOfModelAttributesSet</code></a></li><li><a href="../../reference/models/#MathOptInterface.ListOfVariableAttributesSet"><code>ListOfVariableAttributesSet</code></a></li><li><a href="../../reference/models/#MathOptInterface.ListOfVariableIndices"><code>ListOfVariableIndices</code></a></li><li><a href="../../reference/models/#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints</code></a></li><li><a href="../../reference/models/#MathOptInterface.NumberOfVariables"><code>NumberOfVariables</code></a></li><li><a href="../../reference/models/#MathOptInterface.Name"><code>Name</code></a></li><li><a href="../../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a></li><li><a href="../../reference/models/#MathOptInterface.ObjectiveFunctionType"><code>ObjectiveFunctionType</code></a></li><li><a href="../../reference/models/#MathOptInterface.ObjectiveSense"><code>ObjectiveSense</code></a></li></ul><h2 id="AbstractOptimizer-API"><a class="docs-heading-anchor" href="#AbstractOptimizer-API">AbstractOptimizer API</a><a id="AbstractOptimizer-API-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractOptimizer-API" title="Permalink"></a></h2><p>The following attributes are available:</p><ul><li><a href="../../reference/models/#MathOptInterface.DualStatus"><code>DualStatus</code></a></li><li><a href="../../reference/models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a></li><li><a href="../../reference/models/#MathOptInterface.RawStatusString"><code>RawStatusString</code></a></li><li><a href="../../reference/models/#MathOptInterface.ResultCount"><code>ResultCount</code></a></li><li><a href="../../reference/models/#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a></li><li><a href="../../reference/models/#MathOptInterface.BarrierIterations"><code>BarrierIterations</code></a></li><li><a href="../../reference/models/#MathOptInterface.DualObjectiveValue"><code>DualObjectiveValue</code></a></li><li><a href="../../reference/models/#MathOptInterface.NodeCount"><code>NodeCount</code></a></li><li><a href="../../reference/models/#MathOptInterface.NumberOfThreads"><code>NumberOfThreads</code></a></li><li><a href="../../reference/models/#MathOptInterface.ObjectiveBound"><code>ObjectiveBound</code></a></li><li><a href="../../reference/models/#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a></li><li><a href="../../reference/models/#MathOptInterface.RelativeGap"><code>RelativeGap</code></a></li><li><a href="../../reference/models/#MathOptInterface.RawOptimizerAttribute"><code>RawOptimizerAttribute</code></a></li><li><a href="../../reference/models/#MathOptInterface.RawSolver"><code>RawSolver</code></a></li><li><a href="../../reference/models/#MathOptInterface.Silent"><code>Silent</code></a></li><li><a href="../../reference/models/#MathOptInterface.SimplexIterations"><code>SimplexIterations</code></a></li><li><a href="../../reference/models/#MathOptInterface.SolverName"><code>SolverName</code></a></li><li><a href="../../reference/models/#MathOptInterface.SolverVersion"><code>SolverVersion</code></a></li><li><a href="../../reference/models/#MathOptInterface.SolveTimeSec"><code>SolveTimeSec</code></a></li><li><a href="../../reference/models/#MathOptInterface.TimeLimitSec"><code>TimeLimitSec</code></a></li><li><a href="../../reference/models/#MathOptInterface.ObjectiveLimit"><code>ObjectiveLimit</code></a></li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../standard_form/">« Standard form problem</a><a class="docs-footer-nextpage" href="../variables/">Variables »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/manual/modification/index.html b/previews/PR3525/moi/manual/modification/index.html
new file mode 100644
index 00000000000..db43a64e0ec
--- /dev/null
+++ b/previews/PR3525/moi/manual/modification/index.html
@@ -0,0 +1,129 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Problem modification · JuMP</title><meta name="title" content="Problem modification · JuMP"/><meta property="og:title" content="Problem modification · JuMP"/><meta property="twitter:title" content="Problem modification · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li class="is-active"><a class="tocitem" href>Problem modification</a><ul class="internal"><li><a class="tocitem" href="#Modify-the-set-of-a-constraint"><span>Modify the set of a constraint</span></a></li><li><a class="tocitem" href="#Modify-the-function-of-a-constraint"><span>Modify the function of a constraint</span></a></li><li><a class="tocitem" href="#Modify-constant-term-in-a-scalar-function"><span>Modify constant term in a scalar function</span></a></li><li><a class="tocitem" href="#Modify-constant-terms-in-a-vector-function"><span>Modify constant terms in a vector function</span></a></li><li><a class="tocitem" href="#Modify-affine-coefficients-in-a-scalar-function"><span>Modify affine coefficients in a scalar function</span></a></li><li><a class="tocitem" href="#Modify-affine-coefficients-in-a-vector-function"><span>Modify affine coefficients in a vector function</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Problem modification</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Problem modification</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/modification.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Problem-modification"><a class="docs-heading-anchor" href="#Problem-modification">Problem modification</a><a id="Problem-modification-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-modification" title="Permalink"></a></h1><p>In addition to adding and deleting constraints and variables, MathOptInterface supports modifying, in-place, coefficients in the constraints and the objective function of a model.</p><p>These modifications can be grouped into two categories:</p><ul><li>modifications which replace the set of function of a constraint with a new set or function</li><li>modifications which change, in-place, a component of a function</li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Some <code>ModelLike</code> objects do not support problem modification.</p></div></div><h2 id="Modify-the-set-of-a-constraint"><a class="docs-heading-anchor" href="#Modify-the-set-of-a-constraint">Modify the set of a constraint</a><a id="Modify-the-set-of-a-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-the-set-of-a-constraint" title="Permalink"></a></h2><p>Use <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a> and <a href="../../reference/constraints/#MathOptInterface.ConstraintSet"><code>ConstraintSet</code></a> to modify the set of a constraint by replacing it with a new instance of the same type.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(
+           model,
+           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),
+           MOI.EqualTo(1.0),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)
+
+julia&gt; MOI.set(model, MOI.ConstraintSet(), c, MOI.EqualTo(2.0));
+
+julia&gt; MOI.get(model, MOI.ConstraintSet(), c) == MOI.EqualTo(2.0)
+true</code></pre><p>However, the following will fail as the new set is of a different type to the original set:</p><pre><code class="language-julia hljs">julia&gt; MOI.set(model, MOI.ConstraintSet(), c, MOI.GreaterThan(2.0))
+ERROR: [...]</code></pre><h3 id="Special-cases:-set-transforms"><a class="docs-heading-anchor" href="#Special-cases:-set-transforms">Special cases: set transforms</a><a id="Special-cases:-set-transforms-1"></a><a class="docs-heading-anchor-permalink" href="#Special-cases:-set-transforms" title="Permalink"></a></h3><p>If our constraint is an affine inequality, then this corresponds to modifying the right-hand side of a constraint in linear programming.</p><p>In some special cases, solvers may support efficiently changing the set of a constraint (for example, from <a href="../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a> to <a href="../../reference/standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a>). For these cases, MathOptInterface provides the <a href="../../reference/constraints/#MathOptInterface.transform"><code>transform</code></a> method.</p><p>The <a href="../../reference/constraints/#MathOptInterface.transform"><code>transform</code></a> function returns a new constraint index, and the old constraint index (that is, <code>c</code>) is no longer valid.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(
+           model,
+           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),
+           MOI.LessThan(1.0),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}(1)
+
+julia&gt; new_c = MOI.transform(model, c, MOI.GreaterThan(2.0));
+
+julia&gt; MOI.is_valid(model, c)
+false
+
+julia&gt; MOI.is_valid(model, new_c)
+true</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p><a href="../../reference/constraints/#MathOptInterface.transform"><code>transform</code></a> cannot be called with a set of the same type. Use <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a> instead.</p></div></div><h2 id="Modify-the-function-of-a-constraint"><a class="docs-heading-anchor" href="#Modify-the-function-of-a-constraint">Modify the function of a constraint</a><a id="Modify-the-function-of-a-constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-the-function-of-a-constraint" title="Permalink"></a></h2><p>Use <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a> and <a href="../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> to modify the function of a constraint by replacing it with a new instance of the same type.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(
+           model,
+           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),
+           MOI.EqualTo(1.0),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)
+
+julia&gt; new_f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(2.0, x)], 1.0);
+
+julia&gt; MOI.set(model, MOI.ConstraintFunction(), c, new_f);
+
+julia&gt; MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f
+true</code></pre><p>However, the following will fail as the new function is of a different type to the original function:</p><pre><code class="language-julia hljs">julia&gt; MOI.set(model, MOI.ConstraintFunction(), c, x)
+ERROR: [...]</code></pre><h2 id="Modify-constant-term-in-a-scalar-function"><a class="docs-heading-anchor" href="#Modify-constant-term-in-a-scalar-function">Modify constant term in a scalar function</a><a id="Modify-constant-term-in-a-scalar-function-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-constant-term-in-a-scalar-function" title="Permalink"></a></h2><p>Use <a href="../../reference/modification/#MathOptInterface.modify"><code>modify</code></a> and <a href="../../reference/modification/#MathOptInterface.ScalarConstantChange"><code>ScalarConstantChange</code></a> to modify the constant term in a <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> or <a href="../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(
+           model,
+           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),
+           MOI.EqualTo(1.0),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)
+
+julia&gt; MOI.modify(model, c, MOI.ScalarConstantChange(1.0));
+
+julia&gt; new_f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 1.0);
+
+julia&gt; MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f
+true</code></pre><p><a href="../../reference/modification/#MathOptInterface.ScalarConstantChange"><code>ScalarConstantChange</code></a> can also be used to modify the objective function by passing an instance of <a href="../../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.set(
+           model,
+           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),
+           new_f,
+       );
+
+julia&gt; MOI.modify(
+           model,
+           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),
+           MOI.ScalarConstantChange(-1.0)
+       );
+
+julia&gt; MOI.get(
+           model,
+           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),
+       ) ≈ MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], -1.0)
+true</code></pre><h2 id="Modify-constant-terms-in-a-vector-function"><a class="docs-heading-anchor" href="#Modify-constant-terms-in-a-vector-function">Modify constant terms in a vector function</a><a id="Modify-constant-terms-in-a-vector-function-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-constant-terms-in-a-vector-function" title="Permalink"></a></h2><p>Use <a href="../../reference/modification/#MathOptInterface.modify"><code>modify</code></a> and <a href="../../reference/modification/#MathOptInterface.VectorConstantChange"><code>VectorConstantChange</code></a> to modify the constant vector in a <a href="../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a> or <a href="../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(
+           model,
+           MOI.VectorAffineFunction([
+                   MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, x)),
+                   MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(2.0, x)),
+               ],
+               [0.0, 0.0],
+           ),
+           MOI.Nonnegatives(2),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1)
+
+julia&gt; MOI.modify(model, c, MOI.VectorConstantChange([3.0, 4.0]));
+
+julia&gt; new_f = MOI.VectorAffineFunction(
+           [
+        MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, x)),
+        MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(2.0, x)),
+           ],
+           [3.0, 4.0],
+       );
+
+julia&gt; MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f
+true</code></pre><h2 id="Modify-affine-coefficients-in-a-scalar-function"><a class="docs-heading-anchor" href="#Modify-affine-coefficients-in-a-scalar-function">Modify affine coefficients in a scalar function</a><a id="Modify-affine-coefficients-in-a-scalar-function-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-affine-coefficients-in-a-scalar-function" title="Permalink"></a></h2><p>Use <a href="../../reference/modification/#MathOptInterface.modify"><code>modify</code></a> and <a href="../../reference/modification/#MathOptInterface.ScalarCoefficientChange"><code>ScalarCoefficientChange</code></a> to modify the affine coefficient of a <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> or <a href="../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(
+           model,
+           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),
+           MOI.EqualTo(1.0),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)
+
+julia&gt; MOI.modify(model, c, MOI.ScalarCoefficientChange(x, 2.0));
+
+julia&gt; new_f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(2.0, x)], 0.0);
+
+julia&gt; MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f
+true</code></pre><p><a href="../../reference/modification/#MathOptInterface.ScalarCoefficientChange"><code>ScalarCoefficientChange</code></a> can also be used to modify the objective function by passing an instance of <a href="../../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a>.</p><h2 id="Modify-affine-coefficients-in-a-vector-function"><a class="docs-heading-anchor" href="#Modify-affine-coefficients-in-a-vector-function">Modify affine coefficients in a vector function</a><a id="Modify-affine-coefficients-in-a-vector-function-1"></a><a class="docs-heading-anchor-permalink" href="#Modify-affine-coefficients-in-a-vector-function" title="Permalink"></a></h2><p>Use <a href="../../reference/modification/#MathOptInterface.modify"><code>modify</code></a> and <a href="../../reference/modification/#MathOptInterface.MultirowChange"><code>MultirowChange</code></a> to modify a vector of affine coefficients in a <a href="../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a> or a <a href="../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; c = MOI.add_constraint(
+           model,
+           MOI.VectorAffineFunction([
+                   MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, x)),
+                   MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(2.0, x)),
+               ],
+               [0.0, 0.0],
+           ),
+           MOI.Nonnegatives(2),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1)
+
+julia&gt; MOI.modify(model, c, MOI.MultirowChange(x, [(1, 3.0), (2, 4.0)]));
+
+julia&gt; new_f = MOI.VectorAffineFunction(
+           [
+        MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(3.0, x)),
+        MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(4.0, x)),
+           ],
+           [0.0, 0.0],
+       );
+
+julia&gt; MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f
+true</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../solutions/">« Solutions</a><a class="docs-footer-nextpage" href="../../background/duality/">Duality »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/manual/solutions/index.html b/previews/PR3525/moi/manual/solutions/index.html
new file mode 100644
index 00000000000..bec4317cc00
--- /dev/null
+++ b/previews/PR3525/moi/manual/solutions/index.html
@@ -0,0 +1,39 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Solutions · JuMP</title><meta name="title" content="Solutions · JuMP"/><meta property="og:title" content="Solutions · JuMP"/><meta property="twitter:title" content="Solutions · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li class="is-active"><a class="tocitem" href>Solutions</a><ul class="internal"><li><a class="tocitem" href="#Solving-and-retrieving-the-results"><span>Solving and retrieving the results</span></a></li><li><a class="tocitem" href="#Why-did-the-solver-stop?"><span>Why did the solver stop?</span></a></li><li><a class="tocitem" href="#Primal-solutions"><span>Primal solutions</span></a></li><li><a class="tocitem" href="#Dual-solutions"><span>Dual solutions</span></a></li><li><a class="tocitem" href="#Common-status-situations"><span>Common status situations</span></a></li><li><a class="tocitem" href="#Querying-solution-attributes"><span>Querying solution attributes</span></a></li></ul></li><li><a class="tocitem" href="../modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Solutions</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Solutions</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/solutions.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="manual_solutions"><a class="docs-heading-anchor" href="#manual_solutions">Solutions</a><a id="manual_solutions-1"></a><a class="docs-heading-anchor-permalink" href="#manual_solutions" title="Permalink"></a></h1><h2 id="Solving-and-retrieving-the-results"><a class="docs-heading-anchor" href="#Solving-and-retrieving-the-results">Solving and retrieving the results</a><a id="Solving-and-retrieving-the-results-1"></a><a class="docs-heading-anchor-permalink" href="#Solving-and-retrieving-the-results" title="Permalink"></a></h2><p>Once an optimizer is loaded with the objective function and all of the constraints, we can ask the solver to solve the model by calling <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a>.</p><pre><code class="language-julia hljs">MOI.optimize!(optimizer)</code></pre><h2 id="Why-did-the-solver-stop?"><a class="docs-heading-anchor" href="#Why-did-the-solver-stop?">Why did the solver stop?</a><a id="Why-did-the-solver-stop?-1"></a><a class="docs-heading-anchor-permalink" href="#Why-did-the-solver-stop?" title="Permalink"></a></h2><p>The optimization procedure may stop for a number of reasons. The <a href="../../reference/models/#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a> attribute of the optimizer returns a <a href="../../reference/models/#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> object which explains why the solver stopped.</p><p>The termination statuses distinguish between proofs of optimality, infeasibility, local convergence, limits, and termination because of something unexpected like invalid problem data or failure to converge.</p><p>A typical usage of the <a href="../../reference/models/#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a> attribute is as follows:</p><pre><code class="language-julia hljs">status = MOI.get(optimizer, TerminationStatus())
+if status == MOI.OPTIMAL
+    # Ok, we solved the problem!
+else
+    # Handle other cases.
+end</code></pre><p>After checking the <a href="../../reference/models/#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a>, check <a href="../../reference/models/#MathOptInterface.ResultCount"><code>ResultCount</code></a>. This attribute returns the number of results that the solver has available to return. <em>A result is defined as a primal-dual pair, but either the primal or the dual may be missing from the result.</em> While the <code>OPTIMAL</code> termination status normally implies that at least one result is available, other statuses do not. For example, in the case of infeasibility, a solver may return no result or a proof of infeasibility. The <a href="../../reference/models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute distinguishes between these two cases.</p><h2 id="Primal-solutions"><a class="docs-heading-anchor" href="#Primal-solutions">Primal solutions</a><a id="Primal-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Primal-solutions" title="Permalink"></a></h2><p>Use the <a href="../../reference/models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> optimizer attribute to return a <a href="../../reference/models/#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> describing the status of the primal solution.</p><p>Common returns are described below in the <a href="#Common-status-situations">Common status situations</a> section.</p><p>Query the primal solution using the <a href="../../reference/variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a> and <a href="../../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>ConstraintPrimal</code></a> attributes.</p><p>Query the objective function value using the <a href="../../reference/models/#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a> attribute.</p><h2 id="Dual-solutions"><a class="docs-heading-anchor" href="#Dual-solutions">Dual solutions</a><a id="Dual-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Dual-solutions" title="Permalink"></a></h2><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>See <a href="../../background/duality/#Duality">Duality</a> for a discussion of the MOI conventions for primal-dual pairs and certificates.</p></div></div><p>Use the <a href="../../reference/models/#MathOptInterface.DualStatus"><code>DualStatus</code></a> optimizer attribute to return a <a href="../../reference/models/#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> describing the status of the dual solution.</p><p>Query the dual solution using the <a href="../../reference/constraints/#MathOptInterface.ConstraintDual"><code>ConstraintDual</code></a> attribute.</p><p>Query the dual objective function value using the <a href="../../reference/models/#MathOptInterface.DualObjectiveValue"><code>DualObjectiveValue</code></a> attribute.</p><h2 id="Common-status-situations"><a class="docs-heading-anchor" href="#Common-status-situations">Common status situations</a><a id="Common-status-situations-1"></a><a class="docs-heading-anchor-permalink" href="#Common-status-situations" title="Permalink"></a></h2><p>The sections below describe how to interpret typical or interesting status cases for three common classes of solvers. The example cases are illustrative, not comprehensive. Solver wrappers may provide additional information on how the solver&#39;s statuses map to MOI statuses.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p><code>*</code> in the tables indicate that multiple different values are possible.</p></div></div><h3 id="Primal-dual-convex-solver"><a class="docs-heading-anchor" href="#Primal-dual-convex-solver">Primal-dual convex solver</a><a id="Primal-dual-convex-solver-1"></a><a class="docs-heading-anchor-permalink" href="#Primal-dual-convex-solver" title="Permalink"></a></h3><p>Linear programming and conic optimization solvers fall into this category.</p><table><tr><th style="text-align: right">What happened?</th><th style="text-align: right"><code>TerminationStatus</code></th><th style="text-align: right"><code>ResultCount</code></th><th style="text-align: right"><code>PrimalStatus</code></th><th style="text-align: right"><code>DualStatus</code></th></tr><tr><td style="text-align: right">Proved optimality</td><td style="text-align: right"><code>OPTIMAL</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>FEASIBLE_POINT</code></td><td style="text-align: right"><code>FEASIBLE_POINT</code></td></tr><tr><td style="text-align: right">Proved infeasible</td><td style="text-align: right"><code>INFEASIBLE</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>NO_SOLUTION</code></td><td style="text-align: right"><code>INFEASIBILITY_CERTIFICATE</code></td></tr><tr><td style="text-align: right">Optimal within relaxed tolerances</td><td style="text-align: right"><code>ALMOST_OPTIMAL</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>FEASIBLE_POINT</code></td><td style="text-align: right"><code>FEASIBLE_POINT</code></td></tr><tr><td style="text-align: right">Optimal within relaxed tolerances</td><td style="text-align: right"><code>ALMOST_OPTIMAL</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>ALMOST_FEASIBLE_POINT</code></td><td style="text-align: right"><code>ALMOST_FEASIBLE_POINT</code></td></tr><tr><td style="text-align: right">Detected an unbounded ray of the primal</td><td style="text-align: right"><code>DUAL_INFEASIBLE</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>INFEASIBILITY_CERTIFICATE</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr><tr><td style="text-align: right">Stall</td><td style="text-align: right"><code>SLOW_PROGRESS</code></td><td style="text-align: right">1</td><td style="text-align: right">*</td><td style="text-align: right">*</td></tr></table><h3 id="Global-branch-and-bound-solvers"><a class="docs-heading-anchor" href="#Global-branch-and-bound-solvers">Global branch-and-bound solvers</a><a id="Global-branch-and-bound-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Global-branch-and-bound-solvers" title="Permalink"></a></h3><p>Mixed-integer programming solvers fall into this category.</p><table><tr><th style="text-align: right">What happened?</th><th style="text-align: right"><code>TerminationStatus</code></th><th style="text-align: right"><code>ResultCount</code></th><th style="text-align: right"><code>PrimalStatus</code></th><th style="text-align: right"><code>DualStatus</code></th></tr><tr><td style="text-align: right">Proved optimality</td><td style="text-align: right"><code>OPTIMAL</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>FEASIBLE_POINT</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr><tr><td style="text-align: right">Presolve detected infeasibility or unboundedness</td><td style="text-align: right"><code>INFEASIBLE_OR_UNBOUNDED</code></td><td style="text-align: right">0</td><td style="text-align: right"><code>NO_SOLUTION</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr><tr><td style="text-align: right">Proved infeasibility</td><td style="text-align: right"><code>INFEASIBLE</code></td><td style="text-align: right">0</td><td style="text-align: right"><code>NO_SOLUTION</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr><tr><td style="text-align: right">Timed out (no solution)</td><td style="text-align: right"><code>TIME_LIMIT</code></td><td style="text-align: right">0</td><td style="text-align: right"><code>NO_SOLUTION</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr><tr><td style="text-align: right">Timed out (with a solution)</td><td style="text-align: right"><code>TIME_LIMIT</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>FEASIBLE_POINT</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr><tr><td style="text-align: right"><code>CPXMIP_OPTIMAL_INFEAS</code></td><td style="text-align: right"><code>ALMOST_OPTIMAL</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>INFEASIBLE_POINT</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr></table><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p><a href="https://www.ibm.com/docs/en/SSSA5P_12.6.1/ilog.odms.cplex.help/refcallablelibrary/macros/CPXMIP_OPTIMAL_INFEAS.html"><code>CPXMIP_OPTIMAL_INFEAS</code></a> is a CPLEX status that indicates that a preprocessed problem was solved to optimality, but the solver was unable to recover a feasible solution to the original problem. Handling this status was one of the motivating drivers behind the design of MOI.</p></div></div><h3 id="Local-search-solvers"><a class="docs-heading-anchor" href="#Local-search-solvers">Local search solvers</a><a id="Local-search-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Local-search-solvers" title="Permalink"></a></h3><p>Nonlinear programming solvers fall into this category. It also includes non-global tree search solvers like <a href="https://github.com/lanl-ansi/Juniper.jl">Juniper</a>.</p><table><tr><th style="text-align: right">What happened?</th><th style="text-align: right"><code>TerminationStatus</code></th><th style="text-align: right"><code>ResultCount</code></th><th style="text-align: right"><code>PrimalStatus</code></th><th style="text-align: right"><code>DualStatus</code></th></tr><tr><td style="text-align: right">Converged to a stationary point</td><td style="text-align: right"><code>LOCALLY_SOLVED</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>FEASIBLE_POINT</code></td><td style="text-align: right"><code>FEASIBLE_POINT</code></td></tr><tr><td style="text-align: right">Completed a non-global tree search (with a solution)</td><td style="text-align: right"><code>LOCALLY_SOLVED</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>FEASIBLE_POINT</code></td><td style="text-align: right"><code>FEASIBLE_POINT</code></td></tr><tr><td style="text-align: right">Converged to an infeasible point</td><td style="text-align: right"><code>LOCALLY_INFEASIBLE</code></td><td style="text-align: right">1</td><td style="text-align: right"><code>INFEASIBLE_POINT</code></td><td style="text-align: right">*</td></tr><tr><td style="text-align: right">Completed a non-global tree search (no solution found)</td><td style="text-align: right"><code>LOCALLY_INFEASIBLE</code></td><td style="text-align: right">0</td><td style="text-align: right"><code>NO_SOLUTION</code></td><td style="text-align: right"><code>NO_SOLUTION</code></td></tr><tr><td style="text-align: right">Iteration limit</td><td style="text-align: right"><code>ITERATION_LIMIT</code></td><td style="text-align: right">1</td><td style="text-align: right">*</td><td style="text-align: right">*</td></tr><tr><td style="text-align: right">Diverging iterates</td><td style="text-align: right"><code>NORM_LIMIT</code> or <code>OBJECTIVE_LIMIT</code></td><td style="text-align: right">1</td><td style="text-align: right">*</td><td style="text-align: right">*</td></tr></table><h2 id="Querying-solution-attributes"><a class="docs-heading-anchor" href="#Querying-solution-attributes">Querying solution attributes</a><a id="Querying-solution-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Querying-solution-attributes" title="Permalink"></a></h2><p>Some solvers will not implement every solution attribute. Therefore, a call like <code>MOI.get(model, MOI.SolveTimeSec())</code> may throw an <a href="../../reference/errors/#MathOptInterface.UnsupportedAttribute"><code>UnsupportedAttribute</code></a> error.</p><p>If you need to write code that is agnostic to the solver (for example, you are writing a library that an end-user passes their choice of solver to), you can work-around this problem using a <code>try-catch</code>:</p><pre><code class="language-julia hljs">function get_solve_time(model)
+    try
+        return MOI.get(model, MOI.SolveTimeSec())
+    catch err
+        if err isa MOI.UnsupportedAttribute
+            return NaN  # Solver doesn&#39;t support. Return a placeholder value.
+        end
+        rethrow(err)  # Something else went wrong. Rethrow the error
+    end
+end</code></pre><p>If, <em>after careful profiling</em>, you find that the <code>try-catch</code> is taking a significant portion of your runtime, you can improve performance by caching the result of the <code>try-catch</code>:</p><pre><code class="language-julia hljs">mutable struct CachedSolveTime{M}
+    model::M
+    supports_solve_time::Bool
+    CachedSolveTime(model::M) where {M} = new(model, true)
+end
+
+function get_solve_time(model::CachedSolveTime)
+    if !model.supports_solve_time
+        return NaN
+    end
+    try
+        return MOI.get(model, MOI.SolveTimeSec())
+    catch err
+        if err isa MOI.UnsupportedAttribute
+            model.supports_solve_time = false
+            return NaN
+        end
+        rethrow(err)  # Something else went wrong. Rethrow the error
+    end
+end</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../constraints/">« Constraints</a><a class="docs-footer-nextpage" href="../modification/">Problem modification »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/manual/standard_form/index.html b/previews/PR3525/moi/manual/standard_form/index.html
new file mode 100644
index 00000000000..c2f416e0298
--- /dev/null
+++ b/previews/PR3525/moi/manual/standard_form/index.html
@@ -0,0 +1,10 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Standard form problem · JuMP</title><meta name="title" content="Standard form problem · JuMP"/><meta property="og:title" content="Standard form problem · JuMP"/><meta property="twitter:title" content="Standard form problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Standard form problem</a><ul class="internal"><li><a class="tocitem" href="#Functions"><span>Functions</span></a></li><li><a class="tocitem" href="#One-dimensional-sets"><span>One-dimensional sets</span></a></li><li><a class="tocitem" href="#Vector-cones"><span>Vector cones</span></a></li><li><a class="tocitem" href="#Matrix-cones"><span>Matrix cones</span></a></li><li><a class="tocitem" href="#Multi-dimensional-sets-with-combinatorial-structure"><span>Multi-dimensional sets with combinatorial structure</span></a></li></ul></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Standard form problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Standard form problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/standard_form.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Standard-form-problem"><a class="docs-heading-anchor" href="#Standard-form-problem">Standard form problem</a><a id="Standard-form-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Standard-form-problem" title="Permalink"></a></h1><p>MathOptInterface represents optimization problems in the standard form:</p><p class="math-container">\[\begin{align}
+    &amp; \min_{x \in \mathbb{R}^n} &amp; f_0(x)
+    \\
+    &amp; \;\;\text{s.t.} &amp; f_i(x) &amp; \in \mathcal{S}_i &amp; i = 1 \ldots m
+\end{align}\]</p><p>where:</p><ul><li>the functions <span>$f_0, f_1, \ldots, f_m$</span> are specified by <a href="../../reference/standard_form/#MathOptInterface.AbstractFunction"><code>AbstractFunction</code></a> objects</li><li>the sets <span>$\mathcal{S}_1, \ldots, \mathcal{S}_m$</span> are specified by <a href="../../reference/standard_form/#MathOptInterface.AbstractSet"><code>AbstractSet</code></a> objects</li></ul><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>For more information on this standard form, read <a href="https://arxiv.org/pdf/2002.03447.pdf">our paper</a>.</p></div></div><p>MOI defines some commonly used functions and sets, but the interface is extensible to other sets recognized by the solver.</p><h2 id="Functions"><a class="docs-heading-anchor" href="#Functions">Functions</a><a id="Functions-1"></a><a class="docs-heading-anchor-permalink" href="#Functions" title="Permalink"></a></h2><p>The function types implemented in MathOptInterface.jl are:</p><table><tr><th style="text-align: left">Function</th><th style="text-align: left">Description</th></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a></td><td style="text-align: left"><span>$x_j$</span>, the projection onto a single coordinate defined by a variable index <span>$j$</span>.</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>VectorOfVariables</code></a></td><td style="text-align: left">The projection onto multiple coordinates (that is, extracting a sub-vector).</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a></td><td style="text-align: left"><span>$a^T x + b$</span>, where <span>$a$</span> is a vector and <span>$b$</span> scalar.</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a></td><td style="text-align: left"><span>$f(x)$</span>, where <span>$f$</span> is a nonlinear function.</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a></td><td style="text-align: left"><span>$A x + b$</span>, where <span>$A$</span> is a matrix and <span>$b$</span> is a vector.</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a></td><td style="text-align: left"><span>$\frac{1}{2} x^T Q x + a^T x + b$</span>, where <span>$Q$</span> is a symmetric matrix, <span>$a$</span> is a vector, and <span>$b$</span> is a constant.</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a></td><td style="text-align: left">A vector of scalar-valued quadratic functions.</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.VectorNonlinearFunction"><code>VectorNonlinearFunction</code></a></td><td style="text-align: left"><span>$f(x)$</span>, where <span>$f$</span> is a vector-valued nonlinear function.</td></tr></table><p>Extensions for nonlinear programming are present but not yet well documented.</p><h2 id="One-dimensional-sets"><a class="docs-heading-anchor" href="#One-dimensional-sets">One-dimensional sets</a><a id="One-dimensional-sets-1"></a><a class="docs-heading-anchor-permalink" href="#One-dimensional-sets" title="Permalink"></a></h2><p>The one-dimensional set types implemented in MathOptInterface.jl are:</p><table><tr><th style="text-align: left">Set</th><th style="text-align: left">Description</th></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan(u)</code></a></td><td style="text-align: left"><span>$(-\infty, u]$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan(l)</code></a></td><td style="text-align: left"><span>$[l, \infty)$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.GreaterThan"><code>EqualTo(v)</code></a></td><td style="text-align: left"><span>$\{v\}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Interval"><code>Interval(l, u)</code></a></td><td style="text-align: left"><span>$[l, u]$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Integer"><code>Integer()</code></a></td><td style="text-align: left"><span>$\mathbb{Z}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.ZeroOne"><code>ZeroOne()</code></a></td><td style="text-align: left"><span>$\{ 0, 1 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Semicontinuous"><code>Semicontinuous(l, u)</code></a></td><td style="text-align: left"><span>$\{ 0\} \cup [l, u]$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Semiinteger"><code>Semiinteger(l, u)</code></a></td><td style="text-align: left"><span>$\{ 0\} \cup \{l,l+1,\ldots,u-1,u\}$</span></td></tr></table><h2 id="Vector-cones"><a class="docs-heading-anchor" href="#Vector-cones">Vector cones</a><a id="Vector-cones-1"></a><a class="docs-heading-anchor-permalink" href="#Vector-cones" title="Permalink"></a></h2><p>The vector-valued set types implemented in MathOptInterface.jl are:</p><table><tr><th style="text-align: left">Set</th><th style="text-align: left">Description</th></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Reals"><code>Reals(d)</code></a></td><td style="text-align: left"><span>$\mathbb{R}^{d}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Zeros"><code>Zeros(d)</code></a></td><td style="text-align: left"><span>$0^{d}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>Nonnegatives(d)</code></a></td><td style="text-align: left"><span>$\{ x \in \mathbb{R}^{d} : x \ge 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Nonpositives"><code>Nonpositives(d)</code></a></td><td style="text-align: left"><span>$\{ x \in \mathbb{R}^{d} : x \le 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>SecondOrderCone(d)</code></a></td><td style="text-align: left"><span>$\{ (t,x) \in \mathbb{R}^{d} : t \ge \lVert x \rVert_2 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>RotatedSecondOrderCone(d)</code></a></td><td style="text-align: left"><span>$\{ (t,u,x) \in \mathbb{R}^{d} : 2tu \ge \lVert x \rVert_2^2, t \ge 0,u \ge 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.ExponentialCone"><code>ExponentialCone()</code></a></td><td style="text-align: left"><span>$\{ (x,y,z) \in \mathbb{R}^3 : y \exp (x/y) \le z, y &gt; 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.DualExponentialCone"><code>DualExponentialCone()</code></a></td><td style="text-align: left"><span>$\{ (u,v,w) \in \mathbb{R}^3 : -u \exp (v/u) \le \exp(1) w, u &lt; 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>GeometricMeanCone(d)</code></a></td><td style="text-align: left"><span>$\{ (t,x) \in \mathbb{R}^{1+n} : x \ge 0, t \le \sqrt[n]{x_1 x_2 \cdots x_n} \}$</span> where <span>$n$</span> is <span>$d - 1$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.PowerCone"><code>PowerCone(α)</code></a></td><td style="text-align: left"><span>$\{ (x,y,z) \in \mathbb{R}^3 : x^{\alpha} y^{1-\alpha} \ge |z|, x \ge 0,y \ge 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.DualPowerCone"><code>DualPowerCone(α)</code></a></td><td style="text-align: left"><span>$\{ (u,v,w) \in \mathbb{R}^3 : \left(\frac{u}{\alpha}\right(^{\alpha}\left(\frac{v}{1-\alpha}\right)^{1-\alpha} \ge |w|, u,v \ge 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.NormOneCone"><code>NormOneCone(d)</code></a></td><td style="text-align: left"><span>$\{ (t,x) \in \mathbb{R}^{d} : t \ge \sum_i \lvert x_i \rvert \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.NormInfinityCone"><code>NormInfinityCone(d)</code></a></td><td style="text-align: left"><span>$\{ (t,x) \in \mathbb{R}^{d} : t \ge \max_i \lvert x_i \rvert \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.RelativeEntropyCone"><code>RelativeEntropyCone(d)</code></a></td><td style="text-align: left"><span>$\{ (u, v, w) \in \mathbb{R}^{d} : u \ge \sum_i w_i \log (\frac{w_i}{v_i}), v_i \ge 0, w_i \ge 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.HyperRectangle"><code>HyperRectangle(l, u)</code></a></td><td style="text-align: left"><span>$\{x \in \bar{\mathbb{R}}^d: x_i \in [l_i, u_i] \forall i=1,\ldots,d\}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.NormCone"><code>NormCone(p, d)</code></a></td><td style="text-align: left">``{ (t,x) \in \mathbb{R}^{d} : t \ge \left(\sum\limits_i</td></tr></table><h2 id="Matrix-cones"><a class="docs-heading-anchor" href="#Matrix-cones">Matrix cones</a><a id="Matrix-cones-1"></a><a class="docs-heading-anchor-permalink" href="#Matrix-cones" title="Permalink"></a></h2><p>The matrix-valued set types implemented in MathOptInterface.jl are:</p><table><tr><th style="text-align: left">Set</th><th style="text-align: left">Description</th></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>RootDetConeTriangle(d)</code></a></td><td style="text-align: left"><span>$\{ (t,X) \in \mathbb{R}^{1+d(1+d)/2} : t \le \det(X)^{1/d}, X \mbox{ is the upper triangle of a PSD matrix} \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.RootDetConeSquare"><code>RootDetConeSquare(d)</code></a></td><td style="text-align: left"><span>$\{ (t,X) \in \mathbb{R}^{1+d^2} : t \le \det(X)^{1/d}, X \mbox{ is a PSD matrix} \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle(d)</code></a></td><td style="text-align: left"><span>$\{ X \in \mathbb{R}^{d(d+1)/2} : X \mbox{ is the upper triangle of a PSD matrix} \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare(d)</code></a></td><td style="text-align: left"><span>$\{ X \in \mathbb{R}^{d^2} : X \mbox{ is a PSD matrix} \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.LogDetConeTriangle"><code>LogDetConeTriangle(d)</code></a></td><td style="text-align: left"><span>$\{ (t,u,X) \in \mathbb{R}^{2+d(1+d)/2} : t \le u\log(\det(X/u)), X \mbox{ is the upper triangle of a PSD matrix}, u &gt; 0  \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.LogDetConeSquare"><code>LogDetConeSquare(d)</code></a></td><td style="text-align: left"><span>$\{ (t,u,X) \in \mathbb{R}^{2+d^2} : t \le u \log(\det(X/u)), X \mbox{ is a PSD matrix}, u &gt; 0 \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.NormSpectralCone"><code>NormSpectralCone(r, c)</code></a></td><td style="text-align: left"><span>$\{ (t, X) \in \mathbb{R}^{1 + r \times c} : t \ge \sigma_1(X), X \mbox{ is a } r\times c\mbox{ matrix} \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.NormNuclearCone"><code>NormNuclearCone(r, c)</code></a></td><td style="text-align: left"><span>$\{ (t, X) \in \mathbb{R}^{1 + r \times c} : t \ge \sum_i \sigma_i(X), X \mbox{ is a } r\times c\mbox{ matrix} \}$</span></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>HermitianPositiveSemidefiniteConeTriangle(d)</code></a></td><td style="text-align: left">The cone of Hermitian positive semidefinite matrices, with</td></tr><tr><td style="text-align: left"><code>side_dimension</code> rows and columns.</td><td style="text-align: left"></td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Scaled"><code>Scaled(S)</code></a></td><td style="text-align: left">The set <code>S</code> scaled so that <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.set_dot"><code>Utilities.set_dot</code></a> corresponds to <code>LinearAlgebra.dot</code></td></tr></table><p>Some of these cones can take two forms: <code>XXXConeTriangle</code> and <code>XXXConeSquare</code>.</p><p>In <code>XXXConeTriangle</code> sets, the matrix is assumed to be symmetric, and the elements are provided by a vector, in which the entries of the upper-right triangular part of the matrix are given column by column (or equivalently, the entries of the lower-left triangular part are given row by row).</p><p>In <code>XXXConeSquare</code> sets, the entries of the matrix are given column by column (or equivalently, row by row), and the matrix is constrained to be symmetric. As an example, given a 2-by-2 matrix of variables <code>X</code> and a one-dimensional variable <code>t</code>, we can specify a root-det constraint as <code>[t, X11, X12, X22] ∈ RootDetConeTriangle</code> or <code>[t, X11, X12, X21, X22] ∈ RootDetConeSquare</code>.</p><p>We provide both forms to enable flexibility for solvers who may natively support one or the other. Transformations between <code>XXXConeTriangle</code> and <code>XXXConeSquare</code> are handled by bridges, which removes the chance of conversion mistakes by users or solver developers.</p><h2 id="Multi-dimensional-sets-with-combinatorial-structure"><a class="docs-heading-anchor" href="#Multi-dimensional-sets-with-combinatorial-structure">Multi-dimensional sets with combinatorial structure</a><a id="Multi-dimensional-sets-with-combinatorial-structure-1"></a><a class="docs-heading-anchor-permalink" href="#Multi-dimensional-sets-with-combinatorial-structure" title="Permalink"></a></h2><p>Other sets are vector-valued, with a particular combinatorial structure. Read their docstrings for more information on how to interpret them.</p><table><tr><th style="text-align: left">Set</th><th style="text-align: left">Description</th></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.SOS1"><code>SOS1</code></a></td><td style="text-align: left">A Special Ordered Set (SOS) of Type I</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.SOS2"><code>SOS2</code></a></td><td style="text-align: left">A Special Ordered Set (SOS) of Type II</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Indicator"><code>Indicator</code></a></td><td style="text-align: left">A set to specify an indicator constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Complements"><code>Complements</code></a></td><td style="text-align: left">A set to specify a mixed complementarity constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.AllDifferent"><code>AllDifferent</code></a></td><td style="text-align: left">The <code>all_different</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.BinPacking"><code>BinPacking</code></a></td><td style="text-align: left">The <code>bin_packing</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Circuit"><code>Circuit</code></a></td><td style="text-align: left">The <code>circuit</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.CountAtLeast"><code>CountAtLeast</code></a></td><td style="text-align: left">The <code>at_least</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.CountBelongs"><code>CountBelongs</code></a></td><td style="text-align: left">The <code>nvalue</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.CountDistinct"><code>CountDistinct</code></a></td><td style="text-align: left">The <code>distinct</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.CountGreaterThan"><code>CountGreaterThan</code></a></td><td style="text-align: left">The <code>count_gt</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Cumulative"><code>Cumulative</code></a></td><td style="text-align: left">The <code>cumulative</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Path"><code>Path</code></a></td><td style="text-align: left">The <code>path</code> global constraint</td></tr><tr><td style="text-align: left"><a href="../../reference/standard_form/#MathOptInterface.Table"><code>Table</code></a></td><td style="text-align: left">The <code>table</code> global constraint</td></tr></table></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../tutorials/latency/">« Latency</a><a class="docs-footer-nextpage" href="../models/">Models »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/manual/variables/index.html b/previews/PR3525/moi/manual/variables/index.html
new file mode 100644
index 00000000000..0778729fd5d
--- /dev/null
+++ b/previews/PR3525/moi/manual/variables/index.html
@@ -0,0 +1,17 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Variables · JuMP</title><meta name="title" content="Variables · JuMP"/><meta property="og:title" content="Variables · JuMP"/><meta property="twitter:title" content="Variables · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../models/">Models</a></li><li class="is-active"><a class="tocitem" href>Variables</a><ul class="internal"><li><a class="tocitem" href="#Add-a-variable"><span>Add a variable</span></a></li><li><a class="tocitem" href="#Delete-a-variable"><span>Delete a variable</span></a></li><li><a class="tocitem" href="#Variable-attributes"><span>Variable attributes</span></a></li></ul></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../solutions/">Solutions</a></li><li><a class="tocitem" href="../modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Manual</a></li><li class="is-active"><a href>Variables</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Variables</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/variables.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Variables"><a class="docs-heading-anchor" href="#Variables">Variables</a><a id="Variables-1"></a><a class="docs-heading-anchor-permalink" href="#Variables" title="Permalink"></a></h1><h2 id="Add-a-variable"><a class="docs-heading-anchor" href="#Add-a-variable">Add a variable</a><a id="Add-a-variable-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-variable" title="Permalink"></a></h2><p>Use <a href="../../reference/variables/#MathOptInterface.add_variable"><code>add_variable</code></a> to add a single variable.</p><pre><code class="language-julia-repl hljs">julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)</code></pre><p><a href="../../reference/variables/#MathOptInterface.add_variable"><code>add_variable</code></a> returns a <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> type, which is used to refer to the added variable in other calls.</p><p>Check if a <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> is valid using <a href="../../reference/constraints/#MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.ConstraintIndex}"><code>is_valid</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.is_valid(model, x)
+true</code></pre><p>Use <a href="../../reference/variables/#MathOptInterface.add_variables"><code>add_variables</code></a> to add a number of variables.</p><pre><code class="language-julia-repl hljs">julia&gt; y = MOI.add_variables(model, 2)
+2-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The integer does not necessarily correspond to the column inside an optimizer.</p></div></div><h2 id="Delete-a-variable"><a class="docs-heading-anchor" href="#Delete-a-variable">Delete a variable</a><a id="Delete-a-variable-1"></a><a class="docs-heading-anchor-permalink" href="#Delete-a-variable" title="Permalink"></a></h2><p>Delete a variable using <a href="../../reference/variables/#MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}"><code>delete</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.delete(model, x)
+
+julia&gt; MOI.is_valid(model, x)
+false</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Not all <code>ModelLike</code> models support deleting variables. A <a href="../../reference/errors/#MathOptInterface.DeleteNotAllowed"><code>DeleteNotAllowed</code></a> error is thrown if this is not supported.</p></div></div><h2 id="Variable-attributes"><a class="docs-heading-anchor" href="#Variable-attributes">Variable attributes</a><a id="Variable-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Variable-attributes" title="Permalink"></a></h2><p>The following attributes are available for variables:</p><ul><li><a href="../../reference/variables/#MathOptInterface.VariableName"><code>VariableName</code></a></li><li><a href="../../reference/variables/#MathOptInterface.VariablePrimalStart"><code>VariablePrimalStart</code></a></li><li><a href="../../reference/variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a></li></ul><p>Get and set these attributes using <a href="../../reference/models/#MathOptInterface.get"><code>get</code></a> and <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.set(model, MOI.VariableName(), x, &quot;var_x&quot;)
+
+julia&gt; MOI.get(model, MOI.VariableName(), x)
+&quot;var_x&quot;</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../models/">« Models</a><a class="docs-footer-nextpage" href="../constraints/">Constraints »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/callbacks/index.html b/previews/PR3525/moi/reference/callbacks/index.html
new file mode 100644
index 00000000000..a4a933ac47f
--- /dev/null
+++ b/previews/PR3525/moi/reference/callbacks/index.html
@@ -0,0 +1,36 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Callbacks · JuMP</title><meta name="title" content="Callbacks · JuMP"/><meta property="og:title" content="Callbacks · JuMP"/><meta property="twitter:title" content="Callbacks · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../modification/">Modifications</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear programming</a></li><li class="is-active"><a class="tocitem" href>Callbacks</a><ul class="internal"><li><a class="tocitem" href="#Attributes"><span>Attributes</span></a></li><li><a class="tocitem" href="#Lazy-constraints"><span>Lazy constraints</span></a></li><li><a class="tocitem" href="#User-cuts"><span>User cuts</span></a></li><li><a class="tocitem" href="#Heuristic-solutions"><span>Heuristic solutions</span></a></li></ul></li><li><a class="tocitem" href="../errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Callbacks</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Callbacks</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/callbacks.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Callbacks"><a class="docs-heading-anchor" href="#Callbacks">Callbacks</a><a id="Callbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Callbacks" title="Permalink"></a></h1><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractCallback" href="#MathOptInterface.AbstractCallback"><code>MathOptInterface.AbstractCallback</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractCallback &lt;: AbstractModelAttribute end</code></pre><p>Abstract type for a model attribute representing a callback function. The value set to subtypes of <code>AbstractCallback</code> is a function that may be called during <a href="../models/#MathOptInterface.optimize!"><code>optimize!</code></a>. As <a href="../models/#MathOptInterface.optimize!"><code>optimize!</code></a> is in progress, the result attributes (i.e, the attributes <code>attr</code> such that <code>is_set_by_optimize(attr)</code>) may not be accessible from the callback, hence trying to get result attributes might throw a <a href="../errors/#MathOptInterface.OptimizeInProgress"><code>OptimizeInProgress</code></a> error.</p><p>At most one callback of each type can be registered. If an optimizer already has a function for a callback type, and the user registers a new function, then the old one is replaced.</p><p>The value of the attribute should be a function taking only one argument, commonly called <code>callback_data</code>, that can be used for instance in <a href="#MathOptInterface.LazyConstraintCallback"><code>LazyConstraintCallback</code></a>, <a href="#MathOptInterface.HeuristicCallback"><code>HeuristicCallback</code></a> and <a href="#MathOptInterface.UserCutCallback"><code>UserCutCallback</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L967-L985">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractSubmittable" href="#MathOptInterface.AbstractSubmittable"><code>MathOptInterface.AbstractSubmittable</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractSubmittable</code></pre><p>Abstract supertype for objects that can be submitted to the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L132-L136">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.submit" href="#MathOptInterface.submit"><code>MathOptInterface.submit</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">submit(
+    optimizer::AbstractOptimizer,
+    sub::AbstractSubmittable,
+    values...,
+)::Nothing</code></pre><p>Submit <code>values</code> to the submittable <code>sub</code> of the optimizer <code>optimizer</code>.</p><p>An <a href="../errors/#MathOptInterface.UnsupportedSubmittable"><code>UnsupportedSubmittable</code></a> error is thrown if <code>model</code> does not support the attribute <code>attr</code> (see <a href="../models/#MathOptInterface.supports"><code>supports</code></a>) and a <a href="../errors/#MathOptInterface.SubmitNotAllowed"><code>SubmitNotAllowed</code></a> error is thrown if it supports the submittable <code>sub</code> but it cannot be submitted.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L612-L624">source</a></section></article><h2 id="Attributes"><a class="docs-heading-anchor" href="#Attributes">Attributes</a><a id="Attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CallbackNodeStatus" href="#MathOptInterface.CallbackNodeStatus"><code>MathOptInterface.CallbackNodeStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CallbackNodeStatus(callback_data)</code></pre><p>An optimizer attribute describing the (in)feasibility of the primal solution available from <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> during a callback identified by <code>callback_data</code>.</p><p>Returns a <a href="#MathOptInterface.CallbackNodeStatusCode"><code>CallbackNodeStatusCode</code></a> Enum.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L766-L774">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CallbackVariablePrimal" href="#MathOptInterface.CallbackVariablePrimal"><code>MathOptInterface.CallbackVariablePrimal</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CallbackVariablePrimal(callback_data)</code></pre><p>A variable attribute for the assignment to some primal variable&#39;s value during the callback identified by <code>callback_data</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1471-L1476">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CallbackNodeStatusCode" href="#MathOptInterface.CallbackNodeStatusCode"><code>MathOptInterface.CallbackNodeStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CallbackNodeStatusCode</code></pre><p>An Enum of possible return values from calling <a href="../models/#MathOptInterface.get"><code>get</code></a> with <a href="#MathOptInterface.CallbackNodeStatus"><code>CallbackNodeStatus</code></a>.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="#MathOptInterface.CALLBACK_NODE_STATUS_INTEGER"><code>CALLBACK_NODE_STATUS_INTEGER</code></a>: the primal solution available from <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> is integer feasible.</li><li><a href="#MathOptInterface.CALLBACK_NODE_STATUS_FRACTIONAL"><code>CALLBACK_NODE_STATUS_FRACTIONAL</code></a>: the primal solution available from <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> is integer infeasible.</li><li><a href="#MathOptInterface.CALLBACK_NODE_STATUS_UNKNOWN"><code>CALLBACK_NODE_STATUS_UNKNOWN</code></a>: the primal solution available from <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> might be integer feasible or infeasible.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L750-L763">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CALLBACK_NODE_STATUS_INTEGER" href="#MathOptInterface.CALLBACK_NODE_STATUS_INTEGER"><code>MathOptInterface.CALLBACK_NODE_STATUS_INTEGER</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">CALLBACK_NODE_STATUS_INTEGER::CallbackNodeStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.CallbackNodeStatusCode"><code>CallbackNodeStatusCode</code></a> enum.</p><p><code>CALLBACK_NODE_STATUS_INTEGER</code>: the primal solution available from <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> is integer feasible.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L750-L756">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CALLBACK_NODE_STATUS_FRACTIONAL" href="#MathOptInterface.CALLBACK_NODE_STATUS_FRACTIONAL"><code>MathOptInterface.CALLBACK_NODE_STATUS_FRACTIONAL</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">CALLBACK_NODE_STATUS_FRACTIONAL::CallbackNodeStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.CallbackNodeStatusCode"><code>CallbackNodeStatusCode</code></a> enum.</p><p><code>CALLBACK_NODE_STATUS_FRACTIONAL</code>: the primal solution available from <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> is integer infeasible.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L750-L756">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CALLBACK_NODE_STATUS_UNKNOWN" href="#MathOptInterface.CALLBACK_NODE_STATUS_UNKNOWN"><code>MathOptInterface.CALLBACK_NODE_STATUS_UNKNOWN</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">CALLBACK_NODE_STATUS_UNKNOWN::CallbackNodeStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.CallbackNodeStatusCode"><code>CallbackNodeStatusCode</code></a> enum.</p><p><code>CALLBACK_NODE_STATUS_UNKNOWN</code>: the primal solution available from <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> might be integer feasible or infeasible.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L750-L756">source</a></section></article><h2 id="Lazy-constraints"><a class="docs-heading-anchor" href="#Lazy-constraints">Lazy constraints</a><a id="Lazy-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Lazy-constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LazyConstraintCallback" href="#MathOptInterface.LazyConstraintCallback"><code>MathOptInterface.LazyConstraintCallback</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LazyConstraintCallback() &lt;: AbstractCallback</code></pre><p>The callback can be used to reduce the feasible set given the current primal solution by submitting a <a href="#MathOptInterface.LazyConstraint"><code>LazyConstraint</code></a>. For instance, it may be called at an incumbent of a mixed-integer problem. Note that there is no guarantee that the callback is called at <em>every</em> feasible primal solution.</p><p>The current primal solution is accessed through <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a>. Trying to access other result attributes will throw <a href="../errors/#MathOptInterface.OptimizeInProgress"><code>OptimizeInProgress</code></a> as discussed in <a href="#MathOptInterface.AbstractCallback"><code>AbstractCallback</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">x = MOI.add_variables(optimizer, 8)
+MOI.set(optimizer, MOI.LazyConstraintCallback(), callback_data -&gt; begin
+    sol = MOI.get(optimizer, MOI.CallbackVariablePrimal(callback_data), x)
+    if # should add a lazy constraint
+        func = # computes function
+        set = # computes set
+        MOI.submit(optimizer, MOI.LazyConstraint(callback_data), func, set)
+    end
+end)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L990-L1016">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LazyConstraint" href="#MathOptInterface.LazyConstraint"><code>MathOptInterface.LazyConstraint</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LazyConstraint(callback_data)</code></pre><p>Lazy constraint <code>func</code>-in-<code>set</code> submitted as <code>func, set</code>. The optimal solution returned by <a href="../variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a> will satisfy all lazy constraints that have been submitted.</p><p>This can be submitted only from the <a href="#MathOptInterface.LazyConstraintCallback"><code>LazyConstraintCallback</code></a>. The field <code>callback_data</code> is a solver-specific callback type that is passed as the argument to the feasible solution callback.</p><p><strong>Examples</strong></p><p>Suppose <code>x</code> and <code>y</code> are <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>s of <code>optimizer</code>. To add a <code>LazyConstraint</code> for <code>2x + 3y &lt;= 1</code>, write</p><pre><code class="language-julia hljs">func = 2.0x + 3.0y
+set = MOI.LessThan(1.0)
+MOI.submit(optimizer, MOI.LazyConstraint(callback_data), func, set)</code></pre><p>inside a <a href="#MathOptInterface.LazyConstraintCallback"><code>LazyConstraintCallback</code></a> of data <code>callback_data</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L642-L663">source</a></section></article><h2 id="User-cuts"><a class="docs-heading-anchor" href="#User-cuts">User cuts</a><a id="User-cuts-1"></a><a class="docs-heading-anchor-permalink" href="#User-cuts" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UserCutCallback" href="#MathOptInterface.UserCutCallback"><code>MathOptInterface.UserCutCallback</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UserCutCallback() &lt;: AbstractCallback</code></pre><p>The callback can be used to submit <a href="#MathOptInterface.UserCut"><code>UserCut</code></a> given the current primal solution. For instance, it may be called at fractional (i.e., non-integer) nodes in the branch and bound tree of a mixed-integer problem. Note that there is not guarantee that the callback is called <em>everytime</em> the solver has an infeasible solution.</p><p>The infeasible solution is accessed through <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a>. Trying to access other result attributes will throw <a href="../errors/#MathOptInterface.OptimizeInProgress"><code>OptimizeInProgress</code></a> as discussed in <a href="#MathOptInterface.AbstractCallback"><code>AbstractCallback</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">x = MOI.add_variables(optimizer, 8)
+MOI.set(optimizer, MOI.UserCutCallback(), callback_data -&gt; begin
+    sol = MOI.get(optimizer, MOI.CallbackVariablePrimal(callback_data), x)
+    if # can find a user cut
+        func = # computes function
+        set = # computes set
+        MOI.submit(optimizer, MOI.UserCut(callback_data), func, set)
+    end
+end</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1048-L1074">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UserCut" href="#MathOptInterface.UserCut"><code>MathOptInterface.UserCut</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UserCut(callback_data)</code></pre><p>Constraint <code>func</code>-to-<code>set</code> suggested to help the solver detect the solution given by <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> as infeasible. The cut is submitted as <code>func, set</code>. Typically <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a> will violate integrality constraints, and a cut would be of the form <a href="../standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>-in-<a href="../standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a> or <a href="../standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>-in-<a href="../standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a>. Note that, as opposed to <a href="#MathOptInterface.LazyConstraint"><code>LazyConstraint</code></a>, the provided constraint cannot modify the feasible set, the constraint should be redundant, e.g., it may be a consequence of affine and integrality constraints.</p><p>This can be submitted only from the <a href="#MathOptInterface.UserCutCallback"><code>UserCutCallback</code></a>. The field <code>callback_data</code> is a solver-specific callback type that is passed as the argument to the infeasible solution callback.</p><p>Note that the solver may silently ignore the provided constraint.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L704-L722">source</a></section></article><h2 id="Heuristic-solutions"><a class="docs-heading-anchor" href="#Heuristic-solutions">Heuristic solutions</a><a id="Heuristic-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Heuristic-solutions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HeuristicCallback" href="#MathOptInterface.HeuristicCallback"><code>MathOptInterface.HeuristicCallback</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HeuristicCallback() &lt;: AbstractCallback</code></pre><p>The callback can be used to submit <a href="#MathOptInterface.HeuristicSolution"><code>HeuristicSolution</code></a> given the current primal solution. For example, it may be called at fractional (i.e., non-integer) nodes in the branch and bound tree of a mixed-integer problem. Note that there is no guarantee that the callback is called <em>every</em> time the solver has an infeasible solution.</p><p>The current primal solution is accessed through <a href="#MathOptInterface.CallbackVariablePrimal"><code>CallbackVariablePrimal</code></a>. Trying to access other result attributes will throw <a href="../errors/#MathOptInterface.OptimizeInProgress"><code>OptimizeInProgress</code></a> as discussed in <a href="#MathOptInterface.AbstractCallback"><code>AbstractCallback</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">x = MOI.add_variables(optimizer, 8)
+MOI.set(optimizer, MOI.HeuristicCallback(), callback_data -&gt; begin
+    sol = MOI.get(optimizer, MOI.CallbackVariablePrimal(callback_data), x)
+    if # can find a heuristic solution
+        values = # computes heuristic solution
+        MOI.submit(optimizer, MOI.HeuristicSolution(callback_data), x,
+                   values)
+    end
+end</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1019-L1045">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HeuristicSolution" href="#MathOptInterface.HeuristicSolution"><code>MathOptInterface.HeuristicSolution</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HeuristicSolution(callback_data)</code></pre><p>Heuristically obtained feasible solution. The solution is submitted as <code>variables, values</code> where <code>values[i]</code> gives the value of <code>variables[i]</code>, similarly to <a href="../models/#MathOptInterface.set"><code>set</code></a>. The <a href="#MathOptInterface.submit"><code>submit</code></a> call returns a <a href="#MathOptInterface.HeuristicSolutionStatus"><code>HeuristicSolutionStatus</code></a> indicating whether the provided solution was accepted or rejected.</p><p>This can be submitted only from the <a href="#MathOptInterface.HeuristicCallback"><code>HeuristicCallback</code></a>. The field <code>callback_data</code> is a solver-specific callback type that is passed as the argument to the heuristic callback.</p><p>Some solvers require a complete solution, others only partial solutions.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L685-L699">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HeuristicSolutionStatus" href="#MathOptInterface.HeuristicSolutionStatus"><code>MathOptInterface.HeuristicSolutionStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HeuristicSolutionStatus</code></pre><p>An Enum of possible return values for <a href="#MathOptInterface.submit"><code>submit</code></a> with <a href="#MathOptInterface.HeuristicSolution"><code>HeuristicSolution</code></a>. This informs whether the heuristic solution was accepted or rejected.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="#MathOptInterface.HEURISTIC_SOLUTION_ACCEPTED"><code>HEURISTIC_SOLUTION_ACCEPTED</code></a>: The heuristic solution was accepted</li><li><a href="#MathOptInterface.HEURISTIC_SOLUTION_REJECTED"><code>HEURISTIC_SOLUTION_REJECTED</code></a>: The heuristic solution was rejected</li><li><a href="#MathOptInterface.HEURISTIC_SOLUTION_UNKNOWN"><code>HEURISTIC_SOLUTION_UNKNOWN</code></a>: No information available on the acceptance</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L668-L682">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HEURISTIC_SOLUTION_ACCEPTED" href="#MathOptInterface.HEURISTIC_SOLUTION_ACCEPTED"><code>MathOptInterface.HEURISTIC_SOLUTION_ACCEPTED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">HEURISTIC_SOLUTION_ACCEPTED::HeuristicSolutionStatus</code></pre><p>An instance of the <a href="#MathOptInterface.HeuristicSolutionStatus"><code>HeuristicSolutionStatus</code></a> enum.</p><p><code>HEURISTIC_SOLUTION_ACCEPTED</code>: The heuristic solution was accepted</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L668-L674">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HEURISTIC_SOLUTION_REJECTED" href="#MathOptInterface.HEURISTIC_SOLUTION_REJECTED"><code>MathOptInterface.HEURISTIC_SOLUTION_REJECTED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">HEURISTIC_SOLUTION_REJECTED::HeuristicSolutionStatus</code></pre><p>An instance of the <a href="#MathOptInterface.HeuristicSolutionStatus"><code>HeuristicSolutionStatus</code></a> enum.</p><p><code>HEURISTIC_SOLUTION_REJECTED</code>: The heuristic solution was rejected</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L668-L674">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HEURISTIC_SOLUTION_UNKNOWN" href="#MathOptInterface.HEURISTIC_SOLUTION_UNKNOWN"><code>MathOptInterface.HEURISTIC_SOLUTION_UNKNOWN</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">HEURISTIC_SOLUTION_UNKNOWN::HeuristicSolutionStatus</code></pre><p>An instance of the <a href="#MathOptInterface.HeuristicSolutionStatus"><code>HeuristicSolutionStatus</code></a> enum.</p><p><code>HEURISTIC_SOLUTION_UNKNOWN</code>: No information available on the acceptance</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L668-L674">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../nonlinear/">« Nonlinear programming</a><a class="docs-footer-nextpage" href="../errors/">Errors »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/constraints/index.html b/previews/PR3525/moi/reference/constraints/index.html
new file mode 100644
index 00000000000..1996502a4c2
--- /dev/null
+++ b/previews/PR3525/moi/reference/constraints/index.html
@@ -0,0 +1,16 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Constraints · JuMP</title><meta name="title" content="Constraints · JuMP"/><meta property="og:title" content="Constraints · JuMP"/><meta property="twitter:title" content="Constraints · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li class="is-active"><a class="tocitem" href>Constraints</a><ul class="internal"><li><a class="tocitem" href="#Types"><span>Types</span></a></li><li><a class="tocitem" href="#Functions"><span>Functions</span></a></li><li><a class="tocitem" href="#Attributes"><span>Attributes</span></a></li></ul></li><li><a class="tocitem" href="../modification/">Modifications</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Constraints</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Constraints</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/constraints.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="constraints_ref"><a class="docs-heading-anchor" href="#constraints_ref">Constraints</a><a id="constraints_ref-1"></a><a class="docs-heading-anchor-permalink" href="#constraints_ref" title="Permalink"></a></h1><h2 id="Types"><a class="docs-heading-anchor" href="#Types">Types</a><a id="Types-1"></a><a class="docs-heading-anchor-permalink" href="#Types" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintIndex" href="#MathOptInterface.ConstraintIndex"><code>MathOptInterface.ConstraintIndex</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintIndex{F, S}</code></pre><p>A type-safe wrapper for <code>Int64</code> for use in referencing <code>F</code>-in-<code>S</code> constraints in a model. The parameter <code>F</code> is the type of the function in the constraint, and the parameter <code>S</code> is the type of set in the constraint. To allow for deletion, indices need not be consecutive. Indices within a constraint type (i.e. <code>F</code>-in-<code>S</code>) must be unique, but non-unique indices across different constraint types are allowed. If <code>F</code> is <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> then the index is equal to the index of the variable. That is for an <code>index::ConstraintIndex{VariableIndex}</code>, we always have</p><pre><code class="language-julia hljs">index.value == MOI.get(model, MOI.ConstraintFunction(), index).value</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L48-L63">source</a></section></article><h2 id="Functions"><a class="docs-heading-anchor" href="#Functions">Functions</a><a id="Functions-1"></a><a class="docs-heading-anchor-permalink" href="#Functions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.ConstraintIndex}" href="#MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.ConstraintIndex}"><code>MathOptInterface.is_valid</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">is_valid(model::ModelLike, index::Index)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether this index refers to a valid object in the model <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L95-L99">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.add_constraint" href="#MathOptInterface.add_constraint"><code>MathOptInterface.add_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_constraint(model::ModelLike, func::F, set::S)::ConstraintIndex{F,S} where {F,S}</code></pre><p>Add the constraint <span>$f(x) \in \mathcal{S}$</span> where <span>$f$</span> is defined by <code>func</code>, and <span>$\mathcal{S}$</span> is defined by <code>set</code>.</p><pre><code class="nohighlight hljs">add_constraint(model::ModelLike, v::VariableIndex, set::S)::ConstraintIndex{VariableIndex,S} where {S}
+add_constraint(model::ModelLike, vec::Vector{VariableIndex}, set::S)::ConstraintIndex{VectorOfVariables,S} where {S}</code></pre><p>Add the constraint <span>$v \in \mathcal{S}$</span> where <span>$v$</span> is the variable (or vector of variables) referenced by <code>v</code> and <span>$\mathcal{S}$</span> is defined by <code>set</code>.</p><ul><li>An <a href="../errors/#MathOptInterface.UnsupportedConstraint"><code>UnsupportedConstraint</code></a> error is thrown if <code>model</code> does not support <code>F</code>-in-<code>S</code> constraints,</li><li>a <a href="../errors/#MathOptInterface.AddConstraintNotAllowed"><code>AddConstraintNotAllowed</code></a> error is thrown if it supports <code>F</code>-in-<code>S</code> constraints but it cannot add the constraint(s) in its current state and</li><li>a <a href="../errors/#MathOptInterface.ScalarFunctionConstantNotZero"><code>ScalarFunctionConstantNotZero</code></a> error may be thrown if <code>func</code> is an <code>AbstractScalarFunction</code> with nonzero constant and <code>set</code> is <a href="../standard_form/#MathOptInterface.EqualTo"><code>EqualTo</code></a>, <a href="../standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a>, <a href="../standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a> or <a href="../standard_form/#MathOptInterface.Interval"><code>Interval</code></a>.</li><li>a <a href="../errors/#MathOptInterface.LowerBoundAlreadySet"><code>LowerBoundAlreadySet</code></a> error is thrown if <code>F</code> is a <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> and a constraint was already added to this variable that sets a lower bound.</li><li>a <a href="../errors/#MathOptInterface.UpperBoundAlreadySet"><code>UpperBoundAlreadySet</code></a> error is thrown if <code>F</code> is a <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> and a constraint was already added to this variable that sets an upper bound.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L125-L149">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.add_constraints" href="#MathOptInterface.add_constraints"><code>MathOptInterface.add_constraints</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_constraints(model::ModelLike, funcs::Vector{F}, sets::Vector{S})::Vector{ConstraintIndex{F,S}} where {F,S}</code></pre><p>Add the set of constraints specified by each function-set pair in <code>funcs</code> and <code>sets</code>. <code>F</code> and <code>S</code> should be concrete types. This call is equivalent to <code>add_constraint.(model, funcs, sets)</code> but may be more efficient.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L231-L236">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.transform" href="#MathOptInterface.transform"><code>MathOptInterface.transform</code></a> — <span class="docstring-category">Function</span></header><section><div><p><strong>Transform Constraint Set</strong></p><pre><code class="nohighlight hljs">transform(model::ModelLike, c::ConstraintIndex{F,S1}, newset::S2)::ConstraintIndex{F,S2}</code></pre><p>Replace the set in constraint <code>c</code> with <code>newset</code>. The constraint index <code>c</code> will no longer be valid, and the function returns a new constraint index with the correct type.</p><p>Solvers may only support a subset of constraint transforms that they perform efficiently (for example, changing from a <code>LessThan</code> to <code>GreaterThan</code> set). In addition, set modification (where <code>S1 = S2</code>) should be performed via the <code>modify</code> function.</p><p>Typically, the user should delete the constraint and add a new one.</p><p><strong>Examples</strong></p><p>If <code>c</code> is a <code>ConstraintIndex{ScalarAffineFunction{Float64},LessThan{Float64}}</code>,</p><pre><code class="language-julia hljs">c2 = transform(model, c, GreaterThan(0.0))
+transform(model, c, LessThan(0.0)) # errors</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L288-L313">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.supports_constraint" href="#MathOptInterface.supports_constraint"><code>MathOptInterface.supports_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_constraint(
+    model::ModelLike,
+    ::Type{F},
+    ::Type{S},
+)::Bool where {F&lt;:AbstractFunction,S&lt;:AbstractSet}</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports <code>F</code>-in-<code>S</code> constraints, that is, <code>copy_to(model, src)</code> does not throw <a href="../errors/#MathOptInterface.UnsupportedConstraint"><code>UnsupportedConstraint</code></a> when <code>src</code> contains <code>F</code>-in-<code>S</code> constraints. If <code>F</code>-in-<code>S</code> constraints are only not supported in specific circumstances, e.g. <code>F</code>-in-<code>S</code> constraints cannot be combined with another type of constraint, it should still return <code>true</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L7-L19">source</a></section><section><div><pre><code class="language-julia hljs">MOI.supports_constraint(
+    BT::Type{&lt;:AbstractBridge},
+    F::Type{&lt;:MOI.AbstractFunction},
+    S::Type{&lt;:MOI.AbstractSet},
+)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether the bridges of type <code>BT</code> support bridging <code>F</code>-in-<code>S</code> constraints.</p><p><strong>Implementation notes</strong></p><ul><li>This method depends only on the type of the inputs, not the runtime values.</li><li>There is a default fallback, so you need only implement this method for constraint types that the bridge implements.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridge.jl#L22-L37">source</a></section></article><h2 id="Attributes"><a class="docs-heading-anchor" href="#Attributes">Attributes</a><a id="Attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractConstraintAttribute" href="#MathOptInterface.AbstractConstraintAttribute"><code>MathOptInterface.AbstractConstraintAttribute</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractConstraintAttribute</code></pre><p>Abstract supertype for attribute objects that can be used to set or get attributes (properties) of constraints in the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L38-L43">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintName" href="#MathOptInterface.ConstraintName"><code>MathOptInterface.ConstraintName</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintName()</code></pre><p>A constraint attribute for a string identifying the constraint.</p><p>It is <em>valid</em> for constraints variables to have the same name; however, constraints with duplicate names cannot be looked up using <a href="../models/#MathOptInterface.get"><code>get</code></a>, regardless of whether they have the same <code>F</code>-in-<code>S</code> type.</p><p><code>ConstraintName</code> has a default value of <code>&quot;&quot;</code> if not set.</p><p><strong>Notes</strong></p><p>You should <em>not</em> implement <code>ConstraintName</code> for <code>VariableIndex</code> constraints.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1555-L1569">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintPrimalStart" href="#MathOptInterface.ConstraintPrimalStart"><code>MathOptInterface.ConstraintPrimalStart</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintPrimalStart()</code></pre><p>A constraint attribute for the initial assignment to some constraint&#39;s <a href="#MathOptInterface.ConstraintPrimal"><code>ConstraintPrimal</code></a> that the optimizer may use to warm-start the solve.</p><p>May be <code>nothing</code> (unset), a number for <a href="../standard_form/#MathOptInterface.AbstractScalarFunction"><code>AbstractScalarFunction</code></a>, or a vector for <a href="../standard_form/#MathOptInterface.AbstractVectorFunction"><code>AbstractVectorFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1595-L1603">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintDualStart" href="#MathOptInterface.ConstraintDualStart"><code>MathOptInterface.ConstraintDualStart</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintDualStart()</code></pre><p>A constraint attribute for the initial assignment to some constraint&#39;s <a href="#MathOptInterface.ConstraintDual"><code>ConstraintDual</code></a> that the optimizer may use to warm-start the solve.</p><p>May be <code>nothing</code> (unset), a number for <a href="../standard_form/#MathOptInterface.AbstractScalarFunction"><code>AbstractScalarFunction</code></a>, or a vector for <a href="../standard_form/#MathOptInterface.AbstractVectorFunction"><code>AbstractVectorFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1606-L1614">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintPrimal" href="#MathOptInterface.ConstraintPrimal"><code>MathOptInterface.ConstraintPrimal</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintPrimal(result_index::Int = 1)</code></pre><p>A constraint attribute for the assignment to some constraint&#39;s primal value(s) in result <code>result_index</code>.</p><p>If the constraint is <code>f(x) in S</code>, then in most cases the <code>ConstraintPrimal</code> is the value of <code>f</code>, evaluated at the correspondng <a href="../variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a> solution.</p><p>However, some conic solvers reformulate <code>b - Ax in S</code> to <code>s = b - Ax, s in S</code>. These solvers may return the value of <code>s</code> for <code>ConstraintPrimal</code>, rather than <code>b - Ax</code>. (Although these are constrained by an equality constraint, due to numerical tolerances they may not be identical.)</p><p>If the solver does not have a primal value for the constraint because the <code>result_index</code> is beyond the available solutions (whose number is indicated by the <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute), getting this attribute must throw a <a href="../errors/#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check <a href="../models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> before accessing the <code>ConstraintPrimal</code> attribute.</p><p>If <code>result_index</code> is omitted, it is 1 by default. See <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1617-L1642">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintDual" href="#MathOptInterface.ConstraintDual"><code>MathOptInterface.ConstraintDual</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintDual(result_index::Int = 1)</code></pre><p>A constraint attribute for the assignment to some constraint&#39;s dual value(s) in result <code>result_index</code>. If <code>result_index</code> is omitted, it is 1 by default.</p><p>If the solver does not have a dual value for the variable because the <code>result_index</code> is beyond the available solutions (whose number is indicated by the <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute), getting this attribute must throw a <a href="../errors/#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>. Otherwise, if the result is unavailable for another reason (for instance, only a primal solution is available), the result is undefined. Users should first check <a href="../models/#MathOptInterface.DualStatus"><code>DualStatus</code></a> before accessing the <code>ConstraintDual</code> attribute.</p><p>See <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1648-L1663">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintBasisStatus" href="#MathOptInterface.ConstraintBasisStatus"><code>MathOptInterface.ConstraintBasisStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintBasisStatus(result_index::Int = 1)</code></pre><p>A constraint attribute for the <code>BasisStatusCode</code> of some constraint in result <code>result_index</code>, with respect to an available optimal solution basis. If <code>result_index</code> is omitted, it is 1 by default.</p><p>If the solver does not have a basis statue for the constraint because the <code>result_index</code> is beyond the available solutions (whose number is indicated by the <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute), getting this attribute must throw a <a href="../errors/#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check <a href="../models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> before accessing the <code>ConstraintBasisStatus</code> attribute.</p><p>See <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p><p><strong>Notes</strong></p><p>For the basis status of a variable, query <a href="../variables/#MathOptInterface.VariableBasisStatus"><code>VariableBasisStatus</code></a>.</p><p><code>ConstraintBasisStatus</code> does not apply to <code>VariableIndex</code> constraints. You can infer the basis status of a <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraint by looking at the result of <a href="../variables/#MathOptInterface.VariableBasisStatus"><code>VariableBasisStatus</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1669-L1693">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintFunction" href="#MathOptInterface.ConstraintFunction"><code>MathOptInterface.ConstraintFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintFunction()</code></pre><p>A constraint attribute for the <code>AbstractFunction</code> object used to define the constraint.</p><p>It is guaranteed to be equivalent but not necessarily identical to the function provided by the user.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1755-L1763">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CanonicalConstraintFunction" href="#MathOptInterface.CanonicalConstraintFunction"><code>MathOptInterface.CanonicalConstraintFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CanonicalConstraintFunction()</code></pre><p>A constraint attribute for a canonical representation of the <a href="../standard_form/#MathOptInterface.AbstractFunction"><code>AbstractFunction</code></a> object used to define the constraint.</p><p>Getting this attribute is guaranteed to return a function that is equivalent but not necessarily identical to the function provided by the user.</p><p>By default, <code>MOI.get(model, MOI.CanonicalConstraintFunction(), ci)</code> fallbacks to <code>MOI.Utilities.canonical(MOI.get(model, MOI.ConstraintFunction(), ci))</code>. However, if <code>model</code> knows that the constraint function is canonical then it can implement a specialized method that directly return the function without calling <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.canonical"><code>Utilities.canonical</code></a>. Therefore, the value returned <strong>cannot</strong> be assumed to be a copy of the function stored in <code>model</code>. Moreover, <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.Model"><code>Utilities.Model</code></a> checks with <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.is_canonical"><code>Utilities.is_canonical</code></a> whether the function stored internally is already canonical and if it&#39;s the case, then it returns the function stored internally instead of a copy.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1712-L1730">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintSet" href="#MathOptInterface.ConstraintSet"><code>MathOptInterface.ConstraintSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintSet()</code></pre><p>A constraint attribute for the <code>AbstractSet</code> object used to define the constraint.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1800-L1805">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.BasisStatusCode" href="#MathOptInterface.BasisStatusCode"><code>MathOptInterface.BasisStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">BasisStatusCode</code></pre><p>An Enum of possible values for the <a href="#MathOptInterface.ConstraintBasisStatus"><code>ConstraintBasisStatus</code></a> and <a href="../variables/#MathOptInterface.VariableBasisStatus"><code>VariableBasisStatus</code></a> attributes, explaining the status of a given element with respect to an optimal solution basis.</p><p><strong>Notes</strong></p><ul><li><code>NONBASIC_AT_LOWER</code> and <code>NONBASIC_AT_UPPER</code> should be used only for</li></ul><p>constraints with the <code>Interval</code> set. In this case, they are necessary to distinguish which side of the constraint is active. One-sided constraints (e.g., <code>LessThan</code> and <code>GreaterThan</code>) should use <code>NONBASIC</code> instead of the <code>NONBASIC_AT_*</code> values. This restriction does not apply to <a href="../variables/#MathOptInterface.VariableBasisStatus"><code>VariableBasisStatus</code></a>, which should return <code>NONBASIC_AT_*</code> regardless of whether the alternative bound exists.</p><ul><li>In linear programs, <code>SUPER_BASIC</code> occurs when a variable with no bounds is not</li></ul><p>in the basis.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="#MathOptInterface.BASIC"><code>BASIC</code></a>: element is in the basis</li><li><a href="#MathOptInterface.NONBASIC"><code>NONBASIC</code></a>: element is not in the basis</li><li><a href="#MathOptInterface.NONBASIC_AT_LOWER"><code>NONBASIC_AT_LOWER</code></a>: element is not in the basis and is at its lower bound</li><li><a href="#MathOptInterface.NONBASIC_AT_UPPER"><code>NONBASIC_AT_UPPER</code></a>: element is not in the basis and is at its upper bound</li><li><a href="#MathOptInterface.SUPER_BASIC"><code>SUPER_BASIC</code></a>: element is not in the basis but is also not at one of its bounds</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1482-L1511">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.BASIC" href="#MathOptInterface.BASIC"><code>MathOptInterface.BASIC</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">BASIC::BasisStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.BasisStatusCode"><code>BasisStatusCode</code></a> enum.</p><p><code>BASIC</code>: element is in the basis</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1482-L1488">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NONBASIC" href="#MathOptInterface.NONBASIC"><code>MathOptInterface.NONBASIC</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NONBASIC::BasisStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.BasisStatusCode"><code>BasisStatusCode</code></a> enum.</p><p><code>NONBASIC</code>: element is not in the basis</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1482-L1488">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NONBASIC_AT_LOWER" href="#MathOptInterface.NONBASIC_AT_LOWER"><code>MathOptInterface.NONBASIC_AT_LOWER</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NONBASIC_AT_LOWER::BasisStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.BasisStatusCode"><code>BasisStatusCode</code></a> enum.</p><p><code>NONBASIC_AT_LOWER</code>: element is not in the basis and is at its lower bound</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1482-L1488">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NONBASIC_AT_UPPER" href="#MathOptInterface.NONBASIC_AT_UPPER"><code>MathOptInterface.NONBASIC_AT_UPPER</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NONBASIC_AT_UPPER::BasisStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.BasisStatusCode"><code>BasisStatusCode</code></a> enum.</p><p><code>NONBASIC_AT_UPPER</code>: element is not in the basis and is at its upper bound</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1482-L1488">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SUPER_BASIC" href="#MathOptInterface.SUPER_BASIC"><code>MathOptInterface.SUPER_BASIC</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">SUPER_BASIC::BasisStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.BasisStatusCode"><code>BasisStatusCode</code></a> enum.</p><p><code>SUPER_BASIC</code>: element is not in the basis but is also not at one of its bounds</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1482-L1488">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../variables/">« Variables</a><a class="docs-footer-nextpage" href="../modification/">Modifications »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/errors/index.html b/previews/PR3525/moi/reference/errors/index.html
new file mode 100644
index 00000000000..24ccfcc1ffc
--- /dev/null
+++ b/previews/PR3525/moi/reference/errors/index.html
@@ -0,0 +1,52 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Errors · JuMP</title><meta name="title" content="Errors · JuMP"/><meta property="og:title" content="Errors · JuMP"/><meta property="twitter:title" content="Errors · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../modification/">Modifications</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li class="is-active"><a class="tocitem" href>Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Errors</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Errors</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/errors.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Errors"><a class="docs-heading-anchor" href="#Errors">Errors</a><a id="Errors-1"></a><a class="docs-heading-anchor-permalink" href="#Errors" title="Permalink"></a></h1><p>When an MOI call fails on a model, precise errors should be thrown when possible instead of simply calling <code>error</code> with a message. The docstrings for the respective methods describe the errors that the implementation should throw in certain situations. This error-reporting system allows code to distinguish between internal errors (that should be shown to the user) and unsupported operations which may have automatic workarounds.</p><p>When an invalid index is used in an MOI call, an <a href="#MathOptInterface.InvalidIndex"><code>InvalidIndex</code></a> is thrown:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.InvalidIndex" href="#MathOptInterface.InvalidIndex"><code>MathOptInterface.InvalidIndex</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct InvalidIndex{IndexType&lt;:Index} &lt;: Exception
+    index::IndexType
+end</code></pre><p>An error indicating that the index <code>index</code> is invalid.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L78-L84">source</a></section></article><p>When an invalid result index is used to retrieve an attribute, a <a href="#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a> is thrown:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ResultIndexBoundsError" href="#MathOptInterface.ResultIndexBoundsError"><code>MathOptInterface.ResultIndexBoundsError</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ResultIndexBoundsError{AttrType} &lt;: Exception
+    attr::AttrType
+    result_count::Int
+end</code></pre><p>An error indicating that the requested attribute <code>attr</code> could not be retrieved, because the solver returned too few results compared to what was requested. For instance, the user tries to retrieve <code>VariablePrimal(2)</code> when only one solution is available, or when the model is infeasible and has no solution.</p><p>See also: <a href="#MathOptInterface.check_result_index_bounds"><code>check_result_index_bounds</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L178-L190">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.check_result_index_bounds" href="#MathOptInterface.check_result_index_bounds"><code>MathOptInterface.check_result_index_bounds</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">check_result_index_bounds(model::ModelLike, attr)</code></pre><p>This function checks whether enough results are available in the <code>model</code> for the requested <code>attr</code>, using its <code>result_index</code> field. If the model does not have sufficient results to answer the query, it throws a <a href="#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L196-L203">source</a></section></article><p>As discussed in <a href="../../manual/constraints/#JuMP-mapping">JuMP mapping</a>, for scalar constraint with a nonzero function constant, a <a href="#MathOptInterface.ScalarFunctionConstantNotZero"><code>ScalarFunctionConstantNotZero</code></a> exception may be thrown:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarFunctionConstantNotZero" href="#MathOptInterface.ScalarFunctionConstantNotZero"><code>MathOptInterface.ScalarFunctionConstantNotZero</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ScalarFunctionConstantNotZero{T, F, S} &lt;: Exception
+    constant::T
+end</code></pre><p>An error indicating that the constant part of the function in the constraint <code>F</code>-in-<code>S</code> is nonzero.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L64-L71">source</a></section></article><p>Some <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraints cannot be combined on the same variable:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LowerBoundAlreadySet" href="#MathOptInterface.LowerBoundAlreadySet"><code>MathOptInterface.LowerBoundAlreadySet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LowerBoundAlreadySet{S1, S2}</code></pre><p>Error thrown when setting a <code>VariableIndex</code>-in-<code>S2</code> when a <code>VariableIndex</code>-in-<code>S1</code> has already been added and the sets <code>S1</code>, <code>S2</code> both set a lower bound, i.e. they are <a href="../standard_form/#MathOptInterface.EqualTo"><code>EqualTo</code></a>, <a href="../standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a>, <a href="../standard_form/#MathOptInterface.Interval"><code>Interval</code></a>, <a href="../standard_form/#MathOptInterface.Semicontinuous"><code>Semicontinuous</code></a> or <a href="../standard_form/#MathOptInterface.Semiinteger"><code>Semiinteger</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L244-L251">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UpperBoundAlreadySet" href="#MathOptInterface.UpperBoundAlreadySet"><code>MathOptInterface.UpperBoundAlreadySet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UpperBoundAlreadySet{S1, S2}</code></pre><p>Error thrown when setting a <code>VariableIndex</code>-in-<code>S2</code> when a <code>VariableIndex</code>-in-<code>S1</code> has already been added and the sets <code>S1</code>, <code>S2</code> both set an upper bound, i.e. they are <a href="../standard_form/#MathOptInterface.EqualTo"><code>EqualTo</code></a>, <a href="../standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a>, <a href="../standard_form/#MathOptInterface.Interval"><code>Interval</code></a>, <a href="../standard_form/#MathOptInterface.Semicontinuous"><code>Semicontinuous</code></a> or <a href="../standard_form/#MathOptInterface.Semiinteger"><code>Semiinteger</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L266-L273">source</a></section></article><p>As discussed in <a href="../callbacks/#MathOptInterface.AbstractCallback"><code>AbstractCallback</code></a>, trying to <a href="../models/#MathOptInterface.get"><code>get</code></a> attributes inside a callback may throw:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OptimizeInProgress" href="#MathOptInterface.OptimizeInProgress"><code>MathOptInterface.OptimizeInProgress</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct OptimizeInProgress{AttrType&lt;:AnyAttribute} &lt;: Exception
+    attr::AttrType
+end</code></pre><p>Error thrown from optimizer when <code>MOI.get(optimizer, attr)</code> is called inside an <a href="../callbacks/#MathOptInterface.AbstractCallback"><code>AbstractCallback</code></a> while it is only defined once <a href="../models/#MathOptInterface.optimize!"><code>optimize!</code></a> has completed. This can only happen when <code>is_set_by_optimize(attr)</code> is <code>true</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L945-L953">source</a></section></article><p>Trying to submit the wrong type of <a href="../callbacks/#MathOptInterface.AbstractSubmittable"><code>AbstractSubmittable</code></a> inside an <a href="../callbacks/#MathOptInterface.AbstractCallback"><code>AbstractCallback</code></a> (for example, a <a href="../callbacks/#MathOptInterface.UserCut"><code>UserCut</code></a> inside a <a href="../callbacks/#MathOptInterface.LazyConstraintCallback"><code>LazyConstraintCallback</code></a>) will throw:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.InvalidCallbackUsage" href="#MathOptInterface.InvalidCallbackUsage"><code>MathOptInterface.InvalidCallbackUsage</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct InvalidCallbackUsage{C, S} &lt;: Exception
+    callback::C
+    submittable::S
+end</code></pre><p>An error indicating that <code>submittable</code> cannot be submitted inside <code>callback</code>.</p><p>For example, <a href="../callbacks/#MathOptInterface.UserCut"><code>UserCut</code></a> cannot be submitted inside <a href="../callbacks/#MathOptInterface.LazyConstraintCallback"><code>LazyConstraintCallback</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L727-L737">source</a></section></article><p>The rest of the errors defined in MOI fall in two categories represented by the following two abstract types:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UnsupportedError" href="#MathOptInterface.UnsupportedError"><code>MathOptInterface.UnsupportedError</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UnsupportedError &lt;: Exception</code></pre><p>Abstract type for error thrown when an element is not supported by the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/error.jl#L7-L11">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NotAllowedError" href="#MathOptInterface.NotAllowedError"><code>MathOptInterface.NotAllowedError</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NotAllowedError &lt;: Exception</code></pre><p>Abstract type for error thrown when an operation is supported but cannot be applied in the current state of the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/error.jl#L37-L42">source</a></section></article><p>The different <a href="#MathOptInterface.UnsupportedError"><code>UnsupportedError</code></a> and <a href="#MathOptInterface.NotAllowedError"><code>NotAllowedError</code></a> are the following errors:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UnsupportedAttribute" href="#MathOptInterface.UnsupportedAttribute"><code>MathOptInterface.UnsupportedAttribute</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct UnsupportedAttribute{AttrType} &lt;: UnsupportedError
+    attr::AttrType
+    message::String
+end</code></pre><p>An error indicating that the attribute <code>attr</code> is not supported by the model, i.e. that <a href="../models/#MathOptInterface.supports"><code>supports</code></a> returns <code>false</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L72-L80">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SetAttributeNotAllowed" href="#MathOptInterface.SetAttributeNotAllowed"><code>MathOptInterface.SetAttributeNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct SetAttributeNotAllowed{AttrType} &lt;: NotAllowedError
+    attr::AttrType
+    message::String
+end</code></pre><p>An error indicating that the attribute <code>attr</code> is supported (see <a href="../models/#MathOptInterface.supports"><code>supports</code></a>) but cannot be set for some reason (see the error string).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L90-L98">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AddVariableNotAllowed" href="#MathOptInterface.AddVariableNotAllowed"><code>MathOptInterface.AddVariableNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct AddVariableNotAllowed &lt;: NotAllowedError
+    message::String # Human-friendly explanation why the attribute cannot be set
+end</code></pre><p>An error indicating that variables cannot be added to the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L9-L15">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UnsupportedConstraint" href="#MathOptInterface.UnsupportedConstraint"><code>MathOptInterface.UnsupportedConstraint</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct UnsupportedConstraint{F&lt;:AbstractFunction, S&lt;:AbstractSet} &lt;: UnsupportedError
+    message::String # Human-friendly explanation why the attribute cannot be set
+end</code></pre><p>An error indicating that constraints of type <code>F</code>-in-<code>S</code> are not supported by the model, i.e. that <a href="../constraints/#MathOptInterface.supports_constraint"><code>supports_constraint</code></a> returns <code>false</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L28-L35">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AddConstraintNotAllowed" href="#MathOptInterface.AddConstraintNotAllowed"><code>MathOptInterface.AddConstraintNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct AddConstraintNotAllowed{F&lt;:AbstractFunction, S&lt;:AbstractSet} &lt;: NotAllowedError
+    message::String # Human-friendly explanation why the attribute cannot be set
+end</code></pre><p>An error indicating that constraints of type <code>F</code>-in-<code>S</code> are supported (see <a href="../constraints/#MathOptInterface.supports_constraint"><code>supports_constraint</code></a>) but cannot be added.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/constraints.jl#L46-L53">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ModifyConstraintNotAllowed" href="#MathOptInterface.ModifyConstraintNotAllowed"><code>MathOptInterface.ModifyConstraintNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ModifyConstraintNotAllowed{F&lt;:AbstractFunction, S&lt;:AbstractSet,
+                                  C&lt;:AbstractFunctionModification} &lt;: NotAllowedError
+    constraint_index::ConstraintIndex{F, S}
+    change::C
+    message::String
+end</code></pre><p>An error indicating that the constraint modification <code>change</code> cannot be applied to the constraint of index <code>ci</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/modifications.jl#L7-L17">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ModifyObjectiveNotAllowed" href="#MathOptInterface.ModifyObjectiveNotAllowed"><code>MathOptInterface.ModifyObjectiveNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ModifyObjectiveNotAllowed{C&lt;:AbstractFunctionModification} &lt;: NotAllowedError
+    change::C
+    message::String
+end</code></pre><p>An error indicating that the objective modification <code>change</code> cannot be applied to the objective.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/modifications.jl#L42-L50">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.DeleteNotAllowed" href="#MathOptInterface.DeleteNotAllowed"><code>MathOptInterface.DeleteNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct DeleteNotAllowed{IndexType &lt;: Index} &lt;: NotAllowedError
+    index::IndexType
+    message::String
+end</code></pre><p>An error indicating that the index <code>index</code> cannot be deleted.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L114-L121">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UnsupportedSubmittable" href="#MathOptInterface.UnsupportedSubmittable"><code>MathOptInterface.UnsupportedSubmittable</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct UnsupportedSubmittable{SubmitType} &lt;: UnsupportedError
+    sub::SubmitType
+    message::String
+end</code></pre><p>An error indicating that the submittable <code>sub</code> is not supported by the model, i.e. that <a href="../models/#MathOptInterface.supports"><code>supports</code></a> returns <code>false</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L143-L151">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SubmitNotAllowed" href="#MathOptInterface.SubmitNotAllowed"><code>MathOptInterface.SubmitNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct SubmitNotAllowed{SubmitTyp&lt;:AbstractSubmittable} &lt;: NotAllowedError
+    sub::SubmitType
+    message::String
+end</code></pre><p>An error indicating that the submittable <code>sub</code> is supported (see <a href="../models/#MathOptInterface.supports"><code>supports</code></a>) but cannot be added for some reason (see the error string).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L158-L166">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UnsupportedNonlinearOperator" href="#MathOptInterface.UnsupportedNonlinearOperator"><code>MathOptInterface.UnsupportedNonlinearOperator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UnsupportedNonlinearOperator(head::Symbol[, message::String]) &lt;: UnsupportedError</code></pre><p>An error thrown by optimizers if they do not support the operator <code>head</code> in a <a href="../standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; throw(MOI.UnsupportedNonlinearOperator(:black_box))
+ERROR: MathOptInterface.UnsupportedNonlinearOperator: The nonlinear operator `:black_box` is not supported by the model.
+Stacktrace:
+[...]</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L366-L382">source</a></section></article><p>Note that setting the <a href="../constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> of a <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraint is not allowed:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SettingVariableIndexNotAllowed" href="#MathOptInterface.SettingVariableIndexNotAllowed"><code>MathOptInterface.SettingVariableIndexNotAllowed</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SettingVariableIndexNotAllowed()</code></pre><p>Error type that should be thrown when the user calls <a href="../models/#MathOptInterface.set"><code>set</code></a> to change the <a href="../constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> of a <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraint.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L604-L609">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../callbacks/">« Callbacks</a><a class="docs-footer-nextpage" href="../../submodules/Benchmarks/overview/">Overview »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/models/index.html b/previews/PR3525/moi/reference/models/index.html
new file mode 100644
index 00000000000..57e8ad60d9a
--- /dev/null
+++ b/previews/PR3525/moi/reference/models/index.html
@@ -0,0 +1,136 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Models · JuMP</title><meta name="title" content="Models · JuMP"/><meta property="og:title" content="Models · JuMP"/><meta property="twitter:title" content="Models · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form</a></li><li class="is-active"><a class="tocitem" href>Models</a><ul class="internal"><li><a class="tocitem" href="#Attribute-interface"><span>Attribute interface</span></a></li><li><a class="tocitem" href="#Model-interface"><span>Model interface</span></a></li><li><a class="tocitem" href="#Model-attributes"><span>Model attributes</span></a></li><li><a class="tocitem" href="#Optimizer-interface"><span>Optimizer interface</span></a></li><li><a class="tocitem" href="#Optimizer-attributes"><span>Optimizer attributes</span></a></li></ul></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../modification/">Modifications</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Models</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Models</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/models.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Models"><a class="docs-heading-anchor" href="#Models">Models</a><a id="Models-1"></a><a class="docs-heading-anchor-permalink" href="#Models" title="Permalink"></a></h1><h2 id="Attribute-interface"><a class="docs-heading-anchor" href="#Attribute-interface">Attribute interface</a><a id="Attribute-interface-1"></a><a class="docs-heading-anchor-permalink" href="#Attribute-interface" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.is_set_by_optimize" href="#MathOptInterface.is_set_by_optimize"><code>MathOptInterface.is_set_by_optimize</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_set_by_optimize(::AnyAttribute)</code></pre><p>Return a <code>Bool</code> indicating whether the value of the attribute is modified during an <a href="#MathOptInterface.optimize!"><code>optimize!</code></a> call, that is, the attribute is used to query the result of the optimization.</p><p><strong>Important note when defining new attributes</strong></p><p>This function returns <code>false</code> by default so it should be implemented for attributes that are modified by <a href="#MathOptInterface.optimize!"><code>optimize!</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2158-L2169">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.is_copyable" href="#MathOptInterface.is_copyable"><code>MathOptInterface.is_copyable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_copyable(::AnyAttribute)</code></pre><p>Return a <code>Bool</code> indicating whether the value of the attribute may be copied during <a href="#MathOptInterface.copy_to"><code>copy_to</code></a> using <a href="#MathOptInterface.set"><code>set</code></a>.</p><p><strong>Important note when defining new attributes</strong></p><p>By default <code>is_copyable(attr)</code> returns <code>!is_set_by_optimize(attr)</code>. A specific method should be defined for attributes which are copied indirectly during <a href="#MathOptInterface.copy_to"><code>copy_to</code></a>. For instance, both <code>is_copyable</code> and <a href="#MathOptInterface.is_set_by_optimize"><code>is_set_by_optimize</code></a> return <code>false</code> for the following attributes:</p><ul><li><a href="#MathOptInterface.ListOfOptimizerAttributesSet"><code>ListOfOptimizerAttributesSet</code></a>, <a href="#MathOptInterface.ListOfModelAttributesSet"><code>ListOfModelAttributesSet</code></a>, <a href="#MathOptInterface.ListOfConstraintAttributesSet"><code>ListOfConstraintAttributesSet</code></a> and <a href="#MathOptInterface.ListOfVariableAttributesSet"><code>ListOfVariableAttributesSet</code></a>.</li><li><a href="#MathOptInterface.SolverName"><code>SolverName</code></a> and <a href="#MathOptInterface.RawSolver"><code>RawSolver</code></a>: these attributes cannot be set.</li><li><a href="#MathOptInterface.NumberOfVariables"><code>NumberOfVariables</code></a> and <a href="#MathOptInterface.ListOfVariableIndices"><code>ListOfVariableIndices</code></a>: these attributes are set indirectly by <a href="../variables/#MathOptInterface.add_variable"><code>add_variable</code></a> and <a href="../variables/#MathOptInterface.add_variables"><code>add_variables</code></a>.</li><li><a href="#MathOptInterface.ObjectiveFunctionType"><code>ObjectiveFunctionType</code></a>: this attribute is set indirectly when setting the <a href="#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> attribute.</li><li><a href="#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints</code></a>, <a href="#MathOptInterface.ListOfConstraintIndices"><code>ListOfConstraintIndices</code></a>, <a href="#MathOptInterface.ListOfConstraintTypesPresent"><code>ListOfConstraintTypesPresent</code></a>, <a href="../constraints/#MathOptInterface.CanonicalConstraintFunction"><code>CanonicalConstraintFunction</code></a>, <a href="../constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> and <a href="../constraints/#MathOptInterface.ConstraintSet"><code>ConstraintSet</code></a>: these attributes are set indirectly by <a href="../constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a> and <a href="../constraints/#MathOptInterface.add_constraints"><code>add_constraints</code></a>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2200-L2227">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.get" href="#MathOptInterface.get"><code>MathOptInterface.get</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">get(optimizer::AbstractOptimizer, attr::AbstractOptimizerAttribute)</code></pre><p>Return an attribute <code>attr</code> of the optimizer <code>optimizer</code>.</p><pre><code class="language-julia hljs">get(model::ModelLike, attr::AbstractModelAttribute)</code></pre><p>Return an attribute <code>attr</code> of the model <code>model</code>.</p><pre><code class="language-julia hljs">get(model::ModelLike, attr::AbstractVariableAttribute, v::VariableIndex)</code></pre><p>If the attribute <code>attr</code> is set for the variable <code>v</code> in the model <code>model</code>, return its value, return <code>nothing</code> otherwise. If the attribute <code>attr</code> is not supported by <code>model</code> then an error should be thrown instead of returning <code>nothing</code>.</p><pre><code class="language-julia hljs">get(model::ModelLike, attr::AbstractVariableAttribute, v::Vector{VariableIndex})</code></pre><p>Return a vector of attributes corresponding to each variable in the collection <code>v</code> in the model <code>model</code>.</p><pre><code class="language-julia hljs">get(model::ModelLike, attr::AbstractConstraintAttribute, c::ConstraintIndex)</code></pre><p>If the attribute <code>attr</code> is set for the constraint <code>c</code> in the model <code>model</code>, return its value, return <code>nothing</code> otherwise. If the attribute <code>attr</code> is not supported by <code>model</code> then an error should be thrown instead of returning <code>nothing</code>.</p><pre><code class="language-julia hljs">get(
+    model::ModelLike,
+    attr::AbstractConstraintAttribute,
+    c::Vector{ConstraintIndex{F,S}},
+) where {F,S}</code></pre><p>Return a vector of attributes corresponding to each constraint in the collection <code>c</code> in the model <code>model</code>.</p><pre><code class="language-julia hljs">get(model::ModelLike, ::Type{VariableIndex}, name::String)</code></pre><p>If a variable with name <code>name</code> exists in the model <code>model</code>, return the corresponding index, otherwise return <code>nothing</code>. Errors if two variables have the same name.</p><pre><code class="language-julia hljs">get(
+    model::ModelLike,
+    ::Type{ConstraintIndex{F,S}},
+    name::String,
+) where {F,S}</code></pre><p>If an <code>F</code>-in-<code>S</code> constraint with name <code>name</code> exists in the model <code>model</code>, return the corresponding index, otherwise return <code>nothing</code>. Errors if two constraints have the same name.</p><pre><code class="language-julia hljs">get(model::ModelLike, ::Type{ConstraintIndex}, name::String)</code></pre><p>If <em>any</em> constraint with name <code>name</code> exists in the model <code>model</code>, return the corresponding index, otherwise return <code>nothing</code>. This version is available for convenience but may incur a performance penalty because it is not type stable. Errors if two constraints have the same name.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L314-L388">source</a></section><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.NumberOfVariables)::Int64</code></pre><p>Return the number of variables created by the bridge <code>b</code> in the model.</p><p>See also <a href="#MathOptInterface.NumberOfConstraints"><code>MOI.NumberOfConstraints</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this if the bridge adds new variables.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L42-L53">source</a></section><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.ListOfVariableIndices)</code></pre><p>Return the list of variables created by the bridge <code>b</code>.</p><p>See also <a href="#MathOptInterface.ListOfVariableIndices"><code>MOI.ListOfVariableIndices</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this if the bridge adds new variables.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L56-L67">source</a></section><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.NumberOfConstraints{F,S})::Int64 where {F,S}</code></pre><p>Return the number of constraints of the type <code>F</code>-in-<code>S</code> created by the bridge <code>b</code>.</p><p>See also <a href="#MathOptInterface.NumberOfConstraints"><code>MOI.NumberOfConstraints</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this for the constraint types returned by <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.added_constraint_types"><code>added_constraint_types</code></a>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L70-L82">source</a></section><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.ListOfConstraintIndices{F,S}) where {F,S}</code></pre><p>Return a  <code>Vector{ConstraintIndex{F,S}}</code> with indices of all constraints of type <code>F</code>-in-<code>S</code> created by the bride <code>b</code>.</p><p>See also <a href="#MathOptInterface.ListOfConstraintIndices"><code>MOI.ListOfConstraintIndices</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this for the constraint types returned by <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.added_constraint_types"><code>added_constraint_types</code></a>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L85-L97">source</a></section><section><div><pre><code class="language-julia hljs">function MOI.get(
+    model::MOI.ModelLike,
+    attr::MOI.AbstractConstraintAttribute,
+    bridge::AbstractBridge,
+)</code></pre><p>Return the value of the attribute <code>attr</code> of the model <code>model</code> for the constraint bridged by <code>bridge</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L122-L131">source</a></section><section><div><pre><code class="language-julia hljs">get(model::GenericModel, attr::MathOptInterface.AbstractOptimizerAttribute)</code></pre><p>Return the value of the attribute <code>attr</code> from the model&#39;s MOI backend.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L624-L628">source</a></section><section><div><pre><code class="language-julia hljs">get(model::GenericModel, attr::MathOptInterface.AbstractModelAttribute)</code></pre><p>Return the value of the attribute <code>attr</code> from the model&#39;s MOI backend.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L666-L670">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.get!" href="#MathOptInterface.get!"><code>MathOptInterface.get!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">get!(output, model::ModelLike, args...)</code></pre><p>An in-place version of <a href="#MathOptInterface.get"><code>get</code></a>.</p><p>The signature matches that of <a href="#MathOptInterface.get"><code>get</code></a> except that the the result is placed in the vector <code>output</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L449-L456">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.set" href="#MathOptInterface.set"><code>MathOptInterface.set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set(optimizer::AbstractOptimizer, attr::AbstractOptimizerAttribute, value)</code></pre><p>Assign <code>value</code> to the attribute <code>attr</code> of the optimizer <code>optimizer</code>.</p><pre><code class="language-julia hljs">set(model::ModelLike, attr::AbstractModelAttribute, value)</code></pre><p>Assign <code>value</code> to the attribute <code>attr</code> of the model <code>model</code>.</p><pre><code class="language-julia hljs">set(model::ModelLike, attr::AbstractVariableAttribute, v::VariableIndex, value)</code></pre><p>Assign <code>value</code> to the attribute <code>attr</code> of variable <code>v</code> in model <code>model</code>.</p><pre><code class="language-julia hljs">set(
+    model::ModelLike,
+    attr::AbstractVariableAttribute,
+    v::Vector{VariableIndex},
+    vector_of_values,
+)</code></pre><p>Assign a value respectively to the attribute <code>attr</code> of each variable in the collection <code>v</code> in model <code>model</code>.</p><pre><code class="language-julia hljs">set(
+    model::ModelLike,
+    attr::AbstractConstraintAttribute,
+    c::ConstraintIndex,
+    value,
+)</code></pre><p>Assign a value to the attribute <code>attr</code> of constraint <code>c</code> in model <code>model</code>.</p><pre><code class="language-julia hljs">set(
+    model::ModelLike,
+    attr::AbstractConstraintAttribute,
+    c::Vector{ConstraintIndex{F,S}},
+    vector_of_values,
+) where {F,S}</code></pre><p>Assign a value respectively to the attribute <code>attr</code> of each constraint in the collection <code>c</code> in model <code>model</code>.</p><p>An <a href="../errors/#MathOptInterface.UnsupportedAttribute"><code>UnsupportedAttribute</code></a> error is thrown if <code>model</code> does not support the attribute <code>attr</code> (see <a href="#MathOptInterface.supports"><code>supports</code></a>) and a <a href="../errors/#MathOptInterface.SetAttributeNotAllowed"><code>SetAttributeNotAllowed</code></a> error is thrown if it supports the attribute <code>attr</code> but it cannot be set.</p><pre><code class="language-julia hljs">set(
+    model::ModelLike,
+    ::ConstraintSet,
+    c::ConstraintIndex{F,S},
+    set::S,
+) where {F,S}</code></pre><p>Change the set of constraint <code>c</code> to the new set <code>set</code> which should be of the same type as the original set.</p><pre><code class="language-julia hljs">set(
+    model::ModelLike,
+    ::ConstraintFunction,
+    c::ConstraintIndex{F,S},
+    func::F,
+) where {F,S}</code></pre><p>Replace the function in constraint <code>c</code> with <code>func</code>. <code>F</code> must match the original function type used to define the constraint.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Setting the constraint function is not allowed if <code>F</code> is <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>; a <a href="../errors/#MathOptInterface.SettingVariableIndexNotAllowed"><code>SettingVariableIndexNotAllowed</code></a> error is thrown instead. This is because, it would require changing the index <code>c</code> since the index of <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraints must be the same as the index of the variable.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L462-L548">source</a></section><section><div><pre><code class="language-julia hljs">function MOI.set(
+    model::MOI.ModelLike,
+    attr::MOI.AbstractConstraintAttribute,
+    bridge::AbstractBridge,
+    value,
+)</code></pre><p>Set the value of the attribute <code>attr</code> of the model <code>model</code> for the constraint bridged by <code>bridge</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L161-L171">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.supports" href="#MathOptInterface.supports"><code>MathOptInterface.supports</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports(model::ModelLike, sub::AbstractSubmittable)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports the submittable <code>sub</code>.</p><pre><code class="language-julia hljs">supports(model::ModelLike, attr::AbstractOptimizerAttribute)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports the optimizer attribute <code>attr</code>. That is, it returns <code>false</code> if <code>copy_to(model, src)</code> shows a warning in case <code>attr</code> is in the <a href="#MathOptInterface.ListOfOptimizerAttributesSet"><code>ListOfOptimizerAttributesSet</code></a> of <code>src</code>; see <a href="#MathOptInterface.copy_to"><code>copy_to</code></a> for more details on how unsupported optimizer attributes are handled in copy.</p><pre><code class="language-julia hljs">supports(model::ModelLike, attr::AbstractModelAttribute)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports the model attribute <code>attr</code>. That is, it returns <code>false</code> if <code>copy_to(model, src)</code> cannot be performed in case <code>attr</code> is in the <a href="#MathOptInterface.ListOfModelAttributesSet"><code>ListOfModelAttributesSet</code></a> of <code>src</code>.</p><pre><code class="language-julia hljs">supports(
+    model::ModelLike,
+    attr::AbstractVariableAttribute,
+    ::Type{VariableIndex},
+)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports the variable attribute <code>attr</code>. That is, it returns <code>false</code> if <code>copy_to(model, src)</code> cannot be performed in case <code>attr</code> is in the <a href="#MathOptInterface.ListOfVariableAttributesSet"><code>ListOfVariableAttributesSet</code></a> of <code>src</code>.</p><pre><code class="language-julia hljs">supports(
+    model::ModelLike,
+    attr::AbstractConstraintAttribute,
+    ::Type{ConstraintIndex{F,S}},
+)::Bool where {F,S}</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports the constraint attribute <code>attr</code> applied to an <code>F</code>-in-<code>S</code> constraint. That is, it returns <code>false</code> if <code>copy_to(model, src)</code> cannot be performed in case <code>attr</code> is in the <a href="#MathOptInterface.ListOfConstraintAttributesSet"><code>ListOfConstraintAttributesSet</code></a> of <code>src</code>.</p><p>For all five methods, if the attribute is only not supported in specific circumstances, it should still return <code>true</code>.</p><p>Note that <code>supports</code> is only defined for attributes for which <a href="#MathOptInterface.is_copyable"><code>is_copyable</code></a> returns <code>true</code> as other attributes do not appear in the list of attributes set obtained by <code>ListOf...AttributesSet</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L220-L274">source</a></section><section><div><pre><code class="language-julia hljs">MOI.supports(
+    model::MOI.ModelLike,
+    attr::MOI.AbstractConstraintAttribute,
+    BT::Type{&lt;:AbstractBridge},
+)</code></pre><p>Return a <code>Bool</code> indicating whether <code>BT</code> supports setting <code>attr</code> to <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L105-L113">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.attribute_value_type" href="#MathOptInterface.attribute_value_type"><code>MathOptInterface.attribute_value_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">attribute_value_type(attr::AnyAttribute)</code></pre><p>Given an attribute <code>attr</code>, return the type of value expected by <a href="#MathOptInterface.get"><code>get</code></a>, or returned by <a href="#MathOptInterface.set"><code>set</code></a>.</p><p><strong>Notes</strong></p><ul><li>Only implement this if it make sense to do so. If un-implemented, the default is <code>Any</code>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L59-L69">source</a></section></article><h2 id="Model-interface"><a class="docs-heading-anchor" href="#Model-interface">Model interface</a><a id="Model-interface-1"></a><a class="docs-heading-anchor-permalink" href="#Model-interface" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ModelLike" href="#MathOptInterface.ModelLike"><code>MathOptInterface.ModelLike</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ModelLike</code></pre><p>Abstract supertype for objects that implement the &quot;Model&quot; interface for defining an optimization problem.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L11-L16">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.is_empty" href="#MathOptInterface.is_empty"><code>MathOptInterface.is_empty</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_empty(model::ModelLike)</code></pre><p>Returns <code>false</code> if the <code>model</code> has any model attribute set or has any variables or constraints.</p><p>Note that an empty model can have optimizer attributes set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L155-L162">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.empty!" href="#MathOptInterface.empty!"><code>MathOptInterface.empty!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">empty!(model::ModelLike)</code></pre><p>Empty the model, that is, remove all variables, constraints and model attributes but not optimizer attributes.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L165-L170">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.write_to_file" href="#MathOptInterface.write_to_file"><code>MathOptInterface.write_to_file</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">write_to_file(model::ModelLike, filename::String)</code></pre><p>Write the current model to the file at <code>filename</code>.</p><p>Supported file types depend on the model type.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L116-L122">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.read_from_file" href="#MathOptInterface.read_from_file"><code>MathOptInterface.read_from_file</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">read_from_file(model::ModelLike, filename::String)</code></pre><p>Read the file <code>filename</code> into the model <code>model</code>. If <code>model</code> is non-empty, this may throw an error.</p><p>Supported file types depend on the model type.</p><p><strong>Note</strong></p><p>Once the contents of the file are loaded into the model, users can query the variables via <code>get(model, ListOfVariableIndices())</code>. However, some filetypes, such as LP files, do not maintain an explicit ordering of the variables. Therefore, the returned list may be in an arbitrary order.</p><p>To avoid depending on the order of the indices, look up each variable index by name using  <code>get(model, VariableIndex, &quot;name&quot;)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L130-L147">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.supports_incremental_interface" href="#MathOptInterface.supports_incremental_interface"><code>MathOptInterface.supports_incremental_interface</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_incremental_interface(model::ModelLike)</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports building incrementally via <a href="../variables/#MathOptInterface.add_variable"><code>add_variable</code></a> and <a href="../constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a>.</p><p>The main purpose of this function is to determine whether a model can be loaded into <code>model</code> incrementally or whether it should be cached and copied at once instead.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L173-L182">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.copy_to" href="#MathOptInterface.copy_to"><code>MathOptInterface.copy_to</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">copy_to(dest::ModelLike, src::ModelLike)::IndexMap</code></pre><p>Copy the model from <code>src</code> into <code>dest</code>.</p><p>The target <code>dest</code> is emptied, and all previous indices to variables and constraints in <code>dest</code> are invalidated.</p><p>Returns an <a href="#MathOptInterface.IndexMap"><code>IndexMap</code></a> object that translates variable and constraint indices from the <code>src</code> model to the corresponding indices in the <code>dest</code> model.</p><p><strong>Notes</strong></p><ul><li>If a constraint that in <code>src</code> is not supported by <code>dest</code>, then an <a href="../errors/#MathOptInterface.UnsupportedConstraint"><code>UnsupportedConstraint</code></a> error is thrown.</li><li>If an <a href="#MathOptInterface.AbstractModelAttribute"><code>AbstractModelAttribute</code></a>, <a href="../variables/#MathOptInterface.AbstractVariableAttribute"><code>AbstractVariableAttribute</code></a>, or <a href="../constraints/#MathOptInterface.AbstractConstraintAttribute"><code>AbstractConstraintAttribute</code></a> is set in <code>src</code> but not supported by <code>dest</code>, then an <a href="../errors/#MathOptInterface.UnsupportedAttribute"><code>UnsupportedAttribute</code></a> error is thrown.</li></ul><p><a href="#MathOptInterface.AbstractOptimizerAttribute"><code>AbstractOptimizerAttribute</code></a>s are <em>not</em> copied  to the <code>dest</code> model.</p><p><strong>IndexMap</strong></p><p>Implementations of <code>copy_to</code> must return an <a href="#MathOptInterface.IndexMap"><code>IndexMap</code></a>. For technical reasons, this type is defined in the Utilities submodule as <code>MOI.Utilities.IndexMap</code>. However, since it is an integral part of the MOI API, we provide <code>MOI.IndexMap</code> as an alias.</p><p><strong>Example</strong></p><pre><code class="language-julia hljs"># Given empty `ModelLike` objects `src` and `dest`.
+
+x = add_variable(src)
+
+is_valid(src, x)   # true
+is_valid(dest, x)  # false (`dest` has no variables)
+
+index_map = copy_to(dest, src)
+is_valid(dest, x) # false (unless index_map[x] == x)
+is_valid(dest, index_map[x]) # true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L185-L227">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.IndexMap" href="#MathOptInterface.IndexMap"><code>MathOptInterface.IndexMap</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndexMap()</code></pre><p>The dictionary-like object returned by <a href="#MathOptInterface.copy_to"><code>copy_to</code></a>.</p><p><strong>IndexMap</strong></p><p>Implementations of <a href="#MathOptInterface.copy_to"><code>copy_to</code></a> must return an <a href="#MathOptInterface.IndexMap"><code>IndexMap</code></a>. For technical reasons, the <code>IndexMap</code> type is defined in the Utilities submodule as <code>MOI.Utilities.IndexMap</code>. However, since it is an integral part of the MOI API, we provide this <code>MOI.IndexMap</code> as an alias.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L344-L355">source</a></section></article><h2 id="Model-attributes"><a class="docs-heading-anchor" href="#Model-attributes">Model attributes</a><a id="Model-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Model-attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractModelAttribute" href="#MathOptInterface.AbstractModelAttribute"><code>MathOptInterface.AbstractModelAttribute</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractModelAttribute</code></pre><p>Abstract supertype for attribute objects that can be used to set or get attributes (properties) of the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L22-L27">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Name" href="#MathOptInterface.Name"><code>MathOptInterface.Name</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Name()</code></pre><p>A model attribute for the string identifying the model. It has a default value of <code>&quot;&quot;</code> if not set`.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1088-L1093">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ObjectiveFunction" href="#MathOptInterface.ObjectiveFunction"><code>MathOptInterface.ObjectiveFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveFunction{F&lt;:AbstractScalarFunction}()</code></pre><p>A model attribute for the objective function which has a type <code>F&lt;:AbstractScalarFunction</code>.</p><p><code>F</code> should be guaranteed to be equivalent but not necessarily identical to the function type provided by the user.</p><p>Throws an <code>InexactError</code> if the objective function cannot be converted to <code>F</code>, e.g., the objective function is quadratic and <code>F</code> is <code>ScalarAffineFunction{Float64}</code> or it has non-integer coefficient and <code>F</code> is <code>ScalarAffineFunction{Int}</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1206-L1219">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ObjectiveFunctionType" href="#MathOptInterface.ObjectiveFunctionType"><code>MathOptInterface.ObjectiveFunctionType</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveFunctionType()</code></pre><p>A model attribute for the type <code>F</code> of the objective function set using the <a href="#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction{F}</code></a> attribute.</p><p><strong>Examples</strong></p><p>In the following code, <code>attr</code> should be equal to <code>MOI.VariableIndex</code>:</p><pre><code class="language-julia hljs">x = MOI.add_variable(model)
+MOI.set(model, MOI.ObjectiveFunction{MOI.VariableIndex}(), x)
+attr = MOI.get(model, MOI.ObjectiveFunctionType())</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1224-L1238">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ObjectiveSense" href="#MathOptInterface.ObjectiveSense"><code>MathOptInterface.ObjectiveSense</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveSense()</code></pre><p>A model attribute for the objective sense of the objective function, which must be an <a href="#MathOptInterface.OptimizationSense"><code>OptimizationSense</code></a>: <code>MIN_SENSE</code>, <code>MAX_SENSE</code>, or <code>FEASIBILITY_SENSE</code>. The default is <code>FEASIBILITY_SENSE</code>.</p><p><strong>Interaction with <code>ObjectiveFunction</code></strong></p><p>Setting the sense to <code>FEASIBILITY_SENSE</code> unsets the <a href="#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> attribute. That is, if you first set <a href="#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> and then set <code>ObjectiveSense</code> to be <code>FEASIBILITY_SENSE</code>, no objective function will be passed to the solver.</p><p>In addition, some reformulations of <a href="#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> via bridges rely on the value of <code>ObjectiveSense</code>. Therefore, you should set <code>ObjectiveSense</code> before setting <a href="#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1113-L1130">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OptimizationSense" href="#MathOptInterface.OptimizationSense"><code>MathOptInterface.OptimizationSense</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">OptimizationSense</code></pre><p>An enum for the value of the <a href="#MathOptInterface.ObjectiveSense"><code>ObjectiveSense</code></a> attribute.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="#MathOptInterface.MIN_SENSE"><code>MIN_SENSE</code></a>: the goal is to minimize the objective function</li><li><a href="#MathOptInterface.MAX_SENSE"><code>MAX_SENSE</code></a>: the goal is to maximize the objective function</li><li><a href="#MathOptInterface.FEASIBILITY_SENSE"><code>FEASIBILITY_SENSE</code></a>: the model does not have an objective function</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1098-L1110">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.MIN_SENSE" href="#MathOptInterface.MIN_SENSE"><code>MathOptInterface.MIN_SENSE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">MIN_SENSE::OptimizationSense</code></pre><p>An instance of the <a href="#MathOptInterface.OptimizationSense"><code>OptimizationSense</code></a> enum.</p><p><code>MIN_SENSE</code>: the goal is to minimize the objective function</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1098-L1104">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.MAX_SENSE" href="#MathOptInterface.MAX_SENSE"><code>MathOptInterface.MAX_SENSE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">MAX_SENSE::OptimizationSense</code></pre><p>An instance of the <a href="#MathOptInterface.OptimizationSense"><code>OptimizationSense</code></a> enum.</p><p><code>MAX_SENSE</code>: the goal is to maximize the objective function</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1098-L1104">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FEASIBILITY_SENSE" href="#MathOptInterface.FEASIBILITY_SENSE"><code>MathOptInterface.FEASIBILITY_SENSE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">FEASIBILITY_SENSE::OptimizationSense</code></pre><p>An instance of the <a href="#MathOptInterface.OptimizationSense"><code>OptimizationSense</code></a> enum.</p><p><code>FEASIBILITY_SENSE</code>: the model does not have an objective function</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1098-L1104">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NumberOfVariables" href="#MathOptInterface.NumberOfVariables"><code>MathOptInterface.NumberOfVariables</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NumberOfVariables()</code></pre><p>A model attribute for the number of variables in the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1135-L1139">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfVariableIndices" href="#MathOptInterface.ListOfVariableIndices"><code>MathOptInterface.ListOfVariableIndices</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfVariableIndices()</code></pre><p>A model attribute for the <code>Vector{VariableIndex}</code> of all variable indices present in the model (i.e., of length equal to the value of <a href="#MathOptInterface.NumberOfVariables"><code>NumberOfVariables</code></a> in the order in which they were added.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1144-L1150">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfConstraintTypesPresent" href="#MathOptInterface.ListOfConstraintTypesPresent"><code>MathOptInterface.ListOfConstraintTypesPresent</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfConstraintTypesPresent()</code></pre><p>A model attribute for the list of tuples of the form <code>(F,S)</code>, where <code>F</code> is a function type and <code>S</code> is a set type indicating that the attribute <a href="#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints{F,S}</code></a> has a value greater than zero.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1197-L1203">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NumberOfConstraints" href="#MathOptInterface.NumberOfConstraints"><code>MathOptInterface.NumberOfConstraints</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NumberOfConstraints{F,S}()</code></pre><p>A model attribute for the number of constraints of the type <code>F</code>-in-<code>S</code> present in the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1176-L1181">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfConstraintIndices" href="#MathOptInterface.ListOfConstraintIndices"><code>MathOptInterface.ListOfConstraintIndices</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfConstraintIndices{F,S}()</code></pre><p>A model attribute for the <code>Vector{ConstraintIndex{F,S}}</code> of all constraint indices of type <code>F</code>-in-<code>S</code> in the model (i.e., of length equal to the value of <a href="#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints{F,S}</code></a>) in the order in which they were added.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1153-L1159">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfOptimizerAttributesSet" href="#MathOptInterface.ListOfOptimizerAttributesSet"><code>MathOptInterface.ListOfOptimizerAttributesSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfOptimizerAttributesSet()</code></pre><p>An optimizer attribute for the <code>Vector{AbstractOptimizerAttribute}</code> of all optimizer attributes that were set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L783-L788">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfModelAttributesSet" href="#MathOptInterface.ListOfModelAttributesSet"><code>MathOptInterface.ListOfModelAttributesSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfModelAttributesSet()</code></pre><p>A model attribute for the <code>Vector{AbstractModelAttribute}</code> of all model attributes <code>attr</code> such that:</p><ol><li><code>is_copyable(attr)</code> returns <code>true</code>, and</li><li>the attribute was set to the model</li></ol></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1077-L1085">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfVariableAttributesSet" href="#MathOptInterface.ListOfVariableAttributesSet"><code>MathOptInterface.ListOfVariableAttributesSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfVariableAttributesSet()</code></pre><p>A model attribute for the <code>Vector{AbstractVariableAttribute}</code> of all variable attributes <code>attr</code> such that 1) <code>is_copyable(attr)</code> returns <code>true</code> and 2) the attribute was set to variables.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1420-L1426">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfConstraintAttributesSet" href="#MathOptInterface.ListOfConstraintAttributesSet"><code>MathOptInterface.ListOfConstraintAttributesSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfConstraintAttributesSet{F, S}()</code></pre><p>A model attribute for the <code>Vector{AbstractConstraintAttribute}</code> of all constraint attributes <code>attr</code> such that:</p><ol><li><code>is_copyable(attr)</code> returns <code>true</code> and</li><li>the attribute was set to <code>F</code>-in-<code>S</code> constraints.</li></ol><p><strong>Note</strong></p><p>The attributes <a href="../constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> and <a href="../constraints/#MathOptInterface.ConstraintSet"><code>ConstraintSet</code></a> should not be included in the list even if then have been set with <a href="#MathOptInterface.set"><code>set</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1539-L1552">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UserDefinedFunction" href="#MathOptInterface.UserDefinedFunction"><code>MathOptInterface.UserDefinedFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UserDefinedFunction(name::Symbol, arity::Int) &lt;: AbstractModelAttribute</code></pre><p>Set this attribute to register a user-defined function by the name of <code>name</code> with <code>arity</code> arguments.</p><p>Once registered, <code>name</code> will appear in <a href="#MathOptInterface.ListOfSupportedNonlinearOperators"><code>ListOfSupportedNonlinearOperators</code></a>.</p><p>You cannot register multiple <code>UserDefinedFunction</code>s with the same <code>name</code> but different <code>arity</code>.</p><p><strong>Value type</strong></p><p>The value to be set is a tuple containing one, two, or three functions to evaluate the function, the first-order derivative, and the second-order derivative respectively. Both derivatives are optional, but if you pass the second-order derivative you must also pass the first-order derivative.</p><p>For univariate functions with <code>arity == 1</code>, the functions in the tuple must have the form:</p><ul><li><code>f(x::T)::T</code>: returns the value of the function at <code>x</code></li><li><code>∇f(x::T)::T</code>: returns the first-order derivative of <code>f</code> with respect to <code>x</code></li><li><code>∇²f(x::T)::T</code>: returns the second-order derivative of <code>f</code> with respect to <code>x</code>.</li></ul><p>For multivariate functions with <code>arity &gt; 1</code>, the functions in the tuple must have the form:</p><ul><li><code>f(x::T...)::T</code>: returns the value of the function at <code>x</code></li><li><code>∇f(g::AbstractVector{T}, x::T...)::Nothing</code>: fills the components of <code>g</code>, with <code>g[i]</code> being the first-order partial derivative of <code>f</code> with respect to <code>x[i]</code></li><li><code>∇²f(H::AbstractMatrix{T}, x::T...)::Nothing</code>: fills the non-zero components  of <code>H</code>, with <code>H[i, j]</code> being the second-order partial derivative of <code>f</code> with  respect to <code>x[i]</code> and then <code>x[j]</code>. <code>H</code> is initialized to the zero matrix,  so you do not need to set any zero elements.</li></ul><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; f(x, y) = x^2 + y^2
+f (generic function with 1 method)
+
+julia&gt; function ∇f(g, x, y)
+           g .= 2 * x, 2 * y
+           return
+       end
+∇f (generic function with 1 method)
+
+julia&gt; function ∇²f(H, x...)
+           H[1, 1] = H[2, 2] = 2.0
+           return
+       end
+∇²f (generic function with 1 method)
+
+julia&gt; model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())
+MOIU.UniversalFallback{MOIU.Model{Float64}}
+fallback for MOIU.Model{Float64}
+
+julia&gt; MOI.set(model, MOI.UserDefinedFunction(:f, 2), (f,))
+
+julia&gt; MOI.set(model, MOI.UserDefinedFunction(:g, 2), (f, ∇f))
+
+julia&gt; MOI.set(model, MOI.UserDefinedFunction(:h, 2), (f, ∇f, ∇²f))
+
+julia&gt; x = MOI.add_variables(model, 2)
+2-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+
+julia&gt; MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)
+
+julia&gt; obj_f = MOI.ScalarNonlinearFunction(:f, Any[x[1], x[2]])
+f(MOI.VariableIndex(1), MOI.VariableIndex(2))
+
+julia&gt; MOI.set(model, MOI.ObjectiveFunction{typeof(obj_f)}(), obj_f)
+
+julia&gt; print(model)
+Minimize ScalarNonlinearFunction:
+ f(v[1], v[2])
+
+Subject to:
+</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1872-L1959">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ListOfSupportedNonlinearOperators" href="#MathOptInterface.ListOfSupportedNonlinearOperators"><code>MathOptInterface.ListOfSupportedNonlinearOperators</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfSupportedNonlinearOperators() &lt;: AbstractModelAttribute</code></pre><p>When queried with <a href="#MathOptInterface.get"><code>get</code></a>, return a <code>Vector{Symbol}</code> listing the operators supported by the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1965-L1970">source</a></section></article><h2 id="Optimizer-interface"><a class="docs-heading-anchor" href="#Optimizer-interface">Optimizer interface</a><a id="Optimizer-interface-1"></a><a class="docs-heading-anchor-permalink" href="#Optimizer-interface" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractOptimizer" href="#MathOptInterface.AbstractOptimizer"><code>MathOptInterface.AbstractOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractOptimizer &lt;: ModelLike</code></pre><p>Abstract supertype for objects representing an instance of an optimization problem tied to a particular solver. This is typically a solver&#39;s in-memory representation. In addition to <code>ModelLike</code>, <code>AbstractOptimizer</code> objects let you solve the model and query the solution.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L27-L34">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OptimizerWithAttributes" href="#MathOptInterface.OptimizerWithAttributes"><code>MathOptInterface.OptimizerWithAttributes</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct OptimizerWithAttributes
+    optimizer_constructor
+    params::Vector{Pair{AbstractOptimizerAttribute,&lt;:Any}}
+end</code></pre><p>Object grouping an optimizer constructor and a list of optimizer attributes. Instances are created with <a href="#MathOptInterface.instantiate"><code>instantiate</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/instantiate.jl#L7-L15">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.optimize!" href="#MathOptInterface.optimize!"><code>MathOptInterface.optimize!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">optimize!(optimizer::AbstractOptimizer)</code></pre><p>Optimize the problem contained in <code>optimizer</code>.</p><p>Before calling <code>optimize!</code>, the problem should first be constructed using the incremental interface (see <a href="#MathOptInterface.supports_incremental_interface"><code>supports_incremental_interface</code></a>) or <a href="#MathOptInterface.copy_to"><code>copy_to</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L37-L44">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.optimize!-Tuple{MathOptInterface.ModelLike, MathOptInterface.ModelLike}" href="#MathOptInterface.optimize!-Tuple{MathOptInterface.ModelLike, MathOptInterface.ModelLike}"><code>MathOptInterface.optimize!</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">optimize!(dest::AbstractOptimizer, src::ModelLike)::Tuple{IndexMap,Bool}</code></pre><p>A &quot;one-shot&quot; call that copies the problem from <code>src</code> into <code>dest</code> and then uses <code>dest</code> to optimize the problem.</p><p>Returns a tuple of an <a href="#MathOptInterface.IndexMap"><code>IndexMap</code></a> and a <code>Bool</code> <code>copied</code>.</p><ul><li>The <a href="#MathOptInterface.IndexMap"><code>IndexMap</code></a> object translates variable and constraint indices from the <code>src</code> model to the corresponding indices in the <code>dest</code> optimizer. See <a href="#MathOptInterface.copy_to"><code>copy_to</code></a> for details.</li><li>If <code>copied == true</code>, <code>src</code> was copied to <code>dest</code> and then cached, allowing incremental modification if supported by the solver.</li><li>If <code>copied == false</code>, a cache of the model was <em>not</em> kept in <code>dest</code>. Therefore, only the solution information (attributes for which <a href="#MathOptInterface.is_set_by_optimize"><code>is_set_by_optimize</code></a> is true) is available to query.</li></ul><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The main purpose of <code>optimize!</code> method with two arguments is for use in <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a>.</p></div></div><p><strong>Relationship to the single-argument <code>optimize!</code></strong></p><p>The default fallback of <code>optimize!(dest::AbstractOptimizer, src::ModelLike)</code> is</p><pre><code class="language-julia hljs">function optimize!(dest::AbstractOptimizer, src::ModelLike)
+    index_map = copy_to(dest, src)
+    optimize!(dest)
+    return index_map, true
+end</code></pre><p>Therefore, subtypes of <a href="#MathOptInterface.AbstractOptimizer"><code>AbstractOptimizer</code></a> should either implement this two-argument method, or implement both <a href="#MathOptInterface.copy_to"><code>copy_to(::Optimizer, ::ModelLike)</code></a> and <code>optimize!(::Optimizer)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L47-L81">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.instantiate" href="#MathOptInterface.instantiate"><code>MathOptInterface.instantiate</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">instantiate(
+    optimizer_constructor,
+    with_cache_type::Union{Nothing,Type} = nothing,
+    with_bridge_type::Union{Nothing,Type} = nothing,
+)</code></pre><p>Create an instance of an optimizer by either:</p><ul><li>calling <code>optimizer_constructor.optimizer_constructor()</code> and setting the parameters in <code>optimizer_constructor.params</code> if <code>optimizer_constructor</code> is a <a href="#MathOptInterface.OptimizerWithAttributes"><code>OptimizerWithAttributes</code></a></li><li>calling <code>optimizer_constructor()</code> if <code>optimizer_constructor</code> is callable.</li></ul><p><strong>with<em>cache</em>type</strong></p><p>If <code>with_cache_type</code> is not <code>nothing</code>, then the optimizer is wrapped in a <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a> to store a cache of the model. This is most useful if the optimizer you are constructing does not support the incremental interface (see <a href="#MathOptInterface.supports_incremental_interface"><code>supports_incremental_interface</code></a>).</p><p><strong>with<em>bridge</em>type</strong></p><p>If <code>with_bridge_type</code> is not <code>nothing</code>, the optimizer is wrapped in a <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>Bridges.full_bridge_optimizer</code></a>, enabling all the bridges defined in the MOI.Bridges submodule with coefficient type <code>with_bridge_type</code>.</p><p>In addition, if the optimizer created by <code>optimizer_constructor</code> does not support the incremental interface (see <a href="#MathOptInterface.supports_incremental_interface"><code>supports_incremental_interface</code></a>), then, irrespective of <code>with_cache_type</code>, the optimizer is wrapped in a <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a> to store a cache of the bridged model.</p><p>If <code>with_cache_type</code> and <code>with_bridge_type</code> are both not <code>nothing</code>, then they must be the same type.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/instantiate.jl#L125-L159">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.default_cache" href="#MathOptInterface.default_cache"><code>MathOptInterface.default_cache</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">default_cache(optimizer::ModelLike, ::Type{T}) where {T}</code></pre><p>Return a new instance of the default model type to be used as cache for <code>optimizer</code> in a <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a> for holding constraints of coefficient type <code>T</code>. By default, this returns <code>Utilities.UniversalFallback(Utilities.Model{T}())</code>. If copying from a instance of a given model type is faster for <code>optimizer</code> then a new method returning an instance of this model type should be defined.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/instantiate.jl#L192-L201">source</a></section></article><h2 id="Optimizer-attributes"><a class="docs-heading-anchor" href="#Optimizer-attributes">Optimizer attributes</a><a id="Optimizer-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Optimizer-attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractOptimizerAttribute" href="#MathOptInterface.AbstractOptimizerAttribute"><code>MathOptInterface.AbstractOptimizerAttribute</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractOptimizerAttribute</code></pre><p>Abstract supertype for attribute objects that can be used to set or get attributes (properties) of the optimizer.</p><p><strong>Notes</strong></p><p>The difference between <code>AbstractOptimizerAttribute</code> and <a href="#MathOptInterface.AbstractModelAttribute"><code>AbstractModelAttribute</code></a> lies in the behavior of <a href="#MathOptInterface.is_empty"><code>is_empty</code></a>, <a href="#MathOptInterface.empty!"><code>empty!</code></a> and <a href="#MathOptInterface.copy_to"><code>copy_to</code></a>. Typically optimizer attributes affect only how the model is solved.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L7-L19">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SolverName" href="#MathOptInterface.SolverName"><code>MathOptInterface.SolverName</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SolverName()</code></pre><p>An optimizer attribute for the string identifying the solver/optimizer.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L791-L795">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SolverVersion" href="#MathOptInterface.SolverVersion"><code>MathOptInterface.SolverVersion</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SolverVersion()</code></pre><p>An optimizer attribute for the string identifying the version of the solver.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>For solvers supporting <a href="https://semver.org">semantic versioning</a>, the <code>SolverVersion</code> should be a string of the form &quot;vMAJOR.MINOR.PATCH&quot;, so that it can be converted to a Julia <code>VersionNumber</code> (e.g., `VersionNumber(&quot;v1.2.3&quot;)).</p><p>We do not require Semantic Versioning because some solvers use alternate versioning systems. For example, CPLEX uses Calendar Versioning, so <code>SolverVersion</code> will return a string like <code>&quot;202001&quot;</code>.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L800-L814">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Silent" href="#MathOptInterface.Silent"><code>MathOptInterface.Silent</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Silent()</code></pre><p>An optimizer attribute for silencing the output of an optimizer. When <code>set</code> to <code>true</code>, it takes precedence over any other attribute controlling verbosity and requires the solver to produce no output. The default value is <code>false</code> which has no effect. In this case the verbosity is controlled by other attributes.</p><p><strong>Note</strong></p><p>Every optimizer should have verbosity on by default. For instance, if a solver has a solver-specific log level attribute, the MOI implementation should set it to <code>1</code> by default. If the user sets <code>Silent</code> to <code>true</code>, then the log level should be set to <code>0</code>, even if the user specifically sets a value of log level. If the value of <code>Silent</code> is <code>false</code> then the log level set to the solver is the value given by the user for this solver-specific parameter or <code>1</code> if none is given.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L819-L837">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.TimeLimitSec" href="#MathOptInterface.TimeLimitSec"><code>MathOptInterface.TimeLimitSec</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">TimeLimitSec()</code></pre><p>An optimizer attribute for setting a time limit (in seconnds) for an optimization. When <code>set</code> to <code>nothing</code>, it deactivates the solver time limit. The default value is <code>nothing</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L842-L848">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ObjectiveLimit" href="#MathOptInterface.ObjectiveLimit"><code>MathOptInterface.ObjectiveLimit</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveLimit()</code></pre><p>An optimizer attribute for setting a limit on the objective value.</p><p>The provided limit must be a <code>Union{Real,Nothing}</code>.</p><p>When <code>set</code> to <code>nothing</code>, the limit reverts to the solver&#39;s default.</p><p>The default value is <code>nothing</code>.</p><p>The solver may stop when the <a href="#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a> is better (lower for minimization, higher for maximization) than the <code>ObjectiveLimit</code>. If stopped, the <a href="#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a> should be <code>OBJECTIVE_LIMIT</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L853-L867">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RawOptimizerAttribute" href="#MathOptInterface.RawOptimizerAttribute"><code>MathOptInterface.RawOptimizerAttribute</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RawOptimizerAttribute(name::String)</code></pre><p>An optimizer attribute for the solver-specific parameter identified by <code>name</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L870-L874">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NumberOfThreads" href="#MathOptInterface.NumberOfThreads"><code>MathOptInterface.NumberOfThreads</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NumberOfThreads()</code></pre><p>An optimizer attribute for setting the number of threads used for an optimization. When set to <code>nothing</code> uses solver default. Values are positive integers. The default value is <code>nothing</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L879-L885">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RawSolver" href="#MathOptInterface.RawSolver"><code>MathOptInterface.RawSolver</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RawSolver()</code></pre><p>A model attribute for the object that may be used to access a solver-specific API for this optimizer.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1349-L1354">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbsoluteGapTolerance" href="#MathOptInterface.AbsoluteGapTolerance"><code>MathOptInterface.AbsoluteGapTolerance</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbsoluteGapTolerance()</code></pre><p>An optimizer attribute for setting the absolute gap tolerance for an optimization. This is an <em>optimizer</em> attribute, and should be set before calling <a href="#MathOptInterface.optimize!"><code>optimize!</code></a>. When set to <code>nothing</code> (if supported), uses solver default.</p><p>To set a <em>relative</em> gap tolerance, see <a href="#MathOptInterface.RelativeGapTolerance"><code>RelativeGapTolerance</code></a>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The mathematical definition of &quot;absolute gap&quot;, and its treatment during the optimization, are solver-dependent. However, assuming no other limit nor issue is encountered during the optimization, most solvers that implement this attribute will stop once <span>$|f - b| ≤ g_{abs}$</span>, where <span>$b$</span> is the best bound, <span>$f$</span> is the best feasible objective value, and <span>$g_{abs}$</span> is the absolute gap.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L924-L940">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RelativeGapTolerance" href="#MathOptInterface.RelativeGapTolerance"><code>MathOptInterface.RelativeGapTolerance</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RelativeGapTolerance()</code></pre><p>An optimizer attribute for setting the relative gap tolerance for an optimization. This is an <em>optimizer</em> attribute, and should be set before calling <a href="#MathOptInterface.optimize!"><code>optimize!</code></a>. When set to <code>nothing</code> (if supported), uses solver default.</p><p>If you are looking for the relative gap of the current best solution, see <a href="#MathOptInterface.RelativeGap"><code>RelativeGap</code></a>. If no limit nor issue is encountered during the optimization, the value of <a href="#MathOptInterface.RelativeGap"><code>RelativeGap</code></a> should be at most as large as <code>RelativeGapTolerance</code>.</p><pre><code class="language-julia hljs"># Before optimizing: set relative gap tolerance
+# set 0.1% relative gap tolerance
+MOI.set(model, MOI.RelativeGapTolerance(), 1e-3)
+MOI.optimize!(model)
+
+# After optimizing (assuming all went well)
+# The relative gap tolerance has not changed...
+MOI.get(model, MOI.RelativeGapTolerance())  # returns 1e-3
+# ... and the relative gap of the obtained solution is smaller or equal to the
+# tolerance
+MOI.get(model, MOI.RelativeGap())  # should return something ≤ 1e-3</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The mathematical definition of &quot;relative gap&quot;, and its allowed range, are solver-dependent. Typically, solvers expect a value between <code>0.0</code> and <code>1.0</code>.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L890-L919">source</a></section></article><p>List of attributes useful for optimizers</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.TerminationStatus" href="#MathOptInterface.TerminationStatus"><code>MathOptInterface.TerminationStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">TerminationStatus()</code></pre><p>A model attribute for the <code>TerminationStatusCode</code> explaining why the optimizer stopped.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2038-L2043">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.TerminationStatusCode" href="#MathOptInterface.TerminationStatusCode"><code>MathOptInterface.TerminationStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">TerminationStatusCode</code></pre><p>An Enum of possible values for the <code>TerminationStatus</code> attribute. This attribute is meant to explain the reason why the optimizer stopped executing in the most recent call to <a href="#MathOptInterface.optimize!"><code>optimize!</code></a>.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="#MathOptInterface.OPTIMIZE_NOT_CALLED"><code>OPTIMIZE_NOT_CALLED</code></a>: The algorithm has not started.</li><li><a href="#MathOptInterface.OPTIMAL"><code>OPTIMAL</code></a>: The algorithm found a globally optimal solution.</li><li><a href="#MathOptInterface.INFEASIBLE"><code>INFEASIBLE</code></a>: The algorithm concluded that no feasible solution exists.</li><li><a href="#MathOptInterface.DUAL_INFEASIBLE"><code>DUAL_INFEASIBLE</code></a>: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.</li><li><a href="#MathOptInterface.LOCALLY_SOLVED"><code>LOCALLY_SOLVED</code></a>: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.</li><li><a href="#MathOptInterface.LOCALLY_INFEASIBLE"><code>LOCALLY_INFEASIBLE</code></a>: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.</li><li><a href="#MathOptInterface.INFEASIBLE_OR_UNBOUNDED"><code>INFEASIBLE_OR_UNBOUNDED</code></a>: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.</li><li><a href="#MathOptInterface.ALMOST_OPTIMAL"><code>ALMOST_OPTIMAL</code></a>: The algorithm found a globally optimal solution to relaxed tolerances.</li><li><a href="#MathOptInterface.ALMOST_INFEASIBLE"><code>ALMOST_INFEASIBLE</code></a>: The algorithm concluded that no feasible solution exists within relaxed tolerances.</li><li><a href="#MathOptInterface.ALMOST_DUAL_INFEASIBLE"><code>ALMOST_DUAL_INFEASIBLE</code></a>: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.</li><li><a href="#MathOptInterface.ALMOST_LOCALLY_SOLVED"><code>ALMOST_LOCALLY_SOLVED</code></a>: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.</li><li><a href="#MathOptInterface.ITERATION_LIMIT"><code>ITERATION_LIMIT</code></a>: An iterative algorithm stopped after conducting the maximum number of iterations.</li><li><a href="#MathOptInterface.TIME_LIMIT"><code>TIME_LIMIT</code></a>: The algorithm stopped after a user-specified computation time.</li><li><a href="#MathOptInterface.NODE_LIMIT"><code>NODE_LIMIT</code></a>: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.</li><li><a href="#MathOptInterface.SOLUTION_LIMIT"><code>SOLUTION_LIMIT</code></a>: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.</li><li><a href="#MathOptInterface.MEMORY_LIMIT"><code>MEMORY_LIMIT</code></a>: The algorithm stopped because it ran out of memory.</li><li><a href="#MathOptInterface.OBJECTIVE_LIMIT"><code>OBJECTIVE_LIMIT</code></a>: The algorithm stopped because it found a solution better than a minimum limit set by the user.</li><li><a href="#MathOptInterface.NORM_LIMIT"><code>NORM_LIMIT</code></a>: The algorithm stopped because the norm of an iterate became too large.</li><li><a href="#MathOptInterface.OTHER_LIMIT"><code>OTHER_LIMIT</code></a>: The algorithm stopped due to a limit not covered by one of the <code>_LIMIT_</code> statuses above.</li><li><a href="#MathOptInterface.SLOW_PROGRESS"><code>SLOW_PROGRESS</code></a>: The algorithm stopped because it was unable to continue making progress towards the solution.</li><li><a href="#MathOptInterface.NUMERICAL_ERROR"><code>NUMERICAL_ERROR</code></a>: The algorithm stopped because it encountered unrecoverable numerical error.</li><li><a href="#MathOptInterface.INVALID_MODEL"><code>INVALID_MODEL</code></a>: The algorithm stopped because the model is invalid.</li><li><a href="#MathOptInterface.INVALID_OPTION"><code>INVALID_OPTION</code></a>: The algorithm stopped because it was provided an invalid option.</li><li><a href="#MathOptInterface.INTERRUPTED"><code>INTERRUPTED</code></a>: The algorithm stopped because of an interrupt signal.</li><li><a href="#MathOptInterface.OTHER_ERROR"><code>OTHER_ERROR</code></a>: The algorithm stopped because of an error not covered by one of the statuses defined above.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L2009">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OPTIMIZE_NOT_CALLED" href="#MathOptInterface.OPTIMIZE_NOT_CALLED"><code>MathOptInterface.OPTIMIZE_NOT_CALLED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OPTIMIZE_NOT_CALLED::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OPTIMIZE_NOT_CALLED</code>: The algorithm has not started.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OPTIMAL" href="#MathOptInterface.OPTIMAL"><code>MathOptInterface.OPTIMAL</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OPTIMAL::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OPTIMAL</code>: The algorithm found a globally optimal solution.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.INFEASIBLE" href="#MathOptInterface.INFEASIBLE"><code>MathOptInterface.INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INFEASIBLE</code>: The algorithm concluded that no feasible solution exists.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.DUAL_INFEASIBLE" href="#MathOptInterface.DUAL_INFEASIBLE"><code>MathOptInterface.DUAL_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">DUAL_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>DUAL_INFEASIBLE</code>: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LOCALLY_SOLVED" href="#MathOptInterface.LOCALLY_SOLVED"><code>MathOptInterface.LOCALLY_SOLVED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">LOCALLY_SOLVED::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>LOCALLY_SOLVED</code>: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LOCALLY_INFEASIBLE" href="#MathOptInterface.LOCALLY_INFEASIBLE"><code>MathOptInterface.LOCALLY_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">LOCALLY_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>LOCALLY_INFEASIBLE</code>: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.INFEASIBLE_OR_UNBOUNDED" href="#MathOptInterface.INFEASIBLE_OR_UNBOUNDED"><code>MathOptInterface.INFEASIBLE_OR_UNBOUNDED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBLE_OR_UNBOUNDED::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INFEASIBLE_OR_UNBOUNDED</code>: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ALMOST_OPTIMAL" href="#MathOptInterface.ALMOST_OPTIMAL"><code>MathOptInterface.ALMOST_OPTIMAL</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_OPTIMAL::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_OPTIMAL</code>: The algorithm found a globally optimal solution to relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ALMOST_INFEASIBLE" href="#MathOptInterface.ALMOST_INFEASIBLE"><code>MathOptInterface.ALMOST_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_INFEASIBLE</code>: The algorithm concluded that no feasible solution exists within relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ALMOST_DUAL_INFEASIBLE" href="#MathOptInterface.ALMOST_DUAL_INFEASIBLE"><code>MathOptInterface.ALMOST_DUAL_INFEASIBLE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_DUAL_INFEASIBLE::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_DUAL_INFEASIBLE</code>: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ALMOST_LOCALLY_SOLVED" href="#MathOptInterface.ALMOST_LOCALLY_SOLVED"><code>MathOptInterface.ALMOST_LOCALLY_SOLVED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ALMOST_LOCALLY_SOLVED::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ALMOST_LOCALLY_SOLVED</code>: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ITERATION_LIMIT" href="#MathOptInterface.ITERATION_LIMIT"><code>MathOptInterface.ITERATION_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">ITERATION_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>ITERATION_LIMIT</code>: An iterative algorithm stopped after conducting the maximum number of iterations.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.TIME_LIMIT" href="#MathOptInterface.TIME_LIMIT"><code>MathOptInterface.TIME_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">TIME_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>TIME_LIMIT</code>: The algorithm stopped after a user-specified computation time.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NODE_LIMIT" href="#MathOptInterface.NODE_LIMIT"><code>MathOptInterface.NODE_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NODE_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>NODE_LIMIT</code>: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SOLUTION_LIMIT" href="#MathOptInterface.SOLUTION_LIMIT"><code>MathOptInterface.SOLUTION_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">SOLUTION_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>SOLUTION_LIMIT</code>: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.MEMORY_LIMIT" href="#MathOptInterface.MEMORY_LIMIT"><code>MathOptInterface.MEMORY_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">MEMORY_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>MEMORY_LIMIT</code>: The algorithm stopped because it ran out of memory.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OBJECTIVE_LIMIT" href="#MathOptInterface.OBJECTIVE_LIMIT"><code>MathOptInterface.OBJECTIVE_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OBJECTIVE_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OBJECTIVE_LIMIT</code>: The algorithm stopped because it found a solution better than a minimum limit set by the user.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NORM_LIMIT" href="#MathOptInterface.NORM_LIMIT"><code>MathOptInterface.NORM_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NORM_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>NORM_LIMIT</code>: The algorithm stopped because the norm of an iterate became too large.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OTHER_LIMIT" href="#MathOptInterface.OTHER_LIMIT"><code>MathOptInterface.OTHER_LIMIT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OTHER_LIMIT::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OTHER_LIMIT</code>: The algorithm stopped due to a limit not covered by one of the <code>_LIMIT_</code> statuses above.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SLOW_PROGRESS" href="#MathOptInterface.SLOW_PROGRESS"><code>MathOptInterface.SLOW_PROGRESS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">SLOW_PROGRESS::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>SLOW_PROGRESS</code>: The algorithm stopped because it was unable to continue making progress towards the solution.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NUMERICAL_ERROR" href="#MathOptInterface.NUMERICAL_ERROR"><code>MathOptInterface.NUMERICAL_ERROR</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NUMERICAL_ERROR::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>NUMERICAL_ERROR</code>: The algorithm stopped because it encountered unrecoverable numerical error.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.INVALID_MODEL" href="#MathOptInterface.INVALID_MODEL"><code>MathOptInterface.INVALID_MODEL</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INVALID_MODEL::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INVALID_MODEL</code>: The algorithm stopped because the model is invalid.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.INVALID_OPTION" href="#MathOptInterface.INVALID_OPTION"><code>MathOptInterface.INVALID_OPTION</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INVALID_OPTION::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INVALID_OPTION</code>: The algorithm stopped because it was provided an invalid option.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.INTERRUPTED" href="#MathOptInterface.INTERRUPTED"><code>MathOptInterface.INTERRUPTED</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INTERRUPTED::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>INTERRUPTED</code>: The algorithm stopped because of an interrupt signal.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OTHER_ERROR" href="#MathOptInterface.OTHER_ERROR"><code>MathOptInterface.OTHER_ERROR</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OTHER_ERROR::TerminationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.TerminationStatusCode"><code>TerminationStatusCode</code></a> enum.</p><p><code>OTHER_ERROR</code>: The algorithm stopped because of an error not covered by one of the statuses defined above.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1973-L1979">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.PrimalStatus" href="#MathOptInterface.PrimalStatus"><code>MathOptInterface.PrimalStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">PrimalStatus(result_index::Int = 1)</code></pre><p>A model attribute for the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> of the primal result <code>result_index</code>. If <code>result_index</code> is omitted, it defaults to 1.</p><p>See <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p><p>If <code>result_index</code> is larger than the value of <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> then <code>NO_SOLUTION</code> is returned.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2090-L2100">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.DualStatus" href="#MathOptInterface.DualStatus"><code>MathOptInterface.DualStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">DualStatus(result_index::Int = 1)</code></pre><p>A model attribute for the <code>ResultStatusCode</code> of the dual result <code>result_index</code>. If <code>result_index</code> is omitted, it defaults to 1.</p><p>See <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p><p>If <code>result_index</code> is larger than the value of <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> then <code>NO_SOLUTION</code> is returned.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2108-L2118">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RawStatusString" href="#MathOptInterface.RawStatusString"><code>MathOptInterface.RawStatusString</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RawStatusString()</code></pre><p>A model attribute for a solver specific string explaining why the optimizer stopped.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2048-L2053">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ResultCount" href="#MathOptInterface.ResultCount"><code>MathOptInterface.ResultCount</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ResultCount()</code></pre><p>A model attribute for the number of results available.</p><p><strong>Order of solutions</strong></p><p>A number of attributes contain an index, <code>result_index</code>, which is used to refer to one of the available results. Thus, <code>result_index</code> must be an integer between <code>1</code> and the number of available results.</p><p>As a general rule, the first result (<code>result_index=1</code>) is the most important result (e.g., an optimal solution or an infeasibility certificate). Other results will typically be alternate solutions that the solver found during the search for the first result.</p><p>If a (local) optimal solution is available, i.e., <a href="#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a> is <code>OPTIMAL</code> or <code>LOCALLY_SOLVED</code>, the first result must correspond to the (locally) optimal solution. Other results may be alternative optimal solutions, or they may be other suboptimal solutions; use <a href="#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a> to distingiush between them.</p><p>If a primal or dual infeasibility certificate is available, i.e., <a href="#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a> is <code>INFEASIBLE</code> or <code>DUAL_INFEASIBLE</code> and the corresponding <a href="#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> or <a href="#MathOptInterface.DualStatus"><code>DualStatus</code></a> is <code>INFEASIBILITY_CERTIFICATE</code>, then the first result must be a certificate. Other results may be alternate certificates, or infeasible points.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1357-L1384">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ObjectiveValue" href="#MathOptInterface.ObjectiveValue"><code>MathOptInterface.ObjectiveValue</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveValue(result_index::Int = 1)</code></pre><p>A model attribute for the objective value of the primal solution <code>result_index</code>.</p><p>If the solver does not have a primal value for the objective because the <code>result_index</code> is beyond the available solutions (whose number is indicated by the <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute), getting this attribute must throw a <a href="../errors/#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check <a href="#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> before accessing the <code>ObjectiveValue</code> attribute.</p><p>See <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1243-L1257">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.DualObjectiveValue" href="#MathOptInterface.DualObjectiveValue"><code>MathOptInterface.DualObjectiveValue</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">DualObjectiveValue(result_index::Int = 1)</code></pre><p>A model attribute for the value of the objective function of the dual problem for the <code>result_index</code>th dual result.</p><p>If the solver does not have a dual value for the objective because the <code>result_index</code> is beyond the available solutions (whose number is indicated by the <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute), getting this attribute must throw a <a href="../errors/#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>. Otherwise, if the result is unavailable for another reason (for instance, only a primal solution is available), the result is undefined. Users should first check <a href="#MathOptInterface.DualStatus"><code>DualStatus</code></a> before accessing the <code>DualObjectiveValue</code> attribute.</p><p>See <a href="#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1263-L1278">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ObjectiveBound" href="#MathOptInterface.ObjectiveBound"><code>MathOptInterface.ObjectiveBound</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveBound()</code></pre><p>A model attribute for the best known bound on the optimal objective value.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1284-L1288">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RelativeGap" href="#MathOptInterface.RelativeGap"><code>MathOptInterface.RelativeGap</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RelativeGap()</code></pre><p>A model attribute for the final relative optimality gap.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The definition of this gap is solver-dependent. However, most solvers implementing this attribute define the relative gap as some variation of <span>$\frac{|b-f|}{|f|}$</span>, where <span>$b$</span> is the best bound and <span>$f$</span> is the best feasible objective value.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1291-L1301">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SolveTimeSec" href="#MathOptInterface.SolveTimeSec"><code>MathOptInterface.SolveTimeSec</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SolveTimeSec()</code></pre><p>A model attribute for the total elapsed solution time (in seconds) as reported by the optimizer.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1306-L1311">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SimplexIterations" href="#MathOptInterface.SimplexIterations"><code>MathOptInterface.SimplexIterations</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SimplexIterations()</code></pre><p>A model attribute for the cumulative number of simplex iterations during the optimization process.</p><p>For a mixed-integer program (MIP), the return value is the total simplex iterations for all nodes.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1316-L1324">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.BarrierIterations" href="#MathOptInterface.BarrierIterations"><code>MathOptInterface.BarrierIterations</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">BarrierIterations()</code></pre><p>A model attribute for the cumulative number of barrier iterations while solving a problem.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1329-L1334">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NodeCount" href="#MathOptInterface.NodeCount"><code>MathOptInterface.NodeCount</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NodeCount()</code></pre><p>A model attribute for the total number of branch-and-bound nodes explored while solving a mixed-integer program (MIP).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1339-L1344">source</a></section></article><h3 id="ResultStatusCode"><a class="docs-heading-anchor" href="#ResultStatusCode">ResultStatusCode</a><a id="ResultStatusCode-1"></a><a class="docs-heading-anchor-permalink" href="#ResultStatusCode" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ResultStatusCode" href="#MathOptInterface.ResultStatusCode"><code>MathOptInterface.ResultStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ResultStatusCode</code></pre><p>An Enum of possible values for the <a href="#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> and <a href="#MathOptInterface.DualStatus"><code>DualStatus</code></a> attributes.</p><p>The values indicate how to interpret the result vector.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="#MathOptInterface.NO_SOLUTION"><code>NO_SOLUTION</code></a>: the result vector is empty.</li><li><a href="#MathOptInterface.FEASIBLE_POINT"><code>FEASIBLE_POINT</code></a>: the result vector is a feasible point.</li><li><a href="#MathOptInterface.NEARLY_FEASIBLE_POINT"><code>NEARLY_FEASIBLE_POINT</code></a>: the result vector is feasible if some constraint tolerances are relaxed.</li><li><a href="#MathOptInterface.INFEASIBLE_POINT"><code>INFEASIBLE_POINT</code></a>: the result vector is an infeasible point.</li><li><a href="#MathOptInterface.INFEASIBILITY_CERTIFICATE"><code>INFEASIBILITY_CERTIFICATE</code></a>: the result vector is an infeasibility certificate. If the <code>PrimalStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the primal result vector is a certificate of dual infeasibility. If the <code>DualStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the dual result vector is a proof of primal infeasibility.</li><li><a href="#MathOptInterface.NEARLY_INFEASIBILITY_CERTIFICATE"><code>NEARLY_INFEASIBILITY_CERTIFICATE</code></a>: the result satisfies a relaxed criterion for a certificate of infeasibility.</li><li><a href="#MathOptInterface.REDUCTION_CERTIFICATE"><code>REDUCTION_CERTIFICATE</code></a>: the result vector is an ill-posed certificate; see <a href="https://arxiv.org/abs/1408.4685">this article</a> for details. If the <code>PrimalStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the primal result vector is a proof that the dual problem is ill-posed. If the <code>DualStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the dual result vector is a proof that the primal is ill-posed.</li><li><a href="#MathOptInterface.NEARLY_REDUCTION_CERTIFICATE"><code>NEARLY_REDUCTION_CERTIFICATE</code></a>: the result satisfies a relaxed criterion for an ill-posed certificate.</li><li><a href="#MathOptInterface.UNKNOWN_RESULT_STATUS"><code>UNKNOWN_RESULT_STATUS</code></a>: the result vector contains a solution with an unknown interpretation.</li><li><a href="#MathOptInterface.OTHER_RESULT_STATUS"><code>OTHER_RESULT_STATUS</code></a>: the result vector contains a solution with an interpretation not covered by one of the statuses defined above</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2080">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NO_SOLUTION" href="#MathOptInterface.NO_SOLUTION"><code>MathOptInterface.NO_SOLUTION</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NO_SOLUTION::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NO_SOLUTION</code>: the result vector is empty.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FEASIBLE_POINT" href="#MathOptInterface.FEASIBLE_POINT"><code>MathOptInterface.FEASIBLE_POINT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">FEASIBLE_POINT::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>FEASIBLE_POINT</code>: the result vector is a feasible point.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NEARLY_FEASIBLE_POINT" href="#MathOptInterface.NEARLY_FEASIBLE_POINT"><code>MathOptInterface.NEARLY_FEASIBLE_POINT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NEARLY_FEASIBLE_POINT::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NEARLY_FEASIBLE_POINT</code>: the result vector is feasible if some constraint tolerances are relaxed.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.INFEASIBLE_POINT" href="#MathOptInterface.INFEASIBLE_POINT"><code>MathOptInterface.INFEASIBLE_POINT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBLE_POINT::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>INFEASIBLE_POINT</code>: the result vector is an infeasible point.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.INFEASIBILITY_CERTIFICATE" href="#MathOptInterface.INFEASIBILITY_CERTIFICATE"><code>MathOptInterface.INFEASIBILITY_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">INFEASIBILITY_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>INFEASIBILITY_CERTIFICATE</code>: the result vector is an infeasibility certificate. If the <code>PrimalStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the primal result vector is a certificate of dual infeasibility. If the <code>DualStatus</code> is <code>INFEASIBILITY_CERTIFICATE</code>, then the dual result vector is a proof of primal infeasibility.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NEARLY_INFEASIBILITY_CERTIFICATE" href="#MathOptInterface.NEARLY_INFEASIBILITY_CERTIFICATE"><code>MathOptInterface.NEARLY_INFEASIBILITY_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NEARLY_INFEASIBILITY_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NEARLY_INFEASIBILITY_CERTIFICATE</code>: the result satisfies a relaxed criterion for a certificate of infeasibility.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.REDUCTION_CERTIFICATE" href="#MathOptInterface.REDUCTION_CERTIFICATE"><code>MathOptInterface.REDUCTION_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">REDUCTION_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>REDUCTION_CERTIFICATE</code>: the result vector is an ill-posed certificate; see <a href="https://arxiv.org/abs/1408.4685">this article</a> for details. If the <code>PrimalStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the primal result vector is a proof that the dual problem is ill-posed. If the <code>DualStatus</code> is <code>REDUCTION_CERTIFICATE</code>, then the dual result vector is a proof that the primal is ill-posed.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NEARLY_REDUCTION_CERTIFICATE" href="#MathOptInterface.NEARLY_REDUCTION_CERTIFICATE"><code>MathOptInterface.NEARLY_REDUCTION_CERTIFICATE</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NEARLY_REDUCTION_CERTIFICATE::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>NEARLY_REDUCTION_CERTIFICATE</code>: the result satisfies a relaxed criterion for an ill-posed certificate.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.UNKNOWN_RESULT_STATUS" href="#MathOptInterface.UNKNOWN_RESULT_STATUS"><code>MathOptInterface.UNKNOWN_RESULT_STATUS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">UNKNOWN_RESULT_STATUS::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>UNKNOWN_RESULT_STATUS</code>: the result vector contains a solution with an unknown interpretation.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.OTHER_RESULT_STATUS" href="#MathOptInterface.OTHER_RESULT_STATUS"><code>MathOptInterface.OTHER_RESULT_STATUS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">OTHER_RESULT_STATUS::ResultStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ResultStatusCode"><code>ResultStatusCode</code></a> enum.</p><p><code>OTHER_RESULT_STATUS</code>: the result vector contains a solution with an interpretation not covered by one of the statuses defined above</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L2058-L2064">source</a></section></article><h3 id="Conflict-Status"><a class="docs-heading-anchor" href="#Conflict-Status">Conflict Status</a><a id="Conflict-Status-1"></a><a class="docs-heading-anchor-permalink" href="#Conflict-Status" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.compute_conflict!" href="#MathOptInterface.compute_conflict!"><code>MathOptInterface.compute_conflict!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">compute_conflict!(optimizer::AbstractOptimizer)</code></pre><p>Computes a minimal subset of constraints such that the model with the other constraint removed is still infeasible.</p><p>Some solvers call a set of conflicting constraints an Irreducible Inconsistent Subsystem (IIS).</p><p>See also <a href="#MathOptInterface.ConflictStatus"><code>ConflictStatus</code></a> and <a href="#MathOptInterface.ConstraintConflictStatus"><code>ConstraintConflictStatus</code></a>.</p><p><strong>Note</strong></p><p>If the model is modified after a call to <code>compute_conflict!</code>, the implementor is not obliged to purge the conflict. Any calls to the above attributes may return values for the original conflict without a warning. Similarly, when modifying the model, the conflict can be discarded.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/MathOptInterface.jl#L89-L106">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConflictStatus" href="#MathOptInterface.ConflictStatus"><code>MathOptInterface.ConflictStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConflictStatus()</code></pre><p>A model attribute for the <a href="#MathOptInterface.ConflictStatusCode"><code>ConflictStatusCode</code></a> explaining why the conflict refiner stopped when computing the conflict.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1410-L1415">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConstraintConflictStatus" href="#MathOptInterface.ConstraintConflictStatus"><code>MathOptInterface.ConstraintConflictStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintConflictStatus()</code></pre><p>A constraint attribute indicating whether the constraint participates in the conflict. Its type is <a href="#MathOptInterface.ConflictParticipationStatusCode"><code>ConflictParticipationStatusCode</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1860-L1865">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConflictStatusCode" href="#MathOptInterface.ConflictStatusCode"><code>MathOptInterface.ConflictStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConflictStatusCode</code></pre><p>An Enum of possible values for the <code>ConflictStatus</code> attribute. This attribute is meant to explain the reason why the conflict finder stopped executing in the most recent call to <a href="#MathOptInterface.compute_conflict!"><code>compute_conflict!</code></a>.</p><p>Possible values are:</p><ul><li><code>COMPUTE_CONFLICT_NOT_CALLED</code>: the function <a href="#MathOptInterface.compute_conflict!"><code>compute_conflict!</code></a> has not yet been called</li><li><code>NO_CONFLICT_EXISTS</code>: there is no conflict because the problem is feasible</li><li><code>NO_CONFLICT_FOUND</code>: the solver could not find a conflict</li><li><code>CONFLICT_FOUND</code>: at least one conflict could be found</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1389-L1402">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ConflictParticipationStatusCode" href="#MathOptInterface.ConflictParticipationStatusCode"><code>MathOptInterface.ConflictParticipationStatusCode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConflictParticipationStatusCode</code></pre><p>An Enum of possible values for the <a href="#MathOptInterface.ConstraintConflictStatus"><code>ConstraintConflictStatus</code></a> attribute. This attribute is meant to indicate whether a given constraint participates or not in the last computed conflict.</p><p><strong>Values</strong></p><p>Possible values are:</p><ul><li><a href="#MathOptInterface.NOT_IN_CONFLICT"><code>NOT_IN_CONFLICT</code></a>: the constraint does not participate in the conflict</li><li><a href="#MathOptInterface.IN_CONFLICT"><code>IN_CONFLICT</code></a>: the constraint participates in the conflict</li><li><a href="#MathOptInterface.MAYBE_IN_CONFLICT"><code>MAYBE_IN_CONFLICT</code></a>: the constraint may participate in the conflict, the solver was not able to prove that the constraint can be excluded from the conflict</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1843-L1857">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NOT_IN_CONFLICT" href="#MathOptInterface.NOT_IN_CONFLICT"><code>MathOptInterface.NOT_IN_CONFLICT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">NOT_IN_CONFLICT::ConflictParticipationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ConflictParticipationStatusCode"><code>ConflictParticipationStatusCode</code></a> enum.</p><p><code>NOT_IN_CONFLICT</code>: the constraint does not participate in the conflict</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1843-L1849">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.IN_CONFLICT" href="#MathOptInterface.IN_CONFLICT"><code>MathOptInterface.IN_CONFLICT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">IN_CONFLICT::ConflictParticipationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ConflictParticipationStatusCode"><code>ConflictParticipationStatusCode</code></a> enum.</p><p><code>IN_CONFLICT</code>: the constraint participates in the conflict</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1843-L1849">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.MAYBE_IN_CONFLICT" href="#MathOptInterface.MAYBE_IN_CONFLICT"><code>MathOptInterface.MAYBE_IN_CONFLICT</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">MAYBE_IN_CONFLICT::ConflictParticipationStatusCode</code></pre><p>An instance of the <a href="#MathOptInterface.ConflictParticipationStatusCode"><code>ConflictParticipationStatusCode</code></a> enum.</p><p><code>MAYBE_IN_CONFLICT</code>: the constraint may participate in the conflict, the solver was not able to prove that the constraint can be excluded from the conflict</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1843-L1849">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../standard_form/">« Standard form</a><a class="docs-footer-nextpage" href="../variables/">Variables »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/modification/index.html b/previews/PR3525/moi/reference/modification/index.html
new file mode 100644
index 00000000000..37df230ae25
--- /dev/null
+++ b/previews/PR3525/moi/reference/modification/index.html
@@ -0,0 +1,31 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Modifications · JuMP</title><meta name="title" content="Modifications · JuMP"/><meta property="og:title" content="Modifications · JuMP"/><meta property="twitter:title" content="Modifications · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li class="is-active"><a class="tocitem" href>Modifications</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Modifications</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Modifications</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/modification.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Modifications"><a class="docs-heading-anchor" href="#Modifications">Modifications</a><a id="Modifications-1"></a><a class="docs-heading-anchor-permalink" href="#Modifications" title="Permalink"></a></h1><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.modify" href="#MathOptInterface.modify"><code>MathOptInterface.modify</code></a> — <span class="docstring-category">Function</span></header><section><div><p><strong>Constraint Function</strong></p><pre><code class="nohighlight hljs">modify(model::ModelLike, ci::ConstraintIndex, change::AbstractFunctionModification)</code></pre><p>Apply the modification specified by <code>change</code> to the function of constraint <code>ci</code>.</p><p>An <a href="../errors/#MathOptInterface.ModifyConstraintNotAllowed"><code>ModifyConstraintNotAllowed</code></a> error is thrown if modifying constraints is not supported by the model <code>model</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">modify(model, ci, ScalarConstantChange(10.0))</code></pre><p><strong>Objective Function</strong></p><pre><code class="nohighlight hljs">modify(model::ModelLike, ::ObjectiveFunction, change::AbstractFunctionModification)</code></pre><p>Apply the modification specified by <code>change</code> to the objective function of <code>model</code>. To change the function completely, call <code>set</code> instead.</p><p>An <a href="../errors/#MathOptInterface.ModifyObjectiveNotAllowed"><code>ModifyObjectiveNotAllowed</code></a> error is thrown if modifying objectives is not supported by the model <code>model</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">modify(model, ObjectiveFunction{ScalarAffineFunction{Float64}}(), ScalarConstantChange(10.0))</code></pre><p><strong>Multiple modifications in Constraint Functions</strong></p><pre><code class="nohighlight hljs">modify(
+    model::ModelLike,
+    cis::AbstractVector{&lt;:ConstraintIndex},
+    changes::AbstractVector{&lt;:AbstractFunctionModification},
+)</code></pre><p>Apply multiple modifications specified by <code>changes</code> to the functions of constraints <code>cis</code>.</p><p>A <a href="../errors/#MathOptInterface.ModifyConstraintNotAllowed"><code>ModifyConstraintNotAllowed</code></a> error is thrown if modifying constraints is not supported by <code>model</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">modify(
+    model,
+    [ci, ci],
+    [
+        ScalarCoefficientChange{Float64}(VariableIndex(1), 1.0),
+        ScalarCoefficientChange{Float64}(VariableIndex(2), 0.5),
+    ],
+)</code></pre><p><strong>Multiple modifications in the Objective Function</strong></p><pre><code class="nohighlight hljs">modify(
+    model::ModelLike,
+    attr::ObjectiveFunction,
+    changes::AbstractVector{&lt;:AbstractFunctionModification},
+)</code></pre><p>Apply multiple modifications specified by <code>changes</code> to the functions of constraints <code>cis</code>.</p><p>A <a href="../errors/#MathOptInterface.ModifyObjectiveNotAllowed"><code>ModifyObjectiveNotAllowed</code></a> error is thrown if modifying objective coefficients is not supported by <code>model</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">modify(
+    model,
+    ObjectiveFunction{ScalarAffineFunction{Float64}}(),
+    [
+        ScalarCoefficientChange{Float64}(VariableIndex(1), 1.0),
+        ScalarCoefficientChange{Float64}(VariableIndex(2), 0.5),
+    ],
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/modifications.jl#L67-L150">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractFunctionModification" href="#MathOptInterface.AbstractFunctionModification"><code>MathOptInterface.AbstractFunctionModification</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractFunctionModification</code></pre><p>An abstract supertype for structs which specify partial modifications to functions, to be used for making small modifications instead of replacing the functions entirely.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L739-L745">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarConstantChange" href="#MathOptInterface.ScalarConstantChange"><code>MathOptInterface.ScalarConstantChange</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarConstantChange{T}(new_constant::T)</code></pre><p>A struct used to request a change in the constant term of a scalar-valued function.</p><p>Applicable to <a href="../standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> and <a href="../standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L748-L755">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VectorConstantChange" href="#MathOptInterface.VectorConstantChange"><code>MathOptInterface.VectorConstantChange</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorConstantChange{T}(new_constant::Vector{T})</code></pre><p>A struct used to request a change in the constant vector of a vector-valued function.</p><p>Applicable to <a href="../standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a> and <a href="../standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L760-L767">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarCoefficientChange" href="#MathOptInterface.ScalarCoefficientChange"><code>MathOptInterface.ScalarCoefficientChange</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarCoefficientChange{T}(variable::VariableIndex, new_coefficient::T)</code></pre><p>A struct used to request a change in the linear coefficient of a single variable in a scalar-valued function.</p><p>Applicable to <a href="../standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> and <a href="../standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L772-L779">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.MultirowChange" href="#MathOptInterface.MultirowChange"><code>MathOptInterface.MultirowChange</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">MultirowChange{T}(
+    variable::VariableIndex,
+    new_coefficients::Vector{Tuple{Int64,T}},
+) where {T}</code></pre><p>A struct used to request a change in the linear coefficients of a single variable in a vector-valued function.</p><p>New coefficients are specified by <code>(output_index, coefficient)</code> tuples.</p><p>Applicable to <a href="../standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a> and <a href="../standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L793-L805">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../constraints/">« Constraints</a><a class="docs-footer-nextpage" href="../nonlinear/">Nonlinear programming »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/nonlinear/index.html b/previews/PR3525/moi/reference/nonlinear/index.html
new file mode 100644
index 00000000000..de3b12f39a5
--- /dev/null
+++ b/previews/PR3525/moi/reference/nonlinear/index.html
@@ -0,0 +1,82 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Nonlinear programming · JuMP</title><meta name="title" content="Nonlinear programming · JuMP"/><meta property="og:title" content="Nonlinear programming · JuMP"/><meta property="twitter:title" content="Nonlinear programming · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form</a></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../modification/">Modifications</a></li><li class="is-active"><a class="tocitem" href>Nonlinear programming</a><ul class="internal"><li><a class="tocitem" href="#Types"><span>Types</span></a></li><li><a class="tocitem" href="#Attributes"><span>Attributes</span></a></li><li><a class="tocitem" href="#Functions"><span>Functions</span></a></li></ul></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Nonlinear programming</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Nonlinear programming</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/nonlinear.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Nonlinear-programming"><a class="docs-heading-anchor" href="#Nonlinear-programming">Nonlinear programming</a><a id="Nonlinear-programming-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear-programming" title="Permalink"></a></h1><h2 id="Types"><a class="docs-heading-anchor" href="#Types">Types</a><a id="Types-1"></a><a class="docs-heading-anchor-permalink" href="#Types" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractNLPEvaluator" href="#MathOptInterface.AbstractNLPEvaluator"><code>MathOptInterface.AbstractNLPEvaluator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractNLPEvaluator</code></pre><p>Abstract supertype for the callback object that is used to query function values, derivatives, and expression graphs.</p><p>It is used in <a href="#MathOptInterface.NLPBlockData"><code>NLPBlockData</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L15-L22">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NLPBoundsPair" href="#MathOptInterface.NLPBoundsPair"><code>MathOptInterface.NLPBoundsPair</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NLPBoundsPair(lower::Float64, upper::Float64)</code></pre><p>A struct holding a pair of lower and upper bounds.</p><p><code>-Inf</code> and <code>Inf</code> can be used to indicate no lower or upper bound, respectively.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L58-L64">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NLPBlockData" href="#MathOptInterface.NLPBlockData"><code>MathOptInterface.NLPBlockData</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct NLPBlockData
+    constraint_bounds::Vector{NLPBoundsPair}
+    evaluator::AbstractNLPEvaluator
+    has_objective::Bool
+end</code></pre><p>A struct encoding a set of nonlinear constraints of the form <span>$lb \le g(x) \le ub$</span> and, if <code>has_objective == true</code>, a nonlinear objective function <span>$f(x)$</span>.</p><p>Nonlinear objectives <em>override</em> any objective set by using the <a href="../models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> attribute.</p><p>The <code>evaluator</code> is a callback object that is used to query function values, derivatives, and expression graphs. If <code>has_objective == false</code>, then it is an error to query properties of the objective function, and in Hessian-of-the-Lagrangian queries, <code>σ</code> must be set to zero.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Throughout the evaluator, all variables are ordered according to <a href="../models/#MathOptInterface.ListOfVariableIndices"><code>ListOfVariableIndices</code></a>. Hence, MOI copies of nonlinear problems must not re-order variables.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L70-L93">source</a></section></article><h2 id="Attributes"><a class="docs-heading-anchor" href="#Attributes">Attributes</a><a id="Attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NLPBlock" href="#MathOptInterface.NLPBlock"><code>MathOptInterface.NLPBlock</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NLPBlock()</code></pre><p>An <a href="../models/#MathOptInterface.AbstractModelAttribute"><code>AbstractModelAttribute</code></a> that stores an <a href="#MathOptInterface.NLPBlockData"><code>NLPBlockData</code></a>, representing a set of nonlinear constraints, and optionally a nonlinear objective.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L25-L31">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NLPBlockDual" href="#MathOptInterface.NLPBlockDual"><code>MathOptInterface.NLPBlockDual</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NLPBlockDual(result_index::Int = 1)</code></pre><p>An <a href="../models/#MathOptInterface.AbstractModelAttribute"><code>AbstractModelAttribute</code></a> for the Lagrange multipliers on the constraints from the <a href="#MathOptInterface.NLPBlock"><code>NLPBlock</code></a> in result <code>result_index</code>.</p><p>If <code>result_index</code> is omitted, it is <code>1</code> by default.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L34-L41">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NLPBlockDualStart" href="#MathOptInterface.NLPBlockDualStart"><code>MathOptInterface.NLPBlockDualStart</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NLPBlockDualStart()</code></pre><p>An <a href="../models/#MathOptInterface.AbstractModelAttribute"><code>AbstractModelAttribute</code></a> for the initial assignment of the Lagrange multipliers on the constraints from the <a href="#MathOptInterface.NLPBlock"><code>NLPBlock</code></a> that the solver may use to warm-start the solve.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L49-L55">source</a></section></article><h2 id="Functions"><a class="docs-heading-anchor" href="#Functions">Functions</a><a id="Functions-1"></a><a class="docs-heading-anchor-permalink" href="#Functions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.initialize" href="#MathOptInterface.initialize"><code>MathOptInterface.initialize</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">initialize(
+    d::AbstractNLPEvaluator,
+    requested_features::Vector{Symbol},
+)::Nothing</code></pre><p>Initialize <code>d</code> with the set of features in <code>requested_features</code>. Check <a href="#MathOptInterface.features_available"><code>features_available</code></a> before calling <code>initialize</code> to see what features are supported by <code>d</code>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This method must be called before any other methods.</p></div></div><p><strong>Features</strong></p><p>The following features are defined:</p><ul><li><code>:Grad</code>: enables <a href="#MathOptInterface.eval_objective_gradient"><code>eval_objective_gradient</code></a></li><li><code>:Jac</code>: enables <a href="#MathOptInterface.eval_constraint_jacobian"><code>eval_constraint_jacobian</code></a></li><li><code>:JacVec</code>: enables <a href="#MathOptInterface.eval_constraint_jacobian_product"><code>eval_constraint_jacobian_product</code></a> and <a href="#MathOptInterface.eval_constraint_jacobian_transpose_product"><code>eval_constraint_jacobian_transpose_product</code></a></li><li><code>:Hess</code>: enables <a href="#MathOptInterface.eval_hessian_lagrangian"><code>eval_hessian_lagrangian</code></a></li><li><code>:HessVec</code>: enables <a href="#MathOptInterface.eval_hessian_lagrangian_product"><code>eval_hessian_lagrangian_product</code></a></li><li><code>:ExprGraph</code>: enables <a href="#MathOptInterface.objective_expr"><code>objective_expr</code></a> and <a href="#MathOptInterface.constraint_expr"><code>constraint_expr</code></a>.</li></ul><p>In all cases, including when <code>requested_features</code> is empty, <a href="#MathOptInterface.eval_objective"><code>eval_objective</code></a> and <a href="#MathOptInterface.eval_constraint"><code>eval_constraint</code></a> are supported.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">MOI.initialize(d, Symbol[])
+MOI.initialize(d, [:ExprGraph])
+MOI.initialize(d, MOI.features_available(d))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L100-L135">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.features_available" href="#MathOptInterface.features_available"><code>MathOptInterface.features_available</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">features_available(d::AbstractNLPEvaluator)::Vector{Symbol}</code></pre><p>Returns the subset of features available for this problem instance.</p><p>See <a href="#MathOptInterface.initialize"><code>initialize</code></a> for the list of defined features.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L138-L144">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_objective" href="#MathOptInterface.eval_objective"><code>MathOptInterface.eval_objective</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_objective(d::AbstractNLPEvaluator, x::AbstractVector{T})::T where {T}</code></pre><p>Evaluate the objective <span>$f(x)$</span>, returning a scalar value.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L147-L151">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_constraint" href="#MathOptInterface.eval_constraint"><code>MathOptInterface.eval_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_constraint(d::AbstractNLPEvaluator,
+    g::AbstractVector{T},
+    x::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>Given a set of vector-valued constraints <span>$l \le g(x) \le u$</span>, evaluate the constraint function <span>$g(x)$</span>, storing the result in the vector <code>g</code>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>g</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L154-L168">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_objective_gradient" href="#MathOptInterface.eval_objective_gradient"><code>MathOptInterface.eval_objective_gradient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_objective_gradient(
+    d::AbstractNLPEvaluator,
+    grad::AbstractVector{T},
+    x::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>Evaluate the gradient of the objective function <span>$grad = \nabla f(x)$</span> as a dense vector, storing the result in the vector <code>grad</code>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>grad</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L171-L186">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.jacobian_structure" href="#MathOptInterface.jacobian_structure"><code>MathOptInterface.jacobian_structure</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">jacobian_structure(d::AbstractNLPEvaluator)::Vector{Tuple{Int64,Int64}}</code></pre><p>Returns a vector of tuples, <code>(row, column)</code>, where each indicates the position of a structurally nonzero element in the Jacobian matrix: <span>$J_g(x) = \left[ \begin{array}{c} \nabla g_1(x) \\ \nabla g_2(x) \\ \vdots \\ \nabla g_m(x) \end{array}\right],$</span> where <span>$g_i$</span> is the <span>$i\text{th}$</span> component of the nonlinear constraints <span>$g(x)$</span>.</p><p>The indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.</p><p>The sparsity structure is assumed to be independent of the point <span>$x$</span>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L189-L203">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_constraint_gradient" href="#MathOptInterface.eval_constraint_gradient"><code>MathOptInterface.eval_constraint_gradient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_constraint_gradient(
+    d::AbstractNLPEvaluator,
+    ∇g::AbstractVector{T},
+    x::AbstractVector{T},
+    i::Int,
+)::Nothing where {T}</code></pre><p>Evaluate the gradient of constraint <code>i</code>, <span>$\nabla g_i(x)$</span>, and store the non-zero values in <code>∇g</code>, corresponding to the structure returned by <a href="#MathOptInterface.constraint_gradient_structure"><code>constraint_gradient_structure</code></a>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>∇g</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L281-L298">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.constraint_gradient_structure" href="#MathOptInterface.constraint_gradient_structure"><code>MathOptInterface.constraint_gradient_structure</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constraint_gradient_structure(d::AbstractNLPEvaluator, i::Int)::Vector{Int64}</code></pre><p>Returns a vector of indices, where each element indicates the position of a structurally nonzero element in the gradient of constraint <span>$\nabla g_i(x)$</span>.</p><p>The indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.</p><p>The sparsity structure is assumed to be independent of the point <span>$x$</span>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L267-L278">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_constraint_jacobian" href="#MathOptInterface.eval_constraint_jacobian"><code>MathOptInterface.eval_constraint_jacobian</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_constraint_jacobian(d::AbstractNLPEvaluator,
+    J::AbstractVector{T},
+    x::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>Evaluates the sparse Jacobian matrix <span>$J_g(x) = \left[ \begin{array}{c} \nabla g_1(x) \\ \nabla g_2(x) \\ \vdots \\ \nabla g_m(x) \end{array}\right]$</span>.</p><p>The result is stored in the vector <code>J</code> in the same order as the indices returned by <a href="#MathOptInterface.jacobian_structure"><code>jacobian_structure</code></a>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>J</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L301-L318">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_constraint_jacobian_product" href="#MathOptInterface.eval_constraint_jacobian_product"><code>MathOptInterface.eval_constraint_jacobian_product</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_constraint_jacobian_product(
+    d::AbstractNLPEvaluator,
+    y::AbstractVector{T},
+    x::AbstractVector{T},
+    w::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>Computes the Jacobian-vector product <span>$y = J_g(x)w$</span>, storing the result in the vector <code>y</code>.</p><p>The vectors have dimensions such that <code>length(w) == length(x)</code>, and <code>length(y)</code> is the number of nonlinear constraints.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>y</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L321-L340">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_constraint_jacobian_transpose_product" href="#MathOptInterface.eval_constraint_jacobian_transpose_product"><code>MathOptInterface.eval_constraint_jacobian_transpose_product</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_constraint_jacobian_transpose_product(
+    d::AbstractNLPEvaluator,
+    y::AbstractVector{T},
+    x::AbstractVector{T},
+    w::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>Computes the Jacobian-transpose-vector product <span>$y = J_g(x)^Tw$</span>, storing the result in the vector <code>y</code>.</p><p>The vectors have dimensions such that <code>length(y) == length(x)</code>, and <code>length(w)</code> is the number of nonlinear constraints.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>y</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L343-L362">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.hessian_lagrangian_structure" href="#MathOptInterface.hessian_lagrangian_structure"><code>MathOptInterface.hessian_lagrangian_structure</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">hessian_lagrangian_structure(
+    d::AbstractNLPEvaluator,
+)::Vector{Tuple{Int64,Int64}}</code></pre><p>Returns a vector of tuples, <code>(row, column)</code>, where each indicates the position of a structurally nonzero element in the Hessian-of-the-Lagrangian matrix: <span>$\nabla^2 f(x) + \sum_{i=1}^m \nabla^2 g_i(x)$</span>.</p><p>The indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.</p><p>Any mix of lower and upper-triangular indices is valid. Elements <code>(i,j)</code> and <code>(j,i)</code>, if both present, should be treated as duplicates.</p><p>The sparsity structure is assumed to be independent of the point <span>$x$</span>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L247-L264">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.hessian_objective_structure" href="#MathOptInterface.hessian_objective_structure"><code>MathOptInterface.hessian_objective_structure</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">hessian_objective_structure(
+    d::AbstractNLPEvaluator,
+)::Vector{Tuple{Int64,Int64}}</code></pre><p>Returns a vector of tuples, <code>(row, column)</code>, where each indicates the position of a structurally nonzero element in the Hessian matrix: <span>$\nabla^2 f(x)$</span>.</p><p>The indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.</p><p>Any mix of lower and upper-triangular indices is valid. Elements <code>(i,j)</code> and <code>(j,i)</code>, if both present, should be treated as duplicates.</p><p>The sparsity structure is assumed to be independent of the point <span>$x$</span>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L206-L223">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.hessian_constraint_structure" href="#MathOptInterface.hessian_constraint_structure"><code>MathOptInterface.hessian_constraint_structure</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">hessian_constraint_structure(
+    d::AbstractNLPEvaluator,
+    i::Int64,
+)::Vector{Tuple{Int64,Int64}}</code></pre><p>Returns a vector of tuples, <code>(row, column)</code>, where each indicates the position of a structurally nonzero element in the Hessian matrix: <span>$\nabla^2 g_i(x)$</span>.</p><p>The indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.</p><p>Any mix of lower and upper-triangular indices is valid. Elements <code>(i,j)</code> and <code>(j,i)</code>, if both present, should be treated as duplicates.</p><p>The sparsity structure is assumed to be independent of the point <span>$x$</span>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L226-L244">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_hessian_objective" href="#MathOptInterface.eval_hessian_objective"><code>MathOptInterface.eval_hessian_objective</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_hessian_objective(
+    d::AbstractNLPEvaluator,
+    H::AbstractVector{T},
+    x::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>This function computes the sparse Hessian matrix: <span>$\nabla^2 f(x)$</span>, storing the result in the vector <code>H</code> in the same order as the indices returned by <a href="#MathOptInterface.hessian_objective_structure"><code>hessian_objective_structure</code></a>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>H</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L390-L407">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_hessian_constraint" href="#MathOptInterface.eval_hessian_constraint"><code>MathOptInterface.eval_hessian_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_hessian_constraint(
+    d::AbstractNLPEvaluator,
+    H::AbstractVector{T},
+    x::AbstractVector{T},
+    i::Int64,
+)::Nothing where {T}</code></pre><p>This function computes the sparse Hessian matrix: <span>$\nabla^2 g_i(x)$</span>, storing the result in the vector <code>H</code> in the same order as the indices returned by <a href="#MathOptInterface.hessian_constraint_structure"><code>hessian_constraint_structure</code></a>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>H</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L410-L428">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_hessian_lagrangian" href="#MathOptInterface.eval_hessian_lagrangian"><code>MathOptInterface.eval_hessian_lagrangian</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_hessian_lagrangian(
+    d::AbstractNLPEvaluator,
+    H::AbstractVector{T},
+    x::AbstractVector{T},
+    σ::T,
+    μ::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>Given scalar weight <code>σ</code> and vector of constraint weights <code>μ</code>, this function computes the sparse Hessian-of-the-Lagrangian matrix: <span>$\sigma\nabla^2 f(x) + \sum_{i=1}^m \mu_i \nabla^2 g_i(x)$</span>, storing the result in the vector <code>H</code> in the same order as the indices returned by <a href="#MathOptInterface.hessian_lagrangian_structure"><code>hessian_lagrangian_structure</code></a>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>H</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L431-L451">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.eval_hessian_lagrangian_product" href="#MathOptInterface.eval_hessian_lagrangian_product"><code>MathOptInterface.eval_hessian_lagrangian_product</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_hessian_lagrangian_product(
+    d::AbstractNLPEvaluator,
+    h::AbstractVector{T},
+    x::AbstractVector{T},
+    v::AbstractVector{T},
+    σ::T,
+    μ::AbstractVector{T},
+)::Nothing where {T}</code></pre><p>Given scalar weight <code>σ</code> and vector of constraint weights <code>μ</code>, computes the Hessian-of-the-Lagrangian-vector product <span>$h = \left(\sigma\nabla^2 f(x) + \sum_{i=1}^m \mu_i \nabla^2 g_i(x)\right)v$</span>, storing the result in the vector <code>h</code>.</p><p>The vectors have dimensions such that <code>length(h) == length(x) == length(v)</code>.</p><p><strong>Implementation notes</strong></p><p>When implementing this method, you must not assume that <code>h</code> is <code>Vector{Float64}</code>, but you may assume that it supports <code>setindex!</code> and <code>length</code>. For example, it may be the <code>view</code> of a vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L365-L387">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.objective_expr" href="#MathOptInterface.objective_expr"><code>MathOptInterface.objective_expr</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">objective_expr(d::AbstractNLPEvaluator)::Expr</code></pre><p>Returns a Julia <code>Expr</code> object representing the expression graph of the objective function.</p><p><strong>Format</strong></p><p>The expression has a number of limitations, compared with arbitrary Julia expressions:</p><ul><li>All sums and products are flattened out as simple <code>Expr(:+, ...)</code> and <code>Expr(:*, ...)</code> objects.</li><li>All decision variables must be of the form <code>Expr(:ref, :x, MOI.VariableIndex(i))</code>, where <code>i</code> is the <span>$i$</span>th variable in <a href="../models/#MathOptInterface.ListOfVariableIndices"><code>ListOfVariableIndices</code></a>.</li><li>There are currently no restrictions on recognized functions; typically these will be built-in Julia functions like <code>^</code>, <code>exp</code>, <code>log</code>, <code>cos</code>, <code>tan</code>, <code>sqrt</code>, etc., but modeling interfaces may choose to extend these basic functions, or error if they encounter unsupported functions.</li></ul><p><strong>Examples</strong></p><p>The expression <span>$x_1+\sin(x_2/\exp(x_3))$</span> is represented as</p><pre><code class="language-julia hljs">:(x[MOI.VariableIndex(1)] + sin(x[MOI.VariableIndex(2)] / exp(x[MOI.VariableIndex[3]])))</code></pre><p>or equivalently</p><pre><code class="language-julia hljs">Expr(
+    :call,
+    :+,
+    Expr(:ref, :x, MOI.VariableIndex(1)),
+    Expr(
+        :call,
+        :/,
+        Expr(:call, :sin, Expr(:ref, :x, MOI.VariableIndex(2))),
+        Expr(:call, :exp, Expr(:ref, :x, MOI.VariableIndex(3))),
+    ),
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L454-L495">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.constraint_expr" href="#MathOptInterface.constraint_expr"><code>MathOptInterface.constraint_expr</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constraint_expr(d::AbstractNLPEvaluator, i::Integer)::Expr</code></pre><p>Returns a Julia <code>Expr</code> object representing the expression graph for the <span>$i\text{th}$</span> nonlinear constraint.</p><p><strong>Format</strong></p><p>The format is the same as <a href="#MathOptInterface.objective_expr"><code>objective_expr</code></a>, with an additional comparison operator indicating the sense of and bounds on the constraint.</p><p>For single-sided comparisons, the body of the constraint must be on the left-hand side, and the right-hand side must be a constant.</p><p>For double-sided comparisons (that is, <span>$l \le f(x) \le u$</span>), the body of the constraint must be in the middle, and the left- and right-hand sides must be constants.</p><p>The bounds on the constraints must match the <a href="#MathOptInterface.NLPBoundsPair"><code>NLPBoundsPair</code></a>s passed to <a href="#MathOptInterface.NLPBlockData"><code>NLPBlockData</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">:(x[MOI.VariableIndex(1)]^2 &lt;= 1.0)
+:(x[MOI.VariableIndex(1)]^2 &gt;= 2.0)
+:(x[MOI.VariableIndex(1)]^2 == 3.0)
+:(4.0 &lt;= x[MOI.VariableIndex(1)]^2 &lt;= 5.0)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/nlp.jl#L498-L527">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../modification/">« Modifications</a><a class="docs-footer-nextpage" href="../callbacks/">Callbacks »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/standard_form/index.html b/previews/PR3525/moi/reference/standard_form/index.html
new file mode 100644
index 00000000000..1701bede889
--- /dev/null
+++ b/previews/PR3525/moi/reference/standard_form/index.html
@@ -0,0 +1,993 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Standard form · JuMP</title><meta name="title" content="Standard form · JuMP"/><meta property="og:title" content="Standard form · JuMP"/><meta property="twitter:title" content="Standard form · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Standard form</a><ul class="internal"><li><a class="tocitem" href="#Functions"><span>Functions</span></a></li><li><a class="tocitem" href="#Scalar-functions"><span>Scalar functions</span></a></li><li><a class="tocitem" href="#Vector-functions"><span>Vector functions</span></a></li><li><a class="tocitem" href="#Sets"><span>Sets</span></a></li><li><a class="tocitem" href="#Scalar-sets"><span>Scalar sets</span></a></li><li><a class="tocitem" href="#Vector-sets"><span>Vector sets</span></a></li><li><a class="tocitem" href="#Constraint-programming-sets"><span>Constraint programming sets</span></a></li><li><a class="tocitem" href="#Matrix-sets"><span>Matrix sets</span></a></li></ul></li><li><a class="tocitem" href="../models/">Models</a></li><li><a class="tocitem" href="../variables/">Variables</a></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../modification/">Modifications</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Standard form</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Standard form</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/standard_form.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Standard-form"><a class="docs-heading-anchor" href="#Standard-form">Standard form</a><a id="Standard-form-1"></a><a class="docs-heading-anchor-permalink" href="#Standard-form" title="Permalink"></a></h1><h2 id="Functions"><a class="docs-heading-anchor" href="#Functions">Functions</a><a id="Functions-1"></a><a class="docs-heading-anchor-permalink" href="#Functions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractFunction" href="#MathOptInterface.AbstractFunction"><code>MathOptInterface.AbstractFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractFunction</code></pre><p>Abstract supertype for function objects.</p><p><strong>Required methods</strong></p><p>All functions must implement:</p><ul><li><code>Base.copy</code></li><li><code>Base.isapprox</code></li><li><a href="#MathOptInterface.constant"><code>constant</code></a></li></ul><p>Abstract subtypes of <code>AbstractFunction</code> may require additional methods to be implemented.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L7-L22">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.output_dimension" href="#MathOptInterface.output_dimension"><code>MathOptInterface.output_dimension</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">output_dimension(f::AbstractFunction)</code></pre><p>Return 1 if <code>f</code> is an <a href="#MathOptInterface.AbstractScalarFunction"><code>AbstractScalarFunction</code></a>, or the number of output components if <code>f</code> is an <a href="#MathOptInterface.AbstractVectorFunction"><code>AbstractVectorFunction</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L7-L12">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.constant" href="#MathOptInterface.constant"><code>MathOptInterface.constant</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">constant(f::AbstractFunction[, ::Type{T}]) where {T}</code></pre><p>Returns the constant term of a scalar-valued function, or the constant vector of a vector-valued function.</p><p>If <code>f</code> is untyped and <code>T</code> is provided, returns <code>zero(T)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L17-L24">source</a></section><section><div><pre><code class="language-julia hljs">constant(set::Union{EqualTo,GreaterThan,LessThan,Parameter})</code></pre><p>Returns the constant term of the set <code>set</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.constant(MOI.GreaterThan(1.0))
+1.0
+
+julia&gt; MOI.constant(MOI.LessThan(2.5))
+2.5
+
+julia&gt; MOI.constant(MOI.EqualTo(3))
+3
+
+julia&gt; MOI.constant(MOI.Parameter(4.5))
+4.5</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2827-L2849">source</a></section></article><h2 id="Scalar-functions"><a class="docs-heading-anchor" href="#Scalar-functions">Scalar functions</a><a id="Scalar-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Scalar-functions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractScalarFunction" href="#MathOptInterface.AbstractScalarFunction"><code>MathOptInterface.AbstractScalarFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractScalarFunction &lt;: AbstractFunction</code></pre><p>Abstract supertype for scalar-valued <a href="#MathOptInterface.AbstractFunction"><code>AbstractFunction</code></a>s.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L25-L29">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VariableIndex" href="#MathOptInterface.VariableIndex"><code>MathOptInterface.VariableIndex</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariableIndex</code></pre><p>A type-safe wrapper for <code>Int64</code> for use in referencing variables in a model. To allow for deletion, indices need not be consecutive.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L38-L43">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarAffineTerm" href="#MathOptInterface.ScalarAffineTerm"><code>MathOptInterface.ScalarAffineTerm</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarAffineTerm{T}(coefficient::T, variable::VariableIndex) where {T}</code></pre><p>Represents the scalar-valued term <code>coefficient * variable</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.ScalarAffineTerm(2.0, x)
+MathOptInterface.ScalarAffineTerm{Float64}(2.0, MOI.VariableIndex(1))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L73-L89">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarAffineFunction" href="#MathOptInterface.ScalarAffineFunction"><code>MathOptInterface.ScalarAffineFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarAffineFunction{T}(terms::ScalarAffineTerm{T}, constant::T) where {T}</code></pre><p>Represents the scalar-valued affine function <span>$a^\top x + b$</span>, where:</p><ul><li><span>$a^\top x$</span> is represented by the vector of <a href="#MathOptInterface.ScalarAffineTerm"><code>ScalarAffineTerm</code></a>s</li><li><span>$b$</span> is a scalar <code>constant::T</code></li></ul><p><strong>Duplicates</strong></p><p>Duplicate variable indices in <code>terms</code> are accepted, and the corresponding coefficients are summed together.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1)
+MOI.VariableIndex(1)
+
+julia&gt; terms = [MOI.ScalarAffineTerm(2.0, x), MOI.ScalarAffineTerm(3.0, x)]
+2-element Vector{MathOptInterface.ScalarAffineTerm{Float64}}:
+ MathOptInterface.ScalarAffineTerm{Float64}(2.0, MOI.VariableIndex(1))
+ MathOptInterface.ScalarAffineTerm{Float64}(3.0, MOI.VariableIndex(1))
+
+julia&gt; f = MOI.ScalarAffineFunction(terms, 4.0)
+4.0 + 2.0 MOI.VariableIndex(1) + 3.0 MOI.VariableIndex(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L110-L139">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarQuadraticTerm" href="#MathOptInterface.ScalarQuadraticTerm"><code>MathOptInterface.ScalarQuadraticTerm</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarQuadraticTerm{T}(
+    coefficient::T,
+    variable_1::VariableIndex,
+    variable_2::VariableIndex,
+) where {T}</code></pre><p>Represents the scalar-valued term <span>$c x_i x_j$</span> where <span>$c$</span> is <code>coefficient</code>, <span>$x_i$</span> is <code>variable_1</code> and <span>$x_j$</span> is <code>variable_2</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.ScalarQuadraticTerm(2.0, x, x)
+MathOptInterface.ScalarQuadraticTerm{Float64}(2.0, MOI.VariableIndex(1), MOI.VariableIndex(1))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L155-L176">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarQuadraticFunction" href="#MathOptInterface.ScalarQuadraticFunction"><code>MathOptInterface.ScalarQuadraticFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarQuadraticFunction{T}(
+    quadratic_terms::Vector{ScalarQuadraticTerm{T}},
+    affine_terms::Vector{ScalarAffineTerm{T}},
+    constant::T,
+) wher {T}</code></pre><p>The scalar-valued quadratic function <span>$\frac{1}{2}x^\top Q x + a^\top x + b$</span>, where:</p><ul><li><span>$Q$</span> is the symmetric matrix given by the vector of <a href="#MathOptInterface.ScalarQuadraticTerm"><code>ScalarQuadraticTerm</code></a>s</li><li><span>$a^\top x$</span> is a sparse vector given by the vector of <a href="#MathOptInterface.ScalarAffineTerm"><code>ScalarAffineTerm</code></a>s</li><li><span>$b$</span> is the scalar <code>constant::T</code>.</li></ul><p><strong>Duplicates</strong></p><p>Duplicate indices in <code>quadratic_terms</code> or <code>affine_terms</code> are accepted, and the corresponding coefficients are summed together.</p><p>In <code>quadratic_terms</code>, &quot;mirrored&quot; indices, <code>(q, r)</code> and <code>(r, q)</code> where <code>r</code> and <code>q</code> are <a href="#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>es, are considered duplicates; only one needs to be specified.</p><p><strong>The 0.5 factor</strong></p><p>Coupled with the interpretation of mirrored indices, the <code>0.5</code> factor in front of the <span>$Q$</span> matrix is a common source of bugs.</p><p>As a rule, to represent <span>$a * x^2 + b * x * y$</span>:</p><ul><li>The coefficient <span>$a$</span> in front of squared variables (diagonal elements in <span>$Q$</span>) must be doubled when creating a <a href="#MathOptInterface.ScalarQuadraticTerm"><code>ScalarQuadraticTerm</code></a></li><li>The coefficient <span>$b$</span> in front of off-diagonal elements in <span>$Q$</span> should be left as <span>$b$</span>, be cause the mirrored index <span>$b * y * x$</span> will be implicitly added.</li></ul><p><strong>Example</strong></p><p>To represent the function <span>$f(x, y) = 2 * x^2 + 3 * x * y + 4 * x + 5$</span>, do:</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1);
+
+julia&gt; y = MOI.VariableIndex(2);
+
+julia&gt; constant = 5.0;
+
+julia&gt; affine_terms = [MOI.ScalarAffineTerm(4.0, x)];
+
+julia&gt; quadratic_terms = [
+           MOI.ScalarQuadraticTerm(4.0, x, x),  # Note the changed coefficient
+           MOI.ScalarQuadraticTerm(3.0, x, y),
+       ]
+2-element Vector{MathOptInterface.ScalarQuadraticTerm{Float64}}:
+ MathOptInterface.ScalarQuadraticTerm{Float64}(4.0, MOI.VariableIndex(1), MOI.VariableIndex(1))
+ MathOptInterface.ScalarQuadraticTerm{Float64}(3.0, MOI.VariableIndex(1), MOI.VariableIndex(2))
+
+julia&gt; f = MOI.ScalarQuadraticFunction(quadratic_terms, affine_terms, constant)
+5.0 + 4.0 MOI.VariableIndex(1) + 2.0 MOI.VariableIndex(1)² + 3.0 MOI.VariableIndex(1)*MOI.VariableIndex(2)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L201-L264">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ScalarNonlinearFunction" href="#MathOptInterface.ScalarNonlinearFunction"><code>MathOptInterface.ScalarNonlinearFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarNonlinearFunction(head::Symbol, args::Vector{Any})</code></pre><p>The scalar-valued nonlinear function <code>head(args...)</code>, represented as a symbolic expression tree, with the call operator <code>head</code> and ordered arguments in <code>args</code>.</p><p><strong><code>head</code></strong></p><p>The <code>head::Symbol</code> must be an operator supported by the model.</p><p>The default list of supported univariate operators is given by:</p><ul><li><a href="../../submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS"><code>Nonlinear.DEFAULT_UNIVARIATE_OPERATORS</code></a></li></ul><p>and the default list of supported multivariate operators is given by:</p><ul><li><a href="../../submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS"><code>Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS</code></a></li></ul><p>Additional operators can be registered by setting a <a href="../models/#MathOptInterface.UserDefinedFunction"><code>UserDefinedFunction</code></a> attribute.</p><p>See the full list of operators supported by a <a href="../models/#MathOptInterface.ModelLike"><code>ModelLike</code></a> by querying <a href="../models/#MathOptInterface.ListOfSupportedNonlinearOperators"><code>ListOfSupportedNonlinearOperators</code></a>.</p><p><strong><code>args</code></strong></p><p>The vector <code>args</code> contains the arguments to the nonlinear function. If the operator is univariate, it must contain one element. Otherwise, it may contain multiple elements.</p><p>Each element must be one of the following:</p><ul><li>A constant value of type <code>T&lt;:Real</code></li><li>A <a href="#MathOptInterface.VariableIndex"><code>VariableIndex</code></a></li><li>A <a href="#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a></li><li>A <a href="#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a></li><li>A <a href="#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a></li></ul><p><strong>Unsupported operators</strong></p><p>If the optimizer does not support <code>head</code>, an <a href="../errors/#MathOptInterface.UnsupportedNonlinearOperator"><code>UnsupportedNonlinearOperator</code></a> error will be thrown.</p><p>There is no guarantee about when this error will be thrown; it may be thrown when the function is first added to the model, or it may be thrown when <a href="../models/#MathOptInterface.optimize!"><code>optimize!</code></a> is called.</p><p><strong>Example</strong></p><p>To represent the function <span>$f(x) = sin(x)^2$</span>, do:</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.ScalarNonlinearFunction(
+           :^,
+           Any[MOI.ScalarNonlinearFunction(:sin, Any[x]), 2],
+       )
+^(sin(MOI.VariableIndex(1)), (2))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L281-L344">source</a></section></article><h2 id="Vector-functions"><a class="docs-heading-anchor" href="#Vector-functions">Vector functions</a><a id="Vector-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Vector-functions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractVectorFunction" href="#MathOptInterface.AbstractVectorFunction"><code>MathOptInterface.AbstractVectorFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractVectorFunction &lt;: AbstractFunction</code></pre><p>Abstract supertype for vector-valued <a href="#MathOptInterface.AbstractFunction"><code>AbstractFunction</code></a>s.</p><p><strong>Required methods</strong></p><p>All subtypes of <code>AbstractVectorFunction</code> must implement:</p><ul><li><a href="#MathOptInterface.output_dimension"><code>output_dimension</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L396-L406">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VectorOfVariables" href="#MathOptInterface.VectorOfVariables"><code>MathOptInterface.VectorOfVariables</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorOfVariables(variables::Vector{VariableIndex}) &lt;: AbstractVectorFunction</code></pre><p>The vector-valued function <code>f(x) = variables</code>, where <code>variables</code> is a subset of <a href="#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>es in the model.</p><p>The list of <code>variables</code> may contain duplicates.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex.(1:2)
+2-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+
+julia&gt; f = MOI.VectorOfVariables([x[1], x[2], x[1]])
+┌                    ┐
+│MOI.VariableIndex(1)│
+│MOI.VariableIndex(2)│
+│MOI.VariableIndex(1)│
+└                    ┘
+
+julia&gt; MOI.output_dimension(f)
+3</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L409-L437">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VectorAffineTerm" href="#MathOptInterface.VectorAffineTerm"><code>MathOptInterface.VectorAffineTerm</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorAffineTerm{T}(
+    output_index::Int64,
+    scalar_term::ScalarAffineTerm{T},
+) where {T}</code></pre><p>A <code>VectorAffineTerm</code> is a <code>scalar_term</code> that appears in the <code>output_index</code> row of the vector-valued <a href="#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a> or <a href="#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1);
+
+julia&gt; MOI.VectorAffineTerm(Int64(2), MOI.ScalarAffineTerm(3.0, x))
+MathOptInterface.VectorAffineTerm{Float64}(2, MathOptInterface.ScalarAffineTerm{Float64}(3.0, MOI.VariableIndex(1)))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L456-L476">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VectorAffineFunction" href="#MathOptInterface.VectorAffineFunction"><code>MathOptInterface.VectorAffineFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorAffineFunction{T}(
+    terms::Vector{VectorAffineTerm{T}},
+    constants::Vector{T},
+) where {T}</code></pre><p>The vector-valued affine function <span>$A x + b$</span>, where:</p><ul><li><span>$A x$</span> is the sparse matrix given by the vector of <a href="#MathOptInterface.VectorAffineTerm"><code>VectorAffineTerm</code></a>s</li><li><span>$b$</span> is the vector <code>constants</code></li></ul><p><strong>Duplicates</strong></p><p>Duplicate indices in the <span>$A$</span> are accepted, and the corresponding coefficients are summed together.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1);
+
+julia&gt; terms = [
+           MOI.VectorAffineTerm(Int64(1), MOI.ScalarAffineTerm(2.0, x)),
+           MOI.VectorAffineTerm(Int64(2), MOI.ScalarAffineTerm(3.0, x)),
+       ];
+
+julia&gt; f = MOI.VectorAffineFunction(terms, [4.0, 5.0])
+┌                              ┐
+│4.0 + 2.0 MOI.VariableIndex(1)│
+│5.0 + 3.0 MOI.VariableIndex(1)│
+└                              ┘
+
+julia&gt; MOI.output_dimension(f)
+2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L497-L534">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VectorQuadraticTerm" href="#MathOptInterface.VectorQuadraticTerm"><code>MathOptInterface.VectorQuadraticTerm</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorQuadraticTerm{T}(
+    output_index::Int64,
+    scalar_term::ScalarQuadraticTerm{T},
+) where {T}</code></pre><p>A <code>VectorQuadraticTerm</code> is a <a href="#MathOptInterface.ScalarQuadraticTerm"><code>ScalarQuadraticTerm</code></a> <code>scalar_term</code> that appears in the <code>output_index</code> row of the vector-valued <a href="#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1);
+
+julia&gt; MOI.VectorQuadraticTerm(Int64(2), MOI.ScalarQuadraticTerm(3.0, x, x))
+MathOptInterface.VectorQuadraticTerm{Float64}(2, MathOptInterface.ScalarQuadraticTerm{Float64}(3.0, MOI.VariableIndex(1), MOI.VariableIndex(1)))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L556-L576">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VectorQuadraticFunction" href="#MathOptInterface.VectorQuadraticFunction"><code>MathOptInterface.VectorQuadraticFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorQuadraticFunction{T}(
+    quadratic_terms::Vector{VectorQuadraticTerm{T}},
+    affine_terms::Vector{VectorAffineTerm{T}},
+    constants::Vector{T},
+) where {T}</code></pre><p>The vector-valued quadratic function with i<code>th</code> component (&quot;output index&quot;) defined as <span>$\frac{1}{2}x^\top Q_i x + a_i^\top x + b_i$</span>, where:</p><ul><li><span>$\frac{1}{2}x^\top Q_i x$</span> is the symmetric matrix given by the <a href="#MathOptInterface.VectorQuadraticTerm"><code>VectorQuadraticTerm</code></a> elements in <code>quadratic_terms</code> with <code>output_index == i</code></li><li><span>$a_i^\top x$</span> is the sparse vector given by the <a href="#MathOptInterface.VectorAffineTerm"><code>VectorAffineTerm</code></a> elements in <code>affine_terms</code> with <code>output_index == i</code></li><li><span>$b_i$</span> is a scalar given by <code>constants[i]</code></li></ul><p><strong>Duplicates</strong></p><p>Duplicate indices in <code>quadratic_terms</code> and <code>affine_terms</code> with the same <code>output_index</code> are handled in the same manner as duplicates in <a href="#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1);
+
+julia&gt; y = MOI.VariableIndex(2);
+
+julia&gt; constants = [4.0, 5.0];
+
+julia&gt; affine_terms = [
+           MOI.VectorAffineTerm(Int64(1), MOI.ScalarAffineTerm(2.0, x)),
+           MOI.VectorAffineTerm(Int64(2), MOI.ScalarAffineTerm(3.0, x)),
+       ];
+
+julia&gt; quad_terms = [
+        MOI.VectorQuadraticTerm(Int64(1), MOI.ScalarQuadraticTerm(2.0, x, x)),
+        MOI.VectorQuadraticTerm(Int64(2), MOI.ScalarQuadraticTerm(3.0, x, y)),
+           ];
+
+julia&gt; f = MOI.VectorQuadraticFunction(quad_terms, affine_terms, constants)
+┌                                                                              ┐
+│4.0 + 2.0 MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(1)²                    │
+│5.0 + 3.0 MOI.VariableIndex(1) + 3.0 MOI.VariableIndex(1)*MOI.VariableIndex(2)│
+└                                                                              ┘
+
+julia&gt; MOI.output_dimension(f)
+2</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L597-L650">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VectorNonlinearFunction" href="#MathOptInterface.VectorNonlinearFunction"><code>MathOptInterface.VectorNonlinearFunction</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorNonlinearFunction(args::Vector{ScalarNonlinearFunction})</code></pre><p>The vector-valued nonlinear function composed of a vector of <a href="#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a>.</p><p><strong><code>args</code></strong></p><p>The vector <code>args</code> contains the scalar elements of the nonlinear function. Each element must be a <a href="#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a>, but if you pass a <code>Vector{Any}</code>, the elements can be automatically converted from one of the following:</p><ul><li>A constant value of type <code>T&lt;:Real</code></li><li>A <a href="#MathOptInterface.VariableIndex"><code>VariableIndex</code></a></li><li>A <a href="#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a></li><li>A <a href="#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a></li><li>A <a href="#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a></li></ul><p><strong>Example</strong></p><p>To represent the function <span>$f(x) = [sin(x)^2, x]$</span>, do:</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; x = MOI.VariableIndex(1)
+MOI.VariableIndex(1)
+
+julia&gt; g = MOI.ScalarNonlinearFunction(
+           :^,
+           Any[MOI.ScalarNonlinearFunction(:sin, Any[x]), 2.0],
+       )
+^(sin(MOI.VariableIndex(1)), 2.0)
+
+julia&gt; MOI.VectorNonlinearFunction([g, x])
+┌                                 ┐
+│^(sin(MOI.VariableIndex(1)), 2.0)│
+│+(MOI.VariableIndex(1))          │
+└                                 ┘</code></pre><p>Note the automatic conversion from <code>x</code> to <code>+(x)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/functions.jl#L669-L712">source</a></section></article><h2 id="Sets"><a class="docs-heading-anchor" href="#Sets">Sets</a><a id="Sets-1"></a><a class="docs-heading-anchor-permalink" href="#Sets" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractSet" href="#MathOptInterface.AbstractSet"><code>MathOptInterface.AbstractSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractSet</code></pre><p>Abstract supertype for set objects used to encode constraints.</p><p><strong>Required methods</strong></p><p>For sets of type <code>S</code> with <code>isbitstype(S) == false</code>, you must implement:</p><ul><li><code>Base.copy(set::S)</code></li><li><code>Base.:(==)(x::S, y::S)</code></li></ul><p>Subtypes of <code>AbstractSet</code> such as <a href="#MathOptInterface.AbstractScalarSet"><code>AbstractScalarSet</code></a> and <a href="#MathOptInterface.AbstractVectorSet"><code>AbstractVectorSet</code></a> may prescribe additional required methods.</p><p><strong>Optional methods</strong></p><p>You may optionally implement:</p><ul><li><a href="#MathOptInterface.dual_set"><code>dual_set</code></a></li><li><a href="#MathOptInterface.dual_set_type"><code>dual_set_type</code></a></li></ul><p><strong>Note for developers</strong></p><p>When creating a new set, the set struct must not contain any <a href="#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> or <a href="../constraints/#MathOptInterface.ConstraintIndex"><code>ConstraintIndex</code></a> objects.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L9-L35">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractScalarSet" href="#MathOptInterface.AbstractScalarSet"><code>MathOptInterface.AbstractScalarSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractScalarSet</code></pre><p>Abstract supertype for subsets of <span>$\mathbb{R}$</span>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L121-L125">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractVectorSet" href="#MathOptInterface.AbstractVectorSet"><code>MathOptInterface.AbstractVectorSet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractVectorSet</code></pre><p>Abstract supertype for subsets of <span>$\mathbb{R}^n$</span> for some <span>$n$</span>.</p><p><strong>Required methods</strong></p><p>All <code>AbstractVectorSet</code>s of type <code>S</code> must implement:</p><ul><li><a href="#MathOptInterface.dimension"><code>dimension</code></a>, unless the dimension is stored in the <code>set.dimension</code> field</li><li><a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.set_dot"><code>Utilities.set_dot</code></a>, unless the dot product between two vectors in the set is equivalent to <code>LinearAlgebra.dot</code>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L132-L145">source</a></section></article><h3 id="Utilities"><a class="docs-heading-anchor" href="#Utilities">Utilities</a><a id="Utilities-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.dimension" href="#MathOptInterface.dimension"><code>MathOptInterface.dimension</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dimension(set::AbstractSet)</code></pre><p>Return the <a href="#MathOptInterface.output_dimension"><code>output_dimension</code></a> that an <a href="#MathOptInterface.AbstractFunction"><code>AbstractFunction</code></a> should have to be used with the set <code>set</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.dimension(MOI.Reals(4))
+4
+
+julia&gt; MOI.dimension(MOI.LessThan(3.0))
+1
+
+julia&gt; MOI.dimension(MOI.PositiveSemidefiniteConeTriangle(2))
+3</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L45-L65">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.dual_set" href="#MathOptInterface.dual_set"><code>MathOptInterface.dual_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dual_set(set::AbstractSet)</code></pre><p>Return the dual set of <code>set</code>, that is the dual cone of the set. This follows the definition of duality discussed in <a href="../../background/duality/#Duality">Duality</a>.</p><p>See <a href="https://en.wikipedia.org/wiki/Dual_cone_and_polar_cone">Dual cone</a> for more information.</p><p>If the dual cone is not defined it returns an error.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.dual_set(MOI.Reals(4))
+MathOptInterface.Zeros(4)
+
+julia&gt; MOI.dual_set(MOI.SecondOrderCone(5))
+MathOptInterface.SecondOrderCone(5)
+
+julia&gt; MOI.dual_set(MOI.ExponentialCone())
+MathOptInterface.DualExponentialCone()</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L68-L93">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.dual_set_type" href="#MathOptInterface.dual_set_type"><code>MathOptInterface.dual_set_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">dual_set_type(S::Type{&lt;:AbstractSet})</code></pre><p>Return the type of dual set of sets of type <code>S</code>, as returned by <a href="#MathOptInterface.dual_set"><code>dual_set</code></a>. If the dual cone is not defined it returns an error.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.dual_set_type(MOI.Reals)
+MathOptInterface.Zeros
+
+julia&gt; MOI.dual_set_type(MOI.SecondOrderCone)
+MathOptInterface.SecondOrderCone
+
+julia&gt; MOI.dual_set_type(MOI.ExponentialCone)
+MathOptInterface.DualExponentialCone</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L96-L116">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.constant-Tuple{MathOptInterface.EqualTo}" href="#MathOptInterface.constant-Tuple{MathOptInterface.EqualTo}"><code>MathOptInterface.constant</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">constant(set::Union{EqualTo,GreaterThan,LessThan,Parameter})</code></pre><p>Returns the constant term of the set <code>set</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.constant(MOI.GreaterThan(1.0))
+1.0
+
+julia&gt; MOI.constant(MOI.LessThan(2.5))
+2.5
+
+julia&gt; MOI.constant(MOI.EqualTo(3))
+3
+
+julia&gt; MOI.constant(MOI.Parameter(4.5))
+4.5</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2827-L2849">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.supports_dimension_update" href="#MathOptInterface.supports_dimension_update"><code>MathOptInterface.supports_dimension_update</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_dimension_update(S::Type{&lt;:MOI.AbstractVectorSet})</code></pre><p>Return a <code>Bool</code> indicating whether the elimination of any dimension of <code>n</code>-dimensional sets of type <code>S</code> give an <code>n-1</code>-dimensional set <code>S</code>. By default, this function returns <code>false</code> so it should only be implemented for sets that supports dimension update.</p><p>For instance, <code>supports_dimension_update(MOI.Nonnegatives)</code> is <code>true</code> because the elimination of any dimension of the <code>n</code>-dimensional nonnegative orthant gives the <code>n-1</code>-dimensional nonnegative orthant. However <code>supports_dimension_update(MOI.ExponentialCone)</code> is <code>false</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2795-L2807">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.update_dimension" href="#MathOptInterface.update_dimension"><code>MathOptInterface.update_dimension</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">update_dimension(s::AbstractVectorSet, new_dim::Int)</code></pre><p>Returns a set with the dimension modified to <code>new_dim</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2815-L2819">source</a></section></article><h2 id="Scalar-sets"><a class="docs-heading-anchor" href="#Scalar-sets">Scalar sets</a><a id="Scalar-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Scalar-sets" title="Permalink"></a></h2><p>List of recognized scalar sets.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.GreaterThan" href="#MathOptInterface.GreaterThan"><code>MathOptInterface.GreaterThan</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GreaterThan{T&lt;:Real}(lower::T)</code></pre><p>The set <span>$[lower, \infty) \subseteq \mathbb{R}$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.GreaterThan(0.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L150-L169">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LessThan" href="#MathOptInterface.LessThan"><code>MathOptInterface.LessThan</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LessThan{T&lt;:Real}(upper::T)</code></pre><p>The set <span>$(-\infty, upper] \subseteq \mathbb{R}$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.LessThan(2.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.LessThan{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L176-L195">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.EqualTo" href="#MathOptInterface.EqualTo"><code>MathOptInterface.EqualTo</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">EqualTo{T&lt;:Number}(value::T)</code></pre><p>The set containing the single point <span>$\{value\} \subseteq \mathbb{R}$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.EqualTo(2.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.EqualTo{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L202-L221">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Interval" href="#MathOptInterface.Interval"><code>MathOptInterface.Interval</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Interval{T&lt;:Real}(lower::T, upper::T)</code></pre><p>The interval <span>$[lower, upper] \subseteq \mathbb{R} \cup \{-\infty, +\infty\}$</span>.</p><p>If <code>lower</code> or <code>upper</code> is <code>-Inf</code> or <code>Inf</code>, respectively, the set is interpreted as a one-sided interval.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.Interval(1.0, 2.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L266-L288">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Integer" href="#MathOptInterface.Integer"><code>MathOptInterface.Integer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Integer()</code></pre><p>The set of integers, <span>$\mathbb{Z}$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.Integer())
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L330-L349">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ZeroOne" href="#MathOptInterface.ZeroOne"><code>MathOptInterface.ZeroOne</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ZeroOne()</code></pre><p>The set <span>$\{0, 1\}$</span>.</p><p>Variables belonging to the <code>ZeroOne</code> set are also known as &quot;binary&quot; variables.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.ZeroOne())
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L352-L373">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Semicontinuous" href="#MathOptInterface.Semicontinuous"><code>MathOptInterface.Semicontinuous</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Semicontinuous{T&lt;:Real}(lower::T, upper::T)</code></pre><p>The set <span>$\{0\} \cup [lower, upper]$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.Semicontinuous(2.0, 3.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Semicontinuous{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L376-L395">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Semiinteger" href="#MathOptInterface.Semiinteger"><code>MathOptInterface.Semiinteger</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Semiinteger{T&lt;:Real}(lower::T, upper::T)</code></pre><p>The set <span>$\{0\} \cup \{lower, lower+1, \ldots, upper-1, upper\}$</span>.</p><p>Note that if <code>lower</code> and <code>upper</code> are not equivalent to an integer, then the solver may throw an error, or it may round up <code>lower</code> and round down <code>upper</code> to the nearest integers.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(model, x, MOI.Semiinteger(2.0, 3.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Semiinteger{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L405-L428">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Parameter" href="#MathOptInterface.Parameter"><code>MathOptInterface.Parameter</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Parameter{T&lt;:Number}(value::T)</code></pre><p>The set containing the single point <span>$\{value\} \subseteq \mathbb{R}$</span>.</p><p>The <code>Parameter</code> set is conceptually similar to the <a href="#MathOptInterface.EqualTo"><code>EqualTo</code></a> set, except that a variable constrained to the <code>Parameter</code> set cannot have other constraints added to it, and the <code>Parameter</code> set can never be deleted. Thus, solvers are free to treat the variable as a constant, and they need not add it as a decision variable to the model.</p><p>Because of this behavior, you must add parameters using <a href="../variables/#MathOptInterface.add_constrained_variable"><code>add_constrained_variable</code></a>, and solvers should declare <a href="../variables/#MathOptInterface.supports_add_constrained_variable"><code>supports_add_constrained_variable</code></a> and not <a href="../constraints/#MathOptInterface.supports_constraint"><code>supports_constraint</code></a> for the <code>Parameter</code> set.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; p, ci = MOI.add_constrained_variable(model, MOI.Parameter(2.5))
+(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Parameter{Float64}}(1))
+
+julia&gt; MOI.set(model, MOI.ConstraintSet(), ci, MOI.Parameter(3.0))
+
+julia&gt; MOI.get(model, MOI.ConstraintSet(), ci)
+MathOptInterface.Parameter{Float64}(3.0)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L228-L259">source</a></section></article><h2 id="Vector-sets"><a class="docs-heading-anchor" href="#Vector-sets">Vector sets</a><a id="Vector-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Vector-sets" title="Permalink"></a></h2><p>List of recognized vector sets.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Reals" href="#MathOptInterface.Reals"><code>MathOptInterface.Reals</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Reals(dimension::Int)</code></pre><p>The set <span>$\mathbb{R}^{dimension}$</span> (containing all points) of non-negative dimension <code>dimension</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Reals(3))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Reals}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L438-L457">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Zeros" href="#MathOptInterface.Zeros"><code>MathOptInterface.Zeros</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Zeros(dimension::Int)</code></pre><p>The set <span>$\{ 0 \}^{dimension}$</span> (containing only the origin) of non-negative dimension <code>dimension</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Zeros(3))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Zeros}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L475-L494">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonnegatives" href="#MathOptInterface.Nonnegatives"><code>MathOptInterface.Nonnegatives</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Nonnegatives(dimension::Int)</code></pre><p>The nonnegative orthant <span>$\{ x \in \mathbb{R}^{dimension} : x \ge 0 \}$</span> of non-negative dimension <code>dimension</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Nonnegatives(3))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonnegatives}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L512-L531">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonpositives" href="#MathOptInterface.Nonpositives"><code>MathOptInterface.Nonpositives</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Nonpositives(dimension::Int)</code></pre><p>The nonpositive orthant <span>$\{ x \in \mathbb{R}^{dimension} : x \le 0 \}$</span> of non-negative dimension <code>dimension</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Nonpositives(3))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonpositives}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L549-L568">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NormInfinityCone" href="#MathOptInterface.NormInfinityCone"><code>MathOptInterface.NormInfinityCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormInfinityCone(dimension::Int)</code></pre><p>The <span>$\ell_\infty$</span>-norm cone <span>$\{ (t,x) \in \mathbb{R}^{dimension} : t \ge \lVert x \rVert_\infty = \max_i \lvert x_i \rvert \}$</span> of dimension <code>dimension</code>.</p><p>The <code>dimension</code> must be at least <code>1</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.NormInfinityCone(4))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormInfinityCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L586-L610">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NormOneCone" href="#MathOptInterface.NormOneCone"><code>MathOptInterface.NormOneCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormOneCone(dimension::Int)</code></pre><p>The <span>$\ell_1$</span>-norm cone <span>$\{ (t,x) \in \mathbb{R}^{dimension} : t \ge \lVert x \rVert_1 = \sum_i \lvert x_i \rvert \}$</span> of dimension <code>dimension</code>.</p><p>The <code>dimension</code> must be at least <code>1</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.NormOneCone(4))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormOneCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L629-L653">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NormCone" href="#MathOptInterface.NormCone"><code>MathOptInterface.NormCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormCone(p::Float64, dimension::Int)</code></pre><p>The <span>$\ell_p$</span>-norm cone <span>$\{ (t,x) \in \mathbb{R}^{dimension} : t \ge \left(\sum\limits_i |x_i|^p\right)^{\frac{1}{p}} \}$</span> of dimension <code>dimension</code>.</p><p>The <code>dimension</code> must be at least <code>1</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.NormCone(3, 4))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L765-L789">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SecondOrderCone" href="#MathOptInterface.SecondOrderCone"><code>MathOptInterface.SecondOrderCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SecondOrderCone(dimension::Int)</code></pre><p>The second-order cone (or Lorenz cone or <span>$\ell_2$</span>-norm cone) <span>$\{ (t,x) \in \mathbb{R}^{dimension} : t \ge \lVert x \rVert_2 \}$</span> of dimension <code>dimension</code>.</p><p>The <code>dimension</code> must be at least <code>1</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.SecondOrderCone(4))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.SecondOrderCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L671-L696">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RotatedSecondOrderCone" href="#MathOptInterface.RotatedSecondOrderCone"><code>MathOptInterface.RotatedSecondOrderCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RotatedSecondOrderCone(dimension::Int)</code></pre><p>The rotated second-order cone <span>$\{ (t,u,x) \in \mathbb{R}^{dimension} : 2tu \ge \lVert x \rVert_2^2, t,u \ge 0 \}$</span> of dimension <code>dimension</code>.</p><p>The <code>dimension</code> must be at least <code>2</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; u = MOI.add_variable(model)
+MOI.VariableIndex(2)
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; u; x]),
+           MOI.RotatedSecondOrderCone(5),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RotatedSecondOrderCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L715-L746">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.GeometricMeanCone" href="#MathOptInterface.GeometricMeanCone"><code>MathOptInterface.GeometricMeanCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GeometricMeanCone(dimension::Int)</code></pre><p>The geometric mean cone <span>$\{ (t,x) \in \mathbb{R}^{n+1} : x \ge 0, t \le \sqrt[n]{x_1 x_2 \cdots x_n} \}$</span>, where <code>dimension = n + 1 &gt;= 2</code>.</p><p><strong>Duality note</strong></p><p>The dual of the geometric mean cone is <span>$\{ (u, v) \in \mathbb{R}^{n+1} : u \le 0, v \ge 0, -u \le n \sqrt[n]{\prod_i v_i} \}$</span>, where <code>dimension = n + 1 &gt;= 2</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; x]),
+           MOI.GeometricMeanCone(4),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.GeometricMeanCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L813-L846">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.ExponentialCone" href="#MathOptInterface.ExponentialCone"><code>MathOptInterface.ExponentialCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ExponentialCone()</code></pre><p>The 3-dimensional exponential cone <span>$\{ (x,y,z) \in \mathbb{R}^3 : y \exp (x/y) \le z, y &gt; 0 \}$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.ExponentialCone())
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.ExponentialCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L862-L880">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.DualExponentialCone" href="#MathOptInterface.DualExponentialCone"><code>MathOptInterface.DualExponentialCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">DualExponentialCone()</code></pre><p>The 3-dimensional dual exponential cone <span>$\{ (u,v,w) \in \mathbb{R}^3 : -u \exp (v/u) \le \exp(1) w, u &lt; 0 \}$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.DualExponentialCone())
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.DualExponentialCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L888-L906">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.PowerCone" href="#MathOptInterface.PowerCone"><code>MathOptInterface.PowerCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">PowerCone{T&lt;:Real}(exponent::T)</code></pre><p>The 3-dimensional power cone <span>$\{ (x,y,z) \in \mathbb{R}^3 : x^{exponent} y^{1-exponent} \ge |z|, x \ge 0, y \ge 0 \}$</span> with parameter <code>exponent</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.PowerCone(0.5))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.PowerCone{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L914-L933">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.DualPowerCone" href="#MathOptInterface.DualPowerCone"><code>MathOptInterface.DualPowerCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">DualPowerCone{T&lt;:Real}(exponent::T)</code></pre><p>The 3-dimensional power cone <span>$\{ (u,v,w) \in \mathbb{R}^3 : (\frac{u}{exponent})^{exponent} (\frac{v}{1-exponent})^{1-exponent} \ge |w|, u \ge 0, v \ge 0 \}$</span> with parameter <code>exponent</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.DualPowerCone(0.5))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.DualPowerCone{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L947-L966">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RelativeEntropyCone" href="#MathOptInterface.RelativeEntropyCone"><code>MathOptInterface.RelativeEntropyCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RelativeEntropyCone(dimension::Int)</code></pre><p>The relative entropy cone <span>$\{ (u, v, w) \in \mathbb{R}^{1+2n} : u \ge \sum_{i=1}^n w_i \log(\frac{w_i}{v_i}), v_i \ge 0, w_i \ge 0 \}$</span>, where <code>dimension = 2n + 1 &gt;= 1</code>.</p><p><strong>Duality note</strong></p><p>The dual of the relative entropy cone is <span>$\{ (u, v, w) \in \mathbb{R}^{1+2n} : \forall i, w_i \ge u (\log (\frac{u}{v_i}) - 1), v_i \ge 0, u &gt; 0 \}$</span> of dimension <code>dimension</code><span>${}=2n+1$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; u = MOI.add_variable(model);
+
+julia&gt; v = MOI.add_variables(model, 3);
+
+julia&gt; w = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([u; v; w]),
+           MOI.RelativeEntropyCone(7),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RelativeEntropyCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L980-L1014">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NormSpectralCone" href="#MathOptInterface.NormSpectralCone"><code>MathOptInterface.NormSpectralCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormSpectralCone(row_dim::Int, column_dim::Int)</code></pre><p>The epigraph of the matrix spectral norm (maximum singular value function) <span>$\{ (t, X) \in \mathbb{R}^{1 + row_dim \times column_dim} : t \ge \sigma_1(X) \}$</span>, where <span>$\sigma_i$</span> is the <span>$i$</span>th singular value of the matrix <span>$X$</span> of non-negative row dimension <code>row_dim</code> and column dimension <code>column_dim</code>.</p><p>The matrix X is vectorized by stacking the columns, matching the behavior of Julia&#39;s <code>vec</code> function.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; X = reshape(MOI.add_variables(model, 6), 2, 3)
+2×3 Matrix{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(2)  MOI.VariableIndex(4)  MOI.VariableIndex(6)
+ MOI.VariableIndex(3)  MOI.VariableIndex(5)  MOI.VariableIndex(7)
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; vec(X)]),
+           MOI.NormSpectralCone(2, 3),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormSpectralCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1030-L1064">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.NormNuclearCone" href="#MathOptInterface.NormNuclearCone"><code>MathOptInterface.NormNuclearCone</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormNuclearCone(row_dim::Int, column_dim::Int)</code></pre><p>The epigraph of the matrix nuclear norm (sum of singular values function) <span>$\{ (t, X) \in \mathbb{R}^{1 + row_dim \times column_dim} : t \ge \sum_i \sigma_i(X) \}$</span>, where <span>$\sigma_i$</span> is the <span>$i$</span>th singular value of the matrix <span>$X$</span> of non-negative row dimension <code>row_dim</code> and column dimension <code>column_dim</code>.</p><p>The matrix X is vectorized by stacking the columns, matching the behavior of Julia&#39;s <code>vec</code> function.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; X = reshape(MOI.add_variables(model, 6), 2, 3)
+2×3 Matrix{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(2)  MOI.VariableIndex(4)  MOI.VariableIndex(6)
+ MOI.VariableIndex(3)  MOI.VariableIndex(5)  MOI.VariableIndex(7)
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; vec(X)]),
+           MOI.NormNuclearCone(2, 3),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormNuclearCone}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1086-L1120">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SOS1" href="#MathOptInterface.SOS1"><code>MathOptInterface.SOS1</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOS1{T&lt;:Real}(weights::Vector{T})</code></pre><p>The set corresponding to the Special Ordered Set (SOS) constraint of Type I.</p><p>Of the variables in the set, at most one can be nonzero.</p><p>The <code>weights</code> induce an ordering of the variables such that the <em>k</em>th element in the set corresponds to the <em>k</em>th weight in <code>weights</code>. Solvers may use these weights to improve the efficiency of the solution process, but the ordering does not change the set of feasible solutions.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables(x),
+           MOI.SOS1([1.0, 3.0, 2.5]),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.SOS1{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1825-L1854">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.SOS2" href="#MathOptInterface.SOS2"><code>MathOptInterface.SOS2</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOS2{T&lt;:Real}(weights::Vector{T})</code></pre><p>The set corresponding to the Special Ordered Set (SOS) constraint of Type II.</p><p>The <code>weights</code> induce an ordering of the variables such that the <em>k</em>th element in the set corresponds to the <em>k</em>th weight in <code>weights</code>. Therefore, the weights must be unique.</p><p>Of the variables in the set, at most two can be nonzero, and if two are nonzero, they must be adjacent in the ordering of the set.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables(x),
+           MOI.SOS2([1.0, 3.0, 2.5]),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.SOS2{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1865-L1894">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Indicator" href="#MathOptInterface.Indicator"><code>MathOptInterface.Indicator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Indicator{A&lt;:ActivationCondition,S&lt;:AbstractScalarSet}(set::S)</code></pre><p>The set corresponding to an indicator constraint.</p><p>When <code>A</code> is <code>ACTIVATE_ON_ZERO</code>, this means: <span>$\{(y, x) \in \{0, 1\} \times \mathbb{R}^n : y = 0 \implies x \in set\}$</span></p><p>When <code>A</code> is <code>ACTIVATE_ON_ONE</code>, this means: <span>$\{(y, x) \in \{0, 1\} \times \mathbb{R}^n : y = 1 \implies x \in set\}$</span></p><p><strong>Notes</strong></p><p>Most solvers expect that the first row of the function is interpretable as a variable index <code>x_i</code> (e.g., <code>1.0 * x + 0.0</code>). An error will be thrown if this is not the case.</p><p><strong>Example</strong></p><p>The constraint <span>$\{(y, x) \in \{0, 1\} \times \mathbb{R}^2 : y = 1 \implies x_1 + x_2 \leq 9 \}$</span> is defined as</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 2)
+2-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+
+julia&gt; y, _ = MOI.add_constrained_variable(model, MOI.ZeroOne())
+(MOI.VariableIndex(3), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(3))
+
+julia&gt; f = MOI.VectorAffineFunction(
+           [
+               MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, y)),
+               MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(1.0, x[1])),
+               MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(1.0, x[2])),
+           ],
+           [0.0, 0.0],
+       )
+┌                                                         ┐
+│0.0 + 1.0 MOI.VariableIndex(3)                           │
+│0.0 + 1.0 MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(2)│
+└                                                         ┘
+
+julia&gt; s = MOI.Indicator{MOI.ACTIVATE_ON_ONE}(MOI.LessThan(9.0))
+MathOptInterface.Indicator{MathOptInterface.ACTIVATE_ON_ONE, MathOptInterface.LessThan{Float64}}(MathOptInterface.LessThan{Float64}(9.0))
+
+julia&gt; MOI.add_constraint(model, f, s)
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Indicator{MathOptInterface.ACTIVATE_ON_ONE, MathOptInterface.LessThan{Float64}}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1917-L1972">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Complements" href="#MathOptInterface.Complements"><code>MathOptInterface.Complements</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Complements(dimension::Base.Integer)</code></pre><p>The set corresponding to a mixed complementarity constraint.</p><p>Complementarity constraints should be specified with an <a href="#MathOptInterface.AbstractVectorFunction"><code>AbstractVectorFunction</code></a>-in-<code>Complements(dimension)</code> constraint.</p><p>The dimension of the vector-valued function <code>F</code> must be <code>dimension</code>. This defines a complementarity constraint between the scalar function <code>F[i]</code> and the variable in <code>F[i + dimension/2]</code>. Thus, <code>F[i + dimension/2]</code> must be interpretable as a single variable <code>x_i</code> (e.g., <code>1.0 * x + 0.0</code>), and <code>dimension</code> must be even.</p><p>The mixed complementarity problem consists of finding <code>x_i</code> in the interval <code>[lb, ub]</code> (i.e., in the set <code>Interval(lb, ub)</code>), such that the following holds:</p><ol><li><code>F_i(x) == 0</code> if <code>lb_i &lt; x_i &lt; ub_i</code></li><li><code>F_i(x) &gt;= 0</code> if <code>lb_i == x_i</code></li><li><code>F_i(x) &lt;= 0</code> if <code>x_i == ub_i</code></li></ol><p>Classically, the bounding set for <code>x_i</code> is <code>Interval(0, Inf)</code>, which recovers: <code>0 &lt;= F_i(x) ⟂ x_i &gt;= 0</code>, where the <code>⟂</code> operator implies <code>F_i(x) * x_i = 0</code>.</p><p><strong>Example</strong></p><p>The problem:</p><pre><code class="nohighlight hljs">x -in- Interval(-1, 1)
+[-4 * x - 3, x] -in- Complements(2)</code></pre><p>defines the mixed complementarity problem where the following holds:</p><ol><li><code>-4 * x - 3 == 0</code> if <code>-1 &lt; x &lt; 1</code></li><li><code>-4 * x - 3 &gt;= 0</code> if <code>x == -1</code></li><li><code>-4 * x - 3 &lt;= 0</code> if <code>x == 1</code></li></ol><p>There are three solutions:</p><ol><li><code>x = -3/4</code> with <code>F(x) = 0</code></li><li><code>x = -1</code> with <code>F(x) = 1</code></li><li><code>x = 1</code> with <code>F(x) = -7</code></li></ol><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x, _ = MOI.add_constrained_variable(model, MOI.Interval(-1.0, 1.0));
+
+julia&gt; MOI.add_constraint(
+            model,
+            MOI.Utilities.vectorize([-4.0 * x - 3.0, x]),
+            MOI.Complements(2),
+        )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Complements}(1)</code></pre><p>The function <code>F</code> can also be defined in terms of single variables. For example, the problem:</p><pre><code class="nohighlight hljs">[x_3, x_4] -in- Nonnegatives(2)
+[x_1, x_2, x_3, x_4] -in- Complements(4)</code></pre><p>defines the complementarity problem where <code>0 &lt;= x_1 ⟂ x_3 &gt;= 0</code> and <code>0 &lt;= x_2 ⟂ x_4 &gt;= 0</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 4);
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x[3:4]), MOI.Nonnegatives(2))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonnegatives}(1)
+
+julia&gt; MOI.add_constraint(
+            model,
+            MOI.VectorOfVariables(x),
+            MOI.Complements(4),
+        )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Complements}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1984-L2070">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HyperRectangle" href="#MathOptInterface.HyperRectangle"><code>MathOptInterface.HyperRectangle</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HyperRectangle(lower::Vector{T}, upper::Vector{T}) where {T}</code></pre><p>The set <span>$\{x \in \bar{\mathbb{R}}^d: x_i \in [lower_i, upper_i] \forall i=1,\ldots,d\}$</span>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3)
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables(x),
+           MOI.HyperRectangle(zeros(3), ones(3)),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.HyperRectangle{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2685-L2711">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Scaled" href="#MathOptInterface.Scaled"><code>MathOptInterface.Scaled</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct Scaled{S&lt;:AbstractVectorSet} &lt;: AbstractVectorSet
+    set::S
+end</code></pre><p>Given a vector <span>$a \in \mathbb{R}^d$</span> and a <code>set</code> representing the set <span>$\mathcal{S} \in \mathbb{R}^d$</span> such that <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.set_dot"><code>Utilities.set_dot</code></a> for <span>$x \in \mathcal{S}$</span> and <span>$y \in \mathcal{S}^*$</span> is</p><p class="math-container">\[\sum_{i=1}^d a_i x_i y_i\]</p><p>the set <code>Scaled(set)</code> is defined as</p><p class="math-container">\[\{ (\sqrt{a_1} x_1, \sqrt{a_2} x_2, \ldots, \sqrt{a_d} x_d) : x \in S \}\]</p><p><strong>Example</strong></p><p>This can be used to scale a vector of numbers</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; set = MOI.PositiveSemidefiniteConeTriangle(2)
+MathOptInterface.PositiveSemidefiniteConeTriangle(2)
+
+julia&gt; a = MOI.Utilities.SetDotScalingVector{Float64}(set)
+3-element MathOptInterface.Utilities.SetDotScalingVector{Float64, MathOptInterface.PositiveSemidefiniteConeTriangle}:
+ 1.0
+ 1.4142135623730951
+ 1.0
+
+julia&gt; using LinearAlgebra
+
+julia&gt; MOI.Utilities.operate(*, Float64, Diagonal(a), ones(3))
+3-element Vector{Float64}:
+ 1.0
+ 1.4142135623730951
+ 1.0</code></pre><p>It can be also used to scale a vector of function</p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3);
+
+julia&gt; func = MOI.VectorOfVariables(x)
+┌                    ┐
+│MOI.VariableIndex(1)│
+│MOI.VariableIndex(2)│
+│MOI.VariableIndex(3)│
+└                    ┘
+
+julia&gt; set = MOI.PositiveSemidefiniteConeTriangle(2)
+MathOptInterface.PositiveSemidefiniteConeTriangle(2)
+
+julia&gt; MOI.Utilities.operate(*, Float64, Diagonal(a), func)
+┌                                             ┐
+│0.0 + 1.0 MOI.VariableIndex(1)               │
+│0.0 + 1.4142135623730951 MOI.VariableIndex(2)│
+│0.0 + 1.0 MOI.VariableIndex(3)               │
+└                                             ┘</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1711-L1775">source</a></section></article><h2 id="Constraint-programming-sets"><a class="docs-heading-anchor" href="#Constraint-programming-sets">Constraint programming sets</a><a id="Constraint-programming-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-programming-sets" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AllDifferent" href="#MathOptInterface.AllDifferent"><code>MathOptInterface.AllDifferent</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AllDifferent(dimension::Int)</code></pre><p>The set <span>$\{x \in \mathbb{Z}^{d}\}$</span> such that no two elements in <span>$x$</span> take the same value and <code>dimension = d</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>all_different</code> in MiniZinc, and is sometimes also called <code>distinct</code>.</p><p><strong>Example</strong></p><p>To enforce <code>x[1] != x[2]</code> AND <code>x[1] != x[3]</code> AND <code>x[2] != x[3]</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.AllDifferent(3))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.AllDifferent}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2085-L2115">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.BinPacking" href="#MathOptInterface.BinPacking"><code>MathOptInterface.BinPacking</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">BinPacking(c::T, w::Vector{T}) where {T}</code></pre><p>The set <span>$\{x \in \mathbb{Z}^d\}$</span> where <code>d = length(w)</code>, such that each item <code>i</code> in <code>1:d</code> of weight <code>w[i]</code> is put into bin <code>x[i]</code>, and the total weight of each bin does not exceed <code>c</code>.</p><p>There are additional assumptions that the capacity, <code>c</code>, and the weights, <code>w</code>, must all be non-negative.</p><p>The bin numbers depend on the bounds of <code>x</code>, so they may be something other than the integers <code>1:d</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>bin_packing</code> in MiniZinc.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; bins = MOI.add_variables(model, 5)
+5-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+ MOI.VariableIndex(4)
+ MOI.VariableIndex(5)
+
+julia&gt; weights = Float64[1, 1, 2, 2, 3]
+5-element Vector{Float64}:
+ 1.0
+ 1.0
+ 2.0
+ 2.0
+ 3.0
+
+julia&gt; MOI.add_constraint.(model, bins, MOI.Integer())
+5-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}}:
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(3)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(4)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(5)
+
+julia&gt; MOI.add_constraint.(model, bins, MOI.Interval(4.0, 6.0))
+5-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}}:
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(2)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(3)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(4)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(5)
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(bins), MOI.BinPacking(3.0, weights))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.BinPacking{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2126-L2186">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Circuit" href="#MathOptInterface.Circuit"><code>MathOptInterface.Circuit</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Circuit(dimension::Int)</code></pre><p>The set <span>$\{x \in \{1..d\}^d\}$</span> that constraints <span>$x$</span> to be a circuit, such that <span>$x_i = j$</span> means that <span>$j$</span> is the successor of <span>$i$</span>, and <code>dimension = d</code>.</p><p>Graphs with multiple independent circuits, such as <code>[2, 1, 3]</code> and <code>[2, 1, 4, 3]</code>, are not valid.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>circuit</code> in MiniZinc, and it is equivalent to forming a (potentially sub-optimal) tour in the travelling salesperson problem.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Circuit(3))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Circuit}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2211-L2243">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CountAtLeast" href="#MathOptInterface.CountAtLeast"><code>MathOptInterface.CountAtLeast</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountAtLeast(n::Int, d::Vector{Int}, set::Set{Int})</code></pre><p>The set <span>$\{x \in \mathbb{Z}^{d_1 + d_2 + \ldots d_N}\}$</span>, where <code>x</code> is partitioned into <code>N</code> subsets (<span>$\{x_1,  \ldots, x_{d_1}\}$</span>, <span>$\{x_{d_1 + 1},  \ldots, x_{d_1 + d_2}\}$</span> and so on), and at least <span>$n$</span> elements of each subset take one of the values in <code>set</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>at_least</code> in MiniZinc.</p><p><strong>Example</strong></p><p>To ensure that <code>3</code> appears at least once in each of the subsets <code>{a, b}</code> and <code>{b, c}</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; a, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))
+
+julia&gt; b, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(2), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2))
+
+julia&gt; c, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(3), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(3))
+
+julia&gt; x, d, set = [a, b, b, c], [2, 2], [3]
+(MathOptInterface.VariableIndex[MOI.VariableIndex(1), MOI.VariableIndex(2), MOI.VariableIndex(2), MOI.VariableIndex(3)], [2, 2], [3])
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.CountAtLeast(1, d, Set(set)))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountAtLeast}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2254-L2292">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CountBelongs" href="#MathOptInterface.CountBelongs"><code>MathOptInterface.CountBelongs</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountBelongs(dimenson::Int, set::Set{Int})</code></pre><p>The set <span>$\{(n, x) \in \mathbb{Z}^{1+d}\}$</span>, such that <code>n</code> elements of the vector <code>x</code> take on of the values in <code>set</code> and <code>dimension = 1 + d</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>among</code> by MiniZinc.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; n, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))
+
+julia&gt; x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+ MOI.VariableIndex(4)
+
+julia&gt; set = Set([3, 4, 5])
+Set{Int64} with 3 elements:
+  5
+  4
+  3
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([n; x]), MOI.CountBelongs(4, set))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountBelongs}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2319-L2355">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CountDistinct" href="#MathOptInterface.CountDistinct"><code>MathOptInterface.CountDistinct</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountDistinct(dimension::Int)</code></pre><p>The set <span>$\{(n, x) \in \mathbb{Z}^{1+d}\}$</span>, such that the number of distinct values in <code>x</code> is <code>n</code> and <code>dimension = 1 + d</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>nvalues</code> in MiniZinc.</p><p><strong>Example</strong></p><p>To model:</p><ul><li>if <code>n == 1</code>`<code>, then</code>x[1] == x[2] == x[3]`</li><li>if <code>n == 2</code>, then<ul><li><code>x[1] == x[2] != x[3]</code> or</li><li><code>x[1] != x[2] == x[3]</code> or</li><li><code>x[1] == x[3] != x[2]</code></li></ul></li><li>if <code>n == 3</code>, then <code>x[1] != x[2]</code>, <code>x[2] != x[3]</code> and <code>x[3] != x[1]</code></li></ul><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; n, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))
+
+julia&gt; x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+ MOI.VariableIndex(4)
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(vcat(n, x)), MOI.CountDistinct(4))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountDistinct}(1)</code></pre><p><strong>Relationship to AllDifferent</strong></p><p>When the first element is <code>d</code>, <code>CountDistinct</code> is equivalent to an <a href="#MathOptInterface.AllDifferent"><code>AllDifferent</code></a> constraint.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2373-L2417">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.CountGreaterThan" href="#MathOptInterface.CountGreaterThan"><code>MathOptInterface.CountGreaterThan</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountGreaterThan(dimension::Int)</code></pre><p>The set <span>$\{(c, y, x) \in \mathbb{Z}^{1+1+d}\}$</span>, such that <code>c</code> is strictly greater than the number of occurances of <code>y</code> in <code>x</code> and <code>dimension = 1 + 1 + d</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>count_gt</code> in MiniZinc.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; c, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))
+
+julia&gt; y, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(2), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2))
+
+julia&gt; x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(3)
+ MOI.VariableIndex(4)
+ MOI.VariableIndex(5)
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([c; y; x]), MOI.CountGreaterThan(5))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountGreaterThan}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2428-L2461">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Cumulative" href="#MathOptInterface.Cumulative"><code>MathOptInterface.Cumulative</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Cumulative(dimension::Int)</code></pre><p>The set <span>$\{(s, d, r, b) \in \mathbb{Z}^{3n+1}\}$</span>, representing the <code>cumulative</code> global constraint, where <code>n == length(s) == length(r) == length(b)</code> and <code>dimension = 3n + 1</code>.</p><p><code>Cumulative</code> requires that a set of tasks given by start times <span>$s$</span>, durations <span>$d$</span>, and resource requirements <span>$r$</span>, never requires more than the global resource bound <span>$b$</span> at any one time.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>cumulative</code> in MiniZinc.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; s = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+
+julia&gt; d = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(4)
+ MOI.VariableIndex(5)
+ MOI.VariableIndex(6)
+
+julia&gt; r = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(7)
+ MOI.VariableIndex(8)
+ MOI.VariableIndex(9)
+
+julia&gt; b, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(10), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(10))
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([s; d; r; b]), MOI.Cumulative(10))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Cumulative}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2476-L2523">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Path" href="#MathOptInterface.Path"><code>MathOptInterface.Path</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Path(from::Vector{Int}, to::Vector{Int})</code></pre><p>Given a graph comprised of a set of nodes <code>1..N</code> and a set of arcs <code>1..E</code> represented by an edge from node <code>from[i]</code> to node <code>to[i]</code>, <code>Path</code> constrains the set <span>$(s, t, ns, es) \in (1..N)\times(1..E)\times\{0,1\}^N\times\{0,1\}^E$</span>, to form subgraph that is a path from node <code>s</code> to node <code>t</code>, where node <code>n</code> is in the path if <code>ns[n]</code> is <code>1</code>, and edge <code>e</code> is in the path if <code>es[e]</code> is <code>1</code>.</p><p>The path must be acyclic, and it must traverse all nodes <code>n</code> for which <code>ns[n]</code> is <code>1</code>, and all edges <code>e</code> for which <code>es[e]</code> is <code>1</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>path</code> in MiniZinc.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; N, E = 4, 5
+(4, 5)
+
+julia&gt; from = [1, 1, 2, 2, 3]
+5-element Vector{Int64}:
+ 1
+ 1
+ 2
+ 2
+ 3
+
+julia&gt; to = [2, 3, 3, 4, 4]
+5-element Vector{Int64}:
+ 2
+ 3
+ 3
+ 4
+ 4
+
+julia&gt; s, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))
+
+julia&gt; t, _ = MOI.add_constrained_variable(model, MOI.Integer())
+(MOI.VariableIndex(2), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2))
+
+julia&gt; ns = MOI.add_variables(model, N)
+4-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(3)
+ MOI.VariableIndex(4)
+ MOI.VariableIndex(5)
+ MOI.VariableIndex(6)
+
+julia&gt; MOI.add_constraint.(model, ns, MOI.ZeroOne())
+4-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}}:
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(3)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(4)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(5)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(6)
+
+julia&gt; es = MOI.add_variables(model, E)
+5-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(7)
+ MOI.VariableIndex(8)
+ MOI.VariableIndex(9)
+ MOI.VariableIndex(10)
+ MOI.VariableIndex(11)
+
+julia&gt; MOI.add_constraint.(model, es, MOI.ZeroOne())
+5-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}}:
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(7)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(8)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(9)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(10)
+ MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(11)
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables([s; t; ns; es]), MOI.Path(from, to))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Path}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2534-L2617">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Reified" href="#MathOptInterface.Reified"><code>MathOptInterface.Reified</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Reified(set::AbstractSet)</code></pre><p>The constraint <span>$[z; f(x)] \in Reified(S)$</span> ensures that <span>$f(x) \in S$</span> if and only if <span>$z == 1$</span>, where <span>$z \in \{0, 1\}$</span>.</p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())
+MOIU.UniversalFallback{MOIU.Model{Float64}}
+fallback for MOIU.Model{Float64}
+
+julia&gt; z, _ = MOI.add_constrained_variable(model, MOI.ZeroOne())
+(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(1))
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(2)
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.Utilities.vectorize([z, 2.0 * x]),
+           MOI.Reified(MOI.GreaterThan(1.0)),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Reified{MathOptInterface.GreaterThan{Float64}}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2738-L2764">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Table" href="#MathOptInterface.Table"><code>MathOptInterface.Table</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Table(table::Matrix{T}) where {T}</code></pre><p>The set <span>$\{x \in \mathbb{R}^d\}$</span> where <code>d = size(table, 2)</code>, such that <code>x</code> belongs to one row of <code>table</code>. That is, there exists some <code>j</code> in <code>1:size(table, 1)</code>, such that <code>x[i] = table[j, i]</code> for all <code>i=1:size(table, 2)</code>.</p><p><strong>Also known as</strong></p><p>This constraint is called <code>table</code> in MiniZinc.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variables(model, 3)
+3-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+ MOI.VariableIndex(2)
+ MOI.VariableIndex(3)
+
+julia&gt; table = Float64[1 1 0; 0 1 1; 1 0 1; 1 1 1]
+4×3 Matrix{Float64}:
+ 1.0  1.0  0.0
+ 0.0  1.0  1.0
+ 1.0  0.0  1.0
+ 1.0  1.0  1.0
+
+julia&gt; MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Table(table))
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Table{Float64}}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L2639-L2674">source</a></section></article><h2 id="Matrix-sets"><a class="docs-heading-anchor" href="#Matrix-sets">Matrix sets</a><a id="Matrix-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Matrix-sets" title="Permalink"></a></h2><p>Matrix sets are vectorized to be subtypes of <a href="#MathOptInterface.AbstractVectorSet"><code>AbstractVectorSet</code></a>.</p><p>For sets of symmetric matrices, storing both the <code>(i, j)</code> and <code>(j, i)</code> elements is redundant. Use the <a href="#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>AbstractSymmetricMatrixSetTriangle</code></a> set to represent only the vectorization of the upper triangular part of the matrix.</p><p>When the matrix of expressions constrained to be in the set is not symmetric, and hence additional constraints are needed to force the equality of the <code>(i, j)</code> and <code>(j, i)</code> elements, use the <a href="#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>AbstractSymmetricMatrixSetSquare</code></a> set.</p><p>The <a href="../../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SquareBridge"><code>Bridges.Constraint.SquareBridge</code></a> can transform a set from the square form to the <a href="#MathOptInterface.triangular_form"><code>triangular_form</code></a> by adding appropriate constraints if the <code>(i, j)</code> and <code>(j, i)</code> expressions are different.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractSymmetricMatrixSetTriangle" href="#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>MathOptInterface.AbstractSymmetricMatrixSetTriangle</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractSymmetricMatrixSetTriangle &lt;: AbstractVectorSet end</code></pre><p>Abstract supertype for subsets of the (vectorized) cone of symmetric matrices, with <a href="#MathOptInterface.side_dimension"><code>side_dimension</code></a> rows and columns. The entries of the upper-right triangular part of the matrix are given column by column (or equivalently, the entries of the lower-left triangular part are given row by row). A vectorized cone of <a href="#MathOptInterface.dimension"><code>dimension</code></a> <span>$n$</span> corresponds to a square matrix with side dimension <span>$\sqrt{1/4 + 2 n} - 1/2$</span>. (Because a <span>$d \times d$</span> matrix has <span>$d(d + 1) / 2$</span> elements in the upper or lower triangle.)</p><p><strong>Example</strong></p><p>The matrix</p><p class="math-container">\[\begin{bmatrix}
+  1 &amp; 2 &amp; 4\\
+  2 &amp; 3 &amp; 5\\
+  4 &amp; 5 &amp; 6
+\end{bmatrix}\]</p><p>has <a href="#MathOptInterface.side_dimension"><code>side_dimension</code></a> 3 and vectorization <span>$(1, 2, 3, 4, 5, 6)$</span>.</p><p><strong>Note</strong></p><p>Two packed storage formats exist for symmetric matrices, the respective orders of the entries are:</p><ul><li>upper triangular column by column (or lower triangular row by row);</li><li>lower triangular column by column (or upper triangular row by row).</li></ul><p>The advantage of the first format is the mapping between the <code>(i, j)</code> matrix indices and the <code>k</code> index of the vectorized form. It is simpler and does not depend on the side dimension of the matrix. Indeed,</p><ul><li>the entry of matrix indices <code>(i, j)</code> has vectorized index <code>k = div((j - 1) * j, 2) + i</code> if <span>$i \leq j$</span> and <code>k = div((i - 1) * i, 2) + j</code> if <span>$j \leq i$</span>;</li><li>and the entry with vectorized index <code>k</code> has matrix indices <code>i = div(1 + isqrt(8k - 7), 2)</code> and <code>j = k - div((i - 1) * i, 2)</code> or <code>j = div(1 + isqrt(8k - 7), 2)</code> and <code>i = k - div((j - 1) * j, 2)</code>.</li></ul><p><strong>Duality note</strong></p><p>The scalar product for the symmetric matrix in its vectorized form is the sum of the pairwise product of the diagonal entries plus twice the sum of the pairwise product of the upper diagonal entries; see [p. 634, 1]. This has important consequence for duality.</p><p>Consider for example the following problem (<a href="#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a> is a subtype of <a href="#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>AbstractSymmetricMatrixSetTriangle</code></a>)</p><p class="math-container">\[\begin{align*}
+    &amp; \max_{x \in \mathbb{R}} &amp; x
+    \\
+    &amp; \;\;\text{s.t.} &amp;
+    (1, -x, 1) &amp; \in \text{PositiveSemidefiniteConeTriangle}(2).
+\end{align*}\]</p><p>The dual is the following problem</p><p class="math-container">\[\begin{align*}
+    &amp; \min_{x \in \mathbb{R}^3} &amp; y_1 + y_3
+    \\
+    &amp; \;\;\text{s.t.} &amp; 2y_2 &amp; = 1\\
+    &amp; &amp; y &amp; \in \text{PositiveSemidefiniteConeTriangle}(2).
+\end{align*}\]</p><p>Why do we use <span>$2y_2$</span> in the dual constraint instead of <span>$y_2$</span> ? The reason is that <span>$2y_2$</span> is the scalar product between <span>$y$</span> and the symmetric matrix whose vectorized form is <span>$(0, 1, 0)$</span>. Indeed, with our modified scalar products we have</p><p class="math-container">\[\langle
+(0, 1, 0),
+(y_1, y_2, y_3)
+\rangle
+=
+\mathrm{trace}
+\begin{pmatrix}
+  0 &amp; 1\\
+  1 &amp; 0
+\end{pmatrix}
+\begin{pmatrix}
+  y_1 &amp; y_2\\
+  y_2 &amp; y_3
+\end{pmatrix}
+= 2y_2.\]</p><p><strong>References</strong></p><p>[1] Boyd, S. and Vandenberghe, L.. <em>Convex optimization</em>. Cambridge university     press, 2004.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1142-L1236">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractSymmetricMatrixSetSquare" href="#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>MathOptInterface.AbstractSymmetricMatrixSetSquare</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractSymmetricMatrixSetSquare &lt;: AbstractVectorSet end</code></pre><p>Abstract supertype for subsets of the (vectorized) cone of symmetric matrices, with <a href="#MathOptInterface.side_dimension"><code>side_dimension</code></a> rows and columns. The entries of the matrix are given column by column (or equivalently, row by row). The matrix is both constrained to be symmetric and to have its <a href="#MathOptInterface.triangular_form"><code>triangular_form</code></a> belong to the corresponding set. That is, if the functions in entries <span>$(i, j)$</span> and <span>$(j, i)$</span> are different, then a constraint will be added to make sure that the entries are equal.</p><p><strong>Example</strong></p><p><a href="#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a> is a subtype of <a href="#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>AbstractSymmetricMatrixSetSquare</code></a> and constraining the matrix</p><p class="math-container">\[\begin{bmatrix}
+  1 &amp; -y\\
+  -z &amp; 0\\
+\end{bmatrix}\]</p><p>to be symmetric positive semidefinite can be achieved by constraining the vector <span>$(1, -z, -y, 0)$</span> (or <span>$(1, -y, -z, 0)$</span>) to belong to the <code>PositiveSemidefiniteConeSquare(2)</code>. It both constrains <span>$y = z$</span> and <span>$(1, -y, 0)$</span> (or <span>$(1, -z, 0)$</span>) to be in <code>PositiveSemidefiniteConeTriangle(2)</code>, since <code>triangular_form(PositiveSemidefiniteConeSquare)</code> is <code>PositiveSemidefiniteConeTriangle</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1246-L1274">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.side_dimension" href="#MathOptInterface.side_dimension"><code>MathOptInterface.side_dimension</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">side_dimension(
+    set::Union{
+        AbstractSymmetricMatrixSetTriangle,
+        AbstractSymmetricMatrixSetSquare,
+        HermitianPositiveSemidefiniteConeTriangle,
+    },
+)</code></pre><p>Side dimension of the matrices in <code>set</code>.</p><p><strong>Convention</strong></p><p>By convention, the side dimension should be stored in the <code>side_dimension</code> field. If this is not the case for a subtype of <a href="#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>AbstractSymmetricMatrixSetTriangle</code></a>, or <a href="#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>AbstractSymmetricMatrixSetSquare</code></a> you must implement this method.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1429-L1445">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.triangular_form" href="#MathOptInterface.triangular_form"><code>MathOptInterface.triangular_form</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">triangular_form(S::Type{&lt;:AbstractSymmetricMatrixSetSquare})
+triangular_form(set::AbstractSymmetricMatrixSetSquare)</code></pre><p>Return the <a href="#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>AbstractSymmetricMatrixSetTriangle</code></a> corresponding to the vectorization of the upper triangular part of matrices in the <a href="#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>AbstractSymmetricMatrixSetSquare</code></a> set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1279-L1286">source</a></section></article><p>List of recognized matrix sets.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.PositiveSemidefiniteConeTriangle" href="#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MathOptInterface.PositiveSemidefiniteConeTriangle</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">PositiveSemidefiniteConeTriangle(side_dimension::Int) &lt;: AbstractSymmetricMatrixSetTriangle</code></pre><p>The (vectorized) cone of symmetric positive semidefinite matrices, with non-negative <code>side_dimension</code> rows and columns.</p><p>See <a href="#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>AbstractSymmetricMatrixSetTriangle</code></a> for more details on the vectorized form.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1293-L1301">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.PositiveSemidefiniteConeSquare" href="#MathOptInterface.PositiveSemidefiniteConeSquare"><code>MathOptInterface.PositiveSemidefiniteConeSquare</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">PositiveSemidefiniteConeSquare(side_dimension::Int) &lt;: AbstractSymmetricMatrixSetSquare</code></pre><p>The cone of symmetric positive semidefinite matrices, with non-negative side length <code>side_dimension</code>.</p><p>See <a href="#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>AbstractSymmetricMatrixSetSquare</code></a> for more details on the vectorized form.</p><p>The entries of the matrix are given column by column (or equivalently, row by row).</p><p>The matrix is both constrained to be symmetric and to be positive semidefinite. That is, if the functions in entries <span>$(i, j)$</span> and <span>$(j, i)$</span> are different, then a constraint will be added to make sure that the entries are equal.</p><p><strong>Example</strong></p><p>Constraining the matrix</p><p class="math-container">\[\begin{bmatrix}
+  1 &amp; -y\\
+  -z &amp; 0\\
+\end{bmatrix}\]</p><p>to be symmetric positive semidefinite can be achieved by constraining the vector <span>$(1, -z, -y, 0)$</span> (or <span>$(1, -y, -z, 0)$</span>) to belong to the <code>PositiveSemidefiniteConeSquare(2)</code>.</p><p>It both constrains <span>$y = z$</span> and <span>$(1, -y, 0)$</span> (or <span>$(1, -z, 0)$</span>) to be in <code>PositiveSemidefiniteConeTriangle(2)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1322-L1353">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.HermitianPositiveSemidefiniteConeTriangle" href="#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>MathOptInterface.HermitianPositiveSemidefiniteConeTriangle</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HermitianPositiveSemidefiniteConeTriangle(side_dimension::Int) &lt;: AbstractVectorSet</code></pre><p>The (vectorized) cone of Hermitian positive semidefinite matrices, with non-negative <code>side_dimension</code> rows and columns.</p><p>Becaue the matrix is Hermitian, the diagonal elements are real, and the complex-valued lower triangular entries are obtained as the conjugate of corresponding upper triangular entries.</p><p><strong>Vectorization format</strong></p><p>The vectorized form starts with real part of the entries of the upper triangular part of the matrix, given column by column as explained in <a href="#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>AbstractSymmetricMatrixSetSquare</code></a>.</p><p>It is then followed by the imaginary part of the off-diagonal entries of the upper triangular part, also given column by column.</p><p>For example, the matrix</p><p class="math-container">\[\begin{bmatrix}
+  1 &amp; 2 + 7im &amp; 4 + 8im\\
+  2 - 7im &amp; 3 &amp; 5 + 9im\\
+  4 - 8im &amp; 5 - 9im &amp; 6
+\end{bmatrix}\]</p><p>has <a href="#MathOptInterface.side_dimension"><code>side_dimension</code></a> 3 and is represented as the vector <span>$[1, 2, 3, 4, 5, 6, 7, 8, 9]$</span>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1389-L1418">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LogDetConeTriangle" href="#MathOptInterface.LogDetConeTriangle"><code>MathOptInterface.LogDetConeTriangle</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LogDetConeTriangle(side_dimension::Int)</code></pre><p>The log-determinant cone <span>$\{ (t, u, X) \in \mathbb{R}^{2 + d(d+1)/2} : t \le u \log(\det(X/u)), u &gt; 0 \}$</span>, where the matrix <code>X</code> is represented in the same symmetric packed format as in the <code>PositiveSemidefiniteConeTriangle</code>.</p><p>The non-negative argument <code>side_dimension</code> is the side dimension of the matrix <code>X</code>, i.e., its number of rows or columns.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; X = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; X]),
+           MOI.LogDetConeTriangle(2),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.LogDetConeTriangle}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1456-L1487">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.LogDetConeSquare" href="#MathOptInterface.LogDetConeSquare"><code>MathOptInterface.LogDetConeSquare</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LogDetConeSquare(side_dimension::Int)</code></pre><p>The log-determinant cone <span>$\{ (t, u, X) \in \mathbb{R}^{2 + d^2} : t \le u \log(\det(X/u)), X \text{ symmetric}, u &gt; 0 \}$</span>, where the matrix <code>X</code> is represented in the same format as in the <a href="#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a>.</p><p>Similarly to <a href="#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a>, constraints are added to ensure that <code>X</code> is symmetric.</p><p>The non-negative argument <code>side_dimension</code> is the side dimension of the matrix <code>X</code>, i.e., its number of rows or columns.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; X = reshape(MOI.add_variables(model, 4), 2, 2)
+2×2 Matrix{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(2)  MOI.VariableIndex(4)
+ MOI.VariableIndex(3)  MOI.VariableIndex(5)
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; vec(X)]),
+           MOI.LogDetConeSquare(2),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.LogDetConeSquare}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1507-L1544">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RootDetConeTriangle" href="#MathOptInterface.RootDetConeTriangle"><code>MathOptInterface.RootDetConeTriangle</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RootDetConeTriangle(side_dimension::Int)</code></pre><p>The root-determinant cone <span>$\{ (t, X) \in \mathbb{R}^{1 + d(d+1)/2} : t \le \det(X)^{1/d} \}$</span>, where the matrix <code>X</code> is represented in the same symmetric packed format as in the <a href="#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>PositiveSemidefiniteConeTriangle</code></a>.</p><p>The non-negative argument <code>side_dimension</code> is the side dimension of the matrix <code>X</code>, i.e., its number of rows or columns.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; X = MOI.add_variables(model, 3);
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; X]),
+           MOI.RootDetConeTriangle(2),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RootDetConeTriangle}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1568-L1599">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.RootDetConeSquare" href="#MathOptInterface.RootDetConeSquare"><code>MathOptInterface.RootDetConeSquare</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RootDetConeSquare(side_dimension::Int)</code></pre><p>The root-determinant cone <span>$\{ (t, X) \in \mathbb{R}^{1 + d^2} : t \le \det(X)^{1/d}, X \text{ symmetric} \}$</span>, where the matrix <code>X</code> is represented in the same format as <a href="#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a>.</p><p>Similarly to <a href="#MathOptInterface.PositiveSemidefiniteConeSquare"><code>PositiveSemidefiniteConeSquare</code></a>, constraints are added to ensure that <code>X</code> is symmetric.</p><p>The non-negative argument <code>side_dimension</code> is the side dimension of the matrix <code>X</code>, i.e., its number of rows or columns.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; t = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; X = reshape(MOI.add_variables(model, 4), 2, 2)
+2×2 Matrix{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(2)  MOI.VariableIndex(4)
+ MOI.VariableIndex(3)  MOI.VariableIndex(5)
+
+julia&gt; MOI.add_constraint(
+           model,
+           MOI.VectorOfVariables([t; vec(X)]),
+           MOI.RootDetConeSquare(2),
+       )
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RootDetConeSquare}(1)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/sets.jl#L1619-L1656">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../background/naming_conventions/">« Naming conventions</a><a class="docs-footer-nextpage" href="../models/">Models »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/reference/variables/index.html b/previews/PR3525/moi/reference/variables/index.html
new file mode 100644
index 00000000000..4cd20c22fd7
--- /dev/null
+++ b/previews/PR3525/moi/reference/variables/index.html
@@ -0,0 +1,28 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Variables · JuMP</title><meta name="title" content="Variables · JuMP"/><meta property="og:title" content="Variables · JuMP"/><meta property="twitter:title" content="Variables · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../standard_form/">Standard form</a></li><li><a class="tocitem" href="../models/">Models</a></li><li class="is-active"><a class="tocitem" href>Variables</a><ul class="internal"><li><a class="tocitem" href="#Functions"><span>Functions</span></a></li><li><a class="tocitem" href="#Attributes"><span>Attributes</span></a></li></ul></li><li><a class="tocitem" href="../constraints/">Constraints</a></li><li><a class="tocitem" href="../modification/">Modifications</a></li><li><a class="tocitem" href="../nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">API Reference</a></li><li class="is-active"><a href>Variables</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Variables</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/variables.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Variables"><a class="docs-heading-anchor" href="#Variables">Variables</a><a id="Variables-1"></a><a class="docs-heading-anchor-permalink" href="#Variables" title="Permalink"></a></h1><h2 id="Functions"><a class="docs-heading-anchor" href="#Functions">Functions</a><a id="Functions-1"></a><a class="docs-heading-anchor-permalink" href="#Functions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.add_variable" href="#MathOptInterface.add_variable"><code>MathOptInterface.add_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_variable(model::ModelLike)::VariableIndex</code></pre><p>Add a scalar variable to the model, returning a variable index.</p><p>A <a href="../errors/#MathOptInterface.AddVariableNotAllowed"><code>AddVariableNotAllowed</code></a> error is thrown if adding variables cannot be done in the current state of the model <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L35-L42">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.add_variables" href="#MathOptInterface.add_variables"><code>MathOptInterface.add_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_variables(model::ModelLike, n::Int)::Vector{VariableIndex}</code></pre><p>Add <code>n</code> scalar variables to the model, returning a vector of variable indices.</p><p>A <a href="../errors/#MathOptInterface.AddVariableNotAllowed"><code>AddVariableNotAllowed</code></a> error is thrown if adding variables cannot be done in the current state of the model <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L23-L30">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.add_constrained_variable" href="#MathOptInterface.add_constrained_variable"><code>MathOptInterface.add_constrained_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_constrained_variable(
+    model::ModelLike,
+    set::AbstractScalarSet
+)::Tuple{MOI.VariableIndex,
+         MOI.ConstraintIndex{MOI.VariableIndex, typeof(set)}}</code></pre><p>Add to <code>model</code> a scalar variable constrained to belong to <code>set</code>, returning the index of the variable created and the index of the constraint constraining the variable to belong to <code>set</code>.</p><p>By default, this function falls back to creating a free variable with <a href="#MathOptInterface.add_variable"><code>add_variable</code></a> and then constraining it to belong to <code>set</code> with <a href="../constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L81-L95">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.add_constrained_variables" href="#MathOptInterface.add_constrained_variables"><code>MathOptInterface.add_constrained_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_constrained_variables(
+    model::ModelLike,
+    sets::AbstractVector{&lt;:AbstractScalarSet}
+)::Tuple{
+    Vector{MOI.VariableIndex},
+    Vector{MOI.ConstraintIndex{MOI.VariableIndex,eltype(sets)}},
+}</code></pre><p>Add to <code>model</code> scalar variables constrained to belong to <code>sets</code>, returning the indices of the variables created and the indices of the constraints constraining the variables to belong to each set in <code>sets</code>. That is, if it returns <code>variables</code> and <code>constraints</code>, <code>constraints[i]</code> is the index of the constraint constraining <code>variable[i]</code> to belong to <code>sets[i]</code>.</p><p>By default, this function falls back to calling <a href="#MathOptInterface.add_constrained_variable"><code>add_constrained_variable</code></a> on each set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L159-L176">source</a></section><section><div><pre><code class="language-julia hljs">add_constrained_variables(
+    model::ModelLike,
+    set::AbstractVectorSet,
+)::Tuple{
+    Vector{MOI.VariableIndex},
+    MOI.ConstraintIndex{MOI.VectorOfVariables,typeof(set)},
+}</code></pre><p>Add to <code>model</code> a vector of variables constrained to belong to <code>set</code>, returning the indices of the variables created and the index of the constraint constraining the vector of variables to belong to <code>set</code>.</p><p>By default, this function falls back to creating free variables with <a href="#MathOptInterface.add_variables"><code>add_variables</code></a> and then constraining it to belong to <code>set</code> with <a href="../constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L190-L206">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.supports_add_constrained_variable" href="#MathOptInterface.supports_add_constrained_variable"><code>MathOptInterface.supports_add_constrained_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_add_constrained_variable(
+    model::ModelLike,
+    S::Type{&lt;:AbstractScalarSet}
+)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports constraining a variable to belong to a set of type <code>S</code> either on creation of the variable with <a href="#MathOptInterface.add_constrained_variable"><code>add_constrained_variable</code></a> or after the variable is created with <a href="../constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a>.</p><p>By default, this function falls back to <code>supports_add_constrained_variables(model, Reals) &amp;&amp; supports_constraint(model, MOI.VariableIndex, S)</code> which is the correct definition for most models.</p><p><strong>Example</strong></p><p>Suppose that a solver supports only two kind of variables: binary variables and continuous variables with a lower bound. If the solver decides not to support <code>VariableIndex</code>-in-<code>Binary</code> and <code>VariableIndex</code>-in-<code>GreaterThan</code> constraints, it only has to implement <code>add_constrained_variable</code> for these two sets which prevents the user to add both a binary constraint and a lower bound on the same variable. Moreover, if the user adds a <code>VariableIndex</code>-in-<code>GreaterThan</code> constraint, implementing this interface (i.e., <code>supports_add_constrained_variables</code>) enables the constraint to be transparently bridged into a supported constraint.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L45-L72">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.supports_add_constrained_variables" href="#MathOptInterface.supports_add_constrained_variables"><code>MathOptInterface.supports_add_constrained_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_add_constrained_variables(
+    model::ModelLike,
+    S::Type{&lt;:AbstractVectorSet}
+)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether <code>model</code> supports constraining a vector of variables to belong to a set of type <code>S</code> either on creation of the vector of variables with <a href="#MathOptInterface.add_constrained_variables"><code>add_constrained_variables</code></a> or after the variable is created with <a href="../constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a>.</p><p>By default, if <code>S</code> is <code>Reals</code> then this function returns <code>true</code> and otherwise, it falls back to <code>supports_add_constrained_variables(model, Reals) &amp;&amp; supports_constraint(model, MOI.VectorOfVariables, S)</code> which is the correct definition for most models.</p><p><strong>Example</strong></p><p>In the standard conic form (see <a href="../../background/duality/#Duality">Duality</a>), the variables are grouped into several cones and the constraints are affine equality constraints. If <code>Reals</code> is not one of the cones supported by the solvers then it needs to implement <code>supports_add_constrained_variables(::Optimizer, ::Type{Reals}) = false</code> as free variables are not supported. The solvers should then implement <code>supports_add_constrained_variables(::Optimizer, ::Type{&lt;:SupportedCones}) = true</code> where <code>SupportedCones</code> is the union of all cone types that are supported; it does not have to implement the method <code>supports_constraint(::Type{VectorOfVariables}, Type{&lt;:SupportedCones})</code> as it should return <code>false</code> and it&#39;s the default. This prevents the user to constrain the same variable in two different cones. When a <code>VectorOfVariables</code>-in-<code>S</code> is added, the variables of the vector have already been created so they already belong to given cones. If bridges are enabled, the constraint will therefore be bridged by adding slack variables in <code>S</code> and equality constraints ensuring that the slack variables are equal to the corresponding variables of the given constraint function.</p><p>Note that there may also be sets for which <code>!supports_add_constrained_variables(model, S)</code> and <code>supports_constraint(model, MOI.VectorOfVariables, S)</code>. For instance, suppose a solver supports positive semidefinite variable constraints and two types of variables: binary variables and nonnegative variables. Then the solver should support adding <code>VectorOfVariables</code>-in-<code>PositiveSemidefiniteConeTriangle</code> constraints, but it should not support creating variables constrained to belong to the <code>PositiveSemidefiniteConeTriangle</code> because the variables in <code>PositiveSemidefiniteConeTriangle</code> should first be created as either binary or non-negative.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/variables.jl#L102-L149">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}" href="#MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}"><code>MathOptInterface.is_valid</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">is_valid(model::ModelLike, index::Index)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether this index refers to a valid object in the model <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L95-L99">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}" href="#MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}"><code>MathOptInterface.delete</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">delete(model::ModelLike, index::Index)</code></pre><p>Delete the referenced object from the model. Throw <a href="../errors/#MathOptInterface.DeleteNotAllowed"><code>DeleteNotAllowed</code></a> if if <code>index</code> cannot be deleted.</p><p>The following modifications also take effect if <code>Index</code> is <a href="../standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>:</p><ul><li>If <code>index</code> used in the objective function, it is removed from the function, i.e., it is substituted for zero.</li><li>For each <code>func</code>-in-<code>set</code> constraint of the model:<ul><li>If <code>func isa VariableIndex</code> and <code>func == index</code> then the constraint is deleted.</li><li>If <code>func isa VectorOfVariables</code> and <code>index in func.variables</code> then<ul><li>if <code>length(func.variables) == 1</code> is one, the constraint is deleted;</li><li>if <code>length(func.variables) &gt; 1</code> and <code>supports_dimension_update(set)</code> then then the variable is removed from <code>func</code> and <code>set</code> is replaced by <code>update_dimension(set, MOI.dimension(set) - 1)</code>.</li><li>Otherwise, a <a href="../errors/#MathOptInterface.DeleteNotAllowed"><code>DeleteNotAllowed</code></a> error is thrown.</li></ul></li><li>Otherwise, the variable is removed from <code>func</code>, i.e., it is substituted for zero.</li></ul></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L132-L152">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, Vector{MathOptInterface.VariableIndex}}" href="#MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, Vector{MathOptInterface.VariableIndex}}"><code>MathOptInterface.delete</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">delete(model::ModelLike, indices::Vector{R&lt;:Index}) where {R}</code></pre><p>Delete the referenced objects in the vector <code>indices</code> from the model. It may be assumed that <code>R</code> is a concrete type. The default fallback sequentially deletes the individual items in <code>indices</code>, although specialized implementations may be more efficient.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/indextypes.jl#L155-L162">source</a></section></article><h2 id="Attributes"><a class="docs-heading-anchor" href="#Attributes">Attributes</a><a id="Attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Attributes" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.AbstractVariableAttribute" href="#MathOptInterface.AbstractVariableAttribute"><code>MathOptInterface.AbstractVariableAttribute</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractVariableAttribute</code></pre><p>Abstract supertype for attribute objects that can be used to set or get attributes (properties) of variables in the model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L30-L35">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VariableName" href="#MathOptInterface.VariableName"><code>MathOptInterface.VariableName</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariableName()</code></pre><p>A variable attribute for a string identifying the variable. It is <em>valid</em> for two variables to have the same name; however, variables with duplicate names cannot be looked up using <a href="../models/#MathOptInterface.get"><code>get</code></a>. It has a default value of <code>&quot;&quot;</code> if not set`.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1429-L1436">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VariablePrimalStart" href="#MathOptInterface.VariablePrimalStart"><code>MathOptInterface.VariablePrimalStart</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariablePrimalStart()</code></pre><p>A variable attribute for the initial assignment to some primal variable&#39;s value that the optimizer may use to warm-start the solve. May be a number or <code>nothing</code> (unset).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1441-L1447">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VariablePrimal" href="#MathOptInterface.VariablePrimal"><code>MathOptInterface.VariablePrimal</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariablePrimal(result_index::Int = 1)</code></pre><p>A variable attribute for the assignment to some primal variable&#39;s value in result <code>result_index</code>. If <code>result_index</code> is omitted, it is 1 by default.</p><p>If the solver does not have a primal value for the variable because the <code>result_index</code> is beyond the available solutions (whose number is indicated by the <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute), getting this attribute must throw a <a href="../errors/#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check <a href="../models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> before accessing the <code>VariablePrimal</code> attribute.</p><p>See <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1450-L1465">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.VariableBasisStatus" href="#MathOptInterface.VariableBasisStatus"><code>MathOptInterface.VariableBasisStatus</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariableBasisStatus(result_index::Int = 1)</code></pre><p>A variable attribute for the <code>BasisStatusCode</code> of a variable in result <code>result_index</code>, with respect to an available optimal solution basis.</p><p>If the solver does not have a basis statue for the variable because the <code>result_index</code> is beyond the available solutions (whose number is indicated by the <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> attribute), getting this attribute must throw a <a href="../errors/#MathOptInterface.ResultIndexBoundsError"><code>ResultIndexBoundsError</code></a>. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check <a href="../models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a> before accessing the <code>VariableBasisStatus</code> attribute.</p><p>See <a href="../models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> for information on how the results are ordered.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/attributes.jl#L1516-L1531">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../models/">« Models</a><a class="docs-footer-nextpage" href="../constraints/">Constraints »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/release_notes/index.html b/previews/PR3525/moi/release_notes/index.html
new file mode 100644
index 00000000000..84a3dfe3e8d
--- /dev/null
+++ b/previews/PR3525/moi/release_notes/index.html
@@ -0,0 +1,34 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Release notes · JuMP</title><meta name="title" content="Release notes · JuMP"/><meta property="og:title" content="Release notes · JuMP"/><meta property="twitter:title" content="Release notes · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../">Introduction</a></li><li><a class="tocitem" href="../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../manual/models/">Models</a></li><li><a class="tocitem" href="../manual/variables/">Variables</a></li><li><a class="tocitem" href="../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../background/duality/">Duality</a></li><li><a class="tocitem" href="../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../reference/models/">Models</a></li><li><a class="tocitem" href="../reference/variables/">Variables</a></li><li><a class="tocitem" href="../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../developer/checklists/">Checklists</a></li></ul></li><li class="is-active"><a class="tocitem" href>Release notes</a><ul class="internal"><li><a class="tocitem" href="#[v1.20.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.20.0)-(September-7,-2023)"><span>v1.20.0 (September 7, 2023)</span></a></li><li><a class="tocitem" href="#[v1.19.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.19.0)-(August-15,-2023)"><span>v1.19.0 (August 15, 2023)</span></a></li><li><a class="tocitem" href="#[v1.18.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.18.0)-(June-23,-2023)"><span>v1.18.0 (June 23, 2023)</span></a></li><li><a class="tocitem" href="#[v1.17.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.1)-(June-6,-2023)"><span>v1.17.1 (June 6, 2023)</span></a></li><li><a class="tocitem" href="#[v1.17.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.0)-(June-1,-2023)"><span>v1.17.0 (June 1, 2023)</span></a></li><li><a class="tocitem" href="#[v1.16.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.16.0)-(May-16,-2023)"><span>v1.16.0 (May 16, 2023)</span></a></li><li><a class="tocitem" href="#[v1.15.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.1)-(April-25,-2023)"><span>v1.15.1 (April 25, 2023)</span></a></li><li><a class="tocitem" href="#[v1.15.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.0)-(April-19,-2023)"><span>v1.15.0 (April 19, 2023)</span></a></li><li><a class="tocitem" href="#[v1.14.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.1)-(April-6,-2023)"><span>v1.14.1 (April 6, 2023)</span></a></li><li><a class="tocitem" href="#[v1.14.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.0)-(April-4,-2023)"><span>v1.14.0 (April 4, 2023)</span></a></li><li><a class="tocitem" href="#[v1.13.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.2)-(March-21,-2023)"><span>v1.13.2 (March 21, 2023)</span></a></li><li><a class="tocitem" href="#[v1.13.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.1)-(March-3,-2023)"><span>v1.13.1 (March 3, 2023)</span></a></li><li><a class="tocitem" href="#[v1.13.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.0)-(February-28,-2023)"><span>v1.13.0 (February 28, 2023)</span></a></li><li><a class="tocitem" href="#[v1.12.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.12.0)-(February-10,-2023)"><span>v1.12.0 (February 10, 2023)</span></a></li><li><a class="tocitem" href="#[v1.11.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.5)-(January-24,-2023)"><span>v1.11.5 (January 24, 2023)</span></a></li><li><a class="tocitem" href="#[v1.11.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.4)-(January-12,-2023)"><span>v1.11.4 (January 12, 2023)</span></a></li><li><a class="tocitem" href="#[v1.11.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.3)-(January-12,-2023)"><span>v1.11.3 (January 12, 2023)</span></a></li><li><a class="tocitem" href="#[v1.11.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.2)-(January-2,-2023)"><span>v1.11.2 (January 2, 2023)</span></a></li><li><a class="tocitem" href="#[v1.11.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.1)-(December-22,-2022)"><span>v1.11.1 (December 22, 2022)</span></a></li><li><a class="tocitem" href="#[v1.11.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.0)-(December-2,-2022)"><span>v1.11.0 (December 2, 2022)</span></a></li><li><a class="tocitem" href="#[v1.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.10.0)-(November-22,-2022)"><span>v1.10.0 (November 22, 2022)</span></a></li><li><a class="tocitem" href="#[v1.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.9.0)-(October-29,-2022)"><span>v1.9.0 (October 29, 2022)</span></a></li><li><a class="tocitem" href="#[v1.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.2)-(September-20,-2022)"><span>v1.8.2 (September 20, 2022)</span></a></li><li><a class="tocitem" href="#[v1.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.1)-(September-12,-2022)"><span>v1.8.1 (September 12, 2022)</span></a></li><li><a class="tocitem" href="#[v1.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.0)-(September-1,-2022)"><span>v1.8.0 (September 1, 2022)</span></a></li><li><a class="tocitem" href="#[v1.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.7.0)-(August-16,-2022)"><span>v1.7.0 (August 16, 2022)</span></a></li><li><a class="tocitem" href="#[v1.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.1)-(July-23,-2022)"><span>v1.6.1 (July 23, 2022)</span></a></li><li><a class="tocitem" href="#[v1.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.0)-(July-2,-2022)"><span>v1.6.0 (July 2, 2022)</span></a></li><li><a class="tocitem" href="#[v1.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.5.0)-(June-27,-2022)"><span>v1.5.0 (June 27, 2022)</span></a></li><li><a class="tocitem" href="#[v1.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.4.0)-(June-9,-2022)"><span>v1.4.0 (June 9, 2022)</span></a></li><li><a class="tocitem" href="#[v1.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.3.0)-(May-27,-2022)"><span>v1.3.0 (May 27, 2022)</span></a></li><li><a class="tocitem" href="#[v1.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.2.0)-(April-25,-2022)"><span>v1.2.0 (April 25, 2022)</span></a></li><li><a class="tocitem" href="#[v1.1.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.2)-(March-31,-2022)"><span>v1.1.2 (March 31, 2022)</span></a></li><li><a class="tocitem" href="#[v1.1.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.1)-(March-23,-2022)"><span>v1.1.1 (March 23, 2022)</span></a></li><li><a class="tocitem" href="#[v1.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.0)-(March-2,-2022)"><span>v1.1.0 (March 2, 2022)</span></a></li><li><a class="tocitem" href="#[v1.0.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.2)-(March-1,-2022)"><span>v1.0.2 (March 1, 2022)</span></a></li><li><a class="tocitem" href="#[v1.0.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.1)-(February-25,-2022)"><span>v1.0.1 (February 25, 2022)</span></a></li><li><a class="tocitem" href="#[v1.0.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.0)-(February-17,-2022)"><span>v1.0.0 (February 17, 2022)</span></a></li><li><a class="tocitem" href="#[v0.10.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.9)-(February-16,-2022)"><span>v0.10.9 (February 16, 2022)</span></a></li><li><a class="tocitem" href="#[v0.10.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.8)-(February-3,-2022)"><span>v0.10.8 (February 3, 2022)</span></a></li><li><a class="tocitem" href="#[v0.10.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.7)-(January-5,-2022)"><span>v0.10.7 (January 5, 2022)</span></a></li><li><a class="tocitem" href="#[v0.10.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.6)-(November-30,-2021)"><span>v0.10.6 (November 30, 2021)</span></a></li><li><a class="tocitem" href="#[v0.10.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.5)-(November-7,-2021)"><span>v0.10.5 (November 7, 2021)</span></a></li><li><a class="tocitem" href="#[v0.10.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.4)-(October-26,-2021)"><span>v0.10.4 (October 26, 2021)</span></a></li><li><a class="tocitem" href="#[v0.10.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.3)-(September-18,-2021)"><span>v0.10.3 (September 18, 2021)</span></a></li><li><a class="tocitem" href="#[v0.10.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.2)-(September-16,-2021)"><span>v0.10.2 (September 16, 2021)</span></a></li><li><a class="tocitem" href="#[v0.10.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.1)-(September-8,-2021)"><span>v0.10.1 (September 8, 2021)</span></a></li><li><a class="tocitem" href="#[v0.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.0)-(September-6,-2021)"><span>v0.10.0 (September 6, 2021)</span></a></li><li><a class="tocitem" href="#[v0.9.22](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.22)-(May-22,-2021)"><span>v0.9.22 (May 22, 2021)</span></a></li><li><a class="tocitem" href="#[v0.9.21](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.21)-(April-23,-2021)"><span>v0.9.21 (April 23, 2021)</span></a></li><li><a class="tocitem" href="#[v0.9.20](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.20)-(February-20,-2021)"><span>v0.9.20 (February 20, 2021)</span></a></li><li><a class="tocitem" href="#[v0.9.19](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.19)-(December-1,-2020)"><span>v0.9.19 (December 1, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.18](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.18)-(November-3,-2020)"><span>v0.9.18 (November 3, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.17](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.17)-(September-21,-2020)"><span>v0.9.17 (September 21, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.16](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.16)-(September-17,-2020)"><span>v0.9.16 (September 17, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.15](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.15)-(September-14,-2020)"><span>v0.9.15 (September 14, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.14](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.14)-(May-30,-2020)"><span>v0.9.14 (May 30, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.13](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.13)-(March-24,-2020)"><span>v0.9.13 (March 24, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.12](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.12)-(February-28,-2020)"><span>v0.9.12 (February 28, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.11](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.11)-(February-21,-2020)"><span>v0.9.11 (February 21, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.10](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.10)-(January-31,-2020)"><span>v0.9.10 (January 31, 2020)</span></a></li><li><a class="tocitem" href="#[v0.9.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.9)-(December-29,-2019)"><span>v0.9.9 (December 29, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.8)-(December-19,-2019)"><span>v0.9.8 (December 19, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.7)-(October-30,-2019)"><span>v0.9.7 (October 30, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.6)-(October-25,-2019)"><span>v0.9.6 (October 25, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.5)-(October-9,-2019)"><span>v0.9.5 (October 9, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.4)-(October-2,-2019)"><span>v0.9.4 (October 2, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.3)-(September-20,-2019)"><span>v0.9.3 (September 20, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.2)-(September-5,-2019)"><span>v0.9.2 (September 5, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.1)-(August-22,-2019)"><span>v0.9.1 (August 22, 2019)</span></a></li><li><a class="tocitem" href="#[v0.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.0)-(August-13,-2019)"><span>v0.9.0 (August 13, 2019)</span></a></li><li><a class="tocitem" href="#[v0.8.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.4)-(March-13,-2019)"><span>v0.8.4 (March 13, 2019)</span></a></li><li><a class="tocitem" href="#[v0.8.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.3)-(March-6,-2019)"><span>v0.8.3 (March 6, 2019)</span></a></li><li><a class="tocitem" href="#[v0.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.2)-(February-7,-2019)"><span>v0.8.2 (February 7, 2019)</span></a></li><li><a class="tocitem" href="#[v0.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.1)-(January-7,-2019)"><span>v0.8.1 (January 7, 2019)</span></a></li><li><a class="tocitem" href="#[v0.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.0)-(December-18,-2018)"><span>v0.8.0 (December 18, 2018)</span></a></li><li><a class="tocitem" href="#[v0.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.7.0)-(December-13,-2018)"><span>v0.7.0 (December 13, 2018)</span></a></li><li><a class="tocitem" href="#[v0.6.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.4)-(November-27,-2018)"><span>v0.6.4 (November 27, 2018)</span></a></li><li><a class="tocitem" href="#[v0.6.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.3)-(November-16,-2018)"><span>v0.6.3 (November 16, 2018)</span></a></li><li><a class="tocitem" href="#[v0.6.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.2)-(October-26,-2018)"><span>v0.6.2 (October 26, 2018)</span></a></li><li><a class="tocitem" href="#[v0.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.1)-(September-22,-2018)"><span>v0.6.1 (September 22, 2018)</span></a></li><li><a class="tocitem" href="#[v0.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.0)-(August-30,-2018)"><span>v0.6.0 (August 30, 2018)</span></a></li><li><a class="tocitem" href="#[v0.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.5.0)-(August-5,-2018)"><span>v0.5.0 (August 5, 2018)</span></a></li><li><a class="tocitem" href="#[v0.4.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.1)-(June-28,-2018)"><span>v0.4.1 (June 28, 2018)</span></a></li><li><a class="tocitem" href="#[v0.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.0)-(June-23,-2018)"><span>v0.4.0 (June 23, 2018)</span></a></li><li><a class="tocitem" href="#[v0.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.3.0)-(May-25,-2018)"><span>v0.3.0 (May 25, 2018)</span></a></li><li><a class="tocitem" href="#[v0.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.2.0)-(April-24,-2018)"><span>v0.2.0 (April 24, 2018)</span></a></li><li><a class="tocitem" href="#[v0.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.1.0)-(February-28,-2018)"><span>v0.1.0 (February 28, 2018)</span></a></li></ul></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li class="is-active"><a href>Release notes</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Release notes</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/release_notes.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Release-notes"><a class="docs-heading-anchor" href="#Release-notes">Release notes</a><a id="Release-notes-1"></a><a class="docs-heading-anchor-permalink" href="#Release-notes" title="Permalink"></a></h1><p>The format is based on <a href="https://keepachangelog.com/en/1.0.0/">Keep a Changelog</a>, and this project adheres to <a href="https://semver.org/spec/v2.0.0.html">Semantic Versioning</a>.</p><h2 id="[v1.20.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.20.0)-(September-7,-2023)"><a class="docs-heading-anchor" href="#[v1.20.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.20.0)-(September-7,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.20.0">v1.20.0</a> (September 7, 2023)</a><a id="[v1.20.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.20.0)-(September-7,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.20.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.20.0)-(September-7,-2023)" title="Permalink"></a></h2><h3 id="Added"><a class="docs-heading-anchor" href="#Added">Added</a><a id="Added-1"></a><a class="docs-heading-anchor-permalink" href="#Added" title="Permalink"></a></h3><ul><li>Added <a href="../reference/standard_form/#MathOptInterface.Scaled"><code>Scaled</code></a> set (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2237">#2237</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2263">#2263</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2264">#2264</a>)</li><li>Added <a href="../reference/models/#MathOptInterface.ObjectiveLimit"><code>ObjectiveLimit</code></a> attribute (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2257">#2257</a>)</li></ul><h3 id="Other"><a class="docs-heading-anchor" href="#Other">Other</a><a id="Other-1"></a><a class="docs-heading-anchor-permalink" href="#Other" title="Permalink"></a></h3><ul><li>Updated dependencies (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2258">#2258</a>)</li><li>Improved performance of <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> utilities (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2259">#2259</a>)</li><li>Fixed docstrings (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2261">#2261</a>)</li></ul><h2 id="[v1.19.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.19.0)-(August-15,-2023)"><a class="docs-heading-anchor" href="#[v1.19.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.19.0)-(August-15,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.19.0">v1.19.0</a> (August 15, 2023)</a><a id="[v1.19.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.19.0)-(August-15,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.19.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.19.0)-(August-15,-2023)" title="Permalink"></a></h2><h3 id="Added-2"><a class="docs-heading-anchor" href="#Added-2">Added</a><a class="docs-heading-anchor-permalink" href="#Added-2" title="Permalink"></a></h3><ul><li>Added <a href="../reference/standard_form/#MathOptInterface.VectorNonlinearFunction"><code>VectorNonlinearFunction</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2201">#2201</a>)</li><li>Added new bridges<ul><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IntegerToZeroOneBridge"><code>Bridges.Constraint.IntegerToZeroOneBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2205">#2205</a>)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToScalarQuadraticBridge"><code>Bridges.Constraint.ToScalarQuadraticBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2235">#2235</a>)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToVectorQuadraticBridge"><code>Bridges.Constraint.ToVectorQuadraticBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2235">#2235</a>)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToScalarNonlinearBridge"><code>Bridges.Constraint.ToScalarNonlinearBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2233">#2233</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2235">#2235</a>)</li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.FunctionConversionBridge"><code>Bridges.Constraint.FunctionConversionBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2235">#2235</a>)</li></ul></li><li>Added <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.bridging_cost"><code>Bridges.bridging_cost</code></a> for setting a bridge-specific cost in the bridging graph (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2235">#2235</a>)</li><li>Added <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.eval_variables"><code>Utilities.eval_variables</code></a> support for <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2218">#2218</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2219">#2219</a>)</li><li>Added support for <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> in <code>FileFormats.NL</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2228">#2228</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2231">#2231</a>)</li><li>Added support for writing non-<code>Float64</code> functions in <code>FileFormats.MOF</code></li><li>Added <code>Utilities.lazy_map</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2254">#2254</a>)</li></ul><h3 id="Fixed"><a class="docs-heading-anchor" href="#Fixed">Fixed</a><a id="Fixed-1"></a><a class="docs-heading-anchor-permalink" href="#Fixed" title="Permalink"></a></h3><ul><li>Fixed method ambiguities in <code>operate</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2224">#2224</a>)</li><li>Fixed reading LP file with a <code>-infinity &lt;= x &lt;= +infinity</code> variable (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2225">#2225</a>)</li><li>Fixed missing <code>require</code> in <code>Test.test_nonlinear_duals</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2230">#2230</a>)</li><li>Fixed bug in <a href="../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>ConstraintPrimal</code></a> of <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.QuadtoSOCBridge"><code>Bridges.Constraint.QuadtoSOCBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2240">#2240</a>)</li></ul><h3 id="Other-2"><a class="docs-heading-anchor" href="#Other-2">Other</a><a class="docs-heading-anchor-permalink" href="#Other-2" title="Permalink"></a></h3><ul><li>Added extensions to <code>solver-tests.yml</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2229">#2229</a>)</li><li>Refactored <code>test/Benchmarks</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2234">#2234</a>)</li><li>Fixed warnings in tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2241">#2241</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2243">#2243</a>)</li><li>Small refactoring of bridges for upcoming <code>VectorNonlinearFunction</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2244">#2244</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2245">#2245</a>)</li><li>Fixed various typos (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2251">#2251</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2255">#2255</a>)</li><li>Partitioned how we run the tests on GitHub actions (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2252">#2252</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2253">#2253</a>)</li></ul><h2 id="[v1.18.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.18.0)-(June-23,-2023)"><a class="docs-heading-anchor" href="#[v1.18.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.18.0)-(June-23,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.18.0">v1.18.0</a> (June 23, 2023)</a><a id="[v1.18.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.18.0)-(June-23,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.18.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.18.0)-(June-23,-2023)" title="Permalink"></a></h2><h3 id="Added-3"><a class="docs-heading-anchor" href="#Added-3">Added</a><a class="docs-heading-anchor-permalink" href="#Added-3" title="Permalink"></a></h3><ul><li>Added <code>Bridges.Objective.SlackBridgePrimalDualStart</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2194">#2194</a>)</li><li>Added <a href="../reference/nonlinear/#MathOptInterface.constraint_gradient_structure"><code>constraint_gradient_structure</code></a> and <a href="../reference/nonlinear/#MathOptInterface.eval_constraint_gradient"><code>eval_constraint_gradient</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2200">#2200</a>)</li></ul><h3 id="Fixed-2"><a class="docs-heading-anchor" href="#Fixed-2">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-2" title="Permalink"></a></h3><ul><li>Fixed a missing <code>@require</code> in <code>MOI.Test</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2195">#2195</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2196">#2196</a>)</li><li>Fixed incorrect usage of <code>Utilities.operate!</code> in bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2207">#2207</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2216">#2216</a>)</li><li>Fixed splatting nonlinear expression with univariate operator (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2221">#2221</a>)</li></ul><h3 id="Other-3"><a class="docs-heading-anchor" href="#Other-3">Other</a><a class="docs-heading-anchor-permalink" href="#Other-3" title="Permalink"></a></h3><ul><li>Removed unused argument names (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2199">#2199</a>)</li><li>Reduced memory requirement for tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2204">#2204</a>)</li><li>Refactored <code>Utilities.promote_operation</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2206">#2206</a>)</li><li>Improved code style in <code>Utilities/mutable_arithmetics.jl</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2209">#2209</a>)</li><li>Refactored various methods in <code>Utilities/functions.jl</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2208">#2208</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2212">#2212</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2213">#2213</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2214">#2214</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2215">#2215</a>)</li></ul><h2 id="[v1.17.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.1)-(June-6,-2023)"><a class="docs-heading-anchor" href="#[v1.17.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.1)-(June-6,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.1">v1.17.1</a> (June 6, 2023)</a><a id="[v1.17.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.1)-(June-6,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.17.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.1)-(June-6,-2023)" title="Permalink"></a></h2><h3 id="Fixed-3"><a class="docs-heading-anchor" href="#Fixed-3">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-3" title="Permalink"></a></h3><ul><li>Fixed comparison of <a href="../reference/errors/#MathOptInterface.ScalarFunctionConstantNotZero"><code>ScalarFunctionConstantNotZero</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2190">#2190</a>)</li></ul><h3 id="Other-4"><a class="docs-heading-anchor" href="#Other-4">Other</a><a class="docs-heading-anchor-permalink" href="#Other-4" title="Permalink"></a></h3><ul><li>Added documentation for enum instances (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2186">#2186</a>)</li><li>Updated chatroom links in documentation (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2188">#2188</a>)</li><li>Changed the documentation to build on Julia v1.9 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2191">#2191</a>)</li></ul><h2 id="[v1.17.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.0)-(June-1,-2023)"><a class="docs-heading-anchor" href="#[v1.17.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.0)-(June-1,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.0">v1.17.0</a> (June 1, 2023)</a><a id="[v1.17.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.0)-(June-1,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.17.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.0)-(June-1,-2023)" title="Permalink"></a></h2><h3 id="Added-4"><a class="docs-heading-anchor" href="#Added-4">Added</a><a class="docs-heading-anchor-permalink" href="#Added-4" title="Permalink"></a></h3><ul><li>Added <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2059">#2059</a>)</li><li>Added a variety of tests for <a href="../reference/standard_form/#MathOptInterface.NormSpectralCone"><code>NormSpectralCone</code></a>, <a href="../reference/standard_form/#MathOptInterface.NormNuclearCone"><code>NormNuclearCone</code></a>, and <a href="../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>HermitianPositiveSemidefiniteConeTriangle</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2174">#2174</a>)</li><li>Added <code>Utilities.value_type</code> for computing the value type of a function (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2176">#2176</a>)</li></ul><h3 id="Fixed-4"><a class="docs-heading-anchor" href="#Fixed-4">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-4" title="Permalink"></a></h3><ul><li>Fixed support for external sets in <code>Utilities.loadfromstring!</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2177">#2177</a>)</li><li>Fixed <code>promote_operation</code> for <a href="../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>ScalarNonlinearFunction</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2179">#2179</a>)</li><li>Fixed two issues in <code>FileFormats.LP</code> when reading files with quadratic functions (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2182">#2182</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2184">#2184</a>)</li></ul><h2 id="[v1.16.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.16.0)-(May-16,-2023)"><a class="docs-heading-anchor" href="#[v1.16.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.16.0)-(May-16,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.16.0">v1.16.0</a> (May 16, 2023)</a><a id="[v1.16.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.16.0)-(May-16,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.16.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.16.0)-(May-16,-2023)" title="Permalink"></a></h2><h3 id="Added-5"><a class="docs-heading-anchor" href="#Added-5">Added</a><a class="docs-heading-anchor-permalink" href="#Added-5" title="Permalink"></a></h3><ul><li>Added support for MathOptFormat v1.3 and v1.4 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2158">#2158</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2169">#2169</a>)</li><li>Added new method to the nonlinear API (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2162">#2162</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2164">#2164</a>)<ul><li><a href="../reference/nonlinear/#MathOptInterface.eval_hessian_constraint"><code>eval_hessian_constraint</code></a></li><li><a href="../reference/nonlinear/#MathOptInterface.eval_hessian_objective"><code>eval_hessian_objective</code></a></li><li><a href="../reference/nonlinear/#MathOptInterface.hessian_constraint_structure"><code>hessian_constraint_structure</code></a></li><li><a href="../reference/nonlinear/#MathOptInterface.hessian_objective_structure"><code>hessian_objective_structure</code></a></li></ul></li><li>Added new sets<ul><li><a href="../reference/standard_form/#MathOptInterface.NormCone"><code>NormCone</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2119">#2119</a>)</li><li><code>ScaledPositiveSemidefiniteConeTriangle</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2154">#2154</a>)</li></ul></li></ul><h3 id="Fixed-5"><a class="docs-heading-anchor" href="#Fixed-5">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-5" title="Permalink"></a></h3><ul><li>Fixed support for Julia v1.9 to work around a bug in the upstream Julia compiler (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2161">#2161</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2163">#2163</a>)</li><li>Fixed a correctness bug in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge"><code>Bridges.Constraint.HermitianToSymmetricPSDBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2171">#2171</a>)</li><li>Fixed <code>convert(::VariableIndex, ::ScalarAffineFunction)</code> when the function has terms with <code>0</code>coefficients (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2173">#2173</a>)</li></ul><h3 id="Other-5"><a class="docs-heading-anchor" href="#Other-5">Other</a><a class="docs-heading-anchor-permalink" href="#Other-5" title="Permalink"></a></h3><ul><li>Fixed <code>solver-tests.yml</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2157">#2157</a>)</li><li>Updated documentation links to developer chatroom (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2160">#2160</a>)</li><li>Added various tests for bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2156">#2156</a>)</li><li>Added checklists to the developer documentation (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2167">#2167</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2168">#2168</a>)</li></ul><h2 id="[v1.15.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.1)-(April-25,-2023)"><a class="docs-heading-anchor" href="#[v1.15.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.1)-(April-25,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.1">v1.15.1</a> (April 25, 2023)</a><a id="[v1.15.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.1)-(April-25,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.15.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.1)-(April-25,-2023)" title="Permalink"></a></h2><h3 id="Fixed-6"><a class="docs-heading-anchor" href="#Fixed-6">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-6" title="Permalink"></a></h3><ul><li>Fixed deleting a variable in a bridged objective (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2150">#2150</a>)</li></ul><h2 id="[v1.15.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.0)-(April-19,-2023)"><a class="docs-heading-anchor" href="#[v1.15.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.0)-(April-19,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.0">v1.15.0</a> (April 19, 2023)</a><a id="[v1.15.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.0)-(April-19,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.15.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.0)-(April-19,-2023)" title="Permalink"></a></h2><h3 id="Added-6"><a class="docs-heading-anchor" href="#Added-6">Added</a><a class="docs-heading-anchor-permalink" href="#Added-6" title="Permalink"></a></h3><ul><li>Added <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.VectorFunctionizeBridge"><code>Bridges.Objective.VectorFunctionizeBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2139">#2139</a>)</li></ul><h3 id="Fixed-7"><a class="docs-heading-anchor" href="#Fixed-7">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-7" title="Permalink"></a></h3><ul><li>Fixed support for <code>Rational</code> in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitIntervalBridge"><code>Bridges.Constraint.SplitIntervalBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2137">#2137</a>)</li><li>Fixed printing of LaTeX models (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2141">#2141</a>)</li><li>Fixed <a href="../reference/modification/#MathOptInterface.modify"><code>modify</code></a> in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.VectorSlackBridge"><code>Bridges.Objective.VectorSlackBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2144">#2144</a>)</li><li>Fixed NAME record with spaces in <code>FileFormats.MPS</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2146">#2146</a>)</li><li>Fixed deleting a variable in a bridged objective (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2147">#2147</a>)</li></ul><h3 id="Other-6"><a class="docs-heading-anchor" href="#Other-6">Other</a><a class="docs-heading-anchor-permalink" href="#Other-6" title="Permalink"></a></h3><ul><li>Add a test for variables in one-sided open <a href="../reference/standard_form/#MathOptInterface.Interval"><code>Interval</code></a> sets (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2133">#2133</a>)</li><li>Minor style fixes in the source code (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2148">#2148</a>)</li></ul><h2 id="[v1.14.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.1)-(April-6,-2023)"><a class="docs-heading-anchor" href="#[v1.14.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.1)-(April-6,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.1">v1.14.1</a> (April 6, 2023)</a><a id="[v1.14.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.1)-(April-6,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.14.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.1)-(April-6,-2023)" title="Permalink"></a></h2><h3 id="Fixed-8"><a class="docs-heading-anchor" href="#Fixed-8">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-8" title="Permalink"></a></h3><ul><li>Fixed a bug in <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.print_active_bridges"><code>Bridges.print_active_bridges</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2135">#2135</a>)</li></ul><h3 id="Other-7"><a class="docs-heading-anchor" href="#Other-7">Other</a><a class="docs-heading-anchor-permalink" href="#Other-7" title="Permalink"></a></h3><ul><li>Added a warning when an ambiguous string is passed to <code>exclude</code> in <a href="../submodules/Test/reference/#MathOptInterface.Test.runtests"><code>Test.runtests</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2136">#2136</a>)</li></ul><h2 id="[v1.14.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.0)-(April-4,-2023)"><a class="docs-heading-anchor" href="#[v1.14.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.0)-(April-4,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.0">v1.14.0</a> (April 4, 2023)</a><a id="[v1.14.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.0)-(April-4,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.14.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.0)-(April-4,-2023)" title="Permalink"></a></h2><h3 id="Added-7"><a class="docs-heading-anchor" href="#Added-7">Added</a><a class="docs-heading-anchor-permalink" href="#Added-7" title="Permalink"></a></h3><ul><li>Added support for starting values in <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.QuadtoSOCBridge"><code>Bridges.Constraint.QuadtoSOCBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2115">#2115</a>)</li><li>Added support for <code>Regex</code> in the <code>include</code> and <code>exclude</code> arguments to <a href="../submodules/Test/reference/#MathOptInterface.Test.runtests"><code>Test.runtests</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2129">#2129</a>)</li><li>Added <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.print_active_bridges"><code>Bridges.print_active_bridges</code></a> methods for individual objectives and constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2128">#2128</a>)</li></ul><h3 id="Fixed-9"><a class="docs-heading-anchor" href="#Fixed-9">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-9" title="Permalink"></a></h3><ul><li>Fixed <a href="../reference/models/#MathOptInterface.ResultCount"><code>ResultCount</code></a> when parsing <code>.sol</code> files in <code>FileFormats.NL</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2130">#2130</a>)</li></ul><h2 id="[v1.13.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.2)-(March-21,-2023)"><a class="docs-heading-anchor" href="#[v1.13.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.2)-(March-21,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.2">v1.13.2</a> (March 21, 2023)</a><a id="[v1.13.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.2)-(March-21,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.13.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.2)-(March-21,-2023)" title="Permalink"></a></h2><h3 id="Fixed-10"><a class="docs-heading-anchor" href="#Fixed-10">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-10" title="Permalink"></a></h3><ul><li>Fixed splatting of containers in <code>MOI.Nonlinear</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2120">#2120</a>)</li><li>Fixed a bug reading LP files with default bounds (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2121">#2121</a>)</li><li>Fixed a bug in which <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge"><code>Bridges.Constraint.HermitianToSymmetricPSDBridge</code></a> was not enabled by default (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2123">#2123</a>)</li></ul><h3 id="Other-8"><a class="docs-heading-anchor" href="#Other-8">Other</a><a class="docs-heading-anchor-permalink" href="#Other-8" title="Permalink"></a></h3><ul><li>Fixed typos in the documentation (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2114">#2114</a>)</li><li>Functions now print to the REPL in algebraic form. This is potentially breaking if you have tests which rely on a specific <code>String</code> form of MOI functions. (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2112">#2112</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2126">#2126</a>)</li></ul><h2 id="[v1.13.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.1)-(March-3,-2023)"><a class="docs-heading-anchor" href="#[v1.13.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.1)-(March-3,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.1">v1.13.1</a> (March 3, 2023)</a><a id="[v1.13.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.1)-(March-3,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.13.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.1)-(March-3,-2023)" title="Permalink"></a></h2><h3 id="Other-9"><a class="docs-heading-anchor" href="#Other-9">Other</a><a class="docs-heading-anchor-permalink" href="#Other-9" title="Permalink"></a></h3><ul><li>Added the Google style guide to the documentation linter Vale, and fixed the resulting warnings (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2110">#2110</a>)</li><li>Improved the docstrings in <code>src/functions.jl</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2108">#2108</a>)</li></ul><h2 id="[v1.13.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.0)-(February-28,-2023)"><a class="docs-heading-anchor" href="#[v1.13.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.0)-(February-28,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.0">v1.13.0</a> (February 28, 2023)</a><a id="[v1.13.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.0)-(February-28,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.13.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.0)-(February-28,-2023)" title="Permalink"></a></h2><h3 id="Added-8"><a class="docs-heading-anchor" href="#Added-8">Added</a><a class="docs-heading-anchor-permalink" href="#Added-8" title="Permalink"></a></h3><ul><li>Added <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NumberConversionBridge"><code>Bridges.Constraint.NumberConversionBridge</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2091">#2091</a>)</li><li>Added <a href="../reference/standard_form/#MathOptInterface.Parameter"><code>Parameter</code></a> set (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2095">#2095</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2105">#2105</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2106">#2106</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2109">#2109</a>)</li><li>Added <code>with_cache_type</code> argument to <a href="../reference/models/#MathOptInterface.instantiate"><code>instantiate</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2097">#2097</a>)</li><li>Added support for <a href="../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>HermitianPositiveSemidefiniteConeTriangle</code></a> in <code>Utilities.Model</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2100">#2100</a>)</li></ul><h3 id="Fixed-11"><a class="docs-heading-anchor" href="#Fixed-11">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-11" title="Permalink"></a></h3><ul><li>Fixed bug when <code>Utilities.@product_of_sets</code> is empty (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2101">#2101</a>)</li><li>Fixed <a href="../submodules/Bridges/reference/#MathOptInterface.Bridges.print_active_bridges"><code>Bridges.print_active_bridges</code></a> when variable bridge is an <a href="../reference/standard_form/#MathOptInterface.AbstractScalarSet"><code>AbstractScalarSet</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2107">#2107</a>)</li></ul><h3 id="Other-10"><a class="docs-heading-anchor" href="#Other-10">Other</a><a class="docs-heading-anchor-permalink" href="#Other-10" title="Permalink"></a></h3><ul><li>Added tests for vector-valued objective functions in <code>FileFormats.MOF</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2093">#2093</a>)</li><li>Used and documented preference for <code>import MathOptInterface as MOI</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2096">#2096</a>)</li><li>Fix and test links in the documentation with <code>linkcheck = true</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2098">#2098</a>)</li><li>Improved docstrings of sets in <code>src/sets.jl</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2099">#2099</a>)</li><li>Skip checking flakey links in documentation with <code>linkcheck_ignore</code>  (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2103">#2103</a>)</li></ul><h2 id="[v1.12.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.12.0)-(February-10,-2023)"><a class="docs-heading-anchor" href="#[v1.12.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.12.0)-(February-10,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.12.0">v1.12.0</a> (February 10, 2023)</a><a id="[v1.12.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.12.0)-(February-10,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.12.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.12.0)-(February-10,-2023)" title="Permalink"></a></h2><h3 id="Added-9"><a class="docs-heading-anchor" href="#Added-9">Added</a><a class="docs-heading-anchor-permalink" href="#Added-9" title="Permalink"></a></h3><ul><li>Added support for vector-valued objective functions (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2070">#2070</a>)</li><li>Added a <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.distance_to_set"><code>Utilities.distance_to_set</code></a> method for <a href="../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>SecondOrderCone</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2060">#2060</a>)</li></ul><h3 id="Fixed-12"><a class="docs-heading-anchor" href="#Fixed-12">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-12" title="Permalink"></a></h3><ul><li>Fixed a number of constraint bridges so that <code>Bridges.final_touch</code> can be called multiple times without forcing a rebuild of the reformulation (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2089">#2089</a>)</li></ul><h3 id="Other-11"><a class="docs-heading-anchor" href="#Other-11">Other</a><a class="docs-heading-anchor-permalink" href="#Other-11" title="Permalink"></a></h3><ul><li>Added new tests that a set <a href="../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> appears in <a href="../reference/models/#MathOptInterface.ListOfModelAttributesSet"><code>ListOfModelAttributesSet</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2085">#2085</a>)</li><li>Improved the docstrings of a number of constraint-programming related sets (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2087">#2087</a>)</li></ul><h2 id="[v1.11.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.5)-(January-24,-2023)"><a class="docs-heading-anchor" href="#[v1.11.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.5)-(January-24,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.5">v1.11.5</a> (January 24, 2023)</a><a id="[v1.11.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.5)-(January-24,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.11.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.5)-(January-24,-2023)" title="Permalink"></a></h2><h3 id="Fixed-13"><a class="docs-heading-anchor" href="#Fixed-13">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-13" title="Permalink"></a></h3><ul><li>Fixed a bug writing <code>.lp</code> files with an off-diagonal quadratic objective (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2082">#2082</a>)</li></ul><h3 id="Other-12"><a class="docs-heading-anchor" href="#Other-12">Other</a><a class="docs-heading-anchor-permalink" href="#Other-12" title="Permalink"></a></h3><ul><li>Added <code>SnoopPrecompile</code> directives for reduced time-to-first-X in Julia v1.9 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2080">#2080</a>)</li></ul><h2 id="[v1.11.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.4)-(January-12,-2023)"><a class="docs-heading-anchor" href="#[v1.11.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.4)-(January-12,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.4">v1.11.4</a> (January 12, 2023)</a><a id="[v1.11.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.4)-(January-12,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.11.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.4)-(January-12,-2023)" title="Permalink"></a></h2><h3 id="Fixed-14"><a class="docs-heading-anchor" href="#Fixed-14">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-14" title="Permalink"></a></h3><ul><li>Fixed a bug reading <code>.lp</code> files with an <code>Integer</code> section (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2078">#2078</a>)</li></ul><h2 id="[v1.11.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.3)-(January-12,-2023)"><a class="docs-heading-anchor" href="#[v1.11.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.3)-(January-12,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.3">v1.11.3</a> (January 12, 2023)</a><a id="[v1.11.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.3)-(January-12,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.11.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.3)-(January-12,-2023)" title="Permalink"></a></h2><h3 id="Fixed-15"><a class="docs-heading-anchor" href="#Fixed-15">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-15" title="Permalink"></a></h3><ul><li>Fixed a performance bug when deleting a vector of constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2072">#2072</a>)</li><li>Fixed a bug reading <code>.lp</code> files with terms like <code>x -1 y</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2076">#2076</a>)</li></ul><h3 id="Other-13"><a class="docs-heading-anchor" href="#Other-13">Other</a><a class="docs-heading-anchor-permalink" href="#Other-13" title="Permalink"></a></h3><ul><li>Documented the two-argument method of <a href="../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2074">#2074</a>)</li></ul><h2 id="[v1.11.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.2)-(January-2,-2023)"><a class="docs-heading-anchor" href="#[v1.11.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.2)-(January-2,-2023)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.2">v1.11.2</a> (January 2, 2023)</a><a id="[v1.11.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.2)-(January-2,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.11.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.2)-(January-2,-2023)" title="Permalink"></a></h2><h3 id="Fixed-16"><a class="docs-heading-anchor" href="#Fixed-16">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-16" title="Permalink"></a></h3><ul><li>Fixed a bug reading <code>.mof.json</code> files with <code>ConstraintName</code> set for <code>VariableIndex</code> constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2066">#2066</a>)</li><li>Fixed a bug reading <code>.mof.json</code> files with nonlinear objectives and no constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2068">#2068</a>)</li></ul><h2 id="[v1.11.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.1)-(December-22,-2022)"><a class="docs-heading-anchor" href="#[v1.11.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.1)-(December-22,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.1">v1.11.1</a> (December 22, 2022)</a><a id="[v1.11.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.1)-(December-22,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.11.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.1)-(December-22,-2022)" title="Permalink"></a></h2><h3 id="Fixed-17"><a class="docs-heading-anchor" href="#Fixed-17">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-17" title="Permalink"></a></h3><ul><li>Fixed a bug reading <code>.mof.json</code> files with integer coefficients for affine and quadratic functions (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2063">#2063</a>)</li></ul><h2 id="[v1.11.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.0)-(December-2,-2022)"><a class="docs-heading-anchor" href="#[v1.11.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.0)-(December-2,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.0">v1.11.0</a> (December 2, 2022)</a><a id="[v1.11.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.0)-(December-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.11.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.0)-(December-2,-2022)" title="Permalink"></a></h2><h3 id="Added-10"><a class="docs-heading-anchor" href="#Added-10">Added</a><a class="docs-heading-anchor-permalink" href="#Added-10" title="Permalink"></a></h3><ul><li>Added <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.PenaltyRelaxation"><code>Utilities.PenaltyRelaxation</code></a> and <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>Utilities.ScalarPenaltyRelaxation</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1995">#1995</a>)</li><li>Added <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.distance_to_set"><code>Utilities.distance_to_set</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2048">#2048</a>)</li><li>Added support for <a href="../reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>ConstraintPrimalStart</code></a> and <a href="../reference/constraints/#MathOptInterface.ConstraintDualStart"><code>ConstraintDualStart</code></a> in <code>FileFormats.MOF</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2056">#2056</a>)</li></ul><h3 id="Other-14"><a class="docs-heading-anchor" href="#Other-14">Other</a><a class="docs-heading-anchor-permalink" href="#Other-14" title="Permalink"></a></h3><ul><li>Tidied these release notes (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2055">#2055</a>)</li></ul><h2 id="[v1.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.10.0)-(November-22,-2022)"><a class="docs-heading-anchor" href="#[v1.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.10.0)-(November-22,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.10.0">v1.10.0</a> (November 22, 2022)</a><a id="[v1.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.10.0)-(November-22,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.10.0)-(November-22,-2022)" title="Permalink"></a></h2><h3 id="Added-11"><a class="docs-heading-anchor" href="#Added-11">Added</a><a class="docs-heading-anchor-permalink" href="#Added-11" title="Permalink"></a></h3><ul><li>Added new methods <code>set(::OptimizerWithAttributes, ::RawOptimizerAttribute, value)</code> and <code>get(::OptimizerWithAttributes, ::RawOptimizerAttribute)</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2049">#2049</a>)</li><li>Added new methods <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.outer_keys"><code>Utilities.DoubleDicts.outer_keys</code></a> and <a href="../submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys"><code>Utilities.DoubleDicts.nonempty_outer_keys</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2052">#2052</a>)</li></ul><h3 id="Fixed-18"><a class="docs-heading-anchor" href="#Fixed-18">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-18" title="Permalink"></a></h3><ul><li>Fixed <a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.SlackBridge"><code>Bridges.Objective.SlackBridge</code></a> when the objective function is complex-valued (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2036">#2036</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2038">#2038</a>)</li><li>Fixed docstring of <a href="../submodules/Test/reference/#MathOptInterface.Test.runtests"><code>Test.runtests</code></a> to clarify the <code>warn_unsupported</code> argument (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2037">#2037</a>)</li><li>Fixed reading of free variables in <code>FileFormats.LP</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2044">#2044</a>)</li><li>Fixed numerous edge cases reading files from QPLIB using <code>FileFormats.LP</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2042">#2042</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2044">#2044</a>)</li><li>Fixed situations in which <code>x^y</code> returns a complex value in <code>Nonlinear</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2050">#2050</a>)</li></ul><h3 id="Other-15"><a class="docs-heading-anchor" href="#Other-15">Other</a><a class="docs-heading-anchor-permalink" href="#Other-15" title="Permalink"></a></h3><ul><li>Improved the error message thrown when a user-defined nonlinear function does not accept splatted input (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2032">#2032</a>)</li><li>Removed specialized iterators for <code>keys</code> and <code>values</code> in <code>Utilities.CleverDicts</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2051">#2051</a>)</li></ul><h2 id="[v1.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.9.0)-(October-29,-2022)"><a class="docs-heading-anchor" href="#[v1.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.9.0)-(October-29,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.9.0">v1.9.0</a> (October 29, 2022)</a><a id="[v1.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.9.0)-(October-29,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.9.0)-(October-29,-2022)" title="Permalink"></a></h2><h3 id="Added-12"><a class="docs-heading-anchor" href="#Added-12">Added</a><a class="docs-heading-anchor-permalink" href="#Added-12" title="Permalink"></a></h3><ul><li>Added default fallback for getting <a href="../reference/models/#MathOptInterface.ListOfConstraintIndices"><code>ListOfConstraintIndices</code></a> and <a href="../reference/models/#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints</code></a> when the constraint type is unsupported by the model (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2021">#2021</a>)</li><li>Added support for <code>min</code> and <code>max</code> in nonlinear expressions (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2023">#2023</a>)</li><li>Added support for <code>Indicator{EqualTo{T}}</code> constraints in <code>FileFormats.MPS</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2022">#2022</a>)</li><li>Added default fallback for <a href="../reference/models/#MathOptInterface.write_to_file"><code>write_to_file</code></a> and <a href="../reference/models/#MathOptInterface.read_from_file"><code>read_from_file</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2029">#2029</a>)</li></ul><h3 id="Fixed-19"><a class="docs-heading-anchor" href="#Fixed-19">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-19" title="Permalink"></a></h3><ul><li>Fixed <code>Constraint.ZeroOneBridge</code> by adding new bounds as affine constraints instead of variable bounds (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1879">#1879</a>)</li><li>Fixed reading free rows in <code>FileFormats.MPS</code> files (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2009">#2009</a>)</li><li>Fixed parsing of <code>OBJSENSE</code> blocks in <code>FileFormats.MPS</code> files (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2016">#2016</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2019">#2019</a>)</li><li>Fixed the parsing of deeply nested nonlinear expressions by removing the use of recursion (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2020">#2020</a>)</li><li>Fixed the requirements check in <code>Test.test_constrainnt_get_ConstraintIndex</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2024">#2024</a>)</li></ul><h2 id="[v1.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.2)-(September-20,-2022)"><a class="docs-heading-anchor" href="#[v1.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.2)-(September-20,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.2">v1.8.2</a> (September 20, 2022)</a><a id="[v1.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.2)-(September-20,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.2)-(September-20,-2022)" title="Permalink"></a></h2><h3 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h3><ul><li>Added <a href="https://vale.sh">vale</a> as a documentation linter (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2002">#2002</a>)</li><li>Improved styling of code blocks in the PDF (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1999">#1999</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2000">#2000</a>)</li><li>Fixed a number of typos in the documentation (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2001">#2001</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/2003">#2003</a>)</li></ul><h2 id="[v1.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.1)-(September-12,-2022)"><a class="docs-heading-anchor" href="#[v1.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.1)-(September-12,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.1">v1.8.1</a> (September 12, 2022)</a><a id="[v1.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.1)-(September-12,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.1)-(September-12,-2022)" title="Permalink"></a></h2><h3 id="Fixed-20"><a class="docs-heading-anchor" href="#Fixed-20">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-20" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>supports(::AbstractBridgeOptimizer</code> for constraint attributes (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1991">#1991</a>) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1992">#1992</a>)</li></ul><h2 id="[v1.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.0)-(September-1,-2022)"><a class="docs-heading-anchor" href="#[v1.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.0)-(September-1,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.0">v1.8.0</a> (September 1, 2022)</a><a id="[v1.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.0)-(September-1,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.0)-(September-1,-2022)" title="Permalink"></a></h2><h3 id="Added-13"><a class="docs-heading-anchor" href="#Added-13">Added</a><a class="docs-heading-anchor-permalink" href="#Added-13" title="Permalink"></a></h3><ul><li>Added new sets<ul><li><a href="../reference/standard_form/#MathOptInterface.HyperRectangle"><code>HyperRectangle</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1961">#1961</a>)</li><li><a href="../reference/standard_form/#MathOptInterface.Reified"><code>Reified</code></a> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1955">#1955</a>)</li></ul></li><li>Added new bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1955">#1955</a>)<ul><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge"><code>Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ReifiedCountDistinctToMILPBridge"><code>Bridges.Constraint.ReifiedCountDistinctToMILPBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitHyperRectangleBridge"><code>Bridges.Constraint.SplitHyperRectangleBridge</code></a></li></ul></li><li>Added support for <code>atan(y, x)</code> in <code>Nonlinear</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1987">#1987</a>)</li></ul><h3 id="Fixed-21"><a class="docs-heading-anchor" href="#Fixed-21">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-21" title="Permalink"></a></h3><ul><li>Lazily construct expressions in <code>Nonlinear</code> so that expressions are updated when <code>Nonlinear.Parameter</code> values are updated (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1984">#1984</a>)</li><li>Allow <code>NORM_LIMIT</code> as a <code>TerminationStatus</code> for unbounded problems in <code>Test</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1990">#1990</a>)</li></ul><h2 id="[v1.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.7.0)-(August-16,-2022)"><a class="docs-heading-anchor" href="#[v1.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.7.0)-(August-16,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.7.0">v1.7.0</a> (August 16, 2022)</a><a id="[v1.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.7.0)-(August-16,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.7.0)-(August-16,-2022)" title="Permalink"></a></h2><h3 id="Added-14"><a class="docs-heading-anchor" href="#Added-14">Added</a><a class="docs-heading-anchor-permalink" href="#Added-14" title="Permalink"></a></h3><ul><li>Added new sets<ul><li><a href="../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>HermitianPositiveSemidefiniteConeTriangle</code></a></li></ul></li><li>Added new optimizer-independent options<ul><li><a href="../reference/models/#MathOptInterface.RelativeGapTolerance"><code>RelativeGapTolerance</code></a></li><li><a href="../reference/models/#MathOptInterface.AbsoluteGapTolerance"><code>AbsoluteGapTolerance</code></a></li></ul></li><li>Added new bridges<ul><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge"><code>Bridges.Constraint.GeoMeanToPowerBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge"><code>Bridges.Constraint.HermitianToSymmetricPSDBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorGreaterToLessThanBridge"><code>Bridges.Constraint.IndicatorGreaterToLessThanBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorLessToGreaterThanBridge"><code>Bridges.Constraint.IndicatorLessToGreaterThanBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitComplexZerosBridge"><code>Bridges.Constraint.SplitComplexZerosBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitComplexEqualToBridge"><code>Bridges.Constraint.SplitComplexEqualToBridge</code></a></li><li><a href="../submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.QuadratizeBridge"><code>Bridges.Objective.QuadratizeBridge</code></a></li></ul></li><li>Added support for generic number types in <code>Utilities.loadfromstring!</code></li><li>Updated <code>FileFormats.MOF</code> to MathOptFormat v1.1, enabling support for constraint programming sets in the <code>MOF</code> file format</li><li>Added support in various <code>FileFormats</code> for<ul><li>indicator constraints in <code>FileFormats.MPS</code></li><li>quadratic constraints and an objective in <code>FileFormats.LP</code></li><li>quadratic constraints and an objective in <code>FileFormats.MPS</code></li></ul></li></ul><h3 id="Fixed-22"><a class="docs-heading-anchor" href="#Fixed-22">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-22" title="Permalink"></a></h3><ul><li>Fixed some missing promotion rules</li></ul><h3 id="Other-16"><a class="docs-heading-anchor" href="#Other-16">Other</a><a class="docs-heading-anchor-permalink" href="#Other-16" title="Permalink"></a></h3><ul><li>Improved the performance of Jacobian products in <code>Nonlinear</code></li><li>Removed an un-needed <code>copy</code> in <code>Utilities.modify_function</code></li><li>Various clean-ups in <code>Bridges/bridge_optimizer.jl</code></li></ul><h2 id="[v1.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.1)-(July-23,-2022)"><a class="docs-heading-anchor" href="#[v1.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.1)-(July-23,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.1">v1.6.1</a> (July 23, 2022)</a><a id="[v1.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.1)-(July-23,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.1)-(July-23,-2022)" title="Permalink"></a></h2><h3 id="Fixed-23"><a class="docs-heading-anchor" href="#Fixed-23">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-23" title="Permalink"></a></h3><ul><li>Added support for <code>ExponentialCone</code> in <code>MatrixOfConstraints</code></li><li>Fix <code>PSDSquare_3</code> test to reflect a previously fixed bug getting the <code>ConstraintDual</code> of a <code>PositiveSemidefiniteConeSquare</code> constraint</li></ul><h2 id="[v1.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.0)-(July-2,-2022)"><a class="docs-heading-anchor" href="#[v1.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.0)-(July-2,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.0">v1.6.0</a> (July 2, 2022)</a><a id="[v1.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.0)-(July-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.0)-(July-2,-2022)" title="Permalink"></a></h2><h3 id="Added-15"><a class="docs-heading-anchor" href="#Added-15">Added</a><a class="docs-heading-anchor-permalink" href="#Added-15" title="Permalink"></a></h3><ul><li>Added <code>Bridges.needs_final_touch</code> and <code>Bridges.final_touch</code></li><li>Added new bridges from constraint programming sets to mixed-integer linear programs:<ul><li><code>AllDifferentToCountDistinctBridge</code></li><li><code>CountAtLeastToCountBelongsBridge</code></li><li><code>CountBelongsToMILPBridge</code></li><li><code>CountDistinctToMILPBridge</code></li><li><code>CountGreaterThanToMILPBridge</code></li><li><code>CircuitToMILPBridge</code></li></ul></li></ul><h3 id="Fixed-24"><a class="docs-heading-anchor" href="#Fixed-24">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-24" title="Permalink"></a></h3><ul><li>Relax an instance of <code>::Vector</code> to <code>::AbstractVector</code> in <code>MOI.Nonlinear</code></li><li>Fix <code>BinPackingToMILPBridge</code> to respect variable bounds</li><li>Fix <code>SemiToBinaryBridge</code> to throw error if other bounds are set</li></ul><h2 id="[v1.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.5.0)-(June-27,-2022)"><a class="docs-heading-anchor" href="#[v1.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.5.0)-(June-27,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.5.0">v1.5.0</a> (June 27, 2022)</a><a id="[v1.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.5.0)-(June-27,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.5.0)-(June-27,-2022)" title="Permalink"></a></h2><h3 id="Added-16"><a class="docs-heading-anchor" href="#Added-16">Added</a><a class="docs-heading-anchor-permalink" href="#Added-16" title="Permalink"></a></h3><ul><li>Added <code>GetAttributeNotAllowed</code> for solvers to indicate when getting an attribute encounters an error</li><li>Added <code>Utilities.get_fallback</code> support for <code>ObjectiveValue</code> and <code>DualObjectiveValue</code></li><li>Added new bridges:<ul><li><code>RootDetConeSquare</code> to <code>RootDetConeTriangle</code></li><li><code>LogDetConeSquare</code> to <code>LogDetConeTriangle</code></li><li><code>BinPacking</code> to a mixed-integer linear program</li><li><code>Table</code> to a mixed-integer linear program</li></ul></li><li>Added <code>Bridges.print_active_bridges</code> to display the current optimal hyper-path in a <code>Bridges.LazyBridgeOptimizer</code></li></ul><h3 id="Fixed-25"><a class="docs-heading-anchor" href="#Fixed-25">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-25" title="Permalink"></a></h3><ul><li>Fixed <code>ZeroOne</code> tests with lower and upper bounds</li><li>Fixed error in <code>FileFormats.LP</code> when reading a malformed file</li><li>Fixed reading of nonlinear programs in <code>FileFormats.MOF</code></li><li>Fixed bug in <code>ConstraintDual</code> when using <code>SquareBridge</code></li></ul><h3 id="Other-17"><a class="docs-heading-anchor" href="#Other-17">Other</a><a class="docs-heading-anchor-permalink" href="#Other-17" title="Permalink"></a></h3><ul><li>Improved documentation of nonlinear API</li><li>Documented duality convention for <code>PositiveSemidefiniteConeSquare</code> sets</li><li>Fixed typo in <code>Bridges.Constraint.QuadToSOCBridge</code> docstring</li></ul><h2 id="[v1.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.4.0)-(June-9,-2022)"><a class="docs-heading-anchor" href="#[v1.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.4.0)-(June-9,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.4.0">v1.4.0</a> (June 9, 2022)</a><a id="[v1.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.4.0)-(June-9,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.4.0)-(June-9,-2022)" title="Permalink"></a></h2><h3 id="Added-17"><a class="docs-heading-anchor" href="#Added-17">Added</a><a class="docs-heading-anchor-permalink" href="#Added-17" title="Permalink"></a></h3><ul><li>Added a number of sets for constraint programming:<ul><li><code>AllDifferent</code></li><li><code>BinPacking</code></li><li><code>Circuit</code></li><li><code>CountAtLeast</code></li><li><code>CountBelongs</code></li><li><code>CountDistinct</code></li><li><code>CountGreaterThan</code></li><li><code>Cumulative</code></li><li><code>Path</code></li><li><code>Table</code></li></ul></li><li>Added support for user-defined hessians in <code>Nonlinear</code></li><li>Added <code>Bridges.runtests</code> to simplify the testing of bridge implementations</li></ul><h3 id="Fixed-26"><a class="docs-heading-anchor" href="#Fixed-26">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-26" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>FileFormats.NL</code> when writing univariate <code>*</code></li></ul><h3 id="Other-18"><a class="docs-heading-anchor" href="#Other-18">Other</a><a class="docs-heading-anchor-permalink" href="#Other-18" title="Permalink"></a></h3><ul><li>Began a large refactoring of the <code>Bridges</code> submodule, with greatly improved documentation.</li></ul><h2 id="[v1.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.3.0)-(May-27,-2022)"><a class="docs-heading-anchor" href="#[v1.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.3.0)-(May-27,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.3.0">v1.3.0</a> (May 27, 2022)</a><a id="[v1.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.3.0)-(May-27,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.3.0)-(May-27,-2022)" title="Permalink"></a></h2><h3 id="Added-18"><a class="docs-heading-anchor" href="#Added-18">Added</a><a class="docs-heading-anchor-permalink" href="#Added-18" title="Permalink"></a></h3><ul><li>Add <code>MOI.Nonlinear</code> submodule. This is a large new submodule that has been refactored from code that was in JuMP. For now, it should be considered experimental.</li><li>Add <code>FileFormats.NL.SolFileResults(::IO, ::Model)</code></li><li>Add <code>FileFormats.NL.read!(::IO, ::Model)</code></li><li>Add <code>MOI.modify</code> that accepts a vector of modifications</li></ul><h3 id="Fixed-27"><a class="docs-heading-anchor" href="#Fixed-27">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-27" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>Test</code> which attempted to include non-<code>.jl</code> files</li><li>Fixed a bug in <code>FileFormats</code> for models with open interval constraints</li></ul><h3 id="Other-19"><a class="docs-heading-anchor" href="#Other-19">Other</a><a class="docs-heading-anchor-permalink" href="#Other-19" title="Permalink"></a></h3><ul><li>Fixed a performance issue in <code>Utilities.DoubleDict</code></li><li>Various minor improvements to the documentation</li></ul><h2 id="[v1.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.2.0)-(April-25,-2022)"><a class="docs-heading-anchor" href="#[v1.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.2.0)-(April-25,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.2.0">v1.2.0</a> (April 25, 2022)</a><a id="[v1.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.2.0)-(April-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.2.0)-(April-25,-2022)" title="Permalink"></a></h2><h3 id="Added-19"><a class="docs-heading-anchor" href="#Added-19">Added</a><a class="docs-heading-anchor-permalink" href="#Added-19" title="Permalink"></a></h3><ul><li>Add support for the <code>FORMAT_REW</code>/<code>.rew</code> file format in <code>FileFormats</code>.</li></ul><h3 id="Fixed-28"><a class="docs-heading-anchor" href="#Fixed-28">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-28" title="Permalink"></a></h3><ul><li>Fix bug handling of default variable bounds in <code>FileFormats.LP</code></li><li>Fix <code>FileFormats.MPS</code> to not write <code>OBJSENSE</code> by default since this is only supported by some readers.</li></ul><h2 id="[v1.1.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.2)-(March-31,-2022)"><a class="docs-heading-anchor" href="#[v1.1.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.2)-(March-31,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.2">v1.1.2</a> (March 31, 2022)</a><a id="[v1.1.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.2)-(March-31,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.1.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.2)-(March-31,-2022)" title="Permalink"></a></h2><h3 id="Fixed-29"><a class="docs-heading-anchor" href="#Fixed-29">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-29" title="Permalink"></a></h3><ul><li>Fix a range of bugs in <code>FileFormats.LP</code></li><li>Fix reading of problem dimensions in <code>FileFormats.SDPA</code></li></ul><h2 id="[v1.1.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.1)-(March-23,-2022)"><a class="docs-heading-anchor" href="#[v1.1.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.1)-(March-23,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.1">v1.1.1</a> (March 23, 2022)</a><a id="[v1.1.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.1)-(March-23,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.1.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.1)-(March-23,-2022)" title="Permalink"></a></h2><h3 id="Fixed-30"><a class="docs-heading-anchor" href="#Fixed-30">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-30" title="Permalink"></a></h3><ul><li>Fix bug in <code>test_model_UpperBoundAlreadySet</code></li><li>Fix bug in <code>test_infeasible_</code> tests</li><li>Fix bug in <code>test_objective_ObjectiveFunction_blank</code></li><li>Relax restriction of <code>MOI.AbstractOptimizer</code> to <code>MOI.ModelLike</code> in <code>Utilities.CachingOptimizer</code> and <code>instantiate</code>.</li></ul><h3 id="New-tests"><a class="docs-heading-anchor" href="#New-tests">New tests</a><a id="New-tests-1"></a><a class="docs-heading-anchor-permalink" href="#New-tests" title="Permalink"></a></h3><ul><li>Add <code>test_conic_empty_matrix</code> that checks conic solvers support problems with no variables.</li></ul><h2 id="[v1.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.0)-(March-2,-2022)"><a class="docs-heading-anchor" href="#[v1.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.0)-(March-2,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.0">v1.1.0</a> (March 2, 2022)</a><a id="[v1.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.0)-(March-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.0)-(March-2,-2022)" title="Permalink"></a></h2><h3 id="Added-20"><a class="docs-heading-anchor" href="#Added-20">Added</a><a class="docs-heading-anchor-permalink" href="#Added-20" title="Permalink"></a></h3><ul><li>Added <code>MOI.Utilities.throw_unsupported(::UniversalFallback)</code> for simplifying solver wrappers which copy from a <code>UniversalFallback</code>.</li></ul><h2 id="[v1.0.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.2)-(March-1,-2022)"><a class="docs-heading-anchor" href="#[v1.0.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.2)-(March-1,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.2">v1.0.2</a> (March 1, 2022)</a><a id="[v1.0.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.2)-(March-1,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.0.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.2)-(March-1,-2022)" title="Permalink"></a></h2><h3 id="Fixed-31"><a class="docs-heading-anchor" href="#Fixed-31">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-31" title="Permalink"></a></h3><ul><li>Fixed a bug in the <code>test_model_ScalarFunctionConstantNotZero</code> test</li><li>Fixed the error type when an <code>AbstractFunctionConversionBridge</code> cannot get or set an attribute</li><li>Identified a correctness bug in <code>RSOCtoPSDBridge</code>. We now thrown an error instead of returning an incorrect result.</li></ul><h2 id="[v1.0.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.1)-(February-25,-2022)"><a class="docs-heading-anchor" href="#[v1.0.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.1)-(February-25,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.1">v1.0.1</a> (February 25, 2022)</a><a id="[v1.0.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.1)-(February-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.0.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.1)-(February-25,-2022)" title="Permalink"></a></h2><h3 id="Fixed-32"><a class="docs-heading-anchor" href="#Fixed-32">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-32" title="Permalink"></a></h3><ul><li>Fixed a bug in which OptimizerAttributes were not copied in CachingOptimizer</li><li>Fixed a bug in which <code>shift_constant</code> did not promote mixed types of coefficients</li><li>Fixed a bug in which deleting a constraint of a bridged variable threw <code>ErrorException</code> instead of <code>MOI.DeleteNotAllowed</code></li><li>Fixed a bug in which <code>add_constraint</code> in <code>MatrixOfConstraints</code> did not canonicalize the function</li><li>Fixed a bug when modifying scalar constants of a function containing a bridged variable</li><li>Fixed a bug in which <code>final_touch</code> was not always called with a <code>CachingOptimizer</code></li></ul><h2 id="[v1.0.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.0)-(February-17,-2022)"><a class="docs-heading-anchor" href="#[v1.0.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.0)-(February-17,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.0">v1.0.0</a> (February 17, 2022)</a><a id="[v1.0.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.0)-(February-17,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v1.0.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.0)-(February-17,-2022)" title="Permalink"></a></h2><p>Although tagged as a breaking release, v1.0.0 is v0.10.9 with deprecations removed, similar to how Julia 1.0 was Julia 0.7 with deprecations removed.</p><h3 id="Breaking"><a class="docs-heading-anchor" href="#Breaking">Breaking</a><a id="Breaking-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking" title="Permalink"></a></h3><ul><li>Julia 1.6 is now the minimum supported version</li><li>All deprecations have been removed</li></ul><h3 id="Troubleshooting-problems-when-updating"><a class="docs-heading-anchor" href="#Troubleshooting-problems-when-updating">Troubleshooting problems when updating</a><a id="Troubleshooting-problems-when-updating-1"></a><a class="docs-heading-anchor-permalink" href="#Troubleshooting-problems-when-updating" title="Permalink"></a></h3><p>If you experience problems when updating, you are likely using previously deprecated features. (By default, Julia does not warn when you use deprecated features.)</p><p>To find the deprecated features you are using, start Julia with <code>--depwarn=yes</code>:</p><pre><code class="nohighlight hljs">$ julia --depwarn=yes</code></pre><p>Then install MathOptInterface v0.10.9:</p><pre><code class="language-julia hljs">julia&gt; using Pkg
+julia&gt; pkg&quot;add MathOptInterface@0.10&quot;</code></pre><p>And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.</p><h2 id="[v0.10.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.9)-(February-16,-2022)"><a class="docs-heading-anchor" href="#[v0.10.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.9)-(February-16,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.9">v0.10.9</a> (February 16, 2022)</a><a id="[v0.10.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.9)-(February-16,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.9)-(February-16,-2022)" title="Permalink"></a></h2><h3 id="Added-21"><a class="docs-heading-anchor" href="#Added-21">Added</a><a class="docs-heading-anchor-permalink" href="#Added-21" title="Permalink"></a></h3><ul><li>Added <code>MOI.Utilities.FreeVariables</code> as a new <code>VariablesConstrainer</code> for conic solvers</li><li>Added <code>MOI.default_cache</code> for specifying the model used in <code>CachingOptimizer</code></li></ul><h3 id="Fixed-33"><a class="docs-heading-anchor" href="#Fixed-33">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-33" title="Permalink"></a></h3><ul><li>Fixed LaTeX printing of <code>MOI.Interval</code> sets</li></ul><h3 id="Other-20"><a class="docs-heading-anchor" href="#Other-20">Other</a><a class="docs-heading-anchor-permalink" href="#Other-20" title="Permalink"></a></h3><ul><li>Added Aqua.jl as a CI check, and fixed suggested issues</li><li>The constructors of GeoMeanBridge, StructOfConstraints, and CachingOptimizer were changed from outer to inner constructors. This change is technically breaking, but does not impact users who followed the documented API.</li></ul><h2 id="[v0.10.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.8)-(February-3,-2022)"><a class="docs-heading-anchor" href="#[v0.10.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.8)-(February-3,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.8">v0.10.8</a> (February 3, 2022)</a><a id="[v0.10.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.8)-(February-3,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.8)-(February-3,-2022)" title="Permalink"></a></h2><h3 id="Added-22"><a class="docs-heading-anchor" href="#Added-22">Added</a><a class="docs-heading-anchor-permalink" href="#Added-22" title="Permalink"></a></h3><ul><li>Added a <code>Base.read!</code> for <code>FileFormats.LP</code>.</li></ul><h3 id="Fixed-34"><a class="docs-heading-anchor" href="#Fixed-34">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-34" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>MutableSparseMatrix</code></li><li>Fixed a bug when calling <code>operate!(vcat, ...)</code> with <code>Number</code> arguments</li><li>Removed unintended export of deprecated symbols</li><li>Fixed a bug with <code>PowerCone</code> and <code>DualPowerCone</code> in <code>MatrixOfConstraints</code>.</li></ul><h2 id="[v0.10.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.7)-(January-5,-2022)"><a class="docs-heading-anchor" href="#[v0.10.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.7)-(January-5,-2022)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.7">v0.10.7</a> (January 5, 2022)</a><a id="[v0.10.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.7)-(January-5,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.7)-(January-5,-2022)" title="Permalink"></a></h2><h3 id="Added-23"><a class="docs-heading-anchor" href="#Added-23">Added</a><a class="docs-heading-anchor-permalink" href="#Added-23" title="Permalink"></a></h3><ul><li>Added test for modifying the constant vector in a <code>VectorAffineFunction-in-Zeros</code> constraint.</li></ul><h3 id="Fixed-35"><a class="docs-heading-anchor" href="#Fixed-35">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-35" title="Permalink"></a></h3><ul><li>Fixed the order in which sets are added to a <code>LazyBridgeOptimizer</code>. Compared to v0.10.6, this may result in bridged models being created with a different number (and order) of variables and constraints. However, it was necessary to fix cases which were previously rejected as unsupported, even though there was a valid bridge transformation.</li><li>Fixed an error message in <code>FileFormats.CBF</code></li><li>Fixed comparison in <code>test_linear_integration_Interval</code></li><li>Fixed errors for <code>ConstraintPrimal</code> in a <code>CachingOptimizer</code></li><li>Fixed printing of models with non-<code>Float64</code> coefficients.</li></ul><h3 id="Other-21"><a class="docs-heading-anchor" href="#Other-21">Other</a><a class="docs-heading-anchor-permalink" href="#Other-21" title="Permalink"></a></h3><ul><li>Various improvements to reduce time-to-first-solve latency</li><li>Improved error message when an optimizer does not support <code>compute_conflict!</code></li></ul><h2 id="[v0.10.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.6)-(November-30,-2021)"><a class="docs-heading-anchor" href="#[v0.10.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.6)-(November-30,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.6">v0.10.6</a> (November 30, 2021)</a><a id="[v0.10.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.6)-(November-30,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.6)-(November-30,-2021)" title="Permalink"></a></h2><h3 id="Added-24"><a class="docs-heading-anchor" href="#Added-24">Added</a><a class="docs-heading-anchor-permalink" href="#Added-24" title="Permalink"></a></h3><ul><li>Added new documentation and tests for infeasibility certificates</li><li>Added a version control system for the tests in <code>MOI.Test.runtests</code>. Pass <code>exclude_tests_after = v&quot;0.10.5&quot;</code> to run tests added in v0.10.5 and earlier.</li><li><code>MOI.Test.runtests</code> now supports generic number types. To specify the number type <code>T</code>, pass <code>MOI.Test.Config(T)</code>.</li><li>Added <code>infeasible_status</code> to <code>MOI.Test.Config</code> for solvers which return <code>LOCALLY_INFEASIBLE</code></li><li>CachingOptimizers now use a fallback for <code>ConstraintPrimal</code>. This should enable solvers using a CachingOptimizer to pass tests requiring <code>ConstraintPrimal</code>.</li></ul><h3 id="Fixed-36"><a class="docs-heading-anchor" href="#Fixed-36">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-36" title="Permalink"></a></h3><ul><li>Fixed a StackOverflow bug in <code>copy_to</code></li><li>Fixed error thrown when nonconvex quadratic constraints cannot be bridged</li><li>Fixed a bug in <code>copy_to</code> for <code>FileFormats.NL.Model</code></li><li>Fixed a bug in <code>FileFormats.NL</code> when printing large integers</li><li>Remove a common test failure for <code>LowerBoundAlreadySet</code> tests</li><li><code>Utilities.num_rows</code> is now exported</li><li>Remove parts of failing <code>test_model_copy_to_xxx</code> tests due to bridges</li></ul><h2 id="[v0.10.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.5)-(November-7,-2021)"><a class="docs-heading-anchor" href="#[v0.10.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.5)-(November-7,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.5">v0.10.5</a> (November 7, 2021)</a><a id="[v0.10.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.5)-(November-7,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.5)-(November-7,-2021)" title="Permalink"></a></h2><h3 id="Fixed-37"><a class="docs-heading-anchor" href="#Fixed-37">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-37" title="Permalink"></a></h3><ul><li>Fixed getter in <code>UniversalFallback</code></li><li>Fixed <code>test_solve_conflict_zeroone_ii</code></li></ul><h3 id="Other-22"><a class="docs-heading-anchor" href="#Other-22">Other</a><a class="docs-heading-anchor-permalink" href="#Other-22" title="Permalink"></a></h3><ul><li>Make <code>normalize_and_add_constraint</code> more flexible</li><li>Update paper BibTeX</li></ul><h2 id="[v0.10.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.4)-(October-26,-2021)"><a class="docs-heading-anchor" href="#[v0.10.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.4)-(October-26,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.4">v0.10.4</a> (October 26, 2021)</a><a id="[v0.10.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.4)-(October-26,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.4)-(October-26,-2021)" title="Permalink"></a></h2><h3 id="Added-25"><a class="docs-heading-anchor" href="#Added-25">Added</a><a class="docs-heading-anchor-permalink" href="#Added-25" title="Permalink"></a></h3><ul><li>Add <code>SolverVersion</code> attribute</li><li>Add new tests:<ul><li><code>test_solve_conflict_zeroone_ii</code></li><li><code>test_nonlinear_objective</code></li></ul></li><li><code>Utilities.VariablesContainer</code> now supports <code>ConstraintFunction</code> and <code>ConstraintSet</code></li><li>The documentation is now available as a PDF</li></ul><h3 id="Other-23"><a class="docs-heading-anchor" href="#Other-23">Other</a><a class="docs-heading-anchor-permalink" href="#Other-23" title="Permalink"></a></h3><ul><li>Update to MutableArithmetics 0.3</li><li>Various improvements to the documentation</li></ul><h2 id="[v0.10.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.3)-(September-18,-2021)"><a class="docs-heading-anchor" href="#[v0.10.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.3)-(September-18,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.3">v0.10.3</a> (September 18, 2021)</a><a id="[v0.10.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.3)-(September-18,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.3)-(September-18,-2021)" title="Permalink"></a></h2><h3 id="Fixed-38"><a class="docs-heading-anchor" href="#Fixed-38">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-38" title="Permalink"></a></h3><ul><li>Fixed bug which prevented callbacks from working through a CachingOptimizer</li><li>Fixed bug in <code>Test</code> submodule</li></ul><h2 id="[v0.10.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.2)-(September-16,-2021)"><a class="docs-heading-anchor" href="#[v0.10.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.2)-(September-16,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.2">v0.10.2</a> (September 16, 2021)</a><a id="[v0.10.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.2)-(September-16,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.2)-(September-16,-2021)" title="Permalink"></a></h2><ul><li>Updated MathOptFormat to v1.0</li><li>Updated JSONSchema to v1.0</li><li>Added <code>Utilities.set_with_dimension</code></li><li>Added two-argument <code>optimize!(::AbstractOptimizer, ::ModelLike)</code></li><li>The experimental feature <code>copy_to_and_optimize!</code> has been removed</li><li><code>Det</code> bridges now support getting <code>ConstraintFunction</code> and <code>ConstraintSet</code></li><li>Various minor bug fixes identified by improved testing</li></ul><h2 id="[v0.10.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.1)-(September-8,-2021)"><a class="docs-heading-anchor" href="#[v0.10.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.1)-(September-8,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.1">v0.10.1</a> (September 8, 2021)</a><a id="[v0.10.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.1)-(September-8,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.1)-(September-8,-2021)" title="Permalink"></a></h2><ul><li>Various fixes to <code>MOI.Test</code></li></ul><h2 id="[v0.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.0)-(September-6,-2021)"><a class="docs-heading-anchor" href="#[v0.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.0)-(September-6,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.0">v0.10.0</a> (September 6, 2021)</a><a id="[v0.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.0)-(September-6,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.0)-(September-6,-2021)" title="Permalink"></a></h2><p><strong>MOI v0.10 is a significant breaking release. There are a large number of user-visible breaking changes and code refactors, as well as a substantial number of new features.</strong></p><h3 id="Breaking-in-MOI"><a class="docs-heading-anchor" href="#Breaking-in-MOI">Breaking in MOI</a><a id="Breaking-in-MOI-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-MOI" title="Permalink"></a></h3><ul><li><code>SingleVariable</code> has been removed; use <code>VariableIndex</code> instead</li><li><code>SingleVariableConstraintNameError</code> has been renamed to <code>VariableIndexConstraintNameError</code></li><li><code>SettingSingleVariableFunctionNotAllowed</code> has been renamed to <code>SettingVariableIndexFunctionNotAllowed</code></li><li><code>VariableIndex</code> constraints should not support <code>ConstraintName</code></li><li><code>VariableIndex</code> constraints should not support <code>ConstraintBasisStatus</code>; implement <code>VariableBasisStatus</code> instead</li><li><code>ListOfConstraints</code> has been renamed to <code>ListOfConstraintTypesPresent</code></li><li><code>ListOfConstraintTypesPresent</code> should now return <code>Tuple{Type,Type}</code> instead of <code>Tuple{DataType,DataType}</code></li><li><code>SolveTime</code> has been renamed to <code>SolveTimeSec</code></li><li><code>IndicatorSet</code> has been renamed to <code>Indicator</code></li><li><code>RawParameter</code> has been renamed to <code>RawOptimizerAttribute</code> and now takes <code>String</code> instead of <code>Any</code> as the only argument</li><li>The <code>.N</code> field in result attributes has been renamed to <code>.result_index</code></li><li>The <code>.variable_index</code> field in <code>ScalarAffineTerm</code> has been renamed to <code>.variable</code></li><li>The <code>.variable_index_1</code> field in <code>ScalarQuadraticTerm</code> has been renamed to <code>.variable_1</code></li><li>The <code>.variable_index_2</code> field in <code>ScalarQuadraticTerm</code> has been renamed to <code>.variable_2</code></li><li>The order of <code>affine_terms</code> and <code>quadratic_terms</code> in <code>ScalarQuadraticFunction</code> and <code>VectorQuadraticFunction</code> have been reversed. Both functions now accept quadratic, affine, and constant terms in that order.</li><li>The <code>index_value</code> function has been removed. Use <code>.value</code> instead.</li><li><code>isapprox</code> has been removed for <code>SOS1</code> and <code>SOS2</code>.</li><li>The <code>dimension</code> argument to <code>Complements(dimension::Int)</code> should now be the length of the corresponding function, instead of half the length. An <code>ArgumentError</code> is thrown if <code>dimension</code> is not even.</li><li><code>copy_to</code> no longer takes keyword arguments:<ul><li><code>copy_names</code>: now copy names if they are supported by the destination solver</li><li><code>filter_constraints</code>: use <code>Utilities.ModelFilter</code> instead</li><li><code>warn_attributes</code>: never warn about optimizer attributes</li></ul></li></ul><h3 id="Breaking-in-Bridges"><a class="docs-heading-anchor" href="#Breaking-in-Bridges">Breaking in <code>Bridges</code></a><a id="Breaking-in-Bridges-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-Bridges" title="Permalink"></a></h3><ul><li><code>Constraint.RSOCBridge</code> has been renamed to <code>Constraint.RSOCtoSOCBridge</code></li><li><code>Constraint.SOCRBridge</code> has been renamed to <code>Constraint.SOCtoRSOCBridge</code></li><li>Bridges now return vectors that can be modified by the user. Previously, some bridges returned views instead of copies.</li><li><code>Bridges.IndexInVector</code> has been unified into a single type. Previously, there was a different type for each submodule within <code>Bridges</code></li><li>The signature of indicator bridges has been fixed. Use <code>MOI.Bridges.Constraint.IndicatortoSOS1{Float64}(model)</code>.</li></ul><h3 id="Breaking-in-FileFormats"><a class="docs-heading-anchor" href="#Breaking-in-FileFormats">Breaking in <code>FileFormats</code></a><a id="Breaking-in-FileFormats-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-FileFormats" title="Permalink"></a></h3><ul><li><code>FileFormats.MOF.Model</code> no longer accepts <code>validate</code> argument. Use the JSONSchema package to validate the MOF file. See the documentation for more information.</li></ul><h3 id="Breaking-in-Utilities"><a class="docs-heading-anchor" href="#Breaking-in-Utilities">Breaking in <code>Utilities</code></a><a id="Breaking-in-Utilities-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-Utilities" title="Permalink"></a></h3><ul><li>The datastructure of <code>Utilities.Model</code> (and models created with <code>Utilities.@model</code>) has been significantly refactored in a breaking way. This includes the way that objective functions and variable-related information is stored.</li><li><code>Utilities.supports_default_copy</code> has been renamed to <code>supports_incremental_interface</code></li><li><code>Utilities.automatic_copy_to</code> has been renamed to <code>Utilities.default_copy_to</code></li><li>The allocate-load API has been removed</li><li><code>CachingOptimizer</code>s are now initialized as <code>EMPTY_OPTIMIZER</code> instead of <code>ATTACHED_OPTIMIZER</code>. If your code relies on the optimizer being attached, call <code>MOIU.attach_optimizer(model)</code> after creation.</li><li>The field names of <code>Utilities.IndexMap</code> have been renamed to <code>var_map</code> and <code>con_map</code>. Accessing these fields directly is considered a private detail that may change. Use the public <code>getindex</code> and <code>setindex!</code> API instead.</li><li>The size argument to <code>Utilities.CleverDicts.CleverDict(::Integer)</code> has been removed.</li><li>The size argument to <code>Utilities.IndexMap(::Integer)</code> has been removed.</li><li><code>Utilities.DoubleDicts</code> have been significantly refactored. Consult the source code for details.</li><li><code>Utilities.test_models_equal</code> has been moved to <code>MOI.Test</code></li></ul><h3 id="Breaking-in-Test"><a class="docs-heading-anchor" href="#Breaking-in-Test">Breaking in <code>Test</code></a><a id="Breaking-in-Test-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking-in-Test" title="Permalink"></a></h3><ul><li><code>MOI.Test</code> has been renamed to <code>MOI.DeprecatedTest</code></li><li>An entirely new <code>MOI.Test</code> submodule has been written. See the documentation for details. The new <code>MOI.Test</code> submodule may find many bugs in the implementations of existing solvers that were previously untested.</li></ul><h3 id="Other-changes:"><a class="docs-heading-anchor" href="#Other-changes:">Other changes:</a><a id="Other-changes:-1"></a><a class="docs-heading-anchor-permalink" href="#Other-changes:" title="Permalink"></a></h3><ul><li><code>attribute_value_type</code> has been added</li><li><code>copy_to_and_optimize!</code> has been added</li><li><code>VariableBasisStatus</code> has been added</li><li><code>print(model)</code> now prints a human-readable description of the model</li><li>Various improvements to the <code>FileFormats</code> submodule<ul><li><code>FileFormats.CBF</code> was refactored and received bugfixes</li><li>Support for MathOptFormat v0.6 was added in <code>FileFormats.MOF</code></li><li><code>FileFormats.MPS</code> has had bugfixes and support for more features such as <code>OBJSENSE</code> and objective constants.</li><li><code>FileFormats.NL</code> has been added to support nonlinear files</li></ul></li><li>Improved type inference throughout to reduce latency</li></ul><h3 id="Updating"><a class="docs-heading-anchor" href="#Updating">Updating</a><a id="Updating-1"></a><a class="docs-heading-anchor-permalink" href="#Updating" title="Permalink"></a></h3><p>A helpful script when updating is:</p><pre><code class="language-julia hljs">for (root, dirs, files) in walkdir(&quot;.&quot;)
+    for file in files
+        if !endswith(file, &quot;.jl&quot;)
+            continue
+        end
+        path = joinpath(root, file)
+        s = read(path, String)
+        for pair in [
+            &quot;.variable_index&quot; =&gt; &quot;.variable&quot;,
+            &quot;RawParameter&quot; =&gt; &quot;RawOptimizerAttribute&quot;,
+            &quot;ListOfConstraints&quot; =&gt; &quot;ListOfConstraintTypesPresent&quot;,
+            &quot;TestConfig&quot; =&gt; &quot;Config&quot;,
+            &quot;attr.N&quot; =&gt; &quot;attr.result_index&quot;,
+            &quot;SolveTime&quot; =&gt; &quot;SolveTimeSec&quot;,
+            &quot;DataType&quot; =&gt; &quot;Type&quot;,
+            &quot;Utilities.supports_default_copy_to&quot; =&gt;
+                &quot;supports_incremental_interface&quot;,
+            &quot;SingleVariableConstraintNameError&quot; =&gt;
+                &quot;VariableIndexConstraintNameError&quot;,
+            &quot;SettingSingleVariableFunctionNotAllowed&quot; =&gt;
+                &quot;SettingVariableIndexFunctionNotAllowed&quot;,
+            &quot;automatic_copy_to&quot; =&gt; &quot;default_copy_to&quot;,
+        ]
+            s = replace(s, pair)
+        end
+        write(path, s)
+    end
+end</code></pre><h2 id="[v0.9.22](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.22)-(May-22,-2021)"><a class="docs-heading-anchor" href="#[v0.9.22](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.22)-(May-22,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.22">v0.9.22</a> (May 22, 2021)</a><a id="[v0.9.22](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.22)-(May-22,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.22](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.22)-(May-22,-2021)" title="Permalink"></a></h2><p>This release contains backports from the ongoing development of the v0.10 release.</p><ul><li>Improved type inference in <code>Utilities</code>, <code>Bridges</code> and <code>FileFormats</code> submodules to reduce latency.</li><li>Improved performance of <code>Utilities.is_canonical</code>.</li><li>Fixed <code>Utilities.pass_nonvariable_constraints</code> with bridged variables.</li><li>Fixed performance regression of <code>Utilities.Model</code>.</li><li>Fixed ordering of objective setting in parser.</li></ul><h2 id="[v0.9.21](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.21)-(April-23,-2021)"><a class="docs-heading-anchor" href="#[v0.9.21](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.21)-(April-23,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.21">v0.9.21</a> (April 23, 2021)</a><a id="[v0.9.21](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.21)-(April-23,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.21](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.21)-(April-23,-2021)" title="Permalink"></a></h2><ul><li>Added <code>supports_shift_constant</code>.</li><li>Improve performance of bridging quadratic constraints.</li><li>Add precompilation statements.</li><li>Large improvements to the documentation.</li><li>Fix a variety of inference issues, benefiting precompilation and reducing initial latency.</li><li><code>RawParameter</code>s are now ignored when resetting a <code>CachingOptimizer</code>. Previously, changing the underlying optimizer after <code>RawParameter</code>s were set would throw an error.</li><li><code>Utilities.AbstractModel</code> is being refactored. This may break users interacting with private fields of a model generated using <code>@model</code>.</li></ul><h2 id="[v0.9.20](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.20)-(February-20,-2021)"><a class="docs-heading-anchor" href="#[v0.9.20](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.20)-(February-20,-2021)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.20">v0.9.20</a> (February 20, 2021)</a><a id="[v0.9.20](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.20)-(February-20,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.20](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.20)-(February-20,-2021)" title="Permalink"></a></h2><ul><li>Improved performance of <code>Utilities.ScalarFunctionIterator</code></li><li>Added support for <code>compute_conflict</code> to MOI layers</li><li>Added test with zero off-diagonal quadratic term in objective</li><li>Fixed double deletion of nested bridged <code>SingleVariable</code>/<code>VectorOfVariables</code> constraints</li><li>Fixed modification of un-set objective</li><li>Fixed function modification with duplicate terms</li><li>Made unit tests abort without failing if the problem class is not supported</li><li>Formatted code with JuliaFormatter</li><li>Clarified <code>BasisStatusCode</code>&#39;s docstring</li></ul><h2 id="[v0.9.19](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.19)-(December-1,-2020)"><a class="docs-heading-anchor" href="#[v0.9.19](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.19)-(December-1,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.19">v0.9.19</a> (December 1, 2020)</a><a id="[v0.9.19](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.19)-(December-1,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.19](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.19)-(December-1,-2020)" title="Permalink"></a></h2><ul><li>Added <code>CallbackNodeStatus</code> attribute</li><li>Added bridge from <code>GreaterThan</code> or <code>LessThan</code> to <code>Interval</code></li><li>Added tests for infeasibility certificates and double optimize</li><li>Fixed support for Julia v1.6</li><li>Re-organized MOI docs and added documentation for adding a test</li></ul><h2 id="[v0.9.18](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.18)-(November-3,-2020)"><a class="docs-heading-anchor" href="#[v0.9.18](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.18)-(November-3,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.18">v0.9.18</a> (November 3, 2020)</a><a id="[v0.9.18](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.18)-(November-3,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.18](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.18)-(November-3,-2020)" title="Permalink"></a></h2><ul><li>Various improvements for working with complex numbers</li><li>Added <code>GeoMeantoRelEntrBridge</code> to bridge a <code>GeometricMeanCone</code> constraint to a relative entropy constraint</li></ul><h2 id="[v0.9.17](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.17)-(September-21,-2020)"><a class="docs-heading-anchor" href="#[v0.9.17](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.17)-(September-21,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.17">v0.9.17</a> (September 21, 2020)</a><a id="[v0.9.17](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.17)-(September-21,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.17](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.17)-(September-21,-2020)" title="Permalink"></a></h2><ul><li>Fixed <code>CleverDict</code> with variable of negative index value</li><li>Implement <code>supports_add_constrained_variable</code> for <code>MockOptimizer</code></li></ul><h2 id="[v0.9.16](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.16)-(September-17,-2020)"><a class="docs-heading-anchor" href="#[v0.9.16](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.16)-(September-17,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.16">v0.9.16</a> (September 17, 2020)</a><a id="[v0.9.16](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.16)-(September-17,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.16](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.16)-(September-17,-2020)" title="Permalink"></a></h2><ul><li>Various fixes:<ul><li>32-bit support</li><li><code>CleverDict</code> with abstract value type</li><li>Checks in test suite</li></ul></li></ul><h2 id="[v0.9.15](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.15)-(September-14,-2020)"><a class="docs-heading-anchor" href="#[v0.9.15](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.15)-(September-14,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.15">v0.9.15</a> (September 14, 2020)</a><a id="[v0.9.15](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.15)-(September-14,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.15](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.15)-(September-14,-2020)" title="Permalink"></a></h2><ul><li>Bridges improvements:<ul><li>(R)SOCtoNonConvexQuad bridge</li><li>ZeroOne bridge</li><li>Use <code>supports_add_constrained_variable</code> in <code>LazyBridgeOptimizer</code></li><li>Exposed <code>VariableBridgeCost</code> and <code>ConstraintBridgeCost</code> attributes</li><li>Prioritize constraining variables on creation according to these costs</li><li>Refactor bridge debugging</li></ul></li><li>Large performance improvements across all submodules</li><li>Lots of documentation improvements</li><li>FileFormats improvements:<ul><li>Update MathOptFormat to v0.5</li><li>Fix supported objectives in <code>FileFormats</code></li></ul></li><li>Testing improvements:<ul><li>Add name option for <code>basic_constraint_test</code></li></ul></li><li>Bug fixes and missing methods<ul><li>Add <code>length</code> for iterators</li><li>Fix bug with duplicate terms</li><li>Fix order of <code>LinearOfConstraintIndices</code></li></ul></li></ul><h2 id="[v0.9.14](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.14)-(May-30,-2020)"><a class="docs-heading-anchor" href="#[v0.9.14](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.14)-(May-30,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.14">v0.9.14</a> (May 30, 2020)</a><a id="[v0.9.14](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.14)-(May-30,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.14](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.14)-(May-30,-2020)" title="Permalink"></a></h2><ul><li>Add a solver-independent interface for accessing the set of conflicting constraints an Irreducible Inconsistent Subsystem (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1056">#1056</a>).</li><li>Bump JSONSchema dependency from v0.2 to v0.3 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1090">#1090</a>).</li><li>Documentation improvements:<ul><li>Fix typos (#1054, #1060, #1061, #1064, #1069, #1070).</li><li>Remove the outdated recommendation for a package implementing MOI for a solver <code>XXX</code> to be called <code>MathOptInterfaceXXX</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1087">#1087</a>).</li></ul></li><li>Utilities improvements:<ul><li>Fix <code>is_canonical</code> for quadratic functions (#1081, #1089).</li><li>Implement <code>add_constrained_variable[s]</code> for <code>CachingOptimizer</code> so that it is added as constrained variables to the underlying optimizer (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1084">#1084</a>).</li><li>Add support for custom objective functions for <code>UniversalFallback</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1086">#1086</a>).</li><li>Deterministic ordering of constraints in <code>UniversalFallback</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1088">#1088</a>).</li></ul></li><li>Testing improvements:<ul><li>Add <code>NormOneCone</code>/<code>NormInfinityCone</code> tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1045">#1045</a>).</li></ul></li><li>Bridges improvements:<ul><li>Add bridges from <code>Semiinteger</code> and <code>Semicontinuous</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1059">#1059</a>).</li><li>Implement getting <code>ConstraintSet</code> for <code>Variable.FlipSignBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1066">#1066</a>).</li><li>Fix setting <code>ConstraintFunction</code> for <code>Constraint.ScalarizeBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1093">#1093</a>).</li><li>Fix <code>NormOne</code>/<code>NormInf</code> bridges with nonzero constants (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1045">#1045</a>).</li><li>Fix StackOverflow in <code>debug</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1063">#1063</a>).</li></ul></li><li>FileFormats improvements:<ul><li>[SDPA] Implement the extension for integer variables (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1079">#1079</a>).</li><li>[SDPA] Ignore comments after <code>m</code> and <code>nblocks</code> and detect <code>dat-s</code> extension (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1077">#1077</a>).</li><li>[SDPA] No scaling of off-diagonal coefficient (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1076">#1076</a>).</li><li>[SDPA] Add missing negation of constant (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1075">#1075</a>).</li></ul></li></ul><h2 id="[v0.9.13](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.13)-(March-24,-2020)"><a class="docs-heading-anchor" href="#[v0.9.13](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.13)-(March-24,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.13">v0.9.13</a> (March 24, 2020)</a><a id="[v0.9.13](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.13)-(March-24,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.13](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.13)-(March-24,-2020)" title="Permalink"></a></h2><ul><li>Added tests for <code>Semicontinuous</code> and <code>Semiinteger</code> variables (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1033">#1033</a>).</li><li>Added tests for using <code>ExprGraph</code>s from NLP evaluators (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1043">#1043</a>).</li><li>Update version compatibilities of dependencies (#1034, #1051, #1052).</li><li>Fixed typos in documentation (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1044">#1044</a>).</li></ul><h2 id="[v0.9.12](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.12)-(February-28,-2020)"><a class="docs-heading-anchor" href="#[v0.9.12](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.12)-(February-28,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.12">v0.9.12</a> (February 28, 2020)</a><a id="[v0.9.12](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.12)-(February-28,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.12](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.12)-(February-28,-2020)" title="Permalink"></a></h2><ul><li>Fixed writing <code>NLPBlock</code> in MathOptFormat (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1037">#1037</a>).</li><li>Fixed <code>MockOptimizer</code> for result attributes with non-one result index (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1039">#1039</a>).</li><li>Updated test template with <code>instantiate</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1032">#1032</a>).</li></ul><h2 id="[v0.9.11](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.11)-(February-21,-2020)"><a class="docs-heading-anchor" href="#[v0.9.11](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.11)-(February-21,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.11">v0.9.11</a> (February 21, 2020)</a><a id="[v0.9.11](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.11)-(February-21,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.11](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.11)-(February-21,-2020)" title="Permalink"></a></h2><ul><li>Add an option for the model created by <code>Utilities.@model</code> to be a subtype of <code>AbstractOptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1031">#1031</a>).</li><li>Described dual cone in docstrings of <code>GeoMeanCone</code> and <code>RelativeEntropyCone</code> (#1018, #1028).</li><li>Fixed typos in documentation (#1022, #1024).</li><li>Fixed warning of unsupported attribute (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1027">#1027</a>).</li><li>Added more rootdet/logdet conic tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1026">#1026</a>).</li><li>Implemented <code>ConstraintDual</code> for <code>Constraint.GeoMeanBridge</code>, <code>Constraint.RootDetBridge</code> and <code>Constraint.LogDetBridge</code> and test duals in tests with <code>GeoMeanCone</code> and <code>RootDetConeTriangle</code> and <code>LogDetConeTriangle</code> cones (#1025, #1026).</li></ul><h2 id="[v0.9.10](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.10)-(January-31,-2020)"><a class="docs-heading-anchor" href="#[v0.9.10](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.10)-(January-31,-2020)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.10">v0.9.10</a> (January 31, 2020)</a><a id="[v0.9.10](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.10)-(January-31,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.10](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.10)-(January-31,-2020)" title="Permalink"></a></h2><ul><li>Added <code>OptimizerWithAttributes</code> grouping an optimizer constructor and a list of optimizer attributes (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1008">#1008</a>).</li><li>Added <code>RelativeEntropyCone</code> with corresponding bridge into exponential cone constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/993">#993</a>).</li><li>Added <code>NormSpectralCone</code> and <code>NormNuclearCone</code> with corresponding bridges into positive semidefinite constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/976">#976</a>).</li><li>Added <code>supports_constrained_variable(s)</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1004">#1004</a>).</li><li>Added <code>dual_set_type</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1002">#1002</a>).</li><li>Added tests for vector specialized version of <code>delete</code> (#989, #1011).</li><li>Added PSD3 test (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1007">#1007</a>).</li><li>Clarified dual solution of <code>Tests.pow1v</code> and <code>Tests.pow1f</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1013">#1013</a>).</li><li>Added support for <code>EqualTo</code> and <code>Zero</code> in <code>Bridges.Constraint.SplitIntervalBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1005">#1005</a>).</li><li>Fixed <code>Utilities.vectorize</code> for empty vector (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1003">#1003</a>).</li><li>Fixed free variables in LP writer (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1006">#1006</a>).</li></ul><h2 id="[v0.9.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.9)-(December-29,-2019)"><a class="docs-heading-anchor" href="#[v0.9.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.9)-(December-29,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.9">v0.9.9</a> (December 29, 2019)</a><a id="[v0.9.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.9)-(December-29,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.9)-(December-29,-2019)" title="Permalink"></a></h2><ul><li>Incorporated MathOptFormat.jl as the FileFormats submodule. FileFormats provides readers and writers for a number of standard file formats and MOF, a file format specialized for MOI (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/969">#969</a>).</li><li>Improved performance of deletion of vector of variables in <code>MOI.Utilities.Model</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/983">#983</a>).</li><li>Updated to MutableArithmetics v0.2 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/981">#981</a>).</li><li>Added <code>MutableArithmetics.promote_operation</code> allocation tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/975">#975</a>).</li><li>Fixed inference issue on Julia v1.1 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/982">#982</a>).</li></ul><h2 id="[v0.9.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.8)-(December-19,-2019)"><a class="docs-heading-anchor" href="#[v0.9.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.8)-(December-19,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.8">v0.9.8</a> (December 19, 2019)</a><a id="[v0.9.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.8)-(December-19,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.8)-(December-19,-2019)" title="Permalink"></a></h2><ul><li>Implemented MutableArithmetics API (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/924">#924</a>).</li><li>Fixed callbacks with <code>CachingOptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/959">#959</a>).</li><li>Fixed <code>MOI.dimension</code> for <code>MOI.Complements</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/948">#948</a>).</li><li>Added fallback for <code>add_variables</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/972">#972</a>).</li><li>Added <code>is_diagonal_vectorized_index</code> utility (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/965">#965</a>).</li><li>Improved linear constraints display in manual (#963, #964).</li><li>Bridges improvements:<ul><li>Added <code>IndicatorSet</code> to <code>SOS1</code> bridge (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/877">#877</a>).</li><li>Added support for starting values for <code>Variable.VectorizeBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/944">#944</a>).</li><li>Fixed <code>MOI.add_constraints</code> with non-bridged variable constraint on bridged variable (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/951">#951</a>).</li><li>Fixed corner cases and docstring of <code>GeoMeanBridge</code> (#961, #962, #966).</li><li>Fixed choice between variable or constraint bridges for constrained variables (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/973">#973</a>).</li><li>Improve performance of bridge shortest path (#945, #946, #956).</li><li>Added docstring for <code>test_delete_bridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/954">#954</a>).</li><li>Added Variable bridge tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/952">#952</a>).</li></ul></li></ul><h2 id="[v0.9.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.7)-(October-30,-2019)"><a class="docs-heading-anchor" href="#[v0.9.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.7)-(October-30,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.7">v0.9.7</a> (October 30, 2019)</a><a id="[v0.9.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.7)-(October-30,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.7)-(October-30,-2019)" title="Permalink"></a></h2><ul><li>Implemented <code>_result_index_field</code> for <code>NLPBlockDual</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/934">#934</a>).</li><li>Fixed copy of model with starting values for vector constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/941">#941</a>).</li><li>Bridges improvements:<ul><li>Improved performance of <code>add_bridge</code> and added <code>has_bridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/935">#935</a>).</li><li>Added <code>AbstractSetMapBridge</code> for bridges between sets <code>S1</code>, <code>S2</code> such that there is a linear map <code>A</code> such that <code>A*S1 = S2</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/933">#933</a>).</li><li>Added support for starting values for <code>FlipSignBridge</code>, <code>VectorizeBridge</code>, <code>ScalarizeBridge</code>, <code>SlackBridge</code>, <code>SplitIntervalBridge</code>, <code>RSOCBridge</code>, <code>SOCRBridge</code> <code>NormInfinityBridge</code>, <code>SOCtoPSDBridge</code> and <code>RSOCtoPSDBridge</code> (#933, #936, #937, #938, #939).</li></ul></li></ul><h2 id="[v0.9.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.6)-(October-25,-2019)"><a class="docs-heading-anchor" href="#[v0.9.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.6)-(October-25,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.6">v0.9.6</a> (October 25, 2019)</a><a id="[v0.9.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.6)-(October-25,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.6)-(October-25,-2019)" title="Permalink"></a></h2><ul><li>Added complementarity constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/913">#913</a>).</li><li>Allowed <code>ModelLike</code> objects as value of attributes (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/928">#928</a>).</li><li>Testing improvements:<ul><li>Added <code>dual_objective_value</code> option to <code>MOI.Test.TestConfig</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/922">#922</a>).</li><li>Added <code>InvalidIndex</code> tests in <code>basic_constraint_tests</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/921">#921</a>).</li><li>Added tests for the constant term in indicator constraint (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/929">#929</a>).</li></ul></li><li>Bridges improvements:<ul><li>Added support for starting values for <code>Functionize</code> bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/923">#923</a>).</li><li>Added variable indices context to variable bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/920">#920</a>).</li><li>Fixed a typo in printing o <code>debug_supports</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/927">#927</a>).</li></ul></li></ul><h2 id="[v0.9.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.5)-(October-9,-2019)"><a class="docs-heading-anchor" href="#[v0.9.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.5)-(October-9,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.5">v0.9.5</a> (October 9, 2019)</a><a id="[v0.9.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.5)-(October-9,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.5)-(October-9,-2019)" title="Permalink"></a></h2><ul><li>Clarified <code>PrimalStatus</code>/<code>DualStatus</code> to be <code>NO_SOLUTION</code> if <code>result_index</code> is out of bounds (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/912">#912</a>).</li><li>Added tolerance for checks  and use <code>ResultCount</code> + 1 for the <code>result_index</code> in <code>MOI.Test.solve_result_status</code> (#910, #917).</li><li>Use <code>0.5</code> instead of <code>2.0</code> for power in PowerCone in <code>basic_constraint_test</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/916">#916</a>).</li><li>Bridges improvements:<ul><li>Added debug utilities for unsupported variable/constraint/objective (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/861">#861</a>).</li><li>Fixed deletion of variables in bridged <code>VectorOfVariables</code> constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/909">#909</a>).</li><li>Fixed <code>result_index</code> with objective bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/911">#911</a>).</li></ul></li></ul><h2 id="[v0.9.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.4)-(October-2,-2019)"><a class="docs-heading-anchor" href="#[v0.9.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.4)-(October-2,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.4">v0.9.4</a> (October 2, 2019)</a><a id="[v0.9.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.4)-(October-2,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.4)-(October-2,-2019)" title="Permalink"></a></h2><ul><li>Added solver-independent MIP callbacks (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/782">#782</a>).</li><li>Implements submit for <code>Utilities.CachingOptimizer</code> and <code>Bridges.AbstractBridgeOptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/906">#906</a>).</li><li>Added tests for result count of solution attributes (#901, #904).</li><li>Added <code>NumberOfThreads</code> attribute (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/892">#892</a>).</li><li>Added <code>Utilities.get_bounds</code> to get the bounds on a variable (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/890">#890</a>).</li><li>Added a note on duplicate coefficients in documentation (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/581">#581</a>).</li><li>Added result index in <code>ConstraintBasisStatus</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/898">#898</a>).</li><li>Added extension dictionary to <code>Utilities.Model</code> (#884, #895).</li><li>Fixed deletion of constrained variables for CachingOptimizer (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/905">#905</a>).</li><li>Implemented <code>Utilities.shift_constraint</code> for <code>Test.UnknownScalarSet</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/896">#896</a>).</li><li>Bridges improvements:<ul><li>Added Variable.RSOCtoSOCBridge (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/907">#907</a>).</li><li>Implemented <code>MOI.get</code> for <code>ConstraintFunction</code>/<code>ConstraintSet</code> for <code>Bridges.Constraint.SquareBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/899">#899</a>).</li></ul></li></ul><h2 id="[v0.9.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.3)-(September-20,-2019)"><a class="docs-heading-anchor" href="#[v0.9.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.3)-(September-20,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.3">v0.9.3</a> (September 20, 2019)</a><a id="[v0.9.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.3)-(September-20,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.3)-(September-20,-2019)" title="Permalink"></a></h2><ul><li>Fixed ambiguity detected in Julia v1.3 (#891, #893).</li><li>Fixed missing sets from <code>ListOfSupportedConstraints</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/880">#880</a>).</li><li>Fixed copy of <code>VectorOfVariables</code> constraints with duplicate indices (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/886">#886</a>).</li><li>Added extension dictionary to MOIU.Model (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/884">#884</a>).</li><li>Implemented <code>MOI.get</code> for function and set for <code>GeoMeanBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/888">#888</a>).</li><li>Updated documentation for SingleVariable indices and bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/885">#885</a>).</li><li>Testing improvements:<ul><li>Added more comprehensive tests for names (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/882">#882</a>).</li><li>Added tests for <code>SingleVariable</code> duals (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/883">#883</a>).</li><li>Added tests for <code>DualExponentialCone</code> and <code>DualPowerCone</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/873">#873</a>).</li></ul></li><li>Improvements for arbitrary coefficient type:<ul><li>Fixed <code>==</code> for sets with mutable fields (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/887">#887</a>).</li><li>Removed some <code>Float64</code> assumptions in bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/878">#878</a>).</li><li>Automatic selection of <code>Constraint.[Scalar|Vector]FunctionizeBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/889">#889</a>).</li></ul></li></ul><h2 id="[v0.9.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.2)-(September-5,-2019)"><a class="docs-heading-anchor" href="#[v0.9.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.2)-(September-5,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.2">v0.9.2</a> (September 5, 2019)</a><a id="[v0.9.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.2)-(September-5,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.2)-(September-5,-2019)" title="Permalink"></a></h2><ul><li>Implemented model printing for <code>MOI.ModelLike</code> and specialized it for models defined in MOI (864).</li><li>Generalized <code>contlinear</code> tests for arbitrary coefficient type (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/855">#855</a>).</li><li>Fixed <code>supports_constraint</code> for <code>Semiinteger</code> and <code>Semicontinuous</code> and <code>supports</code> for <code>ObjectiveFunction</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/859">#859</a>).</li><li>Fixed Allocate-Load copy for single variable constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/856">#856</a>).</li><li>Bridges improvements:<ul><li>Add objective bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/789">#789</a>).</li><li>Fixed <code>Variable.RSOCtoPSDBridge</code> for dimension 2 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/869">#869</a>).</li><li>Added <code>Variable.SOCtoRSOCBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/865">#865</a>).</li><li>Added <code>Constraint.SOCRBridge</code> and disable <code>MOI.Bridges.Constraint.SOCtoPSDBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/751">#751</a>).</li><li>Fixed <code>added_constraint_types</code> for <code>Contraint.LogDetBridge</code> and <code>Constraint.RootDetBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/870">#870</a>).</li></ul></li></ul><h2 id="[v0.9.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.1)-(August-22,-2019)"><a class="docs-heading-anchor" href="#[v0.9.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.1)-(August-22,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.1">v0.9.1</a> (August 22, 2019)</a><a id="[v0.9.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.1)-(August-22,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.1)-(August-22,-2019)" title="Permalink"></a></h2><ul><li>Fix support for Julia v1.2 (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/834">#834</a>).</li><li>L<em>1 and L</em>∞ norm epigraph cones and corresponding bridges to LP were added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/818">#818</a>).</li><li>Added tests to <code>MOI.Test.nametest</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/833">#833</a>).</li><li>Fix <code>MOI.Test.soc3test</code> for solvers not supporting infeasibility certificates (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/839">#839</a>).</li><li>Implements <code>operate</code> for operators <code>*</code> and <code>/</code> between vector function and constant (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/837">#837</a>).</li><li>Implements <code>show</code> for <code>MOI.Utilities.IndexMap</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/847">#847</a>).</li><li>Fix corner cases for mapping of variables in <code>MOI.Utilities.CachingOptimizer</code> and substitution of variables in <code>MOI.Bridges.AbstractBridgeOptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/848">#848</a>).</li><li>Fix transformation of constant terms for <code>MOI.Bridges.Constraint.SOCtoPSDBridge</code> and <code>MOI.Bridges.Constraint.RSOCtoPSDBridge</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/840">#840</a>).</li></ul><h2 id="[v0.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.0)-(August-13,-2019)"><a class="docs-heading-anchor" href="#[v0.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.0)-(August-13,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.0">v0.9.0</a> (August 13, 2019)</a><a id="[v0.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.0)-(August-13,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.0)-(August-13,-2019)" title="Permalink"></a></h2><ul><li>Support for Julia v0.6 and v0.7 was dropped (#714, #717).</li><li>A <code>MOI.Utilities.Model</code> implementation of <code>ModelLike</code>, this should replace most use cases of <code>MOI.Utilities.@model</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/781">#781</a>).</li><li><code>add_constrained_variable</code> and <code>add_constrained_variables</code> were added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/759">#759</a>).</li><li>Support for indicator constraints was added (#709, #712).</li><li><code>DualObjectiveValue</code> attribute was added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/473">#473</a>).</li><li><code>RawParameter</code> attribute was added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/733">#733</a>).</li><li>A <code>dual_set</code> function was added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/804">#804</a>).</li><li>A <code>Benchmarks</code> submodule was added to facilitate solver benchmarking (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/769">#769</a>).</li><li>A <code>submit</code> function was added, this may for instance allow the user to submit solutions or cuts to the solver from a callback (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/775">#775</a>).</li><li>The field of <code>ObjectiveValue</code> was renamed to <code>result_index</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/729">#729</a>).</li><li>The <code>_constant</code> and <code>Utilities.getconstant</code> function were renamed to <code>constant</code></li><li><code>REDUCTION_CERTIFICATE</code> result status was added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/734">#734</a>).</li><li>Abstract matrix sets were added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/731">#731</a>).</li><li>Testing improvements:<ul><li>The testing guideline was updated (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/728">#728</a>).</li><li>Quadratic tests were added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/697">#697</a>).</li><li>Unit tests for <code>RawStatusString</code>, <code>SolveTime</code>, <code>Silent</code> and <code>SolverName</code> were added (#726, #741).</li><li>A rotated second-order cone test was added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/759">#759</a>).</li><li>A power cone test was added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/768">#768</a>).</li><li>Tests for <code>ZeroOne</code> variables with variable bounds were added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/772">#772</a>).</li><li>An unbounded test was added (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/773">#773</a>).</li><li>Existing tests had a few updates (#702, #703, #763).</li></ul></li><li>Documentation improvements:<ul><li>Added a section on <code>CachingOptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/777">#777</a>).</li><li>Added a section on <code>UniversalFallback</code>, <code>Model</code> and <code>@model</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/762">#762</a>).</li><li>Transition the knapsack example to a doctest with MockOptimizer (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/786">#786</a>).</li></ul></li><li>Utilities improvements:<ul><li>A <code>CleverDict</code> utility was added for a vector that automatically transform into a dictionary once a first index is removed (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/767">#767</a>).</li><li>The <code>Utilities.constant</code> function was renamed to <code>Utilities.constant_vector</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/740">#740</a>).</li><li>Implement optimizer attributes for CachingOptimizer (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/745">#745</a>).</li><li>Rename <code>Utilities.add_scalar_constraint</code> to <code>Utilities.normalize_and_add_constraint</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/801">#801</a>).</li><li><code>operate</code> with <code>vcat</code>, <code>SingleVariable</code> and <code>VectorOfVariables</code> now returns a <code>VectorOfVariables</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/616">#616</a>).</li><li>Fix a type piracy of <code>operate</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/784">#784</a>).</li><li>The <code>load_constraint</code> fallback signature was fixed (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/760">#760</a>).</li><li>The <code>set_dot</code> function was extended to work with sparse arrays (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/805">#805</a>).</li></ul></li><li>Bridges improvements:<ul><li>The bridges no longer store the constraint function and set before it is bridged, the bridges now have to implement <code>ConstraintFunction</code> and <code>ConstraintSet</code> if the user wants to recover them. As a consequence, the <code>@bridge</code> macro was removed (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/722">#722</a>).</li><li>Bridge are now instantiated with a <code>bridge_constraint</code> function instead of using a constructor (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/730">#730</a>).</li><li>Fix constraint attributes for bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/699">#699</a>).</li><li>Constraint bridges were moved to the <code>Bridges/Constraint</code> submodule so they should now inherit from <code>MOI.Bridges.Constraint.Abstract</code> and should implement <code>MOI.Bridges.Constraint.concrete_bridge_type</code> instead of <code>MOI.Bridges.concrete_bridge_type</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/756">#756</a>).</li><li>Variable bridges were added in (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/759">#759</a>).</li><li>Various improvements (#746, #747).</li></ul></li></ul><h2 id="[v0.8.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.4)-(March-13,-2019)"><a class="docs-heading-anchor" href="#[v0.8.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.4)-(March-13,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.4">v0.8.4</a> (March 13, 2019)</a><a id="[v0.8.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.4)-(March-13,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.8.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.4)-(March-13,-2019)" title="Permalink"></a></h2><ul><li>Performance improvement in <code>default_copy_to</code> and bridge optimizer (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/696">#696</a>).</li><li>Add <code>Silent</code> and implement setting optimizer attributes in caching and mock optimizers (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/695">#695</a>).</li><li>Add <code>Functionize</code> bridges (SingleVariable and VectorOfVariables) (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/659">#659</a>).</li><li>Minor typo fixes (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/694">#694</a>).</li></ul><h2 id="[v0.8.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.3)-(March-6,-2019)"><a class="docs-heading-anchor" href="#[v0.8.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.3)-(March-6,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.3">v0.8.3</a> (March 6, 2019)</a><a id="[v0.8.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.3)-(March-6,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.8.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.3)-(March-6,-2019)" title="Permalink"></a></h2><ul><li>Use zero constant in scalar constraint function of <code>MOI.Test.copytest</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/691">#691</a>).</li><li>Fix variable deletion with <code>SingleVariable</code> objective function (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/690">#690</a>).</li><li>Fix <code>LazyBridgeOptimizer</code> with bridges that add no constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/689">#689</a>).</li><li>Error message improvements (#673, #685, #686, #688).</li><li>Documentation improvements (#682, #683, #687).</li><li>Basis status:<ul><li>Remove <code>VariableBasisStatus</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/679">#679</a>).</li><li>Test <code>ConstraintBasisStatus</code> and implement it in bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/678">#678</a>).</li></ul></li><li>Fix inference of <code>NumberOfVariables</code> and <code>NumberOfConstraints</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/677">#677</a>).</li><li>Implement division between a quadratic function and a number (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/675">#675</a>).</li></ul><h2 id="[v0.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.2)-(February-7,-2019)"><a class="docs-heading-anchor" href="#[v0.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.2)-(February-7,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.2">v0.8.2</a> (February 7, 2019)</a><a id="[v0.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.2)-(February-7,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.2)-(February-7,-2019)" title="Permalink"></a></h2><ul><li>Add <code>RawStatusString</code> attribute (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/629">#629</a>).</li><li>Do not set names to the optimizer but only to the cache in <code>CachingOptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/638">#638</a>).</li><li>Make scalar MOI functions act as scalars in broadcast (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/646">#646</a>).</li><li>Add function utilities:<ul><li>Implement <code>Base.zero</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/634">#634</a>), <code>Base.iszero</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/643">#643</a>), add missing arithmetic operations (#644, #645) and fix division (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/648">#648</a>).</li><li>Add a <code>vectorize</code> function that turns a vector of <code>ScalarAffineFunction</code> into a <code>VectorAffineFunction</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/642">#642</a>).</li></ul></li><li>Improve support for starting values:<ul><li>Show a warning in copy when starting values are not supported instead of throwing an error (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/630">#630</a>).</li><li>Fix <code>UniversalFallback</code> for getting an variable or constraint attribute set to no indices (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/623">#623</a>).</li><li>Add a test in <code>contlineartest</code> with partially set <code>VariablePrimalStart</code>.</li></ul></li><li>Bridges improvements:<ul><li>Fix <code>StackOverFlow</code> in <code>LazyBridgeOptimizer</code> when there is a cycle in the graph of bridges.</li><li>Add <code>Slack</code> bridges (#610, #650).</li><li>Add <code>FlipSign</code> bridges (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/658">#658</a>).</li></ul></li><li>Add tests with duplicate coefficients in <code>ScalarAffineFunction</code> and <code>VectorAffineFunction</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/639">#639</a>).</li><li>Use tolerance to compare <code>VariablePrimal</code> in <code>rotatedsoc1</code> test (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/632">#632</a>).</li><li>Use a zero constant in <code>ScalarAffineFunction</code> of constraints in <code>psdt2</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/622">#622</a>).</li></ul><h2 id="[v0.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.1)-(January-7,-2019)"><a class="docs-heading-anchor" href="#[v0.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.1)-(January-7,-2019)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.1">v0.8.1</a> (January 7, 2019)</a><a id="[v0.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.1)-(January-7,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.1)-(January-7,-2019)" title="Permalink"></a></h2><ul><li>Adding an NLP objective now overrides any objective set using the <code>ObjectiveFunction</code> attribute (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/619">#619</a>).</li><li>Rename <code>fullbridgeoptimizer</code> into <code>full_bridge_optimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/621">#621</a>).</li><li>Allow custom constraint types with <code>full_bridge_optimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/617">#617</a>).</li><li>Add <code>Vectorize</code> bridge which transforms scalar linear constraints into vector linear constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/615">#615</a>).</li></ul><h2 id="[v0.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.0)-(December-18,-2018)"><a class="docs-heading-anchor" href="#[v0.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.0)-(December-18,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.0">v0.8.0</a> (December 18, 2018)</a><a id="[v0.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.0)-(December-18,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.0)-(December-18,-2018)" title="Permalink"></a></h2><ul><li>Rename all enum values to follow the JuMP naming guidelines for constants, for example, <code>Optimal</code> becomes <code>OPTIMAL</code>, and <code>DualInfeasible</code> becomes <code>DUAL_INFEASIBLE</code>.</li><li>Rename CachingOptimizer methods for style compliance.</li><li>Add an <code>MOI.TerminationStatusCode</code> called <code>ALMOST_DUAL_INFEASIBLE</code>.</li></ul><h2 id="[v0.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.7.0)-(December-13,-2018)"><a class="docs-heading-anchor" href="#[v0.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.7.0)-(December-13,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.7.0">v0.7.0</a> (December 13, 2018)</a><a id="[v0.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.7.0)-(December-13,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.7.0)-(December-13,-2018)" title="Permalink"></a></h2><ul><li>Test that <code>MOI.TerminationStatus</code> is <code>MOI.OptimizeNotCalled</code> before <code>MOI.optimize!</code> is called.</li><li>Check <code>supports_default_copy_to</code> in tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/594">#594</a>).</li><li>Key pieces of information like optimality, infeasibility, etc., are now reported through <code>TerminationStatusCode</code>. It is typically no longer necessary to check the result statuses in addition to the termination status.</li><li>Add perspective dimension to log-det cone (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/593">#593</a>).</li></ul><h2 id="[v0.6.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.4)-(November-27,-2018)"><a class="docs-heading-anchor" href="#[v0.6.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.4)-(November-27,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.4">v0.6.4</a> (November 27, 2018)</a><a id="[v0.6.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.4)-(November-27,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.6.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.4)-(November-27,-2018)" title="Permalink"></a></h2><ul><li>Add <code>OptimizeNotCalled</code> termination status (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/577">#577</a>) and improve documentation of other statuses (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/575">#575</a>).</li><li>Add a solver naming guideline (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/578">#578</a>).</li><li>Make <code>FeasibilitySense</code> the default <code>ObjectiveSense</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/579">#579</a>).</li><li>Fix <code>Utilities.@model</code> and <code>Bridges.@bridge</code> macros for functions and sets defined outside MOI (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/582">#582</a>).</li><li>Document solver-specific attributes (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/580">#580</a>) and implement them in <code>Utilities.CachingOptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/565">#565</a>).</li></ul><h2 id="[v0.6.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.3)-(November-16,-2018)"><a class="docs-heading-anchor" href="#[v0.6.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.3)-(November-16,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.3">v0.6.3</a> (November 16, 2018)</a><a id="[v0.6.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.3)-(November-16,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.6.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.3)-(November-16,-2018)" title="Permalink"></a></h2><ul><li>Variables and constraints are now allowed to have duplicate names. An error is thrown only on lookup. This change breaks some existing tests. (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/549">#549</a>)</li><li>Attributes may now be partially set (some values could be <code>nothing</code>). (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/563">#563</a>)</li><li>Performance improvements in Utilities.Model (#549, #567, #568)</li><li>Fix bug in QuadtoSOC (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/558">#558</a>).</li><li>New <code>supports_default_copy_to</code> method that optimizers should implement to control caching behavior.</li><li>Documentation improvements.</li></ul><h2 id="[v0.6.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.2)-(October-26,-2018)"><a class="docs-heading-anchor" href="#[v0.6.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.2)-(October-26,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.2">v0.6.2</a> (October 26, 2018)</a><a id="[v0.6.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.2)-(October-26,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.6.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.2)-(October-26,-2018)" title="Permalink"></a></h2><ul><li>Improve hygiene of <code>@model</code> macro (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/544">#544</a>).</li><li>Fix bug in copy tests (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/543">#543</a>).</li><li>Fix bug in UniversalFallback attribute getter (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/540">#540</a>).</li><li>Allow all correct solutions for <code>solve_blank_obj</code> unit test (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/537">#537</a>).</li><li>Add errors for Allocate-Load and bad constraints (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/534">#534</a>).</li><li>[performance] Add specialized implementation of <code>hash</code> for <code>VariableIndex</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/533">#533</a>).</li><li>[performance] Construct the name to object dictionaries lazily in model (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/535">#535</a>).</li><li>Add the <code>QuadtoSOC</code> bridge which transforms <code>ScalarQuadraticFunction</code> constraints into <code>RotatedSecondOrderCone</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/483">#483</a>).</li></ul><h2 id="[v0.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.1)-(September-22,-2018)"><a class="docs-heading-anchor" href="#[v0.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.1)-(September-22,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.1">v0.6.1</a> (September 22, 2018)</a><a id="[v0.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.1)-(September-22,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.1)-(September-22,-2018)" title="Permalink"></a></h2><ul><li>Enable <code>PositiveSemidefiniteConeSquare</code> set and quadratic functions in <code>MOIB.fullbridgeoptimizer</code> (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/524">#524</a>).</li><li>Add warning in the bridge between <code>PositiveSemidefiniteConeSquare</code> and <code>PositiveSemidefiniteConeTriangle</code> when the matrix is almost symmetric (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/522">#522</a>).</li><li>Modify <code>MOIT.copytest</code> to not add multiples constraints on the same variable (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/521">#521</a>).</li><li>Add missing keyword argument in one of <code>MOIU.add_scalar_constraint</code> methods (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/520">#520</a>).</li></ul><h2 id="[v0.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.0)-(August-30,-2018)"><a class="docs-heading-anchor" href="#[v0.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.0)-(August-30,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.0">v0.6.0</a> (August 30, 2018)</a><a id="[v0.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.0)-(August-30,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.0)-(August-30,-2018)" title="Permalink"></a></h2><ul><li>The <code>MOIU.@model</code> and <code>MOIB.@bridge</code> macros now support functions and sets defined in external modules. As a consequence, function and set names in the macro arguments need to be prefixed by module name.</li><li>Rename functions according to the <a href="https://jump.dev/JuMP.jl/stable/developers/style/">JuMP style guide</a>:<ul><li><code>copy!</code> with keyword arguments <code>copynames</code> and <code>warnattributes</code> -&gt; <code>copy_to</code> with keyword arguments <code>copy_names</code> and <code>warn_attributes</code>;</li><li><code>set!</code> -&gt; <code>set</code>;</li><li><code>addvariable[s]!</code> -&gt; <code>add_variable[s]</code>;</li><li><code>supportsconstraint</code> -&gt; <code>supports_constraint</code>;</li><li><code>addconstraint[s]!</code> -&gt; <code>add_constraint[s]</code>;</li><li><code>isvalid</code> -&gt; <code>is_valid</code>;</li><li><code>isempty</code> -&gt; <code>is_empty</code>;</li><li><code>Base.delete!</code> -&gt; <code>delete</code>;</li><li><code>modify!</code> -&gt; <code>modify</code>;</li><li><code>transform!</code> -&gt; <code>transform</code>;</li><li><code>initialize!</code> -&gt; <code>initialize</code>;</li><li><code>write</code> -&gt; <code>write_to_file</code>; and</li><li><code>read!</code> -&gt; <code>read_from_file</code>.</li></ul></li><li>Remove <code>free!</code> (use <code>Base.finalize</code> instead).</li><li>Add the <code>SquarePSD</code> bridge which transforms <code>PositiveSemidefiniteConeTriangle</code> constraints into <code>PositiveSemidefiniteConeTriangle</code>.</li><li>Add result fallback for <code>ConstraintDual</code> of variable-wise constraint, <code>ConstraintPrimal</code> and <code>ObjectiveValue</code>.</li><li>Add tests for <code>ObjectiveBound</code>.</li><li>Add test for empty rows in vector linear constraint.</li><li>Rework errors: <code>CannotError</code> has been renamed <code>NotAllowedError</code> and the distinction between <code>UnsupportedError</code> and <code>NotAllowedError</code> is now about whether the element is not supported (for example, it cannot be copied a model containing this element) or the operation is not allowed (either because it is not implemented, because it cannot be performed in the current state of the model, or because it cannot be performed for a specific index)</li><li><code>canget</code> is removed. <code>NoSolution</code> is added as a result status to indicate that the solver does not have either a primal or dual solution available (See #479).</li></ul><h2 id="[v0.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.5.0)-(August-5,-2018)"><a class="docs-heading-anchor" href="#[v0.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.5.0)-(August-5,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.5.0">v0.5.0</a> (August 5, 2018)</a><a id="[v0.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.5.0)-(August-5,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.5.0)-(August-5,-2018)" title="Permalink"></a></h2><ul><li>Fix names with CachingOptimizer.</li><li>Cleanup thanks to @mohamed82008.</li><li>Added a universal fallback for constraints.</li><li>Fast utilities for function canonicalization thanks to @rdeits.</li><li>Renamed <code>dimension</code> field to <code>side_dimension</code> in the context of matrix-like sets.</li><li>New and improved tests for cases like duplicate terms and <code>ObjectiveBound</code>.</li><li>Removed <code>cantransform</code>, <code>canaddconstraint</code>, <code>canaddvariable</code>, <code>canset</code>, <code>canmodify</code>, and <code>candelete</code> functions from the API. They are replaced by a new set of errors that are thrown: Subtypes of <code>UnsupportedError</code> indicate unsupported operations, while subtypes of <code>CannotError</code> indicate operations that cannot be performed in the current state.</li><li>The API for <code>copy!</code> is updated to remove the CopyResult type.</li><li>Updates for the new JuMP style guide.</li></ul><h2 id="[v0.4.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.1)-(June-28,-2018)"><a class="docs-heading-anchor" href="#[v0.4.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.1)-(June-28,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.1">v0.4.1</a> (June 28, 2018)</a><a id="[v0.4.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.1)-(June-28,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.4.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.1)-(June-28,-2018)" title="Permalink"></a></h2><ul><li>Fixes vector function modification on 32 bits.</li><li>Fixes Bellman-Ford algorithm for bridges.</li><li>Added an NLP test with <code>FeasibilitySense</code>.</li><li>Update modification documentation.</li></ul><h2 id="[v0.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.0)-(June-23,-2018)"><a class="docs-heading-anchor" href="#[v0.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.0)-(June-23,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.0">v0.4.0</a> (June 23, 2018)</a><a id="[v0.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.0)-(June-23,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.0)-(June-23,-2018)" title="Permalink"></a></h2><ul><li>Helper constructors for <code>VectorAffineTerm</code> and <code>VectorQuadraticTerm</code>.</li><li>Added <code>modify_lhs</code> to <code>TestConfig</code>.</li><li>Additional unit tests for optimizers.</li><li>Added a type parameter to <code>CachingOptimizer</code> for the <code>optimizer</code> field.</li><li>New API for problem modification (<a href="https://github.com/jump-dev/MathOptInterface.jl/issues/388">#388</a>)</li><li>Tests pass without deprecation warnings on Julia 0.7.</li><li>Small fixes and documentation updates.</li></ul><h2 id="[v0.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.3.0)-(May-25,-2018)"><a class="docs-heading-anchor" href="#[v0.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.3.0)-(May-25,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.3.0">v0.3.0</a> (May 25, 2018)</a><a id="[v0.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.3.0)-(May-25,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.3.0)-(May-25,-2018)" title="Permalink"></a></h2><ul><li>Functions have been redefined to use arrays-of-structs instead of structs-of-arrays.</li><li>Improvements to <code>MockOptimizer</code>.</li><li>Significant changes to <code>Bridges</code>.</li><li>New and improved unit tests.</li><li>Fixes for Julia 0.7.</li></ul><h2 id="[v0.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.2.0)-(April-24,-2018)"><a class="docs-heading-anchor" href="#[v0.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.2.0)-(April-24,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.2.0">v0.2.0</a> (April 24, 2018)</a><a id="[v0.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.2.0)-(April-24,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.2.0)-(April-24,-2018)" title="Permalink"></a></h2><ul><li>Improvements to and better coverage of <code>Tests</code>.</li><li>Documentation fixes.</li><li><code>SolverName</code> attribute.</li><li>Changes to the NLP interface (new definition of variable order and arrays of structs for bound pairs and sparsity patterns).</li><li>Addition of NLP tests.</li><li>Introduction of <code>UniversalFallback</code>.</li><li><code>copynames</code> keyword argument to <code>MOI.copy!</code>.</li><li>Add Bridges submodule.</li></ul><h2 id="[v0.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.1.0)-(February-28,-2018)"><a class="docs-heading-anchor" href="#[v0.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.1.0)-(February-28,-2018)"><a href="https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.1.0">v0.1.0</a> (February 28, 2018)</a><a id="[v0.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.1.0)-(February-28,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[v0.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.1.0)-(February-28,-2018)" title="Permalink"></a></h2><ul><li>Initial public release.</li><li>The framework for MOI was developed at the JuMP-dev workshop at MIT in June 2017 as a sorely needed replacement for MathProgBase.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../developer/checklists/">« Checklists</a><a class="docs-footer-nextpage" href="../../release_notes/">Release notes »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Benchmarks/overview/index.html b/previews/PR3525/moi/submodules/Benchmarks/overview/index.html
new file mode 100644
index 00000000000..2b8983051af
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Benchmarks/overview/index.html
@@ -0,0 +1,24 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Overview · JuMP</title><meta name="title" content="Overview · JuMP"/><meta property="og:title" content="Overview · JuMP"/><meta property="twitter:title" content="Overview · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Overview</a></li><li><a class="tocitem" href="../reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Benchmarks</a></li><li class="is-active"><a href>Overview</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Overview</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Benchmarks/overview.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-Benchmarks-submodule"><a class="docs-heading-anchor" href="#The-Benchmarks-submodule">The <code>Benchmarks</code> submodule</a><a id="The-Benchmarks-submodule-1"></a><a class="docs-heading-anchor-permalink" href="#The-Benchmarks-submodule" title="Permalink"></a></h1><p>To aid the development of efficient solver wrappers, MathOptInterface provides benchmarking capability. Benchmarking a wrapper follows a two-step process.</p><p>First, prior to making changes, create a baseline for the benchmark results on a given benchmark suite as follows:</p><pre><code class="language-julia hljs">using SolverPackage  # Replace with your choice of solver.
+import MathOptInterface as MOI
+
+suite = MOI.Benchmarks.suite() do
+    SolverPackage.Optimizer()
+end
+
+MOI.Benchmarks.create_baseline(
+    suite, &quot;current&quot;; directory = &quot;/tmp&quot;, verbose = true
+)</code></pre><p>Use the <code>exclude</code> argument to <a href="../reference/#MathOptInterface.Benchmarks.suite"><code>Benchmarks.suite</code></a> to exclude benchmarks that the solver doesn&#39;t support.</p><p>Second, after making changes to the package, re-run the benchmark suite and compare to the prior saved results:</p><pre><code class="language-julia hljs">using SolverPackage
+import MathOptInterface as MOI
+
+suite = MOI.Benchmarks.suite() do
+    SolverPackage.Optimizer()
+end
+
+MOI.Benchmarks.compare_against_baseline(
+    suite, &quot;current&quot;; directory = &quot;/tmp&quot;, verbose = true
+)</code></pre><p>This comparison will create a report detailing improvements and regressions.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../../reference/errors/">« Errors</a><a class="docs-footer-nextpage" href="../reference/">API Reference »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Benchmarks/reference/index.html b/previews/PR3525/moi/submodules/Benchmarks/reference/index.html
new file mode 100644
index 00000000000..b3d4cc69fda
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Benchmarks/reference/index.html
@@ -0,0 +1,21 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>API Reference · JuMP</title><meta name="title" content="API Reference · JuMP"/><meta property="og:title" content="API Reference · JuMP"/><meta property="twitter:title" content="API Reference · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../overview/">Overview</a></li><li class="is-active"><a class="tocitem" href>API Reference</a><ul class="internal"><li><a class="tocitem" href="#Benchmarks"><span>Benchmarks</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Benchmarks</a></li><li class="is-active"><a href>API Reference</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>API Reference</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Benchmarks/reference.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h2 id="Benchmarks"><a class="docs-heading-anchor" href="#Benchmarks">Benchmarks</a><a id="Benchmarks-1"></a><a class="docs-heading-anchor-permalink" href="#Benchmarks" title="Permalink"></a></h2><p>Functions to help benchmark the performance of solver wrappers. See <a href="../overview/#The-Benchmarks-submodule">The Benchmarks submodule</a> for more details.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Benchmarks.suite" href="#MathOptInterface.Benchmarks.suite"><code>MathOptInterface.Benchmarks.suite</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">suite(
+    new_model::Function;
+    exclude::Vector{Regex} = Regex[]
+)</code></pre><p>Create a suite of benchmarks. <code>new_model</code> should be a function that takes no arguments, and returns a new instance of the optimizer you wish to benchmark.</p><p>Use <code>exclude</code> to exclude a subset of benchmarks.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">suite() do
+    GLPK.Optimizer()
+end
+suite(exclude = [r&quot;delete&quot;]) do
+    Gurobi.Optimizer(OutputFlag=0)
+end</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Benchmarks/Benchmarks.jl#L14-L35">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Benchmarks.create_baseline" href="#MathOptInterface.Benchmarks.create_baseline"><code>MathOptInterface.Benchmarks.create_baseline</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">create_baseline(suite, name::String; directory::String = &quot;&quot;; kwargs...)</code></pre><p>Run all benchmarks in <code>suite</code> and save to files called <code>name</code> in <code>directory</code>.</p><p>Extra <code>kwargs</code> are based to <code>BenchmarkTools.run</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">my_suite = suite(() -&gt; GLPK.Optimizer())
+create_baseline(my_suite, &quot;glpk_master&quot;; directory = &quot;/tmp&quot;, verbose = true)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Benchmarks/Benchmarks.jl#L45-L58">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Benchmarks.compare_against_baseline" href="#MathOptInterface.Benchmarks.compare_against_baseline"><code>MathOptInterface.Benchmarks.compare_against_baseline</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">compare_against_baseline(
+    suite, name::String; directory::String = &quot;&quot;,
+    report_filename::String = &quot;report.txt&quot;
+)</code></pre><p>Run all benchmarks in <code>suite</code> and compare against files called <code>name</code> in <code>directory</code> that were created by a call to <code>create_baseline</code>.</p><p>A report summarizing the comparison is written to <code>report_filename</code> in <code>directory</code>.</p><p>Extra <code>kwargs</code> are based to <code>BenchmarkTools.run</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">my_suite = suite(() -&gt; GLPK.Optimizer())
+compare_against_baseline(
+    my_suite, &quot;glpk_master&quot;; directory = &quot;/tmp&quot;, verbose = true
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Benchmarks/Benchmarks.jl#L75-L97">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../overview/">« Overview</a><a class="docs-footer-nextpage" href="../../Bridges/overview/">Overview »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Bridges/list_of_bridges/index.html b/previews/PR3525/moi/submodules/Bridges/list_of_bridges/index.html
new file mode 100644
index 00000000000..0b98283c095
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Bridges/list_of_bridges/index.html
@@ -0,0 +1,114 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>List of bridges · JuMP</title><meta name="title" content="List of bridges · JuMP"/><meta property="og:title" content="List of bridges · JuMP"/><meta property="twitter:title" content="List of bridges · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../overview/">Overview</a></li><li class="is-active"><a class="tocitem" href>List of bridges</a><ul class="internal"><li><a class="tocitem" href="#constraint_bridges_ref"><span>Constraint bridges</span></a></li><li><a class="tocitem" href="#objective_bridges_ref"><span>Objective bridges</span></a></li><li><a class="tocitem" href="#variable_bridges_ref"><span>Variable bridges</span></a></li></ul></li><li><a class="tocitem" href="../reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Bridges</a></li><li class="is-active"><a href>List of bridges</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>List of bridges</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Bridges/list_of_bridges.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="List-of-bridges"><a class="docs-heading-anchor" href="#List-of-bridges">List of bridges</a><a id="List-of-bridges-1"></a><a class="docs-heading-anchor-permalink" href="#List-of-bridges" title="Permalink"></a></h1><p>This section describes the <a href="../reference/#MathOptInterface.Bridges.AbstractBridge"><code>Bridges.AbstractBridge</code></a>s that are implemented in MathOptInterface.</p><h2 id="constraint_bridges_ref"><a class="docs-heading-anchor" href="#constraint_bridges_ref">Constraint bridges</a><a id="constraint_bridges_ref-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_bridges_ref" title="Permalink"></a></h2><p>These bridges are subtypes of <a href="../reference/#MathOptInterface.Bridges.Constraint.AbstractBridge"><code>Bridges.Constraint.AbstractBridge</code></a>.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.GreaterToIntervalBridge" href="#MathOptInterface.Bridges.Constraint.GreaterToIntervalBridge"><code>MathOptInterface.Bridges.Constraint.GreaterToIntervalBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GreaterToIntervalBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>GreaterToIntervalBridge</code> implements the following reformulations:</p><ul><li><span>$f(x) \ge l$</span> into <span>$f(x) \in [l, \infty)$</span></li></ul><p><strong>Source node</strong></p><p><code>GreaterToIntervalBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>GreaterToIntervalBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/ltgt_to_interval.jl#L47-L65">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.LessToIntervalBridge" href="#MathOptInterface.Bridges.Constraint.LessToIntervalBridge"><code>MathOptInterface.Bridges.Constraint.LessToIntervalBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LessToIntervalBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>LessToIntervalBridge</code> implements the following reformulations:</p><ul><li><span>$f(x) \le u$</span> into <span>$f(x) \in (-\infty, u]$</span></li></ul><p><strong>Source node</strong></p><p><code>LessToIntervalBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>LessToIntervalBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/ltgt_to_interval.jl#L96-L114">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.GreaterToLessBridge" href="#MathOptInterface.Bridges.Constraint.GreaterToLessBridge"><code>MathOptInterface.Bridges.Constraint.GreaterToLessBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GreaterToLessBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>GreaterToLessBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \ge l$</span> into <span>$-f(x) \le -l$</span></li></ul><p><strong>Source node</strong></p><p><code>GreaterToLessBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>GreaterToLessBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/flip_sign.jl#L86-L104">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.LessToGreaterBridge" href="#MathOptInterface.Bridges.Constraint.LessToGreaterBridge"><code>MathOptInterface.Bridges.Constraint.LessToGreaterBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LessToGreaterBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>LessToGreaterBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \le u$</span> into <span>$-f(x) \ge -u$</span></li></ul><p><strong>Source node</strong></p><p><code>LessToGreaterBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>LessToGreaterBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/flip_sign.jl#L139-L157">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NonnegToNonposBridge" href="#MathOptInterface.Bridges.Constraint.NonnegToNonposBridge"><code>MathOptInterface.Bridges.Constraint.NonnegToNonposBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonnegToNonposBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NonnegToNonposBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \in \mathbb{R}_+$</span> into <span>$-f(x) \in \mathbb{R}_-$</span></li></ul><p><strong>Source node</strong></p><p><code>NonnegToNonposBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NonnegToNonposBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/flip_sign.jl#L189-L207">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NonposToNonnegBridge" href="#MathOptInterface.Bridges.Constraint.NonposToNonnegBridge"><code>MathOptInterface.Bridges.Constraint.NonposToNonnegBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonposToNonnegBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NonposToNonnegBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \in \mathbb{R}_-$</span> into <span>$-f(x) \in \mathbb{R}_+$</span></li></ul><p><strong>Source node</strong></p><p><code>NonposToNonnegBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NonposToNonnegBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/flip_sign.jl#L242-L260">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.VectorizeBridge" href="#MathOptInterface.Bridges.Constraint.VectorizeBridge"><code>MathOptInterface.Bridges.Constraint.VectorizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorizeBridge{T,F,S,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>VectorizeBridge</code> implements the following reformulations:</p><ul><li><span>$g(x) \ge a$</span> into <span>$[g(x) - a] \in \mathbb{R}_+$</span></li><li><span>$g(x) \le a$</span> into <span>$[g(x) - a] \in \mathbb{R}_-$</span></li><li><span>$g(x) == a$</span> into <span>$[g(x) - a] \in \{0\}$</span></li></ul><p>where <code>T</code> is the coefficient type of <code>g(x) - a</code>.</p><p><strong>Source node</strong></p><p><code>VectorizeBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>VectorizeBridge</code> creates:</p><ul><li><code>F</code> in <code>S</code>, where <code>S</code> is one of <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a> depending on the type of the input set.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/vectorize.jl#L7-L33">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ScalarizeBridge" href="#MathOptInterface.Bridges.Constraint.ScalarizeBridge"><code>MathOptInterface.Bridges.Constraint.ScalarizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarizeBridge{T,F,S}</code></pre><p><code>ScalarizeBridge</code> implements the following reformulations:</p><ul><li><span>$f(x) - a \in \mathbb{R}_+$</span> into <span>$f_i(x) \ge a_i$</span> for all <span>$i$</span></li><li><span>$f(x) - a \in \mathbb{R}_-$</span> into <span>$f_i(x) \le a_i$</span> for all <span>$i$</span></li><li><span>$f(x) - a \in \{0\}$</span> into <span>$f_i(x) == a_i$</span> for all <span>$i$</span></li></ul><p><strong>Source node</strong></p><p><code>ScalarizeBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives{T}</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives{T}</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>ScalarizeBridge</code> creates:</p><ul><li><code>F</code> in <code>S</code>, where <code>S</code> is one of <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a>, and <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a>, depending on the type of the input set.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/scalarize.jl#L7-L31">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ScalarSlackBridge" href="#MathOptInterface.Bridges.Constraint.ScalarSlackBridge"><code>MathOptInterface.Bridges.Constraint.ScalarSlackBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarSlackBridge{T,F,S} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>ScalarSlackBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \in S$</span> into <span>$f(x) - y == 0$</span> and <span>$y \in S$</span></li></ul><p><strong>Source node</strong></p><p><code>ScalarSlackBridge</code> supports:</p><ul><li><code>G</code> in <code>S</code>, where <code>G</code> is not <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> and <code>S</code> is not <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>ScalarSlackBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/slack.jl#L188-L208">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.VectorSlackBridge" href="#MathOptInterface.Bridges.Constraint.VectorSlackBridge"><code>MathOptInterface.Bridges.Constraint.VectorSlackBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorSlackBridge{T,F,S} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>VectorSlackBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \in S$</span> into <span>$f(x) - y \in \{0\}$</span> and <span>$y \in S$</span></li></ul><p><strong>Source node</strong></p><p><code>VectorSlackBridge</code> supports:</p><ul><li><code>G</code> in <code>S</code>, where <code>G</code> is not <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> and <code>S</code> is not <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>VectorSlackBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/slack.jl#L319-L339">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ScalarFunctionizeBridge" href="#MathOptInterface.Bridges.Constraint.ScalarFunctionizeBridge"><code>MathOptInterface.Bridges.Constraint.ScalarFunctionizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarFunctionizeBridge{T,S} = FunctionConversionBridge{T,MOI.ScalarAffineFunction{T},MOI.VariableIndex,S}</code></pre><p><code>ScalarFunctionizeBridge</code> implements the following reformulations:</p><ul><li><span>$x \in S$</span> into <span>$1x + 0 \in S$</span></li></ul><p><strong>Source node</strong></p><p><code>ScalarFunctionizeBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <code>S</code></li></ul><p><strong>Target nodes</strong></p><p><code>ScalarFunctionizeBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L296-L315">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.VectorFunctionizeBridge" href="#MathOptInterface.Bridges.Constraint.VectorFunctionizeBridge"><code>MathOptInterface.Bridges.Constraint.VectorFunctionizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorFunctionizeBridge{T,S} = FunctionConversionBridge{T,MOI.VectorAffineFunction{T},S}</code></pre><p><code>VectorFunctionizeBridge</code> implements the following reformulations:</p><ul><li><span>$x \in S$</span> into <span>$Ix + 0 \in S$</span></li></ul><p><strong>Source node</strong></p><p><code>VectorFunctionizeBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <code>S</code></li></ul><p><strong>Target nodes</strong></p><p><code>VectorFunctionizeBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L324-L342">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ToScalarQuadraticBridge" href="#MathOptInterface.Bridges.Constraint.ToScalarQuadraticBridge"><code>MathOptInterface.Bridges.Constraint.ToScalarQuadraticBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ToScalarQuadraticBridge{T,G,S} &lt;: AbstractFunctionConversionBridge{G,S}</code></pre><p><code>ToScalarQuadraticBridge</code> implements the following reformulation:</p><ul><li><span>$g(x) \in S$</span> into <span>$f(x) \in S$</span></li></ul><p>where <code>g</code> is an abstract scalar function and <code>f</code> is a <a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction</code></a>.</p><p><strong>Source node</strong></p><p><code>ToScalarQuadraticBridge</code> supports:</p><ul><li><code>G&lt;:AbstractScalarFunction</code> in <code>S</code></li></ul><p><strong>Target nodes</strong></p><p><code>ToScalarQuadraticBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction</code></a> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L355-L376">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ToVectorQuadraticBridge" href="#MathOptInterface.Bridges.Constraint.ToVectorQuadraticBridge"><code>MathOptInterface.Bridges.Constraint.ToVectorQuadraticBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ToVectorQuadraticBridge{T,G,S} &lt;: AbstractFunctionConversionBridge{G,S}</code></pre><p><code>ToVectorQuadraticBridge</code> implements the following reformulation:</p><ul><li><span>$g(x) \in S$</span> into <span>$f(x) \in S$</span></li></ul><p>where <code>g</code> is an abstract vector function and <code>f</code> is a <a href="../../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>MOI.VectorQuadraticFunction</code></a>.</p><p><strong>Source node</strong></p><p><code>ToVectorQuadraticBridge</code> supports:</p><ul><li><code>G&lt;:AbstractVectorFunction</code> in <code>S</code></li></ul><p><strong>Target nodes</strong></p><p><code>ToVectorQuadraticBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>MOI.VectorQuadraticFunction</code></a> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L385-L406">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ToScalarNonlinearBridge" href="#MathOptInterface.Bridges.Constraint.ToScalarNonlinearBridge"><code>MathOptInterface.Bridges.Constraint.ToScalarNonlinearBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ToScalarNonlinearBridge{T,G,S} &lt;: AbstractFunctionConversionBridge{G,S}</code></pre><p><code>ToScalarNonlinearBridge</code> implements the following reformulation:</p><ul><li><span>$g(x) \in S$</span> into <span>$f(x) \in S$</span></li></ul><p>where <code>g</code> is an abstract scalar function and <code>f</code> is a <a href="../../../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>MOI.ScalarNonlinearFunction</code></a>.</p><p><strong>Source node</strong></p><p><code>ToScalarNonlinearBridge</code> supports:</p><ul><li><code>G&lt;:AbstractScalarFunction</code> in <code>S</code></li></ul><p><strong>Target nodes</strong></p><p><code>ToScalarNonlinearBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>MOI.ScalarNonlinearFunction</code></a> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L415-L436">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.FunctionConversionBridge" href="#MathOptInterface.Bridges.Constraint.FunctionConversionBridge"><code>MathOptInterface.Bridges.Constraint.FunctionConversionBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">FunctionConversionBridge{T,F,G,S} &lt;: AbstractFunctionConversionBridge{G,S}</code></pre><p><code>FunctionConversionBridge</code> implements the following reformulations:</p><ul><li><span>$g(x) \in S$</span> into <span>$f(x) \in S$</span></li></ul><p>for these pairs of functions:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a><code>to [</code>MOI.ScalarQuadraticFunction`](@ref)</li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction</code></a>  to <a href="../../../reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>MOI.ScalarNonlinearFunction</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a> to <a href="../../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>MOI.VectorQuadraticFunction</code></a></li></ul><p><strong>Source node</strong></p><p><code>FunctionConversionBridge</code> supports:</p><ul><li><code>G</code> in <code>S</code></li></ul><p><strong>Target nodes</strong></p><p><code>FunctionConversionBridge</code> creates:</p><ul><li><code>F</code> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L159-L183">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SplitComplexEqualToBridge" href="#MathOptInterface.Bridges.Constraint.SplitComplexEqualToBridge"><code>MathOptInterface.Bridges.Constraint.SplitComplexEqualToBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SplitComplexEqualToBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SplitComplexEqualToBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) + g(x) * im = a + b * im$</span> into <span>$f(x) = a$</span> and <span>$g(x) = b$</span></li></ul><p><strong>Source node</strong></p><p><code>SplitComplexEqualToBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Complex{T}}</code></a></li></ul><p>where <code>G</code> is a function with <code>Complex</code> coefficients.</p><p><strong>Target nodes</strong></p><p><code>SplitComplexEqualToBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul><p>where <code>F</code> is the type of the real/imaginary part of <code>G</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/split_complex_equalto.jl#L7-L29">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SplitComplexZerosBridge" href="#MathOptInterface.Bridges.Constraint.SplitComplexZerosBridge"><code>MathOptInterface.Bridges.Constraint.SplitComplexZerosBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SplitComplexZerosBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SplitComplexZerosBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \in \{0\}^n$</span> into <span>$\text{Re}(f(x)) \in \{0\}^n$</span> and <span>$\text{Im}(f(x)) \in \{0\}^n$</span></li></ul><p><strong>Source node</strong></p><p><code>SplitComplexZerosBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li></ul><p>where <code>G</code> is a function with <code>Complex</code> coefficients.</p><p><strong>Target nodes</strong></p><p><code>SplitComplexZerosBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li></ul><p>where <code>F</code> is the type of the real/imaginary part of <code>G</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/split_complex_zeros.jl#L7-L30">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SplitHyperRectangleBridge" href="#MathOptInterface.Bridges.Constraint.SplitHyperRectangleBridge"><code>MathOptInterface.Bridges.Constraint.SplitHyperRectangleBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SplitHyperRectangleBridge{T,G,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SplitHyperRectangleBridge</code> implements the following reformulation:</p><ul><li><span>$f(x) \in \textsf{HyperRectangle}(l, u)$</span> to <span>$[f(x) - l; u - f(x)] \in \mathbb{R}_+$</span>.</li></ul><p><strong>Source node</strong></p><p><code>SplitHyperRectangleBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.HyperRectangle"><code>MOI.HyperRectangle</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>SplitHyperRectangleBridge</code> creates:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/split_hyperrectangle.jl#L7-L26">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SplitIntervalBridge" href="#MathOptInterface.Bridges.Constraint.SplitIntervalBridge"><code>MathOptInterface.Bridges.Constraint.SplitIntervalBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SplitIntervalBridge{T,F,S,LS,US} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SplitIntervalBridge</code> implements the following reformulations:</p><ul><li><span>$l \le f(x) \le u$</span> into <span>$f(x) \ge l$</span> and <span>$f(x) \le u$</span></li><li><span>$f(x) = b$</span> into <span>$f(x) \ge b$</span> and <span>$f(x) \le b$</span></li><li><span>$f(x) \in \{0\}$</span> into <span>$f(x) \in \mathbb{R}_+$</span> and <span>$f(x) \in \mathbb{R}_-$</span></li></ul><p><strong>Source node</strong></p><p><code>SplitIntervalBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{T}</code></a></li><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>SplitIntervalBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul><p>or</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives</code></a></li></ul><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>If <code>T&lt;:AbstractFloat</code> and <code>S</code> is <code>MOI.Interval{T}</code> then no lower (resp. upper) bound constraint is created if the lower (resp. upper) bound is <code>typemin(T)</code> (resp. <code>typemax(T)</code>). Similarly, when <a href="../../../reference/constraints/#MathOptInterface.ConstraintSet"><code>MOI.ConstraintSet</code></a> is set, a lower or upper bound constraint may be deleted or created accordingly.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/interval.jl#L7-L44">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SOCtoRSOCBridge" href="#MathOptInterface.Bridges.Constraint.SOCtoRSOCBridge"><code>MathOptInterface.Bridges.Constraint.SOCtoRSOCBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOCtoRSOCBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SOCtoRSOCBridge</code> implements the following reformulation:</p><ul><li><span>$||x||_2 \le t$</span> into <span>$(t+x_1)(t-x_1)\ge ||(x_2\ldots,x_N)||_2^2$</span></li></ul><p><strong>Assumptions</strong></p><ul><li><code>SOCtoRSOCBridge</code> assumes that the length of <code>x</code> is at least one.</li></ul><p><strong>Source node</strong></p><p><code>SOCtoRSOCBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p><strong>Target node</strong></p><p><code>SOCtoRSOCBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/soc_rsoc.jl#L7-L29">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.RSOCtoSOCBridge" href="#MathOptInterface.Bridges.Constraint.RSOCtoSOCBridge"><code>MathOptInterface.Bridges.Constraint.RSOCtoSOCBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RSOCtoSOCBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>RSOCtoSOCBridge</code> implements the following reformulation:</p><ul><li><span>$||x||_2^2 \le 2tu$</span> into <span>$||\frac{t - u}{\sqrt 2}, x||_2 \le  \frac{t + u}{\sqrt 2}$</span></li></ul><p><strong>Source node</strong></p><p><code>RSOCtoSOCBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul><p><strong>Target node</strong></p><p><code>RSOCtoSOCBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/rsoc_soc.jl#L7-L26">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge" href="#MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge"><code>MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOCtoNonConvexQuadBridge{T} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SOCtoNonConvexQuadBridge</code> implements the following reformulations:</p><ul><li><span>$||x||_2 \le t$</span> into <span>$\sum x^2 - t^2 \le 0$</span> and <span>$1t + 0 \ge 0$</span></li></ul><p>The <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a> <span>$1t + 0$</span> is used in case the variable has other bound constraints.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This transformation starts from a convex constraint and creates a non-convex constraint. Unless the solver has explicit support for detecting second-order cones in quadratic form, this may (wrongly) be interpreted by the solver as being non-convex. Therefore, this bridge is not added automatically by <a href="../reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>MOI.Bridges.full_bridge_optimizer</code></a>. Care is recommended when adding this bridge to a optimizer.</p></div></div><p><strong>Source node</strong></p><p><code>SOCtoNonConvexQuadBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>SOCtoNonConvexQuadBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/soc_to_nonconvex_quad.jl#L83-L113">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.RSOCtoNonConvexQuadBridge" href="#MathOptInterface.Bridges.Constraint.RSOCtoNonConvexQuadBridge"><code>MathOptInterface.Bridges.Constraint.RSOCtoNonConvexQuadBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RSOCtoNonConvexQuadBridge{T} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>RSOCtoNonConvexQuadBridge</code> implements the following reformulations:</p><ul><li><span>$||x||_2^2 \le 2tu$</span> into <span>$\sum x^2 - 2tu \le 0$</span>, <span>$1t + 0 \ge 0$</span>, and <span>$1u + 0 \ge 0$</span>.</li></ul><p>The <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a>s <span>$1t + 0$</span> and <span>$1u + 0$</span> are used in case the variables have other bound constraints.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This transformation starts from a convex constraint and creates a non-convex constraint. Unless the solver has explicit support for detecting rotated second-order cones in quadratic form, this may (wrongly) be interpreted by the solver as being non-convex. Therefore, this bridge is not added automatically by <a href="../reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>MOI.Bridges.full_bridge_optimizer</code></a>. Care is recommended when adding this bridge to a optimizer.</p></div></div><p><strong>Source node</strong></p><p><code>RSOCtoNonConvexQuadBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>RSOCtoNonConvexQuadBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/soc_to_nonconvex_quad.jl#L166-L197">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.QuadtoSOCBridge" href="#MathOptInterface.Bridges.Constraint.QuadtoSOCBridge"><code>MathOptInterface.Bridges.Constraint.QuadtoSOCBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">QuadtoSOCBridge{T} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>QuadtoSOCBridge</code> converts quadratic inequalities</p><p class="math-container">\[\frac{1}{2}x^T Q x + a^T x \le ub\]</p><p>into <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a> constraints, but it only applies when <span>$Q$</span> is positive definite.</p><p>This is because, if <code>Q</code> is positive definite, there exists <code>U</code> such that <span>$Q = U^T U$</span>, and so the inequality can then be rewritten as;</p><p class="math-container">\[\|U x\|_2^2 \le 2 (-a^T x + ub)\]</p><p>Therefore, <code>QuadtoSOCBridge</code> implements the following reformulations:</p><ul><li><span>$\frac{1}{2}x^T Q x + a^T x \le ub$</span> into <span>$(1, -a^T x + ub, Ux) \in RotatedSecondOrderCone$</span> where <span>$Q = U^T U$</span></li><li><span>$\frac{1}{2}x^T Q x + a^T x \ge lb$</span> into <span>$(1, a^T x - lb, Ux) \in RotatedSecondOrderCone$</span> where <span>$-Q = U^T U$</span></li></ul><p><strong>Source node</strong></p><p><code>QuadtoSOCBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>RelativeEntropyBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul><p><strong>Errors</strong></p><p>This bridge errors if <code>Q</code> is not positive definite.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/quad_to_soc.jl#L7-L48">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SOCtoPSDBridge" href="#MathOptInterface.Bridges.Constraint.SOCtoPSDBridge"><code>MathOptInterface.Bridges.Constraint.SOCtoPSDBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOCtoPSDBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SOCtoPSDBridge</code> implements the following reformulation:</p><ul><li><span>$||x||_2 \le t$</span> into <span>$\left[\begin{array}{c c}t &amp; x^\top \\ x &amp; t \mathbf{I}\end{array}\right]\succeq 0$</span></li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This bridge is not added by default by <a href="../reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>MOI.Bridges.full_bridge_optimizer</code></a> because bridging second order cone constraints to semidefinite constraints can be achieved by the <a href="#MathOptInterface.Bridges.Constraint.SOCtoRSOCBridge"><code>SOCtoRSOCBridge</code></a> followed by the <a href="#MathOptInterface.Bridges.Constraint.RSOCtoPSDBridge"><code>RSOCtoPSDBridge</code></a>, while creating a smaller semidefinite constraint.</p></div></div><p><strong>Source node</strong></p><p><code>SOCtoPSDBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>SOCtoPSDBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/soc_to_psd.jl#L41-L66">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.RSOCtoPSDBridge" href="#MathOptInterface.Bridges.Constraint.RSOCtoPSDBridge"><code>MathOptInterface.Bridges.Constraint.RSOCtoPSDBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RSOCtoPSDBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>RSOCtoPSDBridge</code> implements the following reformulation:</p><ul><li><span>$||x||_2^2 \le 2t\cdot u$</span> into <span>$\left[\begin{array}{c c}t &amp; x^\top \\ x &amp; 2tu \mathbf{I}\end{array}\right]\succeq 0$</span></li></ul><p><strong>Source node</strong></p><p><code>RSOCtoPSDBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>RSOCtoPSDBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/soc_to_psd.jl#L153-L172">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NormInfinityBridge" href="#MathOptInterface.Bridges.Constraint.NormInfinityBridge"><code>MathOptInterface.Bridges.Constraint.NormInfinityBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormInfinityBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NormInfinityBridge</code> implements the following reformulation:</p><ul><li><span>$|x|_\infty \le t$</span> into <span>$[t - x_i, t + x_i] \in \mathbb{R}_+$</span>.</li></ul><p><strong>Source node</strong></p><p><code>NormInfinityBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormInfinityCone"><code>MOI.NormInfinityCone{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NormInfinityBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_infinity.jl#L7-L25">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NormOneBridge" href="#MathOptInterface.Bridges.Constraint.NormOneBridge"><code>MathOptInterface.Bridges.Constraint.NormOneBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormOneBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NormOneBridge</code> implements the following reformulation:</p><ul><li><span>$\sum |x_i| \le t$</span> into <span>$[t - \sum y_i, y_i - x_i, y_i + x_i] \in \mathbb{R}_+$</span>.</li></ul><p><strong>Source node</strong></p><p><code>NormOneBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormOneCone"><code>MOI.NormOneCone{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NormOneBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_one.jl#L7-L26">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NormToPowerBridge" href="#MathOptInterface.Bridges.Constraint.NormToPowerBridge"><code>MathOptInterface.Bridges.Constraint.NormToPowerBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormToPowerBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NormToPowerBridge</code> implements the following reformulation:</p><ul><li><span>$(t, x) \in NormCone(p, 1+d)$</span> into <span>$(r_i, t, x_i) \in PowerCone(1 / p)$</span> for all <span>$i$</span>, and <span>$\sum\limits_i r_i == t$</span>.</li></ul><p>For details, see Alizadeh, F., and Goldfarb, D. (2001). &quot;Second-order cone programming.&quot; Mathematical Programming, Series B, 95:3-51.</p><p><strong>Source node</strong></p><p><code>NormToPowerBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormCone"><code>MOI.NormCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NormToPowerBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PowerCone"><code>MOI.PowerCone{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_to_power.jl#L7-L30">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NormOneConeToNormConeBridge" href="#MathOptInterface.Bridges.Constraint.NormOneConeToNormConeBridge"><code>MathOptInterface.Bridges.Constraint.NormOneConeToNormConeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormOneConeToNormConeBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NormOneConeToNormConeBridge</code> implements the following reformulations:</p><ul><li><span>$(t, x) in NormOneCone(d)$</span> into <span>$(t, x) in NormCone(1, d)$</span></li></ul><p><strong>Source node</strong></p><p><code>NormOneConeToNormConeBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormOneCone"><code>MOI.NormOneCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NormOneConeToNormConeBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormCone"><code>MOI.NormCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_special_case.jl#L41-L59">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SecondOrderConeToNormConeBridge" href="#MathOptInterface.Bridges.Constraint.SecondOrderConeToNormConeBridge"><code>MathOptInterface.Bridges.Constraint.SecondOrderConeToNormConeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SecondOrderConeToNormConeBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SecondOrderConeToNormConeBridge</code> implements the following reformulations:</p><ul><li><span>$(t, x) in SecondOrderCone(d)$</span> into <span>$(t, x) in NormCone(2, d)$</span></li></ul><p><strong>Source node</strong></p><p><code>SecondOrderConeToNormConeBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>SecondOrderConeToNormConeBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormCone"><code>MOI.NormCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_special_case.jl#L66-L84">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NormInfinityConeToNormConeBridge" href="#MathOptInterface.Bridges.Constraint.NormInfinityConeToNormConeBridge"><code>MathOptInterface.Bridges.Constraint.NormInfinityConeToNormConeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormInfinityConeToNormConeBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NormInfinityConeToNormConeBridge</code> implements the following reformulations:</p><ul><li><span>$(t, x) in NormInfinityCone(d)$</span> into <span>$(t, x) in NormCone(Inf, d)$</span></li></ul><p><strong>Source node</strong></p><p><code>NormInfinityConeToNormConeBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormInfinityCone"><code>MOI.NormInfinityCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NormInfinityConeToNormConeBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormCone"><code>MOI.NormCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_special_case.jl#L91-L109">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.GeoMeantoRelEntrBridge" href="#MathOptInterface.Bridges.Constraint.GeoMeantoRelEntrBridge"><code>MathOptInterface.Bridges.Constraint.GeoMeantoRelEntrBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GeoMeantoRelEntrBridge{T,F,G,H} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>GeoMeantoRelEntrBridge</code> implements the following reformulation:</p><ul><li><span>$(u, w) \in GeometricMeanCone$</span> into <span>$(0, w, (u + y)\mathbf{1})\in RelativeEntropyCone$</span> and <span>$y \ge 0$</span></li></ul><p><strong>Source node</strong></p><p><code>GeoMeantoRelEntrBridge</code> supports:</p><ul><li><code>H</code> in <a href="../../../reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>GeoMeantoRelEntrBridge</code> creates:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.RelativeEntropyCone"><code>MOI.RelativeEntropyCone</code></a></li><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul><p><strong>Derivation</strong></p><p>The derivation of the bridge is as follows:</p><p class="math-container">\[\begin{aligned}
+(u, w) \in GeometricMeanCone \iff &amp; u \le \left(\prod_{i=1}^n w_i\right)^{1/n} \\
+\iff &amp; 0 \le u + y \le \left(\prod_{i=1}^n w_i\right)^{1/n}, y \ge 0 \\
+\iff &amp; 1 \le \frac{\left(\prod_{i=1}^n w_i\right)^{1/n}}{u + y}, y \ge 0 \\
+\iff &amp; 1 \le \left(\prod_{i=1}^n \frac{w_i}{u + y}\right)^{1/n}, y \ge 0 \\
+\iff &amp; 0 \le \sum_{i=1}^n \log\left(\frac{w_i}{u + y}\right), y \ge 0 \\
+\iff &amp; 0 \ge \sum_{i=1}^n \log\left(\frac{u + y}{w_i}\right), y \ge 0 \\
+\iff &amp; 0 \ge \sum_{i=1}^n (u + y) \log\left(\frac{u + y}{w_i}\right), y \ge 0 \\
+\iff &amp; (0, w, (u + y)\mathbf{1}) \in RelativeEntropyCone, y \ge 0 \\
+\end{aligned}\]</p><p>This derivation assumes that <span>$u + y &gt; 0$</span>, which is enforced by the relative entropy cone.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/geomean_to_relentr.jl#L7-L47">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge" href="#MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge"><code>MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GeoMeanToPowerBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>GeoMeanToPowerBridge</code> implements the following reformulation:</p><ul><li><span>$(y, x...) \in GeometricMeanCone(1+d)$</span> into <span>$(x_1, t, y) \in PowerCone(1/d)$</span> and <span>$(t, x_2, ..., x_d) in GeometricMeanCone(d)$</span>, which is then recursively expanded into more <code>PowerCone</code> constraints.</li></ul><p><strong>Source node</strong></p><p><code>GeoMeanToPowerBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>GeoMeanToPowerBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PowerCone"><code>MOI.PowerCone{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/geomean_to_power.jl#L7-L28">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.GeoMeanBridge" href="#MathOptInterface.Bridges.Constraint.GeoMeanBridge"><code>MathOptInterface.Bridges.Constraint.GeoMeanBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">GeoMeanBridge{T,F,G,H} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>GeoMeanBridge</code> implements a reformulation from <a href="../../../reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a> into <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a>.</p><p>The reformulation is best described in an example.</p><p>Consider the cone of dimension 4:</p><p class="math-container">\[t \le \sqrt[3]{x_1 x_2 x_3}\]</p><p>This can be rewritten as <span>$\exists y \ge 0$</span> such that:</p><p class="math-container">\[\begin{align*}
+  t &amp; \le y,\\
+  y^4 &amp; \le x_1 x_2 x_3 y.
+\end{align*}\]</p><p>Note that we need to create <span>$y$</span> and not use <span>$t^4$</span> directly because <span>$t$</span> is allowed to be negative.</p><p>This is equivalent to:</p><p class="math-container">\[\begin{align*}
+    t &amp; \le \frac{y_1}{\sqrt{4}},\\
+    y_1^2 &amp; \le 2y_2 y_3,\\
+    y_2^2 &amp; \le 2x_1 x_2, \\
+    y_3^2 &amp; \le 2x_3(y_1/\sqrt{4}) \\
+    y     &amp; \ge 0.
+\end{align*}\]</p><p>More generally, you can show how the geometric mean code is recursively expanded into a set of new variables <span>$y$</span> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a>, a set of <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a> constraints, and a <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan</code></a> constraint between <span>$t$</span> and <span>$y_1$</span>.</p><p><strong>Source node</strong></p><p><code>GeoMeanBridge</code> supports:</p><ul><li><code>H</code> in <a href="../../../reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>GeoMeanBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/geomean.jl#L7-L58">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.RelativeEntropyBridge" href="#MathOptInterface.Bridges.Constraint.RelativeEntropyBridge"><code>MathOptInterface.Bridges.Constraint.RelativeEntropyBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RelativeEntropyBridge{T,F,G,H} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>RelativeEntropyBridge</code> implements the following reformulation that converts a <a href="../../../reference/standard_form/#MathOptInterface.RelativeEntropyCone"><code>MOI.RelativeEntropyCone</code></a> into an <a href="../../../reference/standard_form/#MathOptInterface.ExponentialCone"><code>MOI.ExponentialCone</code></a>:</p><ul><li><span>$u \ge \sum_{i=1}^n w_i \log \left(\frac{w_i}{v_i}\right)$</span> into <span>$y_i \ge 0$</span>, <span>$u \ge \sum_{i=1}^n y_i$</span>, and <span>$(-y_i, w_i, v_i) \in ExponentialCone$</span>.</li></ul><p><strong>Source node</strong></p><p><code>RelativeEntropyBridge</code> supports:</p><ul><li><code>H</code> in <a href="../../../reference/standard_form/#MathOptInterface.RelativeEntropyCone"><code>MOI.RelativeEntropyCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>RelativeEntropyBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.ExponentialCone"><code>MOI.ExponentialCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/relentr_to_exp.jl#L7-L29">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NormSpectralBridge" href="#MathOptInterface.Bridges.Constraint.NormSpectralBridge"><code>MathOptInterface.Bridges.Constraint.NormSpectralBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormSpectralBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NormSpectralBridge</code> implements the following reformulation:</p><ul><li><span>$t \ge \sigma_1(X)$</span> into <span>$\left[\begin{array}{c c}t\mathbf{I} &amp; X^\top \\ X &amp; t \mathbf{I}\end{array}\right] \succeq 0$</span></li></ul><p><strong>Source node</strong></p><p><code>NormSpectralBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormSpectralCone"><code>MOI.NormSpectralCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NormSpectralBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_spec_nuc_to_psd.jl#L7-L26">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NormNuclearBridge" href="#MathOptInterface.Bridges.Constraint.NormNuclearBridge"><code>MathOptInterface.Bridges.Constraint.NormNuclearBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NormNuclearBridge{T,F,G,H} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NormNuclearBridge</code> implements the following reformulation:</p><ul><li><span>$t \ge \sum_i \sigma_i (X)$</span> into <span>$\left[\begin{array}{c c}U &amp; X^\top \\ X &amp; V\end{array}\right] \succeq 0$</span> and <span>$2t \ge tr(U) + tr(V)$</span>.</li></ul><p><strong>Source node</strong></p><p><code>NormNuclearBridge</code> supports:</p><ul><li><code>H</code> in <a href="../../../reference/standard_form/#MathOptInterface.NormNuclearCone"><code>MOI.NormNuclearCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NormNuclearBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/norm_spec_nuc_to_psd.jl#L218-L239">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SquareBridge" href="#MathOptInterface.Bridges.Constraint.SquareBridge"><code>MathOptInterface.Bridges.Constraint.SquareBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SquareBridge{T,F,G,TT,ST} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SquareBridge</code> implements the following reformulations:</p><ul><li><span>$(t, u, X) \in LogDetConeSquare$</span> into <span>$(t, u, Y) in LogDetConeTriangle$</span></li><li><span>$(t, X) \in RootDetConeSquare$</span> into <span>$(t, Y) in RootDetConeTriangle$</span></li><li><span>$X \in AbstractSymmetricMatrixSetSquare$</span> into <span>$Y in AbstractSymmetricMatrixSetTriangle$</span></li></ul><p>where <span>$Y$</span> is the upper triangluar component of <span>$X$</span>.</p><p>In addition, constraints are added as necessary to constrain the matrix <span>$X$</span> to be symmetric. For example, the constraint for the matrix:</p><p class="math-container">\[\begin{pmatrix}
+  1      &amp; 1 + x &amp; 2 - 3x\\
+  1 +  x &amp; 2 + x &amp; 3 -  x\\
+  2 - 3x &amp; 2 + x &amp;     2x
+\end{pmatrix}\]</p><p>can be broken down to the constraint of the symmetric matrix</p><p class="math-container">\[\begin{pmatrix}
+  1      &amp; 1 + x &amp; 2 - 3x\\
+  \cdot &amp; 2 + x &amp; 3 -  x\\
+  \cdot &amp; \cdot &amp;    2x
+\end{pmatrix}\]</p><p>and the equality constraint between the off-diagonal entries (2, 3) and (3, 2) <span>$3 - x == 2 + x$</span>. Note that no symmetrization constraint needs to be added between the off-diagonal entries (1, 2) and (2, 1) or between (1, 3) and (3, 1) because the expressions are the same.</p><p><strong>Source node</strong></p><p><code>SquareBridge</code> supports:</p><ul><li><code>F</code> in <code>ST</code></li></ul><p><strong>Target nodes</strong></p><p><code>SquareBridge</code> creates:</p><ul><li><code>G</code> in <code>TT</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/square.jl#L7-L52">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge" href="#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge"><code>MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HermitianToSymmetricPSDBridge{T,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>HermitianToSymmetricPSDBridge</code> implements the following reformulation:</p><ul><li>Hermitian positive semidefinite <code>n x n</code> complex matrix to a symmetric positive semidefinite <code>2n x 2n</code> real matrix.</li></ul><p>See also <a href="#MathOptInterface.Bridges.Variable.HermitianToSymmetricPSDBridge"><code>MOI.Bridges.Variable.HermitianToSymmetricPSDBridge</code></a>.</p><p><strong>Source node</strong></p><p><code>HermitianToSymmetricPSDBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>MOI.HermitianPositiveSemidefiniteConeTriangle</code></a></li></ul><p><strong>Target node</strong></p><p><code>HermitianToSymmetricPSDBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li></ul><p><strong>Reformulation</strong></p><p>The reformulation is best described by example.</p><p>The Hermitian matrix:</p><p class="math-container">\[\begin{bmatrix}
+  x_{11}            &amp; x_{12} + y_{12}im &amp; x_{13} + y_{13}im\\
+  x_{12} - y_{12}im &amp; x_{22}            &amp; x_{23} + y_{23}im\\
+  x_{13} - y_{13}im &amp; x_{23} - y_{23}im &amp; x_{33}
+\end{bmatrix}\]</p><p>is positive semidefinite if and only if the symmetric matrix:</p><p class="math-container">\[\begin{bmatrix}
+    x_{11} &amp; x_{12} &amp; x_{13} &amp; 0       &amp; y_{12}  &amp; y_{13} \\
+           &amp; x_{22} &amp; x_{23} &amp; -y_{12} &amp; 0       &amp; y_{23} \\
+           &amp;        &amp; x_{33} &amp; -y_{13} &amp; -y_{23} &amp; 0      \\
+           &amp;        &amp;        &amp; x_{11}  &amp; x_{12}  &amp; x_{13} \\
+           &amp;        &amp;        &amp;         &amp; x_{22}  &amp; x_{23} \\
+           &amp;        &amp;        &amp;         &amp;         &amp; x_{33}
+\end{bmatrix}\]</p><p>is positive semidefinite.</p><p>The bridge achieves this reformulation by constraining the above matrix to belong to the <code>MOI.PositiveSemidefiniteConeTriangle(6)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/hermitian.jl#L7-L56">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SetDotScalingBridge" href="#MathOptInterface.Bridges.Constraint.SetDotScalingBridge"><code>MathOptInterface.Bridges.Constraint.SetDotScalingBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SetDotScalingBridge{T,S,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SetDotScalingBridge</code> implements the reformulation from constraints in <code>S</code> to constraints in <a href="../../../reference/standard_form/#MathOptInterface.Scaled"><code>MOI.Scaled{S}</code></a>.</p><p><strong>Source node</strong></p><p><code>SetDotScalingBridge</code> supports:</p><ul><li><code>G</code> in <code>S</code></li></ul><p><strong>Target node</strong></p><p><code>SetDotScalingBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Scaled"><code>MOI.Scaled{S}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/set_dot_scaling.jl#L7-L24">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SetDotInverseScalingBridge" href="#MathOptInterface.Bridges.Constraint.SetDotInverseScalingBridge"><code>MathOptInterface.Bridges.Constraint.SetDotInverseScalingBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SetDotInverseScalingBridge{T,S,F,G} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SetDotInverseScalingBridge</code> implements the reformulation from constraints in the <code>MOI.Scaled{S}</code> to constraints in the <code>S</code>.</p><p><strong>Source node</strong></p><p><code>SetDotInverseScalingBridge</code> supports:</p><ul><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.Scaled"><code>MOI.Scaled{S}</code></a></li></ul><p><strong>Target node</strong></p><p><code>SetDotInverseScalingBridge</code> creates:</p><ul><li><code>F</code> in <code>S</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/set_dot_scaling.jl#L111-L128">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.RootDetBridge" href="#MathOptInterface.Bridges.Constraint.RootDetBridge"><code>MathOptInterface.Bridges.Constraint.RootDetBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RootDetBridge{T,F,G,H} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p>The <a href="../../../reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>MOI.RootDetConeTriangle</code></a> is representable by <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> and <a href="../../../reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a> constraints, see [1, p. 149].</p><p>Indeed, <span>$t \le \det(X)^{1/n}$</span> if and only if there exists a lower triangular matrix <span>$Δ$</span> such that:</p><p class="math-container">\[\begin{align*}
+  \begin{pmatrix}
+    X &amp; Δ\\
+    Δ^\top &amp; \mathrm{Diag}(Δ)
+  \end{pmatrix} &amp; \succeq 0\\
+  (t, \mathrm{Diag}(Δ)) &amp; \in GeometricMeanCone
+\end{align*}\]</p><p><strong>Source node</strong></p><p><code>RootDetBridge</code> supports:</p><ul><li><code>I</code> in <a href="../../../reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>MOI.RootDetConeTriangle</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>RootDetBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a></li></ul><p>[1] Ben-Tal, Aharon, and Arkadi Nemirovski. <em>Lectures on modern convex     optimization: analysis, algorithms, and engineering applications</em>. Society     for Industrial and Applied Mathematics, 2001.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/det.jl#L298-L333">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.LogDetBridge" href="#MathOptInterface.Bridges.Constraint.LogDetBridge"><code>MathOptInterface.Bridges.Constraint.LogDetBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LogDetBridge{T,F,G,H,I} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p>The <a href="../../../reference/standard_form/#MathOptInterface.LogDetConeTriangle"><code>MOI.LogDetConeTriangle</code></a> is representable by <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> and <a href="../../../reference/standard_form/#MathOptInterface.ExponentialCone"><code>MOI.ExponentialCone</code></a> constraints.</p><p>Indeed, <span>$\log\det(X) = \sum\limits_{i=1}^n \log(\delta_i)$</span> where <span>$\delta_i$</span> are the eigenvalues of <span>$X$</span>.</p><p>Adapting the method from [1, p. 149], we see that <span>$t \le u \log(\det(X/u))$</span> for <span>$u &gt; 0$</span> if and only if there exists a lower triangular matrix <span>$Δ$</span> such that</p><p class="math-container">\[\begin{align*}
+  \begin{pmatrix}
+    X &amp; Δ\\
+    Δ^\top &amp; \mathrm{Diag}(Δ)
+  \end{pmatrix} &amp; \succeq 0\\
+  t - \sum_{i=1}^n u \log\left(\frac{Δ_{ii}}{u}\right) &amp; \le 0
+\end{align*}\]</p><p>Which we reformulate further into</p><p class="math-container">\[\begin{align*}
+  \begin{pmatrix}
+    X &amp; Δ\\
+    Δ^\top &amp; \mathrm{Diag}(Δ)
+  \end{pmatrix} &amp; \succeq 0\\
+  (l_i, u , Δ_{ii}) &amp; \in ExponentialCone\quad \forall i \\
+  t - \sum_{i=1}^n l_i &amp; \le 0
+\end{align*}\]</p><p><strong>Source node</strong></p><p><code>LogDetBridge</code> supports:</p><ul><li><code>I</code> in <a href="../../../reference/standard_form/#MathOptInterface.LogDetConeTriangle"><code>MOI.LogDetConeTriangle</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>LogDetBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li><li><code>G</code> in <a href="../../../reference/standard_form/#MathOptInterface.ExponentialCone"><code>MOI.ExponentialCone</code></a></li><li><code>H</code> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul><p>[1] Ben-Tal, Aharon, and Arkadi Nemirovski. <em>Lectures on modern convex     optimization: analysis, algorithms, and engineering applications</em>. Society     for Industrial and Applied Mathematics, 2001.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/det.jl#L51-L102">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.IndicatorActiveOnFalseBridge" href="#MathOptInterface.Bridges.Constraint.IndicatorActiveOnFalseBridge"><code>MathOptInterface.Bridges.Constraint.IndicatorActiveOnFalseBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndicatorActiveOnFalseBridge{T,F,S} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>IndicatorActiveOnFalseBridge</code> implements the following reformulation:</p><ul><li><span>$\neg z \implies {f(x) \in S}$</span> into <span>$y \implies {f(x) \in S}$</span>, <span>$z + y = 1$</span>, and <span>$y \in \{0, 1\}$</span></li></ul><p><strong>Source node</strong></p><p><code>IndicatorActiveOnFalseBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Indicator"><code>MOI.Indicator{MOI.ACTIVATE_ON_ZERO,S}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>IndicatorActiveOnFalseBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Indicator"><code>MOI.Indicator{MOI.ACTIVATE_ON_ONE,S}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/indicator_activate_on_zero.jl#L7-L31">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.IndicatorGreaterToLessThanBridge" href="#MathOptInterface.Bridges.Constraint.IndicatorGreaterToLessThanBridge"><code>MathOptInterface.Bridges.Constraint.IndicatorGreaterToLessThanBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndicatorGreaterToLessThanBridge{T,A} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>IndicatorGreaterToLessThanBridge</code> implements the following reformulation:</p><ul><li><span>$z \implies {f(x) \ge l}$</span> into <span>$z \implies {-f(x) \le -l}$</span></li></ul><p><strong>Source node</strong></p><p><code>IndicatorGreaterToLessThanBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Indicator"><code>MOI.Indicator{A,MOI.GreaterThan{T}}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>IndicatorGreaterToLessThanBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Indicator"><code>MOI.Indicator{A,MOI.LessThan{T}}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/indicator_flipsign.jl#L115-L135">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.IndicatorLessToGreaterThanBridge" href="#MathOptInterface.Bridges.Constraint.IndicatorLessToGreaterThanBridge"><code>MathOptInterface.Bridges.Constraint.IndicatorLessToGreaterThanBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndicatorLessToGreaterThanBridge{T,A} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>IndicatorLessToGreaterThanBridge</code> implements the following reformulations:</p><ul><li><span>$z \implies {f(x) \le u}$</span> into <span>$z \implies {-f(x) \ge -u}$</span></li></ul><p><strong>Source node</strong></p><p><code>IndicatorLessToGreaterThanBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Indicator"><code>MOI.Indicator{A,MOI.LessThan{T}}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>IndicatorLessToGreaterThanBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Indicator"><code>MOI.Indicator{A,MOI.GreaterThan{T}}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/indicator_flipsign.jl#L151-L171">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.IndicatorSOS1Bridge" href="#MathOptInterface.Bridges.Constraint.IndicatorSOS1Bridge"><code>MathOptInterface.Bridges.Constraint.IndicatorSOS1Bridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndicatorSOS1Bridge{T,S} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>IndicatorSOS1Bridge</code> implements the following reformulation:</p><ul><li><span>$z \implies {f(x) \in S}$</span> into <span>$f(x) + y \in S$</span>, <span>$SOS1(y, z)$</span></li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This bridge assumes that the solver supports <a href="../../../reference/standard_form/#MathOptInterface.SOS1"><code>MOI.SOS1{T}</code></a> constraints in which one of the variables (<span>$y$</span>) is continuous.</p></div></div><p><strong>Source node</strong></p><p><code>IndicatorSOS1Bridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Indicator"><code>MOI.Indicator{MOI.ACTIVATE_ON_ONE,S}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>IndicatorSOS1Bridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <code>S</code></li><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.SOS1"><code>MOI.SOS1{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/indicator_sos.jl#L7-L32">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SemiToBinaryBridge" href="#MathOptInterface.Bridges.Constraint.SemiToBinaryBridge"><code>MathOptInterface.Bridges.Constraint.SemiToBinaryBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SemiToBinaryBridge{T,S} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>SemiToBinaryBridge</code> implements the following reformulations:</p><ul><li><span>$x \in \{0\} \cup [l, u]$</span> into<p class="math-container">\[\begin{aligned}
+x \leq z u \\
+x \geq z l \\
+z \in \{0, 1\}
+\end{aligned}\]</p></li><li><span>$x \in \{0\} \cup \{l, \ldots, u\}$</span> into<p class="math-container">\[\begin{aligned}
+x \leq z u \\
+x \geq z l \\
+z \in \{0, 1\} \\
+x \in \mathbb{Z}
+\end{aligned}\]</p></li></ul><p><strong>Source node</strong></p><p><code>SemiToBinaryBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Semicontinuous"><code>MOI.Semicontinuous{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>SemiToBinaryBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a> (if <code>S</code> is <a href="../../../reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/semi_to_binary.jl#L7-L46">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ZeroOneBridge" href="#MathOptInterface.Bridges.Constraint.ZeroOneBridge"><code>MathOptInterface.Bridges.Constraint.ZeroOneBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ZeroOneBridge{T} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>ZeroOneBridge</code> implements the following reformulation:</p><ul><li><span>$x \in \{0, 1\}$</span> into <span>$x \in \mathbb{Z}$</span>, <span>$1x \in [0, 1]$</span>.</li></ul><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p><code>ZeroOneBridge</code> adds a linear constraint instead of adding variable bounds to avoid conflicting with bounds set by the user.</p></div></div><p><strong>Source node</strong></p><p><code>ZeroOneBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>ZeroOneBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/zero_one.jl#L7-L30">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.IntegerToZeroOneBridge" href="#MathOptInterface.Bridges.Constraint.IntegerToZeroOneBridge"><code>MathOptInterface.Bridges.Constraint.IntegerToZeroOneBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IntegerToZeroOneBridge{T} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>IntegerToZeroOneBridge</code> implements the following reformulation:</p><ul><li><span>$x \in \mathbf{Z}$</span> into <span>$y_i \in \{0, 1\}$</span>, <span>$x == lb + \sum 2^{i-1} y_i$</span>.</li></ul><p><strong>Source node</strong></p><p><code>IntegerToZeroOneBridge</code> supports:</p><ul><li><code>VariableIndex</code> in <a href="../../../reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>IntegerToZeroOneBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul><p><strong>Developer note</strong></p><p>This bridge is implemented as a constraint bridge instead of a variable bridge because we don&#39;t want to substitute the linear combination of <code>y</code> for every instance of <code>x</code>. Doing so would be expensive and greatly reduce the sparsity of the constraints.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/integer_to_zeroone.jl#L7-L34">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.NumberConversionBridge" href="#MathOptInterface.Bridges.Constraint.NumberConversionBridge"><code>MathOptInterface.Bridges.Constraint.NumberConversionBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NumberConversionBridge{T,F1,S1,F2,S2} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>NumberConversionBridge</code> implements the following reformulation:</p><ul><li><span>$f1(x) \in S1$</span> to <span>$f2(x) \in S2$</span></li></ul><p>where <code>f</code> and <code>S</code> are the same functional form, but differ in their coefficient type.</p><p><strong>Source node</strong></p><p><code>NumberConversionBridge</code> supports:</p><ul><li><code>F1</code> in <code>S1</code></li></ul><p><strong>Target node</strong></p><p><code>NumberConversionBridge</code> creates:</p><ul><li><code>F2</code> in <code>S2</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/number_conversion.jl#L7-L28">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.AllDifferentToCountDistinctBridge" href="#MathOptInterface.Bridges.Constraint.AllDifferentToCountDistinctBridge"><code>MathOptInterface.Bridges.Constraint.AllDifferentToCountDistinctBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AllDifferentToCountDistinctBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>AllDifferentToCountDistinctBridge</code> implements the following reformulations:</p><ul><li><span>$x \in \textsf{AllDifferent}(d)$</span> to <span>$(n, x) \in \textsf{CountDistinct}(1+d)$</span> and <span>$n = d$</span></li><li><span>$f(x) \in \textsf{AllDifferent}(d)$</span> to <span>$(d, f(x)) \in \textsf{CountDistinct}(1+d)$</span></li></ul><p><strong>Source node</strong></p><p><code>AllDifferentToCountDistinctBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.AllDifferent"><code>MOI.AllDifferent</code></a></li></ul><p>where <code>F</code> is <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a>.</p><p><strong>Target nodes</strong></p><p><code>AllDifferentToCountDistinctBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.CountDistinct"><code>MOI.CountDistinct</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/all_different.jl#L7-L31">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge" href="#MathOptInterface.Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge"><code>MathOptInterface.Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ReifiedAllDifferentToCountDistinctBridge{T,F} &lt;:
+Bridges.Constraint.AbstractBridge</code></pre><p><code>ReifiedAllDifferentToCountDistinctBridge</code> implements the following reformulations:</p><ul><li><span>$r \iff x \in \textsf{AllDifferent}(d)$</span> to <span>$r \iff (n, x) \in \textsf{CountDistinct}(1+d)$</span> and <span>$n = d$</span></li><li><span>$r \iff f(x) \in \textsf{AllDifferent}(d)$</span> to <span>$r \iff (d, f(x)) \in \textsf{CountDistinct}(1+d)$</span></li></ul><p><strong>Source node</strong></p><p><code>ReifiedAllDifferentToCountDistinctBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Reified"><code>MOI.Reified{MOI.AllDifferent}</code></a></li></ul><p>where <code>F</code> is <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a>.</p><p><strong>Target nodes</strong></p><p><code>ReifiedAllDifferentToCountDistinctBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Reified"><code>MOI.Reified{MOI.CountDistinct}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/all_different_reif.jl#L7-L34">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.BinPackingToMILPBridge" href="#MathOptInterface.Bridges.Constraint.BinPackingToMILPBridge"><code>MathOptInterface.Bridges.Constraint.BinPackingToMILPBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">BinPackingToMILPBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>BinPackingToMILPBridge</code> implements the following reformulation:</p><ul><li><span>$x \in BinPacking(c, w)$</span> into a mixed-integer linear program.</li></ul><p><strong>Reformulation</strong></p><p>The reformulation is non-trivial, and it depends on the finite domain of each variable <span>$x_i$</span>, which we as define <span>$S_i = \{l_i,\ldots,u_i\}$</span>.</p><p>First, we introduce new binary variables <span>$z_{ij}$</span>, which are <span>$1$</span> if variable <span>$x_i$</span> takes the value <span>$j$</span> in the optimal solution and <span>$0$</span> otherwise:</p><p class="math-container">\[\begin{aligned}
+z_{ij} \in \{0, 1\}                              &amp; \;\; \forall i \in 1\ldots d, j \in S_i  \\
+x_i - \sum\limits_{j\in S_i} j \cdot z_{ij} = 0 &amp; \;\; \forall i \in 1\ldots d              \\
+\sum\limits_{j\in S_i} z_{ij} = 1                &amp; \;\; \forall i \in 1\ldots d              \\
+\end{aligned}\]</p><p>Then, we add the capacity constraint for all possible bins <span>$j$</span>:</p><p class="math-container">\[\sum\limits_{i} w_i z_{ij} \le c \forall j \in \bigcup_{i=1,\ldots,d} S_i\]</p><p><strong>Source node</strong></p><p><code>BinPackingToMILPBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.BinPacking"><code>MOI.BinPacking{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>BinPackingToMILPBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/bin_packing.jl#L7-L47">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.CircuitToMILPBridge" href="#MathOptInterface.Bridges.Constraint.CircuitToMILPBridge"><code>MathOptInterface.Bridges.Constraint.CircuitToMILPBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CircuitToMILPBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>CircuitToMILPBridge</code> implements the following reformulation:</p><ul><li><span>$x \in \textsf{Circuit}(d)$</span> to the Miller-Tucker-Zemlin formulation of the Traveling Salesperson Problem.</li></ul><p><strong>Source node</strong></p><p><code>CircuitToMILPBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Circuit"><code>MOI.Circuit</code></a></li></ul><p>where <code>F</code> is <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a>.</p><p><strong>Target nodes</strong></p><p><code>CircuitToMILPBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/circuit.jl#L7-L33">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.CountAtLeastToCountBelongsBridge" href="#MathOptInterface.Bridges.Constraint.CountAtLeastToCountBelongsBridge"><code>MathOptInterface.Bridges.Constraint.CountAtLeastToCountBelongsBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountAtLeastToCountBelongsBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>CountAtLeastToCountBelongsBridge</code> implements the following reformulation:</p><ul><li><span>$x \in \textsf{CountAtLeast}(n, d, \mathcal{S})$</span> to <span>$(n_i, x_{d_i}) \in \textsf{CountBelongs}(1+d, \mathcal{S})$</span> and <span>$n_i \ge n$</span> for all <span>$i$</span>.</li></ul><p><strong>Source node</strong></p><p><code>CountAtLeastToCountBelongsBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.CountAtLeast"><code>MOI.CountAtLeast</code></a></li></ul><p>where <code>F</code> is <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a>.</p><p><strong>Target nodes</strong></p><p><code>CountAtLeastToCountBelongsBridge</code> creates:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.CountBelongs"><code>MOI.CountBelongs</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/count_at_least.jl#L7-L31">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.CountBelongsToMILPBridge" href="#MathOptInterface.Bridges.Constraint.CountBelongsToMILPBridge"><code>MathOptInterface.Bridges.Constraint.CountBelongsToMILPBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountBelongsToMILPBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>CountBelongsToMILPBridge</code> implements the following reformulation:</p><ul><li><span>$(n, x) \in \textsf{CountBelongs}(1+d, \mathcal{S})$</span> into a mixed-integer linear program.</li></ul><p><strong>Reformulation</strong></p><p>The reformulation is non-trivial, and it depends on the finite domain of each variable <span>$x_i$</span>, which we as define <span>$S_i = \{l_i,\ldots,u_i\}$</span>.</p><p>First, we introduce new binary variables <span>$z_{ij}$</span>, which are <span>$1$</span> if variable <span>$x_i$</span> takes the value <span>$j$</span> in the optimal solution and <span>$0$</span> otherwise:</p><p class="math-container">\[\begin{aligned}
+z_{ij} \in \{0, 1\}                              &amp; \;\; \forall i \in 1\ldots d, j \in S_i  \\
+x_i - \sum\limits_{j\in S_i} j \cdot z_{ij} = 0 &amp; \;\; \forall i \in 1\ldots d              \\
+\sum\limits_{j\in S_i} z_{ij} = 1                &amp; \;\; \forall i \in 1\ldots d              \\
+\end{aligned}\]</p><p>Finally, <span>$n$</span> is constrained to be the number of <span>$z_{ij}$</span> elements that are in <span>$\mathcal{S}$</span>:</p><p class="math-container">\[n - \sum\limits_{i\in 1\ldots d, j \in \mathcal{S}} z_{ij} = 0\]</p><p><strong>Source node</strong></p><p><code>CountBelongsToMILPBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.CountBelongs"><code>MOI.CountBelongs</code></a></li></ul><p>where <code>F</code> is <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a>.</p><p><strong>Target nodes</strong></p><p><code>CountBelongsToMILPBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/count_belongs.jl#L7-L51">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.CountDistinctToMILPBridge" href="#MathOptInterface.Bridges.Constraint.CountDistinctToMILPBridge"><code>MathOptInterface.Bridges.Constraint.CountDistinctToMILPBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountDistinctToMILPBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>CountDistinctToMILPBridge</code> implements the following reformulation:</p><ul><li><span>$(n, x) \in \textsf{CountDistinct}(1+d)$</span> into a mixed-integer linear program.</li></ul><p><strong>Reformulation</strong></p><p>The reformulation is non-trivial, and it depends on the finite domain of each variable <span>$x_i$</span>, which we as define <span>$S_i = \{l_i,\ldots,u_i\}$</span>.</p><p>First, we introduce new binary variables <span>$z_{ij}$</span>, which are <span>$1$</span> if variable <span>$x_i$</span> takes the value <span>$j$</span> in the optimal solution and <span>$0$</span> otherwise:</p><p class="math-container">\[\begin{aligned}
+z_{ij} \in \{0, 1\}                              &amp; \;\; \forall i \in 1\ldots d, j \in S_i  \\
+x_i - \sum\limits_{j\in S_i} j \cdot z_{ij} = 0 &amp; \;\; \forall i \in 1\ldots d              \\
+\sum\limits_{j\in S_i} z_{ij} = 1                &amp; \;\; \forall i \in 1\ldots d              \\
+\end{aligned}\]</p><p>Then, we introduce new binary variables <span>$y_j$</span>, which are <span>$1$</span> if a variable takes the value <span>$j$</span> in the optimal solution and <span>$0$</span> otherwise.</p><p class="math-container">\[\begin{aligned}
+y_{j} \in \{0, 1\}                        &amp; \;\; \forall j \in \bigcup_{i=1,\ldots,d} S_i \\
+y_j \le \sum\limits_{i \in 1\ldots d: j \in S_i} z_{ij} \le M y_j &amp; \;\; \forall j \in \bigcup_{i=1,\ldots,d} S_i\\
+\end{aligned}\]</p><p>Finally, <span>$n$</span> is constrained to be the number of <span>$y_j$</span> elements that are non-zero:</p><p class="math-container">\[n - \sum\limits_{j \in \bigcup_{i=1,\ldots,d} S_i} y_{j} = 0\]</p><p><strong>Source node</strong></p><p><code>CountDistinctToMILPBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.CountDistinct"><code>MOI.CountDistinct</code></a></li></ul><p>where <code>F</code> is <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a>.</p><p><strong>Target nodes</strong></p><p><code>CountDistinctToMILPBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/count_distinct.jl#L7-L60">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.ReifiedCountDistinctToMILPBridge" href="#MathOptInterface.Bridges.Constraint.ReifiedCountDistinctToMILPBridge"><code>MathOptInterface.Bridges.Constraint.ReifiedCountDistinctToMILPBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ReifiedCountDistinctToMILPBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>ReifiedCountDistinctToMILPBridge</code> implements the following reformulation:</p><ul><li><span>$r \iff (n, x) \in \textsf{CountDistinct}(1+d)$</span> into a mixed-integer linear program.</li></ul><p><strong>Reformulation</strong></p><p>The reformulation is non-trivial, and it depends on the finite domain of each variable <span>$x_i$</span>, which we as define <span>$S_i = \{l_i,\ldots,u_i\}$</span>.</p><p>First, we introduce new binary variables <span>$z_{ij}$</span>, which are <span>$1$</span> if variable <span>$x_i$</span> takes the value <span>$j$</span> in the optimal solution and <span>$0$</span> otherwise:</p><p class="math-container">\[\begin{aligned}
+z_{ij} \in \{0, 1\}                              &amp; \;\; \forall i \in 1\ldots d, j \in S_i  \\
+x_i - \sum\limits_{j\in S_i} j \cdot z_{ij} = 0 &amp; \;\; \forall i \in 1\ldots d              \\
+\sum\limits_{j\in S_i} z_{ij} = 1                &amp; \;\; \forall i \in 1\ldots d              \\
+\end{aligned}\]</p><p>Then, we introduce new binary variables <span>$y_j$</span>, which are <span>$1$</span> if a variable takes the value <span>$j$</span> in the optimal solution and <span>$0$</span> otherwise.</p><p class="math-container">\[\begin{aligned}
+y_{j} \in \{0, 1\}                        &amp; \;\; \forall j \in \bigcup_{i=1,\ldots,d} S_i \\
+y_j \le \sum\limits_{i \in 1\ldots d: j \in S_i} z_{ij} \le M y_j &amp; \;\; \forall j \in \bigcup_{i=1,\ldots,d} S_i\\
+\end{aligned}\]</p><p>Finally, <span>$n$</span> is constrained to be the number of <span>$y_j$</span> elements that are non-zero, with some slack:</p><p class="math-container">\[n - \sum\limits_{j \in \bigcup_{i=1,\ldots,d} S_i} y_{j} = \delta^+ - \delta^-\]</p><p>And then the slack is constrained to respect the reif variable <span>$r$</span>:</p><p class="math-container">\[\begin{aligned}
+d_1 \le \delta^+ \le M d_1 \\
+d_2 \le \delta^- \le M d_s \\
+d_1 + d_2 + r = 1             \\
+d_1, d_2 \in \{0, 1\}
+\end{aligned}\]</p><p><strong>Source node</strong></p><p><code>ReifiedCountDistinctToMILPBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Reified"><code>MOI.Reified{MOI.CountDistinct}</code></a></li></ul><p>where <code>F</code> is <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{T}</code></a>.</p><p><strong>Target nodes</strong></p><p><code>ReifiedCountDistinctToMILPBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/count_distinct_reif.jl#L7-L70">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.CountGreaterThanToMILPBridge" href="#MathOptInterface.Bridges.Constraint.CountGreaterThanToMILPBridge"><code>MathOptInterface.Bridges.Constraint.CountGreaterThanToMILPBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CountGreaterThanToMILPBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>CountGreaterThanToMILPBridge</code> implements the following reformulation:</p><ul><li><span>$(c, y, x) \in CountGreaterThan()$</span> into a mixed-integer linear program.</li></ul><p><strong>Source node</strong></p><p><code>CountGreaterThanToMILPBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.CountGreaterThan"><code>MOI.CountGreaterThan</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>CountGreaterThanToMILPBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/count_greater_than.jl#L7-L27">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.TableToMILPBridge" href="#MathOptInterface.Bridges.Constraint.TableToMILPBridge"><code>MathOptInterface.Bridges.Constraint.TableToMILPBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">TableToMILPBridge{T,F} &lt;: Bridges.Constraint.AbstractBridge</code></pre><p><code>TableToMILPBridge</code> implements the following reformulation:</p><ul><li><span>$x \in Table(t)$</span> into<p class="math-container">\[\begin{aligned}
+z_{j} \in \{0, 1\}                     &amp; \quad \forall i, j \\
+\sum\limits_{j=1}^n z_{j} = 1                           \\
+\sum\limits_{j=1}^n t_{ij} z_{j} = x_i &amp; \quad \forall i
+\end{aligned}\]</p></li></ul><p><strong>Source node</strong></p><p><code>TableToMILPBridge</code> supports:</p><ul><li><code>F</code> in <a href="../../../reference/standard_form/#MathOptInterface.Table"><code>MOI.Table{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>TableToMILPBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/table.jl#L7-L33">source</a></section></article><h2 id="objective_bridges_ref"><a class="docs-heading-anchor" href="#objective_bridges_ref">Objective bridges</a><a id="objective_bridges_ref-1"></a><a class="docs-heading-anchor-permalink" href="#objective_bridges_ref" title="Permalink"></a></h2><p>These bridges are subtypes of <a href="../reference/#MathOptInterface.Bridges.Objective.AbstractBridge"><code>Bridges.Objective.AbstractBridge</code></a>.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.FunctionizeBridge" href="#MathOptInterface.Bridges.Objective.FunctionizeBridge"><code>MathOptInterface.Bridges.Objective.FunctionizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">FunctionizeBridge{T}</code></pre><p><code>FunctionizeBridge</code> implements the following reformulations:</p><ul><li><span>$\min \{x\}$</span> into <span>$\min\{1x + 0\}$</span></li><li><span>$\max \{x\}$</span> into <span>$\max\{1x + 0\}$</span></li></ul><p>where <code>T</code> is the coefficient type of <code>1</code> and <code>0</code>.</p><p><strong>Source node</strong></p><p><code>FunctionizeBridge</code> supports:</p><ul><li><a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>FunctionizeBridge</code> creates:</p><ul><li>One objective node: <a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{T}}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridges/functionize.jl#L7-L28">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.QuadratizeBridge" href="#MathOptInterface.Bridges.Objective.QuadratizeBridge"><code>MathOptInterface.Bridges.Objective.QuadratizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">QuadratizeBridge{T}</code></pre><p><code>QuadratizeBridge</code> implements the following reformulations:</p><ul><li><span>$\min \{a^\top x + b\}$</span> into <span>$\min\{x^\top \mathbf{0} x + a^\top x + b\}$</span></li><li><span>$\max \{a^\top x + b\}$</span> into <span>$\max\{x^\top \mathbf{0} x + a^\top x + b\}$</span></li></ul><p>where <code>T</code> is the coefficient type of <code>0</code>.</p><p><strong>Source node</strong></p><p><code>QuadratizeBridge</code> supports:</p><ul><li><a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{T}}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>QuadratizeBridge</code> creates:</p><ul><li>One objective node: <a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{T}}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridges/quadratize.jl#L7-L28">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.SlackBridge" href="#MathOptInterface.Bridges.Objective.SlackBridge"><code>MathOptInterface.Bridges.Objective.SlackBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SlackBridge{T,F,G}</code></pre><p><code>SlackBridge</code> implements the following reformulations:</p><ul><li><span>$\min\{f(x)\}$</span> into <span>$\min\{y\;|\; f(x) - y \le 0\}$</span></li><li><span>$\max\{f(x)\}$</span> into <span>$\max\{y\;|\; f(x) - y \ge 0\}$</span></li></ul><p>where <code>F</code> is the type of <code>f(x) - y</code>, <code>G</code> is the type of <code>f(x)</code>, and <code>T</code> is the coefficient type of <code>f(x)</code>.</p><p><strong>Source node</strong></p><p><code>SlackBridge</code> supports:</p><ul><li><a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{G}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>SlackBridge</code> creates:</p><ul><li>One variable node: <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li><li>One objective node: <a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li><li>One constraint node, that depends on the <a href="../../../reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense</code></a>:<ul><li><code>F</code>-in-<a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan</code></a> if <code>MIN_SENSE</code></li><li><code>F</code>-in-<a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a> if <code>MAX_SENSE</code></li></ul></li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>When using this bridge, changing the optimization sense is not supported. Set the sense to <code>MOI.FEASIBILITY_SENSE</code> first to delete the bridge, then set <a href="../../../reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense</code></a> and re-add the objective.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridges/slack.jl#L7-L38">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.VectorFunctionizeBridge" href="#MathOptInterface.Bridges.Objective.VectorFunctionizeBridge"><code>MathOptInterface.Bridges.Objective.VectorFunctionizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorFunctionizeBridge{T}</code></pre><p><code>VectorFunctionizeBridge</code> implements the following reformulations:</p><ul><li><span>$\min \{x\}$</span> into <span>$\min\{1x + 0\}$</span></li><li><span>$\max \{x\}$</span> into <span>$\max\{1x + 0\}$</span></li></ul><p>where <code>T</code> is the coefficient type of <code>1</code> and <code>0</code>.</p><p><strong>Source node</strong></p><p><code>VectorFunctionizeBridge</code> supports:</p><ul><li><a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VectorOfVariables}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>VectorFunctionizeBridge</code> creates:</p><ul><li>One objective node: <a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VectorAffineFunction{T}}</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridges/vector_functionize.jl#L7-L28">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.VectorSlackBridge" href="#MathOptInterface.Bridges.Objective.VectorSlackBridge"><code>MathOptInterface.Bridges.Objective.VectorSlackBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorSlackBridge{T,F,G}</code></pre><p><code>VectorSlackBridge</code> implements the following reformulations:</p><ul><li><span>$\min\{f(x)\}$</span> into <span>$\min\{y\;|\; y - f(x) \in \mathbb{R}_+ \}$</span></li><li><span>$\max\{f(x)\}$</span> into <span>$\max\{y\;|\; f(x) - y \in \mathbb{R}_+ \}$</span></li></ul><p>where <code>F</code> is the type of <code>f(x) - y</code>, <code>G</code> is the type of <code>f(x)</code>, and <code>T</code> is the coefficient type of <code>f(x)</code>.</p><p><strong>Source node</strong></p><p><code>VectorSlackBridge</code> supports:</p><ul><li><a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{G}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>VectorSlackBridge</code> creates:</p><ul><li>One variable node: <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li><li>One objective node: <a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VectorOfVariables}</code></a></li><li>One constraint node: <code>F</code>-in-<a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>When using this bridge, changing the optimization sense is not supported. Set the sense to <code>MOI.FEASIBILITY_SENSE</code> first to delete the bridge, then set <a href="../../../reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense</code></a> and re-add the objective.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridges/vector_slack.jl#L7-L36">source</a></section></article><h2 id="variable_bridges_ref"><a class="docs-heading-anchor" href="#variable_bridges_ref">Variable bridges</a><a id="variable_bridges_ref-1"></a><a class="docs-heading-anchor-permalink" href="#variable_bridges_ref" title="Permalink"></a></h2><p>These bridges are subtypes of <a href="../reference/#MathOptInterface.Bridges.Variable.AbstractBridge"><code>Bridges.Variable.AbstractBridge</code></a>.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.FreeBridge" href="#MathOptInterface.Bridges.Variable.FreeBridge"><code>MathOptInterface.Bridges.Variable.FreeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">FreeBridge{T} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>FreeBridge</code> implements the following reformulation:</p><ul><li><span>$x \in \mathbb{R}$</span> into <span>$y, z \ge 0$</span> with the substitution rule <span>$x = y - z$</span>,</li></ul><p>where <code>T</code> is the coefficient type of <code>y - z</code>.</p><p><strong>Source node</strong></p><p><code>FreeBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>FreeBridge</code> creates:</p><ul><li>One variable node: <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/free.jl#L7-L28">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.NonposToNonnegBridge" href="#MathOptInterface.Bridges.Variable.NonposToNonnegBridge"><code>MathOptInterface.Bridges.Variable.NonposToNonnegBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NonposToNonnegBridge{T} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>NonposToNonnegBridge</code> implements the following reformulation:</p><ul><li><span>$x \in \mathbb{R}_-$</span> into <span>$y \in \mathbb{R}_+$</span> with the substitution rule <span>$x = -y$</span>,</li></ul><p>where <code>T</code> is the coefficient type of <code>-y</code>.</p><p><strong>Source node</strong></p><p><code>NonposToNonnegBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>NonposToNonnegBridge</code> creates:</p><ul><li>One variable node: <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a>,</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/flip_sign.jl#L47-L68">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.RSOCtoPSDBridge" href="#MathOptInterface.Bridges.Variable.RSOCtoPSDBridge"><code>MathOptInterface.Bridges.Variable.RSOCtoPSDBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RSOCtoPSDBridge{T} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>RSOCtoPSDBridge</code> implements the following reformulation:</p><ul><li><span>$||x||_2^2 \le 2tu$</span> where <span>$t, u \ge 0$</span> into <span>$Y \succeq 0$</span>, with the substitution rule: <span>$Y = \left[\begin{array}{c c}t &amp; x^\top \\ x &amp; 2u \mathbf{I}\end{array}\right].$</span></li></ul><p>Additional bounds are added to ensure the off-diagonals of the <span>$2uI$</span> submatrix are <code>0</code>, and linear constraints are added to ennsure the diagonal of <span>$2uI$</span> takes the same values.</p><p>As a special case, if <span>$|x|| = 0$</span>, then <code>RSOCtoPSDBridge</code> reformulates into <span>$(t, u) \in \mathbb{R}_+$</span>.</p><p><strong>Source node</strong></p><p><code>RSOCtoPSDBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>RSOCtoPSDBridge</code> creates:</p><ul><li>One variable node that depends on the input dimension:<ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a> if dimension is <code>1</code> or <code>2</code></li><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in</li></ul><a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> otherwise</li><li>The constraint node <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a></li><li>The constrant node <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/rsoc_to_psd.jl#L7-L40">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.RSOCtoSOCBridge" href="#MathOptInterface.Bridges.Variable.RSOCtoSOCBridge"><code>MathOptInterface.Bridges.Variable.RSOCtoSOCBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RSOCtoSOCBridge{T} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>RSOCtoSOCBridge</code> implements the following reformulation:</p><ul><li><span>$||x||_2^2 \le 2tu$</span> into <span>$||v||_2 \le w$</span>, with the substitution rules <span>$t = \frac{w}{\sqrt 2} + \frac{v_1}{\sqrt 2}$</span>, <span>$u = \frac{w}{\sqrt 2} - \frac{v_1}{\sqrt 2}$</span>, and <span>$x = (v_2,\ldots,v_N)$</span>.</li></ul><p><strong>Source node</strong></p><p><code>RSOCtoSOCBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul><p><strong>Target node</strong></p><p><code>RSOCtoSOCBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/rsoc_soc.jl#L7-L28">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.SOCtoRSOCBridge" href="#MathOptInterface.Bridges.Variable.SOCtoRSOCBridge"><code>MathOptInterface.Bridges.Variable.SOCtoRSOCBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SOCtoRSOCBridge{T} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>SOCtoRSOCBridge</code> implements the following reformulation:</p><ul><li><span>$||x||_2 \le t$</span> into <span>$2uv \ge ||w||_2^2$</span>, with the substitution rules <span>$t = \frac{u}{\sqrt 2} + \frac{v}{\sqrt 2}$</span>, <span>$x = (\frac{u}{\sqrt 2} - \frac{v}{\sqrt 2}, w)$</span>.</li></ul><p><strong>Assumptions</strong></p><ul><li><code>SOCtoRSOCBridge</code> assumes that <span>$|x| \ge 1$</span>.</li></ul><p><strong>Source node</strong></p><p><code>SOCtoRSOCBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p><strong>Target node</strong></p><p><code>SOCtoRSOCBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/soc_rsoc.jl#L7-L31">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.VectorizeBridge" href="#MathOptInterface.Bridges.Variable.VectorizeBridge"><code>MathOptInterface.Bridges.Variable.VectorizeBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VectorizeBridge{T,S} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>VectorizeBridge</code> implements the following reformulations:</p><ul><li><span>$x \ge a$</span> into <span>$[y] \in \mathbb{R}_+$</span> with the substitution rule <span>$x = a + y$</span></li><li><span>$x \le a$</span> into <span>$[y] \in \mathbb{R}_-$</span> with the substitution rule <span>$x = a + y$</span></li><li><span>$x == a$</span> into <span>$[y] \in \{0\}$</span> with the substitution rule <span>$x = a + y$</span></li></ul><p>where <code>T</code> is the coefficient type of <code>a + y</code>.</p><p><strong>Source node</strong></p><p><code>VectorizeBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>VectorizeBridge</code> creates:</p><ul><li>One variable node: <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <code>S</code>, where <code>S</code> is one of <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.Nonpositives"><code>MOI.Nonpositives</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a> depending on the type of <span>$S$</span>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/vectorize.jl#L7-L36">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.ZerosBridge" href="#MathOptInterface.Bridges.Variable.ZerosBridge"><code>MathOptInterface.Bridges.Variable.ZerosBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ZerosBridge{T} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>ZerosBridge</code> implements the following reformulation:</p><ul><li><span>$x \in \{0\}$</span> into the substitution rule <span>$x = 0$</span>,</li></ul><p>where <code>T</code> is the coefficient type of <code>0</code>.</p><p><strong>Source node</strong></p><p><code>ZerosBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li></ul><p><strong>Target nodes</strong></p><p><code>ZerosBridge</code> does not create target nodes. It replaces all instances of <code>x</code> with <code>0</code> via substitution. This means that no variables are created in the underlying model.</p><p><strong>Caveats</strong></p><p>The bridged variables are similar to parameters with zero values. Parameters with non-zero values can be created with constrained variables in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a> by combining a <a href="#MathOptInterface.Bridges.Variable.VectorizeBridge"><code>VectorizeBridge</code></a> and this bridge.</p><p>However, functions modified by <code>ZerosBridge</code> cannot be unbridged. That is, for a given function, we cannot determine if the bridged variables were used.</p><p>A related implication is that this bridge does not support <a href="../../../reference/constraints/#MathOptInterface.ConstraintDual"><code>MOI.ConstraintDual</code></a>. However, if a <a href="../../Utilities/reference/#MathOptInterface.Utilities.CachingOptimizer"><code>MOI.Utilities.CachingOptimizer</code></a> is used, the dual can be determined by the bridged optimizer using <a href="../../Utilities/reference/#MathOptInterface.Utilities.get_fallback"><code>MOI.Utilities.get_fallback</code></a> because the caching optimizer records the unbridged function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/zeros.jl#L7-L42">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.HermitianToSymmetricPSDBridge" href="#MathOptInterface.Bridges.Variable.HermitianToSymmetricPSDBridge"><code>MathOptInterface.Bridges.Variable.HermitianToSymmetricPSDBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">HermitianToSymmetricPSDBridge{T} &lt;: Bridges.Variable.AbstractBridge</code></pre><p><code>HermitianToSymmetricPSDBridge</code> implements the following reformulation:</p><ul><li>Hermitian positive semidefinite <code>n x n</code> complex matrix to a symmetric positive semidefinite <code>2n x 2n</code> real matrix satisfying equality constraints described below.</li></ul><p><strong>Source node</strong></p><p><code>HermitianToSymmetricPSDBridge</code> supports:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle"><code>MOI.HermitianPositiveSemidefiniteConeTriangle</code></a></li></ul><p><strong>Target node</strong></p><p><code>HermitianToSymmetricPSDBridge</code> creates:</p><ul><li><a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li><li><a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li></ul><p><strong>Reformulation</strong></p><p>The reformulation is best described by example.</p><p>The Hermitian matrix:</p><p class="math-container">\[\begin{bmatrix}
+  x_{11}            &amp; x_{12} + y_{12}im &amp; x_{13} + y_{13}im\\
+  x_{12} - y_{12}im &amp; x_{22}            &amp; x_{23} + y_{23}im\\
+  x_{13} - y_{13}im &amp; x_{23} - y_{23}im &amp; x_{33}
+\end{bmatrix}\]</p><p>is positive semidefinite if and only if the symmetric matrix:</p><p class="math-container">\[\begin{bmatrix}
+    x_{11} &amp; x_{12} &amp; x_{13} &amp; 0       &amp; y_{12}  &amp; y_{13} \\
+           &amp; x_{22} &amp; x_{23} &amp; -y_{12} &amp; 0       &amp; y_{23} \\
+           &amp;        &amp; x_{33} &amp; -y_{13} &amp; -y_{23} &amp; 0      \\
+           &amp;        &amp;        &amp; x_{11}  &amp; x_{12}  &amp; x_{13} \\
+           &amp;        &amp;        &amp;         &amp; x_{22}  &amp; x_{23} \\
+           &amp;        &amp;        &amp;         &amp;         &amp; x_{33}
+\end{bmatrix}\]</p><p>is positive semidefinite.</p><p>The bridge achieves this reformulation by adding a new set of variables in <code>MOI.PositiveSemidefiniteConeTriangle(6)</code>, and then adding three groups of equality constraints to:</p><ul><li>constrain the two <code>x</code> blocks to be equal</li><li>force the diagonal of the <code>y</code> blocks to be <code>0</code></li><li>force the lower triangular of the <code>y</code> block to be the negative of the upper triangle.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/hermitian.jl#L7-L63">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../overview/">« Overview</a><a class="docs-footer-nextpage" href="../reference/">API Reference »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Bridges/overview/index.html b/previews/PR3525/moi/submodules/Bridges/overview/index.html
new file mode 100644
index 00000000000..ff14a0cc06d
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Bridges/overview/index.html
@@ -0,0 +1,57 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Overview · JuMP</title><meta name="title" content="Overview · JuMP"/><meta property="og:title" content="Overview · JuMP"/><meta property="twitter:title" content="Overview · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Overview</a><ul class="internal"><li><a class="tocitem" href="#Why-bridges?"><span>Why bridges?</span></a></li><li><a class="tocitem" href="#The-three-types-of-bridges"><span>The three types of bridges</span></a></li><li><a class="tocitem" href="#Bridges.full_bridge_optimizer"><span>Bridges.<code>full_bridge_optimizer</code></span></a></li><li><a class="tocitem" href="#Add-a-single-bridge"><span>Add a single bridge</span></a></li><li><a class="tocitem" href="#Bridges.LazyBridgeOptimizer"><span>Bridges.LazyBridgeOptimizer</span></a></li></ul></li><li><a class="tocitem" href="../list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Bridges</a></li><li class="is-active"><a href>Overview</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Overview</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Bridges/overview.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-Bridges-submodule"><a class="docs-heading-anchor" href="#The-Bridges-submodule">The Bridges submodule</a><a id="The-Bridges-submodule-1"></a><a class="docs-heading-anchor-permalink" href="#The-Bridges-submodule" title="Permalink"></a></h1><p>The <code>Bridges</code> module simplifies the process of converting models between equivalent formulations.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p><a href="https://arxiv.org/abs/2002.03447">Read our paper</a> for more details on how bridges are implemented.</p></div></div><h2 id="Why-bridges?"><a class="docs-heading-anchor" href="#Why-bridges?">Why bridges?</a><a id="Why-bridges?-1"></a><a class="docs-heading-anchor-permalink" href="#Why-bridges?" title="Permalink"></a></h2><p>A constraint can often be written in a number of equivalent formulations. For example, the constraint <span>$l \le a^\top x \le u$</span> (<a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>Interval</code></a>) could be re-formulated as two constraints: <span>$a^\top x \ge l$</span> (<a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a>) and <span>$a^\top x \le u$</span> (<code>ScalarAffineFunction</code>-in-<code>LessThan</code>). An alternative re-formulation is to add a dummy variable <code>y</code> with the constraints <span>$l \le y \le u$</span> (<a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>Interval</code></a>) and <span>$a^\top x - y = 0$</span> (<a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>EqualTo</code></a>).</p><p>To avoid each solver having to code these transformations manually, MathOptInterface provides <em>bridges</em>.</p><p>A bridge is a small transformation from one constraint type to another (potentially collection of) constraint type.</p><p>Because these bridges are included in MathOptInterface, they can be re-used by any optimizer. Some bridges also implement constraint modifications and constraint primal and dual translations.</p><p>Several bridges can be used in combination to transform a single constraint into a form that the solver may understand. Choosing the bridges to use takes the form of finding a shortest path in the hyper-graph of bridges. The methodology is detailed in <a href="https://arxiv.org/abs/2002.03447">the MOI paper</a>.</p><h2 id="The-three-types-of-bridges"><a class="docs-heading-anchor" href="#The-three-types-of-bridges">The three types of bridges</a><a id="The-three-types-of-bridges-1"></a><a class="docs-heading-anchor-permalink" href="#The-three-types-of-bridges" title="Permalink"></a></h2><p>There are three types of bridges in MathOptInterface:</p><ol><li>Constraint bridges</li><li>Variable bridges</li><li>Objective bridges</li></ol><h3 id="Constraint-bridges"><a class="docs-heading-anchor" href="#Constraint-bridges">Constraint bridges</a><a id="Constraint-bridges-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-bridges" title="Permalink"></a></h3><p>Constraint bridges convert constraints formulated by the user into an equivalent form supported by the solver. Constraint bridges are subtypes of <a href="../reference/#MathOptInterface.Bridges.Constraint.AbstractBridge"><code>Bridges.Constraint.AbstractBridge</code></a>.</p><p>The equivalent formulation may add constraints (and possibly also variables) in the underlying model.</p><p>In particular, constraint bridges can focus on rewriting the function of a constraint, and do not change the set. Function bridges are subtypes of <a href="../reference/#MathOptInterface.Bridges.Constraint.AbstractFunctionConversionBridge"><code>Bridges.Constraint.AbstractFunctionConversionBridge</code></a>.</p><p>Read the <a href="../list_of_bridges/#constraint_bridges_ref">list of implemented constraint bridges</a> for more details on the types of transformations that are available. Function bridges are <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.ScalarFunctionizeBridge"><code>Bridges.Constraint.ScalarFunctionizeBridge</code></a> and <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.VectorFunctionizeBridge"><code>Bridges.Constraint.VectorFunctionizeBridge</code></a>.</p><h3 id="variable_bridges"><a class="docs-heading-anchor" href="#variable_bridges">Variable bridges</a><a id="variable_bridges-1"></a><a class="docs-heading-anchor-permalink" href="#variable_bridges" title="Permalink"></a></h3><p>Variable bridges convert variables added by the user, either free with <a href="../../../reference/variables/#MathOptInterface.add_variable"><code>add_variable</code></a>/<a href="../../../reference/variables/#MathOptInterface.add_variables"><code>add_variables</code></a>, or constrained with <a href="../../../reference/variables/#MathOptInterface.add_constrained_variable"><code>add_constrained_variable</code></a>/<a href="../../../reference/variables/#MathOptInterface.add_constrained_variables"><code>add_constrained_variables</code></a>, into an equivalent form supported by the solver. Variable bridges are subtypes of <a href="../reference/#MathOptInterface.Bridges.Variable.AbstractBridge"><code>Bridges.Variable.AbstractBridge</code></a>.</p><p>The equivalent formulation may add constraints (and possibly also variables) in the underlying model.</p><p>Read the <a href="../list_of_bridges/#variable_bridges_ref">list of implemented variable bridges</a> for more details on the types of transformations that are available.</p><h3 id="Objective-bridges"><a class="docs-heading-anchor" href="#Objective-bridges">Objective bridges</a><a id="Objective-bridges-1"></a><a class="docs-heading-anchor-permalink" href="#Objective-bridges" title="Permalink"></a></h3><p>Objective bridges convert the <a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> set by the user into an equivalent form supported by the solver. Objective bridges are subtypes of <a href="../reference/#MathOptInterface.Bridges.Objective.AbstractBridge"><code>Bridges.Objective.AbstractBridge</code></a>.</p><p>The equivalent formulation may add constraints (and possibly also variables) in the underlying model.</p><p>Read the <a href="../list_of_bridges/#objective_bridges_ref">list of implemented objective bridges</a> for more details on the types of transformations that are available.</p><h2 id="Bridges.full_bridge_optimizer"><a class="docs-heading-anchor" href="#Bridges.full_bridge_optimizer">Bridges.<code>full_bridge_optimizer</code></a><a id="Bridges.full_bridge_optimizer-1"></a><a class="docs-heading-anchor-permalink" href="#Bridges.full_bridge_optimizer" title="Permalink"></a></h2><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Unless you have an advanced use-case, this is probably the only function you need to care about.</p></div></div><p>To enable the full power of MathOptInterface&#39;s bridges, wrap an <code>optimizer</code> in a <a href="../reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>Bridges.full_bridge_optimizer</code></a>.</p><pre><code class="language-julia-repl hljs">julia&gt; inner_optimizer = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; optimizer = MOI.Bridges.full_bridge_optimizer(inner_optimizer, Float64)
+MOIB.LazyBridgeOptimizer{MOIU.Model{Float64}}
+with 0 variable bridges
+with 0 constraint bridges
+with 0 objective bridges
+with inner model MOIU.Model{Float64}</code></pre><p>Now, use <code>optimizer</code> as normal, and bridging will happen lazily behind the scenes. By lazily, we mean that bridging will happen if and only if the constraint is not supported by the <code>inner_optimizer</code>.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Most bridges are added by default in <a href="../reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>Bridges.full_bridge_optimizer</code></a>. However, for technical reasons, some bridges are not added by default. Three examples include <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoPSDBridge"><code>Bridges.Constraint.SOCtoPSDBridge</code></a>, <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge"><code>Bridges.Constraint.SOCtoNonConvexQuadBridge</code></a> and <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.RSOCtoNonConvexQuadBridge"><code>Bridges.Constraint.RSOCtoNonConvexQuadBridge</code></a>. See the docs of those bridges for more information.</p></div></div><h2 id="Add-a-single-bridge"><a class="docs-heading-anchor" href="#Add-a-single-bridge">Add a single bridge</a><a id="Add-a-single-bridge-1"></a><a class="docs-heading-anchor-permalink" href="#Add-a-single-bridge" title="Permalink"></a></h2><p>If you don&#39;t want to use <a href="../reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>Bridges.full_bridge_optimizer</code></a>, you can wrap an optimizer in a single bridge.</p><p>However, this will force the constraint to be bridged, even if the <code>inner_optimizer</code> supports it.</p><pre><code class="language-julia-repl hljs">julia&gt; inner_optimizer = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; optimizer = MOI.Bridges.Constraint.SplitInterval{Float64}(inner_optimizer)
+MOIB.Constraint.SingleBridgeOptimizer{MOIB.Constraint.SplitIntervalBridge{Float64}, MOIU.Model{Float64}}
+with 0 constraint bridges
+with inner model MOIU.Model{Float64}
+
+julia&gt; x = MOI.add_variable(optimizer)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(optimizer, x, MOI.Interval(0.0, 1.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)
+
+julia&gt; MOI.get(optimizer, MOI.ListOfConstraintTypesPresent())
+1-element Vector{Tuple{Type, Type}}:
+ (MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64})
+
+julia&gt; MOI.get(inner_optimizer, MOI.ListOfConstraintTypesPresent())
+2-element Vector{Tuple{Type, Type}}:
+ (MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64})
+ (MathOptInterface.VariableIndex, MathOptInterface.LessThan{Float64})</code></pre><h2 id="Bridges.LazyBridgeOptimizer"><a class="docs-heading-anchor" href="#Bridges.LazyBridgeOptimizer">Bridges.LazyBridgeOptimizer</a><a id="Bridges.LazyBridgeOptimizer-1"></a><a class="docs-heading-anchor-permalink" href="#Bridges.LazyBridgeOptimizer" title="Permalink"></a></h2><p>If you don&#39;t want to use <a href="../reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>Bridges.full_bridge_optimizer</code></a>, but you need more than a single bridge (or you want the bridging to happen lazily), you can manually construct a <a href="../reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>Bridges.LazyBridgeOptimizer</code></a>.</p><p>First, wrap an inner optimizer:</p><pre><code class="language-julia-repl hljs">julia&gt; inner_optimizer = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; optimizer = MOI.Bridges.LazyBridgeOptimizer(inner_optimizer)
+MOIB.LazyBridgeOptimizer{MOIU.Model{Float64}}
+with 0 variable bridges
+with 0 constraint bridges
+with 0 objective bridges
+with inner model MOIU.Model{Float64}</code></pre><p>Then use <a href="../reference/#MathOptInterface.Bridges.add_bridge"><code>Bridges.add_bridge</code></a> to add individual bridges:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Bridges.add_bridge(optimizer, MOI.Bridges.Constraint.SplitIntervalBridge{Float64})
+
+julia&gt; MOI.Bridges.add_bridge(optimizer, MOI.Bridges.Objective.FunctionizeBridge{Float64})</code></pre><p>Now the constraints will be bridged only if needed:</p><pre><code class="language-julia-repl hljs">julia&gt; x = MOI.add_variable(optimizer)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.add_constraint(optimizer, x, MOI.Interval(0.0, 1.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)
+
+julia&gt; MOI.get(optimizer, MOI.ListOfConstraintTypesPresent())
+1-element Vector{Tuple{Type, Type}}:
+ (MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64})
+
+julia&gt; MOI.get(inner_optimizer, MOI.ListOfConstraintTypesPresent())
+1-element Vector{Tuple{Type, Type}}:
+ (MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64})</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../Benchmarks/reference/">« API Reference</a><a class="docs-footer-nextpage" href="../list_of_bridges/">List of bridges »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Bridges/reference/index.html b/previews/PR3525/moi/submodules/Bridges/reference/index.html
new file mode 100644
index 00000000000..d1b5f7e01f8
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Bridges/reference/index.html
@@ -0,0 +1,199 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>API Reference · JuMP</title><meta name="title" content="API Reference · JuMP"/><meta property="og:title" content="API Reference · JuMP"/><meta property="twitter:title" content="API Reference · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../overview/">Overview</a></li><li><a class="tocitem" href="../list_of_bridges/">List of bridges</a></li><li class="is-active"><a class="tocitem" href>API Reference</a><ul class="internal"><li><a class="tocitem" href="#AbstractBridge-API"><span>AbstractBridge API</span></a></li><li><a class="tocitem" href="#Constraint-bridge-API"><span>Constraint bridge API</span></a></li><li><a class="tocitem" href="#Objective-bridge-API"><span>Objective bridge API</span></a></li><li><a class="tocitem" href="#ref_variable_bridges"><span>Variable bridge API</span></a></li><li><a class="tocitem" href="#AbstractBridgeOptimizer-API"><span>AbstractBridgeOptimizer API</span></a></li><li><a class="tocitem" href="#LazyBridgeOptimizer-API"><span>LazyBridgeOptimizer API</span></a></li><li><a class="tocitem" href="#constraint_set_map"><span>SetMap API</span></a></li><li><a class="tocitem" href="#Bridging-graph-API"><span>Bridging graph API</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Bridges</a></li><li class="is-active"><a href>API Reference</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>API Reference</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Bridges/reference.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Bridges"><a class="docs-heading-anchor" href="#Bridges">Bridges</a><a id="Bridges-1"></a><a class="docs-heading-anchor-permalink" href="#Bridges" title="Permalink"></a></h1><h2 id="AbstractBridge-API"><a class="docs-heading-anchor" href="#AbstractBridge-API">AbstractBridge API</a><a id="AbstractBridge-API-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractBridge-API" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.AbstractBridge" href="#MathOptInterface.Bridges.AbstractBridge"><code>MathOptInterface.Bridges.AbstractBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractBridge end</code></pre><p>An abstract type representing a bridged constraint or variable in a <a href="#MathOptInterface.Bridges.AbstractBridgeOptimizer"><code>MOI.Bridges.AbstractBridgeOptimizer</code></a>.</p><p>All bridges must implement:</p><ul><li><a href="#MathOptInterface.Bridges.added_constrained_variable_types"><code>added_constrained_variable_types</code></a></li><li><a href="#MathOptInterface.Bridges.added_constraint_types"><code>added_constraint_types</code></a></li><li><a href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfVariables}"><code>MOI.get(::AbstractBridge, ::MOI.NumberOfVariables)</code></a></li><li><a href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfVariableIndices}"><code>MOI.get(::AbstractBridge, ::MOI.ListOfVariableIndices)</code></a></li><li><a href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfConstraints}"><code>MOI.get(::AbstractBridge, ::MOI.NumberOfConstraints)</code></a></li><li><a href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfConstraintIndices}"><code>MOI.get(::AbstractBridge, ::MOI.ListOfConstraintIndices)</code></a></li></ul><p>Subtypes of <code>AbstractBridge</code> may have additional requirements. Consult their docstrings for details.</p><p>In addition, all subtypes may optionally implement the following constraint attributes with the bridge in place of the constraint index:</p><ul><li><a href="../../../reference/constraints/#MathOptInterface.ConstraintDual"><code>MOI.ConstraintDual</code></a></li><li><a href="../../../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>MOI.ConstraintPrimal</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L7-L30">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.added_constrained_variable_types" href="#MathOptInterface.Bridges.added_constrained_variable_types"><code>MathOptInterface.Bridges.added_constrained_variable_types</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">added_constrained_variable_types(
+    BT::Type{&lt;:AbstractBridge},
+)::Vector{Tuple{Type}}</code></pre><p>Return a list of the types of constrained variables that bridges of concrete type <code>BT</code> add.</p><p><strong>Implementation notes</strong></p><ul><li>This method depends only on the type of the bridge, not the runtime value. If the bridge <em>may</em> add a constrained variable, the type <em>must</em> be included in the return vector.</li><li>If the bridge adds a free variable via <a href="../../../reference/variables/#MathOptInterface.add_variable"><code>MOI.add_variable</code></a> or <a href="../../../reference/variables/#MathOptInterface.add_variables"><code>MOI.add_variables</code></a>, the return vector <em>must</em> include <code>(MOI.Reals,)</code>.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Bridges.added_constrained_variable_types(
+           MOI.Bridges.Variable.NonposToNonnegBridge{Float64},
+       )
+1-element Vector{Tuple{Type}}:
+ (MathOptInterface.Nonnegatives,)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L185-L210">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.added_constraint_types" href="#MathOptInterface.Bridges.added_constraint_types"><code>MathOptInterface.Bridges.added_constraint_types</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">added_constraint_types(
+    BT::Type{&lt;:AbstractBridge},
+)::Vector{Tuple{Type,Type}}</code></pre><p>Return a list of the types of constraints that bridges of concrete type <code>BT</code> add.</p><p><strong>Implementation notes</strong></p><ul><li>This method depends only on the type of the bridge, not the runtime value. If the bridge <em>may</em> add a constraint, the type <em>must</em> be included in the return vector.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Bridges.added_constraint_types(
+           MOI.Bridges.Constraint.ZeroOneBridge{Float64},
+       )
+2-element Vector{Tuple{Type, Type}}:
+ (MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.Interval{Float64})
+ (MathOptInterface.VariableIndex, MathOptInterface.Integer)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L213-L237">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfVariables}" href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfVariables}"><code>MathOptInterface.get</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.NumberOfVariables)::Int64</code></pre><p>Return the number of variables created by the bridge <code>b</code> in the model.</p><p>See also <a href="../../../reference/models/#MathOptInterface.NumberOfConstraints"><code>MOI.NumberOfConstraints</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this if the bridge adds new variables.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L42-L53">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfVariableIndices}" href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfVariableIndices}"><code>MathOptInterface.get</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.ListOfVariableIndices)</code></pre><p>Return the list of variables created by the bridge <code>b</code>.</p><p>See also <a href="../../../reference/models/#MathOptInterface.ListOfVariableIndices"><code>MOI.ListOfVariableIndices</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this if the bridge adds new variables.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L56-L67">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfConstraints}" href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfConstraints}"><code>MathOptInterface.get</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.NumberOfConstraints{F,S})::Int64 where {F,S}</code></pre><p>Return the number of constraints of the type <code>F</code>-in-<code>S</code> created by the bridge <code>b</code>.</p><p>See also <a href="../../../reference/models/#MathOptInterface.NumberOfConstraints"><code>MOI.NumberOfConstraints</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this for the constraint types returned by <a href="#MathOptInterface.Bridges.added_constraint_types"><code>added_constraint_types</code></a>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L70-L82">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfConstraintIndices}" href="#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfConstraintIndices}"><code>MathOptInterface.get</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOI.get(b::AbstractBridge, ::MOI.ListOfConstraintIndices{F,S}) where {F,S}</code></pre><p>Return a  <code>Vector{ConstraintIndex{F,S}}</code> with indices of all constraints of type <code>F</code>-in-<code>S</code> created by the bride <code>b</code>.</p><p>See also <a href="../../../reference/models/#MathOptInterface.ListOfConstraintIndices"><code>MOI.ListOfConstraintIndices</code></a>.</p><p><strong>Implementation notes</strong></p><ul><li>There is a default fallback, so you need only implement this for the constraint types returned by <a href="#MathOptInterface.Bridges.added_constraint_types"><code>added_constraint_types</code></a>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L85-L97">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.needs_final_touch" href="#MathOptInterface.Bridges.needs_final_touch"><code>MathOptInterface.Bridges.needs_final_touch</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">needs_final_touch(bridge::AbstractBridge)::Bool</code></pre><p>Return whether <a href="#MathOptInterface.Bridges.final_touch"><code>final_touch</code></a> is implemented by <code>bridge</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L263-L267">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.final_touch" href="#MathOptInterface.Bridges.final_touch"><code>MathOptInterface.Bridges.final_touch</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">final_touch(bridge::AbstractBridge, model::MOI.ModelLike)::Nothing</code></pre><p>A function that is called immediately prior to <a href="../../../reference/models/#MathOptInterface.optimize!"><code>MOI.optimize!</code></a> to allow bridges to modify their reformulations with repsect to other variables and constraints in <code>model</code>.</p><p>For example, if the correctness of <code>bridge</code> depends on the bounds of a variable or the fact that variables are integer, then the bridge can implement <a href="#MathOptInterface.Bridges.final_touch"><code>final_touch</code></a> to check assumptions immediately before a call to <a href="../../../reference/models/#MathOptInterface.optimize!"><code>MOI.optimize!</code></a>.</p><p>If you implement this method, you must also implement <a href="#MathOptInterface.Bridges.needs_final_touch"><code>needs_final_touch</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L270-L284">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.bridging_cost" href="#MathOptInterface.Bridges.bridging_cost"><code>MathOptInterface.Bridges.bridging_cost</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridging_cost(b::AbstractBridgeOptimizer, S::Type{&lt;:MOI.AbstractSet}})</code></pre><p>Return the cost of bridging variables constrained in <code>S</code> on creation, <code>is_bridged(b, S)</code> is assumed to be <code>true</code>.</p><pre><code class="nohighlight hljs">bridging_cost(
+    b::AbstractBridgeOptimizer,
+    F::Type{&lt;:MOI.AbstractFunction},
+    S::Type{&lt;:MOI.AbstractSet},
+)</code></pre><p>Return the cost of bridging <code>F</code>-in-<code>S</code> constraints.</p><p><code>is_bridged(b, S)</code> is assumed to be <code>true</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge_optimizer.jl#L921-L936">source</a></section></article><h2 id="Constraint-bridge-API"><a class="docs-heading-anchor" href="#Constraint-bridge-API">Constraint bridge API</a><a id="Constraint-bridge-API-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-bridge-API" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.AbstractBridge" href="#MathOptInterface.Bridges.Constraint.AbstractBridge"><code>MathOptInterface.Bridges.Constraint.AbstractBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractBridge &lt;: MOI.Bridges.AbstractType</code></pre><p>Subtype of <a href="#MathOptInterface.Bridges.AbstractBridge"><code>MOI.Bridges.AbstractBridge</code></a> for constraint bridges.</p><p>In addition to the required implementation described in <a href="#MathOptInterface.Bridges.AbstractBridge"><code>MOI.Bridges.AbstractBridge</code></a>, subtypes of <code>AbstractBridge</code> must additionally implement:</p><ul><li><a href="#MathOptInterface.supports_constraint-Tuple{Type{&lt;:MathOptInterface.Bridges.Constraint.AbstractBridge}, Type{&lt;:MathOptInterface.AbstractFunction}, Type{&lt;:MathOptInterface.AbstractSet}}"><code>MOI.supports_constraint(::Type{&lt;:AbstractBridge}, ::Type{&lt;:MOI.AbstractFunction}, ::Type{&lt;:MOI.AbstractSet})</code></a></li><li><a href="#MathOptInterface.Bridges.Constraint.concrete_bridge_type"><code>concrete_bridge_type</code></a></li><li><a href="#MathOptInterface.Bridges.Constraint.bridge_constraint"><code>bridge_constraint</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridge.jl#L7-L19">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.supports_constraint-Tuple{Type{&lt;:MathOptInterface.Bridges.Constraint.AbstractBridge}, Type{&lt;:MathOptInterface.AbstractFunction}, Type{&lt;:MathOptInterface.AbstractSet}}" href="#MathOptInterface.supports_constraint-Tuple{Type{&lt;:MathOptInterface.Bridges.Constraint.AbstractBridge}, Type{&lt;:MathOptInterface.AbstractFunction}, Type{&lt;:MathOptInterface.AbstractSet}}"><code>MathOptInterface.supports_constraint</code></a> — <span class="docstring-category">Method</span></header><section><div><pre><code class="language-julia hljs">MOI.supports_constraint(
+    BT::Type{&lt;:AbstractBridge},
+    F::Type{&lt;:MOI.AbstractFunction},
+    S::Type{&lt;:MOI.AbstractSet},
+)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether the bridges of type <code>BT</code> support bridging <code>F</code>-in-<code>S</code> constraints.</p><p><strong>Implementation notes</strong></p><ul><li>This method depends only on the type of the inputs, not the runtime values.</li><li>There is a default fallback, so you need only implement this method for constraint types that the bridge implements.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridge.jl#L22-L37">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.concrete_bridge_type" href="#MathOptInterface.Bridges.Constraint.concrete_bridge_type"><code>MathOptInterface.Bridges.Constraint.concrete_bridge_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">concrete_bridge_type(
+    BT::Type{&lt;:AbstractBridge},
+    F::Type{&lt;:MOI.AbstractFunction},
+    S::Type{&lt;:MOI.AbstractSet}
+)::Type</code></pre><p>Return the concrete type of the bridge supporting <code>F</code>-in-<code>S</code> constraints.</p><p>This function can only be called if <code>MOI.supports_constraint(BT, F, S)</code> is <code>true</code>.</p><p><strong>Example</strong></p><p>The <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitIntervalBridge"><code>SplitIntervalBridge</code></a> bridges a <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval</code></a> constraint into a <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a> and a <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan</code></a> constraint.</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Bridges.Constraint.concrete_bridge_type(
+           MOI.Bridges.Constraint.SplitIntervalBridge{Float64},
+           MOI.VariableIndex,
+           MOI.Interval{Float64},
+       )
+MathOptInterface.Bridges.Constraint.SplitIntervalBridge{Float64, MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}, MathOptInterface.GreaterThan{Float64}, MathOptInterface.LessThan{Float64}}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridge.jl#L46-L72">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.bridge_constraint" href="#MathOptInterface.Bridges.Constraint.bridge_constraint"><code>MathOptInterface.Bridges.Constraint.bridge_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridge_constraint(
+    BT::Type{&lt;:AbstractBridge},
+    model::MOI.ModelLike,
+    func::AbstractFunction,
+    set::MOI.AbstractSet,
+)::BT</code></pre><p>Bridge the constraint <code>func</code>-in-<code>set</code> using bridge <code>BT</code> to <code>model</code> and returns a bridge object of type <code>BT</code>.</p><p><strong>Implementation notes</strong></p><ul><li>The bridge type <code>BT</code> should be a concrete type, that is, all the type parameters of the bridge must be set.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridge.jl#L89-L104">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.AbstractFunctionConversionBridge" href="#MathOptInterface.Bridges.Constraint.AbstractFunctionConversionBridge"><code>MathOptInterface.Bridges.Constraint.AbstractFunctionConversionBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractFunctionConversionBridge{F,S} &lt;: AbstractBridge end</code></pre><p>Abstract type to support writing bridges in which the function changes but the set does not.</p><p>By convention, the transformed function is stored in the <code>.constraint</code> field.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L7-L14">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SingleBridgeOptimizer" href="#MathOptInterface.Bridges.Constraint.SingleBridgeOptimizer"><code>MathOptInterface.Bridges.Constraint.SingleBridgeOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SingleBridgeOptimizer{BT&lt;:AbstractBridge}(model::MOI.ModelLike)</code></pre><p>Return <code>AbstractBridgeOptimizer</code> that always bridges any objective function supported by the bridge <code>BT</code>.</p><p>This is in contrast with the <a href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a>, which only bridges the objective function if it is supported by the bridge <code>BT</code> and unsupported by <code>model</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; struct MyNewBridge{T} &lt;: MOI.Bridges.Constraint.AbstractBridge end
+
+julia&gt; bridge = MOI.Bridges.Constraint.SingleBridgeOptimizer{MyNewBridge{Float64}}(
+           MOI.Utilities.Model{Float64}(),
+       )
+MOIB.Constraint.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}
+with 0 constraint bridges
+with inner model MOIU.Model{Float64}</code></pre><p><strong>Implementation notes</strong></p><p>All bridges should simplify the creation of <code>SingleBridgeOptimizer</code>s by defining a constant that wraps the bridge in a <code>SingleBridgeOptimizer</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; const MyNewBridgeModel{T,OT&lt;:MOI.ModelLike} =
+           MOI.Bridges.Constraint.SingleBridgeOptimizer{MyNewBridge{T},OT};</code></pre><p>This enables users to create bridged models as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; MyNewBridgeModel{Float64}(MOI.Utilities.Model{Float64}())
+MOIB.Constraint.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}
+with 0 constraint bridges
+with inner model MOIU.Model{Float64}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/single_bridge_optimizer.jl#L7-L45">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.add_all_bridges" href="#MathOptInterface.Bridges.Constraint.add_all_bridges"><code>MathOptInterface.Bridges.Constraint.add_all_bridges</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_all_bridges(bridged_model, ::Type{T}) where {T}</code></pre><p>Add all bridges defined in the <code>Bridges.Constraint</code> submodule to <code>bridged_model</code>. The coefficient type used is <code>T</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/Constraint.jl#L66-L71">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.FlipSignBridge" href="#MathOptInterface.Bridges.Constraint.FlipSignBridge"><code>MathOptInterface.Bridges.Constraint.FlipSignBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">FlipSignBridge{T,S1,S2,F,G}</code></pre><p>An abstract type that simplifies the creation of other bridges.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/flip_sign.jl#L7-L11">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.AbstractToIntervalBridge" href="#MathOptInterface.Bridges.Constraint.AbstractToIntervalBridge"><code>MathOptInterface.Bridges.Constraint.AbstractToIntervalBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractToIntervalBridge{T&lt;:AbstractFloat,S,F}</code></pre><p>An abstract type that simplifies the creation of other bridges.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><code>T</code> must be a <code>AbstractFloat</code> type because otherwise <code>typemin</code> and <code>typemax</code> would either be not implemented (e.g. <code>BigInt</code>), or would not give infinite value (e.g. <code>Int</code>). For this reason, this bridge is only added to <a href="#MathOptInterface.Bridges.full_bridge_optimizer"><code>MOI.Bridges.full_bridge_optimizer</code></a> when <code>T</code> is a subtype of <code>AbstractFloat</code>.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/ltgt_to_interval.jl#L7-L18">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.SetMapBridge" href="#MathOptInterface.Bridges.Constraint.SetMapBridge"><code>MathOptInterface.Bridges.Constraint.SetMapBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type SetMapBridge{T,S2,S1,F,G} &lt;: AbstractBridge end</code></pre><p>Consider two type of sets, <code>S1</code> and <code>S2</code>, and a linear mapping <code>A</code> such that the image of a set of type <code>S1</code> under <code>A</code> is a set of type <code>S2</code>.</p><p>A <code>SetMapBridge{T,S2,S1,F,G}</code> is a bridge that maps <code>G</code>-in-<code>S1</code> constraints into <code>F</code>-in-<code>S2</code> by mapping the function through <code>A</code>.</p><p>The linear map <code>A</code> is described by;</p><ul><li><a href="#MathOptInterface.Bridges.map_set"><code>MOI.Bridges.map_set</code></a></li><li><a href="#MathOptInterface.Bridges.map_function"><code>MOI.Bridges.map_function</code></a>.</li></ul><p>Implementing a method for these two functions is sufficient to bridge constraints. However, in order for the getters and setters of attributes such as dual solutions and starting values to work as well, a method for the following functions must be implemented:</p><ul><li><a href="#MathOptInterface.Bridges.inverse_map_set"><code>MOI.Bridges.inverse_map_set</code></a></li><li><a href="#MathOptInterface.Bridges.inverse_map_function"><code>MOI.Bridges.inverse_map_function</code></a></li><li><a href="#MathOptInterface.Bridges.adjoint_map_function"><code>MOI.Bridges.adjoint_map_function</code></a></li><li><a href="#MathOptInterface.Bridges.inverse_adjoint_map_function"><code>MOI.Bridges.inverse_adjoint_map_function</code></a></li></ul><p>See the docstrings of each function to see which feature would be missing if it was not implemented for a given bridge.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/set_map.jl#L7-L33">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Constraint.conversion_cost" href="#MathOptInterface.Bridges.Constraint.conversion_cost"><code>MathOptInterface.Bridges.Constraint.conversion_cost</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">conversion_cost(
+    F::Type{&lt;:MOI.AbstractFunction},
+    G::Type{&lt;:MOI.AbstractFunction},
+)::Float64</code></pre><p>Return a <code>Float64</code> returning the cost of converting any function of type <code>G</code> to a function of type <code>F</code> with <code>convert</code>.</p><p>This cost is used to compute <a href="#MathOptInterface.Bridges.bridging_cost"><code>MOI.Bridges.bridging_cost</code></a>.</p><p>The default cost is <code>Inf</code>, which means that <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.FunctionConversionBridge"><code>MOI.Bridges.Constraint.FunctionConversionBridge</code></a> should not attempt the conversion.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Constraint/bridges/functionize.jl#L201-L215">source</a></section></article><h2 id="Objective-bridge-API"><a class="docs-heading-anchor" href="#Objective-bridge-API">Objective bridge API</a><a id="Objective-bridge-API-1"></a><a class="docs-heading-anchor-permalink" href="#Objective-bridge-API" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.AbstractBridge" href="#MathOptInterface.Bridges.Objective.AbstractBridge"><code>MathOptInterface.Bridges.Objective.AbstractBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractBridge &lt;: MOI.Bridges.AbstractBridge end</code></pre><p>Subtype of <a href="#MathOptInterface.Bridges.AbstractBridge"><code>MOI.Bridges.AbstractBridge</code></a> for objective bridges.</p><p>In addition to the required implementation described in <a href="#MathOptInterface.Bridges.AbstractBridge"><code>MOI.Bridges.AbstractBridge</code></a>, subtypes of <code>AbstractBridge</code> must additionally implement:</p><ul><li><a href="#MathOptInterface.Bridges.Objective.supports_objective_function"><code>supports_objective_function</code></a></li><li><a href="#MathOptInterface.Bridges.Objective.concrete_bridge_type"><code>concrete_bridge_type</code></a></li><li><a href="#MathOptInterface.Bridges.Objective.bridge_objective"><code>bridge_objective</code></a></li><li><a href="#MathOptInterface.Bridges.set_objective_function_type"><code>MOI.Bridges.set_objective_function_type</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridge.jl#L7-L20">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.supports_objective_function" href="#MathOptInterface.Bridges.Objective.supports_objective_function"><code>MathOptInterface.Bridges.Objective.supports_objective_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_objective_function(
+    BT::Type{&lt;:MOI.Bridges.Objective.AbstractBridge},
+    F::Type{&lt;:MOI.AbstractFunction},
+)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether the bridges of type <code>BT</code> support bridging objective functions of type <code>F</code>.</p><p><strong>Implementation notes</strong></p><ul><li>This method depends only on the type of the inputs, not the runtime values.</li><li>There is a default fallback, so you need only implement this method For objective functions that the bridge implements.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridge.jl#L23-L37">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.set_objective_function_type" href="#MathOptInterface.Bridges.set_objective_function_type"><code>MathOptInterface.Bridges.set_objective_function_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_objective_function_type(
+    BT::Type{&lt;:Objective.AbstractBridge},
+)::Type{&lt;:MOI.AbstractFunction}</code></pre><p>Return the type of objective function that bridges of concrete type <code>BT</code> set.</p><p><strong>Implementation notes</strong></p><ul><li>This method depends only on the type of the bridge, not the runtime value.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Bridges.set_objective_function_type(
+           MOI.Bridges.Objective.FunctionizeBridge{Float64},
+       )
+MathOptInterface.ScalarAffineFunction{Float64}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge.jl#L240-L260">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.concrete_bridge_type" href="#MathOptInterface.Bridges.Objective.concrete_bridge_type"><code>MathOptInterface.Bridges.Objective.concrete_bridge_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">concrete_bridge_type(
+    BT::Type{&lt;:MOI.Bridges.Objective.AbstractBridge},
+    F::Type{&lt;:MOI.AbstractFunction},
+)::Type</code></pre><p>Return the concrete type of the bridge supporting objective functions of type <code>F</code>.</p><p>This function can only be called if <code>MOI.supports_objective_function(BT, F)</code> is <code>true</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridge.jl#L45-L56">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.bridge_objective" href="#MathOptInterface.Bridges.Objective.bridge_objective"><code>MathOptInterface.Bridges.Objective.bridge_objective</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridge_objective(
+    BT::Type{&lt;:MOI.Bridges.Objective.AbstractBridge},
+    model::MOI.ModelLike,
+    func::MOI.AbstractFunction,
+)::BT</code></pre><p>Bridge the objective function <code>func</code> using bridge <code>BT</code> to <code>model</code> and returns a bridge object of type <code>BT</code>.</p><p><strong>Implementation notes</strong></p><ul><li>The bridge type <code>BT</code> must be a concrete type, that is, all the type parameters of the bridge must be set.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/bridge.jl#L71-L85">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.SingleBridgeOptimizer" href="#MathOptInterface.Bridges.Objective.SingleBridgeOptimizer"><code>MathOptInterface.Bridges.Objective.SingleBridgeOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SingleBridgeOptimizer{BT&lt;:AbstractBridge}(model::MOI.ModelLike)</code></pre><p>Return <code>AbstractBridgeOptimizer</code> that always bridges any objective function supported by the bridge <code>BT</code>.</p><p>This is in contrast with the <a href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a>, which only bridges the objective function if it is supported by the bridge <code>BT</code> and unsupported by <code>model</code>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; struct MyNewBridge{T} &lt;: MOI.Bridges.Objective.AbstractBridge end
+
+julia&gt; bridge = MOI.Bridges.Objective.SingleBridgeOptimizer{MyNewBridge{Float64}}(
+           MOI.Utilities.Model{Float64}(),
+       )
+MOIB.Objective.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}
+with 0 objective bridges
+with inner model MOIU.Model{Float64}</code></pre><p><strong>Implementation notes</strong></p><p>All bridges should simplify the creation of <code>SingleBridgeOptimizer</code>s by defining a constant that wraps the bridge in a <code>SingleBridgeOptimizer</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; const MyNewBridgeModel{T,OT&lt;:MOI.ModelLike} =
+           MOI.Bridges.Objective.SingleBridgeOptimizer{MyNewBridge{T},OT};</code></pre><p>This enables users to create bridged models as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; MyNewBridgeModel{Float64}(MOI.Utilities.Model{Float64}())
+MOIB.Objective.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}
+with 0 objective bridges
+with inner model MOIU.Model{Float64}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/single_bridge_optimizer.jl#L7-L45">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Objective.add_all_bridges" href="#MathOptInterface.Bridges.Objective.add_all_bridges"><code>MathOptInterface.Bridges.Objective.add_all_bridges</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_all_bridges(model, ::Type{T}) where {T}</code></pre><p>Add all bridges defined in the <code>Bridges.Objective</code> submodule to <code>model</code>.</p><p>The coefficient type used is <code>T</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Objective/Objective.jl#L21-L27">source</a></section></article><h2 id="ref_variable_bridges"><a class="docs-heading-anchor" href="#ref_variable_bridges">Variable bridge API</a><a id="ref_variable_bridges-1"></a><a class="docs-heading-anchor-permalink" href="#ref_variable_bridges" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.AbstractBridge" href="#MathOptInterface.Bridges.Variable.AbstractBridge"><code>MathOptInterface.Bridges.Variable.AbstractBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractBridge &lt;: MOI.Bridges.AbstractBridge end</code></pre><p>Subtype of <a href="#MathOptInterface.Bridges.AbstractBridge"><code>MOI.Bridges.AbstractBridge</code></a> for variable bridges.</p><p>In addition to the required implementation described in <a href="#MathOptInterface.Bridges.AbstractBridge"><code>MOI.Bridges.AbstractBridge</code></a>, subtypes of <code>AbstractBridge</code> must additionally implement:</p><ul><li><a href="#MathOptInterface.Bridges.Variable.supports_constrained_variable"><code>supports_constrained_variable</code></a></li><li><a href="#MathOptInterface.Bridges.Variable.concrete_bridge_type"><code>concrete_bridge_type</code></a></li><li><a href="#MathOptInterface.Bridges.Variable.bridge_constrained_variable"><code>bridge_constrained_variable</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridge.jl#L7-L19">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.supports_constrained_variable" href="#MathOptInterface.Bridges.Variable.supports_constrained_variable"><code>MathOptInterface.Bridges.Variable.supports_constrained_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_constrained_variable(
+    BT::Type{&lt;:AbstractBridge},
+    S::Type{&lt;:MOI.AbstractSet},
+)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether the bridges of type <code>BT</code> support bridging constrained variables in <code>S</code>. That is, it returns <code>true</code> if the bridge of type <code>BT</code> converts constrained variables of type <code>S</code> into a form supported by the solver.</p><p><strong>Implementation notes</strong></p><ul><li>This method depends only on the type of the bridge and set, not the runtime values.</li><li>There is a default fallback, so you need only implement this method for sets that the bridge implements.</li></ul><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Bridges.Variable.supports_constrained_variable(
+           MOI.Bridges.Variable.NonposToNonnegBridge{Float64},
+           MOI.Nonpositives,
+       )
+true
+
+julia&gt; MOI.Bridges.Variable.supports_constrained_variable(
+           MOI.Bridges.Variable.NonposToNonnegBridge{Float64},
+           MOI.Nonnegatives,
+       )
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridge.jl#L22-L55">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.concrete_bridge_type" href="#MathOptInterface.Bridges.Variable.concrete_bridge_type"><code>MathOptInterface.Bridges.Variable.concrete_bridge_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">concrete_bridge_type(
+    BT::Type{&lt;:AbstractBridge},
+    S::Type{&lt;:MOI.AbstractSet},
+)::Type</code></pre><p>Return the concrete type of the bridge supporting variables in <code>S</code> constraints.</p><p>This function can only be called if <code>MOI.supports_constrained_variable(BT, S)</code> is <code>true</code>.</p><p><strong>Examples</strong></p><p>As a variable in <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a> is bridged into variables in <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a> by the <a href="../list_of_bridges/#MathOptInterface.Bridges.Variable.VectorizeBridge"><code>VectorizeBridge</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Bridges.Variable.concrete_bridge_type(
+           MOI.Bridges.Variable.VectorizeBridge{Float64},
+           MOI.GreaterThan{Float64},
+       )
+MathOptInterface.Bridges.Variable.VectorizeBridge{Float64, MathOptInterface.Nonnegatives}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridge.jl#L63-L87">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.bridge_constrained_variable" href="#MathOptInterface.Bridges.Variable.bridge_constrained_variable"><code>MathOptInterface.Bridges.Variable.bridge_constrained_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridge_constrained_variable(
+    BT::Type{&lt;:AbstractBridge},
+    model::MOI.ModelLike,
+    set::MOI.AbstractSet,
+)::BT</code></pre><p>Bridge the constrained variable in <code>set</code> using bridge <code>BT</code> to <code>model</code> and returns a bridge object of type <code>BT</code>.</p><p><strong>Implementation notes</strong></p><ul><li>The bridge type <code>BT</code> must be a concrete type, that is, all the type parameters of the bridge must be set.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridge.jl#L97-L111">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.SingleBridgeOptimizer" href="#MathOptInterface.Bridges.Variable.SingleBridgeOptimizer"><code>MathOptInterface.Bridges.Variable.SingleBridgeOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SingleBridgeOptimizer{BT&lt;:AbstractBridge}(model::MOI.ModelLike)</code></pre><p>Return <a href="#MathOptInterface.Bridges.AbstractBridgeOptimizer"><code>MOI.Bridges.AbstractBridgeOptimizer</code></a> that always bridges any variables constrained on creation supported by the bridge <code>BT</code>.</p><p>This is in contrast with the <a href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a>, which only bridges the variables constrained on creation if they are supported by the bridge <code>BT</code> and unsupported by <code>model</code>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Two <code>SingleBridgeOptimizer</code>s cannot be used together as both of them assume that the underlying model only returns variable indices with nonnegative values. Use <a href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MOI.Bridges.LazyBridgeOptimizer</code></a> instead.</p></div></div><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; struct MyNewBridge{T} &lt;: MOI.Bridges.Variable.AbstractBridge end
+
+julia&gt; bridge = MOI.Bridges.Variable.SingleBridgeOptimizer{MyNewBridge{Float64}}(
+           MOI.Utilities.Model{Float64}(),
+       )
+MOIB.Variable.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}
+with 0 variable bridges
+with inner model MOIU.Model{Float64}</code></pre><p><strong>Implementation notes</strong></p><p>All bridges should simplify the creation of <code>SingleBridgeOptimizer</code>s by defining a constant that wraps the bridge in a <code>SingleBridgeOptimizer</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; const MyNewBridgeModel{T,OT&lt;:MOI.ModelLike} =
+           MOI.Bridges.Variable.SingleBridgeOptimizer{MyNewBridge{T},OT};</code></pre><p>This enables users to create bridged models as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; MyNewBridgeModel{Float64}(MOI.Utilities.Model{Float64}())
+MOIB.Variable.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}
+with 0 variable bridges
+with inner model MOIU.Model{Float64}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/single_bridge_optimizer.jl#L7-L50">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.add_all_bridges" href="#MathOptInterface.Bridges.Variable.add_all_bridges"><code>MathOptInterface.Bridges.Variable.add_all_bridges</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_all_bridges(model, ::Type{T}) where {T}</code></pre><p>Add all bridges defined in the <code>Bridges.Variable</code> submodule to <code>model</code>.</p><p>The coefficient type used is <code>T</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/Variable.jl#L26-L32">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.FlipSignBridge" href="#MathOptInterface.Bridges.Variable.FlipSignBridge"><code>MathOptInterface.Bridges.Variable.FlipSignBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type FlipSignBridge{T,S1,S2} &lt;: SetMapBridge{T,S2,S1} end</code></pre><p>An abstract type that simplifies the creation of other bridges.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridges/flip_sign.jl#L7-L11">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.SetMapBridge" href="#MathOptInterface.Bridges.Variable.SetMapBridge"><code>MathOptInterface.Bridges.Variable.SetMapBridge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type SetMapBridge{T,S1,S2} &lt;: AbstractBridge end</code></pre><p>Consider two type of sets, <code>S1</code> and <code>S2</code>, and a linear mapping <code>A</code> such that the image of a set of type <code>S1</code> under <code>A</code> is a set of type <code>S2</code>.</p><p>A <code>SetMapBridge{T,S1,S2}</code> is a bridge that substitutes constrained variables in <code>S2</code> into the image through <code>A</code> of constrained variables in <code>S1</code>.</p><p>The linear map <code>A</code> is described by:</p><ul><li><a href="#MathOptInterface.Bridges.map_set"><code>MOI.Bridges.map_set</code></a></li><li><a href="#MathOptInterface.Bridges.map_function"><code>MOI.Bridges.map_function</code></a></li></ul><p>Implementing a method for these two functions is sufficient to bridge constrained variables. However, in order for the getters and setters of attributes such as dual solutions and starting values to work as well, a method for the following functions must be implemented:</p><ul><li><a href="#MathOptInterface.Bridges.inverse_map_set"><code>MOI.Bridges.inverse_map_set</code></a></li><li><a href="#MathOptInterface.Bridges.inverse_map_function"><code>MOI.Bridges.inverse_map_function</code></a></li><li><a href="#MathOptInterface.Bridges.adjoint_map_function"><code>MOI.Bridges.adjoint_map_function</code></a></li><li><a href="#MathOptInterface.Bridges.inverse_adjoint_map_function"><code>MOI.Bridges.inverse_adjoint_map_function</code></a>.</li></ul><p>See the docstrings of each function to see which feature would be missing if it was not implemented for a given bridge.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/set_map.jl#L7-L33">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Variable.unbridged_map" href="#MathOptInterface.Bridges.Variable.unbridged_map"><code>MathOptInterface.Bridges.Variable.unbridged_map</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unbridged_map(
+   bridge::MOI.Bridges.Variable.AbstractBridge,
+    vi::MOI.VariableIndex,
+)</code></pre><p>For a bridged variable in a scalar set, return a tuple of pairs mapping the variables created by the bridge to an affine expression in terms of the bridged variable <code>vi</code>.</p><pre><code class="language-julia hljs">unbridged_map(
+    bridge::MOI.Bridges.Variable.AbstractBridge,
+    vis::Vector{MOI.VariableIndex},
+)</code></pre><p>For a bridged variable in a vector set, return a tuple of pairs mapping the variables created by the bridge to an affine expression in terms of the bridged variable <code>vis</code>. If this method is not implemented, it falls back to calling the following method for every variable of <code>vis</code>.</p><pre><code class="language-julia hljs">unbridged_map(
+    bridge::MOI.Bridges.Variable.AbstractBridge,
+    vi::MOI.VariableIndex,
+    i::MOI.Bridges.IndexInVector,
+)</code></pre><p>For a bridged variable in a vector set, return a tuple of pairs mapping the variables created by the bridge to an affine expression in terms of the bridged variable <code>vi</code> corresponding to the <code>i</code>th variable of the vector.</p><p>If there is no way to recover the expression in terms of the bridged variable(s) <code>vi(s)</code>, return <code>nothing</code>. See <a href="../list_of_bridges/#MathOptInterface.Bridges.Variable.ZerosBridge"><code>ZerosBridge</code></a> for an example of bridge returning <code>nothing</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Variable/bridge.jl#L163-L200">source</a></section></article><h2 id="AbstractBridgeOptimizer-API"><a class="docs-heading-anchor" href="#AbstractBridgeOptimizer-API">AbstractBridgeOptimizer API</a><a id="AbstractBridgeOptimizer-API-1"></a><a class="docs-heading-anchor-permalink" href="#AbstractBridgeOptimizer-API" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.AbstractBridgeOptimizer" href="#MathOptInterface.Bridges.AbstractBridgeOptimizer"><code>MathOptInterface.Bridges.AbstractBridgeOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractBridgeOptimizer &lt;: MOI.AbstractOptimizer end</code></pre><p>An abstract type that implements generic functions for bridges.</p><p><strong>Implementation notes</strong></p><p>By convention, the inner optimizer should be stored in a <code>model</code> field. If not, the optimizer must implement <a href="../../../reference/models/#MathOptInterface.optimize!"><code>MOI.optimize!</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge_optimizer.jl#L7-L16">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.bridged_variable_function" href="#MathOptInterface.Bridges.bridged_variable_function"><code>MathOptInterface.Bridges.bridged_variable_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridged_variable_function(
+    b::AbstractBridgeOptimizer,
+    vi::MOI.VariableIndex,
+)</code></pre><p>Return a <code>MOI.AbstractScalarFunction</code> of variables of <code>b.model</code> that equals <code>vi</code>. That is, if the variable <code>vi</code> is bridged, it returns its expression in terms of the variables of <code>b.model</code>. Otherwise, it returns <code>vi</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge_optimizer.jl#L2003-L2013">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.unbridged_variable_function" href="#MathOptInterface.Bridges.unbridged_variable_function"><code>MathOptInterface.Bridges.unbridged_variable_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unbridged_variable_function(
+    b::AbstractBridgeOptimizer,
+    vi::MOI.VariableIndex,
+)</code></pre><p>Return a <code>MOI.AbstractScalarFunction</code> of variables of <code>b</code> that equals <code>vi</code>. That is, if the variable <code>vi</code> is an internal variable of <code>b.model</code> created by a bridge but not visible to the user, it returns its expression in terms of the variables of bridged variables. Otherwise, it returns <code>vi</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge_optimizer.jl#L2070-L2080">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.bridged_function" href="#MathOptInterface.Bridges.bridged_function"><code>MathOptInterface.Bridges.bridged_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridged_function(b::AbstractBridgeOptimizer, value)::typeof(value)</code></pre><p>Substitute any bridged <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <code>value</code> by an equivalent expression in terms of variables of <code>b.model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge_optimizer.jl#L2031-L2036">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.supports_constraint_bridges" href="#MathOptInterface.Bridges.supports_constraint_bridges"><code>MathOptInterface.Bridges.supports_constraint_bridges</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_constraint_bridges(b::AbstractBridgeOptimizer)::Bool</code></pre><p>Return a <code>Bool</code> indicating if <code>b</code> supports <a href="#MathOptInterface.Bridges.Constraint.AbstractBridge"><code>MOI.Bridges.Constraint.AbstractBridge</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge_optimizer.jl#L42-L47">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.recursive_model" href="#MathOptInterface.Bridges.recursive_model"><code>MathOptInterface.Bridges.recursive_model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">recursive_model(b::AbstractBridgeOptimizer)</code></pre><p>If a variable, constraint, or objective is bridged, return the context of the inner variables. For most optimizers, this should be <code>b.model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/bridge_optimizer.jl#L34-L39">source</a></section></article><h2 id="LazyBridgeOptimizer-API"><a class="docs-heading-anchor" href="#LazyBridgeOptimizer-API">LazyBridgeOptimizer API</a><a id="LazyBridgeOptimizer-API-1"></a><a class="docs-heading-anchor-permalink" href="#LazyBridgeOptimizer-API" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.LazyBridgeOptimizer" href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>MathOptInterface.Bridges.LazyBridgeOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">LazyBridgeOptimizer(model::MOI.ModelLike)</code></pre><p>The <code>LazyBridgeOptimizer</code> is a bridge optimizer that supports multiple bridges, and only bridges things which are not supported by the internal model.</p><p>Internally, the <code>LazyBridgeOptimizer</code> solves a shortest hyper-path problem to determine which bridges to use.</p><p>In general, you should use <a href="#MathOptInterface.Bridges.full_bridge_optimizer"><code>full_bridge_optimizer</code></a> instead of this constructor because <a href="#MathOptInterface.Bridges.full_bridge_optimizer"><code>full_bridge_optimizer</code></a> automatically adds a large number of supported bridges.</p><p>See also: <a href="#MathOptInterface.Bridges.add_bridge"><code>add_bridge</code></a>, <a href="#MathOptInterface.Bridges.remove_bridge"><code>remove_bridge</code></a>, <a href="#MathOptInterface.Bridges.has_bridge"><code>has_bridge</code></a> and <a href="#MathOptInterface.Bridges.full_bridge_optimizer"><code>full_bridge_optimizer</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Bridges.LazyBridgeOptimizer(MOI.Utilities.Model{Float64}())
+MOIB.LazyBridgeOptimizer{MOIU.Model{Float64}}
+with 0 variable bridges
+with 0 constraint bridges
+with 0 objective bridges
+with inner model MOIU.Model{Float64}
+
+julia&gt; MOI.Bridges.add_bridge(model, MOI.Bridges.Variable.FreeBridge{Float64})
+
+julia&gt; MOI.Bridges.has_bridge(model, MOI.Bridges.Variable.FreeBridge{Float64})
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/lazy_bridge_optimizer.jl#L9-L40">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.full_bridge_optimizer" href="#MathOptInterface.Bridges.full_bridge_optimizer"><code>MathOptInterface.Bridges.full_bridge_optimizer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">full_bridge_optimizer(model::MOI.ModelLike, ::Type{T}) where {T}</code></pre><p>Returns a <a href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>LazyBridgeOptimizer</code></a> bridging <code>model</code> for every bridge defined in this package (see below for the few exceptions) and for the coefficient type <code>T</code>, as well as the bridges in the list returned by the <a href="#MathOptInterface.Bridges.ListOfNonstandardBridges"><code>ListOfNonstandardBridges</code></a> attribute.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}();
+
+julia&gt; bridged_model = MOI.Bridges.full_bridge_optimizer(model, Float64);</code></pre><p><strong>Exceptions</strong></p><p>The following bridges are not added by <code>full_bridge_optimizer</code>, except if they are in the list returned by the <a href="#MathOptInterface.Bridges.ListOfNonstandardBridges"><code>ListOfNonstandardBridges</code></a> attribute:</p><ul><li><a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge"><code>Constraint.SOCtoNonConvexQuadBridge</code></a></li><li><code>Constraint.RSOCtoNonConvexQuadBridge</code>](@ref)</li><li><a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoPSDBridge"><code>Constraint.SOCtoPSDBridge</code></a></li><li>If <code>T</code> is not a subtype of <code>AbstractFloat</code>, subtypes of <a href="#MathOptInterface.Bridges.Constraint.AbstractToIntervalBridge"><code>Constraint.AbstractToIntervalBridge</code></a><ul><li><a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.GreaterToIntervalBridge"><code>Constraint.GreaterToIntervalBridge</code></a></li><li><a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.LessToIntervalBridge"><code>Constraint.LessToIntervalBridge</code></a>)</li></ul></li></ul><p>See the docstring of the each bridge for the reason they are not added.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Bridges.jl#L25-L55">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.ListOfNonstandardBridges" href="#MathOptInterface.Bridges.ListOfNonstandardBridges"><code>MathOptInterface.Bridges.ListOfNonstandardBridges</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ListOfNonstandardBridges{T}() &lt;: MOI.AbstractOptimizerAttribute</code></pre><p>Any optimizer can be wrapped in a <a href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>LazyBridgeOptimizer</code></a> using <a href="#MathOptInterface.Bridges.full_bridge_optimizer"><code>full_bridge_optimizer</code></a>. However, by default <a href="#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>LazyBridgeOptimizer</code></a> uses a limited set of bridges that are:</p><ol><li>implemented in <code>MOI.Bridges</code></li><li>generally applicable for all optimizers.</li></ol><p>For some optimizers however, it is useful to add additional bridges, such as those that are implemented in external packages (e.g., within the solver package itself) or only apply in certain circumstances (e.g., <a href="../list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge"><code>Constraint.SOCtoNonConvexQuadBridge</code></a>).</p><p>Such optimizers should implement the <code>ListOfNonstandardBridges</code> attribute to return a vector of bridge types that are added by <a href="#MathOptInterface.Bridges.full_bridge_optimizer"><code>full_bridge_optimizer</code></a> in addition to the list of default bridges.</p><p>Note that optimizers implementing <code>ListOfNonstandardBridges</code> may require package-specific functions or sets to be used if the non-standard bridges are not added. Therefore, you are recommended to use <code>model = MOI.instantiate(Package.Optimizer; with_bridge_type = T)</code> instead of <code>model = MOI.instantiate(Package.Optimizer)</code>. See <a href="../../../reference/models/#MathOptInterface.instantiate"><code>MOI.instantiate</code></a>.</p><p><strong>Examples</strong></p><p><strong>An optimizer using a non-default bridge in <code>MOI.Bridges</code></strong></p><p>Solvers supporting <a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction</code></a> can support <a href="../../../reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a> and <a href="../../../reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a> by defining:</p><pre><code class="language-julia hljs">function MOI.get(::MyQuadraticOptimizer, ::ListOfNonstandardBridges{Float64})
+    return Type[
+        MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge{Float64},
+        MOI.Bridges.Constraint.RSOCtoNonConvexQuadBridge{Float64},
+    ]
+end</code></pre><p><strong>An optimizer defining an internal bridge</strong></p><p>Suppose an optimizer can exploit specific structure of a constraint, e.g., it can exploit the structure of the matrix <code>A</code> in the linear system of equations <code>A * x = b</code>.</p><p>The optimizer can define the function:</p><pre><code class="language-julia hljs">struct MatrixAffineFunction{T} &lt;: MOI.AbstractVectorFunction
+    A::SomeStructuredMatrixType{T}
+    b::Vector{T}
+end</code></pre><p>and then a bridge</p><pre><code class="language-julia hljs">struct MatrixAffineFunctionBridge{T} &lt;: MOI.Constraint.AbstractBridge
+    # ...
+end
+# ...</code></pre><p>from <code>VectorAffineFunction{T}</code> to the <code>MatrixAffineFunction</code>. Finally, it defines:</p><pre><code class="language-julia hljs">function MOI.get(::Optimizer{T}, ::ListOfNonstandardBridges{T}) where {T}
+    return Type[MatrixAffineFunctionBridge{T}]
+end</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/Bridges.jl#L67-L136">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.add_bridge" href="#MathOptInterface.Bridges.add_bridge"><code>MathOptInterface.Bridges.add_bridge</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_bridge(b::LazyBridgeOptimizer, BT::Type{&lt;:AbstractBridge})</code></pre><p>Enable the use of the bridges of type <code>BT</code> by <code>b</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/lazy_bridge_optimizer.jl#L367-L371">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.remove_bridge" href="#MathOptInterface.Bridges.remove_bridge"><code>MathOptInterface.Bridges.remove_bridge</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">remove_bridge(b::LazyBridgeOptimizer, BT::Type{&lt;:AbstractBridge})</code></pre><p>Disable the use of the bridges of type <code>BT</code> by <code>b</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/lazy_bridge_optimizer.jl#L380-L384">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.has_bridge" href="#MathOptInterface.Bridges.has_bridge"><code>MathOptInterface.Bridges.has_bridge</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">has_bridge(b::LazyBridgeOptimizer, BT::Type{&lt;:AbstractBridge})</code></pre><p>Return a <code>Bool</code> indicating whether the bridges of type <code>BT</code> are used by <code>b</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/lazy_bridge_optimizer.jl#L399-L403">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.print_active_bridges" href="#MathOptInterface.Bridges.print_active_bridges"><code>MathOptInterface.Bridges.print_active_bridges</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">print_active_bridges([io::IO=stdout,] b::MOI.Bridges.LazyBridgeOptimizer)</code></pre><p>Print the set of bridges that are active in the model <code>b</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/debug.jl#L538-L542">source</a></section><section><div><pre><code class="language-julia hljs">print_active_bridges(
+    [io::IO=stdout,]
+    b::MOI.Bridges.LazyBridgeOptimizer,
+    F::Type{&lt;:MOI.AbstractFunction}
+)</code></pre><p>Print the set of bridges required for an objective function of type <code>F</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/debug.jl#L580-L588">source</a></section><section><div><pre><code class="language-julia hljs">print_active_bridges(
+    [io::IO=stdout,]
+    b::MOI.Bridges.LazyBridgeOptimizer,
+    F::Type{&lt;:MOI.AbstractFunction},
+    S::Type{&lt;:MOI.AbstractSet},
+)</code></pre><p>Print the set of bridges required for a constraint of type <code>F</code>-in-<code>S</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/debug.jl#L624-L633">source</a></section><section><div><pre><code class="language-julia hljs">print_active_bridges(
+    [io::IO=stdout,]
+    b::MOI.Bridges.LazyBridgeOptimizer,
+    S::Type{&lt;:MOI.AbstractSet}
+)</code></pre><p>Print the set of bridges required for a variable constrained to set <code>S</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/debug.jl#L686-L694">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.print_graph" href="#MathOptInterface.Bridges.print_graph"><code>MathOptInterface.Bridges.print_graph</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">print_graph([io::IO = stdout,] b::LazyBridgeOptimizer)</code></pre><p>Print the hyper-graph containing all variable, constraint, and objective types that could be obtained by bridging the variables, constraints, and objectives that are present in the model by all the bridges added to <code>b</code>.</p><p>Each node in the hyper-graph corresponds to a variable, constraint, or objective type.</p><ul><li>Variable nodes are indicated by <code>[ ]</code></li><li>Constraint nodes are indicated by <code>( )</code></li><li>Objective nodes are indicated by <code>| |</code></li></ul><p>The number inside each pair of brackets is an index of the node in the hyper-graph.</p><p>Note that this hyper-graph is the full list of possible transformations. When the bridged model is created, we select the shortest hyper-path(s) from this graph, so many nodes may be un-used.</p><p>To see which nodes are used, call <a href="#MathOptInterface.Bridges.print_active_bridges"><code>print_active_bridges</code></a>.</p><p>For more information, see Legat, B., Dowson, O., Garcia, J., and Lubin, M. (2020).  &quot;MathOptInterface: a data structure for mathematical optimization problems.&quot; URL: <a href="https://arxiv.org/abs/2002.03447">https://arxiv.org/abs/2002.03447</a></p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/debug.jl#L79-L105">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.debug_supports_constraint" href="#MathOptInterface.Bridges.debug_supports_constraint"><code>MathOptInterface.Bridges.debug_supports_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">debug_supports_constraint(
+    b::LazyBridgeOptimizer,
+    F::Type{&lt;:MOI.AbstractFunction},
+    S::Type{&lt;:MOI.AbstractSet};
+    io::IO = Base.stdout,
+)</code></pre><p>Prints to <code>io</code> explanations for the value of <a href="../../../reference/constraints/#MathOptInterface.supports_constraint"><code>MOI.supports_constraint</code></a> with the same arguments.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/debug.jl#L479-L489">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.debug_supports" href="#MathOptInterface.Bridges.debug_supports"><code>MathOptInterface.Bridges.debug_supports</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">debug_supports(
+    b::LazyBridgeOptimizer,
+    ::MOI.ObjectiveFunction{F};
+    io::IO = Base.stdout,
+) where F</code></pre><p>Prints to <code>io</code> explanations for the value of <a href="../../../reference/models/#MathOptInterface.supports"><code>MOI.supports</code></a> with the same arguments.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/debug.jl#L519-L528">source</a></section></article><h2 id="constraint_set_map"><a class="docs-heading-anchor" href="#constraint_set_map">SetMap API</a><a id="constraint_set_map-1"></a><a class="docs-heading-anchor-permalink" href="#constraint_set_map" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.map_set" href="#MathOptInterface.Bridges.map_set"><code>MathOptInterface.Bridges.map_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">map_set(::Type{BT}, set) where {BT}</code></pre><p>Return the image of <code>set</code> through the linear map <code>A</code> defined in <a href="#MathOptInterface.Bridges.Variable.SetMapBridge"><code>Variable.SetMapBridge</code></a> and <a href="#MathOptInterface.Bridges.Constraint.SetMapBridge"><code>Constraint.SetMapBridge</code></a>. This is used for bridging the constraint and setting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintSet"><code>MOI.ConstraintSet</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/set_map.jl#L7-L14">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.inverse_map_set" href="#MathOptInterface.Bridges.inverse_map_set"><code>MathOptInterface.Bridges.inverse_map_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">inverse_map_set(::Type{BT}, set) where {BT}</code></pre><p>Return the preimage of <code>set</code> through the linear map <code>A</code> defined in <a href="#MathOptInterface.Bridges.Variable.SetMapBridge"><code>Variable.SetMapBridge</code></a> and <a href="#MathOptInterface.Bridges.Constraint.SetMapBridge"><code>Constraint.SetMapBridge</code></a>. This is used for getting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintSet"><code>MOI.ConstraintSet</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/set_map.jl#L17-L23">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.map_function" href="#MathOptInterface.Bridges.map_function"><code>MathOptInterface.Bridges.map_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">map_function(::Type{BT}, func) where {BT}</code></pre><p>Return the image of <code>func</code> through the linear map <code>A</code> defined in <a href="#MathOptInterface.Bridges.Variable.SetMapBridge"><code>Variable.SetMapBridge</code></a> and <a href="#MathOptInterface.Bridges.Constraint.SetMapBridge"><code>Constraint.SetMapBridge</code></a>. This is used for getting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>MOI.ConstraintPrimal</code></a> of variable bridges. For constraint bridges, this is used for bridging the constraint, setting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>MOI.ConstraintFunction</code></a> and <a href="../../../reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>MOI.ConstraintPrimalStart</code></a> and modifying the function with <a href="../../../reference/modification/#MathOptInterface.modify"><code>MOI.modify</code></a>.</p><pre><code class="nohighlight hljs">map_function(::Type{BT}, func, i::IndexInVector) where {BT}</code></pre><p>Return the scalar function at the <code>i</code>th index of the vector function that would be returned by <code>map_function(BT, func)</code> except that it may compute the <code>i</code>th element. This is used by <a href="#MathOptInterface.Bridges.bridged_function"><code>bridged_function</code></a> and for getting the <a href="../../../reference/variables/#MathOptInterface.VariablePrimal"><code>MOI.VariablePrimal</code></a> and <a href="../../../reference/variables/#MathOptInterface.VariablePrimalStart"><code>MOI.VariablePrimalStart</code></a> of variable bridges.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/set_map.jl#L26-L44">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.inverse_map_function" href="#MathOptInterface.Bridges.inverse_map_function"><code>MathOptInterface.Bridges.inverse_map_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">inverse_map_function(::Type{BT}, func) where {BT}</code></pre><p>Return the image of <code>func</code> through the inverse of the linear map <code>A</code> defined in <a href="#MathOptInterface.Bridges.Variable.SetMapBridge"><code>Variable.SetMapBridge</code></a> and <a href="#MathOptInterface.Bridges.Constraint.SetMapBridge"><code>Constraint.SetMapBridge</code></a>. This is used by <a href="#MathOptInterface.Bridges.Variable.unbridged_map"><code>Variable.unbridged_map</code></a> and for setting the <a href="../../../reference/variables/#MathOptInterface.VariablePrimalStart"><code>MOI.VariablePrimalStart</code></a> of variable bridges and for getting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>MOI.ConstraintFunction</code></a>, the <a href="../../../reference/constraints/#MathOptInterface.ConstraintPrimal"><code>MOI.ConstraintPrimal</code></a> and the <a href="../../../reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>MOI.ConstraintPrimalStart</code></a> of constraint bridges.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/set_map.jl#L51-L61">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.adjoint_map_function" href="#MathOptInterface.Bridges.adjoint_map_function"><code>MathOptInterface.Bridges.adjoint_map_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">adjoint_map_function(::Type{BT}, func) where {BT}</code></pre><p>Return the image of <code>func</code> through the adjoint of the linear map <code>A</code> defined in <a href="#MathOptInterface.Bridges.Variable.SetMapBridge"><code>Variable.SetMapBridge</code></a> and <a href="#MathOptInterface.Bridges.Constraint.SetMapBridge"><code>Constraint.SetMapBridge</code></a>. This is used for getting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintDual"><code>MOI.ConstraintDual</code></a> and <a href="../../../reference/constraints/#MathOptInterface.ConstraintDualStart"><code>MOI.ConstraintDualStart</code></a> of constraint bridges.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/set_map.jl#L64-L71">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.inverse_adjoint_map_function" href="#MathOptInterface.Bridges.inverse_adjoint_map_function"><code>MathOptInterface.Bridges.inverse_adjoint_map_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">inverse_adjoint_map_function(::Type{BT}, func) where {BT}</code></pre><p>Return the image of <code>func</code> through the inverse of the adjoint of the linear map <code>A</code> defined in <a href="#MathOptInterface.Bridges.Variable.SetMapBridge"><code>Variable.SetMapBridge</code></a> and <a href="#MathOptInterface.Bridges.Constraint.SetMapBridge"><code>Constraint.SetMapBridge</code></a>. This is used for getting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintDual"><code>MOI.ConstraintDual</code></a> of variable bridges and setting the <a href="../../../reference/constraints/#MathOptInterface.ConstraintDualStart"><code>MOI.ConstraintDualStart</code></a> of constraint bridges.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/set_map.jl#L74-L82">source</a></section></article><h2 id="Bridging-graph-API"><a class="docs-heading-anchor" href="#Bridging-graph-API">Bridging graph API</a><a id="Bridging-graph-API-1"></a><a class="docs-heading-anchor-permalink" href="#Bridging-graph-API" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Graph" href="#MathOptInterface.Bridges.Graph"><code>MathOptInterface.Bridges.Graph</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Graph()</code></pre><p>A type-stable datastructure for computing the shortest hyperpath problem.</p><p><strong>Nodes</strong></p><p>There are three types of nodes in the graph:</p><ul><li><a href="#MathOptInterface.Bridges.VariableNode"><code>VariableNode</code></a></li><li><a href="#MathOptInterface.Bridges.ConstraintNode"><code>ConstraintNode</code></a></li><li><a href="#MathOptInterface.Bridges.ObjectiveNode"><code>ObjectiveNode</code></a></li></ul><p>Add nodes to the graph using <a href="#MathOptInterface.Bridges.add_node"><code>add_node</code></a>.</p><p><strong>Edges</strong></p><p>There are two types of edges in the graph:</p><ul><li><a href="#MathOptInterface.Bridges.Edge"><code>Edge</code></a></li><li><a href="#MathOptInterface.Bridges.ObjectiveEdge"><code>ObjectiveEdge</code></a></li></ul><p>Add edges to the graph using <a href="#MathOptInterface.Bridges.add_edge"><code>add_edge</code></a>.</p><p>For the ability to add a variable constrained  on creation as a free variable followed by a constraint, use <a href="#MathOptInterface.Bridges.set_variable_constraint_node"><code>set_variable_constraint_node</code></a>.</p><p><strong>Optimal hyper-edges</strong></p><p>Use <a href="#MathOptInterface.Bridges.bridge_index"><code>bridge_index</code></a> to compute the minimum-cost bridge leaving a node.</p><p>Note that <a href="#MathOptInterface.Bridges.bridge_index"><code>bridge_index</code></a> lazy runs a Bellman-Ford algorithm to compute the set of minimum cost edges. Thus, the first call to <a href="#MathOptInterface.Bridges.bridge_index"><code>bridge_index</code></a> after adding new nodes or edges will take longer than subsequent calls.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L102-L136">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.VariableNode" href="#MathOptInterface.Bridges.VariableNode"><code>MathOptInterface.Bridges.VariableNode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">VariableNode(index::Int)</code></pre><p>A node in <a href="#MathOptInterface.Bridges.Graph"><code>Graph</code></a> representing a variable constrained on creation.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L12-L16">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.ConstraintNode" href="#MathOptInterface.Bridges.ConstraintNode"><code>MathOptInterface.Bridges.ConstraintNode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintNode(index::Int)</code></pre><p>A node in <a href="#MathOptInterface.Bridges.Graph"><code>Graph</code></a> representing a constraint.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L21-L25">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.ObjectiveNode" href="#MathOptInterface.Bridges.ObjectiveNode"><code>MathOptInterface.Bridges.ObjectiveNode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveNode(index::Int)</code></pre><p>A node in <a href="#MathOptInterface.Bridges.Graph"><code>Graph</code></a> representing an objective function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L30-L34">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.Edge" href="#MathOptInterface.Bridges.Edge"><code>MathOptInterface.Bridges.Edge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Edge(
+    bridge_index::Int,
+    added_variables::Vector{VariableNode},
+    added_constraints::Vector{ConstraintNode},
+    cost::Float64 = 1.0,
+)</code></pre><p>Return a new datastructure representing an edge in <a href="#MathOptInterface.Bridges.Graph"><code>Graph</code></a> that starts at a <a href="#MathOptInterface.Bridges.VariableNode"><code>VariableNode</code></a> or a <a href="#MathOptInterface.Bridges.ConstraintNode"><code>ConstraintNode</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L41-L51">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.ObjectiveEdge" href="#MathOptInterface.Bridges.ObjectiveEdge"><code>MathOptInterface.Bridges.ObjectiveEdge</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveEdge(
+    bridge_index::Int,
+    added_variables::Vector{VariableNode},
+    added_constraints::Vector{ConstraintNode},
+)</code></pre><p>Return a new datastructure representing an edge in <a href="#MathOptInterface.Bridges.Graph"><code>Graph</code></a> that starts at an <a href="#MathOptInterface.Bridges.ObjectiveNode"><code>ObjectiveNode</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L68-L77">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.add_node" href="#MathOptInterface.Bridges.add_node"><code>MathOptInterface.Bridges.add_node</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_node(graph::Graph, ::Type{VariableNode})::VariableNode
+add_node(graph::Graph, ::Type{ConstraintNode})::ConstraintNode
+add_node(graph::Graph, ::Type{ObjectiveNode})::ObjectiveNode</code></pre><p>Add a new node to <code>graph</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L238-L244">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.add_edge" href="#MathOptInterface.Bridges.add_edge"><code>MathOptInterface.Bridges.add_edge</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_edge(graph::Graph, node::VariableNode, edge::Edge)::Nothing
+add_edge(graph::Graph, node::ConstraintNode, edge::Edge)::Nothing
+add_edge(graph::Graph, node::ObjectiveNode, edge::ObjectiveEdge)::Nothing</code></pre><p>Add <code>edge</code> to <code>graph</code>, where <code>edge</code> starts at <code>node</code> and connects to the nodes defined in <code>edge</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L215-L222">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.set_variable_constraint_node" href="#MathOptInterface.Bridges.set_variable_constraint_node"><code>MathOptInterface.Bridges.set_variable_constraint_node</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_variable_constraint_node(
+    graph::Graph,
+    variable_node::VariableNode,
+    constraint_node::ConstraintNode,
+    cost::Int,
+)</code></pre><p>As an alternative to <code>variable_node</code>, add a virtual edge to <code>graph</code> that represents adding a free variable, followed by a constraint of type <code>constraint_node</code>, with bridging cost <code>cost</code>.</p><p><strong>Why is this needed?</strong></p><p>Variables can either be added as a variable constrained on creation, or as a free variable which then has a constraint added to it.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L270-L286">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.bridge_index" href="#MathOptInterface.Bridges.bridge_index"><code>MathOptInterface.Bridges.bridge_index</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">bridge_index(graph::Graph, node::VariableNode)::Int
+bridge_index(graph::Graph, node::ConstraintNode)::Int
+bridge_index(graph::Graph, node::ObjectiveNode)::Int</code></pre><p>Return the optimal index of the bridge to chose from <code>node</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L303-L309">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Bridges.is_variable_edge_best" href="#MathOptInterface.Bridges.is_variable_edge_best"><code>MathOptInterface.Bridges.is_variable_edge_best</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_variable_edge_best(graph::Graph, node::VariableNode)::Bool</code></pre><p>Return a <code>Bool</code> indicating whether <code>node</code> should be added as a variable constrained on creation, or as a free variable followed by a constraint.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Bridges/graph.jl#L325-L330">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../list_of_bridges/">« List of bridges</a><a class="docs-footer-nextpage" href="../../FileFormats/overview/">Overview »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/FileFormats/overview/index.html b/previews/PR3525/moi/submodules/FileFormats/overview/index.html
new file mode 100644
index 00000000000..13e49038bcb
--- /dev/null
+++ b/previews/PR3525/moi/submodules/FileFormats/overview/index.html
@@ -0,0 +1,141 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Overview · JuMP</title><meta name="title" content="Overview · JuMP"/><meta property="og:title" content="Overview · JuMP"/><meta property="twitter:title" content="Overview · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Overview</a><ul class="internal"><li><a class="tocitem" href="#Supported-file-types"><span>Supported file types</span></a></li><li><a class="tocitem" href="#Write-to-file"><span>Write to file</span></a></li><li><a class="tocitem" href="#Read-from-file"><span>Read from file</span></a></li><li><a class="tocitem" href="#Detecting-the-file-type-automatically"><span>Detecting the file-type automatically</span></a></li><li><a class="tocitem" href="#Unsupported-constraints"><span>Unsupported constraints</span></a></li><li><a class="tocitem" href="#Read-and-write-to-io"><span>Read and write to <code>io</code></span></a></li><li><a class="tocitem" href="#Validating-MOF-files"><span>Validating MOF files</span></a></li></ul></li><li><a class="tocitem" href="../reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">FileFormats</a></li><li class="is-active"><a href>Overview</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Overview</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/FileFormats/overview.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-FileFormats-submodule"><a class="docs-heading-anchor" href="#The-FileFormats-submodule">The FileFormats submodule</a><a id="The-FileFormats-submodule-1"></a><a class="docs-heading-anchor-permalink" href="#The-FileFormats-submodule" title="Permalink"></a></h1><p>The <code>FileFormats</code> module provides functions for reading and writing MOI models using <a href="../../../reference/models/#MathOptInterface.write_to_file"><code>write_to_file</code></a> and <a href="../../../reference/models/#MathOptInterface.read_from_file"><code>read_from_file</code></a>.</p><h2 id="Supported-file-types"><a class="docs-heading-anchor" href="#Supported-file-types">Supported file types</a><a id="Supported-file-types-1"></a><a class="docs-heading-anchor-permalink" href="#Supported-file-types" title="Permalink"></a></h2><p>You must read and write files to a <a href="../reference/#MathOptInterface.FileFormats.Model"><code>FileFormats.Model</code></a> object. Specific the file-type by passing a <a href="../reference/#MathOptInterface.FileFormats.FileFormat"><code>FileFormats.FileFormat</code></a> enum. For example:</p><p><strong>The Conic Benchmark Format</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_CBF)
+A Conic Benchmark Format (CBF) model</code></pre><p><strong>The LP file format</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_LP)
+A .LP-file model</code></pre><p><strong>The MathOptFormat file format</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MOF)
+A MathOptFormat Model</code></pre><p><strong>The MPS file format</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MPS)
+A Mathematical Programming System (MPS) model</code></pre><p><strong>The NL file format</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_NL)
+An AMPL (.nl) model</code></pre><p><strong>The REW file format</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_REW)
+A Mathematical Programming System (MPS) model</code></pre><p>Note that the <a href="https://www.gurobi.com/documentation/9.5/refman/rew_format.html">REW format</a> is identical to the MPS file format, except that all names are replaced with generic identifiers.</p><p><strong>The SDPA file format</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_SDPA)
+A SemiDefinite Programming Algorithm Format (SDPA) model</code></pre><h2 id="Write-to-file"><a class="docs-heading-anchor" href="#Write-to-file">Write to file</a><a id="Write-to-file-1"></a><a class="docs-heading-anchor-permalink" href="#Write-to-file" title="Permalink"></a></h2><p>To write a model <code>src</code> to a <a href="https://jump.dev/MathOptFormat/">MathOptFormat file</a>, use:</p><pre><code class="language-julia-repl hljs">julia&gt; src = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; MOI.add_variable(src)
+MOI.VariableIndex(1)
+
+julia&gt; dest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MOF)
+A MathOptFormat Model
+
+julia&gt; MOI.copy_to(dest, src)
+MathOptInterface.Utilities.IndexMap with 1 entry:
+  MOI.VariableIndex(1) =&gt; MOI.VariableIndex(1)
+
+julia&gt; MOI.write_to_file(dest, &quot;file.mof.json&quot;)
+
+julia&gt; print(read(&quot;file.mof.json&quot;, String))
+{
+  &quot;name&quot;: &quot;MathOptFormat Model&quot;,
+  &quot;version&quot;: {
+    &quot;major&quot;: 1,
+    &quot;minor&quot;: 5
+  },
+  &quot;variables&quot;: [
+    {
+      &quot;name&quot;: &quot;x1&quot;
+    }
+  ],
+  &quot;objective&quot;: {
+    &quot;sense&quot;: &quot;feasibility&quot;
+  },
+  &quot;constraints&quot;: []
+}</code></pre><h2 id="Read-from-file"><a class="docs-heading-anchor" href="#Read-from-file">Read from file</a><a id="Read-from-file-1"></a><a class="docs-heading-anchor-permalink" href="#Read-from-file" title="Permalink"></a></h2><p>To read a MathOptFormat file, use:</p><pre><code class="language-julia-repl hljs">julia&gt; dest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MOF)
+A MathOptFormat Model
+
+julia&gt; MOI.read_from_file(dest, &quot;file.mof.json&quot;)
+
+julia&gt; MOI.get(dest, MOI.ListOfVariableIndices())
+1-element Vector{MathOptInterface.VariableIndex}:
+ MOI.VariableIndex(1)
+
+julia&gt; rm(&quot;file.mof.json&quot;)  # Clean up after ourselves.</code></pre><h2 id="Detecting-the-file-type-automatically"><a class="docs-heading-anchor" href="#Detecting-the-file-type-automatically">Detecting the file-type automatically</a><a id="Detecting-the-file-type-automatically-1"></a><a class="docs-heading-anchor-permalink" href="#Detecting-the-file-type-automatically" title="Permalink"></a></h2><p>Instead of the <code>format</code> keyword, you can also use the <code>filename</code> keyword argument to <a href="../reference/#MathOptInterface.FileFormats.Model"><code>FileFormats.Model</code></a>. This will attempt to automatically guess the format from the file extension. For example:</p><pre><code class="language-julia-repl hljs">julia&gt; src = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; dest = MOI.FileFormats.Model(filename = &quot;file.cbf.gz&quot;)
+A Conic Benchmark Format (CBF) model
+
+julia&gt; MOI.copy_to(dest, src)
+MathOptInterface.Utilities.IndexMap()
+
+julia&gt; MOI.write_to_file(dest, &quot;file.cbf.gz&quot;)
+
+julia&gt; src_2 = MOI.FileFormats.Model(filename = &quot;file.cbf.gz&quot;)
+A Conic Benchmark Format (CBF) model
+
+julia&gt; src = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; dest = MOI.FileFormats.Model(filename = &quot;file.cbf.gz&quot;)
+A Conic Benchmark Format (CBF) model
+
+julia&gt; MOI.copy_to(dest, src)
+MathOptInterface.Utilities.IndexMap()
+
+julia&gt; MOI.write_to_file(dest, &quot;file.cbf.gz&quot;)
+
+julia&gt; src_2 = MOI.FileFormats.Model(filename = &quot;file.cbf.gz&quot;)
+A Conic Benchmark Format (CBF) model
+
+julia&gt; MOI.read_from_file(src_2, &quot;file.cbf.gz&quot;)
+
+julia&gt; rm(&quot;file.cbf.gz&quot;)  # Clean up after ourselves.</code></pre><p>Note how the compression format (GZip) is also automatically detected from the filename.</p><h2 id="Unsupported-constraints"><a class="docs-heading-anchor" href="#Unsupported-constraints">Unsupported constraints</a><a id="Unsupported-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Unsupported-constraints" title="Permalink"></a></h2><p>In some cases <code>src</code> may contain constraints that are not supported by the file format (for example, the CBF format supports integer variables but not binary). If so, copy <code>src</code> to a bridged model using <a href="../../Bridges/reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>Bridges.full_bridge_optimizer</code></a>:</p><pre><code class="language-julia hljs">src = MOI.Utilities.Model{Float64}()
+x = MOI.add_variable(model)
+MOI.add_constraint(model, x, MOI.ZeroOne())
+dest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_CBF)
+bridged = MOI.Bridges.full_bridge_optimizer(dest, Float64)
+MOI.copy_to(bridged, src)
+MOI.write_to_file(dest, &quot;my_model.cbf&quot;)</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Even after bridging, it may still not be possible to write the model to file because of unsupported constraints (for example, PSD variables in the LP file format).</p></div></div><h2 id="Read-and-write-to-io"><a class="docs-heading-anchor" href="#Read-and-write-to-io">Read and write to <code>io</code></a><a id="Read-and-write-to-io-1"></a><a class="docs-heading-anchor-permalink" href="#Read-and-write-to-io" title="Permalink"></a></h2><p>In addition to <a href="../../../reference/models/#MathOptInterface.write_to_file"><code>write_to_file</code></a> and <a href="../../../reference/models/#MathOptInterface.read_from_file"><code>read_from_file</code></a>, you can read and write directly from <code>IO</code> streams using <code>Base.write</code> and <code>Base.read!</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; src = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}
+
+julia&gt; dest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MPS)
+A Mathematical Programming System (MPS) model
+
+julia&gt; MOI.copy_to(dest, src)
+MathOptInterface.Utilities.IndexMap()
+
+julia&gt; io = IOBuffer();
+
+julia&gt; write(io, dest)
+
+julia&gt; seekstart(io);
+
+julia&gt; src_2 = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MPS)
+A Mathematical Programming System (MPS) model
+
+julia&gt; read!(io, src_2);</code></pre><h2 id="Validating-MOF-files"><a class="docs-heading-anchor" href="#Validating-MOF-files">Validating MOF files</a><a id="Validating-MOF-files-1"></a><a class="docs-heading-anchor-permalink" href="#Validating-MOF-files" title="Permalink"></a></h2><p>MathOptFormat files are governed by a schema. Use <a href="https://github.com/fredo-dedup/JSONSchema.jl">JSONSchema.jl</a> to check if a <code>.mof.json</code> file satisfies the schema.</p><p>First, construct the schema object as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; import JSON, JSONSchema
+
+julia&gt; schema = JSONSchema.Schema(JSON.parsefile(MOI.FileFormats.MOF.SCHEMA_PATH))
+A JSONSchema</code></pre><p>Then, check if a model file is valid using <code>isvalid</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; good_model = JSON.parse(&quot;&quot;&quot;
+       {
+         &quot;version&quot;: {
+           &quot;major&quot;: 1,
+           &quot;minor&quot;: 5
+         },
+         &quot;variables&quot;: [{&quot;name&quot;: &quot;x&quot;}],
+         &quot;objective&quot;: {&quot;sense&quot;: &quot;feasibility&quot;},
+         &quot;constraints&quot;: []
+       }
+       &quot;&quot;&quot;);
+
+julia&gt; isvalid(schema, good_model)
+true</code></pre><p>If we construct an invalid file, for example by mis-typing <code>name</code> as <code>NaMe</code>, the validation fails:</p><pre><code class="language-julia-repl hljs">julia&gt; bad_model = JSON.parse(&quot;&quot;&quot;
+       {
+         &quot;version&quot;: {
+           &quot;major&quot;: 1,
+           &quot;minor&quot;: 5
+         },
+         &quot;variables&quot;: [{&quot;NaMe&quot;: &quot;x&quot;}],
+         &quot;objective&quot;: {&quot;sense&quot;: &quot;feasibility&quot;},
+         &quot;constraints&quot;: []
+       }
+       &quot;&quot;&quot;);
+
+julia&gt; isvalid(schema, bad_model)
+false</code></pre><p>Use <code>JSONSchema.validate</code> to obtain more insight into why the validation failed:</p><pre><code class="language-julia-repl hljs">julia&gt; JSONSchema.validate(schema, bad_model)
+Validation failed:
+path:         [variables][1]
+instance:     Dict{String, Any}(&quot;NaMe&quot; =&gt; &quot;x&quot;)
+schema key:   required
+schema value: Any[&quot;name&quot;]</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../Bridges/reference/">« API Reference</a><a class="docs-footer-nextpage" href="../reference/">API Reference »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/FileFormats/reference/index.html b/previews/PR3525/moi/submodules/FileFormats/reference/index.html
new file mode 100644
index 00000000000..436032fdf15
--- /dev/null
+++ b/previews/PR3525/moi/submodules/FileFormats/reference/index.html
@@ -0,0 +1,22 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>API Reference · JuMP</title><meta name="title" content="API Reference · JuMP"/><meta property="og:title" content="API Reference · JuMP"/><meta property="twitter:title" content="API Reference · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../overview/">Overview</a></li><li class="is-active"><a class="tocitem" href>API Reference</a><ul class="internal"><li><a class="tocitem" href="#File-Formats"><span>File Formats</span></a></li><li><a class="tocitem" href="#Other-helpers"><span>Other helpers</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">FileFormats</a></li><li class="is-active"><a href>API Reference</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>API Reference</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/FileFormats/reference.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h2 id="File-Formats"><a class="docs-heading-anchor" href="#File-Formats">File Formats</a><a id="File-Formats-1"></a><a class="docs-heading-anchor-permalink" href="#File-Formats" title="Permalink"></a></h2><p>Functions to help read and write MOI models to/from various file formats. See <a href="../overview/#The-FileFormats-submodule">The FileFormats submodule</a> for more details.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.Model" href="#MathOptInterface.FileFormats.Model"><code>MathOptInterface.FileFormats.Model</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">Model(
+    ;
+    format::FileFormat = FORMAT_AUTOMATIC,
+    filename::Union{Nothing, String} = nothing,
+    kwargs...
+)</code></pre><p>Return model corresponding to the <code>FileFormat</code> <code>format</code>, or, if <code>format == FORMAT_AUTOMATIC</code>, guess the format from <code>filename</code>.</p><p>The <code>filename</code> argument is only needed if <code>format == FORMAT_AUTOMATIC</code>.</p><p><code>kwargs</code> are passed to the underlying model constructor.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/FileFormats.jl#L49-L63">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.FileFormat" href="#MathOptInterface.FileFormats.FileFormat"><code>MathOptInterface.FileFormats.FileFormat</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">FileFormat</code></pre><p>List of accepted export formats.</p><ul><li><code>FORMAT_AUTOMATIC</code>: try to detect the file format based on the file name</li><li><code>FORMAT_CBF</code>: the Conic Benchmark format</li><li><code>FORMAT_LP</code>: the LP file format</li><li><code>FORMAT_MOF</code>: the MathOptFormat file format</li><li><code>FORMAT_MPS</code>: the MPS file format</li><li><code>FORMAT_NL</code>: the AMPL .nl file format</li><li><code>FORMAT_REW</code>: the .rew file format, which is MPS with generic names</li><li><code>FORMAT_SDPA</code>: the SemiDefinite Programming Algorithm format</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/FileFormats.jl#L23-L36">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.CBF.Model" href="#MathOptInterface.FileFormats.CBF.Model"><code>MathOptInterface.FileFormats.CBF.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model()</code></pre><p>Create an empty instance of <code>FileFormats.CBF.Model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/CBF/CBF.jl#L57-L61">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.LP.Model" href="#MathOptInterface.FileFormats.LP.Model"><code>MathOptInterface.FileFormats.LP.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model(; kwargs...)</code></pre><p>Create an empty instance of FileFormats.LP.Model.</p><p>Keyword arguments are:</p><ul><li><p><code>maximum_length::Int=255</code>: the maximum length for the name of a variable. lp_solve 5.0 allows only 16 characters, while CPLEX 12.5+ allow 255.</p></li><li><p><code>warn::Bool=false</code>: print a warning when variables or constraints are renamed.</p></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/LP/LP.jl#L52-L63">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.MOF.Model" href="#MathOptInterface.FileFormats.MOF.Model"><code>MathOptInterface.FileFormats.MOF.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model(; kwargs...)</code></pre><p>Create an empty instance of FileFormats.MOF.Model.</p><p>Keyword arguments are:</p><ul><li><code>print_compact::Bool=false</code>: print the JSON file in a compact format without spaces or newlines.</li><li><code>warn::Bool=false</code>: print a warning when variables or constraints are renamed</li><li><code>differentiation_backend::MOI.Nonlinear.AbstractAutomaticDifferentiation = MOI.Nonlinear.SparseReverseMode()</code>: automatic differentiation backend to use when reading models with nonlinear constraints and objectives.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/MOF/MOF.jl#L113-L126">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.MPS.Model" href="#MathOptInterface.FileFormats.MPS.Model"><code>MathOptInterface.FileFormats.MPS.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model(; kwargs...)</code></pre><p>Create an empty instance of FileFormats.MPS.Model.</p><p>Keyword arguments are:</p><ul><li><code>warn::Bool=false</code>: print a warning when variables or constraints are renamed.</li><li><code>print_objsense::Bool=false</code>: print the OBJSENSE section when writing</li><li><code>generic_names::Bool=false</code>: strip all names in the model and replace them with the generic names <code>C$i</code> and <code>R$i</code> for the i&#39;th column and row respectively.</li><li><code>quadratic_format::QuadraticFormat = kQuadraticFormatGurobi</code>: specify the solver-specific extension used when writing the quadratic components of the model. Options are <code>kQuadraticFormatGurobi</code>, <code>kQuadraticFormatCPLEX</code>, and <code>kQuadraticFormatMosek</code>.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/MPS/MPS.jl#L113-L129">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.NL.Model" href="#MathOptInterface.FileFormats.NL.Model"><code>MathOptInterface.FileFormats.NL.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model(; use_nlp_block::Bool = true)</code></pre><p>Create a new Optimizer object.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/NL/NL.jl#L121-L125">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.SDPA.Model" href="#MathOptInterface.FileFormats.SDPA.Model"><code>MathOptInterface.FileFormats.SDPA.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model(; number_type::Type = Float64)</code></pre><p>Create an empty instance of <code>FileFormats.SDPA.Model{number_type}</code>.</p><p>It is important to be aware that the SDPA file format is interpreted in <em>geometric</em> form and not <em>standard conic</em> form. The <em>standard conic</em> form and <em>geometric conic</em> form are two dual standard forms for semidefinite programs (SDPs). The <em>geometric conic</em> form of an SDP is as follows:</p><p class="math-container">\[\begin{align}
+&amp; \min_{y \in \mathbb{R}^m} &amp; b^T y
+\\
+&amp; \;\;\text{s.t.} &amp; \sum_{i=1}^m A_i y_i - C &amp; \in \mathbb{K}
+\end{align}\]</p><p>where <span>$\mathcal{K}$</span> is a cartesian product of nonnegative orthant and positive semidefinite matrices that align with a block diagonal structure shared with the matrices <code>A_i</code> and <code>C</code>.</p><p>In other words, the geometric conic form contains free variables and affine constraints in either the nonnegative orthant or the positive semidefinite cone. That is, in the MathOptInterface&#39;s terminology, <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a> and <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> constraints.</p><p>The corresponding <em>standard conic</em> form of the dual SDP is as follows:</p><p class="math-container">\[\begin{align}
+&amp; \max_{X \in \mathbb{K}} &amp; \text{tr}(CX)
+\\
+&amp; \;\;\text{s.t.} &amp; \text{tr}(A_iX) &amp; = b_i &amp; i = 1, \ldots, m.
+\end{align}\]</p><p>In other words, the standard conic form contains nonnegative and positive semidefinite variables with equality constraints. That is, in the MathOptInterface&#39;s terminology, <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> and <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a> constraints.</p><p>If a model is in standard conic form, use <code>Dualization.jl</code> to transform it into the geometric conic form before writting it. Otherwise, the nonnegative (resp. positive semidefinite) variables will be bridged into free variables with affine constraints constraining them to belong to the nonnegative orthant (resp. positive semidefinite cone) by the <a href="../../Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.VectorFunctionizeBridge"><code>MOI.Bridges.Constraint.VectorFunctionizeBridge</code></a>. Moreover, equality constraints will be bridged into pairs of affine constraints in the nonnegative orthant by the <a href="../../Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitIntervalBridge"><code>MOI.Bridges.Constraint.SplitIntervalBridge</code></a> and then the <a href="../../Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.VectorizeBridge"><code>MOI.Bridges.Constraint.VectorizeBridge</code></a>.</p><p>If a solver is in standard conic form, use <code>Dualization.jl</code> to transform the model read into standard conic form before copying it to the solver. Otherwise, the free variables will be bridged into pairs of variables in the nonnegative orthant by the <a href="../../Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.FreeBridge"><code>MOI.Bridges.Variable.FreeBridge</code></a> and affine constraints will be bridged into equality constraints by creating a slack variable by the <a href="../../Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.VectorSlackBridge"><code>MOI.Bridges.Constraint.VectorSlackBridge</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/SDPA/SDPA.jl#L56-L123">source</a></section></article><h2 id="Other-helpers"><a class="docs-heading-anchor" href="#Other-helpers">Other helpers</a><a id="Other-helpers-1"></a><a class="docs-heading-anchor-permalink" href="#Other-helpers" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.FileFormats.NL.SolFileResults" href="#MathOptInterface.FileFormats.NL.SolFileResults"><code>MathOptInterface.FileFormats.NL.SolFileResults</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SolFileResults(filename::String, model::Model)</code></pre><p>Parse the <code>.sol</code> file <code>filename</code> created by solving <code>model</code> and return a <code>SolFileResults</code> struct.</p><p>The returned struct supports the <code>MOI.get</code> API for querying result attributes such as <a href="../../../reference/models/#MathOptInterface.TerminationStatus"><code>MOI.TerminationStatus</code></a>, <a href="../../../reference/variables/#MathOptInterface.VariablePrimal"><code>MOI.VariablePrimal</code></a>, and <a href="../../../reference/constraints/#MathOptInterface.ConstraintDual"><code>MOI.ConstraintDual</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/NL/sol.jl#L21-L30">source</a></section><section><div><pre><code class="language-julia hljs">SolFileResults(
+    raw_status::String,
+    termination_status::MOI.TerminationStatusCode,
+)</code></pre><p>Return a <code>SolFileResults</code> struct with <a href="../../../reference/models/#MathOptInterface.RawStatusString"><code>MOI.RawStatusString</code></a> set to <code>raw_status</code>, <a href="../../../reference/models/#MathOptInterface.TerminationStatus"><code>MOI.TerminationStatus</code></a> set to <code>termination_status</code>, and <a href="../../../reference/models/#MathOptInterface.PrimalStatus"><code>MOI.PrimalStatus</code></a> and <a href="../../../reference/models/#MathOptInterface.DualStatus"><code>MOI.DualStatus</code></a> set to <code>NO_SOLUTION</code>.</p><p>All other attributes are un-set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/FileFormats/NL/sol.jl#L35-L46">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../overview/">« Overview</a><a class="docs-footer-nextpage" href="../../Nonlinear/overview/">Overview »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Nonlinear/overview/index.html b/previews/PR3525/moi/submodules/Nonlinear/overview/index.html
new file mode 100644
index 00000000000..9ef49a0aa3a
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Nonlinear/overview/index.html
@@ -0,0 +1,187 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Overview · JuMP</title><meta name="title" content="Overview · JuMP"/><meta property="og:title" content="Overview · JuMP"/><meta property="twitter:title" content="Overview · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Overview</a><ul class="internal"><li><a class="tocitem" href="#Standard-form"><span>Standard form</span></a></li><li><a class="tocitem" href="#API-overview"><span>API overview</span></a></li><li><a class="tocitem" href="#Expression-graph-representation"><span>Expression-graph representation</span></a></li><li><a class="tocitem" href="#ReverseAD"><span>ReverseAD</span></a></li></ul></li><li><a class="tocitem" href="../reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Nonlinear</a></li><li class="is-active"><a href>Overview</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Overview</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Nonlinear/overview.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="nonlinear_developers"><a class="docs-heading-anchor" href="#nonlinear_developers">Nonlinear</a><a id="nonlinear_developers-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_developers" title="Permalink"></a></h1><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The Nonlinear submodule is experimental. Until this message is removed, breaking changes may be introduced in any minor or patch release of MathOptInterface.</p></div></div><p>The <code>Nonlinear</code> submodule contains data structures and functions for working with a nonlinear optimization problem in the form of an expression graph. This page explains the API and describes the rationale behind its design.</p><h2 id="Standard-form"><a class="docs-heading-anchor" href="#Standard-form">Standard form</a><a id="Standard-form-1"></a><a class="docs-heading-anchor-permalink" href="#Standard-form" title="Permalink"></a></h2><p><a href="https://en.wikipedia.org/wiki/Nonlinear_programming">Nonlinear programs (NLPs)</a> are a class of optimization problems in which some of the constraints or the objective function are nonlinear:</p><p class="math-container">\[\begin{align}
+    \min_{x \in \mathbb{R}^n} &amp; f_0(x) \\
+    \;\;\text{s.t.} &amp; l_j \le f_j(x) \le u_j &amp; j = 1 \ldots m
+\end{align}\]</p><p>There may be additional constraints, as well as things like variable bounds and integrality restrictions, but we do not consider them here because they are best dealt with by other components of MathOptInterface.</p><h2 id="API-overview"><a class="docs-heading-anchor" href="#API-overview">API overview</a><a id="API-overview-1"></a><a class="docs-heading-anchor-permalink" href="#API-overview" title="Permalink"></a></h2><p>The core element of the <code>Nonlinear</code> submodule is <a href="../reference/#MathOptInterface.Nonlinear.Model"><code>Nonlinear.Model</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; const Nonlinear = MOI.Nonlinear;
+
+julia&gt; model = Nonlinear.Model()
+A Nonlinear.Model with:
+ 0 objectives
+ 0 parameters
+ 0 expressions
+ 0 constraints</code></pre><p><a href="../reference/#MathOptInterface.Nonlinear.Model"><code>Nonlinear.Model</code></a> is a mutable struct that stores all of the nonlinear information added to the model.</p><h3 id="Decision-variables"><a class="docs-heading-anchor" href="#Decision-variables">Decision variables</a><a id="Decision-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Decision-variables" title="Permalink"></a></h3><p>Decision variables are represented by <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>es. The user is responsible for creating these using <code>MOI.VariableIndex(i)</code>, where <code>i</code> is the column associated with the variable.</p><h3 id="Nonlinear_Expressions"><a class="docs-heading-anchor" href="#Nonlinear_Expressions">Expressions</a><a id="Nonlinear_Expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear_Expressions" title="Permalink"></a></h3><p>The input data structure is a Julia <code>Expr</code>. The input expressions can incorporate <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>es, but these must be interpolated into the expression with <code>$</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; x = MOI.VariableIndex(1)
+MOI.VariableIndex(1)
+
+julia&gt; input = :(1 + sin($x)^2)
+:(1 + sin(MathOptInterface.VariableIndex(1)) ^ 2)</code></pre><p>There are a number of restrictions on the input <code>Expr</code>:</p><ul><li>It cannot contain macros</li><li>It cannot contain broadcasting</li><li>It cannot contain splatting (except in limited situations)</li><li>It cannot contain linear algebra, such as matrix-vector products</li><li>It cannot contain generator expressions, including <code>sum(i for i in S)</code></li></ul><p>Given an input expression, add an expression using <a href="../reference/#MathOptInterface.Nonlinear.add_expression"><code>Nonlinear.add_expression</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; expr = Nonlinear.add_expression(model, input)
+MathOptInterface.Nonlinear.ExpressionIndex(1)</code></pre><p>The return value, <code>expr</code>, is a <a href="../reference/#MathOptInterface.Nonlinear.ExpressionIndex"><code>Nonlinear.ExpressionIndex</code></a> that can then be interpolated into other input expressions.</p><p>Looking  again at <code>model</code>, we see:</p><pre><code class="language-julia-repl hljs">julia&gt; model
+A Nonlinear.Model with:
+ 0 objectives
+ 0 parameters
+ 1 expression
+ 0 constraints</code></pre><h3 id="Nonlinear_Parameters"><a class="docs-heading-anchor" href="#Nonlinear_Parameters">Parameters</a><a id="Nonlinear_Parameters-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear_Parameters" title="Permalink"></a></h3><p>In addition to constant literals like <code>1</code> or <code>1.23</code>, you can create parameters. Parameters are placeholders whose values can change before passing the expression to the solver. Create a parameter using <a href="../reference/#MathOptInterface.Nonlinear.add_parameter"><code>Nonlinear.add_parameter</code></a>, which accepts a default value:</p><pre><code class="language-julia-repl hljs">julia&gt; p = Nonlinear.add_parameter(model, 1.23)
+MathOptInterface.Nonlinear.ParameterIndex(1)</code></pre><p>The return value, <code>p</code>, is a <a href="../reference/#MathOptInterface.Nonlinear.ParameterIndex"><code>Nonlinear.ParameterIndex</code></a> that can then be interpolated into other input expressions.</p><p>Looking  again at <code>model</code>, we see:</p><pre><code class="language-julia-repl hljs">julia&gt; model
+A Nonlinear.Model with:
+ 0 objectives
+ 1 parameter
+ 1 expression
+ 0 constraints</code></pre><p>Update a parameter as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model[p]
+1.23
+
+julia&gt; model[p] = 4.56
+4.56
+
+julia&gt; model[p]
+4.56</code></pre><h3 id="Nonlinear_Objectives"><a class="docs-heading-anchor" href="#Nonlinear_Objectives">Objectives</a><a id="Nonlinear_Objectives-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear_Objectives" title="Permalink"></a></h3><p>Set a nonlinear objective using <a href="../reference/#MathOptInterface.Nonlinear.set_objective"><code>Nonlinear.set_objective</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; Nonlinear.set_objective(model, :($p + $expr + $x))
+
+julia&gt; model
+A Nonlinear.Model with:
+ 1 objective
+ 1 parameter
+ 1 expression
+ 0 constraints</code></pre><p>Clear a nonlinear objective by passing <code>nothing</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; Nonlinear.set_objective(model, nothing)
+
+julia&gt; model
+A Nonlinear.Model with:
+ 0 objectives
+ 1 parameter
+ 1 expression
+ 0 constraints</code></pre><p>But we&#39;ll re-add the objective for later:</p><pre><code class="language-julia-repl hljs">julia&gt; Nonlinear.set_objective(model, :($p + $expr + $x));</code></pre><h3 id="Nonlinear_Constraints"><a class="docs-heading-anchor" href="#Nonlinear_Constraints">Constraints</a><a id="Nonlinear_Constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear_Constraints" title="Permalink"></a></h3><p>Add a constraint using <a href="../reference/#MathOptInterface.Nonlinear.add_constraint"><code>Nonlinear.add_constraint</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; c = Nonlinear.add_constraint(model, :(1 + sqrt($x)), MOI.LessThan(2.0))
+MathOptInterface.Nonlinear.ConstraintIndex(1)
+
+julia&gt; model
+A Nonlinear.Model with:
+ 1 objective
+ 1 parameter
+ 1 expression
+ 1 constraint</code></pre><p>The return value, <code>c</code>, is a <a href="../reference/#MathOptInterface.Nonlinear.ConstraintIndex"><code>Nonlinear.ConstraintIndex</code></a> that is a unique identifier for the constraint. Interval constraints are also supported:</p><pre><code class="language-julia-repl hljs">julia&gt; c2 = Nonlinear.add_constraint(model, :(1 + sqrt($x)), MOI.Interval(-1.0, 2.0))
+MathOptInterface.Nonlinear.ConstraintIndex(2)
+
+julia&gt; model
+A Nonlinear.Model with:
+ 1 objective
+ 1 parameter
+ 1 expression
+ 2 constraints</code></pre><p>Delete a constraint using <a href="../reference/#MathOptInterface.Nonlinear.delete"><code>Nonlinear.delete</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; Nonlinear.delete(model, c2)
+
+julia&gt; model
+A Nonlinear.Model with:
+ 1 objective
+ 1 parameter
+ 1 expression
+ 1 constraint</code></pre><h3 id="User-defined-operators"><a class="docs-heading-anchor" href="#User-defined-operators">User-defined operators</a><a id="User-defined-operators-1"></a><a class="docs-heading-anchor-permalink" href="#User-defined-operators" title="Permalink"></a></h3><p>By default, <code>Nonlinear</code> supports a wide range of univariate and multivariate operators. However, you can also define your own operators by <em>registering</em> them.</p><h4 id="Univariate-operators"><a class="docs-heading-anchor" href="#Univariate-operators">Univariate operators</a><a id="Univariate-operators-1"></a><a class="docs-heading-anchor-permalink" href="#Univariate-operators" title="Permalink"></a></h4><p>Register a univariate user-defined operator using <a href="../reference/#MathOptInterface.Nonlinear.register_operator"><code>Nonlinear.register_operator</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; f(x) = 1 + sin(x)^2
+f (generic function with 1 method)
+
+julia&gt; Nonlinear.register_operator(model, :my_f, 1, f)</code></pre><p>Now, you can use <code>:my_f</code> in expressions:</p><pre><code class="language-julia-repl hljs">julia&gt; new_expr = Nonlinear.add_expression(model, :(my_f($x + 1)))
+MathOptInterface.Nonlinear.ExpressionIndex(2)</code></pre><p>By default, <code>Nonlinear</code> will compute first- and second-derivatives of the registered operator using <a href="https://github.com/JuliaDiff/ForwardDiff.jl">ForwardDiff.jl</a>. Override this by passing functions which compute the respective derivative:</p><pre><code class="language-julia-repl hljs">julia&gt; f′(x) = 2 * sin(x) * cos(x)
+f′ (generic function with 1 method)
+
+julia&gt; Nonlinear.register_operator(model, :my_f2, 1, f, f′)</code></pre><p>or</p><pre><code class="language-julia-repl hljs">julia&gt; f′′(x) = 2 * (cos(x)^2 - sin(x)^2)
+f′′ (generic function with 1 method)
+
+julia&gt; Nonlinear.register_operator(model, :my_f3, 1, f, f′, f′′)</code></pre><h4 id="Multivariate-operators"><a class="docs-heading-anchor" href="#Multivariate-operators">Multivariate operators</a><a id="Multivariate-operators-1"></a><a class="docs-heading-anchor-permalink" href="#Multivariate-operators" title="Permalink"></a></h4><p>Register a multivariate user-defined operator using <a href="../reference/#MathOptInterface.Nonlinear.register_operator"><code>Nonlinear.register_operator</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; g(x...) = x[1]^2 + x[1] * x[2] + x[2]^2
+g (generic function with 1 method)
+
+julia&gt; Nonlinear.register_operator(model, :my_g, 2, g)</code></pre><p>Now, you can use <code>:my_g</code> in expressions:</p><pre><code class="language-julia-repl hljs">julia&gt; new_expr = Nonlinear.add_expression(model, :(my_g($x + 1, $x)))
+MathOptInterface.Nonlinear.ExpressionIndex(3)</code></pre><p>By default, <code>Nonlinear</code> will compute the gradient of the registered operator using <a href="https://github.com/JuliaDiff/ForwardDiff.jl">ForwardDiff.jl</a>. (Hessian information is not supported.) Override this by passing a function to compute the gradient:</p><pre><code class="language-julia-repl hljs">julia&gt; function ∇g(ret, x...)
+           ret[1] = 2 * x[1] + x[2]
+           ret[2] = x[1] + 2 * x[2]
+           return
+       end
+∇g (generic function with 1 method)
+
+julia&gt; Nonlinear.register_operator(model, :my_g2, 2, g, ∇g)</code></pre><h3 id="Nonlinear_MOI_interface"><a class="docs-heading-anchor" href="#Nonlinear_MOI_interface">MathOptInterface</a><a id="Nonlinear_MOI_interface-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear_MOI_interface" title="Permalink"></a></h3><p>MathOptInterface communicates the nonlinear portion of an optimization problem to solvers using concrete subtypes of <a href="../../../reference/nonlinear/#MathOptInterface.AbstractNLPEvaluator"><code>AbstractNLPEvaluator</code></a>, which implement the <a href="../../../reference/nonlinear/#Nonlinear-programming">Nonlinear programming</a> API.</p><p>Create an <a href="../../../reference/nonlinear/#MathOptInterface.AbstractNLPEvaluator"><code>AbstractNLPEvaluator</code></a> from <a href="../reference/#MathOptInterface.Nonlinear.Model"><code>Nonlinear.Model</code></a> using <a href="../reference/#MathOptInterface.Nonlinear.Evaluator"><code>Nonlinear.Evaluator</code></a>.</p><p><a href="../reference/#MathOptInterface.Nonlinear.Evaluator"><code>Nonlinear.Evaluator</code></a> requires an <a href="../reference/#MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation"><code>Nonlinear.AbstractAutomaticDifferentiation</code></a> backend and an ordered list of the variables that are included in the model.</p><p>There following backends are available to choose from within MOI, although other packages may add more options by sub-typing <a href="../reference/#MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation"><code>Nonlinear.AbstractAutomaticDifferentiation</code></a>:</p><ul><li><a href="../reference/#MathOptInterface.Nonlinear.ExprGraphOnly"><code>Nonlinear.ExprGraphOnly</code></a></li><li><a href="../reference/#MathOptInterface.Nonlinear.SparseReverseMode"><code>Nonlinear.SparseReverseMode</code></a>.</li></ul><pre><code class="language-julia-repl hljs">julia&gt; evaluator = Nonlinear.Evaluator(model, Nonlinear.ExprGraphOnly(), [x])
+Nonlinear.Evaluator with available features:
+  * :ExprGraph</code></pre><p>The functions of the <a href="../../../reference/nonlinear/#Nonlinear-programming">Nonlinear programming</a> API implemented by <a href="../reference/#MathOptInterface.Nonlinear.Evaluator"><code>Nonlinear.Evaluator</code></a> depends upon the chosen <a href="../reference/#MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation"><code>Nonlinear.AbstractAutomaticDifferentiation</code></a> backend.</p><p>The <code>:ExprGraph</code> feature means we can call <a href="../../../reference/nonlinear/#MathOptInterface.objective_expr"><code>objective_expr</code></a> and <a href="../../../reference/nonlinear/#MathOptInterface.constraint_expr"><code>constraint_expr</code></a> to retrieve the expression graph of the problem. However, we cannot call gradient terms such as <a href="../../../reference/nonlinear/#MathOptInterface.eval_objective_gradient"><code>eval_objective_gradient</code></a> because <a href="../reference/#MathOptInterface.Nonlinear.ExprGraphOnly"><code>Nonlinear.ExprGraphOnly</code></a> does not have the capability to differentiate a nonlinear expression.</p><p>If, instead, we pass <a href="../reference/#MathOptInterface.Nonlinear.SparseReverseMode"><code>Nonlinear.SparseReverseMode</code></a>, then we get access to <code>:Grad</code>, the gradient of the objective function, <code>:Jac</code>, the Jacobian matrix of the constraints, <code>:JacVec</code>, the ability to compute Jacobian-vector products, and <code>:ExprGraph</code>.</p><pre><code class="language-julia-repl hljs">julia&gt; evaluator = Nonlinear.Evaluator(
+           model,
+           Nonlinear.SparseReverseMode(),
+           [x],
+       )
+Nonlinear.Evaluator with available features:
+  * :Grad
+  * :Jac
+  * :JacVec
+  * :ExprGraph</code></pre><p>However, before using the evaluator, we need to call <a href="../../../reference/nonlinear/#MathOptInterface.initialize"><code>initialize</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.initialize(evaluator, [:Grad, :Jac, :JacVec, :ExprGraph])</code></pre><p>Now we can call methods like <a href="../../../reference/nonlinear/#MathOptInterface.eval_objective"><code>eval_objective</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; x = [1.0]
+1-element Vector{Float64}:
+ 1.0
+
+julia&gt; MOI.eval_objective(evaluator, x)
+7.268073418273571</code></pre><p>and <a href="../../../reference/nonlinear/#MathOptInterface.eval_objective_gradient"><code>eval_objective_gradient</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; grad = [0.0]
+1-element Vector{Float64}:
+ 0.0
+
+julia&gt; MOI.eval_objective_gradient(evaluator, grad, x)
+
+julia&gt; grad
+1-element Vector{Float64}:
+ 1.909297426825682</code></pre><p>Instead of passing <a href="../reference/#MathOptInterface.Nonlinear.Evaluator"><code>Nonlinear.Evaluator</code></a> directly to solvers, solvers query the <a href="../../../reference/nonlinear/#MathOptInterface.NLPBlock"><code>NLPBlock</code></a> attribute, which returns an <a href="../../../reference/nonlinear/#MathOptInterface.NLPBlockData"><code>NLPBlockData</code></a>. This object wraps an <a href="../reference/#MathOptInterface.Nonlinear.Evaluator"><code>Nonlinear.Evaluator</code></a> and includes other information such as constraint bounds and whether the evaluator has a nonlinear objective. Create and set <a href="../../../reference/nonlinear/#MathOptInterface.NLPBlockData"><code>NLPBlockData</code></a> as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; block = MOI.NLPBlockData(evaluator);
+
+julia&gt; model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}());
+
+julia&gt; MOI.set(model, MOI.NLPBlock(), block);</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Only call <a href="../../../reference/nonlinear/#MathOptInterface.NLPBlockData"><code>NLPBlockData</code></a> once you have finished modifying the problem in <code>model</code>.</p></div></div><p>Putting everything together, you can create a nonlinear optimization problem in MathOptInterface as follows:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI
+
+function build_model(
+    model::MOI.ModelLike;
+    backend::MOI.Nonlinear.AbstractAutomaticDifferentiation,
+)
+    x = MOI.add_variable(model)
+    y = MOI.add_variable(model)
+    MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)
+    nl_model = MOI.Nonlinear.Model()
+    MOI.Nonlinear.set_objective(nl_model, :($x^2 + $y^2))
+    evaluator = MOI.Nonlinear.Evaluator(nl_model, backend, [x, y])
+    MOI.set(model, MOI.NLPBlock(), MOI.NLPBlockData(evaluator))
+    return
+end
+
+# Replace `model` and `backend` with your optimizer and backend of choice.
+model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())
+build_model(model; backend = MOI.Nonlinear.SparseReverseMode())</code></pre><h2 id="Expression-graph-representation"><a class="docs-heading-anchor" href="#Expression-graph-representation">Expression-graph representation</a><a id="Expression-graph-representation-1"></a><a class="docs-heading-anchor-permalink" href="#Expression-graph-representation" title="Permalink"></a></h2><p><a href="../reference/#MathOptInterface.Nonlinear.Model"><code>Nonlinear.Model</code></a> stores nonlinear expressions in <a href="../reference/#MathOptInterface.Nonlinear.Expression"><code>Nonlinear.Expression</code></a>s. This section explains the design of the expression graph data structure in <a href="../reference/#MathOptInterface.Nonlinear.Expression"><code>Nonlinear.Expression</code></a>.</p><p>Given a nonlinear function like <code>f(x) = sin(x)^2 + x</code>, a conceptual aid for thinking about the graph representation of the expression is to convert it into <a href="https://en.wikipedia.org/wiki/Polish_notation">Polish prefix notation</a>:</p><pre><code class="nohighlight hljs">f(x, y) = (+ (^ (sin x) 2) x)</code></pre><p>This format identifies each operator (function), as well as a list of arguments. Operators can be univariate, like <code>sin</code>, or multivariate, like <code>+</code>.</p><p>A common way of representing Polish prefix notation in code is as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; x = MOI.VariableIndex(1);
+
+julia&gt; struct ExprNode
+           op::Symbol
+           children::Vector{Union{ExprNode,Float64,MOI.VariableIndex}}
+       end
+
+julia&gt; expr = ExprNode(:+, [ExprNode(:^, [ExprNode(:sin, [x]), 2.0]), x]);</code></pre><p>This data structure follows our Polish prefix notation very closely, and we can easily identify the arguments to an operator. However, it has a significant draw-back: each node in the graph requires a <code>Vector</code>, which is heap-allocated and tracked by Julia&#39;s garbage collector (GC). For large models, we can expect to have millions of nodes in the expression graph, so this overhead quickly becomes prohibitive for computation.</p><p>An alternative is to record the expression as a linear tape:</p><pre><code class="language-julia-repl hljs">julia&gt; expr = Any[:+, 2, :^, 2, :sin, 1, x, 2.0, x]
+9-element Vector{Any}:
+  :+
+ 2
+  :^
+ 2
+  :sin
+ 1
+  MOI.VariableIndex(1)
+ 2.0
+  MOI.VariableIndex(1)</code></pre><p>The <code>Int</code> after each operator <code>Symbol</code> specifies the number of arguments.</p><p>This data-structure is a single vector, which resolves our problem with the GC, but each element is the abstract type, <code>Any</code>, and so any operations on it will lead to slower dynamic dispatch. It&#39;s also hard to identify the children of each operation without reading the entire tape.</p><p>To summarize, representing expression graphs in Julia has the following challenges:</p><ul><li>Nodes in the expression graph should not contain a heap-allocated object</li><li>All data-structures should be concretely typed</li><li>It should be easy to identify the children of a node</li></ul><h3 id="Sketch-of-the-design-in-Nonlinear"><a class="docs-heading-anchor" href="#Sketch-of-the-design-in-Nonlinear">Sketch of the design in Nonlinear</a><a id="Sketch-of-the-design-in-Nonlinear-1"></a><a class="docs-heading-anchor-permalink" href="#Sketch-of-the-design-in-Nonlinear" title="Permalink"></a></h3><p><code>Nonlinear</code> overcomes these problems by decomposing the data structure into a number of different concrete-typed vectors.</p><p>First, we create vectors of the supported uni- and multivariate operators.</p><pre><code class="language-julia-repl hljs">julia&gt; const UNIVARIATE_OPERATORS = [:sin];
+
+julia&gt; const MULTIVARIATE_OPERATORS = [:+, :^];</code></pre><p>In practice, there are many more supported operations than the ones listed here.</p><p>Second, we create an enum to represent the different types of nodes present in the expression graph:</p><pre><code class="language-julia-repl hljs">julia&gt; @enum(
+           NodeType,
+           NODE_CALL_MULTIVARIATE,
+           NODE_CALL_UNIVARIATE,
+           NODE_VARIABLE,
+           NODE_VALUE,
+       )</code></pre><p>In practice, there are node types other than the ones listed here.</p><p>Third, we create two concretely typed structs as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; struct Node
+           type::NodeType
+           parent::Int
+           index::Int
+       end
+
+julia&gt; struct Expression
+           nodes::Vector{Node}
+           values::Vector{Float64}
+       end</code></pre><p>For each node <code>node</code> in the <code>.nodes</code> field, if <code>node.type</code> is:</p><ul><li><code>NODE_CALL_MULTIVARIATE</code>, we look up <code>MULTIVARIATE_OPERATORS[node.index]</code> to retrieve the operator</li><li><code>NODE_CALL_UNIVARIATE</code>, we look up <code>UNIVARIATE_OPERATORS[node.index]</code> to retrieve the operator</li><li><code>NODE_VARIABLE</code>, we create <code>MOI.VariableIndex(node.index)</code></li><li><code>NODE_VALUE</code>, we look up <code>values[node.index]</code></li></ul><p>The <code>.parent</code> field of each node is the integer index of the parent node in <code>.nodes</code>. For the first node, the parent is <code>-1</code> by convention.</p><p>Therefore, we can represent our function as:</p><pre><code class="language-julia-repl hljs">julia&gt; expr = Expression(
+           [
+               Node(NODE_CALL_MULTIVARIATE, -1, 1),
+               Node(NODE_CALL_MULTIVARIATE, 1, 2),
+               Node(NODE_CALL_UNIVARIATE, 2, 1),
+               Node(NODE_VARIABLE, 3, 1),
+               Node(NODE_VALUE, 2, 1),
+               Node(NODE_VARIABLE, 1, 1),
+           ],
+           [2.0],
+       );</code></pre><p>This is less readable than the other options, but does this data structure meet our design goals?</p><p>Instead of a heap-allocated object for each node, we only have two <code>Vector</code>s for each expression, <code>nodes</code> and <code>values</code>, as well as two constant vectors for the <code>OPERATORS</code>. In addition, all fields are concretely typed, and there are no <code>Union</code> or <code>Any</code> types.</p><p>For our third goal, it is not easy to identify the children of a node, but it is easy to identify the <em>parent</em> of any node. Therefore, we can use <a href="../reference/#MathOptInterface.Nonlinear.adjacency_matrix"><code>Nonlinear.adjacency_matrix</code></a> to compute a sparse matrix that maps parents to their children.</p><p>The tape is also ordered topologically, so that a reverse pass of the nodes evaluates all children nodes before their parent.</p><h3 id="The-design-in-practice"><a class="docs-heading-anchor" href="#The-design-in-practice">The design in practice</a><a id="The-design-in-practice-1"></a><a class="docs-heading-anchor-permalink" href="#The-design-in-practice" title="Permalink"></a></h3><p>In practice, <code>Node</code> and <code>Expression</code> are exactly <a href="../reference/#MathOptInterface.Nonlinear.Node"><code>Nonlinear.Node</code></a> and <a href="../reference/#MathOptInterface.Nonlinear.Expression"><code>Nonlinear.Expression</code></a>. However, <a href="../reference/#MathOptInterface.Nonlinear.NodeType"><code>Nonlinear.NodeType</code></a> has more fields to account for comparison operators such as <code>:&gt;=</code> and <code>:&lt;=</code>, logic operators such as <code>:&amp;&amp;</code> and <code>:||</code>, nonlinear parameters, and nested subexpressions.</p><p>Moreover, instead of storing the operators as global constants, they are stored in <a href="../reference/#MathOptInterface.Nonlinear.OperatorRegistry"><code>Nonlinear.OperatorRegistry</code></a>, and it also stores a vector of logic operators and a vector of comparison operators. In addition to <a href="../reference/#MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS"><code>Nonlinear.DEFAULT_UNIVARIATE_OPERATORS</code></a> and <a href="../reference/#MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS"><code>Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS</code></a>, you can register user-defined functions using <a href="../reference/#MathOptInterface.Nonlinear.register_operator"><code>Nonlinear.register_operator</code></a>.</p><p><a href="../reference/#MathOptInterface.Nonlinear.Model"><code>Nonlinear.Model</code></a> is a struct that stores the <a href="../reference/#MathOptInterface.Nonlinear.OperatorRegistry"><code>Nonlinear.OperatorRegistry</code></a>, as well as a list of parameters and subexpressions in the model.</p><h2 id="ReverseAD"><a class="docs-heading-anchor" href="#ReverseAD">ReverseAD</a><a id="ReverseAD-1"></a><a class="docs-heading-anchor-permalink" href="#ReverseAD" title="Permalink"></a></h2><p><code>Nonlinear.ReverseAD</code> is a submodule for computing derivatives of a nonlinear optimization problem using sparse reverse-mode automatic differentiation (AD).</p><p>This section does not attempt to explain how sparse reverse-mode AD works, but instead explains why MOI contains its own implementation, and highlights notable differences from similar packages.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Don&#39;t use the API in <code>ReverseAD</code> to compute derivatives. Instead, create a <a href="../reference/#MathOptInterface.Nonlinear.Evaluator"><code>Nonlinear.Evaluator</code></a> object with <a href="../reference/#MathOptInterface.Nonlinear.SparseReverseMode"><code>Nonlinear.SparseReverseMode</code></a> as the backend, and then query the MOI API methods.</p></div></div><h3 id="Design-goals"><a class="docs-heading-anchor" href="#Design-goals">Design goals</a><a id="Design-goals-1"></a><a class="docs-heading-anchor-permalink" href="#Design-goals" title="Permalink"></a></h3><p>The JuliaDiff organization maintains a <a href="https://juliadiff.org">list of packages</a> for doing AD in Julia. At last count, there were at least ten packages——not including <code>ReverseAD</code>——for reverse-mode AD in Julia. <code>ReverseAD</code> exists because it has a different set of design goals.</p><ul><li><strong>Goal: handle scale and sparsity.</strong> The types of nonlinear optimization problems that MOI represents can be large scale (10^5 or more functions across 10^5 or more variables) with very sparse derivatives. The ability to compute a sparse Hessian matrix is essential. To the best of our knowledge, <code>ReverseAD</code> is the only reverse-mode AD system in Julia that handles sparsity by default.</li><li><strong>Goal: limit the scope to improve robustness.</strong> Most other AD packages accept arbitrary Julia functions as input and then trace an expression graph using operator overloading. This means they must deal (or detect and ignore) with control flow, I/O, and other vagaries of Julia. In contrast, <code>ReverseAD</code> only accepts functions in the form of <a href="../reference/#MathOptInterface.Nonlinear.Expression"><code>Nonlinear.Expression</code></a>, which greatly limits the range of syntax that it must deal with. By reducing the scope of what we accept as input to functions relevant for mathematical optimization, we can provide a simpler implementation with various performance optimizations.</li><li><strong>Goal: provide outputs which match what solvers expect.</strong> Other AD packages focus on differentiating individual Julia functions. In contrast, <code>ReverseAD</code> has a very specific use-case: to generate outputs needed by the MOI nonlinear API. This means it needs to efficiently compute sparse Hessians, and it needs subexpression handling to avoid recomputing subexpressions that are shared between functions.</li></ul><h3 id="History"><a class="docs-heading-anchor" href="#History">History</a><a id="History-1"></a><a class="docs-heading-anchor-permalink" href="#History" title="Permalink"></a></h3><p><code>ReverseAD</code> started life as <a href="https://github.com/mlubin/ReverseDiffSparse.jl">ReverseDiffSparse.jl</a>, development of which began in early 2014(!). This was well before the other AD packages started development. Because we had a well-tested, working AD in JuMP, there was less motivation to contribute to and explore other AD packages. The lack of historical interaction also meant that other packages were not optimized for the types of problems that JuMP is built for (that is, large-scale sparse problems). When we first created MathOptInterface, we kept the AD in JuMP to simplify the transition, and post-poned the development of a first-class nonlinear interface in MathOptInterface.</p><p>Prior to the introduction of <code>Nonlinear</code>, JuMP&#39;s nonlinear implementation was a confusing mix of functions and types spread across the code base and in the private <code>_Derivatives</code> submodule. This made it hard to swap the AD system for another. The main motivation for refactoring JuMP to create the <code>Nonlinear</code> submodule in MathOptInterface was to abstract the interface between JuMP and the AD system, allowing us to swap-in and test new AD systems in the future.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../FileFormats/reference/">« API Reference</a><a class="docs-footer-nextpage" href="../reference/">API Reference »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Nonlinear/reference/index.html b/previews/PR3525/moi/submodules/Nonlinear/reference/index.html
new file mode 100644
index 00000000000..55403d208ff
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Nonlinear/reference/index.html
@@ -0,0 +1,151 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>API Reference · JuMP</title><meta name="title" content="API Reference · JuMP"/><meta property="og:title" content="API Reference · JuMP"/><meta property="twitter:title" content="API Reference · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../overview/">Overview</a></li><li class="is-active"><a class="tocitem" href>API Reference</a><ul class="internal"><li><a class="tocitem" href="#nonlinear_api_expressions"><span>Expressions</span></a></li><li><a class="tocitem" href="#nonlinear_api_parameters"><span>Parameters</span></a></li><li><a class="tocitem" href="#nonlinear_api_objectives"><span>Objectives</span></a></li><li><a class="tocitem" href="#nonlinear_api_constraints"><span>Constraints</span></a></li><li><a class="tocitem" href="#nonlinear_api_operators"><span>User-defined operators</span></a></li><li><a class="tocitem" href="#Automatic-differentiation-backends"><span>Automatic-differentiation backends</span></a></li><li><a class="tocitem" href="#Data-structure"><span>Data-structure</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Nonlinear</a></li><li class="is-active"><a href>API Reference</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>API Reference</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Nonlinear/reference.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="NonlinearAPI"><a class="docs-heading-anchor" href="#NonlinearAPI">Nonlinear Modeling</a><a id="NonlinearAPI-1"></a><a class="docs-heading-anchor-permalink" href="#NonlinearAPI" title="Permalink"></a></h1><p>More information can be found in the <a href="../overview/#nonlinear_developers">Nonlinear</a> section of the manual.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear" href="#MathOptInterface.Nonlinear"><code>MathOptInterface.Nonlinear</code></a> — <span class="docstring-category">Module</span></header><section><div><pre><code class="language-julia hljs">Nonlinear</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The Nonlinear submodule is experimental. Until this message is removed, breaking changes may be introduced in any minor or patch release of MathOptInterface.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/Nonlinear.jl#L7-L14">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.Model" href="#MathOptInterface.Nonlinear.Model"><code>MathOptInterface.Nonlinear.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Model()</code></pre><p>The core datastructure for representing a nonlinear optimization problem.</p><p>It has the following fields:</p><ul><li><code>objective::Union{Nothing,Expression}</code> : holds the nonlinear objective function, if one exists, otherwise <code>nothing</code>.</li><li><code>expressions::Vector{Expression}</code> : a vector of expressions in the model.</li><li><code>constraints::OrderedDict{ConstraintIndex,Constraint}</code> : a map from <a href="#MathOptInterface.Nonlinear.ConstraintIndex"><code>ConstraintIndex</code></a> to the corresponding <a href="#MathOptInterface.Nonlinear.Constraint"><code>Constraint</code></a>. An <code>OrderedDict</code> is used instead of a <code>Vector</code> to support constraint deletion.</li><li><code>parameters::Vector{Float64}</code> : holds the current values of the parameters.</li><li><code>operators::OperatorRegistry</code> : stores the operators used in the model.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L145-L159">source</a></section></article><h2 id="nonlinear_api_expressions"><a class="docs-heading-anchor" href="#nonlinear_api_expressions">Expressions</a><a id="nonlinear_api_expressions-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_api_expressions" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.ExpressionIndex" href="#MathOptInterface.Nonlinear.ExpressionIndex"><code>MathOptInterface.Nonlinear.ExpressionIndex</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ExpressionIndex</code></pre><p>An index to a nonlinear expression that is returned by <a href="#MathOptInterface.Nonlinear.add_expression"><code>add_expression</code></a>.</p><p>Given <code>data::Model</code> and <code>ex::ExpressionIndex</code>, use <code>data[ex]</code> to retrieve the corresponding <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L121-L128">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.add_expression" href="#MathOptInterface.Nonlinear.add_expression"><code>MathOptInterface.Nonlinear.add_expression</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_expression(model::Model, expr)::ExpressionIndex</code></pre><p>Parse <code>expr</code> into a <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a> and add to <code>model</code>. Returns an <a href="#MathOptInterface.Nonlinear.ExpressionIndex"><code>ExpressionIndex</code></a> that can be interpolated into other input expressions.</p><p><code>expr</code> must be a type that is supported by <a href="#MathOptInterface.Nonlinear.parse_expression"><code>parse_expression</code></a>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">model = Model()
+x = MOI.VariableIndex(1)
+ex = add_expression(model, :($x^2 + 1))
+set_objective(model, :(sqrt($ex)))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/model.jl#L52-L68">source</a></section></article><h2 id="nonlinear_api_parameters"><a class="docs-heading-anchor" href="#nonlinear_api_parameters">Parameters</a><a id="nonlinear_api_parameters-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_api_parameters" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.ParameterIndex" href="#MathOptInterface.Nonlinear.ParameterIndex"><code>MathOptInterface.Nonlinear.ParameterIndex</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ParameterIndex</code></pre><p>An index to a nonlinear parameter that is returned by <a href="#MathOptInterface.Nonlinear.add_parameter"><code>add_parameter</code></a>. Given <code>data::Model</code> and <code>p::ParameterIndex</code>, use <code>data[p]</code> to retrieve the current value of the parameter and <code>data[p] = value</code> to set a new value.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L110-L116">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.add_parameter" href="#MathOptInterface.Nonlinear.add_parameter"><code>MathOptInterface.Nonlinear.add_parameter</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_parameter(model::Model, value::Float64)::ParameterIndex</code></pre><p>Add a new parameter to <code>model</code> with the default value <code>value</code>. Returns a <a href="#MathOptInterface.Nonlinear.ParameterIndex"><code>ParameterIndex</code></a> that can be interpolated into other input expressions and used to modify the value of the parameter.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">model = Model()
+x = MOI.VariableIndex(1)
+p = add_parameter(model, 1.2)
+c = add_constraint(model, :($x^2 - $p), MOI.LessThan(0.0))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/model.jl#L146-L161">source</a></section></article><h2 id="nonlinear_api_objectives"><a class="docs-heading-anchor" href="#nonlinear_api_objectives">Objectives</a><a id="nonlinear_api_objectives-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_api_objectives" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.set_objective" href="#MathOptInterface.Nonlinear.set_objective"><code>MathOptInterface.Nonlinear.set_objective</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_objective(model::Model, obj)::Nothing</code></pre><p>Parse <code>obj</code> into a <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a> and set as the objective function of <code>model</code>.</p><p><code>obj</code> must be a type that is supported by <a href="#MathOptInterface.Nonlinear.parse_expression"><code>parse_expression</code></a>.</p><p>To remove the objective, pass <code>nothing</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">model = Model()
+x = MOI.VariableIndex(1)
+set_objective(model, :($x^2 + 1))
+set_objective(model, x)
+set_objective(model, nothing)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/model.jl#L22-L41">source</a></section></article><h2 id="nonlinear_api_constraints"><a class="docs-heading-anchor" href="#nonlinear_api_constraints">Constraints</a><a id="nonlinear_api_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_api_constraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.ConstraintIndex" href="#MathOptInterface.Nonlinear.ConstraintIndex"><code>MathOptInterface.Nonlinear.ConstraintIndex</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ConstraintIndex</code></pre><p>An index to a nonlinear constraint that is returned by <a href="#MathOptInterface.Nonlinear.add_constraint"><code>add_constraint</code></a>.</p><p>Given <code>data::Model</code> and <code>c::ConstraintIndex</code>, use <code>data[c]</code> to retrieve the corresponding <a href="#MathOptInterface.Nonlinear.Constraint"><code>Constraint</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L133-L140">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.add_constraint" href="#MathOptInterface.Nonlinear.add_constraint"><code>MathOptInterface.Nonlinear.add_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_constraint(
+    model::Model,
+    func,
+    set::Union{
+        MOI.GreaterThan{Float64},
+        MOI.LessThan{Float64},
+        MOI.Interval{Float64},
+        MOI.EqualTo{Float64},
+    },
+)</code></pre><p>Parse <code>func</code> and <code>set</code> into a <a href="#MathOptInterface.Nonlinear.Constraint"><code>Constraint</code></a> and add to <code>model</code>. Returns a <a href="#MathOptInterface.Nonlinear.ConstraintIndex"><code>ConstraintIndex</code></a> that can be used to delete the constraint or query solution information.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">model = Model()
+x = MOI.VariableIndex(1)
+c = add_constraint(model, :($x^2), MOI.LessThan(1.0))</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/model.jl#L78-L101">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.delete" href="#MathOptInterface.Nonlinear.delete"><code>MathOptInterface.Nonlinear.delete</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">delete(model::Model, c::ConstraintIndex)::Nothing</code></pre><p>Delete the constraint index <code>c</code> from <code>model</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">model = Model()
+x = MOI.VariableIndex(1)
+c = add_constraint(model, :($x^2), MOI.LessThan(1.0))
+delete(model, c)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/model.jl#L119-L132">source</a></section></article><h2 id="nonlinear_api_operators"><a class="docs-heading-anchor" href="#nonlinear_api_operators">User-defined operators</a><a id="nonlinear_api_operators-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_api_operators" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.OperatorRegistry" href="#MathOptInterface.Nonlinear.OperatorRegistry"><code>MathOptInterface.Nonlinear.OperatorRegistry</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">OperatorRegistry()</code></pre><p>Create a new <code>OperatorRegistry</code> to store and evaluate univariate and multivariate operators.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L135-L140">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS" href="#MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS"><code>MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">DEFAULT_UNIVARIATE_OPERATORS</code></pre><p>The list of univariate operators that are supported by default.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS
+72-element Vector{Symbol}:
+ :+
+ :-
+ :abs
+ :sqrt
+ :cbrt
+ :abs2
+ :inv
+ :log
+ :log10
+ :log2
+ ⋮
+ :airybi
+ :airyaiprime
+ :airybiprime
+ :besselj0
+ :besselj1
+ :bessely0
+ :bessely1
+ :erfcx
+ :dawson</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L73-L106">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS" href="#MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS"><code>MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS</code></a> — <span class="docstring-category">Constant</span></header><section><div><pre><code class="language-julia hljs">DEFAULT_MULTIVARIATE_OPERATORS</code></pre><p>The list of multivariate operators that are supported by default.</p><p><strong>Example</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; MOI.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS
+9-element Vector{Symbol}:
+ :+
+ :-
+ :*
+ :^
+ :/
+ :ifelse
+ :atan
+ :min
+ :max</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L109-L131">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.register_operator" href="#MathOptInterface.Nonlinear.register_operator"><code>MathOptInterface.Nonlinear.register_operator</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">register_operator(
+    model::Model,
+    op::Symbol,
+    nargs::Int,
+    f::Function,
+    [∇f::Function],
+    [∇²f::Function],
+)</code></pre><p>Register the user-defined operator <code>op</code> with <code>nargs</code> input arguments in <code>model</code>.</p><p><strong>Univariate functions</strong></p><ul><li><code>f(x::T)::T</code> must be a function that takes a single input argument <code>x</code> and returns the function evaluated at <code>x</code>. If <code>∇f</code> and <code>∇²f</code> are not provided, <code>f</code> must support any <code>Real</code> input type <code>T</code>.</li><li><code>∇f(x::T)::T</code> is a function that takes a single input argument <code>x</code> and returns the first derivative of <code>f</code> with respect to <code>x</code>. If <code>∇²f</code> is not provided, <code>∇f</code> must support any <code>Real</code> input type <code>T</code>.</li><li><code>∇²f(x::T)::T</code> is a function that takes a single input argument <code>x</code> and returns the second derivative of <code>f</code> with respect to <code>x</code>.</li></ul><p><strong>Multivariate functions</strong></p><ul><li><code>f(x::T...)::T</code> must be a function that takes a <code>nargs</code> input arguments <code>x</code> and returns the function evaluated at <code>x</code>. If <code>∇f</code> and <code>∇²f</code> are not provided, <code>f</code> must support any <code>Real</code> input type <code>T</code>.</li><li><code>∇f(g::AbstractVector{T}, x::T...)::T</code> is a function that takes a cache vector <code>g</code> of length <code>length(x)</code>, and fills each element <code>g[i]</code> with the partial derivative of <code>f</code> with respect to <code>x[i]</code>.</li><li><code>∇²f(H::AbstractMatrix, x::T...)::T</code> is a function that takes a matrix <code>H</code> and fills the lower-triangular components <code>H[i, j]</code> with the Hessian of <code>f</code> with respect to <code>x[i]</code> and <code>x[j]</code> for <code>i &gt;= j</code>.</li></ul><p><strong>Notes for multivariate Hessians</strong></p><ul><li><code>H</code> has <code>size(H) == (length(x), length(x))</code>, but you must not access elements <code>H[i, j]</code> for <code>i &gt; j</code>.</li><li><code>H</code> is dense, but you do not need to fill structural zeros.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/model.jl#L175-L215">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.register_operator_if_needed" href="#MathOptInterface.Nonlinear.register_operator_if_needed"><code>MathOptInterface.Nonlinear.register_operator_if_needed</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">register_operator_if_needed(
+    registry::OperatorRegistry,
+    op::Symbol,
+    nargs::Int,
+    f::Function;
+)</code></pre><p>Similar to <a href="#MathOptInterface.Nonlinear.register_operator"><code>register_operator</code></a>, but this function warns if the function is not registered, and skips silently if it already is.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L444-L454">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.assert_registered" href="#MathOptInterface.Nonlinear.assert_registered"><code>MathOptInterface.Nonlinear.assert_registered</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">assert_registered(registry::OperatorRegistry, op::Symbol, nargs::Int)</code></pre><p>Throw an error if <code>op</code> is not registered in <code>registry</code> with <code>nargs</code> arguments.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L468-L472">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.check_return_type" href="#MathOptInterface.Nonlinear.check_return_type"><code>MathOptInterface.Nonlinear.check_return_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">check_return_type(::Type{T}, ret::S) where {T,S}</code></pre><p>Overload this method for new types <code>S</code> to throw an informative error if a user-defined function returns the type <code>S</code> instead of <code>T</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L492-L497">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_univariate_function" href="#MathOptInterface.Nonlinear.eval_univariate_function"><code>MathOptInterface.Nonlinear.eval_univariate_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_univariate_function(
+    registry::OperatorRegistry,
+    op::Symbol,
+    x::T,
+) where {T}</code></pre><p>Evaluate the operator <code>op(x)::T</code>, where <code>op</code> is a univariate function in <code>registry</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L507-L516">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_univariate_gradient" href="#MathOptInterface.Nonlinear.eval_univariate_gradient"><code>MathOptInterface.Nonlinear.eval_univariate_gradient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_univariate_gradient(
+    registry::OperatorRegistry,
+    op::Symbol,
+    x::T,
+) where {T}</code></pre><p>Evaluate the first-derivative of the operator <code>op(x)::T</code>, where <code>op</code> is a univariate function in <code>registry</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L534-L543">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_univariate_hessian" href="#MathOptInterface.Nonlinear.eval_univariate_hessian"><code>MathOptInterface.Nonlinear.eval_univariate_hessian</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_univariate_hessian(
+    registry::OperatorRegistry,
+    op::Symbol,
+    x::T,
+) where {T}</code></pre><p>Evaluate the second-derivative of the operator <code>op(x)::T</code>, where <code>op</code> is a univariate function in <code>registry</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L561-L570">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_multivariate_function" href="#MathOptInterface.Nonlinear.eval_multivariate_function"><code>MathOptInterface.Nonlinear.eval_multivariate_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_multivariate_function(
+    registry::OperatorRegistry,
+    op::Symbol,
+    x::AbstractVector{T},
+) where {T}</code></pre><p>Evaluate the operator <code>op(x)::T</code>, where <code>op</code> is a multivariate function in <code>registry</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L597-L606">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_multivariate_gradient" href="#MathOptInterface.Nonlinear.eval_multivariate_gradient"><code>MathOptInterface.Nonlinear.eval_multivariate_gradient</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_multivariate_gradient(
+    registry::OperatorRegistry,
+    op::Symbol,
+    g::AbstractVector{T},
+    x::AbstractVector{T},
+) where {T}</code></pre><p>Evaluate the gradient of operator <code>g .= ∇op(x)</code>, where <code>op</code> is a multivariate function in <code>registry</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L647-L657">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_multivariate_hessian" href="#MathOptInterface.Nonlinear.eval_multivariate_hessian"><code>MathOptInterface.Nonlinear.eval_multivariate_hessian</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_multivariate_hessian(
+    registry::OperatorRegistry,
+    op::Symbol,
+    H::AbstractMatrix,
+    x::AbstractVector{T},
+) where {T}</code></pre><p>Evaluate the Hessian of operator <code>∇²op(x)</code>, where <code>op</code> is a multivariate function in <code>registry</code>.</p><p>The Hessian is stored in the lower-triangular part of the matrix <code>H</code>.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Implementations of the Hessian operators will not fill structural zeros. Therefore, before calling this function you should pre-populate the matrix <code>H</code> with <code>0</code>.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L737-L754">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_logic_function" href="#MathOptInterface.Nonlinear.eval_logic_function"><code>MathOptInterface.Nonlinear.eval_logic_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_logic_function(
+    registry::OperatorRegistry,
+    op::Symbol,
+    lhs::T,
+    rhs::T,
+)::Bool where {T}</code></pre><p>Evaluate <code>(lhs op rhs)::Bool</code>, where <code>op</code> is a logic operator in <code>registry</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L844-L853">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.eval_comparison_function" href="#MathOptInterface.Nonlinear.eval_comparison_function"><code>MathOptInterface.Nonlinear.eval_comparison_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_comparison_function(
+    registry::OperatorRegistry,
+    op::Symbol,
+    lhs::T,
+    rhs::T,
+)::Bool where {T}</code></pre><p>Evaluate <code>(lhs op rhs)::Bool</code>, where <code>op</code> is a comparison operator in <code>registry</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/operators.jl#L868-L878">source</a></section></article><h2 id="Automatic-differentiation-backends"><a class="docs-heading-anchor" href="#Automatic-differentiation-backends">Automatic-differentiation backends</a><a id="Automatic-differentiation-backends-1"></a><a class="docs-heading-anchor-permalink" href="#Automatic-differentiation-backends" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.Evaluator" href="#MathOptInterface.Nonlinear.Evaluator"><code>MathOptInterface.Nonlinear.Evaluator</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Evaluator(
+    model::Model,
+    backend::AbstractAutomaticDifferentiation,
+    ordered_variables::Vector{MOI.VariableIndex},
+)</code></pre><p>Create <code>Evaluator</code>, a subtype of <code>MOI.AbstractNLPEvaluator</code>, from <code>Model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L187-L195">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation" href="#MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation"><code>MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">AbstractAutomaticDifferentiation</code></pre><p>An abstract type for extending <a href="#MathOptInterface.Nonlinear.Evaluator"><code>Evaluator</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L180-L184">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.ExprGraphOnly" href="#MathOptInterface.Nonlinear.ExprGraphOnly"><code>MathOptInterface.Nonlinear.ExprGraphOnly</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ExprGraphOnly() &lt;: AbstractAutomaticDifferentiation</code></pre><p>The default implementation of <code>AbstractAutomaticDifferentiation</code>. The only supported feature is <code>:ExprGraph</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L257-L262">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.SparseReverseMode" href="#MathOptInterface.Nonlinear.SparseReverseMode"><code>MathOptInterface.Nonlinear.SparseReverseMode</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">SparseReverseMode() &lt;: AbstractAutomaticDifferentiation</code></pre><p>An implementation of <code>AbstractAutomaticDifferentiation</code> that uses sparse reverse-mode automatic differentiation to compute derivatives. Supports all features in the MOI nonlinear interface.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L269-L275">source</a></section></article><h2 id="Data-structure"><a class="docs-heading-anchor" href="#Data-structure">Data-structure</a><a id="Data-structure-1"></a><a class="docs-heading-anchor-permalink" href="#Data-structure" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.Node" href="#MathOptInterface.Nonlinear.Node"><code>MathOptInterface.Nonlinear.Node</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct Node
+    type::NodeType
+    index::Int
+    parent::Int
+end</code></pre><p>A single node in a nonlinear expression tree. Used by <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a>.</p><p>See the MathOptInterface documentation for information on how the nodes and values form an expression tree.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L47-L59">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.NodeType" href="#MathOptInterface.Nonlinear.NodeType"><code>MathOptInterface.Nonlinear.NodeType</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">NodeType</code></pre><p>An enum describing the possible node types. Each <a href="#MathOptInterface.Nonlinear.Node"><code>Node</code></a> has a <code>.index</code> field, which should be interpreted as follows:</p><ul><li><code>NODE_CALL_MULTIVARIATE</code>: the index into <code>operators.multivariate_operators</code></li><li><code>NODE_CALL_UNIVARIATE</code>: the index into <code>operators.univariate_operators</code></li><li><code>NODE_LOGIC</code>: the index into <code>operators.logic_operators</code></li><li><code>NODE_COMPARISON</code>: the index into <code>operators.comparison_operators</code></li><li><code>NODE_MOI_VARIABLE</code>: the value of <code>MOI.VariableIndex(index)</code> in the user&#39;s space of the model.</li><li><code>NODE_VARIABLE</code>: the 1-based index of the internal vector</li><li><code>NODE_VALUE</code>: the index into the <code>.values</code> field of <code>Expression</code></li><li><code>NODE_PARAMETER</code>: the index into <code>data.parameters</code></li><li><code>NODE_SUBEXPRESSION</code>:  the index into <code>data.expressions</code></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L7-L23">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.Expression" href="#MathOptInterface.Nonlinear.Expression"><code>MathOptInterface.Nonlinear.Expression</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct Expression
+    nodes::Vector{Node}
+    values::Vector{Float64}
+end</code></pre><p>The core type that represents a nonlinear expression. See the MathOptInterface documentation for information on how the nodes and values form an expression tree.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L66-L75">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.Constraint" href="#MathOptInterface.Nonlinear.Constraint"><code>MathOptInterface.Nonlinear.Constraint</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct Constraint
+    expression::Expression
+    set::Union{
+        MOI.LessThan{Float64},
+        MOI.GreaterThan{Float64},
+        MOI.EqualTo{Float64},
+        MOI.Interval{Float64},
+    }
+end</code></pre><p>A type to hold information relating to the nonlinear constraint <code>f(x) in S</code>, where <code>f(x)</code> is defined by <code>.expression</code>, and <code>S</code> is <code>.set</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/types.jl#L86-L99">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.adjacency_matrix" href="#MathOptInterface.Nonlinear.adjacency_matrix"><code>MathOptInterface.Nonlinear.adjacency_matrix</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">adjacency_matrix(nodes::Vector{Node})</code></pre><p>Compute the sparse adjacency matrix describing the parent-child relationships in <code>nodes</code>.</p><p>The element <code>(i, j)</code> is <code>true</code> if there is an edge <em>from</em> <code>node[j]</code> to <code>node[i]</code>. Since we get a column-oriented matrix, this gives us a fast way to look up the edges leaving any node (i.e., the children).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/evaluator.jl#L227-L236">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.parse_expression" href="#MathOptInterface.Nonlinear.parse_expression"><code>MathOptInterface.Nonlinear.parse_expression</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">parse_expression(data::Model, input)::Expression</code></pre><p>Parse <code>input</code> into a <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/parse.jl#L7-L11">source</a></section><section><div><pre><code class="language-julia hljs">parse_expression(
+    data::Model,
+    expr::Expression,
+    input::Any,
+    parent_index::Int,
+)::Expression</code></pre><p>Parse <code>input</code> into a <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a>, and add it to <code>expr</code> as a child of <code>expr.nodes[parent_index]</code>. Existing subexpressions and parameters are stored in <code>data</code>.</p><p>You can extend parsing support to new types of objects by overloading this method with a different type on <code>input::Any</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/parse.jl#L18-L32">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.convert_to_expr" href="#MathOptInterface.Nonlinear.convert_to_expr"><code>MathOptInterface.Nonlinear.convert_to_expr</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">convert_to_expr(data::Model, expr::Expression)</code></pre><p>Convert the <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a> <code>expr</code> into a Julia <code>Expr</code>.</p><ul><li>subexpressions are represented by a <a href="#MathOptInterface.Nonlinear.ExpressionIndex"><code>ExpressionIndex</code></a> object.</li><li>parameters are represented by a <a href="#MathOptInterface.Nonlinear.ParameterIndex"><code>ParameterIndex</code></a> object.</li><li>variables are represennted by an <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> object.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/parse.jl#L325-L333">source</a></section><section><div><pre><code class="language-julia hljs">convert_to_expr(
+    evaluator::Evaluator,
+    expr::Expression;
+    moi_output_format::Bool,
+)</code></pre><p>Convert the <a href="#MathOptInterface.Nonlinear.Expression"><code>Expression</code></a> <code>expr</code> into a Julia <code>Expr</code>.</p><p>If <code>moi_output_format = true</code>:</p><ul><li>subexpressions will be converted to Julia <code>Expr</code> and substituted into the output expression.</li><li>the current value of each parameter will be interpolated into the expression</li><li>variables will be represented in the form <code>x[MOI.VariableIndex(i)]</code></li></ul><p>If <code>moi_output_format = false</code>:</p><ul><li>subexpressions will be represented by a <a href="#MathOptInterface.Nonlinear.ExpressionIndex"><code>ExpressionIndex</code></a> object.</li><li>parameters will be represented by a <a href="#MathOptInterface.Nonlinear.ParameterIndex"><code>ParameterIndex</code></a> object.</li><li>variables will be represennted by an <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> object.</li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>To use <code>moi_output_format = true</code>, you must have first called <a href="../../../reference/nonlinear/#MathOptInterface.initialize"><code>MOI.initialize</code></a> with <code>:ExprGraph</code> as a requested feature.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/evaluator.jl#L254-L277">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Nonlinear.ordinal_index" href="#MathOptInterface.Nonlinear.ordinal_index"><code>MathOptInterface.Nonlinear.ordinal_index</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">ordinal_index(evaluator::Evaluator, c::ConstraintIndex)::Int</code></pre><p>Return the 1-indexed value of the constraint index <code>c</code> in <code>evaluator</code>.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">model = Model()
+x = MOI.VariableIndex(1)
+c1 = add_constraint(model, :($x^2), MOI.LessThan(1.0))
+c2 = add_constraint(model, :($x^2), MOI.LessThan(1.0))
+evaluator = Evaluator(model)
+MOI.initialize(evaluator, Symbol[])
+ordinal_index(evaluator, c2)  # Returns 2
+delete(model, c1)
+evaluator = Evaluator(model)
+MOI.initialize(evaluator, Symbol[])
+ordinal_index(model, c2)  # Returns 1</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Nonlinear/evaluator.jl#L11-L31">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../overview/">« Overview</a><a class="docs-footer-nextpage" href="../../Utilities/overview/">Overview »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Test/overview/index.html b/previews/PR3525/moi/submodules/Test/overview/index.html
new file mode 100644
index 00000000000..4b65a60f4e6
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Test/overview/index.html
@@ -0,0 +1,170 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Overview · JuMP</title><meta name="title" content="Overview · JuMP"/><meta property="og:title" content="Overview · JuMP"/><meta property="twitter:title" content="Overview · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Overview</a><ul class="internal"><li><a class="tocitem" href="#How-to-test-a-solver"><span>How to test a solver</span></a></li><li><a class="tocitem" href="#How-to-debug-a-failing-test"><span>How to debug a failing test</span></a></li><li><a class="tocitem" href="#How-to-add-a-test"><span>How to add a test</span></a></li></ul></li><li><a class="tocitem" href="../reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Test</a></li><li class="is-active"><a href>Overview</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Overview</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Test/overview.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="test_module"><a class="docs-heading-anchor" href="#test_module">The Test submodule</a><a id="test_module-1"></a><a class="docs-heading-anchor-permalink" href="#test_module" title="Permalink"></a></h1><p>The <code>Test</code> submodule provides tools to help solvers implement unit tests in order to ensure they implement the MathOptInterface API correctly, and to check for solver-correctness.</p><p>We use a centralized repository of tests, so that if we find a bug in one solver, instead of adding a test to that particular repository, we add it here so that all solvers can benefit.</p><h2 id="How-to-test-a-solver"><a class="docs-heading-anchor" href="#How-to-test-a-solver">How to test a solver</a><a id="How-to-test-a-solver-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-test-a-solver" title="Permalink"></a></h2><p>The skeleton below can be used for the wrapper test file of a solver named <code>FooBar</code>.</p><pre><code class="language-julia hljs"># ============================ /test/MOI_wrapper.jl ============================
+module TestFooBar
+
+import FooBar
+using Test
+
+import MathOptInterface as MOI
+
+const OPTIMIZER = MOI.instantiate(
+    MOI.OptimizerWithAttributes(FooBar.Optimizer, MOI.Silent() =&gt; true),
+)
+
+const BRIDGED = MOI.instantiate(
+    MOI.OptimizerWithAttributes(FooBar.Optimizer, MOI.Silent() =&gt; true),
+    with_bridge_type = Float64,
+)
+
+# See the docstring of MOI.Test.Config for other arguments.
+const CONFIG = MOI.Test.Config(
+    # Modify tolerances as necessary.
+    atol = 1e-6,
+    rtol = 1e-6,
+    # Use MOI.LOCALLY_SOLVED for local solvers.
+    optimal_status = MOI.OPTIMAL,
+    # Pass attributes or MOI functions to `exclude` to skip tests that
+    # rely on this functionality.
+    exclude = Any[MOI.VariableName, MOI.delete],
+)
+
+&quot;&quot;&quot;
+    runtests()
+
+This function runs all functions in the this Module starting with `test_`.
+&quot;&quot;&quot;
+function runtests()
+    for name in names(@__MODULE__; all = true)
+        if startswith(&quot;$(name)&quot;, &quot;test_&quot;)
+            @testset &quot;$(name)&quot; begin
+                getfield(@__MODULE__, name)()
+            end
+        end
+    end
+end
+
+&quot;&quot;&quot;
+    test_runtests()
+
+This function runs all the tests in MathOptInterface.Test.
+
+Pass arguments to `exclude` to skip tests for functionality that is not
+implemented or that your solver doesn&#39;t support.
+&quot;&quot;&quot;
+function test_runtests()
+    MOI.Test.runtests(
+        BRIDGED,
+        CONFIG,
+        exclude = [
+            &quot;test_attribute_NumberOfThreads&quot;,
+            &quot;test_quadratic_&quot;,
+        ],
+        # This argument is useful to prevent tests from failing on future
+        # releases of MOI that add new tests. Don&#39;t let this number get too far
+        # behind the current MOI release though. You should periodically check
+        # for new tests to fix bugs and implement new features.
+        exclude_tests_after = v&quot;0.10.5&quot;,
+    )
+    return
+end
+
+&quot;&quot;&quot;
+    test_SolverName()
+
+You can also write new tests for solver-specific functionality. Write each new
+test as a function with a name beginning with `test_`.
+&quot;&quot;&quot;
+function test_SolverName()
+    @test MOI.get(FooBar.Optimizer(), MOI.SolverName()) == &quot;FooBar&quot;
+    return
+end
+
+end # module TestFooBar
+
+# This line at tne end of the file runs all the tests!
+TestFooBar.runtests()</code></pre><p>Then modify your <code>runtests.jl</code> file to include the <code>MOI_wrapper.jl</code> file:</p><pre><code class="language-julia hljs"># ============================ /test/runtests.jl ============================
+
+using Test
+
+@testset &quot;MOI&quot; begin
+    include(&quot;test/MOI_wrapper.jl&quot;)
+end</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The optimizer <code>BRIDGED</code> constructed with <a href="../../../reference/models/#MathOptInterface.instantiate"><code>instantiate</code></a> automatically bridges constraints that are not supported by <code>OPTIMIZER</code> using the bridges listed in <a href="../../Bridges/reference/#Bridges">Bridges</a>. It is recommended for an implementation of MOI to only support constraints that are natively supported by the solver and let bridges transform the constraint to the appropriate form. For this reason it is expected that tests may not pass if <code>OPTIMIZER</code> is used instead of <code>BRIDGED</code>.</p></div></div><h2 id="How-to-debug-a-failing-test"><a class="docs-heading-anchor" href="#How-to-debug-a-failing-test">How to debug a failing test</a><a id="How-to-debug-a-failing-test-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-debug-a-failing-test" title="Permalink"></a></h2><p>When writing a solver, it&#39;s likely that you will initially fail many tests. Some failures will be bugs, but other failures you may choose to exclude.</p><p>There are two ways to exclude tests:</p><ul><li><p>Exclude tests whose names contain a string using:</p><pre><code class="language-julia hljs">MOI.Test.runtests(
+    model,
+    config;
+    exclude = String[&quot;test_to_exclude&quot;, &quot;test_conic_&quot;],
+)</code></pre><p>This will exclude tests whose name contains either of the two strings provided.</p></li><li><p>Exclude tests which rely on specific functionality using:</p><pre><code class="language-julia hljs">MOI.Test.Config(exclude = Any[MOI.VariableName, MOI.optimize!])</code></pre><p>This will exclude tests which use the <code>MOI.VariableName</code> attribute, or which call <code>MOI.optimize!</code>.</p></li></ul><p>Each test that fails can be independently called as:</p><pre><code class="language-julia hljs">model = FooBar.Optimizer()
+config = MOI.Test.Config()
+MOI.empty!(model)
+MOI.Test.test_category_name_that_failed(model, config)</code></pre><p>You can look-up the source code of the test that failed by searching for it in the <code>src/Test/test_category.jl</code> file.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Each test function also has a docstring that explains what the test is for. Use <code>? MOI.Test.test_category_name_that_failed</code> from the REPL to read it.</p></div></div><p>Periodically, you should re-run excluded tests to see if they now pass. The easiest way to do this is to swap the <code>exclude</code> keyword argument of <code>runtests</code> to <code>include</code>. For example:</p><pre><code class="language-julia hljs">MOI.Test.runtests(
+    model,
+    config;
+    exclude = String[&quot;test_to_exclude&quot;, &quot;test_conic_&quot;],
+)</code></pre><p>becomes</p><pre><code class="language-julia hljs">MOI.Test.runtests(
+    model,
+    config;
+    include = String[&quot;test_to_exclude&quot;, &quot;test_conic_&quot;],
+)</code></pre><h2 id="How-to-add-a-test"><a class="docs-heading-anchor" href="#How-to-add-a-test">How to add a test</a><a id="How-to-add-a-test-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-add-a-test" title="Permalink"></a></h2><p>To detect bugs in solvers, we add new tests to <code>MOI.Test</code>.</p><p>As an example, ECOS errored calling <a href="../../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a> twice in a row. (See <a href="https://github.com/jump-dev/ECOS.jl/pull/72">ECOS.jl PR #72</a>.) We could add a test to ECOS.jl, but that would only stop us from re-introducing the bug to ECOS.jl in the future, but it would not catch other solvers in the ecosystem with the same bug. Instead, if we add a test to <code>MOI.Test</code>, then all solvers will also check that they handle a double optimize call.</p><p>For this test, we care about correctness, rather than performance. therefore, we don&#39;t expect solvers to efficiently decide that they have already solved the problem, only that calling <a href="../../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a> twice doesn&#39;t throw an error or give the wrong answer.</p><p><strong>Step 1</strong></p><p>Install the <code>MathOptInterface</code> julia package in <a href="https://pkgdocs.julialang.org/v1/managing-packages/#developing-1"><code>dev</code> mode</a>:</p><pre><code class="language-julia hljs">julia&gt; ]
+(@v1.6) pkg&gt; dev MathOptInterface</code></pre><p><strong>Step 2</strong></p><p>From here on, proceed with making the following changes in the <code>~/.julia/dev/MathOptInterface</code> folder (or equivalent <code>dev</code> path on your machine).</p><p><strong>Step 3</strong></p><p>Since the double-optimize error involves solving an optimization problem, add a new test to <a href="https://github.com/jump-dev/MathOptInterface.jl/blob/master/src/Test/test_solve.jl">src/Test/test_solve.jl</a>:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    test_unit_optimize!_twice(model::MOI.ModelLike, config::Config)
+
+Test that calling `MOI.optimize!` twice does not error.
+
+This problem was first detected in ECOS.jl PR#72:
+https://github.com/jump-dev/ECOS.jl/pull/72
+&quot;&quot;&quot;
+function test_unit_optimize!_twice(
+    model::MOI.ModelLike,
+    config::Config{T},
+) where {T}
+    # Use the `@requires` macro to check conditions that the test function
+    # requires to run. Models failing this `@requires` check will silently skip
+    # the test.
+    @requires MOI.supports_constraint(
+        model,
+        MOI.VariableIndex,
+        MOI.GreaterThan{Float64},
+    )
+    @requires _supports(config, MOI.optimize!)
+    # If needed, you can test that the model is empty at the start of the test.
+    # You can assume that this will be the case for tests run via `runtests`.
+    # User&#39;s calling tests individually need to call `MOI.empty!` themselves.
+    @test MOI.is_empty(model)
+    # Create a simple model. Try to make this as simple as possible so that the
+    # majority of solvers can run the test.
+    x = MOI.add_variable(model)
+    MOI.add_constraint(model, x, MOI.GreaterThan(one(T)))
+    MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)
+    MOI.set(
+        model,
+        MOI.ObjectiveFunction{MOI.VariableIndex}(),
+        x,
+    )
+    # The main component of the test: does calling `optimize!` twice error?
+    MOI.optimize!(model)
+    MOI.optimize!(model)
+    # Check we have a solution.
+    @test MOI.get(model, MOI.TerminationStatus()) == MOI.OPTIMAL
+    # There is a three-argument version of `Base.isapprox` for checking
+    # approximate equality based on the tolerances defined in `config`:
+    @test isapprox(MOI.get(model, MOI.VariablePrimal(), x), one(T), config)
+    # For code-style, these tests should always `return` `nothing`.
+    return
+end</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Make sure the function is agnostic to the number type <code>T</code>; don&#39;t assume it is a <code>Float64</code> capable solver.</p></div></div><p>We also need to write a test for the test. Place this function immediately below the test you just wrote in the same file:</p><pre><code class="language-julia hljs">function setup_test(
+    ::typeof(test_unit_optimize!_twice),
+    model::MOI.Utilities.MockOptimizer,
+    ::Config,
+)
+    MOI.Utilities.set_mock_optimize!(
+        model,
+        (mock::MOI.Utilities.MockOptimizer) -&gt; MOIU.mock_optimize!(
+            mock,
+            MOI.OPTIMAL,
+            (MOI.FEASIBLE_POINT, [1.0]),
+        ),
+    )
+    return
+end</code></pre><p>Finally, you also need to implement <a href="../reference/#MathOptInterface.Test.version_added"><code>Test.version_added</code></a>. If we added this test when the latest released version of MOI was <code>v0.10.5</code>, define:</p><pre><code class="language-julia hljs">version_added(::typeof(test_unit_optimize!_twice)) = v&quot;0.10.6&quot;</code></pre><p><strong>Step 6</strong></p><p>Commit the changes to git from <code>~/.julia/dev/MathOptInterface</code> and submit the PR for review.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you need help writing a test, <a href="https://github.com/jump-dev/MathOptInterface.jl/issues/new">open an issue on GitHub</a>, or ask the <a href="https://jump.dev/chatroom/">Developer Chatroom</a>.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../Utilities/reference/">« API Reference</a><a class="docs-footer-nextpage" href="../reference/">API Reference »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Test/reference/index.html b/previews/PR3525/moi/submodules/Test/reference/index.html
new file mode 100644
index 00000000000..f519e02bb37
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Test/reference/index.html
@@ -0,0 +1,54 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>API Reference · JuMP</title><meta name="title" content="API Reference · JuMP"/><meta property="og:title" content="API Reference · JuMP"/><meta property="twitter:title" content="API Reference · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../overview/">Overview</a></li><li class="is-active"><a class="tocitem" href>API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Test</a></li><li class="is-active"><a href>API Reference</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>API Reference</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Test/reference.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="test_reference"><a class="docs-heading-anchor" href="#test_reference">The Test submodule</a><a id="test_reference-1"></a><a class="docs-heading-anchor-permalink" href="#test_reference" title="Permalink"></a></h1><p>Functions to help test implementations of MOI. See <a href="../overview/#test_module">The Test submodule</a> for more details.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Test.Config" href="#MathOptInterface.Test.Config"><code>MathOptInterface.Test.Config</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">Config(
+    ::Type{T} = Float64;
+    atol::Real = Base.rtoldefault(T),
+    rtol::Real = Base.rtoldefault(T),
+    optimal_status::MOI.TerminationStatusCode = MOI.OPTIMAL,
+    infeasible_status::MOI.TerminationStatusCode = MOI.INFEASIBLE,
+    exclude::Vector{Any} = Any[],
+) where {T}</code></pre><p>Return an object that is used to configure various tests.</p><p><strong>Configuration arguments</strong></p><ul><li><code>atol::Real = Base.rtoldefault(T)</code>: Control the absolute tolerance used when comparing solutions.</li><li><code>rtol::Real = Base.rtoldefault(T)</code>: Control the relative tolerance used when comparing solutions.</li><li><code>optimal_status = MOI.OPTIMAL</code>: Set to <code>MOI.LOCALLY_SOLVED</code> if the solver cannot prove global optimality.</li><li><code>infeasible_status = MOI.INFEASIBLE</code>: Set to <code>MOI.LOCALLY_INFEASIBLE</code> if the solver cannot prove global infeasibility.</li><li><code>exclude = Vector{Any}</code>: Pass attributes or functions to <code>exclude</code> to skip parts of tests that require certain functionality. Common arguments include:<ul><li><code>MOI.delete</code> to skip deletion-related tests</li><li><code>MOI.optimize!</code> to skip optimize-related tests</li><li><code>MOI.ConstraintDual</code> to skip dual-related tests</li><li><code>MOI.VariableName</code> to skip setting variable names</li><li><code>MOI.ConstraintName</code> to skip setting constraint names</li></ul></li></ul><p><strong>Examples</strong></p><p>For a nonlinear solver that finds local optima and does not support finding dual variables or constraint names:</p><pre><code class="language-julia hljs">Config(
+    Float64;
+    optimal_status = MOI.LOCALLY_SOLVED,
+    exclude = Any[
+        MOI.ConstraintDual,
+        MOI.VariableName,
+        MOI.ConstraintName,
+        MOI.delete,
+    ],
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Test/Test.jl#L50-L96">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Test.runtests" href="#MathOptInterface.Test.runtests"><code>MathOptInterface.Test.runtests</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">runtests(
+    model::MOI.ModelLike,
+    config::Config;
+    include::Vector{Union{String,Regex}} = String[],
+    exclude::Vector{Union{String,Regex}} = String[],
+    warn_unsupported::Bool = false,
+    exclude_tests_after::VersionNumber = v&quot;999.0.0&quot;,
+)</code></pre><p>Run all tests in <code>MathOptInterface.Test</code> on <code>model</code>.</p><p><strong>Configuration arguments</strong></p><ul><li><code>config</code> is a <a href="#MathOptInterface.Test.Config"><code>Test.Config</code></a> object that can be used to modify the behavior of tests.</li><li>If <code>include</code> is not empty, only run tests if an element from <code>include</code> <code>occursin</code> the name of the test.</li><li>If <code>exclude</code> is not empty, skip tests if an element from <code>exclude</code> <code>occursin</code> the name of the test.</li><li><code>exclude</code> takes priority over <code>include</code>.</li><li>If <code>warn_unsupported</code> is <code>false</code>, <code>runtests</code> will silently skip tests that fail with a <code>MOI.NotAllowedError</code>, <code>MOI.UnsupportedError</code>, or <code>RequirementUnmet</code> error. (The latter is thrown when an <code>@requires</code> statement returns <code>false</code>.) When <code>warn_unsupported</code> is <code>true</code>, a warning will be printed. For most cases the default behavior, <code>false</code>, is what you want, since these tests likely test functionality that is not supported by <code>model</code>. However, it can be useful to run  <code>warn_unsupported = true</code> to check you are not skipping tests due to a missing <code>supports_constraint</code> method or equivalent.</li><li><code>exclude_tests_after</code> is a version number that excludes any tests to MOI added after that version number. This is useful for solvers who can declare a fixed set of tests, and not cause their tests to break if a new patch of MOI is released with a new test.</li></ul><p>See also: <a href="#MathOptInterface.Test.setup_test"><code>setup_test</code></a>.</p><p><strong>Example</strong></p><pre><code class="language-julia hljs">config = MathOptInterface.Test.Config()
+MathOptInterface.Test.runtests(
+    model,
+    config;
+    include = [&quot;test_linear_&quot;, r&quot;^test_model_Name$&quot;],
+    exclude = [&quot;VariablePrimalStart&quot;],
+    warn_unsupported = true,
+    exclude_tests_after = v&quot;0.10.5&quot;,
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Test/Test.jl#L169-L219">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Test.setup_test" href="#MathOptInterface.Test.setup_test"><code>MathOptInterface.Test.setup_test</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">setup_test(::typeof(f), model::MOI.ModelLike, config::Config)</code></pre><p>Overload this method to modify <code>model</code> before running the test function <code>f</code> on <code>model</code> with <code>config</code>. You can also modify the fields in <code>config</code> (e.g., to loosen the default tolerances).</p><p>This function should either return <code>nothing</code>, or return a function which, when called with zero arguments, undoes the setup to return the model to its previous state. You do not need to undo any modifications to <code>config</code>.</p><p>This function is most useful when writing new tests of the tests for MOI, but it can also be used to set test-specific tolerances, etc.</p><p>See also: <a href="#MathOptInterface.Test.runtests"><code>runtests</code></a></p><p><strong>Example</strong></p><pre><code class="language-julia hljs">function MOI.Test.setup_test(
+    ::typeof(MOI.Test.test_linear_VariablePrimalStart_partial),
+    mock::MOIU.MockOptimizer,
+    ::MOI.Test.Config,
+)
+    MOIU.set_mock_optimize!(
+        mock,
+        (mock::MOIU.MockOptimizer) -&gt; MOIU.mock_optimize!(mock, [1.0, 0.0]),
+    )
+    mock.eval_variable_constraint_dual = false
+
+    function reset_function()
+        mock.eval_variable_constraint_dual = true
+        return
+    end
+    return reset_function
+end</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Test/Test.jl#L118-L155">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Test.version_added" href="#MathOptInterface.Test.version_added"><code>MathOptInterface.Test.version_added</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">version_added(::typeof(function_name))</code></pre><p>Returns the version of MOI in which the test <code>function_name</code> was added.</p><p>This method should be implemented for all new tests.</p><p>See the <code>exclude_tests_after</code> keyword of <a href="#MathOptInterface.Test.runtests"><code>runtests</code></a> for more details.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Test/Test.jl#L158-L166">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Test.@requires" href="#MathOptInterface.Test.@requires"><code>MathOptInterface.Test.@requires</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@requires(x)</code></pre><p>Check that the condition <code>x</code> is <code>true</code>. Otherwise, throw an <a href="#MathOptInterface.Test.RequirementUnmet"><code>RequirementUnmet</code></a> error to indicate that the model does not support something required by the test function.</p><p><strong>Examples</strong></p><pre><code class="language-julia hljs">@requires MOI.supports(model, MOI.Silent())
+@test MOI.get(model, MOI.Silent())</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Test/Test.jl#L288-L301">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Test.RequirementUnmet" href="#MathOptInterface.Test.RequirementUnmet"><code>MathOptInterface.Test.RequirementUnmet</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">RequirementUnmet(msg::String) &lt;: Exception</code></pre><p>An error for throwing in tests to indicate that the model does not support some requirement expected by the test function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Test/Test.jl#L273-L278">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../overview/">« Overview</a><a class="docs-footer-nextpage" href="../../../developer/checklists/">Checklists »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Utilities/overview/index.html b/previews/PR3525/moi/submodules/Utilities/overview/index.html
new file mode 100644
index 00000000000..cabc45004f8
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Utilities/overview/index.html
@@ -0,0 +1,254 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Overview · JuMP</title><meta name="title" content="Overview · JuMP"/><meta property="og:title" content="Overview · JuMP"/><meta property="twitter:title" content="Overview · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Overview</a><ul class="internal"><li><a class="tocitem" href="#Utilities.Model"><span>Utilities.Model</span></a></li><li><a class="tocitem" href="#Utilities.UniversalFallback"><span>Utilities.UniversalFallback</span></a></li><li><a class="tocitem" href="#Utilities.@model"><span>Utilities.@model</span></a></li><li><a class="tocitem" href="#Utilities.CachingOptimizer"><span>Utilities.CachingOptimizer</span></a></li><li><a class="tocitem" href="#Printing"><span>Printing</span></a></li><li><a class="tocitem" href="#Utilities.PenaltyRelaxation"><span>Utilities.PenaltyRelaxation</span></a></li><li><a class="tocitem" href="#Utilities.MatrixOfConstraints"><span>Utilities.MatrixOfConstraints</span></a></li><li><a class="tocitem" href="#ModelFilter"><span>ModelFilter</span></a></li><li><a class="tocitem" href="#Fallbacks"><span>Fallbacks</span></a></li><li><a class="tocitem" href="#DoubleDicts"><span>DoubleDicts</span></a></li></ul></li><li><a class="tocitem" href="../reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Utilities</a></li><li class="is-active"><a href>Overview</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Overview</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Utilities/overview.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-Utilities-submodule"><a class="docs-heading-anchor" href="#The-Utilities-submodule">The Utilities submodule</a><a id="The-Utilities-submodule-1"></a><a class="docs-heading-anchor-permalink" href="#The-Utilities-submodule" title="Permalink"></a></h1><p>The Utilities submodule provides a variety of functions and datastructures for managing <code>MOI.ModelLike</code> objects.</p><h2 id="Utilities.Model"><a class="docs-heading-anchor" href="#Utilities.Model">Utilities.Model</a><a id="Utilities.Model-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.Model" title="Permalink"></a></h2><p><a href="../reference/#MathOptInterface.Utilities.Model"><code>Utilities.Model</code></a> provides an implementation of a <a href="../../../reference/models/#MathOptInterface.ModelLike"><code>ModelLike</code></a> that efficiently supports all functions and sets defined within MOI. However, given the extensibility of MOI, this might not cover all use cases.</p><p>Create a model as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}</code></pre><h2 id="Utilities.UniversalFallback"><a class="docs-heading-anchor" href="#Utilities.UniversalFallback">Utilities.UniversalFallback</a><a id="Utilities.UniversalFallback-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.UniversalFallback" title="Permalink"></a></h2><p><a href="../reference/#MathOptInterface.Utilities.UniversalFallback"><code>Utilities.UniversalFallback</code></a> is a layer that sits on top of any <a href="../../../reference/models/#MathOptInterface.ModelLike"><code>ModelLike</code></a> and provides non-specialized (slower) fallbacks for constraints and attributes that the underlying <a href="../../../reference/models/#MathOptInterface.ModelLike"><code>ModelLike</code></a> does not support.</p><p>For example, <a href="../reference/#MathOptInterface.Utilities.Model"><code>Utilities.Model</code></a> doesn&#39;t support some variable attributes like <a href="../../../reference/variables/#MathOptInterface.VariablePrimalStart"><code>VariablePrimalStart</code></a>, so JuMP uses a combination of Universal fallback and <a href="../reference/#MathOptInterface.Utilities.Model"><code>Utilities.Model</code></a> as a generic problem cache:</p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())
+MOIU.UniversalFallback{MOIU.Model{Float64}}
+fallback for MOIU.Model{Float64}</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Adding a UniversalFallback means that your <code>model</code> will now support all constraints, even if the inner-model does not. This can lead to unexpected behavior.</p></div></div><h2 id="Utilities.@model"><a class="docs-heading-anchor" href="#Utilities.@model">Utilities.@model</a><a id="Utilities.@model-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.@model" title="Permalink"></a></h2><p>For advanced use cases that need efficient support for functions and sets defined outside of MOI (but still known at compile time), we provide the <a href="../reference/#MathOptInterface.Utilities.@model"><code>Utilities.@model</code></a> macro.</p><p>The <code>@model</code> macro takes a name (for a new type, which must not exist yet), eight tuples specifying the types of constraints that are supported, and then a <code>Bool</code> indicating the type is a subtype of <code>MOI.AbstractOptimizer</code> (if <code>true</code>) or <code>MOI.ModelLike</code> (if <code>false</code>).</p><p>The eight tuples are in the following order:</p><ol><li>Un-typed scalar sets, for example, <a href="../../../reference/standard_form/#MathOptInterface.Integer"><code>Integer</code></a></li><li>Typed scalar sets, for example, <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a></li><li>Un-typed vector sets, for example, <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>Nonnegatives</code></a></li><li>Typed vector sets, for example, <a href="../../../reference/standard_form/#MathOptInterface.PowerCone"><code>PowerCone</code></a></li><li>Un-typed scalar functions, for example, <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a></li><li>Typed scalar functions, for example, <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a></li><li>Un-typed vector functions, for example, <a href="../../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>VectorOfVariables</code></a></li><li>Typed vector functions, for example, <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a></li></ol><p>The tuples can contain more than one element. Typed-sets must be specified without their type parameter, for example, <code>MOI.LessThan</code>, not <code>MOI.LessThan{Float64}</code>.</p><p>Here is an example:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.@model(
+           MyNewModel,
+           (MOI.Integer,),                  # Un-typed scalar sets
+           (MOI.GreaterThan,),              # Typed scalar sets
+           (MOI.Nonnegatives,),             # Un-typed vector sets
+           (MOI.PowerCone,),                # Typed vector sets
+           (MOI.VariableIndex,),            # Un-typed scalar functions
+           (MOI.ScalarAffineFunction,),     # Typed scalar functions
+           (MOI.VectorOfVariables,),        # Un-typed vector functions
+           (MOI.VectorAffineFunction,),     # Typed vector functions
+           true,                            # &lt;:MOI.AbstractOptimizer?
+       )
+MathOptInterface.Utilities.GenericOptimizer{T, MathOptInterface.Utilities.ObjectiveContainer{T}, MathOptInterface.Utilities.VariablesContainer{T}, MyNewModelFunctionConstraints{T}} where T
+
+julia&gt; model = MyNewModel{Float64}()
+MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MyNewModelFunctionConstraints{Float64}}</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><code>MyNewModel</code> supports every <code>VariableIndex</code>-in-Set constraint, as well as <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>, and <a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a> objective functions. Implement <code>MOI.supports</code> as needed to forbid constraint and objective function combinations.</p></div></div><p>As another example, <a href="https://github.com/chkwon/PATHSolver.jl">PATHSolver</a>, which only supports <a href="../../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a>-in-<a href="../../../reference/standard_form/#MathOptInterface.Complements"><code>Complements</code></a> defines its optimizer as:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.@model(
+           PathOptimizer,
+           (),  # Scalar sets
+           (),  # Typed scalar sets
+           (MOI.Complements,),  # Vector sets
+           (),  # Typed vector sets
+           (),  # Scalar functions
+           (),  # Typed scalar functions
+           (),  # Vector functions
+           (MOI.VectorAffineFunction,),  # Typed vector functions
+           true,  # is_optimizer
+       )
+MathOptInterface.Utilities.GenericOptimizer{T, MathOptInterface.Utilities.ObjectiveContainer{T}, MathOptInterface.Utilities.VariablesContainer{T}, MathOptInterface.Utilities.VectorOfConstraints{MathOptInterface.VectorAffineFunction{T}, MathOptInterface.Complements}} where T</code></pre><p>However, <code>PathOptimizer</code> does not support some <code>VariableIndex</code>-in-Set constraints, so we must explicitly define:</p><pre><code class="language-julia-repl hljs">julia&gt; function MOI.supports_constraint(
+           ::PathOptimizer,
+           ::Type{MOI.VariableIndex},
+           ::Type{Union{&lt;:MOI.Semiinteger,MOI.Semicontinuous,MOI.ZeroOne,MOI.Integer}}
+       )
+           return false
+       end</code></pre><p>Finally, PATH doesn&#39;t support an objective function, so we need to add:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.supports(::PathOptimizer, ::MOI.ObjectiveFunction) = false</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This macro creates a new type, so it must be called from the top-level of a module, for example, it cannot be called from inside a function.</p></div></div><h2 id="Utilities.CachingOptimizer"><a class="docs-heading-anchor" href="#Utilities.CachingOptimizer">Utilities.CachingOptimizer</a><a id="Utilities.CachingOptimizer-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.CachingOptimizer" title="Permalink"></a></h2><p>A [<code>Utilities.CachingOptimizer</code>] is an MOI layer that abstracts the difference between solvers that support incremental modification (for example, they support adding variables one-by-one), and solvers that require the entire problem in a single API call (for example, they only accept the <code>A</code>, <code>b</code> and <code>c</code> matrices of a linear program).</p><p>It has two parts:</p><ol><li>A cache, where the model can be built and modified incrementally</li><li>An optimizer, which is used to solve the problem</li></ol><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.CachingOptimizer(
+           MOI.Utilities.Model{Float64}(),
+           PathOptimizer{Float64}(),
+       )
+MOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}
+in state EMPTY_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.Model{Float64}
+with optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}</code></pre><p>A <a href="../reference/#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a> may be in one of three possible states:</p><ul><li><code>NO_OPTIMIZER</code>: The CachingOptimizer does not have any optimizer.</li><li><code>EMPTY_OPTIMIZER</code>: The CachingOptimizer has an empty optimizer, and it is not synchronized with the cached model. Modifications are forwarded to the cache, but <em>not</em> to the optimizer.</li><li><code>ATTACHED_OPTIMIZER</code>: The CachingOptimizer has an optimizer, and it is synchronized with the cached model. Modifications are forwarded to the optimizer. If the optimizer does not support modifications, and error will be thrown.</li></ul><p>Use <a href="../reference/#MathOptInterface.Utilities.attach_optimizer"><code>Utilities.attach_optimizer</code></a> to go from <code>EMPTY_OPTIMIZER</code> to <code>ATTACHED_OPTIMIZER</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.attach_optimizer(model)
+
+julia&gt; model
+MOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}
+in state ATTACHED_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.Model{Float64}
+with optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>You must be in <code>ATTACHED_OPTIMIZER</code> to use <a href="../../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a>.</p></div></div><p>Use <a href="../reference/#MathOptInterface.Utilities.reset_optimizer"><code>Utilities.reset_optimizer</code></a> to go from <code>ATTACHED_OPTIMIZER</code> to <code>EMPTY_OPTIMIZER</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.reset_optimizer(model)
+
+julia&gt; model
+MOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}
+in state EMPTY_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.Model{Float64}
+with optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Calling <code>MOI.empty!(model)</code> also resets the state to <code>EMPTY_OPTIMIZER</code>. So after emptying a model, the modification will only be applied to the cache.</p></div></div><p>Use <a href="../reference/#MathOptInterface.Utilities.drop_optimizer"><code>Utilities.drop_optimizer</code></a> to go from any state to <code>NO_OPTIMIZER</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.drop_optimizer(model)
+
+julia&gt; model
+MOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}
+in state NO_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.Model{Float64}
+with optimizer nothing</code></pre><p>Pass an empty optimizer to <a href="../reference/#MathOptInterface.Utilities.reset_optimizer"><code>Utilities.reset_optimizer</code></a> to go from <code>NO_OPTIMIZER</code> to <code>EMPTY_OPTIMIZER</code>:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.reset_optimizer(model, PathOptimizer{Float64}())
+
+julia&gt; model
+MOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}
+in state EMPTY_OPTIMIZER
+in mode AUTOMATIC
+with model cache MOIU.Model{Float64}
+with optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}</code></pre><p>Deciding when to attach and reset the optimizer is tedious, and you will often write code like this:</p><pre><code class="language-julia hljs">try
+    # modification
+catch
+    MOI.Utilities.reset_optimizer(model)
+    # Re-try modification
+end</code></pre><p>To make this easier, <a href="../reference/#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a> has two modes of operation:</p><ul><li><code>AUTOMATIC</code>: The <code>CachingOptimizer</code> changes its state when necessary. Attempting to add a constraint or perform a modification not supported by the optimizer results in a drop to <code>EMPTY_OPTIMIZER</code> mode.</li><li><code>MANUAL</code>: The user must change the state of the <code>CachingOptimizer</code>. Attempting to perform an operation in the incorrect state results in an error.</li></ul><p>By default, <code>AUTOMATIC</code> mode is chosen. However, you can create a <code>CachingOptimizer</code> in <code>MANUAL</code> mode as follows:</p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.CachingOptimizer(
+           MOI.Utilities.Model{Float64}(),
+           MOI.Utilities.MANUAL,
+       )
+MOIU.CachingOptimizer{MOI.AbstractOptimizer, MOIU.Model{Float64}}
+in state NO_OPTIMIZER
+in mode MANUAL
+with model cache MOIU.Model{Float64}
+with optimizer nothing
+
+julia&gt; MOI.Utilities.reset_optimizer(model, PathOptimizer{Float64}())
+
+julia&gt; model
+MOIU.CachingOptimizer{MOI.AbstractOptimizer, MOIU.Model{Float64}}
+in state EMPTY_OPTIMIZER
+in mode MANUAL
+with model cache MOIU.Model{Float64}
+with optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}</code></pre><h2 id="Printing"><a class="docs-heading-anchor" href="#Printing">Printing</a><a id="Printing-1"></a><a class="docs-heading-anchor-permalink" href="#Printing" title="Permalink"></a></h2><p>Use <code>print</code> to print the formulation of the model.</p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}();
+
+julia&gt; x = MOI.add_variable(model)
+MOI.VariableIndex(1)
+
+julia&gt; MOI.set(model, MOI.VariableName(), x, &quot;x_var&quot;)
+
+julia&gt; MOI.add_constraint(model, x, MOI.ZeroOne())
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(1)
+
+julia&gt; MOI.set(model, MOI.ObjectiveFunction{typeof(x)}(), x)
+
+julia&gt; MOI.set(model, MOI.ObjectiveSense(), MOI.MAX_SENSE)
+
+julia&gt; print(model)
+Maximize VariableIndex:
+ x_var
+
+Subject to:
+
+VariableIndex-in-ZeroOne
+ x_var ∈ {0, 1}</code></pre><p>Use <a href="../reference/#MathOptInterface.Utilities.latex_formulation"><code>Utilities.latex_formulation</code></a> to display the model in LaTeX form:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.latex_formulation(model)
+$$ \begin{aligned}
+\max\quad &amp; x\_var \\
+\text{Subject to}\\
+ &amp; \text{VariableIndex-in-ZeroOne} \\
+ &amp; x\_var \in \{0, 1\} \\
+\end{aligned} $$</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>In IJulia, calling <code>print</code> or ending a cell with <a href="../reference/#MathOptInterface.Utilities.latex_formulation"><code>Utilities.latex_formulation</code></a> will render the model in LaTeX.</p></div></div><h2 id="Utilities.PenaltyRelaxation"><a class="docs-heading-anchor" href="#Utilities.PenaltyRelaxation">Utilities.PenaltyRelaxation</a><a id="Utilities.PenaltyRelaxation-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.PenaltyRelaxation" title="Permalink"></a></h2><p>Pass <a href="../reference/#MathOptInterface.Utilities.PenaltyRelaxation"><code>Utilities.PenaltyRelaxation</code></a> to <a href="../../../reference/modification/#MathOptInterface.modify"><code>modify</code></a> to relax the problem by adding penalized slack variables to the constraints. This is helpful when debugging sources of infeasible models.</p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}();
+
+julia&gt; x = MOI.add_variable(model);
+
+julia&gt; MOI.set(model, MOI.VariableName(), x, &quot;x&quot;)
+
+julia&gt; c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));
+
+julia&gt; map = MOI.modify(model, MOI.Utilities.PenaltyRelaxation(Dict(c =&gt; 2.0)));
+
+julia&gt; print(model)
+Minimize ScalarAffineFunction{Float64}:
+ 0.0 + 2.0 v[2]
+
+Subject to:
+
+ScalarAffineFunction{Float64}-in-LessThan{Float64}
+ 0.0 + 1.0 x - 1.0 v[2] &lt;= 2.0
+
+VariableIndex-in-GreaterThan{Float64}
+ v[2] &gt;= 0.0
+
+julia&gt; map[c]
+0.0 + 1.0 MOI.VariableIndex(2)</code></pre><p>You can also modify a single constraint using <a href="../reference/#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>Utilities.ScalarPenaltyRelaxation</code></a>:</p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}();
+
+julia&gt; x = MOI.add_variable(model);
+
+julia&gt; MOI.set(model, MOI.VariableName(), x, &quot;x&quot;)
+
+julia&gt; c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));
+
+julia&gt; f = MOI.modify(model, c, MOI.Utilities.ScalarPenaltyRelaxation(2.0));
+
+julia&gt; print(model)
+Minimize ScalarAffineFunction{Float64}:
+ 0.0 + 2.0 v[2]
+
+Subject to:
+
+ScalarAffineFunction{Float64}-in-LessThan{Float64}
+ 0.0 + 1.0 x - 1.0 v[2] &lt;= 2.0
+
+VariableIndex-in-GreaterThan{Float64}
+ v[2] &gt;= 0.0
+
+julia&gt; f
+0.0 + 1.0 MOI.VariableIndex(2)</code></pre><h2 id="Utilities.MatrixOfConstraints"><a class="docs-heading-anchor" href="#Utilities.MatrixOfConstraints">Utilities.MatrixOfConstraints</a><a id="Utilities.MatrixOfConstraints-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.MatrixOfConstraints" title="Permalink"></a></h2><p>The constraints of <a href="../reference/#MathOptInterface.Utilities.Model"><code>Utilities.Model</code></a> are stored as a vector of tuples of function and set in a <code>Utilities.VectorOfConstraints</code>. Other representations can be used by parameterizing the type <a href="../reference/#MathOptInterface.Utilities.GenericModel"><code>Utilities.GenericModel</code></a> (resp. <a href="../reference/#MathOptInterface.Utilities.GenericOptimizer"><code>Utilities.GenericOptimizer</code></a>). For instance, if all non-<code>VariableIndex</code> constraints are affine, the coefficients of all the constraints can be stored in a single sparse matrix using <a href="../reference/#MathOptInterface.Utilities.MatrixOfConstraints"><code>Utilities.MatrixOfConstraints</code></a>. The constraints storage can even be customized up to a point where it exactly matches the storage of the solver of interest, in which case <a href="../../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a> can be implemented for the solver by calling <a href="../../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a> to this custom model.</p><p>For instance, <a href="https://github.com/jump-dev/Clp.jl">Clp</a> defines the following model:</p><pre><code class="language-julia hljs">MOI.Utilities.@product_of_scalar_sets(LP, MOI.EqualTo{T}, MOI.LessThan{T}, MOI.GreaterThan{T})
+const Model = MOI.Utilities.GenericModel{
+    Float64,
+    MOI.Utilities.MatrixOfConstraints{
+        Float64,
+        MOI.Utilities.MutableSparseMatrixCSC{Float64,Cint,MOI.Utilities.ZeroBasedIndexing},
+        MOI.Utilities.Hyperrectangle{Float64},
+        LP{Float64},
+    },
+}</code></pre><p>The <a href="../../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a> operation can now be implemented as follows:</p><pre><code class="language-julia hljs">function _copy_to(dest::Optimizer, src::Model)
+    @assert MOI.is_empty(dest)
+    A = src.constraints.coefficients
+    row_bounds = src.constraints.constants
+    Clp_loadProblem(
+        dest,
+        A.n,
+        A.m,
+        A.colptr,
+        A.rowval,
+        A.nzval,
+        src.lower_bound,
+        src.upper_bound,
+        # (...) objective vector (omitted),
+        row_bounds.lower,
+        row_bounds.upper,
+    )
+    # Set objective sense and constant (omitted)
+    return
+end
+
+function MOI.copy_to(dest::Optimizer, src::Model)
+    _copy_to(dest, src)
+    return MOI.Utilities.identity_index_map(src)
+end
+
+function MOI.copy_to(
+    dest::Optimizer,
+    src::MOI.Utilities.UniversalFallback{Model},
+)
+    # Copy attributes from `src` to `dest` and error in case any unsupported
+    # constraints or attributes are set in `UniversalFallback`.
+    return MOI.copy_to(dest, src.model)
+end
+
+function MOI.copy_to(
+    dest::Optimizer,
+    src::MOI.ModelLike,
+)
+    model = Model()
+    index_map = MOI.copy_to(model, src)
+    _copy_to(dest, model)
+    return index_map
+end</code></pre><h2 id="ModelFilter"><a class="docs-heading-anchor" href="#ModelFilter">ModelFilter</a><a id="ModelFilter-1"></a><a class="docs-heading-anchor-permalink" href="#ModelFilter" title="Permalink"></a></h2><p>Utilities provides <a href="../reference/#MathOptInterface.Utilities.ModelFilter"><code>Utilities.ModelFilter</code></a> as a useful tool to copy a subset of a model. For example, given an infeasible model, we can copy the irreducible infeasible subsystem (for models implementing <a href="../../../reference/models/#MathOptInterface.ConstraintConflictStatus"><code>ConstraintConflictStatus</code></a>) as follows:</p><pre><code class="language-julia hljs">my_filter(::Any) = true
+function my_filter(ci::MOI.ConstraintIndex)
+    status = MOI.get(dest, MOI.ConstraintConflictStatus(), ci)
+    return status != MOI.NOT_IN_CONFLICT
+end
+filtered_src = MOI.Utilities.ModelFilter(my_filter, src)
+index_map = MOI.copy_to(dest, filtered_src)</code></pre><h2 id="Fallbacks"><a class="docs-heading-anchor" href="#Fallbacks">Fallbacks</a><a id="Fallbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Fallbacks" title="Permalink"></a></h2><p>The value of some attributes can be inferred from the value of other attributes.</p><p>For example, the value of <a href="../../../reference/models/#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a> can be computed using <a href="../../../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> and <a href="../../../reference/variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a>.</p><p>When a solver gives direct access to an attribute, it is better to return this value. However, if this is not the case, <a href="../reference/#MathOptInterface.Utilities.get_fallback"><code>Utilities.get_fallback</code></a> can be used instead. For example:</p><pre><code class="language-julia hljs">function MOI.get(model::Optimizer, attr::MOI.ObjectiveFunction)
+    return MOI.Utilities.get_fallback(model, attr)
+end</code></pre><h2 id="DoubleDicts"><a class="docs-heading-anchor" href="#DoubleDicts">DoubleDicts</a><a id="DoubleDicts-1"></a><a class="docs-heading-anchor-permalink" href="#DoubleDicts" title="Permalink"></a></h2><p>When writing MOI interfaces, we often need to handle situations in which we map <a href="../../../reference/constraints/#MathOptInterface.ConstraintIndex"><code>ConstraintIndex</code></a>s to different values. For example, to a string for <a href="../../../reference/constraints/#MathOptInterface.ConstraintName"><code>ConstraintName</code></a>.</p><p>One option is to use a dictionary like <code>Dict{MOI.ConstraintIndex,String}</code>. However, this incurs a performance cost because the key is not a concrete type.</p><p>The DoubleDicts submodule helps this situation by providing two types main types <a href="../reference/#MathOptInterface.Utilities.DoubleDicts.DoubleDict"><code>Utilities.DoubleDicts.DoubleDict</code></a> and <a href="../reference/#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict"><code>Utilities.DoubleDicts.IndexDoubleDict</code></a>. These types act like normal dictionaries, but internally they use more efficient dictionaries specialized to the type of the function-set pair.</p><p>The most common usage of a <code>DoubleDict</code> is in the <code>index_map</code> returned by <a href="../../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a>. Performance can be improved, by using a function barrier. That is, instead of code like:</p><pre><code class="language-julia hljs">index_map = MOI.copy_to(dest, src)
+for (F, S) in MOI.get(src, MOI.ListOfConstraintTypesPresent())
+    for ci in MOI.get(src, MOI.ListOfConstraintIndices{F,S}())
+        dest_ci = index_map[ci]
+        # ...
+    end
+end</code></pre><p>use instead:</p><pre><code class="language-julia hljs">function function_barrier(
+    dest,
+    src,
+    index_map::MOI.Utilities.DoubleDicts.IndexDoubleDictInner{F,S},
+) where {F,S}
+    for ci in MOI.get(src, MOI.ListOfConstraintIndices{F,S}())
+        dest_ci = index_map[ci]
+        # ...
+    end
+    return
+end
+
+index_map = MOI.copy_to(dest, src)
+for (F, S) in MOI.get(src, MOI.ListOfConstraintTypesPresent())
+    function_barrier(dest, src, index_map[F, S])
+end</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../Nonlinear/reference/">« API Reference</a><a class="docs-footer-nextpage" href="../reference/">API Reference »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/submodules/Utilities/reference/index.html b/previews/PR3525/moi/submodules/Utilities/reference/index.html
new file mode 100644
index 00000000000..0eff54d13f7
--- /dev/null
+++ b/previews/PR3525/moi/submodules/Utilities/reference/index.html
@@ -0,0 +1,268 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>API Reference · JuMP</title><meta name="title" content="API Reference · JuMP"/><meta property="og:title" content="API Reference · JuMP"/><meta property="twitter:title" content="API Reference · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../../assets/documenter.js"></script><script src="../../../../search_index.js"></script><script src="../../../../siteinfo.js"></script><script src="../../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../../assets/themeswap.js"></script><link href="../../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../../"><img class="docs-light-only" src="../../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../">Introduction</a></li><li><a class="tocitem" href="../../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../reference/models/">Models</a></li><li><a class="tocitem" href="../../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox" checked/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../overview/">Overview</a></li><li class="is-active"><a class="tocitem" href>API Reference</a><ul class="internal"><li><a class="tocitem" href="#Utilities.Model"><span>Utilities.Model</span></a></li><li><a class="tocitem" href="#Utilities.UniversalFallback"><span>Utilities.UniversalFallback</span></a></li><li><a class="tocitem" href="#Utilities.@model"><span>Utilities.@model</span></a></li><li><a class="tocitem" href="#Caching-optimizer"><span>Caching optimizer</span></a></li><li><a class="tocitem" href="#Mock-optimizer"><span>Mock optimizer</span></a></li><li><a class="tocitem" href="#Printing"><span>Printing</span></a></li><li><a class="tocitem" href="#Copy-utilities"><span>Copy utilities</span></a></li><li><a class="tocitem" href="#Penalty-relaxation"><span>Penalty relaxation</span></a></li><li><a class="tocitem" href="#MatrixOfConstraints"><span>MatrixOfConstraints</span></a></li><li><a class="tocitem" href="#Fallbacks"><span>Fallbacks</span></a></li><li><a class="tocitem" href="#Function-utilities"><span>Function utilities</span></a></li><li><a class="tocitem" href="#Constraint-utilities"><span>Constraint utilities</span></a></li><li><a class="tocitem" href="#Set-utilities"><span>Set utilities</span></a></li><li><a class="tocitem" href="#DoubleDicts"><span>DoubleDicts</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../Test/overview/">Overview</a></li><li><a class="tocitem" href="../../Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Submodules</a></li><li><a class="is-disabled">Utilities</a></li><li class="is-active"><a href>API Reference</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>API Reference</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Utilities/reference.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h2 id="Utilities.Model"><a class="docs-heading-anchor" href="#Utilities.Model">Utilities.Model</a><a id="Utilities.Model-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.Model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.Model" href="#MathOptInterface.Utilities.Model"><code>MathOptInterface.Utilities.Model</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">MOI.Utilities.Model{T}() where {T}</code></pre><p>An implementation of <code>ModelLike</code> that supports all functions and sets defined in MOI. It is parameterized by the coefficient type.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; import MathOptInterface as MOI
+
+julia&gt; model = MOI.Utilities.Model{Float64}()
+MOIU.Model{Float64}</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/model.jl#L823-L837">source</a></section></article><h2 id="Utilities.UniversalFallback"><a class="docs-heading-anchor" href="#Utilities.UniversalFallback">Utilities.UniversalFallback</a><a id="Utilities.UniversalFallback-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.UniversalFallback" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.UniversalFallback" href="#MathOptInterface.Utilities.UniversalFallback"><code>MathOptInterface.Utilities.UniversalFallback</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">UniversalFallback</code></pre><p>The <code>UniversalFallback</code> can be applied on a <a href="../../../reference/models/#MathOptInterface.ModelLike"><code>MOI.ModelLike</code></a> <code>model</code> to create the model <code>UniversalFallback(model)</code> supporting <em>any</em> constraint and attribute. This allows to have a specialized implementation in <code>model</code> for performance critical constraints and attributes while still supporting other attributes with a small performance penalty. Note that <code>model</code> is unaware of constraints and attributes stored by <code>UniversalFallback</code> so this is not appropriate if <code>model</code> is an optimizer (for this reason, <a href="../../../reference/models/#MathOptInterface.optimize!"><code>MOI.optimize!</code></a> has not been implemented). In that case, optimizer bridges should be used instead.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/universalfallback.jl#L7-L19">source</a></section></article><h2 id="Utilities.@model"><a class="docs-heading-anchor" href="#Utilities.@model">Utilities.@model</a><a id="Utilities.@model-1"></a><a class="docs-heading-anchor-permalink" href="#Utilities.@model" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.@model" href="#MathOptInterface.Utilities.@model"><code>MathOptInterface.Utilities.@model</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">macro model(
+    model_name,
+    scalar_sets,
+    typed_scalar_sets,
+    vector_sets,
+    typed_vector_sets,
+    scalar_functions,
+    typed_scalar_functions,
+    vector_functions,
+    typed_vector_functions,
+    is_optimizer = false
+)</code></pre><p>Creates a type <code>model_name</code> implementing the MOI model interface and supporting all combinations of the provided functions and sets.</p><p>Each <code>typed_</code> <code>scalar</code>/<code>vector</code> <code>sets</code>/<code>functions</code> argument is a tuple of types. A type is &quot;typed&quot; if it has a coefficient <code>{T}</code> as the first type parameter.</p><p><strong>Tuple syntax</strong></p><p>To give no set/function, write <code>()</code>. To give one set or function <code>X</code>, write <code>(X,)</code>.</p><p><strong><code>is_optimizer</code></strong></p><p>If <code>is_optimizer = true</code>, the resulting struct is a of <a href="#MathOptInterface.Utilities.GenericOptimizer"><code>GenericOptimizer</code></a>, which is a subtype of <a href="../../../reference/models/#MathOptInterface.AbstractOptimizer"><code>MOI.AbstractOptimizer</code></a>, otherwise, it is a <a href="#MathOptInterface.Utilities.GenericModel"><code>GenericModel</code></a>, which is a subtype of <a href="../../../reference/models/#MathOptInterface.ModelLike"><code>MOI.ModelLike</code></a>.</p><p><strong>VariableIndex</strong></p><ul><li>The function <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> must not be given in <code>scalar_functions</code>.</li><li>The model supports <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a>-in-<code>S</code> constraints where <code>S</code> is <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.Semicontinuous"><code>MOI.Semicontinuous</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger</code></a>.</li><li>The sets supported with <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> cannot be controlled from the macro; use <a href="#MathOptInterface.Utilities.UniversalFallback"><code>UniversalFallback</code></a> to support more sets.</li></ul><p><strong>Examples</strong></p><p>The model describing a linear program would be:</p><pre><code class="language-julia hljs">@model(
+    LPModel,                                          # model_name
+    (),                                               # untyped scalar sets
+    (MOI.EqualTo, MOI.GreaterThan, MOI.LessThan, MOI.Interval), #   typed scalar sets
+    (MOI.Zeros, MOI.Nonnegatives, MOI.Nonpositives),  # untyped vector sets
+    (),                                               #   typed vector sets
+    (),                                               # untyped scalar functions
+    (MOI.ScalarAffineFunction,),                      #   typed scalar functions
+    (MOI.VectorOfVariables,),                         # untyped vector functions
+    (MOI.VectorAffineFunction,),                      #   typed vector functions
+    false,                                            # is_optimizer
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/model.jl#L621-L680">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.GenericModel" href="#MathOptInterface.Utilities.GenericModel"><code>MathOptInterface.Utilities.GenericModel</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct GenericModel{T,O,V,C} &lt;: AbstractModelLike{T}</code></pre><p>Implements a model supporting coefficients of type <code>T</code> and:</p><ul><li>An objective function stored in <code>.objective::O</code></li><li>Variables and <code>VariableIndex</code> constraints stored in <code>.variable_bounds::V</code></li><li><code>F</code>-in-<code>S</code> constraints (excluding <code>VariableIndex</code> constraints) stored in <code>.constraints::C</code></li></ul><p>All interactions take place via the MOI interface, so the types <code>O</code>, <code>V</code>, and <code>C</code> must implement the API as needed for their functionality.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/model.jl#L9-L21">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.GenericOptimizer" href="#MathOptInterface.Utilities.GenericOptimizer"><code>MathOptInterface.Utilities.GenericOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct GenericOptimizer{T,O,V,C} &lt;: AbstractOptimizer{T}</code></pre><p>Implements a model supporting coefficients of type <code>T</code> and:</p><ul><li>An objective function stored in <code>.objective::O</code></li><li>Variables and <code>VariableIndex</code> constraints stored in <code>.variable_bounds::V</code></li><li><code>F</code>-in-<code>S</code> constraints (excluding <code>VariableIndex</code> constraints) stored in <code>.constraints::C</code></li></ul><p>All interactions take place via the MOI interface, so the types <code>O</code>, <code>V</code>, and <code>C</code> must implement the API as needed for their functionality.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/model.jl#L52-L64">source</a></section></article><h3 id=".objective"><a class="docs-heading-anchor" href="#.objective"><code>.objective</code></a><a id=".objective-1"></a><a class="docs-heading-anchor-permalink" href="#.objective" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.ObjectiveContainer" href="#MathOptInterface.Utilities.ObjectiveContainer"><code>MathOptInterface.Utilities.ObjectiveContainer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ObjectiveContainer{T}</code></pre><p>A helper struct to simplify the handling of objective functions in Utilities.Model.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/objective_container.jl#L7-L12">source</a></section></article><h3 id=".variables"><a class="docs-heading-anchor" href="#.variables"><code>.variables</code></a><a id=".variables-1"></a><a class="docs-heading-anchor-permalink" href="#.variables" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.VariablesContainer" href="#MathOptInterface.Utilities.VariablesContainer"><code>MathOptInterface.Utilities.VariablesContainer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct VariablesContainer{T} &lt;: AbstractVectorBounds
+    set_mask::Vector{UInt16}
+    lower::Vector{T}
+    upper::Vector{T}
+end</code></pre><p>A struct for storing variables and VariableIndex-related constraints. Used in <code>MOI.Utilities.Model</code> by default.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/variables_container.jl#L180-L189">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.FreeVariables" href="#MathOptInterface.Utilities.FreeVariables"><code>MathOptInterface.Utilities.FreeVariables</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct FreeVariables &lt;: MOI.ModelLike
+    n::Int64
+    FreeVariables() = new(0)
+end</code></pre><p>A struct for storing free variables that can be used as the <code>variables</code> field of <a href="#MathOptInterface.Utilities.GenericModel"><code>GenericModel</code></a> or <a href="#MathOptInterface.Utilities.GenericModel"><code>GenericModel</code></a>. It represents a model that does not support any constraint nor objective function.</p><p><strong>Example</strong></p><p>The following model type represents a conic model in geometric form. As opposed to <a href="#MathOptInterface.Utilities.VariablesContainer"><code>VariablesContainer</code></a>, <code>FreeVariables</code> does not support constraint bounds so they are bridged into an affine constraint in the <a href="../../../reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a> cone as expected for the geometric conic form.</p><pre><code class="language-jldocstest hljs">julia&gt; MOI.Utilities.@product_of_sets(
+    Cones,
+    MOI.Zeros,
+    MOI.Nonnegatives,
+    MOI.SecondOrderCone,
+    MOI.PositiveSemidefiniteConeTriangle,
+);
+
+julia&gt; const ConicModel{T} = MOI.Utilities.GenericOptimizer{
+    T,
+    MOI.Utilities.ObjectiveContainer{T},
+    MOI.Utilities.FreeVariables,
+    MOI.Utilities.MatrixOfConstraints{
+        T,
+        MOI.Utilities.MutableSparseMatrixCSC{
+            T,
+            Int,
+            MOI.Utilities.OneBasedIndexing,
+        },
+        Vector{T},
+        Cones{T},
+    },
+};
+
+julia&gt; model = MOI.instantiate(ConicModel{Float64}, with_bridge_type=Float64);
+
+julia&gt; x = MOI.add_variable(model)
+MathOptInterface.VariableIndex(1)
+
+julia&gt; c = MOI.add_constraint(model, x, MOI.GreaterThan(1.0))
+MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(1)
+
+julia&gt; MOI.Bridges.is_bridged(model, c)
+true
+
+julia&gt; bridge = MOI.Bridges.bridge(model, c)
+MathOptInterface.Bridges.Constraint.VectorizeBridge{Float64, MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives, MathOptInterface.VariableIndex}(MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1), 1.0)
+
+julia&gt; bridge.vector_constraint
+MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1)
+
+julia&gt; MOI.Bridges.is_bridged(model, bridge.vector_constraint)
+false</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/free_variables.jl#L7-L69">source</a></section></article><h3 id=".constraints"><a class="docs-heading-anchor" href="#.constraints"><code>.constraints</code></a><a id=".constraints-1"></a><a class="docs-heading-anchor-permalink" href="#.constraints" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.VectorOfConstraints" href="#MathOptInterface.Utilities.VectorOfConstraints"><code>MathOptInterface.Utilities.VectorOfConstraints</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct VectorOfConstraints{
+    F&lt;:MOI.AbstractFunction,
+    S&lt;:MOI.AbstractSet,
+} &lt;: MOI.ModelLike
+    constraints::CleverDicts.CleverDict{
+        MOI.ConstraintIndex{F,S},
+        Tuple{F,S},
+        typeof(CleverDicts.key_to_index),
+        typeof(CleverDicts.index_to_key),
+    }
+end</code></pre><p>A struct storing <code>F</code>-in-<code>S</code> constraints as a mapping between the constraint indices to the corresponding tuple of function and set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/vector_of_constraints.jl#L24-L39">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.StructOfConstraints" href="#MathOptInterface.Utilities.StructOfConstraints"><code>MathOptInterface.Utilities.StructOfConstraints</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type StructOfConstraints &lt;: MOI.ModelLike end</code></pre><p>A struct storing a subfields other structs storing constraints of different types.</p><p>See <a href="#MathOptInterface.Utilities.@struct_of_constraints_by_function_types"><code>Utilities.@struct_of_constraints_by_function_types</code></a> and <a href="#MathOptInterface.Utilities.@struct_of_constraints_by_set_types"><code>Utilities.@struct_of_constraints_by_set_types</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/struct_of_constraints.jl#L7-L15">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.@struct_of_constraints_by_function_types" href="#MathOptInterface.Utilities.@struct_of_constraints_by_function_types"><code>MathOptInterface.Utilities.@struct_of_constraints_by_function_types</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">Utilities.@struct_of_constraints_by_function_types(name, func_types...)</code></pre><p>Given a vector of <code>n</code> function types <code>(F1, F2,..., Fn)</code> in <code>func_types</code>, defines a subtype of <code>StructOfConstraints</code> of name <code>name</code> and which type parameters <code>{T, C1, C2, ..., Cn}</code>. It contains <code>n</code> field where the <code>i</code>th field has type <code>Ci</code> and stores the constraints of function type <code>Fi</code>.</p><p>The expression <code>Fi</code> can also be a union in which case any constraint for which the function type is in the union is stored in the field with type <code>Ci</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/struct_of_constraints.jl#L396-L407">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.@struct_of_constraints_by_set_types" href="#MathOptInterface.Utilities.@struct_of_constraints_by_set_types"><code>MathOptInterface.Utilities.@struct_of_constraints_by_set_types</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">Utilities.@struct_of_constraints_by_set_types(name, func_types...)</code></pre><p>Given a vector of <code>n</code> set types <code>(S1, S2,..., Sn)</code> in <code>func_types</code>, defines a subtype of <code>StructOfConstraints</code> of name <code>name</code> and which type parameters <code>{T, C1, C2, ..., Cn}</code>. It contains <code>n</code> field where the <code>i</code>th field has type <code>Ci</code> and stores the constraints of set type <code>Si</code>. The expression <code>Si</code> can also be a union in which case any constraint for which the set type is in the union is stored in the field with type <code>Ci</code>. This can be useful if <code>Ci</code> is a <a href="#MathOptInterface.Utilities.MatrixOfConstraints"><code>MatrixOfConstraints</code></a> in order to concatenate the coefficients of constraints of several different set types in the same matrix.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/struct_of_constraints.jl#L413-L426">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.struct_of_constraint_code" href="#MathOptInterface.Utilities.struct_of_constraint_code"><code>MathOptInterface.Utilities.struct_of_constraint_code</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">struct_of_constraint_code(struct_name, types, field_types = nothing)</code></pre><p>Given a vector of <code>n</code> <code>Union{SymbolFun,_UnionSymbolFS{SymbolFun}}</code> or <code>Union{SymbolSet,_UnionSymbolFS{SymbolSet}}</code> in <code>types</code>, defines a subtype of <code>StructOfConstraints</code> of name <code>name</code> and which type parameters <code>{T, F1, F2, ..., Fn}</code> if <code>field_types</code> is <code>nothing</code> and a <code>{T}</code> otherwise. It contains <code>n</code> field where the <code>i</code>th field has type <code>Ci</code> if <code>field_types</code> is <code>nothing</code> and type <code>field_types[i]</code> otherwise. If <code>types</code> is vector of <code>Union{SymbolFun,_UnionSymbolFS{SymbolFun}}</code> (resp. <code>Union{SymbolSet,_UnionSymbolFS{SymbolSet}}</code>) then the constraints of that function (resp. set) type are stored in the corresponding field.</p><p>This function is used by the macros <a href="#MathOptInterface.Utilities.@model"><code>@model</code></a>, <a href="#MathOptInterface.Utilities.@struct_of_constraints_by_function_types"><code>@struct_of_constraints_by_function_types</code></a> and <a href="#MathOptInterface.Utilities.@struct_of_constraints_by_set_types"><code>@struct_of_constraints_by_set_types</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/struct_of_constraints.jl#L263-L280">source</a></section></article><h2 id="Caching-optimizer"><a class="docs-heading-anchor" href="#Caching-optimizer">Caching optimizer</a><a id="Caching-optimizer-1"></a><a class="docs-heading-anchor-permalink" href="#Caching-optimizer" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.CachingOptimizer" href="#MathOptInterface.Utilities.CachingOptimizer"><code>MathOptInterface.Utilities.CachingOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">CachingOptimizer</code></pre><p><code>CachingOptimizer</code> is an intermediate layer that stores a cache of the model and links it with an optimizer. It supports incremental model construction and modification even when the optimizer doesn&#39;t.</p><p><strong>Constructors</strong></p><pre><code class="language-julia hljs">    CachingOptimizer(cache::MOI.ModelLike, optimizer::AbstractOptimizer)</code></pre><p>Creates a <code>CachingOptimizer</code> in <code>AUTOMATIC</code> mode, with the optimizer <code>optimizer</code>.</p><p>The type of the optimizer returned is <code>CachingOptimizer{typeof(optimizer), typeof(cache)}</code> so it does not support the function <code>reset_optimizer(::CachingOptimizer, new_optimizer)</code> if the type of <code>new_optimizer</code> is different from the type of <code>optimizer</code>.</p><pre><code class="language-julia hljs">    CachingOptimizer(cache::MOI.ModelLike, mode::CachingOptimizerMode)</code></pre><p>Creates a <code>CachingOptimizer</code> in the <code>NO_OPTIMIZER</code> state and mode <code>mode</code>.</p><p>The type of the optimizer returned is <code>CachingOptimizer{MOI.AbstractOptimizer,typeof(cache)}</code> so it <em>does</em> support the function <code>reset_optimizer(::CachingOptimizer, new_optimizer)</code> if the type of <code>new_optimizer</code> is different from the type of <code>optimizer</code>.</p><p><strong>About the type</strong></p><p><strong>States</strong></p><p>A <code>CachingOptimizer</code> may be in one of three possible states (<code>CachingOptimizerState</code>):</p><ul><li><code>NO_OPTIMIZER</code>: The CachingOptimizer does not have any optimizer.</li><li><code>EMPTY_OPTIMIZER</code>: The CachingOptimizer has an empty optimizer. The optimizer is not synchronized with the cached model.</li><li><code>ATTACHED_OPTIMIZER</code>: The CachingOptimizer has an optimizer, and it is synchronized with the cached model.</li></ul><p><strong>Modes</strong></p><p>A <code>CachingOptimizer</code> has two modes of operation (<code>CachingOptimizerMode</code>):</p><ul><li><code>MANUAL</code>: The only methods that change the state of the <code>CachingOptimizer</code> are <a href="#MathOptInterface.Utilities.reset_optimizer"><code>Utilities.reset_optimizer</code></a>, <a href="#MathOptInterface.Utilities.drop_optimizer"><code>Utilities.drop_optimizer</code></a>, and <a href="#MathOptInterface.Utilities.attach_optimizer"><code>Utilities.attach_optimizer</code></a>. Attempting to perform an operation in the incorrect state results in an error.</li><li><code>AUTOMATIC</code>: The <code>CachingOptimizer</code> changes its state when necessary. For example, <code>optimize!</code> will automatically call <code>attach_optimizer</code> (an optimizer must have been previously set). Attempting to add a constraint or perform a modification not supported by the optimizer results in a drop to <code>EMPTY_OPTIMIZER</code> mode.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/cachingoptimizer.jl#L18-L76">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.attach_optimizer" href="#MathOptInterface.Utilities.attach_optimizer"><code>MathOptInterface.Utilities.attach_optimizer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">attach_optimizer(model::CachingOptimizer)</code></pre><p>Attaches the optimizer to <code>model</code>, copying all model data into it. Can be called only from the <code>EMPTY_OPTIMIZER</code> state. If the copy succeeds, the <code>CachingOptimizer</code> will be in state <code>ATTACHED_OPTIMIZER</code> after the call, otherwise an error is thrown; see <a href="../../../reference/models/#MathOptInterface.copy_to"><code>MOI.copy_to</code></a> for more details on which errors can be thrown.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/cachingoptimizer.jl#L208-L216">source</a></section><section><div><pre><code class="language-julia hljs">MOIU.attach_optimizer(model::GenericModel)</code></pre><p>Call <code>MOIU.attach_optimizer</code> on the backend of <code>model</code>.</p><p>Cannot be called in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L313-L319">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.reset_optimizer" href="#MathOptInterface.Utilities.reset_optimizer"><code>MathOptInterface.Utilities.reset_optimizer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">reset_optimizer(m::CachingOptimizer, optimizer::MOI.AbstractOptimizer)</code></pre><p>Sets or resets <code>m</code> to have the given empty optimizer <code>optimizer</code>.</p><p>Can be called from any state. An assertion error will be thrown if <code>optimizer</code> is not empty.</p><p>The <code>CachingOptimizer</code> <code>m</code> will be in state <code>EMPTY_OPTIMIZER</code> after the call.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/cachingoptimizer.jl#L163-L172">source</a></section><section><div><pre><code class="language-julia hljs">reset_optimizer(m::CachingOptimizer)</code></pre><p>Detaches and empties the current optimizer. Can be called from <code>ATTACHED_OPTIMIZER</code> or <code>EMPTY_OPTIMIZER</code> state. The <code>CachingOptimizer</code> will be in state <code>EMPTY_OPTIMIZER</code> after the call.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/cachingoptimizer.jl#L181-L187">source</a></section><section><div><pre><code class="language-julia hljs">MOIU.reset_optimizer(model::GenericModel, optimizer::MOI.AbstractOptimizer)</code></pre><p>Call <code>MOIU.reset_optimizer</code> on the backend of <code>model</code>.</p><p>Cannot be called in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L268-L274">source</a></section><section><div><pre><code class="language-julia hljs">MOIU.reset_optimizer(model::GenericModel)</code></pre><p>Call <code>MOIU.reset_optimizer</code> on the backend of <code>model</code>.</p><p>Cannot be called in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L285-L291">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.drop_optimizer" href="#MathOptInterface.Utilities.drop_optimizer"><code>MathOptInterface.Utilities.drop_optimizer</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">drop_optimizer(m::CachingOptimizer)</code></pre><p>Drops the optimizer, if one is present. Can be called from any state. The <code>CachingOptimizer</code> will be in state <code>NO_OPTIMIZER</code> after the call.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/cachingoptimizer.jl#L196-L201">source</a></section><section><div><pre><code class="language-julia hljs">MOIU.drop_optimizer(model::GenericModel)</code></pre><p>Call <code>MOIU.drop_optimizer</code> on the backend of <code>model</code>.</p><p>Cannot be called in direct mode.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/JuMP.jl/blob/a2e6a0212d7f46a88949d2d5488d41f51734ea3c/src/optimizer_interface.jl#L300-L306">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.state" href="#MathOptInterface.Utilities.state"><code>MathOptInterface.Utilities.state</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">state(m::CachingOptimizer)::CachingOptimizerState</code></pre><p>Returns the state of the CachingOptimizer <code>m</code>. See <a href="#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/cachingoptimizer.jl#L149-L153">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.mode" href="#MathOptInterface.Utilities.mode"><code>MathOptInterface.Utilities.mode</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">mode(m::CachingOptimizer)::CachingOptimizerMode</code></pre><p>Returns the operating mode of the CachingOptimizer <code>m</code>. See <a href="#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/cachingoptimizer.jl#L156-L160">source</a></section></article><h2 id="Mock-optimizer"><a class="docs-heading-anchor" href="#Mock-optimizer">Mock optimizer</a><a id="Mock-optimizer-1"></a><a class="docs-heading-anchor-permalink" href="#Mock-optimizer" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.MockOptimizer" href="#MathOptInterface.Utilities.MockOptimizer"><code>MathOptInterface.Utilities.MockOptimizer</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">MockOptimizer</code></pre><p><code>MockOptimizer</code> is a fake optimizer especially useful for testing. Its main feature is that it can store the values that should be returned for each attribute.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/mockoptimizer.jl#L17-L23">source</a></section></article><h2 id="Printing"><a class="docs-heading-anchor" href="#Printing">Printing</a><a id="Printing-1"></a><a class="docs-heading-anchor-permalink" href="#Printing" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.latex_formulation" href="#MathOptInterface.Utilities.latex_formulation"><code>MathOptInterface.Utilities.latex_formulation</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">latex_formulation(model::MOI.ModelLike; kwargs...)</code></pre><p>Wrap <code>model</code> in a type so that it can be pretty-printed as <code>text/latex</code> in a notebook like IJulia, or in Documenter.</p><p>To render the model, end the cell with <code>latex_formulation(model)</code>, or call <code>display(latex_formulation(model))</code> in to force the display of the model from inside a function.</p><p>Possible keyword arguments are:</p><ul><li><code>simplify_coefficients</code> : Simplify coefficients if possible by omitting them or removing trailing zeros.</li><li><code>default_name</code> : The name given to variables with an empty name.</li><li><code>print_types</code> : Print the MOI type of each function and set for clarity.</li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/print.jl#L632-L648">source</a></section></article><h2 id="Copy-utilities"><a class="docs-heading-anchor" href="#Copy-utilities">Copy utilities</a><a id="Copy-utilities-1"></a><a class="docs-heading-anchor-permalink" href="#Copy-utilities" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.default_copy_to" href="#MathOptInterface.Utilities.default_copy_to"><code>MathOptInterface.Utilities.default_copy_to</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">default_copy_to(dest::MOI.ModelLike, src::MOI.ModelLike)</code></pre><p>A default implementation of <code>MOI.copy_to(dest, src)</code> for models that implement the incremental interface, i.e., <a href="../../../reference/models/#MathOptInterface.supports_incremental_interface"><code>MOI.supports_incremental_interface</code></a> returns <code>true</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/copy.jl#L471-L477">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.IndexMap" href="#MathOptInterface.Utilities.IndexMap"><code>MathOptInterface.Utilities.IndexMap</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndexMap()</code></pre><p>The dictionary-like object returned by <a href="../../../reference/models/#MathOptInterface.copy_to"><code>MOI.copy_to</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/copy/index_map.jl#L25-L29">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.identity_index_map" href="#MathOptInterface.Utilities.identity_index_map"><code>MathOptInterface.Utilities.identity_index_map</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">identity_index_map(model::MOI.ModelLike)</code></pre><p>Return an <a href="#MathOptInterface.Utilities.IndexMap"><code>IndexMap</code></a> that maps all variable and constraint indices of <code>model</code> to themselves.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/copy/index_map.jl#L46-L51">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.ModelFilter" href="#MathOptInterface.Utilities.ModelFilter"><code>MathOptInterface.Utilities.ModelFilter</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ModelFilter(filter::Function, model::MOI.ModelLike)</code></pre><p>A layer to filter out various components of <code>model</code>.</p><p>The filter function takes a single argument, which is each element from the list returned by the attributes below. It returns <code>true</code> if the element should be visible in the filtered model and <code>false</code> otherwise.</p><p>The components that are filtered are:</p><ul><li>Entire constraint types via:<ul><li><code>MOI.ListOfConstraintTypesPresent</code></li></ul></li><li>Individual constraints via:<ul><li><code>MOI.ListOfConstraintIndices{F,S}</code></li></ul></li><li>Specific attributes via:<ul><li><code>MOI.ListOfModelAttributesSet</code></li><li><code>MOI.ListOfConstraintAttributesSet</code></li><li><code>MOI.ListOfVariableAttributesSet</code></li></ul></li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The list of attributes filtered may change in a future release. You should write functions that are generic and not limited to the five types listed above. Thus, you should probably define a fallback <code>filter(::Any) = true</code>.</p></div></div><p>See below for examples of how this works.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>This layer has a limited scope. It is intended by be used in conjunction with <code>MOI.copy_to</code>.</p></div></div><p><strong>Example: copy model excluding integer constraints</strong></p><p>Use the <code>do</code> syntax to provide a single function.</p><pre><code class="language-julia hljs">filtered_src = MOI.Utilities.ModelFilter(src) do item
+    return item != (MOI.VariableIndex, MOI.Integer)
+end
+MOI.copy_to(dest, filtered_src)</code></pre><p><strong>Example: copy model excluding names</strong></p><p>Use type dispatch to simplify the implementation:</p><pre><code class="language-julia hljs">my_filter(::Any) = true  # Note the generic fallback!
+my_filter(::MOI.VariableName) = false
+my_filter(::MOI.ConstraintName) = false
+filtered_src = MOI.Utilities.ModelFilter(my_filter, src)
+MOI.copy_to(dest, filtered_src)</code></pre><p><strong>Example: copy irreducible infeasible subsystem</strong></p><pre><code class="language-julia hljs">my_filter(::Any) = true  # Note the generic fallback!
+function my_filter(ci::MOI.ConstraintIndex)
+    status = MOI.get(dest, MOI.ConstraintConflictStatus(), ci)
+    return status != MOI.NOT_IN_CONFLICT
+end
+filtered_src = MOI.Utilities.ModelFilter(my_filter, src)
+MOI.copy_to(dest, filtered_src)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/copy.jl#L512-L577">source</a></section></article><h2 id="Penalty-relaxation"><a class="docs-heading-anchor" href="#Penalty-relaxation">Penalty relaxation</a><a id="Penalty-relaxation-1"></a><a class="docs-heading-anchor-permalink" href="#Penalty-relaxation" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.PenaltyRelaxation" href="#MathOptInterface.Utilities.PenaltyRelaxation"><code>MathOptInterface.Utilities.PenaltyRelaxation</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">PenaltyRelaxation(
+    penalties = Dict{MOI.ConstraintIndex,Float64}();
+    default::Union{Nothing,T} = 1.0,
+)</code></pre><p>A problem modifier that, when passed to <a href="../../../reference/modification/#MathOptInterface.modify"><code>MOI.modify</code></a>, destructively modifies the model in-place to create a penalized relaxation of the constraints.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This is a destructive routine that modifies the model in-place. If you don&#39;t want to modify the original model, use <code>JuMP.copy_model</code> to create a copy before calling <a href="../../../reference/modification/#MathOptInterface.modify"><code>MOI.modify</code></a>.</p></div></div><p><strong>Reformulation</strong></p><p>See <a href="#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>Utilities.ScalarPenaltyRelaxation</code></a> for details of the reformulation.</p><p>For each constraint <code>ci</code>, the penalty passed to <a href="#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>Utilities.ScalarPenaltyRelaxation</code></a> is <code>get(penalties, ci, default)</code>. If the value is <code>nothing</code>, because <code>ci</code> does not exist in <code>penalties</code> and <code>default = nothing</code>, then the constraint is skipped.</p><p><strong>Return value</strong></p><p><code>MOI.modify(model, PenaltyRelaxation())</code> returns a <code>Dict{MOI.ConstraintIndex,MOI.ScalarAffineFunction}</code> that maps each constraint index to the corresponding <code>y + z</code> as a <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a>. In an optimal solution, query the value of these functions to compute the violation of each constraint.</p><p><strong>Relax a subset of constraints</strong></p><p>To relax a subset of constraints, pass a <code>penalties</code> dictionary and set <code>default = nothing</code>.</p><p><strong>Supported constraint types</strong></p><p>The penalty relaxation is currently limited to modifying <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a> and <a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction</code></a> constraints in the linear sets <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a>, <a href="../../../reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo</code></a> and <a href="../../../reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval</code></a>.</p><p>It does not include variable bound or integrality constraints, because these cannot be modified in-place.</p><p>To modify variable bounds, rewrite them as linear constraints.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}();
+
+julia&gt; x = MOI.add_variable(model);
+
+julia&gt; c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));
+
+julia&gt; map = MOI.modify(model, MOI.Utilities.PenaltyRelaxation(default = 2.0));
+
+julia&gt; print(model)
+Minimize ScalarAffineFunction{Float64}:
+ 0.0 + 2.0 v[2]
+
+Subject to:
+
+ScalarAffineFunction{Float64}-in-LessThan{Float64}
+ 0.0 + 1.0 v[1] - 1.0 v[2] &lt;= 2.0
+
+VariableIndex-in-GreaterThan{Float64}
+ v[2] &gt;= 0.0
+
+julia&gt; map[c] isa MOI.ScalarAffineFunction{Float64}
+true</code></pre><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}();
+
+julia&gt; x = MOI.add_variable(model);
+
+julia&gt; c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));
+
+julia&gt; map = MOI.modify(model, MOI.Utilities.PenaltyRelaxation(Dict(c =&gt; 3.0)));
+
+julia&gt; print(model)
+Minimize ScalarAffineFunction{Float64}:
+ 0.0 + 3.0 v[2]
+
+Subject to:
+
+ScalarAffineFunction{Float64}-in-LessThan{Float64}
+ 0.0 + 1.0 v[1] - 1.0 v[2] &lt;= 2.0
+
+VariableIndex-in-GreaterThan{Float64}
+ v[2] &gt;= 0.0
+
+julia&gt; map[c] isa MOI.ScalarAffineFunction{Float64}
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/penalty_relaxation.jl#L141-L241">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.ScalarPenaltyRelaxation" href="#MathOptInterface.Utilities.ScalarPenaltyRelaxation"><code>MathOptInterface.Utilities.ScalarPenaltyRelaxation</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ScalarPenaltyRelaxation(penalty::T) where {T}</code></pre><p>A problem modifier that, when passed to <a href="../../../reference/modification/#MathOptInterface.modify"><code>MOI.modify</code></a>, destructively modifies the constraint in-place to create a penalized relaxation of the constraint.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This is a destructive routine that modifies the constraint in-place. If you don&#39;t want to modify the original model, use <code>JuMP.copy_model</code> to create a copy before calling <a href="../../../reference/modification/#MathOptInterface.modify"><code>MOI.modify</code></a>.</p></div></div><p><strong>Reformulation</strong></p><p>The penalty relaxation modifies constraints of the form <span>$f(x) \in S$</span> into <span>$f(x) + y - z \in S$</span>, where <span>$y, z \ge 0$</span>, and then it introduces a penalty term into the objective of <span>$a \times (y + z)$</span> (if minimizing, else <span>$-a$</span>), where <span>$a$</span> is <code>penalty</code></p><p>When <code>S</code> is <a href="../../../reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a>, we omit <code>y</code> or <code>z</code> respectively as a performance optimization.</p><p><strong>Return value</strong></p><p><code>MOI.modify(model, ci, ScalarPenaltyRelaxation(penalty))</code> returns <code>y + z</code> as a <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a>. In an optimal solution, query the value of this function to compute the violation of the constraint.</p><p><strong>Examples</strong></p><pre><code class="language-julia-repl hljs">julia&gt; model = MOI.Utilities.Model{Float64}();
+
+julia&gt; x = MOI.add_variable(model);
+
+julia&gt; c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));
+
+julia&gt; f = MOI.modify(model, c, MOI.Utilities.ScalarPenaltyRelaxation(2.0));
+
+julia&gt; print(model)
+Minimize ScalarAffineFunction{Float64}:
+ 0.0 + 2.0 v[2]
+
+Subject to:
+
+ScalarAffineFunction{Float64}-in-LessThan{Float64}
+ 0.0 + 1.0 v[1] - 1.0 v[2] &lt;= 2.0
+
+VariableIndex-in-GreaterThan{Float64}
+ v[2] &gt;= 0.0
+
+julia&gt; f isa MOI.ScalarAffineFunction{Float64}
+true</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/penalty_relaxation.jl#L7-L61">source</a></section></article><h2 id="MatrixOfConstraints"><a class="docs-heading-anchor" href="#MatrixOfConstraints">MatrixOfConstraints</a><a id="MatrixOfConstraints-1"></a><a class="docs-heading-anchor-permalink" href="#MatrixOfConstraints" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.MatrixOfConstraints" href="#MathOptInterface.Utilities.MatrixOfConstraints"><code>MathOptInterface.Utilities.MatrixOfConstraints</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct MatrixOfConstraints{T,AT,BT,ST} &lt;: MOI.ModelLike
+    coefficients::AT
+    constants::BT
+    sets::ST
+    caches::Vector{Any}
+    are_indices_mapped::Vector{BitSet}
+    final_touch::Bool
+end</code></pre><p>Represent <code>ScalarAffineFunction</code> and <code>VectorAffinefunction</code> constraints in a matrix form where the linear coefficients of the functions are stored in the <code>coefficients</code> field, the constants of the functions or sets are stored in the <code>constants</code> field. Additional information about the sets are stored in the <code>sets</code> field.</p><p>This model can only be used as the <code>constraints</code> field of a <code>MOI.Utilities.AbstractModel</code>.</p><p>When the constraints are added, they are stored in the <code>caches</code> field. They are only loaded in the <code>coefficients</code> and <code>constants</code> fields once <code>MOI.Utilities.final_touch</code> is called. For this reason, <code>MatrixOfConstraints</code> should not be used by an incremental interface. Use <code>MOI.copy_to</code> instead.</p><p>The constraints can be added in two different ways:</p><ol><li>With <code>add_constraint</code>, in which case a canonicalized copy of the function is stored in <code>caches</code>.</li><li>With <code>pass_nonvariable_constraints</code>, in which case the functions and sets are stored themselves in <code>caches</code> without mapping the variable indices. The corresponding index in <code>caches</code> is added in <code>are_indices_mapped</code>. This avoids doing a copy of the function in case the getter of <code>CanonicalConstraintFunction</code> does not make a copy for the source model, e.g., this is the case of <code>VectorOfConstraints</code>.</li></ol><p>We illustrate this with an example. Suppose a model is copied from a <code>src::MOI.Utilities.Model</code> to a bridged model with a <code>MatrixOfConstraints</code>. For all the types that are not bridged, the constraints will be copied with <code>pass_nonvariable_constraints</code>. Hence the functions stored in <code>caches</code> are exactly the same as the ones stored in <code>src</code>. This is ok since this is only during the <code>copy_to</code> operation during which <code>src</code> cannot be modified. On the other hand, for the types that are bridged, the functions added may contain duplicates even if the functions did not contain duplicates in <code>src</code> so duplicates are removed with <code>MOI.Utilities.canonical</code>.</p><p><strong>Interface</strong></p><p>The <code>.coefficients::AT</code> type must implement:</p><ul><li><code>AT()</code></li><li><code>MOI.empty(::AT)!</code></li><li><a href="#MathOptInterface.Utilities.add_column"><code>MOI.Utilities.add_column</code></a></li><li><a href="#MathOptInterface.Utilities.set_number_of_rows"><code>MOI.Utilities.set_number_of_rows</code></a></li><li><a href="#MathOptInterface.Utilities.allocate_terms"><code>MOI.Utilities.allocate_terms</code></a></li><li><a href="#MathOptInterface.Utilities.load_terms"><code>MOI.Utilities.load_terms</code></a></li><li><a href="#MathOptInterface.Utilities.final_touch"><code>MOI.Utilities.final_touch</code></a></li></ul><p>The <code>.constants::BT</code> type must implement:</p><ul><li><code>BT()</code></li><li><code>Base.empty!(::BT)</code></li><li><code>Base.resize(::BT)</code></li><li><a href="#MathOptInterface.Utilities.load_constants"><code>MOI.Utilities.load_constants</code></a></li><li><a href="#MathOptInterface.Utilities.function_constants"><code>MOI.Utilities.function_constants</code></a></li><li><a href="#MathOptInterface.Utilities.set_from_constants"><code>MOI.Utilities.set_from_constants</code></a></li></ul><p>The <code>.sets::ST</code> type must implement:</p><ul><li><code>ST()</code></li><li><code>MOI.is_empty(::ST)</code></li><li><code>MOI.empty(::ST)</code></li><li><code>MOI.dimension(::ST)</code></li><li><code>MOI.is_valid(::ST, ::MOI.ConstraintIndex)</code></li><li><code>MOI.get(::ST, ::MOI.ListOfConstraintTypesPresent)</code></li><li><code>MOI.get(::ST, ::MOI.NumberOfConstraints)</code></li><li><code>MOI.get(::ST, ::MOI.ListOfConstraintIndices)</code></li><li><a href="#MathOptInterface.Utilities.set_types"><code>MOI.Utilities.set_types</code></a></li><li><a href="#MathOptInterface.Utilities.set_index"><code>MOI.Utilities.set_index</code></a></li><li><a href="#MathOptInterface.Utilities.add_set"><code>MOI.Utilities.add_set</code></a></li><li><a href="#MathOptInterface.Utilities.rows"><code>MOI.Utilities.rows</code></a></li><li><a href="#MathOptInterface.Utilities.final_touch"><code>MOI.Utilities.final_touch</code></a></li></ul></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L7-L87">source</a></section></article><h3 id=".coefficients"><a class="docs-heading-anchor" href="#.coefficients"><code>.coefficients</code></a><a id=".coefficients-1"></a><a class="docs-heading-anchor-permalink" href="#.coefficients" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.add_column" href="#MathOptInterface.Utilities.add_column"><code>MathOptInterface.Utilities.add_column</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_column(coefficients)::Nothing</code></pre><p>Tell <code>coefficients</code> to pre-allocate datastructures as needed to store one column.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L106-L111">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.allocate_terms" href="#MathOptInterface.Utilities.allocate_terms"><code>MathOptInterface.Utilities.allocate_terms</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">allocate_terms(coefficients, index_map, func)::Nothing</code></pre><p>Tell <code>coefficients</code> that the terms of the function <code>func</code> where the variable indices are mapped with <code>index_map</code> will be loaded with <a href="#MathOptInterface.Utilities.load_terms"><code>load_terms</code></a>.</p><p>The function <code>func</code> must be canonicalized before calling <code>allocate_terms</code>. See <a href="#MathOptInterface.Utilities.is_canonical"><code>is_canonical</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L121-L129">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.set_number_of_rows" href="#MathOptInterface.Utilities.set_number_of_rows"><code>MathOptInterface.Utilities.set_number_of_rows</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_number_of_rows(coefficients, n)::Nothing</code></pre><p>Tell <code>coefficients</code> to pre-allocate datastructures as needed to store <code>n</code> rows.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L114-L118">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.load_terms" href="#MathOptInterface.Utilities.load_terms"><code>MathOptInterface.Utilities.load_terms</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">load_terms(coefficients, index_map, func, offset)::Nothing</code></pre><p>Loads the terms of <code>func</code> to <code>coefficients</code>, mapping the variable indices with <code>index_map</code>.</p><p>The <code>i</code>th dimension of <code>func</code> is loaded at the <code>(offset + i)</code>th row of <code>coefficients</code>.</p><p>The function must be allocated first with <a href="#MathOptInterface.Utilities.allocate_terms"><code>allocate_terms</code></a>.</p><p>The function <code>func</code> must be canonicalized, see <a href="#MathOptInterface.Utilities.is_canonical"><code>is_canonical</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L132-L144">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.final_touch" href="#MathOptInterface.Utilities.final_touch"><code>MathOptInterface.Utilities.final_touch</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">final_touch(coefficients)::Nothing</code></pre><p>Informs the <code>coefficients</code> that all functions have been added with <code>load_terms</code>. No more modification is allowed unless <code>MOI.empty!</code> is called.</p><pre><code class="nohighlight hljs">final_touch(sets)::Nothing</code></pre><p>Informs the <code>sets</code> that all functions have been added with <code>add_set</code>. No more modification is allowed unless <code>MOI.empty!</code> is called.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L147-L157">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.extract_function" href="#MathOptInterface.Utilities.extract_function"><code>MathOptInterface.Utilities.extract_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">extract_function(coefficients, row::Integer, constant::T) where {T}</code></pre><p>Return the <code>MOI.ScalarAffineFunction{T}</code> function corresponding to row <code>row</code> in <code>coefficients</code>.</p><pre><code class="nohighlight hljs">extract_function(
+    coefficients,
+    rows::UnitRange,
+    constants::Vector{T},
+) where{T}</code></pre><p>Return the <code>MOI.VectorAffineFunction{T}</code> function corresponding to rows <code>rows</code> in <code>coefficients</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L160-L174">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.MutableSparseMatrixCSC" href="#MathOptInterface.Utilities.MutableSparseMatrixCSC"><code>MathOptInterface.Utilities.MutableSparseMatrixCSC</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">mutable struct MutableSparseMatrixCSC{Tv,Ti&lt;:Integer,I&lt;:AbstractIndexing}
+    indexing::I
+    m::Int
+    n::Int
+    colptr::Vector{Ti}
+    rowval::Vector{Ti}
+    nzval::Vector{Tv}
+    nz_added::Vector{Ti}
+end</code></pre><p>Matrix type loading sparse matrices in the Compressed Sparse Column format. The indexing used is <code>indexing</code>, see <a href="#MathOptInterface.Utilities.AbstractIndexing"><code>AbstractIndexing</code></a>. The other fields have the same meaning than for <code>SparseArrays.SparseMatrixCSC</code> except that the indexing is different unless <code>indexing</code> is <code>OneBasedIndexing</code>. In addition, <code>nz_added</code> is used to cache the number of non-zero terms that have been added to each column due to the incremental nature of <code>load_terms</code>.</p><p>The matrix is loaded in 5 steps:</p><ol><li><code>MOI.empty!</code> is called.</li><li><code>MOI.Utilities.add_column</code> and <code>MOI.Utilities.allocate_terms</code> are called in any order.</li><li><code>MOI.Utilities.set_number_of_rows</code> is called.</li><li><code>MOI.Utilities.load_terms</code> is called for each affine function.</li><li><code>MOI.Utilities.final_touch</code> is called.</li></ol></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sparse_matrix.jl#L44-L69">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.AbstractIndexing" href="#MathOptInterface.Utilities.AbstractIndexing"><code>MathOptInterface.Utilities.AbstractIndexing</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractIndexing end</code></pre><p>Indexing to be used for storing the row and column indices of <code>MutableSparseMatrixCSC</code>. See <a href="#MathOptInterface.Utilities.ZeroBasedIndexing"><code>ZeroBasedIndexing</code></a> and <a href="#MathOptInterface.Utilities.OneBasedIndexing"><code>OneBasedIndexing</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sparse_matrix.jl#L9-L15">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.ZeroBasedIndexing" href="#MathOptInterface.Utilities.ZeroBasedIndexing"><code>MathOptInterface.Utilities.ZeroBasedIndexing</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ZeroBasedIndexing &lt;: AbstractIndexing end</code></pre><p>Zero-based indexing: the <code>i</code>th row or column has index <code>i - 1</code>. This is useful when the vectors of row and column indices need to be communicated to a library using zero-based indexing such as C libraries.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sparse_matrix.jl#L18-L24">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.OneBasedIndexing" href="#MathOptInterface.Utilities.OneBasedIndexing"><code>MathOptInterface.Utilities.OneBasedIndexing</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct ZeroBasedIndexing &lt;: AbstractIndexing end</code></pre><p>One-based indexing: the <code>i</code>th row or column has index <code>i</code>. This enables an allocation-free conversion of <a href="#MathOptInterface.Utilities.MutableSparseMatrixCSC"><code>MutableSparseMatrixCSC</code></a> to <code>SparseArrays.SparseMatrixCSC</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sparse_matrix.jl#L27-L33">source</a></section></article><h3 id=".constants"><a class="docs-heading-anchor" href="#.constants"><code>.constants</code></a><a id=".constants-1"></a><a class="docs-heading-anchor-permalink" href="#.constants" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.load_constants" href="#MathOptInterface.Utilities.load_constants"><code>MathOptInterface.Utilities.load_constants</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">load_constants(constants, offset, func_or_set)::Nothing</code></pre><p>This function loads the constants of <code>func_or_set</code> in <code>constants</code> at an offset of <code>offset</code>. Where <code>offset</code> is the sum of the dimensions of the constraints already loaded. The storage should be preallocated with <code>resize!</code> before calling this function.</p><p>This function should be implemented to be usable as storage of constants for <a href="#MathOptInterface.Utilities.MatrixOfConstraints"><code>MatrixOfConstraints</code></a>.</p><p>The constants are loaded in three steps:</p><ol><li><code>Base.empty!</code> is called.</li><li><code>Base.resize!</code> is called with the sum of the dimensions of all constraints.</li><li><code>MOI.Utilities.load_constants</code> is called for each function for vector constraint or set for scalar constraint.</li></ol></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L181-L197">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.function_constants" href="#MathOptInterface.Utilities.function_constants"><code>MathOptInterface.Utilities.function_constants</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">function_constants(constants, rows)</code></pre><p>This function returns the function constants that were loaded with <a href="#MathOptInterface.Utilities.load_constants"><code>load_constants</code></a> at the rows <code>rows</code>.</p><p>This function should be implemented to be usable as storage of constants for <a href="#MathOptInterface.Utilities.MatrixOfConstraints"><code>MatrixOfConstraints</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L200-L208">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.set_from_constants" href="#MathOptInterface.Utilities.set_from_constants"><code>MathOptInterface.Utilities.set_from_constants</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_from_constants(constants, S::Type, rows)::S</code></pre><p>This function returns an instance of the set <code>S</code> for which the constants where loaded with <a href="#MathOptInterface.Utilities.load_constants"><code>load_constants</code></a> at the rows <code>rows</code>.</p><p>This function should be implemented to be usable as storage of constants for <a href="#MathOptInterface.Utilities.MatrixOfConstraints"><code>MatrixOfConstraints</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L211-L219">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.Hyperrectangle" href="#MathOptInterface.Utilities.Hyperrectangle"><code>MathOptInterface.Utilities.Hyperrectangle</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">struct Hyperrectangle{T} &lt;: AbstractVectorBounds
+    lower::Vector{T}
+    upper::Vector{T}
+end</code></pre><p>A struct for the .constants field in MatrixOfConstraints.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/variables_container.jl#L453-L460">source</a></section></article><h3 id=".sets"><a class="docs-heading-anchor" href="#.sets"><code>.sets</code></a><a id=".sets-1"></a><a class="docs-heading-anchor-permalink" href="#.sets" title="Permalink"></a></h3><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.set_index" href="#MathOptInterface.Utilities.set_index"><code>MathOptInterface.Utilities.set_index</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_index(sets, ::Type{S})::Union{Int,Nothing} where {S&lt;:MOI.AbstractSet}</code></pre><p>Return an integer corresponding to the index of the set type in the list given by <a href="#MathOptInterface.Utilities.set_types"><code>set_types</code></a>.</p><p>If <code>S</code> is not part of the list, return <code>nothing</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L238-L245">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.set_types" href="#MathOptInterface.Utilities.set_types"><code>MathOptInterface.Utilities.set_types</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_types(sets)::Vector{Type}</code></pre><p>Return the list of the types of the sets allowed in <code>sets</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L226-L230">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.add_set" href="#MathOptInterface.Utilities.add_set"><code>MathOptInterface.Utilities.add_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">add_set(sets, i)::Int64</code></pre><p>Add a scalar set of type index <code>i</code>.</p><pre><code class="nohighlight hljs">add_set(sets, i, dim)::Int64</code></pre><p>Add a vector set of type index <code>i</code> and dimension <code>dim</code>.</p><p>Both methods return a unique <code>Int64</code> of the set that can be used to reference this set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L248-L259">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.rows" href="#MathOptInterface.Utilities.rows"><code>MathOptInterface.Utilities.rows</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">rows(sets, ci::MOI.ConstraintIndex)::Union{Int,UnitRange{Int}}</code></pre><p>Return the rows in <code>1:MOI.dimension(sets)</code> corresponding to the set of id <code>ci.value</code>.</p><p>For scalar sets, this returns an <code>Int</code>. For vector sets, this returns an <code>UnitRange{Int}</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L262-L270">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.num_rows" href="#MathOptInterface.Utilities.num_rows"><code>MathOptInterface.Utilities.num_rows</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">num_rows(sets::OrderedProductOfSets, ::Type{S}) where {S}</code></pre><p>Return the number of rows corresponding to a set of type <code>S</code>. That is, it is the sum of the dimensions of the sets of type <code>S</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/product_of_sets.jl#L263-L268">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.set_with_dimension" href="#MathOptInterface.Utilities.set_with_dimension"><code>MathOptInterface.Utilities.set_with_dimension</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_with_dimension(::Type{S}, dim) where {S&lt;:MOI.AbstractVectorSet}</code></pre><p>Returns the instance of <code>S</code> of <a href="../../../reference/standard_form/#MathOptInterface.dimension"><code>MOI.dimension</code></a> <code>dim</code>. This needs to be implemented for sets of type <code>S</code> to be useable with <a href="#MathOptInterface.Utilities.MatrixOfConstraints"><code>MatrixOfConstraints</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/matrix_of_constraints.jl#L558-L564">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.ProductOfSets" href="#MathOptInterface.Utilities.ProductOfSets"><code>MathOptInterface.Utilities.ProductOfSets</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type ProductOfSets{T} end</code></pre><p>Represents a cartesian product of sets of given types.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/product_of_sets.jl#L7-L11">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.MixOfScalarSets" href="#MathOptInterface.Utilities.MixOfScalarSets"><code>MathOptInterface.Utilities.MixOfScalarSets</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type MixOfScalarSets{T} &lt;: ProductOfSets{T} end</code></pre><p>Product of scalar sets in the order the constraints are added, mixing the constraints of different types.</p><p>Use <a href="#MathOptInterface.Utilities.@mix_of_scalar_sets"><code>@mix_of_scalar_sets</code></a> to generate a new subtype.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/product_of_sets.jl#L41-L48">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.@mix_of_scalar_sets" href="#MathOptInterface.Utilities.@mix_of_scalar_sets"><code>MathOptInterface.Utilities.@mix_of_scalar_sets</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@mix_of_scalar_sets(name, set_types...)</code></pre><p>Generate a new <a href="#MathOptInterface.Utilities.MixOfScalarSets"><code>MixOfScalarSets</code></a> subtype.</p><p><strong>Example</strong></p><pre><code class="language-julia hljs">@mix_of_scalar_sets(
+    MixedIntegerLinearProgramSets,
+    MOI.GreaterThan{T},
+    MOI.LessThan{T},
+    MOI.EqualTo{T},
+    MOI.Integer,
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/product_of_sets.jl#L51-L67">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.OrderedProductOfSets" href="#MathOptInterface.Utilities.OrderedProductOfSets"><code>MathOptInterface.Utilities.OrderedProductOfSets</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type OrderedProductOfSets{T} &lt;: ProductOfSets{T} end</code></pre><p>Product of sets in the order the constraints are added, grouping the constraints of the same types contiguously.</p><p>Use <a href="#MathOptInterface.Utilities.@product_of_sets"><code>@product_of_sets</code></a> to generate new subtypes.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/product_of_sets.jl#L139-L146">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.@product_of_sets" href="#MathOptInterface.Utilities.@product_of_sets"><code>MathOptInterface.Utilities.@product_of_sets</code></a> — <span class="docstring-category">Macro</span></header><section><div><pre><code class="language-julia hljs">@product_of_sets(name, set_types...)</code></pre><p>Generate a new <a href="#MathOptInterface.Utilities.OrderedProductOfSets"><code>OrderedProductOfSets</code></a> subtype.</p><p><strong>Example</strong></p><pre><code class="language-julia hljs">@product_of_sets(
+    LinearOrthants,
+    MOI.Zeros,
+    MOI.Nonnegatives,
+    MOI.Nonpositives,
+    MOI.ZeroOne,
+)</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/product_of_sets.jl#L149-L165">source</a></section></article><h2 id="Fallbacks"><a class="docs-heading-anchor" href="#Fallbacks">Fallbacks</a><a id="Fallbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Fallbacks" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.get_fallback" href="#MathOptInterface.Utilities.get_fallback"><code>MathOptInterface.Utilities.get_fallback</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">get_fallback(model::MOI.ModelLike, ::MOI.ObjectiveValue)</code></pre><p>Compute the objective function value using the <code>VariablePrimal</code> results and the <code>ObjectiveFunction</code> value.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/results.jl#L19-L24">source</a></section><section><div><pre><code class="language-julia hljs">get_fallback(model::MOI.ModelLike, ::MOI.DualObjectiveValue, T::Type)::T</code></pre><p>Compute the dual objective value of type <code>T</code> using the <code>ConstraintDual</code> results and the <code>ConstraintFunction</code> and <code>ConstraintSet</code> values. Note that the nonlinear part of the model is ignored.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/results.jl#L146-L152">source</a></section><section><div><pre><code class="language-julia hljs">get_fallback(model::MOI.ModelLike, ::MOI.ConstraintPrimal,
+             constraint_index::MOI.ConstraintIndex)</code></pre><p>Compute the value of the function of the constraint of index <code>constraint_index</code> using the <code>VariablePrimal</code> results and the <code>ConstraintFunction</code> values.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/results.jl#L175-L181">source</a></section><section><div><pre><code class="language-julia hljs">get_fallback(model::MOI.ModelLike, attr::MOI.ConstraintDual,
+             ci::MOI.ConstraintIndex{Union{MOI.VariableIndex,
+                                           MOI.VectorOfVariables}})</code></pre><p>Compute the dual of the constraint of index <code>ci</code> using the <code>ConstraintDual</code> of other constraints and the <code>ConstraintFunction</code> values. Throws an error if some constraints are quadratic or if there is one another <code>MOI.VariableIndex</code>-in-<code>S</code> or <code>MOI.VectorOfVariables</code>-in-<code>S</code> constraint with one of the variables in the function of the constraint <code>ci</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/results.jl#L471-L481">source</a></section></article><h2 id="Function-utilities"><a class="docs-heading-anchor" href="#Function-utilities">Function utilities</a><a id="Function-utilities-1"></a><a class="docs-heading-anchor-permalink" href="#Function-utilities" title="Permalink"></a></h2><p>The following utilities are available for functions:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.eval_variables" href="#MathOptInterface.Utilities.eval_variables"><code>MathOptInterface.Utilities.eval_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eval_variables(value_fn::Function, f::MOI.AbstractFunction)</code></pre><p>Returns the value of function <code>f</code> if each variable index <code>vi</code> is evaluated as <code>value_fn(vi)</code>.</p><p>Note that <code>value_fn</code> must return a Number. See <a href="#MathOptInterface.Utilities.substitute_variables"><code>substitute_variables</code></a> for a similar function where <code>value_fn</code> returns an <a href="../../../reference/standard_form/#MathOptInterface.AbstractScalarFunction"><code>MOI.AbstractScalarFunction</code></a>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>The two-argument version of <code>eval_variables</code> is deprecated and may be removed in MOI v2.0.0. Use the three-argument method <code>eval_variables(::Function, ::MOI.ModelLike, ::MOI.AbstractFunction)</code> instead.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L104-L119">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.map_indices" href="#MathOptInterface.Utilities.map_indices"><code>MathOptInterface.Utilities.map_indices</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">map_indices(index_map::Function, attr::MOI.AnyAttribute, x::X)::X where {X}</code></pre><p>Substitute any <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> (resp. <a href="../../../reference/constraints/#MathOptInterface.ConstraintIndex"><code>MOI.ConstraintIndex</code></a>) in <code>x</code> by the <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> (resp. <a href="../../../reference/constraints/#MathOptInterface.ConstraintIndex"><code>MOI.ConstraintIndex</code></a>) of the same type given by <code>index_map(x)</code>.</p><p><strong>When to implement this method for new types <code>X</code></strong></p><p>This function is used by implementations of <a href="../../../reference/models/#MathOptInterface.copy_to"><code>MOI.copy_to</code></a> on constraint functions, attribute values and submittable values. If you define a new attribute whose values <code>x::X</code> contain variable or constraint indices, you must also implement this function.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L210-L223">source</a></section><section><div><pre><code class="language-julia hljs">map_indices(
+    variable_map::AbstractDict{T,T},
+    x::X,
+)::X where {T&lt;:MOI.Index,X}</code></pre><p>Shortcut for <code>map_indices(vi -&gt; variable_map[vi], x)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L229-L236">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.substitute_variables" href="#MathOptInterface.Utilities.substitute_variables"><code>MathOptInterface.Utilities.substitute_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">substitute_variables(variable_map::Function, x)</code></pre><p>Substitute any <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <code>x</code> by <code>variable_map(x)</code>. The <code>variable_map</code> function returns either <a href="../../../reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a>, see <a href="#MathOptInterface.Utilities.eval_variables"><code>eval_variables</code></a> for a similar function where <code>variable_map</code> returns a number.</p><p>This function is used by bridge optimizers on constraint functions, attribute values and submittable values when at least one variable bridge is used hence it needs to be implemented for custom types that are meant to be used as attribute or submittable value.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>When implementing a new method, don&#39;t use <code>substitute_variables(::Function</code>, because Julia will not specialize on it. Use instead <code>substitute_variables(::F, ...) where {F&lt;:Function}</code>.</p></div></div></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L412-L429">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.filter_variables" href="#MathOptInterface.Utilities.filter_variables"><code>MathOptInterface.Utilities.filter_variables</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">filter_variables(keep::Function, f::AbstractFunction)</code></pre><p>Return a new function <code>f</code> with the variable <code>vi</code> such that <code>!keep(vi)</code> removed.</p><p>WARNING: Don&#39;t define <code>filter_variables(::Function, ...)</code> because Julia will not specialize on this. Define instead <code>filter_variables(::F, ...) where {F&lt;:Function}</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1197-L1205">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.remove_variable" href="#MathOptInterface.Utilities.remove_variable"><code>MathOptInterface.Utilities.remove_variable</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">remove_variable(f::AbstractFunction, vi::VariableIndex)</code></pre><p>Return a new function <code>f</code> with the variable vi removed.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1273-L1277">source</a></section><section><div><pre><code class="language-julia hljs">remove_variable(
+    f::MOI.AbstractFunction,
+    s::MOI.AbstractSet,
+    vi::MOI.VariableIndex,
+)</code></pre><p>Return a tuple <code>(g, t)</code> representing the constraint <code>f</code>-in-<code>s</code> with the variable <code>vi</code> removed. That is, the terms containing the variable <code>vi</code> in the function <code>f</code> are removed and the dimension of the set <code>s</code> is updated if needed (e.g. when <code>f</code> is a <code>VectorOfVariables</code> with <code>vi</code> being one of the variables).</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/vector_of_constraints.jl#L179-L191">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.all_coefficients" href="#MathOptInterface.Utilities.all_coefficients"><code>MathOptInterface.Utilities.all_coefficients</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">all_coefficients(p::Function, f::MOI.AbstractFunction)</code></pre><p>Determine whether predicate <code>p</code> returns <code>true</code> for all coefficients of <code>f</code>, returning <code>false</code> as soon as the first coefficient of <code>f</code> for which <code>p</code> returns <code>false</code> is encountered (short-circuiting). Similar to <code>all</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1107-L1113">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.unsafe_add" href="#MathOptInterface.Utilities.unsafe_add"><code>MathOptInterface.Utilities.unsafe_add</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">unsafe_add(t1::MOI.ScalarAffineTerm, t2::MOI.ScalarAffineTerm)</code></pre><p>Sums the coefficients of <code>t1</code> and <code>t2</code> and returns an output <code>MOI.ScalarAffineTerm</code>. It is unsafe because it uses the <code>variable</code> of <code>t1</code> as the <code>variable</code> of the output without checking that it is equal to that of <code>t2</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L883-L889">source</a></section><section><div><pre><code class="language-julia hljs">unsafe_add(t1::MOI.ScalarQuadraticTerm, t2::MOI.ScalarQuadraticTerm)</code></pre><p>Sums the coefficients of <code>t1</code> and <code>t2</code> and returns an output <code>MOI.ScalarQuadraticTerm</code>. It is unsafe because it uses the <code>variable</code>&#39;s of <code>t1</code> as the <code>variable</code>&#39;s of the output without checking that they are the same (up to permutation) to those of <code>t2</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L894-L901">source</a></section><section><div><pre><code class="language-julia hljs">unsafe_add(t1::MOI.VectorAffineTerm, t2::MOI.VectorAffineTerm)</code></pre><p>Sums the coefficients of <code>t1</code> and <code>t2</code> and returns an output <code>MOI.VectorAffineTerm</code>. It is unsafe because it uses the <code>output_index</code> and <code>variable</code> of <code>t1</code> as the <code>output_index</code> and <code>variable</code> of the output term without checking that they are equal to those of <code>t2</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L910-L917">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.isapprox_zero" href="#MathOptInterface.Utilities.isapprox_zero"><code>MathOptInterface.Utilities.isapprox_zero</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">isapprox_zero(f::MOI.AbstractFunction, tol)</code></pre><p>Return a <code>Bool</code> indicating whether the function <code>f</code> is approximately zero using <code>tol</code> as a tolerance.</p><p><strong>Important note</strong></p><p>This function assumes that <code>f</code> does not contain any duplicate terms, you might want to first call <a href="#MathOptInterface.Utilities.canonical"><code>canonical</code></a> if that is not guaranteed. For instance, given</p><pre><code class="language-julia hljs">f = MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.([1, -1], [x, x]), 0)</code></pre><p>then <code>isapprox_zero(f)</code> is <code>false</code> but <code>isapprox_zero(MOIU.canonical(f))</code> is <code>true</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1126-L1142">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.modify_function" href="#MathOptInterface.Utilities.modify_function"><code>MathOptInterface.Utilities.modify_function</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">modify_function(f::AbstractFunction, change::AbstractFunctionModification)</code></pre><p>Return a copy of the function <code>f</code>, modified according to <code>change</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1291-L1295">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.zero_with_output_dimension" href="#MathOptInterface.Utilities.zero_with_output_dimension"><code>MathOptInterface.Utilities.zero_with_output_dimension</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">zero_with_output_dimension(::Type{T}, output_dimension::Integer) where {T}</code></pre><p>Create an instance of type <code>T</code> with the output dimension <code>output_dimension</code>.</p><p>This is mostly useful in Bridges, when code needs to be agnostic to the type of vector-valued function that is passed in.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L851-L858">source</a></section></article><p>The following functions can be used to canonicalize a function:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.is_canonical" href="#MathOptInterface.Utilities.is_canonical"><code>MathOptInterface.Utilities.is_canonical</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_canonical(f::Union{ScalarAffineFunction, VectorAffineFunction})</code></pre><p>Returns a Bool indicating whether the function is in canonical form. See <a href="#MathOptInterface.Utilities.canonical"><code>canonical</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L952-L957">source</a></section><section><div><pre><code class="language-julia hljs">is_canonical(f::Union{ScalarQuadraticFunction, VectorQuadraticFunction})</code></pre><p>Returns a Bool indicating whether the function is in canonical form. See <a href="#MathOptInterface.Utilities.canonical"><code>canonical</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L964-L969">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.canonical" href="#MathOptInterface.Utilities.canonical"><code>MathOptInterface.Utilities.canonical</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">canonical(f::MOI.AbstractFunction)</code></pre><p>Returns the function in a canonical form, i.e.</p><ul><li>A term appear only once.</li><li>The coefficients are nonzero.</li><li>The terms appear in increasing order of variable where there the order of the variables is the order of their value.</li><li>For a <code>AbstractVectorFunction</code>, the terms are sorted in ascending order of output index.</li></ul><p>The output of <code>canonical</code> can be assumed to be a copy of <code>f</code>, even for <code>VectorOfVariables</code>.</p><p><strong>Examples</strong></p><p>If <code>x</code> (resp. <code>y</code>, <code>z</code>) is <code>VariableIndex(1)</code> (resp. 2, 3). The canonical representation of <code>ScalarAffineFunction([y, x, z, x, z], [2, 1, 3, -2, -3], 5)</code> is <code>ScalarAffineFunction([x, y], [-1, 2], 5)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1002-L1022">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.canonicalize!" href="#MathOptInterface.Utilities.canonicalize!"><code>MathOptInterface.Utilities.canonicalize!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">canonicalize!(f::Union{ScalarAffineFunction, VectorAffineFunction})</code></pre><p>Convert a function to canonical form in-place, without allocating a copy to hold the result. See <a href="#MathOptInterface.Utilities.canonical"><code>canonical</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1033-L1038">source</a></section><section><div><pre><code class="language-julia hljs">canonicalize!(f::Union{ScalarQuadraticFunction, VectorQuadraticFunction})</code></pre><p>Convert a function to canonical form in-place, without allocating a copy to hold the result. See <a href="#MathOptInterface.Utilities.canonical"><code>canonical</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L1062-L1067">source</a></section></article><p>The following functions can be used to manipulate functions with basic algebra:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.scalar_type" href="#MathOptInterface.Utilities.scalar_type"><code>MathOptInterface.Utilities.scalar_type</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">scalar_type(F::Type{&lt;:MOI.AbstractVectorFunction})</code></pre><p>Type of functions obtained by indexing objects obtained by calling <code>eachscalar</code> on functions of type <code>F</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L578-L583">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.scalarize" href="#MathOptInterface.Utilities.scalarize"><code>MathOptInterface.Utilities.scalarize</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">scalarize(func::MOI.VectorOfVariables, ignore_constants::Bool = false)</code></pre><p>Returns a vector of scalar functions making up the vector function in the form of a <code>Vector{MOI.SingleVariable}</code>.</p><p>See also <a href="#MathOptInterface.Utilities.eachscalar"><code>eachscalar</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L2197-L2204">source</a></section><section><div><pre><code class="language-julia hljs">scalarize(func::MOI.VectorAffineFunction{T}, ignore_constants::Bool = false)</code></pre><p>Returns a vector of scalar functions making up the vector function in the form of a <code>Vector{MOI.ScalarAffineFunction{T}}</code>.</p><p>See also <a href="#MathOptInterface.Utilities.eachscalar"><code>eachscalar</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L2209-L2216">source</a></section><section><div><pre><code class="language-julia hljs">scalarize(func::MOI.VectorQuadraticFunction{T}, ignore_constants::Bool = false)</code></pre><p>Returns a vector of scalar functions making up the vector function in the form of a <code>Vector{MOI.ScalarQuadraticFunction{T}}</code>.</p><p>See also <a href="#MathOptInterface.Utilities.eachscalar"><code>eachscalar</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L2236-L2243">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.eachscalar" href="#MathOptInterface.Utilities.eachscalar"><code>MathOptInterface.Utilities.eachscalar</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">eachscalar(f::MOI.AbstractVectorFunction)</code></pre><p>Returns an iterator for the scalar components of the vector function.</p><p>See also <a href="#MathOptInterface.Utilities.scalarize"><code>scalarize</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L706-L712">source</a></section><section><div><pre><code class="language-julia hljs">eachscalar(f::MOI.AbstractVector)</code></pre><p>Returns an iterator for the scalar components of the vector.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L715-L719">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.promote_operation" href="#MathOptInterface.Utilities.promote_operation"><code>MathOptInterface.Utilities.promote_operation</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">promote_operation(
+    op::Function,
+    ::Type{T},
+    ArgsTypes::Type{&lt;:Union{T,AbstractVector{T},MOI.AbstractFunction}}...,
+) where {T&lt;:Number}</code></pre><p>Compute the return type of the call <code>operate(op, T, args...)</code>, where the types of the arguments <code>args</code> are <code>ArgsTypes</code>.</p><p>One assumption is that the element type <code>T</code> is invariant under each operation. That is, <code>op(::T, ::T)::T</code> where <code>op</code> is a <code>+</code>, <code>-</code>, <code>*</code>, and <code>/</code>.</p><p>There are six methods for which we implement <code>Utilities.promote_operation</code>:</p><ol><li><code>+</code> a. <code>promote_operation(::typeof(+), ::Type{T}, ::Type{F1}, ::Type{F2})</code></li><li><code>-</code> a. <code>promote_operation(::typeof(-), ::Type{T}, ::Type{F})</code> b. <code>promote_operation(::typeof(-), ::Type{T}, ::Type{F1}, ::Type{F2})</code></li><li><code>*</code> a. <code>promote_operation(::typeof(*), ::Type{T}, ::Type{T}, ::Type{F})</code> b. <code>promote_operation(::typeof(*), ::Type{T}, ::Type{F}, ::Type{T})</code> c. <code>promote_operation(::typeof(*), ::Type{T}, ::Type{F1}, ::Type{F2})</code>    where <code>F1</code> and <code>F2</code> are <code>VariableIndex</code> or <code>ScalarAffineFunction</code> d. <code>promote_operation(::typeof(*), ::Type{T}, ::Type{&lt;:Diagonal{T}}, ::Type{F}</code></li><li><code>/</code> a. <code>promote_operation(::typeof(/), ::Type{T}, ::Type{F}, ::Type{T})</code></li><li><code>vcat</code> a. <code>promote_operation(::typeof(vcat), ::Type{T}, ::Type{F}...)</code></li><li><code>imag</code> a. <code>promote_operation(::typeof(imag), ::Type{T}, ::Type{F})</code>    where <code>F</code> is <code>VariableIndex</code> or <code>VectorOfVariables</code></li></ol><p>In each case, <code>F</code> (or <code>F1</code> and <code>F2</code>) is one of the ten supported types, with a restriction that the mathematical operation makes sense, for example, we don&#39;t define <code>promote_operation(-, T, F1, F2)</code> where <code>F1</code> is a scalar-valued function and  <code>F2</code> is a vector-valued function. The ten supported types are:</p><ol><li>::T</li><li>::VariableIndex</li><li>::ScalarAffineFunction{T}</li><li>::ScalarQuadraticFunction{T}</li><li>::ScalarNonlinearFunction</li><li>::AbstractVector{T}</li><li>::VectorOfVariables</li><li>::VectorAffineFunction{T}</li><li>::VectorQuadraticFunction{T}</li><li>::VectorNonlinearFunction</li></ol></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/promote_operation.jl#L7-L56">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.operate" href="#MathOptInterface.Utilities.operate"><code>MathOptInterface.Utilities.operate</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">operate(
+    op::Function,
+    ::Type{T},
+    args::Union{T,MOI.AbstractFunction}...,
+)::MOI.AbstractFunction where {T&lt;:Number}</code></pre><p>Returns an <code>MOI.AbstractFunction</code> representing the function resulting from the operation <code>op(args...)</code> on functions of coefficient type <code>T</code>.</p><p>No argument can be modified.</p><p><strong>Methods</strong></p><ol><li><code>+</code> a. <code>operate(::typeof(+), ::Type{T}, ::F1)</code> b. <code>operate(::typeof(+), ::Type{T}, ::F1, ::F2)</code> c. <code>operate(::typeof(+), ::Type{T}, ::F1...)</code></li><li><code>-</code> a. <code>operate(::typeof(-), ::Type{T}, ::F)</code> b. <code>operate(::typeof(-), ::Type{T}, ::F1, ::F2)</code></li><li><code>*</code> a. <code>operate(::typeof(*), ::Type{T}, ::T, ::F)</code> b. <code>operate(::typeof(*), ::Type{T}, ::F, ::T)</code> c. <code>operate(::typeof(*), ::Type{T}, ::F1, ::F2)</code>    where <code>F1</code> and <code>F2</code> are <code>VariableIndex</code> or <code>ScalarAffineFunction</code> d. <code>operate(::typeof(*), ::Type{T}, ::Diagonal{T}, ::F)</code></li><li><code>/</code> a. <code>operate(::typeof(/), ::Type{T}, ::F, ::T)</code></li><li><code>vcat</code> a. <code>operate(::typeof(vcat), ::Type{T}, ::F...)</code></li><li><code>imag</code> a. <code>operate(::typeof(imag), ::Type{T}, ::F)</code>    where <code>F</code> is <code>VariableIndex</code> or <code>VectorOfVariables</code></li></ol><p>One assumption is that the element type <code>T</code> is invariant under each operation. That is, <code>op(::T, ::T)::T</code> where <code>op</code> is a <code>+</code>, <code>-</code>, <code>*</code>, and <code>/</code>.</p><p>In each case, <code>F</code> (or <code>F1</code> and <code>F2</code>) is one of the ten supported types, with a restriction that the mathematical operation makes sense, for example, we don&#39;t define <code>promote_operation(-, T, F1, F2)</code> where <code>F1</code> is a scalar-valued function and  <code>F2</code> is a vector-valued function. The ten supported types are:</p><ol><li>::T</li><li>::VariableIndex</li><li>::ScalarAffineFunction{T}</li><li>::ScalarQuadraticFunction{T}</li><li>::ScalarNonlinearFunction</li><li>::AbstractVector{T}</li><li>::VectorOfVariables</li><li>::VectorAffineFunction{T}</li><li>::VectorQuadraticFunction{T}</li><li>::VectorNonlinearFunction</li></ol></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/operate.jl#L7-L60">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.operate!" href="#MathOptInterface.Utilities.operate!"><code>MathOptInterface.Utilities.operate!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">operate!(
+    op::Function,
+    ::Type{T},
+    args::Union{T,MOI.AbstractFunction}...,
+)::MOI.AbstractFunction where {T&lt;:Number}</code></pre><p>Returns an <code>MOI.AbstractFunction</code> representing the function resulting from the operation <code>op(args...)</code> on functions of coefficient type <code>T</code>.</p><p>The first argument may be modified, in which case the return value is identical to the first argument. For operations which cannot be implemented in-place, this function returns a new object.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/operate.jl#L1036-L1049">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.operate_output_index!" href="#MathOptInterface.Utilities.operate_output_index!"><code>MathOptInterface.Utilities.operate_output_index!</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">operate_output_index!(
+    op::Union{typeof(+),typeof(-)},
+    ::Type{T},
+    output_index::Integer,
+    f::Union{AbstractVector{T},MOI.AbstractVectorFunction}
+    g::Union{T,MOI.AbstractScalarFunction}...
+) where {T&lt;:Number}</code></pre><p>Return an <code>MOI.AbstractVectorFunction</code> in which the scalar function in row <code>output_index</code> is the result of <code>op(f[output_index], g)</code>.</p><p>The functions at output index different to <code>output_index</code> are the same as the functions at the same output index in <code>func</code>. The first argument may be modified.</p><p><strong>Methods</strong></p><ol><li><code>+</code> a. <code>operate_output_index!(+, ::Type{T}, ::Int, ::VectorF, ::ScalarF)</code></li><li><code>-</code> a. <code>operate_output_index!(-, ::Type{T}, ::Int, ::VectorF, ::ScalarF)</code></li></ol></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/operate.jl#L1612-L1634">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.vectorize" href="#MathOptInterface.Utilities.vectorize"><code>MathOptInterface.Utilities.vectorize</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">vectorize(x::AbstractVector{&lt;:Number})</code></pre><p>Returns <code>x</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L2111-L2115">source</a></section><section><div><pre><code class="language-julia hljs">vectorize(x::AbstractVector{MOI.VariableIndex})</code></pre><p>Returns the vector of scalar affine functions in the form of a <code>MOI.VectorAffineFunction{T}</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L2118-L2123">source</a></section><section><div><pre><code class="language-julia hljs">vectorize(funcs::AbstractVector{MOI.ScalarAffineFunction{T}}) where T</code></pre><p>Returns the vector of scalar affine functions in the form of a <code>MOI.VectorAffineFunction{T}</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L2134-L2139">source</a></section><section><div><pre><code class="language-julia hljs">vectorize(funcs::AbstractVector{MOI.ScalarQuadraticFunction{T}}) where T</code></pre><p>Returns the vector of scalar quadratic functions in the form of a <code>MOI.VectorQuadraticFunction{T}</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/functions.jl#L2151-L2156">source</a></section></article><h2 id="Constraint-utilities"><a class="docs-heading-anchor" href="#Constraint-utilities">Constraint utilities</a><a id="Constraint-utilities-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-utilities" title="Permalink"></a></h2><p>The following utilities are available for moving the function constant to the set for scalar constraints:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.shift_constant" href="#MathOptInterface.Utilities.shift_constant"><code>MathOptInterface.Utilities.shift_constant</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">shift_constant(set::MOI.AbstractScalarSet, offset)</code></pre><p>Returns a new scalar set <code>new_set</code> such that <code>func</code>-in-<code>set</code> is equivalent to <code>func + offset</code>-in-<code>new_set</code>.</p><p>Only define this function if it makes sense to!</p><p>Use <a href="#MathOptInterface.Utilities.supports_shift_constant"><code>supports_shift_constant</code></a> to check if the set supports shifting:</p><pre><code class="language-Julia hljs">if supports_shift_constant(typeof(old_set))
+    new_set = shift_constant(old_set, offset)
+    f.constant = 0
+    add_constraint(model, f, new_set)
+else
+    add_constraint(model, f, old_set)
+end</code></pre><p>See also <a href="#MathOptInterface.Utilities.supports_shift_constant"><code>supports_shift_constant</code></a>.</p><p><strong>Examples</strong></p><p>The call <code>shift_constant(MOI.Interval(-2, 3), 1)</code> is equal to <code>MOI.Interval(-1, 4)</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sets.jl#L7-L32">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.supports_shift_constant" href="#MathOptInterface.Utilities.supports_shift_constant"><code>MathOptInterface.Utilities.supports_shift_constant</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">supports_shift_constant(::Type{S}) where {S&lt;:MOI.AbstractSet}</code></pre><p>Return <code>true</code> if <a href="#MathOptInterface.Utilities.shift_constant"><code>shift_constant</code></a> is defined for set <code>S</code>.</p><p>See also <a href="#MathOptInterface.Utilities.shift_constant"><code>shift_constant</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sets.jl#L35-L41">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.normalize_and_add_constraint" href="#MathOptInterface.Utilities.normalize_and_add_constraint"><code>MathOptInterface.Utilities.normalize_and_add_constraint</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">normalize_and_add_constraint(
+    model::MOI.ModelLike,
+    func::MOI.AbstractScalarFunction,
+    set::MOI.AbstractScalarSet;
+    allow_modify_function::Bool = false,
+)</code></pre><p>Adds the scalar constraint obtained by moving the constant term in <code>func</code> to the set in <code>model</code>. If <code>allow_modify_function</code> is <code>true</code> then the function <code>func</code> can be modified.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/constraints.jl#L7-L18">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.normalize_constant" href="#MathOptInterface.Utilities.normalize_constant"><code>MathOptInterface.Utilities.normalize_constant</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">normalize_constant(
+    func::MOI.AbstractScalarFunction,
+    set::MOI.AbstractScalarSet;
+    allow_modify_function::Bool = false,
+)</code></pre><p>Return the <code>func</code>-in-<code>set</code> constraint in normalized form. That is, if <code>func</code> is <a href="../../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction</code></a> or <a href="../../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a>, the constant is moved to the set. If <code>allow_modify_function</code> is <code>true</code> then the function <code>func</code> can be modified.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/constraints.jl#L35-L46">source</a></section></article><p>The following utility identifies those constraints imposing bounds on a given variable, and returns those bound values:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.get_bounds" href="#MathOptInterface.Utilities.get_bounds"><code>MathOptInterface.Utilities.get_bounds</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">get_bounds(model::MOI.ModelLike, ::Type{T}, x::MOI.VariableIndex)</code></pre><p>Return a tuple <code>(lb, ub)</code> of type <code>Tuple{T, T}</code>, where <code>lb</code> and <code>ub</code> are lower  and upper bounds, respectively, imposed on <code>x</code> in <code>model</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/variables.jl#L7-L12">source</a></section></article><p>The following utilities are useful when working with symmetric matrix cones.</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.is_diagonal_vectorized_index" href="#MathOptInterface.Utilities.is_diagonal_vectorized_index"><code>MathOptInterface.Utilities.is_diagonal_vectorized_index</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">is_diagonal_vectorized_index(index::Base.Integer)</code></pre><p>Return whether <code>index</code> is the index of a diagonal element in a <a href="../../../reference/standard_form/#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>MOI.AbstractSymmetricMatrixSetTriangle</code></a> set.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sets.jl#L119-L124">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.side_dimension_for_vectorized_dimension" href="#MathOptInterface.Utilities.side_dimension_for_vectorized_dimension"><code>MathOptInterface.Utilities.side_dimension_for_vectorized_dimension</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">side_dimension_for_vectorized_dimension(n::Integer)</code></pre><p>Return the dimension <code>d</code> such that <code>MOI.dimension(MOI.PositiveSemidefiniteConeTriangle(d))</code> is <code>n</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/sets.jl#L131-L136">source</a></section></article><h2 id="Set-utilities"><a class="docs-heading-anchor" href="#Set-utilities">Set utilities</a><a id="Set-utilities-1"></a><a class="docs-heading-anchor-permalink" href="#Set-utilities" title="Permalink"></a></h2><p>The following utilities are available for sets:</p><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.AbstractDistance" href="#MathOptInterface.Utilities.AbstractDistance"><code>MathOptInterface.Utilities.AbstractDistance</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">abstract type AbstractDistance end</code></pre><p>An abstract type used to enabble dispatch of <a href="#MathOptInterface.Utilities.distance_to_set"><code>Utilities.distance_to_set</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/distance_to_set.jl#L7-L12">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.ProjectionUpperBoundDistance" href="#MathOptInterface.Utilities.ProjectionUpperBoundDistance"><code>MathOptInterface.Utilities.ProjectionUpperBoundDistance</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">ProjectionUpperBoundDistance() &lt;: AbstractDistance</code></pre><p>An upper bound on the minimum distance between <code>point</code> and the closest feasible point in <code>set</code>.</p><p><strong>Definition of distance</strong></p><p>The minimum distance is computed as:</p><p class="math-container">\[d(x, \mathcal{K}) = \min_{y \in \mathcal{K}} || x - y ||\]</p><p>where <span>$x$</span> is <code>point</code> and <span>$\mathcal{K}$</span> is <code>set</code>. The norm is computed as:</p><p class="math-container">\[||x|| = \sqrt{f(x, x, \mathcal{K})}\]</p><p>where <span>$f$</span> is <a href="#MathOptInterface.Utilities.set_dot"><code>Utilities.set_dot</code></a>.</p><p>In the default case, where the set does not have a specialized method for <a href="#MathOptInterface.Utilities.set_dot"><code>Utilities.set_dot</code></a>, the norm is equivalent to the Euclidean norm <span>$||x|| = \sqrt{\sum x_i^2}$</span>.</p><p><strong>Why an upper bound?</strong></p><p>In most cases, <code>distance_to_set</code> should return the smallest upper bound, but it may return a larger value if the smallest upper bound is expensive to compute.</p><p>For example, given an epigraph from of a conic set, <span>$\{(t, x) | f(x) \le t\}$</span>, it may be simpler to return <span>$\delta$</span> such that <span>$f(x) \le t + \delta$</span>, rather than computing the nearest projection onto the set.</p><p>If the distance is not the smallest upper bound, the docstring of the appropriate <code>distance_to_set</code> method <em>must</em> describe the way that the distance is computed.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/distance_to_set.jl#L15-L49">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.distance_to_set" href="#MathOptInterface.Utilities.distance_to_set"><code>MathOptInterface.Utilities.distance_to_set</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">distance_to_set(
+    [d::AbstractDistance = ProjectionUpperBoundDistance()],]
+    point::T,
+    set::MOI.AbstractScalarSet,
+) where {T}
+
+distance_to_set(
+    [d::AbstractDistance = ProjectionUpperBoundDistance(),]
+    point::AbstractVector{T},
+    set::MOI.AbstractVectorSet,
+) where {T}</code></pre><p>Compute the distance between <code>point</code> and <code>set</code> using the distance metric <code>d</code>. If <code>point</code> is in the set <code>set</code>, this function <em>must</em> return <code>zero(T)</code>.</p><p>If <code>d</code> is omitted, the default distance is <a href="#MathOptInterface.Utilities.ProjectionUpperBoundDistance"><code>Utilities.ProjectionUpperBoundDistance</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/distance_to_set.jl#L52-L69">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.set_dot" href="#MathOptInterface.Utilities.set_dot"><code>MathOptInterface.Utilities.set_dot</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">set_dot(x::AbstractVector, y::AbstractVector, set::AbstractVectorSet)</code></pre><p>Return the scalar product between a vector <code>x</code> of the set <code>set</code> and a vector <code>y</code> of the dual of the set <code>s</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/set_dot.jl#L11-L16">source</a></section><section><div><pre><code class="language-julia hljs">set_dot(x, y, set::AbstractScalarSet)</code></pre><p>Return the scalar product between a number <code>x</code> of the set <code>set</code> and a number <code>y</code> of the dual of the set <code>s</code>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/set_dot.jl#L21-L26">source</a></section></article><h2 id="DoubleDicts"><a class="docs-heading-anchor" href="#DoubleDicts">DoubleDicts</a><a id="DoubleDicts-1"></a><a class="docs-heading-anchor-permalink" href="#DoubleDicts" title="Permalink"></a></h2><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.DoubleDicts.DoubleDict" href="#MathOptInterface.Utilities.DoubleDicts.DoubleDict"><code>MathOptInterface.Utilities.DoubleDicts.DoubleDict</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">DoubleDict{V}</code></pre><p>An optimized dictionary to map <code>MOI.ConstraintIndex</code> to values of type <code>V</code>.</p><p>Works as a <code>AbstractDict{MOI.ConstraintIndex,V}</code> with minimal differences.</p><p>If <code>V</code> is also a <code>MOI.ConstraintIndex</code>, use <a href="#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict"><code>IndexDoubleDict</code></a>.</p><p>Note that <code>MOI.ConstraintIndex</code> is not a concrete type, opposed to <code>MOI.ConstraintIndex{MOI.VariableIndex, MOI.Integers}</code>, which is a concrete type.</p><p>When looping through multiple keys of the same Function-in-Set type, use</p><pre><code class="language-julia hljs">inner = dict[F, S]</code></pre><p>to return a type-stable <a href="#MathOptInterface.Utilities.DoubleDicts.DoubleDictInner"><code>DoubleDictInner</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/DoubleDicts.jl#L25-L43">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.DoubleDicts.DoubleDictInner" href="#MathOptInterface.Utilities.DoubleDicts.DoubleDictInner"><code>MathOptInterface.Utilities.DoubleDicts.DoubleDictInner</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">DoubleDictInner{F,S,V}</code></pre><p>A type stable inner dictionary of <a href="#MathOptInterface.Utilities.DoubleDicts.DoubleDict"><code>DoubleDict</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/DoubleDicts.jl#L51-L55">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict" href="#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict"><code>MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndexDoubleDict</code></pre><p>A specialized version of [<code>DoubleDict</code>] in which the values are of type <code>MOI.ConstraintIndex</code></p><p>When looping through multiple keys of the same Function-in-Set type, use</p><pre><code class="language-julia hljs">inner = dict[F, S]</code></pre><p>to return a type-stable <a href="#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDictInner"><code>IndexDoubleDictInner</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/DoubleDicts.jl#L70-L81">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.DoubleDicts.IndexDoubleDictInner" href="#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDictInner"><code>MathOptInterface.Utilities.DoubleDicts.IndexDoubleDictInner</code></a> — <span class="docstring-category">Type</span></header><section><div><pre><code class="language-julia hljs">IndexDoubleDictInner{F,S}</code></pre><p>A type stable inner dictionary of <a href="#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict"><code>IndexDoubleDict</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/DoubleDicts.jl#L89-L93">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.DoubleDicts.outer_keys" href="#MathOptInterface.Utilities.DoubleDicts.outer_keys"><code>MathOptInterface.Utilities.DoubleDicts.outer_keys</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">outer_keys(d::AbstractDoubleDict)</code></pre><p>Return an iterator over the outer keys of the <code>AbstractDoubleDict</code> <code>d</code>. Each outer key is a <code>Tuple{Type,Type}</code> so that a double loop can be easily used:</p><pre><code class="language-julia hljs">for (F, S) in DoubleDicts.outer_keys(dict)
+    for (k, v) in dict[F, S]
+        # ...
+    end
+end</code></pre><p>For performance, it is recommended that the inner loop lies in a separate function to gurantee type-stability. Some outer keys <code>(F, S)</code> might lead to an empty <code>dict[F, S]</code>. If you want only nonempty <code>dict[F, S]</code>, use <a href="#MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys"><code>nonempty_outer_keys</code></a>.</p></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/DoubleDicts.jl#L300-L318">source</a></section></article><article class="docstring"><header><a class="docstring-article-toggle-button fa-solid fa-chevron-down" href="javascript:;" title="Collapse docstring"></a><a class="docstring-binding" id="MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys" href="#MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys"><code>MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys</code></a> — <span class="docstring-category">Function</span></header><section><div><pre><code class="language-julia hljs">nonempty_outer_keys(d::AbstractDoubleDict)</code></pre><p>Return a vector of outer keys of the <code>AbstractDoubleDict</code> <code>d</code>.</p><p>Only outer keys that have a nonempty set of inner keys will be returned.</p><p>Each outer key is a <code>Tuple{Type,Type}</code> so that a double loop can be easily used</p><pre><code class="language-julia hljs">for (F, S) in DoubleDicts.nonempty_outer_keys(dict)
+    for (k, v) in dict[F, S]
+        # ...
+    end
+end
+For performance, it is recommended that the inner loop lies in a separate
+function to gurantee type-stability.
+
+If you want an iterator of all current outer keys, use [`outer_keys`](@ref).</code></pre></div><a class="docs-sourcelink" target="_blank" href="https://github.com/jump-dev/MathOptInterface.jl/blob/c059b1786745e5f6bf87f5e0be66dc2a5cf3621d/src/Utilities/DoubleDicts.jl#L321-L341">source</a></section></article></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../overview/">« Overview</a><a class="docs-footer-nextpage" href="../../Test/overview/">Overview »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/tutorials/bridging_constraint/index.html b/previews/PR3525/moi/tutorials/bridging_constraint/index.html
new file mode 100644
index 00000000000..4dee52e197c
--- /dev/null
+++ b/previews/PR3525/moi/tutorials/bridging_constraint/index.html
@@ -0,0 +1,106 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Implementing a constraint bridge · JuMP</title><meta name="title" content="Implementing a constraint bridge · JuMP"/><meta property="og:title" content="Implementing a constraint bridge · JuMP"/><meta property="twitter:title" content="Implementing a constraint bridge · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../mathprogbase/">Transitioning from MathProgBase</a></li><li class="is-active"><a class="tocitem" href>Implementing a constraint bridge</a><ul class="internal"><li><a class="tocitem" href="#Preliminaries"><span>Preliminaries</span></a></li><li><a class="tocitem" href="#Four-mandatory-parts-in-a-constraint-bridge"><span>Four mandatory parts in a constraint bridge</span></a></li><li><a class="tocitem" href="#Bridge-registration"><span>Bridge registration</span></a></li><li><a class="tocitem" href="#Bridge-improvements"><span>Bridge improvements</span></a></li></ul></li><li><a class="tocitem" href="../manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Tutorials</a></li><li class="is-active"><a href>Implementing a constraint bridge</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Implementing a constraint bridge</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/bridging_constraint.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Implementing-a-constraint-bridge"><a class="docs-heading-anchor" href="#Implementing-a-constraint-bridge">Implementing a constraint bridge</a><a id="Implementing-a-constraint-bridge-1"></a><a class="docs-heading-anchor-permalink" href="#Implementing-a-constraint-bridge" title="Permalink"></a></h1><p>This guide outlines the basic steps to create a new bridge from a constraint expressed in the formalism <code>Function</code>-in-<code>Set</code>.</p><h2 id="Preliminaries"><a class="docs-heading-anchor" href="#Preliminaries">Preliminaries</a><a id="Preliminaries-1"></a><a class="docs-heading-anchor-permalink" href="#Preliminaries" title="Permalink"></a></h2><p>First, decide on the set you want to bridge. Then, study its properties: the most important one is whether the set is scalar or vector, which impacts the dimensionality of the functions that can be used with the set.</p><ul><li>A scalar function only has one dimension. MOI defines three types of scalar functions: a variable (<a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>), an affine function (<a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>), or a quadratic function (<a href="../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a>).</li><li>A vector function has several dimensions (at least one). MOI defines three types of vector functions: several variables (<a href="../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>VectorOfVariables</code></a>), an affine function (<a href="../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a>), or a quadratic function (<a href="../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>). The main difference with scalar functions is that the order of dimensions can be very important: for instance, in an indicator constraint (<a href="../../reference/standard_form/#MathOptInterface.Indicator"><code>Indicator</code></a>), the first dimension indicates whether the constraint about the second dimension is active.</li></ul><p>To explain how to implement a bridge, we present the example of <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.FlipSignBridge"><code>Bridges.Constraint.FlipSignBridge</code></a>. This bridge maps <code>&lt;=</code> (<a href="../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a>) constraints to <code>&gt;=</code> (<a href="../../reference/standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a>) constraints. This corresponds to reversing the sign of the inequality. We focus on scalar affine functions (we disregard the cases of a single variable or of quadratic functions). This example is a simplified version of the code included in MOI.</p><h2 id="Four-mandatory-parts-in-a-constraint-bridge"><a class="docs-heading-anchor" href="#Four-mandatory-parts-in-a-constraint-bridge">Four mandatory parts in a constraint bridge</a><a id="Four-mandatory-parts-in-a-constraint-bridge-1"></a><a class="docs-heading-anchor-permalink" href="#Four-mandatory-parts-in-a-constraint-bridge" title="Permalink"></a></h2><p>The first part of a constraint bridge is a new concrete subtype of <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.AbstractBridge"><code>Bridges.Constraint.AbstractBridge</code></a>. This type must have fields to store all the new variables and constraints that the bridge will add. Typically, these types are parametrized by the type of the coefficients in the model.</p><p>Then, three sets of functions must be defined:</p><ol><li><a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.bridge_constraint"><code>Bridges.Constraint.bridge_constraint</code></a>: this function implements the bridge and creates the required variables and constraints.</li><li><a href="../../reference/constraints/#MathOptInterface.supports_constraint"><code>supports_constraint</code></a>: these functions must return <code>true</code> when the combination of function and set is supported by the bridge. By default, the base implementation always returns <code>false</code> and the bridge does not have to provide this implementation.</li><li><a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.added_constrained_variable_types"><code>Bridges.added_constrained_variable_types</code></a> and <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.added_constraint_types"><code>Bridges.added_constraint_types</code></a>: these functions return the types of variables and constraints that this bridge adds. They are used to compute the set of other bridges that are required to use the one you are defining, if need be.</li></ol><p>More functions can be implemented, for instance to retrieve properties from the bridge or deleting a bridged constraint.</p><h3 id=".-Structure-for-the-bridge"><a class="docs-heading-anchor" href="#.-Structure-for-the-bridge">1. Structure for the bridge</a><a id=".-Structure-for-the-bridge-1"></a><a class="docs-heading-anchor-permalink" href="#.-Structure-for-the-bridge" title="Permalink"></a></h3><p>A typical <code>struct</code> behind a bridge depends on the type of the coefficients that are used for the model (typically <code>Float64</code>, but coefficients might also be integers or complex numbers).</p><p>This structure must hold a reference to all the variables and the constraints that are created as part of the bridge.</p><p>The type of this structure is used throughout MOI as an identifier for the bridge. It is passed as argument to most functions related to bridges.</p><p>The best practice is to have the name of this type end with <code>Bridge</code>.</p><p>In our example, the bridge maps any <code>ScalarAffineFunction{T}</code>-in-<code>LessThan{T}</code> constraint to a single <code>ScalarAffineFunction{T}</code>-in-<code>GreaterThan{T}</code> constraint. The affine function has coefficients of type <code>T</code>. The bridge is parametrized with <code>T</code>, so that the constraint that the bridge creates also has coefficients of type <code>T</code>.</p><pre><code class="language-julia hljs">struct SignBridge{T&lt;:Number} &lt;: Bridges.Constraint.AbstractBridge
+    constraint::ConstraintIndex{ScalarAffineFunction{T}, GreaterThan{T}}
+end</code></pre><h3 id=".-Bridge-creation"><a class="docs-heading-anchor" href="#.-Bridge-creation">2. Bridge creation</a><a id=".-Bridge-creation-1"></a><a class="docs-heading-anchor-permalink" href="#.-Bridge-creation" title="Permalink"></a></h3><p>The function <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.bridge_constraint"><code>Bridges.Constraint.bridge_constraint</code></a> is called whenever the bridge is instantiated for a specific model, with the given function and set. The arguments to <code>bridge_constraint</code> are similar to <a href="../../reference/constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a>, with the exception of the first argument: it is the <code>Type</code> of the struct defined in the first step (for our example, <code>Type{SignBridge{T}}</code>).</p><p><code>bridge_constraint</code> returns an instance of the struct defined in the first step. the first step.</p><p>In our example, the bridge constraint could be defined as:</p><pre><code class="language-julia hljs">function Bridges.Constraint.bridge_constraint(
+    ::Type{SignBridge{T}}, # Bridge to use.
+    model::ModelLike, # Model to which the constraint is being added.
+    f::ScalarAffineFunction{T}, # Function to rewrite.
+    s::LessThan{T}, # Set to rewrite.
+) where {T}
+    # Create the variables and constraints required for the bridge.
+    con = add_constraint(model, -f, GreaterThan(-s.upper))
+
+    # Return an instance of the bridge type with a reference to all the
+    # variables and constraints that were created in this function.
+    return SignBridge(con)
+end</code></pre><h3 id=".-Supported-constraint-types"><a class="docs-heading-anchor" href="#.-Supported-constraint-types">3. Supported constraint types</a><a id=".-Supported-constraint-types-1"></a><a class="docs-heading-anchor-permalink" href="#.-Supported-constraint-types" title="Permalink"></a></h3><p>The function <a href="../../reference/constraints/#MathOptInterface.supports_constraint"><code>supports_constraint</code></a> determines whether the bridge type supports a given combination of function and set.</p><p>This function must closely match <code>bridge_constraint</code>, because it will not be called if <code>supports_constraint</code> returns <code>false</code>.</p><pre><code class="language-julia hljs">function supports_constraint(
+    ::Type{SignBridge{T}}, # Bridge to use.
+    ::Type{ScalarAffineFunction{T}}, # Function to rewrite.
+    ::Type{LessThan{T}}, # Set to rewrite.
+) where {T}
+    # Do some computation to ensure that the constraint is supported.
+    # Typically, you can directly return true.
+    return true
+end</code></pre><h3 id=".-Metadata-about-the-bridge"><a class="docs-heading-anchor" href="#.-Metadata-about-the-bridge">4. Metadata about the bridge</a><a id=".-Metadata-about-the-bridge-1"></a><a class="docs-heading-anchor-permalink" href="#.-Metadata-about-the-bridge" title="Permalink"></a></h3><p>To determine whether a bridge can be used, MOI uses a shortest-path algorithm that uses the variable types and the constraints that the bridge can create. This information is communicated from the bridge to MOI using the functions <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.added_constrained_variable_types"><code>Bridges.added_constrained_variable_types</code></a> and <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.added_constraint_types"><code>Bridges.added_constraint_types</code></a>. Both return lists of tuples: either a list of 1-tuples containing the variable types (typically, <code>ZeroOne</code> or <code>Integer</code>) or a list of 2-tuples contained the functions and sets (like <code>ScalarAffineFunction{T}</code>-<code>GreaterThan</code>).</p><p>For our example, the bridge does not create any constrained variables, and only <code>ScalarAffineFunction{T}</code>-in-<code>GreaterThan{T}</code> constraints:</p><pre><code class="language-julia hljs">function Bridges.added_constrained_variable_types(::Type{SignBridge{T}}) where {T}
+    # The bridge does not create variables, return an empty list of tuples:
+    return Tuple{Type}[]
+end
+
+function Bridges.added_constraint_types(::Type{SignBridge{T}}) where {T}
+    return Tuple{Type,Type}[
+        # One element per F-in-S the bridge creates.
+        (ScalarAffineFunction{T}, GreaterThan{T}),
+    ]
+end</code></pre><p>A bridge that creates binary variables would rather have this definition of <code>added_constrained_variable_types</code>:</p><pre><code class="language-julia hljs">function Bridges.added_constrained_variable_types(::Type{SomeBridge{T}}) where {T}
+    # The bridge only creates binary variables:
+    return Tuple{Type}[(ZeroOne,)]
+end</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>If you declare the creation of constrained variables in <code>added_constrained_variable_types</code>, the corresponding constraint type <code>VariableIndex</code> must not be indicated in <code>added_constraint_types</code>. This would restrict the use of the bridge to solvers that can add such a constraint after the variable is created.</p><p>More concretely, <em>if</em> you declare in <code>added_constrained_variable_types</code> that your bridge creates binary variables (<code>ZeroOne</code>), <em>and if</em> you never add such a constraint afterward (you do not call <code>add_constraint(model, var, ZeroOne())</code>), then you must <em>not</em> list <code>(VariableIndex, ZeroOne)</code> in <code>added_constraint_types</code>.</p></div></div><p>Typically, the function <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.concrete_bridge_type"><code>Bridges.Constraint.concrete_bridge_type</code></a> does not have to be defined for most bridges.</p><h2 id="Bridge-registration"><a class="docs-heading-anchor" href="#Bridge-registration">Bridge registration</a><a id="Bridge-registration-1"></a><a class="docs-heading-anchor-permalink" href="#Bridge-registration" title="Permalink"></a></h2><p>For a bridge to be used by MOI, it must be known by MOI.</p><h3 id="SingleBridgeOptimizer"><a class="docs-heading-anchor" href="#SingleBridgeOptimizer"><code>SingleBridgeOptimizer</code></a><a id="SingleBridgeOptimizer-1"></a><a class="docs-heading-anchor-permalink" href="#SingleBridgeOptimizer" title="Permalink"></a></h3><p>The first way to do so is to create a single-bridge optimizer. This type of optimizer wraps another optimizer and adds the possibility to use only one bridge. It is especially useful when unit testing bridges.</p><p>It is common practice to use the same name as the type defined for the bridge (<code>SignBridge</code>, in our example) without the suffix <code>Bridge</code>.</p><pre><code class="language-julia hljs">const Sign{T,OT&lt;: ModelLike} =
+    SingleBridgeOptimizer{SignBridge{T}, OT}</code></pre><p>In the context of unit tests, this bridge is used in conjunction with a <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.MockOptimizer"><code>Utilities.MockOptimizer</code></a>:</p><pre><code class="language-julia hljs">mock = Utilities.MockOptimizer(
+    Utilities.UniversalFallback(Utilities.Model{Float64}()),
+)
+bridged_mock = Sign{Float64}(mock)</code></pre><h3 id="New-bridge-for-a-LazyBridgeOptimizer"><a class="docs-heading-anchor" href="#New-bridge-for-a-LazyBridgeOptimizer">New bridge for a <code>LazyBridgeOptimizer</code></a><a id="New-bridge-for-a-LazyBridgeOptimizer-1"></a><a class="docs-heading-anchor-permalink" href="#New-bridge-for-a-LazyBridgeOptimizer" title="Permalink"></a></h3><p>Typical user-facing models for MOI are based on <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>Bridges.LazyBridgeOptimizer</code></a>. For instance, this type of model is returned by <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.full_bridge_optimizer"><code>Bridges.full_bridge_optimizer</code></a>. These models can be added more bridges by using <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.add_bridge"><code>Bridges.add_bridge</code></a>:</p><pre><code class="language-julia hljs">inner_optimizer = Utilities.Model{Float64}()
+optimizer = Bridges.full_bridge_optimizer(inner_optimizer, Float64)
+Bridges.add_bridge(optimizer, SignBridge{Float64})</code></pre><h2 id="Bridge-improvements"><a class="docs-heading-anchor" href="#Bridge-improvements">Bridge improvements</a><a id="Bridge-improvements-1"></a><a class="docs-heading-anchor-permalink" href="#Bridge-improvements" title="Permalink"></a></h2><h3 id="Attribute-retrieval"><a class="docs-heading-anchor" href="#Attribute-retrieval">Attribute retrieval</a><a id="Attribute-retrieval-1"></a><a class="docs-heading-anchor-permalink" href="#Attribute-retrieval" title="Permalink"></a></h3><p>Like models, bridges have attributes that can be retrieved using <a href="../../reference/models/#MathOptInterface.get"><code>get</code></a> and <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a>. The most important ones are the number of variables and constraints, but also the lists of variables and constraints.</p><p>In our example, we only have one constraint and only have to implement the <a href="../../reference/models/#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints</code></a> and <a href="../../reference/models/#MathOptInterface.ListOfConstraintIndices"><code>ListOfConstraintIndices</code></a> attributes:</p><pre><code class="language-julia hljs">function get(
+    ::SignBridge{T},
+    ::NumberOfConstraints{
+        ScalarAffineFunction{T},
+        GreaterThan{T},
+    },
+) where {T}
+    return 1
+end
+
+function get(
+    bridge::SignBridge{T},
+    ::ListOfConstraintIndices{
+        ScalarAffineFunction{T},
+        GreaterThan{T},
+    },
+) where {T}
+    return [bridge.constraint]
+end</code></pre><p>You must implement one such pair of functions for each type of constraint the bridge adds to the model.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Avoid returning a list from the bridge object without copying it. Users must be able to change the contents of the returned list without altering the bridge object.</p></div></div><p>For variables, the situation is simpler. If your bridge creates new variables, you must implement the <a href="../../reference/models/#MathOptInterface.NumberOfVariables"><code>NumberOfVariables</code></a> and <a href="../../reference/models/#MathOptInterface.ListOfVariableIndices"><code>ListOfVariableIndices</code></a> attributes. However, these attributes do not have parameters, unlike their constraint counterparts. Only two functions suffice:</p><pre><code class="language-julia hljs">function get(
+    ::SignBridge{T},
+    ::NumberOfVariables,
+) where {T}
+    return 0
+end
+
+function get(
+    ::SignBridge{T},
+    ::ListOfVariableIndices,
+) where {T}
+    return VariableIndex[]
+end</code></pre><p>In order for the user to be able to access the function and set of the original constraint, the bridge needs to implement getters for the <a href="../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> and <a href="../../reference/constraints/#MathOptInterface.ConstraintSet"><code>ConstraintSet</code></a> attributes:</p><pre><code class="language-julia hljs">function get(
+    model::MOI.ModelLike,
+    attr::MOI.ConstraintFunction,
+    bridge::SignBridge,
+)
+    return -MOI.get(model, attr, bridge.constraint)
+end
+
+function get(
+    model::MOI.ModelLike,
+    attr::MOI.ConstraintSet,
+    bridge::SignBridge,
+)
+    set = MOI.get(model, attr, bridge.constraint)
+    return MOI.LessThan(-set.lower)
+end</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Alternatively, one could store the original function and set in <code>SignBridge</code> during <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.bridge_constraint"><code>Bridges.Constraint.bridge_constraint</code></a> to make these getters simpler and more efficient. On the other hand, this will increase the memory footprint of the bridges as the garbage collector won&#39;t be able to delete that object. The convention is to not store the function in the bridge and not care too much about the efficiency of these getters. If the user needs efficient getters for <a href="../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> then they should use a <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.CachingOptimizer"><code>Utilities.CachingOptimizer</code></a>.</p></div></div><h3 id="Model-modifications"><a class="docs-heading-anchor" href="#Model-modifications">Model modifications</a><a id="Model-modifications-1"></a><a class="docs-heading-anchor-permalink" href="#Model-modifications" title="Permalink"></a></h3><p>To avoid copying the model when the user request to change a constraint, MOI provides <a href="../../reference/modification/#MathOptInterface.modify"><code>modify</code></a>. Bridges can also implement this API to allow certain changes, such as coefficient changes.</p><p>In our case, a modification of a coefficient in the original constraint (for example, replacing the value of the coefficient of a variable in the affine function) must be transmitted to the constraint created by the bridge, but with a sign change.</p><pre><code class="language-julia hljs">function modify(
+    model::ModelLike,
+    bridge::SignBridge,
+    change::ScalarCoefficientChange,
+)
+    modify(
+        model,
+        bridge.constraint,
+        ScalarCoefficientChange(change.variable, -change.new_coefficient),
+    )
+    return
+end</code></pre><h3 id="Bridge-deletion"><a class="docs-heading-anchor" href="#Bridge-deletion">Bridge deletion</a><a id="Bridge-deletion-1"></a><a class="docs-heading-anchor-permalink" href="#Bridge-deletion" title="Permalink"></a></h3><p>When a bridge is deleted, the constraints it added must be deleted too.</p><pre><code class="language-julia hljs">function delete(model::ModelLike, bridge::SignBridge)
+    delete(model, bridge.constraint)
+    return
+end</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../mathprogbase/">« Transitioning from MathProgBase</a><a class="docs-footer-nextpage" href="../manipulating_expressions/">Manipulating expressions »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/tutorials/example/index.html b/previews/PR3525/moi/tutorials/example/index.html
new file mode 100644
index 00000000000..02d6f4ce162
--- /dev/null
+++ b/previews/PR3525/moi/tutorials/example/index.html
@@ -0,0 +1,49 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Solving a problem using MathOptInterface · JuMP</title><meta name="title" content="Solving a problem using MathOptInterface · JuMP"/><meta property="og:title" content="Solving a problem using MathOptInterface · JuMP"/><meta property="twitter:title" content="Solving a problem using MathOptInterface · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Solving a problem using MathOptInterface</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Define-the-data"><span>Define the data</span></a></li><li><a class="tocitem" href="#Add-the-variables"><span>Add the variables</span></a></li><li><a class="tocitem" href="#set_objective_example"><span>Set the objective</span></a></li><li><a class="tocitem" href="#Add-the-constraints"><span>Add the constraints</span></a></li><li><a class="tocitem" href="#Optimize-the-model"><span>Optimize the model</span></a></li><li><a class="tocitem" href="#Understand-why-the-solver-stopped"><span>Understand why the solver stopped</span></a></li><li><a class="tocitem" href="#Understand-what-solution-was-returned"><span>Understand what solution was returned</span></a></li><li><a class="tocitem" href="#Query-the-objective"><span>Query the objective</span></a></li><li><a class="tocitem" href="#Query-the-primal-solution"><span>Query the primal solution</span></a></li></ul></li><li><a class="tocitem" href="../implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Tutorials</a></li><li class="is-active"><a href>Solving a problem using MathOptInterface</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Solving a problem using MathOptInterface</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/example.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Solving-a-problem-using-MathOptInterface"><a class="docs-heading-anchor" href="#Solving-a-problem-using-MathOptInterface">Solving a problem using MathOptInterface</a><a id="Solving-a-problem-using-MathOptInterface-1"></a><a class="docs-heading-anchor-permalink" href="#Solving-a-problem-using-MathOptInterface" title="Permalink"></a></h1><p>In this tutorial we demonstrate how to use MathOptInterface to solve the binary-constrained knapsack problem:</p><p class="math-container">\[\begin{aligned}
+\max \; &amp; c^\top x       \\
+s.t. \; &amp; w^\top x \le C \\
+        &amp; x_i \in \{0,1\},\quad \forall i=1,\ldots,n
+\end{aligned}\]</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>Load the MathOptInterface module and define the shorthand <code>MOI</code>:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre><p>As an optimizer, we choose GLPK:</p><pre><code class="language-julia hljs">using GLPK
+optimizer = GLPK.Optimizer()</code></pre><h2 id="Define-the-data"><a class="docs-heading-anchor" href="#Define-the-data">Define the data</a><a id="Define-the-data-1"></a><a class="docs-heading-anchor-permalink" href="#Define-the-data" title="Permalink"></a></h2><p>We first define the constants of the problem:</p><pre><code class="language-julia-repl hljs">julia&gt; c = [1.0, 2.0, 3.0]
+3-element Vector{Float64}:
+ 1.0
+ 2.0
+ 3.0
+
+julia&gt; w = [0.3, 0.5, 1.0]
+3-element Vector{Float64}:
+ 0.3
+ 0.5
+ 1.0
+
+julia&gt; C = 3.2
+3.2</code></pre><h2 id="Add-the-variables"><a class="docs-heading-anchor" href="#Add-the-variables">Add the variables</a><a id="Add-the-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Add-the-variables" title="Permalink"></a></h2><pre><code class="language-julia-repl hljs">julia&gt; x = MOI.add_variables(optimizer, length(c));</code></pre><h2 id="set_objective_example"><a class="docs-heading-anchor" href="#set_objective_example">Set the objective</a><a id="set_objective_example-1"></a><a class="docs-heading-anchor-permalink" href="#set_objective_example" title="Permalink"></a></h2><pre><code class="language-julia-repl hljs">julia&gt; MOI.set(
+           optimizer,
+           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),
+           MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(c, x), 0.0),
+       );
+
+julia&gt; MOI.set(optimizer, MOI.ObjectiveSense(), MOI.MAX_SENSE)</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p><code>MOI.ScalarAffineTerm.(c, x)</code> is a shortcut for <code>[MOI.ScalarAffineTerm(c[i], x[i]) for i = 1:3]</code>. This is Julia&#39;s broadcast syntax in action, and is used quite often throughout MOI.</p></div></div><h2 id="Add-the-constraints"><a class="docs-heading-anchor" href="#Add-the-constraints">Add the constraints</a><a id="Add-the-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Add-the-constraints" title="Permalink"></a></h2><p>We add the knapsack constraint and integrality constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.add_constraint(
+           optimizer,
+           MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(w, x), 0.0),
+           MOI.LessThan(C),
+       );</code></pre><p>Add integrality constraints:</p><pre><code class="language-julia-repl hljs">julia&gt; for x_i in x
+           MOI.add_constraint(optimizer, x_i, MOI.ZeroOne())
+       end</code></pre><h2 id="Optimize-the-model"><a class="docs-heading-anchor" href="#Optimize-the-model">Optimize the model</a><a id="Optimize-the-model-1"></a><a class="docs-heading-anchor-permalink" href="#Optimize-the-model" title="Permalink"></a></h2><pre><code class="language-julia-repl hljs">julia&gt; MOI.optimize!(optimizer)</code></pre><h2 id="Understand-why-the-solver-stopped"><a class="docs-heading-anchor" href="#Understand-why-the-solver-stopped">Understand why the solver stopped</a><a id="Understand-why-the-solver-stopped-1"></a><a class="docs-heading-anchor-permalink" href="#Understand-why-the-solver-stopped" title="Permalink"></a></h2><p>The first thing to check after optimization is why the solver stopped, for example, did it stop because of a time limit or did it stop because it found the optimal solution?</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.get(optimizer, MOI.TerminationStatus())
+OPTIMAL::TerminationStatusCode = 1</code></pre><p>Looks like we found an optimal solution.</p><h2 id="Understand-what-solution-was-returned"><a class="docs-heading-anchor" href="#Understand-what-solution-was-returned">Understand what solution was returned</a><a id="Understand-what-solution-was-returned-1"></a><a class="docs-heading-anchor-permalink" href="#Understand-what-solution-was-returned" title="Permalink"></a></h2><pre><code class="language-julia-repl hljs">julia&gt; MOI.get(optimizer, MOI.ResultCount())
+1
+
+julia&gt; MOI.get(optimizer, MOI.PrimalStatus())
+FEASIBLE_POINT::ResultStatusCode = 1
+
+julia&gt; MOI.get(optimizer, MOI.DualStatus())
+NO_SOLUTION::ResultStatusCode = 0</code></pre><h2 id="Query-the-objective"><a class="docs-heading-anchor" href="#Query-the-objective">Query the objective</a><a id="Query-the-objective-1"></a><a class="docs-heading-anchor-permalink" href="#Query-the-objective" title="Permalink"></a></h2><p>What is its objective value?</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.get(optimizer, MOI.ObjectiveValue())
+6.0</code></pre><h2 id="Query-the-primal-solution"><a class="docs-heading-anchor" href="#Query-the-primal-solution">Query the primal solution</a><a id="Query-the-primal-solution-1"></a><a class="docs-heading-anchor-permalink" href="#Query-the-primal-solution" title="Permalink"></a></h2><p>And what is the value of the variables <code>x</code>?</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.get(optimizer, MOI.VariablePrimal(), x)
+3-element Vector{Float64}:
+ 1.0
+ 1.0
+ 1.0</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../background/motivation/">« Motivation</a><a class="docs-footer-nextpage" href="../implementing/">Implementing a solver interface »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/tutorials/implementing/index.html b/previews/PR3525/moi/tutorials/implementing/index.html
new file mode 100644
index 00000000000..a3347cb7d79
--- /dev/null
+++ b/previews/PR3525/moi/tutorials/implementing/index.html
@@ -0,0 +1,118 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Implementing a solver interface · JuMP</title><meta name="title" content="Implementing a solver interface · JuMP"/><meta property="og:title" content="Implementing a solver interface · JuMP"/><meta property="twitter:title" content="Implementing a solver interface · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../example/">Solving a problem using MathOptInterface</a></li><li class="is-active"><a class="tocitem" href>Implementing a solver interface</a><ul class="internal"><li><a class="tocitem" href="#A-note-on-the-API"><span>A note on the API</span></a></li><li><a class="tocitem" href="#Preliminaries"><span>Preliminaries</span></a></li><li><a class="tocitem" href="#Structuring-the-package"><span>Structuring the package</span></a></li><li><a class="tocitem" href="#Initial-code"><span>Initial code</span></a></li><li><a class="tocitem" href="#The-big-decision:-incremental-modification?"><span>The big decision: incremental modification?</span></a></li><li><a class="tocitem" href="#implement_names"><span>Names</span></a></li><li><a class="tocitem" href="#Solutions"><span>Solutions</span></a></li><li><a class="tocitem" href="#Other-tips"><span>Other tips</span></a></li></ul></li><li><a class="tocitem" href="../mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Tutorials</a></li><li class="is-active"><a href>Implementing a solver interface</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Implementing a solver interface</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/implementing.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Implementing-a-solver-interface"><a class="docs-heading-anchor" href="#Implementing-a-solver-interface">Implementing a solver interface</a><a id="Implementing-a-solver-interface-1"></a><a class="docs-heading-anchor-permalink" href="#Implementing-a-solver-interface" title="Permalink"></a></h1><p>This guide outlines the basic steps to implement an interface to MathOptInterface for a new solver.</p><div class="admonition is-danger"><header class="admonition-header">Danger</header><div class="admonition-body"><p>Implementing an interface to MathOptInterface for a new solver is a lot of work. Before starting, we recommend that you join the <a href="https://jump.dev/chatroom/">Developer chatroom</a> and explain a little bit about the solver you are wrapping. If you have questions that are not answered by this guide, please ask them in the <a href="https://jump.dev/chatroom/">Developer chatroom</a> so we can improve this guide.</p></div></div><h2 id="A-note-on-the-API"><a class="docs-heading-anchor" href="#A-note-on-the-API">A note on the API</a><a id="A-note-on-the-API-1"></a><a class="docs-heading-anchor-permalink" href="#A-note-on-the-API" title="Permalink"></a></h2><p>The API of MathOptInterface is large and varied. In order to support the diversity of solvers and use-cases, we make heavy use of <a href="https://en.wikipedia.org/wiki/Duck_typing">duck-typing</a>. That is, solvers are not expected to implement the full API, nor is there a well-defined minimal subset of what must be implemented. Instead, you should implement the API as necessary to make the solver function as you require.</p><p>The main reason for using duck-typing is that solvers work in different ways and target different use-cases.</p><p>For example:</p><ul><li>Some solvers support incremental problem construction, support modification after a solve, and have native support for things like variable names.</li><li>Other solvers are &quot;one-shot&quot; solvers that require all of the problem data to construct and solve the problem in a single function call. They do not support modification or things like variable names.</li><li>Other &quot;solvers&quot; are not solvers at all, but things like file readers. These may only support functions like <a href="../../reference/models/#MathOptInterface.read_from_file"><code>read_from_file</code></a>, and may not even support the ability to add variables or constraints directly.</li><li>Finally, some &quot;solvers&quot; are layers which take a problem as input, transform it according to some rules, and pass the transformed problem to an inner solver.</li></ul><h2 id="Preliminaries"><a class="docs-heading-anchor" href="#Preliminaries">Preliminaries</a><a id="Preliminaries-1"></a><a class="docs-heading-anchor-permalink" href="#Preliminaries" title="Permalink"></a></h2><p>Before starting on your wrapper, you should do some background research and make the solver accessible via Julia.</p><h3 id="Decide-if-MathOptInterface-is-right-for-you"><a class="docs-heading-anchor" href="#Decide-if-MathOptInterface-is-right-for-you">Decide if MathOptInterface is right for you</a><a id="Decide-if-MathOptInterface-is-right-for-you-1"></a><a class="docs-heading-anchor-permalink" href="#Decide-if-MathOptInterface-is-right-for-you" title="Permalink"></a></h3><p>The first step in writing a wrapper is to decide whether implementing an interface is the right thing to do.</p><p>MathOptInterface is an abstraction layer for unifying <em>constrained</em> mathematical optimization solvers. If your solver doesn&#39;t fit in the category, for example, it implements a derivative-free algorithm for unconstrained objective functions, MathOptInterface may not be the right tool for the job.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you&#39;re not sure whether you should write an interface, ask in the <a href="https://jump.dev/chatroom/">Developer chatroom</a>.</p></div></div><h3 id="Find-a-similar-solver-already-wrapped"><a class="docs-heading-anchor" href="#Find-a-similar-solver-already-wrapped">Find a similar solver already wrapped</a><a id="Find-a-similar-solver-already-wrapped-1"></a><a class="docs-heading-anchor-permalink" href="#Find-a-similar-solver-already-wrapped" title="Permalink"></a></h3><p>The next step is to find (if possible) a similar solver that is already wrapped. Although not strictly necessary, this will be a good place to look for inspiration when implementing your wrapper.</p><p>The <a href="https://jump.dev/JuMP.jl/stable/installation/#Supported-solvers">JuMP documentation</a> has a good list of solvers, along with the problem classes they support.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you&#39;re not sure which solver is most similar, ask in the <a href="https://jump.dev/chatroom/">Developer chatroom</a>.</p></div></div><h3 id="Create-a-low-level-interface"><a class="docs-heading-anchor" href="#Create-a-low-level-interface">Create a low-level interface</a><a id="Create-a-low-level-interface-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-low-level-interface" title="Permalink"></a></h3><p>Before writing a MathOptInterface wrapper, you first need to be able to call the solver from Julia.</p><h4 id="Wrapping-solvers-written-in-Julia"><a class="docs-heading-anchor" href="#Wrapping-solvers-written-in-Julia">Wrapping solvers written in Julia</a><a id="Wrapping-solvers-written-in-Julia-1"></a><a class="docs-heading-anchor-permalink" href="#Wrapping-solvers-written-in-Julia" title="Permalink"></a></h4><p>If your solver is written in Julia, there&#39;s nothing to do here. Go to the next section.</p><h4 id="Wrapping-solvers-written-in-C"><a class="docs-heading-anchor" href="#Wrapping-solvers-written-in-C">Wrapping solvers written in C</a><a id="Wrapping-solvers-written-in-C-1"></a><a class="docs-heading-anchor-permalink" href="#Wrapping-solvers-written-in-C" title="Permalink"></a></h4><p>Julia is well suited to wrapping solvers written in C.</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This is <em>not</em> true for C++. If you have a solver written in C++, first write a C interface, then wrap the C interface.</p></div></div><p>Before writing a MathOptInterface wrapper, there are a few extra steps.</p><h5 id="Create-a-JLL"><a class="docs-heading-anchor" href="#Create-a-JLL">Create a JLL</a><a id="Create-a-JLL-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-JLL" title="Permalink"></a></h5><p>If the C code is publicly available under an open source license, create a JLL package via <a href="https://github.com/JuliaPackaging/Yggdrasil">Yggdrasil</a>. The easiest way to do this is to copy an existing solver. Good examples to follow are the <a href="https://github.com/JuliaPackaging/Yggdrasil/tree/master/C/Coin-OR">COIN-OR solvers</a>.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Building the solver via Yggdrasil is non-trivial. Please ask the <a href="https://jump.dev/chatroom/">Developer chatroom</a> for help.</p></div></div><p>If the code is commercial or not publicly available, the user will need to manually install the solver. See <a href="https://github.com/jump-dev/Gurobi.jl">Gurobi.jl</a> or <a href="https://github.com/jump-dev/CPLEX.jl">CPLEX.jl</a> for examples of how to structure this.</p><h5 id="implement_clang"><a class="docs-heading-anchor" href="#implement_clang">Use Clang.jl to wrap the C API</a><a id="implement_clang-1"></a><a class="docs-heading-anchor-permalink" href="#implement_clang" title="Permalink"></a></h5><p>The next step is to use <a href="https://github.com/JuliaInterop/Clang.jl">Clang.jl</a> to automatically wrap the C API. The easiest way to do this is to follow an example. Good examples to follow are <a href="https://github.com/jump-dev/Cbc.jl/blob/master/scripts/clang.jl">Cbc.jl</a> and <a href="https://github.com/jump-dev/HiGHS.jl/blob/master/gen/gen.jl">HiGHS.jl</a>.</p><p>Sometimes, you will need to make manual modifications to the resulting files.</p><h4 id="Solvers-written-in-other-languages"><a class="docs-heading-anchor" href="#Solvers-written-in-other-languages">Solvers written in other languages</a><a id="Solvers-written-in-other-languages-1"></a><a class="docs-heading-anchor-permalink" href="#Solvers-written-in-other-languages" title="Permalink"></a></h4><p>Ask the <a href="https://jump.dev/chatroom/">Developer chatroom</a> for advice. You may be able to use one of the <a href="https://github.com/JuliaInterop">JuliaInterop</a> packages to call out to the solver.</p><p>For example, <a href="https://github.com/jump-dev/SeDuMi.jl">SeDuMi.jl</a> uses <a href="https://github.com/JuliaInterop/MATLAB.jl">MATLAB.jl</a> to call the SeDuMi solver written in MATLAB.</p><h2 id="Structuring-the-package"><a class="docs-heading-anchor" href="#Structuring-the-package">Structuring the package</a><a id="Structuring-the-package-1"></a><a class="docs-heading-anchor-permalink" href="#Structuring-the-package" title="Permalink"></a></h2><p>Structure your wrapper as a Julia package. Consult the <a href="https://pkgdocs.julialang.org/v1/creating-packages/">Julia documentation</a> if you haven&#39;t done this before.</p><p>MOI solver interfaces may be in the same package as the solver itself (either the C wrapper if the solver is accessible through C, or the Julia code if the solver is written in Julia, for example), or in a separate package which depends on the solver package.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The JuMP <a href="https://jump.dev/pages/governance/#core-contributors">core contributors</a> request that you do not use &quot;JuMP&quot; in the name of your package without prior consent.</p></div></div><p>Your package should have the following structure:</p><pre><code class="nohighlight hljs">/.github
+    /workflows
+        ci.yml
+        format_check.yml
+        TagBot.yml
+/gen
+    gen.jl  # Code to wrap the C API
+/src
+    NewSolver.jl
+    /gen
+        libnewsolver_api.jl
+        libnewsolver_common.jl
+    /MOI_wrapper
+        MOI_wrapper.jl
+        other_files.jl
+/test
+    runtests.jl
+    /MOI_wrapper
+        MOI_wrapper.jl
+.gitignore
+.JuliaFormatter.toml
+README.md
+LICENSE.md
+Project.toml</code></pre><ul><li>The <code>/.github</code> folder contains the scripts for GitHub actions. The easiest way to write these is to copy the ones from an existing solver.</li><li>The <code>/gen</code> and <code>/src/gen</code> folders are only needed if you are wrapping a <a href="#implement_clang">solver written in C</a>.</li><li>The <code>/src/MOI_wrapper</code> folder contains the Julia code for the MOI wrapper.</li><li>The <code>/test</code> folder contains code for testing your package. See <a href="#Setup-tests">Setup tests</a> for more information.</li><li>The <code>.JuliaFormatter.toml</code> and <code>.github/workflows/format_check.yml</code> enforce code formatting using <a href="https://github.com/domluna/JuliaFormatter.jl">JuliaFormatter.jl</a>. Check existing solvers or JuMP.jl for details.</li></ul><h3 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h3><p>Your package must include documentation explaining how to use the package. The easiest approach is to include documentation in your <code>README.md</code>. A more involved option is to use <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a>.</p><p>Examples of packages with README-based documentation include:</p><ul><li><a href="https://github.com/jump-dev/Cbc.jl">Cbc.jl</a></li><li><a href="https://github.com/jump-dev/HiGHS.jl">HiGHS.jl</a></li><li><a href="https://github.com/jump-dev/SCS.jl">SCS.jl</a></li></ul><p>Examples of packages with Documenter-based documentation include:</p><ul><li><a href="https://github.com/lanl-ansi/Alpine.jl">Alpine.jl</a></li><li><a href="https://github.com/oxfordcontrol/COSMO.jl">COSMO.jl</a></li><li><a href="https://github.com/lanl-ansi/Juniper.jl">Juniper.jl</a></li></ul><h3 id="Setup-tests"><a class="docs-heading-anchor" href="#Setup-tests">Setup tests</a><a id="Setup-tests-1"></a><a class="docs-heading-anchor-permalink" href="#Setup-tests" title="Permalink"></a></h3><p>The best way to implement an interface to MathOptInterface is via <a href="https://en.wikipedia.org/wiki/Test-driven_development">test-driven development</a>.</p><p>The <a href="../../submodules/Test/overview/#test_module"><code>MOI.Test</code> submodule</a> contains a large test suite to help check that you have implemented things correctly.</p><p>Follow the guide <a href="../../submodules/Test/overview/#How-to-test-a-solver">How to test a solver</a> to set up the tests for your package.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Run the tests frequently when developing. However, at the start there is going to be a lot of errors. Start by excluding large classes of tests (for example, <code>exclude = [&quot;test_basic_&quot;, &quot;test_model_&quot;]</code>, implement any missing methods until the tests pass, then remove an exclusion and repeat.</p></div></div><h2 id="Initial-code"><a class="docs-heading-anchor" href="#Initial-code">Initial code</a><a id="Initial-code-1"></a><a class="docs-heading-anchor-permalink" href="#Initial-code" title="Permalink"></a></h2><p>By this point, you should have a package setup with tests, formatting, and access to the underlying solver. Now it&#39;s time to start writing the wrapper.</p><h3 id="The-Optimizer-object"><a class="docs-heading-anchor" href="#The-Optimizer-object">The <code>Optimizer</code> object</a><a id="The-Optimizer-object-1"></a><a class="docs-heading-anchor-permalink" href="#The-Optimizer-object" title="Permalink"></a></h3><p>The first object to create is a subtype of <a href="../../reference/models/#MathOptInterface.AbstractOptimizer"><code>AbstractOptimizer</code></a>. This type is going to store everything related to the problem.</p><p>By convention, these optimizers should not be exported and should be named <code>PackageName.Optimizer</code>.</p><pre><code class="language-julia hljs">import MathOptInterface as MOI
+
+struct Optimizer &lt;: MOI.AbstractOptimizer
+    # Fields go here
+end</code></pre><h4 id="Optimizer-objects-for-C-solvers"><a class="docs-heading-anchor" href="#Optimizer-objects-for-C-solvers">Optimizer objects for C solvers</a><a id="Optimizer-objects-for-C-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Optimizer-objects-for-C-solvers" title="Permalink"></a></h4><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This section is important if you wrap a solver written in C.</p></div></div><p>Wrapping a solver written in C will require the use of pointers, and for you to manually free the solver&#39;s memory when the <code>Optimizer</code> is garbage collected by Julia.</p><p><strong>Never pass a pointer directly to a Julia <code>ccall</code> function.</strong></p><p>Instead, store the pointer as a field in your <code>Optimizer</code>, and implement <code>Base.cconvert</code> and <code>Base.unsafe_convert</code>. Then you can pass <code>Optimizer</code> to any <code>ccall</code> function that expects the pointer.</p><p>In addition, make sure you implement a <code>finalizer</code> for each model you create.</p><p>If <code>newsolver_createProblem()</code> is the low-level function that creates the problem pointer in C, and <code>newsolver_freeProblem(::Ptr{Cvoid})</code> is the low-level function that frees memory associated with the pointer, your <code>Optimizer()</code> function should look like this:</p><pre><code class="language-julia hljs">struct Optimizer &lt;: MOI.AbstractOptimizer
+    ptr::Ptr{Cvoid}
+
+    function Optimizer()
+        ptr = newsolver_createProblem()
+        model = Optimizer(ptr)
+        finalizer(model) do m
+            newsolver_freeProblem(m)
+            return
+        end
+        return model
+    end
+end
+
+Base.cconvert(::Type{Ptr{Cvoid}}, model::Optimizer) = model
+Base.unsafe_convert(::Type{Ptr{Cvoid}}, model::Optimizer) = model.ptr</code></pre><h3 id="Implement-methods-for-Optimizer"><a class="docs-heading-anchor" href="#Implement-methods-for-Optimizer">Implement methods for <code>Optimizer</code></a><a id="Implement-methods-for-Optimizer-1"></a><a class="docs-heading-anchor-permalink" href="#Implement-methods-for-Optimizer" title="Permalink"></a></h3><p>All <code>Optimizer</code>s must implement the following methods:</p><ul><li><a href="../../reference/models/#MathOptInterface.empty!"><code>empty!</code></a></li><li><a href="../../reference/models/#MathOptInterface.is_empty"><code>is_empty</code></a></li></ul><p>Other methods, detailed below, are optional or depend on how you implement the interface.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>For this and all future methods, read the docstrings to understand what each method does, what it expects as input, and what it produces as output. If it isn&#39;t clear, let us know and we will improve the docstrings. It is also very helpful to look at an existing wrapper for a similar solver.</p></div></div><p>You should also implement <code>Base.show(::IO, ::Optimizer)</code> to print a nice string when someone prints your model. For example</p><pre><code class="language-julia hljs">function Base.show(io::IO, model::Optimizer)
+    return print(io, &quot;NewSolver with the pointer $(model.ptr)&quot;)
+end</code></pre><h3 id="Implement-attributes"><a class="docs-heading-anchor" href="#Implement-attributes">Implement attributes</a><a id="Implement-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Implement-attributes" title="Permalink"></a></h3><p>MathOptInterface uses attributes to manage different aspects of the problem.</p><p>For each attribute</p><ul><li><a href="../../reference/models/#MathOptInterface.get"><code>get</code></a> gets the current value of the attribute</li><li><a href="../../reference/models/#MathOptInterface.set"><code>set</code></a> sets a new value of the attribute. Not all attributes can be set. For example, the user can&#39;t modify the <a href="../../reference/models/#MathOptInterface.SolverName"><code>SolverName</code></a>.</li><li><a href="../../reference/models/#MathOptInterface.supports"><code>supports</code></a> returns a <code>Bool</code> indicating whether the solver supports the attribute.</li></ul><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Use <a href="../../reference/models/#MathOptInterface.attribute_value_type"><code>attribute_value_type</code></a> to check the value expected by a given attribute. You should make sure that your <a href="../../reference/models/#MathOptInterface.get"><code>get</code></a> function correctly infers to this type (or a subtype of it).</p></div></div><p>Each column in the table indicates whether you need to implement the particular method for each attribute.</p><table><tr><th style="text-align: right">Attribute</th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.get"><code>get</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.set"><code>set</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.supports"><code>supports</code></a></th></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.SolverName"><code>SolverName</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.SolverVersion"><code>SolverVersion</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.RawSolver"><code>RawSolver</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.Name"><code>Name</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.Silent"><code>Silent</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.TimeLimitSec"><code>TimeLimitSec</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ObjectiveLimit"><code>ObjectiveLimit</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.RawOptimizerAttribute"><code>RawOptimizerAttribute</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.NumberOfThreads"><code>NumberOfThreads</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.AbsoluteGapTolerance"><code>AbsoluteGapTolerance</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.RelativeGapTolerance"><code>RelativeGapTolerance</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr></table><p>For example:</p><pre><code class="language-julia hljs">function MOI.get(model::Optimizer, ::MOI.Silent)
+    return # true if MOI.Silent is set
+end
+
+function MOI.set(model::Optimizer, ::MOI.Silent, v::Bool)
+    if v
+        # Set a parameter to turn off printing
+    else
+        # Restore the default printing
+    end
+    return
+end
+
+MOI.supports(::Optimizer, ::MOI.Silent) = true</code></pre><h3 id="Define-supports_constraint"><a class="docs-heading-anchor" href="#Define-supports_constraint">Define <code>supports_constraint</code></a><a id="Define-supports_constraint-1"></a><a class="docs-heading-anchor-permalink" href="#Define-supports_constraint" title="Permalink"></a></h3><p>The next step is to define which constraints and objective functions you plan to support.</p><p>For each function-set constraint pair, define <a href="../../reference/constraints/#MathOptInterface.supports_constraint"><code>supports_constraint</code></a>:</p><pre><code class="language-julia hljs">function MOI.supports_constraint(
+    ::Optimizer,
+    ::Type{MOI.VariableIndex},
+    ::Type{MOI.ZeroOne},
+)
+    return true
+end</code></pre><p>To make this easier, you may want to use <code>Union</code>s:</p><pre><code class="language-julia hljs">function MOI.supports_constraint(
+    ::Optimizer,
+    ::Type{MOI.VariableIndex},
+    ::Type{&lt;:Union{MOI.LessThan,MOI.GreaterThan,MOI.EqualTo}},
+)
+    return true
+end</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Only support a constraint if your solver has native support for it.</p></div></div><h2 id="The-big-decision:-incremental-modification?"><a class="docs-heading-anchor" href="#The-big-decision:-incremental-modification?">The big decision: incremental modification?</a><a id="The-big-decision:-incremental-modification?-1"></a><a class="docs-heading-anchor-permalink" href="#The-big-decision:-incremental-modification?" title="Permalink"></a></h2><p>Now you need to decide whether to support incremental modification or not.</p><p>Incremental modification means that the user can add variables and constraints one-by-one without needing to rebuild the entire problem, and they can modify the problem data after an <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a> call. Supporting incremental modification means implementing functions like <a href="../../reference/variables/#MathOptInterface.add_variable"><code>add_variable</code></a> and <a href="../../reference/constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a>.</p><p>The alternative is to accept the problem data in a single <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a> or <a href="../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a> function call. Because these functions see all of the data at once, it can typically call a more efficient function to load data into the underlying solver.</p><p>Good examples of solvers supporting incremental modification are MILP solvers like <a href="https://github.com/jump-dev/GLPK.jl">GLPK.jl</a> and <a href="https://github.com/jump-dev/Gurobi.jl">Gurobi.jl</a>. Examples of non-incremental solvers are <a href="https://github.com/jump-dev/AmplNLWriter.jl">AmplNLWriter.jl</a> and <a href="https://github.com/jump-dev/SCS.jl">SCS.jl</a></p><p>It is possible for a solver to implement both approaches, but you should probably start with one for simplicity.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Only support incremental modification if your solver has native support for it.</p></div></div><p>In general, supporting incremental modification is more work, and it usually requires some extra book-keeping. However, it provides a more efficient interface to the solver if the problem is going to be resolved multiple times with small modifications. Moreover, once you&#39;ve implemented incremental modification, it&#39;s usually not much extra work to add a <a href="../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a> interface. The converse is not true.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If this is your first time writing an interface, start with the one-shot <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a>.</p></div></div><h3 id="The-non-incremental-interface"><a class="docs-heading-anchor" href="#The-non-incremental-interface">The non-incremental interface</a><a id="The-non-incremental-interface-1"></a><a class="docs-heading-anchor-permalink" href="#The-non-incremental-interface" title="Permalink"></a></h3><p>There are two ways to implement the non-incremental interface. The first uses a two-argument version of <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a>, the second implements <a href="../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a> followed by the one-argument version of <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a>.</p><p>If your solver does not support modification, and requires all data to solve the problem in a single function call, you should implement the &quot;one-shot&quot; <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a>.</p><ul><li><a href="../../reference/models/#MathOptInterface.optimize!-Tuple{MathOptInterface.ModelLike, MathOptInterface.ModelLike}"><code>optimize!(::ModelLike, ::ModelLike)</code></a></li></ul><p>If your solver separates data loading and the actual optimization into separate steps, implement the <a href="../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a> interface.</p><ul><li><a href="../../reference/models/#MathOptInterface.copy_to"><code>copy_to(::ModelLike, ::ModelLike)</code></a></li><li><a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!(::ModelLike)</code></a></li></ul><h3 id="The-incremental-interface"><a class="docs-heading-anchor" href="#The-incremental-interface">The incremental interface</a><a id="The-incremental-interface-1"></a><a class="docs-heading-anchor-permalink" href="#The-incremental-interface" title="Permalink"></a></h3><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Writing this interface is a lot of work. The easiest way is to consult the source code of a similar solver.</p></div></div><p>To implement the incremental interface, implement the following functions:</p><ul><li><a href="../../reference/variables/#MathOptInterface.add_variable"><code>add_variable</code></a></li><li><a href="../../reference/variables/#MathOptInterface.add_variables"><code>add_variables</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.add_constraint"><code>add_constraint</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.add_constraints"><code>add_constraints</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.ConstraintIndex}"><code>is_valid</code></a></li><li><a href="../../reference/variables/#MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}"><code>delete</code></a></li><li><a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!(::ModelLike)</code></a></li></ul><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Solvers do not have to support <code>AbstractScalarFunction</code> in <code>GreaterThan</code>, <code>LessThan</code>, <code>EqualTo</code>, or <code>Interval</code> with a nonzero constant in the function. Throw <a href="../../reference/errors/#MathOptInterface.ScalarFunctionConstantNotZero"><code>ScalarFunctionConstantNotZero</code></a> if the function constant is not zero.</p></div></div><p>In addition, you should implement the following model attributes:</p><table><tr><th style="text-align: right">Attribute</th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.get"><code>get</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.set"><code>set</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.supports"><code>supports</code></a></th></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ListOfModelAttributesSet"><code>ListOfModelAttributesSet</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ObjectiveFunctionType"><code>ObjectiveFunctionType</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ObjectiveSense"><code>ObjectiveSense</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.Name"><code>Name</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr></table><p>Variable-related attributes:</p><table><tr><th style="text-align: right">Attribute</th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.get"><code>get</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.set"><code>set</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.supports"><code>supports</code></a></th></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ListOfVariableAttributesSet"><code>ListOfVariableAttributesSet</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.NumberOfVariables"><code>NumberOfVariables</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ListOfVariableIndices"><code>ListOfVariableIndices</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr></table><p>Constraint-related attributes:</p><table><tr><th style="text-align: right">Attribute</th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.get"><code>get</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.set"><code>set</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.supports"><code>supports</code></a></th></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ListOfConstraintAttributesSet"><code>ListOfConstraintAttributesSet</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.NumberOfConstraints"><code>NumberOfConstraints</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/models/#MathOptInterface.ListOfConstraintTypesPresent"><code>ListOfConstraintTypesPresent</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">No</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">No</td></tr><tr><td style="text-align: right"><a href="../../reference/constraints/#MathOptInterface.ConstraintSet"><code>ConstraintSet</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">No</td></tr></table><h4 id="Modifications"><a class="docs-heading-anchor" href="#Modifications">Modifications</a><a id="Modifications-1"></a><a class="docs-heading-anchor-permalink" href="#Modifications" title="Permalink"></a></h4><p>If your solver supports modifying data in-place, implement <a href="../../reference/modification/#MathOptInterface.modify"><code>modify</code></a> for the following <code>AbstractModification</code>s:</p><ul><li><a href="../../reference/modification/#MathOptInterface.ScalarConstantChange"><code>ScalarConstantChange</code></a></li><li><a href="../../reference/modification/#MathOptInterface.ScalarCoefficientChange"><code>ScalarCoefficientChange</code></a></li><li><a href="../../reference/modification/#MathOptInterface.VectorConstantChange"><code>VectorConstantChange</code></a></li><li><a href="../../reference/modification/#MathOptInterface.MultirowChange"><code>MultirowChange</code></a></li></ul><h4 id="Variables-constrained-on-creation"><a class="docs-heading-anchor" href="#Variables-constrained-on-creation">Variables constrained on creation</a><a id="Variables-constrained-on-creation-1"></a><a class="docs-heading-anchor-permalink" href="#Variables-constrained-on-creation" title="Permalink"></a></h4><p>Some solvers require variables be associated with a set <em>when they are created</em>. This conflicts with the incremental modification approach, since you cannot first add a free variable and then constrain it to the set.</p><p>If this is the case, implement:</p><ul><li><a href="../../reference/variables/#MathOptInterface.add_constrained_variable"><code>add_constrained_variable</code></a></li><li><a href="../../reference/variables/#MathOptInterface.add_constrained_variables"><code>add_constrained_variables</code></a></li><li><a href="../../reference/variables/#MathOptInterface.supports_add_constrained_variables"><code>supports_add_constrained_variables</code></a></li></ul><p>By default, MathOptInterface assumes solvers support free variables. If your solver does not support free variables, define:</p><pre><code class="language-julia hljs">MOI.supports_add_constrained_variables(::Optimizer, ::Type{Reals}) = false</code></pre><h3 id="Incremental-and-copy_to"><a class="docs-heading-anchor" href="#Incremental-and-copy_to">Incremental and <code>copy_to</code></a><a id="Incremental-and-copy_to-1"></a><a class="docs-heading-anchor-permalink" href="#Incremental-and-copy_to" title="Permalink"></a></h3><p>If you implement the incremental interface, you have the option of also implementing <a href="../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a>.</p><p>If you don&#39;t want to implement <a href="../../reference/models/#MathOptInterface.copy_to"><code>copy_to</code></a>, for example, because the solver has no API for building the problem in a single function call, define the following fallback:</p><pre><code class="language-julia hljs">MOI.supports_incremental_interface(::Optimizer) = true
+
+function MOI.copy_to(dest::Optimizer, src::MOI.ModelLike)
+    return MOI.Utilities.default_copy_to(dest, src)
+end</code></pre><h2 id="implement_names"><a class="docs-heading-anchor" href="#implement_names">Names</a><a id="implement_names-1"></a><a class="docs-heading-anchor-permalink" href="#implement_names" title="Permalink"></a></h2><p>Regardless of which interface you implement, you have the option of implementing the <code>Name</code> attribute for variables and constraints:</p><table><tr><th style="text-align: right">Attribute</th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.get"><code>get</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.set"><code>set</code></a></th><th style="text-align: right"><a href="../../reference/models/#MathOptInterface.supports"><code>supports</code></a></th></tr><tr><td style="text-align: right"><a href="../../reference/variables/#MathOptInterface.VariableName"><code>VariableName</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr><tr><td style="text-align: right"><a href="../../reference/constraints/#MathOptInterface.ConstraintName"><code>ConstraintName</code></a></td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td><td style="text-align: right">Yes</td></tr></table><p>If you implement names, you must also implement the following three methods:</p><pre><code class="language-julia hljs">function MOI.get(model::Optimizer, ::Type{MOI.VariableIndex}, name::String)
+    return # The variable named `name`.
+end
+
+function MOI.get(model::Optimizer, ::Type{MOI.ConstraintIndex}, name::String)
+    return # The constraint any type named `name`.
+end
+
+function MOI.get(
+    model::Optimizer,
+    ::Type{MOI.ConstraintIndex{F,S}},
+    name::String,
+) where {F,S}
+    return # The constraint of type F-in-S named `name`.
+end</code></pre><p>These methods have the following rules:</p><ul><li>If there is no variable or constraint with the name, return <code>nothing</code></li><li>If there is a single variable or constraint with that name, return the variable or constraint</li><li>If there are multiple variables or constraints with the name, throw an error.</li></ul><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>You should <em>not</em> implement <a href="../../reference/constraints/#MathOptInterface.ConstraintName"><code>ConstraintName</code></a> for <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraints. If you implement <a href="../../reference/constraints/#MathOptInterface.ConstraintName"><code>ConstraintName</code></a> for other constraints, you can add the following two methods to disable <a href="../../reference/constraints/#MathOptInterface.ConstraintName"><code>ConstraintName</code></a> for <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraints.</p><pre><code class="language-julia hljs">function MOI.supports(
+    ::Optimizer,
+    ::MOI.ConstraintName,
+    ::Type{&lt;:MOI.ConstraintIndex{MOI.VariableIndex,&lt;:MOI.AbstractScalarSet}},
+)
+    return throw(MOI.VariableIndexConstraintNameError())
+end
+function MOI.set(
+    ::Optimizer,
+    ::MOI.ConstraintName,
+    ::MOI.ConstraintIndex{MOI.VariableIndex,&lt;:MOI.AbstractScalarSet},
+    ::String,
+)
+    return throw(MOI.VariableIndexConstraintNameError())
+end</code></pre></div></div><h2 id="Solutions"><a class="docs-heading-anchor" href="#Solutions">Solutions</a><a id="Solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Solutions" title="Permalink"></a></h2><p>Implement <a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a> to solve the model:</p><ul><li><a href="../../reference/models/#MathOptInterface.optimize!"><code>optimize!</code></a></li></ul><p>All <code>Optimizer</code>s must implement the following attributes:</p><ul><li><a href="../../reference/models/#MathOptInterface.DualStatus"><code>DualStatus</code></a></li><li><a href="../../reference/models/#MathOptInterface.PrimalStatus"><code>PrimalStatus</code></a></li><li><a href="../../reference/models/#MathOptInterface.RawStatusString"><code>RawStatusString</code></a></li><li><a href="../../reference/models/#MathOptInterface.ResultCount"><code>ResultCount</code></a></li><li><a href="../../reference/models/#MathOptInterface.TerminationStatus"><code>TerminationStatus</code></a></li></ul><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>You only need to implement <a href="../../reference/models/#MathOptInterface.get"><code>get</code></a> for solution attributes. Don&#39;t implement <a href="../../reference/models/#MathOptInterface.set"><code>set</code></a> or <a href="../../reference/models/#MathOptInterface.supports"><code>supports</code></a>.</p></div></div><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>Solver wrappers should document how the low-level statuses map to the MOI statuses. Statuses like <code>NEARLY_FEASIBLE_POINT</code> and <code>INFEASIBLE_POINT</code>, are designed to be used when the solver explicitly indicates that relaxed tolerances are satisfied or the returned point is infeasible, respectively.</p></div></div><p>You should also implement the following attributes:</p><ul><li><a href="../../reference/models/#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a></li><li><a href="../../reference/models/#MathOptInterface.SolveTimeSec"><code>SolveTimeSec</code></a></li><li><a href="../../reference/variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a></li></ul><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Attributes like <a href="../../reference/variables/#MathOptInterface.VariablePrimal"><code>VariablePrimal</code></a> and <a href="../../reference/models/#MathOptInterface.ObjectiveValue"><code>ObjectiveValue</code></a> are indexed by the result count. Use <code>MOI.check_result_index_bounds(model, attr)</code> to throw an error if the attribute is not available.</p></div></div><p>If your solver returns dual solutions, implement:</p><ul><li><a href="../../reference/constraints/#MathOptInterface.ConstraintDual"><code>ConstraintDual</code></a></li><li><a href="../../reference/models/#MathOptInterface.DualObjectiveValue"><code>DualObjectiveValue</code></a></li></ul><p>For integer solvers, implement:</p><ul><li><a href="../../reference/models/#MathOptInterface.ObjectiveBound"><code>ObjectiveBound</code></a></li><li><a href="../../reference/models/#MathOptInterface.RelativeGap"><code>RelativeGap</code></a></li></ul><p>If applicable, implement:</p><ul><li><a href="../../reference/models/#MathOptInterface.SimplexIterations"><code>SimplexIterations</code></a></li><li><a href="../../reference/models/#MathOptInterface.BarrierIterations"><code>BarrierIterations</code></a></li><li><a href="../../reference/models/#MathOptInterface.NodeCount"><code>NodeCount</code></a></li></ul><p>If your solver uses the Simplex method, implement:</p><ul><li><a href="../../reference/constraints/#MathOptInterface.ConstraintBasisStatus"><code>ConstraintBasisStatus</code></a></li></ul><p>If your solver accepts primal or dual warm-starts, implement:</p><ul><li><a href="../../reference/variables/#MathOptInterface.VariablePrimalStart"><code>VariablePrimalStart</code></a></li><li><a href="../../reference/constraints/#MathOptInterface.ConstraintDualStart"><code>ConstraintDualStart</code></a></li></ul><h2 id="Other-tips"><a class="docs-heading-anchor" href="#Other-tips">Other tips</a><a id="Other-tips-1"></a><a class="docs-heading-anchor-permalink" href="#Other-tips" title="Permalink"></a></h2><p>Here are some other points to be aware of when writing your wrapper.</p><h3 id="Unsupported-constraints-at-runtime"><a class="docs-heading-anchor" href="#Unsupported-constraints-at-runtime">Unsupported constraints at runtime</a><a id="Unsupported-constraints-at-runtime-1"></a><a class="docs-heading-anchor-permalink" href="#Unsupported-constraints-at-runtime" title="Permalink"></a></h3><p>In some cases, your solver may support a particular type of constraint (for example, quadratic constraints), but only if the data meets some condition (for example, it is convex).</p><p>In this case, declare that you <code>support</code> the constraint, and throw <a href="../../reference/errors/#MathOptInterface.AddConstraintNotAllowed"><code>AddConstraintNotAllowed</code></a>.</p><h3 id="Dealing-with-multiple-variable-bounds"><a class="docs-heading-anchor" href="#Dealing-with-multiple-variable-bounds">Dealing with multiple variable bounds</a><a id="Dealing-with-multiple-variable-bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Dealing-with-multiple-variable-bounds" title="Permalink"></a></h3><p>MathOptInterface uses <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> constraints to represent variable bounds. Defining multiple variable bounds on a single variable is not allowed.</p><p>Throw <a href="../../reference/errors/#MathOptInterface.LowerBoundAlreadySet"><code>LowerBoundAlreadySet</code></a> or <a href="../../reference/errors/#MathOptInterface.UpperBoundAlreadySet"><code>UpperBoundAlreadySet</code></a> if the user adds a constraint that results in multiple bounds.</p><p>Only throw if the constraints conflict. It is okay to add <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>-in-<a href="../../reference/standard_form/#MathOptInterface.GreaterThan"><code>GreaterThan</code></a> and then <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>-in-<a href="../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a>, but not <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>-in-<a href="../../reference/standard_form/#MathOptInterface.Interval"><code>Interval</code></a> and then <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a>-in-<a href="../../reference/standard_form/#MathOptInterface.LessThan"><code>LessThan</code></a>,</p><h3 id="Expect-duplicate-coefficients"><a class="docs-heading-anchor" href="#Expect-duplicate-coefficients">Expect duplicate coefficients</a><a id="Expect-duplicate-coefficients-1"></a><a class="docs-heading-anchor-permalink" href="#Expect-duplicate-coefficients" title="Permalink"></a></h3><p>Solvers must expect that functions such as <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> and <a href="../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a> may contain duplicate coefficients.</p><p>For example, <code>ScalarAffineFunction([ScalarAffineTerm(x, 1), ScalarAffineTerm(x, 1)], 0.0)</code>.</p><p>Use <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.canonical"><code>Utilities.canonical</code></a> to return a new function with the duplicate coefficients aggregated together.</p><h3 id="Don&#39;t-modify-user-data"><a class="docs-heading-anchor" href="#Don&#39;t-modify-user-data">Don&#39;t modify user-data</a><a id="Don&#39;t-modify-user-data-1"></a><a class="docs-heading-anchor-permalink" href="#Don&#39;t-modify-user-data" title="Permalink"></a></h3><p>All data passed to the solver must be copied immediately to internal data structures. Solvers may not modify any input vectors and must assume that input vectors will not be modified by users in the future.</p><p>This applies, for example, to the <code>terms</code> vector in <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>. Vectors returned to the user, for example, via <a href="../../reference/models/#MathOptInterface.ObjectiveFunction"><code>ObjectiveFunction</code></a> or <a href="../../reference/constraints/#MathOptInterface.ConstraintFunction"><code>ConstraintFunction</code></a> attributes, must not be modified by the solver afterwards. The in-place version of <a href="../../reference/models/#MathOptInterface.get!"><code>get!</code></a> can be used by users to avoid extra copies in this case.</p><h3 id="Column-Generation"><a class="docs-heading-anchor" href="#Column-Generation">Column Generation</a><a id="Column-Generation-1"></a><a class="docs-heading-anchor-permalink" href="#Column-Generation" title="Permalink"></a></h3><p>There is no special interface for column generation. If the solver has a special API for setting coefficients in existing constraints when adding a new variable, it is possible to queue modifications and new variables and then call the solver&#39;s API once all of the new coefficients are known.</p><h3 id="Solver-specific-attributes"><a class="docs-heading-anchor" href="#Solver-specific-attributes">Solver-specific attributes</a><a id="Solver-specific-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Solver-specific-attributes" title="Permalink"></a></h3><p>You don&#39;t need to restrict yourself to the attributes defined in the MathOptInterface.jl package.</p><p>Solver-specific attributes should be specified by creating an appropriate subtype of <a href="../../reference/models/#MathOptInterface.AbstractModelAttribute"><code>AbstractModelAttribute</code></a>, <a href="../../reference/models/#MathOptInterface.AbstractOptimizerAttribute"><code>AbstractOptimizerAttribute</code></a>, <a href="../../reference/variables/#MathOptInterface.AbstractVariableAttribute"><code>AbstractVariableAttribute</code></a>, or <a href="../../reference/constraints/#MathOptInterface.AbstractConstraintAttribute"><code>AbstractConstraintAttribute</code></a>.</p><p>For example, Gurobi.jl adds attributes for multiobjective optimization by <a href="https://github.com/jump-dev/Gurobi.jl/blob/d9cebe4ec05a102df8917ff2602e6c38abdac090/src/MOI_multi_objective.jl#L1-L15">defining</a>:</p><pre><code class="language-julia hljs">struct NumberOfObjectives &lt;: MOI.AbstractModelAttribute end
+
+function MOI.set(model::Optimizer, ::NumberOfObjectives, n::Integer)
+    # Code to set NumberOfObjectives
+    return
+end
+
+function MOI.get(model::Optimizer, ::NumberOfObjectives)
+    n = # Code to get NumberOfObjectives
+    return n
+end</code></pre><p>Then, the user can write:</p><pre><code class="language-julia hljs">model = Gurobi.Optimizer()
+MOI.set(model, Gurobi.NumberofObjectives(), 3)</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../example/">« Solving a problem using MathOptInterface</a><a class="docs-footer-nextpage" href="../mathprogbase/">Transitioning from MathProgBase »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/tutorials/latency/index.html b/previews/PR3525/moi/tutorials/latency/index.html
new file mode 100644
index 00000000000..2ba0a58de82
--- /dev/null
+++ b/previews/PR3525/moi/tutorials/latency/index.html
@@ -0,0 +1,133 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Latency · JuMP</title><meta name="title" content="Latency · JuMP"/><meta property="og:title" content="Latency · JuMP"/><meta property="twitter:title" content="Latency · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../manipulating_expressions/">Manipulating expressions</a></li><li class="is-active"><a class="tocitem" href>Latency</a><ul class="internal"><li><a class="tocitem" href="#Background"><span>Background</span></a></li><li><a class="tocitem" href="#Causes"><span>Causes</span></a></li><li><a class="tocitem" href="#Resolutions"><span>Resolutions</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Tutorials</a></li><li class="is-active"><a href>Latency</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Latency</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/latency.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Latency"><a class="docs-heading-anchor" href="#Latency">Latency</a><a id="Latency-1"></a><a class="docs-heading-anchor-permalink" href="#Latency" title="Permalink"></a></h1><p>MathOptInterface suffers the &quot;time-to-first-solve&quot; problem of start-up latency.</p><p>This hurts both the user- and developer-experience of MathOptInterface. In the first case, because simple models have a multi-second delay before solving, and in the latter, because our tests take so long to run.</p><p>This page contains some advice on profiling and fixing latency-related problems in the MathOptInterface.jl repository.</p><h2 id="Background"><a class="docs-heading-anchor" href="#Background">Background</a><a id="Background-1"></a><a class="docs-heading-anchor-permalink" href="#Background" title="Permalink"></a></h2><p>Before reading this part of the documentation, you should familiarize yourself with the reasons for latency in Julia and how to fix them.</p><ul><li>Read the blogposts on julialang.org on <a href="https://julialang.org/blog/2021/01/precompile_tutorial/">precompilation</a> and <a href="https://julialang.org/blog/2021/01/snoopi_deep/">SnoopCompile</a></li><li>Read the <a href="https://timholy.github.io/SnoopCompile.jl/stable/">SnoopCompile</a> documentation.</li><li>Watch Tim Holy&#39;s <a href="https://www.youtube.com/watch?v=rVBgrWYKLHY">talk at JuliaCon 2021</a></li><li>Watch the <a href="https://www.youtube.com/watch?v=wXRMwJdEjX4">package development workshop at JuliaCon 2021</a></li></ul><h2 id="Causes"><a class="docs-heading-anchor" href="#Causes">Causes</a><a id="Causes-1"></a><a class="docs-heading-anchor-permalink" href="#Causes" title="Permalink"></a></h2><p>There are three main causes of latency in MathOptInterface:</p><ol><li>A large number of types</li><li>Lack of method ownership</li><li>Type-instability in the bridge layer</li></ol><h3 id="A-large-number-of-types"><a class="docs-heading-anchor" href="#A-large-number-of-types">A large number of types</a><a id="A-large-number-of-types-1"></a><a class="docs-heading-anchor-permalink" href="#A-large-number-of-types" title="Permalink"></a></h3><p>Julia is very good at specializing method calls based on the input type. Each specialization has a compilation cost, but the benefit of faster run-time performance.</p><p>The best-case scenario is for a method to be called a large number of times with a single set of argument types. The worst-case scenario is for a method to be called a single time for a large set of argument types.</p><p>Because of MathOptInterface&#39;s <em>function-in-set</em> formulation, we fall into the worst-case situation.</p><p>This is a fundamental limitation of Julia, so there isn&#39;t much we can do about it. However, if we can precompile MathOptInterface, much of the cost can be shifted from start-up latency to the time it takes to precompile a package on installation.</p><p>However, there are two things which make MathOptInterface hard to precompile.</p><h3 id="Lack-of-method-ownership"><a class="docs-heading-anchor" href="#Lack-of-method-ownership">Lack of method ownership</a><a id="Lack-of-method-ownership-1"></a><a class="docs-heading-anchor-permalink" href="#Lack-of-method-ownership" title="Permalink"></a></h3><p>Lack of method ownership happens when a call is made using a mix of structs and methods from different modules. Because of this, no single module &quot;owns&quot; the method that is being dispatched, and so it cannot be precompiled.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This is a slightly simplified explanation. Read the <a href="https://julialang.org/blog/2021/01/precompile_tutorial/">precompilation tutorial</a> for a more in-depth discussion on back-edges.</p></div></div><p>Unfortunately, the design of MOI means that this is a frequent occurrence: we have a bunch of types in <code>MOI.Utilities</code> that wrap types defined in external packages (for example, the <code>Optimizer</code>s), which implement methods of functions defined in <code>MOI</code> (for example, <code>add_variable</code>, <code>add_constraint</code>).</p><p>Here&#39;s a simple example of method-ownership in practice:</p><pre><code class="language-julia hljs">module MyMOI
+struct Wrapper{T}
+    inner::T
+end
+optimize!(x::Wrapper) = optimize!(x.inner)
+end  # MyMOI
+
+module MyOptimizer
+using ..MyMOI
+struct Optimizer end
+MyMOI.optimize!(x::Optimizer) = 1
+end  # MyOptimizer
+
+using SnoopCompile
+model = MyMOI.Wrapper(MyOptimizer.Optimizer())
+
+julia&gt; tinf = @snoopi_deep MyMOI.optimize!(model)
+InferenceTimingNode: 0.008256/0.008543 on InferenceFrameInfo for Core.Compiler.Timings.ROOT() with 1 direct children</code></pre><p>The result is that there was one method that required type inference. If we visualize <code>tinf</code>:</p><pre><code class="language-julia hljs">using ProfileView
+ProfileView.view(flamegraph(tinf))</code></pre><p>we see a flamegraph with a large red-bar indicating that the method <code>MyMOI.optimize(MyMOI.Wrapper{MyOptimizer.Optimizer})</code> cannot be precompiled.</p><p>To fix this, we need to designate a module to &quot;own&quot; that method (that is, create a back-edge). The easiest way to do this is for <code>MyOptimizer</code> to call <code>MyMOI.optimize(MyMOI.Wrapper{MyOptimizer.Optimizer})</code> during <code>using MyOptimizer</code>. Let&#39;s see that in practice:</p><pre><code class="language-julia hljs">module MyMOI
+struct Wrapper{T}
+    inner::T
+end
+optimize(x::Wrapper) = optimize(x.inner)
+end  # MyMOI
+
+module MyOptimizer
+using ..MyMOI
+struct Optimizer end
+MyMOI.optimize(x::Optimizer) = 1
+# The syntax of this let-while loop is very particular:
+#  * `let ... end` keeps everything local to avoid polluting the MyOptimizer
+#    namespace
+#  * `while true ... break end` runs the code once, and forces Julia to compile
+#    the inner loop, rather than interpret it.
+let
+    while true
+        model = MyMOI.Wrapper(Optimizer())
+        MyMOI.optimize(model)
+        break
+    end
+end
+end  # MyOptimizer
+
+using SnoopCompile
+model = MyMOI.Wrapper(MyOptimizer.Optimizer())
+
+julia&gt; tinf = @snoopi_deep MyMOI.optimize(model)
+InferenceTimingNode: 0.006822/0.006822 on InferenceFrameInfo for Core.Compiler.Timings.ROOT() with 0 direct children</code></pre><p>There are now <code>0</code> direct children that required type inference because the method was already stored in <code>MyOptimizer</code>!</p><p>Unfortunately, this trick only works if the call-chain is fully inferrable. If there are breaks (due to type instability), then the benefit of doing this is reduced. And unfortunately for us, the design of MathOptInterface has a lot of type instabilities.</p><h3 id="Type-instability-in-the-bridge-layer"><a class="docs-heading-anchor" href="#Type-instability-in-the-bridge-layer">Type instability in the bridge layer</a><a id="Type-instability-in-the-bridge-layer-1"></a><a class="docs-heading-anchor-permalink" href="#Type-instability-in-the-bridge-layer" title="Permalink"></a></h3><p>Most of MathOptInterface is pretty good at ensuring type-stability. However, a key component is not type stable, and that is the bridging layer.</p><p>In particular, the bridging layer defines <a href="../../submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer"><code>Bridges.LazyBridgeOptimizer</code></a>, which has fields like:</p><pre><code class="language-julia hljs">struct LazyBridgeOptimizer
+    constraint_bridge_types::Vector{Any}
+    constraint_node::Dict{Tuple{Type,Type},ConstraintNode}
+    constraint_types::Vector{Tuple{Type,Type}}
+end</code></pre><p>This is because the <code>LazyBridgeOptimizer</code> needs to be able to deal with any <em>function-in-set</em> type passed to it, and we also allow users to pass additional bridges that they defined in external packages.</p><p>So to recap, MathOptInterface suffers package latency because:</p><ol><li>there are a large number of types and functions</li><li>and these are split between multiple modules, including external packages</li><li>and there are type-instabilities like those in the bridging layer.</li></ol><h2 id="Resolutions"><a class="docs-heading-anchor" href="#Resolutions">Resolutions</a><a id="Resolutions-1"></a><a class="docs-heading-anchor-permalink" href="#Resolutions" title="Permalink"></a></h2><p>There are no magic solutions to reduce latency. <a href="https://github.com/jump-dev/MathOptInterface.jl/issues/1313">Issue #1313</a> tracks progress on reducing latency in MathOptInterface.</p><p>A useful script is the following (replace GLPK as needed):</p><pre><code class="language-julia hljs">import GLPK
+import MathOptInterface as MOI
+
+function example_diet(optimizer, bridge)
+    category_data = [
+        1800.0 2200.0;
+          91.0    Inf;
+           0.0   65.0;
+           0.0 1779.0
+    ]
+    cost = [2.49, 2.89, 1.50, 1.89, 2.09, 1.99, 2.49, 0.89, 1.59]
+    food_data = [
+        410 24 26 730;
+        420 32 10 1190;
+        560 20 32 1800;
+        380  4 19 270;
+        320 12 10 930;
+        320 15 12 820;
+        320 31 12 1230;
+        100  8 2.5 125;
+        330  8 10 180
+    ]
+    bridge_model = if bridge
+        MOI.instantiate(optimizer; with_bridge_type=Float64)
+    else
+        MOI.instantiate(optimizer)
+    end
+    model = MOI.Utilities.CachingOptimizer(
+        MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}()),
+        MOI.Utilities.AUTOMATIC,
+    )
+    MOI.Utilities.reset_optimizer(model, bridge_model)
+    MOI.set(model, MOI.Silent(), true)
+    nutrition = MOI.add_variables(model, size(category_data, 1))
+    for (i, v) in enumerate(nutrition)
+        MOI.add_constraint(model, v, MOI.GreaterThan(category_data[i, 1]))
+        MOI.add_constraint(model, v, MOI.LessThan(category_data[i, 2]))
+    end
+    buy = MOI.add_variables(model, size(food_data, 1))
+    MOI.add_constraint.(model, buy, MOI.GreaterThan(0.0))
+    MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)
+    f = MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(cost, buy), 0.0)
+    MOI.set(model, MOI.ObjectiveFunction{typeof(f)}(), f)
+    for (j, n) in enumerate(nutrition)
+        f = MOI.ScalarAffineFunction(
+            MOI.ScalarAffineTerm.(food_data[:, j], buy),
+            0.0,
+        )
+        push!(f.terms, MOI.ScalarAffineTerm(-1.0, n))
+        MOI.add_constraint(model, f, MOI.EqualTo(0.0))
+    end
+    MOI.optimize!(model)
+    term_status = MOI.get(model, MOI.TerminationStatus())
+    @assert term_status == MOI.OPTIMAL
+    MOI.add_constraint(
+        model,
+        MOI.ScalarAffineFunction(
+            MOI.ScalarAffineTerm.(1.0, [buy[end-1], buy[end]]),
+            0.0,
+        ),
+        MOI.LessThan(6.0),
+    )
+    MOI.optimize!(model)
+    @assert MOI.get(model, MOI.TerminationStatus()) == MOI.INFEASIBLE
+    return
+end
+
+if length(ARGS) &gt; 0
+    bridge = get(ARGS, 2, &quot;&quot;) != &quot;--no-bridge&quot;
+    println(&quot;Running: $(ARGS[1]) $(get(ARGS, 2, &quot;&quot;))&quot;)
+    @time example_diet(GLPK.Optimizer, bridge)
+    @time example_diet(GLPK.Optimizer, bridge)
+    exit(0)
+end</code></pre><p>You can create a flame-graph via</p><pre><code class="language-julia hljs">using SnoopComile
+tinf = @snoopi_deep example_diet(GLPK.Optimizer, true)
+using ProfileView
+ProfileView.view(flamegraph(tinf))</code></pre><p>Here&#39;s how things looked in mid-August 2021: <img src="../../assets/latency.png" alt="flamegraph"/></p><p>There are a few opportunities for improvement (non-red flames, particularly on the right). But the main problem is a large red (non-precompilable due to method ownership) flame.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../manipulating_expressions/">« Manipulating expressions</a><a class="docs-footer-nextpage" href="../../manual/standard_form/">Standard form problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/tutorials/manipulating_expressions/index.html b/previews/PR3525/moi/tutorials/manipulating_expressions/index.html
new file mode 100644
index 00000000000..29b7f16f072
--- /dev/null
+++ b/previews/PR3525/moi/tutorials/manipulating_expressions/index.html
@@ -0,0 +1,26 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Manipulating expressions · JuMP</title><meta name="title" content="Manipulating expressions · JuMP"/><meta property="og:title" content="Manipulating expressions · JuMP"/><meta property="twitter:title" content="Manipulating expressions · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../bridging_constraint/">Implementing a constraint bridge</a></li><li class="is-active"><a class="tocitem" href>Manipulating expressions</a><ul class="internal"><li><a class="tocitem" href="#Creating-functions"><span>Creating functions</span></a></li><li><a class="tocitem" href="#Canonicalizing-functions"><span>Canonicalizing functions</span></a></li><li><a class="tocitem" href="#Exploring-functions"><span>Exploring functions</span></a></li></ul></li><li><a class="tocitem" href="../latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Tutorials</a></li><li class="is-active"><a href>Manipulating expressions</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Manipulating expressions</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/manipulating_expressions.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Manipulating-expressions"><a class="docs-heading-anchor" href="#Manipulating-expressions">Manipulating expressions</a><a id="Manipulating-expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Manipulating-expressions" title="Permalink"></a></h1><p>This guide highlights a syntactically appealing way to build expressions at the MOI level, but also to look at their contents. It may be especially useful when writing models or bridge code.</p><h2 id="Creating-functions"><a class="docs-heading-anchor" href="#Creating-functions">Creating functions</a><a id="Creating-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Creating-functions" title="Permalink"></a></h2><p>This section details the ways to create functions with MathOptInterface.</p><h3 id="Creating-scalar-affine-functions"><a class="docs-heading-anchor" href="#Creating-scalar-affine-functions">Creating scalar affine functions</a><a id="Creating-scalar-affine-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Creating-scalar-affine-functions" title="Permalink"></a></h3><p>The simplest scalar function is simply a variable:</p><pre><code class="language-julia-repl hljs">julia&gt; x = MOI.add_variable(model) # Create the variable x
+MOI.VariableIndex(1)</code></pre><p>This type of function is extremely simple; to express more complex functions, other types must be used. For instance, a <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> is a sum of linear terms (a factor times a variable) and a constant. Such an object can be built using the standard constructor:</p><pre><code class="language-julia-repl hljs">julia&gt; f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1, x)], 2) # x + 2
+(2) + (1) MOI.VariableIndex(1)</code></pre><p>However, you can also use operators to build the same scalar function:</p><pre><code class="language-julia-repl hljs">julia&gt; f = x + 2
+(2) + (1) MOI.VariableIndex(1)</code></pre><h3 id="Creating-scalar-quadratic-functions"><a class="docs-heading-anchor" href="#Creating-scalar-quadratic-functions">Creating scalar quadratic functions</a><a id="Creating-scalar-quadratic-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Creating-scalar-quadratic-functions" title="Permalink"></a></h3><p>Scalar quadratic functions are stored in <a href="../../reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>ScalarQuadraticFunction</code></a> objects, in a way that is highly similar to scalar affine functions. You can obtain a quadratic function as a product of affine functions:</p><pre><code class="language-julia-repl hljs">julia&gt; 1 * x * x
+(0) + 1.0 MOI.VariableIndex(1)²
+
+julia&gt; f * f  # (x + 2)²
+(4) + (2) MOI.VariableIndex(1) + (2) MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(1)²
+
+julia&gt; f^2  # (x + 2)² too
+(4) + (2) MOI.VariableIndex(1) + (2) MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(1)²</code></pre><h3 id="Creating-vector-functions"><a class="docs-heading-anchor" href="#Creating-vector-functions">Creating vector functions</a><a id="Creating-vector-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Creating-vector-functions" title="Permalink"></a></h3><p>A vector function is a function with several values, irrespective of the number of input variables. Similarly to scalar functions, there are three main types of vector functions: <a href="../../reference/standard_form/#MathOptInterface.VectorOfVariables"><code>VectorOfVariables</code></a>, <a href="../../reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>VectorAffineFunction</code></a>, and <a href="../../reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>VectorQuadraticFunction</code></a>.</p><p>The easiest way to create a vector function is to stack several scalar functions using <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.vectorize"><code>Utilities.vectorize</code></a>. It takes a vector as input, and the generated vector function (of the most appropriate type) has each dimension corresponding to a dimension of the vector.</p><pre><code class="language-julia-repl hljs">julia&gt; g = MOI.Utilities.vectorize([f, 2 * f])
+┌                              ┐
+│(2) + (1) MOI.VariableIndex(1)│
+│(4) + (2) MOI.VariableIndex(1)│
+└                              ┘</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p><a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.vectorize"><code>Utilities.vectorize</code></a> only takes a vector of similar scalar functions: you cannot mix <a href="../../reference/standard_form/#MathOptInterface.VariableIndex"><code>VariableIndex</code></a> and <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a>, for instance. In practice, it means that <code>Utilities.vectorize([x, f])</code> does not work; you should rather use <code>Utilities.vectorize([1 * x, f])</code> instead to only have <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> objects.</p></div></div><h2 id="Canonicalizing-functions"><a class="docs-heading-anchor" href="#Canonicalizing-functions">Canonicalizing functions</a><a id="Canonicalizing-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Canonicalizing-functions" title="Permalink"></a></h2><p>In more advanced use cases, you might need to ensure that a function is &quot;canonical.&quot; Functions are stored as an array of terms, but there is no check that these terms are redundant: a <a href="../../reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>ScalarAffineFunction</code></a> object might have two terms with the same variable, like <code>x + x + 1</code>. These terms could be merged without changing the semantics of the function: <code>2x + 1</code>.</p><p>Working with these objects might be cumbersome. Canonicalization helps maintain redundancy to zero.</p><p><a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.is_canonical"><code>Utilities.is_canonical</code></a> checks whether a function is already in its canonical form:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.is_canonical(f + f) # (x + 2) + (x + 2) is stored as x + x + 4
+false</code></pre><p><a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.canonical"><code>Utilities.canonical</code></a> returns the equivalent canonical version of the function:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.canonical(f + f) # Returns 2x + 4
+(4) + (2) MOI.VariableIndex(1)</code></pre><h2 id="Exploring-functions"><a class="docs-heading-anchor" href="#Exploring-functions">Exploring functions</a><a id="Exploring-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Exploring-functions" title="Permalink"></a></h2><p>At some point, you might need to dig into a function, for instance to map it into solver constructs.</p><h3 id="Vector-functions"><a class="docs-heading-anchor" href="#Vector-functions">Vector functions</a><a id="Vector-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Vector-functions" title="Permalink"></a></h3><p><a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.scalarize"><code>Utilities.scalarize</code></a> returns a vector of scalar functions from a vector function:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.Utilities.scalarize(g) # Returns a vector [f, 2 * f].
+2-element Vector{MathOptInterface.ScalarAffineFunction{Int64}}:
+ (2) + (1) MOI.VariableIndex(1)
+ (4) + (2) MOI.VariableIndex(1)</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p><a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.eachscalar"><code>Utilities.eachscalar</code></a> returns an iterator on the dimensions, which serves the same purpose as <a href="../../submodules/Utilities/reference/#MathOptInterface.Utilities.scalarize"><code>Utilities.scalarize</code></a>.</p></div></div><p><a href="../../reference/standard_form/#MathOptInterface.output_dimension"><code>output_dimension</code></a> returns the number of dimensions of the output of a function:</p><pre><code class="language-julia-repl hljs">julia&gt; MOI.output_dimension(g)
+2</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../bridging_constraint/">« Implementing a constraint bridge</a><a class="docs-footer-nextpage" href="../latency/">Latency »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/moi/tutorials/mathprogbase/index.html b/previews/PR3525/moi/tutorials/mathprogbase/index.html
new file mode 100644
index 00000000000..a9313d45101
--- /dev/null
+++ b/previews/PR3525/moi/tutorials/mathprogbase/index.html
@@ -0,0 +1,58 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Transitioning from MathProgBase · JuMP</title><meta name="title" content="Transitioning from MathProgBase · JuMP"/><meta property="og:title" content="Transitioning from MathProgBase · JuMP"/><meta property="twitter:title" content="Transitioning from MathProgBase · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox" checked/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox" checked/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../implementing/">Implementing a solver interface</a></li><li class="is-active"><a class="tocitem" href>Transitioning from MathProgBase</a><ul class="internal"><li><a class="tocitem" href="#Transitioning-a-solver-interface"><span>Transitioning a solver interface</span></a></li><li><a class="tocitem" href="#Transitioning-the-high-level-functions"><span>Transitioning the high-level functions</span></a></li></ul></li><li><a class="tocitem" href="../bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/duality/">Duality</a></li><li><a class="tocitem" href="../../background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../reference/models/">Models</a></li><li><a class="tocitem" href="../../reference/variables/">Variables</a></li><li><a class="tocitem" href="../../reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">MathOptInterface</a></li><li><a class="is-disabled">Tutorials</a></li><li class="is-active"><a href>Transitioning from MathProgBase</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Transitioning from MathProgBase</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/mathprogbase.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Transitioning-from-MathProgBase"><a class="docs-heading-anchor" href="#Transitioning-from-MathProgBase">Transitioning from MathProgBase</a><a id="Transitioning-from-MathProgBase-1"></a><a class="docs-heading-anchor-permalink" href="#Transitioning-from-MathProgBase" title="Permalink"></a></h1><p>MathOptInterface is a replacement for <a href="https://github.com/JuliaOpt/MathProgBase.jl">MathProgBase.jl</a>. However, it is not a direct replacement.</p><h2 id="Transitioning-a-solver-interface"><a class="docs-heading-anchor" href="#Transitioning-a-solver-interface">Transitioning a solver interface</a><a id="Transitioning-a-solver-interface-1"></a><a class="docs-heading-anchor-permalink" href="#Transitioning-a-solver-interface" title="Permalink"></a></h2><p>MathOptInterface is more extensive than MathProgBase which may make its implementation seem daunting at first. There are however numerous utilities in MathOptInterface that the simplify implementation process.</p><p>For more information, read <a href="../implementing/#Implementing-a-solver-interface">Implementing a solver interface</a>.</p><h2 id="Transitioning-the-high-level-functions"><a class="docs-heading-anchor" href="#Transitioning-the-high-level-functions">Transitioning the high-level functions</a><a id="Transitioning-the-high-level-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Transitioning-the-high-level-functions" title="Permalink"></a></h2><p>MathOptInterface doesn&#39;t provide replacements for the high-level interfaces in MathProgBase. We recommend you use <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> as a modeling interface instead.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you haven&#39;t used JuMP before, start with the tutorial <a href="https://jump.dev/JuMP.jl/stable/tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></p></div></div><h3 id="linprog"><a class="docs-heading-anchor" href="#linprog"><code>linprog</code></a><a id="linprog-1"></a><a class="docs-heading-anchor-permalink" href="#linprog" title="Permalink"></a></h3><p>Here is one way of transitioning from <code>linprog</code>:</p><pre><code class="language-julia hljs">using JuMP
+
+function linprog(c, A, sense, b, l, u, solver)
+    N = length(c)
+    model = Model(solver)
+    @variable(model, l[i] &lt;= x[i=1:N] &lt;= u[i])
+    @objective(model, Min, c&#39; * x)
+    eq_rows, ge_rows, le_rows = sense .== &#39;=&#39;, sense .== &#39;&gt;&#39;, sense .== &#39;&lt;&#39;
+    @constraint(model, A[eq_rows, :] * x .== b[eq_rows])
+    @constraint(model, A[ge_rows, :] * x .&gt;= b[ge_rows])
+    @constraint(model, A[le_rows, :] * x .&lt;= b[le_rows])
+    optimize!(model)
+    return (
+        status = termination_status(model),
+        objval = objective_value(model),
+        sol = value.(x)
+    )
+end</code></pre><h3 id="mixintprog"><a class="docs-heading-anchor" href="#mixintprog"><code>mixintprog</code></a><a id="mixintprog-1"></a><a class="docs-heading-anchor-permalink" href="#mixintprog" title="Permalink"></a></h3><p>Here is one way of transitioning from <code>mixintprog</code>:</p><pre><code class="language-julia hljs">using JuMP
+
+function mixintprog(c, A, rowlb, rowub, vartypes, lb, ub, solver)
+    N = length(c)
+    model = Model(solver)
+    @variable(model, lb[i] &lt;= x[i=1:N] &lt;= ub[i])
+    for i in 1:N
+        if vartypes[i] == :Bin
+            set_binary(x[i])
+        elseif vartypes[i] == :Int
+            set_integer(x[i])
+        end
+    end
+    @objective(model, Min, c&#39; * x)
+    @constraint(model, rowlb .&lt;= A * x .&lt;= rowub)
+    optimize!(model)
+    return (
+        status = termination_status(model),
+        objval = objective_value(model),
+        sol = value.(x)
+    )
+end</code></pre><h3 id="quadprog"><a class="docs-heading-anchor" href="#quadprog"><code>quadprog</code></a><a id="quadprog-1"></a><a class="docs-heading-anchor-permalink" href="#quadprog" title="Permalink"></a></h3><p>Here is one way of transitioning from <code>quadprog</code>:</p><pre><code class="language-julia hljs">using JuMP
+
+function quadprog(c, Q, A, rowlb, rowub, lb, ub, solver)
+    N = length(c)
+    model = Model(solver)
+    @variable(model, lb[i] &lt;= x[i=1:N] &lt;= ub[i])
+    @objective(model, Min, c&#39; * x + 0.5 * x&#39; * Q * x)
+    @constraint(model, rowlb .&lt;= A * x .&lt;= rowub)
+    optimize!(model)
+    return (
+        status = termination_status(model),
+        objval = objective_value(model),
+        sol = value.(x)
+    )
+end</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../implementing/">« Implementing a solver interface</a><a class="docs-footer-nextpage" href="../bridging_constraint/">Implementing a constraint bridge »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Alpine/index.html b/previews/PR3525/packages/Alpine/index.html
new file mode 100644
index 00000000000..34521681e74
--- /dev/null
+++ b/previews/PR3525/packages/Alpine/index.html
@@ -0,0 +1,49 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>lanl-ansi/Alpine.jl · JuMP</title><meta name="title" content="lanl-ansi/Alpine.jl · JuMP"/><meta property="og:title" content="lanl-ansi/Alpine.jl · JuMP"/><meta property="twitter:title" content="lanl-ansi/Alpine.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li class="is-active"><a class="tocitem" href>lanl-ansi/Alpine.jl</a><ul class="internal"><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Support-problem-types"><span>Support problem types</span></a></li><li><a class="tocitem" href="#Underlying-solvers"><span>Underlying solvers</span></a></li><li><a class="tocitem" href="#Bug-reports-and-support"><span>Bug reports and support</span></a></li><li><a class="tocitem" href="#Challenging-Problems"><span>Challenging Problems</span></a></li><li><a class="tocitem" href="#Citing-Alpine"><span>Citing Alpine</span></a></li></ul></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>lanl-ansi/Alpine.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>lanl-ansi/Alpine.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/lanl-ansi/Alpine.jl/blob/32757b2568acfe71257e41ed4122a120717a6460/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Alpine,-a-global-solver-for-non-convex-MINLPs"><a class="docs-heading-anchor" href="#Alpine,-a-global-solver-for-non-convex-MINLPs">Alpine, a global solver for non-convex MINLPs</a><a id="Alpine,-a-global-solver-for-non-convex-MINLPs-1"></a><a class="docs-heading-anchor-permalink" href="#Alpine,-a-global-solver-for-non-convex-MINLPs" title="Permalink"></a></h1><p><a href="https://github.com/lanl-ansi/Alpine.jl/actions/workflows/ci.yml"><img src="https://github.com/lanl-ansi/Alpine.jl/actions/workflows/ci.yml/badge.svg" alt="CI"/></a> <a href="https://codecov.io/gh/lanl-ansi/Alpine.jl"><img src="https://codecov.io/gh/lanl-ansi/Alpine.jl/branch/master/badge.svg" alt="codecov"/></a> <a href="https://lanl-ansi.github.io/Alpine.jl/latest/"><img src="https://github.com/lanl-ansi/Alpine.jl/actions/workflows/documentation.yml/badge.svg" alt="Documentation"/></a>  <a href="https://juliahub.com/ui/Packages/Alpine/TRSJF"><img src="https://juliahub.com/docs/Alpine/version.svg" alt="version"/></a></p><p>ALPINE (glob(AL) o(P)timization for mixed-(I)nteger programs with (N)onlinear (E)quations), is a novel global optimization solver that uses an adaptive, piecewise convexification scheme and constraint programming methods to solve non-convex Mixed-Integer Non-Linear Programs (MINLPs) efficiently. MINLPs are typically &quot;hard&quot; optimization problems which appear in numerous applications (see <a href="https://github.com/lanl-ansi/MINLPLib.jl">MINLPLib.jl</a>). </p><p>Alpine is entirely built upon <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> and <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a> in Julia, which provides incredible flexibility for usage and further development.</p><p>Alpine globally solves a given MINLP by:</p><ul><li><p>Analyzing the problem&#39;s expressions (objective &amp; constraints) and applies appropriate convex relaxations and polyhedral outer-approximations</p></li><li><p>Performing sequential optimization-based bound tightening (OBBT) and an iterative MIP-based adaptive partitioning scheme via piecewise polyhedral relaxations with a guarantee of global convergence</p></li></ul><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Alpine using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Alpine&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use Alpine with JuMP as follows:</p><pre><code class="language-julia hljs">using JuMP, Alpine, Ipopt, HiGHS
+ipopt = optimizer_with_attributes(Ipopt.Optimizer, &quot;print_level&quot; =&gt; 0)
+highs = optimizer_with_attributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; false)
+model = Model(
+    optimizer_with_attributes(
+        Juniper.Optimizer,
+        &quot;nlp_solver&quot; =&gt; ipopt,
+        &quot;mip_solver&quot; =&gt; highs,
+    ),
+)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>For more details, see the <a href="https://lanl-ansi.github.io/Alpine.jl/latest/">online documentation</a>.</p><h2 id="Support-problem-types"><a class="docs-heading-anchor" href="#Support-problem-types">Support problem types</a><a id="Support-problem-types-1"></a><a class="docs-heading-anchor-permalink" href="#Support-problem-types" title="Permalink"></a></h2><p>Alpine can currently handle MINLPs with polynomials in constraints and/or in the objective. Currently, there is no support for exponential cones and Positive Semi-Definite (PSD) cones in MINLPs. Alpine is also a good fit for subsets of the MINLP family, for example, Mixed-Integer Quadratically Constrained Quadratic Programs (MIQCQPs), Non-Linear Programs (NLPs), etc.</p><p>For more details, check out this <a href="https://www.youtube.com/watch?v=mwkhiEIS5JA">video</a> on Alpine.jl at <a href="http://www.juliaopt.org/meetings/bordeaux2018/">JuMP-dev 2018</a>.</p><h2 id="Underlying-solvers"><a class="docs-heading-anchor" href="#Underlying-solvers">Underlying solvers</a><a id="Underlying-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Underlying-solvers" title="Permalink"></a></h2><p>Though the MIP-based bounding algorithm implemented in Alpine is quite involved, most of the computational bottleneck arises in the underlying MIP solvers. Since every iteration of Alpine solves an MIP sub-problem, which is typically a convex MILP/MIQCQP, Alpine&#39;s run time heavily depends on the run-time of these solvers. For the best performance of Alpine, we recommend using the commercial solver <a href="https://www.gurobi.com">Gurobi</a>, which is available <a href="https://www.gurobi.com/academia/academic-program-and-licenses/">free</a> for academic purposes. However, due to the flexibility offered by <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>, the following MIP and NLP solvers are supported in Alpine: </p><table><tr><th style="text-align: right">Solver</th><th style="text-align: right">Julia Package</th></tr><tr><td style="text-align: right"><a href="https://www.ibm.com/analytics/cplex-optimizer">CPLEX</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/CPLEX.jl">CPLEX.jl</a></td></tr><tr><td style="text-align: right"><a href="https://projects.coin-or.org/Cbc">Cbc</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Cbc.jl">Cbc.jl</a></td></tr><tr><td style="text-align: right"><a href="http://gurobi.com/">Gurobi</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Gurobi.jl">Gurobi.jl</a></td></tr><tr><td style="text-align: right"><a href="https://projects.coin-or.org/Ipopt">Ipopt</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Ipopt.jl">Ipopt.jl</a></td></tr><tr><td style="text-align: right"><a href="https://projects.coin-or.org/Bonmin">Bonmin</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/AmplNLWriter.jl">Bonmin.jl</a></td></tr><tr><td style="text-align: right"><a href="http://artelys.com/en/optimization-tools/knitro">Artelys KNITRO</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/KNITRO.jl">KNITRO.jl</a></td></tr><tr><td style="text-align: right"><a href="https://www.fico.com/en/products/fico-xpress-optimization">Xpress</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/Xpress.jl">Xpress.jl</a></td></tr><tr><td style="text-align: right"><a href="https://highs.dev/">HiGHS</a></td><td style="text-align: right"><a href="https://github.com/jump-dev/HiGHS.jl">HiGHS.jl</a></td></tr></table><h2 id="Bug-reports-and-support"><a class="docs-heading-anchor" href="#Bug-reports-and-support">Bug reports and support</a><a id="Bug-reports-and-support-1"></a><a class="docs-heading-anchor-permalink" href="#Bug-reports-and-support" title="Permalink"></a></h2><p>Please report any issues via the GitHub <a href="https://github.com/lanl-ansi/Alpine.jl/issues">issue tracker</a>. All types of issues are welcome and encouraged; this includes bug reports, documentation typos, feature requests, etc. </p><h2 id="Challenging-Problems"><a class="docs-heading-anchor" href="#Challenging-Problems">Challenging Problems</a><a id="Challenging-Problems-1"></a><a class="docs-heading-anchor-permalink" href="#Challenging-Problems" title="Permalink"></a></h2><p>We are seeking out hard benchmark instances for MINLPs. Please get in touch either by opening an issue or <a href="https://harshangrjn.github.io/#contact">privately</a> if you would like to share any hard instances.</p><h2 id="Citing-Alpine"><a class="docs-heading-anchor" href="#Citing-Alpine">Citing Alpine</a><a id="Citing-Alpine-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-Alpine" title="Permalink"></a></h2><p>If you find Alpine useful in your work, we kindly request that you cite the following papers (<a href="http://harshangrjn.github.io/pdf/JOGO_2018.pdf">PDF</a>, <a href="http://harshangrjn.github.io/pdf/CP_2016.pdf">PDF</a>)</p><pre><code class="language-bibtex hljs">@article{alpine_JOGO2019,
+  title = {An adaptive, multivariate partitioning algorithm for global optimization of nonconvex programs},
+  author = {Nagarajan, Harsha and Lu, Mowen and Wang, Site and Bent, Russell and Sundar, Kaarthik},
+  journal = {Journal of Global Optimization},
+  year = {2019},
+  issn = {1573-2916},
+  doi = {10.1007/s10898-018-00734-1},
+}
+
+@inproceedings{alpine_CP2016,
+  title = {Tightening {McCormick} relaxations for nonlinear programs via dynamic multivariate partitioning},
+  author = {Nagarajan, Harsha and Lu, Mowen and Yamangil, Emre and Bent, Russell},
+  booktitle = {International Conference on Principles and Practice of Constraint Programming},
+  pages = {369--387},
+  year = {2016},
+  organization = {Springer},
+  doi = {10.1007/978-3-319-44953-1_24},
+}</code></pre><p>If you find the underlying piecewise polyhedral formulations implemented in Alpine useful in your work, we kindly request that you cite the following papers (<a href="https://doi.org/10.1016/j.orl.2020.12.002">link-1</a>, <a href="http://www.optimization-online.org/DB_HTML/2022/07/8974.html">link-2</a>): </p><pre><code class="language-bibtex hljs">@article{alpine_ORL2021,
+  title = {Piecewise polyhedral formulations for a multilinear term},
+  author = {Sundar, Kaarthik and Nagarajan, Harsha and Linderoth, Jeff and Wang, Site and Bent, Russell},
+  journal = {Operations Research Letters},
+  volume = {49},
+  number = {1},
+  pages = {144--149},
+  year = {2021},
+  publisher = {Elsevier}
+}
+
+@article{alpine_OptOnline2022,
+  title={Piecewise Polyhedral Relaxations of Multilinear Optimization},
+  author={Kim, Jongeun and Richard, Jean-Philippe P. and Tawarmalani, Mohit},
+  eprinttype={Optimization Online},
+  date={2022}
+}</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Loraine/">« kocvara/Loraine.jl</a><a class="docs-footer-nextpage" href="../Juniper/">lanl-ansi/Juniper.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/AmplNLWriter/index.html b/previews/PR3525/packages/AmplNLWriter/index.html
new file mode 100644
index 00000000000..98a80a2613b
--- /dev/null
+++ b/previews/PR3525/packages/AmplNLWriter/index.html
@@ -0,0 +1,15 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/AmplNLWriter.jl · JuMP</title><meta name="title" content="jump-dev/AmplNLWriter.jl · JuMP"/><meta property="og:title" content="jump-dev/AmplNLWriter.jl · JuMP"/><meta property="twitter:title" content="jump-dev/AmplNLWriter.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li class="is-active"><a class="tocitem" href>jump-dev/AmplNLWriter.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Precompiled-binaries"><span>Precompiled binaries</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li></ul></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/AmplNLWriter.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/AmplNLWriter.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/AmplNLWriter.jl/blob/v1.2.0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="AmplNLWriter.jl"><a class="docs-heading-anchor" href="#AmplNLWriter.jl">AmplNLWriter.jl</a><a id="AmplNLWriter.jl-1"></a><a class="docs-heading-anchor-permalink" href="#AmplNLWriter.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/AmplNLWriter.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/AmplNLWriter.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://github.com/jump-dev/AmplNLWriter.jl/actions?query=workflow%3AMINLPTests"><img src="https://github.com/jump-dev/AmplNLWriter.jl/workflows/MINLPTests/badge.svg?branch=master" alt="MINLPTests"/></a> <a href="https://codecov.io/gh/jump-dev/AmplNLWriter.jl"><img src="https://codecov.io/gh/jump-dev/AmplNLWriter.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/AmplNLWriter.jl">AmplNLWriter.jl</a> is an interface between <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface.jl</a> and <a href="http://ampl.com/products/solvers/all-solvers-for-ampl/">AMPL-enabled solvers</a>.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and has no official connection with the AMPL modeling language or AMPL Optimization Inc.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install AmplNLWriter using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;AmplNLWriter&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>AmplNLWriter requires an AMPL compatible solver binary to function.</p><p>Pass a string pointing to any AMPL-compatible solver binary as the first positional argument to <code>AmplNLWriter</code>.</p><p>For example, if the <code>bonmin</code> executable is on the system path, use:</p><pre><code class="language-julia hljs">using JuMP, AmplNLWriter
+model = Model(() -&gt; AmplNLWriter.Optimizer(&quot;bonmin&quot;))</code></pre><p>If the solver is not on the system path, pass the full path to the solver:</p><pre><code class="language-julia hljs">using JuMP, AmplNLWriter
+model = Model(() -&gt; AmplNLWriter.Optimizer(&quot;/Users/Oscar/ampl.macos64/bonmin&quot;))</code></pre><h2 id="Precompiled-binaries"><a class="docs-heading-anchor" href="#Precompiled-binaries">Precompiled binaries</a><a id="Precompiled-binaries-1"></a><a class="docs-heading-anchor-permalink" href="#Precompiled-binaries" title="Permalink"></a></h2><p>To simplify the process of installing solver binaries, a number of Julia packages provide precompiled binaries that are compatible with AmplNLWriter. These are generally the name of the solver, followed by <code>_jll</code>. For example, <code>bomin</code> is provided by the <code>Bonmin_jll</code> package.</p><p>To call Bonmin via AmplNLWriter.jl, install the <code>Bonmin_jll</code> package, then run:</p><pre><code class="language-julia hljs">using JuMP, AmplNLWriter, Bonmin_jll
+model = Model(() -&gt; AmplNLWriter.Optimizer(Bonmin_jll.amplexe))</code></pre><p>Supported packages include:</p><table><tr><th style="text-align: right">Solver</th><th style="text-align: right">Julia Package</th><th style="text-align: right">Executable</th></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/Bonmin">Bonmin</a></td><td style="text-align: right"><code>Bonmin_jll.jl</code></td><td style="text-align: right"><code>Bomin_jll.amplexe</code></td></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/Couenne">Couenne</a></td><td style="text-align: right"><code>Couenne_jll.jl</code></td><td style="text-align: right"><code>Couenne_jll.amplexe</code></td></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/Ipopt">Ipopt</a></td><td style="text-align: right"><code>Ipopt_jll.jl</code></td><td style="text-align: right"><code>Ipopt_jll.amplexe</code></td></tr><tr><td style="text-align: right"><a href="https://github.com/coin-or/SHOT">SHOT</a></td><td style="text-align: right"><code>SHOT_jll.jl</code></td><td style="text-align: right"><code>SHOT_jll.amplexe</code></td></tr><tr><td style="text-align: right"><a href="https://github.comjump-dev/KNITRO.jl">KNITRO</a></td><td style="text-align: right"><code>KNITRO.jl</code></td><td style="text-align: right"><code>KNITRO.amplexe</code></td></tr></table><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The AmplNLWriter optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlock"><code>MOI.NLPBlock()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><p>Note that some solver executables may not support the full list of constraint types. For example, <code>Ipopt_jll</code> does not support <code>MOI.Integer</code> or <code>MOI.ZeroOne</code> constraints.</p><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>A list of available options for each solver can be found here:</p><ul><li><a href="https://github.com/coin-or/Bonmin/blob/master/test/bonmin.opt">Bonmin</a> (plus Ipopt options)</li><li><a href="https://github.com/coin-or/Couenne/blob/master/src/couenne.opt">Couenne</a> (plus Ipopt and Bonmin options)</li><li><a href="https://coin-or.github.io/Ipopt/OPTIONS.html">Ipopt</a></li><li><a href="https://shotsolver.dev/shot/using-shot/solver-options">SHOT</a></li></ul><p>Set an option using <a href="../../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a>. For example, to set the <code>&quot;bonmin.nlp_log_level&quot;</code> option to 0 in Bonmin, use:</p><pre><code class="language-julia hljs">using JuMP
+import AmplNLWriter
+import Bonmin_jll
+model = Model(() -&gt; AmplNLWriter.Optimizer(Bonmin_jll.amplexe))
+set_attribute(model, &quot;bonmin.nlp_log_level&quot;, 0)</code></pre><h3 id="opt-files"><a class="docs-heading-anchor" href="#opt-files">opt files</a><a id="opt-files-1"></a><a class="docs-heading-anchor-permalink" href="#opt-files" title="Permalink"></a></h3><p>Some options need to be specified via an <code>.opt</code> file.</p><p>This file must be located in the current working directory whenever the model is solved.</p><p>The <code>.opt</code> file must be named after the name of the solver, for example, <code>bonmin.opt</code>, and each line must contain an option name and the desired value, separated by a space.</p><p>For example, to set the absolute and relative tolerances in Couenne to <code>1</code> and <code>0.05</code> respectively, the <code>couenne.opt</code> file should contain:</p><pre><code class="language-raw hljs">allowable_gap 1
+allowable_fraction_gap 0.05</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../solvers/">« Introduction</a><a class="docs-footer-nextpage" href="../BARON/">jump-dev/BARON.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/BARON/index.html b/previews/PR3525/packages/BARON/index.html
new file mode 100644
index 00000000000..3526be764e2
--- /dev/null
+++ b/previews/PR3525/packages/BARON/index.html
@@ -0,0 +1,9 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/BARON.jl · JuMP</title><meta name="title" content="jump-dev/BARON.jl · JuMP"/><meta property="og:title" content="jump-dev/BARON.jl · JuMP"/><meta property="twitter:title" content="jump-dev/BARON.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/BARON.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li></ul></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/BARON.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/BARON.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/BARON.jl/blob/v0.8.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="BARON.jl"><a class="docs-heading-anchor" href="#BARON.jl">BARON.jl</a><a id="BARON.jl-1"></a><a class="docs-heading-anchor-permalink" href="#BARON.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/BARON.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/BARON.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/BARON.jl"><img src="https://codecov.io/gh/jump-dev/BARON.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/BARON.jl">BARON.jl</a> is a wrapper for <a href="http://minlp.com/baron">BARON by The Optimization Firm</a>.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not officially supported by The Optimization Firm.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>BARON.jl</code> is licensed under the <a href="https://github.com/jump-dev/BARON.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver is a closed-source commercial product for which you must obtain a license from <a href="http://minlp.com">The Optimization Firm</a>, although a small trial version is available for free.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, download a copy of <a href="http://minlp.com/">the BARON solver</a> and unpack the executable in a location of your choosing.</p><p>Once installed, set the <code>BARON_EXEC</code> environment variable pointing to the BARON executable (full path, including file name as it differs across platforms), and run <code>Pkg.add(&quot;BARON&quot;)</code>. For example:</p><pre><code class="language-julia hljs">ENV[&quot;BARON_EXEC&quot;] = &quot;/path/to/baron.exe&quot;
+using Pkg
+Pkg.add(&quot;BARON&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><pre><code class="language-julia hljs">using JuMP, BARON
+model = Model(BARON.Optimizer)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The BARON optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarNonlinearFunction}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>MOI.ScalarNonlinearFunction</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>MOI.ScalarNonlinearFunction</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>MOI.ScalarNonlinearFunction</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarNonlinearFunction"><code>MOI.ScalarNonlinearFunction</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlock"><code>MOI.NLPBlock()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../AmplNLWriter/">« jump-dev/AmplNLWriter.jl</a><a class="docs-footer-nextpage" href="../CPLEX/">jump-dev/CPLEX.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/BilevelJuMP/index.html b/previews/PR3525/packages/BilevelJuMP/index.html
new file mode 100644
index 00000000000..f66dd48392b
--- /dev/null
+++ b/previews/PR3525/packages/BilevelJuMP/index.html
@@ -0,0 +1,37 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>joaquimg/BilevelJuMP.jl · JuMP</title><meta name="title" content="joaquimg/BilevelJuMP.jl · JuMP"/><meta property="og:title" content="joaquimg/BilevelJuMP.jl · JuMP"/><meta property="twitter:title" content="joaquimg/BilevelJuMP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li class="is-active"><a class="tocitem" href>joaquimg/BilevelJuMP.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Help"><span>Help</span></a></li><li><a class="tocitem" href="#Example"><span>Example</span></a></li></ul></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>joaquimg/BilevelJuMP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>joaquimg/BilevelJuMP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/joaquimg/BilevelJuMP.jl/blob/v0.6.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="BilevelJuMP.jl"><a class="docs-heading-anchor" href="#BilevelJuMP.jl">BilevelJuMP.jl</a><a id="BilevelJuMP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#BilevelJuMP.jl" title="Permalink"></a></h1><p><a href="https://github.com/joaquimg/BilevelJuMP.jl/actions?query=workflow%3ACI"><img src="https://github.com/joaquimg/BilevelJuMP.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="http://codecov.io/github/joaquimg/BilevelJuMP.jl?branch=master"><img src="http://codecov.io/github/joaquimg/BilevelJuMP.jl/coverage.svg?branch=master" alt="Codecov branch"/></a> <a href="https://joaquimg.github.io/BilevelJuMP.jl/dev/"><img src="https://img.shields.io/badge/docs-latest-blue.svg" alt="Docs dev"/></a> <a href="https://joaquimg.github.io/BilevelJuMP.jl/stable/"><img src="https://img.shields.io/badge/docs-stable-blue.svg" alt="Docs stable"/></a></p><p><a href="https://github.com/joaquimg/BilevelJuMP.jl">BilevelJuMP.jl</a> is a <a href="https://github.com/JuMP-dev/JuMP.jl">JuMP</a> extension for modeling and solving bilevel optimization problems.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>BilevelJuMP.jl</code> is licensed under the <a href="https://github.com/joaquimg/BilevelJuMP.jl/blob/master/LICENSE">MIT license</a>.</p><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>You can find the documentation at <a href="https://joaquimg.github.io/BilevelJuMP.jl/stable">https://joaquimg.github.io/BilevelJuMP.jl/stable/</a>.</p><h2 id="Help"><a class="docs-heading-anchor" href="#Help">Help</a><a id="Help-1"></a><a class="docs-heading-anchor-permalink" href="#Help" title="Permalink"></a></h2><p>If you need help, please <a href="https://github.com/joaquimg/BilevelJuMP.jl/issues/new">open a GitHub issue</a>.</p><h2 id="Example"><a class="docs-heading-anchor" href="#Example">Example</a><a id="Example-1"></a><a class="docs-heading-anchor-permalink" href="#Example" title="Permalink"></a></h2><h3 id="Install"><a class="docs-heading-anchor" href="#Install">Install</a><a id="Install-1"></a><a class="docs-heading-anchor-permalink" href="#Install" title="Permalink"></a></h3><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;BilevelJuMP&quot;)
+Pkg.add(&quot;HiGHS&quot;)</code></pre><h3 id="Run"><a class="docs-heading-anchor" href="#Run">Run</a><a id="Run-1"></a><a class="docs-heading-anchor-permalink" href="#Run" title="Permalink"></a></h3><pre><code class="language-julia hljs">using BilevelJuMP, HiGHS
+
+model = BilevelModel(
+    HiGHS.Optimizer,
+    mode = BilevelJuMP.FortunyAmatMcCarlMode(primal_big_M = 100, dual_big_M = 100)
+)
+
+@variable(Lower(model), x)
+@variable(Upper(model), y)
+
+@objective(Upper(model), Min, 3x + y)
+@constraints(Upper(model), begin
+    x &lt;= 5
+    y &lt;= 8
+    y &gt;= 0
+end)
+
+@objective(Lower(model), Min, -x)
+@constraints(Lower(model), begin
+     x +  y &lt;= 8
+    4x +  y &gt;= 8
+    2x +  y &lt;= 13
+    2x - 7y &lt;= 0
+end)
+
+optimize!(model)
+
+objective_value(model) # = 3 * (3.5 * 8/15) + 8/15 # = 6.13...
+value(x) # = 3.5 * 8/15 # = 1.86...
+value(y) # = 8/15 # = 0.53...</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../InfiniteOpt/">« infiniteopt/InfiniteOpt.jl</a><a class="docs-footer-nextpage" href="../SDDP/">odow/SDDP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/CDCS/index.html b/previews/PR3525/packages/CDCS/index.html
new file mode 100644
index 00000000000..2f4a2c2696d
--- /dev/null
+++ b/previews/PR3525/packages/CDCS/index.html
@@ -0,0 +1,30 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>oxfordcontrol/CDCS.jl · JuMP</title><meta name="title" content="oxfordcontrol/CDCS.jl · JuMP"/><meta property="og:title" content="oxfordcontrol/CDCS.jl · JuMP"/><meta property="twitter:title" content="oxfordcontrol/CDCS.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li class="is-active"><a class="tocitem" href>oxfordcontrol/CDCS.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Troubleshooting"><span>Troubleshooting</span></a></li></ul></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>oxfordcontrol/CDCS.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>oxfordcontrol/CDCS.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/oxfordcontrol/CDCS.jl/blob/6d165e8e8f59b14e36e99c8efb19f0e775bbd292/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="CDCS.jl"><a class="docs-heading-anchor" href="#CDCS.jl">CDCS.jl</a><a id="CDCS.jl-1"></a><a class="docs-heading-anchor-permalink" href="#CDCS.jl" title="Permalink"></a></h1><p><a href="https://github.com/oxfordcontrol/CDCS.jl">CDCS.jl</a> is a wrapper for the <a href="https://github.com/oxfordcontrol/CDCS">CDCS</a> solver. </p><p>The wrapper has two components:</p><ul><li>an exported <code>cdcs</code> function that is a thin wrapper on top of the <code>cdcs</code> MATLAB function</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>CDCS.jl is licensed under the <a href="https://github.com/oxfordcontrol/CDCS.jl/blob/master/LICENSE.md">MIT license</a>.</p><p>The underlying solver <a href="https://github.com/oxfordcontrol/CDCS">oxfordcontrol/CDCS</a> is licensed under the <a href="https://github.com/oxfordcontrol/CDCS/blob/master/LICENCE.txt">LGPL v3 license</a>.</p><p>In addition, CDCS requires an installation of MATLAB, which is a closed-source commercial product for which you must <a href="https://www.mathworks.com/products/matlab.html">obtain a license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, make sure that you satisfy the requirements of the <a href="https://github.com/JuliaInterop/MATLAB.jl">MATLAB.jl</a> Julia package and that the CDCS software is installed in your <a href="http://www.mathworks.com/products/matlab/">MATLAB™</a> installation.</p><p>Then, install <code>CDCS.jl</code> using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;CDCS&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use CDCS with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>, do:</p><pre><code class="language-julia hljs">using JuMP, CDCS
+model = Model(CDCS.Optimizer)
+set_attribute(model, &quot;verbose&quot;, 0)</code></pre><h2 id="Troubleshooting"><a class="docs-heading-anchor" href="#Troubleshooting">Troubleshooting</a><a id="Troubleshooting-1"></a><a class="docs-heading-anchor-permalink" href="#Troubleshooting" title="Permalink"></a></h2><p>If you get the error:</p><pre><code class="nohighlight hljs">Undefined function or variable &#39;cdcs&#39;.
+
+Error using save
+Variable &#39;jx_cdcs_arg_out_1&#39; not found.
+
+Linear Programming example: Error During Test at /home/blegat/.julia/dev/CDCS/test/lp.jl:5
+  Got exception outside of a @test
+  MATLAB.MEngineError(&quot;failed to get variable jx_cdcs_arg_out_1 from MATLAB session&quot;)
+  Stacktrace:
+    [1] get_mvariable(session::MATLAB.MSession, name::Symbol)
+      @ MATLAB ~/.julia/packages/MATLAB/SVjnA/src/engine.jl:164
+    [2] mxcall(::MATLAB.MSession, ::Symbol, ::Int64, ::Matrix{Float64}, ::Vararg{Any})
+      @ MATLAB ~/.julia/packages/MATLAB/SVjnA/src/engine.jl:297
+    [3] mxcall
+      @ ~/.julia/packages/MATLAB/SVjnA/src/engine.jl:317 [inlined]
+    [4] cdcs(A::Matrix{Float64}, b::Vector{Float64}, c::Vector{Float64}, K::CDCS.Cone; kws::Base.Pairs{Symbol, Int64, Tuple{Symbol}, NamedTuple{(:verbose,), Tuple{Int64}}})</code></pre><p>The error means that we could not find the <code>cdcs</code> function with one output argument using the MATLAB C API.</p><p>This most likely means that you did not add CDCS to the MATLAB&#39;s path (that is, the <code>toolbox/local/pathdef.m</code> file).</p><p>If modifying <code>toolbox/local/pathdef.m</code> does not work, the following should work where <code>/path/to/CDCS/</code> is the directory where the <code>CDCS</code> folder is located:</p><pre><code class="language-julia hljs">julia&gt; using MATLAB
+
+julia&gt; cd(&quot;/path/to/CDCS/&quot;) do
+           mat&quot;cdcsInstall&quot;
+       end
+
+julia&gt; mat&quot;savepath&quot;</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../OSQP/">« osqp/OSQP.jl</a><a class="docs-footer-nextpage" href="../COSMO/">oxfordcontrol/COSMO.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/CDDLib/index.html b/previews/PR3525/packages/CDDLib/index.html
new file mode 100644
index 00000000000..24454c80d67
--- /dev/null
+++ b/previews/PR3525/packages/CDDLib/index.html
@@ -0,0 +1,9 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>JuliaPolyhedra/CDDLib.jl · JuMP</title><meta name="title" content="JuliaPolyhedra/CDDLib.jl · JuMP"/><meta property="og:title" content="JuliaPolyhedra/CDDLib.jl · JuMP"/><meta property="twitter:title" content="JuliaPolyhedra/CDDLib.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li class="is-active"><a class="tocitem" href>JuliaPolyhedra/CDDLib.jl</a><ul class="internal"><li><a class="tocitem" href="#Problem-description"><span>Problem description</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Debugging"><span>Debugging</span></a></li></ul></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>JuliaPolyhedra/CDDLib.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>JuliaPolyhedra/CDDLib.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/JuliaPolyhedra/CDDLib.jl/blob/9700752f849845bc8d8477928baf3df920daadd6/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="CDDLib"><a class="docs-heading-anchor" href="#CDDLib">CDDLib</a><a id="CDDLib-1"></a><a class="docs-heading-anchor-permalink" href="#CDDLib" title="Permalink"></a></h1><p><a href="https://github.com/JuliaPolyhedra/CDDLib.jl/actions?query=workflow%3ACI"><img src="https://github.com/JuliaPolyhedra/CDDLib.jl/workflows/CI/badge.svg?branch=master" alt/></a> <a href="http://codecov.io/github/JuliaPolyhedra/CDDLib.jl?branch=master"><img src="http://codecov.io/github/JuliaPolyhedra/CDDLib.jl/coverage.svg?branch=master" alt/></a> <a href="https://doi.org/10.5281/zenodo.1214581"><img src="https://zenodo.org/badge/DOI/10.5281/zenodo.1214581.svg" alt/></a></p><p><a href="https://github.com/JuliaPolyhedra/CDDLib.jl">CDDLib.jl</a> is a wrapper for <a href="https://www.inf.ethz.ch/personal/fukudak/cdd_home/">cddlib</a>.</p><p>CDDLib.jl can be used with C API of cddlib, the higher level interface of <a href="https://github.com/JuliaPolyhedra/Polyhedra.jl">Polyhedra.jl</a>, or as a linear programming solver with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> or <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a>.</p><h2 id="Problem-description"><a class="docs-heading-anchor" href="#Problem-description">Problem description</a><a id="Problem-description-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-description" title="Permalink"></a></h2><p>As written in the <a href="https://github.com/cddlib/cddlib">README of cddlib</a>:</p><blockquote><p>The C-library  cddlib is a C implementation of the Double Description Method of Motzkin et al. for generating all vertices (that is, extreme points) and extreme rays of a general convex polyhedron in R^d given by a system of linear inequalities:</p><pre><code class="nohighlight hljs">P = { x=(x1, ..., xd)^T :  b - A  x  &gt;= 0 }</code></pre><p>where  A  is a given m x d real matrix, b is a given m-vector and 0 is the m-vector of all zeros.</p><p>The program can be used for the reverse operation (that is, convex hull computation).  This means that  one can move back and forth between an inequality representation  and a generator (that is, vertex and ray) representation of a polyhedron with cdd.  Also, cdd can solve a linear programming problem, that is, a problem of maximizing and minimizing a linear function over P.</p></blockquote><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>CDDLib.jl is licensed under the <a href="https://github.com/JuliaPolyhedra/CDDLib.jl/blob/master/LICENSE.md">GPL v2 license</a>.</p><p>The underlying solver, <a href="https://github.com/cddlib/cddlib">cddlib/cddlib</a> is also licensed under the <a href="https://github.com/cddlib/cddlib/blob/master/COPYING">GPL v2 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install CDDLib.jl using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;CDDLib&quot;)</code></pre><p>Building the package will download binaries of <a href="https://github.com/cddlib/cddlib">cddlib</a> that are provided by <a href="https://github.com/JuliaBinaryWrappers/cddlib_jll.jl">cddlib_jll.jl</a>.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use <code>CDDLib.Optimizer{Float64}</code> to use CDDLib.jl with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>:</p><pre><code class="language-julia hljs">using JuMP, CDDLib
+model = Model(CDDLib.Optimizer{Float64})</code></pre><p>When using CDDLib.jl with <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a>, you can pass a different number type:</p><pre><code class="language-julia hljs">using MathOptInterface, CDDLib
+model = CDDLib.Optimizer{Rational{BigInt}}()</code></pre><h2 id="Debugging"><a class="docs-heading-anchor" href="#Debugging">Debugging</a><a id="Debugging-1"></a><a class="docs-heading-anchor-permalink" href="#Debugging" title="Permalink"></a></h2><p>CDDLib.jl uses two global Boolean variables to enable debugging outputs: <code>debug</code> and <code>log</code>.</p><p>You can query the value of <code>debug</code> and <code>log</code> with <code>get_debug</code> and <code>get_log</code>, and set their values with <code>set_debug</code> and <code>set_log</code>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../NLopt/">« JuliaOpt/NLopt.jl</a><a class="docs-footer-nextpage" href="../MadNLP/">MadNLP/MadNLP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/COPT/index.html b/previews/PR3525/packages/COPT/index.html
new file mode 100644
index 00000000000..cb10089a6d2
--- /dev/null
+++ b/previews/PR3525/packages/COPT/index.html
@@ -0,0 +1,42 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>COPT-Public/COPT.jl · JuMP</title><meta name="title" content="COPT-Public/COPT.jl · JuMP"/><meta property="og:title" content="COPT-Public/COPT.jl · JuMP"/><meta property="twitter:title" content="COPT-Public/COPT.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li class="is-active"><a class="tocitem" href>COPT-Public/COPT.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li></ul></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>COPT-Public/COPT.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>COPT-Public/COPT.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/COPT-Public/COPT.jl/blob/v1.1.7/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="COPT.jl"><a class="docs-heading-anchor" href="#COPT.jl">COPT.jl</a><a id="COPT.jl-1"></a><a class="docs-heading-anchor-permalink" href="#COPT.jl" title="Permalink"></a></h1><p><a href="https://github.com/COPT-Public/COPT.jl">COPT.jl</a> is a wrapper for the <a href="https://www.shanshu.ai/copt">COPT (Cardinal Optimizer)</a>, a mathematical optimization solver for large-scale optimization problems.</p><p>COPT includes high-performance solvers for LP, MIP, SOCP, convex QP/QCP and SDP.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>COPT.jl is licensed under the <a href="https://github.com/COPT-Public/COPT.jl/blob/main/LICENSE">MIT license</a>.</p><p>The underlying solver is a closed-source commercial product for which you must <a href="https://www.shanshu.ai/copt">obtain a license</a>.</p><h3 id="Note"><a class="docs-heading-anchor" href="#Note">Note</a><a id="Note-1"></a><a class="docs-heading-anchor-permalink" href="#Note" title="Permalink"></a></h3><p>When COPT is upgraded to a newer version, you may see an error message such as <code>ERROR: COPT error 4: Unable to create COPT environment</code>, which indicates that you will need to reapply and upgrade your COPT license files as well.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install COPT using the Julia package manager</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;COPT&quot;)</code></pre><p>When there is no local version of COPT installed, installing COPT.jl will automatically download the necessary solver binaries.</p><p>Without a license, you can solve small models for non-commercial purpose. We strongly recommend that you apply for a license by following the link above.</p><h3 id="Note-2"><a class="docs-heading-anchor" href="#Note-2">Note</a><a class="docs-heading-anchor-permalink" href="#Note-2" title="Permalink"></a></h3><p><strong>MacOS Apple M1/ARM:</strong> on MacOS with Apple M1 chips, Intel based programs can run via <em>Rosetta</em>. When you installed the COPT binaries manually, then please make sure that the <em>COPT build</em> matches the <em>Julia build</em>. We recommend the Intel based COPT and Julia build, as the Apple M1/ARM build of Julia is experimental.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use COPT with JuMP, use <code>COPT.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP
+using COPT
+model = Model(COPT.Optimizer)
+@variable(model, x &gt;= 0)
+@variable(model, 0 &lt;= y &lt;= 3)
+@objective(model, Min, 12x + 20y)
+@constraint(model, c1, 6x + 8y &gt;= 100)
+@constraint(model, c2, 7x + 12y &gt;= 120)
+print(model)
+optimize!(model)
+@show termination_status(model)
+@show primal_status(model)
+@show dual_status(model)
+@show objective_value(model)
+@show value(x)
+@show value(y)
+@show shadow_price(c1)
+@show shadow_price(c2)</code></pre><p>To use the semidefinite programming solver in COPT with JuMP, use <code>COPT.ConeOptimizer</code>:</p><pre><code class="language-julia hljs">using JuMP
+using COPT
+using LinearAlgebra
+model = Model(COPT.ConeOptimizer)
+C = [1.0 -1.0; -1.0 2.0]
+@variable(model, X[1:2, 1:2], PSD)
+@variable(model, z[1:2] &gt;= 0)
+@objective(model, Min, C ⋅ X)
+@constraint(model, c1, X[1, 1] - z[1] == 1)
+@constraint(model, c2, X[2, 2] - z[2] == 1)
+optimize!(model)
+@show termination_status(model)
+@show primal_status(model)
+@show dual_status(model)
+@show objective_value(model)
+@show value.(X)
+@show value.(z)
+@show shadow_price(c1)
+@show shadow_price(c2)</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Xpress/">« jump-dev/Xpress.jl</a><a class="docs-footer-nextpage" href="../GAMS/">GAMS-dev/GAMS.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/COSMO/index.html b/previews/PR3525/packages/COSMO/index.html
new file mode 100644
index 00000000000..fa6fd2b9952
--- /dev/null
+++ b/previews/PR3525/packages/COSMO/index.html
@@ -0,0 +1,37 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>oxfordcontrol/COSMO.jl · JuMP</title><meta name="title" content="oxfordcontrol/COSMO.jl · JuMP"/><meta property="og:title" content="oxfordcontrol/COSMO.jl · JuMP"/><meta property="twitter:title" content="oxfordcontrol/COSMO.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li class="is-active"><a class="tocitem" href>oxfordcontrol/COSMO.jl</a><ul class="internal"><li><a class="tocitem" href="#Features"><span>Features</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Citing"><span>Citing</span></a></li><li><a class="tocitem" href="#Contributing"><span>Contributing</span></a></li><li><a class="tocitem" href="#Python-Interface"><span>Python - Interface</span></a></li><li><a class="tocitem" href="#Licence"><span>Licence 🔍</span></a></li></ul></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>oxfordcontrol/COSMO.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>oxfordcontrol/COSMO.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/oxfordcontrol/COSMO.jl/blob/v0.8.8/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><div style="display: flex; justify-content: flex-end" margin=0px>
+    <img src="https://github.com/migarstka/COSMO_assets/blob/master/star_badge_3.png?raw=true" align="right" width=6%>
+</div><h1 align="center" margin=0px>
+  <img src="https://github.com/migarstka/COSMO_assets/blob/master/cosmo_rocket_with_convex_set.png?raw=true" width=70%>
+</h1><p align="center">
+   <a href="https://github.com/oxfordcontrol/COSMO.jl/actions"><img src="https://github.com/oxfordcontrol/COSMO.jl/workflows/ci/badge.svg?branch=master"></a>
+  <a href="https://codecov.io/gh/oxfordcontrol/COSMO.jl"><img src="https://codecov.io/gh/oxfordcontrol/COSMO.jl/branch/master/graph/badge.svg?raw=true"></a>
+  <a href="https://oxfordcontrol.github.io/COSMO.jl/stable"><img src="https://img.shields.io/badge/Documentation-stable-purple.svg?raw=true"></a>
+  <a href="https://opensource.org/licenses/Apache-2.0"><img src="https://img.shields.io/badge/License-Apache%202.0-blue.svg?raw=true"></a>
+  <a href="https://github.com/oxfordcontrol/COSMO.jl/releases"><img src="https://img.shields.io/badge/Release-v0.8.8-blue.svg?raw=true"></a>
+</p><p align="center">
+  <a href="#features">Features</a> •
+  <a href="#installation">Installation</a> •
+  <a href="NEWS.md">News</a> •
+  <a href="#citing-">Citing</a> •
+  <a href="#contributing">Contributing</a>
+</p><p>This is a Julia implementation of the <em>Conic operator splitting method</em> (COSMO) solver. It can solve large convex conic optimization problems of the following form:</p><p align="center">
+<img src="https://github.com/migarstka/COSMO_assets/blob/master/cosmo_format.png?raw=true" width=220px>
+</p><p>with decision variables <code>x ϵ R^n</code>, <code>s ϵ R^m</code> and data matrices <code>P=P&#39;&gt;=0</code>, <code>q ϵ R^n</code>, <code>A ϵ R^(m×n)</code>, and <code>b ϵ R^m</code>. The convex set <code>K</code> is a composition of convex sets and cones.</p><p><strong>For more information take a look at the COSMO.jl Documentation (<a href="https://oxfordcontrol.github.io/COSMO.jl/stable">stable</a> |  <a href="https://oxfordcontrol.github.io/COSMO.jl/dev">dev</a>).</strong></p><h2 id="Features"><a class="docs-heading-anchor" href="#Features">Features</a><a id="Features-1"></a><a class="docs-heading-anchor-permalink" href="#Features" title="Permalink"></a></h2><ul><li><strong>Versatile</strong>: COSMO solves linear programs, quadratic programs, second-order cone programs, semidefinite programs and problems involving exponential and power cones</li><li><strong>Quad SDPs</strong>: Positive semidefinite programs with quadratic objective functions are natively supported</li><li><strong>Safeguarded acceleration</strong>: robust and faster convergence to higher precision using <a href="https://github.com/oxfordcontrol/COSMOAccelerators.jl">COSMOAccelerators</a></li><li><strong>Infeasibility detection</strong>: Infeasible problems are detected without a homogeneous self-dual embedding of the problem</li><li><strong>JuMP / Convex.jl support</strong>: We provide an interface to MathOptInterface (MOI), which allows you to describe your problem in <a href="https://github.com/JuliaOpt/JuMP.jl">JuMP</a> and <a href="https://github.com/JuliaOpt/Convex.jl">Convex.jl</a>.</li><li><strong>Warm starting</strong>: COSMO supports warm starting of the decision variables</li><li><strong>Custom sets and linear solver</strong>: Customize COSMO&#39;s components by defining your own convex constraint sets and by choosing from a number of direct and indirect linear system solvers, for example, <a href="https://github.com/oxfordcontrol/qdldl">QDLDL</a>, <a href="https://github.com/JuliaSparse/Pardiso.jl">Pardiso</a>, <a href="https://juliamath.github.io/IterativeSolvers.jl/dev/">Conjugate Gradient</a> and <a href="https://juliamath.github.io/IterativeSolvers.jl/dev/">MINRES</a></li><li><strong>Arbitrary precision types</strong>: You can solve problems with any floating point precision.</li><li><strong>Open Source</strong>: Our code is free to use and distributed under the Apache 2.0 Licence</li><li><strong>Chordal decomposition</strong>: COSMO tries to decompose large structured PSD constraints using chordal decomposition techniques. This often results in a significant speedup compared to the original problem.</li><li><strong>Smart clique merging</strong>: After an initial decomposition of a structured SDP, COSMO recombines overlapping cliques/blocks to speed up the algorithm.</li></ul><div align="center" margin=0px>
+  <img src="https://github.com/oxfordcontrol/COSMO.jl/blob/master/docs/src/assets/example_decomposition.gif?raw=true" width=45%>
+</div><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><ul><li><code>COSMO</code> can be added via the Julia package manager (type <code>]</code>): <code>pkg&gt; add COSMO</code></li></ul><h2 id="Citing"><a class="docs-heading-anchor" href="#Citing">Citing</a><a id="Citing-1"></a><a class="docs-heading-anchor-permalink" href="#Citing" title="Permalink"></a></h2><p>If you find COSMO useful in your project, we kindly request that you cite the following paper:</p><pre><code class="nohighlight hljs">@Article{Garstka_2021,
+  author  = {Michael Garstka and Mark Cannon and Paul Goulart},
+  journal = {Journal of Optimization Theory and Applications},
+  title   = {{COSMO}: A Conic Operator Splitting Method for Convex Conic Problems},
+  volume  = {190},
+  number  = {3},
+  pages   = {779--810},
+  year    = {2021},
+  publisher = {Springer},
+  doi     = {10.1007/s10957-021-01896-x},
+  url     = {https://doi.org/10.1007/s10957-021-01896-x}
+}</code></pre><p>The article is available under Open Access <a href="https://link.springer.com/article/10.1007/s10957-021-01896-x">here</a>.</p><h2 id="Contributing"><a class="docs-heading-anchor" href="#Contributing">Contributing</a><a id="Contributing-1"></a><a class="docs-heading-anchor-permalink" href="#Contributing" title="Permalink"></a></h2><ul><li>Contributions are always welcome. Our style guide can be found <a href="https://github.com/oxfordcontrol/COSMO.jl/wiki/Code-Style-Guide">here</a>.</li><li>Current issues, tasks and future ideas are listed in <a href="https://github.com/oxfordcontrol/ossdp/issues">Issues</a>. Please report any issues or bugs that you encounter.</li><li>As an open source project we are also interested in any projects and applications that use COSMO. Please let us know by opening a GitHub issue.</li></ul><h2 id="Python-Interface"><a class="docs-heading-anchor" href="#Python-Interface">Python - Interface</a><a id="Python-Interface-1"></a><a class="docs-heading-anchor-permalink" href="#Python-Interface" title="Permalink"></a></h2><p>COSMO can also be called from Python. Take a look at: <a href="https://github.com/oxfordcontrol/cosmo-python">cosmo-python</a></p><h2 id="Licence"><a class="docs-heading-anchor" href="#Licence">Licence 🔍</a><a id="Licence-1"></a><a class="docs-heading-anchor-permalink" href="#Licence" title="Permalink"></a></h2><p>This project is licensed under the Apache License - see the <a href="https://github.com/oxfordcontrol/COSMO.jl/blob/master/LICENSE.md">LICENSE.md</a> file for details.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../CDCS/">« oxfordcontrol/CDCS.jl</a><a class="docs-footer-nextpage" href="../Clarabel/">oxfordcontrol/Clarabel.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/CPLEX/index.html b/previews/PR3525/packages/CPLEX/index.html
new file mode 100644
index 00000000000..82548c4bed6
--- /dev/null
+++ b/previews/PR3525/packages/CPLEX/index.html
@@ -0,0 +1,164 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/CPLEX.jl · JuMP</title><meta name="title" content="jump-dev/CPLEX.jl · JuMP"/><meta property="og:title" content="jump-dev/CPLEX.jl · JuMP"/><meta property="twitter:title" content="jump-dev/CPLEX.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/CPLEX.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Callbacks"><span>Callbacks</span></a></li><li><a class="tocitem" href="#Annotations-for-automatic-Benders&#39;-decomposition"><span>Annotations for automatic Benders&#39; decomposition</span></a></li></ul></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/CPLEX.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/CPLEX.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/CPLEX.jl/blob/v1.0.0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="CPLEX.jl"><a class="docs-heading-anchor" href="#CPLEX.jl">CPLEX.jl</a><a id="CPLEX.jl-1"></a><a class="docs-heading-anchor-permalink" href="#CPLEX.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/CPLEX.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/CPLEX.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/CPLEX.jl"><img src="https://codecov.io/gh/jump-dev/CPLEX.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/CPLEX.jl">CPLEX.jl</a> is a wrapper for the <a href="https://www.ibm.com/products/ilog-cplex-optimization-studio">IBM® ILOG® CPLEX® Optimization Studio</a>.</p><p>CPLEX.jl has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><p>The C API can be accessed via <code>CPLEX.CPXxx</code> functions, where the names and arguments are identical to the C API. See the <a href="https://www.ibm.com/support/knowledgecenter/SSSA5P_12.10.0/COS_KC_home.html">CPLEX documentation</a> for details.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not officially supported by IBM. However, we thank IBM for providing us with a CPLEX license to test <code>CPLEX.jl</code> on GitHub. If you are a commercial customer interested in official support for CPLEX in Julia, let them know.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>CPLEX.jl</code> is licensed under the <a href="https://github.com/jump-dev/CPLEX.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver is a closed-source commercial product for which you must <a href="https://www.ibm.com/analytics/cplex-optimizer">purchase a license</a>.</p><p>Free CPLEX licenses are available for <a href="http://ibm.biz/Bdzvqw">academics and students</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>CPLEX.jl requires CPLEX version 12.10, 20.1, or 22.1.</p><p>First, obtain a license of CPLEX and install CPLEX solver, following the instructions on <a href="https://www.ibm.com/analytics/cplex-optimizer">IBM&#39;s website</a>.</p><p>Once installed, set the <code>CPLEX_STUDIO_BINARIES</code> environment variable as appropriate and run <code>Pkg.add(&quot;CPLEX&quot;)</code>. For example:</p><pre><code class="language-julia hljs"># On Windows, this might be:
+ENV[&quot;CPLEX_STUDIO_BINARIES&quot;] = &quot;C:\\Program Files\\CPLEX_Studio1210\\cplex\\bin\\x86-64_win\\&quot;
+# On OSX, this might be:
+ENV[&quot;CPLEX_STUDIO_BINARIES&quot;] = &quot;/Applications/CPLEX_Studio1210/cplex/bin/x86-64_osx/&quot;
+# On Unix, this might be:
+ENV[&quot;CPLEX_STUDIO_BINARIES&quot;] = &quot;/opt/CPLEX_Studio1210/cplex/bin/x86-64_linux/&quot;
+
+import Pkg
+Pkg.add(&quot;CPLEX&quot;)</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The exact path may differ. Check which folder you installed CPLEX in, and update the path accordingly.</p></div></div><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use <code>CPLEX.jl</code> with JuMP as follows:</p><pre><code class="language-julia hljs">using JuMP, CPLEX
+model = Model(CPLEX.Optimizer)
+set_attribute(model, &quot;CPX_PARAM_EPINT&quot;, 1e-8)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The CPLEX optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VectorAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semicontinuous"><code>MOI.Semicontinuous{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS1"><code>MOI.SOS1{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS2"><code>MOI.SOS2{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ConflictStatus"><code>MOI.ConflictStatus()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.HeuristicCallback"><code>MOI.HeuristicCallback()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.LazyConstraintCallback"><code>MOI.LazyConstraintCallback()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.UserCutCallback"><code>MOI.UserCutCallback()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>Options match those of the C API in the <a href="https://www.ibm.com/support/knowledgecenter/SSSA5P_12.10.0/ilog.odms.cplex.help/CPLEX/Parameters/topics/introListAlpha.html">CPLEX documentation</a>.</p><h2 id="Callbacks"><a class="docs-heading-anchor" href="#Callbacks">Callbacks</a><a id="Callbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Callbacks" title="Permalink"></a></h2><p>CPLEX.jl provides a solver-specific callback to CPLEX:</p><pre><code class="language-julia hljs">using JuMP, CPLEX, Test
+
+model = direct_model(CPLEX.Optimizer())
+set_silent(model)
+
+# This is very, very important!!! Only use callbacks in single-threaded mode.
+MOI.set(model, MOI.NumberOfThreads(), 1)
+
+@variable(model, 0 &lt;= x &lt;= 2.5, Int)
+@variable(model, 0 &lt;= y &lt;= 2.5, Int)
+@objective(model, Max, y)
+cb_calls = Clong[]
+function my_callback_function(cb_data::CPLEX.CallbackContext, context_id::Clong)
+    # You can reference variables outside the function as normal
+    push!(cb_calls, context_id)
+    # You can select where the callback is run
+    if context_id != CPX_CALLBACKCONTEXT_CANDIDATE
+        return
+    end
+    ispoint_p = Ref{Cint}()
+    ret = CPXcallbackcandidateispoint(cb_data, ispoint_p)
+    if ret != 0 || ispoint_p[] == 0
+        return  # No candidate point available or error
+    end
+    # You can query CALLBACKINFO items
+    valueP = Ref{Cdouble}()
+    ret = CPXcallbackgetinfodbl(cb_data, CPXCALLBACKINFO_BEST_BND, valueP)
+    @info &quot;Best bound is currently: $(valueP[])&quot;
+    # As well as any other C API
+    x_p = Vector{Cdouble}(undef, 2)
+    obj_p = Ref{Cdouble}()
+    ret = CPXcallbackgetincumbent(cb_data, x_p, 0, 1, obj_p)
+    if ret == 0
+        @info &quot;Objective incumbent is: $(obj_p[])&quot;
+        @info &quot;Incumbent solution is: $(x_p)&quot;
+        # Use CPLEX.column to map between variable references and the 1-based
+        # column.
+        x_col = CPLEX.column(cb_data, index(x))
+        @info &quot;x = $(x_p[x_col])&quot;
+    else
+        # Unable to query incumbent.
+    end
+
+    # Before querying `callback_value`, you must call:
+    CPLEX.load_callback_variable_primal(cb_data, context_id)
+    x_val = callback_value(cb_data, x)
+    y_val = callback_value(cb_data, y)
+    # You can submit solver-independent MathOptInterface attributes such as
+    # lazy constraints, user-cuts, and heuristic solutions.
+    if y_val - x_val &gt; 1 + 1e-6
+        con = @build_constraint(y - x &lt;= 1)
+        MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+    elseif y_val + x_val &gt; 3 + 1e-6
+        con = @build_constraint(y + x &lt;= 3)
+        MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+    end
+end
+MOI.set(model, CPLEX.CallbackFunction(), my_callback_function)
+optimize!(model)
+@test termination_status(model) == MOI.OPTIMAL
+@test primal_status(model) == MOI.FEASIBLE_POINT
+@test value(x) == 1
+@test value(y) == 2</code></pre><h2 id="Annotations-for-automatic-Benders&#39;-decomposition"><a class="docs-heading-anchor" href="#Annotations-for-automatic-Benders&#39;-decomposition">Annotations for automatic Benders&#39; decomposition</a><a id="Annotations-for-automatic-Benders&#39;-decomposition-1"></a><a class="docs-heading-anchor-permalink" href="#Annotations-for-automatic-Benders&#39;-decomposition" title="Permalink"></a></h2><p>Here is an example of using the annotation feature for automatic Benders&#39; decomposition:</p><pre><code class="language-julia hljs">using JuMP, CPLEX
+
+function add_annotation(
+    model::JuMP.Model,
+    variable_classification::Dict;
+    all_variables::Bool = true,
+)
+    num_variables = sum(length(it) for it in values(variable_classification))
+    if all_variables
+        @assert num_variables == JuMP.num_variables(model)
+    end
+    indices, annotations = CPXINT[], CPXLONG[]
+    for (key, value) in variable_classification
+        for variable_ref in value
+            push!(indices, variable_ref.index.value - 1)
+            push!(annotations, CPX_BENDERS_MASTERVALUE + key)
+        end
+    end
+    cplex = backend(model)
+    index_p = Ref{CPXINT}()
+    CPXnewlongannotation(
+        cplex.env,
+        cplex.lp,
+        CPX_BENDERS_ANNOTATION,
+        CPX_BENDERS_MASTERVALUE,
+    )
+    CPXgetlongannotationindex(
+        cplex.env,
+        cplex.lp,
+        CPX_BENDERS_ANNOTATION,
+        index_p,
+    )
+    CPXsetlongannotations(
+        cplex.env,
+        cplex.lp,
+        index_p[],
+        CPX_ANNOTATIONOBJ_COL,
+        length(indices),
+        indices,
+        annotations,
+    )
+    return
+end
+
+# Problem
+
+function illustrate_full_annotation()
+    c_1, c_2 = [1, 4], [2, 3]
+    dim_x, dim_y = length(c_1), length(c_2)
+    b = [-2; -3]
+    A_1, A_2 = [1 -3; -1 -3], [1 -2; -1 -1]
+    model = JuMP.direct_model(CPLEX.Optimizer())
+    set_optimizer_attribute(model, &quot;CPXPARAM_Benders_Strategy&quot;, 1)
+    @variable(model, x[1:dim_x] &gt;= 0, Bin)
+    @variable(model, y[1:dim_y] &gt;= 0)
+    variable_classification = Dict(0 =&gt; [x[1], x[2]], 1 =&gt; [y[1], y[2]])
+    @constraint(model, A_2 * y + A_1 * x .&lt;= b)
+    @objective(model, Min, c_1&#39; * x + c_2&#39; * y)
+    add_annotation(model, variable_classification)
+    optimize!(model)
+    x_optimal = value.(x)
+    y_optimal = value.(y)
+    println(&quot;x: $(x_optimal), y: $(y_optimal)&quot;)
+end
+
+function illustrate_partial_annotation()
+    c_1, c_2 = [1, 4], [2, 3]
+    dim_x, dim_y = length(c_1), length(c_2)
+    b = [-2; -3]
+    A_1, A_2 = [1 -3; -1 -3], [1 -2; -1 -1]
+    model = JuMP.direct_model(CPLEX.Optimizer())
+    # Note that the &quot;CPXPARAM_Benders_Strategy&quot; has to be set to 2 if partial
+    # annotation is provided. If &quot;CPXPARAM_Benders_Strategy&quot; is set to 1, then
+    # the following error will be thrown:
+    # `CPLEX Error  2002: Invalid Benders decomposition.`
+    set_optimizer_attribute(model, &quot;CPXPARAM_Benders_Strategy&quot;, 2)
+    @variable(model, x[1:dim_x] &gt;= 0, Bin)
+    @variable(model, y[1:dim_y] &gt;= 0)
+    variable_classification = Dict(0 =&gt; [x[1]], 1 =&gt; [y[1], y[2]])
+    @constraint(model, A_2 * y + A_1 * x .&lt;= b)
+    @objective(model, Min, c_1&#39; * x + c_2&#39; * y)
+    add_annotation(model, variable_classification; all_variables = false)
+    optimize!(model)
+    x_optimal = value.(x)
+    y_optimal = value.(y)
+    println(&quot;x: $(x_optimal), y: $(y_optimal)&quot;)
+end</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../BARON/">« jump-dev/BARON.jl</a><a class="docs-footer-nextpage" href="../CSDP/">jump-dev/CSDP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/CSDP/index.html b/previews/PR3525/packages/CSDP/index.html
new file mode 100644
index 00000000000..c8e265b8e89
--- /dev/null
+++ b/previews/PR3525/packages/CSDP/index.html
@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/CSDP.jl · JuMP</title><meta name="title" content="jump-dev/CSDP.jl · JuMP"/><meta property="og:title" content="jump-dev/CSDP.jl · JuMP"/><meta property="twitter:title" content="jump-dev/CSDP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/CSDP.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Problem-representation"><span>Problem representation</span></a></li><li><a class="tocitem" href="#Termination-criteria"><span>Termination criteria</span></a></li><li><a class="tocitem" href="#Status"><span>Status</span></a></li></ul></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/CSDP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/CSDP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/CSDP.jl/blob/3a2e4c7c48af7660a5ba258a4cb6549ab0970366/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><p><img src="https://www.coin-or.org/wordpress/wp-content/uploads/2014/08/COINOR.png" alt/></p><h1 id="CSDP.jl"><a class="docs-heading-anchor" href="#CSDP.jl">CSDP.jl</a><a id="CSDP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#CSDP.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/CSDP.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/CSDP.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/CSDP.jl"><img src="https://codecov.io/gh/jump-dev/CSDP.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/CSDP.jl"><code>CSDP.jl</code></a> is a wrapper for the <a href="https://projects.coin-or.org/Csdp">COIN-OR SemiDefinite Programming</a> solver.</p><p>The wrapper has two components:</p><ul><li>a thin wrapper around the low-level C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not a COIN-OR project.</p><p>The original algorithm is described by B. Borchers (1999). <em>CSDP, A C Library for Semidefinite Programming</em>. Optimization Methods and Software. 11(1), 613-623. <a href="http://euler.nmt.edu/~brian/csdppaper.pdf">[preprint]</a></p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>CSDP.jl</code> is licensed under the <a href="https://github.com/jump-dev/CSDP.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/coin-or/Csdp">coin-or/Csdp</a>, is licensed under the <a href="https://github.com/coin-or/Csdp/blob/master/LICENSE">Eclipse public license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install CSDP using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;CSDP&quot;)</code></pre><p>In addition to installing the CSDP.jl package, this will also download and install the CSDP binaries. You do not need to install CSDP separately.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use CSDP with JuMP, use <code>CSDP.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, CSDP
+model = Model(CSDP.Optimizer)
+set_attribute(model, &quot;maxiter&quot;, 1000)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The CSDP optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>The CSDP options are listed in the table below.</p><table><tr><th style="text-align: right">Name</th><th style="text-align: right">Default Value</th><th style="text-align: right">Explanation</th></tr><tr><td style="text-align: right"><code>axtol</code></td><td style="text-align: right"><code>1.0e-8</code></td><td style="text-align: right">Tolerance for primal feasibility</td></tr><tr><td style="text-align: right"><code>atytol</code></td><td style="text-align: right"><code>1.0e-8</code></td><td style="text-align: right">Tolerance for dual feasibility</td></tr><tr><td style="text-align: right"><code>objtol</code></td><td style="text-align: right"><code>1.0e-8</code></td><td style="text-align: right">Tolerance for relative duality gap</td></tr><tr><td style="text-align: right"><code>pinftol</code></td><td style="text-align: right"><code>1.0e8</code></td><td style="text-align: right">Tolerance for determining primal infeasibility</td></tr><tr><td style="text-align: right"><code>dinftol</code></td><td style="text-align: right"><code>1.0e8</code></td><td style="text-align: right">Tolerance for determining dual infeasibility</td></tr><tr><td style="text-align: right"><code>maxiter</code></td><td style="text-align: right"><code>100</code></td><td style="text-align: right">Limit for the total number of iterations</td></tr><tr><td style="text-align: right"><code>minstepfrac</code></td><td style="text-align: right"><code>0.90</code></td><td style="text-align: right">The <code>minstepfrac</code> and <code>maxstepfrac</code> parameters determine how close to the edge of the feasible region CSDP will step</td></tr><tr><td style="text-align: right"><code>maxstepfrac</code></td><td style="text-align: right"><code>0.97</code></td><td style="text-align: right">The <code>minstepfrac</code> and <code>maxstepfrac</code> parameters determine how close to the edge of the feasible region CSDP will step</td></tr><tr><td style="text-align: right"><code>minstepp</code></td><td style="text-align: right"><code>1.0e-8</code></td><td style="text-align: right">If the primal step is shorter than <code>minstepp</code> then CSDP declares a line search failure</td></tr><tr><td style="text-align: right"><code>minstepd</code></td><td style="text-align: right"><code>1.0e-8</code></td><td style="text-align: right">If the dual step is shorter than <code>minstepd</code> then CSDP declares a line search failure</td></tr><tr><td style="text-align: right"><code>usexzgap</code></td><td style="text-align: right"><code>1</code></td><td style="text-align: right">If <code>usexzgap</code> is <code>0</code> then CSDP will use the objective duality gap <code>d - p</code> instead of the XY duality gap <code>⟨Z, X⟩</code></td></tr><tr><td style="text-align: right"><code>tweakgap</code></td><td style="text-align: right"><code>0</code></td><td style="text-align: right">If <code>tweakgap</code> is set to <code>1</code>, and <code>usexzgap</code> is set to <code>0</code>, then CSDP will attempt to &quot;fix&quot; negative duality gaps</td></tr><tr><td style="text-align: right"><code>affine</code></td><td style="text-align: right"><code>0</code></td><td style="text-align: right">If <code>affine</code> is set to <code>1</code>, then CSDP will take only primal-dual affine steps and not make use of the barrier term. This can be useful for some problems that do not have feasible solutions that are strictly in the interior of the cone of semidefinite matrices</td></tr><tr><td style="text-align: right"><code>perturbobj</code></td><td style="text-align: right"><code>1</code></td><td style="text-align: right">The <code>perturbobj</code> parameter determines whether the objective function will be perturbed to help deal with problems that have unbounded optimal solution sets. If <code>perturbobj</code> is <code>0</code>, then the objective will not be perturbed. If <code>perturbobj</code> is <code>1</code>, then the objective function will be perturbed by a default amount. Larger values of <code>perturbobj</code> (for example, <code>100</code>) increase the size of the perturbation. This can be helpful in solving some difficult problems.</td></tr><tr><td style="text-align: right"><code>fastmode</code></td><td style="text-align: right"><code>0</code></td><td style="text-align: right">The <code>fastmode</code> parameter determines whether or not CSDP will skip certain time consuming operations that slightly improve the accuracy of the solutions. If <code>fastmode</code> is set to <code>1</code>, then CSDP may be somewhat faster, but also somewhat less accurate</td></tr><tr><td style="text-align: right"><code>printlevel</code></td><td style="text-align: right"><code>1</code></td><td style="text-align: right">The <code>printlevel</code> parameter determines how much debugging information is output. Use a <code>printlevel</code> of <code>0</code> for no output and a <code>printlevel</code> of <code>1</code> for normal output. Higher values of <code>printlevel</code> will generate more debugging output</td></tr></table><h2 id="Problem-representation"><a class="docs-heading-anchor" href="#Problem-representation">Problem representation</a><a id="Problem-representation-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-representation" title="Permalink"></a></h2><p>The primal is represented internally by CSDP as follows:</p><pre><code class="nohighlight hljs">max ⟨C, X⟩
+      A(X) = a
+         X ⪰ 0</code></pre><p>where <code>A(X) = [⟨A_1, X⟩, ..., ⟨A_m, X⟩]</code>. The corresponding dual is:</p><pre><code class="nohighlight hljs">min ⟨a, y⟩
+     A&#39;(y) - C = Z
+             Z ⪰ 0</code></pre><p>where <code>A&#39;(y) = y_1A_1 + ... + y_mA_m</code></p><h2 id="Termination-criteria"><a class="docs-heading-anchor" href="#Termination-criteria">Termination criteria</a><a id="Termination-criteria-1"></a><a class="docs-heading-anchor-permalink" href="#Termination-criteria" title="Permalink"></a></h2><p>CSDP will terminate successfully (or partially) in the following cases:</p><ul><li>If CSDP finds <code>X, Z ⪰ 0</code> such that the following 3 tolerances are satisfied:<ul><li>primal feasibility tolerance: <code>||A(x) - a||_2 / (1 + ||a||_2) &lt; axtol</code></li><li>dual feasibility tolerance: <code>||A&#39;(y) - C - Z||_F / (1 + ||C||_F) &lt; atytol</code></li><li>relative duality gap tolerance: <code>gap / (1 + |⟨a, y⟩| + |⟨C, X⟩|) &lt; objtol</code><ul><li>objective duality gap: if <code>usexygap</code> is <code>0</code>, <code>gap = ⟨a, y⟩ - ⟨C, X⟩</code></li><li>XY duality gap: if <code>usexygap</code> is <code>1</code>, <code>gap = ⟨Z, X⟩</code></li></ul></li></ul>then it returns <code>0</code>.</li><li>If CSDP finds <code>y</code> and <code>Z ⪰ 0</code> such that <code>-⟨a, y⟩ / ||A&#39;(y) - Z||_F &gt; pinftol</code>, it returns <code>1</code> with <code>y</code> such that <code>⟨a, y⟩ = -1</code>.</li><li>If CSDP finds <code>X ⪰ 0</code> such that <code>⟨C, X⟩ / ||A(X)||_2 &gt; dinftol</code>, it returns <code>2</code> with <code>X</code> such that <code>⟨C, X⟩ = 1</code>.</li><li>If CSDP finds <code>X, Z ⪰ 0</code> such that the following 3 tolerances are satisfied with <code>1000*axtol</code>, <code>1000*atytol</code> and <code>1000*objtol</code> but at least one of them is not satisfied with <code>axtol</code>, <code>atytol</code> and <code>objtol</code> and cannot make progress, then it returns <code>3</code>.</li></ul><p>In addition, if the <code>printlevel</code> option is at least <code>1</code>, the following will be printed:</p><ul><li>If the return code is <code>1</code>, CSDP will print <code>⟨a, y⟩</code> and <code>||A&#39;(y) - Z||_F</code></li><li>If the return code is <code>2</code>, CSDP will print <code>⟨C, X⟩</code> and <code>||A(X)||_F</code></li><li>Otherwise, CSDP will print<ul><li>the primal/dual objective value,</li><li>the relative primal/dual infeasibility,</li><li>the objective duality gap <code>⟨a, y⟩ - ⟨C, X⟩</code> and objective relative duality gap <code>(⟨a, y⟩ - ⟨C, X⟩) / (1 + |⟨a, y⟩| + |⟨C, X⟩|)</code>,</li><li>the XY duality gap <code>⟨Z, X⟩</code> and XY relative duality gap <code>⟨Z, X⟩ / (1 + |⟨a, y⟩| + |⟨C, X⟩|)</code></li><li>and the DIMACS error measures.</li></ul></li></ul><p>In theory, for feasible primal and dual solutions, <code>⟨a, y⟩ - ⟨C, X⟩ = ⟨Z, X⟩</code>, so the objective and XY duality gap should be equivalent. However, in practice, there are sometimes solution which satisfy primal and dual feasibility tolerances but have objective duality gap which are not close to XY duality gap. In some cases, the objective duality gap may even become negative (hence the <code>tweakgap</code> option). This is the reason <code>usexygap</code> is <code>1</code> by default.</p><p>CSDP considers that <code>X ⪰ 0</code> (resp. <code>Z ⪰ 0</code>) is satisfied when the Cholesky factorizations can be computed. In practice, this is somewhat more conservative than simply requiring all eigenvalues to be nonnegative.</p><h2 id="Status"><a class="docs-heading-anchor" href="#Status">Status</a><a id="Status-1"></a><a class="docs-heading-anchor-permalink" href="#Status" title="Permalink"></a></h2><p>The table below shows how the different CSDP statuses are converted to the MathOptInterface statuses.</p><table><tr><th style="text-align: right">CSDP code</th><th style="text-align: right">State</th><th style="text-align: right">Description</th><th style="text-align: right">MOI status</th></tr><tr><td style="text-align: right"><code>0</code></td><td style="text-align: right">Success</td><td style="text-align: right">SDP solved</td><td style="text-align: right"><code>MOI.OPTIMAL</code></td></tr><tr><td style="text-align: right"><code>1</code></td><td style="text-align: right">Success</td><td style="text-align: right">The problem is primal infeasible, and we have a certificate</td><td style="text-align: right"><code>MOI.INFEASIBLE</code></td></tr><tr><td style="text-align: right"><code>2</code></td><td style="text-align: right">Success</td><td style="text-align: right">The problem is dual infeasible, and we have a certificate</td><td style="text-align: right"><code>MOI.DUAL_INFEASIBLE</code></td></tr><tr><td style="text-align: right"><code>3</code></td><td style="text-align: right">Partial Success</td><td style="text-align: right">A solution has been found, but full accuracy was not achieved</td><td style="text-align: right"><code>MOI.ALMOST_OPTIMAL</code></td></tr><tr><td style="text-align: right"><code>4</code></td><td style="text-align: right">Failure</td><td style="text-align: right">Maximum iterations reached</td><td style="text-align: right"><code>MOI.ITERATION_LIMIT</code></td></tr><tr><td style="text-align: right"><code>5</code></td><td style="text-align: right">Failure</td><td style="text-align: right">Stuck at edge of primal feasibility</td><td style="text-align: right"><code>MOI.SLOW_PROGRESS</code></td></tr><tr><td style="text-align: right"><code>6</code></td><td style="text-align: right">Failure</td><td style="text-align: right">Stuck at edge of dual infeasibility</td><td style="text-align: right"><code>MOI.SLOW_PROGRESS</code></td></tr><tr><td style="text-align: right"><code>7</code></td><td style="text-align: right">Failure</td><td style="text-align: right">Lack of progress</td><td style="text-align: right"><code>MOI.SLOW_PROGRESS</code></td></tr><tr><td style="text-align: right"><code>8</code></td><td style="text-align: right">Failure</td><td style="text-align: right"><code>X</code>, <code>Z</code>, or <code>O</code> was singular</td><td style="text-align: right"><code>MOI.NUMERICAL_ERROR</code></td></tr><tr><td style="text-align: right"><code>9</code></td><td style="text-align: right">Failure</td><td style="text-align: right">Detected <code>NaN</code> or <code>Inf</code> values</td><td style="text-align: right"><code>MOI.NUMERICAL_ERROR</code></td></tr></table></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../CPLEX/">« jump-dev/CPLEX.jl</a><a class="docs-footer-nextpage" href="../Cbc/">jump-dev/Cbc.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Cbc/index.html b/previews/PR3525/packages/Cbc/index.html
new file mode 100644
index 00000000000..9c8ff49dfe8
--- /dev/null
+++ b/previews/PR3525/packages/Cbc/index.html
@@ -0,0 +1,12 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Cbc.jl · JuMP</title><meta name="title" content="jump-dev/Cbc.jl · JuMP"/><meta property="og:title" content="jump-dev/Cbc.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Cbc.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Cbc.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li></ul></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Cbc.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Cbc.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Cbc.jl/blob/v1.2.0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><p><a href="https://www.coin-or.org"><img src="https://www.coin-or.org/wordpress/wp-content/uploads/2014/08/COINOR.png" alt/></a></p><h1 id="Cbc.jl"><a class="docs-heading-anchor" href="#Cbc.jl">Cbc.jl</a><a id="Cbc.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Cbc.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Cbc.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/Cbc.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Cbc.jl"><img src="https://codecov.io/gh/jump-dev/Cbc.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/Cbc.jl">Cbc.jl</a> is a wrapper for the <a href="https://projects.coin-or.org/Cbc">COIN-OR Branch and Cut (Cbc)</a> solver.</p><p>The wrapper has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not a COIN-OR project.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Cbc.jl</code> is licensed under the <a href="https://github.com/jump-dev/Cbc.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/coin-or/Cbc">coin-or/Cbc</a>, is licensed under the <a href="https://github.com/coin-or/Cbc/blob/master/LICENSE">Eclipse public license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Cbc using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Cbc&quot;)</code></pre><p>In addition to installing the Cbc.jl package, this will also download and install the Cbc binaries. You do not need to install Cbc separately.</p><p>To use a custom binary, read the <a href="https://jump.dev/JuMP.jl/stable/developers/custom_solver_binaries/">Custom solver binaries</a> section of the JuMP documentation.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use Cbc with JuMP, use <code>Cbc.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, Cbc
+model = Model(Cbc.Optimizer)
+set_attribute(model, &quot;logLevel&quot;, 1)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The COIN Branch-and-Cut (Cbc) optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS1"><code>MOI.SOS1{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS2"><code>MOI.SOS2{Float64}</code></a></li></ul><p>List of supported model attributes:</p><ul><li><code>Cbc.Status</code></li><li><code>Cbc.SecondaryStatus</code></li><li><a href="../../moi/reference/models/#MathOptInterface.DualStatus"><code>MOI.DualStatus</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.NodeCount"><code>MOI.NodeCount</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.NumberOfVariables"><code>MOI.NumberOfVariables</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveBound"><code>MOI.ObjectiveBound</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveValue"><code>MOI.ObjectiveValue</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.PrimalStatus"><code>MOI.PrimalStatus</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.RelativeGap"><code>MOI.RelativeGap</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ResultCount"><code>MOI.ResultCount</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.SolveTimeSec"><code>MOI.SolveTimeSec</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.TerminationStatus"><code>MOI.TerminationStatus</code></a></li></ul><p>List of supported optimizer attributes:</p><ul><li><code>Cbc.SetVariableNames</code></li><li><a href="../../moi/reference/models/#MathOptInterface.AbsoluteGapTolerance"><code>MOI.AbsoluteGapTolerance</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.NumberOfThreads"><code>MOI.NumberOfThreads</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.RawOptimizerAttribute"><code>MOI.RawOptimizerAttribute</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.RelativeGapTolerance"><code>MOI.RelativeGapTolerance</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.Silent"><code>MOI.Silent</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.SolverName"><code>MOI.SolverName</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.SolverVersion"><code>MOI.SolverVersion</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.TimeLimitSec"><code>MOI.TimeLimitSec</code></a></li></ul><p>List of supported variable attributes:</p><ul><li><a href="../../moi/reference/variables/#MathOptInterface.VariablePrimal"><code>MOI.VariablePrimal</code></a></li><li><a href="../../moi/reference/variables/#MathOptInterface.VariablePrimalStart"><code>MOI.VariablePrimalStart</code></a></li><li><a href="../../moi/reference/variables/#MathOptInterface.VariableName"><code>MOI.VariableName</code></a></li></ul><p>List of supported constraint attributes:</p><ul><li><a href="../../moi/reference/constraints/#MathOptInterface.ConstraintPrimal"><code>MOI.ConstraintPrimal</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>Options are, unfortunately, not well documented.</p><p>The following options are likely to be the most useful:</p><table><tr><th style="text-align: right">Parameter</th><th style="text-align: right">Example</th><th style="text-align: right">Explanation</th></tr><tr><td style="text-align: right"><code>seconds</code></td><td style="text-align: right"><code>60.0</code></td><td style="text-align: right">Solution timeout limit</td></tr><tr><td style="text-align: right"><code>logLevel</code></td><td style="text-align: right"><code>2</code></td><td style="text-align: right">Set to 0 to disable solution output</td></tr><tr><td style="text-align: right"><code>maxSolutions</code></td><td style="text-align: right"><code>1</code></td><td style="text-align: right">Terminate after this many feasible solutions have been found</td></tr><tr><td style="text-align: right"><code>maxNodes</code></td><td style="text-align: right"><code>1</code></td><td style="text-align: right">Terminate after this many branch-and-bound nodes have been evaluated</td></tr><tr><td style="text-align: right"><code>allowableGap</code></td><td style="text-align: right"><code>0.05</code></td><td style="text-align: right">Terminate after optimality gap is less than this value (on an absolute scale)</td></tr><tr><td style="text-align: right"><code>ratioGap</code></td><td style="text-align: right"><code>0.05</code></td><td style="text-align: right">Terminate after optimality gap is smaller than this relative fraction</td></tr><tr><td style="text-align: right"><code>threads</code></td><td style="text-align: right"><code>1</code></td><td style="text-align: right">Set the number of threads to use for parallel branch &amp; bound</td></tr></table><p>The complete list of parameters can be found by running the <code>cbc</code> executable and typing <code>?</code> at the prompt.</p><p>Start the <code>cbc</code> executable from Julia as follows:</p><pre><code class="language-julia hljs">using Cbc_jll
+Cbc_jll.cbc() do exe
+    run(`$(exe)`)
+end</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../CSDP/">« jump-dev/CSDP.jl</a><a class="docs-footer-nextpage" href="../Clp/">jump-dev/Clp.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Clarabel/index.html b/previews/PR3525/packages/Clarabel/index.html
new file mode 100644
index 00000000000..92b496b6926
--- /dev/null
+++ b/previews/PR3525/packages/Clarabel/index.html
@@ -0,0 +1,26 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>oxfordcontrol/Clarabel.jl · JuMP</title><meta name="title" content="oxfordcontrol/Clarabel.jl · JuMP"/><meta property="og:title" content="oxfordcontrol/Clarabel.jl · JuMP"/><meta property="twitter:title" content="oxfordcontrol/Clarabel.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li class="is-active"><a class="tocitem" href>oxfordcontrol/Clarabel.jl</a><ul class="internal"><li><a class="tocitem" href="#Features"><span>Features</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#License"><span>License 🔍</span></a></li></ul></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>oxfordcontrol/Clarabel.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>oxfordcontrol/Clarabel.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/oxfordcontrol/Clarabel.jl/blob/v0.5.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 align="center" margin=0px>
+  <img src="https://github.com/oxfordcontrol/ClarabelDocs/blob/main/docs/src/assets/logo-banner-light-jl.png?raw=true" class="display-light-only" width=60%>
+  <img src="https://github.com/oxfordcontrol/ClarabelDocs/blob/main/docs/src/assets/logo-banner-dark-jl.png?raw=true" class="display-dark-only"   width=60%>
+  <br>
+Interior Point Conic Optimization for Julia
+</h1><p align="center">
+   <a href="https://github.com/oxfordcontrol/Clarabel.jl/actions"><img src="https://github.com/oxfordcontrol/Clarabel.jl/workflows/ci/badge.svg?branch=main"></a>
+  <a href="https://codecov.io/gh/oxfordcontrol/Clarabel.jl"><img src="https://codecov.io/gh/oxfordcontrol/Clarabel.jl/branch/main/graph/badge.svg?raw=true"></a>
+  <a href="https://oxfordcontrol.github.io/ClarabelDocs/stable"><img src="https://img.shields.io/badge/Documentation-stable-purple.svg?raw=true"></a>
+  <a href="https://opensource.org/licenses/Apache-2.0"><img src="https://img.shields.io/badge/License-Apache%202.0-blue.svg?raw=true"></a>
+  <a href="https://github.com/oxfordcontrol/Clarabel.jl/releases"><img src="https://img.shields.io/badge/Release-v0.5.1-blue.svg?raw=true"></a>
+</p><p align="center">
+  <a href="#features">Features</a> •
+  <a href="#installation">Installation</a> •
+  <a href="#license-">License</a> •
+  <a href="https://oxfordcontrol.github.io/ClarabelDocs/stable">Documentation</a>
+</p><p><strong>Clarabel.jl</strong> is a Julia implementation of an interior point numerical solver for convex optimization problems using a novel homogeneous embedding.  Clarabel.jl solves the following problem:</p><p class="math-container">\[\begin{array}{r}
+\text{minimize} &amp; \frac{1}{2}x^T P x + q^T x\\\\[2ex]
+\text{subject to} &amp; Ax + s = b \\\\[1ex]
+        &amp; s \in \mathcal{K}
+\end{array}\]</p><p>with decision variables <span>$x \in \mathbb{R}^n$</span>, <span>$s \in \mathbb{R}^m$</span> and data matrices <span>$P=P^\top \succeq 0$</span>, <span>$q \in \mathbb{R}^n$</span>, <span>$A \in \mathbb{R}^{m \times n}$</span>, and <span>$b \in \mathbb{R}^m$</span>. The convex set <span>$\mathcal{K}$</span> is a composition of convex cones.</p><p><strong>For more information see the Clarabel Documentation (<a href="https://oxfordcontrol.github.io/ClarabelDocs/stable">stable</a> |  <a href="https://oxfordcontrol.github.io/ClarabelDocs/dev">dev</a>).</strong></p><p>Clarabel is also available in a Rust / Python implementation.  See <a href="https://github.com/oxfordcontrol/Clarabel.rs">here</a>.</p><h2 id="Features"><a class="docs-heading-anchor" href="#Features">Features</a><a id="Features-1"></a><a class="docs-heading-anchor-permalink" href="#Features" title="Permalink"></a></h2><ul><li><strong>Versatile</strong>: Clarabel.jl solves linear programs (LPs), quadratic programs (QPs), second-order cone programs (SOCPs) and semidefinite programs (SDPs).  It also solves problems with exponential and power cone constraints.</li><li><strong>Quadratic objectives</strong>: Unlike interior point solvers based on the standard homogeneous self-dual embedding (HSDE), Clarabel.jl handles quadratic objectives without requiring any epigraphical reformulation of the objective.   It can therefore be significantly faster than other HSDE-based solvers for problems with quadratic objective functions.</li><li><strong>Infeasibility detection</strong>: Infeasible problems are detected using a homogeneous embedding technique.</li><li><strong>JuMP / Convex.jl support</strong>: We provide an interface to <a href="https://jump.dev/JuMP.jl/stable/moi/">MathOptInterface</a> (MOI), which allows you to describe your problem in <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> and <a href="https://github.com/jump-dev/Convex.jl">Convex.jl</a>.</li><li><strong>Arbitrary precision types</strong>: You can solve problems with any floating point precision, for example, Float32 or Julia&#39;s BigFloat type, using either the native interface, or via MathOptInterface / Convex.jl.</li><li><strong>Open Source</strong>: Our code is available on <a href="https://github.com/oxfordcontrol/Clarabel.jl">GitHub</a> and distributed under the Apache 2.0 License</li></ul><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><ul><li><strong>Clarabel.jl</strong> can be added via the Julia package manager (type <code>]</code>): <code>pkg&gt; add Clarabel</code></li></ul><h2 id="License"><a class="docs-heading-anchor" href="#License">License 🔍</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>This project is licensed under the Apache License  2.0 - see the <a href="https://github.com/oxfordcontrol/Clarabel.jl/blob/main/LICENSE.md">LICENSE.md</a> file for details.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../COSMO/">« oxfordcontrol/COSMO.jl</a><a class="docs-footer-nextpage" href="../SCIP/">scipopt/SCIP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Clp/index.html b/previews/PR3525/packages/Clp/index.html
new file mode 100644
index 00000000000..283601c476b
--- /dev/null
+++ b/previews/PR3525/packages/Clp/index.html
@@ -0,0 +1,10 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Clp.jl · JuMP</title><meta name="title" content="jump-dev/Clp.jl · JuMP"/><meta property="og:title" content="jump-dev/Clp.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Clp.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Clp.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#C-API"><span>C API</span></a></li></ul></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Clp.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Clp.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Clp.jl/blob/v1.0.3/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><p><img src="https://www.coin-or.org/wordpress/wp-content/uploads/2014/08/COINOR.png" alt/></p><h1 id="Clp.jl"><a class="docs-heading-anchor" href="#Clp.jl">Clp.jl</a><a id="Clp.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Clp.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Clp.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/Clp.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Clp.jl"><img src="https://codecov.io/gh/jump-dev/Clp.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/Clp.jl">Clp.jl</a> is a wrapper for the <a href="https://projects.coin-or.org/Clp">COIN-OR Linear Programming</a> solver.</p><p>The wrapper has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not a COIN-OR project.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Clp.jl</code> is licensed under the <a href="https://github.com/jump-dev/Clp.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/coin-or/Clp">coin-or/Clp</a>, is licensed under the <a href="https://github.com/coin-or/Clp/blob/master/LICENSE">Eclipse public license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Clp using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Clp&quot;)</code></pre><p>In addition to installing the Clp.jl package, this will also download and install the Clp binaries. You do not need to install Clp separately.</p><p>To use a custom binary, read the <a href="https://jump.dev/JuMP.jl/stable/developers/custom_solver_binaries/">Custom solver binaries</a> section of the JuMP documentation.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use Clp with JuMP, use <code>Clp.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, Clp
+model = Model(Clp.Optimizer)
+set_attribute(model, &quot;LogLevel&quot;, 1)
+set_attribute(model, &quot;Algorithm&quot;, 4)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The Clp optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>Options are, unfortunately, not well documented.</p><p>The following options are likely to be the most useful:</p><table><tr><th style="text-align: right">Parameter</th><th style="text-align: right">Example</th><th style="text-align: right">Explanation</th></tr><tr><td style="text-align: right"><code>PrimalTolerance</code></td><td style="text-align: right"><code>1e-7</code></td><td style="text-align: right">Primal feasibility tolerance</td></tr><tr><td style="text-align: right"><code>DualTolerance</code></td><td style="text-align: right"><code>1e-7</code></td><td style="text-align: right">Dual feasibility tolerance</td></tr><tr><td style="text-align: right"><code>DualObjectiveLimit</code></td><td style="text-align: right"><code>1e308</code></td><td style="text-align: right">When using dual simplex (where the objective is monotonically changing), terminate when the objective exceeds this limit</td></tr><tr><td style="text-align: right"><code>MaximumIterations</code></td><td style="text-align: right"><code>2147483647</code></td><td style="text-align: right">Terminate after performing this number of simplex iterations</td></tr><tr><td style="text-align: right"><code>MaximumSeconds</code></td><td style="text-align: right"><code>-1.0</code></td><td style="text-align: right">Terminate after this many seconds have passed. A negative value means no time limit</td></tr><tr><td style="text-align: right"><code>LogLevel</code></td><td style="text-align: right"><code>1</code></td><td style="text-align: right">Set to 1, 2, 3, or 4 for increasing output. Set to <code>0</code> to disable output</td></tr><tr><td style="text-align: right"><code>PresolveType</code></td><td style="text-align: right"><code>0</code></td><td style="text-align: right">Set to <code>1</code> to disable presolve</td></tr><tr><td style="text-align: right"><code>SolveType</code></td><td style="text-align: right"><code>5</code></td><td style="text-align: right">Solution method: dual simplex (<code>0</code>), primal simplex (<code>1</code>), sprint (<code>2</code>), barrier with crossover (<code>3</code>), barrier without crossover (<code>4</code>), automatic (<code>5</code>)</td></tr><tr><td style="text-align: right"><code>InfeasibleReturn</code></td><td style="text-align: right"><code>0</code></td><td style="text-align: right">Set to 1 to return as soon as the problem is found to be infeasible (by default, an infeasibility proof is computed as well)</td></tr><tr><td style="text-align: right"><code>Scaling</code></td><td style="text-align: right"><code>3</code></td><td style="text-align: right"><code>0</code> -off, <code>1</code> equilibrium, <code>2</code> geometric, <code>3</code> auto, <code>4</code> dynamic(later)</td></tr><tr><td style="text-align: right"><code>Perturbation</code></td><td style="text-align: right"><code>100</code></td><td style="text-align: right">switch on perturbation (<code>50</code>), automatic (<code>100</code>), don&#39;t try perturbing (<code>102</code>)</td></tr></table><h2 id="C-API"><a class="docs-heading-anchor" href="#C-API">C API</a><a id="C-API-1"></a><a class="docs-heading-anchor-permalink" href="#C-API" title="Permalink"></a></h2><p>The C API can be accessed via <code>Clp.Clp_XXX</code> functions, where the names and arguments are identical to the C API.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Cbc/">« jump-dev/Cbc.jl</a><a class="docs-footer-nextpage" href="../Dualization/">jump-dev/Dualization.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/DAQP/index.html b/previews/PR3525/packages/DAQP/index.html
new file mode 100644
index 00000000000..439482bc8d8
--- /dev/null
+++ b/previews/PR3525/packages/DAQP/index.html
@@ -0,0 +1,8 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>darnstrom/DAQP.jl · JuMP</title><meta name="title" content="darnstrom/DAQP.jl · JuMP"/><meta property="og:title" content="darnstrom/DAQP.jl · JuMP"/><meta property="twitter:title" content="darnstrom/DAQP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li class="is-active"><a class="tocitem" href>darnstrom/DAQP.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li></ul></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>darnstrom/DAQP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>darnstrom/DAQP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/darnstrom/DAQP.jl/blob/v0.5.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="DAQP.jl"><a class="docs-heading-anchor" href="#DAQP.jl">DAQP.jl</a><a id="DAQP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#DAQP.jl" title="Permalink"></a></h1><p><a href="https://github.com/darnstrom/DAQP.jl/actions"><img src="https://github.com/darnstrom/DAQP.jl/workflows/CI/badge.svg" alt/></a> <a href="http://codecov.io/github/darnstrom/DAQP.jl"><img src="http://codecov.io/gh/darnstrom/DAQP.jl/graphs/badge.svg" alt/></a></p><p><a href="https://github.com/darnstrom/DAQP.jl">DAQP.jl</a> is a Julia wrapper for the Quadratic Programming solver <a href="https://github.com/darnstrom/daqp">DAQP</a>.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>DAQP.jl is licensed under the <a href="https://github.com/darnstrom/DAQP.jl/blob/main/LICENSE">MIT license</a>.</p><p>The underlying solver, <a href="https://github.com/darnstrom/daqp">darnstrom/daqp</a> is licensed under the <a href="https://github.com/darnstrom/daqp/blob/master/LICENSE">MIT license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install DAQP.jl using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;DAQP&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use DAQP with JuMP, do:</p><pre><code class="language-julia hljs">using JuMP, DAQP
+model = Model(DAQP.Optimizer)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>General information about the solver is available at <a href="https://darnstrom.github.io/daqp/">https://darnstrom.github.io/daqp/</a>, and specifics for the Julia interface are available at <a href="https://darnstrom.github.io/daqp/start/julia">https://darnstrom.github.io/daqp/start/julia</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Hypatia/">« chriscoey/Hypatia.jl</a><a class="docs-footer-nextpage" href="../Tulip/">ds4dm/Tulip.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/DiffOpt/index.html b/previews/PR3525/packages/DiffOpt/index.html
new file mode 100644
index 00000000000..b4bfc0f77b5
--- /dev/null
+++ b/previews/PR3525/packages/DiffOpt/index.html
@@ -0,0 +1,23 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/DiffOpt.jl · JuMP</title><meta name="title" content="jump-dev/DiffOpt.jl · JuMP"/><meta property="og:title" content="jump-dev/DiffOpt.jl · JuMP"/><meta property="twitter:title" content="jump-dev/DiffOpt.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li class="is-active"><a class="tocitem" href>jump-dev/DiffOpt.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#GSOC2020"><span>GSOC2020</span></a></li></ul></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>jump-dev/DiffOpt.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/DiffOpt.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/DiffOpt.jl/blob/v0.4.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="DiffOpt.jl"><a class="docs-heading-anchor" href="#DiffOpt.jl">DiffOpt.jl</a><a id="DiffOpt.jl-1"></a><a class="docs-heading-anchor-permalink" href="#DiffOpt.jl" title="Permalink"></a></h1><p><a href="https://jump.dev/DiffOpt.jl/stable"><img src="https://img.shields.io/badge/docs-stable-blue.svg" alt="stable docs"/></a> <a href="https://jump.dev/DiffOpt.jl/dev"><img src="https://img.shields.io/badge/docs-dev-blue.svg" alt="development docs"/></a> <a href="https://github.com/jump-dev/DiffOpt.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/DiffOpt.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/DiffOpt.jl"><img src="https://codecov.io/gh/jump-dev/DiffOpt.jl/branch/master/graph/badge.svg" alt="Coverage"/></a></p><p><a href="https://github.com/jump-dev/DiffOpt.jl">DiffOpt.jl</a> is a package for differentiating convex optimization programs with respect to the program parameters. DiffOpt currently supports linear, quadratic, and conic programs.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>DiffOpt.jl</code> is licensed under the <a href="https://github.com/jump-dev/DiffOpt.jl/blob/master/LICENSE.md">MIT License</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install DiffOpt using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;DiffOpt&quot;)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>The <a href="https://jump.dev/DiffOpt.jl/stable/">documentation for DiffOpt.jl</a> includes a detailed description of the theory behind the package, along with examples, tutorials, and an API reference.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use DiffOpt with JuMP by following this brief example:</p><pre><code class="language-julia hljs">using JuMP, DiffOpt, HiGHS
+# Create a model using the wrapper
+model = Model(() -&gt; DiffOpt.diff_optimizer(HiGHS.Optimizer))
+# Define your model and solve it
+@variable(model, x)
+@constraint(model, cons, x &gt;= 3)
+@objective(model, Min, 2x)
+optimize!(model)
+# Choose the problem parameters to differentiate with respect to, and set their
+# perturbations.
+MOI.set(model, DiffOpt.ReverseVariablePrimal(), x, 1.0)
+# Differentiate the model
+DiffOpt.reverse_differentiate!(model)
+# fetch the gradients
+grad_exp = MOI.get(model, DiffOpt.ReverseConstraintFunction(), cons)  # -3 x - 1
+constant(grad_exp)        # -1
+coefficient(grad_exp, x)  # -3</code></pre><h2 id="GSOC2020"><a class="docs-heading-anchor" href="#GSOC2020">GSOC2020</a><a id="GSOC2020-1"></a><a class="docs-heading-anchor-permalink" href="#GSOC2020" title="Permalink"></a></h2><p>DiffOpt began as a <a href="https://summerofcode.withgoogle.com/organizations/4727917315096576/?sp-page=2#5232064888045568">NumFOCUS sponsored Google Summer of Code (2020) project</a></p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../extensions/introduction/">« Introduction</a><a class="docs-footer-nextpage" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Dualization/index.html b/previews/PR3525/packages/Dualization/index.html
new file mode 100644
index 00000000000..67b384294a9
--- /dev/null
+++ b/previews/PR3525/packages/Dualization/index.html
@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Dualization.jl · JuMP</title><meta name="title" content="jump-dev/Dualization.jl · JuMP"/><meta property="og:title" content="jump-dev/Dualization.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Dualization.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Dualization.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li></ul></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Dualization.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Dualization.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Dualization.jl/blob/v0.5.8/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Dualization.jl"><a class="docs-heading-anchor" href="#Dualization.jl">Dualization.jl</a><a id="Dualization.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Dualization.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Dualization.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/Dualization.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Dualization.jl"><img src="https://codecov.io/gh/jump-dev/Dualization.jl/branch/master/graph/badge.svg" alt="codecov"/></a> <a href="https://zenodo.org/badge/latestdoi/182854997"><img src="https://zenodo.org/badge/182854997.svg" alt="DOI"/></a></p><p><a href="https://github.com/jump-dev/Dualization.jl">Dualization.jl</a> is an extension package for <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface.jl</a> that formulates the dual of conic optimization problems.</p><p>Dualization.jl has two main features:</p><ul><li>The <code>Dualization.dualize</code> function that computes the dual formulation of either a <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface.jl</a> or a <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> model.</li><li>The <code>Dualization.dual_optimizer</code> function that creates a MathOptInterface-compatible optimizer that solves the dual of the problem instead of the primal.</li></ul><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Dualization.jl</code> is licensed under the <a href="https://github.com/jump-dev/Dualization.jl/blob/master/LICENSE.md">MIT License</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Dualization using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Dualization&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To compute the dual formulation of a JuMP model, use <code>dualize</code>:</p><pre><code class="language-julia hljs">using JuMP, Dualization
+model = Model()
+# ... build model ...
+dual_model = dualize(model)</code></pre><p>To solve the dual formulation of a JuMP model, create a <code>dual_optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, Dualization, SCS
+model = Model(dual_optimizer(SCS.Optimizer))
+# ... build model ...
+optimize!(model)  # Solves the dual instead of the primal</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>The <a href="https://jump.dev/Dualization.jl/stable/">documentation for Dualization.jl</a> includes a detailed description of the dual reformulation, along with examples and an API reference.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Clp/">« jump-dev/Clp.jl</a><a class="docs-footer-nextpage" href="../ECOS/">jump-dev/ECOS.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/ECOS/index.html b/previews/PR3525/packages/ECOS/index.html
new file mode 100644
index 00000000000..8a68acf090a
--- /dev/null
+++ b/previews/PR3525/packages/ECOS/index.html
@@ -0,0 +1,9 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/ECOS.jl · JuMP</title><meta name="title" content="jump-dev/ECOS.jl · JuMP"/><meta property="og:title" content="jump-dev/ECOS.jl · JuMP"/><meta property="twitter:title" content="jump-dev/ECOS.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/ECOS.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li></ul></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/ECOS.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/ECOS.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/ECOS.jl/blob/v1.1.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="ECOS.jl"><a class="docs-heading-anchor" href="#ECOS.jl">ECOS.jl</a><a id="ECOS.jl-1"></a><a class="docs-heading-anchor-permalink" href="#ECOS.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/ECOS.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/ECOS.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/ECOS.jl"><img src="https://codecov.io/gh/jump-dev/ECOS.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/ECOS.jl">ECOS.jl</a> is a wrapper for the <a href="https://github.com/embotech/ecos">ECOS</a> solver.</p><p>The wrapper has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not a product of Embotech AG.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>ECOS.jl</code> is licensed under the <a href="https://github.com/jump-dev/ECOS.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/embotech/ecos">embotech/ecos</a>, is licensed under the <a href="https://github.com/embotech/ecos/blob/develop/COPYING">GPL v3 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install ECOS.jl using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;ECOS&quot;)</code></pre><p>In addition to installing the ECOS.jl package, this will also download and install the ECOS binaries. You do not need to install ECOS separately.</p><p>To use a custom binary, read the <a href="https://jump.dev/JuMP.jl/stable/developers/custom_solver_binaries/">Custom solver binaries</a> section of the JuMP documentation.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use ECOS with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>, use <code>ECOS.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, ECOS
+model = Model(ECOS.Optimizer)
+set_attribute(model, &quot;maxit&quot;, 100)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The ECOS optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>The following options are supported:</p><table><tr><th style="text-align: right">Parameter</th><th style="text-align: right">Explanation</th></tr><tr><td style="text-align: right"><code>gamma</code></td><td style="text-align: right">scaling the final step length</td></tr><tr><td style="text-align: right"><code>delta</code></td><td style="text-align: right">regularization parameter</td></tr><tr><td style="text-align: right"><code>eps</code></td><td style="text-align: right">regularization threshold</td></tr><tr><td style="text-align: right"><code>feastol</code></td><td style="text-align: right">primal/dual infeasibility tolerance</td></tr><tr><td style="text-align: right"><code>abstol</code></td><td style="text-align: right">absolute tolerance on duality gap</td></tr><tr><td style="text-align: right"><code>reltol</code></td><td style="text-align: right">relative tolerance on duality gap</td></tr><tr><td style="text-align: right"><code>feastol_inacc</code></td><td style="text-align: right">primal/dual infeasibility relaxed tolerance</td></tr><tr><td style="text-align: right"><code>abstol_inacc</code></td><td style="text-align: right">absolute relaxed tolerance on duality gap</td></tr><tr><td style="text-align: right"><code>reltol_inacc</code></td><td style="text-align: right">relative relaxed tolerance on duality gap</td></tr><tr><td style="text-align: right"><code>nitref</code></td><td style="text-align: right">number of iterative refinement steps</td></tr><tr><td style="text-align: right"><code>maxit</code></td><td style="text-align: right">maximum number of iterations</td></tr><tr><td style="text-align: right"><code>verbose</code></td><td style="text-align: right">verbosity <code>bool</code> for <code>PRINTLEVEL &lt; 3</code></td></tr></table></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Dualization/">« jump-dev/Dualization.jl</a><a class="docs-footer-nextpage" href="../GLPK/">jump-dev/GLPK.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/GAMS/index.html b/previews/PR3525/packages/GAMS/index.html
new file mode 100644
index 00000000000..a02e68db388
--- /dev/null
+++ b/previews/PR3525/packages/GAMS/index.html
@@ -0,0 +1,25 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>GAMS-dev/GAMS.jl · JuMP</title><meta name="title" content="GAMS-dev/GAMS.jl · JuMP"/><meta property="og:title" content="GAMS-dev/GAMS.jl · JuMP"/><meta property="twitter:title" content="GAMS-dev/GAMS.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li class="is-active"><a class="tocitem" href>GAMS-dev/GAMS.jl</a><ul class="internal"><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Usage"><span>Usage</span></a></li></ul></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>GAMS-dev/GAMS.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>GAMS-dev/GAMS.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/GAMS-dev/GAMS.jl/blob/c5dee9f929e9d2f4433ae09fa92b8d872c9c43e0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="GAMS.jl"><a class="docs-heading-anchor" href="#GAMS.jl">GAMS.jl</a><a id="GAMS.jl-1"></a><a class="docs-heading-anchor-permalink" href="#GAMS.jl" title="Permalink"></a></h1><p>GAMS.jl provides a <a href="https://github.com/JuliaOpt/MathOptInterface.jl">MathOptInterface</a> Optimizer to solve <a href="https://github.com/JuliaOpt/JuMP.jl">JuMP</a> models using <a href="https://www.gams.com/">GAMS</a>.</p><p>GAMS comes with dozens of supported solvers. Among them are: <a href="https://www.gams.com/latest/docs/S_ALPHAECP.html">ALPHAECP</a>, <a href="https://www.gams.com/latest/docs/S_ANTIGONE.html">ANTIGONE</a>, <a href="https://www.gams.com/latest/docs/S_BARON.html">BARON</a>, <a href="https://www.gams.com/latest/docs/S_CBC.html">CBC</a>, <a href="https://www.gams.com/latest/docs/S_CONOPT.html">CONOPT</a>, <a href="https://www.gams.com/latest/docs/S_CPLEX.html">CPLEX</a>, <a href="https://www.gams.com/latest/docs/S_DICOPT.html">DICOPT</a>, <a href="https://www.gams.com/latest/docs/S_GUROBI.html">GUROBI</a>, <a href="https://www.gams.com/latest/docs/S_IPOPT.html">IPOPT</a>, <a href="https://www.gams.com/latest/docs/S_KNITRO.html">KNITRO</a>, <a href="https://www.gams.com/latest/docs/S_LINDO.html">LINDO</a>, <a href="https://www.gams.com/latest/docs/S_LINDO.html">LINDOGLOBAL</a>, <a href="https://www.gams.com/latest/docs/S_MINOS.html">MINOS</a>, <a href="https://www.gams.com/latest/docs/S_MOSEK.html">MOSEK</a>, <a href="https://www.gams.com/latest/docs/S_NLPEC.html">NLPEC</a>, <a href="https://www.gams.com/latest/docs/S_PATH.html">PATH</a>, <a href="https://www.gams.com/latest/docs/S_MINOS.html">QUADMINOS</a>, <a href="https://www.gams.com/latest/docs/S_SBB.html">SBB</a>, <a href="https://www.gams.com/latest/docs/S_SHOT.html">SHOT</a>, <a href="https://www.gams.com/latest/docs/S_SCIP.html">SCIP</a>, <a href="https://www.gams.com/latest/docs/S_SNOPT.html">SNOPT</a>, <a href="https://www.gams.com/latest/docs/S_SOPLEX.html">SOPLEX</a>, <a href="https://www.gams.com/latest/docs/S_XPRESS.html">XPRESS</a>. Find a complete list <a href="https://www.gams.com/latest/docs/S_MAIN.html">here</a>.</p><p>GAMS.jl supports the following JuMP features:</p><ul><li>linear, quadratic and nonlinear (convex and non-convex) objective and constraints</li><li>continuous, binary, integer, semi-continuous and semi-integer variables</li><li>SOS1 and SOS2 sets</li><li>complementarity constraints</li></ul><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><ol><li><a href="https://www.gams.com/download/">Download GAMS</a> and obtain a GAMS license. Please note that GAMS also offers a <a href="https://www.gams.com/latest/docs/UG_License.html#GAMS_Community_Licenses">free community license</a>.</li><li>(optional) Add the GAMS system directory to the <code>PATH</code> variable in order to find GAMS automatically.</li><li>Install GAMS.jl using the Julia package manager:<pre><code class="language-julia hljs">using Pkg
+Pkg.add(&quot;GAMS&quot;)</code></pre></li></ol><h2 id="Usage"><a class="docs-heading-anchor" href="#Usage">Usage</a><a id="Usage-1"></a><a class="docs-heading-anchor-permalink" href="#Usage" title="Permalink"></a></h2><p>Using GAMS as optimizer for your JuMP model:</p><pre><code class="language-julia hljs">using GAMS, JuMP
+model = Model(GAMS.Optimizer)</code></pre><h3 id="GAMS-System"><a class="docs-heading-anchor" href="#GAMS-System">GAMS System</a><a id="GAMS-System-1"></a><a class="docs-heading-anchor-permalink" href="#GAMS-System" title="Permalink"></a></h3><p>If the GAMS system directory has been added to the <code>PATH</code> variable (you can check this with <code>print(ENV[&quot;PATH&quot;])</code>), GAMS.jl will find it automatically. Otherwise, or if you like to switch between systems, the system directory can be specified by (one of the following):</p><pre><code class="language-julia hljs">set_optimizer_attribute(model, &quot;SysDir&quot;, &quot;&lt;gams_system_dir&gt;&quot;)
+set_optimizer_attribute(model, GAMS.SysDir(), &quot;&lt;gams_system_dir&gt;&quot;)</code></pre><p>Analogously, you can specify a working directory with <code>&quot;WorkDir&quot;</code> or <code>GAMS.WorkDir()</code>. If no working directory has been set, GAMS.jl will create a temporary one.</p><p>If you want to use the same GAMS workspace (same system and working directory) for multiple models, you can create a <code>GAMSWorkspace</code> first with either of the following</p><pre><code class="language-julia hljs">ws = GAMS.GAMSWorkspace()
+ws = GAMS.GAMSWorkspace(&quot;&lt;gams_system_dir&gt;&quot;)
+ws = GAMS.GAMSWorkspace(&quot;&lt;gams_system_dir&gt;&quot;, &quot;&lt;gams_working_dir&gt;&quot;)</code></pre><p>and then pass it to your models:</p><pre><code class="language-julia hljs">model = Model(() -&gt; GAMS.Optimizer(ws))</code></pre><h3 id="GAMS-Options"><a class="docs-heading-anchor" href="#GAMS-Options">GAMS Options</a><a id="GAMS-Options-1"></a><a class="docs-heading-anchor-permalink" href="#GAMS-Options" title="Permalink"></a></h3><p><a href="https://www.gams.com/latest/docs/UG_GamsCall.html#UG_GamsCall_ListOfCommandLineParameters">GAMS command line options</a> can be specified by</p><pre><code class="language-julia hljs">set_optimizer_attribute(model, &quot;&lt;option&gt;&quot;, &quot;&lt;solver_name&gt;&quot;)
+set_optimizer_attribute(model, GAMS.&lt;option&gt;(), &quot;&lt;solver_name&gt;&quot;)</code></pre><p>where <code>&lt;option&gt;</code> is either <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOholdfixed">HoldFixed</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOiterlim">IterLim</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOlicense">License</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOlogoption">LogOption</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOnodlim">NodLim</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOoptca">OptCA</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOoptcr">OptCR</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOreslim">ResLim</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOsolver">Solver</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOthreads">Threads</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOtrace">Trace</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOtraceopt">TraceOpt</a> as well as <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOlp">LP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOmip">MIP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOrmip">RMIP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOnlp">NLP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOdnlp">DNLP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOcns">CNS</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOminlp">MINLP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOrminlp">RMINLP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOqcp">QCP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOmiqcp">MIQCP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOrmiqcp">RMIQCP</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOmcp">MCP</a> or <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOmpec">MPEC</a>. Note that <code>GAMS.ResLim()</code> is equivalent to <code>MOI.TimeLimitSec()</code> and <code>GAMS.Threads()</code> to <code>MOI.NumberOfThreads()</code>. Options <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOlimcol">LimCol</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOlimrow">LimRow</a>, <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOsolprint">SolPrint</a> and <a href="https://www.gams.com/latest/docs/UG_GamsCall.html#GAMSAOsolvelink">SolveLink</a> cannot be changed and are set to <code>0</code>, <code>0</code>, <code>0</code> and <code>5</code>, respectively.</p><h4 id="Model-Type"><a class="docs-heading-anchor" href="#Model-Type">Model Type</a><a id="Model-Type-1"></a><a class="docs-heading-anchor-permalink" href="#Model-Type" title="Permalink"></a></h4><p>GAMS.jl will automatically choose a <a href="https://www.gams.com/latest/docs/UG_ModelSolve.html#UG_ModelSolve_ModelClassificationOfModels">GAMS model type</a> for you. Choosing a different model type:</p><pre><code class="language-julia hljs">set_optimizer_attribute(model, GAMS.ModelType(), &quot;&lt;model_type&gt;&quot;)</code></pre><h3 id="GAMS-Solver-Options"><a class="docs-heading-anchor" href="#GAMS-Solver-Options">GAMS Solver Options</a><a id="GAMS-Solver-Options-1"></a><a class="docs-heading-anchor-permalink" href="#GAMS-Solver-Options" title="Permalink"></a></h3><p>Specifying GAMS solver options:</p><pre><code class="language-julia hljs">set_optimizer_attribute(model, &quot;&lt;solver_option_name&gt;&quot;, &lt;option_value&gt;)</code></pre><p>Note that passing a solver option is only valid when explicitly choosing a GAMS solver and not using the default.</p><h3 id="GAMS-Names-vs.-JuMP-Names"><a class="docs-heading-anchor" href="#GAMS-Names-vs.-JuMP-Names">GAMS Names vs. JuMP Names</a><a id="GAMS-Names-vs.-JuMP-Names-1"></a><a class="docs-heading-anchor-permalink" href="#GAMS-Names-vs.-JuMP-Names" title="Permalink"></a></h3><p>GAMS uses generated variable and constraint names although it is possible to pass the JuMP names to the GAMS optimizer, because GAMS is more restrictive when it comes to variable and constraint naming. Use the attributes <code>GeneratedVariableName</code>, <code>GeneratedConstraintName</code>, <code>OriginalVariableName</code>, <code>OriginalConstraintName</code> to query a GAMS symbol name from a JuMP symbol and vice versa. This can help for debugging, for example in case of GAMS compilation errors. For example:</p><pre><code class="language-julia hljs">using GAMS
+
+model = direct_model(GAMS.Optimizer())
+
+@variable(model, x[1:2,1:3] &gt;= 0)
+@constraint(model, c[i = 1:2], sum(x[i,j] for j = 1:3) &lt;= 10)
+
+MOI.get(model, GAMS.GeneratedVariableName(), x[2,2]) # returns x4
+MOI.get(model, GAMS.OriginalVariableName(&quot;x6&quot;))      # returns x[2,3]
+MOI.get(model, GAMS.OriginalVariableName(&quot;x10&quot;))     # returns nothing
+
+MOI.get(model, GAMS.GeneratedConstraintName(), c[2]) # returns eq2
+MOI.get(model, GAMS.OriginalConstraintName(&quot;eq1&quot;))   # returns c[1]
+MOI.get(model, GAMS.OriginalConstraintName(&quot;eq10&quot;))  # returns nothing</code></pre><p>Note that JuMP direct-mode is used.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../COPT/">« COPT-Public/COPT.jl</a><a class="docs-footer-nextpage" href="../NLopt/">JuliaOpt/NLopt.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/GLPK/index.html b/previews/PR3525/packages/GLPK/index.html
new file mode 100644
index 00000000000..9099deccd9a
--- /dev/null
+++ b/previews/PR3525/packages/GLPK/index.html
@@ -0,0 +1,39 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/GLPK.jl · JuMP</title><meta name="title" content="jump-dev/GLPK.jl · JuMP"/><meta property="og:title" content="jump-dev/GLPK.jl · JuMP"/><meta property="twitter:title" content="jump-dev/GLPK.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/GLPK.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Callbacks"><span>Callbacks</span></a></li><li><a class="tocitem" href="#C-API"><span>C API</span></a></li></ul></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/GLPK.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/GLPK.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/GLPK.jl/blob/v1.1.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="GLPK.jl"><a class="docs-heading-anchor" href="#GLPK.jl">GLPK.jl</a><a id="GLPK.jl-1"></a><a class="docs-heading-anchor-permalink" href="#GLPK.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/GLPK.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/GLPK.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/GLPK.jl"><img src="https://codecov.io/gh/jump-dev/GLPK.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/GLPK.jl">GLPK.jl</a> is a wrapper for the <a href="https://www.gnu.org/software/glpk">GNU Linear Programming Kit library</a>.</p><p>The wrapper has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not an GNU project.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>GLPK.jl</code> is licensed under the <a href="https://github.com/jump-dev/GLPK.jl/blob/master/LICENSE.md">GPL v3 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install GLPK using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;GLPK&quot;)</code></pre><p>In addition to installing the GLPK.jl package, this will also download and install the GLPK binaries. You do not need to install GLPK separately.</p><p>To use a custom binary, read the <a href="https://jump.dev/JuMP.jl/stable/developers/custom_solver_binaries/">Custom solver binaries</a> section of the JuMP documentation.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use GLPK with JuMP, use <code>GLPK.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, GLPK
+model = Model(GLPK.Optimizer)
+set_attribute(model, &quot;tm_lim&quot;, 60 * 1_000)
+set_attribute(model, &quot;msg_lev&quot;, GLPK.GLP_MSG_OFF)</code></pre><p>If the model is primal or dual infeasible, GLPK will attempt to find a certificate of infeasibility. This can be expensive, particularly if you do not intend to use the certificate. If this is the case, use:</p><pre><code class="language-julia hljs">model = Model(() -&gt; GLPK.Optimizer(; want_infeasibility_certificates = false))</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The GLPK optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/callbacks/#MathOptInterface.HeuristicCallback"><code>MOI.HeuristicCallback()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.LazyConstraintCallback"><code>MOI.LazyConstraintCallback()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.UserCutCallback"><code>MOI.UserCutCallback()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>Options for GLPK are comprehensively documented in the <a href="https://github.com/jump-dev/GLPK.jl/files/11143880/glpk.pdf">PDF documentation</a>, but they are hard to find.</p><ul><li>Options when solving a linear program are defined in Section 2.8.1</li><li>Options when solving a mixed-integer program are defined in Section 2.10.5</li></ul><p>However, the following options are likely to be the most useful:</p><table><tr><th style="text-align: right">Parameter</th><th style="text-align: right">Example</th><th style="text-align: right">Explanation</th></tr><tr><td style="text-align: right"><code>msg_lev</code></td><td style="text-align: right"><code>GLPK.GLP_MSG_ALL</code></td><td style="text-align: right">Message level for terminal output</td></tr><tr><td style="text-align: right"><code>presolve</code></td><td style="text-align: right"><code>GLPK.GLP_ON</code></td><td style="text-align: right">Turn presolve on or off</td></tr><tr><td style="text-align: right"><code>tol_int</code></td><td style="text-align: right"><code>1e-5</code></td><td style="text-align: right">Absolute tolerance for integer feasibility</td></tr><tr><td style="text-align: right"><code>tol_obj</code></td><td style="text-align: right"><code>1e-7</code></td><td style="text-align: right">Relative objective tolerance for mixed-integer programs</td></tr></table><h2 id="Callbacks"><a class="docs-heading-anchor" href="#Callbacks">Callbacks</a><a id="Callbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Callbacks" title="Permalink"></a></h2><p>Here is an example using GLPK&#39;s solver-specific callbacks.</p><pre><code class="language-julia hljs">using JuMP, GLPK, Test
+
+model = Model(GLPK.Optimizer)
+@variable(model, 0 &lt;= x &lt;= 2.5, Int)
+@variable(model, 0 &lt;= y &lt;= 2.5, Int)
+@objective(model, Max, y)
+reasons = UInt8[]
+function my_callback_function(cb_data)
+    reason = GLPK.glp_ios_reason(cb_data.tree)
+    push!(reasons, reason)
+    if reason != GLPK.GLP_IROWGEN
+        return
+    end
+    x_val = callback_value(cb_data, x)
+    y_val = callback_value(cb_data, y)
+    if y_val - x_val &gt; 1 + 1e-6
+        con = @build_constraint(y - x &lt;= 1)
+        MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+    elseif y_val + x_val &gt; 3 + 1e-6
+        con = @build_constraint(y - x &lt;= 1)
+        MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+    end
+end
+MOI.set(model, GLPK.CallbackFunction(), my_callback_function)
+optimize!(model)
+@test termination_status(model) == MOI.OPTIMAL
+@test primal_status(model) == MOI.FEASIBLE_POINT
+@test value(x) == 1
+@test value(y) == 2
+@show reasons</code></pre><h2 id="C-API"><a class="docs-heading-anchor" href="#C-API">C API</a><a id="C-API-1"></a><a class="docs-heading-anchor-permalink" href="#C-API" title="Permalink"></a></h2><p>The C API can be accessed via <code>GLPK.glp_XXX</code> functions, where the names and arguments are identical to the C API. See the <code>/tests</code> folder for inspiration.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../ECOS/">« jump-dev/ECOS.jl</a><a class="docs-footer-nextpage" href="../Gurobi/">jump-dev/Gurobi.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Gurobi/index.html b/previews/PR3525/packages/Gurobi/index.html
new file mode 100644
index 00000000000..ea78212d848
--- /dev/null
+++ b/previews/PR3525/packages/Gurobi/index.html
@@ -0,0 +1,97 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Gurobi.jl · JuMP</title><meta name="title" content="jump-dev/Gurobi.jl · JuMP"/><meta property="og:title" content="jump-dev/Gurobi.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Gurobi.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Gurobi.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#C-API"><span>C API</span></a></li><li><a class="tocitem" href="#Reusing-the-same-Gurobi-environment-for-multiple-solves"><span>Reusing the same Gurobi environment for multiple solves</span></a></li><li><a class="tocitem" href="#Accessing-Gurobi-specific-attributes"><span>Accessing Gurobi-specific attributes</span></a></li><li><a class="tocitem" href="#Callbacks"><span>Callbacks</span></a></li><li><a class="tocitem" href="#Common-errors"><span>Common errors</span></a></li></ul></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Gurobi.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Gurobi.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Gurobi.jl/blob/v1.0.3/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Gurobi.jl"><a class="docs-heading-anchor" href="#Gurobi.jl">Gurobi.jl</a><a id="Gurobi.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Gurobi.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Gurobi.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/Gurobi.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Gurobi.jl"><img src="https://codecov.io/gh/jump-dev/Gurobi.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/Gurobi.jl">Gurobi.jl</a> is a wrapper for the <a href="https://www.gurobi.com">Gurobi Optimizer</a>.</p><p>It has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not officially supported by Gurobi. However, we thank Gurobi for providing us with a license to test Gurobi.jl on GitHub. If you are a commercial customer interested in official support for Gurobi in Julia, let them know.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Gurobi.jl</code> is licensed under the <a href="https://github.com/jump-dev/Gurobi.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver is a closed-source commercial product for which you must <a href="https://www.gurobi.com">obtain a license</a>.</p><p>Free Gurobi licenses are available for <a href="https://www.gurobi.com/academia/academic-program-and-licenses/">academics and students</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, obtain a license of Gurobi and install Gurobi solver.</p><p>Then, set the <code>GUROBI_HOME</code> environment variable as appropriate and run <code>Pkg.add(&quot;Gurobi&quot;)</code>:</p><pre><code class="language-julia hljs"># On Windows, this might be
+ENV[&quot;GUROBI_HOME&quot;] = &quot;C:\\Program Files\\gurobi1000\\win64&quot;
+# ... or perhaps ...
+ENV[&quot;GUROBI_HOME&quot;] = &quot;C:\\gurobi1000\\win64&quot;
+# On Mac, this might be
+ENV[&quot;GUROBI_HOME&quot;] = &quot;/Library/gurobi1000/mac64&quot;
+
+import Pkg
+Pkg.add(&quot;Gurobi&quot;)</code></pre><p><strong>Note: your path may differ. Check which folder you installed Gurobi in, and update the path accordingly.</strong></p><p>By default, building Gurobi.jl will fail if the Gurobi library is not found. This may not be desirable in certain cases, for example when part of a package&#39;s test suite uses Gurobi as an optional test dependency, but Gurobi cannot be installed on a CI server running the test suite. To support this use case, the <code>GUROBI_JL_SKIP_LIB_CHECK</code> environment variable may be set (to any value) to make Gurobi.jl installable (but not usable).</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use Gurobi with JuMP, use <code>Gurobi.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, Gurobi
+model = Model(Gurobi.Optimizer)
+set_attribute(model, &quot;TimeLimit&quot;, 100)
+set_attribute(model, &quot;Presolve&quot;, 0)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The Gurobi optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VectorAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semicontinuous"><code>MOI.Semicontinuous{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS1"><code>MOI.SOS1{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS2"><code>MOI.SOS2{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/callbacks/#MathOptInterface.HeuristicCallback"><code>MOI.HeuristicCallback()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.LazyConstraintCallback"><code>MOI.LazyConstraintCallback()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.UserCutCallback"><code>MOI.UserCutCallback()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>See the <a href="https://www.gurobi.com/documentation/current/refman/parameters.html">Gurobi Documentation</a> for a list and description of allowable parameters.</p><h2 id="C-API"><a class="docs-heading-anchor" href="#C-API">C API</a><a id="C-API-1"></a><a class="docs-heading-anchor-permalink" href="#C-API" title="Permalink"></a></h2><p>The C API can be accessed via <code>Gurobi.GRBxx</code> functions, where the names and arguments are identical to the C API.</p><p>See the <a href="https://www.gurobi.com/documentation/current/refman/c_api_details.html">Gurobi documentation</a> for details.</p><h2 id="Reusing-the-same-Gurobi-environment-for-multiple-solves"><a class="docs-heading-anchor" href="#Reusing-the-same-Gurobi-environment-for-multiple-solves">Reusing the same Gurobi environment for multiple solves</a><a id="Reusing-the-same-Gurobi-environment-for-multiple-solves-1"></a><a class="docs-heading-anchor-permalink" href="#Reusing-the-same-Gurobi-environment-for-multiple-solves" title="Permalink"></a></h2><p>When using this package via other packages such as <a href="https://github.com/jump-dev/JuMP.jl">JuMP.jl</a>, the default behavior is to obtain a new Gurobi license token every time a model is created. If you are using Gurobi in a setting where the number of concurrent Gurobi uses is limited (for example, <a href="http://www.gurobi.com/products/licensing-pricing/licensing-overview">&quot;Single-Use&quot; or &quot;Floating-Use&quot; licenses</a>), you might instead prefer to obtain a single license token that is shared by all models that your program solves.</p><p>You can do this by passing a <code>Gurobi.Env()</code> object as the first parameter to <code>Gurobi.Optimizer</code>. For example:</p><pre><code class="language-julia hljs">using JuMP, Gurobi
+const GRB_ENV = Gurobi.Env()
+
+model_1 = Model(() -&gt; Gurobi.Optimizer(GRB_ENV))
+
+# The solvers can have different options too
+model_2 = direct_model(Gurobi.Optimizer(GRB_ENV))
+set_attribute(model_2, &quot;OutputFlag&quot;, 0)</code></pre><h2 id="Accessing-Gurobi-specific-attributes"><a class="docs-heading-anchor" href="#Accessing-Gurobi-specific-attributes">Accessing Gurobi-specific attributes</a><a id="Accessing-Gurobi-specific-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Accessing-Gurobi-specific-attributes" title="Permalink"></a></h2><p>Get and set Gurobi-specific variable, constraint, and model attributes as follows:</p><pre><code class="language-julia hljs">using JuMP, Gurobi
+model = direct_model(Gurobi.Optimizer())
+@variable(model, x &gt;= 0)
+@constraint(model, c, 2x &gt;= 1)
+@objective(model, Min, x)
+MOI.set(model, Gurobi.ConstraintAttribute(&quot;Lazy&quot;), c, 2)
+optimize!(model)
+MOI.get(model, Gurobi.VariableAttribute(&quot;LB&quot;), x)  # Returns 0.0
+MOI.get(model, Gurobi.ModelAttribute(&quot;NumConstrs&quot;)) # Returns 1</code></pre><p>A complete list of supported Gurobi attributes can be found in <a href="https://www.gurobi.com/documentation/current/refman/attributes.html">their online documentation</a>.</p><h2 id="Callbacks"><a class="docs-heading-anchor" href="#Callbacks">Callbacks</a><a id="Callbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Callbacks" title="Permalink"></a></h2><p>Here is an example using Gurobi&#39;s solver-specific callbacks.</p><pre><code class="language-julia hljs">using JuMP, Gurobi, Test
+
+model = direct_model(Gurobi.Optimizer())
+@variable(model, 0 &lt;= x &lt;= 2.5, Int)
+@variable(model, 0 &lt;= y &lt;= 2.5, Int)
+@objective(model, Max, y)
+cb_calls = Cint[]
+function my_callback_function(cb_data, cb_where::Cint)
+    # You can reference variables outside the function as normal
+    push!(cb_calls, cb_where)
+    # You can select where the callback is run
+    if cb_where != GRB_CB_MIPSOL &amp;&amp; cb_where != GRB_CB_MIPNODE
+        return
+    end
+    # You can query a callback attribute using GRBcbget
+    if cb_where == GRB_CB_MIPNODE
+        resultP = Ref{Cint}()
+        GRBcbget(cb_data, cb_where, GRB_CB_MIPNODE_STATUS, resultP)
+        if resultP[] != GRB_OPTIMAL
+            return  # Solution is something other than optimal.
+        end
+    end
+    # Before querying `callback_value`, you must call:
+    Gurobi.load_callback_variable_primal(cb_data, cb_where)
+    x_val = callback_value(cb_data, x)
+    y_val = callback_value(cb_data, y)
+    # You can submit solver-independent MathOptInterface attributes such as
+    # lazy constraints, user-cuts, and heuristic solutions.
+    if y_val - x_val &gt; 1 + 1e-6
+        con = @build_constraint(y - x &lt;= 1)
+        MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+    elseif y_val + x_val &gt; 3 + 1e-6
+        con = @build_constraint(y + x &lt;= 3)
+        MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+    end
+    if rand() &lt; 0.1
+        # You can terminate the callback as follows:
+        GRBterminate(backend(model))
+    end
+    return
+end
+# You _must_ set this parameter if using lazy constraints.
+MOI.set(model, MOI.RawOptimizerAttribute(&quot;LazyConstraints&quot;), 1)
+MOI.set(model, Gurobi.CallbackFunction(), my_callback_function)
+optimize!(model)
+@test termination_status(model) == MOI.OPTIMAL
+@test primal_status(model) == MOI.FEASIBLE_POINT
+@test value(x) == 1
+@test value(y) == 2</code></pre><p>See the <a href="https://www.gurobi.com/documentation/current/refman/cb_codes.html">Gurobi documentation</a> for other information that can be queried with <code>GRBcbget</code>.</p><h3 id="Common-Performance-Pitfall-with-JuMP"><a class="docs-heading-anchor" href="#Common-Performance-Pitfall-with-JuMP">Common Performance Pitfall with JuMP</a><a id="Common-Performance-Pitfall-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Common-Performance-Pitfall-with-JuMP" title="Permalink"></a></h3><p>Gurobi&#39;s API works differently than most solvers. Any changes to the model are not applied immediately, but instead go sit in a internal buffer (making any modifications appear to be instantaneous) waiting for a call to <a href="https://www.gurobi.com/documentation/9.0/refman/c_updatemodel.html"><code>GRBupdatemodel</code></a> (where the work is  done).</p><p>This leads to a common performance pitfall that has the following message as its main symptom:</p><pre><code class="nohighlight hljs">Warning: excessive time spent in model updates. Consider calling update less frequently.</code></pre><p>This often means the JuMP program was structured in such a way that Gurobi.jl ends up calling <code>GRBupdatemodel</code> in each iteration of a loop.</p><p>Usually, it is possible (and easy) to restructure the JuMP program in a way it stays ssolver-agnostic and has a close-to-ideal performance with Gurobi.</p><p>To guide such restructuring it is good to keep in mind the following bits of information:</p><ol><li><code>GRBupdatemodel</code> is only called if changes were done since last <code>GRBupdatemodel</code> (that is, if the internal buffer is not empty).</li><li><code>GRBupdatemodel</code> is called when <code>JuMP.optimize!</code> is called, but this often is not the source of the problem.</li><li><code>GRBupdatemodel</code> <em>may</em> be called when <em>any</em> model attribute is queried, <em>even if</em> that specific attribute was not changed. This often the source of the problem.</li></ol><p>The worst-case scenario is, therefore, a loop of modify-query-modify-query, even if what is being modified and what is being queried are two completely distinct things.</p><p>As an example, instead of:</p><pre><code class="language-julia hljs">model = Model(Gurobi.Optimizer)
+@variable(model, x[1:100] &gt;= 0)
+for i in 1:100
+    set_upper_bound(x[i], i)
+    # `GRBupdatemodel` called on each iteration of this loop.
+    println(lower_bound(x[i]))
+end</code></pre><p>do</p><pre><code class="language-julia hljs">model = Model(Gurobi.Optimizer)
+@variable(model, x[1:100] &gt;= 0)
+# All modifications are done before any queries.
+for i in 1:100
+    set_upper_bound(x[i], i)
+end
+for i in 1:100
+    # Only the first `lower_bound` query may trigger an `GRBupdatemodel`.
+    println(lower_bound(x[i]))
+end</code></pre><h2 id="Common-errors"><a class="docs-heading-anchor" href="#Common-errors">Common errors</a><a id="Common-errors-1"></a><a class="docs-heading-anchor-permalink" href="#Common-errors" title="Permalink"></a></h2><h3 id="Using-Gurobi-v9.0-and-you-got-an-error-like-Q-not-PSD?"><a class="docs-heading-anchor" href="#Using-Gurobi-v9.0-and-you-got-an-error-like-Q-not-PSD?">Using Gurobi v9.0 and you got an error like <code>Q not PSD</code>?</a><a id="Using-Gurobi-v9.0-and-you-got-an-error-like-Q-not-PSD?-1"></a><a class="docs-heading-anchor-permalink" href="#Using-Gurobi-v9.0-and-you-got-an-error-like-Q-not-PSD?" title="Permalink"></a></h3><p>You need to set the NonConvex parameter:</p><pre><code class="language-julia hljs">model = Model(Gurobi.Optimizer)
+set_optimizer_attribute(model, &quot;NonConvex&quot;, 2)</code></pre><h3 id="Gurobi-Error-1009:-Version-number-is-XX.X,-license-is-for-version-XX.X"><a class="docs-heading-anchor" href="#Gurobi-Error-1009:-Version-number-is-XX.X,-license-is-for-version-XX.X">Gurobi Error 1009: Version number is XX.X, license is for version XX.X</a><a id="Gurobi-Error-1009:-Version-number-is-XX.X,-license-is-for-version-XX.X-1"></a><a class="docs-heading-anchor-permalink" href="#Gurobi-Error-1009:-Version-number-is-XX.X,-license-is-for-version-XX.X" title="Permalink"></a></h3><p>Make sure that your license is correct for your Gurobi version. See the <a href="https://support.gurobi.com/hc/en-us/articles/360034784572-How-do-I-check-for-a-valid-license-file-">Gurobi documentation</a> for details.</p><p>Once you are sure that the license and Gurobi versions match, re-install Gurobi.jl by running:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.build(&quot;Gurobi&quot;)</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../GLPK/">« jump-dev/GLPK.jl</a><a class="docs-footer-nextpage" href="../HiGHS/">jump-dev/HiGHS.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/HiGHS/index.html b/previews/PR3525/packages/HiGHS/index.html
new file mode 100644
index 00000000000..d0eea432aae
--- /dev/null
+++ b/previews/PR3525/packages/HiGHS/index.html
@@ -0,0 +1,10 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/HiGHS.jl · JuMP</title><meta name="title" content="jump-dev/HiGHS.jl · JuMP"/><meta property="og:title" content="jump-dev/HiGHS.jl · JuMP"/><meta property="twitter:title" content="jump-dev/HiGHS.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/HiGHS.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#C-API"><span>C API</span></a></li></ul></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/HiGHS.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/HiGHS.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/HiGHS.jl/blob/v1.6.0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="HiGHS.jl"><a class="docs-heading-anchor" href="#HiGHS.jl">HiGHS.jl</a><a id="HiGHS.jl-1"></a><a class="docs-heading-anchor-permalink" href="#HiGHS.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/HiGHS.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/HiGHS.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/HiGHS.jl"><img src="https://codecov.io/gh/jump-dev/HiGHS.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/HiGHS.jl">HiGHS.jl</a> is a wrapper for the <a href="https://highs.dev">HiGHS</a> solver.</p><p>It has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not an official project of the HiGHS developers.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>HiGHS.jl</code> is licensed under the <a href="https://github.com/jump-dev/HiGHS.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/ERGO-Code/HiGHS">ERGO-Code/HiGHS</a>, is licensed under the <a href="https://github.com/ERGO-Code/HiGHS/blob/master/LICENSE">MIT license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install HiGHS as follows:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;HiGHS&quot;)</code></pre><p>In addition to installing the HiGHS.jl package, this will also download and install the HiGHS binaries. You do not need to install HiGHS separately.</p><p>To use a custom binary, read the <a href="https://jump.dev/JuMP.jl/stable/developers/custom_solver_binaries/">Custom solver binaries</a> section of the JuMP documentation.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use HiGHS with JuMP, use <code>HiGHS.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, HiGHS
+model = Model(HiGHS.Optimizer)
+set_attribute(model, &quot;presolve&quot;, &quot;on&quot;)
+set_attribute(model, &quot;time_limit&quot;, 60.0)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The HiGHS optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semicontinuous"><code>MOI.Semicontinuous{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>See the <a href="https://ergo-code.github.io/HiGHS/dev/options/definitions/">HiGHS documentation</a> for a full list of the available options.</p><h2 id="C-API"><a class="docs-heading-anchor" href="#C-API">C API</a><a id="C-API-1"></a><a class="docs-heading-anchor-permalink" href="#C-API" title="Permalink"></a></h2><p>The C API can be accessed via <code>HiGHS.Highs_xxx</code> functions, where the names and arguments are identical to the C API.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Gurobi/">« jump-dev/Gurobi.jl</a><a class="docs-footer-nextpage" href="../Ipopt/">jump-dev/Ipopt.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Hypatia/index.html b/previews/PR3525/packages/Hypatia/index.html
new file mode 100644
index 00000000000..54a86e9dd0c
--- /dev/null
+++ b/previews/PR3525/packages/Hypatia/index.html
@@ -0,0 +1,45 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>chriscoey/Hypatia.jl · JuMP</title><meta name="title" content="chriscoey/Hypatia.jl · JuMP"/><meta property="og:title" content="chriscoey/Hypatia.jl · JuMP"/><meta property="twitter:title" content="chriscoey/Hypatia.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li class="is-active"><a class="tocitem" href>chriscoey/Hypatia.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Usage"><span>Usage</span></a></li><li><a class="tocitem" href="#Contributing"><span>Contributing</span></a></li><li><a class="tocitem" href="#Acknowledgements"><span>Acknowledgements</span></a></li><li><a class="tocitem" href="#Citing-Hypatia"><span>Citing Hypatia</span></a></li></ul></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>chriscoey/Hypatia.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>chriscoey/Hypatia.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/chriscoey/Hypatia.jl/blob/v0.7.3/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><img src="https://github.com/chriscoey/Hypatia.jl/wiki/hypatia_logo.png?raw=true" alt="Hypatia logo" width="358"/><p><a href="https://github.com/chriscoey/Hypatia.jl/actions?query=workflow%3ACI+branch%3Amaster"><img src="https://github.com/chriscoey/Hypatia.jl/workflows/CI/badge.svg" alt="Build Status"/></a> <a href="https://codecov.io/gh/chriscoey/Hypatia.jl"><img src="https://codecov.io/gh/chriscoey/Hypatia.jl/branch/master/graph/badge.svg?token=x7G2wQeKJF" alt="codecov"/></a> <a href="https://chriscoey.github.io/Hypatia.jl/stable"><img src="https://img.shields.io/badge/docs-stable-blue.svg" alt/></a> <a href="https://chriscoey.github.io/Hypatia.jl/dev"><img src="https://img.shields.io/badge/docs-dev-blue.svg" alt/></a></p><p><a href="https://github.com/chriscoey/Hypatia.jl">Hypatia</a> is a highly customizable open source interior point solver for generic conic optimization problems, written in <a href="https://julialang.org/">Julia</a>.</p><p>For more information on Hypatia, please see:</p><ul><li><a href="https://chriscoey.github.io/Hypatia.jl/dev">documentation</a> for Hypatia&#39;s conic form, predefined cones, and interfaces</li><li><a href="https://github.com/chriscoey/Hypatia.jl/wiki/files/coneref.pdf">cones reference</a> for cone definitions and oracles</li><li><a href="https://github.com/chriscoey/Hypatia.jl/tree/master/examples">examples folder</a> for applied examples and instances</li><li><a href="https://github.com/chriscoey/Hypatia.jl/tree/master/benchmarks">benchmarks folder</a> for scripts used to run and analyze various computational benchmarks</li></ul><p>and preprints of our papers:</p><ul><li><a href="https://arxiv.org/abs/2005.01136">Solving natural conic formulations with Hypatia.jl</a> for computational arguments for expanding the class of cones recognized by conic solvers</li><li><a href="https://arxiv.org/abs/2107.04262">Performance enhancements for a generic conic interior point algorithm</a> for a description of Hypatia&#39;s algorithm and our enhanced stepping procedures</li><li><a href="https://arxiv.org/abs/2103.11499">Sum of squares generalizations for conic sets</a> for barriers and computational techniques for our generalized polynomial sum of squares cones</li><li><a href="https://arxiv.org/abs/2103.04104">Conic optimization with spectral functions on Euclidean Jordan algebras</a> for barriers and computational techniques for many of our epigraph/hypograph cones</li></ul><p>and corresponding <a href="https://github.com/chriscoey/Hypatia.jl/wiki">raw results CSV files</a> generated by our run scripts in the benchmarks folder.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>Hypatia is licensed under the MIT License (see <a href="https://github.com/chriscoey/Hypatia.jl/blob/master/LICENSE.md">LICENSE</a>).</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>To use Hypatia, install <a href="https://julialang.org/downloads/">Julia</a>, then at the Julia REPL, type:</p><pre><code class="language-julia hljs">using Hypatia
+using Pkg
+Pkg.add(&quot;Hypatia&quot;)</code></pre><p>Hypatia is an experimental solver and a work in progress, and may not run with older releases of Julia. Default options/parameters are not well-tuned, so we encourage you to experiment with these.</p><h2 id="Usage"><a class="docs-heading-anchor" href="#Usage">Usage</a><a id="Usage-1"></a><a class="docs-heading-anchor-permalink" href="#Usage" title="Permalink"></a></h2><p>Hypatia can be accessed through a low-level native Julia interface or through open-source modeling tools such as <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> and <a href="https://github.com/jump-dev/Convex.jl">Convex.jl</a>. The native interface is more expressive, allowing Hypatia to solve conic models expressed with generic real floating point types and structured matrices or linear operators, for example. However, it is typically sufficient and more convenient to use JuMP.</p><p>Using JuMP, we can model a simple D-optimal experiment design problem and call Hypatia:</p><pre><code class="language-julia hljs">using LinearAlgebra
+using JuMP
+using Hypatia
+
+model = Model(() -&gt; Hypatia.Optimizer(verbose = false))
+@variable(model, x[1:3] &gt;= 0)
+@constraint(model, sum(x) == 5)
+@variable(model, hypo)
+@objective(model, Max, hypo)
+V = rand(2, 3)
+Q = V * diagm(x) * V&#39;
+aff = vcat(hypo, [Q[i, j] for i in 1:2 for j in 1:i]...)
+@constraint(model, aff in MOI.RootDetConeTriangle(2))
+
+# solve and query solution
+optimize!(model)
+termination_status(model)
+objective_value(model)
+value.(x)</code></pre><p>See our <a href="https://github.com/chriscoey/Hypatia.jl/blob/master/examples/doptimaldesign/JuMP.jl">D-optimal design</a> example for more information and references.</p><p>Many more examples using the native interface or JuMP can be found in the <a href="https://github.com/chriscoey/Hypatia.jl/tree/master/examples">examples folder</a>.</p><h2 id="Contributing"><a class="docs-heading-anchor" href="#Contributing">Contributing</a><a id="Contributing-1"></a><a class="docs-heading-anchor-permalink" href="#Contributing" title="Permalink"></a></h2><p>Comments, questions, suggestions, and improvements/extensions to the code or documentation are welcomed. Please reach out on <a href="https://discourse.julialang.org/c/domain/opt">Discourse</a>, or submit an issue or contribute a PR on our <a href="https://github.com/chriscoey/Hypatia.jl">GitHub</a>. If contributing code, try to maintain consistent style and add docstrings or comments for clarity. New examples are welcomed and should be implemented similarly to the <a href="https://github.com/chriscoey/Hypatia.jl/tree/master/examples">existing examples</a>.</p><h2 id="Acknowledgements"><a class="docs-heading-anchor" href="#Acknowledgements">Acknowledgements</a><a id="Acknowledgements-1"></a><a class="docs-heading-anchor-permalink" href="#Acknowledgements" title="Permalink"></a></h2><p>This work has been partially funded by the National Science Foundation under grant OAC-1835443 and the Office of Naval Research under grant N00014-18-1-2079.</p><h2 id="Citing-Hypatia"><a class="docs-heading-anchor" href="#Citing-Hypatia">Citing Hypatia</a><a id="Citing-Hypatia-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-Hypatia" title="Permalink"></a></h2><p>If you find Hypatia solver useful, please cite our <a href="https://pubsonline.informs.org/doi/abs/10.1287/ijoc.2022.1202">solver paper</a>:</p><pre><code class="language-bibtex hljs">@article{coey2022solving,
+    title={Solving natural conic formulations with {H}ypatia.jl},
+    author={Chris Coey and Lea Kapelevich and Juan Pablo Vielma},
+    year={2022},
+    journal={INFORMS Journal on Computing},
+    publisher={INFORMS},
+    volume={34},
+    number={5},
+    pages={2686--2699},
+    doi={https://doi.org/10.1287/ijoc.2022.1202}
+}</code></pre><p>If you find aspects of Hypatia&#39;s IPM implementation useful, please cite our <a href="https://link.springer.com/article/10.1007/s12532-022-00226-0">algorithm paper</a>:</p><pre><code class="language-bibtex hljs">@article{coey2022performance,
+    title={Performance enhancements for a generic conic interior point algorithm},
+    author={Chris Coey and Lea Kapelevich and Juan Pablo Vielma},
+    year={2023},
+    journal={Mathematical Programming Computation},
+    publisher={Springer},
+    volume={15},
+    pages={53--101},
+    doi={https://doi.org/10.1007/s12532-022-00226-0}
+}</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../PATHSolver/">« chkwon/PATHSolver.jl</a><a class="docs-footer-nextpage" href="../DAQP/">darnstrom/DAQP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/InfiniteOpt/index.html b/previews/PR3525/packages/InfiniteOpt/index.html
new file mode 100644
index 00000000000..0a8bf03ca5a
--- /dev/null
+++ b/previews/PR3525/packages/InfiniteOpt/index.html
@@ -0,0 +1,15 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>infiniteopt/InfiniteOpt.jl · JuMP</title><meta name="title" content="infiniteopt/InfiniteOpt.jl · JuMP"/><meta property="og:title" content="infiniteopt/InfiniteOpt.jl · JuMP"/><meta property="twitter:title" content="infiniteopt/InfiniteOpt.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li class="is-active"><a class="tocitem" href>infiniteopt/InfiniteOpt.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Questions"><span>Questions</span></a></li><li><a class="tocitem" href="#Citing"><span>Citing</span></a></li></ul></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>infiniteopt/InfiniteOpt.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>infiniteopt/InfiniteOpt.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/infiniteopt/InfiniteOpt.jl/blob/v0.5.7/docs/jump/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><p><img src="https://raw.githubusercontent.com/infiniteopt/InfiniteOpt.jl/master/full_logo.png" alt="Logo"/></p><p><a href="https://github.com/infiniteopt/InfiniteOpt.jl"><code>InfiniteOpt.jl</code></a> is a <code>JuMP</code> extension for expressing and solving infinite-dimensional optimization problems. Such areas include <a href="https://en.wikipedia.org/wiki/Stochastic_programming">stochastic programming</a>, <a href="https://en.wikipedia.org/wiki/Dynamic_programming">dynamic programming</a>, space-time optimization, and more. <code>InfiniteOpt</code> serves as an easy-to-use modeling interface for these advanced problem types that can be used by those with little background in these areas. It also it contains a wealth of capabilities making it a powerful and convenient tool for advanced users.  </p><table><tr><th style="text-align: center"><strong>Current Version</strong></th><th style="text-align: center"><strong>Documentation</strong></th><th style="text-align: center"><strong>Build Status</strong></th><th style="text-align: center"><strong>Citation</strong></th></tr><tr><td style="text-align: center"><a href="https://juliahub.com/ui/Packages/InfiniteOpt/p3GvY"><img src="https://docs.juliahub.com/InfiniteOpt/version.svg" alt/></a></td><td style="text-align: center"><a href="https://infiniteopt.github.io/InfiniteOpt.jl/stable"><img src="https://img.shields.io/badge/docs-stable-blue.svg" alt/></a></td><td style="text-align: center"><a href="https://github.com/infiniteopt/InfiniteOpt.jl/actions?query=workflow%3ACI"><img src="https://github.com/infiniteopt/InfiniteOpt.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/github/infiniteopt/InfiniteOpt.jl?branch=master"><img src="https://codecov.io/github/infiniteopt/InfiniteOpt.jl/coverage.svg?branch=master" alt="codecov.io"/></a></td><td style="text-align: center"><a href="https://doi.org/10.1016/j.compchemeng.2021.107567"><img src="https://img.shields.io/badge/Elsevier-CompChemEng%3A107567-yellow.svg" alt="DOI"/></a></td></tr></table><p><code>InfiniteOpt</code> builds upon <code>JuMP</code> to add support for many complex modeling objects which  include:</p><ul><li>Infinite parameters (for example, time, space, and/or uncertainty)</li><li>Finite parameters (similar to <code>ParameterJuMP</code>)</li><li>Infinite variables (decision functions) (for example, <span>$y(t, x)$</span>)</li><li>Derivatives (for example, <span>$\frac{\partial y(t, x)}{\partial t}$</span>)</li><li>Measures (for example, <span>$\int_{t \in T}y(t, x)dt$</span> and <span>$\mathbb{E}[y(\xi)]$</span>)</li><li>First class nonlinear modeling</li></ul><p>The unifying modeling abstraction behind <code>InfiniteOpt</code> captures a wide spectrum  of disciplines which include dynamic, PDE, stochastic, and semi-infinite  optimization. Moreover, we facilitate transferring techniques between these  to synthesize new optimization paradigms.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>InfiniteOpt</code> is licensed under the <a href="https://github.com/infiniteopt/InfiniteOpt.jl/blob/master/LICENSE">MIT &quot;Expat&quot; license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p><code>InfiniteOpt.jl</code> is a registered <a href="https://julialang.org/">Julia</a> package and  can be installed by entering the following in the REPL.</p><pre><code class="language-julia hljs">julia&gt; import Pkg; Pkg.add(&quot;InfiniteOpt&quot;)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>Please visit our <a href="https://infiniteopt.github.io/InfiniteOpt.jl/stable">documentation pages</a>  to learn more. These pages are quite extensive and feature overviews, guides, manuals, tutorials, examples, and more.</p><h2 id="Questions"><a class="docs-heading-anchor" href="#Questions">Questions</a><a id="Questions-1"></a><a class="docs-heading-anchor-permalink" href="#Questions" title="Permalink"></a></h2><p>For additional help please visit and post in our  <a href="https://github.com/infiniteopt/InfiniteOpt.jl/discussions">discussion forum</a>.</p><h2 id="Citing"><a class="docs-heading-anchor" href="#Citing">Citing</a><a id="Citing-1"></a><a class="docs-heading-anchor-permalink" href="#Citing" title="Permalink"></a></h2><p><a href="https://doi.org/10.1016/j.compchemeng.2021.107567"><img src="https://img.shields.io/badge/Elsevier-CompChemEng%3A107567-yellow.svg" alt="DOI"/></a>  <a href="https://arxiv.org/abs/2106.12689"><img src="https://img.shields.io/badge/math.OC-arXiv%3A2106.12689-B31B1B.svg" alt="DOI"/></a></p><p>If you use InfiniteOpt.jl in your research, we would greatly appreciate your  citing it.</p><pre><code class="language-latex hljs">@article{pulsipher2022unifying,
+      title = {A unifying modeling abstraction for infinite-dimensional optimization},
+      journal = {Computers &amp; Chemical Engineering},
+      volume = {156},
+      year = {2022},
+      issn = {0098-1354},
+      doi = {https://doi.org/10.1016/j.compchemeng.2021.107567},
+      url = {https://www.sciencedirect.com/science/article/pii/S0098135421003458},
+      author = {Joshua L. Pulsipher and Weiqi Zhang and Tyler J. Hongisto and Victor M. Zavala},
+}</code></pre><p>A pre-print version is freely available though <a href="https://arxiv.org/abs/2106.12689">arXiv</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../SumOfSquares/">« jump-dev/SumOfSquares.jl</a><a class="docs-footer-nextpage" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Ipopt/index.html b/previews/PR3525/packages/Ipopt/index.html
new file mode 100644
index 00000000000..10e7e4c2b17
--- /dev/null
+++ b/previews/PR3525/packages/Ipopt/index.html
@@ -0,0 +1,118 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Ipopt.jl · JuMP</title><meta name="title" content="jump-dev/Ipopt.jl · JuMP"/><meta property="og:title" content="jump-dev/Ipopt.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Ipopt.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Ipopt.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Solver-specific-callbacks"><span>Solver-specific callbacks</span></a></li><li><a class="tocitem" href="#C-API"><span>C API</span></a></li><li><a class="tocitem" href="#INVALID_MODEL-error"><span><code>INVALID_MODEL</code> error</span></a></li><li><a class="tocitem" href="#Linear-Solvers"><span>Linear Solvers</span></a></li></ul></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Ipopt.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Ipopt.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Ipopt.jl/blob/v1.4.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><p><img src="https://www.coin-or.org/wordpress/wp-content/uploads/2014/08/COINOR.png" alt/></p><h1 id="Ipopt.jl"><a class="docs-heading-anchor" href="#Ipopt.jl">Ipopt.jl</a><a id="Ipopt.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Ipopt.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Ipopt.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/Ipopt.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Ipopt.jl"><img src="https://codecov.io/gh/jump-dev/Ipopt.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/Ipopt.jl">Ipopt.jl</a> is a wrapper for the <a href="https://github.com/coin-or/ipopt">Ipopt</a> solver.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not a COIN-OR project.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Ipopt.jl</code> is licensed under the <a href="https://github.com/jump-dev/Ipopt.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/coin-or/Ipopt">coin-or/Ipopt</a>, is licensed under the <a href="https://github.com/coin-or/Ipopt/blob/master/LICENSE">Eclipse public license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install <code>Ipopt.jl</code> using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Ipopt&quot;)</code></pre><p>In addition to installing the <code>Ipopt.jl</code> package, this will also download and install the Ipopt binaries. You do not need to install Ipopt separately.</p><p>To use a custom binary, read the <a href="https://jump.dev/JuMP.jl/stable/developers/custom_solver_binaries/">Custom solver binaries</a> section of the JuMP documentation.</p><p>For details on using a different linear solver, see the <code>Linear Solvers</code> section below. You do not need a custom binary to change the linear solver.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>You can use Ipopt with JuMP as follows:</p><pre><code class="language-julia hljs">using JuMP, Ipopt
+model = Model(Ipopt.Optimizer)
+set_attribute(model, &quot;max_cpu_time&quot;, 60.0)
+set_attribute(model, &quot;print_level&quot;, 0)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The Ipopt optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlock"><code>MOI.NLPBlock()</code></a></li><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlockDualStart"><code>MOI.NLPBlockDualStart()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>Supported options are listed in the <a href="https://coin-or.github.io/Ipopt/OPTIONS.html#OPTIONS_REF">Ipopt documentation</a>.</p><h2 id="Solver-specific-callbacks"><a class="docs-heading-anchor" href="#Solver-specific-callbacks">Solver-specific callbacks</a><a id="Solver-specific-callbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Solver-specific-callbacks" title="Permalink"></a></h2><p>Ipopt provides a callback that can be used to log the status of the optimization during a solve. It can also be used to terminate the optimization by returning <code>false</code>. Here is an example:</p><pre><code class="language-julia hljs">using JuMP, Ipopt, Test
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x &gt;= 1)
+@objective(model, Min, x + 0.5)
+x_vals = Float64[]
+function my_callback(
+   alg_mod::Cint,
+   iter_count::Cint,
+   obj_value::Float64,
+   inf_pr::Float64,
+   inf_du::Float64,
+   mu::Float64,
+   d_norm::Float64,
+   regularization_size::Float64,
+   alpha_du::Float64,
+   alpha_pr::Float64,
+   ls_trials::Cint,
+)
+   push!(x_vals, callback_value(model, x))
+   @test isapprox(obj_value, 1.0 * x_vals[end] + 0.5, atol = 1e-1)
+   # return `true` to keep going, or `false` to terminate the optimization.
+   return iter_count &lt; 1
+end
+MOI.set(model, Ipopt.CallbackFunction(), my_callback)
+optimize!(model)
+@test MOI.get(model, MOI.TerminationStatus()) == MOI.INTERRUPTED
+@test length(x_vals) == 2</code></pre><p>See the <a href="https://coin-or.github.io/Ipopt/OUTPUT.html">Ipopt documentation</a> for an explanation of the arguments to the callback. They are identical to the output contained in the logging table printed to the screen.</p><p>To access the current solution and primal, dual, and complementarity violations of each iteration, use <code>Ipopt.GetIpoptCurrentViolations</code> and <code>Ipopt.GetIpoptCurrentIterate</code>. The two functions are identical to the ones in the <a href="https://coin-or.github.io/Ipopt/INTERFACES.html">Ipopt C interface</a>.</p><h2 id="C-API"><a class="docs-heading-anchor" href="#C-API">C API</a><a id="C-API-1"></a><a class="docs-heading-anchor-permalink" href="#C-API" title="Permalink"></a></h2><p>Ipopt.jl wraps the <a href="https://coin-or.github.io/Ipopt/INTERFACES.html">Ipopt C interface</a> with minimal modifications.</p><p>A complete example is available in the <code>test/C_wrapper.jl</code> file.</p><p>For simplicity, the five callbacks required by Ipopt are slightly different to the C interface. They are as follows:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+   eval_f(x::Vector{Float64})::Float64
+
+Returns the objective value `f(x)`.
+&quot;&quot;&quot;
+function eval_f end
+
+&quot;&quot;&quot;
+   eval_grad_f(x::Vector{Float64}, grad_f::Vector{Float64})::Nothing
+
+Fills `grad_f` in-place with the gradient of the objective function evaluated at
+`x`.
+&quot;&quot;&quot;
+function eval_grad_f end
+
+&quot;&quot;&quot;
+   eval_g(x::Vector{Float64}, g::Vector{Float64})::Nothing
+
+Fills `g` in-place with the value of the constraints evaluated at `x`.
+&quot;&quot;&quot;
+function eval_g end
+
+&quot;&quot;&quot;
+   eval_jac_g(
+      x::Vector{Float64},
+      rows::Vector{Cint},
+      cols::Vector{Cint},
+      values::Union{Nothing,Vector{Float64}},
+   )::Nothing
+
+Compute the Jacobian matrix.
+
+* If `values === nothing`
+   - Fill `rows` and `cols` with the 1-indexed sparsity structure
+* Otherwise:
+   - Fill `values` with the elements of the Jacobian matrix according to the
+     sparsity structure.
+
+!!! warning
+    If `values === nothing`, `x` is an undefined object. Accessing any elements
+    in it will cause Julia to segfault.
+&quot;&quot;&quot;
+function eval_jac_g end
+
+&quot;&quot;&quot;
+   eval_h(
+      x::Vector{Float64},
+      rows::Vector{Cint},
+      cols::Vector{Cint},
+      obj_factor::Float64,
+      lambda::Float64,
+      values::Union{Nothing,Vector{Float64}},
+   )::Nothing
+
+Compute the Hessian-of-the-Lagrangian matrix.
+
+* If `values === nothing`
+   - Fill `rows` and `cols` with the 1-indexed sparsity structure
+* Otherwise:
+   - Fill `values` with the Hessian matrix according to the sparsity structure.
+
+!!! warning
+    If `values === nothing`, `x` is an undefined object. Accessing any elements
+    in it will cause Julia to segfault.
+&quot;&quot;&quot;
+function eval_h end</code></pre><h2 id="INVALID_MODEL-error"><a class="docs-heading-anchor" href="#INVALID_MODEL-error"><code>INVALID_MODEL</code> error</a><a id="INVALID_MODEL-error-1"></a><a class="docs-heading-anchor-permalink" href="#INVALID_MODEL-error" title="Permalink"></a></h2><p>If you get a termination status <code>MOI.INVALID_MODEL</code>, it is probably because you have some undefined value in your model, for example, a division by zero. Fix this by removing the division, or by imposing variable bounds so that you cut off the undefined region.</p><p>Instead of</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+@variable(model, x)
+@NLobjective(model, 1 / x)</code></pre><p>do</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+@variable(model, x &gt;= 0.0001)
+@NLobjective(model, 1 / x)</code></pre><h2 id="Linear-Solvers"><a class="docs-heading-anchor" href="#Linear-Solvers">Linear Solvers</a><a id="Linear-Solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Linear-Solvers" title="Permalink"></a></h2><p>To improve performance, Ipopt supports a number of linear solvers.</p><h3 id="HSL"><a class="docs-heading-anchor" href="#HSL">HSL</a><a id="HSL-1"></a><a class="docs-heading-anchor-permalink" href="#HSL" title="Permalink"></a></h3><p>Obtain a license and download <code>HSL_jll.jl</code> from <a href="https://licences.stfc.ac.uk/product/julia-hsl">https://licences.stfc.ac.uk/product/julia-hsl</a>.</p><p>There are two versions available: LBT and OpenBLAS. LBT is the recommended option for Julia ≥ v1.9.</p><p>Install this download into your current environment using:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.develop(path = &quot;/full/path/to/HSL_jll.jl&quot;)</code></pre><p>Then, use a linear solver in HSL by setting the <code>hsllib</code> and <code>linear_solver</code> attributes:</p><pre><code class="language-julia hljs">using JuMP, Ipopt
+import HSL_jll
+model = Model(Ipopt.Optimizer)
+set_attribute(model, &quot;hsllib&quot;, HSL_jll.libhsl_path)
+set_attribute(model, &quot;linear_solver&quot;, &quot;ma86&quot;)</code></pre><h4 id="macOS-users"><a class="docs-heading-anchor" href="#macOS-users">macOS users</a><a id="macOS-users-1"></a><a class="docs-heading-anchor-permalink" href="#macOS-users" title="Permalink"></a></h4><p>Due to the security policy of macOS, Mac users may need to delete the quarantine attribute of the ZIP archive before extracting. For example:</p><pre><code class="language-raw hljs">xattr -d com.apple.quarantine lbt_HSL_jll.jl-2023.5.26.zip
+xattr -d com.apple.quarantine openblas_HSL_jll.jl-2023.5.26.zip</code></pre><h3 id="Pardiso"><a class="docs-heading-anchor" href="#Pardiso">Pardiso</a><a id="Pardiso-1"></a><a class="docs-heading-anchor-permalink" href="#Pardiso" title="Permalink"></a></h3><p>Download Pardiso from <a href="https://www.pardiso-project.org">https://www.pardiso-project.org</a>. Save the shared library somewhere, and record the filename.</p><p>Then, use Pardiso by setting the <code>pardisolib</code> and <code>linear_solver</code> attributes:</p><pre><code class="language-julia hljs">using JuMP, Ipopt
+model = Model(Ipopt.Optimizer)
+set_attribute(model, &quot;pardisolib&quot;, &quot;/full/path/to/libpardiso&quot;)
+set_attribute(model, &quot;linear_solver&quot;, &quot;pardiso&quot;)</code></pre><h3 id="SPRAL"><a class="docs-heading-anchor" href="#SPRAL">SPRAL</a><a id="SPRAL-1"></a><a class="docs-heading-anchor-permalink" href="#SPRAL" title="Permalink"></a></h3><p>If you use Ipopt.jl with Julia ≥ v1.9, the linear solver <a href="https://github.com/ralna/spral">SPRAL</a> is available. You can use it by setting the <code>linear_solver</code> attribute:</p><pre><code class="language-julia hljs">using JuMP, Ipopt
+model = Model(Ipopt.Optimizer)
+set_attribute(model, &quot;linear_solver&quot;, &quot;spral&quot;)</code></pre><p>Note that the following environment variables must be set before starting Julia:</p><pre><code class="language-raw hljs">export OMP_CANCELLATION=TRUE
+export OMP_PROC_BIND=TRUE</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../HiGHS/">« jump-dev/HiGHS.jl</a><a class="docs-footer-nextpage" href="../KNITRO/">jump-dev/KNITRO.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Juniper/index.html b/previews/PR3525/packages/Juniper/index.html
new file mode 100644
index 00000000000..7970b2fc8a6
--- /dev/null
+++ b/previews/PR3525/packages/Juniper/index.html
@@ -0,0 +1,36 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>lanl-ansi/Juniper.jl · JuMP</title><meta name="title" content="lanl-ansi/Juniper.jl · JuMP"/><meta property="og:title" content="lanl-ansi/Juniper.jl · JuMP"/><meta property="twitter:title" content="lanl-ansi/Juniper.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li class="is-active"><a class="tocitem" href>lanl-ansi/Juniper.jl</a><ul class="internal"><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Feasibility-pump"><span>Feasibility pump</span></a></li><li><a class="tocitem" href="#Citing-Juniper"><span>Citing Juniper</span></a></li></ul></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>lanl-ansi/Juniper.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>lanl-ansi/Juniper.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/lanl-ansi/Juniper.jl/blob/62532341586d447f19c7360715333ba62a42bea9/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Juniper"><a class="docs-heading-anchor" href="#Juniper">Juniper</a><a id="Juniper-1"></a><a class="docs-heading-anchor-permalink" href="#Juniper" title="Permalink"></a></h1><p><a href="https://github.com/lanl-ansi/Juniper.jl/actions?query=workflow%3ACI"><img src="https://github.com/lanl-ansi/Juniper.jl/workflows/CI/badge.svg" alt="CI"/></a> <a href="https://codecov.io/gh/lanl-ansi/Juniper.jl"><img src="https://codecov.io/gh/lanl-ansi/Juniper.jl/branch/master/graph/badge.svg" alt="codecov"/></a> <a href="https://lanl-ansi.github.io/Juniper.jl/stable"><img src="https://img.shields.io/badge/docs-stable-blue.svg" alt="Documentation"/></a></p><p><a href="http://github.com/lanl-ansi/Juniper.jl">Juniper</a> (Jump Nonlinear Integer Program solver) is a solver for mixed-integer nonlinear programs. </p><p>It is a heuristic which is not guaranteed to find the global optimum. If you need the global optimum, check out <a href="http://github.com/lanl-ansi/Alpine.jl">Alpine</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Juniper using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;JuMP&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use Juniper with JuMP as follows:</p><pre><code class="language-julia hljs">using JuMP, Juniper, Ipopt
+ipopt = optimizer_with_attributes(Ipopt.Optimizer, &quot;print_level&quot;=&gt;0)
+optimizer = optimizer_with_attributes(Juniper.Optimizer, &quot;nl_solver&quot;=&gt;ipopt)
+model = Model(optimizer)
+v = [10, 20, 12, 23, 42]
+w = [12, 45, 12, 22, 21]
+@variable(model, x[1:5], Bin)
+@objective(model, Max, v&#39; * x)
+@constraint(model, sum(w[i]*x[i]^2 for i in 1:5) &lt;= 45)
+optimize!(model)
+println(termination_status(model))
+println(objective_value(model))
+println(value.(x))</code></pre><p>The <code>nl_solver</code> is used by Juniper to solve continuous nonlinear sub-problems while Juniper searches for acceptable assignments to the discrete variables. A common choice is <a href="https://github.com/jump-dev/Ipopt.jl">Ipopt</a>, but any optimizer that supports the continuous relaxation of the model may be used.</p><p>To solve problems with more complex nonlinear functions, use the <code>@NLconstraint</code> and <code>@NLobjective</code> JuMP macros.</p><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>The online documentation is available at <a href="https://lanl-ansi.github.io/Juniper.jl/stable/">https://lanl-ansi.github.io/Juniper.jl/stable/</a>.</p><h2 id="Feasibility-pump"><a class="docs-heading-anchor" href="#Feasibility-pump">Feasibility pump</a><a id="Feasibility-pump-1"></a><a class="docs-heading-anchor-permalink" href="#Feasibility-pump" title="Permalink"></a></h2><p>If Juniper has difficulty finding feasible solutions on your model, try adding a solver that supports integer variables (for example, <a href="https://github.com/jump-dev/HiGHS.jl">HiGHS</a>) to run a <em>feasibility pump</em>:</p><pre><code class="language-julia hljs">using JuMP, Juniper, Ipopt, HiGHS
+ipopt = optimizer_with_attributes(Ipopt.Optimizer, &quot;print_level&quot; =&gt; 0)
+highs = optimizer_with_attributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; false)
+model = Model(
+    optimizer_with_attributes(
+        Juniper.Optimizer,
+        &quot;nl_solver&quot; =&gt; ipopt,
+        &quot;mip_solver&quot; =&gt; highs,
+    ),
+)</code></pre><p>The feasibility pump is used at the start of Juniper to find a feasible solution before the branch and bound part starts.  For some classes of problems this can be a highly effective pre-processor.</p><h2 id="Citing-Juniper"><a class="docs-heading-anchor" href="#Citing-Juniper">Citing Juniper</a><a id="Citing-Juniper-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-Juniper" title="Permalink"></a></h2><p>If you find Juniper useful in your work, we kindly request that you cite the following <a href="https://link.springer.com/chapter/10.1007/978-3-319-93031-2_27">paper</a> or <a href="https://arxiv.org/abs/1804.07332">technical report</a>:</p><pre><code class="nohighlight hljs">@inproceedings{juniper,
+     Author = {Ole Kröger and Carleton Coffrin and Hassan Hijazi and Harsha Nagarajan},
+     Title = {Juniper: An Open-Source Nonlinear Branch-and-Bound Solver in Julia},
+     booktitle=&quot;Integration of Constraint Programming, Artificial Intelligence, and Operations Research&quot;,
+     pages=&quot;377--386&quot;,
+     year=&quot;2018&quot;,
+     publisher=&quot;Springer International Publishing&quot;,
+     isbn=&quot;978-3-319-93031-2&quot;
+}</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Alpine/">« lanl-ansi/Alpine.jl</a><a class="docs-footer-nextpage" href="../ProxSDP/">mariohsouto/ProxSDP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/KNITRO/index.html b/previews/PR3525/packages/KNITRO/index.html
new file mode 100644
index 00000000000..c40783beebc
--- /dev/null
+++ b/previews/PR3525/packages/KNITRO/index.html
@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/KNITRO.jl · JuMP</title><meta name="title" content="jump-dev/KNITRO.jl · JuMP"/><meta property="og:title" content="jump-dev/KNITRO.jl · JuMP"/><meta property="twitter:title" content="jump-dev/KNITRO.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/KNITRO.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Use-with-AMPL"><span>Use with AMPL</span></a></li><li><a class="tocitem" href="#Use-with-other-packages"><span>Use with other packages</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Low-level-wrapper"><span>Low-level wrapper</span></a></li><li><a class="tocitem" href="#Multi-threading"><span>Multi-threading</span></a></li></ul></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/KNITRO.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/KNITRO.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/KNITRO.jl/blob/4c56de7684c42dd3c83c5fbc515ae1f424eed524/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="KNITRO.jl"><a class="docs-heading-anchor" href="#KNITRO.jl">KNITRO.jl</a><a id="KNITRO.jl-1"></a><a class="docs-heading-anchor-permalink" href="#KNITRO.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/KNITRO.jl">KNITRO.jl</a> is a wrapper for the <a href="https://www.artelys.com/knitro">Artelys Knitro solver</a>.</p><p>It has two components:</p><ul><li>a thin wrapper around the <a href="https://www.artelys.com/tools/knitro_doc/3_referenceManual/callableLibraryAPI.html">C API</a></li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a>.</li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community with help from Artelys.</p><p>Contact <a href="mailto:support-knitro@artelys.com">Artelys support</a> if you encounter any problem with this interface or the solver.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>KNITRO.jl</code> is licensed under the <a href="https://github.com/jump-dev/KNITRO.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver is a closed-source commercial product for which you must <a href="https://www.artelys.com/knitro">purchase a license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, obtain a license and install a copy of KNITRO from <a href="https://www.artelys.com/knitro">Artelys</a>.</p><p>Then, install <code>KNITRO.jl</code> using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;KNITRO&quot;)</code></pre><p>If you are having trouble installing KNITRO.jl, here are several things to try:</p><ul><li>Make sure that you have defined your global variables correctly, for example with <code>export KNITRODIR=&quot;/path/to/knitro-vXXX-$OS-64&quot;</code> and <code>export LD_LIBRARY_PATH=&quot;$LD_LIBRARY_PATH:$KNITRODIR/lib&quot;</code>. You can check that <code>KNITRO.jl</code> sees your library with <code>using KNITRO; KNITRO.has_knitro()</code>.</li><li>If <code>KNITRO.has_knitro()</code> returns <code>false</code> but you are confident that your paths are correct, try running <code>Pkg.build(&quot;KNITRO&quot;)</code> and restarting Julia. In at least one user&#39;s experience, installing and using KNITRO in a temporary Julia environment (activated with <code>] activate --temp</code>) does not work and the need to manually build is likely the reason why.</li></ul><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use KNITRO with JuMP, use <code>KNITRO.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, KNITRO
+model = Model(KNITRO.Optimizer)
+set_attribute(model, &quot;outlev&quot;, 1)
+set_attribute(model, &quot;algorithm&quot;, 4)</code></pre><h2 id="Use-with-AMPL"><a class="docs-heading-anchor" href="#Use-with-AMPL">Use with AMPL</a><a id="Use-with-AMPL-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-AMPL" title="Permalink"></a></h2><p>To use KNITRO with <a href="https://github.com/jump-dev/AmplNLWriter.jl">AmplNLWriter.jl</a>, use <code>KNITRO.amplexe</code>:</p><pre><code class="language-julia hljs">using JuMP
+import AmplNLWriter
+import KNITRO
+model = Model(() -&gt; AmplNLWriter.Optimizer(KNITRO.amplexe, [&quot;outlev=3&quot;]))</code></pre><h2 id="Use-with-other-packages"><a class="docs-heading-anchor" href="#Use-with-other-packages">Use with other packages</a><a id="Use-with-other-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-other-packages" title="Permalink"></a></h2><p>A variety of packages extend KNITRO.jl to support other optimization modeling systems. These include:</p><ul><li><a href="https://github.com/JuliaSmoothOptimizers/NLPModelsKnitro.jl">NLPModelsKnitro</a></li><li><a href="http://optimization.sciml.ai/stable/">Optimization.jl</a></li></ul><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The Knitro optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Complements"><code>MOI.Complements</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlock"><code>MOI.NLPBlock()</code></a></li><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlockDualStart"><code>MOI.NLPBlockDualStart()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>A list of available options is provided in the <a href="https://www.artelys.com/docs/knitro/3_referenceManual/userOptions.html">KNITRO reference manual</a>.</p><h2 id="Low-level-wrapper"><a class="docs-heading-anchor" href="#Low-level-wrapper">Low-level wrapper</a><a id="Low-level-wrapper-1"></a><a class="docs-heading-anchor-permalink" href="#Low-level-wrapper" title="Permalink"></a></h2><p>KNITRO.jl implements most of Knitro&#39;s functionalities. If you aim at using part of Knitro&#39;s API that are not implemented in the MathOptInterface/JuMP ecosystem, you can refer to the low-level API, which wraps Knitro&#39;s C API (whose templates are specified in the file <code>knitro.h</code>).</p><p>Extensive examples using the C wrapper can be found in <code>examples/</code>.</p><h2 id="Multi-threading"><a class="docs-heading-anchor" href="#Multi-threading">Multi-threading</a><a id="Multi-threading-1"></a><a class="docs-heading-anchor-permalink" href="#Multi-threading" title="Permalink"></a></h2><p>Due to limitations in the interaction between Julia and C, KNITRO.jl disables multi-threading if the problem is nonlinear. This will override any options such as <code>par_numthreads</code> that you may have set. Read <a href="https://github.com/jump-dev/KNITRO.jl/issues/93">GitHub issue #93</a> for more details.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Ipopt/">« jump-dev/Ipopt.jl</a><a class="docs-footer-nextpage" href="../MiniZinc/">jump-dev/MiniZinc.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Loraine/index.html b/previews/PR3525/packages/Loraine/index.html
new file mode 100644
index 00000000000..aa8effaf28c
--- /dev/null
+++ b/previews/PR3525/packages/Loraine/index.html
@@ -0,0 +1,32 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>kocvara/Loraine.jl · JuMP</title><meta name="title" content="kocvara/Loraine.jl · JuMP"/><meta property="og:title" content="kocvara/Loraine.jl · JuMP"/><meta property="twitter:title" content="kocvara/Loraine.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li class="is-active"><a class="tocitem" href>kocvara/Loraine.jl</a><ul class="internal"><li><a class="tocitem" href="#License-and-Original-Contributors"><span>License and Original Contributors</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Rank-one-data"><span>Rank-one data</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Citing"><span>Citing</span></a></li></ul></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>kocvara/Loraine.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>kocvara/Loraine.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/kocvara/Loraine.jl/blob/v0.1.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Loraine.jl"><a class="docs-heading-anchor" href="#Loraine.jl">Loraine.jl</a><a id="Loraine.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Loraine.jl" title="Permalink"></a></h1><p><em>Sweet Lor(r)aine, let the party carry on<sup class="footnote-reference"><a id="citeref-1" href="#footnote-1">[1]</a></sup>...</em></p><p>Loraine.jl is a Julia implementation of an interior point method algorithm for linear semidefinite optimization problems. </p><p>The special feature of Loraine is the iterative solver for linear systems. This is to be used for problems with (very) low rank solution matrix.</p><p>Standard (non-low-rank) problems and linear programs can be solved using the direct solver; then the user gets a standard IP method akin SDPT3.</p><p>There is also a MATLAB version of the code at <a href="https://github.com/kocvara/Loraine.m">kocvara/Loraine.m</a>.</p><h2 id="License-and-Original-Contributors"><a class="docs-heading-anchor" href="#License-and-Original-Contributors">License and Original Contributors</a><a id="License-and-Original-Contributors-1"></a><a class="docs-heading-anchor-permalink" href="#License-and-Original-Contributors" title="Permalink"></a></h2><p>Loraine is licensed under the <a href="https://github.com/kocvara/Loraine.jl/blob/main/LICENSE.md">MIT License</a>.</p><p>Loraine was developed by Soodeh Habibi and Michal Kočvara, University of Birmingham, and Michael Stingl, University of Erlangen, for H2020 ITN POEMA. </p><p>The JuMP interface was provided by Benoît Legat. His help is greatly acknowledged.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install <code>Loraine</code> using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Loraine&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use Loraine with JuMP, use <code>Loraine.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, Loraine
+model = Model(Loraine.Optimizer)
+set_attribute(model, &quot;maxit&quot;, 100)</code></pre><p>To solve an SDP problem stored in SDPA format, do</p><pre><code class="language-julia hljs">using JuMP, Loraine
+model = read_from_file(&quot;examples/data/theta1.dat-s&quot;)
+set_optimizer(model, Loraine.Optimizer)
+optimize!(model)</code></pre><p>For more examples, the folder <a href="https://github.com/kocvara/Loraine.jl/tree/main/examples"><code>examples</code></a> includes a few examples of how to use Loraine via JuMP; in particular, <code>solve_sdpa.jl</code> reads an SDP in the SDPA input format and solves it by Loraine. A few sample problems can be found in folder <code>examples/data</code>.</p><h2 id="Rank-one-data"><a class="docs-heading-anchor" href="#Rank-one-data">Rank-one data</a><a id="Rank-one-data-1"></a><a class="docs-heading-anchor-permalink" href="#Rank-one-data" title="Permalink"></a></h2><p>If the solution does not have low rank, it is recommended to use a direct  solver <code>kit = 0</code>. However, if you know that your data matrices are all rank-one,  use the option <code>datarank = -1</code> to get a significant reduction in the complexity  (and CPU time). Examples of such problems are <code>maxG11</code> and <code>thetaG11</code> from the  <code>SDPLIB</code> collection.</p><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>The list of options:</p><pre><code class="language-julia hljs">kit             # kit = 0 for direct solver; kit = 1 for CG [0]
+tol_cg          # tolerance for CG solver [1.0e-2]
+tol_cg_up       # tolerance update [0.5]
+tol_cg_min      # minimal tolerance for CG solver [1.0e-6]
+eDIMACS         # epsilon for DIMACS error stopping criterion [1.0e-5]
+preconditioner  # 0...no; 1...H_alpha; 2...H_beta; 4...hybrid [1]
+erank           # estimated rank [1]
+aamat           # 0..A^TA; 1..diag(A^TA); 2..identity [2]
+verb            # 2..full output; 1..short output; 0..no output [1]
+datarank        # 0..full rank matrices expected [0]
+                # -1..rank-1 matrices expected, converted to vectors, if possible
+                # (TBD) 1..vectors expected for low-rank data matrices
+initpoint       # 0..Loraine heuristics, 1..SDPT3-like heuristics [0]
+timing          # 1..yes, 0..no
+maxit           # maximal number of global iterations [200]</code></pre><h2 id="Citing"><a class="docs-heading-anchor" href="#Citing">Citing</a><a id="Citing-1"></a><a class="docs-heading-anchor-permalink" href="#Citing" title="Permalink"></a></h2><p>If you find Loraine useful, please cite the following paper:</p><pre><code class="language-bibtex hljs">@article{loraine2023,
+  title={Loraine-An interior-point solver for low-rank semidefinite programming},
+  author={Habibi, Soodeh and Ko{\v{c}}vara, Michal and Stingl, Michael},
+  www={https://hal.science/hal-04076509/}
+  note={Preprint hal-04076509}
+  year={2023}
+}</code></pre><section class="footnotes is-size-7"><ul><li class="footnote" id="footnote-1"><a class="tag is-link" href="#citeref-1">1</a>https://www.youtube.com/watch?v=0D2wNf1lVrI</li></ul></section></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Tulip/">« ds4dm/Tulip.jl</a><a class="docs-footer-nextpage" href="../Alpine/">lanl-ansi/Alpine.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/MadNLP/index.html b/previews/PR3525/packages/MadNLP/index.html
new file mode 100644
index 00000000000..823462d60d4
--- /dev/null
+++ b/previews/PR3525/packages/MadNLP/index.html
@@ -0,0 +1,48 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>MadNLP/MadNLP.jl · JuMP</title><meta name="title" content="MadNLP/MadNLP.jl · JuMP"/><meta property="og:title" content="MadNLP/MadNLP.jl · JuMP"/><meta property="twitter:title" content="MadNLP/MadNLP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li class="is-active"><a class="tocitem" href>MadNLP/MadNLP.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Usage"><span>Usage</span></a></li><li><a class="tocitem" href="#Citing-MadNLP.jl"><span>Citing MadNLP.jl</span></a></li><li><a class="tocitem" href="#Bug-reports-and-support"><span>Bug reports and support</span></a></li></ul></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>MadNLP/MadNLP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>MadNLP/MadNLP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/MadNLP/MadNLP.jl/blob/389561354a774441cd5e6b3aa5fffee102ed222e/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><img src="https://github.com/MadNLP/MadNLP.jl/blob/master/logo-full.svg?raw=true"/><table><tr><th style="text-align: center"><strong>Documentation</strong></th><th style="text-align: center"><strong>Build Status</strong></th><th style="text-align: center"><strong>Coverage</strong></th><th style="text-align: center"><strong>DOI</strong></th></tr><tr><td style="text-align: center"><a href="https://madnlp.github.io/MadNLP.jl/dev"><img src="https://img.shields.io/badge/docs-dev-blue.svg" alt="doc"/></a></td><td style="text-align: center"><a href="https://github.com/MadNLP/MadNLP.jl/actions?query=workflow%3Abuild"><img src="https://github.com/MadNLP/MadNLP.jl/workflows/build/badge.svg?branch=dev%2Fgithub_actions" alt="build"/></a></td><td style="text-align: center"><a href="https://codecov.io/gh/MadNLP/MadNLP.jl"><img src="https://codecov.io/gh/MadNLP/MadNLP.jl/branch/master/graph/badge.svg?token=MBxH2AAu8Z" alt="codecov"/></a></td><td style="text-align: center"><a href="https://doi.org/10.5281/zenodo.5825776"><img src="https://zenodo.org/badge/DOI/10.5281/zenodo.5825776.svg" alt="DOI"/></a></td></tr></table><p>MadNLP is a <a href="https://en.wikipedia.org/wiki/Nonlinear_programming">nonlinear programming</a> (NLP) solver, purely implemented in <a href="https://julialang.org/">Julia</a>. MadNLP implements a filter line-search algorithm, as that used in <a href="https://github.com/coin-or/Ipopt">Ipopt</a>. MadNLP seeks to streamline the development of modeling and algorithmic paradigms in order to exploit structures and to make efficient use of high-performance computers.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>MadNLP is available under the <a href="https://github.com/MadNLP/MadNLP.jl/blob/master/LICENSE">MIT license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><pre><code class="language-julia hljs">pkg&gt; add MadNLP</code></pre><p>Optionally, various extension packages can be installed together:</p><pre><code class="language-julia hljs">pkg&gt; add MadNLPHSL, MadNLPPardiso, MadNLPMumps, MadNLPGPU, MadNLPGraph, MadNLPKrylov</code></pre><p>These packages are stored in the <code>lib</code> subdirectory within the main MadNLP repository. Some extension packages may require additional dependencies or specific hardware. For the instructions for the build procedure, see the following links:</p><ul><li><a href="https://github.com/MadNLP/MadNLP.jl/tree/master/lib/MadNLPHSL">MadNLPHSL</a></li><li><a href="https://github.com/MadNLP/MadNLP.jl/tree/master/lib/MadNLPHSL">MadNLPPardiso</a></li><li><a href="https://github.com/MadNLP/MadNLP.jl/tree/master/lib/MadNLPGPU">MadNLPGPU</a></li></ul><h2 id="Usage"><a class="docs-heading-anchor" href="#Usage">Usage</a><a id="Usage-1"></a><a class="docs-heading-anchor-permalink" href="#Usage" title="Permalink"></a></h2><h3 id="Interfaces"><a class="docs-heading-anchor" href="#Interfaces">Interfaces</a><a id="Interfaces-1"></a><a class="docs-heading-anchor-permalink" href="#Interfaces" title="Permalink"></a></h3><p>MadNLP is interfaced with modeling packages:</p><ul><li><a href="https://github.com/jump-dev/JuMP.jl">JuMP</a></li><li><a href="https://github.com/zavalab/Plasmo.jl">Plasmo</a></li><li><a href="https://github.com/JuliaSmoothOptimizers/NLPModels.jl">NLPModels</a>.</li></ul><p>Users can pass various options to MadNLP also through the modeling packages. The interface-specific syntax are shown below. To see the list of MadNLP solver options, check the <a href="https://github.com/MadNLP/MadNLP/blob/master/OPTIONS.md">OPTIONS.md</a> file.</p><h4 id="JuMP-interface"><a class="docs-heading-anchor" href="#JuMP-interface">JuMP interface</a><a id="JuMP-interface-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-interface" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, JuMP
+model = Model(()-&gt;MadNLP.Optimizer(print_level=MadNLP.INFO, max_iter=100))
+@variable(model, x, start = 0.0)
+@variable(model, y, start = 0.0)
+@NLobjective(model, Min, (1 - x)^2 + 100 * (y - x^2)^2)
+optimize!(model)</code></pre><h4 id="NLPModels-interface"><a class="docs-heading-anchor" href="#NLPModels-interface">NLPModels interface</a><a id="NLPModels-interface-1"></a><a class="docs-heading-anchor-permalink" href="#NLPModels-interface" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, CUTEst
+model = CUTEstModel(&quot;PRIMALC1&quot;)
+madnlp(model, print_level=MadNLP.WARN, max_wall_time=3600)</code></pre><h4 id="Plasmo-interface-(requires-extension-MadNLPGraph)"><a class="docs-heading-anchor" href="#Plasmo-interface-(requires-extension-MadNLPGraph)">Plasmo interface (requires extension <code>MadNLPGraph</code>)</a><a id="Plasmo-interface-(requires-extension-MadNLPGraph)-1"></a><a class="docs-heading-anchor-permalink" href="#Plasmo-interface-(requires-extension-MadNLPGraph)" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, MadNLPGraph, Plasmo
+graph = OptiGraph()
+@optinode(graph,n1)
+@optinode(graph,n2)
+@variable(n1,0 &lt;= x &lt;= 2)
+@variable(n1,0 &lt;= y &lt;= 3)
+@constraint(n1,x+y &lt;= 4)
+@objective(n1,Min,x)
+@variable(n2,x)
+@NLnodeconstraint(n2,exp(x) &gt;= 2)
+@linkconstraint(graph,n1[:x] == n2[:x])
+MadNLP.optimize!(graph; print_level=MadNLP.DEBUG, max_iter=100)</code></pre><h3 id="Linear-Solvers"><a class="docs-heading-anchor" href="#Linear-Solvers">Linear Solvers</a><a id="Linear-Solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Linear-Solvers" title="Permalink"></a></h3><p>MadNLP is interfaced with non-Julia sparse/dense linear solvers:</p><ul><li><a href="https://people.engr.tamu.edu/davis/suitesparse.html">Umfpack</a></li><li><a href="https://software.intel.com/content/www/us/en/develop/documentation/mkl-developer-reference-fortran/top/sparse-solver-routines/intel-mkl-pardiso-parallel-direct-sparse-solver-interface.html">MKL-Pardiso</a></li><li><a href="https://software.intel.com/content/www/us/en/develop/documentation/mkl-developer-reference-fortran/top/lapack-routines.html">MKL-Lapack</a></li><li><a href="http://www.hsl.rl.ac.uk/ipopt/">HSL solvers</a> (requires extension)</li><li><a href="https://www.pardiso-project.org/">Pardiso</a> (requires extension)</li><li><a href="http://mumps.enseeiht.fr/">Mumps</a>  (requires extension)</li><li><a href="https://docs.nvidia.com/cuda/cusolver/index.html">cuSOLVER</a> (requires extension)</li></ul><p>Each linear solver in MadNLP is a Julia type, and the <code>linear_solver</code> option should be specified by the actual type. Note that the linear solvers are always exported to <code>Main</code>.</p><h4 id="Built-in-Solvers:-Umfpack,-PardisoMKL,-LapackCPU"><a class="docs-heading-anchor" href="#Built-in-Solvers:-Umfpack,-PardisoMKL,-LapackCPU">Built-in Solvers: Umfpack, PardisoMKL, LapackCPU</a><a id="Built-in-Solvers:-Umfpack,-PardisoMKL,-LapackCPU-1"></a><a class="docs-heading-anchor-permalink" href="#Built-in-Solvers:-Umfpack,-PardisoMKL,-LapackCPU" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, JuMP
+# ...
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=UmfpackSolver)) # default
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=LapackCPUSolver))</code></pre><h4 id="HSL-(requires-extension-MadNLPHSL)"><a class="docs-heading-anchor" href="#HSL-(requires-extension-MadNLPHSL)">HSL (requires extension <code>MadNLPHSL</code>)</a><a id="HSL-(requires-extension-MadNLPHSL)-1"></a><a class="docs-heading-anchor-permalink" href="#HSL-(requires-extension-MadNLPHSL)" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, MadNLPHSL, JuMP
+# ...
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=Ma27Solver))
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=Ma57Solver))
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=Ma77Solver))
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=Ma86Solver))
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=Ma97Solver))</code></pre><h4 id="Mumps-(requires-extension-MadNLPMumps)"><a class="docs-heading-anchor" href="#Mumps-(requires-extension-MadNLPMumps)">Mumps (requires extension <code>MadNLPMumps</code>)</a><a id="Mumps-(requires-extension-MadNLPMumps)-1"></a><a class="docs-heading-anchor-permalink" href="#Mumps-(requires-extension-MadNLPMumps)" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, MadNLPMumps, JuMP
+# ...
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=MumpsSolver))</code></pre><h4 id="Pardiso-(requires-extension-MadNLPPardiso)"><a class="docs-heading-anchor" href="#Pardiso-(requires-extension-MadNLPPardiso)">Pardiso (requires extension <code>MadNLPPardiso</code>)</a><a id="Pardiso-(requires-extension-MadNLPPardiso)-1"></a><a class="docs-heading-anchor-permalink" href="#Pardiso-(requires-extension-MadNLPPardiso)" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, MadNLPPardiso, JuMP
+# ...
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=PardisoSolver))
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=PardisoMKLSolver))</code></pre><h4 id="LapackGPU-(requires-extension-MadNLPGPU)"><a class="docs-heading-anchor" href="#LapackGPU-(requires-extension-MadNLPGPU)">LapackGPU (requires extension <code>MadNLPGPU</code>)</a><a id="LapackGPU-(requires-extension-MadNLPGPU)-1"></a><a class="docs-heading-anchor-permalink" href="#LapackGPU-(requires-extension-MadNLPGPU)" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, MadNLPGPU, JuMP
+# ...
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=LapackGPUSolver))</code></pre><h4 id="Schur-and-Schwarz-(requires-extension-MadNLPGraph)"><a class="docs-heading-anchor" href="#Schur-and-Schwarz-(requires-extension-MadNLPGraph)">Schur and Schwarz (requires extension <code>MadNLPGraph</code>)</a><a id="Schur-and-Schwarz-(requires-extension-MadNLPGraph)-1"></a><a class="docs-heading-anchor-permalink" href="#Schur-and-Schwarz-(requires-extension-MadNLPGraph)" title="Permalink"></a></h4><pre><code class="language-julia hljs">using MadNLP, MadNLPGraph, JuMP
+# ...
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=MadNLPSchwarz))
+model = Model(()-&gt;MadNLP.Optimizer(linear_solver=MadNLPSchur))</code></pre><p>The solvers in <code>MadNLPGraph</code> (<code>Schur</code> and <code>Schwarz</code>) use multi-thread parallelism; thus, Julia session should be started with <code>-t</code> flag.</p><pre><code class="language-sh hljs">julia -t 16 # to use 16 threads</code></pre><h2 id="Citing-MadNLP.jl"><a class="docs-heading-anchor" href="#Citing-MadNLP.jl">Citing MadNLP.jl</a><a id="Citing-MadNLP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-MadNLP.jl" title="Permalink"></a></h2><p>If you use MadNLP.jl in your research, we would greatly appreciate your citing it.</p><pre><code class="language-bibtex hljs">@article{shin2020graph,
+  title={Graph-Based Modeling and Decomposition of Energy Infrastructures},
+  author={Shin, Sungho and Coffrin, Carleton and Sundar, Kaarthik and Zavala, Victor M},
+  journal={arXiv preprint arXiv:2010.02404},
+  year={2020}
+}</code></pre><h2 id="Bug-reports-and-support"><a class="docs-heading-anchor" href="#Bug-reports-and-support">Bug reports and support</a><a id="Bug-reports-and-support-1"></a><a class="docs-heading-anchor-permalink" href="#Bug-reports-and-support" title="Permalink"></a></h2><p>Please report issues and feature requests via the <a href="https://github.com/MadNLP/MadNLP.jl/issues">GitHub issue tracker</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../CDDLib/">« JuliaPolyhedra/CDDLib.jl</a><a class="docs-footer-nextpage" href="../PATHSolver/">chkwon/PATHSolver.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/MiniZinc/index.html b/previews/PR3525/packages/MiniZinc/index.html
new file mode 100644
index 00000000000..a2ade744c58
--- /dev/null
+++ b/previews/PR3525/packages/MiniZinc/index.html
@@ -0,0 +1,57 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/MiniZinc.jl · JuMP</title><meta name="title" content="jump-dev/MiniZinc.jl · JuMP"/><meta property="og:title" content="jump-dev/MiniZinc.jl · JuMP"/><meta property="twitter:title" content="jump-dev/MiniZinc.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/MiniZinc.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Install"><span>Install</span></a></li><li><a class="tocitem" href="#Use-with-MathOptInterface"><span>Use with MathOptInterface</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li></ul></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/MiniZinc.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/MiniZinc.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MiniZinc.jl/blob/v0.3.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="MiniZinc.jl"><a class="docs-heading-anchor" href="#MiniZinc.jl">MiniZinc.jl</a><a id="MiniZinc.jl-1"></a><a class="docs-heading-anchor-permalink" href="#MiniZinc.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/MiniZinc.jl">MiniZinc.jl</a> is a wrapper for the <a href="https://www.minizinc.org">MiniZinc</a> constraint modeling language.</p><p>It provides a way to write MathOptInterface models to <code>.mzn</code> files, and a way to interact with <code>libminizinc</code>.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not part of the MiniZinc project.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>MiniZinc.jl</code> is licensed under the <a href="https://github.com/jump-dev/MiniZinc.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying project, <a href="https://github.com/MiniZinc/libminizinc">MiniZinc/libminizinc</a>, is licensed under the <a href="https://github.com/MiniZinc/libminizinc/blob/master/LICENSE.txt">MPL 2.0 license</a>.</p><h2 id="Install"><a class="docs-heading-anchor" href="#Install">Install</a><a id="Install-1"></a><a class="docs-heading-anchor-permalink" href="#Install" title="Permalink"></a></h2><p>Install MiniZinc.jl using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;MiniZinc&quot;)</code></pre><p><strong>Windows</strong></p><p>On Linux and macOS, this package automatically installs <code>libminizinc</code>. However, we&#39;re still working out problems with the install on Windows. To use MiniZinc.jl, you&#39;ll need to manually install a copy of <code>libminizinc</code> from <a href="https://www.minizinc.org">minizinc.org</a> or compile one yourself from <a href="https://github.com/MiniZinc/libminizinc">MiniZinc/libminizinc</a>.</p><p>To teach MiniZinc.jl where to look for <code>libminizinc</code>, set the <code>JULIA_LIBMINIZINC_DIR</code> environment variable. For example:</p><pre><code class="language-julia hljs">ENV[&quot;JULIA_LIBMINIZINC_DIR&quot;] = &quot;C:\\Program Files\\MiniZinc&quot;</code></pre><h2 id="Use-with-MathOptInterface"><a class="docs-heading-anchor" href="#Use-with-MathOptInterface">Use with MathOptInterface</a><a id="Use-with-MathOptInterface-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-MathOptInterface" title="Permalink"></a></h2><p>MiniZinc.jl supports the <a href="https://jump.dev/MathOptInterface.jl/stable/reference/standard_form/#Constraint-programming-sets">constraint programming sets</a> defined in MathOptInterface, as well as (in)equality constraints.</p><p>The following example solves the following constraint program:</p><pre><code class="nohighlight hljs">xᵢ ∈ {1, 2, 3} ∀i=1,2,3
+zⱼ ∈ {0, 1}    ∀j=1,2
+z₁ &lt;-&gt; x₁ != x₂
+z₂ &lt;-&gt; x₂ != x₃
+z₁ + z₂ = 1</code></pre><pre><code class="language-julia hljs">julia&gt; import MiniZinc
+
+julia&gt; const MOI = MiniZinc.MOI
+MathOptInterface
+
+julia&gt; function main()
+           model = MOI.Utilities.CachingOptimizer(
+               MiniZinc.Model{Int}(),
+               MiniZinc.Optimizer{Int}(&quot;chuffed&quot;),
+           )
+           # xᵢ ∈ {1, 2, 3} ∀i=1,2,3
+           x = MOI.add_variables(model, 3)
+           MOI.add_constraint.(model, x, MOI.Interval(1, 3))
+           MOI.add_constraint.(model, x, MOI.Integer())
+           # zⱼ ∈ {0, 1}    ∀j=1,2
+           z = MOI.add_variables(model, 2)
+           MOI.add_constraint.(model, z, MOI.ZeroOne())
+           # z₁ &lt;-&gt; x₁ != x₂
+           MOI.add_constraint(
+               model,
+               MOI.VectorOfVariables([z[1], x[1], x[2]]),
+               MOI.Reified(MOI.AllDifferent(2)),
+           )
+           # z₂ &lt;-&gt; x₂ != x₃
+           MOI.add_constraint(
+               model,
+               MOI.VectorOfVariables([z[2], x[2], x[3]]),
+               MOI.Reified(MOI.AllDifferent(2)),
+           )
+           # z₁ + z₂ = 1
+           MOI.add_constraint(model, 1 * z[1] + x[2], MOI.EqualTo(1))
+           MOI.optimize!(model)
+           x_star = MOI.get(model, MOI.VariablePrimal(), x)
+           z_star = MOI.get(model, MOI.VariablePrimal(), z)
+           return x_star, z_star
+       end
+main (generic function with 1 method)
+
+julia&gt; main()
+([1, 1, 3], [0, 1])</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>You can also call MiniZinc from JuMP, using any solver that <code>libminizinc</code> supports. By default, MiniZinc.jl is compiled with <code>&quot;highs&quot;</code>:</p><pre><code class="language-julia hljs">using JuMP
+import MiniZinc
+model = Model(() -&gt; MiniZinc.Optimizer{Float64}(&quot;highs&quot;))
+@variable(model, 1 &lt;= x[1:3] &lt;= 3, Int)
+@constraint(model, x in MOI.AllDifferent(3))
+@objective(model, Max, sum(i * x[i] for i in 1:3))
+optimize!(model)
+@show value.(x)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The MiniZinc <code>Optimizer{T}</code> supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{T}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{T}}</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.VariableIndex}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{T}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Parameter"><code>MOI.Parameter{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semicontinuous"><code>MOI.Semicontinuous{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Semiinteger"><code>MOI.Semiinteger{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.AllDifferent"><code>MOI.AllDifferent</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.BinPacking"><code>MOI.BinPacking{T}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Circuit"><code>MOI.Circuit</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.CountAtLeast"><code>MOI.CountAtLeast</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.CountBelongs"><code>MOI.CountBelongs</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.CountDistinct"><code>MOI.CountDistinct</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.CountGreaterThan"><code>MOI.CountGreaterThan</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Cumulative"><code>MOI.Cumulative</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Path"><code>MOI.Path</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Table"><code>MOI.Table{T}</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlock"><code>MOI.NLPBlock()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>Set options using <a href="../../moi/reference/models/#MathOptInterface.RawOptimizerAttribute"><code>MOI.RawOptimizerAttribute</code></a> in MOI or <a href="../../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a> in JuMP.</p><p><code>MiniZinc.jl</code> supports the following options:</p><ul><li><code>model_filename::String = &quot;&quot;</code>: the location at which to write out the <code>.mzn</code> file during optimization. This option can be helpful during debugging. If left empty, a temporary file will be used instead.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../KNITRO/">« jump-dev/KNITRO.jl</a><a class="docs-footer-nextpage" href="../MosekTools/">jump-dev/MosekTools.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/MosekTools/index.html b/previews/PR3525/packages/MosekTools/index.html
new file mode 100644
index 00000000000..83edc6f717b
--- /dev/null
+++ b/previews/PR3525/packages/MosekTools/index.html
@@ -0,0 +1,10 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/MosekTools.jl · JuMP</title><meta name="title" content="jump-dev/MosekTools.jl · JuMP"/><meta property="og:title" content="jump-dev/MosekTools.jl · JuMP"/><meta property="twitter:title" content="jump-dev/MosekTools.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/MosekTools.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li></ul></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/MosekTools.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/MosekTools.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MosekTools.jl/blob/v0.15.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="MosekTools.jl"><a class="docs-heading-anchor" href="#MosekTools.jl">MosekTools.jl</a><a id="MosekTools.jl-1"></a><a class="docs-heading-anchor-permalink" href="#MosekTools.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/MosekTools.jl">MosekTools.jl</a> is the <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface.jl</a> implementation for the MOSEK solver.</p><p>The low-level solver API for MOSEK is found in the package <a href="https://github.com/MOSEK/Mosek.jl">Mosek.jl</a>.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>MosekTools.jl is maintained by the JuMP community and is not officially supported by MOSEK. However, Mosek.jl <em>is</em> an officially supported product of MOSEK.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>MosekTools.jl</code> is licensed under the <a href="https://github.com/jump-dev/MosekTools.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver is a closed-source commercial product for which you must <a href="https://www.mosek.com">obtain a license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>The latest release of this package and the <code>master</code> branch are to be used with the latest release of Mosek.jl (which uses MOSEK v10).</p><p>To use MOSEK v9 (resp. v8), use the v0.12.x (resp. v0.7.x) releases of this package, and the <code>mosekv9</code> (resp. <code>mosekv8</code>) branch and v1.2.x (resp. v0.9.x) releases of Mosek.jl.</p><p>See the following table for a summary:</p><table><tr><th style="text-align: right">MOSEK</th><th style="text-align: right">Mosek.jl</th><th style="text-align: right">MosekTools.jl release</th><th style="text-align: right">MosekTools.jl branch</th></tr><tr><td style="text-align: right">v10</td><td style="text-align: right">v10</td><td style="text-align: right">v0.13</td><td style="text-align: right">master</td></tr><tr><td style="text-align: right">v9</td><td style="text-align: right">v0.12</td><td style="text-align: right">v0.12</td><td style="text-align: right">mosekv9</td></tr><tr><td style="text-align: right">v8</td><td style="text-align: right">v0.9</td><td style="text-align: right">v0.7</td><td style="text-align: right">mosekv8</td></tr></table><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><pre><code class="language-julia hljs">using JuMP
+using MosekTools
+model = Model(Mosek.Optimizer)
+set_attribute(model, &quot;QUIET&quot;, true)
+set_attribute(model, &quot;INTPNT_CO_TOL_DFEAS&quot;, 1e-7)</code></pre><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>The parameter <code>QUIET</code> is a special parameter that when set to <code>true</code> disables all Mosek printing output.</p><p>All other parameters can be found in the <a href="https://docs.mosek.com/8.1/capi/param-groups.html#doc-param-groups">Mosek documentation</a>.</p><p>Note that the prefix <code>MSK_IPAR_</code> (for integer parameters), <code>MSK_DPAR_</code> (for floating point parameters) or <code>MSK_SPAR_</code> (for string parameters) are optional. If they are not given, they are inferred from the type of the value. For example, in the example above, as <code>1e-7</code> is a floating point number, the parameters name used is <code>MSK_DPAR_INTPNT_CO_TOL_DFEAS</code>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../MiniZinc/">« jump-dev/MiniZinc.jl</a><a class="docs-footer-nextpage" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/MultiObjectiveAlgorithms/index.html b/previews/PR3525/packages/MultiObjectiveAlgorithms/index.html
new file mode 100644
index 00000000000..e54d3bda8f3
--- /dev/null
+++ b/previews/PR3525/packages/MultiObjectiveAlgorithms/index.html
@@ -0,0 +1,12 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/MultiObjectiveAlgorithms.jl · JuMP</title><meta name="title" content="jump-dev/MultiObjectiveAlgorithms.jl · JuMP"/><meta property="og:title" content="jump-dev/MultiObjectiveAlgorithms.jl · JuMP"/><meta property="twitter:title" content="jump-dev/MultiObjectiveAlgorithms.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/MultiObjectiveAlgorithms.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Algorithm"><span>Algorithm</span></a></li><li><a class="tocitem" href="#Other-optimizer-attributes"><span>Other optimizer attributes</span></a></li></ul></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/MultiObjectiveAlgorithms.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/MultiObjectiveAlgorithms.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/MultiObjectiveAlgorithms.jl/blob/v1.2.0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><img src="https://raw.githubusercontent.com/jump-dev/MultiObjectiveAlgorithms.jl/master/moa.png?raw=true" alt="An image of the Moa bird. Licensed into the Public Domain by https://freesvg.org/moa" width="100px"/><h1 id="MultiObjectiveAlgorithms.jl"><a class="docs-heading-anchor" href="#MultiObjectiveAlgorithms.jl">MultiObjectiveAlgorithms.jl</a><a id="MultiObjectiveAlgorithms.jl-1"></a><a class="docs-heading-anchor-permalink" href="#MultiObjectiveAlgorithms.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/MultiObjectiveAlgorithms.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/MultiObjectiveAlgorithms.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/MultiObjectiveAlgorithms.jl"><img src="https://codecov.io/gh/jump-dev/MultiObjectiveAlgorithms.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/MultiObjectiveAlgorithms.jl">MultiObjectiveAlgorithms.jl</a> (MOA) is a collection of algorithms for multi-objective optimization.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>MultiObjectiveAlgorithms.jl</code> is licensed under the <a href="https://github.com/jump-dev/MultiObjectiveAlgorithms.jl/blob/master/LICENSE.md">MPL 2.0 License</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install MOA using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;MultiObjectiveAlgorithms&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use <code>MultiObjectiveAlgorithms</code> with JuMP as follows:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+import MultiObjectiveAlgorithms as MOA
+model = JuMP.Model(() -&gt; MOA.Optimizer(HiGHS.Optimizer))
+set_attribute(model, MOA.Algorithm(), MOA.Dichotomy())
+set_attribute(model, MOA.SolutionLimit(), 4)</code></pre><p>Replace <code>HiGHS.Optimizer</code> with an optimizer capable of solving a single-objective instance of your optimization problem.</p><p>You may set additional optimizer attributes, the supported attributes depend on the choice of solution algorithm.</p><h2 id="Algorithm"><a class="docs-heading-anchor" href="#Algorithm">Algorithm</a><a id="Algorithm-1"></a><a class="docs-heading-anchor-permalink" href="#Algorithm" title="Permalink"></a></h2><p>Set the algorithm using the <code>MOA.Algorithm()</code> attribute.</p><p>The value must be one of the algorithms supported by MOA:</p><ul><li><code>MOA.Chalmet()</code></li><li><code>MOA.Dichotomy()</code></li><li><code>MOA.DominguezRios()</code></li><li><code>MOA.EpsilonConstraint()</code></li><li><code>MOA.Hierarchical()</code></li><li><code>MOA.KirlikSayin()</code></li><li><code>MOA.Lexicographic()</code> [default]</li><li><code>MOA.TambyVanderpooten()</code></li></ul><p>Consult their docstrings for details.</p><h2 id="Other-optimizer-attributes"><a class="docs-heading-anchor" href="#Other-optimizer-attributes">Other optimizer attributes</a><a id="Other-optimizer-attributes-1"></a><a class="docs-heading-anchor-permalink" href="#Other-optimizer-attributes" title="Permalink"></a></h2><p>There are a number of optimizer attributes supported by the algorithms in MOA.</p><p>Each algorithm supports only a subset of the attributes. Consult the algorithm&#39;s docstring for details on which attributes it supports, and how it uses them in the solution process.</p><ul><li><code>MOA.EpsilonConstraintStep()</code></li><li><code>MOA.LexicographicAllPermutations()</code></li><li><code>MOA.ObjectiveAbsoluteTolerance(index::Int)</code></li><li><code>MOA.ObjectivePriority(index::Int)</code></li><li><code>MOA.ObjectiveRelativeTolerance(index::Int)</code></li><li><code>MOA.ObjectiveWeight(index::Int)</code></li><li><code>MOA.SolutionLimit()</code></li><li><code>MOI.TimeLimitSec()</code></li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../MosekTools/">« jump-dev/MosekTools.jl</a><a class="docs-footer-nextpage" href="../Pajarito/">jump-dev/Pajarito.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/NEOSServer/index.html b/previews/PR3525/packages/NEOSServer/index.html
new file mode 100644
index 00000000000..7eaf2e1b35f
--- /dev/null
+++ b/previews/PR3525/packages/NEOSServer/index.html
@@ -0,0 +1,31 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>odow/NEOSServer.jl · JuMP</title><meta name="title" content="odow/NEOSServer.jl · JuMP"/><meta property="og:title" content="odow/NEOSServer.jl · JuMP"/><meta property="twitter:title" content="odow/NEOSServer.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li class="is-active"><a class="tocitem" href>odow/NEOSServer.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#The-NEOS-API"><span>The NEOS API</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#NEOS-Limits"><span>NEOS Limits</span></a></li></ul></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>odow/NEOSServer.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>odow/NEOSServer.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/odow/NEOSServer.jl/blob/v1.0.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="NEOSServer.jl"><a class="docs-heading-anchor" href="#NEOSServer.jl">NEOSServer.jl</a><a id="NEOSServer.jl-1"></a><a class="docs-heading-anchor-permalink" href="#NEOSServer.jl" title="Permalink"></a></h1><p><a href="https://github.com/odow/NEOSServer.jl/actions?query=workflow%3ACI"><img src="https://github.com/odow/NEOSServer.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/odow/NEOSServer.jl"><img src="https://codecov.io/gh/odow/NEOSServer.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/odow/NEOSServer.jl">NEOSServer.jl</a> is a wrapper for the <a href="http://www.neos-server.org/neos">NEOS Server</a>, a free internet-based service for solving numerical optimization problems.</p><p>See <a href="http://www.neos-server.org/neos/solvers/index.html">here</a> for the full list of solvers and input formats that NEOS supports.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>NEOSServer.jl</code> is licensed under the <a href="https://github.com/odow/NEOSServer.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>However, use of the <a href="http://www.neos-server.org/neos">NEOS Server</a> requires you to comply with <a href="http://www.neos-server.org/neos/termofuse.html">NEOS Server terms of use</a>. In particular, the commercial solvers are to be used solely for academic, non-commercial research purposes.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install NEOSServer.jl using the package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;NEOSServer&quot;)</code></pre><h2 id="The-NEOS-API"><a class="docs-heading-anchor" href="#The-NEOS-API">The NEOS API</a><a id="The-NEOS-API-1"></a><a class="docs-heading-anchor-permalink" href="#The-NEOS-API" title="Permalink"></a></h2><p>This package contains an interface for the <a href="http://www.neos-server.org/neos/NEOS-API.html">NEOS XML-RPC API</a>.</p><p>The following example shows how you can interact with the API. Wrapped XML-RPC functions begin with <code>neos_</code> and are exported.</p><pre><code class="language-julia hljs">using NEOSServer
+
+# Create a server. You must supply a valid email:
+server = NEOSServer.Server(&quot;me@mydomain.com&quot;)
+
+# Print the NEOS welcome message:
+println(neos_welcome(server))
+
+# Get an XML template:
+xml_string = neos_getSolverTemplate(server, &quot;milp&quot;, &quot;Cbc&quot;, &quot;AMPL&quot;)
+
+# Modify template with problem data...
+
+# Submit the XML job to NEOS:
+job = neos_submitJob(server, xml_string)
+
+# Get the status of the Job from NEOS:
+status = neos_getJobStatus(server, job)
+
+# Get the final results:
+results = neos_getFinalResults(server, job)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use NEOSServer.jl with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> as follows:</p><pre><code class="language-julia hljs">using JuMP, NEOSServer
+
+model = Model() do
+    NEOSServer.Optimizer(email=&quot;me@mydomain.com&quot;, solver=&quot;Ipopt&quot;)
+end</code></pre><p><strong>Note: <code>NEOSServer.Optimizer</code> is limited to the following solvers:</strong></p><ul><li><code>&quot;CPLEX&quot;</code></li><li><code>&quot;FICO-Xpress&quot;</code></li><li><code>&quot;Ipopt&quot;</code></li><li><code>&quot;Knitro&quot;</code></li><li><code>&quot;MOSEK&quot;</code></li><li><code>&quot;OCTERACT&quot;</code></li><li><code>&quot;SNOPT&quot;</code></li></ul><h2 id="NEOS-Limits"><a class="docs-heading-anchor" href="#NEOS-Limits">NEOS Limits</a><a id="NEOS-Limits-1"></a><a class="docs-heading-anchor-permalink" href="#NEOS-Limits" title="Permalink"></a></h2><p>NEOS currently limits jobs to an 8 hour time limit, 3 GB of memory, and a 16 MB submission file. If your model exceeds these limits, NEOSServer.jl may be unable to return useful information to the user.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../ProxSDP/">« mariohsouto/ProxSDP.jl</a><a class="docs-footer-nextpage" href="../OSQP/">osqp/OSQP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/NLopt/index.html b/previews/PR3525/packages/NLopt/index.html
new file mode 100644
index 00000000000..2f0025685e7
--- /dev/null
+++ b/previews/PR3525/packages/NLopt/index.html
@@ -0,0 +1,9 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>JuliaOpt/NLopt.jl · JuMP</title><meta name="title" content="JuliaOpt/NLopt.jl · JuMP"/><meta property="og:title" content="JuliaOpt/NLopt.jl · JuMP"/><meta property="twitter:title" content="JuliaOpt/NLopt.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li class="is-active"><a class="tocitem" href>JuliaOpt/NLopt.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li></ul></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>JuliaOpt/NLopt.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>JuliaOpt/NLopt.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/packages/NLopt.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="NLopt.jl"><a class="docs-heading-anchor" href="#NLopt.jl">NLopt.jl</a><a id="NLopt.jl-1"></a><a class="docs-heading-anchor-permalink" href="#NLopt.jl" title="Permalink"></a></h1><p><a href="https://github.com/JuliaOpt/NLopt.jl/actions?query=workflow%3ACI"><img src="https://github.com/JuliaOpt/NLopt.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/JuliaOpt/NLopt.jl"><img src="https://codecov.io/gh/JuliaOpt/NLopt.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/JuliaOpt/NLopt.jl">NLopt.jl</a> is a wrapper for the <a href="https://nlopt.readthedocs.io/en/latest/">NLopt</a> library.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>NLopt.jl</code> is licensed under the <a href="https://github.com/JuliaOpt/NLopt.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/stevengj/nlopt">stevengj/nlopt</a>, is licensed under the <a href="https://github.com/stevengj/nlopt/blob/master/COPYING">LGPL v3.0 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install <code>NLopt.jl</code> using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;NLopt&quot;)</code></pre><p>In addition to installing the <code>NLopt.jl</code> package, this will also download and install the NLopt binaries. You do not need to install NLopt separately.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>You can use NLopt with JuMP as follows:</p><pre><code class="language-julia hljs">using JuMP, NLopt
+model = Model(NLopt.Optimizer)
+set_attribute(model, &quot;algorithm&quot;, :LD_MMA)</code></pre><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>The <code>algorithm</code> attribute is required. The value must be one of the supported <a href="https://nlopt.readthedocs.io/en/latest/NLopt_Algorithms/">NLopt algorithms</a>.</p><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>For more details, see the <a href="https://github.com/JuliaOpt/NLopt.jl/blob/master/README.md">NLopt.jl README</a> or the <a href="https://nlopt.readthedocs.io/en/latest/">NLopt documentation</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../GAMS/">« GAMS-dev/GAMS.jl</a><a class="docs-footer-nextpage" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/OSQP/index.html b/previews/PR3525/packages/OSQP/index.html
new file mode 100644
index 00000000000..3f5fe67872e
--- /dev/null
+++ b/previews/PR3525/packages/OSQP/index.html
@@ -0,0 +1,9 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>osqp/OSQP.jl · JuMP</title><meta name="title" content="osqp/OSQP.jl · JuMP"/><meta property="og:title" content="osqp/OSQP.jl · JuMP"/><meta property="twitter:title" content="osqp/OSQP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li class="is-active"><a class="tocitem" href>osqp/OSQP.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Problem-class"><span>Problem class</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li></ul></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>osqp/OSQP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>osqp/OSQP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/osqp/OSQP.jl/blob/443706e34c2619acbe65281c60bbe850ca4a8fac/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="OSQP.jl"><a class="docs-heading-anchor" href="#OSQP.jl">OSQP.jl</a><a id="OSQP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#OSQP.jl" title="Permalink"></a></h1><p><a href="https://github.com/osqp/OSQP.jl/actions"><img src="https://github.com/osqp/OSQP.jl/workflows/CI/badge.svg" alt="Build Status"/></a>  <a href="http://codecov.io/github/osqp/OSQP.jl?branch=master"><img src="http://codecov.io/github/osqp/OSQP.jl/coverage.svg?branch=master" alt="codecov.io"/></a></p><p><a href="https://github.com/osqp/OSQP.jl">OSQP.jl</a> is a Julia wrapper for <a href="https://osqp.org/">OSQP</a>: the Operator Splitting QP Solver.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>OSQP.jl is licensed under the <a href="https://github.com/osqp/OSQP.jl/blob/master/LICENSE.md">Apache-2.0 license</a>.</p><p>The upstream solver, <a href="https://github.com/osqp/osqp">osqp/osqp</a> is also licensed under the <a href="https://github.com/osqp/osqp/blob/master/LICENSE">Apache-2.0 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install OSQP.jl using the Julia package manager</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;OSQP&quot;)</code></pre><h2 id="Problem-class"><a class="docs-heading-anchor" href="#Problem-class">Problem class</a><a id="Problem-class-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-class" title="Permalink"></a></h2><p>The OSQP (Operator Splitting Quadratic Program) solver is a numerical optimization package for solving problems in the form</p><pre><code class="nohighlight hljs">minimize        0.5 x&#39; P x + q&#39; x
+
+subject to      l &lt;= A x &lt;= u</code></pre><p>where <code>x in R^n</code> is the optimization variable. The objective function is defined by a positive semidefinite matrix <code>P in S^n_+</code> and vector <code>q in R^n</code>. The linear constraints are defined by matrix <code>A in R^{m x n}</code> and vectors <code>l in R^m U {-inf}^m</code>, <code>u in R^m U {+inf}^m</code>.</p><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>Detailed documentation is available at <a href="https://osqp.org/">https://osqp.org/</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../NEOSServer/">« odow/NEOSServer.jl</a><a class="docs-footer-nextpage" href="../CDCS/">oxfordcontrol/CDCS.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/PATHSolver/index.html b/previews/PR3525/packages/PATHSolver/index.html
new file mode 100644
index 00000000000..906d9ae1032
--- /dev/null
+++ b/previews/PR3525/packages/PATHSolver/index.html
@@ -0,0 +1,168 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>chkwon/PATHSolver.jl · JuMP</title><meta name="title" content="chkwon/PATHSolver.jl · JuMP"/><meta property="og:title" content="chkwon/PATHSolver.jl · JuMP"/><meta property="twitter:title" content="chkwon/PATHSolver.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li class="is-active"><a class="tocitem" href>chkwon/PATHSolver.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Use-with-the-C-API"><span>Use with the C API</span></a></li><li><a class="tocitem" href="#Thread-safety"><span>Thread safety</span></a></li><li><a class="tocitem" href="#Factorization-methods"><span>Factorization methods</span></a></li><li><a class="tocitem" href="#Manual-installation"><span>Manual installation</span></a></li></ul></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>chkwon/PATHSolver.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>chkwon/PATHSolver.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/chkwon/PATHSolver.jl/blob/v1.6.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="PATHSolver.jl"><a class="docs-heading-anchor" href="#PATHSolver.jl">PATHSolver.jl</a><a id="PATHSolver.jl-1"></a><a class="docs-heading-anchor-permalink" href="#PATHSolver.jl" title="Permalink"></a></h1><p><a href="https://github.com/chkwon/PATHSolver.jl/actions?query=workflow%3ACI"><img src="https://github.com/chkwon/PATHSolver.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/chkwon/PATHSolver.jl"><img src="https://codecov.io/gh/chkwon/PATHSolver.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/chkwon/PATHSolver.jl">PATHSolver.jl</a> is a wrapper for the <a href="http://pages.cs.wisc.edu/~ferris/path.html">PATH solver</a>.</p><p>The wrapper has two components:</p><ul><li>a thin wrapper around the C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><p>You can solve any complementarity problem using the wrapper around the C API, although you must manually provide the callback functions, including the Jacobian.</p><p>The MathOptInterface wrapper is more limited, supporting only linear complementarity problems, but it enables PATHSolver to be used with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not an official wrapper of PATH. However, we are in close contact with the PATH developers, and they have given us permission to re-distribute the PATH binaries for automatic installation.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>PATHSolver.jl</code> is licensed under the <a href="https://github.com/chkwon/PATHSolver.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://pages.cs.wisc.edu/~ferris/path.html">path</a> is closed source and requires a license.</p><p>Without a license, the PATH Solver can solve problem instances up to with up to 300 variables and 2000 non-zeros. For larger problems, <a href="http://pages.cs.wisc.edu/~ferris/path/julia/LICENSE">this web page</a> provides a temporary license that is valid for a year.</p><p>You can either store the license in the <code>PATH_LICENSE_STRING</code> environment variable, or you can use the <code>PATHSolver.c_api_License_SetString</code> function immediately after importing the <code>PATHSolver</code> package:</p><pre><code class="language-julia hljs">import PATHSolver
+PATHSolver.c_api_License_SetString(&quot;&lt;LICENSE STRING&gt;&quot;)</code></pre><p>where <code>&lt;LICENSE STRING&gt;</code> is replaced by the current license string.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install <code>PATHSolver.jl</code> as follows:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;PATHSolver&quot;)</code></pre><p>By default, <code>PATHSolver.jl</code> will download a copy of the underlying PATH solver. To use a different version of PATH, see the Manual Installation section below.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><pre><code class="language-julia hljs">julia&gt; using JuMP, PATHSolver
+
+julia&gt; M = [
+           0  0 -1 -1
+           0  0  1 -2
+           1 -1  2 -2
+           1  2 -2  4
+       ]
+4×4 Array{Int64,2}:
+ 0   0  -1  -1
+ 0   0   1  -2
+ 1  -1   2  -2
+ 1   2  -2   4
+
+julia&gt; q = [2, 2, -2, -6]
+4-element Array{Int64,1}:
+  2
+  2
+ -2
+ -6
+
+julia&gt; model = Model(PATHSolver.Optimizer)
+A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: Path 5.0.00
+
+julia&gt; set_optimizer_attribute(model, &quot;output&quot;, &quot;no&quot;)
+
+julia&gt; @variable(model, x[1:4] &gt;= 0)
+4-element Array{VariableRef,1}:
+ x[1]
+ x[2]
+ x[3]
+ x[4]
+
+julia&gt; @constraint(model, M * x .+ q ⟂ x)
+[-x[3] - x[4] + 2, x[3] - 2 x[4] + 2, x[1] - x[2] + 2 x[3] - 2 x[4] - 2, x[1] + 2 x[2] - 2 x[3] + 4 x[4] - 6, x[1], x[2], x[3], x[4]] ∈ MOI.Complements(4)
+
+julia&gt; optimize!(model)
+Reading options file /var/folders/bg/dzq_hhvx1dxgy6gb5510pxj80000gn/T/tmpiSsCRO
+Read of options file complete.
+
+Path 5.0.00 (Mon Aug 19 10:57:18 2019)
+Written by Todd Munson, Steven Dirkse, Youngdae Kim, and Michael Ferris
+
+julia&gt; value.(x)
+4-element Array{Float64,1}:
+ 2.8
+ 0.0
+ 0.7999999999999998
+ 1.2
+
+julia&gt; termination_status(model)
+LOCALLY_SOLVED::TerminationStatusCode = 4</code></pre><p>Note that options are set using <code>JuMP.set_optimizer_attribute</code>.</p><p>The list of options supported by PATH can be found here: https://pages.cs.wisc.edu/~ferris/path/options.pdf</p><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The Path 5.0.03 optimizer supports the following constraints and attributes.</p><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Complements"><code>MOI.Complements</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Complements"><code>MOI.Complements</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorQuadraticFunction"><code>MOI.VectorQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Complements"><code>MOI.Complements</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorNonlinearFunction"><code>MOI.VectorNonlinearFunction</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Complements"><code>MOI.Complements</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.Name"><code>MOI.Name()</code></a></li></ul><h2 id="Use-with-the-C-API"><a class="docs-heading-anchor" href="#Use-with-the-C-API">Use with the C API</a><a id="Use-with-the-C-API-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-the-C-API" title="Permalink"></a></h2><p><code>PATHSolver.jl</code> wraps the PATH C API using <code>PATHSolver.c_api_XXX</code> for the C method <code>XXX</code>. However, using the C API directly from Julia can be challenging, particularly with respect to avoiding issues with Julia&#39;s garbage collector.</p><p>Instead, we recommend that you use the <code>PATHSolver.solve_mcp</code> function, which wrappers the C API into a single call. See the docstring of <code>PATHSolver.solve_mcp</code> for a detailed description of the arguments.</p><p>Here is the same example using <code>PATHSolver.solve_mcp</code>. Note that you must manually construct the sparse Jacobian callback.</p><pre><code class="language-julia hljs">julia&gt; import PATHSolver
+
+julia&gt; M = [
+           0  0 -1 -1
+           0  0  1 -2
+           1 -1  2 -2
+           1  2 -2  4
+       ]
+4×4 Matrix{Int64}:
+ 0   0  -1  -1
+ 0   0   1  -2
+ 1  -1   2  -2
+ 1   2  -2   4
+
+julia&gt; q = [2, 2, -2, -6]
+4-element Vector{Int64}:
+  2
+  2
+ -2
+ -6
+
+julia&gt; function F(n::Cint, x::Vector{Cdouble}, f::Vector{Cdouble})
+           @assert n == length(x) == length(f)
+           f .= M * x .+ q
+           return Cint(0)
+       end
+F (generic function with 1 method)
+
+julia&gt; function J(
+           n::Cint,
+           nnz::Cint,
+           x::Vector{Cdouble},
+           col::Vector{Cint},
+           len::Vector{Cint},
+           row::Vector{Cint},
+           data::Vector{Cdouble},
+       )
+           @assert n == length(x) == length(col) == length(len) == 4
+           @assert nnz == length(row) == length(data)
+           i = 1
+           for c in 1:n
+               col[c], len[c] = i, 0
+               for r in 1:n
+                   if !iszero(M[r, c])
+                       row[i], data[i] = r, M[r, c]
+                       len[c] += 1
+                       i += 1
+                   end
+               end
+           end
+           return Cint(0)
+       end
+J (generic function with 1 method)
+
+julia&gt; status, z, info = PATHSolver.solve_mcp(
+           F,
+           J,
+           fill(0.0, 4),  # Lower bounds
+           fill(Inf, 4),  # Upper bounds
+           fill(0.0, 4);  # Starting point
+           nnz = 12,      # Number of nonzeros in the Jacobian
+           output = &quot;yes&quot;,
+       )
+Reading options file /var/folders/bg/dzq_hhvx1dxgy6gb5510pxj80000gn/T/jl_iftYBS
+ &gt; output yes
+Read of options file complete.
+
+Path 5.0.03 (Fri Jun 26 09:58:07 2020)
+Written by Todd Munson, Steven Dirkse, Youngdae Kim, and Michael Ferris
+
+Crash Log
+major  func  diff  size  residual    step       prox   (label)
+    0     0             1.2649e+01             0.0e+00 (f[    4])
+    1     2     4     2 1.0535e+01  8.0e-01    0.0e+00 (f[    1])
+    2     3     2     4 8.4815e-01  1.0e+00    0.0e+00 (f[    4])
+    3     4     0     3 4.4409e-16  1.0e+00    0.0e+00 (f[    3])
+pn_search terminated: no basis change.
+
+Major Iteration Log
+major minor  func  grad  residual    step  type prox    inorm  (label)
+    0     0     5     4 4.4409e-16           I 0.0e+00 4.4e-16 (f[    3])
+
+Major Iterations. . . . 0
+Minor Iterations. . . . 0
+Restarts. . . . . . . . 0
+Crash Iterations. . . . 3
+Gradient Steps. . . . . 0
+Function Evaluations. . 5
+Gradient Evaluations. . 4
+Basis Time. . . . . . . 0.000016
+Total Time. . . . . . . 0.044383
+Residual. . . . . . . . 4.440892e-16
+(PATHSolver.MCP_Solved, [2.8, 0.0, 0.8, 1.2], PATHSolver.Information(4.4408920985006247e-16, 0.0, 0.0, 0.044383, 1.6e-5, 0.0, 0, 0, 3, 5, 4, 0, 0, 0, 0, false, false, false, true, false, false, false))
+
+julia&gt; status
+MCP_Solved::MCP_Termination = 1
+
+julia&gt; z
+4-element Vector{Float64}:
+ 2.8
+ 0.0
+ 0.8
+ 1.2</code></pre><h2 id="Thread-safety"><a class="docs-heading-anchor" href="#Thread-safety">Thread safety</a><a id="Thread-safety-1"></a><a class="docs-heading-anchor-permalink" href="#Thread-safety" title="Permalink"></a></h2><p>PATH is not thread-safe and there are no known work-arounds. Do not run it in parallel using <code>Threads.@threads</code>. See <a href="https://github.com/chkwon/PATHSolver.jl/issues/62">issue #62</a> for more details.</p><h2 id="Factorization-methods"><a class="docs-heading-anchor" href="#Factorization-methods">Factorization methods</a><a id="Factorization-methods-1"></a><a class="docs-heading-anchor-permalink" href="#Factorization-methods" title="Permalink"></a></h2><p>By default, <code>PATHSolver.jl</code> will download the <a href="https://web.stanford.edu/group/SOL/software/lusol/">LUSOL</a> shared library. To use LUSOL, set the following options:</p><pre><code class="language-julia hljs">model = Model(PATHSolver.Optimizer)
+set_optimizer_attribute(model, &quot;factorization_method&quot;, &quot;blu_lusol&quot;)
+set_optimizer_attribute(model, &quot;factorization_library_name&quot;, PATHSolver.LUSOL_LIBRARY_PATH)</code></pre><p>To use <code>factorization_method umfpack</code> you will need the <code>umfpack</code> shared library that is available directly from the <a href="http://faculty.cse.tamu.edu/davis/suitesparse.html">developers of that code for academic use</a>.</p><h2 id="Manual-installation"><a class="docs-heading-anchor" href="#Manual-installation">Manual installation</a><a id="Manual-installation-1"></a><a class="docs-heading-anchor-permalink" href="#Manual-installation" title="Permalink"></a></h2><p>By default <code>PATHSolver.jl</code> will download a copy of the <code>libpath</code> library. If you already have one installed and want to use that, set the <code>PATH_JL_LOCATION</code> environment variable to point to the <code>libpath50.xx</code> library.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../MadNLP/">« MadNLP/MadNLP.jl</a><a class="docs-footer-nextpage" href="../Hypatia/">chriscoey/Hypatia.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Pajarito/index.html b/previews/PR3525/packages/Pajarito/index.html
new file mode 100644
index 00000000000..5160801c6b7
--- /dev/null
+++ b/previews/PR3525/packages/Pajarito/index.html
@@ -0,0 +1,30 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Pajarito.jl · JuMP</title><meta name="title" content="jump-dev/Pajarito.jl · JuMP"/><meta property="og:title" content="jump-dev/Pajarito.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Pajarito.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Pajarito.jl</a><ul class="internal"><li><a class="tocitem" href="#Pavito"><span>Pavito</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#MIP-and-continuous-solvers"><span>MIP and continuous solvers</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Cone-interface"><span>Cone interface</span></a></li><li><a class="tocitem" href="#Bug-reports-and-support"><span>Bug reports and support</span></a></li><li><a class="tocitem" href="#References"><span>References</span></a></li></ul></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Pajarito.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Pajarito.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Pajarito.jl/blob/4c7efa915bdc900b51d1c8290dc771403b11230a/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Pajarito.jl"><a class="docs-heading-anchor" href="#Pajarito.jl">Pajarito.jl</a><a id="Pajarito.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Pajarito.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Pajarito.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/Pajarito.jl/workflows/CI/badge.svg?branch=main" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Pajarito.jl"><img src="https://codecov.io/gh/jump-dev/Pajarito.jl/branch/main/graph/badge.svg" alt="codecov"/></a></p><p>Pajarito is a <strong>mixed-integer convex programming</strong> (MICP) solver package written in <a href="http://julialang.org/">Julia</a>.</p><p>MICP problems are convex except for restrictions that some variables take binary or integer values.</p><p>Pajarito solves MICP problems in conic form, by constructing sequential polyhedral outer approximations of the conic feasible set.</p><p>The underlying algorithm has theoretical finite-time convergence under reasonable assumptions.</p><p>Pajarito accesses state-of-the-art MILP solvers and continuous conic solvers through <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a>.</p><h2 id="Pavito"><a class="docs-heading-anchor" href="#Pavito">Pavito</a><a id="Pavito-1"></a><a class="docs-heading-anchor-permalink" href="#Pavito" title="Permalink"></a></h2><p>For algorithms that use a derivative-based nonlinear programming (NLP) solver (for example, Ipopt) instead of a conic solver, use <a href="https://github.com/jump-dev/Pavito.jl">Pavito</a>.</p><p>Pavito is a convex mixed-integer nonlinear programming (convex MINLP) solver. Because Pavito relies on gradient cuts, it can fail near points of non-differentiability. Pajarito may be more robust than Pavito on non-smooth problems.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Pajarito.jl</code> is licensed under the <a href="https://github.com/jump-dev/Pajarito.jl/blob/master/LICENSE.md">MPL 2.0 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Pajarito using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Pajarito&quot;)</code></pre><h2 id="MIP-and-continuous-solvers"><a class="docs-heading-anchor" href="#MIP-and-continuous-solvers">MIP and continuous solvers</a><a id="MIP-and-continuous-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#MIP-and-continuous-solvers" title="Permalink"></a></h2><p>The algorithm implemented by Pajarito itself is relatively simple, and most of the hard work is performed by the MIP outer approximation (OA) solver and the continuous conic solver.</p><p>Therefore, in addition to installing Pajarito, you must also install a mixed-integer linear programming solver and a continuous conic solver.</p><p><strong>The performance of Pajarito depends on these two types of solvers.</strong></p><p>The OA solver (typically a mixed-integer linear solver) is specified by the <code>oa_solver</code> option. You must first load the Julia package that provides this solver, for example, <code>using Gurobi</code>. The continuous conic solver is specified by the <code>conic_solver</code> option.</p><p>See JuMP&#39;s <a href="https://jump.dev/JuMP.jl/stable/installation/#Supported-solvers">list of supported solvers</a>.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use Pajarito with JuMP, use:</p><pre><code class="language-julia hljs">using JuMP, Pajarito, HiGHS, Hypatia
+model = Model(
+    optimizer_with_attributes(
+        Pajarito.Optimizer,
+        &quot;oa_solver&quot; =&gt; optimizer_with_attributes(
+            HiGHS.Optimizer,
+            MOI.Silent() =&gt; true,
+            &quot;mip_feasibility_tolerance&quot; =&gt; 1e-8,
+            &quot;mip_rel_gap&quot; =&gt; 1e-6,
+        ),
+        &quot;conic_solver&quot; =&gt;
+            optimizer_with_attributes(Hypatia.Optimizer, MOI.Silent() =&gt; true),
+    )
+)
+set_attribute(model, &quot;time_limit&quot;, 60)</code></pre><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>We list Pajarito&#39;s options below.</p><ul><li><code>verbose::Bool</code> toggles printing</li><li><code>tol_feas::Float64</code> is the feasibility tolerance for conic constraints</li><li><code>tol_rel_gap::Float64</code> is the relative optimality gap tolerance</li><li><code>tol_abs_gap::Float64</code> is the absolute optimality gap tolerance</li><li><code>time_limit::Float64</code> sets the time limit (in seconds)</li><li><code>iteration_limit::Int</code> sets the iteration limit (for the iterative method)</li><li><code>use_iterative_method::Union{Nothing,Bool}</code> toggles the iterative algorithm; if <code>nothing</code> is specified, Pajarito defaults to the OA-solver-driven (single tree) algorithm if lazy callbacks are supported by the OA solver</li><li><code>use_extended_form::Bool</code> toggles the use of extended formulations for the second-order cone</li><li><code>solve_relaxation::Bool</code> toggles solution of the continuous conic relaxation</li><li><code>solve_subproblems::Bool</code> toggles solution of the continuous conic subproblems</li><li><code>use_init_fixed_oa::Bool</code> toggles initial fixed OA cuts</li><li><code>oa_solver::Union{Nothing,MOI.OptimizerWithAttributes}</code> is the OA solver</li><li><code>conic_solver::Union{Nothing,MOI.OptimizerWithAttributes}</code> is the conic solver</li></ul><p><strong>Pajarito may require tuning of parameters to improve convergence.</strong></p><p>For example, it often helps to tighten the OA solver&#39;s integrality tolerance. OA solver and conic solver options must be specified directly to those solvers.</p><p>Note: if <code>solve_subproblems</code> is true, Pajarito usually returns a solution constructed from one of the conic solver&#39;s feasible solutions; since the conic solver is not subject to the same feasibility tolerances as the OA solver, Pajarito&#39;s solution will not necessarily satisfy <code>tol_feas</code>.</p><h2 id="Cone-interface"><a class="docs-heading-anchor" href="#Cone-interface">Cone interface</a><a id="Cone-interface-1"></a><a class="docs-heading-anchor-permalink" href="#Cone-interface" title="Permalink"></a></h2><p>Pajarito has a generic cone interface (see the <a href="https://github.com/jump-dev/Pajarito.jl/tree/main/src/Cones">cones folder</a> that allows the user to add support for new convex cones.</p><p>To illustrate, in the experimental package <a href="https://github.com/chriscoey/PajaritoExtras.jl">PajaritoExtras</a> we have extended Pajarito by adding support for several cones recognized by <a href="https://github.com/chriscoey/Hypatia.jl">Hypatia.jl</a> (a continuous conic solver with its own generic cone interface).</p><p>The examples folder of PajaritoExtras also contains many applied mixed-integer convex problems that are solved using Pajarito.</p><h2 id="Bug-reports-and-support"><a class="docs-heading-anchor" href="#Bug-reports-and-support">Bug reports and support</a><a id="Bug-reports-and-support-1"></a><a class="docs-heading-anchor-permalink" href="#Bug-reports-and-support" title="Permalink"></a></h2><p>Please report any issues via the GitHub <a href="https://github.com/jump-dev/Pajarito.jl/issues">issue tracker</a>.</p><p>All types of issues are welcome and encouraged; this includes bug reports, documentation typos, and feature requests. The <a href="https://discourse.julialang.org/c/domain/opt">Optimization (Mathematical)</a> category on Discourse is appropriate for general discussion.</p><h2 id="References"><a class="docs-heading-anchor" href="#References">References</a><a id="References-1"></a><a class="docs-heading-anchor-permalink" href="#References" title="Permalink"></a></h2><p>If you find Pajarito useful in your work, we kindly request that you cite the following paper (<a href="http://arxiv.org/abs/1808.05290">arXiv preprint</a>), which is recommended reading for advanced users:</p><pre><code class="nohighlight hljs">@article{CoeyLubinVielma2020,
+    title={Outer approximation with conic certificates for mixed-integer convex problems},
+    author={Coey, Chris and Lubin, Miles and Vielma, Juan Pablo},
+    journal={Mathematical Programming Computation},
+    volume={12},
+    number={2},
+    pages={249--293},
+    year={2020},
+    publisher={Springer}
+}</code></pre><p>Note this paper describes a legacy MathProgBase version of Pajarito, which is available on the <a href="https://github.com/jump-dev/Pajarito.jl/tree/mathprogbase"><code>mathprogbase</code> branch</a> of this repository. Starting with version v0.8.0, Pajarito supports MathOptInterface instead of MathProgBase.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../MultiObjectiveAlgorithms/">« jump-dev/MultiObjectiveAlgorithms.jl</a><a class="docs-footer-nextpage" href="../Pavito/">jump-dev/Pavito.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/ParametricOptInterface/index.html b/previews/PR3525/packages/ParametricOptInterface/index.html
new file mode 100644
index 00000000000..b0cf63275f7
--- /dev/null
+++ b/previews/PR3525/packages/ParametricOptInterface/index.html
@@ -0,0 +1,16 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/ParametricOptInterface.jl · JuMP</title><meta name="title" content="jump-dev/ParametricOptInterface.jl · JuMP"/><meta property="og:title" content="jump-dev/ParametricOptInterface.jl · JuMP"/><meta property="twitter:title" content="jump-dev/ParametricOptInterface.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/ParametricOptInterface.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#GSOC2020"><span>GSOC2020</span></a></li></ul></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>jump-dev/ParametricOptInterface.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/ParametricOptInterface.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/ParametricOptInterface.jl/blob/1721e7b07bd1991238802a438ecb5af566d4b555/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="ParametricOptInterface.jl"><a class="docs-heading-anchor" href="#ParametricOptInterface.jl">ParametricOptInterface.jl</a><a id="ParametricOptInterface.jl-1"></a><a class="docs-heading-anchor-permalink" href="#ParametricOptInterface.jl" title="Permalink"></a></h1><p><a href="https://jump.dev/ParametricOptInterface.jl/stable"><img src="https://img.shields.io/badge/docs-stable-blue.svg" alt="stable docs"/></a> <a href="https://jump.dev/ParametricOptInterface.jl/dev"><img src="https://img.shields.io/badge/docs-dev-blue.svg" alt="development docs"/></a> <a href="https://github.com/jump-dev/ParametricOptInterface.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/ParametricOptInterface.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/ParametricOptInterface.jl"><img src="https://codecov.io/gh/jump-dev/ParametricOptInterface.jl/branch/master/graph/badge.svg" alt="Coverage"/></a></p><p><a href="https://github.com/jump-dev/ParametricOptInterface.jl">ParametricOptInterface.jl</a> is a package that adds parameters to models in JuMP and MathOptInterface.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>ParametricOptInterface.jl</code> is licensed under the <a href="https://github.com/jump-dev/ParametricOptInterface.jl/blob/master/LICENSE.md">MIT License</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install ParametricOptInterface using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;ParametricOptInterface&quot;)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>The <a href="https://jump.dev/ParametricOptInterface.jl/stable/">documentation for ParametricOptInterface.jl</a> includes a detailed description of the theory behind the package, along with examples, tutorials, and an API reference.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use ParametricOptInterface with JuMP by following this brief example:</p><pre><code class="language-julia hljs">using JuMP, HiGHS
+import ParametricOptInterface as POI
+model = Model(() -&gt; POI.Optimizer(HiGHS.Optimizer()))
+@variable(model, x)
+@variable(model, p in POI.Parameter(1.0))
+@constraint(model, cons, x + p &gt;= 3)
+@objective(model, Min, 2x)
+optimize!(model)
+MOI.set(model, POI.ParameterValue(), p, 2.0)
+optimize!(model)</code></pre><h2 id="GSOC2020"><a class="docs-heading-anchor" href="#GSOC2020">GSOC2020</a><a id="GSOC2020-1"></a><a class="docs-heading-anchor-permalink" href="#GSOC2020" title="Permalink"></a></h2><p>ParametricOptInterface began as a <a href="https://summerofcode.withgoogle.com/archive/2020/projects/4959861055422464">NumFOCUS sponsored Google Summer of Code (2020) project</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../DiffOpt/">« jump-dev/DiffOpt.jl</a><a class="docs-footer-nextpage" href="../PolyJuMP/">jump-dev/PolyJuMP.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Pavito/index.html b/previews/PR3525/packages/Pavito/index.html
new file mode 100644
index 00000000000..9030e427835
--- /dev/null
+++ b/previews/PR3525/packages/Pavito/index.html
@@ -0,0 +1,16 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Pavito.jl · JuMP</title><meta name="title" content="jump-dev/Pavito.jl · JuMP"/><meta property="og:title" content="jump-dev/Pavito.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Pavito.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Pavito.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Bug-reports-and-support"><span>Bug reports and support</span></a></li></ul></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Pavito.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Pavito.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Pavito.jl/blob/v0.3.7/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Pavito.jl"><a class="docs-heading-anchor" href="#Pavito.jl">Pavito.jl</a><a id="Pavito.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Pavito.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Pavito.jl/actions"><img src="https://github.com/jump-dev/Pavito.jl/workflows/CI/badge.svg" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Pavito.jl"><img src="https://codecov.io/gh/jump-dev/Pavito.jl/branch/master/graph/badge.svg" alt="Coverage"/></a></p><p><a href="https://github.com/jump-dev/Pavito.jl">Pavito.jl</a> is a mixed-integer convex programming (MICP) solver package written in <a href="http://julialang.org/">Julia</a>.</p><p>MICP problems are convex, except for restrictions that some variables take binary or integer values.</p><p>Pavito solves MICP problems by constructing sequential polyhedral outer-approximations of the convex feasible set, similar to <a href="https://projects.coin-or.org/Bonmin">Bonmin</a>.</p><p>Pavito accesses state-of-the-art MILP solvers and continuous, derivative-based nonlinear programming (NLP) solvers through <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a>.</p><p>For algorithms that use a conic solver instead of an NLP solver, use <a href="https://github.com/jump-dev/Pajarito.jl">Pajarito</a>. Pajarito is a robust mixed-integer conic solver that can handle such established problem classes as mixed-integer second-order cone programming (MISOCP) and mixed-integer semidefinite programming (MISDP).</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Pavito.jl</code> is licensed under the <a href="https://github.com/jump-dev/Pavito.jl/blob/master/LICENSE.md">MPL 2.0 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Pavito using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Pavito&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use Pavito with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>, use <code>Pavito.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP, Pavito
+import GLPK, Ipopt
+model = Model(
+    optimizer_with_attributes(
+        Pavito.Optimizer,
+        &quot;mip_solver&quot; =&gt; optimizer_with_attributes(GLPK.Optimizer),
+        &quot;cont_solver&quot; =&gt;
+            optimizer_with_attributes(Ipopt.Optimizer, &quot;print_level&quot; =&gt; 0),
+    ),
+)</code></pre><p>The algorithm implemented by Pavito itself is relatively simple; most of the hard work is performed by the MILP solver passed as <code>mip_solver</code> and the NLP solver passed as <code>cont_solver</code>.</p><p><strong>The performance of Pavito depends on these two types of solvers.</strong></p><p>For better performance, you should use a commercial MILP solver such as CPLEX or Gurobi.</p><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>The following optimizer attributes can set to a <code>Pavito.Optimizer</code> to modify its behavior:</p><ul><li><code>log_level::Int</code> Verbosity flag: 0 for quiet, higher for basic solve info</li><li><code>timeout::Float64</code> Time limit for algorithm (in seconds)</li><li><code>rel_gap::Float64</code> Relative optimality gap termination condition</li><li><code>mip_solver_drives::Bool</code> Let MILP solver manage convergence (&quot;branch and cut&quot;)</li><li><code>mip_solver::MOI.OptimizerWithAttributes</code> MILP solver</li><li><code>cont_solver::MOI.OptimizerWithAttributes</code> Continuous NLP solver</li></ul><p><strong>Pavito is not yet numerically robust and may require tuning of parameters to improve convergence.</strong></p><p>If the default parameters don&#39;t work for you, please let us know by opening an issue.</p><p>For improved Pavito performance, MILP solver integrality tolerance and feasibility tolerances should typically be tightened, for example to <code>1e-8</code>.</p><h2 id="Bug-reports-and-support"><a class="docs-heading-anchor" href="#Bug-reports-and-support">Bug reports and support</a><a id="Bug-reports-and-support-1"></a><a class="docs-heading-anchor-permalink" href="#Bug-reports-and-support" title="Permalink"></a></h2><p>Please report any issues via the <a href="https://github.com/jump-dev/Pavito.jl/issues">GitHub issue tracker</a>. All types of issues are welcome and encouraged; this includes bug reports, documentation typos, feature requests, etc. The <a href="https://discourse.julialang.org/c/domain/opt">Optimization (Mathematical)</a> category on Discourse is appropriate for general discussion.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Pajarito/">« jump-dev/Pajarito.jl</a><a class="docs-footer-nextpage" href="../SCS/">jump-dev/SCS.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Plasmo/index.html b/previews/PR3525/packages/Plasmo/index.html
new file mode 100644
index 00000000000..95cf7570e63
--- /dev/null
+++ b/previews/PR3525/packages/Plasmo/index.html
@@ -0,0 +1,49 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>plasmo-dev/Plasmo.jl · JuMP</title><meta name="title" content="plasmo-dev/Plasmo.jl · JuMP"/><meta property="og:title" content="plasmo-dev/Plasmo.jl · JuMP"/><meta property="twitter:title" content="plasmo-dev/Plasmo.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li class="is-active"><a class="tocitem" href>plasmo-dev/Plasmo.jl</a><ul class="internal"><li><a class="tocitem" href="#Overview"><span>Overview</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Simple-Example"><span>Simple Example</span></a></li><li><a class="tocitem" href="#Acknowledgments"><span>Acknowledgments</span></a></li><li><a class="tocitem" href="#Citing-Plasmo.jl"><span>Citing Plasmo.jl</span></a></li></ul></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>plasmo-dev/Plasmo.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>plasmo-dev/Plasmo.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/plasmo-dev/Plasmo.jl/blob/37f80fe96781dc647b3a5e0ed26b9eb0a994bc4a/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><img src="https://github.com/plasmo-dev/Plasmo.jl/blob/main/docs/plasmo_logo.svg?raw=true"/><p><a href="https://github.com/plasmo-dev/Plasmo.jl/actions"><img src="https://github.com/plasmo-dev/Plasmo.jl/workflows/CI/badge.svg" alt="CI"/></a> <a href="https://codecov.io/gh/plasmo-dev/Plasmo.jl"><img src="https://codecov.io/gh/plasmo-dev/Plasmo.jl/branch/master/graph/badge.svg" alt="codecov"/></a> <a href="https://plasmo-dev.github.io/Plasmo.jl/dev/"><img src="https://img.shields.io/badge/docs-latest-blue.svg" alt/></a> <a href="https://zenodo.org/badge/latestdoi/96967382"><img src="https://zenodo.org/badge/96967382.svg" alt="DOI"/></a></p><h1 id="Plasmo.jl"><a class="docs-heading-anchor" href="#Plasmo.jl">Plasmo.jl</a><a id="Plasmo.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Plasmo.jl" title="Permalink"></a></h1><p><a href="https://github.com/plasmo-dev/Plasmo.jl">Plasmo.jl</a> (Platform for Scalable Modeling and Optimization) is a graph-based algebraic modeling framework that adopts a modular style to create mathematical optimization problems and manage distributed and hierarchical structures. The package has been developed as a <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> extension and consequently supports  most JuMP syntax and functions. </p><h2 id="Overview"><a class="docs-heading-anchor" href="#Overview">Overview</a><a id="Overview-1"></a><a class="docs-heading-anchor-permalink" href="#Overview" title="Permalink"></a></h2><p>The core data structure in Plasmo.jl is the <code>OptiGraph</code>. The optigraph contains a set of optinodes which represent self-contained optimization problems and optiedges that represent coupling between optinodes (which produces an underlying <a href="https://en.wikipedia.org/wiki/Hypergraph">hypergraph</a> structure of optinodes and optiedges). Optigraphs can further be embedded within other optigraphs to create nested hierarchical graph structures. The graph structures obtained using Plasmo.jl can be used for simple model and data management, but they can also be used to perform graph partitioning or develop interfaces to structured optimization solvers.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>Plasmo is licensed under the <a href="https://github.com/plasmo-dev/Plasmo.jl/blob/main/LICENSE.md">MPL 2.0 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Plasmo using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Plasmo&quot;)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>The latest documentation is available through <a href="https://plasmo-dev.github.io/Plasmo.jl/dev/">GitHub Pages</a>. Additional examples can be found in the <a href="https://github.com/plasmo-dev/Plasmo.jl/tree/main/examples">examples</a> folder.</p><h2 id="Simple-Example"><a class="docs-heading-anchor" href="#Simple-Example">Simple Example</a><a id="Simple-Example-1"></a><a class="docs-heading-anchor-permalink" href="#Simple-Example" title="Permalink"></a></h2><pre><code class="language-julia hljs">using Plasmo
+using Ipopt
+
+#create an optigraph
+graph = OptiGraph()
+
+#add nodes to an optigraph
+@optinode(graph, n1)
+@optinode(graph, n2)
+
+#add variables, constraints, and objective functions to nodes
+@variable(n1, 0 &lt;= x &lt;= 2)
+@variable(n1, 0 &lt;= y &lt;= 3)
+@constraint(n1, x+y &lt;= 4)
+@objective(n1, Min, x)
+
+@variable(n2,x)
+@NLconstraint(n2, exp(x) &gt;= 2)
+
+#add a linkconstraint to couple nodes
+@linkconstraint(graph, n1[:x] == n2[:x])
+
+#optimize with Ipopt
+set_optimizer(graph, Ipopt.Optimizer)
+optimize!(graph)
+
+#Print solution values
+println(&quot;n1[:x] = &quot;, value(n1[:x]))
+println(&quot;n2[:x] = &quot;, value(n2[:x]))</code></pre><h2 id="Acknowledgments"><a class="docs-heading-anchor" href="#Acknowledgments">Acknowledgments</a><a id="Acknowledgments-1"></a><a class="docs-heading-anchor-permalink" href="#Acknowledgments" title="Permalink"></a></h2><p>This code is based on work supported by the following funding agencies:</p><ul><li>U.S. Department of Energy (DOE), Office of Science, under Contract No. DE-AC02-06CH11357</li><li>DOE Office of Electricity Delivery and Energy Reliability’s Advanced Grid Research and Development program at Argonne National Laboratory</li><li>National Science Foundation under award NSF-EECS-1609183 and under award CBET-1748516</li></ul><p>The primary developer is Jordan Jalving (@jalving) with support from the following contributors.  </p><ul><li>Victor Zavala (University of Wisconsin-Madison)</li><li>Yankai Cao (University of British Columbia)</li><li>Kibaek Kim (Argonne National Laboratory)</li><li>Sungho Shin (University of Wisconsin-Madison)</li></ul><h2 id="Citing-Plasmo.jl"><a class="docs-heading-anchor" href="#Citing-Plasmo.jl">Citing Plasmo.jl</a><a id="Citing-Plasmo.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-Plasmo.jl" title="Permalink"></a></h2><p>If you find Plasmo.jl useful for your work, you may cite the <a href="https://link.springer.com/article/10.1007/s12532-022-00223-3">manuscript</a> as:</p><pre><code class="nohighlight hljs">@article{JalvingShinZavala2022,
+  title={A Graph-Based Modeling Abstraction for Optimization: Concepts and Implementation in Plasmo.jl},
+  author={Jordan Jalving and Sungho Shin and Victor M. Zavala},
+  journal={Mathematical Programming Computation},
+  year={2022},
+  volume={14},
+  pages={699 - 747}
+}</code></pre><p>There is also a freely available <a href="https://arxiv.org/abs/2006.05378">pre-print</a>:</p><pre><code class="nohighlight hljs">@misc{JalvingShinZavala2020,
+title = {A Graph-Based Modeling Abstraction for Optimization: Concepts and Implementation in Plasmo.jl},
+author = {Jordan Jalving and Sungho Shin and Victor M. Zavala},
+year = {2020},
+eprint = {2006.05378},
+archivePrefix = {arXiv},
+primaryClass = {math.OC}
+}</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../SDDP/">« odow/SDDP.jl</a><a class="docs-footer-nextpage" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/PolyJuMP/index.html b/previews/PR3525/packages/PolyJuMP/index.html
new file mode 100644
index 00000000000..96b2ddb31ae
--- /dev/null
+++ b/previews/PR3525/packages/PolyJuMP/index.html
@@ -0,0 +1,7 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/PolyJuMP.jl · JuMP</title><meta name="title" content="jump-dev/PolyJuMP.jl · JuMP"/><meta property="og:title" content="jump-dev/PolyJuMP.jl · JuMP"/><meta property="twitter:title" content="jump-dev/PolyJuMP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/PolyJuMP.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li></ul></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>jump-dev/PolyJuMP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/PolyJuMP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/PolyJuMP.jl/blob/v0.7.0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="PolyJuMP.jl"><a class="docs-heading-anchor" href="#PolyJuMP.jl">PolyJuMP.jl</a><a id="PolyJuMP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#PolyJuMP.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/PolyJuMP.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/PolyJuMP.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/PolyJuMP.jl"><img src="https://codecov.io/gh/jump-dev/PolyJuMP.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/PolyJuMP.jl">PolyJuMP.jl</a> is a <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> extension for formulating and solving polynomial optimization problems.</p><p>These problems can then be solved using <a href="https://github.com/jump-dev/SumOfSquares.jl">SumOfSquares.jl</a>.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>PolyJuMP.jl</code> is licensed under the <a href="https://github.com/jump-dev/PolyJuMP.jl/blob/master/LICENSE.md">MIT license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install <code>PolyJuMP</code> using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;PolyJuMP&quot;)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>Documentation for <code>PolyJuMP.jl</code> is included in the <a href="https://jump.dev/SumOfSquares.jl/stable">documentation for SumOfSquares.jl</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../ParametricOptInterface/">« jump-dev/ParametricOptInterface.jl</a><a class="docs-footer-nextpage" href="../SumOfSquares/">jump-dev/SumOfSquares.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/ProxSDP/index.html b/previews/PR3525/packages/ProxSDP/index.html
new file mode 100644
index 00000000000..99c9b271966
--- /dev/null
+++ b/previews/PR3525/packages/ProxSDP/index.html
@@ -0,0 +1,59 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>mariohsouto/ProxSDP.jl · JuMP</title><meta name="title" content="mariohsouto/ProxSDP.jl · JuMP"/><meta property="og:title" content="mariohsouto/ProxSDP.jl · JuMP"/><meta property="twitter:title" content="mariohsouto/ProxSDP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li class="is-active"><a class="tocitem" href>mariohsouto/ProxSDP.jl</a><ul class="internal"><li><a class="tocitem" href="#Overview-of-problems-ProxSDP-can-solve"><span>Overview of problems ProxSDP can solve</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Usage"><span>Usage</span></a></li><li><a class="tocitem" href="#Citing-this-package"><span>Citing this package</span></a></li><li><a class="tocitem" href="#Disclaimer"><span>Disclaimer</span></a></li><li><a class="tocitem" href="#ROAD-MAP"><span>ROAD MAP</span></a></li></ul></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>mariohsouto/ProxSDP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>mariohsouto/ProxSDP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/mariohsouto/ProxSDP.jl/blob/de648be90e63a69b8d8e4f82371760065d242fe9/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><img src="https://raw.githubusercontent.com/mariohsouto/ProxSDP_aux/19f261a912e29c77cb68ee2957a5b41128824f22/logo_proxSDP.png?raw=true" alt="logo"/><p><a href="https://github.com/mariohsouto/ProxSDP.jl/actions?query=workflow%3ACI"><img src="https://github.com/mariohsouto/ProxSDP.jl/workflows/CI/badge.svg?branch=master" alt/></a> <a href="http://codecov.io/github/mariohsouto/ProxSDP.jl?branch=master"><img src="http://codecov.io/github/mariohsouto/ProxSDP.jl/coverage.svg?branch=master" alt/></a> <a href="https://mariohsouto.github.io/ProxSDP.jl/latest/"><img src="https://img.shields.io/badge/docs-latest-blue.svg" alt/></a></p><p><a href="https://github.com/mariohsouto/ProxSDP.jl">ProxSDP</a> is an open-source semidefinite programming (<a href="https://en.wikipedia.org/wiki/Semidefinite_programming">SDP</a>) solver based on the paper <a href="https://arxiv.org/abs/1810.05231">&quot;Exploiting Low-Rank Structure in Semidefinite Programming by Approximate Operator Splitting&quot;</a>.</p><p>The main advantage of ProxSDP over other state-of-the-art solvers is the ability to exploit the <strong>low-rank</strong> structure inherent to several SDP problems.</p><h2 id="Overview-of-problems-ProxSDP-can-solve"><a class="docs-heading-anchor" href="#Overview-of-problems-ProxSDP-can-solve">Overview of problems ProxSDP can solve</a><a id="Overview-of-problems-ProxSDP-can-solve-1"></a><a class="docs-heading-anchor-permalink" href="#Overview-of-problems-ProxSDP-can-solve" title="Permalink"></a></h2><ul><li>General conic convex optimization problems with the presence of the <a href="https://web.stanford.edu/~boyd/papers/pdf/semidef_prog.pdf">positive semidefinite cone</a>, <a href="https://web.stanford.edu/~boyd/papers/pdf/socp.pdf">second-order cone</a> and <a href="https://www.math.ucla.edu/~tom/LP.pdf">positive orthant</a>;</li><li>Semidefinite relaxation of nonconvex problems, for example, <a href="http://www-math.mit.edu/~goemans/PAPERS/maxcut-jacm.pdf">max-cut</a>, <a href="https://arxiv.org/pdf/cs/0606083.pdf">binary MIMO</a>, <a href="http://authorstest.library.caltech.edu/141/1/TPS_OPF_2_tech.pdf">optimal power flow</a>, <a href="https://web.stanford.edu/~boyd/papers/pdf/sensor_selection.pdf">sensor localization</a>, <a href="https://en.wikipedia.org/wiki/Sum-of-squares_optimization">sum-of-squares</a>;</li><li>Control theory problems with <a href="https://en.wikipedia.org/wiki/Linear_matrix_inequality">LMI</a> constraints;</li><li>Nuclear norm minimization problems, for example, <a href="https://statweb.stanford.edu/~candes/papers/MatrixCompletion.pdf">matrix completion</a>;</li></ul><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>ProxSDP is licensed under the <a href="https://github.com/mariohsouto/ProxSDP.jl/blob/master/LICENSE">MIT License</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install ProxSDP using Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;ProxSDP&quot;)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>The online documentation is available at <a href="https://mariohsouto.github.io/ProxSDP.jl/latest/">https://mariohsouto.github.io/ProxSDP.jl/latest/</a>.</p><h2 id="Usage"><a class="docs-heading-anchor" href="#Usage">Usage</a><a id="Usage-1"></a><a class="docs-heading-anchor-permalink" href="#Usage" title="Permalink"></a></h2><p>Consider the semidefinite programming relaxation of the <a href="http://www-math.mit.edu/~goemans/PAPERS/maxcut-jacm.pdf">max-cut</a> problem, as in:</p><pre><code class="nohighlight hljs">max   0.25 * W•X
+s.t.  diag(X) = 1,
+      X ≽ 0,</code></pre><h3 id="JuMP"><a class="docs-heading-anchor" href="#JuMP">JuMP</a><a id="JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP" title="Permalink"></a></h3><p>This problem can be solved by the following code using <strong>ProxSDP</strong> and <a href="https://github.com/JuliaOpt/JuMP.jl">JuMP</a>.</p><pre><code class="language-julia hljs"># Load packages
+using ProxSDP, JuMP, LinearAlgebra
+# Number of vertices
+n = 4
+# Graph weights
+W = [
+    18.0  -5.0  -7.0  -6.0
+    -5.0   6.0   0.0  -1.0
+    -7.0   0.0   8.0  -1.0
+    -6.0  -1.0  -1.0   8.0
+]
+# Build Max-Cut SDP relaxation via JuMP
+model = Model(ProxSDP.Optimizer)
+set_optimizer_attribute(model, &quot;log_verbose&quot;, true)
+set_optimizer_attribute(model, &quot;tol_gap&quot;, 1e-4)
+set_optimizer_attribute(model, &quot;tol_feasibility&quot;, 1e-4)
+@variable(model, X[1:n, 1:n], PSD)
+@objective(model, Max, 0.25 * LinearAlgebra.dot(W, X))
+@constraint(model, LinearAlgebra.diag(X) .== 1)
+# Solve optimization problem with ProxSDP
+optimize!(model)
+# Retrieve solution
+Xsol = value.(X)</code></pre><h3 id="Convex.jl"><a class="docs-heading-anchor" href="#Convex.jl">Convex.jl</a><a id="Convex.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Convex.jl" title="Permalink"></a></h3><p>Another alternative is to use <strong>ProxSDP</strong> via <a href="https://github.com/jump-dev/Convex.jl">Convex.jl</a>:</p><pre><code class="language-julia hljs"># Load packages
+using Convex, ProxSDP
+# Number of vertices
+n = 4
+# Graph weights
+W = [
+    18.0  -5.0  -7.0  -6.0
+    -5.0   6.0   0.0  -1.0
+    -7.0   0.0   8.0  -1.0
+    -6.0  -1.0  -1.0   8.0
+]
+# Define optimization problem
+X = Semidefinite(n)
+problem = maximize(0.25 * dot(W, X), diag(X) == 1)
+# Solve optimization problem with ProxSDP
+solve!(problem, ProxSDP.Optimizer(log_verbose=true, tol_gap=1e-4, tol_feasibility=1e-4))
+# Get the objective value
+problem.optval
+# Retrieve solution
+evaluate(X)</code></pre><h2 id="Citing-this-package"><a class="docs-heading-anchor" href="#Citing-this-package">Citing this package</a><a id="Citing-this-package-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-this-package" title="Permalink"></a></h2><p>The published version of the paper can be found <a href="https://doi.org/10.1080/02331934.2020.1823387">here</a> and the arXiv version <a href="https://arxiv.org/pdf/1810.05231.pdf">here</a>.</p><p>We kindly request that you cite the paper as:</p><pre><code class="language-bibtex hljs">@article{souto2020exploiting,
+  author = {Mario Souto and Joaquim D. Garcia and \&#39;Alvaro Veiga},
+  title = {Exploiting low-rank structure in semidefinite programming by approximate operator splitting},
+  journal = {Optimization},
+  pages = {1-28},
+  year  = {2020},
+  publisher = {Taylor &amp; Francis},
+  doi = {10.1080/02331934.2020.1823387},
+  URL = {https://doi.org/10.1080/02331934.2020.1823387}
+}</code></pre><p>The preprint version of the paper can be found <a href="https://arxiv.org/abs/1810.05231">here</a>.</p><h2 id="Disclaimer"><a class="docs-heading-anchor" href="#Disclaimer">Disclaimer</a><a id="Disclaimer-1"></a><a class="docs-heading-anchor-permalink" href="#Disclaimer" title="Permalink"></a></h2><ul><li>ProxSDP is a research software, therefore it should not be used in production.</li><li>Please open an issue if you find any problems, developers will try to fix and find alternatives.</li><li>There is no continuous development for 32-bit systems, the package should work, but might reach some issues.</li><li>ProxSDP assumes primal and dual feasibility.</li></ul><h2 id="ROAD-MAP"><a class="docs-heading-anchor" href="#ROAD-MAP">ROAD MAP</a><a id="ROAD-MAP-1"></a><a class="docs-heading-anchor-permalink" href="#ROAD-MAP" title="Permalink"></a></h2><ul><li>Support for exponential and power cones</li><li>Warm start</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Juniper/">« lanl-ansi/Juniper.jl</a><a class="docs-footer-nextpage" href="../NEOSServer/">odow/NEOSServer.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SCIP/index.html b/previews/PR3525/packages/SCIP/index.html
new file mode 100644
index 00000000000..fc226a54c3b
--- /dev/null
+++ b/previews/PR3525/packages/SCIP/index.html
@@ -0,0 +1,13 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>scipopt/SCIP.jl · JuMP</title><meta name="title" content="scipopt/SCIP.jl · JuMP"/><meta property="og:title" content="scipopt/SCIP.jl · JuMP"/><meta property="twitter:title" content="scipopt/SCIP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li class="is-active"><a class="tocitem" href>scipopt/SCIP.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Custom-installations"><span>Custom installations</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Design-considerations"><span>Design considerations</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>scipopt/SCIP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>scipopt/SCIP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/scipopt/SCIP.jl/blob/v0.11.14/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="SCIP.jl"><a class="docs-heading-anchor" href="#SCIP.jl">SCIP.jl</a><a id="SCIP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SCIP.jl" title="Permalink"></a></h1><p><a href="https://github.com/scipopt/SCIP.jl/actions?query=workflow%3ACI"><img src="https://github.com/scipopt/SCIP.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/scipopt/SCIP.jl"><img src="https://codecov.io/gh/scipopt/SCIP.jl/branch/master/graph/badge.svg" alt="codecov"/></a> <a href="https://pkgs.genieframework.com?packages=SCIP"><img src="https://shields.io/endpoint?url=https://pkgs.genieframework.com/api/v1/badge/SCIP" alt="Genie Downloads"/></a></p><p><a href="https://github.com/scipopt/SCIP.jl">SCIP.jl</a> is a Julia interface to the <a href="https://scipopt.org">SCIP</a> solver.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the <a href="https://www.scipopt.org/">SCIP project</a> with the help of the JuMP community.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SCIP.jl</code> is licensed under the <a href="https://github.com/scipopt/SCIP.jl/blob/master/LICENSE">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/scipopt/scip">scipopt/scip</a>, is licensed under the <a href="https://github.com/scipopt/scip/blob/master/LICENSE">Apache 2.0 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install SCIP.jl using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;SCIP&quot;)</code></pre><p>On MacOS and Linux, installing the SCIP Julia package will work out of the box and install the <a href="https://github.com/JuliaBinaryWrappers/SCIP_jll.jl"><code>SCIP_jll.jl</code></a> and <a href="https://github.com/JuliaBinaryWrappers/SCIP_PaPILO_jll.jl"><code>SCIP_PaPILO_jll.jl</code></a> dependencies.</p><p>On Windows, a separate installation of SCIP is still mandatory, as detailed in the &quot;Custom installation&quot; section below.</p><h2 id="Custom-installations"><a class="docs-heading-anchor" href="#Custom-installations">Custom installations</a><a id="Custom-installations-1"></a><a class="docs-heading-anchor-permalink" href="#Custom-installations" title="Permalink"></a></h2><p>If you use an older Julia version, Windows, or you want a custom SCIP installation, you must manually install the SCIP binaries.</p><p>Once installed, set the <code>SCIPOPTDIR</code> environment variable to point to the installation path, that is, depending on your operating system, <code>$SCIPOPTDIR/lib/libscip.so</code>, <code>$SCIPOPTDIR/lib/libscip.dylib</code>, or <code>$SCIPOPTDIR/bin/scip.dll</code> must exist.</p><p>Then, install <code>SCIP.jl</code> using <code>Pkg.add</code> and <code>Pkg.build</code>:</p><pre><code class="language-julia hljs">ENV[&quot;SCIPOPTDIR&quot;] = &quot;/Users/Oscar/code/SCIP&quot;
+import Pkg
+Pkg.add(&quot;SCIP&quot;)
+Pkg.build(&quot;SCIP&quot;)</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>Use SCIP with JuMP as follows:</p><pre><code class="language-julia hljs">using JuMP, SCIP
+model = Model(SCIP.Optimizer)
+set_attribute(model, &quot;display/verblevel&quot;, 0)
+set_attribute(model, &quot;limits/gap&quot;, 0.05)</code></pre><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>See the <a href="https://scip.zib.de/doc-8.0.0/html/PARAMETERS.php">SCIP documentation</a> for a list of supported options.</p><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The SCIP optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction"><code>MOI.ScalarQuadraticFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Integer"><code>MOI.Integer</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Interval"><code>MOI.Interval{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.LessThan"><code>MOI.LessThan{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VariableIndex"><code>MOI.VariableIndex</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ZeroOne"><code>MOI.ZeroOne</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS1"><code>MOI.SOS1{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SOS2"><code>MOI.SOS2{Float64}</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/nonlinear/#MathOptInterface.NLPBlock"><code>MOI.NLPBlock()</code></a></li><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li><li><a href="../../moi/reference/callbacks/#MathOptInterface.UserCutCallback"><code>MOI.UserCutCallback()</code></a></li></ul><h2 id="Design-considerations"><a class="docs-heading-anchor" href="#Design-considerations">Design considerations</a><a id="Design-considerations-1"></a><a class="docs-heading-anchor-permalink" href="#Design-considerations" title="Permalink"></a></h2><h3 id="Wrapping-the-public-API"><a class="docs-heading-anchor" href="#Wrapping-the-public-API">Wrapping the public API</a><a id="Wrapping-the-public-API-1"></a><a class="docs-heading-anchor-permalink" href="#Wrapping-the-public-API" title="Permalink"></a></h3><p>All of the public API methods are wrapped and available within the <code>SCIP</code> package. This includes the <code>scip_*.h</code> and <code>pub_*.h</code> headers that are collected in <code>scip.h</code>, as well as all default constraint handlers (<code>cons_*.h</code>.) </p><p>The wrapped functions do not transform any data structures and work on the <em>raw</em> pointers (for example, <code>SCIP*</code> in C, <code>Ptr{SCIP_}</code> in Julia). Convenience wrapper functions based on Julia types are added as needed.</p><h3 id="Memory-management"><a class="docs-heading-anchor" href="#Memory-management">Memory management</a><a id="Memory-management-1"></a><a class="docs-heading-anchor-permalink" href="#Memory-management" title="Permalink"></a></h3><p>Programming with SCIP requires dealing with variable and constraint objects that use <a href="https://scip.zib.de/doc-8.0.0/html/OBJ.php">reference counting</a> for memory management.</p><p>The <code>SCIP.Optimizer</code> wrapper type collects lists of <code>SCIP_VAR*</code> and <code>SCIP_CONS*</code> under the hood, and it releases all references when it is garbage collected itself (via <code>finalize</code>).</p><p>When adding a variable (<code>add_variable</code>) or a constraint (<code>add_linear_constraint</code>), an integer index is returned. This index can be used to retrieve the <code>SCIP_VAR*</code> or <code>SCIP_CONS*</code> pointer via <code>get_var</code> and <code>get_cons</code> respectively.</p><h3 id="Supported-nonlinear-operators"><a class="docs-heading-anchor" href="#Supported-nonlinear-operators">Supported nonlinear operators</a><a id="Supported-nonlinear-operators-1"></a><a class="docs-heading-anchor-permalink" href="#Supported-nonlinear-operators" title="Permalink"></a></h3><p>Supported operators in nonlinear expressions are as follows:</p><ul><li><code>+</code></li><li><code>-</code></li><li><code>*</code></li><li><code>/</code></li><li><code>^</code></li><li><code>sqrt</code></li><li><code>exp</code></li><li><code>log</code></li><li><code>abs</code></li><li><code>cos</code></li><li><code>sin</code></li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Clarabel/">« oxfordcontrol/Clarabel.jl</a><a class="docs-footer-nextpage" href="../../extensions/introduction/">Introduction »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SCS/index.html b/previews/PR3525/packages/SCS/index.html
new file mode 100644
index 00000000000..b118312576b
--- /dev/null
+++ b/previews/PR3525/packages/SCS/index.html
@@ -0,0 +1,71 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/SCS.jl · JuMP</title><meta name="title" content="jump-dev/SCS.jl · JuMP"/><meta property="og:title" content="jump-dev/SCS.jl · JuMP"/><meta property="twitter:title" content="jump-dev/SCS.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/SCS.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-Convex.jl"><span>Use with Convex.jl</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Linear-solvers"><span>Linear solvers</span></a></li><li><a class="tocitem" href="#Low-level-wrapper"><span>Low-level wrapper</span></a></li></ul></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/SCS.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/SCS.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/SCS.jl/blob/v1.3.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="SCS.jl"><a class="docs-heading-anchor" href="#SCS.jl">SCS.jl</a><a id="SCS.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SCS.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/SCS.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/SCS.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/SCS.jl"><img src="https://codecov.io/gh/jump-dev/SCS.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/SCS.jl">SCS.jl</a> is a wrapper for the <a href="https://github.com/cvxgrp/scs">SCS</a> splitting cone solver.</p><p>SCS can solve linear programs, second-order cone programs, semidefinite programs, exponential cone programs, and power cone programs.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not a project of the SCS developers.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SCS.jl</code> is licensed under the <a href="https://github.com/jump-dev/SCS.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/cvxgrp/scs">cvxgrp/scs</a>, is licensed under the <a href="https://github.com/cvxgrp/scs/blob/master/LICENSE.txt">MIT License</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install SCS.jl using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;SCS&quot;)</code></pre><p>In addition to installing the <code>SCS.jl</code> package, this will also download and install the SCS binaries. (You do not need to install SCS separately.)</p><p>To use a custom binary, read the <a href="https://jump.dev/JuMP.jl/stable/developers/custom_solver_binaries/">Custom solver binaries</a> section of the JuMP documentation.</p><h2 id="Use-with-Convex.jl"><a class="docs-heading-anchor" href="#Use-with-Convex.jl">Use with Convex.jl</a><a id="Use-with-Convex.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-Convex.jl" title="Permalink"></a></h2><p>This example shows how we can model a simple knapsack problem with Convex and use SCS to solve it.</p><pre><code class="language-julia hljs">using Convex, SCS
+items  = [:Gold, :Silver, :Bronze]
+values = [5.0, 3.0, 1.0]
+weights = [2.0, 1.5, 0.3]
+
+# Define a variable of size 3, each index representing an item
+x = Variable(3)
+p = maximize(x&#39; * values, 0 &lt;= x, x &lt;= 1, x&#39; * weights &lt;= 3)
+solve!(p, SCS.Optimizer)
+println([items x.value])
+# [:Gold 0.9999971880377178
+#  :Silver 0.46667637765641057
+#  :Bronze 0.9999998036351865]</code></pre><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>This example shows how we can model a simple knapsack problem with JuMP and use SCS to solve it.</p><pre><code class="language-julia hljs">using JuMP, SCS
+items  = [:Gold, :Silver, :Bronze]
+values = Dict(:Gold =&gt; 5.0,  :Silver =&gt; 3.0,  :Bronze =&gt; 1.0)
+weight = Dict(:Gold =&gt; 2.0,  :Silver =&gt; 1.5,  :Bronze =&gt; 0.3)
+
+model = Model(SCS.Optimizer)
+@variable(model, 0 &lt;= take[items] &lt;= 1)  # Define a variable for each item
+@objective(model, Max, sum(values[item] * take[item] for item in items))
+@constraint(model, sum(weight[item] * take[item] for item in items) &lt;= 3)
+optimize!(model)
+println(value.(take))
+# 1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+#     Dimension 1, Symbol[:Gold, :Silver, :Bronze]
+# And data, a 3-element Array{Float64,1}:
+#  1.0000002002226671
+#  0.4666659513182934
+#  1.0000007732744878</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The SCS optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Reals"><code>MOI.Reals</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.DualExponentialCone"><code>MOI.DualExponentialCone</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.DualPowerCone"><code>MOI.DualPowerCone{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.ExponentialCone"><code>MOI.ExponentialCone</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.PowerCone"><code>MOI.PowerCone{Float64}</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.Zeros"><code>MOI.Zeros</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction{Float64}</code></a> in <code>SCS.ScaledPSDCone</code></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>All SCS solver options can be set through <code>Convex.jl</code> or <code>MathOptInterface.jl</code>.</p><p>For example:</p><pre><code class="language-julia hljs">model = Model(optimizer_with_attributes(SCS.Optimizer, &quot;max_iters&quot; =&gt; 10))
+
+# via MathOptInterface:
+optimizer = SCS.Optimizer()
+MOI.set(optimizer, MOI.RawOptimizerAttribute(&quot;max_iters&quot;), 10)
+MOI.set(optimizer, MOI.RawOptimizerAttribute(&quot;verbose&quot;), 0)</code></pre><p>Common options are:</p><ul><li><code>max_iters</code>: the maximum number of iterations to take</li><li><code>verbose</code>: turn printing on (<code>1</code>) or off (<code>0</code>)</li></ul><p>See the <a href="https://github.com/cvxgrp/scs/blob/3aaa93c7aa04c7001df5e51b81f21b126dfa99b3/include/glbopts.h#L35"><code>glbopts.h</code> header</a> for other options.</p><h2 id="Linear-solvers"><a class="docs-heading-anchor" href="#Linear-solvers">Linear solvers</a><a id="Linear-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Linear-solvers" title="Permalink"></a></h2><p><code>SCS</code> uses a linear solver internally, see <a href="https://www.cvxgrp.org/scs/linear_solver/index.html#linear-system-solver">this section</a> of <code>SCS</code> documentation. <code>SCS.jl</code> ships with</p><ul><li><code>SCS.DirectSolver</code> (sparse direct, the default) and</li><li><code>SCS.LinearSolver</code> (sparse indirect, by conjugate gradient)</li></ul><p>enabled.</p><p>The find currently available linear solvers one can inspect <code>SCS.available_solvers</code>:</p><pre><code class="language-julia hljs">julia&gt; using SCS
+
+julia&gt; SCS.available_solvers
+2-element Vector{DataType}:
+ SCS.DirectSolver
+ SCS.IndirectSolver</code></pre><p>To select the linear solver of choice:</p><ul><li>pass the <code>linear_solver</code> option to <a href="../../api/JuMP/#JuMP.optimizer_with_attributes"><code>optimizer_with_attributes</code></a>, or to <a href="../../moi/reference/models/#MathOptInterface.OptimizerWithAttributes"><code>MOI.OptimizerWithAttributes</code></a>;</li><li>specify the solver as the first argument when using <code>scs_solve</code> directly (see the low-level wrapper section below).</li></ul><h3 id="SCS-with-MKL-Pardiso-linear-solver"><a class="docs-heading-anchor" href="#SCS-with-MKL-Pardiso-linear-solver">SCS with MKL Pardiso linear solver</a><a id="SCS-with-MKL-Pardiso-linear-solver-1"></a><a class="docs-heading-anchor-permalink" href="#SCS-with-MKL-Pardiso-linear-solver" title="Permalink"></a></h3><p>To enable the MKL Pardiso (direct sparse) solver one needs to load <code>MKL_jll</code> <strong>before</strong> <code>SCS</code>:</p><pre><code class="language-julia hljs">julia&gt; import Pkg
+
+julia&gt; Pkg.add(Pkg.PackageSpec(name = &quot;MKL_jll&quot;, version = &quot;2022.2&quot;))
+
+julia&gt; using MKL_jll    # This must be called before `using SCS`.
+
+julia&gt; using SCS
+
+julia&gt; SCS.available_solvers
+3-element Vector{DataType}:
+ SCS.DirectSolver
+ SCS.IndirectSolver
+ SCS.MKLDirectSolver</code></pre><p>The <code>MKLDirectSolver</code> is available on <code>Linux x86_64</code> platform only.</p><h3 id="SCS-with-Sparse-GPU-indirect-solver-(CUDA-only)"><a class="docs-heading-anchor" href="#SCS-with-Sparse-GPU-indirect-solver-(CUDA-only)">SCS with Sparse GPU indirect solver (CUDA only)</a><a id="SCS-with-Sparse-GPU-indirect-solver-(CUDA-only)-1"></a><a class="docs-heading-anchor-permalink" href="#SCS-with-Sparse-GPU-indirect-solver-(CUDA-only)" title="Permalink"></a></h3><p>To enable the indirect linear solver on GPU one needs to load <code>CUDA_jll</code> <strong>before</strong> <code>SCS</code>:</p><pre><code class="language-julia hljs">julia&gt; import Pkg
+
+julia&gt; Pkg.add(Pkg.PackageSpec(name = &quot;CUDA_jll&quot;, version = &quot;10.1&quot;))
+
+julia&gt; using CUDA_jll  # This must be called before `using SCS`.
+
+julia&gt; using SCS
+
+julia&gt; SCS.available_solvers
+3-element Array{DataType,1}:
+ SCS.DirectSolver
+ SCS.IndirectSolver
+ SCS.GpuIndirectSolver</code></pre><p>The <code>GpuIndirectSolver</code> is available on <code>Linux x86_64</code> platform only.</p><h2 id="Low-level-wrapper"><a class="docs-heading-anchor" href="#Low-level-wrapper">Low-level wrapper</a><a id="Low-level-wrapper-1"></a><a class="docs-heading-anchor-permalink" href="#Low-level-wrapper" title="Permalink"></a></h2><p>SCS.jl provides a low-level interface to solve a problem directly, without interfacing through MathOptInterface.</p><p><strong>This is an advanced interface with a risk of incorrect usage. For new users, we recommend that you use the JuMP or Convex interfaces instead.</strong></p><p>SCS solves a problem of the form:</p><pre><code class="nohighlight hljs">minimize        1/2 * x&#39; * P * x + c&#39; * x
+subject to      A * x + s = b
+                s in K</code></pre><p>where <code>K</code> is a product cone of:</p><ul><li>zero cone</li><li>positive orthant <code>{ x | x ≥ 0 }</code></li><li>box cone <code>{ (t,x) | t*l ≤ x ≤ t*u}</code></li><li>second-order cone (SOC) <code>{ (t,x) | ||x||_2 ≤ t }</code></li><li>semi-definite cone (SDC) <code>{ X | X is psd }</code></li><li>exponential cone <code>{ (x,y,z) | y e^(x/y) ≤ z, y&gt;0 }</code></li><li>power cone <code>{ (x,y,z) | x^a * y^(1-a) ≥ |z|, x ≥ 0, y ≥ 0 }</code></li><li>dual power cone <code>{ (u,v,w) | (u/a)^a * (v/(1-a))^(1-a) ≥ |w|, u ≥ 0, v ≥ 0 }</code>.</li></ul><p>To solve this problem with SCS, call <code>SCS.scs_solve</code>; see the docstring for details.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../Pavito/">« jump-dev/Pavito.jl</a><a class="docs-footer-nextpage" href="../SDPA/">jump-dev/SDPA.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SDDP/index.html b/previews/PR3525/packages/SDDP/index.html
new file mode 100644
index 00000000000..70de9834dc2
--- /dev/null
+++ b/previews/PR3525/packages/SDDP/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>odow/SDDP.jl · JuMP</title><meta name="title" content="odow/SDDP.jl · JuMP"/><meta property="og:title" content="odow/SDDP.jl · JuMP"/><meta property="twitter:title" content="odow/SDDP.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li class="is-active"><a class="tocitem" href>odow/SDDP.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Help"><span>Help</span></a></li></ul></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>odow/SDDP.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>odow/SDDP.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/odow/SDDP.jl/blob/v1.6.3/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><img src="https://raw.githubusercontent.com/odow/SDDP.jl/e9de84e0a4b57374bd9e0c95148da1501816e4c5/docs/src/assets/logo.png?raw=true" alt="logo" width="100px"/><h1 id="SDDP.jl"><a class="docs-heading-anchor" href="#SDDP.jl">SDDP.jl</a><a id="SDDP.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SDDP.jl" title="Permalink"></a></h1><p><a href="https://github.com/odow/SDDP.jl/actions?query=workflow%3ACI"><img src="https://github.com/odow/SDDP.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/odow/SDDP.jl"><img src="https://codecov.io/gh/odow/SDDP.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/odow/SDDP.jl">SDDP.jl</a> is a JuMP extension for solving large convex multistage stochastic programming problems using stochastic dual dynamic programming.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SDDP.jl</code> is licensed under the <a href="https://github.com/odow/SDDP.jl/blob/master/LICENSE.md">MPL 2.0 license</a>.</p><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>You can find the documentation at <a href="https://odow.github.io/SDDP.jl/stable">https://odow.github.io/SDDP.jl/stable/</a>.</p><h2 id="Help"><a class="docs-heading-anchor" href="#Help">Help</a><a id="Help-1"></a><a class="docs-heading-anchor-permalink" href="#Help" title="Permalink"></a></h2><p>If you need help, please <a href="https://github.com/odow/SDDP.jl/issues/new">open a GitHub issue</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../BilevelJuMP/">« joaquimg/BilevelJuMP.jl</a><a class="docs-footer-nextpage" href="../Plasmo/">plasmo-dev/Plasmo.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SDPA/index.html b/previews/PR3525/packages/SDPA/index.html
new file mode 100644
index 00000000000..7f06fccbf53
--- /dev/null
+++ b/previews/PR3525/packages/SDPA/index.html
@@ -0,0 +1,15 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/SDPA.jl · JuMP</title><meta name="title" content="jump-dev/SDPA.jl · JuMP"/><meta property="og:title" content="jump-dev/SDPA.jl · JuMP"/><meta property="twitter:title" content="jump-dev/SDPA.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/SDPA.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#MathOptInterface-API"><span>MathOptInterface API</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li></ul></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/SDPA.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/SDPA.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/SDPA.jl/blob/v0.5.0/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="SDPA.jl"><a class="docs-heading-anchor" href="#SDPA.jl">SDPA.jl</a><a id="SDPA.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SDPA.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/SDPA.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/SDPA.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/SDPA.jl"><img src="https://codecov.io/gh/jump-dev/SDPA.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/SDPA.jl">SDPA.jl</a> is a wrapper for the <a href="http://sdpa.sourceforge.net/">SDPA</a> semidefinite programming solver in double precision floating point arithmetic.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not a product of the SDPA developers.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SDPA.jl</code> is licensed under the <a href="https://github.com/jump-dev/SDPA.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="http://sdpa.sourceforge.net/">SDPA</a> is licensed under the <a href="https://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html">GPL v2 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install SDPA using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;SDPA&quot;)</code></pre><p>In addition to installing the SDPA.jl package, this will also download and install the SDPA binaries. (You do not need to install SDPA separately.)</p><p>If you see an error similar to:</p><pre><code class="language-julia hljs">INFO: Precompiling module GZip.
+ERROR: LoadError: LoadError: error compiling anonymous: could not load library &quot;libz&quot;</code></pre><p>please see <a href="https://github.com/JuliaIO/GZip.jl/issues/54">GZip.jl#54</a> or <a href="https://github.com/FluxML/Flux.jl/issues/343">Flux.jl#343</a>. In particular, in Ubuntu this issue may be resolved by running</p><pre><code class="language-bash hljs">sudo apt-get install zlib1g-dev</code></pre><p>See <a href="https://github.com/ericphanson/SDPAFamily.jl">SDPAFamily</a> for the other solvers, SDPA-GMP, SDPA-DD, and SDPA-QD of the family.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><pre><code class="language-julia hljs">using JuMP, SDPA
+model = Model(SDPA.Optimizer)
+set_attribute(model, &quot;Mode&quot;, SDPA.PARAMETER_DEFAULT)</code></pre><h2 id="MathOptInterface-API"><a class="docs-heading-anchor" href="#MathOptInterface-API">MathOptInterface API</a><a id="MathOptInterface-API-1"></a><a class="docs-heading-anchor-permalink" href="#MathOptInterface-API" title="Permalink"></a></h2><p>The SDPA optimizer supports the following constraints and attributes.</p><p>List of supported objective functions:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveFunction"><code>MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}</code></a></li></ul><p>List of supported variable types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></li><li><a href="../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a></li></ul><p>List of supported constraint types:</p><ul><li><a href="../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction{Float64}</code></a> in <a href="../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{Float64}</code></a></li></ul><p>List of supported model attributes:</p><ul><li><a href="../../moi/reference/models/#MathOptInterface.ObjectiveSense"><code>MOI.ObjectiveSense()</code></a></li></ul><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>SDPA has three modes that give default value to all ten parameters.</p><p>The following table gives the default values for each parameter and mode.</p><table><tr><th style="text-align: right">Parameter</th><th style="text-align: right"><code>PARAMETER_DEFAULT</code></th><th style="text-align: right"><code>PARAMETER_UNSTABLE_BUT_FAST</code></th><th style="text-align: right"><code>PARAMETER_STABLE_BUT_SLOW</code></th></tr><tr><td style="text-align: right">MaxIteration</td><td style="text-align: right">100</td><td style="text-align: right">100</td><td style="text-align: right">1000</td></tr><tr><td style="text-align: right">EpsilonStar</td><td style="text-align: right">1.0e-7</td><td style="text-align: right">1.0e-7</td><td style="text-align: right">1.0e-7</td></tr><tr><td style="text-align: right">LambdaStar</td><td style="text-align: right">1.0e+2</td><td style="text-align: right">1.0e+2</td><td style="text-align: right">1.0e+4</td></tr><tr><td style="text-align: right">OmegaStar</td><td style="text-align: right">2.0</td><td style="text-align: right">2.0</td><td style="text-align: right">2.0</td></tr><tr><td style="text-align: right">LowerBound</td><td style="text-align: right">1.0e+5</td><td style="text-align: right">1.0e+5</td><td style="text-align: right">1.0e+5</td></tr><tr><td style="text-align: right">UpperBound</td><td style="text-align: right">1.0e+5</td><td style="text-align: right">1.0e+5</td><td style="text-align: right">1.0e+5</td></tr><tr><td style="text-align: right">BetaStar</td><td style="text-align: right">0.1</td><td style="text-align: right">0.01</td><td style="text-align: right">0.1</td></tr><tr><td style="text-align: right">BetaBar</td><td style="text-align: right">0.2</td><td style="text-align: right">0.02</td><td style="text-align: right">0.3</td></tr><tr><td style="text-align: right">GammaStar</td><td style="text-align: right">0.9</td><td style="text-align: right">0.95</td><td style="text-align: right">0.8</td></tr><tr><td style="text-align: right">EpsilonDash</td><td style="text-align: right">1.0e-7</td><td style="text-align: right">1.0e-7</td><td style="text-align: right">1.0e-7</td></tr></table><p>By default, we put SDPA in the <code>SDPA.PARAMETER_DEFAULT</code> mode.</p><p>Change the mode using the <code>&quot;Mode&quot;</code> option:</p><pre><code class="language-julia hljs">using JuMP, SDPA
+model = Model(SDPA.Optimizer)
+set_attribute(model, &quot;Mode&quot;, SDPA.PARAMETER_STABLE_BUT_SLOW)</code></pre><p>Note that the parameters are set in the order they are given, so you can set a mode and then modify parameters from this mode.</p><pre><code class="language-julia hljs">using JuMP, SDPA
+model = Model(SDPA.Optimizer)
+set_attribute(model, &quot;Mode&quot;, SDPA.PARAMETER_STABLE_BUT_SLOW)
+set_attribute(model, &quot;MaxIteration&quot;, 100)</code></pre><p>The choice of parameter mode has a large impact on the performance and stability of SDPA, and not necessarily in the way implied by the names of the modes; for example, <code>PARAMETER_UNSTABLE_BUT_FAST</code> can be more stable than the other modes for some problems. You should try each mode to see how it performs on your specific problem. See <a href="https://github.com/jump-dev/SDPA.jl/issues/17">SDPA.jl#17</a> for more details.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../SCS/">« jump-dev/SCS.jl</a><a class="docs-footer-nextpage" href="../SDPNAL/">jump-dev/SDPNAL.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SDPNAL/index.html b/previews/PR3525/packages/SDPNAL/index.html
new file mode 100644
index 00000000000..6870a52fa65
--- /dev/null
+++ b/previews/PR3525/packages/SDPNAL/index.html
@@ -0,0 +1,21 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/SDPNAL.jl · JuMP</title><meta name="title" content="jump-dev/SDPNAL.jl · JuMP"/><meta property="og:title" content="jump-dev/SDPNAL.jl · JuMP"/><meta property="twitter:title" content="jump-dev/SDPNAL.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/SDPNAL.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li></ul></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/SDPNAL.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/SDPNAL.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/SDPNAL.jl/blob/00a3fa19f4e1235587948113b0b681da17f4dab5/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="SDPNAL.jl"><a class="docs-heading-anchor" href="#SDPNAL.jl">SDPNAL.jl</a><a id="SDPNAL.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SDPNAL.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/SDPNAL.jl">SDPNAL.jl</a> is wrapper for the <a href="https://blog.nus.edu.sg/mattohkc/softwares/sdpnalplus">SDPNALplus</a> solver.</p><p>The wrapper has two components:</p><ul><li>an exported <code>sdpnalplus</code> function that is a thin wrapper on top of the <code>sdpnalplus</code> MATLAB function</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not an official wrapper of SDPNALplus.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SDPNAL.jl</code> is licensed under the <a href="https://github.com/jump-dev/SDPNAL.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://blog.nus.edu.sg/mattohkc/softwares/sdpnalplus/">SDPNALplus</a> is licensed under the <a href="https://creativecommons.org/licenses/by-sa/4.0/">Creative Commons Attribution-ShareAlike 4.0 International Public License</a>.</p><p>In addition, SDPNAL requires an installation of MATLAB, which is a closed-source commercial product for which you must <a href="https://www.mathworks.com/products/matlab.html">obtain a license</a>.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use SDPNAL with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>, do:</p><pre><code class="language-julia hljs">using JuMP, SDPNAL
+model = Model(SDPNAL.Optimizer)
+set_attribute(model, &quot;printlevel&quot;, 0)</code></pre><p>Note that, contrary to implementation of other solver-independent interfaces, using SDPNAL from JuMP or MOI fully exploits the particular structures of the SDPNAL interface and does not create superfluous slack variables and equality constraints as discussed in <a href="https://arxiv.org/pdf/1710.10604.pdf">the SDPNAL guide</a>:</p><blockquote><p>A new interface is necessary to facilitate the modeling of an SDP problem for SDPNAL+ because of latter’s flexibility to directly accept inequality constraints of the form “l ≤ B(X) ≤ u”, and bound constraints of the form “L ≤ X ≤ U”. The flexibility can significantly simplify the generation of the data in the SDPNAL+ format as compared to what need to be done in CVX or YALMIP to reformulate them as equality constraints through introducing extra variables. In addition, the final number of equality constraints present in the data input to SDPNAL+ can also be substantially fewer than those present in CVX or YALMIP. It is important to note here that the number of equality constraints present in the generated problem data can greatly affect the computational efficiency of the solvers, especially for interior-point based solvers.</p></blockquote><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, make sure that you satisfy the requirements of the <a href="https://github.com/JuliaInterop/MATLAB.jl">MATLAB.jl</a> Julia package, and that the SDPNALplus software is installed in your <a href="http://www.mathworks.com/products/matlab/">MATLAB™</a> installation.</p><p>Then, install <code>SDPNAL.jl</code> using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;SDPNAL&quot;)</code></pre><p>There is a <code>startup.m</code> file at the root of the SDPNAL folder. This adds all subdirectories recursively when MATLAB starts. However, the <code>interface</code> directory contains a <code>.git</code> subdirectory which contains a very large number of files. Because of this, MATLAB crashes if SDPNAL is in its path because the <code>startup.m</code> requests MATLAB to try to parse all the files in the <code>.git</code> folder. To resolve this problem, delete the <code>startup.m</code> file and <code>.git</code> folder, and add the subdirectories manually your <code>toolbox/local/pathdef.m</code> file as follows:</p><pre><code class="nohighlight hljs">function p = pathdef
+
+% (...)
+
+p = [...
+%%% BEGIN ENTRIES %%%
+&#39;/path/to/SDPNALv1.0:&#39;, ...
+&#39;/path/to/SDPNALv1.0/interface:&#39;, ...
+&#39;/path/to/SDPNALv1.0/mexfun:&#39;, ...
+&#39;/path/to/SDPNALv1.0/solver:&#39;, ...
+&#39;/path/to/SDPNALv1.0/solver_main_default:&#39;, ...
+&#39;/path/to/SDPNALv1.0/util:&#39;, ...
+% (...)</code></pre><p>If you have <a href="https://github.com/jump-dev/SDPT3.jl">SDPT3</a> in addition to SDPNAL in the MATLAB path (that is, the <code>toolbox/local/pathdef.m</code> file) then you might have issues because both solvers define a <code>validate</code> function, and this might make SDPNAL call SDPT3&#39;s <code>validate</code> function instead of SDPT3&#39;s <code>validate</code> function.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../SDPA/">« jump-dev/SDPA.jl</a><a class="docs-footer-nextpage" href="../SDPT3/">jump-dev/SDPT3.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SDPT3/index.html b/previews/PR3525/packages/SDPT3/index.html
new file mode 100644
index 00000000000..2845d663bfb
--- /dev/null
+++ b/previews/PR3525/packages/SDPT3/index.html
@@ -0,0 +1,32 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/SDPT3.jl · JuMP</title><meta name="title" content="jump-dev/SDPT3.jl · JuMP"/><meta property="og:title" content="jump-dev/SDPT3.jl · JuMP"/><meta property="twitter:title" content="jump-dev/SDPT3.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/SDPT3.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li></ul></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/SDPT3.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/SDPT3.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/SDPT3.jl/blob/b565aac2a58818090d521f2340e71f597688e4fb/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="SDPT3.jl"><a class="docs-heading-anchor" href="#SDPT3.jl">SDPT3.jl</a><a id="SDPT3.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SDPT3.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/SDPT3.jl">SDPT3.jl</a> is wrapper for the <a href="https://blog.nus.edu.sg/mattohkc/softwares/sdpt3">SDPT3</a> solver.</p><p>The wrapper has two components:</p><ul><li>an exported <code>sdpt3</code> function that is a thin wrapper on top of the <code>sdpt3</code> MATLAB function</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not an official wrapper of SDPT3.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SDPT3.jl</code> is licensed under the <a href="https://github.com/jump-dev/SDPT3.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://blog.nus.edu.sg/mattohkc/softwares/sdpt3/">SDPT3</a> is licensed under the <a href="https://www.gnu.org/licenses/old-licenses/gpl-2.0.en.html">GPL v2 License</a>.</p><p>In addition, SDPT3 requires an installation of MATLAB, which is a closed-source commercial product for which you must <a href="https://www.mathworks.com/products/matlab.html">obtain a license</a>.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use SDPT3 with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>, do:</p><pre><code class="language-julia hljs">using JuMP, SDPT3
+model = Model(SDPT3.Optimizer)
+set_attribute(model, &quot;printlevel&quot;, 0)</code></pre><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, make sure that you satisfy the requirements of the <a href="https://github.com/JuliaInterop/MATLAB.jl">MATLAB.jl</a> Julia package, and that the SeDuMi software is installed in your <a href="http://www.mathworks.com/products/matlab/">MATLAB™</a> installation.</p><p>Then, install <code>SDPT3.jl</code> using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;SDPT3&quot;)</code></pre><h3 id="SDPT3-not-in-PATH"><a class="docs-heading-anchor" href="#SDPT3-not-in-PATH">SDPT3 not in PATH</a><a id="SDPT3-not-in-PATH-1"></a><a class="docs-heading-anchor-permalink" href="#SDPT3-not-in-PATH" title="Permalink"></a></h3><p>If you get the error:</p><pre><code class="nohighlight hljs">Error using save
+Variable &#39;jx_sdpt3_arg_out_1&#39; not found.
+
+ERROR: LoadError: MATLAB.MEngineError(&quot;failed to get variable jx_sdpt3_arg_out_1 from MATLAB session&quot;)
+Stacktrace:
+[...]</code></pre><p>The error means that we could not find the <code>sdpt3</code> function with one output argument using the MATLAB C API. This most likely means that you did not add SDPT3 to the MATLAB&#39;s path (that is, the <code>toolbox/local/pathdef.m</code> file).</p><p>If modifying <code>toolbox/local/pathdef.m</code> does not work, the following should work, where <code>/path/to/sdpt3/</code> is the directory where the <code>sdpt3</code> folder is located:</p><pre><code class="language-julia hljs">julia&gt; using MATLAB
+
+julia&gt; cd(&quot;/path/to/sdpt3/&quot;) do
+           MATLAB.mat&quot;install_sdpt3&quot;
+       end
+
+julia&gt; MATLAB.mat&quot;savepath&quot;</code></pre><p>An alternative fix is suggested in the <a href="https://github.com/jump-dev/SDPT3.jl/issues/9#issuecomment-855509257">following issue</a>.</p><h3 id="Error-in-validate"><a class="docs-heading-anchor" href="#Error-in-validate">Error in validate</a><a id="Error-in-validate-1"></a><a class="docs-heading-anchor-permalink" href="#Error-in-validate" title="Permalink"></a></h3><p>If you get the error:</p><pre><code class="nohighlight hljs">Brace indexing is not supported for variables of this type.
+
+Error in validate
+
+Error in sdpt3 (line 171)
+   [blk,At,C,b,blkdim,numblk,parbarrier] = validate(blk,At,C,b,par,parbarrier);
+
+Error using save
+Variable &#39;jx_sdpt3_arg_out_1&#39; not found.</code></pre><p>It might mean that you have added <a href="https://github.com/jump-dev/SDPNAL.jl">SDPNAL</a> in addition to SDPT3 in the MATLAB&#39;s path (that is, the <code>toolbox/local/pathdef.m</code> file). Because SDPNAL also defines a <code>validate</code> function, this can make <code>sdpt3</code> call SDPNAL&#39;s <code>validate</code> function instead of SDPT3&#39;s <code>validate</code> function, which causes the issue.</p><p>One way to fix this from the Julia REPL is to reset the search path to the factory-installed state using <code>restoredefaultpath</code>:</p><pre><code class="language-julia hljs">julia&gt; using MATLAB
+
+julia&gt; MATLAB.restoredefaultpath()
+
+julia&gt; MATLAB.mat&quot;savepath&quot;</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../SDPNAL/">« jump-dev/SDPNAL.jl</a><a class="docs-footer-nextpage" href="../SeDuMi/">jump-dev/SeDuMi.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SeDuMi/index.html b/previews/PR3525/packages/SeDuMi/index.html
new file mode 100644
index 00000000000..ea2f88d99a9
--- /dev/null
+++ b/previews/PR3525/packages/SeDuMi/index.html
@@ -0,0 +1,20 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/SeDuMi.jl · JuMP</title><meta name="title" content="jump-dev/SeDuMi.jl · JuMP"/><meta property="og:title" content="jump-dev/SeDuMi.jl · JuMP"/><meta property="twitter:title" content="jump-dev/SeDuMi.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/SeDuMi.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li></ul></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/SeDuMi.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/SeDuMi.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/SeDuMi.jl/blob/v0.4.2/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="SeDuMi.jl"><a class="docs-heading-anchor" href="#SeDuMi.jl">SeDuMi.jl</a><a id="SeDuMi.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SeDuMi.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/SeDuMi.jl">SeDuMi.jl</a> is wrapper for the <a href="http://sedumi.ie.lehigh.edu/">SeDuMi</a> solver.</p><p>The wrapper has two components:</p><ul><li>an exported <code>sedumi</code> function that is a thin wrapper on top of the <code>sedumi</code> MATLAB function</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>This wrapper is maintained by the JuMP community and is not an official wrapper of SeDuMi.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SeDuMi.jl</code> is licensed under the <a href="https://github.com/jump-dev/SeDuMi.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver, <a href="https://github.com/sqlp/sedumi">sqlp/sedumi</a> is licensed under the <a href="https://github.com/sqlp/sedumi/blob/master/COPYING">GPL v2 license</a>.</p><p>In addition, SeDuMi requires an installation of MATLAB, which is a closed-source commercial product for which you must <a href="https://www.mathworks.com/products/matlab.html">obtain a license</a>.</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use SeDuMi with <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>, do:</p><pre><code class="language-julia hljs">using JuMP, SeDuMi
+model = Model(SeDuMi.Optimizer)
+set_attribute(model, &quot;fid&quot;, 0)</code></pre><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, make sure that you satisfy the requirements of the <a href="https://github.com/JuliaInterop/MATLAB.jl">MATLAB.jl</a> Julia package, and that the SeDuMi software is installed in your <a href="http://www.mathworks.com/products/matlab/">MATLAB™</a> installation.</p><p>Then, install <code>SeDuMi.jl</code> using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;SeDuMi&quot;)</code></pre><p>If you get the error:</p><pre><code class="language-raw hljs">Undefined function or variable &#39;sedumi&#39;.
+
+Error using save
+Variable &#39;jx_sedumi_arg_out_1&#39; not found.
+
+ERROR: LoadError: MATLAB.MEngineError(&quot;failed to get variable jx_sedumi_arg_out_1 from MATLAB session&quot;)</code></pre><p>The error means that we couldn&#39;t find the <code>sedumi</code> function with one output argument using the MATLAB C API.</p><p>This most likely means that you did not add SeDuMi to the MATLAB&#39;s path, that is, the <code>toolbox/local/pathdef.m</code> file.</p><p>If modifying <code>toolbox/local/pathdef.m</code> does not work, the following should work, where <code>/path/to/sedumi/</code> is the directory where the <code>sedumi</code> folder is located:</p><pre><code class="language-julia hljs">julia&gt; import MATLAB
+
+julia&gt; cd(&quot;/path/to/sedumi/&quot;) do
+           MATLAB.mat&quot;install_sedumi&quot;
+       end
+
+julia&gt; MATLAB.mat&quot;savepath&quot;</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../SDPT3/">« jump-dev/SDPT3.jl</a><a class="docs-footer-nextpage" href="../Xpress/">jump-dev/Xpress.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/SumOfSquares/index.html b/previews/PR3525/packages/SumOfSquares/index.html
new file mode 100644
index 00000000000..1f31d5d0375
--- /dev/null
+++ b/previews/PR3525/packages/SumOfSquares/index.html
@@ -0,0 +1,7 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/SumOfSquares.jl · JuMP</title><meta name="title" content="jump-dev/SumOfSquares.jl · JuMP"/><meta property="og:title" content="jump-dev/SumOfSquares.jl · JuMP"/><meta property="twitter:title" content="jump-dev/SumOfSquares.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox" checked/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/SumOfSquares.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li><li><a class="tocitem" href="#Presentations"><span>Presentations</span></a></li><li><a class="tocitem" href="#Citing"><span>Citing</span></a></li></ul></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Extensions</a></li><li class="is-active"><a href>jump-dev/SumOfSquares.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/SumOfSquares.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/SumOfSquares.jl/blob/v0.7.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="SumOfSquares.jl"><a class="docs-heading-anchor" href="#SumOfSquares.jl">SumOfSquares.jl</a><a id="SumOfSquares.jl-1"></a><a class="docs-heading-anchor-permalink" href="#SumOfSquares.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/SumOfSquares.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/SumOfSquares.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/SumOfSquares.jl"><img src="https://codecov.io/gh/jump-dev/SumOfSquares.jl/branch/master/graph/badge.svg" alt="codecov"/></a> <a href="https://doi.org/10.5281/zenodo.1208672"><img src="https://zenodo.org/badge/DOI/10.5281/zenodo.1208672.svg" alt/></a></p><p><a href="https://github.com/jump-dev/SumOfSquares.jl">SumOfSquares.jl</a> is a JuMP extension that, when used in conjunction with <a href="https://github.com/JuliaAlgebra/MultivariatePolynomials.jl">MultivariatePolynomial</a> and <a href="https://github.com/jump-dev/PolyJuMP.jl">PolyJuMP</a>, implements a sum of squares reformulation for polynomial optimization.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>SumOfSquares.jl</code> is licensed under the <a href="https://github.com/jump-dev/SumOfSquares.jl/blob/master/LICENSE.md">MIT license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install <code>SumOfSquares</code> using <code>Pkg.add</code>:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;SumOfSquares&quot;)</code></pre><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>See <a href="https://jump.dev/SumOfSquares.jl/stable">https://jump.dev/SumOfSquares.jl/stable</a> for the most recently tagged version of the documentation.</p><p>See <a href="https://jump.dev/SumOfSquares.jl/dev">https://jump.dev/SumOfSquares.jl/dev</a> for the in-development version of the documentation.</p><h2 id="Presentations"><a class="docs-heading-anchor" href="#Presentations">Presentations</a><a id="Presentations-1"></a><a class="docs-heading-anchor-permalink" href="#Presentations" title="Permalink"></a></h2><p>Some presentations on, or using, SumOfSquares (see <a href="https://github.com/blegat/SumOfSquaresSlides">blegat/SumOfSquaresSlides</a> for the source code of the presentations):</p><ul><li>Benoît Legat, Marek Kaluba and Tillmann Weisser at INFORMS 2022 [<a href="https://drive.google.com/file/d/1rlsIxgcnKWT436k4MNenjHfgH0UYRLAB/view?usp=share_link">Slides</a>]</li><li>Benoît Legat at <a href="http://poema-network.eu/index.php/news-and-events/project-workshops/13-poema-learning-week-2">POEMA Learning Week 2</a></li><li>Benoît Legat, Marek Kaluba and Tillmann Weisser at <a href="https://pretalx.com/juliacon2021/talk/L8DTE3/">JuMP-dev 2021</a> [<a href="https://drive.google.com/file/d/1HtArDFNMQ6IYUqRjSWR3JviJp9xLtSlB/view?usp=sharing">Slides</a>]</li><li>Benoît Legat at <a href="http://meetings2.informs.org/wordpress/annual2020/">INFORMS 2020</a> [<a href="https://drive.google.com/file/d/1lb8NtOWCikTYm6KRUZCSLYgaUjqIsSyV/view?usp=sharing">Slides</a>]</li><li>Tillmann Weisser, Benoît Legat, Chris Coey, Lea Kapelevich and Juan Pablo Vielma at <a href="https://juliacon.org/2019/">JuliaCon 2019</a> [<a href="https://drive.google.com/open?id=1HiA-praFyejE0Z3nVSpFEv938TAcPjA9">Slides</a>] [<a href="https://www.youtube.com/watch?v=cTmqmPcroFo">Video</a>]</li><li>Benoît Legat at <a href="https://cnls.lanl.gov/External/showtalksummary.php?selection=7768">CNLS 2019</a> [<a href="https://drive.google.com/open?id=1kNF18C7RY2zi7jcZBMO1PRXtHuvVTFPn">Slides</a>]</li><li>Benoît Legat at <a href="https://www.euro2019dublin.com/">EURO 2019</a> [<a href="https://drive.google.com/open?id=1Wry56NzzL4QBRSwuhP4AlKOe2i2FL7dk">Slides</a>]</li><li>Benoît Legat at <a href="https://julialang.univ-nantes.fr/programme/">juliaday Nantes 2019</a> [<a href="https://drive.google.com/open?id=1pN3G9Pr8jbzK9EEaJ9a6p_qKwSbxb2bo">Slides</a>]</li><li>Benoît Legat at <a href="https://www.mis.mpg.de/calendar/conferences/2018/nc2018.html">Summer School on Numerical Computing in Algebraic Geometry 2018</a> [<a href="https://drive.google.com/open?id=1pf9rdoVEjAnD164rptLki1AG0AH4i88M">Poster</a>]</li><li>Benoît Legat at <a href="https://jump.dev/meetings/mit2017/">The First Annual JuMP-dev Workshop 2017</a> [<a href="https://drive.google.com/file/d/1ea5eSMvMB3jXPuljzNGmMKied-n50YIo/view?usp=sharing">Slides</a>] [<a href="https://youtu.be/kyo72yWYr54">Video</a>]</li><li><a href="https://docs.google.com/presentation/d/1ASfjB1TdLJmYxT0b6rnyGh9eLbMc-66bTOt3_3yvc90/edit?usp=sharing">Joey Huchette at SIAM Opt 2017</a></li></ul><h2 id="Citing"><a class="docs-heading-anchor" href="#Citing">Citing</a><a id="Citing-1"></a><a class="docs-heading-anchor-permalink" href="#Citing" title="Permalink"></a></h2><p>See <a href="https://github.com/jump-dev/SumOfSquares.jl/blob/master/CITATION.bib">CITATION.bib</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../PolyJuMP/">« jump-dev/PolyJuMP.jl</a><a class="docs-footer-nextpage" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Tulip/index.html b/previews/PR3525/packages/Tulip/index.html
new file mode 100644
index 00000000000..a222b1c5441
--- /dev/null
+++ b/previews/PR3525/packages/Tulip/index.html
@@ -0,0 +1,31 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>ds4dm/Tulip.jl · JuMP</title><meta name="title" content="ds4dm/Tulip.jl · JuMP"/><meta property="og:title" content="ds4dm/Tulip.jl · JuMP"/><meta property="twitter:title" content="ds4dm/Tulip.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li class="is-active"><a class="tocitem" href>ds4dm/Tulip.jl</a><ul class="internal"><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Usage"><span>Usage</span></a></li><li><a class="tocitem" href="#Solver-parameters"><span>Solver parameters</span></a></li><li><a class="tocitem" href="#Command-line-executable"><span>Command-line executable</span></a></li><li><a class="tocitem" href="#Citing-Tulip.jl"><span>Citing <code>Tulip.jl</code></span></a></li></ul></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>ds4dm/Tulip.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>ds4dm/Tulip.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/ds4dm/Tulip.jl/blob/069cdda082dfde31ef7ee935c96e833a3b96b33c/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Tulip"><a class="docs-heading-anchor" href="#Tulip">Tulip</a><a id="Tulip-1"></a><a class="docs-heading-anchor-permalink" href="#Tulip" title="Permalink"></a></h1><p><a href="https://zenodo.org/badge/latestdoi/131298750"><img src="https://zenodo.org/badge/131298750.svg" alt="DOI"/></a> <a href="https://github.com/ds4dm/Tulip.jl/actions?query=workflow%3ACI"><img src="https://github.com/ds4dm/Tulip.jl/workflows/CI/badge.svg?branch=master" alt/></a> <a href="https://codecov.io/github/ds4dm/Tulip.jl?branch=master"><img src="https://codecov.io/github/ds4dm/Tulip.jl/coverage.svg?branch=master" alt/></a></p><p><a href="https://github.com/ds4dm/Tulip.jl">Tulip</a> is an open-source interior-point solver for linear optimization, written in pure Julia. It implements the homogeneous primal-dual interior-point algorithm with multiple centrality corrections, and therefore handles unbounded and infeasible problems. Tulip’s main feature is that its algorithmic framework is disentangled from linear algebra implementations. This allows to seamlessly integrate specialized routines for structured problems.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p>Tulip is licensed under the <a href="https://github.com/ds4dm/Tulip.jl/blob/master/LICENSE.md">MPL 2.0 license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>Install Tulip using the Julia package manager:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Tulip&quot;)</code></pre><h2 id="Usage"><a class="docs-heading-anchor" href="#Usage">Usage</a><a id="Usage-1"></a><a class="docs-heading-anchor-permalink" href="#Usage" title="Permalink"></a></h2><p>The recommended way of using Tulip is through <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a> or <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a> (MOI).</p><p>The low-level interface is still under development and is likely change in the future. The MOI interface is more stable.</p><h3 id="Using-with-JuMP"><a class="docs-heading-anchor" href="#Using-with-JuMP">Using with JuMP</a><a id="Using-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Using-with-JuMP" title="Permalink"></a></h3><p>Tulip follows the syntax convention <code>PackageName.Optimizer</code>:</p><pre><code class="language-julia hljs">using JuMP
+import Tulip
+model = Model(Tulip.Optimizer)</code></pre><p>Linear objectives, linear constraints and lower/upper bounds on variables are supported.</p><h3 id="Using-with-MOI"><a class="docs-heading-anchor" href="#Using-with-MOI">Using with MOI</a><a id="Using-with-MOI-1"></a><a class="docs-heading-anchor-permalink" href="#Using-with-MOI" title="Permalink"></a></h3><p>The type <code>Tulip.Optimizer</code> is parametrized by the model&#39;s arithmetic, for example, <code>Float64</code> or <code>BigFloat</code>. This allows to solve problem in higher numerical precision. See the documentation for more details.</p><pre><code class="language-julia hljs">import MathOptInterface as MOI
+import Tulip
+model = Tulip.Optimizer{Float64}()   # Create a model in Float64 precision
+model = Tulip.Optimizer()            # Defaults to the above call
+model = Tulip.Optimizer{BigFloat}()  # Create a model in BigFloat precision</code></pre><h2 id="Solver-parameters"><a class="docs-heading-anchor" href="#Solver-parameters">Solver parameters</a><a id="Solver-parameters-1"></a><a class="docs-heading-anchor-permalink" href="#Solver-parameters" title="Permalink"></a></h2><p>See the <a href="https://ds4dm.github.io/Tulip.jl/stable/reference/options/">documentation</a> for a full list of parameters.</p><p>To set parameters in JuMP, use:</p><pre><code class="language-julia hljs">using JuMP, Tulip
+model = Model(Tulip.Optimizer)
+set_attribute(model, &quot;IPM_IterationsLimit&quot;, 200)</code></pre><p>To set parameters in MathOptInterface, use:</p><pre><code class="language-julia hljs">using Tulip
+import MathOptInterface as MOI
+model = Tulip.Optimizer{Float64}()
+MOI.set(model, MOI.RawOptimizerAttribute(&quot;IPM_IterationsLimit&quot;), 200)</code></pre><p>To set parameters in the Tulip API, use:</p><pre><code class="language-julia hljs">using Tulip
+model = Tulip.Model{Float64}()
+Tulip.set_parameter(model, &quot;IPM_IterationsLimit&quot;, 200)</code></pre><h2 id="Command-line-executable"><a class="docs-heading-anchor" href="#Command-line-executable">Command-line executable</a><a id="Command-line-executable-1"></a><a class="docs-heading-anchor-permalink" href="#Command-line-executable" title="Permalink"></a></h2><p>See <a href="https://github.com/ds4dm/Tulip.jl/blob/master/app/README.md">app building instructions</a>.</p><h2 id="Citing-Tulip.jl"><a class="docs-heading-anchor" href="#Citing-Tulip.jl">Citing <code>Tulip.jl</code></a><a id="Citing-Tulip.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Citing-Tulip.jl" title="Permalink"></a></h2><p>If you use Tulip in your work, we kindly ask that you cite the following <a href="https://doi.org/10.1007/s12532-020-00200-8">reference</a> (preprint available <a href="https://arxiv.org/abs/2006.08814">here</a>).</p><pre><code class="nohighlight hljs">@Article{Tulip.jl,
+  author   = {Tanneau, Mathieu and Anjos, Miguel F. and Lodi, Andrea},
+  journal  = {Mathematical Programming Computation},
+  title    = {Design and implementation of a modular interior-point solver for linear optimization},
+  year     = {2021},
+  issn     = {1867-2957},
+  month    = feb,
+  doi      = {10.1007/s12532-020-00200-8},
+  language = {en},
+  url      = {https://doi.org/10.1007/s12532-020-00200-8},
+  urldate  = {2021-03-07},
+}</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../DAQP/">« darnstrom/DAQP.jl</a><a class="docs-footer-nextpage" href="../Loraine/">kocvara/Loraine.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/Xpress/index.html b/previews/PR3525/packages/Xpress/index.html
new file mode 100644
index 00000000000..8b0d9caacb2
--- /dev/null
+++ b/previews/PR3525/packages/Xpress/index.html
@@ -0,0 +1,11 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>jump-dev/Xpress.jl · JuMP</title><meta name="title" content="jump-dev/Xpress.jl · JuMP"/><meta property="og:title" content="jump-dev/Xpress.jl · JuMP"/><meta property="twitter:title" content="jump-dev/Xpress.jl · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../solvers/">Introduction</a></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li class="is-active"><a class="tocitem" href>jump-dev/Xpress.jl</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#License"><span>License</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#Use-with-JuMP"><span>Use with JuMP</span></a></li><li><a class="tocitem" href="#Options"><span>Options</span></a></li><li><a class="tocitem" href="#Callbacks"><span>Callbacks</span></a></li><li><a class="tocitem" href="#Environment-variables"><span>Environment variables</span></a></li><li><a class="tocitem" href="#Skipping-Xpress.postsolve"><span>Skipping Xpress.postsolve</span></a></li><li><a class="tocitem" href="#C-API"><span>C API</span></a></li><li><a class="tocitem" href="#Documentation"><span>Documentation</span></a></li></ul></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>jump-dev/Xpress.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>jump-dev/Xpress.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/Xpress.jl/blob/v0.16.1/README.md" title="View source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Xpress.jl"><a class="docs-heading-anchor" href="#Xpress.jl">Xpress.jl</a><a id="Xpress.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Xpress.jl" title="Permalink"></a></h1><p><a href="https://github.com/jump-dev/Xpress.jl/actions?query=workflow%3ACI"><img src="https://github.com/jump-dev/Xpress.jl/workflows/CI/badge.svg?branch=master" alt="Build Status"/></a> <a href="https://codecov.io/gh/jump-dev/Xpress.jl"><img src="https://codecov.io/gh/jump-dev/Xpress.jl/branch/master/graph/badge.svg" alt="codecov"/></a></p><p><a href="https://github.com/jump-dev/Xpress.jl">Xpress.jl</a> is a wrapper for the <a href="https://www.fico.com/products/fico-xpress-solver">FICO Xpress Solver</a>.</p><p>It has two components:</p><ul><li>a thin wrapper around the complete C API</li><li>an interface to <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a></li></ul><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>The Xpress wrapper for Julia is community driven and not officially supported by FICO Xpress. If you are a commercial customer interested in official support for Julia from FICO Xpress, let them know.</p><h2 id="License"><a class="docs-heading-anchor" href="#License">License</a><a id="License-1"></a><a class="docs-heading-anchor-permalink" href="#License" title="Permalink"></a></h2><p><code>Xpress.jl</code> is licensed under the <a href="https://github.com/jump-dev/Xpress.jl/blob/master/LICENSE.md">MIT License</a>.</p><p>The underlying solver is a closed-source commercial product for which you must <a href="https://www.fico.com/products/fico-xpress-solver">purchase a license</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>First, obtain a license of Xpress and install Xpress solver, following the <a href="https://www.fico.com/products/fico-xpress-solver">instructions on the FICO website</a>.</p><p>Then, install this package using:</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;Xpress&quot;)</code></pre><p>If you encounter an error, make sure that the <code>XPRESSDIR</code> environmental variable is set to the path of the Xpress directory. This should be part of a standard installation. The Xpress library will be searched for in <code>XPRESSDIR/lib</code> on Unix platforms and <code>XPRESSDIR/bin</code> on Windows.</p><p>For example, on macOS, you may need:</p><pre><code class="language-julia hljs">ENV[&quot;XPRESSDIR&quot;] = &quot;/Applications/FICO Xpress/xpressmp/&quot;
+import Pkg
+Pkg.add(&quot;Xpress&quot;)</code></pre><p>By default, building Xpress.jl will fail if the Xpress library is not found. This may not be desirable in certain cases, for example when part of a package&#39;s test suite uses Xpress as an optional test dependency, but Xpress cannot be installed on a CI server running the test suite. To support this use case, the <code>XPRESS_JL_SKIP_LIB_CHECK</code> environment variable may be set (to any value) to make Xpress.jl installable (but not usable).</p><h2 id="Use-with-JuMP"><a class="docs-heading-anchor" href="#Use-with-JuMP">Use with JuMP</a><a id="Use-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Use-with-JuMP" title="Permalink"></a></h2><p>To use Xpress with JuMP, use:</p><pre><code class="language-julia hljs">using JuMP, Xpress
+model = Model(Xpress.Optimizer)
+set_optimizer(model, &quot;PRESOLVE&quot;, 0)</code></pre><h2 id="Options"><a class="docs-heading-anchor" href="#Options">Options</a><a id="Options-1"></a><a class="docs-heading-anchor-permalink" href="#Options" title="Permalink"></a></h2><p>For other parameters use <a href="https://www.fico.com/fico-xpress-optimization/docs/latest/solver/optimizer/HTML/">Xpress Optimizer manual</a> or type <code>julia -e &quot;using Xpress; println(keys(Xpress.XPRS_ATTRIBUTES))&quot;</code>.</p><p>If <code>logfile</code> is set to <code>&quot;&quot;</code>, the log file is disabled and output is printed to the console (<a href="https://www.fico.com/fico-xpress-optimization/docs/latest/solver/optimizer/HTML/OUTPUTLOG.html">there might be issues with console output on windows (it is manually implemented with callbacks)</a>). If <code>logfile</code> is set to a file&#39;s path, output is printed to that file. By default, <code>logfile = &quot;&quot;</code> (console).</p><h2 id="Callbacks"><a class="docs-heading-anchor" href="#Callbacks">Callbacks</a><a id="Callbacks-1"></a><a class="docs-heading-anchor-permalink" href="#Callbacks" title="Permalink"></a></h2><p>Solver specific and solver independent callbacks are working in <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a> and, consequently, in <a href="https://github.com/jump-dev/JuMP.jl">JuMP</a>. However, the current implementation should be considered experimental.</p><h2 id="Environment-variables"><a class="docs-heading-anchor" href="#Environment-variables">Environment variables</a><a id="Environment-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Environment-variables" title="Permalink"></a></h2><ul><li><code>XPRESS_JL_SKIP_LIB_CHECK</code>: Used to skip build lib check as previously described.</li><li><code>XPRESS_JL_NO_INFO</code>: Disable license info log.</li><li><code>XPRESS_JL_NO_DEPS_ERROR</code>: Disable error when do deps.jl file is found.</li><li><code>XPRESS_JL_NO_AUTO_INIT</code>: Disable automatic run of <code>Xpress.initialize()</code>.</li></ul><p>Specially useful for explicitly loading the dynamic library.</p><h2 id="Skipping-Xpress.postsolve"><a class="docs-heading-anchor" href="#Skipping-Xpress.postsolve">Skipping Xpress.postsolve</a><a id="Skipping-Xpress.postsolve-1"></a><a class="docs-heading-anchor-permalink" href="#Skipping-Xpress.postsolve" title="Permalink"></a></h2><p>In older versions of Xpress, the command <code>XPRSpostsolve</code> throws an error in infeasible models. In these older versions the post solve should not be executed. To do this, one can use the <code>MOI.RawOptimizerAttribute(&quot;MOI_POST_SOLVE&quot;)</code> to skip this routine.</p><h2 id="C-API"><a class="docs-heading-anchor" href="#C-API">C API</a><a id="C-API-1"></a><a class="docs-heading-anchor-permalink" href="#C-API" title="Permalink"></a></h2><p>The C API can be accessed via <code>Xpress.Lib.XPRSxx</code> functions, where the names and arguments are identical to the C API. </p><p>See the <a href="https://www.fico.com/fico-xpress-optimization/docs/latest/solver/optimizer/HTML">Xpress documentation</a> for details.</p><h2 id="Documentation"><a class="docs-heading-anchor" href="#Documentation">Documentation</a><a id="Documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Documentation" title="Permalink"></a></h2><p>For more information, consult the <a href="https://www.fico.com/fico-xpress-optimization/docs/latest/solver/optimizer/HTML">FICO optimizer manual</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../SeDuMi/">« jump-dev/SeDuMi.jl</a><a class="docs-footer-nextpage" href="../COPT/">COPT-Public/COPT.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/packages/solvers/index.html b/previews/PR3525/packages/solvers/index.html
new file mode 100644
index 00000000000..313d0ab954d
--- /dev/null
+++ b/previews/PR3525/packages/solvers/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../assets/documenter.js"></script><script src="../../search_index.js"></script><script src="../../siteinfo.js"></script><script src="../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../assets/themeswap.js"></script><link href="../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../"><img class="docs-light-only" src="../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../">Introduction</a></li><li><a class="tocitem" href="../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../manual/models/">Models</a></li><li><a class="tocitem" href="../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox" checked/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#Affiliation"><span>Affiliation</span></a></li><li><a class="tocitem" href="#Adding-new-solvers"><span>Adding new solvers</span></a></li></ul></li><li><a class="tocitem" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/">Introduction</a></li><li><a class="tocitem" href="../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Solvers</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/packages/solvers.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Introduction"><a class="docs-heading-anchor" href="#Introduction">Introduction</a><a id="Introduction-1"></a><a class="docs-heading-anchor-permalink" href="#Introduction" title="Permalink"></a></h1><p>This section of the documentation contains brief documentation for some of the solvers that JuMP supports. The list of solvers is not exhaustive, but instead is intended to help you discover commonly used solvers.</p><h2 id="Affiliation"><a class="docs-heading-anchor" href="#Affiliation">Affiliation</a><a id="Affiliation-1"></a><a class="docs-heading-anchor-permalink" href="#Affiliation" title="Permalink"></a></h2><p>Packages beginning with <code>jump-dev/</code> are developed and maintained by the JuMP developers. In many cases, these packages wrap external solvers that are not developed by the JuMP developers and, while the Julia packages are all open-source, in some cases the solvers themselves are closed source commercial products.</p><p>Packages that do not begin with <code>jump-dev/</code> are developed independently. The developers of these packages requested or consented to the inclusion of their README contents in the JuMP documentation for the benefit of users.</p><h2 id="Adding-new-solvers"><a class="docs-heading-anchor" href="#Adding-new-solvers">Adding new solvers</a><a id="Adding-new-solvers-1"></a><a class="docs-heading-anchor-permalink" href="#Adding-new-solvers" title="Permalink"></a></h2><p>Written a solver? Add it to this section of the JuMP documentation by making a pull request to the <a href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/packages.toml"><code>docs/packages.toml</code></a> file.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../developers/checklists/">« Checklists</a><a class="docs-footer-nextpage" href="../AmplNLWriter/">jump-dev/AmplNLWriter.jl »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/release_notes/index.html b/previews/PR3525/release_notes/index.html
new file mode 100644
index 00000000000..ba2cb8e68f6
--- /dev/null
+++ b/previews/PR3525/release_notes/index.html
@@ -0,0 +1,15 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Release notes · JuMP</title><meta name="title" content="Release notes · JuMP"/><meta property="og:title" content="Release notes · JuMP"/><meta property="twitter:title" content="Release notes · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL=".."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../assets/documenter.js"></script><script src="../search_index.js"></script><script src="../siteinfo.js"></script><script src="../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../assets/themeswap.js"></script><link href="../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../"><img class="docs-light-only" src="../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../">Introduction</a></li><li><a class="tocitem" href="../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../manual/models/">Models</a></li><li><a class="tocitem" href="../manual/variables/">Variables</a></li><li><a class="tocitem" href="../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../manual/containers/">Containers</a></li><li><a class="tocitem" href="../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/">Introduction</a></li><li><a class="tocitem" href="../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../moi/release_notes/">Release notes</a></li></ul></li><li class="is-active"><a class="tocitem" href>Release notes</a><ul class="internal"><li><a class="tocitem" href="#[Version-1.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.1)-(September-23,-2023)"><span>Version 1.15.1 (September 23, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.0)-(September-15,-2023)"><span>Version 1.15.0 (September 15, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.1)-(September-2,-2023)"><span>Version 1.14.1 (September 2, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.0)-(August-27,-2023)"><span>Version 1.14.0 (August 27, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.13.0)-(July-27,-2023)"><span>Version 1.13.0 (July 27, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.12.0)-(June-19,-2023)"><span>Version 1.12.0 (June 19, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.1)-(May-19,-2023)"><span>Version 1.11.1 (May 19, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.0)-(May-3,-2023)"><span>Version 1.11.0 (May 3, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.10.0)-(April-3,-2023)"><span>Version 1.10.0 (April 3, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.9.0)-(March-7,-2023)"><span>Version 1.9.0 (March 7, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.8.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.2)-(February-27,-2023)"><span>Version 1.8.2 (February 27, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.8.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.1)-(February-23,-2023)"><span>Version 1.8.1 (February 23, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.0)-(February-16,-2023)"><span>Version 1.8.0 (February 16, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.7.0)-(January-25,-2023)"><span>Version 1.7.0 (January 25, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.6.0)-(January-1,-2023)"><span>Version 1.6.0 (January 1, 2023)</span></a></li><li><a class="tocitem" href="#[Version-1.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.5.0)-(December-8,-2022)"><span>Version 1.5.0 (December 8, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.4.0)-(October-29,-2022)"><span>Version 1.4.0 (October 29, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.1)-(September-28,-2022)"><span>Version 1.3.1 (September 28, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.0)-(September-5,-2022)"><span>Version 1.3.0 (September 5, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.2.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.1)-(August-22,-2022)"><span>Version 1.2.1 (August 22, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.0)-(August-16,-2022)"><span>Version 1.2.0 (August 16, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.1)-(June-14,-2022)"><span>Version 1.1.1 (June 14, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.0)-(May-25,-2022)"><span>Version 1.1.0 (May 25, 2022)</span></a></li><li><a class="tocitem" href="#[Version-1.0.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.0.0)-(March-24,-2022)"><span>Version 1.0.0 (March 24, 2022)</span></a></li><li><a class="tocitem" href="#[Version-0.23.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.2)-(March-14,-2022)"><span>Version 0.23.2 (March 14, 2022)</span></a></li><li><a class="tocitem" href="#[Version-0.23.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.1)-(March-2,-2022)"><span>Version 0.23.1 (March 2, 2022)</span></a></li><li><a class="tocitem" href="#[Version-0.23.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.0)-(February-25,-2022)"><span>Version 0.23.0 (February 25, 2022)</span></a></li><li><a class="tocitem" href="#[Version-0.22.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.3)-(February-10,-2022)"><span>Version 0.22.3 (February 10, 2022)</span></a></li><li><a class="tocitem" href="#[Version-0.22.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.2)-(January-10,-2022)"><span>Version 0.22.2 (January 10, 2022)</span></a></li><li><a class="tocitem" href="#[Version-0.22.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.1)-(November-29,-2021)"><span>Version 0.22.1 (November 29, 2021)</span></a></li><li><a class="tocitem" href="#[Version-0.22.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.0)-(November-10,-2021)"><span>Version 0.22.0 (November 10, 2021)</span></a></li><li><a class="tocitem" href="#[Version-0.21.10](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.10)-(September-4,-2021)"><span>Version 0.21.10 (September 4, 2021)</span></a></li><li><a class="tocitem" href="#[Version-0.21.9](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.9)-(August-1,-2021)"><span>Version 0.21.9 (August 1, 2021)</span></a></li><li><a class="tocitem" href="#[Version-0.21.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.8)-(May-8,-2021)"><span>Version 0.21.8 (May 8, 2021)</span></a></li><li><a class="tocitem" href="#[Version-0.21.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.7)-(April-12,-2021)"><span>Version 0.21.7 (April 12, 2021)</span></a></li><li><a class="tocitem" href="#[Version-0.21.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.6)-(January-29,-2021)"><span>Version 0.21.6 (January 29, 2021)</span></a></li><li><a class="tocitem" href="#[Version-0.21.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.5)-(September-18,-2020)"><span>Version 0.21.5 (September 18, 2020)</span></a></li><li><a class="tocitem" href="#[Version-0.21.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.4)-(September-14,-2020)"><span>Version 0.21.4 (September 14, 2020)</span></a></li><li><a class="tocitem" href="#[Version-0.21.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.3)-(June-18,-2020)"><span>Version 0.21.3 (June 18, 2020)</span></a></li><li><a class="tocitem" href="#[Version-0.21.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.2)-(April-2,-2020)"><span>Version 0.21.2 (April 2, 2020)</span></a></li><li><a class="tocitem" href="#[Version-0.21.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.1)-(Feb-18,-2020)"><span>Version 0.21.1 (Feb 18, 2020)</span></a></li><li><a class="tocitem" href="#[Version-0.21.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.0)-(Feb-16,-2020)"><span>Version 0.21.0 (Feb 16, 2020)</span></a></li><li><a class="tocitem" href="#[Version-0.20.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.1)-(Oct-18,-2019)"><span>Version 0.20.1 (Oct 18, 2019)</span></a></li><li><a class="tocitem" href="#[Version-0.20.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.0)-(Aug-24,-2019)"><span>Version 0.20.0 (Aug 24, 2019)</span></a></li><li><a class="tocitem" href="#[Version-0.19.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.2)-(June-8,-2019)"><span>Version 0.19.2 (June 8, 2019)</span></a></li><li><a class="tocitem" href="#[Version-0.19.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.1)-(May-12,-2019)"><span>Version 0.19.1 (May 12, 2019)</span></a></li><li><a class="tocitem" href="#[Version-0.19.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.0)-(February-15,-2019)"><span>Version 0.19.0 (February 15, 2019)</span></a></li><li><a class="tocitem" href="#[Version-0.18.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.5)-(December-1,-2018)"><span>Version 0.18.5 (December 1, 2018)</span></a></li><li><a class="tocitem" href="#[Version-0.18.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.4)-(October-8,-2018)"><span>Version 0.18.4 (October 8, 2018)</span></a></li><li><a class="tocitem" href="#[Version-0.18.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.3)-(October-1,-2018)"><span>Version 0.18.3 (October 1, 2018)</span></a></li><li><a class="tocitem" href="#[Version-0.18.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.2)-(June-10,-2018)"><span>Version 0.18.2 (June 10, 2018)</span></a></li><li><a class="tocitem" href="#[Version-0.18.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.1)-(April-9,-2018)"><span>Version 0.18.1 (April 9, 2018)</span></a></li><li><a class="tocitem" href="#[Version-0.18.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.0)-(July-27,-2017)"><span>Version 0.18.0 (July 27, 2017)</span></a></li><li><a class="tocitem" href="#[Version-0.17.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.1)-(June-9,-2017)"><span>Version 0.17.1 (June 9, 2017)</span></a></li><li><a class="tocitem" href="#[Version-0.17.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.0)-(May-27,-2017)"><span>Version 0.17.0 (May 27, 2017)</span></a></li><li><a class="tocitem" href="#[Version-0.16.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.2)-(March-28,-2017)"><span>Version 0.16.2 (March 28, 2017)</span></a></li><li><a class="tocitem" href="#[Version-0.16.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.1)-(March-7,-2017)"><span>Version 0.16.1 (March 7, 2017)</span></a></li><li><a class="tocitem" href="#[Version-0.16.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.0)-(February-23,-2017)"><span>Version 0.16.0 (February 23, 2017)</span></a></li><li><a class="tocitem" href="#[Version-0.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.1)-(January-31,-2017)"><span>Version 0.15.1 (January 31, 2017)</span></a></li><li><a class="tocitem" href="#[Version-0.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.0)-(December-22,-2016)"><span>Version 0.15.0 (December 22, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.14.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.2)-(December-12,-2016)"><span>Version 0.14.2 (December 12, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.1)-(September-12,-2016)"><span>Version 0.14.1 (September 12, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.0)-(August-7,-2016)"><span>Version 0.14.0 (August 7, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.13.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.2)-(May-16,-2016)"><span>Version 0.13.2 (May 16, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.13.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.1)-(May-3,-2016)"><span>Version 0.13.1 (May 3, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.0)-(April-29,-2016)"><span>Version 0.13.0 (April 29, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.12.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.2)-(March-9,-2016)"><span>Version 0.12.2 (March 9, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.12.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.1)-(March-1,-2016)"><span>Version 0.12.1 (March 1, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.0)-(February-27,-2016)"><span>Version 0.12.0 (February 27, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.11.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.3)-(February-4,-2016)"><span>Version 0.11.3 (February 4, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.11.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.2)-(January-14,-2016)"><span>Version 0.11.2 (January 14, 2016)</span></a></li><li><a class="tocitem" href="#[Version-0.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.1)-(December-1,-2015)"><span>Version 0.11.1 (December 1, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.0)-(November-30,-2015)"><span>Version 0.11.0 (November 30, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.10.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.3)-(November-20,-2015)"><span>Version 0.10.3 (November 20, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.10.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.2)-(September-28,-2015)"><span>Version 0.10.2 (September 28, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.10.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.1)-(September-3,-2015)"><span>Version 0.10.1 (September 3, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.0)-(August-31,-2015)"><span>Version 0.10.0 (August 31, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.9.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.3)-(August-11,-2015)"><span>Version 0.9.3 (August 11, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.9.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.2)-(June-27,-2015)"><span>Version 0.9.2 (June 27, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.9.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.1)-(April-25,-2015)"><span>Version 0.9.1 (April 25, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.0)-(April-18,-2015)"><span>Version 0.9.0 (April 18, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.8.0)-(February-17,-2015)"><span>Version 0.8.0 (February 17, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.7.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.4)-(February-4,-2015)"><span>Version 0.7.4 (February 4, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.7.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.3)-(January-14,-2015)"><span>Version 0.7.3 (January 14, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.7.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.2)-(January-9,-2015)"><span>Version 0.7.2 (January 9, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.7.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.1)-(January-2,-2015)"><span>Version 0.7.1 (January 2, 2015)</span></a></li><li><a class="tocitem" href="#[Version-0.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.0)-(December-29,-2014)"><span>Version 0.7.0 (December 29, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.6.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.3)-(October-19,-2014)"><span>Version 0.6.3 (October 19, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.6.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.2)-(October-11,-2014)"><span>Version 0.6.2 (October 11, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.6.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.1)-(September-19,-2014)"><span>Version 0.6.1 (September 19, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.0)-(September-9,-2014)"><span>Version 0.6.0 (September 9, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.8)-(September-24,-2014)"><span>Version 0.5.8 (September 24, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.7)-(September-5,-2014)"><span>Version 0.5.7 (September 5, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.6)-(September-2,-2014)"><span>Version 0.5.6 (September 2, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.5)-(July-6,-2014)"><span>Version 0.5.5 (July 6, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.4)-(June-19,-2014)"><span>Version 0.5.4 (June 19, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.3)-(May-21,-2014)"><span>Version 0.5.3 (May 21, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.2)-(May-9,-2014)"><span>Version 0.5.2 (May 9, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.1)-(May-5,-2014)"><span>Version 0.5.1 (May 5, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.0)-(May-2,-2014)"><span>Version 0.5.0 (May 2, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.4.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.1)-(March-24,-2014)"><span>Version 0.4.1 (March 24, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.0)-(March-10,-2014)"><span>Version 0.4.0 (March 10, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.3.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.2)-(February-17,-2014)"><span>Version 0.3.2 (February 17, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.1)-(January-30,-2014)"><span>Version 0.3.1 (January 30, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.0)-(January-21,-2014)"><span>Version 0.3.0 (January 21, 2014)</span></a></li><li><a class="tocitem" href="#[Version-0.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.2.0)-(December-15,-2013)"><span>Version 0.2.0 (December 15, 2013)</span></a></li><li><a class="tocitem" href="#[Version-0.1.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.2)-(November-16,-2013)"><span>Version 0.1.2 (November 16, 2013)</span></a></li><li><a class="tocitem" href="#[Version-0.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.1)-(October-23,-2013)"><span>Version 0.1.1 (October 23, 2013)</span></a></li><li><a class="tocitem" href="#[Version-0.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.0)-(October-3,-2013)"><span>Version 0.1.0 (October 3, 2013)</span></a></li></ul></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li class="is-active"><a href>Release notes</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Release notes</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/changelog.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Release-notes"><a class="docs-heading-anchor" href="#Release-notes">Release notes</a><a id="Release-notes-1"></a><a class="docs-heading-anchor-permalink" href="#Release-notes" title="Permalink"></a></h1><p>The format is based on <a href="https://keepachangelog.com/en/1.0.0/">Keep a Changelog</a>, and this project adheres to <a href="https://semver.org/spec/v2.0.0.html">Semantic Versioning</a>.</p><h2 id="[Version-1.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.1)-(September-23,-2023)"><a class="docs-heading-anchor" href="#[Version-1.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.1)-(September-23,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.1">Version 1.15.1</a> (September 23, 2023)</a><a id="[Version-1.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.1)-(September-23,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.1)-(September-23,-2023)" title="Permalink"></a></h2><h3 id="Fixed"><a class="docs-heading-anchor" href="#Fixed">Fixed</a><a id="Fixed-1"></a><a class="docs-heading-anchor-permalink" href="#Fixed" title="Permalink"></a></h3><ul><li>Fixed support for single argument <code>min</code> and <code>max</code> operators (<a href="https://github.com/jump-dev/JuMP.jl/issues/3522">#3522</a>)</li><li>Fixed error message for <a href="../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a> when called with a <a href="../api/JuMP/#JuMP.GenericNonlinearExpr"><code>GenericNonlinearExpr</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3506">#3506</a>)</li><li>Fixed constraint tags with broadcasted constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/3515">#3515</a>)</li><li>Fixed MethodError in <code>MA.scaling</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3518">#3518</a>)</li><li>Fixed support for arrays of <a href="../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a> variables (<a href="https://github.com/jump-dev/JuMP.jl/issues/3524">#3524</a>)</li></ul><h3 id="Other"><a class="docs-heading-anchor" href="#Other">Other</a><a id="Other-1"></a><a class="docs-heading-anchor-permalink" href="#Other" title="Permalink"></a></h3><ul><li>Updated to Documenter@1 (<a href="https://github.com/jump-dev/JuMP.jl/issues/3501">#3501</a>)</li><li>Fixed links to data in tutorials (<a href="https://github.com/jump-dev/JuMP.jl/issues/3512">#3512</a>)</li><li>Fixed typo in TSP tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3516">#3516</a>)</li><li>Improved error message for <a href="../api/JuMP/#JuMP.VariableNotOwned"><code>VariableNotOwned</code></a> errors (<a href="https://github.com/jump-dev/JuMP.jl/issues/3520">#3520</a>)</li><li>Fixed various JET errors (<a href="https://github.com/jump-dev/JuMP.jl/issues/3519">#3519</a>)</li></ul><h2 id="[Version-1.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.0)-(September-15,-2023)"><a class="docs-heading-anchor" href="#[Version-1.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.0)-(September-15,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.0">Version 1.15.0</a> (September 15, 2023)</a><a id="[Version-1.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.0)-(September-15,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.0)-(September-15,-2023)" title="Permalink"></a></h2><p>This is a large minor release because it adds an entirely new data structure and API path for working with nonlinear programs. The previous nonlinear interface remains unchanged and is documented at <a href="../manual/nlp/#Nonlinear-Modeling-(Legacy)">Nonlinear Modeling (Legacy)</a>. The new interface is a treated as a non-breaking feature addition and is documented at <a href="../manual/nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a>.</p><h3 id="Breaking"><a class="docs-heading-anchor" href="#Breaking">Breaking</a><a id="Breaking-1"></a><a class="docs-heading-anchor-permalink" href="#Breaking" title="Permalink"></a></h3><p>Although the new nonlinear interface is a feature addition, there are two changes which might be breaking for a very small number of users.</p><ul><li>The syntax inside JuMP macros is parsed using a different code path, even for linear and quadratic expressions. We made this change to unify how we parse linear, quadratic, and nonlinear expressions. In all cases, the new code returns equivalent expressions, but because of the different order of operations, there are three changes to be aware of when updating:<ul><li>The printed form of the expression may change, for example from <code>x * y</code> to <code>y * x</code>. This can cause tests which test the <code>String</code> representation of a model to fail.</li><li>Some coefficients may change slightly due to floating point round-off error.</li><li>Particularly when working with a JuMP extension, you may encounter a <code>MethodError</code> due to a missing or ambiguous method. These errors are due to previously existing bugs that were not triggered by the previous parsing code. If you encounter such an error, please open a GitHub issue.</li></ul></li><li>The methods for <code>Base.:^(x::VariableRef, n::Integer)</code> and <code>Base.:^(x::AffExpr, n::Integer)</code> have changed. Previously, these methods supported only <code>n = 0, 1, 2</code> and they always returned a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a>, even for the case when <code>n = 0</code> or <code>n = 1</code>. Now:<ul><li><code>x^0</code> returns <code>one(T)</code>, where <code>T</code> is the <a href="../api/JuMP/#JuMP.value_type"><code>value_type</code></a> of the model (defaults to <code>Float64</code>)</li><li><code>x^1</code> returns <code>x</code></li><li><code>x^2</code> returns a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a></li><li><code>x^n</code> where <code>!(0 &lt;= n &lt;= 2)</code> returns a <a href="../api/JuMP/#JuMP.NonlinearExpr"><code>NonlinearExpr</code></a>.</li></ul>We made this change to support nonlinear expressions and to align the mathematical definition of the operation with their return type. (Previously, users were surprised that <code>x^1</code> returned a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a>.) As a consequence of this change, the methods are now not type-stable. This means that the compiler cannot prove that <code>x^2</code> returns a <a href="../api/JuMP/#JuMP.QuadExpr"><code>QuadExpr</code></a>. If benchmarking shows that this is a performance problem, you can use the type-stable <code>x * x</code> instead of <code>x^2</code>.</li></ul><h3 id="Added"><a class="docs-heading-anchor" href="#Added">Added</a><a id="Added-1"></a><a class="docs-heading-anchor-permalink" href="#Added" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP/#JuMP.triangle_vec"><code>triangle_vec</code></a> which simplifies adding <a href="../moi/reference/standard_form/#MathOptInterface.LogDetConeTriangle"><code>MOI.LogDetConeTriangle</code></a> and <a href="../moi/reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>MOI.RootDetConeTriangle</code></a> constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/3456">#3456</a>)</li><li>Added the new nonlinear interface. This is a very large change. See the documentation at <a href="../manual/nonlinear/#Nonlinear-Modeling">Nonlinear Modeling</a> and the (long) discussion in <a href="https://github.com/jump-dev/JuMP.jl/pull/3106">JuMP.jl#3106</a>. Related PRs are (<a href="https://github.com/jump-dev/JuMP.jl/issues/3468">#3468</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3472">#3472</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3475">#3475</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3483">#3483</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3487">#3487</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3488">#3488</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3489">#3489</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3504">#3504</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3509">#3509</a>)</li></ul><h3 id="Fixed-2"><a class="docs-heading-anchor" href="#Fixed-2">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-2" title="Permalink"></a></h3><ul><li>Fixed uses of <code>@nospecialize</code> which cause precompilation failures in Julia v1.6.0 and v1.6.1. (<a href="https://github.com/jump-dev/JuMP.jl/issues/3464">#3464</a>)</li><li>Fixed adding a container of <a href="../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3473">#3473</a>)</li><li>Fixed return type of <code>x^0</code> and <code>x^1</code> to no longer return <code>QuadExpr</code> (see note in <code>Breaking</code> section above) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3474">#3474</a>)</li><li>Fixed error messages in <a href="../api/JuMP/#JuMP.LowerBoundRef"><code>LowerBoundRef</code></a>, <a href="../api/JuMP/#JuMP.UpperBoundRef"><code>UpperBoundRef</code></a>, <a href="../api/JuMP/#JuMP.FixRef"><code>FixRef</code></a>, <a href="../api/JuMP/#JuMP.IntegerRef"><code>IntegerRef</code></a>, <a href="../api/JuMP/#JuMP.BinaryRef"><code>BinaryRef</code></a>, <a href="../api/JuMP/#JuMP.ParameterRef"><code>ParameterRef</code></a> and related functions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3494">#3494</a>)</li><li>Fixed type inference of empty containers in JuMP macros (<a href="https://github.com/jump-dev/JuMP.jl/issues/3500">#3500</a>)</li></ul><h3 id="Other-2"><a class="docs-heading-anchor" href="#Other-2">Other</a><a class="docs-heading-anchor-permalink" href="#Other-2" title="Permalink"></a></h3><ul><li>Added GAMS to solver documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3357">#3357</a>)</li><li>Updated various tutorials (<a href="https://github.com/jump-dev/JuMP.jl/issues/3459">#3459</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3460">#3460</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3462">#3462</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3463">#3463</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3465">#3465</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3490">#3490</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3492">#3492</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3503">#3503</a>)</li><li>Added <a href="../tutorials/linear/multi_commodity_network/#The-network-multi-commodity-flow-problem">The network multi-commodity flow problem</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3491">#3491</a>)</li><li>Added <a href="../tutorials/applications/two_stage_stochastic/#Two-stage-stochastic-programs">Two-stage stochastic programs</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3466">#3466</a>)</li><li>Added better error messages for unsupported operations in <code>LinearAlgebra</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3476">#3476</a>)</li><li>Updated to the latest version of Documenter (<a href="https://github.com/jump-dev/JuMP.jl/issues/3484">#3484</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3495">#3495</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3497">#3497</a>)</li><li>Updated GitHub action versions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3507">#3507</a>)</li></ul><h2 id="[Version-1.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.1)-(September-2,-2023)"><a class="docs-heading-anchor" href="#[Version-1.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.1)-(September-2,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.1">Version 1.14.1</a> (September 2, 2023)</a><a id="[Version-1.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.1)-(September-2,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.1)-(September-2,-2023)" title="Permalink"></a></h2><h3 id="Fixed-3"><a class="docs-heading-anchor" href="#Fixed-3">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-3" title="Permalink"></a></h3><ul><li>Fix links in Documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3478">#3478</a>)</li></ul><h2 id="[Version-1.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.0)-(August-27,-2023)"><a class="docs-heading-anchor" href="#[Version-1.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.0)-(August-27,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.0">Version 1.14.0</a> (August 27, 2023)</a><a id="[Version-1.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.0)-(August-27,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.0)-(August-27,-2023)" title="Permalink"></a></h2><h3 id="Added-2"><a class="docs-heading-anchor" href="#Added-2">Added</a><a class="docs-heading-anchor-permalink" href="#Added-2" title="Permalink"></a></h3><ul><li>Added <a href="../extensions/DimensionalData/#DimensionalData.jl">DimensionalData.jl</a> extension (<a href="https://github.com/jump-dev/JuMP.jl/issues/3413">#3413</a>)</li><li>Added syntactic sugar for the <a href="../moi/reference/standard_form/#MathOptInterface.Parameter"><code>MOI.Parameter</code></a> set (<a href="https://github.com/jump-dev/JuMP.jl/issues/3443">#3443</a>)<ul><li><a href="../api/JuMP/#JuMP.Parameter"><code>Parameter</code></a></li><li><a href="../api/JuMP/#JuMP.ParameterRef"><code>ParameterRef</code></a></li><li><a href="../api/JuMP/#JuMP.is_parameter"><code>is_parameter</code></a></li><li><a href="../api/JuMP/#JuMP.parameter_value"><code>parameter_value</code></a></li><li><a href="../api/JuMP/#JuMP.set_parameter_value"><code>set_parameter_value</code></a></li></ul></li></ul><h3 id="Fixed-4"><a class="docs-heading-anchor" href="#Fixed-4">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-4" title="Permalink"></a></h3><ul><li>Fixed <code>model_convert</code> for <code>BridgeableConstraint</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3437">#3437</a>)</li><li>Fixed printing models with integer coefficients larger than <code>typemax(Int)</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3447">#3447</a>)</li><li>Fixed support for constant left-hand side functions in a complementarity constraint (<a href="https://github.com/jump-dev/JuMP.jl/issues/3452">#3452</a>)</li></ul><h3 id="Other-3"><a class="docs-heading-anchor" href="#Other-3">Other</a><a class="docs-heading-anchor-permalink" href="#Other-3" title="Permalink"></a></h3><ul><li>Updated packages used in documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3444">#3444</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3455">#3455</a>)</li><li>Fixed docstring tests (<a href="https://github.com/jump-dev/JuMP.jl/issues/3445">#3445</a>)</li><li>Fixed printing change for MathOptInterface (<a href="https://github.com/jump-dev/JuMP.jl/issues/3446">#3446</a>)</li><li>Fixed typos in documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3448">#3448</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3457">#3457</a>)</li><li>Added SCIP to callback documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3449">#3449</a>)</li></ul><h2 id="[Version-1.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.13.0)-(July-27,-2023)"><a class="docs-heading-anchor" href="#[Version-1.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.13.0)-(July-27,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.13.0">Version 1.13.0</a> (July 27, 2023)</a><a id="[Version-1.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.13.0)-(July-27,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.13.0)-(July-27,-2023)" title="Permalink"></a></h2><h3 id="Added-3"><a class="docs-heading-anchor" href="#Added-3">Added</a><a class="docs-heading-anchor-permalink" href="#Added-3" title="Permalink"></a></h3><ul><li>Added support for generic number types (<a href="https://github.com/jump-dev/JuMP.jl/issues/3377">#3377</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3385">#3385</a>)</li><li>Added fallback for <a href="../moi/reference/standard_form/#MathOptInterface.AbstractSymmetricMatrixSetTriangle"><code>MOI.AbstractSymmetricMatrixSetTriangle</code></a> and <a href="../moi/reference/standard_form/#MathOptInterface.AbstractSymmetricMatrixSetSquare"><code>MOI.AbstractSymmetricMatrixSetSquare</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3424">#3424</a>)</li></ul><h3 id="Fixed-5"><a class="docs-heading-anchor" href="#Fixed-5">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-5" title="Permalink"></a></h3><ul><li>Fixed <a href="../api/JuMP/#JuMP.set_start_values"><code>set_start_values</code></a> with <a href="../moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.SlackBridge"><code>MOI.Bridges.Objective.SlackBridge</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3422">#3422</a>)</li><li>Fixed flakey doctest in <code>variables.md</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3425">#3425</a>)</li><li>Fixed names on <code>CITATION.bib</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3423">#3423</a>)</li></ul><h3 id="Other-4"><a class="docs-heading-anchor" href="#Other-4">Other</a><a class="docs-heading-anchor-permalink" href="#Other-4" title="Permalink"></a></h3><ul><li>Added Loraine.jl to the installation table (<a href="https://github.com/jump-dev/JuMP.jl/issues/3426">#3426</a>)</li><li>Removed Penopt.jl from packages.toml (<a href="https://github.com/jump-dev/JuMP.jl/issues/3428">#3428</a>)</li><li>Improved problem statement in cannery example of tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3430">#3430</a>)</li><li>Minor cleanups in <a href="../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> implementation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3429">#3429</a>)</li><li>Changed <code>nested_problems.jl</code>: outer/inner to upper/lower (<a href="https://github.com/jump-dev/JuMP.jl/issues/3433">#3433</a>)</li><li>Removed second SDP relaxation in OPF tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3432">#3432</a>)</li></ul><h2 id="[Version-1.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.12.0)-(June-19,-2023)"><a class="docs-heading-anchor" href="#[Version-1.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.12.0)-(June-19,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.12.0">Version 1.12.0</a> (June 19, 2023)</a><a id="[Version-1.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.12.0)-(June-19,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.12.0)-(June-19,-2023)" title="Permalink"></a></h2><h3 id="Added-4"><a class="docs-heading-anchor" href="#Added-4">Added</a><a class="docs-heading-anchor-permalink" href="#Added-4" title="Permalink"></a></h3><ul><li>Added <code>coefficient_type</code> keyword argument to <a href="../api/JuMP/#JuMP.add_bridge"><code>add_bridge</code></a> and <a href="../api/JuMP/#JuMP.remove_bridge"><code>remove_bridge</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3394">#3394</a>)</li></ul><h3 id="Fixed-6"><a class="docs-heading-anchor" href="#Fixed-6">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-6" title="Permalink"></a></h3><ul><li>Fixed error message for matrix in <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3369">#3369</a>)</li><li>Fixed <code>EditURL</code> for custom documentation pages (<a href="https://github.com/jump-dev/JuMP.jl/issues/3373">#3373</a>)</li><li>Fixed return type annotations for <a href="../moi/reference/constraints/#MathOptInterface.ConstraintPrimal"><code>MOI.ConstraintPrimal</code></a> and <a href="../moi/reference/constraints/#MathOptInterface.ConstraintDual"><code>MOI.ConstraintDual</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3381">#3381</a>)</li><li>Fixed printing change in Julia nightly (<a href="https://github.com/jump-dev/JuMP.jl/issues/3391">#3391</a>)</li><li>Fixed printing of <code>Complex</code> coefficients (<a href="https://github.com/jump-dev/JuMP.jl/issues/3397">#3397</a>)</li><li>Fixed printing of constraints in <code>text/latex</code> mode (<a href="https://github.com/jump-dev/JuMP.jl/issues/3405">#3405</a>)</li><li>Fixed performance issue in <a href="../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3410">#3410</a>)</li><li>Fixed bug when variables added to set of wrong dimension (<a href="https://github.com/jump-dev/JuMP.jl/issues/3411">#3411</a>)</li></ul><h3 id="Other-5"><a class="docs-heading-anchor" href="#Other-5">Other</a><a class="docs-heading-anchor-permalink" href="#Other-5" title="Permalink"></a></h3><ul><li>Added more solver READMEs to the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3358">#3358</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3360">#3360</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3364">#3364</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3365">#3365</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3366">#3366</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3368">#3368</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3372">#3372</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3374">#3374</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3376">#3376</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3379">#3379</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3387">#3387</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3389">#3389</a>)</li><li>Added StatusSwitchingQP.jl to the installation table (<a href="https://github.com/jump-dev/JuMP.jl/issues/3354">#3354</a>)</li><li>Updated checklist for adding a new solver (<a href="https://github.com/jump-dev/JuMP.jl/issues/3370">#3370</a>)</li><li>Updated <code>extension-tests.yml</code> action (<a href="https://github.com/jump-dev/JuMP.jl/issues/3371">#3371</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3375">#3375</a>)</li><li>Color logs in GitHub actions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3392">#3392</a>)</li><li>Added new tutorials<ul><li><a href="../tutorials/applications/optimal_power_flow/#Optimal-power-flow">Optimal power flow</a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3395">#3395</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3412">#3412</a>)</li><li><a href="../tutorials/conic/simple_examples/#Lovász-numbers">Lovász numbers</a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3399">#3399</a>)</li><li><a href="../tutorials/conic/dualization/#Dualization">Dualization</a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3402">#3402</a>)</li></ul></li><li>Updated JuMP paper citation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3400">#3400</a>)</li><li>Changed GitHub action to upload LaTeX logs when building documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3403">#3403</a>)</li><li>Fixed printing of SCS log in documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3406">#3406</a>)</li><li>Updated solver versions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3407">#3407</a>)</li><li>Updated documentation to use Julia v1.9 (<a href="https://github.com/jump-dev/JuMP.jl/issues/3398">#3398</a>)</li><li>Replaced <code>_value_type</code> with <code>MOI.Utilities.value_type</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3414">#3414</a>)</li><li>Fixed a typo in docstring (<a href="https://github.com/jump-dev/JuMP.jl/issues/3415">#3415</a>)</li><li>Refactored API documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3386">#3386</a>)</li><li>Updated SCIP license (<a href="https://github.com/jump-dev/JuMP.jl/issues/3420">#3420</a>)</li></ul><h2 id="[Version-1.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.1)-(May-19,-2023)"><a class="docs-heading-anchor" href="#[Version-1.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.1)-(May-19,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.1">Version 1.11.1</a> (May 19, 2023)</a><a id="[Version-1.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.1)-(May-19,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.1)-(May-19,-2023)" title="Permalink"></a></h2><h3 id="Fixed-7"><a class="docs-heading-anchor" href="#Fixed-7">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-7" title="Permalink"></a></h3><ul><li>Fixed a poor error message when <code>sum(::DenseAxisArray; dims)</code> was called (<a href="https://github.com/jump-dev/JuMP.jl/issues/3338">#3338</a>)</li><li>Fixed support for dependent sets in the <a href="../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro (<a href="https://github.com/jump-dev/JuMP.jl/issues/3344">#3344</a>)</li><li>Fixed a performance bug in constraints with sparse symmetric matrices (<a href="https://github.com/jump-dev/JuMP.jl/issues/3349">#3349</a>)</li></ul><h3 id="Other-6"><a class="docs-heading-anchor" href="#Other-6">Other</a><a class="docs-heading-anchor-permalink" href="#Other-6" title="Permalink"></a></h3><ul><li>Improved the printing of complex numbers (<a href="https://github.com/jump-dev/JuMP.jl/issues/3332">#3332</a>)</li><li>When printing, sets which contain constants ending in <code>.0</code> now print as integers. This follows the behavior of constants in functions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3341">#3341</a>)</li><li>Added <code>InfiniteOpt</code> to the extensions documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3343">#3343</a>)</li><li>Added more documentation for the exponential cone (<a href="https://github.com/jump-dev/JuMP.jl/issues/3345">#3345</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3347">#3347</a>)</li><li>Added checklists for developers (<a href="https://github.com/jump-dev/JuMP.jl/issues/3346">#3346</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3355">#3355</a>)</li><li>Fixed test support upcoming Julia nightly (<a href="https://github.com/jump-dev/JuMP.jl/issues/3351">#3351</a>)</li><li>Fixed <code>extension-tests.yml</code> action (<a href="https://github.com/jump-dev/JuMP.jl/issues/3353">#3353</a>)</li><li>Add more solvers to the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3359">#3359</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3361">#3361</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3362">#3362</a>)</li></ul><h2 id="[Version-1.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.0)-(May-3,-2023)"><a class="docs-heading-anchor" href="#[Version-1.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.0)-(May-3,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.0">Version 1.11.0</a> (May 3, 2023)</a><a id="[Version-1.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.0)-(May-3,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.0)-(May-3,-2023)" title="Permalink"></a></h2><h3 id="Added-5"><a class="docs-heading-anchor" href="#Added-5">Added</a><a class="docs-heading-anchor-permalink" href="#Added-5" title="Permalink"></a></h3><ul><li>Added new methods to <a href="../api/JuMP/#JuMP.print_active_bridges"><code>print_active_bridges</code></a> for printing a particular objective, constraint, or variable (<a href="https://github.com/jump-dev/JuMP.jl/issues/3316">#3316</a>)</li></ul><h3 id="Fixed-8"><a class="docs-heading-anchor" href="#Fixed-8">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-8" title="Permalink"></a></h3><ul><li>Fixed tests for MOI v1.14.0 release (<a href="https://github.com/jump-dev/JuMP.jl/issues/3312">#3312</a>)</li><li>Fixed indexing containers when an axis is <code>Vector{Any}</code> that contains a <code>Vector{Any}</code> element (<a href="https://github.com/jump-dev/JuMP.jl/issues/3280">#3280</a>)</li><li>Fixed <code>getindex(::AbstractJuMPScalar)</code> which is called for an expression like <code>x[]</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3314">#3314</a>)</li><li>Fixed bug in <code>set_string_names_on_creation</code> with a vector of variables (<a href="https://github.com/jump-dev/JuMP.jl/issues/3322">#3322</a>)</li><li>Fixed bug in <code>memoize</code> function in nonlinear documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3337">#3337</a>)</li></ul><h3 id="Other-7"><a class="docs-heading-anchor" href="#Other-7">Other</a><a class="docs-heading-anchor-permalink" href="#Other-7" title="Permalink"></a></h3><ul><li>Fixed typos in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3317">#3317</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3318">#3318</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3328">#3328</a>)</li><li>Added a test for the order of setting start values (<a href="https://github.com/jump-dev/JuMP.jl/issues/3315">#3315</a>)</li><li>Added READMEs of solvers and extensions to the docs (<a href="https://github.com/jump-dev/JuMP.jl/issues/3309">#3309</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3320">#3320</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3327">#3327</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3329">#3329</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3333">#3333</a>)</li><li>Style improvements to <code>src/variables.jl</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3324">#3324</a>)</li><li>Clarify that column generation does not find global optimum (<a href="https://github.com/jump-dev/JuMP.jl/issues/3325">#3325</a>)</li><li>Add a GitHub actions workflow for testing extensions prior to release (<a href="https://github.com/jump-dev/JuMP.jl/issues/3331">#3331</a>)</li><li>Document the release process for JuMP (<a href="https://github.com/jump-dev/JuMP.jl/issues/3334">#3334</a>)</li><li>Fix links to discourse and chatroom (<a href="https://github.com/jump-dev/JuMP.jl/issues/3335">#3335</a>)</li></ul><h2 id="[Version-1.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.10.0)-(April-3,-2023)"><a class="docs-heading-anchor" href="#[Version-1.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.10.0)-(April-3,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.10.0">Version 1.10.0</a> (April 3, 2023)</a><a id="[Version-1.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.10.0)-(April-3,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.10.0)-(April-3,-2023)" title="Permalink"></a></h2><h3 id="Added-6"><a class="docs-heading-anchor" href="#Added-6">Added</a><a class="docs-heading-anchor-permalink" href="#Added-6" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP/#JuMP.Nonnegatives"><code>Nonnegatives</code></a>, <a href="../api/JuMP/#JuMP.Nonpositives"><code>Nonpositives</code></a> and <a href="../api/JuMP/#JuMP.Zeros"><code>Zeros</code></a>, and support vector-valued inequality syntax in the JuMP macros (<a href="https://github.com/jump-dev/JuMP.jl/issues/3273">#3273</a>)</li><li>Added special support for <code>LinearAlgebra.Symmetric</code> and <code>LinearAlgebra.Hermitian</code> matrices in <a href="../api/JuMP/#JuMP.Zeros"><code>Zeros</code></a> constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/3281">#3281</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3296">#3296</a>)</li><li>Added <a href="../api/JuMP/#JuMP.HermitianMatrixSpace"><code>HermitianMatrixSpace</code></a> and the <code>Hermitian</code> tag for generating a matrix of variables that is Hermitian (<a href="https://github.com/jump-dev/JuMP.jl/issues/3292">#3292</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3293">#3293</a>)</li><li>Added <a href="../api/JuMP/#JuMP.Semicontinuous"><code>Semicontinuous</code></a> and <a href="../api/JuMP/#JuMP.Semiinteger"><code>Semiinteger</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3302">#3302</a>)</li><li>Added support for keyword indexing of containers (<a href="https://github.com/jump-dev/JuMP.jl/issues/3237">#3237</a>)</li></ul><h3 id="Fixed-9"><a class="docs-heading-anchor" href="#Fixed-9">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-9" title="Permalink"></a></h3><ul><li>Fixed <code>[compat]</code> bound for MathOptInterface in <code>Project.toml</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3272">#3272</a>)</li></ul><h3 id="Other-8"><a class="docs-heading-anchor" href="#Other-8">Other</a><a class="docs-heading-anchor-permalink" href="#Other-8" title="Permalink"></a></h3><ul><li>Split out the <a href="../tutorials/nonlinear/nested_problems/#Nested-optimization-problems">Nested optimization problems</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3274">#3274</a>)</li><li>Updated doctests to ensure none have hidden state (<a href="https://github.com/jump-dev/JuMP.jl/issues/3275">#3275</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3276">#3276</a>)</li><li>Clarified how lazy constraints may revisit points (<a href="https://github.com/jump-dev/JuMP.jl/issues/3278">#3278</a>)</li><li>Added <a href="../tutorials/conic/tips_and_tricks/#P-Norm">P-Norm</a> example (<a href="https://github.com/jump-dev/JuMP.jl/issues/3282">#3282</a>)</li><li>Clarified docs that macros create new bindings (<a href="https://github.com/jump-dev/JuMP.jl/issues/3284">#3284</a>)</li><li>Fixed threading example (<a href="https://github.com/jump-dev/JuMP.jl/issues/3283">#3283</a>)</li><li>Added plot to <a href="../tutorials/conic/simple_examples/#The-minimum-distortion-problem">The minimum distortion problem</a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3288">#3288</a>)</li><li>Added Google style rules for Vale and fixed warnings (<a href="https://github.com/jump-dev/JuMP.jl/issues/3285">#3285</a>)</li><li>Added citation for the JuMP 1.0 paper (<a href="https://github.com/jump-dev/JuMP.jl/issues/3294">#3294</a>)</li><li>Updated package versions in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3298">#3298</a>)</li><li>Added comment for the order in which start values must be set (<a href="https://github.com/jump-dev/JuMP.jl/issues/3303">#3303</a>)</li><li>Improved error message for unrecognized constraint operators (<a href="https://github.com/jump-dev/JuMP.jl/issues/3311">#3311</a>)</li></ul><h2 id="[Version-1.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.9.0)-(March-7,-2023)"><a class="docs-heading-anchor" href="#[Version-1.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.9.0)-(March-7,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.9.0">Version 1.9.0</a> (March 7, 2023)</a><a id="[Version-1.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.9.0)-(March-7,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.9.0)-(March-7,-2023)" title="Permalink"></a></h2><h3 id="Added-7"><a class="docs-heading-anchor" href="#Added-7">Added</a><a class="docs-heading-anchor-permalink" href="#Added-7" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP/#JuMP.get_attribute"><code>get_attribute</code></a> and <a href="../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a>. These replace <a href="../api/JuMP/#JuMP.get_optimizer_attribute"><code>get_optimizer_attribute</code></a> and <a href="../api/JuMP/#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a>, although the <code>_optimizer_</code> functions remain for backward compatibility. (<a href="https://github.com/jump-dev/JuMP.jl/issues/3219">#3219</a>)</li><li>Added <a href="../api/JuMP/#JuMP.set_start_values"><code>set_start_values</code></a> for setting all supported start values in a model (<a href="https://github.com/jump-dev/JuMP.jl/issues/3238">#3238</a>)</li><li>Add <a href="../api/JuMP/#JuMP.remove_bridge"><code>remove_bridge</code></a> and <a href="../api/JuMP/#JuMP.print_active_bridges"><code>print_active_bridges</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3259">#3259</a>)</li></ul><h3 id="Fixed-10"><a class="docs-heading-anchor" href="#Fixed-10">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-10" title="Permalink"></a></h3><ul><li>The matrix returned by a variable in <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> is now a <code>LinearAlgebra.Hermitian</code> matrix. This is potentially breaking if you have written code to assume the return is a <code>Matrix</code>. (<a href="https://github.com/jump-dev/JuMP.jl/issues/3245">#3245</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3246">#3246</a>)</li><li>Fixed missing support for <code>Base.isreal</code> of expressions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3252">#3252</a>)</li></ul><h3 id="Other-9"><a class="docs-heading-anchor" href="#Other-9">Other</a><a class="docs-heading-anchor-permalink" href="#Other-9" title="Permalink"></a></h3><ul><li>Fixed a thread safety issue in the <a href="../tutorials/algorithms/parallelism/#Parallelism">Parallelism</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3240">#3240</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3243">#3243</a>)</li><li>Improved the error message when unsupported operators are used in <code>@NL</code> macros (<a href="https://github.com/jump-dev/JuMP.jl/issues/3236">#3236</a>)</li><li>Clarified the documentation to say that matrices in <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> must be <code>LinearAlgebra.Hermitian</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3241">#3241</a>)</li><li>Minor style fixes to internal macro code (<a href="https://github.com/jump-dev/JuMP.jl/issues/3247">#3247</a>)</li><li>Add <a href="../tutorials/conic/quantum_discrimination/#Quantum-state-discrimination">Quantum state discrimination</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3250">#3250</a>)</li><li>Improve error message when <code>begin...end</code> not passed to plural macros (<a href="https://github.com/jump-dev/JuMP.jl/issues/3255">#3255</a>)</li><li>Document how to register function with varying number of input arguments (<a href="https://github.com/jump-dev/JuMP.jl/issues/3258">#3258</a>)</li><li>Tidy tests by removing unneeded <code>JuMP.</code> prefixes (<a href="https://github.com/jump-dev/JuMP.jl/issues/3260">#3260</a>)</li><li>Clarified the introduction to the <a href="../manual/complex/#Complex-number-support">Complex number support</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3262">#3262</a>)</li><li>Fixed typos in the Documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3263">#3263</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3266">#3266</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3268">#3268</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3269">#3269</a>)</li></ul><h2 id="[Version-1.8.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.2)-(February-27,-2023)"><a class="docs-heading-anchor" href="#[Version-1.8.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.2)-(February-27,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.2">Version 1.8.2</a> (February 27, 2023)</a><a id="[Version-1.8.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.2)-(February-27,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.8.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.2)-(February-27,-2023)" title="Permalink"></a></h2><h3 id="Fixed-11"><a class="docs-heading-anchor" href="#Fixed-11">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-11" title="Permalink"></a></h3><ul><li>Fixed dot product between complex JuMP expression and number (<a href="https://github.com/jump-dev/JuMP.jl/issues/3244">#3244</a>)</li></ul><h3 id="Other-10"><a class="docs-heading-anchor" href="#Other-10">Other</a><a class="docs-heading-anchor-permalink" href="#Other-10" title="Permalink"></a></h3><ul><li>Polish simple SDP examples (<a href="https://github.com/jump-dev/JuMP.jl/issues/3232">#3232</a>)</li></ul><h2 id="[Version-1.8.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.1)-(February-23,-2023)"><a class="docs-heading-anchor" href="#[Version-1.8.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.1)-(February-23,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.1">Version 1.8.1</a> (February 23, 2023)</a><a id="[Version-1.8.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.1)-(February-23,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.8.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.1)-(February-23,-2023)" title="Permalink"></a></h2><h3 id="Fixed-12"><a class="docs-heading-anchor" href="#Fixed-12">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-12" title="Permalink"></a></h3><ul><li>Fixed support for <code>init</code> in nonlinear generator expressions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3226">#3226</a>)</li></ul><h3 id="Other-11"><a class="docs-heading-anchor" href="#Other-11">Other</a><a class="docs-heading-anchor-permalink" href="#Other-11" title="Permalink"></a></h3><ul><li>Use and document <code>import MathOptInterface as MOI</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3222">#3222</a>)</li><li>Removed references in documentation to multiobjective optimization being unsupported (<a href="https://github.com/jump-dev/JuMP.jl/issues/3223">#3223</a>)</li><li>Added tutorial on multi-objective portfolio optimization (<a href="https://github.com/jump-dev/JuMP.jl/issues/3227">#3227</a>)</li><li>Refactored some of the conic tutorials (<a href="https://github.com/jump-dev/JuMP.jl/issues/3229">#3229</a>)</li><li>Fixed typos in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3230">#3230</a>)</li><li>Added tutorial on parallelism (<a href="https://github.com/jump-dev/JuMP.jl/issues/3231">#3231</a>)</li></ul><h2 id="[Version-1.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.0)-(February-16,-2023)"><a class="docs-heading-anchor" href="#[Version-1.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.0)-(February-16,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.0">Version 1.8.0</a> (February 16, 2023)</a><a id="[Version-1.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.0)-(February-16,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.0)-(February-16,-2023)" title="Permalink"></a></h2><h3 id="Added-8"><a class="docs-heading-anchor" href="#Added-8">Added</a><a class="docs-heading-anchor-permalink" href="#Added-8" title="Permalink"></a></h3><ul><li>Added <code>--&gt;</code> syntax support for indicator constraints. The old syntax of <code>=&gt;</code> remains supported (<a href="https://github.com/jump-dev/JuMP.jl/issues/3207">#3207</a>)</li><li>Added <code>&lt;--&gt;</code> syntax for reified constraints. For now, few solvers support reified constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/3206">#3206</a>)</li><li>Added <a href="../api/JuMP/#JuMP.fix_discrete_variables"><code>fix_discrete_variables</code></a>. This is most useful for computing the dual of a mixed-integer program (<a href="https://github.com/jump-dev/JuMP.jl/issues/3208">#3208</a>)</li><li>Added support for vector-valued objectives. For details, see the <a href="../tutorials/linear/multi_objective_knapsack/#Multi-objective-knapsack">Multi-objective knapsack</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3176">#3176</a>)</li></ul><h3 id="Fixed-13"><a class="docs-heading-anchor" href="#Fixed-13">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-13" title="Permalink"></a></h3><ul><li>Fixed a bug in <a href="../api/JuMP/#JuMP.lp_sensitivity_report"><code>lp_sensitivity_report</code></a> by switching to an explicit LU factorization of the basis matrix (<a href="https://github.com/jump-dev/JuMP.jl/issues/3182">#3182</a>)</li><li>Fixed a bug that prevented <code>[; kwarg]</code> arguments in macros (<a href="https://github.com/jump-dev/JuMP.jl/issues/3220">#3220</a>)</li></ul><h3 id="Other-12"><a class="docs-heading-anchor" href="#Other-12">Other</a><a class="docs-heading-anchor-permalink" href="#Other-12" title="Permalink"></a></h3><ul><li>Minor fixes to the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3200">#3200</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3201">#3201</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3203">#3203</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3210">#3210</a>)</li><li>Added tutorial <a href="../tutorials/linear/constraint_programming/#Constraint-programming">Constraint programming</a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3202">#3202</a>)</li><li>Added more examples to <a href="../tutorials/conic/tips_and_tricks/#conic_tips_and_tricks">Tips and Tricks</a></li><li>Remove <code>_distance_to_set</code> in favor of <code>MOI.Utilities.distance_to_set</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3209">#3209</a>)</li><li>Improve <a href="../tutorials/linear/diet/#The-diet-problem">The diet problem</a> tutorial by adding the variable as a column in the dataframe (<a href="https://github.com/jump-dev/JuMP.jl/issues/3213">#3213</a>)</li><li>Improve <a href="../tutorials/linear/knapsack/#The-knapsack-problem-example">The knapsack problem example</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3216">#3216</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3217">#3217</a>)</li><li>Added the <a href="../tutorials/conic/ellipse_approx/#Ellipsoid-approximation">Ellipsoid approximation</a> tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3218">#3218</a>)</li></ul><h2 id="[Version-1.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.7.0)-(January-25,-2023)"><a class="docs-heading-anchor" href="#[Version-1.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.7.0)-(January-25,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.7.0">Version 1.7.0</a> (January 25, 2023)</a><a id="[Version-1.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.7.0)-(January-25,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.7.0)-(January-25,-2023)" title="Permalink"></a></h2><h3 id="Added-9"><a class="docs-heading-anchor" href="#Added-9">Added</a><a class="docs-heading-anchor-permalink" href="#Added-9" title="Permalink"></a></h3><ul><li>Added support for <code>view</code> of a <a href="../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3152">#3152</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3180">#3180</a>)</li><li>Added support for containers of variables in <a href="../api/JuMP/#JuMP.ComplexPlane"><code>ComplexPlane</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3184">#3184</a>)</li><li>Added support for <code>minimum</code> and <code>maximum</code> generators in nonlinear expressions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3189">#3189</a>)</li><li>Added <code>SnoopPrecompile</code> statements that reduce the time-to-first-solve in Julia 1.9 (<a href="https://github.com/jump-dev/JuMP.jl/issues/3193">#3193</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3195">#3195</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3196">#3196</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3197">#3197</a>)</li></ul><h3 id="Other-13"><a class="docs-heading-anchor" href="#Other-13">Other</a><a class="docs-heading-anchor-permalink" href="#Other-13" title="Permalink"></a></h3><ul><li>Large refactoring of the tests (<a href="https://github.com/jump-dev/JuMP.jl/issues/3166">#3166</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3167">#3167</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3168">#3168</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3169">#3169</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3170">#3170</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3171">#3171</a>)</li><li>Remove unreachable code due to <code>VERSION</code> checks (<a href="https://github.com/jump-dev/JuMP.jl/issues/3172">#3172</a>)</li><li>Document how to test JuMP extensions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3174">#3174</a>)</li><li>Fix method ambiguities in <code>Containers</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3173">#3173</a>)</li><li>Improve error message that is thrown when <code>=</code> is used instead of <code>==</code> in the <a href="../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macro (<a href="https://github.com/jump-dev/JuMP.jl/issues/3178">#3178</a>)</li><li>Improve the error message when <code>Bool</code> is used instead of <code>Bin</code> in the <a href="../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro (<a href="https://github.com/jump-dev/JuMP.jl/issues/3180">#3180</a>)</li><li>Update versions of the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3185">#3185</a>)</li><li>Tidy the import of packages and remove unnecessary prefixes (<a href="https://github.com/jump-dev/JuMP.jl/issues/3186">#3186</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3187">#3187</a>)</li><li>Refactor <code>src/JuMP.jl</code> by moving methods into more relevant files (<a href="https://github.com/jump-dev/JuMP.jl/issues/3188">#3188</a>)</li><li>Fix docstring of <a href="../api/JuMP/#JuMP.Model"><code>Model</code></a> not appearing in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3198">#3198</a>)</li></ul><h2 id="[Version-1.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.6.0)-(January-1,-2023)"><a class="docs-heading-anchor" href="#[Version-1.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.6.0)-(January-1,-2023)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.6.0">Version 1.6.0</a> (January 1, 2023)</a><a id="[Version-1.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.6.0)-(January-1,-2023)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.6.0)-(January-1,-2023)" title="Permalink"></a></h2><h3 id="Added-10"><a class="docs-heading-anchor" href="#Added-10">Added</a><a class="docs-heading-anchor-permalink" href="#Added-10" title="Permalink"></a></h3><ul><li>Added a <code>result</code> keyword argument to <a href="../api/JuMP/#JuMP.solution_summary"><code>solution_summary</code></a> to allow summarizing models with multiple solutions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3138">#3138</a>)</li><li>Added <a href="../api/JuMP/#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a>, which is a useful tool when debugging infeasible models (<a href="https://github.com/jump-dev/JuMP.jl/issues/3140">#3140</a>)</li><li>Added <a href="../api/JuMP/#JuMP.has_start_value"><code>has_start_value</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3157">#3157</a>)</li><li>Added support for <a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> in constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/3154">#3154</a>)</li></ul><h3 id="Fixed-14"><a class="docs-heading-anchor" href="#Fixed-14">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-14" title="Permalink"></a></h3><ul><li>Fixed promotion of complex expressions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3150">#3150</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3164">#3164</a>)</li></ul><h3 id="Other-14"><a class="docs-heading-anchor" href="#Other-14">Other</a><a class="docs-heading-anchor-permalink" href="#Other-14" title="Permalink"></a></h3><ul><li>Added Benders tutorial with in-place resolves (<a href="https://github.com/jump-dev/JuMP.jl/issues/3145">#3145</a>)</li><li>Added more <a href="../tutorials/linear/tips_and_tricks/#linear_tips_and_tricks">Tips and tricks</a> for linear programs (<a href="https://github.com/jump-dev/JuMP.jl/issues/3144">#3144</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3163">#3163</a>)</li><li>Clarified documentation that <code>start</code> can depend on the indices of a variable container (<a href="https://github.com/jump-dev/JuMP.jl/issues/3148">#3148</a>)</li><li>Replace instances of <code>length</code> and <code>size</code> by the recommended <code>eachindex</code> and <code>axes</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3149">#3149</a>)</li><li>Added a warning explaining why the model is dirty when accessing solution results from a modified model (<a href="https://github.com/jump-dev/JuMP.jl/issues/3156">#3156</a>)</li><li>Clarify documentation that <code>PSD</code> ensures a symmetric matrix (<a href="https://github.com/jump-dev/JuMP.jl/issues/3159">#3159</a>)</li><li>Maintenance of the JuMP test suite (<a href="https://github.com/jump-dev/JuMP.jl/issues/3146">#3146</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3158">#3158</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3162">#3162</a>)</li></ul><h2 id="[Version-1.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.5.0)-(December-8,-2022)"><a class="docs-heading-anchor" href="#[Version-1.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.5.0)-(December-8,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.5.0">Version 1.5.0</a> (December 8, 2022)</a><a id="[Version-1.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.5.0)-(December-8,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.5.0)-(December-8,-2022)" title="Permalink"></a></h2><h3 id="Added-11"><a class="docs-heading-anchor" href="#Added-11">Added</a><a class="docs-heading-anchor-permalink" href="#Added-11" title="Permalink"></a></h3><ul><li>Add support for complex-valued variables:<ul><li><a href="../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3109">#3109</a>)</li><li><a href="../api/JuMP/#JuMP.ComplexPlane"><code>ComplexPlane</code></a> and <a href="../api/JuMP/#JuMP.ComplexVariable"><code>ComplexVariable</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3134">#3134</a>)</li></ul></li><li>Add support for <a href="../moi/reference/models/#MathOptInterface.OptimizerWithAttributes"><code>MOI.OptimizerWithAttributes</code></a> in <a href="../api/JuMP/#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a> and <a href="../api/JuMP/#JuMP.get_optimizer_attribute"><code>get_optimizer_attribute</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3129">#3129</a>)</li></ul><h3 id="Fixed-15"><a class="docs-heading-anchor" href="#Fixed-15">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-15" title="Permalink"></a></h3><ul><li>Fixed error message for vectorized interval constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/3123">#3123</a>)</li><li>Fixed passing <code>AbstractString</code> to <a href="../api/JuMP/#JuMP.set_optimizer_attribute"><code>set_optimizer_attribute</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3127">#3127</a>)</li></ul><h3 id="Other-15"><a class="docs-heading-anchor" href="#Other-15">Other</a><a class="docs-heading-anchor-permalink" href="#Other-15" title="Permalink"></a></h3><ul><li>Update package versions used in docs (<a href="https://github.com/jump-dev/JuMP.jl/issues/3119">#3119</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3133">#3133</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3139">#3139</a>)</li><li>Fixed output of diet tutorial (<a href="https://github.com/jump-dev/JuMP.jl/issues/3120">#3120</a>)</li><li>Explain how to use <code>Dates.period</code> in <a href="../api/JuMP/#JuMP.set_time_limit_sec"><code>set_time_limit_sec</code></a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3121">#3121</a>)</li><li>Update to <code>JuliaFormatter</code> v1.0.15 (<a href="https://github.com/jump-dev/JuMP.jl/issues/3130">#3130</a>)</li><li>Fixed <code>HTTP</code> server example in web_app.jl (<a href="https://github.com/jump-dev/JuMP.jl/issues/3131">#3131</a>)</li><li>Update docs to build with <code>Documenter#master</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3094">#3094</a>)</li><li>Add tests for <code>LinearAlgebra</code> operations (<a href="https://github.com/jump-dev/JuMP.jl/issues/3132">#3132</a>)</li><li>Tidy these release notes (<a href="https://github.com/jump-dev/JuMP.jl/issues/3135">#3135</a>)</li><li>Added documentation for <a href="../manual/complex/#Complex-number-support">Complex number support</a> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3141">#3141</a>)</li><li>Removed the &quot;workforce scheduling&quot; and &quot;steelT3&quot; tutorials (<a href="https://github.com/jump-dev/JuMP.jl/issues/3143">#3143</a>)</li></ul><h2 id="[Version-1.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.4.0)-(October-29,-2022)"><a class="docs-heading-anchor" href="#[Version-1.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.4.0)-(October-29,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.4.0">Version 1.4.0</a> (October 29, 2022)</a><a id="[Version-1.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.4.0)-(October-29,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.4.0)-(October-29,-2022)" title="Permalink"></a></h2><h3 id="Added-12"><a class="docs-heading-anchor" href="#Added-12">Added</a><a class="docs-heading-anchor-permalink" href="#Added-12" title="Permalink"></a></h3><ul><li>Added <a href="../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> which converts a container into a vector of <code>NamedTuple</code>s to support the Tables.jl interface. This simplifies converting <a href="../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> and <a href="../api/JuMP.Containers/#JuMP.Containers.SparseAxisArray"><code>Containers.SparseAxisArray</code></a> objects into tabular forms such as a DataFrame (<a href="https://github.com/jump-dev/JuMP.jl/issues/3104">#3104</a>)</li><li>Added a new method to <a href="../api/JuMP.Containers/#JuMP.Containers.container"><code>Containers.container</code></a> so that index names are passed to the container (<a href="https://github.com/jump-dev/JuMP.jl/issues/3088">#3088</a>)</li></ul><h3 id="Fixed-16"><a class="docs-heading-anchor" href="#Fixed-16">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-16" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>copy_to(dest::Model, src::MOI.ModelLike)</code> when <code>src</code> has nonlinear components (<a href="https://github.com/jump-dev/JuMP.jl/issues/3101">#3101</a>)</li><li>Fixed the printing of <code>(-1.0 + 0.0im)</code> coefficients in complex expressions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3112">#3112</a>)</li><li>Fixed a parsing bug in nonlinear expressions with generator statements that contain multiple <code>for</code> statements (<a href="https://github.com/jump-dev/JuMP.jl/issues/3116">#3116</a>)</li></ul><h3 id="Other-16"><a class="docs-heading-anchor" href="#Other-16">Other</a><a class="docs-heading-anchor-permalink" href="#Other-16" title="Permalink"></a></h3><ul><li>Converted the multi-commodity flow tutorial to use an SQLite database (<a href="https://github.com/jump-dev/JuMP.jl/issues/3098">#3098</a>)</li><li>Fixed a number of typos in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3103">#3103</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3107">#3107</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3018">#3018</a>)</li><li>Improved various style aspects of the PDF documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3095">#3095</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3098">#3098</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3102">#3102</a>)</li></ul><h2 id="[Version-1.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.1)-(September-28,-2022)"><a class="docs-heading-anchor" href="#[Version-1.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.1)-(September-28,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.1">Version 1.3.1</a> (September 28, 2022)</a><a id="[Version-1.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.1)-(September-28,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.1)-(September-28,-2022)" title="Permalink"></a></h2><h3 id="Fixed-17"><a class="docs-heading-anchor" href="#Fixed-17">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-17" title="Permalink"></a></h3><ul><li>Fixed a performance issue in <code>relax_integrality</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3087">#3087</a>)</li><li>Fixed the type stability of operators with <code>Complex</code> arguments (<a href="https://github.com/jump-dev/JuMP.jl/issues/3072">#3072</a>)</li><li>Fixed a bug which added additional <code>+()</code> terms to some nonlinear expressions (<a href="https://github.com/jump-dev/JuMP.jl/issues/3091">#3091</a>)</li><li>Fixed potential method ambiguities with <code>AffExpr</code> and <code>QuadExpr</code> objects (<a href="https://github.com/jump-dev/JuMP.jl/issues/3092">#3092</a>)</li></ul><h3 id="Other-17"><a class="docs-heading-anchor" href="#Other-17">Other</a><a class="docs-heading-anchor-permalink" href="#Other-17" title="Permalink"></a></h3><ul><li>Added <a href="https://vale.sh">vale</a> as a linter for the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3080">#3080</a>)</li><li>Added a tutorial on debugging JuMP models (<a href="https://github.com/jump-dev/JuMP.jl/issues/3043">#3043</a>)</li><li>Fixed a number of typos in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3079">#3079</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3083">#3083</a>)</li><li>Many other small tweaks to the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3068">#3068</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3073">#3073</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3074">#3074</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3075">#3075</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3076">#3076</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3077">#3077</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3078">#3078</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3081">#3081</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3082">#3082</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3084">#3084</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3085">#3085</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3089">#3089</a>)</li></ul><h2 id="[Version-1.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.0)-(September-5,-2022)"><a class="docs-heading-anchor" href="#[Version-1.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.0)-(September-5,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.0">Version 1.3.0</a> (September 5, 2022)</a><a id="[Version-1.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.0)-(September-5,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.0)-(September-5,-2022)" title="Permalink"></a></h2><h3 id="Added-13"><a class="docs-heading-anchor" href="#Added-13">Added</a><a class="docs-heading-anchor-permalink" href="#Added-13" title="Permalink"></a></h3><ul><li>Support slicing in <code>SparseAxisArray</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/3031">#3031</a>)</li></ul><h3 id="Fixed-18"><a class="docs-heading-anchor" href="#Fixed-18">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-18" title="Permalink"></a></h3><ul><li>Fixed a bug introduced in v1.2.0 that prevented <code>DenseAxisArray</code>s with <code>Vector</code> keys (<a href="https://github.com/jump-dev/JuMP.jl/issues/3064">#3064</a>)</li></ul><h3 id="Other-18"><a class="docs-heading-anchor" href="#Other-18">Other</a><a class="docs-heading-anchor-permalink" href="#Other-18" title="Permalink"></a></h3><ul><li>Released the JuMP logos under the CC BY 4.0 license (<a href="https://github.com/jump-dev/JuMP.jl/issues/3063">#3063</a>)</li><li>Minor tweaks to the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/3054">#3054</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3056">#3056</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3057">#3057</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3060">#3060</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3061">#3061</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3065">#3065</a>)</li><li>Improved code coverage of a number of files (<a href="https://github.com/jump-dev/JuMP.jl/issues/3048">#3048</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3049">#3049</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3050">#3050</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3051">#3051</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3052">#3052</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3053">#3053</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3058">#3058</a>) (<a href="https://github.com/jump-dev/JuMP.jl/issues/3059">#3059</a>)</li></ul><h2 id="[Version-1.2.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.1)-(August-22,-2022)"><a class="docs-heading-anchor" href="#[Version-1.2.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.1)-(August-22,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.1">Version 1.2.1</a> (August 22, 2022)</a><a id="[Version-1.2.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.1)-(August-22,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.2.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.1)-(August-22,-2022)" title="Permalink"></a></h2><h3 id="Fixed-19"><a class="docs-heading-anchor" href="#Fixed-19">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-19" title="Permalink"></a></h3><ul><li>Fixed a bug when parsing two-sided nonlinear constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/3045">#3045</a>)</li></ul><h2 id="[Version-1.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.0)-(August-16,-2022)"><a class="docs-heading-anchor" href="#[Version-1.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.0)-(August-16,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.0">Version 1.2.0</a> (August 16, 2022)</a><a id="[Version-1.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.0)-(August-16,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.0)-(August-16,-2022)" title="Permalink"></a></h2><h3 id="Breaking-2"><a class="docs-heading-anchor" href="#Breaking-2">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-2" title="Permalink"></a></h3><p>This is a large minor release because it significantly refactors the internal code for handling nonlinear programs to use the <code>MathOptInterface.Nonlinear</code> submodule that was introduced in MathOptInterface v1.3.0. As a consequence, the internal datastructure in <code>model.nlp_data</code> has been removed, as has the <code>JuMP._Derivatives</code> submodule. Despite the changes, the public API for nonlinear programming has not changed, and any code that uses only the public API and that worked with v1.1.1 will continue to work with v1.2.0.</p><h3 id="Added-14"><a class="docs-heading-anchor" href="#Added-14">Added</a><a class="docs-heading-anchor-permalink" href="#Added-14" title="Permalink"></a></h3><ul><li>Added <code>all_constraints(model; include_variable_in_set_constraints)</code> which simplifies returning a list of all constraint indices in the model.</li><li>Added the ability to delete nonlinear constraints via <code>delete(::Model, ::NonlinearConstraintRef)</code>.</li><li>Added the ability to provide an explicit Hessian for a multivariate user-defined function.</li><li>Added support for querying the primal value of a nonlinear constraint via <code>value(::NonlinearConstraintRef)</code></li></ul><h3 id="Fixed-20"><a class="docs-heading-anchor" href="#Fixed-20">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-20" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>Containers.DenseAxisArray</code> so that it now supports indexing with keys that hash to the same value, even if they are different types, for example, <code>Int32</code> and <code>Int64</code>.</li><li>Fixed a bug printing the model when the solver does not support <code>MOI.Name</code>.</li></ul><h3 id="Other-19"><a class="docs-heading-anchor" href="#Other-19">Other</a><a class="docs-heading-anchor-permalink" href="#Other-19" title="Permalink"></a></h3><ul><li>Added a constraint programming formulation to the Sudoku tutorial.</li><li>Added newly supported solvers Pajarito, Clarabel, and COPT to the installation table.</li><li>Fixed a variety of other miscellaneous issues in the documentation.</li></ul><h2 id="[Version-1.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.1)-(June-14,-2022)"><a class="docs-heading-anchor" href="#[Version-1.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.1)-(June-14,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.1">Version 1.1.1</a> (June 14, 2022)</a><a id="[Version-1.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.1)-(June-14,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.1)-(June-14,-2022)" title="Permalink"></a></h2><h3 id="Other-20"><a class="docs-heading-anchor" href="#Other-20">Other</a><a class="docs-heading-anchor-permalink" href="#Other-20" title="Permalink"></a></h3><ul><li>Fixed problem displaying LaTeX in the documentation</li><li>Minor updates to the style guide</li><li>Updated to MOI v1.4.0 in the documentation</li></ul><h2 id="[Version-1.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.0)-(May-25,-2022)"><a class="docs-heading-anchor" href="#[Version-1.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.0)-(May-25,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.0">Version 1.1.0</a> (May 25, 2022)</a><a id="[Version-1.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.0)-(May-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.0)-(May-25,-2022)" title="Permalink"></a></h2><h3 id="Added-15"><a class="docs-heading-anchor" href="#Added-15">Added</a><a class="docs-heading-anchor-permalink" href="#Added-15" title="Permalink"></a></h3><ul><li>Added <code>num_constraints(::Model; count_variable_in_set_constraints)</code> to simplify the process of counting the number of constraints in a model</li><li>Added <code>VariableRef(::ConstraintRef)</code> for querying the variable associated with a bound or integrality constraint.</li><li>Added <code>set_normalized_coefficients</code> for modifying the variable coefficients of a vector-valued constraint.</li><li>Added <code>set_string_names_on_creation</code> to disable creating <code>String</code> names for variables and constraints. This can improve performance.</li></ul><h3 id="Fixed-21"><a class="docs-heading-anchor" href="#Fixed-21">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-21" title="Permalink"></a></h3><ul><li>Fixed a bug passing <code>nothing</code> to the <code>start</code> keyword of <code>@variable</code></li></ul><h3 id="Other-21"><a class="docs-heading-anchor" href="#Other-21">Other</a><a class="docs-heading-anchor-permalink" href="#Other-21" title="Permalink"></a></h3><ul><li>New tutorials:<ul><li>Sensitivity analysis of a linear program</li><li>Serving web apps</li></ul></li><li>Minimal ellipse SDP tutorial refactored and improved</li><li>Docs updated to the latest version of each package</li><li>Lots of minor fixes and improvements to the documentation</li></ul><h2 id="[Version-1.0.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.0.0)-(March-24,-2022)"><a class="docs-heading-anchor" href="#[Version-1.0.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.0.0)-(March-24,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v1.0.0">Version 1.0.0</a> (March 24, 2022)</a><a id="[Version-1.0.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.0.0)-(March-24,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-1.0.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.0.0)-(March-24,-2022)" title="Permalink"></a></h2><p><strong>Read more about this release, along with an acknowledgement of all the contributors in our <a href="https://jump.dev/blog/1.0.0-release/">JuMP 1.0.0 is released</a> blog post.</strong></p><h3 id="Breaking-3"><a class="docs-heading-anchor" href="#Breaking-3">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-3" title="Permalink"></a></h3><ul><li>The previously deprecated functions (v0.23.0, v0.23.1) have been removed. Deprecation was to improve consistency of function names:<ul><li><code>num_nl_constraints</code> (see <code>num_nonlinear_constraints</code>)</li><li><code>all_nl_constraints</code> (see <code>all_nonlinear_constraints</code>)</li><li><code>add_NL_expression</code> (see <code>add_nonlinear_expression</code>)</li><li><code>set_NL_objective</code>  (see <code>set_nonlinear_objective</code>)</li><li><code>add_NL_constraint</code> (see <code>add_nonlinear_constraint</code>)</li><li><code>nl_expr_string</code> (see <code>nonlinear_expr_string</code>)</li><li><code>nl_constraint_string</code> (see <code>nonlinear_constraint_string</code>)</li><li><code>SymMatrixSpace</code> (see <code>SymmetricMatrixSpace</code>)</li></ul></li><li>The unintentionally exported variable <code>JuMP.op_hint</code> has been renamed to the unexported <code>JuMP._OP_HINT</code></li></ul><h3 id="Fixed-22"><a class="docs-heading-anchor" href="#Fixed-22">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-22" title="Permalink"></a></h3><ul><li>Fixed a bug writing .nl files</li><li>Fixed a bug broadcasting <code>SparseAxisArray</code>s</li></ul><h2 id="[Version-0.23.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.2)-(March-14,-2022)"><a class="docs-heading-anchor" href="#[Version-0.23.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.2)-(March-14,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.2">Version 0.23.2</a> (March 14, 2022)</a><a id="[Version-0.23.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.2)-(March-14,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.23.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.2)-(March-14,-2022)" title="Permalink"></a></h2><h3 id="Added-16"><a class="docs-heading-anchor" href="#Added-16">Added</a><a class="docs-heading-anchor-permalink" href="#Added-16" title="Permalink"></a></h3><ul><li>Added <code>relative_gap</code> to <code>solution_summary</code></li><li><code>register</code> now throws an informative error if the function is not differentiable using ForwardDiff. In some cases, the check in <code>register</code> will encounter a false negative, and the informative error will be thrown at run-time. This usually happens when the function is non-differentiable in a subset of the domain.</li></ul><h3 id="Fixed-23"><a class="docs-heading-anchor" href="#Fixed-23">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-23" title="Permalink"></a></h3><ul><li>Fixed a scoping issue when extending the <code>container</code> keyword of containers</li></ul><h3 id="Other-22"><a class="docs-heading-anchor" href="#Other-22">Other</a><a class="docs-heading-anchor-permalink" href="#Other-22" title="Permalink"></a></h3><ul><li>Docs updated to the latest version of each package</li></ul><h2 id="[Version-0.23.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.1)-(March-2,-2022)"><a class="docs-heading-anchor" href="#[Version-0.23.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.1)-(March-2,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.1">Version 0.23.1</a> (March 2, 2022)</a><a id="[Version-0.23.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.1)-(March-2,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.23.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.1)-(March-2,-2022)" title="Permalink"></a></h2><h3 id="Deprecated"><a class="docs-heading-anchor" href="#Deprecated">Deprecated</a><a id="Deprecated-1"></a><a class="docs-heading-anchor-permalink" href="#Deprecated" title="Permalink"></a></h3><ul><li><code>nl_expr_string</code> and <code>nl_constraint_string</code> have been renamed to <code>nonlinear_expr_string</code> and <code>nonlinear_constraint_string</code>. The old methods still exist with deprecation warnings. This change should impact very few users because to call them you must rely on private internals of the nonlinear API. Users are encouraged to use <code>sprint(show, x)</code> instead, where <code>x</code> is the nonlinear expression or constraint of interest.</li></ul><h3 id="Added-17"><a class="docs-heading-anchor" href="#Added-17">Added</a><a class="docs-heading-anchor-permalink" href="#Added-17" title="Permalink"></a></h3><ul><li>Added support for <code>Base.abs2(x)</code> where <code>x</code> is a variable or affine expression. This is mainly useful for complex-valued constraints.</li></ul><h3 id="Fixed-24"><a class="docs-heading-anchor" href="#Fixed-24">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-24" title="Permalink"></a></h3><ul><li>Fixed addition of complex and real affine expressions</li><li>Fixed arithmetic for Complex-valued quadratic expressions</li><li>Fixed variable bounds passed as <code>Rational{Int}(Inf)</code></li><li>Fixed printing of the coefficient <code>(0 + 1im)</code></li><li>Fixed a bug when <code>solution_summary</code> is called prior to <code>optimize!</code></li></ul><h2 id="[Version-0.23.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.0)-(February-25,-2022)"><a class="docs-heading-anchor" href="#[Version-0.23.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.0)-(February-25,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.0">Version 0.23.0</a> (February 25, 2022)</a><a id="[Version-0.23.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.0)-(February-25,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.23.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.0)-(February-25,-2022)" title="Permalink"></a></h2><p><strong>JuMP v0.23.0 is a breaking release. It is also a release-candidate for JuMP v1.0.0. That is, if no issues are found with the v0.23.0 release, then it will be re-tagged as v1.0.0.</strong></p><h3 id="Breaking-4"><a class="docs-heading-anchor" href="#Breaking-4">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-4" title="Permalink"></a></h3><ul><li>Julia 1.6 is now the minimum supported version</li><li>MathOptInterface has been updated to v1.0.0</li><li>All previously deprecated functionality has been removed</li><li><code>PrintMode</code>, <code>REPLMode</code> and <code>IJuliaMode</code> have been removed in favor of the MIME types <code>MIME&quot;text/plain&quot;</code> and <code>MIME&quot;text/latex&quot;</code>. Replace instances of <code>::Type{REPLMode}</code> with <code>::MIME&quot;text/plain&quot;</code>, <code>REPLMode</code> with <code>MIME(&quot;text/plain&quot;)</code>, <code>::Type{IJuliaMode}</code> with <code>::MIME&quot;text/latex&quot;</code>, and <code>IJuliaMode</code> with <code>MIME(&quot;text/latex&quot;)</code>.</li><li>Functions containing the <code>nl_</code> acronym have been renamed to the more explicit <code>nonlinear_</code>. For example, <code>num_nl_constraints</code> is now <code>num_nonlinear_constraints</code> and <code>set_NL_objective</code> is now <code>set_nonlinear_objective</code>. Calls to the old functions throw an error explaining the new name.</li><li><code>SymMatrixSpace</code> has been renamed to <code>SymmetricMatrixSpace</code></li></ul><h3 id="Added-18"><a class="docs-heading-anchor" href="#Added-18">Added</a><a class="docs-heading-anchor-permalink" href="#Added-18" title="Permalink"></a></h3><ul><li>Added <code>nonlinear_dual_start_value</code> and <code>set_nonlinear_dual_start_value</code></li><li>Added preliminary support for <code>Complex</code> coefficient types</li></ul><h3 id="Fixed-25"><a class="docs-heading-anchor" href="#Fixed-25">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-25" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>solution_summary</code></li></ul><h3 id="Other-23"><a class="docs-heading-anchor" href="#Other-23">Other</a><a class="docs-heading-anchor-permalink" href="#Other-23" title="Permalink"></a></h3><ul><li>MILP examples have been migrated from GLPK to HiGHS</li><li>Fixed various typos</li><li>Improved section on setting constraint start values</li></ul><h3 id="Troubleshooting-problems-when-updating"><a class="docs-heading-anchor" href="#Troubleshooting-problems-when-updating">Troubleshooting problems when updating</a><a id="Troubleshooting-problems-when-updating-1"></a><a class="docs-heading-anchor-permalink" href="#Troubleshooting-problems-when-updating" title="Permalink"></a></h3><p>If you experience problems when updating, you are likely using previously deprecated functionality. (By default, Julia does not warn when you use deprecated features.)</p><p>To find the deprecated features you are using, start Julia with <code>--depwarn=yes</code>:</p><pre><code class="nohighlight hljs">$ julia --depwarn=yes</code></pre><p>Then install JuMP v0.22.3:</p><pre><code class="language-julia hljs">julia&gt; using Pkg
+julia&gt; pkg&quot;add JuMP@0.22.3&quot;</code></pre><p>And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.</p><h2 id="[Version-0.22.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.3)-(February-10,-2022)"><a class="docs-heading-anchor" href="#[Version-0.22.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.3)-(February-10,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.3">Version 0.22.3</a> (February 10, 2022)</a><a id="[Version-0.22.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.3)-(February-10,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.22.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.3)-(February-10,-2022)" title="Permalink"></a></h2><h3 id="Fixed-26"><a class="docs-heading-anchor" href="#Fixed-26">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-26" title="Permalink"></a></h3><ul><li>Fixed a reproducibility issue in the TSP tutorial</li><li>Fixed a reproducibility issue in the <code>max_cut_sdp</code> tutorial</li><li>Fixed a bug broadcasting an empty SparseAxisArray</li></ul><h3 id="Other-24"><a class="docs-heading-anchor" href="#Other-24">Other</a><a class="docs-heading-anchor-permalink" href="#Other-24" title="Permalink"></a></h3><ul><li>Added a warning and improved documentation for the modify-then-query case</li><li>Fixed a typo in the docstring of <code>RotatedSecondOrderCone</code></li><li>Added Aqua.jl as a check for code health</li><li>Added introductions to each section of the tutorials</li><li>Improved the column generation and Benders decomposition tutorials</li><li>Updated documentation to MOI v0.10.8</li><li>Updated JuliaFormatter to v0.22.2</li></ul><h2 id="[Version-0.22.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.2)-(January-10,-2022)"><a class="docs-heading-anchor" href="#[Version-0.22.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.2)-(January-10,-2022)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.2">Version 0.22.2</a> (January 10, 2022)</a><a id="[Version-0.22.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.2)-(January-10,-2022)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.22.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.2)-(January-10,-2022)" title="Permalink"></a></h2><h3 id="Added-19"><a class="docs-heading-anchor" href="#Added-19">Added</a><a class="docs-heading-anchor-permalink" href="#Added-19" title="Permalink"></a></h3><ul><li>The function <code>all_nl_constraints</code> now returns all nonlinear constraints in a model</li><li><code>start_value</code> and <code>set_start_value</code> can now be used to get and set the primal start for constraint references</li><li>Plural macros now return a tuple containing the elements that were defined instead of <code>nothing</code></li><li>Anonymous variables are now printed as <code>_[i]</code> where <code>i</code> is the index of the variable instead of <code>noname</code>. Calling <code>name(x)</code> still returns <code>&quot;&quot;</code> so this is non-breaking.</li></ul><h3 id="Fixed-27"><a class="docs-heading-anchor" href="#Fixed-27">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-27" title="Permalink"></a></h3><ul><li>Fixed handling of <code>min</code> and <code>max</code> in nonlinear expressions</li><li>CartesianIndex is no longer allowed as a key for DenseAxisArrays.</li></ul><h3 id="Other-25"><a class="docs-heading-anchor" href="#Other-25">Other</a><a class="docs-heading-anchor-permalink" href="#Other-25" title="Permalink"></a></h3><ul><li>Improved the performance of GenericAffExpr</li><li>Added a tutorial on the Travelling Salesperson Problem</li><li>Added a tutorial on querying the Hessian of a nonlinear program</li><li>Added documentation on using custom solver binaries.</li></ul><h2 id="[Version-0.22.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.1)-(November-29,-2021)"><a class="docs-heading-anchor" href="#[Version-0.22.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.1)-(November-29,-2021)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.1">Version 0.22.1</a> (November 29, 2021)</a><a id="[Version-0.22.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.1)-(November-29,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.22.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.1)-(November-29,-2021)" title="Permalink"></a></h2><h3 id="Added-20"><a class="docs-heading-anchor" href="#Added-20">Added</a><a class="docs-heading-anchor-permalink" href="#Added-20" title="Permalink"></a></h3><ul><li>Export <code>OptimizationSense</code> enum, with instances: <code>MIN_SENSE</code>, <code>MAX_SENSE</code>, and <code>FEASIBILITY_SENSE</code></li><li>Add <code>Base.isempty(::Model)</code> to match <code>Base.empty(::Model)</code></li></ul><h3 id="Fixed-28"><a class="docs-heading-anchor" href="#Fixed-28">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-28" title="Permalink"></a></h3><ul><li>Fix bug in container with tuples as indices</li><li>Fix bug in <code>set_time_limit_sec</code></li></ul><h3 id="Other-26"><a class="docs-heading-anchor" href="#Other-26">Other</a><a class="docs-heading-anchor-permalink" href="#Other-26" title="Permalink"></a></h3><ul><li>Add tutorial &quot;Design patterns for larger models&quot;</li><li>Remove release notes section from PDF</li><li>General edits of the documentation and error messages</li></ul><h2 id="[Version-0.22.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.0)-(November-10,-2021)"><a class="docs-heading-anchor" href="#[Version-0.22.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.0)-(November-10,-2021)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.0">Version 0.22.0</a> (November 10, 2021)</a><a id="[Version-0.22.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.0)-(November-10,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.22.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.0)-(November-10,-2021)" title="Permalink"></a></h2><p><strong>JuMP v0.22 is a breaking release</strong></p><h3 id="Breaking-5"><a class="docs-heading-anchor" href="#Breaking-5">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-5" title="Permalink"></a></h3><p>JuMP 0.22 contains a number of breaking changes. However, these should be invisible for the majority of users. You will mostly encounter these breaking changes if you: wrote a JuMP extension, accessed <code>backend(model)</code>, or called <code>@SDconstraint</code>.</p><p>The breaking changes are as follows:</p><ul><li>MathOptInterface has been updated to v0.10.4. For users who have interacted with the MOI backend, this contains a large number of breaking changes. Read the <a href="https://jump.dev/MathOptInterface.jl/v0.10/release_notes/#v0.10.0-(September-6,-2021)">MathOptInterface release notes</a> for more details.</li><li>The <code>bridge_constraints</code> keyword argument to <code>Model</code> and <code>set_optimizer</code> has been renamed <code>add_bridges</code> to reflect that more thing were bridged than just constraints.</li><li>The <code>backend(model)</code> field now contains a concrete instance of a <code>MOI.Utilities.CachingOptimizer</code> instead of one with an abstractly typed optimizer field. In most cases, this will lead to improved performance. However, calling <code>set_optimizer</code> after <code>backend</code> invalidates the old backend. For example:<pre><code class="language-julia hljs">model = Model()
+b = backend(model)
+set_optimizer(model, GLPK.Optimizer)
+@variable(model, x)
+# b is not updated with `x`! Get a new b by calling `backend` again.
+new_b = backend(model)</code></pre></li><li>All usages of <code>@SDconstraint</code> are deprecated. The new syntax is <code>@constraint(model, X &gt;= Y, PSDCone())</code>.</li><li>Creating a <code>DenseAxisArray</code> with a <code>Number</code> as an axis will now display a warning. This catches a common error in which users write <code>@variable(model, x[length(S)])</code> instead of <code>@variable(model, x[1:length(S)])</code>.</li><li>The <code>caching_mode</code> argument to <code>Model</code>, for example, <code>Model(caching_mode = MOIU.MANUAL)</code> mode has been removed. For more control over the optimizer, use <code>direct_model</code> instead.</li><li>The previously deprecated <code>lp_objective_perturbation_range</code> and <code>lp_rhs_perturbation_range</code> functions have been removed. Use <code>lp_sensitivity_report</code> instead.</li><li>The <code>.m</code> fields of <code>NonlinearExpression</code> and <code>NonlinearParameter</code> have been renamed to <code>.model</code>.</li><li>Infinite variable bounds are now ignored. Thus, <code>@variable(model, x &lt;= Inf)</code> will show <code>has_upper_bound(x) == false</code>. Previously, these bounds were passed through to the solvers which caused numerical issues for solvers expecting finite bounds.</li><li>The <code>variable_type</code> and <code>constraint_type</code> functions were removed. This should only affect users who previously wrote JuMP extensions. The functions can be deleted without consequence.</li><li>The internal functions <code>moi_mode</code>, <code>moi_bridge_constraints</code>, <code>moi_add_constraint</code>, and <code>moi_add_to_function_constant</code> are no longer exported.</li><li>The un-used method <code>Containers.generate_container</code> has been deleted.</li><li>The <code>Containers</code> API has been refactored, and <code>_build_ref_sets</code> is now public as <code>Containers.build_ref_sets</code>.</li><li>The <code>parse_constraint_</code> methods for extending <code>@constraint</code> at parse time have been refactored in a breaking way. Consult the Extensions documentation for more details and examples.</li></ul><h3 id="Added-21"><a class="docs-heading-anchor" href="#Added-21">Added</a><a class="docs-heading-anchor-permalink" href="#Added-21" title="Permalink"></a></h3><ul><li>The <code>TerminationStatusCode</code> and <code>ResultStatusCode</code> enums are now exported by JuMP. Prefer <code>termination_status(model) == OPTIMAL</code> instead of <code>== MOI.OPTIMAL</code>, although the <code>MOI.</code> prefix way still works.</li><li>Copy a <code>x::DenseAxisArray</code> to an <code>Array</code> by calling <code>Array(x)</code>.</li><li><code>NonlinearExpression</code> is now a subtype of <code>AbstractJuMPScalar</code></li><li>Constraints such as <code>@constraint(model, x + 1 in MOI.Integer())</code> are now supported.</li><li><code>primal_feasibility_report</code> now accepts a function as the first argument.</li><li>Scalar variables <code>@variable(model, x[1:2] in MOI.Integer())</code> creates two variables, both of which are constrained to be in the set <code>MOI.Integer</code>.</li><li>Conic constraints can now be specified as inequalities under a different partial ordering. So <code>@constraint(model, x - y in MOI.Nonnegatives())</code> can now be written as <code>@constraint(model, x &gt;= y, MOI.Nonnegatives())</code>.</li><li>Names are now set for vectorized constraints.</li></ul><h3 id="Fixed-29"><a class="docs-heading-anchor" href="#Fixed-29">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-29" title="Permalink"></a></h3><ul><li>Fixed a performance issue when <code>show</code> was called on a <code>SparseAxisArray</code> with a large number of elements.</li><li>Fixed a bug displaying barrier and simplex iterations in <code>solution_summary</code>.</li><li>Fixed a bug by implementing <code>hash</code> for <code>DenseAxisArray</code> and <code>SparseAxisArray</code>.</li><li>Names are now only set if the solver supports them. Previously, this prevented solvers such as Ipopt from being used with <code>direct_model</code>.</li><li><code>MutableArithmetics.Zero</code> is converted into a <code>0.0</code> before being returned to the user. Previously, some calls to <code>@expression</code> would return the undocumented <code>MutableArithmetics.Zero()</code> object. One example is summing over an empty set <code>@expression(model, sum(x[i] for i in 1:0))</code>. You will now get <code>0.0</code> instead.</li><li><code>AffExpr</code> and <code>QuadExpr</code> can now be used with <code>== 0</code> instead of <code>iszero</code>. This fixes a number of issues relating to Julia standard libraries such as <code>LinearAlgebra</code> and <code>SparseArrays</code>.</li><li>Fixed a bug when registering a user-defined function with splatting.</li></ul><h3 id="Other-27"><a class="docs-heading-anchor" href="#Other-27">Other</a><a class="docs-heading-anchor-permalink" href="#Other-27" title="Permalink"></a></h3><ul><li>The documentation is now available as a PDF.</li><li>The documentation now includes a full copy of the MathOptInterface documentation to make it easy to link concepts between the docs. (The MathOptInterface documentation has also been significantly improved.)</li><li>The documentation contains a large number of improvements and clarifications on a range of topics. Thanks to @sshin23, @DilumAluthge, and @jlwether.</li><li>The documentation is now built with Julia 1.6 instead of 1.0.</li><li>Various error messages have been improved to be more readable.</li></ul><h2 id="[Version-0.21.10](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.10)-(September-4,-2021)"><a class="docs-heading-anchor" href="#[Version-0.21.10](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.10)-(September-4,-2021)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.10">Version 0.21.10</a> (September 4, 2021)</a><a id="[Version-0.21.10](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.10)-(September-4,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.10](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.10)-(September-4,-2021)" title="Permalink"></a></h2><h3 id="Added-22"><a class="docs-heading-anchor" href="#Added-22">Added</a><a class="docs-heading-anchor-permalink" href="#Added-22" title="Permalink"></a></h3><ul><li>Added <code>add_NL_expression</code></li><li><code>add_NL_xxx</code> functions now support <code>AffExpr</code> and <code>QuadExpr</code> as terms</li></ul><h3 id="Fixed-30"><a class="docs-heading-anchor" href="#Fixed-30">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-30" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>solution_summary</code></li><li>Fixed a bug in <code>relax_integrality</code></li></ul><h3 id="Other-28"><a class="docs-heading-anchor" href="#Other-28">Other</a><a class="docs-heading-anchor-permalink" href="#Other-28" title="Permalink"></a></h3><ul><li>Improved error message in <code>lp_sensitivity_report</code></li></ul><h2 id="[Version-0.21.9](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.9)-(August-1,-2021)"><a class="docs-heading-anchor" href="#[Version-0.21.9](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.9)-(August-1,-2021)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.9">Version 0.21.9</a> (August 1, 2021)</a><a id="[Version-0.21.9](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.9)-(August-1,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.9](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.9)-(August-1,-2021)" title="Permalink"></a></h2><h3 id="Added-23"><a class="docs-heading-anchor" href="#Added-23">Added</a><a class="docs-heading-anchor-permalink" href="#Added-23" title="Permalink"></a></h3><ul><li>Containers now support arbitrary container types by passing the type to the <code>container</code> keyword and overloading <code>Containers.container</code>.</li><li><code>is_valid</code> now supports nonlinear constraints</li><li>Added <code>unsafe_backend</code> for querying the inner-most optimizer of a JuMP model.</li><li>Nonlinear parameters now support the plural <code>@NLparameters</code> macro.</li><li>Containers (for example, <code>DenseAxisArray</code>) can now be used in vector-valued constraints.</li></ul><h3 id="Other-29"><a class="docs-heading-anchor" href="#Other-29">Other</a><a class="docs-heading-anchor-permalink" href="#Other-29" title="Permalink"></a></h3><ul><li>Various improvements to the documentation.</li></ul><h2 id="[Version-0.21.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.8)-(May-8,-2021)"><a class="docs-heading-anchor" href="#[Version-0.21.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.8)-(May-8,-2021)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.8">Version 0.21.8</a> (May 8, 2021)</a><a id="[Version-0.21.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.8)-(May-8,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.8)-(May-8,-2021)" title="Permalink"></a></h2><h3 id="Added-24"><a class="docs-heading-anchor" href="#Added-24">Added</a><a class="docs-heading-anchor-permalink" href="#Added-24" title="Permalink"></a></h3><ul><li>The <code>@constraint</code> macro is now extendable in the same way as <code>@variable</code>.</li><li><code>AffExpr</code> and <code>QuadExpr</code> can now be used in nonlinear macros.</li></ul><h3 id="Fixed-31"><a class="docs-heading-anchor" href="#Fixed-31">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-31" title="Permalink"></a></h3><ul><li>Fixed a bug in <code>lp_sensitivity_report</code>.</li><li>Fixed an inference issue when creating empty <code>SparseAxisArray</code>s.</li></ul><h2 id="[Version-0.21.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.7)-(April-12,-2021)"><a class="docs-heading-anchor" href="#[Version-0.21.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.7)-(April-12,-2021)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.7">Version 0.21.7</a> (April 12, 2021)</a><a id="[Version-0.21.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.7)-(April-12,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.7)-(April-12,-2021)" title="Permalink"></a></h2><h3 id="Added-25"><a class="docs-heading-anchor" href="#Added-25">Added</a><a class="docs-heading-anchor-permalink" href="#Added-25" title="Permalink"></a></h3><ul><li>Added <code>primal_feasibility_report</code>, which can be used to check whether a primal point satisfies primal feasibility.</li><li>Added <code>coefficient</code>, which returns the coefficient associated with a variable in affine and quadratic expressions.</li><li>Added <code>copy_conflict</code>, which returns the IIS of an infeasible model.</li><li>Added <code>solution_summary</code>, which returns (and prints) a struct containing a summary of the solution.</li><li>Allow <code>AbstractVector</code> in vector constraints instead of just <code>Vector</code>.</li><li>Added <code>latex_formulation(model)</code> which returns an object representing the latex formulation of a model. Use <code>print(latex_formulation(model))</code> to print the formulation as a string.</li><li>User-defined functions in nonlinear expressions are now automatically registered to aid quick model prototyping. However, a warning is printed to encourage the manual registration.</li><li>DenseAxisArray&#39;s now support broadcasting over multiple arrays.</li><li>Container indices can now be iterators of <code>Base.SizeUnknown</code>.</li></ul><h3 id="Fixed-32"><a class="docs-heading-anchor" href="#Fixed-32">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-32" title="Permalink"></a></h3><ul><li>Fixed bug in <code>rad2deg</code> and <code>deg2rad</code> in nonlinear expressions.</li><li>Fixed a MethodError bug in <code>Containers</code> when forcing container type.</li><li>Allow partial slicing of a DenseAxisArray, resolving an issue from 2014.</li><li>Fixed a bug printing variable names in IJulia.</li><li>Ending an IJulia cell with <code>model</code> now prints a summary of the model (like in the REPL) not the latex formulation. Use <code>print(model)</code> to print the latex formulation.</li><li>Fixed a bug when copying models containing nested arrays.</li></ul><h3 id="Other-30"><a class="docs-heading-anchor" href="#Other-30">Other</a><a class="docs-heading-anchor-permalink" href="#Other-30" title="Permalink"></a></h3><ul><li>Tutorials are now part of the documentation, and more refactoring has taken place.</li><li>Added JuliaFormatter added as a code formatter.</li><li>Added some precompilation statements to reduce initial latency.</li><li>Various improvements to error messages to make them more helpful.</li><li>Improved performance of <code>value(::NonlinearExpression)</code>.</li><li>Improved performance of <code>fix(::VariableRef)</code>.</li></ul><h2 id="[Version-0.21.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.6)-(January-29,-2021)"><a class="docs-heading-anchor" href="#[Version-0.21.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.6)-(January-29,-2021)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.6">Version 0.21.6</a> (January 29, 2021)</a><a id="[Version-0.21.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.6)-(January-29,-2021)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.6)-(January-29,-2021)" title="Permalink"></a></h2><h3 id="Added-26"><a class="docs-heading-anchor" href="#Added-26">Added</a><a class="docs-heading-anchor-permalink" href="#Added-26" title="Permalink"></a></h3><ul><li>Added support for skew symmetric variables via <code>@variable(model, X[1:2, 1:2] in SkewSymmetricMatrixSpace())</code>.</li><li><code>lp_sensitivity_report</code> has been added which significantly improves the performance of querying the sensitivity summary of an LP. <code>lp_objective_perturbation_range</code> and <code>lp_rhs_perturbation_range</code> are deprecated.</li><li>Dual warm-starts are now supported with <code>set_dual_start_value</code> and <code>dual_start_value</code>.</li><li><code>∈</code> (<code>\in&lt;tab&gt;</code>) can now be used in macros instead of <code>=</code> or <code>in</code>.</li><li>Use <code>haskey(model::Model, key::Symbol)</code> to check if a name <code>key</code> is registered in a model.</li><li>Added <code>unregister(model::Model, key::Symbol)</code> to unregister a name <code>key</code> from <code>model</code>.</li><li>Added <code>callback_node_status</code> for use in callbacks.</li><li>Added <code>print_bridge_graph</code> to visualize the bridging graph generated by MathOptInterface.</li><li>Improved error message for containers with duplicate indices.</li></ul><h3 id="Fixed-33"><a class="docs-heading-anchor" href="#Fixed-33">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-33" title="Permalink"></a></h3><ul><li>Various fixes to pass tests on Julia 1.6.</li><li>Fixed a bug in the printing of nonlinear expressions in IJulia.</li><li>Fixed a bug when nonlinear expressions are passed to user-defined functions.</li><li>Some internal functions that were previously exported are now no longer exported.</li><li>Fixed a bug when relaxing a fixed binary variable.</li><li>Fixed a <code>StackOverflowError</code> that occurred when <code>SparseAxisArray</code>s had a large number of elements.</li><li>Removed an unnecessary type assertion in <code>list_of_constraint_types</code>.</li><li>Fixed a bug when copying models with registered expressions.</li></ul><h3 id="Other-31"><a class="docs-heading-anchor" href="#Other-31">Other</a><a class="docs-heading-anchor-permalink" href="#Other-31" title="Permalink"></a></h3><ul><li>The documentation has been significantly overhauled. It now has distinct sections for the manual, API reference, and examples. The existing examples in <code>/examples</code> have now been moved to <code>/docs/src/examples</code> and rewritten using <code>Literate.jl</code>, and they are now included in the documentation.</li><li>JuliaFormatter has been applied to most of the codebase. This will continue to roll out over time, as we fix upstream issues in the formatter, and will eventually become compulsory.</li><li>The root cause of a large number of method invalidations has been resolved.</li><li>We switched continuous integration from Travis and Appveyor to GitHub Actions.</li></ul><h2 id="[Version-0.21.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.5)-(September-18,-2020)"><a class="docs-heading-anchor" href="#[Version-0.21.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.5)-(September-18,-2020)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.5">Version 0.21.5</a> (September 18, 2020)</a><a id="[Version-0.21.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.5)-(September-18,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.5)-(September-18,-2020)" title="Permalink"></a></h2><h3 id="Fixed-34"><a class="docs-heading-anchor" href="#Fixed-34">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-34" title="Permalink"></a></h3><ul><li>Fixed deprecation warnings</li><li>Throw <code>DimensionMismatch</code> for incompatibly sized functions and sets</li><li>Unify treatment of <code>keys(x)</code> on JuMP containers</li></ul><h2 id="[Version-0.21.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.4)-(September-14,-2020)"><a class="docs-heading-anchor" href="#[Version-0.21.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.4)-(September-14,-2020)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.4">Version 0.21.4</a> (September 14, 2020)</a><a id="[Version-0.21.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.4)-(September-14,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.4)-(September-14,-2020)" title="Permalink"></a></h2><h3 id="Added-27"><a class="docs-heading-anchor" href="#Added-27">Added</a><a class="docs-heading-anchor-permalink" href="#Added-27" title="Permalink"></a></h3><ul><li>Add debug info when adding unsupported constraints</li><li>Add <code>relax_integrality</code> for solving continuous relaxation</li><li>Allow querying constraint conflicts</li></ul><h3 id="Fixed-35"><a class="docs-heading-anchor" href="#Fixed-35">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-35" title="Permalink"></a></h3><ul><li>Dispatch on <code>Real</code> for <code>MOI.submit</code></li><li>Implement <code>copy</code> for <code>CustomSet</code> in tests</li><li>Don&#39;t export private macros</li><li>Fix invalid assertion in nonlinear</li><li>Error if constraint has <code>NaN</code> right-hand side</li><li>Improve speed of tests</li><li>Lots of work modularizing files in <code>/test</code></li><li>Improve line numbers in macro error messages</li><li>Print nonlinear subexpressions</li><li>Various documentation updates</li><li>Dependency updates:<ul><li>Datastructures 0.18</li><li>MathOptFormat v0.5</li><li>Prep for MathOptInterface 0.9.15</li></ul></li></ul><h2 id="[Version-0.21.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.3)-(June-18,-2020)"><a class="docs-heading-anchor" href="#[Version-0.21.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.3)-(June-18,-2020)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.3">Version 0.21.3</a> (June 18, 2020)</a><a id="[Version-0.21.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.3)-(June-18,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.3)-(June-18,-2020)" title="Permalink"></a></h2><ul><li>Added Special Order Sets (SOS1 and SOS2) to JuMP with default weights to ease the creation of such constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/2212">#2212</a>).</li><li>Added functions <code>simplex_iterations</code>, <code>barrier_iterations</code> and <code>node_count</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2201">#2201</a>).</li><li>Added function <code>reduced_cost</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2205">#2205</a>).</li><li>Implemented <code>callback_value</code> for affine and quadratic expressions (<a href="https://github.com/jump-dev/JuMP.jl/issues/2231">#2231</a>).</li><li>Support <code>MutableArithmetics.Zero</code> in objective and constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/2219">#2219</a>).</li><li>Documentation improvements:<ul><li>Mention tutorials in the docs (<a href="https://github.com/jump-dev/JuMP.jl/issues/2223">#2223</a>).</li><li>Update COIN-OR links (<a href="https://github.com/jump-dev/JuMP.jl/issues/2242">#2242</a>).</li><li>Explicit link to the documentation of <code>MOI.FileFormats</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2253">#2253</a>).</li><li>Typo fixes (<a href="https://github.com/jump-dev/JuMP.jl/issues/2261">#2261</a>).</li></ul></li><li>Containers improvements:<ul><li>Fix <code>Base.map</code> for <code>DenseAxisArray</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2235">#2235</a>).</li><li>Throw <code>BoundsError</code> if number of indices is incorrect for <code>DenseAxisArray</code> and <code>SparseAxisArray</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2240">#2240</a>).</li></ul></li><li>Extensibility improvements:<ul><li>Implement a <code>set_objective</code> method fallback that redirects to <code>set_objective_sense</code> and <code>set_objective_function</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2247">#2247</a>).</li><li>Add <code>parse_constraint</code> method with arbitrary number of arguments (<a href="https://github.com/jump-dev/JuMP.jl/issues/2051">#2051</a>).</li><li>Add <code>parse_constraint_expr</code> and <code>parse_constraint_head</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2228">#2228</a>).</li></ul></li></ul><h2 id="[Version-0.21.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.2)-(April-2,-2020)"><a class="docs-heading-anchor" href="#[Version-0.21.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.2)-(April-2,-2020)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.2">Version 0.21.2</a> (April 2, 2020)</a><a id="[Version-0.21.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.2)-(April-2,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.2)-(April-2,-2020)" title="Permalink"></a></h2><ul><li>Added <code>relative_gap()</code> to access <code>MOI.RelativeGap()</code> attribute (<a href="https://github.com/jump-dev/JuMP.jl/issues/2199">#2199</a>).</li><li>Documentation fixes:<ul><li>Added link to source for docstrings in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/2207">#2207</a>).</li><li>Added docstring for <code>@variables</code> macro (<a href="https://github.com/jump-dev/JuMP.jl/issues/2216">#2216</a>).</li><li>Typo fixes (#2177, #2184, #2182).</li></ul></li><li>Implementation of methods for Base functions:<ul><li>Implemented <code>Base.empty!</code> for <code>JuMP.Model</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2198">#2198</a>).</li><li>Implemented <code>Base.conj</code> for JuMP scalar types (<a href="https://github.com/jump-dev/JuMP.jl/issues/2209">#2209</a>).</li></ul></li></ul><h3 id="Fixed-36"><a class="docs-heading-anchor" href="#Fixed-36">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-36" title="Permalink"></a></h3><ul><li>Fixed sum of expression with scalar product in macro (<a href="https://github.com/jump-dev/JuMP.jl/issues/2178">#2178</a>).</li><li>Fixed writing of nonlinear models to MathOptFormat (<a href="https://github.com/jump-dev/JuMP.jl/issues/2181">#2181</a>).</li><li>Fixed construction of empty SparseAxisArray (<a href="https://github.com/jump-dev/JuMP.jl/issues/2179">#2179</a>).</li><li>Fixed constraint with zero function (<a href="https://github.com/jump-dev/JuMP.jl/issues/2188">#2188</a>).</li></ul><h2 id="[Version-0.21.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.1)-(Feb-18,-2020)"><a class="docs-heading-anchor" href="#[Version-0.21.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.1)-(Feb-18,-2020)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.1">Version 0.21.1</a> (Feb 18, 2020)</a><a id="[Version-0.21.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.1)-(Feb-18,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.1)-(Feb-18,-2020)" title="Permalink"></a></h2><ul><li>Improved the clarity of the <code>with_optimizer</code> deprecation warning.</li></ul><h2 id="[Version-0.21.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.0)-(Feb-16,-2020)"><a class="docs-heading-anchor" href="#[Version-0.21.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.0)-(Feb-16,-2020)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.0">Version 0.21.0</a> (Feb 16, 2020)</a><a id="[Version-0.21.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.0)-(Feb-16,-2020)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.21.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.0)-(Feb-16,-2020)" title="Permalink"></a></h2><h3 id="Breaking-6"><a class="docs-heading-anchor" href="#Breaking-6">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-6" title="Permalink"></a></h3><ul><li><p>Deprecated <code>with_optimizer</code> (#2090, #2084, #2141). You can replace <code>with_optimizer</code> by either nothing, <code>optimizer_with_attributes</code> or a closure:</p><ul><li>replace <code>with_optimizer(Ipopt.Optimizer)</code> by <code>Ipopt.Optimizer</code>.</li><li>replace <code>with_optimizer(Ipopt.Optimizer, max_cpu_time=60.0)</code> by <code>optimizer_with_attributes(Ipopt.Optimizer, &quot;max_cpu_time&quot; =&gt; 60.0)</code>.</li><li>replace <code>with_optimizer(Gurobi.Optimizer, env)</code> by <code>() -&gt; Gurobi.Optimizer(env)</code>.</li><li>replace <code>with_optimizer(Gurobi.Optimizer, env, Presolve=0)</code> by <code>optimizer_with_attributes(() -&gt; Gurobi.Optimizer(env), &quot;Presolve&quot; =&gt; 0)</code>.</li></ul><p>alternatively to <code>optimizer_with_attributes</code>, you can also set the attributes separately with <code>set_optimizer_attribute</code>.</p></li><li><p>Renamed <code>set_parameter</code> and <code>set_parameters</code> to <code>set_optimizer_attribute</code> and <code>set_optimizer_attributes</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2150">#2150</a>).</p></li><li><p>Broadcast should now be explicit inside macros. <code>@SDconstraint(model, x &gt;= 1)</code> and <code>@constraint(model, x + 1 in SecondOrderCone())</code> now throw an error instead of broadcasting <code>1</code> along the dimension of <code>x</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2107">#2107</a>).</p></li><li><p><code>@SDconstraint(model, x &gt;= 0)</code> is now equivalent to <code>@constraint(model, x in PSDCone())</code> instead of <code>@constraint(model, (x .- 0) in PSDCone())</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/2107">#2107</a>).</p></li><li><p>The macros now create the containers with <code>map</code> instead of <code>for</code> loops, as a consequence, containers created by <code>@expression</code> can now have any element type and containers of constraint references now have concrete element types when possible. This fixes a long-standing issue where <code>@expression</code> could only be used to generate a collection of linear expressions. Now it works for quadratic expressions as well (<a href="https://github.com/jump-dev/JuMP.jl/issues/2070">#2070</a>).</p></li><li><p>Calling <code>deepcopy(::AbstractModel)</code> now throws an error.</p></li><li><p>The constraint name is now printed in the model string (<a href="https://github.com/jump-dev/JuMP.jl/issues/2108">#2108</a>).</p></li></ul><h3 id="Added-28"><a class="docs-heading-anchor" href="#Added-28">Added</a><a class="docs-heading-anchor-permalink" href="#Added-28" title="Permalink"></a></h3><ul><li>Added support for solver-independent and solver-specific callbacks (<a href="https://github.com/jump-dev/JuMP.jl/issues/2101">#2101</a>).</li><li>Added <code>write_to_file</code> and <code>read_from_file</code>, supported formats are CBF, LP, MathOptFormat, MPS and SDPA (<a href="https://github.com/jump-dev/JuMP.jl/issues/2114">#2114</a>).</li><li>Added support for complementarity constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/2132">#2132</a>).</li><li>Added support for indicator constraints (<a href="https://github.com/jump-dev/JuMP.jl/issues/2092">#2092</a>).</li><li>Added support for querying multiple solutions with the <code>result</code> keyword (<a href="https://github.com/jump-dev/JuMP.jl/issues/2100">#2100</a>).</li><li>Added support for constraining variables on creation (<a href="https://github.com/jump-dev/JuMP.jl/issues/2128">#2128</a>).</li><li>Added method <code>delete</code> that deletes a vector of variables at once if it is supported by the underlying solver (<a href="https://github.com/jump-dev/JuMP.jl/issues/2135">#2135</a>).</li><li>The arithmetic between JuMP expression has be refactored into the MutableArithmetics package (<a href="https://github.com/jump-dev/JuMP.jl/issues/2107">#2107</a>).</li><li>Improved error on complex values in NLP (<a href="https://github.com/jump-dev/JuMP.jl/issues/1978">#1978</a>).</li><li>Added an example of column generation (<a href="https://github.com/jump-dev/JuMP.jl/issues/2010">#2010</a>).</li></ul><h3 id="Fixed-37"><a class="docs-heading-anchor" href="#Fixed-37">Fixed</a><a class="docs-heading-anchor-permalink" href="#Fixed-37" title="Permalink"></a></h3><ul><li>Incorrect coefficients generated when using Symmetric variables (<a href="https://github.com/jump-dev/JuMP.jl/issues/2102">#2102</a>)</li></ul><h2 id="[Version-0.20.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.1)-(Oct-18,-2019)"><a class="docs-heading-anchor" href="#[Version-0.20.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.1)-(Oct-18,-2019)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.1">Version 0.20.1</a> (Oct 18, 2019)</a><a id="[Version-0.20.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.1)-(Oct-18,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.20.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.1)-(Oct-18,-2019)" title="Permalink"></a></h2><ul><li>Add sections on <code>@variables</code> and <code>@constraints</code> in the documentation (<a href="https://github.com/jump-dev/JuMP.jl/issues/2062">#2062</a>).</li><li>Fixed product of sparse matrices for Julia v1.3 (<a href="https://github.com/jump-dev/JuMP.jl/issues/2063">#2063</a>).</li><li>Added <code>set_objective_coefficient</code> to modify the coefficient of a linear term of the objective function (<a href="https://github.com/jump-dev/JuMP.jl/issues/2008">#2008</a>).</li><li>Added <code>set_time_limit_sec</code>, <code>unset_time_limit_sec</code> and <code>time_limit_sec</code> to set and query the time limit for the solver in seconds (<a href="https://github.com/jump-dev/JuMP.jl/issues/2053">#2053</a>).</li></ul><h2 id="[Version-0.20.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.0)-(Aug-24,-2019)"><a class="docs-heading-anchor" href="#[Version-0.20.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.0)-(Aug-24,-2019)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.0">Version 0.20.0</a> (Aug 24, 2019)</a><a id="[Version-0.20.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.0)-(Aug-24,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.20.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.0)-(Aug-24,-2019)" title="Permalink"></a></h2><ul><li>Documentation updates.</li><li>Numerous bug fixes.</li><li>Better error messages (#1977, #1978, #1997, #2017).</li><li>Performance improvements (#1947, #2032).</li><li>Added LP sensitivity summary functions <code>lp_objective_perturbation_range</code> and <code>lp_rhs_perturbation_range</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/1917">#1917</a>).</li><li>Added functions <code>dual_objective_value</code>, <code>raw_status</code> and <code>set_parameter</code>.</li><li>Added function <code>set_objective_coefficient</code> to modify the coefficient of a linear term of the objective (<a href="https://github.com/jump-dev/JuMP.jl/issues/2008">#2008</a>).</li><li>Added functions <code>set_normalized_rhs</code>, <code>normalized_rhs</code>, and <code>add_to_function_constant</code> to modify and get the constant part of a constraint (#1935, #1960).</li><li>Added functions <code>set_normalized_coefficient</code> and <code>normalized_coefficient</code> to modify and get the coefficient of a linear term of a constraint (#1935, #1960).</li><li>Numerous other improvements in MOI 0.9, see the <code>NEWS.md</code> file of MOI for more details.</li></ul><h2 id="[Version-0.19.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.2)-(June-8,-2019)"><a class="docs-heading-anchor" href="#[Version-0.19.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.2)-(June-8,-2019)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.2">Version 0.19.2</a> (June 8, 2019)</a><a id="[Version-0.19.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.2)-(June-8,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.19.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.2)-(June-8,-2019)" title="Permalink"></a></h2><ul><li>Fix a bug in derivatives that could arise in models with nested nonlinear subexpressions.</li></ul><h2 id="[Version-0.19.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.1)-(May-12,-2019)"><a class="docs-heading-anchor" href="#[Version-0.19.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.1)-(May-12,-2019)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.1">Version 0.19.1</a> (May 12, 2019)</a><a id="[Version-0.19.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.1)-(May-12,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.19.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.1)-(May-12,-2019)" title="Permalink"></a></h2><ul><li>Usability and performance improvements.</li><li>Bug fixes.</li></ul><h2 id="[Version-0.19.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.0)-(February-15,-2019)"><a class="docs-heading-anchor" href="#[Version-0.19.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.0)-(February-15,-2019)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.0">Version 0.19.0</a> (February 15, 2019)</a><a id="[Version-0.19.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.0)-(February-15,-2019)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.19.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.0)-(February-15,-2019)" title="Permalink"></a></h2><p><strong>JuMP 0.19 contains significant breaking changes.</strong></p><h3 id="Breaking-7"><a class="docs-heading-anchor" href="#Breaking-7">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-7" title="Permalink"></a></h3><ul><li><p>JuMP&#39;s abstraction layer for communicating with solvers changed from <a href="https://github.com/JuliaOpt/MathProgBase.jl">MathProgBase</a> (MPB) to <a href="https://github.com/JuliaOpt/MathOptInterface.jl">MathOptInterface</a> (MOI). MOI addresses many longstanding design issues. (See @mlubin&#39;s <a href="https://www.juliaopt.org/meetings/bordeaux2018/lubin.pdf">slides</a> from JuMP-dev 2018.) JuMP 0.19 is compatible only with solvers that have been updated for MOI. See the <a href="https://www.juliaopt.org/JuMP.jl/dev/installation/">installation guide</a> for a list of solvers that have and have not yet been updated.</p></li><li><p>Most solvers have been renamed to <code>PackageName.Optimizer</code>. For example, <code>GurobiSolver()</code> is now <code>Gurobi.Optimizer</code>.</p></li><li><p>Solvers are no longer added to a model via <code>Model(solver = XXX(kwargs...))</code>. Instead use <code>Model(with_optimizer(XXX, kwargs...))</code>. For example, <code>Model(with_optimizer(Gurobi.Optimizer, OutputFlag=0))</code>.</p></li><li><p>JuMP containers (for example, the objects returned by <code>@variable</code>) have been redesigned. <code>Containers.SparseAxisArray</code> replaces <code>JuMPDict</code>, <code>JuMPArray</code> was rewritten (inspired by <code>AxisArrays</code>) and renamed <code>Containers.DenseAxisArray</code>, and you can now request a container type with the <code>container=</code> keyword to the macros. See the corresponding <a href="https://www.juliaopt.org/JuMP.jl/dev/variables/#Variable-containers-1">documentation</a> for more details.</p></li><li><p>The statuses returned by solvers have changed. See the possible status values <a href="https://www.juliaopt.org/MathOptInterface.jl/stable/apireference.html#Termination-Status-1">here</a>. The MOI statuses are much richer than the MPB statuses and can be used to distinguish between previously indistinguishable cases (for example, did the solver have a feasible solution when it stopped because of the time limit?).</p></li><li><p>Starting values are separate from result values. Use <code>value</code> to query the value of a variable in a solution. Use <code>start_value</code> and <code>set_start_value</code> to get and set an initial starting point provided to the solver. The solutions from previous solves are no longer automatically set as the starting points for the next solve.</p></li><li><p>The data structures for affine and quadratic expressions <code>AffExpr</code> and <code>QuadExpr</code> have changed. Internally, terms are stored in dictionaries instead of lists. Duplicate coefficients can no longer exist. Accessors and iteration methods have changed.</p></li><li><p><code>JuMPNLPEvaluator</code> no longer includes the linear and quadratic parts of the model in the evaluation calls. These are now handled separately to allow NLP solvers that support various types of constraints.</p></li><li><p>JuMP solver-independent callbacks have been replaced by solver-specific callbacks. See your favorite solver for more details. (See the note below: No solver-specific callbacks are implemented yet.)</p></li><li><p>The <code>norm()</code> syntax is no longer recognized inside macros. Use the <code>SecondOrderCone()</code> set instead.</p></li><li><p>JuMP no longer performs automatic transformation between special quadratic forms and second-order cone constraints. Support for these constraint classes depends on the solver.</p></li><li><p>The symbols <code>:Min</code> and <code>:Max</code> are no longer used as optimization senses. Instead, JuMP uses the <code>OptimizationSense</code> enum from MathOptInterface. <code>@objective(model, Max, ...)</code>, <code>@objective(model, Min, ...)</code>, <code>@NLobjective(model, Max, ...)</code>, and <code>@objective(model, Min, ...)</code> remain valid, but <code>@objective(m, :Max, ...)</code> is no longer accepted.</p></li><li><p>The sign conventions for duals has changed in some cases for consistency with conic duality (see the <a href="https://www.juliaopt.org/MathOptInterface.jl/v0.6.2/apimanual.html#Duals-1">documentation</a>). The <code>shadow_price</code> helper method returns duals with signs that match conventional LP interpretations of dual values as sensitivities of the objective value to relaxations of constraints.</p></li><li><p><code>@constraintref</code> is no longer defined. Instead, create the appropriate container to hold constraint references manually. For example,</p><pre><code class="language-julia hljs">constraints = Dict() # Optionally, specify types for improved performance.
+for i in 1:N
+  constraints[i] = @constraint(model, ...)
+end</code></pre></li><li><p>The <code>lowerbound</code>, <code>upperbound</code>, and <code>basename</code> keyword arguments to the <code>@variable</code> macro have been renamed to <code>lower_bound</code>, <code>upper_bound</code>, and <code>base_name</code>, for consistency with JuMP&#39;s new <a href="https://www.juliaopt.org/JuMP.jl/dev/style/">style recommendations</a>.</p></li><li><p>We rely on broadcasting syntax to apply accessors to collections of variables, for example, <code>value.(x)</code> instead of <code>getvalue(x)</code> for collections. (Use <code>value(x)</code> when <code>x</code> is a scalar object.)</p></li></ul><h3 id="Added-29"><a class="docs-heading-anchor" href="#Added-29">Added</a><a class="docs-heading-anchor-permalink" href="#Added-29" title="Permalink"></a></h3><ul><li><p>Splatting (like <code>f(x...)</code>) is recognized in restricted settings in nonlinear expressions.</p></li><li><p>Support for deleting constraints and variables.</p></li><li><p>The documentation has been completely rewritten using docstrings and Documenter.</p></li><li><p>Support for modeling mixed conic and quadratic models (for example, conic models with quadratic objectives and bi-linear matrix inequalities).</p></li><li><p>Significantly improved support for modeling new types of constraints and for extending JuMP&#39;s macros.</p></li><li><p>Support for providing dual warm starts.</p></li><li><p>Improved support for accessing solver-specific attributes (for example, the irreducible inconsistent subsystem).</p></li><li><p>Explicit control of whether symmetry-enforcing constraints are added to PSD constraints.</p></li><li><p>Support for modeling exponential cones.</p></li><li><p>Significant improvements in internal code quality and testing.</p></li><li><p>Style and naming guidelines.</p></li><li><p>Direct mode and manual mode provide explicit control over when copies of a model are stored or regenerated. See the corresponding <a href="https://www.juliaopt.org/JuMP.jl/dev/solvers/">documentation</a>.</p></li></ul><h3 id="Regressions"><a class="docs-heading-anchor" href="#Regressions">Regressions</a><a id="Regressions-1"></a><a class="docs-heading-anchor-permalink" href="#Regressions" title="Permalink"></a></h3><p>There are known regressions from JuMP 0.18 that will be addressed in a future release (0.19.x or later):</p><ul><li><p>Performance regressions in model generation (<a href="https://github.com/JuliaOpt/JuMP.jl/issues/1403">issue</a>). Please file an issue anyway if you notice a significant performance regression. We have plans to address a number of performance issues, but we might not be aware of all of them.</p></li><li><p>Fast incremental NLP solves are not yet reimplemented (<a href="https://github.com/JuliaOpt/JuMP.jl/issues/1185">issue</a>).</p></li><li><p>We do not yet have an implementation of solver-specific callbacks.</p></li><li><p>The column generation syntax in <code>@variable</code> has been removed (that is, the <code>objective</code>, <code>coefficients</code>, and <code>inconstraints</code> keyword arguments). Support for column generation will be re-introduced in a future release.</p></li><li><p>The ability to solve the continuous relaxation (that is, via <code>solve(model; relaxation = true)</code>) is not yet reimplemented (<a href="https://github.com/JuliaOpt/JuMP.jl/issues/1611">issue</a>).</p></li></ul><h2 id="[Version-0.18.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.5)-(December-1,-2018)"><a class="docs-heading-anchor" href="#[Version-0.18.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.5)-(December-1,-2018)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.5">Version 0.18.5</a> (December 1, 2018)</a><a id="[Version-0.18.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.5)-(December-1,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.18.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.5)-(December-1,-2018)" title="Permalink"></a></h2><ul><li>Support views in some derivative evaluation functions.</li><li>Improved compatibility with PackageCompiler.</li></ul><h2 id="[Version-0.18.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.4)-(October-8,-2018)"><a class="docs-heading-anchor" href="#[Version-0.18.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.4)-(October-8,-2018)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.4">Version 0.18.4</a> (October 8, 2018)</a><a id="[Version-0.18.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.4)-(October-8,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.18.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.4)-(October-8,-2018)" title="Permalink"></a></h2><ul><li>Fix a bug in model printing on Julia 0.7 and 1.0.</li></ul><h2 id="[Version-0.18.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.3)-(October-1,-2018)"><a class="docs-heading-anchor" href="#[Version-0.18.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.3)-(October-1,-2018)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.3">Version 0.18.3</a> (October 1, 2018)</a><a id="[Version-0.18.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.3)-(October-1,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.18.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.3)-(October-1,-2018)" title="Permalink"></a></h2><ul><li>Add support for Julia v1.0 (Thanks @ExpandingMan)</li><li>Fix matrix expressions with quadratic functions (<a href="https://github.com/jump-dev/JuMP.jl/issues/1508">#1508</a>)</li></ul><h2 id="[Version-0.18.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.2)-(June-10,-2018)"><a class="docs-heading-anchor" href="#[Version-0.18.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.2)-(June-10,-2018)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.2">Version 0.18.2</a> (June 10, 2018)</a><a id="[Version-0.18.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.2)-(June-10,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.18.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.2)-(June-10,-2018)" title="Permalink"></a></h2><ul><li>Fix a bug in second-order derivatives when expressions are present (<a href="https://github.com/jump-dev/JuMP.jl/issues/1319">#1319</a>)</li><li>Fix a bug in <code>@constraintref</code> (<a href="https://github.com/jump-dev/JuMP.jl/issues/1330">#1330</a>)</li></ul><h2 id="[Version-0.18.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.1)-(April-9,-2018)"><a class="docs-heading-anchor" href="#[Version-0.18.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.1)-(April-9,-2018)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.1">Version 0.18.1</a> (April 9, 2018)</a><a id="[Version-0.18.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.1)-(April-9,-2018)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.18.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.1)-(April-9,-2018)" title="Permalink"></a></h2><ul><li>Fix for nested tuple destructuring (<a href="https://github.com/jump-dev/JuMP.jl/issues/1193">#1193</a>)</li><li>Preserve internal model when relaxation=true (<a href="https://github.com/jump-dev/JuMP.jl/issues/1209">#1209</a>)</li><li>Minor bug fixes and updates for example</li></ul><h2 id="[Version-0.18.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.0)-(July-27,-2017)"><a class="docs-heading-anchor" href="#[Version-0.18.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.0)-(July-27,-2017)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.0">Version 0.18.0</a> (July 27, 2017)</a><a id="[Version-0.18.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.0)-(July-27,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.18.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.0)-(July-27,-2017)" title="Permalink"></a></h2><ul><li>Drop support for Julia 0.5.</li><li>Update for ForwardDiff 0.5.</li><li>Minor bug fixes.</li></ul><h2 id="[Version-0.17.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.1)-(June-9,-2017)"><a class="docs-heading-anchor" href="#[Version-0.17.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.1)-(June-9,-2017)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.1">Version 0.17.1</a> (June 9, 2017)</a><a id="[Version-0.17.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.1)-(June-9,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.17.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.1)-(June-9,-2017)" title="Permalink"></a></h2><ul><li>Use of <code>constructconstraint!</code> in <code>@SDconstraint</code>.</li><li>Minor bug fixes.</li></ul><h2 id="[Version-0.17.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.0)-(May-27,-2017)"><a class="docs-heading-anchor" href="#[Version-0.17.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.0)-(May-27,-2017)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.0">Version 0.17.0</a> (May 27, 2017)</a><a id="[Version-0.17.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.0)-(May-27,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.17.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.0)-(May-27,-2017)" title="Permalink"></a></h2><ul><li><strong>Breaking change</strong>: Mixing quadratic and conic constraints is no longer supported.</li><li><strong>Breaking change</strong>: The <code>getvariable</code> and <code>getconstraint</code> functions are replaced by indexing on the corresponding symbol. For instance, to access the variable with name <code>x</code>, one should now write <code>m[:x]</code> instead of <code>getvariable(m, :x)</code>. As a consequence, creating a variable and constraint with the same name now triggers a warning, and accessing one of them afterwards throws an error. This change is breaking only in the latter case.</li><li>Addition of the <code>getobjectivebound</code> function that mirrors the functionality of the MathProgBase <code>getobjbound</code> function except that it takes into account transformations performed by JuMP.</li><li>Minor bug fixes.</li></ul><p>The following changes are primarily of interest to developers of JuMP extensions:</p><ul><li>The new syntax <code>@constraint(model, expr in Cone)</code> creates the constraint ensuring that <code>expr</code> is inside <code>Cone</code>. The <code>Cone</code> argument is passed to <code>constructconstraint!</code> which enables the call to the dispatched to an extension.</li><li>The <code>@variable</code> macro now calls <code>constructvariable!</code> instead of directly calling the <code>Variable</code> constructor. Extra arguments and keyword arguments passed to <code>@variable</code> are passed to <code>constructvariable!</code> which enables the call to be dispatched to an extension.</li><li>Refactor the internal function <code>conicdata</code> (used build the MathProgBase conic model) into smaller sub-functions to make these parts reusable by extensions.</li></ul><h2 id="[Version-0.16.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.2)-(March-28,-2017)"><a class="docs-heading-anchor" href="#[Version-0.16.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.2)-(March-28,-2017)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.2">Version 0.16.2</a> (March 28, 2017)</a><a id="[Version-0.16.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.2)-(March-28,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.16.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.2)-(March-28,-2017)" title="Permalink"></a></h2><ul><li>Minor bug fixes and printing tweaks</li><li>Address deprecation warnings for Julia 0.6</li></ul><h2 id="[Version-0.16.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.1)-(March-7,-2017)"><a class="docs-heading-anchor" href="#[Version-0.16.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.1)-(March-7,-2017)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.1">Version 0.16.1</a> (March 7, 2017)</a><a id="[Version-0.16.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.1)-(March-7,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.16.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.1)-(March-7,-2017)" title="Permalink"></a></h2><ul><li>Better support for <code>AbstractArray</code> in JuMP (Thanks @tkoolen)</li><li>Minor bug fixes</li></ul><h2 id="[Version-0.16.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.0)-(February-23,-2017)"><a class="docs-heading-anchor" href="#[Version-0.16.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.0)-(February-23,-2017)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.0">Version 0.16.0</a> (February 23, 2017)</a><a id="[Version-0.16.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.0)-(February-23,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.16.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.0)-(February-23,-2017)" title="Permalink"></a></h2><ul><li><strong>Breaking change</strong>: JuMP no longer has a mechanism for selecting solvers by default (the previous mechanism was flawed and incompatible with Julia 0.6). Not specifying a solver before calling <code>solve()</code> will result in an error.</li><li><strong>Breaking change</strong>: User-defined functions are no longer global. The first argument to <code>JuMP.register</code> is now a JuMP <code>Model</code> object within whose scope the function will be registered. Calling <code>JuMP.register</code> without a <code>Model</code> now produces an error.</li><li><strong>Breaking change</strong>: Use the new <code>JuMP.fix</code> method to fix a variable to a value or to update the value to which a variable is fixed. Calling <code>setvalue</code> on a fixed variable now results in an error in order to avoid silent behavior changes. (Thanks @joaquimg)</li><li>Nonlinear expressions now print out similarly to linear/quadratic expressions (useful for debugging!)</li><li>New <code>category</code> keyword to <code>@variable</code>. Used for specifying categories of anonymous variables.</li><li>Compatibility with Julia 0.6-dev.</li><li>Minor fixes and improvements (Thanks @cossio, @ccoffrin, @blegat)</li></ul><h2 id="[Version-0.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.1)-(January-31,-2017)"><a class="docs-heading-anchor" href="#[Version-0.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.1)-(January-31,-2017)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.1">Version 0.15.1</a> (January 31, 2017)</a><a id="[Version-0.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.1)-(January-31,-2017)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.1)-(January-31,-2017)" title="Permalink"></a></h2><ul><li>Bugfix for <code>@LinearConstraints</code> and friends</li></ul><h2 id="[Version-0.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.0)-(December-22,-2016)"><a class="docs-heading-anchor" href="#[Version-0.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.0)-(December-22,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.0">Version 0.15.0</a> (December 22, 2016)</a><a id="[Version-0.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.0)-(December-22,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.0)-(December-22,-2016)" title="Permalink"></a></h2><ul><li>Julia 0.5.0 is the minimum required version for this release.</li><li>Document support for BARON solver</li><li>Enable info callbacks in more states than before, for example, for recording solutions. New <code>when</code> argument to <code>addinfocallback</code> (<a href="https://github.com/JuliaOpt/JuMP.jl/pull/814">#814</a>, thanks @yeesian)</li><li>Improved support for anonymous variables. This includes new warnings for potentially confusing use of the traditional non-anonymous syntax:<ul><li>When multiple variables in a model are given the same name</li><li>When non-symbols are used as names, for example, <code>@variable(m, x[1][1:N])</code></li></ul></li><li>Improvements in iterating over JuMP containers (<a href="https://github.com/JuliaOpt/JuMP.jl/pull/836">#836</a>, thanks @IssamT)</li><li>Support for writing variable names in .lp file output (Thanks @leethargo)</li><li>Support for querying duals to SDP problems (Thanks @blegat)</li><li>The comprehension syntax with curly braces <code>sum{}</code>, <code>prod{}</code>, and <code>norm2{}</code> has been deprecated in favor of Julia&#39;s native comprehension syntax <code>sum()</code>, <code>prod()</code> and <code>norm()</code> as previously announced. (For early adopters of the new syntax, <code>norm2()</code> was renamed to <code>norm()</code> without deprecation.)</li><li>Unit tests rewritten to use Base.Test instead of FactCheck</li><li>Improved support for operations with matrices of JuMP types (Thanks @ExpandingMan)</li><li>The syntax to halt a solver from inside a callback has changed from <code>throw(CallbackAbort())</code> to <code>return JuMP.StopTheSolver</code></li><li>Minor bug fixes</li></ul><h2 id="[Version-0.14.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.2)-(December-12,-2016)"><a class="docs-heading-anchor" href="#[Version-0.14.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.2)-(December-12,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.2">Version 0.14.2</a> (December 12, 2016)</a><a id="[Version-0.14.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.2)-(December-12,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.14.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.2)-(December-12,-2016)" title="Permalink"></a></h2><ul><li>Allow singleton anonymous variables (includes bugfix)</li></ul><h2 id="[Version-0.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.1)-(September-12,-2016)"><a class="docs-heading-anchor" href="#[Version-0.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.1)-(September-12,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.1">Version 0.14.1</a> (September 12, 2016)</a><a id="[Version-0.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.1)-(September-12,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.1)-(September-12,-2016)" title="Permalink"></a></h2><ul><li>More consistent handling of states in informational callbacks, includes a new <code>when</code> parameter to <code>addinfocallback</code> for specifying in which state an informational callback should be called.</li></ul><h2 id="[Version-0.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.0)-(August-7,-2016)"><a class="docs-heading-anchor" href="#[Version-0.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.0)-(August-7,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.0">Version 0.14.0</a> (August 7, 2016)</a><a id="[Version-0.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.0)-(August-7,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.0)-(August-7,-2016)" title="Permalink"></a></h2><ul><li>Compatibility with Julia 0.5 and ForwardDiff 0.2</li><li>Support for &quot;anonymous&quot; variables, constraints, expressions, and parameters, for example, <code>x = @variable(m, [1:N])</code> instead of <code>@variable(m, x[1:N])</code></li><li>Support for retrieving constraints from a model by name via <code>getconstraint</code></li><li><code>@NLconstraint</code> now returns constraint references (as expected).</li><li>Support for vectorized expressions within lazy constraints</li><li>On Julia 0.5, parse new comprehension syntax <code>sum(x[i] for i in 1:N if isodd(i))</code> instead of <code>sum{ x[i], i in 1:N; isodd(i) }</code>. The old syntax with curly braces will be deprecated in JuMP 0.15.</li><li>Now possible to provide nonlinear expressions as &quot;raw&quot; Julia <code>Expr</code> objects instead of using JuMP&#39;s nonlinear macros. This input format is useful for programmatically generated expressions.</li><li><code>s/Mathematical Programming/Mathematical Optimization/</code></li><li>Support for local cuts (Thanks to @madanim, Mehdi Madani)</li><li>Document Xpress interface developed by @joaquimg, Joaquim Dias Garcia</li><li>Minor bug and deprecation fixes (Thanks @odow, @jrevels)</li></ul><h2 id="[Version-0.13.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.2)-(May-16,-2016)"><a class="docs-heading-anchor" href="#[Version-0.13.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.2)-(May-16,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.2">Version 0.13.2</a> (May 16, 2016)</a><a id="[Version-0.13.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.2)-(May-16,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.13.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.2)-(May-16,-2016)" title="Permalink"></a></h2><ul><li>Compatibility update for MathProgBase</li></ul><h2 id="[Version-0.13.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.1)-(May-3,-2016)"><a class="docs-heading-anchor" href="#[Version-0.13.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.1)-(May-3,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.1">Version 0.13.1</a> (May 3, 2016)</a><a id="[Version-0.13.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.1)-(May-3,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.13.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.1)-(May-3,-2016)" title="Permalink"></a></h2><ul><li>Fix broken deprecation for <code>registerNLfunction</code>.</li></ul><h2 id="[Version-0.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.0)-(April-29,-2016)"><a class="docs-heading-anchor" href="#[Version-0.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.0)-(April-29,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.0">Version 0.13.0</a> (April 29, 2016)</a><a id="[Version-0.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.0)-(April-29,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.0)-(April-29,-2016)" title="Permalink"></a></h2><ul><li>Most exported methods and macros have been renamed to avoid camelCase. See the list of changes <a href="https://github.com/JuliaOpt/JuMP.jl/blob/e53d0db67cde2a4b80d0c1281f4b49eb0128a1f5/src/deprecated.jl#L30">here</a>. There is a 1-1 mapping from the old names to the new, and it is safe to simply replace the names to update existing models.</li><li>Specify variable lower/upper bounds in <code>@variable</code> using the <code>lowerbound</code> and <code>upperbound</code> keyword arguments.</li><li>Change name printed for variable using the <code>basename</code> keyword argument to <code>@variable</code>.</li><li>New <code>@variables</code> macro allows multi-line declaration of groups of variables.</li><li>A number of solver methods previously available only through MathProgBase are now exposed directly in JuMP. The fix was <a href="https://youtu.be/qF1lZPJ3a5A">recorded</a> live.</li><li>Compatibility fixes with Julia 0.5.</li><li>The &quot;end&quot; indexing syntax is no longer supported within JuMPArrays which do not use 1-based indexing until upstream issues are resolved, see <a href="https://github.com/JuliaOpt/JuMP.jl/issues/730">here</a>.</li></ul><h2 id="[Version-0.12.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.2)-(March-9,-2016)"><a class="docs-heading-anchor" href="#[Version-0.12.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.2)-(March-9,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.2">Version 0.12.2</a> (March 9, 2016)</a><a id="[Version-0.12.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.2)-(March-9,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.12.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.2)-(March-9,-2016)" title="Permalink"></a></h2><ul><li>Small fixes for nonlinear optimization</li></ul><h2 id="[Version-0.12.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.1)-(March-1,-2016)"><a class="docs-heading-anchor" href="#[Version-0.12.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.1)-(March-1,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.1">Version 0.12.1</a> (March 1, 2016)</a><a id="[Version-0.12.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.1)-(March-1,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.12.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.1)-(March-1,-2016)" title="Permalink"></a></h2><ul><li>Fix a regression in slicing for JuMPArrays (when not using 1-based indexing)</li></ul><h2 id="[Version-0.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.0)-(February-27,-2016)"><a class="docs-heading-anchor" href="#[Version-0.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.0)-(February-27,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.0">Version 0.12.0</a> (February 27, 2016)</a><a id="[Version-0.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.0)-(February-27,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.0)-(February-27,-2016)" title="Permalink"></a></h2><ul><li>The automatic differentiation functionality has been completely rewritten with a number of user-facing changes:<ul><li><code>@defExpr</code> and <code>@defNLExpr</code> now take the model as the first argument. The previous one-argument version of <code>@defExpr</code> is deprecated; all expressions should be named. For example, replace <code>@defExpr(2x+y)</code> with <code>@defExpr(jump_model, my_expr, 2x+y)</code>.</li><li>JuMP no longer uses Julia&#39;s variable binding rules for efficiently re-solving a sequence of nonlinear models. Instead, we have introduced nonlinear parameters. This is a breaking change, so we have added a warning message when we detect models that may depend on the old behavior.</li><li>Support for user-defined functions integrated within nonlinear JuMP expressions.</li></ul></li><li>Replaced iteration over <code>AffExpr</code> with <code>Number</code>-like scalar iteration; previous iteration behavior is now available via <code>linearterms(::AffExpr)</code>.</li><li>Stopping the solver via <code>throw(CallbackAbort())</code> from a callback no longer triggers an exception. Instead, <code>solve()</code> returns <code>UserLimit</code> status.</li><li><code>getDual()</code> now works for conic problems (Thanks @emreyamangil.)</li></ul><h2 id="[Version-0.11.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.3)-(February-4,-2016)"><a class="docs-heading-anchor" href="#[Version-0.11.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.3)-(February-4,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.3">Version 0.11.3</a> (February 4, 2016)</a><a id="[Version-0.11.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.3)-(February-4,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.11.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.3)-(February-4,-2016)" title="Permalink"></a></h2><ul><li>Bug-fix for problems with quadratic objectives and semidefinite constraints</li></ul><h2 id="[Version-0.11.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.2)-(January-14,-2016)"><a class="docs-heading-anchor" href="#[Version-0.11.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.2)-(January-14,-2016)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.2">Version 0.11.2</a> (January 14, 2016)</a><a id="[Version-0.11.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.2)-(January-14,-2016)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.11.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.2)-(January-14,-2016)" title="Permalink"></a></h2><ul><li>Compatibility update for Mosek</li></ul><h2 id="[Version-0.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.1)-(December-1,-2015)"><a class="docs-heading-anchor" href="#[Version-0.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.1)-(December-1,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.1">Version 0.11.1</a> (December 1, 2015)</a><a id="[Version-0.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.1)-(December-1,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.1)-(December-1,-2015)" title="Permalink"></a></h2><ul><li>Remove usage of <code>@compat</code> in tests.</li><li>Fix updating quadratic objectives for nonlinear models.</li></ul><h2 id="[Version-0.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.0)-(November-30,-2015)"><a class="docs-heading-anchor" href="#[Version-0.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.0)-(November-30,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.0">Version 0.11.0</a> (November 30, 2015)</a><a id="[Version-0.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.0)-(November-30,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.0)-(November-30,-2015)" title="Permalink"></a></h2><ul><li>Julia 0.4.0 is the minimum required version for this release.</li><li>Fix for scoping semantics of index variables in sum{}. Index variables no longer leak into the surrounding scope.</li><li>Addition of the <code>solve(m::Model, relaxation=true)</code> keyword argument to solve the standard continuous relaxation of model <code>m</code></li><li>The <code>getConstraintBounds()</code> method allows access to the lower and upper bounds of all constraints in a (nonlinear) model.</li><li>Update for breaking changes in MathProgBase</li></ul><h2 id="[Version-0.10.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.3)-(November-20,-2015)"><a class="docs-heading-anchor" href="#[Version-0.10.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.3)-(November-20,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.3">Version 0.10.3</a> (November 20, 2015)</a><a id="[Version-0.10.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.3)-(November-20,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.10.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.3)-(November-20,-2015)" title="Permalink"></a></h2><ul><li>Fix a rare error when parsing quadratic expressions</li><li>Fix <code>Variable()</code> constructor with default arguments</li><li>Detect unrecognized keywords in <code>solve()</code></li></ul><h2 id="[Version-0.10.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.2)-(September-28,-2015)"><a class="docs-heading-anchor" href="#[Version-0.10.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.2)-(September-28,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.2">Version 0.10.2</a> (September 28, 2015)</a><a id="[Version-0.10.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.2)-(September-28,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.10.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.2)-(September-28,-2015)" title="Permalink"></a></h2><ul><li>Fix for deprecation warnings</li></ul><h2 id="[Version-0.10.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.1)-(September-3,-2015)"><a class="docs-heading-anchor" href="#[Version-0.10.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.1)-(September-3,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.1">Version 0.10.1</a> (September 3, 2015)</a><a id="[Version-0.10.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.1)-(September-3,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.10.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.1)-(September-3,-2015)" title="Permalink"></a></h2><ul><li>Fixes for ambiguity warnings.</li><li>Fix for breaking change in precompilation syntax in Julia 0.4-pre</li></ul><h2 id="[Version-0.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.0)-(August-31,-2015)"><a class="docs-heading-anchor" href="#[Version-0.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.0)-(August-31,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.0">Version 0.10.0</a> (August 31, 2015)</a><a id="[Version-0.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.0)-(August-31,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.0)-(August-31,-2015)" title="Permalink"></a></h2><ul><li>Support (on Julia 0.4 and later) for conditions in indexing <code>@defVar</code> and <code>@addConstraint</code> constructs, for example, <code>@defVar(m, x[i=1:5,j=1:5; i+j &gt;= 3])</code></li><li>Support for vectorized operations on Variables and expressions. See the documentation for details.</li><li>New <code>getVar()</code> method to access variables in a model by name</li><li>Support for semidefinite programming.</li><li>Dual solutions are now available for general nonlinear problems. You may call <code>getDual</code> on a reference object for a nonlinear constraint, and <code>getDual</code> on a variable object for Lagrange multipliers from active bounds.</li><li>Introduce warnings for two common performance traps: too many calls to <code>getValue()</code> on a collection of variables and use of the <code>+</code> operator in a loop to sum expressions.</li><li>Second-order cone constraints can be written directly with the <code>norm()</code> and <code>norm2{}</code> syntax.</li><li>Implement MathProgBase interface for querying Hessian-vector products.</li><li>Iteration over <code>JuMPContainer</code>s is deprecated; instead, use the <code>keys</code> and <code>values</code> functions, and <code>zip(keys(d),values(d))</code> for the old behavior.</li><li><code>@defVar</code> returns <code>Array{Variable,N}</code> when each of <code>N</code> index sets are of the form <code>1:nᵢ</code>.</li><li>Module precompilation: on Julia 0.4 and later, <code>using JuMP</code> is now much faster.</li></ul><h2 id="[Version-0.9.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.3)-(August-11,-2015)"><a class="docs-heading-anchor" href="#[Version-0.9.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.3)-(August-11,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.3">Version 0.9.3</a> (August 11, 2015)</a><a id="[Version-0.9.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.3)-(August-11,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.9.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.3)-(August-11,-2015)" title="Permalink"></a></h2><ul><li>Fixes for FactCheck testing on julia v0.4.</li></ul><h2 id="[Version-0.9.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.2)-(June-27,-2015)"><a class="docs-heading-anchor" href="#[Version-0.9.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.2)-(June-27,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.2">Version 0.9.2</a> (June 27, 2015)</a><a id="[Version-0.9.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.2)-(June-27,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.9.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.2)-(June-27,-2015)" title="Permalink"></a></h2><ul><li>Fix bug in @addConstraints.</li></ul><h2 id="[Version-0.9.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.1)-(April-25,-2015)"><a class="docs-heading-anchor" href="#[Version-0.9.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.1)-(April-25,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.1">Version 0.9.1</a> (April 25, 2015)</a><a id="[Version-0.9.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.1)-(April-25,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.9.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.1)-(April-25,-2015)" title="Permalink"></a></h2><ul><li>Fix for Julia 0.4-dev.</li><li>Small infrastructure improvements for extensions.</li></ul><h2 id="[Version-0.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.0)-(April-18,-2015)"><a class="docs-heading-anchor" href="#[Version-0.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.0)-(April-18,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.0">Version 0.9.0</a> (April 18, 2015)</a><a id="[Version-0.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.0)-(April-18,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.0)-(April-18,-2015)" title="Permalink"></a></h2><ul><li>Comparison operators for constructing constraints (for example, <code>2x &gt;= 1</code>) have been deprecated. Instead, construct the constraints explicitly in the <code>@addConstraint</code> macro to add them to the model, or in the <code>@LinearConstraint</code> macro to create a stand-alone linear constraint instance.</li><li><code>getValue()</code> method implemented to compute the value of a nonlinear subexpression</li><li>JuMP is now released under the Mozilla Public License version 2.0 (was previously LGPL). MPL is a copyleft license which is less restrictive than LGPL, especially for embedding JuMP within other applications.</li><li>A number of performance improvements in ReverseDiffSparse for computing derivatives.</li><li><code>MathProgBase.getsolvetime(m)</code> now returns the solution time reported by the solver, if available. (Thanks @odow, Oscar Dowson)</li><li>Formatting fix for LP format output. (Thanks @sbebo, Leonardo Taccari).</li></ul><h2 id="[Version-0.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.8.0)-(February-17,-2015)"><a class="docs-heading-anchor" href="#[Version-0.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.8.0)-(February-17,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.8.0">Version 0.8.0</a> (February 17, 2015)</a><a id="[Version-0.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.8.0)-(February-17,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.8.0)-(February-17,-2015)" title="Permalink"></a></h2><ul><li>Nonlinear subexpressions now supported with the <code>@defNLExpr</code> macro.</li><li>SCS supported for solving second-order conic problems.</li><li><code>setXXXCallback</code> family deprecated in favor of <code>addXXXCallback</code>.</li><li>Multiple callbacks of the same type can be registered.</li><li>Added support for informational callbacks via <code>addInfoCallback</code>.</li><li>A <code>CallbackAbort</code> exception can be thrown from callback to safely exit optimization.</li></ul><h2 id="[Version-0.7.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.4)-(February-4,-2015)"><a class="docs-heading-anchor" href="#[Version-0.7.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.4)-(February-4,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.4">Version 0.7.4</a> (February 4, 2015)</a><a id="[Version-0.7.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.4)-(February-4,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.7.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.4)-(February-4,-2015)" title="Permalink"></a></h2><ul><li>Reduced costs and linear constraint duals are now accessible when quadratic constraints are present.</li><li>Two-sided nonlinear constraints are supported.</li><li>Methods for accessing the number of variables and constraints in a model are renamed.</li><li>New default procedure for setting initial values in nonlinear optimization: project zero onto the variable bounds.</li><li>Small bug fixes.</li></ul><h2 id="[Version-0.7.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.3)-(January-14,-2015)"><a class="docs-heading-anchor" href="#[Version-0.7.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.3)-(January-14,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.3">Version 0.7.3</a> (January 14, 2015)</a><a id="[Version-0.7.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.3)-(January-14,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.7.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.3)-(January-14,-2015)" title="Permalink"></a></h2><ul><li>Fix a method ambiguity conflict with Compose.jl (cosmetic fix)</li></ul><h2 id="[Version-0.7.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.2)-(January-9,-2015)"><a class="docs-heading-anchor" href="#[Version-0.7.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.2)-(January-9,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.2">Version 0.7.2</a> (January 9, 2015)</a><a id="[Version-0.7.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.2)-(January-9,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.7.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.2)-(January-9,-2015)" title="Permalink"></a></h2><ul><li>Fix a bug in <code>sum(::JuMPDict)</code></li><li>Added the <code>setCategory</code> function to change a variables category (for example, continuous or binary)</li></ul><p>after construction, and <code>getCategory</code> to retrieve the variable category.</p><h2 id="[Version-0.7.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.1)-(January-2,-2015)"><a class="docs-heading-anchor" href="#[Version-0.7.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.1)-(January-2,-2015)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.1">Version 0.7.1</a> (January 2, 2015)</a><a id="[Version-0.7.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.1)-(January-2,-2015)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.7.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.1)-(January-2,-2015)" title="Permalink"></a></h2><ul><li>Fix a bug in parsing linear expressions in macros. Affects only Julia 0.4 and later.</li></ul><h2 id="[Version-0.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.0)-(December-29,-2014)"><a class="docs-heading-anchor" href="#[Version-0.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.0)-(December-29,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.0">Version 0.7.0</a> (December 29, 2014)</a><a id="[Version-0.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.0)-(December-29,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.0)-(December-29,-2014)" title="Permalink"></a></h2><h3 id="Linear/quadratic/conic-programming"><a class="docs-heading-anchor" href="#Linear/quadratic/conic-programming">Linear/quadratic/conic programming</a><a id="Linear/quadratic/conic-programming-1"></a><a class="docs-heading-anchor-permalink" href="#Linear/quadratic/conic-programming" title="Permalink"></a></h3><ul><li><strong>Breaking change</strong>: The syntax for column-wise model generation has been changed to use keyword arguments in <code>@defVar</code>.</li><li>On Julia 0.4 and later, variables and coefficients may be multiplied in any order within macros. That is, variable*coefficient is now valid syntax.</li><li>ECOS supported for solving second-order conic problems.</li></ul><h3 id="_nonlinear_programming_release_notes"><a class="docs-heading-anchor" href="#_nonlinear_programming_release_notes">Nonlinear programming</a><a id="_nonlinear_programming_release_notes-1"></a><a class="docs-heading-anchor-permalink" href="#_nonlinear_programming_release_notes" title="Permalink"></a></h3><ul><li>Support for skipping model generation when solving a sequence of nonlinear models with changing data.</li><li>Fix a memory leak when solving a sequence of nonlinear models.</li><li>The <code>@addNLConstraint</code> macro now supports the three-argument version to define sets of nonlinear constraints.</li><li>KNITRO supported as a nonlinear solver.</li><li>Speed improvements for model generation.</li><li>The <code>@addNLConstraints</code> macro supports adding multiple (groups of) constraints at once. Syntax is similar to <code>@addConstraints</code>.</li><li>Discrete variables allowed in nonlinear problems for solvers which support them (currently only KNITRO).</li></ul><h3 id="General"><a class="docs-heading-anchor" href="#General">General</a><a id="General-1"></a><a class="docs-heading-anchor-permalink" href="#General" title="Permalink"></a></h3><ul><li>Starting values for variables may now be specified with <code>@defVar(m, x, start=value)</code>.</li><li>The <code>setSolver</code> function allows users to change the solver subsequent to model creation.</li><li>Support for &quot;fixed&quot; variables via the <code>@defVar(m, x == 1)</code> syntax.</li><li>Unit tests rewritten to use FactCheck.jl, improved testing across solvers.</li></ul><h2 id="[Version-0.6.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.3)-(October-19,-2014)"><a class="docs-heading-anchor" href="#[Version-0.6.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.3)-(October-19,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.3">Version 0.6.3</a> (October 19, 2014)</a><a id="[Version-0.6.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.3)-(October-19,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.6.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.3)-(October-19,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug in multiplying two AffExpr objects.</li></ul><h2 id="[Version-0.6.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.2)-(October-11,-2014)"><a class="docs-heading-anchor" href="#[Version-0.6.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.2)-(October-11,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.2">Version 0.6.2</a> (October 11, 2014)</a><a id="[Version-0.6.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.2)-(October-11,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.6.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.2)-(October-11,-2014)" title="Permalink"></a></h2><ul><li>Further improvements and bug fixes for printing.</li><li>Fixed a bug in <code>@defExpr</code>.</li><li>Support for accessing expression graphs through the MathProgBase NLP interface.</li></ul><h2 id="[Version-0.6.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.1)-(September-19,-2014)"><a class="docs-heading-anchor" href="#[Version-0.6.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.1)-(September-19,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.1">Version 0.6.1</a> (September 19, 2014)</a><a id="[Version-0.6.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.1)-(September-19,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.6.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.1)-(September-19,-2014)" title="Permalink"></a></h2><ul><li>Improvements and bug fixes for printing.</li></ul><h2 id="[Version-0.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.0)-(September-9,-2014)"><a class="docs-heading-anchor" href="#[Version-0.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.0)-(September-9,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.0">Version 0.6.0</a> (September 9, 2014)</a><a id="[Version-0.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.0)-(September-9,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.0)-(September-9,-2014)" title="Permalink"></a></h2><ul><li>Julia 0.3.0 is the minimum required version for this release.</li><li><code>buildInternalModel(m::Model)</code> added to build solver-level model in memory without optimizing.</li><li>Deprecate <code>load_model_only</code> keyword argument to <code>solve</code>.</li><li>Add groups of constraints with <code>@addConstraints</code> macro.</li><li>Unicode operators now supported, including <code>∑</code> for <code>sum</code>, <code>∏</code> for <code>prod</code>, and <code>≤</code>/<code>≥</code></li><li>Quadratic constraints supported in <code>@addConstraint</code> macro.</li><li>Quadratic objectives supported in <code>@setObjective</code> macro.</li><li>MathProgBase solver-independent interface replaces Ipopt-specific interface for nonlinear problems<ul><li><strong>Breaking change</strong>: <code>IpoptOptions</code> no longer supported to specify solver options, use <code>m = Model(solver=IpoptSolver(options...))</code> instead.</li></ul></li><li>New solver interfaces: ECOS, NLopt, and nonlinear support for MOSEK</li><li>New option to control whether the lazy constraint callback is executed at each node in the B&amp;B tree or just when feasible solutions are found</li><li>Add support for semicontinuous and semi-integer variables for those solvers that support them.</li><li>Add support for index dependencies (for example, triangular indexing) in <code>@defVar</code>, <code>@addConstraint</code>, and <code>@defExpr</code> (for example, <code>@defVar(m, x[i=1:10,j=i:10])</code>).<ul><li>This required some changes to the internal structure of JuMP containers, which may break code that explicitly stored <code>JuMPDict</code> objects.</li></ul></li></ul><h2 id="[Version-0.5.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.8)-(September-24,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.8)-(September-24,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.8">Version 0.5.8</a> (September 24, 2014)</a><a id="[Version-0.5.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.8)-(September-24,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.8)-(September-24,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug with specifying solvers (affects Julia 0.2 only)</li></ul><h2 id="[Version-0.5.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.7)-(September-5,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.7)-(September-5,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.7">Version 0.5.7</a> (September 5, 2014)</a><a id="[Version-0.5.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.7)-(September-5,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.7)-(September-5,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug in printing models</li></ul><h2 id="[Version-0.5.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.6)-(September-2,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.6)-(September-2,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.6">Version 0.5.6</a> (September 2, 2014)</a><a id="[Version-0.5.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.6)-(September-2,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.6)-(September-2,-2014)" title="Permalink"></a></h2><ul><li>Add support for semicontinuous and semi-integer variables for those solvers that support them.<ul><li><strong>Breaking change</strong>: Syntax for <code>Variable()</code> constructor has changed (use of this interface remains discouraged)</li></ul></li><li>Update for breaking changes in MathProgBase</li></ul><h2 id="[Version-0.5.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.5)-(July-6,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.5)-(July-6,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.5">Version 0.5.5</a> (July 6, 2014)</a><a id="[Version-0.5.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.5)-(July-6,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.5)-(July-6,-2014)" title="Permalink"></a></h2><ul><li>Fix bug with problem modification: adding variables that did not appear in existing constraints or objective.</li></ul><h2 id="[Version-0.5.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.4)-(June-19,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.4)-(June-19,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.4">Version 0.5.4</a> (June 19, 2014)</a><a id="[Version-0.5.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.4)-(June-19,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.4)-(June-19,-2014)" title="Permalink"></a></h2><ul><li>Update for breaking change in MathProgBase which reduces loading times for <code>using JuMP</code></li><li>Fix error when MIPs not solved to optimality</li></ul><h2 id="[Version-0.5.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.3)-(May-21,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.3)-(May-21,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.3">Version 0.5.3</a> (May 21, 2014)</a><a id="[Version-0.5.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.3)-(May-21,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.3)-(May-21,-2014)" title="Permalink"></a></h2><ul><li>Update for breaking change in ReverseDiffSparse</li></ul><h2 id="[Version-0.5.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.2)-(May-9,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.2)-(May-9,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.2">Version 0.5.2</a> (May 9, 2014)</a><a id="[Version-0.5.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.2)-(May-9,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.2)-(May-9,-2014)" title="Permalink"></a></h2><ul><li>Fix compatibility with Julia 0.3 prerelease</li></ul><h2 id="[Version-0.5.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.1)-(May-5,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.1)-(May-5,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.1">Version 0.5.1</a> (May 5, 2014)</a><a id="[Version-0.5.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.1)-(May-5,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.1)-(May-5,-2014)" title="Permalink"></a></h2><ul><li>Fix a bug in coefficient handling inside lazy constraints and user cuts</li></ul><h2 id="[Version-0.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.0)-(May-2,-2014)"><a class="docs-heading-anchor" href="#[Version-0.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.0)-(May-2,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.0">Version 0.5.0</a> (May 2, 2014)</a><a id="[Version-0.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.0)-(May-2,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.0)-(May-2,-2014)" title="Permalink"></a></h2><ul><li>Support for nonlinear optimization with exact, sparse second-order derivatives automatically computed. Ipopt is currently the only solver supported.</li><li><code>getValue</code> for <code>AffExpr</code> and <code>QuadExpr</code></li><li><strong>Breaking change</strong>: <code>getSolverModel</code> replaced by <code>getInternalModel</code>, which returns the internal MathProgBase-level model</li><li>Groups of constraints can be specified with <code>@addConstraint</code> (see documentation for details). This is not a breaking change.</li><li><code>dot(::JuMPDict{Variable},::JuMPDict{Variable})</code> now returns the corresponding quadratic expression.</li></ul><h2 id="[Version-0.4.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.1)-(March-24,-2014)"><a class="docs-heading-anchor" href="#[Version-0.4.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.1)-(March-24,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.1">Version 0.4.1</a> (March 24, 2014)</a><a id="[Version-0.4.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.1)-(March-24,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.4.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.1)-(March-24,-2014)" title="Permalink"></a></h2><ul><li>Fix bug where change in objective sense was ignored when re-solving a model.</li><li>Fix issue with handling zero coefficients in AffExpr.</li></ul><h2 id="[Version-0.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.0)-(March-10,-2014)"><a class="docs-heading-anchor" href="#[Version-0.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.0)-(March-10,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.0">Version 0.4.0</a> (March 10, 2014)</a><a id="[Version-0.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.0)-(March-10,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.0)-(March-10,-2014)" title="Permalink"></a></h2><ul><li>Support for SOS1 and SOS2 constraints.</li><li>Solver-independent callback for user heuristics.</li><li><code>dot</code> and <code>sum</code> implemented for <code>JuMPDict</code> objects. Now you can say <code>@addConstraint(m, dot(a,x) &lt;= b)</code>.</li><li>Developers: support for extensions to JuMP. See definition of Model in <code>src/JuMP.jl</code> for more details.</li><li>Option to construct the low-level model before optimizing.</li></ul><h2 id="[Version-0.3.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.2)-(February-17,-2014)"><a class="docs-heading-anchor" href="#[Version-0.3.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.2)-(February-17,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.2">Version 0.3.2</a> (February 17, 2014)</a><a id="[Version-0.3.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.2)-(February-17,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.3.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.2)-(February-17,-2014)" title="Permalink"></a></h2><ul><li>Improved model printing<ul><li>Preliminary support for IJulia output</li></ul></li></ul><h2 id="[Version-0.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.1)-(January-30,-2014)"><a class="docs-heading-anchor" href="#[Version-0.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.1)-(January-30,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.1">Version 0.3.1</a> (January 30, 2014)</a><a id="[Version-0.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.1)-(January-30,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.1)-(January-30,-2014)" title="Permalink"></a></h2><ul><li>Documentation updates</li><li>Support for MOSEK</li><li>CPLEXLink renamed to CPLEX</li></ul><h2 id="[Version-0.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.0)-(January-21,-2014)"><a class="docs-heading-anchor" href="#[Version-0.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.0)-(January-21,-2014)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.0">Version 0.3.0</a> (January 21, 2014)</a><a id="[Version-0.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.0)-(January-21,-2014)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.0)-(January-21,-2014)" title="Permalink"></a></h2><ul><li>Unbounded/infeasibility rays: <code>getValue()</code> will return the corresponding components of an unbounded ray when a model is unbounded, if supported by the selected solver. <code>getDual()</code> will return an infeasibility ray (Farkas proof) if a model is infeasible and the selected solver supports this feature.</li><li>Solver-independent callbacks for user generated cuts.</li><li>Use new interface for solver-independent QCQP.</li><li><code>setlazycallback</code> renamed to <code>setLazyCallback</code> for consistency.</li></ul><h2 id="[Version-0.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.2.0)-(December-15,-2013)"><a class="docs-heading-anchor" href="#[Version-0.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.2.0)-(December-15,-2013)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.2.0">Version 0.2.0</a> (December 15, 2013)</a><a id="[Version-0.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.2.0)-(December-15,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.2.0)-(December-15,-2013)" title="Permalink"></a></h2><h3 id="Breaking-8"><a class="docs-heading-anchor" href="#Breaking-8">Breaking</a><a class="docs-heading-anchor-permalink" href="#Breaking-8" title="Permalink"></a></h3><ul><li>Objective sense is specified in <code>setObjective</code> instead of in the <code>Model</code> constructor.</li><li><code>lpsolver</code> and <code>mipsolver</code> merged into single <code>solver</code> option.</li></ul><h3 id="Added-30"><a class="docs-heading-anchor" href="#Added-30">Added</a><a class="docs-heading-anchor-permalink" href="#Added-30" title="Permalink"></a></h3><ul><li>Problem modification with efficient LP restarts and MIP warm-starts.</li><li>Relatedly, column-wise modeling now supported.</li><li>Solver-independent callbacks supported. Currently we support only a &quot;lazy constraint&quot; callback, which works with Gurobi, CPLEX, and GLPK. More callbacks coming soon.</li></ul><h2 id="[Version-0.1.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.2)-(November-16,-2013)"><a class="docs-heading-anchor" href="#[Version-0.1.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.2)-(November-16,-2013)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.2">Version 0.1.2</a> (November 16, 2013)</a><a id="[Version-0.1.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.2)-(November-16,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.1.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.2)-(November-16,-2013)" title="Permalink"></a></h2><ul><li>Bug fixes for printing, improved error messages.</li><li>Allow <code>AffExpr</code> to be used in macros; for example, <code>ex = y + z; @addConstraint(m, x + 2*ex &lt;= 3)</code></li></ul><h2 id="[Version-0.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.1)-(October-23,-2013)"><a class="docs-heading-anchor" href="#[Version-0.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.1)-(October-23,-2013)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.1">Version 0.1.1</a> (October 23, 2013)</a><a id="[Version-0.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.1)-(October-23,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.1)-(October-23,-2013)" title="Permalink"></a></h2><ul><li>Update for solver specification API changes in MathProgBase.</li></ul><h2 id="[Version-0.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.0)-(October-3,-2013)"><a class="docs-heading-anchor" href="#[Version-0.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.0)-(October-3,-2013)"><a href="https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.0">Version 0.1.0</a> (October 3, 2013)</a><a id="[Version-0.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.0)-(October-3,-2013)-1"></a><a class="docs-heading-anchor-permalink" href="#[Version-0.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.0)-(October-3,-2013)" title="Permalink"></a></h2><ul><li>Initial public release.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../moi/release_notes/">« Release notes</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/search_index.js b/previews/PR3525/search_index.js
new file mode 100644
index 00000000000..8bd063266e3
--- /dev/null
+++ b/previews/PR3525/search_index.js
@@ -0,0 +1,3 @@
+var documenterSearchIndex = {"docs":
+[{"location":"moi/reference/nonlinear/","page":"Nonlinear programming","title":"Nonlinear programming","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/nonlinear.md\"","category":"page"},{"location":"moi/reference/nonlinear/","page":"Nonlinear programming","title":"Nonlinear programming","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/nonlinear/#Nonlinear-programming","page":"Nonlinear programming","title":"Nonlinear programming","text":"","category":"section"},{"location":"moi/reference/nonlinear/#Types","page":"Nonlinear programming","title":"Types","text":"","category":"section"},{"location":"moi/reference/nonlinear/","page":"Nonlinear programming","title":"Nonlinear programming","text":"AbstractNLPEvaluator\nNLPBoundsPair\nNLPBlockData","category":"page"},{"location":"moi/reference/nonlinear/#MathOptInterface.AbstractNLPEvaluator","page":"Nonlinear programming","title":"MathOptInterface.AbstractNLPEvaluator","text":"AbstractNLPEvaluator\n\nAbstract supertype for the callback object that is used to query function values, derivatives, and expression graphs.\n\nIt is used in NLPBlockData.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/nonlinear/#MathOptInterface.NLPBoundsPair","page":"Nonlinear programming","title":"MathOptInterface.NLPBoundsPair","text":"NLPBoundsPair(lower::Float64, upper::Float64)\n\nA struct holding a pair of lower and upper bounds.\n\n-Inf and Inf can be used to indicate no lower or upper bound, respectively.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/nonlinear/#MathOptInterface.NLPBlockData","page":"Nonlinear programming","title":"MathOptInterface.NLPBlockData","text":"struct NLPBlockData\n    constraint_bounds::Vector{NLPBoundsPair}\n    evaluator::AbstractNLPEvaluator\n    has_objective::Bool\nend\n\nA struct encoding a set of nonlinear constraints of the form lb le g(x) le ub and, if has_objective == true, a nonlinear objective function f(x).\n\nNonlinear objectives override any objective set by using the ObjectiveFunction attribute.\n\nThe evaluator is a callback object that is used to query function values, derivatives, and expression graphs. If has_objective == false, then it is an error to query properties of the objective function, and in Hessian-of-the-Lagrangian queries, σ must be set to zero.\n\nnote: Note\nThroughout the evaluator, all variables are ordered according to ListOfVariableIndices. Hence, MOI copies of nonlinear problems must not re-order variables.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/nonlinear/#Attributes","page":"Nonlinear programming","title":"Attributes","text":"","category":"section"},{"location":"moi/reference/nonlinear/","page":"Nonlinear programming","title":"Nonlinear programming","text":"NLPBlock\nNLPBlockDual\nNLPBlockDualStart","category":"page"},{"location":"moi/reference/nonlinear/#MathOptInterface.NLPBlock","page":"Nonlinear programming","title":"MathOptInterface.NLPBlock","text":"NLPBlock()\n\nAn AbstractModelAttribute that stores an NLPBlockData, representing a set of nonlinear constraints, and optionally a nonlinear objective.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/nonlinear/#MathOptInterface.NLPBlockDual","page":"Nonlinear programming","title":"MathOptInterface.NLPBlockDual","text":"NLPBlockDual(result_index::Int = 1)\n\nAn AbstractModelAttribute for the Lagrange multipliers on the constraints from the NLPBlock in result result_index.\n\nIf result_index is omitted, it is 1 by default.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/nonlinear/#MathOptInterface.NLPBlockDualStart","page":"Nonlinear programming","title":"MathOptInterface.NLPBlockDualStart","text":"NLPBlockDualStart()\n\nAn AbstractModelAttribute for the initial assignment of the Lagrange multipliers on the constraints from the NLPBlock that the solver may use to warm-start the solve.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/nonlinear/#Functions","page":"Nonlinear programming","title":"Functions","text":"","category":"section"},{"location":"moi/reference/nonlinear/","page":"Nonlinear programming","title":"Nonlinear programming","text":"initialize\nfeatures_available\neval_objective\neval_constraint\neval_objective_gradient\njacobian_structure\neval_constraint_gradient\nconstraint_gradient_structure\neval_constraint_jacobian\neval_constraint_jacobian_product\neval_constraint_jacobian_transpose_product\nhessian_lagrangian_structure\nhessian_objective_structure\nhessian_constraint_structure\neval_hessian_objective\neval_hessian_constraint\neval_hessian_lagrangian\neval_hessian_lagrangian_product\nobjective_expr\nconstraint_expr","category":"page"},{"location":"moi/reference/nonlinear/#MathOptInterface.initialize","page":"Nonlinear programming","title":"MathOptInterface.initialize","text":"initialize(\n    d::AbstractNLPEvaluator,\n    requested_features::Vector{Symbol},\n)::Nothing\n\nInitialize d with the set of features in requested_features. Check features_available before calling initialize to see what features are supported by d.\n\nwarning: Warning\nThis method must be called before any other methods.\n\nFeatures\n\nThe following features are defined:\n\n:Grad: enables eval_objective_gradient\n:Jac: enables eval_constraint_jacobian\n:JacVec: enables eval_constraint_jacobian_product and eval_constraint_jacobian_transpose_product\n:Hess: enables eval_hessian_lagrangian\n:HessVec: enables eval_hessian_lagrangian_product\n:ExprGraph: enables objective_expr and constraint_expr.\n\nIn all cases, including when requested_features is empty, eval_objective and eval_constraint are supported.\n\nExamples\n\nMOI.initialize(d, Symbol[])\nMOI.initialize(d, [:ExprGraph])\nMOI.initialize(d, MOI.features_available(d))\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.features_available","page":"Nonlinear programming","title":"MathOptInterface.features_available","text":"features_available(d::AbstractNLPEvaluator)::Vector{Symbol}\n\nReturns the subset of features available for this problem instance.\n\nSee initialize for the list of defined features.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_objective","page":"Nonlinear programming","title":"MathOptInterface.eval_objective","text":"eval_objective(d::AbstractNLPEvaluator, x::AbstractVector{T})::T where {T}\n\nEvaluate the objective f(x), returning a scalar value.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_constraint","page":"Nonlinear programming","title":"MathOptInterface.eval_constraint","text":"eval_constraint(d::AbstractNLPEvaluator,\n    g::AbstractVector{T},\n    x::AbstractVector{T},\n)::Nothing where {T}\n\nGiven a set of vector-valued constraints l le g(x) le u, evaluate the constraint function g(x), storing the result in the vector g.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that g is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_objective_gradient","page":"Nonlinear programming","title":"MathOptInterface.eval_objective_gradient","text":"eval_objective_gradient(\n    d::AbstractNLPEvaluator,\n    grad::AbstractVector{T},\n    x::AbstractVector{T},\n)::Nothing where {T}\n\nEvaluate the gradient of the objective function grad = nabla f(x) as a dense vector, storing the result in the vector grad.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that grad is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.jacobian_structure","page":"Nonlinear programming","title":"MathOptInterface.jacobian_structure","text":"jacobian_structure(d::AbstractNLPEvaluator)::Vector{Tuple{Int64,Int64}}\n\nReturns a vector of tuples, (row, column), where each indicates the position of a structurally nonzero element in the Jacobian matrix: J_g(x) = left beginarrayc nabla g_1(x)  nabla g_2(x)  vdots  nabla g_m(x) endarrayright where g_i is the itextth component of the nonlinear constraints g(x).\n\nThe indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.\n\nThe sparsity structure is assumed to be independent of the point x.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_constraint_gradient","page":"Nonlinear programming","title":"MathOptInterface.eval_constraint_gradient","text":"eval_constraint_gradient(\n    d::AbstractNLPEvaluator,\n    ∇g::AbstractVector{T},\n    x::AbstractVector{T},\n    i::Int,\n)::Nothing where {T}\n\nEvaluate the gradient of constraint i, nabla g_i(x), and store the non-zero values in ∇g, corresponding to the structure returned by constraint_gradient_structure.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that ∇g is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.constraint_gradient_structure","page":"Nonlinear programming","title":"MathOptInterface.constraint_gradient_structure","text":"constraint_gradient_structure(d::AbstractNLPEvaluator, i::Int)::Vector{Int64}\n\nReturns a vector of indices, where each element indicates the position of a structurally nonzero element in the gradient of constraint nabla g_i(x).\n\nThe indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.\n\nThe sparsity structure is assumed to be independent of the point x.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_constraint_jacobian","page":"Nonlinear programming","title":"MathOptInterface.eval_constraint_jacobian","text":"eval_constraint_jacobian(d::AbstractNLPEvaluator,\n    J::AbstractVector{T},\n    x::AbstractVector{T},\n)::Nothing where {T}\n\nEvaluates the sparse Jacobian matrix J_g(x) = left beginarrayc nabla g_1(x)  nabla g_2(x)  vdots  nabla g_m(x) endarrayright.\n\nThe result is stored in the vector J in the same order as the indices returned by jacobian_structure.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that J is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_constraint_jacobian_product","page":"Nonlinear programming","title":"MathOptInterface.eval_constraint_jacobian_product","text":"eval_constraint_jacobian_product(\n    d::AbstractNLPEvaluator,\n    y::AbstractVector{T},\n    x::AbstractVector{T},\n    w::AbstractVector{T},\n)::Nothing where {T}\n\nComputes the Jacobian-vector product y = J_g(x)w, storing the result in the vector y.\n\nThe vectors have dimensions such that length(w) == length(x), and length(y) is the number of nonlinear constraints.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that y is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_constraint_jacobian_transpose_product","page":"Nonlinear programming","title":"MathOptInterface.eval_constraint_jacobian_transpose_product","text":"eval_constraint_jacobian_transpose_product(\n    d::AbstractNLPEvaluator,\n    y::AbstractVector{T},\n    x::AbstractVector{T},\n    w::AbstractVector{T},\n)::Nothing where {T}\n\nComputes the Jacobian-transpose-vector product y = J_g(x)^Tw, storing the result in the vector y.\n\nThe vectors have dimensions such that length(y) == length(x), and length(w) is the number of nonlinear constraints.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that y is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.hessian_lagrangian_structure","page":"Nonlinear programming","title":"MathOptInterface.hessian_lagrangian_structure","text":"hessian_lagrangian_structure(\n    d::AbstractNLPEvaluator,\n)::Vector{Tuple{Int64,Int64}}\n\nReturns a vector of tuples, (row, column), where each indicates the position of a structurally nonzero element in the Hessian-of-the-Lagrangian matrix: nabla^2 f(x) + sum_i=1^m nabla^2 g_i(x).\n\nThe indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.\n\nAny mix of lower and upper-triangular indices is valid. Elements (i,j) and (j,i), if both present, should be treated as duplicates.\n\nThe sparsity structure is assumed to be independent of the point x.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.hessian_objective_structure","page":"Nonlinear programming","title":"MathOptInterface.hessian_objective_structure","text":"hessian_objective_structure(\n    d::AbstractNLPEvaluator,\n)::Vector{Tuple{Int64,Int64}}\n\nReturns a vector of tuples, (row, column), where each indicates the position of a structurally nonzero element in the Hessian matrix: nabla^2 f(x).\n\nThe indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.\n\nAny mix of lower and upper-triangular indices is valid. Elements (i,j) and (j,i), if both present, should be treated as duplicates.\n\nThe sparsity structure is assumed to be independent of the point x.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.hessian_constraint_structure","page":"Nonlinear programming","title":"MathOptInterface.hessian_constraint_structure","text":"hessian_constraint_structure(\n    d::AbstractNLPEvaluator,\n    i::Int64,\n)::Vector{Tuple{Int64,Int64}}\n\nReturns a vector of tuples, (row, column), where each indicates the position of a structurally nonzero element in the Hessian matrix: nabla^2 g_i(x).\n\nThe indices are not required to be sorted and can contain duplicates, in which case the solver should combine the corresponding elements by adding them together.\n\nAny mix of lower and upper-triangular indices is valid. Elements (i,j) and (j,i), if both present, should be treated as duplicates.\n\nThe sparsity structure is assumed to be independent of the point x.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_hessian_objective","page":"Nonlinear programming","title":"MathOptInterface.eval_hessian_objective","text":"eval_hessian_objective(\n    d::AbstractNLPEvaluator,\n    H::AbstractVector{T},\n    x::AbstractVector{T},\n)::Nothing where {T}\n\nThis function computes the sparse Hessian matrix: nabla^2 f(x), storing the result in the vector H in the same order as the indices returned by hessian_objective_structure.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that H is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_hessian_constraint","page":"Nonlinear programming","title":"MathOptInterface.eval_hessian_constraint","text":"eval_hessian_constraint(\n    d::AbstractNLPEvaluator,\n    H::AbstractVector{T},\n    x::AbstractVector{T},\n    i::Int64,\n)::Nothing where {T}\n\nThis function computes the sparse Hessian matrix: nabla^2 g_i(x), storing the result in the vector H in the same order as the indices returned by hessian_constraint_structure.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that H is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_hessian_lagrangian","page":"Nonlinear programming","title":"MathOptInterface.eval_hessian_lagrangian","text":"eval_hessian_lagrangian(\n    d::AbstractNLPEvaluator,\n    H::AbstractVector{T},\n    x::AbstractVector{T},\n    σ::T,\n    μ::AbstractVector{T},\n)::Nothing where {T}\n\nGiven scalar weight σ and vector of constraint weights μ, this function computes the sparse Hessian-of-the-Lagrangian matrix: sigmanabla^2 f(x) + sum_i=1^m mu_i nabla^2 g_i(x), storing the result in the vector H in the same order as the indices returned by hessian_lagrangian_structure.\n\nImplementation notes\n\nWhen implementing this method, you must not assume that H is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.eval_hessian_lagrangian_product","page":"Nonlinear programming","title":"MathOptInterface.eval_hessian_lagrangian_product","text":"eval_hessian_lagrangian_product(\n    d::AbstractNLPEvaluator,\n    h::AbstractVector{T},\n    x::AbstractVector{T},\n    v::AbstractVector{T},\n    σ::T,\n    μ::AbstractVector{T},\n)::Nothing where {T}\n\nGiven scalar weight σ and vector of constraint weights μ, computes the Hessian-of-the-Lagrangian-vector product h = left(sigmanabla^2 f(x) + sum_i=1^m mu_i nabla^2 g_i(x)right)v, storing the result in the vector h.\n\nThe vectors have dimensions such that length(h) == length(x) == length(v).\n\nImplementation notes\n\nWhen implementing this method, you must not assume that h is Vector{Float64}, but you may assume that it supports setindex! and length. For example, it may be the view of a vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.objective_expr","page":"Nonlinear programming","title":"MathOptInterface.objective_expr","text":"objective_expr(d::AbstractNLPEvaluator)::Expr\n\nReturns a Julia Expr object representing the expression graph of the objective function.\n\nFormat\n\nThe expression has a number of limitations, compared with arbitrary Julia expressions:\n\nAll sums and products are flattened out as simple Expr(:+, ...) and Expr(:*, ...) objects.\nAll decision variables must be of the form Expr(:ref, :x, MOI.VariableIndex(i)), where i is the ith variable in ListOfVariableIndices.\nThere are currently no restrictions on recognized functions; typically these will be built-in Julia functions like ^, exp, log, cos, tan, sqrt, etc., but modeling interfaces may choose to extend these basic functions, or error if they encounter unsupported functions.\n\nExamples\n\nThe expression x_1+sin(x_2exp(x_3)) is represented as\n\n:(x[MOI.VariableIndex(1)] + sin(x[MOI.VariableIndex(2)] / exp(x[MOI.VariableIndex[3]])))\n\nor equivalently\n\nExpr(\n    :call,\n    :+,\n    Expr(:ref, :x, MOI.VariableIndex(1)),\n    Expr(\n        :call,\n        :/,\n        Expr(:call, :sin, Expr(:ref, :x, MOI.VariableIndex(2))),\n        Expr(:call, :exp, Expr(:ref, :x, MOI.VariableIndex(3))),\n    ),\n)\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/nonlinear/#MathOptInterface.constraint_expr","page":"Nonlinear programming","title":"MathOptInterface.constraint_expr","text":"constraint_expr(d::AbstractNLPEvaluator, i::Integer)::Expr\n\nReturns a Julia Expr object representing the expression graph for the itextth nonlinear constraint.\n\nFormat\n\nThe format is the same as objective_expr, with an additional comparison operator indicating the sense of and bounds on the constraint.\n\nFor single-sided comparisons, the body of the constraint must be on the left-hand side, and the right-hand side must be a constant.\n\nFor double-sided comparisons (that is, l le f(x) le u), the body of the constraint must be in the middle, and the left- and right-hand sides must be constants.\n\nThe bounds on the constraints must match the NLPBoundsPairs passed to NLPBlockData.\n\nExamples\n\n:(x[MOI.VariableIndex(1)]^2 <= 1.0)\n:(x[MOI.VariableIndex(1)]^2 >= 2.0)\n:(x[MOI.VariableIndex(1)]^2 == 3.0)\n:(4.0 <= x[MOI.VariableIndex(1)]^2 <= 5.0)\n\n\n\n\n\n","category":"function"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"EditURL = \"https://github.com/jump-dev/Cbc.jl/blob/v1.2.0/README.md\"","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"(Image: )","category":"page"},{"location":"packages/Cbc/#Cbc.jl","page":"jump-dev/Cbc.jl","title":"Cbc.jl","text":"","category":"section"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Cbc.jl is a wrapper for the COIN-OR Branch and Cut (Cbc) solver.","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/Cbc/#Affiliation","page":"jump-dev/Cbc.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"This wrapper is maintained by the JuMP community and is not a COIN-OR project.","category":"page"},{"location":"packages/Cbc/#License","page":"jump-dev/Cbc.jl","title":"License","text":"","category":"section"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Cbc.jl is licensed under the MIT License.","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"The underlying solver, coin-or/Cbc, is licensed under the Eclipse public license.","category":"page"},{"location":"packages/Cbc/#Installation","page":"jump-dev/Cbc.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Install Cbc using Pkg.add:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"import Pkg\nPkg.add(\"Cbc\")","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"In addition to installing the Cbc.jl package, this will also download and install the Cbc binaries. You do not need to install Cbc separately.","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"To use a custom binary, read the Custom solver binaries section of the JuMP documentation.","category":"page"},{"location":"packages/Cbc/#Use-with-JuMP","page":"jump-dev/Cbc.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"To use Cbc with JuMP, use Cbc.Optimizer:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"using JuMP, Cbc\nmodel = Model(Cbc.Optimizer)\nset_attribute(model, \"logLevel\", 1)","category":"page"},{"location":"packages/Cbc/#MathOptInterface-API","page":"jump-dev/Cbc.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"The COIN Branch-and-Cut (Cbc) optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"MOI.Reals","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.ZeroOne\nMOI.VectorOfVariables in MOI.SOS1{Float64}\nMOI.VectorOfVariables in MOI.SOS2{Float64}","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Cbc.Status\nCbc.SecondaryStatus\nMOI.DualStatus\nMOI.NodeCount\nMOI.NumberOfVariables\nMOI.ObjectiveBound\nMOI.ObjectiveSense\nMOI.ObjectiveValue\nMOI.PrimalStatus\nMOI.RelativeGap\nMOI.ResultCount\nMOI.SolveTimeSec\nMOI.TerminationStatus","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"List of supported optimizer attributes:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Cbc.SetVariableNames\nMOI.AbsoluteGapTolerance\nMOI.NumberOfThreads\nMOI.RawOptimizerAttribute\nMOI.RelativeGapTolerance\nMOI.Silent\nMOI.SolverName\nMOI.SolverVersion\nMOI.TimeLimitSec","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"List of supported variable attributes:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"MOI.VariablePrimal\nMOI.VariablePrimalStart\nMOI.VariableName","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"List of supported constraint attributes:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"MOI.ConstraintPrimal","category":"page"},{"location":"packages/Cbc/#Options","page":"jump-dev/Cbc.jl","title":"Options","text":"","category":"section"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Options are, unfortunately, not well documented.","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"The following options are likely to be the most useful:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Parameter Example Explanation\nseconds 60.0 Solution timeout limit\nlogLevel 2 Set to 0 to disable solution output\nmaxSolutions 1 Terminate after this many feasible solutions have been found\nmaxNodes 1 Terminate after this many branch-and-bound nodes have been evaluated\nallowableGap 0.05 Terminate after optimality gap is less than this value (on an absolute scale)\nratioGap 0.05 Terminate after optimality gap is smaller than this relative fraction\nthreads 1 Set the number of threads to use for parallel branch & bound","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"The complete list of parameters can be found by running the cbc executable and typing ? at the prompt.","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"Start the cbc executable from Julia as follows:","category":"page"},{"location":"packages/Cbc/","page":"jump-dev/Cbc.jl","title":"jump-dev/Cbc.jl","text":"using Cbc_jll\nCbc_jll.cbc() do exe\n    run(`$(exe)`)\nend","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"EditURL = \"geographic_clustering.jl\"","category":"page"},{"location":"tutorials/linear/geographic_clustering/#Geographical-clustering","page":"Geographical clustering","title":"Geographical clustering","text":"","category":"section"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"This tutorial was originally contributed by Matthew Helm and Mathieu Tanneau.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"The goal of this exercise is to cluster n cities into k groups, minimizing the total pairwise distance between cities and ensuring that the variance in the total populations of each group is relatively small.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"using JuMP\nimport DataFrames\nimport HiGHS\nimport LinearAlgebra","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"For this example, we'll use the 20 most populous cities in the United States.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"cities = DataFrames.DataFrame(\n    Union{String,Float64}[\n        \"New York, NY\" 8.405 40.7127 -74.0059\n        \"Los Angeles, CA\" 3.884 34.0522 -118.2436\n        \"Chicago, IL\" 2.718 41.8781 -87.6297\n        \"Houston, TX\" 2.195 29.7604 -95.3698\n        \"Philadelphia, PA\" 1.553 39.9525 -75.1652\n        \"Phoenix, AZ\" 1.513 33.4483 -112.0740\n        \"San Antonio, TX\" 1.409 29.4241 -98.4936\n        \"San Diego, CA\" 1.355 32.7157 -117.1610\n        \"Dallas, TX\" 1.257 32.7766 -96.7969\n        \"San Jose, CA\" 0.998 37.3382 -121.8863\n        \"Austin, TX\" 0.885 30.2671 -97.7430\n        \"Indianapolis, IN\" 0.843 39.7684 -86.1580\n        \"Jacksonville, FL\" 0.842 30.3321 -81.6556\n        \"San Francisco, CA\" 0.837 37.7749 -122.4194\n        \"Columbus, OH\" 0.822 39.9611 -82.9987\n        \"Charlotte, NC\" 0.792 35.2270 -80.8431\n        \"Fort Worth, TX\" 0.792 32.7554 -97.3307\n        \"Detroit, MI\" 0.688 42.3314 -83.0457\n        \"El Paso, TX\" 0.674 31.7775 -106.4424\n        \"Memphis, TN\" 0.653 35.1495 -90.0489\n    ],\n    [\"city\", \"population\", \"lat\", \"lon\"],\n)","category":"page"},{"location":"tutorials/linear/geographic_clustering/#Model-Specifics","page":"Geographical clustering","title":"Model Specifics","text":"","category":"section"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"We will cluster these 20 cities into 3 different groups and we will assume that the ideal or target population P for a group is simply the total population of the 20 cities divided by 3:","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"n = size(cities, 1)\nk = 3\nP = sum(cities.population) / k","category":"page"},{"location":"tutorials/linear/geographic_clustering/#Obtaining-the-distances-between-each-city","page":"Geographical clustering","title":"Obtaining the distances between each city","text":"","category":"section"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"Let's compute the pairwise Haversine distance between each of the cities in our data set and store the result in a variable we'll call dm:","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"\"\"\"\n    haversine(lat1, long1, lat2, long2, r = 6372.8)\n\nCompute the haversine distance between two points on a sphere of radius `r`,\nwhere the points are given by the latitude/longitude pairs `lat1/long1` and\n`lat2/long2` (in degrees).\n\"\"\"\nfunction haversine(lat1, long1, lat2, long2, r = 6372.8)\n    lat1, long1 = deg2rad(lat1), deg2rad(long1)\n    lat2, long2 = deg2rad(lat2), deg2rad(long2)\n    hav(a, b) = sin((b - a) / 2)^2\n    inner_term = hav(lat1, lat2) + cos(lat1) * cos(lat2) * hav(long1, long2)\n    d = 2 * r * asin(sqrt(inner_term))\n    # Round distance to nearest kilometer.\n    return round(Int, d)\nend","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"Our distance matrix is symmetric so we'll convert it to a LowerTriangular matrix so that we can better interpret the objective value of our model:","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"dm = LinearAlgebra.LowerTriangular([\n    haversine(cities.lat[i], cities.lon[i], cities.lat[j], cities.lon[j])\n    for i in 1:n, j in 1:n\n])","category":"page"},{"location":"tutorials/linear/geographic_clustering/#Build-the-model","page":"Geographical clustering","title":"Build the model","text":"","category":"section"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"Now that we have the basics taken  care of, we can set up our model, create decision variables, add constraints, and then solve.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"First, we'll set up a model that leverages the Cbc solver. Next, we'll set up a binary variable x_ik that takes the value 1 if city i is in group k and 0 otherwise. Each city must be in a group, so we'll add the constraint sum_k x_ik = 1 for every i.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x[1:n, 1:k], Bin)\n@constraint(model, [i = 1:n], sum(x[i, :]) == 1);\n# To reduce symmetry, we fix the first city to belong to the first group.\nfix(x[1, 1], 1; force = true)","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"The total population of a group k is Q_k = sum_ix_ikq_i where q_i is simply the i-th value from the population column in our cities DataFrame. Let's add constraints so that alpha leq (Q_k - P) leq beta. We'll set alpha equal to -3 million and beta equal to 3. By adjusting these thresholds you'll find that there is a tradeoff between having relatively even populations between groups and having geographically close cities within each group. In other words, the larger the absolute values of alpha and beta, the closer together the cities in a group will be but the variance between the group populations will be higher.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"@variable(model, -3 <= population_diff[1:k] <= 3)\n@constraint(model, population_diff .== x' * cities.population .- P)","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"Now we need to add one last binary variable z_ij to our model that we'll use to compute the total distance between the cities in our groups, defined as sum_ijd_ijz_ij. Variable z_ij will equal 1 if cities i and j are in the same group, and 0 if they are not in the same group.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"To ensure that z_ij = 1 if and only if cities i and j are in the same group, we add the constraints z_ij geq x_ik + x_jk - 1 for every pair ij and every k:","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"@variable(model, z[i = 1:n, j = 1:i], Bin)\nfor k in 1:k, i in 1:n, j in 1:i\n    @constraint(model, z[i, j] >= x[i, k] + x[j, k] - 1)\nend","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"We can now add an objective to our model which will simply be to minimize the dot product of z and our distance matrix, dm.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"@objective(model, Min, sum(dm[i, j] * z[i, j] for i in 1:n, j in 1:i));\nnothing #hide","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"We can then call optimize! and review the results.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"optimize!(model)","category":"page"},{"location":"tutorials/linear/geographic_clustering/#Reviewing-the-Results","page":"Geographical clustering","title":"Reviewing the Results","text":"","category":"section"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"Now that we have results, we can add a column to our cities DataFrame for the group and then loop through our x variable to assign each city to its group. Once we have that, we can look at the total population for each group and also look at the cities in each group to verify that they are grouped by geographic proximity.","category":"page"},{"location":"tutorials/linear/geographic_clustering/","page":"Geographical clustering","title":"Geographical clustering","text":"cities.group = zeros(n)\n\nfor i in 1:n, j in 1:k\n    if round(Int, value(x[i, j])) == 1\n        cities.group[i] = j\n    end\nend\n\nfor group in DataFrames.groupby(cities, :group)\n    @show group\n    println(\"\")\n    @show sum(group.population)\n    println(\"\")\nend","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/objective/#Objectives","page":"Objectives","title":"Objectives","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"This page describes macros and functions related to linear and quadratic objective functions only, unless otherwise indicated. For nonlinear objective functions, see Nonlinear Modeling.","category":"page"},{"location":"manual/objective/#Set-a-linear-objective","page":"Objectives","title":"Set a linear objective","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use the @objective macro to set a linear objective function.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use Min to create a minimization objective:","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x + 1)\n2 x + 1","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use Max to create a maximization objective:","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Max, 2x + 1)\n2 x + 1","category":"page"},{"location":"manual/objective/#Set-a-quadratic-objective","page":"Objectives","title":"Set a quadratic objective","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use the @objective macro to set a quadratic objective function.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use ^2 to have a variable squared:","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, x^2 + 2x + 1)\nx² + 2 x + 1","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"You can also have bilinear terms between variables:","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> @objective(model, Max, x * y + x + y)\nx*y + x + y","category":"page"},{"location":"manual/objective/#Set-a-nonlinear-objective","page":"Objectives","title":"Set a nonlinear objective","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use the @objective macro to set a nonlinear objective function:","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x <= 1);\n\njulia> @objective(model, Max, log(x))\nlog(x)","category":"page"},{"location":"manual/objective/#Query-the-objective-function","page":"Objectives","title":"Query the objective function","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use objective_function to return the current objective function.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x + 1)\n2 x + 1\n\njulia> objective_function(model)\n2 x + 1","category":"page"},{"location":"manual/objective/#Evaluate-the-objective-function-at-a-point","page":"Objectives","title":"Evaluate the objective function at a point","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use value to evaluate an objective function at a point specifying values for variables.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @objective(model, Min, 2x[1]^2 + x[1] + 0.5*x[2])\n2 x[1]² + x[1] + 0.5 x[2]\n\njulia> f = objective_function(model)\n2 x[1]² + x[1] + 0.5 x[2]\n\njulia> point = Dict(x[1] => 2.0, x[2] => 1.0);\n\njulia> value(z -> point[z], f)\n10.5","category":"page"},{"location":"manual/objective/#Query-the-objective-sense","page":"Objectives","title":"Query the objective sense","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use objective_sense to return the current objective sense.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x + 1)\n2 x + 1\n\njulia> objective_sense(model)\nMIN_SENSE::OptimizationSense = 0","category":"page"},{"location":"manual/objective/#Modify-an-objective","page":"Objectives","title":"Modify an objective","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"To modify an objective, call @objective with the new objective function.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x)\n2 x\n\njulia> @objective(model, Max, -2x)\n-2 x","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Alternatively, use set_objective_function.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x)\n2 x\n\njulia> new_objective = @expression(model, -2 * x)\n-2 x\n\njulia> set_objective_function(model, new_objective)","category":"page"},{"location":"manual/objective/#Modify-an-objective-coefficient","page":"Objectives","title":"Modify an objective coefficient","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use set_objective_coefficient to modify an objective coefficient.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x)\n2 x\n\njulia> set_objective_coefficient(model, x, 3)\n\njulia> objective_function(model)\n3 x","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"info: Info\nThere is no way to modify the coefficient of a quadratic term. Set a new objective instead.","category":"page"},{"location":"manual/objective/#Modify-the-objective-sense","page":"Objectives","title":"Modify the objective sense","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Use set_objective_sense to modify the objective sense.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x)\n2 x\n\njulia> objective_sense(model)\nMIN_SENSE::OptimizationSense = 0\n\njulia> set_objective_sense(model, MAX_SENSE);\n\njulia> objective_sense(model)\nMAX_SENSE::OptimizationSense = 1","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Alternatively, call @objective and pass the existing objective function.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @objective(model, Min, 2x)\n2 x\n\njulia> @objective(model, Max, objective_function(model))\n2 x","category":"page"},{"location":"manual/objective/#Set-a-vector-valued-objective","page":"Objectives","title":"Set a vector-valued objective","text":"","category":"section"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Define a multi-objective optimization problem by passing a vector of objectives:","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @objective(model, Min, [1 + x[1], 2 * x[2]])\n2-element Vector{AffExpr}:\n x[1] + 1\n 2 x[2]\n\njulia> f = objective_function(model)\n2-element Vector{AffExpr}:\n x[1] + 1\n 2 x[2]","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"tip: Tip\nThe Multi-objective knapsack tutorial provides an example of solving a multi-objective integer program.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"In most cases, multi-objective optimization solvers will return multiple solutions, corresponding to points on the Pareto frontier. See Multiple solutions for information on how to query and work with multiple solutions.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Note that you must set a single objective sense, that is, you cannot have both minimization and maximization objectives. Work around this limitation by choosing Min and negating any objectives you want to maximize:","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @expression(model, obj1, 1 + x[1])\nx[1] + 1\n\njulia> @expression(model, obj2, 2 * x[1])\n2 x[1]\n\njulia> @objective(model, Min, [obj1, -obj2])\n2-element Vector{AffExpr}:\n x[1] + 1\n -2 x[1]","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"Defining your objectives as expressions allows flexibility in how you can solve variations of the same problem, with some objectives removed and constrained to be no worse that a fixed value.","category":"page"},{"location":"manual/objective/","page":"Objectives","title":"Objectives","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @expression(model, obj1, 1 + x[1])\nx[1] + 1\n\njulia> @expression(model, obj2, 2 * x[1])\n2 x[1]\n\njulia> @expression(model, obj3, x[1] + x[2])\nx[1] + x[2]\n\njulia> @objective(model, Min, [obj1, obj2, obj3])  # Three-objective problem\n3-element Vector{AffExpr}:\n x[1] + 1\n 2 x[1]\n x[1] + x[2]\n\njulia> # optimize!(model), look at the solution, talk to stakeholders, then\n       # decide you want to solve a new problem where the third objective is\n       # removed and constrained to be better than 2.0.\n       nothing\n\njulia> @objective(model, Min, [obj1, obj2])   # Two-objective problem\n2-element Vector{AffExpr}:\n x[1] + 1\n 2 x[1]\n\njulia> @constraint(model, obj3 <= 2.0)\nx[1] + x[2] ≤ 2","category":"page"},{"location":"moi/reference/callbacks/","page":"Callbacks","title":"Callbacks","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/callbacks.md\"","category":"page"},{"location":"moi/reference/callbacks/","page":"Callbacks","title":"Callbacks","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/callbacks/#Callbacks","page":"Callbacks","title":"Callbacks","text":"","category":"section"},{"location":"moi/reference/callbacks/","page":"Callbacks","title":"Callbacks","text":"AbstractCallback\nAbstractSubmittable\nsubmit","category":"page"},{"location":"moi/reference/callbacks/#MathOptInterface.AbstractCallback","page":"Callbacks","title":"MathOptInterface.AbstractCallback","text":"abstract type AbstractCallback <: AbstractModelAttribute end\n\nAbstract type for a model attribute representing a callback function. The value set to subtypes of AbstractCallback is a function that may be called during optimize!. As optimize! is in progress, the result attributes (i.e, the attributes attr such that is_set_by_optimize(attr)) may not be accessible from the callback, hence trying to get result attributes might throw a OptimizeInProgress error.\n\nAt most one callback of each type can be registered. If an optimizer already has a function for a callback type, and the user registers a new function, then the old one is replaced.\n\nThe value of the attribute should be a function taking only one argument, commonly called callback_data, that can be used for instance in LazyConstraintCallback, HeuristicCallback and UserCutCallback.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.AbstractSubmittable","page":"Callbacks","title":"MathOptInterface.AbstractSubmittable","text":"AbstractSubmittable\n\nAbstract supertype for objects that can be submitted to the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.submit","page":"Callbacks","title":"MathOptInterface.submit","text":"submit(\n    optimizer::AbstractOptimizer,\n    sub::AbstractSubmittable,\n    values...,\n)::Nothing\n\nSubmit values to the submittable sub of the optimizer optimizer.\n\nAn UnsupportedSubmittable error is thrown if model does not support the attribute attr (see supports) and a SubmitNotAllowed error is thrown if it supports the submittable sub but it cannot be submitted.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/callbacks/#Attributes","page":"Callbacks","title":"Attributes","text":"","category":"section"},{"location":"moi/reference/callbacks/","page":"Callbacks","title":"Callbacks","text":"CallbackNodeStatus\nCallbackVariablePrimal\nCallbackNodeStatusCode\nCALLBACK_NODE_STATUS_INTEGER\nCALLBACK_NODE_STATUS_FRACTIONAL\nCALLBACK_NODE_STATUS_UNKNOWN","category":"page"},{"location":"moi/reference/callbacks/#MathOptInterface.CallbackNodeStatus","page":"Callbacks","title":"MathOptInterface.CallbackNodeStatus","text":"CallbackNodeStatus(callback_data)\n\nAn optimizer attribute describing the (in)feasibility of the primal solution available from CallbackVariablePrimal during a callback identified by callback_data.\n\nReturns a CallbackNodeStatusCode Enum.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.CallbackVariablePrimal","page":"Callbacks","title":"MathOptInterface.CallbackVariablePrimal","text":"CallbackVariablePrimal(callback_data)\n\nA variable attribute for the assignment to some primal variable's value during the callback identified by callback_data.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.CallbackNodeStatusCode","page":"Callbacks","title":"MathOptInterface.CallbackNodeStatusCode","text":"CallbackNodeStatusCode\n\nAn Enum of possible return values from calling get with CallbackNodeStatus.\n\nValues\n\nPossible values are:\n\nCALLBACK_NODE_STATUS_INTEGER: the primal solution available from CallbackVariablePrimal is integer feasible.\nCALLBACK_NODE_STATUS_FRACTIONAL: the primal solution available from CallbackVariablePrimal is integer infeasible.\nCALLBACK_NODE_STATUS_UNKNOWN: the primal solution available from CallbackVariablePrimal might be integer feasible or infeasible.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.CALLBACK_NODE_STATUS_INTEGER","page":"Callbacks","title":"MathOptInterface.CALLBACK_NODE_STATUS_INTEGER","text":"CALLBACK_NODE_STATUS_INTEGER::CallbackNodeStatusCode\n\nAn instance of the CallbackNodeStatusCode enum.\n\nCALLBACK_NODE_STATUS_INTEGER: the primal solution available from CallbackVariablePrimal is integer feasible.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/callbacks/#MathOptInterface.CALLBACK_NODE_STATUS_FRACTIONAL","page":"Callbacks","title":"MathOptInterface.CALLBACK_NODE_STATUS_FRACTIONAL","text":"CALLBACK_NODE_STATUS_FRACTIONAL::CallbackNodeStatusCode\n\nAn instance of the CallbackNodeStatusCode enum.\n\nCALLBACK_NODE_STATUS_FRACTIONAL: the primal solution available from CallbackVariablePrimal is integer infeasible.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/callbacks/#MathOptInterface.CALLBACK_NODE_STATUS_UNKNOWN","page":"Callbacks","title":"MathOptInterface.CALLBACK_NODE_STATUS_UNKNOWN","text":"CALLBACK_NODE_STATUS_UNKNOWN::CallbackNodeStatusCode\n\nAn instance of the CallbackNodeStatusCode enum.\n\nCALLBACK_NODE_STATUS_UNKNOWN: the primal solution available from CallbackVariablePrimal might be integer feasible or infeasible.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/callbacks/#Lazy-constraints","page":"Callbacks","title":"Lazy constraints","text":"","category":"section"},{"location":"moi/reference/callbacks/","page":"Callbacks","title":"Callbacks","text":"LazyConstraintCallback\nLazyConstraint","category":"page"},{"location":"moi/reference/callbacks/#MathOptInterface.LazyConstraintCallback","page":"Callbacks","title":"MathOptInterface.LazyConstraintCallback","text":"LazyConstraintCallback() <: AbstractCallback\n\nThe callback can be used to reduce the feasible set given the current primal solution by submitting a LazyConstraint. For instance, it may be called at an incumbent of a mixed-integer problem. Note that there is no guarantee that the callback is called at every feasible primal solution.\n\nThe current primal solution is accessed through CallbackVariablePrimal. Trying to access other result attributes will throw OptimizeInProgress as discussed in AbstractCallback.\n\nExamples\n\nx = MOI.add_variables(optimizer, 8)\nMOI.set(optimizer, MOI.LazyConstraintCallback(), callback_data -> begin\n    sol = MOI.get(optimizer, MOI.CallbackVariablePrimal(callback_data), x)\n    if # should add a lazy constraint\n        func = # computes function\n        set = # computes set\n        MOI.submit(optimizer, MOI.LazyConstraint(callback_data), func, set)\n    end\nend)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.LazyConstraint","page":"Callbacks","title":"MathOptInterface.LazyConstraint","text":"LazyConstraint(callback_data)\n\nLazy constraint func-in-set submitted as func, set. The optimal solution returned by VariablePrimal will satisfy all lazy constraints that have been submitted.\n\nThis can be submitted only from the LazyConstraintCallback. The field callback_data is a solver-specific callback type that is passed as the argument to the feasible solution callback.\n\nExamples\n\nSuppose x and y are VariableIndexs of optimizer. To add a LazyConstraint for 2x + 3y <= 1, write\n\nfunc = 2.0x + 3.0y\nset = MOI.LessThan(1.0)\nMOI.submit(optimizer, MOI.LazyConstraint(callback_data), func, set)\n\ninside a LazyConstraintCallback of data callback_data.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#User-cuts","page":"Callbacks","title":"User cuts","text":"","category":"section"},{"location":"moi/reference/callbacks/","page":"Callbacks","title":"Callbacks","text":"UserCutCallback\nUserCut","category":"page"},{"location":"moi/reference/callbacks/#MathOptInterface.UserCutCallback","page":"Callbacks","title":"MathOptInterface.UserCutCallback","text":"UserCutCallback() <: AbstractCallback\n\nThe callback can be used to submit UserCut given the current primal solution. For instance, it may be called at fractional (i.e., non-integer) nodes in the branch and bound tree of a mixed-integer problem. Note that there is not guarantee that the callback is called everytime the solver has an infeasible solution.\n\nThe infeasible solution is accessed through CallbackVariablePrimal. Trying to access other result attributes will throw OptimizeInProgress as discussed in AbstractCallback.\n\nExamples\n\nx = MOI.add_variables(optimizer, 8)\nMOI.set(optimizer, MOI.UserCutCallback(), callback_data -> begin\n    sol = MOI.get(optimizer, MOI.CallbackVariablePrimal(callback_data), x)\n    if # can find a user cut\n        func = # computes function\n        set = # computes set\n        MOI.submit(optimizer, MOI.UserCut(callback_data), func, set)\n    end\nend\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.UserCut","page":"Callbacks","title":"MathOptInterface.UserCut","text":"UserCut(callback_data)\n\nConstraint func-to-set suggested to help the solver detect the solution given by CallbackVariablePrimal as infeasible. The cut is submitted as func, set. Typically CallbackVariablePrimal will violate integrality constraints, and a cut would be of the form ScalarAffineFunction-in-LessThan or ScalarAffineFunction-in-GreaterThan. Note that, as opposed to LazyConstraint, the provided constraint cannot modify the feasible set, the constraint should be redundant, e.g., it may be a consequence of affine and integrality constraints.\n\nThis can be submitted only from the UserCutCallback. The field callback_data is a solver-specific callback type that is passed as the argument to the infeasible solution callback.\n\nNote that the solver may silently ignore the provided constraint.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#Heuristic-solutions","page":"Callbacks","title":"Heuristic solutions","text":"","category":"section"},{"location":"moi/reference/callbacks/","page":"Callbacks","title":"Callbacks","text":"HeuristicCallback\nHeuristicSolution\nHeuristicSolutionStatus\nHEURISTIC_SOLUTION_ACCEPTED\nHEURISTIC_SOLUTION_REJECTED\nHEURISTIC_SOLUTION_UNKNOWN","category":"page"},{"location":"moi/reference/callbacks/#MathOptInterface.HeuristicCallback","page":"Callbacks","title":"MathOptInterface.HeuristicCallback","text":"HeuristicCallback() <: AbstractCallback\n\nThe callback can be used to submit HeuristicSolution given the current primal solution. For example, it may be called at fractional (i.e., non-integer) nodes in the branch and bound tree of a mixed-integer problem. Note that there is no guarantee that the callback is called every time the solver has an infeasible solution.\n\nThe current primal solution is accessed through CallbackVariablePrimal. Trying to access other result attributes will throw OptimizeInProgress as discussed in AbstractCallback.\n\nExamples\n\nx = MOI.add_variables(optimizer, 8)\nMOI.set(optimizer, MOI.HeuristicCallback(), callback_data -> begin\n    sol = MOI.get(optimizer, MOI.CallbackVariablePrimal(callback_data), x)\n    if # can find a heuristic solution\n        values = # computes heuristic solution\n        MOI.submit(optimizer, MOI.HeuristicSolution(callback_data), x,\n                   values)\n    end\nend\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.HeuristicSolution","page":"Callbacks","title":"MathOptInterface.HeuristicSolution","text":"HeuristicSolution(callback_data)\n\nHeuristically obtained feasible solution. The solution is submitted as variables, values where values[i] gives the value of variables[i], similarly to set. The submit call returns a HeuristicSolutionStatus indicating whether the provided solution was accepted or rejected.\n\nThis can be submitted only from the HeuristicCallback. The field callback_data is a solver-specific callback type that is passed as the argument to the heuristic callback.\n\nSome solvers require a complete solution, others only partial solutions.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.HeuristicSolutionStatus","page":"Callbacks","title":"MathOptInterface.HeuristicSolutionStatus","text":"HeuristicSolutionStatus\n\nAn Enum of possible return values for submit with HeuristicSolution. This informs whether the heuristic solution was accepted or rejected.\n\nValues\n\nPossible values are:\n\nHEURISTIC_SOLUTION_ACCEPTED: The heuristic solution was accepted\nHEURISTIC_SOLUTION_REJECTED: The heuristic solution was rejected\nHEURISTIC_SOLUTION_UNKNOWN: No information available on the acceptance\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/callbacks/#MathOptInterface.HEURISTIC_SOLUTION_ACCEPTED","page":"Callbacks","title":"MathOptInterface.HEURISTIC_SOLUTION_ACCEPTED","text":"HEURISTIC_SOLUTION_ACCEPTED::HeuristicSolutionStatus\n\nAn instance of the HeuristicSolutionStatus enum.\n\nHEURISTIC_SOLUTION_ACCEPTED: The heuristic solution was accepted\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/callbacks/#MathOptInterface.HEURISTIC_SOLUTION_REJECTED","page":"Callbacks","title":"MathOptInterface.HEURISTIC_SOLUTION_REJECTED","text":"HEURISTIC_SOLUTION_REJECTED::HeuristicSolutionStatus\n\nAn instance of the HeuristicSolutionStatus enum.\n\nHEURISTIC_SOLUTION_REJECTED: The heuristic solution was rejected\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/callbacks/#MathOptInterface.HEURISTIC_SOLUTION_UNKNOWN","page":"Callbacks","title":"MathOptInterface.HEURISTIC_SOLUTION_UNKNOWN","text":"HEURISTIC_SOLUTION_UNKNOWN::HeuristicSolutionStatus\n\nAn instance of the HeuristicSolutionStatus enum.\n\nHEURISTIC_SOLUTION_UNKNOWN: No information available on the acceptance\n\n\n\n\n\n","category":"constant"},{"location":"packages/SDDP/","page":"odow/SDDP.jl","title":"odow/SDDP.jl","text":"EditURL = \"https://github.com/odow/SDDP.jl/blob/v1.6.3/README.md\"","category":"page"},{"location":"packages/SDDP/","page":"odow/SDDP.jl","title":"odow/SDDP.jl","text":"<img src=\"https://raw.githubusercontent.com/odow/SDDP.jl/e9de84e0a4b57374bd9e0c95148da1501816e4c5/docs/src/assets/logo.png?raw=true\" alt=\"logo\" width=\"100px\"/>","category":"page"},{"location":"packages/SDDP/#SDDP.jl","page":"odow/SDDP.jl","title":"SDDP.jl","text":"","category":"section"},{"location":"packages/SDDP/","page":"odow/SDDP.jl","title":"odow/SDDP.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/SDDP/","page":"odow/SDDP.jl","title":"odow/SDDP.jl","text":"SDDP.jl is a JuMP extension for solving large convex multistage stochastic programming problems using stochastic dual dynamic programming.","category":"page"},{"location":"packages/SDDP/#License","page":"odow/SDDP.jl","title":"License","text":"","category":"section"},{"location":"packages/SDDP/","page":"odow/SDDP.jl","title":"odow/SDDP.jl","text":"SDDP.jl is licensed under the MPL 2.0 license.","category":"page"},{"location":"packages/SDDP/#Documentation","page":"odow/SDDP.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/SDDP/","page":"odow/SDDP.jl","title":"odow/SDDP.jl","text":"You can find the documentation at https://odow.github.io/SDDP.jl/stable/.","category":"page"},{"location":"packages/SDDP/#Help","page":"odow/SDDP.jl","title":"Help","text":"","category":"section"},{"location":"packages/SDDP/","page":"odow/SDDP.jl","title":"odow/SDDP.jl","text":"If you need help, please open a GitHub issue.","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"EditURL = \"https://github.com/plasmo-dev/Plasmo.jl/blob/37f80fe96781dc647b3a5e0ed26b9eb0a994bc4a/README.md\"","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"<img src=\"https://github.com/plasmo-dev/Plasmo.jl/blob/main/docs/plasmo_logo.svg?raw=true\"/>","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"(Image: CI) (Image: codecov) (Image: ) (Image: DOI)","category":"page"},{"location":"packages/Plasmo/#Plasmo.jl","page":"plasmo-dev/Plasmo.jl","title":"Plasmo.jl","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"Plasmo.jl (Platform for Scalable Modeling and Optimization) is a graph-based algebraic modeling framework that adopts a modular style to create mathematical optimization problems and manage distributed and hierarchical structures. The package has been developed as a JuMP extension and consequently supports  most JuMP syntax and functions. ","category":"page"},{"location":"packages/Plasmo/#Overview","page":"plasmo-dev/Plasmo.jl","title":"Overview","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"The core data structure in Plasmo.jl is the OptiGraph. The optigraph contains a set of optinodes which represent self-contained optimization problems and optiedges that represent coupling between optinodes (which produces an underlying hypergraph structure of optinodes and optiedges). Optigraphs can further be embedded within other optigraphs to create nested hierarchical graph structures. The graph structures obtained using Plasmo.jl can be used for simple model and data management, but they can also be used to perform graph partitioning or develop interfaces to structured optimization solvers.","category":"page"},{"location":"packages/Plasmo/#License","page":"plasmo-dev/Plasmo.jl","title":"License","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"Plasmo is licensed under the MPL 2.0 license.","category":"page"},{"location":"packages/Plasmo/#Installation","page":"plasmo-dev/Plasmo.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"Install Plasmo using Pkg.add:","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"import Pkg\nPkg.add(\"Plasmo\")","category":"page"},{"location":"packages/Plasmo/#Documentation","page":"plasmo-dev/Plasmo.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"The latest documentation is available through GitHub Pages. Additional examples can be found in the examples folder.","category":"page"},{"location":"packages/Plasmo/#Simple-Example","page":"plasmo-dev/Plasmo.jl","title":"Simple Example","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"using Plasmo\nusing Ipopt\n\n#create an optigraph\ngraph = OptiGraph()\n\n#add nodes to an optigraph\n@optinode(graph, n1)\n@optinode(graph, n2)\n\n#add variables, constraints, and objective functions to nodes\n@variable(n1, 0 <= x <= 2)\n@variable(n1, 0 <= y <= 3)\n@constraint(n1, x+y <= 4)\n@objective(n1, Min, x)\n\n@variable(n2,x)\n@NLconstraint(n2, exp(x) >= 2)\n\n#add a linkconstraint to couple nodes\n@linkconstraint(graph, n1[:x] == n2[:x])\n\n#optimize with Ipopt\nset_optimizer(graph, Ipopt.Optimizer)\noptimize!(graph)\n\n#Print solution values\nprintln(\"n1[:x] = \", value(n1[:x]))\nprintln(\"n2[:x] = \", value(n2[:x]))","category":"page"},{"location":"packages/Plasmo/#Acknowledgments","page":"plasmo-dev/Plasmo.jl","title":"Acknowledgments","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"This code is based on work supported by the following funding agencies:","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"U.S. Department of Energy (DOE), Office of Science, under Contract No. DE-AC02-06CH11357\nDOE Office of Electricity Delivery and Energy Reliability’s Advanced Grid Research and Development program at Argonne National Laboratory\nNational Science Foundation under award NSF-EECS-1609183 and under award CBET-1748516","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"The primary developer is Jordan Jalving (@jalving) with support from the following contributors.  ","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"Victor Zavala (University of Wisconsin-Madison)\nYankai Cao (University of British Columbia)\nKibaek Kim (Argonne National Laboratory)\nSungho Shin (University of Wisconsin-Madison)","category":"page"},{"location":"packages/Plasmo/#Citing-Plasmo.jl","page":"plasmo-dev/Plasmo.jl","title":"Citing Plasmo.jl","text":"","category":"section"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"If you find Plasmo.jl useful for your work, you may cite the manuscript as:","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"@article{JalvingShinZavala2022,\n  title={A Graph-Based Modeling Abstraction for Optimization: Concepts and Implementation in Plasmo.jl},\n  author={Jordan Jalving and Sungho Shin and Victor M. Zavala},\n  journal={Mathematical Programming Computation},\n  year={2022},\n  volume={14},\n  pages={699 - 747}\n}","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"There is also a freely available pre-print:","category":"page"},{"location":"packages/Plasmo/","page":"plasmo-dev/Plasmo.jl","title":"plasmo-dev/Plasmo.jl","text":"@misc{JalvingShinZavala2020,\ntitle = {A Graph-Based Modeling Abstraction for Optimization: Concepts and Implementation in Plasmo.jl},\nauthor = {Jordan Jalving and Sungho Shin and Victor M. Zavala},\nyear = {2020},\neprint = {2006.05378},\narchivePrefix = {arXiv},\nprimaryClass = {math.OC}\n}","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"EditURL = \"https://github.com/jump-dev/MiniZinc.jl/blob/v0.3.2/README.md\"","category":"page"},{"location":"packages/MiniZinc/#MiniZinc.jl","page":"jump-dev/MiniZinc.jl","title":"MiniZinc.jl","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MiniZinc.jl is a wrapper for the MiniZinc constraint modeling language.","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"It provides a way to write MathOptInterface models to .mzn files, and a way to interact with libminizinc.","category":"page"},{"location":"packages/MiniZinc/#Affiliation","page":"jump-dev/MiniZinc.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"This wrapper is maintained by the JuMP community and is not part of the MiniZinc project.","category":"page"},{"location":"packages/MiniZinc/#License","page":"jump-dev/MiniZinc.jl","title":"License","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MiniZinc.jl is licensed under the MIT License.","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"The underlying project, MiniZinc/libminizinc, is licensed under the MPL 2.0 license.","category":"page"},{"location":"packages/MiniZinc/#Install","page":"jump-dev/MiniZinc.jl","title":"Install","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"Install MiniZinc.jl using the Julia package manager:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"import Pkg\nPkg.add(\"MiniZinc\")","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"Windows","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"On Linux and macOS, this package automatically installs libminizinc. However, we're still working out problems with the install on Windows. To use MiniZinc.jl, you'll need to manually install a copy of libminizinc from minizinc.org or compile one yourself from MiniZinc/libminizinc.","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"To teach MiniZinc.jl where to look for libminizinc, set the JULIA_LIBMINIZINC_DIR environment variable. For example:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"ENV[\"JULIA_LIBMINIZINC_DIR\"] = \"C:\\\\Program Files\\\\MiniZinc\"","category":"page"},{"location":"packages/MiniZinc/#Use-with-MathOptInterface","page":"jump-dev/MiniZinc.jl","title":"Use with MathOptInterface","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MiniZinc.jl supports the constraint programming sets defined in MathOptInterface, as well as (in)equality constraints.","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"The following example solves the following constraint program:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"xᵢ ∈ {1, 2, 3} ∀i=1,2,3\nzⱼ ∈ {0, 1}    ∀j=1,2\nz₁ <-> x₁ != x₂\nz₂ <-> x₂ != x₃\nz₁ + z₂ = 1","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"julia> import MiniZinc\n\njulia> const MOI = MiniZinc.MOI\nMathOptInterface\n\njulia> function main()\n           model = MOI.Utilities.CachingOptimizer(\n               MiniZinc.Model{Int}(),\n               MiniZinc.Optimizer{Int}(\"chuffed\"),\n           )\n           # xᵢ ∈ {1, 2, 3} ∀i=1,2,3\n           x = MOI.add_variables(model, 3)\n           MOI.add_constraint.(model, x, MOI.Interval(1, 3))\n           MOI.add_constraint.(model, x, MOI.Integer())\n           # zⱼ ∈ {0, 1}    ∀j=1,2\n           z = MOI.add_variables(model, 2)\n           MOI.add_constraint.(model, z, MOI.ZeroOne())\n           # z₁ <-> x₁ != x₂\n           MOI.add_constraint(\n               model,\n               MOI.VectorOfVariables([z[1], x[1], x[2]]),\n               MOI.Reified(MOI.AllDifferent(2)),\n           )\n           # z₂ <-> x₂ != x₃\n           MOI.add_constraint(\n               model,\n               MOI.VectorOfVariables([z[2], x[2], x[3]]),\n               MOI.Reified(MOI.AllDifferent(2)),\n           )\n           # z₁ + z₂ = 1\n           MOI.add_constraint(model, 1 * z[1] + x[2], MOI.EqualTo(1))\n           MOI.optimize!(model)\n           x_star = MOI.get(model, MOI.VariablePrimal(), x)\n           z_star = MOI.get(model, MOI.VariablePrimal(), z)\n           return x_star, z_star\n       end\nmain (generic function with 1 method)\n\njulia> main()\n([1, 1, 3], [0, 1])","category":"page"},{"location":"packages/MiniZinc/#Use-with-JuMP","page":"jump-dev/MiniZinc.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"You can also call MiniZinc from JuMP, using any solver that libminizinc supports. By default, MiniZinc.jl is compiled with \"highs\":","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"using JuMP\nimport MiniZinc\nmodel = Model(() -> MiniZinc.Optimizer{Float64}(\"highs\"))\n@variable(model, 1 <= x[1:3] <= 3, Int)\n@constraint(model, x in MOI.AllDifferent(3))\n@objective(model, Max, sum(i * x[i] for i in 1:3))\noptimize!(model)\n@show value.(x)","category":"page"},{"location":"packages/MiniZinc/#MathOptInterface-API","page":"jump-dev/MiniZinc.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"The MiniZinc Optimizer{T} supports the following constraints and attributes.","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{T}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{T}}\nMOI.ObjectiveFunction{MOI.VariableIndex}","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MOI.Reals","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\nMOI.ScalarAffineFunction{T} in MOI.GreaterThan{T}\nMOI.ScalarAffineFunction{T} in MOI.Integer\nMOI.ScalarAffineFunction{T} in MOI.Interval{T}\nMOI.ScalarAffineFunction{T} in MOI.LessThan{T}\nMOI.ScalarAffineFunction{T} in MOI.ZeroOne\nMOI.VariableIndex in MOI.EqualTo{T}\nMOI.VariableIndex in MOI.GreaterThan{T}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{T}\nMOI.VariableIndex in MOI.LessThan{T}\nMOI.VariableIndex in MOI.Parameter{T}\nMOI.VariableIndex in MOI.Semicontinuous{T}\nMOI.VariableIndex in MOI.Semiinteger{T}\nMOI.VariableIndex in MOI.ZeroOne\nMOI.VectorOfVariables in MOI.AllDifferent\nMOI.VectorOfVariables in MOI.BinPacking{T}\nMOI.VectorOfVariables in MOI.Circuit\nMOI.VectorOfVariables in MOI.CountAtLeast\nMOI.VectorOfVariables in MOI.CountBelongs\nMOI.VectorOfVariables in MOI.CountDistinct\nMOI.VectorOfVariables in MOI.CountGreaterThan\nMOI.VectorOfVariables in MOI.Cumulative\nMOI.VectorOfVariables in MOI.Path\nMOI.VectorOfVariables in MOI.Table{T}","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MOI.NLPBlock()\nMOI.Name()\nMOI.ObjectiveSense()","category":"page"},{"location":"packages/MiniZinc/#Options","page":"jump-dev/MiniZinc.jl","title":"Options","text":"","category":"section"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"Set options using MOI.RawOptimizerAttribute in MOI or set_attribute in JuMP.","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"MiniZinc.jl supports the following options:","category":"page"},{"location":"packages/MiniZinc/","page":"jump-dev/MiniZinc.jl","title":"jump-dev/MiniZinc.jl","text":"model_filename::String = \"\": the location at which to write out the .mzn file during optimization. This option can be helpful during debugging. If left empty, a temporary file will be used instead.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"EditURL = \"transp.jl\"","category":"page"},{"location":"tutorials/linear/transp/#The-transportation-problem","page":"The transportation problem","title":"The transportation problem","text":"","category":"section"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"This tutorial was originally contributed by Louis Luangkesorn.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"This tutorial is an adaptation of the transportation problem described in AMPL: A Modeling Language for Mathematical Programming, by R. Fourer, D.M. Gay and B.W. Kernighan.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"The purpose of this tutorial is to demonstrate how to create a JuMP model from an ad-hoc structured text file.","category":"page"},{"location":"tutorials/linear/transp/#Required-packages","page":"The transportation problem","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"using JuMP\nimport DelimitedFiles\nimport HiGHS","category":"page"},{"location":"tutorials/linear/transp/#Formulation","page":"The transportation problem","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"Suppose that we have a set of factories that produce pogo sticks, and a set of retail stores in which to sell them. Each factory has a maximum number of pogo sticks that it can produce, and each retail store has a demand of pogo sticks that it can sell.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"In the transportation problem, we want to choose the number of pogo sticks to make and ship from each factory to each retail store that minimizes the total shipping cost.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"Mathematically, we represent our set of factories by a set of origins i in O and our retail stores by a set of destinations j in D. The maximum supply at each factory is s_i and the demand from each retail store is d_j. The cost of shipping one pogo stick from i to j is c_ij.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"With a little effort, we can model the transportation problem as the following linear program:","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"beginaligned\nmin   sum_i in O j in D c_ij x_ij \nst   sum_j in D x_i j le s_i  forall i in O \n       sum_i in O x_i j = d_j  forall j in D \n       x_i j ge 0  forall i in O j in D\nendaligned","category":"page"},{"location":"tutorials/linear/transp/#Data","page":"The transportation problem","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"We assume our data is in the form of a text file that has the following form. In practice, we would obtain this text file from the user as input, but for the purpose of this tutorial we're going to create it from Julia.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"open(joinpath(@__DIR__, \"transp.txt\"), \"w\") do io\n    print(\n        io,\n        \"\"\"\n             . FRA  DET LAN WIN  STL  FRE  LAF SUPPLY\n          GARY  39   14  11  14   16   82    8   1400\n          CLEV  27    .  12   .   26   95   17   2600\n          PITT  24   14  17  13   28   99   20   2900\n        DEMAND 900 1200 600 400 1700 1100 1000      0\n        \"\"\",\n    )\n    return\nend","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"Here the rows are the origins, the columns are the destinations, and the values are the cost of shipping one pogo stick from the origin to the destination. If pogo stick cannot be transported from a source to a destination, then the value is .. The final row and final column are the demand and supply of each location respectively.","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"We didn't account for arcs which do not exist in our formulation, but we can make a small change and fix x_ij = 0 if c_ij = .","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"Our first step is to convert this text format into an appropriate Julia datastructure that we can work with. Since our data is tabular with named rows and columns, one option is JuMP's Containers.DenseAxisArray object:","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"function read_data(filename::String)\n    data = DelimitedFiles.readdlm(filename)\n    rows, columns = data[2:end, 1], data[1, 2:end]\n    return Containers.DenseAxisArray(data[2:end, 2:end], rows, columns)\nend\n\ndata = read_data(joinpath(@__DIR__, \"transp.txt\"))","category":"page"},{"location":"tutorials/linear/transp/#JuMP-formulation","page":"The transportation problem","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"Following Design patterns for larger models, we code our JuMP model as a function which takes in an input. In this example, we print the output to stdout:","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"function solve_transportation_problem(data::Containers.DenseAxisArray)\n    # Get the set of supplies and demands\n    O, D = axes(data)\n    # Drop the SUPPLY and DEMAND nodes from our sets\n    O, D = setdiff(O, [\"DEMAND\"]), setdiff(D, [\"SUPPLY\"])\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    @variable(model, x[o in O, d in D] >= 0)\n    # Remove arcs with \".\" cost by fixing them to 0.0.\n    for o in O, d in D\n        if data[o, d] == \".\"\n            fix(x[o, d], 0.0; force = true)\n        end\n    end\n    @objective(\n        model,\n        Min,\n        sum(data[o, d] * x[o, d] for o in O, d in D if data[o, d] != \".\"),\n    )\n    @constraint(model, [o in O], sum(x[o, :]) <= data[o, \"SUPPLY\"])\n    @constraint(model, [d in D], sum(x[:, d]) == data[\"DEMAND\", d])\n    optimize!(model)\n    # Pretty print the solution in the format of the input\n    print(\"    \", join(lpad.(D, 7, ' ')))\n    for o in O\n        print(\"\\n\", o)\n        for d in D\n            if isapprox(value(x[o, d]), 0.0; atol = 1e-6)\n                print(\"      .\")\n            else\n                print(\" \", lpad(value(x[o, d]), 6, ' '))\n            end\n        end\n    end\n    return\nend","category":"page"},{"location":"tutorials/linear/transp/#Solution","page":"The transportation problem","title":"Solution","text":"","category":"section"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"Let's solve and view the solution:","category":"page"},{"location":"tutorials/linear/transp/","page":"The transportation problem","title":"The transportation problem","text":"solve_transportation_problem(data)","category":"page"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/developer/checklists.md\"","category":"page"},{"location":"moi/developer/checklists/#Checklists","page":"Checklists","title":"Checklists","text":"","category":"section"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"The purpose of this page is to collate a series of checklists for commonly performed changes to the source code of MathOptInterface.","category":"page"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"In each case, copy the checklist into the description of the pull request.","category":"page"},{"location":"moi/developer/checklists/#Making-a-release","page":"Checklists","title":"Making a release","text":"","category":"section"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"Use this checklist when making a release of the MathOptInterface repository.","category":"page"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"## Basic\n\n - [ ] `version` field of `Project.toml` has been updated\n       - If a breaking change, increment the MAJOR field and reset others to 0\n       - If adding new features, increment the MINOR field and reset PATCH to 0\n       - If adding bug fixes or documentation changes, increment the PATCH field\n\n## Documentation\n\n - [ ] Add a new entry to `docs/src/changelog.md`, following existing style\n\n## Tests\n\n - [ ] The `solver-tests.yml` GitHub action does not have unexpected failures.\n       To run the action, go to:\n       https://github.com/jump-dev/MathOptInterface.jl/actions/workflows/solver-tests.yml\n       and click \"Run workflow\"","category":"page"},{"location":"moi/developer/checklists/#Adding-a-new-set","page":"Checklists","title":"Adding a new set","text":"","category":"section"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"Use this checklist when adding a new set to the MathOptInterface repository.","category":"page"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"## Basic\n\n - [ ] Add a new `AbstractScalarSet` or `AbstractVectorSet` to `src/sets.jl`\n - [ ] If `isbitstype(S) == false`, implement `Base.copy(set::S)`\n - [ ] If `isbitstype(S) == false`, implement `Base.:(==)(x::S, y::S)`\n - [ ] If an `AbstractVectorSet`, implement `dimension(set::S)`, unless the\n       dimension is given by `set.dimension`.\n\n## Utilities\n\n - [ ] If an `AbstractVectorSet`, implement `Utilities.set_dot`,\n       unless the dot product between two vectors in the set is equivalent to\n       `LinearAlgebra.dot`\n - [ ] If an `AbstractVectorSet`, implement `Utilities.set_with_dimension` in\n       `src/Utilities/matrix_of_constraints.jl`\n - [ ] Add the set to the `@model` macro at the bottom of `src/Utilities.model.jl`\n\n## Documentation\n\n - [ ] Add a docstring, which gives the mathematical definition of the set,\n       along with an `## Example` block containing a `jldoctest`\n - [ ] Add the docstring to `docs/src/reference/standard_form.md`\n - [ ] Add the set to the relevant table in `docs/src/manual/standard_form.md`\n\n## Tests\n\n - [ ] Define a new `_set(::Type{S})` method in `src/Test/test_basic_constraint.jl`\n       and add the name of the set to the list at the bottom of that files\n - [ ] If the set has any checks in its constructor, add tests to `test/sets.jl`\n\n## MathOptFormat\n\n - [ ] Open an issue at `https://github.com/jump-dev/MathOptFormat` to add\n       support for the new set {{ replace with link to the issue }}\n\n## Optional\n\n - [ ] Implement `dual_set(::S)` and `dual_set_type(::Type{S})`\n - [ ] Add new tests to the `Test` submodule exercising your new set\n - [ ] Add new bridges to convert your set into more commonly used sets","category":"page"},{"location":"moi/developer/checklists/#Adding-a-new-bridge","page":"Checklists","title":"Adding a new bridge","text":"","category":"section"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"Use this checklist when adding a new bridge to the MathOptInterface repository.","category":"page"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"The steps are mostly the same, but locations depend on whether the bridge is a Constraint, Objective, or Variable bridge. In each case below, replace XXX with the appropriate type of bridge.","category":"page"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"## Basic\n\n - [ ] Create a new file in `src/Bridges/XXX/bridges`\n - [ ] Define the bridge, following existing examples. The name of the bridge\n       struct must end in `Bridge`\n - [ ] Check if your bridge can be a subtype of [`MOI.Bridges.Constraint.SetMapBridge`](@ref)\n - [ ] Define a new `const` that is a `SingleBridgeOptimizer` wrapping the\n       new bridge. The name of the const must be the name of the bridge, less\n       the `Bridge` suffix\n - [ ] `include` the file in `src/Bridges/XXX/bridges/XXX.jl`\n - [ ] If the bridge should be enabled by default, add the bridge to\n       `add_all_bridges` at the bottom of `src/Bridges/XXX/XXX.jl`\n\n## Tests\n\n - [ ] Create a new file in the appropriate subdirectory of `tests/Bridges/XXX`\n - [ ] Use `MOI.Bridges.runtests` to test various inputs and outputs of the\n       bridge\n - [ ] If, after opening the pull request to add the bridge, some lines are not\n       covered by the tests, add additional bridge-specific tests to cover the\n       untested lines.\n\n## Documentation\n\n - [ ] Add a docstring which uses the same template as existing bridges.\n - [ ] Add the docstring to `docs/src/submodules/Bridges/list_of_bridges.md`\n\n## Final touch\n\nIf the bridge depends on run-time values of other variables and constraints in\nthe model:\n\n - [ ] Implement `MOI.Utilities.needs_final_touch(::Bridge)`\n - [ ] Implement `MOI.Utilities.final_touch(::Bridge, ::MOI.ModelLike)`\n - [ ] Ensure that `final_touch` can be called multiple times in a row","category":"page"},{"location":"moi/developer/checklists/#Updating-MathOptFormat","page":"Checklists","title":"Updating MathOptFormat","text":"","category":"section"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"Use this checklist when updating the version of MathOptFormat.","category":"page"},{"location":"moi/developer/checklists/","page":"Checklists","title":"Checklists","text":"## Basic\n\n - [ ] The file at `src/FileFormats/MOF/mof.X.Y.schema.json` is updated\n - [ ] The constants `SCHEMA_PATH`, `VERSION`, and `SUPPORTED_VERSIONS` are\n       updated in `src/FileFormats/MOF/MOF.jl`\n\n## New sets\n\n - [ ] New sets are added to the `@model` in `src/FileFormats/MOF/MOF.jl`\n - [ ] New sets are added to the `@enum` in `src/FileFormats/MOF/read.jl`\n - [ ] `set_to_moi` is defined for each set in `src/FileFormats/MOF/read.jl`\n - [ ] `head_name` is defined for each set in `src/FileFormats/MOF/write.jl`\n - [ ] A new unit test calling `_test_model_equality` is aded to\n       `test/FileFormats/MOF/MOF.jl`\n\n## Tests\n\n - [ ] The version field in `test/FileFormats/MOF/nlp.mof.json` is updated\n\n## Documentation\n\n - [ ] The version fields are updated in `docs/src/submodules/FileFormats/overview.md`","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"DocTestSetup = quote\n    using JuMP\nend","category":"page"},{"location":"manual/expressions/#Expressions","page":"Expressions","title":"Expressions","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"JuMP has three types of expressions: affine, quadratic, and nonlinear. These expressions can be inserted into constraints or into the objective. This is particularly useful if an expression is used in multiple places in the model.","category":"page"},{"location":"manual/expressions/#Affine-expressions","page":"Expressions","title":"Affine expressions","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"There are four ways of constructing an affine expression in JuMP: with the @expression macro, with operator overloading, with the AffExpr constructor, and with add_to_expression!.","category":"page"},{"location":"manual/expressions/#Macros","page":"Expressions","title":"Macros","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"The recommended way to create an affine expression is via the @expression macro.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> ex = @expression(model, 2x + y - 1)\n2 x + y - 1","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"This expression can be used in the objective or added to a constraint. For example:","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> @objective(model, Min, 2 * ex - 1)\n4 x + 2 y - 3\n\njulia> objective_function(model)\n4 x + 2 y - 3","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Just like variables and constraints, named expressions can also be created. For example","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x[i = 1:3]);\n\njulia> @expression(model, expr[i = 1:3], i * sum(x[j] for j in i:3));\n\njulia> expr\n3-element Vector{AffExpr}:\n x[1] + x[2] + x[3]\n 2 x[2] + 2 x[3]\n 3 x[3]","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"tip: Tip\nYou can read more about containers in the Containers section.","category":"page"},{"location":"manual/expressions/#Operator-overloading","page":"Expressions","title":"Operator overloading","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Expressions can also be created without macros. However, note that in some cases, this can be much slower that constructing an expression using macros.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> ex = 2x + y - 1\n2 x + y - 1","category":"page"},{"location":"manual/expressions/#Constructors","page":"Expressions","title":"Constructors","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"A third way to create an affine expression is by the AffExpr constructor. The first argument is the constant term, and the remaining arguments are variable-coefficient pairs.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> ex = AffExpr(-1.0, x => 2.0, y => 1.0)\n2 x + y - 1","category":"page"},{"location":"manual/expressions/#add_to_expression!","page":"Expressions","title":"add_to_expression!","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"The fourth way to create an affine expression is by using add_to_expression!. Compared to the operator overloading method, this approach is faster because it avoids constructing temporary objects. The @expression macro uses add_to_expression! behind-the-scenes.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> ex = AffExpr(-1.0)\n-1\n\njulia> add_to_expression!(ex, 2.0, x)\n2 x - 1\n\njulia> add_to_expression!(ex, 1.0, y)\n2 x + y - 1","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"warning: Warning\nRead the section Initializing arrays for some cases to be careful about when using add_to_expression!.","category":"page"},{"location":"manual/expressions/#Removing-zero-terms","page":"Expressions","title":"Removing zero terms","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Use drop_zeros! to remove terms from an affine expression with a 0 coefficient.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @expression(model, ex, x + 1 - x)\n0 x + 1\n\njulia> drop_zeros!(ex)\n\njulia> ex\n1","category":"page"},{"location":"manual/expressions/#Coefficients","page":"Expressions","title":"Coefficients","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Use coefficient to return the coefficient associated with a variable in an affine expression.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> @expression(model, ex, 2x + 1)\n2 x + 1\n\njulia> coefficient(ex, x)\n2.0\n\njulia> coefficient(ex, y)\n0.0","category":"page"},{"location":"manual/expressions/#Quadratic-expressions","page":"Expressions","title":"Quadratic expressions","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Like affine expressions, there are four ways of constructing a quadratic expression in JuMP: macros, operator overloading, constructors, and add_to_expression!.","category":"page"},{"location":"manual/expressions/#Macros-2","page":"Expressions","title":"Macros","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"The @expression macro can be used to create quadratic expressions by including quadratic terms.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> ex = @expression(model, x^2 + 2 * x * y + y^2 + x + y - 1)\nx² + 2 x*y + y² + x + y - 1","category":"page"},{"location":"manual/expressions/#Operator-overloading-2","page":"Expressions","title":"Operator overloading","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Operator overloading can also be used to create quadratic expressions. The same performance warning (discussed in the affine expression section) applies.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> ex = x^2 + 2 * x * y + y^2 + x + y - 1\nx² + 2 x*y + y² + x + y - 1","category":"page"},{"location":"manual/expressions/#Constructors-2","page":"Expressions","title":"Constructors","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Quadratic expressions can also be created using the QuadExpr constructor. The first argument is an affine expression, and the remaining arguments are pairs, where the first term is a JuMP.UnorderedPair and the second term is the coefficient.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> aff_expr = AffExpr(-1.0, x => 1.0, y => 1.0)\nx + y - 1\n\njulia> quad_expr = QuadExpr(\n           aff_expr,\n           UnorderedPair(x, x) => 1.0,\n           UnorderedPair(x, y) => 2.0,\n           UnorderedPair(y, y) => 1.0,\n       )\nx² + 2 x*y + y² + x + y - 1","category":"page"},{"location":"manual/expressions/#add_to_expression!-2","page":"Expressions","title":"add_to_expression!","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Finally, add_to_expression! can also be used to add quadratic terms.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> ex = QuadExpr(x + y - 1.0)\nx + y - 1\n\njulia> add_to_expression!(ex, 1.0, x, x)\nx² + x + y - 1\n\njulia> add_to_expression!(ex, 2.0, x, y)\nx² + 2 x*y + x + y - 1\n\njulia> add_to_expression!(ex, 1.0, y, y)\nx² + 2 x*y + y² + x + y - 1","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"warning: Warning\nRead the section Initializing arrays for some cases to be careful about when using add_to_expression!.","category":"page"},{"location":"manual/expressions/#Removing-zero-terms-2","page":"Expressions","title":"Removing zero terms","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Use drop_zeros! to remove terms from a quadratic expression with a 0 coefficient.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @expression(model, ex, x^2 + x + 1 - x^2)\n0 x² + x + 1\n\njulia> drop_zeros!(ex)\n\njulia> ex\nx + 1","category":"page"},{"location":"manual/expressions/#Coefficients-2","page":"Expressions","title":"Coefficients","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Use coefficient to return the coefficient associated with a pair of variables in a quadratic expression.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> @expression(model, ex, 2*x*y + 3*x)\n2 x*y + 3 x\n\njulia> coefficient(ex, x, y)\n2.0\n\njulia> coefficient(ex, x, x)\n0.0\n\njulia> coefficient(ex, y, x)\n2.0\n\njulia> coefficient(ex, x)\n3.0","category":"page"},{"location":"manual/expressions/#Nonlinear-expressions","page":"Expressions","title":"Nonlinear expressions","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Nonlinear expressions in JuMP are represented by a NonlinearExpr object. See Nonlinear expressions in detail for more details.","category":"page"},{"location":"manual/expressions/#Initializing-arrays","page":"Expressions","title":"Initializing arrays","text":"","category":"section"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"JuMP implements zero(AffExpr) and one(AffExpr) to support various functions in LinearAlgebra (for example, accessing the off-diagonal of a Diagonal matrix).","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> zero(AffExpr)\n0\n\njulia> one(AffExpr)\n1","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"However, this can result in a subtle bug if you call add_to_expression! or the MutableArithmetics API on an element created by zeros or ones:","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> x = zeros(AffExpr, 2)\n2-element Vector{AffExpr}:\n 0\n 0\n\njulia> add_to_expression!(x[1], 1.1)\n1.1\n\njulia> x\n2-element Vector{AffExpr}:\n 1.1\n 1.1","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Notice how we modified x[1], but we also changed x[2]!","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"This happened because zeros(AffExpr, 2) calls zero(AffExpr) once to obtain a zero element, and then creates an appropriately sized array filled with the same element.","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"This also happens with broadcasting calls containing a conversion of 0 or 1:","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> x = Vector{AffExpr}(undef, 2)\n2-element Vector{AffExpr}:\n #undef\n #undef\n\njulia> x .= 0\n2-element Vector{AffExpr}:\n 0\n 0\n\njulia> add_to_expression!(x[1], 1.1)\n1.1\n\njulia> x\n2-element Vector{AffExpr}:\n 1.1\n 1.1","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"The recommended way to create an array of empty expressions is as follows:","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> x = Vector{AffExpr}(undef, 2)\n2-element Vector{AffExpr}:\n #undef\n #undef\n\njulia> for i in eachindex(x)\n           x[i] = AffExpr(0.0)\n       end\n\njulia> add_to_expression!(x[1], 1.1)\n1.1\n\njulia> x\n2-element Vector{AffExpr}:\n 1.1\n 0","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Alternatively, use non-mutating operation to avoid updating x[1] in-place:","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"julia> x = zeros(AffExpr, 2)\n2-element Vector{AffExpr}:\n 0\n 0\n\njulia> x[1] += 1.1\n1.1\n\njulia> x\n2-element Vector{AffExpr}:\n 1.1\n 0","category":"page"},{"location":"manual/expressions/","page":"Expressions","title":"Expressions","text":"Note that for large expressions this will be slower due to the allocation of additional temporary objects.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"EditURL = \"factory_schedule.jl\"","category":"page"},{"location":"tutorials/linear/factory_schedule/#The-factory-schedule-example","page":"The factory schedule example","title":"The factory schedule example","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"This tutorial was originally contributed by @Crghilardi.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"This tutorial is a Julia translation of Part 5 from Introduction to Linear Programming with Python.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"The purpose of this tutorial is to demonstrate how to use DataFrames and delimited files, and to structure your code that is robust to infeasibilities and permits running with different datasets.","category":"page"},{"location":"tutorials/linear/factory_schedule/#Required-packages","page":"The factory schedule example","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"using JuMP\nimport CSV\nimport DataFrames\nimport HiGHS\nimport StatsPlots","category":"page"},{"location":"tutorials/linear/factory_schedule/#Formulation","page":"The factory schedule example","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"The Factory Scheduling Problem assumes we are optimizing the production of a good from factories f in F over the course of 12 months m in M.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"If a factory f runs during a month m, a fixed cost of a_f is incurred, the factory must produce x_mf units that is within some minimum and maximum production levels l_f and u_f respectively, and each unit of production incurs a variable cost c_f. Otherwise, the factory can be shut for the month with zero production and no fixed-cost is incurred. We denote the run/not-run decision by z_mf in 0 1, where z_mf is 1 if factory f runs in month m. The factory must produce enough units to satisfy demand d_m.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"With a little effort, we can formulate our problem as the following linear program:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"beginaligned\nmin  sumlimits_f in F m in M a_f z_mf + c_f x_mf \ntextst  x_mf le u_f z_mf  forall f in F m in M \n             x_mf ge l_f z_mf  forall f in F m in M \n             sumlimits_fin F x_mf = d_m  forall f in F m in M \n             z_mf in 0 1  forall f in F m in M\nendaligned","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"However, this formulation has a problem: if demand is too high, we may be unable to satisfy the demand constraint, and the problem will be infeasible.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"tip: Tip\nWhen modeling, consider ways to formulate your model such that it always has a feasible solution. This greatly simplifies debugging data errors that would otherwise result in an infeasible solution. In practice, most practical decisions have a feasible solution. In our case, we could satisfy demand (at a high cost) by buying replacement items for the buyer, or running the factories in overtime to make up the difference.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"We can improve our model by adding a new variable, delta_m, which represents the quantity of unmet demand in each month m. We penalize delta_m by an arbitrarily large value of $10,000/unit in the objective.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"beginaligned\nmin  sumlimits_f in F m in M a_f z_mf + c_f x_mf + sumlimits_m in M10000 delta_m \ntextst  x_mf le u_f z_mf  forall f in F m in M \n             x_mf ge l_f z_mf  forall f in F m in M \n             sumlimits_fin F x_mf - delta_m = d_m  forall f in F m in M \n             z_mf in 0 1  forall f in F m in M \n             delta_m ge 0  forall m in M\nendaligned","category":"page"},{"location":"tutorials/linear/factory_schedule/#Data","page":"The factory schedule example","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"The JuMP GitHub repository contains two text files with the data we need for this tutorial.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"The first file contains a dataset of our factories, A and B, with their production and cost levels for each month. For the documentation, the file is located at:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"factories_filename = joinpath(@__DIR__, \"factory_schedule_factories.txt\");\nnothing #hide","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"To run locally, download factory_schedule_factories.txt and update factories_filename appropriately.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"The file has the following contents:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"print(read(factories_filename, String))","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"We use the CSV and DataFrames packages to read it into Julia:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"factory_df = CSV.read(\n    factories_filename,\n    DataFrames.DataFrame;\n    delim = ' ',\n    ignorerepeated = true,\n)","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"The second file contains the demand data by month:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"demand_filename = joinpath(@__DIR__, \"factory_schedule_demand.txt\");\nnothing #hide","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"To run locally, download factory_schedule_demand.txt and update demand_filename appropriately.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"demand_df = CSV.read(\n    demand_filename,\n    DataFrames.DataFrame;\n    delim = ' ',\n    ignorerepeated = true,\n)","category":"page"},{"location":"tutorials/linear/factory_schedule/#Data-validation","page":"The factory schedule example","title":"Data validation","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Before moving on, it's always good practice to validate the data you read from external sources. The more effort you spend here, the fewer issues you will have later. The following function contains a few simple checks, but we could add more. For example, you might want to check that none of the values are too large (or too small), which might indicate a typo or a unit conversion issue (perhaps the variable costs are in $/1000 units instead of $/unit).","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"function valiate_data(\n    demand_df::DataFrames.DataFrame,\n    factory_df::DataFrames.DataFrame,\n)\n    # Minimum production must not exceed maximum production.\n    @assert all(factory_df.min_production .<= factory_df.max_production)\n    # Demand, minimum production, fixed costs, and variable costs must all be\n    # non-negative.\n    @assert all(demand_df.demand .>= 0)\n    @assert all(factory_df.min_production .>= 0)\n    @assert all(factory_df.fixed_cost .>= 0)\n    @assert all(factory_df.variable_cost .>= 0)\n    return\nend\n\nvaliate_data(demand_df, factory_df)","category":"page"},{"location":"tutorials/linear/factory_schedule/#JuMP-formulation","page":"The factory schedule example","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Next, we need to code our JuMP formulation. As shown in Design patterns for larger models, it's always good practice to code your model in a function that accepts well-defined input and returns well-defined output.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"function solve_factory_scheduling(\n    demand_df::DataFrames.DataFrame,\n    factory_df::DataFrames.DataFrame,\n)\n    # Even though we validated the data above, it's good practice to do it here\n    # too.\n    valiate_data(demand_df, factory_df)\n    months, factories = unique(factory_df.month), unique(factory_df.factory)\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    @variable(model, status[months, factories], Bin)\n    @variable(model, production[months, factories], Int)\n    @variable(model, unmet_demand[months] >= 0)\n    # We use `eachrow` to loop through the rows of the dataframe and add the\n    # relevant constraints.\n    for r in eachrow(factory_df)\n        m, f = r.month, r.factory\n        @constraint(model, production[m, f] <= r.max_production * status[m, f])\n        @constraint(model, production[m, f] >= r.min_production * status[m, f])\n    end\n    @constraint(\n        model,\n        [r in eachrow(demand_df)],\n        sum(production[r.month, :]) + unmet_demand[r.month] == r.demand,\n    )\n    @objective(\n        model,\n        Min,\n        10_000 * sum(unmet_demand) + sum(\n            r.fixed_cost * status[r.month, r.factory] +\n            r.variable_cost * production[r.month, r.factory] for\n            r in eachrow(factory_df)\n        )\n    )\n    optimize!(model)\n    schedules = Dict{Symbol,Vector{Float64}}(\n        Symbol(f) => value.(production[:, f]) for f in factories\n    )\n    schedules[:unmet_demand] = value.(unmet_demand)\n    return (\n        termination_status = termination_status(model),\n        cost = objective_value(model),\n        # This `select` statement re-orders the columns in the DataFrame.\n        schedules = DataFrames.select(\n            DataFrames.DataFrame(schedules),\n            [:unmet_demand, :A, :B],\n        ),\n    )\nend","category":"page"},{"location":"tutorials/linear/factory_schedule/#Solution","page":"The factory schedule example","title":"Solution","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Now we can call our solve_factory_scheduling function using the data we read in above.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"solution = solve_factory_scheduling(demand_df, factory_df);\nnothing #hide","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Let's see what solution contains:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"solution.termination_status","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"solution.cost","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"solution.schedules","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"These schedules will be easier to visualize as a graph:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"StatsPlots.groupedbar(\n    Matrix(solution.schedules);\n    bar_position = :stack,\n    labels = [\"unmet demand\" \"A\" \"B\"],\n    xlabel = \"Month\",\n    ylabel = \"Production\",\n    legend = :topleft,\n    color = [\"#20326c\" \"#4063d8\" \"#a0b1ec\"],\n)","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Note that we don't have any unmet demand.","category":"page"},{"location":"tutorials/linear/factory_schedule/#What-happens-if-demand-increases?","page":"The factory schedule example","title":"What happens if demand increases?","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Let's run an experiment by increasing the demand by 50% in all time periods:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"demand_df.demand .*= 1.5","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Now we resolve the problem:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"high_demand_solution = solve_factory_scheduling(demand_df, factory_df);\nnothing #hide","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"and visualize the solution:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"StatsPlots.groupedbar(\n    Matrix(high_demand_solution.schedules);\n    bar_position = :stack,\n    labels = [\"unmet demand\" \"A\" \"B\"],\n    xlabel = \"Month\",\n    ylabel = \"Production\",\n    legend = :topleft,\n    color = [\"#20326c\" \"#4063d8\" \"#a0b1ec\"],\n)","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Uh oh, we can't satisfy all of the demand.","category":"page"},{"location":"tutorials/linear/factory_schedule/#How-sensitive-is-the-solution-to-changes-in-variable-cost?","page":"The factory schedule example","title":"How sensitive is the solution to changes in variable cost?","text":"","category":"section"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Let's run another experiment, this time seeing how the optimal objective value changes as we vary the variable costs of each factory.","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"First though, let's reset the demand to it's original level:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"demand_df.demand ./= 1.5;\nnothing #hide","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"For our experiment, we're going to scale the variable costs of both factories by a set of values from 0.0 to 1.5:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"scale_factors = 0:0.1:1.5","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"At a high level, we're going to loop over the scale factors for A, then the scale factors for B, rescale the input data, call our solve_factory_scheduling example, and then store the optimal objective value in the following cost matrix:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"cost = zeros(length(scale_factors), length(scale_factors));\nnothing #hide","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Because we're modifying factory_df in-place, we need to store the original variable costs in a new column:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"factory_df[!, :old_variable_cost] = copy(factory_df.variable_cost);\nnothing #hide","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Then, we need a function to scale the :variable_cost column for a particular factory by a value scale:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"function scale_variable_cost(df, factory, scale)\n    rows = df.factory .== factory\n    df[rows, :variable_cost] .=\n        round.(Int, df[rows, :old_variable_cost] .* scale)\n    return\nend","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Our experiment is just a nested for-loop, modifying A and B and storing the cost:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"for (j, a) in enumerate(scale_factors)\n    scale_variable_cost(factory_df, \"A\", a)\n    for (i, b) in enumerate(scale_factors)\n        scale_variable_cost(factory_df, \"B\", b)\n        cost[i, j] = solve_factory_scheduling(demand_df, factory_df).cost\n    end\nend","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"Let's visualize the cost matrix:","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"StatsPlots.contour(\n    scale_factors,\n    scale_factors,\n    cost;\n    xlabel = \"Scale of factory A\",\n    ylabel = \"Scale of factory B\",\n)","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"What can you infer from the solution?","category":"page"},{"location":"tutorials/linear/factory_schedule/","page":"The factory schedule example","title":"The factory schedule example","text":"info: Info\nThe Power Systems tutorial explains a number of other ways you can structure a problem to perform a parametric analysis of the solution. In particular, you can use in-place modification to reduce the time it takes to build and solve the resulting models.","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"EditURL = \"changelog.md\"","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"CurrentModule = JuMP","category":"page"},{"location":"release_notes/#Release-notes","page":"Release notes","title":"Release notes","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.","category":"page"},{"location":"release_notes/#[Version-1.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.1)-(September-23,-2023)","page":"Release notes","title":"Version 1.15.1 (September 23, 2023)","text":"","category":"section"},{"location":"release_notes/#Fixed","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed support for single argument min and max operators (#3522)\nFixed error message for add_to_expression! when called with a GenericNonlinearExpr (#3506)\nFixed constraint tags with broadcasted constraints (#3515)\nFixed MethodError in MA.scaling (#3518)\nFixed support for arrays of Parameter variables (#3524)","category":"page"},{"location":"release_notes/#Other","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Updated to Documenter@1 (#3501)\nFixed links to data in tutorials (#3512)\nFixed typo in TSP tutorial (#3516)\nImproved error message for VariableNotOwned errors (#3520)\nFixed various JET errors (#3519)","category":"page"},{"location":"release_notes/#[Version-1.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.15.0)-(September-15,-2023)","page":"Release notes","title":"Version 1.15.0 (September 15, 2023)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"This is a large minor release because it adds an entirely new data structure and API path for working with nonlinear programs. The previous nonlinear interface remains unchanged and is documented at Nonlinear Modeling (Legacy). The new interface is a treated as a non-breaking feature addition and is documented at Nonlinear Modeling.","category":"page"},{"location":"release_notes/#Breaking","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Although the new nonlinear interface is a feature addition, there are two changes which might be breaking for a very small number of users.","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The syntax inside JuMP macros is parsed using a different code path, even for linear and quadratic expressions. We made this change to unify how we parse linear, quadratic, and nonlinear expressions. In all cases, the new code returns equivalent expressions, but because of the different order of operations, there are three changes to be aware of when updating:\nThe printed form of the expression may change, for example from x * y to y * x. This can cause tests which test the String representation of a model to fail.\nSome coefficients may change slightly due to floating point round-off error.\nParticularly when working with a JuMP extension, you may encounter a MethodError due to a missing or ambiguous method. These errors are due to previously existing bugs that were not triggered by the previous parsing code. If you encounter such an error, please open a GitHub issue.\nThe methods for Base.:^(x::VariableRef, n::Integer) and Base.:^(x::AffExpr, n::Integer) have changed. Previously, these methods supported only n = 0, 1, 2 and they always returned a QuadExpr, even for the case when n = 0 or n = 1. Now:\nx^0 returns one(T), where T is the value_type of the model (defaults to Float64)\nx^1 returns x\nx^2 returns a QuadExpr\nx^n where !(0 <= n <= 2) returns a NonlinearExpr.\nWe made this change to support nonlinear expressions and to align the mathematical definition of the operation with their return type. (Previously, users were surprised that x^1 returned a QuadExpr.) As a consequence of this change, the methods are now not type-stable. This means that the compiler cannot prove that x^2 returns a QuadExpr. If benchmarking shows that this is a performance problem, you can use the type-stable x * x instead of x^2.","category":"page"},{"location":"release_notes/#Added","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added triangle_vec which simplifies adding MOI.LogDetConeTriangle and MOI.RootDetConeTriangle constraints (#3456)\nAdded the new nonlinear interface. This is a very large change. See the documentation at Nonlinear Modeling and the (long) discussion in JuMP.jl#3106. Related PRs are (#3468) (#3472) (#3475) (#3483) (#3487) (#3488) (#3489) (#3504) (#3509)","category":"page"},{"location":"release_notes/#Fixed-2","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed uses of @nospecialize which cause precompilation failures in Julia v1.6.0 and v1.6.1. (#3464)\nFixed adding a container of Parameter (#3473)\nFixed return type of x^0 and x^1 to no longer return QuadExpr (see note in Breaking section above) (#3474)\nFixed error messages in LowerBoundRef, UpperBoundRef, FixRef, IntegerRef, BinaryRef, ParameterRef and related functions (#3494)\nFixed type inference of empty containers in JuMP macros (#3500)","category":"page"},{"location":"release_notes/#Other-2","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added GAMS to solver documentation (#3357)\nUpdated various tutorials (#3459) (#3460) (#3462) (#3463) (#3465) (#3490) (#3492) (#3503)\nAdded The network multi-commodity flow problem tutorial (#3491)\nAdded Two-stage stochastic programs tutorial (#3466)\nAdded better error messages for unsupported operations in LinearAlgebra (#3476)\nUpdated to the latest version of Documenter (#3484) (#3495) (#3497)\nUpdated GitHub action versions (#3507)","category":"page"},{"location":"release_notes/#[Version-1.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.1)-(September-2,-2023)","page":"Release notes","title":"Version 1.14.1 (September 2, 2023)","text":"","category":"section"},{"location":"release_notes/#Fixed-3","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix links in Documentation (#3478)","category":"page"},{"location":"release_notes/#[Version-1.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.14.0)-(August-27,-2023)","page":"Release notes","title":"Version 1.14.0 (August 27, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-2","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added DimensionalData.jl extension (#3413)\nAdded syntactic sugar for the MOI.Parameter set (#3443)\nParameter\nParameterRef\nis_parameter\nparameter_value\nset_parameter_value","category":"page"},{"location":"release_notes/#Fixed-4","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed model_convert for BridgeableConstraint (#3437)\nFixed printing models with integer coefficients larger than typemax(Int) (#3447)\nFixed support for constant left-hand side functions in a complementarity constraint (#3452)","category":"page"},{"location":"release_notes/#Other-3","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Updated packages used in documentation (#3444) (#3455)\nFixed docstring tests (#3445)\nFixed printing change for MathOptInterface (#3446)\nFixed typos in documentation (#3448) (#3457)\nAdded SCIP to callback documentation (#3449)","category":"page"},{"location":"release_notes/#[Version-1.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.13.0)-(July-27,-2023)","page":"Release notes","title":"Version 1.13.0 (July 27, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-3","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added support for generic number types (#3377) (#3385)\nAdded fallback for MOI.AbstractSymmetricMatrixSetTriangle and MOI.AbstractSymmetricMatrixSetSquare (#3424)","category":"page"},{"location":"release_notes/#Fixed-5","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed set_start_values with MOI.Bridges.Objective.SlackBridge (#3422)\nFixed flakey doctest in variables.md (#3425)\nFixed names on CITATION.bib (#3423)","category":"page"},{"location":"release_notes/#Other-4","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added Loraine.jl to the installation table (#3426)\nRemoved Penopt.jl from packages.toml (#3428)\nImproved problem statement in cannery example of tutorial (#3430)\nMinor cleanups in Containers.DenseAxisArray implementation (#3429)\nChanged nested_problems.jl: outer/inner to upper/lower (#3433)\nRemoved second SDP relaxation in OPF tutorial (#3432)","category":"page"},{"location":"release_notes/#[Version-1.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.12.0)-(June-19,-2023)","page":"Release notes","title":"Version 1.12.0 (June 19, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-4","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added coefficient_type keyword argument to add_bridge and remove_bridge (#3394)","category":"page"},{"location":"release_notes/#Fixed-6","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed error message for matrix in HermitianPSDCone (#3369)\nFixed EditURL for custom documentation pages (#3373)\nFixed return type annotations for MOI.ConstraintPrimal and MOI.ConstraintDual (#3381)\nFixed printing change in Julia nightly (#3391)\nFixed printing of Complex coefficients (#3397)\nFixed printing of constraints in text/latex mode (#3405)\nFixed performance issue in Containers.rowtable (#3410)\nFixed bug when variables added to set of wrong dimension (#3411)","category":"page"},{"location":"release_notes/#Other-5","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added more solver READMEs to the documentation (#3358) (#3360) (#3364) (#3365) (#3366) (#3368) (#3372) (#3374) (#3376) (#3379) (#3387) (#3389)\nAdded StatusSwitchingQP.jl to the installation table (#3354)\nUpdated checklist for adding a new solver (#3370)\nUpdated extension-tests.yml action (#3371) (#3375)\nColor logs in GitHub actions (#3392)\nAdded new tutorials\nOptimal power flow (#3395) (#3412)\nLovász numbers (#3399)\nDualization (#3402)\nUpdated JuMP paper citation (#3400)\nChanged GitHub action to upload LaTeX logs when building documentation (#3403)\nFixed printing of SCS log in documentation (#3406)\nUpdated solver versions (#3407)\nUpdated documentation to use Julia v1.9 (#3398)\nReplaced _value_type with MOI.Utilities.value_type (#3414)\nFixed a typo in docstring (#3415)\nRefactored API documentation (#3386)\nUpdated SCIP license (#3420)","category":"page"},{"location":"release_notes/#[Version-1.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.1)-(May-19,-2023)","page":"Release notes","title":"Version 1.11.1 (May 19, 2023)","text":"","category":"section"},{"location":"release_notes/#Fixed-7","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a poor error message when sum(::DenseAxisArray; dims) was called (#3338)\nFixed support for dependent sets in the @variable macro (#3344)\nFixed a performance bug in constraints with sparse symmetric matrices (#3349)","category":"page"},{"location":"release_notes/#Other-6","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Improved the printing of complex numbers (#3332)\nWhen printing, sets which contain constants ending in .0 now print as integers. This follows the behavior of constants in functions (#3341)\nAdded InfiniteOpt to the extensions documentation (#3343)\nAdded more documentation for the exponential cone (#3345) (#3347)\nAdded checklists for developers (#3346) (#3355)\nFixed test support upcoming Julia nightly (#3351)\nFixed extension-tests.yml action (#3353)\nAdd more solvers to the documentation (#3359) (#3361) (#3362)","category":"page"},{"location":"release_notes/#[Version-1.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.11.0)-(May-3,-2023)","page":"Release notes","title":"Version 1.11.0 (May 3, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-5","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added new methods to print_active_bridges for printing a particular objective, constraint, or variable (#3316)","category":"page"},{"location":"release_notes/#Fixed-8","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed tests for MOI v1.14.0 release (#3312)\nFixed indexing containers when an axis is Vector{Any} that contains a Vector{Any} element (#3280)\nFixed getindex(::AbstractJuMPScalar) which is called for an expression like x[] (#3314)\nFixed bug in set_string_names_on_creation with a vector of variables (#3322)\nFixed bug in memoize function in nonlinear documentation (#3337)","category":"page"},{"location":"release_notes/#Other-7","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed typos in the documentation (#3317) (#3318) (#3328)\nAdded a test for the order of setting start values (#3315)\nAdded READMEs of solvers and extensions to the docs (#3309) (#3320) (#3327) (#3329) (#3333)\nStyle improvements to src/variables.jl (#3324)\nClarify that column generation does not find global optimum (#3325)\nAdd a GitHub actions workflow for testing extensions prior to release (#3331)\nDocument the release process for JuMP (#3334)\nFix links to discourse and chatroom (#3335)","category":"page"},{"location":"release_notes/#[Version-1.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.10.0)-(April-3,-2023)","page":"Release notes","title":"Version 1.10.0 (April 3, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-6","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added Nonnegatives, Nonpositives and Zeros, and support vector-valued inequality syntax in the JuMP macros (#3273)\nAdded special support for LinearAlgebra.Symmetric and LinearAlgebra.Hermitian matrices in Zeros constraints (#3281) (#3296)\nAdded HermitianMatrixSpace and the Hermitian tag for generating a matrix of variables that is Hermitian (#3292) (#3293)\nAdded Semicontinuous and Semiinteger (#3302)\nAdded support for keyword indexing of containers (#3237)","category":"page"},{"location":"release_notes/#Fixed-9","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed [compat] bound for MathOptInterface in Project.toml (#3272)","category":"page"},{"location":"release_notes/#Other-8","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Split out the Nested optimization problems tutorial (#3274)\nUpdated doctests to ensure none have hidden state (#3275) (#3276)\nClarified how lazy constraints may revisit points (#3278)\nAdded P-Norm example (#3282)\nClarified docs that macros create new bindings (#3284)\nFixed threading example (#3283)\nAdded plot to The minimum distortion problem (#3288)\nAdded Google style rules for Vale and fixed warnings (#3285)\nAdded citation for the JuMP 1.0 paper (#3294)\nUpdated package versions in the documentation (#3298)\nAdded comment for the order in which start values must be set (#3303)\nImproved error message for unrecognized constraint operators (#3311)","category":"page"},{"location":"release_notes/#[Version-1.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.9.0)-(March-7,-2023)","page":"Release notes","title":"Version 1.9.0 (March 7, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-7","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added get_attribute and set_attribute. These replace get_optimizer_attribute and set_optimizer_attribute, although the _optimizer_ functions remain for backward compatibility. (#3219)\nAdded set_start_values for setting all supported start values in a model (#3238)\nAdd remove_bridge and print_active_bridges (#3259)","category":"page"},{"location":"release_notes/#Fixed-10","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The matrix returned by a variable in HermitianPSDCone is now a LinearAlgebra.Hermitian matrix. This is potentially breaking if you have written code to assume the return is a Matrix. (#3245) (#3246)\nFixed missing support for Base.isreal of expressions (#3252)","category":"page"},{"location":"release_notes/#Other-9","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a thread safety issue in the Parallelism tutorial (#3240) (#3243)\nImproved the error message when unsupported operators are used in @NL macros (#3236)\nClarified the documentation to say that matrices in HermitianPSDCone must be LinearAlgebra.Hermitian (#3241)\nMinor style fixes to internal macro code (#3247)\nAdd Quantum state discrimination tutorial (#3250)\nImprove error message when begin...end not passed to plural macros (#3255)\nDocument how to register function with varying number of input arguments (#3258)\nTidy tests by removing unneeded JuMP. prefixes (#3260)\nClarified the introduction to the Complex number support tutorial (#3262)\nFixed typos in the Documentation (#3263) (#3266) (#3268) (#3269)","category":"page"},{"location":"release_notes/#[Version-1.8.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.2)-(February-27,-2023)","page":"Release notes","title":"Version 1.8.2 (February 27, 2023)","text":"","category":"section"},{"location":"release_notes/#Fixed-11","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed dot product between complex JuMP expression and number (#3244)","category":"page"},{"location":"release_notes/#Other-10","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Polish simple SDP examples (#3232)","category":"page"},{"location":"release_notes/#[Version-1.8.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.1)-(February-23,-2023)","page":"Release notes","title":"Version 1.8.1 (February 23, 2023)","text":"","category":"section"},{"location":"release_notes/#Fixed-12","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed support for init in nonlinear generator expressions (#3226)","category":"page"},{"location":"release_notes/#Other-11","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Use and document import MathOptInterface as MOI (#3222)\nRemoved references in documentation to multiobjective optimization being unsupported (#3223)\nAdded tutorial on multi-objective portfolio optimization (#3227)\nRefactored some of the conic tutorials (#3229)\nFixed typos in the documentation (#3230)\nAdded tutorial on parallelism (#3231)","category":"page"},{"location":"release_notes/#[Version-1.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.8.0)-(February-16,-2023)","page":"Release notes","title":"Version 1.8.0 (February 16, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-8","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added --> syntax support for indicator constraints. The old syntax of => remains supported (#3207)\nAdded <--> syntax for reified constraints. For now, few solvers support reified constraints (#3206)\nAdded fix_discrete_variables. This is most useful for computing the dual of a mixed-integer program (#3208)\nAdded support for vector-valued objectives. For details, see the Multi-objective knapsack tutorial (#3176)","category":"page"},{"location":"release_notes/#Fixed-13","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in lp_sensitivity_report by switching to an explicit LU factorization of the basis matrix (#3182)\nFixed a bug that prevented [; kwarg] arguments in macros (#3220)","category":"page"},{"location":"release_notes/#Other-12","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Minor fixes to the documentation (#3200) (#3201) (#3203) (#3210)\nAdded tutorial Constraint programming (#3202)\nAdded more examples to Tips and Tricks\nRemove _distance_to_set in favor of MOI.Utilities.distance_to_set (#3209)\nImprove The diet problem tutorial by adding the variable as a column in the dataframe (#3213)\nImprove The knapsack problem example tutorial (#3216) (#3217)\nAdded the Ellipsoid approximation tutorial (#3218)","category":"page"},{"location":"release_notes/#[Version-1.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.7.0)-(January-25,-2023)","page":"Release notes","title":"Version 1.7.0 (January 25, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-9","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added support for view of a Containers.DenseAxisArray (#3152) (#3180)\nAdded support for containers of variables in ComplexPlane (#3184)\nAdded support for minimum and maximum generators in nonlinear expressions (#3189)\nAdded SnoopPrecompile statements that reduce the time-to-first-solve in Julia 1.9 (#3193) (#3195) (#3196) (#3197)","category":"page"},{"location":"release_notes/#Other-13","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Large refactoring of the tests (#3166) (#3167) (#3168) (#3169) (#3170) (#3171)\nRemove unreachable code due to VERSION checks (#3172)\nDocument how to test JuMP extensions (#3174)\nFix method ambiguities in Containers (#3173)\nImprove error message that is thrown when = is used instead of == in the @constraint macro (#3178)\nImprove the error message when Bool is used instead of Bin in the @variable macro (#3180)\nUpdate versions of the documentation (#3185)\nTidy the import of packages and remove unnecessary prefixes (#3186) (#3187)\nRefactor src/JuMP.jl by moving methods into more relevant files (#3188)\nFix docstring of Model not appearing in the documentation (#3198)","category":"page"},{"location":"release_notes/#[Version-1.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.6.0)-(January-1,-2023)","page":"Release notes","title":"Version 1.6.0 (January 1, 2023)","text":"","category":"section"},{"location":"release_notes/#Added-10","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added a result keyword argument to solution_summary to allow summarizing models with multiple solutions (#3138)\nAdded relax_with_penalty!, which is a useful tool when debugging infeasible models (#3140)\nAdded has_start_value (#3157)\nAdded support for HermitianPSDCone in constraints (#3154)","category":"page"},{"location":"release_notes/#Fixed-14","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed promotion of complex expressions (#3150) (#3164)","category":"page"},{"location":"release_notes/#Other-14","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added Benders tutorial with in-place resolves (#3145)\nAdded more Tips and tricks for linear programs (#3144) (#3163)\nClarified documentation that start can depend on the indices of a variable container (#3148)\nReplace instances of length and size by the recommended eachindex and axes (#3149)\nAdded a warning explaining why the model is dirty when accessing solution results from a modified model (#3156)\nClarify documentation that PSD ensures a symmetric matrix (#3159)\nMaintenance of the JuMP test suite (#3146) (#3158) (#3162)","category":"page"},{"location":"release_notes/#[Version-1.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.5.0)-(December-8,-2022)","page":"Release notes","title":"Version 1.5.0 (December 8, 2022)","text":"","category":"section"},{"location":"release_notes/#Added-11","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Add support for complex-valued variables:\nHermitianPSDCone (#3109)\nComplexPlane and ComplexVariable (#3134)\nAdd support for MOI.OptimizerWithAttributes in set_optimizer_attribute and get_optimizer_attribute (#3129)","category":"page"},{"location":"release_notes/#Fixed-15","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed error message for vectorized interval constraints (#3123)\nFixed passing AbstractString to set_optimizer_attribute (#3127)","category":"page"},{"location":"release_notes/#Other-15","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Update package versions used in docs (#3119) (#3133) (#3139)\nFixed output of diet tutorial (#3120)\nExplain how to use Dates.period in set_time_limit_sec (#3121)\nUpdate to JuliaFormatter v1.0.15 (#3130)\nFixed HTTP server example in web_app.jl (#3131)\nUpdate docs to build with Documenter#master (#3094)\nAdd tests for LinearAlgebra operations (#3132)\nTidy these release notes (#3135)\nAdded documentation for Complex number support (#3141)\nRemoved the \"workforce scheduling\" and \"steelT3\" tutorials (#3143)","category":"page"},{"location":"release_notes/#[Version-1.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.4.0)-(October-29,-2022)","page":"Release notes","title":"Version 1.4.0 (October 29, 2022)","text":"","category":"section"},{"location":"release_notes/#Added-12","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added Containers.rowtable which converts a container into a vector of NamedTuples to support the Tables.jl interface. This simplifies converting Containers.DenseAxisArray and Containers.SparseAxisArray objects into tabular forms such as a DataFrame (#3104)\nAdded a new method to Containers.container so that index names are passed to the container (#3088)","category":"page"},{"location":"release_notes/#Fixed-16","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in copy_to(dest::Model, src::MOI.ModelLike) when src has nonlinear components (#3101)\nFixed the printing of (-1.0 + 0.0im) coefficients in complex expressions (#3112)\nFixed a parsing bug in nonlinear expressions with generator statements that contain multiple for statements (#3116)","category":"page"},{"location":"release_notes/#Other-16","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Converted the multi-commodity flow tutorial to use an SQLite database (#3098)\nFixed a number of typos in the documentation (#3103) (#3107) (#3018)\nImproved various style aspects of the PDF documentation (#3095) (#3098) (#3102)","category":"page"},{"location":"release_notes/#[Version-1.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.1)-(September-28,-2022)","page":"Release notes","title":"Version 1.3.1 (September 28, 2022)","text":"","category":"section"},{"location":"release_notes/#Fixed-17","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a performance issue in relax_integrality (#3087)\nFixed the type stability of operators with Complex arguments (#3072)\nFixed a bug which added additional +() terms to some nonlinear expressions (#3091)\nFixed potential method ambiguities with AffExpr and QuadExpr objects (#3092)","category":"page"},{"location":"release_notes/#Other-17","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added vale as a linter for the documentation (#3080)\nAdded a tutorial on debugging JuMP models (#3043)\nFixed a number of typos in the documentation (#3079) (#3083)\nMany other small tweaks to the documentation (#3068) (#3073) (#3074) (#3075) (#3076) (#3077) (#3078) (#3081) (#3082) (#3084) (#3085) (#3089)","category":"page"},{"location":"release_notes/#[Version-1.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.3.0)-(September-5,-2022)","page":"Release notes","title":"Version 1.3.0 (September 5, 2022)","text":"","category":"section"},{"location":"release_notes/#Added-13","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Support slicing in SparseAxisArray (#3031)","category":"page"},{"location":"release_notes/#Fixed-18","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug introduced in v1.2.0 that prevented DenseAxisArrays with Vector keys (#3064)","category":"page"},{"location":"release_notes/#Other-18","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Released the JuMP logos under the CC BY 4.0 license (#3063)\nMinor tweaks to the documentation (#3054) (#3056) (#3057) (#3060) (#3061) (#3065)\nImproved code coverage of a number of files (#3048) (#3049) (#3050) (#3051) (#3052) (#3053) (#3058) (#3059)","category":"page"},{"location":"release_notes/#[Version-1.2.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.1)-(August-22,-2022)","page":"Release notes","title":"Version 1.2.1 (August 22, 2022)","text":"","category":"section"},{"location":"release_notes/#Fixed-19","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug when parsing two-sided nonlinear constraints (#3045)","category":"page"},{"location":"release_notes/#[Version-1.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.2.0)-(August-16,-2022)","page":"Release notes","title":"Version 1.2.0 (August 16, 2022)","text":"","category":"section"},{"location":"release_notes/#Breaking-2","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"This is a large minor release because it significantly refactors the internal code for handling nonlinear programs to use the MathOptInterface.Nonlinear submodule that was introduced in MathOptInterface v1.3.0. As a consequence, the internal datastructure in model.nlp_data has been removed, as has the JuMP._Derivatives submodule. Despite the changes, the public API for nonlinear programming has not changed, and any code that uses only the public API and that worked with v1.1.1 will continue to work with v1.2.0.","category":"page"},{"location":"release_notes/#Added-14","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added all_constraints(model; include_variable_in_set_constraints) which simplifies returning a list of all constraint indices in the model.\nAdded the ability to delete nonlinear constraints via delete(::Model, ::NonlinearConstraintRef).\nAdded the ability to provide an explicit Hessian for a multivariate user-defined function.\nAdded support for querying the primal value of a nonlinear constraint via value(::NonlinearConstraintRef)","category":"page"},{"location":"release_notes/#Fixed-20","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in Containers.DenseAxisArray so that it now supports indexing with keys that hash to the same value, even if they are different types, for example, Int32 and Int64.\nFixed a bug printing the model when the solver does not support MOI.Name.","category":"page"},{"location":"release_notes/#Other-19","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added a constraint programming formulation to the Sudoku tutorial.\nAdded newly supported solvers Pajarito, Clarabel, and COPT to the installation table.\nFixed a variety of other miscellaneous issues in the documentation.","category":"page"},{"location":"release_notes/#[Version-1.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.1)-(June-14,-2022)","page":"Release notes","title":"Version 1.1.1 (June 14, 2022)","text":"","category":"section"},{"location":"release_notes/#Other-20","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed problem displaying LaTeX in the documentation\nMinor updates to the style guide\nUpdated to MOI v1.4.0 in the documentation","category":"page"},{"location":"release_notes/#[Version-1.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.1.0)-(May-25,-2022)","page":"Release notes","title":"Version 1.1.0 (May 25, 2022)","text":"","category":"section"},{"location":"release_notes/#Added-15","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added num_constraints(::Model; count_variable_in_set_constraints) to simplify the process of counting the number of constraints in a model\nAdded VariableRef(::ConstraintRef) for querying the variable associated with a bound or integrality constraint.\nAdded set_normalized_coefficients for modifying the variable coefficients of a vector-valued constraint.\nAdded set_string_names_on_creation to disable creating String names for variables and constraints. This can improve performance.","category":"page"},{"location":"release_notes/#Fixed-21","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug passing nothing to the start keyword of @variable","category":"page"},{"location":"release_notes/#Other-21","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"New tutorials:\nSensitivity analysis of a linear program\nServing web apps\nMinimal ellipse SDP tutorial refactored and improved\nDocs updated to the latest version of each package\nLots of minor fixes and improvements to the documentation","category":"page"},{"location":"release_notes/#[Version-1.0.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v1.0.0)-(March-24,-2022)","page":"Release notes","title":"Version 1.0.0 (March 24, 2022)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Read more about this release, along with an acknowledgement of all the contributors in our JuMP 1.0.0 is released blog post.","category":"page"},{"location":"release_notes/#Breaking-3","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The previously deprecated functions (v0.23.0, v0.23.1) have been removed. Deprecation was to improve consistency of function names:\nnum_nl_constraints (see num_nonlinear_constraints)\nall_nl_constraints (see all_nonlinear_constraints)\nadd_NL_expression (see add_nonlinear_expression)\nset_NL_objective  (see set_nonlinear_objective)\nadd_NL_constraint (see add_nonlinear_constraint)\nnl_expr_string (see nonlinear_expr_string)\nnl_constraint_string (see nonlinear_constraint_string)\nSymMatrixSpace (see SymmetricMatrixSpace)\nThe unintentionally exported variable JuMP.op_hint has been renamed to the unexported JuMP._OP_HINT","category":"page"},{"location":"release_notes/#Fixed-22","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug writing .nl files\nFixed a bug broadcasting SparseAxisArrays","category":"page"},{"location":"release_notes/#[Version-0.23.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.2)-(March-14,-2022)","page":"Release notes","title":"Version 0.23.2 (March 14, 2022)","text":"","category":"section"},{"location":"release_notes/#Added-16","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added relative_gap to solution_summary\nregister now throws an informative error if the function is not differentiable using ForwardDiff. In some cases, the check in register will encounter a false negative, and the informative error will be thrown at run-time. This usually happens when the function is non-differentiable in a subset of the domain.","category":"page"},{"location":"release_notes/#Fixed-23","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a scoping issue when extending the container keyword of containers","category":"page"},{"location":"release_notes/#Other-22","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Docs updated to the latest version of each package","category":"page"},{"location":"release_notes/#[Version-0.23.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.1)-(March-2,-2022)","page":"Release notes","title":"Version 0.23.1 (March 2, 2022)","text":"","category":"section"},{"location":"release_notes/#Deprecated","page":"Release notes","title":"Deprecated","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"nl_expr_string and nl_constraint_string have been renamed to nonlinear_expr_string and nonlinear_constraint_string. The old methods still exist with deprecation warnings. This change should impact very few users because to call them you must rely on private internals of the nonlinear API. Users are encouraged to use sprint(show, x) instead, where x is the nonlinear expression or constraint of interest.","category":"page"},{"location":"release_notes/#Added-17","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added support for Base.abs2(x) where x is a variable or affine expression. This is mainly useful for complex-valued constraints.","category":"page"},{"location":"release_notes/#Fixed-24","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed addition of complex and real affine expressions\nFixed arithmetic for Complex-valued quadratic expressions\nFixed variable bounds passed as Rational{Int}(Inf)\nFixed printing of the coefficient (0 + 1im)\nFixed a bug when solution_summary is called prior to optimize!","category":"page"},{"location":"release_notes/#[Version-0.23.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.23.0)-(February-25,-2022)","page":"Release notes","title":"Version 0.23.0 (February 25, 2022)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"JuMP v0.23.0 is a breaking release. It is also a release-candidate for JuMP v1.0.0. That is, if no issues are found with the v0.23.0 release, then it will be re-tagged as v1.0.0.","category":"page"},{"location":"release_notes/#Breaking-4","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Julia 1.6 is now the minimum supported version\nMathOptInterface has been updated to v1.0.0\nAll previously deprecated functionality has been removed\nPrintMode, REPLMode and IJuliaMode have been removed in favor of the MIME types MIME\"text/plain\" and MIME\"text/latex\". Replace instances of ::Type{REPLMode} with ::MIME\"text/plain\", REPLMode with MIME(\"text/plain\"), ::Type{IJuliaMode} with ::MIME\"text/latex\", and IJuliaMode with MIME(\"text/latex\").\nFunctions containing the nl_ acronym have been renamed to the more explicit nonlinear_. For example, num_nl_constraints is now num_nonlinear_constraints and set_NL_objective is now set_nonlinear_objective. Calls to the old functions throw an error explaining the new name.\nSymMatrixSpace has been renamed to SymmetricMatrixSpace","category":"page"},{"location":"release_notes/#Added-18","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added nonlinear_dual_start_value and set_nonlinear_dual_start_value\nAdded preliminary support for Complex coefficient types","category":"page"},{"location":"release_notes/#Fixed-25","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in solution_summary","category":"page"},{"location":"release_notes/#Other-23","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"MILP examples have been migrated from GLPK to HiGHS\nFixed various typos\nImproved section on setting constraint start values","category":"page"},{"location":"release_notes/#Troubleshooting-problems-when-updating","page":"Release notes","title":"Troubleshooting problems when updating","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"If you experience problems when updating, you are likely using previously deprecated functionality. (By default, Julia does not warn when you use deprecated features.)","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"To find the deprecated features you are using, start Julia with --depwarn=yes:","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"$ julia --depwarn=yes","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Then install JuMP v0.22.3:","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"julia> using Pkg\njulia> pkg\"add JuMP@0.22.3\"","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.","category":"page"},{"location":"release_notes/#[Version-0.22.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.3)-(February-10,-2022)","page":"Release notes","title":"Version 0.22.3 (February 10, 2022)","text":"","category":"section"},{"location":"release_notes/#Fixed-26","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a reproducibility issue in the TSP tutorial\nFixed a reproducibility issue in the max_cut_sdp tutorial\nFixed a bug broadcasting an empty SparseAxisArray","category":"page"},{"location":"release_notes/#Other-24","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added a warning and improved documentation for the modify-then-query case\nFixed a typo in the docstring of RotatedSecondOrderCone\nAdded Aqua.jl as a check for code health\nAdded introductions to each section of the tutorials\nImproved the column generation and Benders decomposition tutorials\nUpdated documentation to MOI v0.10.8\nUpdated JuliaFormatter to v0.22.2","category":"page"},{"location":"release_notes/#[Version-0.22.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.2)-(January-10,-2022)","page":"Release notes","title":"Version 0.22.2 (January 10, 2022)","text":"","category":"section"},{"location":"release_notes/#Added-19","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The function all_nl_constraints now returns all nonlinear constraints in a model\nstart_value and set_start_value can now be used to get and set the primal start for constraint references\nPlural macros now return a tuple containing the elements that were defined instead of nothing\nAnonymous variables are now printed as _[i] where i is the index of the variable instead of noname. Calling name(x) still returns \"\" so this is non-breaking.","category":"page"},{"location":"release_notes/#Fixed-27","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed handling of min and max in nonlinear expressions\nCartesianIndex is no longer allowed as a key for DenseAxisArrays.","category":"page"},{"location":"release_notes/#Other-25","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Improved the performance of GenericAffExpr\nAdded a tutorial on the Travelling Salesperson Problem\nAdded a tutorial on querying the Hessian of a nonlinear program\nAdded documentation on using custom solver binaries.","category":"page"},{"location":"release_notes/#[Version-0.22.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.1)-(November-29,-2021)","page":"Release notes","title":"Version 0.22.1 (November 29, 2021)","text":"","category":"section"},{"location":"release_notes/#Added-20","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Export OptimizationSense enum, with instances: MIN_SENSE, MAX_SENSE, and FEASIBILITY_SENSE\nAdd Base.isempty(::Model) to match Base.empty(::Model)","category":"page"},{"location":"release_notes/#Fixed-28","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix bug in container with tuples as indices\nFix bug in set_time_limit_sec","category":"page"},{"location":"release_notes/#Other-26","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Add tutorial \"Design patterns for larger models\"\nRemove release notes section from PDF\nGeneral edits of the documentation and error messages","category":"page"},{"location":"release_notes/#[Version-0.22.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.22.0)-(November-10,-2021)","page":"Release notes","title":"Version 0.22.0 (November 10, 2021)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"JuMP v0.22 is a breaking release","category":"page"},{"location":"release_notes/#Breaking-5","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"JuMP 0.22 contains a number of breaking changes. However, these should be invisible for the majority of users. You will mostly encounter these breaking changes if you: wrote a JuMP extension, accessed backend(model), or called @SDconstraint.","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The breaking changes are as follows:","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"MathOptInterface has been updated to v0.10.4. For users who have interacted with the MOI backend, this contains a large number of breaking changes. Read the MathOptInterface release notes for more details.\nThe bridge_constraints keyword argument to Model and set_optimizer has been renamed add_bridges to reflect that more thing were bridged than just constraints.\nThe backend(model) field now contains a concrete instance of a MOI.Utilities.CachingOptimizer instead of one with an abstractly typed optimizer field. In most cases, this will lead to improved performance. However, calling set_optimizer after backend invalidates the old backend. For example:\nmodel = Model()\nb = backend(model)\nset_optimizer(model, GLPK.Optimizer)\n@variable(model, x)\n# b is not updated with `x`! Get a new b by calling `backend` again.\nnew_b = backend(model)\nAll usages of @SDconstraint are deprecated. The new syntax is @constraint(model, X >= Y, PSDCone()).\nCreating a DenseAxisArray with a Number as an axis will now display a warning. This catches a common error in which users write @variable(model, x[length(S)]) instead of @variable(model, x[1:length(S)]).\nThe caching_mode argument to Model, for example, Model(caching_mode = MOIU.MANUAL) mode has been removed. For more control over the optimizer, use direct_model instead.\nThe previously deprecated lp_objective_perturbation_range and lp_rhs_perturbation_range functions have been removed. Use lp_sensitivity_report instead.\nThe .m fields of NonlinearExpression and NonlinearParameter have been renamed to .model.\nInfinite variable bounds are now ignored. Thus, @variable(model, x <= Inf) will show has_upper_bound(x) == false. Previously, these bounds were passed through to the solvers which caused numerical issues for solvers expecting finite bounds.\nThe variable_type and constraint_type functions were removed. This should only affect users who previously wrote JuMP extensions. The functions can be deleted without consequence.\nThe internal functions moi_mode, moi_bridge_constraints, moi_add_constraint, and moi_add_to_function_constant are no longer exported.\nThe un-used method Containers.generate_container has been deleted.\nThe Containers API has been refactored, and _build_ref_sets is now public as Containers.build_ref_sets.\nThe parse_constraint_ methods for extending @constraint at parse time have been refactored in a breaking way. Consult the Extensions documentation for more details and examples.","category":"page"},{"location":"release_notes/#Added-21","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The TerminationStatusCode and ResultStatusCode enums are now exported by JuMP. Prefer termination_status(model) == OPTIMAL instead of == MOI.OPTIMAL, although the MOI. prefix way still works.\nCopy a x::DenseAxisArray to an Array by calling Array(x).\nNonlinearExpression is now a subtype of AbstractJuMPScalar\nConstraints such as @constraint(model, x + 1 in MOI.Integer()) are now supported.\nprimal_feasibility_report now accepts a function as the first argument.\nScalar variables @variable(model, x[1:2] in MOI.Integer()) creates two variables, both of which are constrained to be in the set MOI.Integer.\nConic constraints can now be specified as inequalities under a different partial ordering. So @constraint(model, x - y in MOI.Nonnegatives()) can now be written as @constraint(model, x >= y, MOI.Nonnegatives()).\nNames are now set for vectorized constraints.","category":"page"},{"location":"release_notes/#Fixed-29","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a performance issue when show was called on a SparseAxisArray with a large number of elements.\nFixed a bug displaying barrier and simplex iterations in solution_summary.\nFixed a bug by implementing hash for DenseAxisArray and SparseAxisArray.\nNames are now only set if the solver supports them. Previously, this prevented solvers such as Ipopt from being used with direct_model.\nMutableArithmetics.Zero is converted into a 0.0 before being returned to the user. Previously, some calls to @expression would return the undocumented MutableArithmetics.Zero() object. One example is summing over an empty set @expression(model, sum(x[i] for i in 1:0)). You will now get 0.0 instead.\nAffExpr and QuadExpr can now be used with == 0 instead of iszero. This fixes a number of issues relating to Julia standard libraries such as LinearAlgebra and SparseArrays.\nFixed a bug when registering a user-defined function with splatting.","category":"page"},{"location":"release_notes/#Other-27","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The documentation is now available as a PDF.\nThe documentation now includes a full copy of the MathOptInterface documentation to make it easy to link concepts between the docs. (The MathOptInterface documentation has also been significantly improved.)\nThe documentation contains a large number of improvements and clarifications on a range of topics. Thanks to @sshin23, @DilumAluthge, and @jlwether.\nThe documentation is now built with Julia 1.6 instead of 1.0.\nVarious error messages have been improved to be more readable.","category":"page"},{"location":"release_notes/#[Version-0.21.10](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.10)-(September-4,-2021)","page":"Release notes","title":"Version 0.21.10 (September 4, 2021)","text":"","category":"section"},{"location":"release_notes/#Added-22","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added add_NL_expression\nadd_NL_xxx functions now support AffExpr and QuadExpr as terms","category":"page"},{"location":"release_notes/#Fixed-30","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in solution_summary\nFixed a bug in relax_integrality","category":"page"},{"location":"release_notes/#Other-28","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Improved error message in lp_sensitivity_report","category":"page"},{"location":"release_notes/#[Version-0.21.9](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.9)-(August-1,-2021)","page":"Release notes","title":"Version 0.21.9 (August 1, 2021)","text":"","category":"section"},{"location":"release_notes/#Added-23","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Containers now support arbitrary container types by passing the type to the container keyword and overloading Containers.container.\nis_valid now supports nonlinear constraints\nAdded unsafe_backend for querying the inner-most optimizer of a JuMP model.\nNonlinear parameters now support the plural @NLparameters macro.\nContainers (for example, DenseAxisArray) can now be used in vector-valued constraints.","category":"page"},{"location":"release_notes/#Other-29","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Various improvements to the documentation.","category":"page"},{"location":"release_notes/#[Version-0.21.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.8)-(May-8,-2021)","page":"Release notes","title":"Version 0.21.8 (May 8, 2021)","text":"","category":"section"},{"location":"release_notes/#Added-24","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The @constraint macro is now extendable in the same way as @variable.\nAffExpr and QuadExpr can now be used in nonlinear macros.","category":"page"},{"location":"release_notes/#Fixed-31","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in lp_sensitivity_report.\nFixed an inference issue when creating empty SparseAxisArrays.","category":"page"},{"location":"release_notes/#[Version-0.21.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.7)-(April-12,-2021)","page":"Release notes","title":"Version 0.21.7 (April 12, 2021)","text":"","category":"section"},{"location":"release_notes/#Added-25","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added primal_feasibility_report, which can be used to check whether a primal point satisfies primal feasibility.\nAdded coefficient, which returns the coefficient associated with a variable in affine and quadratic expressions.\nAdded copy_conflict, which returns the IIS of an infeasible model.\nAdded solution_summary, which returns (and prints) a struct containing a summary of the solution.\nAllow AbstractVector in vector constraints instead of just Vector.\nAdded latex_formulation(model) which returns an object representing the latex formulation of a model. Use print(latex_formulation(model)) to print the formulation as a string.\nUser-defined functions in nonlinear expressions are now automatically registered to aid quick model prototyping. However, a warning is printed to encourage the manual registration.\nDenseAxisArray's now support broadcasting over multiple arrays.\nContainer indices can now be iterators of Base.SizeUnknown.","category":"page"},{"location":"release_notes/#Fixed-32","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed bug in rad2deg and deg2rad in nonlinear expressions.\nFixed a MethodError bug in Containers when forcing container type.\nAllow partial slicing of a DenseAxisArray, resolving an issue from 2014.\nFixed a bug printing variable names in IJulia.\nEnding an IJulia cell with model now prints a summary of the model (like in the REPL) not the latex formulation. Use print(model) to print the latex formulation.\nFixed a bug when copying models containing nested arrays.","category":"page"},{"location":"release_notes/#Other-30","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Tutorials are now part of the documentation, and more refactoring has taken place.\nAdded JuliaFormatter added as a code formatter.\nAdded some precompilation statements to reduce initial latency.\nVarious improvements to error messages to make them more helpful.\nImproved performance of value(::NonlinearExpression).\nImproved performance of fix(::VariableRef).","category":"page"},{"location":"release_notes/#[Version-0.21.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.6)-(January-29,-2021)","page":"Release notes","title":"Version 0.21.6 (January 29, 2021)","text":"","category":"section"},{"location":"release_notes/#Added-26","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added support for skew symmetric variables via @variable(model, X[1:2, 1:2] in SkewSymmetricMatrixSpace()).\nlp_sensitivity_report has been added which significantly improves the performance of querying the sensitivity summary of an LP. lp_objective_perturbation_range and lp_rhs_perturbation_range are deprecated.\nDual warm-starts are now supported with set_dual_start_value and dual_start_value.\n∈ (\\in<tab>) can now be used in macros instead of = or in.\nUse haskey(model::Model, key::Symbol) to check if a name key is registered in a model.\nAdded unregister(model::Model, key::Symbol) to unregister a name key from model.\nAdded callback_node_status for use in callbacks.\nAdded print_bridge_graph to visualize the bridging graph generated by MathOptInterface.\nImproved error message for containers with duplicate indices.","category":"page"},{"location":"release_notes/#Fixed-33","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Various fixes to pass tests on Julia 1.6.\nFixed a bug in the printing of nonlinear expressions in IJulia.\nFixed a bug when nonlinear expressions are passed to user-defined functions.\nSome internal functions that were previously exported are now no longer exported.\nFixed a bug when relaxing a fixed binary variable.\nFixed a StackOverflowError that occurred when SparseAxisArrays had a large number of elements.\nRemoved an unnecessary type assertion in list_of_constraint_types.\nFixed a bug when copying models with registered expressions.","category":"page"},{"location":"release_notes/#Other-31","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The documentation has been significantly overhauled. It now has distinct sections for the manual, API reference, and examples. The existing examples in /examples have now been moved to /docs/src/examples and rewritten using Literate.jl, and they are now included in the documentation.\nJuliaFormatter has been applied to most of the codebase. This will continue to roll out over time, as we fix upstream issues in the formatter, and will eventually become compulsory.\nThe root cause of a large number of method invalidations has been resolved.\nWe switched continuous integration from Travis and Appveyor to GitHub Actions.","category":"page"},{"location":"release_notes/#[Version-0.21.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.5)-(September-18,-2020)","page":"Release notes","title":"Version 0.21.5 (September 18, 2020)","text":"","category":"section"},{"location":"release_notes/#Fixed-34","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed deprecation warnings\nThrow DimensionMismatch for incompatibly sized functions and sets\nUnify treatment of keys(x) on JuMP containers","category":"page"},{"location":"release_notes/#[Version-0.21.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.4)-(September-14,-2020)","page":"Release notes","title":"Version 0.21.4 (September 14, 2020)","text":"","category":"section"},{"location":"release_notes/#Added-27","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Add debug info when adding unsupported constraints\nAdd relax_integrality for solving continuous relaxation\nAllow querying constraint conflicts","category":"page"},{"location":"release_notes/#Fixed-35","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Dispatch on Real for MOI.submit\nImplement copy for CustomSet in tests\nDon't export private macros\nFix invalid assertion in nonlinear\nError if constraint has NaN right-hand side\nImprove speed of tests\nLots of work modularizing files in /test\nImprove line numbers in macro error messages\nPrint nonlinear subexpressions\nVarious documentation updates\nDependency updates:\nDatastructures 0.18\nMathOptFormat v0.5\nPrep for MathOptInterface 0.9.15","category":"page"},{"location":"release_notes/#[Version-0.21.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.3)-(June-18,-2020)","page":"Release notes","title":"Version 0.21.3 (June 18, 2020)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added Special Order Sets (SOS1 and SOS2) to JuMP with default weights to ease the creation of such constraints (#2212).\nAdded functions simplex_iterations, barrier_iterations and node_count (#2201).\nAdded function reduced_cost (#2205).\nImplemented callback_value for affine and quadratic expressions (#2231).\nSupport MutableArithmetics.Zero in objective and constraints (#2219).\nDocumentation improvements:\nMention tutorials in the docs (#2223).\nUpdate COIN-OR links (#2242).\nExplicit link to the documentation of MOI.FileFormats (#2253).\nTypo fixes (#2261).\nContainers improvements:\nFix Base.map for DenseAxisArray (#2235).\nThrow BoundsError if number of indices is incorrect for DenseAxisArray and SparseAxisArray (#2240).\nExtensibility improvements:\nImplement a set_objective method fallback that redirects to set_objective_sense and set_objective_function (#2247).\nAdd parse_constraint method with arbitrary number of arguments (#2051).\nAdd parse_constraint_expr and parse_constraint_head (#2228).","category":"page"},{"location":"release_notes/#[Version-0.21.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.2)-(April-2,-2020)","page":"Release notes","title":"Version 0.21.2 (April 2, 2020)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added relative_gap() to access MOI.RelativeGap() attribute (#2199).\nDocumentation fixes:\nAdded link to source for docstrings in the documentation (#2207).\nAdded docstring for @variables macro (#2216).\nTypo fixes (#2177, #2184, #2182).\nImplementation of methods for Base functions:\nImplemented Base.empty! for JuMP.Model (#2198).\nImplemented Base.conj for JuMP scalar types (#2209).","category":"page"},{"location":"release_notes/#Fixed-36","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixed sum of expression with scalar product in macro (#2178).\nFixed writing of nonlinear models to MathOptFormat (#2181).\nFixed construction of empty SparseAxisArray (#2179).\nFixed constraint with zero function (#2188).","category":"page"},{"location":"release_notes/#[Version-0.21.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.1)-(Feb-18,-2020)","page":"Release notes","title":"Version 0.21.1 (Feb 18, 2020)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Improved the clarity of the with_optimizer deprecation warning.","category":"page"},{"location":"release_notes/#[Version-0.21.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.21.0)-(Feb-16,-2020)","page":"Release notes","title":"Version 0.21.0 (Feb 16, 2020)","text":"","category":"section"},{"location":"release_notes/#Breaking-6","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Deprecated with_optimizer (#2090, #2084, #2141). You can replace with_optimizer by either nothing, optimizer_with_attributes or a closure:\nreplace with_optimizer(Ipopt.Optimizer) by Ipopt.Optimizer.\nreplace with_optimizer(Ipopt.Optimizer, max_cpu_time=60.0) by optimizer_with_attributes(Ipopt.Optimizer, \"max_cpu_time\" => 60.0).\nreplace with_optimizer(Gurobi.Optimizer, env) by () -> Gurobi.Optimizer(env).\nreplace with_optimizer(Gurobi.Optimizer, env, Presolve=0) by optimizer_with_attributes(() -> Gurobi.Optimizer(env), \"Presolve\" => 0).\nalternatively to optimizer_with_attributes, you can also set the attributes separately with set_optimizer_attribute.\nRenamed set_parameter and set_parameters to set_optimizer_attribute and set_optimizer_attributes (#2150).\nBroadcast should now be explicit inside macros. @SDconstraint(model, x >= 1) and @constraint(model, x + 1 in SecondOrderCone()) now throw an error instead of broadcasting 1 along the dimension of x (#2107).\n@SDconstraint(model, x >= 0) is now equivalent to @constraint(model, x in PSDCone()) instead of @constraint(model, (x .- 0) in PSDCone()) (#2107).\nThe macros now create the containers with map instead of for loops, as a consequence, containers created by @expression can now have any element type and containers of constraint references now have concrete element types when possible. This fixes a long-standing issue where @expression could only be used to generate a collection of linear expressions. Now it works for quadratic expressions as well (#2070).\nCalling deepcopy(::AbstractModel) now throws an error.\nThe constraint name is now printed in the model string (#2108).","category":"page"},{"location":"release_notes/#Added-28","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Added support for solver-independent and solver-specific callbacks (#2101).\nAdded write_to_file and read_from_file, supported formats are CBF, LP, MathOptFormat, MPS and SDPA (#2114).\nAdded support for complementarity constraints (#2132).\nAdded support for indicator constraints (#2092).\nAdded support for querying multiple solutions with the result keyword (#2100).\nAdded support for constraining variables on creation (#2128).\nAdded method delete that deletes a vector of variables at once if it is supported by the underlying solver (#2135).\nThe arithmetic between JuMP expression has be refactored into the MutableArithmetics package (#2107).\nImproved error on complex values in NLP (#1978).\nAdded an example of column generation (#2010).","category":"page"},{"location":"release_notes/#Fixed-37","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Incorrect coefficients generated when using Symmetric variables (#2102)","category":"page"},{"location":"release_notes/#[Version-0.20.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.1)-(Oct-18,-2019)","page":"Release notes","title":"Version 0.20.1 (Oct 18, 2019)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Add sections on @variables and @constraints in the documentation (#2062).\nFixed product of sparse matrices for Julia v1.3 (#2063).\nAdded set_objective_coefficient to modify the coefficient of a linear term of the objective function (#2008).\nAdded set_time_limit_sec, unset_time_limit_sec and time_limit_sec to set and query the time limit for the solver in seconds (#2053).","category":"page"},{"location":"release_notes/#[Version-0.20.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.20.0)-(Aug-24,-2019)","page":"Release notes","title":"Version 0.20.0 (Aug 24, 2019)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Documentation updates.\nNumerous bug fixes.\nBetter error messages (#1977, #1978, #1997, #2017).\nPerformance improvements (#1947, #2032).\nAdded LP sensitivity summary functions lp_objective_perturbation_range and lp_rhs_perturbation_range (#1917).\nAdded functions dual_objective_value, raw_status and set_parameter.\nAdded function set_objective_coefficient to modify the coefficient of a linear term of the objective (#2008).\nAdded functions set_normalized_rhs, normalized_rhs, and add_to_function_constant to modify and get the constant part of a constraint (#1935, #1960).\nAdded functions set_normalized_coefficient and normalized_coefficient to modify and get the coefficient of a linear term of a constraint (#1935, #1960).\nNumerous other improvements in MOI 0.9, see the NEWS.md file of MOI for more details.","category":"page"},{"location":"release_notes/#[Version-0.19.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.2)-(June-8,-2019)","page":"Release notes","title":"Version 0.19.2 (June 8, 2019)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in derivatives that could arise in models with nested nonlinear subexpressions.","category":"page"},{"location":"release_notes/#[Version-0.19.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.1)-(May-12,-2019)","page":"Release notes","title":"Version 0.19.1 (May 12, 2019)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Usability and performance improvements.\nBug fixes.","category":"page"},{"location":"release_notes/#[Version-0.19.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.19.0)-(February-15,-2019)","page":"Release notes","title":"Version 0.19.0 (February 15, 2019)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"JuMP 0.19 contains significant breaking changes.","category":"page"},{"location":"release_notes/#Breaking-7","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"JuMP's abstraction layer for communicating with solvers changed from MathProgBase (MPB) to MathOptInterface (MOI). MOI addresses many longstanding design issues. (See @mlubin's slides from JuMP-dev 2018.) JuMP 0.19 is compatible only with solvers that have been updated for MOI. See the installation guide for a list of solvers that have and have not yet been updated.\nMost solvers have been renamed to PackageName.Optimizer. For example, GurobiSolver() is now Gurobi.Optimizer.\nSolvers are no longer added to a model via Model(solver = XXX(kwargs...)). Instead use Model(with_optimizer(XXX, kwargs...)). For example, Model(with_optimizer(Gurobi.Optimizer, OutputFlag=0)).\nJuMP containers (for example, the objects returned by @variable) have been redesigned. Containers.SparseAxisArray replaces JuMPDict, JuMPArray was rewritten (inspired by AxisArrays) and renamed Containers.DenseAxisArray, and you can now request a container type with the container= keyword to the macros. See the corresponding documentation for more details.\nThe statuses returned by solvers have changed. See the possible status values here. The MOI statuses are much richer than the MPB statuses and can be used to distinguish between previously indistinguishable cases (for example, did the solver have a feasible solution when it stopped because of the time limit?).\nStarting values are separate from result values. Use value to query the value of a variable in a solution. Use start_value and set_start_value to get and set an initial starting point provided to the solver. The solutions from previous solves are no longer automatically set as the starting points for the next solve.\nThe data structures for affine and quadratic expressions AffExpr and QuadExpr have changed. Internally, terms are stored in dictionaries instead of lists. Duplicate coefficients can no longer exist. Accessors and iteration methods have changed.\nJuMPNLPEvaluator no longer includes the linear and quadratic parts of the model in the evaluation calls. These are now handled separately to allow NLP solvers that support various types of constraints.\nJuMP solver-independent callbacks have been replaced by solver-specific callbacks. See your favorite solver for more details. (See the note below: No solver-specific callbacks are implemented yet.)\nThe norm() syntax is no longer recognized inside macros. Use the SecondOrderCone() set instead.\nJuMP no longer performs automatic transformation between special quadratic forms and second-order cone constraints. Support for these constraint classes depends on the solver.\nThe symbols :Min and :Max are no longer used as optimization senses. Instead, JuMP uses the OptimizationSense enum from MathOptInterface. @objective(model, Max, ...), @objective(model, Min, ...), @NLobjective(model, Max, ...), and @objective(model, Min, ...) remain valid, but @objective(m, :Max, ...) is no longer accepted.\nThe sign conventions for duals has changed in some cases for consistency with conic duality (see the documentation). The shadow_price helper method returns duals with signs that match conventional LP interpretations of dual values as sensitivities of the objective value to relaxations of constraints.\n@constraintref is no longer defined. Instead, create the appropriate container to hold constraint references manually. For example,\nconstraints = Dict() # Optionally, specify types for improved performance.\nfor i in 1:N\n  constraints[i] = @constraint(model, ...)\nend\nThe lowerbound, upperbound, and basename keyword arguments to the @variable macro have been renamed to lower_bound, upper_bound, and base_name, for consistency with JuMP's new style recommendations.\nWe rely on broadcasting syntax to apply accessors to collections of variables, for example, value.(x) instead of getvalue(x) for collections. (Use value(x) when x is a scalar object.)","category":"page"},{"location":"release_notes/#Added-29","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Splatting (like f(x...)) is recognized in restricted settings in nonlinear expressions.\nSupport for deleting constraints and variables.\nThe documentation has been completely rewritten using docstrings and Documenter.\nSupport for modeling mixed conic and quadratic models (for example, conic models with quadratic objectives and bi-linear matrix inequalities).\nSignificantly improved support for modeling new types of constraints and for extending JuMP's macros.\nSupport for providing dual warm starts.\nImproved support for accessing solver-specific attributes (for example, the irreducible inconsistent subsystem).\nExplicit control of whether symmetry-enforcing constraints are added to PSD constraints.\nSupport for modeling exponential cones.\nSignificant improvements in internal code quality and testing.\nStyle and naming guidelines.\nDirect mode and manual mode provide explicit control over when copies of a model are stored or regenerated. See the corresponding documentation.","category":"page"},{"location":"release_notes/#Regressions","page":"Release notes","title":"Regressions","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"There are known regressions from JuMP 0.18 that will be addressed in a future release (0.19.x or later):","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Performance regressions in model generation (issue). Please file an issue anyway if you notice a significant performance regression. We have plans to address a number of performance issues, but we might not be aware of all of them.\nFast incremental NLP solves are not yet reimplemented (issue).\nWe do not yet have an implementation of solver-specific callbacks.\nThe column generation syntax in @variable has been removed (that is, the objective, coefficients, and inconstraints keyword arguments). Support for column generation will be re-introduced in a future release.\nThe ability to solve the continuous relaxation (that is, via solve(model; relaxation = true)) is not yet reimplemented (issue).","category":"page"},{"location":"release_notes/#[Version-0.18.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.5)-(December-1,-2018)","page":"Release notes","title":"Version 0.18.5 (December 1, 2018)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Support views in some derivative evaluation functions.\nImproved compatibility with PackageCompiler.","category":"page"},{"location":"release_notes/#[Version-0.18.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.4)-(October-8,-2018)","page":"Release notes","title":"Version 0.18.4 (October 8, 2018)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in model printing on Julia 0.7 and 1.0.","category":"page"},{"location":"release_notes/#[Version-0.18.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.3)-(October-1,-2018)","page":"Release notes","title":"Version 0.18.3 (October 1, 2018)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Add support for Julia v1.0 (Thanks @ExpandingMan)\nFix matrix expressions with quadratic functions (#1508)","category":"page"},{"location":"release_notes/#[Version-0.18.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.2)-(June-10,-2018)","page":"Release notes","title":"Version 0.18.2 (June 10, 2018)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in second-order derivatives when expressions are present (#1319)\nFix a bug in @constraintref (#1330)","category":"page"},{"location":"release_notes/#[Version-0.18.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.1)-(April-9,-2018)","page":"Release notes","title":"Version 0.18.1 (April 9, 2018)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix for nested tuple destructuring (#1193)\nPreserve internal model when relaxation=true (#1209)\nMinor bug fixes and updates for example","category":"page"},{"location":"release_notes/#[Version-0.18.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.18.0)-(July-27,-2017)","page":"Release notes","title":"Version 0.18.0 (July 27, 2017)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Drop support for Julia 0.5.\nUpdate for ForwardDiff 0.5.\nMinor bug fixes.","category":"page"},{"location":"release_notes/#[Version-0.17.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.1)-(June-9,-2017)","page":"Release notes","title":"Version 0.17.1 (June 9, 2017)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Use of constructconstraint! in @SDconstraint.\nMinor bug fixes.","category":"page"},{"location":"release_notes/#[Version-0.17.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.17.0)-(May-27,-2017)","page":"Release notes","title":"Version 0.17.0 (May 27, 2017)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Breaking change: Mixing quadratic and conic constraints is no longer supported.\nBreaking change: The getvariable and getconstraint functions are replaced by indexing on the corresponding symbol. For instance, to access the variable with name x, one should now write m[:x] instead of getvariable(m, :x). As a consequence, creating a variable and constraint with the same name now triggers a warning, and accessing one of them afterwards throws an error. This change is breaking only in the latter case.\nAddition of the getobjectivebound function that mirrors the functionality of the MathProgBase getobjbound function except that it takes into account transformations performed by JuMP.\nMinor bug fixes.","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The following changes are primarily of interest to developers of JuMP extensions:","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The new syntax @constraint(model, expr in Cone) creates the constraint ensuring that expr is inside Cone. The Cone argument is passed to constructconstraint! which enables the call to the dispatched to an extension.\nThe @variable macro now calls constructvariable! instead of directly calling the Variable constructor. Extra arguments and keyword arguments passed to @variable are passed to constructvariable! which enables the call to be dispatched to an extension.\nRefactor the internal function conicdata (used build the MathProgBase conic model) into smaller sub-functions to make these parts reusable by extensions.","category":"page"},{"location":"release_notes/#[Version-0.16.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.2)-(March-28,-2017)","page":"Release notes","title":"Version 0.16.2 (March 28, 2017)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Minor bug fixes and printing tweaks\nAddress deprecation warnings for Julia 0.6","category":"page"},{"location":"release_notes/#[Version-0.16.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.1)-(March-7,-2017)","page":"Release notes","title":"Version 0.16.1 (March 7, 2017)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Better support for AbstractArray in JuMP (Thanks @tkoolen)\nMinor bug fixes","category":"page"},{"location":"release_notes/#[Version-0.16.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.16.0)-(February-23,-2017)","page":"Release notes","title":"Version 0.16.0 (February 23, 2017)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Breaking change: JuMP no longer has a mechanism for selecting solvers by default (the previous mechanism was flawed and incompatible with Julia 0.6). Not specifying a solver before calling solve() will result in an error.\nBreaking change: User-defined functions are no longer global. The first argument to JuMP.register is now a JuMP Model object within whose scope the function will be registered. Calling JuMP.register without a Model now produces an error.\nBreaking change: Use the new JuMP.fix method to fix a variable to a value or to update the value to which a variable is fixed. Calling setvalue on a fixed variable now results in an error in order to avoid silent behavior changes. (Thanks @joaquimg)\nNonlinear expressions now print out similarly to linear/quadratic expressions (useful for debugging!)\nNew category keyword to @variable. Used for specifying categories of anonymous variables.\nCompatibility with Julia 0.6-dev.\nMinor fixes and improvements (Thanks @cossio, @ccoffrin, @blegat)","category":"page"},{"location":"release_notes/#[Version-0.15.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.1)-(January-31,-2017)","page":"Release notes","title":"Version 0.15.1 (January 31, 2017)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Bugfix for @LinearConstraints and friends","category":"page"},{"location":"release_notes/#[Version-0.15.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.15.0)-(December-22,-2016)","page":"Release notes","title":"Version 0.15.0 (December 22, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Julia 0.5.0 is the minimum required version for this release.\nDocument support for BARON solver\nEnable info callbacks in more states than before, for example, for recording solutions. New when argument to addinfocallback (#814, thanks @yeesian)\nImproved support for anonymous variables. This includes new warnings for potentially confusing use of the traditional non-anonymous syntax:\nWhen multiple variables in a model are given the same name\nWhen non-symbols are used as names, for example, @variable(m, x[1][1:N])\nImprovements in iterating over JuMP containers (#836, thanks @IssamT)\nSupport for writing variable names in .lp file output (Thanks @leethargo)\nSupport for querying duals to SDP problems (Thanks @blegat)\nThe comprehension syntax with curly braces sum{}, prod{}, and norm2{} has been deprecated in favor of Julia's native comprehension syntax sum(), prod() and norm() as previously announced. (For early adopters of the new syntax, norm2() was renamed to norm() without deprecation.)\nUnit tests rewritten to use Base.Test instead of FactCheck\nImproved support for operations with matrices of JuMP types (Thanks @ExpandingMan)\nThe syntax to halt a solver from inside a callback has changed from throw(CallbackAbort()) to return JuMP.StopTheSolver\nMinor bug fixes","category":"page"},{"location":"release_notes/#[Version-0.14.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.2)-(December-12,-2016)","page":"Release notes","title":"Version 0.14.2 (December 12, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Allow singleton anonymous variables (includes bugfix)","category":"page"},{"location":"release_notes/#[Version-0.14.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.1)-(September-12,-2016)","page":"Release notes","title":"Version 0.14.1 (September 12, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"More consistent handling of states in informational callbacks, includes a new when parameter to addinfocallback for specifying in which state an informational callback should be called.","category":"page"},{"location":"release_notes/#[Version-0.14.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.14.0)-(August-7,-2016)","page":"Release notes","title":"Version 0.14.0 (August 7, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Compatibility with Julia 0.5 and ForwardDiff 0.2\nSupport for \"anonymous\" variables, constraints, expressions, and parameters, for example, x = @variable(m, [1:N]) instead of @variable(m, x[1:N])\nSupport for retrieving constraints from a model by name via getconstraint\n@NLconstraint now returns constraint references (as expected).\nSupport for vectorized expressions within lazy constraints\nOn Julia 0.5, parse new comprehension syntax sum(x[i] for i in 1:N if isodd(i)) instead of sum{ x[i], i in 1:N; isodd(i) }. The old syntax with curly braces will be deprecated in JuMP 0.15.\nNow possible to provide nonlinear expressions as \"raw\" Julia Expr objects instead of using JuMP's nonlinear macros. This input format is useful for programmatically generated expressions.\ns/Mathematical Programming/Mathematical Optimization/\nSupport for local cuts (Thanks to @madanim, Mehdi Madani)\nDocument Xpress interface developed by @joaquimg, Joaquim Dias Garcia\nMinor bug and deprecation fixes (Thanks @odow, @jrevels)","category":"page"},{"location":"release_notes/#[Version-0.13.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.2)-(May-16,-2016)","page":"Release notes","title":"Version 0.13.2 (May 16, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Compatibility update for MathProgBase","category":"page"},{"location":"release_notes/#[Version-0.13.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.1)-(May-3,-2016)","page":"Release notes","title":"Version 0.13.1 (May 3, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix broken deprecation for registerNLfunction.","category":"page"},{"location":"release_notes/#[Version-0.13.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.13.0)-(April-29,-2016)","page":"Release notes","title":"Version 0.13.0 (April 29, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Most exported methods and macros have been renamed to avoid camelCase. See the list of changes here. There is a 1-1 mapping from the old names to the new, and it is safe to simply replace the names to update existing models.\nSpecify variable lower/upper bounds in @variable using the lowerbound and upperbound keyword arguments.\nChange name printed for variable using the basename keyword argument to @variable.\nNew @variables macro allows multi-line declaration of groups of variables.\nA number of solver methods previously available only through MathProgBase are now exposed directly in JuMP. The fix was recorded live.\nCompatibility fixes with Julia 0.5.\nThe \"end\" indexing syntax is no longer supported within JuMPArrays which do not use 1-based indexing until upstream issues are resolved, see here.","category":"page"},{"location":"release_notes/#[Version-0.12.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.2)-(March-9,-2016)","page":"Release notes","title":"Version 0.12.2 (March 9, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Small fixes for nonlinear optimization","category":"page"},{"location":"release_notes/#[Version-0.12.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.1)-(March-1,-2016)","page":"Release notes","title":"Version 0.12.1 (March 1, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a regression in slicing for JuMPArrays (when not using 1-based indexing)","category":"page"},{"location":"release_notes/#[Version-0.12.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.12.0)-(February-27,-2016)","page":"Release notes","title":"Version 0.12.0 (February 27, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"The automatic differentiation functionality has been completely rewritten with a number of user-facing changes:\n@defExpr and @defNLExpr now take the model as the first argument. The previous one-argument version of @defExpr is deprecated; all expressions should be named. For example, replace @defExpr(2x+y) with @defExpr(jump_model, my_expr, 2x+y).\nJuMP no longer uses Julia's variable binding rules for efficiently re-solving a sequence of nonlinear models. Instead, we have introduced nonlinear parameters. This is a breaking change, so we have added a warning message when we detect models that may depend on the old behavior.\nSupport for user-defined functions integrated within nonlinear JuMP expressions.\nReplaced iteration over AffExpr with Number-like scalar iteration; previous iteration behavior is now available via linearterms(::AffExpr).\nStopping the solver via throw(CallbackAbort()) from a callback no longer triggers an exception. Instead, solve() returns UserLimit status.\ngetDual() now works for conic problems (Thanks @emreyamangil.)","category":"page"},{"location":"release_notes/#[Version-0.11.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.3)-(February-4,-2016)","page":"Release notes","title":"Version 0.11.3 (February 4, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Bug-fix for problems with quadratic objectives and semidefinite constraints","category":"page"},{"location":"release_notes/#[Version-0.11.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.2)-(January-14,-2016)","page":"Release notes","title":"Version 0.11.2 (January 14, 2016)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Compatibility update for Mosek","category":"page"},{"location":"release_notes/#[Version-0.11.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.1)-(December-1,-2015)","page":"Release notes","title":"Version 0.11.1 (December 1, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Remove usage of @compat in tests.\nFix updating quadratic objectives for nonlinear models.","category":"page"},{"location":"release_notes/#[Version-0.11.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.11.0)-(November-30,-2015)","page":"Release notes","title":"Version 0.11.0 (November 30, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Julia 0.4.0 is the minimum required version for this release.\nFix for scoping semantics of index variables in sum{}. Index variables no longer leak into the surrounding scope.\nAddition of the solve(m::Model, relaxation=true) keyword argument to solve the standard continuous relaxation of model m\nThe getConstraintBounds() method allows access to the lower and upper bounds of all constraints in a (nonlinear) model.\nUpdate for breaking changes in MathProgBase","category":"page"},{"location":"release_notes/#[Version-0.10.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.3)-(November-20,-2015)","page":"Release notes","title":"Version 0.10.3 (November 20, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a rare error when parsing quadratic expressions\nFix Variable() constructor with default arguments\nDetect unrecognized keywords in solve()","category":"page"},{"location":"release_notes/#[Version-0.10.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.2)-(September-28,-2015)","page":"Release notes","title":"Version 0.10.2 (September 28, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix for deprecation warnings","category":"page"},{"location":"release_notes/#[Version-0.10.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.1)-(September-3,-2015)","page":"Release notes","title":"Version 0.10.1 (September 3, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixes for ambiguity warnings.\nFix for breaking change in precompilation syntax in Julia 0.4-pre","category":"page"},{"location":"release_notes/#[Version-0.10.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.10.0)-(August-31,-2015)","page":"Release notes","title":"Version 0.10.0 (August 31, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Support (on Julia 0.4 and later) for conditions in indexing @defVar and @addConstraint constructs, for example, @defVar(m, x[i=1:5,j=1:5; i+j >= 3])\nSupport for vectorized operations on Variables and expressions. See the documentation for details.\nNew getVar() method to access variables in a model by name\nSupport for semidefinite programming.\nDual solutions are now available for general nonlinear problems. You may call getDual on a reference object for a nonlinear constraint, and getDual on a variable object for Lagrange multipliers from active bounds.\nIntroduce warnings for two common performance traps: too many calls to getValue() on a collection of variables and use of the + operator in a loop to sum expressions.\nSecond-order cone constraints can be written directly with the norm() and norm2{} syntax.\nImplement MathProgBase interface for querying Hessian-vector products.\nIteration over JuMPContainers is deprecated; instead, use the keys and values functions, and zip(keys(d),values(d)) for the old behavior.\n@defVar returns Array{Variable,N} when each of N index sets are of the form 1:nᵢ.\nModule precompilation: on Julia 0.4 and later, using JuMP is now much faster.","category":"page"},{"location":"release_notes/#[Version-0.9.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.3)-(August-11,-2015)","page":"Release notes","title":"Version 0.9.3 (August 11, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fixes for FactCheck testing on julia v0.4.","category":"page"},{"location":"release_notes/#[Version-0.9.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.2)-(June-27,-2015)","page":"Release notes","title":"Version 0.9.2 (June 27, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix bug in @addConstraints.","category":"page"},{"location":"release_notes/#[Version-0.9.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.1)-(April-25,-2015)","page":"Release notes","title":"Version 0.9.1 (April 25, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix for Julia 0.4-dev.\nSmall infrastructure improvements for extensions.","category":"page"},{"location":"release_notes/#[Version-0.9.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.9.0)-(April-18,-2015)","page":"Release notes","title":"Version 0.9.0 (April 18, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Comparison operators for constructing constraints (for example, 2x >= 1) have been deprecated. Instead, construct the constraints explicitly in the @addConstraint macro to add them to the model, or in the @LinearConstraint macro to create a stand-alone linear constraint instance.\ngetValue() method implemented to compute the value of a nonlinear subexpression\nJuMP is now released under the Mozilla Public License version 2.0 (was previously LGPL). MPL is a copyleft license which is less restrictive than LGPL, especially for embedding JuMP within other applications.\nA number of performance improvements in ReverseDiffSparse for computing derivatives.\nMathProgBase.getsolvetime(m) now returns the solution time reported by the solver, if available. (Thanks @odow, Oscar Dowson)\nFormatting fix for LP format output. (Thanks @sbebo, Leonardo Taccari).","category":"page"},{"location":"release_notes/#[Version-0.8.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.8.0)-(February-17,-2015)","page":"Release notes","title":"Version 0.8.0 (February 17, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Nonlinear subexpressions now supported with the @defNLExpr macro.\nSCS supported for solving second-order conic problems.\nsetXXXCallback family deprecated in favor of addXXXCallback.\nMultiple callbacks of the same type can be registered.\nAdded support for informational callbacks via addInfoCallback.\nA CallbackAbort exception can be thrown from callback to safely exit optimization.","category":"page"},{"location":"release_notes/#[Version-0.7.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.4)-(February-4,-2015)","page":"Release notes","title":"Version 0.7.4 (February 4, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Reduced costs and linear constraint duals are now accessible when quadratic constraints are present.\nTwo-sided nonlinear constraints are supported.\nMethods for accessing the number of variables and constraints in a model are renamed.\nNew default procedure for setting initial values in nonlinear optimization: project zero onto the variable bounds.\nSmall bug fixes.","category":"page"},{"location":"release_notes/#[Version-0.7.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.3)-(January-14,-2015)","page":"Release notes","title":"Version 0.7.3 (January 14, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a method ambiguity conflict with Compose.jl (cosmetic fix)","category":"page"},{"location":"release_notes/#[Version-0.7.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.2)-(January-9,-2015)","page":"Release notes","title":"Version 0.7.2 (January 9, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in sum(::JuMPDict)\nAdded the setCategory function to change a variables category (for example, continuous or binary)","category":"page"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"after construction, and getCategory to retrieve the variable category.","category":"page"},{"location":"release_notes/#[Version-0.7.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.1)-(January-2,-2015)","page":"Release notes","title":"Version 0.7.1 (January 2, 2015)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in parsing linear expressions in macros. Affects only Julia 0.4 and later.","category":"page"},{"location":"release_notes/#[Version-0.7.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.7.0)-(December-29,-2014)","page":"Release notes","title":"Version 0.7.0 (December 29, 2014)","text":"","category":"section"},{"location":"release_notes/#Linear/quadratic/conic-programming","page":"Release notes","title":"Linear/quadratic/conic programming","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Breaking change: The syntax for column-wise model generation has been changed to use keyword arguments in @defVar.\nOn Julia 0.4 and later, variables and coefficients may be multiplied in any order within macros. That is, variable*coefficient is now valid syntax.\nECOS supported for solving second-order conic problems.","category":"page"},{"location":"release_notes/#_nonlinear_programming_release_notes","page":"Release notes","title":"Nonlinear programming","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Support for skipping model generation when solving a sequence of nonlinear models with changing data.\nFix a memory leak when solving a sequence of nonlinear models.\nThe @addNLConstraint macro now supports the three-argument version to define sets of nonlinear constraints.\nKNITRO supported as a nonlinear solver.\nSpeed improvements for model generation.\nThe @addNLConstraints macro supports adding multiple (groups of) constraints at once. Syntax is similar to @addConstraints.\nDiscrete variables allowed in nonlinear problems for solvers which support them (currently only KNITRO).","category":"page"},{"location":"release_notes/#General","page":"Release notes","title":"General","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Starting values for variables may now be specified with @defVar(m, x, start=value).\nThe setSolver function allows users to change the solver subsequent to model creation.\nSupport for \"fixed\" variables via the @defVar(m, x == 1) syntax.\nUnit tests rewritten to use FactCheck.jl, improved testing across solvers.","category":"page"},{"location":"release_notes/#[Version-0.6.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.3)-(October-19,-2014)","page":"Release notes","title":"Version 0.6.3 (October 19, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in multiplying two AffExpr objects.","category":"page"},{"location":"release_notes/#[Version-0.6.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.2)-(October-11,-2014)","page":"Release notes","title":"Version 0.6.2 (October 11, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Further improvements and bug fixes for printing.\nFixed a bug in @defExpr.\nSupport for accessing expression graphs through the MathProgBase NLP interface.","category":"page"},{"location":"release_notes/#[Version-0.6.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.1)-(September-19,-2014)","page":"Release notes","title":"Version 0.6.1 (September 19, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Improvements and bug fixes for printing.","category":"page"},{"location":"release_notes/#[Version-0.6.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.6.0)-(September-9,-2014)","page":"Release notes","title":"Version 0.6.0 (September 9, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Julia 0.3.0 is the minimum required version for this release.\nbuildInternalModel(m::Model) added to build solver-level model in memory without optimizing.\nDeprecate load_model_only keyword argument to solve.\nAdd groups of constraints with @addConstraints macro.\nUnicode operators now supported, including ∑ for sum, ∏ for prod, and ≤/≥\nQuadratic constraints supported in @addConstraint macro.\nQuadratic objectives supported in @setObjective macro.\nMathProgBase solver-independent interface replaces Ipopt-specific interface for nonlinear problems\nBreaking change: IpoptOptions no longer supported to specify solver options, use m = Model(solver=IpoptSolver(options...)) instead.\nNew solver interfaces: ECOS, NLopt, and nonlinear support for MOSEK\nNew option to control whether the lazy constraint callback is executed at each node in the B&B tree or just when feasible solutions are found\nAdd support for semicontinuous and semi-integer variables for those solvers that support them.\nAdd support for index dependencies (for example, triangular indexing) in @defVar, @addConstraint, and @defExpr (for example, @defVar(m, x[i=1:10,j=i:10])).\nThis required some changes to the internal structure of JuMP containers, which may break code that explicitly stored JuMPDict objects.","category":"page"},{"location":"release_notes/#[Version-0.5.8](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.8)-(September-24,-2014)","page":"Release notes","title":"Version 0.5.8 (September 24, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug with specifying solvers (affects Julia 0.2 only)","category":"page"},{"location":"release_notes/#[Version-0.5.7](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.7)-(September-5,-2014)","page":"Release notes","title":"Version 0.5.7 (September 5, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in printing models","category":"page"},{"location":"release_notes/#[Version-0.5.6](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.6)-(September-2,-2014)","page":"Release notes","title":"Version 0.5.6 (September 2, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Add support for semicontinuous and semi-integer variables for those solvers that support them.\nBreaking change: Syntax for Variable() constructor has changed (use of this interface remains discouraged)\nUpdate for breaking changes in MathProgBase","category":"page"},{"location":"release_notes/#[Version-0.5.5](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.5)-(July-6,-2014)","page":"Release notes","title":"Version 0.5.5 (July 6, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix bug with problem modification: adding variables that did not appear in existing constraints or objective.","category":"page"},{"location":"release_notes/#[Version-0.5.4](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.4)-(June-19,-2014)","page":"Release notes","title":"Version 0.5.4 (June 19, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Update for breaking change in MathProgBase which reduces loading times for using JuMP\nFix error when MIPs not solved to optimality","category":"page"},{"location":"release_notes/#[Version-0.5.3](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.3)-(May-21,-2014)","page":"Release notes","title":"Version 0.5.3 (May 21, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Update for breaking change in ReverseDiffSparse","category":"page"},{"location":"release_notes/#[Version-0.5.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.2)-(May-9,-2014)","page":"Release notes","title":"Version 0.5.2 (May 9, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix compatibility with Julia 0.3 prerelease","category":"page"},{"location":"release_notes/#[Version-0.5.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.1)-(May-5,-2014)","page":"Release notes","title":"Version 0.5.1 (May 5, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix a bug in coefficient handling inside lazy constraints and user cuts","category":"page"},{"location":"release_notes/#[Version-0.5.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.5.0)-(May-2,-2014)","page":"Release notes","title":"Version 0.5.0 (May 2, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Support for nonlinear optimization with exact, sparse second-order derivatives automatically computed. Ipopt is currently the only solver supported.\ngetValue for AffExpr and QuadExpr\nBreaking change: getSolverModel replaced by getInternalModel, which returns the internal MathProgBase-level model\nGroups of constraints can be specified with @addConstraint (see documentation for details). This is not a breaking change.\ndot(::JuMPDict{Variable},::JuMPDict{Variable}) now returns the corresponding quadratic expression.","category":"page"},{"location":"release_notes/#[Version-0.4.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.1)-(March-24,-2014)","page":"Release notes","title":"Version 0.4.1 (March 24, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Fix bug where change in objective sense was ignored when re-solving a model.\nFix issue with handling zero coefficients in AffExpr.","category":"page"},{"location":"release_notes/#[Version-0.4.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.4.0)-(March-10,-2014)","page":"Release notes","title":"Version 0.4.0 (March 10, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Support for SOS1 and SOS2 constraints.\nSolver-independent callback for user heuristics.\ndot and sum implemented for JuMPDict objects. Now you can say @addConstraint(m, dot(a,x) <= b).\nDevelopers: support for extensions to JuMP. See definition of Model in src/JuMP.jl for more details.\nOption to construct the low-level model before optimizing.","category":"page"},{"location":"release_notes/#[Version-0.3.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.2)-(February-17,-2014)","page":"Release notes","title":"Version 0.3.2 (February 17, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Improved model printing\nPreliminary support for IJulia output","category":"page"},{"location":"release_notes/#[Version-0.3.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.1)-(January-30,-2014)","page":"Release notes","title":"Version 0.3.1 (January 30, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Documentation updates\nSupport for MOSEK\nCPLEXLink renamed to CPLEX","category":"page"},{"location":"release_notes/#[Version-0.3.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.3.0)-(January-21,-2014)","page":"Release notes","title":"Version 0.3.0 (January 21, 2014)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Unbounded/infeasibility rays: getValue() will return the corresponding components of an unbounded ray when a model is unbounded, if supported by the selected solver. getDual() will return an infeasibility ray (Farkas proof) if a model is infeasible and the selected solver supports this feature.\nSolver-independent callbacks for user generated cuts.\nUse new interface for solver-independent QCQP.\nsetlazycallback renamed to setLazyCallback for consistency.","category":"page"},{"location":"release_notes/#[Version-0.2.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.2.0)-(December-15,-2013)","page":"Release notes","title":"Version 0.2.0 (December 15, 2013)","text":"","category":"section"},{"location":"release_notes/#Breaking-8","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Objective sense is specified in setObjective instead of in the Model constructor.\nlpsolver and mipsolver merged into single solver option.","category":"page"},{"location":"release_notes/#Added-30","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Problem modification with efficient LP restarts and MIP warm-starts.\nRelatedly, column-wise modeling now supported.\nSolver-independent callbacks supported. Currently we support only a \"lazy constraint\" callback, which works with Gurobi, CPLEX, and GLPK. More callbacks coming soon.","category":"page"},{"location":"release_notes/#[Version-0.1.2](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.2)-(November-16,-2013)","page":"Release notes","title":"Version 0.1.2 (November 16, 2013)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Bug fixes for printing, improved error messages.\nAllow AffExpr to be used in macros; for example, ex = y + z; @addConstraint(m, x + 2*ex <= 3)","category":"page"},{"location":"release_notes/#[Version-0.1.1](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.1)-(October-23,-2013)","page":"Release notes","title":"Version 0.1.1 (October 23, 2013)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Update for solver specification API changes in MathProgBase.","category":"page"},{"location":"release_notes/#[Version-0.1.0](https://github.com/jump-dev/JuMP.jl/releases/tag/v0.1.0)-(October-3,-2013)","page":"Release notes","title":"Version 0.1.0 (October 3, 2013)","text":"","category":"section"},{"location":"release_notes/","page":"Release notes","title":"Release notes","text":"Initial public release.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\n    import HiGHS\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/constraints/#jump_constraints","page":"Constraints","title":"Constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"JuMP is based on the MathOptInterface (MOI) API. Because of this, JuMP uses the following standard form to represent problems:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"beginalign\n     min_x in mathbbR^n  f_0(x)\n    \n     textst  f_i(x)  in mathcalS_i  i = 1 ldots m\nendalign","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Each constraint, f_i(x) in mathcalS_i, is composed of a function and a set. For example, instead of calling a^top x le b a less-than-or-equal-to constraint, we say that it is a scalar-affine-in-less-than constraint, where the function a^top x belongs to the less-than set (-infty b. We use the shorthand function-in-set to refer to constraints composed of different types of functions and sets.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"This page explains how to write various types of constraints in JuMP. For nonlinear constraints, see Nonlinear Modeling instead.","category":"page"},{"location":"manual/constraints/#Add-a-constraint","page":"Constraints","title":"Add a constraint","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Add a constraint to a JuMP model using the @constraint macro. The syntax to use depends on the type of constraint you wish to add.","category":"page"},{"location":"manual/constraints/#Add-a-linear-constraint","page":"Constraints","title":"Add a linear constraint","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Create linear constraints using the @constraint macro:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3]);\n\njulia> @constraint(model, c1, sum(x) <= 1)\nc1 : x[1] + x[2] + x[3] ≤ 1\n\njulia> @constraint(model, c2, x[1] + 2 * x[3] >= 2)\nc2 : x[1] + 2 x[3] ≥ 2\n\njulia> @constraint(model, c3, sum(i * x[i] for i in 1:3) == 3)\nc3 : x[1] + 2 x[2] + 3 x[3] = 3\n\njulia> @constraint(model, c4, 4 <= 2 * x[2] <= 5)\nc4 : 2 x[2] ∈ [4, 5]","category":"page"},{"location":"manual/constraints/#Normalization","page":"Constraints","title":"Normalization","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"JuMP normalizes constraints by moving all of the terms containing variables to the left-hand side and all of the constant terms to the right-hand side. Thus, we get:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, c, 2x + 1 <= 4x + 4)\nc : -2 x ≤ 3","category":"page"},{"location":"manual/constraints/#quad_constraints","page":"Constraints","title":"Add a quadratic constraint","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In addition to affine functions, JuMP also supports constraints with quadratic terms. For example:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> @variable(model, t >= 0)\nt\n\njulia> @constraint(model, my_q, x[1]^2 + x[2]^2 <= t^2)\nmy_q : x[1]² + x[2]² - t² ≤ 0","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"tip: Tip\nBecause solvers can take advantage of the knowledge that a constraint is quadratic, prefer adding quadratic constraints using @constraint, rather than @NLconstraint.","category":"page"},{"location":"manual/constraints/#Vectorized-constraints","page":"Constraints","title":"Vectorized constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"You can also add constraints to JuMP using vectorized linear algebra. For example:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> A = [1 2; 3 4]\n2×2 Matrix{Int64}:\n 1  2\n 3  4\n\njulia> b = [5, 6]\n2-element Vector{Int64}:\n 5\n 6\n\njulia> @constraint(model, con_vector, A * x == b)\ncon_vector : [x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Zeros(2)\n\njulia> @constraint(model, con_scalar, A * x .== b)\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}, ScalarShape}}:\n con_scalar : x[1] + 2 x[2] = 5\n con_scalar : 3 x[1] + 4 x[2] = 6","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The two constraints, == and .== are similar, but subtly different. The first creates a single constraint that is a MOI.VectorAffineFunction in MOI.Zeros constraint. The second creates a vector of MOI.ScalarAffineFunction in MOI.EqualTo constraints.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Which formulation to choose depends on the solver, and what you want to do with the constraint object con_vector or con_scalar.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"If you are using a conic solver, expect the dual of con_vector to be a Vector{Float64}, and do not intend to delete a row in the constraint, choose the == formulation.\nIf you are using a solver that expects a list of scalar constraints, for example HiGHS, or you wish to delete part of the constraint or access a single row of the constraint, for example, dual(con_scalar[2]), then use the broadcast .==.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"JuMP reformulates both constraints into the other form if needed by the solver, but choosing the right format for a particular solver is more efficient.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"You can also use <=, .<= , >=, and .>= as comparison operators in the constraint.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, A * x <= b)\n[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonpositives(2)\n\njulia> @constraint(model, A * x .<= b)\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n x[1] + 2 x[2] ≤ 5\n 3 x[1] + 4 x[2] ≤ 6\n\njulia> @constraint(model, A * x >= b)\n[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonnegatives(2)\n\njulia> @constraint(model, A * x .>= b)\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:\n x[1] + 2 x[2] ≥ 5\n 3 x[1] + 4 x[2] ≥ 6","category":"page"},{"location":"manual/constraints/#Vectorized-matrix-constraints","page":"Constraints","title":"Vectorized matrix constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In most cases, you cannot use the non-broadcasting syntax for general matrices. For example:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, X[1:2, 1:2])\n2×2 Matrix{VariableRef}:\n X[1,1]  X[1,2]\n X[2,1]  X[2,2]\n\njulia> @constraint(model, X >= 0)\nERROR: At none:1: `@constraint(model, X >= 0)`: Unsupported matrix in vector-valued set. Did you mean to use the broadcasting syntax `.>=` instead? Alternatively, perhaps you are missing a set argument like `@constraint(model, X >= 0, PSDCone())` or `@constraint(model, X >= 0, HermmitianPSDCone())`.\nStacktrace:\n[...]","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Instead, to represent matrix inequalities you must always use the element-wise broadcasting .==, .>=, or .<=, or use the Set inequality syntax.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"There are two exceptions: if the result of the left-hand side minus the right-hand side is a LinearAlgebra.Symmetric matrix or a LinearAlgebra.Hermitian matrix, you may use the non-broadcasting equality syntax:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> using LinearAlgebra\n\njulia> model = Model();\n\njulia> @variable(model, X[1:2, 1:2], Symmetric)\n2×2 Symmetric{VariableRef, Matrix{VariableRef}}:\n X[1,1]  X[1,2]\n X[1,2]  X[2,2]\n\njulia> @constraint(model, X == LinearAlgebra.I)\n[X[1,1] - 1  X[1,2];\n X[1,2]      X[2,2] - 1] ∈ Zeros()","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Despite the model showing the matrix in Zeros, this will add only three rows to the constraint matrix because the symmetric constraints are redundant. In contrast, the broadcasting syntax adds four linear constraints:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, X .== LinearAlgebra.I)\n2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}, ScalarShape}}:\n X[1,1] = 1  X[1,2] = 0\n X[1,2] = 0  X[2,2] = 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The same holds for LinearAlgebra.Hermitian matrices:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> using LinearAlgebra\n\njulia> model = Model();\n\njulia> @variable(model, X[1:2, 1:2] in HermitianPSDCone())\n2×2 Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(X[1,1])                    real(X[1,2]) + imag(X[1,2]) im\n real(X[1,2]) - imag(X[1,2]) im  real(X[2,2])\n\njulia> @constraint(model, X == LinearAlgebra.I)\n[real(X[1,1]) - 1                real(X[1,2]) + imag(X[1,2]) im;\n real(X[1,2]) - imag(X[1,2]) im  real(X[2,2]) - 1] ∈ Zeros()\n\njulia> @constraint(model, X .== LinearAlgebra.I)\n2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{ComplexF64}, MathOptInterface.EqualTo{ComplexF64}}, ScalarShape}}:\n real(X[1,1]) = 1                    real(X[1,2]) + imag(X[1,2]) im = 0\n real(X[1,2]) - imag(X[1,2]) im = 0  real(X[2,2]) = 1","category":"page"},{"location":"manual/constraints/#Containers-of-constraints","page":"Constraints","title":"Containers of constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The @constraint macro supports creating collections of constraints. We'll cover some brief syntax here; read the Constraint containers section for more details:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Create arrays of constraints:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3]);\n\njulia> @constraint(model, c[i=1:3], x[i] <= i^2)\n3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n c[1] : x[1] ≤ 1\n c[2] : x[2] ≤ 4\n c[3] : x[3] ≤ 9\n\njulia> c[2]\nc[2] : x[2] ≤ 4","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Sets can be any Julia type that supports iteration:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3]);\n\njulia> @constraint(model, c[i=2:3, [\"red\", \"blue\"]], x[i] <= i^2)\n2-dimensional DenseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},2,...} with index sets:\n    Dimension 1, 2:3\n    Dimension 2, [\"red\", \"blue\"]\nAnd data, a 2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n c[2,red] : x[2] ≤ 4  c[2,blue] : x[2] ≤ 4\n c[3,red] : x[3] ≤ 9  c[3,blue] : x[3] ≤ 9\n\njulia> c[2, \"red\"]\nc[2,red] : x[2] ≤ 4","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Sets can depend upon previous indices:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3]);\n\njulia> @constraint(model, c[i=1:3, j=i:3], x[i] <= j)\nJuMP.Containers.SparseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, 2, Tuple{Int64, Int64}} with 6 entries:\n  [1, 1]  =  c[1,1] : x[1] ≤ 1\n  [1, 2]  =  c[1,2] : x[1] ≤ 2\n  [1, 3]  =  c[1,3] : x[1] ≤ 3\n  [2, 2]  =  c[2,2] : x[2] ≤ 2\n  [2, 3]  =  c[2,3] : x[2] ≤ 3\n  [3, 3]  =  c[3,3] : x[3] ≤ 3","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"and you can filter elements in the sets using the ; syntax:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:9]);\n\njulia> @constraint(model, c[i=1:9; mod(i, 3) == 0], x[i] <= i)\nJuMP.Containers.SparseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, 1, Tuple{Int64}} with 3 entries:\n  [3]  =  c[3] : x[3] ≤ 3\n  [6]  =  c[6] : x[6] ≤ 6\n  [9]  =  c[9] : x[9] ≤ 9","category":"page"},{"location":"manual/constraints/#Registered-constraints","page":"Constraints","title":"Registered constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"When you create constraints, JuMP registers them inside the model using their corresponding symbol. Get a registered name using model[:key]:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model()\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, my_c, 2x <= 1)\nmy_c : 2 x ≤ 1\n\njulia> model\nA JuMP Model\nFeasibility problem with:\nVariable: 1\n`AffExpr`-in-`MathOptInterface.LessThan{Float64}`: 1 constraint\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\nNames registered in the model: my_c, x\n\njulia> model[:my_c] === my_c\ntrue","category":"page"},{"location":"manual/constraints/#Anonymous-constraints","page":"Constraints","title":"Anonymous constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To reduce the likelihood of accidental bugs, and because JuMP registers constraints inside a model, creating two constraints with the same name is an error:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, c, 2x <= 1)\nc : 2 x ≤ 1\n\njulia> @constraint(model, c, 2x <= 1)\nERROR: An object of name c is already attached to this model. If this\n    is intended, consider using the anonymous construction syntax, e.g.,\n    `x = @variable(model, [1:N], ...)` where the name of the object does\n    not appear inside the macro.\n\n    Alternatively, use `unregister(model, :c)` to first unregister\n    the existing name from the model. Note that this will not delete the\n    object; it will just remove the reference at `model[:c]`.\n[...]","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"A common reason for encountering this error is adding constraints in a loop.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"As a work-around, JuMP provides anonymous constraints. Create an anonymous constraint by omitting the name argument:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> c = @constraint(model, 2x <= 1)\n2 x ≤ 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Create a container of anonymous constraints by dropping the name in front of the [:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3]);\n\njulia> c = @constraint(model, [i = 1:3], x[i] <= i)\n3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n x[1] ≤ 1\n x[2] ≤ 2\n x[3] ≤ 3","category":"page"},{"location":"manual/constraints/#Constraint-names","page":"Constraints","title":"Constraint names","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In addition to the symbol that constraints are registered with, constraints have a String name that is used for printing and writing to file formats.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Get and set the name of a constraint using name(::JuMP.ConstraintRef) and set_name(::JuMP.ConstraintRef, ::String):","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model(); @variable(model, x);\n\njulia> @constraint(model, con, x <= 1)\ncon : x ≤ 1\n\njulia> name(con)\n\"con\"\n\njulia> set_name(con, \"my_con_name\")\n\njulia> con\nmy_con_name : x ≤ 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Override the default choice of name using the base_name keyword:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model(); @variable(model, x);\n\njulia> con = @constraint(model, [i=1:2], x <= i, base_name = \"my_con\")\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n my_con[1] : x ≤ 1\n my_con[2] : x ≤ 2","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Note that names apply to each element of the container, not to the container of constraints:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> name(con[1])\n\"my_con[1]\"\n\njulia> set_name(con[1], \"c\")\n\njulia> con\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n c : x ≤ 1\n my_con[2] : x ≤ 2","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"tip: Tip\nFor some models, setting the string name of each constraint can take a non-trivial portion of the total time required to build the model. Turn off String names by passing set_string_name = false to @constraint:julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con, x <= 2, set_string_name = false)\nx ≤ 2See Disable string names for more information.","category":"page"},{"location":"manual/constraints/#Retrieve-a-constraint-by-name","page":"Constraints","title":"Retrieve a constraint by name","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Retrieve a constraint from a model using constraint_by_name:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> constraint_by_name(model, \"c\")\nc : x ≤ 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"If the name is not present, nothing will be returned:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> constraint_by_name(model, \"bad_name\")","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"You can only look up individual constraints using constraint_by_name. Something like this will not work:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model(); @variable(model, x);\n\njulia> con = @constraint(model, [i=1:2], x <= i, base_name = \"my_con\")\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n my_con[1] : x ≤ 1\n my_con[2] : x ≤ 2\n\njulia> constraint_by_name(model, \"my_con\")","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To look up a collection of constraints, do not use constraint_by_name. Instead, register them using the model[:key] = value syntax:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model(); @variable(model, x);\n\njulia> model[:con] = @constraint(model, [i=1:2], x <= i, base_name = \"my_con\")\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n my_con[1] : x ≤ 1\n my_con[2] : x ≤ 2\n\njulia> model[:con]\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n my_con[1] : x ≤ 1\n my_con[2] : x ≤ 2","category":"page"},{"location":"manual/constraints/#String-names,-symbolic-names,-and-bindings","page":"Constraints","title":"String names, symbolic names, and bindings","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"It's common for new users to experience confusion relating to constraints. Part of the problem is the difference between the name that a constraint is registered under and the String name used for printing.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Here's a summary of the differences:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Constraints are created using @constraint.\nConstraints can be named or anonymous.\nNamed constraints have the form @constraint(model, c, expr). For named constraints:\nThe String name of the constraint is set to \"c\".\nA Julia variable c is created that binds c to  the JuMP constraint.\nThe name :c is registered as a key in the model with the value c.\nAnonymous constraints have the form c = @constraint(model, expr). For anonymous constraints:\nThe String name of the constraint is set to \"\".\nYou control the name of the Julia variable used as the binding.\nNo name is registered as a key in the model.\nThe base_name keyword can override the String name of the constraint.\nYou can manually register names in the model via model[:key] = value.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Here's an example of the differences:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> c_binding = @constraint(model, 2x <= 1, base_name = \"c\")\nc : 2 x ≤ 1\n\njulia> model\nA JuMP Model\nFeasibility problem with:\nVariable: 1\n`AffExpr`-in-`MathOptInterface.LessThan{Float64}`: 1 constraint\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\nNames registered in the model: x\n\njulia> c\nERROR: UndefVarError: `c` not defined\n\njulia> c_binding\nc : 2 x ≤ 1\n\njulia> name(c_binding)\n\"c\"\n\njulia> model[:c_register] = c_binding\nc : 2 x ≤ 1\n\njulia> model\nA JuMP Model\nFeasibility problem with:\nVariable: 1\n`AffExpr`-in-`MathOptInterface.LessThan{Float64}`: 1 constraint\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\nNames registered in the model: c_register, x\n\njulia> model[:c_register]\nc : 2 x ≤ 1\n\njulia> model[:c_register] === c_binding\ntrue\n\njulia> c\nERROR: UndefVarError: `c` not defined","category":"page"},{"location":"manual/constraints/#The-@constraints-macro","page":"Constraints","title":"The @constraints macro","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"If you have many @constraint calls, use the @constraints macro to improve readability:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraints(model, begin\n           2x <= 1\n           c, x >= -1\n       end)\n(2 x ≤ 1, c : x ≥ -1)\n\njulia> print(model)\nFeasibility\nSubject to\n c : x ≥ -1\n 2 x ≤ 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The @constraints macro returns a tuple of the constraints that were defined.","category":"page"},{"location":"manual/constraints/#constraint_duality","page":"Constraints","title":"Duality","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"JuMP adopts the notion of conic duality from MathOptInterface. For linear programs, a feasible dual on a >= constraint is nonnegative and a feasible dual on a <= constraint is nonpositive. If the constraint is an equality constraint, it depends on which direction is binding.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"warning: Warning\nJuMP's definition of duality is independent of the objective sense. That is, the sign of feasible duals associated with a constraint depends on the direction of the constraint and not whether the problem is maximization or minimization. This is a different convention from linear programming duality in some common textbooks. If you have a linear program, and you want the textbook definition, you probably want to use shadow_price and reduced_cost instead.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The dual value associated with a constraint in the most recent solution can be accessed using the dual function. Use has_duals to check if the model has a dual solution available to query. For example:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, con, x <= 1)\ncon : x ≤ 1\n\njulia> @objective(model, Min, -2x)\n-2 x\n\njulia> has_duals(model)\nfalse\n\njulia> optimize!(model)\n\njulia> has_duals(model)\ntrue\n\njulia> dual(con)\n-2.0\n\njulia> @objective(model, Max, 2x)\n2 x\n\njulia> optimize!(model)\n\njulia> dual(con)\n-2.0","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To help users who may be less familiar with conic duality, JuMP provides shadow_price, which returns a value that can be interpreted as the improvement in the objective in response to an infinitesimal relaxation (on the scale of one unit) in the right-hand side of the constraint. shadow_price can be used only on linear constraints with a <=, >=, or == comparison operator.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In the example above, dual(con) returned -2.0 regardless of the optimization sense. However, in the second case when the optimization sense is Max, shadow_price returns:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> shadow_price(con)\n2.0","category":"page"},{"location":"manual/constraints/#Duals-of-variable-bounds","page":"Constraints","title":"Duals of variable bounds","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To query the dual variables associated with a variable bound, first obtain a constraint reference using one of UpperBoundRef, LowerBoundRef, or FixRef, and then call dual on the returned constraint reference. The reduced_cost function may simplify this process as it returns the shadow price of an active bound of a variable (or zero, if no active bound exists).","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x <= 1)\nx\n\njulia> @objective(model, Min, -2x)\n-2 x\n\njulia> optimize!(model)\n\njulia> dual(UpperBoundRef(x))\n-2.0\n\njulia> reduced_cost(x)\n-2.0","category":"page"},{"location":"manual/constraints/#Modify-a-constant-term","page":"Constraints","title":"Modify a constant term","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"This section explains how to modify the constant term in a constraint. There are multiple ways to achieve this goal; we explain three options.","category":"page"},{"location":"manual/constraints/#Option-1:-change-the-right-hand-side","page":"Constraints","title":"Option 1: change the right-hand side","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Use set_normalized_rhs to modify the right-hand side (constant) term of a linear or quadratic  constraint. Use normalized_rhs to query the right-hand side term.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con, 2x <= 1)\ncon : 2 x ≤ 1\n\njulia> set_normalized_rhs(con, 3)\n\njulia> con\ncon : 2 x ≤ 3\n\njulia> normalized_rhs(con)\n3.0","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"warning: Warning\nset_normalized_rhs sets the right-hand side term of the normalized constraint. See Normalization for more details.","category":"page"},{"location":"manual/constraints/#Option-2:-use-fixed-variables","page":"Constraints","title":"Option 2: use fixed variables","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"If constraints are complicated, for example, they are composed of a number of components, each of which has a constant term, then it may be difficult to calculate what the right-hand side term is in the standard form.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"For this situation, JuMP includes the ability to fix variables to a value using the fix function. Fixing a variable sets its lower and upper bound to the same value. Thus, changes in a constant term can be simulated by adding a new variable and fixing it to different values. Here is an example:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, const_term)\nconst_term\n\njulia> @constraint(model, con, 2x <= const_term + 1)\ncon : 2 x - const_term ≤ 1\n\njulia> fix(const_term, 1.0)","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The constraint con is now equivalent to 2x <= 2.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"warning: Warning\nFixed variables are not replaced with constants when communicating the problem to a solver. Therefore, even though const_term is fixed, it is still a decision variable, and so const_term * x is bilinear.","category":"page"},{"location":"manual/constraints/#Option-3:-modify-the-function's-constant-term","page":"Constraints","title":"Option 3: modify the function's constant term","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The third option is to use add_to_function_constant. The constant given is added to the function of a func-in-set constraint. In the following example, adding 2 to the function has the effect of removing 2 to the right-hand side:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con, 2x <= 1)\ncon : 2 x ≤ 1\n\njulia> add_to_function_constant(con, 2)\n\njulia> con\ncon : 2 x ≤ -1\n\njulia> normalized_rhs(con)\n-1.0","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In the case of interval constraints, the constant is removed from each bound:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con, 0 <= 2x + 1 <= 2)\ncon : 2 x ∈ [-1, 1]\n\njulia> add_to_function_constant(con, 3)\n\njulia> con\ncon : 2 x ∈ [-4, -2]","category":"page"},{"location":"manual/constraints/#Modify-a-variable-coefficient","page":"Constraints","title":"Modify a variable coefficient","text":"","category":"section"},{"location":"manual/constraints/#Scalar-constraints","page":"Constraints","title":"Scalar constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To modify the coefficients for a linear term (modifying the coefficient of a quadratic term is not supported) in a constraint, use set_normalized_coefficient. To query the current coefficient, use normalized_coefficient.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @constraint(model, con, 2x[1] + x[2] <= 1)\ncon : 2 x[1] + x[2] ≤ 1\n\njulia> set_normalized_coefficient(con, x[2], 0)\n\njulia> con\ncon : 2 x[1] ≤ 1\n\njulia> normalized_coefficient(con, x[2])\n0.0","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"warning: Warning\nset_normalized_coefficient sets the coefficient of the normalized constraint. See Normalization for more details.","category":"page"},{"location":"manual/constraints/#Vector-constraints","page":"Constraints","title":"Vector constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To modify the coefficients of a vector-valued constraint, use set_normalized_coefficients.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, con, [2x + 3x, 4x] in MOI.Nonnegatives(2))\ncon : [5 x, 4 x] ∈ MathOptInterface.Nonnegatives(2)\n\njulia> set_normalized_coefficients(con, x, [(1, 3.0)])\n\njulia> con\ncon : [3 x, 4 x] ∈ MathOptInterface.Nonnegatives(2)\n\njulia> set_normalized_coefficients(con, x, [(1, 2.0), (2, 5.0)])\n\njulia> con\ncon : [2 x, 5 x] ∈ MathOptInterface.Nonnegatives(2)","category":"page"},{"location":"manual/constraints/#Delete-a-constraint","page":"Constraints","title":"Delete a constraint","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Use delete to delete a constraint from a model. Use is_valid to check if a constraint belongs to a model and has not been deleted.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con, 2x <= 1)\ncon : 2 x ≤ 1\n\njulia> is_valid(model, con)\ntrue\n\njulia> delete(model, con)\n\njulia> is_valid(model, con)\nfalse","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Deleting a constraint does not unregister the symbolic reference from the model. Therefore, creating a new constraint of the same name will throw an error:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, con, 2x <= 1)\nERROR: An object of name con is already attached to this model. If this\n    is intended, consider using the anonymous construction syntax, e.g.,\n    `x = @variable(model, [1:N], ...)` where the name of the object does\n    not appear inside the macro.\n\n    Alternatively, use `unregister(model, :con)` to first unregister\n    the existing name from the model. Note that this will not delete the\n    object; it will just remove the reference at `model[:con]`.\n[...]","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"After calling delete, call unregister to remove the symbolic reference:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> unregister(model, :con)\n\njulia> @constraint(model, con, 2x <= 1)\ncon : 2 x ≤ 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"info: Info\ndelete does not automatically unregister because we do not distinguish between names that are automatically registered by JuMP macros, and names that are manually registered by the user by setting values in object_dictionary. In addition, deleting a constraint and then adding a new constraint of the same name is an easy way to introduce bugs into your code.","category":"page"},{"location":"manual/constraints/#Start-values","page":"Constraints","title":"Start values","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Provide a starting value (also called warmstart) for a constraint's primal and dual solutions using set_start_value and set_dual_start_value.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Query the starting value for a constraint's primal and dual solution using start_value and dual_start_value. If no start value has been set, the methods will return nothing.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, con, x >= 10)\ncon : x ≥ 10\n\njulia> start_value(con)\n\njulia> set_start_value(con, 10.0)\n\njulia> start_value(con)\n10.0\n\njulia> dual_start_value(con)\n\njulia> set_dual_start_value(con, 2)\n\njulia> dual_start_value(con)\n2.0","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Vector-valued constraints require a vector:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3])\n3-element Vector{VariableRef}:\n x[1]\n x[2]\n x[3]\n\njulia> @constraint(model, con, x in SecondOrderCone())\ncon : [x[1], x[2], x[3]] in MathOptInterface.SecondOrderCone(3)\n\njulia> dual_start_value(con)\n\njulia> set_dual_start_value(con, [1.0, 2.0, 3.0])\n\njulia> dual_start_value(con)\n3-element Vector{Float64}:\n 1.0\n 2.0\n 3.0","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"tip: Tip\nTo simplify setting start values for all variables and constraints in a model, see set_start_values. The Primal and dual warm-starts tutorial also gives a detailed description of how to iterate over constraints in the model to set custom start values.","category":"page"},{"location":"manual/constraints/#Constraint-containers","page":"Constraints","title":"Constraint containers","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Like Variable containers, JuMP provides a mechanism for building groups of constraints compactly. References to these groups of constraints are returned in containers. Three types of constraint containers are supported: Arrays, DenseAxisArrays, and SparseAxisArrays. We explain each of these in the following.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"tip: Tip\nYou can read more about containers in the Containers section.","category":"page"},{"location":"manual/constraints/#constraint_arrays","page":"Constraints","title":"Arrays","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"One way of adding a group of constraints compactly is the following:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con[i = 1:3], i * x <= i + 1)\n3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n con[1] : x ≤ 2\n con[2] : 2 x ≤ 3\n con[3] : 3 x ≤ 4","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"JuMP returns references to the three constraints in an Array that is bound to the Julia variable con. This array can be accessed and sliced as you would with any Julia array:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> con[1]\ncon[1] : x ≤ 2\n\njulia> con[2:3]\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n con[2] : 2 x ≤ 3\n con[3] : 3 x ≤ 4","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Anonymous containers can also be constructed by dropping the name (for example, con) before the square brackets:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> con = @constraint(model, [i = 1:2], i * x <= i + 1)\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n x ≤ 2\n 2 x ≤ 3","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Just like @variable, JuMP will form an Array of constraints when it can determine at parse time that the indices are one-based integer ranges. Therefore con[1:b] will create an Array, but con[a:b] will not. A special case is con[Base.OneTo(n)] which will produce an Array. If JuMP cannot determine that the indices are one-based integer ranges (for example, in the case of con[a:b]), JuMP will create a DenseAxisArray instead.","category":"page"},{"location":"manual/constraints/#DenseAxisArrays","page":"Constraints","title":"DenseAxisArrays","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The syntax for constructing a DenseAxisArray of constraints is very similar to the syntax for constructing a DenseAxisArray of variables.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con[i = 1:2, j = 2:3], i * x <= j + 1)\n2-dimensional DenseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},2,...} with index sets:\n    Dimension 1, Base.OneTo(2)\n    Dimension 2, 2:3\nAnd data, a 2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n con[1,2] : x ≤ 3    con[1,3] : x ≤ 4\n con[2,2] : 2 x ≤ 3  con[2,3] : 2 x ≤ 4","category":"page"},{"location":"manual/constraints/#SparseAxisArrays","page":"Constraints","title":"SparseAxisArrays","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The syntax for constructing a SparseAxisArray of constraints is very similar to the syntax for constructing a SparseAxisArray of variables.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con[i = 1:2, j = 1:2; i != j], i * x <= j + 1)\nJuMP.Containers.SparseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, 2, Tuple{Int64, Int64}} with 2 entries:\n  [1, 2]  =  con[1,2] : x ≤ 3\n  [2, 1]  =  con[2,1] : 2 x ≤ 2","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"warning: Warning\nIf you have many index dimensions and a large amount of sparsity, read Performance considerations.","category":"page"},{"location":"manual/constraints/#Forcing-the-container-type","page":"Constraints","title":"Forcing the container type","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"When creating a container of constraints, JuMP will attempt to choose the tightest container type that can store the constraints. However, because this happens at parse time, it does not always make the best choice. Just like in @variable, you can force the type of container using the container keyword. For syntax and the reason behind this, take a look at the variable docs.","category":"page"},{"location":"manual/constraints/#Constraints-with-similar-indices","page":"Constraints","title":"Constraints with similar indices","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Containers are often used to create constraints over a set of indices. However, you'll often have cases in which you are repeating the indices:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @variable(model, y[1:2]);\n\njulia> @constraints(model, begin\n           [i=1:2, j=1:2, k=1:2], i * x[j] <= k\n           [i=1:2, j=1:2, k=1:2], i * y[j] <= k\n       end);","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"This is hard to read and leads to a lot of copy-paste. A more readable way is to use a for-loop:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> for i=1:2, j=1:2, k=1:2\n           @constraints(model, begin\n               i * x[j] <= k\n               i * y[j] <= k\n           end)\n       end","category":"page"},{"location":"manual/constraints/#Accessing-constraints-from-a-model","page":"Constraints","title":"Accessing constraints from a model","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Query the types of function-in-set constraints in a model using list_of_constraint_types:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:2] >= i, Int);\n\njulia> @constraint(model, x[1] + x[2] <= 1);\n\njulia> list_of_constraint_types(model)\n3-element Vector{Tuple{Type, Type}}:\n (AffExpr, MathOptInterface.LessThan{Float64})\n (VariableRef, MathOptInterface.GreaterThan{Float64})\n (VariableRef, MathOptInterface.Integer)","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"For a given combination of function and set type, use num_constraints to access the number of constraints and all_constraints to access a list of their references:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> num_constraints(model, VariableRef, MOI.Integer)\n2\n\njulia> cons = all_constraints(model, VariableRef, MOI.Integer)\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}, ScalarShape}}:\n x[1] integer\n x[2] integer","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"You can also count the total number of constraints in the model, but you must explicitly choose whether to count VariableRef constraints such as bound and integrality constraints:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> num_constraints(model; count_variable_in_set_constraints = true)\n5\n\njulia> num_constraints(model; count_variable_in_set_constraints = false)\n1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The same also applies for all_constraints:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> all_constraints(model; include_variable_in_set_constraints = true)\n5-element Vector{ConstraintRef}:\n x[1] + x[2] ≤ 1\n x[1] ≥ 1\n x[2] ≥ 2\n x[1] integer\n x[2] integer\n\njulia> all_constraints(model; include_variable_in_set_constraints = false)\n1-element Vector{ConstraintRef}:\n x[1] + x[2] ≤ 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"If you need finer-grained control on which constraints to include, use a variant of:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> sum(\n           num_constraints(model, F, S) for\n           (F, S) in list_of_constraint_types(model) if F != VariableRef\n       )\n1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Use constraint_object to get an instance of an AbstractConstraint object that stores the constraint data:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> con = constraint_object(cons[1])\nScalarConstraint{VariableRef, MathOptInterface.Integer}(x[1], MathOptInterface.Integer())\n\njulia> con.func\nx[1]\n\njulia> con.set\nMathOptInterface.Integer()","category":"page"},{"location":"manual/constraints/#MathOptInterface-constraints","page":"Constraints","title":"MathOptInterface constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Because JuMP is based on MathOptInterface, you can add any constraints supported by MathOptInterface using the function-in-set syntax. For a list of supported functions and sets, read Standard form problem.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"note: Note\nWe use MOI as an alias for the MathOptInterface module. This alias is defined by using JuMP. You may also define it in your code as follows:import MathOptInterface as MOI","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"For example, the following two constraints are equivalent:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3]);\n\njulia> @constraint(model, 2 * x[1] <= 1)\n2 x[1] ≤ 1\n\njulia> @constraint(model, 2 * x[1] in MOI.LessThan(1.0))\n2 x[1] ≤ 1","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"You can also use any set defined by MathOptInterface:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, x - [1; 2; 3] in MOI.Nonnegatives(3))\n[x[1] - 1, x[2] - 2, x[3] - 3] ∈ MathOptInterface.Nonnegatives(3)\n\njulia> @constraint(model, x in MOI.ExponentialCone())\n[x[1], x[2], x[3]] ∈ MathOptInterface.ExponentialCone()","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"info: Info\nSimilar to how JuMP defines the <= and >= syntax as a convenience way to specify MOI.LessThan and MOI.GreaterThan constraints, the remaining sections in this page describe functions and syntax that have been added for the convenience of common modeling situations.","category":"page"},{"location":"manual/constraints/#Set-inequality-syntax","page":"Constraints","title":"Set inequality syntax","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"For modeling convenience, the syntax @constraint(model, x >= y, Set()) is short-hand for @constraint(model, x - y in Set()). Therefore, the following calls are equivalent:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> y = [0.5, 0.75];\n\njulia> @constraint(model, x >= y, MOI.Nonnegatives(2))\n[x[1] - 0.5, x[2] - 0.75] ∈ MathOptInterface.Nonnegatives(2)\n\njulia> @constraint(model, y <= x, MOI.Nonnegatives(2))\n[x[1] - 0.5, x[2] - 0.75] ∈ MathOptInterface.Nonnegatives(2)\n\njulia> @constraint(model, x - y in MOI.Nonnegatives(2))\n[x[1] - 0.5, x[2] - 0.75] ∈ MathOptInterface.Nonnegatives(2)","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Non-zero constants are not supported in this syntax:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, x >= 1, MOI.Nonnegatives(2))\nERROR: Operation `sub_mul` between `Vector{VariableRef}` and `Int64` is not allowed. This most often happens when you write a constraint like `x >= y` where `x` is an array and `y` is a constant. Use the broadcast syntax `x .- y >= 0` instead.\nStacktrace:\n[...]","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Use instead:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, x .- 1 >= 0, MOI.Nonnegatives(2))\n[x[1] - 1, x[2] - 1] ∈ MathOptInterface.Nonnegatives(2)","category":"page"},{"location":"manual/constraints/#Second-order-cone-constraints","page":"Constraints","title":"Second-order cone constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"A SecondOrderCone constrains the variables t and x to the set:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"x_2 le t","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"and t ge 0. It can be added as follows:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, t)\nt\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> @constraint(model, [t; x] in SecondOrderCone())\n[t, x[1], x[2]] ∈ MathOptInterface.SecondOrderCone(3)","category":"page"},{"location":"manual/constraints/#Rotated-second-order-cone-constraints","page":"Constraints","title":"Rotated second-order cone constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"A RotatedSecondOrderCone constrains the variables t, u, and x to the set:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"x_2^2 le 2 t cdot u","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"and t u ge 0. It can be added as follows:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, t)\nt\n\njulia> @variable(model, u)\nu\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> @constraint(model, [t; u; x] in RotatedSecondOrderCone())\n[t, u, x[1], x[2]] ∈ MathOptInterface.RotatedSecondOrderCone(4)","category":"page"},{"location":"manual/constraints/#Semi-integer-and-semi-continuous-variables","page":"Constraints","title":"Semi-integer and semi-continuous variables","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Semi-continuous variables are constrained to the set x in 0 cup l u.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Create a semi-continuous variable using the Semicontinuous set:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, x in Semicontinuous(1.5, 3.5))\nx in MathOptInterface.Semicontinuous{Float64}(1.5, 3.5)","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Semi-integer variables  are constrained to the set x in 0 cup l l+1 dots u.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Create a semi-integer variable using the Semiinteger set:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, x in Semiinteger(1.0, 3.0))\nx in MathOptInterface.Semiinteger{Float64}(1.0, 3.0)","category":"page"},{"location":"manual/constraints/#Special-Ordered-Sets-of-Type-1","page":"Constraints","title":"Special Ordered Sets of Type 1","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In a Special Ordered Set of Type 1 (often denoted SOS-I or SOS1), at most one element can take a non-zero value.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Construct SOS-I constraints using the SOS1 set:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3])\n3-element Vector{VariableRef}:\n x[1]\n x[2]\n x[3]\n\njulia> @constraint(model, x in SOS1())\n[x[1], x[2], x[3]] in MathOptInterface.SOS1{Float64}([1.0, 2.0, 3.0])","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Although not required for feasibility, solvers can benefit from an ordering of the variables (for example, the variables represent different factories to build, at most one factory can be built, and the factories can be ordered according to cost). To induce an ordering, a vector of weights can be provided, and the variables are ordered according to their corresponding weight.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"For example, in the constraint:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, x in SOS1([3.1, 1.2, 2.3]))\n[x[1], x[2], x[3]] in MathOptInterface.SOS1{Float64}([3.1, 1.2, 2.3])","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"the variables x have precedence x[2], x[3], x[1].","category":"page"},{"location":"manual/constraints/#Special-Ordered-Sets-of-Type-2","page":"Constraints","title":"Special Ordered Sets of Type 2","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In a Special Ordered Set of Type 2 (SOS-II), at most two elements can be non-zero, and if there are two non-zeros, they must be consecutive according to the ordering induced by a weight vector.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Construct SOS-II constraints using the SOS2 set:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, x in SOS2([3.0, 1.0, 2.0]))\n[x[1], x[2], x[3]] in MathOptInterface.SOS2{Float64}([3.0, 1.0, 2.0])","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The possible non-zero pairs are (x[1], x[3]) and (x[2], x[3]):","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"If the weight vector is omitted, JuMP induces an ordering from 1:length(x):","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, x in SOS2())\n[x[1], x[2], x[3]] in MathOptInterface.SOS2{Float64}([1.0, 2.0, 3.0])","category":"page"},{"location":"manual/constraints/#Indicator-constraints","page":"Constraints","title":"Indicator constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Indicator constraints consist of a binary variable and a linear constraint. The constraint holds when the binary variable takes the value 1. The constraint may or may not hold when the binary variable takes the value 0.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To enforce the constraint x + y <= 1 when the binary variable a is 1, use:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> @variable(model, a, Bin)\na\n\njulia> @constraint(model, a => {x + y <= 1})\na => {x + y ≤ 1}","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"If the constraint must hold when a is zero, add ! or ¬ before the binary variable;","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, !a => {x + y <= 1})\n!a => {x + y ≤ 1}","category":"page"},{"location":"manual/constraints/#Semidefinite-constraints","page":"Constraints","title":"Semidefinite constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To constrain a matrix to be positive semidefinite (PSD), use PSDCone:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, X[1:2, 1:2])\n2×2 Matrix{VariableRef}:\n X[1,1]  X[1,2]\n X[2,1]  X[2,2]\n\njulia> @constraint(model, X >= 0, PSDCone())\n[X[1,1]  X[1,2];\n X[2,1]  X[2,2]] ∈ PSDCone()","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"tip: Tip\nWhere possible, prefer constructing a matrix of Semidefinite variables using the @variable macro, rather than adding a constraint like @constraint(model, X >= 0, PSDCone()). In some solvers, adding the constraint via @constraint is less efficient, and can result in additional intermediate variables and constraints being added to the model.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The inequality X >= Y between two square matrices X and Y is understood as constraining X - Y to be positive semidefinite.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> Y = [1 2; 2 1]\n2×2 Matrix{Int64}:\n 1  2\n 2  1\n\njulia> @constraint(model, X >= Y, PSDCone())\n[X[1,1] - 1  X[1,2] - 2;\n X[2,1] - 2  X[2,2] - 1] ∈ PSDCone()","category":"page"},{"location":"manual/constraints/#Symmetry","page":"Constraints","title":"Symmetry","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Solvers supporting PSD constraints usually expect to be given a matrix that is symbolically symmetric, that is, for which the expression in corresponding off-diagonal entries are the same. In our example, the expressions of entries (1, 2) and (2, 1) are respectively X[1,2] - 2 and X[2,1] - 2 which are different.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"To bridge the gap between the constraint modeled and what the solver expects, solvers may add an equality constraint X[1,2] - 2 == X[2,1] - 2 to force symmetry. Use LinearAlgebra.Symmetric to explicitly tell the solver that the matrix is symmetric:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> import LinearAlgebra\n\njulia> Z = [X[1, 1] X[1, 2]; X[1, 2] X[2, 2]]\n2×2 Matrix{VariableRef}:\n X[1,1]  X[1,2]\n X[1,2]  X[2,2]\n\njulia> @constraint(model, LinearAlgebra.Symmetric(Z) >= 0, PSDCone())\n[X[1,1]  X[1,2];\n X[1,2]  X[2,2]] ∈ PSDCone()","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Note that the lower triangular entries are ignored even if they are different so use it with caution:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, LinearAlgebra.Symmetric(X) >= 0, PSDCone())\n[X[1,1]  X[1,2];\n X[1,2]  X[2,2]] ∈ PSDCone()","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"(Note the (2, 1) element of the constraint is X[1,2], not X[2,1].)","category":"page"},{"location":"manual/constraints/#Complementarity-constraints","page":"Constraints","title":"Complementarity constraints","text":"","category":"section"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"A mixed complementarity constraint F(x) ⟂ x consists of finding x in the interval [lb, ub], such that the following holds:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"F(x) == 0 if lb < x < ub\nF(x) >= 0 if lb == x\nF(x) <= 0 if x == ub","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"JuMP supports mixed complementarity constraints via complements(F(x), x) or F(x) ⟂ x in the @constraint macro. The interval set [lb, ub] is obtained from the variable bounds on x.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"For example, to define the problem 2x - 1 ⟂ x with x ∈ [0, ∞), do:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> model = Model();\n\njulia> @variable(model, x >= 0)\nx\n\njulia> @constraint(model, 2x - 1 ⟂ x)\n[2 x - 1, x] ∈ MathOptInterface.Complements(2)","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"This problem has a unique solution at x = 0.5.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"The perp operator ⟂ can be entered in most editors (and the Julia REPL) by typing \\perp<tab>.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"An alternative approach that does not require the ⟂ symbol uses the complements function as follows:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @constraint(model, complements(2x - 1, x))\n[2 x - 1, x] ∈ MathOptInterface.Complements(2)","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"In both cases, the mapping F(x) is supplied as the first argument, and the matching variable x is supplied as the second.","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"Vector-valued complementarity constraints are also supported:","category":"page"},{"location":"manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> @variable(model, -2 <= y[1:2] <= 2)\n2-element Vector{VariableRef}:\n y[1]\n y[2]\n\njulia> M = [1 2; 3 4]\n2×2 Matrix{Int64}:\n 1  2\n 3  4\n\njulia> q = [5, 6]\n2-element Vector{Int64}:\n 5\n 6\n\njulia> @constraint(model, M * y + q ⟂ y)\n[y[1] + 2 y[2] + 5, 3 y[1] + 4 y[2] + 6, y[1], y[2]] ∈ MathOptInterface.Complements(4)","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"EditURL = \"power_systems.jl\"","category":"page"},{"location":"tutorials/applications/power_systems/#Power-Systems","page":"Power Systems","title":"Power Systems","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"This tutorial was originally contributed by Yury Dvorkin and Miles Lubin.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"This tutorial demonstrates how to formulate basic power systems engineering models in JuMP.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"We will consider basic \"economic dispatch\" and \"unit commitment\" models without taking into account transmission constraints.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"For this tutorial, we use the following packages:","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"using JuMP\nimport DataFrames\nimport HiGHS\nimport Plots\nimport StatsPlots","category":"page"},{"location":"tutorials/applications/power_systems/#Economic-dispatch","page":"Power Systems","title":"Economic dispatch","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Economic dispatch (ED) is an optimization problem that minimizes the cost of supplying energy demand subject to operational constraints on power system assets. In its simplest modification, ED is an LP problem solved for an aggregated load and wind forecast and for a single infinitesimal moment.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Mathematically, the ED problem can be written as follows:","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"min sum_i in I c^g_i cdot g_i + c^w cdot w","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"where c_i and g_i are the incremental cost ($/MWh) and power output (MW) of the i^th generator, respectively, and c^w and w are the incremental cost ($/MWh) and wind power injection (MW), respectively.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Subject to the constraints:","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Minimum (g^min) and maximum (g^max) limits on power outputs of generators: g^min_i leq g_i leq g^max_i\nConstraint on the wind power injection: 0 leq w leq w^f where w and w^f are the wind power injection and wind power forecast, respectively.\nPower balance constraint: sum_i in I g_i + w = d^f where d^f is the demand forecast.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Further reading on ED models can be found in A. J. Wood, B. F. Wollenberg, and G. B. Sheblé, \"Power Generation, Operation and Control,\" Wiley, 2013.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Define some input data about the test system.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"We define some thermal generators:","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"function ThermalGenerator(\n    min::Float64,\n    max::Float64,\n    fixed_cost::Float64,\n    variable_cost::Float64,\n)\n    return (\n        min = min,\n        max = max,\n        fixed_cost = fixed_cost,\n        variable_cost = variable_cost,\n    )\nend\n\ngenerators = [\n    ThermalGenerator(0.0, 1000.0, 1000.0, 50.0),\n    ThermalGenerator(300.0, 1000.0, 0.0, 100.0),\n]","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"A wind generator","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"WindGenerator(variable_cost::Float64) = (variable_cost = variable_cost,)\n\nwind_generator = WindGenerator(50.0)","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"And a scenario","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"function Scenario(demand::Float64, wind::Float64)\n    return (demand = demand, wind = wind)\nend\n\nscenario = Scenario(1500.0, 200.0)","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Create a function solve_economic_dispatch, which solves the economic dispatch problem for a given set of input parameters.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"function solve_economic_dispatch(generators::Vector, wind, scenario)\n    # Define the economic dispatch (ED) model\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    # Define decision variables\n    # power output of generators\n    N = length(generators)\n    @variable(model, generators[i].min <= g[i = 1:N] <= generators[i].max)\n    # wind power injection\n    @variable(model, 0 <= w <= scenario.wind)\n    # Define the objective function\n    @objective(\n        model,\n        Min,\n        sum(generators[i].variable_cost * g[i] for i in 1:N) +\n        wind.variable_cost * w,\n    )\n    # Define the power balance constraint\n    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)\n    # Solve statement\n    optimize!(model)\n    # return the optimal value of the objective function and its minimizers\n    return (\n        g = value.(g),\n        w = value(w),\n        wind_spill = scenario.wind - value(w),\n        total_cost = objective_value(model),\n    )\nend","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Solve the economic dispatch problem","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"solution = solve_economic_dispatch(generators, wind_generator, scenario);\n\nprintln(\"Dispatch of Generators: \", solution.g, \" MW\")\nprintln(\"Dispatch of Wind: \", solution.w, \" MW\")\nprintln(\"Wind spillage: \", solution.wind_spill, \" MW\")\nprintln(\"Total cost: \\$\", solution.total_cost)","category":"page"},{"location":"tutorials/applications/power_systems/#Economic-dispatch-with-adjustable-incremental-costs","page":"Power Systems","title":"Economic dispatch with adjustable incremental costs","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"In the following exercise we adjust the incremental cost of generator G1 and observe its impact on the total cost.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"function scale_generator_cost(g, scale)\n    return ThermalGenerator(g.min, g.max, g.fixed_cost, scale * g.variable_cost)\nend\n\nstart = time()\nc_g_scale_df = DataFrames.DataFrame(;\n    # Scale factor\n    scale = Float64[],\n    # Dispatch of Generator 1 [MW]\n    dispatch_G1 = Float64[],\n    # Dispatch of Generator 2 [MW]\n    dispatch_G2 = Float64[],\n    # Dispatch of Wind [MW]\n    dispatch_wind = Float64[],\n    # Spillage of Wind [MW]\n    spillage_wind = Float64[],\n    # Total cost [$]\n    total_cost = Float64[],\n)\nfor c_g1_scale in 0.5:0.1:3.0\n    # Update the incremental cost of the first generator at every iteration.\n    new_generators = scale_generator_cost.(generators, [c_g1_scale, 1.0])\n    # Solve the economic-dispatch problem with the updated incremental cost\n    sol = solve_economic_dispatch(new_generators, wind_generator, scenario)\n    push!(\n        c_g_scale_df,\n        (c_g1_scale, sol.g[1], sol.g[2], sol.w, sol.wind_spill, sol.total_cost),\n    )\nend\nprint(string(\"elapsed time: \", time() - start, \" seconds\"))","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"c_g_scale_df","category":"page"},{"location":"tutorials/applications/power_systems/#Modifying-the-JuMP-model-in-place","page":"Power Systems","title":"Modifying the JuMP model in-place","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Note that in the previous exercise we entirely rebuilt the optimization model at every iteration of the internal loop, which incurs an additional computational burden. This burden can be alleviated if instead of re-building the entire model, we modify the constraints or objective function, as it shown in the example below.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Compare the computing time in case of the above and below models.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"function solve_economic_dispatch_inplace(\n    generators::Vector,\n    wind,\n    scenario,\n    scale::AbstractVector{Float64},\n)\n    obj_out = Float64[]\n    w_out = Float64[]\n    g1_out = Float64[]\n    g2_out = Float64[]\n    # This function only works for two generators\n    @assert length(generators) == 2\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    N = length(generators)\n    @variable(model, generators[i].min <= g[i = 1:N] <= generators[i].max)\n    @variable(model, 0 <= w <= scenario.wind)\n    @objective(\n        model,\n        Min,\n        sum(generators[i].variable_cost * g[i] for i in 1:N) +\n        wind.variable_cost * w,\n    )\n    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)\n    for c_g1_scale in scale\n        @objective(\n            model,\n            Min,\n            c_g1_scale * generators[1].variable_cost * g[1] +\n            generators[2].variable_cost * g[2] +\n            wind.variable_cost * w,\n        )\n        optimize!(model)\n        push!(obj_out, objective_value(model))\n        push!(w_out, value(w))\n        push!(g1_out, value(g[1]))\n        push!(g2_out, value(g[2]))\n    end\n    df = DataFrames.DataFrame(;\n        scale = scale,\n        dispatch_G1 = g1_out,\n        dispatch_G2 = g2_out,\n        dispatch_wind = w_out,\n        spillage_wind = scenario.wind .- w_out,\n        total_cost = obj_out,\n    )\n    return df\nend\n\nstart = time()\ninplace_df = solve_economic_dispatch_inplace(\n    generators,\n    wind_generator,\n    scenario,\n    0.5:0.1:3.0,\n)\nprint(string(\"elapsed time: \", time() - start, \" seconds\"))","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"For small models, adjusting specific constraints or the objective function is sometimes faster and sometimes slower than re-building the entire model. However, as the problem size increases, updating the model in-place is usually faster.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"inplace_df","category":"page"},{"location":"tutorials/applications/power_systems/#Inefficient-usage-of-wind-generators","page":"Power Systems","title":"Inefficient usage of wind generators","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"The economic dispatch problem does not perform commitment decisions and, thus, assumes that all generators must be dispatched at least at their minimum power output limit. This approach is not cost efficient and may lead to absurd decisions. For example, if d = sum_i in I g^min_i, the wind power injection must be zero, that is, all available wind generation is spilled, to meet the minimum power output constraints on generators.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"In the following example, we adjust the total demand and observed how it affects wind spillage.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"demand_scale_df = DataFrames.DataFrame(;\n    demand = Float64[],\n    dispatch_G1 = Float64[],\n    dispatch_G2 = Float64[],\n    dispatch_wind = Float64[],\n    spillage_wind = Float64[],\n    total_cost = Float64[],\n)\n\nfunction scale_demand(scenario, scale)\n    return Scenario(scale * scenario.demand, scenario.wind)\nend\n\nfor demand_scale in 0.2:0.1:1.4\n    new_scenario = scale_demand(scenario, demand_scale)\n    sol = solve_economic_dispatch(generators, wind_generator, new_scenario)\n    push!(\n        demand_scale_df,\n        (\n            new_scenario.demand,\n            sol.g[1],\n            sol.g[2],\n            sol.w,\n            sol.wind_spill,\n            sol.total_cost,\n        ),\n    )\nend\n\ndemand_scale_df","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"dispatch_plot = StatsPlots.@df(\n    demand_scale_df,\n    Plots.plot(\n        :demand,\n        [:dispatch_G1, :dispatch_G2],\n        labels = [\"G1\" \"G2\"],\n        title = \"Thermal Dispatch\",\n        legend = :bottomright,\n        linewidth = 3,\n        xlabel = \"Demand\",\n        ylabel = \"Dispatch [MW]\",\n    ),\n)\n\nwind_plot = StatsPlots.@df(\n    demand_scale_df,\n    Plots.plot(\n        :demand,\n        [:dispatch_wind, :spillage_wind],\n        labels = [\"Dispatch\" \"Spillage\"],\n        title = \"Wind\",\n        legend = :bottomright,\n        linewidth = 3,\n        xlabel = \"Demand [MW]\",\n        ylabel = \"Energy [MW]\",\n    ),\n)\n\nPlots.plot(dispatch_plot, wind_plot)","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"This particular drawback can be overcome by introducing binary decisions on the \"on/off\" status of generators. This model is called unit commitment and considered later in these notes.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"For further reading on the interplay between wind generation and the minimum power output constraints of generators, we refer interested readers to R. Baldick, \"Wind and energy markets: a case study of Texas,\" IEEE Systems Journal, vol. 6, pp. 27-34, 2012.","category":"page"},{"location":"tutorials/applications/power_systems/#Unit-commitment","page":"Power Systems","title":"Unit commitment","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"The Unit Commitment (UC) model can be obtained from ED model by introducing binary variable associated with each generator. This binary variable can attain two values: if it is \"1,\" the generator is synchronized and, thus, can be dispatched, otherwise, that is, if the binary variable is \"0,\" that generator is not synchronized and its power output is set to 0.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"To obtain the mathematical formulation of the UC model, we will modify the constraints of the ED model as follows:","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"g^min_i cdot u_ti leq g_i leq g^max_i cdot u_ti","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"where u_i in 01 In this constraint, if u_i = 0, then g_i  = 0. On the other hand, if u_i = 1, then g^min_i leq g_i leq g^max_i.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"For further reading on the UC problem we refer interested readers to G. Morales-Espana, J. M. Latorre, and A. Ramos, \"Tight and Compact MILP Formulation for the Thermal Unit Commitment Problem,\" IEEE Transactions on Power Systems, vol. 28, pp. 4897-4908, 2013.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"In the following example we convert the ED model explained above to the UC model.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"function solve_unit_commitment(generators::Vector, wind, scenario)\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    N = length(generators)\n    @variable(model, 0 <= g[i = 1:N] <= generators[i].max)\n    @variable(model, 0 <= w <= scenario.wind)\n    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)\n    # !!! New: add binary on-off variables for each generator\n    @variable(model, u[i = 1:N], Bin)\n    @constraint(model, [i = 1:N], g[i] <= generators[i].max * u[i])\n    @constraint(model, [i = 1:N], g[i] >= generators[i].min * u[i])\n    @objective(\n        model,\n        Min,\n        sum(generators[i].variable_cost * g[i] for i in 1:N) +\n        wind.variable_cost * w +\n        # !!! new\n        sum(generators[i].fixed_cost * u[i] for i in 1:N)\n    )\n    optimize!(model)\n    status = termination_status(model)\n    if status != OPTIMAL\n        return (status = status,)\n    end\n    return (\n        status = status,\n        g = value.(g),\n        w = value(w),\n        wind_spill = scenario.wind - value(w),\n        u = value.(u),\n        total_cost = objective_value(model),\n    )\nend","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Solve the unit commitment problem","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"solution = solve_unit_commitment(generators, wind_generator, scenario)\n\nprintln(\"Dispatch of Generators: \", solution.g, \" MW\")\nprintln(\"Commitments of Generators: \", solution.u)\nprintln(\"Dispatch of Wind: \", solution.w, \" MW\")\nprintln(\"Wind spillage: \", solution.wind_spill, \" MW\")\nprintln(\"Total cost: \\$\", solution.total_cost)","category":"page"},{"location":"tutorials/applications/power_systems/#Unit-commitment-as-a-function-of-demand","page":"Power Systems","title":"Unit commitment as a function of demand","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"After implementing the unit commitment model, we can now assess the interplay between the minimum power output constraints on generators and wind generation.","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"uc_df = DataFrames.DataFrame(;\n    demand = Float64[],\n    commitment_G1 = Float64[],\n    commitment_G2 = Float64[],\n    dispatch_G1 = Float64[],\n    dispatch_G2 = Float64[],\n    dispatch_wind = Float64[],\n    spillage_wind = Float64[],\n    total_cost = Float64[],\n)\n\nfor demand_scale in 0.2:0.1:1.4\n    new_scenario = scale_demand(scenario, demand_scale)\n    sol = solve_unit_commitment(generators, wind_generator, new_scenario)\n    if sol.status == OPTIMAL\n        push!(\n            uc_df,\n            (\n                new_scenario.demand,\n                sol.u[1],\n                sol.u[2],\n                sol.g[1],\n                sol.g[2],\n                sol.w,\n                sol.wind_spill,\n                sol.total_cost,\n            ),\n        )\n    end\n    println(\"Status: $(sol.status) for demand_scale = $(demand_scale)\")\nend","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"uc_df","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"commitment_plot = StatsPlots.@df(\n    uc_df,\n    Plots.plot(\n        :demand,\n        [:commitment_G1, :commitment_G2],\n        labels = [\"G1\" \"G2\"],\n        title = \"Commitment\",\n        legend = :bottomright,\n        linewidth = 3,\n        xlabel = \"Demand [MW]\",\n        ylabel = \"Commitment decision {0, 1}\",\n    ),\n)\n\ndispatch_plot = StatsPlots.@df(\n    uc_df,\n    Plots.plot(\n        :demand,\n        [:dispatch_G1, :dispatch_G2, :dispatch_wind],\n        labels = [\"G1\" \"G2\" \"Wind\"],\n        title = \"Dispatch [MW]\",\n        legend = :bottomright,\n        linewidth = 3,\n        xlabel = \"Demand\",\n        ylabel = \"Dispatch [MW]\",\n    ),\n)\n\nPlots.plot(commitment_plot, dispatch_plot)","category":"page"},{"location":"tutorials/applications/power_systems/#Nonlinear-economic-dispatch","page":"Power Systems","title":"Nonlinear economic dispatch","text":"","category":"section"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"As a final example, we modify our economic dispatch problem in two ways:","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"The thermal cost function is user-defined\nThe output of the wind is only the square-root of the dispatch","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"import Ipopt\n\n\"\"\"\n    thermal_cost_function(g)\n\nA user-defined thermal cost function in pure-Julia! You can include\nnonlinearities, and even things like control flow.\n\n!!! warning\n    It's still up to you to make sure that the function has a meaningful\n    derivative.\n\"\"\"\nfunction thermal_cost_function(g)\n    if g <= 500\n        return g\n    else\n        return g + 1e-2 * (g - 500)^2\n    end\nend\n\nfunction solve_nonlinear_economic_dispatch(\n    generators::Vector,\n    wind,\n    scenario;\n    silent::Bool = false,\n)\n    model = Model(Ipopt.Optimizer)\n    if silent\n        set_silent(model)\n    end\n    @operator(model, op_tcf, 1, thermal_cost_function)\n    N = length(generators)\n    @variable(model, generators[i].min <= g[i = 1:N] <= generators[i].max)\n    @variable(model, 0 <= w <= scenario.wind)\n    @objective(\n        model,\n        Min,\n        sum(generators[i].variable_cost * op_tcf(g[i]) for i in 1:N) +\n        wind.variable_cost * w,\n    )\n    @constraint(model, sum(g[i] for i in 1:N) + sqrt(w) == scenario.demand)\n    optimize!(model)\n    return (\n        g = value.(g),\n        w = value(w),\n        wind_spill = scenario.wind - value(w),\n        total_cost = objective_value(model),\n    )\nend\n\nsolution =\n    solve_nonlinear_economic_dispatch(generators, wind_generator, scenario)","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"Now let's see how the wind is dispatched as a function of the cost:","category":"page"},{"location":"tutorials/applications/power_systems/","page":"Power Systems","title":"Power Systems","text":"wind_cost = 0.0:1:100\nwind_dispatch = Float64[]\nfor c in wind_cost\n    sol = solve_nonlinear_economic_dispatch(\n        generators,\n        WindGenerator(c),\n        scenario;\n        silent = true,\n    )\n    push!(wind_dispatch, sol.w)\nend\n\nPlots.plot(\n    wind_cost,\n    wind_dispatch;\n    xlabel = \"Cost\",\n    ylabel = \"Dispatch [MW]\",\n    label = false,\n)","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"EditURL = \"logistic_regression.jl\"","category":"page"},{"location":"tutorials/conic/logistic_regression/#Logistic-regression","page":"Logistic regression","title":"Logistic regression","text":"","category":"section"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"This tutorial was originally contributed by François Pacaud.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"This tutorial shows how to solve a logistic regression problem with JuMP. Logistic regression is a well known method in machine learning, useful when we want to classify binary variables with the help of a given set of features. To this goal, we find the optimal combination of features maximizing the (log)-likelihood onto a training set. From a modern optimization glance, the resulting problem is convex and differentiable. On a modern optimization glance, it is even conic representable.","category":"page"},{"location":"tutorials/conic/logistic_regression/#Formulating-the-logistic-regression-problem","page":"Logistic regression","title":"Formulating the logistic regression problem","text":"","category":"section"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"Suppose we have a set of training data-point i = 1 cdots n, where for each i we have a vector of features x_i in mathbbR^p and a categorical observation y_i in -1 1.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"The log-likelihood is given by","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"l(theta) = sum_i=1^n log(dfrac11 + exp(-y_i theta^top x_i))","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"and the optimal theta minimizes the logistic loss function:","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"min_theta sum_i=1^n log(1 + exp(-y_i theta^top x_i))","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"Most of the time, instead of solving directly the previous optimization problem, we prefer to add a regularization term:","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"min_theta sum_i=1^n log(1 + exp(-y_i theta^top x_i)) + lambda  theta ","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"with lambda in mathbbR_+ a penalty and  a norm function. By adding such a regularization term, we avoid overfitting on the training set and usually achieve a greater score in cross-validation.","category":"page"},{"location":"tutorials/conic/logistic_regression/#Reformulation-as-a-conic-optimization-problem","page":"Logistic regression","title":"Reformulation as a conic optimization problem","text":"","category":"section"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"By introducing auxiliary variables t_1 cdots t_n and r, the optimization problem is equivalent to","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"beginaligned\nmin_t r theta  sum_i=1^n t_i + lambda r \ntextsubject to   quad t_i geq log(1 + exp(- y_i theta^top x_i)) \n                    quad r geq theta\nendaligned","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"Now, the trick is to reformulate the constraints t_i geq log(1 + exp(- y_i theta^top x_i)) with the help of the exponential cone","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"K_exp =  (x y z) in mathbbR^3   y exp(x  y) leq z  ","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"Indeed, by passing to the exponential, we see that for all i=1 cdots n, the constraint t_i geq log(1 + exp(- y_i theta^top x_i)) is equivalent to","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"exp(-t_i) + exp(u_i - t_i) leq 1","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"with u_i = -y_i theta^top x_i. Then, by adding two auxiliary variables z_i1 and z_i2 such that z_i1 geq exp(u_i-t_i) and z_i2 geq exp(-t_i), we get the equivalent formulation","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"left\nbeginaligned\n(u_i -t_i  1 z_i1)  in  K_exp  \n(-t_i  1 z_i2)  in  K_exp  \nz_i1 + z_i2  leq  1\nendaligned\nright","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"In this setting, the conic version of the logistic regression problems writes out","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"beginaligned\nmin_t z r theta   sum_i=1^n t_i + lambda r \ntextsubject to   quad  (u_i -t_i  1 z_i1)  in  K_exp  \n                    quad  (-t_i  1 z_i2)  in  K_exp  \n                    quad  z_i1 + z_i2  leq  1 \n                    quad u_i = -y_i x_i^top theta \n                    quad r geq theta\nendaligned","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"and thus encompasses 3n + p + 1 variables and 3n + 1 constraints (u_i = -y_i theta^top x_i is only a virtual constraint used to clarify the notation). Thus, if n gg 1, we get a large number of variables and constraints.","category":"page"},{"location":"tutorials/conic/logistic_regression/#Fitting-logistic-regression-with-a-conic-solver","page":"Logistic regression","title":"Fitting logistic regression with a conic solver","text":"","category":"section"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"It is now time to pass to the implementation. We choose SCS as a conic solver.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"using JuMP\nimport Random\nimport SCS","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"info: Info\nThis tutorial uses sets from MathOptInterface. By default, JuMP exports the MOI symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:import MathOptInterface as MOI","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"Random.seed!(2713);\nnothing #hide","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"We start by implementing a function to generate a fake dataset, and where we could tune the correlation between the feature variables. The function is a direct transcription of the one used in this blog post.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"function generate_dataset(n_samples = 100, n_features = 10; shift = 0.0)\n    X = randn(n_samples, n_features)\n    w = randn(n_features)\n    y = sign.(X * w)\n    X .+= 0.8 * randn(n_samples, n_features) # add noise\n    X .+= shift # shift the points in the feature space\n    X = hcat(X, ones(n_samples, 1))\n    return X, y\nend","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"We write a softplus function to formulate each constraint t geq log(1 + exp(u)) with two exponential cones.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"function softplus(model, t, u)\n    z = @variable(model, [1:2], lower_bound = 0.0)\n    @constraint(model, sum(z) <= 1.0)\n    @constraint(model, [u - t, 1, z[1]] in MOI.ExponentialCone())\n    @constraint(model, [-t, 1, z[2]] in MOI.ExponentialCone())\nend","category":"page"},{"location":"tutorials/conic/logistic_regression/#\\ell_2-regularized-logistic-regression","page":"Logistic regression","title":"ell_2 regularized logistic regression","text":"","category":"section"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"Then, with the help of the softplus function, we could write our optimization model. In the ell_2 regularization case, the constraint r geq theta_2 rewrites as a second order cone constraint.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"function build_logit_model(X, y, λ)\n    n, p = size(X)\n    model = Model()\n    @variable(model, θ[1:p])\n    @variable(model, t[1:n])\n    for i in 1:n\n        u = -(X[i, :]' * θ) * y[i]\n        softplus(model, t[i], u)\n    end\n    # Add ℓ2 regularization\n    @variable(model, 0.0 <= reg)\n    @constraint(model, [reg; θ] in SecondOrderCone())\n    # Define objective\n    @objective(model, Min, sum(t) + λ * reg)\n    return model\nend","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"We generate the dataset.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"warning: Warning\nBe careful here, for large n and p SCS could fail to converge.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"n, p = 200, 10\nX, y = generate_dataset(n, p; shift = 10.0);\n\n# We could now solve the logistic regression problem\nλ = 10.0\nmodel = build_logit_model(X, y, λ)\nset_optimizer(model, SCS.Optimizer)\nset_silent(model)\nJuMP.optimize!(model)","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"θ♯ = JuMP.value.(model[:θ])","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"It appears that the speed of convergence is not that impacted by the correlation of the dataset, nor by the penalty lambda.","category":"page"},{"location":"tutorials/conic/logistic_regression/#\\ell_1-regularized-logistic-regression","page":"Logistic regression","title":"ell_1 regularized logistic regression","text":"","category":"section"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"We now formulate the logistic problem with a ell_1 regularization term. The ell_1 regularization ensures sparsity in the optimal solution of the resulting optimization problem. Luckily, the ell_1 norm is implemented as a set in MathOptInterface. Thus, we could formulate the sparse logistic regression problem with the help of a MOI.NormOneCone set.","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"function build_sparse_logit_model(X, y, λ)\n    n, p = size(X)\n    model = Model()\n    @variable(model, θ[1:p])\n    @variable(model, t[1:n])\n    for i in 1:n\n        u = -(X[i, :]' * θ) * y[i]\n        softplus(model, t[i], u)\n    end\n    # Add ℓ1 regularization\n    @variable(model, 0.0 <= reg)\n    @constraint(model, [reg; θ] in MOI.NormOneCone(p + 1))\n    # Define objective\n    @objective(model, Min, sum(t) + λ * reg)\n    return model\nend\n\n# Auxiliary function to count non-null components:\ncount_nonzero(v::Vector; tol = 1e-6) = sum(abs.(v) .>= tol)\n\n# We solve the sparse logistic regression problem on the same dataset as before.\nλ = 10.0\nsparse_model = build_sparse_logit_model(X, y, λ)\nset_optimizer(sparse_model, SCS.Optimizer)\nset_silent(sparse_model)\nJuMP.optimize!(sparse_model)","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"θ♯ = JuMP.value.(sparse_model[:θ])\nprintln(\n    \"Number of non-zero components: \",\n    count_nonzero(θ♯),\n    \" (out of \",\n    p,\n    \" features)\",\n)","category":"page"},{"location":"tutorials/conic/logistic_regression/#Extensions","page":"Logistic regression","title":"Extensions","text":"","category":"section"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"A direct extension would be to consider the sparse logistic regression with hard thresholding, which, on contrary to the soft version using a ell_1 regularization, adds an explicit cardinality constraint in its formulation:","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"beginaligned\nmin_theta   sum_i=1^n log(1 + exp(-y_i theta^top x_i)) + lambda  theta _2^2 \ntextsubject to   quad  theta _0 = k\nendaligned","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"where k is the maximum number of non-zero components in the vector theta, and _0 is the ell_0 pseudo-norm:","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":" x_0 = i   x_i neq 0","category":"page"},{"location":"tutorials/conic/logistic_regression/","page":"Logistic regression","title":"Logistic regression","text":"The cardinality constraint theta_0 leq k could be reformulated with binary variables. Thus the hard sparse regression problem could be solved by any solver supporting mixed integer conic problems.","category":"page"},{"location":"moi/background/motivation/","page":"Motivation","title":"Motivation","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/motivation.md\"","category":"page"},{"location":"moi/background/motivation/#Motivation","page":"Motivation","title":"Motivation","text":"","category":"section"},{"location":"moi/background/motivation/","page":"Motivation","title":"Motivation","text":"MathOptInterface (MOI) is a replacement for MathProgBase, the first-generation abstraction layer for mathematical optimization previously used by JuMP and Convex.jl.","category":"page"},{"location":"moi/background/motivation/","page":"Motivation","title":"Motivation","text":"To address a number of limitations of MathProgBase, MOI is designed to:","category":"page"},{"location":"moi/background/motivation/","page":"Motivation","title":"Motivation","text":"Be simple and extensible\nunifying linear, quadratic, and conic optimization,\nseamlessly facilitating extensions to essentially arbitrary constraints and functions (for example, indicator constraints, complementarity constraints, and piecewise-linear functions)\nBe fast\nby allowing access to a solver's in-memory representation of a problem without writing intermediate files (when possible)\nby using multiple dispatch and avoiding requiring containers of non-concrete types\nAllow a solver to return multiple results (for example, a pool of solutions)\nAllow a solver to return extra arbitrary information via attributes (for example, variable- and constraint-wise membership in an irreducible inconsistent subset for infeasibility analysis)\nProvide a greatly expanded set of status codes explaining what happened during the optimization procedure\nEnable a solver to more precisely specify which problem classes it supports\nEnable both primal and dual warm starts\nEnable adding and removing both variables and constraints by indices that are not required to be consecutive\nEnable any modification that the solver supports to an existing model\nAvoid requiring the solver wrapper to store an additional copy of the problem data","category":"page"},{"location":"tutorials/conic/introduction/#Introduction","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"tutorials/conic/introduction/","page":"Introduction","title":"Introduction","text":"Conic programs are a class of convex nonlinear optimization problems which use cones to represent the nonlinearities. They have the form:","category":"page"},{"location":"tutorials/conic/introduction/","page":"Introduction","title":"Introduction","text":"beginalign\n     min_x in mathbbR^n  f_0(x) \n     textst  f_j(x) in mathcalS_j   j = 1 ldots m\nendalign","category":"page"},{"location":"tutorials/conic/introduction/","page":"Introduction","title":"Introduction","text":"Mixed-integer conic programs (MICPs) are extensions of conic programs in which some (or all) of the decision variables take discrete values.","category":"page"},{"location":"tutorials/conic/introduction/#How-to-choose-a-solver","page":"Introduction","title":"How to choose a solver","text":"","category":"section"},{"location":"tutorials/conic/introduction/","page":"Introduction","title":"Introduction","text":"JuMP supports a range of conic solvers, although support differs on what types of cones each solver supports. In the list of Supported solvers, \"SOCP\" denotes solvers supporting second-order cones and \"SDP\" denotes solvers supporting semidefinite cones. In addition, solvers such as SCS and Mosek have support for the exponential cone. Moreover, due to the bridging system in MathOptInterface, many of these solvers support a much wider range of exotic cones than they natively support. Solvers supporting discrete variables start with \"(MI)\" in the list of Supported solvers.","category":"page"},{"location":"tutorials/conic/introduction/","page":"Introduction","title":"Introduction","text":"tip: Tip\nDuality plays a large role in solving conic optimization models. Depending on the solver, it can be more efficient to solve the dual instead of the primal. If performance is an issue, see the Dualization tutorial for more details.","category":"page"},{"location":"tutorials/conic/introduction/#How-these-tutorials-are-structured","page":"Introduction","title":"How these tutorials are structured","text":"","category":"section"},{"location":"tutorials/conic/introduction/","page":"Introduction","title":"Introduction","text":"Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.","category":"page"},{"location":"tutorials/conic/introduction/","page":"Introduction","title":"Introduction","text":"The following tutorials are worked examples that present a problem in words, then formulate it in mathematics, and then solve it in JuMP. This usually involves some sort of visualization of the solution. Start here if you are new to JuMP.\nExperiment design\nLogistic regression\nThe Tips and tricks tutorial contains a number of helpful reformulations and tricks you can use when modeling conic programs. Look here if you are stuck trying to formulate a problem as a conic program.\nThe remaining tutorials are less verbose and styled in the form of short code examples. These tutorials have less explanation, but may contain useful code snippets, particularly if they are similar to a problem you are trying to solve.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/manipulating_expressions.md\"","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Manipulating-expressions","page":"Manipulating expressions","title":"Manipulating expressions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"This guide highlights a syntactically appealing way to build expressions at the MOI level, but also to look at their contents. It may be especially useful when writing models or bridge code.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Creating-functions","page":"Manipulating expressions","title":"Creating functions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"This section details the ways to create functions with MathOptInterface.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Creating-scalar-affine-functions","page":"Manipulating expressions","title":"Creating scalar affine functions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"The simplest scalar function is simply a variable:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> x = MOI.add_variable(model) # Create the variable x\nMOI.VariableIndex(1)","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"This type of function is extremely simple; to express more complex functions, other types must be used. For instance, a ScalarAffineFunction is a sum of linear terms (a factor times a variable) and a constant. Such an object can be built using the standard constructor:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1, x)], 2) # x + 2\n(2) + (1) MOI.VariableIndex(1)","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"However, you can also use operators to build the same scalar function:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> f = x + 2\n(2) + (1) MOI.VariableIndex(1)","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Creating-scalar-quadratic-functions","page":"Manipulating expressions","title":"Creating scalar quadratic functions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"Scalar quadratic functions are stored in ScalarQuadraticFunction objects, in a way that is highly similar to scalar affine functions. You can obtain a quadratic function as a product of affine functions:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> 1 * x * x\n(0) + 1.0 MOI.VariableIndex(1)²\n\njulia> f * f  # (x + 2)²\n(4) + (2) MOI.VariableIndex(1) + (2) MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(1)²\n\njulia> f^2  # (x + 2)² too\n(4) + (2) MOI.VariableIndex(1) + (2) MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(1)²","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Creating-vector-functions","page":"Manipulating expressions","title":"Creating vector functions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"A vector function is a function with several values, irrespective of the number of input variables. Similarly to scalar functions, there are three main types of vector functions: VectorOfVariables, VectorAffineFunction, and VectorQuadraticFunction.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"The easiest way to create a vector function is to stack several scalar functions using Utilities.vectorize. It takes a vector as input, and the generated vector function (of the most appropriate type) has each dimension corresponding to a dimension of the vector.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> g = MOI.Utilities.vectorize([f, 2 * f])\n┌                              ┐\n│(2) + (1) MOI.VariableIndex(1)│\n│(4) + (2) MOI.VariableIndex(1)│\n└                              ┘","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"warning: Warning\nUtilities.vectorize only takes a vector of similar scalar functions: you cannot mix VariableIndex and ScalarAffineFunction, for instance. In practice, it means that Utilities.vectorize([x, f]) does not work; you should rather use Utilities.vectorize([1 * x, f]) instead to only have ScalarAffineFunction objects.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Canonicalizing-functions","page":"Manipulating expressions","title":"Canonicalizing functions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"In more advanced use cases, you might need to ensure that a function is \"canonical.\" Functions are stored as an array of terms, but there is no check that these terms are redundant: a ScalarAffineFunction object might have two terms with the same variable, like x + x + 1. These terms could be merged without changing the semantics of the function: 2x + 1.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"Working with these objects might be cumbersome. Canonicalization helps maintain redundancy to zero.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"Utilities.is_canonical checks whether a function is already in its canonical form:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> MOI.Utilities.is_canonical(f + f) # (x + 2) + (x + 2) is stored as x + x + 4\nfalse","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"Utilities.canonical returns the equivalent canonical version of the function:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> MOI.Utilities.canonical(f + f) # Returns 2x + 4\n(4) + (2) MOI.VariableIndex(1)","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Exploring-functions","page":"Manipulating expressions","title":"Exploring functions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"At some point, you might need to dig into a function, for instance to map it into solver constructs.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/#Vector-functions","page":"Manipulating expressions","title":"Vector functions","text":"","category":"section"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"Utilities.scalarize returns a vector of scalar functions from a vector function:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> MOI.Utilities.scalarize(g) # Returns a vector [f, 2 * f].\n2-element Vector{MathOptInterface.ScalarAffineFunction{Int64}}:\n (2) + (1) MOI.VariableIndex(1)\n (4) + (2) MOI.VariableIndex(1)","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"note: Note\nUtilities.eachscalar returns an iterator on the dimensions, which serves the same purpose as Utilities.scalarize.","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"output_dimension returns the number of dimensions of the output of a function:","category":"page"},{"location":"moi/tutorials/manipulating_expressions/","page":"Manipulating expressions","title":"Manipulating expressions","text":"julia> MOI.output_dimension(g)\n2","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"EditURL = \"https://github.com/jump-dev/MosekTools.jl/blob/v0.15.1/README.md\"","category":"page"},{"location":"packages/MosekTools/#MosekTools.jl","page":"jump-dev/MosekTools.jl","title":"MosekTools.jl","text":"","category":"section"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"MosekTools.jl is the MathOptInterface.jl implementation for the MOSEK solver.","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"The low-level solver API for MOSEK is found in the package Mosek.jl.","category":"page"},{"location":"packages/MosekTools/#Affiliation","page":"jump-dev/MosekTools.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"MosekTools.jl is maintained by the JuMP community and is not officially supported by MOSEK. However, Mosek.jl is an officially supported product of MOSEK.","category":"page"},{"location":"packages/MosekTools/#License","page":"jump-dev/MosekTools.jl","title":"License","text":"","category":"section"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"MosekTools.jl is licensed under the MIT License.","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"The underlying solver is a closed-source commercial product for which you must obtain a license.","category":"page"},{"location":"packages/MosekTools/#Installation","page":"jump-dev/MosekTools.jl","title":"Installation","text":"","category":"section"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"The latest release of this package and the master branch are to be used with the latest release of Mosek.jl (which uses MOSEK v10).","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"To use MOSEK v9 (resp. v8), use the v0.12.x (resp. v0.7.x) releases of this package, and the mosekv9 (resp. mosekv8) branch and v1.2.x (resp. v0.9.x) releases of Mosek.jl.","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"See the following table for a summary:","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"MOSEK Mosek.jl MosekTools.jl release MosekTools.jl branch\nv10 v10 v0.13 master\nv9 v0.12 v0.12 mosekv9\nv8 v0.9 v0.7 mosekv8","category":"page"},{"location":"packages/MosekTools/#Use-with-JuMP","page":"jump-dev/MosekTools.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"using JuMP\nusing MosekTools\nmodel = Model(Mosek.Optimizer)\nset_attribute(model, \"QUIET\", true)\nset_attribute(model, \"INTPNT_CO_TOL_DFEAS\", 1e-7)","category":"page"},{"location":"packages/MosekTools/#Options","page":"jump-dev/MosekTools.jl","title":"Options","text":"","category":"section"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"The parameter QUIET is a special parameter that when set to true disables all Mosek printing output.","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"All other parameters can be found in the Mosek documentation.","category":"page"},{"location":"packages/MosekTools/","page":"jump-dev/MosekTools.jl","title":"jump-dev/MosekTools.jl","text":"Note that the prefix MSK_IPAR_ (for integer parameters), MSK_DPAR_ (for floating point parameters) or MSK_SPAR_ (for string parameters) are optional. If they are not given, they are inferred from the type of the value. For example, in the example above, as 1e-7 is a floating point number, the parameters name used is MSK_DPAR_INTPNT_CO_TOL_DFEAS.","category":"page"},{"location":"developers/style/#Style-guide-and-design-principles","page":"Style Guide","title":"Style guide and design principles","text":"","category":"section"},{"location":"developers/style/#Style-guide","page":"Style Guide","title":"Style guide","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"This section describes the coding style rules that apply to JuMP code and that we recommend for JuMP models and surrounding Julia code. The motivations for a style guide include:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"conveying best practices for writing readable and maintainable code\nreducing the amount of time spent on bike-shedding by establishing basic naming and formatting conventions\nlowering the barrier for new contributors by codifying the existing practices (for example, you can be more confident your code will pass review if you follow the style guide)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"In some cases, the JuMP style guide diverges from the Julia style guide. All such cases will be explicitly noted and justified.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"The JuMP style guide adopts many recommendations from the Google style guides.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"info: Info\nThe style guide is always a work in progress, and not all JuMP code follows the rules. When modifying JuMP, please fix the style violations of the surrounding code (that is, leave the code tidier than when you started). If large changes are needed, consider separating them into another PR.","category":"page"},{"location":"developers/style/#JuliaFormatter","page":"Style Guide","title":"JuliaFormatter","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"JuMP uses JuliaFormatter.jl as an auto-formatting tool.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"We use the options contained in .JuliaFormatter.toml.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"To format code, cd to the JuMP directory, then run:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"] add JuliaFormatter@1\nusing JuliaFormatter\nformat(\"docs\")\nformat(\"src\")\nformat(\"test\")","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"info: Info\nA continuous integration check verifies that all PRs made to JuMP have passed the formatter.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"The following sections outline extra style guide points that are not fixed automatically by JuliaFormatter.","category":"page"},{"location":"developers/style/#Abstract-types-and-composition","page":"Style Guide","title":"Abstract types and composition","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Specifying types for method arguments is mostly optional in Julia. The benefit of abstract method arguments is that it enables functions and types from one package to be used with functions and types from another package via multiple dispatch.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"However, abstractly typed methods have two main drawbacks:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"It's possible to find out that you are working with unexpected types deep in the call chain, potentially leading to hard-to-diagnose MethodErrors.\nUntyped function arguments can lead to correctness problems if the user's choice of input type does not satisfy the assumptions made by the author of the function.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"As a motivating example, consider the following function:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> function my_sum(x)\n           y = 0.0\n           for i in 1:length(x)\n               y += x[i]\n           end\n           return y\n       end\nmy_sum (generic function with 1 method)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"This function contains a number of implicit assumptions about the type of x:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"x supports 1-based getindex and implements length\nThe element type of x supports addition with 0.0, and then with the result of x + 0.0.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"info: Info\nAs a motivating example for the second point, VariableRef plus Float64 produces an AffExpr. Do not assume that +(::A, ::B) produces an instance of the type A or B.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"my_sum works as expected if the user passes in Vector{Float64}:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> my_sum([1.0, 2.0, 3.0])\n6.0","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"but it doesn't respect input types, for example returning a Float64 if the user passes Vector{Int}:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> my_sum([1, 2, 3])\n6.0","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"but it throws a MethodError if the user passes String:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> my_sum(\"abc\")\nERROR: MethodError: no method matching +(::Float64, ::Char)\n[...]","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"This particular MethodError is hard to debug, particularly for new users, because it mentions +, Float64, and Char, none of which were called or passed by the user.","category":"page"},{"location":"developers/style/#Dealing-with-MethodErrors","page":"Style Guide","title":"Dealing with MethodErrors","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"This section diverges from the Julia style guide, as well as other common guides like SciML. The following suggestions are intended to provide a friendlier experience for novice Julia programmers, at the cost of limiting the power and flexibility of advanced Julia programmers.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Code should follow the MethodError principle:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"info: The MethodError principle\nA user should see a MethodError only for methods that they called directly.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Bad:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"_internal_function(x::Integer) = x + 1\n# The user sees a MethodError for _internal_function when calling\n# public_function(\"a string\"). This is not very helpful.\npublic_function(x) = _internal_function(x)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Good:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"_internal_function(x::Integer) = x + 1\n# The user sees a MethodError for public_function when calling\n# public_function(\"a string\"). This is easy to understand.\npublic_function(x::Integer) = _internal_function(x)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"If it is hard to provide an error message at the top of the call chain, then the following pattern is also ok:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"_internal_function(x::Integer) = x + 1\nfunction _internal_function(x)\n    error(\n        \"Internal error. This probably means that you called \" *\n        \"`public_function()`s with the wrong type.\",\n    )\nend\npublic_function(x) = _internal_function(x)","category":"page"},{"location":"developers/style/#Dealing-with-correctness","page":"Style Guide","title":"Dealing with correctness","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Dealing with correctness is harder, because Julia has no way of formally specifying interfaces that abstract types must implement. Instead, here are two options that you can use when writing and interacting with generic code:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Option 1: use concrete types and let users extend new methods.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"In this option, explicitly restrict input arguments to concrete types that are tested and have been validated for correctness. For example:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> function my_sum_option_1(x::Vector{Float64})\n           y = 0.0\n           for i in 1:length(x)\n               y += x[i]\n           end\n           return y\n       end\nmy_sum_option_1 (generic function with 1 method)\n\njulia> my_sum_option_1([1.0, 2.0, 3.0])\n6.0","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Using concrete types satisfies the MethodError principle:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> my_sum_option_1(\"abc\")\nERROR: MethodError: no method matching my_sum_option_1(::String)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"and it allows other types to be supported in future by defining new methods:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> function my_sum_option_1(x::Array{T,N}) where {T<:Number,N}\n           y = zero(T)\n           for i in eachindex(x)\n               y += x[i]\n           end\n           return y\n       end\nmy_sum_option_1 (generic function with 2 methods)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Importantly, these methods do not have to be defined in the original package.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"info: Info\nSome usage of abstract types is okay. For example, in my_sum_option_1, we allowed the element type, T, to be a subtype of Number. This is fairly safe, but it still has an implicit assumption that T supports zero(T) and +(::T, ::T).","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Option 2: program defensively, and validate all assumptions.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"An alternative is to program defensively, and to rigorously document and validate all assumptions that the code makes. In particular:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"All assumptions on abstract types that aren't guaranteed by the definition of the abstract type (for example, optional methods without a fallback) should be documented.\nIf practical, the assumptions should be checked in code, and informative error messages should be provided to the user if the assumptions are not met. In general, these checks may be expensive, so you should prefer to do this once, at the highest level of the call-chain.\nTests should cover for a range of corner cases and argument types.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"For example:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"\"\"\"\n    test_my_sum_defensive_assumptions(x::AbstractArray{T}) where {T}\n\nTest the assumptions made by `my_sum_defensive`.\n\"\"\"\nfunction test_my_sum_defensive_assumptions(x::AbstractArray{T}) where {T}\n    try\n        # Some types may not define zero.\n        @assert zero(T) isa T\n        # Check iteration supported\n        @assert iterate(x) isa Union{Nothing,Tuple{T,Int}}\n        # Check that + is defined\n        @assert +(zero(T), zero(T)) isa Any\n    catch err\n        error(\n            \"Unable to call my_sum_defensive(::$(typeof(x))) because \" *\n            \"it failed an internal assumption\",\n        )\n    end\n    return\nend\n\n\"\"\"\n    my_sum_defensive(x::AbstractArray{T}) where {T}\n\nReturn the sum of the elements in the abstract array `x`.\n\n## Assumptions\n\nThis function makes the following assumptions:\n\n * That `zero(T)` is defined\n * That `x` supports the iteration interface\n * That  `+(::T, ::T)` is defined\n\"\"\"\nfunction my_sum_defensive(x::AbstractArray{T}) where {T}\n    test_my_sum_defensive_assumptions(x)\n    y = zero(T)\n    for xi in x\n        y += xi\n    end\n    return y\nend\n\n# output\n\nmy_sum_defensive","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"This function works on Vector{Float64}:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> my_sum_defensive([1.0, 2.0, 3.0])\n6.0","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"as well as Matrix{Rational{Int}}:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> my_sum_defensive([(1//2) + (4//3)im; (6//5) + (7//11)im])\n17//10 + 65//33*im","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"and it throws an error when the assumptions aren't met:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"julia> my_sum_defensive(['a', 'b', 'c'])\nERROR: Unable to call my_sum_defensive(::Vector{Char}) because it failed an internal assumption\n[...]","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"As an alternative, you may choose not to call test_my_sum_defensive_assumptions within my_sum_defensive, and instead ask users of my_sum_defensive to call it in their tests.","category":"page"},{"location":"developers/style/#Juxtaposed-multiplication","page":"Style Guide","title":"Juxtaposed multiplication","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Only use juxtaposed multiplication when the right-hand side is a symbol.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Good:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"2x  # Acceptable if there are space constraints.\n2 * x  # This is preferred if space is not an issue.\n2 * (x + 1)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Bad:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"2(x + 1)","category":"page"},{"location":"developers/style/#Empty-vectors","page":"Style Guide","title":"Empty vectors","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"For a type T, T[] and Vector{T}() are equivalent ways to create an empty vector with element type T. Prefer T[] because it is more concise.","category":"page"},{"location":"developers/style/#Comments","page":"Style Guide","title":"Comments","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"For non-native speakers and for general clarity, comments in code must be proper English sentences with appropriate punctuation.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Good:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"# This is a comment demonstrating a good comment.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Bad:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"# a bad comment","category":"page"},{"location":"developers/style/#JuMP-macro-syntax","page":"Style Guide","title":"JuMP macro syntax","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"For consistency, always use parentheses.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Good:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variable(model, x >= 0)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Bad:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variable model x >= 0","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"For consistency, always use constant * variable as opposed to variable * constant. This makes it easier to read models in ambiguous cases like a * x.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Good:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"a = 4\n@constraint(model, 3 * x <= 1)\n@constraint(model, a * x <= 1)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Bad:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"a = 4\n@constraint(model, x * 3 <= 1)\n@constraint(model, x * a <= 1)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"In order to reduce boilerplate code, prefer the plural form of macros over lots of repeated calls to singular forms.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Good:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variables(model, begin\n    x >= 0\n    y >= 1\n    z <= 2\nend)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Bad:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variable(model, x >= 0)\n@variable(model, y >= 1)\n@variable(model, z <= 2)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"An exception is made for calls with many keyword arguments, since these need to be enclosed in parentheses in order to parse properly.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Acceptable:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variable(model, x >= 0, start = 0.0, base_name = \"my_x\")\n@variable(model, y >= 1, start = 2.0)\n@variable(model, z <= 2, start = -1.0)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Also acceptable:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variables(model, begin\n    x >= 0, (start = 0.0, base_name = \"my_x\")\n    y >= 1, (start = 2.0)\n    z <= 2, (start = -1.0)\nend)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"While we always use in for for-loops, it is acceptable to use = in the container declarations of JuMP macros.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Okay:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variable(model, x[i=1:3])","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Also okay:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@variable(model, x[i in 1:3])","category":"page"},{"location":"developers/style/#Naming","page":"Style Guide","title":"Naming","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"module SomeModule end\nfunction some_function end\nconst SOME_CONSTANT = ...\nstruct SomeStruct\n  some_field::SomeType\nend\n@enum SomeEnum ENUM_VALUE_A ENUM_VALUE_B\nsome_local_variable = ...\nsome_file.jl # Except for ModuleName.jl.","category":"page"},{"location":"developers/style/#Exported-and-non-exported-names","page":"Style Guide","title":"Exported and non-exported names","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Begin private module level functions and constants with an underscore. All other objects in the scope of a module should be exported. (See JuMP.jl for an example of how to do this.)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Names beginning with an underscore should only be used for distinguishing between exported (public) and non-exported (private) objects. Therefore, never begin the name of a local variable with an underscore.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"module MyModule\n\nexport public_function, PUBLIC_CONSTANT\n\nfunction _private_function()\n    local_variable = 1\n    return\nend\n\nfunction public_function end\n\nconst _PRIVATE_CONSTANT = 3.14159\nconst PUBLIC_CONSTANT = 1.41421\n\nend","category":"page"},{"location":"developers/style/#Use-of-underscores-within-names","page":"Style Guide","title":"Use of underscores within names","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"The Julia style guide recommends avoiding underscores \"when readable,\" for example, haskey, isequal, remotecall, and remotecall_fetch. This convention creates the potential for unnecessary bikeshedding and also forces the user to recall the presence/absence of an underscore, for example, \"was that argument named basename or base_name?\". For consistency, always use underscores in variable names and function names to separate words.","category":"page"},{"location":"developers/style/#Use-of-!","page":"Style Guide","title":"Use of !","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Julia has a convention of appending ! to a function name if the function modifies its arguments. We recommend to:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Omit ! when the name itself makes it clear that modification is taking place, for example, add_constraint and set_name. We depart from the Julia style guide because ! does not provide a reader with any additional information in this case, and adherence to this convention is not uniform even in base Julia itself (consider Base.println and Base.finalize).\nUse ! in all other cases. In particular it can be used to distinguish between modifying and non-modifying variants of the same function like scale and scale!.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Note that ! is not a self-documenting feature because it is still ambiguous which arguments are modified when multiple arguments are present. Be sure to document which arguments are modified in the method's docstring.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"See also the Julia style guide recommendations for ordering of function arguments.","category":"page"},{"location":"developers/style/#Abbreviations","page":"Style Guide","title":"Abbreviations","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Abbreviate names to make the code more readable, not to save typing. Don't arbitrarily delete letters from a word to abbreviate it (for example, indx). Use abbreviations consistently within a body of code (for example, do not mix con and constr, idx and indx).","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Common abbreviations:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"num for number\ncon for constraint","category":"page"},{"location":"developers/style/#No-one-letter-variable-names","page":"Style Guide","title":"No one-letter variable names","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Where possible, avoid one-letter variable names.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Use model = Model() instead of m = Model()","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Exceptions are made for indices in loops.","category":"page"},{"location":"developers/style/#@enum-vs.-Symbol","page":"Style Guide","title":"@enum vs. Symbol","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"The @enum macro lets you define types with a finite number of values that are explicitly enumerated (like enum in C/C++). Symbols are lightweight strings that are used to represent identifiers in Julia (for example, :x).","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"@enum provides type safety and can have docstrings attached to explain the possible values. Use @enums when applicable, for example, for reporting statuses. Use strings to provide long-form additional information like error messages.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Use of Symbol should typically be reserved for identifiers, for example, for lookup in the JuMP model (model[:my_variable]).","category":"page"},{"location":"developers/style/#using-vs.-import","page":"Style Guide","title":"using vs. import","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"using ModuleName brings all symbols exported by the module ModuleName into scope, while import ModuleName brings only the module itself into scope. (See the Julia manual) for examples and more details.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"For the same reason that from <module> import * is not recommended in python (PEP 8), avoid using ModuleName except in throw-away scripts or at the REPL. The using statement makes it harder to track where symbols come from and exposes the code to ambiguities when two modules export the same symbol.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Prefer using ModuleName: x, p to import ModuleName.x, ModuleName.p and import MyModule: x, p because the import versions allow method extension without qualifying with the module name.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Similarly, using ModuleName: ModuleName is an acceptable substitute for import ModuleName, because it does not bring all symbols exported by ModuleName into scope. However, we prefer import ModuleName for consistency.","category":"page"},{"location":"developers/style/#Documentation","page":"Style Guide","title":"Documentation","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"This section describes the writing style that should be used when writing documentation for JuMP (and supporting packages).","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"We can recommend the documentation style guides by Divio, Google, and Write the Docs as general reading for those writing documentation. This guide delegates a thorough handling of the topic to those guides and instead elaborates on the points more specific to Julia and documentation that use Documenter.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Be concise\nUse lists instead of long sentences\nUse numbered lists when describing a sequence, for example, (1) do X, (2) then Y\nUse bullet points when the items are not ordered\nExample code should be covered by doctests\nWhen a word is a Julia symbol and not an English word, enclose it with backticks. In addition, if it has a docstring in this doc add a link using @ref. If it is a plural, add the \"s\" after the closing backtick. For example,\n[`VariableRef`](@ref)s\nUse @meta blocks for TODOs and other comments that shouldn't be visible to readers. For example,\n```@meta\n# TODO: Mention also X, Y, and Z.\n```","category":"page"},{"location":"developers/style/#Docstrings","page":"Style Guide","title":"Docstrings","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Every exported object needs a docstring\nAll examples in docstrings should be jldoctests\nAlways use complete English sentences with proper punctuation\nDo not terminate lists with punctuation (for example, as in this doc)","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Here is an example:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"\"\"\"\n    signature(args; kwargs...)\n\nShort sentence describing the function.\n\nOptional: add a slightly longer paragraph describing the function.\n\n## Notes\n\n - List any notes that the user should be aware of\n\n## Examples\n\n```jldoctest\njulia> 1 + 1\n2\n```\n\"\"\"","category":"page"},{"location":"developers/style/#Testing","page":"Style Guide","title":"Testing","text":"","category":"section"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Use a module to encapsulate tests, and structure all tests as functions. This avoids leaking local variables between tests.","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Here is a basic skeleton:","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"module TestPkg\n\nusing Test\n\nfunction runtests()\n    for name in names(@__MODULE__; all = true)\n        if startswith(\"$(name)\", \"test_\")\n            @testset \"$(name)\" begin\n                getfield(@__MODULE__, name)()\n            end\n        end\n    end\n    return\nend\n\n_helper_function() = 2\n\nfunction test_addition()\n    @test 1 + 1 == _helper_function()\n    return\nend\n\nend # module TestPkg\n\nTestPkg.runtests()","category":"page"},{"location":"developers/style/","page":"Style Guide","title":"Style Guide","text":"Break the tests into multiple files, with one module per file, so that subsets of the codebase can be tested by calling include with the relevant file.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Test/overview.md\"","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Test/overview/#test_module","page":"Overview","title":"The Test submodule","text":"","category":"section"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"The Test submodule provides tools to help solvers implement unit tests in order to ensure they implement the MathOptInterface API correctly, and to check for solver-correctness.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"We use a centralized repository of tests, so that if we find a bug in one solver, instead of adding a test to that particular repository, we add it here so that all solvers can benefit.","category":"page"},{"location":"moi/submodules/Test/overview/#How-to-test-a-solver","page":"Overview","title":"How to test a solver","text":"","category":"section"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"The skeleton below can be used for the wrapper test file of a solver named FooBar.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"# ============================ /test/MOI_wrapper.jl ============================\nmodule TestFooBar\n\nimport FooBar\nusing Test\n\nimport MathOptInterface as MOI\n\nconst OPTIMIZER = MOI.instantiate(\n    MOI.OptimizerWithAttributes(FooBar.Optimizer, MOI.Silent() => true),\n)\n\nconst BRIDGED = MOI.instantiate(\n    MOI.OptimizerWithAttributes(FooBar.Optimizer, MOI.Silent() => true),\n    with_bridge_type = Float64,\n)\n\n# See the docstring of MOI.Test.Config for other arguments.\nconst CONFIG = MOI.Test.Config(\n    # Modify tolerances as necessary.\n    atol = 1e-6,\n    rtol = 1e-6,\n    # Use MOI.LOCALLY_SOLVED for local solvers.\n    optimal_status = MOI.OPTIMAL,\n    # Pass attributes or MOI functions to `exclude` to skip tests that\n    # rely on this functionality.\n    exclude = Any[MOI.VariableName, MOI.delete],\n)\n\n\"\"\"\n    runtests()\n\nThis function runs all functions in the this Module starting with `test_`.\n\"\"\"\nfunction runtests()\n    for name in names(@__MODULE__; all = true)\n        if startswith(\"$(name)\", \"test_\")\n            @testset \"$(name)\" begin\n                getfield(@__MODULE__, name)()\n            end\n        end\n    end\nend\n\n\"\"\"\n    test_runtests()\n\nThis function runs all the tests in MathOptInterface.Test.\n\nPass arguments to `exclude` to skip tests for functionality that is not\nimplemented or that your solver doesn't support.\n\"\"\"\nfunction test_runtests()\n    MOI.Test.runtests(\n        BRIDGED,\n        CONFIG,\n        exclude = [\n            \"test_attribute_NumberOfThreads\",\n            \"test_quadratic_\",\n        ],\n        # This argument is useful to prevent tests from failing on future\n        # releases of MOI that add new tests. Don't let this number get too far\n        # behind the current MOI release though. You should periodically check\n        # for new tests to fix bugs and implement new features.\n        exclude_tests_after = v\"0.10.5\",\n    )\n    return\nend\n\n\"\"\"\n    test_SolverName()\n\nYou can also write new tests for solver-specific functionality. Write each new\ntest as a function with a name beginning with `test_`.\n\"\"\"\nfunction test_SolverName()\n    @test MOI.get(FooBar.Optimizer(), MOI.SolverName()) == \"FooBar\"\n    return\nend\n\nend # module TestFooBar\n\n# This line at tne end of the file runs all the tests!\nTestFooBar.runtests()","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Then modify your runtests.jl file to include the MOI_wrapper.jl file:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"# ============================ /test/runtests.jl ============================\n\nusing Test\n\n@testset \"MOI\" begin\n    include(\"test/MOI_wrapper.jl\")\nend","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"info: Info\nThe optimizer BRIDGED constructed with instantiate automatically bridges constraints that are not supported by OPTIMIZER using the bridges listed in Bridges. It is recommended for an implementation of MOI to only support constraints that are natively supported by the solver and let bridges transform the constraint to the appropriate form. For this reason it is expected that tests may not pass if OPTIMIZER is used instead of BRIDGED.","category":"page"},{"location":"moi/submodules/Test/overview/#How-to-debug-a-failing-test","page":"Overview","title":"How to debug a failing test","text":"","category":"section"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"When writing a solver, it's likely that you will initially fail many tests. Some failures will be bugs, but other failures you may choose to exclude.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"There are two ways to exclude tests:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Exclude tests whose names contain a string using:\nMOI.Test.runtests(\n    model,\n    config;\n    exclude = String[\"test_to_exclude\", \"test_conic_\"],\n)\nThis will exclude tests whose name contains either of the two strings provided.\nExclude tests which rely on specific functionality using:\nMOI.Test.Config(exclude = Any[MOI.VariableName, MOI.optimize!])\nThis will exclude tests which use the MOI.VariableName attribute, or which call MOI.optimize!.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Each test that fails can be independently called as:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"model = FooBar.Optimizer()\nconfig = MOI.Test.Config()\nMOI.empty!(model)\nMOI.Test.test_category_name_that_failed(model, config)","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"You can look-up the source code of the test that failed by searching for it in the src/Test/test_category.jl file.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"tip: Tip\nEach test function also has a docstring that explains what the test is for. Use ? MOI.Test.test_category_name_that_failed from the REPL to read it.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Periodically, you should re-run excluded tests to see if they now pass. The easiest way to do this is to swap the exclude keyword argument of runtests to include. For example:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"MOI.Test.runtests(\n    model,\n    config;\n    exclude = String[\"test_to_exclude\", \"test_conic_\"],\n)","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"becomes","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"MOI.Test.runtests(\n    model,\n    config;\n    include = String[\"test_to_exclude\", \"test_conic_\"],\n)","category":"page"},{"location":"moi/submodules/Test/overview/#How-to-add-a-test","page":"Overview","title":"How to add a test","text":"","category":"section"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"To detect bugs in solvers, we add new tests to MOI.Test.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"As an example, ECOS errored calling optimize! twice in a row. (See ECOS.jl PR #72.) We could add a test to ECOS.jl, but that would only stop us from re-introducing the bug to ECOS.jl in the future, but it would not catch other solvers in the ecosystem with the same bug. Instead, if we add a test to MOI.Test, then all solvers will also check that they handle a double optimize call.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"For this test, we care about correctness, rather than performance. therefore, we don't expect solvers to efficiently decide that they have already solved the problem, only that calling optimize! twice doesn't throw an error or give the wrong answer.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Step 1","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Install the MathOptInterface julia package in dev mode:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"julia> ]\n(@v1.6) pkg> dev MathOptInterface","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Step 2","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"From here on, proceed with making the following changes in the ~/.julia/dev/MathOptInterface folder (or equivalent dev path on your machine).","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Step 3","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Since the double-optimize error involves solving an optimization problem, add a new test to src/Test/test_solve.jl:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"\"\"\"\n    test_unit_optimize!_twice(model::MOI.ModelLike, config::Config)\n\nTest that calling `MOI.optimize!` twice does not error.\n\nThis problem was first detected in ECOS.jl PR#72:\nhttps://github.com/jump-dev/ECOS.jl/pull/72\n\"\"\"\nfunction test_unit_optimize!_twice(\n    model::MOI.ModelLike,\n    config::Config{T},\n) where {T}\n    # Use the `@requires` macro to check conditions that the test function\n    # requires to run. Models failing this `@requires` check will silently skip\n    # the test.\n    @requires MOI.supports_constraint(\n        model,\n        MOI.VariableIndex,\n        MOI.GreaterThan{Float64},\n    )\n    @requires _supports(config, MOI.optimize!)\n    # If needed, you can test that the model is empty at the start of the test.\n    # You can assume that this will be the case for tests run via `runtests`.\n    # User's calling tests individually need to call `MOI.empty!` themselves.\n    @test MOI.is_empty(model)\n    # Create a simple model. Try to make this as simple as possible so that the\n    # majority of solvers can run the test.\n    x = MOI.add_variable(model)\n    MOI.add_constraint(model, x, MOI.GreaterThan(one(T)))\n    MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)\n    MOI.set(\n        model,\n        MOI.ObjectiveFunction{MOI.VariableIndex}(),\n        x,\n    )\n    # The main component of the test: does calling `optimize!` twice error?\n    MOI.optimize!(model)\n    MOI.optimize!(model)\n    # Check we have a solution.\n    @test MOI.get(model, MOI.TerminationStatus()) == MOI.OPTIMAL\n    # There is a three-argument version of `Base.isapprox` for checking\n    # approximate equality based on the tolerances defined in `config`:\n    @test isapprox(MOI.get(model, MOI.VariablePrimal(), x), one(T), config)\n    # For code-style, these tests should always `return` `nothing`.\n    return\nend","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"info: Info\nMake sure the function is agnostic to the number type T; don't assume it is a Float64 capable solver.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"We also need to write a test for the test. Place this function immediately below the test you just wrote in the same file:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"function setup_test(\n    ::typeof(test_unit_optimize!_twice),\n    model::MOI.Utilities.MockOptimizer,\n    ::Config,\n)\n    MOI.Utilities.set_mock_optimize!(\n        model,\n        (mock::MOI.Utilities.MockOptimizer) -> MOIU.mock_optimize!(\n            mock,\n            MOI.OPTIMAL,\n            (MOI.FEASIBLE_POINT, [1.0]),\n        ),\n    )\n    return\nend","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Finally, you also need to implement Test.version_added. If we added this test when the latest released version of MOI was v0.10.5, define:","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"version_added(::typeof(test_unit_optimize!_twice)) = v\"0.10.6\"","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Step 6","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"Commit the changes to git from ~/.julia/dev/MathOptInterface and submit the PR for review.","category":"page"},{"location":"moi/submodules/Test/overview/","page":"Overview","title":"Overview","text":"tip: Tip\nIf you need help writing a test, open an issue on GitHub, or ask the Developer Chatroom.","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Utilities/reference.md\"","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Utilities/reference/#Utilities.Model","page":"API Reference","title":"Utilities.Model","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.Model","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.Model","page":"API Reference","title":"MathOptInterface.Utilities.Model","text":"MOI.Utilities.Model{T}() where {T}\n\nAn implementation of ModelLike that supports all functions and sets defined in MOI. It is parameterized by the coefficient type.\n\nExamples\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#Utilities.UniversalFallback","page":"API Reference","title":"Utilities.UniversalFallback","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.UniversalFallback","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.UniversalFallback","page":"API Reference","title":"MathOptInterface.Utilities.UniversalFallback","text":"UniversalFallback\n\nThe UniversalFallback can be applied on a MOI.ModelLike model to create the model UniversalFallback(model) supporting any constraint and attribute. This allows to have a specialized implementation in model for performance critical constraints and attributes while still supporting other attributes with a small performance penalty. Note that model is unaware of constraints and attributes stored by UniversalFallback so this is not appropriate if model is an optimizer (for this reason, MOI.optimize! has not been implemented). In that case, optimizer bridges should be used instead.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#Utilities.@model","page":"API Reference","title":"Utilities.@model","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.@model\nUtilities.GenericModel\nUtilities.GenericOptimizer","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.@model","page":"API Reference","title":"MathOptInterface.Utilities.@model","text":"macro model(\n    model_name,\n    scalar_sets,\n    typed_scalar_sets,\n    vector_sets,\n    typed_vector_sets,\n    scalar_functions,\n    typed_scalar_functions,\n    vector_functions,\n    typed_vector_functions,\n    is_optimizer = false\n)\n\nCreates a type model_name implementing the MOI model interface and supporting all combinations of the provided functions and sets.\n\nEach typed_ scalar/vector sets/functions argument is a tuple of types. A type is \"typed\" if it has a coefficient {T} as the first type parameter.\n\nTuple syntax\n\nTo give no set/function, write (). To give one set or function X, write (X,).\n\nis_optimizer\n\nIf is_optimizer = true, the resulting struct is a of GenericOptimizer, which is a subtype of MOI.AbstractOptimizer, otherwise, it is a GenericModel, which is a subtype of MOI.ModelLike.\n\nVariableIndex\n\nThe function MOI.VariableIndex must not be given in scalar_functions.\nThe model supports MOI.VariableIndex-in-S constraints where S is MOI.EqualTo, MOI.GreaterThan, MOI.LessThan, MOI.Interval, MOI.Integer, MOI.ZeroOne, MOI.Semicontinuous or MOI.Semiinteger.\nThe sets supported with MOI.VariableIndex cannot be controlled from the macro; use UniversalFallback to support more sets.\n\nExamples\n\nThe model describing a linear program would be:\n\n@model(\n    LPModel,                                          # model_name\n    (),                                               # untyped scalar sets\n    (MOI.EqualTo, MOI.GreaterThan, MOI.LessThan, MOI.Interval), #   typed scalar sets\n    (MOI.Zeros, MOI.Nonnegatives, MOI.Nonpositives),  # untyped vector sets\n    (),                                               #   typed vector sets\n    (),                                               # untyped scalar functions\n    (MOI.ScalarAffineFunction,),                      #   typed scalar functions\n    (MOI.VectorOfVariables,),                         # untyped vector functions\n    (MOI.VectorAffineFunction,),                      #   typed vector functions\n    false,                                            # is_optimizer\n)\n\n\n\n\n\n","category":"macro"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.GenericModel","page":"API Reference","title":"MathOptInterface.Utilities.GenericModel","text":"mutable struct GenericModel{T,O,V,C} <: AbstractModelLike{T}\n\nImplements a model supporting coefficients of type T and:\n\nAn objective function stored in .objective::O\nVariables and VariableIndex constraints stored in .variable_bounds::V\nF-in-S constraints (excluding VariableIndex constraints) stored in .constraints::C\n\nAll interactions take place via the MOI interface, so the types O, V, and C must implement the API as needed for their functionality.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.GenericOptimizer","page":"API Reference","title":"MathOptInterface.Utilities.GenericOptimizer","text":"mutable struct GenericOptimizer{T,O,V,C} <: AbstractOptimizer{T}\n\nImplements a model supporting coefficients of type T and:\n\nAn objective function stored in .objective::O\nVariables and VariableIndex constraints stored in .variable_bounds::V\nF-in-S constraints (excluding VariableIndex constraints) stored in .constraints::C\n\nAll interactions take place via the MOI interface, so the types O, V, and C must implement the API as needed for their functionality.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#.objective","page":"API Reference","title":".objective","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.ObjectiveContainer","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ObjectiveContainer","page":"API Reference","title":"MathOptInterface.Utilities.ObjectiveContainer","text":"ObjectiveContainer{T}\n\nA helper struct to simplify the handling of objective functions in Utilities.Model.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#.variables","page":"API Reference","title":".variables","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.VariablesContainer\nUtilities.FreeVariables","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.VariablesContainer","page":"API Reference","title":"MathOptInterface.Utilities.VariablesContainer","text":"struct VariablesContainer{T} <: AbstractVectorBounds\n    set_mask::Vector{UInt16}\n    lower::Vector{T}\n    upper::Vector{T}\nend\n\nA struct for storing variables and VariableIndex-related constraints. Used in MOI.Utilities.Model by default.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.FreeVariables","page":"API Reference","title":"MathOptInterface.Utilities.FreeVariables","text":"mutable struct FreeVariables <: MOI.ModelLike\n    n::Int64\n    FreeVariables() = new(0)\nend\n\nA struct for storing free variables that can be used as the variables field of GenericModel or GenericModel. It represents a model that does not support any constraint nor objective function.\n\nExample\n\nThe following model type represents a conic model in geometric form. As opposed to VariablesContainer, FreeVariables does not support constraint bounds so they are bridged into an affine constraint in the MOI.Nonnegatives cone as expected for the geometric conic form.\n\njulia> MOI.Utilities.@product_of_sets(\n    Cones,\n    MOI.Zeros,\n    MOI.Nonnegatives,\n    MOI.SecondOrderCone,\n    MOI.PositiveSemidefiniteConeTriangle,\n);\n\njulia> const ConicModel{T} = MOI.Utilities.GenericOptimizer{\n    T,\n    MOI.Utilities.ObjectiveContainer{T},\n    MOI.Utilities.FreeVariables,\n    MOI.Utilities.MatrixOfConstraints{\n        T,\n        MOI.Utilities.MutableSparseMatrixCSC{\n            T,\n            Int,\n            MOI.Utilities.OneBasedIndexing,\n        },\n        Vector{T},\n        Cones{T},\n    },\n};\n\njulia> model = MOI.instantiate(ConicModel{Float64}, with_bridge_type=Float64);\n\njulia> x = MOI.add_variable(model)\nMathOptInterface.VariableIndex(1)\n\njulia> c = MOI.add_constraint(model, x, MOI.GreaterThan(1.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(1)\n\njulia> MOI.Bridges.is_bridged(model, c)\ntrue\n\njulia> bridge = MOI.Bridges.bridge(model, c)\nMathOptInterface.Bridges.Constraint.VectorizeBridge{Float64, MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives, MathOptInterface.VariableIndex}(MathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1), 1.0)\n\njulia> bridge.vector_constraint\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1)\n\njulia> MOI.Bridges.is_bridged(model, bridge.vector_constraint)\nfalse\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#.constraints","page":"API Reference","title":".constraints","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.VectorOfConstraints\nUtilities.StructOfConstraints\nUtilities.@struct_of_constraints_by_function_types\nUtilities.@struct_of_constraints_by_set_types\nUtilities.struct_of_constraint_code","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.VectorOfConstraints","page":"API Reference","title":"MathOptInterface.Utilities.VectorOfConstraints","text":"mutable struct VectorOfConstraints{\n    F<:MOI.AbstractFunction,\n    S<:MOI.AbstractSet,\n} <: MOI.ModelLike\n    constraints::CleverDicts.CleverDict{\n        MOI.ConstraintIndex{F,S},\n        Tuple{F,S},\n        typeof(CleverDicts.key_to_index),\n        typeof(CleverDicts.index_to_key),\n    }\nend\n\nA struct storing F-in-S constraints as a mapping between the constraint indices to the corresponding tuple of function and set.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.StructOfConstraints","page":"API Reference","title":"MathOptInterface.Utilities.StructOfConstraints","text":"abstract type StructOfConstraints <: MOI.ModelLike end\n\nA struct storing a subfields other structs storing constraints of different types.\n\nSee Utilities.@struct_of_constraints_by_function_types and Utilities.@struct_of_constraints_by_set_types.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.@struct_of_constraints_by_function_types","page":"API Reference","title":"MathOptInterface.Utilities.@struct_of_constraints_by_function_types","text":"Utilities.@struct_of_constraints_by_function_types(name, func_types...)\n\nGiven a vector of n function types (F1, F2,..., Fn) in func_types, defines a subtype of StructOfConstraints of name name and which type parameters {T, C1, C2, ..., Cn}. It contains n field where the ith field has type Ci and stores the constraints of function type Fi.\n\nThe expression Fi can also be a union in which case any constraint for which the function type is in the union is stored in the field with type Ci.\n\n\n\n\n\n","category":"macro"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.@struct_of_constraints_by_set_types","page":"API Reference","title":"MathOptInterface.Utilities.@struct_of_constraints_by_set_types","text":"Utilities.@struct_of_constraints_by_set_types(name, func_types...)\n\nGiven a vector of n set types (S1, S2,..., Sn) in func_types, defines a subtype of StructOfConstraints of name name and which type parameters {T, C1, C2, ..., Cn}. It contains n field where the ith field has type Ci and stores the constraints of set type Si. The expression Si can also be a union in which case any constraint for which the set type is in the union is stored in the field with type Ci. This can be useful if Ci is a MatrixOfConstraints in order to concatenate the coefficients of constraints of several different set types in the same matrix.\n\n\n\n\n\n","category":"macro"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.struct_of_constraint_code","page":"API Reference","title":"MathOptInterface.Utilities.struct_of_constraint_code","text":"struct_of_constraint_code(struct_name, types, field_types = nothing)\n\nGiven a vector of n Union{SymbolFun,_UnionSymbolFS{SymbolFun}} or Union{SymbolSet,_UnionSymbolFS{SymbolSet}} in types, defines a subtype of StructOfConstraints of name name and which type parameters {T, F1, F2, ..., Fn} if field_types is nothing and a {T} otherwise. It contains n field where the ith field has type Ci if field_types is nothing and type field_types[i] otherwise. If types is vector of Union{SymbolFun,_UnionSymbolFS{SymbolFun}} (resp. Union{SymbolSet,_UnionSymbolFS{SymbolSet}}) then the constraints of that function (resp. set) type are stored in the corresponding field.\n\nThis function is used by the macros @model, @struct_of_constraints_by_function_types and @struct_of_constraints_by_set_types.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#Caching-optimizer","page":"API Reference","title":"Caching optimizer","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.CachingOptimizer\nUtilities.attach_optimizer\nUtilities.reset_optimizer\nUtilities.drop_optimizer\nUtilities.state\nUtilities.mode","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.CachingOptimizer","page":"API Reference","title":"MathOptInterface.Utilities.CachingOptimizer","text":"CachingOptimizer\n\nCachingOptimizer is an intermediate layer that stores a cache of the model and links it with an optimizer. It supports incremental model construction and modification even when the optimizer doesn't.\n\nConstructors\n\n    CachingOptimizer(cache::MOI.ModelLike, optimizer::AbstractOptimizer)\n\nCreates a CachingOptimizer in AUTOMATIC mode, with the optimizer optimizer.\n\nThe type of the optimizer returned is CachingOptimizer{typeof(optimizer), typeof(cache)} so it does not support the function reset_optimizer(::CachingOptimizer, new_optimizer) if the type of new_optimizer is different from the type of optimizer.\n\n    CachingOptimizer(cache::MOI.ModelLike, mode::CachingOptimizerMode)\n\nCreates a CachingOptimizer in the NO_OPTIMIZER state and mode mode.\n\nThe type of the optimizer returned is CachingOptimizer{MOI.AbstractOptimizer,typeof(cache)} so it does support the function reset_optimizer(::CachingOptimizer, new_optimizer) if the type of new_optimizer is different from the type of optimizer.\n\nAbout the type\n\nStates\n\nA CachingOptimizer may be in one of three possible states (CachingOptimizerState):\n\nNO_OPTIMIZER: The CachingOptimizer does not have any optimizer.\nEMPTY_OPTIMIZER: The CachingOptimizer has an empty optimizer. The optimizer is not synchronized with the cached model.\nATTACHED_OPTIMIZER: The CachingOptimizer has an optimizer, and it is synchronized with the cached model.\n\nModes\n\nA CachingOptimizer has two modes of operation (CachingOptimizerMode):\n\nMANUAL: The only methods that change the state of the CachingOptimizer are Utilities.reset_optimizer, Utilities.drop_optimizer, and Utilities.attach_optimizer. Attempting to perform an operation in the incorrect state results in an error.\nAUTOMATIC: The CachingOptimizer changes its state when necessary. For example, optimize! will automatically call attach_optimizer (an optimizer must have been previously set). Attempting to add a constraint or perform a modification not supported by the optimizer results in a drop to EMPTY_OPTIMIZER mode.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.attach_optimizer","page":"API Reference","title":"MathOptInterface.Utilities.attach_optimizer","text":"attach_optimizer(model::CachingOptimizer)\n\nAttaches the optimizer to model, copying all model data into it. Can be called only from the EMPTY_OPTIMIZER state. If the copy succeeds, the CachingOptimizer will be in state ATTACHED_OPTIMIZER after the call, otherwise an error is thrown; see MOI.copy_to for more details on which errors can be thrown.\n\n\n\n\n\nMOIU.attach_optimizer(model::GenericModel)\n\nCall MOIU.attach_optimizer on the backend of model.\n\nCannot be called in direct mode.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.reset_optimizer","page":"API Reference","title":"MathOptInterface.Utilities.reset_optimizer","text":"reset_optimizer(m::CachingOptimizer, optimizer::MOI.AbstractOptimizer)\n\nSets or resets m to have the given empty optimizer optimizer.\n\nCan be called from any state. An assertion error will be thrown if optimizer is not empty.\n\nThe CachingOptimizer m will be in state EMPTY_OPTIMIZER after the call.\n\n\n\n\n\nreset_optimizer(m::CachingOptimizer)\n\nDetaches and empties the current optimizer. Can be called from ATTACHED_OPTIMIZER or EMPTY_OPTIMIZER state. The CachingOptimizer will be in state EMPTY_OPTIMIZER after the call.\n\n\n\n\n\nMOIU.reset_optimizer(model::GenericModel, optimizer::MOI.AbstractOptimizer)\n\nCall MOIU.reset_optimizer on the backend of model.\n\nCannot be called in direct mode.\n\n\n\n\n\nMOIU.reset_optimizer(model::GenericModel)\n\nCall MOIU.reset_optimizer on the backend of model.\n\nCannot be called in direct mode.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.drop_optimizer","page":"API Reference","title":"MathOptInterface.Utilities.drop_optimizer","text":"drop_optimizer(m::CachingOptimizer)\n\nDrops the optimizer, if one is present. Can be called from any state. The CachingOptimizer will be in state NO_OPTIMIZER after the call.\n\n\n\n\n\nMOIU.drop_optimizer(model::GenericModel)\n\nCall MOIU.drop_optimizer on the backend of model.\n\nCannot be called in direct mode.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.state","page":"API Reference","title":"MathOptInterface.Utilities.state","text":"state(m::CachingOptimizer)::CachingOptimizerState\n\nReturns the state of the CachingOptimizer m. See Utilities.CachingOptimizer.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.mode","page":"API Reference","title":"MathOptInterface.Utilities.mode","text":"mode(m::CachingOptimizer)::CachingOptimizerMode\n\nReturns the operating mode of the CachingOptimizer m. See Utilities.CachingOptimizer.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#Mock-optimizer","page":"API Reference","title":"Mock optimizer","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.MockOptimizer","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.MockOptimizer","page":"API Reference","title":"MathOptInterface.Utilities.MockOptimizer","text":"MockOptimizer\n\nMockOptimizer is a fake optimizer especially useful for testing. Its main feature is that it can store the values that should be returned for each attribute.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#Printing","page":"API Reference","title":"Printing","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.latex_formulation","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.latex_formulation","page":"API Reference","title":"MathOptInterface.Utilities.latex_formulation","text":"latex_formulation(model::MOI.ModelLike; kwargs...)\n\nWrap model in a type so that it can be pretty-printed as text/latex in a notebook like IJulia, or in Documenter.\n\nTo render the model, end the cell with latex_formulation(model), or call display(latex_formulation(model)) in to force the display of the model from inside a function.\n\nPossible keyword arguments are:\n\nsimplify_coefficients : Simplify coefficients if possible by omitting them or removing trailing zeros.\ndefault_name : The name given to variables with an empty name.\nprint_types : Print the MOI type of each function and set for clarity.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#Copy-utilities","page":"API Reference","title":"Copy utilities","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.default_copy_to\nUtilities.IndexMap\nUtilities.identity_index_map\nUtilities.ModelFilter","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.default_copy_to","page":"API Reference","title":"MathOptInterface.Utilities.default_copy_to","text":"default_copy_to(dest::MOI.ModelLike, src::MOI.ModelLike)\n\nA default implementation of MOI.copy_to(dest, src) for models that implement the incremental interface, i.e., MOI.supports_incremental_interface returns true.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.IndexMap","page":"API Reference","title":"MathOptInterface.Utilities.IndexMap","text":"IndexMap()\n\nThe dictionary-like object returned by MOI.copy_to.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.identity_index_map","page":"API Reference","title":"MathOptInterface.Utilities.identity_index_map","text":"identity_index_map(model::MOI.ModelLike)\n\nReturn an IndexMap that maps all variable and constraint indices of model to themselves.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ModelFilter","page":"API Reference","title":"MathOptInterface.Utilities.ModelFilter","text":"ModelFilter(filter::Function, model::MOI.ModelLike)\n\nA layer to filter out various components of model.\n\nThe filter function takes a single argument, which is each element from the list returned by the attributes below. It returns true if the element should be visible in the filtered model and false otherwise.\n\nThe components that are filtered are:\n\nEntire constraint types via:\nMOI.ListOfConstraintTypesPresent\nIndividual constraints via:\nMOI.ListOfConstraintIndices{F,S}\nSpecific attributes via:\nMOI.ListOfModelAttributesSet\nMOI.ListOfConstraintAttributesSet\nMOI.ListOfVariableAttributesSet\n\nwarning: Warning\nThe list of attributes filtered may change in a future release. You should write functions that are generic and not limited to the five types listed above. Thus, you should probably define a fallback filter(::Any) = true.\n\nSee below for examples of how this works.\n\nnote: Note\nThis layer has a limited scope. It is intended by be used in conjunction with MOI.copy_to.\n\nExample: copy model excluding integer constraints\n\nUse the do syntax to provide a single function.\n\nfiltered_src = MOI.Utilities.ModelFilter(src) do item\n    return item != (MOI.VariableIndex, MOI.Integer)\nend\nMOI.copy_to(dest, filtered_src)\n\nExample: copy model excluding names\n\nUse type dispatch to simplify the implementation:\n\nmy_filter(::Any) = true  # Note the generic fallback!\nmy_filter(::MOI.VariableName) = false\nmy_filter(::MOI.ConstraintName) = false\nfiltered_src = MOI.Utilities.ModelFilter(my_filter, src)\nMOI.copy_to(dest, filtered_src)\n\nExample: copy irreducible infeasible subsystem\n\nmy_filter(::Any) = true  # Note the generic fallback!\nfunction my_filter(ci::MOI.ConstraintIndex)\n    status = MOI.get(dest, MOI.ConstraintConflictStatus(), ci)\n    return status != MOI.NOT_IN_CONFLICT\nend\nfiltered_src = MOI.Utilities.ModelFilter(my_filter, src)\nMOI.copy_to(dest, filtered_src)\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#Penalty-relaxation","page":"API Reference","title":"Penalty relaxation","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.PenaltyRelaxation\nUtilities.ScalarPenaltyRelaxation","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.PenaltyRelaxation","page":"API Reference","title":"MathOptInterface.Utilities.PenaltyRelaxation","text":"PenaltyRelaxation(\n    penalties = Dict{MOI.ConstraintIndex,Float64}();\n    default::Union{Nothing,T} = 1.0,\n)\n\nA problem modifier that, when passed to MOI.modify, destructively modifies the model in-place to create a penalized relaxation of the constraints.\n\nwarning: Warning\nThis is a destructive routine that modifies the model in-place. If you don't want to modify the original model, use JuMP.copy_model to create a copy before calling MOI.modify.\n\nReformulation\n\nSee Utilities.ScalarPenaltyRelaxation for details of the reformulation.\n\nFor each constraint ci, the penalty passed to Utilities.ScalarPenaltyRelaxation is get(penalties, ci, default). If the value is nothing, because ci does not exist in penalties and default = nothing, then the constraint is skipped.\n\nReturn value\n\nMOI.modify(model, PenaltyRelaxation()) returns a Dict{MOI.ConstraintIndex,MOI.ScalarAffineFunction} that maps each constraint index to the corresponding y + z as a MOI.ScalarAffineFunction. In an optimal solution, query the value of these functions to compute the violation of each constraint.\n\nRelax a subset of constraints\n\nTo relax a subset of constraints, pass a penalties dictionary and set default = nothing.\n\nSupported constraint types\n\nThe penalty relaxation is currently limited to modifying MOI.ScalarAffineFunction and MOI.ScalarQuadraticFunction constraints in the linear sets MOI.LessThan, MOI.GreaterThan, MOI.EqualTo and MOI.Interval.\n\nIt does not include variable bound or integrality constraints, because these cannot be modified in-place.\n\nTo modify variable bounds, rewrite them as linear constraints.\n\nExamples\n\njulia> model = MOI.Utilities.Model{Float64}();\n\njulia> x = MOI.add_variable(model);\n\njulia> c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));\n\njulia> map = MOI.modify(model, MOI.Utilities.PenaltyRelaxation(default = 2.0));\n\njulia> print(model)\nMinimize ScalarAffineFunction{Float64}:\n 0.0 + 2.0 v[2]\n\nSubject to:\n\nScalarAffineFunction{Float64}-in-LessThan{Float64}\n 0.0 + 1.0 v[1] - 1.0 v[2] <= 2.0\n\nVariableIndex-in-GreaterThan{Float64}\n v[2] >= 0.0\n\njulia> map[c] isa MOI.ScalarAffineFunction{Float64}\ntrue\n\njulia> model = MOI.Utilities.Model{Float64}();\n\njulia> x = MOI.add_variable(model);\n\njulia> c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));\n\njulia> map = MOI.modify(model, MOI.Utilities.PenaltyRelaxation(Dict(c => 3.0)));\n\njulia> print(model)\nMinimize ScalarAffineFunction{Float64}:\n 0.0 + 3.0 v[2]\n\nSubject to:\n\nScalarAffineFunction{Float64}-in-LessThan{Float64}\n 0.0 + 1.0 v[1] - 1.0 v[2] <= 2.0\n\nVariableIndex-in-GreaterThan{Float64}\n v[2] >= 0.0\n\njulia> map[c] isa MOI.ScalarAffineFunction{Float64}\ntrue\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ScalarPenaltyRelaxation","page":"API Reference","title":"MathOptInterface.Utilities.ScalarPenaltyRelaxation","text":"ScalarPenaltyRelaxation(penalty::T) where {T}\n\nA problem modifier that, when passed to MOI.modify, destructively modifies the constraint in-place to create a penalized relaxation of the constraint.\n\nwarning: Warning\nThis is a destructive routine that modifies the constraint in-place. If you don't want to modify the original model, use JuMP.copy_model to create a copy before calling MOI.modify.\n\nReformulation\n\nThe penalty relaxation modifies constraints of the form f(x) in S into f(x) + y - z in S, where y z ge 0, and then it introduces a penalty term into the objective of a times (y + z) (if minimizing, else -a), where a is penalty\n\nWhen S is MOI.LessThan or MOI.GreaterThan, we omit y or z respectively as a performance optimization.\n\nReturn value\n\nMOI.modify(model, ci, ScalarPenaltyRelaxation(penalty)) returns y + z as a MOI.ScalarAffineFunction. In an optimal solution, query the value of this function to compute the violation of the constraint.\n\nExamples\n\njulia> model = MOI.Utilities.Model{Float64}();\n\njulia> x = MOI.add_variable(model);\n\njulia> c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));\n\njulia> f = MOI.modify(model, c, MOI.Utilities.ScalarPenaltyRelaxation(2.0));\n\njulia> print(model)\nMinimize ScalarAffineFunction{Float64}:\n 0.0 + 2.0 v[2]\n\nSubject to:\n\nScalarAffineFunction{Float64}-in-LessThan{Float64}\n 0.0 + 1.0 v[1] - 1.0 v[2] <= 2.0\n\nVariableIndex-in-GreaterThan{Float64}\n v[2] >= 0.0\n\njulia> f isa MOI.ScalarAffineFunction{Float64}\ntrue\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MatrixOfConstraints","page":"API Reference","title":"MatrixOfConstraints","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.MatrixOfConstraints","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.MatrixOfConstraints","page":"API Reference","title":"MathOptInterface.Utilities.MatrixOfConstraints","text":"mutable struct MatrixOfConstraints{T,AT,BT,ST} <: MOI.ModelLike\n    coefficients::AT\n    constants::BT\n    sets::ST\n    caches::Vector{Any}\n    are_indices_mapped::Vector{BitSet}\n    final_touch::Bool\nend\n\nRepresent ScalarAffineFunction and VectorAffinefunction constraints in a matrix form where the linear coefficients of the functions are stored in the coefficients field, the constants of the functions or sets are stored in the constants field. Additional information about the sets are stored in the sets field.\n\nThis model can only be used as the constraints field of a MOI.Utilities.AbstractModel.\n\nWhen the constraints are added, they are stored in the caches field. They are only loaded in the coefficients and constants fields once MOI.Utilities.final_touch is called. For this reason, MatrixOfConstraints should not be used by an incremental interface. Use MOI.copy_to instead.\n\nThe constraints can be added in two different ways:\n\nWith add_constraint, in which case a canonicalized copy of the function is stored in caches.\nWith pass_nonvariable_constraints, in which case the functions and sets are stored themselves in caches without mapping the variable indices. The corresponding index in caches is added in are_indices_mapped. This avoids doing a copy of the function in case the getter of CanonicalConstraintFunction does not make a copy for the source model, e.g., this is the case of VectorOfConstraints.\n\nWe illustrate this with an example. Suppose a model is copied from a src::MOI.Utilities.Model to a bridged model with a MatrixOfConstraints. For all the types that are not bridged, the constraints will be copied with pass_nonvariable_constraints. Hence the functions stored in caches are exactly the same as the ones stored in src. This is ok since this is only during the copy_to operation during which src cannot be modified. On the other hand, for the types that are bridged, the functions added may contain duplicates even if the functions did not contain duplicates in src so duplicates are removed with MOI.Utilities.canonical.\n\nInterface\n\nThe .coefficients::AT type must implement:\n\nAT()\nMOI.empty(::AT)!\nMOI.Utilities.add_column\nMOI.Utilities.set_number_of_rows\nMOI.Utilities.allocate_terms\nMOI.Utilities.load_terms\nMOI.Utilities.final_touch\n\nThe .constants::BT type must implement:\n\nBT()\nBase.empty!(::BT)\nBase.resize(::BT)\nMOI.Utilities.load_constants\nMOI.Utilities.function_constants\nMOI.Utilities.set_from_constants\n\nThe .sets::ST type must implement:\n\nST()\nMOI.is_empty(::ST)\nMOI.empty(::ST)\nMOI.dimension(::ST)\nMOI.is_valid(::ST, ::MOI.ConstraintIndex)\nMOI.get(::ST, ::MOI.ListOfConstraintTypesPresent)\nMOI.get(::ST, ::MOI.NumberOfConstraints)\nMOI.get(::ST, ::MOI.ListOfConstraintIndices)\nMOI.Utilities.set_types\nMOI.Utilities.set_index\nMOI.Utilities.add_set\nMOI.Utilities.rows\nMOI.Utilities.final_touch\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#.coefficients","page":"API Reference","title":".coefficients","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.add_column\nUtilities.allocate_terms\nUtilities.set_number_of_rows\nUtilities.load_terms\nUtilities.final_touch\nUtilities.extract_function","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.add_column","page":"API Reference","title":"MathOptInterface.Utilities.add_column","text":"add_column(coefficients)::Nothing\n\nTell coefficients to pre-allocate datastructures as needed to store one column.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.allocate_terms","page":"API Reference","title":"MathOptInterface.Utilities.allocate_terms","text":"allocate_terms(coefficients, index_map, func)::Nothing\n\nTell coefficients that the terms of the function func where the variable indices are mapped with index_map will be loaded with load_terms.\n\nThe function func must be canonicalized before calling allocate_terms. See is_canonical.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.set_number_of_rows","page":"API Reference","title":"MathOptInterface.Utilities.set_number_of_rows","text":"set_number_of_rows(coefficients, n)::Nothing\n\nTell coefficients to pre-allocate datastructures as needed to store n rows.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.load_terms","page":"API Reference","title":"MathOptInterface.Utilities.load_terms","text":"load_terms(coefficients, index_map, func, offset)::Nothing\n\nLoads the terms of func to coefficients, mapping the variable indices with index_map.\n\nThe ith dimension of func is loaded at the (offset + i)th row of coefficients.\n\nThe function must be allocated first with allocate_terms.\n\nThe function func must be canonicalized, see is_canonical.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.final_touch","page":"API Reference","title":"MathOptInterface.Utilities.final_touch","text":"final_touch(coefficients)::Nothing\n\nInforms the coefficients that all functions have been added with load_terms. No more modification is allowed unless MOI.empty! is called.\n\nfinal_touch(sets)::Nothing\n\nInforms the sets that all functions have been added with add_set. No more modification is allowed unless MOI.empty! is called.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.extract_function","page":"API Reference","title":"MathOptInterface.Utilities.extract_function","text":"extract_function(coefficients, row::Integer, constant::T) where {T}\n\nReturn the MOI.ScalarAffineFunction{T} function corresponding to row row in coefficients.\n\nextract_function(\n    coefficients,\n    rows::UnitRange,\n    constants::Vector{T},\n) where{T}\n\nReturn the MOI.VectorAffineFunction{T} function corresponding to rows rows in coefficients.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.MutableSparseMatrixCSC\nUtilities.AbstractIndexing\nUtilities.ZeroBasedIndexing\nUtilities.OneBasedIndexing","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.MutableSparseMatrixCSC","page":"API Reference","title":"MathOptInterface.Utilities.MutableSparseMatrixCSC","text":"mutable struct MutableSparseMatrixCSC{Tv,Ti<:Integer,I<:AbstractIndexing}\n    indexing::I\n    m::Int\n    n::Int\n    colptr::Vector{Ti}\n    rowval::Vector{Ti}\n    nzval::Vector{Tv}\n    nz_added::Vector{Ti}\nend\n\nMatrix type loading sparse matrices in the Compressed Sparse Column format. The indexing used is indexing, see AbstractIndexing. The other fields have the same meaning than for SparseArrays.SparseMatrixCSC except that the indexing is different unless indexing is OneBasedIndexing. In addition, nz_added is used to cache the number of non-zero terms that have been added to each column due to the incremental nature of load_terms.\n\nThe matrix is loaded in 5 steps:\n\nMOI.empty! is called.\nMOI.Utilities.add_column and MOI.Utilities.allocate_terms are called in any order.\nMOI.Utilities.set_number_of_rows is called.\nMOI.Utilities.load_terms is called for each affine function.\nMOI.Utilities.final_touch is called.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.AbstractIndexing","page":"API Reference","title":"MathOptInterface.Utilities.AbstractIndexing","text":"abstract type AbstractIndexing end\n\nIndexing to be used for storing the row and column indices of MutableSparseMatrixCSC. See ZeroBasedIndexing and OneBasedIndexing.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ZeroBasedIndexing","page":"API Reference","title":"MathOptInterface.Utilities.ZeroBasedIndexing","text":"struct ZeroBasedIndexing <: AbstractIndexing end\n\nZero-based indexing: the ith row or column has index i - 1. This is useful when the vectors of row and column indices need to be communicated to a library using zero-based indexing such as C libraries.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.OneBasedIndexing","page":"API Reference","title":"MathOptInterface.Utilities.OneBasedIndexing","text":"struct ZeroBasedIndexing <: AbstractIndexing end\n\nOne-based indexing: the ith row or column has index i. This enables an allocation-free conversion of MutableSparseMatrixCSC to SparseArrays.SparseMatrixCSC.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#.constants","page":"API Reference","title":".constants","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.load_constants\nUtilities.function_constants\nUtilities.set_from_constants","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.load_constants","page":"API Reference","title":"MathOptInterface.Utilities.load_constants","text":"load_constants(constants, offset, func_or_set)::Nothing\n\nThis function loads the constants of func_or_set in constants at an offset of offset. Where offset is the sum of the dimensions of the constraints already loaded. The storage should be preallocated with resize! before calling this function.\n\nThis function should be implemented to be usable as storage of constants for MatrixOfConstraints.\n\nThe constants are loaded in three steps:\n\nBase.empty! is called.\nBase.resize! is called with the sum of the dimensions of all constraints.\nMOI.Utilities.load_constants is called for each function for vector constraint or set for scalar constraint.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.function_constants","page":"API Reference","title":"MathOptInterface.Utilities.function_constants","text":"function_constants(constants, rows)\n\nThis function returns the function constants that were loaded with load_constants at the rows rows.\n\nThis function should be implemented to be usable as storage of constants for MatrixOfConstraints.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.set_from_constants","page":"API Reference","title":"MathOptInterface.Utilities.set_from_constants","text":"set_from_constants(constants, S::Type, rows)::S\n\nThis function returns an instance of the set S for which the constants where loaded with load_constants at the rows rows.\n\nThis function should be implemented to be usable as storage of constants for MatrixOfConstraints.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.Hyperrectangle","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.Hyperrectangle","page":"API Reference","title":"MathOptInterface.Utilities.Hyperrectangle","text":"struct Hyperrectangle{T} <: AbstractVectorBounds\n    lower::Vector{T}\n    upper::Vector{T}\nend\n\nA struct for the .constants field in MatrixOfConstraints.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#.sets","page":"API Reference","title":".sets","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.set_index\nUtilities.set_types\nUtilities.add_set\nUtilities.rows\nUtilities.num_rows\nUtilities.set_with_dimension","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.set_index","page":"API Reference","title":"MathOptInterface.Utilities.set_index","text":"set_index(sets, ::Type{S})::Union{Int,Nothing} where {S<:MOI.AbstractSet}\n\nReturn an integer corresponding to the index of the set type in the list given by set_types.\n\nIf S is not part of the list, return nothing.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.set_types","page":"API Reference","title":"MathOptInterface.Utilities.set_types","text":"set_types(sets)::Vector{Type}\n\nReturn the list of the types of the sets allowed in sets.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.add_set","page":"API Reference","title":"MathOptInterface.Utilities.add_set","text":"add_set(sets, i)::Int64\n\nAdd a scalar set of type index i.\n\nadd_set(sets, i, dim)::Int64\n\nAdd a vector set of type index i and dimension dim.\n\nBoth methods return a unique Int64 of the set that can be used to reference this set.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.rows","page":"API Reference","title":"MathOptInterface.Utilities.rows","text":"rows(sets, ci::MOI.ConstraintIndex)::Union{Int,UnitRange{Int}}\n\nReturn the rows in 1:MOI.dimension(sets) corresponding to the set of id ci.value.\n\nFor scalar sets, this returns an Int. For vector sets, this returns an UnitRange{Int}.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.num_rows","page":"API Reference","title":"MathOptInterface.Utilities.num_rows","text":"num_rows(sets::OrderedProductOfSets, ::Type{S}) where {S}\n\nReturn the number of rows corresponding to a set of type S. That is, it is the sum of the dimensions of the sets of type S.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.set_with_dimension","page":"API Reference","title":"MathOptInterface.Utilities.set_with_dimension","text":"set_with_dimension(::Type{S}, dim) where {S<:MOI.AbstractVectorSet}\n\nReturns the instance of S of MOI.dimension dim. This needs to be implemented for sets of type S to be useable with MatrixOfConstraints.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.ProductOfSets\nUtilities.MixOfScalarSets\nUtilities.@mix_of_scalar_sets\nUtilities.OrderedProductOfSets\nUtilities.@product_of_sets","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ProductOfSets","page":"API Reference","title":"MathOptInterface.Utilities.ProductOfSets","text":"abstract type ProductOfSets{T} end\n\nRepresents a cartesian product of sets of given types.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.MixOfScalarSets","page":"API Reference","title":"MathOptInterface.Utilities.MixOfScalarSets","text":"abstract type MixOfScalarSets{T} <: ProductOfSets{T} end\n\nProduct of scalar sets in the order the constraints are added, mixing the constraints of different types.\n\nUse @mix_of_scalar_sets to generate a new subtype.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.@mix_of_scalar_sets","page":"API Reference","title":"MathOptInterface.Utilities.@mix_of_scalar_sets","text":"@mix_of_scalar_sets(name, set_types...)\n\nGenerate a new MixOfScalarSets subtype.\n\nExample\n\n@mix_of_scalar_sets(\n    MixedIntegerLinearProgramSets,\n    MOI.GreaterThan{T},\n    MOI.LessThan{T},\n    MOI.EqualTo{T},\n    MOI.Integer,\n)\n\n\n\n\n\n","category":"macro"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.OrderedProductOfSets","page":"API Reference","title":"MathOptInterface.Utilities.OrderedProductOfSets","text":"abstract type OrderedProductOfSets{T} <: ProductOfSets{T} end\n\nProduct of sets in the order the constraints are added, grouping the constraints of the same types contiguously.\n\nUse @product_of_sets to generate new subtypes.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.@product_of_sets","page":"API Reference","title":"MathOptInterface.Utilities.@product_of_sets","text":"@product_of_sets(name, set_types...)\n\nGenerate a new OrderedProductOfSets subtype.\n\nExample\n\n@product_of_sets(\n    LinearOrthants,\n    MOI.Zeros,\n    MOI.Nonnegatives,\n    MOI.Nonpositives,\n    MOI.ZeroOne,\n)\n\n\n\n\n\n","category":"macro"},{"location":"moi/submodules/Utilities/reference/#Fallbacks","page":"API Reference","title":"Fallbacks","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.get_fallback","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.get_fallback","page":"API Reference","title":"MathOptInterface.Utilities.get_fallback","text":"get_fallback(model::MOI.ModelLike, ::MOI.ObjectiveValue)\n\nCompute the objective function value using the VariablePrimal results and the ObjectiveFunction value.\n\n\n\n\n\nget_fallback(model::MOI.ModelLike, ::MOI.DualObjectiveValue, T::Type)::T\n\nCompute the dual objective value of type T using the ConstraintDual results and the ConstraintFunction and ConstraintSet values. Note that the nonlinear part of the model is ignored.\n\n\n\n\n\nget_fallback(model::MOI.ModelLike, ::MOI.ConstraintPrimal,\n             constraint_index::MOI.ConstraintIndex)\n\nCompute the value of the function of the constraint of index constraint_index using the VariablePrimal results and the ConstraintFunction values.\n\n\n\n\n\nget_fallback(model::MOI.ModelLike, attr::MOI.ConstraintDual,\n             ci::MOI.ConstraintIndex{Union{MOI.VariableIndex,\n                                           MOI.VectorOfVariables}})\n\nCompute the dual of the constraint of index ci using the ConstraintDual of other constraints and the ConstraintFunction values. Throws an error if some constraints are quadratic or if there is one another MOI.VariableIndex-in-S or MOI.VectorOfVariables-in-S constraint with one of the variables in the function of the constraint ci.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#Function-utilities","page":"API Reference","title":"Function utilities","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"The following utilities are available for functions:","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.eval_variables\nUtilities.map_indices\nUtilities.substitute_variables\nUtilities.filter_variables\nUtilities.remove_variable\nUtilities.all_coefficients\nUtilities.unsafe_add\nUtilities.isapprox_zero\nUtilities.modify_function\nUtilities.zero_with_output_dimension","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.eval_variables","page":"API Reference","title":"MathOptInterface.Utilities.eval_variables","text":"eval_variables(value_fn::Function, f::MOI.AbstractFunction)\n\nReturns the value of function f if each variable index vi is evaluated as value_fn(vi).\n\nNote that value_fn must return a Number. See substitute_variables for a similar function where value_fn returns an MOI.AbstractScalarFunction.\n\nwarning: Warning\nThe two-argument version of eval_variables is deprecated and may be removed in MOI v2.0.0. Use the three-argument method eval_variables(::Function, ::MOI.ModelLike, ::MOI.AbstractFunction) instead.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.map_indices","page":"API Reference","title":"MathOptInterface.Utilities.map_indices","text":"map_indices(index_map::Function, attr::MOI.AnyAttribute, x::X)::X where {X}\n\nSubstitute any MOI.VariableIndex (resp. MOI.ConstraintIndex) in x by the MOI.VariableIndex (resp. MOI.ConstraintIndex) of the same type given by index_map(x).\n\nWhen to implement this method for new types X\n\nThis function is used by implementations of MOI.copy_to on constraint functions, attribute values and submittable values. If you define a new attribute whose values x::X contain variable or constraint indices, you must also implement this function.\n\n\n\n\n\nmap_indices(\n    variable_map::AbstractDict{T,T},\n    x::X,\n)::X where {T<:MOI.Index,X}\n\nShortcut for map_indices(vi -> variable_map[vi], x).\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.substitute_variables","page":"API Reference","title":"MathOptInterface.Utilities.substitute_variables","text":"substitute_variables(variable_map::Function, x)\n\nSubstitute any MOI.VariableIndex in x by variable_map(x). The variable_map function returns either MOI.VariableIndex or MOI.ScalarAffineFunction, see eval_variables for a similar function where variable_map returns a number.\n\nThis function is used by bridge optimizers on constraint functions, attribute values and submittable values when at least one variable bridge is used hence it needs to be implemented for custom types that are meant to be used as attribute or submittable value.\n\nnote: Note\nWhen implementing a new method, don't use substitute_variables(::Function, because Julia will not specialize on it. Use instead substitute_variables(::F, ...) where {F<:Function}.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.filter_variables","page":"API Reference","title":"MathOptInterface.Utilities.filter_variables","text":"filter_variables(keep::Function, f::AbstractFunction)\n\nReturn a new function f with the variable vi such that !keep(vi) removed.\n\nWARNING: Don't define filter_variables(::Function, ...) because Julia will not specialize on this. Define instead filter_variables(::F, ...) where {F<:Function}.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.remove_variable","page":"API Reference","title":"MathOptInterface.Utilities.remove_variable","text":"remove_variable(f::AbstractFunction, vi::VariableIndex)\n\nReturn a new function f with the variable vi removed.\n\n\n\n\n\nremove_variable(\n    f::MOI.AbstractFunction,\n    s::MOI.AbstractSet,\n    vi::MOI.VariableIndex,\n)\n\nReturn a tuple (g, t) representing the constraint f-in-s with the variable vi removed. That is, the terms containing the variable vi in the function f are removed and the dimension of the set s is updated if needed (e.g. when f is a VectorOfVariables with vi being one of the variables).\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.all_coefficients","page":"API Reference","title":"MathOptInterface.Utilities.all_coefficients","text":"all_coefficients(p::Function, f::MOI.AbstractFunction)\n\nDetermine whether predicate p returns true for all coefficients of f, returning false as soon as the first coefficient of f for which p returns false is encountered (short-circuiting). Similar to all.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.unsafe_add","page":"API Reference","title":"MathOptInterface.Utilities.unsafe_add","text":"unsafe_add(t1::MOI.ScalarAffineTerm, t2::MOI.ScalarAffineTerm)\n\nSums the coefficients of t1 and t2 and returns an output MOI.ScalarAffineTerm. It is unsafe because it uses the variable of t1 as the variable of the output without checking that it is equal to that of t2.\n\n\n\n\n\nunsafe_add(t1::MOI.ScalarQuadraticTerm, t2::MOI.ScalarQuadraticTerm)\n\nSums the coefficients of t1 and t2 and returns an output MOI.ScalarQuadraticTerm. It is unsafe because it uses the variable's of t1 as the variable's of the output without checking that they are the same (up to permutation) to those of t2.\n\n\n\n\n\nunsafe_add(t1::MOI.VectorAffineTerm, t2::MOI.VectorAffineTerm)\n\nSums the coefficients of t1 and t2 and returns an output MOI.VectorAffineTerm. It is unsafe because it uses the output_index and variable of t1 as the output_index and variable of the output term without checking that they are equal to those of t2.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.isapprox_zero","page":"API Reference","title":"MathOptInterface.Utilities.isapprox_zero","text":"isapprox_zero(f::MOI.AbstractFunction, tol)\n\nReturn a Bool indicating whether the function f is approximately zero using tol as a tolerance.\n\nImportant note\n\nThis function assumes that f does not contain any duplicate terms, you might want to first call canonical if that is not guaranteed. For instance, given\n\nf = MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.([1, -1], [x, x]), 0)\n\nthen isapprox_zero(f) is false but isapprox_zero(MOIU.canonical(f)) is true.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.modify_function","page":"API Reference","title":"MathOptInterface.Utilities.modify_function","text":"modify_function(f::AbstractFunction, change::AbstractFunctionModification)\n\nReturn a copy of the function f, modified according to change.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.zero_with_output_dimension","page":"API Reference","title":"MathOptInterface.Utilities.zero_with_output_dimension","text":"zero_with_output_dimension(::Type{T}, output_dimension::Integer) where {T}\n\nCreate an instance of type T with the output dimension output_dimension.\n\nThis is mostly useful in Bridges, when code needs to be agnostic to the type of vector-valued function that is passed in.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"The following functions can be used to canonicalize a function:","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.is_canonical\nUtilities.canonical\nUtilities.canonicalize!","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.is_canonical","page":"API Reference","title":"MathOptInterface.Utilities.is_canonical","text":"is_canonical(f::Union{ScalarAffineFunction, VectorAffineFunction})\n\nReturns a Bool indicating whether the function is in canonical form. See canonical.\n\n\n\n\n\nis_canonical(f::Union{ScalarQuadraticFunction, VectorQuadraticFunction})\n\nReturns a Bool indicating whether the function is in canonical form. See canonical.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.canonical","page":"API Reference","title":"MathOptInterface.Utilities.canonical","text":"canonical(f::MOI.AbstractFunction)\n\nReturns the function in a canonical form, i.e.\n\nA term appear only once.\nThe coefficients are nonzero.\nThe terms appear in increasing order of variable where there the order of the variables is the order of their value.\nFor a AbstractVectorFunction, the terms are sorted in ascending order of output index.\n\nThe output of canonical can be assumed to be a copy of f, even for VectorOfVariables.\n\nExamples\n\nIf x (resp. y, z) is VariableIndex(1) (resp. 2, 3). The canonical representation of ScalarAffineFunction([y, x, z, x, z], [2, 1, 3, -2, -3], 5) is ScalarAffineFunction([x, y], [-1, 2], 5).\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.canonicalize!","page":"API Reference","title":"MathOptInterface.Utilities.canonicalize!","text":"canonicalize!(f::Union{ScalarAffineFunction, VectorAffineFunction})\n\nConvert a function to canonical form in-place, without allocating a copy to hold the result. See canonical.\n\n\n\n\n\ncanonicalize!(f::Union{ScalarQuadraticFunction, VectorQuadraticFunction})\n\nConvert a function to canonical form in-place, without allocating a copy to hold the result. See canonical.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"The following functions can be used to manipulate functions with basic algebra:","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.scalar_type\nUtilities.scalarize\nUtilities.eachscalar\nUtilities.promote_operation\nUtilities.operate\nUtilities.operate!\nUtilities.operate_output_index!\nUtilities.vectorize","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.scalar_type","page":"API Reference","title":"MathOptInterface.Utilities.scalar_type","text":"scalar_type(F::Type{<:MOI.AbstractVectorFunction})\n\nType of functions obtained by indexing objects obtained by calling eachscalar on functions of type F.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.scalarize","page":"API Reference","title":"MathOptInterface.Utilities.scalarize","text":"scalarize(func::MOI.VectorOfVariables, ignore_constants::Bool = false)\n\nReturns a vector of scalar functions making up the vector function in the form of a Vector{MOI.SingleVariable}.\n\nSee also eachscalar.\n\n\n\n\n\nscalarize(func::MOI.VectorAffineFunction{T}, ignore_constants::Bool = false)\n\nReturns a vector of scalar functions making up the vector function in the form of a Vector{MOI.ScalarAffineFunction{T}}.\n\nSee also eachscalar.\n\n\n\n\n\nscalarize(func::MOI.VectorQuadraticFunction{T}, ignore_constants::Bool = false)\n\nReturns a vector of scalar functions making up the vector function in the form of a Vector{MOI.ScalarQuadraticFunction{T}}.\n\nSee also eachscalar.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.eachscalar","page":"API Reference","title":"MathOptInterface.Utilities.eachscalar","text":"eachscalar(f::MOI.AbstractVectorFunction)\n\nReturns an iterator for the scalar components of the vector function.\n\nSee also scalarize.\n\n\n\n\n\neachscalar(f::MOI.AbstractVector)\n\nReturns an iterator for the scalar components of the vector.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.promote_operation","page":"API Reference","title":"MathOptInterface.Utilities.promote_operation","text":"promote_operation(\n    op::Function,\n    ::Type{T},\n    ArgsTypes::Type{<:Union{T,AbstractVector{T},MOI.AbstractFunction}}...,\n) where {T<:Number}\n\nCompute the return type of the call operate(op, T, args...), where the types of the arguments args are ArgsTypes.\n\nOne assumption is that the element type T is invariant under each operation. That is, op(::T, ::T)::T where op is a +, -, *, and /.\n\nThere are six methods for which we implement Utilities.promote_operation:\n\n+ a. promote_operation(::typeof(+), ::Type{T}, ::Type{F1}, ::Type{F2})\n- a. promote_operation(::typeof(-), ::Type{T}, ::Type{F}) b. promote_operation(::typeof(-), ::Type{T}, ::Type{F1}, ::Type{F2})\n* a. promote_operation(::typeof(*), ::Type{T}, ::Type{T}, ::Type{F}) b. promote_operation(::typeof(*), ::Type{T}, ::Type{F}, ::Type{T}) c. promote_operation(::typeof(*), ::Type{T}, ::Type{F1}, ::Type{F2})    where F1 and F2 are VariableIndex or ScalarAffineFunction d. promote_operation(::typeof(*), ::Type{T}, ::Type{<:Diagonal{T}}, ::Type{F}\n/ a. promote_operation(::typeof(/), ::Type{T}, ::Type{F}, ::Type{T})\nvcat a. promote_operation(::typeof(vcat), ::Type{T}, ::Type{F}...)\nimag a. promote_operation(::typeof(imag), ::Type{T}, ::Type{F})    where F is VariableIndex or VectorOfVariables\n\nIn each case, F (or F1 and F2) is one of the ten supported types, with a restriction that the mathematical operation makes sense, for example, we don't define promote_operation(-, T, F1, F2) where F1 is a scalar-valued function and  F2 is a vector-valued function. The ten supported types are:\n\n::T\n::VariableIndex\n::ScalarAffineFunction{T}\n::ScalarQuadraticFunction{T}\n::ScalarNonlinearFunction\n::AbstractVector{T}\n::VectorOfVariables\n::VectorAffineFunction{T}\n::VectorQuadraticFunction{T}\n::VectorNonlinearFunction\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.operate","page":"API Reference","title":"MathOptInterface.Utilities.operate","text":"operate(\n    op::Function,\n    ::Type{T},\n    args::Union{T,MOI.AbstractFunction}...,\n)::MOI.AbstractFunction where {T<:Number}\n\nReturns an MOI.AbstractFunction representing the function resulting from the operation op(args...) on functions of coefficient type T.\n\nNo argument can be modified.\n\nMethods\n\n+ a. operate(::typeof(+), ::Type{T}, ::F1) b. operate(::typeof(+), ::Type{T}, ::F1, ::F2) c. operate(::typeof(+), ::Type{T}, ::F1...)\n- a. operate(::typeof(-), ::Type{T}, ::F) b. operate(::typeof(-), ::Type{T}, ::F1, ::F2)\n* a. operate(::typeof(*), ::Type{T}, ::T, ::F) b. operate(::typeof(*), ::Type{T}, ::F, ::T) c. operate(::typeof(*), ::Type{T}, ::F1, ::F2)    where F1 and F2 are VariableIndex or ScalarAffineFunction d. operate(::typeof(*), ::Type{T}, ::Diagonal{T}, ::F)\n/ a. operate(::typeof(/), ::Type{T}, ::F, ::T)\nvcat a. operate(::typeof(vcat), ::Type{T}, ::F...)\nimag a. operate(::typeof(imag), ::Type{T}, ::F)    where F is VariableIndex or VectorOfVariables\n\nOne assumption is that the element type T is invariant under each operation. That is, op(::T, ::T)::T where op is a +, -, *, and /.\n\nIn each case, F (or F1 and F2) is one of the ten supported types, with a restriction that the mathematical operation makes sense, for example, we don't define promote_operation(-, T, F1, F2) where F1 is a scalar-valued function and  F2 is a vector-valued function. The ten supported types are:\n\n::T\n::VariableIndex\n::ScalarAffineFunction{T}\n::ScalarQuadraticFunction{T}\n::ScalarNonlinearFunction\n::AbstractVector{T}\n::VectorOfVariables\n::VectorAffineFunction{T}\n::VectorQuadraticFunction{T}\n::VectorNonlinearFunction\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.operate!","page":"API Reference","title":"MathOptInterface.Utilities.operate!","text":"operate!(\n    op::Function,\n    ::Type{T},\n    args::Union{T,MOI.AbstractFunction}...,\n)::MOI.AbstractFunction where {T<:Number}\n\nReturns an MOI.AbstractFunction representing the function resulting from the operation op(args...) on functions of coefficient type T.\n\nThe first argument may be modified, in which case the return value is identical to the first argument. For operations which cannot be implemented in-place, this function returns a new object.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.operate_output_index!","page":"API Reference","title":"MathOptInterface.Utilities.operate_output_index!","text":"operate_output_index!(\n    op::Union{typeof(+),typeof(-)},\n    ::Type{T},\n    output_index::Integer,\n    f::Union{AbstractVector{T},MOI.AbstractVectorFunction}\n    g::Union{T,MOI.AbstractScalarFunction}...\n) where {T<:Number}\n\nReturn an MOI.AbstractVectorFunction in which the scalar function in row output_index is the result of op(f[output_index], g).\n\nThe functions at output index different to output_index are the same as the functions at the same output index in func. The first argument may be modified.\n\nMethods\n\n+ a. operate_output_index!(+, ::Type{T}, ::Int, ::VectorF, ::ScalarF)\n- a. operate_output_index!(-, ::Type{T}, ::Int, ::VectorF, ::ScalarF)\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.vectorize","page":"API Reference","title":"MathOptInterface.Utilities.vectorize","text":"vectorize(x::AbstractVector{<:Number})\n\nReturns x.\n\n\n\n\n\nvectorize(x::AbstractVector{MOI.VariableIndex})\n\nReturns the vector of scalar affine functions in the form of a MOI.VectorAffineFunction{T}.\n\n\n\n\n\nvectorize(funcs::AbstractVector{MOI.ScalarAffineFunction{T}}) where T\n\nReturns the vector of scalar affine functions in the form of a MOI.VectorAffineFunction{T}.\n\n\n\n\n\nvectorize(funcs::AbstractVector{MOI.ScalarQuadraticFunction{T}}) where T\n\nReturns the vector of scalar quadratic functions in the form of a MOI.VectorQuadraticFunction{T}.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#Constraint-utilities","page":"API Reference","title":"Constraint utilities","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"The following utilities are available for moving the function constant to the set for scalar constraints:","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.shift_constant\nUtilities.supports_shift_constant\nUtilities.normalize_and_add_constraint\nUtilities.normalize_constant","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.shift_constant","page":"API Reference","title":"MathOptInterface.Utilities.shift_constant","text":"shift_constant(set::MOI.AbstractScalarSet, offset)\n\nReturns a new scalar set new_set such that func-in-set is equivalent to func + offset-in-new_set.\n\nOnly define this function if it makes sense to!\n\nUse supports_shift_constant to check if the set supports shifting:\n\nif supports_shift_constant(typeof(old_set))\n    new_set = shift_constant(old_set, offset)\n    f.constant = 0\n    add_constraint(model, f, new_set)\nelse\n    add_constraint(model, f, old_set)\nend\n\nSee also supports_shift_constant.\n\nExamples\n\nThe call shift_constant(MOI.Interval(-2, 3), 1) is equal to MOI.Interval(-1, 4).\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.supports_shift_constant","page":"API Reference","title":"MathOptInterface.Utilities.supports_shift_constant","text":"supports_shift_constant(::Type{S}) where {S<:MOI.AbstractSet}\n\nReturn true if shift_constant is defined for set S.\n\nSee also shift_constant.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.normalize_and_add_constraint","page":"API Reference","title":"MathOptInterface.Utilities.normalize_and_add_constraint","text":"normalize_and_add_constraint(\n    model::MOI.ModelLike,\n    func::MOI.AbstractScalarFunction,\n    set::MOI.AbstractScalarSet;\n    allow_modify_function::Bool = false,\n)\n\nAdds the scalar constraint obtained by moving the constant term in func to the set in model. If allow_modify_function is true then the function func can be modified.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.normalize_constant","page":"API Reference","title":"MathOptInterface.Utilities.normalize_constant","text":"normalize_constant(\n    func::MOI.AbstractScalarFunction,\n    set::MOI.AbstractScalarSet;\n    allow_modify_function::Bool = false,\n)\n\nReturn the func-in-set constraint in normalized form. That is, if func is MOI.ScalarQuadraticFunction or MOI.ScalarAffineFunction, the constant is moved to the set. If allow_modify_function is true then the function func can be modified.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"The following utility identifies those constraints imposing bounds on a given variable, and returns those bound values:","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.get_bounds","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.get_bounds","page":"API Reference","title":"MathOptInterface.Utilities.get_bounds","text":"get_bounds(model::MOI.ModelLike, ::Type{T}, x::MOI.VariableIndex)\n\nReturn a tuple (lb, ub) of type Tuple{T, T}, where lb and ub are lower  and upper bounds, respectively, imposed on x in model.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"The following utilities are useful when working with symmetric matrix cones.","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.is_diagonal_vectorized_index\nUtilities.side_dimension_for_vectorized_dimension","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.is_diagonal_vectorized_index","page":"API Reference","title":"MathOptInterface.Utilities.is_diagonal_vectorized_index","text":"is_diagonal_vectorized_index(index::Base.Integer)\n\nReturn whether index is the index of a diagonal element in a MOI.AbstractSymmetricMatrixSetTriangle set.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.side_dimension_for_vectorized_dimension","page":"API Reference","title":"MathOptInterface.Utilities.side_dimension_for_vectorized_dimension","text":"side_dimension_for_vectorized_dimension(n::Integer)\n\nReturn the dimension d such that MOI.dimension(MOI.PositiveSemidefiniteConeTriangle(d)) is n.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#Set-utilities","page":"API Reference","title":"Set utilities","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"The following utilities are available for sets:","category":"page"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.AbstractDistance\nUtilities.ProjectionUpperBoundDistance\nUtilities.distance_to_set\nUtilities.set_dot","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.AbstractDistance","page":"API Reference","title":"MathOptInterface.Utilities.AbstractDistance","text":"abstract type AbstractDistance end\n\nAn abstract type used to enabble dispatch of Utilities.distance_to_set.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.ProjectionUpperBoundDistance","page":"API Reference","title":"MathOptInterface.Utilities.ProjectionUpperBoundDistance","text":"ProjectionUpperBoundDistance() <: AbstractDistance\n\nAn upper bound on the minimum distance between point and the closest feasible point in set.\n\nDefinition of distance\n\nThe minimum distance is computed as:\n\nd(x mathcalK) = min_y in mathcalK  x - y \n\nwhere x is point and mathcalK is set. The norm is computed as:\n\nx = sqrtf(x x mathcalK)\n\nwhere f is Utilities.set_dot.\n\nIn the default case, where the set does not have a specialized method for Utilities.set_dot, the norm is equivalent to the Euclidean norm x = sqrtsum x_i^2.\n\nWhy an upper bound?\n\nIn most cases, distance_to_set should return the smallest upper bound, but it may return a larger value if the smallest upper bound is expensive to compute.\n\nFor example, given an epigraph from of a conic set, (t x)  f(x) le t, it may be simpler to return delta such that f(x) le t + delta, rather than computing the nearest projection onto the set.\n\nIf the distance is not the smallest upper bound, the docstring of the appropriate distance_to_set method must describe the way that the distance is computed.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.distance_to_set","page":"API Reference","title":"MathOptInterface.Utilities.distance_to_set","text":"distance_to_set(\n    [d::AbstractDistance = ProjectionUpperBoundDistance()],]\n    point::T,\n    set::MOI.AbstractScalarSet,\n) where {T}\n\ndistance_to_set(\n    [d::AbstractDistance = ProjectionUpperBoundDistance(),]\n    point::AbstractVector{T},\n    set::MOI.AbstractVectorSet,\n) where {T}\n\nCompute the distance between point and set using the distance metric d. If point is in the set set, this function must return zero(T).\n\nIf d is omitted, the default distance is Utilities.ProjectionUpperBoundDistance.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.set_dot","page":"API Reference","title":"MathOptInterface.Utilities.set_dot","text":"set_dot(x::AbstractVector, y::AbstractVector, set::AbstractVectorSet)\n\nReturn the scalar product between a vector x of the set set and a vector y of the dual of the set s.\n\n\n\n\n\nset_dot(x, y, set::AbstractScalarSet)\n\nReturn the scalar product between a number x of the set set and a number y of the dual of the set s.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#DoubleDicts","page":"API Reference","title":"DoubleDicts","text":"","category":"section"},{"location":"moi/submodules/Utilities/reference/","page":"API Reference","title":"API Reference","text":"Utilities.DoubleDicts.DoubleDict\nUtilities.DoubleDicts.DoubleDictInner\nUtilities.DoubleDicts.IndexDoubleDict\nUtilities.DoubleDicts.IndexDoubleDictInner\nUtilities.DoubleDicts.outer_keys\nUtilities.DoubleDicts.nonempty_outer_keys","category":"page"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.DoubleDict","page":"API Reference","title":"MathOptInterface.Utilities.DoubleDicts.DoubleDict","text":"DoubleDict{V}\n\nAn optimized dictionary to map MOI.ConstraintIndex to values of type V.\n\nWorks as a AbstractDict{MOI.ConstraintIndex,V} with minimal differences.\n\nIf V is also a MOI.ConstraintIndex, use IndexDoubleDict.\n\nNote that MOI.ConstraintIndex is not a concrete type, opposed to MOI.ConstraintIndex{MOI.VariableIndex, MOI.Integers}, which is a concrete type.\n\nWhen looping through multiple keys of the same Function-in-Set type, use\n\ninner = dict[F, S]\n\nto return a type-stable DoubleDictInner.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.DoubleDictInner","page":"API Reference","title":"MathOptInterface.Utilities.DoubleDicts.DoubleDictInner","text":"DoubleDictInner{F,S,V}\n\nA type stable inner dictionary of DoubleDict.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict","page":"API Reference","title":"MathOptInterface.Utilities.DoubleDicts.IndexDoubleDict","text":"IndexDoubleDict\n\nA specialized version of [DoubleDict] in which the values are of type MOI.ConstraintIndex\n\nWhen looping through multiple keys of the same Function-in-Set type, use\n\ninner = dict[F, S]\n\nto return a type-stable IndexDoubleDictInner.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.IndexDoubleDictInner","page":"API Reference","title":"MathOptInterface.Utilities.DoubleDicts.IndexDoubleDictInner","text":"IndexDoubleDictInner{F,S}\n\nA type stable inner dictionary of IndexDoubleDict.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.outer_keys","page":"API Reference","title":"MathOptInterface.Utilities.DoubleDicts.outer_keys","text":"outer_keys(d::AbstractDoubleDict)\n\nReturn an iterator over the outer keys of the AbstractDoubleDict d. Each outer key is a Tuple{Type,Type} so that a double loop can be easily used:\n\nfor (F, S) in DoubleDicts.outer_keys(dict)\n    for (k, v) in dict[F, S]\n        # ...\n    end\nend\n\nFor performance, it is recommended that the inner loop lies in a separate function to gurantee type-stability. Some outer keys (F, S) might lead to an empty dict[F, S]. If you want only nonempty dict[F, S], use nonempty_outer_keys.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Utilities/reference/#MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys","page":"API Reference","title":"MathOptInterface.Utilities.DoubleDicts.nonempty_outer_keys","text":"nonempty_outer_keys(d::AbstractDoubleDict)\n\nReturn a vector of outer keys of the AbstractDoubleDict d.\n\nOnly outer keys that have a nonempty set of inner keys will be returned.\n\nEach outer key is a Tuple{Type,Type} so that a double loop can be easily used\n\nfor (F, S) in DoubleDicts.nonempty_outer_keys(dict)\n    for (k, v) in dict[F, S]\n        # ...\n    end\nend\nFor performance, it is recommended that the inner loop lies in a separate\nfunction to gurantee type-stability.\n\nIf you want an iterator of all current outer keys, use [`outer_keys`](@ref).\n\n\n\n\n\n","category":"function"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"EditURL = \"https://github.com/jump-dev/Clp.jl/blob/v1.0.3/README.md\"","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"(Image: )","category":"page"},{"location":"packages/Clp/#Clp.jl","page":"jump-dev/Clp.jl","title":"Clp.jl","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"Clp.jl is a wrapper for the COIN-OR Linear Programming solver.","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/Clp/#Affiliation","page":"jump-dev/Clp.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"This wrapper is maintained by the JuMP community and is not a COIN-OR project.","category":"page"},{"location":"packages/Clp/#License","page":"jump-dev/Clp.jl","title":"License","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"Clp.jl is licensed under the MIT License.","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"The underlying solver, coin-or/Clp, is licensed under the Eclipse public license.","category":"page"},{"location":"packages/Clp/#Installation","page":"jump-dev/Clp.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"Install Clp using Pkg.add:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"import Pkg\nPkg.add(\"Clp\")","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"In addition to installing the Clp.jl package, this will also download and install the Clp binaries. You do not need to install Clp separately.","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"To use a custom binary, read the Custom solver binaries section of the JuMP documentation.","category":"page"},{"location":"packages/Clp/#Use-with-JuMP","page":"jump-dev/Clp.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"To use Clp with JuMP, use Clp.Optimizer:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"using JuMP, Clp\nmodel = Model(Clp.Optimizer)\nset_attribute(model, \"LogLevel\", 1)\nset_attribute(model, \"Algorithm\", 4)","category":"page"},{"location":"packages/Clp/#MathOptInterface-API","page":"jump-dev/Clp.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"The Clp optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"MOI.Reals","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"MOI.ObjectiveSense()","category":"page"},{"location":"packages/Clp/#Options","page":"jump-dev/Clp.jl","title":"Options","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"Options are, unfortunately, not well documented.","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"The following options are likely to be the most useful:","category":"page"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"Parameter Example Explanation\nPrimalTolerance 1e-7 Primal feasibility tolerance\nDualTolerance 1e-7 Dual feasibility tolerance\nDualObjectiveLimit 1e308 When using dual simplex (where the objective is monotonically changing), terminate when the objective exceeds this limit\nMaximumIterations 2147483647 Terminate after performing this number of simplex iterations\nMaximumSeconds -1.0 Terminate after this many seconds have passed. A negative value means no time limit\nLogLevel 1 Set to 1, 2, 3, or 4 for increasing output. Set to 0 to disable output\nPresolveType 0 Set to 1 to disable presolve\nSolveType 5 Solution method: dual simplex (0), primal simplex (1), sprint (2), barrier with crossover (3), barrier without crossover (4), automatic (5)\nInfeasibleReturn 0 Set to 1 to return as soon as the problem is found to be infeasible (by default, an infeasibility proof is computed as well)\nScaling 3 0 -off, 1 equilibrium, 2 geometric, 3 auto, 4 dynamic(later)\nPerturbation 100 switch on perturbation (50), automatic (100), don't try perturbing (102)","category":"page"},{"location":"packages/Clp/#C-API","page":"jump-dev/Clp.jl","title":"C API","text":"","category":"section"},{"location":"packages/Clp/","page":"jump-dev/Clp.jl","title":"jump-dev/Clp.jl","text":"The C API can be accessed via Clp.Clp_XXX functions, where the names and arguments are identical to the C API.","category":"page"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/index.md\"","category":"page"},{"location":"moi/#moi_documentation","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"warning: Warning\nThis documentation in this section is a copy of the official MathOptInterface documentation available at https://jump.dev/MathOptInterface.jl/v1.20.0. It is included here to make it easier to link concepts between JuMP and MathOptInterface.","category":"page"},{"location":"moi/#What-is-MathOptInterface?","page":"Introduction","title":"What is MathOptInterface?","text":"","category":"section"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"MathOptInterface.jl (MOI) is an abstraction layer designed to provide a unified interface to mathematical optimization solvers so that users do not need to understand multiple solver-specific APIs.","category":"page"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"tip: Tip\nThis documentation is aimed at developers writing software interfaces to solvers and modeling languages using the MathOptInterface API. If you are a user interested in solving optimization problems, we encourage you instead to use MOI through a higher-level modeling interface like JuMP or Convex.jl.","category":"page"},{"location":"moi/#How-the-documentation-is-structured","page":"Introduction","title":"How the documentation is structured","text":"","category":"section"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"Having a high-level overview of how this documentation is structured will help you know where to look for certain things.","category":"page"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"The Tutorials section contains articles on how to use and implement the MathOptInteraface API. Look here if you want to write a model in MOI, or write an interface to a new solver.\nThe Manual contains short code-snippets that explain how to use the MOI API. Look here for more details on particular areas of MOI.\nThe Background section contains articles on the theory behind MathOptInterface. Look here if you want to understand why, rather than how.\nThe API Reference contains a complete list of functions and types that comprise the MOI API. Look here is you want to know how to use (or implement) a particular function.\nThe Submodules section contains stand-alone documentation for each of the submodules within MOI. These submodules are not required to interface a solver with MOI, but they make the job much easier.","category":"page"},{"location":"moi/#Citing-MathOptInterface","page":"Introduction","title":"Citing MathOptInterface","text":"","category":"section"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"A paper describing the design and features of MathOptInterface is available on arXiv.","category":"page"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"If you find MathOptInterface useful in your work, we kindly request that you cite the following paper:","category":"page"},{"location":"moi/","page":"Introduction","title":"Introduction","text":"@article{legat2021mathoptinterface,\n    title={{MathOptInterface}: a data structure for mathematical optimization problems},\n    author={Legat, Beno{\\^\\i}t and Dowson, Oscar and Garcia, Joaquim Dias and Lubin, Miles},\n    journal={INFORMS Journal on Computing},\n    year={2021},\n    doi={10.1287/ijoc.2021.1067},\n    publisher={INFORMS}\n}","category":"page"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"EditURL = \"https://github.com/osqp/OSQP.jl/blob/443706e34c2619acbe65281c60bbe850ca4a8fac/README.md\"","category":"page"},{"location":"packages/OSQP/#OSQP.jl","page":"osqp/OSQP.jl","title":"OSQP.jl","text":"","category":"section"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"(Image: Build Status)  (Image: codecov.io)","category":"page"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"OSQP.jl is a Julia wrapper for OSQP: the Operator Splitting QP Solver.","category":"page"},{"location":"packages/OSQP/#License","page":"osqp/OSQP.jl","title":"License","text":"","category":"section"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"OSQP.jl is licensed under the Apache-2.0 license.","category":"page"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"The upstream solver, osqp/osqp is also licensed under the Apache-2.0 license.","category":"page"},{"location":"packages/OSQP/#Installation","page":"osqp/OSQP.jl","title":"Installation","text":"","category":"section"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"Install OSQP.jl using the Julia package manager","category":"page"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"import Pkg\nPkg.add(\"OSQP\")","category":"page"},{"location":"packages/OSQP/#Problem-class","page":"osqp/OSQP.jl","title":"Problem class","text":"","category":"section"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"The OSQP (Operator Splitting Quadratic Program) solver is a numerical optimization package for solving problems in the form","category":"page"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"minimize        0.5 x' P x + q' x\n\nsubject to      l <= A x <= u","category":"page"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"where x in R^n is the optimization variable. The objective function is defined by a positive semidefinite matrix P in S^n_+ and vector q in R^n. The linear constraints are defined by matrix A in R^{m x n} and vectors l in R^m U {-inf}^m, u in R^m U {+inf}^m.","category":"page"},{"location":"packages/OSQP/#Documentation","page":"osqp/OSQP.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/OSQP/","page":"osqp/OSQP.jl","title":"osqp/OSQP.jl","text":"Detailed documentation is available at https://osqp.org/.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP, HiGHS, SCS\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/models/#jump_models","page":"Models","title":"Models","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"JuMP models are the fundamental building block that we use to construct optimization problems. They hold things like the variables and constraints, as well as which solver to use and even solution information.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"info: Info\nJuMP uses \"optimizer\" as a synonym for \"solver.\" Our convention is to use \"solver\" to refer to the underlying software, and use \"optimizer\" to refer to the Julia object that wraps the solver. For example, HiGHS is a solver, and HiGHS.Optimizer is an optimizer.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"tip: Tip\nSee Supported solvers for a list of available solvers.","category":"page"},{"location":"manual/models/#Create-a-model","page":"Models","title":"Create a model","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"Create a model by passing an optimizer to Model:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(HiGHS.Optimizer)\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: EMPTY_OPTIMIZER\nSolver name: HiGHS","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"If you don't know which optimizer you will be using at creation time, create a model without an optimizer, and then call set_optimizer at any time prior to optimize!:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model()\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\n\njulia> set_optimizer(model, HiGHS.Optimizer)","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"tip: Tip\nDon't know what the fields Model mode and CachingOptimizer state mean? Read the Backends section.","category":"page"},{"location":"manual/models/#What-is-the-difference?","page":"Models","title":"What is the difference?","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"For most models, there is no difference between passing the optimizer to Model, and calling set_optimizer.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"However, if an optimizer does not support a constraint in the model, the timing of when an error will be thrown can differ:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"If you pass an optimizer, an error will be thrown when you try to add the constraint.\nIf you call set_optimizer, an error will be thrown when you try to solve the model via optimize!.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Therefore, most users should pass an optimizer to Model because it provides the earliest warning that your solver is not suitable for the model you are trying to build. However, if you are modifying a problem by adding and deleting different constraint types, you may need to use set_optimizer. See Switching optimizer for the relaxed problem for an example of when this is useful.","category":"page"},{"location":"manual/models/#Reducing-time-to-first-solve-latency","page":"Models","title":"Reducing time-to-first-solve latency","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"By default, JuMP uses bridges to reformulate the model you are building into an equivalent model supported by the solver.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"However, if your model is already supported by the solver, bridges add latency (read The \"time-to-first-solve\" issue). This is particularly noticeable for small models.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"To reduce the \"time-to-first-solve,s\" try passing add_bridges = false.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(HiGHS.Optimizer; add_bridges = false);","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"or","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model();\n\njulia> set_optimizer(model, HiGHS.Optimizer; add_bridges = false)","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"However, be wary. If your model and solver combination needs bridges, an error will be thrown:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(SCS.Optimizer; add_bridges = false);\n\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, 2x <= 1)\nERROR: Constraints of type MathOptInterface.ScalarAffineFunction{Float64}-in-MathOptInterface.LessThan{Float64} are not supported by the solver.\n\nIf you expected the solver to support your problem, you may have an error in your formulation. Otherwise, consider using a different solver.\n\nThe list of available solvers, along with the problem types they support, is available at https://jump.dev/JuMP.jl/stable/installation/#Supported-solvers.\n[...]","category":"page"},{"location":"manual/models/#Solvers-which-expect-environments","page":"Models","title":"Solvers which expect environments","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"Some solvers accept (or require) positional arguments such as a license environment or a path to a binary executable. For these solvers, you can pass a function to Model which takes zero arguments and returns an instance of the optimizer.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"A common use-case for this is passing an environment or sub-solver to the optimizer:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> import HiGHS\n\njulia> import MultiObjectiveAlgorithms as MOA\n\njulia> model = Model(() -> MOA.Optimizer(HiGHS.Optimizer))\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: EMPTY_OPTIMIZER\nSolver name: MOA[algorithm=MultiObjectiveAlgorithms.Lexicographic, optimizer=HiGHS]","category":"page"},{"location":"manual/models/#solver_options","page":"Models","title":"Solver options","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"JuMP uses \"attribute\" as a synonym for \"option.\" Use optimizer_with_attributes to create an optimizer with some attributes initialized:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(\n           optimizer_with_attributes(HiGHS.Optimizer, \"output_flag\" => false),\n       )\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: EMPTY_OPTIMIZER\nSolver name: HiGHS","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Alternatively, use set_attribute to set an attribute after the model has been created:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_attribute(model, \"output_flag\", false)\n\njulia> get_attribute(model, \"output_flag\")\nfalse","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"You can also modify attributes within an optimizer_with_attributes object:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> solver = optimizer_with_attributes(HiGHS.Optimizer, \"output_flag\" => true);\n\njulia> get_attribute(solver, \"output_flag\")\ntrue\n\njulia> set_attribute(solver, \"output_flag\", false)\n\njulia> get_attribute(solver, \"output_flag\")\nfalse\n\njulia> model = Model(solver);","category":"page"},{"location":"manual/models/#Changing-the-number-types","page":"Models","title":"Changing the number types","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"By default, the coefficients of affine and quadratic expressions are numbers of type either Float64 or Complex{Float64} (see Complex number support).","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"The type Float64 can be changed using the GenericModel constructor:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = GenericModel{Rational{BigInt}}();\n\njulia> @variable(model, x)\nx\n\njulia> @expression(model, expr, 1 // 3 * x)\n1//3 x\n\njulia> typeof(expr)\nGenericAffExpr{Rational{BigInt}, GenericVariableRef{Rational{BigInt}}}","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Using a value_type other than Float64 is an advanced operation and should be used only if the underlying solver actually solves the problem using the provided value type.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"warning: Warning\nNonlinear Modeling is currently restricted to the Float64 number type.","category":"page"},{"location":"manual/models/#Print-the-model","page":"Models","title":"Print the model","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"By default, show(model) will print a summary of the problem:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(); @variable(model, x >= 0); @objective(model, Max, x);\n\njulia> model\nA JuMP Model\nMaximization problem with:\nVariable: 1\nObjective function type: VariableRef\n`VariableRef`-in-`MathOptInterface.GreaterThan{Float64}`: 1 constraint\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\nNames registered in the model: x","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Use print to print the formulation of the model (in IJulia, this will render as LaTeX.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> print(model)\nMax x\nSubject to\n x ≥ 0","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"warning: Warning\nThis format is specific to JuMP and may change in any future release. It is not intended to be an instance format. To write the model to a file, use write_to_file instead.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Use latex_formulation to display the model in LaTeX form.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> latex_formulation(model)\n$$ \\begin{aligned}\n\\max\\quad & x\\\\\n\\text{Subject to} \\quad & x \\geq 0\\\\\n\\end{aligned} $$","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"In IJulia (and Documenter), ending a cell in with latex_formulation will render the model in LaTeX:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"using JuMP                # hide\nmodel = Model()           # hide\n@variable(model, x >= 0)  # hide\n@objective(model, Max, x) # hide\nlatex_formulation(model)","category":"page"},{"location":"manual/models/#Turn-off-output","page":"Models","title":"Turn off output","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"Use set_silent and unset_silent to disable or enable printing output from the solver.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> unset_silent(model)","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"tip: Tip\nMost solvers will also have a solver-specific option to provide finer-grained control over the output. Consult their README's for details.","category":"page"},{"location":"manual/models/#Set-a-time-limit","page":"Models","title":"Set a time limit","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"Use set_time_limit_sec, unset_time_limit_sec, and time_limit_sec to manage time limits.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_time_limit_sec(model, 60.0)\n\n\njulia> time_limit_sec(model)\n60.0\n\njulia> unset_time_limit_sec(model)\n\njulia> limit = time_limit_sec(model)\n\njulia> limit === nothing\ntrue","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"If your time limit is encoded as a Dates.Period object, use the following code to convert it to Float64 for set_time_limit_sec:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> import Dates\n\njulia> seconds(x::Dates.Period) = 1e-3 * Dates.value(round(x, Dates.Millisecond))\nseconds (generic function with 1 method)\n\njulia> set_time_limit_sec(model, seconds(Dates.Hour(1)))\n\njulia> time_limit_sec(model)\n3600.0","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"info: Info\nSome solvers do not support time limits. In these cases, an error will be thrown.","category":"page"},{"location":"manual/models/#Write-a-model-to-file","page":"Models","title":"Write a model to file","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"JuMP can write models to a variety of file-formats using write_to_file and Base.write.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"For most common file formats, the file type will be detected from the extension.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"For example, here is how to write an MPS file:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model();\n\njulia> write_to_file(model, \"model.mps\")","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Other supported file formats include:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":".cbf for the Conic Benchmark Format\n.lp for the LP file format\n.mof.json for the MathOptFormat\n.nl for AMPL's NL file format\n.rew for the REW file format\n.sdpa and \".dat-s\" for the SDPA file format","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"To write to a specific io::IO, use Base.write. Specify the file type by passing a MOI.FileFormats.FileFormat enum.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model();\n\njulia> io = IOBuffer();\n\njulia> write(io, model; format = MOI.FileFormats.FORMAT_MPS)","category":"page"},{"location":"manual/models/#Read-a-model-from-file","page":"Models","title":"Read a model from file","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"JuMP models can be created from file formats using read_from_file and Base.read.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = read_from_file(\"model.mps\")\nA JuMP Model\nMinimization problem with:\nVariables: 0\nObjective function type: AffExpr\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\n\njulia> seekstart(io);\n\njulia> model2 = read(io, Model; format = MOI.FileFormats.FORMAT_MPS)\nA JuMP Model\nMinimization problem with:\nVariables: 0\nObjective function type: AffExpr\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"note: Note\nBecause file formats do not serialize the containers of JuMP variables and constraints, the names in the model will not be registered. Therefore, you cannot access named variables and constraints via model[:x]. Instead, use variable_by_name or constraint_by_name to access specific variables or constraints.","category":"page"},{"location":"manual/models/#Relax-integrality","page":"Models","title":"Relax integrality","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"Use relax_integrality to remove any integrality constraints from the model, such as integer and binary restrictions on variables. relax_integrality returns a function that can be later called with zero arguments to re-add the removed constraints:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model();\n\njulia> @variable(model, x, Int)\nx\n\njulia> num_constraints(model, VariableRef, MOI.Integer)\n1\n\njulia> undo = relax_integrality(model);\n\njulia> num_constraints(model, VariableRef, MOI.Integer)\n0\n\njulia> undo()\n\njulia> num_constraints(model, VariableRef, MOI.Integer)\n1","category":"page"},{"location":"manual/models/#Switching-optimizer-for-the-relaxed-problem","page":"Models","title":"Switching optimizer for the relaxed problem","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"A common reason for relaxing integrality is to compute dual variables of the relaxed problem. However, some mixed-integer linear solvers (for example, Cbc) do not return dual solutions, even if the problem does not have integrality restrictions.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Therefore, after relax_integrality you should call set_optimizer with a solver that does support dual solutions, such as Clp.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"For example, instead of:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"using JuMP, Cbc\nmodel = Model(Cbc.Optimizer)\n@variable(model, x, Int)\nundo = relax_integrality(model)\noptimize!(model)\nreduced_cost(x)  # Errors","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"do:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"using JuMP, Cbc, Clp\nmodel = Model(Cbc.Optimizer)\n@variable(model, x, Int)\nundo = relax_integrality(model)\nset_optimizer(model, Clp.Optimizer)\noptimize!(model)\nreduced_cost(x)  # Works","category":"page"},{"location":"manual/models/#Backends","page":"Models","title":"Backends","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"info: Info\nThis section discusses advanced features of JuMP. For new users, you may want to skip this section. You don't need to know how JuMP manages problems behind the scenes to create and solve JuMP models.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"A JuMP Model is a thin layer around a backend of type MOI.ModelLike that stores the optimization problem and acts as the optimization solver.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"However, if you construct a model like Model(HiGHS.Optimizer), the backend is not a HiGHS.Optimizer, but a more complicated object.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"From JuMP, the MOI backend can be accessed using the backend function. Let's see what the backend of a JuMP Model is:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> b = backend(model)\nMOIU.CachingOptimizer{MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}, MOIU.UniversalFallback{MOIU.Model{Float64}}}\nin state EMPTY_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.UniversalFallback{MOIU.Model{Float64}}\n  fallback for MOIU.Model{Float64}\nwith optimizer MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}\n  with 0 variable bridges\n  with 0 constraint bridges\n  with 0 objective bridges\n  with inner model A HiGHS model with 0 columns and 0 rows.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Uh oh. Even though we passed a HiGHS.Optimizer, the backend is a much more complicated object.","category":"page"},{"location":"manual/models/#CachingOptimizer","page":"Models","title":"CachingOptimizer","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"A MOIU.CachingOptimizer is a layer that abstracts the difference between solvers that support incremental modification (for example, they support adding variables one-by-one), and solvers that require the entire problem in a single API call (for example, they only accept the A, b and c matrices of a linear program).","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"It has two parts:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"A cache, where the model can be built and modified incrementally\njulia> b.model_cache\nMOIU.UniversalFallback{MOIU.Model{Float64}}\nfallback for MOIU.Model{Float64}\nAn optimizer, which is used to solve the problem\njulia> b.optimizer\nMOIB.LazyBridgeOptimizer{HiGHS.Optimizer}\nwith 0 variable bridges\nwith 0 constraint bridges\nwith 0 objective bridges\nwith inner model A HiGHS model with 0 columns and 0 rows.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"info: Info\nThe LazyBridgeOptimizer section explains what a LazyBridgeOptimizer is.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"The CachingOptimizer has logic to decide when to copy the problem from the cache to the optimizer, and when it can efficiently update the optimizer in-place.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"A CachingOptimizer may be in one of three possible states:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"NO_OPTIMIZER: The CachingOptimizer does not have any optimizer.\nEMPTY_OPTIMIZER: The CachingOptimizer has an empty optimizer, and it is not synchronized with the cached model.\nATTACHED_OPTIMIZER: The CachingOptimizer has an optimizer, and it is synchronized with the cached model.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"A CachingOptimizer has two modes of operation:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"AUTOMATIC: The CachingOptimizer changes its state when necessary. For example, optimize! will automatically call attach_optimizer (an optimizer must have been previously set). Attempting to add a constraint or perform a modification not supported by the optimizer results in a drop to EMPTY_OPTIMIZER mode.\nMANUAL: The user must change the state of the CachingOptimizer using MOIU.reset_optimizer(::JuMP.Model), MOIU.drop_optimizer(::JuMP.Model), and MOIU.attach_optimizer(::JuMP.Model). Attempting to perform an operation in the incorrect state results in an error.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"By default Model will create a CachingOptimizer in AUTOMATIC mode.","category":"page"},{"location":"manual/models/#LazyBridgeOptimizer","page":"Models","title":"LazyBridgeOptimizer","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"The second layer that JuMP applies automatically is a MOI.Bridges.LazyBridgeOptimizer. A MOI.Bridges.LazyBridgeOptimizer is an MOI layer that attempts to transform the problem from the formulation provided by the user into an equivalent problem supported by the solver. This may involve adding new variables and constraints to the optimizer. The transformations are selected from a set of known recipes called bridges.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"A common example of a bridge is one that splits an interval constraint like @constraint(model, 1 <= x + y <= 2) into two constraints, @constraint(model, x + y >= 1) and @constraint(model, x + y <= 2).","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Use the add_bridges = false keyword to remove the bridging layer:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = Model(HiGHS.Optimizer; add_bridges = false)\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: EMPTY_OPTIMIZER\nSolver name: HiGHS\n\njulia> backend(model)\nMOIU.CachingOptimizer{HiGHS.Optimizer, MOIU.UniversalFallback{MOIU.Model{Float64}}}\nin state EMPTY_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.UniversalFallback{MOIU.Model{Float64}}\n  fallback for MOIU.Model{Float64}\nwith optimizer A HiGHS model with 0 columns and 0 rows.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"Bridges can be added and removed from a MOI.Bridges.LazyBridgeOptimizer using add_bridge and remove_bridge. Use print_active_bridges to see which bridges are used to reformulate the model. Read the Ellipsoid approximation tutorial for more details.","category":"page"},{"location":"manual/models/#Unsafe-backend","page":"Models","title":"Unsafe backend","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"In some advanced use-cases, it is necessary to work with the inner optimization model directly. To access this model, use unsafe_backend:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> backend(model)\nMOIU.CachingOptimizer{MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}, MOIU.UniversalFallback{MOIU.Model{Float64}}}\nin state EMPTY_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.UniversalFallback{MOIU.Model{Float64}}\n  fallback for MOIU.Model{Float64}\nwith optimizer MOIB.LazyBridgeOptimizer{HiGHS.Optimizer}\n  with 0 variable bridges\n  with 0 constraint bridges\n  with 0 objective bridges\n  with inner model A HiGHS model with 0 columns and 0 rows.\n\njulia> unsafe_backend(model)\nA HiGHS model with 0 columns and 0 rows.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"warning: Warning\nbackend and unsafe_backend are advanced routines. Read their docstrings to understand the caveats of their usage, and only call them if you wish to access low-level solver-specific functions.","category":"page"},{"location":"manual/models/#Direct-mode","page":"Models","title":"Direct mode","text":"","category":"section"},{"location":"manual/models/","page":"Models","title":"Models","text":"Using a CachingOptimizer results in an additional copy of the model being stored by JuMP in the .model_cache field. To avoid this overhead, create a JuMP model using direct_model:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = direct_model(HiGHS.Optimizer())\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: DIRECT\nSolver name: HiGHS","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"warning: Warning\nSolvers that do not support incremental modification do not support direct_model. An error will be thrown, telling you to use a CachingOptimizer instead.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"The benefit of using direct_model is that there are no extra layers (for example, Cachingoptimizer or LazyBridgeOptimizer) between model and the provided optimizer:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> backend(model)\nA HiGHS model with 0 columns and 0 rows.","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"A downside of direct mode is that there is no bridging layer. Therefore, only constraints which are natively supported by the solver are supported. For example, HiGHS.jl does not implement quadratic constraints:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"julia> model = direct_model(HiGHS.Optimizer());\n\njulia> set_silent(model)\n\njulia> @variable(model, x[1:2]);\n\njulia> @constraint(model, x[1]^2 + x[2]^2 <= 2)\nERROR: Constraints of type MathOptInterface.ScalarQuadraticFunction{Float64}-in-MathOptInterface.LessThan{Float64} are not supported by the solver.\n\nIf you expected the solver to support your problem, you may have an error in your formulation. Otherwise, consider using a different solver.\n\nThe list of available solvers, along with the problem types they support, is available at https://jump.dev/JuMP.jl/stable/installation/#Supported-solvers.\nStacktrace:","category":"page"},{"location":"manual/models/","page":"Models","title":"Models","text":"warning: Warning\nAnother downside of direct mode is that the behavior of querying solution information after modifying the problem is solver-specific. This can lead to errors, or the solver silently returning an incorrect value. See OptimizeNotCalled errors for more information.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/duality.md\"","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/background/duality/#Duality","page":"Duality","title":"Duality","text":"","category":"section"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Conic duality is the starting point for MOI's duality conventions. When all functions are affine (or coordinate projections), and all constraint sets are closed convex cones, the model may be called a conic optimization problem.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"For a minimization problem in geometric conic form, the primal is:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n min_x in mathbbR^n  a_0^T x + b_0\n\n textst  A_i x + b_i  in mathcalC_i  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"and the dual is a maximization problem in standard conic form:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n max_y_1 ldots y_m  -sum_i=1^m b_i^T y_i + b_0\n\n textst  a_0 - sum_i=1^m A_i^T y_i  = 0\n\n  y_i  in mathcalC_i^*  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"where each mathcalC_i is a closed convex cone and mathcalC_i^* is its dual cone.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"For a maximization problem in geometric conic form, the primal is:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n max_x in mathbbR^n  a_0^T x + b_0\n\n textst  A_i x + b_i  in mathcalC_i  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"and the dual is a minimization problem in standard conic form:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n min_y_1 ldots y_m  sum_i=1^m b_i^T y_i + b_0\n\n textst  a_0 + sum_i=1^m A_i^T y_i  = 0\n\n  y_i  in mathcalC_i^*  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"A linear inequality constraint a^T x + b ge c is equivalent to a^T x + b - c in mathbbR_+, and a^T x + b le c is equivalent to a^T x + b - c in mathbbR_-. Variable-wise constraints are affine constraints with the appropriate identity mapping in place of A_i.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"For the special case of minimization LPs, the MOI primal form can be stated as:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n min_x in mathbbR^n  a_0^T x + b_0\n\n textst\nA_1 x  ge b_1\n A_2 x  le b_2\n A_3 x  = b_3\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"By applying the stated transformations to conic form, taking the dual, and transforming back into linear inequality form, one obtains the following dual:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n max_y_1y_2y_3  b_1^Ty_1 + b_2^Ty_2 + b_3^Ty_3 + b_0\n\n textst\nA_1^Ty_1 + A_2^Ty_2 + A_3^Ty_3  = a_0\n y_1 ge 0\n y_2 le 0\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"For maximization LPs, the MOI primal form can be stated as:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n max_x in mathbbR^n  a_0^T x + b_0\n\n textst\nA_1 x  ge b_1\n A_2 x  le b_2\n A_3 x  = b_3\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"and similarly, the dual is:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n min_y_1y_2y_3  -b_1^Ty_1 - b_2^Ty_2 - b_3^Ty_3 + b_0\n\n textst\nA_1^Ty_1 + A_2^Ty_2 + A_3^Ty_3  = -a_0\n y_1 ge 0\n y_2 le 0\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"warning: Warning\nFor the LP case, the signs of the feasible dual variables depend only on the sense of the corresponding primal inequality and not on the objective sense.","category":"page"},{"location":"moi/background/duality/#Duality-and-scalar-product","page":"Duality","title":"Duality and scalar product","text":"","category":"section"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"The scalar product is different from the canonical one for the sets PositiveSemidefiniteConeTriangle, LogDetConeTriangle, RootDetConeTriangle.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"If the set C_i of the section Duality is one of these three cones, then the rows of the matrix A_i corresponding to off-diagonal entries are twice the value of the coefficients field in the VectorAffineFunction for the corresponding rows. See PositiveSemidefiniteConeTriangle for details.","category":"page"},{"location":"moi/background/duality/#Dual-for-problems-with-quadratic-functions","page":"Duality","title":"Dual for problems with quadratic functions","text":"","category":"section"},{"location":"moi/background/duality/#Quadratic-Programs-(QPs)","page":"Duality","title":"Quadratic Programs (QPs)","text":"","category":"section"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"For quadratic programs with only affine conic constraints,","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign*\n min_x in mathbbR^n  frac12x^TQ_0x + a_0^T x + b_0\n\n textst  A_i x + b_i  in mathcalC_i  i = 1 ldots m\nendalign*","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"with cones mathcalC_i subseteq mathbbR^m_i for i = 1 ldots m, consider the Lagrangian function","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"L(x y) = frac12x^TQ_0x + a_0^T x + b_0 - sum_i = 1^m y_i^T (A_i x + b_i)","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Let z(y) denote sum_i = 1^m A_i^T y_i - a_0, the Lagrangian can be rewritten as","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"L(x y) = frac12x^TQ_0x - z(y)^T x + b_0 - sum_i = 1^m y_i^T b_i","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"The condition nabla_x L(x y) = 0 gives","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"0 = nabla_x L(x y) = Q_0x + a_0 - sum_i = 1^m y_i^T b_i","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"which gives Q_0x = z(y). This allows to obtain that","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"min_x in mathbbR^n L(x y) = -frac12x^TQ_0x + b_0 - sum_i = 1^m y_i^T b_i","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"so the dual problem is","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"max_y_i in mathcalC_i^* min_x in mathbbR^n -frac12x^TQ_0x + b_0 - sum_i = 1^m y_i^T b_i","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"If Q_0 is invertible, we have x = Q_0^-1z(y) hence","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"min_x in mathbbR^n L(x y) = -frac12z(y)^TQ_0^-1z(y) + b_0 - sum_i = 1^m y_i^T b_i","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"so the dual problem is","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"max_y_i in mathcalC_i^* -frac12z(y)^TQ_0^-1z(y) + b_0 - sum_i = 1^m y_i^T b_i","category":"page"},{"location":"moi/background/duality/#Quadratically-Constrained-Quadratic-Programs-(QCQPs)","page":"Duality","title":"Quadratically Constrained Quadratic Programs (QCQPs)","text":"","category":"section"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Given a problem with both quadratic function and quadratic objectives:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign*\n min_x in mathbbR^n  frac12x^TQ_0x + a_0^T x + b_0\n\n textst  frac12x^TQ_ix + a_i^T x + b_i  in mathcalC_i  i = 1 ldots m\nendalign*","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"with cones mathcalC_i subseteq mathbbR for i = 1 ldots m, consider the Lagrangian function","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"L(x y) = frac12x^TQ_0x + a_0^T x + b_0 - sum_i = 1^m y_i (frac12x^TQ_ix + a_i^T x + b_i)","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"A pair of primal-dual variables (x^star y^star) is optimal if","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"x^star is a minimizer of\nmin_x in mathbbR^n L(x y^star)\nThat is,\n0 = nabla_x L(x y^star) = Q_0x + a_0 - sum_i = 1^m y_i^star (Q_ix + a_i)\nand y^star is a maximizer of\nmax_y_i in mathcalC_i^* L(x^star y)\nThat is, for all i = 1 ldots m, frac12x^TQ_ix + a_i^T x + b_i is either zero or in the normal cone of mathcalC_i^* at y^star. For instance, if mathcalC_i is  z in mathbbR  z le 0 , this means that if frac12x^TQ_ix + a_i^T x + b_i is nonzero at x^star then y_i^star = 0. This is the classical complementary slackness condition.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"If mathcalC_i is a vector set, the discussion remains valid with y_i(frac12x^TQ_ix + a_i^T x + b_i) replaced with the scalar product between y_i and the vector of scalar-valued quadratic functions.","category":"page"},{"location":"moi/background/duality/#Dual-for-square-semidefinite-matrices","page":"Duality","title":"Dual for square semidefinite matrices","text":"","category":"section"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"The set PositiveSemidefiniteConeTriangle is a self-dual. That is, querying ConstraintDual of a PositiveSemidefiniteConeTriangle constraint returns a vector that is itself a member of PositiveSemidefiniteConeTriangle.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"However, the dual of PositiveSemidefiniteConeSquare is not so straight forward. This section explains the duality convention we use, and how it is derived.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"info: Info\nIf you have a PositiveSemidefiniteConeSquare constraint, the result matrix A from ConstraintDual is not positive semidefinite. However, A + A^top is positive semidefinite.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Let mathcalS_+ be the cone of symmetric semidefinite matrices in the fracn(n+1)2 dimensional space of symmetric mathbbR^n times n matrices. That is, mathcalS_+ is the set PositiveSemidefiniteConeTriangle. It is well known that mathcalS_+ is a self-dual proper cone.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Let mathcalP_+ be the cone of symmetric semidefinite matrices in the n^2 dimensional space of mathbbR^n times n matrices. That is mathcalP_+ is the set PositiveSemidefiniteConeSquare.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"In addition, let mathcalD_+ be the cone of matrices A such that A+A^top in mathcalP_+.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"mathcalP_+ is not proper because it is not solid (it is not n^2 dimensional), so it is not necessarily true that mathcalP_+^** = mathcalP_+.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"However, this is the case, because we will show that mathcalP_+^* = mathcalD_+ and mathcalD_+^* = mathcalP_+.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"First, let us see why mathcalP_+^* = mathcalD_+.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"If B is symmetric, then","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"langle AB rangle = langle A^top B^top rangle = langle A^top Brangle","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"so","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"2langle A B rangle = langle A B rangle + langle A^top B rangle = langle A + A^top  B rangle","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Therefore, langle ABrangle ge 0 for all B in mathcalP_+ if and only if langle A+A^topBrangle ge 0 for all B in mathcalP_+. Since A+A^top is symmetric, and we know that mathcalS_+ is self-dual, we have shown that mathcalP_+^* is the set of matrices A such that A+A^top in mathcalP_+.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Second, let us see why mathcalD_+^* = mathcalP_+.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Since A in mathcalD_+ implies that A^top in mathcalD_+, B in mathcalD_+^* means that langle A+A^topBrangle ge 0 for all A in mathcalD_+, and hence B in mathcalP_+.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"To see why it should be symmetric, simply notice that if B_ij  B_ji, then langle ABrangle can be made arbitrarily small by setting A_ij = A_ij + s and A_ji = A_ji - s, with s arbitrarily large, and A stays in mathcalD_+ because A+A^top does not change.","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"Typically, the primal/dual pair for semidefinite programs is presented as:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n       min  langle C X rangle \ntextst   langle A_k Xrangle = b_k forall k \n             X                     in mathcalS_+\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"with the dual","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n       max  sum_k b_k y_k \ntextst   C - sum A_k y_k in mathcalS_+\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"If we allow A_k to be non-symmetric, we should instead use:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n       min  langle C X rangle \ntextst   langle A_k Xrangle = b_k forall k \n             X                     in mathcalD_+\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"with the dual","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n       max  sum b_k y_k \ntextst   C - sum A_k y_k in mathcalP_+\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"This is implemented as:","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n       min  langle C   Z rangle + langle C - C^top S rangle \ntextst   langle A_k Z rangle + langle A_k - A_k^top S rangle = b_k forall k \n             Z in mathcalS_+\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"with the dual","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"beginalign\n       max  sum b_k y_k \ntextst   C+C^top - sum (A_k+A_k^top) y_k in mathcalS_+ \n             C-C^top - sum(A_k-A_k^top) y_k = 0\nendalign","category":"page"},{"location":"moi/background/duality/","page":"Duality","title":"Duality","text":"and we recover Z = X + X^top.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"EditURL = \"tips_and_tricks.jl\"","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#conic_tips_and_tricks","page":"Tips and Tricks","title":"Tips and Tricks","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"This tutorial was originally contributed by Arpit Bhatia.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"This tutorial is aimed at providing a simplistic introduction to conic programming using JuMP.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"It uses the following packages:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"using JuMP\nimport SCS\nimport LinearAlgebra","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"info: Info\nThis tutorial uses sets from MathOptInterface. By default, JuMP exports the MOI symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:import MathOptInterface as MOI","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"import Random      # hide\nRandom.seed!(1234) # hide\nnothing            # hide","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"tip: Tip\nA good resource for learning more about functions which can be modeled using cones is the MOSEK Modeling Cookbook.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Background-theory","page":"Tips and Tricks","title":"Background theory","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"A subset C of a vector space V is a cone if forall x in C and positive scalars lambda  0, the product lambda x in C.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"A cone C is a convex cone if lambda x + (1 - lambda) y in C, for any lambda in 0 1, and any x y in C.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"Conic programming problems are convex optimization problems in which a convex function is minimized over the intersection of an affine subspace and a convex cone. An example of a conic-form minimization problems, in the primal form is:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"beginaligned\n min_x in mathbbR^n  a_0^T x + b_0 \n textst  A_i x + b_i  in mathcalC_i  i = 1 ldots m\nendaligned","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The corresponding dual problem is:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"beginaligned\n max_y_1 ldots y_m  -sum_i=1^m b_i^T y_i + b_0 \n textst  a_0 - sum_i=1^m A_i^T y_i  = 0 \n  y_i  in mathcalC_i^*  i = 1 ldots m\nendaligned","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"where each mathcalC_i is a closed convex cone and mathcalC_i^* is its dual cone.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Second-Order-Cone","page":"Tips and Tricks","title":"Second-Order Cone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The SecondOrderCone (or Lorentz Cone) of dimension n is a cone of the form:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"K_soc =  (t x) in mathbbR^n  t ge x_2 ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"It is most commonly used to represent the L2-norm of the vector x:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, x[1:3])\n@variable(model, t)\n@constraint(model, sum(x) == 1)\n@constraint(model, [t; x] in SecondOrderCone())\n@objective(model, Min, t)\noptimize!(model)\nvalue(t), value.(x)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Rotated-Second-Order-Cone","page":"Tips and Tricks","title":"Rotated Second-Order Cone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"A Second-Order Cone rotated by pi4 in the (x_1x_2) plane is called a RotatedSecondOrderCone. It is a cone of the form:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"K_rsoc =  (tux) in mathbbR^n  2tu ge x_2^2 tu ge 0 ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"When u = 0.5, it represents the sum of squares of a vector x:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"data = [1.0, 2.0, 3.0, 4.0]\ntarget = [0.45, 1.04, 1.51, 1.97]\nmodel = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, θ)\n@variable(model, t)\n@expression(model, residuals, θ * data .- target)\n@constraint(model, [t; 0.5; residuals] in RotatedSecondOrderCone())\n@objective(model, Min, t)\noptimize!(model)\nvalue(θ), value(t)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Exponential-Cone","page":"Tips and Tricks","title":"Exponential Cone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The MOI.ExponentialCone is a set of the form:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"K_exp =  (xyz) in mathbbR^3  y exp (xy) le z y  0 ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"It can be used to model problems involving log and exp.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Exponential","page":"Tips and Tricks","title":"Exponential","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"To model exp(x) le z, use (x, 1, z):","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, x == 1.5)\n@variable(model, z)\n@objective(model, Min, z)\n@constraint(model, [x, 1, z] in MOI.ExponentialCone())\noptimize!(model)\nvalue(z), exp(1.5)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Logarithm","page":"Tips and Tricks","title":"Logarithm","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"To model x le log(z), use (x, 1, z):","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, x)\n@variable(model, z == 1.5)\n@objective(model, Max, x)\n@constraint(model, [x, 1, z] in MOI.ExponentialCone())\noptimize!(model)\nvalue(x), log(1.5)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Log-sum-exp","page":"Tips and Tricks","title":"Log-sum-exp","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"To model t ge logleft(sum e^x_iright), use:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"N = 3\nx0 = rand(N)\nmodel = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, x[i = 1:N] == x0[i])\n@variable(model, t)\n@objective(model, Min, t)\n@variable(model, u[1:N])\n@constraint(model, sum(u) <= 1)\n@constraint(model, [i = 1:N], [x[i] - t, 1, u[i]] in MOI.ExponentialCone())\noptimize!(model)\nvalue(t), log(sum(exp.(x0)))","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Entropy","page":"Tips and Tricks","title":"Entropy","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The entropy maximization problem consists of maximizing the entropy function, H(x) = -xlogx subject to linear inequality constraints.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"beginaligned\n max  - sum_i=1^n x_i log x_i \n textst  mathbf1^top x = 1 \n  Ax leq b\nendaligned","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"We can model this problem using an exponential cone by using the following transformation:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"tleq -xlogx iff tleq xlog(1x)  iff (t x 1) in K_exp","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"Thus, our problem becomes,","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"beginaligned\n max  1^Tt \n textst  Ax leq b \n  1^T x = 1 \n  (t_i x_i 1) in K_exp  forall i = 1 ldots n \nendaligned","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"m, n = 10, 15\nA, b = randn(m, n), rand(m, 1)\nmodel = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t[1:n])\n@variable(model, x[1:n])\n@objective(model, Max, sum(t))\n@constraint(model, sum(x) == 1)\n@constraint(model, A * x .<= b)\n@constraint(model, [i = 1:n], [t[i], x[i], 1] in MOI.ExponentialCone())\noptimize!(model)\nobjective_value(model)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The MOI.ExponentialCone has a dual, the MOI.DualExponentialCone, that offers an alternative formulation that can be more efficient for some formulations.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"There is also the MOI.RelativeEntropyCone for explicitly encoding the relative entropy function","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t)\n@variable(model, x[1:n])\n@objective(model, Max, -t)\n@constraint(model, sum(x) == 1)\n@constraint(model, A * x .<= b)\n@constraint(model, [t; ones(n); x] in MOI.RelativeEntropyCone(2n + 1))\noptimize!(model)\nobjective_value(model)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#PowerCone","page":"Tips and Tricks","title":"PowerCone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The MOI.PowerCone is a three-dimensional set parameterized by a scalar value α. It has the form:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"K_p =  (xyz) in mathbbR^3  x^alpha y^1-alpha ge z x ge 0 y ge 0 ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The power cone permits a number of reformulations. For example, when p  1, we can model t ge x^p using the power cone (t 1 x) with alpha = 1  p. Thus, to model t ge x^3 with x ge 0","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t)\n@variable(model, x >= 1.5)\n@constraint(model, [t, 1, x] in MOI.PowerCone(1 / 3))\n@objective(model, Min, t)\noptimize!(model)\nvalue(t), value(x)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The MOI.PowerCone has a dual, the MOI.DualPowerCone, that offers an alternative formulation that can be more efficient for some formulations.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#P-Norm","page":"Tips and Tricks","title":"P-Norm","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The p-norm x_p = left(sumlimits_i x_i^pright)^frac1p can be modeled using MOI.PowerCones. See the Mosek Modeling Cookbook for the derivation.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"function p_norm(x::Vector, p)\n    N = length(x)\n    model = Model(SCS.Optimizer)\n    set_silent(model)\n    @variable(model, r[1:N])\n    @variable(model, t)\n    @constraint(model, [i = 1:N], [r[i], t, x[i]] in MOI.PowerCone(1 / p))\n    @constraint(model, sum(r) == t)\n    @objective(model, Min, t)\n    optimize!(model)\n    return value(t)\nend\n\nx = rand(5);\nLinearAlgebra.norm(x, 4), p_norm(x, 4)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Positive-Semidefinite-Cone","page":"Tips and Tricks","title":"Positive Semidefinite Cone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The set of positive semidefinite matrices (PSD) of dimension n form a cone in mathbbR^n. We write this set mathematically as:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"mathcalS_+^n =  X in mathcalS^n mid z^T X z geq 0  forall zin mathbbR^n ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"A PSD cone is represented in JuMP using the MOI sets PositiveSemidefiniteConeTriangle (for upper triangle of a PSD matrix) and PositiveSemidefiniteConeSquare (for a complete PSD matrix). However, it is preferable to use the PSDCone shortcut as illustrated below.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Example:-largest-eigenvalue-of-a-symmetric-matrix","page":"Tips and Tricks","title":"Example: largest eigenvalue of a symmetric matrix","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"Suppose A has eigenvalues lambda_1 geq lambda_2 ldots geq lambda_n. Then the matrix t I-A has eigenvalues t-lambda_1 t-lambda_2 ldots t-lambda_n. Note that t I-A is PSD exactly when all these eigenvalues are non-negative, and this happens for values t geq lambda_1. Thus, we can model the problem of finding the largest eigenvalue of a symmetric matrix as:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"beginaligned\nlambda_1 = min t \ntext  st  t I-A succeq 0\nendaligned","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"A = [3 2 4; 2 0 2; 4 2 3]\nI = Matrix{Float64}(LinearAlgebra.I, 3, 3)\nmodel = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t)\n@objective(model, Min, t)\n@constraint(model, t .* I - A in PSDCone())\noptimize!(model)\nobjective_value(model)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#GeometricMeanCone","page":"Tips and Tricks","title":"GeometricMeanCone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The MOI.GeometricMeanCone is a cone of the form:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"K_geo =  (t x) in mathbbR^n  x ge 0 t le sqrtn-1x_1 x_2 cdots x_n-1 ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, x[1:4])\n@variable(model, t)\n@constraint(model, sum(x) == 1)\n@constraint(model, [t; x] in MOI.GeometricMeanCone(5))\noptimize!(model)\nvalue(t), value.(x)","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#RootDetCone","page":"Tips and Tricks","title":"RootDetCone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The MOI.RootDetConeSquare is a cone of the form:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"K =  (t X) in mathbbR^1+d^2  t le det(X)^frac1d ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t)\n@variable(model, X[1:2, 1:2])\n@objective(model, Max, t)\n@constraint(model, [t; vec(X)] in MOI.RootDetConeSquare(2))\n@constraint(model, X .== [2 1; 1 3])\noptimize!(model)\nvalue(t), sqrt(LinearAlgebra.det(value.(X)))","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"If X is symmetric, then you can use MOI.RootDetConeTriangle instead. This can be more efficient because the solver does not need to add additional constraints to ensure X is symmetric.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"When forming the function, use triangle_vec to obtain the column-wise upper triangle of the matrix as a vector in the order that JuMP requires.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t)\n@variable(model, X[1:2, 1:2], Symmetric)\n@objective(model, Max, t)\n@constraint(model, [t; triangle_vec(X)] in MOI.RootDetConeTriangle(2))\n@constraint(model, X .== [2 1; 1 3])\noptimize!(model)\nvalue(t), sqrt(LinearAlgebra.det(value.(X)))","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#LogDetCone","page":"Tips and Tricks","title":"LogDetCone","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"The MOI.LogDetConeSquare is a cone of the form:","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"K =  (t u X) in mathbbR^2+d^2  t le u log(det(X  u)) ","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t)\n@variable(model, u)\n@variable(model, X[1:2, 1:2])\n@objective(model, Max, t)\n@constraint(model, [t; u; vec(X)] in MOI.LogDetConeSquare(2))\n@constraint(model, X .== [2 1; 1 3])\n@constraint(model, u == 0.5)\noptimize!(model)\nvalue(t), 0.5 * log(LinearAlgebra.det(value.(X) ./ 0.5))","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"If X is symmetric, then you can use MOI.LogDetConeTriangle instead. This can be more efficient because the solver does not need to add additional constraints to ensure X is symmetric.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"When forming the function, use triangle_vec to obtain the column-wise upper triangle of the matrix as a vector in the order that JuMP requires.","category":"page"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"model = Model(SCS.Optimizer)\nset_silent(model)\n@variable(model, t)\n@variable(model, u)\n@variable(model, X[1:2, 1:2], Symmetric)\n@objective(model, Max, t)\n@constraint(model, [t; u; triangle_vec(X)] in MOI.LogDetConeTriangle(2))\n@constraint(model, X .== [2 1; 1 3])\n@constraint(model, u == 0.5)\noptimize!(model)\nvalue(t), 0.5 * log(LinearAlgebra.det(value.(X) ./ 0.5))","category":"page"},{"location":"tutorials/conic/tips_and_tricks/#Other-Cones-and-Functions","page":"Tips and Tricks","title":"Other Cones and Functions","text":"","category":"section"},{"location":"tutorials/conic/tips_and_tricks/","page":"Tips and Tricks","title":"Tips and Tricks","text":"For other cones supported by JuMP, check out the MathOptInterface Manual.","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"EditURL = \"https://github.com/jump-dev/BARON.jl/blob/v0.8.2/README.md\"","category":"page"},{"location":"packages/BARON/#BARON.jl","page":"jump-dev/BARON.jl","title":"BARON.jl","text":"","category":"section"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"BARON.jl is a wrapper for BARON by The Optimization Firm.","category":"page"},{"location":"packages/BARON/#Affiliation","page":"jump-dev/BARON.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"This wrapper is maintained by the JuMP community and is not officially supported by The Optimization Firm.","category":"page"},{"location":"packages/BARON/#License","page":"jump-dev/BARON.jl","title":"License","text":"","category":"section"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"BARON.jl is licensed under the MIT License.","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"The underlying solver is a closed-source commercial product for which you must obtain a license from The Optimization Firm, although a small trial version is available for free.","category":"page"},{"location":"packages/BARON/#Installation","page":"jump-dev/BARON.jl","title":"Installation","text":"","category":"section"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"First, download a copy of the BARON solver and unpack the executable in a location of your choosing.","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"Once installed, set the BARON_EXEC environment variable pointing to the BARON executable (full path, including file name as it differs across platforms), and run Pkg.add(\"BARON\"). For example:","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"ENV[\"BARON_EXEC\"] = \"/path/to/baron.exe\"\nusing Pkg\nPkg.add(\"BARON\")","category":"page"},{"location":"packages/BARON/#Use-with-JuMP","page":"jump-dev/BARON.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"using JuMP, BARON\nmodel = Model(BARON.Optimizer)","category":"page"},{"location":"packages/BARON/#MathOptInterface-API","page":"jump-dev/BARON.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"The BARON optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarNonlinearFunction}","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"MOI.Reals","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarNonlinearFunction in MOI.EqualTo{Float64}\nMOI.ScalarNonlinearFunction in MOI.GreaterThan{Float64}\nMOI.ScalarNonlinearFunction in MOI.Interval{Float64}\nMOI.ScalarNonlinearFunction in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.ZeroOne","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/BARON/","page":"jump-dev/BARON.jl","title":"jump-dev/BARON.jl","text":"MOI.NLPBlock()\nMOI.ObjectiveSense()","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"EditURL = \"getting_started_with_julia.jl\"","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Getting-started-with-Julia","page":"Getting started with Julia","title":"Getting started with Julia","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Because JuMP is embedded in Julia, knowing some basic Julia is important before you start learning JuMP.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nThis tutorial is designed to provide a minimalist crash course in the basics of Julia. You can find resources that provide a more comprehensive introduction to Julia here.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Installing-Julia","page":"Getting started with Julia","title":"Installing Julia","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"To install Julia, download the latest stable release, then follow the platform specific install instructions.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nUnless you know otherwise, you probably want the 64-bit version.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Next, you need an IDE to develop in. VS Code is a popular choice, so follow these install instructions.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Julia can also be used with Jupyter notebooks or the reactive notebooks of Pluto.jl.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#The-Julia-REPL","page":"Getting started with Julia","title":"The Julia REPL","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"The main way of interacting with Julia is via its REPL (Read Evaluate Print Loop). To access the REPL, start the Julia executable to arrive at the julia> prompt, and then start coding:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"1 + 1","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"As your programs become larger, write a script as a text file, and then run that file using:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"julia> include(\"path/to/file.jl\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"warning: Warning\nBecause of Julia's startup latency, running scripts from the command line like the following is slow:$ julia path/to/file.jlUse the REPL or a notebook instead, and read The \"time-to-first-solve\" issue for more information.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Code-blocks-in-this-documentation","page":"Getting started with Julia","title":"Code blocks in this documentation","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"In this documentation you'll see a mix of code examples with and without the julia>.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"The Julia prompt is mostly used to demonstrate short code snippets, and the output is exactly what you will see if run from the REPL.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Blocks without the julia> can be copy-pasted into the REPL, but they are used because they enable richer output like plots or LaTeX to be displayed in the online and PDF versions of the documentation. If you run them from the REPL you may see different output.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Where-to-get-help","page":"Getting started with Julia","title":"Where to get help","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Read the documentation\nJuMP https://jump.dev/JuMP.jl/stable/\nJulia https://docs.julialang.org/en/v1/\nAsk (or browse) the Julia community forum: https://discourse.julialang.org\nIf the question is JuMP-related, ask in the Optimization (Mathematical) section, or tag your question with \"jump\"","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"To access the built-in help at the REPL, type ? to enter help-mode, followed by the name of the function to lookup:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"help?> print\nsearch: print println printstyled sprint isprint prevind parentindices precision escape_string\n\n  print([io::IO], xs...)\n\n  Write to io (or to the default output stream stdout if io is not given) a canonical\n  (un-decorated) text representation. The representation used by print includes minimal formatting\n  and tries to avoid Julia-specific details.\n\n  print falls back to calling show, so most types should just define show. Define print if your\n  type has a separate \"plain\" representation. For example, show displays strings with quotes, and\n  print displays strings without quotes.\n\n  string returns the output of print as a string.\n\n  Examples\n  ≡≡≡≡≡≡≡≡≡≡\n\n  julia> print(\"Hello World!\")\n  Hello World!\n  julia> io = IOBuffer();\n\n  julia> print(io, \"Hello\", ' ', :World!)\n\n  julia> String(take!(io))\n  \"Hello World!\"","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Numbers-and-arithmetic","page":"Getting started with Julia","title":"Numbers and arithmetic","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Since we want to solve optimization problems, we're going to be using a lot of math. Luckily, Julia is great for math, with all the usual operators:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"1 + 1\n1 - 2\n2 * 2\n4 / 5\n3^2","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Did you notice how Julia didn't print .0 after some of the numbers? Julia is a dynamic language, which means you never have to explicitly declare the type of a variable. However, in the background, Julia is giving each variable a type. Check the type of something using the typeof function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"typeof(1)\ntypeof(1.0)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Here 1 is an Int64, which is an integer with 64 bits of precision, and 1.0 is a Float64, which is a floating point number with 64-bits of precision.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nIf you aren't familiar with floating point numbers, make sure to read the Floating point numbers section.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"We create complex numbers using im:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"x = 2 + 1im\nreal(x)\nimag(x)\ntypeof(x)\nx * (1 - 2im)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"info: Info\nThe curly brackets surround what we call the parameters of a type. You can read Complex{Int64}  as \"a complex number, where the real and imaginary parts are represented by Int64.\" If we call typeof(1.0 + 2.0im) it will be Complex{Float64}, which a complex number with the parts represented by Float64.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"There are also some cool things like an irrational representation of π.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"π","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nTo make π (and most other Greek letters), type \\pi and then press [TAB].","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"typeof(π)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"However, if we do math with irrational numbers, they get converted to Float64:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"typeof(2π / 3)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Floating-point-numbers","page":"Getting started with Julia","title":"Floating point numbers","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"warning: Warning\nIf you aren't familiar with floating point numbers, make sure to read this section carefully.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A Float64 is a floating point approximation of a real number using 64-bits of information.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Because it is an approximation, things we know hold true in mathematics don't hold true in a computer. For example:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"0.1 * 3 == 0.3","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A more complicated example is:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"sin(2π / 3) == √3 / 2","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nGet √ by typing \\sqrt then press [TAB].","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Let's see what the differences are:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"0.1 * 3 - 0.3\nsin(2π / 3) - √3 / 2","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"They are small, but not zero.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"One way of explaining this difference is to consider how we would write 1 / 3 and 2 / 3 using only four digits after the decimal point. We would write 1 / 3 as 0.3333, and 2 / 3 as 0.6667. So, despite the fact that 2 * (1 / 3) == 2 / 3, 2 * 0.3333 == 0.6666 != 0.6667.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Let's try that again using ≈ (\\approx + [TAB]) instead of ==:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"0.1 * 3 ≈ 0.3\nsin(2π / 3) ≈ √3 / 2","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"≈ is a clever way of calling the isapprox function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"isapprox(sin(2π / 3), √3 / 2; atol = 1e-8)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"warning: Warning\nFloating point is the reason solvers use tolerances when they solve optimization models. A common mistake you're likely to make is checking whether a binary variable is 0 using value(z) == 0. Always remember to use something like isapprox when comparing floating point numbers.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Note that isapprox will always return false if one of the number being compared is 0 and atol is zero (its default value).","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"1e-300 ≈ 0.0","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"so always set a nonzero value of atol if one of the arguments can be zero.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"isapprox(1e-9, 0.0; atol = 1e-8)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nGurobi has a good series of articles on the implications of floating point in optimization if you want to read more.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"If you aren't careful, floating point arithmetic can throw up all manner of issues. For example:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"1 + 1e-16 == 1","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"It even turns out that floating point numbers aren't associative:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"(1 + 1e-16) - 1e-16 == 1 + (1e-16 - 1e-16)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"It's important to note that this issue isn't Julia-specific. It happens in every programming language (try it out in Python).","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Vectors,-matrices-and-arrays","page":"Getting started with Julia","title":"Vectors, matrices and arrays","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Similar to MATLAB, Julia has native support for vectors, matrices and tensors; all of which are represented by arrays of different dimensions. Vectors are constructed by comma-separated elements surrounded by square brackets:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"b = [5, 6]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Matrices can be constructed with spaces separating the columns, and semicolons separating the rows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A = [1.0 2.0; 3.0 4.0]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"We can do linear algebra:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"x = A \\ b","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"info: Info\nHere is floating point at work again; x is approximately [-4, 4.5].","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A * x\nA * x ≈ b","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Note that when multiplying vectors and matrices, dimensions matter. For example, you can't multiply a vector by a vector:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"try                         #hide\n    b * b\ncatch err                   #hide\n    showerror(stderr, err)  #hide\nend                         #hide","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"But multiplying transposes works:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"b' * b\nb * b'","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Other-common-types","page":"Getting started with Julia","title":"Other common types","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/#Comments","page":"Getting started with Julia","title":"Comments","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Although not technically a type, code comments begin with the # character:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"1 + 1  # This is a comment","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Multiline comments begin with #= and end with =#:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"#=\nHere is a\nmultiline comment\n=#","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Comments can even be nested inside expressions. This is sometimes helpful when documenting inputs to functions:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"isapprox(\n    sin(π),\n    0.0;\n    #= We need an explicit atol here because we are comparing with 0 =#\n    atol = 0.001,\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Strings","page":"Getting started with Julia","title":"Strings","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Double quotes are used for strings:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"typeof(\"This is Julia\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Unicode is fine in strings:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"typeof(\"π is about 3.1415\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Use println to print a string:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"println(\"Hello, World!\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Use $() to interpolate values into a string:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"x = 123\nprintln(\"The value of x is: $(x)\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Use triple-quotes for multiline strings:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"s = \"\"\"\nHere is\na\nmultiline string\n\"\"\"\n\nprintln(s)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Symbols","page":"Getting started with Julia","title":"Symbols","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Julia Symbols are a data structure from the compiler that represent Julia identifiers (that is, variable names).","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"println(\"The value of x is: $(eval(:x))\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"warning: Warning\nWe used eval here to demonstrate how Julia links Symbols to variables. However, avoid calling eval in your code. It is usually a sign that your code is doing something that could be more easily achieved a different way. The Community Forum is a good place to ask for advice on alternative approaches.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"typeof(:x)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"You can think of a Symbol as a String that takes up less memory, and that can't be modified.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Convert between String and Symbol using their constructors:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"String(:abc)\nSymbol(\"abc\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nSymbols are often (ab)used to stand in for a String or an Enum, when one of the latter is likely a better choice. The JuMP Style guide recommends reserving Symbols for identifiers. See @enum vs. Symbol for more.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Tuples","page":"Getting started with Julia","title":"Tuples","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Julia makes extensive use of a simple data structure called Tuples. Tuples are immutable collections of values. For example:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"t = (\"hello\", 1.2, :foo)\ntypeof(t)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Tuples can be accessed by index, similar to arrays:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"t[2]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"And they can be \"unpacked\" like so:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"a, b, c = t\nb","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"The values can also be given names, which is a convenient way of making light-weight data structures.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"t = (word = \"hello\", num = 1.2, sym = :foo)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Values can be accessed using dot syntax:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"t.word","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Dictionaries","page":"Getting started with Julia","title":"Dictionaries","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Similar to Python, Julia has native support for dictionaries. Dictionaries provide a very generic way of mapping keys to values.  For example, a map of integers to strings:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"d1 = Dict(1 => \"A\", 2 => \"B\", 4 => \"D\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"info: Info\nType-stuff again: Dict{Int64,String} is a dictionary with Int64 keys and String values.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Looking up a value uses the bracket syntax:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"d1[2]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Dictionaries support non-integer keys and can mix data types:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Dict(\"A\" => 1, \"B\" => 2.5, \"D\" => 2 - 3im)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"info: Info\nJulia types form a hierarchy. Here the value type of the dictionary is Number, which is a generalization of Int64, Float64, and Complex{Int}. Leaf nodes in this hierarchy are called \"concrete\" types, and all others are called \"Abstract.\" In general, having variables with abstract types like Number can lead to slower code, so you should try to make sure every element in a dictionary or vector is the same type. For example, in this case we could represent every element as a Complex{Float64}:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Dict(\"A\" => 1.0 + 0.0im, \"B\" => 2.5 + 0.0im, \"D\" => 2.0 - 3.0im)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Dictionaries can be nested:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"d2 = Dict(\"A\" => 1, \"B\" => 2, \"D\" => Dict(:foo => 3, :bar => 4))\nd2[\"B\"]\nd2[\"D\"][:foo]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Structs","page":"Getting started with Julia","title":"Structs","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"You can define custom datastructures with struct:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"struct MyStruct\n    x::Int\n    y::String\n    z::Dict{Int,Int}\nend\n\na = MyStruct(1, \"a\", Dict(2 => 3))\na.x","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"By default, these are not mutable","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"try                         #hide\n    a.x = 2\ncatch err                   #hide\n    showerror(stderr, err)  #hide\nend                         #hide","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"However, you can declare a mutable struct which is mutable:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"mutable struct MyStructMutable\n    x::Int\n    y::String\n    z::Dict{Int,Int}\nend\n\na = MyStructMutable(1, \"a\", Dict(2 => 3))\na.x\na.x = 2\na","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Loops","page":"Getting started with Julia","title":"Loops","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Julia has native support for for-each style loops with the syntax for <value> in <collection> end:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"for i in 1:5\n    println(i)\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"info: Info\nRanges are constructed as start:stop, or start:step:stop.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"for i in 1.2:1.1:5.6\n    println(i)\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"This for-each loop also works with dictionaries:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"for (key, value) in Dict(\"A\" => 1, \"B\" => 2.5, \"D\" => 2 - 3im)\n    println(\"$(key): $(value)\")\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Note that in contrast to vector languages like MATLAB and R, loops do not result in a significant performance degradation in Julia.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Control-flow","page":"Getting started with Julia","title":"Control flow","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Julia control flow is similar to MATLAB, using the keywords if-elseif-else-end, and the logical operators || and && for or and and respectively:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"for i in 0:5:15\n    if i < 5\n        println(\"$(i) is less than 5\")\n    elseif i < 10\n        println(\"$(i) is less than 10\")\n    else\n        if i == 10\n            println(\"the value is 10\")\n        else\n            println(\"$(i) is bigger than 10\")\n        end\n    end\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Comprehensions","page":"Getting started with Julia","title":"Comprehensions","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Similar to languages like Haskell and Python, Julia supports the use of simple loops in the construction of arrays and dictionaries, called comprehensions.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A list of increasing integers:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"[i for i in 1:5]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Matrices can be built by including multiple indices:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"[i * j for i in 1:5, j in 5:10]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Conditional statements can be used to filter out some values:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"[i for i in 1:10 if i % 2 == 1]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A similar syntax can be used for building dictionaries:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Dict(\"$(i)\" => i for i in 1:10 if i % 2 == 1)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Functions","page":"Getting started with Julia","title":"Functions","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A simple function is defined as follows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"function print_hello()\n    return println(\"hello\")\nend\nprint_hello()","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Arguments can be added to a function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"function print_it(x)\n    return println(x)\nend\nprint_it(\"hello\")\nprint_it(1.234)\nprint_it(:my_id)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Optional keyword arguments are also possible:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"function print_it(x; prefix = \"value:\")\n    return println(\"$(prefix) $(x)\")\nend\nprint_it(1.234)\nprint_it(1.234; prefix = \"val:\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"The keyword return is used to specify the return values of a function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"function mult(x; y = 2.0)\n    return x * y\nend\n\nmult(4.0)\nmult(4.0; y = 5.0)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Anonymous-functions","page":"Getting started with Julia","title":"Anonymous functions","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"The syntax input -> output creates an anonymous function. These are most useful when passed to other functions. For example:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"f = x -> x^2\nf(2)\nmap(x -> x^2, 1:4)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Type-parameters","page":"Getting started with Julia","title":"Type parameters","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"We can constrain the inputs to a function using type parameters, which are :: followed by the type of the input we want. For example:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"function foo(x::Int)\n    return x^2\nend\n\nfunction foo(x::Float64)\n    return exp(x)\nend\n\nfunction foo(x::Number)\n    return x + 1\nend\n\nfoo(2)\nfoo(2.0)\nfoo(1 + 1im)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"But what happens if we call foo with something we haven't defined it for?","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"try                         #hide\n    foo([1, 2, 3])\ncatch err                   #hide\n    showerror(stderr, err)  #hide\nend                         #hide","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"A MethodError means that you passed a function something that didn't match the type that it was expecting. In this case, the error message says that it doesn't know how to handle an Vector{Int64}, but it does know how to handle Float64, Int64, and Number.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nRead the \"Closest candidates\" part of the error message carefully to get a hint as to what was expected.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Broadcasting","page":"Getting started with Julia","title":"Broadcasting","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"In the example above, we didn't define what to do if f was passed a Vector. Luckily, Julia provides a convenient syntax for mapping f element-wise over arrays. Just add a . between the name of the function and the opening (. This works for any function, including functions with multiple arguments. For example:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"f.([1, 2, 3])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nGet a MethodError when calling a function that takes a Vector, Matrix, or Array? Try broadcasting.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Mutable-vs-immutable-objects","page":"Getting started with Julia","title":"Mutable vs immutable objects","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Some types in Julia are mutable, which means you can change the values inside them. A good example is an array. You can modify the contents of an array without having to make a new array.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"In contrast, types like Float64 are immutable. You cannot modify the contents of a Float64.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"This is something to be aware of when passing types into functions. For example:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"function mutability_example(mutable_type::Vector{Int}, immutable_type::Int)\n    mutable_type[1] += 1\n    immutable_type += 1\n    return\nend\n\nmutable_type = [1, 2, 3]\nimmutable_type = 1\n\nmutability_example(mutable_type, immutable_type)\n\nprintln(\"mutable_type: $(mutable_type)\")\nprintln(\"immutable_type: $(immutable_type)\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Because Vector{Int} is a mutable type, modifying the variable inside the function changed the value outside of the function. In contrast, the change to immutable_type didn't modify the value outside the function.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"You can check mutability with the isimmutable function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"isimmutable([1, 2, 3])\nisimmutable(1)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#The-package-manager","page":"Getting started with Julia","title":"The package manager","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/#Installing-packages","page":"Getting started with Julia","title":"Installing packages","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"No matter how wonderful Julia's base language is, at some point you will want to use an extension package.  Some of these are built-in, for example random number generation is available in the Random package in the standard library. These packages are loaded with the commands using and import.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"using Random  # The equivalent of Python's `from Random import *`\nimport Random  # The equivalent of Python's `import Random`\n\nRandom.seed!(33)\n\n[rand() for i in 1:10]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"The Package Manager is used to install packages that are not part of Julia's standard library.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"For example the following can be used to install JuMP,","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"using Pkg\nPkg.add(\"JuMP\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"For a complete list of registered Julia packages see the package listing at JuliaHub.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"From time to you may wish to use a Julia package that is not registered.  In this case a git repository URL can be used to install the package.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"using Pkg\nPkg.add(\"https://github.com/user-name/MyPackage.jl.git\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/#Package-environments","page":"Getting started with Julia","title":"Package environments","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"By default, Pkg.add will add packages to Julia's global environment. However, Julia also has built-in support for virtual environments.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"Activate a virtual environment with:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"import Pkg; Pkg.activate(\"/path/to/environment\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"You can see what packages are installed in the current environment with Pkg.status().","category":"page"},{"location":"tutorials/getting_started/getting_started_with_julia/","page":"Getting started with Julia","title":"Getting started with Julia","text":"tip: Tip\nWe strongly recommend you create a Pkg environment for each project that you create in Julia, and add only the packages that you need, instead of adding lots of packages to the global environment. The Pkg manager documentation has more information on this topic.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"EditURL = \"tips_and_tricks.jl\"","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#linear_tips_and_tricks","page":"Tips and tricks","title":"Tips and tricks","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This tutorial was originally contributed by Arpit Bhatia.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"tip: Tip\nA good source of tips is the Mosek Modeling Cookbook.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This tutorial collates some tips and tricks you can use when formulating mixed-integer programs. It uses the following packages:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"using JuMP","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Absolute-value","page":"Tips and tricks","title":"Absolute value","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"To model the absolute value function t ge x, there are a few options. In all cases, these reformulations only work if you are minimizing t \"down\" into x. They do not work if you are trying to maximize x.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Option-1","page":"Tips and tricks","title":"Option 1","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This option adds two linear inequality constraints:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x)\n@variable(model, t)\n@constraint(model, t >= x)\n@constraint(model, t >= -x)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Option-2","page":"Tips and tricks","title":"Option 2","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This option uses two non-negative variables and forms expressions for x and t:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, z[1:2] >= 0)\n@expression(model, t, z[1] + z[2])\n@expression(model, x, z[1] - z[2])","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Option-3","page":"Tips and tricks","title":"Option 3","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This option uses MOI.NormOneCone and lets JuMP choose the reformulation:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x)\n@variable(model, t)\n@constraint(model, [t; x] in MOI.NormOneCone(2))","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#L1-norm","page":"Tips and tricks","title":"L1-norm","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"To model min x_1, that is, min sumlimits_i x_i, use the MOI.NormOneCone:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:3])\n@variable(model, t)\n@constraint(model, [t; x] in MOI.NormOneCone(1 + length(x)))\n@objective(model, Min, t)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Infinity-norm","page":"Tips and tricks","title":"Infinity-norm","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"To model min x_infty, that is, min maxlimits_i x_i, use the MOI.NormInfinityCone:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:3])\n@variable(model, t)\n@constraint(model, [t; x] in MOI.NormInfinityCone(1 + length(x)))\n@objective(model, Min, t)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Max","page":"Tips and tricks","title":"Max","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"To model t ge maxx y, do:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, t)\n@variable(model, x)\n@variable(model, y)\n@constraint(model, t >= x)\n@constraint(model, t >= y)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This reformulation does not work for t ge minx y.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Min","page":"Tips and tricks","title":"Min","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"To model t le minx y, do:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, t)\n@variable(model, x)\n@variable(model, y)\n@constraint(model, t <= x)\n@constraint(model, t <= y)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This reformulation does not work for t le maxx y.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Boolean-operators","page":"Tips and tricks","title":"Boolean operators","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Binary variables can be used to construct logical operators. Here are some example.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Or","page":"Tips and tricks","title":"Or","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"x_3 = x_1 lor x_2","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:3], Bin)\n@constraints(model, begin\n    x[1] <= x[3]\n    x[2] <= x[3]\n    x[3] <= x[1] + x[2]\nend)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#And","page":"Tips and tricks","title":"And","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"x_3 = x_1 land x_2","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:3], Bin)\n@constraints(model, begin\n    x[3] <= x[1]\n    x[3] <= x[2]\n    x[3] >= x[1] + x[2] - 1\nend)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Not","page":"Tips and tricks","title":"Not","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"x_1 neg x_2","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:2], Bin)\n@constraint(model, x[1] == 1 - x[2])","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Implies","page":"Tips and tricks","title":"Implies","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"x_1 implies x_2","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:2], Bin)\n@constraint(model, x[1] <= x[2])","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Disjunctions","page":"Tips and tricks","title":"Disjunctions","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/#Problem","page":"Tips and tricks","title":"Problem","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Suppose that we have two constraints a^top x leq b and c^top x leq d, and we want at least one to hold.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Trick","page":"Tips and tricks","title":"Trick","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Introduce a \"big-M\" multiplied by a binary variable to relax one of the constraints.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Example Either x_1 leq 1 or x_2 leq 2.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:2])\n@variable(model, y, Bin)\nM = 100\n@constraint(model, x[1] <= 1 + M * y)\n@constraint(model, x[2] <= 2 + M * (1 - y))","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"warning: Warning\nIf M is too small, the solution may be suboptimal. If M is too big, the solver may encounter numerical issues. Try to use domain knowledge to choose an M that is just right. Gurobi has a good documentation section on this topic.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Indicator-constraints","page":"Tips and tricks","title":"Indicator constraints","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/#Problem-2","page":"Tips and tricks","title":"Problem","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Suppose we want to model that a certain linear inequality must be satisfied when some other event occurs, that is, for a binary variable z, we want to model the implication:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"z = 1 implies a^top x leq b","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Trick-1","page":"Tips and tricks","title":"Trick 1","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Some solvers have native support for indicator constraints.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Example x_1 + x_2 leq 1 if z = 1.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:2])\n@variable(model, z, Bin)\n@constraint(model, z => {sum(x) <= 1})","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Example x_1 + x_2 leq 1 if z = 0.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:2])\n@variable(model, z, Bin)\n@constraint(model, !z => {sum(x) <= 1})","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Trick-2","page":"Tips and tricks","title":"Trick 2","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"If the solver doesn't support indicator constraints, you an use the big-M trick.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Example x_1 + x_2 leq 1 if z = 1.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:2])\n@variable(model, z, Bin)\nM = 100\n@constraint(model, sum(x) <= 1 + M * (1 - z))","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Example x_1 + x_2 leq 1 if z = 0.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:2])\n@variable(model, z, Bin)\nM = 100\n@constraint(model, sum(x) <= 1 + M * z)","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Semi-continuous-variables","page":"Tips and tricks","title":"Semi-continuous variables","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"info: Info\nThis section uses sets from MathOptInterface. By default, JuMP exports the MOI symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:import MathOptInterface as MOI","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"A semi-continuous variable is a continuous variable between bounds lu that also can assume the value zero, that is: x in 0 cup lu","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Example x in 0cup 1 2","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x in Semicontinuous(1.0, 2.0))","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Semi-integer-variables","page":"Tips and tricks","title":"Semi-integer variables","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"A semi-integer variable is a variable which assumes integer values between bounds lu and can also assume the value zero: x in 0 cup l u cap mathbbZ","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x in Semiinteger(5.0, 10.0))","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Special-Ordered-Sets-of-Type-1","page":"Tips and tricks","title":"Special Ordered Sets of Type 1","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"A Special Ordered Set of Type 1 is a set of variables, at most one of which can take a non-zero value, all others being at 0.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"They most frequently apply where a set of variables are actually binary variables. In other words, we have to choose at most one from a set of possibilities.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:3], Bin)\n@constraint(model, x in SOS1())","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"You can optionally pass SOS1 a weight vector like","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"@constraint(model, x in SOS1([0.2, 0.5, 0.3]))","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"If the decision variables are related and have a physical ordering, then the weight vector, although not used directly in the constraint, can help the solver make a better decision in the solution process.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#tip_sos2","page":"Tips and tricks","title":"Special Ordered Sets of Type 2","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"A Special Ordered Set of type 2 is a set of non-negative variables, of which at most two can be non-zero, and if two are non-zero these must be consecutive in their ordering.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model();\n@variable(model, x[1:3])\n@constraint(model, x in SOS2([3.0, 1.0, 2.0]))","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"The ordering provided by the weight vector is more important in this case as the variables need to be consecutive according to the ordering. For example, in the above constraint, the possible pairs are:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Consecutive\n(x[1] and x[3]) as they correspond to 3 and 2 resp. and thus can be non-zero\n(x[2] and x[3]) as they correspond to 1 and 2 resp. and thus can be non-zero\nNon-consecutive\n(x[1] and x[2]) as they correspond to 3 and 1 resp. and thus cannot be non-zero","category":"page"},{"location":"tutorials/linear/tips_and_tricks/#Piecewise-linear-approximations","page":"Tips and tricks","title":"Piecewise linear approximations","text":"","category":"section"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"SOSII constraints are most often used to form piecewise linear approximations of a function.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Given a set of points for x:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"x̂ = -1:0.5:2","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"and a set of corresponding points for y:","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"ŷ = x̂ .^ 2","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"the piecewise linear approximation is constructed by representing x and y as convex combinations of x̂ and ŷ.","category":"page"},{"location":"tutorials/linear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"N = length(x̂)\nmodel = Model();\n@variable(model, -1 <= x <= 2)\n@variable(model, y)\n@variable(model, 0 <= λ[1:N] <= 1)\n@objective(model, Max, y)\n@constraints(model, begin\n    x == sum(x̂[i] * λ[i] for i in 1:N)\n    y == sum(ŷ[i] * λ[i] for i in 1:N)\n    sum(λ) == 1\n    λ in SOS2()\nend)","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"EditURL = \"web_app.jl\"","category":"page"},{"location":"tutorials/applications/web_app/#Serving-web-apps","page":"Serving web apps","title":"Serving web apps","text":"","category":"section"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"This tutorial demonstrates how to setup and serve JuMP models via a REST API.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"In the example app we are building, we solve a trivial mixed-integer program, which is parameterized by the lower bound of a variable. To call the service, users send an HTTP POST request with JSON contents indicating the lower bound. The returned value is the solution of the mixed-integer program as JSON.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"First, we need JuMP and a solver:","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"We also need HTTP.jl to act as our REST server, and JSON.jl to marshal data.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"import HTTP\nimport JSON","category":"page"},{"location":"tutorials/applications/web_app/#The-server-side","page":"Serving web apps","title":"The server side","text":"","category":"section"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"The core components of our REST server are endpoints. These are functions which accept a Dict{String,Any} of input parameters, and return a Dict{String,Any} as output. The types are Dict{String,Any} because we're going to read these to and from JSON.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"Here's a very simple endpoint: it accepts params as input, formulates and solves a trivial mixed-integer program, and then returns a dictionary with the result.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"function endpoint_solve(params::Dict{String,Any})\n    if !haskey(params, \"lower_bound\")\n        return Dict{String,Any}(\n            \"status\" => \"failure\",\n            \"reason\" => \"missing lower_bound param\",\n        )\n    elseif !(params[\"lower_bound\"] isa Real)\n        return Dict{String,Any}(\n            \"status\" => \"failure\",\n            \"reason\" => \"lower_bound is not a number\",\n        )\n    end\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    @variable(model, x >= params[\"lower_bound\"], Int)\n    optimize!(model)\n    ret = Dict{String,Any}(\n        \"status\" => \"okay\",\n        \"terminaton_status\" => termination_status(model),\n        \"primal_status\" => primal_status(model),\n    )\n    # Only include the `x` key if it has a value.\n    if has_values(model)\n        ret[\"x\"] = value(x)\n    end\n    return ret\nend","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"When we call this, we get:","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"endpoint_solve(Dict{String,Any}(\"lower_bound\" => 1.2))","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"endpoint_solve(Dict{String,Any}())","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"For a second function, we need a function that accepts an HTTP.Request object and returns an HTTP.Response object.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"function serve_solve(request::HTTP.Request)\n    data = JSON.parse(String(request.body))\n    solution = endpoint_solve(data)\n    return HTTP.Response(200, JSON.json(solution))\nend","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"Finally, we need an HTTP server. There are a variety of ways you can do this in HTTP.jl. We use an explicit Sockets.listen so we have manual control of when we shutdown the server.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"function setup_server(host, port)\n    server = HTTP.Sockets.listen(host, port)\n    HTTP.serve!(host, port; server = server) do request\n        try\n            # Extend the server by adding other endpoints here.\n            if request.target == \"/api/solve\"\n                return serve_solve(request)\n            else\n                return HTTP.Response(404, \"target $(request.target) not found\")\n            end\n        catch err\n            # Log details about the exception server-side\n            @info \"Unhandled exception: $err\"\n            # Return a response to the client\n            return HTTP.Response(500, \"internal error\")\n        end\n    end\n    return server\nend","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"warning: Warning\nHTTP.jl does not serve requests on a separate thread. Therefore, a long-running job will block the main thread, preventing concurrent users from submitting requests. To work-around this, read HTTP.jl issue 798 or watch Building Microservices and Applications in Julia from JuliaCon 2020.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"server = setup_server(HTTP.ip\"127.0.0.1\", 8080)","category":"page"},{"location":"tutorials/applications/web_app/#The-client-side","page":"Serving web apps","title":"The client side","text":"","category":"section"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"Now that we have a server, we can send it requests via this function:","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"function send_request(data::Dict; endpoint::String = \"solve\")\n    ret = HTTP.request(\n        \"POST\",\n        # This should match the URL and endpoint we defined for our server.\n        \"http://127.0.0.1:8080/api/$endpoint\",\n        [\"Content-Type\" => \"application/json\"],\n        JSON.json(data),\n    )\n    if ret.status != 200\n        # This could happen if there are time-outs, network errors, etc.\n        return Dict(\n            \"status\" => \"failure\",\n            \"code\" => ret.status,\n            \"body\" => String(ret.body),\n        )\n    end\n    return JSON.parse(String(ret.body))\nend","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"Let's see what happens:","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"send_request(Dict(\"lower_bound\" => 0))","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"send_request(Dict(\"lower_bound\" => 1.2))","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"If we don't send a lower_bound, we get:","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"send_request(Dict(\"invalid_param\" => 1.2))","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"If we don't send a lower_bound that is a number, we get:","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"send_request(Dict(\"lower_bound\" => \"1.2\"))","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"Finally, we can shutdown our HTTP server:","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"close(server)","category":"page"},{"location":"tutorials/applications/web_app/#Next-steps","page":"Serving web apps","title":"Next steps","text":"","category":"section"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"For more complicated examples relating to HTTP servers, consult the HTTP.jl documentation.","category":"page"},{"location":"tutorials/applications/web_app/","page":"Serving web apps","title":"Serving web apps","text":"To see how you can integrate this with a larger JuMP model, read Design patterns for larger models.","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"EditURL = \"https://github.com/jump-dev/SDPT3.jl/blob/b565aac2a58818090d521f2340e71f597688e4fb/README.md\"","category":"page"},{"location":"packages/SDPT3/#SDPT3.jl","page":"jump-dev/SDPT3.jl","title":"SDPT3.jl","text":"","category":"section"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"SDPT3.jl is wrapper for the SDPT3 solver.","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"an exported sdpt3 function that is a thin wrapper on top of the sdpt3 MATLAB function\nan interface to MathOptInterface","category":"page"},{"location":"packages/SDPT3/#Affiliation","page":"jump-dev/SDPT3.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"This wrapper is maintained by the JuMP community and is not an official wrapper of SDPT3.","category":"page"},{"location":"packages/SDPT3/#License","page":"jump-dev/SDPT3.jl","title":"License","text":"","category":"section"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"SDPT3.jl is licensed under the MIT License.","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"The underlying solver, SDPT3 is licensed under the GPL v2 License.","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"In addition, SDPT3 requires an installation of MATLAB, which is a closed-source commercial product for which you must obtain a license.","category":"page"},{"location":"packages/SDPT3/#Use-with-JuMP","page":"jump-dev/SDPT3.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"To use SDPT3 with JuMP, do:","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"using JuMP, SDPT3\nmodel = Model(SDPT3.Optimizer)\nset_attribute(model, \"printlevel\", 0)","category":"page"},{"location":"packages/SDPT3/#Installation","page":"jump-dev/SDPT3.jl","title":"Installation","text":"","category":"section"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"First, make sure that you satisfy the requirements of the MATLAB.jl Julia package, and that the SeDuMi software is installed in your MATLAB™ installation.","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"Then, install SDPT3.jl using Pkg.add:","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"import Pkg\nPkg.add(\"SDPT3\")","category":"page"},{"location":"packages/SDPT3/#SDPT3-not-in-PATH","page":"jump-dev/SDPT3.jl","title":"SDPT3 not in PATH","text":"","category":"section"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"If you get the error:","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"Error using save\nVariable 'jx_sdpt3_arg_out_1' not found.\n\nERROR: LoadError: MATLAB.MEngineError(\"failed to get variable jx_sdpt3_arg_out_1 from MATLAB session\")\nStacktrace:\n[...]","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"The error means that we could not find the sdpt3 function with one output argument using the MATLAB C API. This most likely means that you did not add SDPT3 to the MATLAB's path (that is, the toolbox/local/pathdef.m file).","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"If modifying toolbox/local/pathdef.m does not work, the following should work, where /path/to/sdpt3/ is the directory where the sdpt3 folder is located:","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"julia> using MATLAB\n\njulia> cd(\"/path/to/sdpt3/\") do\n           MATLAB.mat\"install_sdpt3\"\n       end\n\njulia> MATLAB.mat\"savepath\"","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"An alternative fix is suggested in the following issue.","category":"page"},{"location":"packages/SDPT3/#Error-in-validate","page":"jump-dev/SDPT3.jl","title":"Error in validate","text":"","category":"section"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"If you get the error:","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"Brace indexing is not supported for variables of this type.\n\nError in validate\n\nError in sdpt3 (line 171)\n   [blk,At,C,b,blkdim,numblk,parbarrier] = validate(blk,At,C,b,par,parbarrier);\n\nError using save\nVariable 'jx_sdpt3_arg_out_1' not found.","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"It might mean that you have added SDPNAL in addition to SDPT3 in the MATLAB's path (that is, the toolbox/local/pathdef.m file). Because SDPNAL also defines a validate function, this can make sdpt3 call SDPNAL's validate function instead of SDPT3's validate function, which causes the issue.","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"One way to fix this from the Julia REPL is to reset the search path to the factory-installed state using restoredefaultpath:","category":"page"},{"location":"packages/SDPT3/","page":"jump-dev/SDPT3.jl","title":"jump-dev/SDPT3.jl","text":"julia> using MATLAB\n\njulia> MATLAB.restoredefaultpath()\n\njulia> MATLAB.mat\"savepath\"","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"EditURL = \"dualization.jl\"","category":"page"},{"location":"tutorials/conic/dualization/#Dualization","page":"Dualization","title":"Dualization","text":"","category":"section"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The purpose of this tutorial is to explain how to use Dualization.jl to improve the performance of some conic optimization models. There are two important takeaways:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"JuMP reformulates problems to meet the input requirements of the solver, potentially increasing the problem size by adding slack variables and constraints.\nSolving the dual of a conic model can be more efficient than solving the primal.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"Dualization.jl is a package which fixes these problems, allowing you to solve the dual instead of the primal with a one-line change to your code.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"This tutorial uses the following packages","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"using JuMP\nimport Dualization\nimport SCS","category":"page"},{"location":"tutorials/conic/dualization/#Background","page":"Dualization","title":"Background","text":"","category":"section"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"Conic optimization solvers typically accept one of two input formulations.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The first is the standard conic form:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"beginalign\n    min_x in mathbbR^n   c^top x \n              textst   A x = b  \n                               x in mathcalK\nendalign","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"in which we have a set of linear equality constraints Ax = b and the variables belong to a cone mathcalK.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The second is the geometric conic form:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"beginalign\n    min_x in mathbbR^n   c^top x \n              textst   A x - b in mathcalK\nendalign","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"in which an affine function Ax - b belongs to a cone mathcalK and the variables are free.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"It is trivial to convert between these two representations, for example, to go from the geometric conic form to the standard conic form we introduce slack variables y:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"beginalign\n    min_x in mathbbR^n   c^top x \n              textst   beginbmatrixA  -Iendbmatrix beginbmatrixxyendbmatrix = b \n                               beginbmatrixxyendbmatrix in mathbbR^n times mathcalK\nendalign","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"and to go from the standard conic form to the geometric conic form, we can rewrite the equality constraint as a function belonging to the {0} cone:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"beginalign\n    min_x in mathbbR^n   c^top x \n              textst  beginbmatrixAIendbmatrix x - beginbmatrixb0endbmatrix in 0 times mathcalK\nendalign","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"From a theoretical perspective, the two formulations are equivalent, and if you implement a model in the standard conic form and pass it to a geometric conic form solver (or vice versa), then JuMP will automatically reformulate the problem into the correct formulation.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"From a practical perspective though, the reformulations are problematic because the additional slack variables and constraints can make the problem much larger and therefore harder to solve.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"You should also note many problems contain a mix of conic constraints and variables, and so they do not neatly fall into one of the two formulations. In these cases, JuMP reformulates only the variables and constraints as necessary to convert the problem into the desired form.","category":"page"},{"location":"tutorials/conic/dualization/#Primal-and-dual-formulations","page":"Dualization","title":"Primal and dual formulations","text":"","category":"section"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"Duality plays a large role in conic optimization. For a detailed description of conic duality, see Duality.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"A useful observation is that if the primal problem is in standard conic form, then the dual problem is in geometric conic form, and vice versa. Moreover, the primal and dual may have a different number of variables and constraints, although which one is smaller depends on the problem. Therefore, instead of reformulating the problem from one form to the other, it can be more efficient to solve the dual instead of the primal.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"To demonstrate, we use a variation of the Maximum cut via SDP example.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The primal formulation (in standard conic form) is:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"model_primal = Model()\n@variable(model_primal, X[1:2, 1:2], PSD)\n@objective(model_primal, Max, sum([1 -1; -1 1] .* X))\n@constraint(model_primal, primal_c[i = 1:2], 1 - X[i, i] == 0)\nprint(model_primal)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"This problem has three scalar decision variables (the matrix X is symmetric), two scalar equality constraints, and a constraint that X is positive semidefinite.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The dual of model_primal is:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"model_dual = Model()\n@variable(model_dual, y[1:2])\n@objective(model_dual, Min, sum(y))\n@constraint(model_dual, dual_c, [y[1]-1 1; 1 y[2]-1] in PSDCone())\nprint(model_dual)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"This problem has two scalar decision variables, and a 2x2 positive semidefinite matrix constraint.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"tip: Tip\nIf you haven't seen conic duality before, try deriving the dual problem based on the description in Duality. You'll need to know that the dual cone of PSDCone is the PSDCone.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"When we solve model_primal with SCS.Optimizer, SCS reports three variables (variables n: 3), five rows in the constraint matrix (constraints m: 5), and five non-zeros in the matrix (nnz(A): 5):","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"set_optimizer(model_primal, SCS.Optimizer)\noptimize!(model_primal)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"(There are five rows in the constraint matrix because SCS expects problems in geometric conic form, and so JuMP has reformulated the X, PSD variable constraint into the affine constraint X .+ 0 in PSDCone().)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The solution we obtain is:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"value.(X)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"dual.(primal_c)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"objective_value(model_primal)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"When we solve model_dual with SCS.Optimizer, SCS reports two variables (variables n: 2), three rows in the constraint matrix (constraints m: 3), and two non-zeros in the matrix (nnz(A): 2):","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"set_optimizer(model_dual, SCS.Optimizer)\noptimize!(model_dual)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"and the solution we obtain is:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"dual.(dual_c)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"value.(y)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"objective_value(model_dual)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"This particular problem is small enough that it isn't meaningful to compare the solve times, but in general, we should expect model_dual to solve faster than model_primal because it contains fewer variables and constraints. The difference is particularly noticeable on large-scale optimization problems.","category":"page"},{"location":"tutorials/conic/dualization/#dual_optimizer","page":"Dualization","title":"dual_optimizer","text":"","category":"section"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"Manually deriving the conic dual is difficult and error-prone. The package Dualization.jl provides the Dualization.dual_optimizer meta-solver, which wraps any MathOptInterface-compatible solver in an interface that automatically formulates and solves the dual of an input problem.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"To demonstrate, we use Dualization.dual_optimizer to solve model_primal:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"set_optimizer(model_primal, Dualization.dual_optimizer(SCS.Optimizer))\noptimize!(model_primal)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The performance is the same as if we solved model_dual, and the correct solution is returned to X:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"value.(X)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"dual.(primal_c)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"Moreover, if we use dual_optimizer on model_dual, then we get the same performance as if we had solved model_primal:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"set_optimizer(model_dual, Dualization.dual_optimizer(SCS.Optimizer))\noptimize!(model_dual)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"dual.(dual_c)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"value.(y)","category":"page"},{"location":"tutorials/conic/dualization/#A-mixed-example","page":"Dualization","title":"A mixed example","text":"","category":"section"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"The Maximum cut via SDP example is nicely defined because the primal is in standard conic form and the dual is in geometric conic form. However, many practical models contain a mix of the two formulations. One example is The minimum distortion problem:","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"D = [0 1 1 1; 1 0 2 2; 1 2 0 2; 1 2 2 0]\nmodel = Model()\n@variable(model, c²)\n@variable(model, Q[1:4, 1:4], PSD)\n@objective(model, Min, c²)\nfor i in 1:4, j in (i+1):4\n    @constraint(model, D[i, j]^2 <= Q[i, i] + Q[j, j] - 2 * Q[i, j])\n    @constraint(model, Q[i, i] + Q[j, j] - 2 * Q[i, j] <= c² * D[i, j]^2)\nend\n@constraint(model, Q[1, 1] == 0)\n@constraint(model, c² >= 1)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"In this formulation, the Q variable is of the form xinmathcalK, but there is also a free variable, c², a linear equality constraint, Q[1, 1] == 0, and some linear inequality constraints. Rather than attempting to derive the formulation that JuMP would pass to SCS and its dual, the simplest solution is to try solving the problem with and without dual_optimizer to see which formulation is most efficient.","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"set_optimizer(model, SCS.Optimizer)\noptimize!(model)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"set_optimizer(model, Dualization.dual_optimizer(SCS.Optimizer))\noptimize!(model)","category":"page"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"For this problem, SCS reports that the primal has variables n: 11, constraints m: 24 and that the dual has variables n: 14, constraints m: 24. Therefore, we should probably use the primal formulation because it has fewer variables and the same number of constraints.","category":"page"},{"location":"tutorials/conic/dualization/#When-to-use-dual_optimizer","page":"Dualization","title":"When to use dual_optimizer","text":"","category":"section"},{"location":"tutorials/conic/dualization/","page":"Dualization","title":"Dualization","text":"Because it can make the problem larger or smaller, depending on the problem and the choice of solver, there is no definitive rule on when you should use dual_optimizer. However, you should try dual_optimizer if your conic optimization problem takes a long time to solve, or if you need to repeatedly solve similarly structured problems with different data. In some cases solving the dual instead of the primal can make a large difference.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"EditURL = \"ellipse_approx.jl\"","category":"page"},{"location":"tutorials/conic/ellipse_approx/#Ellipsoid-approximation","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"","category":"section"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"This tutorial considers the problem of computing extremal ellipsoids: finding ellipsoids that best approximate a given set. As an extension, we show how to use JuMP to inspect the bridges that were used, and how to explore alternative formulations.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"The model comes from Section 4.9 \"Applications VII: extremal ellipsoids\" of the book Lectures on Modern Convex Optimization by Ben-Tal and Nemirovski (2001).","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"For a related example, see also the Minimal ellipses tutorial.","category":"page"},{"location":"tutorials/conic/ellipse_approx/#Problem-formulation","page":"Ellipsoid approximation","title":"Problem formulation","text":"","category":"section"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Suppose that we are given a set mathcalS consisting of m points in n-dimensional space:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"mathcalS =  x_1 ldots x_m  subset mathbbR^n","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Our goal is to determine an optimal vector c in  mathbbR^n and an optimal n times n real symmetric matrix D such that the ellipse:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"E(D c) =  x  (x - c)^top D ( x - c) leq 1 ","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"contains mathcalS and has the smallest possible volume.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"The optimal D and c are given by the optimization problem:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"beginaligned\nmax        quad  t                                                                    \ntextst quad  Z succeq 0 \n                   beginbmatrix s  z^top  z  Z  endbmatrix succeq 0 \n                   x_i^top Z x_i - 2x_i^top z + s leq 1     quad i=1 ldots m \n                   t le sqrtndet(Z)\nendaligned","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"where D = Z_* and c = Z_*^-1 z_*.","category":"page"},{"location":"tutorials/conic/ellipse_approx/#Required-packages","page":"Ellipsoid approximation","title":"Required packages","text":"","category":"section"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"using JuMP\nimport LinearAlgebra\nimport Plots\nimport Random\nimport SCS","category":"page"},{"location":"tutorials/conic/ellipse_approx/#Data","page":"Ellipsoid approximation","title":"Data","text":"","category":"section"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"We first need to generate some points to work with.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"function generate_point_cloud(\n    m;            # number of 2-dimensional points\n    a = 10,       # scaling in x direction\n    b = 2,        # scaling in y direction\n    rho = π / 6,  # rotation of points around origin\n    random_seed = 1,\n)\n    rng = Random.MersenneTwister(random_seed)\n    P = randn(rng, Float64, m, 2)\n    Phi = [a*cos(rho) a*sin(rho); -b*sin(rho) b*cos(rho)]\n    S = P * Phi\n    return S\nend","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"For the sake of this example, let's take m = 600:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"S = generate_point_cloud(600);\nnothing #hide","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"We will visualise the points (and ellipse) using the Plots package:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"r = 1.1 * maximum(abs.(S))\nplot = Plots.scatter(\n    S[:, 1],\n    S[:, 2];\n    xlim = (-r, r),\n    ylim = (-r, r),\n    label = nothing,\n    c = :green,\n    shape = :x,\n    size = (600, 600),\n)","category":"page"},{"location":"tutorials/conic/ellipse_approx/#JuMP-formulation","page":"Ellipsoid approximation","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Now let's build and the JuMP model. We'll compute D and c after the solve.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"model = Model(SCS.Optimizer)\n# We need to use a tighter tolerance for this example, otherwise the bounding\n# ellipse won't actually be bounding...\nset_attribute(model, \"eps_rel\", 1e-6)\nset_silent(model)\nm, n = size(S)\n@variable(model, z[1:n])\n@variable(model, Z[1:n, 1:n], PSD)\n@variable(model, s)\n@variable(model, t)\n@constraint(model, [s z'; z Z] >= 0, PSDCone())\n@constraint(\n    model,\n    [i in 1:m],\n    S[i, :]' * Z * S[i, :] - 2 * S[i, :]' * z + s <= 1,\n)\n@constraint(model, [t; vec(Z)] in MOI.RootDetConeSquare(n))\n@objective(model, Max, t)\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/#Results","page":"Ellipsoid approximation","title":"Results","text":"","category":"section"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"After solving the model to optimality we can recover the solution in terms of D and c:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"D = value.(Z)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"c = D \\ value.(z)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Finally, overlaying the solution in the plot we see the minimal volume approximating ellipsoid:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"P = sqrt(D)\nq = -P * c\ndata = [tuple(P \\ [cos(θ) - q[1], sin(θ) - q[2]]...) for θ in 0:0.05:(2pi+0.05)]\nPlots.plot!(plot, data; c = :crimson, label = nothing)","category":"page"},{"location":"tutorials/conic/ellipse_approx/#Alternative-formulations","page":"Ellipsoid approximation","title":"Alternative formulations","text":"","category":"section"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"The formulation of model uses MOI.RootDetConeSquare. However, because SCS does not natively support this cone, JuMP automatically reformulates the problem into an equivalent problem that SCS does support. You can see the reformulation that JuMP chose using print_active_bridges:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"print_active_bridges(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"There's a lot going on here, but the first bullet is:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"* Unsupported objective: MOI.VariableIndex\n|  bridged by:\n|   MOIB.Objective.FunctionizeBridge{Float64}\n|  introduces:\n|   * Supported objective: MOI.ScalarAffineFunction{Float64}","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"This says that SCS does not support a MOI.VariableIndex objective function, and that JuMP used a MOI.Bridges.Objective.FunctionizeBridge to convert it into a MOI.ScalarAffineFunction{Float64} objective function.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"We can leave JuMP to do the reformulation, or we can rewrite our model to have an objective function that SCS natively supports:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"@objective(model, Max, 1.0 * t + 0.0);\nnothing #hide","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Re-printing the active bridges:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"print_active_bridges(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"we get * Supported objective: MOI.ScalarAffineFunction{Float64}.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"We can manually implement some other reformulations to change our model to something that SCS more closely supports by:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Replacing the MOI.VectorOfVariables in MOI.PositiveSemidefiniteConeTriangle constraint @variable(model, Z[1:n, 1:n], PSD) with the MOI.VectorAffineFunction in MOI.PositiveSemidefiniteConeTriangle @constraint(model, Z >= 0, PSDCone()).\nReplacing the MOI.VectorOfVariables in MOI.PositiveSemidefiniteConeSquare constraint [s z'; z Z] >= 0, PSDCone() with the MOI.VectorAffineFunction in MOI.PositiveSemidefiniteConeTriangle @constraint(model, LinearAlgebra.Symmetric([s z'; z Z]) >= 0, PSDCone()).\nReplacing the MOI.ScalarAffineFunction in MOI.GreaterThan constraints with the vectorized equivalent of MOI.VectorAffineFunction in MOI.Nonnegatives\nReplacing the MOI.VectorOfVariables in MOI.RootDetConeSquare constraint with MOI.VectorAffineFunction in MOI.RootDetConeTriangle.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Note that we still need to bridge MOI.PositiveSemidefiniteConeTriangle constraints because SCS uses an internal SCS.ScaledPSDCone set instead.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"model = Model(SCS.Optimizer)\nset_attribute(model, \"eps_rel\", 1e-6)\nset_silent(model)\n@variable(model, z[1:n])\n@variable(model, s)\n@variable(model, t)\n# The former @variable(model, Z[1:n, 1:n], PSD)\n@variable(model, Z[1:n, 1:n], Symmetric)\n@constraint(model, Z >= 0, PSDCone())\n# The former [s z'; z Z] >= 0, PSDCone()\n@constraint(model, LinearAlgebra.Symmetric([s z'; z Z]) >= 0, PSDCone())\n# The former constraint S[i, :]' * Z * S[i, :] - 2 * S[i, :]' * z + s <= 1\nf = [1 - S[i, :]' * Z * S[i, :] + 2 * S[i, :]' * z - s for i in 1:m]\n@constraint(model, f in MOI.Nonnegatives(m))\n# The former constraint [t; vec(Z)] in MOI.RootDetConeSquare(n)\n@constraint(model, 1 * [t; triangle_vec(Z)] .+ 0 in MOI.RootDetConeTriangle(n))\n# The former @objective(model, Max, t)\n@objective(model, Max, 1 * t + 0)\noptimize!(model)\nsolve_time_1 = solve_time(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"This formulation gives the much smaller graph:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"print_active_bridges(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"The last bullet shows how JuMP reformulated the MOI.RootDetConeTriangle constraint by adding a mix of MOI.PositiveSemidefiniteConeTriangle and MOI.GeometricMeanCone constraints.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Because SCS doesn't natively support the MOI.GeometricMeanCone, these constraints were further bridged using a MOI.Bridges.Constraint.GeoMeanToPowerBridge to a series of MOI.PowerCone constraints.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"However, there are many other ways that a MOI.GeometricMeanCone can be reformulated into something that SCS supports. Let's see what happens if we use remove_bridge to remove the MOI.Bridges.Constraint.GeoMeanToPowerBridge:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"remove_bridge(model, MOI.Bridges.Constraint.GeoMeanToPowerBridge)\noptimize!(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"This time, the solve took:","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"solve_time_2 = solve_time(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"where previously it took","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"solve_time_1","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Why was the solve time different?","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"print_active_bridges(model)","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"This time, JuMP used a MOI.Bridges.Constraint.GeoMeanBridge to reformulate the constraint into a set of MOI.RotatedSecondOrderCone constraints, which were further reformulated into a set of supported MOI.SecondOrderCone constraints.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"Since the two models are equivalent, we can conclude that for this particular model, the MOI.SecondOrderCone formulation is more efficient.","category":"page"},{"location":"tutorials/conic/ellipse_approx/","page":"Ellipsoid approximation","title":"Ellipsoid approximation","text":"In general though, the performance of a particular reformulation is problem- and solver-specific. Therefore, JuMP chooses to minimize the number of bridges in the default reformulation, leaving you to explore alternative formulations using the tools and techniques shown in this tutorial.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"DocTestSetup = quote\n    using JuMP\nend","category":"page"},{"location":"manual/containers/#Containers","page":"Containers","title":"Containers","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"JuMP provides specialized containers similar to AxisArrays that enable multi-dimensional arrays with non-integer indices.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"These containers are created automatically by JuMP's macros. Each macro has the same basic syntax:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"@macroname(model, name[key1=index1, index2; optional_condition], other stuff)","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"The containers are generated by the name[key1=index1, index2; optional_condition] syntax. Everything else is specific to the particular macro.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Containers can be named, for example, name[key=index], or unnamed, for example, [key=index]. We call unnamed containers anonymous.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"We call the bits inside the square brackets and before the ; the index sets. The index sets can be named, for example, [i = 1:4], or they can be unnamed, for example, [1:4].","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"We call the bit inside the square brackets and after the ; the condition. Conditions are optional.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"In addition to the standard JuMP macros like @variable and @constraint, which construct containers of variables and constraints respectively, you can use Containers.@container to construct containers with arbitrary elements.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"We will use this macro to explain the three types of containers that are natively supported by JuMP: Array, Containers.DenseAxisArray, and Containers.SparseAxisArray.","category":"page"},{"location":"manual/containers/#Array","page":"Containers","title":"Array","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"An Array is created when the index sets are rectangular and the index sets are of the form 1:n.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> Containers.@container(x[i = 1:2, j = 1:3], (i, j))\n2×3 Matrix{Tuple{Int64, Int64}}:\n (1, 1)  (1, 2)  (1, 3)\n (2, 1)  (2, 2)  (2, 3)","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"The result is a normal Julia Array, so you can do all the usual things.","category":"page"},{"location":"manual/containers/#Slicing","page":"Containers","title":"Slicing","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Arrays can be sliced","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> x[:, 1]\n2-element Vector{Tuple{Int64, Int64}}:\n (1, 1)\n (2, 1)\n\njulia> x[2, :]\n3-element Vector{Tuple{Int64, Int64}}:\n (2, 1)\n (2, 2)\n (2, 3)","category":"page"},{"location":"manual/containers/#Looping","page":"Containers","title":"Looping","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use eachindex to loop over the elements:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> for key in eachindex(x)\n           println(x[key])\n       end\n(1, 1)\n(2, 1)\n(1, 2)\n(2, 2)\n(1, 3)\n(2, 3)","category":"page"},{"location":"manual/containers/#Get-the-index-sets","page":"Containers","title":"Get the index sets","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use axes to obtain the index sets:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> axes(x)\n(Base.OneTo(2), Base.OneTo(3))","category":"page"},{"location":"manual/containers/#Broadcasting","page":"Containers","title":"Broadcasting","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Broadcasting over an Array returns an Array","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> swap(x::Tuple) = (last(x), first(x))\nswap (generic function with 1 method)\n\njulia> swap.(x)\n2×3 Matrix{Tuple{Int64, Int64}}:\n (1, 1)  (2, 1)  (3, 1)\n (1, 2)  (2, 2)  (3, 2)","category":"page"},{"location":"manual/containers/#Tables","page":"Containers","title":"Tables","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use Containers.rowtable to convert the Array into a Tables.jl compatible Vector{<:NamedTuple}:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> table = Containers.rowtable(x; header = [:I, :J, :value])\n6-element Vector{NamedTuple{(:I, :J, :value), Tuple{Int64, Int64, Tuple{Int64, Int64}}}}:\n (I = 1, J = 1, value = (1, 1))\n (I = 2, J = 1, value = (2, 1))\n (I = 1, J = 2, value = (1, 2))\n (I = 2, J = 2, value = (2, 2))\n (I = 1, J = 3, value = (1, 3))\n (I = 2, J = 3, value = (2, 3))","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Because it supports the Tables.jl interface, you can pass it to any function which accepts a table as input:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> import DataFrames;\n\njulia> DataFrames.DataFrame(table)\n6×3 DataFrame\n Row │ I      J      value\n     │ Int64  Int64  Tuple…\n─────┼──────────────────────\n   1 │     1      1  (1, 1)\n   2 │     2      1  (2, 1)\n   3 │     1      2  (1, 2)\n   4 │     2      2  (2, 2)\n   5 │     1      3  (1, 3)\n   6 │     2      3  (2, 3)","category":"page"},{"location":"manual/containers/#DenseAxisArray","page":"Containers","title":"DenseAxisArray","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"A Containers.DenseAxisArray is created when the index sets are rectangular, but not of the form 1:n. The index sets can be of any type.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> x = Containers.@container([i = 1:2, j = [:A, :B]], (i, j))\n2-dimensional DenseAxisArray{Tuple{Int64, Symbol},2,...} with index sets:\n    Dimension 1, Base.OneTo(2)\n    Dimension 2, [:A, :B]\nAnd data, a 2×2 Matrix{Tuple{Int64, Symbol}}:\n (1, :A)  (1, :B)\n (2, :A)  (2, :B)","category":"page"},{"location":"manual/containers/#Slicing-2","page":"Containers","title":"Slicing","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"DenseAxisArrays can be sliced","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> x[:, :A]\n1-dimensional DenseAxisArray{Tuple{Int64, Symbol},1,...} with index sets:\n    Dimension 1, Base.OneTo(2)\nAnd data, a 2-element Vector{Tuple{Int64, Symbol}}:\n (1, :A)\n (2, :A)\n\njulia> x[1, :]\n1-dimensional DenseAxisArray{Tuple{Int64, Symbol},1,...} with index sets:\n    Dimension 1, [:A, :B]\nAnd data, a 2-element Vector{Tuple{Int64, Symbol}}:\n (1, :A)\n (1, :B)","category":"page"},{"location":"manual/containers/#Looping-2","page":"Containers","title":"Looping","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use eachindex to loop over the elements:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> for key in eachindex(x)\n           println(x[key])\n       end\n(1, :A)\n(2, :A)\n(1, :B)\n(2, :B)","category":"page"},{"location":"manual/containers/#Get-the-index-sets-2","page":"Containers","title":"Get the index sets","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use axes to obtain the index sets:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> axes(x)\n(Base.OneTo(2), [:A, :B])","category":"page"},{"location":"manual/containers/#Broadcasting-2","page":"Containers","title":"Broadcasting","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Broadcasting over a DenseAxisArray returns a DenseAxisArray","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> swap(x::Tuple) = (last(x), first(x))\nswap (generic function with 1 method)\n\njulia> swap.(x)\n2-dimensional DenseAxisArray{Tuple{Symbol, Int64},2,...} with index sets:\n    Dimension 1, Base.OneTo(2)\n    Dimension 2, [:A, :B]\nAnd data, a 2×2 Matrix{Tuple{Symbol, Int64}}:\n (:A, 1)  (:B, 1)\n (:A, 2)  (:B, 2)","category":"page"},{"location":"manual/containers/#Access-internal-data","page":"Containers","title":"Access internal data","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use Array(x) to copy the internal data array into a new Array:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> Array(x)\n2×2 Matrix{Tuple{Int64, Symbol}}:\n (1, :A)  (1, :B)\n (2, :A)  (2, :B)","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"To access the internal data without a copy, use x.data.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> x.data\n2×2 Matrix{Tuple{Int64, Symbol}}:\n (1, :A)  (1, :B)\n (2, :A)  (2, :B)","category":"page"},{"location":"manual/containers/#Tables-2","page":"Containers","title":"Tables","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use Containers.rowtable to convert the DenseAxisArray into a Tables.jl compatible Vector{<:NamedTuple}:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> table = Containers.rowtable(x; header = [:I, :J, :value])\n4-element Vector{NamedTuple{(:I, :J, :value), Tuple{Int64, Symbol, Tuple{Int64, Symbol}}}}:\n (I = 1, J = :A, value = (1, :A))\n (I = 2, J = :A, value = (2, :A))\n (I = 1, J = :B, value = (1, :B))\n (I = 2, J = :B, value = (2, :B))","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Because it supports the Tables.jl interface, you can pass it to any function which accepts a table as input:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> import DataFrames;\n\njulia> DataFrames.DataFrame(table)\n4×3 DataFrame\n Row │ I      J       value\n     │ Int64  Symbol  Tuple…\n─────┼────────────────────────\n   1 │     1  A       (1, :A)\n   2 │     2  A       (2, :A)\n   3 │     1  B       (1, :B)\n   4 │     2  B       (2, :B)","category":"page"},{"location":"manual/containers/#Keyword-indexing","page":"Containers","title":"Keyword indexing","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"If all axes are named, you can use keyword indexing:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> x[i = 2, j = :A]\n(2, :A)\n\njulia> x[i = :, j = :B]\n1-dimensional DenseAxisArray{Tuple{Int64, Symbol},1,...} with index sets:\n    Dimension 1, Base.OneTo(2)\nAnd data, a 2-element Vector{Tuple{Int64, Symbol}}:\n (1, :B)\n (2, :B)","category":"page"},{"location":"manual/containers/#SparseAxisArray","page":"Containers","title":"SparseAxisArray","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"A Containers.SparseAxisArray is created when the index sets are non-rectangular. This occurs in two circumstances:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"An index depends on a prior index:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> Containers.@container([i = 1:2, j = i:2], (i, j))\nJuMP.Containers.SparseAxisArray{Tuple{Int64, Int64}, 2, Tuple{Int64, Int64}} with 3 entries:\n  [1, 1]  =  (1, 1)\n  [1, 2]  =  (1, 2)\n  [2, 2]  =  (2, 2)","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"The [indices; condition] syntax is used:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> x = Containers.@container([i = 1:3, j = [:A, :B]; i > 1], (i, j))\nJuMP.Containers.SparseAxisArray{Tuple{Int64, Symbol}, 2, Tuple{Int64, Symbol}} with 4 entries:\n  [2, A]  =  (2, :A)\n  [2, B]  =  (2, :B)\n  [3, A]  =  (3, :A)\n  [3, B]  =  (3, :B)","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Here we have the index sets i = 1:3, j = [:A, :B], followed by ;, and then a condition, which evaluates to true or false: i > 1.","category":"page"},{"location":"manual/containers/#Slicing-3","page":"Containers","title":"Slicing","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Slicing is supported:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> y = x[:, :B]\nJuMP.Containers.SparseAxisArray{Tuple{Int64, Symbol}, 1, Tuple{Int64}} with 2 entries:\n  [2]  =  (2, :B)\n  [3]  =  (3, :B)","category":"page"},{"location":"manual/containers/#Looping-3","page":"Containers","title":"Looping","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use eachindex to loop over the elements:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> for key in eachindex(y)\n           println(y[key])\n       end\n(2, :B)\n(3, :B)","category":"page"},{"location":"manual/containers/#Broadcasting-3","page":"Containers","title":"Broadcasting","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Broadcasting over a SparseAxisArray returns a SparseAxisArray","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> swap(x::Tuple) = (last(x), first(x))\nswap (generic function with 1 method)\n\njulia> swap.(y)\nJuMP.Containers.SparseAxisArray{Tuple{Symbol, Int64}, 1, Tuple{Int64}} with 2 entries:\n  [2]  =  (:B, 2)\n  [3]  =  (:B, 3)","category":"page"},{"location":"manual/containers/#Tables-3","page":"Containers","title":"Tables","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Use Containers.rowtable to convert the SparseAxisArray into a Tables.jl compatible Vector{<:NamedTuple}:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> table = Containers.rowtable(x; header = [:I, :J, :value])\n4-element Vector{NamedTuple{(:I, :J, :value), Tuple{Int64, Symbol, Tuple{Int64, Symbol}}}}:\n (I = 2, J = :A, value = (2, :A))\n (I = 2, J = :B, value = (2, :B))\n (I = 3, J = :A, value = (3, :A))\n (I = 3, J = :B, value = (3, :B))","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Because it supports the Tables.jl interface, you can pass it to any function which accepts a table as input:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> import DataFrames;\n\njulia> DataFrames.DataFrame(table)\n4×3 DataFrame\n Row │ I      J       value\n     │ Int64  Symbol  Tuple…\n─────┼────────────────────────\n   1 │     2  A       (2, :A)\n   2 │     2  B       (2, :B)\n   3 │     3  A       (3, :A)\n   4 │     3  B       (3, :B)","category":"page"},{"location":"manual/containers/#Keyword-indexing-2","page":"Containers","title":"Keyword indexing","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"If all axes are named, you can use keyword indexing:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> x[i = 2, j = :A]\n(2, :A)\n\njulia> x[i = :, j = :B]\nJuMP.Containers.SparseAxisArray{Tuple{Int64, Symbol}, 1, Tuple{Int64}} with 2 entries:\n  [2]  =  (2, :B)\n  [3]  =  (3, :B)","category":"page"},{"location":"manual/containers/#Forcing-the-container-type","page":"Containers","title":"Forcing the container type","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Pass container = T to use T as the container. For example:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> Containers.@container([i = 1:2, j = 1:2], i + j, container = Array)\n2×2 Matrix{Int64}:\n 2  3\n 3  4\n\njulia> Containers.@container([i = 1:2, j = 1:2], i + j, container = Dict)\nDict{Tuple{Int64, Int64}, Int64} with 4 entries:\n  (1, 2) => 3\n  (1, 1) => 2\n  (2, 2) => 4\n  (2, 1) => 3","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"You can also pass DenseAxisArray or SparseAxisArray.","category":"page"},{"location":"manual/containers/#How-different-container-types-are-chosen","page":"Containers","title":"How different container types are chosen","text":"","category":"section"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"If the compiler can prove at compile time that the index sets are rectangular, and indexed by a compact set of integers that start at 1, Containers.@container will return an array. This is the case if your index sets are visible to the macro as 1:n:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> Containers.@container([i=1:3, j=1:5], i + j)\n3×5 Matrix{Int64}:\n 2  3  4  5  6\n 3  4  5  6  7\n 4  5  6  7  8","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"or an instance of Base.OneTo:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> set = Base.OneTo(3)\nBase.OneTo(3)\n\njulia> Containers.@container([i=set, j=1:5], i + j)\n3×5 Matrix{Int64}:\n 2  3  4  5  6\n 3  4  5  6  7\n 4  5  6  7  8","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"If the compiler can prove that the index set is rectangular, but not necessarily of the form 1:n at compile time, then a Containers.DenseAxisArray will be constructed instead:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> set = 1:3\n1:3\n\njulia> Containers.@container([i=set, j=1:5], i + j)\n2-dimensional DenseAxisArray{Int64,2,...} with index sets:\n    Dimension 1, 1:3\n    Dimension 2, Base.OneTo(5)\nAnd data, a 3×5 Matrix{Int64}:\n 2  3  4  5  6\n 3  4  5  6  7\n 4  5  6  7  8","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"info: Info\nWhat happened here? Although we know that set contains 1:3, at compile time the typeof(set) is a UnitRange{Int}. Therefore, Julia can't prove that the range starts at 1 (it only finds this out at runtime), and it defaults to a  DenseAxisArray. The case where we explicitly wrote i = 1:3 worked because the macro can \"see\" the 1 at compile time.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"However, if you know that the indices do form an Array, you can force the container type with container = Array:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> set = 1:3\n1:3\n\njulia> Containers.@container([i=set, j=1:5], i + j, container = Array)\n3×5 Matrix{Int64}:\n 2  3  4  5  6\n 3  4  5  6  7\n 4  5  6  7  8","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Here's another example with something similar:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> a = 1\n1\n\njulia> Containers.@container([i=a:3, j=1:5], i + j)\n2-dimensional DenseAxisArray{Int64,2,...} with index sets:\n    Dimension 1, 1:3\n    Dimension 2, Base.OneTo(5)\nAnd data, a 3×5 Matrix{Int64}:\n 2  3  4  5  6\n 3  4  5  6  7\n 4  5  6  7  8\n\njulia> Containers.@container([i=1:a, j=1:5], i + j)\n1×5 Matrix{Int64}:\n 2  3  4  5  6","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"Finally, if the compiler cannot prove that the index set is rectangular, a Containers.SparseAxisArray will be created.","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"This occurs when some indices depend on a previous one:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> Containers.@container([i=1:3, j=1:i], i + j)\nJuMP.Containers.SparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 6 entries:\n  [1, 1]  =  2\n  [2, 1]  =  3\n  [2, 2]  =  4\n  [3, 1]  =  4\n  [3, 2]  =  5\n  [3, 3]  =  6","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"or if there is a condition on the index sets:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> Containers.@container([i = 1:5; isodd(i)], i^2)\nJuMP.Containers.SparseAxisArray{Int64, 1, Tuple{Int64}} with 3 entries:\n  [1]  =  1\n  [3]  =  9\n  [5]  =  25","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"The condition can depend on multiple indices, the only requirement is that it is an expression that returns true or false:","category":"page"},{"location":"manual/containers/","page":"Containers","title":"Containers","text":"julia> condition(i, j) = isodd(i) && iseven(j)\ncondition (generic function with 1 method)\n\njulia> Containers.@container([i = 1:2, j = 1:4; condition(i, j)], i + j)\nJuMP.Containers.SparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 2 entries:\n  [1, 2]  =  3\n  [1, 4]  =  5","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/solutions.md\"","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/manual/solutions/#manual_solutions","page":"Solutions","title":"Solutions","text":"","category":"section"},{"location":"moi/manual/solutions/#Solving-and-retrieving-the-results","page":"Solutions","title":"Solving and retrieving the results","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Once an optimizer is loaded with the objective function and all of the constraints, we can ask the solver to solve the model by calling optimize!.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"MOI.optimize!(optimizer)","category":"page"},{"location":"moi/manual/solutions/#Why-did-the-solver-stop?","page":"Solutions","title":"Why did the solver stop?","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"The optimization procedure may stop for a number of reasons. The TerminationStatus attribute of the optimizer returns a TerminationStatusCode object which explains why the solver stopped.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"The termination statuses distinguish between proofs of optimality, infeasibility, local convergence, limits, and termination because of something unexpected like invalid problem data or failure to converge.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"A typical usage of the TerminationStatus attribute is as follows:","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"status = MOI.get(optimizer, TerminationStatus())\nif status == MOI.OPTIMAL\n    # Ok, we solved the problem!\nelse\n    # Handle other cases.\nend","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"After checking the TerminationStatus, check ResultCount. This attribute returns the number of results that the solver has available to return. A result is defined as a primal-dual pair, but either the primal or the dual may be missing from the result. While the OPTIMAL termination status normally implies that at least one result is available, other statuses do not. For example, in the case of infeasibility, a solver may return no result or a proof of infeasibility. The ResultCount attribute distinguishes between these two cases.","category":"page"},{"location":"moi/manual/solutions/#Primal-solutions","page":"Solutions","title":"Primal solutions","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Use the PrimalStatus optimizer attribute to return a ResultStatusCode describing the status of the primal solution.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Common returns are described below in the Common status situations section.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Query the primal solution using the VariablePrimal and ConstraintPrimal attributes.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Query the objective function value using the ObjectiveValue attribute.","category":"page"},{"location":"moi/manual/solutions/#Dual-solutions","page":"Solutions","title":"Dual solutions","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"warning: Warning\nSee Duality for a discussion of the MOI conventions for primal-dual pairs and certificates.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Use the DualStatus optimizer attribute to return a ResultStatusCode describing the status of the dual solution.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Query the dual solution using the ConstraintDual attribute.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Query the dual objective function value using the DualObjectiveValue attribute.","category":"page"},{"location":"moi/manual/solutions/#Common-status-situations","page":"Solutions","title":"Common status situations","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"The sections below describe how to interpret typical or interesting status cases for three common classes of solvers. The example cases are illustrative, not comprehensive. Solver wrappers may provide additional information on how the solver's statuses map to MOI statuses.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"info: Info\n* in the tables indicate that multiple different values are possible.","category":"page"},{"location":"moi/manual/solutions/#Primal-dual-convex-solver","page":"Solutions","title":"Primal-dual convex solver","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Linear programming and conic optimization solvers fall into this category.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"What happened? TerminationStatus ResultCount PrimalStatus DualStatus\nProved optimality OPTIMAL 1 FEASIBLE_POINT FEASIBLE_POINT\nProved infeasible INFEASIBLE 1 NO_SOLUTION INFEASIBILITY_CERTIFICATE\nOptimal within relaxed tolerances ALMOST_OPTIMAL 1 FEASIBLE_POINT FEASIBLE_POINT\nOptimal within relaxed tolerances ALMOST_OPTIMAL 1 ALMOST_FEASIBLE_POINT ALMOST_FEASIBLE_POINT\nDetected an unbounded ray of the primal DUAL_INFEASIBLE 1 INFEASIBILITY_CERTIFICATE NO_SOLUTION\nStall SLOW_PROGRESS 1 * *","category":"page"},{"location":"moi/manual/solutions/#Global-branch-and-bound-solvers","page":"Solutions","title":"Global branch-and-bound solvers","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Mixed-integer programming solvers fall into this category.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"What happened? TerminationStatus ResultCount PrimalStatus DualStatus\nProved optimality OPTIMAL 1 FEASIBLE_POINT NO_SOLUTION\nPresolve detected infeasibility or unboundedness INFEASIBLE_OR_UNBOUNDED 0 NO_SOLUTION NO_SOLUTION\nProved infeasibility INFEASIBLE 0 NO_SOLUTION NO_SOLUTION\nTimed out (no solution) TIME_LIMIT 0 NO_SOLUTION NO_SOLUTION\nTimed out (with a solution) TIME_LIMIT 1 FEASIBLE_POINT NO_SOLUTION\nCPXMIP_OPTIMAL_INFEAS ALMOST_OPTIMAL 1 INFEASIBLE_POINT NO_SOLUTION","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"info: Info\nCPXMIP_OPTIMAL_INFEAS is a CPLEX status that indicates that a preprocessed problem was solved to optimality, but the solver was unable to recover a feasible solution to the original problem. Handling this status was one of the motivating drivers behind the design of MOI.","category":"page"},{"location":"moi/manual/solutions/#Local-search-solvers","page":"Solutions","title":"Local search solvers","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Nonlinear programming solvers fall into this category. It also includes non-global tree search solvers like Juniper.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"What happened? TerminationStatus ResultCount PrimalStatus DualStatus\nConverged to a stationary point LOCALLY_SOLVED 1 FEASIBLE_POINT FEASIBLE_POINT\nCompleted a non-global tree search (with a solution) LOCALLY_SOLVED 1 FEASIBLE_POINT FEASIBLE_POINT\nConverged to an infeasible point LOCALLY_INFEASIBLE 1 INFEASIBLE_POINT *\nCompleted a non-global tree search (no solution found) LOCALLY_INFEASIBLE 0 NO_SOLUTION NO_SOLUTION\nIteration limit ITERATION_LIMIT 1 * *\nDiverging iterates NORM_LIMIT or OBJECTIVE_LIMIT 1 * *","category":"page"},{"location":"moi/manual/solutions/#Querying-solution-attributes","page":"Solutions","title":"Querying solution attributes","text":"","category":"section"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"Some solvers will not implement every solution attribute. Therefore, a call like MOI.get(model, MOI.SolveTimeSec()) may throw an UnsupportedAttribute error.","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"If you need to write code that is agnostic to the solver (for example, you are writing a library that an end-user passes their choice of solver to), you can work-around this problem using a try-catch:","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"function get_solve_time(model)\n    try\n        return MOI.get(model, MOI.SolveTimeSec())\n    catch err\n        if err isa MOI.UnsupportedAttribute\n            return NaN  # Solver doesn't support. Return a placeholder value.\n        end\n        rethrow(err)  # Something else went wrong. Rethrow the error\n    end\nend","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"If, after careful profiling, you find that the try-catch is taking a significant portion of your runtime, you can improve performance by caching the result of the try-catch:","category":"page"},{"location":"moi/manual/solutions/","page":"Solutions","title":"Solutions","text":"mutable struct CachedSolveTime{M}\n    model::M\n    supports_solve_time::Bool\n    CachedSolveTime(model::M) where {M} = new(model, true)\nend\n\nfunction get_solve_time(model::CachedSolveTime)\n    if !model.supports_solve_time\n        return NaN\n    end\n    try\n        return MOI.get(model, MOI.SolveTimeSec())\n    catch err\n        if err isa MOI.UnsupportedAttribute\n            model.supports_solve_time = false\n            return NaN\n        end\n        rethrow(err)  # Something else went wrong. Rethrow the error\n    end\nend","category":"page"},{"location":"tutorials/getting_started/introduction/#Introduction","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"tutorials/getting_started/introduction/","page":"Introduction","title":"Introduction","text":"The purpose of these \"Getting started\" tutorials is to teach new users the basics of Julia and JuMP.","category":"page"},{"location":"tutorials/getting_started/introduction/#How-these-tutorials-are-structured","page":"Introduction","title":"How these tutorials are structured","text":"","category":"section"},{"location":"tutorials/getting_started/introduction/","page":"Introduction","title":"Introduction","text":"Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.","category":"page"},{"location":"tutorials/getting_started/introduction/","page":"Introduction","title":"Introduction","text":"The \"Getting started with\" tutorials are basic introductions to different aspects of JuMP and Julia. If you are new to JuMP and Julia, start by reading them in the following order:\nGetting started with Julia\nGetting started with JuMP\nGetting started with sets and indexing\nGetting started with data and plotting\nJulia has a reputation for being \"fast.\" Unfortunately, it is also easy to write slow Julia code. Performance tips contains a number of important tips on how to improve the performance of models you write in JuMP.\nDesign patterns for larger models is a more advanced tutorial that is aimed at users writing large JuMP models. It's in the \"Getting started\" section to give you an early preview of how JuMP makes it easy to structure larger models. If you are new to JuMP you may want to skip or briefly skim this tutorial, and come back to it once you have written a few JuMP models.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"EditURL = \"two_stage_stochastic.jl\"","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/#Two-stage-stochastic-programs","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"","category":"section"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"The purpose of this tutorial is to demonstrate how to model and solve a two-stage stochastic program.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"This tutorial uses the following packages","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"using JuMP\nimport Distributions\nimport HiGHS\nimport Plots\nimport StatsPlots\nimport Statistics","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/#Background","page":"Two-stage stochastic programs","title":"Background","text":"","category":"section"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"During the week, you are a busy practitioner of Operations Research. To escape the drudgery of mathematics, you decide to open a side business selling creamy mushroom pies with puff pastry. After a few weeks, it quickly becomes apparent that operating a food business is not so easy.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"The pies must be prepared in the morning, before you open for the day and can gauge the level of demand. If you bake too many, the unsold pies at the end of the day must be discarded and you have wasted time and money on their production. But if you bake too few, then there may be un-served customers and you could have made more money by baking more pies.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"After a few weeks of poor decision making, you decide to put your knowledge of Operations Research to good use, starting with some data collection.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"Each pie costs you $2 to make, and you sell them at $5 each. Disposal of an unsold pie costs $0.10. Based on three weeks of data collected, in which you made 200 pies each week, you sold 150, 190, and 200 pies. Thus, as a guess, you assume a triangular distribution of demand with a minimum of 150, a median of 200, and a maximum of 250.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"We can model this problem by a two-stage stochastic program. In the first stage, we decide a quantity of pies to make x. We make this decision before we observe the demand d_omega. In the second stage, we sell y_omega pies, and incur any costs for unsold pies.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"We can formulate this problem as follows:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"beginaligned\nmaxlimits_xy_omega   -2x + mathbbE_omega5y_omega - 01(x - y_omega) \n   y_omega le x               quad forall omega in Omega \n   0 le y_omega le d_omega  quad forall omega in Omega \n   x ge 0\nendaligned","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/#Sample-Average-approximation","page":"Two-stage stochastic programs","title":"Sample Average approximation","text":"","category":"section"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"If the distribution of demand is continuous, then our problem has an infinite number of variables and constraints. To form a computationally tractable problem, we instead use a finite set of samples drawn from the distribution. This is called sample average approximation (SAA).","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"D = Distributions.TriangularDist(150.0, 250.0, 200.0)\nN = 100\nd = sort!(rand(D, N));\nΩ = 1:N\nP = fill(1 / N, N);\nStatsPlots.histogram(d; bins = 20, label = \"\", xlabel = \"Demand\")","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/#JuMP-model","page":"Two-stage stochastic programs","title":"JuMP model","text":"","category":"section"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"The implementation of our two-stage stochastic program in JuMP is:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x >= 0)\n@variable(model, 0 <= y[ω in Ω] <= d[ω])\n@constraint(model, [ω in Ω], y[ω] <= x)\n@expression(model, z[ω in Ω], 5y[ω] - 0.1 * (x - y[ω]))\n@objective(model, Max, -2x + sum(P[ω] * z[ω] for ω in Ω))\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"The optimal number of pies to make is:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"value(x)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"The distribution of total profit is:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"total_profit = [-2 * value(x) + value(z[ω]) for ω in Ω]","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"Let's plot it:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"\"\"\"\n    bin_distribution(x::Vector{Float64}, N::Int)\n\nA helper function that discretizes `x` into bins of width `N`.\n\"\"\"\nbin_distribution(x, N) = N * (floor(minimum(x) / N):ceil(maximum(x) / N))\n\nplot = StatsPlots.histogram(\n    total_profit;\n    bins = bin_distribution(total_profit, 25),\n    label = \"\",\n    xlabel = \"Profit [\\$]\",\n    ylabel = \"Number of outcomes\",\n)\nμ = Statistics.mean(total_profit)\nPlots.vline!(\n    plot,\n    [μ];\n    label = \"Expected profit (\\$$(round(Int, μ)))\",\n    linewidth = 3,\n)\nplot","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/#Risk-measures","page":"Two-stage stochastic programs","title":"Risk measures","text":"","category":"section"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"A risk measure is a function which maps a random variable to a real number. Common risk measures include the mean (expectation), median, mode, and maximum. We need a risk measure to convert the distribution of second stage costs into a single number that can be optimized.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"Our model currently uses the expectation risk measure, but others are possible too. One popular risk measure is the conditional value at risk (CVaR).","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"CVaR has a parameter gamma, and it computes the expectation of the worst gamma fraction of outcomes.","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"If we are maximizing, so that small outcomes are bad, the definition of CVaR is:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"CVaR_gammaZ = maxlimits_xi  xi - frac1gammamathbbE_omegaleft(xi - Z)_+right","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"which can be formulated as the linear program:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"beginaligned\nCVaR_gammaZ = maxlimits_xi z_omega   xi - frac1gammasum P_omega z_omega\n  z_omega ge xi - Z_omega  quad forall omega \n  z_omega ge 0  quad forall omega\nendaligned","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"function CVaR(Z::Vector{Float64}, P::Vector{Float64}; γ::Float64)\n    @assert 0 < γ <= 1\n    N = length(Z)\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    @variable(model, ξ)\n    @variable(model, z[1:N] >= 0)\n    @constraint(model, [i in 1:N], z[i] >= ξ - Z[i])\n    @objective(model, Max, ξ - 1 / γ * sum(P[i] * z[i] for i in 1:N))\n    optimize!(model)\n    return objective_value(model)\nend","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"When γ is 1.0, we compute the mean of the profit:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"cvar_10 = CVaR(total_profit, P; γ = 1.0)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"Statistics.mean(total_profit)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"As γ approaches 0.0, we compute the worst-case (minimum) profit:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"cvar_00 = CVaR(total_profit, P; γ = 0.0001)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"minimum(total_profit)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"By varying γ between 0 and 1 we can compute some trade-off of these two extremes:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"cvar_05 = CVaR(total_profit, P; γ = 0.5)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"Let's plot these outcomes on our distribution:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"plot = StatsPlots.histogram(\n    total_profit;\n    bins = bin_distribution(total_profit, 25),\n    label = \"\",\n    xlabel = \"Profit [\\$]\",\n    ylabel = \"Number of outcomes\",\n)\nPlots.vline!(\n    plot,\n    [cvar_10 cvar_05 cvar_00];\n    label = [\"γ = 1.0\" \"γ = 0.5\" \"γ = 0.0\"],\n    linewidth = 3,\n)\nplot","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/#Risk-averse-sample-average-approximation","page":"Two-stage stochastic programs","title":"Risk averse sample average approximation","text":"","category":"section"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"Because CVaR can be formulated as a linear program, we can form a risk averse sample average approximation model by combining the two formulations:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"γ = 0.4\nmodel = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x >= 0)\n@variable(model, 0 <= y[ω in Ω] <= d[ω])\n@constraint(model, [ω in Ω], y[ω] <= x)\n@expression(model, Z[ω in Ω], 5 * y[ω] - 0.1(x - y[ω]))\n@variable(model, ξ)\n@variable(model, z[ω in Ω] >= 0)\n@constraint(model, [ω in Ω], z[ω] >= ξ - Z[ω])\n@objective(model, Max, -2x + ξ - 1 / γ * sum(P[ω] * z[ω] for ω in Ω))\noptimize!(model)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"When gamma = 04, the optimal number of pies to bake is:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"value(x)","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"The distribution of total profit is:","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"risk_averse_total_profit = [value(-2x + Z[ω]) for ω in Ω]\nbins = bin_distribution([total_profit; risk_averse_total_profit], 25)\nplot = StatsPlots.histogram(total_profit; label = \"Expectation\", bins = bins)\nStatsPlots.histogram!(\n    plot,\n    risk_averse_total_profit;\n    label = \"CV@R\",\n    bins = bins,\n    alpha = 0.5,\n)\nplot","category":"page"},{"location":"tutorials/applications/two_stage_stochastic/#Next-steps","page":"Two-stage stochastic programs","title":"Next steps","text":"","category":"section"},{"location":"tutorials/applications/two_stage_stochastic/","page":"Two-stage stochastic programs","title":"Two-stage stochastic programs","text":"Try solving this problem for different numbers of samples and different distributions.\nRefactor the example to avoid hard-coding the costs. What happens to the solution if the cost of disposing unsold pies increases?\nPlot the optimal number of pies to make for different values of the risk aversion parameter gamma. What is the relationship?","category":"page"},{"location":"packages/solvers/#Introduction","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"packages/solvers/","page":"Introduction","title":"Introduction","text":"This section of the documentation contains brief documentation for some of the solvers that JuMP supports. The list of solvers is not exhaustive, but instead is intended to help you discover commonly used solvers.","category":"page"},{"location":"packages/solvers/#Affiliation","page":"Introduction","title":"Affiliation","text":"","category":"section"},{"location":"packages/solvers/","page":"Introduction","title":"Introduction","text":"Packages beginning with jump-dev/ are developed and maintained by the JuMP developers. In many cases, these packages wrap external solvers that are not developed by the JuMP developers and, while the Julia packages are all open-source, in some cases the solvers themselves are closed source commercial products.","category":"page"},{"location":"packages/solvers/","page":"Introduction","title":"Introduction","text":"Packages that do not begin with jump-dev/ are developed independently. The developers of these packages requested or consented to the inclusion of their README contents in the JuMP documentation for the benefit of users.","category":"page"},{"location":"packages/solvers/#Adding-new-solvers","page":"Introduction","title":"Adding new solvers","text":"","category":"section"},{"location":"packages/solvers/","page":"Introduction","title":"Introduction","text":"Written a solver? Add it to this section of the JuMP documentation by making a pull request to the docs/packages.toml file.","category":"page"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/models.md\"","category":"page"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/models/#Models","page":"Models","title":"Models","text":"","category":"section"},{"location":"moi/reference/models/#Attribute-interface","page":"Models","title":"Attribute interface","text":"","category":"section"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"is_set_by_optimize\nis_copyable\nget\nget!\nset\nsupports\nattribute_value_type","category":"page"},{"location":"moi/reference/models/#MathOptInterface.is_set_by_optimize","page":"Models","title":"MathOptInterface.is_set_by_optimize","text":"is_set_by_optimize(::AnyAttribute)\n\nReturn a Bool indicating whether the value of the attribute is modified during an optimize! call, that is, the attribute is used to query the result of the optimization.\n\nImportant note when defining new attributes\n\nThis function returns false by default so it should be implemented for attributes that are modified by optimize!.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.is_copyable","page":"Models","title":"MathOptInterface.is_copyable","text":"is_copyable(::AnyAttribute)\n\nReturn a Bool indicating whether the value of the attribute may be copied during copy_to using set.\n\nImportant note when defining new attributes\n\nBy default is_copyable(attr) returns !is_set_by_optimize(attr). A specific method should be defined for attributes which are copied indirectly during copy_to. For instance, both is_copyable and is_set_by_optimize return false for the following attributes:\n\nListOfOptimizerAttributesSet, ListOfModelAttributesSet, ListOfConstraintAttributesSet and ListOfVariableAttributesSet.\nSolverName and RawSolver: these attributes cannot be set.\nNumberOfVariables and ListOfVariableIndices: these attributes are set indirectly by add_variable and add_variables.\nObjectiveFunctionType: this attribute is set indirectly when setting the ObjectiveFunction attribute.\nNumberOfConstraints, ListOfConstraintIndices, ListOfConstraintTypesPresent, CanonicalConstraintFunction, ConstraintFunction and ConstraintSet: these attributes are set indirectly by add_constraint and add_constraints.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.get","page":"Models","title":"MathOptInterface.get","text":"get(optimizer::AbstractOptimizer, attr::AbstractOptimizerAttribute)\n\nReturn an attribute attr of the optimizer optimizer.\n\nget(model::ModelLike, attr::AbstractModelAttribute)\n\nReturn an attribute attr of the model model.\n\nget(model::ModelLike, attr::AbstractVariableAttribute, v::VariableIndex)\n\nIf the attribute attr is set for the variable v in the model model, return its value, return nothing otherwise. If the attribute attr is not supported by model then an error should be thrown instead of returning nothing.\n\nget(model::ModelLike, attr::AbstractVariableAttribute, v::Vector{VariableIndex})\n\nReturn a vector of attributes corresponding to each variable in the collection v in the model model.\n\nget(model::ModelLike, attr::AbstractConstraintAttribute, c::ConstraintIndex)\n\nIf the attribute attr is set for the constraint c in the model model, return its value, return nothing otherwise. If the attribute attr is not supported by model then an error should be thrown instead of returning nothing.\n\nget(\n    model::ModelLike,\n    attr::AbstractConstraintAttribute,\n    c::Vector{ConstraintIndex{F,S}},\n) where {F,S}\n\nReturn a vector of attributes corresponding to each constraint in the collection c in the model model.\n\nget(model::ModelLike, ::Type{VariableIndex}, name::String)\n\nIf a variable with name name exists in the model model, return the corresponding index, otherwise return nothing. Errors if two variables have the same name.\n\nget(\n    model::ModelLike,\n    ::Type{ConstraintIndex{F,S}},\n    name::String,\n) where {F,S}\n\nIf an F-in-S constraint with name name exists in the model model, return the corresponding index, otherwise return nothing. Errors if two constraints have the same name.\n\nget(model::ModelLike, ::Type{ConstraintIndex}, name::String)\n\nIf any constraint with name name exists in the model model, return the corresponding index, otherwise return nothing. This version is available for convenience but may incur a performance penalty because it is not type stable. Errors if two constraints have the same name.\n\n\n\n\n\nMOI.get(b::AbstractBridge, ::MOI.NumberOfVariables)::Int64\n\nReturn the number of variables created by the bridge b in the model.\n\nSee also MOI.NumberOfConstraints.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this if the bridge adds new variables.\n\n\n\n\n\nMOI.get(b::AbstractBridge, ::MOI.ListOfVariableIndices)\n\nReturn the list of variables created by the bridge b.\n\nSee also MOI.ListOfVariableIndices.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this if the bridge adds new variables.\n\n\n\n\n\nMOI.get(b::AbstractBridge, ::MOI.NumberOfConstraints{F,S})::Int64 where {F,S}\n\nReturn the number of constraints of the type F-in-S created by the bridge b.\n\nSee also MOI.NumberOfConstraints.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this for the constraint types returned by added_constraint_types.\n\n\n\n\n\nMOI.get(b::AbstractBridge, ::MOI.ListOfConstraintIndices{F,S}) where {F,S}\n\nReturn a  Vector{ConstraintIndex{F,S}} with indices of all constraints of type F-in-S created by the bride b.\n\nSee also MOI.ListOfConstraintIndices.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this for the constraint types returned by added_constraint_types.\n\n\n\n\n\nfunction MOI.get(\n    model::MOI.ModelLike,\n    attr::MOI.AbstractConstraintAttribute,\n    bridge::AbstractBridge,\n)\n\nReturn the value of the attribute attr of the model model for the constraint bridged by bridge.\n\n\n\n\n\nget(model::GenericModel, attr::MathOptInterface.AbstractOptimizerAttribute)\n\nReturn the value of the attribute attr from the model's MOI backend.\n\n\n\n\n\nget(model::GenericModel, attr::MathOptInterface.AbstractModelAttribute)\n\nReturn the value of the attribute attr from the model's MOI backend.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.get!","page":"Models","title":"MathOptInterface.get!","text":"get!(output, model::ModelLike, args...)\n\nAn in-place version of get.\n\nThe signature matches that of get except that the the result is placed in the vector output.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.set","page":"Models","title":"MathOptInterface.set","text":"set(optimizer::AbstractOptimizer, attr::AbstractOptimizerAttribute, value)\n\nAssign value to the attribute attr of the optimizer optimizer.\n\nset(model::ModelLike, attr::AbstractModelAttribute, value)\n\nAssign value to the attribute attr of the model model.\n\nset(model::ModelLike, attr::AbstractVariableAttribute, v::VariableIndex, value)\n\nAssign value to the attribute attr of variable v in model model.\n\nset(\n    model::ModelLike,\n    attr::AbstractVariableAttribute,\n    v::Vector{VariableIndex},\n    vector_of_values,\n)\n\nAssign a value respectively to the attribute attr of each variable in the collection v in model model.\n\nset(\n    model::ModelLike,\n    attr::AbstractConstraintAttribute,\n    c::ConstraintIndex,\n    value,\n)\n\nAssign a value to the attribute attr of constraint c in model model.\n\nset(\n    model::ModelLike,\n    attr::AbstractConstraintAttribute,\n    c::Vector{ConstraintIndex{F,S}},\n    vector_of_values,\n) where {F,S}\n\nAssign a value respectively to the attribute attr of each constraint in the collection c in model model.\n\nAn UnsupportedAttribute error is thrown if model does not support the attribute attr (see supports) and a SetAttributeNotAllowed error is thrown if it supports the attribute attr but it cannot be set.\n\nset(\n    model::ModelLike,\n    ::ConstraintSet,\n    c::ConstraintIndex{F,S},\n    set::S,\n) where {F,S}\n\nChange the set of constraint c to the new set set which should be of the same type as the original set.\n\nset(\n    model::ModelLike,\n    ::ConstraintFunction,\n    c::ConstraintIndex{F,S},\n    func::F,\n) where {F,S}\n\nReplace the function in constraint c with func. F must match the original function type used to define the constraint.\n\nnote: Note\nSetting the constraint function is not allowed if F is VariableIndex; a SettingVariableIndexNotAllowed error is thrown instead. This is because, it would require changing the index c since the index of VariableIndex constraints must be the same as the index of the variable.\n\n\n\n\n\nfunction MOI.set(\n    model::MOI.ModelLike,\n    attr::MOI.AbstractConstraintAttribute,\n    bridge::AbstractBridge,\n    value,\n)\n\nSet the value of the attribute attr of the model model for the constraint bridged by bridge.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.supports","page":"Models","title":"MathOptInterface.supports","text":"supports(model::ModelLike, sub::AbstractSubmittable)::Bool\n\nReturn a Bool indicating whether model supports the submittable sub.\n\nsupports(model::ModelLike, attr::AbstractOptimizerAttribute)::Bool\n\nReturn a Bool indicating whether model supports the optimizer attribute attr. That is, it returns false if copy_to(model, src) shows a warning in case attr is in the ListOfOptimizerAttributesSet of src; see copy_to for more details on how unsupported optimizer attributes are handled in copy.\n\nsupports(model::ModelLike, attr::AbstractModelAttribute)::Bool\n\nReturn a Bool indicating whether model supports the model attribute attr. That is, it returns false if copy_to(model, src) cannot be performed in case attr is in the ListOfModelAttributesSet of src.\n\nsupports(\n    model::ModelLike,\n    attr::AbstractVariableAttribute,\n    ::Type{VariableIndex},\n)::Bool\n\nReturn a Bool indicating whether model supports the variable attribute attr. That is, it returns false if copy_to(model, src) cannot be performed in case attr is in the ListOfVariableAttributesSet of src.\n\nsupports(\n    model::ModelLike,\n    attr::AbstractConstraintAttribute,\n    ::Type{ConstraintIndex{F,S}},\n)::Bool where {F,S}\n\nReturn a Bool indicating whether model supports the constraint attribute attr applied to an F-in-S constraint. That is, it returns false if copy_to(model, src) cannot be performed in case attr is in the ListOfConstraintAttributesSet of src.\n\nFor all five methods, if the attribute is only not supported in specific circumstances, it should still return true.\n\nNote that supports is only defined for attributes for which is_copyable returns true as other attributes do not appear in the list of attributes set obtained by ListOf...AttributesSet.\n\n\n\n\n\nMOI.supports(\n    model::MOI.ModelLike,\n    attr::MOI.AbstractConstraintAttribute,\n    BT::Type{<:AbstractBridge},\n)\n\nReturn a Bool indicating whether BT supports setting attr to model.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.attribute_value_type","page":"Models","title":"MathOptInterface.attribute_value_type","text":"attribute_value_type(attr::AnyAttribute)\n\nGiven an attribute attr, return the type of value expected by get, or returned by set.\n\nNotes\n\nOnly implement this if it make sense to do so. If un-implemented, the default is Any.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#Model-interface","page":"Models","title":"Model interface","text":"","category":"section"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"ModelLike\nis_empty\nempty!\nwrite_to_file\nread_from_file\nsupports_incremental_interface\ncopy_to\nIndexMap","category":"page"},{"location":"moi/reference/models/#MathOptInterface.ModelLike","page":"Models","title":"MathOptInterface.ModelLike","text":"ModelLike\n\nAbstract supertype for objects that implement the \"Model\" interface for defining an optimization problem.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.is_empty","page":"Models","title":"MathOptInterface.is_empty","text":"is_empty(model::ModelLike)\n\nReturns false if the model has any model attribute set or has any variables or constraints.\n\nNote that an empty model can have optimizer attributes set.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.empty!","page":"Models","title":"MathOptInterface.empty!","text":"empty!(model::ModelLike)\n\nEmpty the model, that is, remove all variables, constraints and model attributes but not optimizer attributes.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.write_to_file","page":"Models","title":"MathOptInterface.write_to_file","text":"write_to_file(model::ModelLike, filename::String)\n\nWrite the current model to the file at filename.\n\nSupported file types depend on the model type.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.read_from_file","page":"Models","title":"MathOptInterface.read_from_file","text":"read_from_file(model::ModelLike, filename::String)\n\nRead the file filename into the model model. If model is non-empty, this may throw an error.\n\nSupported file types depend on the model type.\n\nNote\n\nOnce the contents of the file are loaded into the model, users can query the variables via get(model, ListOfVariableIndices()). However, some filetypes, such as LP files, do not maintain an explicit ordering of the variables. Therefore, the returned list may be in an arbitrary order.\n\nTo avoid depending on the order of the indices, look up each variable index by name using  get(model, VariableIndex, \"name\").\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.supports_incremental_interface","page":"Models","title":"MathOptInterface.supports_incremental_interface","text":"supports_incremental_interface(model::ModelLike)\n\nReturn a Bool indicating whether model supports building incrementally via add_variable and add_constraint.\n\nThe main purpose of this function is to determine whether a model can be loaded into model incrementally or whether it should be cached and copied at once instead.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.copy_to","page":"Models","title":"MathOptInterface.copy_to","text":"copy_to(dest::ModelLike, src::ModelLike)::IndexMap\n\nCopy the model from src into dest.\n\nThe target dest is emptied, and all previous indices to variables and constraints in dest are invalidated.\n\nReturns an IndexMap object that translates variable and constraint indices from the src model to the corresponding indices in the dest model.\n\nNotes\n\nIf a constraint that in src is not supported by dest, then an UnsupportedConstraint error is thrown.\nIf an AbstractModelAttribute, AbstractVariableAttribute, or AbstractConstraintAttribute is set in src but not supported by dest, then an UnsupportedAttribute error is thrown.\n\nAbstractOptimizerAttributes are not copied  to the dest model.\n\nIndexMap\n\nImplementations of copy_to must return an IndexMap. For technical reasons, this type is defined in the Utilities submodule as MOI.Utilities.IndexMap. However, since it is an integral part of the MOI API, we provide MOI.IndexMap as an alias.\n\nExample\n\n# Given empty `ModelLike` objects `src` and `dest`.\n\nx = add_variable(src)\n\nis_valid(src, x)   # true\nis_valid(dest, x)  # false (`dest` has no variables)\n\nindex_map = copy_to(dest, src)\nis_valid(dest, x) # false (unless index_map[x] == x)\nis_valid(dest, index_map[x]) # true\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.IndexMap","page":"Models","title":"MathOptInterface.IndexMap","text":"IndexMap()\n\nThe dictionary-like object returned by copy_to.\n\nIndexMap\n\nImplementations of copy_to must return an IndexMap. For technical reasons, the IndexMap type is defined in the Utilities submodule as MOI.Utilities.IndexMap. However, since it is an integral part of the MOI API, we provide this MOI.IndexMap as an alias.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#Model-attributes","page":"Models","title":"Model attributes","text":"","category":"section"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"AbstractModelAttribute\nName\nObjectiveFunction\nObjectiveFunctionType\nObjectiveSense\nOptimizationSense\nMIN_SENSE\nMAX_SENSE\nFEASIBILITY_SENSE\nNumberOfVariables\nListOfVariableIndices\nListOfConstraintTypesPresent\nNumberOfConstraints\nListOfConstraintIndices\nListOfOptimizerAttributesSet\nListOfModelAttributesSet\nListOfVariableAttributesSet\nListOfConstraintAttributesSet\nUserDefinedFunction\nListOfSupportedNonlinearOperators","category":"page"},{"location":"moi/reference/models/#MathOptInterface.AbstractModelAttribute","page":"Models","title":"MathOptInterface.AbstractModelAttribute","text":"AbstractModelAttribute\n\nAbstract supertype for attribute objects that can be used to set or get attributes (properties) of the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.Name","page":"Models","title":"MathOptInterface.Name","text":"Name()\n\nA model attribute for the string identifying the model. It has a default value of \"\" if not set`.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ObjectiveFunction","page":"Models","title":"MathOptInterface.ObjectiveFunction","text":"ObjectiveFunction{F<:AbstractScalarFunction}()\n\nA model attribute for the objective function which has a type F<:AbstractScalarFunction.\n\nF should be guaranteed to be equivalent but not necessarily identical to the function type provided by the user.\n\nThrows an InexactError if the objective function cannot be converted to F, e.g., the objective function is quadratic and F is ScalarAffineFunction{Float64} or it has non-integer coefficient and F is ScalarAffineFunction{Int}.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ObjectiveFunctionType","page":"Models","title":"MathOptInterface.ObjectiveFunctionType","text":"ObjectiveFunctionType()\n\nA model attribute for the type F of the objective function set using the ObjectiveFunction{F} attribute.\n\nExamples\n\nIn the following code, attr should be equal to MOI.VariableIndex:\n\nx = MOI.add_variable(model)\nMOI.set(model, MOI.ObjectiveFunction{MOI.VariableIndex}(), x)\nattr = MOI.get(model, MOI.ObjectiveFunctionType())\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ObjectiveSense","page":"Models","title":"MathOptInterface.ObjectiveSense","text":"ObjectiveSense()\n\nA model attribute for the objective sense of the objective function, which must be an OptimizationSense: MIN_SENSE, MAX_SENSE, or FEASIBILITY_SENSE. The default is FEASIBILITY_SENSE.\n\nInteraction with ObjectiveFunction\n\nSetting the sense to FEASIBILITY_SENSE unsets the ObjectiveFunction attribute. That is, if you first set ObjectiveFunction and then set ObjectiveSense to be FEASIBILITY_SENSE, no objective function will be passed to the solver.\n\nIn addition, some reformulations of ObjectiveFunction via bridges rely on the value of ObjectiveSense. Therefore, you should set ObjectiveSense before setting ObjectiveFunction.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.OptimizationSense","page":"Models","title":"MathOptInterface.OptimizationSense","text":"OptimizationSense\n\nAn enum for the value of the ObjectiveSense attribute.\n\nValues\n\nPossible values are:\n\nMIN_SENSE: the goal is to minimize the objective function\nMAX_SENSE: the goal is to maximize the objective function\nFEASIBILITY_SENSE: the model does not have an objective function\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.MIN_SENSE","page":"Models","title":"MathOptInterface.MIN_SENSE","text":"MIN_SENSE::OptimizationSense\n\nAn instance of the OptimizationSense enum.\n\nMIN_SENSE: the goal is to minimize the objective function\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.MAX_SENSE","page":"Models","title":"MathOptInterface.MAX_SENSE","text":"MAX_SENSE::OptimizationSense\n\nAn instance of the OptimizationSense enum.\n\nMAX_SENSE: the goal is to maximize the objective function\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.FEASIBILITY_SENSE","page":"Models","title":"MathOptInterface.FEASIBILITY_SENSE","text":"FEASIBILITY_SENSE::OptimizationSense\n\nAn instance of the OptimizationSense enum.\n\nFEASIBILITY_SENSE: the model does not have an objective function\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.NumberOfVariables","page":"Models","title":"MathOptInterface.NumberOfVariables","text":"NumberOfVariables()\n\nA model attribute for the number of variables in the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfVariableIndices","page":"Models","title":"MathOptInterface.ListOfVariableIndices","text":"ListOfVariableIndices()\n\nA model attribute for the Vector{VariableIndex} of all variable indices present in the model (i.e., of length equal to the value of NumberOfVariables in the order in which they were added.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfConstraintTypesPresent","page":"Models","title":"MathOptInterface.ListOfConstraintTypesPresent","text":"ListOfConstraintTypesPresent()\n\nA model attribute for the list of tuples of the form (F,S), where F is a function type and S is a set type indicating that the attribute NumberOfConstraints{F,S} has a value greater than zero.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.NumberOfConstraints","page":"Models","title":"MathOptInterface.NumberOfConstraints","text":"NumberOfConstraints{F,S}()\n\nA model attribute for the number of constraints of the type F-in-S present in the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfConstraintIndices","page":"Models","title":"MathOptInterface.ListOfConstraintIndices","text":"ListOfConstraintIndices{F,S}()\n\nA model attribute for the Vector{ConstraintIndex{F,S}} of all constraint indices of type F-in-S in the model (i.e., of length equal to the value of NumberOfConstraints{F,S}) in the order in which they were added.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfOptimizerAttributesSet","page":"Models","title":"MathOptInterface.ListOfOptimizerAttributesSet","text":"ListOfOptimizerAttributesSet()\n\nAn optimizer attribute for the Vector{AbstractOptimizerAttribute} of all optimizer attributes that were set.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfModelAttributesSet","page":"Models","title":"MathOptInterface.ListOfModelAttributesSet","text":"ListOfModelAttributesSet()\n\nA model attribute for the Vector{AbstractModelAttribute} of all model attributes attr such that:\n\nis_copyable(attr) returns true, and\nthe attribute was set to the model\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfVariableAttributesSet","page":"Models","title":"MathOptInterface.ListOfVariableAttributesSet","text":"ListOfVariableAttributesSet()\n\nA model attribute for the Vector{AbstractVariableAttribute} of all variable attributes attr such that 1) is_copyable(attr) returns true and 2) the attribute was set to variables.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfConstraintAttributesSet","page":"Models","title":"MathOptInterface.ListOfConstraintAttributesSet","text":"ListOfConstraintAttributesSet{F, S}()\n\nA model attribute for the Vector{AbstractConstraintAttribute} of all constraint attributes attr such that:\n\nis_copyable(attr) returns true and\nthe attribute was set to F-in-S constraints.\n\nNote\n\nThe attributes ConstraintFunction and ConstraintSet should not be included in the list even if then have been set with set.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.UserDefinedFunction","page":"Models","title":"MathOptInterface.UserDefinedFunction","text":"UserDefinedFunction(name::Symbol, arity::Int) <: AbstractModelAttribute\n\nSet this attribute to register a user-defined function by the name of name with arity arguments.\n\nOnce registered, name will appear in ListOfSupportedNonlinearOperators.\n\nYou cannot register multiple UserDefinedFunctions with the same name but different arity.\n\nValue type\n\nThe value to be set is a tuple containing one, two, or three functions to evaluate the function, the first-order derivative, and the second-order derivative respectively. Both derivatives are optional, but if you pass the second-order derivative you must also pass the first-order derivative.\n\nFor univariate functions with arity == 1, the functions in the tuple must have the form:\n\nf(x::T)::T: returns the value of the function at x\n∇f(x::T)::T: returns the first-order derivative of f with respect to x\n∇²f(x::T)::T: returns the second-order derivative of f with respect to x.\n\nFor multivariate functions with arity > 1, the functions in the tuple must have the form:\n\nf(x::T...)::T: returns the value of the function at x\n∇f(g::AbstractVector{T}, x::T...)::Nothing: fills the components of g, with g[i] being the first-order partial derivative of f with respect to x[i]\n∇²f(H::AbstractMatrix{T}, x::T...)::Nothing: fills the non-zero components  of H, with H[i, j] being the second-order partial derivative of f with  respect to x[i] and then x[j]. H is initialized to the zero matrix,  so you do not need to set any zero elements.\n\nExamples\n\njulia> import MathOptInterface as MOI\n\njulia> f(x, y) = x^2 + y^2\nf (generic function with 1 method)\n\njulia> function ∇f(g, x, y)\n           g .= 2 * x, 2 * y\n           return\n       end\n∇f (generic function with 1 method)\n\njulia> function ∇²f(H, x...)\n           H[1, 1] = H[2, 2] = 2.0\n           return\n       end\n∇²f (generic function with 1 method)\n\njulia> model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())\nMOIU.UniversalFallback{MOIU.Model{Float64}}\nfallback for MOIU.Model{Float64}\n\njulia> MOI.set(model, MOI.UserDefinedFunction(:f, 2), (f,))\n\njulia> MOI.set(model, MOI.UserDefinedFunction(:g, 2), (f, ∇f))\n\njulia> MOI.set(model, MOI.UserDefinedFunction(:h, 2), (f, ∇f, ∇²f))\n\njulia> x = MOI.add_variables(model, 2)\n2-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n\njulia> MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)\n\njulia> obj_f = MOI.ScalarNonlinearFunction(:f, Any[x[1], x[2]])\nf(MOI.VariableIndex(1), MOI.VariableIndex(2))\n\njulia> MOI.set(model, MOI.ObjectiveFunction{typeof(obj_f)}(), obj_f)\n\njulia> print(model)\nMinimize ScalarNonlinearFunction:\n f(v[1], v[2])\n\nSubject to:\n\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ListOfSupportedNonlinearOperators","page":"Models","title":"MathOptInterface.ListOfSupportedNonlinearOperators","text":"ListOfSupportedNonlinearOperators() <: AbstractModelAttribute\n\nWhen queried with get, return a Vector{Symbol} listing the operators supported by the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#Optimizer-interface","page":"Models","title":"Optimizer interface","text":"","category":"section"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"AbstractOptimizer\nOptimizerWithAttributes\noptimize!\noptimize!(::ModelLike, ::ModelLike)\ninstantiate\ndefault_cache","category":"page"},{"location":"moi/reference/models/#MathOptInterface.AbstractOptimizer","page":"Models","title":"MathOptInterface.AbstractOptimizer","text":"AbstractOptimizer <: ModelLike\n\nAbstract supertype for objects representing an instance of an optimization problem tied to a particular solver. This is typically a solver's in-memory representation. In addition to ModelLike, AbstractOptimizer objects let you solve the model and query the solution.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.OptimizerWithAttributes","page":"Models","title":"MathOptInterface.OptimizerWithAttributes","text":"struct OptimizerWithAttributes\n    optimizer_constructor\n    params::Vector{Pair{AbstractOptimizerAttribute,<:Any}}\nend\n\nObject grouping an optimizer constructor and a list of optimizer attributes. Instances are created with instantiate.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.optimize!","page":"Models","title":"MathOptInterface.optimize!","text":"optimize!(optimizer::AbstractOptimizer)\n\nOptimize the problem contained in optimizer.\n\nBefore calling optimize!, the problem should first be constructed using the incremental interface (see supports_incremental_interface) or copy_to.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.optimize!-Tuple{MathOptInterface.ModelLike, MathOptInterface.ModelLike}","page":"Models","title":"MathOptInterface.optimize!","text":"optimize!(dest::AbstractOptimizer, src::ModelLike)::Tuple{IndexMap,Bool}\n\nA \"one-shot\" call that copies the problem from src into dest and then uses dest to optimize the problem.\n\nReturns a tuple of an IndexMap and a Bool copied.\n\nThe IndexMap object translates variable and constraint indices from the src model to the corresponding indices in the dest optimizer. See copy_to for details.\nIf copied == true, src was copied to dest and then cached, allowing incremental modification if supported by the solver.\nIf copied == false, a cache of the model was not kept in dest. Therefore, only the solution information (attributes for which is_set_by_optimize is true) is available to query.\n\nnote: Note\nThe main purpose of optimize! method with two arguments is for use in Utilities.CachingOptimizer.\n\nRelationship to the single-argument optimize!\n\nThe default fallback of optimize!(dest::AbstractOptimizer, src::ModelLike) is\n\nfunction optimize!(dest::AbstractOptimizer, src::ModelLike)\n    index_map = copy_to(dest, src)\n    optimize!(dest)\n    return index_map, true\nend\n\nTherefore, subtypes of AbstractOptimizer should either implement this two-argument method, or implement both copy_to(::Optimizer, ::ModelLike) and optimize!(::Optimizer).\n\n\n\n\n\n","category":"method"},{"location":"moi/reference/models/#MathOptInterface.instantiate","page":"Models","title":"MathOptInterface.instantiate","text":"instantiate(\n    optimizer_constructor,\n    with_cache_type::Union{Nothing,Type} = nothing,\n    with_bridge_type::Union{Nothing,Type} = nothing,\n)\n\nCreate an instance of an optimizer by either:\n\ncalling optimizer_constructor.optimizer_constructor() and setting the parameters in optimizer_constructor.params if optimizer_constructor is a OptimizerWithAttributes\ncalling optimizer_constructor() if optimizer_constructor is callable.\n\nwithcachetype\n\nIf with_cache_type is not nothing, then the optimizer is wrapped in a Utilities.CachingOptimizer to store a cache of the model. This is most useful if the optimizer you are constructing does not support the incremental interface (see supports_incremental_interface).\n\nwithbridgetype\n\nIf with_bridge_type is not nothing, the optimizer is wrapped in a Bridges.full_bridge_optimizer, enabling all the bridges defined in the MOI.Bridges submodule with coefficient type with_bridge_type.\n\nIn addition, if the optimizer created by optimizer_constructor does not support the incremental interface (see supports_incremental_interface), then, irrespective of with_cache_type, the optimizer is wrapped in a Utilities.CachingOptimizer to store a cache of the bridged model.\n\nIf with_cache_type and with_bridge_type are both not nothing, then they must be the same type.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.default_cache","page":"Models","title":"MathOptInterface.default_cache","text":"default_cache(optimizer::ModelLike, ::Type{T}) where {T}\n\nReturn a new instance of the default model type to be used as cache for optimizer in a Utilities.CachingOptimizer for holding constraints of coefficient type T. By default, this returns Utilities.UniversalFallback(Utilities.Model{T}()). If copying from a instance of a given model type is faster for optimizer then a new method returning an instance of this model type should be defined.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#Optimizer-attributes","page":"Models","title":"Optimizer attributes","text":"","category":"section"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"AbstractOptimizerAttribute\nSolverName\nSolverVersion\nSilent\nTimeLimitSec\nObjectiveLimit\nRawOptimizerAttribute\nNumberOfThreads\nRawSolver\nAbsoluteGapTolerance\nRelativeGapTolerance","category":"page"},{"location":"moi/reference/models/#MathOptInterface.AbstractOptimizerAttribute","page":"Models","title":"MathOptInterface.AbstractOptimizerAttribute","text":"AbstractOptimizerAttribute\n\nAbstract supertype for attribute objects that can be used to set or get attributes (properties) of the optimizer.\n\nNotes\n\nThe difference between AbstractOptimizerAttribute and AbstractModelAttribute lies in the behavior of is_empty, empty! and copy_to. Typically optimizer attributes affect only how the model is solved.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.SolverName","page":"Models","title":"MathOptInterface.SolverName","text":"SolverName()\n\nAn optimizer attribute for the string identifying the solver/optimizer.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.SolverVersion","page":"Models","title":"MathOptInterface.SolverVersion","text":"SolverVersion()\n\nAn optimizer attribute for the string identifying the version of the solver.\n\nnote: Note\nFor solvers supporting semantic versioning, the SolverVersion should be a string of the form \"vMAJOR.MINOR.PATCH\", so that it can be converted to a Julia VersionNumber (e.g., `VersionNumber(\"v1.2.3\")).We do not require Semantic Versioning because some solvers use alternate versioning systems. For example, CPLEX uses Calendar Versioning, so SolverVersion will return a string like \"202001\".\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.Silent","page":"Models","title":"MathOptInterface.Silent","text":"Silent()\n\nAn optimizer attribute for silencing the output of an optimizer. When set to true, it takes precedence over any other attribute controlling verbosity and requires the solver to produce no output. The default value is false which has no effect. In this case the verbosity is controlled by other attributes.\n\nNote\n\nEvery optimizer should have verbosity on by default. For instance, if a solver has a solver-specific log level attribute, the MOI implementation should set it to 1 by default. If the user sets Silent to true, then the log level should be set to 0, even if the user specifically sets a value of log level. If the value of Silent is false then the log level set to the solver is the value given by the user for this solver-specific parameter or 1 if none is given.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.TimeLimitSec","page":"Models","title":"MathOptInterface.TimeLimitSec","text":"TimeLimitSec()\n\nAn optimizer attribute for setting a time limit (in seconnds) for an optimization. When set to nothing, it deactivates the solver time limit. The default value is nothing.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ObjectiveLimit","page":"Models","title":"MathOptInterface.ObjectiveLimit","text":"ObjectiveLimit()\n\nAn optimizer attribute for setting a limit on the objective value.\n\nThe provided limit must be a Union{Real,Nothing}.\n\nWhen set to nothing, the limit reverts to the solver's default.\n\nThe default value is nothing.\n\nThe solver may stop when the ObjectiveValue is better (lower for minimization, higher for maximization) than the ObjectiveLimit. If stopped, the TerminationStatus should be OBJECTIVE_LIMIT.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.RawOptimizerAttribute","page":"Models","title":"MathOptInterface.RawOptimizerAttribute","text":"RawOptimizerAttribute(name::String)\n\nAn optimizer attribute for the solver-specific parameter identified by name.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.NumberOfThreads","page":"Models","title":"MathOptInterface.NumberOfThreads","text":"NumberOfThreads()\n\nAn optimizer attribute for setting the number of threads used for an optimization. When set to nothing uses solver default. Values are positive integers. The default value is nothing.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.RawSolver","page":"Models","title":"MathOptInterface.RawSolver","text":"RawSolver()\n\nA model attribute for the object that may be used to access a solver-specific API for this optimizer.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.AbsoluteGapTolerance","page":"Models","title":"MathOptInterface.AbsoluteGapTolerance","text":"AbsoluteGapTolerance()\n\nAn optimizer attribute for setting the absolute gap tolerance for an optimization. This is an optimizer attribute, and should be set before calling optimize!. When set to nothing (if supported), uses solver default.\n\nTo set a relative gap tolerance, see RelativeGapTolerance.\n\nwarning: Warning\nThe mathematical definition of \"absolute gap\", and its treatment during the optimization, are solver-dependent. However, assuming no other limit nor issue is encountered during the optimization, most solvers that implement this attribute will stop once f - b  g_abs, where b is the best bound, f is the best feasible objective value, and g_abs is the absolute gap.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.RelativeGapTolerance","page":"Models","title":"MathOptInterface.RelativeGapTolerance","text":"RelativeGapTolerance()\n\nAn optimizer attribute for setting the relative gap tolerance for an optimization. This is an optimizer attribute, and should be set before calling optimize!. When set to nothing (if supported), uses solver default.\n\nIf you are looking for the relative gap of the current best solution, see RelativeGap. If no limit nor issue is encountered during the optimization, the value of RelativeGap should be at most as large as RelativeGapTolerance.\n\n# Before optimizing: set relative gap tolerance\n# set 0.1% relative gap tolerance\nMOI.set(model, MOI.RelativeGapTolerance(), 1e-3)\nMOI.optimize!(model)\n\n# After optimizing (assuming all went well)\n# The relative gap tolerance has not changed...\nMOI.get(model, MOI.RelativeGapTolerance())  # returns 1e-3\n# ... and the relative gap of the obtained solution is smaller or equal to the\n# tolerance\nMOI.get(model, MOI.RelativeGap())  # should return something ≤ 1e-3\n\nwarning: Warning\nThe mathematical definition of \"relative gap\", and its allowed range, are solver-dependent. Typically, solvers expect a value between 0.0 and 1.0.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"List of attributes useful for optimizers","category":"page"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"TerminationStatus\nTerminationStatusCode\nOPTIMIZE_NOT_CALLED\nOPTIMAL\nINFEASIBLE\nDUAL_INFEASIBLE\nLOCALLY_SOLVED\nLOCALLY_INFEASIBLE\nINFEASIBLE_OR_UNBOUNDED\nALMOST_OPTIMAL\nALMOST_INFEASIBLE\nALMOST_DUAL_INFEASIBLE\nALMOST_LOCALLY_SOLVED\nITERATION_LIMIT\nTIME_LIMIT\nNODE_LIMIT\nSOLUTION_LIMIT\nMEMORY_LIMIT\nOBJECTIVE_LIMIT\nNORM_LIMIT\nOTHER_LIMIT\nSLOW_PROGRESS\nNUMERICAL_ERROR\nINVALID_MODEL\nINVALID_OPTION\nINTERRUPTED\nOTHER_ERROR\nPrimalStatus\nDualStatus\nRawStatusString\nResultCount\nObjectiveValue\nDualObjectiveValue\nObjectiveBound\nRelativeGap\nSolveTimeSec\nSimplexIterations\nBarrierIterations\nNodeCount","category":"page"},{"location":"moi/reference/models/#MathOptInterface.TerminationStatus","page":"Models","title":"MathOptInterface.TerminationStatus","text":"TerminationStatus()\n\nA model attribute for the TerminationStatusCode explaining why the optimizer stopped.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.TerminationStatusCode","page":"Models","title":"MathOptInterface.TerminationStatusCode","text":"TerminationStatusCode\n\nAn Enum of possible values for the TerminationStatus attribute. This attribute is meant to explain the reason why the optimizer stopped executing in the most recent call to optimize!.\n\nValues\n\nPossible values are:\n\nOPTIMIZE_NOT_CALLED: The algorithm has not started.\nOPTIMAL: The algorithm found a globally optimal solution.\nINFEASIBLE: The algorithm concluded that no feasible solution exists.\nDUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.\nLOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.\nLOCALLY_INFEASIBLE: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.\nINFEASIBLE_OR_UNBOUNDED: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.\nALMOST_OPTIMAL: The algorithm found a globally optimal solution to relaxed tolerances.\nALMOST_INFEASIBLE: The algorithm concluded that no feasible solution exists within relaxed tolerances.\nALMOST_DUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.\nALMOST_LOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.\nITERATION_LIMIT: An iterative algorithm stopped after conducting the maximum number of iterations.\nTIME_LIMIT: The algorithm stopped after a user-specified computation time.\nNODE_LIMIT: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.\nSOLUTION_LIMIT: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.\nMEMORY_LIMIT: The algorithm stopped because it ran out of memory.\nOBJECTIVE_LIMIT: The algorithm stopped because it found a solution better than a minimum limit set by the user.\nNORM_LIMIT: The algorithm stopped because the norm of an iterate became too large.\nOTHER_LIMIT: The algorithm stopped due to a limit not covered by one of the _LIMIT_ statuses above.\nSLOW_PROGRESS: The algorithm stopped because it was unable to continue making progress towards the solution.\nNUMERICAL_ERROR: The algorithm stopped because it encountered unrecoverable numerical error.\nINVALID_MODEL: The algorithm stopped because the model is invalid.\nINVALID_OPTION: The algorithm stopped because it was provided an invalid option.\nINTERRUPTED: The algorithm stopped because of an interrupt signal.\nOTHER_ERROR: The algorithm stopped because of an error not covered by one of the statuses defined above.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.OPTIMIZE_NOT_CALLED","page":"Models","title":"MathOptInterface.OPTIMIZE_NOT_CALLED","text":"OPTIMIZE_NOT_CALLED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOPTIMIZE_NOT_CALLED: The algorithm has not started.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.OPTIMAL","page":"Models","title":"MathOptInterface.OPTIMAL","text":"OPTIMAL::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOPTIMAL: The algorithm found a globally optimal solution.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.INFEASIBLE","page":"Models","title":"MathOptInterface.INFEASIBLE","text":"INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINFEASIBLE: The algorithm concluded that no feasible solution exists.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.DUAL_INFEASIBLE","page":"Models","title":"MathOptInterface.DUAL_INFEASIBLE","text":"DUAL_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nDUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.LOCALLY_SOLVED","page":"Models","title":"MathOptInterface.LOCALLY_SOLVED","text":"LOCALLY_SOLVED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nLOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.LOCALLY_INFEASIBLE","page":"Models","title":"MathOptInterface.LOCALLY_INFEASIBLE","text":"LOCALLY_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nLOCALLY_INFEASIBLE: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.INFEASIBLE_OR_UNBOUNDED","page":"Models","title":"MathOptInterface.INFEASIBLE_OR_UNBOUNDED","text":"INFEASIBLE_OR_UNBOUNDED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINFEASIBLE_OR_UNBOUNDED: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.ALMOST_OPTIMAL","page":"Models","title":"MathOptInterface.ALMOST_OPTIMAL","text":"ALMOST_OPTIMAL::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_OPTIMAL: The algorithm found a globally optimal solution to relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.ALMOST_INFEASIBLE","page":"Models","title":"MathOptInterface.ALMOST_INFEASIBLE","text":"ALMOST_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_INFEASIBLE: The algorithm concluded that no feasible solution exists within relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.ALMOST_DUAL_INFEASIBLE","page":"Models","title":"MathOptInterface.ALMOST_DUAL_INFEASIBLE","text":"ALMOST_DUAL_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_DUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.ALMOST_LOCALLY_SOLVED","page":"Models","title":"MathOptInterface.ALMOST_LOCALLY_SOLVED","text":"ALMOST_LOCALLY_SOLVED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_LOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.ITERATION_LIMIT","page":"Models","title":"MathOptInterface.ITERATION_LIMIT","text":"ITERATION_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nITERATION_LIMIT: An iterative algorithm stopped after conducting the maximum number of iterations.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.TIME_LIMIT","page":"Models","title":"MathOptInterface.TIME_LIMIT","text":"TIME_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nTIME_LIMIT: The algorithm stopped after a user-specified computation time.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.NODE_LIMIT","page":"Models","title":"MathOptInterface.NODE_LIMIT","text":"NODE_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nNODE_LIMIT: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.SOLUTION_LIMIT","page":"Models","title":"MathOptInterface.SOLUTION_LIMIT","text":"SOLUTION_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nSOLUTION_LIMIT: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.MEMORY_LIMIT","page":"Models","title":"MathOptInterface.MEMORY_LIMIT","text":"MEMORY_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nMEMORY_LIMIT: The algorithm stopped because it ran out of memory.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.OBJECTIVE_LIMIT","page":"Models","title":"MathOptInterface.OBJECTIVE_LIMIT","text":"OBJECTIVE_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOBJECTIVE_LIMIT: The algorithm stopped because it found a solution better than a minimum limit set by the user.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.NORM_LIMIT","page":"Models","title":"MathOptInterface.NORM_LIMIT","text":"NORM_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nNORM_LIMIT: The algorithm stopped because the norm of an iterate became too large.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.OTHER_LIMIT","page":"Models","title":"MathOptInterface.OTHER_LIMIT","text":"OTHER_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOTHER_LIMIT: The algorithm stopped due to a limit not covered by one of the _LIMIT_ statuses above.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.SLOW_PROGRESS","page":"Models","title":"MathOptInterface.SLOW_PROGRESS","text":"SLOW_PROGRESS::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nSLOW_PROGRESS: The algorithm stopped because it was unable to continue making progress towards the solution.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.NUMERICAL_ERROR","page":"Models","title":"MathOptInterface.NUMERICAL_ERROR","text":"NUMERICAL_ERROR::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nNUMERICAL_ERROR: The algorithm stopped because it encountered unrecoverable numerical error.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.INVALID_MODEL","page":"Models","title":"MathOptInterface.INVALID_MODEL","text":"INVALID_MODEL::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINVALID_MODEL: The algorithm stopped because the model is invalid.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.INVALID_OPTION","page":"Models","title":"MathOptInterface.INVALID_OPTION","text":"INVALID_OPTION::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINVALID_OPTION: The algorithm stopped because it was provided an invalid option.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.INTERRUPTED","page":"Models","title":"MathOptInterface.INTERRUPTED","text":"INTERRUPTED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINTERRUPTED: The algorithm stopped because of an interrupt signal.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.OTHER_ERROR","page":"Models","title":"MathOptInterface.OTHER_ERROR","text":"OTHER_ERROR::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOTHER_ERROR: The algorithm stopped because of an error not covered by one of the statuses defined above.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.PrimalStatus","page":"Models","title":"MathOptInterface.PrimalStatus","text":"PrimalStatus(result_index::Int = 1)\n\nA model attribute for the ResultStatusCode of the primal result result_index. If result_index is omitted, it defaults to 1.\n\nSee ResultCount for information on how the results are ordered.\n\nIf result_index is larger than the value of ResultCount then NO_SOLUTION is returned.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.DualStatus","page":"Models","title":"MathOptInterface.DualStatus","text":"DualStatus(result_index::Int = 1)\n\nA model attribute for the ResultStatusCode of the dual result result_index. If result_index is omitted, it defaults to 1.\n\nSee ResultCount for information on how the results are ordered.\n\nIf result_index is larger than the value of ResultCount then NO_SOLUTION is returned.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.RawStatusString","page":"Models","title":"MathOptInterface.RawStatusString","text":"RawStatusString()\n\nA model attribute for a solver specific string explaining why the optimizer stopped.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ResultCount","page":"Models","title":"MathOptInterface.ResultCount","text":"ResultCount()\n\nA model attribute for the number of results available.\n\nOrder of solutions\n\nA number of attributes contain an index, result_index, which is used to refer to one of the available results. Thus, result_index must be an integer between 1 and the number of available results.\n\nAs a general rule, the first result (result_index=1) is the most important result (e.g., an optimal solution or an infeasibility certificate). Other results will typically be alternate solutions that the solver found during the search for the first result.\n\nIf a (local) optimal solution is available, i.e., TerminationStatus is OPTIMAL or LOCALLY_SOLVED, the first result must correspond to the (locally) optimal solution. Other results may be alternative optimal solutions, or they may be other suboptimal solutions; use ObjectiveValue to distingiush between them.\n\nIf a primal or dual infeasibility certificate is available, i.e., TerminationStatus is INFEASIBLE or DUAL_INFEASIBLE and the corresponding PrimalStatus or DualStatus is INFEASIBILITY_CERTIFICATE, then the first result must be a certificate. Other results may be alternate certificates, or infeasible points.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ObjectiveValue","page":"Models","title":"MathOptInterface.ObjectiveValue","text":"ObjectiveValue(result_index::Int = 1)\n\nA model attribute for the objective value of the primal solution result_index.\n\nIf the solver does not have a primal value for the objective because the result_index is beyond the available solutions (whose number is indicated by the ResultCount attribute), getting this attribute must throw a ResultIndexBoundsError. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check PrimalStatus before accessing the ObjectiveValue attribute.\n\nSee ResultCount for information on how the results are ordered.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.DualObjectiveValue","page":"Models","title":"MathOptInterface.DualObjectiveValue","text":"DualObjectiveValue(result_index::Int = 1)\n\nA model attribute for the value of the objective function of the dual problem for the result_indexth dual result.\n\nIf the solver does not have a dual value for the objective because the result_index is beyond the available solutions (whose number is indicated by the ResultCount attribute), getting this attribute must throw a ResultIndexBoundsError. Otherwise, if the result is unavailable for another reason (for instance, only a primal solution is available), the result is undefined. Users should first check DualStatus before accessing the DualObjectiveValue attribute.\n\nSee ResultCount for information on how the results are ordered.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ObjectiveBound","page":"Models","title":"MathOptInterface.ObjectiveBound","text":"ObjectiveBound()\n\nA model attribute for the best known bound on the optimal objective value.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.RelativeGap","page":"Models","title":"MathOptInterface.RelativeGap","text":"RelativeGap()\n\nA model attribute for the final relative optimality gap.\n\nwarning: Warning\nThe definition of this gap is solver-dependent. However, most solvers implementing this attribute define the relative gap as some variation of fracb-ff, where b is the best bound and f is the best feasible objective value.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.SolveTimeSec","page":"Models","title":"MathOptInterface.SolveTimeSec","text":"SolveTimeSec()\n\nA model attribute for the total elapsed solution time (in seconds) as reported by the optimizer.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.SimplexIterations","page":"Models","title":"MathOptInterface.SimplexIterations","text":"SimplexIterations()\n\nA model attribute for the cumulative number of simplex iterations during the optimization process.\n\nFor a mixed-integer program (MIP), the return value is the total simplex iterations for all nodes.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.BarrierIterations","page":"Models","title":"MathOptInterface.BarrierIterations","text":"BarrierIterations()\n\nA model attribute for the cumulative number of barrier iterations while solving a problem.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.NodeCount","page":"Models","title":"MathOptInterface.NodeCount","text":"NodeCount()\n\nA model attribute for the total number of branch-and-bound nodes explored while solving a mixed-integer program (MIP).\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#ResultStatusCode","page":"Models","title":"ResultStatusCode","text":"","category":"section"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"ResultStatusCode\nNO_SOLUTION\nFEASIBLE_POINT\nNEARLY_FEASIBLE_POINT\nINFEASIBLE_POINT\nINFEASIBILITY_CERTIFICATE\nNEARLY_INFEASIBILITY_CERTIFICATE\nREDUCTION_CERTIFICATE\nNEARLY_REDUCTION_CERTIFICATE\nUNKNOWN_RESULT_STATUS\nOTHER_RESULT_STATUS","category":"page"},{"location":"moi/reference/models/#MathOptInterface.ResultStatusCode","page":"Models","title":"MathOptInterface.ResultStatusCode","text":"ResultStatusCode\n\nAn Enum of possible values for the PrimalStatus and DualStatus attributes.\n\nThe values indicate how to interpret the result vector.\n\nValues\n\nPossible values are:\n\nNO_SOLUTION: the result vector is empty.\nFEASIBLE_POINT: the result vector is a feasible point.\nNEARLY_FEASIBLE_POINT: the result vector is feasible if some constraint tolerances are relaxed.\nINFEASIBLE_POINT: the result vector is an infeasible point.\nINFEASIBILITY_CERTIFICATE: the result vector is an infeasibility certificate. If the PrimalStatus is INFEASIBILITY_CERTIFICATE, then the primal result vector is a certificate of dual infeasibility. If the DualStatus is INFEASIBILITY_CERTIFICATE, then the dual result vector is a proof of primal infeasibility.\nNEARLY_INFEASIBILITY_CERTIFICATE: the result satisfies a relaxed criterion for a certificate of infeasibility.\nREDUCTION_CERTIFICATE: the result vector is an ill-posed certificate; see this article for details. If the PrimalStatus is REDUCTION_CERTIFICATE, then the primal result vector is a proof that the dual problem is ill-posed. If the DualStatus is REDUCTION_CERTIFICATE, then the dual result vector is a proof that the primal is ill-posed.\nNEARLY_REDUCTION_CERTIFICATE: the result satisfies a relaxed criterion for an ill-posed certificate.\nUNKNOWN_RESULT_STATUS: the result vector contains a solution with an unknown interpretation.\nOTHER_RESULT_STATUS: the result vector contains a solution with an interpretation not covered by one of the statuses defined above\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.NO_SOLUTION","page":"Models","title":"MathOptInterface.NO_SOLUTION","text":"NO_SOLUTION::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNO_SOLUTION: the result vector is empty.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.FEASIBLE_POINT","page":"Models","title":"MathOptInterface.FEASIBLE_POINT","text":"FEASIBLE_POINT::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nFEASIBLE_POINT: the result vector is a feasible point.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.NEARLY_FEASIBLE_POINT","page":"Models","title":"MathOptInterface.NEARLY_FEASIBLE_POINT","text":"NEARLY_FEASIBLE_POINT::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNEARLY_FEASIBLE_POINT: the result vector is feasible if some constraint tolerances are relaxed.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.INFEASIBLE_POINT","page":"Models","title":"MathOptInterface.INFEASIBLE_POINT","text":"INFEASIBLE_POINT::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nINFEASIBLE_POINT: the result vector is an infeasible point.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.INFEASIBILITY_CERTIFICATE","page":"Models","title":"MathOptInterface.INFEASIBILITY_CERTIFICATE","text":"INFEASIBILITY_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nINFEASIBILITY_CERTIFICATE: the result vector is an infeasibility certificate. If the PrimalStatus is INFEASIBILITY_CERTIFICATE, then the primal result vector is a certificate of dual infeasibility. If the DualStatus is INFEASIBILITY_CERTIFICATE, then the dual result vector is a proof of primal infeasibility.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.NEARLY_INFEASIBILITY_CERTIFICATE","page":"Models","title":"MathOptInterface.NEARLY_INFEASIBILITY_CERTIFICATE","text":"NEARLY_INFEASIBILITY_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNEARLY_INFEASIBILITY_CERTIFICATE: the result satisfies a relaxed criterion for a certificate of infeasibility.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.REDUCTION_CERTIFICATE","page":"Models","title":"MathOptInterface.REDUCTION_CERTIFICATE","text":"REDUCTION_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nREDUCTION_CERTIFICATE: the result vector is an ill-posed certificate; see this article for details. If the PrimalStatus is REDUCTION_CERTIFICATE, then the primal result vector is a proof that the dual problem is ill-posed. If the DualStatus is REDUCTION_CERTIFICATE, then the dual result vector is a proof that the primal is ill-posed.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.NEARLY_REDUCTION_CERTIFICATE","page":"Models","title":"MathOptInterface.NEARLY_REDUCTION_CERTIFICATE","text":"NEARLY_REDUCTION_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNEARLY_REDUCTION_CERTIFICATE: the result satisfies a relaxed criterion for an ill-posed certificate.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.UNKNOWN_RESULT_STATUS","page":"Models","title":"MathOptInterface.UNKNOWN_RESULT_STATUS","text":"UNKNOWN_RESULT_STATUS::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nUNKNOWN_RESULT_STATUS: the result vector contains a solution with an unknown interpretation.\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.OTHER_RESULT_STATUS","page":"Models","title":"MathOptInterface.OTHER_RESULT_STATUS","text":"OTHER_RESULT_STATUS::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nOTHER_RESULT_STATUS: the result vector contains a solution with an interpretation not covered by one of the statuses defined above\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#Conflict-Status","page":"Models","title":"Conflict Status","text":"","category":"section"},{"location":"moi/reference/models/","page":"Models","title":"Models","text":"compute_conflict!\nConflictStatus\nConstraintConflictStatus\nConflictStatusCode\nConflictParticipationStatusCode\nNOT_IN_CONFLICT\nIN_CONFLICT\nMAYBE_IN_CONFLICT","category":"page"},{"location":"moi/reference/models/#MathOptInterface.compute_conflict!","page":"Models","title":"MathOptInterface.compute_conflict!","text":"compute_conflict!(optimizer::AbstractOptimizer)\n\nComputes a minimal subset of constraints such that the model with the other constraint removed is still infeasible.\n\nSome solvers call a set of conflicting constraints an Irreducible Inconsistent Subsystem (IIS).\n\nSee also ConflictStatus and ConstraintConflictStatus.\n\nNote\n\nIf the model is modified after a call to compute_conflict!, the implementor is not obliged to purge the conflict. Any calls to the above attributes may return values for the original conflict without a warning. Similarly, when modifying the model, the conflict can be discarded.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/models/#MathOptInterface.ConflictStatus","page":"Models","title":"MathOptInterface.ConflictStatus","text":"ConflictStatus()\n\nA model attribute for the ConflictStatusCode explaining why the conflict refiner stopped when computing the conflict.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ConstraintConflictStatus","page":"Models","title":"MathOptInterface.ConstraintConflictStatus","text":"ConstraintConflictStatus()\n\nA constraint attribute indicating whether the constraint participates in the conflict. Its type is ConflictParticipationStatusCode.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ConflictStatusCode","page":"Models","title":"MathOptInterface.ConflictStatusCode","text":"ConflictStatusCode\n\nAn Enum of possible values for the ConflictStatus attribute. This attribute is meant to explain the reason why the conflict finder stopped executing in the most recent call to compute_conflict!.\n\nPossible values are:\n\nCOMPUTE_CONFLICT_NOT_CALLED: the function compute_conflict! has not yet been called\nNO_CONFLICT_EXISTS: there is no conflict because the problem is feasible\nNO_CONFLICT_FOUND: the solver could not find a conflict\nCONFLICT_FOUND: at least one conflict could be found\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.ConflictParticipationStatusCode","page":"Models","title":"MathOptInterface.ConflictParticipationStatusCode","text":"ConflictParticipationStatusCode\n\nAn Enum of possible values for the ConstraintConflictStatus attribute. This attribute is meant to indicate whether a given constraint participates or not in the last computed conflict.\n\nValues\n\nPossible values are:\n\nNOT_IN_CONFLICT: the constraint does not participate in the conflict\nIN_CONFLICT: the constraint participates in the conflict\nMAYBE_IN_CONFLICT: the constraint may participate in the conflict, the solver was not able to prove that the constraint can be excluded from the conflict\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/models/#MathOptInterface.NOT_IN_CONFLICT","page":"Models","title":"MathOptInterface.NOT_IN_CONFLICT","text":"NOT_IN_CONFLICT::ConflictParticipationStatusCode\n\nAn instance of the ConflictParticipationStatusCode enum.\n\nNOT_IN_CONFLICT: the constraint does not participate in the conflict\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.IN_CONFLICT","page":"Models","title":"MathOptInterface.IN_CONFLICT","text":"IN_CONFLICT::ConflictParticipationStatusCode\n\nAn instance of the ConflictParticipationStatusCode enum.\n\nIN_CONFLICT: the constraint participates in the conflict\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/models/#MathOptInterface.MAYBE_IN_CONFLICT","page":"Models","title":"MathOptInterface.MAYBE_IN_CONFLICT","text":"MAYBE_IN_CONFLICT::ConflictParticipationStatusCode\n\nAn instance of the ConflictParticipationStatusCode enum.\n\nMAYBE_IN_CONFLICT: the constraint may participate in the conflict, the solver was not able to prove that the constraint can be excluded from the conflict\n\n\n\n\n\n","category":"constant"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Nonlinear/reference.md\"","category":"page"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#NonlinearAPI","page":"API Reference","title":"Nonlinear Modeling","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"More information can be found in the Nonlinear section of the manual.","category":"page"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear\nNonlinear.Model","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear","page":"API Reference","title":"MathOptInterface.Nonlinear","text":"Nonlinear\n\nwarning: Warning\nThe Nonlinear submodule is experimental. Until this message is removed, breaking changes may be introduced in any minor or patch release of MathOptInterface.\n\n\n\n\n\n","category":"module"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Model","page":"API Reference","title":"MathOptInterface.Nonlinear.Model","text":"Model()\n\nThe core datastructure for representing a nonlinear optimization problem.\n\nIt has the following fields:\n\nobjective::Union{Nothing,Expression} : holds the nonlinear objective function, if one exists, otherwise nothing.\nexpressions::Vector{Expression} : a vector of expressions in the model.\nconstraints::OrderedDict{ConstraintIndex,Constraint} : a map from ConstraintIndex to the corresponding Constraint. An OrderedDict is used instead of a Vector to support constraint deletion.\nparameters::Vector{Float64} : holds the current values of the parameters.\noperators::OperatorRegistry : stores the operators used in the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#nonlinear_api_expressions","page":"API Reference","title":"Expressions","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear.ExpressionIndex\nNonlinear.add_expression","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.ExpressionIndex","page":"API Reference","title":"MathOptInterface.Nonlinear.ExpressionIndex","text":"ExpressionIndex\n\nAn index to a nonlinear expression that is returned by add_expression.\n\nGiven data::Model and ex::ExpressionIndex, use data[ex] to retrieve the corresponding Expression.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.add_expression","page":"API Reference","title":"MathOptInterface.Nonlinear.add_expression","text":"add_expression(model::Model, expr)::ExpressionIndex\n\nParse expr into a Expression and add to model. Returns an ExpressionIndex that can be interpolated into other input expressions.\n\nexpr must be a type that is supported by parse_expression.\n\nExamples\n\nmodel = Model()\nx = MOI.VariableIndex(1)\nex = add_expression(model, :($x^2 + 1))\nset_objective(model, :(sqrt($ex)))\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#nonlinear_api_parameters","page":"API Reference","title":"Parameters","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear.ParameterIndex\nNonlinear.add_parameter","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.ParameterIndex","page":"API Reference","title":"MathOptInterface.Nonlinear.ParameterIndex","text":"ParameterIndex\n\nAn index to a nonlinear parameter that is returned by add_parameter. Given data::Model and p::ParameterIndex, use data[p] to retrieve the current value of the parameter and data[p] = value to set a new value.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.add_parameter","page":"API Reference","title":"MathOptInterface.Nonlinear.add_parameter","text":"add_parameter(model::Model, value::Float64)::ParameterIndex\n\nAdd a new parameter to model with the default value value. Returns a ParameterIndex that can be interpolated into other input expressions and used to modify the value of the parameter.\n\nExamples\n\nmodel = Model()\nx = MOI.VariableIndex(1)\np = add_parameter(model, 1.2)\nc = add_constraint(model, :($x^2 - $p), MOI.LessThan(0.0))\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#nonlinear_api_objectives","page":"API Reference","title":"Objectives","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear.set_objective","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.set_objective","page":"API Reference","title":"MathOptInterface.Nonlinear.set_objective","text":"set_objective(model::Model, obj)::Nothing\n\nParse obj into a Expression and set as the objective function of model.\n\nobj must be a type that is supported by parse_expression.\n\nTo remove the objective, pass nothing.\n\nExamples\n\nmodel = Model()\nx = MOI.VariableIndex(1)\nset_objective(model, :($x^2 + 1))\nset_objective(model, x)\nset_objective(model, nothing)\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#nonlinear_api_constraints","page":"API Reference","title":"Constraints","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear.ConstraintIndex\nNonlinear.add_constraint\nNonlinear.delete","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.ConstraintIndex","page":"API Reference","title":"MathOptInterface.Nonlinear.ConstraintIndex","text":"ConstraintIndex\n\nAn index to a nonlinear constraint that is returned by add_constraint.\n\nGiven data::Model and c::ConstraintIndex, use data[c] to retrieve the corresponding Constraint.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.add_constraint","page":"API Reference","title":"MathOptInterface.Nonlinear.add_constraint","text":"add_constraint(\n    model::Model,\n    func,\n    set::Union{\n        MOI.GreaterThan{Float64},\n        MOI.LessThan{Float64},\n        MOI.Interval{Float64},\n        MOI.EqualTo{Float64},\n    },\n)\n\nParse func and set into a Constraint and add to model. Returns a ConstraintIndex that can be used to delete the constraint or query solution information.\n\nExamples\n\nmodel = Model()\nx = MOI.VariableIndex(1)\nc = add_constraint(model, :($x^2), MOI.LessThan(1.0))\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.delete","page":"API Reference","title":"MathOptInterface.Nonlinear.delete","text":"delete(model::Model, c::ConstraintIndex)::Nothing\n\nDelete the constraint index c from model.\n\nExamples\n\nmodel = Model()\nx = MOI.VariableIndex(1)\nc = add_constraint(model, :($x^2), MOI.LessThan(1.0))\ndelete(model, c)\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#nonlinear_api_operators","page":"API Reference","title":"User-defined operators","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear.OperatorRegistry\nNonlinear.DEFAULT_UNIVARIATE_OPERATORS\nNonlinear.DEFAULT_MULTIVARIATE_OPERATORS\nNonlinear.register_operator\nNonlinear.register_operator_if_needed\nNonlinear.assert_registered\nNonlinear.check_return_type\nNonlinear.eval_univariate_function\nNonlinear.eval_univariate_gradient\nNonlinear.eval_univariate_hessian\nNonlinear.eval_multivariate_function\nNonlinear.eval_multivariate_gradient\nNonlinear.eval_multivariate_hessian\nNonlinear.eval_logic_function\nNonlinear.eval_comparison_function","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.OperatorRegistry","page":"API Reference","title":"MathOptInterface.Nonlinear.OperatorRegistry","text":"OperatorRegistry()\n\nCreate a new OperatorRegistry to store and evaluate univariate and multivariate operators.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS","page":"API Reference","title":"MathOptInterface.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS","text":"DEFAULT_UNIVARIATE_OPERATORS\n\nThe list of univariate operators that are supported by default.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS\n72-element Vector{Symbol}:\n :+\n :-\n :abs\n :sqrt\n :cbrt\n :abs2\n :inv\n :log\n :log10\n :log2\n ⋮\n :airybi\n :airyaiprime\n :airybiprime\n :besselj0\n :besselj1\n :bessely0\n :bessely1\n :erfcx\n :dawson\n\n\n\n\n\n","category":"constant"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS","page":"API Reference","title":"MathOptInterface.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS","text":"DEFAULT_MULTIVARIATE_OPERATORS\n\nThe list of multivariate operators that are supported by default.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS\n9-element Vector{Symbol}:\n :+\n :-\n :*\n :^\n :/\n :ifelse\n :atan\n :min\n :max\n\n\n\n\n\n","category":"constant"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.register_operator","page":"API Reference","title":"MathOptInterface.Nonlinear.register_operator","text":"register_operator(\n    model::Model,\n    op::Symbol,\n    nargs::Int,\n    f::Function,\n    [∇f::Function],\n    [∇²f::Function],\n)\n\nRegister the user-defined operator op with nargs input arguments in model.\n\nUnivariate functions\n\nf(x::T)::T must be a function that takes a single input argument x and returns the function evaluated at x. If ∇f and ∇²f are not provided, f must support any Real input type T.\n∇f(x::T)::T is a function that takes a single input argument x and returns the first derivative of f with respect to x. If ∇²f is not provided, ∇f must support any Real input type T.\n∇²f(x::T)::T is a function that takes a single input argument x and returns the second derivative of f with respect to x.\n\nMultivariate functions\n\nf(x::T...)::T must be a function that takes a nargs input arguments x and returns the function evaluated at x. If ∇f and ∇²f are not provided, f must support any Real input type T.\n∇f(g::AbstractVector{T}, x::T...)::T is a function that takes a cache vector g of length length(x), and fills each element g[i] with the partial derivative of f with respect to x[i].\n∇²f(H::AbstractMatrix, x::T...)::T is a function that takes a matrix H and fills the lower-triangular components H[i, j] with the Hessian of f with respect to x[i] and x[j] for i >= j.\n\nNotes for multivariate Hessians\n\nH has size(H) == (length(x), length(x)), but you must not access elements H[i, j] for i > j.\nH is dense, but you do not need to fill structural zeros.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.register_operator_if_needed","page":"API Reference","title":"MathOptInterface.Nonlinear.register_operator_if_needed","text":"register_operator_if_needed(\n    registry::OperatorRegistry,\n    op::Symbol,\n    nargs::Int,\n    f::Function;\n)\n\nSimilar to register_operator, but this function warns if the function is not registered, and skips silently if it already is.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.assert_registered","page":"API Reference","title":"MathOptInterface.Nonlinear.assert_registered","text":"assert_registered(registry::OperatorRegistry, op::Symbol, nargs::Int)\n\nThrow an error if op is not registered in registry with nargs arguments.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.check_return_type","page":"API Reference","title":"MathOptInterface.Nonlinear.check_return_type","text":"check_return_type(::Type{T}, ret::S) where {T,S}\n\nOverload this method for new types S to throw an informative error if a user-defined function returns the type S instead of T.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_univariate_function","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_univariate_function","text":"eval_univariate_function(\n    registry::OperatorRegistry,\n    op::Symbol,\n    x::T,\n) where {T}\n\nEvaluate the operator op(x)::T, where op is a univariate function in registry.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_univariate_gradient","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_univariate_gradient","text":"eval_univariate_gradient(\n    registry::OperatorRegistry,\n    op::Symbol,\n    x::T,\n) where {T}\n\nEvaluate the first-derivative of the operator op(x)::T, where op is a univariate function in registry.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_univariate_hessian","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_univariate_hessian","text":"eval_univariate_hessian(\n    registry::OperatorRegistry,\n    op::Symbol,\n    x::T,\n) where {T}\n\nEvaluate the second-derivative of the operator op(x)::T, where op is a univariate function in registry.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_multivariate_function","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_multivariate_function","text":"eval_multivariate_function(\n    registry::OperatorRegistry,\n    op::Symbol,\n    x::AbstractVector{T},\n) where {T}\n\nEvaluate the operator op(x)::T, where op is a multivariate function in registry.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_multivariate_gradient","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_multivariate_gradient","text":"eval_multivariate_gradient(\n    registry::OperatorRegistry,\n    op::Symbol,\n    g::AbstractVector{T},\n    x::AbstractVector{T},\n) where {T}\n\nEvaluate the gradient of operator g .= ∇op(x), where op is a multivariate function in registry.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_multivariate_hessian","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_multivariate_hessian","text":"eval_multivariate_hessian(\n    registry::OperatorRegistry,\n    op::Symbol,\n    H::AbstractMatrix,\n    x::AbstractVector{T},\n) where {T}\n\nEvaluate the Hessian of operator ∇²op(x), where op is a multivariate function in registry.\n\nThe Hessian is stored in the lower-triangular part of the matrix H.\n\nnote: Note\nImplementations of the Hessian operators will not fill structural zeros. Therefore, before calling this function you should pre-populate the matrix H with 0.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_logic_function","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_logic_function","text":"eval_logic_function(\n    registry::OperatorRegistry,\n    op::Symbol,\n    lhs::T,\n    rhs::T,\n)::Bool where {T}\n\nEvaluate (lhs op rhs)::Bool, where op is a logic operator in registry.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.eval_comparison_function","page":"API Reference","title":"MathOptInterface.Nonlinear.eval_comparison_function","text":"eval_comparison_function(\n    registry::OperatorRegistry,\n    op::Symbol,\n    lhs::T,\n    rhs::T,\n)::Bool where {T}\n\nEvaluate (lhs op rhs)::Bool, where op is a comparison operator in registry.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#Automatic-differentiation-backends","page":"API Reference","title":"Automatic-differentiation backends","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear.Evaluator\nNonlinear.AbstractAutomaticDifferentiation\nNonlinear.ExprGraphOnly\nNonlinear.SparseReverseMode","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Evaluator","page":"API Reference","title":"MathOptInterface.Nonlinear.Evaluator","text":"Evaluator(\n    model::Model,\n    backend::AbstractAutomaticDifferentiation,\n    ordered_variables::Vector{MOI.VariableIndex},\n)\n\nCreate Evaluator, a subtype of MOI.AbstractNLPEvaluator, from Model.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation","page":"API Reference","title":"MathOptInterface.Nonlinear.AbstractAutomaticDifferentiation","text":"AbstractAutomaticDifferentiation\n\nAn abstract type for extending Evaluator.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.ExprGraphOnly","page":"API Reference","title":"MathOptInterface.Nonlinear.ExprGraphOnly","text":"ExprGraphOnly() <: AbstractAutomaticDifferentiation\n\nThe default implementation of AbstractAutomaticDifferentiation. The only supported feature is :ExprGraph.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.SparseReverseMode","page":"API Reference","title":"MathOptInterface.Nonlinear.SparseReverseMode","text":"SparseReverseMode() <: AbstractAutomaticDifferentiation\n\nAn implementation of AbstractAutomaticDifferentiation that uses sparse reverse-mode automatic differentiation to compute derivatives. Supports all features in the MOI nonlinear interface.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#Data-structure","page":"API Reference","title":"Data-structure","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/reference/","page":"API Reference","title":"API Reference","text":"Nonlinear.Node\nNonlinear.NodeType\nNonlinear.Expression\nNonlinear.Constraint\nNonlinear.adjacency_matrix\nNonlinear.parse_expression\nNonlinear.convert_to_expr\nNonlinear.ordinal_index","category":"page"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Node","page":"API Reference","title":"MathOptInterface.Nonlinear.Node","text":"struct Node\n    type::NodeType\n    index::Int\n    parent::Int\nend\n\nA single node in a nonlinear expression tree. Used by Expression.\n\nSee the MathOptInterface documentation for information on how the nodes and values form an expression tree.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.NodeType","page":"API Reference","title":"MathOptInterface.Nonlinear.NodeType","text":"NodeType\n\nAn enum describing the possible node types. Each Node has a .index field, which should be interpreted as follows:\n\nNODE_CALL_MULTIVARIATE: the index into operators.multivariate_operators\nNODE_CALL_UNIVARIATE: the index into operators.univariate_operators\nNODE_LOGIC: the index into operators.logic_operators\nNODE_COMPARISON: the index into operators.comparison_operators\nNODE_MOI_VARIABLE: the value of MOI.VariableIndex(index) in the user's space of the model.\nNODE_VARIABLE: the 1-based index of the internal vector\nNODE_VALUE: the index into the .values field of Expression\nNODE_PARAMETER: the index into data.parameters\nNODE_SUBEXPRESSION:  the index into data.expressions\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Expression","page":"API Reference","title":"MathOptInterface.Nonlinear.Expression","text":"struct Expression\n    nodes::Vector{Node}\n    values::Vector{Float64}\nend\n\nThe core type that represents a nonlinear expression. See the MathOptInterface documentation for information on how the nodes and values form an expression tree.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Constraint","page":"API Reference","title":"MathOptInterface.Nonlinear.Constraint","text":"struct Constraint\n    expression::Expression\n    set::Union{\n        MOI.LessThan{Float64},\n        MOI.GreaterThan{Float64},\n        MOI.EqualTo{Float64},\n        MOI.Interval{Float64},\n    }\nend\n\nA type to hold information relating to the nonlinear constraint f(x) in S, where f(x) is defined by .expression, and S is .set.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.adjacency_matrix","page":"API Reference","title":"MathOptInterface.Nonlinear.adjacency_matrix","text":"adjacency_matrix(nodes::Vector{Node})\n\nCompute the sparse adjacency matrix describing the parent-child relationships in nodes.\n\nThe element (i, j) is true if there is an edge from node[j] to node[i]. Since we get a column-oriented matrix, this gives us a fast way to look up the edges leaving any node (i.e., the children).\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.parse_expression","page":"API Reference","title":"MathOptInterface.Nonlinear.parse_expression","text":"parse_expression(data::Model, input)::Expression\n\nParse input into a Expression.\n\n\n\n\n\nparse_expression(\n    data::Model,\n    expr::Expression,\n    input::Any,\n    parent_index::Int,\n)::Expression\n\nParse input into a Expression, and add it to expr as a child of expr.nodes[parent_index]. Existing subexpressions and parameters are stored in data.\n\nYou can extend parsing support to new types of objects by overloading this method with a different type on input::Any.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.convert_to_expr","page":"API Reference","title":"MathOptInterface.Nonlinear.convert_to_expr","text":"convert_to_expr(data::Model, expr::Expression)\n\nConvert the Expression expr into a Julia Expr.\n\nsubexpressions are represented by a ExpressionIndex object.\nparameters are represented by a ParameterIndex object.\nvariables are represennted by an MOI.VariableIndex object.\n\n\n\n\n\nconvert_to_expr(\n    evaluator::Evaluator,\n    expr::Expression;\n    moi_output_format::Bool,\n)\n\nConvert the Expression expr into a Julia Expr.\n\nIf moi_output_format = true:\n\nsubexpressions will be converted to Julia Expr and substituted into the output expression.\nthe current value of each parameter will be interpolated into the expression\nvariables will be represented in the form x[MOI.VariableIndex(i)]\n\nIf moi_output_format = false:\n\nsubexpressions will be represented by a ExpressionIndex object.\nparameters will be represented by a ParameterIndex object.\nvariables will be represennted by an MOI.VariableIndex object.\n\nwarning: Warning\nTo use moi_output_format = true, you must have first called MOI.initialize with :ExprGraph as a requested feature.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.ordinal_index","page":"API Reference","title":"MathOptInterface.Nonlinear.ordinal_index","text":"ordinal_index(evaluator::Evaluator, c::ConstraintIndex)::Int\n\nReturn the 1-indexed value of the constraint index c in evaluator.\n\nExamples\n\nmodel = Model()\nx = MOI.VariableIndex(1)\nc1 = add_constraint(model, :($x^2), MOI.LessThan(1.0))\nc2 = add_constraint(model, :($x^2), MOI.LessThan(1.0))\nevaluator = Evaluator(model)\nMOI.initialize(evaluator, Symbol[])\nordinal_index(evaluator, c2)  # Returns 2\ndelete(model, c1)\nevaluator = Evaluator(model)\nMOI.initialize(evaluator, Symbol[])\nordinal_index(model, c2)  # Returns 1\n\n\n\n\n\n","category":"function"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"EditURL = \"https://github.com/lanl-ansi/Juniper.jl/blob/62532341586d447f19c7360715333ba62a42bea9/README.md\"","category":"page"},{"location":"packages/Juniper/#Juniper","page":"lanl-ansi/Juniper.jl","title":"Juniper","text":"","category":"section"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"(Image: CI) (Image: codecov) (Image: Documentation)","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"Juniper (Jump Nonlinear Integer Program solver) is a solver for mixed-integer nonlinear programs. ","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"It is a heuristic which is not guaranteed to find the global optimum. If you need the global optimum, check out Alpine.","category":"page"},{"location":"packages/Juniper/#Installation","page":"lanl-ansi/Juniper.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"Install Juniper using the Julia package manager:","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"import Pkg\nPkg.add(\"JuMP\")","category":"page"},{"location":"packages/Juniper/#Use-with-JuMP","page":"lanl-ansi/Juniper.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"Use Juniper with JuMP as follows:","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"using JuMP, Juniper, Ipopt\nipopt = optimizer_with_attributes(Ipopt.Optimizer, \"print_level\"=>0)\noptimizer = optimizer_with_attributes(Juniper.Optimizer, \"nl_solver\"=>ipopt)\nmodel = Model(optimizer)\nv = [10, 20, 12, 23, 42]\nw = [12, 45, 12, 22, 21]\n@variable(model, x[1:5], Bin)\n@objective(model, Max, v' * x)\n@constraint(model, sum(w[i]*x[i]^2 for i in 1:5) <= 45)\noptimize!(model)\nprintln(termination_status(model))\nprintln(objective_value(model))\nprintln(value.(x))","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"The nl_solver is used by Juniper to solve continuous nonlinear sub-problems while Juniper searches for acceptable assignments to the discrete variables. A common choice is Ipopt, but any optimizer that supports the continuous relaxation of the model may be used.","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"To solve problems with more complex nonlinear functions, use the @NLconstraint and @NLobjective JuMP macros.","category":"page"},{"location":"packages/Juniper/#Documentation","page":"lanl-ansi/Juniper.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"The online documentation is available at https://lanl-ansi.github.io/Juniper.jl/stable/.","category":"page"},{"location":"packages/Juniper/#Feasibility-pump","page":"lanl-ansi/Juniper.jl","title":"Feasibility pump","text":"","category":"section"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"If Juniper has difficulty finding feasible solutions on your model, try adding a solver that supports integer variables (for example, HiGHS) to run a feasibility pump:","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"using JuMP, Juniper, Ipopt, HiGHS\nipopt = optimizer_with_attributes(Ipopt.Optimizer, \"print_level\" => 0)\nhighs = optimizer_with_attributes(HiGHS.Optimizer, \"output_flag\" => false)\nmodel = Model(\n    optimizer_with_attributes(\n        Juniper.Optimizer,\n        \"nl_solver\" => ipopt,\n        \"mip_solver\" => highs,\n    ),\n)","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"The feasibility pump is used at the start of Juniper to find a feasible solution before the branch and bound part starts.  For some classes of problems this can be a highly effective pre-processor.","category":"page"},{"location":"packages/Juniper/#Citing-Juniper","page":"lanl-ansi/Juniper.jl","title":"Citing Juniper","text":"","category":"section"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"If you find Juniper useful in your work, we kindly request that you cite the following paper or technical report:","category":"page"},{"location":"packages/Juniper/","page":"lanl-ansi/Juniper.jl","title":"lanl-ansi/Juniper.jl","text":"@inproceedings{juniper,\n     Author = {Ole Kröger and Carleton Coffrin and Hassan Hijazi and Harsha Nagarajan},\n     Title = {Juniper: An Open-Source Nonlinear Branch-and-Bound Solver in Julia},\n     booktitle=\"Integration of Constraint Programming, Artificial Intelligence, and Operations Research\",\n     pages=\"377--386\",\n     year=\"2018\",\n     publisher=\"Springer International Publishing\",\n     isbn=\"978-3-319-93031-2\"\n}","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"EditURL = \"diet.jl\"","category":"page"},{"location":"tutorials/linear/diet/#The-diet-problem","page":"The diet problem","title":"The diet problem","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"The purpose of this tutorial is to demonstrate how to incorporate DataFrames into a JuMP model. As an example, we use classic Stigler diet problem.","category":"page"},{"location":"tutorials/linear/diet/#Required-packages","page":"The diet problem","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"using JuMP\nimport CSV\nimport DataFrames\nimport HiGHS\nimport Test  #hide","category":"page"},{"location":"tutorials/linear/diet/#Formulation","page":"The diet problem","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"We wish to cook a nutritionally balanced meal by choosing the quantity of each food f to eat from a set of foods F in our kitchen.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Each food f has a cost, c_f, as well as a macro-nutrient profile a_mf for each macro-nutrient m in M.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Because we care about a nutritionally balanced meal, we set some minimum and maximum limits for each nutrient, which we denote l_m and u_m respectively.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Furthermore, because we are optimizers, we seek the minimum cost solution.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"With a little effort, we can formulate our dinner problem as the following linear program:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"beginaligned\nmin  sumlimits_f in F c_f x_f \ntextst   l_m le sumlimits_f in F a_mf x_f le u_m  forall m in M \n x_f ge 0  forall f in F\nendaligned","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"In the rest of this tutorial, we will create and solve this problem in JuMP, and learn what we should cook for dinner.","category":"page"},{"location":"tutorials/linear/diet/#Data","page":"The diet problem","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"First, we need some data for the problem. For this tutorial, we'll write CSV files to a temporary directory from Julia. If you have existing files, you could change the filenames to point to them instead.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"dir = mktempdir()","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"The first file is a list of foods with their macro-nutrient profile:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"food_csv_filename = joinpath(dir, \"diet_foods.csv\")\nopen(food_csv_filename, \"w\") do io\n    write(\n        io,\n        \"\"\"\n        name,cost,calories,protein,fat,sodium\n        hamburger,2.49,410,24,26,730\n        chicken,2.89,420,32,10,1190\n        hot dog,1.50,560,20,32,1800\n        fries,1.89,380,4,19,270\n        macaroni,2.09,320,12,10,930\n        pizza,1.99,320,15,12,820\n        salad,2.49,320,31,12,1230\n        milk,0.89,100,8,2.5,125\n        ice cream,1.59,330,8,10,180\n        \"\"\",\n    )\n    return\nend\nfoods = CSV.read(food_csv_filename, DataFrames.DataFrame)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Here, F is foods.name and c_f is foods.cost. (We're also playing a bit loose the term \"macro-nutrient\" by including calories and sodium.)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"We also need our minimum and maximum limits:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"nutrient_csv_filename = joinpath(dir, \"diet_nutrient.csv\")\nopen(nutrient_csv_filename, \"w\") do io\n    write(\n        io,\n        \"\"\"\n        nutrient,min,max\n        calories,1800,2200\n        protein,91,\n        fat,0,65\n        sodium,0,1779\n        \"\"\",\n    )\n    return\nend\nlimits = CSV.read(nutrient_csv_filename, DataFrames.DataFrame)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Protein is missing data for the maximum. Let's fix that using coalesce:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"limits.max = coalesce.(limits.max, Inf)\nlimits","category":"page"},{"location":"tutorials/linear/diet/#JuMP-formulation","page":"The diet problem","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Now we're ready to convert our mathematical formulation into a JuMP model.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"First, create a new JuMP model. Since we have a linear program, we'll use HiGHS as our optimizer:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Next, we create a set of decision variables x, with one element for each row in the DataFrame, and each x has a lower bound of 0:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"@variable(model, x[foods.name] >= 0)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"To simplify things later on, we store the vector as a new column x in the DataFrame foods. Since x is a DenseAxisArray, we first need to convert it to an Array:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"foods.x = Array(x)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Our objective is to minimize the total cost of purchasing food:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"@objective(model, Min, sum(foods.cost .* foods.x));\nnothing #hide","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"For the next component, we need to add a constraint that our total intake of each component is within the limits contained in the limits DataFrame:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"@constraint(\n    model,\n    [row in eachrow(limits)],\n    row.min <= sum(foods[!, row.nutrient] .* foods.x) <= row.max,\n);\nnothing #hide","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"What does our model look like?","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"print(model)","category":"page"},{"location":"tutorials/linear/diet/#Solution","page":"The diet problem","title":"Solution","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"Let's optimize and take a look at the solution:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"optimize!(model)\nTest.@test primal_status(model) == FEASIBLE_POINT        #hide\nTest.@test objective_value(model) ≈ 11.8288 atol = 1e-4  #hide\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"We found an optimal solution. Let's see what the optimal solution is:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"for row in eachrow(foods)\n    println(row.name, \" = \", value(row.x))\nend","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"That's a lot of milk and ice cream, and sadly, we only get 0.6 of a hamburger.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"We can also use the function Containers.rowtable to easily convert the result into a DataFrame:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"table = Containers.rowtable(value, x; header = [:food, :quantity])\nsolution = DataFrames.DataFrame(table)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"This makes it easy to perform analyses our solution:","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"filter!(row -> row.quantity > 0.0, solution)","category":"page"},{"location":"tutorials/linear/diet/#Problem-modification","page":"The diet problem","title":"Problem modification","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"JuMP makes it easy to take an existing model and modify it by adding extra constraints. Let's see what happens if we add a constraint that we can buy at most 6 units of milk or ice cream combined.","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"dairy_foods = [\"milk\", \"ice cream\"]\nis_dairy = map(name -> name in dairy_foods, foods.name)\ndairy_constraint = @constraint(model, sum(foods[is_dairy, :x]) <= 6)\noptimize!(model)\nTest.@test termination_status(model) == INFEASIBLE  #hide\nTest.@test primal_status(model) == NO_SOLUTION      #hide\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"There exists no feasible solution to our problem. Looks like we're stuck eating ice cream for dinner.","category":"page"},{"location":"tutorials/linear/diet/#Next-steps","page":"The diet problem","title":"Next steps","text":"","category":"section"},{"location":"tutorials/linear/diet/","page":"The diet problem","title":"The diet problem","text":"You can delete a constraint using delete(model, dairy_constraint). Can you add a different constraint to provide a diet with less dairy?\nSome food items (like hamburgers) are discrete. You can use set_integer to force a variable to take integer values. What happens to the solution if you do?","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"EditURL = \"n-queens.jl\"","category":"page"},{"location":"tutorials/linear/n-queens/#N-Queens","page":"N-Queens","title":"N-Queens","text":"","category":"section"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"This tutorial was originally contributed by Matthew Helm and Mathieu Tanneau.","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"The N-Queens problem involves placing N queens on an N x N chessboard such that none of the queens attacks another. In chess, a queen can move vertically, horizontally, and diagonally so there cannot be more than one queen on any given row, column, or diagonal.","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"(Image: Four Queens)","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"Note that none of the queens above are able to attack any other as a result of their careful placement.","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"using JuMP\nimport HiGHS\nimport LinearAlgebra","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"N-Queens","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"N = 8\n\nmodel = Model(HiGHS.Optimizer)\nset_silent(model)","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"Next, let's create an N x N chessboard of binary values. 0 will represent an empty space on the board and 1 will represent a space occupied by one of our queens:","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"@variable(model, x[1:N, 1:N], Bin);\nnothing #hide","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"Now we can add our constraints:","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"There must be exactly one queen in a given row/column","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"for i in 1:N\n    @constraint(model, sum(x[i, :]) == 1)\n    @constraint(model, sum(x[:, i]) == 1)\nend","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"There can only be one queen on any given diagonal","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"for i in -(N - 1):(N-1)\n    @constraint(model, sum(LinearAlgebra.diag(x, i)) <= 1)\n    @constraint(model, sum(LinearAlgebra.diag(reverse(x; dims = 1), i)) <= 1)\nend","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"We are ready to put our model to work and see if it is able to find a feasible solution:","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"optimize!(model)","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"We can now review the solution that our model found:","category":"page"},{"location":"tutorials/linear/n-queens/","page":"N-Queens","title":"N-Queens","text":"solution = round.(Int, value.(x))","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"EditURL = \"https://github.com/jump-dev/SDPNAL.jl/blob/00a3fa19f4e1235587948113b0b681da17f4dab5/README.md\"","category":"page"},{"location":"packages/SDPNAL/#SDPNAL.jl","page":"jump-dev/SDPNAL.jl","title":"SDPNAL.jl","text":"","category":"section"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"SDPNAL.jl is wrapper for the SDPNALplus solver.","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"an exported sdpnalplus function that is a thin wrapper on top of the sdpnalplus MATLAB function\nan interface to MathOptInterface","category":"page"},{"location":"packages/SDPNAL/#Affiliation","page":"jump-dev/SDPNAL.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"This wrapper is maintained by the JuMP community and is not an official wrapper of SDPNALplus.","category":"page"},{"location":"packages/SDPNAL/#License","page":"jump-dev/SDPNAL.jl","title":"License","text":"","category":"section"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"SDPNAL.jl is licensed under the MIT License.","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"The underlying solver, SDPNALplus is licensed under the Creative Commons Attribution-ShareAlike 4.0 International Public License.","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"In addition, SDPNAL requires an installation of MATLAB, which is a closed-source commercial product for which you must obtain a license.","category":"page"},{"location":"packages/SDPNAL/#Use-with-JuMP","page":"jump-dev/SDPNAL.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"To use SDPNAL with JuMP, do:","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"using JuMP, SDPNAL\nmodel = Model(SDPNAL.Optimizer)\nset_attribute(model, \"printlevel\", 0)","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"Note that, contrary to implementation of other solver-independent interfaces, using SDPNAL from JuMP or MOI fully exploits the particular structures of the SDPNAL interface and does not create superfluous slack variables and equality constraints as discussed in the SDPNAL guide:","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"A new interface is necessary to facilitate the modeling of an SDP problem for SDPNAL+ because of latter’s flexibility to directly accept inequality constraints of the form “l ≤ B(X) ≤ u”, and bound constraints of the form “L ≤ X ≤ U”. The flexibility can significantly simplify the generation of the data in the SDPNAL+ format as compared to what need to be done in CVX or YALMIP to reformulate them as equality constraints through introducing extra variables. In addition, the final number of equality constraints present in the data input to SDPNAL+ can also be substantially fewer than those present in CVX or YALMIP. It is important to note here that the number of equality constraints present in the generated problem data can greatly affect the computational efficiency of the solvers, especially for interior-point based solvers.","category":"page"},{"location":"packages/SDPNAL/#Installation","page":"jump-dev/SDPNAL.jl","title":"Installation","text":"","category":"section"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"First, make sure that you satisfy the requirements of the MATLAB.jl Julia package, and that the SDPNALplus software is installed in your MATLAB™ installation.","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"Then, install SDPNAL.jl using Pkg.add:","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"import Pkg\nPkg.add(\"SDPNAL\")","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"There is a startup.m file at the root of the SDPNAL folder. This adds all subdirectories recursively when MATLAB starts. However, the interface directory contains a .git subdirectory which contains a very large number of files. Because of this, MATLAB crashes if SDPNAL is in its path because the startup.m requests MATLAB to try to parse all the files in the .git folder. To resolve this problem, delete the startup.m file and .git folder, and add the subdirectories manually your toolbox/local/pathdef.m file as follows:","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"function p = pathdef\n\n% (...)\n\np = [...\n%%% BEGIN ENTRIES %%%\n'/path/to/SDPNALv1.0:', ...\n'/path/to/SDPNALv1.0/interface:', ...\n'/path/to/SDPNALv1.0/mexfun:', ...\n'/path/to/SDPNALv1.0/solver:', ...\n'/path/to/SDPNALv1.0/solver_main_default:', ...\n'/path/to/SDPNALv1.0/util:', ...\n% (...)","category":"page"},{"location":"packages/SDPNAL/","page":"jump-dev/SDPNAL.jl","title":"jump-dev/SDPNAL.jl","text":"If you have SDPT3 in addition to SDPNAL in the MATLAB path (that is, the toolbox/local/pathdef.m file) then you might have issues because both solvers define a validate function, and this might make SDPNAL call SDPT3's validate function instead of SDPT3's validate function.","category":"page"},{"location":"should_i_use/#Should-you-use-JuMP?","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"JuMP is an algebraic modeling language for mathematical optimization written in the Julia language.","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"This page explains when you should consider using JuMP, and importantly, when you should not use JuMP.","category":"page"},{"location":"should_i_use/#When-should-you-use-JuMP?","page":"Should you use JuMP?","title":"When should you use JuMP?","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"You should use JuMP if you have a constrained optimization problem for which you can formulate using the language of mathematical programming, that is:","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"a set of decision variables\na scalar- or vector-valued objective function\na set of constraints.","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Key reasons to use JuMP include:","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"User friendliness\nJuMP has syntax that mimics natural mathematical expressions. (See the section on algebraic modeling languages.)\nSolver independence\nJuMP uses a generic solver-independent interface provided by the MathOptInterface package, making it easy to change between a number of open-source and commercial optimization software packages (\"solvers\"). The Supported solvers section contains a table of the currently supported solvers.\nEase of embedding\nJuMP itself is written purely in Julia. Solvers are the only binary dependencies.\nJuMP provides automatic installation of many open-source solvers. This is different to modeling languages in Python which require you to download and install a solver yourself.\nBecause it is embedded in a general-purpose programming language, JuMP makes it easy to solve optimization problems as part of a larger workflow, for example, inside a simulation, behind a web server, or as a subproblem in a decomposition algorithm. As a trade-off, JuMP's syntax is constrained by the syntax and functionality available in Julia.\nJuMP is MPL licensed, meaning that it can be embedded in commercial software that complies with the terms of the license.\nSpeed\nBenchmarking has shown that JuMP can create problems at similar speeds to special-purpose modeling languages such as AMPL.\nJuMP communicates with most solvers in memory, avoiding the need to write intermediary files.\nAccess to advanced algorithmic techniques\nJuMP supports efficient in-memory re-solves of linear programs, which previously required using solver-specific or low-level C++ libraries.\nJuMP provides access to solver-independent and solver-dependent Callbacks.","category":"page"},{"location":"should_i_use/#When-should-you-not-use-JuMP?","page":"Should you use JuMP?","title":"When should you not use JuMP?","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"JuMP supports a broad range of optimization classes. However, there are still some that it doesn't support, or that are better supported by other software packages.","category":"page"},{"location":"should_i_use/#You-want-to-optimize-a-complicated-Julia-function","page":"Should you use JuMP?","title":"You want to optimize a complicated Julia function","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Packages in Julia compose well. It's common for people to pick two unrelated packages and use them in conjunction to create novel behavior. JuMP isn't one of those packages.","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"If you want to optimize an ordinary differential equation from DifferentialEquations.jl or tune a neural network from Flux.jl, consider using other packages such as:","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Optim.jl\nOptimization.jl\nNLPModels.jl\nNonconvex.jl","category":"page"},{"location":"should_i_use/#Black-box,-derivative-free,-or-unconstrained-optimization","page":"Should you use JuMP?","title":"Black-box, derivative free, or unconstrained optimization","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"JuMP does support nonlinear programs with constraints and objectives containing user-defined operators. However, the functions must be automatically differentiable, or need to provide explicit derivatives. (See User-defined operators for more information.)","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"If your function is a black-box that is non-differentiable (for example, it is the output of a simulation written in C++), JuMP is not the right tool for the job. This also applies if you want to use a derivative free method.","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Even if your problem is differentiable, if it is unconstrained there is limited benefit (and downsides in the form of more overhead) to using JuMP over tools which are only concerned with function minimization.","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Alternatives to consider are:","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Optim.jl\nOptimization.jl\nNLopt.jl","category":"page"},{"location":"should_i_use/#Optimal-control-problems","page":"Should you use JuMP?","title":"Optimal control problems","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"JuMP supports formulating optimal control problems as large nonlinear programs (see, for example, Optimal control for a Space Shuttle reentry trajectory). However, the nonlinear interface has a number of limitations (for example, the need to write out the dynamics in algebraic form) that mean JuMP might not be the right tool for the job.","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Alternatives to consider are:","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"CasADi [MATLAB/Python], CasADi.jl\nInfiniteOpt.jl\npyomo.DAE [Python]","category":"page"},{"location":"should_i_use/#Disciplined-convex-programming","page":"Should you use JuMP?","title":"Disciplined convex programming","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"JuMP does not support disciplined convex programming (DCP).","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Alternatives to consider are:","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"Convex.jl\nCVXPY [Python]\nYALMIP [MATLAB]","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"note: Note\nConvex.jl is also built on MathOptInterface, and shares the same set of underlying solvers. However, you input problems differently, and Convex.jl checks that the problem is DCP.","category":"page"},{"location":"should_i_use/#Stochastic-programming","page":"Should you use JuMP?","title":"Stochastic programming","text":"","category":"section"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"JuMP requires deterministic input data.","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"If you have stochastic input data, consider using a JuMP extension such as:","category":"page"},{"location":"should_i_use/","page":"Should you use JuMP?","title":"Should you use JuMP?","text":"InfiniteOpt.jl\nStochasticPrograms.jl\nSDDP.jl","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"EditURL = \"design_patterns_for_larger_models.jl\"","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Design-patterns-for-larger-models","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"JuMP makes it easy to build and solve optimization models. However, once you start to construct larger models, and especially ones that interact with external data sources or have customizable sets of variables and constraints based on client choices, you may find that your scripts become unwieldy. This tutorial demonstrates a variety of ways in which you can structure larger JuMP models to improve their readability and maintainability.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"tip: Tip\nThis tutorial is more advanced than the other \"Getting started\" tutorials. It's in the \"Getting started\" section to give you an early preview of how JuMP makes it easy to structure larger models. However, if you are new to JuMP you may want to briefly skim the tutorial, and come back to it once you have written a few JuMP models.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Overview","page":"Design patterns for larger models","title":"Overview","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"This tutorial uses explanation-by-example. We're going to start with a simple knapsack model, and then expand it to add various features and structure.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#A-simple-script","page":"Design patterns for larger models","title":"A simple script","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Your first prototype of a JuMP model is probably a script that uses a small set of hard-coded data.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"using JuMP, HiGHS\nprofit = [5, 3, 2, 7, 4]\nweight = [2, 8, 4, 2, 5]\ncapacity = 10\nN = 5\nmodel = Model(HiGHS.Optimizer)\n@variable(model, x[1:N], Bin)\n@objective(model, Max, sum(profit[i] * x[i] for i in 1:N))\n@constraint(model, sum(weight[i] * x[i] for i in 1:N) <= capacity)\noptimize!(model)\nvalue.(x)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"The benefits of this approach are:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"it is quick to code\nit is quick to make changes.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"The downsides include:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"all variables are global (read Performance tips)\nit is easy to introduce errors, for example, having profit and weight be vectors of different lengths, or not match N\nthe solution, x[i], is hard to interpret without knowing the order in which we provided the data.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Wrap-the-model-in-a-function","page":"Design patterns for larger models","title":"Wrap the model in a function","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"A good next step is to wrap your model in a function. This is useful for a few reasons:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"it removes global variables\nit encapsulates the JuMP model and forces you to clarify your inputs and outputs\nwe can add some error checking.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function solve_knapsack_1(profit::Vector, weight::Vector, capacity::Real)\n    if length(profit) != length(weight)\n        throw(DimensionMismatch(\"profit and weight are different sizes\"))\n    end\n    N = length(weight)\n    model = Model(HiGHS.Optimizer)\n    @variable(model, x[1:N], Bin)\n    @objective(model, Max, sum(profit[i] * x[i] for i in 1:N))\n    @constraint(model, sum(weight[i] * x[i] for i in 1:N) <= capacity)\n    optimize!(model)\n    return value.(x)\nend\n\nsolve_knapsack_1([5, 3, 2, 7, 4], [2, 8, 4, 2, 5], 10)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Create-better-data-structures","page":"Design patterns for larger models","title":"Create better data structures","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Although we can check for errors like mis-matched vector lengths, if you start to develop models with a lot of data, keeping track of vectors and lengths and indices is fragile and a common source of bugs. A good solution is to use Julia's type system to create an abstraction over your data.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"For example, we can create a struct that represents a single object, with a constructor that lets us validate assumptions on the input data:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"struct KnapsackObject\n    profit::Float64\n    weight::Float64\n    function KnapsackObject(profit::Float64, weight::Float64)\n        if weight < 0\n            throw(DomainError(\"Weight of object cannot be negative\"))\n        end\n        return new(profit, weight)\n    end\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"as well as a struct that holds a dictionary of objects and the knapsack's capacity:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"struct KnapsackData\n    objects::Dict{String,KnapsackObject}\n    capacity::Float64\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Here's what our data might look like now:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"objects = Dict(\n    \"apple\" => KnapsackObject(5.0, 2.0),\n    \"banana\" => KnapsackObject(3.0, 8.0),\n    \"cherry\" => KnapsackObject(2.0, 4.0),\n    \"date\" => KnapsackObject(7.0, 2.0),\n    \"eggplant\" => KnapsackObject(4.0, 5.0),\n)\ndata = KnapsackData(objects, 10.0)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"If you want, you can add custom printing to make it easier to visualize:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function Base.show(io::IO, data::KnapsackData)\n    println(io, \"A knapsack with capacity $(data.capacity) and possible items:\")\n    for (k, v) in data.objects\n        println(\n            io,\n            \"  $(rpad(k, 8)) : profit = $(v.profit), weight = $(v.weight)\",\n        )\n    end\n    return\nend\n\ndata","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Then, we can re-write our solve_knapsack function to take our KnapsackData as input:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function solve_knapsack_2(data::KnapsackData)\n    model = Model(HiGHS.Optimizer)\n    @variable(model, x[keys(data.objects)], Bin)\n    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))\n    @constraint(\n        model,\n        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,\n    )\n    optimize!(model)\n    return value.(x)\nend\n\nsolve_knapsack_2(data)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Read-in-data-from-files","page":"Design patterns for larger models","title":"Read in data from files","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Having a data structure is a good step. But it is still annoying that we have to hard-code the data into Julia. A good next step is to separate the data into an external file format; JSON is a common choice.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"json_data = \"\"\"\n{\n    \"objects\": {\n        \"apple\": {\"profit\": 5.0, \"weight\": 2.0},\n        \"banana\": {\"profit\": 3.0, \"weight\": 8.0},\n        \"cherry\": {\"profit\": 2.0, \"weight\": 4.0},\n        \"date\": {\"profit\": 7.0, \"weight\": 2.0},\n        \"eggplant\": {\"profit\": 4.0, \"weight\": 5.0}\n    },\n    \"capacity\": 10.0\n}\n\"\"\"\ntemp_dir = mktempdir()\nknapsack_json_filename = joinpath(temp_dir, \"knapsack.json\")\n# Instead of writing a new file here you could replace `knapsack_json_filename`\n# with the path to a local file.\nwrite(knapsack_json_filename, json_data);\nnothing #hide","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Now let's write a function that reads this file and builds a KnapsackData object:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"import JSON\n\nfunction read_data(filename)\n    d = JSON.parsefile(filename)\n    return KnapsackData(\n        Dict(\n            k => KnapsackObject(v[\"profit\"], v[\"weight\"]) for\n            (k, v) in d[\"objects\"]\n        ),\n        d[\"capacity\"],\n    )\nend\n\ndata = read_data(knapsack_json_filename)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Add-options-via-if-else","page":"Design patterns for larger models","title":"Add options via if-else","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"At this point, we have data in a file format which we can load and solve a single problem. For many users, this might be sufficient. However, at some point you may be asked to add features like \"but what if we want to take more than one of a particular item?\"","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"If this is the first time that you've been asked to add a feature, adding options via if-else statements is a good approach. For example, we might write:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function solve_knapsack_3(data::KnapsackData; binary_knapsack::Bool)\n    model = Model(HiGHS.Optimizer)\n    if binary_knapsack\n        @variable(model, x[keys(data.objects)], Bin)\n    else\n        @variable(model, x[keys(data.objects)] >= 0, Int)\n    end\n    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))\n    @constraint(\n        model,\n        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,\n    )\n    optimize!(model)\n    return value.(x)\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Now we can solve the binary knapsack:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"solve_knapsack_3(data; binary_knapsack = true)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"And an integer knapsack where we can take more than one copy of each item:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"solve_knapsack_3(data; binary_knapsack = false)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Add-configuration-options-via-dispatch","page":"Design patterns for larger models","title":"Add configuration options via dispatch","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"If you get repeated requests to add different options, you'll quickly find yourself in a mess of different flags and if-else statements. It's hard to write, hard to read, and hard to ensure you haven't introduced any bugs. A good solution is to use Julia's type dispatch to control the configuration of the model. The easiest way to explain this is by example.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"First, start by defining a new abstract type, as well as new subtypes for each of our options. These types are going to control the configuration of the knapsack model.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"abstract type AbstractConfiguration end\n\nstruct BinaryKnapsackConfig <: AbstractConfiguration end\n\nstruct IntegerKnapsackConfig <: AbstractConfiguration end","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Then, we rewrite our solve_knapsack function to take a config argument, and we introduce an add_knapsack_variables function to abstract the creation of our variables.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function solve_knapsack_4(data::KnapsackData, config::AbstractConfiguration)\n    model = Model(HiGHS.Optimizer)\n    x = add_knapsack_variables(model, data, config)\n    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))\n    @constraint(\n        model,\n        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,\n    )\n    optimize!(model)\n    return value.(x)\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"For the binary knapsack problem, add_knapsack_variables looks like this:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function add_knapsack_variables(\n    model::Model,\n    data::KnapsackData,\n    ::BinaryKnapsackConfig,\n)\n    return @variable(model, x[keys(data.objects)], Bin)\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"For the integer knapsack problem, add_knapsack_variables looks like this:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function add_knapsack_variables(\n    model::Model,\n    data::KnapsackData,\n    ::IntegerKnapsackConfig,\n)\n    return @variable(model, x[keys(data.objects)] >= 0, Int)\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Now we can solve the binary knapsack:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"solve_knapsack_4(data, BinaryKnapsackConfig())","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"and the integer knapsack problem:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"solve_knapsack_4(data, IntegerKnapsackConfig())","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"The main benefit of the dispatch approach is that you can quickly add new options without needing to modify the existing code. For example:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"struct UpperBoundedKnapsackConfig <: AbstractConfiguration\n    limit::Int\nend\n\nfunction add_knapsack_variables(\n    model::Model,\n    data::KnapsackData,\n    config::UpperBoundedKnapsackConfig,\n)\n    return @variable(model, 0 <= x[keys(data.objects)] <= config.limit, Int)\nend\n\nsolve_knapsack_4(data, UpperBoundedKnapsackConfig(3))","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Generalize-constraints-and-objectives","page":"Design patterns for larger models","title":"Generalize constraints and objectives","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"It's easy to extend the dispatch approach to constraints and objectives as well. The key points to notice in the next two functions are that:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"we can access registered variables via model[:x]\nwe can define generic functions which accept any AbstractConfiguration as a configuration argument. That means we can implement a single method and have it apply to multiple configuration types.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function add_knapsack_constraints(\n    model::Model,\n    data::KnapsackData,\n    ::AbstractConfiguration,\n)\n    x = model[:x]\n    @constraint(\n        model,\n        capacity_constraint,\n        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,\n    )\n    return\nend\n\nfunction add_knapsack_objective(\n    model::Model,\n    data::KnapsackData,\n    ::AbstractConfiguration,\n)\n    x = model[:x]\n    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))\n    return\nend\n\nfunction solve_knapsack_5(data::KnapsackData, config::AbstractConfiguration)\n    model = Model(HiGHS.Optimizer)\n    add_knapsack_variables(model, data, config)\n    add_knapsack_constraints(model, data, config)\n    add_knapsack_objective(model, data, config)\n    optimize!(model)\n    return value.(model[:x])\nend\n\nsolve_knapsack_5(data, BinaryKnapsackConfig())","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Remove-solver-dependence,-add-error-checks","page":"Design patterns for larger models","title":"Remove solver dependence, add error checks","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Compared to where we started, our knapsack model is now significantly different. We've wrapped it in a function, defined some data types, and introduced configuration options to control the variables and constraints that get added. There are a few other steps we can do to further improve things:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"remove the dependence on HiGHS\nadd checks that we found an optimal solution\nadd a helper function to avoid the need to explicitly construct the data.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"function solve_knapsack_6(\n    optimizer,\n    data::KnapsackData,\n    config::AbstractConfiguration,\n)\n    model = Model(optimizer)\n    add_knapsack_variables(model, data, config)\n    add_knapsack_constraints(model, data, config)\n    add_knapsack_objective(model, data, config)\n    optimize!(model)\n    if termination_status(model) != OPTIMAL\n        @warn(\"Model not solved to optimality\")\n        return nothing\n    end\n    return value.(model[:x])\nend\n\nfunction solve_knapsack_6(\n    optimizer,\n    data::String,\n    config::AbstractConfiguration,\n)\n    return solve_knapsack_6(optimizer, read_data(data), config)\nend\n\nsolution = solve_knapsack_6(\n    HiGHS.Optimizer,\n    knapsack_json_filename,\n    BinaryKnapsackConfig(),\n)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Create-a-module","page":"Design patterns for larger models","title":"Create a module","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Now we're ready to expose our model to the wider world. That might be as part of a larger Julia project that we're contributing to, or as a stand-alone script that we can run on-demand. In either case, it's good practice to wrap everything in a module. This further encapsulates our code into a single namespace, and we can add documentation in the form of docstrings.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Some good rules to follow when creating a module are:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"use import in a module instead of using to make it clear which functions are from which packages\nuse _ to start function and type names that are considered private\nadd docstrings to all public variables and functions.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"module KnapsackModel\n\nimport JuMP\nimport JSON\n\nstruct _KnapsackObject\n    profit::Float64\n    weight::Float64\n    function _KnapsackObject(profit::Float64, weight::Float64)\n        if weight < 0\n            throw(DomainError(\"Weight of object cannot be negative\"))\n        end\n        return new(profit, weight)\n    end\nend\n\nstruct _KnapsackData\n    objects::Dict{String,_KnapsackObject}\n    capacity::Float64\nend\n\nfunction _read_data(filename)\n    d = JSON.parsefile(filename)\n    return _KnapsackData(\n        Dict(\n            k => _KnapsackObject(v[\"profit\"], v[\"weight\"]) for\n            (k, v) in d[\"objects\"]\n        ),\n        d[\"capacity\"],\n    )\nend\n\nabstract type _AbstractConfiguration end\n\n\"\"\"\n    BinaryKnapsackConfig()\n\nCreate a binary knapsack problem where each object can be taken 0 or 1 times.\n\"\"\"\nstruct BinaryKnapsackConfig <: _AbstractConfiguration end\n\n\"\"\"\n    IntegerKnapsackConfig()\n\nCreate an integer knapsack problem where each object can be taken any number of\ntimes.\n\"\"\"\nstruct IntegerKnapsackConfig <: _AbstractConfiguration end\n\nfunction _add_knapsack_variables(\n    model::JuMP.Model,\n    data::_KnapsackData,\n    ::BinaryKnapsackConfig,\n)\n    return JuMP.@variable(model, x[keys(data.objects)], Bin)\nend\n\nfunction _add_knapsack_variables(\n    model::JuMP.Model,\n    data::_KnapsackData,\n    ::IntegerKnapsackConfig,\n)\n    return JuMP.@variable(model, x[keys(data.objects)] >= 0, Int)\nend\n\nfunction _add_knapsack_constraints(\n    model::JuMP.Model,\n    data::_KnapsackData,\n    ::_AbstractConfiguration,\n)\n    x = model[:x]\n    JuMP.@constraint(\n        model,\n        capacity_constraint,\n        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,\n    )\n    return\nend\n\nfunction _add_knapsack_objective(\n    model::JuMP.Model,\n    data::_KnapsackData,\n    ::_AbstractConfiguration,\n)\n    x = model[:x]\n    JuMP.@objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))\n    return\nend\n\nfunction _solve_knapsack(\n    optimizer,\n    data::_KnapsackData,\n    config::_AbstractConfiguration,\n)\n    model = JuMP.Model(optimizer)\n    _add_knapsack_variables(model, data, config)\n    _add_knapsack_constraints(model, data, config)\n    _add_knapsack_objective(model, data, config)\n    JuMP.optimize!(model)\n    if JuMP.termination_status(model) != JuMP.OPTIMAL\n        @warn(\"Model not solved to optimality\")\n        return nothing\n    end\n    return JuMP.value.(model[:x])\nend\n\n\"\"\"\n    solve_knapsack(\n        optimizer,\n        knapsack_json_filename::String,\n        config::_AbstractConfiguration,\n    )\n\nSolve the knapsack problem and return the optimal primal solution\n\n# Arguments\n\n * `optimizer` : an object that can be passed to `JuMP.Model` to construct a new\n   JuMP model.\n * `knapsack_json_filename` : the filename of a JSON file containing the data for the\n   problem.\n * `config` : an object to control the type of knapsack model constructed.\n   Valid options are:\n    * `BinaryKnapsackConfig()`\n    * `IntegerKnapsackConfig()`\n\n# Returns\n\n * If an optimal solution exists: a `JuMP.DenseAxisArray` that maps the `String`\n   name of each object to the number of objects to pack into the knapsack.\n * Otherwise, `nothing`, indicating that the problem does not have an optimal\n   solution.\n\n# Examples\n\n```julia\nsolution = solve_knapsack(\n    HiGHS.Optimizer,\n    \"path/to/data.json\",\n    BinaryKnapsackConfig(),\n)\n```\n\n```julia\nsolution = solve_knapsack(\n    MOI.OptimizerWithAttributes(HiGHS.Optimizer, \"output_flag\" => false),\n    \"path/to/data.json\",\n    IntegerKnapsackConfig(),\n)\n```\n\"\"\"\nfunction solve_knapsack(\n    optimizer,\n    knapsack_json_filename::String,\n    config::_AbstractConfiguration,\n)\n    data = _read_data(knapsack_json_filename)\n    return _solve_knapsack(optimizer, data, config)\nend\n\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Finally, you can call your model:","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"import .KnapsackModel\n\nKnapsackModel.solve_knapsack(\n    HiGHS.Optimizer,\n    knapsack_json_filename,\n    KnapsackModel.BinaryKnapsackConfig(),\n)","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"note: Note\nThe . in .KnapsackModel denotes that it is a submodule and not a separate package that we installed with Pkg.add. If you put the KnapsackModel in a separate file, load it with:include(\"path/to/KnapsackModel.jl\")\nimport .KnapsackModel","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Add-tests","page":"Design patterns for larger models","title":"Add tests","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"As a final step, you should add tests for your model. This often means testing on a small problem for which you can work out the optimal solution by hand. The Julia standard library Test has good unit-testing functionality.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"import .KnapsackModel\nusing Test\n\n@testset \"KnapsackModel\" begin\n    @testset \"feasible_binary_knapsack\" begin\n        x = KnapsackModel.solve_knapsack(\n            HiGHS.Optimizer,\n            knapsack_json_filename,\n            KnapsackModel.BinaryKnapsackConfig(),\n        )\n        @test isapprox(x[\"apple\"], 1, atol = 1e-5)\n        @test isapprox(x[\"banana\"], 0, atol = 1e-5)\n        @test isapprox(x[\"cherry\"], 0, atol = 1e-5)\n        @test isapprox(x[\"date\"], 1, atol = 1e-5)\n        @test isapprox(x[\"eggplant\"], 1, atol = 1e-5)\n    end\n    @testset \"feasible_integer_knapsack\" begin\n        x = KnapsackModel.solve_knapsack(\n            HiGHS.Optimizer,\n            knapsack_json_filename,\n            KnapsackModel.IntegerKnapsackConfig(),\n        )\n        @test isapprox(x[\"apple\"], 0, atol = 1e-5)\n        @test isapprox(x[\"banana\"], 0, atol = 1e-5)\n        @test isapprox(x[\"cherry\"], 0, atol = 1e-5)\n        @test isapprox(x[\"date\"], 5, atol = 1e-5)\n        @test isapprox(x[\"eggplant\"], 0, atol = 1e-5)\n    end\n    @testset \"infeasible_binary_knapsack\" begin\n        dir = mktempdir()\n        infeasible_filename = joinpath(dir, \"infeasible.json\")\n        write(\n            infeasible_filename,\n            \"\"\"{\n                \"objects\": {\n                    \"apple\": {\"profit\": 5.0, \"weight\": 2.0},\n                    \"banana\": {\"profit\": 3.0, \"weight\": 8.0},\n                    \"cherry\": {\"profit\": 2.0, \"weight\": 4.0},\n                    \"date\": {\"profit\": 7.0, \"weight\": 2.0},\n                    \"eggplant\": {\"profit\": 4.0, \"weight\": 5.0}\n                },\n                \"capacity\": -10.0\n            }\"\"\",\n        )\n        x = KnapsackModel.solve_knapsack(\n            HiGHS.Optimizer,\n            infeasible_filename,\n            KnapsackModel.BinaryKnapsackConfig(),\n        )\n        @test x === nothing\n    end\nend","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"tip: Tip\nPlace these tests in a separate file test_knapsack_model.jl so that you can run the tests by adding include(\"test_knapsack_model.jl\") to any file where needed.","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/#Next-steps","page":"Design patterns for larger models","title":"Next steps","text":"","category":"section"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"We've only briefly scratched the surface of ways to create and structure large JuMP models, so consider this tutorial a starting point, rather than a comprehensive list of all the possible ways to structure JuMP models.  If you are embarking on a large project that uses JuMP, a good next step is to look at ways people have written large JuMP projects \"in the wild.\"","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"Here are some good examples (all co-incidentally related to energy):","category":"page"},{"location":"tutorials/getting_started/design_patterns_for_larger_models/","page":"Design patterns for larger models","title":"Design patterns for larger models","text":"AnyMOD.jl\nJuMP-dev 2021 talk\nsource code\nPowerModels.jl\nJuMP-dev 2021 talk\nsource code\nPowerSimulations.jl\nJuliaCon 2021 talk\nsource code\nUnitCommitment.jl\nJuMP-dev 2021 talk\nsource code","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"EditURL = \"multi.jl\"","category":"page"},{"location":"tutorials/linear/multi/#The-multi-commodity-flow-problem","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"","category":"section"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"This tutorial was originally contributed by Louis Luangkesorn.","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"This tutorial is a JuMP implementation of the multi-commodity transportation model described in AMPL: A Modeling Language for Mathematical Programming, by R. Fourer, D.M. Gay and B.W. Kernighan.","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"The purpose of this tutorial is to demonstrate creating a JuMP model from an SQLite database.","category":"page"},{"location":"tutorials/linear/multi/#Required-packages","page":"The multi-commodity flow problem","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"This tutorial uses the following packages","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"using JuMP\nimport DataFrames\nimport HiGHS\nimport SQLite\nimport Tables\n\nconst DBInterface = SQLite.DBInterface","category":"page"},{"location":"tutorials/linear/multi/#Formulation","page":"The multi-commodity flow problem","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"The multi-commondity flow problem is a simple extension of The transportation problem to multiple types of products. Briefly, we start with the formulation of the transportation problem:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"beginaligned\nmin   sum_i in O j in D c_ij x_ij \nst   sum_j in D x_i j le s_i  forall i in O \n       sum_i in O x_i j = d_j  forall j in D \n       x_i j ge 0  forall i in O j in D\nendaligned","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"but introduce a set of products P, resulting in:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"beginaligned\nmin   sum_i in O j in D k in P c_ijk x_ijk \nst   sum_j in D x_i j k le s_ik  forall i in O k in P \n       sum_i in O x_i j k = d_jk  forall j in D k in P \n       x_i jk ge 0  forall i in O j in D k in P \n       sum_k in P x_i j k le u_ij  forall i in O j in D\nendaligned","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"Note that the last constraint is new; it says that there is a maximum quantity of goods (of any type) that can be transported from origin i to destination j.","category":"page"},{"location":"tutorials/linear/multi/#Data","page":"The multi-commodity flow problem","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"For the purpose of this tutorial, the JuMP repository contains an example database called multi.sqlite.","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"filename = joinpath(@__DIR__, \"multi.sqlite\");\nnothing #hide","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"To run locally, download multi.sqlite and update filename appropriately.","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"Load the database using SQLite.DB:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"db = SQLite.DB(filename)","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"A quick way to see the schema of the database is via SQLite.tables:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"SQLite.tables(db)","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"We interact with the database by executing queries, and then piping the results to an appropriate table. One example is a DataFrame:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"DBInterface.execute(db, \"SELECT * FROM locations\") |> DataFrames.DataFrame","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"But other table types are supported, such as Tables.rowtable:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"DBInterface.execute(db, \"SELECT * FROM locations\") |> Tables.rowtable","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"A rowtable is a Vector of NamedTuples.","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"You can construct more complicated SQL queries:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"origins =\n    DBInterface.execute(\n        db,\n        \"SELECT location FROM locations WHERE type = \\\"origin\\\"\",\n    ) |> Tables.rowtable","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"But for our purpose, we just want the list of strings:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"origins = map(y -> y.location, origins)","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"We can compose these two operations to get a list of destinations:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"destinations =\n    DBInterface.execute(\n        db,\n        \"SELECT location FROM locations WHERE type = \\\"destination\\\"\",\n    ) |>\n    Tables.rowtable |>\n    x -> map(y -> y.location, x)","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"And a list of products from our products table:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"products =\n    DBInterface.execute(db, \"SELECT product FROM products\") |>\n    Tables.rowtable |>\n    x -> map(y -> y.product, x)","category":"page"},{"location":"tutorials/linear/multi/#JuMP-formulation","page":"The multi-commodity flow problem","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"We start by creating a model and our decision variables:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x[origins, destinations, products] >= 0)","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"One approach when working with databases is to extract all of the data into a Julia datastructure. For example, let's pull the cost table into a DataFrame and then construct our objective by iterating over the rows of the DataFrame:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"cost = DBInterface.execute(db, \"SELECT * FROM cost\") |> DataFrames.DataFrame\n@objective(\n    model,\n    Max,\n    sum(r.cost * x[r.origin, r.destination, r.product] for r in eachrow(cost)),\n);\nnothing #hide","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"If we don't want to use a DataFrame, we can use a Tables.rowtable instead:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"supply = DBInterface.execute(db, \"SELECT * FROM supply\") |> Tables.rowtable\nfor r in supply\n    @constraint(model, sum(x[r.origin, :, r.product]) <= r.supply)\nend","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"Another approach is to execute the query, and then to iterate through the rows of the query using Tables.rows:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"demand = DBInterface.execute(db, \"SELECT * FROM demand\")\nfor r in Tables.rows(demand)\n    @constraint(model, sum(x[:, r.destination, r.product]) == r.demand)\nend","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"warning: Warning\nIterating through the rows of a query result works by incrementing a cursor inside the database. As a consequence, you cannot call Tables.rows twice on the same query result.","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"The SQLite queries can be arbitrarily complex. For example, here's a query which builds every possible origin-destination pair:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"od_pairs = DBInterface.execute(\n    db,\n    \"\"\"\n    SELECT a.location as 'origin',\n           b.location as 'destination'\n    FROM locations a\n    INNER JOIN locations b\n    ON a.type = 'origin' AND b.type = 'destination'\n    \"\"\",\n)","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"With a constraint that we cannot send more than 625 units between each pair:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"for r in Tables.rows(od_pairs)\n    @constraint(model, sum(x[r.origin, r.destination, :]) <= 625)\nend","category":"page"},{"location":"tutorials/linear/multi/#Solution","page":"The multi-commodity flow problem","title":"Solution","text":"","category":"section"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"Finally, we can optimize the model:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"and print the solution:","category":"page"},{"location":"tutorials/linear/multi/","page":"The multi-commodity flow problem","title":"The multi-commodity flow problem","text":"begin\n    println(\"         \", join(products, ' '))\n    for o in origins, d in destinations\n        v = lpad.([round(Int, value(x[o, d, p])) for p in products], 5)\n        println(o, \" \", d, \" \", join(replace.(v, \"   0\" => \"  . \"), \" \"))\n    end\nend","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"EditURL = \"finance.jl\"","category":"page"},{"location":"tutorials/linear/finance/#Financial-modeling-problems","page":"Financial modeling problems","title":"Financial modeling problems","text":"","category":"section"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"This tutorial was originally contributed by Arpit Bhatia.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Optimization models play an increasingly important role in financial decisions. Many computational finance problems can be solved efficiently using modern optimization techniques.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"In this tutorial we will discuss two such examples taken from the book Optimization Methods in Finance.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"This tutorial uses the following packages","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/linear/finance/#Short-term-financing","page":"Financial modeling problems","title":"Short-term financing","text":"","category":"section"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Corporations routinely face the problem of financing short term cash commitments such as the following:","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Month Jan Feb Mar Apr May Jun\nNet Cash Flow -150 -100 200 -200 50 300","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Net cash flow requirements are given in thousands of dollars. The company has the following sources of funds:","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"A line of credit of up to $100K at an interest rate of 1% per month,\nIn any one of the first three months, it can issue 90-day commercial paper bearing a total interest of 2% for the 3-month period,\nExcess funds can be invested at an interest rate of 0.3% per month.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Our task is to find out the most economical way to use these 3 sources such that we end up with the most amount of money at the end of June.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"We model this problem in the following manner:","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"We will use the following decision variables:","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"the amount u_i drawn from the line of credit in month i\nthe amount v_i of commercial paper issued in month i\nthe excess funds w_i in month i","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Here we have three types of constraints:","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"for every month, cash inflow = cash outflow for each month\nupper bounds on u_i\nnonnegativity of the decision variables u_i, v_i and w_i.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Our objective will be to simply maximize the company's wealth in June, which say we represent with the variable m.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"financing = Model(HiGHS.Optimizer)\n\n@variables(financing, begin\n    0 <= u[1:5] <= 100\n    0 <= v[1:3]\n    0 <= w[1:5]\n    m\nend)\n\n@objective(financing, Max, m)\n\n@constraints(\n    financing,\n    begin\n        u[1] + v[1] - w[1] == 150 # January\n        u[2] + v[2] - w[2] - 1.01u[1] + 1.003w[1] == 100 # February\n        u[3] + v[3] - w[3] - 1.01u[2] + 1.003w[2] == -200 # March\n        u[4] - w[4] - 1.02v[1] - 1.01u[3] + 1.003w[3] == 200 # April\n        u[5] - w[5] - 1.02v[2] - 1.01u[4] + 1.003w[4] == -50 # May\n        -m - 1.02v[3] - 1.01u[5] + 1.003w[5] == -300 # June\n    end\n)\n\noptimize!(financing)\n\nobjective_value(financing)","category":"page"},{"location":"tutorials/linear/finance/#Combinatorial-auctions","page":"Financial modeling problems","title":"Combinatorial auctions","text":"","category":"section"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"In many auctions, the value that a bidder has for a set of items may not be the sum of the values that he has for individual items.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Examples are equity trading, electricity markets, pollution right auctions and auctions for airport landing slots.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"To take this into account, combinatorial auctions allow the bidders to submit bids on combinations of items.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Let M=12 ldots m be the set of items that the auctioneer has to sell. A bid is a pair B_j=left(S_j p_jright) where S_j subseteq M is a nonempty set of items and p_j is the price offer for this set.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"Suppose that the auctioneer has received n bids B_1 B_2 ldots B_n The goal of this problem is to help an auctioneer determine the winners in order to maximize his revenue.","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"We model this problem by taking a decision variable y_j for every bid. We add a constraint that each item i is sold at most once. This gives us the following model:","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"beginaligned\nmax  sum_i=1^n p_j y_j \ntext  st    sum_j  i in S_j y_j leq 1  forall i=12 ldots m \n y_j in01  forall j in12 ldots n\nendaligned","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"bid_values = [6 3 12 12 8 16]\nbid_items = [[1], [2], [3 4], [1 3], [2 4], [1 3 4]]\n\nauction = Model(HiGHS.Optimizer)\n@variable(auction, y[1:6], Bin)\n@objective(auction, Max, sum(y' .* bid_values))\nfor i in 1:6\n    @constraint(auction, sum(y[j] for j in 1:6 if i in bid_items[j]) <= 1)\nend\n\noptimize!(auction)\n\nobjective_value(auction)","category":"page"},{"location":"tutorials/linear/finance/","page":"Financial modeling problems","title":"Financial modeling problems","text":"value.(y)","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"EditURL = \"https://github.com/GAMS-dev/GAMS.jl/blob/c5dee9f929e9d2f4433ae09fa92b8d872c9c43e0/README.md\"","category":"page"},{"location":"packages/GAMS/#GAMS.jl","page":"GAMS-dev/GAMS.jl","title":"GAMS.jl","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"GAMS.jl provides a MathOptInterface Optimizer to solve JuMP models using GAMS.","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"GAMS comes with dozens of supported solvers. Among them are: ALPHAECP, ANTIGONE, BARON, CBC, CONOPT, CPLEX, DICOPT, GUROBI, IPOPT, KNITRO, LINDO, LINDOGLOBAL, MINOS, MOSEK, NLPEC, PATH, QUADMINOS, SBB, SHOT, SCIP, SNOPT, SOPLEX, XPRESS. Find a complete list here.","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"GAMS.jl supports the following JuMP features:","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"linear, quadratic and nonlinear (convex and non-convex) objective and constraints\ncontinuous, binary, integer, semi-continuous and semi-integer variables\nSOS1 and SOS2 sets\ncomplementarity constraints","category":"page"},{"location":"packages/GAMS/#Installation","page":"GAMS-dev/GAMS.jl","title":"Installation","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"Download GAMS and obtain a GAMS license. Please note that GAMS also offers a free community license.\n(optional) Add the GAMS system directory to the PATH variable in order to find GAMS automatically.\nInstall GAMS.jl using the Julia package manager:\nusing Pkg\nPkg.add(\"GAMS\")","category":"page"},{"location":"packages/GAMS/#Usage","page":"GAMS-dev/GAMS.jl","title":"Usage","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"Using GAMS as optimizer for your JuMP model:","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"using GAMS, JuMP\nmodel = Model(GAMS.Optimizer)","category":"page"},{"location":"packages/GAMS/#GAMS-System","page":"GAMS-dev/GAMS.jl","title":"GAMS System","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"If the GAMS system directory has been added to the PATH variable (you can check this with print(ENV[\"PATH\"])), GAMS.jl will find it automatically. Otherwise, or if you like to switch between systems, the system directory can be specified by (one of the following):","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"set_optimizer_attribute(model, \"SysDir\", \"<gams_system_dir>\")\nset_optimizer_attribute(model, GAMS.SysDir(), \"<gams_system_dir>\")","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"Analogously, you can specify a working directory with \"WorkDir\" or GAMS.WorkDir(). If no working directory has been set, GAMS.jl will create a temporary one.","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"If you want to use the same GAMS workspace (same system and working directory) for multiple models, you can create a GAMSWorkspace first with either of the following","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"ws = GAMS.GAMSWorkspace()\nws = GAMS.GAMSWorkspace(\"<gams_system_dir>\")\nws = GAMS.GAMSWorkspace(\"<gams_system_dir>\", \"<gams_working_dir>\")","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"and then pass it to your models:","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"model = Model(() -> GAMS.Optimizer(ws))","category":"page"},{"location":"packages/GAMS/#GAMS-Options","page":"GAMS-dev/GAMS.jl","title":"GAMS Options","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"GAMS command line options can be specified by","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"set_optimizer_attribute(model, \"<option>\", \"<solver_name>\")\nset_optimizer_attribute(model, GAMS.<option>(), \"<solver_name>\")","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"where <option> is either HoldFixed, IterLim, License, LogOption, NodLim, OptCA, OptCR, ResLim, Solver, Threads, Trace, TraceOpt as well as LP, MIP, RMIP, NLP, DNLP, CNS, MINLP, RMINLP, QCP, MIQCP, RMIQCP, MCP or MPEC. Note that GAMS.ResLim() is equivalent to MOI.TimeLimitSec() and GAMS.Threads() to MOI.NumberOfThreads(). Options LimCol, LimRow, SolPrint and SolveLink cannot be changed and are set to 0, 0, 0 and 5, respectively.","category":"page"},{"location":"packages/GAMS/#Model-Type","page":"GAMS-dev/GAMS.jl","title":"Model Type","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"GAMS.jl will automatically choose a GAMS model type for you. Choosing a different model type:","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"set_optimizer_attribute(model, GAMS.ModelType(), \"<model_type>\")","category":"page"},{"location":"packages/GAMS/#GAMS-Solver-Options","page":"GAMS-dev/GAMS.jl","title":"GAMS Solver Options","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"Specifying GAMS solver options:","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"set_optimizer_attribute(model, \"<solver_option_name>\", <option_value>)","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"Note that passing a solver option is only valid when explicitly choosing a GAMS solver and not using the default.","category":"page"},{"location":"packages/GAMS/#GAMS-Names-vs.-JuMP-Names","page":"GAMS-dev/GAMS.jl","title":"GAMS Names vs. JuMP Names","text":"","category":"section"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"GAMS uses generated variable and constraint names although it is possible to pass the JuMP names to the GAMS optimizer, because GAMS is more restrictive when it comes to variable and constraint naming. Use the attributes GeneratedVariableName, GeneratedConstraintName, OriginalVariableName, OriginalConstraintName to query a GAMS symbol name from a JuMP symbol and vice versa. This can help for debugging, for example in case of GAMS compilation errors. For example:","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"using GAMS\n\nmodel = direct_model(GAMS.Optimizer())\n\n@variable(model, x[1:2,1:3] >= 0)\n@constraint(model, c[i = 1:2], sum(x[i,j] for j = 1:3) <= 10)\n\nMOI.get(model, GAMS.GeneratedVariableName(), x[2,2]) # returns x4\nMOI.get(model, GAMS.OriginalVariableName(\"x6\"))      # returns x[2,3]\nMOI.get(model, GAMS.OriginalVariableName(\"x10\"))     # returns nothing\n\nMOI.get(model, GAMS.GeneratedConstraintName(), c[2]) # returns eq2\nMOI.get(model, GAMS.OriginalConstraintName(\"eq1\"))   # returns c[1]\nMOI.get(model, GAMS.OriginalConstraintName(\"eq10\"))  # returns nothing","category":"page"},{"location":"packages/GAMS/","page":"GAMS-dev/GAMS.jl","title":"GAMS-dev/GAMS.jl","text":"Note that JuMP direct-mode is used.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"EditURL = \"https://github.com/jump-dev/Pavito.jl/blob/v0.3.7/README.md\"","category":"page"},{"location":"packages/Pavito/#Pavito.jl","page":"jump-dev/Pavito.jl","title":"Pavito.jl","text":"","category":"section"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"(Image: Build Status) (Image: Coverage)","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"Pavito.jl is a mixed-integer convex programming (MICP) solver package written in Julia.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"MICP problems are convex, except for restrictions that some variables take binary or integer values.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"Pavito solves MICP problems by constructing sequential polyhedral outer-approximations of the convex feasible set, similar to Bonmin.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"Pavito accesses state-of-the-art MILP solvers and continuous, derivative-based nonlinear programming (NLP) solvers through MathOptInterface.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"For algorithms that use a conic solver instead of an NLP solver, use Pajarito. Pajarito is a robust mixed-integer conic solver that can handle such established problem classes as mixed-integer second-order cone programming (MISOCP) and mixed-integer semidefinite programming (MISDP).","category":"page"},{"location":"packages/Pavito/#License","page":"jump-dev/Pavito.jl","title":"License","text":"","category":"section"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"Pavito.jl is licensed under the MPL 2.0 license.","category":"page"},{"location":"packages/Pavito/#Installation","page":"jump-dev/Pavito.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"Install Pavito using Pkg.add:","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"import Pkg\nPkg.add(\"Pavito\")","category":"page"},{"location":"packages/Pavito/#Use-with-JuMP","page":"jump-dev/Pavito.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"To use Pavito with JuMP, use Pavito.Optimizer:","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"using JuMP, Pavito\nimport GLPK, Ipopt\nmodel = Model(\n    optimizer_with_attributes(\n        Pavito.Optimizer,\n        \"mip_solver\" => optimizer_with_attributes(GLPK.Optimizer),\n        \"cont_solver\" =>\n            optimizer_with_attributes(Ipopt.Optimizer, \"print_level\" => 0),\n    ),\n)","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"The algorithm implemented by Pavito itself is relatively simple; most of the hard work is performed by the MILP solver passed as mip_solver and the NLP solver passed as cont_solver.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"The performance of Pavito depends on these two types of solvers.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"For better performance, you should use a commercial MILP solver such as CPLEX or Gurobi.","category":"page"},{"location":"packages/Pavito/#Options","page":"jump-dev/Pavito.jl","title":"Options","text":"","category":"section"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"The following optimizer attributes can set to a Pavito.Optimizer to modify its behavior:","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"log_level::Int Verbosity flag: 0 for quiet, higher for basic solve info\ntimeout::Float64 Time limit for algorithm (in seconds)\nrel_gap::Float64 Relative optimality gap termination condition\nmip_solver_drives::Bool Let MILP solver manage convergence (\"branch and cut\")\nmip_solver::MOI.OptimizerWithAttributes MILP solver\ncont_solver::MOI.OptimizerWithAttributes Continuous NLP solver","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"Pavito is not yet numerically robust and may require tuning of parameters to improve convergence.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"If the default parameters don't work for you, please let us know by opening an issue.","category":"page"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"For improved Pavito performance, MILP solver integrality tolerance and feasibility tolerances should typically be tightened, for example to 1e-8.","category":"page"},{"location":"packages/Pavito/#Bug-reports-and-support","page":"jump-dev/Pavito.jl","title":"Bug reports and support","text":"","category":"section"},{"location":"packages/Pavito/","page":"jump-dev/Pavito.jl","title":"jump-dev/Pavito.jl","text":"Please report any issues via the GitHub issue tracker. All types of issues are welcome and encouraged; this includes bug reports, documentation typos, feature requests, etc. The Optimization (Mathematical) category on Discourse is appropriate for general discussion.","category":"page"},{"location":"developers/contributing/#How-to-contribute-to-JuMP","page":"Contributing","title":"How to contribute to JuMP","text":"","category":"section"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Welcome, this document explains some ways you can contribute to JuMP.","category":"page"},{"location":"developers/contributing/#Code-of-Conduct","page":"Contributing","title":"Code of Conduct","text":"","category":"section"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"This project and everyone participating in it is governed by the JuMP Code of Conduct. By participating, you are expected to uphold this code.","category":"page"},{"location":"developers/contributing/#Join-the-community-forum","page":"Contributing","title":"Join the community forum","text":"","category":"section"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"First up, join the community forum.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"The forum is a good place to ask questions about how to use JuMP. You can also use the forum to discuss possible feature requests and bugs before raising a GitHub issue (more on this below).","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Aside from asking questions, the easiest way you can contribute to JuMP is to help answer questions on the forum.","category":"page"},{"location":"developers/contributing/#Join-the-developer-chatroom","page":"Contributing","title":"Join the developer chatroom","text":"","category":"section"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"If you're interested in contributing code to JuMP, the next place to join is the developer chatroom. Let us know what you have in mind, and we can point you in the right direction.","category":"page"},{"location":"developers/contributing/#Improve-the-documentation","page":"Contributing","title":"Improve the documentation","text":"","category":"section"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Chances are, if you asked (or answered) a question on the community forum, then it is a sign that the documentation could be improved. Moreover, since it is your question, you are probably the best-placed person to improve it.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"The docs are written in Markdown and are built using Documenter.jl. You can find the source of all the docs here.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"If your change is small (like fixing typos, or one or two sentence corrections), the easiest way to do this is via GitHub's online editor. (GitHub has help on how to do this.)","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"If your change is larger, or touches multiple files, you will need to make the change locally and then use Git to submit a pull request. (See Contribute code to JuMP below for more on this.)","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"tip: Tip\nIf you need any help, come join the developer chatroom and we will walk you through the process.","category":"page"},{"location":"developers/contributing/#File-a-bug-report","page":"Contributing","title":"File a bug report","text":"","category":"section"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Another way to contribute to JuMP is to file bug reports.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Make sure you read the info in the box where you write the body of the issue before posting. You can also find a copy of that info here.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"tip: Tip\nIf you're unsure whether you have a real bug, post on the community forum first. Someone will either help you fix the problem, or let you know the most appropriate place to open a bug report.","category":"page"},{"location":"developers/contributing/#Contribute-code-to-JuMP","page":"Contributing","title":"Contribute code to JuMP","text":"","category":"section"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Finally, you can also contribute code to JuMP.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"warning: Warning\nIf you do not have experience with Git, GitHub, and Julia development, the first steps can be a little daunting. However, there are lots of tutorials available online, including these for:GitHub\nGit and GitHub\nGit\nJulia package developmentIf you need any help, come join the developer chatroom and we will walk you through the process.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Once you are familiar with Git and GitHub, the workflow for contributing code to JuMP is similar to the following:","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 1: decide what to work on","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"The first step is to find an open issue (or open a new one) for the problem you want to solve. Then, before spending too much time on it, discuss what you are planning to do in the issue to see if other contributors are fine with your proposed changes. Getting feedback early can improve code quality, and avoid time spent writing code that does not get merged into JuMP.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"tip: Tip\nAt this point, remember to be patient and polite; you may get a lot of comments on your issue. However, do not be afraid. Comments mean that people are willing to help you improve the code that you are contributing to JuMP.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 2: fork JuMP","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Go to https://github.com/jump-dev/JuMP.jl and click the \"Fork\" button in the top-right corner. This will create a copy of JuMP under your GitHub account.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 3: install JuMP locally","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Open Julia and run:","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"] dev JuMP","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"This will download the JuMP Git repository to ~/.julia/dev/JuMP. If you're on Windows, this will be C:\\\\Users\\\\<my_name>\\\\.julia\\\\dev\\\\JuMP.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"warning: Warning\n] command means \"first type ] to enter the Julia pkg mode, then type the rest. Don't copy-paste the code directly.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 4: checkout a new branch","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"note: Note\nIn the following, replace any instance of GITHUB_ACCOUNT with your GitHub user name.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"The next step is to checkout a development branch. In a terminal (or command prompt on Windows), run:","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"$ cd ~/.julia/dev/JuMP\n\n$ git remote add GITHUB_ACCOUNT https://github.com/GITHUB_ACCOUNT/JuMP.jl.git\n\n$ git checkout master\n\n$ git pull\n\n$ git checkout -b my_new_branch","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"tip: Tip\nLines starting with $ mean \"run these in a terminal (command prompt on Windows).\"","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 5: make changes","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Now make any changes to the source code inside the ~/.julia/dev/JuMP directory.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Make sure you:","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Follow the Style guide and run JuliaFormatter\nAdd tests and documentation for any changes or new features","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"tip: Tip\nWhen you change the source code, you'll need to restart Julia for the changes to take effect. This is a pain, so install Revise.jl.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 6a: test your code changes","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"To test that your changes work, run the JuMP test-suite by opening Julia and running:","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"cd(\"~/.julia/dev/JuMP\")\n] activate .\n] test","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"warning: Warning\nRunning the tests might take a long time (~10–15 minutes).","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"tip: Tip\nIf you're using Revise.jl, you can also run the tests by calling include:include(\"test/runtests.jl\")This can be faster if you want to re-run the tests multiple times.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 6b: test your documentation changes","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Open Julia, then run:","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"cd(\"~/.julia/dev/JuMP/docs\")\n] activate .\ninclude(\"src/make.jl\")","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"warning: Warning\nBuilding the documentation might take a long time (~10 minutes).","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"tip: Tip\nIf there's a problem with the tests that you don't know how to fix, don't worry. Continue to step 5, and one of the JuMP contributors will comment on your pull request telling you how to fix things.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 7: make a pull request","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Once you've made changes, you're ready to push the changes to GitHub. Run:","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"$ cd ~/.julia/dev/JuMP\n\n$ git add .\n\n$ git commit -m \"A descriptive message of the changes\"\n\n$ git push -u GITHUB_ACCOUNT my_new_branch","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Then go to https://github.com/jump-dev/JuMP.jl and follow the instructions that pop up to open a pull request.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 8: respond to comments","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"At this point, remember to be patient and polite; you may get a lot of comments on your pull request. However, do not be afraid. A lot of comments means that people are willing to help you improve the code that you are contributing to JuMP.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"To respond to the comments, go back to step 5, make any changes, test the changes in step 6, and then make a new commit in step 7. Your PR will automatically update.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Step 9: cleaning up","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"Once the PR is merged, clean-up your Git repository ready for the next contribution.","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"$ cd ~/.julia/dev/JuMP\n\n$ git checkout master\n\n$ git pull","category":"page"},{"location":"developers/contributing/","page":"Contributing","title":"Contributing","text":"note: Note\nIf you have suggestions to improve this guide, please make a pull request. It's particularly helpful if you do this after your first pull request because you'll know all the parts that could be explained better.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"EditURL = \"knapsack.jl\"","category":"page"},{"location":"tutorials/linear/knapsack/#The-knapsack-problem-example","page":"The knapsack problem example","title":"The knapsack problem example","text":"","category":"section"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"The purpose of this tutorial is to demonstrate how to formulate and solve a simple optimization problem.","category":"page"},{"location":"tutorials/linear/knapsack/#Required-packages","page":"The knapsack problem example","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/linear/knapsack/#Formulation","page":"The knapsack problem example","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"The knapsack problem is a classical optimization problem: given a set of items and a container with a fixed capacity, choose a subset of items having the greatest combined value that will fit within the container without exceeding the capacity.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"The name of the problem suggests its analogy to packing for a trip, where the baggage weight limit is the capacity and the goal is to pack the most profitable combination of belongings.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"We can formulate the knapsack problem as the integer linear program:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"beginaligned\nmax   sum_i=1^n c_i x_i      \nst   sum_i=1^n w_i x_i le C \n         x_i in 01quad forall i=1ldotsn\nendaligned","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"where C is the capacity, and there is a choice between n items, with item i having weight w_i, profit c_i. Decision variable x_i is equal to 1 if the item is chosen and 0 if not.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"This formulation can be written more compactly as:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"beginaligned\nmax   c^top x       \nst   w^top x le C \n         x text binary \nendaligned","category":"page"},{"location":"tutorials/linear/knapsack/#Data","page":"The knapsack problem example","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"The data for the problem consists of two vectors (one for the profits and one for the weights) along with a capacity.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"There are five objects:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"n = 5;\nnothing #hide","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"For our example, we use a capacity of 10 units:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"capacity = 10.0;\nnothing #hide","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"and the profit and cost data:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"profit = [5.0, 3.0, 2.0, 7.0, 4.0];\nweight = [2.0, 8.0, 4.0, 2.0, 5.0];\nnothing #hide","category":"page"},{"location":"tutorials/linear/knapsack/#JuMP-formulation","page":"The knapsack problem example","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"Let's begin constructing the JuMP model for our knapsack problem.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"First, we'll create a Model object for holding model elements as we construct each part. We'll also set the solver that will ultimately be called to solve the model, once it's constructed.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"model = Model(HiGHS.Optimizer)","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"Next we need the decision variables representing which items are chosen:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"@variable(model, x[1:n], Bin)","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"We now want to constrain those variables so that their combined weight is less than or equal to the given capacity:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"@constraint(model, sum(weight[i] * x[i] for i in 1:n) <= capacity)","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"Finally, our objective is to maximize the combined profit of the chosen items:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"@objective(model, Max, sum(profit[i] * x[i] for i in 1:n))","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"Let's print a human-readable description of the model and check that the model looks as expected:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"print(model)","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"We can now solve the optimization problem and inspect the results.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"The items chosen are","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"items_chosen = [i for i in 1:n if value(x[i]) > 0.5]","category":"page"},{"location":"tutorials/linear/knapsack/#Writing-a-function","page":"The knapsack problem example","title":"Writing a function","text":"","category":"section"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"After working interactively, it is good practice to implement your model in a function.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"The function can be used to ensure that the model is given well-defined input data with validation checks, and that the solution process went as expected.","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"function solve_knapsack_problem(;\n    profit::Vector{Float64},\n    weight::Vector{Float64},\n    capacity::Float64,\n)\n    n = length(weight)\n    # The profit and weight vectors must be of equal length.\n    @assert length(profit) == n\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    @variable(model, x[1:n], Bin)\n    @objective(model, Max, profit' * x)\n    @constraint(model, weight' * x <= capacity)\n    optimize!(model)\n    @assert termination_status(model) == OPTIMAL\n    @assert primal_status(model) == FEASIBLE_POINT\n    println(\"Objective is: \", objective_value(model))\n    println(\"Solution is:\")\n    for i in 1:n\n        print(\"x[$i] = \", round(Int, value(x[i])))\n        println(\", c[$i] / w[$i] = \", profit[i] / weight[i])\n    end\n    chosen_items = [i for i in 1:n if value(x[i]) > 0.5]\n    return return chosen_items\nend\n\nsolve_knapsack_problem(; profit = profit, weight = weight, capacity = capacity)","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"We observe that the chosen items (1, 4, and 5) have the best profit to weight ratio in this particular example.","category":"page"},{"location":"tutorials/linear/knapsack/#Next-steps","page":"The knapsack problem example","title":"Next steps","text":"","category":"section"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"Here are some things to try next:","category":"page"},{"location":"tutorials/linear/knapsack/","page":"The knapsack problem example","title":"The knapsack problem example","text":"Call the function with different data. What happens as the capacity increases?\nWhat happens if the profit and weight vectors are different lengths?\nInstead of creating a binary variable with Bin, we could have written @variable(model, 0 <= x[1:n] <= 1, Int). Verify that this formulation finds the same solution. What happens if we are allowed to take more than one of each item?","category":"page"},{"location":"developers/checklists/#Checklists","page":"Checklists","title":"Checklists","text":"","category":"section"},{"location":"developers/checklists/","page":"Checklists","title":"Checklists","text":"The purpose of this page is to collate a series of checklists for commonly performed changes to the source code of JuMP.","category":"page"},{"location":"developers/checklists/","page":"Checklists","title":"Checklists","text":"In each case, copy the checklist into the description of the pull request.","category":"page"},{"location":"developers/checklists/#Making-a-release","page":"Checklists","title":"Making a release","text":"","category":"section"},{"location":"developers/checklists/","page":"Checklists","title":"Checklists","text":"In preparation for a release, use the following checklist. These steps can be done in the same commit, or separately. The last commit should have the message \"Prep for vX.Y.Z.\"","category":"page"},{"location":"developers/checklists/","page":"Checklists","title":"Checklists","text":"## Pre-release\n\n - [ ] Check that the pinned packages in `docs/Project.toml` are updated. We pin\n       the versions so that changes in the solvers (changes in printing, small\n       numeric changes) do not break the printing of the JuMP docs in arbitrary\n       commits.\n - [ ] Check that the `rev` fields in `docs/packages.toml` are updated. We pin\n       the versions of solvers and extensions to ensure that changes to their\n       READMEs do not break the JuMP docs in arbitrary commits, and to ensure\n       that the versions are compatible with the latest JuMP and\n       MathOptInterface releases.\n - [ ] Update `docs/src/changelog.md`\n - [ ] Run https://github.com/jump-dev/JuMP.jl/actions/workflows/extension-tests.yml\n       using a `workflow_dispatch` trigger to check for any changes in JuMP that\n       broke extensions.\n - [ ] Change the version number in `Project.toml`\n - [ ] Update the links in README.md\n - [ ] The commit messages in this PR do not contain `[ci skip]`\n\n## The release\n\n - [ ] After merging this pull request, comment `[at]JuliaRegistrator register` in\n       the GitHub commit. This should automatically publish a new version to the\n       Julia registry, as well as create a tag, and rebuild the documentation\n       for this tag.\n\n       These steps can take quite a bit of time (1 hour or more), so don't be\n       surprised if the new documentation takes a while to appear. In addition,\n       the links in the README will be broken until JuliaHub fetches the new\n       version on their servers.\n\n## Post-release\n\n - [ ] Once the tag is created, update the relevant `release-` branch. The latest\n       release branch at the time of writing is `release-1.0` (we haven't\n       back-ported any patches that needed to create a `release-1.Y` branch). To\n       to update the release branch with the v1.10.0 tag, do:\n       ```\n       git checkout release-1.0\n       git pull\n       git merge v1.10.0\n       git push\n       ```","category":"page"},{"location":"developers/checklists/#Adding-a-new-solver-to-the-documentation","page":"Checklists","title":"Adding a new solver to the documentation","text":"","category":"section"},{"location":"developers/checklists/","page":"Checklists","title":"Checklists","text":"Use the following checklist when adding a new solver to the JuMP documentation.","category":"page"},{"location":"developers/checklists/","page":"Checklists","title":"Checklists","text":"## Basic\n\n - [ ] Check that the solver is a registered Julia package\n - [ ] Check that the solver supports the long-term support release of Julia\n - [ ] Check that the solver has a MathOptInterface wrapper\n - [ ] Check that the tests call `MOI.Test.runtests`. Some test excludes are\n       permissible, but the reason for skipping a particular test should be\n       documented.\n - [ ] Check that the README and/or documentation provides an example of how to\n       use the solver with JuMP\n\n## Documentation\n\n - [ ] Add a new row to the table in `docs/src/installation.md`\n\n## Optional\n\n - [ ] Add package metadata to `docs/packages.toml`","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"EditURL = \"simple_examples.jl\"","category":"page"},{"location":"tutorials/conic/simple_examples/#Simple-semidefinite-programming-examples","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"","category":"section"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"This tutorial is a collection of examples of small conic programs from the field of semidefinite programming (SDP).","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"This tutorial makes use of the following packages:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"using JuMP\nimport LinearAlgebra\nimport Plots\nimport Random\nimport SCS\nimport Test","category":"page"},{"location":"tutorials/conic/simple_examples/#Maximum-cut-via-SDP","page":"Simple semidefinite programming examples","title":"Maximum cut via SDP","text":"","category":"section"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"The maximum cut problem is a classical example in graph theory, where we seek to partition a graph into two complementary sets, such that the weight of edges between the two sets is maximized. This problem is NP-hard, but it is possible to obtain an approximate solution using the semidefinite programming relaxation:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"    textmax  quad   025 LX \n    text    st quad   mathrmdiag(X) = e \n                 qquad X succeq 0","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"where L is the weighted graph Laplacian and e is a vector of ones. For more details, see:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Goemans, M. X., & Williamson, D. P. (1995). Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite programming. Journal of the ACM (JACM), 42(6), 1115-1145.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"\"\"\"\n    svd_cholesky(X::AbstractMatrix, rtol)\n\nReturn the matrix `U` of the Cholesky decomposition of `X` as `U' * U`.\nNote that we do not use the `LinearAlgebra.cholesky` function because it\nrequires the matrix to be positive definite while `X` may be only\npositive *semi*definite.\n\nWe use the convention `U' * U` instead of `U * U'` to be consistent with\n`LinearAlgebra.cholesky`.\n\"\"\"\nfunction svd_cholesky(X::AbstractMatrix)\n    F = LinearAlgebra.svd(X)\n    # We now have `X ≈ `F.U * D² * F.U'` where:\n    D = LinearAlgebra.Diagonal(sqrt.(F.S))\n    # So `X ≈ U' * U` where `U` is:\n    return (F.U * D)'\nend\n\nfunction solve_max_cut_sdp(weights)\n    N = size(weights, 1)\n    # Calculate the (weighted) Laplacian of the graph: L = D - W.\n    L = LinearAlgebra.diagm(0 => weights * ones(N)) - weights\n    model = Model(SCS.Optimizer)\n    set_silent(model)\n    @variable(model, X[1:N, 1:N], PSD)\n    for i in 1:N\n        set_start_value(X[i, i], 1.0)\n    end\n    @objective(model, Max, 0.25 * LinearAlgebra.dot(L, X))\n    @constraint(model, LinearAlgebra.diag(X) .== 1)\n    optimize!(model)\n    V = svd_cholesky(value(X))\n    Random.seed!(N)\n    r = rand(N)\n    r /= LinearAlgebra.norm(r)\n    cut = [LinearAlgebra.dot(r, V[:, i]) > 0 for i in 1:N]\n    S = findall(cut)\n    T = findall(.!cut)\n    println(\"Solution:\")\n    println(\" (S, T) = ({\", join(S, \", \"), \"}, {\", join(T, \", \"), \"})\")\n    return S, T\nend","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Given the graph","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"[1] --- 5 --- [2]","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"The solution is (S, T)  = ({1}, {2})","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"S, T = solve_max_cut_sdp([0 5; 5 0])","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Given the graph","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"[1] --- 5 --- [2]\n |  \\          |\n |    \\        |\n 7      6      1\n |        \\    |\n |          \\  |\n[3] --- 1 --- [4]","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"The solution is (S, T)  = ({1}, {2, 3, 4})","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"S, T = solve_max_cut_sdp([0 5 7 6; 5 0 0 1; 7 0 0 1; 6 1 1 0])","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Given the graph","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"[1] --- 1 --- [2]\n |             |\n |             |\n 5             9\n |             |\n |             |\n[3] --- 2 --- [4]","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"The solution is (S, T)  = ({1, 4}, {2, 3})","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"S, T = solve_max_cut_sdp([0 1 5 0; 1 0 0 9; 5 0 0 2; 0 9 2 0])","category":"page"},{"location":"tutorials/conic/simple_examples/#K-means-clustering-via-SDP","page":"Simple semidefinite programming examples","title":"K-means clustering via SDP","text":"","category":"section"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Given a set of points a_1 ldots a_m  in mathbbR^n, allocate them to k clusters.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"For more details, see:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Peng, J., & Wei, Y. (2007). Approximating k-means-type clustering via semidefinite programming. SIAM Journal on Optimization, 18(1), 186-205.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"function example_k_means_clustering()\n    a = [[2.0, 2.0], [2.5, 2.1], [7.0, 7.0], [2.2, 2.3], [6.8, 7.0], [7.2, 7.5]]\n    m = length(a)\n    num_clusters = 2\n    W = zeros(m, m)\n    for i in 1:m, j in i+1:m\n        W[i, j] = W[j, i] = exp(-LinearAlgebra.norm(a[i] - a[j]) / 1.0)\n    end\n    model = Model(SCS.Optimizer)\n    set_silent(model)\n    @variable(model, Z[1:m, 1:m] >= 0, PSD)\n    @objective(model, Min, LinearAlgebra.tr(W * (LinearAlgebra.I - Z)))\n    @constraint(model, [i = 1:m], sum(Z[i, :]) .== 1)\n    @constraint(model, LinearAlgebra.tr(Z) == num_clusters)\n    optimize!(model)\n    Z_val = value.(Z)\n    current_cluster, visited = 0, Set{Int}()\n    for i in 1:m\n        if !(i in visited)\n            current_cluster += 1\n            println(\"Cluster $current_cluster\")\n            for j in i:m\n                if isapprox(Z_val[i, i], Z_val[i, j]; atol = 1e-3)\n                    println(a[j])\n                    push!(visited, j)\n                end\n            end\n        end\n    end\n    return\nend\n\nexample_k_means_clustering()","category":"page"},{"location":"tutorials/conic/simple_examples/#The-correlation-problem","page":"Simple semidefinite programming examples","title":"The correlation problem","text":"","category":"section"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Given three random variables A, B, and C, and given bounds on two of the three correlation coefficients:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"    -02 leq ρ_AB leq -01 \n    04  leq ρ_BC leq  05","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"our problem is to determine upper and lower bounds on other correlation coefficient ρ_AC.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"We solve an SDP to make use of the following positive semidefinite property of the correlation matrix:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"beginbmatrix\n      1     ρ_AB   ρ_AC \n     ρ_AB   1      ρ_BC  \n     ρ_AC   ρ_BC   1\nendbmatrix succeq 0","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"function example_correlation_problem()\n    model = Model(SCS.Optimizer)\n    set_silent(model)\n    @variable(model, X[1:3, 1:3], PSD)\n    S = [\"A\", \"B\", \"C\"]\n    ρ = Containers.DenseAxisArray(X, S, S)\n    @constraint(model, [i in S], ρ[i, i] == 1)\n    @constraint(model, -0.2 <= ρ[\"A\", \"B\"] <= -0.1)\n    @constraint(model, 0.4 <= ρ[\"B\", \"C\"] <= 0.5)\n    @objective(model, Max, ρ[\"A\", \"C\"])\n    optimize!(model)\n    println(\"An upper bound for ρ_AC is $(value(ρ[\"A\", \"C\"]))\")\n    @objective(model, Min, ρ[\"A\", \"C\"])\n    optimize!(model)\n    println(\"A lower bound for ρ_AC is $(value(ρ[\"A\", \"C\"]))\")\n    return\nend\n\nexample_correlation_problem()","category":"page"},{"location":"tutorials/conic/simple_examples/#The-minimum-distortion-problem","page":"Simple semidefinite programming examples","title":"The minimum distortion problem","text":"","category":"section"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"This example arises from computational geometry, in particular the problem of embedding a general finite metric space into a Euclidean space.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"It is known that the 4-point metric space defined by the star graph","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"  [1]\n    \\\n     1\n      \\\n      [0] —- 1 -- [2]\n      /\n     1\n    /\n  [3]","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"cannot be exactly embedded into a Euclidean space of any dimension, where distances are computed by length of the shortest path between vertices. A distance-preserving embedding would require the three leaf nodes to form an equilateral triangle of side length 2, with the centre node (0) mapped to an equidistant point at distance 1; this is impossible since the triangle inequality in Euclidean space implies all points would need to be simultaneously collinear.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Here we will formulate and solve an SDP to compute the best possible embedding, that is, the embedding f assigning each vertex v to a vector f(v) that minimizes the distortion c such that","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"    Da b leq f(a) - f(b) leq c  Da b","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"for all edges (a b) in the graph, where Da b is the distance in the graph metric space.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Any embedding f can be characterized by a Gram matrix Q, which is PSD and such that","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"    f(a) - f(b)^2 = Qa a + Qb b - 2 Qa b","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"The matrix entry Qab represents the inner product of f(a) with f(b).","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"We therefore impose the constraint","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"    Da b^2 leq Qa a + Qb b - 2 Qa b leq c^2  Da b^2","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"for all edges (a b) in the graph and minimize c^2, which gives us the SDP formulation below. Since we may choose any point to be the origin, we fix the first vertex at 0.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"For more details, see:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"J. Matoušek (2002), Lectures on discrete geometry, Springer, pp. 378-379","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"N. Linial (2002), Finite metric spaces–combinatorics, geometry and algorithms, Proceedings of the ICM, Vol. 3, 573-586","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"function example_minimum_distortion()\n    model = Model(SCS.Optimizer)\n    set_silent(model)\n    D = [\n        0.0 1.0 1.0 1.0\n        1.0 0.0 2.0 2.0\n        1.0 2.0 0.0 2.0\n        1.0 2.0 2.0 0.0\n    ]\n    @variable(model, c² >= 1.0)\n    @variable(model, Q[1:4, 1:4], PSD)\n    for i in 1:4, j in (i+1):4\n        @constraint(model, D[i, j]^2 <= Q[i, i] + Q[j, j] - 2 * Q[i, j])\n        @constraint(model, Q[i, i] + Q[j, j] - 2 * Q[i, j] <= c² * D[i, j]^2)\n    end\n    fix(Q[1, 1], 0)\n    @objective(model, Min, c²)\n    optimize!(model)\n    Test.@test termination_status(model) == OPTIMAL\n    Test.@test primal_status(model) == FEASIBLE_POINT\n    Test.@test objective_value(model) ≈ 4 / 3 atol = 1e-4\n    # Recover the minimal distorted embedding:\n    X = [zeros(3) sqrt(value.(Q)[2:end, 2:end])]\n    return Plots.plot(\n        X[1, :],\n        X[2, :],\n        X[3, :];\n        seriestype = :mesh3d,\n        connections = ([0, 0, 0, 1], [1, 2, 3, 2], [2, 3, 1, 3]),\n        legend = false,\n        fillalpha = 0.1,\n        lw = 3,\n        ratio = :equal,\n        xlim = (-1.1, 1.1),\n        ylim = (-1.1, 1.1),\n        zlim = (-1.5, 1.0),\n        zticks = -1:1,\n        camera = (60, 30),\n    )\nend\n\nexample_minimum_distortion()","category":"page"},{"location":"tutorials/conic/simple_examples/#Lovász-numbers","page":"Simple semidefinite programming examples","title":"Lovász numbers","text":"","category":"section"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"The Lovász number of a graph, also known as Lovász's theta-function, is a number that lies between two important and related numbers that are computationally hard to determine, namely the chromatic and clique numbers of the graph. It is possible however to efficient compute the Lovász number as the optimal value of a semidefinite program.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Consider the pentagon graph:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"     [5]\n    /   \\\n   /     \\\n [1]     [4]\n  |       |\n  |       |\n [2] --- [3]","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"with five vertices and edges. Its Lovász number is known to be precisely sqrt5 approx 2236, lying between 2 (the largest clique size) and 3 (the smallest number needed for a vertex coloring).","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Let i  j be integers such that 1 leq i  j leq 5. We define A^ij to be the 5 times 5 symmetric matrix with entries (ij) and (ji) equal to 1, with all other entries 0. Let  E be the graph's edge set; in this example, E contains (1,2), (2,3), (3,4), (4,5), (5,1) and their transposes. The Lovász number can be computed from the program","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"beginalign\n    textmax       quad    J  X  \n    text    st  quad A^ij  X = 0 text for all  (ij) notin E   \n                     quad I  X = 1 \n                     quad X succeq 0\nendalign","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"where J is the matrix filled with ones, and I is the identity matrix.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"For more details, see:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Barvinok, A. (2002). A course in convexity, American Mathematical Society (Vol. 54), pp. 182-185.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Knuth, D. E. (1994), The sandwich theorem, Electronic Journal of Combinatorics, Volume 1, Issue 1, A1.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"function example_theta_problem()\n    model = Model(SCS.Optimizer)\n    set_silent(model)\n    E = [(1, 2), (2, 3), (3, 4), (4, 5), (5, 1)]\n    @variable(model, X[1:5, 1:5], PSD)\n    for i in 1:5\n        for j in (i+1):5\n            if !((i, j) in E || (j, i) in E)\n                A = zeros(Int, 5, 5)\n                A[i, j] = 1\n                A[j, i] = 1\n                @constraint(model, LinearAlgebra.dot(A, X) == 0)\n            end\n        end\n    end\n    @constraint(model, LinearAlgebra.tr(LinearAlgebra.I * X) == 1)\n    J = ones(Int, 5, 5)\n    @objective(model, Max, LinearAlgebra.dot(J, X))\n    optimize!(model)\n    Test.@test termination_status(model) == OPTIMAL\n    Test.@test primal_status(model) == FEASIBLE_POINT\n    Test.@test objective_value(model) ≈ sqrt(5) rtol = 1e-4\n    println(\"The Lovász number is: $(objective_value(model))\")\n    return\nend\n\nexample_theta_problem()","category":"page"},{"location":"tutorials/conic/simple_examples/#Robust-uncertainty-sets","page":"Simple semidefinite programming examples","title":"Robust uncertainty sets","text":"","category":"section"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"This example computes the Value at Risk for a data-driven uncertainty set. Closed-form expressions for the optimal value are available. For more details, see:","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"Bertsimas, D., Gupta, V., & Kallus, N. (2018). Data-driven robust optimization. Mathematical Programming, 167, 235-292.","category":"page"},{"location":"tutorials/conic/simple_examples/","page":"Simple semidefinite programming examples","title":"Simple semidefinite programming examples","text":"function example_robust_uncertainty_sets()\n    R, d, 𝛿, ɛ = 1, 3, 0.05, 0.05\n    N = ceil((2 + 2 * log(2 / 𝛿))^2) + 1\n    c, μhat, M = randn(d), rand(d), rand(d, d)\n    Σhat = 1 / (d - 1) * (M - ones(d) * μhat')' * (M - ones(d) * μhat')\n    Γ1(𝛿, N) = R / sqrt(N) * (2 + sqrt(2 * log(1 / 𝛿)))\n    Γ2(𝛿, N) = 2 * R^2 / sqrt(N) * (2 + sqrt(2 * log(2 / 𝛿)))\n    model = Model(SCS.Optimizer)\n    set_silent(model)\n    @variable(model, Σ[1:d, 1:d], PSD)\n    @variable(model, u[1:d])\n    @variable(model, μ[1:d])\n    @constraint(model, [Γ1(𝛿 / 2, N); μ - μhat] in SecondOrderCone())\n    @constraint(model, [Γ2(𝛿 / 2, N); vec(Σ - Σhat)] in SecondOrderCone())\n    @constraint(model, [((1-ɛ)/ɛ) (u - μ)'; (u-μ) Σ] >= 0, PSDCone())\n    @objective(model, Max, c' * u)\n    optimize!(model)\n    exact =\n        μhat' * c +\n        Γ1(𝛿 / 2, N) * LinearAlgebra.norm(c) +\n        sqrt((1 - ɛ) / ɛ) *\n        sqrt(c' * (Σhat + Γ2(𝛿 / 2, N) * LinearAlgebra.I) * c)\n    Test.@test objective_value(model) ≈ exact atol = 1e-2\n    return\nend\n\nexample_robust_uncertainty_sets()","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/callbacks/#callbacks_manual","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"","category":"section"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"Many mixed-integer (linear, conic, and nonlinear) programming solvers offer the ability to modify the solve process. Examples include changing branching decisions in branch-and-bound, adding custom cutting planes, providing custom heuristics to find feasible solutions, or implementing on-demand separators to add new constraints only when they are violated by the current solution (also known as lazy constraints).","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"While historically this functionality has been limited to solver-specific interfaces, JuMP provides solver-independent support for three types of callbacks:","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"lazy constraints\nuser-cuts\nheuristic solutions","category":"page"},{"location":"manual/callbacks/#Available-solvers","page":"Solver-independent Callbacks","title":"Available solvers","text":"","category":"section"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"Solver-independent callback support is limited to a few solvers. This includes CPLEX, GLPK, Gurobi, Xpress, and SCIP.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"warning: Warning\nWhile JuMP provides a solver-independent way of accessing callbacks, you should not assume that you will see identical behavior when running the same code on different solvers. For example, some solvers may ignore user-cuts for various reasons, while other solvers may add every user-cut. Read the underlying solver's callback documentation to understand details specific to each solver.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"tip: Tip\nThis page discusses solver-independent callbacks. However, each solver listed above also provides a solver-dependent callback to provide access to the full range of solver-specific features. Consult the solver's README for an example of how to use the solver-dependent callback. This will require you to understand the C interface of the solver.","category":"page"},{"location":"manual/callbacks/#Things-you-can-and-cannot-do-during-solver-independent-callbacks","page":"Solver-independent Callbacks","title":"Things you can and cannot do during solver-independent callbacks","text":"","category":"section"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"There is a limited range of things you can do during a callback. Only use the functions and macros explicitly stated in this page of the documentation, or in the Callbacks tutorial.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"Using any other part of the JuMP API (for example, adding a constraint with @constraint or modifying a variable bound with set_lower_bound) is undefined behavior, and your solver may throw an error, return an incorrect solution, or result in a segfault that aborts Julia.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"In each of the three solver-independent callbacks, there are two things you may query:","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"callback_node_status returns an MOI.CallbackNodeStatusCode enum indicating if the current primal solution is integer feasible.\ncallback_value returns the current primal solution of a variable.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"If you need to query any other information, use a solver-dependent callback instead. Each solver supporting a solver-dependent callback has information on how to use it in the README of their GitHub repository.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"If you want to modify the problem in a callback, you must use a lazy constraint.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"warning: Warning\nYou can only set each callback once. Calling set twice will over-write the earlier callback. In addition, if you use a solver-independent callback, you cannot set a solver-dependent callback.","category":"page"},{"location":"manual/callbacks/#Lazy-constraints","page":"Solver-independent Callbacks","title":"Lazy constraints","text":"","category":"section"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"Lazy constraints are useful when the full set of constraints is too large to explicitly include in the initial formulation. When a MIP solver reaches a new solution, for example with a heuristic or by solving a problem at a node in the branch-and-bound tree, it will give the user the chance to provide constraints that would make the current solution infeasible. For some more information about lazy constraints, see this blog post by Paul Rubin.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"A lazy constraint callback can be set using the following syntax:","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"julia> import GLPK\n\njulia> model = Model(GLPK.Optimizer);\n\njulia> @variable(model, x <= 10, Int)\nx\n\njulia> @objective(model, Max, x)\nx\n\njulia> function my_callback_function(cb_data)\n           status = callback_node_status(cb_data, model)\n           if status == MOI.CALLBACK_NODE_STATUS_FRACTIONAL\n               # `callback_value(cb_data, x)` is not integer (to some tolerance).\n               # If, for example, your lazy constraint generator requires an\n               # integer-feasible primal solution, you can add a `return` here.\n               return\n           elseif status == MOI.CALLBACK_NODE_STATUS_INTEGER\n               # `callback_value(cb_data, x)` is integer (to some tolerance).\n           else\n               @assert status == MOI.CALLBACK_NODE_STATUS_UNKNOWN\n               # `callback_value(cb_data, x)` might be fractional or integer.\n           end\n           x_val = callback_value(cb_data, x)\n           if x_val > 2 + 1e-6\n               con = @build_constraint(x <= 2)\n               MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n           end\n       end\nmy_callback_function (generic function with 1 method)\n\njulia> set_attribute(model, MOI.LazyConstraintCallback(), my_callback_function)","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"info: Info\nThe lazy constraint callback may be called at fractional or integer nodes in the branch-and-bound tree. There is no guarantee that the callback is called at every primal solution.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"warning: Warning\nOnly add a lazy constraint if your primal solution violates the constraint. Adding the lazy constraint irrespective of feasibility may result in the solver returning an incorrect solution, or lead to many constraints being added, slowing down the solution process.model = Model(GLPK.Optimizer)\n@variable(model, x <= 10, Int)\n@objective(model, Max, x)\nfunction bad_callback_function(cb_data)\n    # Don't do this!\n    con = @build_constraint(x <= 2)\n    MOI.submit(model, MOI.LazyConstraint(cb_data), con)\nend\nfunction good_callback_function(cb_data)\n    if callback_value(x) > 2\n        con = @build_constraint(x <= 2)\n        MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n    end\nend\nset_attribute(model, MOI.LazyConstraintCallback(), good_callback_function)","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"warning: Warning\nDuring the solve, a solver may visit a point that was cut off by a previous lazy constraint, for example, because the earlier lazy constraint was removed during presolve. If this happens, you must re-add the lazy constraint.","category":"page"},{"location":"manual/callbacks/#User-cuts","page":"Solver-independent Callbacks","title":"User cuts","text":"","category":"section"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"User cuts, or simply cuts, provide a way for the user to tighten the LP relaxation using problem-specific knowledge that the solver cannot or is unable to infer from the model. Just like with lazy constraints, when a MIP solver reaches a new node in the branch-and-bound tree, it will give the user the chance to provide cuts to make the current relaxed (fractional) solution infeasible in the hopes of obtaining an integer solution. For more details about the difference between user cuts and lazy constraints see the aforementioned blog post.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"A user-cut callback can be set using the following syntax:","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"julia> import GLPK\n\njulia> model = Model(GLPK.Optimizer);\n\njulia> @variable(model, x <= 10.5, Int)\nx\n\njulia> @objective(model, Max, x)\nx\n\njulia> function my_callback_function(cb_data)\n           x_val = callback_value(cb_data, x)\n           con = @build_constraint(x <= floor(x_val))\n           MOI.submit(model, MOI.UserCut(cb_data), con)\n       end\nmy_callback_function (generic function with 1 method)\n\njulia> set_attribute(model, MOI.UserCutCallback(), my_callback_function)","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"warning: Warning\nUser cuts must not change the set of integer feasible solutions. Equivalently, user cuts can only remove fractional solutions. If you add a cut that removes an integer solution (even one that is not optimal), the solver may return an incorrect solution.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"info: Info\nThe user-cut callback may be called at fractional nodes in the branch-and-bound tree. There is no guarantee that the callback is called at every fractional primal solution.","category":"page"},{"location":"manual/callbacks/#Heuristic-solutions","page":"Solver-independent Callbacks","title":"Heuristic solutions","text":"","category":"section"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"Integer programming solvers frequently include heuristics that run at the nodes of the branch-and-bound tree. They aim to find integer solutions quicker than plain branch-and-bound would to tighten the bound, allowing us to fathom nodes quicker and to tighten the integrality gap.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"Some heuristics take integer solutions and explore their \"local neighborhood\" (for example, flipping binary variables, fix some variables and solve a smaller MILP) and others take fractional solutions and attempt to round them in an intelligent way.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"You may want to add a heuristic of your own if you have some special insight into the problem structure that the solver is not aware of, for example, you can consistently take fractional solutions and intelligently guess integer solutions from them.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"A heuristic solution callback can be set using the following syntax:","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"julia> import GLPK\n\njulia> model = Model(GLPK.Optimizer);\n\njulia> @variable(model, x <= 10.5, Int)\nx\n\njulia> @objective(model, Max, x)\nx\n\njulia> function my_callback_function(cb_data)\n           x_val = callback_value(cb_data, x)\n           status = MOI.submit(\n               model, MOI.HeuristicSolution(cb_data), [x], [floor(Int, x_val)]\n           )\n           println(\"I submitted a heuristic solution, and the status was: \", status)\n       end\nmy_callback_function (generic function with 1 method)\n\njulia> set_attribute(model, MOI.HeuristicCallback(), my_callback_function)","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"The third argument to submit is a vector of JuMP variables, and the fourth argument is a vector of values corresponding to each variable.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"MOI.submit returns an enum that depends on whether the solver accepted the solution. The possible return codes are:","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"MOI.HEURISTIC_SOLUTION_ACCEPTED\nMOI.HEURISTIC_SOLUTION_REJECTED\nMOI.HEURISTIC_SOLUTION_UNKNOWN","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"warning: Warning\nSome solvers may accept partial solutions. Others require a feasible integer solution for every variable. If in doubt, provide a complete solution.","category":"page"},{"location":"manual/callbacks/","page":"Solver-independent Callbacks","title":"Solver-independent Callbacks","text":"info: Info\nThe heuristic solution callback may be called at fractional nodes in the branch-and-bound tree. There is no guarantee that the callback is called at every fractional primal solution.","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/infeasibility_certificates.md\"","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/background/infeasibility_certificates/#Infeasibility-certificates","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"","category":"section"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"When given a conic problem that is infeasible or unbounded, some solvers can produce a certificate of infeasibility. This page explains what a certificate of infeasibility is, and the related conventions that MathOptInterface adopts.","category":"page"},{"location":"moi/background/infeasibility_certificates/#Conic-duality","page":"Infeasibility certificates","title":"Conic duality","text":"","category":"section"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"MathOptInterface uses conic duality to define infeasibility certificates. A full explanation is given in the section Duality, but here is a brief overview.","category":"page"},{"location":"moi/background/infeasibility_certificates/#Minimization-problems","page":"Infeasibility certificates","title":"Minimization problems","text":"","category":"section"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"For a minimization problem in geometric conic form, the primal is:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"beginalign\n min_x in mathbbR^n  a_0^top x + b_0\n\n textst  A_i x + b_i  in mathcalC_i  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"and the dual is a maximization problem in standard conic form:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"beginalign\n max_y_1 ldots y_m  -sum_i=1^m b_i^top y_i + b_0\n\n textst  a_0 - sum_i=1^m A_i^top y_i  = 0\n\n  y_i  in mathcalC_i^*  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"where each mathcalC_i is a closed convex cone and mathcalC_i^* is its dual cone.","category":"page"},{"location":"moi/background/infeasibility_certificates/#Maximization-problems","page":"Infeasibility certificates","title":"Maximization problems","text":"","category":"section"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"For a maximization problem in geometric conic form, the primal is:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"beginalign\n max_x in mathbbR^n  a_0^top x + b_0\n\n textst  A_i x + b_i  in mathcalC_i  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"and the dual is a minimization problem in standard conic form:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"beginalign\n min_y_1 ldots y_m  sum_i=1^m b_i^top y_i + b_0\n\n textst  a_0 + sum_i=1^m A_i^top y_i  = 0\n\n  y_i  in mathcalC_i^*  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/infeasibility_certificates/#Unbounded-problems","page":"Infeasibility certificates","title":"Unbounded problems","text":"","category":"section"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"A problem is unbounded if and only if:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"there exists a feasible primal solution\nthe dual is infeasible.","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"A feasible primal solution—if one exists—can be obtained by setting ObjectiveSense to FEASIBILITY_SENSE before optimizing. Therefore, most solvers stop after they prove the dual is infeasible via a certificate of dual infeasibility, but before they have found a feasible primal solution. This is also the reason that MathOptInterface defines the DUAL_INFEASIBLE status instead of UNBOUNDED.","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"A certificate of dual infeasibility is an improving ray of the primal problem. That is, there exists some vector d such that for all eta  0:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"A_i (x + eta d) + b_i in mathcalC_i  i = 1 ldots m","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"and (for minimization problems):","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"a_0^top (x + eta d) + b_0  a_0^top x + b_0","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"for any feasible point x. The latter simplifies to a_0^top d  0. For maximization problems, the inequality is reversed, so that a_0^top d  0.","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"If the solver has found a certificate of dual infeasibility:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"TerminationStatus must be DUAL_INFEASIBLE\nPrimalStatus must be INFEASIBILITY_CERTIFICATE\nVariablePrimal must be the corresponding value of d\nConstraintPrimal must be the corresponding value of A_i d\nObjectiveValue must be the value a_0^top d. Note that this is the value of the objective function at d, ignoring the constant b_0.","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"note: Note\nThe choice of whether to scale the ray d to have magnitude 1 is left to the solver.","category":"page"},{"location":"moi/background/infeasibility_certificates/#Infeasible-problems","page":"Infeasibility certificates","title":"Infeasible problems","text":"","category":"section"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"A certificate of primal infeasibility is an improving ray of the dual problem. However, because infeasibility is independent of the objective function, we first homogenize the primal problem by removing its objective.","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"For a minimization problem, a dual improving ray is some vector d such that for all eta  0:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"beginalign\n-sum_i=1^m A_i^top (y_i + eta d_i)  = 0 \n(y_i + eta d_i)  in mathcalC_i^*  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"and:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"-sum_i=1^m b_i^top (y_i + eta d_i)  -sum_i=1^m b_i^top y_i","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"for any feasible dual solution y. The latter simplifies to -sum_i=1^m b_i^top d_i  0. For a maximization problem, the inequality is sum_i=1^m b_i^top d_i  0. (Note that these are the same inequality, modulo a - sign.)","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"If the solver has found a certificate of primal infeasibility:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"TerminationStatus must be INFEASIBLE\nDualStatus must be INFEASIBILITY_CERTIFICATE\nConstraintDual must be the corresponding value of d\nDualObjectiveValue must be the value -sum_i=1^m b_i^top d_i for minimization problems and sum_i=1^m b_i^top d_i for maximization problems.","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"note: Note\nThe choice of whether to scale the ray d to have magnitude 1 is left to the solver.","category":"page"},{"location":"moi/background/infeasibility_certificates/#Infeasibility-certificates-of-variable-bounds","page":"Infeasibility certificates","title":"Infeasibility certificates of variable bounds","text":"","category":"section"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"Many linear solvers (for example, Gurobi) do not provide explicit access to the primal infeasibility certificate of a variable bound. However, given a set of linear constraints:","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"beginalign\nl_A le A x le u_A \nl_x le x le u_x\nendalign","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"the primal certificate of the variable bounds can be computed using the primal certificate associated with the affine constraints, d. (Note that d will have one element for each row of the A matrix, and that some or all of the elements in the vectors l_A and u_A may be pm infty. If both l_A and u_A are finite for some row, the corresponding element in `d  must be 0.)","category":"page"},{"location":"moi/background/infeasibility_certificates/","page":"Infeasibility certificates","title":"Infeasibility certificates","text":"Given d, compute bard = d^top A. If the bound is finite, a certificate for the lower variable bound of x_i is maxbard_i 0, and a certificate for the upper variable bound is minbard_i 0.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"EditURL = \"debugging.jl\"","category":"page"},{"location":"tutorials/getting_started/debugging/#Debugging","page":"Debugging","title":"Debugging","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Dealing with bugs is an unavoidable part of coding optimization models in any framework, including JuMP. Sources of bugs include not only generic coding errors (method errors, typos, off-by-one issues), but also semantic mistakes in the formulation of an optimization problem and the incorrect use of a solver.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"This tutorial explains some common sources of bugs and modeling issues that you might encounter when writing models in JuMP, and it suggests a variety of strategies to deal with them.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"tip: Tip\nThis tutorial is more advanced than the other \"Getting started\" tutorials. It's in the \"Getting started\" section to give you an early preview of how to debug JuMP models. However, if you are new to JuMP, you may want to briefly skim the tutorial, and come back to it once you have written a few JuMP models.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/getting_started/debugging/#Getting-help","page":"Debugging","title":"Getting help","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Debugging can be a frustrating part of modeling, particularly if you're new to optimization and programming. If you're stuck, join the community forum to search for answers to commonly asked questions.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Before asking a new question, make sure to read the post Make it easier to help you, which contains a number of tips on how to ask a good question.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Above all else, take time to simplify your code as much as possible. The fewer lines of code you can post that reproduces the same issue, the faster someone can answer your question.","category":"page"},{"location":"tutorials/getting_started/debugging/#Debugging-Julia-code","page":"Debugging","title":"Debugging Julia code","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Read the Debugging chapter in the book ThinkJulia.jl. It has a number of great tips and tricks for debugging Julia code.","category":"page"},{"location":"tutorials/getting_started/debugging/#Solve-failures","page":"Debugging","title":"Solve failures","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"When a solver experiences an issue that prevents it from finding an optimal solution (or proving that one does not exist), JuMP may return one of a number of termination_statuses.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"For example, if the solver found a solution, but experienced numerical imprecision, it may return a status such as ALMOST_OPTIMAL or ALMOST_LOCALLY_SOLVED indicating that the problem was solved to a relaxed set of tolerances. Alternatively, the solver may return a problematic status such as NUMERICAL_ERROR, SLOW_PROGRESS, or OTHER_ERROR, indicating that it could not find a solution to the problem.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Most solvers can experience numerical imprecision because they use floating-point arithmetic to perform operations such as addition, subtraction, and multiplication. These operations aren't exact, and small errors can accrue between the theoretical value and the value that the computer computes. For example:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"0.1 * 3 == 0.3","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"tip: Tip\nRead the Guidelines for numerical issues section of the Gurobi documentation, along with the Debugging numerical problems section of the YALMIP documentation.","category":"page"},{"location":"tutorials/getting_started/debugging/#Common-sources","page":"Debugging","title":"Common sources","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Common sources of solve failures are:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Very large numbers and very small numbers as problem coefficients. Exactly what \"large\" is depends on the solver and the problem, but in general, values above 1e6 or smaller than 1e-6 cause problems.\nNonlinear problems with functions that are not defined in parts of their domain. For example, minimizing log(x) where x >= 0 is undefined when x = 0 (a common starting value).","category":"page"},{"location":"tutorials/getting_started/debugging/#Strategies","page":"Debugging","title":"Strategies","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Strategies to debug sources of solve failures include:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Rescale variables in the problem and their associated coefficients to make the magnitudes of all coefficients in the 1e-4 to 1e4 range. For example, that might mean rescaling a variable from measuring distance in centimeters to kilometers.\nTry a different solver. Some solvers might be more robust than others for a particular problem.\nRead the documentation of your solver, and try settings that encourage numerical robustness.\nSet bounds or add constraints so that all nonlinear functions are defined across all of the feasible region. This particularly applies for functions like 1 / x and log(x) which are not defined for x = 0.","category":"page"},{"location":"tutorials/getting_started/debugging/#Incorrect-results","page":"Debugging","title":"Incorrect results","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Sometimes, you might find that the solver returns an \"optimal\" solution that is incorrect according to the model you are trying to solve (perhaps the solution is suboptimal, or it doesn't satisfy some of the constraints).","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Incorrect results can be hard to detect and debug, because the solver gives no hints that there is a problem. Indeed, the termination_status will likely be OPTIMAL and a solution will be available.","category":"page"},{"location":"tutorials/getting_started/debugging/#Common-sources-2","page":"Debugging","title":"Common sources","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Common sources of incorrect results are:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"A modeling error, so that your JuMP model does not match the formulation you have on paper\nNot accounting for the tolerances that solvers use (for example, if x is binary, a value like x = 1.0000001 may still be considered feasible)\nA bug in JuMP or the solver.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"The probability of the issue being a bug in JuMP or the solver is much smaller than a modeling error. When in doubt, first assume there is a bug in your code before assuming that there is a bug in JuMP.","category":"page"},{"location":"tutorials/getting_started/debugging/#Strategies-2","page":"Debugging","title":"Strategies","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Strategies to debug sources of incorrect results include:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Print your JuMP model to see if it matches the formulation you have on paper. Look out for incorrect signs + instead of -, and off-by-one errors such as x[t] instead of x[t-1].\nCheck that you are not using exact comparisons like value(x) == 1.0; always use isapprox(value(x), 1.0; atol = 1e-6) where you manually specify the comparison tolerance.\nTry a different solver. If one solver succeeds where another doesn't this is a sign that the problem is a numerical issue or a bug in the solver.","category":"page"},{"location":"tutorials/getting_started/debugging/#Debugging-an-infeasible-model","page":"Debugging","title":"Debugging an infeasible model","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"A model is infeasible if there is no primal solution that satisfies all of the constraints. In general, an infeasible model means one of two things:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Your problem really has no feasible solution\nThere is a mistake in your model.","category":"page"},{"location":"tutorials/getting_started/debugging/#Example","page":"Debugging","title":"Example","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"A simple example of an infeasible model is:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"model = Model(HiGHS.Optimizer);\nset_silent(model)\n@variable(model, x >= 0)\n@objective(model, Max, 2x + 1)\n@constraint(model, con, 2x - 1 <= -2)","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"because the bound says that x >= 0, but we can rewrite the constraint to be x <= -1/2. When the problem is infeasible, JuMP may return one of a number of statuses. The most common is INFEASIBLE:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"optimize!(model)\ntermination_status(model)","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Depending on the solver, you may also receive INFEASIBLE_OR_UNBOUNDED or LOCALLY_INFEASIBLE.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"A termination status of INFEASIBLE_OR_UNBOUNDED means that the solver could not prove if the solver was infeasible or unbounded, only that the model does not have a finite feasible optimal solution.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Nonlinear optimizers such as Ipopt may return the status LOCALLY_INFEASIBLE. This does not mean that the solver proved no feasible solution exists, only that it could not find one. If you know a primal feasible point, try providing it as a starting point using set_start_value and re-optimize.","category":"page"},{"location":"tutorials/getting_started/debugging/#Common-sources-3","page":"Debugging","title":"Common sources","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Common sources of infeasibility are:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Incorrect units, for example, using a lower bound of megawatts and an upper bound of kilowatts\nUsing + instead of - in a constraint\nOff-by-one and related errors, for example, using x[t] instead of x[t-1] in part of a constraint\nOtherwise invalid mathematical formulations","category":"page"},{"location":"tutorials/getting_started/debugging/#Strategies-3","page":"Debugging","title":"Strategies","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Strategies to debug sources of infeasibility include:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Iteratively comment out a constraint (or block of constraints) and re-solve the problem. When you find a constraint that makes the problem infeasible when added, check the constraint carefully for errors.\nIf the problem is still infeasible with all constraints commented out, check all variable bounds. Do they use the right data?\nIf you have a known feasible solution, use primal_feasibility_report to evaluate the constraints and check for violations. You'll probably find that you have a typo in one of the constraints.\nTry a different solver. Sometimes, solvers have bugs, and they can incorrectly report a problem as infeasible when it isn't. If you find such a case where one solver reports the problem is infeasible and another can find an optimal solution, please report it by opening an issue on the GitHub repository of the solver that reports infeasibility.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"tip: Tip\nSome solvers also have specialized support for debugging sources of infeasibility via an irreducible infeasible subsystem. To see if your solver has support, try calling compute_conflict!:julia> compute_conflict!(model)\nERROR: ArgumentError: The optimizer HiGHS.Optimizer does not support `compute_conflict!`In this case, HiGHS does not support computing conflicts, but other solvers such as Gurobi and CPLEX do. If the solver does support computing conflicts, read Conflicts for more details.","category":"page"},{"location":"tutorials/getting_started/debugging/#Penalty-relaxation","page":"Debugging","title":"Penalty relaxation","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Another strategy to debug sources of infeasibility is the relax_with_penalty! function.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"The penalty relaxation modifies constraints of the form f(x) in S into f(x) + y - z in S, where y z ge 0, and then it introduces a penalty term into the objective of a times (y + z) (if minimizing, else -a), where a is a penalty.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"map = relax_with_penalty!(model)","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Here map is a dictionary which maps constraint indices to an affine expression representing (y + z).","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"If we optimize the relaxed model, this time we get a feasible solution:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"optimize!(model)\ntermination_status(model)","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Iterate over the contents of map to see which constraints are violated:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"for (con, penalty) in map\n    violation = value(penalty)\n    if violation > 0\n        println(\"Constraint `$(name(con))` is violated by $violation\")\n    end\nend","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Once you find a violated constraint in the relaxed problem, take a look to see if there is a typo or other common mistake in that particular constraint.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Consult the docstring relax_with_penalty! for information on how to modify the penalty cost term a, either for every constraint in the model or a particular subset of the constraints.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"When using relax_with_penalty!, you should be aware that:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Variable bounds and integrality restrictions are not relaxed. If the problem is still infeasible after calling relax_with_penalty!, check the variable bounds.\nYou cannot undo the penalty relaxation. If you need an unmodified model, rebuild the problem, or call copy_model before calling relax_with_penalty!.","category":"page"},{"location":"tutorials/getting_started/debugging/#Debugging-an-unbounded-model","page":"Debugging","title":"Debugging an unbounded model","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"A model is unbounded if there is no limit on how good the objective value can get. Most often, an unbounded model means that you have an error in your modeling, because all physical systems have limits. (You cannot make an infinite amount of profit.)","category":"page"},{"location":"tutorials/getting_started/debugging/#Example-2","page":"Debugging","title":"Example","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"A simple example of an unbounded model is:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"model = Model(HiGHS.Optimizer);\nset_silent(model)\n@variable(model, x >= 0)\n@objective(model, Max, 2x + 1)","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"because we can increase x without limit, and the objective value 2x + 1 gets better as x increases.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"When the problem is unbounded, JuMP may return one of a number of statuses. The most common is DUAL_INFEASIBLE:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"optimize!(model)\ntermination_status(model)","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Depending on the solver, you may also receive INFEASIBLE_OR_UNBOUNDED or an error code like NORM_LIMIT.","category":"page"},{"location":"tutorials/getting_started/debugging/#Common-sources-4","page":"Debugging","title":"Common sources","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Common sources of unboundedness are:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Using Max instead of Min\nOmitting variable bounds, such as 0 <= x <= 1\nUsing + instead of - in a term of the objective function.","category":"page"},{"location":"tutorials/getting_started/debugging/#Strategies-4","page":"Debugging","title":"Strategies","text":"","category":"section"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Strategies to debug sources of unboundedness include:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"Double check whether you intended Min or Max in the @objective line.\nPrint the objective function with print(objective_function(model)) and verify that the value and sign of each coefficient is as you expect.\nAdd large bounds to all variables that are free or have one-sided bounds, then re-solve the problem. Because all variables are now bounded, the problem will have a finite optimal solution. Look at the value of each variable in the optimal solution to see if it is at one of the new bounds. If it is, you either need to specify a better bound for that variable, or there might be a mistake in the objective function associated with that variable (for example, a + instead of a -).","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"If there are too many variables to add bounds to, or there are too many terms to examine by hand, another strategy is to create a new variable with a large upper bound (if maximizing, lower bound if minimizing) and a constraint that the variable must be less-than or equal to the expression of the objective function. For example:","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"model = Model(HiGHS.Optimizer);\nset_silent(model)\n@variable(model, x >= 0)\n# @objective(model, Max, 2x + 1)\n@variable(model, objective <= 10_000)\n@constraint(model, objective <= 2x + 1)\n@objective(model, Max, objective)","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"This new model has a finite optimal solution, so we can solve it and then look for variables with large positive or negative values in the optimal solution.","category":"page"},{"location":"tutorials/getting_started/debugging/","page":"Debugging","title":"Debugging","text":"optimize!(model)\nfor var in all_variables(model)\n    if var == objective\n        continue\n    end\n    if abs(value(var)) > 1e3\n        println(\"Variable `$(name(var))` may be unbounded\")\n    end\nend","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"EditURL = \"https://github.com/jump-dev/ECOS.jl/blob/v1.1.1/README.md\"","category":"page"},{"location":"packages/ECOS/#ECOS.jl","page":"jump-dev/ECOS.jl","title":"ECOS.jl","text":"","category":"section"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"ECOS.jl is a wrapper for the ECOS solver.","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/ECOS/#Affiliation","page":"jump-dev/ECOS.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"This wrapper is maintained by the JuMP community and is not a product of Embotech AG.","category":"page"},{"location":"packages/ECOS/#License","page":"jump-dev/ECOS.jl","title":"License","text":"","category":"section"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"ECOS.jl is licensed under the MIT License.","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"The underlying solver, embotech/ecos, is licensed under the GPL v3 license.","category":"page"},{"location":"packages/ECOS/#Installation","page":"jump-dev/ECOS.jl","title":"Installation","text":"","category":"section"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"Install ECOS.jl using Pkg.add:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"import Pkg\nPkg.add(\"ECOS\")","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"In addition to installing the ECOS.jl package, this will also download and install the ECOS binaries. You do not need to install ECOS separately.","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"To use a custom binary, read the Custom solver binaries section of the JuMP documentation.","category":"page"},{"location":"packages/ECOS/#Use-with-JuMP","page":"jump-dev/ECOS.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"To use ECOS with JuMP, use ECOS.Optimizer:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"using JuMP, ECOS\nmodel = Model(ECOS.Optimizer)\nset_attribute(model, \"maxit\", 100)","category":"page"},{"location":"packages/ECOS/#MathOptInterface-API","page":"jump-dev/ECOS.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"The ECOS optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"MOI.Reals","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"MOI.VectorAffineFunction{Float64} in MOI.Nonnegatives\nMOI.VectorAffineFunction{Float64} in MOI.SecondOrderCone\nMOI.VectorAffineFunction{Float64} in MOI.Zeros","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"MOI.ObjectiveSense()","category":"page"},{"location":"packages/ECOS/#Options","page":"jump-dev/ECOS.jl","title":"Options","text":"","category":"section"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"The following options are supported:","category":"page"},{"location":"packages/ECOS/","page":"jump-dev/ECOS.jl","title":"jump-dev/ECOS.jl","text":"Parameter Explanation\ngamma scaling the final step length\ndelta regularization parameter\neps regularization threshold\nfeastol primal/dual infeasibility tolerance\nabstol absolute tolerance on duality gap\nreltol relative tolerance on duality gap\nfeastol_inacc primal/dual infeasibility relaxed tolerance\nabstol_inacc absolute relaxed tolerance on duality gap\nreltol_inacc relative relaxed tolerance on duality gap\nnitref number of iterative refinement steps\nmaxit maximum number of iterations\nverbose verbosity bool for PRINTLEVEL < 3","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Nonlinear/overview.md\"","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#nonlinear_developers","page":"Overview","title":"Nonlinear","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"warning: Warning\nThe Nonlinear submodule is experimental. Until this message is removed, breaking changes may be introduced in any minor or patch release of MathOptInterface.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The Nonlinear submodule contains data structures and functions for working with a nonlinear optimization problem in the form of an expression graph. This page explains the API and describes the rationale behind its design.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Standard-form","page":"Overview","title":"Standard form","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear programs (NLPs) are a class of optimization problems in which some of the constraints or the objective function are nonlinear:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"beginalign\n    min_x in mathbbR^n  f_0(x) \n    textst  l_j le f_j(x) le u_j  j = 1 ldots m\nendalign","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"There may be additional constraints, as well as things like variable bounds and integrality restrictions, but we do not consider them here because they are best dealt with by other components of MathOptInterface.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#API-overview","page":"Overview","title":"API overview","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The core element of the Nonlinear submodule is Nonlinear.Model:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> const Nonlinear = MOI.Nonlinear;\n\njulia> model = Nonlinear.Model()\nA Nonlinear.Model with:\n 0 objectives\n 0 parameters\n 0 expressions\n 0 constraints","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear.Model is a mutable struct that stores all of the nonlinear information added to the model.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Decision-variables","page":"Overview","title":"Decision variables","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Decision variables are represented by VariableIndexes. The user is responsible for creating these using MOI.VariableIndex(i), where i is the column associated with the variable.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Nonlinear_Expressions","page":"Overview","title":"Expressions","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The input data structure is a Julia Expr. The input expressions can incorporate VariableIndexes, but these must be interpolated into the expression with $:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> x = MOI.VariableIndex(1)\nMOI.VariableIndex(1)\n\njulia> input = :(1 + sin($x)^2)\n:(1 + sin(MathOptInterface.VariableIndex(1)) ^ 2)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"There are a number of restrictions on the input Expr:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"It cannot contain macros\nIt cannot contain broadcasting\nIt cannot contain splatting (except in limited situations)\nIt cannot contain linear algebra, such as matrix-vector products\nIt cannot contain generator expressions, including sum(i for i in S)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Given an input expression, add an expression using Nonlinear.add_expression:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> expr = Nonlinear.add_expression(model, input)\nMathOptInterface.Nonlinear.ExpressionIndex(1)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The return value, expr, is a Nonlinear.ExpressionIndex that can then be interpolated into other input expressions.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Looking  again at model, we see:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> model\nA Nonlinear.Model with:\n 0 objectives\n 0 parameters\n 1 expression\n 0 constraints","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Nonlinear_Parameters","page":"Overview","title":"Parameters","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"In addition to constant literals like 1 or 1.23, you can create parameters. Parameters are placeholders whose values can change before passing the expression to the solver. Create a parameter using Nonlinear.add_parameter, which accepts a default value:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> p = Nonlinear.add_parameter(model, 1.23)\nMathOptInterface.Nonlinear.ParameterIndex(1)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The return value, p, is a Nonlinear.ParameterIndex that can then be interpolated into other input expressions.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Looking  again at model, we see:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> model\nA Nonlinear.Model with:\n 0 objectives\n 1 parameter\n 1 expression\n 0 constraints","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Update a parameter as follows:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> model[p]\n1.23\n\njulia> model[p] = 4.56\n4.56\n\njulia> model[p]\n4.56","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Nonlinear_Objectives","page":"Overview","title":"Objectives","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Set a nonlinear objective using Nonlinear.set_objective:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> Nonlinear.set_objective(model, :($p + $expr + $x))\n\njulia> model\nA Nonlinear.Model with:\n 1 objective\n 1 parameter\n 1 expression\n 0 constraints","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Clear a nonlinear objective by passing nothing:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> Nonlinear.set_objective(model, nothing)\n\njulia> model\nA Nonlinear.Model with:\n 0 objectives\n 1 parameter\n 1 expression\n 0 constraints","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"But we'll re-add the objective for later:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> Nonlinear.set_objective(model, :($p + $expr + $x));","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Nonlinear_Constraints","page":"Overview","title":"Constraints","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Add a constraint using Nonlinear.add_constraint:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> c = Nonlinear.add_constraint(model, :(1 + sqrt($x)), MOI.LessThan(2.0))\nMathOptInterface.Nonlinear.ConstraintIndex(1)\n\njulia> model\nA Nonlinear.Model with:\n 1 objective\n 1 parameter\n 1 expression\n 1 constraint","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The return value, c, is a Nonlinear.ConstraintIndex that is a unique identifier for the constraint. Interval constraints are also supported:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> c2 = Nonlinear.add_constraint(model, :(1 + sqrt($x)), MOI.Interval(-1.0, 2.0))\nMathOptInterface.Nonlinear.ConstraintIndex(2)\n\njulia> model\nA Nonlinear.Model with:\n 1 objective\n 1 parameter\n 1 expression\n 2 constraints","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Delete a constraint using Nonlinear.delete:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> Nonlinear.delete(model, c2)\n\njulia> model\nA Nonlinear.Model with:\n 1 objective\n 1 parameter\n 1 expression\n 1 constraint","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#User-defined-operators","page":"Overview","title":"User-defined operators","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"By default, Nonlinear supports a wide range of univariate and multivariate operators. However, you can also define your own operators by registering them.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Univariate-operators","page":"Overview","title":"Univariate operators","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Register a univariate user-defined operator using Nonlinear.register_operator:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> f(x) = 1 + sin(x)^2\nf (generic function with 1 method)\n\njulia> Nonlinear.register_operator(model, :my_f, 1, f)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Now, you can use :my_f in expressions:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> new_expr = Nonlinear.add_expression(model, :(my_f($x + 1)))\nMathOptInterface.Nonlinear.ExpressionIndex(2)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"By default, Nonlinear will compute first- and second-derivatives of the registered operator using ForwardDiff.jl. Override this by passing functions which compute the respective derivative:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> f′(x) = 2 * sin(x) * cos(x)\nf′ (generic function with 1 method)\n\njulia> Nonlinear.register_operator(model, :my_f2, 1, f, f′)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"or","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> f′′(x) = 2 * (cos(x)^2 - sin(x)^2)\nf′′ (generic function with 1 method)\n\njulia> Nonlinear.register_operator(model, :my_f3, 1, f, f′, f′′)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Multivariate-operators","page":"Overview","title":"Multivariate operators","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Register a multivariate user-defined operator using Nonlinear.register_operator:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> g(x...) = x[1]^2 + x[1] * x[2] + x[2]^2\ng (generic function with 1 method)\n\njulia> Nonlinear.register_operator(model, :my_g, 2, g)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Now, you can use :my_g in expressions:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> new_expr = Nonlinear.add_expression(model, :(my_g($x + 1, $x)))\nMathOptInterface.Nonlinear.ExpressionIndex(3)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"By default, Nonlinear will compute the gradient of the registered operator using ForwardDiff.jl. (Hessian information is not supported.) Override this by passing a function to compute the gradient:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> function ∇g(ret, x...)\n           ret[1] = 2 * x[1] + x[2]\n           ret[2] = x[1] + 2 * x[2]\n           return\n       end\n∇g (generic function with 1 method)\n\njulia> Nonlinear.register_operator(model, :my_g2, 2, g, ∇g)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Nonlinear_MOI_interface","page":"Overview","title":"MathOptInterface","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"MathOptInterface communicates the nonlinear portion of an optimization problem to solvers using concrete subtypes of AbstractNLPEvaluator, which implement the Nonlinear programming API.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Create an AbstractNLPEvaluator from Nonlinear.Model using Nonlinear.Evaluator.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear.Evaluator requires an Nonlinear.AbstractAutomaticDifferentiation backend and an ordered list of the variables that are included in the model.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"There following backends are available to choose from within MOI, although other packages may add more options by sub-typing Nonlinear.AbstractAutomaticDifferentiation:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear.ExprGraphOnly\nNonlinear.SparseReverseMode.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> evaluator = Nonlinear.Evaluator(model, Nonlinear.ExprGraphOnly(), [x])\nNonlinear.Evaluator with available features:\n  * :ExprGraph","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The functions of the Nonlinear programming API implemented by Nonlinear.Evaluator depends upon the chosen Nonlinear.AbstractAutomaticDifferentiation backend.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The :ExprGraph feature means we can call objective_expr and constraint_expr to retrieve the expression graph of the problem. However, we cannot call gradient terms such as eval_objective_gradient because Nonlinear.ExprGraphOnly does not have the capability to differentiate a nonlinear expression.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"If, instead, we pass Nonlinear.SparseReverseMode, then we get access to :Grad, the gradient of the objective function, :Jac, the Jacobian matrix of the constraints, :JacVec, the ability to compute Jacobian-vector products, and :ExprGraph.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> evaluator = Nonlinear.Evaluator(\n           model,\n           Nonlinear.SparseReverseMode(),\n           [x],\n       )\nNonlinear.Evaluator with available features:\n  * :Grad\n  * :Jac\n  * :JacVec\n  * :ExprGraph","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"However, before using the evaluator, we need to call initialize:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> MOI.initialize(evaluator, [:Grad, :Jac, :JacVec, :ExprGraph])","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Now we can call methods like eval_objective:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> x = [1.0]\n1-element Vector{Float64}:\n 1.0\n\njulia> MOI.eval_objective(evaluator, x)\n7.268073418273571","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"and eval_objective_gradient:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> grad = [0.0]\n1-element Vector{Float64}:\n 0.0\n\njulia> MOI.eval_objective_gradient(evaluator, grad, x)\n\njulia> grad\n1-element Vector{Float64}:\n 1.909297426825682","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Instead of passing Nonlinear.Evaluator directly to solvers, solvers query the NLPBlock attribute, which returns an NLPBlockData. This object wraps an Nonlinear.Evaluator and includes other information such as constraint bounds and whether the evaluator has a nonlinear objective. Create and set NLPBlockData as follows:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> block = MOI.NLPBlockData(evaluator);\n\njulia> model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}());\n\njulia> MOI.set(model, MOI.NLPBlock(), block);","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"warning: Warning\nOnly call NLPBlockData once you have finished modifying the problem in model.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Putting everything together, you can create a nonlinear optimization problem in MathOptInterface as follows:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"import MathOptInterface as MOI\n\nfunction build_model(\n    model::MOI.ModelLike;\n    backend::MOI.Nonlinear.AbstractAutomaticDifferentiation,\n)\n    x = MOI.add_variable(model)\n    y = MOI.add_variable(model)\n    MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)\n    nl_model = MOI.Nonlinear.Model()\n    MOI.Nonlinear.set_objective(nl_model, :($x^2 + $y^2))\n    evaluator = MOI.Nonlinear.Evaluator(nl_model, backend, [x, y])\n    MOI.set(model, MOI.NLPBlock(), MOI.NLPBlockData(evaluator))\n    return\nend\n\n# Replace `model` and `backend` with your optimizer and backend of choice.\nmodel = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())\nbuild_model(model; backend = MOI.Nonlinear.SparseReverseMode())","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Expression-graph-representation","page":"Overview","title":"Expression-graph representation","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear.Model stores nonlinear expressions in Nonlinear.Expressions. This section explains the design of the expression graph data structure in Nonlinear.Expression.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Given a nonlinear function like f(x) = sin(x)^2 + x, a conceptual aid for thinking about the graph representation of the expression is to convert it into Polish prefix notation:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"f(x, y) = (+ (^ (sin x) 2) x)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"This format identifies each operator (function), as well as a list of arguments. Operators can be univariate, like sin, or multivariate, like +.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"A common way of representing Polish prefix notation in code is as follows:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> x = MOI.VariableIndex(1);\n\njulia> struct ExprNode\n           op::Symbol\n           children::Vector{Union{ExprNode,Float64,MOI.VariableIndex}}\n       end\n\njulia> expr = ExprNode(:+, [ExprNode(:^, [ExprNode(:sin, [x]), 2.0]), x]);","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"This data structure follows our Polish prefix notation very closely, and we can easily identify the arguments to an operator. However, it has a significant draw-back: each node in the graph requires a Vector, which is heap-allocated and tracked by Julia's garbage collector (GC). For large models, we can expect to have millions of nodes in the expression graph, so this overhead quickly becomes prohibitive for computation.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"An alternative is to record the expression as a linear tape:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> expr = Any[:+, 2, :^, 2, :sin, 1, x, 2.0, x]\n9-element Vector{Any}:\n  :+\n 2\n  :^\n 2\n  :sin\n 1\n  MOI.VariableIndex(1)\n 2.0\n  MOI.VariableIndex(1)","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The Int after each operator Symbol specifies the number of arguments.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"This data-structure is a single vector, which resolves our problem with the GC, but each element is the abstract type, Any, and so any operations on it will lead to slower dynamic dispatch. It's also hard to identify the children of each operation without reading the entire tape.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"To summarize, representing expression graphs in Julia has the following challenges:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nodes in the expression graph should not contain a heap-allocated object\nAll data-structures should be concretely typed\nIt should be easy to identify the children of a node","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Sketch-of-the-design-in-Nonlinear","page":"Overview","title":"Sketch of the design in Nonlinear","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear overcomes these problems by decomposing the data structure into a number of different concrete-typed vectors.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"First, we create vectors of the supported uni- and multivariate operators.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> const UNIVARIATE_OPERATORS = [:sin];\n\njulia> const MULTIVARIATE_OPERATORS = [:+, :^];","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"In practice, there are many more supported operations than the ones listed here.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Second, we create an enum to represent the different types of nodes present in the expression graph:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> @enum(\n           NodeType,\n           NODE_CALL_MULTIVARIATE,\n           NODE_CALL_UNIVARIATE,\n           NODE_VARIABLE,\n           NODE_VALUE,\n       )","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"In practice, there are node types other than the ones listed here.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Third, we create two concretely typed structs as follows:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> struct Node\n           type::NodeType\n           parent::Int\n           index::Int\n       end\n\njulia> struct Expression\n           nodes::Vector{Node}\n           values::Vector{Float64}\n       end","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"For each node node in the .nodes field, if node.type is:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"NODE_CALL_MULTIVARIATE, we look up MULTIVARIATE_OPERATORS[node.index] to retrieve the operator\nNODE_CALL_UNIVARIATE, we look up UNIVARIATE_OPERATORS[node.index] to retrieve the operator\nNODE_VARIABLE, we create MOI.VariableIndex(node.index)\nNODE_VALUE, we look up values[node.index]","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The .parent field of each node is the integer index of the parent node in .nodes. For the first node, the parent is -1 by convention.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Therefore, we can represent our function as:","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"julia> expr = Expression(\n           [\n               Node(NODE_CALL_MULTIVARIATE, -1, 1),\n               Node(NODE_CALL_MULTIVARIATE, 1, 2),\n               Node(NODE_CALL_UNIVARIATE, 2, 1),\n               Node(NODE_VARIABLE, 3, 1),\n               Node(NODE_VALUE, 2, 1),\n               Node(NODE_VARIABLE, 1, 1),\n           ],\n           [2.0],\n       );","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"This is less readable than the other options, but does this data structure meet our design goals?","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Instead of a heap-allocated object for each node, we only have two Vectors for each expression, nodes and values, as well as two constant vectors for the OPERATORS. In addition, all fields are concretely typed, and there are no Union or Any types.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"For our third goal, it is not easy to identify the children of a node, but it is easy to identify the parent of any node. Therefore, we can use Nonlinear.adjacency_matrix to compute a sparse matrix that maps parents to their children.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The tape is also ordered topologically, so that a reverse pass of the nodes evaluates all children nodes before their parent.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#The-design-in-practice","page":"Overview","title":"The design in practice","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"In practice, Node and Expression are exactly Nonlinear.Node and Nonlinear.Expression. However, Nonlinear.NodeType has more fields to account for comparison operators such as :>= and :<=, logic operators such as :&& and :||, nonlinear parameters, and nested subexpressions.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Moreover, instead of storing the operators as global constants, they are stored in Nonlinear.OperatorRegistry, and it also stores a vector of logic operators and a vector of comparison operators. In addition to Nonlinear.DEFAULT_UNIVARIATE_OPERATORS and Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS, you can register user-defined functions using Nonlinear.register_operator.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear.Model is a struct that stores the Nonlinear.OperatorRegistry, as well as a list of parameters and subexpressions in the model.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#ReverseAD","page":"Overview","title":"ReverseAD","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Nonlinear.ReverseAD is a submodule for computing derivatives of a nonlinear optimization problem using sparse reverse-mode automatic differentiation (AD).","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"This section does not attempt to explain how sparse reverse-mode AD works, but instead explains why MOI contains its own implementation, and highlights notable differences from similar packages.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"warning: Warning\nDon't use the API in ReverseAD to compute derivatives. Instead, create a Nonlinear.Evaluator object with Nonlinear.SparseReverseMode as the backend, and then query the MOI API methods.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#Design-goals","page":"Overview","title":"Design goals","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"The JuliaDiff organization maintains a list of packages for doing AD in Julia. At last count, there were at least ten packages——not including ReverseAD——for reverse-mode AD in Julia. ReverseAD exists because it has a different set of design goals.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Goal: handle scale and sparsity. The types of nonlinear optimization problems that MOI represents can be large scale (10^5 or more functions across 10^5 or more variables) with very sparse derivatives. The ability to compute a sparse Hessian matrix is essential. To the best of our knowledge, ReverseAD is the only reverse-mode AD system in Julia that handles sparsity by default.\nGoal: limit the scope to improve robustness. Most other AD packages accept arbitrary Julia functions as input and then trace an expression graph using operator overloading. This means they must deal (or detect and ignore) with control flow, I/O, and other vagaries of Julia. In contrast, ReverseAD only accepts functions in the form of Nonlinear.Expression, which greatly limits the range of syntax that it must deal with. By reducing the scope of what we accept as input to functions relevant for mathematical optimization, we can provide a simpler implementation with various performance optimizations.\nGoal: provide outputs which match what solvers expect. Other AD packages focus on differentiating individual Julia functions. In contrast, ReverseAD has a very specific use-case: to generate outputs needed by the MOI nonlinear API. This means it needs to efficiently compute sparse Hessians, and it needs subexpression handling to avoid recomputing subexpressions that are shared between functions.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/#History","page":"Overview","title":"History","text":"","category":"section"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"ReverseAD started life as ReverseDiffSparse.jl, development of which began in early 2014(!). This was well before the other AD packages started development. Because we had a well-tested, working AD in JuMP, there was less motivation to contribute to and explore other AD packages. The lack of historical interaction also meant that other packages were not optimized for the types of problems that JuMP is built for (that is, large-scale sparse problems). When we first created MathOptInterface, we kept the AD in JuMP to simplify the transition, and post-poned the development of a first-class nonlinear interface in MathOptInterface.","category":"page"},{"location":"moi/submodules/Nonlinear/overview/","page":"Overview","title":"Overview","text":"Prior to the introduction of Nonlinear, JuMP's nonlinear implementation was a confusing mix of functions and types spread across the code base and in the private _Derivatives submodule. This made it hard to swap the AD system for another. The main motivation for refactoring JuMP to create the Nonlinear submodule in MathOptInterface was to abstract the interface between JuMP and the AD system, allowing us to swap-in and test new AD systems in the future.","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"EditURL = \"https://github.com/ds4dm/Tulip.jl/blob/069cdda082dfde31ef7ee935c96e833a3b96b33c/README.md\"","category":"page"},{"location":"packages/Tulip/#Tulip","page":"ds4dm/Tulip.jl","title":"Tulip","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"(Image: DOI) (Image: ) (Image: )","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"Tulip is an open-source interior-point solver for linear optimization, written in pure Julia. It implements the homogeneous primal-dual interior-point algorithm with multiple centrality corrections, and therefore handles unbounded and infeasible problems. Tulip’s main feature is that its algorithmic framework is disentangled from linear algebra implementations. This allows to seamlessly integrate specialized routines for structured problems.","category":"page"},{"location":"packages/Tulip/#License","page":"ds4dm/Tulip.jl","title":"License","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"Tulip is licensed under the MPL 2.0 license.","category":"page"},{"location":"packages/Tulip/#Installation","page":"ds4dm/Tulip.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"Install Tulip using the Julia package manager:","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"import Pkg\nPkg.add(\"Tulip\")","category":"page"},{"location":"packages/Tulip/#Usage","page":"ds4dm/Tulip.jl","title":"Usage","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"The recommended way of using Tulip is through JuMP or MathOptInterface (MOI).","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"The low-level interface is still under development and is likely change in the future. The MOI interface is more stable.","category":"page"},{"location":"packages/Tulip/#Using-with-JuMP","page":"ds4dm/Tulip.jl","title":"Using with JuMP","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"Tulip follows the syntax convention PackageName.Optimizer:","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"using JuMP\nimport Tulip\nmodel = Model(Tulip.Optimizer)","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"Linear objectives, linear constraints and lower/upper bounds on variables are supported.","category":"page"},{"location":"packages/Tulip/#Using-with-MOI","page":"ds4dm/Tulip.jl","title":"Using with MOI","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"The type Tulip.Optimizer is parametrized by the model's arithmetic, for example, Float64 or BigFloat. This allows to solve problem in higher numerical precision. See the documentation for more details.","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"import MathOptInterface as MOI\nimport Tulip\nmodel = Tulip.Optimizer{Float64}()   # Create a model in Float64 precision\nmodel = Tulip.Optimizer()            # Defaults to the above call\nmodel = Tulip.Optimizer{BigFloat}()  # Create a model in BigFloat precision","category":"page"},{"location":"packages/Tulip/#Solver-parameters","page":"ds4dm/Tulip.jl","title":"Solver parameters","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"See the documentation for a full list of parameters.","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"To set parameters in JuMP, use:","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"using JuMP, Tulip\nmodel = Model(Tulip.Optimizer)\nset_attribute(model, \"IPM_IterationsLimit\", 200)","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"To set parameters in MathOptInterface, use:","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"using Tulip\nimport MathOptInterface as MOI\nmodel = Tulip.Optimizer{Float64}()\nMOI.set(model, MOI.RawOptimizerAttribute(\"IPM_IterationsLimit\"), 200)","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"To set parameters in the Tulip API, use:","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"using Tulip\nmodel = Tulip.Model{Float64}()\nTulip.set_parameter(model, \"IPM_IterationsLimit\", 200)","category":"page"},{"location":"packages/Tulip/#Command-line-executable","page":"ds4dm/Tulip.jl","title":"Command-line executable","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"See app building instructions.","category":"page"},{"location":"packages/Tulip/#Citing-Tulip.jl","page":"ds4dm/Tulip.jl","title":"Citing Tulip.jl","text":"","category":"section"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"If you use Tulip in your work, we kindly ask that you cite the following reference (preprint available here).","category":"page"},{"location":"packages/Tulip/","page":"ds4dm/Tulip.jl","title":"ds4dm/Tulip.jl","text":"@Article{Tulip.jl,\n  author   = {Tanneau, Mathieu and Anjos, Miguel F. and Lodi, Andrea},\n  journal  = {Mathematical Programming Computation},\n  title    = {Design and implementation of a modular interior-point solver for linear optimization},\n  year     = {2021},\n  issn     = {1867-2957},\n  month    = feb,\n  doi      = {10.1007/s12532-020-00200-8},\n  language = {en},\n  url      = {https://doi.org/10.1007/s12532-020-00200-8},\n  urldate  = {2021-03-07},\n}","category":"page"},{"location":"tutorials/algorithms/parallelism/#Parallelism","page":"Parallelism","title":"Parallelism","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"The purpose of this tutorial is to give a brief overview of parallelism in Julia as it pertains to JuMP, and to explain some of the things to be aware of when writing parallel algorithms involving JuMP models.","category":"page"},{"location":"tutorials/algorithms/parallelism/#Multi-threading-and-distributed-computing","page":"Parallelism","title":"Multi-threading and distributed computing","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"There are two main types of parallelism in Julia:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Multi-threading\nDistributed computing","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"In multi-threading, multiple tasks are run in a single Julia process and share the same memory. In distributed computing, tasks are run in multiple Julia processes with independent memory spaces. This can include processes across multiple physical machines, such as in a high-performance computing cluster.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Choosing and understanding the type of parallelism you are using is important because the code you write for each type is different, and there are different limitations and benefits to each approach. However, the best choice is highly problem dependent, so you may want to experiment with both approaches to determine what works for your situation.","category":"page"},{"location":"tutorials/algorithms/parallelism/#Multi-threading","page":"Parallelism","title":"Multi-threading","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"To use multi-threading with Julia, you must either start Julia with the command line flag --threads=N, or you must set the JULIA_NUM_THREADS environment variable before launching Julia. For this documentation, we set the environment variable to:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> ENV[\"JULIA_NUM_THREADS\"]\n\"4\"","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"You can check how many threads are available using:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> Threads.nthreads()\n4","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"The easiest way to use multi-threading in Julia is by placing the Threads.@threads macro in front of a for-loop:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> @time begin\n           ids = Int[]\n           my_lock = Threads.ReentrantLock()\n           Threads.@threads for i in 1:Threads.nthreads()\n               global ids, my_lock\n               Threads.lock(my_lock) do\n                   push!(ids, Threads.threadid())\n               end\n               sleep(1.0)\n           end\n       end\n  1.037087 seconds (31.32 k allocations: 1.836 MiB, 2.02% compilation time)","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"This for-loop sleeps for 1 second on each iteration. Thus, if it had executed sequentially, it should have taken the same number of seconds as there are threads available. Instead, it took only 1 second, showing that the iterations were executed simultaneously. We can verify this by checking the Threads.threadid() of the thread that executed each iteration:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> ids\n4-element Vector{Int64}:\n 2\n 4\n 1\n 3","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"warning: Warning\nWhen working with threads, you need to avoid race conditions, in which two threads attempt to write to the same variable at the same time. In the above example we avoided a race condition by using ReentrantLock. See the Multi-threading section of the Julia documentation for more details.","category":"page"},{"location":"tutorials/algorithms/parallelism/#Distributed-computing","page":"Parallelism","title":"Distributed computing","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"To use distributed computing with Julia, use the Distributed package:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> import Distributed","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Like multi-threading, we need to tell Julia how many processes to add. We can do this either by starting Julia with the -p N command line argument, or by using Distributed.addprocs:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> import Pkg\n\njulia> project = Pkg.project();\n\njulia> workers = Distributed.addprocs(4; exeflags = \"--project=$(project.path)\")\n4-element Vector{Int64}:\n 2\n 3\n 4\n 5","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"warning: Warning\nNot loading the parent environment with --project is a common mistake.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"The added processes are \"worker\" processes that we can use to do computation with. They are orchestrated by the process with the id 1. You can check what process the code is currently running on using Distributed.myid()","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> Distributed.myid()\n1","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"As a general rule, to get maximum performance you should add as many processes as you have logical cores available.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Unlike the for-loop approach of multi-threading, distributed computing extends the Julia map function to a \"parallel-map\" function Distributed.pmap. For each element in the list of arguments to map over, Julia will copy the element to an idle worker process and evaluate the function, passing the element as an input argument.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> function hard_work(i::Int)\n           sleep(1.0)\n           return Distributed.myid()\n       end\nhard_work (generic function with 1 method)\n\njulia> Distributed.pmap(hard_work, 1:4)\nERROR: On worker 2:\nUndefVarError: #hard_work not defined\nStacktrace:\n[...]","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Unfortunately, if you try this code directly, you will get an error message that says On worker 2: UndefVarError: hard_work not defined. The error is thrown because, although process 1 knows what the hard_work function is, the worker processes do not.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"To fix the error, we need to use Distributed.@everywhere, which evaluates the code on every process:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> Distributed.@everywhere begin\n           function hard_work(i::Int)\n               sleep(1.0)\n               return Distributed.myid()\n           end\n       end","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Now if we run pmap, we see that it took only 1 second instead of 4, and that it executed on each of the worker processes:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> @time ids = Distributed.pmap(hard_work, 1:4)\n  1.202006 seconds (216.39 k allocations: 13.301 MiB, 4.07% compilation time)\n4-element Vector{Int64}:\n 2\n 3\n 5\n 4","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"tip: Tip\nFor more information, read the Julia documentation Distributed Computing.","category":"page"},{"location":"tutorials/algorithms/parallelism/#Using-parallelism-the-wrong-way","page":"Parallelism","title":"Using parallelism the wrong way","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/#When-building-a-JuMP-model","page":"Parallelism","title":"When building a JuMP model","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"For large problems, building the model in JuMP can be a bottleneck, and you may consider trying to write code that builds the model in parallel, for example, by wrapping a for-loop that adds constraints with Threads.@threads.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Unfortunately, you cannot build a JuMP model in parallel, and attempting to do so may error or produce incorrect results.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"In most cases, we find that the reason for the bottleneck is not JuMP, but in how you are constructing the problem data, and that with changes, it is possible to build a model in a way that is not the bottleneck in the solution process.","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"tip: Tip\nLooking for help to make your code run faster? Ask for help on the community forum. Make sure to include a reproducible example of your code.","category":"page"},{"location":"tutorials/algorithms/parallelism/#With-a-single-JuMP-model","page":"Parallelism","title":"With a single JuMP model","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"A common approach people try is to use parallelism with a single JuMP model. For example, to optimize a model over multiple right-hand side vectors, you may try:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"using JuMP\nimport HiGHS\nmodel = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x)\n@objective(model, Min, x)\nsolutions = Pair{Int,Float64}[]\nmy_lock = Threads.ReentrantLock()\nThreads.@threads for i in 1:10\n    set_lower_bound(x, i)\n    optimize!(model)\n    Threads.lock(my_lock) do\n        push!(solutions, i => objective_value(model))\n    end\nend","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"This will not work, and attempting to do so may error, crash Julia or produce incorrect results.","category":"page"},{"location":"tutorials/algorithms/parallelism/#Using-parallelism-the-right-way","page":"Parallelism","title":"Using parallelism the right way","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"To use parallelism with JuMP, the simplest rule to remember is that each worker must have its own instance of a JuMP model.","category":"page"},{"location":"tutorials/algorithms/parallelism/#With-multi-threading","page":"Parallelism","title":"With multi-threading","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"With multi-threading, create a new instance of model in each iteration of the for-loop:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> using JuMP\n\njulia> import HiGHS\n\njulia> solutions = Pair{Int,Float64}[]\n\njulia> my_lock = Threads.ReentrantLock();\n\njulia> Threads.@threads for i in 1:10\n           model = Model(HiGHS.Optimizer)\n           set_silent(model)\n           set_attribute(model, MOI.NumberOfThreads(), 1)\n           @variable(model, x)\n           @objective(model, Min, x)\n           set_lower_bound(x, i)\n           optimize!(model)\n           Threads.lock(my_lock) do\n               push!(solutions, i => objective_value(model))\n           end\n       end\n\njulia> solutions\n10-element Vector{Pair{Int64, Float64}}:\n  7 => 7.0\n  4 => 4.0\n  1 => 1.0\n  9 => 9.0\n  5 => 5.0\n  8 => 8.0\n 10 => 10.0\n  2 => 2.0\n  6 => 6.0\n  3 => 3.0","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"warning: Warning\nFor some solvers, it may be necessary to limit the number of threads used internally by the solver to 1 by setting the MOI.NumberOfThreads attribute.","category":"page"},{"location":"tutorials/algorithms/parallelism/#With-distributed-computing","page":"Parallelism","title":"With distributed computing","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"With distributed computing, remember to evaluate all of the code on all of the processes using Distributed.@everywhere, and then write a function which creates a new instance of the model on every evaluation:","category":"page"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"julia> Distributed.@everywhere begin\n           using JuMP\n           import HiGHS\n       end\n\njulia> Distributed.@everywhere begin\n           function solve_model_with_right_hand_side(i)\n               model = Model(HiGHS.Optimizer)\n               set_silent(model)\n               @variable(model, x)\n               @objective(model, Min, x)\n               set_lower_bound(x, i)\n               optimize!(model)\n               return objective_value(model)\n           end\n       end\n\njulia> solutions = Distributed.pmap(solve_model_with_right_hand_side, 1:10)\n10-element Vector{Float64}:\n  1.0\n  2.0\n  3.0\n  4.0\n  5.0\n  6.0\n  7.0\n  8.0\n  9.0\n 10.0","category":"page"},{"location":"tutorials/algorithms/parallelism/#Other-types-of-parallelism","page":"Parallelism","title":"Other types of parallelism","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/#GPU","page":"Parallelism","title":"GPU","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"JuMP does not support GPU programming, and few solvers support execution on a GPU.","category":"page"},{"location":"tutorials/algorithms/parallelism/#Parallelism-within-the-solver","page":"Parallelism","title":"Parallelism within the solver","text":"","category":"section"},{"location":"tutorials/algorithms/parallelism/","page":"Parallelism","title":"Parallelism","text":"Many solvers use parallelism internally. For example, commercial solvers like Gurobi and CPLEX both parallelize the search in branch-and-bound. Solvers supporting internal parallelism will typically support the MOI.NumberOfThreads attribute, which you can set using set_attribute.","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/variables.md\"","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/manual/variables/#Variables","page":"Variables","title":"Variables","text":"","category":"section"},{"location":"moi/manual/variables/#Add-a-variable","page":"Variables","title":"Add a variable","text":"","category":"section"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"Use add_variable to add a single variable.","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"julia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"add_variable returns a VariableIndex type, which is used to refer to the added variable in other calls.","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"Check if a VariableIndex is valid using is_valid.","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"julia> MOI.is_valid(model, x)\ntrue","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"Use add_variables to add a number of variables.","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"julia> y = MOI.add_variables(model, 2)\n2-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"warning: Warning\nThe integer does not necessarily correspond to the column inside an optimizer.","category":"page"},{"location":"moi/manual/variables/#Delete-a-variable","page":"Variables","title":"Delete a variable","text":"","category":"section"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"Delete a variable using delete.","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"julia> MOI.delete(model, x)\n\njulia> MOI.is_valid(model, x)\nfalse","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"warning: Warning\nNot all ModelLike models support deleting variables. A DeleteNotAllowed error is thrown if this is not supported.","category":"page"},{"location":"moi/manual/variables/#Variable-attributes","page":"Variables","title":"Variable attributes","text":"","category":"section"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"The following attributes are available for variables:","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"VariableName\nVariablePrimalStart\nVariablePrimal","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"Get and set these attributes using get and set.","category":"page"},{"location":"moi/manual/variables/","page":"Variables","title":"Variables","text":"julia> MOI.set(model, MOI.VariableName(), x, \"var_x\")\n\njulia> MOI.get(model, MOI.VariableName(), x)\n\"var_x\"","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"EditURL = \"complementarity.jl\"","category":"page"},{"location":"tutorials/nonlinear/complementarity/#Mixed-complementarity-problems","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"","category":"section"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"This tutorial is a collection of mixed complementarity programs.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"using JuMP\nimport PATHSolver","category":"page"},{"location":"tutorials/nonlinear/complementarity/#Background","page":"Mixed complementarity problems","title":"Background","text":"","category":"section"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"A mixed complementarity problem has the form:","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"beginalign\n    F_i(x) perp x_i  i = 1 ldots n \n    l_i le x_i le u_i  i = 1 ldots n\nendalign","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"where the perp constraint enforces the following relations:","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"If l_i  x_i  u_i, then F_i(x) = 0\nIf l_i = x_i, then F_i(x) ge 0\nIf x_i = u_i, then F_i(x) le 0","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"You may have seen a complementarity problem written as 0 le F(x) perp x ge 0. This is a special case of a mixed complementarity problem in which l_i = 0 and u_i = infty.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"Importantly, a mixed complementarity problem does not have an objective, and no other constraint types are present.","category":"page"},{"location":"tutorials/nonlinear/complementarity/#Linear-complementarity","page":"Mixed complementarity problems","title":"Linear complementarity","text":"","category":"section"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"Form a mixed complementarity problem using the perp symbol ⟂ (type \\perp<tab> in the REPL).","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"M = [0 0 -1 -1; 0 0 1 -2; 1 -1 2 -2; 1 2 -2 4]\nq = [2, 2, -2, -6]\nmodel = Model(PATHSolver.Optimizer)\nset_silent(model)\n@variable(model, 0 <= x[1:4] <= 10, start = 0)\n@constraint(model, M * x + q ⟂ x)\noptimize!(model)\nvalue.(x)","category":"page"},{"location":"tutorials/nonlinear/complementarity/#Other-ways-of-writing-linear-complementarity-problems","page":"Mixed complementarity problems","title":"Other ways of writing linear complementarity problems","text":"","category":"section"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"You do not need to use a single vector of variables, and the complementarity constraints can be given in any order. In addition, you can use the perp symbol, the complements(F, x) syntax, or the MOI.Complements set.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"model = Model(PATHSolver.Optimizer)\nset_silent(model)\n@variable(model, 0 <= w <= 10, start = 0)\n@variable(model, 0 <= x <= 10, start = 0)\n@variable(model, 0 <= y <= 10, start = 0)\n@variable(model, 0 <= z <= 10, start = 0)\n@constraint(model, complements(y - 2z + 2, x))\n@constraint(model, [-y - z + 2, w] in MOI.Complements(2))\n@constraint(model, w + 2x - 2y + 4z - 6 ⟂ z)\n@constraint(model, w - x + 2y - 2z - 2 ⟂ y)\noptimize!(model)\nvalue.([w, x, y, z])","category":"page"},{"location":"tutorials/nonlinear/complementarity/#Transportation","page":"Mixed complementarity problems","title":"Transportation","text":"","category":"section"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"This example is a reformulation of the transportation problem from Chapter 3.3 of Dantzig, G.B. (1963). Linear Programming and Extensions. Princeton University Press, Princeton, New Jersey. It is based on the GAMS model gamslib_transmcp.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"capacity = Dict(\"seattle\" => 350, \"san-diego\" => 600)\ndemand = Dict(\"new-york\" => 325, \"chicago\" => 300, \"topeka\" => 275)\ncost = Dict(\n    (\"seattle\" => \"new-york\") => 90 * 2.5 / 1_000,\n    (\"seattle\" => \"chicago\") => 90 * 1.7 / 1_000,\n    (\"seattle\" => \"topeka\") => 90 * 1.8 / 1_000,\n    (\"san-diego\" => \"new-york\") => 90 * 2.5 / 1_000,\n    (\"san-diego\" => \"chicago\") => 90 * 1.8 / 1_000,\n    (\"san-diego\" => \"topeka\") => 90 * 1.4 / 1_000,\n)\nplants, markets = keys(capacity), keys(demand)\nmodel = Model(PATHSolver.Optimizer)\nset_silent(model)\n@variable(model, w[i in plants] >= 0)\n@variable(model, p[j in markets] >= 0)\n@variable(model, x[i in plants, j in markets] >= 0)\n@constraints(\n    model,\n    begin\n        [i in plants, j in markets], w[i] + cost[i=>j] - p[j] ⟂ x[i, j]\n        [i in plants], capacity[i] - sum(x[i, :]) ⟂ w[i]\n        [j in markets], sum(x[:, j]) - demand[j] ⟂ p[j]\n    end\n)\noptimize!(model)\nvalue.(p)","category":"page"},{"location":"tutorials/nonlinear/complementarity/#Expected-utility-of-insurance","page":"Mixed complementarity problems","title":"Expected utility of insurance","text":"","category":"section"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"This example is taken from a lecture of the course AAE706, given by Thomas F. Rutherford at the University of Wisconsin, Madison. It models the expected coverage of insurance K that a rational actor would obtain to insure a risk that occurs with probability pi and results in a loss of L.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"pi = 0.01  # Probability of a bad outcome\nL = 0.5    # Loss with a bad outcome\nγ = 0.02   # Premium for coverage\nσ = 0.5    # Elasticity\nρ = -1     # Risk exponent\nU(C) = C^ρ / ρ\nMU(C) = C^(ρ - 1)\nmodel = Model(PATHSolver.Optimizer)\nset_silent(model)\n@variable(model, EU, start = 1)   # Expected utilitiy\n@variable(model, EV, start = 1)   # Equivalent variation in income\n@variable(model, C_G, start = 1)  # Consumption on a good day\n@variable(model, C_B, start = 1)  # Consumption on a bad day\n@variable(model, K, start = 1)    # Coverage\n@constraints(\n    model,\n    begin\n        (1 - pi) * U(C_G) + pi * U(C_B) - EU ⟂ EU\n        100 * (((1 - pi) * C_G^ρ + pi * C_B^ρ)^(1 / ρ) - 1) - EV ⟂ EV\n        1 - γ * K - C_G ⟂ C_G\n        1 - L + (1 - γ) * K - C_B ⟂ C_B\n        γ * ((1 - pi) * MU(C_G) + pi * MU(C_B)) - pi * MU(C_B) ⟂ K\n    end\n)\noptimize!(model)\nvalue(K)","category":"page"},{"location":"tutorials/nonlinear/complementarity/#Electricity-consumption","page":"Mixed complementarity problems","title":"Electricity consumption","text":"","category":"section"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"This example is a mixed complementarity formulation of example 3.3.1 from D’Aertrycke, G., Ehrenmann, A., Ralph, D., & Smeers, Y. (2017). Risk trading in capacity equilibrium models.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"This example models a risk neutral competitive equilibrium between a producer and a consumer of electricity.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"In our example, we assume a producer is looking to invest in a new power plant with capacity x [MW]. This plant has an annualized capital cost of I [€/MW] and an operating cost of C [€/MWh]. There are 8760 hours in a year.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"After making the capital investment, there are five possible consumption scenarios, omega, which occur with probability theta_omega. In each scenario , the producer makes Y_ω MW of electricity.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"There is one consumer in the model, who has a quadratic utility function, U(Q_ω) = A_ω Q_ω + fracB_ω Q_ω^22.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"We now build and solve the mixed complementarity problem with a few brief comments. The economic justification for the model would require a larger tutorial than the space available here. Consult the original text for details.","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"I = 90_000                     # Annualized capital cost\nC = 60                         # Operation cost per MWh\nτ = 8_760                      # Hours per year\nθ = [0.2, 0.2, 0.2, 0.2, 0.2]  # Scenario probabilities\nA = [300, 350, 400, 450, 500]  # Utility function coefficients\nB = 1                          # Utility function coefficients\nmodel = Model(PATHSolver.Optimizer)\nset_silent(model)\n@variable(model, x >= 0, start = 1)           # Installed capacity\n@variable(model, Q[ω = 1:5] >= 0, start = 1)  # Consumption\n@variable(model, Y[ω = 1:5] >= 0, start = 1)  # Production\n@variable(model, P[ω = 1:5], start = 1)       # Electricity price\n@variable(model, μ[ω = 1:5] >= 0, start = 1)  # Capital scarcity margin\n# Unit investment cost equals annualized scarcity profit or investment is 0\n@constraint(model, I - τ * θ' * μ ⟂ x)\n# Difference between price and scarcity margin is equal to operation cost\n@constraint(model, [ω = 1:5], C - (P[ω] - μ[ω]) ⟂ Y[ω])\n# Price is equal to consumer's marginal utility\n@constraint(model, [ω = 1:5], P[ω] - (A[ω] - B * Q[ω]) ⟂ Q[ω])\n# Production is equal to consumption\n@constraint(model, [ω = 1:5], Y[ω] - Q[ω] ⟂ P[ω])\n# Production does not exceed capacity\n@constraint(model, [ω = 1:5], x - Y[ω] ⟂ μ[ω])\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"An equilibrium solution is to build 389 MW:","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"value(x)","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"The production in each scenario is:","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"value.(Q)","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"The price in each scenario is:","category":"page"},{"location":"tutorials/nonlinear/complementarity/","page":"Mixed complementarity problems","title":"Mixed complementarity problems","text":"value.(P)","category":"page"},{"location":"developers/custom_solver_binaries/#How-to-use-a-custom-binary","page":"Custom binaries","title":"How to use a custom binary","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Many solvers are not written in Julia, but instead in languages like C or C++. JuMP interacts with these solvers through binary dependencies.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"For many open-source solvers, we automatically install the appropriate binary when you run Pkg.add(\"Solver\"). For example, Pkg.add(\"ECOS\") will also install the ECOS binary.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"This page explains how this installation works, and how you can use a custom binary.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"compat: Compat\nThese instructions require Julia 1.6 or later.","category":"page"},{"location":"developers/custom_solver_binaries/#Background","page":"Custom binaries","title":"Background","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Each solver that JuMP supports is structured as a Julia package. For example, the interface for the ECOS solver is provided by the ECOS.jl package.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"tip: Tip\nThis page uses the example of ECOS.jl because it is simple to compile. Other solvers follow similar conventions. For example, the interface to the Clp solver is provided by Clp.jl.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"The ECOS.jl package provides an interface between the C API of ECOS and MathOptInterface. However, it does not handle the installation of the solver binary; that is the job for a JLL package.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"A JLL is a Julia package that wraps a pre-compiled binary. Binaries are built using Yggdrasil (for example, ECOS) and hosted in the JuliaBinaryWrappers GitHub repository (for example, ECOS_jll.jl).","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"JLL packages contain little code. Their only job is to dlopen a dynamic library, along with any dependencies.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"JLL packages manage their binary dependencies using Julia's artifact system. Each JLL package has an Artifacts.toml file which describes where to find each binary artifact for each different platform that it might be installed on. Here is the Artifacts.toml file for ECOS_jll.jl.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"The binaries installed by the JLL package should be sufficient for most users. In rare cases, however, you may require a custom binary. The two main reasons to use a custom binary are:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"You want a binary with custom compilation settings (for example, debugging)\nYou want a binary with a set of dependencies that are not available on Yggdrasil (for example, a commercial solver like Gurobi or CPLEX).","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"The following sections explain how to replace the binaries provided by a JLL package with the custom ones you have compiled. As a reminder, we use ECOS as an example for simplicity, but the steps are the same for other solvers.","category":"page"},{"location":"developers/custom_solver_binaries/#jll_structure","page":"Custom binaries","title":"Explore the JLL you want to override","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"The first step is to explore the structure and filenames of the JLL package we want to override.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Find the location of the files using .artifact_dir:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"julia> using ECOS_jll\n\njulia> ECOS_jll.artifact_dir\n\"/Users/oscar/.julia/artifacts/2addb75332eff5a1657b46bb6bf30d2410bc7ecf\"","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"tip: Tip\nThis path may be different on other machines.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Here is what it contains:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"julia> readdir(ECOS_jll.artifact_dir)\n4-element Vector{String}:\n \"include\"\n \"lib\"\n \"logs\"\n \"share\"\n\njulia> readdir(joinpath(ECOS_jll.artifact_dir, \"lib\"))\n1-element Vector{String}:\n \"libecos.dylib\"","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Other solvers may have a bin directory containing executables. To use a custom binary of ECOS, we need to replace /lib/libecos.dylib with our custom binary.","category":"page"},{"location":"developers/custom_solver_binaries/#compile_ecos","page":"Custom binaries","title":"Compile a custom binary","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"The next step is to compile a custom binary. Because ECOS is written in C with no dependencies, this is easy to do if you have a C compiler:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"oscar@Oscars-MBP jll_example % git clone https://github.com/embotech/ecos.git\n[... lines omitted ...]\noscar@Oscars-MBP jll_example % cd ecos\noscar@Oscars-MBP ecos % make shared\n[... many lines omitted...]\noscar@Oscars-MBP ecos % mkdir lib\noscar@Oscars-MBP ecos % cp libecos.dylib lib","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"warning: Warning\nCompiling custom solver binaries is an advanced operation. Due to the complexities of compiling various solvers, the JuMP community is unable to help you diagnose and fix compilation issues.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"After this compilation step, we now have a folder /tmp/jll_example/ecos that contains lib and include directories with the same files as ECOS_jll:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"julia> readdir(joinpath(\"ecos\", \"lib\"))\n1-element Vector{String}:\n \"libecos.dylib\"","category":"page"},{"location":"developers/custom_solver_binaries/#Overriding-a-single-library","page":"Custom binaries","title":"Overriding a single library","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"To override the libecos library, we need to know what ECOS_jll calls it. (In most cases, it will also be libecos, but not always.)","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"There are two ways you can check.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Check the bottom of the JLL's GitHub README. For example, ECOS_jll has a single LibraryProduct called libecos.\nType ECOS_jll. and the press the [TAB] key twice to auto-complete available options:\njulia> ECOS_jll.\nLIBPATH           PATH_list          best_wrapper       get_libecos_path   libecos_handle\nLIBPATH_list      __init__           dev_jll            is_available       libecos_path\nPATH              artifact_dir       find_artifact_dir  libecos\nHere you can see there is libecos, and more usefully for us, libecos_path.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Once you know the name of the variable to override (the one that ends in _path), use  Preferences.jl to specify a new path:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"using Preferences\nset_preferences!(\n    \"LocalPreferences.toml\",\n    \"ECOS_jll\",\n    \"libecos_path\" => \"/tmp/jll_example/ecos/lib/libecos\"\n)","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"This will create a file in your current directory called LocalPreferences.toml with the contents:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"[ECOS_jll]\nlibecos_path = \"/tmp/jll_example/ecos/lib/libecos\"","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Now if you restart Julia, you will see:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"julia> using ECOS_jll\n\njulia> ECOS_jll.libecos\n\"/tmp/jll_example/ecos/lib/libecos\"","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"To go back to using the default library, just delete the LocalPreferences.toml file.","category":"page"},{"location":"developers/custom_solver_binaries/#Overriding-an-entire-artifact","page":"Custom binaries","title":"Overriding an entire artifact","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Sometimes a solver may provide a number of libraries and executables, and specifying the path for each of the becomes tedious. In this case, we can use Julia's Override.toml to replace an entire artifact.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Overriding an entire artifact requires you to replicate the structure and contents of the JLL package that we explored above.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"In most cases you need only reproduce the include, lib, and bin directories (if they exist). You can safely ignore any logs or share directories. Take careful note of what files each directory contains and what they are called.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"For our ECOS example, we already reproduced the structure when we compiled ECOS.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"So, now we need to tell Julia to use our custom installation instead of the default. We can do this by making an override file at ~/.julia/artifacts/Overrides.toml.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Overrides.toml has the following content:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"# Override for ECOS_jll\n2addb75332eff5a1657b46bb6bf30d2410bc7ecf = \"/tmp/jll_example/ecos\"","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"where 2addb75332eff5a1657b46bb6bf30d2410bc7ecf is the folder from the original ECOS_jll.artifact_dir and \"/tmp/jll_example/ecos\" is the location of our new installation. Replace these as appropriate for your system.","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"If you restart Julia after creating the override file, you will see:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"julia> using ECOS_jll\n\njulia> ECOS_jll.artifact_dir\n\"/tmp/jll_example/ecos\"","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Now when we use ECOS it will use our custom binary.","category":"page"},{"location":"developers/custom_solver_binaries/#Using-Cbc-with-a-custom-binary","page":"Custom binaries","title":"Using Cbc with a custom binary","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"As a second example, we demonstrate how to use Cbc.jl with a custom binary.","category":"page"},{"location":"developers/custom_solver_binaries/#Explore-the-JLL-you-want-to-override","page":"Custom binaries","title":"Explore the JLL you want to override","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"First, let's check where Cbc_jll is installed:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"julia> using Cbc_jll\n\njulia> Cbc_jll.artifact_dir\n\"/Users/oscar/.julia/artifacts/e481bc81db5e229ba1f52b2b4bd57484204b1b06\"\n\njulia> readdir(Cbc_jll.artifact_dir)\n5-element Vector{String}:\n \"bin\"\n \"include\"\n \"lib\"\n \"logs\"\n \"share\"\n\njulia> readdir(joinpath(Cbc_jll.artifact_dir, \"bin\"))\n1-element Vector{String}:\n \"cbc\"\n\njulia> readdir(joinpath(Cbc_jll.artifact_dir, \"lib\"))\n10-element Vector{String}:\n \"libCbc.3.10.5.dylib\"\n \"libCbc.3.dylib\"\n \"libCbc.dylib\"\n \"libCbcSolver.3.10.5.dylib\"\n \"libCbcSolver.3.dylib\"\n \"libCbcSolver.dylib\"\n \"libOsiCbc.3.10.5.dylib\"\n \"libOsiCbc.3.dylib\"\n \"libOsiCbc.dylib\"\n \"pkgconfig\"","category":"page"},{"location":"developers/custom_solver_binaries/#Compile-a-custom-binary","page":"Custom binaries","title":"Compile a custom binary","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Next, we need to compile Cbc. Cbc can be difficult to compile (it has a lot of dependencies), but for macOS users there is a homebrew recipe:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"(base) oscar@Oscars-MBP jll_example % brew install cbc\n[ ... lines omitted ... ]\n(base) oscar@Oscars-MBP jll_example % brew list cbc\n/usr/local/Cellar/cbc/2.10.5/bin/cbc\n/usr/local/Cellar/cbc/2.10.5/include/cbc/ (76 files)\n/usr/local/Cellar/cbc/2.10.5/lib/libCbc.3.10.5.dylib\n/usr/local/Cellar/cbc/2.10.5/lib/libCbcSolver.3.10.5.dylib\n/usr/local/Cellar/cbc/2.10.5/lib/libOsiCbc.3.10.5.dylib\n/usr/local/Cellar/cbc/2.10.5/lib/pkgconfig/ (2 files)\n/usr/local/Cellar/cbc/2.10.5/lib/ (6 other files)\n/usr/local/Cellar/cbc/2.10.5/share/cbc/ (59 files)\n/usr/local/Cellar/cbc/2.10.5/share/coin/ (4 files)","category":"page"},{"location":"developers/custom_solver_binaries/#Override-single-libraries","page":"Custom binaries","title":"Override single libraries","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"To use Preferences.jl to override specific libraries we first check the names of each library in Cbc_jll:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"julia> Cbc_jll.\nLIBPATH               cbc                    get_libcbcsolver_path  libOsiCbc_path\nLIBPATH_list          cbc_path               is_available           libcbcsolver\nPATH                  dev_jll                libCbc                 libcbcsolver_handle\nPATH_list             find_artifact_dir      libCbc_handle          libcbcsolver_path\n__init__              get_cbc_path           libCbc_path\nartifact_dir          get_libCbc_path        libOsiCbc\nbest_wrapper          get_libOsiCbc_path     libOsiCbc_handle","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"Then we add the following to LocalPreferences.toml:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"[Cbc_jll]\ncbc_path = \"/usr/local/Cellar/cbc/2.10.5/bin/cbc\"\nlibCbc_path = \"/usr/local/Cellar/cbc/2.10.5/lib/libCbc.3.10.5\"\nlibOsiCbc_path = \"/usr/local/Cellar/cbc/2.10.5/lib/libOsiCbc.3.10.5\"\nlibcbcsolver_path = \"/usr/local/Cellar/cbc/2.10.5/lib/libCbcSolver.3.10.5\"","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"info: Info\nNote that capitalization matters, so libcbcsolver_path corresponds to libCbcSolver.3.10.5.","category":"page"},{"location":"developers/custom_solver_binaries/#Override-entire-artifact","page":"Custom binaries","title":"Override entire artifact","text":"","category":"section"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"To use the homebrew install as our custom binary we add the following to ~/.julia/artifacts/Overrides.toml:","category":"page"},{"location":"developers/custom_solver_binaries/","page":"Custom binaries","title":"Custom binaries","text":"# Override for Cbc_jll\ne481bc81db5e229ba1f52b2b4bd57484204b1b06 = \"/usr/local/Cellar/cbc/2.10.5\"","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"EditURL = \"benders_decomposition.jl\"","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/#benders_decomposition_classical","page":"Benders decomposition","title":"Benders decomposition","text":"","category":"section"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"This tutorial was originally contributed by Shuvomoy Das Gupta.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"This tutorial describes how to implement Benders decomposition in JuMP. It uses the following packages:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"using JuMP\nimport GLPK\nimport Printf","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/#Theory","page":"Benders decomposition","title":"Theory","text":"","category":"section"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Benders decomposition is a useful algorithm for solving convex optimization problems with a large number of variables. It works best when a larger problem can be decomposed into two (or more) smaller problems that are individually much easier to solve. This tutorial demonstrates Benders decomposition on the following mixed-integer linear program:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"beginaligned\ntextmin          c_1^top x+c_2^top y   \ntextsubject to   A_1 x+ A_2 y le b      \n                     x ge 0                 \n                     y ge 0                 \n                     x in mathbbZ^n\nendaligned","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"where b in mathbbR^m, A_1 in mathbbR^m times n, A_2 in mathbbR^m times p and mathbbZ is the set of integers.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Any mixed integer programming problem can be written in the form above.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"If there are relatively few integer variables, and many more continuous variables, then it may be beneficial to decompose the problem into a small problem containing only integer variables and a linear program containing only continuous variables. Hopefully, the linear program will be much easier to solve in isolation than in the full mixed-integer linear program.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"For example, if we knew a feasible solution for x, we could obtain a solution for y by solving:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"beginaligned\nV_2(x) =  textmin            c_2^top y                        \n          textsubject to     A_2 y le b - A_1 x  quad pi \n                                 y ge 0\nendaligned","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"where pi is the dual variable associated with the constraints. Because this is a linear program, it is easy to solve.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Replacing the c_2^top y component of the objective in our original problem with V_2 yields:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"beginaligned\ntextmin          c_1^top x + V_2(x) \ntextsubject to   x ge 0             \n                     x in mathbbZ^n\nendaligned","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"This problem looks a lot simpler to solve, but we need to do something else with V_2 first.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Because x is a constant that appears on the right-hand side of the constraints, V_2 is a convex function with respect to x, and the dual variable pi can be multiplied by -A_1 to obtain a subgradient of V_2(x) with respect to x. Therefore, if we have a candidate solution x_k, then we can solve V_2(x_k) and obtain a feasible dual vector pi_k. Using these values, we can construct a first-order Taylor-series approximation of V_2 about the point x_k:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"V_2(x) ge V_2(x_k) + -pi_k^top A_1 (x - x_k)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"By convexity, we know that this inequality holds for all x, and we call these inequalities cuts.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Benders decomposition is an iterative technique that replaces V_2(x) with a new decision variable theta, and approximates it from below using cuts:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"beginaligned\nV_1^K =  textmin           c_1^top x + theta  \n          textsubject to   x ge 0              \n                              x in mathbbZ^n   \n                              theta ge M         \n                              theta ge V_2(x_k) + pi_k^top(x - x_k)  quad forall k = 1ldotsK\nendaligned","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"This integer program is called the first-stage subproblem.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"To generate cuts, we solve V_1^K to obtain a candidate first-stage solution x_k, then we use that solution to solve V_2(x_k). Then, using the optimal objective value and dual solution from V_2, we add a new cut to form V_1^K+1 and repeat.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/#Bounds","page":"Benders decomposition","title":"Bounds","text":"","category":"section"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Due to convexity, we know that V_2(x) ge theta for all x. Therefore, the optimal objective value of V_1^K provides a valid lower bound on the objective value of the full problem. In addition, if we take a feasible solution for x from the first-stage problem, then c_1^top x + V_2(x) is a valid upper bound on the objective value of the full problem.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Benders decomposition uses the lower and upper bounds to determine when it has found the global optimal solution.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/#Input-data","page":"Benders decomposition","title":"Input data","text":"","category":"section"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"As an example for this tutorial, we use the input data is from page 139 of Garfinkel, R. & Nemhauser, G. L. Integer programming. (Wiley, 1972).","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"c_1 = [1, 4]\nc_2 = [2, 3]\ndim_x = length(c_1)\ndim_y = length(c_2)\nb = [-2; -3]\nA_1 = [1 -3; -1 -3]\nA_2 = [1 -2; -1 -1]\nM = -1000;\nnothing #hide","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/#benders_iterative","page":"Benders decomposition","title":"Iterative method","text":"","category":"section"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"warning: Warning\nThis is a basic implementation for pedagogical purposes. We haven't discussed Benders feasibility cuts, or any of the computational tricks that are required to build a performant implementation for large-scale problems. See In-place iterative method for one improvement that helps computation time.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"We start by formulating the first-stage subproblem:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"model = Model(GLPK.Optimizer)\n@variable(model, x[1:dim_x] >= 0, Int)\n@variable(model, θ >= M)\n@objective(model, Min, c_1' * x + θ)\nprint(model)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"For the next step, we need a function that takes a first-stage candidate solution x and returns the optimal solution from the second-stage subproblem:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"function solve_subproblem(x)\n    model = Model(GLPK.Optimizer)\n    @variable(model, y[1:dim_y] >= 0)\n    con = @constraint(model, A_2 * y .<= b - A_1 * x)\n    @objective(model, Min, c_2' * y)\n    optimize!(model)\n    @assert termination_status(model) == OPTIMAL\n    return (obj = objective_value(model), y = value.(y), π = dual.(con))\nend","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Note that solve_subproblem returns a NamedTuple of the objective value, the optimal primal solution for y, and the optimal dual solution for π.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"We're almost ready for our optimization loop, but first, here's a helpful function for logging:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"function print_iteration(k, args...)\n    f(x) = Printf.@sprintf(\"%12.4e\", x)\n    println(lpad(k, 9), \" \", join(f.(args), \" \"))\n    return\nend","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"We also need to put a limit on the number of iterations before termination:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"MAXIMUM_ITERATIONS = 100","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"And a way to check if the lower and upper bounds are close-enough to terminate:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"ABSOLUTE_OPTIMALITY_GAP = 1e-6","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Now we're ready to iterate Benders decomposition:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"println(\"Iteration  Lower Bound  Upper Bound          Gap\")\nfor k in 1:MAXIMUM_ITERATIONS\n    optimize!(model)\n    lower_bound = objective_value(model)\n    x_k = value.(x)\n    ret = solve_subproblem(x_k)\n    upper_bound = c_1' * x_k + ret.obj\n    gap = (upper_bound - lower_bound) / upper_bound\n    print_iteration(k, lower_bound, upper_bound, gap)\n    if gap < ABSOLUTE_OPTIMALITY_GAP\n        println(\"Terminating with the optimal solution\")\n        break\n    end\n    cut = @constraint(model, θ >= ret.obj + -ret.π' * A_1 * (x .- x_k))\n    @info \"Adding the cut $(cut)\"\nend","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Finally, we can obtain the optimal solution","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"optimize!(model)\nx_optimal = value.(x)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"optimal_ret = solve_subproblem(x_optimal)\ny_optimal = optimal_ret.y","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/#Callback-method","page":"Benders decomposition","title":"Callback method","text":"","category":"section"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"The Iterative method section implemented Benders decomposition using a loop. In each iteration, we re-solved the first-stage subproblem to generate a candidate solution. However, modern MILP solvers such as CPLEX, Gurobi, and GLPK provide lazy constraint callbacks which allow us to add new cuts while the solver is running. This can be more efficient than an iterative method because we can avoid repeating work such as solving the root node of the first-stage MILP at each iteration.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"tip: Tip\nFor more information on callbacks, read the page Solver-independent callbacks.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"As before, we construct the same first-stage subproblem:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"lazy_model = Model(GLPK.Optimizer)\n@variable(lazy_model, x[1:dim_x] >= 0, Int)\n@variable(lazy_model, θ >= M)\n@objective(lazy_model, Min, θ)\nprint(lazy_model)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"What differs is that we write a callback function instead of a loop:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"k = 0\n\n\"\"\"\n    my_callback(cb_data)\n\nA callback that implements Benders decomposition. Note how similar it is to the\ninner loop of the iterative method.\n\"\"\"\nfunction my_callback(cb_data)\n    global k += 1\n    x_k = callback_value.(cb_data, x)\n    θ_k = callback_value(cb_data, θ)\n    lower_bound = c_1' * x_k + θ_k\n    ret = solve_subproblem(x_k)\n    upper_bound = c_1' * x_k + c_2' * ret.y\n    gap = (upper_bound - lower_bound) / upper_bound\n    print_iteration(k, lower_bound, upper_bound, gap)\n    if gap < ABSOLUTE_OPTIMALITY_GAP\n        println(\"Terminating with the optimal solution\")\n        return\n    end\n    cut = @build_constraint(θ >= ret.obj + -ret.π' * A_1 * (x .- x_k))\n    MOI.submit(model, MOI.LazyConstraint(cb_data), cut)\n    return\nend\n\nset_attribute(lazy_model, MOI.LazyConstraintCallback(), my_callback)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Now when we optimize!, our callback is run:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"optimize!(lazy_model)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Note how this problem also takes 4 iterations to converge, but the sequence of bounds is different compared to the iterative method.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Finally, we can obtain the optimal solution:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"x_optimal = value.(x)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"optimal_ret = solve_subproblem(x_optimal)\ny_optimal = optimal_ret.y","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/#In-place-iterative-method","page":"Benders decomposition","title":"In-place iterative method","text":"","category":"section"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Our implementation of the iterative method has a problem: every time we need to solve the subproblem, we must rebuild it from scratch. This is expensive, and it can be the bottleneck in the solution process. We can improve our implementation by using re-using the subproblem between solves.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"First, we create our first-stage problem as usual:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"model = Model(GLPK.Optimizer)\n@variable(model, x[1:dim_x] >= 0, Int)\n@variable(model, θ >= M)\n@objective(model, Min, c_1' * x + θ)\nprint(model)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Then, instead of building the subproblem in a function, we build it once here:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"subproblem = Model(GLPK.Optimizer)\n@variable(subproblem, x_copy[1:dim_x])\n@variable(subproblem, y[1:dim_y] >= 0)\n@constraint(subproblem, A_1 * x_copy + A_2 * y .<= b)\n@objective(subproblem, Min, c_2' * y)\nprint(subproblem)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"This formulation is slightly different. We have included a copy of the x variables, x_copy, and used x_copy in the left-hand side of the constraints.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Our function to solve the subproblem is also slightly different. First, we need to fix the value of the x_copy variables to the value of x from the first-stage problem, and second, we compute the dual using the reduced_cost of x_copy, not the dual of the linear constraints:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"function solve_subproblem(model, x)\n    fix.(model[:x_copy], x)\n    optimize!(model)\n    @assert termination_status(model) == OPTIMAL\n    return (\n        obj = objective_value(model),\n        y = value.(model[:y]),\n        π = reduced_cost.(model[:x_copy]),\n    )\nend","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Now we're ready to iterate our in-place Benders decomposition, but this time the cut computation is slightly different. Because we used reduced_cost on the x_copy variables, the value of π is a valid subgradient on the objective of subproblem with respect to x. Therefore, we don't need to multiply the duals by -A_1, and so our cut uses ret.π' instead of -ret.π' * A_1:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"println(\"Iteration  Lower Bound  Upper Bound          Gap\")\nfor k in 1:MAXIMUM_ITERATIONS\n    optimize!(model)\n    lower_bound = objective_value(model)\n    x_k = value.(x)\n    ret = solve_subproblem(subproblem, x_k)\n    upper_bound = c_1' * x_k + ret.obj\n    gap = (upper_bound - lower_bound) / upper_bound\n    print_iteration(k, lower_bound, upper_bound, gap)\n    if gap < ABSOLUTE_OPTIMALITY_GAP\n        println(\"Terminating with the optimal solution\")\n        break\n    end\n    cut = @constraint(model, θ >= ret.obj + ret.π' * (x .- x_k))\n    @info \"Adding the cut $(cut)\"\nend","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"Finally, we can obtain the optimal solution:","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"optimize!(model)\nx_optimal = value.(x)","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"optimal_ret = solve_subproblem(subproblem, x_optimal)\ny_optimal = optimal_ret.y","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"For larger problems, the benefit of re-using the same subproblem and not needing to multiply the duals by A_1 in the cut coefficient usually outweights the cost of needing a copy of the x variables in the subproblem.","category":"page"},{"location":"tutorials/algorithms/benders_decomposition/","page":"Benders decomposition","title":"Benders decomposition","text":"As a secondary benefit, because we no longer need an explicit representation of A_1 in the cut, we can build the model and subproblem formulations using arbitrary JuMP syntax; they do not need to be in matrix form.","category":"page"},{"location":"tutorials/nonlinear/introduction/#Introduction","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"tutorials/nonlinear/introduction/","page":"Introduction","title":"Introduction","text":"Nonlinear programs (NLPs) are a class of optimization problems in which some of the constraints or the objective function are nonlinear:","category":"page"},{"location":"tutorials/nonlinear/introduction/","page":"Introduction","title":"Introduction","text":"beginalign\n    min_x in mathbbR^n  f_0(x) \n    textst  l_j le f_j(x) le u_j  j = 1 ldots m \n     l_i le x_i le u_i  i = 1 ldots n\nendalign","category":"page"},{"location":"tutorials/nonlinear/introduction/","page":"Introduction","title":"Introduction","text":"Mixed-integer nonlinear linear programs (MINLPs) are extensions of nonlinear programs in which some (or all) of the decision variables take discrete values.","category":"page"},{"location":"tutorials/nonlinear/introduction/#How-to-choose-a-solver","page":"Introduction","title":"How to choose a solver","text":"","category":"section"},{"location":"tutorials/nonlinear/introduction/","page":"Introduction","title":"Introduction","text":"JuMP supports a range of nonlinear solvers; look for \"NLP\" in the list of Supported solvers. However, very few solvers support mixed-integer nonlinear linear programs. Solvers supporting discrete variables start with \"(MI)\" in the list of Supported solvers.","category":"page"},{"location":"tutorials/nonlinear/introduction/","page":"Introduction","title":"Introduction","text":"If the only nonlinearities in your model are quadratic terms (that is, multiplication between two decision variables), you can also use second-order cone solvers, which are indicated by \"SOCP.\" In most cases, these solvers are restricted to convex quadratic problems and will error if you pass a nonconvex quadratic function; however, Gurobi has the ability to solve nonconvex quadratic terms.","category":"page"},{"location":"tutorials/nonlinear/introduction/#How-these-tutorials-are-structured","page":"Introduction","title":"How these tutorials are structured","text":"","category":"section"},{"location":"tutorials/nonlinear/introduction/","page":"Introduction","title":"Introduction","text":"Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.","category":"page"},{"location":"tutorials/nonlinear/introduction/","page":"Introduction","title":"Introduction","text":"The following tutorials are worked examples that present a problem in words, then formulate it in mathematics, and then solve it in JuMP. This usually involves some sort of visualization of the solution. Start here if you are new to JuMP.\nRocket Control\nOptimal control for a Space Shuttle reentry trajectory\nPortfolio optimization\nThe Tips and tricks tutorial contains a number of helpful reformulations and tricks you can use when modeling nonlinear programs. Look here if you are stuck trying to formulate a problem as a nonlinear program.\nThe Computing Hessians is an advanced tutorial which explains how to compute the Hessian of the Lagrangian of a nonlinear program. This is useful only in particular cases.\nThe remaining tutorials are less verbose and styled in the form of short code examples. These tutorials have less explanation, but may contain useful code snippets, particularly if they are similar to a problem you are trying to solve.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"EditURL = \"https://github.com/jump-dev/Pajarito.jl/blob/4c7efa915bdc900b51d1c8290dc771403b11230a/README.md\"","category":"page"},{"location":"packages/Pajarito/#Pajarito.jl","page":"jump-dev/Pajarito.jl","title":"Pajarito.jl","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Pajarito is a mixed-integer convex programming (MICP) solver package written in Julia.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"MICP problems are convex except for restrictions that some variables take binary or integer values.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Pajarito solves MICP problems in conic form, by constructing sequential polyhedral outer approximations of the conic feasible set.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"The underlying algorithm has theoretical finite-time convergence under reasonable assumptions.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Pajarito accesses state-of-the-art MILP solvers and continuous conic solvers through MathOptInterface.","category":"page"},{"location":"packages/Pajarito/#Pavito","page":"jump-dev/Pajarito.jl","title":"Pavito","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"For algorithms that use a derivative-based nonlinear programming (NLP) solver (for example, Ipopt) instead of a conic solver, use Pavito.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Pavito is a convex mixed-integer nonlinear programming (convex MINLP) solver. Because Pavito relies on gradient cuts, it can fail near points of non-differentiability. Pajarito may be more robust than Pavito on non-smooth problems.","category":"page"},{"location":"packages/Pajarito/#License","page":"jump-dev/Pajarito.jl","title":"License","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Pajarito.jl is licensed under the MPL 2.0 license.","category":"page"},{"location":"packages/Pajarito/#Installation","page":"jump-dev/Pajarito.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Install Pajarito using Pkg.add:","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"import Pkg\nPkg.add(\"Pajarito\")","category":"page"},{"location":"packages/Pajarito/#MIP-and-continuous-solvers","page":"jump-dev/Pajarito.jl","title":"MIP and continuous solvers","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"The algorithm implemented by Pajarito itself is relatively simple, and most of the hard work is performed by the MIP outer approximation (OA) solver and the continuous conic solver.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Therefore, in addition to installing Pajarito, you must also install a mixed-integer linear programming solver and a continuous conic solver.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"The performance of Pajarito depends on these two types of solvers.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"The OA solver (typically a mixed-integer linear solver) is specified by the oa_solver option. You must first load the Julia package that provides this solver, for example, using Gurobi. The continuous conic solver is specified by the conic_solver option.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"See JuMP's list of supported solvers.","category":"page"},{"location":"packages/Pajarito/#Use-with-JuMP","page":"jump-dev/Pajarito.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"To use Pajarito with JuMP, use:","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"using JuMP, Pajarito, HiGHS, Hypatia\nmodel = Model(\n    optimizer_with_attributes(\n        Pajarito.Optimizer,\n        \"oa_solver\" => optimizer_with_attributes(\n            HiGHS.Optimizer,\n            MOI.Silent() => true,\n            \"mip_feasibility_tolerance\" => 1e-8,\n            \"mip_rel_gap\" => 1e-6,\n        ),\n        \"conic_solver\" =>\n            optimizer_with_attributes(Hypatia.Optimizer, MOI.Silent() => true),\n    )\n)\nset_attribute(model, \"time_limit\", 60)","category":"page"},{"location":"packages/Pajarito/#Options","page":"jump-dev/Pajarito.jl","title":"Options","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"We list Pajarito's options below.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"verbose::Bool toggles printing\ntol_feas::Float64 is the feasibility tolerance for conic constraints\ntol_rel_gap::Float64 is the relative optimality gap tolerance\ntol_abs_gap::Float64 is the absolute optimality gap tolerance\ntime_limit::Float64 sets the time limit (in seconds)\niteration_limit::Int sets the iteration limit (for the iterative method)\nuse_iterative_method::Union{Nothing,Bool} toggles the iterative algorithm; if nothing is specified, Pajarito defaults to the OA-solver-driven (single tree) algorithm if lazy callbacks are supported by the OA solver\nuse_extended_form::Bool toggles the use of extended formulations for the second-order cone\nsolve_relaxation::Bool toggles solution of the continuous conic relaxation\nsolve_subproblems::Bool toggles solution of the continuous conic subproblems\nuse_init_fixed_oa::Bool toggles initial fixed OA cuts\noa_solver::Union{Nothing,MOI.OptimizerWithAttributes} is the OA solver\nconic_solver::Union{Nothing,MOI.OptimizerWithAttributes} is the conic solver","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Pajarito may require tuning of parameters to improve convergence.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"For example, it often helps to tighten the OA solver's integrality tolerance. OA solver and conic solver options must be specified directly to those solvers.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Note: if solve_subproblems is true, Pajarito usually returns a solution constructed from one of the conic solver's feasible solutions; since the conic solver is not subject to the same feasibility tolerances as the OA solver, Pajarito's solution will not necessarily satisfy tol_feas.","category":"page"},{"location":"packages/Pajarito/#Cone-interface","page":"jump-dev/Pajarito.jl","title":"Cone interface","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Pajarito has a generic cone interface (see the cones folder that allows the user to add support for new convex cones.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"To illustrate, in the experimental package PajaritoExtras we have extended Pajarito by adding support for several cones recognized by Hypatia.jl (a continuous conic solver with its own generic cone interface).","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"The examples folder of PajaritoExtras also contains many applied mixed-integer convex problems that are solved using Pajarito.","category":"page"},{"location":"packages/Pajarito/#Bug-reports-and-support","page":"jump-dev/Pajarito.jl","title":"Bug reports and support","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Please report any issues via the GitHub issue tracker.","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"All types of issues are welcome and encouraged; this includes bug reports, documentation typos, and feature requests. The Optimization (Mathematical) category on Discourse is appropriate for general discussion.","category":"page"},{"location":"packages/Pajarito/#References","page":"jump-dev/Pajarito.jl","title":"References","text":"","category":"section"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"If you find Pajarito useful in your work, we kindly request that you cite the following paper (arXiv preprint), which is recommended reading for advanced users:","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"@article{CoeyLubinVielma2020,\n    title={Outer approximation with conic certificates for mixed-integer convex problems},\n    author={Coey, Chris and Lubin, Miles and Vielma, Juan Pablo},\n    journal={Mathematical Programming Computation},\n    volume={12},\n    number={2},\n    pages={249--293},\n    year={2020},\n    publisher={Springer}\n}","category":"page"},{"location":"packages/Pajarito/","page":"jump-dev/Pajarito.jl","title":"jump-dev/Pajarito.jl","text":"Note this paper describes a legacy MathProgBase version of Pajarito, which is available on the mathprogbase branch of this repository. Starting with version v0.8.0, Pajarito supports MathOptInterface instead of MathProgBase.","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/models.md\"","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/manual/models/#Models","page":"Models","title":"Models","text":"","category":"section"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"The most significant part of MOI is the definition of the model API that is used to specify an instance of an optimization problem (for example, by adding variables and constraints). Objects that implement the model API must inherit from the ModelLike abstract type.","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"Notably missing from the model API is the method to solve an optimization problem. ModelLike objects may store an instance (for example, in memory or backed by a file format) without being linked to a particular solver. In addition to the model API, MOI defines AbstractOptimizer and provides methods to solve the model and interact with solutions. See the Solutions section for more details.","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"info: Info\nThroughout the rest of the manual, model is used as a generic ModelLike, and optimizer is used as a generic AbstractOptimizer.","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"tip: Tip\nMOI does not export functions, but for brevity we often omit qualifying names with the MOI module. Best practice is to haveimport MathOptInterface as MOIand prefix all MOI methods with MOI. in user code. If a name is also available in base Julia, we always explicitly use the module prefix, for example, with MOI.get.","category":"page"},{"location":"moi/manual/models/#Attributes","page":"Models","title":"Attributes","text":"","category":"section"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"Attributes are properties of the model that can be queried and modified. These include constants such as the number of variables in a model NumberOfVariables), and properties of variables and constraints such as the name of a variable (VariableName).","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"There are four types of attributes:","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"Model attributes (subtypes of AbstractModelAttribute) refer to properties of a model.\nOptimizer attributes (subtypes of AbstractOptimizerAttribute) refer to properties of an optimizer.\nConstraint attributes (subtypes of AbstractConstraintAttribute) refer to properties of an individual constraint.\nVariable attributes (subtypes of AbstractVariableAttribute) refer to properties of an individual variable.","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"Some attributes are values that can be queried by the user but not modified, while other attributes can be modified by the user.","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"All interactions with attributes occur through the get and set functions.","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"Consult the docstrings of each attribute for information on what it represents.","category":"page"},{"location":"moi/manual/models/#ModelLike-API","page":"Models","title":"ModelLike API","text":"","category":"section"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"The following attributes are available:","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"ListOfConstraintAttributesSet\nListOfConstraintIndices\nListOfConstraintTypesPresent\nListOfModelAttributesSet\nListOfVariableAttributesSet\nListOfVariableIndices\nNumberOfConstraints\nNumberOfVariables\nName\nObjectiveFunction\nObjectiveFunctionType\nObjectiveSense","category":"page"},{"location":"moi/manual/models/#AbstractOptimizer-API","page":"Models","title":"AbstractOptimizer API","text":"","category":"section"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"The following attributes are available:","category":"page"},{"location":"moi/manual/models/","page":"Models","title":"Models","text":"DualStatus\nPrimalStatus\nRawStatusString\nResultCount\nTerminationStatus\nBarrierIterations\nDualObjectiveValue\nNodeCount\nNumberOfThreads\nObjectiveBound\nObjectiveValue\nRelativeGap\nRawOptimizerAttribute\nRawSolver\nSilent\nSimplexIterations\nSolverName\nSolverVersion\nSolveTimeSec\nTimeLimitSec\nObjectiveLimit","category":"page"},{"location":"packages/NLopt/#NLopt.jl","page":"JuliaOpt/NLopt.jl","title":"NLopt.jl","text":"","category":"section"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"NLopt.jl is a wrapper for the NLopt library.","category":"page"},{"location":"packages/NLopt/#License","page":"JuliaOpt/NLopt.jl","title":"License","text":"","category":"section"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"NLopt.jl is licensed under the MIT License.","category":"page"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"The underlying solver, stevengj/nlopt, is licensed under the LGPL v3.0 license.","category":"page"},{"location":"packages/NLopt/#Installation","page":"JuliaOpt/NLopt.jl","title":"Installation","text":"","category":"section"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"Install NLopt.jl using the Julia package manager:","category":"page"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"import Pkg\nPkg.add(\"NLopt\")","category":"page"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"In addition to installing the NLopt.jl package, this will also download and install the NLopt binaries. You do not need to install NLopt separately.","category":"page"},{"location":"packages/NLopt/#Use-with-JuMP","page":"JuliaOpt/NLopt.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"You can use NLopt with JuMP as follows:","category":"page"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"using JuMP, NLopt\nmodel = Model(NLopt.Optimizer)\nset_attribute(model, \"algorithm\", :LD_MMA)","category":"page"},{"location":"packages/NLopt/#Options","page":"JuliaOpt/NLopt.jl","title":"Options","text":"","category":"section"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"The algorithm attribute is required. The value must be one of the supported NLopt algorithms.","category":"page"},{"location":"packages/NLopt/#Documentation","page":"JuliaOpt/NLopt.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/NLopt/","page":"JuliaOpt/NLopt.jl","title":"JuliaOpt/NLopt.jl","text":"For more details, see the NLopt.jl README or the NLopt documentation.","category":"page"},{"location":"tutorials/linear/introduction/#Introduction","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"tutorials/linear/introduction/","page":"Introduction","title":"Introduction","text":"Linear programs (LPs) are a fundamental class of optimization problems of the form:","category":"page"},{"location":"tutorials/linear/introduction/","page":"Introduction","title":"Introduction","text":"beginalign\n    min_x in mathbbR^n  sumlimits_i=1^n c_i x_i \n    textst  l_j le sumlimits_i=1^n a_ij x_i le u_j  j = 1 ldots m \n     l_i le x_i le u_i  i = 1 ldots n\nendalign","category":"page"},{"location":"tutorials/linear/introduction/","page":"Introduction","title":"Introduction","text":"The most important thing to note is that all terms are of the form coefficient * variable, and that there are no nonlinear terms or multiplications between variables.","category":"page"},{"location":"tutorials/linear/introduction/","page":"Introduction","title":"Introduction","text":"Mixed-integer linear programs (MILPs) are extensions of linear programs in which some (or all) of the decision variables take discrete values.","category":"page"},{"location":"tutorials/linear/introduction/#How-to-choose-a-solver","page":"Introduction","title":"How to choose a solver","text":"","category":"section"},{"location":"tutorials/linear/introduction/","page":"Introduction","title":"Introduction","text":"Almost all solvers support linear programs; look for \"LP\" in the list of Supported solvers. However, fewer solvers support mixed-integer linear programs. Solvers supporting discrete variables start with \"(MI)\" in the list of Supported solvers.","category":"page"},{"location":"tutorials/linear/introduction/#How-these-tutorials-are-structured","page":"Introduction","title":"How these tutorials are structured","text":"","category":"section"},{"location":"tutorials/linear/introduction/","page":"Introduction","title":"Introduction","text":"Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.","category":"page"},{"location":"tutorials/linear/introduction/","page":"Introduction","title":"Introduction","text":"The following tutorials are worked examples that present a problem in words, then formulate it in mathematics, and then solve it in JuMP. This usually involves some sort of visualization of the solution. Start here if you are new to JuMP.\nThe diet problem\nThe cannery problem\nThe facility location problem\nFinancial modeling problems\nNetwork flow problems\nN-Queens\nSudoku\nThe Tips and tricks tutorial contains a number of helpful reformulations and tricks you can use when modeling linear programs. Look here if you are stuck trying to formulate a problem as a linear program.\nThe Sensitivity analysis of a linear program tutorial explains how to create sensitivity reports like those produced by the Excel Solver.\nThe Callbacks tutorial explains how to write a variety of solver-independent callbacks. Look here if you want to write a callback.\nThe remaining tutorials are less verbose and styled in the form of short code examples. These tutorials have less explanation, but may contain useful code snippets, particularly if they are similar to a problem you are trying to solve.","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"EditURL = \"tips_and_tricks.jl\"","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/#nonlinear_tips_and_tricks","page":"Tips and tricks","title":"Tips and tricks","text":"","category":"section"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This example collates some tips and tricks you can use when formulating nonlinear programs. It uses the following packages:","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"using JuMP\nimport Ipopt\nimport Test","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/#User-defined-operators-with-vector-outputs","page":"Tips and tricks","title":"User-defined operators with vector outputs","text":"","category":"section"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"A common situation is to have a user-defined operator like the following that returns multiple outputs (we define function_calls to keep track of how many times we call this method):","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"function_calls = 0\nfunction foo(x, y)\n    global function_calls += 1\n    common_term = x^2 + y^2\n    term_1 = sqrt(1 + common_term)\n    term_2 = common_term\n    return term_1, term_2\nend","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"For example, the first term might be used in the objective, and the second term might be used in a constraint, and often they share work that is expensive to evaluate.","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"This is a problem for JuMP, because it requires user-defined operators to return a single number. One option is to define two separate functions, the first returning the first argument, and the second returning the second argument.","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"foo_1(x, y) = foo(x, y)[1]\nfoo_2(x, y) = foo(x, y)[2]","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"However, if the common term is expensive to compute, this approach is wasteful because it will evaluate the expensive term twice. Let's have a look at how many times we evaluate x^2 + y^2 during a solve:","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model(Ipopt.Optimizer)\nset_silent(model)\n@variable(model, x[1:2] >= 0, start = 0.1)\n@operator(model, op_foo_1, 2, foo_1)\n@operator(model, op_foo_2, 2, foo_2)\n@objective(model, Max, op_foo_1(x[1], x[2]))\n@constraint(model, op_foo_2(x[1], x[2]) <= 2)\nfunction_calls = 0\noptimize!(model)\nTest.@test objective_value(model) ≈ √3 atol = 1e-4\nTest.@test value.(x) ≈ [1.0, 1.0] atol = 1e-4\nprintln(\"Naive approach: function calls = $(function_calls)\")","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"An alternative approach is to use memoization, which uses a cache to store the result of function evaluations. We can write a memoization function as follows:","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"\"\"\"\n    memoize(foo::Function, n_outputs::Int)\n\nTake a function `foo` and return a vector of length `n_outputs`, where element\n`i` is a function that returns the equivalent of `foo(x...)[i]`.\n\nTo avoid duplication of work, cache the most-recent evaluations of `foo`.\nBecause `foo_i` is auto-differentiated with ForwardDiff, our cache needs to\nwork when `x` is a `Float64` and a `ForwardDiff.Dual`.\n\"\"\"\nfunction memoize(foo::Function, n_outputs::Int)\n    last_x, last_f = nothing, nothing\n    last_dx, last_dfdx = nothing, nothing\n    function foo_i(i, x::T...) where {T<:Real}\n        if T == Float64\n            if x !== last_x\n                last_x, last_f = x, foo(x...)\n            end\n            return last_f[i]::T\n        else\n            if x !== last_dx\n                last_dx, last_dfdx = x, foo(x...)\n            end\n            return last_dfdx[i]::T\n        end\n    end\n    return [(x...) -> foo_i(i, x...) for i in 1:n_outputs]\nend","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Let's see how it works. First, construct the memoized versions of foo:","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"memoized_foo = memoize(foo, 2)","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Now try evaluating the first element of memoized_foo.","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"function_calls = 0\nmemoized_foo[1](1.0, 1.0)\nprintln(\"function_calls = \", function_calls)","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"As expected, this evaluated the function once. However, if we call the function again, we hit the cache instead of needing to re-compute foo and function_calls is still 1!","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"memoized_foo[1](1.0, 1.0)\nprintln(\"function_calls = \", function_calls)","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Now let's see how this works during a real solve:","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"model = Model(Ipopt.Optimizer)\nset_silent(model)\n@variable(model, x[1:2] >= 0, start = 0.1)\n@operator(model, op_foo_1, 2, memoized_foo[1])\n@operator(model, op_foo_2, 2, memoized_foo[2])\n@objective(model, Max, op_foo_1(x[1], x[2]))\n@constraint(model, op_foo_2(x[1], x[2]) <= 2)\nfunction_calls = 0\noptimize!(model)\nTest.@test objective_value(model) ≈ √3 atol = 1e-4\nTest.@test value.(x) ≈ [1.0, 1.0] atol = 1e-4\nprintln(\"Memoized approach: function_calls = $(function_calls)\")","category":"page"},{"location":"tutorials/nonlinear/tips_and_tricks/","page":"Tips and tricks","title":"Tips and tricks","text":"Compared to the naive approach, the memoized approach requires half as many function evaluations.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"EditURL = \"rocket_control.jl\"","category":"page"},{"location":"tutorials/nonlinear/rocket_control/#Rocket-Control","page":"Rocket Control","title":"Rocket Control","text":"","category":"section"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"This tutorial was originally contributed by Iain Dunning.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"The purpose of this tutorial is to demonstrate how to setup and solve a nonlinear optimization problem.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"The example is an optimal control problem of a nonlinear rocket.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"using JuMP\nimport Ipopt\nimport Plots","category":"page"},{"location":"tutorials/nonlinear/rocket_control/#Overview","page":"Rocket Control","title":"Overview","text":"","category":"section"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Our goal is to maximize the final altitude of a vertically launched rocket.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"We can control the thrust of the rocket, and must take account of the rocket mass, fuel consumption rate, gravity, and aerodynamic drag.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Let us consider the basic description of the model (for the full description, including parameters for the rocket, see COPS3).","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"There are three state variables in our model:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Velocity: x_v(t)\nAltitude: x_h(t)\nMass of rocket and remaining fuel, x_m(t)","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"and a single control variable:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Thrust: u_t(t).","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"There are three equations that control the dynamics of the rocket:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Rate of ascent: fracd x_hdt = x_v\nAcceleration: fracd x_vdt = fracu_t - D(x_h x_v)x_m - g(x_h)\nRate of mass loss: fracd x_mdt = -fracu_tc","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"where drag D(x_h x_v) is a function of altitude and velocity, gravity g(x_h) is a function of altitude, and c is a constant.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"These forces are defined as:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"D(x_h x_v) = D_c cdot x_v^2 cdot e^-h_c left( fracx_h-x_h(0)x_h(0) right)","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"and g(x_h) = g_0 cdot left( fracx_h(0)x_h right)^2","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"We use a discretized model of time, with a fixed number of time steps, T.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Our goal is thus to maximize x_h(T).","category":"page"},{"location":"tutorials/nonlinear/rocket_control/#Data","page":"Rocket Control","title":"Data","text":"","category":"section"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"All parameters in this model have been normalized to be dimensionless, and they are taken from COPS3.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"h_0 = 1                      # Initial height\nv_0 = 0                      # Initial velocity\nm_0 = 1.0                    # Initial mass\nm_T = 0.6                    # Final mass\ng_0 = 1                      # Gravity at the surface\nh_c = 500                    # Used for drag\nc = 0.5 * sqrt(g_0 * h_0)    # Thrust-to-fuel mass\nD_c = 0.5 * 620 * m_0 / g_0  # Drag scaling\nu_t_max = 3.5 * g_0 * m_0    # Maximum thrust\nT_max = 0.2                  # Number of seconds\nT = 1_000                    # Number of time steps\nΔt = 0.2 / T;                # Time per discretized step\nnothing #hide","category":"page"},{"location":"tutorials/nonlinear/rocket_control/#JuMP-formulation","page":"Rocket Control","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"First, we create a model and choose an optimizer. Since this is a nonlinear program, we need to use a nonlinear solver like Ipopt. We cannot use a linear solver like HiGHS.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"model = Model(Ipopt.Optimizer)\nset_silent(model)","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Next, we create our state and control variables, which are each indexed by t. It is good practice for nonlinear programs to always provide a starting solution for each variable.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"@variable(model, x_v[1:T] >= 0, start = v_0)           # Velocity\n@variable(model, x_h[1:T] >= 0, start = h_0)           # Height\n@variable(model, x_m[1:T] >= m_T, start = m_0)         # Mass\n@variable(model, 0 <= u_t[1:T] <= u_t_max, start = 0); # Thrust\nnothing #hide","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"We implement boundary conditions by fixing variables to values.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"fix(x_v[1], v_0; force = true)\nfix(x_h[1], h_0; force = true)\nfix(x_m[1], m_0; force = true)\nfix(u_t[T], 0.0; force = true)","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"The objective is to maximize altitude at end of time of flight.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"@objective(model, Max, x_h[T])","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Forces are defined as functions:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"D(x_h, x_v) = D_c * x_v^2 * exp(-h_c * (x_h - h_0) / h_0)\ng(x_h) = g_0 * (h_0 / x_h)^2","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"The dynamical equations are implemented as constraints.","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"ddt(x::Vector, t::Int) = (x[t] - x[t-1]) / Δt\n@constraint(model, [t in 2:T], ddt(x_h, t) == x_v[t-1])\n@constraint(\n    model,\n    [t in 2:T],\n    ddt(x_v, t) == (u_t[t-1] - D(x_h[t-1], x_v[t-1])) / x_m[t-1] - g(x_h[t-1]),\n)\n@constraint(model, [t in 2:T], ddt(x_m, t) == -u_t[t-1] / c);\nnothing #hide","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Now we optimize the model and check that we found a solution:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Finally, we plot the solution:","category":"page"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"function plot_trajectory(y; kwargs...)\n    return Plots.plot(\n        (1:T) * Δt,\n        value.(y);\n        xlabel = \"Time (s)\",\n        legend = false,\n        kwargs...,\n    )\nend\n\nPlots.plot(\n    plot_trajectory(x_h; ylabel = \"Altitude\"),\n    plot_trajectory(x_m; ylabel = \"Mass\"),\n    plot_trajectory(x_v; ylabel = \"Velocity\"),\n    plot_trajectory(u_t; ylabel = \"Thrust\");\n    layout = (2, 2),\n)","category":"page"},{"location":"tutorials/nonlinear/rocket_control/#Next-steps","page":"Rocket Control","title":"Next steps","text":"","category":"section"},{"location":"tutorials/nonlinear/rocket_control/","page":"Rocket Control","title":"Rocket Control","text":"Experiment with different values for the constants. How does the solution change? In particular, what happens if you change T_max?\nThe dynamical equations use rectangular integration for the right-hand side terms. Modify the equations to use the Trapezoidal rule instead. (As an example, x_v[t-1] would become 0.5 * (x_v[t-1] + x_v[t]).) Is there a difference?","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"EditURL = \"querying_hessians.jl\"","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/#Computing-Hessians","page":"Computing Hessians","title":"Computing Hessians","text":"","category":"section"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"The purpose of this tutorial is to demonstrate how to compute the Hessian of the Lagrangian of a nonlinear program.","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"warning: Warning\nThis is an advanced tutorial that interacts with the low-level nonlinear interface of MathOptInterface.By default, JuMP exports the MOI symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:import MathOptInterface as MOI","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Given a nonlinear program:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"beginalign\n min_x in mathbbR^n  f(x) \n textst  l le g_i(x) le u\nendalign","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"the Hessian of the Lagrangian is computed as:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"H(x sigma mu) = sigmanabla^2 f(x) + sum_i=1^m mu_i nabla^2 g_i(x)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"where x is a primal point, sigma is a scalar (typically 1), and mu is a vector of weights corresponding to the Lagrangian dual of the constraints.","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"using JuMP\nimport Ipopt\nimport LinearAlgebra\nimport Random\nimport SparseArrays","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/#The-basic-model","page":"Computing Hessians","title":"The basic model","text":"","category":"section"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"To demonstrate how to interact with the lower-level nonlinear interface, we need an example model. The exact model isn't important; we use the model from The Rosenbrock function tutorial, with some additional constraints to demonstrate various features of the lower-level interface.","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"model = Model(Ipopt.Optimizer)\nset_silent(model)\n@variable(model, x[i = 1:2], start = -i)\n@constraint(model, g_1, x[1]^2 <= 1)\n@constraint(model, g_2, (x[1] + x[2])^2 <= 2)\n@objective(model, Min, (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2)\noptimize!(model)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/#The-analytic-solution","page":"Computing Hessians","title":"The analytic solution","text":"","category":"section"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"With a little work, it is possible to analytically derive the correct hessian:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"function analytic_hessian(x, σ, μ)\n    g_1_H = [2.0 0.0; 0.0 0.0]\n    g_2_H = [2.0 2.0; 2.0 2.0]\n    f_H = zeros(2, 2)\n    f_H[1, 1] = 2.0 + 1200.0 * x[1]^2 - 400.0 * x[2]\n    f_H[1, 2] = f_H[2, 1] = -400.0 * x[1]\n    f_H[2, 2] = 200.0\n    return σ * f_H + μ' * [g_1_H, g_2_H]\nend","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Here are various points:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"analytic_hessian([1, 1], 0, [0, 0])","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"analytic_hessian([1, 1], 0, [1, 0])","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"analytic_hessian([1, 1], 0, [0, 1])","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"analytic_hessian([1, 1], 1, [0, 0])","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/#Create-a-nonlinear-model","page":"Computing Hessians","title":"Create a nonlinear model","text":"","category":"section"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"JuMP delegates automatic differentiation to the MOI.Nonlinear submodule. Therefore, to compute the Hessian of the Lagrangian, we need to create a MOI.Nonlinear.Model object:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"rows = Any[]\nnlp = MOI.Nonlinear.Model()\nfor (F, S) in list_of_constraint_types(model)\n    if F <: VariableRef\n        continue  # Skip variable bounds\n    end\n    for ci in all_constraints(model, F, S)\n        push!(rows, ci)\n        object = constraint_object(ci)\n        MOI.Nonlinear.add_constraint(nlp, object.func, object.set)\n    end\nend\nMOI.Nonlinear.set_objective(nlp, objective_function(model))\nnlp","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"It is important that we save the constraint indices in a vector rows, so that we know the order of the constraints in the nonlinear model.","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Next, we need to convert our model into an MOI.Nonlinear.Evaluator, specifying an automatic differentiation backend. In this case, we use MOI.Nonlinear.SparseReverseMode:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"evaluator = MOI.Nonlinear.Evaluator(\n    nlp,\n    MOI.Nonlinear.SparseReverseMode(),\n    index.(all_variables(model)),\n)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Before computing anything with the evaluator, we need to initialize it. Use MOI.features_available to see what we can query:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"MOI.features_available(evaluator)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Consult the MOI documentation for specifics, but to obtain the Hessian matrix, we need to initialize :Hess:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"MOI.initialize(evaluator, [:Hess])","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"MOI represents the Hessian as a sparse matrix. Get the sparsity pattern as follows:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"hessian_sparsity = MOI.hessian_lagrangian_structure(evaluator)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"The sparsity pattern has a few properties of interest:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Each element (i, j) indicates a structural non-zero in row i and column j\nThe list may contain duplicates, in which case we should add the values together\nThe list does not need to be sorted\nThe list may contain any mix of lower- or upper-triangular indices","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"This format matches Julia's sparse-triplet form of a SparseArray, so we can convert from the sparse Hessian representation to a Julia SparseArray as follows:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"I = [i for (i, _) in hessian_sparsity]\nJ = [j for (_, j) in hessian_sparsity]\nV = zeros(length(hessian_sparsity))\nn = num_variables(model)\nH = SparseArrays.sparse(I, J, V, n, n)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Of course, knowing where the zeros are isn't very interesting. We really want to compute the value of the Hessian matrix at a point.","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"MOI.eval_hessian_lagrangian(evaluator, V, ones(n), 1.0, ones(length(rows)))\nH = SparseArrays.sparse(I, J, V, n, n)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"In practice, we often want to compute the value of the hessian at the optimal solution.","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"First, we compute the primal solution. To do so, we need a vector of the variables in the order that they were passed to the solver:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"x = all_variables(model)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Here x[1] is the variable that corresponds to column 1, and so on. Here's the optimal primal solution:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"x_optimal = value.(x)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Next, we need the optimal dual solution associated with the nonlinear constraints (this is where it is important to record the order of the constraints as we added them to nlp):","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"y_optimal = dual.(rows)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Now we can compute the Hessian at the optimal primal-dual point:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"MOI.eval_hessian_lagrangian(evaluator, V, x_optimal, 1.0, y_optimal)\nH = SparseArrays.sparse(I, J, V, n, n)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"However, this Hessian isn't quite right because it isn't symmetric. We can fix this by filling in the appropriate off-diagonal terms:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"function fill_off_diagonal(H)\n    ret = H + H'\n    row_vals = SparseArrays.rowvals(ret)\n    non_zeros = SparseArrays.nonzeros(ret)\n    for col in 1:size(ret, 2)\n        for i in SparseArrays.nzrange(ret, col)\n            if col == row_vals[i]\n                non_zeros[i] /= 2\n            end\n        end\n    end\n    return ret\nend\n\nfill_off_diagonal(H)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"Putting everything together:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"function compute_optimal_hessian(model::Model)\n    rows = Any[]\n    nlp = MOI.Nonlinear.Model()\n    for (F, S) in list_of_constraint_types(model)\n        for ci in all_constraints(model, F, S)\n            push!(rows, ci)\n            object = constraint_object(ci)\n            MOI.Nonlinear.add_constraint(nlp, object.func, object.set)\n        end\n    end\n    MOI.Nonlinear.set_objective(nlp, objective_function(model))\n    x = all_variables(model)\n    backend = MOI.Nonlinear.SparseReverseMode()\n    evaluator = MOI.Nonlinear.Evaluator(nlp, backend, index.(x))\n    MOI.initialize(evaluator, [:Hess])\n    hessian_sparsity = MOI.hessian_lagrangian_structure(evaluator)\n    I = [i for (i, _) in hessian_sparsity]\n    J = [j for (_, j) in hessian_sparsity]\n    V = zeros(length(hessian_sparsity))\n    MOI.eval_hessian_lagrangian(evaluator, V, value.(x), 1.0, dual.(rows))\n    H = SparseArrays.sparse(I, J, V, length(x), length(x))\n    return Matrix(fill_off_diagonal(H))\nend\n\nH_star = compute_optimal_hessian(model)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"If we compare our solution against the analytical solution:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"analytic_hessian(value.(x), 1.0, dual.([g_1, g_2]))","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"If we look at the eigenvalues of the Hessian:","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"LinearAlgebra.eigvals(H_star)","category":"page"},{"location":"tutorials/nonlinear/querying_hessians/","page":"Computing Hessians","title":"Computing Hessians","text":"we see that they are all positive. Therefore, the Hessian is positive definite, and so the solution found by Ipopt is a local minimizer.","category":"page"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"EditURL = \"https://github.com/jump-dev/SumOfSquares.jl/blob/v0.7.1/README.md\"","category":"page"},{"location":"packages/SumOfSquares/#SumOfSquares.jl","page":"jump-dev/SumOfSquares.jl","title":"SumOfSquares.jl","text":"","category":"section"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"(Image: Build Status) (Image: codecov) (Image: )","category":"page"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"SumOfSquares.jl is a JuMP extension that, when used in conjunction with MultivariatePolynomial and PolyJuMP, implements a sum of squares reformulation for polynomial optimization.","category":"page"},{"location":"packages/SumOfSquares/#License","page":"jump-dev/SumOfSquares.jl","title":"License","text":"","category":"section"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"SumOfSquares.jl is licensed under the MIT license.","category":"page"},{"location":"packages/SumOfSquares/#Installation","page":"jump-dev/SumOfSquares.jl","title":"Installation","text":"","category":"section"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"Install SumOfSquares using Pkg.add:","category":"page"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"import Pkg\nPkg.add(\"SumOfSquares\")","category":"page"},{"location":"packages/SumOfSquares/#Documentation","page":"jump-dev/SumOfSquares.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"See https://jump.dev/SumOfSquares.jl/stable for the most recently tagged version of the documentation.","category":"page"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"See https://jump.dev/SumOfSquares.jl/dev for the in-development version of the documentation.","category":"page"},{"location":"packages/SumOfSquares/#Presentations","page":"jump-dev/SumOfSquares.jl","title":"Presentations","text":"","category":"section"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"Some presentations on, or using, SumOfSquares (see blegat/SumOfSquaresSlides for the source code of the presentations):","category":"page"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"Benoît Legat, Marek Kaluba and Tillmann Weisser at INFORMS 2022 [Slides]\nBenoît Legat at POEMA Learning Week 2\nBenoît Legat, Marek Kaluba and Tillmann Weisser at JuMP-dev 2021 [Slides]\nBenoît Legat at INFORMS 2020 [Slides]\nTillmann Weisser, Benoît Legat, Chris Coey, Lea Kapelevich and Juan Pablo Vielma at JuliaCon 2019 [Slides] [Video]\nBenoît Legat at CNLS 2019 [Slides]\nBenoît Legat at EURO 2019 [Slides]\nBenoît Legat at juliaday Nantes 2019 [Slides]\nBenoît Legat at Summer School on Numerical Computing in Algebraic Geometry 2018 [Poster]\nBenoît Legat at The First Annual JuMP-dev Workshop 2017 [Slides] [Video]\nJoey Huchette at SIAM Opt 2017","category":"page"},{"location":"packages/SumOfSquares/#Citing","page":"jump-dev/SumOfSquares.jl","title":"Citing","text":"","category":"section"},{"location":"packages/SumOfSquares/","page":"jump-dev/SumOfSquares.jl","title":"jump-dev/SumOfSquares.jl","text":"See CITATION.bib.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"EditURL = \"cutting_stock_column_generation.jl\"","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Column-generation","page":"Column generation","title":"Column generation","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"The purpose of this tutorial is to demonstrate the column generation algorithm. As an example, it solves the Cutting stock problem.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"using JuMP\nimport DataFrames\nimport HiGHS\nimport Plots\nimport SparseArrays","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Background","page":"Column generation","title":"Background","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"The cutting stock problem is about cutting large rolls of paper into smaller pieces.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"We denote the set of possible sized pieces that a roll can be cut into by iin 1ldotsI. Each piece i has a width, w_i, and a demand, d_i. The width of the large roll is W.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Our objective is to minimize the number of rolls needed to meet all demand.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Here's the data that we are going to use in this tutorial:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"struct Piece\n    w::Float64\n    d::Int\nend\n\nstruct Data\n    pieces::Vector{Piece}\n    W::Float64\nend\n\nfunction Base.show(io::IO, d::Data)\n    println(io, \"Data for the cutting stock problem:\")\n    println(io, \"  W = $(d.W)\")\n    println(io, \"with pieces:\")\n    println(io, \"   i   w_i d_i\")\n    println(io, \"  ------------\")\n    for (i, p) in enumerate(d.pieces)\n        println(io, lpad(i, 4), \" \", lpad(p.w, 5), \" \", lpad(p.d, 3))\n    end\n    return\nend\n\nfunction get_data()\n    data = [\n        75.0 38\n        75.0 44\n        75.0 30\n        75.0 41\n        75.0 36\n        53.8 33\n        53.0 36\n        51.0 41\n        50.2 35\n        32.2 37\n        30.8 44\n        29.8 49\n        20.1 37\n        16.2 36\n        14.5 42\n        11.0 33\n        8.6 47\n        8.2 35\n        6.6 49\n        5.1 42\n    ]\n    return Data([Piece(data[i, 1], data[i, 2]) for i in axes(data, 1)], 100.0)\nend\n\ndata = get_data()","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Mathematical-formulation","page":"Column generation","title":"Mathematical formulation","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"To formulate the cutting stock problem as a mixed-integer linear program, we assume that there is a set of large rolls j=1ldotsJ to use. Then, we introduce two classes of decision variables:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"x_ij ge 0 textinteger  forall i=1ldotsI j=1ldotsJ\ny_j in 0 1 forall j=1ldotsJ","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"y_j is a binary variable that indicates if we use roll j, and x_ij counts how many pieces of size i that we cut from roll j.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Our mixed-integer linear program is therefore:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"beginalign\n    min  sumlimits_j=1^J y_j \n    textst  sumlimits_i=1^N w_i x_ij le W y_j  forall j=1ldotsJ \n          sumlimits_j=1^J x_ij ge d_i  forall i=1ldotsI \n          x_ij ge 0  forall i=1ldotsN j=1ldotsJ \n          x_ij in mathbbZ  forall i=1ldotsI j=1ldotsJ \n          y_j in 0 1  forall j=1ldotsJ \nendalign","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"The objective is to minimize the number of rolls that we use, and the two constraints ensure that we respect the total width of each large roll and that we satisfy demand exactly.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"The JuMP formulation of this model is:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"I = length(data.pieces)\nJ = 1_000  # Some large number\nmodel = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x[1:I, 1:J] >= 0, Int)\n@variable(model, y[1:J], Bin)\n@objective(model, Min, sum(y))\n@constraint(model, [i in 1:I], sum(x[i, :]) >= data.pieces[i].d)\n@constraint(\n    model,\n    [j in 1:J],\n    sum(data.pieces[i].w * x[i, j] for i in 1:I) <= data.W * y[j],\n);\nnothing #hide","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Unfortunately, we can't solve this formulation for realistic instances because it takes a very long time to solve. (Try removing the time limit.)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"set_time_limit_sec(model, 5.0)\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"However, there is a formulation that solves much faster, and that is to use a column generation scheme.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Column-generation-theory","page":"Column generation","title":"Column generation theory","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"The key insight for column generation is to recognize that feasible columns in the x matrix of variables encode cutting patterns.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"For example, if we look only at the roll j=1, then a feasible solution is:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"x_11 = 1 (1 unit of piece #1)\nx_131 = 1 (1 unit of piece #13)\nAll other x_i1 = 0","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Another solution is","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"x_201 = 19 (19 unit of piece #20)\nAll other x_i1 = 0","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Cutting patterns like x_11 = 1 and x_21 = 1 are infeasible because the combined length is greater than W.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Since there are a finite number of ways that we could cut a roll into a valid cutting pattern, we could create a set of all possible cutting patterns p = 1ldotsP, with data a_ip indicating how many units of piece i we cut in pattern p. Then, we can formulate our mixed-integer linear program as:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"beginalign\n    min  sumlimits_p=1^P x_p \n    textst  sumlimits_p=1^P a_ip x_p ge d_i  forall i=1ldotsI \n          x_p ge 0  forall p=1ldotsP \n          x_p in mathbbZ  forall p=1ldotsP\nendalign","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Unfortunately, there will be a very large number of these patterns, so it is often intractable to enumerate all columns p=1ldotsP.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Column generation is an iterative algorithm that starts with a small set of initial patterns, and then cleverly chooses new columns to add to the main MILP so that we find the optimal solution without having to enumerate every column.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Choosing-the-initial-set-of-patterns","page":"Column generation","title":"Choosing the initial set of patterns","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"For the initial set of patterns, we create a trivial cutting pattern which cuts as many units of piece i as will fit.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"patterns = map(1:I) do i\n    n_pieces = floor(Int, data.W / data.pieces[i].w)\n    return SparseArrays.sparsevec([i], [n_pieces], I)\nend","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"We can visualize the patterns as follows:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"\"\"\"\n    cutting_locations(data::Data, pattern::SparseArrays.SparseVector)\n\nA function which returns a vector of the locations along the roll at which to\ncut in order to produce pattern `pattern`.\n\"\"\"\nfunction cutting_locations(data::Data, pattern::SparseArrays.SparseVector)\n    locations = Float64[]\n    offset = 0.0\n    for (i, c) in zip(SparseArrays.findnz(pattern)...)\n        for _ in 1:c\n            offset += data.pieces[i].w\n            push!(locations, offset)\n        end\n    end\n    return locations\nend\n\nfunction plot_patterns(data::Data, patterns)\n    plot = Plots.bar(;\n        xlims = (0, length(patterns) + 1),\n        ylims = (0, data.W),\n        xlabel = \"Pattern\",\n        ylabel = \"Roll length\",\n    )\n    for (i, p) in enumerate(patterns)\n        locations = cutting_locations(data, p)\n        Plots.bar!(\n            plot,\n            fill(i, length(locations)),\n            reverse(locations);\n            bar_width = 0.6,\n            label = false,\n            color = \"#90caf9\",\n        )\n    end\n    return plot\nend\n\nplot_patterns(data, patterns)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#The-base-problem","page":"Column generation","title":"The base problem","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Using the initial set of patterns, we can create and optimize our base model:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x[1:length(patterns)] >= 0, Int)\n@objective(model, Min, sum(x))\n@constraint(model, demand[i in 1:I], patterns[i]' * x >= data.pieces[i].d)\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"This solution requires 421 rolls. This solution is sub-optimal because the model does not contain the full set of possible patterns.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"How do we find a new column that leads to an improved solution?","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Choosing-new-columns","page":"Column generation","title":"Choosing new columns","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Column generation chooses a new column by relaxing the integrality constraint on x and looking at the dual variable pi_i associated with demand constraint i.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"For example, the dual of demand[13] is:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"unset_integer.(x)\noptimize!(model)\nπ_13 = dual(demand[13])","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Using the economic interpretation of the dual variable, we can say that a one unit increase in demand for piece i will cost an extra pi_i rolls. Alternatively, we can say that a one unit increase in the left-hand side (for example, due to a new cutting pattern) will save us pi_i rolls. Therefore, we want a new column that maximizes the savings associated with the dual variables, while respecting the total width of the roll:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"beginalign\n    max  sumlimits_i=1^I pi_i y_i \n    textst  sumlimits_i=1^I w_i y_i le W \n          y_i ge 0  forall i=1ldotsI \n          y_i in mathbbZ  forall i=1ldotsI \nendalign","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"If this problem, called the pricing problem, has an objective value greater than 1, then we estimate than adding y as the coefficients of a new column will decrease the objective by more than the cost of an extra roll.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Here is code to solve the pricing problem:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"function solve_pricing(data::Data, π::Vector{Float64})\n    I = length(π)\n    model = Model(HiGHS.Optimizer)\n    set_silent(model)\n    @variable(model, y[1:I] >= 0, Int)\n    @constraint(model, sum(data.pieces[i].w * y[i] for i in 1:I) <= data.W)\n    @objective(model, Max, sum(π[i] * y[i] for i in 1:I))\n    optimize!(model)\n    number_of_rolls_saved = objective_value(model)\n    if number_of_rolls_saved > 1 + 1e-8\n        # Benefit of pattern is more than the cost of a new roll plus some\n        # tolerance\n        return SparseArrays.sparse(round.(Int, value.(y)))\n    end\n    return nothing\nend","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"If we solve the pricing problem with an artificial dual vector:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"solve_pricing(data, [1.0 / i for i in 1:I])","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"the solution is a roll with 1 unit of piece #1, 1 unit of piece #17, and 3 units of piece #20.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"If we solve the pricing problem with a dual vector of zeros, then the benefit of the new pattern is less than the cost of a roll, and so the function returns nothing:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"solve_pricing(data, zeros(I))","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Iterative-algorithm","page":"Column generation","title":"Iterative algorithm","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Now we can combine our base model with the pricing subproblem in an iterative column generation scheme:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"while true\n    # Solve the linear relaxation\n    optimize!(model)\n    # Obtain a new dual vector\n    π = dual.(demand)\n    # Solve the pricing problem\n    new_pattern = solve_pricing(data, π)\n    # Stop iterating if there is no new pattern\n    if new_pattern === nothing\n        @info \"No new patterns, terminating the algorithm.\"\n        break\n    end\n    push!(patterns, new_pattern)\n    # Create a new column\n    push!(x, @variable(model, lower_bound = 0))\n    # Update the objective coefficient of the new column\n    set_objective_coefficient(model, x[end], 1.0)\n    # Update the non-zeros in the coefficient matrix\n    for (i, count) in zip(SparseArrays.findnz(new_pattern)...)\n        set_normalized_coefficient(demand[i], x[end], count)\n    end\n    println(\"Found new pattern. Total patterns = $(length(patterns))\")\nend","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"We found lots of new patterns. Here's pattern 21:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"patterns[21]","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Let's have a look at the patterns now:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"plot_patterns(data, patterns)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Looking-at-the-solution","page":"Column generation","title":"Looking at the solution","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Let's see how many of each column we need:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"solution = DataFrames.DataFrame([\n    (pattern = p, rolls = value(x_p)) for (p, x_p) in enumerate(x)\n])\nfilter!(row -> row.rolls > 0, solution)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Since we solved a linear program, some of our columns have fractional solutions. We can create a integer feasible solution by rounding up the orders. This requires 341 rolls:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"sum(ceil.(Int, solution.rolls))","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Alternatively, we can re-introduce the integrality constraints and resolve the problem:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"set_integer.(x)\noptimize!(model)\nsolution = DataFrames.DataFrame([\n    (pattern = p, rolls = value(x_p)) for (p, x_p) in enumerate(x)\n])\nfilter!(row -> row.rolls > 0, solution)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"This now requires 334 rolls:","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"sum(solution.rolls)","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Note that this may not be the global minimum because we are not adding new columns during the solution of the mixed-integer problem model (an algorithm known as branch and price). Nevertheless, the column generation algorithm typically finds good integer feasible solutions to an otherwise intractable optimization problem.","category":"page"},{"location":"tutorials/algorithms/cutting_stock_column_generation/#Next-steps","page":"Column generation","title":"Next steps","text":"","category":"section"},{"location":"tutorials/algorithms/cutting_stock_column_generation/","page":"Column generation","title":"Column generation","text":"Our objective function is to minimize the total number of rolls. What is the total length of waste? How does that compare to the total demand?\nWriting the optimization algorithm is only part of the challenge. Can you develop a better way to communicate the solution to stakeholders?","category":"page"},{"location":"moi/reference/constraints/","page":"Constraints","title":"Constraints","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/constraints.md\"","category":"page"},{"location":"moi/reference/constraints/","page":"Constraints","title":"Constraints","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/constraints/#constraints_ref","page":"Constraints","title":"Constraints","text":"","category":"section"},{"location":"moi/reference/constraints/#Types","page":"Constraints","title":"Types","text":"","category":"section"},{"location":"moi/reference/constraints/","page":"Constraints","title":"Constraints","text":"ConstraintIndex","category":"page"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintIndex","page":"Constraints","title":"MathOptInterface.ConstraintIndex","text":"ConstraintIndex{F, S}\n\nA type-safe wrapper for Int64 for use in referencing F-in-S constraints in a model. The parameter F is the type of the function in the constraint, and the parameter S is the type of set in the constraint. To allow for deletion, indices need not be consecutive. Indices within a constraint type (i.e. F-in-S) must be unique, but non-unique indices across different constraint types are allowed. If F is VariableIndex then the index is equal to the index of the variable. That is for an index::ConstraintIndex{VariableIndex}, we always have\n\nindex.value == MOI.get(model, MOI.ConstraintFunction(), index).value\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#Functions","page":"Constraints","title":"Functions","text":"","category":"section"},{"location":"moi/reference/constraints/","page":"Constraints","title":"Constraints","text":"is_valid(::ModelLike,::ConstraintIndex)\nadd_constraint\nadd_constraints\ntransform\nsupports_constraint","category":"page"},{"location":"moi/reference/constraints/#MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.ConstraintIndex}","page":"Constraints","title":"MathOptInterface.is_valid","text":"is_valid(model::ModelLike, index::Index)::Bool\n\nReturn a Bool indicating whether this index refers to a valid object in the model model.\n\n\n\n\n\n","category":"method"},{"location":"moi/reference/constraints/#MathOptInterface.add_constraint","page":"Constraints","title":"MathOptInterface.add_constraint","text":"add_constraint(model::ModelLike, func::F, set::S)::ConstraintIndex{F,S} where {F,S}\n\nAdd the constraint f(x) in mathcalS where f is defined by func, and mathcalS is defined by set.\n\nadd_constraint(model::ModelLike, v::VariableIndex, set::S)::ConstraintIndex{VariableIndex,S} where {S}\nadd_constraint(model::ModelLike, vec::Vector{VariableIndex}, set::S)::ConstraintIndex{VectorOfVariables,S} where {S}\n\nAdd the constraint v in mathcalS where v is the variable (or vector of variables) referenced by v and mathcalS is defined by set.\n\nAn UnsupportedConstraint error is thrown if model does not support F-in-S constraints,\na AddConstraintNotAllowed error is thrown if it supports F-in-S constraints but it cannot add the constraint(s) in its current state and\na ScalarFunctionConstantNotZero error may be thrown if func is an AbstractScalarFunction with nonzero constant and set is EqualTo, GreaterThan, LessThan or Interval.\na LowerBoundAlreadySet error is thrown if F is a VariableIndex and a constraint was already added to this variable that sets a lower bound.\na UpperBoundAlreadySet error is thrown if F is a VariableIndex and a constraint was already added to this variable that sets an upper bound.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/constraints/#MathOptInterface.add_constraints","page":"Constraints","title":"MathOptInterface.add_constraints","text":"add_constraints(model::ModelLike, funcs::Vector{F}, sets::Vector{S})::Vector{ConstraintIndex{F,S}} where {F,S}\n\nAdd the set of constraints specified by each function-set pair in funcs and sets. F and S should be concrete types. This call is equivalent to add_constraint.(model, funcs, sets) but may be more efficient.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/constraints/#MathOptInterface.transform","page":"Constraints","title":"MathOptInterface.transform","text":"Transform Constraint Set\n\ntransform(model::ModelLike, c::ConstraintIndex{F,S1}, newset::S2)::ConstraintIndex{F,S2}\n\nReplace the set in constraint c with newset. The constraint index c will no longer be valid, and the function returns a new constraint index with the correct type.\n\nSolvers may only support a subset of constraint transforms that they perform efficiently (for example, changing from a LessThan to GreaterThan set). In addition, set modification (where S1 = S2) should be performed via the modify function.\n\nTypically, the user should delete the constraint and add a new one.\n\nExamples\n\nIf c is a ConstraintIndex{ScalarAffineFunction{Float64},LessThan{Float64}},\n\nc2 = transform(model, c, GreaterThan(0.0))\ntransform(model, c, LessThan(0.0)) # errors\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/constraints/#MathOptInterface.supports_constraint","page":"Constraints","title":"MathOptInterface.supports_constraint","text":"supports_constraint(\n    model::ModelLike,\n    ::Type{F},\n    ::Type{S},\n)::Bool where {F<:AbstractFunction,S<:AbstractSet}\n\nReturn a Bool indicating whether model supports F-in-S constraints, that is, copy_to(model, src) does not throw UnsupportedConstraint when src contains F-in-S constraints. If F-in-S constraints are only not supported in specific circumstances, e.g. F-in-S constraints cannot be combined with another type of constraint, it should still return true.\n\n\n\n\n\nMOI.supports_constraint(\n    BT::Type{<:AbstractBridge},\n    F::Type{<:MOI.AbstractFunction},\n    S::Type{<:MOI.AbstractSet},\n)::Bool\n\nReturn a Bool indicating whether the bridges of type BT support bridging F-in-S constraints.\n\nImplementation notes\n\nThis method depends only on the type of the inputs, not the runtime values.\nThere is a default fallback, so you need only implement this method for constraint types that the bridge implements.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/constraints/#Attributes","page":"Constraints","title":"Attributes","text":"","category":"section"},{"location":"moi/reference/constraints/","page":"Constraints","title":"Constraints","text":"AbstractConstraintAttribute\nConstraintName\nConstraintPrimalStart\nConstraintDualStart\nConstraintPrimal\nConstraintDual\nConstraintBasisStatus\nConstraintFunction\nCanonicalConstraintFunction\nConstraintSet\nBasisStatusCode\nBASIC\nNONBASIC\nNONBASIC_AT_LOWER\nNONBASIC_AT_UPPER\nSUPER_BASIC","category":"page"},{"location":"moi/reference/constraints/#MathOptInterface.AbstractConstraintAttribute","page":"Constraints","title":"MathOptInterface.AbstractConstraintAttribute","text":"AbstractConstraintAttribute\n\nAbstract supertype for attribute objects that can be used to set or get attributes (properties) of constraints in the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintName","page":"Constraints","title":"MathOptInterface.ConstraintName","text":"ConstraintName()\n\nA constraint attribute for a string identifying the constraint.\n\nIt is valid for constraints variables to have the same name; however, constraints with duplicate names cannot be looked up using get, regardless of whether they have the same F-in-S type.\n\nConstraintName has a default value of \"\" if not set.\n\nNotes\n\nYou should not implement ConstraintName for VariableIndex constraints.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintPrimalStart","page":"Constraints","title":"MathOptInterface.ConstraintPrimalStart","text":"ConstraintPrimalStart()\n\nA constraint attribute for the initial assignment to some constraint's ConstraintPrimal that the optimizer may use to warm-start the solve.\n\nMay be nothing (unset), a number for AbstractScalarFunction, or a vector for AbstractVectorFunction.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintDualStart","page":"Constraints","title":"MathOptInterface.ConstraintDualStart","text":"ConstraintDualStart()\n\nA constraint attribute for the initial assignment to some constraint's ConstraintDual that the optimizer may use to warm-start the solve.\n\nMay be nothing (unset), a number for AbstractScalarFunction, or a vector for AbstractVectorFunction.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintPrimal","page":"Constraints","title":"MathOptInterface.ConstraintPrimal","text":"ConstraintPrimal(result_index::Int = 1)\n\nA constraint attribute for the assignment to some constraint's primal value(s) in result result_index.\n\nIf the constraint is f(x) in S, then in most cases the ConstraintPrimal is the value of f, evaluated at the correspondng VariablePrimal solution.\n\nHowever, some conic solvers reformulate b - Ax in S to s = b - Ax, s in S. These solvers may return the value of s for ConstraintPrimal, rather than b - Ax. (Although these are constrained by an equality constraint, due to numerical tolerances they may not be identical.)\n\nIf the solver does not have a primal value for the constraint because the result_index is beyond the available solutions (whose number is indicated by the ResultCount attribute), getting this attribute must throw a ResultIndexBoundsError. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check PrimalStatus before accessing the ConstraintPrimal attribute.\n\nIf result_index is omitted, it is 1 by default. See ResultCount for information on how the results are ordered.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintDual","page":"Constraints","title":"MathOptInterface.ConstraintDual","text":"ConstraintDual(result_index::Int = 1)\n\nA constraint attribute for the assignment to some constraint's dual value(s) in result result_index. If result_index is omitted, it is 1 by default.\n\nIf the solver does not have a dual value for the variable because the result_index is beyond the available solutions (whose number is indicated by the ResultCount attribute), getting this attribute must throw a ResultIndexBoundsError. Otherwise, if the result is unavailable for another reason (for instance, only a primal solution is available), the result is undefined. Users should first check DualStatus before accessing the ConstraintDual attribute.\n\nSee ResultCount for information on how the results are ordered.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintBasisStatus","page":"Constraints","title":"MathOptInterface.ConstraintBasisStatus","text":"ConstraintBasisStatus(result_index::Int = 1)\n\nA constraint attribute for the BasisStatusCode of some constraint in result result_index, with respect to an available optimal solution basis. If result_index is omitted, it is 1 by default.\n\nIf the solver does not have a basis statue for the constraint because the result_index is beyond the available solutions (whose number is indicated by the ResultCount attribute), getting this attribute must throw a ResultIndexBoundsError. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check PrimalStatus before accessing the ConstraintBasisStatus attribute.\n\nSee ResultCount for information on how the results are ordered.\n\nNotes\n\nFor the basis status of a variable, query VariableBasisStatus.\n\nConstraintBasisStatus does not apply to VariableIndex constraints. You can infer the basis status of a VariableIndex constraint by looking at the result of VariableBasisStatus.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintFunction","page":"Constraints","title":"MathOptInterface.ConstraintFunction","text":"ConstraintFunction()\n\nA constraint attribute for the AbstractFunction object used to define the constraint.\n\nIt is guaranteed to be equivalent but not necessarily identical to the function provided by the user.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.CanonicalConstraintFunction","page":"Constraints","title":"MathOptInterface.CanonicalConstraintFunction","text":"CanonicalConstraintFunction()\n\nA constraint attribute for a canonical representation of the AbstractFunction object used to define the constraint.\n\nGetting this attribute is guaranteed to return a function that is equivalent but not necessarily identical to the function provided by the user.\n\nBy default, MOI.get(model, MOI.CanonicalConstraintFunction(), ci) fallbacks to MOI.Utilities.canonical(MOI.get(model, MOI.ConstraintFunction(), ci)). However, if model knows that the constraint function is canonical then it can implement a specialized method that directly return the function without calling Utilities.canonical. Therefore, the value returned cannot be assumed to be a copy of the function stored in model. Moreover, Utilities.Model checks with Utilities.is_canonical whether the function stored internally is already canonical and if it's the case, then it returns the function stored internally instead of a copy.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.ConstraintSet","page":"Constraints","title":"MathOptInterface.ConstraintSet","text":"ConstraintSet()\n\nA constraint attribute for the AbstractSet object used to define the constraint.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.BasisStatusCode","page":"Constraints","title":"MathOptInterface.BasisStatusCode","text":"BasisStatusCode\n\nAn Enum of possible values for the ConstraintBasisStatus and VariableBasisStatus attributes, explaining the status of a given element with respect to an optimal solution basis.\n\nNotes\n\nNONBASIC_AT_LOWER and NONBASIC_AT_UPPER should be used only for\n\nconstraints with the Interval set. In this case, they are necessary to distinguish which side of the constraint is active. One-sided constraints (e.g., LessThan and GreaterThan) should use NONBASIC instead of the NONBASIC_AT_* values. This restriction does not apply to VariableBasisStatus, which should return NONBASIC_AT_* regardless of whether the alternative bound exists.\n\nIn linear programs, SUPER_BASIC occurs when a variable with no bounds is not\n\nin the basis.\n\nValues\n\nPossible values are:\n\nBASIC: element is in the basis\nNONBASIC: element is not in the basis\nNONBASIC_AT_LOWER: element is not in the basis and is at its lower bound\nNONBASIC_AT_UPPER: element is not in the basis and is at its upper bound\nSUPER_BASIC: element is not in the basis but is also not at one of its bounds\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/constraints/#MathOptInterface.BASIC","page":"Constraints","title":"MathOptInterface.BASIC","text":"BASIC::BasisStatusCode\n\nAn instance of the BasisStatusCode enum.\n\nBASIC: element is in the basis\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/constraints/#MathOptInterface.NONBASIC","page":"Constraints","title":"MathOptInterface.NONBASIC","text":"NONBASIC::BasisStatusCode\n\nAn instance of the BasisStatusCode enum.\n\nNONBASIC: element is not in the basis\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/constraints/#MathOptInterface.NONBASIC_AT_LOWER","page":"Constraints","title":"MathOptInterface.NONBASIC_AT_LOWER","text":"NONBASIC_AT_LOWER::BasisStatusCode\n\nAn instance of the BasisStatusCode enum.\n\nNONBASIC_AT_LOWER: element is not in the basis and is at its lower bound\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/constraints/#MathOptInterface.NONBASIC_AT_UPPER","page":"Constraints","title":"MathOptInterface.NONBASIC_AT_UPPER","text":"NONBASIC_AT_UPPER::BasisStatusCode\n\nAn instance of the BasisStatusCode enum.\n\nNONBASIC_AT_UPPER: element is not in the basis and is at its upper bound\n\n\n\n\n\n","category":"constant"},{"location":"moi/reference/constraints/#MathOptInterface.SUPER_BASIC","page":"Constraints","title":"MathOptInterface.SUPER_BASIC","text":"SUPER_BASIC::BasisStatusCode\n\nAn instance of the BasisStatusCode enum.\n\nSUPER_BASIC: element is not in the basis but is also not at one of its bounds\n\n\n\n\n\n","category":"constant"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"EditURL = \"tsp_lazy_constraints.jl\"","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#tsp_lazy","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"This tutorial was originally contributed by Daniel Schermer.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"This tutorial describes how to implement the Traveling Salesperson Problem in JuMP using solver-independent lazy constraints that dynamically separate subtours. To be more precise, we use lazy constraints to cut off infeasible subtours only when necessary and not before needed.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"It uses the following packages:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"using JuMP\nimport GLPK\nimport Random\nimport Plots","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#tsp_model","page":"Traveling Salesperson Problem","title":"Mathematical Formulation","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Assume that we are given a complete graph mathcalG(VE) where V is the set of vertices (or cities) and E is the set of edges (or roads). For each pair of vertices i j in V i neq j the edge (ij) in E is associated with a weight (or distance) d_ij in mathbbR^+.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"For this tutorial, we assume the problem to be symmetric, that is, d_ij = d_ji  forall ij in V.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"In the Traveling Salesperson Problem, we are tasked with finding a tour with minimal length that visits every vertex exactly once and then returns to the point of origin, that is, a Hamiltonian cycle with minimal weight.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"To model the problem, we introduce a binary variable, x_ij in 01  forall i j in V, that indicates if edge (ij) is part of the tour or not. Using these variables, the Traveling Salesperson Problem can be modeled as the following integer linear program.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#tsp_objective","page":"Traveling Salesperson Problem","title":"Objective Function","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"The objective is to minimize the length of the tour (due to the assumed symmetry, the second sum only contains ij):","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"textmin  sum_i in V  sum_j in V i  j d_ij x_ij","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Note that it is also possible to use the following objective function instead:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"textmin  sum_i in V  sum_j in V dfracd_ij x_ij2","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#tsp_constraints","page":"Traveling Salesperson Problem","title":"Constraints","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"There are four classes of constraints in our formulation.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"First, due to the presumed symmetry, the following constraints must hold:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"x_ij = x_ji quad forall ij in V","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Second, for each vertex i, exactly two edges must be selected that connect it to other vertices j in the graph G:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"sum_j in V x_ij = 2 quad forall i in V","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Third, we do not permit loops to occur:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"x_ii = 0 quad forall i in V","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"The fourth constraint is more complicated. A major difficulty of the Traveling Salesperson Problem arises from the fact that we need to prevent subtours, that is, several distinct Hamiltonian cycles existing on subgraphs of G.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Note that the previous constraints do not guarantee that the solution will be free of subtours. To this end, by S we label a subset of vertices. Then, for each proper subset S subset V, the following constraints guarantee that no subtour may occur:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"sum_i in S sum_j in S i  j x_ij leq vert S vert - 1 quad forall S subset V","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Problematically, we require exponentially many of these constraints as vert V vert increases. Therefore, we will add these constraints only when necessary.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#tsp_implementation","page":"Traveling Salesperson Problem","title":"Implementation","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"There are two ways we can eliminate subtours in JuMP, both of which will be shown in what follows:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"iteratively solving a new model that incorporates previously identified subtours,\nor adding violated subtours as lazy constraints.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#Data","page":"Traveling Salesperson Problem","title":"Data","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"The vertices are assumed to be randomly distributed in the Euclidean space; thus, the weight (distance) of each edge is defined as follows.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"function generate_distance_matrix(n; random_seed = 1)\n    rng = Random.MersenneTwister(random_seed)\n    X = 100 * rand(rng, n)\n    Y = 100 * rand(rng, n)\n    d = [sqrt((X[i] - X[j])^2 + (Y[i] - Y[j])^2) for i in 1:n, j in 1:n]\n    return X, Y, d\nend\n\nn = 40\nX, Y, d = generate_distance_matrix(n)","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"For the JuMP model, we first initialize the model object. Then, we create the binary decision variables and add the objective function and constraints. By defining the x matrix as Symmetric, we do not need to add explicit constraints that x[i, j] == x[j, i].","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"function build_tsp_model(d, n)\n    model = Model(GLPK.Optimizer)\n    @variable(model, x[1:n, 1:n], Bin, Symmetric)\n    @objective(model, Min, sum(d .* x) / 2)\n    @constraint(model, [i in 1:n], sum(x[i, :]) == 2)\n    @constraint(model, [i in 1:n], x[i, i] == 0)\n    return model\nend","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"To search for violated constraints, based on the edges that are currently in the solution (that is, those that have value x_ij = 1), we identify the shortest cycle through the function subtour(). Whenever a subtour has been identified, a constraint corresponding to the form above can be added to the model.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"function subtour(edges::Vector{Tuple{Int,Int}}, n)\n    shortest_subtour, unvisited = collect(1:n), Set(collect(1:n))\n    while !isempty(unvisited)\n        this_cycle, neighbors = Int[], unvisited\n        while !isempty(neighbors)\n            current = pop!(neighbors)\n            push!(this_cycle, current)\n            if length(this_cycle) > 1\n                pop!(unvisited, current)\n            end\n            neighbors =\n                [j for (i, j) in edges if i == current && j in unvisited]\n        end\n        if length(this_cycle) < length(shortest_subtour)\n            shortest_subtour = this_cycle\n        end\n    end\n    return shortest_subtour\nend","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Let us declare a helper function selected_edges() that will be repeatedly used in what follows.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"function selected_edges(x::Matrix{Float64}, n)\n    return Tuple{Int,Int}[(i, j) for i in 1:n, j in 1:n if x[i, j] > 0.5]\nend","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Other helper functions for computing subtours:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"subtour(x::Matrix{Float64}) = subtour(selected_edges(x, size(x, 1)), size(x, 1))\nsubtour(x::AbstractMatrix{VariableRef}) = subtour(value.(x))","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#Iterative-method","page":"Traveling Salesperson Problem","title":"Iterative method","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"An iterative way of eliminating subtours is the following.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"However, it is reasonable to assume that this is not the most efficient way: whenever a new subtour elimination constraint is added to the model, the optimization has to start from the very beginning.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"That way, the solver will repeatedly discard useful information encountered during previous solves (for example, all cuts, the incumbent solution, or lower bounds).","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"info: Info\nNote that, in principle, it would also be feasible to add all subtours (instead of just the shortest one) to the model. However, preventing just the shortest cycle is often sufficient for breaking other subtours and will keep the model size smaller.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"iterative_model = build_tsp_model(d, n)\noptimize!(iterative_model)\ntime_iterated = solve_time(iterative_model)\ncycle = subtour(iterative_model[:x])\nwhile 1 < length(cycle) < n\n    println(\"Found cycle of length $(length(cycle))\")\n    S = [(i, j) for (i, j) in Iterators.product(cycle, cycle) if i < j]\n    @constraint(\n        iterative_model,\n        sum(iterative_model[:x][i, j] for (i, j) in S) <= length(cycle) - 1,\n    )\n    optimize!(iterative_model)\n    global time_iterated += solve_time(iterative_model)\n    global cycle = subtour(iterative_model[:x])\nend\n\nobjective_value(iterative_model)","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"time_iterated","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"As a quick sanity check, we visualize the optimal tour to verify that no subtour is present:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"function plot_tour(X, Y, x)\n    plot = Plots.plot()\n    for (i, j) in selected_edges(x, size(x, 1))\n        Plots.plot!([X[i], X[j]], [Y[i], Y[j]]; legend = false)\n    end\n    return plot\nend\n\nplot_tour(X, Y, value.(iterative_model[:x]))","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/#Lazy-constraint-method","page":"Traveling Salesperson Problem","title":"Lazy constraint method","text":"","category":"section"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"A more sophisticated approach makes use of lazy constraints. To be more precise, we do this through the subtour_elimination_callback() below, which is only run whenever we encounter a new integer-feasible solution.","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"lazy_model = build_tsp_model(d, n)\nfunction subtour_elimination_callback(cb_data)\n    status = callback_node_status(cb_data, lazy_model)\n    if status != MOI.CALLBACK_NODE_STATUS_INTEGER\n        return  # Only run at integer solutions\n    end\n    cycle = subtour(callback_value.(cb_data, lazy_model[:x]))\n    if !(1 < length(cycle) < n)\n        return  # Only add a constraint if there is a cycle\n    end\n    println(\"Found cycle of length $(length(cycle))\")\n    S = [(i, j) for (i, j) in Iterators.product(cycle, cycle) if i < j]\n    con = @build_constraint(\n        sum(lazy_model[:x][i, j] for (i, j) in S) <= length(cycle) - 1,\n    )\n    MOI.submit(lazy_model, MOI.LazyConstraint(cb_data), con)\n    return\nend\nset_attribute(\n    lazy_model,\n    MOI.LazyConstraintCallback(),\n    subtour_elimination_callback,\n)\noptimize!(lazy_model)\nobjective_value(lazy_model)","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"This finds the same optimal tour:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"plot_tour(X, Y, value.(lazy_model[:x]))","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"Surprisingly, for this particular model with GLPK, the solution time is worse than the iterative method:","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"time_lazy = solve_time(lazy_model)","category":"page"},{"location":"tutorials/algorithms/tsp_lazy_constraints/","page":"Traveling Salesperson Problem","title":"Traveling Salesperson Problem","text":"In most other cases and solvers, however, the lazy time should be faster than the iterative method.","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"CurrentModule = JuMP","category":"page"},{"location":"changelog/#Release-notes","page":"Release notes","title":"Release notes","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.","category":"page"},{"location":"changelog/#Version-1.15.1-(September-23,-2023)","page":"Release notes","title":"Version 1.15.1 (September 23, 2023)","text":"","category":"section"},{"location":"changelog/#Fixed","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed support for single argument min and max operators (#3522)\nFixed error message for add_to_expression! when called with a GenericNonlinearExpr (#3506)\nFixed constraint tags with broadcasted constraints (#3515)\nFixed MethodError in MA.scaling (#3518)\nFixed support for arrays of Parameter variables (#3524)","category":"page"},{"location":"changelog/#Other","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Updated to Documenter@1 (#3501)\nFixed links to data in tutorials (#3512)\nFixed typo in TSP tutorial (#3516)\nImproved error message for VariableNotOwned errors (#3520)\nFixed various JET errors (#3519)","category":"page"},{"location":"changelog/#Version-1.15.0-(September-15,-2023)","page":"Release notes","title":"Version 1.15.0 (September 15, 2023)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"This is a large minor release because it adds an entirely new data structure and API path for working with nonlinear programs. The previous nonlinear interface remains unchanged and is documented at Nonlinear Modeling (Legacy). The new interface is a treated as a non-breaking feature addition and is documented at Nonlinear Modeling.","category":"page"},{"location":"changelog/#Breaking","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Although the new nonlinear interface is a feature addition, there are two changes which might be breaking for a very small number of users.","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The syntax inside JuMP macros is parsed using a different code path, even for linear and quadratic expressions. We made this change to unify how we parse linear, quadratic, and nonlinear expressions. In all cases, the new code returns equivalent expressions, but because of the different order of operations, there are three changes to be aware of when updating:\nThe printed form of the expression may change, for example from x * y to y * x. This can cause tests which test the String representation of a model to fail.\nSome coefficients may change slightly due to floating point round-off error.\nParticularly when working with a JuMP extension, you may encounter a MethodError due to a missing or ambiguous method. These errors are due to previously existing bugs that were not triggered by the previous parsing code. If you encounter such an error, please open a GitHub issue.\nThe methods for Base.:^(x::VariableRef, n::Integer) and Base.:^(x::AffExpr, n::Integer) have changed. Previously, these methods supported only n = 0, 1, 2 and they always returned a QuadExpr, even for the case when n = 0 or n = 1. Now:\nx^0 returns one(T), where T is the value_type of the model (defaults to Float64)\nx^1 returns x\nx^2 returns a QuadExpr\nx^n where !(0 <= n <= 2) returns a NonlinearExpr.\nWe made this change to support nonlinear expressions and to align the mathematical definition of the operation with their return type. (Previously, users were surprised that x^1 returned a QuadExpr.) As a consequence of this change, the methods are now not type-stable. This means that the compiler cannot prove that x^2 returns a QuadExpr. If benchmarking shows that this is a performance problem, you can use the type-stable x * x instead of x^2.","category":"page"},{"location":"changelog/#Added","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added triangle_vec which simplifies adding MOI.LogDetConeTriangle and MOI.RootDetConeTriangle constraints (#3456)\nAdded the new nonlinear interface. This is a very large change. See the documentation at Nonlinear Modeling and the (long) discussion in JuMP.jl#3106. Related PRs are (#3468) (#3472) (#3475) (#3483) (#3487) (#3488) (#3489) (#3504) (#3509)","category":"page"},{"location":"changelog/#Fixed-2","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed uses of @nospecialize which cause precompilation failures in Julia v1.6.0 and v1.6.1. (#3464)\nFixed adding a container of Parameter (#3473)\nFixed return type of x^0 and x^1 to no longer return QuadExpr (see note in Breaking section above) (#3474)\nFixed error messages in LowerBoundRef, UpperBoundRef, FixRef, IntegerRef, BinaryRef, ParameterRef and related functions (#3494)\nFixed type inference of empty containers in JuMP macros (#3500)","category":"page"},{"location":"changelog/#Other-2","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added GAMS to solver documentation (#3357)\nUpdated various tutorials (#3459) (#3460) (#3462) (#3463) (#3465) (#3490) (#3492) (#3503)\nAdded The network multi-commodity flow problem tutorial (#3491)\nAdded Two-stage stochastic programs tutorial (#3466)\nAdded better error messages for unsupported operations in LinearAlgebra (#3476)\nUpdated to the latest version of Documenter (#3484) (#3495) (#3497)\nUpdated GitHub action versions (#3507)","category":"page"},{"location":"changelog/#Version-1.14.1-(September-2,-2023)","page":"Release notes","title":"Version 1.14.1 (September 2, 2023)","text":"","category":"section"},{"location":"changelog/#Fixed-3","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix links in Documentation (#3478)","category":"page"},{"location":"changelog/#Version-1.14.0-(August-27,-2023)","page":"Release notes","title":"Version 1.14.0 (August 27, 2023)","text":"","category":"section"},{"location":"changelog/#Added-2","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added DimensionalData.jl extension (#3413)\nAdded syntactic sugar for the MOI.Parameter set (#3443)\nParameter\nParameterRef\nis_parameter\nparameter_value\nset_parameter_value","category":"page"},{"location":"changelog/#Fixed-4","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed model_convert for BridgeableConstraint (#3437)\nFixed printing models with integer coefficients larger than typemax(Int) (#3447)\nFixed support for constant left-hand side functions in a complementarity constraint (#3452)","category":"page"},{"location":"changelog/#Other-3","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Updated packages used in documentation (#3444) (#3455)\nFixed docstring tests (#3445)\nFixed printing change for MathOptInterface (#3446)\nFixed typos in documentation (#3448) (#3457)\nAdded SCIP to callback documentation (#3449)","category":"page"},{"location":"changelog/#Version-1.13.0-(July-27,-2023)","page":"Release notes","title":"Version 1.13.0 (July 27, 2023)","text":"","category":"section"},{"location":"changelog/#Added-3","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added support for generic number types (#3377) (#3385)\nAdded fallback for MOI.AbstractSymmetricMatrixSetTriangle and MOI.AbstractSymmetricMatrixSetSquare (#3424)","category":"page"},{"location":"changelog/#Fixed-5","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed set_start_values with MOI.Bridges.Objective.SlackBridge (#3422)\nFixed flakey doctest in variables.md (#3425)\nFixed names on CITATION.bib (#3423)","category":"page"},{"location":"changelog/#Other-4","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added Loraine.jl to the installation table (#3426)\nRemoved Penopt.jl from packages.toml (#3428)\nImproved problem statement in cannery example of tutorial (#3430)\nMinor cleanups in Containers.DenseAxisArray implementation (#3429)\nChanged nested_problems.jl: outer/inner to upper/lower (#3433)\nRemoved second SDP relaxation in OPF tutorial (#3432)","category":"page"},{"location":"changelog/#Version-1.12.0-(June-19,-2023)","page":"Release notes","title":"Version 1.12.0 (June 19, 2023)","text":"","category":"section"},{"location":"changelog/#Added-4","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added coefficient_type keyword argument to add_bridge and remove_bridge (#3394)","category":"page"},{"location":"changelog/#Fixed-6","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed error message for matrix in HermitianPSDCone (#3369)\nFixed EditURL for custom documentation pages (#3373)\nFixed return type annotations for MOI.ConstraintPrimal and MOI.ConstraintDual (#3381)\nFixed printing change in Julia nightly (#3391)\nFixed printing of Complex coefficients (#3397)\nFixed printing of constraints in text/latex mode (#3405)\nFixed performance issue in Containers.rowtable (#3410)\nFixed bug when variables added to set of wrong dimension (#3411)","category":"page"},{"location":"changelog/#Other-5","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added more solver READMEs to the documentation (#3358) (#3360) (#3364) (#3365) (#3366) (#3368) (#3372) (#3374) (#3376) (#3379) (#3387) (#3389)\nAdded StatusSwitchingQP.jl to the installation table (#3354)\nUpdated checklist for adding a new solver (#3370)\nUpdated extension-tests.yml action (#3371) (#3375)\nColor logs in GitHub actions (#3392)\nAdded new tutorials\nOptimal power flow (#3395) (#3412)\nLovász numbers (#3399)\nDualization (#3402)\nUpdated JuMP paper citation (#3400)\nChanged GitHub action to upload LaTeX logs when building documentation (#3403)\nFixed printing of SCS log in documentation (#3406)\nUpdated solver versions (#3407)\nUpdated documentation to use Julia v1.9 (#3398)\nReplaced _value_type with MOI.Utilities.value_type (#3414)\nFixed a typo in docstring (#3415)\nRefactored API documentation (#3386)\nUpdated SCIP license (#3420)","category":"page"},{"location":"changelog/#Version-1.11.1-(May-19,-2023)","page":"Release notes","title":"Version 1.11.1 (May 19, 2023)","text":"","category":"section"},{"location":"changelog/#Fixed-7","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a poor error message when sum(::DenseAxisArray; dims) was called (#3338)\nFixed support for dependent sets in the @variable macro (#3344)\nFixed a performance bug in constraints with sparse symmetric matrices (#3349)","category":"page"},{"location":"changelog/#Other-6","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Improved the printing of complex numbers (#3332)\nWhen printing, sets which contain constants ending in .0 now print as integers. This follows the behavior of constants in functions (#3341)\nAdded InfiniteOpt to the extensions documentation (#3343)\nAdded more documentation for the exponential cone (#3345) (#3347)\nAdded checklists for developers (#3346) (#3355)\nFixed test support upcoming Julia nightly (#3351)\nFixed extension-tests.yml action (#3353)\nAdd more solvers to the documentation (#3359) (#3361) (#3362)","category":"page"},{"location":"changelog/#Version-1.11.0-(May-3,-2023)","page":"Release notes","title":"Version 1.11.0 (May 3, 2023)","text":"","category":"section"},{"location":"changelog/#Added-5","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added new methods to print_active_bridges for printing a particular objective, constraint, or variable (#3316)","category":"page"},{"location":"changelog/#Fixed-8","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed tests for MOI v1.14.0 release (#3312)\nFixed indexing containers when an axis is Vector{Any} that contains a Vector{Any} element (#3280)\nFixed getindex(::AbstractJuMPScalar) which is called for an expression like x[] (#3314)\nFixed bug in set_string_names_on_creation with a vector of variables (#3322)\nFixed bug in memoize function in nonlinear documentation (#3337)","category":"page"},{"location":"changelog/#Other-7","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed typos in the documentation (#3317) (#3318) (#3328)\nAdded a test for the order of setting start values (#3315)\nAdded READMEs of solvers and extensions to the docs (#3309) (#3320) (#3327) (#3329) (#3333)\nStyle improvements to src/variables.jl (#3324)\nClarify that column generation does not find global optimum (#3325)\nAdd a GitHub actions workflow for testing extensions prior to release (#3331)\nDocument the release process for JuMP (#3334)\nFix links to discourse and chatroom (#3335)","category":"page"},{"location":"changelog/#Version-1.10.0-(April-3,-2023)","page":"Release notes","title":"Version 1.10.0 (April 3, 2023)","text":"","category":"section"},{"location":"changelog/#Added-6","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added Nonnegatives, Nonpositives and Zeros, and support vector-valued inequality syntax in the JuMP macros (#3273)\nAdded special support for LinearAlgebra.Symmetric and LinearAlgebra.Hermitian matrices in Zeros constraints (#3281) (#3296)\nAdded HermitianMatrixSpace and the Hermitian tag for generating a matrix of variables that is Hermitian (#3292) (#3293)\nAdded Semicontinuous and Semiinteger (#3302)\nAdded support for keyword indexing of containers (#3237)","category":"page"},{"location":"changelog/#Fixed-9","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed [compat] bound for MathOptInterface in Project.toml (#3272)","category":"page"},{"location":"changelog/#Other-8","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Split out the Nested optimization problems tutorial (#3274)\nUpdated doctests to ensure none have hidden state (#3275) (#3276)\nClarified how lazy constraints may revisit points (#3278)\nAdded P-Norm example (#3282)\nClarified docs that macros create new bindings (#3284)\nFixed threading example (#3283)\nAdded plot to The minimum distortion problem (#3288)\nAdded Google style rules for Vale and fixed warnings (#3285)\nAdded citation for the JuMP 1.0 paper (#3294)\nUpdated package versions in the documentation (#3298)\nAdded comment for the order in which start values must be set (#3303)\nImproved error message for unrecognized constraint operators (#3311)","category":"page"},{"location":"changelog/#Version-1.9.0-(March-7,-2023)","page":"Release notes","title":"Version 1.9.0 (March 7, 2023)","text":"","category":"section"},{"location":"changelog/#Added-7","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added get_attribute and set_attribute. These replace get_optimizer_attribute and set_optimizer_attribute, although the _optimizer_ functions remain for backward compatibility. (#3219)\nAdded set_start_values for setting all supported start values in a model (#3238)\nAdd remove_bridge and print_active_bridges (#3259)","category":"page"},{"location":"changelog/#Fixed-10","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The matrix returned by a variable in HermitianPSDCone is now a LinearAlgebra.Hermitian matrix. This is potentially breaking if you have written code to assume the return is a Matrix. (#3245) (#3246)\nFixed missing support for Base.isreal of expressions (#3252)","category":"page"},{"location":"changelog/#Other-9","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a thread safety issue in the Parallelism tutorial (#3240) (#3243)\nImproved the error message when unsupported operators are used in @NL macros (#3236)\nClarified the documentation to say that matrices in HermitianPSDCone must be LinearAlgebra.Hermitian (#3241)\nMinor style fixes to internal macro code (#3247)\nAdd Quantum state discrimination tutorial (#3250)\nImprove error message when begin...end not passed to plural macros (#3255)\nDocument how to register function with varying number of input arguments (#3258)\nTidy tests by removing unneeded JuMP. prefixes (#3260)\nClarified the introduction to the Complex number support tutorial (#3262)\nFixed typos in the Documentation (#3263) (#3266) (#3268) (#3269)","category":"page"},{"location":"changelog/#Version-1.8.2-(February-27,-2023)","page":"Release notes","title":"Version 1.8.2 (February 27, 2023)","text":"","category":"section"},{"location":"changelog/#Fixed-11","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed dot product between complex JuMP expression and number (#3244)","category":"page"},{"location":"changelog/#Other-10","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Polish simple SDP examples (#3232)","category":"page"},{"location":"changelog/#Version-1.8.1-(February-23,-2023)","page":"Release notes","title":"Version 1.8.1 (February 23, 2023)","text":"","category":"section"},{"location":"changelog/#Fixed-12","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed support for init in nonlinear generator expressions (#3226)","category":"page"},{"location":"changelog/#Other-11","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Use and document import MathOptInterface as MOI (#3222)\nRemoved references in documentation to multiobjective optimization being unsupported (#3223)\nAdded tutorial on multi-objective portfolio optimization (#3227)\nRefactored some of the conic tutorials (#3229)\nFixed typos in the documentation (#3230)\nAdded tutorial on parallelism (#3231)","category":"page"},{"location":"changelog/#Version-1.8.0-(February-16,-2023)","page":"Release notes","title":"Version 1.8.0 (February 16, 2023)","text":"","category":"section"},{"location":"changelog/#Added-8","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added --> syntax support for indicator constraints. The old syntax of => remains supported (#3207)\nAdded <--> syntax for reified constraints. For now, few solvers support reified constraints (#3206)\nAdded fix_discrete_variables. This is most useful for computing the dual of a mixed-integer program (#3208)\nAdded support for vector-valued objectives. For details, see the Multi-objective knapsack tutorial (#3176)","category":"page"},{"location":"changelog/#Fixed-13","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in lp_sensitivity_report by switching to an explicit LU factorization of the basis matrix (#3182)\nFixed a bug that prevented [; kwarg] arguments in macros (#3220)","category":"page"},{"location":"changelog/#Other-12","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Minor fixes to the documentation (#3200) (#3201) (#3203) (#3210)\nAdded tutorial Constraint programming (#3202)\nAdded more examples to Tips and Tricks\nRemove _distance_to_set in favor of MOI.Utilities.distance_to_set (#3209)\nImprove The diet problem tutorial by adding the variable as a column in the dataframe (#3213)\nImprove The knapsack problem example tutorial (#3216) (#3217)\nAdded the Ellipsoid approximation tutorial (#3218)","category":"page"},{"location":"changelog/#Version-1.7.0-(January-25,-2023)","page":"Release notes","title":"Version 1.7.0 (January 25, 2023)","text":"","category":"section"},{"location":"changelog/#Added-9","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added support for view of a Containers.DenseAxisArray (#3152) (#3180)\nAdded support for containers of variables in ComplexPlane (#3184)\nAdded support for minimum and maximum generators in nonlinear expressions (#3189)\nAdded SnoopPrecompile statements that reduce the time-to-first-solve in Julia 1.9 (#3193) (#3195) (#3196) (#3197)","category":"page"},{"location":"changelog/#Other-13","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Large refactoring of the tests (#3166) (#3167) (#3168) (#3169) (#3170) (#3171)\nRemove unreachable code due to VERSION checks (#3172)\nDocument how to test JuMP extensions (#3174)\nFix method ambiguities in Containers (#3173)\nImprove error message that is thrown when = is used instead of == in the @constraint macro (#3178)\nImprove the error message when Bool is used instead of Bin in the @variable macro (#3180)\nUpdate versions of the documentation (#3185)\nTidy the import of packages and remove unnecessary prefixes (#3186) (#3187)\nRefactor src/JuMP.jl by moving methods into more relevant files (#3188)\nFix docstring of Model not appearing in the documentation (#3198)","category":"page"},{"location":"changelog/#Version-1.6.0-(January-1,-2023)","page":"Release notes","title":"Version 1.6.0 (January 1, 2023)","text":"","category":"section"},{"location":"changelog/#Added-10","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added a result keyword argument to solution_summary to allow summarizing models with multiple solutions (#3138)\nAdded relax_with_penalty!, which is a useful tool when debugging infeasible models (#3140)\nAdded has_start_value (#3157)\nAdded support for HermitianPSDCone in constraints (#3154)","category":"page"},{"location":"changelog/#Fixed-14","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed promotion of complex expressions (#3150) (#3164)","category":"page"},{"location":"changelog/#Other-14","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added Benders tutorial with in-place resolves (#3145)\nAdded more Tips and tricks for linear programs (#3144) (#3163)\nClarified documentation that start can depend on the indices of a variable container (#3148)\nReplace instances of length and size by the recommended eachindex and axes (#3149)\nAdded a warning explaining why the model is dirty when accessing solution results from a modified model (#3156)\nClarify documentation that PSD ensures a symmetric matrix (#3159)\nMaintenance of the JuMP test suite (#3146) (#3158) (#3162)","category":"page"},{"location":"changelog/#Version-1.5.0-(December-8,-2022)","page":"Release notes","title":"Version 1.5.0 (December 8, 2022)","text":"","category":"section"},{"location":"changelog/#Added-11","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Add support for complex-valued variables:\nHermitianPSDCone (#3109)\nComplexPlane and ComplexVariable (#3134)\nAdd support for MOI.OptimizerWithAttributes in set_optimizer_attribute and get_optimizer_attribute (#3129)","category":"page"},{"location":"changelog/#Fixed-15","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed error message for vectorized interval constraints (#3123)\nFixed passing AbstractString to set_optimizer_attribute (#3127)","category":"page"},{"location":"changelog/#Other-15","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Update package versions used in docs (#3119) (#3133) (#3139)\nFixed output of diet tutorial (#3120)\nExplain how to use Dates.period in set_time_limit_sec (#3121)\nUpdate to JuliaFormatter v1.0.15 (#3130)\nFixed HTTP server example in web_app.jl (#3131)\nUpdate docs to build with Documenter#master (#3094)\nAdd tests for LinearAlgebra operations (#3132)\nTidy these release notes (#3135)\nAdded documentation for Complex number support (#3141)\nRemoved the \"workforce scheduling\" and \"steelT3\" tutorials (#3143)","category":"page"},{"location":"changelog/#Version-1.4.0-(October-29,-2022)","page":"Release notes","title":"Version 1.4.0 (October 29, 2022)","text":"","category":"section"},{"location":"changelog/#Added-12","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added Containers.rowtable which converts a container into a vector of NamedTuples to support the Tables.jl interface. This simplifies converting Containers.DenseAxisArray and Containers.SparseAxisArray objects into tabular forms such as a DataFrame (#3104)\nAdded a new method to Containers.container so that index names are passed to the container (#3088)","category":"page"},{"location":"changelog/#Fixed-16","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in copy_to(dest::Model, src::MOI.ModelLike) when src has nonlinear components (#3101)\nFixed the printing of (-1.0 + 0.0im) coefficients in complex expressions (#3112)\nFixed a parsing bug in nonlinear expressions with generator statements that contain multiple for statements (#3116)","category":"page"},{"location":"changelog/#Other-16","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Converted the multi-commodity flow tutorial to use an SQLite database (#3098)\nFixed a number of typos in the documentation (#3103) (#3107) (#3018)\nImproved various style aspects of the PDF documentation (#3095) (#3098) (#3102)","category":"page"},{"location":"changelog/#Version-1.3.1-(September-28,-2022)","page":"Release notes","title":"Version 1.3.1 (September 28, 2022)","text":"","category":"section"},{"location":"changelog/#Fixed-17","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a performance issue in relax_integrality (#3087)\nFixed the type stability of operators with Complex arguments (#3072)\nFixed a bug which added additional +() terms to some nonlinear expressions (#3091)\nFixed potential method ambiguities with AffExpr and QuadExpr objects (#3092)","category":"page"},{"location":"changelog/#Other-17","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added vale as a linter for the documentation (#3080)\nAdded a tutorial on debugging JuMP models (#3043)\nFixed a number of typos in the documentation (#3079) (#3083)\nMany other small tweaks to the documentation (#3068) (#3073) (#3074) (#3075) (#3076) (#3077) (#3078) (#3081) (#3082) (#3084) (#3085) (#3089)","category":"page"},{"location":"changelog/#Version-1.3.0-(September-5,-2022)","page":"Release notes","title":"Version 1.3.0 (September 5, 2022)","text":"","category":"section"},{"location":"changelog/#Added-13","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Support slicing in SparseAxisArray (#3031)","category":"page"},{"location":"changelog/#Fixed-18","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug introduced in v1.2.0 that prevented DenseAxisArrays with Vector keys (#3064)","category":"page"},{"location":"changelog/#Other-18","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Released the JuMP logos under the CC BY 4.0 license (#3063)\nMinor tweaks to the documentation (#3054) (#3056) (#3057) (#3060) (#3061) (#3065)\nImproved code coverage of a number of files (#3048) (#3049) (#3050) (#3051) (#3052) (#3053) (#3058) (#3059)","category":"page"},{"location":"changelog/#Version-1.2.1-(August-22,-2022)","page":"Release notes","title":"Version 1.2.1 (August 22, 2022)","text":"","category":"section"},{"location":"changelog/#Fixed-19","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug when parsing two-sided nonlinear constraints (#3045)","category":"page"},{"location":"changelog/#Version-1.2.0-(August-16,-2022)","page":"Release notes","title":"Version 1.2.0 (August 16, 2022)","text":"","category":"section"},{"location":"changelog/#Breaking-2","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"This is a large minor release because it significantly refactors the internal code for handling nonlinear programs to use the MathOptInterface.Nonlinear submodule that was introduced in MathOptInterface v1.3.0. As a consequence, the internal datastructure in model.nlp_data has been removed, as has the JuMP._Derivatives submodule. Despite the changes, the public API for nonlinear programming has not changed, and any code that uses only the public API and that worked with v1.1.1 will continue to work with v1.2.0.","category":"page"},{"location":"changelog/#Added-14","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added all_constraints(model; include_variable_in_set_constraints) which simplifies returning a list of all constraint indices in the model.\nAdded the ability to delete nonlinear constraints via delete(::Model, ::NonlinearConstraintRef).\nAdded the ability to provide an explicit Hessian for a multivariate user-defined function.\nAdded support for querying the primal value of a nonlinear constraint via value(::NonlinearConstraintRef)","category":"page"},{"location":"changelog/#Fixed-20","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in Containers.DenseAxisArray so that it now supports indexing with keys that hash to the same value, even if they are different types, for example, Int32 and Int64.\nFixed a bug printing the model when the solver does not support MOI.Name.","category":"page"},{"location":"changelog/#Other-19","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added a constraint programming formulation to the Sudoku tutorial.\nAdded newly supported solvers Pajarito, Clarabel, and COPT to the installation table.\nFixed a variety of other miscellaneous issues in the documentation.","category":"page"},{"location":"changelog/#Version-1.1.1-(June-14,-2022)","page":"Release notes","title":"Version 1.1.1 (June 14, 2022)","text":"","category":"section"},{"location":"changelog/#Other-20","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed problem displaying LaTeX in the documentation\nMinor updates to the style guide\nUpdated to MOI v1.4.0 in the documentation","category":"page"},{"location":"changelog/#Version-1.1.0-(May-25,-2022)","page":"Release notes","title":"Version 1.1.0 (May 25, 2022)","text":"","category":"section"},{"location":"changelog/#Added-15","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added num_constraints(::Model; count_variable_in_set_constraints) to simplify the process of counting the number of constraints in a model\nAdded VariableRef(::ConstraintRef) for querying the variable associated with a bound or integrality constraint.\nAdded set_normalized_coefficients for modifying the variable coefficients of a vector-valued constraint.\nAdded set_string_names_on_creation to disable creating String names for variables and constraints. This can improve performance.","category":"page"},{"location":"changelog/#Fixed-21","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug passing nothing to the start keyword of @variable","category":"page"},{"location":"changelog/#Other-21","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"New tutorials:\nSensitivity analysis of a linear program\nServing web apps\nMinimal ellipse SDP tutorial refactored and improved\nDocs updated to the latest version of each package\nLots of minor fixes and improvements to the documentation","category":"page"},{"location":"changelog/#Version-1.0.0-(March-24,-2022)","page":"Release notes","title":"Version 1.0.0 (March 24, 2022)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Read more about this release, along with an acknowledgement of all the contributors in our JuMP 1.0.0 is released blog post.","category":"page"},{"location":"changelog/#Breaking-3","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The previously deprecated functions (v0.23.0, v0.23.1) have been removed. Deprecation was to improve consistency of function names:\nnum_nl_constraints (see num_nonlinear_constraints)\nall_nl_constraints (see all_nonlinear_constraints)\nadd_NL_expression (see add_nonlinear_expression)\nset_NL_objective  (see set_nonlinear_objective)\nadd_NL_constraint (see add_nonlinear_constraint)\nnl_expr_string (see nonlinear_expr_string)\nnl_constraint_string (see nonlinear_constraint_string)\nSymMatrixSpace (see SymmetricMatrixSpace)\nThe unintentionally exported variable JuMP.op_hint has been renamed to the unexported JuMP._OP_HINT","category":"page"},{"location":"changelog/#Fixed-22","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug writing .nl files\nFixed a bug broadcasting SparseAxisArrays","category":"page"},{"location":"changelog/#Version-0.23.2-(March-14,-2022)","page":"Release notes","title":"Version 0.23.2 (March 14, 2022)","text":"","category":"section"},{"location":"changelog/#Added-16","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added relative_gap to solution_summary\nregister now throws an informative error if the function is not differentiable using ForwardDiff. In some cases, the check in register will encounter a false negative, and the informative error will be thrown at run-time. This usually happens when the function is non-differentiable in a subset of the domain.","category":"page"},{"location":"changelog/#Fixed-23","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a scoping issue when extending the container keyword of containers","category":"page"},{"location":"changelog/#Other-22","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Docs updated to the latest version of each package","category":"page"},{"location":"changelog/#Version-0.23.1-(March-2,-2022)","page":"Release notes","title":"Version 0.23.1 (March 2, 2022)","text":"","category":"section"},{"location":"changelog/#Deprecated","page":"Release notes","title":"Deprecated","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"nl_expr_string and nl_constraint_string have been renamed to nonlinear_expr_string and nonlinear_constraint_string. The old methods still exist with deprecation warnings. This change should impact very few users because to call them you must rely on private internals of the nonlinear API. Users are encouraged to use sprint(show, x) instead, where x is the nonlinear expression or constraint of interest.","category":"page"},{"location":"changelog/#Added-17","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added support for Base.abs2(x) where x is a variable or affine expression. This is mainly useful for complex-valued constraints.","category":"page"},{"location":"changelog/#Fixed-24","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed addition of complex and real affine expressions\nFixed arithmetic for Complex-valued quadratic expressions\nFixed variable bounds passed as Rational{Int}(Inf)\nFixed printing of the coefficient (0 + 1im)\nFixed a bug when solution_summary is called prior to optimize!","category":"page"},{"location":"changelog/#Version-0.23.0-(February-25,-2022)","page":"Release notes","title":"Version 0.23.0 (February 25, 2022)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"JuMP v0.23.0 is a breaking release. It is also a release-candidate for JuMP v1.0.0. That is, if no issues are found with the v0.23.0 release, then it will be re-tagged as v1.0.0.","category":"page"},{"location":"changelog/#Breaking-4","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Julia 1.6 is now the minimum supported version\nMathOptInterface has been updated to v1.0.0\nAll previously deprecated functionality has been removed\nPrintMode, REPLMode and IJuliaMode have been removed in favor of the MIME types MIME\"text/plain\" and MIME\"text/latex\". Replace instances of ::Type{REPLMode} with ::MIME\"text/plain\", REPLMode with MIME(\"text/plain\"), ::Type{IJuliaMode} with ::MIME\"text/latex\", and IJuliaMode with MIME(\"text/latex\").\nFunctions containing the nl_ acronym have been renamed to the more explicit nonlinear_. For example, num_nl_constraints is now num_nonlinear_constraints and set_NL_objective is now set_nonlinear_objective. Calls to the old functions throw an error explaining the new name.\nSymMatrixSpace has been renamed to SymmetricMatrixSpace","category":"page"},{"location":"changelog/#Added-18","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added nonlinear_dual_start_value and set_nonlinear_dual_start_value\nAdded preliminary support for Complex coefficient types","category":"page"},{"location":"changelog/#Fixed-25","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in solution_summary","category":"page"},{"location":"changelog/#Other-23","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"MILP examples have been migrated from GLPK to HiGHS\nFixed various typos\nImproved section on setting constraint start values","category":"page"},{"location":"changelog/#Troubleshooting-problems-when-updating","page":"Release notes","title":"Troubleshooting problems when updating","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"If you experience problems when updating, you are likely using previously deprecated functionality. (By default, Julia does not warn when you use deprecated features.)","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"To find the deprecated features you are using, start Julia with --depwarn=yes:","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"$ julia --depwarn=yes","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Then install JuMP v0.22.3:","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"julia> using Pkg\njulia> pkg\"add JuMP@0.22.3\"","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.","category":"page"},{"location":"changelog/#Version-0.22.3-(February-10,-2022)","page":"Release notes","title":"Version 0.22.3 (February 10, 2022)","text":"","category":"section"},{"location":"changelog/#Fixed-26","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a reproducibility issue in the TSP tutorial\nFixed a reproducibility issue in the max_cut_sdp tutorial\nFixed a bug broadcasting an empty SparseAxisArray","category":"page"},{"location":"changelog/#Other-24","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added a warning and improved documentation for the modify-then-query case\nFixed a typo in the docstring of RotatedSecondOrderCone\nAdded Aqua.jl as a check for code health\nAdded introductions to each section of the tutorials\nImproved the column generation and Benders decomposition tutorials\nUpdated documentation to MOI v0.10.8\nUpdated JuliaFormatter to v0.22.2","category":"page"},{"location":"changelog/#Version-0.22.2-(January-10,-2022)","page":"Release notes","title":"Version 0.22.2 (January 10, 2022)","text":"","category":"section"},{"location":"changelog/#Added-19","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The function all_nl_constraints now returns all nonlinear constraints in a model\nstart_value and set_start_value can now be used to get and set the primal start for constraint references\nPlural macros now return a tuple containing the elements that were defined instead of nothing\nAnonymous variables are now printed as _[i] where i is the index of the variable instead of noname. Calling name(x) still returns \"\" so this is non-breaking.","category":"page"},{"location":"changelog/#Fixed-27","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed handling of min and max in nonlinear expressions\nCartesianIndex is no longer allowed as a key for DenseAxisArrays.","category":"page"},{"location":"changelog/#Other-25","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Improved the performance of GenericAffExpr\nAdded a tutorial on the Travelling Salesperson Problem\nAdded a tutorial on querying the Hessian of a nonlinear program\nAdded documentation on using custom solver binaries.","category":"page"},{"location":"changelog/#Version-0.22.1-(November-29,-2021)","page":"Release notes","title":"Version 0.22.1 (November 29, 2021)","text":"","category":"section"},{"location":"changelog/#Added-20","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Export OptimizationSense enum, with instances: MIN_SENSE, MAX_SENSE, and FEASIBILITY_SENSE\nAdd Base.isempty(::Model) to match Base.empty(::Model)","category":"page"},{"location":"changelog/#Fixed-28","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix bug in container with tuples as indices\nFix bug in set_time_limit_sec","category":"page"},{"location":"changelog/#Other-26","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Add tutorial \"Design patterns for larger models\"\nRemove release notes section from PDF\nGeneral edits of the documentation and error messages","category":"page"},{"location":"changelog/#Version-0.22.0-(November-10,-2021)","page":"Release notes","title":"Version 0.22.0 (November 10, 2021)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"JuMP v0.22 is a breaking release","category":"page"},{"location":"changelog/#Breaking-5","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"JuMP 0.22 contains a number of breaking changes. However, these should be invisible for the majority of users. You will mostly encounter these breaking changes if you: wrote a JuMP extension, accessed backend(model), or called @SDconstraint.","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The breaking changes are as follows:","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"MathOptInterface has been updated to v0.10.4. For users who have interacted with the MOI backend, this contains a large number of breaking changes. Read the MathOptInterface release notes for more details.\nThe bridge_constraints keyword argument to Model and set_optimizer has been renamed add_bridges to reflect that more thing were bridged than just constraints.\nThe backend(model) field now contains a concrete instance of a MOI.Utilities.CachingOptimizer instead of one with an abstractly typed optimizer field. In most cases, this will lead to improved performance. However, calling set_optimizer after backend invalidates the old backend. For example:\nmodel = Model()\nb = backend(model)\nset_optimizer(model, GLPK.Optimizer)\n@variable(model, x)\n# b is not updated with `x`! Get a new b by calling `backend` again.\nnew_b = backend(model)\nAll usages of @SDconstraint are deprecated. The new syntax is @constraint(model, X >= Y, PSDCone()).\nCreating a DenseAxisArray with a Number as an axis will now display a warning. This catches a common error in which users write @variable(model, x[length(S)]) instead of @variable(model, x[1:length(S)]).\nThe caching_mode argument to Model, for example, Model(caching_mode = MOIU.MANUAL) mode has been removed. For more control over the optimizer, use direct_model instead.\nThe previously deprecated lp_objective_perturbation_range and lp_rhs_perturbation_range functions have been removed. Use lp_sensitivity_report instead.\nThe .m fields of NonlinearExpression and NonlinearParameter have been renamed to .model.\nInfinite variable bounds are now ignored. Thus, @variable(model, x <= Inf) will show has_upper_bound(x) == false. Previously, these bounds were passed through to the solvers which caused numerical issues for solvers expecting finite bounds.\nThe variable_type and constraint_type functions were removed. This should only affect users who previously wrote JuMP extensions. The functions can be deleted without consequence.\nThe internal functions moi_mode, moi_bridge_constraints, moi_add_constraint, and moi_add_to_function_constant are no longer exported.\nThe un-used method Containers.generate_container has been deleted.\nThe Containers API has been refactored, and _build_ref_sets is now public as Containers.build_ref_sets.\nThe parse_constraint_ methods for extending @constraint at parse time have been refactored in a breaking way. Consult the Extensions documentation for more details and examples.","category":"page"},{"location":"changelog/#Added-21","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The TerminationStatusCode and ResultStatusCode enums are now exported by JuMP. Prefer termination_status(model) == OPTIMAL instead of == MOI.OPTIMAL, although the MOI. prefix way still works.\nCopy a x::DenseAxisArray to an Array by calling Array(x).\nNonlinearExpression is now a subtype of AbstractJuMPScalar\nConstraints such as @constraint(model, x + 1 in MOI.Integer()) are now supported.\nprimal_feasibility_report now accepts a function as the first argument.\nScalar variables @variable(model, x[1:2] in MOI.Integer()) creates two variables, both of which are constrained to be in the set MOI.Integer.\nConic constraints can now be specified as inequalities under a different partial ordering. So @constraint(model, x - y in MOI.Nonnegatives()) can now be written as @constraint(model, x >= y, MOI.Nonnegatives()).\nNames are now set for vectorized constraints.","category":"page"},{"location":"changelog/#Fixed-29","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a performance issue when show was called on a SparseAxisArray with a large number of elements.\nFixed a bug displaying barrier and simplex iterations in solution_summary.\nFixed a bug by implementing hash for DenseAxisArray and SparseAxisArray.\nNames are now only set if the solver supports them. Previously, this prevented solvers such as Ipopt from being used with direct_model.\nMutableArithmetics.Zero is converted into a 0.0 before being returned to the user. Previously, some calls to @expression would return the undocumented MutableArithmetics.Zero() object. One example is summing over an empty set @expression(model, sum(x[i] for i in 1:0)). You will now get 0.0 instead.\nAffExpr and QuadExpr can now be used with == 0 instead of iszero. This fixes a number of issues relating to Julia standard libraries such as LinearAlgebra and SparseArrays.\nFixed a bug when registering a user-defined function with splatting.","category":"page"},{"location":"changelog/#Other-27","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The documentation is now available as a PDF.\nThe documentation now includes a full copy of the MathOptInterface documentation to make it easy to link concepts between the docs. (The MathOptInterface documentation has also been significantly improved.)\nThe documentation contains a large number of improvements and clarifications on a range of topics. Thanks to @sshin23, @DilumAluthge, and @jlwether.\nThe documentation is now built with Julia 1.6 instead of 1.0.\nVarious error messages have been improved to be more readable.","category":"page"},{"location":"changelog/#Version-0.21.10-(September-4,-2021)","page":"Release notes","title":"Version 0.21.10 (September 4, 2021)","text":"","category":"section"},{"location":"changelog/#Added-22","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added add_NL_expression\nadd_NL_xxx functions now support AffExpr and QuadExpr as terms","category":"page"},{"location":"changelog/#Fixed-30","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in solution_summary\nFixed a bug in relax_integrality","category":"page"},{"location":"changelog/#Other-28","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Improved error message in lp_sensitivity_report","category":"page"},{"location":"changelog/#Version-0.21.9-(August-1,-2021)","page":"Release notes","title":"Version 0.21.9 (August 1, 2021)","text":"","category":"section"},{"location":"changelog/#Added-23","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Containers now support arbitrary container types by passing the type to the container keyword and overloading Containers.container.\nis_valid now supports nonlinear constraints\nAdded unsafe_backend for querying the inner-most optimizer of a JuMP model.\nNonlinear parameters now support the plural @NLparameters macro.\nContainers (for example, DenseAxisArray) can now be used in vector-valued constraints.","category":"page"},{"location":"changelog/#Other-29","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Various improvements to the documentation.","category":"page"},{"location":"changelog/#Version-0.21.8-(May-8,-2021)","page":"Release notes","title":"Version 0.21.8 (May 8, 2021)","text":"","category":"section"},{"location":"changelog/#Added-24","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The @constraint macro is now extendable in the same way as @variable.\nAffExpr and QuadExpr can now be used in nonlinear macros.","category":"page"},{"location":"changelog/#Fixed-31","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in lp_sensitivity_report.\nFixed an inference issue when creating empty SparseAxisArrays.","category":"page"},{"location":"changelog/#Version-0.21.7-(April-12,-2021)","page":"Release notes","title":"Version 0.21.7 (April 12, 2021)","text":"","category":"section"},{"location":"changelog/#Added-25","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added primal_feasibility_report, which can be used to check whether a primal point satisfies primal feasibility.\nAdded coefficient, which returns the coefficient associated with a variable in affine and quadratic expressions.\nAdded copy_conflict, which returns the IIS of an infeasible model.\nAdded solution_summary, which returns (and prints) a struct containing a summary of the solution.\nAllow AbstractVector in vector constraints instead of just Vector.\nAdded latex_formulation(model) which returns an object representing the latex formulation of a model. Use print(latex_formulation(model)) to print the formulation as a string.\nUser-defined functions in nonlinear expressions are now automatically registered to aid quick model prototyping. However, a warning is printed to encourage the manual registration.\nDenseAxisArray's now support broadcasting over multiple arrays.\nContainer indices can now be iterators of Base.SizeUnknown.","category":"page"},{"location":"changelog/#Fixed-32","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed bug in rad2deg and deg2rad in nonlinear expressions.\nFixed a MethodError bug in Containers when forcing container type.\nAllow partial slicing of a DenseAxisArray, resolving an issue from 2014.\nFixed a bug printing variable names in IJulia.\nEnding an IJulia cell with model now prints a summary of the model (like in the REPL) not the latex formulation. Use print(model) to print the latex formulation.\nFixed a bug when copying models containing nested arrays.","category":"page"},{"location":"changelog/#Other-30","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Tutorials are now part of the documentation, and more refactoring has taken place.\nAdded JuliaFormatter added as a code formatter.\nAdded some precompilation statements to reduce initial latency.\nVarious improvements to error messages to make them more helpful.\nImproved performance of value(::NonlinearExpression).\nImproved performance of fix(::VariableRef).","category":"page"},{"location":"changelog/#Version-0.21.6-(January-29,-2021)","page":"Release notes","title":"Version 0.21.6 (January 29, 2021)","text":"","category":"section"},{"location":"changelog/#Added-26","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added support for skew symmetric variables via @variable(model, X[1:2, 1:2] in SkewSymmetricMatrixSpace()).\nlp_sensitivity_report has been added which significantly improves the performance of querying the sensitivity summary of an LP. lp_objective_perturbation_range and lp_rhs_perturbation_range are deprecated.\nDual warm-starts are now supported with set_dual_start_value and dual_start_value.\n∈ (\\in<tab>) can now be used in macros instead of = or in.\nUse haskey(model::Model, key::Symbol) to check if a name key is registered in a model.\nAdded unregister(model::Model, key::Symbol) to unregister a name key from model.\nAdded callback_node_status for use in callbacks.\nAdded print_bridge_graph to visualize the bridging graph generated by MathOptInterface.\nImproved error message for containers with duplicate indices.","category":"page"},{"location":"changelog/#Fixed-33","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Various fixes to pass tests on Julia 1.6.\nFixed a bug in the printing of nonlinear expressions in IJulia.\nFixed a bug when nonlinear expressions are passed to user-defined functions.\nSome internal functions that were previously exported are now no longer exported.\nFixed a bug when relaxing a fixed binary variable.\nFixed a StackOverflowError that occurred when SparseAxisArrays had a large number of elements.\nRemoved an unnecessary type assertion in list_of_constraint_types.\nFixed a bug when copying models with registered expressions.","category":"page"},{"location":"changelog/#Other-31","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The documentation has been significantly overhauled. It now has distinct sections for the manual, API reference, and examples. The existing examples in /examples have now been moved to /docs/src/examples and rewritten using Literate.jl, and they are now included in the documentation.\nJuliaFormatter has been applied to most of the codebase. This will continue to roll out over time, as we fix upstream issues in the formatter, and will eventually become compulsory.\nThe root cause of a large number of method invalidations has been resolved.\nWe switched continuous integration from Travis and Appveyor to GitHub Actions.","category":"page"},{"location":"changelog/#Version-0.21.5-(September-18,-2020)","page":"Release notes","title":"Version 0.21.5 (September 18, 2020)","text":"","category":"section"},{"location":"changelog/#Fixed-34","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed deprecation warnings\nThrow DimensionMismatch for incompatibly sized functions and sets\nUnify treatment of keys(x) on JuMP containers","category":"page"},{"location":"changelog/#Version-0.21.4-(September-14,-2020)","page":"Release notes","title":"Version 0.21.4 (September 14, 2020)","text":"","category":"section"},{"location":"changelog/#Added-27","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Add debug info when adding unsupported constraints\nAdd relax_integrality for solving continuous relaxation\nAllow querying constraint conflicts","category":"page"},{"location":"changelog/#Fixed-35","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Dispatch on Real for MOI.submit\nImplement copy for CustomSet in tests\nDon't export private macros\nFix invalid assertion in nonlinear\nError if constraint has NaN right-hand side\nImprove speed of tests\nLots of work modularizing files in /test\nImprove line numbers in macro error messages\nPrint nonlinear subexpressions\nVarious documentation updates\nDependency updates:\nDatastructures 0.18\nMathOptFormat v0.5\nPrep for MathOptInterface 0.9.15","category":"page"},{"location":"changelog/#Version-0.21.3-(June-18,-2020)","page":"Release notes","title":"Version 0.21.3 (June 18, 2020)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added Special Order Sets (SOS1 and SOS2) to JuMP with default weights to ease the creation of such constraints (#2212).\nAdded functions simplex_iterations, barrier_iterations and node_count (#2201).\nAdded function reduced_cost (#2205).\nImplemented callback_value for affine and quadratic expressions (#2231).\nSupport MutableArithmetics.Zero in objective and constraints (#2219).\nDocumentation improvements:\nMention tutorials in the docs (#2223).\nUpdate COIN-OR links (#2242).\nExplicit link to the documentation of MOI.FileFormats (#2253).\nTypo fixes (#2261).\nContainers improvements:\nFix Base.map for DenseAxisArray (#2235).\nThrow BoundsError if number of indices is incorrect for DenseAxisArray and SparseAxisArray (#2240).\nExtensibility improvements:\nImplement a set_objective method fallback that redirects to set_objective_sense and set_objective_function (#2247).\nAdd parse_constraint method with arbitrary number of arguments (#2051).\nAdd parse_constraint_expr and parse_constraint_head (#2228).","category":"page"},{"location":"changelog/#Version-0.21.2-(April-2,-2020)","page":"Release notes","title":"Version 0.21.2 (April 2, 2020)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added relative_gap() to access MOI.RelativeGap() attribute (#2199).\nDocumentation fixes:\nAdded link to source for docstrings in the documentation (#2207).\nAdded docstring for @variables macro (#2216).\nTypo fixes (#2177, #2184, #2182).\nImplementation of methods for Base functions:\nImplemented Base.empty! for JuMP.Model (#2198).\nImplemented Base.conj for JuMP scalar types (#2209).","category":"page"},{"location":"changelog/#Fixed-36","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixed sum of expression with scalar product in macro (#2178).\nFixed writing of nonlinear models to MathOptFormat (#2181).\nFixed construction of empty SparseAxisArray (#2179).\nFixed constraint with zero function (#2188).","category":"page"},{"location":"changelog/#Version-0.21.1-(Feb-18,-2020)","page":"Release notes","title":"Version 0.21.1 (Feb 18, 2020)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Improved the clarity of the with_optimizer deprecation warning.","category":"page"},{"location":"changelog/#Version-0.21.0-(Feb-16,-2020)","page":"Release notes","title":"Version 0.21.0 (Feb 16, 2020)","text":"","category":"section"},{"location":"changelog/#Breaking-6","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Deprecated with_optimizer (#2090, #2084, #2141). You can replace with_optimizer by either nothing, optimizer_with_attributes or a closure:\nreplace with_optimizer(Ipopt.Optimizer) by Ipopt.Optimizer.\nreplace with_optimizer(Ipopt.Optimizer, max_cpu_time=60.0) by optimizer_with_attributes(Ipopt.Optimizer, \"max_cpu_time\" => 60.0).\nreplace with_optimizer(Gurobi.Optimizer, env) by () -> Gurobi.Optimizer(env).\nreplace with_optimizer(Gurobi.Optimizer, env, Presolve=0) by optimizer_with_attributes(() -> Gurobi.Optimizer(env), \"Presolve\" => 0).\nalternatively to optimizer_with_attributes, you can also set the attributes separately with set_optimizer_attribute.\nRenamed set_parameter and set_parameters to set_optimizer_attribute and set_optimizer_attributes (#2150).\nBroadcast should now be explicit inside macros. @SDconstraint(model, x >= 1) and @constraint(model, x + 1 in SecondOrderCone()) now throw an error instead of broadcasting 1 along the dimension of x (#2107).\n@SDconstraint(model, x >= 0) is now equivalent to @constraint(model, x in PSDCone()) instead of @constraint(model, (x .- 0) in PSDCone()) (#2107).\nThe macros now create the containers with map instead of for loops, as a consequence, containers created by @expression can now have any element type and containers of constraint references now have concrete element types when possible. This fixes a long-standing issue where @expression could only be used to generate a collection of linear expressions. Now it works for quadratic expressions as well (#2070).\nCalling deepcopy(::AbstractModel) now throws an error.\nThe constraint name is now printed in the model string (#2108).","category":"page"},{"location":"changelog/#Added-28","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Added support for solver-independent and solver-specific callbacks (#2101).\nAdded write_to_file and read_from_file, supported formats are CBF, LP, MathOptFormat, MPS and SDPA (#2114).\nAdded support for complementarity constraints (#2132).\nAdded support for indicator constraints (#2092).\nAdded support for querying multiple solutions with the result keyword (#2100).\nAdded support for constraining variables on creation (#2128).\nAdded method delete that deletes a vector of variables at once if it is supported by the underlying solver (#2135).\nThe arithmetic between JuMP expression has be refactored into the MutableArithmetics package (#2107).\nImproved error on complex values in NLP (#1978).\nAdded an example of column generation (#2010).","category":"page"},{"location":"changelog/#Fixed-37","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Incorrect coefficients generated when using Symmetric variables (#2102)","category":"page"},{"location":"changelog/#Version-0.20.1-(Oct-18,-2019)","page":"Release notes","title":"Version 0.20.1 (Oct 18, 2019)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Add sections on @variables and @constraints in the documentation (#2062).\nFixed product of sparse matrices for Julia v1.3 (#2063).\nAdded set_objective_coefficient to modify the coefficient of a linear term of the objective function (#2008).\nAdded set_time_limit_sec, unset_time_limit_sec and time_limit_sec to set and query the time limit for the solver in seconds (#2053).","category":"page"},{"location":"changelog/#Version-0.20.0-(Aug-24,-2019)","page":"Release notes","title":"Version 0.20.0 (Aug 24, 2019)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Documentation updates.\nNumerous bug fixes.\nBetter error messages (#1977, #1978, #1997, #2017).\nPerformance improvements (#1947, #2032).\nAdded LP sensitivity summary functions lp_objective_perturbation_range and lp_rhs_perturbation_range (#1917).\nAdded functions dual_objective_value, raw_status and set_parameter.\nAdded function set_objective_coefficient to modify the coefficient of a linear term of the objective (#2008).\nAdded functions set_normalized_rhs, normalized_rhs, and add_to_function_constant to modify and get the constant part of a constraint (#1935, #1960).\nAdded functions set_normalized_coefficient and normalized_coefficient to modify and get the coefficient of a linear term of a constraint (#1935, #1960).\nNumerous other improvements in MOI 0.9, see the NEWS.md file of MOI for more details.","category":"page"},{"location":"changelog/#Version-0.19.2-(June-8,-2019)","page":"Release notes","title":"Version 0.19.2 (June 8, 2019)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in derivatives that could arise in models with nested nonlinear subexpressions.","category":"page"},{"location":"changelog/#Version-0.19.1-(May-12,-2019)","page":"Release notes","title":"Version 0.19.1 (May 12, 2019)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Usability and performance improvements.\nBug fixes.","category":"page"},{"location":"changelog/#Version-0.19.0-(February-15,-2019)","page":"Release notes","title":"Version 0.19.0 (February 15, 2019)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"JuMP 0.19 contains significant breaking changes.","category":"page"},{"location":"changelog/#Breaking-7","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"JuMP's abstraction layer for communicating with solvers changed from MathProgBase (MPB) to MathOptInterface (MOI). MOI addresses many longstanding design issues. (See @mlubin's slides from JuMP-dev 2018.) JuMP 0.19 is compatible only with solvers that have been updated for MOI. See the installation guide for a list of solvers that have and have not yet been updated.\nMost solvers have been renamed to PackageName.Optimizer. For example, GurobiSolver() is now Gurobi.Optimizer.\nSolvers are no longer added to a model via Model(solver = XXX(kwargs...)). Instead use Model(with_optimizer(XXX, kwargs...)). For example, Model(with_optimizer(Gurobi.Optimizer, OutputFlag=0)).\nJuMP containers (for example, the objects returned by @variable) have been redesigned. Containers.SparseAxisArray replaces JuMPDict, JuMPArray was rewritten (inspired by AxisArrays) and renamed Containers.DenseAxisArray, and you can now request a container type with the container= keyword to the macros. See the corresponding documentation for more details.\nThe statuses returned by solvers have changed. See the possible status values here. The MOI statuses are much richer than the MPB statuses and can be used to distinguish between previously indistinguishable cases (for example, did the solver have a feasible solution when it stopped because of the time limit?).\nStarting values are separate from result values. Use value to query the value of a variable in a solution. Use start_value and set_start_value to get and set an initial starting point provided to the solver. The solutions from previous solves are no longer automatically set as the starting points for the next solve.\nThe data structures for affine and quadratic expressions AffExpr and QuadExpr have changed. Internally, terms are stored in dictionaries instead of lists. Duplicate coefficients can no longer exist. Accessors and iteration methods have changed.\nJuMPNLPEvaluator no longer includes the linear and quadratic parts of the model in the evaluation calls. These are now handled separately to allow NLP solvers that support various types of constraints.\nJuMP solver-independent callbacks have been replaced by solver-specific callbacks. See your favorite solver for more details. (See the note below: No solver-specific callbacks are implemented yet.)\nThe norm() syntax is no longer recognized inside macros. Use the SecondOrderCone() set instead.\nJuMP no longer performs automatic transformation between special quadratic forms and second-order cone constraints. Support for these constraint classes depends on the solver.\nThe symbols :Min and :Max are no longer used as optimization senses. Instead, JuMP uses the OptimizationSense enum from MathOptInterface. @objective(model, Max, ...), @objective(model, Min, ...), @NLobjective(model, Max, ...), and @objective(model, Min, ...) remain valid, but @objective(m, :Max, ...) is no longer accepted.\nThe sign conventions for duals has changed in some cases for consistency with conic duality (see the documentation). The shadow_price helper method returns duals with signs that match conventional LP interpretations of dual values as sensitivities of the objective value to relaxations of constraints.\n@constraintref is no longer defined. Instead, create the appropriate container to hold constraint references manually. For example,\nconstraints = Dict() # Optionally, specify types for improved performance.\nfor i in 1:N\n  constraints[i] = @constraint(model, ...)\nend\nThe lowerbound, upperbound, and basename keyword arguments to the @variable macro have been renamed to lower_bound, upper_bound, and base_name, for consistency with JuMP's new style recommendations.\nWe rely on broadcasting syntax to apply accessors to collections of variables, for example, value.(x) instead of getvalue(x) for collections. (Use value(x) when x is a scalar object.)","category":"page"},{"location":"changelog/#Added-29","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Splatting (like f(x...)) is recognized in restricted settings in nonlinear expressions.\nSupport for deleting constraints and variables.\nThe documentation has been completely rewritten using docstrings and Documenter.\nSupport for modeling mixed conic and quadratic models (for example, conic models with quadratic objectives and bi-linear matrix inequalities).\nSignificantly improved support for modeling new types of constraints and for extending JuMP's macros.\nSupport for providing dual warm starts.\nImproved support for accessing solver-specific attributes (for example, the irreducible inconsistent subsystem).\nExplicit control of whether symmetry-enforcing constraints are added to PSD constraints.\nSupport for modeling exponential cones.\nSignificant improvements in internal code quality and testing.\nStyle and naming guidelines.\nDirect mode and manual mode provide explicit control over when copies of a model are stored or regenerated. See the corresponding documentation.","category":"page"},{"location":"changelog/#Regressions","page":"Release notes","title":"Regressions","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"There are known regressions from JuMP 0.18 that will be addressed in a future release (0.19.x or later):","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Performance regressions in model generation (issue). Please file an issue anyway if you notice a significant performance regression. We have plans to address a number of performance issues, but we might not be aware of all of them.\nFast incremental NLP solves are not yet reimplemented (issue).\nWe do not yet have an implementation of solver-specific callbacks.\nThe column generation syntax in @variable has been removed (that is, the objective, coefficients, and inconstraints keyword arguments). Support for column generation will be re-introduced in a future release.\nThe ability to solve the continuous relaxation (that is, via solve(model; relaxation = true)) is not yet reimplemented (issue).","category":"page"},{"location":"changelog/#Version-0.18.5-(December-1,-2018)","page":"Release notes","title":"Version 0.18.5 (December 1, 2018)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Support views in some derivative evaluation functions.\nImproved compatibility with PackageCompiler.","category":"page"},{"location":"changelog/#Version-0.18.4-(October-8,-2018)","page":"Release notes","title":"Version 0.18.4 (October 8, 2018)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in model printing on Julia 0.7 and 1.0.","category":"page"},{"location":"changelog/#Version-0.18.3-(October-1,-2018)","page":"Release notes","title":"Version 0.18.3 (October 1, 2018)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Add support for Julia v1.0 (Thanks @ExpandingMan)\nFix matrix expressions with quadratic functions (#1508)","category":"page"},{"location":"changelog/#Version-0.18.2-(June-10,-2018)","page":"Release notes","title":"Version 0.18.2 (June 10, 2018)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in second-order derivatives when expressions are present (#1319)\nFix a bug in @constraintref (#1330)","category":"page"},{"location":"changelog/#Version-0.18.1-(April-9,-2018)","page":"Release notes","title":"Version 0.18.1 (April 9, 2018)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix for nested tuple destructuring (#1193)\nPreserve internal model when relaxation=true (#1209)\nMinor bug fixes and updates for example","category":"page"},{"location":"changelog/#Version-0.18.0-(July-27,-2017)","page":"Release notes","title":"Version 0.18.0 (July 27, 2017)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Drop support for Julia 0.5.\nUpdate for ForwardDiff 0.5.\nMinor bug fixes.","category":"page"},{"location":"changelog/#Version-0.17.1-(June-9,-2017)","page":"Release notes","title":"Version 0.17.1 (June 9, 2017)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Use of constructconstraint! in @SDconstraint.\nMinor bug fixes.","category":"page"},{"location":"changelog/#Version-0.17.0-(May-27,-2017)","page":"Release notes","title":"Version 0.17.0 (May 27, 2017)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Breaking change: Mixing quadratic and conic constraints is no longer supported.\nBreaking change: The getvariable and getconstraint functions are replaced by indexing on the corresponding symbol. For instance, to access the variable with name x, one should now write m[:x] instead of getvariable(m, :x). As a consequence, creating a variable and constraint with the same name now triggers a warning, and accessing one of them afterwards throws an error. This change is breaking only in the latter case.\nAddition of the getobjectivebound function that mirrors the functionality of the MathProgBase getobjbound function except that it takes into account transformations performed by JuMP.\nMinor bug fixes.","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The following changes are primarily of interest to developers of JuMP extensions:","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The new syntax @constraint(model, expr in Cone) creates the constraint ensuring that expr is inside Cone. The Cone argument is passed to constructconstraint! which enables the call to the dispatched to an extension.\nThe @variable macro now calls constructvariable! instead of directly calling the Variable constructor. Extra arguments and keyword arguments passed to @variable are passed to constructvariable! which enables the call to be dispatched to an extension.\nRefactor the internal function conicdata (used build the MathProgBase conic model) into smaller sub-functions to make these parts reusable by extensions.","category":"page"},{"location":"changelog/#Version-0.16.2-(March-28,-2017)","page":"Release notes","title":"Version 0.16.2 (March 28, 2017)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Minor bug fixes and printing tweaks\nAddress deprecation warnings for Julia 0.6","category":"page"},{"location":"changelog/#Version-0.16.1-(March-7,-2017)","page":"Release notes","title":"Version 0.16.1 (March 7, 2017)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Better support for AbstractArray in JuMP (Thanks @tkoolen)\nMinor bug fixes","category":"page"},{"location":"changelog/#Version-0.16.0-(February-23,-2017)","page":"Release notes","title":"Version 0.16.0 (February 23, 2017)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Breaking change: JuMP no longer has a mechanism for selecting solvers by default (the previous mechanism was flawed and incompatible with Julia 0.6). Not specifying a solver before calling solve() will result in an error.\nBreaking change: User-defined functions are no longer global. The first argument to JuMP.register is now a JuMP Model object within whose scope the function will be registered. Calling JuMP.register without a Model now produces an error.\nBreaking change: Use the new JuMP.fix method to fix a variable to a value or to update the value to which a variable is fixed. Calling setvalue on a fixed variable now results in an error in order to avoid silent behavior changes. (Thanks @joaquimg)\nNonlinear expressions now print out similarly to linear/quadratic expressions (useful for debugging!)\nNew category keyword to @variable. Used for specifying categories of anonymous variables.\nCompatibility with Julia 0.6-dev.\nMinor fixes and improvements (Thanks @cossio, @ccoffrin, @blegat)","category":"page"},{"location":"changelog/#Version-0.15.1-(January-31,-2017)","page":"Release notes","title":"Version 0.15.1 (January 31, 2017)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Bugfix for @LinearConstraints and friends","category":"page"},{"location":"changelog/#Version-0.15.0-(December-22,-2016)","page":"Release notes","title":"Version 0.15.0 (December 22, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Julia 0.5.0 is the minimum required version for this release.\nDocument support for BARON solver\nEnable info callbacks in more states than before, for example, for recording solutions. New when argument to addinfocallback (#814, thanks @yeesian)\nImproved support for anonymous variables. This includes new warnings for potentially confusing use of the traditional non-anonymous syntax:\nWhen multiple variables in a model are given the same name\nWhen non-symbols are used as names, for example, @variable(m, x[1][1:N])\nImprovements in iterating over JuMP containers (#836, thanks @IssamT)\nSupport for writing variable names in .lp file output (Thanks @leethargo)\nSupport for querying duals to SDP problems (Thanks @blegat)\nThe comprehension syntax with curly braces sum{}, prod{}, and norm2{} has been deprecated in favor of Julia's native comprehension syntax sum(), prod() and norm() as previously announced. (For early adopters of the new syntax, norm2() was renamed to norm() without deprecation.)\nUnit tests rewritten to use Base.Test instead of FactCheck\nImproved support for operations with matrices of JuMP types (Thanks @ExpandingMan)\nThe syntax to halt a solver from inside a callback has changed from throw(CallbackAbort()) to return JuMP.StopTheSolver\nMinor bug fixes","category":"page"},{"location":"changelog/#Version-0.14.2-(December-12,-2016)","page":"Release notes","title":"Version 0.14.2 (December 12, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Allow singleton anonymous variables (includes bugfix)","category":"page"},{"location":"changelog/#Version-0.14.1-(September-12,-2016)","page":"Release notes","title":"Version 0.14.1 (September 12, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"More consistent handling of states in informational callbacks, includes a new when parameter to addinfocallback for specifying in which state an informational callback should be called.","category":"page"},{"location":"changelog/#Version-0.14.0-(August-7,-2016)","page":"Release notes","title":"Version 0.14.0 (August 7, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Compatibility with Julia 0.5 and ForwardDiff 0.2\nSupport for \"anonymous\" variables, constraints, expressions, and parameters, for example, x = @variable(m, [1:N]) instead of @variable(m, x[1:N])\nSupport for retrieving constraints from a model by name via getconstraint\n@NLconstraint now returns constraint references (as expected).\nSupport for vectorized expressions within lazy constraints\nOn Julia 0.5, parse new comprehension syntax sum(x[i] for i in 1:N if isodd(i)) instead of sum{ x[i], i in 1:N; isodd(i) }. The old syntax with curly braces will be deprecated in JuMP 0.15.\nNow possible to provide nonlinear expressions as \"raw\" Julia Expr objects instead of using JuMP's nonlinear macros. This input format is useful for programmatically generated expressions.\ns/Mathematical Programming/Mathematical Optimization/\nSupport for local cuts (Thanks to @madanim, Mehdi Madani)\nDocument Xpress interface developed by @joaquimg, Joaquim Dias Garcia\nMinor bug and deprecation fixes (Thanks @odow, @jrevels)","category":"page"},{"location":"changelog/#Version-0.13.2-(May-16,-2016)","page":"Release notes","title":"Version 0.13.2 (May 16, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Compatibility update for MathProgBase","category":"page"},{"location":"changelog/#Version-0.13.1-(May-3,-2016)","page":"Release notes","title":"Version 0.13.1 (May 3, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix broken deprecation for registerNLfunction.","category":"page"},{"location":"changelog/#Version-0.13.0-(April-29,-2016)","page":"Release notes","title":"Version 0.13.0 (April 29, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Most exported methods and macros have been renamed to avoid camelCase. See the list of changes here. There is a 1-1 mapping from the old names to the new, and it is safe to simply replace the names to update existing models.\nSpecify variable lower/upper bounds in @variable using the lowerbound and upperbound keyword arguments.\nChange name printed for variable using the basename keyword argument to @variable.\nNew @variables macro allows multi-line declaration of groups of variables.\nA number of solver methods previously available only through MathProgBase are now exposed directly in JuMP. The fix was recorded live.\nCompatibility fixes with Julia 0.5.\nThe \"end\" indexing syntax is no longer supported within JuMPArrays which do not use 1-based indexing until upstream issues are resolved, see here.","category":"page"},{"location":"changelog/#Version-0.12.2-(March-9,-2016)","page":"Release notes","title":"Version 0.12.2 (March 9, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Small fixes for nonlinear optimization","category":"page"},{"location":"changelog/#Version-0.12.1-(March-1,-2016)","page":"Release notes","title":"Version 0.12.1 (March 1, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a regression in slicing for JuMPArrays (when not using 1-based indexing)","category":"page"},{"location":"changelog/#Version-0.12.0-(February-27,-2016)","page":"Release notes","title":"Version 0.12.0 (February 27, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"The automatic differentiation functionality has been completely rewritten with a number of user-facing changes:\n@defExpr and @defNLExpr now take the model as the first argument. The previous one-argument version of @defExpr is deprecated; all expressions should be named. For example, replace @defExpr(2x+y) with @defExpr(jump_model, my_expr, 2x+y).\nJuMP no longer uses Julia's variable binding rules for efficiently re-solving a sequence of nonlinear models. Instead, we have introduced nonlinear parameters. This is a breaking change, so we have added a warning message when we detect models that may depend on the old behavior.\nSupport for user-defined functions integrated within nonlinear JuMP expressions.\nReplaced iteration over AffExpr with Number-like scalar iteration; previous iteration behavior is now available via linearterms(::AffExpr).\nStopping the solver via throw(CallbackAbort()) from a callback no longer triggers an exception. Instead, solve() returns UserLimit status.\ngetDual() now works for conic problems (Thanks @emreyamangil.)","category":"page"},{"location":"changelog/#Version-0.11.3-(February-4,-2016)","page":"Release notes","title":"Version 0.11.3 (February 4, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Bug-fix for problems with quadratic objectives and semidefinite constraints","category":"page"},{"location":"changelog/#Version-0.11.2-(January-14,-2016)","page":"Release notes","title":"Version 0.11.2 (January 14, 2016)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Compatibility update for Mosek","category":"page"},{"location":"changelog/#Version-0.11.1-(December-1,-2015)","page":"Release notes","title":"Version 0.11.1 (December 1, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Remove usage of @compat in tests.\nFix updating quadratic objectives for nonlinear models.","category":"page"},{"location":"changelog/#Version-0.11.0-(November-30,-2015)","page":"Release notes","title":"Version 0.11.0 (November 30, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Julia 0.4.0 is the minimum required version for this release.\nFix for scoping semantics of index variables in sum{}. Index variables no longer leak into the surrounding scope.\nAddition of the solve(m::Model, relaxation=true) keyword argument to solve the standard continuous relaxation of model m\nThe getConstraintBounds() method allows access to the lower and upper bounds of all constraints in a (nonlinear) model.\nUpdate for breaking changes in MathProgBase","category":"page"},{"location":"changelog/#Version-0.10.3-(November-20,-2015)","page":"Release notes","title":"Version 0.10.3 (November 20, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a rare error when parsing quadratic expressions\nFix Variable() constructor with default arguments\nDetect unrecognized keywords in solve()","category":"page"},{"location":"changelog/#Version-0.10.2-(September-28,-2015)","page":"Release notes","title":"Version 0.10.2 (September 28, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix for deprecation warnings","category":"page"},{"location":"changelog/#Version-0.10.1-(September-3,-2015)","page":"Release notes","title":"Version 0.10.1 (September 3, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixes for ambiguity warnings.\nFix for breaking change in precompilation syntax in Julia 0.4-pre","category":"page"},{"location":"changelog/#Version-0.10.0-(August-31,-2015)","page":"Release notes","title":"Version 0.10.0 (August 31, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Support (on Julia 0.4 and later) for conditions in indexing @defVar and @addConstraint constructs, for example, @defVar(m, x[i=1:5,j=1:5; i+j >= 3])\nSupport for vectorized operations on Variables and expressions. See the documentation for details.\nNew getVar() method to access variables in a model by name\nSupport for semidefinite programming.\nDual solutions are now available for general nonlinear problems. You may call getDual on a reference object for a nonlinear constraint, and getDual on a variable object for Lagrange multipliers from active bounds.\nIntroduce warnings for two common performance traps: too many calls to getValue() on a collection of variables and use of the + operator in a loop to sum expressions.\nSecond-order cone constraints can be written directly with the norm() and norm2{} syntax.\nImplement MathProgBase interface for querying Hessian-vector products.\nIteration over JuMPContainers is deprecated; instead, use the keys and values functions, and zip(keys(d),values(d)) for the old behavior.\n@defVar returns Array{Variable,N} when each of N index sets are of the form 1:nᵢ.\nModule precompilation: on Julia 0.4 and later, using JuMP is now much faster.","category":"page"},{"location":"changelog/#Version-0.9.3-(August-11,-2015)","page":"Release notes","title":"Version 0.9.3 (August 11, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fixes for FactCheck testing on julia v0.4.","category":"page"},{"location":"changelog/#Version-0.9.2-(June-27,-2015)","page":"Release notes","title":"Version 0.9.2 (June 27, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix bug in @addConstraints.","category":"page"},{"location":"changelog/#Version-0.9.1-(April-25,-2015)","page":"Release notes","title":"Version 0.9.1 (April 25, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix for Julia 0.4-dev.\nSmall infrastructure improvements for extensions.","category":"page"},{"location":"changelog/#Version-0.9.0-(April-18,-2015)","page":"Release notes","title":"Version 0.9.0 (April 18, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Comparison operators for constructing constraints (for example, 2x >= 1) have been deprecated. Instead, construct the constraints explicitly in the @addConstraint macro to add them to the model, or in the @LinearConstraint macro to create a stand-alone linear constraint instance.\ngetValue() method implemented to compute the value of a nonlinear subexpression\nJuMP is now released under the Mozilla Public License version 2.0 (was previously LGPL). MPL is a copyleft license which is less restrictive than LGPL, especially for embedding JuMP within other applications.\nA number of performance improvements in ReverseDiffSparse for computing derivatives.\nMathProgBase.getsolvetime(m) now returns the solution time reported by the solver, if available. (Thanks @odow, Oscar Dowson)\nFormatting fix for LP format output. (Thanks @sbebo, Leonardo Taccari).","category":"page"},{"location":"changelog/#Version-0.8.0-(February-17,-2015)","page":"Release notes","title":"Version 0.8.0 (February 17, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Nonlinear subexpressions now supported with the @defNLExpr macro.\nSCS supported for solving second-order conic problems.\nsetXXXCallback family deprecated in favor of addXXXCallback.\nMultiple callbacks of the same type can be registered.\nAdded support for informational callbacks via addInfoCallback.\nA CallbackAbort exception can be thrown from callback to safely exit optimization.","category":"page"},{"location":"changelog/#Version-0.7.4-(February-4,-2015)","page":"Release notes","title":"Version 0.7.4 (February 4, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Reduced costs and linear constraint duals are now accessible when quadratic constraints are present.\nTwo-sided nonlinear constraints are supported.\nMethods for accessing the number of variables and constraints in a model are renamed.\nNew default procedure for setting initial values in nonlinear optimization: project zero onto the variable bounds.\nSmall bug fixes.","category":"page"},{"location":"changelog/#Version-0.7.3-(January-14,-2015)","page":"Release notes","title":"Version 0.7.3 (January 14, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a method ambiguity conflict with Compose.jl (cosmetic fix)","category":"page"},{"location":"changelog/#Version-0.7.2-(January-9,-2015)","page":"Release notes","title":"Version 0.7.2 (January 9, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in sum(::JuMPDict)\nAdded the setCategory function to change a variables category (for example, continuous or binary)","category":"page"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"after construction, and getCategory to retrieve the variable category.","category":"page"},{"location":"changelog/#Version-0.7.1-(January-2,-2015)","page":"Release notes","title":"Version 0.7.1 (January 2, 2015)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in parsing linear expressions in macros. Affects only Julia 0.4 and later.","category":"page"},{"location":"changelog/#Version-0.7.0-(December-29,-2014)","page":"Release notes","title":"Version 0.7.0 (December 29, 2014)","text":"","category":"section"},{"location":"changelog/#Linear/quadratic/conic-programming","page":"Release notes","title":"Linear/quadratic/conic programming","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Breaking change: The syntax for column-wise model generation has been changed to use keyword arguments in @defVar.\nOn Julia 0.4 and later, variables and coefficients may be multiplied in any order within macros. That is, variable*coefficient is now valid syntax.\nECOS supported for solving second-order conic problems.","category":"page"},{"location":"changelog/#_nonlinear_programming_release_notes","page":"Release notes","title":"Nonlinear programming","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Support for skipping model generation when solving a sequence of nonlinear models with changing data.\nFix a memory leak when solving a sequence of nonlinear models.\nThe @addNLConstraint macro now supports the three-argument version to define sets of nonlinear constraints.\nKNITRO supported as a nonlinear solver.\nSpeed improvements for model generation.\nThe @addNLConstraints macro supports adding multiple (groups of) constraints at once. Syntax is similar to @addConstraints.\nDiscrete variables allowed in nonlinear problems for solvers which support them (currently only KNITRO).","category":"page"},{"location":"changelog/#General","page":"Release notes","title":"General","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Starting values for variables may now be specified with @defVar(m, x, start=value).\nThe setSolver function allows users to change the solver subsequent to model creation.\nSupport for \"fixed\" variables via the @defVar(m, x == 1) syntax.\nUnit tests rewritten to use FactCheck.jl, improved testing across solvers.","category":"page"},{"location":"changelog/#Version-0.6.3-(October-19,-2014)","page":"Release notes","title":"Version 0.6.3 (October 19, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in multiplying two AffExpr objects.","category":"page"},{"location":"changelog/#Version-0.6.2-(October-11,-2014)","page":"Release notes","title":"Version 0.6.2 (October 11, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Further improvements and bug fixes for printing.\nFixed a bug in @defExpr.\nSupport for accessing expression graphs through the MathProgBase NLP interface.","category":"page"},{"location":"changelog/#Version-0.6.1-(September-19,-2014)","page":"Release notes","title":"Version 0.6.1 (September 19, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Improvements and bug fixes for printing.","category":"page"},{"location":"changelog/#Version-0.6.0-(September-9,-2014)","page":"Release notes","title":"Version 0.6.0 (September 9, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Julia 0.3.0 is the minimum required version for this release.\nbuildInternalModel(m::Model) added to build solver-level model in memory without optimizing.\nDeprecate load_model_only keyword argument to solve.\nAdd groups of constraints with @addConstraints macro.\nUnicode operators now supported, including ∑ for sum, ∏ for prod, and ≤/≥\nQuadratic constraints supported in @addConstraint macro.\nQuadratic objectives supported in @setObjective macro.\nMathProgBase solver-independent interface replaces Ipopt-specific interface for nonlinear problems\nBreaking change: IpoptOptions no longer supported to specify solver options, use m = Model(solver=IpoptSolver(options...)) instead.\nNew solver interfaces: ECOS, NLopt, and nonlinear support for MOSEK\nNew option to control whether the lazy constraint callback is executed at each node in the B&B tree or just when feasible solutions are found\nAdd support for semicontinuous and semi-integer variables for those solvers that support them.\nAdd support for index dependencies (for example, triangular indexing) in @defVar, @addConstraint, and @defExpr (for example, @defVar(m, x[i=1:10,j=i:10])).\nThis required some changes to the internal structure of JuMP containers, which may break code that explicitly stored JuMPDict objects.","category":"page"},{"location":"changelog/#Version-0.5.8-(September-24,-2014)","page":"Release notes","title":"Version 0.5.8 (September 24, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug with specifying solvers (affects Julia 0.2 only)","category":"page"},{"location":"changelog/#Version-0.5.7-(September-5,-2014)","page":"Release notes","title":"Version 0.5.7 (September 5, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in printing models","category":"page"},{"location":"changelog/#Version-0.5.6-(September-2,-2014)","page":"Release notes","title":"Version 0.5.6 (September 2, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Add support for semicontinuous and semi-integer variables for those solvers that support them.\nBreaking change: Syntax for Variable() constructor has changed (use of this interface remains discouraged)\nUpdate for breaking changes in MathProgBase","category":"page"},{"location":"changelog/#Version-0.5.5-(July-6,-2014)","page":"Release notes","title":"Version 0.5.5 (July 6, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix bug with problem modification: adding variables that did not appear in existing constraints or objective.","category":"page"},{"location":"changelog/#Version-0.5.4-(June-19,-2014)","page":"Release notes","title":"Version 0.5.4 (June 19, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Update for breaking change in MathProgBase which reduces loading times for using JuMP\nFix error when MIPs not solved to optimality","category":"page"},{"location":"changelog/#Version-0.5.3-(May-21,-2014)","page":"Release notes","title":"Version 0.5.3 (May 21, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Update for breaking change in ReverseDiffSparse","category":"page"},{"location":"changelog/#Version-0.5.2-(May-9,-2014)","page":"Release notes","title":"Version 0.5.2 (May 9, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix compatibility with Julia 0.3 prerelease","category":"page"},{"location":"changelog/#Version-0.5.1-(May-5,-2014)","page":"Release notes","title":"Version 0.5.1 (May 5, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix a bug in coefficient handling inside lazy constraints and user cuts","category":"page"},{"location":"changelog/#Version-0.5.0-(May-2,-2014)","page":"Release notes","title":"Version 0.5.0 (May 2, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Support for nonlinear optimization with exact, sparse second-order derivatives automatically computed. Ipopt is currently the only solver supported.\ngetValue for AffExpr and QuadExpr\nBreaking change: getSolverModel replaced by getInternalModel, which returns the internal MathProgBase-level model\nGroups of constraints can be specified with @addConstraint (see documentation for details). This is not a breaking change.\ndot(::JuMPDict{Variable},::JuMPDict{Variable}) now returns the corresponding quadratic expression.","category":"page"},{"location":"changelog/#Version-0.4.1-(March-24,-2014)","page":"Release notes","title":"Version 0.4.1 (March 24, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Fix bug where change in objective sense was ignored when re-solving a model.\nFix issue with handling zero coefficients in AffExpr.","category":"page"},{"location":"changelog/#Version-0.4.0-(March-10,-2014)","page":"Release notes","title":"Version 0.4.0 (March 10, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Support for SOS1 and SOS2 constraints.\nSolver-independent callback for user heuristics.\ndot and sum implemented for JuMPDict objects. Now you can say @addConstraint(m, dot(a,x) <= b).\nDevelopers: support for extensions to JuMP. See definition of Model in src/JuMP.jl for more details.\nOption to construct the low-level model before optimizing.","category":"page"},{"location":"changelog/#Version-0.3.2-(February-17,-2014)","page":"Release notes","title":"Version 0.3.2 (February 17, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Improved model printing\nPreliminary support for IJulia output","category":"page"},{"location":"changelog/#Version-0.3.1-(January-30,-2014)","page":"Release notes","title":"Version 0.3.1 (January 30, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Documentation updates\nSupport for MOSEK\nCPLEXLink renamed to CPLEX","category":"page"},{"location":"changelog/#Version-0.3.0-(January-21,-2014)","page":"Release notes","title":"Version 0.3.0 (January 21, 2014)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Unbounded/infeasibility rays: getValue() will return the corresponding components of an unbounded ray when a model is unbounded, if supported by the selected solver. getDual() will return an infeasibility ray (Farkas proof) if a model is infeasible and the selected solver supports this feature.\nSolver-independent callbacks for user generated cuts.\nUse new interface for solver-independent QCQP.\nsetlazycallback renamed to setLazyCallback for consistency.","category":"page"},{"location":"changelog/#Version-0.2.0-(December-15,-2013)","page":"Release notes","title":"Version 0.2.0 (December 15, 2013)","text":"","category":"section"},{"location":"changelog/#Breaking-8","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Objective sense is specified in setObjective instead of in the Model constructor.\nlpsolver and mipsolver merged into single solver option.","category":"page"},{"location":"changelog/#Added-30","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Problem modification with efficient LP restarts and MIP warm-starts.\nRelatedly, column-wise modeling now supported.\nSolver-independent callbacks supported. Currently we support only a \"lazy constraint\" callback, which works with Gurobi, CPLEX, and GLPK. More callbacks coming soon.","category":"page"},{"location":"changelog/#Version-0.1.2-(November-16,-2013)","page":"Release notes","title":"Version 0.1.2 (November 16, 2013)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Bug fixes for printing, improved error messages.\nAllow AffExpr to be used in macros; for example, ex = y + z; @addConstraint(m, x + 2*ex <= 3)","category":"page"},{"location":"changelog/#Version-0.1.1-(October-23,-2013)","page":"Release notes","title":"Version 0.1.1 (October 23, 2013)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Update for solver specification API changes in MathProgBase.","category":"page"},{"location":"changelog/#Version-0.1.0-(October-3,-2013)","page":"Release notes","title":"Version 0.1.0 (October 3, 2013)","text":"","category":"section"},{"location":"changelog/","page":"Release notes","title":"Release notes","text":"Initial public release.","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"EditURL = \"https://github.com/JuliaPolyhedra/CDDLib.jl/blob/9700752f849845bc8d8477928baf3df920daadd6/README.md\"","category":"page"},{"location":"packages/CDDLib/#CDDLib","page":"JuliaPolyhedra/CDDLib.jl","title":"CDDLib","text":"","category":"section"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"(Image: ) (Image: ) (Image: )","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"CDDLib.jl is a wrapper for cddlib.","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"CDDLib.jl can be used with C API of cddlib, the higher level interface of Polyhedra.jl, or as a linear programming solver with JuMP or MathOptInterface.","category":"page"},{"location":"packages/CDDLib/#Problem-description","page":"JuliaPolyhedra/CDDLib.jl","title":"Problem description","text":"","category":"section"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"As written in the README of cddlib:","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"The C-library  cddlib is a C implementation of the Double Description Method of Motzkin et al. for generating all vertices (that is, extreme points) and extreme rays of a general convex polyhedron in R^d given by a system of linear inequalities:P = { x=(x1, ..., xd)^T :  b - A  x  >= 0 }where  A  is a given m x d real matrix, b is a given m-vector and 0 is the m-vector of all zeros.The program can be used for the reverse operation (that is, convex hull computation).  This means that  one can move back and forth between an inequality representation  and a generator (that is, vertex and ray) representation of a polyhedron with cdd.  Also, cdd can solve a linear programming problem, that is, a problem of maximizing and minimizing a linear function over P.","category":"page"},{"location":"packages/CDDLib/#License","page":"JuliaPolyhedra/CDDLib.jl","title":"License","text":"","category":"section"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"CDDLib.jl is licensed under the GPL v2 license.","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"The underlying solver, cddlib/cddlib is also licensed under the GPL v2 license.","category":"page"},{"location":"packages/CDDLib/#Installation","page":"JuliaPolyhedra/CDDLib.jl","title":"Installation","text":"","category":"section"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"Install CDDLib.jl using the Julia package manager:","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"import Pkg\nPkg.add(\"CDDLib\")","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"Building the package will download binaries of cddlib that are provided by cddlib_jll.jl.","category":"page"},{"location":"packages/CDDLib/#Use-with-JuMP","page":"JuliaPolyhedra/CDDLib.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"Use CDDLib.Optimizer{Float64} to use CDDLib.jl with JuMP:","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"using JuMP, CDDLib\nmodel = Model(CDDLib.Optimizer{Float64})","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"When using CDDLib.jl with MathOptInterface, you can pass a different number type:","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"using MathOptInterface, CDDLib\nmodel = CDDLib.Optimizer{Rational{BigInt}}()","category":"page"},{"location":"packages/CDDLib/#Debugging","page":"JuliaPolyhedra/CDDLib.jl","title":"Debugging","text":"","category":"section"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"CDDLib.jl uses two global Boolean variables to enable debugging outputs: debug and log.","category":"page"},{"location":"packages/CDDLib/","page":"JuliaPolyhedra/CDDLib.jl","title":"JuliaPolyhedra/CDDLib.jl","text":"You can query the value of debug and log with get_debug and get_log, and set their values with set_debug and set_log.","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"EditURL = \"https://github.com/scipopt/SCIP.jl/blob/v0.11.14/README.md\"","category":"page"},{"location":"packages/SCIP/#SCIP.jl","page":"scipopt/SCIP.jl","title":"SCIP.jl","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"(Image: Build Status) (Image: codecov) (Image: Genie Downloads)","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"SCIP.jl is a Julia interface to the SCIP solver.","category":"page"},{"location":"packages/SCIP/#Affiliation","page":"scipopt/SCIP.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"This wrapper is maintained by the SCIP project with the help of the JuMP community.","category":"page"},{"location":"packages/SCIP/#License","page":"scipopt/SCIP.jl","title":"License","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"SCIP.jl is licensed under the MIT License.","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"The underlying solver, scipopt/scip, is licensed under the Apache 2.0 license.","category":"page"},{"location":"packages/SCIP/#Installation","page":"scipopt/SCIP.jl","title":"Installation","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"Install SCIP.jl using Pkg.add:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"import Pkg\nPkg.add(\"SCIP\")","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"On MacOS and Linux, installing the SCIP Julia package will work out of the box and install the SCIP_jll.jl and SCIP_PaPILO_jll.jl dependencies.","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"On Windows, a separate installation of SCIP is still mandatory, as detailed in the \"Custom installation\" section below.","category":"page"},{"location":"packages/SCIP/#Custom-installations","page":"scipopt/SCIP.jl","title":"Custom installations","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"If you use an older Julia version, Windows, or you want a custom SCIP installation, you must manually install the SCIP binaries.","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"Once installed, set the SCIPOPTDIR environment variable to point to the installation path, that is, depending on your operating system, $SCIPOPTDIR/lib/libscip.so, $SCIPOPTDIR/lib/libscip.dylib, or $SCIPOPTDIR/bin/scip.dll must exist.","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"Then, install SCIP.jl using Pkg.add and Pkg.build:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"ENV[\"SCIPOPTDIR\"] = \"/Users/Oscar/code/SCIP\"\nimport Pkg\nPkg.add(\"SCIP\")\nPkg.build(\"SCIP\")","category":"page"},{"location":"packages/SCIP/#Use-with-JuMP","page":"scipopt/SCIP.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"Use SCIP with JuMP as follows:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"using JuMP, SCIP\nmodel = Model(SCIP.Optimizer)\nset_attribute(model, \"display/verblevel\", 0)\nset_attribute(model, \"limits/gap\", 0.05)","category":"page"},{"location":"packages/SCIP/#Options","page":"scipopt/SCIP.jl","title":"Options","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"See the SCIP documentation for a list of supported options.","category":"page"},{"location":"packages/SCIP/#MathOptInterface-API","page":"scipopt/SCIP.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"The SCIP optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"MOI.Reals","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.ZeroOne\nMOI.VectorOfVariables in MOI.SOS1{Float64}\nMOI.VectorOfVariables in MOI.SOS2{Float64}","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"MOI.NLPBlock()\nMOI.ObjectiveSense()\nMOI.UserCutCallback()","category":"page"},{"location":"packages/SCIP/#Design-considerations","page":"scipopt/SCIP.jl","title":"Design considerations","text":"","category":"section"},{"location":"packages/SCIP/#Wrapping-the-public-API","page":"scipopt/SCIP.jl","title":"Wrapping the public API","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"All of the public API methods are wrapped and available within the SCIP package. This includes the scip_*.h and pub_*.h headers that are collected in scip.h, as well as all default constraint handlers (cons_*.h.) ","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"The wrapped functions do not transform any data structures and work on the raw pointers (for example, SCIP* in C, Ptr{SCIP_} in Julia). Convenience wrapper functions based on Julia types are added as needed.","category":"page"},{"location":"packages/SCIP/#Memory-management","page":"scipopt/SCIP.jl","title":"Memory management","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"Programming with SCIP requires dealing with variable and constraint objects that use reference counting for memory management.","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"The SCIP.Optimizer wrapper type collects lists of SCIP_VAR* and SCIP_CONS* under the hood, and it releases all references when it is garbage collected itself (via finalize).","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"When adding a variable (add_variable) or a constraint (add_linear_constraint), an integer index is returned. This index can be used to retrieve the SCIP_VAR* or SCIP_CONS* pointer via get_var and get_cons respectively.","category":"page"},{"location":"packages/SCIP/#Supported-nonlinear-operators","page":"scipopt/SCIP.jl","title":"Supported nonlinear operators","text":"","category":"section"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"Supported operators in nonlinear expressions are as follows:","category":"page"},{"location":"packages/SCIP/","page":"scipopt/SCIP.jl","title":"scipopt/SCIP.jl","text":"+\n-\n*\n/\n^\nsqrt\nexp\nlog\nabs\ncos\nsin","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"EditURL = \"nested_problems.jl\"","category":"page"},{"location":"tutorials/nonlinear/nested_problems/#Nested-optimization-problems","page":"Nested optimization problems","title":"Nested optimization problems","text":"","category":"section"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"In this tutorial we explain how to write a nested optimization problem, where an upper problem uses the results from the optimization of a lower subproblem.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"To model the problem, we define a user-defined operator to handle the decomposition of the lower problem inside the upper one. Finally, we show how to improve the performance by using a cache that avoids resolving the lower problem.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"For a simpler example of writing a user-defined operator, see the User-defined Hessians tutorial.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"using JuMP\nimport Ipopt","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"In the rest of this tutorial, our goal is to solve the bilevel optimization problem:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"beginarrayr l\nminlimits_xz  x_1^2 + x_2^2 + z \nst             beginarrayr l\n                      z = maxlimits_y  x_1^2 y_1 + x_2^2 y_2  - x_1 y_1^4 - 2 x_2 y_2^4 \n                      st                   (y_1 - 10)^2 + (y_2 - 10)^2 le 25\n                  endarray \n                 x ge 0\nendarray","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"This bilevel optimization problem is composed of two nested optimization problems. An upper level, involving variables x, and a lower level, involving variables y. From the perspective of the lower-level problem, the values of x are fixed parameters, and so the model optimizes y given those fixed parameters. Simultaneously, the upper-level problem optimizes x and z given the response of y.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/#Decomposition","page":"Nested optimization problems","title":"Decomposition","text":"","category":"section"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"There are a few ways to solve this problem, but we are going to use a nonlinear decomposition method. The first step is to write a function to compute the lower-level problem:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"beginarrayr l\n  V(x_1 x_2) = maxlimits_y  x_1^2 y_1 + x_2^2 y_2  - x_1 y_1^4 - 2 x_2 y_2^4 \n                           st  (y_1 - 10)^2 + (y_2 - 10)^2 le 25\nendarray","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"function solve_lower_level(x...)\n    model = Model(Ipopt.Optimizer)\n    set_silent(model)\n    @variable(model, y[1:2])\n    @objective(\n        model,\n        Max,\n        x[1]^2 * y[1] + x[2]^2 * y[2] - x[1] * y[1]^4 - 2 * x[2] * y[2]^4,\n    )\n    @constraint(model, (y[1] - 10)^2 + (y[2] - 10)^2 <= 25)\n    optimize!(model)\n    @assert termination_status(model) == LOCALLY_SOLVED\n    return objective_value(model), value.(y)\nend","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"The next function takes a value of x and returns the optimal lower-level objective-value and the optimal response y. The reason why we need both the objective and the optimal y will be made clear shortly, but for now let us define:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"function V(x...)\n    f, _ = solve_lower_level(x...)\n    return f\nend","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"Then, we can substitute V into our full problem to create:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"beginarrayr l\nminlimits_x  x_1^2 + x_2^2 + V(x_1 x_2) \nst             x ge 0\nendarray","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"This looks like a nonlinear optimization problem with a user-defined operator V! However, because V solves an optimization problem internally, we can't use automatic differentiation to compute the first and second derivatives. Instead, we can use JuMP's ability to pass callback functions for the gradient and Hessian instead.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"First up, we need to define the gradient of V with respect to x. In general, this may be difficult to compute, but because x appears only in the objective, we can just differentiate the objective function with respect to x, giving:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"function ∇V(g::AbstractVector, x...)\n    _, y = solve_lower_level(x...)\n    g[1] = 2 * x[1] * y[1] - y[1]^4\n    g[2] = 2 * x[2] * y[2] - 2 * y[2]^4\n    return\nend","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"Second, we need to define the Hessian of V with respect to x. This is a symmetric matrix, but in our example only the diagonal elements are non-zero:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"function ∇²V(H::AbstractMatrix, x...)\n    _, y = solve_lower_level(x...)\n    H[1, 1] = 2 * y[1]\n    H[2, 2] = 2 * y[2]\n    return\nend","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"info: Info\nProviding an explicit Hessian function is optional if first derivatives are already available.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"We now have enough to define our bilevel optimization problem:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"model = Model(Ipopt.Optimizer)\n@variable(model, x[1:2] >= 0)\n@operator(model, op_V, 2, V, ∇V, ∇²V)\n@objective(model, Min, x[1]^2 + x[2]^2 + op_V(x[1], x[2]))\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"The optimal objective value is:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"objective_value(model)","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"and the optimal upper-level decision variables x are:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"value.(x)","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"To compute the optimal lower-level decision variables, we need to call solve_lower_level with the optimal upper-level decision variables:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"_, y = solve_lower_level(value.(x)...)\ny","category":"page"},{"location":"tutorials/nonlinear/nested_problems/#Improving-performance","page":"Nested optimization problems","title":"Improving performance","text":"","category":"section"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"Our solution approach works, but it has a performance problem: every time we need to compute the value, gradient, or Hessian of V, we have to re-solve the lower-level optimization problem. This is wasteful, because we will often call the gradient and Hessian at the same point, and so solving the problem twice with the same input repeats work unnecessarily.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"We can work around this by using a cache:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"mutable struct Cache\n    x::Any\n    f::Float64\n    y::Vector{Float64}\nend","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"with a function to update the cache if needed:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"function _update_if_needed(cache::Cache, x...)\n    if cache.x !== x\n        cache.f, cache.y = solve_lower_level(x...)\n        cache.x = x\n    end\n    return\nend","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"Then, we define cached versions of out three functions which call _updated_if_needed and return values from the cache.","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"function cached_f(cache::Cache, x...)\n    _update_if_needed(cache, x...)\n    return cache.f\nend\n\nfunction cached_∇f(cache::Cache, g::AbstractVector, x...)\n    _update_if_needed(cache, x...)\n    g[1] = 2 * x[1] * cache.y[1] - cache.y[1]^4\n    g[2] = 2 * x[2] * cache.y[2] - 2 * cache.y[2]^4\n    return\nend\n\nfunction cached_∇²f(cache::Cache, H::AbstractMatrix, x...)\n    _update_if_needed(cache, x...)\n    H[1, 1] = 2 * cache.y[1]\n    H[2, 2] = 2 * cache.y[2]\n    return\nend","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"Now we're ready to setup and solve the upper level optimization problem:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"model = Model(Ipopt.Optimizer)\n@variable(model, x[1:2] >= 0)\ncache = Cache(Float64[], NaN, Float64[])\n@operator(\n    model,\n    op_cached_f,\n    2,\n    (x...) -> cached_f(cache, x...),\n    (g, x...) -> cached_∇f(cache, g, x...),\n    (H, x...) -> cached_∇²f(cache, H, x...),\n)\n@objective(model, Min, x[1]^2 + x[2]^2 + op_cached_f(x[1], x[2]))\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"an we can check we get the same objective value:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"objective_value(model)","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"and upper-level decision variable x:","category":"page"},{"location":"tutorials/nonlinear/nested_problems/","page":"Nested optimization problems","title":"Nested optimization problems","text":"value.(x)","category":"page"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"EditURL = \"https://github.com/jump-dev/PolyJuMP.jl/blob/v0.7.0/README.md\"","category":"page"},{"location":"packages/PolyJuMP/#PolyJuMP.jl","page":"jump-dev/PolyJuMP.jl","title":"PolyJuMP.jl","text":"","category":"section"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"PolyJuMP.jl is a JuMP extension for formulating and solving polynomial optimization problems.","category":"page"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"These problems can then be solved using SumOfSquares.jl.","category":"page"},{"location":"packages/PolyJuMP/#License","page":"jump-dev/PolyJuMP.jl","title":"License","text":"","category":"section"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"PolyJuMP.jl is licensed under the MIT license.","category":"page"},{"location":"packages/PolyJuMP/#Installation","page":"jump-dev/PolyJuMP.jl","title":"Installation","text":"","category":"section"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"Install PolyJuMP using Pkg.add:","category":"page"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"import Pkg\nPkg.add(\"PolyJuMP\")","category":"page"},{"location":"packages/PolyJuMP/#Documentation","page":"jump-dev/PolyJuMP.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/PolyJuMP/","page":"jump-dev/PolyJuMP.jl","title":"jump-dev/PolyJuMP.jl","text":"Documentation for PolyJuMP.jl is included in the documentation for SumOfSquares.jl.","category":"page"},{"location":"extensions/introduction/#Introduction","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"extensions/introduction/","page":"Introduction","title":"Introduction","text":"This section of the documentation contains brief documentation for some popular JuMP extensions. The list of extensions is not exhaustive, but instead is intended to help you discover popular JuMP extensions, and to give you an overview of the types of extensions that are possible to write with JuMP.","category":"page"},{"location":"extensions/introduction/#Affiliation","page":"Introduction","title":"Affiliation","text":"","category":"section"},{"location":"extensions/introduction/","page":"Introduction","title":"Introduction","text":"Packages beginning with jump-dev/ are developed and maintained by the JuMP developers.","category":"page"},{"location":"extensions/introduction/","page":"Introduction","title":"Introduction","text":"Packages that do not begin with jump-dev/ are developed independently. The developers of these packages requested or consented to the inclusion of their README contents in the JuMP documentation for the benefit of users.","category":"page"},{"location":"extensions/introduction/#Adding-new-extensions","page":"Introduction","title":"Adding new extensions","text":"","category":"section"},{"location":"extensions/introduction/","page":"Introduction","title":"Introduction","text":"Written an extension? Add it to this section of the JuMP documentation by making a pull request to the docs/packages.toml file.","category":"page"},{"location":"extensions/introduction/#Weak-dependencies","page":"Introduction","title":"Weak dependencies","text":"","category":"section"},{"location":"extensions/introduction/","page":"Introduction","title":"Introduction","text":"Some extensions listed in this section are implemented using the weak dependency feature added to Julia in v1.9. These extensions are activated if and only if you have JuMP and the other package loaded into your current scope with using or import.","category":"page"},{"location":"extensions/introduction/","page":"Introduction","title":"Introduction","text":"compat: Compat\nUsing a weak dependency requires Julia v1.9 or later.","category":"page"},{"location":"installation/#Installation-Guide","page":"Installation Guide","title":"Installation Guide","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"This guide explains how to install Julia and  JuMP. If you have installation troubles, read the Common installation issues section below.","category":"page"},{"location":"installation/#Install-Julia","page":"Installation Guide","title":"Install Julia","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"JuMP is a package for Julia. To use JuMP, first download and install Julia.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"tip: Tip\nIf you are new to Julia, read our Getting started with Julia tutorial.","category":"page"},{"location":"installation/#Choosing-a-version","page":"Installation Guide","title":"Choosing a version","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"You can install the \"Current stable release\" or the \"Long-term support (LTS) release.\"","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"The \"Current stable release\" is the latest release of Julia. It has access to newer features, and is likely faster.\nThe \"Long-term support release\" is an older version of Julia that has continued to receive bug and security fixes. However, it may not have the latest features or performance improvements.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"For most users, you should install the \"Current stable release,\" and whenever Julia releases a new version of the current stable release, you should update your version of Julia. Note that any code you write on one version of the current stable release will continue to work on all subsequent releases.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"For users in restricted software environments (for example, your enterprise IT controls what software you can install), you may be better off installing the long-term support release because you will not have to update Julia as frequently.","category":"page"},{"location":"installation/#Install-JuMP","page":"Installation Guide","title":"Install JuMP","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"From Julia, JuMP is installed using the built-in package manager:","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"import Pkg\nPkg.add(\"JuMP\")","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"tip: Tip\nWe recommend you create a Pkg environment for each project you use JuMP for, instead of adding lots of packages to the global environment. The Pkg manager documentation has more information on this topic.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"When we release a new version of JuMP, you can update with:","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"import Pkg\nPkg.update(\"JuMP\")","category":"page"},{"location":"installation/#Install-a-solver","page":"Installation Guide","title":"Install a solver","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"JuMP depends on solvers to solve optimization problems. Therefore, you will need to install one before you can solve problems with JuMP.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Install a solver using the Julia package manager, replacing \"Clp\" by the Julia package name as appropriate.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"import Pkg\nPkg.add(\"Clp\")","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Once installed, you can use Clp as a solver with JuMP as follows, using set_attribute to set solver-specific options:","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"using JuMP\nusing Clp\nmodel = Model(Clp.Optimizer)\nset_attribute(model, \"LogLevel\" => 1)\nset_attribute(model, \"PrimalTolerance\" => 1e-7)","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"note: Note\nMost packages follow the ModuleName.Optimizer naming convention, but exceptions may exist. See the README of the Julia package's GitHub repository for more details on how to use a particular solver, including any solver-specific options.","category":"page"},{"location":"installation/#Supported-solvers","page":"Installation Guide","title":"Supported solvers","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Most solvers are not written in Julia, and some require commercial licenses to use, so installation is often more complex.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"If a solver has Manual in the Installation column, the solver requires a manual installation step, such as downloading and installing a binary, or obtaining a commercial license. Consult the README of the relevant Julia package for more information.\nIf the solver has Manualᴹ in the Installation column, the solver requires an installation of MATLAB.\nIf the Installation column is missing an entry, installing the Julia package will download and install any relevant solver binaries automatically, and you shouldn't need to do anything other than Pkg.add.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Solvers with a missing entry in the Julia Package column are written in Julia. The link in the Solver column is the corresponding Julia package.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Solver Julia Package Installation License Supports\nAlpine.jl   Triad NS (MI)NLP\nArtelys Knitro KNITRO.jl Manual Comm. (MI)LP, (MI)SOCP, (MI)NLP\nBARON BARON.jl Manual Comm. (MI)NLP\nBonmin AmplNLWriter.jl  EPL (MI)NLP\nCbc Cbc.jl  EPL (MI)LP\nCDCS CDCS.jl Manualᴹ GPL LP, SOCP, SDP\nCDD CDDLib.jl  GPL LP\nClarabel.jl   Apache LP, QP, SOCP, SDP\nClp Clp.jl  EPL LP\nCOPT COPT.jl  Comm. (MI)LP, SOCP, SDP\nCOSMO.jl   Apache LP, QP, SOCP, SDP\nCouenne AmplNLWriter.jl  EPL (MI)NLP\nCPLEX CPLEX.jl Manual Comm. (MI)LP, (MI)SOCP\nCSDP CSDP.jl  EPL LP, SDP\nDAQP DAQP.jl  MIT (Mixed-binary) QP\nEAGO.jl   MIT NLP\nECOS ECOS.jl  GPL LP, SOCP\nFICO Xpress Xpress.jl Manual Comm. (MI)LP, (MI)SOCP\nGLPK GLPK.jl  GPL (MI)LP\nGurobi Gurobi.jl Manual Comm. (MI)LP, (MI)SOCP\nHiGHS HiGHS.jl  MIT (MI)LP, QP\nHypatia.jl   MIT LP, SOCP, SDP\nIpopt Ipopt.jl  EPL LP, QP, NLP\nJuniper.jl   MIT (MI)SOCP, (MI)NLP\nLoraine.jl   MIT LP, SDP\nMadNLP.jl   MIT LP, QP, NLP\nMOSEK MosekTools.jl Manual Comm. (MI)LP, (MI)SOCP, SDP\nNLopt NLopt.jl  GPL LP, QP, NLP\nOSQP OSQP.jl  Apache LP, QP\nPATH PATHSolver.jl  MIT MCP\nPajarito.jl   MPL-2 (MI)NLP, (MI)SOCP, (MI)SDP\nPavito.jl   MPL-2 (MI)NLP\nPenbmi Penopt.jl  Comm. Bilinear SDP\nProxSDP.jl   MIT LP, SOCP, SDP\nRAPOSa AmplNLWriter.jl Manual RAPOSa (MI)NLP\nSCIP SCIP.jl  Apache (MI)LP, (MI)NLP\nSCS SCS.jl  MIT LP, QP, SOCP, SDP\nSDPA SDPA.jl, SDPAFamily.jl  GPL LP, SDP\nSDPNAL SDPNAL.jl Manualᴹ CC BY-SA LP, SDP\nSDPT3 SDPT3.jl Manualᴹ GPL LP, SOCP, SDP\nSeDuMi SeDuMi.jl Manualᴹ GPL LP, SOCP, SDP\nStatusSwitchingQP.jl   MIT LP, QP\nTulip.jl   MPL-2 LP","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Where:","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"LP = Linear programming\nQP = Quadratic programming\nSOCP = Second-order conic programming (including problems with convex quadratic constraints or objective)\nMCP = Mixed-complementarity programming\nNLP = Nonlinear programming\nSDP = Semidefinite programming\n(MI)XXX = Mixed-integer equivalent of problem type XXX","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"note: Note\nDeveloped a solver or solver wrapper? This table is open for new contributions. Edit the installation.md file, and use the checklist Adding a new solver to the documentation when opening the pull request.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"note: Note\nDeveloping a solver or solver wrapper? See Models and the MathOptInterface docs for more details on how JuMP interacts with solvers. Please get in touch via the Developer Chatroom with any questions about connecting new solvers with JuMP.","category":"page"},{"location":"installation/#AMPL-based-solvers","page":"Installation Guide","title":"AMPL-based solvers","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Use AmplNLWriter to access solvers that support the NL format.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Some solvers, such as Bonmin and Couenne can be installed via the Julia package manager. Others need to be manually installed.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Consult the AMPL documentation for a complete list of supported solvers.","category":"page"},{"location":"installation/#GAMS-based-solvers","page":"Installation Guide","title":"GAMS-based solvers","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Use GAMS.jl to access solvers available through GAMS. Such solvers include: AlphaECP, Antigone, BARON, CONOPT, Couenne, LocalSolver, PATHNLP, SHOT, SNOPT, SoPlex. See a complete list here.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"note: Note\nGAMS.jl requires an installation of the commercial software GAMS for which a free community license exists.","category":"page"},{"location":"installation/#NEOS-based-solvers","page":"Installation Guide","title":"NEOS-based solvers","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Use NEOSServer.jl to access solvers available through the NEOS Server.","category":"page"},{"location":"installation/#Common-installation-issues","page":"Installation Guide","title":"Common installation issues","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"tip: Tip\nWhen in doubt, run import Pkg; Pkg.update() to see if updating your packages fixes the issue. Remember you will need to exit Julia and start a new session for the changes to take effect.","category":"page"},{"location":"installation/#Check-the-version-of-your-packages","page":"Installation Guide","title":"Check the version of your packages","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Each package is versioned with a three-part number of the form vX.Y.Z. You can check which versions you have installed with import Pkg; Pkg.status().","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"This should almost always be the most-recent release. You can check the releases of a package by going to the relevant GitHub page, and navigating to the \"releases\" page. For example, the list of JuMP releases is available at: https://github.com/jump-dev/JuMP.jl/releases.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"If you post on the community forum, please include the output of Pkg.status().","category":"page"},{"location":"installation/#Unsatisfiable-requirements-detected","page":"Installation Guide","title":"Unsatisfiable requirements detected","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Did you get an error like Unsatisfiable requirements detected for package JuMP? The Pkg documentation has a section on how to understand and manage these conflicts.","category":"page"},{"location":"installation/#Installing-new-packages-can-make-JuMP-downgrade-to-an-earlier-version","page":"Installation Guide","title":"Installing new packages can make JuMP downgrade to an earlier version","text":"","category":"section"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"Another common complaint is that after adding a new package, code that previously worked no longer works.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"This usually happens because the new package is not compatible with the latest version of JuMP. Therefore, the package manager rolls-back JuMP to an earlier version. Here's an example.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"First, we add JuMP:","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"(jump_example) pkg> add JuMP\n  Resolving package versions...\nUpdating `~/jump_example/Project.toml`\n  [4076af6c] + JuMP v0.21.5\nUpdating `~/jump_example/Manifest.toml`\n  ... lines omitted ...","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"The + JuMP v0.21.5 line indicates that JuMP has been added at version 0.21.5. However, watch what happens when we add JuMPeR:","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"(jump_example) pkg> add JuMPeR\n  Resolving package versions...\nUpdating `~/jump_example/Project.toml`\n  [4076af6c] ↓ JuMP v0.21.5 ⇒ v0.18.6\n  [707a9f91] + JuMPeR v0.6.0\nUpdating `~/jump_example/Manifest.toml`\n  ... lines omitted ...","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"JuMPeR gets added at version 0.6.0 (+ JuMPeR v0.6.0), but JuMP gets downgraded from 0.21.5 to 0.18.6 (↓ JuMP v0.21.5 ⇒ v0.18.6)! The reason for this is that JuMPeR doesn't support a version of JuMP newer than 0.18.6.","category":"page"},{"location":"installation/","page":"Installation Guide","title":"Installation Guide","text":"tip: Tip\nPay careful attention to the output of the package manager when adding new packages, especially when you see a package being downgraded.","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/errors.md\"","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/errors/#Errors","page":"Errors","title":"Errors","text":"","category":"section"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"When an MOI call fails on a model, precise errors should be thrown when possible instead of simply calling error with a message. The docstrings for the respective methods describe the errors that the implementation should throw in certain situations. This error-reporting system allows code to distinguish between internal errors (that should be shown to the user) and unsupported operations which may have automatic workarounds.","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"When an invalid index is used in an MOI call, an InvalidIndex is thrown:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"InvalidIndex","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.InvalidIndex","page":"Errors","title":"MathOptInterface.InvalidIndex","text":"struct InvalidIndex{IndexType<:Index} <: Exception\n    index::IndexType\nend\n\nAn error indicating that the index index is invalid.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"When an invalid result index is used to retrieve an attribute, a ResultIndexBoundsError is thrown:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"ResultIndexBoundsError\ncheck_result_index_bounds","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.ResultIndexBoundsError","page":"Errors","title":"MathOptInterface.ResultIndexBoundsError","text":"struct ResultIndexBoundsError{AttrType} <: Exception\n    attr::AttrType\n    result_count::Int\nend\n\nAn error indicating that the requested attribute attr could not be retrieved, because the solver returned too few results compared to what was requested. For instance, the user tries to retrieve VariablePrimal(2) when only one solution is available, or when the model is infeasible and has no solution.\n\nSee also: check_result_index_bounds.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.check_result_index_bounds","page":"Errors","title":"MathOptInterface.check_result_index_bounds","text":"check_result_index_bounds(model::ModelLike, attr)\n\nThis function checks whether enough results are available in the model for the requested attr, using its result_index field. If the model does not have sufficient results to answer the query, it throws a ResultIndexBoundsError.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"As discussed in JuMP mapping, for scalar constraint with a nonzero function constant, a ScalarFunctionConstantNotZero exception may be thrown:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"ScalarFunctionConstantNotZero","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.ScalarFunctionConstantNotZero","page":"Errors","title":"MathOptInterface.ScalarFunctionConstantNotZero","text":"struct ScalarFunctionConstantNotZero{T, F, S} <: Exception\n    constant::T\nend\n\nAn error indicating that the constant part of the function in the constraint F-in-S is nonzero.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"Some VariableIndex constraints cannot be combined on the same variable:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"LowerBoundAlreadySet\nUpperBoundAlreadySet","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.LowerBoundAlreadySet","page":"Errors","title":"MathOptInterface.LowerBoundAlreadySet","text":"LowerBoundAlreadySet{S1, S2}\n\nError thrown when setting a VariableIndex-in-S2 when a VariableIndex-in-S1 has already been added and the sets S1, S2 both set a lower bound, i.e. they are EqualTo, GreaterThan, Interval, Semicontinuous or Semiinteger.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.UpperBoundAlreadySet","page":"Errors","title":"MathOptInterface.UpperBoundAlreadySet","text":"UpperBoundAlreadySet{S1, S2}\n\nError thrown when setting a VariableIndex-in-S2 when a VariableIndex-in-S1 has already been added and the sets S1, S2 both set an upper bound, i.e. they are EqualTo, LessThan, Interval, Semicontinuous or Semiinteger.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"As discussed in AbstractCallback, trying to get attributes inside a callback may throw:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"OptimizeInProgress","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.OptimizeInProgress","page":"Errors","title":"MathOptInterface.OptimizeInProgress","text":"struct OptimizeInProgress{AttrType<:AnyAttribute} <: Exception\n    attr::AttrType\nend\n\nError thrown from optimizer when MOI.get(optimizer, attr) is called inside an AbstractCallback while it is only defined once optimize! has completed. This can only happen when is_set_by_optimize(attr) is true.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"Trying to submit the wrong type of AbstractSubmittable inside an AbstractCallback (for example, a UserCut inside a LazyConstraintCallback) will throw:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"InvalidCallbackUsage","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.InvalidCallbackUsage","page":"Errors","title":"MathOptInterface.InvalidCallbackUsage","text":"struct InvalidCallbackUsage{C, S} <: Exception\n    callback::C\n    submittable::S\nend\n\nAn error indicating that submittable cannot be submitted inside callback.\n\nFor example, UserCut cannot be submitted inside LazyConstraintCallback.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"The rest of the errors defined in MOI fall in two categories represented by the following two abstract types:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"UnsupportedError\nNotAllowedError","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.UnsupportedError","page":"Errors","title":"MathOptInterface.UnsupportedError","text":"UnsupportedError <: Exception\n\nAbstract type for error thrown when an element is not supported by the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.NotAllowedError","page":"Errors","title":"MathOptInterface.NotAllowedError","text":"NotAllowedError <: Exception\n\nAbstract type for error thrown when an operation is supported but cannot be applied in the current state of the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"The different UnsupportedError and NotAllowedError are the following errors:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"UnsupportedAttribute\nSetAttributeNotAllowed\nAddVariableNotAllowed\nUnsupportedConstraint\nAddConstraintNotAllowed\nModifyConstraintNotAllowed\nModifyObjectiveNotAllowed\nDeleteNotAllowed\nUnsupportedSubmittable\nSubmitNotAllowed\nUnsupportedNonlinearOperator","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.UnsupportedAttribute","page":"Errors","title":"MathOptInterface.UnsupportedAttribute","text":"struct UnsupportedAttribute{AttrType} <: UnsupportedError\n    attr::AttrType\n    message::String\nend\n\nAn error indicating that the attribute attr is not supported by the model, i.e. that supports returns false.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.SetAttributeNotAllowed","page":"Errors","title":"MathOptInterface.SetAttributeNotAllowed","text":"struct SetAttributeNotAllowed{AttrType} <: NotAllowedError\n    attr::AttrType\n    message::String\nend\n\nAn error indicating that the attribute attr is supported (see supports) but cannot be set for some reason (see the error string).\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.AddVariableNotAllowed","page":"Errors","title":"MathOptInterface.AddVariableNotAllowed","text":"struct AddVariableNotAllowed <: NotAllowedError\n    message::String # Human-friendly explanation why the attribute cannot be set\nend\n\nAn error indicating that variables cannot be added to the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.UnsupportedConstraint","page":"Errors","title":"MathOptInterface.UnsupportedConstraint","text":"struct UnsupportedConstraint{F<:AbstractFunction, S<:AbstractSet} <: UnsupportedError\n    message::String # Human-friendly explanation why the attribute cannot be set\nend\n\nAn error indicating that constraints of type F-in-S are not supported by the model, i.e. that supports_constraint returns false.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.AddConstraintNotAllowed","page":"Errors","title":"MathOptInterface.AddConstraintNotAllowed","text":"struct AddConstraintNotAllowed{F<:AbstractFunction, S<:AbstractSet} <: NotAllowedError\n    message::String # Human-friendly explanation why the attribute cannot be set\nend\n\nAn error indicating that constraints of type F-in-S are supported (see supports_constraint) but cannot be added.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.ModifyConstraintNotAllowed","page":"Errors","title":"MathOptInterface.ModifyConstraintNotAllowed","text":"struct ModifyConstraintNotAllowed{F<:AbstractFunction, S<:AbstractSet,\n                                  C<:AbstractFunctionModification} <: NotAllowedError\n    constraint_index::ConstraintIndex{F, S}\n    change::C\n    message::String\nend\n\nAn error indicating that the constraint modification change cannot be applied to the constraint of index ci.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.ModifyObjectiveNotAllowed","page":"Errors","title":"MathOptInterface.ModifyObjectiveNotAllowed","text":"struct ModifyObjectiveNotAllowed{C<:AbstractFunctionModification} <: NotAllowedError\n    change::C\n    message::String\nend\n\nAn error indicating that the objective modification change cannot be applied to the objective.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.DeleteNotAllowed","page":"Errors","title":"MathOptInterface.DeleteNotAllowed","text":"struct DeleteNotAllowed{IndexType <: Index} <: NotAllowedError\n    index::IndexType\n    message::String\nend\n\nAn error indicating that the index index cannot be deleted.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.UnsupportedSubmittable","page":"Errors","title":"MathOptInterface.UnsupportedSubmittable","text":"struct UnsupportedSubmittable{SubmitType} <: UnsupportedError\n    sub::SubmitType\n    message::String\nend\n\nAn error indicating that the submittable sub is not supported by the model, i.e. that supports returns false.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.SubmitNotAllowed","page":"Errors","title":"MathOptInterface.SubmitNotAllowed","text":"struct SubmitNotAllowed{SubmitTyp<:AbstractSubmittable} <: NotAllowedError\n    sub::SubmitType\n    message::String\nend\n\nAn error indicating that the submittable sub is supported (see supports) but cannot be added for some reason (see the error string).\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/#MathOptInterface.UnsupportedNonlinearOperator","page":"Errors","title":"MathOptInterface.UnsupportedNonlinearOperator","text":"UnsupportedNonlinearOperator(head::Symbol[, message::String]) <: UnsupportedError\n\nAn error thrown by optimizers if they do not support the operator head in a ScalarNonlinearFunction.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> throw(MOI.UnsupportedNonlinearOperator(:black_box))\nERROR: MathOptInterface.UnsupportedNonlinearOperator: The nonlinear operator `:black_box` is not supported by the model.\nStacktrace:\n[...]\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"Note that setting the ConstraintFunction of a VariableIndex constraint is not allowed:","category":"page"},{"location":"moi/reference/errors/","page":"Errors","title":"Errors","text":"SettingVariableIndexNotAllowed","category":"page"},{"location":"moi/reference/errors/#MathOptInterface.SettingVariableIndexNotAllowed","page":"Errors","title":"MathOptInterface.SettingVariableIndexNotAllowed","text":"SettingVariableIndexNotAllowed()\n\nError type that should be thrown when the user calls set to change the ConstraintFunction of a VariableIndex constraint.\n\n\n\n\n\n","category":"type"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"EditURL = \"https://github.com/jump-dev/CPLEX.jl/blob/v1.0.0/README.md\"","category":"page"},{"location":"packages/CPLEX/#CPLEX.jl","page":"jump-dev/CPLEX.jl","title":"CPLEX.jl","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"CPLEX.jl is a wrapper for the IBM® ILOG® CPLEX® Optimization Studio.","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"CPLEX.jl has two components:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"The C API can be accessed via CPLEX.CPXxx functions, where the names and arguments are identical to the C API. See the CPLEX documentation for details.","category":"page"},{"location":"packages/CPLEX/#Affiliation","page":"jump-dev/CPLEX.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"This wrapper is maintained by the JuMP community and is not officially supported by IBM. However, we thank IBM for providing us with a CPLEX license to test CPLEX.jl on GitHub. If you are a commercial customer interested in official support for CPLEX in Julia, let them know.","category":"page"},{"location":"packages/CPLEX/#License","page":"jump-dev/CPLEX.jl","title":"License","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"CPLEX.jl is licensed under the MIT License.","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"The underlying solver is a closed-source commercial product for which you must purchase a license.","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"Free CPLEX licenses are available for academics and students.","category":"page"},{"location":"packages/CPLEX/#Installation","page":"jump-dev/CPLEX.jl","title":"Installation","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"CPLEX.jl requires CPLEX version 12.10, 20.1, or 22.1.","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"First, obtain a license of CPLEX and install CPLEX solver, following the instructions on IBM's website.","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"Once installed, set the CPLEX_STUDIO_BINARIES environment variable as appropriate and run Pkg.add(\"CPLEX\"). For example:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"# On Windows, this might be:\nENV[\"CPLEX_STUDIO_BINARIES\"] = \"C:\\\\Program Files\\\\CPLEX_Studio1210\\\\cplex\\\\bin\\\\x86-64_win\\\\\"\n# On OSX, this might be:\nENV[\"CPLEX_STUDIO_BINARIES\"] = \"/Applications/CPLEX_Studio1210/cplex/bin/x86-64_osx/\"\n# On Unix, this might be:\nENV[\"CPLEX_STUDIO_BINARIES\"] = \"/opt/CPLEX_Studio1210/cplex/bin/x86-64_linux/\"\n\nimport Pkg\nPkg.add(\"CPLEX\")","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"note: Note\nThe exact path may differ. Check which folder you installed CPLEX in, and update the path accordingly.","category":"page"},{"location":"packages/CPLEX/#Use-with-JuMP","page":"jump-dev/CPLEX.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"Use CPLEX.jl with JuMP as follows:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"using JuMP, CPLEX\nmodel = Model(CPLEX.Optimizer)\nset_attribute(model, \"CPX_PARAM_EPINT\", 1e-8)","category":"page"},{"location":"packages/CPLEX/#MathOptInterface-API","page":"jump-dev/CPLEX.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"The CPLEX optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}\nMOI.ObjectiveFunction{MOI.VariableIndex}\nMOI.ObjectiveFunction{MOI.VectorAffineFunction{Float64}}","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"MOI.Reals","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.Semicontinuous{Float64}\nMOI.VariableIndex in MOI.Semiinteger{Float64}\nMOI.VariableIndex in MOI.ZeroOne\nMOI.VectorOfVariables in MOI.SOS1{Float64}\nMOI.VectorOfVariables in MOI.SOS2{Float64}\nMOI.VectorOfVariables in MOI.SecondOrderCone","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"MOI.ConflictStatus()\nMOI.HeuristicCallback()\nMOI.LazyConstraintCallback()\nMOI.Name()\nMOI.ObjectiveSense()\nMOI.UserCutCallback()","category":"page"},{"location":"packages/CPLEX/#Options","page":"jump-dev/CPLEX.jl","title":"Options","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"Options match those of the C API in the CPLEX documentation.","category":"page"},{"location":"packages/CPLEX/#Callbacks","page":"jump-dev/CPLEX.jl","title":"Callbacks","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"CPLEX.jl provides a solver-specific callback to CPLEX:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"using JuMP, CPLEX, Test\n\nmodel = direct_model(CPLEX.Optimizer())\nset_silent(model)\n\n# This is very, very important!!! Only use callbacks in single-threaded mode.\nMOI.set(model, MOI.NumberOfThreads(), 1)\n\n@variable(model, 0 <= x <= 2.5, Int)\n@variable(model, 0 <= y <= 2.5, Int)\n@objective(model, Max, y)\ncb_calls = Clong[]\nfunction my_callback_function(cb_data::CPLEX.CallbackContext, context_id::Clong)\n    # You can reference variables outside the function as normal\n    push!(cb_calls, context_id)\n    # You can select where the callback is run\n    if context_id != CPX_CALLBACKCONTEXT_CANDIDATE\n        return\n    end\n    ispoint_p = Ref{Cint}()\n    ret = CPXcallbackcandidateispoint(cb_data, ispoint_p)\n    if ret != 0 || ispoint_p[] == 0\n        return  # No candidate point available or error\n    end\n    # You can query CALLBACKINFO items\n    valueP = Ref{Cdouble}()\n    ret = CPXcallbackgetinfodbl(cb_data, CPXCALLBACKINFO_BEST_BND, valueP)\n    @info \"Best bound is currently: $(valueP[])\"\n    # As well as any other C API\n    x_p = Vector{Cdouble}(undef, 2)\n    obj_p = Ref{Cdouble}()\n    ret = CPXcallbackgetincumbent(cb_data, x_p, 0, 1, obj_p)\n    if ret == 0\n        @info \"Objective incumbent is: $(obj_p[])\"\n        @info \"Incumbent solution is: $(x_p)\"\n        # Use CPLEX.column to map between variable references and the 1-based\n        # column.\n        x_col = CPLEX.column(cb_data, index(x))\n        @info \"x = $(x_p[x_col])\"\n    else\n        # Unable to query incumbent.\n    end\n\n    # Before querying `callback_value`, you must call:\n    CPLEX.load_callback_variable_primal(cb_data, context_id)\n    x_val = callback_value(cb_data, x)\n    y_val = callback_value(cb_data, y)\n    # You can submit solver-independent MathOptInterface attributes such as\n    # lazy constraints, user-cuts, and heuristic solutions.\n    if y_val - x_val > 1 + 1e-6\n        con = @build_constraint(y - x <= 1)\n        MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n    elseif y_val + x_val > 3 + 1e-6\n        con = @build_constraint(y + x <= 3)\n        MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n    end\nend\nMOI.set(model, CPLEX.CallbackFunction(), my_callback_function)\noptimize!(model)\n@test termination_status(model) == MOI.OPTIMAL\n@test primal_status(model) == MOI.FEASIBLE_POINT\n@test value(x) == 1\n@test value(y) == 2","category":"page"},{"location":"packages/CPLEX/#Annotations-for-automatic-Benders'-decomposition","page":"jump-dev/CPLEX.jl","title":"Annotations for automatic Benders' decomposition","text":"","category":"section"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"Here is an example of using the annotation feature for automatic Benders' decomposition:","category":"page"},{"location":"packages/CPLEX/","page":"jump-dev/CPLEX.jl","title":"jump-dev/CPLEX.jl","text":"using JuMP, CPLEX\n\nfunction add_annotation(\n    model::JuMP.Model,\n    variable_classification::Dict;\n    all_variables::Bool = true,\n)\n    num_variables = sum(length(it) for it in values(variable_classification))\n    if all_variables\n        @assert num_variables == JuMP.num_variables(model)\n    end\n    indices, annotations = CPXINT[], CPXLONG[]\n    for (key, value) in variable_classification\n        for variable_ref in value\n            push!(indices, variable_ref.index.value - 1)\n            push!(annotations, CPX_BENDERS_MASTERVALUE + key)\n        end\n    end\n    cplex = backend(model)\n    index_p = Ref{CPXINT}()\n    CPXnewlongannotation(\n        cplex.env,\n        cplex.lp,\n        CPX_BENDERS_ANNOTATION,\n        CPX_BENDERS_MASTERVALUE,\n    )\n    CPXgetlongannotationindex(\n        cplex.env,\n        cplex.lp,\n        CPX_BENDERS_ANNOTATION,\n        index_p,\n    )\n    CPXsetlongannotations(\n        cplex.env,\n        cplex.lp,\n        index_p[],\n        CPX_ANNOTATIONOBJ_COL,\n        length(indices),\n        indices,\n        annotations,\n    )\n    return\nend\n\n# Problem\n\nfunction illustrate_full_annotation()\n    c_1, c_2 = [1, 4], [2, 3]\n    dim_x, dim_y = length(c_1), length(c_2)\n    b = [-2; -3]\n    A_1, A_2 = [1 -3; -1 -3], [1 -2; -1 -1]\n    model = JuMP.direct_model(CPLEX.Optimizer())\n    set_optimizer_attribute(model, \"CPXPARAM_Benders_Strategy\", 1)\n    @variable(model, x[1:dim_x] >= 0, Bin)\n    @variable(model, y[1:dim_y] >= 0)\n    variable_classification = Dict(0 => [x[1], x[2]], 1 => [y[1], y[2]])\n    @constraint(model, A_2 * y + A_1 * x .<= b)\n    @objective(model, Min, c_1' * x + c_2' * y)\n    add_annotation(model, variable_classification)\n    optimize!(model)\n    x_optimal = value.(x)\n    y_optimal = value.(y)\n    println(\"x: $(x_optimal), y: $(y_optimal)\")\nend\n\nfunction illustrate_partial_annotation()\n    c_1, c_2 = [1, 4], [2, 3]\n    dim_x, dim_y = length(c_1), length(c_2)\n    b = [-2; -3]\n    A_1, A_2 = [1 -3; -1 -3], [1 -2; -1 -1]\n    model = JuMP.direct_model(CPLEX.Optimizer())\n    # Note that the \"CPXPARAM_Benders_Strategy\" has to be set to 2 if partial\n    # annotation is provided. If \"CPXPARAM_Benders_Strategy\" is set to 1, then\n    # the following error will be thrown:\n    # `CPLEX Error  2002: Invalid Benders decomposition.`\n    set_optimizer_attribute(model, \"CPXPARAM_Benders_Strategy\", 2)\n    @variable(model, x[1:dim_x] >= 0, Bin)\n    @variable(model, y[1:dim_y] >= 0)\n    variable_classification = Dict(0 => [x[1]], 1 => [y[1], y[2]])\n    @constraint(model, A_2 * y + A_1 * x .<= b)\n    @objective(model, Min, c_1' * x + c_2' * y)\n    add_annotation(model, variable_classification; all_variables = false)\n    optimize!(model)\n    x_optimal = value.(x)\n    y_optimal = value.(y)\n    println(\"x: $(x_optimal), y: $(y_optimal)\")\nend","category":"page"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"EditURL = nothing","category":"page"},{"location":"api/JuMP/#DocumenterReference_JuMP","page":"JuMP","title":"JuMP","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"This page lists the public API of JuMP.","category":"page"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"info: Info\nThis page is an unstructured list of the JuMP API. For a more structured overview, read the Manual or Tutorial parts of this documentation.","category":"page"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"Load all of the public the API into the current scope with:","category":"page"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"using JuMP","category":"page"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"Alternatively, load only the module with:","category":"page"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"import JuMP","category":"page"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"and then prefix all calls with JuMP. to create JuMP.<NAME>.","category":"page"},{"location":"api/JuMP/#@NLconstraint","page":"JuMP","title":"@NLconstraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@NLconstraint","category":"page"},{"location":"api/JuMP/#JuMP.@NLconstraint","page":"JuMP","title":"JuMP.@NLconstraint","text":"@NLconstraint(model::GenericModel, expr)\n\nAdd a constraint described by the nonlinear expression expr. See also @constraint. For example:\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @NLconstraint(model, sin(x) <= 1)\nsin(x) - 1.0 ≤ 0\n\njulia> @NLconstraint(model, [i = 1:3], sin(i * x) <= 1 / i)\n3-element Vector{NonlinearConstraintRef{ScalarShape}}:\n (sin(1.0 * x) - 1.0 / 1.0) - 0.0 ≤ 0\n (sin(2.0 * x) - 1.0 / 2.0) - 0.0 ≤ 0\n (sin(3.0 * x) - 1.0 / 3.0) - 0.0 ≤ 0\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@NLconstraints","page":"JuMP","title":"@NLconstraints","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@NLconstraints","category":"page"},{"location":"api/JuMP/#JuMP.@NLconstraints","page":"JuMP","title":"JuMP.@NLconstraints","text":"@NLconstraints(model, args...)\n\nAdds multiple nonlinear constraints to model at once, in the same fashion as the @NLconstraint macro.\n\nThe model must be the first argument, and multiple constraints can be added on multiple lines wrapped in a begin ... end block.\n\nThe macro returns a tuple containing the constraints that were defined.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, y);\n\njulia> @variable(model, t);\n\njulia> @variable(model, z[1:2]);\n\njulia> a = [4, 5];\n\njulia> @NLconstraints(model, begin\n           t >= sqrt(x^2 + y^2)\n           [i = 1:2], z[i] <= log(a[i])\n       end)\n((t - sqrt(x ^ 2.0 + y ^ 2.0)) - 0.0 ≥ 0, NonlinearConstraintRef{ScalarShape}[(z[1] - log(4.0)) - 0.0 ≤ 0, (z[2] - log(5.0)) - 0.0 ≤ 0])\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@NLexpression","page":"JuMP","title":"@NLexpression","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@NLexpression","category":"page"},{"location":"api/JuMP/#JuMP.@NLexpression","page":"JuMP","title":"JuMP.@NLexpression","text":"@NLexpression(args...)\n\nEfficiently build a nonlinear expression which can then be inserted in other nonlinear constraints and the objective. See also [@expression]. For example:\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> @NLexpression(model, my_expr, sin(x)^2 + cos(x^2))\nsubexpression[1]: sin(x) ^ 2.0 + cos(x ^ 2.0)\n\njulia> @NLconstraint(model, my_expr + y >= 5)\n(subexpression[1] + y) - 5.0 ≥ 0\n\njulia> @NLobjective(model, Min, my_expr)\n\nIndexing over sets and anonymous expressions are also supported:\n\njulia> @NLexpression(model, my_expr_1[i=1:3], sin(i * x))\n3-element Vector{NonlinearExpression}:\n subexpression[2]: sin(1.0 * x)\n subexpression[3]: sin(2.0 * x)\n subexpression[4]: sin(3.0 * x)\n\njulia> my_expr_2 = @NLexpression(model, log(1 + sum(exp(my_expr_1[i]) for i in 1:2)))\nsubexpression[5]: log(1.0 + (exp(subexpression[2]) + exp(subexpression[3])))\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@NLexpressions","page":"JuMP","title":"@NLexpressions","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@NLexpressions","category":"page"},{"location":"api/JuMP/#JuMP.@NLexpressions","page":"JuMP","title":"JuMP.@NLexpressions","text":"@NLexpressions(model, args...)\n\nAdds multiple nonlinear expressions to model at once, in the same fashion as the @NLexpression macro.\n\nThe model must be the first argument, and multiple expressions can be added on multiple lines wrapped in a begin ... end block.\n\nThe macro returns a tuple containing the expressions that were defined.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, y);\n\njulia> @variable(model, z[1:2]);\n\njulia> a = [4, 5];\n\njulia> @NLexpressions(model, begin\n           my_expr, sqrt(x^2 + y^2)\n           my_expr_1[i = 1:2], log(a[i]) - z[i]\n       end)\n(subexpression[1]: sqrt(x ^ 2.0 + y ^ 2.0), NonlinearExpression[subexpression[2]: log(4.0) - z[1], subexpression[3]: log(5.0) - z[2]])\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@NLobjective","page":"JuMP","title":"@NLobjective","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@NLobjective","category":"page"},{"location":"api/JuMP/#JuMP.@NLobjective","page":"JuMP","title":"JuMP.@NLobjective","text":"@NLobjective(model, sense, expression)\n\nAdd a nonlinear objective to model with optimization sense sense. sense must be Max or Min.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @NLobjective(model, Max, 2x + 1 + sin(x))\n\njulia> print(model)\nMax 2.0 * x + 1.0 + sin(x)\nSubject to\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@NLparameter","page":"JuMP","title":"@NLparameter","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@NLparameter","category":"page"},{"location":"api/JuMP/#JuMP.@NLparameter","page":"JuMP","title":"JuMP.@NLparameter","text":"@NLparameter(model, param == value)\n\nCreate and return a nonlinear parameter param attached to the model model with initial value set to value. Nonlinear parameters may be used only in nonlinear expressions.\n\nExample\n\njulia> model = Model();\n\njulia> @NLparameter(model, x == 10)\nx == 10.0\n\njulia> value(x)\n10.0\n\n@NLparameter(model, value = param_value)\n\nCreate and return an anonymous nonlinear parameter param attached to the model model with initial value set to param_value. Nonlinear parameters may be used only in nonlinear expressions.\n\nExample\n\njulia> model = Model();\n\njulia> x = @NLparameter(model, value = 10)\nparameter[1] == 10.0\n\njulia> value(x)\n10.0\n\n@NLparameter(model, param_collection[...] == value_expr)\n\nCreate and return a collection of nonlinear parameters param_collection attached to the model model with initial value set to value_expr (may depend on index sets). Uses the same syntax for specifying index sets as @variable.\n\nExample\n\njulia> model = Model();\n\njulia> @NLparameter(model, y[i = 1:3] == 2 * i)\n3-element Vector{NonlinearParameter}:\n parameter[1] == 2.0\n parameter[2] == 4.0\n parameter[3] == 6.0\n\njulia> value(y[2])\n4.0\n\n@NLparameter(model, [...] == value_expr)\n\nCreate and return an anonymous collection of nonlinear parameters attached to the model model with initial value set to value_expr (may depend on index sets). Uses the same syntax for specifying index sets as @variable.\n\nExample\n\njulia> model = Model();\n\njulia> y = @NLparameter(model, [i = 1:3] == 2 * i)\n3-element Vector{NonlinearParameter}:\n parameter[1] == 2.0\n parameter[2] == 4.0\n parameter[3] == 6.0\n\njulia> value(y[2])\n4.0\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@NLparameters","page":"JuMP","title":"@NLparameters","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@NLparameters","category":"page"},{"location":"api/JuMP/#JuMP.@NLparameters","page":"JuMP","title":"JuMP.@NLparameters","text":" @NLparameters(model, args...)\n\nCreate and return multiple nonlinear parameters attached to model model, in the same fashion as @NLparameter macro.\n\nThe model must be the first argument, and multiple parameters can be added on multiple lines wrapped in a begin ... end block. Distinct parameters need to be placed on separate lines as in the following example.\n\nThe macro returns a tuple containing the parameters that were defined.\n\nExample\n\njulia> model = Model();\n\njulia> @NLparameters(model, begin\n           x == 10\n           b == 156\n       end);\n\njulia> value(x)\n10.0\n\n\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@build_constraint","page":"JuMP","title":"@build_constraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@build_constraint","category":"page"},{"location":"api/JuMP/#JuMP.@build_constraint","page":"JuMP","title":"JuMP.@build_constraint","text":"@build_constraint(constraint_expr)\n\nConstructs a ScalarConstraint or VectorConstraint using the same machinery as @constraint but without adding the constraint to a model.\n\nConstraints using broadcast operators like x .<= 1 are also supported and will create arrays of ScalarConstraint or VectorConstraint.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @build_constraint(2x >= 1)\nScalarConstraint{AffExpr, MathOptInterface.GreaterThan{Float64}}(2 x, MathOptInterface.GreaterThan{Float64}(1.0))\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@constraint","page":"JuMP","title":"@constraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@constraint","category":"page"},{"location":"api/JuMP/#JuMP.@constraint","page":"JuMP","title":"JuMP.@constraint","text":"@constraint(m::GenericModel, expr, kw_args...)\n\nAdd a constraint described by the expression expr.\n\n@constraint(m::GenericModel, ref[i=..., j=..., ...], expr, kw_args...)\n\nAdd a group of constraints described by the expression expr parametrized by i, j, ...\n\nThe expression expr can either be\n\nof the form func in set constraining the function func to belong to the set set which is either a MOI.AbstractSet or one of the JuMP shortcuts SecondOrderCone, RotatedSecondOrderCone and PSDCone, e.g. @constraint(model, [1, x-1, y-2] in SecondOrderCone()) constrains the norm of [x-1, y-2] be less than 1;\nof the form a sign b, where sign is one of ==, ≥, >=, ≤ and <= building the single constraint enforcing the comparison to hold for the expression a and b, e.g. @constraint(m, x^2 + y^2 == 1) constrains x and y to lie on the unit circle;\nof the form a ≤ b ≤ c or a ≥ b ≥ c (where ≤ and <= (resp. ≥ and >=) can be used interchangeably) constraining the paired the expression b to lie between a and c;\nof the forms @constraint(m, a .sign b) or @constraint(m, a .sign b .sign c) which broadcast the constraint creation to each element of the vectors.\n\nThe recognized keyword arguments in kw_args are the following:\n\nbase_name: Sets the name prefix used to generate constraint names. It corresponds to the constraint name for scalar constraints, otherwise, the constraint names are set to base_name[...] for each index ... of the axes axes.\ncontainer: Specify the container type.\nset_string_name::Bool = true: control whether to set the MOI.ConstraintName attribute. Passing set_string_name = false can improve performance.\n\nNote for extending the constraint macro\n\nEach constraint will be created using add_constraint(m, build_constraint(_error, func, set)) where\n\n_error is an error function showing the constraint call in addition to the error message given as argument,\nfunc is the expression that is constrained\nand set is the set in which it is constrained to belong.\n\nFor expr of the first type (i.e. @constraint(m, func in set)), func and set are passed unchanged to build_constraint but for the other types, they are determined from the expressions and signs. For instance, @constraint(m, x^2 + y^2 == 1) is transformed into add_constraint(m, build_constraint(_error, x^2 + y^2, MOI.EqualTo(1.0))).\n\nTo extend JuMP to accept new constraints of this form, it is necessary to add the corresponding methods to build_constraint. Note that this will likely mean that either func or set will be some custom type, rather than e.g. a Symbol, since we will likely want to dispatch on the type of the function or set appearing in the constraint.\n\nFor extensions that need to create constraints with more information than just func and set, an additional positional argument can be specified to @constraint that will then be passed on build_constraint. Hence, we can enable this syntax by defining extensions of build_constraint(_error, func, set, my_arg; kw_args...). This produces the user syntax: @constraint(model, ref[...], expr, my_arg, kw_args...).\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@constraints","page":"JuMP","title":"@constraints","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@constraints","category":"page"},{"location":"api/JuMP/#JuMP.@constraints","page":"JuMP","title":"JuMP.@constraints","text":"@constraints(model, args...)\n\nAdds groups of constraints at once, in the same fashion as the @constraint macro.\n\nThe model must be the first argument, and multiple constraints can be added on multiple lines wrapped in a begin ... end block.\n\nThe macro returns a tuple containing the constraints that were defined.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, w);\n\njulia> @variable(model, x);\n\njulia> @variable(model, y);\n\njulia> @variable(model, z[1:3]);\n\njulia> @constraints(model, begin\n           x >= 1\n           y - w <= 2\n           sum_to_one[i=1:3], z[i] + y == 1\n       end);\n\njulia> print(model)\nFeasibility\nSubject to\n sum_to_one[1] : y + z[1] = 1\n sum_to_one[2] : y + z[2] = 1\n sum_to_one[3] : y + z[3] = 1\n x ≥ 1\n -w + y ≤ 2\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@expression","page":"JuMP","title":"@expression","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@expression","category":"page"},{"location":"api/JuMP/#JuMP.@expression","page":"JuMP","title":"JuMP.@expression","text":"@expression(args...)\n\nEfficiently builds a linear or quadratic expression but does not add to model immediately. Instead, returns the expression which can then be inserted in other constraints.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x[1:5]);\n\njulia> @variable(model, y);\n\njulia> @variable(model, z);\n\njulia> @expression(model, shared, sum(i * x[i] for i in 1:5))\nx[1] + 2 x[2] + 3 x[3] + 4 x[4] + 5 x[5]\n\njulia> @constraint(model, shared + y >= 5)\nx[1] + 2 x[2] + 3 x[3] + 4 x[4] + 5 x[5] + y ≥ 5\n\njulia> @constraint(model, shared + z <= 10)\nx[1] + 2 x[2] + 3 x[3] + 4 x[4] + 5 x[5] + z ≤ 10\n\nThe ref accepts index sets in the same way as @variable, and those indices can be used in the construction of the expressions:\n\njulia> @expression(model, expr[i = 1:3], i * sum(x[j] for j in 1:3))\n3-element Vector{AffExpr}:\n x[1] + x[2] + x[3]\n 2 x[1] + 2 x[2] + 2 x[3]\n 3 x[1] + 3 x[2] + 3 x[3]\n\nAnonymous syntax is also supported:\n\njulia> expr = @expression(model, [i in 1:3], i * sum(x[j] for j in 1:3))\n3-element Vector{AffExpr}:\n x[1] + x[2] + x[3]\n 2 x[1] + 2 x[2] + 2 x[3]\n 3 x[1] + 3 x[2] + 3 x[3]\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@expressions","page":"JuMP","title":"@expressions","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@expressions","category":"page"},{"location":"api/JuMP/#JuMP.@expressions","page":"JuMP","title":"JuMP.@expressions","text":"@expressions(model, args...)\n\nAdds multiple expressions to model at once, in the same fashion as the @expression macro.\n\nThe model must be the first argument, and multiple expressions can be added on multiple lines wrapped in a begin ... end block.\n\nThe macro returns a tuple containing the expressions that were defined.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, y);\n\njulia> @variable(model, z[1:2]);\n\njulia> a = [4, 5];\n\njulia> @expressions(model, begin\n           my_expr, x^2 + y^2\n           my_expr_1[i = 1:2], a[i] - z[i]\n       end)\n(x² + y², AffExpr[-z[1] + 4, -z[2] + 5])\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@objective","page":"JuMP","title":"@objective","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@objective","category":"page"},{"location":"api/JuMP/#JuMP.@objective","page":"JuMP","title":"JuMP.@objective","text":"@objective(model::GenericModel, sense, func)\n\nSet the objective sense to sense and objective function to func. The objective sense can be either Min, Max, MOI.MIN_SENSE, MOI.MAX_SENSE or MOI.FEASIBILITY_SENSE; see MOI.ObjectiveSense.\n\nIn order to set the sense programmatically, i.e., when sense is a Julia variable whose value is the sense, one of the three MOI.ObjectiveSense values should be used.\n\nExample\n\nTo minimize the value of the variable x, do as follows:\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @objective(model, Min, x)\nx\n\nTo maximize the value of the affine expression 2x - 1, do as follows:\n\njulia> @objective(model, Max, 2x - 1)\n2 x - 1\n\nTo set a quadratic objective and set the objective sense programmatically, do as follows:\n\njulia> sense = MIN_SENSE\nMIN_SENSE::OptimizationSense = 0\n\njulia> @objective(model, sense, x^2 - 2x + 1)\nx² - 2 x + 1\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@operator","page":"JuMP","title":"@operator","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@operator","category":"page"},{"location":"api/JuMP/#JuMP.@operator","page":"JuMP","title":"JuMP.@operator","text":"@operator(model, operator, dim, f[, ∇f[, ∇²f]])\n\nAdd the nonlinear operator operator in model with dim arguments, and create a new NonlinearOperator object called operator in the current scope.\n\nThe function f evaluates the operator and must return a scalar.\n\nThe optional function ∇f evaluates the first derivative, and the optional function ∇²f evaluates the second derivative.\n\n∇²f may be provided only if ∇f is also provided.\n\nUnivariate syntax\n\nIf dim == 1, then the method signatures of each function must be:\n\nf(::T)::T where {T<:Real}\n∇f(::T)::T where {T<:Real}\n∇²f(::T)::T where {T<:Real}\n\nMultivariate syntax\n\nIf dim > 1, then the method signatures of each function must be:\n\nf(x::T...)::T where {T<:Real}\n∇f(g::AbstractVector{T}, x::T...)::Nothing where {T<:Real}\n∇²f(H::AbstractMatrix{T}, x::T...)::Nothing where {T<:Real}\n\nWhere the gradient vector g and Hessian matrix H are filled in-place. For the Hessian, you must fill in the non-zero lower-triangular entries only. Setting an off-diagonal upper-triangular element may error.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> f(x::Float64) = x^2\nf (generic function with 1 method)\n\njulia> ∇f(x::Float64) = 2 * x\n∇f (generic function with 1 method)\n\njulia> ∇²f(x::Float64) = 2.0\n∇²f (generic function with 1 method)\n\njulia> @operator(model, op_f, 1, f, ∇f, ∇²f)\nNonlinearOperator(f, :op_f)\n\njulia> @objective(model, Min, op_f(x))\nop_f(x)\n\njulia> op_f(2.0)\n4.0\n\njulia> model[:op_f]\nNonlinearOperator(f, :op_f)\n\njulia> model[:op_f](x)\nop_f(x)\n\nNon-macro version\n\nThis macro is provided as helpful syntax that matches the style of the rest of the JuMP macros. However, you may also add operators outside the macro using add_nonlinear_operator. For example:\n\njulia> model = Model();\n\njulia> f(x) = x^2\nf (generic function with 1 method)\n\njulia> @operator(model, op_f, 1, f)\nNonlinearOperator(f, :op_f)\n\nis equivalent to\n\njulia> model = Model();\n\njulia> f(x) = x^2\nf (generic function with 1 method)\n\njulia> op_f = model[:op_f] = add_nonlinear_operator(model, 1, f; name = :op_f)\nNonlinearOperator(f, :op_f)\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@variable","page":"JuMP","title":"@variable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@variable","category":"page"},{"location":"api/JuMP/#JuMP.@variable","page":"JuMP","title":"JuMP.@variable","text":"@variable(model, expr, args..., kw_args...)\n\nAdd a variable to the model model described by the expression expr, the positional arguments args and the keyword arguments kw_args.\n\nAnonymous and named variables\n\nexpr must be one of the forms:\n\nOmitted, like @variable(model), which creates an anonymous variable\nA single symbol like @variable(model, x)\nA container expression like @variable(model, x[i=1:3])\nAn anoymous container expression like @variable(model, [i=1:3])\n\nBounds\n\nIn addition, the expression can have bounds, such as:\n\n@variable(model, x >= 0)\n@variable(model, x <= 0)\n@variable(model, x == 0)\n@variable(model, 0 <= x <= 1)\n\nand bounds can depend on the indices of the container expressions:\n\n@variable(model, -i <= x[i=1:3] <= i)\n\nSets\n\nYou can explicitly specify the set to which the variable belongs:\n\n@variable(model, x in MOI.Interval(0.0, 1.0))\n\nFor more information on this syntax, read Variables constrained on creation.\n\nPositional arguments\n\nThe recognized positional arguments in args are the following:\n\nBin: restricts the variable to the MOI.ZeroOne set, that is, {0, 1}. For example, @variable(model, x, Bin). Note: you cannot use @variable(model, Bin), use the binary keyword instead.\nInt: restricts the variable to the set of integers, that is, ..., -2, -1,  0, 1, 2, ... For example, @variable(model, x, Int). Note: you cannot use  @variable(model, Int), use the integer keyword instead.\nSymmetric: Only available when creating a square matrix of variables, i.e., when expr is of the form varname[1:n,1:n] or varname[i=1:n,j=1:n], it creates a symmetric matrix of variables.\nPSD: A restrictive extension to Symmetric which constraints a square matrix of variables to Symmetric and constrains to be positive semidefinite.\n\nKeyword arguments\n\nFour keyword arguments are useful in all cases:\n\nbase_name: Sets the name prefix used to generate variable names. It corresponds to the variable name for scalar variable, otherwise, the variable names are set to base_name[...] for each index ... of the axes axes.\nstart::Float64: specify the value passed to set_start_value for each variable\ncontainer: specify the container type. See Forcing the container type for more information.\nset_string_name::Bool = true: control whether to set the MOI.VariableName attribute. Passing set_string_name = false can improve performance.\n\nOther keyword arguments are needed to disambiguate sitations with anonymous variables:\n\nlower_bound::Float64: an alternative to x >= lb, sets the value of the variable lower bound.\nupper_bound::Float64: an alternative to x <= ub, sets the value of the variable upper bound.\nbinary::Bool: an alternative to passing Bin, sets whether the variable is binary or not.\ninteger::Bool: an alternative to passing Int, sets whether the variable is integer or not.\nset::MOI.AbstractSet: an alternative to using x in set\nvariable_type: used by JuMP extensions. See Extend @variable for more information.\n\nExample\n\nThe following are equivalent ways of creating a variable x of name x with lower bound 0:\n\njulia> model = Model();\n\njulia> @variable(model, x >= 0)\nx\n\njulia> model = Model();\n\njulia> @variable(model, x, lower_bound = 0)\nx\n\njulia> model = Model();\n\njulia> x = @variable(model, base_name = \"x\", lower_bound = 0)\nx\n\nOther examples:\n\njulia> model = Model();\n\njulia> @variable(model, x[i=1:3] <= i, Int, start = sqrt(i), lower_bound = -i)\n3-element Vector{VariableRef}:\n x[1]\n x[2]\n x[3]\n\njulia> @variable(model, y[i=1:3], container = DenseAxisArray, set = MOI.ZeroOne())\n1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:\n    Dimension 1, Base.OneTo(3)\nAnd data, a 3-element Vector{VariableRef}:\n y[1]\n y[2]\n y[3]\n\njulia> @variable(model, z[i=1:3], set_string_name = false)\n3-element Vector{VariableRef}:\n _[7]\n _[8]\n _[9]\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#@variables","page":"JuMP","title":"@variables","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.@variables","category":"page"},{"location":"api/JuMP/#JuMP.@variables","page":"JuMP","title":"JuMP.@variables","text":"@variables(model, args...)\n\nAdds multiple variables to model at once, in the same fashion as the @variable macro.\n\nThe model must be the first argument, and multiple variables can be added on multiple lines wrapped in a begin ... end block.\n\nThe macro returns a tuple containing the variables that were defined.\n\nExample\n\njulia> model = Model();\n\njulia> @variables(model, begin\n           x\n           y[i = 1:2] >= 0, (start = i)\n           z, Bin, (start = 0, base_name = \"Z\")\n       end)\n(x, VariableRef[y[1], y[2]], Z)\n\nnote: Note\nKeyword arguments must be contained within parentheses (refer to the example above).\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP/#add_bridge","page":"JuMP","title":"add_bridge","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_bridge","category":"page"},{"location":"api/JuMP/#JuMP.add_bridge","page":"JuMP","title":"JuMP.add_bridge","text":"add_bridge(\n    model::GenericModel{T},\n    BT::Type{<:MOI.Bridges.AbstractBridge};\n    coefficient_type::Type{S} = T,\n) where {T,S}\n\nAdd BT{T} to the list of bridges that can be used to transform unsupported constraints into an equivalent formulation using only constraints supported by the optimizer.\n\nSee also: remove_bridge.\n\nExample\n\njulia> model = Model();\n\njulia> add_bridge(model, MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge)\n\njulia> add_bridge(\n           model,\n           MOI.Bridges.Constraint.NumberConversionBridge;\n           coefficient_type = Complex{Float64}\n       )\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_constraint","page":"JuMP","title":"add_constraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_constraint","category":"page"},{"location":"api/JuMP/#JuMP.add_constraint","page":"JuMP","title":"JuMP.add_constraint","text":"add_constraint(model::GenericModel, con::AbstractConstraint, name::String=\"\")\n\nAdd a constraint con to Model model and sets its name.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_nonlinear_constraint","page":"JuMP","title":"add_nonlinear_constraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_nonlinear_constraint","category":"page"},{"location":"api/JuMP/#JuMP.add_nonlinear_constraint","page":"JuMP","title":"JuMP.add_nonlinear_constraint","text":"add_nonlinear_constraint(model::Model, expr::Expr)\n\nAdd a nonlinear constraint described by the Julia expression ex to model.\n\nThis function is most useful if the expression ex is generated programmatically, and you cannot use @NLconstraint.\n\nNotes\n\nYou must interpolate the variables directly into the expression expr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> add_nonlinear_constraint(model, :($(x) + $(x)^2 <= 1))\n(x + x ^ 2.0) - 1.0 ≤ 0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_nonlinear_expression","page":"JuMP","title":"add_nonlinear_expression","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_nonlinear_expression","category":"page"},{"location":"api/JuMP/#JuMP.add_nonlinear_expression","page":"JuMP","title":"JuMP.add_nonlinear_expression","text":"add_nonlinear_expression(model::Model, expr::Expr)\n\nAdd a nonlinear expression expr to model.\n\nThis function is most useful if the expression expr is generated programmatically, and you cannot use @NLexpression.\n\nNotes\n\nYou must interpolate the variables directly into the expression expr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> add_nonlinear_expression(model, :($(x) + $(x)^2))\nsubexpression[1]: x + x ^ 2.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_nonlinear_operator","page":"JuMP","title":"add_nonlinear_operator","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_nonlinear_operator","category":"page"},{"location":"api/JuMP/#JuMP.add_nonlinear_operator","page":"JuMP","title":"JuMP.add_nonlinear_operator","text":"add_nonlinear_operator(\n    model::Model,\n    dim::Int,\n    f::Function,\n    [∇f::Function,]\n    [∇²f::Function];\n    [name::Symbol = Symbol(f),]\n)\n\nAdd a new nonlinear operator with dim input arguments to model and associate it with the name name.\n\nThe function f evaluates the operator and must return a scalar.\n\nThe optional function ∇f evaluates the first derivative, and the optional function ∇²f evaluates the second derivative.\n\n∇²f may be provided only if ∇f is also provided.\n\nUnivariate syntax\n\nIf dim == 1, then the method signatures of each function must be:\n\nf(::T)::T where {T<:Real}\n∇f(::T)::T where {T<:Real}\n∇²f(::T)::T where {T<:Real}\n\nMultivariate syntax\n\nIf dim > 1, then the method signatures of each function must be:\n\nf(x::T...)::T where {T<:Real}\n∇f(g::AbstractVector{T}, x::T...)::Nothing where {T<:Real}\n∇²f(H::AbstractMatrix{T}, x::T...)::Nothing where {T<:Real}\n\nWhere the gradient vector g and Hessian matrix H are filled in-place. For the Hessian, you must fill in the non-zero lower-triangular entries only. Setting an off-diagonal upper-triangular element may error.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> f(x::Float64) = x^2\nf (generic function with 1 method)\n\njulia> ∇f(x::Float64) = 2 * x\n∇f (generic function with 1 method)\n\njulia> ∇²f(x::Float64) = 2.0\n∇²f (generic function with 1 method)\n\njulia> op_f = add_nonlinear_operator(model, 1, f, ∇f, ∇²f)\nNonlinearOperator(f, :f)\n\njulia> @objective(model, Min, op_f(x))\nf(x)\n\njulia> op_f(2.0)\n4.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_nonlinear_parameter","page":"JuMP","title":"add_nonlinear_parameter","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_nonlinear_parameter","category":"page"},{"location":"api/JuMP/#JuMP.add_nonlinear_parameter","page":"JuMP","title":"JuMP.add_nonlinear_parameter","text":"add_nonlinear_parameter(model::Model, value::Real)\n\nAdd an anonymous parameter to the model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_to_expression!","page":"JuMP","title":"add_to_expression!","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_to_expression!","category":"page"},{"location":"api/JuMP/#JuMP.add_to_expression!","page":"JuMP","title":"JuMP.add_to_expression!","text":"add_to_expression!(expression, terms...)\n\nUpdates expression in place to expression + (*)(terms...). This is typically much more efficient than expression += (*)(terms...). For example, add_to_expression!(expression, a, b) produces the same result as expression += a*b, and add_to_expression!(expression, a) produces the same result as expression += a.\n\nOnly a few methods are defined, mostly for internal use, and only for the cases when (1) they can be implemented efficiently and (2) expression is capable of storing the result. For example, add_to_expression!(::AffExpr, ::GenericVariableRef, ::GenericVariableRef) is not defined because a GenericAffExpr cannot store the product of two variables.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_to_function_constant","page":"JuMP","title":"add_to_function_constant","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_to_function_constant","category":"page"},{"location":"api/JuMP/#JuMP.add_to_function_constant","page":"JuMP","title":"JuMP.add_to_function_constant","text":"add_to_function_constant(constraint::ConstraintRef, value)\n\nAdd value to the function constant term.\n\nNote that for scalar constraints, JuMP will aggregate all constant terms onto the right-hand side of the constraint so instead of modifying the function, the set will be translated by -value. For example, given a constraint 2x <= 3, add_to_function_constant(c, 4) will modify it to 2x <= -1.\n\nExample\n\nFor scalar constraints, the set is translated by -value:\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con, 0 <= 2x - 1 <= 2)\ncon : 2 x ∈ [1, 3]\n\njulia> add_to_function_constant(con, 4)\n\njulia> con\ncon : 2 x ∈ [-3, -1]\n\nFor vector constraints, the constant is added to the function:\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, y);\n\njulia> @constraint(model, con, [x + y, x, y] in SecondOrderCone())\ncon : [x + y, x, y] ∈ MathOptInterface.SecondOrderCone(3)\n\njulia> add_to_function_constant(con, [1, 2, 2])\n\njulia> con\ncon : [x + y + 1, x + 2, y + 2] ∈ MathOptInterface.SecondOrderCone(3)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#add_variable","page":"JuMP","title":"add_variable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.add_variable","category":"page"},{"location":"api/JuMP/#JuMP.add_variable","page":"JuMP","title":"JuMP.add_variable","text":"add_variable(m::GenericModel, v::AbstractVariable, name::String=\"\")\n\nAdd a variable v to Model m and sets its name.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#all_constraints","page":"JuMP","title":"all_constraints","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.all_constraints","category":"page"},{"location":"api/JuMP/#JuMP.all_constraints","page":"JuMP","title":"JuMP.all_constraints","text":"all_constraints(model::GenericModel, function_type, set_type)::Vector{<:ConstraintRef}\n\nReturn a list of all constraints currently in the model where the function has type function_type and the set has type set_type. The constraints are ordered by creation time.\n\nSee also list_of_constraint_types and num_constraints.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x >= 0, Bin);\n\njulia> @constraint(model, 2x <= 1);\n\njulia> all_constraints(model, VariableRef, MOI.GreaterThan{Float64})\n1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:\n x ≥ 0\n\njulia> all_constraints(model, VariableRef, MOI.ZeroOne)\n1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}, ScalarShape}}:\n x binary\n\njulia> all_constraints(model, AffExpr, MOI.LessThan{Float64})\n1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n 2 x ≤ 1\n\n\n\n\n\nall_constraints(\n    model::GenericModel;\n    include_variable_in_set_constraints::Bool,\n)::Vector{ConstraintRef}\n\nReturn a list of all constraints in model.\n\nIf include_variable_in_set_constraints == true, then VariableRef constraints such as VariableRef-in-Integer are included. To return only the structural constraints (e.g., the rows in the constraint matrix of a linear program), pass include_variable_in_set_constraints = false.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x >= 0, Int);\n\njulia> @constraint(model, 2x <= 1);\n\njulia> @NLconstraint(model, x^2 <= 1);\n\njulia> all_constraints(model; include_variable_in_set_constraints = true)\n4-element Vector{ConstraintRef}:\n 2 x ≤ 1\n x ≥ 0\n x integer\n x ^ 2.0 - 1.0 ≤ 0\n\njulia> all_constraints(model; include_variable_in_set_constraints = false)\n2-element Vector{ConstraintRef}:\n 2 x ≤ 1\n x ^ 2.0 - 1.0 ≤ 0\n\nPerformance considerations\n\nNote that this function is type-unstable because it returns an abstractly typed vector. If performance is a problem, consider using list_of_constraint_types and a function barrier. See the Performance tips for extensions section of the documentation for more details.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#all_nonlinear_constraints","page":"JuMP","title":"all_nonlinear_constraints","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.all_nonlinear_constraints","category":"page"},{"location":"api/JuMP/#JuMP.all_nonlinear_constraints","page":"JuMP","title":"JuMP.all_nonlinear_constraints","text":"all_nonlinear_constraints(model::GenericModel)\n\nReturn a vector of all nonlinear constraint references in the model in the order they were added to the model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#all_variables","page":"JuMP","title":"all_variables","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.all_variables","category":"page"},{"location":"api/JuMP/#JuMP.all_variables","page":"JuMP","title":"JuMP.all_variables","text":"all_variables(model::GenericModel{T})::Vector{GenericVariableRef{T}} where {T}\n\nReturns a list of all variables currently in the model. The variables are ordered by creation time.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, y);\n\njulia> all_variables(model)\n2-element Vector{VariableRef}:\n x\n y\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#anonymous_name","page":"JuMP","title":"anonymous_name","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.anonymous_name","category":"page"},{"location":"api/JuMP/#JuMP.anonymous_name","page":"JuMP","title":"JuMP.anonymous_name","text":"anonymous_name(::MIME, x::AbstractVariableRef)\n\nThe name to use for an anonymous variable x when printing.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#backend","page":"JuMP","title":"backend","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.backend","category":"page"},{"location":"api/JuMP/#JuMP.backend","page":"JuMP","title":"JuMP.backend","text":"backend(model::GenericModel)\n\nReturn the lower-level MathOptInterface model that sits underneath JuMP. This model depends on which operating mode JuMP is in (see mode).\n\nIf JuMP is in DIRECT mode (i.e., the model was created using direct_model), the backend will be the optimizer passed to direct_model.\nIf JuMP is in MANUAL or AUTOMATIC mode, the backend is a MOI.Utilities.CachingOptimizer.\n\nThis function should only be used by advanced users looking to access low-level MathOptInterface or solver-specific functionality.\n\nNotes\n\nIf JuMP is not in DIRECT mode, the type returned by backend may change between any JuMP releases. Therefore, only use the public API exposed by MathOptInterface, and do not access internal fields. If you require access to the innermost optimizer, see unsafe_backend. Alternatively, use direct_model to create a JuMP model in DIRECT mode.\n\nSee also: unsafe_backend.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#barrier_iterations","page":"JuMP","title":"barrier_iterations","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.barrier_iterations","category":"page"},{"location":"api/JuMP/#JuMP.barrier_iterations","page":"JuMP","title":"JuMP.barrier_iterations","text":"barrier_iterations(model::GenericModel)\n\nGets the cumulative number of barrier iterations during the most recent optimization.\n\nSolvers must implement MOI.BarrierIterations() to use this function.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#bridge_constraints","page":"JuMP","title":"bridge_constraints","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.bridge_constraints","category":"page"},{"location":"api/JuMP/#JuMP.bridge_constraints","page":"JuMP","title":"JuMP.bridge_constraints","text":"bridge_constraints(model::GenericModel)\n\nWhen in direct mode, return false.\n\nWhen in manual or automatic mode, return a Bool indicating whether the optimizer is set and unsupported constraints are automatically bridged to equivalent supported constraints when an appropriate transformation is available.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#build_constraint","page":"JuMP","title":"build_constraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.build_constraint","category":"page"},{"location":"api/JuMP/#JuMP.build_constraint","page":"JuMP","title":"JuMP.build_constraint","text":"build_constraint(\n    _error::Function,\n    Q::LinearAlgebra.Symmetric{V, M},\n    ::PSDCone,\n) where {V<:AbstractJuMPScalar,M<:AbstractMatrix{V}}\n\nReturn a VectorConstraint of shape SymmetricMatrixShape constraining the matrix Q to be positive semidefinite.\n\nThis function is used by the @constraint macros as follows:\n\njulia> import LinearAlgebra\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2]);\n\njulia> @constraint(model, LinearAlgebra.Symmetric(Q) in PSDCone())\n[Q[1,1]  Q[1,2];\n Q[1,2]  Q[2,2]] ∈ PSDCone()\n\nThe form above is usually used when the entries of Q are affine or quadratic expressions, but it can also be used when the entries are variables to get the reference of the semidefinite constraint, e.g.,\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2], Symmetric)\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n Q[1,1]  Q[1,2]\n Q[1,2]  Q[2,2]\n\njulia> @constraint(model, Q in PSDCone())\n[Q[1,1]  Q[1,2];\n Q[1,2]  Q[2,2]] ∈ PSDCone()\n\n\n\n\n\nbuild_constraint(\n    _error::Function,\n    Q::AbstractMatrix{<:AbstractJuMPScalar},\n    ::PSDCone,\n)\n\nReturn a VectorConstraint of shape SquareMatrixShape constraining the matrix Q to be symmetric and positive semidefinite.\n\nThis function is used by the @constraint macro as follows:\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2]);\n\njulia> @constraint(model, Q in PSDCone())\n[Q[1,1]  Q[1,2];\n Q[2,1]  Q[2,2]] ∈ PSDCone()\n\n\n\n\n\nbuild_constraint(\n    _error::Function,\n    Q::LinearAlgebra.Hermitian{V,M},\n    ::HermitianPSDCone,\n) where {V<:AbstractJuMPScalar,M<:AbstractMatrix{V}}\n\nReturn a VectorConstraint of shape HermitianMatrixShape constraining the matrix Q to be Hermitian positive semidefinite.\n\nThis function is used by the @constraint macros as follows:\n\njulia> import LinearAlgebra\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2]);\n\njulia> @constraint(model, LinearAlgebra.Hermitian(Q) in HermitianPSDCone())\n[Q[1,1]  Q[1,2];\n Q[1,2]  Q[2,2]] ∈ HermitianPSDCone()\n\n\n\n\n\nbuild_constraint(\n    _error::Function,\n    f::AbstractVector{<:AbstractJuMPScalar},\n    ::Nonnegatives,\n    extra::Union{MOI.AbstractVectorSet,AbstractVectorSet},\n)\n\nA helper method that re-writes\n\n@constraint(model, X >= Y, extra)\n\ninto\n\n@constraint(model, X - Y in extra)\n\n\n\n\n\nbuild_constraint(\n    _error::Function,\n    f::AbstractVector{<:AbstractJuMPScalar},\n    ::Nonpositives,\n    extra::Union{MOI.AbstractVectorSet,AbstractVectorSet},\n)\n\nA helper method that re-writes\n\n@constraint(model, Y <= X, extra)\n\ninto\n\n@constraint(model, X - Y in extra)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#build_variable","page":"JuMP","title":"build_variable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.build_variable","category":"page"},{"location":"api/JuMP/#JuMP.build_variable","page":"JuMP","title":"JuMP.build_variable","text":"build_variable(\n    _error::Function,\n    info::VariableInfo,\n    args...;\n    kwargs...,\n)\n\nReturn a new AbstractVariable object.\n\nThis method should only be implemented by developers creating JuMP extensions. It should never be called by users of JuMP.\n\nArguments\n\n_error: a function to call instead of error. _error annotates the error message with additional information for the user.\ninfo: an instance of VariableInfo. This has a variety of fields relating to the variable such as info.lower_bound and info.binary.\nargs: optional additional positional arguments for extending the @variable macro.\nkwargs: optional keyword arguments for extending the @variable macro.\n\nSee also: @variable\n\nwarning: Warning\nExtensions should define a method with ONE positional argument to dispatch the call to a different method. Creating an extension that relies on multiple positional arguments leads to MethodErrors if the user passes the arguments in the wrong order.\n\nExample\n\n@variable(model, x, Foo)\n\nwill call\n\nbuild_variable(_error::Function, info::VariableInfo, ::Type{Foo})\n\nPassing special-case positional arguments such as Bin, Int, and PSD is okay, along with keyword arguments:\n\n@variable(model, x, Int, Foo(), mykwarg = true)\n# or\n@variable(model, x, Foo(), Int, mykwarg = true)\n\nwill call\n\nbuild_variable(_error::Function, info::VariableInfo, ::Foo; mykwarg)\n\nand info.integer will be true.\n\nNote that the order of the positional arguments does not matter.\n\n\n\n\n\nbuild_variable(_error::Function, variables, ::SymmetricMatrixSpace)\n\nReturn a VariablesConstrainedOnCreation of shape SymmetricMatrixShape creating variables in MOI.Reals, i.e. \"free\" variables unless they are constrained after their creation.\n\nThis function is used by the @variable macro as follows:\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2], Symmetric)\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n Q[1,1]  Q[1,2]\n Q[1,2]  Q[2,2]\n\n\n\n\n\nbuild_variable(_error::Function, variables, ::SkewSymmetricMatrixSpace)\n\nReturn a VariablesConstrainedOnCreation of shape SkewSymmetricMatrixShape creating variables in MOI.Reals, i.e. \"free\" variables unless they are constrained after their creation.\n\nThis function is used by the @variable macro as follows:\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2] in SkewSymmetricMatrixSpace())\n2×2 Matrix{AffExpr}:\n 0        Q[1,2]\n -Q[1,2]  0\n\n\n\n\n\nbuild_variable(_error::Function, variables, ::HermitianMatrixSpace)\n\nReturn a VariablesConstrainedOnCreation of shape HermitianMatrixShape creating variables in MOI.Reals, i.e. \"free\" variables unless they are constrained after their creation.\n\nThis function is used by the @variable macro as follows:\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2] in HermitianMatrixSpace())\n2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(Q[1,1])                    real(Q[1,2]) + imag(Q[1,2]) im\n real(Q[1,2]) - imag(Q[1,2]) im  real(Q[2,2])\n\n\n\n\n\nbuild_variable(_error::Function, variables, ::PSDCone)\n\nReturn a VariablesConstrainedOnCreation of shape SymmetricMatrixShape constraining the variables to be positive semidefinite.\n\nThis function is used by the @variable macro as follows:\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2], PSD)\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n Q[1,1]  Q[1,2]\n Q[1,2]  Q[2,2]\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#callback_node_status","page":"JuMP","title":"callback_node_status","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.callback_node_status","category":"page"},{"location":"api/JuMP/#JuMP.callback_node_status","page":"JuMP","title":"JuMP.callback_node_status","text":"callback_node_status(cb_data, model::GenericModel)\n\nReturn an MOI.CallbackNodeStatusCode enum, indicating if the current primal solution available from callback_value is integer feasible.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#callback_value","page":"JuMP","title":"callback_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.callback_value","category":"page"},{"location":"api/JuMP/#JuMP.callback_value","page":"JuMP","title":"JuMP.callback_value","text":"callback_value(cb_data, x::GenericVariableRef)\n\nReturn the primal solution of a variable inside a callback.\n\ncb_data is the argument to the callback function, and the type is dependent on the solver.\n\n\n\n\n\ncallback_value(cb_data, expr::Union{GenericAffExpr, GenericQuadExpr})\n\nReturn the primal solution of an affine or quadratic expression inside a callback by getting the value for each variable appearing in the expression.\n\ncb_data is the argument to the callback function, and the type is dependent on the solver.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#check_belongs_to_model","page":"JuMP","title":"check_belongs_to_model","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.check_belongs_to_model","category":"page"},{"location":"api/JuMP/#JuMP.check_belongs_to_model","page":"JuMP","title":"JuMP.check_belongs_to_model","text":"check_belongs_to_model(func::AbstractJuMPScalar, model::AbstractModel)\n\nThrow VariableNotOwned if the owner_model of one of the variables of the function func is not model.\n\ncheck_belongs_to_model(constraint::AbstractConstraint, model::AbstractModel)\n\nThrow VariableNotOwned if the owner_model of one of the variables of the constraint constraint is not model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#coefficient","page":"JuMP","title":"coefficient","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.coefficient","category":"page"},{"location":"api/JuMP/#JuMP.coefficient","page":"JuMP","title":"JuMP.coefficient","text":"coefficient(v1::GenericVariableRef{T}, v2::GenericVariableRef{T}) where {T}\n\nReturn one(T) if v1 == v2, and zero(T) otherwise.\n\nThis is a fallback for other coefficient methods to simplify code in which the expression may be a single variable.\n\n\n\n\n\ncoefficient(a::GenericAffExpr{C,V}, v::V) where {C,V}\n\nReturn the coefficient associated with variable v in the affine expression a.\n\n\n\n\n\ncoefficient(a::GenericAffExpr{C,V}, v1::V, v2::V) where {C,V}\n\nReturn the coefficient associated with the term v1 * v2 in the quadratic expression a.\n\nNote that coefficient(a, v1, v2) is the same as coefficient(a, v2, v1).\n\n\n\n\n\ncoefficient(a::GenericQuadExpr{C,V}, v::V) where {C,V}\n\nReturn the coefficient associated with variable v in the affine component of a.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#compute_conflict!","page":"JuMP","title":"compute_conflict!","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.compute_conflict!","category":"page"},{"location":"api/JuMP/#JuMP.compute_conflict!","page":"JuMP","title":"JuMP.compute_conflict!","text":"compute_conflict!(model::GenericModel)\n\nCompute a conflict if the model is infeasible. If an optimizer has not been set yet (see set_optimizer), a NoOptimizer error is thrown.\n\nThe status of the conflict can be checked with the MOI.ConflictStatus model attribute. Then, the status for each constraint can be queried with the MOI.ConstraintConflictStatus attribute.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#constant","page":"JuMP","title":"constant","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.constant","category":"page"},{"location":"api/JuMP/#JuMP.constant","page":"JuMP","title":"JuMP.constant","text":"constant(aff::GenericAffExpr{C, V})::C\n\nReturn the constant of the affine expression.\n\n\n\n\n\nconstant(aff::GenericQuadExpr{C, V})::C\n\nReturn the constant of the quadratic expression.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#constraint_by_name","page":"JuMP","title":"constraint_by_name","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.constraint_by_name","category":"page"},{"location":"api/JuMP/#JuMP.constraint_by_name","page":"JuMP","title":"JuMP.constraint_by_name","text":"constraint_by_name(model::AbstractModel,\n                   name::String)::Union{ConstraintRef, Nothing}\n\nReturn the reference of the constraint with name attribute name or Nothing if no constraint has this name attribute. Throws an error if several constraints have name as their name attribute.\n\nconstraint_by_name(model::AbstractModel,\n                   name::String,\n                   F::Type{<:Union{AbstractJuMPScalar,\n                                   Vector{<:AbstractJuMPScalar},\n                                   MOI.AbstactFunction}},\n                   S::Type{<:MOI.AbstractSet})::Union{ConstraintRef, Nothing}\n\nSimilar to the method above, except that it throws an error if the constraint is not an F-in-S contraint where F is either the JuMP or MOI type of the function, and S is the MOI type of the set. This method is recommended if you know the type of the function and set since its returned type can be inferred while for the method above (i.e. without F and S), the exact return type of the constraint index cannot be inferred.\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, con, x^2 == 1)\ncon : x² = 1\n\njulia> constraint_by_name(model, \"kon\")\n\njulia> constraint_by_name(model, \"con\")\ncon : x² = 1\n\njulia> constraint_by_name(model, \"con\", AffExpr, MOI.EqualTo{Float64})\n\njulia> constraint_by_name(model, \"con\", QuadExpr, MOI.EqualTo{Float64})\ncon : x² = 1\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#constraint_object","page":"JuMP","title":"constraint_object","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.constraint_object","category":"page"},{"location":"api/JuMP/#JuMP.constraint_object","page":"JuMP","title":"JuMP.constraint_object","text":"constraint_object(con_ref::ConstraintRef)\n\nReturn the underlying constraint data for the constraint referenced by ref.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#constraint_ref_with_index","page":"JuMP","title":"constraint_ref_with_index","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.constraint_ref_with_index","category":"page"},{"location":"api/JuMP/#JuMP.constraint_ref_with_index","page":"JuMP","title":"JuMP.constraint_ref_with_index","text":"constraint_ref_with_index(model::AbstractModel, index::MOI.ConstraintIndex)\n\nReturn a ConstraintRef of model corresponding to index.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#constraint_string","page":"JuMP","title":"constraint_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.constraint_string","category":"page"},{"location":"api/JuMP/#JuMP.constraint_string","page":"JuMP","title":"JuMP.constraint_string","text":"constraint_string(\n    mode::MIME,\n    ref::ConstraintRef;\n    in_math_mode::Bool = false)\n\nReturn a string representation of the constraint ref, given the mode.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#constraints_string","page":"JuMP","title":"constraints_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.constraints_string","category":"page"},{"location":"api/JuMP/#JuMP.constraints_string","page":"JuMP","title":"JuMP.constraints_string","text":"constraints_string(mode, model::AbstractModel)::Vector{String}\n\nReturn a list of Strings describing each constraint of the model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#copy_conflict","page":"JuMP","title":"copy_conflict","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.copy_conflict","category":"page"},{"location":"api/JuMP/#JuMP.copy_conflict","page":"JuMP","title":"JuMP.copy_conflict","text":"copy_conflict(model::GenericModel)\n\nReturn a copy of the current conflict for the model model and a GenericReferenceMap that can be used to obtain the variable and constraint reference of the new model corresponding to a given model's reference.\n\nThis is a convenience function that provides a filtering function for copy_model.\n\nNote\n\nModel copy is not supported in DIRECT mode, i.e. when a model is constructed using the direct_model constructor instead of the Model constructor. Moreover, independently on whether an optimizer was provided at model construction, the new model will have no optimizer, i.e., an optimizer will have to be provided to the new model in the optimize! call.\n\nExample\n\nIn the following example, a model model is constructed with a variable x and two constraints c1 and c2. This model has no solution, as the two constraints are mutually exclusive. The solver is asked to compute a conflict with compute_conflict!. The parts of model participating in the conflict are then copied into a model iis_model.\n\njulia> using JuMP\n\njulia> import Gurobi\n\njulia> model = Model(Gurobi.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 0)\nx\n\njulia> @constraint(model, c1, x >= 2)\nc1 : x ≥ 2\n\njulia> @constraint(model, c2, x <= 1)\nc2 : x ≤ 1\n\njulia> optimize!(model)\n\njulia> compute_conflict!(model)\n\njulia> if get_attribute(model, MOI.ConflictStatus()) == MOI.CONFLICT_FOUND\n           iis_model, reference_map = copy_conflict(model)\n           print(iis_model)\n       end\nFeasibility\nSubject to\n c1 : x ≥ 2\n c2 : x ≤ 1\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#copy_extension_data","page":"JuMP","title":"copy_extension_data","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.copy_extension_data","category":"page"},{"location":"api/JuMP/#JuMP.copy_extension_data","page":"JuMP","title":"JuMP.copy_extension_data","text":"copy_extension_data(data, new_model::AbstractModel, model::AbstractModel)\n\nReturn a copy of the extension data data of the model model to the extension data of the new model new_model.\n\nA method should be added for any JuMP extension storing data in the ext field.\n\nwarning: Warning\nDo not engage in type piracy by implementing this method for types of data that you did not define! JuMP extensions should store types that they define in model.ext, rather than regular Julia types.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#copy_model","page":"JuMP","title":"copy_model","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.copy_model","category":"page"},{"location":"api/JuMP/#JuMP.copy_model","page":"JuMP","title":"JuMP.copy_model","text":"copy_model(model::GenericModel; filter_constraints::Union{Nothing, Function}=nothing)\n\nReturn a copy of the model model and a GenericReferenceMap that can be used to obtain the variable and constraint reference of the new model corresponding to a given model's reference. A Base.copy(::AbstractModel) method has also been implemented, it is similar to copy_model but does not return the reference map.\n\nIf the filter_constraints argument is given, only the constraints for which this function returns true will be copied. This function is given a constraint reference as argument.\n\nNote\n\nModel copy is not supported in DIRECT mode, i.e. when a model is constructed using the direct_model constructor instead of the Model constructor. Moreover, independently on whether an optimizer was provided at model construction, the new model will have no optimizer, i.e., an optimizer will have to be provided to the new model in the optimize! call.\n\nExample\n\nIn the following example, a model model is constructed with a variable x and a constraint cref. It is then copied into a model new_model with the new references assigned to x_new and cref_new.\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, cref, x == 2)\ncref : x = 2\n\njulia> new_model, reference_map = copy_model(model);\n\njulia> x_new = reference_map[x]\nx\n\njulia> cref_new = reference_map[cref]\ncref : x = 2\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#delete","page":"JuMP","title":"delete","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.delete","category":"page"},{"location":"api/JuMP/#JuMP.delete","page":"JuMP","title":"JuMP.delete","text":"delete(model::GenericModel, con_ref::ConstraintRef)\n\nDelete the constraint associated with constraint_ref from the model model.\n\nNote that delete does not unregister the name from the model, so adding a new constraint of the same name will throw an error. Use unregister to unregister the name after deletion.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, c, 2x <= 1)\nc : 2 x ≤ 1\n\njulia> delete(model, c)\n\njulia> unregister(model, :c)\n\njulia> print(model)\nFeasibility\nSubject to\n\njulia> model[:c]\nERROR: KeyError: key :c not found\nStacktrace:\n[...]\n\n\n\n\n\ndelete(model::GenericModel, con_refs::Vector{<:ConstraintRef})\n\nDelete the constraints associated with con_refs from the model model. Solvers may implement specialized methods for deleting multiple constraints of the same concrete type, i.e., when isconcretetype(eltype(con_refs)). These may be more efficient than repeatedly calling the single constraint delete method.\n\nSee also: unregister\n\n\n\n\n\ndelete(model::GenericModel, variable_ref::GenericVariableRef)\n\nDelete the variable associated with variable_ref from the model model.\n\nNote that delete does not unregister the name from the model, so adding a new variable of the same name will throw an error. Use unregister to unregister the name after deletion.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> delete(model, x)\n\njulia> unregister(model, :x)\n\njulia> print(model)\nFeasibility\nSubject to\n\njulia> model[:x]\nERROR: KeyError: key :x not found\nStacktrace:\n[...]\n\n\n\n\n\ndelete(model::GenericModel, variable_refs::Vector{<:GenericVariableRef})\n\nDelete the variables associated with variable_refs from the model model. Solvers may implement methods for deleting multiple variables that are more efficient than repeatedly calling the single variable delete method.\n\nSee also: unregister\n\n\n\n\n\ndelete(model::Model, c::NonlinearConstraintRef)\n\nDelete the nonlinear constraint c from model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#delete_lower_bound","page":"JuMP","title":"delete_lower_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.delete_lower_bound","category":"page"},{"location":"api/JuMP/#JuMP.delete_lower_bound","page":"JuMP","title":"JuMP.delete_lower_bound","text":"delete_lower_bound(v::GenericVariableRef)\n\nDelete the lower bound constraint of a variable.\n\nSee also LowerBoundRef, has_lower_bound, lower_bound, set_lower_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x >= 1.0);\n\njulia> has_lower_bound(x)\ntrue\n\njulia> delete_lower_bound(x)\n\njulia> has_lower_bound(x)\nfalse\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#delete_upper_bound","page":"JuMP","title":"delete_upper_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.delete_upper_bound","category":"page"},{"location":"api/JuMP/#JuMP.delete_upper_bound","page":"JuMP","title":"JuMP.delete_upper_bound","text":"delete_upper_bound(v::GenericVariableRef)\n\nDelete the upper bound constraint of a variable.\n\nErrors if one does not exist.\n\nSee also UpperBoundRef, has_upper_bound, upper_bound, set_upper_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x <= 1.0);\n\njulia> has_upper_bound(x)\ntrue\n\njulia> delete_upper_bound(x)\n\njulia> has_upper_bound(x)\nfalse\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#direct_generic_model","page":"JuMP","title":"direct_generic_model","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.direct_generic_model","category":"page"},{"location":"api/JuMP/#JuMP.direct_generic_model","page":"JuMP","title":"JuMP.direct_generic_model","text":"direct_generic_model(\n    value_type::Type{T},\n    backend::MOI.ModelLike;\n) where {T<:Real}\n\nReturn a new JuMP model using backend to store the model and solve it.\n\nAs opposed to the Model constructor, no cache of the model is stored outside of backend and no bridges are automatically applied to backend.\n\nNotes\n\nThe absence of a cache reduces the memory footprint but, it is important to bear in mind the following implications of creating models using this direct mode:\n\nWhen backend does not support an operation, such as modifying constraints or adding variables/constraints after solving, an error is thrown. For models created using the Model constructor, such situations can be dealt with by storing the modifications in a cache and loading them into the optimizer when optimize! is called.\nNo constraint bridging is supported by default.\nThe optimizer used cannot be changed the model is constructed.\nThe model created cannot be copied.\n\n\n\n\n\ndirect_generic_model(::Type{T}, factory::MOI.OptimizerWithAttributes)\n\nCreate a direct_generic_model using factory, a MOI.OptimizerWithAttributes object created by optimizer_with_attributes.\n\nExample\n\njulia> import HiGHS\n\njulia> optimizer = optimizer_with_attributes(\n           HiGHS.Optimizer,\n           \"presolve\" => \"off\",\n           MOI.Silent() => true,\n       );\n\njulia> model = direct_generic_model(Float64, optimizer)\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: DIRECT\nSolver name: HiGHS\n\nis equivalent to:\n\njulia> import HiGHS\n\njulia> model = direct_generic_model(Float64, HiGHS.Optimizer())\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: DIRECT\nSolver name: HiGHS\n\njulia> set_attribute(model, \"presolve\", \"off\")\n\njulia> set_attribute(model, MOI.Silent(), true)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#direct_model","page":"JuMP","title":"direct_model","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.direct_model","category":"page"},{"location":"api/JuMP/#JuMP.direct_model","page":"JuMP","title":"JuMP.direct_model","text":"direct_model(backend::MOI.ModelLike)\n\nReturn a new JuMP model using backend to store the model and solve it.\n\nAs opposed to the Model constructor, no cache of the model is stored outside of backend and no bridges are automatically applied to backend.\n\nNotes\n\nThe absence of a cache reduces the memory footprint but, it is important to bear in mind the following implications of creating models using this direct mode:\n\nWhen backend does not support an operation, such as modifying constraints or adding variables/constraints after solving, an error is thrown. For models created using the Model constructor, such situations can be dealt with by storing the modifications in a cache and loading them into the optimizer when optimize! is called.\nNo constraint bridging is supported by default.\nThe optimizer used cannot be changed the model is constructed.\nThe model created cannot be copied.\n\n\n\n\n\ndirect_model(factory::MOI.OptimizerWithAttributes)\n\nCreate a direct_model using factory, a MOI.OptimizerWithAttributes object created by optimizer_with_attributes.\n\nExample\n\njulia> import HiGHS\n\njulia> optimizer = optimizer_with_attributes(\n           HiGHS.Optimizer,\n           \"presolve\" => \"off\",\n           MOI.Silent() => true,\n       );\n\njulia> model = direct_model(optimizer)\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: DIRECT\nSolver name: HiGHS\n\nis equivalent to:\n\njulia> import HiGHS\n\njulia> model = direct_model(HiGHS.Optimizer())\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: DIRECT\nSolver name: HiGHS\n\njulia> set_attribute(model, \"presolve\", \"off\")\n\njulia> set_attribute(model, MOI.Silent(), true)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#drop_zeros!","page":"JuMP","title":"drop_zeros!","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.drop_zeros!","category":"page"},{"location":"api/JuMP/#JuMP.drop_zeros!","page":"JuMP","title":"JuMP.drop_zeros!","text":"drop_zeros!(expr::GenericAffExpr)\n\nRemove terms in the affine expression with 0 coefficients.\n\n\n\n\n\ndrop_zeros!(expr::GenericQuadExpr)\n\nRemove terms in the quadratic expression with 0 coefficients.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#dual","page":"JuMP","title":"dual","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.dual","category":"page"},{"location":"api/JuMP/#JuMP.dual","page":"JuMP","title":"JuMP.dual","text":"dual(con_ref::ConstraintRef; result::Int = 1)\n\nReturn the dual value of constraint con_ref associated with result index result of the most-recent solution returned by the solver.\n\nUse has_dual to check if a result exists before asking for values.\n\nSee also: result_count, shadow_price.\n\n\n\n\n\ndual(c::NonlinearConstraintRef)\n\nReturn the dual of the nonlinear constraint c.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#dual_objective_value","page":"JuMP","title":"dual_objective_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.dual_objective_value","category":"page"},{"location":"api/JuMP/#JuMP.dual_objective_value","page":"JuMP","title":"JuMP.dual_objective_value","text":"dual_objective_value(model::GenericModel; result::Int = 1)\n\nReturn the value of the objective of the dual problem associated with result index result of the most-recent solution returned by the solver.\n\nThrows MOI.UnsupportedAttribute{MOI.DualObjectiveValue} if the solver does not support this attribute.\n\nSee also: result_count.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#dual_shape","page":"JuMP","title":"dual_shape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.dual_shape","category":"page"},{"location":"api/JuMP/#JuMP.dual_shape","page":"JuMP","title":"JuMP.dual_shape","text":"dual_shape(shape::AbstractShape)::AbstractShape\n\nReturns the shape of the dual space of the space of objects of shape shape. By default, the dual_shape of a shape is itself. See the examples section below for an example for which this is not the case.\n\nExample\n\nConsider polynomial constraints for which the dual is moment constraints and moment constraints for which the dual is polynomial constraints. Shapes for polynomials can be defined as follows:\n\nstruct Polynomial\n    coefficients::Vector{Float64}\n    monomials::Vector{Monomial}\nend\nstruct PolynomialShape <: AbstractShape\n    monomials::Vector{Monomial}\nend\nJuMP.reshape_vector(x::Vector, shape::PolynomialShape) = Polynomial(x, shape.monomials)\n\nand a shape for moments can be defined as follows:\n\nstruct Moments\n    coefficients::Vector{Float64}\n    monomials::Vector{Monomial}\nend\nstruct MomentsShape <: AbstractShape\n    monomials::Vector{Monomial}\nend\nJuMP.reshape_vector(x::Vector, shape::MomentsShape) = Moments(x, shape.monomials)\n\nThen dual_shape allows the definition of the shape of the dual of polynomial and moment constraints:\n\ndual_shape(shape::PolynomialShape) = MomentsShape(shape.monomials)\ndual_shape(shape::MomentsShape) = PolynomialShape(shape.monomials)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#dual_start_value","page":"JuMP","title":"dual_start_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.dual_start_value","category":"page"},{"location":"api/JuMP/#JuMP.dual_start_value","page":"JuMP","title":"JuMP.dual_start_value","text":"dual_start_value(con_ref::ConstraintRef)\n\nReturn the dual start value (MOI attribute ConstraintDualStart) of the constraint con_ref.\n\nNote: If no dual start value has been set, dual_start_value will return nothing.\n\nSee also set_dual_start_value.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#dual_status","page":"JuMP","title":"dual_status","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.dual_status","category":"page"},{"location":"api/JuMP/#JuMP.dual_status","page":"JuMP","title":"JuMP.dual_status","text":"dual_status(model::GenericModel; result::Int = 1)\n\nReturn a MOI.ResultStatusCode describing the status of the most recent dual solution of the solver (i.e., the MOI.DualStatus attribute) associated with the result index result.\n\nSee also: result_count.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#error_if_direct_mode","page":"JuMP","title":"error_if_direct_mode","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.error_if_direct_mode","category":"page"},{"location":"api/JuMP/#JuMP.error_if_direct_mode","page":"JuMP","title":"JuMP.error_if_direct_mode","text":"error_if_direct_mode(model::GenericModel, func::Symbol)\n\nErrors if model is in direct mode during a call from the function named func.\n\nUsed internally within JuMP, or by JuMP extensions who do not want to support models in direct mode.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#fix","page":"JuMP","title":"fix","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.fix","category":"page"},{"location":"api/JuMP/#JuMP.fix","page":"JuMP","title":"JuMP.fix","text":"fix(v::GenericVariableRef, value::Number; force::Bool = false)\n\nFix a variable to a value. Update the fixing constraint if one exists, otherwise create a new one.\n\nIf the variable already has variable bounds and force=false, calling fix will throw an error. If force=true, existing variable bounds will be deleted, and the fixing constraint will be added. Note a variable will have no bounds after a call to unfix.\n\nSee also FixRef, is_fixed, fix_value, unfix.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> is_fixed(x)\nfalse\n\njulia> fix(x, 1.0)\n\njulia> is_fixed(x)\ntrue\n\njulia> model = Model();\n\njulia> @variable(model, 0 <= x <= 1);\n\njulia> is_fixed(x)\nfalse\n\njulia> fix(x, 1.0; force = true)\n\njulia> is_fixed(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#fix_discrete_variables","page":"JuMP","title":"fix_discrete_variables","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.fix_discrete_variables","category":"page"},{"location":"api/JuMP/#JuMP.fix_discrete_variables","page":"JuMP","title":"JuMP.fix_discrete_variables","text":"fix_discrete_variables([var_value::Function = value,] model::GenericModel)\n\nModifies model to convert all binary and integer variables to continuous variables with fixed bounds of var_value(x).\n\nReturn\n\nReturns a function that can be called without any arguments to restore the original model. The behavior of this function is undefined if additional changes are made to the affected variables in the meantime.\n\nNotes\n\nAn error is thrown if semi-continuous or semi-integer constraints are present (support may be added for these in the future).\nAll other constraints are ignored (left in place). This includes discrete constraints like SOS and indicator constraints.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x, Bin, start = 1);\n\njulia> @variable(model, 1 <= y <= 10, Int, start = 2);\n\njulia> @objective(model, Min, x + y);\n\njulia> undo_relax = fix_discrete_variables(start_value, model);\n\njulia> print(model)\nMin x + y\nSubject to\n x = 1\n y = 2\n\njulia> undo_relax()\n\njulia> print(model)\nMin x + y\nSubject to\n y ≥ 1\n y ≤ 10\n y integer\n x binary\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#fix_value","page":"JuMP","title":"fix_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.fix_value","category":"page"},{"location":"api/JuMP/#JuMP.fix_value","page":"JuMP","title":"JuMP.fix_value","text":"fix_value(v::GenericVariableRef)\n\nReturn the value to which a variable is fixed.\n\nError if one does not exist.\n\nSee also FixRef, is_fixed, fix, unfix.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x == 1);\n\njulia> fix_value(x)\n1.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#flatten!","page":"JuMP","title":"flatten!","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.flatten!","category":"page"},{"location":"api/JuMP/#JuMP.flatten!","page":"JuMP","title":"JuMP.flatten!","text":"flatten!(expr::GenericNonlinearExpr)\n\nFlatten a nonlinear expression in-place by lifting nested + and * nodes into a single n-ary operation.\n\nMotivation\n\nNonlinear expressions created using operator overloading can be deeply nested and unbalanced. For example, prod(x for i in 1:4) creates *(x, *(x, *(x, x))) instead of the more preferable *(x, x, x, x).\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> y = prod(x for i in 1:4)\n((x²) * x) * x\n\njulia> flatten!(y)\n(x²) * x * x\n\njulia> flatten!(sin(prod(x for i in 1:4)))\nsin((x²) * x * x)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#function_string","page":"JuMP","title":"function_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.function_string","category":"page"},{"location":"api/JuMP/#JuMP.function_string","page":"JuMP","title":"JuMP.function_string","text":"function_string(\n    mode::MIME,\n    func::Union{JuMP.AbstractJuMPScalar,Vector{<:JuMP.AbstractJuMPScalar}},\n)\n\nReturn a String representing the function func using print mode mode.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#get_attribute","page":"JuMP","title":"get_attribute","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.get_attribute","category":"page"},{"location":"api/JuMP/#JuMP.get_attribute","page":"JuMP","title":"JuMP.get_attribute","text":"get_attribute(model::GenericModel, attr::MOI.AbstractModelAttribute)\nget_attribute(x::GenericVariableRef, attr::MOI.AbstractVariableAttribute)\nget_attribute(cr::ConstraintRef, attr::MOI.AbstractConstraintAttribute)\n\nGet the value of a solver-specifc attribute attr.\n\nThis is equivalent to calling MOI.get with the associated MOI model and, for variables and constraints, with the associated MOI.VariableIndex or MOI.ConstraintIndex.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, c, 2 * x <= 1)\nc : 2 x ≤ 1\n\njulia> get_attribute(model, MOI.Name())\n\"\"\n\njulia> get_attribute(x, MOI.VariableName())\n\"x\"\n\njulia> get_attribute(c, MOI.ConstraintName())\n\"c\"\n\n\n\n\n\nget_attribute(\n    model::Union{GenericModel,MOI.OptimizerWithAttributes},\n    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},\n)\n\nGet the value of a solver-specifc attribute attr.\n\nThis is equivalent to calling MOI.get with the associated MOI model.\n\nIf attr is an AbstractString, it is converted to MOI.RawOptimizerAttribute.\n\nExample\n\njulia> import HiGHS\n\njulia> opt = optimizer_with_attributes(HiGHS.Optimizer, \"output_flag\" => true);\n\njulia> model = Model(opt);\n\njulia> get_attribute(model, \"output_flag\")\ntrue\n\njulia> get_attribute(model, MOI.RawOptimizerAttribute(\"output_flag\"))\ntrue\n\njulia> get_attribute(opt, \"output_flag\")\ntrue\n\njulia> get_attribute(opt, MOI.RawOptimizerAttribute(\"output_flag\"))\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#get_optimizer_attribute","page":"JuMP","title":"get_optimizer_attribute","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.get_optimizer_attribute","category":"page"},{"location":"api/JuMP/#JuMP.get_optimizer_attribute","page":"JuMP","title":"JuMP.get_optimizer_attribute","text":"get_optimizer_attribute(\n    model::Union{GenericModel,MOI.OptimizerWithAttributes},\n    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},\n)\n\nReturn the value associated with the solver-specific attribute attr.\n\nIf attr is an AbstractString, this is equivalent to get_optimizer_attribute(model, MOI.RawOptimizerAttribute(name)).\n\ncompat: Compat\nThis method will remain in all v1.X releases of JuMP, but it may be removed in a future v2.0 release. We recommend using get_attribute instead.\n\nSee also: set_optimizer_attribute, set_optimizer_attributes.\n\nExample\n\njulia> import Ipopt\n\njulia> model = Model(Ipopt.Optimizer);\n\njulia> get_optimizer_attribute(model, MOI.Silent())\nfalse\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#has_duals","page":"JuMP","title":"has_duals","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.has_duals","category":"page"},{"location":"api/JuMP/#JuMP.has_duals","page":"JuMP","title":"JuMP.has_duals","text":"has_duals(model::GenericModel; result::Int = 1)\n\nReturn true if the solver has a dual solution in result index result available to query, otherwise return false.\n\nSee also dual, shadow_price, and result_count.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#has_lower_bound","page":"JuMP","title":"has_lower_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.has_lower_bound","category":"page"},{"location":"api/JuMP/#JuMP.has_lower_bound","page":"JuMP","title":"JuMP.has_lower_bound","text":"has_lower_bound(v::GenericVariableRef)\n\nReturn true if v has a lower bound. If true, the lower bound can be queried with lower_bound.\n\nSee also LowerBoundRef, lower_bound, set_lower_bound, delete_lower_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x >= 1.0);\n\njulia> has_lower_bound(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#has_start_value","page":"JuMP","title":"has_start_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.has_start_value","category":"page"},{"location":"api/JuMP/#JuMP.has_start_value","page":"JuMP","title":"JuMP.has_start_value","text":"has_start_value(variable::AbstractVariableRef)\n\nReturn true if the variable has a start value set otherwise return false.\n\nSee also set_start_value.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#has_upper_bound","page":"JuMP","title":"has_upper_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.has_upper_bound","category":"page"},{"location":"api/JuMP/#JuMP.has_upper_bound","page":"JuMP","title":"JuMP.has_upper_bound","text":"has_upper_bound(v::GenericVariableRef)\n\nReturn true if v has a upper bound. If true, the upper bound can be queried with upper_bound.\n\nSee also UpperBoundRef, upper_bound, set_upper_bound, delete_upper_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x <= 1.0);\n\njulia> has_upper_bound(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#has_values","page":"JuMP","title":"has_values","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.has_values","category":"page"},{"location":"api/JuMP/#JuMP.has_values","page":"JuMP","title":"JuMP.has_values","text":"has_values(model::GenericModel; result::Int = 1)\n\nReturn true if the solver has a primal solution in result index result available to query, otherwise return false.\n\nSee also value and result_count.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#in_set_string","page":"JuMP","title":"in_set_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.in_set_string","category":"page"},{"location":"api/JuMP/#JuMP.in_set_string","page":"JuMP","title":"JuMP.in_set_string","text":"in_set_string(mode::MIME, set)\n\nReturn a String representing the membership to the set set using print mode mode.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#index","page":"JuMP","title":"index","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.index","category":"page"},{"location":"api/JuMP/#JuMP.index","page":"JuMP","title":"JuMP.index","text":"index(cr::ConstraintRef)::MOI.ConstraintIndex\n\nReturn the index of the constraint that corresponds to cr in the MOI backend.\n\n\n\n\n\nindex(v::GenericVariableRef)::MOI.VariableIndex\n\nReturn the index of the variable that corresponds to v in the MOI backend.\n\n\n\n\n\nindex(p::NonlinearParameter)::MOI.Nonlinear.ParameterIndex\n\nReturn the index of the nonlinear parameter associated with p.\n\n\n\n\n\nindex(ex::NonlinearExpression)::MOI.Nonlinear.ExpressionIndex\n\nReturn the index of the nonlinear expression associated with ex.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#is_binary","page":"JuMP","title":"is_binary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.is_binary","category":"page"},{"location":"api/JuMP/#JuMP.is_binary","page":"JuMP","title":"JuMP.is_binary","text":"is_binary(v::GenericVariableRef)\n\nReturn true if v is constrained to be binary.\n\nSee also BinaryRef, set_binary, unset_binary.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x, Bin);\n\njulia> is_binary(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#is_fixed","page":"JuMP","title":"is_fixed","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.is_fixed","category":"page"},{"location":"api/JuMP/#JuMP.is_fixed","page":"JuMP","title":"JuMP.is_fixed","text":"is_fixed(v::GenericVariableRef)\n\nReturn true if v is a fixed variable. If true, the fixed value can be queried with fix_value.\n\nSee also FixRef, fix_value, fix, unfix.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> is_fixed(x)\nfalse\n\njulia> fix(x, 1.0)\n\njulia> is_fixed(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#is_integer","page":"JuMP","title":"is_integer","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.is_integer","category":"page"},{"location":"api/JuMP/#JuMP.is_integer","page":"JuMP","title":"JuMP.is_integer","text":"is_integer(v::GenericVariableRef)\n\nReturn true if v is constrained to be integer.\n\nSee also IntegerRef, set_integer, unset_integer.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> is_integer(x)\nfalse\n\njulia> set_integer(x)\n\njulia> is_integer(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#is_parameter","page":"JuMP","title":"is_parameter","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.is_parameter","category":"page"},{"location":"api/JuMP/#JuMP.is_parameter","page":"JuMP","title":"JuMP.is_parameter","text":"is_parameter(x::GenericVariableRef)::Bool\n\nReturn true if x is constrained to be a parameter.\n\nSee also ParameterRef, set_parameter_value, parameter_value.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, p in Parameter(2))\np\n\njulia> is_parameter(p)\ntrue\n\njulia> @variable(model, x)\nx\n\njulia> is_parameter(x)\nfalse\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#is_valid","page":"JuMP","title":"is_valid","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.is_valid","category":"page"},{"location":"api/JuMP/#JuMP.is_valid","page":"JuMP","title":"JuMP.is_valid","text":"is_valid(model::GenericModel, con_ref::ConstraintRef{<:AbstractModel})\n\nReturn true if constraint_ref refers to a valid constraint in model.\n\n\n\n\n\nis_valid(model::GenericModel, variable_ref::GenericVariableRef)\n\nReturn true if variable refers to a valid variable in model.\n\n\n\n\n\nis_valid(model::Model, c::NonlinearConstraintRef)\n\nReturn true if c refers to a valid nonlinear constraint in model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#isequal_canonical","page":"JuMP","title":"isequal_canonical","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.isequal_canonical","category":"page"},{"location":"api/JuMP/#JuMP.isequal_canonical","page":"JuMP","title":"JuMP.isequal_canonical","text":"isequal_canonical(\n    aff::GenericAffExpr{C,V},\n    other::GenericAffExpr{C,V}\n) where {C,V}\n\nReturn true if aff is equal to other after dropping zeros and disregarding the order. Mainly useful for testing.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#jump_function","page":"JuMP","title":"jump_function","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.jump_function","category":"page"},{"location":"api/JuMP/#JuMP.jump_function","page":"JuMP","title":"JuMP.jump_function","text":"jump_function(x)\n\nGiven an MathOptInterface object x, return the JuMP equivalent.\n\nSee also: moi_function.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#jump_function_type","page":"JuMP","title":"jump_function_type","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.jump_function_type","category":"page"},{"location":"api/JuMP/#JuMP.jump_function_type","page":"JuMP","title":"JuMP.jump_function_type","text":"jump_function_type(::Type{T}) where {T}\n\nGiven an MathOptInterface object type T, return the JuMP equivalent.\n\nSee also: moi_function_type.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#latex_formulation","page":"JuMP","title":"latex_formulation","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.latex_formulation","category":"page"},{"location":"api/JuMP/#JuMP.latex_formulation","page":"JuMP","title":"JuMP.latex_formulation","text":"latex_formulation(model::AbstractModel)\n\nWrap model in a type so that it can be pretty-printed as text/latex in a notebook like IJulia, or in Documenter.\n\nTo render the model, end the cell with latex_formulation(model), or call display(latex_formulation(model)) in to force the display of the model from inside a function.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#linear_terms","page":"JuMP","title":"linear_terms","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.linear_terms","category":"page"},{"location":"api/JuMP/#JuMP.linear_terms","page":"JuMP","title":"JuMP.linear_terms","text":"linear_terms(aff::GenericAffExpr{C, V})\n\nProvides an iterator over coefficient-variable tuples (a_i::C, x_i::V) in the linear part of the affine expression.\n\n\n\n\n\nlinear_terms(quad::GenericQuadExpr{C, V})\n\nProvides an iterator over tuples (coefficient::C, variable::V) in the linear part of the quadratic expression.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#list_of_constraint_types","page":"JuMP","title":"list_of_constraint_types","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.list_of_constraint_types","category":"page"},{"location":"api/JuMP/#JuMP.list_of_constraint_types","page":"JuMP","title":"JuMP.list_of_constraint_types","text":"list_of_constraint_types(model::GenericModel)::Vector{Tuple{Type,Type}}\n\nReturn a list of tuples of the form (F, S) where F is a JuMP function type and S is an MOI set type such that all_constraints(model, F, S) returns a nonempty list.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x >= 0, Bin);\n\njulia> @constraint(model, 2x <= 1);\n\njulia> list_of_constraint_types(model)\n3-element Vector{Tuple{Type, Type}}:\n (AffExpr, MathOptInterface.LessThan{Float64})\n (VariableRef, MathOptInterface.GreaterThan{Float64})\n (VariableRef, MathOptInterface.ZeroOne)\n\nPerformance considerations\n\nIterating over the list of function and set types is a type-unstable operation. Consider using a function barrier. See the Performance tips for extensions section of the documentation for more details.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#lower_bound","page":"JuMP","title":"lower_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.lower_bound","category":"page"},{"location":"api/JuMP/#JuMP.lower_bound","page":"JuMP","title":"JuMP.lower_bound","text":"lower_bound(v::GenericVariableRef)\n\nReturn the lower bound of a variable. Error if one does not exist.\n\nSee also LowerBoundRef, has_lower_bound, set_lower_bound, delete_lower_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x >= 1.0);\n\njulia> lower_bound(x)\n1.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#lp_sensitivity_report","page":"JuMP","title":"lp_sensitivity_report","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.lp_sensitivity_report","category":"page"},{"location":"api/JuMP/#JuMP.lp_sensitivity_report","page":"JuMP","title":"JuMP.lp_sensitivity_report","text":"lp_sensitivity_report(model::GenericModel{T}; atol::T = Base.rtoldefault(T))::SensitivityReport{T} where {T}\n\nGiven a linear program model with a current optimal basis, return a SensitivityReport object, which maps:\n\nEvery variable reference to a tuple (d_lo, d_hi)::Tuple{T,T}, explaining how much the objective coefficient of the corresponding variable can change by, such that the original basis remains optimal.\nEvery constraint reference to a tuple (d_lo, d_hi)::Tuple{T,T}, explaining how much the right-hand side of the corresponding constraint can change by, such that the basis remains optimal.\n\nBoth tuples are relative, rather than absolute. So given a objective coefficient of 1.0 and a tuple (-0.5, 0.5), the objective coefficient can range between 1.0 - 0.5 an 1.0 + 0.5.\n\natol is the primal/dual optimality tolerance, and should match the tolerance of the solver used to compute the basis.\n\nNote: interval constraints are NOT supported.\n\nExample\n\njulia> import HiGHS\n\njulia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, -1 <= x <= 2)\nx\n\njulia> @objective(model, Min, x)\nx\n\njulia> optimize!(model)\n\njulia> report = lp_sensitivity_report(model; atol = 1e-7);\n\njulia> dx_lo, dx_hi = report[x]\n(-1.0, Inf)\n\njulia> println(\n           \"The objective coefficient of `x` can decrease by $dx_lo or \" *\n           \"increase by $dx_hi.\"\n       )\nThe objective coefficient of `x` can decrease by -1.0 or increase by Inf.\n\njulia> dRHS_lo, dRHS_hi = report[LowerBoundRef(x)]\n(-Inf, 3.0)\n\njulia> println(\n           \"The lower bound of `x` can decrease by $dRHS_lo or increase \" *\n           \"by $dRHS_hi.\"\n       )\nThe lower bound of `x` can decrease by -Inf or increase by 3.0.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#map_coefficients","page":"JuMP","title":"map_coefficients","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.map_coefficients","category":"page"},{"location":"api/JuMP/#JuMP.map_coefficients","page":"JuMP","title":"JuMP.map_coefficients","text":"map_coefficients(f::Function, a::GenericAffExpr)\n\nApply f to the coefficients and constant term of an GenericAffExpr a and return a new expression.\n\nSee also: map_coefficients_inplace!\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> a = GenericAffExpr(1.0, x => 1.0)\nx + 1\n\njulia> map_coefficients(c -> 2 * c, a)\n2 x + 2\n\njulia> a\nx + 1\n\n\n\n\n\nmap_coefficients(f::Function, a::GenericQuadExpr)\n\nApply f to the coefficients and constant term of an GenericQuadExpr a and return a new expression.\n\nSee also: map_coefficients_inplace!\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> a = @expression(model, x^2 + x + 1)\nx² + x + 1\n\njulia> map_coefficients(c -> 2 * c, a)\n2 x² + 2 x + 2\n\njulia> a\nx² + x + 1\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#map_coefficients_inplace!","page":"JuMP","title":"map_coefficients_inplace!","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.map_coefficients_inplace!","category":"page"},{"location":"api/JuMP/#JuMP.map_coefficients_inplace!","page":"JuMP","title":"JuMP.map_coefficients_inplace!","text":"map_coefficients_inplace!(f::Function, a::GenericAffExpr)\n\nApply f to the coefficients and constant term of an GenericAffExpr a and update them in-place.\n\nSee also: map_coefficients\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> a = GenericAffExpr(1.0, x => 1.0)\nx + 1\n\njulia> map_coefficients_inplace!(c -> 2 * c, a)\n2 x + 2\n\njulia> a\n2 x + 2\n\n\n\n\n\nmap_coefficients_inplace!(f::Function, a::GenericQuadExpr)\n\nApply f to the coefficients and constant term of an GenericQuadExpr a and update them in-place.\n\nSee also: map_coefficients\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> a = @expression(model, x^2 + x + 1)\nx² + x + 1\n\njulia> map_coefficients_inplace!(c -> 2 * c, a)\n2 x² + 2 x + 2\n\njulia> a\n2 x² + 2 x + 2\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#mode","page":"JuMP","title":"mode","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.mode","category":"page"},{"location":"api/JuMP/#JuMP.mode","page":"JuMP","title":"JuMP.mode","text":"mode(model::GenericModel)\n\nReturn the ModelMode (DIRECT, AUTOMATIC, or MANUAL) of model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#model_convert","page":"JuMP","title":"model_convert","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.model_convert","category":"page"},{"location":"api/JuMP/#JuMP.model_convert","page":"JuMP","title":"JuMP.model_convert","text":"model_convert(\n    model::AbstractModel,\n    rhs::Union{\n        AbstractConstraint,\n        Number,\n        AbstractJuMPScalar,\n        MOI.AbstractSet,\n    },\n)\n\nConvert the coefficients and constants of functions and sets in the rhs to the coefficient type value_type(typeof(model)).\n\nPurpose\n\nCreating and adding a constraint is a two-step process. The first step calls build_constraint, and the result of that is passed to add_constraint.\n\nHowever, because build_constraint does not take the model as an argument, the coefficients and constants of the function or set might be different than value_type(typeof(model)).\n\nTherefore, the result of build_constraint is converted in a call to model_convert before the result is passed to add_constraint.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#model_string","page":"JuMP","title":"model_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.model_string","category":"page"},{"location":"api/JuMP/#JuMP.model_string","page":"JuMP","title":"JuMP.model_string","text":"model_string(mode::MIME, model::AbstractModel)\n\nReturn a String representation of model given the mode.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#moi_function","page":"JuMP","title":"moi_function","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.moi_function","category":"page"},{"location":"api/JuMP/#JuMP.moi_function","page":"JuMP","title":"JuMP.moi_function","text":"moi_function(x)\n\nGiven a JuMP object x, return the MathOptInterface equivalent.\n\nSee also: jump_function.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#moi_function_type","page":"JuMP","title":"moi_function_type","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.moi_function_type","category":"page"},{"location":"api/JuMP/#JuMP.moi_function_type","page":"JuMP","title":"JuMP.moi_function_type","text":"moi_function_type(::Type{T}) where {T}\n\nGiven a JuMP object type T, return the MathOptInterface equivalent.\n\nSee also: jump_function_type.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#moi_set","page":"JuMP","title":"moi_set","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.moi_set","category":"page"},{"location":"api/JuMP/#JuMP.moi_set","page":"JuMP","title":"JuMP.moi_set","text":"moi_set(constraint::AbstractConstraint)\n\nReturn the set of the constraint constraint in the function-in-set form as a MathOptInterface.AbstractSet.\n\nmoi_set(s::AbstractVectorSet, dim::Int)\n\nReturns the MOI set of dimension dim corresponding to the JuMP set s.\n\nmoi_set(s::AbstractScalarSet)\n\nReturns the MOI set corresponding to the JuMP set s.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#name","page":"JuMP","title":"name","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.name","category":"page"},{"location":"api/JuMP/#JuMP.name","page":"JuMP","title":"JuMP.name","text":"name(con_ref::ConstraintRef)\n\nGet a constraint's name attribute.\n\n\n\n\n\nname(v::GenericVariableRef)::String\n\nGet a variable's name attribute.\n\n\n\n\n\nname(model::AbstractModel)\n\nReturn the MOI.Name attribute of model's backend, or a default if empty.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#node_count","page":"JuMP","title":"node_count","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.node_count","category":"page"},{"location":"api/JuMP/#JuMP.node_count","page":"JuMP","title":"JuMP.node_count","text":"node_count(model::GenericModel)\n\nGets the total number of branch-and-bound nodes explored during the most recent optimization in a Mixed Integer Program.\n\nSolvers must implement MOI.NodeCount() to use this function.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#nonlinear_constraint_string","page":"JuMP","title":"nonlinear_constraint_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.nonlinear_constraint_string","category":"page"},{"location":"api/JuMP/#JuMP.nonlinear_constraint_string","page":"JuMP","title":"JuMP.nonlinear_constraint_string","text":"nonlinear_constraint_string(\n    model::GenericModel,\n    mode::MIME,\n    c::_NonlinearConstraint,\n)\n\nReturn a string representation of the nonlinear constraint c belonging to model, given the mode.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#nonlinear_dual_start_value","page":"JuMP","title":"nonlinear_dual_start_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.nonlinear_dual_start_value","category":"page"},{"location":"api/JuMP/#JuMP.nonlinear_dual_start_value","page":"JuMP","title":"JuMP.nonlinear_dual_start_value","text":"nonlinear_dual_start_value(model::Model)\n\nReturn the current value of the MOI attribute MOI.NLPBlockDualStart.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#nonlinear_expr_string","page":"JuMP","title":"nonlinear_expr_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.nonlinear_expr_string","category":"page"},{"location":"api/JuMP/#JuMP.nonlinear_expr_string","page":"JuMP","title":"JuMP.nonlinear_expr_string","text":"nonlinear_expr_string(\n    model::GenericModel,\n    mode::MIME,\n    c::MOI.Nonlinear.Expression,\n)\n\nReturn a string representation of the nonlinear expression c belonging to model, given the mode.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#nonlinear_model","page":"JuMP","title":"nonlinear_model","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.nonlinear_model","category":"page"},{"location":"api/JuMP/#JuMP.nonlinear_model","page":"JuMP","title":"JuMP.nonlinear_model","text":"nonlinear_model(\n    model::GenericModel;\n    force::Bool = false,\n)::Union{MOI.Nonlinear.Model,Nothing}\n\nIf model has nonlinear components, return a MOI.Nonlinear.Model, otherwise return nothing.\n\nIf force, always return a MOI.Nonlinear.Model, and if one does not exist for the model, create an empty one.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#normalized_coefficient","page":"JuMP","title":"normalized_coefficient","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.normalized_coefficient","category":"page"},{"location":"api/JuMP/#JuMP.normalized_coefficient","page":"JuMP","title":"JuMP.normalized_coefficient","text":"normalized_coefficient(con_ref::ConstraintRef, variable::GenericVariableRef)\n\nReturn the coefficient associated with variable in constraint after JuMP has normalized the constraint into its standard form. See also set_normalized_coefficient.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#normalized_rhs","page":"JuMP","title":"normalized_rhs","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.normalized_rhs","category":"page"},{"location":"api/JuMP/#JuMP.normalized_rhs","page":"JuMP","title":"JuMP.normalized_rhs","text":"normalized_rhs(con_ref::ConstraintRef)\n\nReturn the right-hand side term of con_ref after JuMP has converted the constraint into its normalized form. See also set_normalized_rhs.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#num_constraints","page":"JuMP","title":"num_constraints","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.num_constraints","category":"page"},{"location":"api/JuMP/#JuMP.num_constraints","page":"JuMP","title":"JuMP.num_constraints","text":"num_constraints(model::GenericModel, function_type, set_type)::Int64\n\nReturn the number of constraints currently in the model where the function has type function_type and the set has type set_type.\n\nSee also list_of_constraint_types and all_constraints.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x >= 0, Bin);\n\njulia> @variable(model, y);\n\njulia> @constraint(model, y in MOI.GreaterThan(1.0));\n\njulia> @constraint(model, y <= 1.0);\n\njulia> @constraint(model, 2x <= 1);\n\njulia> num_constraints(model, VariableRef, MOI.GreaterThan{Float64})\n2\n\njulia> num_constraints(model, VariableRef, MOI.ZeroOne)\n1\n\njulia> num_constraints(model, AffExpr, MOI.LessThan{Float64})\n2\n\n\n\n\n\nnum_constraints(model::GenericModel; count_variable_in_set_constraints::Bool)\n\nReturn the number of constraints in model.\n\nIf count_variable_in_set_constraints == true, then VariableRef constraints such as VariableRef-in-Integer are included. To count only the number of structural constraints (e.g., the rows in the constraint matrix of a linear program), pass count_variable_in_set_constraints = false.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x >= 0, Int);\n\njulia> @constraint(model, 2x <= 1);\n\njulia> num_constraints(model; count_variable_in_set_constraints = true)\n3\n\njulia> num_constraints(model; count_variable_in_set_constraints = false)\n1\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#num_nonlinear_constraints","page":"JuMP","title":"num_nonlinear_constraints","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.num_nonlinear_constraints","category":"page"},{"location":"api/JuMP/#JuMP.num_nonlinear_constraints","page":"JuMP","title":"JuMP.num_nonlinear_constraints","text":"num_nonlinear_constraints(model::GenericModel)\n\nReturns the number of nonlinear constraints associated with the model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#num_variables","page":"JuMP","title":"num_variables","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.num_variables","category":"page"},{"location":"api/JuMP/#JuMP.num_variables","page":"JuMP","title":"JuMP.num_variables","text":"num_variables(model::GenericModel)::Int64\n\nReturns number of variables in model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#object_dictionary","page":"JuMP","title":"object_dictionary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.object_dictionary","category":"page"},{"location":"api/JuMP/#JuMP.object_dictionary","page":"JuMP","title":"JuMP.object_dictionary","text":"object_dictionary(model::GenericModel)\n\nReturn the dictionary that maps the symbol name of a variable, constraint, or expression to the corresponding object.\n\nObjects are registered to a specific symbol in the macros. For example, @variable(model, x[1:2, 1:2]) registers the array of variables x to the symbol :x.\n\nThis method should be defined for any subtype of AbstractModel.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#objective_bound","page":"JuMP","title":"objective_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.objective_bound","category":"page"},{"location":"api/JuMP/#JuMP.objective_bound","page":"JuMP","title":"JuMP.objective_bound","text":"objective_bound(model::GenericModel)\n\nReturn the best known bound on the optimal objective value after a call to optimize!(model).\n\nFor scalar-valued objectives, this function returns a Float64. For vector-valued objectives, it returns a Vector{Float64}.\n\nIn the case of a vector-valued objective, this returns the ideal point, that is, the point obtained if each objective was optimized independently.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#objective_function","page":"JuMP","title":"objective_function","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.objective_function","category":"page"},{"location":"api/JuMP/#JuMP.objective_function","page":"JuMP","title":"JuMP.objective_function","text":"objective_function(\n    model::GenericModel,\n    T::Type = objective_function_type(model),\n)\n\nReturn an object of type T representing the objective function.\n\nError if the objective is not convertible to type T.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @objective(model, Min, 2x + 1)\n2 x + 1\n\njulia> objective_function(model, AffExpr)\n2 x + 1\n\njulia> objective_function(model, QuadExpr)\n2 x + 1\n\njulia> typeof(objective_function(model, QuadExpr))\nQuadExpr (alias for GenericQuadExpr{Float64, GenericVariableRef{Float64}})\n\nWe see with the last two commands that even if the objective function is affine, as it is convertible to a quadratic function, it can be queried as a quadratic function and the result is quadratic.\n\nHowever, it is not convertible to a variable.\n\njulia> objective_function(model, VariableRef)\nERROR: InexactError: convert(MathOptInterface.VariableIndex, 1.0 + 2.0 MOI.VariableIndex(1))\n[...]\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#objective_function_string","page":"JuMP","title":"objective_function_string","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.objective_function_string","category":"page"},{"location":"api/JuMP/#JuMP.objective_function_string","page":"JuMP","title":"JuMP.objective_function_string","text":"objective_function_string(mode, model::AbstractModel)::String\n\nReturn a String describing the objective function of the model.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#objective_function_type","page":"JuMP","title":"objective_function_type","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.objective_function_type","category":"page"},{"location":"api/JuMP/#JuMP.objective_function_type","page":"JuMP","title":"JuMP.objective_function_type","text":"objective_function_type(model::GenericModel)::AbstractJuMPScalar\n\nReturn the type of the objective function.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#objective_sense","page":"JuMP","title":"objective_sense","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.objective_sense","category":"page"},{"location":"api/JuMP/#JuMP.objective_sense","page":"JuMP","title":"JuMP.objective_sense","text":"objective_sense(model::GenericModel)::MOI.OptimizationSense\n\nReturn the objective sense.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#objective_value","page":"JuMP","title":"objective_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.objective_value","category":"page"},{"location":"api/JuMP/#JuMP.objective_value","page":"JuMP","title":"JuMP.objective_value","text":"objective_value(model::GenericModel; result::Int = 1)\n\nReturn the objective value associated with result index result of the most-recent solution returned by the solver.\n\nFor scalar-valued objectives, this function returns a Float64. For vector-valued objectives, it returns a Vector{Float64}.\n\nSee also: result_count.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#op_ifelse","page":"JuMP","title":"op_ifelse","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_ifelse","category":"page"},{"location":"api/JuMP/#JuMP.op_ifelse","page":"JuMP","title":"JuMP.op_ifelse","text":"op_ifelse(a, x, y)\n\nA function that falls back to ifelse(a, x, y), but when called with a JuMP variables or expression in the first argument, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_ifelse(true, 1.0, 2.0)\n1.0\n\njulia> op_ifelse(x, 1.0, 2.0)\nifelse(x, 1.0, 2.0)\n\njulia> op_ifelse(true, x, 2.0)\nx\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#operator_to_set","page":"JuMP","title":"operator_to_set","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.operator_to_set","category":"page"},{"location":"api/JuMP/#JuMP.operator_to_set","page":"JuMP","title":"JuMP.operator_to_set","text":"operator_to_set(_error::Function, ::Val{sense_symbol})\n\nConverts a sense symbol to a set set such that @constraint(model, func sense_symbol 0) is equivalent to @constraint(model, func in set) for any func::AbstractJuMPScalar.\n\nExample\n\nOnce a custom set is defined you can directly create a JuMP constraint with it:\n\njulia> struct CustomSet{T} <: MOI.AbstractScalarSet\n           value::T\n       end\n\njulia> Base.copy(x::CustomSet) = CustomSet(x.value)\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> cref = @constraint(model, x in CustomSet(1.0))\nx ∈ CustomSet{Float64}(1.0)\n\nHowever, there might be an appropriate sign that could be used in order to provide a more convenient syntax:\n\njulia> JuMP.operator_to_set(::Function, ::Val{:⊰}) = CustomSet(0.0)\n\njulia> MOIU.supports_shift_constant(::Type{<:CustomSet}) = true\n\njulia> MOIU.shift_constant(set::CustomSet, value) = CustomSet(set.value + value)\n\njulia> cref = @constraint(model, x ⊰ 1)\nx ∈ CustomSet{Float64}(1.0)\n\nNote that the whole function is first moved to the right-hand side, then the sign is transformed into a set with zero constant and finally the constant is moved to the set with MOIU.shift_constant.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#operator_warn","page":"JuMP","title":"operator_warn","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.operator_warn","category":"page"},{"location":"api/JuMP/#JuMP.operator_warn","page":"JuMP","title":"JuMP.operator_warn","text":"operator_warn(model::AbstractModel)\noperator_warn(model::GenericModel)\n\nThis function is called on the model whenever two affine expressions are added together without using destructive_add!, and at least one of the two expressions has more than 50 terms.\n\nFor the case of Model, if this function is called more than 20,000 times then a warning is generated once.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#optimize!","page":"JuMP","title":"optimize!","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.optimize!","category":"page"},{"location":"api/JuMP/#JuMP.optimize!","page":"JuMP","title":"JuMP.optimize!","text":"optimize!(\n    model::GenericModel;\n    ignore_optimize_hook = (model.optimize_hook === nothing),\n    _differentiation_backend::MOI.Nonlinear.AbstractAutomaticDifferentiation =\n        MOI.Nonlinear.SparseReverseMode(),\n    kwargs...,\n)\n\nOptimize the model.\n\nIf an optimizer has not been set yet (see set_optimizer), a NoOptimizer error is thrown.\n\nIf ignore_optimize_hook == true, the optimize hook is ignored and the model is solved as if the hook was not set. Keyword arguments kwargs are passed to the optimize_hook. An error is thrown if optimize_hook is nothing and keyword arguments are provided.\n\nExperimental features\n\nThese features may change or be removed in any future version of JuMP.\n\nPass _differentiation_backend to set the MOI.Nonlinear.AbstractAutomaticDifferentiation backend used to compute derivatives of nonlinear programs.\n\nIf you require only :ExprGraph, it is more efficient to pass _differentiation_backend = MOI.Nonlinear.ExprGraphOnly().\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#optimizer_index","page":"JuMP","title":"optimizer_index","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.optimizer_index","category":"page"},{"location":"api/JuMP/#JuMP.optimizer_index","page":"JuMP","title":"JuMP.optimizer_index","text":"optimizer_index(x::GenericVariableRef)::MOI.VariableIndex\noptimizer_index(x::ConstraintRef{<:GenericModel})::MOI.ConstraintIndex\n\nReturn the index that corresponds to x in the optimizer model.\n\nThrows NoOptimizer if no optimizer is set, and throws an ErrorException if the optimizer is set but is not attached.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#optimizer_with_attributes","page":"JuMP","title":"optimizer_with_attributes","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.optimizer_with_attributes","category":"page"},{"location":"api/JuMP/#JuMP.optimizer_with_attributes","page":"JuMP","title":"JuMP.optimizer_with_attributes","text":"optimizer_with_attributes(optimizer_constructor, attrs::Pair...)\n\nGroups an optimizer constructor with the list of attributes attrs. Note that it is equivalent to MOI.OptimizerWithAttributes.\n\nWhen provided to the Model constructor or to set_optimizer, it creates an optimizer by calling optimizer_constructor(), and then sets the attributes using set_attribute.\n\nSee also: set_attribute, get_attribute.\n\nNote\n\nThe string names of the attributes are specific to each solver. One should consult the solver's documentation to find the attributes of interest.\n\nExample\n\njulia> import HiGHS\n\njulia> optimizer = optimizer_with_attributes(\n           HiGHS.Optimizer, \"presolve\" => \"off\", MOI.Silent() => true,\n       );\n\njulia> model = Model(optimizer);\n\nis equivalent to:\n\njulia> import HiGHS\n\njulia> model = Model(HiGHS.Optimizer);\n\njulia> set_attribute(model, \"presolve\", \"off\")\n\njulia> set_attribute(model, MOI.Silent(), true)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#owner_model","page":"JuMP","title":"owner_model","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.owner_model","category":"page"},{"location":"api/JuMP/#JuMP.owner_model","page":"JuMP","title":"JuMP.owner_model","text":"owner_model(s::AbstractJuMPScalar)\n\nReturn the model owning the scalar s.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#parameter_value","page":"JuMP","title":"parameter_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.parameter_value","category":"page"},{"location":"api/JuMP/#JuMP.parameter_value","page":"JuMP","title":"JuMP.parameter_value","text":"parameter_value(x::GenericVariableRef)\n\nReturn the value of the parameter x.\n\nErrors if x is not a parameter.\n\nSee also ParameterRef, is_parameter, set_parameter_value.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, p in Parameter(2))\np\n\njulia> parameter_value(p)\n2.0\n\njulia> set_parameter_value(p, 2.5)\n\njulia> parameter_value(p)\n2.5\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#parse_constraint","page":"JuMP","title":"parse_constraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.parse_constraint","category":"page"},{"location":"api/JuMP/#JuMP.parse_constraint","page":"JuMP","title":"JuMP.parse_constraint","text":"parse_constraint(_error::Function, expr::Expr)\n\nThe entry-point for all constraint-related parsing.\n\nArguments\n\nThe _error function is passed everywhere to provide better error messages\nexpr comes from the @constraint macro. There are two possibilities:\n@constraint(model, expr)\n@constraint(model, name[args], expr)\nIn both cases, expr is the main component of the constraint.\n\nSupported syntax\n\nJuMP currently supports the following expr objects:\n\nlhs <= rhs\nlhs == rhs\nlhs >= rhs\nl <= body <= u\nu >= body >= l\nlhs ⟂ rhs\nlhs in rhs\nlhs ∈ rhs\nz => {constraint}\n!z => {constraint}\n\nas well as all broadcasted variants.\n\nExtensions\n\nThe infrastructure behind parse_constraint is extendable. See parse_constraint_head and parse_constraint_call for details.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#parse_constraint_call","page":"JuMP","title":"parse_constraint_call","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.parse_constraint_call","category":"page"},{"location":"api/JuMP/#JuMP.parse_constraint_call","page":"JuMP","title":"JuMP.parse_constraint_call","text":"parse_constraint_call(\n    _error::Function,\n    is_vectorized::Bool,\n    ::Val{op},\n    args...,\n)\n\nImplement this method to intercept the parsing of a :call expression with operator op.\n\nwarning: Warning\nExtending the constraint macro at parse time is an advanced operation and has the potential to interfere with existing JuMP syntax. Please discuss with the developer chatroom before publishing any code that implements these methods.\n\nArguments\n\n_error: a function that accepts a String and throws the string as an error, along with some descriptive information of the macro from which it was thrown.\nis_vectorized: a boolean to indicate if op should be broadcast or not\nop: the first element of the .args field of the Expr to intercept\nargs...: the .args field of the Expr.\n\nReturns\n\nThis function must return:\n\nparse_code::Expr: an expression containing any setup or rewriting code that needs to be called before build_constraint\nbuild_code::Expr: an expression that calls build_constraint( or build_constraint.( depending on is_vectorized.\n\nSee also: parse_constraint_head, build_constraint\n\n\n\n\n\nparse_constraint_call(\n    _error::Function,\n    vectorized::Bool,\n    ::Val{op},\n    lhs,\n    rhs,\n) where {op}\n\nFallback handler for binary operators. These might be infix operators like @constraint(model, lhs op rhs), or normal operators like @constraint(model, op(lhs, rhs)).\n\nIn both cases, we rewrite as lhs - rhs in operator_to_set(_error, op).\n\nSee operator_to_set for details.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#parse_constraint_head","page":"JuMP","title":"parse_constraint_head","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.parse_constraint_head","category":"page"},{"location":"api/JuMP/#JuMP.parse_constraint_head","page":"JuMP","title":"JuMP.parse_constraint_head","text":"parse_constraint_head(_error::Function, ::Val{head}, args...)\n\nImplement this method to intercept the parsing of an expression with head head.\n\nwarning: Warning\nExtending the constraint macro at parse time is an advanced operation and has the potential to interfere with existing JuMP syntax. Please discuss with the developer chatroom before publishing any code that implements these methods.\n\nArguments\n\n_error: a function that accepts a String and throws the string as an error, along with some descriptive information of the macro from which it was thrown.\nhead: the .head field of the Expr to intercept\nargs...: the .args field of the Expr.\n\nReturns\n\nThis function must return:\n\nis_vectorized::Bool: whether the expression represents a broadcasted expression like x .<= 1\nparse_code::Expr: an expression containing any setup or rewriting code that needs to be called before build_constraint\nbuild_code::Expr: an expression that calls build_constraint( or build_constraint.( depending on is_vectorized.\n\nExisting implementations\n\nJuMP currently implements:\n\n::Val{:call}, which forwards calls to parse_constraint_call\n::Val{:comparison}, which handles the special case of l <= body <= u.\n\nSee also: parse_constraint_call, build_constraint\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#parse_one_operator_variable","page":"JuMP","title":"parse_one_operator_variable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.parse_one_operator_variable","category":"page"},{"location":"api/JuMP/#JuMP.parse_one_operator_variable","page":"JuMP","title":"JuMP.parse_one_operator_variable","text":"parse_one_operator_variable(_error::Function, infoexpr::_VariableInfoExpr, sense::Val{S}, value) where S\n\nUpdate infoexr for a variable expression in the @variable macro of the form variable name S value.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#parse_ternary_variable","page":"JuMP","title":"parse_ternary_variable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.parse_ternary_variable","category":"page"},{"location":"api/JuMP/#JuMP.parse_ternary_variable","page":"JuMP","title":"JuMP.parse_ternary_variable","text":"parse_ternary_variable(_error, variable_info, lhs_sense, lhs, rhs_sense, rhs)\n\nA hook for JuMP extensiosn to intercept the parsing of a :comparison expression, which has the form lhs lhs_sense variable rhs_sense rhs.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#parse_variable","page":"JuMP","title":"parse_variable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.parse_variable","category":"page"},{"location":"api/JuMP/#JuMP.parse_variable","page":"JuMP","title":"JuMP.parse_variable","text":"parse_variable(_error::Function, ::_VariableInfoExpr, args...)\n\nA hook for extensions to intercept the parsing of inequality constraints in the @variable macro.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#primal_feasibility_report","page":"JuMP","title":"primal_feasibility_report","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.primal_feasibility_report","category":"page"},{"location":"api/JuMP/#JuMP.primal_feasibility_report","page":"JuMP","title":"JuMP.primal_feasibility_report","text":"primal_feasibility_report(\n    model::GenericModel{T},\n    point::AbstractDict{GenericVariableRef{T},T} = _last_primal_solution(model),\n    atol::T = zero(T),\n    skip_missing::Bool = false,\n)::Dict{Any,T}\n\nGiven a dictionary point, which maps variables to primal values, return a dictionary whose keys are the constraints with an infeasibility greater than the supplied tolerance atol. The value corresponding to each key is the respective infeasibility. Infeasibility is defined as the distance between the primal value of the constraint (see MOI.ConstraintPrimal) and the nearest point by Euclidean distance in the corresponding set.\n\nNotes\n\nIf skip_missing = true, constraints containing variables that are not in point will be ignored.\nIf skip_missing = false and a partial primal solution is provided, an error will be thrown.\nIf no point is provided, the primal solution from the last time the model was solved is used.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, 0.5 <= x <= 1);\n\njulia> primal_feasibility_report(model, Dict(x => 0.2))\nDict{Any, Float64} with 1 entry:\n  x ≥ 0.5 => 0.3\n\n\n\n\n\nprimal_feasibility_report(\n    point::Function,\n    model::GenericModel{T};\n    atol::T = zero(T),\n    skip_missing::Bool = false,\n) where {T}\n\nA form of primal_feasibility_report where a function is passed as the first argument instead of a dictionary as the second argument.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, 0.5 <= x <= 1, start = 1.3);\n\njulia> primal_feasibility_report(model) do v\n           return start_value(v)\n       end\nDict{Any, Float64} with 1 entry:\n  x ≤ 1 => 0.3\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#primal_status","page":"JuMP","title":"primal_status","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.primal_status","category":"page"},{"location":"api/JuMP/#JuMP.primal_status","page":"JuMP","title":"JuMP.primal_status","text":"primal_status(model::GenericModel; result::Int = 1)\n\nReturn a MOI.ResultStatusCode describing the status of the most recent primal solution of the solver (i.e., the MOI.PrimalStatus attribute) associated with the result index result.\n\nSee also: result_count.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#print_active_bridges","page":"JuMP","title":"print_active_bridges","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.print_active_bridges","category":"page"},{"location":"api/JuMP/#JuMP.print_active_bridges","page":"JuMP","title":"JuMP.print_active_bridges","text":"print_active_bridges([io::IO = stdout,] model::GenericModel)\n\nPrint a list of the variable, constraint, and objective bridges that are currently used in the model.\n\n\n\n\n\nprint_active_bridges([io::IO = stdout,] model::GenericModel, ::Type{F}) where {F}\n\nPrint a list of bridges required for an objective function of type F.\n\n\n\n\n\nprint_active_bridges(\n    [io::IO = stdout,]\n    model::GenericModel,\n    F::Type,\n    S::Type{<:MOI.AbstractSet},\n)\n\nPrint a list of bridges required for a constraint of type F-in-S.\n\n\n\n\n\nprint_active_bridges(\n    [io::IO = stdout,]\n    model::GenericModel,\n    S::Type{<:MOI.AbstractSet},\n)\n\nPrint a list of bridges required to add a variable constrained to the set S.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#print_bridge_graph","page":"JuMP","title":"print_bridge_graph","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.print_bridge_graph","category":"page"},{"location":"api/JuMP/#JuMP.print_bridge_graph","page":"JuMP","title":"JuMP.print_bridge_graph","text":" print_bridge_graph([io::IO,] model::GenericModel)\n\nPrint the hyper-graph containing all variable, constraint, and objective types that could be obtained by bridging the variables, constraints, and objectives that are present in the model.\n\nwarning: Warning\nThis function is intended for advanced users. If you want to see only the bridges that are currently used, use print_active_bridges instead.\n\nExplanation of output\n\nEach node in the hyper-graph corresponds to a variable, constraint, or objective type.\n\nVariable nodes are indicated by [ ]\nConstraint nodes are indicated by ( )\nObjective nodes are indicated by | |\n\nThe number inside each pair of brackets is an index of the node in the hyper-graph.\n\nNote that this hyper-graph is the full list of possible transformations. When the bridged model is created, we select the shortest hyper-path(s) from this graph, so many nodes may be un-used.\n\nFor more information, see Legat, B., Dowson, O., Garcia, J., and Lubin, M. (2020).  \"MathOptInterface: a data structure for mathematical optimization problems.\" URL: https://arxiv.org/abs/2002.03447\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#quad_terms","page":"JuMP","title":"quad_terms","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.quad_terms","category":"page"},{"location":"api/JuMP/#JuMP.quad_terms","page":"JuMP","title":"JuMP.quad_terms","text":"quad_terms(quad::GenericQuadExpr{C, V})\n\nProvides an iterator over tuples (coefficient::C, var_1::V, var_2::V) in the quadratic part of the quadratic expression.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#raw_status","page":"JuMP","title":"raw_status","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.raw_status","category":"page"},{"location":"api/JuMP/#JuMP.raw_status","page":"JuMP","title":"JuMP.raw_status","text":"raw_status(model::GenericModel)\n\nReturn the reason why the solver stopped in its own words (i.e., the MathOptInterface model attribute RawStatusString).\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#read_from_file","page":"JuMP","title":"read_from_file","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.read_from_file","category":"page"},{"location":"api/JuMP/#JuMP.read_from_file","page":"JuMP","title":"JuMP.read_from_file","text":"read_from_file(\n    filename::String;\n    format::MOI.FileFormats.FileFormat = MOI.FileFormats.FORMAT_AUTOMATIC,\n    kwargs...,\n)\n\nReturn a JuMP model read from filename in the format format.\n\nIf the filename ends in .gz, it will be uncompressed using Gzip. If the filename ends in .bz2, it will be uncompressed using BZip2.\n\nOther kwargs are passed to the Model constructor of the chosen format.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#reduced_cost","page":"JuMP","title":"reduced_cost","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.reduced_cost","category":"page"},{"location":"api/JuMP/#JuMP.reduced_cost","page":"JuMP","title":"JuMP.reduced_cost","text":"reduced_cost(x::GenericVariableRef{T})::T where {T}\n\nReturn the reduced cost associated with variable x.\n\nEquivalent to querying the shadow price of the active variable bound (if one exists and is active).\n\nSee also: shadow_price.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#register","page":"JuMP","title":"register","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.register","category":"page"},{"location":"api/JuMP/#JuMP.register","page":"JuMP","title":"JuMP.register","text":"register(\n    model::Model,\n    op::Symbol,\n    dimension::Integer,\n    f::Function;\n    autodiff:Bool = false,\n)\n\nRegister the user-defined function f that takes dimension arguments in model as the symbol op.\n\nThe function f must support all subtypes of Real as arguments. Do not assume that the inputs are Float64.\n\nNotes\n\nFor this method, you must explicitly set autodiff = true, because no user-provided gradient function ∇f is given.\nSecond-derivative information is only computed if dimension == 1.\nop does not have to be the same symbol as f, but it is generally more readable if it is.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> f(x::T) where {T<:Real} = x^2\nf (generic function with 1 method)\n\njulia> register(model, :foo, 1, f; autodiff = true)\n\njulia> @NLobjective(model, Min, foo(x))\n\njulia> model = Model();\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> g(x::T, y::T) where {T<:Real} = x * y\ng (generic function with 1 method)\n\njulia> register(model, :g, 2, g; autodiff = true)\n\njulia> @NLobjective(model, Min, g(x[1], x[2]))\n\n\n\n\n\nregister(\n    model::Model,\n    s::Symbol,\n    dimension::Integer,\n    f::Function,\n    ∇f::Function;\n    autodiff:Bool = false,\n)\n\nRegister the user-defined function f that takes dimension arguments in model as the symbol s. In addition, provide a gradient function ∇f.\n\nThe functions fand ∇f must support all subtypes of Real as arguments. Do not assume that the inputs are Float64.\n\nNotes\n\nIf the function f is univariate (i.e., dimension == 1), ∇f must return a number which represents the first-order derivative of the function f.\nIf the function f is multi-variate, ∇f must have a signature matching ∇f(g::AbstractVector{T}, args::T...) where {T<:Real}, where the first argument is a vector g that is modified in-place with the gradient.\nIf autodiff = true and dimension == 1, use automatic differentiation to compute the second-order derivative information. If autodiff = false, only first-order derivative information will be used.\ns does not have to be the same symbol as f, but it is generally more readable if it is.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> f(x::T) where {T<:Real} = x^2\nf (generic function with 1 method)\n\njulia> ∇f(x::T) where {T<:Real} = 2 * x\n∇f (generic function with 1 method)\n\njulia> register(model, :foo, 1, f, ∇f; autodiff = true)\n\njulia> @NLobjective(model, Min, foo(x))\n\njulia> model = Model();\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> g(x::T, y::T) where {T<:Real} = x * y\ng (generic function with 1 method)\n\njulia> function ∇g(g::AbstractVector{T}, x::T, y::T) where {T<:Real}\n           g[1] = y\n           g[2] = x\n           return\n       end\n∇g (generic function with 1 method)\n\njulia> register(model, :g, 2, g, ∇g)\n\njulia> @NLobjective(model, Min, g(x[1], x[2]))\n\n\n\n\n\nregister(\n    model::Model,\n    s::Symbol,\n    dimension::Integer,\n    f::Function,\n    ∇f::Function,\n    ∇²f::Function,\n)\n\nRegister the user-defined function f that takes dimension arguments in model as the symbol s. In addition, provide a gradient function ∇f and a hessian function ∇²f.\n\n∇f and ∇²f must return numbers corresponding to the first- and second-order derivatives of the function f respectively.\n\nNotes\n\nBecause automatic differentiation is not used, you can assume the inputs are all Float64.\nThis method will throw an error if dimension > 1.\ns does not have to be the same symbol as f, but it is generally more readable if it is.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> f(x::Float64) = x^2\nf (generic function with 1 method)\n\njulia> ∇f(x::Float64) = 2 * x\n∇f (generic function with 1 method)\n\njulia> ∇²f(x::Float64) = 2.0\n∇²f (generic function with 1 method)\n\njulia> register(model, :foo, 1, f, ∇f, ∇²f)\n\njulia> @NLobjective(model, Min, foo(x))\n\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#relative_gap","page":"JuMP","title":"relative_gap","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.relative_gap","category":"page"},{"location":"api/JuMP/#JuMP.relative_gap","page":"JuMP","title":"JuMP.relative_gap","text":"relative_gap(model::GenericModel)\n\nReturn the final relative optimality gap after a call to optimize!(model). Exact value depends upon implementation of MathOptInterface.RelativeGap() by the particular solver used for optimization.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#relax_integrality","page":"JuMP","title":"relax_integrality","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.relax_integrality","category":"page"},{"location":"api/JuMP/#JuMP.relax_integrality","page":"JuMP","title":"JuMP.relax_integrality","text":"relax_integrality(model::GenericModel)\n\nModifies model to \"relax\" all binary and integrality constraints on variables. Specifically,\n\nBinary constraints are deleted, and variable bounds are tightened if necessary to ensure the variable is constrained to the interval 0 1.\nIntegrality constraints are deleted without modifying variable bounds.\nAn error is thrown if semi-continuous or semi-integer constraints are present (support may be added for these in the future).\nAll other constraints are ignored (left in place). This includes discrete constraints like SOS and indicator constraints.\n\nReturns a function that can be called without any arguments to restore the original model. The behavior of this function is undefined if additional changes are made to the affected variables in the meantime.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x, Bin);\n\njulia> @variable(model, 1 <= y <= 10, Int);\n\njulia> @objective(model, Min, x + y);\n\njulia> undo_relax = relax_integrality(model);\n\njulia> print(model)\nMin x + y\nSubject to\n x ≥ 0\n y ≥ 1\n x ≤ 1\n y ≤ 10\n\njulia> undo_relax()\n\njulia> print(model)\nMin x + y\nSubject to\n y ≥ 1\n y ≤ 10\n y integer\n x binary\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#relax_with_penalty!","page":"JuMP","title":"relax_with_penalty!","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.relax_with_penalty!","category":"page"},{"location":"api/JuMP/#JuMP.relax_with_penalty!","page":"JuMP","title":"JuMP.relax_with_penalty!","text":"relax_with_penalty!(\n    model::GenericModel{T},\n    [penalties::Dict{ConstraintRef,T}];\n    [default::Union{Nothing,Real} = nothing,]\n) where {T}\n\nDestructively modify the model in-place to create a penalized relaxation of the constraints.\n\nwarning: Warning\nThis is a destructive routine that modifies the model in-place. If you don't want to modify the original model, use copy_model to create a copy before calling relax_with_penalty!.\n\nReformulation\n\nSee MOI.Utilities.ScalarPenaltyRelaxation for details of the reformulation.\n\nFor each constraint ci, the penalty passed to MOI.Utilities.ScalarPenaltyRelaxation is get(penalties, ci, default). If the value is nothing, because ci does not exist in penalties and default = nothing, then the constraint is skipped.\n\nReturn value\n\nThis function returns a Dict{ConstraintRef,AffExpr} that maps each constraint index to the corresponding y + z as an AffExpr. In an optimal solution, query the value of these functions to compute the violation of each constraint.\n\nRelax a subset of constraints\n\nTo relax a subset of constraints, pass a penalties dictionary and set default = nothing.\n\nExample\n\njulia> function new_model()\n           model = Model()\n           @variable(model, x)\n           @objective(model, Max, 2x + 1)\n           @constraint(model, c1, 2x - 1 <= -2)\n           @constraint(model, c2, 3x >= 0)\n           return model\n       end\nnew_model (generic function with 1 method)\n\njulia> model_1 = new_model();\n\njulia> penalty_map = relax_with_penalty!(model_1; default = 2.0);\n\njulia> penalty_map[model_1[:c1]]\n_[3]\n\njulia> penalty_map[model_1[:c2]]\n_[2]\n\njulia> print(model_1)\nMax 2 x - 2 _[2] - 2 _[3] + 1\nSubject to\n c2 : 3 x + _[2] ≥ 0\n c1 : 2 x - _[3] ≤ -1\n _[2] ≥ 0\n _[3] ≥ 0\n\njulia> model_2 = new_model();\n\njulia> relax_with_penalty!(model_2, Dict(model_2[:c2] => 3.0))\nDict{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.GreaterThan{Float64}}, ScalarShape}, AffExpr} with 1 entry:\n  c2 : 3 x + _[2] ≥ 0 => _[2]\n\njulia> print(model_2)\nMax 2 x - 3 _[2] + 1\nSubject to\n c2 : 3 x + _[2] ≥ 0\n c1 : 2 x ≤ -1\n _[2] ≥ 0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#remove_bridge","page":"JuMP","title":"remove_bridge","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.remove_bridge","category":"page"},{"location":"api/JuMP/#JuMP.remove_bridge","page":"JuMP","title":"JuMP.remove_bridge","text":"remove_bridge(\n    model::GenericModel{S},\n    BT::Type{<:MOI.Bridges.AbstractBridge};\n    coefficient_type::Type{T} = S,\n) where {S,T}\n\nRemove BT{T} from the list of bridges that can be used to transform unsupported constraints into an equivalent formulation using only constraints supported by the optimizer.\n\nSee also: add_bridge.\n\nExample\n\njulia> model = Model();\n\njulia> add_bridge(model, MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge)\n\njulia> remove_bridge(model, MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge)\n\njulia> add_bridge(\n           model,\n           MOI.Bridges.Constraint.NumberConversionBridge;\n           coefficient_type = Complex{Float64},\n       )\n\njulia> remove_bridge(\n           model,\n           MOI.Bridges.Constraint.NumberConversionBridge;\n           coefficient_type = Complex{Float64},\n       )\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#reshape_set","page":"JuMP","title":"reshape_set","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.reshape_set","category":"page"},{"location":"api/JuMP/#JuMP.reshape_set","page":"JuMP","title":"JuMP.reshape_set","text":"reshape_set(vectorized_set::MOI.AbstractSet, shape::AbstractShape)\n\nReturn a set in its original shape shape given its vectorized form vectorized_form.\n\nExample\n\nGiven a SymmetricMatrixShape of vectorized form [1, 2, 3] in MOI.PositiveSemidefinieConeTriangle(2), the following code returns the set of the original constraint Symmetric(Matrix[1 2; 2 3]) in PSDCone():\n\njulia> reshape_set(MOI.PositiveSemidefiniteConeTriangle(2), SymmetricMatrixShape(2))\nPSDCone()\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#reshape_vector","page":"JuMP","title":"reshape_vector","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.reshape_vector","category":"page"},{"location":"api/JuMP/#JuMP.reshape_vector","page":"JuMP","title":"JuMP.reshape_vector","text":"reshape_vector(vectorized_form::Vector, shape::AbstractShape)\n\nReturn an object in its original shape shape given its vectorized form vectorized_form.\n\nExample\n\nGiven a SymmetricMatrixShape of vectorized form [1, 2, 3], the following code returns the matrix Symmetric(Matrix[1 2; 2 3]):\n\njulia> reshape_vector([1, 2, 3], SymmetricMatrixShape(2))\n2×2 LinearAlgebra.Symmetric{Int64, Matrix{Int64}}:\n 1  2\n 2  3\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#result_count","page":"JuMP","title":"result_count","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.result_count","category":"page"},{"location":"api/JuMP/#JuMP.result_count","page":"JuMP","title":"JuMP.result_count","text":"result_count(model::GenericModel)\n\nReturn the number of results available to query after a call to optimize!.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#reverse_sense","page":"JuMP","title":"reverse_sense","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.reverse_sense","category":"page"},{"location":"api/JuMP/#JuMP.reverse_sense","page":"JuMP","title":"JuMP.reverse_sense","text":"reverse_sense(::Val{T}) where {T}\n\nGiven an (in)equality symbol T, return a new Val object with the opposite (in)equality symbol.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_attribute","page":"JuMP","title":"set_attribute","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_attribute","category":"page"},{"location":"api/JuMP/#JuMP.set_attribute","page":"JuMP","title":"JuMP.set_attribute","text":"set_attribute(model::GenericModel, attr::MOI.AbstractModelAttribute, value)\nset_attribute(x::GenericVariableRef, attr::MOI.AbstractVariableAttribute, value)\nset_attribute(cr::ConstraintRef, attr::MOI.AbstractConstraintAttribute, value)\n\nSet the value of a solver-specifc attribute attr to value.\n\nThis is equivalent to calling MOI.set with the associated MOI model and, for variables and constraints, with the associated MOI.VariableIndex or MOI.ConstraintIndex.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, c, 2 * x <= 1)\nc : 2 x ≤ 1\n\njulia> set_attribute(model, MOI.Name(), \"model_new\")\n\njulia> set_attribute(x, MOI.VariableName(), \"x_new\")\n\njulia> set_attribute(c, MOI.ConstraintName(), \"c_new\")\n\n\n\n\n\nset_attribute(\n    model::Union{GenericModel,MOI.OptimizerWithAttributes},\n    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},\n    value,\n)\n\nSet the value of a solver-specifc attribute attr to value.\n\nThis is equivalent to calling MOI.set with the associated MOI model.\n\nIf attr is an AbstractString, it is converted to MOI.RawOptimizerAttribute.\n\nExample\n\njulia> import HiGHS\n\njulia> opt = optimizer_with_attributes(HiGHS.Optimizer, \"output_flag\" => false);\n\njulia> model = Model(opt);\n\njulia> set_attribute(model, \"output_flag\", false)\n\njulia> set_attribute(model, MOI.RawOptimizerAttribute(\"output_flag\"), true)\n\njulia> set_attribute(opt, \"output_flag\", true)\n\njulia> set_attribute(opt, MOI.RawOptimizerAttribute(\"output_flag\"), false)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_attributes","page":"JuMP","title":"set_attributes","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_attributes","category":"page"},{"location":"api/JuMP/#JuMP.set_attributes","page":"JuMP","title":"JuMP.set_attributes","text":"set_attributes(\n    destination::Union{\n        GenericModel,\n        MOI.OptimizerWithAttributes,\n        GenericVariableRef,\n        ConstraintRef,\n    },\n    pairs::Pair...,\n)\n\nGiven a list of attribute => value pairs, calls set_attribute(destination, attribute, value) for each pair.\n\nSee also: set_attribute, get_attribute.\n\nExample\n\njulia> import Ipopt\n\njulia> model = Model(Ipopt.Optimizer);\n\njulia> set_attributes(model, \"tol\" => 1e-4, \"max_iter\" => 100)\n\nis equivalent to:\n\njulia> import Ipopt\n\njulia> model = Model(Ipopt.Optimizer);\n\njulia> set_attribute(model, \"tol\", 1e-4)\n\njulia> set_attribute(model, \"max_iter\", 100)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_binary","page":"JuMP","title":"set_binary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_binary","category":"page"},{"location":"api/JuMP/#JuMP.set_binary","page":"JuMP","title":"JuMP.set_binary","text":"set_binary(v::GenericVariableRef)\n\nAdd a constraint on the variable v that it must take values in the set 01.\n\nSee also BinaryRef, is_binary, unset_binary.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> is_binary(x)\nfalse\n\njulia> set_binary(x)\n\njulia> is_binary(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_dual_start_value","page":"JuMP","title":"set_dual_start_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_dual_start_value","category":"page"},{"location":"api/JuMP/#JuMP.set_dual_start_value","page":"JuMP","title":"JuMP.set_dual_start_value","text":"set_dual_start_value(con_ref::ConstraintRef, value)\n\nSet the dual start value (MOI attribute ConstraintDualStart) of the constraint con_ref to value. To remove a dual start value set it to nothing.\n\nSee also dual_start_value.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_integer","page":"JuMP","title":"set_integer","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_integer","category":"page"},{"location":"api/JuMP/#JuMP.set_integer","page":"JuMP","title":"JuMP.set_integer","text":"set_integer(variable_ref::GenericVariableRef)\n\nAdd an integrality constraint on the variable variable_ref.\n\nSee also IntegerRef, is_integer, unset_integer.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> is_integer(x)\nfalse\n\njulia> set_integer(x)\n\njulia> is_integer(x)\ntrue\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_lower_bound","page":"JuMP","title":"set_lower_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_lower_bound","category":"page"},{"location":"api/JuMP/#JuMP.set_lower_bound","page":"JuMP","title":"JuMP.set_lower_bound","text":"set_lower_bound(v::GenericVariableRef, lower::Number)\n\nSet the lower bound of a variable. If one does not exist, create a new lower bound constraint.\n\nSee also LowerBoundRef, has_lower_bound, lower_bound, delete_lower_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x >= 1.0);\n\njulia> lower_bound(x)\n1.0\n\njulia> set_lower_bound(x, 2.0)\n\njulia> lower_bound(x)\n2.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_name","page":"JuMP","title":"set_name","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_name","category":"page"},{"location":"api/JuMP/#JuMP.set_name","page":"JuMP","title":"JuMP.set_name","text":"set_name(con_ref::ConstraintRef, s::AbstractString)\n\nSet a constraint's name attribute.\n\n\n\n\n\nset_name(v::GenericVariableRef, s::AbstractString)\n\nSet a variable's name attribute.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_nonlinear_dual_start_value","page":"JuMP","title":"set_nonlinear_dual_start_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_nonlinear_dual_start_value","category":"page"},{"location":"api/JuMP/#JuMP.set_nonlinear_dual_start_value","page":"JuMP","title":"JuMP.set_nonlinear_dual_start_value","text":"set_nonlinear_dual_start_value(\n    model::Model,\n    start::Union{Nothing,Vector{Float64}},\n)\n\nSet the value of the MOI attribute MOI.NLPBlockDualStart.\n\nThe start vector corresponds to the Lagrangian duals of the nonlinear constraints, in the order given by all_nonlinear_constraints. That is, you must pass a single start vector corresponding to all of the nonlinear constraints in a single function call; you cannot set the dual start value of nonlinear constraints one-by-one. The example below demonstrates how to use all_nonlinear_constraints to create a mapping between the nonlinear constraint references and the start vector.\n\nPass nothing to unset a previous start.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> nl1 = @NLconstraint(model, x[1] <= sqrt(x[2]));\n\njulia> nl2 = @NLconstraint(model, x[1] >= exp(x[2]));\n\njulia> start = Dict(nl1 => -1.0, nl2 => 1.0);\n\njulia> start_vector = [start[con] for con in all_nonlinear_constraints(model)]\n2-element Vector{Float64}:\n -1.0\n  1.0\n\njulia> set_nonlinear_dual_start_value(model, start_vector)\n\njulia> nonlinear_dual_start_value(model)\n2-element Vector{Float64}:\n -1.0\n  1.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_nonlinear_objective","page":"JuMP","title":"set_nonlinear_objective","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_nonlinear_objective","category":"page"},{"location":"api/JuMP/#JuMP.set_nonlinear_objective","page":"JuMP","title":"JuMP.set_nonlinear_objective","text":"set_nonlinear_objective(\n    model::Model,\n    sense::MOI.OptimizationSense,\n    expr::Expr,\n)\n\nSet the nonlinear objective of model to the expression expr, with the optimization sense sense.\n\nThis function is most useful if the expression expr is generated programmatically, and you cannot use @NLobjective.\n\nNotes\n\nYou must interpolate the variables directly into the expression expr.\nYou must use MIN_SENSE or MAX_SENSE instead of Min and Max.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> set_nonlinear_objective(model, MIN_SENSE, :($(x) + $(x)^2))\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_normalized_coefficient","page":"JuMP","title":"set_normalized_coefficient","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_normalized_coefficient","category":"page"},{"location":"api/JuMP/#JuMP.set_normalized_coefficient","page":"JuMP","title":"JuMP.set_normalized_coefficient","text":"set_normalized_coefficient(con_ref::ConstraintRef, variable::GenericVariableRef, value)\n\nSet the coefficient of variable in the constraint constraint to value.\n\nNote that prior to this step, JuMP will aggregate multiple terms containing the same variable. For example, given a constraint 2x + 3x <= 2, set_normalized_coefficient(con, x, 4) will create the constraint 4x <= 2.\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, con, 2x + 3x <= 2)\ncon : 5 x ≤ 2\n\njulia> set_normalized_coefficient(con, x, 4)\n\njulia> con\ncon : 4 x ≤ 2\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_normalized_coefficients","page":"JuMP","title":"set_normalized_coefficients","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_normalized_coefficients","category":"page"},{"location":"api/JuMP/#JuMP.set_normalized_coefficients","page":"JuMP","title":"JuMP.set_normalized_coefficients","text":"set_normalized_coefficients(\n    con_ref::ConstraintRef,\n    variable,\n    new_coefficients::Vector{Tuple{Int64,T}},\n)\n\nSet the coefficients of variable in the constraint con_ref to new_coefficients, where each element in new_coefficients is a tuple which maps the row to a new coefficient.\n\nNote that prior to this step, during constraint creation, JuMP will aggregate multiple terms containing the same variable.\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, con, [2x + 3x, 4x] in MOI.Nonnegatives(2))\ncon : [5 x, 4 x] ∈ MathOptInterface.Nonnegatives(2)\n\njulia> set_normalized_coefficients(con, x, [(1, 2.0), (2, 5.0)])\n\njulia> con\ncon : [2 x, 5 x] ∈ MathOptInterface.Nonnegatives(2)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_normalized_rhs","page":"JuMP","title":"set_normalized_rhs","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_normalized_rhs","category":"page"},{"location":"api/JuMP/#JuMP.set_normalized_rhs","page":"JuMP","title":"JuMP.set_normalized_rhs","text":"set_normalized_rhs(con_ref::ConstraintRef, value)\n\nSet the right-hand side term of constraint to value.\n\nNote that prior to this step, JuMP will aggregate all constant terms onto the right-hand side of the constraint. For example, given a constraint 2x + 1 <= 2, set_normalized_rhs(con, 4) will create the constraint 2x <= 4, not 2x + 1 <= 4.\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @constraint(model, con, 2x + 1 <= 2)\ncon : 2 x ≤ 1\n\njulia> set_normalized_rhs(con, 4)\n\njulia> con\ncon : 2 x ≤ 4\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_objective","page":"JuMP","title":"set_objective","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_objective","category":"page"},{"location":"api/JuMP/#JuMP.set_objective","page":"JuMP","title":"JuMP.set_objective","text":"set_objective(model::AbstractModel, sense::MOI.OptimizationSense, func)\n\nThe functional equivalent of the @objective macro.\n\nSets the objective sense and objective function simultaneously, and is equivalent to calling set_objective_sense and set_objective_function separately.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> set_objective(model, MIN_SENSE, x)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_objective_coefficient","page":"JuMP","title":"set_objective_coefficient","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_objective_coefficient","category":"page"},{"location":"api/JuMP/#JuMP.set_objective_coefficient","page":"JuMP","title":"JuMP.set_objective_coefficient","text":"set_objective_coefficient(model::GenericModel, variable::GenericVariableRef, coefficient::Real)\n\nSet the linear objective coefficient associated with Variable to coefficient.\n\nNote: this function will throw an error if a nonlinear objective is set.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_objective_function","page":"JuMP","title":"set_objective_function","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_objective_function","category":"page"},{"location":"api/JuMP/#JuMP.set_objective_function","page":"JuMP","title":"JuMP.set_objective_function","text":"set_objective_function(model::GenericModel, func::MOI.AbstractFunction)\nset_objective_function(model::GenericModel, func::AbstractJuMPScalar)\nset_objective_function(model::GenericModel, func::Real)\nset_objective_function(model::GenericModel, func::Vector{<:AbstractJuMPScalar})\n\nSets the objective function of the model to the given function. See set_objective_sense to set the objective sense. These are low-level functions; the recommended way to set the objective is with the @objective macro.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_objective_sense","page":"JuMP","title":"set_objective_sense","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_objective_sense","category":"page"},{"location":"api/JuMP/#JuMP.set_objective_sense","page":"JuMP","title":"JuMP.set_objective_sense","text":"set_objective_sense(model::GenericModel, sense::MOI.OptimizationSense)\n\nSets the objective sense of the model to the given sense. See set_objective_function to set the objective function. These are low-level functions; the recommended way to set the objective is with the @objective macro.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_optimize_hook","page":"JuMP","title":"set_optimize_hook","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_optimize_hook","category":"page"},{"location":"api/JuMP/#JuMP.set_optimize_hook","page":"JuMP","title":"JuMP.set_optimize_hook","text":"set_optimize_hook(model::GenericModel, f::Union{Function,Nothing})\n\nSet the function f as the optimize hook for model.\n\nf should have a signature f(model::GenericModel; kwargs...), where the kwargs are those passed to optimize!.\n\nNotes\n\nThe optimize hook should generally modify the model, or some external state in some way, and then call optimize!(model; ignore_optimize_hook = true) to optimize the problem, bypassing the hook.\nUse set_optimize_hook(model, nothing) to unset an optimize hook.\n\nExample\n\njulia> model = Model();\n\njulia> function my_hook(model::Model; kwargs...)\n           println(kwargs)\n           println(\"Calling with `ignore_optimize_hook = true`\")\n           optimize!(model; ignore_optimize_hook = true)\n           return\n       end\nmy_hook (generic function with 1 method)\n\njulia> set_optimize_hook(model, my_hook)\nmy_hook (generic function with 1 method)\n\njulia> optimize!(model; test_arg = true)\nBase.Pairs{Symbol, Bool, Tuple{Symbol}, NamedTuple{(:test_arg,), Tuple{Bool}}}(:test_arg => 1)\nCalling with `ignore_optimize_hook = true`\nERROR: NoOptimizer()\n[...]\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_optimizer","page":"JuMP","title":"set_optimizer","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_optimizer","category":"page"},{"location":"api/JuMP/#JuMP.set_optimizer","page":"JuMP","title":"JuMP.set_optimizer","text":"set_optimizer(\n    model::GenericModel,\n    optimizer_factory;\n    add_bridges::Bool = true,\n)\n\nCreates an empty MathOptInterface.AbstractOptimizer instance by calling optimizer_factory() and sets it as the optimizer of model. Specifically, optimizer_factory must be callable with zero arguments and return an empty MathOptInterface.AbstractOptimizer.\n\nIf add_bridges is true, constraints and objectives that are not supported by the optimizer are automatically bridged to equivalent supported formulation. Passing add_bridges = false can improve performance if the solver natively supports all of the elements in model.\n\nSee set_attribute for setting solver-specific parameters of the optimizer.\n\nExample\n\njulia> import HiGHS\n\njulia> model = Model();\n\njulia> set_optimizer(model, () -> HiGHS.Optimizer())\n\njulia> set_optimizer(model, HiGHS.Optimizer; add_bridges = false)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_optimizer_attribute","page":"JuMP","title":"set_optimizer_attribute","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_optimizer_attribute","category":"page"},{"location":"api/JuMP/#JuMP.set_optimizer_attribute","page":"JuMP","title":"JuMP.set_optimizer_attribute","text":"set_optimizer_attribute(\n    model::Union{GenericModel,MOI.OptimizerWithAttributes},\n    attr::Union{AbstractString,MOI.AbstractOptimizerAttribute},\n    value,\n)\n\nSet the solver-specific attribute attr in model to value.\n\nIf attr is an AbstractString, this is equivalent to set_optimizer_attribute(model, MOI.RawOptimizerAttribute(name), value).\n\ncompat: Compat\nThis method will remain in all v1.X releases of JuMP, but it may be removed in a future v2.0 release. We recommend using set_attribute instead.\n\nSee also: set_optimizer_attributes, get_optimizer_attribute.\n\nExample\n\njulia> model = Model();\n\njulia> set_optimizer_attribute(model, MOI.Silent(), true)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_optimizer_attributes","page":"JuMP","title":"set_optimizer_attributes","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_optimizer_attributes","category":"page"},{"location":"api/JuMP/#JuMP.set_optimizer_attributes","page":"JuMP","title":"JuMP.set_optimizer_attributes","text":"set_optimizer_attributes(\n    model::Union{GenericModel,MOI.OptimizerWithAttributes},\n    pairs::Pair...,\n)\n\nGiven a list of attribute => value pairs, calls set_optimizer_attribute(model, attribute, value) for each pair.\n\ncompat: Compat\nThis method will remain in all v1.X releases of JuMP, but it may be removed in a future v2.0 release. We recommend using set_attributes instead.\n\nSee also: set_optimizer_attribute, get_optimizer_attribute.\n\nExample\n\njulia> import Ipopt\n\njulia> model = Model(Ipopt.Optimizer);\n\njulia> set_optimizer_attributes(model, \"tol\" => 1e-4, \"max_iter\" => 100)\n\nis equivalent to:\n\njulia> import Ipopt\n\njulia> model = Model(Ipopt.Optimizer);\n\njulia> set_optimizer_attribute(model, \"tol\", 1e-4)\n\njulia> set_optimizer_attribute(model, \"max_iter\", 100)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_parameter_value","page":"JuMP","title":"set_parameter_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_parameter_value","category":"page"},{"location":"api/JuMP/#JuMP.set_parameter_value","page":"JuMP","title":"JuMP.set_parameter_value","text":"set_parameter_value(x::GenericVariableRef, value)\n\nUpdate the parameter constraint on the variable x to value.\n\nErrors if x is not a parameter.\n\nSee also ParameterRef, is_parameter, parameter_value.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, p in Parameter(2))\np\n\njulia> parameter_value(p)\n2.0\n\njulia> set_parameter_value(p, 2.5)\n\njulia> parameter_value(p)\n2.5\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_silent","page":"JuMP","title":"set_silent","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_silent","category":"page"},{"location":"api/JuMP/#JuMP.set_silent","page":"JuMP","title":"JuMP.set_silent","text":"set_silent(model::GenericModel)\n\nTakes precedence over any other attribute controlling verbosity and requires the solver to produce no output.\n\nSee also: unset_silent.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_start_value","page":"JuMP","title":"set_start_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_start_value","category":"page"},{"location":"api/JuMP/#JuMP.set_start_value","page":"JuMP","title":"JuMP.set_start_value","text":"set_start_value(con_ref::ConstraintRef, value)\n\nSet the primal start value (MOI.ConstraintPrimalStart) of the constraint con_ref to value. To remove a primal start value set it to nothing.\n\nSee also start_value.\n\n\n\n\n\nset_start_value(variable::GenericVariableRef, value::Union{Real,Nothing})\n\nSet the start value (MOI attribute VariablePrimalStart) of the variable to value.\n\nPass nothing to unset the start value.\n\nNote: VariablePrimalStarts are sometimes called \"MIP-starts\" or \"warmstarts\".\n\nSee also start_value.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_start_values","page":"JuMP","title":"set_start_values","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_start_values","category":"page"},{"location":"api/JuMP/#JuMP.set_start_values","page":"JuMP","title":"JuMP.set_start_values","text":"set_start_values(\n    model::GenericModel;\n    variable_primal_start::Union{Nothing,Function} = value,\n    constraint_primal_start::Union{Nothing,Function} = value,\n    constraint_dual_start::Union{Nothing,Function} = dual,\n    nonlinear_dual_start::Union{Nothing,Function} = nonlinear_dual_start_value,\n)\n\nSet the primal and dual starting values in model using the functions provided.\n\nIf any keyword argument is nothing, the corresponding start value is skipped.\n\nIf the optimizer does not support setting the starting value, the value will be skipped.\n\nvariable_primal_start\n\nThis function controls the primal starting solution for the variables. It is equivalent to calling set_start_value for each variable, or setting the MOI.VariablePrimalStart attribute.\n\nIf it is a function, it must have the form variable_primal_start(x::VariableRef) that maps each variable x to the starting primal value.\n\nThe default is value.\n\nconstraint_primal_start\n\nThis function controls the primal starting solution for the constraints. It is equivalent to calling set_start_value for each constraint, or setting the MOI.ConstraintPrimalStart attribute.\n\nIf it is a function, it must have the form constraint_primal_start(ci::ConstraintRef) that maps each constraint ci to the starting primal value.\n\nThe default is value.\n\nconstraint_dual_start\n\nThis function controls the dual starting solution for the constraints. It is equivalent to calling set_dual_start_value for each constraint, or setting the MOI.ConstraintDualStart attribute.\n\nIf it is a function, it must have the form constraint_dual_start(ci::ConstraintRef) that maps each constraint ci to the starting dual value.\n\nThe default is dual.\n\nnonlinear_dual_start\n\nThis function controls the dual starting solution for the nonlinear constraints It is equivalent to calling set_nonlinear_dual_start_value.\n\nIf it is a function, it must have the form nonlinear_dual_start(model::GenericModel) that returns a vector corresponding to the dual start of the constraints.\n\nThe default is nonlinear_dual_start_value.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_string_names_on_creation","page":"JuMP","title":"set_string_names_on_creation","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_string_names_on_creation","category":"page"},{"location":"api/JuMP/#JuMP.set_string_names_on_creation","page":"JuMP","title":"JuMP.set_string_names_on_creation","text":"set_string_names_on_creation(model::GenericModel, value::Bool)\n\nSet the default argument of the set_string_name keyword in the @variable and @constraint macros to value. This is used to determine whether to assign String names to all variables and constraints in model.\n\nBy default, value is true. However, for larger models calling set_string_names_on_creation(model, false) can improve performance at the cost of reducing the readability of printing and solver log messages.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_time_limit_sec","page":"JuMP","title":"set_time_limit_sec","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_time_limit_sec","category":"page"},{"location":"api/JuMP/#JuMP.set_time_limit_sec","page":"JuMP","title":"JuMP.set_time_limit_sec","text":"set_time_limit_sec(model::GenericModel, limit::Float64)\n\nSet the time limit (in seconds) of the solver.\n\nCan be unset using unset_time_limit_sec or with limit set to nothing.\n\nSee also: unset_time_limit_sec, time_limit_sec.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_upper_bound","page":"JuMP","title":"set_upper_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_upper_bound","category":"page"},{"location":"api/JuMP/#JuMP.set_upper_bound","page":"JuMP","title":"JuMP.set_upper_bound","text":"set_upper_bound(v::GenericVariableRef, upper::Number)\n\nSet the upper bound of a variable. If one does not exist, create an upper bound constraint.\n\nSee also UpperBoundRef, has_upper_bound, upper_bound, delete_upper_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x <= 1.0);\n\njulia> upper_bound(x)\n1.0\n\njulia> set_upper_bound(x, 2.0)\n\njulia> upper_bound(x)\n2.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#set_value","page":"JuMP","title":"set_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.set_value","category":"page"},{"location":"api/JuMP/#JuMP.set_value","page":"JuMP","title":"JuMP.set_value","text":"set_value(p::NonlinearParameter, v::Number)\n\nStore the value v in the nonlinear parameter p.\n\nExample\n\njulia> model = Model();\n\njulia> @NLparameter(model, p == 0)\np == 0.0\n\njulia> set_value(p, 5)\n5\n\njulia> value(p)\n5.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#shadow_price","page":"JuMP","title":"shadow_price","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.shadow_price","category":"page"},{"location":"api/JuMP/#JuMP.shadow_price","page":"JuMP","title":"JuMP.shadow_price","text":"shadow_price(con_ref::ConstraintRef)\n\nReturn the change in the objective from an infinitesimal relaxation of the constraint.\n\nThis value is computed from dual and can be queried only when has_duals is true and the objective sense is MIN_SENSE or MAX_SENSE (not FEASIBILITY_SENSE). For linear constraints, the shadow prices differ at most in sign from the dual value depending on the objective sense.\n\nSee also reduced_cost.\n\nNotes\n\nThe function simply translates signs from dual and does not validate the conditions needed to guarantee the sensitivity interpretation of the shadow price. The caller is responsible, e.g., for checking whether the solver converged to an optimal primal-dual pair or a proof of infeasibility.\nThe computation is based on the current objective sense of the model. If this has changed since the last solve, the results will be incorrect.\nRelaxation of equality constraints (and hence the shadow price) is defined based on which sense of the equality constraint is active.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#shape","page":"JuMP","title":"shape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.shape","category":"page"},{"location":"api/JuMP/#JuMP.shape","page":"JuMP","title":"JuMP.shape","text":"shape(c::AbstractConstraint)::AbstractShape\n\nReturn the shape of the constraint c.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#show_backend_summary","page":"JuMP","title":"show_backend_summary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.show_backend_summary","category":"page"},{"location":"api/JuMP/#JuMP.show_backend_summary","page":"JuMP","title":"JuMP.show_backend_summary","text":"show_backend_summary(io::IO, model::GenericModel)\n\nPrint a summary of the optimizer backing model.\n\nAbstractModels should implement this method.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#show_constraints_summary","page":"JuMP","title":"show_constraints_summary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.show_constraints_summary","category":"page"},{"location":"api/JuMP/#JuMP.show_constraints_summary","page":"JuMP","title":"JuMP.show_constraints_summary","text":"show_constraints_summary(io::IO, model::AbstractModel)\n\nWrite to io a summary of the number of constraints.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#show_objective_function_summary","page":"JuMP","title":"show_objective_function_summary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.show_objective_function_summary","category":"page"},{"location":"api/JuMP/#JuMP.show_objective_function_summary","page":"JuMP","title":"JuMP.show_objective_function_summary","text":"show_objective_function_summary(io::IO, model::AbstractModel)\n\nWrite to io a summary of the objective function type.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#simplex_iterations","page":"JuMP","title":"simplex_iterations","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.simplex_iterations","category":"page"},{"location":"api/JuMP/#JuMP.simplex_iterations","page":"JuMP","title":"JuMP.simplex_iterations","text":"simplex_iterations(model::GenericModel)\n\nGets the cumulative number of simplex iterations during the most-recent optimization.\n\nSolvers must implement MOI.SimplexIterations() to use this function.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#solution_summary","page":"JuMP","title":"solution_summary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.solution_summary","category":"page"},{"location":"api/JuMP/#JuMP.solution_summary","page":"JuMP","title":"JuMP.solution_summary","text":"solution_summary(model::GenericModel; result::Int = 1, verbose::Bool = false)\n\nReturn a struct that can be used print a summary of the solution in result result.\n\nIf verbose=true, write out the primal solution for every variable and the dual solution for every constraint, excluding those with empty names.\n\nExample\n\nWhen called at the REPL, the summary is automatically printed:\n\njulia> model = Model();\n\njulia> solution_summary(model)\n* Solver : No optimizer attached.\n\n* Status\n  Result count       : 0\n  Termination status : OPTIMIZE_NOT_CALLED\n  Message from the solver:\n  \"optimize not called\"\n\n* Candidate solution (result #1)\n  Primal status      : NO_SOLUTION\n  Dual status        : NO_SOLUTION\n\n* Work counters\n\nUse print to force the printing of the summary from inside a function:\n\njulia> model = Model();\n\njulia> function foo(model)\n           print(solution_summary(model))\n           return\n       end\nfoo (generic function with 1 method)\n\njulia> foo(model)\n* Solver : No optimizer attached.\n\n* Status\n  Result count       : 0\n  Termination status : OPTIMIZE_NOT_CALLED\n  Message from the solver:\n  \"optimize not called\"\n\n* Candidate solution (result #1)\n  Primal status      : NO_SOLUTION\n  Dual status        : NO_SOLUTION\n\n* Work counters\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#solve_time","page":"JuMP","title":"solve_time","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.solve_time","category":"page"},{"location":"api/JuMP/#JuMP.solve_time","page":"JuMP","title":"JuMP.solve_time","text":"solve_time(model::GenericModel)\n\nIf available, returns the solve time reported by the solver. Returns \"ArgumentError: ModelLike of type Solver.Optimizer does not support accessing the attribute MathOptInterface.SolveTimeSec()\" if the attribute is not implemented.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#solver_name","page":"JuMP","title":"solver_name","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.solver_name","category":"page"},{"location":"api/JuMP/#JuMP.solver_name","page":"JuMP","title":"JuMP.solver_name","text":"solver_name(model::GenericModel)\n\nIf available, returns the SolverName property of the underlying optimizer.\n\nReturns \"No optimizer attached\" in AUTOMATIC or MANUAL modes when no optimizer is attached.\n\nReturns \"SolverName() attribute not implemented by the optimizer.\" if the attribute is not implemented.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#start_value","page":"JuMP","title":"start_value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.start_value","category":"page"},{"location":"api/JuMP/#JuMP.start_value","page":"JuMP","title":"JuMP.start_value","text":"start_value(con_ref::ConstraintRef)\n\nReturn the primal start value (MOI.ConstraintPrimalStart) of the constraint con_ref.\n\nNote: If no primal start value has been set, start_value will return nothing.\n\nSee also set_start_value.\n\n\n\n\n\nstart_value(v::GenericVariableRef)\n\nReturn the start value (MOI attribute VariablePrimalStart) of the variable v.\n\nNote: VariablePrimalStarts are sometimes called \"MIP-starts\" or \"warmstarts\".\n\nSee also set_start_value.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#termination_status","page":"JuMP","title":"termination_status","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.termination_status","category":"page"},{"location":"api/JuMP/#JuMP.termination_status","page":"JuMP","title":"JuMP.termination_status","text":"termination_status(model::GenericModel)\n\nReturn a MOI.TerminationStatusCode describing why the solver stopped (i.e., the MOI.TerminationStatus attribute).\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#time_limit_sec","page":"JuMP","title":"time_limit_sec","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.time_limit_sec","category":"page"},{"location":"api/JuMP/#JuMP.time_limit_sec","page":"JuMP","title":"JuMP.time_limit_sec","text":"time_limit_sec(model::GenericModel)\n\nReturn the time limit (in seconds) of the model.\n\nReturns nothing if unset.\n\nSee also: set_time_limit_sec, unset_time_limit_sec.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#triangle_vec","page":"JuMP","title":"triangle_vec","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.triangle_vec","category":"page"},{"location":"api/JuMP/#JuMP.triangle_vec","page":"JuMP","title":"JuMP.triangle_vec","text":"triangle_vec(matrix::Matrix)\n\nReturn the upper triangle of a matrix concatenated into a vector in the order required by JuMP and MathOptInterface for Triangle sets.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, X[1:3, 1:3], Symmetric);\n\njulia> @variable(model, t)\nt\n\njulia> @constraint(model, [t; triangle_vec(X)] in MOI.RootDetConeTriangle(3))\n[t, X[1,1], X[1,2], X[2,2], X[1,3], X[2,3], X[3,3]] ∈ MathOptInterface.RootDetConeTriangle(3)\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#unfix","page":"JuMP","title":"unfix","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.unfix","category":"page"},{"location":"api/JuMP/#JuMP.unfix","page":"JuMP","title":"JuMP.unfix","text":"unfix(v::GenericVariableRef)\n\nDelete the fixing constraint of a variable.\n\nError if one does not exist.\n\nSee also FixRef, is_fixed, fix_value, fix.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x == 1);\n\njulia> is_fixed(x)\ntrue\n\njulia> unfix(x)\n\njulia> is_fixed(x)\nfalse\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#unregister","page":"JuMP","title":"unregister","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.unregister","category":"page"},{"location":"api/JuMP/#JuMP.unregister","page":"JuMP","title":"JuMP.unregister","text":"unregister(model::GenericModel, key::Symbol)\n\nUnregister the name key from model so that a new variable, constraint, or expression can be created with the same key.\n\nNote that this will not delete the object model[key]; it will just remove the reference at model[key]. To delete the object, use delete as well.\n\nSee also: delete, object_dictionary.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, x)\nERROR: An object of name x is already attached to this model. If this\n    is intended, consider using the anonymous construction syntax, e.g.,\n    `x = @variable(model, [1:N], ...)` where the name of the object does\n    not appear inside the macro.\n\n    Alternatively, use `unregister(model, :x)` to first unregister\n    the existing name from the model. Note that this will not delete the\n    object; it will just remove the reference at `model[:x]`.\n\nStacktrace:\n[...]\n\njulia> num_variables(model)\n1\n\njulia> unregister(model, :x)\n\njulia> @variable(model, x)\nx\n\njulia> num_variables(model)\n2\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#unsafe_backend","page":"JuMP","title":"unsafe_backend","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.unsafe_backend","category":"page"},{"location":"api/JuMP/#JuMP.unsafe_backend","page":"JuMP","title":"JuMP.unsafe_backend","text":"unsafe_backend(model::GenericModel)\n\nReturn the innermost optimizer associated with the JuMP model model.\n\nThis function should only be used by advanced users looking to access low-level solver-specific functionality. It has a high-risk of incorrect usage. We strongly suggest you use the alternative suggested below.\n\nSee also: backend.\n\nUnsafe behavior\n\nThis function is unsafe for two main reasons.\n\nFirst, the formulation and order of variables and constraints in the unsafe backend may be different to the variables and constraints in model. This can happen because of bridges, or because the solver requires the variables or constraints in a specific order. In addition, the variable or constraint index returned by index at the JuMP level may be different to the index of the corresponding variable or constraint in the unsafe_backend. There is no solution to this. Use the alternative suggested below instead.\n\nSecond, the unsafe_backend may be empty, or lack some modifications made to the JuMP model. Thus, before calling unsafe_backend you should first call MOI.Utilities.attach_optimizer to ensure that the backend is synchronized with the JuMP model.\n\njulia> import HiGHS\n\njulia> model = Model(HiGHS.Optimizer)\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: EMPTY_OPTIMIZER\nSolver name: HiGHS\n\njulia> MOI.Utilities.attach_optimizer(model)\n\njulia> inner = unsafe_backend(model)\nA HiGHS model with 0 columns and 0 rows.\n\nMoreover, if you modify the JuMP model, the reference you have to the backend (i.e., inner in the example above) may be out-dated, and you should call MOI.Utilities.attach_optimizer again.\n\nThis function is also unsafe in the reverse direction: if you modify the unsafe backend, e.g., by adding a new constraint to inner, the changes may be silently discarded by JuMP when the JuMP model is modified or solved.\n\nAlternative\n\nInstead of unsafe_backend, create a model using direct_model and call backend instead.\n\nFor example, instead of:\n\njulia> import HiGHS\n\njulia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 0)\nx\n\njulia> MOI.Utilities.attach_optimizer(model)\n\njulia> highs = unsafe_backend(model)\nA HiGHS model with 1 columns and 0 rows.\n\nUse:\n\njulia> import HiGHS\n\njulia> model = direct_model(HiGHS.Optimizer());\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 0)\nx\n\njulia> highs = backend(model)  # No need to call `attach_optimizer`.\nA HiGHS model with 1 columns and 0 rows.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#unset_binary","page":"JuMP","title":"unset_binary","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.unset_binary","category":"page"},{"location":"api/JuMP/#JuMP.unset_binary","page":"JuMP","title":"JuMP.unset_binary","text":"unset_binary(variable_ref::GenericVariableRef)\n\nRemove the binary constraint on the variable variable_ref.\n\nSee also BinaryRef, is_binary, set_binary.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x, Bin);\n\njulia> is_binary(x)\ntrue\n\njulia> unset_binary(x)\n\njulia> is_binary(x)\nfalse\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#unset_integer","page":"JuMP","title":"unset_integer","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.unset_integer","category":"page"},{"location":"api/JuMP/#JuMP.unset_integer","page":"JuMP","title":"JuMP.unset_integer","text":"unset_integer(variable_ref::GenericVariableRef)\n\nRemove the integrality constraint on the variable variable_ref.\n\nErrors if one does not exist.\n\nSee also IntegerRef, is_integer, set_integer.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x, Int);\n\njulia> is_integer(x)\ntrue\n\njulia> unset_integer(x)\n\njulia> is_integer(x)\nfalse\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#unset_silent","page":"JuMP","title":"unset_silent","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.unset_silent","category":"page"},{"location":"api/JuMP/#JuMP.unset_silent","page":"JuMP","title":"JuMP.unset_silent","text":"unset_silent(model::GenericModel)\n\nNeutralize the effect of the set_silent function and let the solver attributes control the verbosity.\n\nSee also: set_silent.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#unset_time_limit_sec","page":"JuMP","title":"unset_time_limit_sec","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.unset_time_limit_sec","category":"page"},{"location":"api/JuMP/#JuMP.unset_time_limit_sec","page":"JuMP","title":"JuMP.unset_time_limit_sec","text":"unset_time_limit_sec(model::GenericModel)\n\nUnset the time limit of the solver.\n\nSee also: set_time_limit_sec, time_limit_sec.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#upper_bound","page":"JuMP","title":"upper_bound","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.upper_bound","category":"page"},{"location":"api/JuMP/#JuMP.upper_bound","page":"JuMP","title":"JuMP.upper_bound","text":"upper_bound(v::GenericVariableRef)\n\nReturn the upper bound of a variable.\n\nError if one does not exist.\n\nSee also UpperBoundRef, has_upper_bound, set_upper_bound, delete_upper_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x <= 1.0);\n\njulia> upper_bound(x)\n1.0\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#value","page":"JuMP","title":"value","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.value","category":"page"},{"location":"api/JuMP/#JuMP.value","page":"JuMP","title":"JuMP.value","text":"value(con_ref::ConstraintRef; result::Int = 1)\n\nReturn the primal value of constraint con_ref associated with result index result of the most-recent solution returned by the solver.\n\nThat is, if con_ref is the reference of a constraint func-in-set, it returns the value of func evaluated at the value of the variables (given by value(::GenericVariableRef)).\n\nUse has_values to check if a result exists before asking for values.\n\nSee also: result_count.\n\nNote\n\nFor scalar constraints, the constant is moved to the set so it is not taken into account in the primal value of the constraint. For instance, the constraint @constraint(model, 2x + 3y + 1 == 5) is transformed into 2x + 3y-in-MOI.EqualTo(4) so the value returned by this function is the evaluation of 2x + 3y.\n\n\n\n\n\nvalue(var_value::Function, con_ref::ConstraintRef)\n\nEvaluate the primal value of the constraint con_ref using var_value(v) as the value for each variable v.\n\n\n\n\n\nvalue(v::GenericVariableRef; result = 1)\n\nReturn the value of variable v associated with result index result of the most-recent returned by the solver.\n\nUse has_values to check if a result exists before asking for values.\n\nSee also: result_count.\n\n\n\n\n\nvalue(var_value::Function, v::GenericVariableRef)\n\nEvaluate the value of the variable v as var_value(v).\n\n\n\n\n\nvalue(var_value::Function, ex::GenericAffExpr)\n\nEvaluate ex using var_value(v) as the value for each variable v.\n\n\n\n\n\nvalue(v::GenericAffExpr; result::Int = 1)\n\nReturn the value of the GenericAffExpr v associated with result index result of the most-recent solution returned by the solver.\n\nSee also: result_count.\n\n\n\n\n\nvalue(var_value::Function, ex::GenericQuadExpr)\n\nEvaluate ex using var_value(v) as the value for each variable v.\n\n\n\n\n\nvalue(v::GenericQuadExpr; result::Int = 1)\n\nReturn the value of the GenericQuadExpr v associated with result index result of the most-recent solution returned by the solver.\n\nReplaces getvalue for most use cases.\n\nSee also: result_count.\n\n\n\n\n\nvalue(p::NonlinearParameter)\n\nReturn the current value stored in the nonlinear parameter p.\n\nExample\n\njulia> model = Model();\n\njulia> @NLparameter(model, p == 10)\np == 10.0\n\njulia> value(p)\n10.0\n\n\n\n\n\nvalue(ex::NonlinearExpression; result::Int = 1)\n\nReturn the value of the NonlinearExpression ex associated with result index result of the most-recent solution returned by the solver.\n\nReplaces getvalue for most use cases.\n\nSee also: result_count.\n\n\n\n\n\nvalue(var_value::Function, ex::NonlinearExpression)\n\nEvaluate ex using var_value(v) as the value for each variable v.\n\n\n\n\n\nvalue(c::NonlinearConstraintRef; result::Int = 1)\n\nReturn the value of the NonlinearConstraintRef c associated with result index result of the most-recent solution returned by the solver.\n\nSee also: result_count.\n\n\n\n\n\nvalue(var_value::Function, c::NonlinearConstraintRef)\n\nEvaluate c using var_value(v) as the value for each variable v.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#value_type","page":"JuMP","title":"value_type","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.value_type","category":"page"},{"location":"api/JuMP/#JuMP.value_type","page":"JuMP","title":"JuMP.value_type","text":"value_type(::Type{<:Union{AbstractModel,AbstractVariableRef}})\n\nReturn the return type of value for variables of that model. It defaults to Float64 if it is not implemented.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#variable_by_name","page":"JuMP","title":"variable_by_name","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.variable_by_name","category":"page"},{"location":"api/JuMP/#JuMP.variable_by_name","page":"JuMP","title":"JuMP.variable_by_name","text":"variable_by_name(\n    model::AbstractModel,\n    name::String,\n)::Union{AbstractVariableRef,Nothing}\n\nReturns the reference of the variable with name attribute name or Nothing if no variable has this name attribute. Throws an error if several variables have name as their name attribute.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> variable_by_name(model, \"x\")\nx\n\njulia> @variable(model, base_name=\"x\")\nx\n\njulia> variable_by_name(model, \"x\")\nERROR: Multiple variables have the name x.\nStacktrace:\n [1] error(::String) at ./error.jl:33\n [2] get(::MOIU.Model{Float64}, ::Type{MathOptInterface.VariableIndex}, ::String) at /home/blegat/.julia/dev/MathOptInterface/src/Utilities/model.jl:222\n [3] get at /home/blegat/.julia/dev/MathOptInterface/src/Utilities/universalfallback.jl:201 [inlined]\n [4] get(::MathOptInterface.Utilities.CachingOptimizer{MathOptInterface.AbstractOptimizer,MathOptInterface.Utilities.UniversalFallback{MOIU.Model{Float64}}}, ::Type{MathOptInterface.VariableIndex}, ::String) at /home/blegat/.julia/dev/MathOptInterface/src/Utilities/cachingoptimizer.jl:490\n [5] variable_by_name(::GenericModel, ::String) at /home/blegat/.julia/dev/JuMP/src/variables.jl:268\n [6] top-level scope at none:0\n\njulia> var = @variable(model, base_name=\"y\")\ny\n\njulia> variable_by_name(model, \"y\")\ny\n\njulia> set_name(var, \"z\")\n\njulia> variable_by_name(model, \"y\")\n\njulia> variable_by_name(model, \"z\")\nz\n\njulia> @variable(model, u[1:2])\n2-element Vector{VariableRef}:\n u[1]\n u[2]\n\njulia> variable_by_name(model, \"u[2]\")\nu[2]\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#variable_ref_type","page":"JuMP","title":"variable_ref_type","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.variable_ref_type","category":"page"},{"location":"api/JuMP/#JuMP.variable_ref_type","page":"JuMP","title":"JuMP.variable_ref_type","text":"variable_ref_type(::Union{F,Type{F}}) where {F}\n\nA helper function used internally by JuMP and some JuMP extensions. Returns the variable type associated with the model or expression type F.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#vectorize","page":"JuMP","title":"vectorize","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.vectorize","category":"page"},{"location":"api/JuMP/#JuMP.vectorize","page":"JuMP","title":"JuMP.vectorize","text":"vectorize(matrix::AbstractMatrix, ::Shape)\n\nConvert the matrix into a vector according to Shape.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#write_to_file","page":"JuMP","title":"write_to_file","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.write_to_file","category":"page"},{"location":"api/JuMP/#JuMP.write_to_file","page":"JuMP","title":"JuMP.write_to_file","text":"write_to_file(\n    model::GenericModel,\n    filename::String;\n    format::MOI.FileFormats.FileFormat = MOI.FileFormats.FORMAT_AUTOMATIC,\n    kwargs...,\n)\n\nWrite the JuMP model model to filename in the format format.\n\nIf the filename ends in .gz, it will be compressed using Gzip. If the filename ends in .bz2, it will be compressed using BZip2.\n\nOther kwargs are passed to the Model constructor of the chosen format.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#AbstractConstraint","page":"JuMP","title":"AbstractConstraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractConstraint","category":"page"},{"location":"api/JuMP/#JuMP.AbstractConstraint","page":"JuMP","title":"JuMP.AbstractConstraint","text":"abstract type AbstractConstraint\n\nAn abstract base type for all constraint types. AbstractConstraints store the function and set directly, unlike ConstraintRefs that are merely references to constraints stored in a model. AbstractConstraints do not need to be attached to a model.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AbstractJuMPScalar","page":"JuMP","title":"AbstractJuMPScalar","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractJuMPScalar","category":"page"},{"location":"api/JuMP/#JuMP.AbstractJuMPScalar","page":"JuMP","title":"JuMP.AbstractJuMPScalar","text":"AbstractJuMPScalar <: MutableArithmetics.AbstractMutable\n\nAbstract base type for all scalar types\n\nThe subtyping of AbstractMutable will allow calls of some Base functions to be redirected to a method in MA that handles type promotion more carefully (e.g. the promotion in sparse matrix products in SparseArrays usually does not work for JuMP types) and exploits the mutability of AffExpr and QuadExpr.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AbstractModel","page":"JuMP","title":"AbstractModel","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractModel","category":"page"},{"location":"api/JuMP/#JuMP.AbstractModel","page":"JuMP","title":"JuMP.AbstractModel","text":"AbstractModel\n\nAn abstract type that should be subtyped for users creating JuMP extensions.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AbstractScalarSet","page":"JuMP","title":"AbstractScalarSet","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractScalarSet","category":"page"},{"location":"api/JuMP/#JuMP.AbstractScalarSet","page":"JuMP","title":"JuMP.AbstractScalarSet","text":"AbstractScalarSet\n\nAn abstract type for defining new scalar sets in JuMP.\n\nImplement moi_set(::AbstractScalarSet) to convert the type into an MOI set.\n\nSee also: moi_set.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AbstractShape","page":"JuMP","title":"AbstractShape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractShape","category":"page"},{"location":"api/JuMP/#JuMP.AbstractShape","page":"JuMP","title":"JuMP.AbstractShape","text":"AbstractShape\n\nAbstract vectorizable shape. Given a flat vector form of an object of shape shape, the original object can be obtained by reshape_vector.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AbstractVariable","page":"JuMP","title":"AbstractVariable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractVariable","category":"page"},{"location":"api/JuMP/#JuMP.AbstractVariable","page":"JuMP","title":"JuMP.AbstractVariable","text":"AbstractVariable\n\nVariable returned by build_variable. It represents a variable that has not been added yet to any model. It can be added to a given model with add_variable.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AbstractVariableRef","page":"JuMP","title":"AbstractVariableRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractVariableRef","category":"page"},{"location":"api/JuMP/#JuMP.AbstractVariableRef","page":"JuMP","title":"JuMP.AbstractVariableRef","text":"AbstractVariableRef\n\nVariable returned by add_variable. Affine (resp. quadratic) operations with variables of type V<:AbstractVariableRef and coefficients of type T     create a GenericAffExpr{T,V} (resp. GenericQuadExpr{T,V}).\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AbstractVectorSet","page":"JuMP","title":"AbstractVectorSet","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AbstractVectorSet","category":"page"},{"location":"api/JuMP/#JuMP.AbstractVectorSet","page":"JuMP","title":"JuMP.AbstractVectorSet","text":"AbstractVectorSet\n\nAn abstract type for defining new sets in JuMP.\n\nImplement moi_set(::AbstractVectorSet, dim::Int) to convert the type into an MOI set.\n\nSee also: moi_set.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#AffExpr","page":"JuMP","title":"AffExpr","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AffExpr","category":"page"},{"location":"api/JuMP/#JuMP.AffExpr","page":"JuMP","title":"JuMP.AffExpr","text":"AffExpr\n\nAlias for GenericAffExpr{Float64,VariableRef}, the specific GenericAffExpr used by JuMP.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#BinaryRef","page":"JuMP","title":"BinaryRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.BinaryRef","category":"page"},{"location":"api/JuMP/#JuMP.BinaryRef","page":"JuMP","title":"JuMP.BinaryRef","text":"BinaryRef(v::GenericVariableRef)\n\nReturn a constraint reference to the constraint constraining v to be binary. Errors if one does not exist.\n\nSee also is_binary, set_binary, unset_binary.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x, Bin);\n\njulia> BinaryRef(x)\nx binary\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#BridgeableConstraint","page":"JuMP","title":"BridgeableConstraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.BridgeableConstraint","category":"page"},{"location":"api/JuMP/#JuMP.BridgeableConstraint","page":"JuMP","title":"JuMP.BridgeableConstraint","text":"BridgeableConstraint(\n    constraint::C,\n    bridge_type::B;\n    coefficient_type::Type{T} = Float64,\n) where {C<:AbstractConstraint,B<:Type{<:MOI.Bridges.AbstractBridge},T}\n\nAn AbstractConstraint representinng that constraint that can be bridged by the bridge of type bridge_type{coefficient_type}.\n\nAdding a BridgeableConstraint to a model is equivalent to:\n\nadd_bridge(model, bridge_type; coefficient_type = coefficient_type)\nadd_constraint(model, constraint)\n\nExample\n\nGiven a new scalar set type CustomSet with a bridge CustomBridge that can bridge F-in-CustomSet constraints, when the user does:\n\nmodel = Model()\n@variable(model, x)\n@constraint(model, x + 1 in CustomSet())\noptimize!(model)\n\nwith an optimizer that does not support F-in-CustomSet constraints, the constraint will not be bridged unless they first call add_bridge(model, CustomBridge).\n\nIn order to automatically add the CustomBridge to any model to which an F-in-CustomSet is added, add the following method:\n\nfunction JuMP.build_constraint(\n    _error::Function,\n    func::AbstractJuMPScalar,\n    set::CustomSet,\n)\n    constraint = ScalarConstraint(func, set)\n    return BridgeableConstraint(constraint, CustomBridge)\nend\n\nNote\n\nJuMP extensions should extend JuMP.build_constraint only if they also defined CustomSet, for three reasons:\n\nIt is problematic if multiple extensions overload the same JuMP method.\nA missing method will not inform the users that they forgot to load the extension module defining the build_constraint method.\nDefining a method where neither the function nor any of the argument types are defined in the package is called type piracy and is discouraged in the Julia style guide.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ComplexPlane","page":"JuMP","title":"ComplexPlane","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ComplexPlane","category":"page"},{"location":"api/JuMP/#JuMP.ComplexPlane","page":"JuMP","title":"JuMP.ComplexPlane","text":"ComplexPlane\n\nComplex plane object that can be used to create a complex variable in the @variable macro.\n\nExample\n\nConsider the following example:\n\njulia> model = Model();\n\njulia> @variable(model, x in ComplexPlane())\nreal(x) + imag(x) im\n\njulia> all_variables(model)\n2-element Vector{VariableRef}:\n real(x)\n imag(x)\n\nWe see in the output of the last command that two real variables were created. The Julia variable x binds to an affine expression in terms of these two variables that parametrize the complex plane.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ComplexVariable","page":"JuMP","title":"ComplexVariable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ComplexVariable","category":"page"},{"location":"api/JuMP/#JuMP.ComplexVariable","page":"JuMP","title":"JuMP.ComplexVariable","text":"ComplexVariable{S,T,U,V} <: AbstractVariable\n\nA struct used when adding complex variables.\n\nSee also: ComplexPlane.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ConstraintNotOwned","page":"JuMP","title":"ConstraintNotOwned","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ConstraintNotOwned","category":"page"},{"location":"api/JuMP/#JuMP.ConstraintNotOwned","page":"JuMP","title":"JuMP.ConstraintNotOwned","text":"struct ConstraintNotOwned{C <: ConstraintRef} <: Exception\n    constraint_ref::C\nend\n\nThe constraint constraint_ref was used in a model different to owner_model(constraint_ref).\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ConstraintRef","page":"JuMP","title":"ConstraintRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ConstraintRef","category":"page"},{"location":"api/JuMP/#JuMP.ConstraintRef","page":"JuMP","title":"JuMP.ConstraintRef","text":"ConstraintRef\n\nHolds a reference to the model and the corresponding MOI.ConstraintIndex.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#FixRef","page":"JuMP","title":"FixRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.FixRef","category":"page"},{"location":"api/JuMP/#JuMP.FixRef","page":"JuMP","title":"JuMP.FixRef","text":"FixRef(v::GenericVariableRef)\n\nReturn a constraint reference to the constraint fixing the value of v.\n\nErrors if one does not exist.\n\nSee also is_fixed, fix_value, fix, unfix.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x == 1);\n\njulia> FixRef(x)\nx = 1\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#GenericAffExpr","page":"JuMP","title":"GenericAffExpr","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.GenericAffExpr","category":"page"},{"location":"api/JuMP/#JuMP.GenericAffExpr","page":"JuMP","title":"JuMP.GenericAffExpr","text":"mutable struct GenericAffExpr{CoefType,VarType} <: AbstractJuMPScalar\n    constant::CoefType\n    terms::OrderedDict{VarType,CoefType}\nend\n\nAn expression type representing an affine expression of the form: sum a_i x_i + c.\n\nFields\n\n.constant: the constant c in the expression.\n.terms: an OrderedDict, with keys of VarType and values of CoefType describing the sparse vector a.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#GenericModel","page":"JuMP","title":"GenericModel","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.GenericModel","category":"page"},{"location":"api/JuMP/#JuMP.GenericModel","page":"JuMP","title":"JuMP.GenericModel","text":"GenericModel{T}(\n    [optimizer_factory;]\n    add_bridges::Bool = true,\n) where {T<:Real}\n\nCreate a new instance of a JuMP model.\n\nIf optimizer_factory is provided, the model is initialized with the optimizer returned by MOI.instantiate(optimizer_factory).\n\nIf optimizer_factory is not provided, use set_optimizer to set the optimizer before calling optimize!.\n\nIf add_bridges, JuMP adds a MOI.Bridges.LazyBridgeOptimizer to automatically reformulate the problem into a form supported by the optimizer.\n\nValue type T\n\nPassing a type other than Float64 as the value type T is an advanced operation. The value type must match that expected by the chosen optimizer. Consult the optimizers documentation for details.\n\nIf not documented, assume that the optimizer supports only Float64.\n\nChoosing an unsupported value type will throw an MOI.UnsupportedConstraint or an MOI.UnsupportedAttribute error, the timing of which (during the model construction or during a call to optimize!) depends on how the solver is interfaced to JuMP.\n\nExample\n\njulia> model = GenericModel{BigFloat}();\n\njulia> typeof(model)\nGenericModel{BigFloat}\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#GenericNonlinearExpr","page":"JuMP","title":"GenericNonlinearExpr","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.GenericNonlinearExpr","category":"page"},{"location":"api/JuMP/#JuMP.GenericNonlinearExpr","page":"JuMP","title":"JuMP.GenericNonlinearExpr","text":"GenericNonlinearExpr{V}(head::Symbol, args::Vector{Any})\nGenericNonlinearExpr{V}(head::Symbol, args::Any...)\n\nThe scalar-valued nonlinear function head(args...), represented as a symbolic expression tree, with the call operator head and ordered arguments in args.\n\nV is the type of AbstractVariableRef present in the expression, and is used to help dispatch JuMP extensions.\n\nhead\n\nThe head::Symbol must be an operator supported by the model.\n\nThe default list of supported univariate operators is given by:\n\nMOI.Nonlinear.DEFAULT_UNIVARIATE_OPERATORS\n\nand the default list of supported multivariate operators is given by:\n\nMOI.Nonlinear.DEFAULT_MULTIVARIATE_OPERATORS\n\nAdditional operators can be add using @operator.\n\nSee the full list of operators supported by a MOI.ModelLike by querying the MOI.ListOfSupportedNonlinearOperators attribute.\n\nargs\n\nThe vector args contains the arguments to the nonlinear function. If the operator is univariate, it must contain one element. Otherwise, it may contain multiple elements.\n\nGiven a subtype of AbstractVariableRef, V, for GenericNonlinearExpr{V}, each element must be one of the following:\n\nA constant value of type <:Real\nA V\nA GenericAffExpr{T,V}\nA GenericQuadExpr{T,V}\nA GenericNonlinearExpr{V}\n\nwhere T<:Real and T == value_type(V).\n\nUnsupported operators\n\nIf the optimizer does not support head, an MOI.UnsupportedNonlinearOperator error will be thrown.\n\nThere is no guarantee about when this error will be thrown; it may be thrown when the function is first added to the model, or it may be thrown when optimize! is called.\n\nExample\n\nTo represent the function f(x) = sin(x)^2, do:\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> f = sin(x)^2\nsin(x) ^ 2.0\n\njulia> f = GenericNonlinearExpr{VariableRef}(\n           :^,\n           GenericNonlinearExpr{VariableRef}(:sin, x),\n           2.0,\n       )\nsin(x) ^ 2.0\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#GenericQuadExpr","page":"JuMP","title":"GenericQuadExpr","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.GenericQuadExpr","category":"page"},{"location":"api/JuMP/#JuMP.GenericQuadExpr","page":"JuMP","title":"JuMP.GenericQuadExpr","text":"mutable struct GenericQuadExpr{CoefType,VarType} <: AbstractJuMPScalar\n    aff::GenericAffExpr{CoefType,VarType}\n    terms::OrderedDict{UnorderedPair{VarType}, CoefType}\nend\n\nAn expression type representing an quadratic expression of the form: sum q_ij x_i x_j + sum a_i x_i + c.\n\nFields\n\n.aff: an GenericAffExpr representing the affine portion of the expression.\n.terms: an OrderedDict, with keys of UnorderedPair{VarType} and values of CoefType, describing the sparse list of terms q.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#GenericReferenceMap","page":"JuMP","title":"GenericReferenceMap","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.GenericReferenceMap","category":"page"},{"location":"api/JuMP/#JuMP.GenericReferenceMap","page":"JuMP","title":"JuMP.GenericReferenceMap","text":"GenericReferenceMap{T}\n\nMapping between variable and constraint reference of a model and its copy. The reference of the copied model can be obtained by indexing the map with the reference of the corresponding reference of the original model.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#GenericVariableRef","page":"JuMP","title":"GenericVariableRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.GenericVariableRef","category":"page"},{"location":"api/JuMP/#JuMP.GenericVariableRef","page":"JuMP","title":"JuMP.GenericVariableRef","text":"GenericVariableRef{T} <: AbstractVariableRef\n\nHolds a reference to the model and the corresponding MOI.VariableIndex.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#HermitianMatrixShape","page":"JuMP","title":"HermitianMatrixShape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.HermitianMatrixShape","category":"page"},{"location":"api/JuMP/#JuMP.HermitianMatrixShape","page":"JuMP","title":"JuMP.HermitianMatrixShape","text":"HermitianMatrixShape\n\nShape object for a Hermitian square matrix of side_dimension rows and columns. The vectorized form corresponds to MOI.HermitianPositiveSemidefiniteConeTriangle.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#HermitianMatrixSpace","page":"JuMP","title":"HermitianMatrixSpace","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.HermitianMatrixSpace","category":"page"},{"location":"api/JuMP/#JuMP.HermitianMatrixSpace","page":"JuMP","title":"JuMP.HermitianMatrixSpace","text":"HermitianMatrixSpace()\n\nUse in the @variable macro to constrain a matrix of variables to be hermitian.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2] in HermitianMatrixSpace())\n2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(Q[1,1])                    real(Q[1,2]) + imag(Q[1,2]) im\n real(Q[1,2]) - imag(Q[1,2]) im  real(Q[2,2])\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#HermitianPSDCone","page":"JuMP","title":"HermitianPSDCone","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.HermitianPSDCone","category":"page"},{"location":"api/JuMP/#JuMP.HermitianPSDCone","page":"JuMP","title":"JuMP.HermitianPSDCone","text":"HermitianPSDCone\n\nHermitian positive semidefinite cone object that can be used to create a Hermitian positive semidefinite square matrix in the @variable and @constraint macros.\n\nExample\n\nConsider the following example:\n\njulia> model = Model();\n\njulia> @variable(model, H[1:3, 1:3] in HermitianPSDCone())\n3×3 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(H[1,1])                    …  real(H[1,3]) + imag(H[1,3]) im\n real(H[1,2]) - imag(H[1,2]) im     real(H[2,3]) + imag(H[2,3]) im\n real(H[1,3]) - imag(H[1,3]) im     real(H[3,3])\n\njulia> all_variables(model)\n9-element Vector{VariableRef}:\n real(H[1,1])\n real(H[1,2])\n real(H[2,2])\n real(H[1,3])\n real(H[2,3])\n real(H[3,3])\n imag(H[1,2])\n imag(H[1,3])\n imag(H[2,3])\n\njulia> all_constraints(model, Vector{VariableRef}, MOI.HermitianPositiveSemidefiniteConeTriangle)\n1-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.HermitianPositiveSemidefiniteConeTriangle}}}:\n [real(H[1,1]), real(H[1,2]), real(H[2,2]), real(H[1,3]), real(H[2,3]), real(H[3,3]), imag(H[1,2]), imag(H[1,3]), imag(H[2,3])] ∈ MathOptInterface.HermitianPositiveSemidefiniteConeTriangle(3)\n\nWe see in the output of the last commands that 9 real variables were created. The matrix H contrains affine expressions in terms of these 9 variables that parametrize a Hermitian matrix.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#IntegerRef","page":"JuMP","title":"IntegerRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.IntegerRef","category":"page"},{"location":"api/JuMP/#JuMP.IntegerRef","page":"JuMP","title":"JuMP.IntegerRef","text":"IntegerRef(v::GenericVariableRef)\n\nReturn a constraint reference to the constraint constraining v to be integer.\n\nErrors if one does not exist.\n\nSee also is_integer, set_integer, unset_integer.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x, Int);\n\njulia> IntegerRef(x)\nx integer\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#LinearTermIterator","page":"JuMP","title":"LinearTermIterator","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.LinearTermIterator","category":"page"},{"location":"api/JuMP/#JuMP.LinearTermIterator","page":"JuMP","title":"JuMP.LinearTermIterator","text":"LinearTermIterator{GAE<:GenericAffExpr}\n\nA struct that implements the iterate protocol in order to iterate over tuples of (coefficient, variable) in the GenericAffExpr.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#LowerBoundRef","page":"JuMP","title":"LowerBoundRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.LowerBoundRef","category":"page"},{"location":"api/JuMP/#JuMP.LowerBoundRef","page":"JuMP","title":"JuMP.LowerBoundRef","text":"LowerBoundRef(v::GenericVariableRef)\n\nReturn a constraint reference to the lower bound constraint of v.\n\nErrors if one does not exist.\n\nSee also has_lower_bound, lower_bound, set_lower_bound, delete_lower_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x >= 1.0);\n\njulia> LowerBoundRef(x)\nx ≥ 1\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#Model","page":"JuMP","title":"Model","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Model","category":"page"},{"location":"api/JuMP/#JuMP.Model","page":"JuMP","title":"JuMP.Model","text":"Model([optimizer_factory;] add_bridges::Bool = true)\n\nCreate a new instance of a JuMP model.\n\nIf optimizer_factory is provided, the model is initialized with thhe optimizer returned by MOI.instantiate(optimizer_factory).\n\nIf optimizer_factory is not provided, use set_optimizer to set the optimizer before calling optimize!.\n\nIf add_bridges, JuMP adds a MOI.Bridges.LazyBridgeOptimizer to automatically reformulate the problem into a form supported by the optimizer.\n\nExample\n\njulia> import Ipopt\n\njulia> model = Model(Ipopt.Optimizer);\n\njulia> solver_name(model)\n\"Ipopt\"\n\njulia> import HiGHS\n\njulia> import MultiObjectiveAlgorithms as MOA\n\njulia> model = Model(() -> MOA.Optimizer(HiGHS.Optimizer); add_bridges = false);\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ModelMode","page":"JuMP","title":"ModelMode","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ModelMode","category":"page"},{"location":"api/JuMP/#JuMP.ModelMode","page":"JuMP","title":"JuMP.ModelMode","text":"ModelMode\n\nAn enum to describe the state of the CachingOptimizer inside a JuMP model.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#NLPEvaluator","page":"JuMP","title":"NLPEvaluator","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NLPEvaluator","category":"page"},{"location":"api/JuMP/#JuMP.NLPEvaluator","page":"JuMP","title":"JuMP.NLPEvaluator","text":"NLPEvaluator(\n    model::Model,\n    _differentiation_backend::MOI.Nonlinear.AbstractAutomaticDifferentiation =\n        MOI.Nonlinear.SparseReverseMode(),\n)\n\nReturn an MOI.AbstractNLPEvaluator constructed from model\n\nwarning: Warning\nBefore using, you must initialize the evaluator using MOI.initialize.\n\nExperimental\n\nThese features may change or be removed in any future version of JuMP.\n\nPass _differentiation_backend to specify the differentiation backend used to compute derivatives.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#NoOptimizer","page":"JuMP","title":"NoOptimizer","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NoOptimizer","category":"page"},{"location":"api/JuMP/#JuMP.NoOptimizer","page":"JuMP","title":"JuMP.NoOptimizer","text":"struct NoOptimizer <: Exception end\n\nNo optimizer is set. The optimizer can be provided to the Model constructor or by calling set_optimizer.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#NonlinearConstraintIndex","page":"JuMP","title":"NonlinearConstraintIndex","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NonlinearConstraintIndex","category":"page"},{"location":"api/JuMP/#JuMP.NonlinearConstraintIndex","page":"JuMP","title":"JuMP.NonlinearConstraintIndex","text":"ConstraintIndex\n\nAn index to a nonlinear constraint that is returned by add_constraint.\n\nGiven data::Model and c::ConstraintIndex, use data[c] to retrieve the corresponding Constraint.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#NonlinearConstraintRef","page":"JuMP","title":"NonlinearConstraintRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NonlinearConstraintRef","category":"page"},{"location":"api/JuMP/#JuMP.NonlinearConstraintRef","page":"JuMP","title":"JuMP.NonlinearConstraintRef","text":"NonlinearConstraintRef\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#NonlinearExpr","page":"JuMP","title":"NonlinearExpr","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NonlinearExpr","category":"page"},{"location":"api/JuMP/#JuMP.NonlinearExpr","page":"JuMP","title":"JuMP.NonlinearExpr","text":"NonlinearExpr\n\nAlias for GenericNonlinearExpr{VariableRef}, the specific GenericNonlinearExpr used by JuMP.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#NonlinearExpression","page":"JuMP","title":"NonlinearExpression","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NonlinearExpression","category":"page"},{"location":"api/JuMP/#JuMP.NonlinearExpression","page":"JuMP","title":"JuMP.NonlinearExpression","text":"NonlinearExpression <: AbstractJuMPScalar\n\nA struct to represent a nonlinear expression.\n\nCreate an expression using @NLexpression.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#NonlinearOperator","page":"JuMP","title":"NonlinearOperator","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NonlinearOperator","category":"page"},{"location":"api/JuMP/#JuMP.NonlinearOperator","page":"JuMP","title":"JuMP.NonlinearOperator","text":"NonlinearOperator(func::Function, head::Symbol)\n\nA callable struct (functor) representing a function named head.\n\nWhen called with AbstractJuMPScalars, the struct returns a GenericNonlinearExpr.\n\nWhen called with non-JuMP types, the struct returns the evaluation of func(args...).\n\nUnless head is special-cased by the optimizer, the operator must have already been added to the model using add_nonlinear_operator or @operator.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> f(x::Float64) = x^2\nf (generic function with 1 method)\n\njulia> ∇f(x::Float64) = 2 * x\n∇f (generic function with 1 method)\n\njulia> ∇²f(x::Float64) = 2.0\n∇²f (generic function with 1 method)\n\njulia> @operator(model, op_f, 1, f, ∇f, ∇²f)\nNonlinearOperator(f, :op_f)\n\njulia> bar = NonlinearOperator(f, :op_f)\nNonlinearOperator(f, :op_f)\n\njulia> @objective(model, Min, bar(x))\nop_f(x)\n\njulia> bar(2.0)\n4.0\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#NonlinearParameter","page":"JuMP","title":"NonlinearParameter","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NonlinearParameter","category":"page"},{"location":"api/JuMP/#JuMP.NonlinearParameter","page":"JuMP","title":"JuMP.NonlinearParameter","text":"NonlinearParameter <: AbstractJuMPScalar\n\nA struct to represent a nonlinear parameter.\n\nCreate a parameter using @NLparameter.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#Nonnegatives","page":"JuMP","title":"Nonnegatives","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Nonnegatives","category":"page"},{"location":"api/JuMP/#JuMP.Nonnegatives","page":"JuMP","title":"JuMP.Nonnegatives","text":"Nonnegatives()\n\nThe JuMP equivalent of the MOI.Nonnegatives set, in which the dimension is inferred from the corresponding function.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> @constraint(model, x in Nonnegatives())\n[x[1], x[2]] ∈ MathOptInterface.Nonnegatives(2)\n\njulia> A = [1 2; 3 4];\n\njulia> b = [5, 6];\n\njulia> @constraint(model, A * x >= b)\n[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonnegatives(2)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#Nonpositives","page":"JuMP","title":"Nonpositives","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Nonpositives","category":"page"},{"location":"api/JuMP/#JuMP.Nonpositives","page":"JuMP","title":"JuMP.Nonpositives","text":"Nonpositives()\n\nThe JuMP equivalent of the MOI.Nonpositives set, in which the dimension is inferred from the corresponding function.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> @constraint(model, x in Nonpositives())\n[x[1], x[2]] ∈ MathOptInterface.Nonpositives(2)\n\njulia> A = [1 2; 3 4];\n\njulia> b = [5, 6];\n\njulia> @constraint(model, A * x <= b)\n[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Nonpositives(2)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#OptimizationSense","page":"JuMP","title":"OptimizationSense","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OptimizationSense","category":"page"},{"location":"api/JuMP/#JuMP.OptimizationSense","page":"JuMP","title":"JuMP.OptimizationSense","text":"OptimizationSense\n\nAn enum for the value of the ObjectiveSense attribute.\n\nValues\n\nPossible values are:\n\nMIN_SENSE: the goal is to minimize the objective function\nMAX_SENSE: the goal is to maximize the objective function\nFEASIBILITY_SENSE: the model does not have an objective function\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#OptimizeNotCalled","page":"JuMP","title":"OptimizeNotCalled","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OptimizeNotCalled","category":"page"},{"location":"api/JuMP/#JuMP.OptimizeNotCalled","page":"JuMP","title":"JuMP.OptimizeNotCalled","text":"struct OptimizeNotCalled <: Exception end\n\nA result attribute cannot be queried before optimize! is called.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#PSDCone","page":"JuMP","title":"PSDCone","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.PSDCone","category":"page"},{"location":"api/JuMP/#JuMP.PSDCone","page":"JuMP","title":"JuMP.PSDCone","text":"PSDCone\n\nPositive semidefinite cone object that can be used to constrain a square matrix to be positive semidefinite in the @constraint macro. If the matrix has type Symmetric then the columns vectorization (the vector obtained by concatenating the columns) of its upper triangular part is constrained to belong to the MOI.PositiveSemidefiniteConeTriangle set, otherwise its column vectorization is constrained to belong to the MOI.PositiveSemidefiniteConeSquare set.\n\nExample\n\nConsider the following example:\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> a = [ x 2x\n            2x  x];\n\njulia> b = [1 2\n            2 4];\n\njulia> cref = @constraint(model, a >= b, PSDCone())\n[x - 1    2 x - 2;\n 2 x - 2  x - 4] ∈ PSDCone()\n\njulia> jump_function(constraint_object(cref))\n4-element Vector{AffExpr}:\n x - 1\n 2 x - 2\n 2 x - 2\n x - 4\n\njulia> moi_set(constraint_object(cref))\nMathOptInterface.PositiveSemidefiniteConeSquare(2)\n\nWe see in the output of the last command that the vectorization of the matrix is constrained to belong to the PositiveSemidefiniteConeSquare.\n\njulia> using LinearAlgebra # For Symmetric\n\njulia> cref = @constraint(model, Symmetric(a - b) in PSDCone())\n[x - 1    2 x - 2;\n 2 x - 2  x - 4] ∈ PSDCone()\n\njulia> jump_function(constraint_object(cref))\n3-element Vector{AffExpr}:\n x - 1\n 2 x - 2\n x - 4\n\njulia> moi_set(constraint_object(cref))\nMathOptInterface.PositiveSemidefiniteConeTriangle(2)\n\nAs we see in the output of the last command, the vectorization of only the upper triangular part of the matrix is constrained to belong to the PositiveSemidefiniteConeSquare.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#Parameter","page":"JuMP","title":"Parameter","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Parameter","category":"page"},{"location":"api/JuMP/#JuMP.Parameter","page":"JuMP","title":"JuMP.Parameter","text":"Parameter(value)\n\nA short-cut for the MOI.Parameter set.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x in Parameter(2))\nx\n\njulia> print(model)\nFeasibility\nSubject to\n x ∈ MathOptInterface.Parameter{Float64}(2.0)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ParameterRef","page":"JuMP","title":"ParameterRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ParameterRef","category":"page"},{"location":"api/JuMP/#JuMP.ParameterRef","page":"JuMP","title":"JuMP.ParameterRef","text":"ParameterRef(x::GenericVariableRef)\n\nReturn a constraint reference to the constraint constraining x to be a parameter.\n\nErrors if one does not exist.\n\nSee also is_parameter, set_parameter_value, parameter_value.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, p in Parameter(2))\np\n\njulia> ParameterRef(p)\np ∈ MathOptInterface.Parameter{Float64}(2.0)\n\njulia> @variable(model, x);\n\njulia> ParameterRef(x)\nERROR: Variable x is not a parameter.\nStacktrace:\n[...]\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#QuadExpr","page":"JuMP","title":"QuadExpr","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.QuadExpr","category":"page"},{"location":"api/JuMP/#JuMP.QuadExpr","page":"JuMP","title":"JuMP.QuadExpr","text":"QuadExpr\n\nAn alias for GenericQuadExpr{Float64,VariableRef}, the specific GenericQuadExpr used by JuMP.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#QuadTermIterator","page":"JuMP","title":"QuadTermIterator","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.QuadTermIterator","category":"page"},{"location":"api/JuMP/#JuMP.QuadTermIterator","page":"JuMP","title":"JuMP.QuadTermIterator","text":"QuadTermIterator{GQE<:GenericQuadExpr}\n\nA struct that implements the iterate protocol in order to iterate over tuples of (coefficient, variable, variable) in the GenericQuadExpr.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ReferenceMap","page":"JuMP","title":"ReferenceMap","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ReferenceMap","category":"page"},{"location":"api/JuMP/#JuMP.ReferenceMap","page":"JuMP","title":"JuMP.ReferenceMap","text":"GenericReferenceMap{T}\n\nMapping between variable and constraint reference of a model and its copy. The reference of the copied model can be obtained by indexing the map with the reference of the corresponding reference of the original model.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ResultStatusCode","page":"JuMP","title":"ResultStatusCode","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ResultStatusCode","category":"page"},{"location":"api/JuMP/#JuMP.ResultStatusCode","page":"JuMP","title":"JuMP.ResultStatusCode","text":"ResultStatusCode\n\nAn Enum of possible values for the PrimalStatus and DualStatus attributes.\n\nThe values indicate how to interpret the result vector.\n\nValues\n\nPossible values are:\n\nNO_SOLUTION: the result vector is empty.\nFEASIBLE_POINT: the result vector is a feasible point.\nNEARLY_FEASIBLE_POINT: the result vector is feasible if some constraint tolerances are relaxed.\nINFEASIBLE_POINT: the result vector is an infeasible point.\nINFEASIBILITY_CERTIFICATE: the result vector is an infeasibility certificate. If the PrimalStatus is INFEASIBILITY_CERTIFICATE, then the primal result vector is a certificate of dual infeasibility. If the DualStatus is INFEASIBILITY_CERTIFICATE, then the dual result vector is a proof of primal infeasibility.\nNEARLY_INFEASIBILITY_CERTIFICATE: the result satisfies a relaxed criterion for a certificate of infeasibility.\nREDUCTION_CERTIFICATE: the result vector is an ill-posed certificate; see this article for details. If the PrimalStatus is REDUCTION_CERTIFICATE, then the primal result vector is a proof that the dual problem is ill-posed. If the DualStatus is REDUCTION_CERTIFICATE, then the dual result vector is a proof that the primal is ill-posed.\nNEARLY_REDUCTION_CERTIFICATE: the result satisfies a relaxed criterion for an ill-posed certificate.\nUNKNOWN_RESULT_STATUS: the result vector contains a solution with an unknown interpretation.\nOTHER_RESULT_STATUS: the result vector contains a solution with an interpretation not covered by one of the statuses defined above\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#RotatedSecondOrderCone","page":"JuMP","title":"RotatedSecondOrderCone","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.RotatedSecondOrderCone","category":"page"},{"location":"api/JuMP/#JuMP.RotatedSecondOrderCone","page":"JuMP","title":"JuMP.RotatedSecondOrderCone","text":"RotatedSecondOrderCone\n\nRotated second order cone object that can be used to constrain the square of the euclidean norm of a vector x to be less than or equal to 2tu where t and u are nonnegative scalars. This is a shortcut for the MOI.RotatedSecondOrderCone.\n\nExample\n\nThe following constrains (x-1 x-2)^2_2 le 2tx and t x ge 0:\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, t)\nt\n\njulia> @constraint(model, [t, x, x-1, x-2] in RotatedSecondOrderCone())\n[t, x, x - 1, x - 2] ∈ MathOptInterface.RotatedSecondOrderCone(4)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SOS1","page":"JuMP","title":"SOS1","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SOS1","category":"page"},{"location":"api/JuMP/#JuMP.SOS1","page":"JuMP","title":"JuMP.SOS1","text":"SOS1\n\nSOS1 (Special Ordered Sets type 1) object than can be used to constrain a vector x to a set where at most 1 variable can take a non-zero value, all others being at 0. The weights, when specified, induce an ordering of the variables; as such, they should be unique values. The kth element in the set corresponds to the kth weight in weights. See here for a description of SOS constraints and their potential uses. This is a shortcut for the MathOptInterface.SOS1 set.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SOS2","page":"JuMP","title":"SOS2","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SOS2","category":"page"},{"location":"api/JuMP/#JuMP.SOS2","page":"JuMP","title":"JuMP.SOS2","text":"SOS2\n\nSOS1 (Special Ordered Sets type 2) object than can be used to constrain a vector x to a set where at most 2 variables can take a non-zero value, all others being at 0. In addition, if two are non-zero these must be consecutive in their ordering. The weights induce an ordering of the variables; as such, they should be unique values. The kth element in the set corresponds to the kth weight in weights. See here for a description of SOS constraints and their potential uses. This is a shortcut for the MathOptInterface.SOS2 set.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ScalarConstraint","page":"JuMP","title":"ScalarConstraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ScalarConstraint","category":"page"},{"location":"api/JuMP/#JuMP.ScalarConstraint","page":"JuMP","title":"JuMP.ScalarConstraint","text":"struct ScalarConstraint\n\nThe data for a scalar constraint. The func field contains a JuMP object representing the function and the set field contains the MOI set. See also the documentation on JuMP's representation of constraints for more background.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ScalarShape","page":"JuMP","title":"ScalarShape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ScalarShape","category":"page"},{"location":"api/JuMP/#JuMP.ScalarShape","page":"JuMP","title":"JuMP.ScalarShape","text":"ScalarShape\n\nShape of scalar constraints.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ScalarVariable","page":"JuMP","title":"ScalarVariable","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ScalarVariable","category":"page"},{"location":"api/JuMP/#JuMP.ScalarVariable","page":"JuMP","title":"JuMP.ScalarVariable","text":"ScalarVariable{S,T,U,V} <: AbstractVariable\n\nA struct used when adding variables.\n\nSee also: add_variable.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SecondOrderCone","page":"JuMP","title":"SecondOrderCone","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SecondOrderCone","category":"page"},{"location":"api/JuMP/#JuMP.SecondOrderCone","page":"JuMP","title":"JuMP.SecondOrderCone","text":"SecondOrderCone\n\nSecond order cone object that can be used to constrain the euclidean norm of a vector x to be less than or equal to a nonnegative scalar t. This is a shortcut for the MOI.SecondOrderCone.\n\nExample\n\nThe following constrains (x-1 x-2)_2 le t and t ge 0:\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, t)\nt\n\njulia> @constraint(model, [t, x-1, x-2] in SecondOrderCone())\n[t, x - 1, x - 2] ∈ MathOptInterface.SecondOrderCone(3)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#Semicontinuous","page":"JuMP","title":"Semicontinuous","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Semicontinuous","category":"page"},{"location":"api/JuMP/#JuMP.Semicontinuous","page":"JuMP","title":"JuMP.Semicontinuous","text":"Semicontinuous(lower, upper)\n\nA short-cut for the MOI.Semicontinuous set.\n\nThis short-cut is useful because it automatically promotes lower and upper to the same type, and converts them into the element type supported by the JuMP model.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x in Semicontinuous(1, 2))\nx\n\njulia> print(model)\nFeasibility\nSubject to\n x ∈ MathOptInterface.Semicontinuous{Int64}(1, 2)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#Semiinteger","page":"JuMP","title":"Semiinteger","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Semiinteger","category":"page"},{"location":"api/JuMP/#JuMP.Semiinteger","page":"JuMP","title":"JuMP.Semiinteger","text":"Semiinteger(lower, upper)\n\nA short-cut for the MOI.Semiinteger set.\n\nThis short-cut is useful because it automatically promotes lower and upper to the same type, and converts them into the element type supported by the JuMP model.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x in Semiinteger(3, 5))\nx\n\njulia> print(model)\nFeasibility\nSubject to\n x ∈ MathOptInterface.Semiinteger{Int64}(3, 5)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SensitivityReport","page":"JuMP","title":"SensitivityReport","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SensitivityReport","category":"page"},{"location":"api/JuMP/#JuMP.SensitivityReport","page":"JuMP","title":"JuMP.SensitivityReport","text":"SensitivityReport\n\nSee lp_sensitivity_report.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SkewSymmetricMatrixShape","page":"JuMP","title":"SkewSymmetricMatrixShape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SkewSymmetricMatrixShape","category":"page"},{"location":"api/JuMP/#JuMP.SkewSymmetricMatrixShape","page":"JuMP","title":"JuMP.SkewSymmetricMatrixShape","text":"SkewSymmetricMatrixShape\n\nShape object for a skew symmetric square matrix of side_dimension rows and columns. The vectorized form contains the entries of the upper-right triangular part of the matrix (without the diagonal) given column by column (or equivalently, the entries of the lower-left triangular part given row by row). The diagonal is zero.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SkewSymmetricMatrixSpace","page":"JuMP","title":"SkewSymmetricMatrixSpace","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SkewSymmetricMatrixSpace","category":"page"},{"location":"api/JuMP/#JuMP.SkewSymmetricMatrixSpace","page":"JuMP","title":"JuMP.SkewSymmetricMatrixSpace","text":"SkewSymmetricMatrixSpace()\n\nUse in the @variable macro to constrain a matrix of variables to be skew-symmetric.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2] in SkewSymmetricMatrixSpace())\n2×2 Matrix{AffExpr}:\n 0        Q[1,2]\n -Q[1,2]  0\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SquareMatrixShape","page":"JuMP","title":"SquareMatrixShape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SquareMatrixShape","category":"page"},{"location":"api/JuMP/#JuMP.SquareMatrixShape","page":"JuMP","title":"JuMP.SquareMatrixShape","text":"SquareMatrixShape\n\nShape object for a square matrix of side_dimension rows and columns. The vectorized form contains the entries of the the matrix given column by column (or equivalently, the entries of the lower-left triangular part given row by row).\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SymmetricMatrixShape","page":"JuMP","title":"SymmetricMatrixShape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SymmetricMatrixShape","category":"page"},{"location":"api/JuMP/#JuMP.SymmetricMatrixShape","page":"JuMP","title":"JuMP.SymmetricMatrixShape","text":"SymmetricMatrixShape\n\nShape object for a symmetric square matrix of side_dimension rows and columns. The vectorized form contains the entries of the upper-right triangular part of the matrix given column by column (or equivalently, the entries of the lower-left triangular part given row by row).\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#SymmetricMatrixSpace","page":"JuMP","title":"SymmetricMatrixSpace","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SymmetricMatrixSpace","category":"page"},{"location":"api/JuMP/#JuMP.SymmetricMatrixSpace","page":"JuMP","title":"JuMP.SymmetricMatrixSpace","text":"SymmetricMatrixSpace()\n\nUse in the @variable macro to constrain a matrix of variables to be symmetric.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, Q[1:2, 1:2] in SymmetricMatrixSpace())\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n Q[1,1]  Q[1,2]\n Q[1,2]  Q[2,2]\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#TerminationStatusCode","page":"JuMP","title":"TerminationStatusCode","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.TerminationStatusCode","category":"page"},{"location":"api/JuMP/#JuMP.TerminationStatusCode","page":"JuMP","title":"JuMP.TerminationStatusCode","text":"TerminationStatusCode\n\nAn Enum of possible values for the TerminationStatus attribute. This attribute is meant to explain the reason why the optimizer stopped executing in the most recent call to optimize!.\n\nValues\n\nPossible values are:\n\nOPTIMIZE_NOT_CALLED: The algorithm has not started.\nOPTIMAL: The algorithm found a globally optimal solution.\nINFEASIBLE: The algorithm concluded that no feasible solution exists.\nDUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.\nLOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.\nLOCALLY_INFEASIBLE: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.\nINFEASIBLE_OR_UNBOUNDED: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.\nALMOST_OPTIMAL: The algorithm found a globally optimal solution to relaxed tolerances.\nALMOST_INFEASIBLE: The algorithm concluded that no feasible solution exists within relaxed tolerances.\nALMOST_DUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.\nALMOST_LOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.\nITERATION_LIMIT: An iterative algorithm stopped after conducting the maximum number of iterations.\nTIME_LIMIT: The algorithm stopped after a user-specified computation time.\nNODE_LIMIT: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.\nSOLUTION_LIMIT: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.\nMEMORY_LIMIT: The algorithm stopped because it ran out of memory.\nOBJECTIVE_LIMIT: The algorithm stopped because it found a solution better than a minimum limit set by the user.\nNORM_LIMIT: The algorithm stopped because the norm of an iterate became too large.\nOTHER_LIMIT: The algorithm stopped due to a limit not covered by one of the _LIMIT_ statuses above.\nSLOW_PROGRESS: The algorithm stopped because it was unable to continue making progress towards the solution.\nNUMERICAL_ERROR: The algorithm stopped because it encountered unrecoverable numerical error.\nINVALID_MODEL: The algorithm stopped because the model is invalid.\nINVALID_OPTION: The algorithm stopped because it was provided an invalid option.\nINTERRUPTED: The algorithm stopped because of an interrupt signal.\nOTHER_ERROR: The algorithm stopped because of an error not covered by one of the statuses defined above.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#UnorderedPair","page":"JuMP","title":"UnorderedPair","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.UnorderedPair","category":"page"},{"location":"api/JuMP/#JuMP.UnorderedPair","page":"JuMP","title":"JuMP.UnorderedPair","text":"UnorderedPair(a::T, b::T)\n\nA wrapper type used by GenericQuadExpr with fields .a and .b.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#UpperBoundRef","page":"JuMP","title":"UpperBoundRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.UpperBoundRef","category":"page"},{"location":"api/JuMP/#JuMP.UpperBoundRef","page":"JuMP","title":"JuMP.UpperBoundRef","text":"UpperBoundRef(v::GenericVariableRef)\n\nReturn a constraint reference to the upper bound constraint of v.\n\nErrors if one does not exist.\n\nSee also has_upper_bound, upper_bound, set_upper_bound, delete_upper_bound.\n\nExamples\n\njulia> model = Model();\n\njulia> @variable(model, x <= 1.0);\n\njulia> UpperBoundRef(x)\nx ≤ 1\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP/#VariableConstrainedOnCreation","page":"JuMP","title":"VariableConstrainedOnCreation","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.VariableConstrainedOnCreation","category":"page"},{"location":"api/JuMP/#JuMP.VariableConstrainedOnCreation","page":"JuMP","title":"JuMP.VariableConstrainedOnCreation","text":"VariableConstrainedOnCreation <: AbstractVariable\n\nVariable scalar_variables constrained to belong to set.\n\nAdding this variable can be understood as doing:\n\nfunction JuMP.add_variable(\n    model::GenericModel,\n    variable::VariableConstrainedOnCreation,\n    names,\n)\n    var_ref = add_variable(model, variable.scalar_variable, name)\n    add_constraint(model, VectorConstraint(var_ref, variable.set))\n    return var_ref\nend\n\nbut adds the variables with MOI.add_constrained_variable(model, variable.set) instead. See the MOI documentation for the difference between adding the variables with MOI.add_constrained_variable and adding them with MOI.add_variable and adding the constraint separately.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#VariableInfo","page":"JuMP","title":"VariableInfo","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.VariableInfo","category":"page"},{"location":"api/JuMP/#JuMP.VariableInfo","page":"JuMP","title":"JuMP.VariableInfo","text":"VariableInfo{S,T,U,V}\n\nA struct by JuMP internally when creating variables. This may also be used by JuMP extensions to create new types of variables.\n\nSee also: ScalarVariable.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#VariableNotOwned","page":"JuMP","title":"VariableNotOwned","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.VariableNotOwned","category":"page"},{"location":"api/JuMP/#JuMP.VariableNotOwned","page":"JuMP","title":"JuMP.VariableNotOwned","text":"struct VariableNotOwned{V<:AbstractVariableRef} <: Exception\n    variable::V\nend\n\nThe variable variable was used in a model different to owner_model(variable).\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#VariableRef","page":"JuMP","title":"VariableRef","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.VariableRef","category":"page"},{"location":"api/JuMP/#JuMP.VariableRef","page":"JuMP","title":"JuMP.VariableRef","text":"GenericVariableRef{T} <: AbstractVariableRef\n\nHolds a reference to the model and the corresponding MOI.VariableIndex.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#VariablesConstrainedOnCreation","page":"JuMP","title":"VariablesConstrainedOnCreation","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.VariablesConstrainedOnCreation","category":"page"},{"location":"api/JuMP/#JuMP.VariablesConstrainedOnCreation","page":"JuMP","title":"JuMP.VariablesConstrainedOnCreation","text":"VariablesConstrainedOnCreation <: AbstractVariable\n\nVector of variables scalar_variables constrained to belong to set. Adding this variable can be thought as doing:\n\nfunction JuMP.add_variable(\n    model::GenericModel,\n    variable::VariablesConstrainedOnCreation,\n    names,\n)\n    v_names = vectorize(names, variable.shape)\n    var_refs = add_variable.(model, variable.scalar_variables, v_names)\n    add_constraint(model, VectorConstraint(var_refs, variable.set))\n    return reshape_vector(var_refs, variable.shape)\nend\n\nbut adds the variables with MOI.add_constrained_variables(model, variable.set) instead. See the MOI documentation for the difference between adding the variables with MOI.add_constrained_variables and adding them with MOI.add_variables and adding the constraint separately.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#VectorConstraint","page":"JuMP","title":"VectorConstraint","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.VectorConstraint","category":"page"},{"location":"api/JuMP/#JuMP.VectorConstraint","page":"JuMP","title":"JuMP.VectorConstraint","text":"struct VectorConstraint\n\nThe data for a vector constraint. The func field contains a JuMP object representing the function and the set field contains the MOI set. The shape field contains an AbstractShape matching the form in which the constraint was constructed (e.g., by using matrices or flat vectors). See also the documentation on JuMP's representation of constraints.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#VectorShape","page":"JuMP","title":"VectorShape","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.VectorShape","category":"page"},{"location":"api/JuMP/#JuMP.VectorShape","page":"JuMP","title":"JuMP.VectorShape","text":"VectorShape\n\nVector for which the vectorized form corresponds exactly to the vector given.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#Zeros","page":"JuMP","title":"Zeros","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Zeros","category":"page"},{"location":"api/JuMP/#JuMP.Zeros","page":"JuMP","title":"JuMP.Zeros","text":"Zeros()\n\nThe JuMP equivalent of the MOI.Zeros set, in which the dimension is inferred from the corresponding function.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> @constraint(model, x in Zeros())\n[x[1], x[2]] ∈ MathOptInterface.Zeros(2)\n\njulia> A = [1 2; 3 4];\n\njulia> b = [5, 6];\n\njulia> @constraint(model, A * x == b)\n[x[1] + 2 x[2] - 5, 3 x[1] + 4 x[2] - 6] ∈ MathOptInterface.Zeros(2)\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP/#ALMOST_DUAL_INFEASIBLE","page":"JuMP","title":"ALMOST_DUAL_INFEASIBLE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ALMOST_DUAL_INFEASIBLE","category":"page"},{"location":"api/JuMP/#JuMP.ALMOST_DUAL_INFEASIBLE","page":"JuMP","title":"JuMP.ALMOST_DUAL_INFEASIBLE","text":"ALMOST_DUAL_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_DUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem within relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#ALMOST_INFEASIBLE","page":"JuMP","title":"ALMOST_INFEASIBLE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ALMOST_INFEASIBLE","category":"page"},{"location":"api/JuMP/#JuMP.ALMOST_INFEASIBLE","page":"JuMP","title":"JuMP.ALMOST_INFEASIBLE","text":"ALMOST_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_INFEASIBLE: The algorithm concluded that no feasible solution exists within relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#ALMOST_LOCALLY_SOLVED","page":"JuMP","title":"ALMOST_LOCALLY_SOLVED","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ALMOST_LOCALLY_SOLVED","category":"page"},{"location":"api/JuMP/#JuMP.ALMOST_LOCALLY_SOLVED","page":"JuMP","title":"JuMP.ALMOST_LOCALLY_SOLVED","text":"ALMOST_LOCALLY_SOLVED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_LOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, or could not find directions for improvement within relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#ALMOST_OPTIMAL","page":"JuMP","title":"ALMOST_OPTIMAL","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ALMOST_OPTIMAL","category":"page"},{"location":"api/JuMP/#JuMP.ALMOST_OPTIMAL","page":"JuMP","title":"JuMP.ALMOST_OPTIMAL","text":"ALMOST_OPTIMAL::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nALMOST_OPTIMAL: The algorithm found a globally optimal solution to relaxed tolerances.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#AUTOMATIC","page":"JuMP","title":"AUTOMATIC","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.AUTOMATIC","category":"page"},{"location":"api/JuMP/#JuMP.AUTOMATIC","page":"JuMP","title":"JuMP.AUTOMATIC","text":"moi_backend field holds a CachingOptimizer in AUTOMATIC mode.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#DIRECT","page":"JuMP","title":"DIRECT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.DIRECT","category":"page"},{"location":"api/JuMP/#JuMP.DIRECT","page":"JuMP","title":"JuMP.DIRECT","text":"moi_backend field holds an AbstractOptimizer. No extra copy of the model is stored. The moi_backend must support add_constraint etc.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#DUAL_INFEASIBLE","page":"JuMP","title":"DUAL_INFEASIBLE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.DUAL_INFEASIBLE","category":"page"},{"location":"api/JuMP/#JuMP.DUAL_INFEASIBLE","page":"JuMP","title":"JuMP.DUAL_INFEASIBLE","text":"DUAL_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nDUAL_INFEASIBLE: The algorithm concluded that no dual bound exists for the problem. If, additionally, a feasible (primal) solution is known to exist, this status typically implies that the problem is unbounded, with some technical exceptions.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#FEASIBILITY_SENSE","page":"JuMP","title":"FEASIBILITY_SENSE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.FEASIBILITY_SENSE","category":"page"},{"location":"api/JuMP/#JuMP.FEASIBILITY_SENSE","page":"JuMP","title":"JuMP.FEASIBILITY_SENSE","text":"FEASIBILITY_SENSE::OptimizationSense\n\nAn instance of the OptimizationSense enum.\n\nFEASIBILITY_SENSE: the model does not have an objective function\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#FEASIBLE_POINT","page":"JuMP","title":"FEASIBLE_POINT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.FEASIBLE_POINT","category":"page"},{"location":"api/JuMP/#JuMP.FEASIBLE_POINT","page":"JuMP","title":"JuMP.FEASIBLE_POINT","text":"FEASIBLE_POINT::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nFEASIBLE_POINT: the result vector is a feasible point.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#INFEASIBILITY_CERTIFICATE","page":"JuMP","title":"INFEASIBILITY_CERTIFICATE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.INFEASIBILITY_CERTIFICATE","category":"page"},{"location":"api/JuMP/#JuMP.INFEASIBILITY_CERTIFICATE","page":"JuMP","title":"JuMP.INFEASIBILITY_CERTIFICATE","text":"INFEASIBILITY_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nINFEASIBILITY_CERTIFICATE: the result vector is an infeasibility certificate. If the PrimalStatus is INFEASIBILITY_CERTIFICATE, then the primal result vector is a certificate of dual infeasibility. If the DualStatus is INFEASIBILITY_CERTIFICATE, then the dual result vector is a proof of primal infeasibility.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#INFEASIBLE","page":"JuMP","title":"INFEASIBLE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.INFEASIBLE","category":"page"},{"location":"api/JuMP/#JuMP.INFEASIBLE","page":"JuMP","title":"JuMP.INFEASIBLE","text":"INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINFEASIBLE: The algorithm concluded that no feasible solution exists.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#INFEASIBLE_OR_UNBOUNDED","page":"JuMP","title":"INFEASIBLE_OR_UNBOUNDED","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.INFEASIBLE_OR_UNBOUNDED","category":"page"},{"location":"api/JuMP/#JuMP.INFEASIBLE_OR_UNBOUNDED","page":"JuMP","title":"JuMP.INFEASIBLE_OR_UNBOUNDED","text":"INFEASIBLE_OR_UNBOUNDED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINFEASIBLE_OR_UNBOUNDED: The algorithm stopped because it decided that the problem is infeasible or unbounded; this occasionally happens during MIP presolve.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#INFEASIBLE_POINT","page":"JuMP","title":"INFEASIBLE_POINT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.INFEASIBLE_POINT","category":"page"},{"location":"api/JuMP/#JuMP.INFEASIBLE_POINT","page":"JuMP","title":"JuMP.INFEASIBLE_POINT","text":"INFEASIBLE_POINT::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nINFEASIBLE_POINT: the result vector is an infeasible point.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#INTERRUPTED","page":"JuMP","title":"INTERRUPTED","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.INTERRUPTED","category":"page"},{"location":"api/JuMP/#JuMP.INTERRUPTED","page":"JuMP","title":"JuMP.INTERRUPTED","text":"INTERRUPTED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINTERRUPTED: The algorithm stopped because of an interrupt signal.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#INVALID_MODEL","page":"JuMP","title":"INVALID_MODEL","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.INVALID_MODEL","category":"page"},{"location":"api/JuMP/#JuMP.INVALID_MODEL","page":"JuMP","title":"JuMP.INVALID_MODEL","text":"INVALID_MODEL::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINVALID_MODEL: The algorithm stopped because the model is invalid.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#INVALID_OPTION","page":"JuMP","title":"INVALID_OPTION","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.INVALID_OPTION","category":"page"},{"location":"api/JuMP/#JuMP.INVALID_OPTION","page":"JuMP","title":"JuMP.INVALID_OPTION","text":"INVALID_OPTION::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nINVALID_OPTION: The algorithm stopped because it was provided an invalid option.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#ITERATION_LIMIT","page":"JuMP","title":"ITERATION_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.ITERATION_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.ITERATION_LIMIT","page":"JuMP","title":"JuMP.ITERATION_LIMIT","text":"ITERATION_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nITERATION_LIMIT: An iterative algorithm stopped after conducting the maximum number of iterations.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#LOCALLY_INFEASIBLE","page":"JuMP","title":"LOCALLY_INFEASIBLE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.LOCALLY_INFEASIBLE","category":"page"},{"location":"api/JuMP/#JuMP.LOCALLY_INFEASIBLE","page":"JuMP","title":"JuMP.LOCALLY_INFEASIBLE","text":"LOCALLY_INFEASIBLE::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nLOCALLY_INFEASIBLE: The algorithm converged to an infeasible point or otherwise completed its search without finding a feasible solution, without guarantees that no feasible solution exists.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#LOCALLY_SOLVED","page":"JuMP","title":"LOCALLY_SOLVED","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.LOCALLY_SOLVED","category":"page"},{"location":"api/JuMP/#JuMP.LOCALLY_SOLVED","page":"JuMP","title":"JuMP.LOCALLY_SOLVED","text":"LOCALLY_SOLVED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nLOCALLY_SOLVED: The algorithm converged to a stationary point, local optimal solution, could not find directions for improvement, or otherwise completed its search without global guarantees.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#MANUAL","page":"JuMP","title":"MANUAL","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.MANUAL","category":"page"},{"location":"api/JuMP/#JuMP.MANUAL","page":"JuMP","title":"JuMP.MANUAL","text":"moi_backend field holds a CachingOptimizer in MANUAL mode.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#MAX_SENSE","page":"JuMP","title":"MAX_SENSE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.MAX_SENSE","category":"page"},{"location":"api/JuMP/#JuMP.MAX_SENSE","page":"JuMP","title":"JuMP.MAX_SENSE","text":"MAX_SENSE::OptimizationSense\n\nAn instance of the OptimizationSense enum.\n\nMAX_SENSE: the goal is to maximize the objective function\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#MEMORY_LIMIT","page":"JuMP","title":"MEMORY_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.MEMORY_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.MEMORY_LIMIT","page":"JuMP","title":"JuMP.MEMORY_LIMIT","text":"MEMORY_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nMEMORY_LIMIT: The algorithm stopped because it ran out of memory.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#MIN_SENSE","page":"JuMP","title":"MIN_SENSE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.MIN_SENSE","category":"page"},{"location":"api/JuMP/#JuMP.MIN_SENSE","page":"JuMP","title":"JuMP.MIN_SENSE","text":"MIN_SENSE::OptimizationSense\n\nAn instance of the OptimizationSense enum.\n\nMIN_SENSE: the goal is to minimize the objective function\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#NEARLY_FEASIBLE_POINT","page":"JuMP","title":"NEARLY_FEASIBLE_POINT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NEARLY_FEASIBLE_POINT","category":"page"},{"location":"api/JuMP/#JuMP.NEARLY_FEASIBLE_POINT","page":"JuMP","title":"JuMP.NEARLY_FEASIBLE_POINT","text":"NEARLY_FEASIBLE_POINT::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNEARLY_FEASIBLE_POINT: the result vector is feasible if some constraint tolerances are relaxed.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#NEARLY_INFEASIBILITY_CERTIFICATE","page":"JuMP","title":"NEARLY_INFEASIBILITY_CERTIFICATE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NEARLY_INFEASIBILITY_CERTIFICATE","category":"page"},{"location":"api/JuMP/#JuMP.NEARLY_INFEASIBILITY_CERTIFICATE","page":"JuMP","title":"JuMP.NEARLY_INFEASIBILITY_CERTIFICATE","text":"NEARLY_INFEASIBILITY_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNEARLY_INFEASIBILITY_CERTIFICATE: the result satisfies a relaxed criterion for a certificate of infeasibility.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#NEARLY_REDUCTION_CERTIFICATE","page":"JuMP","title":"NEARLY_REDUCTION_CERTIFICATE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NEARLY_REDUCTION_CERTIFICATE","category":"page"},{"location":"api/JuMP/#JuMP.NEARLY_REDUCTION_CERTIFICATE","page":"JuMP","title":"JuMP.NEARLY_REDUCTION_CERTIFICATE","text":"NEARLY_REDUCTION_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNEARLY_REDUCTION_CERTIFICATE: the result satisfies a relaxed criterion for an ill-posed certificate.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#NODE_LIMIT","page":"JuMP","title":"NODE_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NODE_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.NODE_LIMIT","page":"JuMP","title":"JuMP.NODE_LIMIT","text":"NODE_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nNODE_LIMIT: A branch-and-bound algorithm stopped because it explored a maximum number of nodes in the branch-and-bound tree.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#NORM_LIMIT","page":"JuMP","title":"NORM_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NORM_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.NORM_LIMIT","page":"JuMP","title":"JuMP.NORM_LIMIT","text":"NORM_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nNORM_LIMIT: The algorithm stopped because the norm of an iterate became too large.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#NO_SOLUTION","page":"JuMP","title":"NO_SOLUTION","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NO_SOLUTION","category":"page"},{"location":"api/JuMP/#JuMP.NO_SOLUTION","page":"JuMP","title":"JuMP.NO_SOLUTION","text":"NO_SOLUTION::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nNO_SOLUTION: the result vector is empty.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#NUMERICAL_ERROR","page":"JuMP","title":"NUMERICAL_ERROR","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.NUMERICAL_ERROR","category":"page"},{"location":"api/JuMP/#JuMP.NUMERICAL_ERROR","page":"JuMP","title":"JuMP.NUMERICAL_ERROR","text":"NUMERICAL_ERROR::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nNUMERICAL_ERROR: The algorithm stopped because it encountered unrecoverable numerical error.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#OBJECTIVE_LIMIT","page":"JuMP","title":"OBJECTIVE_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OBJECTIVE_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.OBJECTIVE_LIMIT","page":"JuMP","title":"JuMP.OBJECTIVE_LIMIT","text":"OBJECTIVE_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOBJECTIVE_LIMIT: The algorithm stopped because it found a solution better than a minimum limit set by the user.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#OPTIMAL","page":"JuMP","title":"OPTIMAL","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OPTIMAL","category":"page"},{"location":"api/JuMP/#JuMP.OPTIMAL","page":"JuMP","title":"JuMP.OPTIMAL","text":"OPTIMAL::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOPTIMAL: The algorithm found a globally optimal solution.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#OPTIMIZE_NOT_CALLED","page":"JuMP","title":"OPTIMIZE_NOT_CALLED","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OPTIMIZE_NOT_CALLED","category":"page"},{"location":"api/JuMP/#JuMP.OPTIMIZE_NOT_CALLED","page":"JuMP","title":"JuMP.OPTIMIZE_NOT_CALLED","text":"OPTIMIZE_NOT_CALLED::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOPTIMIZE_NOT_CALLED: The algorithm has not started.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#OTHER_ERROR","page":"JuMP","title":"OTHER_ERROR","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OTHER_ERROR","category":"page"},{"location":"api/JuMP/#JuMP.OTHER_ERROR","page":"JuMP","title":"JuMP.OTHER_ERROR","text":"OTHER_ERROR::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOTHER_ERROR: The algorithm stopped because of an error not covered by one of the statuses defined above.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#OTHER_LIMIT","page":"JuMP","title":"OTHER_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OTHER_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.OTHER_LIMIT","page":"JuMP","title":"JuMP.OTHER_LIMIT","text":"OTHER_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nOTHER_LIMIT: The algorithm stopped due to a limit not covered by one of the _LIMIT_ statuses above.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#OTHER_RESULT_STATUS","page":"JuMP","title":"OTHER_RESULT_STATUS","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.OTHER_RESULT_STATUS","category":"page"},{"location":"api/JuMP/#JuMP.OTHER_RESULT_STATUS","page":"JuMP","title":"JuMP.OTHER_RESULT_STATUS","text":"OTHER_RESULT_STATUS::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nOTHER_RESULT_STATUS: the result vector contains a solution with an interpretation not covered by one of the statuses defined above\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#REDUCTION_CERTIFICATE","page":"JuMP","title":"REDUCTION_CERTIFICATE","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.REDUCTION_CERTIFICATE","category":"page"},{"location":"api/JuMP/#JuMP.REDUCTION_CERTIFICATE","page":"JuMP","title":"JuMP.REDUCTION_CERTIFICATE","text":"REDUCTION_CERTIFICATE::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nREDUCTION_CERTIFICATE: the result vector is an ill-posed certificate; see this article for details. If the PrimalStatus is REDUCTION_CERTIFICATE, then the primal result vector is a proof that the dual problem is ill-posed. If the DualStatus is REDUCTION_CERTIFICATE, then the dual result vector is a proof that the primal is ill-posed.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#SLOW_PROGRESS","page":"JuMP","title":"SLOW_PROGRESS","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SLOW_PROGRESS","category":"page"},{"location":"api/JuMP/#JuMP.SLOW_PROGRESS","page":"JuMP","title":"JuMP.SLOW_PROGRESS","text":"SLOW_PROGRESS::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nSLOW_PROGRESS: The algorithm stopped because it was unable to continue making progress towards the solution.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#SOLUTION_LIMIT","page":"JuMP","title":"SOLUTION_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.SOLUTION_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.SOLUTION_LIMIT","page":"JuMP","title":"JuMP.SOLUTION_LIMIT","text":"SOLUTION_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nSOLUTION_LIMIT: The algorithm stopped because it found the required number of solutions. This is often used in MIPs to get the solver to return the first feasible solution it encounters.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#TIME_LIMIT","page":"JuMP","title":"TIME_LIMIT","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.TIME_LIMIT","category":"page"},{"location":"api/JuMP/#JuMP.TIME_LIMIT","page":"JuMP","title":"JuMP.TIME_LIMIT","text":"TIME_LIMIT::TerminationStatusCode\n\nAn instance of the TerminationStatusCode enum.\n\nTIME_LIMIT: The algorithm stopped after a user-specified computation time.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#UNKNOWN_RESULT_STATUS","page":"JuMP","title":"UNKNOWN_RESULT_STATUS","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.UNKNOWN_RESULT_STATUS","category":"page"},{"location":"api/JuMP/#JuMP.UNKNOWN_RESULT_STATUS","page":"JuMP","title":"JuMP.UNKNOWN_RESULT_STATUS","text":"UNKNOWN_RESULT_STATUS::ResultStatusCode\n\nAn instance of the ResultStatusCode enum.\n\nUNKNOWN_RESULT_STATUS: the result vector contains a solution with an unknown interpretation.\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#op_and","page":"JuMP","title":"op_and","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_and","category":"page"},{"location":"api/JuMP/#JuMP.op_and","page":"JuMP","title":"JuMP.op_and","text":"op_and(x, y)\n\nA function that falls back to x & y, but when called with JuMP variables or expressions, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_and(true, false)\nfalse\n\njulia> op_and(true, x)\ntrue && x\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#op_equal_to","page":"JuMP","title":"op_equal_to","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_equal_to","category":"page"},{"location":"api/JuMP/#JuMP.op_equal_to","page":"JuMP","title":"JuMP.op_equal_to","text":"op_equal_to(x, y)\n\nA function that falls back to x == y, but when called with JuMP variables or expressions, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_equal_to(2, 2)\ntrue\n\njulia> op_equal_to(x, 2)\nx == 2\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#op_greater_than_or_equal_to","page":"JuMP","title":"op_greater_than_or_equal_to","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_greater_than_or_equal_to","category":"page"},{"location":"api/JuMP/#JuMP.op_greater_than_or_equal_to","page":"JuMP","title":"JuMP.op_greater_than_or_equal_to","text":"op_greater_than_or_equal_to(x, y)\n\nA function that falls back to x >= y, but when called with JuMP variables or expressions, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_greater_than_or_equal_to(2, 2)\ntrue\n\njulia> op_greater_than_or_equal_to(x, 2)\nx >= 2\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#op_less_than_or_equal_to","page":"JuMP","title":"op_less_than_or_equal_to","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_less_than_or_equal_to","category":"page"},{"location":"api/JuMP/#JuMP.op_less_than_or_equal_to","page":"JuMP","title":"JuMP.op_less_than_or_equal_to","text":"op_less_than_or_equal_to(x, y)\n\nA function that falls back to x <= y, but when called with JuMP variables or expressions, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_less_than_or_equal_to(2, 2)\ntrue\n\njulia> op_less_than_or_equal_to(x, 2)\nx <= 2\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#op_or","page":"JuMP","title":"op_or","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_or","category":"page"},{"location":"api/JuMP/#JuMP.op_or","page":"JuMP","title":"JuMP.op_or","text":"op_or(x, y)\n\nA function that falls back to x | y, but when called with JuMP variables or expressions, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_or(true, false)\ntrue\n\njulia> op_or(true, x)\ntrue || x\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#op_strictly_greater_than","page":"JuMP","title":"op_strictly_greater_than","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_strictly_greater_than","category":"page"},{"location":"api/JuMP/#JuMP.op_strictly_greater_than","page":"JuMP","title":"JuMP.op_strictly_greater_than","text":"op_strictly_greater_than(x, y)\n\nA function that falls back to x > y, but when called with JuMP variables or expressions, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_strictly_greater_than(1, 2)\nfalse\n\njulia> op_strictly_greater_than(x, 2)\nx > 2\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#op_strictly_less_than","page":"JuMP","title":"op_strictly_less_than","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.op_strictly_less_than","category":"page"},{"location":"api/JuMP/#JuMP.op_strictly_less_than","page":"JuMP","title":"JuMP.op_strictly_less_than","text":"op_strictly_less_than(x, y)\n\nA function that falls back to x < y, but when called with JuMP variables or expressions, returns a GenericNonlinearExpr.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x);\n\njulia> op_strictly_less_than(1, 2)\ntrue\n\njulia> op_strictly_less_than(x, 2)\nx < 2\n\n\n\n\n\n","category":"constant"},{"location":"api/JuMP/#Base.empty!(::GenericModel)","page":"JuMP","title":"Base.empty!(::GenericModel)","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Base.empty!(::GenericModel)","category":"page"},{"location":"api/JuMP/#Base.empty!-Tuple{GenericModel}","page":"JuMP","title":"Base.empty!","text":"empty!(model::GenericModel)::GenericModel\n\nEmpty the model, that is, remove all variables, constraints and model attributes but not optimizer attributes. Always return the argument.\n\nNote: removes extensions data.\n\n\n\n\n\n","category":"method"},{"location":"api/JuMP/#Base.isempty(::GenericModel)","page":"JuMP","title":"Base.isempty(::GenericModel)","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Base.isempty(::GenericModel)","category":"page"},{"location":"api/JuMP/#Base.isempty-Tuple{GenericModel}","page":"JuMP","title":"Base.isempty","text":"isempty(model::GenericModel)\n\nVerifies whether the model is empty, that is, whether the MOI backend is empty and whether the model is in the same state as at its creation apart from optimizer attributes.\n\n\n\n\n\n","category":"method"},{"location":"api/JuMP/#Base.copy(::AbstractModel)","page":"JuMP","title":"Base.copy(::AbstractModel)","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Base.copy(::AbstractModel)","category":"page"},{"location":"api/JuMP/#Base.copy-Tuple{AbstractModel}","page":"JuMP","title":"Base.copy","text":"copy(model::AbstractModel)\n\nReturn a copy of the model model. It is similar to copy_model except that it does not return the mapping between the references of model and its copy.\n\nNote\n\nModel copy is not supported in DIRECT mode, i.e. when a model is constructed using the direct_model constructor instead of the Model constructor. Moreover, independently on whether an optimizer was provided at model construction, the new model will have no optimizer, i.e., an optimizer will have to be provided to the new model in the optimize! call.\n\nExample\n\nIn the following example, a model model is constructed with a variable x and a constraint cref. It is then copied into a model new_model with the new references assigned to x_new and cref_new.\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @constraint(model, cref, x == 2)\ncref : x = 2\n\njulia> new_model = copy(model);\n\njulia> x_new = model[:x]\nx\n\njulia> cref_new = model[:cref]\ncref : x = 2\n\n\n\n\n\n","category":"method"},{"location":"api/JuMP/#Base.write(::IO,-::GenericModel;-::MOI.FileFormats.FileFormat)","page":"JuMP","title":"Base.write(::IO, ::GenericModel; ::MOI.FileFormats.FileFormat)","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.Base.write(::IO, ::GenericModel; ::MOI.FileFormats.FileFormat)","category":"page"},{"location":"api/JuMP/#Base.write-Tuple{IO, GenericModel}","page":"JuMP","title":"Base.write","text":"Base.write(\n    io::IO,\n    model::GenericModel;\n    format::MOI.FileFormats.FileFormat = MOI.FileFormats.FORMAT_MOF,\n    kwargs...,\n)\n\nWrite the JuMP model model to io in the format format.\n\nOther kwargs are passed to the Model constructor of the chosen format.\n\n\n\n\n\n","category":"method"},{"location":"api/JuMP/#MOI.Utilities.reset_optimizer(::GenericModel)","page":"JuMP","title":"MOI.Utilities.reset_optimizer(::GenericModel)","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.MOI.Utilities.reset_optimizer(::GenericModel)","category":"page"},{"location":"api/JuMP/#MathOptInterface.Utilities.reset_optimizer-Tuple{GenericModel}","page":"JuMP","title":"MathOptInterface.Utilities.reset_optimizer","text":"MOIU.reset_optimizer(model::GenericModel)\n\nCall MOIU.reset_optimizer on the backend of model.\n\nCannot be called in direct mode.\n\n\n\n\n\n","category":"method"},{"location":"api/JuMP/#MOI.Utilities.drop_optimizer(::GenericModel)","page":"JuMP","title":"MOI.Utilities.drop_optimizer(::GenericModel)","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.MOI.Utilities.drop_optimizer(::GenericModel)","category":"page"},{"location":"api/JuMP/#MathOptInterface.Utilities.drop_optimizer-Tuple{GenericModel}","page":"JuMP","title":"MathOptInterface.Utilities.drop_optimizer","text":"MOIU.drop_optimizer(model::GenericModel)\n\nCall MOIU.drop_optimizer on the backend of model.\n\nCannot be called in direct mode.\n\n\n\n\n\n","category":"method"},{"location":"api/JuMP/#MOI.Utilities.attach_optimizer(::GenericModel)","page":"JuMP","title":"MOI.Utilities.attach_optimizer(::GenericModel)","text":"","category":"section"},{"location":"api/JuMP/","page":"JuMP","title":"JuMP","text":"JuMP.MOI.Utilities.attach_optimizer(::GenericModel)","category":"page"},{"location":"api/JuMP/#MathOptInterface.Utilities.attach_optimizer-Tuple{GenericModel}","page":"JuMP","title":"MathOptInterface.Utilities.attach_optimizer","text":"MOIU.attach_optimizer(model::GenericModel)\n\nCall MOIU.attach_optimizer on the backend of model.\n\nCannot be called in direct mode.\n\n\n\n\n\n","category":"method"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"EditURL = \"constraint_programming.jl\"","category":"page"},{"location":"tutorials/linear/constraint_programming/#Constraint-programming","page":"Constraint programming","title":"Constraint programming","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"JuMP supports a range of constraint-programming type constraints via the corresponding sets in MathOptInterface. For most constraints, there are reformulations built-in that convert the constraint programming constraint into a mixed-integer programming equivalent.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"Because of this reformulation, all variables must be integer, and they must typically have finite bounds. An error will be thrown if the reformulation requires finiteness and you have a variable with non-finite bounds.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/linear/constraint_programming/#AllDifferent","page":"Constraint programming","title":"AllDifferent","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.AllDifferent set ensures that every element in a list takes a different integer value.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, 1 <= x[1:4] <= 4, Int)\n@constraint(model, x in MOI.AllDifferent(4))\noptimize!(model)\nvalue.(x)","category":"page"},{"location":"tutorials/linear/constraint_programming/#BinPacking","page":"Constraint programming","title":"BinPacking","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.BinPacking set can be used to divide up a set of items into different groups, such that the sum of their weights does not exceed the capacity of a bin.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"weights, capacity = Float64[1, 1, 2, 2, 3], 3.0;\nnumber_of_bins = 3\nmodel = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, 1 <= x[1:length(weights)] <= number_of_bins, Int)\n@constraint(model, x in MOI.BinPacking(capacity, weights))\noptimize!(model)\nvalue.(x)","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"Here, the value of x[i] is the bin that item i was placed into.","category":"page"},{"location":"tutorials/linear/constraint_programming/#Circuit","page":"Constraint programming","title":"Circuit","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.Circuit set is used to construct a tour of a list of N variables. They will each be assigned an integer from 1 to N, that describes the successor to each variable in the list:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x[1:4], Int)\n@constraint(model, x in MOI.Circuit(4))\noptimize!(model)","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"Let's see what tour was found, starting at node number 1:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"y = round.(Int, value.(x))\ntour = Int[1]\nwhile length(tour) < length(y)\n    push!(tour, y[tour[end]])\nend\ntour","category":"page"},{"location":"tutorials/linear/constraint_programming/#CountAtLeast","page":"Constraint programming","title":"CountAtLeast","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.CountAtLeast set is used to ensure that at least n elements in a set of variables belong to a set of values.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"For example, here is a model with three variables, constrained between 0 and 5:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, 0 <= x[1:3] <= 5, Int)","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"If we want to ensure that at least one element of each set {x[1], x[2]} and {x[2], x[3]} is in the set {3}, then we create a list of variables by concatenating the sets together:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"variables = [x[1], x[2], x[2], x[3]]","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"Then we need a partition list that contains the number of elements in each set of variables:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"partitions = [2, 2]","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"Finally, we need a set of values that the elements must be a part of:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"values = Set([3])","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"And the number of elements that must be part of the set values:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"n = 1","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The constraint is:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"@constraint(model, variables in MOI.CountAtLeast(n, partitions, values))","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"To ensure the uniqueness of the solution, we'll add a constraint that x[2] must be <= 2. This ensures that the only feasible solution is for x[1] and x[3] to be 3:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"@constraint(model, x[2] <= 2)","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"Let's check that we found a valid solution:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"optimize!(model)\nvalue.(x)","category":"page"},{"location":"tutorials/linear/constraint_programming/#CountBelongs","page":"Constraint programming","title":"CountBelongs","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.CountBelongs set is used to count how many elements in a set of variables belong to a set of values.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"For example, to count how many elements in a set of 4 variables belong to the set {2, 3}, do:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, 0 <= x[i = 1:4] <= i, Int)\n@variable(model, n, Int)\n@objective(model, Max, sum(x))\nset = Set([2, 3])\n@constraint(model, [n; x] in MOI.CountBelongs(1 + length(x), set))\noptimize!(model)\nvalue(n), value.(x)","category":"page"},{"location":"tutorials/linear/constraint_programming/#CountDistinct","page":"Constraint programming","title":"CountDistinct","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.CountDistinct set is used to count the number of distinct elements in a set of variables.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, 0 <= x[i = 1:4] <= i, Int)\n@variable(model, n, Int)\n@objective(model, Max, sum(x))\n@constraint(model, [n; x] in MOI.CountDistinct(1 + length(x)))\noptimize!(model)\nvalue(n), value.(x)","category":"page"},{"location":"tutorials/linear/constraint_programming/#CountGreaterThan","page":"Constraint programming","title":"CountGreaterThan","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.CountGreaterThan set is used to strictly upper-bound the number of distinct elements in a set of variables that have a value equal to another variable.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"For example, to count the number n of times that y appears in the vector x, use:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, 0 <= x[i = 1:4] <= i, Int)\n@variable(model, n, Int)\n@variable(model, 3 <= y <= 4, Int)\n@objective(model, Max, sum(x))\n@constraint(model, [n; y; x] in MOI.CountGreaterThan(1 + 1 + length(x)))\noptimize!(model)\nvalue(n), value(y), value.(x)","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"Here n is strictly greater than the count, and there is no limit on how large n could be. For example, n = 100 is also a feasible solution. The only constraint is that n cannot be equal to or smaller than the number of times that y appears.","category":"page"},{"location":"tutorials/linear/constraint_programming/#Table","page":"Constraint programming","title":"Table","text":"","category":"section"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"The MOI.Table set is used to select a single row from a matrix of values.","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"For example, given a matrix:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"table = Float64[1 1 0; 0 1 1; 1 0 1; 1 1 1]","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"we can constraint a 3-element vector x to equal one of the rows in table via:","category":"page"},{"location":"tutorials/linear/constraint_programming/","page":"Constraint programming","title":"Constraint programming","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x[i = 1:3], Int)\n@constraint(model, x in MOI.Table(table))\noptimize!(model)\nvalue.(x)","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"EditURL = \"https://github.com/oxfordcontrol/COSMO.jl/blob/v0.8.8/README.md\"","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"<div style=\"display: flex; justify-content: flex-end\" margin=0px>\n    <img src=\"https://github.com/migarstka/COSMO_assets/blob/master/star_badge_3.png?raw=true\" align=\"right\" width=6%>\n</div>","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"<h1 align=\"center\" margin=0px>\n  <img src=\"https://github.com/migarstka/COSMO_assets/blob/master/cosmo_rocket_with_convex_set.png?raw=true\" width=70%>\n</h1>","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"<p align=\"center\">\n   <a href=\"https://github.com/oxfordcontrol/COSMO.jl/actions\"><img src=\"https://github.com/oxfordcontrol/COSMO.jl/workflows/ci/badge.svg?branch=master\"></a>\n  <a href=\"https://codecov.io/gh/oxfordcontrol/COSMO.jl\"><img src=\"https://codecov.io/gh/oxfordcontrol/COSMO.jl/branch/master/graph/badge.svg?raw=true\"></a>\n  <a href=\"https://oxfordcontrol.github.io/COSMO.jl/stable\"><img src=\"https://img.shields.io/badge/Documentation-stable-purple.svg?raw=true\"></a>\n  <a href=\"https://opensource.org/licenses/Apache-2.0\"><img src=\"https://img.shields.io/badge/License-Apache%202.0-blue.svg?raw=true\"></a>\n  <a href=\"https://github.com/oxfordcontrol/COSMO.jl/releases\"><img src=\"https://img.shields.io/badge/Release-v0.8.8-blue.svg?raw=true\"></a>\n</p>","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"<p align=\"center\">\n  <a href=\"#features\">Features</a> •\n  <a href=\"#installation\">Installation</a> •\n  <a href=\"NEWS.md\">News</a> •\n  <a href=\"#citing-\">Citing</a> •\n  <a href=\"#contributing\">Contributing</a>\n</p>","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"This is a Julia implementation of the Conic operator splitting method (COSMO) solver. It can solve large convex conic optimization problems of the following form:","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"<p align=\"center\">\n<img src=\"https://github.com/migarstka/COSMO_assets/blob/master/cosmo_format.png?raw=true\" width=220px>\n</p>","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"with decision variables x ϵ R^n, s ϵ R^m and data matrices P=P'>=0, q ϵ R^n, A ϵ R^(m×n), and b ϵ R^m. The convex set K is a composition of convex sets and cones.","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"For more information take a look at the COSMO.jl Documentation (stable |  dev).","category":"page"},{"location":"packages/COSMO/#Features","page":"oxfordcontrol/COSMO.jl","title":"Features","text":"","category":"section"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"Versatile: COSMO solves linear programs, quadratic programs, second-order cone programs, semidefinite programs and problems involving exponential and power cones\nQuad SDPs: Positive semidefinite programs with quadratic objective functions are natively supported\nSafeguarded acceleration: robust and faster convergence to higher precision using COSMOAccelerators\nInfeasibility detection: Infeasible problems are detected without a homogeneous self-dual embedding of the problem\nJuMP / Convex.jl support: We provide an interface to MathOptInterface (MOI), which allows you to describe your problem in JuMP and Convex.jl.\nWarm starting: COSMO supports warm starting of the decision variables\nCustom sets and linear solver: Customize COSMO's components by defining your own convex constraint sets and by choosing from a number of direct and indirect linear system solvers, for example, QDLDL, Pardiso, Conjugate Gradient and MINRES\nArbitrary precision types: You can solve problems with any floating point precision.\nOpen Source: Our code is free to use and distributed under the Apache 2.0 Licence\nChordal decomposition: COSMO tries to decompose large structured PSD constraints using chordal decomposition techniques. This often results in a significant speedup compared to the original problem.\nSmart clique merging: After an initial decomposition of a structured SDP, COSMO recombines overlapping cliques/blocks to speed up the algorithm.","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"<div align=\"center\" margin=0px>\n  <img src=\"https://github.com/oxfordcontrol/COSMO.jl/blob/master/docs/src/assets/example_decomposition.gif?raw=true\" width=45%>\n</div>","category":"page"},{"location":"packages/COSMO/#Installation","page":"oxfordcontrol/COSMO.jl","title":"Installation","text":"","category":"section"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"COSMO can be added via the Julia package manager (type ]): pkg> add COSMO","category":"page"},{"location":"packages/COSMO/#Citing","page":"oxfordcontrol/COSMO.jl","title":"Citing","text":"","category":"section"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"If you find COSMO useful in your project, we kindly request that you cite the following paper:","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"@Article{Garstka_2021,\n  author  = {Michael Garstka and Mark Cannon and Paul Goulart},\n  journal = {Journal of Optimization Theory and Applications},\n  title   = {{COSMO}: A Conic Operator Splitting Method for Convex Conic Problems},\n  volume  = {190},\n  number  = {3},\n  pages   = {779--810},\n  year    = {2021},\n  publisher = {Springer},\n  doi     = {10.1007/s10957-021-01896-x},\n  url     = {https://doi.org/10.1007/s10957-021-01896-x}\n}","category":"page"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"The article is available under Open Access here.","category":"page"},{"location":"packages/COSMO/#Contributing","page":"oxfordcontrol/COSMO.jl","title":"Contributing","text":"","category":"section"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"Contributions are always welcome. Our style guide can be found here.\nCurrent issues, tasks and future ideas are listed in Issues. Please report any issues or bugs that you encounter.\nAs an open source project we are also interested in any projects and applications that use COSMO. Please let us know by opening a GitHub issue.","category":"page"},{"location":"packages/COSMO/#Python-Interface","page":"oxfordcontrol/COSMO.jl","title":"Python - Interface","text":"","category":"section"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"COSMO can also be called from Python. Take a look at: cosmo-python","category":"page"},{"location":"packages/COSMO/#Licence","page":"oxfordcontrol/COSMO.jl","title":"Licence 🔍","text":"","category":"section"},{"location":"packages/COSMO/","page":"oxfordcontrol/COSMO.jl","title":"oxfordcontrol/COSMO.jl","text":"This project is licensed under the Apache License - see the LICENSE.md file for details.","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"EditURL = \"https://github.com/chriscoey/Hypatia.jl/blob/v0.7.3/README.md\"","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"<img src=\"https://github.com/chriscoey/Hypatia.jl/wiki/hypatia_logo.png?raw=true\" alt=\"Hypatia logo\" width=\"358\"/>","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"(Image: Build Status) (Image: codecov) (Image: ) (Image: )","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Hypatia is a highly customizable open source interior point solver for generic conic optimization problems, written in Julia.","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"For more information on Hypatia, please see:","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"documentation for Hypatia's conic form, predefined cones, and interfaces\ncones reference for cone definitions and oracles\nexamples folder for applied examples and instances\nbenchmarks folder for scripts used to run and analyze various computational benchmarks","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"and preprints of our papers:","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Solving natural conic formulations with Hypatia.jl for computational arguments for expanding the class of cones recognized by conic solvers\nPerformance enhancements for a generic conic interior point algorithm for a description of Hypatia's algorithm and our enhanced stepping procedures\nSum of squares generalizations for conic sets for barriers and computational techniques for our generalized polynomial sum of squares cones\nConic optimization with spectral functions on Euclidean Jordan algebras for barriers and computational techniques for many of our epigraph/hypograph cones","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"and corresponding raw results CSV files generated by our run scripts in the benchmarks folder.","category":"page"},{"location":"packages/Hypatia/#License","page":"chriscoey/Hypatia.jl","title":"License","text":"","category":"section"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Hypatia is licensed under the MIT License (see LICENSE).","category":"page"},{"location":"packages/Hypatia/#Installation","page":"chriscoey/Hypatia.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"To use Hypatia, install Julia, then at the Julia REPL, type:","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"using Hypatia\nusing Pkg\nPkg.add(\"Hypatia\")","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Hypatia is an experimental solver and a work in progress, and may not run with older releases of Julia. Default options/parameters are not well-tuned, so we encourage you to experiment with these.","category":"page"},{"location":"packages/Hypatia/#Usage","page":"chriscoey/Hypatia.jl","title":"Usage","text":"","category":"section"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Hypatia can be accessed through a low-level native Julia interface or through open-source modeling tools such as JuMP and Convex.jl. The native interface is more expressive, allowing Hypatia to solve conic models expressed with generic real floating point types and structured matrices or linear operators, for example. However, it is typically sufficient and more convenient to use JuMP.","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Using JuMP, we can model a simple D-optimal experiment design problem and call Hypatia:","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"using LinearAlgebra\nusing JuMP\nusing Hypatia\n\nmodel = Model(() -> Hypatia.Optimizer(verbose = false))\n@variable(model, x[1:3] >= 0)\n@constraint(model, sum(x) == 5)\n@variable(model, hypo)\n@objective(model, Max, hypo)\nV = rand(2, 3)\nQ = V * diagm(x) * V'\naff = vcat(hypo, [Q[i, j] for i in 1:2 for j in 1:i]...)\n@constraint(model, aff in MOI.RootDetConeTriangle(2))\n\n# solve and query solution\noptimize!(model)\ntermination_status(model)\nobjective_value(model)\nvalue.(x)","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"See our D-optimal design example for more information and references.","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Many more examples using the native interface or JuMP can be found in the examples folder.","category":"page"},{"location":"packages/Hypatia/#Contributing","page":"chriscoey/Hypatia.jl","title":"Contributing","text":"","category":"section"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"Comments, questions, suggestions, and improvements/extensions to the code or documentation are welcomed. Please reach out on Discourse, or submit an issue or contribute a PR on our GitHub. If contributing code, try to maintain consistent style and add docstrings or comments for clarity. New examples are welcomed and should be implemented similarly to the existing examples.","category":"page"},{"location":"packages/Hypatia/#Acknowledgements","page":"chriscoey/Hypatia.jl","title":"Acknowledgements","text":"","category":"section"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"This work has been partially funded by the National Science Foundation under grant OAC-1835443 and the Office of Naval Research under grant N00014-18-1-2079.","category":"page"},{"location":"packages/Hypatia/#Citing-Hypatia","page":"chriscoey/Hypatia.jl","title":"Citing Hypatia","text":"","category":"section"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"If you find Hypatia solver useful, please cite our solver paper:","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"@article{coey2022solving,\n    title={Solving natural conic formulations with {H}ypatia.jl},\n    author={Chris Coey and Lea Kapelevich and Juan Pablo Vielma},\n    year={2022},\n    journal={INFORMS Journal on Computing},\n    publisher={INFORMS},\n    volume={34},\n    number={5},\n    pages={2686--2699},\n    doi={https://doi.org/10.1287/ijoc.2022.1202}\n}","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"If you find aspects of Hypatia's IPM implementation useful, please cite our algorithm paper:","category":"page"},{"location":"packages/Hypatia/","page":"chriscoey/Hypatia.jl","title":"chriscoey/Hypatia.jl","text":"@article{coey2022performance,\n    title={Performance enhancements for a generic conic interior point algorithm},\n    author={Chris Coey and Lea Kapelevich and Juan Pablo Vielma},\n    year={2023},\n    journal={Mathematical Programming Computation},\n    publisher={Springer},\n    volume={15},\n    pages={53--101},\n    doi={https://doi.org/10.1007/s12532-022-00226-0}\n}","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"EditURL = \"arbitrary_precision.jl\"","category":"page"},{"location":"tutorials/conic/arbitrary_precision/#Arbitrary-precision-arithmetic","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"","category":"section"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The purpose of this tutorial is to explain how to use a solver which supports arithmetic using a number type other than Float64.","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"using JuMP\nimport CDDLib\nimport Clarabel","category":"page"},{"location":"tutorials/conic/arbitrary_precision/#Higher-precision-arithmetic","page":"Arbitrary precision arithmetic","title":"Higher-precision arithmetic","text":"","category":"section"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"To create a model with a number type other than Float64, use GenericModel with an optimizer which supports the same number type:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"model = GenericModel{BigFloat}(Clarabel.Optimizer{BigFloat})","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The syntax for adding decision variables is the same as a normal JuMP model, except that values are converted to BigFloat:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"@variable(model, -1 <= x[1:2] <= sqrt(big\"2\"))","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"Note that each x is now a GenericVariableRef{BigFloat}, which means that the value of x in a solution will be a BigFloat.","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The lower and upper bounds of the decision variables are also BigFloat:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"lower_bound(x[1])","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"typeof(lower_bound(x[1]))","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"upper_bound(x[2])","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"typeof(upper_bound(x[2]))","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The syntax for adding constraints is the same as a normal JuMP model, except that coefficients are converted to BigFloat:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"@constraint(model, c, x[1] == big\"2\" * x[2])","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The function is a GenericAffExpr with BigFloat for the coefficient and variable types;","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"constraint = constraint_object(c)\ntypeof(constraint.func)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"and the set is a MOI.EqualTo{BigFloat}:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"typeof(constraint.set)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The syntax for adding and objective is the same as a normal JuMP model, except that coefficients are converted to BigFloat:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"@objective(model, Min, 3x[1]^2 + 2x[2]^2 - x[1] - big\"4\" * x[2])","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"typeof(objective_function(model))","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"Here's the model we have built:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"print(model)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"Let's solve and inspect the solution:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The value of each decision variable is a BigFloat:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"value.(x)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"as well as other solution attributes like the objective value:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"objective_value(model)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"and dual solution:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"dual(c)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"This problem has an analytic solution of x = [3//7, 3//14]. Currently, our solution has an error of approximately 1e-9:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"value.(x) .- [3 // 7, 3 // 14]","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"But by reducing the tolerances, we can obtain a more accurate solution:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"set_attribute(model, \"tol_gap_abs\", 1e-32)\nset_attribute(model, \"tol_gap_rel\", 1e-32)\noptimize!(model)\nvalue.(x) .- [3 // 7, 3 // 14]","category":"page"},{"location":"tutorials/conic/arbitrary_precision/#Rational-arithmetic","page":"Arbitrary precision arithmetic","title":"Rational arithmetic","text":"","category":"section"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"In addition to higher-precision floating point number types like BigFloat, JuMP also supports solvers with exact rational arithmetic. One example is CDDLib.jl, which supports the Rational{BigInt} number type:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"model = GenericModel{Rational{BigInt}}(CDDLib.Optimizer{Rational{BigInt}})","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"As before, we can create variables using rational bounds:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"@variable(model, 1 // 7 <= x[1:2] <= 2 // 3)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"lower_bound(x[1])","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"typeof(lower_bound(x[1]))","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"As well as constraints:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"@constraint(model, c1, (2 // 1) * x[1] + x[2] <= 1)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"@constraint(model, c2, x[1] + 3x[2] <= 9 // 4)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"and objective functions:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"@objective(model, Max, sum(x))","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"Here's the model we have built:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"print(model)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"Let's solve and inspect the solution:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"The optimal values are given in exact rational arithmetic:","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"value.(x)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"objective_value(model)","category":"page"},{"location":"tutorials/conic/arbitrary_precision/","page":"Arbitrary precision arithmetic","title":"Arbitrary precision arithmetic","text":"value(c2)","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/standard_form.md\"","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/manual/standard_form/#Standard-form-problem","page":"Standard form problem","title":"Standard form problem","text":"","category":"section"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"MathOptInterface represents optimization problems in the standard form:","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"beginalign\n     min_x in mathbbR^n  f_0(x)\n    \n     textst  f_i(x)  in mathcalS_i  i = 1 ldots m\nendalign","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"where:","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"the functions f_0 f_1 ldots f_m are specified by AbstractFunction objects\nthe sets mathcalS_1 ldots mathcalS_m are specified by AbstractSet objects","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"tip: Tip\nFor more information on this standard form, read our paper.","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"MOI defines some commonly used functions and sets, but the interface is extensible to other sets recognized by the solver.","category":"page"},{"location":"moi/manual/standard_form/#Functions","page":"Standard form problem","title":"Functions","text":"","category":"section"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"The function types implemented in MathOptInterface.jl are:","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Function Description\nVariableIndex x_j, the projection onto a single coordinate defined by a variable index j.\nVectorOfVariables The projection onto multiple coordinates (that is, extracting a sub-vector).\nScalarAffineFunction a^T x + b, where a is a vector and b scalar.\nScalarNonlinearFunction f(x), where f is a nonlinear function.\nVectorAffineFunction A x + b, where A is a matrix and b is a vector.\nScalarQuadraticFunction frac12 x^T Q x + a^T x + b, where Q is a symmetric matrix, a is a vector, and b is a constant.\nVectorQuadraticFunction A vector of scalar-valued quadratic functions.\nVectorNonlinearFunction f(x), where f is a vector-valued nonlinear function.","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Extensions for nonlinear programming are present but not yet well documented.","category":"page"},{"location":"moi/manual/standard_form/#One-dimensional-sets","page":"Standard form problem","title":"One-dimensional sets","text":"","category":"section"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"The one-dimensional set types implemented in MathOptInterface.jl are:","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Set Description\nLessThan(u) (-infty u\nGreaterThan(l) l infty)\nEqualTo(v) v\nInterval(l, u) l u\nInteger() mathbbZ\nZeroOne()  0 1 \nSemicontinuous(l, u)  0 cup l u\nSemiinteger(l, u)  0 cup ll+1ldotsu-1u","category":"page"},{"location":"moi/manual/standard_form/#Vector-cones","page":"Standard form problem","title":"Vector cones","text":"","category":"section"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"The vector-valued set types implemented in MathOptInterface.jl are:","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Set Description\nReals(d) mathbbR^d\nZeros(d) 0^d\nNonnegatives(d)  x in mathbbR^d  x ge 0 \nNonpositives(d)  x in mathbbR^d  x le 0 \nSecondOrderCone(d)  (tx) in mathbbR^d  t ge lVert x rVert_2 \nRotatedSecondOrderCone(d)  (tux) in mathbbR^d  2tu ge lVert x rVert_2^2 t ge 0u ge 0 \nExponentialCone()  (xyz) in mathbbR^3  y exp (xy) le z y  0 \nDualExponentialCone()  (uvw) in mathbbR^3  -u exp (vu) le exp(1) w u  0 \nGeometricMeanCone(d)  (tx) in mathbbR^1+n  x ge 0 t le sqrtnx_1 x_2 cdots x_n  where n is d - 1\nPowerCone(α)  (xyz) in mathbbR^3  x^alpha y^1-alpha ge z x ge 0y ge 0 \nDualPowerCone(α)  (uvw) in mathbbR^3  left(fracualpharight(^alphaleft(fracv1-alpharight)^1-alpha ge w uv ge 0 \nNormOneCone(d)  (tx) in mathbbR^d  t ge sum_i lvert x_i rvert \nNormInfinityCone(d)  (tx) in mathbbR^d  t ge max_i lvert x_i rvert \nRelativeEntropyCone(d)  (u v w) in mathbbR^d  u ge sum_i w_i log (fracw_iv_i) v_i ge 0 w_i ge 0 \nHyperRectangle(l, u) x in barmathbbR^d x_i in l_i u_i forall i=1ldotsd\nNormCone(p, d) ``{ (t,x) \\in \\mathbb{R}^{d} : t \\ge \\left(\\sum\\limits_i","category":"page"},{"location":"moi/manual/standard_form/#Matrix-cones","page":"Standard form problem","title":"Matrix cones","text":"","category":"section"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"The matrix-valued set types implemented in MathOptInterface.jl are:","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Set Description\nRootDetConeTriangle(d)  (tX) in mathbbR^1+d(1+d)2  t le det(X)^1d X mbox is the upper triangle of a PSD matrix \nRootDetConeSquare(d)  (tX) in mathbbR^1+d^2  t le det(X)^1d X mbox is a PSD matrix \nPositiveSemidefiniteConeTriangle(d)  X in mathbbR^d(d+1)2  X mbox is the upper triangle of a PSD matrix \nPositiveSemidefiniteConeSquare(d)  X in mathbbR^d^2  X mbox is a PSD matrix \nLogDetConeTriangle(d)  (tuX) in mathbbR^2+d(1+d)2  t le ulog(det(Xu)) X mbox is the upper triangle of a PSD matrix u  0  \nLogDetConeSquare(d)  (tuX) in mathbbR^2+d^2  t le u log(det(Xu)) X mbox is a PSD matrix u  0 \nNormSpectralCone(r, c)  (t X) in mathbbR^1 + r times c  t ge sigma_1(X) X mbox is a  rtimes cmbox matrix \nNormNuclearCone(r, c)  (t X) in mathbbR^1 + r times c  t ge sum_i sigma_i(X) X mbox is a  rtimes cmbox matrix \nHermitianPositiveSemidefiniteConeTriangle(d) The cone of Hermitian positive semidefinite matrices, with\nside_dimension rows and columns. \nScaled(S) The set S scaled so that Utilities.set_dot corresponds to LinearAlgebra.dot","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Some of these cones can take two forms: XXXConeTriangle and XXXConeSquare.","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"In XXXConeTriangle sets, the matrix is assumed to be symmetric, and the elements are provided by a vector, in which the entries of the upper-right triangular part of the matrix are given column by column (or equivalently, the entries of the lower-left triangular part are given row by row).","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"In XXXConeSquare sets, the entries of the matrix are given column by column (or equivalently, row by row), and the matrix is constrained to be symmetric. As an example, given a 2-by-2 matrix of variables X and a one-dimensional variable t, we can specify a root-det constraint as [t, X11, X12, X22] ∈ RootDetConeTriangle or [t, X11, X12, X21, X22] ∈ RootDetConeSquare.","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"We provide both forms to enable flexibility for solvers who may natively support one or the other. Transformations between XXXConeTriangle and XXXConeSquare are handled by bridges, which removes the chance of conversion mistakes by users or solver developers.","category":"page"},{"location":"moi/manual/standard_form/#Multi-dimensional-sets-with-combinatorial-structure","page":"Standard form problem","title":"Multi-dimensional sets with combinatorial structure","text":"","category":"section"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Other sets are vector-valued, with a particular combinatorial structure. Read their docstrings for more information on how to interpret them.","category":"page"},{"location":"moi/manual/standard_form/","page":"Standard form problem","title":"Standard form problem","text":"Set Description\nSOS1 A Special Ordered Set (SOS) of Type I\nSOS2 A Special Ordered Set (SOS) of Type II\nIndicator A set to specify an indicator constraint\nComplements A set to specify a mixed complementarity constraint\nAllDifferent The all_different global constraint\nBinPacking The bin_packing global constraint\nCircuit The circuit global constraint\nCountAtLeast The at_least global constraint\nCountBelongs The nvalue global constraint\nCountDistinct The distinct global constraint\nCountGreaterThan The count_gt global constraint\nCumulative The cumulative global constraint\nPath The path global constraint\nTable The table global constraint","category":"page"},{"location":"moi/reference/variables/","page":"Variables","title":"Variables","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/variables.md\"","category":"page"},{"location":"moi/reference/variables/","page":"Variables","title":"Variables","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/variables/#Variables","page":"Variables","title":"Variables","text":"","category":"section"},{"location":"moi/reference/variables/#Functions","page":"Variables","title":"Functions","text":"","category":"section"},{"location":"moi/reference/variables/","page":"Variables","title":"Variables","text":"add_variable\nadd_variables\nadd_constrained_variable\nadd_constrained_variables\nsupports_add_constrained_variable\nsupports_add_constrained_variables\nis_valid(::ModelLike,::VariableIndex)\ndelete(::ModelLike, ::VariableIndex)\ndelete(::ModelLike, ::Vector{VariableIndex})","category":"page"},{"location":"moi/reference/variables/#MathOptInterface.add_variable","page":"Variables","title":"MathOptInterface.add_variable","text":"add_variable(model::ModelLike)::VariableIndex\n\nAdd a scalar variable to the model, returning a variable index.\n\nA AddVariableNotAllowed error is thrown if adding variables cannot be done in the current state of the model model.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/variables/#MathOptInterface.add_variables","page":"Variables","title":"MathOptInterface.add_variables","text":"add_variables(model::ModelLike, n::Int)::Vector{VariableIndex}\n\nAdd n scalar variables to the model, returning a vector of variable indices.\n\nA AddVariableNotAllowed error is thrown if adding variables cannot be done in the current state of the model model.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/variables/#MathOptInterface.add_constrained_variable","page":"Variables","title":"MathOptInterface.add_constrained_variable","text":"add_constrained_variable(\n    model::ModelLike,\n    set::AbstractScalarSet\n)::Tuple{MOI.VariableIndex,\n         MOI.ConstraintIndex{MOI.VariableIndex, typeof(set)}}\n\nAdd to model a scalar variable constrained to belong to set, returning the index of the variable created and the index of the constraint constraining the variable to belong to set.\n\nBy default, this function falls back to creating a free variable with add_variable and then constraining it to belong to set with add_constraint.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/variables/#MathOptInterface.add_constrained_variables","page":"Variables","title":"MathOptInterface.add_constrained_variables","text":"add_constrained_variables(\n    model::ModelLike,\n    sets::AbstractVector{<:AbstractScalarSet}\n)::Tuple{\n    Vector{MOI.VariableIndex},\n    Vector{MOI.ConstraintIndex{MOI.VariableIndex,eltype(sets)}},\n}\n\nAdd to model scalar variables constrained to belong to sets, returning the indices of the variables created and the indices of the constraints constraining the variables to belong to each set in sets. That is, if it returns variables and constraints, constraints[i] is the index of the constraint constraining variable[i] to belong to sets[i].\n\nBy default, this function falls back to calling add_constrained_variable on each set.\n\n\n\n\n\nadd_constrained_variables(\n    model::ModelLike,\n    set::AbstractVectorSet,\n)::Tuple{\n    Vector{MOI.VariableIndex},\n    MOI.ConstraintIndex{MOI.VectorOfVariables,typeof(set)},\n}\n\nAdd to model a vector of variables constrained to belong to set, returning the indices of the variables created and the index of the constraint constraining the vector of variables to belong to set.\n\nBy default, this function falls back to creating free variables with add_variables and then constraining it to belong to set with add_constraint.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/variables/#MathOptInterface.supports_add_constrained_variable","page":"Variables","title":"MathOptInterface.supports_add_constrained_variable","text":"supports_add_constrained_variable(\n    model::ModelLike,\n    S::Type{<:AbstractScalarSet}\n)::Bool\n\nReturn a Bool indicating whether model supports constraining a variable to belong to a set of type S either on creation of the variable with add_constrained_variable or after the variable is created with add_constraint.\n\nBy default, this function falls back to supports_add_constrained_variables(model, Reals) && supports_constraint(model, MOI.VariableIndex, S) which is the correct definition for most models.\n\nExample\n\nSuppose that a solver supports only two kind of variables: binary variables and continuous variables with a lower bound. If the solver decides not to support VariableIndex-in-Binary and VariableIndex-in-GreaterThan constraints, it only has to implement add_constrained_variable for these two sets which prevents the user to add both a binary constraint and a lower bound on the same variable. Moreover, if the user adds a VariableIndex-in-GreaterThan constraint, implementing this interface (i.e., supports_add_constrained_variables) enables the constraint to be transparently bridged into a supported constraint.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/variables/#MathOptInterface.supports_add_constrained_variables","page":"Variables","title":"MathOptInterface.supports_add_constrained_variables","text":"supports_add_constrained_variables(\n    model::ModelLike,\n    S::Type{<:AbstractVectorSet}\n)::Bool\n\nReturn a Bool indicating whether model supports constraining a vector of variables to belong to a set of type S either on creation of the vector of variables with add_constrained_variables or after the variable is created with add_constraint.\n\nBy default, if S is Reals then this function returns true and otherwise, it falls back to supports_add_constrained_variables(model, Reals) && supports_constraint(model, MOI.VectorOfVariables, S) which is the correct definition for most models.\n\nExample\n\nIn the standard conic form (see Duality), the variables are grouped into several cones and the constraints are affine equality constraints. If Reals is not one of the cones supported by the solvers then it needs to implement supports_add_constrained_variables(::Optimizer, ::Type{Reals}) = false as free variables are not supported. The solvers should then implement supports_add_constrained_variables(::Optimizer, ::Type{<:SupportedCones}) = true where SupportedCones is the union of all cone types that are supported; it does not have to implement the method supports_constraint(::Type{VectorOfVariables}, Type{<:SupportedCones}) as it should return false and it's the default. This prevents the user to constrain the same variable in two different cones. When a VectorOfVariables-in-S is added, the variables of the vector have already been created so they already belong to given cones. If bridges are enabled, the constraint will therefore be bridged by adding slack variables in S and equality constraints ensuring that the slack variables are equal to the corresponding variables of the given constraint function.\n\nNote that there may also be sets for which !supports_add_constrained_variables(model, S) and supports_constraint(model, MOI.VectorOfVariables, S). For instance, suppose a solver supports positive semidefinite variable constraints and two types of variables: binary variables and nonnegative variables. Then the solver should support adding VectorOfVariables-in-PositiveSemidefiniteConeTriangle constraints, but it should not support creating variables constrained to belong to the PositiveSemidefiniteConeTriangle because the variables in PositiveSemidefiniteConeTriangle should first be created as either binary or non-negative.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/variables/#MathOptInterface.is_valid-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}","page":"Variables","title":"MathOptInterface.is_valid","text":"is_valid(model::ModelLike, index::Index)::Bool\n\nReturn a Bool indicating whether this index refers to a valid object in the model model.\n\n\n\n\n\n","category":"method"},{"location":"moi/reference/variables/#MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, MathOptInterface.VariableIndex}","page":"Variables","title":"MathOptInterface.delete","text":"delete(model::ModelLike, index::Index)\n\nDelete the referenced object from the model. Throw DeleteNotAllowed if if index cannot be deleted.\n\nThe following modifications also take effect if Index is VariableIndex:\n\nIf index used in the objective function, it is removed from the function, i.e., it is substituted for zero.\nFor each func-in-set constraint of the model:\nIf func isa VariableIndex and func == index then the constraint is deleted.\nIf func isa VectorOfVariables and index in func.variables then\nif length(func.variables) == 1 is one, the constraint is deleted;\nif length(func.variables) > 1 and supports_dimension_update(set) then then the variable is removed from func and set is replaced by update_dimension(set, MOI.dimension(set) - 1).\nOtherwise, a DeleteNotAllowed error is thrown.\nOtherwise, the variable is removed from func, i.e., it is substituted for zero.\n\n\n\n\n\n","category":"method"},{"location":"moi/reference/variables/#MathOptInterface.delete-Tuple{MathOptInterface.ModelLike, Vector{MathOptInterface.VariableIndex}}","page":"Variables","title":"MathOptInterface.delete","text":"delete(model::ModelLike, indices::Vector{R<:Index}) where {R}\n\nDelete the referenced objects in the vector indices from the model. It may be assumed that R is a concrete type. The default fallback sequentially deletes the individual items in indices, although specialized implementations may be more efficient.\n\n\n\n\n\n","category":"method"},{"location":"moi/reference/variables/#Attributes","page":"Variables","title":"Attributes","text":"","category":"section"},{"location":"moi/reference/variables/","page":"Variables","title":"Variables","text":"AbstractVariableAttribute\nVariableName\nVariablePrimalStart\nVariablePrimal\nVariableBasisStatus","category":"page"},{"location":"moi/reference/variables/#MathOptInterface.AbstractVariableAttribute","page":"Variables","title":"MathOptInterface.AbstractVariableAttribute","text":"AbstractVariableAttribute\n\nAbstract supertype for attribute objects that can be used to set or get attributes (properties) of variables in the model.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/variables/#MathOptInterface.VariableName","page":"Variables","title":"MathOptInterface.VariableName","text":"VariableName()\n\nA variable attribute for a string identifying the variable. It is valid for two variables to have the same name; however, variables with duplicate names cannot be looked up using get. It has a default value of \"\" if not set`.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/variables/#MathOptInterface.VariablePrimalStart","page":"Variables","title":"MathOptInterface.VariablePrimalStart","text":"VariablePrimalStart()\n\nA variable attribute for the initial assignment to some primal variable's value that the optimizer may use to warm-start the solve. May be a number or nothing (unset).\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/variables/#MathOptInterface.VariablePrimal","page":"Variables","title":"MathOptInterface.VariablePrimal","text":"VariablePrimal(result_index::Int = 1)\n\nA variable attribute for the assignment to some primal variable's value in result result_index. If result_index is omitted, it is 1 by default.\n\nIf the solver does not have a primal value for the variable because the result_index is beyond the available solutions (whose number is indicated by the ResultCount attribute), getting this attribute must throw a ResultIndexBoundsError. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check PrimalStatus before accessing the VariablePrimal attribute.\n\nSee ResultCount for information on how the results are ordered.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/variables/#MathOptInterface.VariableBasisStatus","page":"Variables","title":"MathOptInterface.VariableBasisStatus","text":"VariableBasisStatus(result_index::Int = 1)\n\nA variable attribute for the BasisStatusCode of a variable in result result_index, with respect to an available optimal solution basis.\n\nIf the solver does not have a basis statue for the variable because the result_index is beyond the available solutions (whose number is indicated by the ResultCount attribute), getting this attribute must throw a ResultIndexBoundsError. Otherwise, if the result is unavailable for another reason (for instance, only a dual solution is available), the result is undefined. Users should first check PrimalStatus before accessing the VariableBasisStatus attribute.\n\nSee ResultCount for information on how the results are ordered.\n\n\n\n\n\n","category":"type"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"EditURL = \"https://github.com/jump-dev/Ipopt.jl/blob/v1.4.2/README.md\"","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"(Image: )","category":"page"},{"location":"packages/Ipopt/#Ipopt.jl","page":"jump-dev/Ipopt.jl","title":"Ipopt.jl","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Ipopt.jl is a wrapper for the Ipopt solver.","category":"page"},{"location":"packages/Ipopt/#Affiliation","page":"jump-dev/Ipopt.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"This wrapper is maintained by the JuMP community and is not a COIN-OR project.","category":"page"},{"location":"packages/Ipopt/#License","page":"jump-dev/Ipopt.jl","title":"License","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Ipopt.jl is licensed under the MIT License.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"The underlying solver, coin-or/Ipopt, is licensed under the Eclipse public license.","category":"page"},{"location":"packages/Ipopt/#Installation","page":"jump-dev/Ipopt.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Install Ipopt.jl using the Julia package manager:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"import Pkg\nPkg.add(\"Ipopt\")","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"In addition to installing the Ipopt.jl package, this will also download and install the Ipopt binaries. You do not need to install Ipopt separately.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"To use a custom binary, read the Custom solver binaries section of the JuMP documentation.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"For details on using a different linear solver, see the Linear Solvers section below. You do not need a custom binary to change the linear solver.","category":"page"},{"location":"packages/Ipopt/#Use-with-JuMP","page":"jump-dev/Ipopt.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"You can use Ipopt with JuMP as follows:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"using JuMP, Ipopt\nmodel = Model(Ipopt.Optimizer)\nset_attribute(model, \"max_cpu_time\", 60.0)\nset_attribute(model, \"print_level\", 0)","category":"page"},{"location":"packages/Ipopt/#MathOptInterface-API","page":"jump-dev/Ipopt.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"The Ipopt optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}\nMOI.ObjectiveFunction{MOI.VariableIndex}","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"MOI.Reals","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"MOI.NLPBlock()\nMOI.NLPBlockDualStart()\nMOI.Name()\nMOI.ObjectiveSense()","category":"page"},{"location":"packages/Ipopt/#Options","page":"jump-dev/Ipopt.jl","title":"Options","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Supported options are listed in the Ipopt documentation.","category":"page"},{"location":"packages/Ipopt/#Solver-specific-callbacks","page":"jump-dev/Ipopt.jl","title":"Solver-specific callbacks","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Ipopt provides a callback that can be used to log the status of the optimization during a solve. It can also be used to terminate the optimization by returning false. Here is an example:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"using JuMP, Ipopt, Test\nmodel = Model(Ipopt.Optimizer)\nset_silent(model)\n@variable(model, x >= 1)\n@objective(model, Min, x + 0.5)\nx_vals = Float64[]\nfunction my_callback(\n   alg_mod::Cint,\n   iter_count::Cint,\n   obj_value::Float64,\n   inf_pr::Float64,\n   inf_du::Float64,\n   mu::Float64,\n   d_norm::Float64,\n   regularization_size::Float64,\n   alpha_du::Float64,\n   alpha_pr::Float64,\n   ls_trials::Cint,\n)\n   push!(x_vals, callback_value(model, x))\n   @test isapprox(obj_value, 1.0 * x_vals[end] + 0.5, atol = 1e-1)\n   # return `true` to keep going, or `false` to terminate the optimization.\n   return iter_count < 1\nend\nMOI.set(model, Ipopt.CallbackFunction(), my_callback)\noptimize!(model)\n@test MOI.get(model, MOI.TerminationStatus()) == MOI.INTERRUPTED\n@test length(x_vals) == 2","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"See the Ipopt documentation for an explanation of the arguments to the callback. They are identical to the output contained in the logging table printed to the screen.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"To access the current solution and primal, dual, and complementarity violations of each iteration, use Ipopt.GetIpoptCurrentViolations and Ipopt.GetIpoptCurrentIterate. The two functions are identical to the ones in the Ipopt C interface.","category":"page"},{"location":"packages/Ipopt/#C-API","page":"jump-dev/Ipopt.jl","title":"C API","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Ipopt.jl wraps the Ipopt C interface with minimal modifications.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"A complete example is available in the test/C_wrapper.jl file.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"For simplicity, the five callbacks required by Ipopt are slightly different to the C interface. They are as follows:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"\"\"\"\n   eval_f(x::Vector{Float64})::Float64\n\nReturns the objective value `f(x)`.\n\"\"\"\nfunction eval_f end\n\n\"\"\"\n   eval_grad_f(x::Vector{Float64}, grad_f::Vector{Float64})::Nothing\n\nFills `grad_f` in-place with the gradient of the objective function evaluated at\n`x`.\n\"\"\"\nfunction eval_grad_f end\n\n\"\"\"\n   eval_g(x::Vector{Float64}, g::Vector{Float64})::Nothing\n\nFills `g` in-place with the value of the constraints evaluated at `x`.\n\"\"\"\nfunction eval_g end\n\n\"\"\"\n   eval_jac_g(\n      x::Vector{Float64},\n      rows::Vector{Cint},\n      cols::Vector{Cint},\n      values::Union{Nothing,Vector{Float64}},\n   )::Nothing\n\nCompute the Jacobian matrix.\n\n* If `values === nothing`\n   - Fill `rows` and `cols` with the 1-indexed sparsity structure\n* Otherwise:\n   - Fill `values` with the elements of the Jacobian matrix according to the\n     sparsity structure.\n\n!!! warning\n    If `values === nothing`, `x` is an undefined object. Accessing any elements\n    in it will cause Julia to segfault.\n\"\"\"\nfunction eval_jac_g end\n\n\"\"\"\n   eval_h(\n      x::Vector{Float64},\n      rows::Vector{Cint},\n      cols::Vector{Cint},\n      obj_factor::Float64,\n      lambda::Float64,\n      values::Union{Nothing,Vector{Float64}},\n   )::Nothing\n\nCompute the Hessian-of-the-Lagrangian matrix.\n\n* If `values === nothing`\n   - Fill `rows` and `cols` with the 1-indexed sparsity structure\n* Otherwise:\n   - Fill `values` with the Hessian matrix according to the sparsity structure.\n\n!!! warning\n    If `values === nothing`, `x` is an undefined object. Accessing any elements\n    in it will cause Julia to segfault.\n\"\"\"\nfunction eval_h end","category":"page"},{"location":"packages/Ipopt/#INVALID_MODEL-error","page":"jump-dev/Ipopt.jl","title":"INVALID_MODEL error","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"If you get a termination status MOI.INVALID_MODEL, it is probably because you have some undefined value in your model, for example, a division by zero. Fix this by removing the division, or by imposing variable bounds so that you cut off the undefined region.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Instead of","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"model = Model(Ipopt.Optimizer)\n@variable(model, x)\n@NLobjective(model, 1 / x)","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"do","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"model = Model(Ipopt.Optimizer)\n@variable(model, x >= 0.0001)\n@NLobjective(model, 1 / x)","category":"page"},{"location":"packages/Ipopt/#Linear-Solvers","page":"jump-dev/Ipopt.jl","title":"Linear Solvers","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"To improve performance, Ipopt supports a number of linear solvers.","category":"page"},{"location":"packages/Ipopt/#HSL","page":"jump-dev/Ipopt.jl","title":"HSL","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Obtain a license and download HSL_jll.jl from https://licences.stfc.ac.uk/product/julia-hsl.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"There are two versions available: LBT and OpenBLAS. LBT is the recommended option for Julia ≥ v1.9.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Install this download into your current environment using:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"import Pkg\nPkg.develop(path = \"/full/path/to/HSL_jll.jl\")","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Then, use a linear solver in HSL by setting the hsllib and linear_solver attributes:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"using JuMP, Ipopt\nimport HSL_jll\nmodel = Model(Ipopt.Optimizer)\nset_attribute(model, \"hsllib\", HSL_jll.libhsl_path)\nset_attribute(model, \"linear_solver\", \"ma86\")","category":"page"},{"location":"packages/Ipopt/#macOS-users","page":"jump-dev/Ipopt.jl","title":"macOS users","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Due to the security policy of macOS, Mac users may need to delete the quarantine attribute of the ZIP archive before extracting. For example:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"xattr -d com.apple.quarantine lbt_HSL_jll.jl-2023.5.26.zip\nxattr -d com.apple.quarantine openblas_HSL_jll.jl-2023.5.26.zip","category":"page"},{"location":"packages/Ipopt/#Pardiso","page":"jump-dev/Ipopt.jl","title":"Pardiso","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Download Pardiso from https://www.pardiso-project.org. Save the shared library somewhere, and record the filename.","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Then, use Pardiso by setting the pardisolib and linear_solver attributes:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"using JuMP, Ipopt\nmodel = Model(Ipopt.Optimizer)\nset_attribute(model, \"pardisolib\", \"/full/path/to/libpardiso\")\nset_attribute(model, \"linear_solver\", \"pardiso\")","category":"page"},{"location":"packages/Ipopt/#SPRAL","page":"jump-dev/Ipopt.jl","title":"SPRAL","text":"","category":"section"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"If you use Ipopt.jl with Julia ≥ v1.9, the linear solver SPRAL is available. You can use it by setting the linear_solver attribute:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"using JuMP, Ipopt\nmodel = Model(Ipopt.Optimizer)\nset_attribute(model, \"linear_solver\", \"spral\")","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"Note that the following environment variables must be set before starting Julia:","category":"page"},{"location":"packages/Ipopt/","page":"jump-dev/Ipopt.jl","title":"jump-dev/Ipopt.jl","text":"export OMP_CANCELLATION=TRUE\nexport OMP_PROC_BIND=TRUE","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/release_notes.md\"","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"CurrentModule = MathOptInterface","category":"page"},{"location":"moi/release_notes/#Release-notes","page":"Release notes","title":"Release notes","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.","category":"page"},{"location":"moi/release_notes/#[v1.20.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.20.0)-(September-7,-2023)","page":"Release notes","title":"v1.20.0 (September 7, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added Scaled set (#2237) (#2263) (#2264)\nAdded ObjectiveLimit attribute (#2257)","category":"page"},{"location":"moi/release_notes/#Other","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Updated dependencies (#2258)\nImproved performance of ScalarNonlinearFunction utilities (#2259)\nFixed docstrings (#2261)","category":"page"},{"location":"moi/release_notes/#[v1.19.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.19.0)-(August-15,-2023)","page":"Release notes","title":"v1.19.0 (August 15, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-2","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added VectorNonlinearFunction (#2201)\nAdded new bridges\nBridges.Constraint.IntegerToZeroOneBridge (#2205)\nBridges.Constraint.ToScalarQuadraticBridge (#2235)\nBridges.Constraint.ToVectorQuadraticBridge (#2235)\nBridges.Constraint.ToScalarNonlinearBridge (#2233) (#2235)\nBridges.Constraint.FunctionConversionBridge (#2235)\nAdded Bridges.bridging_cost for setting a bridge-specific cost in the bridging graph (#2235)\nAdded Utilities.eval_variables support for ScalarNonlinearFunction (#2218) (#2219)\nAdded support for ScalarNonlinearFunction in FileFormats.NL (#2228) (#2231)\nAdded support for writing non-Float64 functions in FileFormats.MOF\nAdded Utilities.lazy_map (#2254)","category":"page"},{"location":"moi/release_notes/#Fixed","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed method ambiguities in operate (#2224)\nFixed reading LP file with a -infinity <= x <= +infinity variable (#2225)\nFixed missing require in Test.test_nonlinear_duals (#2230)\nFixed bug in ConstraintPrimal of Bridges.Constraint.QuadtoSOCBridge (#2240)","category":"page"},{"location":"moi/release_notes/#Other-2","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added extensions to solver-tests.yml (#2229)\nRefactored test/Benchmarks (#2234)\nFixed warnings in tests (#2241) (#2243)\nSmall refactoring of bridges for upcoming VectorNonlinearFunction (#2244) (#2245)\nFixed various typos (#2251) (#2255)\nPartitioned how we run the tests on GitHub actions (#2252) (#2253)","category":"page"},{"location":"moi/release_notes/#[v1.18.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.18.0)-(June-23,-2023)","page":"Release notes","title":"v1.18.0 (June 23, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-3","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added Bridges.Objective.SlackBridgePrimalDualStart (#2194)\nAdded constraint_gradient_structure and eval_constraint_gradient (#2200)","category":"page"},{"location":"moi/release_notes/#Fixed-2","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a missing @require in MOI.Test (#2195) (#2196)\nFixed incorrect usage of Utilities.operate! in bridges (#2207) (#2216)\nFixed splatting nonlinear expression with univariate operator (#2221)","category":"page"},{"location":"moi/release_notes/#Other-3","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Removed unused argument names (#2199)\nReduced memory requirement for tests (#2204)\nRefactored Utilities.promote_operation (#2206)\nImproved code style in Utilities/mutable_arithmetics.jl (#2209)\nRefactored various methods in Utilities/functions.jl (#2208) (#2212) (#2213) (#2214) (#2215)","category":"page"},{"location":"moi/release_notes/#[v1.17.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.1)-(June-6,-2023)","page":"Release notes","title":"v1.17.1 (June 6, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-3","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed comparison of ScalarFunctionConstantNotZero (#2190)","category":"page"},{"location":"moi/release_notes/#Other-4","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added documentation for enum instances (#2186)\nUpdated chatroom links in documentation (#2188)\nChanged the documentation to build on Julia v1.9 (#2191)","category":"page"},{"location":"moi/release_notes/#[v1.17.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.17.0)-(June-1,-2023)","page":"Release notes","title":"v1.17.0 (June 1, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-4","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added ScalarNonlinearFunction (#2059)\nAdded a variety of tests for NormSpectralCone, NormNuclearCone, and HermitianPositiveSemidefiniteConeTriangle (#2174)\nAdded Utilities.value_type for computing the value type of a function (#2176)","category":"page"},{"location":"moi/release_notes/#Fixed-4","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed support for external sets in Utilities.loadfromstring! (#2177)\nFixed promote_operation for ScalarNonlinearFunction (#2179)\nFixed two issues in FileFormats.LP when reading files with quadratic functions (#2182) (#2184)","category":"page"},{"location":"moi/release_notes/#[v1.16.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.16.0)-(May-16,-2023)","page":"Release notes","title":"v1.16.0 (May 16, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-5","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added support for MathOptFormat v1.3 and v1.4 (#2158) (#2169)\nAdded new method to the nonlinear API (#2162) (#2164)\neval_hessian_constraint\neval_hessian_objective\nhessian_constraint_structure\nhessian_objective_structure\nAdded new sets\nNormCone (#2119)\nScaledPositiveSemidefiniteConeTriangle (#2154)","category":"page"},{"location":"moi/release_notes/#Fixed-5","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed support for Julia v1.9 to work around a bug in the upstream Julia compiler (#2161) (#2163)\nFixed a correctness bug in Bridges.Constraint.HermitianToSymmetricPSDBridge (#2171)\nFixed convert(::VariableIndex, ::ScalarAffineFunction) when the function has terms with 0coefficients (#2173)","category":"page"},{"location":"moi/release_notes/#Other-5","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed solver-tests.yml (#2157)\nUpdated documentation links to developer chatroom (#2160)\nAdded various tests for bridges (#2156)\nAdded checklists to the developer documentation (#2167) (#2168)","category":"page"},{"location":"moi/release_notes/#[v1.15.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.1)-(April-25,-2023)","page":"Release notes","title":"v1.15.1 (April 25, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-6","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed deleting a variable in a bridged objective (#2150)","category":"page"},{"location":"moi/release_notes/#[v1.15.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.15.0)-(April-19,-2023)","page":"Release notes","title":"v1.15.0 (April 19, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-6","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added Bridges.Objective.VectorFunctionizeBridge (#2139)","category":"page"},{"location":"moi/release_notes/#Fixed-7","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed support for Rational in Bridges.Constraint.SplitIntervalBridge (#2137)\nFixed printing of LaTeX models (#2141)\nFixed modify in Bridges.Objective.VectorSlackBridge (#2144)\nFixed NAME record with spaces in FileFormats.MPS (#2146)\nFixed deleting a variable in a bridged objective (#2147)","category":"page"},{"location":"moi/release_notes/#Other-6","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add a test for variables in one-sided open Interval sets (#2133)\nMinor style fixes in the source code (#2148)","category":"page"},{"location":"moi/release_notes/#[v1.14.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.1)-(April-6,-2023)","page":"Release notes","title":"v1.14.1 (April 6, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-8","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in Bridges.print_active_bridges (#2135)","category":"page"},{"location":"moi/release_notes/#Other-7","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added a warning when an ambiguous string is passed to exclude in Test.runtests (#2136)","category":"page"},{"location":"moi/release_notes/#[v1.14.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.14.0)-(April-4,-2023)","page":"Release notes","title":"v1.14.0 (April 4, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-7","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added support for starting values in Bridges.Constraint.QuadtoSOCBridge (#2115)\nAdded support for Regex in the include and exclude arguments to Test.runtests (#2129)\nAdded Bridges.print_active_bridges methods for individual objectives and constraints (#2128)","category":"page"},{"location":"moi/release_notes/#Fixed-9","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed ResultCount when parsing .sol files in FileFormats.NL (#2130)","category":"page"},{"location":"moi/release_notes/#[v1.13.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.2)-(March-21,-2023)","page":"Release notes","title":"v1.13.2 (March 21, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-10","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed splatting of containers in MOI.Nonlinear (#2120)\nFixed a bug reading LP files with default bounds (#2121)\nFixed a bug in which Bridges.Constraint.HermitianToSymmetricPSDBridge was not enabled by default (#2123)","category":"page"},{"location":"moi/release_notes/#Other-8","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed typos in the documentation (#2114)\nFunctions now print to the REPL in algebraic form. This is potentially breaking if you have tests which rely on a specific String form of MOI functions. (#2112) (#2126)","category":"page"},{"location":"moi/release_notes/#[v1.13.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.1)-(March-3,-2023)","page":"Release notes","title":"v1.13.1 (March 3, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Other-9","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added the Google style guide to the documentation linter Vale, and fixed the resulting warnings (#2110)\nImproved the docstrings in src/functions.jl (#2108)","category":"page"},{"location":"moi/release_notes/#[v1.13.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.13.0)-(February-28,-2023)","page":"Release notes","title":"v1.13.0 (February 28, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-8","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added Bridges.Constraint.NumberConversionBridge (#2091)\nAdded Parameter set (#2095) (#2105) (#2106) (#2109)\nAdded with_cache_type argument to instantiate (#2097)\nAdded support for HermitianPositiveSemidefiniteConeTriangle in Utilities.Model (#2100)","category":"page"},{"location":"moi/release_notes/#Fixed-11","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed bug when Utilities.@product_of_sets is empty (#2101)\nFixed Bridges.print_active_bridges when variable bridge is an AbstractScalarSet (#2107)","category":"page"},{"location":"moi/release_notes/#Other-10","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added tests for vector-valued objective functions in FileFormats.MOF (#2093)\nUsed and documented preference for import MathOptInterface as MOI (#2096)\nFix and test links in the documentation with linkcheck = true (#2098)\nImproved docstrings of sets in src/sets.jl (#2099)\nSkip checking flakey links in documentation with linkcheck_ignore  (#2103)","category":"page"},{"location":"moi/release_notes/#[v1.12.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.12.0)-(February-10,-2023)","page":"Release notes","title":"v1.12.0 (February 10, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Added-9","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added support for vector-valued objective functions (#2070)\nAdded a Utilities.distance_to_set method for SecondOrderCone (#2060)","category":"page"},{"location":"moi/release_notes/#Fixed-12","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a number of constraint bridges so that Bridges.final_touch can be called multiple times without forcing a rebuild of the reformulation (#2089)","category":"page"},{"location":"moi/release_notes/#Other-11","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added new tests that a set ObjectiveFunction appears in ListOfModelAttributesSet (#2085)\nImproved the docstrings of a number of constraint-programming related sets (#2087)","category":"page"},{"location":"moi/release_notes/#[v1.11.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.5)-(January-24,-2023)","page":"Release notes","title":"v1.11.5 (January 24, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-13","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug writing .lp files with an off-diagonal quadratic objective (#2082)","category":"page"},{"location":"moi/release_notes/#Other-12","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added SnoopPrecompile directives for reduced time-to-first-X in Julia v1.9 (#2080)","category":"page"},{"location":"moi/release_notes/#[v1.11.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.4)-(January-12,-2023)","page":"Release notes","title":"v1.11.4 (January 12, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-14","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug reading .lp files with an Integer section (#2078)","category":"page"},{"location":"moi/release_notes/#[v1.11.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.3)-(January-12,-2023)","page":"Release notes","title":"v1.11.3 (January 12, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-15","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a performance bug when deleting a vector of constraints (#2072)\nFixed a bug reading .lp files with terms like x -1 y (#2076)","category":"page"},{"location":"moi/release_notes/#Other-13","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Documented the two-argument method of optimize! (#2074)","category":"page"},{"location":"moi/release_notes/#[v1.11.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.2)-(January-2,-2023)","page":"Release notes","title":"v1.11.2 (January 2, 2023)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-16","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug reading .mof.json files with ConstraintName set for VariableIndex constraints (#2066)\nFixed a bug reading .mof.json files with nonlinear objectives and no constraints (#2068)","category":"page"},{"location":"moi/release_notes/#[v1.11.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.1)-(December-22,-2022)","page":"Release notes","title":"v1.11.1 (December 22, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-17","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug reading .mof.json files with integer coefficients for affine and quadratic functions (#2063)","category":"page"},{"location":"moi/release_notes/#[v1.11.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.11.0)-(December-2,-2022)","page":"Release notes","title":"v1.11.0 (December 2, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-10","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added Utilities.PenaltyRelaxation and Utilities.ScalarPenaltyRelaxation (#1995)\nAdded Utilities.distance_to_set (#2048)\nAdded support for ConstraintPrimalStart and ConstraintDualStart in FileFormats.MOF (#2056)","category":"page"},{"location":"moi/release_notes/#Other-14","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Tidied these release notes (#2055)","category":"page"},{"location":"moi/release_notes/#[v1.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.10.0)-(November-22,-2022)","page":"Release notes","title":"v1.10.0 (November 22, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-11","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added new methods set(::OptimizerWithAttributes, ::RawOptimizerAttribute, value) and get(::OptimizerWithAttributes, ::RawOptimizerAttribute) (#2049)\nAdded new methods Utilities.DoubleDicts.outer_keys and Utilities.DoubleDicts.nonempty_outer_keys (#2052)","category":"page"},{"location":"moi/release_notes/#Fixed-18","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed Bridges.Objective.SlackBridge when the objective function is complex-valued (#2036) (#2038)\nFixed docstring of Test.runtests to clarify the warn_unsupported argument (#2037)\nFixed reading of free variables in FileFormats.LP (#2044)\nFixed numerous edge cases reading files from QPLIB using FileFormats.LP (#2042) (#2044)\nFixed situations in which x^y returns a complex value in Nonlinear (#2050)","category":"page"},{"location":"moi/release_notes/#Other-15","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Improved the error message thrown when a user-defined nonlinear function does not accept splatted input (#2032)\nRemoved specialized iterators for keys and values in Utilities.CleverDicts (#2051)","category":"page"},{"location":"moi/release_notes/#[v1.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.9.0)-(October-29,-2022)","page":"Release notes","title":"v1.9.0 (October 29, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-12","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added default fallback for getting ListOfConstraintIndices and NumberOfConstraints when the constraint type is unsupported by the model (#2021)\nAdded support for min and max in nonlinear expressions (#2023)\nAdded support for Indicator{EqualTo{T}} constraints in FileFormats.MPS (#2022)\nAdded default fallback for write_to_file and read_from_file (#2029)","category":"page"},{"location":"moi/release_notes/#Fixed-19","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed Constraint.ZeroOneBridge by adding new bounds as affine constraints instead of variable bounds (#1879)\nFixed reading free rows in FileFormats.MPS files (#2009)\nFixed parsing of OBJSENSE blocks in FileFormats.MPS files (#2016) (#2019)\nFixed the parsing of deeply nested nonlinear expressions by removing the use of recursion (#2020)\nFixed the requirements check in Test.test_constrainnt_get_ConstraintIndex (#2024)","category":"page"},{"location":"moi/release_notes/#[v1.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.2)-(September-20,-2022)","page":"Release notes","title":"v1.8.2 (September 20, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Documentation","page":"Release notes","title":"Documentation","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added vale as a documentation linter (#2002)\nImproved styling of code blocks in the PDF (#1999) (#2000)\nFixed a number of typos in the documentation (#2001) (#2003)","category":"page"},{"location":"moi/release_notes/#[v1.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.1)-(September-12,-2022)","page":"Release notes","title":"v1.8.1 (September 12, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-20","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in supports(::AbstractBridgeOptimizer for constraint attributes (#1991) (#1992)","category":"page"},{"location":"moi/release_notes/#[v1.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.8.0)-(September-1,-2022)","page":"Release notes","title":"v1.8.0 (September 1, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-13","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added new sets\nHyperRectangle (#1961)\nReified (#1955)\nAdded new bridges (#1955)\nBridges.Constraint.ReifiedAllDifferentToCountDistinctBridge\nBridges.Constraint.ReifiedCountDistinctToMILPBridge\nBridges.Constraint.SplitHyperRectangleBridge\nAdded support for atan(y, x) in Nonlinear (#1987)","category":"page"},{"location":"moi/release_notes/#Fixed-21","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Lazily construct expressions in Nonlinear so that expressions are updated when Nonlinear.Parameter values are updated (#1984)\nAllow NORM_LIMIT as a TerminationStatus for unbounded problems in Test (#1990)","category":"page"},{"location":"moi/release_notes/#[v1.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.7.0)-(August-16,-2022)","page":"Release notes","title":"v1.7.0 (August 16, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-14","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added new sets\nHermitianPositiveSemidefiniteConeTriangle\nAdded new optimizer-independent options\nRelativeGapTolerance\nAbsoluteGapTolerance\nAdded new bridges\nBridges.Constraint.GeoMeanToPowerBridge\nBridges.Constraint.HermitianToSymmetricPSDBridge\nBridges.Constraint.IndicatorGreaterToLessThanBridge\nBridges.Constraint.IndicatorLessToGreaterThanBridge\nBridges.Constraint.SplitComplexZerosBridge\nBridges.Constraint.SplitComplexEqualToBridge\nBridges.Objective.QuadratizeBridge\nAdded support for generic number types in Utilities.loadfromstring!\nUpdated FileFormats.MOF to MathOptFormat v1.1, enabling support for constraint programming sets in the MOF file format\nAdded support in various FileFormats for\nindicator constraints in FileFormats.MPS\nquadratic constraints and an objective in FileFormats.LP\nquadratic constraints and an objective in FileFormats.MPS","category":"page"},{"location":"moi/release_notes/#Fixed-22","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed some missing promotion rules","category":"page"},{"location":"moi/release_notes/#Other-16","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Improved the performance of Jacobian products in Nonlinear\nRemoved an un-needed copy in Utilities.modify_function\nVarious clean-ups in Bridges/bridge_optimizer.jl","category":"page"},{"location":"moi/release_notes/#[v1.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.1)-(July-23,-2022)","page":"Release notes","title":"v1.6.1 (July 23, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-23","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added support for ExponentialCone in MatrixOfConstraints\nFix PSDSquare_3 test to reflect a previously fixed bug getting the ConstraintDual of a PositiveSemidefiniteConeSquare constraint","category":"page"},{"location":"moi/release_notes/#[v1.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.6.0)-(July-2,-2022)","page":"Release notes","title":"v1.6.0 (July 2, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-15","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added Bridges.needs_final_touch and Bridges.final_touch\nAdded new bridges from constraint programming sets to mixed-integer linear programs:\nAllDifferentToCountDistinctBridge\nCountAtLeastToCountBelongsBridge\nCountBelongsToMILPBridge\nCountDistinctToMILPBridge\nCountGreaterThanToMILPBridge\nCircuitToMILPBridge","category":"page"},{"location":"moi/release_notes/#Fixed-24","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Relax an instance of ::Vector to ::AbstractVector in MOI.Nonlinear\nFix BinPackingToMILPBridge to respect variable bounds\nFix SemiToBinaryBridge to throw error if other bounds are set","category":"page"},{"location":"moi/release_notes/#[v1.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.5.0)-(June-27,-2022)","page":"Release notes","title":"v1.5.0 (June 27, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-16","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added GetAttributeNotAllowed for solvers to indicate when getting an attribute encounters an error\nAdded Utilities.get_fallback support for ObjectiveValue and DualObjectiveValue\nAdded new bridges:\nRootDetConeSquare to RootDetConeTriangle\nLogDetConeSquare to LogDetConeTriangle\nBinPacking to a mixed-integer linear program\nTable to a mixed-integer linear program\nAdded Bridges.print_active_bridges to display the current optimal hyper-path in a Bridges.LazyBridgeOptimizer","category":"page"},{"location":"moi/release_notes/#Fixed-25","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed ZeroOne tests with lower and upper bounds\nFixed error in FileFormats.LP when reading a malformed file\nFixed reading of nonlinear programs in FileFormats.MOF\nFixed bug in ConstraintDual when using SquareBridge","category":"page"},{"location":"moi/release_notes/#Other-17","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Improved documentation of nonlinear API\nDocumented duality convention for PositiveSemidefiniteConeSquare sets\nFixed typo in Bridges.Constraint.QuadToSOCBridge docstring","category":"page"},{"location":"moi/release_notes/#[v1.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.4.0)-(June-9,-2022)","page":"Release notes","title":"v1.4.0 (June 9, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-17","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added a number of sets for constraint programming:\nAllDifferent\nBinPacking\nCircuit\nCountAtLeast\nCountBelongs\nCountDistinct\nCountGreaterThan\nCumulative\nPath\nTable\nAdded support for user-defined hessians in Nonlinear\nAdded Bridges.runtests to simplify the testing of bridge implementations","category":"page"},{"location":"moi/release_notes/#Fixed-26","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in FileFormats.NL when writing univariate *","category":"page"},{"location":"moi/release_notes/#Other-18","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Began a large refactoring of the Bridges submodule, with greatly improved documentation.","category":"page"},{"location":"moi/release_notes/#[v1.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.3.0)-(May-27,-2022)","page":"Release notes","title":"v1.3.0 (May 27, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-18","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add MOI.Nonlinear submodule. This is a large new submodule that has been refactored from code that was in JuMP. For now, it should be considered experimental.\nAdd FileFormats.NL.SolFileResults(::IO, ::Model)\nAdd FileFormats.NL.read!(::IO, ::Model)\nAdd MOI.modify that accepts a vector of modifications","category":"page"},{"location":"moi/release_notes/#Fixed-27","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in Test which attempted to include non-.jl files\nFixed a bug in FileFormats for models with open interval constraints","category":"page"},{"location":"moi/release_notes/#Other-19","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a performance issue in Utilities.DoubleDict\nVarious minor improvements to the documentation","category":"page"},{"location":"moi/release_notes/#[v1.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.2.0)-(April-25,-2022)","page":"Release notes","title":"v1.2.0 (April 25, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-19","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add support for the FORMAT_REW/.rew file format in FileFormats.","category":"page"},{"location":"moi/release_notes/#Fixed-28","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fix bug handling of default variable bounds in FileFormats.LP\nFix FileFormats.MPS to not write OBJSENSE by default since this is only supported by some readers.","category":"page"},{"location":"moi/release_notes/#[v1.1.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.2)-(March-31,-2022)","page":"Release notes","title":"v1.1.2 (March 31, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-29","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fix a range of bugs in FileFormats.LP\nFix reading of problem dimensions in FileFormats.SDPA","category":"page"},{"location":"moi/release_notes/#[v1.1.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.1)-(March-23,-2022)","page":"Release notes","title":"v1.1.1 (March 23, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-30","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fix bug in test_model_UpperBoundAlreadySet\nFix bug in test_infeasible_ tests\nFix bug in test_objective_ObjectiveFunction_blank\nRelax restriction of MOI.AbstractOptimizer to MOI.ModelLike in Utilities.CachingOptimizer and instantiate.","category":"page"},{"location":"moi/release_notes/#New-tests","page":"Release notes","title":"New tests","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add test_conic_empty_matrix that checks conic solvers support problems with no variables.","category":"page"},{"location":"moi/release_notes/#[v1.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.1.0)-(March-2,-2022)","page":"Release notes","title":"v1.1.0 (March 2, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-20","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added MOI.Utilities.throw_unsupported(::UniversalFallback) for simplifying solver wrappers which copy from a UniversalFallback.","category":"page"},{"location":"moi/release_notes/#[v1.0.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.2)-(March-1,-2022)","page":"Release notes","title":"v1.0.2 (March 1, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-31","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in the test_model_ScalarFunctionConstantNotZero test\nFixed the error type when an AbstractFunctionConversionBridge cannot get or set an attribute\nIdentified a correctness bug in RSOCtoPSDBridge. We now thrown an error instead of returning an incorrect result.","category":"page"},{"location":"moi/release_notes/#[v1.0.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.1)-(February-25,-2022)","page":"Release notes","title":"v1.0.1 (February 25, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-32","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in which OptimizerAttributes were not copied in CachingOptimizer\nFixed a bug in which shift_constant did not promote mixed types of coefficients\nFixed a bug in which deleting a constraint of a bridged variable threw ErrorException instead of MOI.DeleteNotAllowed\nFixed a bug in which add_constraint in MatrixOfConstraints did not canonicalize the function\nFixed a bug when modifying scalar constants of a function containing a bridged variable\nFixed a bug in which final_touch was not always called with a CachingOptimizer","category":"page"},{"location":"moi/release_notes/#[v1.0.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v1.0.0)-(February-17,-2022)","page":"Release notes","title":"v1.0.0 (February 17, 2022)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Although tagged as a breaking release, v1.0.0 is v0.10.9 with deprecations removed, similar to how Julia 1.0 was Julia 0.7 with deprecations removed.","category":"page"},{"location":"moi/release_notes/#Breaking","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Julia 1.6 is now the minimum supported version\nAll deprecations have been removed","category":"page"},{"location":"moi/release_notes/#Troubleshooting-problems-when-updating","page":"Release notes","title":"Troubleshooting problems when updating","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"If you experience problems when updating, you are likely using previously deprecated features. (By default, Julia does not warn when you use deprecated features.)","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"To find the deprecated features you are using, start Julia with --depwarn=yes:","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"$ julia --depwarn=yes","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Then install MathOptInterface v0.10.9:","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"julia> using Pkg\njulia> pkg\"add MathOptInterface@0.10\"","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.","category":"page"},{"location":"moi/release_notes/#[v0.10.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.9)-(February-16,-2022)","page":"Release notes","title":"v0.10.9 (February 16, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-21","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added MOI.Utilities.FreeVariables as a new VariablesConstrainer for conic solvers\nAdded MOI.default_cache for specifying the model used in CachingOptimizer","category":"page"},{"location":"moi/release_notes/#Fixed-33","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed LaTeX printing of MOI.Interval sets","category":"page"},{"location":"moi/release_notes/#Other-20","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added Aqua.jl as a CI check, and fixed suggested issues\nThe constructors of GeoMeanBridge, StructOfConstraints, and CachingOptimizer were changed from outer to inner constructors. This change is technically breaking, but does not impact users who followed the documented API.","category":"page"},{"location":"moi/release_notes/#[v0.10.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.8)-(February-3,-2022)","page":"Release notes","title":"v0.10.8 (February 3, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-22","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added a Base.read! for FileFormats.LP.","category":"page"},{"location":"moi/release_notes/#Fixed-34","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a bug in MutableSparseMatrix\nFixed a bug when calling operate!(vcat, ...) with Number arguments\nRemoved unintended export of deprecated symbols\nFixed a bug with PowerCone and DualPowerCone in MatrixOfConstraints.","category":"page"},{"location":"moi/release_notes/#[v0.10.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.7)-(January-5,-2022)","page":"Release notes","title":"v0.10.7 (January 5, 2022)","text":"","category":"section"},{"location":"moi/release_notes/#Added-23","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added test for modifying the constant vector in a VectorAffineFunction-in-Zeros constraint.","category":"page"},{"location":"moi/release_notes/#Fixed-35","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed the order in which sets are added to a LazyBridgeOptimizer. Compared to v0.10.6, this may result in bridged models being created with a different number (and order) of variables and constraints. However, it was necessary to fix cases which were previously rejected as unsupported, even though there was a valid bridge transformation.\nFixed an error message in FileFormats.CBF\nFixed comparison in test_linear_integration_Interval\nFixed errors for ConstraintPrimal in a CachingOptimizer\nFixed printing of models with non-Float64 coefficients.","category":"page"},{"location":"moi/release_notes/#Other-21","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Various improvements to reduce time-to-first-solve latency\nImproved error message when an optimizer does not support compute_conflict!","category":"page"},{"location":"moi/release_notes/#[v0.10.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.6)-(November-30,-2021)","page":"Release notes","title":"v0.10.6 (November 30, 2021)","text":"","category":"section"},{"location":"moi/release_notes/#Added-24","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added new documentation and tests for infeasibility certificates\nAdded a version control system for the tests in MOI.Test.runtests. Pass exclude_tests_after = v\"0.10.5\" to run tests added in v0.10.5 and earlier.\nMOI.Test.runtests now supports generic number types. To specify the number type T, pass MOI.Test.Config(T).\nAdded infeasible_status to MOI.Test.Config for solvers which return LOCALLY_INFEASIBLE\nCachingOptimizers now use a fallback for ConstraintPrimal. This should enable solvers using a CachingOptimizer to pass tests requiring ConstraintPrimal.","category":"page"},{"location":"moi/release_notes/#Fixed-36","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed a StackOverflow bug in copy_to\nFixed error thrown when nonconvex quadratic constraints cannot be bridged\nFixed a bug in copy_to for FileFormats.NL.Model\nFixed a bug in FileFormats.NL when printing large integers\nRemove a common test failure for LowerBoundAlreadySet tests\nUtilities.num_rows is now exported\nRemove parts of failing test_model_copy_to_xxx tests due to bridges","category":"page"},{"location":"moi/release_notes/#[v0.10.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.5)-(November-7,-2021)","page":"Release notes","title":"v0.10.5 (November 7, 2021)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-37","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed getter in UniversalFallback\nFixed test_solve_conflict_zeroone_ii","category":"page"},{"location":"moi/release_notes/#Other-22","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Make normalize_and_add_constraint more flexible\nUpdate paper BibTeX","category":"page"},{"location":"moi/release_notes/#[v0.10.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.4)-(October-26,-2021)","page":"Release notes","title":"v0.10.4 (October 26, 2021)","text":"","category":"section"},{"location":"moi/release_notes/#Added-25","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add SolverVersion attribute\nAdd new tests:\ntest_solve_conflict_zeroone_ii\ntest_nonlinear_objective\nUtilities.VariablesContainer now supports ConstraintFunction and ConstraintSet\nThe documentation is now available as a PDF","category":"page"},{"location":"moi/release_notes/#Other-23","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Update to MutableArithmetics 0.3\nVarious improvements to the documentation","category":"page"},{"location":"moi/release_notes/#[v0.10.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.3)-(September-18,-2021)","page":"Release notes","title":"v0.10.3 (September 18, 2021)","text":"","category":"section"},{"location":"moi/release_notes/#Fixed-38","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed bug which prevented callbacks from working through a CachingOptimizer\nFixed bug in Test submodule","category":"page"},{"location":"moi/release_notes/#[v0.10.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.2)-(September-16,-2021)","page":"Release notes","title":"v0.10.2 (September 16, 2021)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Updated MathOptFormat to v1.0\nUpdated JSONSchema to v1.0\nAdded Utilities.set_with_dimension\nAdded two-argument optimize!(::AbstractOptimizer, ::ModelLike)\nThe experimental feature copy_to_and_optimize! has been removed\nDet bridges now support getting ConstraintFunction and ConstraintSet\nVarious minor bug fixes identified by improved testing","category":"page"},{"location":"moi/release_notes/#[v0.10.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.1)-(September-8,-2021)","page":"Release notes","title":"v0.10.1 (September 8, 2021)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Various fixes to MOI.Test","category":"page"},{"location":"moi/release_notes/#[v0.10.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.10.0)-(September-6,-2021)","page":"Release notes","title":"v0.10.0 (September 6, 2021)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"MOI v0.10 is a significant breaking release. There are a large number of user-visible breaking changes and code refactors, as well as a substantial number of new features.","category":"page"},{"location":"moi/release_notes/#Breaking-in-MOI","page":"Release notes","title":"Breaking in MOI","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"SingleVariable has been removed; use VariableIndex instead\nSingleVariableConstraintNameError has been renamed to VariableIndexConstraintNameError\nSettingSingleVariableFunctionNotAllowed has been renamed to SettingVariableIndexFunctionNotAllowed\nVariableIndex constraints should not support ConstraintName\nVariableIndex constraints should not support ConstraintBasisStatus; implement VariableBasisStatus instead\nListOfConstraints has been renamed to ListOfConstraintTypesPresent\nListOfConstraintTypesPresent should now return Tuple{Type,Type} instead of Tuple{DataType,DataType}\nSolveTime has been renamed to SolveTimeSec\nIndicatorSet has been renamed to Indicator\nRawParameter has been renamed to RawOptimizerAttribute and now takes String instead of Any as the only argument\nThe .N field in result attributes has been renamed to .result_index\nThe .variable_index field in ScalarAffineTerm has been renamed to .variable\nThe .variable_index_1 field in ScalarQuadraticTerm has been renamed to .variable_1\nThe .variable_index_2 field in ScalarQuadraticTerm has been renamed to .variable_2\nThe order of affine_terms and quadratic_terms in ScalarQuadraticFunction and VectorQuadraticFunction have been reversed. Both functions now accept quadratic, affine, and constant terms in that order.\nThe index_value function has been removed. Use .value instead.\nisapprox has been removed for SOS1 and SOS2.\nThe dimension argument to Complements(dimension::Int) should now be the length of the corresponding function, instead of half the length. An ArgumentError is thrown if dimension is not even.\ncopy_to no longer takes keyword arguments:\ncopy_names: now copy names if they are supported by the destination solver\nfilter_constraints: use Utilities.ModelFilter instead\nwarn_attributes: never warn about optimizer attributes","category":"page"},{"location":"moi/release_notes/#Breaking-in-Bridges","page":"Release notes","title":"Breaking in Bridges","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Constraint.RSOCBridge has been renamed to Constraint.RSOCtoSOCBridge\nConstraint.SOCRBridge has been renamed to Constraint.SOCtoRSOCBridge\nBridges now return vectors that can be modified by the user. Previously, some bridges returned views instead of copies.\nBridges.IndexInVector has been unified into a single type. Previously, there was a different type for each submodule within Bridges\nThe signature of indicator bridges has been fixed. Use MOI.Bridges.Constraint.IndicatortoSOS1{Float64}(model).","category":"page"},{"location":"moi/release_notes/#Breaking-in-FileFormats","page":"Release notes","title":"Breaking in FileFormats","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"FileFormats.MOF.Model no longer accepts validate argument. Use the JSONSchema package to validate the MOF file. See the documentation for more information.","category":"page"},{"location":"moi/release_notes/#Breaking-in-Utilities","page":"Release notes","title":"Breaking in Utilities","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"The datastructure of Utilities.Model (and models created with Utilities.@model) has been significantly refactored in a breaking way. This includes the way that objective functions and variable-related information is stored.\nUtilities.supports_default_copy has been renamed to supports_incremental_interface\nUtilities.automatic_copy_to has been renamed to Utilities.default_copy_to\nThe allocate-load API has been removed\nCachingOptimizers are now initialized as EMPTY_OPTIMIZER instead of ATTACHED_OPTIMIZER. If your code relies on the optimizer being attached, call MOIU.attach_optimizer(model) after creation.\nThe field names of Utilities.IndexMap have been renamed to var_map and con_map. Accessing these fields directly is considered a private detail that may change. Use the public getindex and setindex! API instead.\nThe size argument to Utilities.CleverDicts.CleverDict(::Integer) has been removed.\nThe size argument to Utilities.IndexMap(::Integer) has been removed.\nUtilities.DoubleDicts have been significantly refactored. Consult the source code for details.\nUtilities.test_models_equal has been moved to MOI.Test","category":"page"},{"location":"moi/release_notes/#Breaking-in-Test","page":"Release notes","title":"Breaking in Test","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"MOI.Test has been renamed to MOI.DeprecatedTest\nAn entirely new MOI.Test submodule has been written. See the documentation for details. The new MOI.Test submodule may find many bugs in the implementations of existing solvers that were previously untested.","category":"page"},{"location":"moi/release_notes/#Other-changes:","page":"Release notes","title":"Other changes:","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"attribute_value_type has been added\ncopy_to_and_optimize! has been added\nVariableBasisStatus has been added\nprint(model) now prints a human-readable description of the model\nVarious improvements to the FileFormats submodule\nFileFormats.CBF was refactored and received bugfixes\nSupport for MathOptFormat v0.6 was added in FileFormats.MOF\nFileFormats.MPS has had bugfixes and support for more features such as OBJSENSE and objective constants.\nFileFormats.NL has been added to support nonlinear files\nImproved type inference throughout to reduce latency","category":"page"},{"location":"moi/release_notes/#Updating","page":"Release notes","title":"Updating","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"A helpful script when updating is:","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"for (root, dirs, files) in walkdir(\".\")\n    for file in files\n        if !endswith(file, \".jl\")\n            continue\n        end\n        path = joinpath(root, file)\n        s = read(path, String)\n        for pair in [\n            \".variable_index\" => \".variable\",\n            \"RawParameter\" => \"RawOptimizerAttribute\",\n            \"ListOfConstraints\" => \"ListOfConstraintTypesPresent\",\n            \"TestConfig\" => \"Config\",\n            \"attr.N\" => \"attr.result_index\",\n            \"SolveTime\" => \"SolveTimeSec\",\n            \"DataType\" => \"Type\",\n            \"Utilities.supports_default_copy_to\" =>\n                \"supports_incremental_interface\",\n            \"SingleVariableConstraintNameError\" =>\n                \"VariableIndexConstraintNameError\",\n            \"SettingSingleVariableFunctionNotAllowed\" =>\n                \"SettingVariableIndexFunctionNotAllowed\",\n            \"automatic_copy_to\" => \"default_copy_to\",\n        ]\n            s = replace(s, pair)\n        end\n        write(path, s)\n    end\nend","category":"page"},{"location":"moi/release_notes/#[v0.9.22](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.22)-(May-22,-2021)","page":"Release notes","title":"v0.9.22 (May 22, 2021)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"This release contains backports from the ongoing development of the v0.10 release.","category":"page"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Improved type inference in Utilities, Bridges and FileFormats submodules to reduce latency.\nImproved performance of Utilities.is_canonical.\nFixed Utilities.pass_nonvariable_constraints with bridged variables.\nFixed performance regression of Utilities.Model.\nFixed ordering of objective setting in parser.","category":"page"},{"location":"moi/release_notes/#[v0.9.21](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.21)-(April-23,-2021)","page":"Release notes","title":"v0.9.21 (April 23, 2021)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added supports_shift_constant.\nImprove performance of bridging quadratic constraints.\nAdd precompilation statements.\nLarge improvements to the documentation.\nFix a variety of inference issues, benefiting precompilation and reducing initial latency.\nRawParameters are now ignored when resetting a CachingOptimizer. Previously, changing the underlying optimizer after RawParameters were set would throw an error.\nUtilities.AbstractModel is being refactored. This may break users interacting with private fields of a model generated using @model.","category":"page"},{"location":"moi/release_notes/#[v0.9.20](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.20)-(February-20,-2021)","page":"Release notes","title":"v0.9.20 (February 20, 2021)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Improved performance of Utilities.ScalarFunctionIterator\nAdded support for compute_conflict to MOI layers\nAdded test with zero off-diagonal quadratic term in objective\nFixed double deletion of nested bridged SingleVariable/VectorOfVariables constraints\nFixed modification of un-set objective\nFixed function modification with duplicate terms\nMade unit tests abort without failing if the problem class is not supported\nFormatted code with JuliaFormatter\nClarified BasisStatusCode's docstring","category":"page"},{"location":"moi/release_notes/#[v0.9.19](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.19)-(December-1,-2020)","page":"Release notes","title":"v0.9.19 (December 1, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added CallbackNodeStatus attribute\nAdded bridge from GreaterThan or LessThan to Interval\nAdded tests for infeasibility certificates and double optimize\nFixed support for Julia v1.6\nRe-organized MOI docs and added documentation for adding a test","category":"page"},{"location":"moi/release_notes/#[v0.9.18](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.18)-(November-3,-2020)","page":"Release notes","title":"v0.9.18 (November 3, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Various improvements for working with complex numbers\nAdded GeoMeantoRelEntrBridge to bridge a GeometricMeanCone constraint to a relative entropy constraint","category":"page"},{"location":"moi/release_notes/#[v0.9.17](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.17)-(September-21,-2020)","page":"Release notes","title":"v0.9.17 (September 21, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed CleverDict with variable of negative index value\nImplement supports_add_constrained_variable for MockOptimizer","category":"page"},{"location":"moi/release_notes/#[v0.9.16](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.16)-(September-17,-2020)","page":"Release notes","title":"v0.9.16 (September 17, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Various fixes:\n32-bit support\nCleverDict with abstract value type\nChecks in test suite","category":"page"},{"location":"moi/release_notes/#[v0.9.15](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.15)-(September-14,-2020)","page":"Release notes","title":"v0.9.15 (September 14, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Bridges improvements:\n(R)SOCtoNonConvexQuad bridge\nZeroOne bridge\nUse supports_add_constrained_variable in LazyBridgeOptimizer\nExposed VariableBridgeCost and ConstraintBridgeCost attributes\nPrioritize constraining variables on creation according to these costs\nRefactor bridge debugging\nLarge performance improvements across all submodules\nLots of documentation improvements\nFileFormats improvements:\nUpdate MathOptFormat to v0.5\nFix supported objectives in FileFormats\nTesting improvements:\nAdd name option for basic_constraint_test\nBug fixes and missing methods\nAdd length for iterators\nFix bug with duplicate terms\nFix order of LinearOfConstraintIndices","category":"page"},{"location":"moi/release_notes/#[v0.9.14](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.14)-(May-30,-2020)","page":"Release notes","title":"v0.9.14 (May 30, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add a solver-independent interface for accessing the set of conflicting constraints an Irreducible Inconsistent Subsystem (#1056).\nBump JSONSchema dependency from v0.2 to v0.3 (#1090).\nDocumentation improvements:\nFix typos (#1054, #1060, #1061, #1064, #1069, #1070).\nRemove the outdated recommendation for a package implementing MOI for a solver XXX to be called MathOptInterfaceXXX (#1087).\nUtilities improvements:\nFix is_canonical for quadratic functions (#1081, #1089).\nImplement add_constrained_variable[s] for CachingOptimizer so that it is added as constrained variables to the underlying optimizer (#1084).\nAdd support for custom objective functions for UniversalFallback (#1086).\nDeterministic ordering of constraints in UniversalFallback (#1088).\nTesting improvements:\nAdd NormOneCone/NormInfinityCone tests (#1045).\nBridges improvements:\nAdd bridges from Semiinteger and Semicontinuous (#1059).\nImplement getting ConstraintSet for Variable.FlipSignBridge (#1066).\nFix setting ConstraintFunction for Constraint.ScalarizeBridge (#1093).\nFix NormOne/NormInf bridges with nonzero constants (#1045).\nFix StackOverflow in debug (#1063).\nFileFormats improvements:\n[SDPA] Implement the extension for integer variables (#1079).\n[SDPA] Ignore comments after m and nblocks and detect dat-s extension (#1077).\n[SDPA] No scaling of off-diagonal coefficient (#1076).\n[SDPA] Add missing negation of constant (#1075).","category":"page"},{"location":"moi/release_notes/#[v0.9.13](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.13)-(March-24,-2020)","page":"Release notes","title":"v0.9.13 (March 24, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added tests for Semicontinuous and Semiinteger variables (#1033).\nAdded tests for using ExprGraphs from NLP evaluators (#1043).\nUpdate version compatibilities of dependencies (#1034, #1051, #1052).\nFixed typos in documentation (#1044).","category":"page"},{"location":"moi/release_notes/#[v0.9.12](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.12)-(February-28,-2020)","page":"Release notes","title":"v0.9.12 (February 28, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed writing NLPBlock in MathOptFormat (#1037).\nFixed MockOptimizer for result attributes with non-one result index (#1039).\nUpdated test template with instantiate (#1032).","category":"page"},{"location":"moi/release_notes/#[v0.9.11](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.11)-(February-21,-2020)","page":"Release notes","title":"v0.9.11 (February 21, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add an option for the model created by Utilities.@model to be a subtype of AbstractOptimizer (#1031).\nDescribed dual cone in docstrings of GeoMeanCone and RelativeEntropyCone (#1018, #1028).\nFixed typos in documentation (#1022, #1024).\nFixed warning of unsupported attribute (#1027).\nAdded more rootdet/logdet conic tests (#1026).\nImplemented ConstraintDual for Constraint.GeoMeanBridge, Constraint.RootDetBridge and Constraint.LogDetBridge and test duals in tests with GeoMeanCone and RootDetConeTriangle and LogDetConeTriangle cones (#1025, #1026).","category":"page"},{"location":"moi/release_notes/#[v0.9.10](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.10)-(January-31,-2020)","page":"Release notes","title":"v0.9.10 (January 31, 2020)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added OptimizerWithAttributes grouping an optimizer constructor and a list of optimizer attributes (#1008).\nAdded RelativeEntropyCone with corresponding bridge into exponential cone constraints (#993).\nAdded NormSpectralCone and NormNuclearCone with corresponding bridges into positive semidefinite constraints (#976).\nAdded supports_constrained_variable(s) (#1004).\nAdded dual_set_type (#1002).\nAdded tests for vector specialized version of delete (#989, #1011).\nAdded PSD3 test (#1007).\nClarified dual solution of Tests.pow1v and Tests.pow1f (#1013).\nAdded support for EqualTo and Zero in Bridges.Constraint.SplitIntervalBridge (#1005).\nFixed Utilities.vectorize for empty vector (#1003).\nFixed free variables in LP writer (#1006).","category":"page"},{"location":"moi/release_notes/#[v0.9.9](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.9)-(December-29,-2019)","page":"Release notes","title":"v0.9.9 (December 29, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Incorporated MathOptFormat.jl as the FileFormats submodule. FileFormats provides readers and writers for a number of standard file formats and MOF, a file format specialized for MOI (#969).\nImproved performance of deletion of vector of variables in MOI.Utilities.Model (#983).\nUpdated to MutableArithmetics v0.2 (#981).\nAdded MutableArithmetics.promote_operation allocation tests (#975).\nFixed inference issue on Julia v1.1 (#982).","category":"page"},{"location":"moi/release_notes/#[v0.9.8](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.8)-(December-19,-2019)","page":"Release notes","title":"v0.9.8 (December 19, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Implemented MutableArithmetics API (#924).\nFixed callbacks with CachingOptimizer (#959).\nFixed MOI.dimension for MOI.Complements (#948).\nAdded fallback for add_variables (#972).\nAdded is_diagonal_vectorized_index utility (#965).\nImproved linear constraints display in manual (#963, #964).\nBridges improvements:\nAdded IndicatorSet to SOS1 bridge (#877).\nAdded support for starting values for Variable.VectorizeBridge (#944).\nFixed MOI.add_constraints with non-bridged variable constraint on bridged variable (#951).\nFixed corner cases and docstring of GeoMeanBridge (#961, #962, #966).\nFixed choice between variable or constraint bridges for constrained variables (#973).\nImprove performance of bridge shortest path (#945, #946, #956).\nAdded docstring for test_delete_bridge (#954).\nAdded Variable bridge tests (#952).","category":"page"},{"location":"moi/release_notes/#[v0.9.7](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.7)-(October-30,-2019)","page":"Release notes","title":"v0.9.7 (October 30, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Implemented _result_index_field for NLPBlockDual (#934).\nFixed copy of model with starting values for vector constraints (#941).\nBridges improvements:\nImproved performance of add_bridge and added has_bridge (#935).\nAdded AbstractSetMapBridge for bridges between sets S1, S2 such that there is a linear map A such that A*S1 = S2 (#933).\nAdded support for starting values for FlipSignBridge, VectorizeBridge, ScalarizeBridge, SlackBridge, SplitIntervalBridge, RSOCBridge, SOCRBridge NormInfinityBridge, SOCtoPSDBridge and RSOCtoPSDBridge (#933, #936, #937, #938, #939).","category":"page"},{"location":"moi/release_notes/#[v0.9.6](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.6)-(October-25,-2019)","page":"Release notes","title":"v0.9.6 (October 25, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added complementarity constraints (#913).\nAllowed ModelLike objects as value of attributes (#928).\nTesting improvements:\nAdded dual_objective_value option to MOI.Test.TestConfig (#922).\nAdded InvalidIndex tests in basic_constraint_tests (#921).\nAdded tests for the constant term in indicator constraint (#929).\nBridges improvements:\nAdded support for starting values for Functionize bridges (#923).\nAdded variable indices context to variable bridges (#920).\nFixed a typo in printing o debug_supports (#927).","category":"page"},{"location":"moi/release_notes/#[v0.9.5](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.5)-(October-9,-2019)","page":"Release notes","title":"v0.9.5 (October 9, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Clarified PrimalStatus/DualStatus to be NO_SOLUTION if result_index is out of bounds (#912).\nAdded tolerance for checks  and use ResultCount + 1 for the result_index in MOI.Test.solve_result_status (#910, #917).\nUse 0.5 instead of 2.0 for power in PowerCone in basic_constraint_test (#916).\nBridges improvements:\nAdded debug utilities for unsupported variable/constraint/objective (#861).\nFixed deletion of variables in bridged VectorOfVariables constraints (#909).\nFixed result_index with objective bridges (#911).","category":"page"},{"location":"moi/release_notes/#[v0.9.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.4)-(October-2,-2019)","page":"Release notes","title":"v0.9.4 (October 2, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Added solver-independent MIP callbacks (#782).\nImplements submit for Utilities.CachingOptimizer and Bridges.AbstractBridgeOptimizer (#906).\nAdded tests for result count of solution attributes (#901, #904).\nAdded NumberOfThreads attribute (#892).\nAdded Utilities.get_bounds to get the bounds on a variable (#890).\nAdded a note on duplicate coefficients in documentation (#581).\nAdded result index in ConstraintBasisStatus (#898).\nAdded extension dictionary to Utilities.Model (#884, #895).\nFixed deletion of constrained variables for CachingOptimizer (#905).\nImplemented Utilities.shift_constraint for Test.UnknownScalarSet (#896).\nBridges improvements:\nAdded Variable.RSOCtoSOCBridge (#907).\nImplemented MOI.get for ConstraintFunction/ConstraintSet for Bridges.Constraint.SquareBridge (#899).","category":"page"},{"location":"moi/release_notes/#[v0.9.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.3)-(September-20,-2019)","page":"Release notes","title":"v0.9.3 (September 20, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixed ambiguity detected in Julia v1.3 (#891, #893).\nFixed missing sets from ListOfSupportedConstraints (#880).\nFixed copy of VectorOfVariables constraints with duplicate indices (#886).\nAdded extension dictionary to MOIU.Model (#884).\nImplemented MOI.get for function and set for GeoMeanBridge (#888).\nUpdated documentation for SingleVariable indices and bridges (#885).\nTesting improvements:\nAdded more comprehensive tests for names (#882).\nAdded tests for SingleVariable duals (#883).\nAdded tests for DualExponentialCone and DualPowerCone (#873).\nImprovements for arbitrary coefficient type:\nFixed == for sets with mutable fields (#887).\nRemoved some Float64 assumptions in bridges (#878).\nAutomatic selection of Constraint.[Scalar|Vector]FunctionizeBridge (#889).","category":"page"},{"location":"moi/release_notes/#[v0.9.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.2)-(September-5,-2019)","page":"Release notes","title":"v0.9.2 (September 5, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Implemented model printing for MOI.ModelLike and specialized it for models defined in MOI (864).\nGeneralized contlinear tests for arbitrary coefficient type (#855).\nFixed supports_constraint for Semiinteger and Semicontinuous and supports for ObjectiveFunction (#859).\nFixed Allocate-Load copy for single variable constraints (#856).\nBridges improvements:\nAdd objective bridges (#789).\nFixed Variable.RSOCtoPSDBridge for dimension 2 (#869).\nAdded Variable.SOCtoRSOCBridge (#865).\nAdded Constraint.SOCRBridge and disable MOI.Bridges.Constraint.SOCtoPSDBridge (#751).\nFixed added_constraint_types for Contraint.LogDetBridge and Constraint.RootDetBridge (#870).","category":"page"},{"location":"moi/release_notes/#[v0.9.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.1)-(August-22,-2019)","page":"Release notes","title":"v0.9.1 (August 22, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fix support for Julia v1.2 (#834).\nL1 and L∞ norm epigraph cones and corresponding bridges to LP were added (#818).\nAdded tests to MOI.Test.nametest (#833).\nFix MOI.Test.soc3test for solvers not supporting infeasibility certificates (#839).\nImplements operate for operators * and / between vector function and constant (#837).\nImplements show for MOI.Utilities.IndexMap (#847).\nFix corner cases for mapping of variables in MOI.Utilities.CachingOptimizer and substitution of variables in MOI.Bridges.AbstractBridgeOptimizer (#848).\nFix transformation of constant terms for MOI.Bridges.Constraint.SOCtoPSDBridge and MOI.Bridges.Constraint.RSOCtoPSDBridge (#840).","category":"page"},{"location":"moi/release_notes/#[v0.9.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.9.0)-(August-13,-2019)","page":"Release notes","title":"v0.9.0 (August 13, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Support for Julia v0.6 and v0.7 was dropped (#714, #717).\nA MOI.Utilities.Model implementation of ModelLike, this should replace most use cases of MOI.Utilities.@model (#781).\nadd_constrained_variable and add_constrained_variables were added (#759).\nSupport for indicator constraints was added (#709, #712).\nDualObjectiveValue attribute was added (#473).\nRawParameter attribute was added (#733).\nA dual_set function was added (#804).\nA Benchmarks submodule was added to facilitate solver benchmarking (#769).\nA submit function was added, this may for instance allow the user to submit solutions or cuts to the solver from a callback (#775).\nThe field of ObjectiveValue was renamed to result_index (#729).\nThe _constant and Utilities.getconstant function were renamed to constant\nREDUCTION_CERTIFICATE result status was added (#734).\nAbstract matrix sets were added (#731).\nTesting improvements:\nThe testing guideline was updated (#728).\nQuadratic tests were added (#697).\nUnit tests for RawStatusString, SolveTime, Silent and SolverName were added (#726, #741).\nA rotated second-order cone test was added (#759).\nA power cone test was added (#768).\nTests for ZeroOne variables with variable bounds were added (#772).\nAn unbounded test was added (#773).\nExisting tests had a few updates (#702, #703, #763).\nDocumentation improvements:\nAdded a section on CachingOptimizer (#777).\nAdded a section on UniversalFallback, Model and @model (#762).\nTransition the knapsack example to a doctest with MockOptimizer (#786).\nUtilities improvements:\nA CleverDict utility was added for a vector that automatically transform into a dictionary once a first index is removed (#767).\nThe Utilities.constant function was renamed to Utilities.constant_vector (#740).\nImplement optimizer attributes for CachingOptimizer (#745).\nRename Utilities.add_scalar_constraint to Utilities.normalize_and_add_constraint (#801).\noperate with vcat, SingleVariable and VectorOfVariables now returns a VectorOfVariables (#616).\nFix a type piracy of operate (#784).\nThe load_constraint fallback signature was fixed (#760).\nThe set_dot function was extended to work with sparse arrays (#805).\nBridges improvements:\nThe bridges no longer store the constraint function and set before it is bridged, the bridges now have to implement ConstraintFunction and ConstraintSet if the user wants to recover them. As a consequence, the @bridge macro was removed (#722).\nBridge are now instantiated with a bridge_constraint function instead of using a constructor (#730).\nFix constraint attributes for bridges (#699).\nConstraint bridges were moved to the Bridges/Constraint submodule so they should now inherit from MOI.Bridges.Constraint.Abstract and should implement MOI.Bridges.Constraint.concrete_bridge_type instead of MOI.Bridges.concrete_bridge_type (#756).\nVariable bridges were added in (#759).\nVarious improvements (#746, #747).","category":"page"},{"location":"moi/release_notes/#[v0.8.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.4)-(March-13,-2019)","page":"Release notes","title":"v0.8.4 (March 13, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Performance improvement in default_copy_to and bridge optimizer (#696).\nAdd Silent and implement setting optimizer attributes in caching and mock optimizers (#695).\nAdd Functionize bridges (SingleVariable and VectorOfVariables) (#659).\nMinor typo fixes (#694).","category":"page"},{"location":"moi/release_notes/#[v0.8.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.3)-(March-6,-2019)","page":"Release notes","title":"v0.8.3 (March 6, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Use zero constant in scalar constraint function of MOI.Test.copytest (#691).\nFix variable deletion with SingleVariable objective function (#690).\nFix LazyBridgeOptimizer with bridges that add no constraints (#689).\nError message improvements (#673, #685, #686, #688).\nDocumentation improvements (#682, #683, #687).\nBasis status:\nRemove VariableBasisStatus (#679).\nTest ConstraintBasisStatus and implement it in bridges (#678).\nFix inference of NumberOfVariables and NumberOfConstraints (#677).\nImplement division between a quadratic function and a number (#675).","category":"page"},{"location":"moi/release_notes/#[v0.8.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.2)-(February-7,-2019)","page":"Release notes","title":"v0.8.2 (February 7, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add RawStatusString attribute (#629).\nDo not set names to the optimizer but only to the cache in CachingOptimizer (#638).\nMake scalar MOI functions act as scalars in broadcast (#646).\nAdd function utilities:\nImplement Base.zero (#634), Base.iszero (#643), add missing arithmetic operations (#644, #645) and fix division (#648).\nAdd a vectorize function that turns a vector of ScalarAffineFunction into a VectorAffineFunction (#642).\nImprove support for starting values:\nShow a warning in copy when starting values are not supported instead of throwing an error (#630).\nFix UniversalFallback for getting an variable or constraint attribute set to no indices (#623).\nAdd a test in contlineartest with partially set VariablePrimalStart.\nBridges improvements:\nFix StackOverFlow in LazyBridgeOptimizer when there is a cycle in the graph of bridges.\nAdd Slack bridges (#610, #650).\nAdd FlipSign bridges (#658).\nAdd tests with duplicate coefficients in ScalarAffineFunction and VectorAffineFunction (#639).\nUse tolerance to compare VariablePrimal in rotatedsoc1 test (#632).\nUse a zero constant in ScalarAffineFunction of constraints in psdt2 (#622).","category":"page"},{"location":"moi/release_notes/#[v0.8.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.1)-(January-7,-2019)","page":"Release notes","title":"v0.8.1 (January 7, 2019)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Adding an NLP objective now overrides any objective set using the ObjectiveFunction attribute (#619).\nRename fullbridgeoptimizer into full_bridge_optimizer (#621).\nAllow custom constraint types with full_bridge_optimizer (#617).\nAdd Vectorize bridge which transforms scalar linear constraints into vector linear constraints (#615).","category":"page"},{"location":"moi/release_notes/#[v0.8.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.8.0)-(December-18,-2018)","page":"Release notes","title":"v0.8.0 (December 18, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Rename all enum values to follow the JuMP naming guidelines for constants, for example, Optimal becomes OPTIMAL, and DualInfeasible becomes DUAL_INFEASIBLE.\nRename CachingOptimizer methods for style compliance.\nAdd an MOI.TerminationStatusCode called ALMOST_DUAL_INFEASIBLE.","category":"page"},{"location":"moi/release_notes/#[v0.7.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.7.0)-(December-13,-2018)","page":"Release notes","title":"v0.7.0 (December 13, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Test that MOI.TerminationStatus is MOI.OptimizeNotCalled before MOI.optimize! is called.\nCheck supports_default_copy_to in tests (#594).\nKey pieces of information like optimality, infeasibility, etc., are now reported through TerminationStatusCode. It is typically no longer necessary to check the result statuses in addition to the termination status.\nAdd perspective dimension to log-det cone (#593).","category":"page"},{"location":"moi/release_notes/#[v0.6.4](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.4)-(November-27,-2018)","page":"Release notes","title":"v0.6.4 (November 27, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Add OptimizeNotCalled termination status (#577) and improve documentation of other statuses (#575).\nAdd a solver naming guideline (#578).\nMake FeasibilitySense the default ObjectiveSense (#579).\nFix Utilities.@model and Bridges.@bridge macros for functions and sets defined outside MOI (#582).\nDocument solver-specific attributes (#580) and implement them in Utilities.CachingOptimizer (#565).","category":"page"},{"location":"moi/release_notes/#[v0.6.3](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.3)-(November-16,-2018)","page":"Release notes","title":"v0.6.3 (November 16, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Variables and constraints are now allowed to have duplicate names. An error is thrown only on lookup. This change breaks some existing tests. (#549)\nAttributes may now be partially set (some values could be nothing). (#563)\nPerformance improvements in Utilities.Model (#549, #567, #568)\nFix bug in QuadtoSOC (#558).\nNew supports_default_copy_to method that optimizers should implement to control caching behavior.\nDocumentation improvements.","category":"page"},{"location":"moi/release_notes/#[v0.6.2](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.2)-(October-26,-2018)","page":"Release notes","title":"v0.6.2 (October 26, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Improve hygiene of @model macro (#544).\nFix bug in copy tests (#543).\nFix bug in UniversalFallback attribute getter (#540).\nAllow all correct solutions for solve_blank_obj unit test (#537).\nAdd errors for Allocate-Load and bad constraints (#534).\n[performance] Add specialized implementation of hash for VariableIndex (#533).\n[performance] Construct the name to object dictionaries lazily in model (#535).\nAdd the QuadtoSOC bridge which transforms ScalarQuadraticFunction constraints into RotatedSecondOrderCone (#483).","category":"page"},{"location":"moi/release_notes/#[v0.6.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.1)-(September-22,-2018)","page":"Release notes","title":"v0.6.1 (September 22, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Enable PositiveSemidefiniteConeSquare set and quadratic functions in MOIB.fullbridgeoptimizer (#524).\nAdd warning in the bridge between PositiveSemidefiniteConeSquare and PositiveSemidefiniteConeTriangle when the matrix is almost symmetric (#522).\nModify MOIT.copytest to not add multiples constraints on the same variable (#521).\nAdd missing keyword argument in one of MOIU.add_scalar_constraint methods (#520).","category":"page"},{"location":"moi/release_notes/#[v0.6.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.6.0)-(August-30,-2018)","page":"Release notes","title":"v0.6.0 (August 30, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"The MOIU.@model and MOIB.@bridge macros now support functions and sets defined in external modules. As a consequence, function and set names in the macro arguments need to be prefixed by module name.\nRename functions according to the JuMP style guide:\ncopy! with keyword arguments copynames and warnattributes -> copy_to with keyword arguments copy_names and warn_attributes;\nset! -> set;\naddvariable[s]! -> add_variable[s];\nsupportsconstraint -> supports_constraint;\naddconstraint[s]! -> add_constraint[s];\nisvalid -> is_valid;\nisempty -> is_empty;\nBase.delete! -> delete;\nmodify! -> modify;\ntransform! -> transform;\ninitialize! -> initialize;\nwrite -> write_to_file; and\nread! -> read_from_file.\nRemove free! (use Base.finalize instead).\nAdd the SquarePSD bridge which transforms PositiveSemidefiniteConeTriangle constraints into PositiveSemidefiniteConeTriangle.\nAdd result fallback for ConstraintDual of variable-wise constraint, ConstraintPrimal and ObjectiveValue.\nAdd tests for ObjectiveBound.\nAdd test for empty rows in vector linear constraint.\nRework errors: CannotError has been renamed NotAllowedError and the distinction between UnsupportedError and NotAllowedError is now about whether the element is not supported (for example, it cannot be copied a model containing this element) or the operation is not allowed (either because it is not implemented, because it cannot be performed in the current state of the model, or because it cannot be performed for a specific index)\ncanget is removed. NoSolution is added as a result status to indicate that the solver does not have either a primal or dual solution available (See #479).","category":"page"},{"location":"moi/release_notes/#[v0.5.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.5.0)-(August-5,-2018)","page":"Release notes","title":"v0.5.0 (August 5, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fix names with CachingOptimizer.\nCleanup thanks to @mohamed82008.\nAdded a universal fallback for constraints.\nFast utilities for function canonicalization thanks to @rdeits.\nRenamed dimension field to side_dimension in the context of matrix-like sets.\nNew and improved tests for cases like duplicate terms and ObjectiveBound.\nRemoved cantransform, canaddconstraint, canaddvariable, canset, canmodify, and candelete functions from the API. They are replaced by a new set of errors that are thrown: Subtypes of UnsupportedError indicate unsupported operations, while subtypes of CannotError indicate operations that cannot be performed in the current state.\nThe API for copy! is updated to remove the CopyResult type.\nUpdates for the new JuMP style guide.","category":"page"},{"location":"moi/release_notes/#[v0.4.1](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.1)-(June-28,-2018)","page":"Release notes","title":"v0.4.1 (June 28, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Fixes vector function modification on 32 bits.\nFixes Bellman-Ford algorithm for bridges.\nAdded an NLP test with FeasibilitySense.\nUpdate modification documentation.","category":"page"},{"location":"moi/release_notes/#[v0.4.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.4.0)-(June-23,-2018)","page":"Release notes","title":"v0.4.0 (June 23, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Helper constructors for VectorAffineTerm and VectorQuadraticTerm.\nAdded modify_lhs to TestConfig.\nAdditional unit tests for optimizers.\nAdded a type parameter to CachingOptimizer for the optimizer field.\nNew API for problem modification (#388)\nTests pass without deprecation warnings on Julia 0.7.\nSmall fixes and documentation updates.","category":"page"},{"location":"moi/release_notes/#[v0.3.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.3.0)-(May-25,-2018)","page":"Release notes","title":"v0.3.0 (May 25, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Functions have been redefined to use arrays-of-structs instead of structs-of-arrays.\nImprovements to MockOptimizer.\nSignificant changes to Bridges.\nNew and improved unit tests.\nFixes for Julia 0.7.","category":"page"},{"location":"moi/release_notes/#[v0.2.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.2.0)-(April-24,-2018)","page":"Release notes","title":"v0.2.0 (April 24, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Improvements to and better coverage of Tests.\nDocumentation fixes.\nSolverName attribute.\nChanges to the NLP interface (new definition of variable order and arrays of structs for bound pairs and sparsity patterns).\nAddition of NLP tests.\nIntroduction of UniversalFallback.\ncopynames keyword argument to MOI.copy!.\nAdd Bridges submodule.","category":"page"},{"location":"moi/release_notes/#[v0.1.0](https://github.com/jump-dev/MathOptInterface.jl/releases/tag/v0.1.0)-(February-28,-2018)","page":"Release notes","title":"v0.1.0 (February 28, 2018)","text":"","category":"section"},{"location":"moi/release_notes/","page":"Release notes","title":"Release notes","text":"Initial public release.\nThe framework for MOI was developed at the JuMP-dev workshop at MIT in June 2017 as a sorely needed replacement for MathProgBase.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"developers/extensions/#extensions_manual","page":"Extensions","title":"Extensions","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"JuMP provides a variety of ways to extend the basic modeling functionality.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"tip: Tip\nThis documentation in this section is still a work-in-progress. The best place to look for ideas and help when writing a new JuMP extension are existing JuMP extensions. Examples include:BilevelJuMP.jl\nColuna.jl\nInfiniteOpt.jl\nPlasmo.jl\nPolyJuMP.jl\nSDDP.jl\nStochasticPrograms.jl\nSumOfSquares.jl\nvOptGeneric.jl","category":"page"},{"location":"developers/extensions/#Compatibility","page":"Extensions","title":"Compatibility","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"When writing JuMP extensions, you should carefully consider the compatibility guarantees that JuMP makes. In particular:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"All functions, structs, and constants which do not begin with an underscore (_) are public. These are always safe to use, and they should all have corresponding documentation.\nAll identifiers beginning with an underscore (_) are private. These are not safe to use, because they may break in any JuMP release, including patch releases.\nUnless explicitly mentioned in the documentation, all fields of a struct are private. These are not safe to use, because they may break in any JuMP release, including patch releases. An example of a field which is safe to use is the model.ext extension dictionary, which is documented in The extension dictionary.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"In general, we strongly encourage you to use only the public API of JuMP. If you are missing a feature, please open a GitHub issue.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"However, if you do use the private API (for example, because your feature request has not been implemented yet), then you must carefully restrict the versions of JuMP that your package is compatible with in the Project.toml file. The easiest way to do this is via the hyphen specifiers. For example, if your package supports all JuMP versions between v1.0.0 and v1.1.1, do:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"JuMP = \"1.0.0 - 1.1.1\"","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Then, whenever JuMP releases a new version, you should check if your package is still compatible and update the bound accordingly.","category":"page"},{"location":"developers/extensions/#Define-a-new-set","page":"Extensions","title":"Define a new set","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"To define a new set for JuMP, subtype MOI.AbstractScalarSet or MOI.AbstractVectorSet and implement Base.copy for the set.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> struct NewMOIVectorSet <: MOI.AbstractVectorSet\n           dimension::Int\n       end\n\njulia> Base.copy(x::NewMOIVectorSet) = x\n\njulia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @constraint(model, x in NewMOIVectorSet(2))\n[x[1], x[2]] ∈ NewMOIVectorSet(2)","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"However, for vector-sets, this requires the user to specify the dimension argument to their set, even though we could infer it from the length of x!","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"You can make a more user-friendly set by subtyping AbstractVectorSet and implementing moi_set.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> struct NewVectorSet <: JuMP.AbstractVectorSet end\n\njulia> JuMP.moi_set(::NewVectorSet, dim::Int) = NewMOIVectorSet(dim)\n\njulia> @constraint(model, x in NewVectorSet())\n[x[1], x[2]] ∈ NewMOIVectorSet(2)","category":"page"},{"location":"developers/extensions/#extend_variable_macro","page":"Extensions","title":"Extend @variable","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Just as Bin and Int create binary and integer variables, you can extend the @variable macro to create new types of variables. Here is an explanation by example, where we create a AddTwice type, that creates a tuple of two JuMP variables instead of a single variable.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"First, create a new struct. This can be anything. Our struct holds a VariableInfo object that stores bound information, and whether the variable is binary or integer.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> struct AddTwice\n           info::JuMP.VariableInfo\n       end","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Second, implement build_variable, which takes ::Type{AddTwice} as an argument, and returns an instance of AddTwice. Note that you can also receive keyword arguments.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> function JuMP.build_variable(\n           _err::Function,\n           info::JuMP.VariableInfo,\n           ::Type{AddTwice};\n           kwargs...\n       )\n           println(\"Can also use $kwargs here.\")\n           return AddTwice(info)\n       end","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Third, implement add_variable, which takes the instance of AddTwice from the previous step, and returns something. Typically, you will want to call add_variable here. For example, our AddTwice call is going to add two JuMP variables.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> function JuMP.add_variable(\n           model::JuMP.Model,\n           duplicate::AddTwice,\n           name::String,\n       )\n           a = JuMP.add_variable(\n               model,\n               JuMP.ScalarVariable(duplicate.info),\n               \"$(name)_a\",\n            )\n           b = JuMP.add_variable(\n               model,\n               JuMP.ScalarVariable(duplicate.info),\n               \"$(name)_b\",\n            )\n           return (a, b)\n       end","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Now AddTwice can be passed to @variable similar to Bin or Int, or through the variable_type keyword. However, now it adds two variables instead of one.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:2], variable_type = AddTwice, kw = i)\nCan also use Base.Pairs(:kw => 1) here.\nCan also use Base.Pairs(:kw => 2) here.\n2-element Vector{Tuple{VariableRef, VariableRef}}:\n (x[1]_a, x[1]_b)\n (x[2]_a, x[2]_b)\n\njulia> num_variables(model)\n4\n\njulia> first(x[1])\nx[1]_a\n\njulia> last(x[2])\nx[2]_b","category":"page"},{"location":"developers/extensions/#Extend-[@constraint](@ref)","page":"Extensions","title":"Extend @constraint","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"The @constraint macro has three steps that can be intercepted and extended: parse time, build time, and add time.","category":"page"},{"location":"developers/extensions/#Parse","page":"Extensions","title":"Parse","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"To extend the @constraint macro at parse time, implement one of the following methods:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"parse_constraint_head\nparse_constraint_call","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"warning: Warning\nExtending the constraint macro at parse time is an advanced operation and has the potential to interfere with existing JuMP syntax. Please discuss with the developer chatroom before publishing any code that implements these methods.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"parse_constraint_head should be implemented to intercept an expression based on the .head field of Base.Expr. For example:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> using JuMP\n\njulia> const MutableArithmetics = JuMP._MA;\n\njulia> model = Model(); @variable(model, x);\n\njulia> function JuMP.parse_constraint_head(\n           _error::Function,\n           ::Val{:(:=)},\n           lhs,\n           rhs,\n       )\n           println(\"Rewriting := as ==\")\n           new_lhs, parse_code = MutableArithmetics.rewrite(lhs)\n           build_code = :(\n               build_constraint($(_error), $(new_lhs), MOI.EqualTo($(rhs)))\n           )\n           return false, parse_code, build_code\n       end\n\njulia> @constraint(model, x + x := 1.0)\nRewriting := as ==\n2 x = 1","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"parse_constraint_call should be implemented to intercept an expression of the form Expr(:call, op, args...). For example:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> using JuMP\n\njulia> const MutableArithmetics = JuMP._MA;\n\njulia> model = Model(); @variable(model, x);\n\njulia> function JuMP.parse_constraint_call(\n           _error::Function,\n           is_vectorized::Bool,\n           ::Val{:my_equal_to},\n           lhs,\n           rhs,\n       )\n           println(\"Rewriting my_equal_to to ==\")\n           new_lhs, parse_code = MutableArithmetics.rewrite(lhs)\n           build_code = if is_vectorized\n               :(build_constraint($(_error), $(new_lhs), MOI.EqualTo($(rhs)))\n           )\n           else\n               :(build_constraint.($(_error), $(new_lhs), MOI.EqualTo($(rhs))))\n           end\n           return parse_code, build_code\n       end\n\njulia> @constraint(model, my_equal_to(x + x, 1.0))\nRewriting my_equal_to to ==\n2 x = 1","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"tip: Tip\nWhen parsing a constraint you can recurse into sub-constraint (for example, the {expr} in z => {x <= 1}) by calling parse_constraint.","category":"page"},{"location":"developers/extensions/#Build","page":"Extensions","title":"Build","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"To extend the @constraint macro at build time, implement a new build_constraint method.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"This may mean implementing a method for a specific function or set created at parse time, or it may mean implementing a method which handles additional positional arguments.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"build_constraint must return an AbstractConstraint, which can either be an AbstractConstraint already supported by JuMP, for example, ScalarConstraint or VectorConstraint, or a custom AbstractConstraint with a corresponding add_constraint method (see Add).","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"tip: Tip\nThe easiest way to extend @constraint is via an additional positional argument to build_constraint.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Here is an example of adding extra arguments to build_constraint:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> model = Model(); @variable(model, x);\n\njulia> struct MyConstrType end\n\njulia> function JuMP.build_constraint(\n            _error::Function,\n            f::JuMP.GenericAffExpr,\n            set::MOI.EqualTo,\n            extra::Type{MyConstrType};\n            d = 0,\n       )\n            new_set = MOI.LessThan(set.value + d)\n            return JuMP.build_constraint(_error, f, new_set)\n       end\n\njulia> @constraint(model, my_con, x == 0, MyConstrType, d = 2)\nmy_con : x ≤ 2","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"note: Note\nOnly a single positional argument can be given to a particular constraint. Extensions that seek to pass multiple arguments (for example, Foo and Bar) should combine them into one argument type (for example, FooBar).","category":"page"},{"location":"developers/extensions/#extension_add_constraint","page":"Extensions","title":"Add","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"build_constraint returns an AbstractConstraint object. To extend @constraint at add time, define a subtype of AbstractConstraint, implement build_constraint to return an instance of the new type, and then implement add_constraint.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Here is an example:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> model = Model(); @variable(model, x);\n\njulia> struct MyTag\n           name::String\n       end\n\njulia> struct MyConstraint{S} <: AbstractConstraint\n           name::String\n           f::AffExpr\n           s::S\n       end\n\njulia> function JuMP.build_constraint(\n            _error::Function,\n            f::AffExpr,\n            set::MOI.AbstractScalarSet,\n            extra::MyTag,\n       )\n            return MyConstraint(extra.name, f, set)\n       end\n\njulia> function JuMP.add_constraint(\n            model::Model,\n            con::MyConstraint,\n            name::String,\n       )\n            return add_constraint(\n                model,\n                ScalarConstraint(con.f, con.s),\n                \"$(con.name)[$(name)]\",\n            )\n       end\n\njulia> @constraint(model, my_con, 2x <= 1, MyTag(\"my_prefix\"))\nmy_prefix[my_con] : 2 x - 1 ≤ 0","category":"page"},{"location":"developers/extensions/#The-extension-dictionary","page":"Extensions","title":"The extension dictionary","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Every JuMP model has a field .ext::Dict{Symbol,Any} that can be used by extensions. This is useful if your extensions to @variable and @constraint need to store information between calls.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"The most common way to initialize a model with information in the .ext dictionary is to provide a new constructor:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> function MyModel()\n           model = Model()\n           model.ext[:MyModel] = 1\n           return model\n       end\nMyModel (generic function with 1 method)\n\njulia> model = MyModel()\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\n\njulia> model.ext\nDict{Symbol, Any} with 1 entry:\n  :MyModel => 1","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"If you define extension data, implement copy_extension_data to support copy_model.","category":"page"},{"location":"developers/extensions/#Defining-new-JuMP-models","page":"Extensions","title":"Defining new JuMP models","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"If extending individual calls to @variable and @constraint is not sufficient, it is possible to implement a new model via a subtype of AbstractModel. You can also define new AbstractVariableRefs to create different types of JuMP variables.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"warning: Warning\nExtending JuMP in this manner is an advanced operation. We strongly encourage you to consider how you can use the methods mentioned in the previous sections to achieve your aims instead of defining new model and variable types. Consult the developer chatroom before starting work on this.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"If you define new types, you will need to implement a considerable number of methods, and doing so will require a detailed understanding of the JuMP internals. Therefore, the list of methods to implement is currently undocumented.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"The easiest way to extend JuMP by defining a new model type is to follow an existing example. A simple example to follow is the JuMPExtension module in the JuMP test suite. The best example of an external JuMP extension that implements an AbstractModel is InfiniteOpt.jl.","category":"page"},{"location":"developers/extensions/#Testing-JuMP-extensions","page":"Extensions","title":"Testing JuMP extensions","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"The JuMP test suite contains a large number of tests for JuMP extensions. You can run these tests by copying the MIT-licensed Kokako.jl file in the JuMP tests into your /test folder, and then adding this snippet to your /test/runtests.jl file:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"using MyJuMPExtension\nimport JuMP\ninclude(\"Kokako.jl\")\nconst MODULES_TO_TEST = Kokako.include_modules_to_test(JuMP)\nKokako.run_tests(\n    MODULES_TO_TEST,\n    MyJuMPExtension.MyModel,\n    MyJuMPExtension.MyVariableRef;\n    test_prefix = \"test_extension_\",\n)","category":"page"},{"location":"developers/extensions/#Set-an-optimize!-hook","page":"Extensions","title":"Set an optimize! hook","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Some extensions require modification to the problem after the user has finished constructing the problem, but before optimize! is called. For these situations, JuMP provides set_optimize_hook, which lets you intercept the optimize! call.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"Here's a simple example of adding an optimize hook that extends optimize! to take a keyword argument silent:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> using JuMP, HiGHS\n\njulia> model = Model(HiGHS.Optimizer);\n\njulia> @variable(model, x >= 1.5, Int);\n\njulia> @objective(model, Min, x);\n\njulia> function silent_hook(model; silent::Bool)\n           if silent\n               set_silent(model)\n           else\n               unset_silent(model)\n           end\n           ## Make sure you set ignore_optimize_hook = true, or we'll\n           ## recursively enter the optimize hook!\n           return optimize!(model; ignore_optimize_hook = true)\n       end\nsilent_hook (generic function with 1 method)\n\njulia> set_optimize_hook(model, silent_hook)\nsilent_hook (generic function with 1 method)\n\njulia> optimize!(model; silent = true)\n\njulia> optimize!(model; silent = false)\nSolution has               num          max          sum\nCol     infeasibilities      0            0            0\nInteger infeasibilities      0            0            0\nRow     infeasibilities      0            0            0\nRow     residuals            0            0            0\nPresolving model\n0 rows, 0 cols, 0 nonzeros\n0 rows, 0 cols, 0 nonzeros\nPresolve: Optimal\n\nSolving report\n  Status            Optimal\n  Primal bound      2\n  Dual bound        2\n  Gap               0% (tolerance: 0.01%)\n  Solution status   feasible\n                    2 (objective)\n                    0 (bound viol.)\n                    0 (int. viol.)\n                    0 (row viol.)\n  Timing            0.00 (total)\n                    0.00 (presolve)\n                    0.00 (postsolve)\n  Nodes             0\n  LP iterations     0 (total)\n                    0 (strong br.)\n                    0 (separation)\n                    0 (heuristics)","category":"page"},{"location":"developers/extensions/#Creating-new-container-types","page":"Extensions","title":"Creating new container types","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"JuMP macros (for example, @variable) accept a container keyword argument to force the type of container that is chosen. By default, JuMP supports container = Array, container = DenseAxisArray, container = SparseAxisArray and container = Auto. You can extend support to user-defined types by implementing Containers.container.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"For example, here is a container that reverses the order of the indices:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"julia> struct Foo end\n\njulia> function Containers.container(f::Function, indices, ::Type{Foo})\n           return reverse([f(i...) for i in indices])\n       end\n\njulia> model = Model();\n\njulia> @variable(model, x[1:3], container = Foo)\n3-element Vector{VariableRef}:\n x[3]\n x[2]\n x[1]\n\njulia> x[1]\nx[3]\n\njulia> @variable(model, y[1:3, 1:2], container = Foo)\n3×2 Matrix{VariableRef}:\n y[3,2]  y[3,1]\n y[2,2]  y[2,1]\n y[1,2]  y[1,1]\n\njulia> y[1, 1]\ny[3,2]\n\njulia> @variable(model, z[i=1:3; isodd(i)], container = Foo)\n2-element Vector{VariableRef}:\n z[3]\n z[1]\n\njulia> z[2]\nz[1]","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"warning: Warning\nIf you are a general user, you should not need to create a new container type. Instead, consider following User-defined containers and create a new container using standard Julia syntax. For example:julia> model = Model();\n\njulia> @variable(model, x[1:3])\n3-element Vector{VariableRef}:\n x[1]\n x[2]\n x[3]\n\njulia> y = reverse(x)\n3-element Vector{VariableRef}:\n x[3]\n x[2]\n x[1]","category":"page"},{"location":"developers/extensions/#Performance-tips-for-extensions","page":"Extensions","title":"Performance tips for extensions","text":"","category":"section"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"The function-in-set design of MathOptInterface causes type stability issues in Julia if you try to iterate over all of the constraints in a model. The easiest way to fix this is to use a function barrier.","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"For example, instead of:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"function all_names_slow(model)\n    names = Set{String}()\n    for ci in all_constraints(model)\n        push!(names, name(ci))\n    end\n    return names\nend","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"use:","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"function _function_barrier(names, model, ::Type{F}, ::Type{S}) where {F,S}\n    for ci in all_constraints(model, F, S)\n        push!(names, name(ci))\n    end\n    return\nend\n\nfunction all_names_fast(model)\n    names = Set{String}()\n    for (F, S) in list_of_constraint_types(model)\n        _function_barrier(names, model, F, S)\n    end\n    return names\nend","category":"page"},{"location":"developers/extensions/","page":"Extensions","title":"Extensions","text":"note: Note\nIt is important to explicitly type the F and S arguments. If you leave them untyped, for example, function _function_barrier(names, model, F, S), Julia will not specialize the function calls and performance will not be improved.","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/changelog.md\"","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"CurrentModule = MathOptInterface","category":"page"},{"location":"moi/changelog/#Release-notes","page":"Release notes","title":"Release notes","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.","category":"page"},{"location":"moi/changelog/#v1.20.0-(September-7,-2023)","page":"Release notes","title":"v1.20.0 (September 7, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added Scaled set (#2237) (#2263) (#2264)\nAdded ObjectiveLimit attribute (#2257)","category":"page"},{"location":"moi/changelog/#Other","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Updated dependencies (#2258)\nImproved performance of ScalarNonlinearFunction utilities (#2259)\nFixed docstrings (#2261)","category":"page"},{"location":"moi/changelog/#v1.19.0-(August-15,-2023)","page":"Release notes","title":"v1.19.0 (August 15, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-2","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added VectorNonlinearFunction (#2201)\nAdded new bridges\nBridges.Constraint.IntegerToZeroOneBridge (#2205)\nBridges.Constraint.ToScalarQuadraticBridge (#2235)\nBridges.Constraint.ToVectorQuadraticBridge (#2235)\nBridges.Constraint.ToScalarNonlinearBridge (#2233) (#2235)\nBridges.Constraint.FunctionConversionBridge (#2235)\nAdded Bridges.bridging_cost for setting a bridge-specific cost in the bridging graph (#2235)\nAdded Utilities.eval_variables support for ScalarNonlinearFunction (#2218) (#2219)\nAdded support for ScalarNonlinearFunction in FileFormats.NL (#2228) (#2231)\nAdded support for writing non-Float64 functions in FileFormats.MOF\nAdded Utilities.lazy_map (#2254)","category":"page"},{"location":"moi/changelog/#Fixed","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed method ambiguities in operate (#2224)\nFixed reading LP file with a -infinity <= x <= +infinity variable (#2225)\nFixed missing require in Test.test_nonlinear_duals (#2230)\nFixed bug in ConstraintPrimal of Bridges.Constraint.QuadtoSOCBridge (#2240)","category":"page"},{"location":"moi/changelog/#Other-2","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added extensions to solver-tests.yml (#2229)\nRefactored test/Benchmarks (#2234)\nFixed warnings in tests (#2241) (#2243)\nSmall refactoring of bridges for upcoming VectorNonlinearFunction (#2244) (#2245)\nFixed various typos (#2251) (#2255)\nPartitioned how we run the tests on GitHub actions (#2252) (#2253)","category":"page"},{"location":"moi/changelog/#v1.18.0-(June-23,-2023)","page":"Release notes","title":"v1.18.0 (June 23, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-3","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added Bridges.Objective.SlackBridgePrimalDualStart (#2194)\nAdded constraint_gradient_structure and eval_constraint_gradient (#2200)","category":"page"},{"location":"moi/changelog/#Fixed-2","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a missing @require in MOI.Test (#2195) (#2196)\nFixed incorrect usage of Utilities.operate! in bridges (#2207) (#2216)\nFixed splatting nonlinear expression with univariate operator (#2221)","category":"page"},{"location":"moi/changelog/#Other-3","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Removed unused argument names (#2199)\nReduced memory requirement for tests (#2204)\nRefactored Utilities.promote_operation (#2206)\nImproved code style in Utilities/mutable_arithmetics.jl (#2209)\nRefactored various methods in Utilities/functions.jl (#2208) (#2212) (#2213) (#2214) (#2215)","category":"page"},{"location":"moi/changelog/#v1.17.1-(June-6,-2023)","page":"Release notes","title":"v1.17.1 (June 6, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-3","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed comparison of ScalarFunctionConstantNotZero (#2190)","category":"page"},{"location":"moi/changelog/#Other-4","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added documentation for enum instances (#2186)\nUpdated chatroom links in documentation (#2188)\nChanged the documentation to build on Julia v1.9 (#2191)","category":"page"},{"location":"moi/changelog/#v1.17.0-(June-1,-2023)","page":"Release notes","title":"v1.17.0 (June 1, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-4","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added ScalarNonlinearFunction (#2059)\nAdded a variety of tests for NormSpectralCone, NormNuclearCone, and HermitianPositiveSemidefiniteConeTriangle (#2174)\nAdded Utilities.value_type for computing the value type of a function (#2176)","category":"page"},{"location":"moi/changelog/#Fixed-4","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed support for external sets in Utilities.loadfromstring! (#2177)\nFixed promote_operation for ScalarNonlinearFunction (#2179)\nFixed two issues in FileFormats.LP when reading files with quadratic functions (#2182) (#2184)","category":"page"},{"location":"moi/changelog/#v1.16.0-(May-16,-2023)","page":"Release notes","title":"v1.16.0 (May 16, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-5","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added support for MathOptFormat v1.3 and v1.4 (#2158) (#2169)\nAdded new method to the nonlinear API (#2162) (#2164)\neval_hessian_constraint\neval_hessian_objective\nhessian_constraint_structure\nhessian_objective_structure\nAdded new sets\nNormCone (#2119)\nScaledPositiveSemidefiniteConeTriangle (#2154)","category":"page"},{"location":"moi/changelog/#Fixed-5","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed support for Julia v1.9 to work around a bug in the upstream Julia compiler (#2161) (#2163)\nFixed a correctness bug in Bridges.Constraint.HermitianToSymmetricPSDBridge (#2171)\nFixed convert(::VariableIndex, ::ScalarAffineFunction) when the function has terms with 0coefficients (#2173)","category":"page"},{"location":"moi/changelog/#Other-5","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed solver-tests.yml (#2157)\nUpdated documentation links to developer chatroom (#2160)\nAdded various tests for bridges (#2156)\nAdded checklists to the developer documentation (#2167) (#2168)","category":"page"},{"location":"moi/changelog/#v1.15.1-(April-25,-2023)","page":"Release notes","title":"v1.15.1 (April 25, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-6","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed deleting a variable in a bridged objective (#2150)","category":"page"},{"location":"moi/changelog/#v1.15.0-(April-19,-2023)","page":"Release notes","title":"v1.15.0 (April 19, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-6","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added Bridges.Objective.VectorFunctionizeBridge (#2139)","category":"page"},{"location":"moi/changelog/#Fixed-7","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed support for Rational in Bridges.Constraint.SplitIntervalBridge (#2137)\nFixed printing of LaTeX models (#2141)\nFixed modify in Bridges.Objective.VectorSlackBridge (#2144)\nFixed NAME record with spaces in FileFormats.MPS (#2146)\nFixed deleting a variable in a bridged objective (#2147)","category":"page"},{"location":"moi/changelog/#Other-6","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add a test for variables in one-sided open Interval sets (#2133)\nMinor style fixes in the source code (#2148)","category":"page"},{"location":"moi/changelog/#v1.14.1-(April-6,-2023)","page":"Release notes","title":"v1.14.1 (April 6, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-8","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in Bridges.print_active_bridges (#2135)","category":"page"},{"location":"moi/changelog/#Other-7","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added a warning when an ambiguous string is passed to exclude in Test.runtests (#2136)","category":"page"},{"location":"moi/changelog/#v1.14.0-(April-4,-2023)","page":"Release notes","title":"v1.14.0 (April 4, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-7","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added support for starting values in Bridges.Constraint.QuadtoSOCBridge (#2115)\nAdded support for Regex in the include and exclude arguments to Test.runtests (#2129)\nAdded Bridges.print_active_bridges methods for individual objectives and constraints (#2128)","category":"page"},{"location":"moi/changelog/#Fixed-9","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed ResultCount when parsing .sol files in FileFormats.NL (#2130)","category":"page"},{"location":"moi/changelog/#v1.13.2-(March-21,-2023)","page":"Release notes","title":"v1.13.2 (March 21, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-10","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed splatting of containers in MOI.Nonlinear (#2120)\nFixed a bug reading LP files with default bounds (#2121)\nFixed a bug in which Bridges.Constraint.HermitianToSymmetricPSDBridge was not enabled by default (#2123)","category":"page"},{"location":"moi/changelog/#Other-8","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed typos in the documentation (#2114)\nFunctions now print to the REPL in algebraic form. This is potentially breaking if you have tests which rely on a specific String form of MOI functions. (#2112) (#2126)","category":"page"},{"location":"moi/changelog/#v1.13.1-(March-3,-2023)","page":"Release notes","title":"v1.13.1 (March 3, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Other-9","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added the Google style guide to the documentation linter Vale, and fixed the resulting warnings (#2110)\nImproved the docstrings in src/functions.jl (#2108)","category":"page"},{"location":"moi/changelog/#v1.13.0-(February-28,-2023)","page":"Release notes","title":"v1.13.0 (February 28, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-8","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added Bridges.Constraint.NumberConversionBridge (#2091)\nAdded Parameter set (#2095) (#2105) (#2106) (#2109)\nAdded with_cache_type argument to instantiate (#2097)\nAdded support for HermitianPositiveSemidefiniteConeTriangle in Utilities.Model (#2100)","category":"page"},{"location":"moi/changelog/#Fixed-11","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed bug when Utilities.@product_of_sets is empty (#2101)\nFixed Bridges.print_active_bridges when variable bridge is an AbstractScalarSet (#2107)","category":"page"},{"location":"moi/changelog/#Other-10","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added tests for vector-valued objective functions in FileFormats.MOF (#2093)\nUsed and documented preference for import MathOptInterface as MOI (#2096)\nFix and test links in the documentation with linkcheck = true (#2098)\nImproved docstrings of sets in src/sets.jl (#2099)\nSkip checking flakey links in documentation with linkcheck_ignore  (#2103)","category":"page"},{"location":"moi/changelog/#v1.12.0-(February-10,-2023)","page":"Release notes","title":"v1.12.0 (February 10, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Added-9","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added support for vector-valued objective functions (#2070)\nAdded a Utilities.distance_to_set method for SecondOrderCone (#2060)","category":"page"},{"location":"moi/changelog/#Fixed-12","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a number of constraint bridges so that Bridges.final_touch can be called multiple times without forcing a rebuild of the reformulation (#2089)","category":"page"},{"location":"moi/changelog/#Other-11","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added new tests that a set ObjectiveFunction appears in ListOfModelAttributesSet (#2085)\nImproved the docstrings of a number of constraint-programming related sets (#2087)","category":"page"},{"location":"moi/changelog/#v1.11.5-(January-24,-2023)","page":"Release notes","title":"v1.11.5 (January 24, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-13","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug writing .lp files with an off-diagonal quadratic objective (#2082)","category":"page"},{"location":"moi/changelog/#Other-12","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added SnoopPrecompile directives for reduced time-to-first-X in Julia v1.9 (#2080)","category":"page"},{"location":"moi/changelog/#v1.11.4-(January-12,-2023)","page":"Release notes","title":"v1.11.4 (January 12, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-14","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug reading .lp files with an Integer section (#2078)","category":"page"},{"location":"moi/changelog/#v1.11.3-(January-12,-2023)","page":"Release notes","title":"v1.11.3 (January 12, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-15","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a performance bug when deleting a vector of constraints (#2072)\nFixed a bug reading .lp files with terms like x -1 y (#2076)","category":"page"},{"location":"moi/changelog/#Other-13","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Documented the two-argument method of optimize! (#2074)","category":"page"},{"location":"moi/changelog/#v1.11.2-(January-2,-2023)","page":"Release notes","title":"v1.11.2 (January 2, 2023)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-16","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug reading .mof.json files with ConstraintName set for VariableIndex constraints (#2066)\nFixed a bug reading .mof.json files with nonlinear objectives and no constraints (#2068)","category":"page"},{"location":"moi/changelog/#v1.11.1-(December-22,-2022)","page":"Release notes","title":"v1.11.1 (December 22, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-17","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug reading .mof.json files with integer coefficients for affine and quadratic functions (#2063)","category":"page"},{"location":"moi/changelog/#v1.11.0-(December-2,-2022)","page":"Release notes","title":"v1.11.0 (December 2, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-10","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added Utilities.PenaltyRelaxation and Utilities.ScalarPenaltyRelaxation (#1995)\nAdded Utilities.distance_to_set (#2048)\nAdded support for ConstraintPrimalStart and ConstraintDualStart in FileFormats.MOF (#2056)","category":"page"},{"location":"moi/changelog/#Other-14","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Tidied these release notes (#2055)","category":"page"},{"location":"moi/changelog/#v1.10.0-(November-22,-2022)","page":"Release notes","title":"v1.10.0 (November 22, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-11","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added new methods set(::OptimizerWithAttributes, ::RawOptimizerAttribute, value) and get(::OptimizerWithAttributes, ::RawOptimizerAttribute) (#2049)\nAdded new methods Utilities.DoubleDicts.outer_keys and Utilities.DoubleDicts.nonempty_outer_keys (#2052)","category":"page"},{"location":"moi/changelog/#Fixed-18","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed Bridges.Objective.SlackBridge when the objective function is complex-valued (#2036) (#2038)\nFixed docstring of Test.runtests to clarify the warn_unsupported argument (#2037)\nFixed reading of free variables in FileFormats.LP (#2044)\nFixed numerous edge cases reading files from QPLIB using FileFormats.LP (#2042) (#2044)\nFixed situations in which x^y returns a complex value in Nonlinear (#2050)","category":"page"},{"location":"moi/changelog/#Other-15","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Improved the error message thrown when a user-defined nonlinear function does not accept splatted input (#2032)\nRemoved specialized iterators for keys and values in Utilities.CleverDicts (#2051)","category":"page"},{"location":"moi/changelog/#v1.9.0-(October-29,-2022)","page":"Release notes","title":"v1.9.0 (October 29, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-12","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added default fallback for getting ListOfConstraintIndices and NumberOfConstraints when the constraint type is unsupported by the model (#2021)\nAdded support for min and max in nonlinear expressions (#2023)\nAdded support for Indicator{EqualTo{T}} constraints in FileFormats.MPS (#2022)\nAdded default fallback for write_to_file and read_from_file (#2029)","category":"page"},{"location":"moi/changelog/#Fixed-19","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed Constraint.ZeroOneBridge by adding new bounds as affine constraints instead of variable bounds (#1879)\nFixed reading free rows in FileFormats.MPS files (#2009)\nFixed parsing of OBJSENSE blocks in FileFormats.MPS files (#2016) (#2019)\nFixed the parsing of deeply nested nonlinear expressions by removing the use of recursion (#2020)\nFixed the requirements check in Test.test_constrainnt_get_ConstraintIndex (#2024)","category":"page"},{"location":"moi/changelog/#v1.8.2-(September-20,-2022)","page":"Release notes","title":"v1.8.2 (September 20, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Documentation","page":"Release notes","title":"Documentation","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added vale as a documentation linter (#2002)\nImproved styling of code blocks in the PDF (#1999) (#2000)\nFixed a number of typos in the documentation (#2001) (#2003)","category":"page"},{"location":"moi/changelog/#v1.8.1-(September-12,-2022)","page":"Release notes","title":"v1.8.1 (September 12, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-20","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in supports(::AbstractBridgeOptimizer for constraint attributes (#1991) (#1992)","category":"page"},{"location":"moi/changelog/#v1.8.0-(September-1,-2022)","page":"Release notes","title":"v1.8.0 (September 1, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-13","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added new sets\nHyperRectangle (#1961)\nReified (#1955)\nAdded new bridges (#1955)\nBridges.Constraint.ReifiedAllDifferentToCountDistinctBridge\nBridges.Constraint.ReifiedCountDistinctToMILPBridge\nBridges.Constraint.SplitHyperRectangleBridge\nAdded support for atan(y, x) in Nonlinear (#1987)","category":"page"},{"location":"moi/changelog/#Fixed-21","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Lazily construct expressions in Nonlinear so that expressions are updated when Nonlinear.Parameter values are updated (#1984)\nAllow NORM_LIMIT as a TerminationStatus for unbounded problems in Test (#1990)","category":"page"},{"location":"moi/changelog/#v1.7.0-(August-16,-2022)","page":"Release notes","title":"v1.7.0 (August 16, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-14","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added new sets\nHermitianPositiveSemidefiniteConeTriangle\nAdded new optimizer-independent options\nRelativeGapTolerance\nAbsoluteGapTolerance\nAdded new bridges\nBridges.Constraint.GeoMeanToPowerBridge\nBridges.Constraint.HermitianToSymmetricPSDBridge\nBridges.Constraint.IndicatorGreaterToLessThanBridge\nBridges.Constraint.IndicatorLessToGreaterThanBridge\nBridges.Constraint.SplitComplexZerosBridge\nBridges.Constraint.SplitComplexEqualToBridge\nBridges.Objective.QuadratizeBridge\nAdded support for generic number types in Utilities.loadfromstring!\nUpdated FileFormats.MOF to MathOptFormat v1.1, enabling support for constraint programming sets in the MOF file format\nAdded support in various FileFormats for\nindicator constraints in FileFormats.MPS\nquadratic constraints and an objective in FileFormats.LP\nquadratic constraints and an objective in FileFormats.MPS","category":"page"},{"location":"moi/changelog/#Fixed-22","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed some missing promotion rules","category":"page"},{"location":"moi/changelog/#Other-16","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Improved the performance of Jacobian products in Nonlinear\nRemoved an un-needed copy in Utilities.modify_function\nVarious clean-ups in Bridges/bridge_optimizer.jl","category":"page"},{"location":"moi/changelog/#v1.6.1-(July-23,-2022)","page":"Release notes","title":"v1.6.1 (July 23, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-23","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added support for ExponentialCone in MatrixOfConstraints\nFix PSDSquare_3 test to reflect a previously fixed bug getting the ConstraintDual of a PositiveSemidefiniteConeSquare constraint","category":"page"},{"location":"moi/changelog/#v1.6.0-(July-2,-2022)","page":"Release notes","title":"v1.6.0 (July 2, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-15","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added Bridges.needs_final_touch and Bridges.final_touch\nAdded new bridges from constraint programming sets to mixed-integer linear programs:\nAllDifferentToCountDistinctBridge\nCountAtLeastToCountBelongsBridge\nCountBelongsToMILPBridge\nCountDistinctToMILPBridge\nCountGreaterThanToMILPBridge\nCircuitToMILPBridge","category":"page"},{"location":"moi/changelog/#Fixed-24","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Relax an instance of ::Vector to ::AbstractVector in MOI.Nonlinear\nFix BinPackingToMILPBridge to respect variable bounds\nFix SemiToBinaryBridge to throw error if other bounds are set","category":"page"},{"location":"moi/changelog/#v1.5.0-(June-27,-2022)","page":"Release notes","title":"v1.5.0 (June 27, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-16","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added GetAttributeNotAllowed for solvers to indicate when getting an attribute encounters an error\nAdded Utilities.get_fallback support for ObjectiveValue and DualObjectiveValue\nAdded new bridges:\nRootDetConeSquare to RootDetConeTriangle\nLogDetConeSquare to LogDetConeTriangle\nBinPacking to a mixed-integer linear program\nTable to a mixed-integer linear program\nAdded Bridges.print_active_bridges to display the current optimal hyper-path in a Bridges.LazyBridgeOptimizer","category":"page"},{"location":"moi/changelog/#Fixed-25","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed ZeroOne tests with lower and upper bounds\nFixed error in FileFormats.LP when reading a malformed file\nFixed reading of nonlinear programs in FileFormats.MOF\nFixed bug in ConstraintDual when using SquareBridge","category":"page"},{"location":"moi/changelog/#Other-17","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Improved documentation of nonlinear API\nDocumented duality convention for PositiveSemidefiniteConeSquare sets\nFixed typo in Bridges.Constraint.QuadToSOCBridge docstring","category":"page"},{"location":"moi/changelog/#v1.4.0-(June-9,-2022)","page":"Release notes","title":"v1.4.0 (June 9, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-17","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added a number of sets for constraint programming:\nAllDifferent\nBinPacking\nCircuit\nCountAtLeast\nCountBelongs\nCountDistinct\nCountGreaterThan\nCumulative\nPath\nTable\nAdded support for user-defined hessians in Nonlinear\nAdded Bridges.runtests to simplify the testing of bridge implementations","category":"page"},{"location":"moi/changelog/#Fixed-26","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in FileFormats.NL when writing univariate *","category":"page"},{"location":"moi/changelog/#Other-18","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Began a large refactoring of the Bridges submodule, with greatly improved documentation.","category":"page"},{"location":"moi/changelog/#v1.3.0-(May-27,-2022)","page":"Release notes","title":"v1.3.0 (May 27, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-18","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add MOI.Nonlinear submodule. This is a large new submodule that has been refactored from code that was in JuMP. For now, it should be considered experimental.\nAdd FileFormats.NL.SolFileResults(::IO, ::Model)\nAdd FileFormats.NL.read!(::IO, ::Model)\nAdd MOI.modify that accepts a vector of modifications","category":"page"},{"location":"moi/changelog/#Fixed-27","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in Test which attempted to include non-.jl files\nFixed a bug in FileFormats for models with open interval constraints","category":"page"},{"location":"moi/changelog/#Other-19","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a performance issue in Utilities.DoubleDict\nVarious minor improvements to the documentation","category":"page"},{"location":"moi/changelog/#v1.2.0-(April-25,-2022)","page":"Release notes","title":"v1.2.0 (April 25, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-19","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add support for the FORMAT_REW/.rew file format in FileFormats.","category":"page"},{"location":"moi/changelog/#Fixed-28","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fix bug handling of default variable bounds in FileFormats.LP\nFix FileFormats.MPS to not write OBJSENSE by default since this is only supported by some readers.","category":"page"},{"location":"moi/changelog/#v1.1.2-(March-31,-2022)","page":"Release notes","title":"v1.1.2 (March 31, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-29","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fix a range of bugs in FileFormats.LP\nFix reading of problem dimensions in FileFormats.SDPA","category":"page"},{"location":"moi/changelog/#v1.1.1-(March-23,-2022)","page":"Release notes","title":"v1.1.1 (March 23, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-30","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fix bug in test_model_UpperBoundAlreadySet\nFix bug in test_infeasible_ tests\nFix bug in test_objective_ObjectiveFunction_blank\nRelax restriction of MOI.AbstractOptimizer to MOI.ModelLike in Utilities.CachingOptimizer and instantiate.","category":"page"},{"location":"moi/changelog/#New-tests","page":"Release notes","title":"New tests","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add test_conic_empty_matrix that checks conic solvers support problems with no variables.","category":"page"},{"location":"moi/changelog/#v1.1.0-(March-2,-2022)","page":"Release notes","title":"v1.1.0 (March 2, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-20","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added MOI.Utilities.throw_unsupported(::UniversalFallback) for simplifying solver wrappers which copy from a UniversalFallback.","category":"page"},{"location":"moi/changelog/#v1.0.2-(March-1,-2022)","page":"Release notes","title":"v1.0.2 (March 1, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-31","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in the test_model_ScalarFunctionConstantNotZero test\nFixed the error type when an AbstractFunctionConversionBridge cannot get or set an attribute\nIdentified a correctness bug in RSOCtoPSDBridge. We now thrown an error instead of returning an incorrect result.","category":"page"},{"location":"moi/changelog/#v1.0.1-(February-25,-2022)","page":"Release notes","title":"v1.0.1 (February 25, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-32","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in which OptimizerAttributes were not copied in CachingOptimizer\nFixed a bug in which shift_constant did not promote mixed types of coefficients\nFixed a bug in which deleting a constraint of a bridged variable threw ErrorException instead of MOI.DeleteNotAllowed\nFixed a bug in which add_constraint in MatrixOfConstraints did not canonicalize the function\nFixed a bug when modifying scalar constants of a function containing a bridged variable\nFixed a bug in which final_touch was not always called with a CachingOptimizer","category":"page"},{"location":"moi/changelog/#v1.0.0-(February-17,-2022)","page":"Release notes","title":"v1.0.0 (February 17, 2022)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Although tagged as a breaking release, v1.0.0 is v0.10.9 with deprecations removed, similar to how Julia 1.0 was Julia 0.7 with deprecations removed.","category":"page"},{"location":"moi/changelog/#Breaking","page":"Release notes","title":"Breaking","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Julia 1.6 is now the minimum supported version\nAll deprecations have been removed","category":"page"},{"location":"moi/changelog/#Troubleshooting-problems-when-updating","page":"Release notes","title":"Troubleshooting problems when updating","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"If you experience problems when updating, you are likely using previously deprecated features. (By default, Julia does not warn when you use deprecated features.)","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"To find the deprecated features you are using, start Julia with --depwarn=yes:","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"$ julia --depwarn=yes","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Then install MathOptInterface v0.10.9:","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"julia> using Pkg\njulia> pkg\"add MathOptInterface@0.10\"","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"And then run your code. Apply any suggestions, or search the release notes below for advice on updating a specific deprecated feature.","category":"page"},{"location":"moi/changelog/#v0.10.9-(February-16,-2022)","page":"Release notes","title":"v0.10.9 (February 16, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-21","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added MOI.Utilities.FreeVariables as a new VariablesConstrainer for conic solvers\nAdded MOI.default_cache for specifying the model used in CachingOptimizer","category":"page"},{"location":"moi/changelog/#Fixed-33","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed LaTeX printing of MOI.Interval sets","category":"page"},{"location":"moi/changelog/#Other-20","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added Aqua.jl as a CI check, and fixed suggested issues\nThe constructors of GeoMeanBridge, StructOfConstraints, and CachingOptimizer were changed from outer to inner constructors. This change is technically breaking, but does not impact users who followed the documented API.","category":"page"},{"location":"moi/changelog/#v0.10.8-(February-3,-2022)","page":"Release notes","title":"v0.10.8 (February 3, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-22","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added a Base.read! for FileFormats.LP.","category":"page"},{"location":"moi/changelog/#Fixed-34","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a bug in MutableSparseMatrix\nFixed a bug when calling operate!(vcat, ...) with Number arguments\nRemoved unintended export of deprecated symbols\nFixed a bug with PowerCone and DualPowerCone in MatrixOfConstraints.","category":"page"},{"location":"moi/changelog/#v0.10.7-(January-5,-2022)","page":"Release notes","title":"v0.10.7 (January 5, 2022)","text":"","category":"section"},{"location":"moi/changelog/#Added-23","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added test for modifying the constant vector in a VectorAffineFunction-in-Zeros constraint.","category":"page"},{"location":"moi/changelog/#Fixed-35","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed the order in which sets are added to a LazyBridgeOptimizer. Compared to v0.10.6, this may result in bridged models being created with a different number (and order) of variables and constraints. However, it was necessary to fix cases which were previously rejected as unsupported, even though there was a valid bridge transformation.\nFixed an error message in FileFormats.CBF\nFixed comparison in test_linear_integration_Interval\nFixed errors for ConstraintPrimal in a CachingOptimizer\nFixed printing of models with non-Float64 coefficients.","category":"page"},{"location":"moi/changelog/#Other-21","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Various improvements to reduce time-to-first-solve latency\nImproved error message when an optimizer does not support compute_conflict!","category":"page"},{"location":"moi/changelog/#v0.10.6-(November-30,-2021)","page":"Release notes","title":"v0.10.6 (November 30, 2021)","text":"","category":"section"},{"location":"moi/changelog/#Added-24","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added new documentation and tests for infeasibility certificates\nAdded a version control system for the tests in MOI.Test.runtests. Pass exclude_tests_after = v\"0.10.5\" to run tests added in v0.10.5 and earlier.\nMOI.Test.runtests now supports generic number types. To specify the number type T, pass MOI.Test.Config(T).\nAdded infeasible_status to MOI.Test.Config for solvers which return LOCALLY_INFEASIBLE\nCachingOptimizers now use a fallback for ConstraintPrimal. This should enable solvers using a CachingOptimizer to pass tests requiring ConstraintPrimal.","category":"page"},{"location":"moi/changelog/#Fixed-36","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed a StackOverflow bug in copy_to\nFixed error thrown when nonconvex quadratic constraints cannot be bridged\nFixed a bug in copy_to for FileFormats.NL.Model\nFixed a bug in FileFormats.NL when printing large integers\nRemove a common test failure for LowerBoundAlreadySet tests\nUtilities.num_rows is now exported\nRemove parts of failing test_model_copy_to_xxx tests due to bridges","category":"page"},{"location":"moi/changelog/#v0.10.5-(November-7,-2021)","page":"Release notes","title":"v0.10.5 (November 7, 2021)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-37","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed getter in UniversalFallback\nFixed test_solve_conflict_zeroone_ii","category":"page"},{"location":"moi/changelog/#Other-22","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Make normalize_and_add_constraint more flexible\nUpdate paper BibTeX","category":"page"},{"location":"moi/changelog/#v0.10.4-(October-26,-2021)","page":"Release notes","title":"v0.10.4 (October 26, 2021)","text":"","category":"section"},{"location":"moi/changelog/#Added-25","page":"Release notes","title":"Added","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add SolverVersion attribute\nAdd new tests:\ntest_solve_conflict_zeroone_ii\ntest_nonlinear_objective\nUtilities.VariablesContainer now supports ConstraintFunction and ConstraintSet\nThe documentation is now available as a PDF","category":"page"},{"location":"moi/changelog/#Other-23","page":"Release notes","title":"Other","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Update to MutableArithmetics 0.3\nVarious improvements to the documentation","category":"page"},{"location":"moi/changelog/#v0.10.3-(September-18,-2021)","page":"Release notes","title":"v0.10.3 (September 18, 2021)","text":"","category":"section"},{"location":"moi/changelog/#Fixed-38","page":"Release notes","title":"Fixed","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed bug which prevented callbacks from working through a CachingOptimizer\nFixed bug in Test submodule","category":"page"},{"location":"moi/changelog/#v0.10.2-(September-16,-2021)","page":"Release notes","title":"v0.10.2 (September 16, 2021)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Updated MathOptFormat to v1.0\nUpdated JSONSchema to v1.0\nAdded Utilities.set_with_dimension\nAdded two-argument optimize!(::AbstractOptimizer, ::ModelLike)\nThe experimental feature copy_to_and_optimize! has been removed\nDet bridges now support getting ConstraintFunction and ConstraintSet\nVarious minor bug fixes identified by improved testing","category":"page"},{"location":"moi/changelog/#v0.10.1-(September-8,-2021)","page":"Release notes","title":"v0.10.1 (September 8, 2021)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Various fixes to MOI.Test","category":"page"},{"location":"moi/changelog/#v0.10.0-(September-6,-2021)","page":"Release notes","title":"v0.10.0 (September 6, 2021)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"MOI v0.10 is a significant breaking release. There are a large number of user-visible breaking changes and code refactors, as well as a substantial number of new features.","category":"page"},{"location":"moi/changelog/#Breaking-in-MOI","page":"Release notes","title":"Breaking in MOI","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"SingleVariable has been removed; use VariableIndex instead\nSingleVariableConstraintNameError has been renamed to VariableIndexConstraintNameError\nSettingSingleVariableFunctionNotAllowed has been renamed to SettingVariableIndexFunctionNotAllowed\nVariableIndex constraints should not support ConstraintName\nVariableIndex constraints should not support ConstraintBasisStatus; implement VariableBasisStatus instead\nListOfConstraints has been renamed to ListOfConstraintTypesPresent\nListOfConstraintTypesPresent should now return Tuple{Type,Type} instead of Tuple{DataType,DataType}\nSolveTime has been renamed to SolveTimeSec\nIndicatorSet has been renamed to Indicator\nRawParameter has been renamed to RawOptimizerAttribute and now takes String instead of Any as the only argument\nThe .N field in result attributes has been renamed to .result_index\nThe .variable_index field in ScalarAffineTerm has been renamed to .variable\nThe .variable_index_1 field in ScalarQuadraticTerm has been renamed to .variable_1\nThe .variable_index_2 field in ScalarQuadraticTerm has been renamed to .variable_2\nThe order of affine_terms and quadratic_terms in ScalarQuadraticFunction and VectorQuadraticFunction have been reversed. Both functions now accept quadratic, affine, and constant terms in that order.\nThe index_value function has been removed. Use .value instead.\nisapprox has been removed for SOS1 and SOS2.\nThe dimension argument to Complements(dimension::Int) should now be the length of the corresponding function, instead of half the length. An ArgumentError is thrown if dimension is not even.\ncopy_to no longer takes keyword arguments:\ncopy_names: now copy names if they are supported by the destination solver\nfilter_constraints: use Utilities.ModelFilter instead\nwarn_attributes: never warn about optimizer attributes","category":"page"},{"location":"moi/changelog/#Breaking-in-Bridges","page":"Release notes","title":"Breaking in Bridges","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Constraint.RSOCBridge has been renamed to Constraint.RSOCtoSOCBridge\nConstraint.SOCRBridge has been renamed to Constraint.SOCtoRSOCBridge\nBridges now return vectors that can be modified by the user. Previously, some bridges returned views instead of copies.\nBridges.IndexInVector has been unified into a single type. Previously, there was a different type for each submodule within Bridges\nThe signature of indicator bridges has been fixed. Use MOI.Bridges.Constraint.IndicatortoSOS1{Float64}(model).","category":"page"},{"location":"moi/changelog/#Breaking-in-FileFormats","page":"Release notes","title":"Breaking in FileFormats","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"FileFormats.MOF.Model no longer accepts validate argument. Use the JSONSchema package to validate the MOF file. See the documentation for more information.","category":"page"},{"location":"moi/changelog/#Breaking-in-Utilities","page":"Release notes","title":"Breaking in Utilities","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"The datastructure of Utilities.Model (and models created with Utilities.@model) has been significantly refactored in a breaking way. This includes the way that objective functions and variable-related information is stored.\nUtilities.supports_default_copy has been renamed to supports_incremental_interface\nUtilities.automatic_copy_to has been renamed to Utilities.default_copy_to\nThe allocate-load API has been removed\nCachingOptimizers are now initialized as EMPTY_OPTIMIZER instead of ATTACHED_OPTIMIZER. If your code relies on the optimizer being attached, call MOIU.attach_optimizer(model) after creation.\nThe field names of Utilities.IndexMap have been renamed to var_map and con_map. Accessing these fields directly is considered a private detail that may change. Use the public getindex and setindex! API instead.\nThe size argument to Utilities.CleverDicts.CleverDict(::Integer) has been removed.\nThe size argument to Utilities.IndexMap(::Integer) has been removed.\nUtilities.DoubleDicts have been significantly refactored. Consult the source code for details.\nUtilities.test_models_equal has been moved to MOI.Test","category":"page"},{"location":"moi/changelog/#Breaking-in-Test","page":"Release notes","title":"Breaking in Test","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"MOI.Test has been renamed to MOI.DeprecatedTest\nAn entirely new MOI.Test submodule has been written. See the documentation for details. The new MOI.Test submodule may find many bugs in the implementations of existing solvers that were previously untested.","category":"page"},{"location":"moi/changelog/#Other-changes:","page":"Release notes","title":"Other changes:","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"attribute_value_type has been added\ncopy_to_and_optimize! has been added\nVariableBasisStatus has been added\nprint(model) now prints a human-readable description of the model\nVarious improvements to the FileFormats submodule\nFileFormats.CBF was refactored and received bugfixes\nSupport for MathOptFormat v0.6 was added in FileFormats.MOF\nFileFormats.MPS has had bugfixes and support for more features such as OBJSENSE and objective constants.\nFileFormats.NL has been added to support nonlinear files\nImproved type inference throughout to reduce latency","category":"page"},{"location":"moi/changelog/#Updating","page":"Release notes","title":"Updating","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"A helpful script when updating is:","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"for (root, dirs, files) in walkdir(\".\")\n    for file in files\n        if !endswith(file, \".jl\")\n            continue\n        end\n        path = joinpath(root, file)\n        s = read(path, String)\n        for pair in [\n            \".variable_index\" => \".variable\",\n            \"RawParameter\" => \"RawOptimizerAttribute\",\n            \"ListOfConstraints\" => \"ListOfConstraintTypesPresent\",\n            \"TestConfig\" => \"Config\",\n            \"attr.N\" => \"attr.result_index\",\n            \"SolveTime\" => \"SolveTimeSec\",\n            \"DataType\" => \"Type\",\n            \"Utilities.supports_default_copy_to\" =>\n                \"supports_incremental_interface\",\n            \"SingleVariableConstraintNameError\" =>\n                \"VariableIndexConstraintNameError\",\n            \"SettingSingleVariableFunctionNotAllowed\" =>\n                \"SettingVariableIndexFunctionNotAllowed\",\n            \"automatic_copy_to\" => \"default_copy_to\",\n        ]\n            s = replace(s, pair)\n        end\n        write(path, s)\n    end\nend","category":"page"},{"location":"moi/changelog/#v0.9.22-(May-22,-2021)","page":"Release notes","title":"v0.9.22 (May 22, 2021)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"This release contains backports from the ongoing development of the v0.10 release.","category":"page"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Improved type inference in Utilities, Bridges and FileFormats submodules to reduce latency.\nImproved performance of Utilities.is_canonical.\nFixed Utilities.pass_nonvariable_constraints with bridged variables.\nFixed performance regression of Utilities.Model.\nFixed ordering of objective setting in parser.","category":"page"},{"location":"moi/changelog/#v0.9.21-(April-23,-2021)","page":"Release notes","title":"v0.9.21 (April 23, 2021)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added supports_shift_constant.\nImprove performance of bridging quadratic constraints.\nAdd precompilation statements.\nLarge improvements to the documentation.\nFix a variety of inference issues, benefiting precompilation and reducing initial latency.\nRawParameters are now ignored when resetting a CachingOptimizer. Previously, changing the underlying optimizer after RawParameters were set would throw an error.\nUtilities.AbstractModel is being refactored. This may break users interacting with private fields of a model generated using @model.","category":"page"},{"location":"moi/changelog/#v0.9.20-(February-20,-2021)","page":"Release notes","title":"v0.9.20 (February 20, 2021)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Improved performance of Utilities.ScalarFunctionIterator\nAdded support for compute_conflict to MOI layers\nAdded test with zero off-diagonal quadratic term in objective\nFixed double deletion of nested bridged SingleVariable/VectorOfVariables constraints\nFixed modification of un-set objective\nFixed function modification with duplicate terms\nMade unit tests abort without failing if the problem class is not supported\nFormatted code with JuliaFormatter\nClarified BasisStatusCode's docstring","category":"page"},{"location":"moi/changelog/#v0.9.19-(December-1,-2020)","page":"Release notes","title":"v0.9.19 (December 1, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added CallbackNodeStatus attribute\nAdded bridge from GreaterThan or LessThan to Interval\nAdded tests for infeasibility certificates and double optimize\nFixed support for Julia v1.6\nRe-organized MOI docs and added documentation for adding a test","category":"page"},{"location":"moi/changelog/#v0.9.18-(November-3,-2020)","page":"Release notes","title":"v0.9.18 (November 3, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Various improvements for working with complex numbers\nAdded GeoMeantoRelEntrBridge to bridge a GeometricMeanCone constraint to a relative entropy constraint","category":"page"},{"location":"moi/changelog/#v0.9.17-(September-21,-2020)","page":"Release notes","title":"v0.9.17 (September 21, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed CleverDict with variable of negative index value\nImplement supports_add_constrained_variable for MockOptimizer","category":"page"},{"location":"moi/changelog/#v0.9.16-(September-17,-2020)","page":"Release notes","title":"v0.9.16 (September 17, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Various fixes:\n32-bit support\nCleverDict with abstract value type\nChecks in test suite","category":"page"},{"location":"moi/changelog/#v0.9.15-(September-14,-2020)","page":"Release notes","title":"v0.9.15 (September 14, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Bridges improvements:\n(R)SOCtoNonConvexQuad bridge\nZeroOne bridge\nUse supports_add_constrained_variable in LazyBridgeOptimizer\nExposed VariableBridgeCost and ConstraintBridgeCost attributes\nPrioritize constraining variables on creation according to these costs\nRefactor bridge debugging\nLarge performance improvements across all submodules\nLots of documentation improvements\nFileFormats improvements:\nUpdate MathOptFormat to v0.5\nFix supported objectives in FileFormats\nTesting improvements:\nAdd name option for basic_constraint_test\nBug fixes and missing methods\nAdd length for iterators\nFix bug with duplicate terms\nFix order of LinearOfConstraintIndices","category":"page"},{"location":"moi/changelog/#v0.9.14-(May-30,-2020)","page":"Release notes","title":"v0.9.14 (May 30, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add a solver-independent interface for accessing the set of conflicting constraints an Irreducible Inconsistent Subsystem (#1056).\nBump JSONSchema dependency from v0.2 to v0.3 (#1090).\nDocumentation improvements:\nFix typos (#1054, #1060, #1061, #1064, #1069, #1070).\nRemove the outdated recommendation for a package implementing MOI for a solver XXX to be called MathOptInterfaceXXX (#1087).\nUtilities improvements:\nFix is_canonical for quadratic functions (#1081, #1089).\nImplement add_constrained_variable[s] for CachingOptimizer so that it is added as constrained variables to the underlying optimizer (#1084).\nAdd support for custom objective functions for UniversalFallback (#1086).\nDeterministic ordering of constraints in UniversalFallback (#1088).\nTesting improvements:\nAdd NormOneCone/NormInfinityCone tests (#1045).\nBridges improvements:\nAdd bridges from Semiinteger and Semicontinuous (#1059).\nImplement getting ConstraintSet for Variable.FlipSignBridge (#1066).\nFix setting ConstraintFunction for Constraint.ScalarizeBridge (#1093).\nFix NormOne/NormInf bridges with nonzero constants (#1045).\nFix StackOverflow in debug (#1063).\nFileFormats improvements:\n[SDPA] Implement the extension for integer variables (#1079).\n[SDPA] Ignore comments after m and nblocks and detect dat-s extension (#1077).\n[SDPA] No scaling of off-diagonal coefficient (#1076).\n[SDPA] Add missing negation of constant (#1075).","category":"page"},{"location":"moi/changelog/#v0.9.13-(March-24,-2020)","page":"Release notes","title":"v0.9.13 (March 24, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added tests for Semicontinuous and Semiinteger variables (#1033).\nAdded tests for using ExprGraphs from NLP evaluators (#1043).\nUpdate version compatibilities of dependencies (#1034, #1051, #1052).\nFixed typos in documentation (#1044).","category":"page"},{"location":"moi/changelog/#v0.9.12-(February-28,-2020)","page":"Release notes","title":"v0.9.12 (February 28, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed writing NLPBlock in MathOptFormat (#1037).\nFixed MockOptimizer for result attributes with non-one result index (#1039).\nUpdated test template with instantiate (#1032).","category":"page"},{"location":"moi/changelog/#v0.9.11-(February-21,-2020)","page":"Release notes","title":"v0.9.11 (February 21, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add an option for the model created by Utilities.@model to be a subtype of AbstractOptimizer (#1031).\nDescribed dual cone in docstrings of GeoMeanCone and RelativeEntropyCone (#1018, #1028).\nFixed typos in documentation (#1022, #1024).\nFixed warning of unsupported attribute (#1027).\nAdded more rootdet/logdet conic tests (#1026).\nImplemented ConstraintDual for Constraint.GeoMeanBridge, Constraint.RootDetBridge and Constraint.LogDetBridge and test duals in tests with GeoMeanCone and RootDetConeTriangle and LogDetConeTriangle cones (#1025, #1026).","category":"page"},{"location":"moi/changelog/#v0.9.10-(January-31,-2020)","page":"Release notes","title":"v0.9.10 (January 31, 2020)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added OptimizerWithAttributes grouping an optimizer constructor and a list of optimizer attributes (#1008).\nAdded RelativeEntropyCone with corresponding bridge into exponential cone constraints (#993).\nAdded NormSpectralCone and NormNuclearCone with corresponding bridges into positive semidefinite constraints (#976).\nAdded supports_constrained_variable(s) (#1004).\nAdded dual_set_type (#1002).\nAdded tests for vector specialized version of delete (#989, #1011).\nAdded PSD3 test (#1007).\nClarified dual solution of Tests.pow1v and Tests.pow1f (#1013).\nAdded support for EqualTo and Zero in Bridges.Constraint.SplitIntervalBridge (#1005).\nFixed Utilities.vectorize for empty vector (#1003).\nFixed free variables in LP writer (#1006).","category":"page"},{"location":"moi/changelog/#v0.9.9-(December-29,-2019)","page":"Release notes","title":"v0.9.9 (December 29, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Incorporated MathOptFormat.jl as the FileFormats submodule. FileFormats provides readers and writers for a number of standard file formats and MOF, a file format specialized for MOI (#969).\nImproved performance of deletion of vector of variables in MOI.Utilities.Model (#983).\nUpdated to MutableArithmetics v0.2 (#981).\nAdded MutableArithmetics.promote_operation allocation tests (#975).\nFixed inference issue on Julia v1.1 (#982).","category":"page"},{"location":"moi/changelog/#v0.9.8-(December-19,-2019)","page":"Release notes","title":"v0.9.8 (December 19, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Implemented MutableArithmetics API (#924).\nFixed callbacks with CachingOptimizer (#959).\nFixed MOI.dimension for MOI.Complements (#948).\nAdded fallback for add_variables (#972).\nAdded is_diagonal_vectorized_index utility (#965).\nImproved linear constraints display in manual (#963, #964).\nBridges improvements:\nAdded IndicatorSet to SOS1 bridge (#877).\nAdded support for starting values for Variable.VectorizeBridge (#944).\nFixed MOI.add_constraints with non-bridged variable constraint on bridged variable (#951).\nFixed corner cases and docstring of GeoMeanBridge (#961, #962, #966).\nFixed choice between variable or constraint bridges for constrained variables (#973).\nImprove performance of bridge shortest path (#945, #946, #956).\nAdded docstring for test_delete_bridge (#954).\nAdded Variable bridge tests (#952).","category":"page"},{"location":"moi/changelog/#v0.9.7-(October-30,-2019)","page":"Release notes","title":"v0.9.7 (October 30, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Implemented _result_index_field for NLPBlockDual (#934).\nFixed copy of model with starting values for vector constraints (#941).\nBridges improvements:\nImproved performance of add_bridge and added has_bridge (#935).\nAdded AbstractSetMapBridge for bridges between sets S1, S2 such that there is a linear map A such that A*S1 = S2 (#933).\nAdded support for starting values for FlipSignBridge, VectorizeBridge, ScalarizeBridge, SlackBridge, SplitIntervalBridge, RSOCBridge, SOCRBridge NormInfinityBridge, SOCtoPSDBridge and RSOCtoPSDBridge (#933, #936, #937, #938, #939).","category":"page"},{"location":"moi/changelog/#v0.9.6-(October-25,-2019)","page":"Release notes","title":"v0.9.6 (October 25, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added complementarity constraints (#913).\nAllowed ModelLike objects as value of attributes (#928).\nTesting improvements:\nAdded dual_objective_value option to MOI.Test.TestConfig (#922).\nAdded InvalidIndex tests in basic_constraint_tests (#921).\nAdded tests for the constant term in indicator constraint (#929).\nBridges improvements:\nAdded support for starting values for Functionize bridges (#923).\nAdded variable indices context to variable bridges (#920).\nFixed a typo in printing o debug_supports (#927).","category":"page"},{"location":"moi/changelog/#v0.9.5-(October-9,-2019)","page":"Release notes","title":"v0.9.5 (October 9, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Clarified PrimalStatus/DualStatus to be NO_SOLUTION if result_index is out of bounds (#912).\nAdded tolerance for checks  and use ResultCount + 1 for the result_index in MOI.Test.solve_result_status (#910, #917).\nUse 0.5 instead of 2.0 for power in PowerCone in basic_constraint_test (#916).\nBridges improvements:\nAdded debug utilities for unsupported variable/constraint/objective (#861).\nFixed deletion of variables in bridged VectorOfVariables constraints (#909).\nFixed result_index with objective bridges (#911).","category":"page"},{"location":"moi/changelog/#v0.9.4-(October-2,-2019)","page":"Release notes","title":"v0.9.4 (October 2, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Added solver-independent MIP callbacks (#782).\nImplements submit for Utilities.CachingOptimizer and Bridges.AbstractBridgeOptimizer (#906).\nAdded tests for result count of solution attributes (#901, #904).\nAdded NumberOfThreads attribute (#892).\nAdded Utilities.get_bounds to get the bounds on a variable (#890).\nAdded a note on duplicate coefficients in documentation (#581).\nAdded result index in ConstraintBasisStatus (#898).\nAdded extension dictionary to Utilities.Model (#884, #895).\nFixed deletion of constrained variables for CachingOptimizer (#905).\nImplemented Utilities.shift_constraint for Test.UnknownScalarSet (#896).\nBridges improvements:\nAdded Variable.RSOCtoSOCBridge (#907).\nImplemented MOI.get for ConstraintFunction/ConstraintSet for Bridges.Constraint.SquareBridge (#899).","category":"page"},{"location":"moi/changelog/#v0.9.3-(September-20,-2019)","page":"Release notes","title":"v0.9.3 (September 20, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixed ambiguity detected in Julia v1.3 (#891, #893).\nFixed missing sets from ListOfSupportedConstraints (#880).\nFixed copy of VectorOfVariables constraints with duplicate indices (#886).\nAdded extension dictionary to MOIU.Model (#884).\nImplemented MOI.get for function and set for GeoMeanBridge (#888).\nUpdated documentation for SingleVariable indices and bridges (#885).\nTesting improvements:\nAdded more comprehensive tests for names (#882).\nAdded tests for SingleVariable duals (#883).\nAdded tests for DualExponentialCone and DualPowerCone (#873).\nImprovements for arbitrary coefficient type:\nFixed == for sets with mutable fields (#887).\nRemoved some Float64 assumptions in bridges (#878).\nAutomatic selection of Constraint.[Scalar|Vector]FunctionizeBridge (#889).","category":"page"},{"location":"moi/changelog/#v0.9.2-(September-5,-2019)","page":"Release notes","title":"v0.9.2 (September 5, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Implemented model printing for MOI.ModelLike and specialized it for models defined in MOI (864).\nGeneralized contlinear tests for arbitrary coefficient type (#855).\nFixed supports_constraint for Semiinteger and Semicontinuous and supports for ObjectiveFunction (#859).\nFixed Allocate-Load copy for single variable constraints (#856).\nBridges improvements:\nAdd objective bridges (#789).\nFixed Variable.RSOCtoPSDBridge for dimension 2 (#869).\nAdded Variable.SOCtoRSOCBridge (#865).\nAdded Constraint.SOCRBridge and disable MOI.Bridges.Constraint.SOCtoPSDBridge (#751).\nFixed added_constraint_types for Contraint.LogDetBridge and Constraint.RootDetBridge (#870).","category":"page"},{"location":"moi/changelog/#v0.9.1-(August-22,-2019)","page":"Release notes","title":"v0.9.1 (August 22, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fix support for Julia v1.2 (#834).\nL1 and L∞ norm epigraph cones and corresponding bridges to LP were added (#818).\nAdded tests to MOI.Test.nametest (#833).\nFix MOI.Test.soc3test for solvers not supporting infeasibility certificates (#839).\nImplements operate for operators * and / between vector function and constant (#837).\nImplements show for MOI.Utilities.IndexMap (#847).\nFix corner cases for mapping of variables in MOI.Utilities.CachingOptimizer and substitution of variables in MOI.Bridges.AbstractBridgeOptimizer (#848).\nFix transformation of constant terms for MOI.Bridges.Constraint.SOCtoPSDBridge and MOI.Bridges.Constraint.RSOCtoPSDBridge (#840).","category":"page"},{"location":"moi/changelog/#v0.9.0-(August-13,-2019)","page":"Release notes","title":"v0.9.0 (August 13, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Support for Julia v0.6 and v0.7 was dropped (#714, #717).\nA MOI.Utilities.Model implementation of ModelLike, this should replace most use cases of MOI.Utilities.@model (#781).\nadd_constrained_variable and add_constrained_variables were added (#759).\nSupport for indicator constraints was added (#709, #712).\nDualObjectiveValue attribute was added (#473).\nRawParameter attribute was added (#733).\nA dual_set function was added (#804).\nA Benchmarks submodule was added to facilitate solver benchmarking (#769).\nA submit function was added, this may for instance allow the user to submit solutions or cuts to the solver from a callback (#775).\nThe field of ObjectiveValue was renamed to result_index (#729).\nThe _constant and Utilities.getconstant function were renamed to constant\nREDUCTION_CERTIFICATE result status was added (#734).\nAbstract matrix sets were added (#731).\nTesting improvements:\nThe testing guideline was updated (#728).\nQuadratic tests were added (#697).\nUnit tests for RawStatusString, SolveTime, Silent and SolverName were added (#726, #741).\nA rotated second-order cone test was added (#759).\nA power cone test was added (#768).\nTests for ZeroOne variables with variable bounds were added (#772).\nAn unbounded test was added (#773).\nExisting tests had a few updates (#702, #703, #763).\nDocumentation improvements:\nAdded a section on CachingOptimizer (#777).\nAdded a section on UniversalFallback, Model and @model (#762).\nTransition the knapsack example to a doctest with MockOptimizer (#786).\nUtilities improvements:\nA CleverDict utility was added for a vector that automatically transform into a dictionary once a first index is removed (#767).\nThe Utilities.constant function was renamed to Utilities.constant_vector (#740).\nImplement optimizer attributes for CachingOptimizer (#745).\nRename Utilities.add_scalar_constraint to Utilities.normalize_and_add_constraint (#801).\noperate with vcat, SingleVariable and VectorOfVariables now returns a VectorOfVariables (#616).\nFix a type piracy of operate (#784).\nThe load_constraint fallback signature was fixed (#760).\nThe set_dot function was extended to work with sparse arrays (#805).\nBridges improvements:\nThe bridges no longer store the constraint function and set before it is bridged, the bridges now have to implement ConstraintFunction and ConstraintSet if the user wants to recover them. As a consequence, the @bridge macro was removed (#722).\nBridge are now instantiated with a bridge_constraint function instead of using a constructor (#730).\nFix constraint attributes for bridges (#699).\nConstraint bridges were moved to the Bridges/Constraint submodule so they should now inherit from MOI.Bridges.Constraint.Abstract and should implement MOI.Bridges.Constraint.concrete_bridge_type instead of MOI.Bridges.concrete_bridge_type (#756).\nVariable bridges were added in (#759).\nVarious improvements (#746, #747).","category":"page"},{"location":"moi/changelog/#v0.8.4-(March-13,-2019)","page":"Release notes","title":"v0.8.4 (March 13, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Performance improvement in default_copy_to and bridge optimizer (#696).\nAdd Silent and implement setting optimizer attributes in caching and mock optimizers (#695).\nAdd Functionize bridges (SingleVariable and VectorOfVariables) (#659).\nMinor typo fixes (#694).","category":"page"},{"location":"moi/changelog/#v0.8.3-(March-6,-2019)","page":"Release notes","title":"v0.8.3 (March 6, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Use zero constant in scalar constraint function of MOI.Test.copytest (#691).\nFix variable deletion with SingleVariable objective function (#690).\nFix LazyBridgeOptimizer with bridges that add no constraints (#689).\nError message improvements (#673, #685, #686, #688).\nDocumentation improvements (#682, #683, #687).\nBasis status:\nRemove VariableBasisStatus (#679).\nTest ConstraintBasisStatus and implement it in bridges (#678).\nFix inference of NumberOfVariables and NumberOfConstraints (#677).\nImplement division between a quadratic function and a number (#675).","category":"page"},{"location":"moi/changelog/#v0.8.2-(February-7,-2019)","page":"Release notes","title":"v0.8.2 (February 7, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add RawStatusString attribute (#629).\nDo not set names to the optimizer but only to the cache in CachingOptimizer (#638).\nMake scalar MOI functions act as scalars in broadcast (#646).\nAdd function utilities:\nImplement Base.zero (#634), Base.iszero (#643), add missing arithmetic operations (#644, #645) and fix division (#648).\nAdd a vectorize function that turns a vector of ScalarAffineFunction into a VectorAffineFunction (#642).\nImprove support for starting values:\nShow a warning in copy when starting values are not supported instead of throwing an error (#630).\nFix UniversalFallback for getting an variable or constraint attribute set to no indices (#623).\nAdd a test in contlineartest with partially set VariablePrimalStart.\nBridges improvements:\nFix StackOverFlow in LazyBridgeOptimizer when there is a cycle in the graph of bridges.\nAdd Slack bridges (#610, #650).\nAdd FlipSign bridges (#658).\nAdd tests with duplicate coefficients in ScalarAffineFunction and VectorAffineFunction (#639).\nUse tolerance to compare VariablePrimal in rotatedsoc1 test (#632).\nUse a zero constant in ScalarAffineFunction of constraints in psdt2 (#622).","category":"page"},{"location":"moi/changelog/#v0.8.1-(January-7,-2019)","page":"Release notes","title":"v0.8.1 (January 7, 2019)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Adding an NLP objective now overrides any objective set using the ObjectiveFunction attribute (#619).\nRename fullbridgeoptimizer into full_bridge_optimizer (#621).\nAllow custom constraint types with full_bridge_optimizer (#617).\nAdd Vectorize bridge which transforms scalar linear constraints into vector linear constraints (#615).","category":"page"},{"location":"moi/changelog/#v0.8.0-(December-18,-2018)","page":"Release notes","title":"v0.8.0 (December 18, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Rename all enum values to follow the JuMP naming guidelines for constants, for example, Optimal becomes OPTIMAL, and DualInfeasible becomes DUAL_INFEASIBLE.\nRename CachingOptimizer methods for style compliance.\nAdd an MOI.TerminationStatusCode called ALMOST_DUAL_INFEASIBLE.","category":"page"},{"location":"moi/changelog/#v0.7.0-(December-13,-2018)","page":"Release notes","title":"v0.7.0 (December 13, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Test that MOI.TerminationStatus is MOI.OptimizeNotCalled before MOI.optimize! is called.\nCheck supports_default_copy_to in tests (#594).\nKey pieces of information like optimality, infeasibility, etc., are now reported through TerminationStatusCode. It is typically no longer necessary to check the result statuses in addition to the termination status.\nAdd perspective dimension to log-det cone (#593).","category":"page"},{"location":"moi/changelog/#v0.6.4-(November-27,-2018)","page":"Release notes","title":"v0.6.4 (November 27, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Add OptimizeNotCalled termination status (#577) and improve documentation of other statuses (#575).\nAdd a solver naming guideline (#578).\nMake FeasibilitySense the default ObjectiveSense (#579).\nFix Utilities.@model and Bridges.@bridge macros for functions and sets defined outside MOI (#582).\nDocument solver-specific attributes (#580) and implement them in Utilities.CachingOptimizer (#565).","category":"page"},{"location":"moi/changelog/#v0.6.3-(November-16,-2018)","page":"Release notes","title":"v0.6.3 (November 16, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Variables and constraints are now allowed to have duplicate names. An error is thrown only on lookup. This change breaks some existing tests. (#549)\nAttributes may now be partially set (some values could be nothing). (#563)\nPerformance improvements in Utilities.Model (#549, #567, #568)\nFix bug in QuadtoSOC (#558).\nNew supports_default_copy_to method that optimizers should implement to control caching behavior.\nDocumentation improvements.","category":"page"},{"location":"moi/changelog/#v0.6.2-(October-26,-2018)","page":"Release notes","title":"v0.6.2 (October 26, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Improve hygiene of @model macro (#544).\nFix bug in copy tests (#543).\nFix bug in UniversalFallback attribute getter (#540).\nAllow all correct solutions for solve_blank_obj unit test (#537).\nAdd errors for Allocate-Load and bad constraints (#534).\n[performance] Add specialized implementation of hash for VariableIndex (#533).\n[performance] Construct the name to object dictionaries lazily in model (#535).\nAdd the QuadtoSOC bridge which transforms ScalarQuadraticFunction constraints into RotatedSecondOrderCone (#483).","category":"page"},{"location":"moi/changelog/#v0.6.1-(September-22,-2018)","page":"Release notes","title":"v0.6.1 (September 22, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Enable PositiveSemidefiniteConeSquare set and quadratic functions in MOIB.fullbridgeoptimizer (#524).\nAdd warning in the bridge between PositiveSemidefiniteConeSquare and PositiveSemidefiniteConeTriangle when the matrix is almost symmetric (#522).\nModify MOIT.copytest to not add multiples constraints on the same variable (#521).\nAdd missing keyword argument in one of MOIU.add_scalar_constraint methods (#520).","category":"page"},{"location":"moi/changelog/#v0.6.0-(August-30,-2018)","page":"Release notes","title":"v0.6.0 (August 30, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"The MOIU.@model and MOIB.@bridge macros now support functions and sets defined in external modules. As a consequence, function and set names in the macro arguments need to be prefixed by module name.\nRename functions according to the JuMP style guide:\ncopy! with keyword arguments copynames and warnattributes -> copy_to with keyword arguments copy_names and warn_attributes;\nset! -> set;\naddvariable[s]! -> add_variable[s];\nsupportsconstraint -> supports_constraint;\naddconstraint[s]! -> add_constraint[s];\nisvalid -> is_valid;\nisempty -> is_empty;\nBase.delete! -> delete;\nmodify! -> modify;\ntransform! -> transform;\ninitialize! -> initialize;\nwrite -> write_to_file; and\nread! -> read_from_file.\nRemove free! (use Base.finalize instead).\nAdd the SquarePSD bridge which transforms PositiveSemidefiniteConeTriangle constraints into PositiveSemidefiniteConeTriangle.\nAdd result fallback for ConstraintDual of variable-wise constraint, ConstraintPrimal and ObjectiveValue.\nAdd tests for ObjectiveBound.\nAdd test for empty rows in vector linear constraint.\nRework errors: CannotError has been renamed NotAllowedError and the distinction between UnsupportedError and NotAllowedError is now about whether the element is not supported (for example, it cannot be copied a model containing this element) or the operation is not allowed (either because it is not implemented, because it cannot be performed in the current state of the model, or because it cannot be performed for a specific index)\ncanget is removed. NoSolution is added as a result status to indicate that the solver does not have either a primal or dual solution available (See #479).","category":"page"},{"location":"moi/changelog/#v0.5.0-(August-5,-2018)","page":"Release notes","title":"v0.5.0 (August 5, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fix names with CachingOptimizer.\nCleanup thanks to @mohamed82008.\nAdded a universal fallback for constraints.\nFast utilities for function canonicalization thanks to @rdeits.\nRenamed dimension field to side_dimension in the context of matrix-like sets.\nNew and improved tests for cases like duplicate terms and ObjectiveBound.\nRemoved cantransform, canaddconstraint, canaddvariable, canset, canmodify, and candelete functions from the API. They are replaced by a new set of errors that are thrown: Subtypes of UnsupportedError indicate unsupported operations, while subtypes of CannotError indicate operations that cannot be performed in the current state.\nThe API for copy! is updated to remove the CopyResult type.\nUpdates for the new JuMP style guide.","category":"page"},{"location":"moi/changelog/#v0.4.1-(June-28,-2018)","page":"Release notes","title":"v0.4.1 (June 28, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Fixes vector function modification on 32 bits.\nFixes Bellman-Ford algorithm for bridges.\nAdded an NLP test with FeasibilitySense.\nUpdate modification documentation.","category":"page"},{"location":"moi/changelog/#v0.4.0-(June-23,-2018)","page":"Release notes","title":"v0.4.0 (June 23, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Helper constructors for VectorAffineTerm and VectorQuadraticTerm.\nAdded modify_lhs to TestConfig.\nAdditional unit tests for optimizers.\nAdded a type parameter to CachingOptimizer for the optimizer field.\nNew API for problem modification (#388)\nTests pass without deprecation warnings on Julia 0.7.\nSmall fixes and documentation updates.","category":"page"},{"location":"moi/changelog/#v0.3.0-(May-25,-2018)","page":"Release notes","title":"v0.3.0 (May 25, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Functions have been redefined to use arrays-of-structs instead of structs-of-arrays.\nImprovements to MockOptimizer.\nSignificant changes to Bridges.\nNew and improved unit tests.\nFixes for Julia 0.7.","category":"page"},{"location":"moi/changelog/#v0.2.0-(April-24,-2018)","page":"Release notes","title":"v0.2.0 (April 24, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Improvements to and better coverage of Tests.\nDocumentation fixes.\nSolverName attribute.\nChanges to the NLP interface (new definition of variable order and arrays of structs for bound pairs and sparsity patterns).\nAddition of NLP tests.\nIntroduction of UniversalFallback.\ncopynames keyword argument to MOI.copy!.\nAdd Bridges submodule.","category":"page"},{"location":"moi/changelog/#v0.1.0-(February-28,-2018)","page":"Release notes","title":"v0.1.0 (February 28, 2018)","text":"","category":"section"},{"location":"moi/changelog/","page":"Release notes","title":"Release notes","text":"Initial public release.\nThe framework for MOI was developed at the JuMP-dev workshop at MIT in June 2017 as a sorely needed replacement for MathProgBase.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Bridges/overview.md\"","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Bridges/overview/#The-Bridges-submodule","page":"Overview","title":"The Bridges submodule","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"The Bridges module simplifies the process of converting models between equivalent formulations.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"tip: Tip\nRead our paper for more details on how bridges are implemented.","category":"page"},{"location":"moi/submodules/Bridges/overview/#Why-bridges?","page":"Overview","title":"Why bridges?","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"A constraint can often be written in a number of equivalent formulations. For example, the constraint l le a^top x le u (ScalarAffineFunction-in-Interval) could be re-formulated as two constraints: a^top x ge l (ScalarAffineFunction-in-GreaterThan) and a^top x le u (ScalarAffineFunction-in-LessThan). An alternative re-formulation is to add a dummy variable y with the constraints l le y le u (VariableIndex-in-Interval) and a^top x - y = 0 (ScalarAffineFunction-in-EqualTo).","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"To avoid each solver having to code these transformations manually, MathOptInterface provides bridges.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"A bridge is a small transformation from one constraint type to another (potentially collection of) constraint type.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Because these bridges are included in MathOptInterface, they can be re-used by any optimizer. Some bridges also implement constraint modifications and constraint primal and dual translations.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Several bridges can be used in combination to transform a single constraint into a form that the solver may understand. Choosing the bridges to use takes the form of finding a shortest path in the hyper-graph of bridges. The methodology is detailed in the MOI paper.","category":"page"},{"location":"moi/submodules/Bridges/overview/#The-three-types-of-bridges","page":"Overview","title":"The three types of bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"There are three types of bridges in MathOptInterface:","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Constraint bridges\nVariable bridges\nObjective bridges","category":"page"},{"location":"moi/submodules/Bridges/overview/#Constraint-bridges","page":"Overview","title":"Constraint bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Constraint bridges convert constraints formulated by the user into an equivalent form supported by the solver. Constraint bridges are subtypes of Bridges.Constraint.AbstractBridge.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"The equivalent formulation may add constraints (and possibly also variables) in the underlying model.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"In particular, constraint bridges can focus on rewriting the function of a constraint, and do not change the set. Function bridges are subtypes of Bridges.Constraint.AbstractFunctionConversionBridge.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Read the list of implemented constraint bridges for more details on the types of transformations that are available. Function bridges are Bridges.Constraint.ScalarFunctionizeBridge and Bridges.Constraint.VectorFunctionizeBridge.","category":"page"},{"location":"moi/submodules/Bridges/overview/#variable_bridges","page":"Overview","title":"Variable bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Variable bridges convert variables added by the user, either free with add_variable/add_variables, or constrained with add_constrained_variable/add_constrained_variables, into an equivalent form supported by the solver. Variable bridges are subtypes of Bridges.Variable.AbstractBridge.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"The equivalent formulation may add constraints (and possibly also variables) in the underlying model.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Read the list of implemented variable bridges for more details on the types of transformations that are available.","category":"page"},{"location":"moi/submodules/Bridges/overview/#Objective-bridges","page":"Overview","title":"Objective bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Objective bridges convert the ObjectiveFunction set by the user into an equivalent form supported by the solver. Objective bridges are subtypes of Bridges.Objective.AbstractBridge.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"The equivalent formulation may add constraints (and possibly also variables) in the underlying model.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Read the list of implemented objective bridges for more details on the types of transformations that are available.","category":"page"},{"location":"moi/submodules/Bridges/overview/#Bridges.full_bridge_optimizer","page":"Overview","title":"Bridges.full_bridge_optimizer","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"tip: Tip\nUnless you have an advanced use-case, this is probably the only function you need to care about.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"To enable the full power of MathOptInterface's bridges, wrap an optimizer in a Bridges.full_bridge_optimizer.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"julia> inner_optimizer = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> optimizer = MOI.Bridges.full_bridge_optimizer(inner_optimizer, Float64)\nMOIB.LazyBridgeOptimizer{MOIU.Model{Float64}}\nwith 0 variable bridges\nwith 0 constraint bridges\nwith 0 objective bridges\nwith inner model MOIU.Model{Float64}","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Now, use optimizer as normal, and bridging will happen lazily behind the scenes. By lazily, we mean that bridging will happen if and only if the constraint is not supported by the inner_optimizer.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"info: Info\nMost bridges are added by default in Bridges.full_bridge_optimizer. However, for technical reasons, some bridges are not added by default. Three examples include Bridges.Constraint.SOCtoPSDBridge, Bridges.Constraint.SOCtoNonConvexQuadBridge and Bridges.Constraint.RSOCtoNonConvexQuadBridge. See the docs of those bridges for more information.","category":"page"},{"location":"moi/submodules/Bridges/overview/#Add-a-single-bridge","page":"Overview","title":"Add a single bridge","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"If you don't want to use Bridges.full_bridge_optimizer, you can wrap an optimizer in a single bridge.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"However, this will force the constraint to be bridged, even if the inner_optimizer supports it.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"julia> inner_optimizer = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> optimizer = MOI.Bridges.Constraint.SplitInterval{Float64}(inner_optimizer)\nMOIB.Constraint.SingleBridgeOptimizer{MOIB.Constraint.SplitIntervalBridge{Float64}, MOIU.Model{Float64}}\nwith 0 constraint bridges\nwith inner model MOIU.Model{Float64}\n\njulia> x = MOI.add_variable(optimizer)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(optimizer, x, MOI.Interval(0.0, 1.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)\n\njulia> MOI.get(optimizer, MOI.ListOfConstraintTypesPresent())\n1-element Vector{Tuple{Type, Type}}:\n (MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64})\n\njulia> MOI.get(inner_optimizer, MOI.ListOfConstraintTypesPresent())\n2-element Vector{Tuple{Type, Type}}:\n (MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64})\n (MathOptInterface.VariableIndex, MathOptInterface.LessThan{Float64})","category":"page"},{"location":"moi/submodules/Bridges/overview/#Bridges.LazyBridgeOptimizer","page":"Overview","title":"Bridges.LazyBridgeOptimizer","text":"","category":"section"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"If you don't want to use Bridges.full_bridge_optimizer, but you need more than a single bridge (or you want the bridging to happen lazily), you can manually construct a Bridges.LazyBridgeOptimizer.","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"First, wrap an inner optimizer:","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"julia> inner_optimizer = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> optimizer = MOI.Bridges.LazyBridgeOptimizer(inner_optimizer)\nMOIB.LazyBridgeOptimizer{MOIU.Model{Float64}}\nwith 0 variable bridges\nwith 0 constraint bridges\nwith 0 objective bridges\nwith inner model MOIU.Model{Float64}","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Then use Bridges.add_bridge to add individual bridges:","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Bridges.add_bridge(optimizer, MOI.Bridges.Constraint.SplitIntervalBridge{Float64})\n\njulia> MOI.Bridges.add_bridge(optimizer, MOI.Bridges.Objective.FunctionizeBridge{Float64})","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"Now the constraints will be bridged only if needed:","category":"page"},{"location":"moi/submodules/Bridges/overview/","page":"Overview","title":"Overview","text":"julia> x = MOI.add_variable(optimizer)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(optimizer, x, MOI.Interval(0.0, 1.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)\n\njulia> MOI.get(optimizer, MOI.ListOfConstraintTypesPresent())\n1-element Vector{Tuple{Type, Type}}:\n (MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64})\n\njulia> MOI.get(inner_optimizer, MOI.ListOfConstraintTypesPresent())\n1-element Vector{Tuple{Type, Type}}:\n (MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64})","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"EditURL = \"getting_started_with_data_and_plotting.jl\"","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Getting-started-with-data-and-plotting","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"In this tutorial we will learn how to read tabular data into Julia, and some of the basics of plotting.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"If you're new to Julia, start by reading Getting started with Julia and Getting started with JuMP first.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"note: Note\nThere are multiple ways to read the same kind of data into Julia. This tutorial focuses on DataFrames.jl because it provides the ecosystem to work with most of the required file types in a straightforward manner.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Before we get started, we need this constant to point to where the data files are.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"import JuMP\nconst DATA_DIR = joinpath(\n    dirname(pathof(JuMP)),\n    joinpath(\"..\", \"docs\", \"src\", \"tutorials\", \"getting_started\", \"data\"),\n);\nnothing #hide","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Where-to-get-help","page":"Getting started with data and plotting","title":"Where to get help","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Read the documentation","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Plots.jl: http://docs.juliaplots.org/latest/\nCSV.jl: http://csv.juliadata.org/stable\nDataFrames.jl: https://dataframes.juliadata.org/stable/","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Preliminaries","page":"Getting started with data and plotting","title":"Preliminaries","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"To get started, we need to install some packages.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#DataFrames.jl","page":"Getting started with data and plotting","title":"DataFrames.jl","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The DataFrames package provides a set of tools for working with tabular data. It is available through the Julia package manager.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"using Pkg\nPkg.add(\"DataFrames\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"import DataFrames","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"info: What is a DataFrame?\nA DataFrame is a data structure like a table or spreadsheet. You can use it for storing and exploring a set of related data values. Think of it as a smarter array for holding tabular data.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Plots.jl","page":"Getting started with data and plotting","title":"Plots.jl","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The Plots package provides a set of tools for plotting. It is available through the Julia package manager.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"using Pkg\nPkg.add(\"Plots\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"import Plots","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#CSV-.jl","page":"Getting started with data and plotting","title":"CSV .jl","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"CSV and other delimited text files can be read by the CSV.jl package.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Pkg.add(\"CSV\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"import CSV","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#DataFrame-basics","page":"Getting started with data and plotting","title":"DataFrame basics","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"To read a CSV file into a DataFrame, we use the CSV.read function.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df = CSV.read(joinpath(DATA_DIR, \"StarWars.csv\"), DataFrames.DataFrame)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Let's try plotting some of this data","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Plots.scatter(\n    csv_df.Weight,\n    csv_df.Height;\n    xlabel = \"Weight\",\n    ylabel = \"Height\",\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"That doesn't look right. What happened? If you look at the dataframe above, it read Weight in as a String column because there are \"NA\" fields. Let's correct that, by telling CSV to consider \"NA\" as missing.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df = CSV.read(\n    joinpath(DATA_DIR, \"StarWars.csv\"),\n    DataFrames.DataFrame;\n    missingstring = \"NA\",\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Then let's re-plot our data","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Plots.scatter(\n    csv_df.Weight,\n    csv_df.Height;\n    title = \"Height vs Weight of StarWars characters\",\n    xlabel = \"Weight\",\n    ylabel = \"Height\",\n    label = false,\n    ylims = (0, 3),\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"That looks better.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"tip: Tip\nRead the CSV documentation for other parsing options.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"DataFrames.jl supports manipulation using functions similar to pandas. For example, split the dataframe into groups based on eye-color:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"by_eyecolor = DataFrames.groupby(csv_df, :Eyecolor)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Then recombine into a single dataframe based on a function operating over the split dataframes:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"eyecolor_count = DataFrames.combine(by_eyecolor) do df\n    return DataFrames.nrow(df)\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We can rename columns:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"DataFrames.rename!(eyecolor_count, :x1 => :count)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Drop some missing rows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"DataFrames.dropmissing!(eyecolor_count, :Eyecolor)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Then we can visualize the data:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"sort!(eyecolor_count, :count; rev = true)\nPlots.bar(\n    eyecolor_count.Eyecolor,\n    eyecolor_count.count;\n    xlabel = \"Eye color\",\n    ylabel = \"Number of characters\",\n    label = false,\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Other-Delimited-Files","page":"Getting started with data and plotting","title":"Other Delimited Files","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We can also use the CSV.jl package to read any other delimited text file format.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"By default, CSV.File will try to detect a file's delimiter from the first 10 lines of the file.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Candidate delimiters include ',', '\\t', ' ', '|', ';', and ':'. If it can't auto-detect the delimiter, it will assume ','.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Let's take the example of space separated data.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"ss_df = CSV.read(joinpath(DATA_DIR, \"Cereal.txt\"), DataFrames.DataFrame)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We can also specify the delimiter as follows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"delim_df = CSV.read(\n    joinpath(DATA_DIR, \"Soccer.txt\"),\n    DataFrames.DataFrame;\n    delim = \"::\",\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Working-with-DataFrames","page":"Getting started with data and plotting","title":"Working with DataFrames","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Now that we have read the required data into a DataFrame, let us look at some basic operations we can perform on it.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Querying-Basic-Information","page":"Getting started with data and plotting","title":"Querying Basic Information","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The size function gets us the dimensions of the DataFrame:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"DataFrames.size(ss_df)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We can also use the nrow and ncol functions to get the number of rows and columns respectively:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"DataFrames.nrow(ss_df), DataFrames.ncol(ss_df)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The describe function gives basic summary statistics of data in a DataFrame:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"DataFrames.describe(ss_df)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Names of every column can be obtained by the names function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"DataFrames.names(ss_df)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Corresponding data types are obtained using the broadcasted eltype function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"eltype.(ss_df)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Accessing-the-Data","page":"Getting started with data and plotting","title":"Accessing the Data","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Similar to regular arrays, we use numerical indexing to access elements of a DataFrame:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[1, 1]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The following are different ways to access a column:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[!, 1]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[!, :Name]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df.Name","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[:, 1] # Note that this creates a copy.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The following are different ways to access a row:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[1:1, :]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[1, :] # This produces a DataFrameRow.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We can change the values just as we normally assign values.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Assign a range to scalar:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[1:3, :Height] .= 1.83","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Assign a vector:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df[4:6, :Height] = [1.8, 1.6, 1.8]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"csv_df","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"tip: Tip\nThere are a lot more things which can be done with a DataFrame. Read the docs for more information.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"For information on dplyr-type syntax:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Read the DataFrames.jl documentation\nCheck out DataFramesMeta.jl","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Example:-the-passport-problem","page":"Getting started with data and plotting","title":"Example: the passport problem","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Let's now apply what we have learned to solve a real problem.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Data-manipulation","page":"Getting started with data and plotting","title":"Data manipulation","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The Passport Index Dataset lists travel visa requirements for 199 countries, in .csv format. Our task is to find the minimum number of passports required to visit all countries.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"passport_data = CSV.read(\n    joinpath(DATA_DIR, \"passport-index-matrix.csv\"),\n    DataFrames.DataFrame,\n);\nnothing #hide","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"In this dataset, the first column represents a passport (=from) and each remaining column represents a foreign country (=to).","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The values in each cell are as follows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"3 = visa-free travel\n2 = eTA is required\n1 = visa can be obtained on arrival\n0 = visa is required\n-1 is for all instances where passport and destination are the same","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Our task is to find out the minimum number of passports needed to visit every country without requiring a visa.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The values we are interested in are -1 and 3. Let's modify the dataframe so that the -1 and 3 are 1 (true), and all others are 0 (false):","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"function modifier(x)\n    if x == -1 || x == 3\n        return 1\n    else\n        return 0\n    end\nend\n\nfor country in passport_data.Passport\n    passport_data[!, country] = modifier.(passport_data[!, country])\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"The values in the cells now represent:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"1 = no visa required for travel\n0 = visa required for travel","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#JuMP-Modeling","page":"Getting started with data and plotting","title":"JuMP Modeling","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"To model the problem as a mixed-integer linear program, we need a binary decision variable x_c for each country c. x_c is 1 if we select passport c and 0 otherwise. Our objective is to minimize the sum sum x_c over all countries.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Since we wish to visit all the countries, for every country, we must own at least one passport that lets us travel to that country visa free. For one destination, this can be mathematically represented as sum_c in C a_cd cdot x_d geq 1, where a is the passport_data dataframe.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Thus, we can represent this problem using the following model:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"beginaligned\nmin  sum_c in C x_c \ntextst  sum_c in C a_cd x_c geq 1  forall d in C \n x_c in 01  forall c in C\nendaligned","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We'll now solve the problem using JuMP:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"First, create the set of countries:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"C = passport_data.Passport","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Then, create the model and initialize the decision variables:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x[C], Bin)\n@objective(model, Min, sum(x))\n@constraint(model, [d in C], passport_data[!, d]' * x >= 1)\nmodel","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Now optimize:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"optimize!(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We can use the solution_summary function to get an overview of the solution:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"solution_summary(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/#Solution","page":"Getting started with data and plotting","title":"Solution","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"Let's  have a look at the solution in more detail:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"println(\"Minimum number of passports needed: \", objective_value(model))","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"println(\"Optimal passports:\")\nfor c in C\n    if value(x[c]) > 0.5\n        println(\" * \", c)\n    end\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"We need some passports, like New Zealand and the United States, which have widespread access to a large number of countries. However, we also need passports like North Korea which only have visa-free access to a very limited number of countries.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_data_and_plotting/","page":"Getting started with data and plotting","title":"Getting started with data and plotting","text":"note: Note\nWe use value(x[c]) > 0.5 rather than value(x[c]) == 1 to avoid excluding solutions like x[c] = 0.99999 that are \"1\" to some tolerance.","category":"page"},{"location":"developers/roadmap/#Development-roadmap","page":"Roadmap","title":"Development roadmap","text":"","category":"section"},{"location":"developers/roadmap/","page":"Roadmap","title":"Roadmap","text":"The JuMP developers have compiled this roadmap document to share their plans and goals with the JuMP community. Contributions to roadmap issues are especially invited.","category":"page"},{"location":"developers/roadmap/","page":"Roadmap","title":"Roadmap","text":"Most of these issues will require changes to both JuMP and MathOptInterface, and are non-trivial in their implementation. They are in no particular order, but represent broad themes that we see as areas in which JuMP could be improved.","category":"page"},{"location":"developers/roadmap/","page":"Roadmap","title":"Roadmap","text":"Done #3106 Make nonlinear programming a first-class citizen. There have been many issues and discussions about this: currently nonlinear constraints are handled through a MOI.NLPBlock and have various limitations and restrictions.\nhttps://github.com/jump-dev/JuMP.jl/issues/1185\nhttps://github.com/jump-dev/JuMP.jl/issues/1198\nhttps://github.com/jump-dev/JuMP.jl/issues/2788\nhttps://github.com/jump-dev/MathOptInterface.jl/issues/846\nhttps://github.com/jump-dev/MathOptInterface.jl/issues/1397\nDone #3385 Add support for coefficient types other than Float64: https://github.com/jump-dev/JuMP.jl/issues/2025 Since the very beginning, JuMP has hard-coded the coefficient type as Float64. This has made it impossible to support solvers which can use other types such as BigFloat or Rational{BigInt}.\nAdd support for constraint programming: https://github.com/jump-dev/JuMP.jl/issues/2227 JuMP has a strong focus on linear, conic and nonlinear optimization problems. We want to add better support for constraint programming.\nDone #3176 Add support for multiobjective problems: https://github.com/jump-dev/JuMP.jl/issues/2099 JuMP is restricted to problems with scalar-valued objectives. We want to extend this to vector-valued problems.\nRefactor the internal code of JuMP's macros. The code in src/macros.jl is some of the oldest part of JuMP and is difficult to read, modify, and extend. We should overhaul the internals of JuMP's macros–-without making user-visible breaking changes–-to improve their long-term maintainability.","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"EditURL = \"simple_examples.jl\"","category":"page"},{"location":"tutorials/nonlinear/simple_examples/#Simple-examples","page":"Simple examples","title":"Simple examples","text":"","category":"section"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"This tutorial is a collection of examples of small nonlinear programs. It uses the following packages:","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"using JuMP\nimport Ipopt\nimport Random\nimport Statistics\nimport Test","category":"page"},{"location":"tutorials/nonlinear/simple_examples/#The-Rosenbrock-function","page":"Simple examples","title":"The Rosenbrock function","text":"","category":"section"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"A nonlinear example of the classical Rosenbrock function.","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"function example_rosenbrock()\n    model = Model(Ipopt.Optimizer)\n    set_silent(model)\n    @variable(model, x)\n    @variable(model, y)\n    @objective(model, Min, (1 - x)^2 + 100 * (y - x^2)^2)\n    optimize!(model)\n    Test.@test termination_status(model) == LOCALLY_SOLVED\n    Test.@test primal_status(model) == FEASIBLE_POINT\n    Test.@test objective_value(model) ≈ 0.0 atol = 1e-10\n    Test.@test value(x) ≈ 1.0\n    Test.@test value(y) ≈ 1.0\n    return\nend\n\nexample_rosenbrock()","category":"page"},{"location":"tutorials/nonlinear/simple_examples/#The-clnlbeam-problem","page":"Simple examples","title":"The clnlbeam problem","text":"","category":"section"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"Based on an AMPL model by Hande Y. Benson","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"Copyright (C) 2001 Princeton University All Rights Reserved","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"Permission to use, copy, modify, and distribute this software and its documentation for any purpose and without fee is hereby granted, provided that the above copyright notice appear in all copies and that the copyright notice and this permission notice appear in all supporting documentation.","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"Source:","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"H. Maurer and H.D. Mittelman, \"The non-linear beam via optimal control with bound state variables,\" Optimal Control Applications and Methods 12, pp. 19-31, 1991.","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"function example_clnlbeam()\n    N = 1000\n    h = 1 / N\n    alpha = 350\n    model = Model(Ipopt.Optimizer)\n    @variables(model, begin\n        -1 <= t[1:(N+1)] <= 1\n        -0.05 <= x[1:(N+1)] <= 0.05\n        u[1:(N+1)]\n    end)\n    @objective(\n        model,\n        Min,\n        sum(\n            0.5 * h * (u[i+1]^2 + u[i]^2) +\n            0.5 * alpha * h * (cos(t[i+1]) + cos(t[i])) for i in 1:N\n        ),\n    )\n    @constraint(\n        model,\n        [i = 1:N],\n        x[i+1] - x[i] - 0.5 * h * (sin(t[i+1]) + sin(t[i])) == 0,\n    )\n    @constraint(\n        model,\n        [i = 1:N],\n        t[i+1] - t[i] - 0.5 * h * u[i+1] - 0.5 * h * u[i] == 0,\n    )\n    optimize!(model)\n    println(\"\"\"\n    termination_status = $(termination_status(model))\n    primal_status      = $(primal_status(model))\n    objective_value    = $(objective_value(model))\n    \"\"\")\n    return\nend\n\nexample_clnlbeam()","category":"page"},{"location":"tutorials/nonlinear/simple_examples/#Maximum-likelihood-estimation","page":"Simple examples","title":"Maximum likelihood estimation","text":"","category":"section"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"This example uses nonlinear optimization to compute the maximum likelihood estimate (MLE) of the parameters of a normal distribution, a.k.a., the sample mean and variance.","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"function example_mle()\n    n = 1_000\n    Random.seed!(1234)\n    data = randn(n)\n    model = Model(Ipopt.Optimizer)\n    set_silent(model)\n    @variable(model, μ, start = 0.0)\n    @variable(model, σ >= 0.0, start = 1.0)\n    @objective(\n        model,\n        Max,\n        n / 2 * log(1 / (2 * π * σ^2)) -\n        sum((data[i] - μ)^2 for i in 1:n) / (2 * σ^2)\n    )\n    optimize!(model)\n    println(\"μ             = \", value(μ))\n    println(\"mean(data)    = \", Statistics.mean(data))\n    println(\"σ^2           = \", value(σ)^2)\n    println(\"var(data)     = \", Statistics.var(data))\n    println(\"MLE objective = \", objective_value(model))\n    Test.@test value(μ) ≈ Statistics.mean(data) atol = 1e-3\n    Test.@test value(σ)^2 ≈ Statistics.var(data) atol = 1e-2\n    # You can even do constrained MLE!\n    @constraint(model, μ == σ^2)\n    optimize!(model)\n    Test.@test value(μ) ≈ value(σ)^2\n    println()\n    println(\"With constraint μ == σ^2:\")\n    println(\"μ                         = \", value(μ))\n    println(\"σ^2                       = \", value(σ)^2)\n    println(\"Constrained MLE objective = \", objective_value(model))\n    return\nend\n\nexample_mle()","category":"page"},{"location":"tutorials/nonlinear/simple_examples/#Quadratically-constrained-programs","page":"Simple examples","title":"Quadratically constrained programs","text":"","category":"section"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"A simple quadratically constrained program based on an example from Gurobi.","category":"page"},{"location":"tutorials/nonlinear/simple_examples/","page":"Simple examples","title":"Simple examples","text":"function example_qcp()\n    model = Model(Ipopt.Optimizer)\n    set_silent(model)\n    @variable(model, x)\n    @variable(model, y >= 0)\n    @variable(model, z >= 0)\n    @objective(model, Max, x)\n    @constraint(model, x + y + z == 1)\n    @constraint(model, x * x + y * y - z * z <= 0)\n    @constraint(model, x * x - y * z <= 0)\n    optimize!(model)\n    print(model)\n    println(\"Objective value: \", objective_value(model))\n    println(\"x = \", value(x))\n    println(\"y = \", value(y))\n    Test.@test termination_status(model) == LOCALLY_SOLVED\n    Test.@test primal_status(model) == FEASIBLE_POINT\n    Test.@test objective_value(model) ≈ 0.32699 atol = 1e-5\n    Test.@test value(x) ≈ 0.32699 atol = 1e-5\n    Test.@test value(y) ≈ 0.25707 atol = 1e-5\n    return\nend\n\nexample_qcp()","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"EditURL = \"https://github.com/jump-dev/GLPK.jl/blob/v1.1.2/README.md\"","category":"page"},{"location":"packages/GLPK/#GLPK.jl","page":"jump-dev/GLPK.jl","title":"GLPK.jl","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"GLPK.jl is a wrapper for the GNU Linear Programming Kit library.","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/GLPK/#Affiliation","page":"jump-dev/GLPK.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"This wrapper is maintained by the JuMP community and is not an GNU project.","category":"page"},{"location":"packages/GLPK/#License","page":"jump-dev/GLPK.jl","title":"License","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"GLPK.jl is licensed under the GPL v3 license.","category":"page"},{"location":"packages/GLPK/#Installation","page":"jump-dev/GLPK.jl","title":"Installation","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"Install GLPK using Pkg.add:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"import Pkg\nPkg.add(\"GLPK\")","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"In addition to installing the GLPK.jl package, this will also download and install the GLPK binaries. You do not need to install GLPK separately.","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"To use a custom binary, read the Custom solver binaries section of the JuMP documentation.","category":"page"},{"location":"packages/GLPK/#Use-with-JuMP","page":"jump-dev/GLPK.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"To use GLPK with JuMP, use GLPK.Optimizer:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"using JuMP, GLPK\nmodel = Model(GLPK.Optimizer)\nset_attribute(model, \"tm_lim\", 60 * 1_000)\nset_attribute(model, \"msg_lev\", GLPK.GLP_MSG_OFF)","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"If the model is primal or dual infeasible, GLPK will attempt to find a certificate of infeasibility. This can be expensive, particularly if you do not intend to use the certificate. If this is the case, use:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"model = Model(() -> GLPK.Optimizer(; want_infeasibility_certificates = false))","category":"page"},{"location":"packages/GLPK/#MathOptInterface-API","page":"jump-dev/GLPK.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"The GLPK optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"MOI.Reals","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.ZeroOne","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"MOI.HeuristicCallback()\nMOI.LazyConstraintCallback()\nMOI.Name()\nMOI.ObjectiveSense()\nMOI.UserCutCallback()","category":"page"},{"location":"packages/GLPK/#Options","page":"jump-dev/GLPK.jl","title":"Options","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"Options for GLPK are comprehensively documented in the PDF documentation, but they are hard to find.","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"Options when solving a linear program are defined in Section 2.8.1\nOptions when solving a mixed-integer program are defined in Section 2.10.5","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"However, the following options are likely to be the most useful:","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"Parameter Example Explanation\nmsg_lev GLPK.GLP_MSG_ALL Message level for terminal output\npresolve GLPK.GLP_ON Turn presolve on or off\ntol_int 1e-5 Absolute tolerance for integer feasibility\ntol_obj 1e-7 Relative objective tolerance for mixed-integer programs","category":"page"},{"location":"packages/GLPK/#Callbacks","page":"jump-dev/GLPK.jl","title":"Callbacks","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"Here is an example using GLPK's solver-specific callbacks.","category":"page"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"using JuMP, GLPK, Test\n\nmodel = Model(GLPK.Optimizer)\n@variable(model, 0 <= x <= 2.5, Int)\n@variable(model, 0 <= y <= 2.5, Int)\n@objective(model, Max, y)\nreasons = UInt8[]\nfunction my_callback_function(cb_data)\n    reason = GLPK.glp_ios_reason(cb_data.tree)\n    push!(reasons, reason)\n    if reason != GLPK.GLP_IROWGEN\n        return\n    end\n    x_val = callback_value(cb_data, x)\n    y_val = callback_value(cb_data, y)\n    if y_val - x_val > 1 + 1e-6\n        con = @build_constraint(y - x <= 1)\n        MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n    elseif y_val + x_val > 3 + 1e-6\n        con = @build_constraint(y - x <= 1)\n        MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n    end\nend\nMOI.set(model, GLPK.CallbackFunction(), my_callback_function)\noptimize!(model)\n@test termination_status(model) == MOI.OPTIMAL\n@test primal_status(model) == MOI.FEASIBLE_POINT\n@test value(x) == 1\n@test value(y) == 2\n@show reasons","category":"page"},{"location":"packages/GLPK/#C-API","page":"jump-dev/GLPK.jl","title":"C API","text":"","category":"section"},{"location":"packages/GLPK/","page":"jump-dev/GLPK.jl","title":"jump-dev/GLPK.jl","text":"The C API can be accessed via GLPK.glp_XXX functions, where the names and arguments are identical to the C API. See the /tests folder for inspiration.","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"EditURL = \"callbacks.jl\"","category":"page"},{"location":"tutorials/linear/callbacks/#callbacks_tutorial","page":"Callbacks","title":"Callbacks","text":"","category":"section"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"The purpose of the tutorial is to demonstrate the various solver-independent and solver-dependent callbacks that are supported by JuMP.","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"The tutorial uses the following packages:","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"using JuMP\nimport GLPK\nimport Random","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"info: Info\nThis tutorial uses the MathOptInterface API. By default, JuMP exports the MOI symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:import MathOptInterface as MOI","category":"page"},{"location":"tutorials/linear/callbacks/#Lazy-constraints","page":"Callbacks","title":"Lazy constraints","text":"","category":"section"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"An example using a lazy constraint callback.","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"function example_lazy_constraint()\n    model = Model(GLPK.Optimizer)\n    @variable(model, 0 <= x <= 2.5, Int)\n    @variable(model, 0 <= y <= 2.5, Int)\n    @objective(model, Max, y)\n    lazy_called = false\n    function my_callback_function(cb_data)\n        lazy_called = true\n        x_val = callback_value(cb_data, x)\n        y_val = callback_value(cb_data, y)\n        println(\"Called from (x, y) = ($x_val, $y_val)\")\n        status = callback_node_status(cb_data, model)\n        if status == MOI.CALLBACK_NODE_STATUS_FRACTIONAL\n            println(\" - Solution is integer infeasible!\")\n        elseif status == MOI.CALLBACK_NODE_STATUS_INTEGER\n            println(\" - Solution is integer feasible!\")\n        else\n            @assert status == MOI.CALLBACK_NODE_STATUS_UNKNOWN\n            println(\" - I don't know if the solution is integer feasible :(\")\n        end\n        if y_val - x_val > 1 + 1e-6\n            con = @build_constraint(y - x <= 1)\n            println(\"Adding $(con)\")\n            MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n        elseif y_val + x_val > 3 + 1e-6\n            con = @build_constraint(y + x <= 3)\n            println(\"Adding $(con)\")\n            MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n        end\n    end\n    set_attribute(model, MOI.LazyConstraintCallback(), my_callback_function)\n    optimize!(model)\n    println(\"Optimal solution (x, y) = ($(value(x)), $(value(y)))\")\n    return\nend\n\nexample_lazy_constraint()","category":"page"},{"location":"tutorials/linear/callbacks/#User-cuts","page":"Callbacks","title":"User-cuts","text":"","category":"section"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"An example using a user-cut callback.","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"function example_user_cut_constraint()\n    Random.seed!(1)\n    N = 30\n    item_weights, item_values = rand(N), rand(N)\n    model = Model(GLPK.Optimizer)\n    @variable(model, x[1:N], Bin)\n    @constraint(model, sum(item_weights[i] * x[i] for i in 1:N) <= 10)\n    @objective(model, Max, sum(item_values[i] * x[i] for i in 1:N))\n    callback_called = false\n    function my_callback_function(cb_data)\n        callback_called = true\n        x_vals = callback_value.(Ref(cb_data), x)\n        accumulated = sum(item_weights[i] for i in 1:N if x_vals[i] > 1e-4)\n        println(\"Called with accumulated = $(accumulated)\")\n        n_terms = sum(1 for i in 1:N if x_vals[i] > 1e-4)\n        if accumulated > 10\n            con = @build_constraint(\n                sum(x[i] for i in 1:N if x_vals[i] > 0.5) <= n_terms - 1\n            )\n            println(\"Adding $(con)\")\n            MOI.submit(model, MOI.UserCut(cb_data), con)\n        end\n    end\n    set_attribute(model, MOI.UserCutCallback(), my_callback_function)\n    optimize!(model)\n    @show callback_called\n    return\nend\n\nexample_user_cut_constraint()","category":"page"},{"location":"tutorials/linear/callbacks/#Heuristic-solutions","page":"Callbacks","title":"Heuristic solutions","text":"","category":"section"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"An example using a heuristic solution callback.","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"function example_heuristic_solution()\n    Random.seed!(1)\n    N = 30\n    item_weights, item_values = rand(N), rand(N)\n    model = Model(GLPK.Optimizer)\n    @variable(model, x[1:N], Bin)\n    @constraint(model, sum(item_weights[i] * x[i] for i in 1:N) <= 10)\n    @objective(model, Max, sum(item_values[i] * x[i] for i in 1:N))\n    callback_called = false\n    function my_callback_function(cb_data)\n        callback_called = true\n        x_vals = callback_value.(Ref(cb_data), x)\n        ret =\n            MOI.submit(model, MOI.HeuristicSolution(cb_data), x, floor.(x_vals))\n        println(\"Heuristic solution status = $(ret)\")\n    end\n    set_attribute(model, MOI.HeuristicCallback(), my_callback_function)\n    optimize!(model)\n    return\nend\n\nexample_heuristic_solution()","category":"page"},{"location":"tutorials/linear/callbacks/#GLPK-solver-dependent-callback","page":"Callbacks","title":"GLPK solver-dependent callback","text":"","category":"section"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"An example using GLPK's solver-dependent callback.","category":"page"},{"location":"tutorials/linear/callbacks/","page":"Callbacks","title":"Callbacks","text":"function example_solver_dependent_callback()\n    model = Model(GLPK.Optimizer)\n    @variable(model, 0 <= x <= 2.5, Int)\n    @variable(model, 0 <= y <= 2.5, Int)\n    @objective(model, Max, y)\n    lazy_called = false\n    function my_callback_function(cb_data)\n        lazy_called = true\n        reason = GLPK.glp_ios_reason(cb_data.tree)\n        println(\"Called from reason = $(reason)\")\n        if reason != GLPK.GLP_IROWGEN\n            return\n        end\n        x_val = callback_value(cb_data, x)\n        y_val = callback_value(cb_data, y)\n        if y_val - x_val > 1 + 1e-6\n            con = @build_constraint(y - x <= 1)\n            println(\"Adding $(con)\")\n            MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n        elseif y_val + x_val > 3 + 1e-6\n            con = @build_constraint(y - x <= 1)\n            println(\"Adding $(con)\")\n            MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n        end\n    end\n    set_attribute(model, GLPK.CallbackFunction(), my_callback_function)\n    optimize!(model)\n    return\nend\n\nexample_solver_dependent_callback()","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/nlp/#Nonlinear-Modeling-(Legacy)","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"warning: Warning\nThis page describes the legacy nonlinear interface to JuMP. It has a number of quirks and limitations that prompted the development of a new nonlinear interface. The new interface is documented at Nonlinear Modeling. This legacy interface will remain for all future v1.X releases of JuMP. The two nonlinear interfaces cannot be combined.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"JuMP has support for general smooth nonlinear (convex and nonconvex) optimization problems. JuMP is able to provide exact, sparse second-order derivatives to solvers. This information can improve solver accuracy and performance.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"There are three main changes to solve nonlinear programs in JuMP.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use @NLobjective instead of @objective\nUse @NLconstraint instead of @constraint\nUse @NLexpression instead of @expression","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"info: Info\nThere are some restrictions on what syntax you can use in the @NLxxx macros. Make sure to read the Syntax notes.","category":"page"},{"location":"manual/nlp/#Set-a-nonlinear-objective","page":"Nonlinear Modeling (Legacy)","title":"Set a nonlinear objective","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use @NLobjective to set a nonlinear objective.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @NLobjective(model, Min, exp(x[1]) - sqrt(x[2]))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"To modify a nonlinear objective, call @NLobjective again.","category":"page"},{"location":"manual/nlp/#Add-a-nonlinear-constraint","page":"Nonlinear Modeling (Legacy)","title":"Add a nonlinear constraint","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use @NLconstraint to add a nonlinear constraint.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @NLconstraint(model, exp(x[1]) <= 1)\nexp(x[1]) - 1.0 ≤ 0\n\njulia> @NLconstraint(model, [i = 1:2], x[i]^i >= i)\n2-element Vector{NonlinearConstraintRef{ScalarShape}}:\n x[1] ^ 1.0 - 1.0 ≥ 0\n x[2] ^ 2.0 - 2.0 ≥ 0\n\njulia> @NLconstraint(model, con[i = 1:2], prod(x[j] for j = 1:i) == i)\n2-element Vector{NonlinearConstraintRef{ScalarShape}}:\n (*)(x[1]) - 1.0 = 0\n x[1] * x[2] - 2.0 = 0","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"info: Info\nYou can only create nonlinear constraints with <=, >=, and ==. More general Nonlinear-in-Set constraints are not supported.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Delete a nonlinear constraint using delete:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> delete(model, con[1])","category":"page"},{"location":"manual/nlp/#Create-a-nonlinear-expression","page":"Nonlinear Modeling (Legacy)","title":"Create a nonlinear expression","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use @NLexpression to create nonlinear expression objects. The syntax is identical to @expression, except that the expression can contain nonlinear terms.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> expr = @NLexpression(model, exp(x[1]) + sqrt(x[2]))\nsubexpression[1]: exp(x[1]) + sqrt(x[2])\n\njulia> my_anon_expr = @NLexpression(model, [i = 1:2], sin(x[i]))\n2-element Vector{NonlinearExpression}:\n subexpression[2]: sin(x[1])\n subexpression[3]: sin(x[2])\n\njulia> @NLexpression(model, my_expr[i = 1:2], sin(x[i]))\n2-element Vector{NonlinearExpression}:\n subexpression[4]: sin(x[1])\n subexpression[5]: sin(x[2])","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Nonlinear expression can be used in @NLobjective, @NLconstraint, and even nested in other @NLexpressions.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> @NLobjective(model, Min, expr^2 + 1)\n\njulia> @NLconstraint(model, [i = 1:2], my_expr[i] <= i)\n2-element Vector{NonlinearConstraintRef{ScalarShape}}:\n subexpression[4] - 1.0 ≤ 0\n subexpression[5] - 2.0 ≤ 0\n\njulia> @NLexpression(model, nested[i = 1:2], sin(my_expr[i]))\n2-element Vector{NonlinearExpression}:\n subexpression[6]: sin(subexpression[4])\n subexpression[7]: sin(subexpression[5])","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use value to query the value of a nonlinear expression:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> set_start_value(x[1], 1.0)\n\njulia> value(start_value, nested[1])\n0.7456241416655579\n\njulia> sin(sin(1.0))\n0.7456241416655579","category":"page"},{"location":"manual/nlp/#Create-a-nonlinear-parameter","page":"Nonlinear Modeling (Legacy)","title":"Create a nonlinear parameter","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"For nonlinear models only, JuMP offers a syntax for explicit \"parameter\" objects, which are constants in the model that can be efficiently updated between solves.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Nonlinear parameters are declared by using the @NLparameter macro and may be indexed by arbitrary sets analogously to JuMP variables and expressions.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The initial value of the parameter must be provided on the right-hand side of the == sign.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @NLparameter(model, p[i = 1:2] == i)\n2-element Vector{NonlinearParameter}:\n parameter[1] == 1.0\n parameter[2] == 2.0","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Create anonymous parameters using the value keyword:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> anon_parameter = @NLparameter(model, value = 1)\nparameter[3] == 1.0","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"info: Info\nA parameter is not an optimization variable. It must be fixed to a value with ==. If you want a parameter that is <= or >=, create a variable instead using @variable.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use value and set_value to query or update the value of a parameter.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> value.(p)\n2-element Vector{Float64}:\n 1.0\n 2.0\n\njulia> set_value(p[2], 3.0)\n3.0\n\njulia> value.(p)\n2-element Vector{Float64}:\n 1.0\n 3.0","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Nonlinear parameters must be used within nonlinear macros only.","category":"page"},{"location":"manual/nlp/#When-to-use-a-parameter","page":"Nonlinear Modeling (Legacy)","title":"When to use a parameter","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Nonlinear parameters are useful when solving nonlinear models in a sequence:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"using JuMP, Ipopt\nmodel = Model(Ipopt.Optimizer)\nset_silent(model)\n@variable(model, z)\n@NLparameter(model, x == 1.0)\n@NLobjective(model, Min, (z - x)^2)\noptimize!(model)\n@show value(z) # Equals 1.0.\n\n# Now, update the value of x to solve a different problem.\nset_value(x, 5.0)\noptimize!(model)\n@show value(z) # Equals 5.0\nnothing #hide","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"info: Info\nUsing nonlinear parameters can be faster than creating a new model from scratch with updated data because JuMP is able to avoid repeating a number of steps in processing the model before handing it off to the solver.","category":"page"},{"location":"manual/nlp/#Syntax-notes","page":"Nonlinear Modeling (Legacy)","title":"Syntax notes","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The syntax accepted in nonlinear macros is more restricted than the syntax for linear and quadratic macros. We note some important points below.","category":"page"},{"location":"manual/nlp/#Scalar-operations-only","page":"Nonlinear Modeling (Legacy)","title":"Scalar operations only","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Except for the splatting syntax discussed below, all expressions must be simple scalar operations. You cannot use dot, matrix-vector products, vector slices, etc.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @variable(model, y);\n\njulia> c = [1, 2];\n\njulia> @NLobjective(model, Min, c' * x + 3y)\nERROR: Unexpected array [1 2] in nonlinear expression. Nonlinear expressions may contain only scalar expressions.\n[...]","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Translate vector operations into explicit sum() operations:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> @NLobjective(model, Min, sum(c[i] * x[i] for i = 1:2) + 3y)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Or use an @expression:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> @expression(model, expr, c' * x)\nx[1] + 2 x[2]\n\njulia> @NLobjective(model, Min, expr + 3y)\n","category":"page"},{"location":"manual/nlp/#Splatting","page":"Nonlinear Modeling (Legacy)","title":"Splatting","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The splatting operator   ... is recognized in a very restricted setting for expanding function   arguments. The expression splatted can be only a symbol. More complex   expressions are not recognized.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x[1:3]);\n\njulia> @NLconstraint(model, *(x...) <= 1.0)\nx[1] * x[2] * x[3] - 1.0 ≤ 0\n\njulia> @NLconstraint(model, *((x / 2)...) <= 0.0)\nERROR: Unsupported use of the splatting operator. JuMP supports splatting only symbols. For example, `x...` is ok, but `(x + 1)...`, `[x; y]...` and `g(f(y)...)` are not.","category":"page"},{"location":"manual/nlp/#User-defined-Functions","page":"Nonlinear Modeling (Legacy)","title":"User-defined Functions","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"JuMP natively supports the set of univariate and multivariate functions recognized by the MOI.Nonlinear submodule. In addition to this list of functions, it is possible to register custom user-defined nonlinear functions. User-defined functions can be used anywhere in @NLobjective, @NLconstraint, and @NLexpression.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"JuMP will attempt to automatically register functions it detects in your nonlinear expressions, which usually means manually registering a function is not needed. Two exceptions are if you want to provide custom derivatives, or if the function is not available in the scope of the nonlinear expression.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"warning: Warning\nUser-defined functions must return a scalar output. For a work-around, see User-defined operators with vector outputs.","category":"page"},{"location":"manual/nlp/#Automatic-differentiation","page":"Nonlinear Modeling (Legacy)","title":"Automatic differentiation","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"JuMP does not support black-box optimization, so all user-defined functions must provide derivatives in some form. Fortunately, JuMP supports automatic differentiation of user-defined functions, a feature to our knowledge not available in any comparable modeling systems.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"info: Info\nAutomatic differentiation is not finite differencing. JuMP's automatically computed derivatives are not subject to approximation error.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"JuMP uses ForwardDiff.jl to perform automatic differentiation; see the ForwardDiff.jl documentation for a description of how to write a function suitable for automatic differentiation.","category":"page"},{"location":"manual/nlp/#Common-mistakes-when-writing-a-user-defined-function","page":"Nonlinear Modeling (Legacy)","title":"Common mistakes when writing a user-defined function","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"warning: Warning\nGet an error like No method matching Float64(::ForwardDiff.Dual)? Read this section, and see the guidelines at ForwardDiff.jl.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The most common error is that your user-defined function is not generic with respect to the number type, that is, don't assume that the input to the function is Float64.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"f(x::Float64) = 2 * x  # This will not work.\nf(x::Real)    = 2 * x  # This is good.\nf(x)          = 2 * x  # This is also good.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Another reason you may encounter this error is if you create arrays inside your function which are Float64.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"function bad_f(x...)\n    y = zeros(length(x))  # This constructs an array of `Float64`!\n    for i = 1:length(x)\n        y[i] = x[i]^i\n    end\n    return sum(y)\nend\n\nfunction good_f(x::T...) where {T<:Real}\n    y = zeros(T, length(x))  # Construct an array of type `T` instead!\n    for i = 1:length(x)\n        y[i] = x[i]^i\n    end\n    return sum(y)\nend","category":"page"},{"location":"manual/nlp/#Register-a-function","page":"Nonlinear Modeling (Legacy)","title":"Register a function","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"To register a user-defined function with derivatives computed by automatic differentiation, use the register method as in the following example:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"using JuMP #hide\nsquare(x) = x^2\nf(x, y) = (x - 1)^2 + (y - 2)^2\n\nmodel = Model()\n\nregister(model, :square, 1, square; autodiff = true)\nregister(model, :my_f, 2, f; autodiff = true)\n\n@variable(model, x[1:2] >= 0.5)\n@NLobjective(model, Min, my_f(x[1], square(x[2])))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The above code creates a JuMP model with the objective function (x[1] - 1)^2 + (x[2]^2 - 2)^2. The arguments to register are:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The model for which the functions are registered.\nA Julia symbol object which serves as the name of the user-defined function in JuMP expressions.\nThe number of input arguments that the function takes.\nThe Julia method which computes the function\nA flag to instruct JuMP to compute exact gradients automatically.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"tip: Tip\nThe symbol :my_f doesn't have to match the name of the function f. However, it's more readable if it does. Make sure you use my_f and not f in the macros.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"warning: Warning\nUser-defined functions cannot be re-registered and will not update if you modify the underlying Julia function. If you want to change a user-defined function between solves, rebuild the model or use a different name. To use a different name programmatically, see Raw expression input.","category":"page"},{"location":"manual/nlp/#Register-a-function-and-gradient","page":"Nonlinear Modeling (Legacy)","title":"Register a function and gradient","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Forward-mode automatic differentiation as implemented by ForwardDiff.jl has a computational cost that scales linearly with the number of input dimensions. As such, it is not the most efficient way to compute gradients of user-defined functions if the number of input arguments is large. In this case, users may want to provide their own routines for evaluating gradients.","category":"page"},{"location":"manual/nlp/#Univariate-functions","page":"Nonlinear Modeling (Legacy)","title":"Univariate functions","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"For univariate functions, the gradient function ∇f returns a number that represents the first-order derivative:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"using JuMP #hide\nf(x) = x^2\n∇f(x) = 2x\nmodel = Model()\nregister(model, :my_square, 1, f, ∇f; autodiff = true)\n@variable(model, x >= 0)\n@NLobjective(model, Min, my_square(x))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"If autodiff = true, JuMP will use automatic differentiation to compute the hessian.","category":"page"},{"location":"manual/nlp/#Multivariate-functions","page":"Nonlinear Modeling (Legacy)","title":"Multivariate functions","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"For multivariate functions, the gradient function ∇f must take a gradient vector as the first argument that is filled in-place:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"using JuMP #hide\nf(x, y) = (x - 1)^2 + (y - 2)^2\nfunction ∇f(g::AbstractVector{T}, x::T, y::T) where {T}\n    g[1] = 2 * (x - 1)\n    g[2] = 2 * (y - 2)\n    return\nend\n\nmodel = Model()\nregister(model, :my_square, 2, f, ∇f)\n@variable(model, x[1:2] >= 0)\n@NLobjective(model, Min, my_square(x[1], x[2]))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"warning: Warning\nMake sure the first argument to ∇f supports an AbstractVector, and do not assume the input is Float64.","category":"page"},{"location":"manual/nlp/#Register-a-function,-gradient,-and-hessian","page":"Nonlinear Modeling (Legacy)","title":"Register a function, gradient, and hessian","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"You can also register a function with the second-order derivative information, which is a scalar for univariate functions, and a symmetric matrix for multivariate functions.","category":"page"},{"location":"manual/nlp/#Univariate-functions-2","page":"Nonlinear Modeling (Legacy)","title":"Univariate functions","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Pass a function which returns a number representing the second-order derivative:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"using JuMP #hide\nf(x) = x^2\n∇f(x) = 2x\n∇²f(x) = 2\nmodel = Model()\nregister(model, :my_square, 1, f, ∇f, ∇²f)\n@variable(model, x >= 0)\n@NLobjective(model, Min, my_square(x))","category":"page"},{"location":"manual/nlp/#Multivariate-functions-2","page":"Nonlinear Modeling (Legacy)","title":"Multivariate functions","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"For multivariate functions, the hessian function ∇²f must take an AbstractMatrix as the first argument, the lower-triangular of which is filled in-place:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"using JuMP #hide\nf(x...) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2\nfunction ∇f(g, x...)\n    g[1] = 400 * x[1]^3 - 400 * x[1] * x[2] + 2 * x[1] - 2\n    g[2] = 200 * (x[2] - x[1]^2)\n    return\nend\nfunction ∇²f(H, x...)\n    H[1, 1] = 1200 * x[1]^2 - 400 * x[2] + 2\n    # H[1, 2] = -400 * x[1]  <-- Not needed. Fill the lower-triangular only.\n    H[2, 1] = -400 * x[1]\n    H[2, 2] = 200.0\n    return\nend\n\nmodel = Model()\nregister(model, :rosenbrock, 2, f, ∇f, ∇²f)\n@variable(model, x[1:2])\n@NLobjective(model, Min, rosenbrock(x[1], x[2]))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"warning: Warning\nYou may assume the Hessian matrix H is initialized with zeros, and because H is symmetric, you need only to fill in the non-zero of the lower-triangular terms. The matrix type passed in as H depends on the automatic differentiation system, so make sure the first argument to the Hessian function supports an AbstractMatrix (it may be something other than Matrix{Float64}). However, you may assume only that H supports size(H) and setindex!. Finally, the matrix is treated as dense, so the performance will be poor on functions with high-dimensional input.","category":"page"},{"location":"manual/nlp/#User-defined-functions-with-vector-inputs","page":"Nonlinear Modeling (Legacy)","title":"User-defined functions with vector inputs","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"User-defined functions which take vectors as input arguments (for example, f(x::Vector)) are not supported. Instead, use Julia's splatting syntax to create a function with scalar arguments. For example, instead of","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"f(x::Vector) = sum(x[i]^i for i in 1:length(x))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"define:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"f(x...) = sum(x[i]^i for i in 1:length(x))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"This function f can be used in a JuMP model as follows:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"using JuMP #hide\nmodel = Model()\n@variable(model, x[1:5] >= 0)\nf(x...) = sum(x[i]^i for i in 1:length(x))\nregister(model, :f, 5, f; autodiff = true)\n@NLobjective(model, Min, f(x...))","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"tip: Tip\nMake sure to read the syntax restrictions of Splatting.","category":"page"},{"location":"manual/nlp/#Factors-affecting-solution-time","page":"Nonlinear Modeling (Legacy)","title":"Factors affecting solution time","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The execution time when solving a nonlinear programming problem can be divided into two parts, the time spent in the optimization algorithm (the solver) and the time spent evaluating the nonlinear functions and corresponding derivatives. Ipopt explicitly displays these two timings in its output, for example:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Total CPU secs in IPOPT (w/o function evaluations)   =      7.412\nTotal CPU secs in NLP function evaluations           =      2.083","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"For Ipopt in particular, one can improve the performance by installing advanced sparse linear algebra packages, see Installation Guide. For other solvers, see their respective documentation for performance tips.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The function evaluation time, on the other hand, is the responsibility of the modeling language. JuMP computes derivatives by using reverse-mode automatic differentiation with graph coloring methods for exploiting sparsity of the Hessian matrix. As a conservative bound, JuMP's performance here currently may be expected to be within a factor of 5 of AMPL's. Our paper in SIAM Review has more details.","category":"page"},{"location":"manual/nlp/#Querying-derivatives-from-a-JuMP-model","page":"Nonlinear Modeling (Legacy)","title":"Querying derivatives from a JuMP model","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"For some advanced use cases, one may want to directly query the derivatives of a JuMP model instead of handing the problem off to a solver. Internally, JuMP implements the MOI.AbstractNLPEvaluator interface. To obtain an NLP evaluator object from a JuMP model, use NLPEvaluator. index returns the MOI.VariableIndex corresponding to a JuMP variable. MOI.VariableIndex itself is a type-safe wrapper for Int64 (stored in the .value field.)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"For example:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> raw_index(v::MOI.VariableIndex) = v.value\nraw_index (generic function with 1 method)\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, y)\ny\n\njulia> @NLobjective(model, Min, sin(x) + sin(y))\n\njulia> values = zeros(2)\n2-element Vector{Float64}:\n 0.0\n 0.0\n\njulia> x_index = raw_index(JuMP.index(x))\n1\n\njulia> y_index = raw_index(JuMP.index(y))\n2\n\njulia> values[x_index] = 2.0\n2.0\n\njulia> values[y_index] = 3.0\n3.0\n\njulia> d = NLPEvaluator(model)\nNonlinear.Evaluator with available features:\n  * :Grad\n  * :Jac\n  * :JacVec\n  * :Hess\n  * :HessVec\n  * :ExprGraph\n\njulia> MOI.initialize(d, [:Grad])\n\njulia> MOI.eval_objective(d, values)\n1.0504174348855488\n\njulia> sin(2.0) + sin(3.0)\n1.0504174348855488\n\njulia> ∇f = zeros(2)\n2-element Vector{Float64}:\n 0.0\n 0.0\n\njulia> MOI.eval_objective_gradient(d, ∇f, values)\n\njulia> ∇f[x_index], ∇f[y_index]\n(-0.4161468365471424, -0.9899924966004454)\n\njulia> cos(2.0), cos(3.0)\n(-0.4161468365471424, -0.9899924966004454)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Only nonlinear constraints (those added with @NLconstraint), and nonlinear objectives (added with @NLobjective) exist in the scope of the NLPEvaluator.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The NLPEvaluator does not evaluate derivatives of linear or quadratic constraints or objectives.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The index method applied to a nonlinear constraint reference object returns its index as a MOI.Nonlinear.ConstraintIndex. For example:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @NLconstraint(model, cons1, sin(x) <= 1);\n\njulia> @NLconstraint(model, cons2, x + 5 == 10);\n\njulia> typeof(cons1)\nNonlinearConstraintRef{ScalarShape} (alias for ConstraintRef{GenericModel{Float64}, MathOptInterface.Nonlinear.ConstraintIndex, ScalarShape})\n\njulia> index(cons1)\nMathOptInterface.Nonlinear.ConstraintIndex(1)\n\njulia> index(cons2)\nMathOptInterface.Nonlinear.ConstraintIndex(2)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"# TODO: Provide a link for how to access the linear and quadratic parts of the\n# model.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Note that for one-sided nonlinear constraints, JuMP subtracts any values on the right-hand side when computing expressions. In other words, one-sided nonlinear constraints are always transformed to have a right-hand side of zero.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"This method of querying derivatives directly from a JuMP model is convenient for interacting with the model in a structured way, for example, for accessing derivatives of specific variables. For example, in statistical maximum likelihood estimation problems, one is often interested in the Hessian matrix at the optimal solution, which can be queried using the NLPEvaluator.","category":"page"},{"location":"manual/nlp/#Raw-expression-input","page":"Nonlinear Modeling (Legacy)","title":"Raw expression input","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"warning: Warning\nThis section requires advanced knowledge of Julia's Expr. You should read the Expressions and evaluation section of the Julia documentation first.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"In addition to the @NLexpression, @NLobjective and @NLconstraint macros, it is also possible to provide Julia Expr objects directly by using add_nonlinear_expression, set_nonlinear_objective and add_nonlinear_constraint.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"This input form may be useful if the expressions are generated programmatically, or if you experience compilation issues with the macro input (see Known performance issues for more information).","category":"page"},{"location":"manual/nlp/#Add-a-nonlinear-expression","page":"Nonlinear Modeling (Legacy)","title":"Add a nonlinear expression","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use add_nonlinear_expression to add a nonlinear expression to the model.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> expr = :($(x) + sin($(x)^2))\n:(x + sin(x ^ 2))\n\njulia> expr_ref = add_nonlinear_expression(model, expr)\nsubexpression[1]: x + sin(x ^ 2.0)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"This is equivalent to","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr_ref = @NLexpression(model, x + sin(x^2))\nsubexpression[1]: x + sin(x ^ 2.0)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"note: Note\nYou must interpolate the variables directly into the expression expr.","category":"page"},{"location":"manual/nlp/#Set-the-objective-function","page":"Nonlinear Modeling (Legacy)","title":"Set the objective function","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use set_nonlinear_objective to set a nonlinear objective.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr = :($(x) + $(x)^2)\n:(x + x ^ 2)\n\njulia> set_nonlinear_objective(model, MIN_SENSE, expr)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"This is equivalent to","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @NLobjective(model, Min, x + x^2)","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"note: Note\nYou must use MIN_SENSE or MAX_SENSE instead of Min and Max.","category":"page"},{"location":"manual/nlp/#Add-a-constraint","page":"Nonlinear Modeling (Legacy)","title":"Add a constraint","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Use add_nonlinear_constraint to add a nonlinear constraint.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr = :($(x) + $(x)^2)\n:(x + x ^ 2)\n\njulia> add_nonlinear_constraint(model, :($(expr) <= 1))\n(x + x ^ 2.0) - 1.0 ≤ 0","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"This is equivalent to","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @NLconstraint(model, Min, x + x^2 <= 1)\n(x + x ^ 2.0) - 1.0 ≤ 0","category":"page"},{"location":"manual/nlp/#More-complicated-examples","page":"Nonlinear Modeling (Legacy)","title":"More complicated examples","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"Raw expression input is most useful when the expressions are generated programmatically, often in conjunction with user-defined functions.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"As an example, we construct a model with the nonlinear constraints f(x) <= 1, where f(x) = x^2 and f(x) = sin(x)^2:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> function main(functions::Vector{Function})\n           model = Model()\n           @variable(model, x)\n           for (i, f) in enumerate(functions)\n               f_sym = Symbol(\"f_$(i)\")\n               register(model, f_sym, 1, f; autodiff = true)\n               add_nonlinear_constraint(model, :($(f_sym)($(x)) <= 1))\n           end\n           print(model)\n           return\n       end\nmain (generic function with 1 method)\n\njulia> main([x -> x^2, x -> sin(x)^2])\nFeasibility\nSubject to\n f_1(x) - 1.0 ≤ 0\n f_2(x) - 1.0 ≤ 0","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"As another example, we construct a model with the constraint x^2 + sin(x)^2 <= 1:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> function main(functions::Vector{Function})\n           model = Model()\n           @variable(model, x)\n           expr = Expr(:call, :+)\n           for (i, f) in enumerate(functions)\n               f_sym = Symbol(\"f_$(i)\")\n               register(model, f_sym, 1, f; autodiff = true)\n               push!(expr.args, :($(f_sym)($(x))))\n           end\n           add_nonlinear_constraint(model, :($(expr) <= 1))\n           print(model)\n           return\n       end\nmain (generic function with 1 method)\n\njulia> main([x -> x^2, x -> sin(x)^2])\nFeasibility\nSubject to\n (f_1(x) + f_2(x)) - 1.0 ≤ 0","category":"page"},{"location":"manual/nlp/#Registered-functions-with-a-variable-number-of-arguments","page":"Nonlinear Modeling (Legacy)","title":"Registered functions with a variable number of arguments","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"User defined functions require a fixed number of input arguments. However, sometimes you will want to use a registered function like:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> f(x...) = sum(exp(x[i]^2) for i in 1:length(x));","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"with different numbers of arguments.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The solution is to register the same function f for each unique number of input arguments, making sure to use a unique name each time. For example:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"julia> A = [[1], [1, 2], [2, 3, 4], [1, 3, 4, 5]];\n\njulia> model = Model();\n\njulia> @variable(model, x[1:5]);\n\njulia> funcs = Set{Symbol}();\n\njulia> for a in A\n           key = Symbol(\"f$(length(a))\")\n           if !(key in funcs)\n               push!(funcs, key)\n               register(model, key, length(a), f; autodiff = true)\n           end\n           add_nonlinear_constraint(model, :($key($(x[a]...)) <= 1))\n       end\n\njulia> print(model)\nFeasibility\nSubject to\n f1(x[1]) - 1.0 ≤ 0\n f2(x[1], x[2]) - 1.0 ≤ 0\n f3(x[2], x[3], x[4]) - 1.0 ≤ 0\n f4(x[1], x[3], x[4], x[5]) - 1.0 ≤ 0","category":"page"},{"location":"manual/nlp/#Known-performance-issues","page":"Nonlinear Modeling (Legacy)","title":"Known performance issues","text":"","category":"section"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The macro-based input to JuMP's nonlinear interface can cause a performance issue if you:","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"write a macro with a large number (hundreds) of terms\ncall that macro from within a function instead of from the top-level in global scope.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The first issue does not depend on the number of resulting terms in the mathematical expression, but rather the number of terms in the Julia Expr representation of that expression. For example, the expression sum(x[i] for i in 1:1_000_000) contains one million mathematical terms, but the Expr representation is just a single sum.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"The most common cause, other than a lot of tedious typing, is if you write a program that automatically writes a JuMP model as a text file, which you later execute. One example is MINLPlib.jl which automatically transpiled models in the GAMS scalar format into JuMP examples.","category":"page"},{"location":"manual/nlp/","page":"Nonlinear Modeling (Legacy)","title":"Nonlinear Modeling (Legacy)","text":"As a rule of thumb, if you are writing programs to automatically generate expressions for the JuMP macros, you should target the Raw expression input instead. For more information, read MathOptInterface Issue#1997.","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"EditURL = \"https://github.com/jump-dev/Dualization.jl/blob/v0.5.8/README.md\"","category":"page"},{"location":"packages/Dualization/#Dualization.jl","page":"jump-dev/Dualization.jl","title":"Dualization.jl","text":"","category":"section"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"(Image: Build Status) (Image: codecov) (Image: DOI)","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"Dualization.jl is an extension package for MathOptInterface.jl that formulates the dual of conic optimization problems.","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"Dualization.jl has two main features:","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"The Dualization.dualize function that computes the dual formulation of either a MathOptInterface.jl or a JuMP model.\nThe Dualization.dual_optimizer function that creates a MathOptInterface-compatible optimizer that solves the dual of the problem instead of the primal.","category":"page"},{"location":"packages/Dualization/#License","page":"jump-dev/Dualization.jl","title":"License","text":"","category":"section"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"Dualization.jl is licensed under the MIT License.","category":"page"},{"location":"packages/Dualization/#Installation","page":"jump-dev/Dualization.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"Install Dualization using Pkg.add:","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"import Pkg\nPkg.add(\"Dualization\")","category":"page"},{"location":"packages/Dualization/#Use-with-JuMP","page":"jump-dev/Dualization.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"To compute the dual formulation of a JuMP model, use dualize:","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"using JuMP, Dualization\nmodel = Model()\n# ... build model ...\ndual_model = dualize(model)","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"To solve the dual formulation of a JuMP model, create a dual_optimizer:","category":"page"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"using JuMP, Dualization, SCS\nmodel = Model(dual_optimizer(SCS.Optimizer))\n# ... build model ...\noptimize!(model)  # Solves the dual instead of the primal","category":"page"},{"location":"packages/Dualization/#Documentation","page":"jump-dev/Dualization.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/Dualization/","page":"jump-dev/Dualization.jl","title":"jump-dev/Dualization.jl","text":"The documentation for Dualization.jl includes a detailed description of the dual reformulation, along with examples and an API reference.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/constraints.md\"","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/manual/constraints/#Constraints","page":"Constraints","title":"Constraints","text":"","category":"section"},{"location":"moi/manual/constraints/#Add-a-constraint","page":"Constraints","title":"Add a constraint","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Use add_constraint to add a single constraint.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> c = MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Nonnegatives(2))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonnegatives}(1)","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"add_constraint returns a ConstraintIndex type, which is used to refer to the added constraint in other calls.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Check if a ConstraintIndex is valid using is_valid.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> MOI.is_valid(model, c)\ntrue","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Use add_constraints to add a number of constraints of the same type.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> c = MOI.add_constraints(\n           model,\n           [x[1], x[2]],\n           [MOI.GreaterThan(0.0), MOI.GreaterThan(1.0)]\n       )\n2-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}}:\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(1)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(2)","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"This time, a vector of ConstraintIndex are returned.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Use supports_constraint to check if the model supports adding a constraint type.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> MOI.supports_constraint(\n           model,\n           MOI.VariableIndex,\n           MOI.GreaterThan{Float64},\n       )\ntrue","category":"page"},{"location":"moi/manual/constraints/#Delete-a-constraint","page":"Constraints","title":"Delete a constraint","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Use delete to delete a constraint.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> MOI.delete(model, c)\n\njulia> MOI.is_valid(model, c)\nfalse","category":"page"},{"location":"moi/manual/constraints/#Constraint-attributes","page":"Constraints","title":"Constraint attributes","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"The following attributes are available for constraints:","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"ConstraintName\nConstraintPrimalStart\nConstraintDualStart\nConstraintPrimal\nConstraintDual\nConstraintBasisStatus\nConstraintFunction\nCanonicalConstraintFunction\nConstraintSet","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Get and set these attributes using get and set.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"julia> MOI.set(model, MOI.ConstraintName(), c, \"con_c\")\n\njulia> MOI.get(model, MOI.ConstraintName(), c)\n\"con_c\"","category":"page"},{"location":"moi/manual/constraints/#Constraints-by-function-set-pairs","page":"Constraints","title":"Constraints by function-set pairs","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Below is a list of common constraint types and how they are represented as function-set pairs in MOI. In the notation below, x is a vector of decision variables, x_i is a scalar decision variable, alpha beta are scalar constants, a b are constant vectors, A is a constant matrix and mathbbR_+ (resp. mathbbR_-) is the set of non-negative (resp. non-positive) real numbers.","category":"page"},{"location":"moi/manual/constraints/#Linear-constraints","page":"Constraints","title":"Linear constraints","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Mathematical Constraint MOI Function MOI Set\na^Tx le beta ScalarAffineFunction LessThan\na^Tx ge alpha ScalarAffineFunction GreaterThan\na^Tx = beta ScalarAffineFunction EqualTo\nalpha le a^Tx le beta ScalarAffineFunction Interval\nx_i le beta VariableIndex LessThan\nx_i ge alpha VariableIndex GreaterThan\nx_i = beta VariableIndex EqualTo\nalpha le x_i le beta VariableIndex Interval\nAx + b in mathbbR_+^n VectorAffineFunction Nonnegatives\nAx + b in mathbbR_-^n VectorAffineFunction Nonpositives\nAx + b = 0 VectorAffineFunction Zeros","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"By convention, solvers are not expected to support nonzero constant terms in the ScalarAffineFunctions the first four rows of the preceding table because they are redundant with the parameters of the sets. For example, encode 2x + 1 le 2 as 2x le 1.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Constraints with VariableIndex in LessThan, GreaterThan, EqualTo, or Interval sets have a natural interpretation as variable bounds. As such, it is typically not natural to impose multiple lower- or upper-bounds on the same variable, and the solver interfaces will throw respectively LowerBoundAlreadySet or UpperBoundAlreadySet.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Moreover, adding two VariableIndex constraints on the same variable with the same set is impossible because they share the same index as it is the index of the variable, see ConstraintIndex.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"It is natural, however, to impose upper- and lower-bounds separately as two different constraints on a single variable. The difference between imposing bounds by using a single Interval constraint and by using separate LessThan and GreaterThan constraints is that the latter will allow the solver to return separate dual multipliers for the two bounds, while the former will allow the solver to return only a single dual for the interval constraint.","category":"page"},{"location":"moi/manual/constraints/#Conic-constraints","page":"Constraints","title":"Conic constraints","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Mathematical Constraint MOI Function MOI Set\nlVert Ax + brVert_2 le c^Tx + d VectorAffineFunction SecondOrderCone\ny ge lVert x rVert_2 VectorOfVariables SecondOrderCone\n2yz ge lVert x rVert_2^2 yz ge 0 VectorOfVariables RotatedSecondOrderCone\n(a_1^Tx + b_1a_2^Tx + b_2a_3^Tx + b_3) in mathcalE VectorAffineFunction ExponentialCone\nA(x) in mathcalS_+ VectorAffineFunction PositiveSemidefiniteConeTriangle\nB(x) in mathcalS_+ VectorAffineFunction PositiveSemidefiniteConeSquare\nx in mathcalS_+ VectorOfVariables PositiveSemidefiniteConeTriangle\nx in mathcalS_+ VectorOfVariables PositiveSemidefiniteConeSquare","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"where mathcalE is the exponential cone (see ExponentialCone), mathcalS_+ is the set of positive semidefinite symmetric matrices, A is an affine map that outputs symmetric matrices and B is an affine map that outputs square matrices.","category":"page"},{"location":"moi/manual/constraints/#Quadratic-constraints","page":"Constraints","title":"Quadratic constraints","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Mathematical Constraint MOI Function MOI Set\nfrac12x^TQx + a^Tx + b ge 0 ScalarQuadraticFunction GreaterThan\nfrac12x^TQx + a^Tx + b le 0 ScalarQuadraticFunction LessThan\nfrac12x^TQx + a^Tx + b = 0 ScalarQuadraticFunction EqualTo\nBilinear matrix inequality VectorQuadraticFunction PositiveSemidefiniteCone...","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"note: Note\nFor more details on the internal format of the quadratic functions see ScalarQuadraticFunction or VectorQuadraticFunction.","category":"page"},{"location":"moi/manual/constraints/#Discrete-and-logical-constraints","page":"Constraints","title":"Discrete and logical constraints","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Mathematical Constraint MOI Function MOI Set\nx_i in mathbbZ VariableIndex Integer\nx_i in 01 VariableIndex ZeroOne\nx_i in 0 cup lu VariableIndex Semicontinuous\nx_i in 0 cup ll+1ldotsu-1u VariableIndex Semiinteger\nAt most one component of x can be nonzero VectorOfVariables SOS1\nAt most two components of x can be nonzero, and if so they must be adjacent components VectorOfVariables SOS2\ny = 1 implies a^T x in S VectorAffineFunction Indicator","category":"page"},{"location":"moi/manual/constraints/#JuMP-mapping","page":"Constraints","title":"JuMP mapping","text":"","category":"section"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"The following bullet points show examples of how JuMP constraints are translated into MOI function-set pairs:","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"@constraint(m, 2x + y <= 10) becomes ScalarAffineFunction-in-LessThan\n@constraint(m, 2x + y >= 10) becomes ScalarAffineFunction-in-GreaterThan\n@constraint(m, 2x + y == 10) becomes ScalarAffineFunction-in-EqualTo\n@constraint(m, 0 <= 2x + y <= 10) becomes ScalarAffineFunction-in-Interval\n@constraint(m, 2x + y in ArbitrarySet()) becomes ScalarAffineFunction-in-ArbitrarySet.","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"Variable bounds are handled in a similar fashion:","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"@variable(m, x <= 1) becomes VariableIndex-in-LessThan\n@variable(m, x >= 1) becomes VariableIndex-in-GreaterThan","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"One notable difference is that a variable with an upper and lower bound is translated into two constraints, rather than an interval, that is:","category":"page"},{"location":"moi/manual/constraints/","page":"Constraints","title":"Constraints","text":"@variable(m, 0 <= x <= 1) becomes VariableIndex-in-LessThan and  VariableIndex-in-GreaterThan.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"EditURL = \"sudoku.jl\"","category":"page"},{"location":"tutorials/linear/sudoku/#Sudoku","page":"Sudoku","title":"Sudoku","text":"","category":"section"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"This tutorial was originally contributed by Iain Dunning.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Sudoku is a popular number puzzle. The goal is to place the digits 1 to 9 on a nine-by-nine grid, with some of the digits already filled in. Your solution must satisfy the following rules:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"The numbers 1 to 9 must appear in each 3x3 square\nThe numbers 1 to 9 must appear in each row\nThe numbers 1 to 9 must appear in each column","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Here is a partially solved Sudoku problem:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"(Image: Partially solved Sudoku)","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Solving a Sudoku isn't an optimization problem with an objective; its actually a feasibility problem: we wish to find a feasible solution that satisfies these rules. You can think of it as an optimization problem with an objective of 0.","category":"page"},{"location":"tutorials/linear/sudoku/#Mixed-integer-linear-programming-formulation","page":"Sudoku","title":"Mixed-integer linear programming formulation","text":"","category":"section"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"We can model this problem using 0-1 integer programming: a problem where all the decision variables are binary. We'll use JuMP to create the model, and then we can solve it with any integer programming solver.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"using JuMP\nusing HiGHS","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"We will define a binary variable (a variable that is either 0 or 1) for each possible number in each possible cell. The meaning of each variable is as follows: x[i,j,k] = 1  if and only if cell (i,j) has number k, where i is the row and j is the column.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Create a model","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"sudoku = Model(HiGHS.Optimizer)\nset_silent(sudoku)","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Create our variables","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"@variable(sudoku, x[i = 1:9, j = 1:9, k = 1:9], Bin);\nnothing #hide","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Now we can begin to add our constraints. We'll actually start with something obvious to us as humans, but what we need to enforce: that there can be only one number per cell.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"for i in 1:9  # For each row\n    for j in 1:9  # and each column\n        # Sum across all the possible digits. One and only one of the digits\n        # can be in this cell, so the sum must be equal to one.\n        @constraint(sudoku, sum(x[i, j, k] for k in 1:9) == 1)\n    end\nend","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Next we'll add the constraints for the rows and the columns. These constraints are all very similar, so much so that we can actually add them at the same time.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"for ind in 1:9  # Each row, OR each column\n    for k in 1:9  # Each digit\n        # Sum across columns (j) - row constraint\n        @constraint(sudoku, sum(x[ind, j, k] for j in 1:9) == 1)\n        # Sum across rows (i) - column constraint\n        @constraint(sudoku, sum(x[i, ind, k] for i in 1:9) == 1)\n    end\nend","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Finally, we have the to enforce the constraint that each digit appears once in each of the nine 3x3 sub-grids. Our strategy will be to index over the top-left corners of each 3x3 square with for loops, then sum over the squares.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"for i in 1:3:7\n    for j in 1:3:7\n        for k in 1:9\n            # i is the top left row, j is the top left column.\n            # We'll sum from i to i+2, for example, i=4, r=4, 5, 6.\n            @constraint(\n                sudoku,\n                sum(x[r, c, k] for r in i:(i+2), c in j:(j+2)) == 1\n            )\n        end\n    end\nend","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"The final step is to add the initial solution as a set of constraints. We'll solve the problem that is in the picture at the start of the tutorial. We'll put a 0 if there is no digit in that location.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"The given digits","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"init_sol = [\n    5 3 0 0 7 0 0 0 0\n    6 0 0 1 9 5 0 0 0\n    0 9 8 0 0 0 0 6 0\n    8 0 0 0 6 0 0 0 3\n    4 0 0 8 0 3 0 0 1\n    7 0 0 0 2 0 0 0 6\n    0 6 0 0 0 0 2 8 0\n    0 0 0 4 1 9 0 0 5\n    0 0 0 0 8 0 0 7 9\n]\nfor i in 1:9\n    for j in 1:9\n        # If the space isn't empty\n        if init_sol[i, j] != 0\n            # Then the corresponding variable for that digit and location must\n            # be 1.\n            fix(x[i, j, init_sol[i, j]], 1; force = true)\n        end\n    end\nend","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"solve problem","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"optimize!(sudoku)","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Extract the values of x","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"x_val = value.(x);\nnothing #hide","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Create a matrix to store the solution","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"sol = zeros(Int, 9, 9)  # 9x9 matrix of integers\nfor i in 1:9\n    for j in 1:9\n        for k in 1:9\n            # Integer programs are solved as a series of linear programs so the\n            # values might not be precisely 0 and 1. We can round them to\n            # the nearest integer to make it easier.\n            if round(Int, x_val[i, j, k]) == 1\n                sol[i, j] = k\n            end\n        end\n    end\nend","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Display the solution","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"sol","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Which is the correct solution:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"(Image: Solved Sudoku)","category":"page"},{"location":"tutorials/linear/sudoku/#Constraint-programming-formulation","page":"Sudoku","title":"Constraint programming formulation","text":"","category":"section"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"We can also model this problem using constraint programming and the all-different constraint, which says that no two elements of a vector can take the same value.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Because of the reformulation system in MathOptInterface, we can still solve this problem using HiGHS.","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n# HiGHS v1.2 has a bug in presolve which causes the problem to be classified as\n# infeasible.\nset_attribute(model, \"presolve\", \"off\")","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Instead of the binary variables, we directly define a 9x9 grid of integer values between 1 and 9:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"@variable(model, 1 <= x[1:9, 1:9] <= 9, Int);\nnothing #hide","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Then, we enforce that the values in each row must be all-different:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"@constraint(model, [i = 1:9], x[i, :] in MOI.AllDifferent(9));\nnothing #hide","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"That the values in each column must be all-different:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"@constraint(model, [j = 1:9], x[:, j] in MOI.AllDifferent(9));\nnothing #hide","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"And that the values in each 3x3 sub-grid must be all-different:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"for i in (0, 3, 6), j in (0, 3, 6)\n    @constraint(model, vec(x[i.+(1:3), j.+(1:3)]) in MOI.AllDifferent(9))\nend","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Finally, as before we set the initial solution and optimize:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"for i in 1:9, j in 1:9\n    if init_sol[i, j] != 0\n        fix(x[i, j], init_sol[i, j]; force = true)\n    end\nend\n\noptimize!(model)","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Display the solution","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"csp_sol = round.(Int, value.(x))","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"Which is the same as we found before:","category":"page"},{"location":"tutorials/linear/sudoku/","page":"Sudoku","title":"Sudoku","text":"sol == csp_sol","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"EditURL = \"performance_tips.jl\"","category":"page"},{"location":"tutorials/getting_started/performance_tips/#Performance-tips","page":"Performance tips","title":"Performance tips","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"By now you should have read the other \"getting started\" tutorials. You're almost ready to write your own models, but before you do so there are some important things to be aware of.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/getting_started/performance_tips/#Read-the-Julia-performance-tips","page":"Performance tips","title":"Read the Julia performance tips","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"The first thing to do is read the Performance tips section of the Julia manual. The most important rule is to avoid global variables. This is particularly important if you're learning JuMP after using a language like MATLAB.","category":"page"},{"location":"tutorials/getting_started/performance_tips/#The-\"time-to-first-solve\"-issue","page":"Performance tips","title":"The \"time-to-first-solve\" issue","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Similar to the infamous time-to-first-plot plotting problem, JuMP suffers from time-to-first-solve latency. This latency occurs because the first time you call JuMP code in each session, Julia needs to compile a lot of code specific to your problem. This issue is actively being worked on, but there are a few things you can do to improve things.","category":"page"},{"location":"tutorials/getting_started/performance_tips/#Suggestion-1:-don't-call-JuMP-from-the-command-line","page":"Performance tips","title":"Suggestion 1: don't call JuMP from the command line","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"In other languages, you might be used to a workflow like:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"$ julia my_script.jl","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"This doesn't work for JuMP, because we have to pay the compilation latency every time you run the script. Instead, use one of the suggested workflows from the Julia documentation.","category":"page"},{"location":"tutorials/getting_started/performance_tips/#Suggestion-2:-disable-bridges-if-none-are-being-used","page":"Performance tips","title":"Suggestion 2: disable bridges if none are being used","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"At present, the majority of the latency problems are caused by JuMP's bridging mechanism. If you only use constraints that are natively supported by the solver, you can disable bridges by passing add_bridges = false to Model.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"model = Model(HiGHS.Optimizer; add_bridges = false)","category":"page"},{"location":"tutorials/getting_started/performance_tips/#Suggestion-3:-use-PackageCompiler","page":"Performance tips","title":"Suggestion 3: use PackageCompiler","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"As an example of compilation latency, consider the following linear program with two variables and two constraints:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"using JuMP, HiGHS\nmodel = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, x >= 0)\n@variable(model, 0 <= y <= 3)\n@objective(model, Min, 12x + 20y)\n@constraint(model, c1, 6x + 8y >= 100)\n@constraint(model, c2, 7x + 12y >= 120)\noptimize!(model)\nopen(\"model.log\", \"w\") do io\n    print(io, solution_summary(model; verbose = true))\n    return\nend","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Saving the problem in model.jl and calling from the command line results in:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"$ time julia model.jl\n15.78s user 0.48s system 100% cpu 16.173 total","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Clearly, 16 seconds is a large overhead to pay for solving this trivial model. However, the compilation latency is independent on the problem size, and so 16 seconds of additional overhead may be tolerable for larger models that take minutes or hours to solve.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"In cases where the compilation latency is intolerable, JuMP is compatible with the PackageCompiler.jl package, which makes it easy to generate a custom sysimage (a binary extension to Julia that caches compiled code) that dramatically reduces the compilation latency. A custom image for our problem can be created as follows:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"using PackageCompiler, Libdl\nPackageCompiler.create_sysimage(\n    [\"JuMP\", \"HiGHS\"],\n    sysimage_path = \"customimage.\" * Libdl.dlext,\n    precompile_execution_file = \"model.jl\",\n)","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"When Julia is run with the custom image, the run time is now 0.7 seconds instead of 16:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"$ time julia --sysimage customimage model.jl\n0.68s user 0.22s system 153% cpu 0.587 total","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Other performance tweaks, such as disabling bridges or using direct mode can reduce this time further.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"note: Note\ncreate_sysimage only needs to be run once, and the same sysimage can be used–to a slight detriment of performance–even if we modify model.jl or run a different file.","category":"page"},{"location":"tutorials/getting_started/performance_tips/#Use-macros-to-build-expressions","page":"Performance tips","title":"Use macros to build expressions","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Use JuMP's macros (or add_to_expression!) to build expressions. Avoid constructing expressions outside the macros.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Constructing an expression outside the macro results in intermediate copies of the expression. For example,","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"x[1] + x[2] + x[3]","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"is equivalent to","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"a = x[1]\nb = a + x[2]\nc = b + x[3]","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Since we only care about c, the a and b expressions are not needed and constructing them slows the program down.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"JuMP's macros rewrite the expressions to operate in-place and avoid these extra copies. Because they allocate less memory, they are faster, particularly for large expressions.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Here's an example.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"model = Model()\n@variable(model, x[1:3])","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Here's what happens if we construct the expression outside the macro:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"@allocated x[1] + x[2] + x[3]","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"info: Info\nThe @allocated measures how many bytes were allocated during the evaluation of an expression. Fewer is better.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"If we use the @expression macro, we get many fewer allocations:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"@allocated @expression(model, x[1] + x[2] + x[3])","category":"page"},{"location":"tutorials/getting_started/performance_tips/#Disable-string-names","page":"Performance tips","title":"Disable string names","text":"","category":"section"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"By default, JuMP creates String names for variables and constraints and passes these to the solver. The benefit of passing names is that it improves the readability of log messages from the solver (for example, \"variable x has invalid bounds\" instead of \"variable v1203 has invalid bounds\"), but for larger models the overhead of passing names can be non-trivial.","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Disable the creation of String names by setting set_string_name = false in the @variable and @constraint macros, or by calling set_string_names_on_creation to disable all names for a particular model:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"model = Model();\nset_string_names_on_creation(model, false)\n@variable(model, x)\n@constraint(model, c, 2x <= 1)","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"Note that this doesn't change how symbolic names and bindings are stored:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"x\nmodel[:x]\nx === model[:x]","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"But you can no longer look up the variable by the string name:","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"variable_by_name(model, \"x\") === nothing","category":"page"},{"location":"tutorials/getting_started/performance_tips/","page":"Performance tips","title":"Performance tips","text":"info: Info\nFor more information on the difference between string names, symbolic names, and bindings, see String names, symbolic names, and bindings.","category":"page"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"EditURL = \"https://github.com/jump-dev/ParametricOptInterface.jl/blob/1721e7b07bd1991238802a438ecb5af566d4b555/README.md\"","category":"page"},{"location":"packages/ParametricOptInterface/#ParametricOptInterface.jl","page":"jump-dev/ParametricOptInterface.jl","title":"ParametricOptInterface.jl","text":"","category":"section"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"(Image: stable docs) (Image: development docs) (Image: Build Status) (Image: Coverage)","category":"page"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"ParametricOptInterface.jl is a package that adds parameters to models in JuMP and MathOptInterface.","category":"page"},{"location":"packages/ParametricOptInterface/#License","page":"jump-dev/ParametricOptInterface.jl","title":"License","text":"","category":"section"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"ParametricOptInterface.jl is licensed under the MIT License.","category":"page"},{"location":"packages/ParametricOptInterface/#Installation","page":"jump-dev/ParametricOptInterface.jl","title":"Installation","text":"","category":"section"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"Install ParametricOptInterface using Pkg.add:","category":"page"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"import Pkg\nPkg.add(\"ParametricOptInterface\")","category":"page"},{"location":"packages/ParametricOptInterface/#Documentation","page":"jump-dev/ParametricOptInterface.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"The documentation for ParametricOptInterface.jl includes a detailed description of the theory behind the package, along with examples, tutorials, and an API reference.","category":"page"},{"location":"packages/ParametricOptInterface/#Use-with-JuMP","page":"jump-dev/ParametricOptInterface.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"Use ParametricOptInterface with JuMP by following this brief example:","category":"page"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"using JuMP, HiGHS\nimport ParametricOptInterface as POI\nmodel = Model(() -> POI.Optimizer(HiGHS.Optimizer()))\n@variable(model, x)\n@variable(model, p in POI.Parameter(1.0))\n@constraint(model, cons, x + p >= 3)\n@objective(model, Min, 2x)\noptimize!(model)\nMOI.set(model, POI.ParameterValue(), p, 2.0)\noptimize!(model)","category":"page"},{"location":"packages/ParametricOptInterface/#GSOC2020","page":"jump-dev/ParametricOptInterface.jl","title":"GSOC2020","text":"","category":"section"},{"location":"packages/ParametricOptInterface/","page":"jump-dev/ParametricOptInterface.jl","title":"jump-dev/ParametricOptInterface.jl","text":"ParametricOptInterface began as a NumFOCUS sponsored Google Summer of Code (2020) project.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"EditURL = \"quantum_discrimination.jl\"","category":"page"},{"location":"tutorials/conic/quantum_discrimination/#Quantum-state-discrimination","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"","category":"section"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"This tutorial solves the problem of quantum state discrimination.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"The purpose of this tutorial to demonstrate how to solve problems involving complex-valued decision variables and the HermitianPSDCone. See Complex number support for more details.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/#Required-packages","page":"Quantum state discrimination","title":"Required packages","text":"","category":"section"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"This tutorial makes use of the following packages:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"using JuMP\nimport LinearAlgebra\nimport SCS","category":"page"},{"location":"tutorials/conic/quantum_discrimination/#Formulation","page":"Quantum state discrimination","title":"Formulation","text":"","category":"section"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"A d-dimensional quantum state, rho, can be defined by a complex-valued Hermitian matrix with a trace of 1. Assume we have N d-dimensional quantum states, rho_i_i=1^N, each of which is equally likely.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"The goal of the quantum state discrimination problem is to choose a positive operator-valued measure (POVM),  E_i _i=1^N, such that if we observe E_i then the most probable state that we are in is rho_i.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"Each POVM element, E_i, is a complex-valued Hermitian matrix, and there is a requirement that sumlimits_i=1^N E_i = mathbfI.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"To choose a POVM, we want to maximize the probability that we guess the quantum state correctly. This can be formulated as the following optimization problem:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"beginaligned\nmaxlimits_E   frac1N sumlimits_i=1^N operatornametr(rho_i  E_i) \ntextst       sumlimits_i=1^N E_i = mathbfI \n                      E_i succeq 0  forall i = 1ldotsN\nendaligned","category":"page"},{"location":"tutorials/conic/quantum_discrimination/#Data","page":"Quantum state discrimination","title":"Data","text":"","category":"section"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"To setup our problem, we need N d-dimensional quantum states. To keep the problem simple, we use N = 2 and d = 2.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"N, d = 2, 2","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"We then generated N random d-dimensional quantum states:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"function random_state(d)\n    x = randn(ComplexF64, (d, d))\n    y = x * x'\n    return LinearAlgebra.Hermitian(y / LinearAlgebra.tr(y))\nend\n\nρ = [random_state(d) for i in 1:N]","category":"page"},{"location":"tutorials/conic/quantum_discrimination/#JuMP-formulation","page":"Quantum state discrimination","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"To model the problem in JuMP, we need a solver that supports positive semidefinite matrices:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"model = Model(SCS.Optimizer)\nset_silent(model)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"Then, we construct our set of E variables:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"E = [@variable(model, [1:d, 1:d] in HermitianPSDCone()) for i in 1:N]","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"Here we have created a vector of matrices. This is different to other modeling languages such as YALMIP, which allow you to create a multi-dimensional array in which 2-dimensional slices of the array are Hermitian matrices.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"We also need to enforce the constraint that sumlimits_i=1^N E_i = mathbfI:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"@constraint(model, sum(E) == LinearAlgebra.I)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"This constraint is a complex-valued equality constraint. In the solver, it will be decomposed onto two types of equality constraints: one to enforce equality of the real components, and one to enforce equality of the imaginary components.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"Our objective is to maximize the expected probability of guessing correctly:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"@objective(\n    model,\n    Max,\n    sum(real(LinearAlgebra.tr(ρ[i] * E[i])) for i in 1:N) / N,\n)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"Now we optimize:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"The probability of guessing correctly is:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"objective_value(model)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"When N = 2, there is a known analytical solution of:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"0.5 + 0.25 * sum(LinearAlgebra.svdvals(ρ[1] - ρ[2]))","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"proving that we found the optimal solution.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"Finally, the optimal POVM is:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"solution = [value.(e) for e in E]","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"tip: Tip\nDuality plays a large role in solving conic optimization models. Depending on the solver, it can be more efficient to solve the dual of this problem instead of the primal. If performance is an issue, see the Dualization tutorial for more details.","category":"page"},{"location":"tutorials/conic/quantum_discrimination/#Alternative-formulation","page":"Quantum state discrimination","title":"Alternative formulation","text":"","category":"section"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"The formulation above includes N Hermitian matrices and a set of linear equality constraints. We can simplify the problem by replacing E_N with E_N = I - sumlimits_i=1^N-1 E_i. This results in:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"model = Model(SCS.Optimizer)\nset_silent(model)\nE = [@variable(model, [1:d, 1:d] in HermitianPSDCone()) for i in 1:N-1]\nE_N = LinearAlgebra.Hermitian(LinearAlgebra.I - sum(E))\n@constraint(model, E_N in HermitianPSDCone())\npush!(E, E_N)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"The objective can also be simplified, by observing that it is equivalent to:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"@objective(model, Max, real(LinearAlgebra.dot(ρ, E)) / N)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"Then we can check that we get the same solution:","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/conic/quantum_discrimination/","page":"Quantum state discrimination","title":"Quantum state discrimination","text":"objective_value(model)","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/implementing.md\"","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/tutorials/implementing/#Implementing-a-solver-interface","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"This guide outlines the basic steps to implement an interface to MathOptInterface for a new solver.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"danger: Danger\nImplementing an interface to MathOptInterface for a new solver is a lot of work. Before starting, we recommend that you join the Developer chatroom and explain a little bit about the solver you are wrapping. If you have questions that are not answered by this guide, please ask them in the Developer chatroom so we can improve this guide.","category":"page"},{"location":"moi/tutorials/implementing/#A-note-on-the-API","page":"Implementing a solver interface","title":"A note on the API","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The API of MathOptInterface is large and varied. In order to support the diversity of solvers and use-cases, we make heavy use of duck-typing. That is, solvers are not expected to implement the full API, nor is there a well-defined minimal subset of what must be implemented. Instead, you should implement the API as necessary to make the solver function as you require.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The main reason for using duck-typing is that solvers work in different ways and target different use-cases.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For example:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Some solvers support incremental problem construction, support modification after a solve, and have native support for things like variable names.\nOther solvers are \"one-shot\" solvers that require all of the problem data to construct and solve the problem in a single function call. They do not support modification or things like variable names.\nOther \"solvers\" are not solvers at all, but things like file readers. These may only support functions like read_from_file, and may not even support the ability to add variables or constraints directly.\nFinally, some \"solvers\" are layers which take a problem as input, transform it according to some rules, and pass the transformed problem to an inner solver.","category":"page"},{"location":"moi/tutorials/implementing/#Preliminaries","page":"Implementing a solver interface","title":"Preliminaries","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Before starting on your wrapper, you should do some background research and make the solver accessible via Julia.","category":"page"},{"location":"moi/tutorials/implementing/#Decide-if-MathOptInterface-is-right-for-you","page":"Implementing a solver interface","title":"Decide if MathOptInterface is right for you","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The first step in writing a wrapper is to decide whether implementing an interface is the right thing to do.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"MathOptInterface is an abstraction layer for unifying constrained mathematical optimization solvers. If your solver doesn't fit in the category, for example, it implements a derivative-free algorithm for unconstrained objective functions, MathOptInterface may not be the right tool for the job.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nIf you're not sure whether you should write an interface, ask in the Developer chatroom.","category":"page"},{"location":"moi/tutorials/implementing/#Find-a-similar-solver-already-wrapped","page":"Implementing a solver interface","title":"Find a similar solver already wrapped","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The next step is to find (if possible) a similar solver that is already wrapped. Although not strictly necessary, this will be a good place to look for inspiration when implementing your wrapper.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The JuMP documentation has a good list of solvers, along with the problem classes they support.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nIf you're not sure which solver is most similar, ask in the Developer chatroom.","category":"page"},{"location":"moi/tutorials/implementing/#Create-a-low-level-interface","page":"Implementing a solver interface","title":"Create a low-level interface","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Before writing a MathOptInterface wrapper, you first need to be able to call the solver from Julia.","category":"page"},{"location":"moi/tutorials/implementing/#Wrapping-solvers-written-in-Julia","page":"Implementing a solver interface","title":"Wrapping solvers written in Julia","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If your solver is written in Julia, there's nothing to do here. Go to the next section.","category":"page"},{"location":"moi/tutorials/implementing/#Wrapping-solvers-written-in-C","page":"Implementing a solver interface","title":"Wrapping solvers written in C","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Julia is well suited to wrapping solvers written in C.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"info: Info\nThis is not true for C++. If you have a solver written in C++, first write a C interface, then wrap the C interface.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Before writing a MathOptInterface wrapper, there are a few extra steps.","category":"page"},{"location":"moi/tutorials/implementing/#Create-a-JLL","page":"Implementing a solver interface","title":"Create a JLL","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If the C code is publicly available under an open source license, create a JLL package via Yggdrasil. The easiest way to do this is to copy an existing solver. Good examples to follow are the COIN-OR solvers.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"warning: Warning\nBuilding the solver via Yggdrasil is non-trivial. Please ask the Developer chatroom for help.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If the code is commercial or not publicly available, the user will need to manually install the solver. See Gurobi.jl or CPLEX.jl for examples of how to structure this.","category":"page"},{"location":"moi/tutorials/implementing/#implement_clang","page":"Implementing a solver interface","title":"Use Clang.jl to wrap the C API","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The next step is to use Clang.jl to automatically wrap the C API. The easiest way to do this is to follow an example. Good examples to follow are Cbc.jl and HiGHS.jl.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Sometimes, you will need to make manual modifications to the resulting files.","category":"page"},{"location":"moi/tutorials/implementing/#Solvers-written-in-other-languages","page":"Implementing a solver interface","title":"Solvers written in other languages","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Ask the Developer chatroom for advice. You may be able to use one of the JuliaInterop packages to call out to the solver.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For example, SeDuMi.jl uses MATLAB.jl to call the SeDuMi solver written in MATLAB.","category":"page"},{"location":"moi/tutorials/implementing/#Structuring-the-package","page":"Implementing a solver interface","title":"Structuring the package","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Structure your wrapper as a Julia package. Consult the Julia documentation if you haven't done this before.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"MOI solver interfaces may be in the same package as the solver itself (either the C wrapper if the solver is accessible through C, or the Julia code if the solver is written in Julia, for example), or in a separate package which depends on the solver package.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"note: Note\nThe JuMP core contributors request that you do not use \"JuMP\" in the name of your package without prior consent.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Your package should have the following structure:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"/.github\n    /workflows\n        ci.yml\n        format_check.yml\n        TagBot.yml\n/gen\n    gen.jl  # Code to wrap the C API\n/src\n    NewSolver.jl\n    /gen\n        libnewsolver_api.jl\n        libnewsolver_common.jl\n    /MOI_wrapper\n        MOI_wrapper.jl\n        other_files.jl\n/test\n    runtests.jl\n    /MOI_wrapper\n        MOI_wrapper.jl\n.gitignore\n.JuliaFormatter.toml\nREADME.md\nLICENSE.md\nProject.toml","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The /.github folder contains the scripts for GitHub actions. The easiest way to write these is to copy the ones from an existing solver.\nThe /gen and /src/gen folders are only needed if you are wrapping a solver written in C.\nThe /src/MOI_wrapper folder contains the Julia code for the MOI wrapper.\nThe /test folder contains code for testing your package. See Setup tests for more information.\nThe .JuliaFormatter.toml and .github/workflows/format_check.yml enforce code formatting using JuliaFormatter.jl. Check existing solvers or JuMP.jl for details.","category":"page"},{"location":"moi/tutorials/implementing/#Documentation","page":"Implementing a solver interface","title":"Documentation","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Your package must include documentation explaining how to use the package. The easiest approach is to include documentation in your README.md. A more involved option is to use Documenter.jl.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Examples of packages with README-based documentation include:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Cbc.jl\nHiGHS.jl\nSCS.jl","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Examples of packages with Documenter-based documentation include:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Alpine.jl\nCOSMO.jl\nJuniper.jl","category":"page"},{"location":"moi/tutorials/implementing/#Setup-tests","page":"Implementing a solver interface","title":"Setup tests","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The best way to implement an interface to MathOptInterface is via test-driven development.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The MOI.Test submodule contains a large test suite to help check that you have implemented things correctly.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Follow the guide How to test a solver to set up the tests for your package.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nRun the tests frequently when developing. However, at the start there is going to be a lot of errors. Start by excluding large classes of tests (for example, exclude = [\"test_basic_\", \"test_model_\"], implement any missing methods until the tests pass, then remove an exclusion and repeat.","category":"page"},{"location":"moi/tutorials/implementing/#Initial-code","page":"Implementing a solver interface","title":"Initial code","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"By this point, you should have a package setup with tests, formatting, and access to the underlying solver. Now it's time to start writing the wrapper.","category":"page"},{"location":"moi/tutorials/implementing/#The-Optimizer-object","page":"Implementing a solver interface","title":"The Optimizer object","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The first object to create is a subtype of AbstractOptimizer. This type is going to store everything related to the problem.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"By convention, these optimizers should not be exported and should be named PackageName.Optimizer.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"import MathOptInterface as MOI\n\nstruct Optimizer <: MOI.AbstractOptimizer\n    # Fields go here\nend","category":"page"},{"location":"moi/tutorials/implementing/#Optimizer-objects-for-C-solvers","page":"Implementing a solver interface","title":"Optimizer objects for C solvers","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"warning: Warning\nThis section is important if you wrap a solver written in C.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Wrapping a solver written in C will require the use of pointers, and for you to manually free the solver's memory when the Optimizer is garbage collected by Julia.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Never pass a pointer directly to a Julia ccall function.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Instead, store the pointer as a field in your Optimizer, and implement Base.cconvert and Base.unsafe_convert. Then you can pass Optimizer to any ccall function that expects the pointer.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"In addition, make sure you implement a finalizer for each model you create.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If newsolver_createProblem() is the low-level function that creates the problem pointer in C, and newsolver_freeProblem(::Ptr{Cvoid}) is the low-level function that frees memory associated with the pointer, your Optimizer() function should look like this:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"struct Optimizer <: MOI.AbstractOptimizer\n    ptr::Ptr{Cvoid}\n\n    function Optimizer()\n        ptr = newsolver_createProblem()\n        model = Optimizer(ptr)\n        finalizer(model) do m\n            newsolver_freeProblem(m)\n            return\n        end\n        return model\n    end\nend\n\nBase.cconvert(::Type{Ptr{Cvoid}}, model::Optimizer) = model\nBase.unsafe_convert(::Type{Ptr{Cvoid}}, model::Optimizer) = model.ptr","category":"page"},{"location":"moi/tutorials/implementing/#Implement-methods-for-Optimizer","page":"Implementing a solver interface","title":"Implement methods for Optimizer","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"All Optimizers must implement the following methods:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"empty!\nis_empty","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Other methods, detailed below, are optional or depend on how you implement the interface.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nFor this and all future methods, read the docstrings to understand what each method does, what it expects as input, and what it produces as output. If it isn't clear, let us know and we will improve the docstrings. It is also very helpful to look at an existing wrapper for a similar solver.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"You should also implement Base.show(::IO, ::Optimizer) to print a nice string when someone prints your model. For example","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"function Base.show(io::IO, model::Optimizer)\n    return print(io, \"NewSolver with the pointer $(model.ptr)\")\nend","category":"page"},{"location":"moi/tutorials/implementing/#Implement-attributes","page":"Implementing a solver interface","title":"Implement attributes","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"MathOptInterface uses attributes to manage different aspects of the problem.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For each attribute","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"get gets the current value of the attribute\nset sets a new value of the attribute. Not all attributes can be set. For example, the user can't modify the SolverName.\nsupports returns a Bool indicating whether the solver supports the attribute.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"info: Info\nUse attribute_value_type to check the value expected by a given attribute. You should make sure that your get function correctly infers to this type (or a subtype of it).","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Each column in the table indicates whether you need to implement the particular method for each attribute.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Attribute get set supports\nSolverName Yes No No\nSolverVersion Yes No No\nRawSolver Yes No No\nName Yes Yes Yes\nSilent Yes Yes Yes\nTimeLimitSec Yes Yes Yes\nObjectiveLimit Yes Yes Yes\nRawOptimizerAttribute Yes Yes Yes\nNumberOfThreads Yes Yes Yes\nAbsoluteGapTolerance Yes Yes Yes\nRelativeGapTolerance Yes Yes Yes","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For example:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"function MOI.get(model::Optimizer, ::MOI.Silent)\n    return # true if MOI.Silent is set\nend\n\nfunction MOI.set(model::Optimizer, ::MOI.Silent, v::Bool)\n    if v\n        # Set a parameter to turn off printing\n    else\n        # Restore the default printing\n    end\n    return\nend\n\nMOI.supports(::Optimizer, ::MOI.Silent) = true","category":"page"},{"location":"moi/tutorials/implementing/#Define-supports_constraint","page":"Implementing a solver interface","title":"Define supports_constraint","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The next step is to define which constraints and objective functions you plan to support.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For each function-set constraint pair, define supports_constraint:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"function MOI.supports_constraint(\n    ::Optimizer,\n    ::Type{MOI.VariableIndex},\n    ::Type{MOI.ZeroOne},\n)\n    return true\nend","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"To make this easier, you may want to use Unions:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"function MOI.supports_constraint(\n    ::Optimizer,\n    ::Type{MOI.VariableIndex},\n    ::Type{<:Union{MOI.LessThan,MOI.GreaterThan,MOI.EqualTo}},\n)\n    return true\nend","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nOnly support a constraint if your solver has native support for it.","category":"page"},{"location":"moi/tutorials/implementing/#The-big-decision:-incremental-modification?","page":"Implementing a solver interface","title":"The big decision: incremental modification?","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Now you need to decide whether to support incremental modification or not.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Incremental modification means that the user can add variables and constraints one-by-one without needing to rebuild the entire problem, and they can modify the problem data after an optimize! call. Supporting incremental modification means implementing functions like add_variable and add_constraint.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"The alternative is to accept the problem data in a single optimize! or copy_to function call. Because these functions see all of the data at once, it can typically call a more efficient function to load data into the underlying solver.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Good examples of solvers supporting incremental modification are MILP solvers like GLPK.jl and Gurobi.jl. Examples of non-incremental solvers are AmplNLWriter.jl and SCS.jl","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"It is possible for a solver to implement both approaches, but you should probably start with one for simplicity.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nOnly support incremental modification if your solver has native support for it.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"In general, supporting incremental modification is more work, and it usually requires some extra book-keeping. However, it provides a more efficient interface to the solver if the problem is going to be resolved multiple times with small modifications. Moreover, once you've implemented incremental modification, it's usually not much extra work to add a copy_to interface. The converse is not true.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nIf this is your first time writing an interface, start with the one-shot optimize!.","category":"page"},{"location":"moi/tutorials/implementing/#The-non-incremental-interface","page":"Implementing a solver interface","title":"The non-incremental interface","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"There are two ways to implement the non-incremental interface. The first uses a two-argument version of optimize!, the second implements copy_to followed by the one-argument version of optimize!.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If your solver does not support modification, and requires all data to solve the problem in a single function call, you should implement the \"one-shot\" optimize!.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"optimize!(::ModelLike, ::ModelLike)","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If your solver separates data loading and the actual optimization into separate steps, implement the copy_to interface.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"copy_to(::ModelLike, ::ModelLike)\noptimize!(::ModelLike)","category":"page"},{"location":"moi/tutorials/implementing/#The-incremental-interface","page":"Implementing a solver interface","title":"The incremental interface","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"warning: Warning\nWriting this interface is a lot of work. The easiest way is to consult the source code of a similar solver.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"To implement the incremental interface, implement the following functions:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"add_variable\nadd_variables\nadd_constraint\nadd_constraints\nis_valid\ndelete\noptimize!(::ModelLike)","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"info: Info\nSolvers do not have to support AbstractScalarFunction in GreaterThan, LessThan, EqualTo, or Interval with a nonzero constant in the function. Throw ScalarFunctionConstantNotZero if the function constant is not zero.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"In addition, you should implement the following model attributes:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Attribute get set supports\nListOfModelAttributesSet Yes No No\nObjectiveFunctionType Yes No No\nObjectiveFunction Yes Yes Yes\nObjectiveSense Yes Yes Yes\nName Yes Yes Yes","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Variable-related attributes:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Attribute get set supports\nListOfVariableAttributesSet Yes No No\nNumberOfVariables Yes No No\nListOfVariableIndices Yes No No","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Constraint-related attributes:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Attribute get set supports\nListOfConstraintAttributesSet Yes No No\nNumberOfConstraints Yes No No\nListOfConstraintTypesPresent Yes No No\nConstraintFunction Yes Yes No\nConstraintSet Yes Yes No","category":"page"},{"location":"moi/tutorials/implementing/#Modifications","page":"Implementing a solver interface","title":"Modifications","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If your solver supports modifying data in-place, implement modify for the following AbstractModifications:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"ScalarConstantChange\nScalarCoefficientChange\nVectorConstantChange\nMultirowChange","category":"page"},{"location":"moi/tutorials/implementing/#Variables-constrained-on-creation","page":"Implementing a solver interface","title":"Variables constrained on creation","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Some solvers require variables be associated with a set when they are created. This conflicts with the incremental modification approach, since you cannot first add a free variable and then constrain it to the set.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If this is the case, implement:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"add_constrained_variable\nadd_constrained_variables\nsupports_add_constrained_variables","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"By default, MathOptInterface assumes solvers support free variables. If your solver does not support free variables, define:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"MOI.supports_add_constrained_variables(::Optimizer, ::Type{Reals}) = false","category":"page"},{"location":"moi/tutorials/implementing/#Incremental-and-copy_to","page":"Implementing a solver interface","title":"Incremental and copy_to","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If you implement the incremental interface, you have the option of also implementing copy_to.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If you don't want to implement copy_to, for example, because the solver has no API for building the problem in a single function call, define the following fallback:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"MOI.supports_incremental_interface(::Optimizer) = true\n\nfunction MOI.copy_to(dest::Optimizer, src::MOI.ModelLike)\n    return MOI.Utilities.default_copy_to(dest, src)\nend","category":"page"},{"location":"moi/tutorials/implementing/#implement_names","page":"Implementing a solver interface","title":"Names","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Regardless of which interface you implement, you have the option of implementing the Name attribute for variables and constraints:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Attribute get set supports\nVariableName Yes Yes Yes\nConstraintName Yes Yes Yes","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If you implement names, you must also implement the following three methods:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"function MOI.get(model::Optimizer, ::Type{MOI.VariableIndex}, name::String)\n    return # The variable named `name`.\nend\n\nfunction MOI.get(model::Optimizer, ::Type{MOI.ConstraintIndex}, name::String)\n    return # The constraint any type named `name`.\nend\n\nfunction MOI.get(\n    model::Optimizer,\n    ::Type{MOI.ConstraintIndex{F,S}},\n    name::String,\n) where {F,S}\n    return # The constraint of type F-in-S named `name`.\nend","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"These methods have the following rules:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If there is no variable or constraint with the name, return nothing\nIf there is a single variable or constraint with that name, return the variable or constraint\nIf there are multiple variables or constraints with the name, throw an error.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"warning: Warning\nYou should not implement ConstraintName for VariableIndex constraints. If you implement ConstraintName for other constraints, you can add the following two methods to disable ConstraintName for VariableIndex constraints.function MOI.supports(\n    ::Optimizer,\n    ::MOI.ConstraintName,\n    ::Type{<:MOI.ConstraintIndex{MOI.VariableIndex,<:MOI.AbstractScalarSet}},\n)\n    return throw(MOI.VariableIndexConstraintNameError())\nend\nfunction MOI.set(\n    ::Optimizer,\n    ::MOI.ConstraintName,\n    ::MOI.ConstraintIndex{MOI.VariableIndex,<:MOI.AbstractScalarSet},\n    ::String,\n)\n    return throw(MOI.VariableIndexConstraintNameError())\nend","category":"page"},{"location":"moi/tutorials/implementing/#Solutions","page":"Implementing a solver interface","title":"Solutions","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Implement optimize! to solve the model:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"optimize!","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"All Optimizers must implement the following attributes:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"DualStatus\nPrimalStatus\nRawStatusString\nResultCount\nTerminationStatus","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"info: Info\nYou only need to implement get for solution attributes. Don't implement set or supports.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"note: Note\nSolver wrappers should document how the low-level statuses map to the MOI statuses. Statuses like NEARLY_FEASIBLE_POINT and INFEASIBLE_POINT, are designed to be used when the solver explicitly indicates that relaxed tolerances are satisfied or the returned point is infeasible, respectively.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"You should also implement the following attributes:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"ObjectiveValue\nSolveTimeSec\nVariablePrimal","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"tip: Tip\nAttributes like VariablePrimal and ObjectiveValue are indexed by the result count. Use MOI.check_result_index_bounds(model, attr) to throw an error if the attribute is not available.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If your solver returns dual solutions, implement:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"ConstraintDual\nDualObjectiveValue","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For integer solvers, implement:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"ObjectiveBound\nRelativeGap","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If applicable, implement:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"SimplexIterations\nBarrierIterations\nNodeCount","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If your solver uses the Simplex method, implement:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"ConstraintBasisStatus","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"If your solver accepts primal or dual warm-starts, implement:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"VariablePrimalStart\nConstraintDualStart","category":"page"},{"location":"moi/tutorials/implementing/#Other-tips","page":"Implementing a solver interface","title":"Other tips","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Here are some other points to be aware of when writing your wrapper.","category":"page"},{"location":"moi/tutorials/implementing/#Unsupported-constraints-at-runtime","page":"Implementing a solver interface","title":"Unsupported constraints at runtime","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"In some cases, your solver may support a particular type of constraint (for example, quadratic constraints), but only if the data meets some condition (for example, it is convex).","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"In this case, declare that you support the constraint, and throw AddConstraintNotAllowed.","category":"page"},{"location":"moi/tutorials/implementing/#Dealing-with-multiple-variable-bounds","page":"Implementing a solver interface","title":"Dealing with multiple variable bounds","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"MathOptInterface uses VariableIndex constraints to represent variable bounds. Defining multiple variable bounds on a single variable is not allowed.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Throw LowerBoundAlreadySet or UpperBoundAlreadySet if the user adds a constraint that results in multiple bounds.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Only throw if the constraints conflict. It is okay to add VariableIndex-in-GreaterThan and then VariableIndex-in-LessThan, but not VariableIndex-in-Interval and then VariableIndex-in-LessThan,","category":"page"},{"location":"moi/tutorials/implementing/#Expect-duplicate-coefficients","page":"Implementing a solver interface","title":"Expect duplicate coefficients","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Solvers must expect that functions such as ScalarAffineFunction and VectorQuadraticFunction may contain duplicate coefficients.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For example, ScalarAffineFunction([ScalarAffineTerm(x, 1), ScalarAffineTerm(x, 1)], 0.0).","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Use Utilities.canonical to return a new function with the duplicate coefficients aggregated together.","category":"page"},{"location":"moi/tutorials/implementing/#Don't-modify-user-data","page":"Implementing a solver interface","title":"Don't modify user-data","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"All data passed to the solver must be copied immediately to internal data structures. Solvers may not modify any input vectors and must assume that input vectors will not be modified by users in the future.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"This applies, for example, to the terms vector in ScalarAffineFunction. Vectors returned to the user, for example, via ObjectiveFunction or ConstraintFunction attributes, must not be modified by the solver afterwards. The in-place version of get! can be used by users to avoid extra copies in this case.","category":"page"},{"location":"moi/tutorials/implementing/#Column-Generation","page":"Implementing a solver interface","title":"Column Generation","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"There is no special interface for column generation. If the solver has a special API for setting coefficients in existing constraints when adding a new variable, it is possible to queue modifications and new variables and then call the solver's API once all of the new coefficients are known.","category":"page"},{"location":"moi/tutorials/implementing/#Solver-specific-attributes","page":"Implementing a solver interface","title":"Solver-specific attributes","text":"","category":"section"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"You don't need to restrict yourself to the attributes defined in the MathOptInterface.jl package.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Solver-specific attributes should be specified by creating an appropriate subtype of AbstractModelAttribute, AbstractOptimizerAttribute, AbstractVariableAttribute, or AbstractConstraintAttribute.","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"For example, Gurobi.jl adds attributes for multiobjective optimization by defining:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"struct NumberOfObjectives <: MOI.AbstractModelAttribute end\n\nfunction MOI.set(model::Optimizer, ::NumberOfObjectives, n::Integer)\n    # Code to set NumberOfObjectives\n    return\nend\n\nfunction MOI.get(model::Optimizer, ::NumberOfObjectives)\n    n = # Code to get NumberOfObjectives\n    return n\nend","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"Then, the user can write:","category":"page"},{"location":"moi/tutorials/implementing/","page":"Implementing a solver interface","title":"Implementing a solver interface","text":"model = Gurobi.Optimizer()\nMOI.set(model, Gurobi.NumberofObjectives(), 3)","category":"page"},{"location":"moi/submodules/Benchmarks/reference/","page":"API Reference","title":"API Reference","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Benchmarks/reference.md\"","category":"page"},{"location":"moi/submodules/Benchmarks/reference/","page":"API Reference","title":"API Reference","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Benchmarks/reference/#Benchmarks","page":"API Reference","title":"Benchmarks","text":"","category":"section"},{"location":"moi/submodules/Benchmarks/reference/","page":"API Reference","title":"API Reference","text":"Functions to help benchmark the performance of solver wrappers. See The Benchmarks submodule for more details.","category":"page"},{"location":"moi/submodules/Benchmarks/reference/","page":"API Reference","title":"API Reference","text":"Benchmarks.suite\nBenchmarks.create_baseline\nBenchmarks.compare_against_baseline","category":"page"},{"location":"moi/submodules/Benchmarks/reference/#MathOptInterface.Benchmarks.suite","page":"API Reference","title":"MathOptInterface.Benchmarks.suite","text":"suite(\n    new_model::Function;\n    exclude::Vector{Regex} = Regex[]\n)\n\nCreate a suite of benchmarks. new_model should be a function that takes no arguments, and returns a new instance of the optimizer you wish to benchmark.\n\nUse exclude to exclude a subset of benchmarks.\n\nExamples\n\nsuite() do\n    GLPK.Optimizer()\nend\nsuite(exclude = [r\"delete\"]) do\n    Gurobi.Optimizer(OutputFlag=0)\nend\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Benchmarks/reference/#MathOptInterface.Benchmarks.create_baseline","page":"API Reference","title":"MathOptInterface.Benchmarks.create_baseline","text":"create_baseline(suite, name::String; directory::String = \"\"; kwargs...)\n\nRun all benchmarks in suite and save to files called name in directory.\n\nExtra kwargs are based to BenchmarkTools.run.\n\nExamples\n\nmy_suite = suite(() -> GLPK.Optimizer())\ncreate_baseline(my_suite, \"glpk_master\"; directory = \"/tmp\", verbose = true)\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Benchmarks/reference/#MathOptInterface.Benchmarks.compare_against_baseline","page":"API Reference","title":"MathOptInterface.Benchmarks.compare_against_baseline","text":"compare_against_baseline(\n    suite, name::String; directory::String = \"\",\n    report_filename::String = \"report.txt\"\n)\n\nRun all benchmarks in suite and compare against files called name in directory that were created by a call to create_baseline.\n\nA report summarizing the comparison is written to report_filename in directory.\n\nExtra kwargs are based to BenchmarkTools.run.\n\nExamples\n\nmy_suite = suite(() -> GLPK.Optimizer())\ncompare_against_baseline(\n    my_suite, \"glpk_master\"; directory = \"/tmp\", verbose = true\n)\n\n\n\n\n\n","category":"function"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/bridging_constraint.md\"","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/tutorials/bridging_constraint/#Implementing-a-constraint-bridge","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"This guide outlines the basic steps to create a new bridge from a constraint expressed in the formalism Function-in-Set.","category":"page"},{"location":"moi/tutorials/bridging_constraint/#Preliminaries","page":"Implementing a constraint bridge","title":"Preliminaries","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"First, decide on the set you want to bridge. Then, study its properties: the most important one is whether the set is scalar or vector, which impacts the dimensionality of the functions that can be used with the set.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"A scalar function only has one dimension. MOI defines three types of scalar functions: a variable (VariableIndex), an affine function (ScalarAffineFunction), or a quadratic function (ScalarQuadraticFunction).\nA vector function has several dimensions (at least one). MOI defines three types of vector functions: several variables (VectorOfVariables), an affine function (VectorAffineFunction), or a quadratic function (VectorQuadraticFunction). The main difference with scalar functions is that the order of dimensions can be very important: for instance, in an indicator constraint (Indicator), the first dimension indicates whether the constraint about the second dimension is active.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"To explain how to implement a bridge, we present the example of Bridges.Constraint.FlipSignBridge. This bridge maps <= (LessThan) constraints to >= (GreaterThan) constraints. This corresponds to reversing the sign of the inequality. We focus on scalar affine functions (we disregard the cases of a single variable or of quadratic functions). This example is a simplified version of the code included in MOI.","category":"page"},{"location":"moi/tutorials/bridging_constraint/#Four-mandatory-parts-in-a-constraint-bridge","page":"Implementing a constraint bridge","title":"Four mandatory parts in a constraint bridge","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"The first part of a constraint bridge is a new concrete subtype of Bridges.Constraint.AbstractBridge. This type must have fields to store all the new variables and constraints that the bridge will add. Typically, these types are parametrized by the type of the coefficients in the model.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"Then, three sets of functions must be defined:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"Bridges.Constraint.bridge_constraint: this function implements the bridge and creates the required variables and constraints.\nsupports_constraint: these functions must return true when the combination of function and set is supported by the bridge. By default, the base implementation always returns false and the bridge does not have to provide this implementation.\nBridges.added_constrained_variable_types and Bridges.added_constraint_types: these functions return the types of variables and constraints that this bridge adds. They are used to compute the set of other bridges that are required to use the one you are defining, if need be.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"More functions can be implemented, for instance to retrieve properties from the bridge or deleting a bridged constraint.","category":"page"},{"location":"moi/tutorials/bridging_constraint/#.-Structure-for-the-bridge","page":"Implementing a constraint bridge","title":"1. Structure for the bridge","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"A typical struct behind a bridge depends on the type of the coefficients that are used for the model (typically Float64, but coefficients might also be integers or complex numbers).","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"This structure must hold a reference to all the variables and the constraints that are created as part of the bridge.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"The type of this structure is used throughout MOI as an identifier for the bridge. It is passed as argument to most functions related to bridges.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"The best practice is to have the name of this type end with Bridge.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"In our example, the bridge maps any ScalarAffineFunction{T}-in-LessThan{T} constraint to a single ScalarAffineFunction{T}-in-GreaterThan{T} constraint. The affine function has coefficients of type T. The bridge is parametrized with T, so that the constraint that the bridge creates also has coefficients of type T.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"struct SignBridge{T<:Number} <: Bridges.Constraint.AbstractBridge\n    constraint::ConstraintIndex{ScalarAffineFunction{T}, GreaterThan{T}}\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/#.-Bridge-creation","page":"Implementing a constraint bridge","title":"2. Bridge creation","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"The function Bridges.Constraint.bridge_constraint is called whenever the bridge is instantiated for a specific model, with the given function and set. The arguments to bridge_constraint are similar to add_constraint, with the exception of the first argument: it is the Type of the struct defined in the first step (for our example, Type{SignBridge{T}}).","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"bridge_constraint returns an instance of the struct defined in the first step. the first step.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"In our example, the bridge constraint could be defined as:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function Bridges.Constraint.bridge_constraint(\n    ::Type{SignBridge{T}}, # Bridge to use.\n    model::ModelLike, # Model to which the constraint is being added.\n    f::ScalarAffineFunction{T}, # Function to rewrite.\n    s::LessThan{T}, # Set to rewrite.\n) where {T}\n    # Create the variables and constraints required for the bridge.\n    con = add_constraint(model, -f, GreaterThan(-s.upper))\n\n    # Return an instance of the bridge type with a reference to all the\n    # variables and constraints that were created in this function.\n    return SignBridge(con)\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/#.-Supported-constraint-types","page":"Implementing a constraint bridge","title":"3. Supported constraint types","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"The function supports_constraint determines whether the bridge type supports a given combination of function and set.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"This function must closely match bridge_constraint, because it will not be called if supports_constraint returns false.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function supports_constraint(\n    ::Type{SignBridge{T}}, # Bridge to use.\n    ::Type{ScalarAffineFunction{T}}, # Function to rewrite.\n    ::Type{LessThan{T}}, # Set to rewrite.\n) where {T}\n    # Do some computation to ensure that the constraint is supported.\n    # Typically, you can directly return true.\n    return true\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/#.-Metadata-about-the-bridge","page":"Implementing a constraint bridge","title":"4. Metadata about the bridge","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"To determine whether a bridge can be used, MOI uses a shortest-path algorithm that uses the variable types and the constraints that the bridge can create. This information is communicated from the bridge to MOI using the functions Bridges.added_constrained_variable_types and Bridges.added_constraint_types. Both return lists of tuples: either a list of 1-tuples containing the variable types (typically, ZeroOne or Integer) or a list of 2-tuples contained the functions and sets (like ScalarAffineFunction{T}-GreaterThan).","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"For our example, the bridge does not create any constrained variables, and only ScalarAffineFunction{T}-in-GreaterThan{T} constraints:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function Bridges.added_constrained_variable_types(::Type{SignBridge{T}}) where {T}\n    # The bridge does not create variables, return an empty list of tuples:\n    return Tuple{Type}[]\nend\n\nfunction Bridges.added_constraint_types(::Type{SignBridge{T}}) where {T}\n    return Tuple{Type,Type}[\n        # One element per F-in-S the bridge creates.\n        (ScalarAffineFunction{T}, GreaterThan{T}),\n    ]\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"A bridge that creates binary variables would rather have this definition of added_constrained_variable_types:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function Bridges.added_constrained_variable_types(::Type{SomeBridge{T}}) where {T}\n    # The bridge only creates binary variables:\n    return Tuple{Type}[(ZeroOne,)]\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"warning: Warning\nIf you declare the creation of constrained variables in added_constrained_variable_types, the corresponding constraint type VariableIndex must not be indicated in added_constraint_types. This would restrict the use of the bridge to solvers that can add such a constraint after the variable is created.More concretely, if you declare in added_constrained_variable_types that your bridge creates binary variables (ZeroOne), and if you never add such a constraint afterward (you do not call add_constraint(model, var, ZeroOne())), then you must not list (VariableIndex, ZeroOne) in added_constraint_types.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"Typically, the function Bridges.Constraint.concrete_bridge_type does not have to be defined for most bridges.","category":"page"},{"location":"moi/tutorials/bridging_constraint/#Bridge-registration","page":"Implementing a constraint bridge","title":"Bridge registration","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"For a bridge to be used by MOI, it must be known by MOI.","category":"page"},{"location":"moi/tutorials/bridging_constraint/#SingleBridgeOptimizer","page":"Implementing a constraint bridge","title":"SingleBridgeOptimizer","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"The first way to do so is to create a single-bridge optimizer. This type of optimizer wraps another optimizer and adds the possibility to use only one bridge. It is especially useful when unit testing bridges.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"It is common practice to use the same name as the type defined for the bridge (SignBridge, in our example) without the suffix Bridge.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"const Sign{T,OT<: ModelLike} =\n    SingleBridgeOptimizer{SignBridge{T}, OT}","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"In the context of unit tests, this bridge is used in conjunction with a Utilities.MockOptimizer:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"mock = Utilities.MockOptimizer(\n    Utilities.UniversalFallback(Utilities.Model{Float64}()),\n)\nbridged_mock = Sign{Float64}(mock)","category":"page"},{"location":"moi/tutorials/bridging_constraint/#New-bridge-for-a-LazyBridgeOptimizer","page":"Implementing a constraint bridge","title":"New bridge for a LazyBridgeOptimizer","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"Typical user-facing models for MOI are based on Bridges.LazyBridgeOptimizer. For instance, this type of model is returned by Bridges.full_bridge_optimizer. These models can be added more bridges by using Bridges.add_bridge:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"inner_optimizer = Utilities.Model{Float64}()\noptimizer = Bridges.full_bridge_optimizer(inner_optimizer, Float64)\nBridges.add_bridge(optimizer, SignBridge{Float64})","category":"page"},{"location":"moi/tutorials/bridging_constraint/#Bridge-improvements","page":"Implementing a constraint bridge","title":"Bridge improvements","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/#Attribute-retrieval","page":"Implementing a constraint bridge","title":"Attribute retrieval","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"Like models, bridges have attributes that can be retrieved using get and set. The most important ones are the number of variables and constraints, but also the lists of variables and constraints.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"In our example, we only have one constraint and only have to implement the NumberOfConstraints and ListOfConstraintIndices attributes:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function get(\n    ::SignBridge{T},\n    ::NumberOfConstraints{\n        ScalarAffineFunction{T},\n        GreaterThan{T},\n    },\n) where {T}\n    return 1\nend\n\nfunction get(\n    bridge::SignBridge{T},\n    ::ListOfConstraintIndices{\n        ScalarAffineFunction{T},\n        GreaterThan{T},\n    },\n) where {T}\n    return [bridge.constraint]\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"You must implement one such pair of functions for each type of constraint the bridge adds to the model.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"warning: Warning\nAvoid returning a list from the bridge object without copying it. Users must be able to change the contents of the returned list without altering the bridge object.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"For variables, the situation is simpler. If your bridge creates new variables, you must implement the NumberOfVariables and ListOfVariableIndices attributes. However, these attributes do not have parameters, unlike their constraint counterparts. Only two functions suffice:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function get(\n    ::SignBridge{T},\n    ::NumberOfVariables,\n) where {T}\n    return 0\nend\n\nfunction get(\n    ::SignBridge{T},\n    ::ListOfVariableIndices,\n) where {T}\n    return VariableIndex[]\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"In order for the user to be able to access the function and set of the original constraint, the bridge needs to implement getters for the ConstraintFunction and ConstraintSet attributes:","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function get(\n    model::MOI.ModelLike,\n    attr::MOI.ConstraintFunction,\n    bridge::SignBridge,\n)\n    return -MOI.get(model, attr, bridge.constraint)\nend\n\nfunction get(\n    model::MOI.ModelLike,\n    attr::MOI.ConstraintSet,\n    bridge::SignBridge,\n)\n    set = MOI.get(model, attr, bridge.constraint)\n    return MOI.LessThan(-set.lower)\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"warning: Warning\nAlternatively, one could store the original function and set in SignBridge during Bridges.Constraint.bridge_constraint to make these getters simpler and more efficient. On the other hand, this will increase the memory footprint of the bridges as the garbage collector won't be able to delete that object. The convention is to not store the function in the bridge and not care too much about the efficiency of these getters. If the user needs efficient getters for ConstraintFunction then they should use a Utilities.CachingOptimizer.","category":"page"},{"location":"moi/tutorials/bridging_constraint/#Model-modifications","page":"Implementing a constraint bridge","title":"Model modifications","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"To avoid copying the model when the user request to change a constraint, MOI provides modify. Bridges can also implement this API to allow certain changes, such as coefficient changes.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"In our case, a modification of a coefficient in the original constraint (for example, replacing the value of the coefficient of a variable in the affine function) must be transmitted to the constraint created by the bridge, but with a sign change.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function modify(\n    model::ModelLike,\n    bridge::SignBridge,\n    change::ScalarCoefficientChange,\n)\n    modify(\n        model,\n        bridge.constraint,\n        ScalarCoefficientChange(change.variable, -change.new_coefficient),\n    )\n    return\nend","category":"page"},{"location":"moi/tutorials/bridging_constraint/#Bridge-deletion","page":"Implementing a constraint bridge","title":"Bridge deletion","text":"","category":"section"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"When a bridge is deleted, the constraints it added must be deleted too.","category":"page"},{"location":"moi/tutorials/bridging_constraint/","page":"Implementing a constraint bridge","title":"Implementing a constraint bridge","text":"function delete(model::ModelLike, bridge::SignBridge)\n    delete(model, bridge.constraint)\n    return\nend","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"EditURL = \"experiment_design.jl\"","category":"page"},{"location":"tutorials/conic/experiment_design/#Experiment-design","page":"Experiment design","title":"Experiment design","text":"","category":"section"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"This tutorial was originally contributed by Arpit Bhatia and Chris Coey.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"This tutorial covers experiment design examples (D-optimal, A-optimal, and E-optimal) from section 7.5 of the book Convex Optimization by Boyd and Vandenberghe.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"The tutorial uses the following packages","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"using JuMP\nimport SCS\nimport LinearAlgebra\nimport Random","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"info: Info\nThis tutorial uses sets from MathOptInterface. By default, JuMP exports the MOI symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:import MathOptInterface as MOI","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"We set a seed so the random numbers are repeatable:","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"Random.seed!(1234)","category":"page"},{"location":"tutorials/conic/experiment_design/#The-relaxed-experiment-design-problem","page":"Experiment design","title":"The relaxed experiment design problem","text":"","category":"section"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"The basic experiment design problem is as follows.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"Given the menu of possible choices for experiments, v_1 ldots v_p, and the total number m of experiments to be carried out, choose the numbers of each type of experiment, that is, m_1 ldots m_p to make the error covariance E small (in  some sense).","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"The variables m_1 ldots m_p must, of course, be integers and sum to m the given total number of experiments. This leads to the optimization problem:","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"beginaligned\nminleft(mathrmwrt mathbfS_+^nright)  E=left(sum_j=1^p m_j v_j v_j^Tright)^-1 \ntextsubject to  m_i geq 0 \n sumlimits_i=1^p m_i = m \n m_i in mathbbZquad i=1ldotsp\nendaligned","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"The basic experiment design problem can be a hard combinatorial problem when m, the total number of experiments, is comparable to n, since in this case the m_i are all small integers.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"In the case when m is large compared to n, however, a good approximate solution can be found by ignoring, or relaxing, the constraint that the m_i are integers.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"Let lambda_i=m_i  m which is the fraction of the total number of experiments for which a_j=v_i or the relative frequency of experiment i. We can express the error covariance in terms of lambda_i as:","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"E=frac1mleft(sum_i=1^p lambda_i v_i v_i^Tright)^-1","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"The vector lambda in mathbfR^p satisfies lambda succeq 0 mathbf1^T lambda=1 and also, each lambda_i is an integer multiple of 1  m. By ignoring this last constraint, we arrive at the problem:","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"beginaligned\nminleft(mathrmwrt mathbfS_+^nright)  E=(1  m)left(sum_i=1^p lambda_i v_i v_i^Tright)^-1 \ntextsubject toquad  lambda succeq 0 \n mathbf1^T lambda=1\nendaligned","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"Several scalarizations have been proposed for the experiment design problem, which is a vector optimization problem over the positive semidefinite cone.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"q = 4 # dimension of estimate space\np = 8 # number of experimental vectors\nn_max = 3 # upper bound on lambda\nn = 12\n\nV = randn(q, p)\n\neye = Matrix{Float64}(LinearAlgebra.I, q, q);\nnothing #hide","category":"page"},{"location":"tutorials/conic/experiment_design/#A-optimal-design","page":"Experiment design","title":"A-optimal design","text":"","category":"section"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"In A-optimal experiment design, we minimize tr E, the trace of the covariance matrix. This objective is simply the mean of the norm of the error squared:","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"mathbfEe_2^2=mathbfE operatornametrleft(e e^Tright)=operatornametr E","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"The A-optimal experiment design problem in SDP form is","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"beginaligned\nmin  mathbf1^T u \ntextsubject to  leftbeginalignedsum_i=1^p lambda_i v_i v_i^T  e_k  e_k^T  u_kendalignedright succeq 0 quad k=1 ldots n \n lambda succeq 0 \n  mathbf1^T lambda=1\nendaligned","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"aOpt = Model(SCS.Optimizer)\nset_silent(aOpt)\n@variable(aOpt, np[1:p], lower_bound = 0, upper_bound = n_max)\n@variable(aOpt, u[1:q], lower_bound = 0)\n@constraint(aOpt, sum(np) <= n)\nfor i in 1:q\n    matrix = [\n        V*LinearAlgebra.diagm(0 => np ./ n)*V' eye[:, i]\n        eye[i, :]' u[i]\n    ]\n    @constraint(aOpt, matrix >= 0, PSDCone())\nend\n@objective(aOpt, Min, sum(u))\noptimize!(aOpt)\nobjective_value(aOpt)","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"value.(np)","category":"page"},{"location":"tutorials/conic/experiment_design/#E-optimal-design","page":"Experiment design","title":"E-optimal design","text":"","category":"section"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"In E -optimal design, we minimize the norm of the error covariance matrix, that is, the maximum eigenvalue of E.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"Since the diameter (twice the longest semi-axis) of the confidence ellipsoid mathcalE is proportional to E_2^1  2, minimizing E_2 can be interpreted geometrically as minimizing the diameter of the confidence ellipsoid.","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"E-optimal design can also be interpreted as minimizing the maximum variance of q^T e, over all q with q_2=1. The E-optimal experiment design problem in SDP form is:","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"beginaligned\nmin  t \ntextsubject to  sum_i=1^p lambda_i v_i v_i^T succeq t I \n lambda succeq 0 \n mathbf1^T lambda=1\nendaligned","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"eOpt = Model(SCS.Optimizer)\nset_silent(eOpt)\n@variable(eOpt, 0 <= np[1:p] <= n_max)\n@variable(eOpt, t)\n@constraint(\n    eOpt,\n    V * LinearAlgebra.diagm(0 => np ./ n) * V' - (t .* eye) >= 0,\n    PSDCone(),\n)\n@constraint(eOpt, sum(np) <= n)\n@objective(eOpt, Max, t)\noptimize!(eOpt)\nobjective_value(eOpt)","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"value.(np)","category":"page"},{"location":"tutorials/conic/experiment_design/#D-optimal-design","page":"Experiment design","title":"D-optimal design","text":"","category":"section"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"The most widely used scalarization is called D -optimal design, in which we minimize the determinant of the error covariance matrix E. This corresponds to designing the experiment to minimize the volume of the resulting confidence ellipsoid (for a fixed confidence level). Ignoring the constant factor 1  m in E, and taking the logarithm of the objective, we can pose this problem as convex optimization problem:","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"beginaligned\nmin  log operatornamedetleft(sum_i=1^p lambda_i v_i v_i^Tright)^-1 \ntextsubject to  lambda succeq 0 \n mathbf1^T lambda=1\nendaligned","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"dOpt = Model(SCS.Optimizer)\nset_silent(dOpt)\n@variable(dOpt, np[1:p], lower_bound = 0, upper_bound = n_max)\n@variable(dOpt, t)\n@objective(dOpt, Max, t)\n@constraint(dOpt, sum(np) <= n)\nE = V * LinearAlgebra.diagm(0 => np ./ n) * V'\n@constraint(dOpt, [t; 1; triangle_vec(E)] in MOI.LogDetConeTriangle(q))\noptimize!(dOpt)\nobjective_value(dOpt)","category":"page"},{"location":"tutorials/conic/experiment_design/","page":"Experiment design","title":"Experiment design","text":"value.(np)","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/nonlinear/#Nonlinear-Modeling","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"warning: Warning\nThis page describes a new nonlinear interface to JuMP. It replaces the legacy @NL interface, which is documented at Nonlinear Modeling (Legacy). The API described below is stable, and it will not break with future 1.X releases of JuMP. However, solver support may be limited, and there may be gaps in functionality compared with the legacy interface. To report a bug, or request a missing feature, please open an issue.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"JuMP has support for nonlinear (convex and nonconvex) optimization problems. JuMP is able to automatically provide exact, sparse second-order derivatives to solvers. This information can improve solver accuracy and performance.","category":"page"},{"location":"manual/nonlinear/#Set-a-nonlinear-objective","page":"Nonlinear Modeling","title":"Set a nonlinear objective","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Use @objective to set a nonlinear objective.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @objective(model, Min, exp(x[1]) - sqrt(x[2]))\nexp(x[1]) - sqrt(x[2])","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"To modify a nonlinear objective, call @objective again.","category":"page"},{"location":"manual/nonlinear/#Add-a-nonlinear-constraint","page":"Nonlinear Modeling","title":"Add a nonlinear constraint","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Use @constraint to add a nonlinear constraint.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> @constraint(model, exp(x[1]) <= 1)\nexp(x[1]) - 1.0 ≤ 0\n\njulia> @constraint(model, con[i = 1:2], 2^x[i] >= i)\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarNonlinearFunction, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:\n con[1] : (2.0 ^ x[1]) - 1.0 ≥ 0\n con[2] : (2.0 ^ x[2]) - 2.0 ≥ 0","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Delete a nonlinear constraint using delete:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> delete(model, con[1])","category":"page"},{"location":"manual/nonlinear/#Create-a-nonlinear-expression","page":"Nonlinear Modeling","title":"Create a nonlinear expression","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Use @expression to create nonlinear expression objects:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2]);\n\njulia> expr = @expression(model, exp(x[1]) + sqrt(x[2]))\nexp(x[1]) + sqrt(x[2])\n\njulia> my_anon_expr = @expression(model, [i = 1:2], sin(x[i]))\n2-element Vector{NonlinearExpr}:\n sin(x[1])\n sin(x[2])\n\njulia> @expression(model, my_expr[i = 1:2], sin(x[i]))\n2-element Vector{NonlinearExpr}:\n sin(x[1])\n sin(x[2])","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"A NonlinearExpr can be used in @objective, @constraint, and even nested in other @expressions.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> @objective(model, Min, expr^2 + 1)\n((exp(x[1]) + sqrt(x[2])) ^ 2.0) + 1.0\n\njulia> @constraint(model, [i = 1:2], my_expr[i] <= i)\n2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarNonlinearFunction, MathOptInterface.LessThan{Float64}}, ScalarShape}}:\n sin(x[1]) - 1.0 ≤ 0\n sin(x[2]) - 2.0 ≤ 0\n\njulia> @expression(model, nested[i = 1:2], sin(my_expr[i]))\n2-element Vector{NonlinearExpr}:\n sin(sin(x[1]))\n sin(sin(x[2]))","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Use value to query the value of a nonlinear expression:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> set_start_value(x[1], 1.0)\n\njulia> value(start_value, nested[1])\n0.7456241416655579\n\njulia> sin(sin(1.0))\n0.7456241416655579","category":"page"},{"location":"manual/nonlinear/#Nonlinear-expressions-in-detail","page":"Nonlinear Modeling","title":"Nonlinear expressions in detail","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Nonlinear expressions in JuMP are represented by a NonlinearExpr object.","category":"page"},{"location":"manual/nonlinear/#Constructors","page":"Nonlinear Modeling","title":"Constructors","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Nonlinear expressions can be created using the NonlinearExpr constructors:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr = NonlinearExpr(:sin, Any[x])\nsin(x)","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"or via operator overloading:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr = sin(x)\nsin(x)","category":"page"},{"location":"manual/nonlinear/#Supported-arguments","page":"Nonlinear Modeling","title":"Supported arguments","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Nonlinear expressions can contain a mix of numbers, AffExpr, QuadExpr, and other NonlinearExpr:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> aff = x + 1;\n\njulia> quad = x^2 + x;\n\njulia> expr = cos(x) * sin(quad) + aff\n(cos(x) * sin(x² + x)) + (x + 1)","category":"page"},{"location":"manual/nonlinear/#Supported-operators","page":"Nonlinear Modeling","title":"Supported operators","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The list of supported operators may vary between solvers. Given an optimizer, query the list of supported operators using MOI.ListOfSupportedNonlinearOperators:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> import Ipopt\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.get(Ipopt.Optimizer(), MOI.ListOfSupportedNonlinearOperators())\n85-element Vector{Symbol}:\n :+\n :-\n :abs\n :sqrt\n :cbrt\n :abs2\n :inv\n :log\n :log10\n :log2\n ⋮\n :min\n :max\n :&&\n :||\n :<=\n :(==)\n :>=\n :<\n :>","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"In some univariate cases, the operator is defined in SpecialFunctions.jl. To use these functions, you must explicitly import SpecialFunctions.jl","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> import Ipopt\n\njulia> op = MOI.get(Ipopt.Optimizer(), MOI.ListOfSupportedNonlinearOperators());\n\njulia> :erfcx in op\ntrue\n\njulia> :dawson in op\ntrue\n\njulia> import SpecialFunctions\n\njulia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @expression(model, SpecialFunctions.erfcx(x))\nerfcx(x)\n\njulia> @expression(model, SpecialFunctions.dawson(x))\ndawson(x)","category":"page"},{"location":"manual/nonlinear/#Limitations","page":"Nonlinear Modeling","title":"Limitations","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Some nonlinear expressions cannot be created via operator overloading. For example, to minimize the likelihood of bugs in user-code, we have not overloaded comparisons such as < and >= between JuMP objects:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> x < 1\nERROR: Cannot evaluate `<` between a variable and a number.\n[...]","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Instead, wrap the expression in the @expression macro:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr = @expression(model, x < 1)\nx < 1","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"For technical reasons, other operators that are not overloaded include ||, &&, and ifelse.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr = @expression(model, ifelse(x < -1 || x >= 1, x^2, 0.0))\nifelse((x < -1) || (x >= 1), x², 0.0)","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"As an alternative, use the JuMP.op_ functions, which fallback to the various comparison and logical operators:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> expr = op_ifelse(\n           op_or(op_strictly_less_than(x, -1), op_greater_than_or_equal_to(x, 1)),\n           x^2,\n           0.0,\n       )\nifelse((x < -1) || (x >= 1), x², 0.0)","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The available functions are:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"JuMP function Julia function\nop_ifelse ifelse\nop_and &&\nop_or ||\nop_greater_than_or_equal_to >=\nop_less_than_or_equal_to <=\nop_equal_to ==\nop_strictly_greater_than >\nop_strictly_less_than <","category":"page"},{"location":"manual/nonlinear/#Fields","page":"Nonlinear Modeling","title":"Fields","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Each NonlinearExpr has two fields.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The .head field is a Symbol that represents the operator being called:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> expr.head\n:sin","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The .args field is a Vector{Any} containing the arguments to the operator:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> expr.args\n1-element Vector{Any}:\n x","category":"page"},{"location":"manual/nonlinear/#jump_user_defined_operators","page":"Nonlinear Modeling","title":"User-defined operators","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"In addition to a standard list of univariate and multivariate operators recognized by the MOI.Nonlinear submodule, JuMP supports user-defined Julia operators.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"warning: Warning\nUser-defined operators must return a scalar output. For a work-around, see User-defined operators with vector outputs.","category":"page"},{"location":"manual/nonlinear/#Add-an-operator","page":"Nonlinear Modeling","title":"Add an operator","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Add a user-defined operator using the @operator macro:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"using JuMP\nsquare(x) = x^2\nf(x, y) = (x - 1)^2 + (y - 2)^2\nmodel = Model();\n@operator(model, op_square, 1, square)\n@operator(model, op_f, 2, f)\n@variable(model, x[1:2]);\n@objective(model, Min, op_f(x[1], op_square(x[2])))","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The arguments to @operator are:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The model to which the operator is added.\nA Julia symbol object which serves as the name of the user-defined operator in JuMP expressions. This name must not be the same as that of the function.\nThe number of scalar input arguments that the function takes.\nA Julia method which computes the function.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"warning: Warning\nUser-defined operators cannot be deleted.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"You can obtain a reference to the operator using the model[:key] syntax:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"using JuMP\nsquare(x) = x^2\nmodel = Model();\n@operator(model, op_square, 1, square)\nop_square_2 = model[:op_square]","category":"page"},{"location":"manual/nonlinear/#Add-an-operator-without-macros","page":"Nonlinear Modeling","title":"Add an operator without macros","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The @operator macro is syntactic sugar for add_nonlinear_operator. Thus, the non-macro version of the preceding example is:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"using JuMP\nsquare(x) = x^2\nf(x, y) = (x - 1)^2 + (y - 2)^2\nmodel = Model();\nop_square = add_nonlinear_operator(model, 1, square; name = :op_square)\nmodel[:op_square] = op_square\nop_f = add_nonlinear_operator(model, 2, f; name = :op_f)\nmodel[:op_f] = op_f\n@variable(model, x[1:2]);\n@objective(model, Min, op_f(x[1], op_square(x[2])))","category":"page"},{"location":"manual/nonlinear/#Operators-with-the-same-name-as-an-existing-function","page":"Nonlinear Modeling","title":"Operators with the same name as an existing function","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"A common error encountered is the following:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> using JuMP\n\njulia> model = Model();\n\njulia> f(x) = x^2\nf (generic function with 1 method)\n\njulia> @operator(model, f, 1, f)\nERROR: Unable to add the nonlinear operator `:f` with the same name as\nan existing function.\n[...]","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"This error occurs because @operator(model, f, 1, f) is equivalent to:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> f = add_nonlinear_operator(model, 1, f; name = :f)","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"but f already exists as a Julia function.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"If you evaluate the function without adding it as an operator, JuMP will trace the function using operator overloading:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> @variable(model, x);\n\njulia> f(x)\nx²","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"To force JuMP to treat f as a user-defined operator and not trace it, add the operator using add_nonlinear_operator and define a new method which manually creates a NonlinearExpr:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"julia> _ = add_nonlinear_operator(model, 1, f; name = :f)\nNonlinearOperator(f, :f)\n\njulia> f(x::AbstractJuMPScalar) = NonlinearExpr(:f, Any[x])\nf (generic function with 2 methods)\n\njulia> @expression(model, log(f(x)))\nlog(f(x))","category":"page"},{"location":"manual/nonlinear/#Gradients-and-Hessians","page":"Nonlinear Modeling","title":"Gradients and Hessians","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"By default, JuMP will use automatic differentiation to compute the gradient and Hessian of user-defined operators. If your function is not amenable to automatic differentiation, or you can compute analytic derivatives, you may pass additional arguments to @operator to compute the first- and second-derivatives.","category":"page"},{"location":"manual/nonlinear/#Univariate-functions","page":"Nonlinear Modeling","title":"Univariate functions","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"For univariate functions, a gradient function ∇f returns a number that represents the first-order derivative. You may, in addition, pass a third function which returns a number representing the second-order derivative:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"using JuMP\nf(x) = x^2\n∇f(x) = 2x\n∇²f(x) = 2\nmodel = Model();\n@operator(model, op_f, 1, f, ∇f, ∇²f)  # Providing ∇²f is optional\n@variable(model, x)\n@objective(model, Min, op_f(x))","category":"page"},{"location":"manual/nonlinear/#Multivariate-functions","page":"Nonlinear Modeling","title":"Multivariate functions","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"For multivariate functions, the gradient function ∇f must take an AbstractVector as the first argument that is filled in-place. The Hessian function, ∇²f, must take an AbstractMatrix as the first argument, the lower-triangular of which is filled in-place:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"using JuMP\nf(x...) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2\nfunction ∇f(g::AbstractVector{T}, x::T...) where {T}\n    g[1] = 400 * x[1]^3 - 400 * x[1] * x[2] + 2 * x[1] - 2\n    g[2] = 200 * (x[2] - x[1]^2)\n    return\nend\nfunction ∇²f(H::AbstractMatrix{T}, x::T...) where {T}\n    H[1, 1] = 1200 * x[1]^2 - 400 * x[2] + 2\n    # H[1, 2] = -400 * x[1]  <-- Not needed. Fill the lower-triangular only.\n    H[2, 1] = -400 * x[1]\n    H[2, 2] = 200.0\n    return\nend\nmodel = Model();\n@operator(model, rosenbrock, 2, f, ∇f, ∇²f)  # Providing ∇²f is optional\n@variable(model, x[1:2])\n@objective(model, Min, rosenbrock(x[1], x[2]))","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"You may assume the Hessian matrix H is initialized with zeros, and because H is symmetric, you need only to fill in the non-zero lower-triangular terms. The matrix type passed in as H depends on the automatic differentiation system, so make sure the first argument to the Hessian function supports an AbstractMatrix (it may be something other than Matrix{Float64}). Moreover, you may assume only that H supports size(H) and setindex!. Finally, the matrix is treated as dense, so the performance will be poor on functions with high-dimensional input.","category":"page"},{"location":"manual/nonlinear/#User-defined-operators-with-vector-inputs","page":"Nonlinear Modeling","title":"User-defined operators with vector inputs","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"User-defined operators which take vectors as input arguments (for example, f(x::Vector)) are not supported. Instead, use Julia's splatting syntax to create a function with scalar arguments. For example, instead of:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"f(x::Vector) = sum(x[i]^i for i in 1:length(x))","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"define:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"f(x...) = sum(x[i]^i for i in 1:length(x))","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Another approach is to define the splatted function as an anonymous function:","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"using JuMP\nmodel = Model();\n@variable(model, x[1:5])\nf(x::Vector) = sum(x[i]^i for i in 1:length(x))\n@operator(model, op_f, 5, (x...) -> f(collect(x)))\n@objective(model, Min, op_f(x...))","category":"page"},{"location":"manual/nonlinear/#Automatic-differentiation","page":"Nonlinear Modeling","title":"Automatic differentiation","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"JuMP does not support black-box optimization, so all user-defined operators must provide derivatives in some form. Fortunately, JuMP supports automatic differentiation of user-defined operators.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"info: Info\nAutomatic differentiation is not finite differencing. JuMP's automatically computed derivatives are not subject to approximation error.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"JuMP uses ForwardDiff.jl to perform automatic differentiation of user-defined operators; see the ForwardDiff.jl documentation for a description of how to write a function suitable for automatic differentiation.","category":"page"},{"location":"manual/nonlinear/#Common-mistakes-when-writing-a-user-defined-operator","page":"Nonlinear Modeling","title":"Common mistakes when writing a user-defined operator","text":"","category":"section"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"warning: Warning\nGet an error like No method matching Float64(::ForwardDiff.Dual)? Read this section, and see the guidelines at ForwardDiff.jl.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"The most common error is that your user-defined operator is not generic with respect to the number type, that is, don't assume that the input to the function is Float64.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"f(x::Float64) = 2 * x  # This will not work.\nf(x::Real)    = 2 * x  # This is good.\nf(x)          = 2 * x  # This is also good.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"Another reason you may encounter this error is if you create arrays inside your function which are Float64.","category":"page"},{"location":"manual/nonlinear/","page":"Nonlinear Modeling","title":"Nonlinear Modeling","text":"function bad_f(x...)\n    y = zeros(length(x))  # This constructs an array of `Float64`!\n    for i in 1:length(x)\n        y[i] = x[i]^i\n    end\n    return sum(y)\nend\n\nfunction good_f(x::T...) where {T<:Real}\n    y = zeros(T, length(x))  # Construct an array of type `T` instead!\n    for i in 1:length(x)\n        y[i] = x[i]^i\n    end\n    return sum(y)\nend","category":"page"},{"location":"moi/background/naming_conventions/","page":"Naming conventions","title":"Naming conventions","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/background/naming_conventions.md\"","category":"page"},{"location":"moi/background/naming_conventions/","page":"Naming conventions","title":"Naming conventions","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/background/naming_conventions/#Naming-conventions","page":"Naming conventions","title":"Naming conventions","text":"","category":"section"},{"location":"moi/background/naming_conventions/","page":"Naming conventions","title":"Naming conventions","text":"MOI follows several conventions for naming functions and structures. These should also be followed by packages extending MOI.","category":"page"},{"location":"moi/background/naming_conventions/#Sets","page":"Naming conventions","title":"Sets","text":"","category":"section"},{"location":"moi/background/naming_conventions/","page":"Naming conventions","title":"Naming conventions","text":"Sets encode the structure of constraints. Their names should follow the following conventions:","category":"page"},{"location":"moi/background/naming_conventions/","page":"Naming conventions","title":"Naming conventions","text":"Abstract types in the set hierarchy should begin with Abstract and end in Set, for example, AbstractScalarSet, AbstractVectorSet.\nVector-valued conic sets should end with Cone, for example, NormInfinityCone, SecondOrderCone.\nVector-valued Cartesian products should be plural and not end in Cone, for example, Nonnegatives, not NonnegativeCone.\nMatrix-valued conic sets should provide two representations: ConeSquare and ConeTriangle, for example, RootDetConeTriangle and RootDetConeSquare. See Matrix cones for more details.\nScalar sets should be singular, not plural, for example, Integer, not Integers.\nAs much as possible, the names should follow established conventions in the domain where this set is used: for instance, convex sets should have names close to those of CVX, and constraint-programming sets should follow MiniZinc's constraints.","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"EditURL = \"https://github.com/oxfordcontrol/CDCS.jl/blob/6d165e8e8f59b14e36e99c8efb19f0e775bbd292/README.md\"","category":"page"},{"location":"packages/CDCS/#CDCS.jl","page":"oxfordcontrol/CDCS.jl","title":"CDCS.jl","text":"","category":"section"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"CDCS.jl is a wrapper for the CDCS solver. ","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"an exported cdcs function that is a thin wrapper on top of the cdcs MATLAB function\nan interface to MathOptInterface","category":"page"},{"location":"packages/CDCS/#License","page":"oxfordcontrol/CDCS.jl","title":"License","text":"","category":"section"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"CDCS.jl is licensed under the MIT license.","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"The underlying solver oxfordcontrol/CDCS is licensed under the LGPL v3 license.","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"In addition, CDCS requires an installation of MATLAB, which is a closed-source commercial product for which you must obtain a license.","category":"page"},{"location":"packages/CDCS/#Installation","page":"oxfordcontrol/CDCS.jl","title":"Installation","text":"","category":"section"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"First, make sure that you satisfy the requirements of the MATLAB.jl Julia package and that the CDCS software is installed in your MATLAB™ installation.","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"Then, install CDCS.jl using Pkg.add:","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"import Pkg\nPkg.add(\"CDCS\")","category":"page"},{"location":"packages/CDCS/#Use-with-JuMP","page":"oxfordcontrol/CDCS.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"To use CDCS with JuMP, do:","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"using JuMP, CDCS\nmodel = Model(CDCS.Optimizer)\nset_attribute(model, \"verbose\", 0)","category":"page"},{"location":"packages/CDCS/#Troubleshooting","page":"oxfordcontrol/CDCS.jl","title":"Troubleshooting","text":"","category":"section"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"If you get the error:","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"Undefined function or variable 'cdcs'.\n\nError using save\nVariable 'jx_cdcs_arg_out_1' not found.\n\nLinear Programming example: Error During Test at /home/blegat/.julia/dev/CDCS/test/lp.jl:5\n  Got exception outside of a @test\n  MATLAB.MEngineError(\"failed to get variable jx_cdcs_arg_out_1 from MATLAB session\")\n  Stacktrace:\n    [1] get_mvariable(session::MATLAB.MSession, name::Symbol)\n      @ MATLAB ~/.julia/packages/MATLAB/SVjnA/src/engine.jl:164\n    [2] mxcall(::MATLAB.MSession, ::Symbol, ::Int64, ::Matrix{Float64}, ::Vararg{Any})\n      @ MATLAB ~/.julia/packages/MATLAB/SVjnA/src/engine.jl:297\n    [3] mxcall\n      @ ~/.julia/packages/MATLAB/SVjnA/src/engine.jl:317 [inlined]\n    [4] cdcs(A::Matrix{Float64}, b::Vector{Float64}, c::Vector{Float64}, K::CDCS.Cone; kws::Base.Pairs{Symbol, Int64, Tuple{Symbol}, NamedTuple{(:verbose,), Tuple{Int64}}})","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"The error means that we could not find the cdcs function with one output argument using the MATLAB C API.","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"This most likely means that you did not add CDCS to the MATLAB's path (that is, the toolbox/local/pathdef.m file).","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"If modifying toolbox/local/pathdef.m does not work, the following should work where /path/to/CDCS/ is the directory where the CDCS folder is located:","category":"page"},{"location":"packages/CDCS/","page":"oxfordcontrol/CDCS.jl","title":"oxfordcontrol/CDCS.jl","text":"julia> using MATLAB\n\njulia> cd(\"/path/to/CDCS/\") do\n           mat\"cdcsInstall\"\n       end\n\njulia> mat\"savepath\"","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"EditURL = \"https://github.com/jump-dev/MultiObjectiveAlgorithms.jl/blob/v1.2.0/README.md\"","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"<img src=\"https://raw.githubusercontent.com/jump-dev/MultiObjectiveAlgorithms.jl/master/moa.png?raw=true\" alt=\"An image of the Moa bird. Licensed into the Public Domain by https://freesvg.org/moa\" width=\"100px\"/>","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/#MultiObjectiveAlgorithms.jl","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"MultiObjectiveAlgorithms.jl","text":"","category":"section"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"MultiObjectiveAlgorithms.jl (MOA) is a collection of algorithms for multi-objective optimization.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/#License","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"License","text":"","category":"section"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"MultiObjectiveAlgorithms.jl is licensed under the MPL 2.0 License.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/#Installation","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"Installation","text":"","category":"section"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"Install MOA using Pkg.add:","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"import Pkg\nPkg.add(\"MultiObjectiveAlgorithms\")","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/#Use-with-JuMP","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"Use MultiObjectiveAlgorithms with JuMP as follows:","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"using JuMP\nimport HiGHS\nimport MultiObjectiveAlgorithms as MOA\nmodel = JuMP.Model(() -> MOA.Optimizer(HiGHS.Optimizer))\nset_attribute(model, MOA.Algorithm(), MOA.Dichotomy())\nset_attribute(model, MOA.SolutionLimit(), 4)","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"Replace HiGHS.Optimizer with an optimizer capable of solving a single-objective instance of your optimization problem.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"You may set additional optimizer attributes, the supported attributes depend on the choice of solution algorithm.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/#Algorithm","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"Algorithm","text":"","category":"section"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"Set the algorithm using the MOA.Algorithm() attribute.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"The value must be one of the algorithms supported by MOA:","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"MOA.Chalmet()\nMOA.Dichotomy()\nMOA.DominguezRios()\nMOA.EpsilonConstraint()\nMOA.Hierarchical()\nMOA.KirlikSayin()\nMOA.Lexicographic() [default]\nMOA.TambyVanderpooten()","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"Consult their docstrings for details.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/#Other-optimizer-attributes","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"Other optimizer attributes","text":"","category":"section"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"There are a number of optimizer attributes supported by the algorithms in MOA.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"Each algorithm supports only a subset of the attributes. Consult the algorithm's docstring for details on which attributes it supports, and how it uses them in the solution process.","category":"page"},{"location":"packages/MultiObjectiveAlgorithms/","page":"jump-dev/MultiObjectiveAlgorithms.jl","title":"jump-dev/MultiObjectiveAlgorithms.jl","text":"MOA.EpsilonConstraintStep()\nMOA.LexicographicAllPermutations()\nMOA.ObjectiveAbsoluteTolerance(index::Int)\nMOA.ObjectivePriority(index::Int)\nMOA.ObjectiveRelativeTolerance(index::Int)\nMOA.ObjectiveWeight(index::Int)\nMOA.SolutionLimit()\nMOI.TimeLimitSec()","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"EditURL = \"multi_commodity_network.jl\"","category":"page"},{"location":"tutorials/linear/multi_commodity_network/#The-network-multi-commodity-flow-problem","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"","category":"section"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"This tutorial is a variation of The multi-commodity flow problem where the graph is a network instead of a bipartite graph.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The purpose of this tutorial is to demonstrate a style of modeling that uses relational algebra.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/#Required-packages","page":"The network multi-commodity flow problem","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"using JuMP\nimport DataFrames\nimport HiGHS\nimport SQLite\nimport SQLite.DBInterface","category":"page"},{"location":"tutorials/linear/multi_commodity_network/#Formulation","page":"The network multi-commodity flow problem","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The network multi-commondity flow problem is an extension of the The multi-commodity flow problem, where instead of having a bipartite graph of supply and demand nodes, the graph can contains a set of nodes, i in mathcalN , which each have a (potentially zero) supply capacity, u^s_ip, and (potentially zero) a demand, d_ip for each commodity p in P. The nodes are connected by a set of edges (i j) in mathcalE, which have a shipment cost c^x_ijp and a total flow capacity of u^x_ij.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"Our take is to choose an optimal supply for each node s_ip, as well as the optimal transshipment x_ijp that minimizes the total cost.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The mathematical formulation is:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"beginaligned\nmin   sum_(ij)inmathcalE p in P c^x_ijp x_ijp + sum_iinmathcalN p in P c^s_ip s_ip \nst   s_ip + sum_(j i) in mathcalE x_jip - sum_(ij) in mathcalE x_ijp = d_ip  forall i in mathcalN p in P \n           x_ijp ge 0            forall (i j) in mathcalE p in P \n           sum_p in P x_ijp le u^x_ij            forall (i j) in mathcalE \n           0 le s_ip le u^s_ip  forall i in mathcalN p in P\nendaligned","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The purpose of this tutorial is to demonstrate how this model can be built using relational algebra instead of a direct math-to-code translation of the summations.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/#Data","page":"The network multi-commodity flow problem","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"For the purpose of this tutorial, the JuMP repository contains an example database called commodity_nz.db:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"filename = joinpath(@__DIR__, \"commodity_nz.db\");\nnothing #hide","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"To run locally, download commodity_nz.db and update filename appropriately.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"Load the database using SQLite.DB:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"db = SQLite.DB(filename)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"A quick way to see the schema of the database is via SQLite.tables:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"SQLite.tables(db)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"We interact with the database by executing queries and then loading the results into a DataFrame:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"function get_table(db, table)\n    query = DBInterface.execute(db, \"SELECT * FROM $table\")\n    return DataFrames.DataFrame(query)\nend","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The shipping table contains the set of arcs and their distances:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_shipping = get_table(db, \"shipping\")","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The products table contains the shipping cost per kilometer of each product:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_products = get_table(db, \"products\")","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The supply table contains the supply capacity of each node, as well as the cost:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_supply = get_table(db, \"supply\")","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"The demand table contains the demand of each node:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_demand = get_table(db, \"demand\")","category":"page"},{"location":"tutorials/linear/multi_commodity_network/#JuMP-formulation","page":"The network multi-commodity flow problem","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"We start by creating a model and our decision variables:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"For the shipping decisions, we create a new column in df_shipping called x_flow, which has one non-negative decision variable for each row:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_shipping.x_flow = @variable(model, x[1:size(df_shipping, 1)] >= 0)\ndf_shipping","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"For the supply, we add a variable to each row, and then set the upper bound to the capacity of each supply node:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_supply.x_supply = @variable(model, s[1:size(df_supply, 1)] >= 0)\nset_upper_bound.(df_supply.x_supply, df_supply.capacity)\ndf_supply","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"Our objective is to minimize the shipping cost plus the supply cost. To compute the flow cost, we need to join the shipping table, which contains distance_km with the products table, which contains cost_per_km:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_cost = DataFrames.leftjoin(df_shipping, df_products; on = [:product])\ndf_cost.flow_cost = df_cost.cost_per_km .* df_cost.distance_km\ndf_cost","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"Then we can use linear algebra to compute the inner product between two columns:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"@objective(\n    model,\n    Min,\n    df_cost.flow_cost' * df_shipping.x_flow +\n    df_supply.cost' * df_supply.x_supply\n);\nnothing #hide","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"For the flow capacities on each arc, we use DataFrames.groupby to partition the flow variables based on :origin and :destination, and then we constrain their sum to be less than a fixed capacity.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"capacity = 30\nfor df in DataFrames.groupby(df_shipping, [:origin, :destination])\n    @constraint(model, sum(df.x_flow) <= capacity)\nend","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"For each node in the graph, we need to compute a mass balance constraint which says that for each product, the supply, plus the flow into the node, and less the flow out of the node is equal to the demand.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"We can compute an expression for the flow out of each node using DataFrames.groupby on the origin and product columns of the df_shipping table:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_flow_out = DataFrames.DataFrame(\n    (node = i.origin, product = i.product, x_flow_out = sum(df.x_flow)) for\n    (i, df) in pairs(DataFrames.groupby(df_shipping, [:origin, :product]))\n)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"We can compute an expression for the flow into each node using DataFrames.groupby on the destination and product columns of the df_shipping table:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_flow_in = DataFrames.DataFrame(\n    (node = i.destination, product = i.product, x_flow_in = sum(df.x_flow))\n    for (i, df) in\n    pairs(DataFrames.groupby(df_shipping, [:destination, :product]))\n)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"We can join the two tables together using DataFrames.outerjoin. We need to use outerjoin here because there might be missing rows.","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df = DataFrames.outerjoin(df_flow_in, df_flow_out; on = [:node, :product])","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"Next, we need to join the supply column:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df = DataFrames.leftjoin(\n    df,\n    DataFrames.select(df_supply, [:origin, :product, :x_supply]);\n    on = [:node => :origin, :product],\n)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"and then the demand column","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df = DataFrames.leftjoin(\n    df,\n    DataFrames.select(df_demand, [:destination, :product, :demand]);\n    on = [:node => :destination, :product],\n)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"Now we're ready to add our mass balance constraint. Because some rows contain missing values, we need to use coalesce to convert any missing into a numeric value:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"@constraint(\n    model,\n    [r in eachrow(df)],\n    coalesce(r.x_supply, 0.0) + coalesce(r.x_flow_in, 0.0) -\n    coalesce(r.x_flow_out, 0.0) == coalesce(r.demand, 0.0),\n);\nnothing #hide","category":"page"},{"location":"tutorials/linear/multi_commodity_network/#Solution","page":"The network multi-commodity flow problem","title":"Solution","text":"","category":"section"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"Finally, we can optimize the model:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"update the solution in the DataFrames:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"df_shipping.x_flow = value.(df_shipping.x_flow)\ndf_supply.x_supply = value.(df_supply.x_supply);\nnothing #hide","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"and display the optimal solution for flows:","category":"page"},{"location":"tutorials/linear/multi_commodity_network/","page":"The network multi-commodity flow problem","title":"The network multi-commodity flow problem","text":"DataFrames.select(\n    filter!(row -> row.x_flow > 0.0, df_shipping),\n    [:origin, :destination, :product, :x_flow],\n)","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/standard_form.md\"","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/standard_form/#Standard-form","page":"Standard form","title":"Standard form","text":"","category":"section"},{"location":"moi/reference/standard_form/#Functions","page":"Standard form","title":"Functions","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"AbstractFunction\noutput_dimension\nconstant","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractFunction","page":"Standard form","title":"MathOptInterface.AbstractFunction","text":"AbstractFunction\n\nAbstract supertype for function objects.\n\nRequired methods\n\nAll functions must implement:\n\nBase.copy\nBase.isapprox\nconstant\n\nAbstract subtypes of AbstractFunction may require additional methods to be implemented.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.output_dimension","page":"Standard form","title":"MathOptInterface.output_dimension","text":"output_dimension(f::AbstractFunction)\n\nReturn 1 if f is an AbstractScalarFunction, or the number of output components if f is an AbstractVectorFunction.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#MathOptInterface.constant","page":"Standard form","title":"MathOptInterface.constant","text":"constant(f::AbstractFunction[, ::Type{T}]) where {T}\n\nReturns the constant term of a scalar-valued function, or the constant vector of a vector-valued function.\n\nIf f is untyped and T is provided, returns zero(T).\n\n\n\n\n\nconstant(set::Union{EqualTo,GreaterThan,LessThan,Parameter})\n\nReturns the constant term of the set set.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.constant(MOI.GreaterThan(1.0))\n1.0\n\njulia> MOI.constant(MOI.LessThan(2.5))\n2.5\n\njulia> MOI.constant(MOI.EqualTo(3))\n3\n\njulia> MOI.constant(MOI.Parameter(4.5))\n4.5\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#Scalar-functions","page":"Standard form","title":"Scalar functions","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"AbstractScalarFunction\nVariableIndex\nScalarAffineTerm\nScalarAffineFunction\nScalarQuadraticTerm\nScalarQuadraticFunction\nScalarNonlinearFunction","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractScalarFunction","page":"Standard form","title":"MathOptInterface.AbstractScalarFunction","text":"abstract type AbstractScalarFunction <: AbstractFunction\n\nAbstract supertype for scalar-valued AbstractFunctions.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.VariableIndex","page":"Standard form","title":"MathOptInterface.VariableIndex","text":"VariableIndex\n\nA type-safe wrapper for Int64 for use in referencing variables in a model. To allow for deletion, indices need not be consecutive.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.ScalarAffineTerm","page":"Standard form","title":"MathOptInterface.ScalarAffineTerm","text":"ScalarAffineTerm{T}(coefficient::T, variable::VariableIndex) where {T}\n\nRepresents the scalar-valued term coefficient * variable.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1)\nMOI.VariableIndex(1)\n\njulia> MOI.ScalarAffineTerm(2.0, x)\nMathOptInterface.ScalarAffineTerm{Float64}(2.0, MOI.VariableIndex(1))\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction","page":"Standard form","title":"MathOptInterface.ScalarAffineFunction","text":"ScalarAffineFunction{T}(terms::ScalarAffineTerm{T}, constant::T) where {T}\n\nRepresents the scalar-valued affine function a^top x + b, where:\n\na^top x is represented by the vector of ScalarAffineTerms\nb is a scalar constant::T\n\nDuplicates\n\nDuplicate variable indices in terms are accepted, and the corresponding coefficients are summed together.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1)\nMOI.VariableIndex(1)\n\njulia> terms = [MOI.ScalarAffineTerm(2.0, x), MOI.ScalarAffineTerm(3.0, x)]\n2-element Vector{MathOptInterface.ScalarAffineTerm{Float64}}:\n MathOptInterface.ScalarAffineTerm{Float64}(2.0, MOI.VariableIndex(1))\n MathOptInterface.ScalarAffineTerm{Float64}(3.0, MOI.VariableIndex(1))\n\njulia> f = MOI.ScalarAffineFunction(terms, 4.0)\n4.0 + 2.0 MOI.VariableIndex(1) + 3.0 MOI.VariableIndex(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.ScalarQuadraticTerm","page":"Standard form","title":"MathOptInterface.ScalarQuadraticTerm","text":"ScalarQuadraticTerm{T}(\n    coefficient::T,\n    variable_1::VariableIndex,\n    variable_2::VariableIndex,\n) where {T}\n\nRepresents the scalar-valued term c x_i x_j where c is coefficient, x_i is variable_1 and x_j is variable_2.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1)\nMOI.VariableIndex(1)\n\njulia> MOI.ScalarQuadraticTerm(2.0, x, x)\nMathOptInterface.ScalarQuadraticTerm{Float64}(2.0, MOI.VariableIndex(1), MOI.VariableIndex(1))\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.ScalarQuadraticFunction","page":"Standard form","title":"MathOptInterface.ScalarQuadraticFunction","text":"ScalarQuadraticFunction{T}(\n    quadratic_terms::Vector{ScalarQuadraticTerm{T}},\n    affine_terms::Vector{ScalarAffineTerm{T}},\n    constant::T,\n) wher {T}\n\nThe scalar-valued quadratic function frac12x^top Q x + a^top x + b, where:\n\nQ is the symmetric matrix given by the vector of ScalarQuadraticTerms\na^top x is a sparse vector given by the vector of ScalarAffineTerms\nb is the scalar constant::T.\n\nDuplicates\n\nDuplicate indices in quadratic_terms or affine_terms are accepted, and the corresponding coefficients are summed together.\n\nIn quadratic_terms, \"mirrored\" indices, (q, r) and (r, q) where r and q are VariableIndexes, are considered duplicates; only one needs to be specified.\n\nThe 0.5 factor\n\nCoupled with the interpretation of mirrored indices, the 0.5 factor in front of the Q matrix is a common source of bugs.\n\nAs a rule, to represent a * x^2 + b * x * y:\n\nThe coefficient a in front of squared variables (diagonal elements in Q) must be doubled when creating a ScalarQuadraticTerm\nThe coefficient b in front of off-diagonal elements in Q should be left as b, be cause the mirrored index b * y * x will be implicitly added.\n\nExample\n\nTo represent the function f(x y) = 2 * x^2 + 3 * x * y + 4 * x + 5, do:\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1);\n\njulia> y = MOI.VariableIndex(2);\n\njulia> constant = 5.0;\n\njulia> affine_terms = [MOI.ScalarAffineTerm(4.0, x)];\n\njulia> quadratic_terms = [\n           MOI.ScalarQuadraticTerm(4.0, x, x),  # Note the changed coefficient\n           MOI.ScalarQuadraticTerm(3.0, x, y),\n       ]\n2-element Vector{MathOptInterface.ScalarQuadraticTerm{Float64}}:\n MathOptInterface.ScalarQuadraticTerm{Float64}(4.0, MOI.VariableIndex(1), MOI.VariableIndex(1))\n MathOptInterface.ScalarQuadraticTerm{Float64}(3.0, MOI.VariableIndex(1), MOI.VariableIndex(2))\n\njulia> f = MOI.ScalarQuadraticFunction(quadratic_terms, affine_terms, constant)\n5.0 + 4.0 MOI.VariableIndex(1) + 2.0 MOI.VariableIndex(1)² + 3.0 MOI.VariableIndex(1)*MOI.VariableIndex(2)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.ScalarNonlinearFunction","page":"Standard form","title":"MathOptInterface.ScalarNonlinearFunction","text":"ScalarNonlinearFunction(head::Symbol, args::Vector{Any})\n\nThe scalar-valued nonlinear function head(args...), represented as a symbolic expression tree, with the call operator head and ordered arguments in args.\n\nhead\n\nThe head::Symbol must be an operator supported by the model.\n\nThe default list of supported univariate operators is given by:\n\nNonlinear.DEFAULT_UNIVARIATE_OPERATORS\n\nand the default list of supported multivariate operators is given by:\n\nNonlinear.DEFAULT_MULTIVARIATE_OPERATORS\n\nAdditional operators can be registered by setting a UserDefinedFunction attribute.\n\nSee the full list of operators supported by a ModelLike by querying ListOfSupportedNonlinearOperators.\n\nargs\n\nThe vector args contains the arguments to the nonlinear function. If the operator is univariate, it must contain one element. Otherwise, it may contain multiple elements.\n\nEach element must be one of the following:\n\nA constant value of type T<:Real\nA VariableIndex\nA ScalarAffineFunction\nA ScalarQuadraticFunction\nA ScalarNonlinearFunction\n\nUnsupported operators\n\nIf the optimizer does not support head, an UnsupportedNonlinearOperator error will be thrown.\n\nThere is no guarantee about when this error will be thrown; it may be thrown when the function is first added to the model, or it may be thrown when optimize! is called.\n\nExample\n\nTo represent the function f(x) = sin(x)^2, do:\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1)\nMOI.VariableIndex(1)\n\njulia> MOI.ScalarNonlinearFunction(\n           :^,\n           Any[MOI.ScalarNonlinearFunction(:sin, Any[x]), 2],\n       )\n^(sin(MOI.VariableIndex(1)), (2))\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#Vector-functions","page":"Standard form","title":"Vector functions","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"AbstractVectorFunction\nVectorOfVariables\nVectorAffineTerm\nVectorAffineFunction\nVectorQuadraticTerm\nVectorQuadraticFunction\nVectorNonlinearFunction","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractVectorFunction","page":"Standard form","title":"MathOptInterface.AbstractVectorFunction","text":"abstract type AbstractVectorFunction <: AbstractFunction\n\nAbstract supertype for vector-valued AbstractFunctions.\n\nRequired methods\n\nAll subtypes of AbstractVectorFunction must implement:\n\noutput_dimension\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.VectorOfVariables","page":"Standard form","title":"MathOptInterface.VectorOfVariables","text":"VectorOfVariables(variables::Vector{VariableIndex}) <: AbstractVectorFunction\n\nThe vector-valued function f(x) = variables, where variables is a subset of VariableIndexes in the model.\n\nThe list of variables may contain duplicates.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex.(1:2)\n2-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n\njulia> f = MOI.VectorOfVariables([x[1], x[2], x[1]])\n┌                    ┐\n│MOI.VariableIndex(1)│\n│MOI.VariableIndex(2)│\n│MOI.VariableIndex(1)│\n└                    ┘\n\njulia> MOI.output_dimension(f)\n3\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.VectorAffineTerm","page":"Standard form","title":"MathOptInterface.VectorAffineTerm","text":"VectorAffineTerm{T}(\n    output_index::Int64,\n    scalar_term::ScalarAffineTerm{T},\n) where {T}\n\nA VectorAffineTerm is a scalar_term that appears in the output_index row of the vector-valued VectorAffineFunction or VectorQuadraticFunction.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1);\n\njulia> MOI.VectorAffineTerm(Int64(2), MOI.ScalarAffineTerm(3.0, x))\nMathOptInterface.VectorAffineTerm{Float64}(2, MathOptInterface.ScalarAffineTerm{Float64}(3.0, MOI.VariableIndex(1)))\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.VectorAffineFunction","page":"Standard form","title":"MathOptInterface.VectorAffineFunction","text":"VectorAffineFunction{T}(\n    terms::Vector{VectorAffineTerm{T}},\n    constants::Vector{T},\n) where {T}\n\nThe vector-valued affine function A x + b, where:\n\nA x is the sparse matrix given by the vector of VectorAffineTerms\nb is the vector constants\n\nDuplicates\n\nDuplicate indices in the A are accepted, and the corresponding coefficients are summed together.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1);\n\njulia> terms = [\n           MOI.VectorAffineTerm(Int64(1), MOI.ScalarAffineTerm(2.0, x)),\n           MOI.VectorAffineTerm(Int64(2), MOI.ScalarAffineTerm(3.0, x)),\n       ];\n\njulia> f = MOI.VectorAffineFunction(terms, [4.0, 5.0])\n┌                              ┐\n│4.0 + 2.0 MOI.VariableIndex(1)│\n│5.0 + 3.0 MOI.VariableIndex(1)│\n└                              ┘\n\njulia> MOI.output_dimension(f)\n2\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.VectorQuadraticTerm","page":"Standard form","title":"MathOptInterface.VectorQuadraticTerm","text":"VectorQuadraticTerm{T}(\n    output_index::Int64,\n    scalar_term::ScalarQuadraticTerm{T},\n) where {T}\n\nA VectorQuadraticTerm is a ScalarQuadraticTerm scalar_term that appears in the output_index row of the vector-valued VectorQuadraticFunction.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1);\n\njulia> MOI.VectorQuadraticTerm(Int64(2), MOI.ScalarQuadraticTerm(3.0, x, x))\nMathOptInterface.VectorQuadraticTerm{Float64}(2, MathOptInterface.ScalarQuadraticTerm{Float64}(3.0, MOI.VariableIndex(1), MOI.VariableIndex(1)))\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.VectorQuadraticFunction","page":"Standard form","title":"MathOptInterface.VectorQuadraticFunction","text":"VectorQuadraticFunction{T}(\n    quadratic_terms::Vector{VectorQuadraticTerm{T}},\n    affine_terms::Vector{VectorAffineTerm{T}},\n    constants::Vector{T},\n) where {T}\n\nThe vector-valued quadratic function with ith component (\"output index\") defined as frac12x^top Q_i x + a_i^top x + b_i, where:\n\nfrac12x^top Q_i x is the symmetric matrix given by the VectorQuadraticTerm elements in quadratic_terms with output_index == i\na_i^top x is the sparse vector given by the VectorAffineTerm elements in affine_terms with output_index == i\nb_i is a scalar given by constants[i]\n\nDuplicates\n\nDuplicate indices in quadratic_terms and affine_terms with the same output_index are handled in the same manner as duplicates in ScalarQuadraticFunction.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1);\n\njulia> y = MOI.VariableIndex(2);\n\njulia> constants = [4.0, 5.0];\n\njulia> affine_terms = [\n           MOI.VectorAffineTerm(Int64(1), MOI.ScalarAffineTerm(2.0, x)),\n           MOI.VectorAffineTerm(Int64(2), MOI.ScalarAffineTerm(3.0, x)),\n       ];\n\njulia> quad_terms = [\n        MOI.VectorQuadraticTerm(Int64(1), MOI.ScalarQuadraticTerm(2.0, x, x)),\n        MOI.VectorQuadraticTerm(Int64(2), MOI.ScalarQuadraticTerm(3.0, x, y)),\n           ];\n\njulia> f = MOI.VectorQuadraticFunction(quad_terms, affine_terms, constants)\n┌                                                                              ┐\n│4.0 + 2.0 MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(1)²                    │\n│5.0 + 3.0 MOI.VariableIndex(1) + 3.0 MOI.VariableIndex(1)*MOI.VariableIndex(2)│\n└                                                                              ┘\n\njulia> MOI.output_dimension(f)\n2\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.VectorNonlinearFunction","page":"Standard form","title":"MathOptInterface.VectorNonlinearFunction","text":"VectorNonlinearFunction(args::Vector{ScalarNonlinearFunction})\n\nThe vector-valued nonlinear function composed of a vector of ScalarNonlinearFunction.\n\nargs\n\nThe vector args contains the scalar elements of the nonlinear function. Each element must be a ScalarNonlinearFunction, but if you pass a Vector{Any}, the elements can be automatically converted from one of the following:\n\nA constant value of type T<:Real\nA VariableIndex\nA ScalarAffineFunction\nA ScalarQuadraticFunction\nA ScalarNonlinearFunction\n\nExample\n\nTo represent the function f(x) = sin(x)^2 x, do:\n\njulia> import MathOptInterface as MOI\n\njulia> x = MOI.VariableIndex(1)\nMOI.VariableIndex(1)\n\njulia> g = MOI.ScalarNonlinearFunction(\n           :^,\n           Any[MOI.ScalarNonlinearFunction(:sin, Any[x]), 2.0],\n       )\n^(sin(MOI.VariableIndex(1)), 2.0)\n\njulia> MOI.VectorNonlinearFunction([g, x])\n┌                                 ┐\n│^(sin(MOI.VariableIndex(1)), 2.0)│\n│+(MOI.VariableIndex(1))          │\n└                                 ┘\n\nNote the automatic conversion from x to +(x).\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#Sets","page":"Standard form","title":"Sets","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"AbstractSet\nAbstractScalarSet\nAbstractVectorSet","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractSet","page":"Standard form","title":"MathOptInterface.AbstractSet","text":"AbstractSet\n\nAbstract supertype for set objects used to encode constraints.\n\nRequired methods\n\nFor sets of type S with isbitstype(S) == false, you must implement:\n\nBase.copy(set::S)\nBase.:(==)(x::S, y::S)\n\nSubtypes of AbstractSet such as AbstractScalarSet and AbstractVectorSet may prescribe additional required methods.\n\nOptional methods\n\nYou may optionally implement:\n\ndual_set\ndual_set_type\n\nNote for developers\n\nWhen creating a new set, the set struct must not contain any VariableIndex or ConstraintIndex objects.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractScalarSet","page":"Standard form","title":"MathOptInterface.AbstractScalarSet","text":"AbstractScalarSet\n\nAbstract supertype for subsets of mathbbR.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractVectorSet","page":"Standard form","title":"MathOptInterface.AbstractVectorSet","text":"AbstractVectorSet\n\nAbstract supertype for subsets of mathbbR^n for some n.\n\nRequired methods\n\nAll AbstractVectorSets of type S must implement:\n\ndimension, unless the dimension is stored in the set.dimension field\nUtilities.set_dot, unless the dot product between two vectors in the set is equivalent to LinearAlgebra.dot.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#Utilities","page":"Standard form","title":"Utilities","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"dimension\ndual_set\ndual_set_type\nconstant(s::EqualTo)\nsupports_dimension_update\nupdate_dimension","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.dimension","page":"Standard form","title":"MathOptInterface.dimension","text":"dimension(set::AbstractSet)\n\nReturn the output_dimension that an AbstractFunction should have to be used with the set set.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.dimension(MOI.Reals(4))\n4\n\njulia> MOI.dimension(MOI.LessThan(3.0))\n1\n\njulia> MOI.dimension(MOI.PositiveSemidefiniteConeTriangle(2))\n3\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#MathOptInterface.dual_set","page":"Standard form","title":"MathOptInterface.dual_set","text":"dual_set(set::AbstractSet)\n\nReturn the dual set of set, that is the dual cone of the set. This follows the definition of duality discussed in Duality.\n\nSee Dual cone for more information.\n\nIf the dual cone is not defined it returns an error.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.dual_set(MOI.Reals(4))\nMathOptInterface.Zeros(4)\n\njulia> MOI.dual_set(MOI.SecondOrderCone(5))\nMathOptInterface.SecondOrderCone(5)\n\njulia> MOI.dual_set(MOI.ExponentialCone())\nMathOptInterface.DualExponentialCone()\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#MathOptInterface.dual_set_type","page":"Standard form","title":"MathOptInterface.dual_set_type","text":"dual_set_type(S::Type{<:AbstractSet})\n\nReturn the type of dual set of sets of type S, as returned by dual_set. If the dual cone is not defined it returns an error.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.dual_set_type(MOI.Reals)\nMathOptInterface.Zeros\n\njulia> MOI.dual_set_type(MOI.SecondOrderCone)\nMathOptInterface.SecondOrderCone\n\njulia> MOI.dual_set_type(MOI.ExponentialCone)\nMathOptInterface.DualExponentialCone\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#MathOptInterface.constant-Tuple{MathOptInterface.EqualTo}","page":"Standard form","title":"MathOptInterface.constant","text":"constant(set::Union{EqualTo,GreaterThan,LessThan,Parameter})\n\nReturns the constant term of the set set.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> MOI.constant(MOI.GreaterThan(1.0))\n1.0\n\njulia> MOI.constant(MOI.LessThan(2.5))\n2.5\n\njulia> MOI.constant(MOI.EqualTo(3))\n3\n\njulia> MOI.constant(MOI.Parameter(4.5))\n4.5\n\n\n\n\n\n","category":"method"},{"location":"moi/reference/standard_form/#MathOptInterface.supports_dimension_update","page":"Standard form","title":"MathOptInterface.supports_dimension_update","text":"supports_dimension_update(S::Type{<:MOI.AbstractVectorSet})\n\nReturn a Bool indicating whether the elimination of any dimension of n-dimensional sets of type S give an n-1-dimensional set S. By default, this function returns false so it should only be implemented for sets that supports dimension update.\n\nFor instance, supports_dimension_update(MOI.Nonnegatives) is true because the elimination of any dimension of the n-dimensional nonnegative orthant gives the n-1-dimensional nonnegative orthant. However supports_dimension_update(MOI.ExponentialCone) is false.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#MathOptInterface.update_dimension","page":"Standard form","title":"MathOptInterface.update_dimension","text":"update_dimension(s::AbstractVectorSet, new_dim::Int)\n\nReturns a set with the dimension modified to new_dim.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#Scalar-sets","page":"Standard form","title":"Scalar sets","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"List of recognized scalar sets.","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"GreaterThan\nLessThan\nEqualTo\nInterval\nInteger\nZeroOne\nSemicontinuous\nSemiinteger\nParameter","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.GreaterThan","page":"Standard form","title":"MathOptInterface.GreaterThan","text":"GreaterThan{T<:Real}(lower::T)\n\nThe set lower infty) subseteq mathbbR.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.GreaterThan(0.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.GreaterThan{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.LessThan","page":"Standard form","title":"MathOptInterface.LessThan","text":"LessThan{T<:Real}(upper::T)\n\nThe set (-infty upper subseteq mathbbR.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.LessThan(2.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.LessThan{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.EqualTo","page":"Standard form","title":"MathOptInterface.EqualTo","text":"EqualTo{T<:Number}(value::T)\n\nThe set containing the single point value subseteq mathbbR.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.EqualTo(2.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.EqualTo{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Interval","page":"Standard form","title":"MathOptInterface.Interval","text":"Interval{T<:Real}(lower::T, upper::T)\n\nThe interval lower upper subseteq mathbbR cup -infty +infty.\n\nIf lower or upper is -Inf or Inf, respectively, the set is interpreted as a one-sided interval.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.Interval(1.0, 2.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Integer","page":"Standard form","title":"MathOptInterface.Integer","text":"Integer()\n\nThe set of integers, mathbbZ.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.Integer())\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.ZeroOne","page":"Standard form","title":"MathOptInterface.ZeroOne","text":"ZeroOne()\n\nThe set 0 1.\n\nVariables belonging to the ZeroOne set are also known as \"binary\" variables.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.ZeroOne())\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Semicontinuous","page":"Standard form","title":"MathOptInterface.Semicontinuous","text":"Semicontinuous{T<:Real}(lower::T, upper::T)\n\nThe set 0 cup lower upper.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.Semicontinuous(2.0, 3.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Semicontinuous{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Semiinteger","page":"Standard form","title":"MathOptInterface.Semiinteger","text":"Semiinteger{T<:Real}(lower::T, upper::T)\n\nThe set 0 cup lower lower+1 ldots upper-1 upper.\n\nNote that if lower and upper are not equivalent to an integer, then the solver may throw an error, or it may round up lower and round down upper to the nearest integers.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.add_constraint(model, x, MOI.Semiinteger(2.0, 3.0))\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Semiinteger{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Parameter","page":"Standard form","title":"MathOptInterface.Parameter","text":"Parameter{T<:Number}(value::T)\n\nThe set containing the single point value subseteq mathbbR.\n\nThe Parameter set is conceptually similar to the EqualTo set, except that a variable constrained to the Parameter set cannot have other constraints added to it, and the Parameter set can never be deleted. Thus, solvers are free to treat the variable as a constant, and they need not add it as a decision variable to the model.\n\nBecause of this behavior, you must add parameters using add_constrained_variable, and solvers should declare supports_add_constrained_variable and not supports_constraint for the Parameter set.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> p, ci = MOI.add_constrained_variable(model, MOI.Parameter(2.5))\n(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Parameter{Float64}}(1))\n\njulia> MOI.set(model, MOI.ConstraintSet(), ci, MOI.Parameter(3.0))\n\njulia> MOI.get(model, MOI.ConstraintSet(), ci)\nMathOptInterface.Parameter{Float64}(3.0)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#Vector-sets","page":"Standard form","title":"Vector sets","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"List of recognized vector sets.","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"Reals\nZeros\nNonnegatives\nNonpositives\nNormInfinityCone\nNormOneCone\nNormCone\nSecondOrderCone\nRotatedSecondOrderCone\nGeometricMeanCone\nExponentialCone\nDualExponentialCone\nPowerCone\nDualPowerCone\nRelativeEntropyCone\nNormSpectralCone\nNormNuclearCone\nSOS1\nSOS2\nIndicator\nComplements\nHyperRectangle\nScaled","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.Reals","page":"Standard form","title":"MathOptInterface.Reals","text":"Reals(dimension::Int)\n\nThe set mathbbR^dimension (containing all points) of non-negative dimension dimension.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Reals(3))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Reals}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Zeros","page":"Standard form","title":"MathOptInterface.Zeros","text":"Zeros(dimension::Int)\n\nThe set  0 ^dimension (containing only the origin) of non-negative dimension dimension.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Zeros(3))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Zeros}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Nonnegatives","page":"Standard form","title":"MathOptInterface.Nonnegatives","text":"Nonnegatives(dimension::Int)\n\nThe nonnegative orthant  x in mathbbR^dimension  x ge 0  of non-negative dimension dimension.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Nonnegatives(3))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonnegatives}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Nonpositives","page":"Standard form","title":"MathOptInterface.Nonpositives","text":"Nonpositives(dimension::Int)\n\nThe nonpositive orthant  x in mathbbR^dimension  x le 0  of non-negative dimension dimension.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Nonpositives(3))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonpositives}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.NormInfinityCone","page":"Standard form","title":"MathOptInterface.NormInfinityCone","text":"NormInfinityCone(dimension::Int)\n\nThe ell_infty-norm cone  (tx) in mathbbR^dimension  t ge lVert x rVert_infty = max_i lvert x_i rvert  of dimension dimension.\n\nThe dimension must be at least 1.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.NormInfinityCone(4))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormInfinityCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.NormOneCone","page":"Standard form","title":"MathOptInterface.NormOneCone","text":"NormOneCone(dimension::Int)\n\nThe ell_1-norm cone  (tx) in mathbbR^dimension  t ge lVert x rVert_1 = sum_i lvert x_i rvert  of dimension dimension.\n\nThe dimension must be at least 1.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.NormOneCone(4))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormOneCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.NormCone","page":"Standard form","title":"MathOptInterface.NormCone","text":"NormCone(p::Float64, dimension::Int)\n\nThe ell_p-norm cone  (tx) in mathbbR^dimension  t ge left(sumlimits_i x_i^pright)^frac1p  of dimension dimension.\n\nThe dimension must be at least 1.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.NormCone(3, 4))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.SecondOrderCone","page":"Standard form","title":"MathOptInterface.SecondOrderCone","text":"SecondOrderCone(dimension::Int)\n\nThe second-order cone (or Lorenz cone or ell_2-norm cone)  (tx) in mathbbR^dimension  t ge lVert x rVert_2  of dimension dimension.\n\nThe dimension must be at least 1.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([t; x]), MOI.SecondOrderCone(4))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.SecondOrderCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.RotatedSecondOrderCone","page":"Standard form","title":"MathOptInterface.RotatedSecondOrderCone","text":"RotatedSecondOrderCone(dimension::Int)\n\nThe rotated second-order cone  (tux) in mathbbR^dimension  2tu ge lVert x rVert_2^2 tu ge 0  of dimension dimension.\n\nThe dimension must be at least 2.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> u = MOI.add_variable(model)\nMOI.VariableIndex(2)\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; u; x]),\n           MOI.RotatedSecondOrderCone(5),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RotatedSecondOrderCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.GeometricMeanCone","page":"Standard form","title":"MathOptInterface.GeometricMeanCone","text":"GeometricMeanCone(dimension::Int)\n\nThe geometric mean cone  (tx) in mathbbR^n+1  x ge 0 t le sqrtnx_1 x_2 cdots x_n , where dimension = n + 1 >= 2.\n\nDuality note\n\nThe dual of the geometric mean cone is  (u v) in mathbbR^n+1  u le 0 v ge 0 -u le n sqrtnprod_i v_i , where dimension = n + 1 >= 2.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; x]),\n           MOI.GeometricMeanCone(4),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.GeometricMeanCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.ExponentialCone","page":"Standard form","title":"MathOptInterface.ExponentialCone","text":"ExponentialCone()\n\nThe 3-dimensional exponential cone  (xyz) in mathbbR^3  y exp (xy) le z y  0 .\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.ExponentialCone())\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.ExponentialCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.DualExponentialCone","page":"Standard form","title":"MathOptInterface.DualExponentialCone","text":"DualExponentialCone()\n\nThe 3-dimensional dual exponential cone  (uvw) in mathbbR^3  -u exp (vu) le exp(1) w u  0 .\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.DualExponentialCone())\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.DualExponentialCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.PowerCone","page":"Standard form","title":"MathOptInterface.PowerCone","text":"PowerCone{T<:Real}(exponent::T)\n\nThe 3-dimensional power cone  (xyz) in mathbbR^3  x^exponent y^1-exponent ge z x ge 0 y ge 0  with parameter exponent.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.PowerCone(0.5))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.PowerCone{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.DualPowerCone","page":"Standard form","title":"MathOptInterface.DualPowerCone","text":"DualPowerCone{T<:Real}(exponent::T)\n\nThe 3-dimensional power cone  (uvw) in mathbbR^3  (fracuexponent)^exponent (fracv1-exponent)^1-exponent ge w u ge 0 v ge 0  with parameter exponent.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.DualPowerCone(0.5))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.DualPowerCone{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.RelativeEntropyCone","page":"Standard form","title":"MathOptInterface.RelativeEntropyCone","text":"RelativeEntropyCone(dimension::Int)\n\nThe relative entropy cone  (u v w) in mathbbR^1+2n  u ge sum_i=1^n w_i log(fracw_iv_i) v_i ge 0 w_i ge 0 , where dimension = 2n + 1 >= 1.\n\nDuality note\n\nThe dual of the relative entropy cone is  (u v w) in mathbbR^1+2n  forall i w_i ge u (log (fracuv_i) - 1) v_i ge 0 u  0  of dimension dimension=2n+1.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> u = MOI.add_variable(model);\n\njulia> v = MOI.add_variables(model, 3);\n\njulia> w = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([u; v; w]),\n           MOI.RelativeEntropyCone(7),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RelativeEntropyCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.NormSpectralCone","page":"Standard form","title":"MathOptInterface.NormSpectralCone","text":"NormSpectralCone(row_dim::Int, column_dim::Int)\n\nThe epigraph of the matrix spectral norm (maximum singular value function)  (t X) in mathbbR^1 + row_dim times column_dim  t ge sigma_1(X) , where sigma_i is the ith singular value of the matrix X of non-negative row dimension row_dim and column dimension column_dim.\n\nThe matrix X is vectorized by stacking the columns, matching the behavior of Julia's vec function.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> X = reshape(MOI.add_variables(model, 6), 2, 3)\n2×3 Matrix{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(2)  MOI.VariableIndex(4)  MOI.VariableIndex(6)\n MOI.VariableIndex(3)  MOI.VariableIndex(5)  MOI.VariableIndex(7)\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; vec(X)]),\n           MOI.NormSpectralCone(2, 3),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormSpectralCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.NormNuclearCone","page":"Standard form","title":"MathOptInterface.NormNuclearCone","text":"NormNuclearCone(row_dim::Int, column_dim::Int)\n\nThe epigraph of the matrix nuclear norm (sum of singular values function)  (t X) in mathbbR^1 + row_dim times column_dim  t ge sum_i sigma_i(X) , where sigma_i is the ith singular value of the matrix X of non-negative row dimension row_dim and column dimension column_dim.\n\nThe matrix X is vectorized by stacking the columns, matching the behavior of Julia's vec function.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> X = reshape(MOI.add_variables(model, 6), 2, 3)\n2×3 Matrix{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(2)  MOI.VariableIndex(4)  MOI.VariableIndex(6)\n MOI.VariableIndex(3)  MOI.VariableIndex(5)  MOI.VariableIndex(7)\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; vec(X)]),\n           MOI.NormNuclearCone(2, 3),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.NormNuclearCone}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.SOS1","page":"Standard form","title":"MathOptInterface.SOS1","text":"SOS1{T<:Real}(weights::Vector{T})\n\nThe set corresponding to the Special Ordered Set (SOS) constraint of Type I.\n\nOf the variables in the set, at most one can be nonzero.\n\nThe weights induce an ordering of the variables such that the kth element in the set corresponds to the kth weight in weights. Solvers may use these weights to improve the efficiency of the solution process, but the ordering does not change the set of feasible solutions.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables(x),\n           MOI.SOS1([1.0, 3.0, 2.5]),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.SOS1{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.SOS2","page":"Standard form","title":"MathOptInterface.SOS2","text":"SOS2{T<:Real}(weights::Vector{T})\n\nThe set corresponding to the Special Ordered Set (SOS) constraint of Type II.\n\nThe weights induce an ordering of the variables such that the kth element in the set corresponds to the kth weight in weights. Therefore, the weights must be unique.\n\nOf the variables in the set, at most two can be nonzero, and if two are nonzero, they must be adjacent in the ordering of the set.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables(x),\n           MOI.SOS2([1.0, 3.0, 2.5]),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.SOS2{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Indicator","page":"Standard form","title":"MathOptInterface.Indicator","text":"Indicator{A<:ActivationCondition,S<:AbstractScalarSet}(set::S)\n\nThe set corresponding to an indicator constraint.\n\nWhen A is ACTIVATE_ON_ZERO, this means: (y x) in 0 1 times mathbbR^n  y = 0 implies x in set\n\nWhen A is ACTIVATE_ON_ONE, this means: (y x) in 0 1 times mathbbR^n  y = 1 implies x in set\n\nNotes\n\nMost solvers expect that the first row of the function is interpretable as a variable index x_i (e.g., 1.0 * x + 0.0). An error will be thrown if this is not the case.\n\nExample\n\nThe constraint (y x) in 0 1 times mathbbR^2  y = 1 implies x_1 + x_2 leq 9  is defined as\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 2)\n2-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n\njulia> y, _ = MOI.add_constrained_variable(model, MOI.ZeroOne())\n(MOI.VariableIndex(3), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(3))\n\njulia> f = MOI.VectorAffineFunction(\n           [\n               MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, y)),\n               MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(1.0, x[1])),\n               MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(1.0, x[2])),\n           ],\n           [0.0, 0.0],\n       )\n┌                                                         ┐\n│0.0 + 1.0 MOI.VariableIndex(3)                           │\n│0.0 + 1.0 MOI.VariableIndex(1) + 1.0 MOI.VariableIndex(2)│\n└                                                         ┘\n\njulia> s = MOI.Indicator{MOI.ACTIVATE_ON_ONE}(MOI.LessThan(9.0))\nMathOptInterface.Indicator{MathOptInterface.ACTIVATE_ON_ONE, MathOptInterface.LessThan{Float64}}(MathOptInterface.LessThan{Float64}(9.0))\n\njulia> MOI.add_constraint(model, f, s)\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Indicator{MathOptInterface.ACTIVATE_ON_ONE, MathOptInterface.LessThan{Float64}}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Complements","page":"Standard form","title":"MathOptInterface.Complements","text":"Complements(dimension::Base.Integer)\n\nThe set corresponding to a mixed complementarity constraint.\n\nComplementarity constraints should be specified with an AbstractVectorFunction-in-Complements(dimension) constraint.\n\nThe dimension of the vector-valued function F must be dimension. This defines a complementarity constraint between the scalar function F[i] and the variable in F[i + dimension/2]. Thus, F[i + dimension/2] must be interpretable as a single variable x_i (e.g., 1.0 * x + 0.0), and dimension must be even.\n\nThe mixed complementarity problem consists of finding x_i in the interval [lb, ub] (i.e., in the set Interval(lb, ub)), such that the following holds:\n\nF_i(x) == 0 if lb_i < x_i < ub_i\nF_i(x) >= 0 if lb_i == x_i\nF_i(x) <= 0 if x_i == ub_i\n\nClassically, the bounding set for x_i is Interval(0, Inf), which recovers: 0 <= F_i(x) ⟂ x_i >= 0, where the ⟂ operator implies F_i(x) * x_i = 0.\n\nExample\n\nThe problem:\n\nx -in- Interval(-1, 1)\n[-4 * x - 3, x] -in- Complements(2)\n\ndefines the mixed complementarity problem where the following holds:\n\n-4 * x - 3 == 0 if -1 < x < 1\n-4 * x - 3 >= 0 if x == -1\n-4 * x - 3 <= 0 if x == 1\n\nThere are three solutions:\n\nx = -3/4 with F(x) = 0\nx = -1 with F(x) = 1\nx = 1 with F(x) = -7\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x, _ = MOI.add_constrained_variable(model, MOI.Interval(-1.0, 1.0));\n\njulia> MOI.add_constraint(\n            model,\n            MOI.Utilities.vectorize([-4.0 * x - 3.0, x]),\n            MOI.Complements(2),\n        )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Complements}(1)\n\nThe function F can also be defined in terms of single variables. For example, the problem:\n\n[x_3, x_4] -in- Nonnegatives(2)\n[x_1, x_2, x_3, x_4] -in- Complements(4)\n\ndefines the complementarity problem where 0 <= x_1 ⟂ x_3 >= 0 and 0 <= x_2 ⟂ x_4 >= 0.\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 4);\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x[3:4]), MOI.Nonnegatives(2))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Nonnegatives}(1)\n\njulia> MOI.add_constraint(\n            model,\n            MOI.VectorOfVariables(x),\n            MOI.Complements(4),\n        )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Complements}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.HyperRectangle","page":"Standard form","title":"MathOptInterface.HyperRectangle","text":"HyperRectangle(lower::Vector{T}, upper::Vector{T}) where {T}\n\nThe set x in barmathbbR^d x_i in lower_i upper_i forall i=1ldotsd.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3)\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables(x),\n           MOI.HyperRectangle(zeros(3), ones(3)),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.HyperRectangle{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Scaled","page":"Standard form","title":"MathOptInterface.Scaled","text":"struct Scaled{S<:AbstractVectorSet} <: AbstractVectorSet\n    set::S\nend\n\nGiven a vector a in mathbbR^d and a set representing the set mathcalS in mathbbR^d such that Utilities.set_dot for x in mathcalS and y in mathcalS^* is\n\nsum_i=1^d a_i x_i y_i\n\nthe set Scaled(set) is defined as\n\n (sqrta_1 x_1 sqrta_2 x_2 ldots sqrta_d x_d)  x in S \n\nExample\n\nThis can be used to scale a vector of numbers\n\njulia> import MathOptInterface as MOI\n\njulia> set = MOI.PositiveSemidefiniteConeTriangle(2)\nMathOptInterface.PositiveSemidefiniteConeTriangle(2)\n\njulia> a = MOI.Utilities.SetDotScalingVector{Float64}(set)\n3-element MathOptInterface.Utilities.SetDotScalingVector{Float64, MathOptInterface.PositiveSemidefiniteConeTriangle}:\n 1.0\n 1.4142135623730951\n 1.0\n\njulia> using LinearAlgebra\n\njulia> MOI.Utilities.operate(*, Float64, Diagonal(a), ones(3))\n3-element Vector{Float64}:\n 1.0\n 1.4142135623730951\n 1.0\n\nIt can be also used to scale a vector of function\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3);\n\njulia> func = MOI.VectorOfVariables(x)\n┌                    ┐\n│MOI.VariableIndex(1)│\n│MOI.VariableIndex(2)│\n│MOI.VariableIndex(3)│\n└                    ┘\n\njulia> set = MOI.PositiveSemidefiniteConeTriangle(2)\nMathOptInterface.PositiveSemidefiniteConeTriangle(2)\n\njulia> MOI.Utilities.operate(*, Float64, Diagonal(a), func)\n┌                                             ┐\n│0.0 + 1.0 MOI.VariableIndex(1)               │\n│0.0 + 1.4142135623730951 MOI.VariableIndex(2)│\n│0.0 + 1.0 MOI.VariableIndex(3)               │\n└                                             ┘\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#Constraint-programming-sets","page":"Standard form","title":"Constraint programming sets","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"AllDifferent\nBinPacking\nCircuit\nCountAtLeast\nCountBelongs\nCountDistinct\nCountGreaterThan\nCumulative\nPath\nReified\nTable","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.AllDifferent","page":"Standard form","title":"MathOptInterface.AllDifferent","text":"AllDifferent(dimension::Int)\n\nThe set x in mathbbZ^d such that no two elements in x take the same value and dimension = d.\n\nAlso known as\n\nThis constraint is called all_different in MiniZinc, and is sometimes also called distinct.\n\nExample\n\nTo enforce x[1] != x[2] AND x[1] != x[3] AND x[2] != x[3]:\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.AllDifferent(3))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.AllDifferent}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.BinPacking","page":"Standard form","title":"MathOptInterface.BinPacking","text":"BinPacking(c::T, w::Vector{T}) where {T}\n\nThe set x in mathbbZ^d where d = length(w), such that each item i in 1:d of weight w[i] is put into bin x[i], and the total weight of each bin does not exceed c.\n\nThere are additional assumptions that the capacity, c, and the weights, w, must all be non-negative.\n\nThe bin numbers depend on the bounds of x, so they may be something other than the integers 1:d.\n\nAlso known as\n\nThis constraint is called bin_packing in MiniZinc.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> bins = MOI.add_variables(model, 5)\n5-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n MOI.VariableIndex(4)\n MOI.VariableIndex(5)\n\njulia> weights = Float64[1, 1, 2, 2, 3]\n5-element Vector{Float64}:\n 1.0\n 1.0\n 2.0\n 2.0\n 3.0\n\njulia> MOI.add_constraint.(model, bins, MOI.Integer())\n5-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}}:\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(3)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(4)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(5)\n\njulia> MOI.add_constraint.(model, bins, MOI.Interval(4.0, 6.0))\n5-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}}:\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(1)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(2)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(3)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(4)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}}(5)\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(bins), MOI.BinPacking(3.0, weights))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.BinPacking{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Circuit","page":"Standard form","title":"MathOptInterface.Circuit","text":"Circuit(dimension::Int)\n\nThe set x in 1d^d that constraints x to be a circuit, such that x_i = j means that j is the successor of i, and dimension = d.\n\nGraphs with multiple independent circuits, such as [2, 1, 3] and [2, 1, 4, 3], are not valid.\n\nAlso known as\n\nThis constraint is called circuit in MiniZinc, and it is equivalent to forming a (potentially sub-optimal) tour in the travelling salesperson problem.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Circuit(3))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Circuit}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.CountAtLeast","page":"Standard form","title":"MathOptInterface.CountAtLeast","text":"CountAtLeast(n::Int, d::Vector{Int}, set::Set{Int})\n\nThe set x in mathbbZ^d_1 + d_2 + ldots d_N, where x is partitioned into N subsets (x_1  ldots x_d_1, x_d_1 + 1  ldots x_d_1 + d_2 and so on), and at least n elements of each subset take one of the values in set.\n\nAlso known as\n\nThis constraint is called at_least in MiniZinc.\n\nExample\n\nTo ensure that 3 appears at least once in each of the subsets {a, b} and {b, c}:\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> a, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))\n\njulia> b, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(2), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2))\n\njulia> c, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(3), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(3))\n\njulia> x, d, set = [a, b, b, c], [2, 2], [3]\n(MathOptInterface.VariableIndex[MOI.VariableIndex(1), MOI.VariableIndex(2), MOI.VariableIndex(2), MOI.VariableIndex(3)], [2, 2], [3])\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.CountAtLeast(1, d, Set(set)))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountAtLeast}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.CountBelongs","page":"Standard form","title":"MathOptInterface.CountBelongs","text":"CountBelongs(dimenson::Int, set::Set{Int})\n\nThe set (n x) in mathbbZ^1+d, such that n elements of the vector x take on of the values in set and dimension = 1 + d.\n\nAlso known as\n\nThis constraint is called among by MiniZinc.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> n, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))\n\njulia> x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n MOI.VariableIndex(4)\n\njulia> set = Set([3, 4, 5])\nSet{Int64} with 3 elements:\n  5\n  4\n  3\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([n; x]), MOI.CountBelongs(4, set))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountBelongs}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.CountDistinct","page":"Standard form","title":"MathOptInterface.CountDistinct","text":"CountDistinct(dimension::Int)\n\nThe set (n x) in mathbbZ^1+d, such that the number of distinct values in x is n and dimension = 1 + d.\n\nAlso known as\n\nThis constraint is called nvalues in MiniZinc.\n\nExample\n\nTo model:\n\nif n == 1`, thenx[1] == x[2] == x[3]`\nif n == 2, then\nx[1] == x[2] != x[3] or\nx[1] != x[2] == x[3] or\nx[1] == x[3] != x[2]\nif n == 3, then x[1] != x[2], x[2] != x[3] and x[3] != x[1]\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> n, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))\n\njulia> x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n MOI.VariableIndex(4)\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(vcat(n, x)), MOI.CountDistinct(4))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountDistinct}(1)\n\nRelationship to AllDifferent\n\nWhen the first element is d, CountDistinct is equivalent to an AllDifferent constraint.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.CountGreaterThan","page":"Standard form","title":"MathOptInterface.CountGreaterThan","text":"CountGreaterThan(dimension::Int)\n\nThe set (c y x) in mathbbZ^1+1+d, such that c is strictly greater than the number of occurances of y in x and dimension = 1 + 1 + d.\n\nAlso known as\n\nThis constraint is called count_gt in MiniZinc.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> c, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))\n\njulia> y, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(2), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2))\n\njulia> x = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(3)\n MOI.VariableIndex(4)\n MOI.VariableIndex(5)\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([c; y; x]), MOI.CountGreaterThan(5))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.CountGreaterThan}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Cumulative","page":"Standard form","title":"MathOptInterface.Cumulative","text":"Cumulative(dimension::Int)\n\nThe set (s d r b) in mathbbZ^3n+1, representing the cumulative global constraint, where n == length(s) == length(r) == length(b) and dimension = 3n + 1.\n\nCumulative requires that a set of tasks given by start times s, durations d, and resource requirements r, never requires more than the global resource bound b at any one time.\n\nAlso known as\n\nThis constraint is called cumulative in MiniZinc.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> s = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n\njulia> d = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(4)\n MOI.VariableIndex(5)\n MOI.VariableIndex(6)\n\njulia> r = [MOI.add_constrained_variable(model, MOI.Integer())[1] for _ in 1:3]\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(7)\n MOI.VariableIndex(8)\n MOI.VariableIndex(9)\n\njulia> b, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(10), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(10))\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([s; d; r; b]), MOI.Cumulative(10))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Cumulative}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Path","page":"Standard form","title":"MathOptInterface.Path","text":"Path(from::Vector{Int}, to::Vector{Int})\n\nGiven a graph comprised of a set of nodes 1..N and a set of arcs 1..E represented by an edge from node from[i] to node to[i], Path constrains the set (s t ns es) in (1N)times(1E)times01^Ntimes01^E, to form subgraph that is a path from node s to node t, where node n is in the path if ns[n] is 1, and edge e is in the path if es[e] is 1.\n\nThe path must be acyclic, and it must traverse all nodes n for which ns[n] is 1, and all edges e for which es[e] is 1.\n\nAlso known as\n\nThis constraint is called path in MiniZinc.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> N, E = 4, 5\n(4, 5)\n\njulia> from = [1, 1, 2, 2, 3]\n5-element Vector{Int64}:\n 1\n 1\n 2\n 2\n 3\n\njulia> to = [2, 3, 3, 4, 4]\n5-element Vector{Int64}:\n 2\n 3\n 3\n 4\n 4\n\njulia> s, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(1))\n\njulia> t, _ = MOI.add_constrained_variable(model, MOI.Integer())\n(MOI.VariableIndex(2), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.Integer}(2))\n\njulia> ns = MOI.add_variables(model, N)\n4-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(3)\n MOI.VariableIndex(4)\n MOI.VariableIndex(5)\n MOI.VariableIndex(6)\n\njulia> MOI.add_constraint.(model, ns, MOI.ZeroOne())\n4-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}}:\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(3)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(4)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(5)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(6)\n\njulia> es = MOI.add_variables(model, E)\n5-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(7)\n MOI.VariableIndex(8)\n MOI.VariableIndex(9)\n MOI.VariableIndex(10)\n MOI.VariableIndex(11)\n\njulia> MOI.add_constraint.(model, es, MOI.ZeroOne())\n5-element Vector{MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}}:\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(7)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(8)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(9)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(10)\n MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(11)\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables([s; t; ns; es]), MOI.Path(from, to))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Path}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Reified","page":"Standard form","title":"MathOptInterface.Reified","text":"Reified(set::AbstractSet)\n\nThe constraint z f(x) in Reified(S) ensures that f(x) in S if and only if z == 1, where z in 0 1.\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())\nMOIU.UniversalFallback{MOIU.Model{Float64}}\nfallback for MOIU.Model{Float64}\n\njulia> z, _ = MOI.add_constrained_variable(model, MOI.ZeroOne())\n(MOI.VariableIndex(1), MathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(1))\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(2)\n\njulia> MOI.add_constraint(\n           model,\n           MOI.Utilities.vectorize([z, 2.0 * x]),\n           MOI.Reified(MOI.GreaterThan(1.0)),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Reified{MathOptInterface.GreaterThan{Float64}}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.Table","page":"Standard form","title":"MathOptInterface.Table","text":"Table(table::Matrix{T}) where {T}\n\nThe set x in mathbbR^d where d = size(table, 2), such that x belongs to one row of table. That is, there exists some j in 1:size(table, 1), such that x[i] = table[j, i] for all i=1:size(table, 2).\n\nAlso known as\n\nThis constraint is called table in MiniZinc.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> x = MOI.add_variables(model, 3)\n3-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n MOI.VariableIndex(2)\n MOI.VariableIndex(3)\n\njulia> table = Float64[1 1 0; 0 1 1; 1 0 1; 1 1 1]\n4×3 Matrix{Float64}:\n 1.0  1.0  0.0\n 0.0  1.0  1.0\n 1.0  0.0  1.0\n 1.0  1.0  1.0\n\njulia> MOI.add_constraint(model, MOI.VectorOfVariables(x), MOI.Table(table))\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.Table{Float64}}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#Matrix-sets","page":"Standard form","title":"Matrix sets","text":"","category":"section"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"Matrix sets are vectorized to be subtypes of AbstractVectorSet.","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"For sets of symmetric matrices, storing both the (i, j) and (j, i) elements is redundant. Use the AbstractSymmetricMatrixSetTriangle set to represent only the vectorization of the upper triangular part of the matrix.","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"When the matrix of expressions constrained to be in the set is not symmetric, and hence additional constraints are needed to force the equality of the (i, j) and (j, i) elements, use the AbstractSymmetricMatrixSetSquare set.","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"The Bridges.Constraint.SquareBridge can transform a set from the square form to the triangular_form by adding appropriate constraints if the (i, j) and (j, i) expressions are different.","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"AbstractSymmetricMatrixSetTriangle\nAbstractSymmetricMatrixSetSquare\nside_dimension\ntriangular_form","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractSymmetricMatrixSetTriangle","page":"Standard form","title":"MathOptInterface.AbstractSymmetricMatrixSetTriangle","text":"abstract type AbstractSymmetricMatrixSetTriangle <: AbstractVectorSet end\n\nAbstract supertype for subsets of the (vectorized) cone of symmetric matrices, with side_dimension rows and columns. The entries of the upper-right triangular part of the matrix are given column by column (or equivalently, the entries of the lower-left triangular part are given row by row). A vectorized cone of dimension n corresponds to a square matrix with side dimension sqrt14 + 2 n - 12. (Because a d times d matrix has d(d + 1)  2 elements in the upper or lower triangle.)\n\nExample\n\nThe matrix\n\nbeginbmatrix\n  1  2  4\n  2  3  5\n  4  5  6\nendbmatrix\n\nhas side_dimension 3 and vectorization (1 2 3 4 5 6).\n\nNote\n\nTwo packed storage formats exist for symmetric matrices, the respective orders of the entries are:\n\nupper triangular column by column (or lower triangular row by row);\nlower triangular column by column (or upper triangular row by row).\n\nThe advantage of the first format is the mapping between the (i, j) matrix indices and the k index of the vectorized form. It is simpler and does not depend on the side dimension of the matrix. Indeed,\n\nthe entry of matrix indices (i, j) has vectorized index k = div((j - 1) * j, 2) + i if i leq j and k = div((i - 1) * i, 2) + j if j leq i;\nand the entry with vectorized index k has matrix indices i = div(1 + isqrt(8k - 7), 2) and j = k - div((i - 1) * i, 2) or j = div(1 + isqrt(8k - 7), 2) and i = k - div((j - 1) * j, 2).\n\nDuality note\n\nThe scalar product for the symmetric matrix in its vectorized form is the sum of the pairwise product of the diagonal entries plus twice the sum of the pairwise product of the upper diagonal entries; see [p. 634, 1]. This has important consequence for duality.\n\nConsider for example the following problem (PositiveSemidefiniteConeTriangle is a subtype of AbstractSymmetricMatrixSetTriangle)\n\nbeginalign*\n     max_x in mathbbR  x\n    \n     textst \n    (1 -x 1)  in textPositiveSemidefiniteConeTriangle(2)\nendalign*\n\nThe dual is the following problem\n\nbeginalign*\n     min_x in mathbbR^3  y_1 + y_3\n    \n     textst  2y_2  = 1\n      y  in textPositiveSemidefiniteConeTriangle(2)\nendalign*\n\nWhy do we use 2y_2 in the dual constraint instead of y_2 ? The reason is that 2y_2 is the scalar product between y and the symmetric matrix whose vectorized form is (0 1 0). Indeed, with our modified scalar products we have\n\nlangle\n(0 1 0)\n(y_1 y_2 y_3)\nrangle\n=\nmathrmtrace\nbeginpmatrix\n  0  1\n  1  0\nendpmatrix\nbeginpmatrix\n  y_1  y_2\n  y_2  y_3\nendpmatrix\n= 2y_2\n\nReferences\n\n[1] Boyd, S. and Vandenberghe, L.. Convex optimization. Cambridge university     press, 2004.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.AbstractSymmetricMatrixSetSquare","page":"Standard form","title":"MathOptInterface.AbstractSymmetricMatrixSetSquare","text":"abstract type AbstractSymmetricMatrixSetSquare <: AbstractVectorSet end\n\nAbstract supertype for subsets of the (vectorized) cone of symmetric matrices, with side_dimension rows and columns. The entries of the matrix are given column by column (or equivalently, row by row). The matrix is both constrained to be symmetric and to have its triangular_form belong to the corresponding set. That is, if the functions in entries (i j) and (j i) are different, then a constraint will be added to make sure that the entries are equal.\n\nExample\n\nPositiveSemidefiniteConeSquare is a subtype of AbstractSymmetricMatrixSetSquare and constraining the matrix\n\nbeginbmatrix\n  1  -y\n  -z  0\nendbmatrix\n\nto be symmetric positive semidefinite can be achieved by constraining the vector (1 -z -y 0) (or (1 -y -z 0)) to belong to the PositiveSemidefiniteConeSquare(2). It both constrains y = z and (1 -y 0) (or (1 -z 0)) to be in PositiveSemidefiniteConeTriangle(2), since triangular_form(PositiveSemidefiniteConeSquare) is PositiveSemidefiniteConeTriangle.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.side_dimension","page":"Standard form","title":"MathOptInterface.side_dimension","text":"side_dimension(\n    set::Union{\n        AbstractSymmetricMatrixSetTriangle,\n        AbstractSymmetricMatrixSetSquare,\n        HermitianPositiveSemidefiniteConeTriangle,\n    },\n)\n\nSide dimension of the matrices in set.\n\nConvention\n\nBy convention, the side dimension should be stored in the side_dimension field. If this is not the case for a subtype of AbstractSymmetricMatrixSetTriangle, or AbstractSymmetricMatrixSetSquare you must implement this method.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/#MathOptInterface.triangular_form","page":"Standard form","title":"MathOptInterface.triangular_form","text":"triangular_form(S::Type{<:AbstractSymmetricMatrixSetSquare})\ntriangular_form(set::AbstractSymmetricMatrixSetSquare)\n\nReturn the AbstractSymmetricMatrixSetTriangle corresponding to the vectorization of the upper triangular part of matrices in the AbstractSymmetricMatrixSetSquare set.\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"List of recognized matrix sets.","category":"page"},{"location":"moi/reference/standard_form/","page":"Standard form","title":"Standard form","text":"PositiveSemidefiniteConeTriangle\nPositiveSemidefiniteConeSquare\nHermitianPositiveSemidefiniteConeTriangle\nLogDetConeTriangle\nLogDetConeSquare\nRootDetConeTriangle\nRootDetConeSquare","category":"page"},{"location":"moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle","page":"Standard form","title":"MathOptInterface.PositiveSemidefiniteConeTriangle","text":"PositiveSemidefiniteConeTriangle(side_dimension::Int) <: AbstractSymmetricMatrixSetTriangle\n\nThe (vectorized) cone of symmetric positive semidefinite matrices, with non-negative side_dimension rows and columns.\n\nSee AbstractSymmetricMatrixSetTriangle for more details on the vectorized form.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeSquare","page":"Standard form","title":"MathOptInterface.PositiveSemidefiniteConeSquare","text":"PositiveSemidefiniteConeSquare(side_dimension::Int) <: AbstractSymmetricMatrixSetSquare\n\nThe cone of symmetric positive semidefinite matrices, with non-negative side length side_dimension.\n\nSee AbstractSymmetricMatrixSetSquare for more details on the vectorized form.\n\nThe entries of the matrix are given column by column (or equivalently, row by row).\n\nThe matrix is both constrained to be symmetric and to be positive semidefinite. That is, if the functions in entries (i j) and (j i) are different, then a constraint will be added to make sure that the entries are equal.\n\nExample\n\nConstraining the matrix\n\nbeginbmatrix\n  1  -y\n  -z  0\nendbmatrix\n\nto be symmetric positive semidefinite can be achieved by constraining the vector (1 -z -y 0) (or (1 -y -z 0)) to belong to the PositiveSemidefiniteConeSquare(2).\n\nIt both constrains y = z and (1 -y 0) (or (1 -z 0)) to be in PositiveSemidefiniteConeTriangle(2).\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.HermitianPositiveSemidefiniteConeTriangle","page":"Standard form","title":"MathOptInterface.HermitianPositiveSemidefiniteConeTriangle","text":"HermitianPositiveSemidefiniteConeTriangle(side_dimension::Int) <: AbstractVectorSet\n\nThe (vectorized) cone of Hermitian positive semidefinite matrices, with non-negative side_dimension rows and columns.\n\nBecaue the matrix is Hermitian, the diagonal elements are real, and the complex-valued lower triangular entries are obtained as the conjugate of corresponding upper triangular entries.\n\nVectorization format\n\nThe vectorized form starts with real part of the entries of the upper triangular part of the matrix, given column by column as explained in AbstractSymmetricMatrixSetSquare.\n\nIt is then followed by the imaginary part of the off-diagonal entries of the upper triangular part, also given column by column.\n\nFor example, the matrix\n\nbeginbmatrix\n  1  2 + 7im  4 + 8im\n  2 - 7im  3  5 + 9im\n  4 - 8im  5 - 9im  6\nendbmatrix\n\nhas side_dimension 3 and is represented as the vector 1 2 3 4 5 6 7 8 9.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.LogDetConeTriangle","page":"Standard form","title":"MathOptInterface.LogDetConeTriangle","text":"LogDetConeTriangle(side_dimension::Int)\n\nThe log-determinant cone  (t u X) in mathbbR^2 + d(d+1)2  t le u log(det(Xu)) u  0 , where the matrix X is represented in the same symmetric packed format as in the PositiveSemidefiniteConeTriangle.\n\nThe non-negative argument side_dimension is the side dimension of the matrix X, i.e., its number of rows or columns.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> X = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; X]),\n           MOI.LogDetConeTriangle(2),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.LogDetConeTriangle}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.LogDetConeSquare","page":"Standard form","title":"MathOptInterface.LogDetConeSquare","text":"LogDetConeSquare(side_dimension::Int)\n\nThe log-determinant cone  (t u X) in mathbbR^2 + d^2  t le u log(det(Xu)) X text symmetric u  0 , where the matrix X is represented in the same format as in the PositiveSemidefiniteConeSquare.\n\nSimilarly to PositiveSemidefiniteConeSquare, constraints are added to ensure that X is symmetric.\n\nThe non-negative argument side_dimension is the side dimension of the matrix X, i.e., its number of rows or columns.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> X = reshape(MOI.add_variables(model, 4), 2, 2)\n2×2 Matrix{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(2)  MOI.VariableIndex(4)\n MOI.VariableIndex(3)  MOI.VariableIndex(5)\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; vec(X)]),\n           MOI.LogDetConeSquare(2),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.LogDetConeSquare}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.RootDetConeTriangle","page":"Standard form","title":"MathOptInterface.RootDetConeTriangle","text":"RootDetConeTriangle(side_dimension::Int)\n\nThe root-determinant cone  (t X) in mathbbR^1 + d(d+1)2  t le det(X)^1d , where the matrix X is represented in the same symmetric packed format as in the PositiveSemidefiniteConeTriangle.\n\nThe non-negative argument side_dimension is the side dimension of the matrix X, i.e., its number of rows or columns.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> X = MOI.add_variables(model, 3);\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; X]),\n           MOI.RootDetConeTriangle(2),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RootDetConeTriangle}(1)\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/standard_form/#MathOptInterface.RootDetConeSquare","page":"Standard form","title":"MathOptInterface.RootDetConeSquare","text":"RootDetConeSquare(side_dimension::Int)\n\nThe root-determinant cone  (t X) in mathbbR^1 + d^2  t le det(X)^1d X text symmetric , where the matrix X is represented in the same format as PositiveSemidefiniteConeSquare.\n\nSimilarly to PositiveSemidefiniteConeSquare, constraints are added to ensure that X is symmetric.\n\nThe non-negative argument side_dimension is the side dimension of the matrix X, i.e., its number of rows or columns.\n\nExample\n\njulia> import MathOptInterface as MOI\n\njulia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> t = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> X = reshape(MOI.add_variables(model, 4), 2, 2)\n2×2 Matrix{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(2)  MOI.VariableIndex(4)\n MOI.VariableIndex(3)  MOI.VariableIndex(5)\n\njulia> MOI.add_constraint(\n           model,\n           MOI.VectorOfVariables([t; vec(X)]),\n           MOI.RootDetConeSquare(2),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorOfVariables, MathOptInterface.RootDetConeSquare}(1)\n\n\n\n\n\n","category":"type"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"EditURL = \"https://github.com/darnstrom/DAQP.jl/blob/v0.5.1/README.md\"","category":"page"},{"location":"packages/DAQP/#DAQP.jl","page":"darnstrom/DAQP.jl","title":"DAQP.jl","text":"","category":"section"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"(Image: ) (Image: )","category":"page"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"DAQP.jl is a Julia wrapper for the Quadratic Programming solver DAQP.","category":"page"},{"location":"packages/DAQP/#License","page":"darnstrom/DAQP.jl","title":"License","text":"","category":"section"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"DAQP.jl is licensed under the MIT license.","category":"page"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"The underlying solver, darnstrom/daqp is licensed under the MIT license.","category":"page"},{"location":"packages/DAQP/#Installation","page":"darnstrom/DAQP.jl","title":"Installation","text":"","category":"section"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"Install DAQP.jl using the Julia package manager:","category":"page"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"import Pkg\nPkg.add(\"DAQP\")","category":"page"},{"location":"packages/DAQP/#Use-with-JuMP","page":"darnstrom/DAQP.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"To use DAQP with JuMP, do:","category":"page"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"using JuMP, DAQP\nmodel = Model(DAQP.Optimizer)","category":"page"},{"location":"packages/DAQP/#Documentation","page":"darnstrom/DAQP.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/DAQP/","page":"darnstrom/DAQP.jl","title":"darnstrom/DAQP.jl","text":"General information about the solver is available at https://darnstrom.github.io/daqp/, and specifics for the Julia interface are available at https://darnstrom.github.io/daqp/start/julia.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"EditURL = \"portfolio.jl\"","category":"page"},{"location":"tutorials/nonlinear/portfolio/#Portfolio-optimization","page":"Portfolio optimization","title":"Portfolio optimization","text":"","category":"section"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"This tutorial was originally contributed by Arpit Bhatia.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Optimization models play an increasingly important role in financial decisions. Many computational finance problems can be solved efficiently using modern optimization techniques.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"This tutorial solves the famous Markowitz Portfolio Optimization problem with data from lecture notes from a course taught at Georgia Tech by Shabbir Ahmed.","category":"page"},{"location":"tutorials/nonlinear/portfolio/#Required-packages","page":"Portfolio optimization","title":"Required packages","text":"","category":"section"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"This tutorial uses the following packages","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"using JuMP\nimport DataFrames\nimport Ipopt\nimport MultiObjectiveAlgorithms as MOA\nimport Plots\nimport Statistics\nimport StatsPlots","category":"page"},{"location":"tutorials/nonlinear/portfolio/#Formulation","page":"Portfolio optimization","title":"Formulation","text":"","category":"section"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Suppose we are considering investing 1000 dollars in three non-dividend paying stocks, IBM (IBM), Walmart (WMT), and Southern Electric (SEHI), for a one-month period.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"We will use the initial money to buy shares of the three stocks at the current market prices, hold these for one month, and sell the shares off at the prevailing market prices at the end of the month.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"As a rational investor, we hope to make some profit out of this endeavor, that is, the return on our investment should be positive.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Suppose we bought a stock at p dollars per share in the beginning of the month, and sold it off at s dollars per share at the end of the month. Then the one-month return on a share of the stock is fracs-pp.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Since the stock prices are quite uncertain, so is the end-of-month return on our investment. Our goal is to invest in such a way that the expected end-of-month return is at least $50 or 5%. Furthermore, we want to make sure that the “risk” of not achieving our desired return is minimum.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Note that we are solving the problem under the following assumptions:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"We can trade any continuum of shares.\nNo short-selling is allowed.\nThere are no transaction costs.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"We model this problem by taking decision variables x_i i=123 denoting the dollars invested in each of the 3 stocks.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Let us denote by tilder_i the random variable corresponding to the monthly return (increase in the stock price) per dollar for stock i.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Then, the return (or profit) on x_i dollars invested in stock i is tilder_i x_i and the total (random) return on our investment is sum_i=1^3 tilder_i x_i The expected return on our investment is then mathbbEleftsum_i=1^3 tilder_i x_iright=sum_i=1^3 overliner_i x_i where overliner_i is the expected value of the tilder_i","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Now we need to quantify the notion of “risk” in our investment.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Markowitz, in his Nobel prize winning work, showed that a rational investor’s notion of minimizing risk can be closely approximated by minimizing the variance of the return of the investment portfolio. This variance is given by:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"operatornameVarleftsum_i=1^3 tilder_i x_iright = sum_i=1^3 sum_j=1^3 x_i x_j sigma_i j","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"where sigma_i j is the covariance of the return of stock i with stock j.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Note that the right hand side of the equation is the most reduced form of the expression and we have not shown the intermediate steps involved in getting to this form. We can also write this equation as:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"operatornameVarleftsum_i=1^3 tilder_i x_iright =x^top Q x","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Where Q is the covariance matrix for the random vector tilder.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Finally, we can write the model as:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"beginaligned\nmin x^top Q x \ntext  st  sum_i=1^3 x_i leq 1000 \noverliner^top x geq 50 \nx geq 0\nendaligned","category":"page"},{"location":"tutorials/nonlinear/portfolio/#Data","page":"Portfolio optimization","title":"Data","text":"","category":"section"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"For the data in our problem, we use the stock prices given below, in monthly values from November 2000, through November 2001.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"df = DataFrames.DataFrame(\n    [\n        93.043 51.826 1.063\n        84.585 52.823 0.938\n        111.453 56.477 1.000\n        99.525 49.805 0.938\n        95.819 50.287 1.438\n        114.708 51.521 1.700\n        111.515 51.531 2.540\n        113.211 48.664 2.390\n        104.942 55.744 3.120\n        99.827 47.916 2.980\n        91.607 49.438 1.900\n        107.937 51.336 1.750\n        115.590 55.081 1.800\n    ],\n    [:IBM, :WMT, :SEHI],\n)","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Next, we compute the percentage return for the stock in each month:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"returns = diff(Matrix(df); dims = 1) ./ Matrix(df[1:end-1, :])","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"The expected monthly return is:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"r = vec(Statistics.mean(returns; dims = 1))","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"and the covariance matrix is:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Q = Statistics.cov(returns)","category":"page"},{"location":"tutorials/nonlinear/portfolio/#JuMP-formulation","page":"Portfolio optimization","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"model = Model(Ipopt.Optimizer)\nset_silent(model)\n@variable(model, x[1:3] >= 0)\n@objective(model, Min, x' * Q * x)\n@constraint(model, sum(x) <= 1000)\n@constraint(model, r' * x >= 50)\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"The optimal allocation of our assets is:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"value.(x)","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"So we spend $497 on IBM, and $503 on SEHI. This results in a variance of:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"scalar_variance = value(x' * Q * x)","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"and an expected return of:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"scalar_return = value(r' * x)","category":"page"},{"location":"tutorials/nonlinear/portfolio/#Multi-objective-portfolio-optimization","page":"Portfolio optimization","title":"Multi-objective portfolio optimization","text":"","category":"section"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"The previous model returned a single solution that minimized the variance, ensuring that our expected return was at least $50. In practice, we might be willing to accept a slightly higher variance if it meant a much increased expected return. To explore this problem space, we can instead formulate our portfolio optimization problem with two objectives:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"to minimize the variance\nto maximize the expected return","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"The solution to this bi-objective problem is the efficient frontier of modern portfolio theory, and each point in the solution is a point with the best return for a fixed level of risk.","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"model = Model(() -> MOA.Optimizer(Ipopt.Optimizer))\nset_silent(model)","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"We also need to choose a solution algorithm for MOA. For our problem, the efficient frontier will have an infinite number of solutions. Since we cannot find all of the solutions, we choose an approximation algorithm and limit the number of solution points that are returned:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"set_optimizer_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())\nset_optimizer_attribute(model, MOA.SolutionLimit(), 50)","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Now we can define the rest of the model:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"@variable(model, x[1:3] >= 0)\n@constraint(model, sum(x) <= 1000)\n@expression(model, variance, x' * Q * x)\n@expression(model, expected_return, r' * x)\n# We want to minimize variance and maximize expected return, but we must pick\n# a single objective sense `Min`, and negate any `Max` objectives:\n@objective(model, Min, [variance, -expected_return])\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"The algorithm found 50 different solutions. Let's plot them to see how they differ:","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"objective_space = Plots.hline(\n    [scalar_return];\n    label = \"Single-objective solution\",\n    linecolor = :red,\n)\nPlots.vline!(objective_space, [scalar_variance]; label = \"\", linecolor = :red)\nPlots.scatter!(\n    objective_space,\n    [value(variance; result = i) for i in 1:result_count(model)],\n    [value(expected_return; result = i) for i in 1:result_count(model)];\n    xlabel = \"Variance\",\n    ylabel = \"Expected Return\",\n    label = \"\",\n    title = \"Objective space\",\n    markercolor = \"white\",\n    markersize = 5,\n    legend = :bottomright,\n)\nfor i in 1:result_count(model)\n    y = objective_value(model; result = i)\n    Plots.annotate!(objective_space, y[1], -y[2], (i, 3))\nend\n\ndecision_space = StatsPlots.groupedbar(\n    vcat([value.(x; result = i)' for i in 1:result_count(model)]...);\n    bar_position = :stack,\n    label = [\"IBM\" \"WMT\" \"SEHI\"],\n    xlabel = \"Solution #\",\n    ylabel = \"Investment (\\$)\",\n    title = \"Decision space\",\n)\nPlots.plot(objective_space, decision_space; layout = (2, 1), size = (600, 600))","category":"page"},{"location":"tutorials/nonlinear/portfolio/","page":"Portfolio optimization","title":"Portfolio optimization","text":"Perhaps our trade-off wasn't so bad after all. Our original solution corresponded to picking a solution #17. If we buy more SEHI, we can increase the return, but the variance also increases. If we buy less SEHI, such as a solution like #5 or #6, then we can achieve the corresponding return without deploying all of our capital. We should also note that at no point should we buy WMT.","category":"page"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"EditURL = \"https://github.com/joaquimg/BilevelJuMP.jl/blob/v0.6.2/README.md\"","category":"page"},{"location":"packages/BilevelJuMP/#BilevelJuMP.jl","page":"joaquimg/BilevelJuMP.jl","title":"BilevelJuMP.jl","text":"","category":"section"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"(Image: Build Status) (Image: Codecov branch) (Image: Docs dev) (Image: Docs stable)","category":"page"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"BilevelJuMP.jl is a JuMP extension for modeling and solving bilevel optimization problems.","category":"page"},{"location":"packages/BilevelJuMP/#License","page":"joaquimg/BilevelJuMP.jl","title":"License","text":"","category":"section"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"BilevelJuMP.jl is licensed under the MIT license.","category":"page"},{"location":"packages/BilevelJuMP/#Documentation","page":"joaquimg/BilevelJuMP.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"You can find the documentation at https://joaquimg.github.io/BilevelJuMP.jl/stable/.","category":"page"},{"location":"packages/BilevelJuMP/#Help","page":"joaquimg/BilevelJuMP.jl","title":"Help","text":"","category":"section"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"If you need help, please open a GitHub issue.","category":"page"},{"location":"packages/BilevelJuMP/#Example","page":"joaquimg/BilevelJuMP.jl","title":"Example","text":"","category":"section"},{"location":"packages/BilevelJuMP/#Install","page":"joaquimg/BilevelJuMP.jl","title":"Install","text":"","category":"section"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"import Pkg\nPkg.add(\"BilevelJuMP\")\nPkg.add(\"HiGHS\")","category":"page"},{"location":"packages/BilevelJuMP/#Run","page":"joaquimg/BilevelJuMP.jl","title":"Run","text":"","category":"section"},{"location":"packages/BilevelJuMP/","page":"joaquimg/BilevelJuMP.jl","title":"joaquimg/BilevelJuMP.jl","text":"using BilevelJuMP, HiGHS\n\nmodel = BilevelModel(\n    HiGHS.Optimizer,\n    mode = BilevelJuMP.FortunyAmatMcCarlMode(primal_big_M = 100, dual_big_M = 100)\n)\n\n@variable(Lower(model), x)\n@variable(Upper(model), y)\n\n@objective(Upper(model), Min, 3x + y)\n@constraints(Upper(model), begin\n    x <= 5\n    y <= 8\n    y >= 0\nend)\n\n@objective(Lower(model), Min, -x)\n@constraints(Lower(model), begin\n     x +  y <= 8\n    4x +  y >= 8\n    2x +  y <= 13\n    2x - 7y <= 0\nend)\n\noptimize!(model)\n\nobjective_value(model) # = 3 * (3.5 * 8/15) + 8/15 # = 6.13...\nvalue(x) # = 3.5 * 8/15 # = 1.86...\nvalue(y) # = 8/15 # = 0.53...","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"EditURL = \"https://github.com/oxfordcontrol/Clarabel.jl/blob/v0.5.1/README.md\"","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"<h1 align=\"center\" margin=0px>\n  <img src=\"https://github.com/oxfordcontrol/ClarabelDocs/blob/main/docs/src/assets/logo-banner-light-jl.png?raw=true\" class=\"display-light-only\" width=60%>\n  <img src=\"https://github.com/oxfordcontrol/ClarabelDocs/blob/main/docs/src/assets/logo-banner-dark-jl.png?raw=true\" class=\"display-dark-only\"   width=60%>\n  <br>\nInterior Point Conic Optimization for Julia\n</h1>","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"<p align=\"center\">\n   <a href=\"https://github.com/oxfordcontrol/Clarabel.jl/actions\"><img src=\"https://github.com/oxfordcontrol/Clarabel.jl/workflows/ci/badge.svg?branch=main\"></a>\n  <a href=\"https://codecov.io/gh/oxfordcontrol/Clarabel.jl\"><img src=\"https://codecov.io/gh/oxfordcontrol/Clarabel.jl/branch/main/graph/badge.svg?raw=true\"></a>\n  <a href=\"https://oxfordcontrol.github.io/ClarabelDocs/stable\"><img src=\"https://img.shields.io/badge/Documentation-stable-purple.svg?raw=true\"></a>\n  <a href=\"https://opensource.org/licenses/Apache-2.0\"><img src=\"https://img.shields.io/badge/License-Apache%202.0-blue.svg?raw=true\"></a>\n  <a href=\"https://github.com/oxfordcontrol/Clarabel.jl/releases\"><img src=\"https://img.shields.io/badge/Release-v0.5.1-blue.svg?raw=true\"></a>\n</p>","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"<p align=\"center\">\n  <a href=\"#features\">Features</a> •\n  <a href=\"#installation\">Installation</a> •\n  <a href=\"#license-\">License</a> •\n  <a href=\"https://oxfordcontrol.github.io/ClarabelDocs/stable\">Documentation</a>\n</p>","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"Clarabel.jl is a Julia implementation of an interior point numerical solver for convex optimization problems using a novel homogeneous embedding.  Clarabel.jl solves the following problem:","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"beginarrayr\ntextminimize  frac12x^T P x + q^T x2ex\ntextsubject to  Ax + s = b 1ex\n         s in mathcalK\nendarray","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"with decision variables x in mathbbR^n, s in mathbbR^m and data matrices P=P^top succeq 0, q in mathbbR^n, A in mathbbR^m times n, and b in mathbbR^m. The convex set mathcalK is a composition of convex cones.","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"For more information see the Clarabel Documentation (stable |  dev).","category":"page"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"Clarabel is also available in a Rust / Python implementation.  See here.","category":"page"},{"location":"packages/Clarabel/#Features","page":"oxfordcontrol/Clarabel.jl","title":"Features","text":"","category":"section"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"Versatile: Clarabel.jl solves linear programs (LPs), quadratic programs (QPs), second-order cone programs (SOCPs) and semidefinite programs (SDPs).  It also solves problems with exponential and power cone constraints.\nQuadratic objectives: Unlike interior point solvers based on the standard homogeneous self-dual embedding (HSDE), Clarabel.jl handles quadratic objectives without requiring any epigraphical reformulation of the objective.   It can therefore be significantly faster than other HSDE-based solvers for problems with quadratic objective functions.\nInfeasibility detection: Infeasible problems are detected using a homogeneous embedding technique.\nJuMP / Convex.jl support: We provide an interface to MathOptInterface (MOI), which allows you to describe your problem in JuMP and Convex.jl.\nArbitrary precision types: You can solve problems with any floating point precision, for example, Float32 or Julia's BigFloat type, using either the native interface, or via MathOptInterface / Convex.jl.\nOpen Source: Our code is available on GitHub and distributed under the Apache 2.0 License","category":"page"},{"location":"packages/Clarabel/#Installation","page":"oxfordcontrol/Clarabel.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"Clarabel.jl can be added via the Julia package manager (type ]): pkg> add Clarabel","category":"page"},{"location":"packages/Clarabel/#License","page":"oxfordcontrol/Clarabel.jl","title":"License 🔍","text":"","category":"section"},{"location":"packages/Clarabel/","page":"oxfordcontrol/Clarabel.jl","title":"oxfordcontrol/Clarabel.jl","text":"This project is licensed under the Apache License  2.0 - see the LICENSE.md file for details.","category":"page"},{"location":"moi/reference/modification/","page":"Modifications","title":"Modifications","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/reference/modification.md\"","category":"page"},{"location":"moi/reference/modification/","page":"Modifications","title":"Modifications","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/reference/modification/#Modifications","page":"Modifications","title":"Modifications","text":"","category":"section"},{"location":"moi/reference/modification/","page":"Modifications","title":"Modifications","text":"modify\nAbstractFunctionModification\nScalarConstantChange\nVectorConstantChange\nScalarCoefficientChange\nMultirowChange","category":"page"},{"location":"moi/reference/modification/#MathOptInterface.modify","page":"Modifications","title":"MathOptInterface.modify","text":"Constraint Function\n\nmodify(model::ModelLike, ci::ConstraintIndex, change::AbstractFunctionModification)\n\nApply the modification specified by change to the function of constraint ci.\n\nAn ModifyConstraintNotAllowed error is thrown if modifying constraints is not supported by the model model.\n\nExamples\n\nmodify(model, ci, ScalarConstantChange(10.0))\n\nObjective Function\n\nmodify(model::ModelLike, ::ObjectiveFunction, change::AbstractFunctionModification)\n\nApply the modification specified by change to the objective function of model. To change the function completely, call set instead.\n\nAn ModifyObjectiveNotAllowed error is thrown if modifying objectives is not supported by the model model.\n\nExamples\n\nmodify(model, ObjectiveFunction{ScalarAffineFunction{Float64}}(), ScalarConstantChange(10.0))\n\nMultiple modifications in Constraint Functions\n\nmodify(\n    model::ModelLike,\n    cis::AbstractVector{<:ConstraintIndex},\n    changes::AbstractVector{<:AbstractFunctionModification},\n)\n\nApply multiple modifications specified by changes to the functions of constraints cis.\n\nA ModifyConstraintNotAllowed error is thrown if modifying constraints is not supported by model.\n\nExamples\n\nmodify(\n    model,\n    [ci, ci],\n    [\n        ScalarCoefficientChange{Float64}(VariableIndex(1), 1.0),\n        ScalarCoefficientChange{Float64}(VariableIndex(2), 0.5),\n    ],\n)\n\nMultiple modifications in the Objective Function\n\nmodify(\n    model::ModelLike,\n    attr::ObjectiveFunction,\n    changes::AbstractVector{<:AbstractFunctionModification},\n)\n\nApply multiple modifications specified by changes to the functions of constraints cis.\n\nA ModifyObjectiveNotAllowed error is thrown if modifying objective coefficients is not supported by model.\n\nExamples\n\nmodify(\n    model,\n    ObjectiveFunction{ScalarAffineFunction{Float64}}(),\n    [\n        ScalarCoefficientChange{Float64}(VariableIndex(1), 1.0),\n        ScalarCoefficientChange{Float64}(VariableIndex(2), 0.5),\n    ],\n)\n\n\n\n\n\n","category":"function"},{"location":"moi/reference/modification/#MathOptInterface.AbstractFunctionModification","page":"Modifications","title":"MathOptInterface.AbstractFunctionModification","text":"AbstractFunctionModification\n\nAn abstract supertype for structs which specify partial modifications to functions, to be used for making small modifications instead of replacing the functions entirely.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/modification/#MathOptInterface.ScalarConstantChange","page":"Modifications","title":"MathOptInterface.ScalarConstantChange","text":"ScalarConstantChange{T}(new_constant::T)\n\nA struct used to request a change in the constant term of a scalar-valued function.\n\nApplicable to ScalarAffineFunction and ScalarQuadraticFunction.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/modification/#MathOptInterface.VectorConstantChange","page":"Modifications","title":"MathOptInterface.VectorConstantChange","text":"VectorConstantChange{T}(new_constant::Vector{T})\n\nA struct used to request a change in the constant vector of a vector-valued function.\n\nApplicable to VectorAffineFunction and VectorQuadraticFunction.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/modification/#MathOptInterface.ScalarCoefficientChange","page":"Modifications","title":"MathOptInterface.ScalarCoefficientChange","text":"ScalarCoefficientChange{T}(variable::VariableIndex, new_coefficient::T)\n\nA struct used to request a change in the linear coefficient of a single variable in a scalar-valued function.\n\nApplicable to ScalarAffineFunction and ScalarQuadraticFunction.\n\n\n\n\n\n","category":"type"},{"location":"moi/reference/modification/#MathOptInterface.MultirowChange","page":"Modifications","title":"MathOptInterface.MultirowChange","text":"MultirowChange{T}(\n    variable::VariableIndex,\n    new_coefficients::Vector{Tuple{Int64,T}},\n) where {T}\n\nA struct used to request a change in the linear coefficients of a single variable in a vector-valued function.\n\nNew coefficients are specified by (output_index, coefficient) tuples.\n\nApplicable to VectorAffineFunction and VectorQuadraticFunction.\n\n\n\n\n\n","category":"type"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"EditURL = \"https://github.com/mariohsouto/ProxSDP.jl/blob/de648be90e63a69b8d8e4f82371760065d242fe9/README.md\"","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"<img src=\"https://raw.githubusercontent.com/mariohsouto/ProxSDP_aux/19f261a912e29c77cb68ee2957a5b41128824f22/logo_proxSDP.png?raw=true\" alt=\"logo\"/>","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"(Image: ) (Image: ) (Image: )","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"ProxSDP is an open-source semidefinite programming (SDP) solver based on the paper \"Exploiting Low-Rank Structure in Semidefinite Programming by Approximate Operator Splitting\".","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"The main advantage of ProxSDP over other state-of-the-art solvers is the ability to exploit the low-rank structure inherent to several SDP problems.","category":"page"},{"location":"packages/ProxSDP/#Overview-of-problems-ProxSDP-can-solve","page":"mariohsouto/ProxSDP.jl","title":"Overview of problems ProxSDP can solve","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"General conic convex optimization problems with the presence of the positive semidefinite cone, second-order cone and positive orthant;\nSemidefinite relaxation of nonconvex problems, for example, max-cut, binary MIMO, optimal power flow, sensor localization, sum-of-squares;\nControl theory problems with LMI constraints;\nNuclear norm minimization problems, for example, matrix completion;","category":"page"},{"location":"packages/ProxSDP/#License","page":"mariohsouto/ProxSDP.jl","title":"License","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"ProxSDP is licensed under the MIT License.","category":"page"},{"location":"packages/ProxSDP/#Installation","page":"mariohsouto/ProxSDP.jl","title":"Installation","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"Install ProxSDP using Julia package manager:","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"import Pkg\nPkg.add(\"ProxSDP\")","category":"page"},{"location":"packages/ProxSDP/#Documentation","page":"mariohsouto/ProxSDP.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"The online documentation is available at https://mariohsouto.github.io/ProxSDP.jl/latest/.","category":"page"},{"location":"packages/ProxSDP/#Usage","page":"mariohsouto/ProxSDP.jl","title":"Usage","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"Consider the semidefinite programming relaxation of the max-cut problem, as in:","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"max   0.25 * W•X\ns.t.  diag(X) = 1,\n      X ≽ 0,","category":"page"},{"location":"packages/ProxSDP/#JuMP","page":"mariohsouto/ProxSDP.jl","title":"JuMP","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"This problem can be solved by the following code using ProxSDP and JuMP.","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"# Load packages\nusing ProxSDP, JuMP, LinearAlgebra\n# Number of vertices\nn = 4\n# Graph weights\nW = [\n    18.0  -5.0  -7.0  -6.0\n    -5.0   6.0   0.0  -1.0\n    -7.0   0.0   8.0  -1.0\n    -6.0  -1.0  -1.0   8.0\n]\n# Build Max-Cut SDP relaxation via JuMP\nmodel = Model(ProxSDP.Optimizer)\nset_optimizer_attribute(model, \"log_verbose\", true)\nset_optimizer_attribute(model, \"tol_gap\", 1e-4)\nset_optimizer_attribute(model, \"tol_feasibility\", 1e-4)\n@variable(model, X[1:n, 1:n], PSD)\n@objective(model, Max, 0.25 * LinearAlgebra.dot(W, X))\n@constraint(model, LinearAlgebra.diag(X) .== 1)\n# Solve optimization problem with ProxSDP\noptimize!(model)\n# Retrieve solution\nXsol = value.(X)","category":"page"},{"location":"packages/ProxSDP/#Convex.jl","page":"mariohsouto/ProxSDP.jl","title":"Convex.jl","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"Another alternative is to use ProxSDP via Convex.jl:","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"# Load packages\nusing Convex, ProxSDP\n# Number of vertices\nn = 4\n# Graph weights\nW = [\n    18.0  -5.0  -7.0  -6.0\n    -5.0   6.0   0.0  -1.0\n    -7.0   0.0   8.0  -1.0\n    -6.0  -1.0  -1.0   8.0\n]\n# Define optimization problem\nX = Semidefinite(n)\nproblem = maximize(0.25 * dot(W, X), diag(X) == 1)\n# Solve optimization problem with ProxSDP\nsolve!(problem, ProxSDP.Optimizer(log_verbose=true, tol_gap=1e-4, tol_feasibility=1e-4))\n# Get the objective value\nproblem.optval\n# Retrieve solution\nevaluate(X)","category":"page"},{"location":"packages/ProxSDP/#Citing-this-package","page":"mariohsouto/ProxSDP.jl","title":"Citing this package","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"The published version of the paper can be found here and the arXiv version here.","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"We kindly request that you cite the paper as:","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"@article{souto2020exploiting,\n  author = {Mario Souto and Joaquim D. Garcia and \\'Alvaro Veiga},\n  title = {Exploiting low-rank structure in semidefinite programming by approximate operator splitting},\n  journal = {Optimization},\n  pages = {1-28},\n  year  = {2020},\n  publisher = {Taylor & Francis},\n  doi = {10.1080/02331934.2020.1823387},\n  URL = {https://doi.org/10.1080/02331934.2020.1823387}\n}","category":"page"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"The preprint version of the paper can be found here.","category":"page"},{"location":"packages/ProxSDP/#Disclaimer","page":"mariohsouto/ProxSDP.jl","title":"Disclaimer","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"ProxSDP is a research software, therefore it should not be used in production.\nPlease open an issue if you find any problems, developers will try to fix and find alternatives.\nThere is no continuous development for 32-bit systems, the package should work, but might reach some issues.\nProxSDP assumes primal and dual feasibility.","category":"page"},{"location":"packages/ProxSDP/#ROAD-MAP","page":"mariohsouto/ProxSDP.jl","title":"ROAD MAP","text":"","category":"section"},{"location":"packages/ProxSDP/","page":"mariohsouto/ProxSDP.jl","title":"mariohsouto/ProxSDP.jl","text":"Support for exponential and power cones\nWarm start","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"EditURL = \"https://github.com/jump-dev/HiGHS.jl/blob/v1.6.0/README.md\"","category":"page"},{"location":"packages/HiGHS/#HiGHS.jl","page":"jump-dev/HiGHS.jl","title":"HiGHS.jl","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"HiGHS.jl is a wrapper for the HiGHS solver.","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"It has two components:","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/HiGHS/#Affiliation","page":"jump-dev/HiGHS.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"This wrapper is maintained by the JuMP community and is not an official project of the HiGHS developers.","category":"page"},{"location":"packages/HiGHS/#License","page":"jump-dev/HiGHS.jl","title":"License","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"HiGHS.jl is licensed under the MIT License.","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"The underlying solver, ERGO-Code/HiGHS, is licensed under the MIT license.","category":"page"},{"location":"packages/HiGHS/#Installation","page":"jump-dev/HiGHS.jl","title":"Installation","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"Install HiGHS as follows:","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"import Pkg\nPkg.add(\"HiGHS\")","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"In addition to installing the HiGHS.jl package, this will also download and install the HiGHS binaries. You do not need to install HiGHS separately.","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"To use a custom binary, read the Custom solver binaries section of the JuMP documentation.","category":"page"},{"location":"packages/HiGHS/#Use-with-JuMP","page":"jump-dev/HiGHS.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"To use HiGHS with JuMP, use HiGHS.Optimizer:","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"using JuMP, HiGHS\nmodel = Model(HiGHS.Optimizer)\nset_attribute(model, \"presolve\", \"on\")\nset_attribute(model, \"time_limit\", 60.0)","category":"page"},{"location":"packages/HiGHS/#MathOptInterface-API","page":"jump-dev/HiGHS.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"The HiGHS optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"MOI.Reals","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.Semicontinuous{Float64}\nMOI.VariableIndex in MOI.Semiinteger{Float64}\nMOI.VariableIndex in MOI.ZeroOne","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"MOI.Name()\nMOI.ObjectiveSense()","category":"page"},{"location":"packages/HiGHS/#Options","page":"jump-dev/HiGHS.jl","title":"Options","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"See the HiGHS documentation for a full list of the available options.","category":"page"},{"location":"packages/HiGHS/#C-API","page":"jump-dev/HiGHS.jl","title":"C API","text":"","category":"section"},{"location":"packages/HiGHS/","page":"jump-dev/HiGHS.jl","title":"jump-dev/HiGHS.jl","text":"The C API can be accessed via HiGHS.Highs_xxx functions, where the names and arguments are identical to the C API.","category":"page"},{"location":"moi/submodules/FileFormats/reference/","page":"API Reference","title":"API Reference","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/FileFormats/reference.md\"","category":"page"},{"location":"moi/submodules/FileFormats/reference/","page":"API Reference","title":"API Reference","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/FileFormats/reference/#File-Formats","page":"API Reference","title":"File Formats","text":"","category":"section"},{"location":"moi/submodules/FileFormats/reference/","page":"API Reference","title":"API Reference","text":"Functions to help read and write MOI models to/from various file formats. See The FileFormats submodule for more details.","category":"page"},{"location":"moi/submodules/FileFormats/reference/","page":"API Reference","title":"API Reference","text":"FileFormats.Model\nFileFormats.FileFormat\nFileFormats.CBF.Model\nFileFormats.LP.Model\nFileFormats.MOF.Model\nFileFormats.MPS.Model\nFileFormats.NL.Model\nFileFormats.SDPA.Model","category":"page"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.Model","page":"API Reference","title":"MathOptInterface.FileFormats.Model","text":"Model(\n    ;\n    format::FileFormat = FORMAT_AUTOMATIC,\n    filename::Union{Nothing, String} = nothing,\n    kwargs...\n)\n\nReturn model corresponding to the FileFormat format, or, if format == FORMAT_AUTOMATIC, guess the format from filename.\n\nThe filename argument is only needed if format == FORMAT_AUTOMATIC.\n\nkwargs are passed to the underlying model constructor.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.FileFormat","page":"API Reference","title":"MathOptInterface.FileFormats.FileFormat","text":"FileFormat\n\nList of accepted export formats.\n\nFORMAT_AUTOMATIC: try to detect the file format based on the file name\nFORMAT_CBF: the Conic Benchmark format\nFORMAT_LP: the LP file format\nFORMAT_MOF: the MathOptFormat file format\nFORMAT_MPS: the MPS file format\nFORMAT_NL: the AMPL .nl file format\nFORMAT_REW: the .rew file format, which is MPS with generic names\nFORMAT_SDPA: the SemiDefinite Programming Algorithm format\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.CBF.Model","page":"API Reference","title":"MathOptInterface.FileFormats.CBF.Model","text":"Model()\n\nCreate an empty instance of FileFormats.CBF.Model.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.LP.Model","page":"API Reference","title":"MathOptInterface.FileFormats.LP.Model","text":"Model(; kwargs...)\n\nCreate an empty instance of FileFormats.LP.Model.\n\nKeyword arguments are:\n\nmaximum_length::Int=255: the maximum length for the name of a variable. lp_solve 5.0 allows only 16 characters, while CPLEX 12.5+ allow 255.\nwarn::Bool=false: print a warning when variables or constraints are renamed.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.MOF.Model","page":"API Reference","title":"MathOptInterface.FileFormats.MOF.Model","text":"Model(; kwargs...)\n\nCreate an empty instance of FileFormats.MOF.Model.\n\nKeyword arguments are:\n\nprint_compact::Bool=false: print the JSON file in a compact format without spaces or newlines.\nwarn::Bool=false: print a warning when variables or constraints are renamed\ndifferentiation_backend::MOI.Nonlinear.AbstractAutomaticDifferentiation = MOI.Nonlinear.SparseReverseMode(): automatic differentiation backend to use when reading models with nonlinear constraints and objectives.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.MPS.Model","page":"API Reference","title":"MathOptInterface.FileFormats.MPS.Model","text":"Model(; kwargs...)\n\nCreate an empty instance of FileFormats.MPS.Model.\n\nKeyword arguments are:\n\nwarn::Bool=false: print a warning when variables or constraints are renamed.\nprint_objsense::Bool=false: print the OBJSENSE section when writing\ngeneric_names::Bool=false: strip all names in the model and replace them with the generic names C$i and R$i for the i'th column and row respectively.\nquadratic_format::QuadraticFormat = kQuadraticFormatGurobi: specify the solver-specific extension used when writing the quadratic components of the model. Options are kQuadraticFormatGurobi, kQuadraticFormatCPLEX, and kQuadraticFormatMosek.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.NL.Model","page":"API Reference","title":"MathOptInterface.FileFormats.NL.Model","text":"Model(; use_nlp_block::Bool = true)\n\nCreate a new Optimizer object.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.SDPA.Model","page":"API Reference","title":"MathOptInterface.FileFormats.SDPA.Model","text":"Model(; number_type::Type = Float64)\n\nCreate an empty instance of FileFormats.SDPA.Model{number_type}.\n\nIt is important to be aware that the SDPA file format is interpreted in geometric form and not standard conic form. The standard conic form and geometric conic form are two dual standard forms for semidefinite programs (SDPs). The geometric conic form of an SDP is as follows:\n\nbeginalign\n min_y in mathbbR^m  b^T y\n\n textst  sum_i=1^m A_i y_i - C  in mathbbK\nendalign\n\nwhere mathcalK is a cartesian product of nonnegative orthant and positive semidefinite matrices that align with a block diagonal structure shared with the matrices A_i and C.\n\nIn other words, the geometric conic form contains free variables and affine constraints in either the nonnegative orthant or the positive semidefinite cone. That is, in the MathOptInterface's terminology, MOI.VectorAffineFunction-in-MOI.Nonnegatives and MOI.VectorAffineFunction-in-MOI.PositiveSemidefiniteConeTriangle constraints.\n\nThe corresponding standard conic form of the dual SDP is as follows:\n\nbeginalign\n max_X in mathbbK  texttr(CX)\n\n textst  texttr(A_iX)  = b_i  i = 1 ldots m\nendalign\n\nIn other words, the standard conic form contains nonnegative and positive semidefinite variables with equality constraints. That is, in the MathOptInterface's terminology, MOI.VectorOfVariables-in-MOI.Nonnegatives, MOI.VectorOfVariables-in-MOI.PositiveSemidefiniteConeTriangle and MOI.ScalarAffineFunction-in-MOI.EqualTo constraints.\n\nIf a model is in standard conic form, use Dualization.jl to transform it into the geometric conic form before writting it. Otherwise, the nonnegative (resp. positive semidefinite) variables will be bridged into free variables with affine constraints constraining them to belong to the nonnegative orthant (resp. positive semidefinite cone) by the MOI.Bridges.Constraint.VectorFunctionizeBridge. Moreover, equality constraints will be bridged into pairs of affine constraints in the nonnegative orthant by the MOI.Bridges.Constraint.SplitIntervalBridge and then the MOI.Bridges.Constraint.VectorizeBridge.\n\nIf a solver is in standard conic form, use Dualization.jl to transform the model read into standard conic form before copying it to the solver. Otherwise, the free variables will be bridged into pairs of variables in the nonnegative orthant by the MOI.Bridges.Variable.FreeBridge and affine constraints will be bridged into equality constraints by creating a slack variable by the MOI.Bridges.Constraint.VectorSlackBridge.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/FileFormats/reference/#Other-helpers","page":"API Reference","title":"Other helpers","text":"","category":"section"},{"location":"moi/submodules/FileFormats/reference/","page":"API Reference","title":"API Reference","text":"FileFormats.NL.SolFileResults","category":"page"},{"location":"moi/submodules/FileFormats/reference/#MathOptInterface.FileFormats.NL.SolFileResults","page":"API Reference","title":"MathOptInterface.FileFormats.NL.SolFileResults","text":"SolFileResults(filename::String, model::Model)\n\nParse the .sol file filename created by solving model and return a SolFileResults struct.\n\nThe returned struct supports the MOI.get API for querying result attributes such as MOI.TerminationStatus, MOI.VariablePrimal, and MOI.ConstraintDual.\n\n\n\n\n\nSolFileResults(\n    raw_status::String,\n    termination_status::MOI.TerminationStatusCode,\n)\n\nReturn a SolFileResults struct with MOI.RawStatusString set to raw_status, MOI.TerminationStatus set to termination_status, and MOI.PrimalStatus and MOI.DualStatus set to NO_SOLUTION.\n\nAll other attributes are un-set.\n\n\n\n\n\n","category":"type"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"EditURL = \"https://github.com/COPT-Public/COPT.jl/blob/v1.1.7/README.md\"","category":"page"},{"location":"packages/COPT/#COPT.jl","page":"COPT-Public/COPT.jl","title":"COPT.jl","text":"","category":"section"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"COPT.jl is a wrapper for the COPT (Cardinal Optimizer), a mathematical optimization solver for large-scale optimization problems.","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"COPT includes high-performance solvers for LP, MIP, SOCP, convex QP/QCP and SDP.","category":"page"},{"location":"packages/COPT/#License","page":"COPT-Public/COPT.jl","title":"License","text":"","category":"section"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"COPT.jl is licensed under the MIT license.","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"The underlying solver is a closed-source commercial product for which you must obtain a license.","category":"page"},{"location":"packages/COPT/#Note","page":"COPT-Public/COPT.jl","title":"Note","text":"","category":"section"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"When COPT is upgraded to a newer version, you may see an error message such as ERROR: COPT error 4: Unable to create COPT environment, which indicates that you will need to reapply and upgrade your COPT license files as well.","category":"page"},{"location":"packages/COPT/#Installation","page":"COPT-Public/COPT.jl","title":"Installation","text":"","category":"section"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"Install COPT using the Julia package manager","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"import Pkg\nPkg.add(\"COPT\")","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"When there is no local version of COPT installed, installing COPT.jl will automatically download the necessary solver binaries.","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"Without a license, you can solve small models for non-commercial purpose. We strongly recommend that you apply for a license by following the link above.","category":"page"},{"location":"packages/COPT/#Note-2","page":"COPT-Public/COPT.jl","title":"Note","text":"","category":"section"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"MacOS Apple M1/ARM: on MacOS with Apple M1 chips, Intel based programs can run via Rosetta. When you installed the COPT binaries manually, then please make sure that the COPT build matches the Julia build. We recommend the Intel based COPT and Julia build, as the Apple M1/ARM build of Julia is experimental.","category":"page"},{"location":"packages/COPT/#Use-with-JuMP","page":"COPT-Public/COPT.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"To use COPT with JuMP, use COPT.Optimizer:","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"using JuMP\nusing COPT\nmodel = Model(COPT.Optimizer)\n@variable(model, x >= 0)\n@variable(model, 0 <= y <= 3)\n@objective(model, Min, 12x + 20y)\n@constraint(model, c1, 6x + 8y >= 100)\n@constraint(model, c2, 7x + 12y >= 120)\nprint(model)\noptimize!(model)\n@show termination_status(model)\n@show primal_status(model)\n@show dual_status(model)\n@show objective_value(model)\n@show value(x)\n@show value(y)\n@show shadow_price(c1)\n@show shadow_price(c2)","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"To use the semidefinite programming solver in COPT with JuMP, use COPT.ConeOptimizer:","category":"page"},{"location":"packages/COPT/","page":"COPT-Public/COPT.jl","title":"COPT-Public/COPT.jl","text":"using JuMP\nusing COPT\nusing LinearAlgebra\nmodel = Model(COPT.ConeOptimizer)\nC = [1.0 -1.0; -1.0 2.0]\n@variable(model, X[1:2, 1:2], PSD)\n@variable(model, z[1:2] >= 0)\n@objective(model, Min, C ⋅ X)\n@constraint(model, c1, X[1, 1] - z[1] == 1)\n@constraint(model, c2, X[2, 2] - z[2] == 1)\noptimize!(model)\n@show termination_status(model)\n@show primal_status(model)\n@show dual_status(model)\n@show objective_value(model)\n@show value.(X)\n@show value.(z)\n@show shadow_price(c1)\n@show shadow_price(c2)","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"EditURL = \"cannery.jl\"","category":"page"},{"location":"tutorials/linear/cannery/#The-cannery-problem","page":"The cannery problem","title":"The cannery problem","text":"","category":"section"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"This tutorial was originally contributed by Louis Luangkesorn.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"This tutorial solves the cannery problem from Dantzig, Linear Programming and Extensions, Princeton University Press, Princeton, NJ, 1963. This class of problem is known as a transshipment problem.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"The purpose of this tutorial is to demonstrate how to use JSON data in the formulation of a JuMP model.","category":"page"},{"location":"tutorials/linear/cannery/#Required-packages","page":"The cannery problem","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"using JuMP\nimport HiGHS\nimport JSON","category":"page"},{"location":"tutorials/linear/cannery/#Formulation","page":"The cannery problem","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"The cannery problem assumes we are optimizing the shipment of cases of cans from production plants p in P to markets m in M.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"Each production plant p has a capacity c_p, and each market m has a demand d_m. The shipping cost per case of cans from plant p to market m is d_pm.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"We wish to find the distribution plan x_pm, the number of cases of cans to ship from plant p to market m, for p in P and m in M that minimizes the shipping costs. We can formulate our problem as the following linear program:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"beginaligned\nmin  sumlimits_p in Psumlimits_m in M d_pm x_pm \ntextst  sumlimits_m in M x_pm le c_p  forall p in P \n             sumlimits_p in P x_pm ge d_m  forall m in M \n             x_pm ge 0  forall p in P m in M\nendaligned","category":"page"},{"location":"tutorials/linear/cannery/#Data","page":"The cannery problem","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"A key feature of the tutorial is to demonstrate how to load data from JSON.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"For simplicity, we've hard-coded it below. But if the data was available as a .json file, we could use data = JSON.parsefile(filename) to read in the data.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"data = JSON.parse(\"\"\"\n{\n    \"plants\": {\n        \"Seattle\": {\"capacity\": 350},\n        \"San-Diego\": {\"capacity\": 600}\n    },\n    \"markets\": {\n        \"New-York\": {\"demand\": 300},\n        \"Chicago\": {\"demand\": 300},\n        \"Topeka\": {\"demand\": 300}\n    },\n    \"distances\": {\n        \"Seattle => New-York\": 2.5,\n        \"Seattle => Chicago\": 1.7,\n        \"Seattle => Topeka\": 1.8,\n        \"San-Diego => New-York\": 2.5,\n        \"San-Diego => Chicago\": 1.8,\n        \"San-Diego => Topeka\": 1.4\n    }\n}\n\"\"\")","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"Create the set of plants:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"P = keys(data[\"plants\"])","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"Create the set of markets:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"M = keys(data[\"markets\"])","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"We also need a function to compute the distance from plant to market:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"distance(p::String, m::String) = data[\"distances\"][\"$(p) => $(m)\"]","category":"page"},{"location":"tutorials/linear/cannery/#JuMP-formulation","page":"The cannery problem","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"Now we're ready to convert our mathematical formulation into a JuMP model.","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"First, create a new JuMP model. Since we have a linear program, we'll use HiGHS as our optimizer:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"model = Model(HiGHS.Optimizer)","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"Our decision variables are indexed over the set of plants and markets:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"@variable(model, x[P, M] >= 0)","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"We need a constraint that each plant can ship no more than its capacity:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"@constraint(model, [p in P], sum(x[p, :]) <= data[\"plants\"][p][\"capacity\"])","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"and each market must receive at least its demand:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"@constraint(model, [m in M], sum(x[:, m]) >= data[\"markets\"][m][\"demand\"])","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"Finally, our objective is to minimize the transportation distance:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"@objective(model, Min, sum(distance(p, m) * x[p, m] for p in P, m in M));\nnothing #hide","category":"page"},{"location":"tutorials/linear/cannery/#Solution","page":"The cannery problem","title":"Solution","text":"","category":"section"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"Let's optimize and look at the solution:","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"What's the optimal shipment?","category":"page"},{"location":"tutorials/linear/cannery/","page":"The cannery problem","title":"The cannery problem","text":"for p in P, m in M\n    println(p, \" => \", m, \": \", value(x[p, m]))\nend","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Utilities/overview.md\"","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Utilities/overview/#The-Utilities-submodule","page":"Overview","title":"The Utilities submodule","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The Utilities submodule provides a variety of functions and datastructures for managing MOI.ModelLike objects.","category":"page"},{"location":"moi/submodules/Utilities/overview/#Utilities.Model","page":"Overview","title":"Utilities.Model","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Utilities.Model provides an implementation of a ModelLike that efficiently supports all functions and sets defined within MOI. However, given the extensibility of MOI, this might not cover all use cases.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Create a model as follows:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}","category":"page"},{"location":"moi/submodules/Utilities/overview/#Utilities.UniversalFallback","page":"Overview","title":"Utilities.UniversalFallback","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Utilities.UniversalFallback is a layer that sits on top of any ModelLike and provides non-specialized (slower) fallbacks for constraints and attributes that the underlying ModelLike does not support.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"For example, Utilities.Model doesn't support some variable attributes like VariablePrimalStart, so JuMP uses a combination of Universal fallback and Utilities.Model as a generic problem cache:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}())\nMOIU.UniversalFallback{MOIU.Model{Float64}}\nfallback for MOIU.Model{Float64}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"warning: Warning\nAdding a UniversalFallback means that your model will now support all constraints, even if the inner-model does not. This can lead to unexpected behavior.","category":"page"},{"location":"moi/submodules/Utilities/overview/#Utilities.@model","page":"Overview","title":"Utilities.@model","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"For advanced use cases that need efficient support for functions and sets defined outside of MOI (but still known at compile time), we provide the Utilities.@model macro.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The @model macro takes a name (for a new type, which must not exist yet), eight tuples specifying the types of constraints that are supported, and then a Bool indicating the type is a subtype of MOI.AbstractOptimizer (if true) or MOI.ModelLike (if false).","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The eight tuples are in the following order:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Un-typed scalar sets, for example, Integer\nTyped scalar sets, for example, LessThan\nUn-typed vector sets, for example, Nonnegatives\nTyped vector sets, for example, PowerCone\nUn-typed scalar functions, for example, VariableIndex\nTyped scalar functions, for example, ScalarAffineFunction\nUn-typed vector functions, for example, VectorOfVariables\nTyped vector functions, for example, VectorAffineFunction","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The tuples can contain more than one element. Typed-sets must be specified without their type parameter, for example, MOI.LessThan, not MOI.LessThan{Float64}.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Here is an example:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Utilities.@model(\n           MyNewModel,\n           (MOI.Integer,),                  # Un-typed scalar sets\n           (MOI.GreaterThan,),              # Typed scalar sets\n           (MOI.Nonnegatives,),             # Un-typed vector sets\n           (MOI.PowerCone,),                # Typed vector sets\n           (MOI.VariableIndex,),            # Un-typed scalar functions\n           (MOI.ScalarAffineFunction,),     # Typed scalar functions\n           (MOI.VectorOfVariables,),        # Un-typed vector functions\n           (MOI.VectorAffineFunction,),     # Typed vector functions\n           true,                            # <:MOI.AbstractOptimizer?\n       )\nMathOptInterface.Utilities.GenericOptimizer{T, MathOptInterface.Utilities.ObjectiveContainer{T}, MathOptInterface.Utilities.VariablesContainer{T}, MyNewModelFunctionConstraints{T}} where T\n\njulia> model = MyNewModel{Float64}()\nMOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MyNewModelFunctionConstraints{Float64}}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"warning: Warning\nMyNewModel supports every VariableIndex-in-Set constraint, as well as VariableIndex, ScalarAffineFunction, and ScalarQuadraticFunction objective functions. Implement MOI.supports as needed to forbid constraint and objective function combinations.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"As another example, PATHSolver, which only supports VectorAffineFunction-in-Complements defines its optimizer as:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Utilities.@model(\n           PathOptimizer,\n           (),  # Scalar sets\n           (),  # Typed scalar sets\n           (MOI.Complements,),  # Vector sets\n           (),  # Typed vector sets\n           (),  # Scalar functions\n           (),  # Typed scalar functions\n           (),  # Vector functions\n           (MOI.VectorAffineFunction,),  # Typed vector functions\n           true,  # is_optimizer\n       )\nMathOptInterface.Utilities.GenericOptimizer{T, MathOptInterface.Utilities.ObjectiveContainer{T}, MathOptInterface.Utilities.VariablesContainer{T}, MathOptInterface.Utilities.VectorOfConstraints{MathOptInterface.VectorAffineFunction{T}, MathOptInterface.Complements}} where T","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"However, PathOptimizer does not support some VariableIndex-in-Set constraints, so we must explicitly define:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> function MOI.supports_constraint(\n           ::PathOptimizer,\n           ::Type{MOI.VariableIndex},\n           ::Type{Union{<:MOI.Semiinteger,MOI.Semicontinuous,MOI.ZeroOne,MOI.Integer}}\n       )\n           return false\n       end","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Finally, PATH doesn't support an objective function, so we need to add:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.supports(::PathOptimizer, ::MOI.ObjectiveFunction) = false","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"warning: Warning\nThis macro creates a new type, so it must be called from the top-level of a module, for example, it cannot be called from inside a function.","category":"page"},{"location":"moi/submodules/Utilities/overview/#Utilities.CachingOptimizer","page":"Overview","title":"Utilities.CachingOptimizer","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"A [Utilities.CachingOptimizer] is an MOI layer that abstracts the difference between solvers that support incremental modification (for example, they support adding variables one-by-one), and solvers that require the entire problem in a single API call (for example, they only accept the A, b and c matrices of a linear program).","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"It has two parts:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"A cache, where the model can be built and modified incrementally\nAn optimizer, which is used to solve the problem","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.Utilities.CachingOptimizer(\n           MOI.Utilities.Model{Float64}(),\n           PathOptimizer{Float64}(),\n       )\nMOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}\nin state EMPTY_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.Model{Float64}\nwith optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"A Utilities.CachingOptimizer may be in one of three possible states:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"NO_OPTIMIZER: The CachingOptimizer does not have any optimizer.\nEMPTY_OPTIMIZER: The CachingOptimizer has an empty optimizer, and it is not synchronized with the cached model. Modifications are forwarded to the cache, but not to the optimizer.\nATTACHED_OPTIMIZER: The CachingOptimizer has an optimizer, and it is synchronized with the cached model. Modifications are forwarded to the optimizer. If the optimizer does not support modifications, and error will be thrown.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Use Utilities.attach_optimizer to go from EMPTY_OPTIMIZER to ATTACHED_OPTIMIZER:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Utilities.attach_optimizer(model)\n\njulia> model\nMOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}\nin state ATTACHED_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.Model{Float64}\nwith optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"info: Info\nYou must be in ATTACHED_OPTIMIZER to use optimize!.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Use Utilities.reset_optimizer to go from ATTACHED_OPTIMIZER to EMPTY_OPTIMIZER:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Utilities.reset_optimizer(model)\n\njulia> model\nMOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}\nin state EMPTY_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.Model{Float64}\nwith optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"info: Info\nCalling MOI.empty!(model) also resets the state to EMPTY_OPTIMIZER. So after emptying a model, the modification will only be applied to the cache.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Use Utilities.drop_optimizer to go from any state to NO_OPTIMIZER:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Utilities.drop_optimizer(model)\n\njulia> model\nMOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}\nin state NO_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.Model{Float64}\nwith optimizer nothing","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Pass an empty optimizer to Utilities.reset_optimizer to go from NO_OPTIMIZER to EMPTY_OPTIMIZER:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Utilities.reset_optimizer(model, PathOptimizer{Float64}())\n\njulia> model\nMOIU.CachingOptimizer{MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}, MOIU.Model{Float64}}\nin state EMPTY_OPTIMIZER\nin mode AUTOMATIC\nwith model cache MOIU.Model{Float64}\nwith optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Deciding when to attach and reset the optimizer is tedious, and you will often write code like this:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"try\n    # modification\ncatch\n    MOI.Utilities.reset_optimizer(model)\n    # Re-try modification\nend","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"To make this easier, Utilities.CachingOptimizer has two modes of operation:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"AUTOMATIC: The CachingOptimizer changes its state when necessary. Attempting to add a constraint or perform a modification not supported by the optimizer results in a drop to EMPTY_OPTIMIZER mode.\nMANUAL: The user must change the state of the CachingOptimizer. Attempting to perform an operation in the incorrect state results in an error.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"By default, AUTOMATIC mode is chosen. However, you can create a CachingOptimizer in MANUAL mode as follows:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.Utilities.CachingOptimizer(\n           MOI.Utilities.Model{Float64}(),\n           MOI.Utilities.MANUAL,\n       )\nMOIU.CachingOptimizer{MOI.AbstractOptimizer, MOIU.Model{Float64}}\nin state NO_OPTIMIZER\nin mode MANUAL\nwith model cache MOIU.Model{Float64}\nwith optimizer nothing\n\njulia> MOI.Utilities.reset_optimizer(model, PathOptimizer{Float64}())\n\njulia> model\nMOIU.CachingOptimizer{MOI.AbstractOptimizer, MOIU.Model{Float64}}\nin state EMPTY_OPTIMIZER\nin mode MANUAL\nwith model cache MOIU.Model{Float64}\nwith optimizer MOIU.GenericOptimizer{Float64, MOIU.ObjectiveContainer{Float64}, MOIU.VariablesContainer{Float64}, MOIU.VectorOfConstraints{MOI.VectorAffineFunction{Float64}, MOI.Complements}}","category":"page"},{"location":"moi/submodules/Utilities/overview/#Printing","page":"Overview","title":"Printing","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Use print to print the formulation of the model.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.Utilities.Model{Float64}();\n\njulia> x = MOI.add_variable(model)\nMOI.VariableIndex(1)\n\njulia> MOI.set(model, MOI.VariableName(), x, \"x_var\")\n\njulia> MOI.add_constraint(model, x, MOI.ZeroOne())\nMathOptInterface.ConstraintIndex{MathOptInterface.VariableIndex, MathOptInterface.ZeroOne}(1)\n\njulia> MOI.set(model, MOI.ObjectiveFunction{typeof(x)}(), x)\n\njulia> MOI.set(model, MOI.ObjectiveSense(), MOI.MAX_SENSE)\n\njulia> print(model)\nMaximize VariableIndex:\n x_var\n\nSubject to:\n\nVariableIndex-in-ZeroOne\n x_var ∈ {0, 1}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Use Utilities.latex_formulation to display the model in LaTeX form:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> MOI.Utilities.latex_formulation(model)\n$$ \\begin{aligned}\n\\max\\quad & x\\_var \\\\\n\\text{Subject to}\\\\\n & \\text{VariableIndex-in-ZeroOne} \\\\\n & x\\_var \\in \\{0, 1\\} \\\\\n\\end{aligned} $$","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"tip: Tip\nIn IJulia, calling print or ending a cell with Utilities.latex_formulation will render the model in LaTeX.","category":"page"},{"location":"moi/submodules/Utilities/overview/#Utilities.PenaltyRelaxation","page":"Overview","title":"Utilities.PenaltyRelaxation","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Pass Utilities.PenaltyRelaxation to modify to relax the problem by adding penalized slack variables to the constraints. This is helpful when debugging sources of infeasible models.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.Utilities.Model{Float64}();\n\njulia> x = MOI.add_variable(model);\n\njulia> MOI.set(model, MOI.VariableName(), x, \"x\")\n\njulia> c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));\n\njulia> map = MOI.modify(model, MOI.Utilities.PenaltyRelaxation(Dict(c => 2.0)));\n\njulia> print(model)\nMinimize ScalarAffineFunction{Float64}:\n 0.0 + 2.0 v[2]\n\nSubject to:\n\nScalarAffineFunction{Float64}-in-LessThan{Float64}\n 0.0 + 1.0 x - 1.0 v[2] <= 2.0\n\nVariableIndex-in-GreaterThan{Float64}\n v[2] >= 0.0\n\njulia> map[c]\n0.0 + 1.0 MOI.VariableIndex(2)","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"You can also modify a single constraint using Utilities.ScalarPenaltyRelaxation:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.Utilities.Model{Float64}();\n\njulia> x = MOI.add_variable(model);\n\njulia> MOI.set(model, MOI.VariableName(), x, \"x\")\n\njulia> c = MOI.add_constraint(model, 1.0 * x, MOI.LessThan(2.0));\n\njulia> f = MOI.modify(model, c, MOI.Utilities.ScalarPenaltyRelaxation(2.0));\n\njulia> print(model)\nMinimize ScalarAffineFunction{Float64}:\n 0.0 + 2.0 v[2]\n\nSubject to:\n\nScalarAffineFunction{Float64}-in-LessThan{Float64}\n 0.0 + 1.0 x - 1.0 v[2] <= 2.0\n\nVariableIndex-in-GreaterThan{Float64}\n v[2] >= 0.0\n\njulia> f\n0.0 + 1.0 MOI.VariableIndex(2)","category":"page"},{"location":"moi/submodules/Utilities/overview/#Utilities.MatrixOfConstraints","page":"Overview","title":"Utilities.MatrixOfConstraints","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The constraints of Utilities.Model are stored as a vector of tuples of function and set in a Utilities.VectorOfConstraints. Other representations can be used by parameterizing the type Utilities.GenericModel (resp. Utilities.GenericOptimizer). For instance, if all non-VariableIndex constraints are affine, the coefficients of all the constraints can be stored in a single sparse matrix using Utilities.MatrixOfConstraints. The constraints storage can even be customized up to a point where it exactly matches the storage of the solver of interest, in which case copy_to can be implemented for the solver by calling copy_to to this custom model.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"For instance, Clp defines the following model:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"MOI.Utilities.@product_of_scalar_sets(LP, MOI.EqualTo{T}, MOI.LessThan{T}, MOI.GreaterThan{T})\nconst Model = MOI.Utilities.GenericModel{\n    Float64,\n    MOI.Utilities.MatrixOfConstraints{\n        Float64,\n        MOI.Utilities.MutableSparseMatrixCSC{Float64,Cint,MOI.Utilities.ZeroBasedIndexing},\n        MOI.Utilities.Hyperrectangle{Float64},\n        LP{Float64},\n    },\n}","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The copy_to operation can now be implemented as follows:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"function _copy_to(dest::Optimizer, src::Model)\n    @assert MOI.is_empty(dest)\n    A = src.constraints.coefficients\n    row_bounds = src.constraints.constants\n    Clp_loadProblem(\n        dest,\n        A.n,\n        A.m,\n        A.colptr,\n        A.rowval,\n        A.nzval,\n        src.lower_bound,\n        src.upper_bound,\n        # (...) objective vector (omitted),\n        row_bounds.lower,\n        row_bounds.upper,\n    )\n    # Set objective sense and constant (omitted)\n    return\nend\n\nfunction MOI.copy_to(dest::Optimizer, src::Model)\n    _copy_to(dest, src)\n    return MOI.Utilities.identity_index_map(src)\nend\n\nfunction MOI.copy_to(\n    dest::Optimizer,\n    src::MOI.Utilities.UniversalFallback{Model},\n)\n    # Copy attributes from `src` to `dest` and error in case any unsupported\n    # constraints or attributes are set in `UniversalFallback`.\n    return MOI.copy_to(dest, src.model)\nend\n\nfunction MOI.copy_to(\n    dest::Optimizer,\n    src::MOI.ModelLike,\n)\n    model = Model()\n    index_map = MOI.copy_to(model, src)\n    _copy_to(dest, model)\n    return index_map\nend","category":"page"},{"location":"moi/submodules/Utilities/overview/#ModelFilter","page":"Overview","title":"ModelFilter","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"Utilities provides Utilities.ModelFilter as a useful tool to copy a subset of a model. For example, given an infeasible model, we can copy the irreducible infeasible subsystem (for models implementing ConstraintConflictStatus) as follows:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"my_filter(::Any) = true\nfunction my_filter(ci::MOI.ConstraintIndex)\n    status = MOI.get(dest, MOI.ConstraintConflictStatus(), ci)\n    return status != MOI.NOT_IN_CONFLICT\nend\nfiltered_src = MOI.Utilities.ModelFilter(my_filter, src)\nindex_map = MOI.copy_to(dest, filtered_src)","category":"page"},{"location":"moi/submodules/Utilities/overview/#Fallbacks","page":"Overview","title":"Fallbacks","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The value of some attributes can be inferred from the value of other attributes.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"For example, the value of ObjectiveValue can be computed using ObjectiveFunction and VariablePrimal.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"When a solver gives direct access to an attribute, it is better to return this value. However, if this is not the case, Utilities.get_fallback can be used instead. For example:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"function MOI.get(model::Optimizer, attr::MOI.ObjectiveFunction)\n    return MOI.Utilities.get_fallback(model, attr)\nend","category":"page"},{"location":"moi/submodules/Utilities/overview/#DoubleDicts","page":"Overview","title":"DoubleDicts","text":"","category":"section"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"When writing MOI interfaces, we often need to handle situations in which we map ConstraintIndexs to different values. For example, to a string for ConstraintName.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"One option is to use a dictionary like Dict{MOI.ConstraintIndex,String}. However, this incurs a performance cost because the key is not a concrete type.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The DoubleDicts submodule helps this situation by providing two types main types Utilities.DoubleDicts.DoubleDict and Utilities.DoubleDicts.IndexDoubleDict. These types act like normal dictionaries, but internally they use more efficient dictionaries specialized to the type of the function-set pair.","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"The most common usage of a DoubleDict is in the index_map returned by copy_to. Performance can be improved, by using a function barrier. That is, instead of code like:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"index_map = MOI.copy_to(dest, src)\nfor (F, S) in MOI.get(src, MOI.ListOfConstraintTypesPresent())\n    for ci in MOI.get(src, MOI.ListOfConstraintIndices{F,S}())\n        dest_ci = index_map[ci]\n        # ...\n    end\nend","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"use instead:","category":"page"},{"location":"moi/submodules/Utilities/overview/","page":"Overview","title":"Overview","text":"function function_barrier(\n    dest,\n    src,\n    index_map::MOI.Utilities.DoubleDicts.IndexDoubleDictInner{F,S},\n) where {F,S}\n    for ci in MOI.get(src, MOI.ListOfConstraintIndices{F,S}())\n        dest_ci = index_map[ci]\n        # ...\n    end\n    return\nend\n\nindex_map = MOI.copy_to(dest, src)\nfor (F, S) in MOI.get(src, MOI.ListOfConstraintTypesPresent())\n    function_barrier(dest, src, index_map[F, S])\nend","category":"page"},{"location":"moi/submodules/Test/reference/","page":"API Reference","title":"API Reference","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Test/reference.md\"","category":"page"},{"location":"moi/submodules/Test/reference/","page":"API Reference","title":"API Reference","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Test/reference/#test_reference","page":"API Reference","title":"The Test submodule","text":"","category":"section"},{"location":"moi/submodules/Test/reference/","page":"API Reference","title":"API Reference","text":"Functions to help test implementations of MOI. See The Test submodule for more details.","category":"page"},{"location":"moi/submodules/Test/reference/","page":"API Reference","title":"API Reference","text":"Test.Config\nTest.runtests\nTest.setup_test\nTest.version_added\nTest.@requires\nTest.RequirementUnmet","category":"page"},{"location":"moi/submodules/Test/reference/#MathOptInterface.Test.Config","page":"API Reference","title":"MathOptInterface.Test.Config","text":"Config(\n    ::Type{T} = Float64;\n    atol::Real = Base.rtoldefault(T),\n    rtol::Real = Base.rtoldefault(T),\n    optimal_status::MOI.TerminationStatusCode = MOI.OPTIMAL,\n    infeasible_status::MOI.TerminationStatusCode = MOI.INFEASIBLE,\n    exclude::Vector{Any} = Any[],\n) where {T}\n\nReturn an object that is used to configure various tests.\n\nConfiguration arguments\n\natol::Real = Base.rtoldefault(T): Control the absolute tolerance used when comparing solutions.\nrtol::Real = Base.rtoldefault(T): Control the relative tolerance used when comparing solutions.\noptimal_status = MOI.OPTIMAL: Set to MOI.LOCALLY_SOLVED if the solver cannot prove global optimality.\ninfeasible_status = MOI.INFEASIBLE: Set to MOI.LOCALLY_INFEASIBLE if the solver cannot prove global infeasibility.\nexclude = Vector{Any}: Pass attributes or functions to exclude to skip parts of tests that require certain functionality. Common arguments include:\nMOI.delete to skip deletion-related tests\nMOI.optimize! to skip optimize-related tests\nMOI.ConstraintDual to skip dual-related tests\nMOI.VariableName to skip setting variable names\nMOI.ConstraintName to skip setting constraint names\n\nExamples\n\nFor a nonlinear solver that finds local optima and does not support finding dual variables or constraint names:\n\nConfig(\n    Float64;\n    optimal_status = MOI.LOCALLY_SOLVED,\n    exclude = Any[\n        MOI.ConstraintDual,\n        MOI.VariableName,\n        MOI.ConstraintName,\n        MOI.delete,\n    ],\n)\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Test/reference/#MathOptInterface.Test.runtests","page":"API Reference","title":"MathOptInterface.Test.runtests","text":"runtests(\n    model::MOI.ModelLike,\n    config::Config;\n    include::Vector{Union{String,Regex}} = String[],\n    exclude::Vector{Union{String,Regex}} = String[],\n    warn_unsupported::Bool = false,\n    exclude_tests_after::VersionNumber = v\"999.0.0\",\n)\n\nRun all tests in MathOptInterface.Test on model.\n\nConfiguration arguments\n\nconfig is a Test.Config object that can be used to modify the behavior of tests.\nIf include is not empty, only run tests if an element from include occursin the name of the test.\nIf exclude is not empty, skip tests if an element from exclude occursin the name of the test.\nexclude takes priority over include.\nIf warn_unsupported is false, runtests will silently skip tests that fail with a MOI.NotAllowedError, MOI.UnsupportedError, or RequirementUnmet error. (The latter is thrown when an @requires statement returns false.) When warn_unsupported is true, a warning will be printed. For most cases the default behavior, false, is what you want, since these tests likely test functionality that is not supported by model. However, it can be useful to run  warn_unsupported = true to check you are not skipping tests due to a missing supports_constraint method or equivalent.\nexclude_tests_after is a version number that excludes any tests to MOI added after that version number. This is useful for solvers who can declare a fixed set of tests, and not cause their tests to break if a new patch of MOI is released with a new test.\n\nSee also: setup_test.\n\nExample\n\nconfig = MathOptInterface.Test.Config()\nMathOptInterface.Test.runtests(\n    model,\n    config;\n    include = [\"test_linear_\", r\"^test_model_Name$\"],\n    exclude = [\"VariablePrimalStart\"],\n    warn_unsupported = true,\n    exclude_tests_after = v\"0.10.5\",\n)\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Test/reference/#MathOptInterface.Test.setup_test","page":"API Reference","title":"MathOptInterface.Test.setup_test","text":"setup_test(::typeof(f), model::MOI.ModelLike, config::Config)\n\nOverload this method to modify model before running the test function f on model with config. You can also modify the fields in config (e.g., to loosen the default tolerances).\n\nThis function should either return nothing, or return a function which, when called with zero arguments, undoes the setup to return the model to its previous state. You do not need to undo any modifications to config.\n\nThis function is most useful when writing new tests of the tests for MOI, but it can also be used to set test-specific tolerances, etc.\n\nSee also: runtests\n\nExample\n\nfunction MOI.Test.setup_test(\n    ::typeof(MOI.Test.test_linear_VariablePrimalStart_partial),\n    mock::MOIU.MockOptimizer,\n    ::MOI.Test.Config,\n)\n    MOIU.set_mock_optimize!(\n        mock,\n        (mock::MOIU.MockOptimizer) -> MOIU.mock_optimize!(mock, [1.0, 0.0]),\n    )\n    mock.eval_variable_constraint_dual = false\n\n    function reset_function()\n        mock.eval_variable_constraint_dual = true\n        return\n    end\n    return reset_function\nend\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Test/reference/#MathOptInterface.Test.version_added","page":"API Reference","title":"MathOptInterface.Test.version_added","text":"version_added(::typeof(function_name))\n\nReturns the version of MOI in which the test function_name was added.\n\nThis method should be implemented for all new tests.\n\nSee the exclude_tests_after keyword of runtests for more details.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Test/reference/#MathOptInterface.Test.@requires","page":"API Reference","title":"MathOptInterface.Test.@requires","text":"@requires(x)\n\nCheck that the condition x is true. Otherwise, throw an RequirementUnmet error to indicate that the model does not support something required by the test function.\n\nExamples\n\n@requires MOI.supports(model, MOI.Silent())\n@test MOI.get(model, MOI.Silent())\n\n\n\n\n\n","category":"macro"},{"location":"moi/submodules/Test/reference/#MathOptInterface.Test.RequirementUnmet","page":"API Reference","title":"MathOptInterface.Test.RequirementUnmet","text":"RequirementUnmet(msg::String) <: Exception\n\nAn error for throwing in tests to indicate that the model does not support some requirement expected by the test function.\n\n\n\n\n\n","category":"type"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"EditURL = \"https://github.com/MadNLP/MadNLP.jl/blob/389561354a774441cd5e6b3aa5fffee102ed222e/README.md\"","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"<img src=\"https://github.com/MadNLP/MadNLP.jl/blob/master/logo-full.svg?raw=true\"/>","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"Documentation Build Status Coverage DOI\n(Image: doc) (Image: build) (Image: codecov) (Image: DOI)","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"MadNLP is a nonlinear programming (NLP) solver, purely implemented in Julia. MadNLP implements a filter line-search algorithm, as that used in Ipopt. MadNLP seeks to streamline the development of modeling and algorithmic paradigms in order to exploit structures and to make efficient use of high-performance computers.","category":"page"},{"location":"packages/MadNLP/#License","page":"MadNLP/MadNLP.jl","title":"License","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"MadNLP is available under the MIT license.","category":"page"},{"location":"packages/MadNLP/#Installation","page":"MadNLP/MadNLP.jl","title":"Installation","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"pkg> add MadNLP","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"Optionally, various extension packages can be installed together:","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"pkg> add MadNLPHSL, MadNLPPardiso, MadNLPMumps, MadNLPGPU, MadNLPGraph, MadNLPKrylov","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"These packages are stored in the lib subdirectory within the main MadNLP repository. Some extension packages may require additional dependencies or specific hardware. For the instructions for the build procedure, see the following links:","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"MadNLPHSL\nMadNLPPardiso\nMadNLPGPU","category":"page"},{"location":"packages/MadNLP/#Usage","page":"MadNLP/MadNLP.jl","title":"Usage","text":"","category":"section"},{"location":"packages/MadNLP/#Interfaces","page":"MadNLP/MadNLP.jl","title":"Interfaces","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"MadNLP is interfaced with modeling packages:","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"JuMP\nPlasmo\nNLPModels.","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"Users can pass various options to MadNLP also through the modeling packages. The interface-specific syntax are shown below. To see the list of MadNLP solver options, check the OPTIONS.md file.","category":"page"},{"location":"packages/MadNLP/#JuMP-interface","page":"MadNLP/MadNLP.jl","title":"JuMP interface","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, JuMP\nmodel = Model(()->MadNLP.Optimizer(print_level=MadNLP.INFO, max_iter=100))\n@variable(model, x, start = 0.0)\n@variable(model, y, start = 0.0)\n@NLobjective(model, Min, (1 - x)^2 + 100 * (y - x^2)^2)\noptimize!(model)","category":"page"},{"location":"packages/MadNLP/#NLPModels-interface","page":"MadNLP/MadNLP.jl","title":"NLPModels interface","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, CUTEst\nmodel = CUTEstModel(\"PRIMALC1\")\nmadnlp(model, print_level=MadNLP.WARN, max_wall_time=3600)","category":"page"},{"location":"packages/MadNLP/#Plasmo-interface-(requires-extension-MadNLPGraph)","page":"MadNLP/MadNLP.jl","title":"Plasmo interface (requires extension MadNLPGraph)","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, MadNLPGraph, Plasmo\ngraph = OptiGraph()\n@optinode(graph,n1)\n@optinode(graph,n2)\n@variable(n1,0 <= x <= 2)\n@variable(n1,0 <= y <= 3)\n@constraint(n1,x+y <= 4)\n@objective(n1,Min,x)\n@variable(n2,x)\n@NLnodeconstraint(n2,exp(x) >= 2)\n@linkconstraint(graph,n1[:x] == n2[:x])\nMadNLP.optimize!(graph; print_level=MadNLP.DEBUG, max_iter=100)","category":"page"},{"location":"packages/MadNLP/#Linear-Solvers","page":"MadNLP/MadNLP.jl","title":"Linear Solvers","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"MadNLP is interfaced with non-Julia sparse/dense linear solvers:","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"Umfpack\nMKL-Pardiso\nMKL-Lapack\nHSL solvers (requires extension)\nPardiso (requires extension)\nMumps  (requires extension)\ncuSOLVER (requires extension)","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"Each linear solver in MadNLP is a Julia type, and the linear_solver option should be specified by the actual type. Note that the linear solvers are always exported to Main.","category":"page"},{"location":"packages/MadNLP/#Built-in-Solvers:-Umfpack,-PardisoMKL,-LapackCPU","page":"MadNLP/MadNLP.jl","title":"Built-in Solvers: Umfpack, PardisoMKL, LapackCPU","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, JuMP\n# ...\nmodel = Model(()->MadNLP.Optimizer(linear_solver=UmfpackSolver)) # default\nmodel = Model(()->MadNLP.Optimizer(linear_solver=LapackCPUSolver))","category":"page"},{"location":"packages/MadNLP/#HSL-(requires-extension-MadNLPHSL)","page":"MadNLP/MadNLP.jl","title":"HSL (requires extension MadNLPHSL)","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, MadNLPHSL, JuMP\n# ...\nmodel = Model(()->MadNLP.Optimizer(linear_solver=Ma27Solver))\nmodel = Model(()->MadNLP.Optimizer(linear_solver=Ma57Solver))\nmodel = Model(()->MadNLP.Optimizer(linear_solver=Ma77Solver))\nmodel = Model(()->MadNLP.Optimizer(linear_solver=Ma86Solver))\nmodel = Model(()->MadNLP.Optimizer(linear_solver=Ma97Solver))","category":"page"},{"location":"packages/MadNLP/#Mumps-(requires-extension-MadNLPMumps)","page":"MadNLP/MadNLP.jl","title":"Mumps (requires extension MadNLPMumps)","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, MadNLPMumps, JuMP\n# ...\nmodel = Model(()->MadNLP.Optimizer(linear_solver=MumpsSolver))","category":"page"},{"location":"packages/MadNLP/#Pardiso-(requires-extension-MadNLPPardiso)","page":"MadNLP/MadNLP.jl","title":"Pardiso (requires extension MadNLPPardiso)","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, MadNLPPardiso, JuMP\n# ...\nmodel = Model(()->MadNLP.Optimizer(linear_solver=PardisoSolver))\nmodel = Model(()->MadNLP.Optimizer(linear_solver=PardisoMKLSolver))","category":"page"},{"location":"packages/MadNLP/#LapackGPU-(requires-extension-MadNLPGPU)","page":"MadNLP/MadNLP.jl","title":"LapackGPU (requires extension MadNLPGPU)","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, MadNLPGPU, JuMP\n# ...\nmodel = Model(()->MadNLP.Optimizer(linear_solver=LapackGPUSolver))","category":"page"},{"location":"packages/MadNLP/#Schur-and-Schwarz-(requires-extension-MadNLPGraph)","page":"MadNLP/MadNLP.jl","title":"Schur and Schwarz (requires extension MadNLPGraph)","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"using MadNLP, MadNLPGraph, JuMP\n# ...\nmodel = Model(()->MadNLP.Optimizer(linear_solver=MadNLPSchwarz))\nmodel = Model(()->MadNLP.Optimizer(linear_solver=MadNLPSchur))","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"The solvers in MadNLPGraph (Schur and Schwarz) use multi-thread parallelism; thus, Julia session should be started with -t flag.","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"julia -t 16 # to use 16 threads","category":"page"},{"location":"packages/MadNLP/#Citing-MadNLP.jl","page":"MadNLP/MadNLP.jl","title":"Citing MadNLP.jl","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"If you use MadNLP.jl in your research, we would greatly appreciate your citing it.","category":"page"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"@article{shin2020graph,\n  title={Graph-Based Modeling and Decomposition of Energy Infrastructures},\n  author={Shin, Sungho and Coffrin, Carleton and Sundar, Kaarthik and Zavala, Victor M},\n  journal={arXiv preprint arXiv:2010.02404},\n  year={2020}\n}","category":"page"},{"location":"packages/MadNLP/#Bug-reports-and-support","page":"MadNLP/MadNLP.jl","title":"Bug reports and support","text":"","category":"section"},{"location":"packages/MadNLP/","page":"MadNLP/MadNLP.jl","title":"MadNLP/MadNLP.jl","text":"Please report issues and feature requests via the GitHub issue tracker.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\nend","category":"page"},{"location":"manual/variables/#jump_variables","page":"Variables","title":"Variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The term variable in mathematical optimization has many meanings. For example, optimization variables (also called decision variables) are the unknowns x that we are solving for in the problem:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"beginalign\n     min_x in mathbbR^n  f_0(x)\n    \n     textst  f_i(x)  in mathcalS_i  i = 1 ldots m\nendalign","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"To complicate things, Julia uses variable to mean a binding between a name and a value. For example, in the statement:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> x = 1\n1","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"x is a variable that stores the value 1.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"JuMP uses variable in a third way, to mean an instance of the VariableRef struct. JuMP variables are the link between Julia and the optimization variables inside a JuMP model.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"This page explains how to create and manage JuMP variables in a variety of contexts.","category":"page"},{"location":"manual/variables/#Create-a-variable","page":"Variables","title":"Create a variable","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Create variables using the @variable macro:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> typeof(x)\nVariableRef (alias for GenericVariableRef{Float64})\n\njulia> num_variables(model)\n1","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Here x is a Julia variable that is bound to a VariableRef object, and we have added 1 decision variable to our model.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"To make the binding more explicit, we could have written:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> x = @variable(model, x)\nx","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"but there is no need to in general; the macro does it for us.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"When creating a variable, you can also specify variable bounds:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x_free)\nx_free\n\njulia> @variable(model, x_lower >= 0)\nx_lower\n\njulia> @variable(model, x_upper <= 1)\nx_upper\n\njulia> @variable(model, 2 <= x_interval <= 3)\nx_interval\n\njulia> @variable(model, x_fixed == 4)\nx_fixed\n\njulia> print(model)\nFeasibility\nSubject to\n x_fixed = 4\n x_lower ≥ 0\n x_interval ≥ 2\n x_upper ≤ 1\n x_interval ≤ 3","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"warning: Warning\nWhen creating a variable with a single lower- or upper-bound, and the value of the bound is not a numeric literal (for example, 1 or 1.0), the name of the variable must appear on the left-hand side. Putting the name on the right-hand side is an error. For example, to create a variable x:a = 1\n@variable(model, x >= 1)      # ✓ Okay\n@variable(model, 1.0 <= x)    # ✓ Okay\n@variable(model, x >= a)      # ✓ Okay\n@variable(model, a <= x)      # × Not okay\n@variable(model, x >= 1 / 2)  # ✓ Okay\n@variable(model, 1 / 2 <= x)  # × Not okay","category":"page"},{"location":"manual/variables/#Containers-of-variables","page":"Variables","title":"Containers of variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The @variable macro also supports creating collections of JuMP variables. We'll cover some brief syntax here; read the Variable containers section for more details.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"You can create arrays of JuMP variables:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2])\n2×2 Matrix{VariableRef}:\n x[1,1]  x[1,2]\n x[2,1]  x[2,2]\n\njulia> x[1, 2]\nx[1,2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Index sets can be named, and bounds can depend on those names:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, sqrt(i) <= x[i = 1:3] <= i^2)\n3-element Vector{VariableRef}:\n x[1]\n x[2]\n x[3]\n\njulia> x[2]\nx[2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Sets can be any Julia type that supports iteration:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[i = 2:3, j = 1:2:3, [\"red\", \"blue\"]] >= 0)\n3-dimensional DenseAxisArray{VariableRef,3,...} with index sets:\n    Dimension 1, 2:3\n    Dimension 2, 1:2:3\n    Dimension 3, [\"red\", \"blue\"]\nAnd data, a 2×2×2 Array{VariableRef, 3}:\n[:, :, \"red\"] =\n x[2,1,red]  x[2,3,red]\n x[3,1,red]  x[3,3,red]\n\n[:, :, \"blue\"] =\n x[2,1,blue]  x[2,3,blue]\n x[3,1,blue]  x[3,3,blue]\n\njulia> x[2, 1, \"red\"]\nx[2,1,red]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Sets can depend upon previous indices:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, u[i = 1:2, j = i:3])\nJuMP.Containers.SparseAxisArray{VariableRef, 2, Tuple{Int64, Int64}} with 5 entries:\n  [1, 1]  =  u[1,1]\n  [1, 2]  =  u[1,2]\n  [1, 3]  =  u[1,3]\n  [2, 2]  =  u[2,2]\n  [2, 3]  =  u[2,3]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"and we can filter elements in the sets using the ; syntax:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, v[i = 1:9; mod(i, 3) == 0])\nJuMP.Containers.SparseAxisArray{VariableRef, 1, Tuple{Int64}} with 3 entries:\n  [3]  =  v[3]\n  [6]  =  v[6]\n  [9]  =  v[9]","category":"page"},{"location":"manual/variables/#Registered-variables","page":"Variables","title":"Registered variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"When you create variables, JuMP registers them inside the model using their corresponding symbol. Get a registered name using model[:key]:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model()\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\n\njulia> @variable(model, x)\nx\n\njulia> model\nA JuMP Model\nFeasibility problem with:\nVariable: 1\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\nNames registered in the model: x\n\njulia> model[:x] === x\ntrue","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Registered names are most useful when you start to write larger models and want to break up the model construction into functions:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> function set_objective(model::Model)\n           @objective(model, Min, 2 * model[:my_x] + 1)\n           return\n       end\nset_objective (generic function with 1 method)\n\njulia> model = Model();\n\njulia> @variable(model, my_x);\n\njulia> set_objective(model)\n\njulia> print(model)\nMin 2 my_x + 1\nSubject to","category":"page"},{"location":"manual/variables/#anonymous_variables","page":"Variables","title":"Anonymous variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"To reduce the likelihood of accidental bugs, and because JuMP registers variables inside a model, creating two variables with the same name is an error:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> @variable(model, x)\nERROR: An object of name x is already attached to this model. If this\n    is intended, consider using the anonymous construction syntax, e.g.,\n    `x = @variable(model, [1:N], ...)` where the name of the object does\n    not appear inside the macro.\n\n    Alternatively, use `unregister(model, :x)` to first unregister\n    the existing name from the model. Note that this will not delete the\n    object; it will just remove the reference at `model[:x]`.\n[...]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"A common reason for encountering this error is adding variables in a loop.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"As a work-around, JuMP provides anonymous variables. Create a scalar valued anonymous variable by omitting the name argument:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> x = @variable(model)\n_[1]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Anonymous variables get printed as an underscore followed by a unique index of the variable.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"warning: Warning\nThe index of the variable may not correspond to the column of the variable in the solver.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Create a container of anonymous JuMP variables by dropping the name in front of the [:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> y = @variable(model, [1:2])\n2-element Vector{VariableRef}:\n _[1]\n _[2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The <= and >= short-hand cannot be used to set bounds on scalar-valued anonymous JuMP variables. Instead, use the lower_bound and upper_bound keywords:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> x_lower = @variable(model, lower_bound = 1.0)\n_[1]\n\njulia> x_upper = @variable(model, upper_bound = 2.0)\n_[2]\n\njulia> x_interval = @variable(model, lower_bound = 3.0, upper_bound = 4.0)\n_[3]","category":"page"},{"location":"manual/variables/#Variable-names","page":"Variables","title":"Variable names","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"In addition to the symbol that variables are registered with, JuMP variables have a String name that is used for printing and writing to file formats.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Get and set the name of a variable using name and set_name:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> name(x)\n\"x\"\n\njulia> set_name(x, \"my_x_name\")\n\njulia> x\nmy_x_name","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Override the default choice of name using the base_name keyword:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:2], base_name = \"my_var\")\n2-element Vector{VariableRef}:\n my_var[1]\n my_var[2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Note that names apply to each element of the container, not to the container of variables:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> name(x[1])\n\"my_var[1]\"\n\njulia> set_name(x[1], \"my_x\")\n\njulia> x\n2-element Vector{VariableRef}:\n my_x\n my_var[2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"tip: Tip\nFor some models, setting the string name of each variable can take a non-trivial portion of the total time required to build the model. Turn off String names by passing set_string_name = false to @variable:julia> model = Model();\n\njulia> @variable(model, x, set_string_name = false)\n_[1]See Disable string names for more information.","category":"page"},{"location":"manual/variables/#Retrieve-a-variable-by-name","page":"Variables","title":"Retrieve a variable by name","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Retrieve a variable from a model using variable_by_name:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> variable_by_name(model, \"my_x\")\nmy_x","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"If the name is not present, nothing will be returned:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> variable_by_name(model, \"bad_name\")","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"You can only look up individual variables using variable_by_name. Something like this will not work:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, [i = 1:2], base_name = \"my_var\")\n2-element Vector{VariableRef}:\n my_var[1]\n my_var[2]\n\njulia> variable_by_name(model, \"my_var\")","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"To look up a collection of variables, do not use variable_by_name. Instead, register them using the model[:key] = value syntax:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> model[:x] = @variable(model, [i = 1:2], base_name = \"my_var\")\n2-element Vector{VariableRef}:\n my_var[1]\n my_var[2]\n\njulia> model[:x]\n2-element Vector{VariableRef}:\n my_var[1]\n my_var[2]","category":"page"},{"location":"manual/variables/#variable_names_and_bindings","page":"Variables","title":"String names, symbolic names, and bindings","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"It's common for new users to experience confusion relating to JuMP variables. Part of the problem is the overloaded use of \"variable\" in mathematical optimization, along with the difference between the name that a variable is registered under and the String name used for printing.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Here's a summary of the differences:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"JuMP variables are created using @variable.\nJuMP variables can be named or anonymous.\nNamed JuMP variables have the form @variable(model, x). For named variables:\nThe String name of the variable is set to \"x\".\nA Julia variable x is created that binds x to  the JuMP variable.\nThe name :x is registered as a key in the model with the value x.\nAnonymous JuMP variables have the form x = @variable(model). For anonymous variables:\nThe String name of the variable is set to \"\". When printed, this is replaced with \"_[i]\" where i is the index of the variable.\nYou control the name of the Julia variable used as the binding.\nNo name is registered as a key in the model.\nThe base_name keyword can override the String name of the variable.\nYou can manually register names in the model via model[:key] = value","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Here's an example that should make things clearer:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> x_binding = @variable(model, base_name = \"x\")\nx\n\njulia> model\nA JuMP Model\nFeasibility problem with:\nVariable: 1\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\n\njulia> x\nERROR: UndefVarError: `x` not defined\n\njulia> x_binding\nx\n\njulia> name(x_binding)\n\"x\"\n\njulia> model[:x_register] = x_binding\nx\n\njulia> model\nA JuMP Model\nFeasibility problem with:\nVariable: 1\nModel mode: AUTOMATIC\nCachingOptimizer state: NO_OPTIMIZER\nSolver name: No optimizer attached.\nNames registered in the model: x_register\n\njulia> model[:x_register]\nx\n\njulia> model[:x_register] === x_binding\ntrue\n\njulia> x\nERROR: UndefVarError: `x` not defined","category":"page"},{"location":"manual/variables/#Create,-delete,-and-modify-variable-bounds","page":"Variables","title":"Create, delete, and modify variable bounds","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Query whether a variable has a bound using has_lower_bound, has_upper_bound, and is_fixed:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> has_lower_bound(x_free)\nfalse\n\njulia> has_upper_bound(x_upper)\ntrue\n\njulia> is_fixed(x_fixed)\ntrue","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"If a variable has a particular bound, query the value of it using lower_bound, upper_bound, and fix_value:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> lower_bound(x_interval)\n2.0\n\njulia> upper_bound(x_interval)\n3.0\n\njulia> fix_value(x_fixed)\n4.0","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Querying the value of a bound that does not exist will result in an error.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Delete variable bounds using delete_lower_bound, delete_upper_bound, and unfix:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> delete_lower_bound(x_lower)\n\njulia> has_lower_bound(x_lower)\nfalse\n\njulia> delete_upper_bound(x_upper)\n\njulia> has_upper_bound(x_upper)\nfalse\n\njulia> unfix(x_fixed)\n\njulia> is_fixed(x_fixed)\nfalse","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Set or update variable bounds using set_lower_bound, set_upper_bound, and fix:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> set_lower_bound(x_lower, 1.1)\n\njulia> set_upper_bound(x_upper, 2.1)\n\njulia> fix(x_fixed, 4.1)","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Fixing a variable with existing bounds will throw an error. To delete the bounds prior to fixing, use fix(variable, value; force = true).","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x >= 1)\nx\n\njulia> fix(x, 2)\nERROR: Unable to fix x to 2 because it has existing variable bounds. Consider calling `JuMP.fix(variable, value; force=true)` which will delete existing bounds before fixing the variable.\n\njulia> fix(x, 2; force = true)\n\n\njulia> fix_value(x)\n2.0","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"tip: Tip\nUse fix instead of @constraint(model, x == 2). The former modifies variable bounds, while the latter adds a new linear constraint to the problem.","category":"page"},{"location":"manual/variables/#Binary-variables","page":"Variables","title":"Binary variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Binary variables are constrained to the set x in 0 1.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Create a binary variable by passing Bin as an optional positional argument:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x, Bin)\nx","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Check if a variable is binary using is_binary:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> is_binary(x)\ntrue","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Delete a binary constraint using unset_binary:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> unset_binary(x)\n\njulia> is_binary(x)\nfalse","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Binary variables can also be created by setting the binary keyword to true:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x, binary=true)\nx","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"or by using set_binary:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> set_binary(x)","category":"page"},{"location":"manual/variables/#Integer-variables","page":"Variables","title":"Integer variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Integer variables are constrained to the set x in mathbbZ.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Create an integer variable by passing Int as an optional positional argument:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x, Int)\nx","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Check if a variable is integer using is_integer:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> is_integer(x)\ntrue","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Delete an integer constraint using unset_integer.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> unset_integer(x)\n\njulia> is_integer(x)\nfalse","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Integer variables can also be created by setting the integer keyword to true:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x, integer=true)\nx","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"or by using set_integer:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> set_integer(x)","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"tip: Tip\nThe relax_integrality function relaxes all integrality constraints in the model, returning a function that can be called to undo the operation later on.","category":"page"},{"location":"manual/variables/#Start-values","page":"Variables","title":"Start values","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"There are two ways to provide a primal starting solution (also called MIP-start or a warmstart) for each variable:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"using the start keyword in the @variable macro\nusing set_start_value","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The starting value of a variable can be queried using start_value. If no start value has been set, start_value will return nothing.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> start_value(x)\n\njulia> @variable(model, y, start = 1)\ny\n\njulia> start_value(y)\n1.0\n\njulia> set_start_value(y, 2)\n\njulia> start_value(y)\n2.0","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The start keyword argument can depend on the indices of a variable container:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, z[i = 1:2], start = i^2)\n2-element Vector{VariableRef}:\n z[1]\n z[2]\n\njulia> start_value.(z)\n2-element Vector{Float64}:\n 1.0\n 4.0","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"warning: Warning\nSome solvers do not support start values. If a solver does not support start values, an MathOptInterface.UnsupportedAttribute{MathOptInterface.VariablePrimalStart} error will be thrown.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"tip: Tip\nTo set the optimal solution from a previous solve as a new starting value, use all_variables to get a vector of all the variables in the model, then run:x = all_variables(model)\nx_solution = value.(x)\nset_start_value.(x, x_solution)Alternatively, use set_start_values.","category":"page"},{"location":"manual/variables/#delete_a_variable","page":"Variables","title":"Delete a variable","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Use delete to delete a variable from a model. Use is_valid to check if a variable belongs to a model and has not been deleted.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> is_valid(model, x)\ntrue\n\njulia> delete(model, x)\n\njulia> is_valid(model, x)\nfalse","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Deleting a variable does not unregister the corresponding name from the model. Therefore, creating a new variable of the same name will throw an error:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> @variable(model, x)\nERROR: An object of name x is already attached to this model. If this\n    is intended, consider using the anonymous construction syntax, e.g.,\n    `x = @variable(model, [1:N], ...)` where the name of the object does\n    not appear inside the macro.\n\n    Alternatively, use `unregister(model, :x)` to first unregister\n    the existing name from the model. Note that this will not delete the\n    object; it will just remove the reference at `model[:x]`.\n[...]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"After calling delete, call unregister to remove the symbolic reference:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> unregister(model, :x)\n\njulia> @variable(model, x)\nx","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"info: Info\ndelete does not automatically unregister because we do not distinguish between names that are automatically registered by JuMP macros and names that are manually registered by the user by setting values in object_dictionary. In addition, deleting a variable and then adding a new variable of the same name is an easy way to introduce bugs into your code.","category":"page"},{"location":"manual/variables/#Variable-containers","page":"Variables","title":"Variable containers","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"JuMP provides a mechanism for creating collections of variables in three types of data structures, which we refer to as containers.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The three types are Arrays, DenseAxisArrays, and SparseAxisArrays. We explain each of these in the following.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"tip: Tip\nYou can read more about containers in the Containers section.","category":"page"},{"location":"manual/variables/#Arrays","page":"Variables","title":"Arrays","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"We have already seen the creation of an array of JuMP variables with the x[1:2] syntax. This can be extended to create multi-dimensional arrays of JuMP variables. For example:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2])\n2×2 Matrix{VariableRef}:\n x[1,1]  x[1,2]\n x[2,1]  x[2,2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Arrays of JuMP variables can be indexed and sliced as follows:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> x[1, 2]\nx[1,2]\n\njulia> x[2, :]\n2-element Vector{VariableRef}:\n x[2,1]\n x[2,2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Variable bounds can depend upon the indices:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:2, j=1:2] >= 2i + j)\n2×2 Matrix{VariableRef}:\n x[1,1]  x[1,2]\n x[2,1]  x[2,2]\n\njulia> lower_bound.(x)\n2×2 Matrix{Float64}:\n 3.0  4.0\n 5.0  6.0","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"JuMP will form an Array of JuMP variables when it can determine at compile time that the indices are one-based integer ranges. Therefore x[1:b] will create an Array of JuMP variables, but x[a:b] will not. If JuMP cannot determine that the indices are one-based integer ranges (for example, in the case of x[a:b]), JuMP will create a DenseAxisArray instead.","category":"page"},{"location":"manual/variables/#variable_jump_arrays","page":"Variables","title":"DenseAxisArrays","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"We often want to create arrays where the indices are not one-based integer ranges. For example, we may want to create a variable indexed by the name of a product or a location. The syntax is the same as that above, except with an arbitrary vector as an index as opposed to a one-based range. The biggest difference is that instead of returning an Array of JuMP variables, JuMP will return a DenseAxisArray. For example:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, [:A,:B]])\n2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:\n    Dimension 1, Base.OneTo(2)\n    Dimension 2, [:A, :B]\nAnd data, a 2×2 Matrix{VariableRef}:\n x[1,A]  x[1,B]\n x[2,A]  x[2,B]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"DenseAxisArrays can be indexed and sliced as follows:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> x[1, :A]\nx[1,A]\n\njulia> x[2, :]\n1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:\n    Dimension 1, [:A, :B]\nAnd data, a 2-element Vector{VariableRef}:\n x[2,A]\n x[2,B]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Bounds can depend upon indices:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[i=2:3, j=1:2:3] >= 0.5i + j)\n2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:\n    Dimension 1, 2:3\n    Dimension 2, 1:2:3\nAnd data, a 2×2 Matrix{VariableRef}:\n x[2,1]  x[2,3]\n x[3,1]  x[3,3]\n\njulia> lower_bound.(x)\n2-dimensional DenseAxisArray{Float64,2,...} with index sets:\n    Dimension 1, 2:3\n    Dimension 2, 1:2:3\nAnd data, a 2×2 Matrix{Float64}:\n 2.0  4.0\n 2.5  4.5","category":"page"},{"location":"manual/variables/#variable_sparseaxisarrays","page":"Variables","title":"SparseAxisArrays","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The third container type that JuMP natively supports is SparseAxisArray. These arrays are created when the indices do not form a rectangular set. For example, this applies when indices have a dependence upon previous indices (called triangular indexing). JuMP supports this as follows:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:2, j=i:2])\nJuMP.Containers.SparseAxisArray{VariableRef, 2, Tuple{Int64, Int64}} with 3 entries:\n  [1, 1]  =  x[1,1]\n  [1, 2]  =  x[1,2]\n  [2, 2]  =  x[2,2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"We can also conditionally create variables via a JuMP-specific syntax. This syntax appends a comparison check that depends upon the named indices and is separated from the indices by a semi-colon (;). For example:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[i=1:4; mod(i, 2)==0])\nJuMP.Containers.SparseAxisArray{VariableRef, 1, Tuple{Int64}} with 2 entries:\n  [2]  =  x[2]\n  [4]  =  x[4]","category":"page"},{"location":"manual/variables/#Performance-considerations","page":"Variables","title":"Performance considerations","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"When using the semi-colon as a filter, JuMP iterates over all indices and evaluates the conditional for each combination. If there are many index dimensions and a large amount of sparsity, this can be inefficient.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"For example:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> N = 10\n10\n\njulia> S = [(1, 1, 1), (N, N, N)]\n2-element Vector{Tuple{Int64, Int64, Int64}}:\n (1, 1, 1)\n (10, 10, 10)\n\njulia> @time @variable(model, x1[i=1:N, j=1:N, k=1:N; (i, j, k) in S])\n  0.203861 seconds (392.22 k allocations: 23.977 MiB, 99.10% compilation time)\nJuMP.Containers.SparseAxisArray{VariableRef, 3, Tuple{Int64, Int64, Int64}} with 2 entries:\n  [1, 1, 1   ]  =  x1[1,1,1]\n  [10, 10, 10]  =  x1[10,10,10]\n\njulia> @time @variable(model, x2[S])\n  0.045407 seconds (65.24 k allocations: 3.771 MiB, 99.15% compilation time)\n1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:\n    Dimension 1, [(1, 1, 1), (10, 10, 10)]\nAnd data, a 2-element Vector{VariableRef}:\n x2[(1, 1, 1)]\n x2[(10, 10, 10)]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The first option is slower because it is equivalent to:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> x1 = Dict{NTuple{3,Int},VariableRef}()\nDict{Tuple{Int64, Int64, Int64}, VariableRef}()\n\njulia> for i in 1:N\n           for j in 1:N\n               for k in 1:N\n                   if (i, j, k) in S\n                       x1[i, j, k] = @variable(model, base_name = \"x1[$i,$j,$k]\")\n                   end\n               end\n           end\n       end\n\njulia> x1\nDict{Tuple{Int64, Int64, Int64}, VariableRef} with 2 entries:\n  (1, 1, 1)    => x1[1,1,1]\n  (10, 10, 10) => x1[10,10,10]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"If performance is a concern, explicitly construct the set of indices instead of using the filtering syntax.","category":"page"},{"location":"manual/variables/#variable_forcing","page":"Variables","title":"Forcing the container type","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"When creating a container of JuMP variables, JuMP will attempt to choose the tightest container type that can store the JuMP variables. Thus, it will prefer to create an Array before a DenseAxisArray and a DenseAxisArray before a SparseAxisArray. However, because this happens at compile time, JuMP does not always make the best choice. To illustrate this, consider the following example:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> A = 1:2\n1:2\n\njulia> @variable(model, x[A])\n1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:\n    Dimension 1, 1:2\nAnd data, a 2-element Vector{VariableRef}:\n x[1]\n x[2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Since the value (and type) of A is unknown at parsing time, JuMP is unable to infer that A is a one-based integer range. Therefore, JuMP creates a DenseAxisArray, even though it could store these two variables in a standard one-dimensional Array.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"We can share our knowledge that it is possible to store these JuMP variables as an array by setting the container keyword:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> @variable(model, y[A], container=Array)\n2-element Vector{VariableRef}:\n y[1]\n y[2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"JuMP now creates a vector of JuMP variables instead of a DenseAxisArray. Choosing an invalid container type will throw an error.","category":"page"},{"location":"manual/variables/#User-defined-containers","page":"Variables","title":"User-defined containers","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"In addition to the built-in container types, you can create your own collections of JuMP variables.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"tip: Tip\nThis is a point that users often overlook: you are not restricted to the built-in container types in JuMP.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"For example, the following code creates a dictionary with symmetric matrices as the values:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> variables = Dict{Symbol,Array{VariableRef,2}}(\n           key => @variable(model, [1:2, 1:2], Symmetric, base_name = \"$(key)\")\n           for key in [:A, :B]\n       )\nDict{Symbol, Matrix{VariableRef}} with 2 entries:\n  :A => [A[1,1] A[1,2]; A[1,2] A[2,2]]\n  :B => [B[1,1] B[1,2]; B[1,2] B[2,2]]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Another common scenario is a request to add variables to existing containers, for example:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"using JuMP\nmodel = Model()\n@variable(model, x[1:2] >= 0)\n# Later I want to add\n@variable(model, x[3:4] >= 0)","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"This is not possible with the built-in JuMP container types. However, you can use regular Julia types instead:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> x = model[:x] = @variable(model, [1:2], lower_bound = 0, base_name = \"x\")\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> append!(x, @variable(model, [1:2], lower_bound = 0, base_name = \"y\"));\n\njulia> model[:x]\n4-element Vector{VariableRef}:\n x[1]\n x[2]\n y[1]\n y[2]","category":"page"},{"location":"manual/variables/#Semidefinite-variables","page":"Variables","title":"Semidefinite variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Declare a square matrix of JuMP variables to be positive semidefinite by passing PSD as an optional positional argument:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2], PSD)\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n x[1,1]  x[1,2]\n x[1,2]  x[2,2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"This will ensure that x is symmetric, and that all of its eigenvalues are nonnegative.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"note: Note\nx must be a square 2-dimensional Array of JuMP variables; it cannot be a DenseAxisArray or a SparseAxisArray.","category":"page"},{"location":"manual/variables/#Symmetric-variables","page":"Variables","title":"Symmetric variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Declare a square matrix of JuMP variables to be symmetric (but not necessarily positive semidefinite) by passing Symmetric  as an optional positional argument:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2], Symmetric)\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n x[1,1]  x[1,2]\n x[1,2]  x[2,2]","category":"page"},{"location":"manual/variables/#variables","page":"Variables","title":"The @variables macro","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"If you have many @variable calls, JuMP provides the macro @variables that can improve readability:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variables(model, begin\n           x\n           y[i=1:2] >= i, (start = i, base_name = \"Y_$i\")\n           z, Bin\n       end)\n(x, VariableRef[Y_1[1], Y_2[2]], z)\n\njulia> print(model)\nFeasibility\nSubject to\n Y_1[1] ≥ 1\n Y_2[2] ≥ 2\n z binary","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The @variables macro returns a tuple of the variables that were defined.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"note: Note\nKeyword arguments must be contained within parentheses.","category":"page"},{"location":"manual/variables/#jump_variables_on_creation","page":"Variables","title":"Variables constrained on creation","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"All uses of the @variable macro documented so far translate into separate calls for variable creation and the adding of any bound or integrality constraints.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"For example, @variable(model, x >= 0, Int), is equivalent to:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x)\nx\n\njulia> set_lower_bound(x, 0.0)\n\njulia> set_integer(x)","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Importantly, the bound and integrality constraints are added after the variable has been created.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"However, some solvers require a set specifying the variable domain to be given when the variable is first created. We say that these variables are constrained on creation.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Use in within @variable to access the special syntax for constraining variables on creation.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"For example, the following creates a vector of variables that belong to the SecondOrderCone:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, y[1:3] in SecondOrderCone())\n3-element Vector{VariableRef}:\n y[1]\n y[2]\n y[3]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"For contrast, the standard syntax is as follows:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> @variable(model, x[1:3])\n3-element Vector{VariableRef}:\n x[1]\n x[2]\n x[3]\n\njulia> @constraint(model, x in SecondOrderCone())\n[x[1], x[2], x[3]] ∈ MathOptInterface.SecondOrderCone(3)","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"An alternate syntax to x in Set is to use the set keyword of @variable. This is most useful when creating anonymous variables:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> x = @variable(model, [1:3], set = SecondOrderCone())\n3-element Vector{VariableRef}:\n _[1]\n _[2]\n _[3]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"note: Note\nYou cannot delete the constraint associated with a variable constrained on creation.","category":"page"},{"location":"manual/variables/#Example:-positive-semidefinite-variables","page":"Variables","title":"Example: positive semidefinite variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"An alternative to the syntax in Semidefinite variables, declare a matrix of JuMP variables to be positive semidefinite using PSDCone:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2] in PSDCone())\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n x[1,1]  x[1,2]\n x[1,2]  x[2,2]","category":"page"},{"location":"manual/variables/#Example:-symmetric-variables","page":"Variables","title":"Example: symmetric variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"As an alternative to the syntax in Symmetric variables, declare a matrix of JuMP variables to be symmetric using SymmetricMatrixSpace:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2] in SymmetricMatrixSpace())\n2×2 LinearAlgebra.Symmetric{VariableRef, Matrix{VariableRef}}:\n x[1,1]  x[1,2]\n x[1,2]  x[2,2]","category":"page"},{"location":"manual/variables/#Example:-skew-symmetric-variables","page":"Variables","title":"Example: skew-symmetric variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Declare a matrix of JuMP variables to be skew-symmetric using SkewSymmetricMatrixSpace:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2] in SkewSymmetricMatrixSpace())\n2×2 Matrix{AffExpr}:\n 0        x[1,2]\n -x[1,2]  0","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"note: Note\nEven though x is a 2 by 2 matrix, only one decision variable is added to model; the remaining elements in x are linear transformations of the single variable.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Because the returned matrix x is Matrix{AffExpr}, you cannot use variable-related functions on its elements:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> set_lower_bound(x[1, 2], 0.0)\nERROR: MethodError: no method matching set_lower_bound(::AffExpr, ::Float64)\n[...]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Instead, you can convert an upper-triangular elements to a variable as follows:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> to_variable(x::AffExpr) = first(keys(x.terms))\nto_variable (generic function with 1 method)\n\njulia> to_variable(x[1, 2])\nx[1,2]\n\njulia> set_lower_bound(to_variable(x[1, 2]), 0.0)","category":"page"},{"location":"manual/variables/#Example:-Hermitian-positive-semidefinite-variables","page":"Variables","title":"Example: Hermitian positive semidefinite variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Declare a matrix of JuMP variables to be Hermitian positive semidefinite using HermitianPSDCone:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, H[1:2, 1:2] in HermitianPSDCone())\n2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(H[1,1])                    real(H[1,2]) + imag(H[1,2]) im\n real(H[1,2]) - imag(H[1,2]) im  real(H[2,2])","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"This adds 4 real variables in the MOI.HermitianPositiveSemidefiniteConeTriangle:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> first(all_constraints(model, Vector{VariableRef}, MOI.HermitianPositiveSemidefiniteConeTriangle))\n[real(H[1,1]), real(H[1,2]), real(H[2,2]), imag(H[1,2])] ∈ MathOptInterface.HermitianPositiveSemidefiniteConeTriangle(2)","category":"page"},{"location":"manual/variables/#Example:-Hermitian-variables","page":"Variables","title":"Example: Hermitian variables","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Declare a matrix of JuMP variables to be Hermitian using the Hermitian tag:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2], Hermitian)\n2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(x[1,1])                    real(x[1,2]) + imag(x[1,2]) im\n real(x[1,2]) - imag(x[1,2]) im  real(x[2,2])","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"This is equivalent to declaring the variable in HermitianMatrixSpace:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2, 1:2] in HermitianMatrixSpace())\n2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(x[1,1])                    real(x[1,2]) + imag(x[1,2]) im\n real(x[1,2]) - imag(x[1,2]) im  real(x[2,2])","category":"page"},{"location":"manual/variables/#Why-use-variables-constrained-on-creation?","page":"Variables","title":"Why use variables constrained on creation?","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"For most users, it does not matter if you use the constrained on creation syntax. Therefore, use whatever syntax you find most convenient.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"However, if you use direct_model, you may be forced to use the constrained on creation syntax.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The technical difference between variables constrained on creation and the standard JuMP syntax is that variables constrained on creation calls MOI.add_constrained_variables, while the standard JuMP syntax calls MOI.add_variables and then MOI.add_constraint.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Consult the implementation of solver package you are using to see if your solver requires MOI.add_constrained_variables.","category":"page"},{"location":"manual/variables/#Parameters","page":"Variables","title":"Parameters","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Some solvers have explicit support for parameters, which are constants in the model that can be efficiently updated between solves.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"JuMP implements parameters by a decision variable constrained on creation to the Parameter set.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, p[i = 1:2] in Parameter(i))\n2-element Vector{VariableRef}:\n p[1]\n p[2]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Create anonymous parameters using the set keyword:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> anon_parameter = @variable(model, set = Parameter(1.0))\n_[4]","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Use parameter_value and set_parameter_value to query or update the value of a parameter.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> parameter_value.(p)\n2-element Vector{Float64}:\n 1.0\n 2.0\n\njulia> set_parameter_value(p[2], 3.0)\n\njulia> parameter_value.(p)\n2-element Vector{Float64}:\n 1.0\n 3.0","category":"page"},{"location":"manual/variables/#Limitations","page":"Variables","title":"Limitations","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Parameters are implemented as decision variables belonging to the Parameter set. If the solver supports the MOI.Parameter set, it may decide to replace all instances of the parameter variable by the associated constant. If the solver does not support parameters, it will add the parameter as a decision variable with fixed bounds.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"The most important implication of this design is that JuMP treats a parameter multiplied by a decision variable as a quadratic expression, even though it is equivalent to a linear expression.","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"julia> model = Model();\n\njulia> @variable(model, x);\n\njulia> @variable(model, p in Parameter(2));\n\njulia> px = @expression(model, p * x)\np*x\n\njulia> typeof(px)\nQuadExpr (alias for GenericQuadExpr{Float64, GenericVariableRef{Float64}})","category":"page"},{"location":"manual/variables/#When-to-use-a-parameter","page":"Variables","title":"When to use a parameter","text":"","category":"section"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Parameters are most useful when solving nonlinear models in a sequence:","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"using JuMP, Ipopt\nmodel = Model(Ipopt.Optimizer);\nset_silent(model)\n@variable(model, x)\n@variable(model, p in Parameter(1.0))\n@objective(model, Min, (x - p)^2)\noptimize!(model)\nvalue(x)\nset_parameter_value(p, 5.0)\noptimize!(model)\nvalue(x)","category":"page"},{"location":"manual/variables/","page":"Variables","title":"Variables","text":"Using parameters can be faster than creating a new model from scratch with updated data because JuMP is able to avoid repeating a number of steps in processing the model before handing it off to the solver.","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/FileFormats/overview.md\"","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/FileFormats/overview/#The-FileFormats-submodule","page":"Overview","title":"The FileFormats submodule","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The FileFormats module provides functions for reading and writing MOI models using write_to_file and read_from_file.","category":"page"},{"location":"moi/submodules/FileFormats/overview/#Supported-file-types","page":"Overview","title":"Supported file types","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"You must read and write files to a FileFormats.Model object. Specific the file-type by passing a FileFormats.FileFormat enum. For example:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The Conic Benchmark Format","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_CBF)\nA Conic Benchmark Format (CBF) model","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The LP file format","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_LP)\nA .LP-file model","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The MathOptFormat file format","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MOF)\nA MathOptFormat Model","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The MPS file format","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MPS)\nA Mathematical Programming System (MPS) model","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The NL file format","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_NL)\nAn AMPL (.nl) model","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The REW file format","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_REW)\nA Mathematical Programming System (MPS) model","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"Note that the REW format is identical to the MPS file format, except that all names are replaced with generic identifiers.","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"The SDPA file format","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> model = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_SDPA)\nA SemiDefinite Programming Algorithm Format (SDPA) model","category":"page"},{"location":"moi/submodules/FileFormats/overview/#Write-to-file","page":"Overview","title":"Write to file","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"To write a model src to a MathOptFormat file, use:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> src = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> MOI.add_variable(src)\nMOI.VariableIndex(1)\n\njulia> dest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MOF)\nA MathOptFormat Model\n\njulia> MOI.copy_to(dest, src)\nMathOptInterface.Utilities.IndexMap with 1 entry:\n  MOI.VariableIndex(1) => MOI.VariableIndex(1)\n\njulia> MOI.write_to_file(dest, \"file.mof.json\")\n\njulia> print(read(\"file.mof.json\", String))\n{\n  \"name\": \"MathOptFormat Model\",\n  \"version\": {\n    \"major\": 1,\n    \"minor\": 5\n  },\n  \"variables\": [\n    {\n      \"name\": \"x1\"\n    }\n  ],\n  \"objective\": {\n    \"sense\": \"feasibility\"\n  },\n  \"constraints\": []\n}","category":"page"},{"location":"moi/submodules/FileFormats/overview/#Read-from-file","page":"Overview","title":"Read from file","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"To read a MathOptFormat file, use:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> dest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MOF)\nA MathOptFormat Model\n\njulia> MOI.read_from_file(dest, \"file.mof.json\")\n\njulia> MOI.get(dest, MOI.ListOfVariableIndices())\n1-element Vector{MathOptInterface.VariableIndex}:\n MOI.VariableIndex(1)\n\njulia> rm(\"file.mof.json\")  # Clean up after ourselves.","category":"page"},{"location":"moi/submodules/FileFormats/overview/#Detecting-the-file-type-automatically","page":"Overview","title":"Detecting the file-type automatically","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"Instead of the format keyword, you can also use the filename keyword argument to FileFormats.Model. This will attempt to automatically guess the format from the file extension. For example:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> src = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> dest = MOI.FileFormats.Model(filename = \"file.cbf.gz\")\nA Conic Benchmark Format (CBF) model\n\njulia> MOI.copy_to(dest, src)\nMathOptInterface.Utilities.IndexMap()\n\njulia> MOI.write_to_file(dest, \"file.cbf.gz\")\n\njulia> src_2 = MOI.FileFormats.Model(filename = \"file.cbf.gz\")\nA Conic Benchmark Format (CBF) model\n\njulia> src = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> dest = MOI.FileFormats.Model(filename = \"file.cbf.gz\")\nA Conic Benchmark Format (CBF) model\n\njulia> MOI.copy_to(dest, src)\nMathOptInterface.Utilities.IndexMap()\n\njulia> MOI.write_to_file(dest, \"file.cbf.gz\")\n\njulia> src_2 = MOI.FileFormats.Model(filename = \"file.cbf.gz\")\nA Conic Benchmark Format (CBF) model\n\njulia> MOI.read_from_file(src_2, \"file.cbf.gz\")\n\njulia> rm(\"file.cbf.gz\")  # Clean up after ourselves.","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"Note how the compression format (GZip) is also automatically detected from the filename.","category":"page"},{"location":"moi/submodules/FileFormats/overview/#Unsupported-constraints","page":"Overview","title":"Unsupported constraints","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"In some cases src may contain constraints that are not supported by the file format (for example, the CBF format supports integer variables but not binary). If so, copy src to a bridged model using Bridges.full_bridge_optimizer:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"src = MOI.Utilities.Model{Float64}()\nx = MOI.add_variable(model)\nMOI.add_constraint(model, x, MOI.ZeroOne())\ndest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_CBF)\nbridged = MOI.Bridges.full_bridge_optimizer(dest, Float64)\nMOI.copy_to(bridged, src)\nMOI.write_to_file(dest, \"my_model.cbf\")","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"note: Note\nEven after bridging, it may still not be possible to write the model to file because of unsupported constraints (for example, PSD variables in the LP file format).","category":"page"},{"location":"moi/submodules/FileFormats/overview/#Read-and-write-to-io","page":"Overview","title":"Read and write to io","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"In addition to write_to_file and read_from_file, you can read and write directly from IO streams using Base.write and Base.read!:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> src = MOI.Utilities.Model{Float64}()\nMOIU.Model{Float64}\n\njulia> dest = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MPS)\nA Mathematical Programming System (MPS) model\n\njulia> MOI.copy_to(dest, src)\nMathOptInterface.Utilities.IndexMap()\n\njulia> io = IOBuffer();\n\njulia> write(io, dest)\n\njulia> seekstart(io);\n\njulia> src_2 = MOI.FileFormats.Model(format = MOI.FileFormats.FORMAT_MPS)\nA Mathematical Programming System (MPS) model\n\njulia> read!(io, src_2);","category":"page"},{"location":"moi/submodules/FileFormats/overview/#Validating-MOF-files","page":"Overview","title":"Validating MOF files","text":"","category":"section"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"MathOptFormat files are governed by a schema. Use JSONSchema.jl to check if a .mof.json file satisfies the schema.","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"First, construct the schema object as follows:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> import JSON, JSONSchema\n\njulia> schema = JSONSchema.Schema(JSON.parsefile(MOI.FileFormats.MOF.SCHEMA_PATH))\nA JSONSchema","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"Then, check if a model file is valid using isvalid:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> good_model = JSON.parse(\"\"\"\n       {\n         \"version\": {\n           \"major\": 1,\n           \"minor\": 5\n         },\n         \"variables\": [{\"name\": \"x\"}],\n         \"objective\": {\"sense\": \"feasibility\"},\n         \"constraints\": []\n       }\n       \"\"\");\n\njulia> isvalid(schema, good_model)\ntrue","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"If we construct an invalid file, for example by mis-typing name as NaMe, the validation fails:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> bad_model = JSON.parse(\"\"\"\n       {\n         \"version\": {\n           \"major\": 1,\n           \"minor\": 5\n         },\n         \"variables\": [{\"NaMe\": \"x\"}],\n         \"objective\": {\"sense\": \"feasibility\"},\n         \"constraints\": []\n       }\n       \"\"\");\n\njulia> isvalid(schema, bad_model)\nfalse","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"Use JSONSchema.validate to obtain more insight into why the validation failed:","category":"page"},{"location":"moi/submodules/FileFormats/overview/","page":"Overview","title":"Overview","text":"julia> JSONSchema.validate(schema, bad_model)\nValidation failed:\npath:         [variables][1]\ninstance:     Dict{String, Any}(\"NaMe\" => \"x\")\nschema key:   required\nschema value: Any[\"name\"]","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"EditURL = \"https://github.com/lanl-ansi/Alpine.jl/blob/32757b2568acfe71257e41ed4122a120717a6460/README.md\"","category":"page"},{"location":"packages/Alpine/#Alpine,-a-global-solver-for-non-convex-MINLPs","page":"lanl-ansi/Alpine.jl","title":"Alpine, a global solver for non-convex MINLPs","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"(Image: CI) (Image: codecov) (Image: Documentation)  (Image: version)","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"ALPINE (glob(AL) o(P)timization for mixed-(I)nteger programs with (N)onlinear (E)quations), is a novel global optimization solver that uses an adaptive, piecewise convexification scheme and constraint programming methods to solve non-convex Mixed-Integer Non-Linear Programs (MINLPs) efficiently. MINLPs are typically \"hard\" optimization problems which appear in numerous applications (see MINLPLib.jl). ","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Alpine is entirely built upon JuMP and MathOptInterface in Julia, which provides incredible flexibility for usage and further development.","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Alpine globally solves a given MINLP by:","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Analyzing the problem's expressions (objective & constraints) and applies appropriate convex relaxations and polyhedral outer-approximations\nPerforming sequential optimization-based bound tightening (OBBT) and an iterative MIP-based adaptive partitioning scheme via piecewise polyhedral relaxations with a guarantee of global convergence","category":"page"},{"location":"packages/Alpine/#Installation","page":"lanl-ansi/Alpine.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Install Alpine using the Julia package manager:","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"import Pkg\nPkg.add(\"Alpine\")","category":"page"},{"location":"packages/Alpine/#Use-with-JuMP","page":"lanl-ansi/Alpine.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Use Alpine with JuMP as follows:","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"using JuMP, Alpine, Ipopt, HiGHS\nipopt = optimizer_with_attributes(Ipopt.Optimizer, \"print_level\" => 0)\nhighs = optimizer_with_attributes(HiGHS.Optimizer, \"output_flag\" => false)\nmodel = Model(\n    optimizer_with_attributes(\n        Juniper.Optimizer,\n        \"nlp_solver\" => ipopt,\n        \"mip_solver\" => highs,\n    ),\n)","category":"page"},{"location":"packages/Alpine/#Documentation","page":"lanl-ansi/Alpine.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"For more details, see the online documentation.","category":"page"},{"location":"packages/Alpine/#Support-problem-types","page":"lanl-ansi/Alpine.jl","title":"Support problem types","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Alpine can currently handle MINLPs with polynomials in constraints and/or in the objective. Currently, there is no support for exponential cones and Positive Semi-Definite (PSD) cones in MINLPs. Alpine is also a good fit for subsets of the MINLP family, for example, Mixed-Integer Quadratically Constrained Quadratic Programs (MIQCQPs), Non-Linear Programs (NLPs), etc.","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"For more details, check out this video on Alpine.jl at JuMP-dev 2018.","category":"page"},{"location":"packages/Alpine/#Underlying-solvers","page":"lanl-ansi/Alpine.jl","title":"Underlying solvers","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Though the MIP-based bounding algorithm implemented in Alpine is quite involved, most of the computational bottleneck arises in the underlying MIP solvers. Since every iteration of Alpine solves an MIP sub-problem, which is typically a convex MILP/MIQCQP, Alpine's run time heavily depends on the run-time of these solvers. For the best performance of Alpine, we recommend using the commercial solver Gurobi, which is available free for academic purposes. However, due to the flexibility offered by JuMP, the following MIP and NLP solvers are supported in Alpine: ","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Solver Julia Package\nCPLEX CPLEX.jl\nCbc Cbc.jl\nGurobi Gurobi.jl\nIpopt Ipopt.jl\nBonmin Bonmin.jl\nArtelys KNITRO KNITRO.jl\nXpress Xpress.jl\nHiGHS HiGHS.jl","category":"page"},{"location":"packages/Alpine/#Bug-reports-and-support","page":"lanl-ansi/Alpine.jl","title":"Bug reports and support","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"Please report any issues via the GitHub issue tracker. All types of issues are welcome and encouraged; this includes bug reports, documentation typos, feature requests, etc. ","category":"page"},{"location":"packages/Alpine/#Challenging-Problems","page":"lanl-ansi/Alpine.jl","title":"Challenging Problems","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"We are seeking out hard benchmark instances for MINLPs. Please get in touch either by opening an issue or privately if you would like to share any hard instances.","category":"page"},{"location":"packages/Alpine/#Citing-Alpine","page":"lanl-ansi/Alpine.jl","title":"Citing Alpine","text":"","category":"section"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"If you find Alpine useful in your work, we kindly request that you cite the following papers (PDF, PDF)","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"@article{alpine_JOGO2019,\n  title = {An adaptive, multivariate partitioning algorithm for global optimization of nonconvex programs},\n  author = {Nagarajan, Harsha and Lu, Mowen and Wang, Site and Bent, Russell and Sundar, Kaarthik},\n  journal = {Journal of Global Optimization},\n  year = {2019},\n  issn = {1573-2916},\n  doi = {10.1007/s10898-018-00734-1},\n}\n\n@inproceedings{alpine_CP2016,\n  title = {Tightening {McCormick} relaxations for nonlinear programs via dynamic multivariate partitioning},\n  author = {Nagarajan, Harsha and Lu, Mowen and Yamangil, Emre and Bent, Russell},\n  booktitle = {International Conference on Principles and Practice of Constraint Programming},\n  pages = {369--387},\n  year = {2016},\n  organization = {Springer},\n  doi = {10.1007/978-3-319-44953-1_24},\n}","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"If you find the underlying piecewise polyhedral formulations implemented in Alpine useful in your work, we kindly request that you cite the following papers (link-1, link-2): ","category":"page"},{"location":"packages/Alpine/","page":"lanl-ansi/Alpine.jl","title":"lanl-ansi/Alpine.jl","text":"@article{alpine_ORL2021,\n  title = {Piecewise polyhedral formulations for a multilinear term},\n  author = {Sundar, Kaarthik and Nagarajan, Harsha and Linderoth, Jeff and Wang, Site and Bent, Russell},\n  journal = {Operations Research Letters},\n  volume = {49},\n  number = {1},\n  pages = {144--149},\n  year = {2021},\n  publisher = {Elsevier}\n}\n\n@article{alpine_OptOnline2022,\n  title={Piecewise Polyhedral Relaxations of Multilinear Optimization},\n  author={Kim, Jongeun and Richard, Jean-Philippe P. and Tawarmalani, Mohit},\n  eprinttype={Optimization Online},\n  date={2022}\n}","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"EditURL = \"https://github.com/odow/NEOSServer.jl/blob/v1.0.2/README.md\"","category":"page"},{"location":"packages/NEOSServer/#NEOSServer.jl","page":"odow/NEOSServer.jl","title":"NEOSServer.jl","text":"","category":"section"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"NEOSServer.jl is a wrapper for the NEOS Server, a free internet-based service for solving numerical optimization problems.","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"See here for the full list of solvers and input formats that NEOS supports.","category":"page"},{"location":"packages/NEOSServer/#License","page":"odow/NEOSServer.jl","title":"License","text":"","category":"section"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"NEOSServer.jl is licensed under the MIT License.","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"However, use of the NEOS Server requires you to comply with NEOS Server terms of use. In particular, the commercial solvers are to be used solely for academic, non-commercial research purposes.","category":"page"},{"location":"packages/NEOSServer/#Installation","page":"odow/NEOSServer.jl","title":"Installation","text":"","category":"section"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"Install NEOSServer.jl using the package manager:","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"import Pkg\nPkg.add(\"NEOSServer\")","category":"page"},{"location":"packages/NEOSServer/#The-NEOS-API","page":"odow/NEOSServer.jl","title":"The NEOS API","text":"","category":"section"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"This package contains an interface for the NEOS XML-RPC API.","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"The following example shows how you can interact with the API. Wrapped XML-RPC functions begin with neos_ and are exported.","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"using NEOSServer\n\n# Create a server. You must supply a valid email:\nserver = NEOSServer.Server(\"me@mydomain.com\")\n\n# Print the NEOS welcome message:\nprintln(neos_welcome(server))\n\n# Get an XML template:\nxml_string = neos_getSolverTemplate(server, \"milp\", \"Cbc\", \"AMPL\")\n\n# Modify template with problem data...\n\n# Submit the XML job to NEOS:\njob = neos_submitJob(server, xml_string)\n\n# Get the status of the Job from NEOS:\nstatus = neos_getJobStatus(server, job)\n\n# Get the final results:\nresults = neos_getFinalResults(server, job)","category":"page"},{"location":"packages/NEOSServer/#Use-with-JuMP","page":"odow/NEOSServer.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"Use NEOSServer.jl with JuMP as follows:","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"using JuMP, NEOSServer\n\nmodel = Model() do\n    NEOSServer.Optimizer(email=\"me@mydomain.com\", solver=\"Ipopt\")\nend","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"Note: NEOSServer.Optimizer is limited to the following solvers:","category":"page"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"\"CPLEX\"\n\"FICO-Xpress\"\n\"Ipopt\"\n\"Knitro\"\n\"MOSEK\"\n\"OCTERACT\"\n\"SNOPT\"","category":"page"},{"location":"packages/NEOSServer/#NEOS-Limits","page":"odow/NEOSServer.jl","title":"NEOS Limits","text":"","category":"section"},{"location":"packages/NEOSServer/","page":"odow/NEOSServer.jl","title":"odow/NEOSServer.jl","text":"NEOS currently limits jobs to an 8 hour time limit, 3 GB of memory, and a 16 MB submission file. If your model exceeds these limits, NEOSServer.jl may be unable to return useful information to the user.","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"EditURL = \"https://github.com/jump-dev/SDPA.jl/blob/v0.5.0/README.md\"","category":"page"},{"location":"packages/SDPA/#SDPA.jl","page":"jump-dev/SDPA.jl","title":"SDPA.jl","text":"","category":"section"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"SDPA.jl is a wrapper for the SDPA semidefinite programming solver in double precision floating point arithmetic.","category":"page"},{"location":"packages/SDPA/#Affiliation","page":"jump-dev/SDPA.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"This wrapper is maintained by the JuMP community and is not a product of the SDPA developers.","category":"page"},{"location":"packages/SDPA/#License","page":"jump-dev/SDPA.jl","title":"License","text":"","category":"section"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"SDPA.jl is licensed under the MIT License.","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"The underlying solver, SDPA is licensed under the GPL v2 license.","category":"page"},{"location":"packages/SDPA/#Installation","page":"jump-dev/SDPA.jl","title":"Installation","text":"","category":"section"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"Install SDPA using Pkg.add:","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"import Pkg\nPkg.add(\"SDPA\")","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"In addition to installing the SDPA.jl package, this will also download and install the SDPA binaries. (You do not need to install SDPA separately.)","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"If you see an error similar to:","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"INFO: Precompiling module GZip.\nERROR: LoadError: LoadError: error compiling anonymous: could not load library \"libz\"","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"please see GZip.jl#54 or Flux.jl#343. In particular, in Ubuntu this issue may be resolved by running","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"sudo apt-get install zlib1g-dev","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"See SDPAFamily for the other solvers, SDPA-GMP, SDPA-DD, and SDPA-QD of the family.","category":"page"},{"location":"packages/SDPA/#Use-with-JuMP","page":"jump-dev/SDPA.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"using JuMP, SDPA\nmodel = Model(SDPA.Optimizer)\nset_attribute(model, \"Mode\", SDPA.PARAMETER_DEFAULT)","category":"page"},{"location":"packages/SDPA/#MathOptInterface-API","page":"jump-dev/SDPA.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"The SDPA optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"MOI.Nonnegatives\nMOI.PositiveSemidefiniteConeTriangle","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"MOI.ObjectiveSense()","category":"page"},{"location":"packages/SDPA/#Options","page":"jump-dev/SDPA.jl","title":"Options","text":"","category":"section"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"SDPA has three modes that give default value to all ten parameters.","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"The following table gives the default values for each parameter and mode.","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"Parameter PARAMETER_DEFAULT PARAMETER_UNSTABLE_BUT_FAST PARAMETER_STABLE_BUT_SLOW\nMaxIteration 100 100 1000\nEpsilonStar 1.0e-7 1.0e-7 1.0e-7\nLambdaStar 1.0e+2 1.0e+2 1.0e+4\nOmegaStar 2.0 2.0 2.0\nLowerBound 1.0e+5 1.0e+5 1.0e+5\nUpperBound 1.0e+5 1.0e+5 1.0e+5\nBetaStar 0.1 0.01 0.1\nBetaBar 0.2 0.02 0.3\nGammaStar 0.9 0.95 0.8\nEpsilonDash 1.0e-7 1.0e-7 1.0e-7","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"By default, we put SDPA in the SDPA.PARAMETER_DEFAULT mode.","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"Change the mode using the \"Mode\" option:","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"using JuMP, SDPA\nmodel = Model(SDPA.Optimizer)\nset_attribute(model, \"Mode\", SDPA.PARAMETER_STABLE_BUT_SLOW)","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"Note that the parameters are set in the order they are given, so you can set a mode and then modify parameters from this mode.","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"using JuMP, SDPA\nmodel = Model(SDPA.Optimizer)\nset_attribute(model, \"Mode\", SDPA.PARAMETER_STABLE_BUT_SLOW)\nset_attribute(model, \"MaxIteration\", 100)","category":"page"},{"location":"packages/SDPA/","page":"jump-dev/SDPA.jl","title":"jump-dev/SDPA.jl","text":"The choice of parameter mode has a large impact on the performance and stability of SDPA, and not necessarily in the way implied by the names of the modes; for example, PARAMETER_UNSTABLE_BUT_FAST can be more stable than the other modes for some problems. You should try each mode to see how it performs on your specific problem. See SDPA.jl#17 for more details.","category":"page"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/mathprogbase.md\"","category":"page"},{"location":"moi/tutorials/mathprogbase/#Transitioning-from-MathProgBase","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"","category":"section"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"MathOptInterface is a replacement for MathProgBase.jl. However, it is not a direct replacement.","category":"page"},{"location":"moi/tutorials/mathprogbase/#Transitioning-a-solver-interface","page":"Transitioning from MathProgBase","title":"Transitioning a solver interface","text":"","category":"section"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"MathOptInterface is more extensive than MathProgBase which may make its implementation seem daunting at first. There are however numerous utilities in MathOptInterface that the simplify implementation process.","category":"page"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"For more information, read Implementing a solver interface.","category":"page"},{"location":"moi/tutorials/mathprogbase/#Transitioning-the-high-level-functions","page":"Transitioning from MathProgBase","title":"Transitioning the high-level functions","text":"","category":"section"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"MathOptInterface doesn't provide replacements for the high-level interfaces in MathProgBase. We recommend you use JuMP as a modeling interface instead.","category":"page"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"tip: Tip\nIf you haven't used JuMP before, start with the tutorial Getting started with JuMP","category":"page"},{"location":"moi/tutorials/mathprogbase/#linprog","page":"Transitioning from MathProgBase","title":"linprog","text":"","category":"section"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"Here is one way of transitioning from linprog:","category":"page"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"using JuMP\n\nfunction linprog(c, A, sense, b, l, u, solver)\n    N = length(c)\n    model = Model(solver)\n    @variable(model, l[i] <= x[i=1:N] <= u[i])\n    @objective(model, Min, c' * x)\n    eq_rows, ge_rows, le_rows = sense .== '=', sense .== '>', sense .== '<'\n    @constraint(model, A[eq_rows, :] * x .== b[eq_rows])\n    @constraint(model, A[ge_rows, :] * x .>= b[ge_rows])\n    @constraint(model, A[le_rows, :] * x .<= b[le_rows])\n    optimize!(model)\n    return (\n        status = termination_status(model),\n        objval = objective_value(model),\n        sol = value.(x)\n    )\nend","category":"page"},{"location":"moi/tutorials/mathprogbase/#mixintprog","page":"Transitioning from MathProgBase","title":"mixintprog","text":"","category":"section"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"Here is one way of transitioning from mixintprog:","category":"page"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"using JuMP\n\nfunction mixintprog(c, A, rowlb, rowub, vartypes, lb, ub, solver)\n    N = length(c)\n    model = Model(solver)\n    @variable(model, lb[i] <= x[i=1:N] <= ub[i])\n    for i in 1:N\n        if vartypes[i] == :Bin\n            set_binary(x[i])\n        elseif vartypes[i] == :Int\n            set_integer(x[i])\n        end\n    end\n    @objective(model, Min, c' * x)\n    @constraint(model, rowlb .<= A * x .<= rowub)\n    optimize!(model)\n    return (\n        status = termination_status(model),\n        objval = objective_value(model),\n        sol = value.(x)\n    )\nend","category":"page"},{"location":"moi/tutorials/mathprogbase/#quadprog","page":"Transitioning from MathProgBase","title":"quadprog","text":"","category":"section"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"Here is one way of transitioning from quadprog:","category":"page"},{"location":"moi/tutorials/mathprogbase/","page":"Transitioning from MathProgBase","title":"Transitioning from MathProgBase","text":"using JuMP\n\nfunction quadprog(c, Q, A, rowlb, rowub, lb, ub, solver)\n    N = length(c)\n    model = Model(solver)\n    @variable(model, lb[i] <= x[i=1:N] <= ub[i])\n    @objective(model, Min, c' * x + 0.5 * x' * Q * x)\n    @constraint(model, rowlb .<= A * x .<= rowub)\n    optimize!(model)\n    return (\n        status = termination_status(model),\n        objval = objective_value(model),\n        sol = value.(x)\n    )\nend","category":"page"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Bridges/reference.md\"","category":"page"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Bridges/reference/#Bridges","page":"API Reference","title":"Bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/#AbstractBridge-API","page":"API Reference","title":"AbstractBridge API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.AbstractBridge\nBridges.added_constrained_variable_types\nBridges.added_constraint_types\nget(::Bridges.AbstractBridge, ::NumberOfVariables)\nget(::Bridges.AbstractBridge, ::ListOfVariableIndices)\nget(::Bridges.AbstractBridge, ::NumberOfConstraints)\nget(::Bridges.AbstractBridge, ::ListOfConstraintIndices)\nBridges.needs_final_touch\nBridges.final_touch\nBridges.bridging_cost","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.AbstractBridge","page":"API Reference","title":"MathOptInterface.Bridges.AbstractBridge","text":"abstract type AbstractBridge end\n\nAn abstract type representing a bridged constraint or variable in a MOI.Bridges.AbstractBridgeOptimizer.\n\nAll bridges must implement:\n\nadded_constrained_variable_types\nadded_constraint_types\nMOI.get(::AbstractBridge, ::MOI.NumberOfVariables)\nMOI.get(::AbstractBridge, ::MOI.ListOfVariableIndices)\nMOI.get(::AbstractBridge, ::MOI.NumberOfConstraints)\nMOI.get(::AbstractBridge, ::MOI.ListOfConstraintIndices)\n\nSubtypes of AbstractBridge may have additional requirements. Consult their docstrings for details.\n\nIn addition, all subtypes may optionally implement the following constraint attributes with the bridge in place of the constraint index:\n\nMOI.ConstraintDual\nMOI.ConstraintPrimal\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.added_constrained_variable_types","page":"API Reference","title":"MathOptInterface.Bridges.added_constrained_variable_types","text":"added_constrained_variable_types(\n    BT::Type{<:AbstractBridge},\n)::Vector{Tuple{Type}}\n\nReturn a list of the types of constrained variables that bridges of concrete type BT add.\n\nImplementation notes\n\nThis method depends only on the type of the bridge, not the runtime value. If the bridge may add a constrained variable, the type must be included in the return vector.\nIf the bridge adds a free variable via MOI.add_variable or MOI.add_variables, the return vector must include (MOI.Reals,).\n\nExample\n\njulia> MOI.Bridges.added_constrained_variable_types(\n           MOI.Bridges.Variable.NonposToNonnegBridge{Float64},\n       )\n1-element Vector{Tuple{Type}}:\n (MathOptInterface.Nonnegatives,)\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.added_constraint_types","page":"API Reference","title":"MathOptInterface.Bridges.added_constraint_types","text":"added_constraint_types(\n    BT::Type{<:AbstractBridge},\n)::Vector{Tuple{Type,Type}}\n\nReturn a list of the types of constraints that bridges of concrete type BT add.\n\nImplementation notes\n\nThis method depends only on the type of the bridge, not the runtime value. If the bridge may add a constraint, the type must be included in the return vector.\n\nExample\n\njulia> MOI.Bridges.added_constraint_types(\n           MOI.Bridges.Constraint.ZeroOneBridge{Float64},\n       )\n2-element Vector{Tuple{Type, Type}}:\n (MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.Interval{Float64})\n (MathOptInterface.VariableIndex, MathOptInterface.Integer)\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfVariables}","page":"API Reference","title":"MathOptInterface.get","text":"MOI.get(b::AbstractBridge, ::MOI.NumberOfVariables)::Int64\n\nReturn the number of variables created by the bridge b in the model.\n\nSee also MOI.NumberOfConstraints.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this if the bridge adds new variables.\n\n\n\n\n\n","category":"method"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfVariableIndices}","page":"API Reference","title":"MathOptInterface.get","text":"MOI.get(b::AbstractBridge, ::MOI.ListOfVariableIndices)\n\nReturn the list of variables created by the bridge b.\n\nSee also MOI.ListOfVariableIndices.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this if the bridge adds new variables.\n\n\n\n\n\n","category":"method"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.NumberOfConstraints}","page":"API Reference","title":"MathOptInterface.get","text":"MOI.get(b::AbstractBridge, ::MOI.NumberOfConstraints{F,S})::Int64 where {F,S}\n\nReturn the number of constraints of the type F-in-S created by the bridge b.\n\nSee also MOI.NumberOfConstraints.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this for the constraint types returned by added_constraint_types.\n\n\n\n\n\n","category":"method"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.get-Tuple{MathOptInterface.Bridges.AbstractBridge, MathOptInterface.ListOfConstraintIndices}","page":"API Reference","title":"MathOptInterface.get","text":"MOI.get(b::AbstractBridge, ::MOI.ListOfConstraintIndices{F,S}) where {F,S}\n\nReturn a  Vector{ConstraintIndex{F,S}} with indices of all constraints of type F-in-S created by the bride b.\n\nSee also MOI.ListOfConstraintIndices.\n\nImplementation notes\n\nThere is a default fallback, so you need only implement this for the constraint types returned by added_constraint_types.\n\n\n\n\n\n","category":"method"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.needs_final_touch","page":"API Reference","title":"MathOptInterface.Bridges.needs_final_touch","text":"needs_final_touch(bridge::AbstractBridge)::Bool\n\nReturn whether final_touch is implemented by bridge.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.final_touch","page":"API Reference","title":"MathOptInterface.Bridges.final_touch","text":"final_touch(bridge::AbstractBridge, model::MOI.ModelLike)::Nothing\n\nA function that is called immediately prior to MOI.optimize! to allow bridges to modify their reformulations with repsect to other variables and constraints in model.\n\nFor example, if the correctness of bridge depends on the bounds of a variable or the fact that variables are integer, then the bridge can implement final_touch to check assumptions immediately before a call to MOI.optimize!.\n\nIf you implement this method, you must also implement needs_final_touch.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.bridging_cost","page":"API Reference","title":"MathOptInterface.Bridges.bridging_cost","text":"bridging_cost(b::AbstractBridgeOptimizer, S::Type{<:MOI.AbstractSet}})\n\nReturn the cost of bridging variables constrained in S on creation, is_bridged(b, S) is assumed to be true.\n\nbridging_cost(\n    b::AbstractBridgeOptimizer,\n    F::Type{<:MOI.AbstractFunction},\n    S::Type{<:MOI.AbstractSet},\n)\n\nReturn the cost of bridging F-in-S constraints.\n\nis_bridged(b, S) is assumed to be true.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#Constraint-bridge-API","page":"API Reference","title":"Constraint bridge API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.Constraint.AbstractBridge\nsupports_constraint(::Type{<:Bridges.Constraint.AbstractBridge}, ::Type{<:AbstractFunction}, ::Type{<:AbstractSet})\nBridges.Constraint.concrete_bridge_type\nBridges.Constraint.bridge_constraint\nBridges.Constraint.AbstractFunctionConversionBridge\nBridges.Constraint.SingleBridgeOptimizer\nBridges.Constraint.add_all_bridges\nBridges.Constraint.FlipSignBridge\nBridges.Constraint.AbstractToIntervalBridge\nBridges.Constraint.SetMapBridge\nBridges.Constraint.conversion_cost","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.AbstractBridge","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.AbstractBridge","text":"abstract type AbstractBridge <: MOI.Bridges.AbstractType\n\nSubtype of MOI.Bridges.AbstractBridge for constraint bridges.\n\nIn addition to the required implementation described in MOI.Bridges.AbstractBridge, subtypes of AbstractBridge must additionally implement:\n\nMOI.supports_constraint(::Type{<:AbstractBridge}, ::Type{<:MOI.AbstractFunction}, ::Type{<:MOI.AbstractSet})\nconcrete_bridge_type\nbridge_constraint\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.supports_constraint-Tuple{Type{<:MathOptInterface.Bridges.Constraint.AbstractBridge}, Type{<:MathOptInterface.AbstractFunction}, Type{<:MathOptInterface.AbstractSet}}","page":"API Reference","title":"MathOptInterface.supports_constraint","text":"MOI.supports_constraint(\n    BT::Type{<:AbstractBridge},\n    F::Type{<:MOI.AbstractFunction},\n    S::Type{<:MOI.AbstractSet},\n)::Bool\n\nReturn a Bool indicating whether the bridges of type BT support bridging F-in-S constraints.\n\nImplementation notes\n\nThis method depends only on the type of the inputs, not the runtime values.\nThere is a default fallback, so you need only implement this method for constraint types that the bridge implements.\n\n\n\n\n\n","category":"method"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.concrete_bridge_type","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.concrete_bridge_type","text":"concrete_bridge_type(\n    BT::Type{<:AbstractBridge},\n    F::Type{<:MOI.AbstractFunction},\n    S::Type{<:MOI.AbstractSet}\n)::Type\n\nReturn the concrete type of the bridge supporting F-in-S constraints.\n\nThis function can only be called if MOI.supports_constraint(BT, F, S) is true.\n\nExample\n\nThe SplitIntervalBridge bridges a MOI.VariableIndex-in-MOI.Interval constraint into a MOI.VariableIndex-in-MOI.GreaterThan and a MOI.VariableIndex-in-MOI.LessThan constraint.\n\njulia> MOI.Bridges.Constraint.concrete_bridge_type(\n           MOI.Bridges.Constraint.SplitIntervalBridge{Float64},\n           MOI.VariableIndex,\n           MOI.Interval{Float64},\n       )\nMathOptInterface.Bridges.Constraint.SplitIntervalBridge{Float64, MathOptInterface.VariableIndex, MathOptInterface.Interval{Float64}, MathOptInterface.GreaterThan{Float64}, MathOptInterface.LessThan{Float64}}\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.bridge_constraint","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.bridge_constraint","text":"bridge_constraint(\n    BT::Type{<:AbstractBridge},\n    model::MOI.ModelLike,\n    func::AbstractFunction,\n    set::MOI.AbstractSet,\n)::BT\n\nBridge the constraint func-in-set using bridge BT to model and returns a bridge object of type BT.\n\nImplementation notes\n\nThe bridge type BT should be a concrete type, that is, all the type parameters of the bridge must be set.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.AbstractFunctionConversionBridge","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.AbstractFunctionConversionBridge","text":"abstract type AbstractFunctionConversionBridge{F,S} <: AbstractBridge end\n\nAbstract type to support writing bridges in which the function changes but the set does not.\n\nBy convention, the transformed function is stored in the .constraint field.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.SingleBridgeOptimizer","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.SingleBridgeOptimizer","text":"SingleBridgeOptimizer{BT<:AbstractBridge}(model::MOI.ModelLike)\n\nReturn AbstractBridgeOptimizer that always bridges any objective function supported by the bridge BT.\n\nThis is in contrast with the MOI.Bridges.LazyBridgeOptimizer, which only bridges the objective function if it is supported by the bridge BT and unsupported by model.\n\nExample\n\njulia> struct MyNewBridge{T} <: MOI.Bridges.Constraint.AbstractBridge end\n\njulia> bridge = MOI.Bridges.Constraint.SingleBridgeOptimizer{MyNewBridge{Float64}}(\n           MOI.Utilities.Model{Float64}(),\n       )\nMOIB.Constraint.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}\nwith 0 constraint bridges\nwith inner model MOIU.Model{Float64}\n\nImplementation notes\n\nAll bridges should simplify the creation of SingleBridgeOptimizers by defining a constant that wraps the bridge in a SingleBridgeOptimizer.\n\njulia> const MyNewBridgeModel{T,OT<:MOI.ModelLike} =\n           MOI.Bridges.Constraint.SingleBridgeOptimizer{MyNewBridge{T},OT};\n\nThis enables users to create bridged models as follows:\n\njulia> MyNewBridgeModel{Float64}(MOI.Utilities.Model{Float64}())\nMOIB.Constraint.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}\nwith 0 constraint bridges\nwith inner model MOIU.Model{Float64}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.add_all_bridges","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.add_all_bridges","text":"add_all_bridges(bridged_model, ::Type{T}) where {T}\n\nAdd all bridges defined in the Bridges.Constraint submodule to bridged_model. The coefficient type used is T.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.FlipSignBridge","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.FlipSignBridge","text":"FlipSignBridge{T,S1,S2,F,G}\n\nAn abstract type that simplifies the creation of other bridges.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.AbstractToIntervalBridge","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.AbstractToIntervalBridge","text":"AbstractToIntervalBridge{T<:AbstractFloat,S,F}\n\nAn abstract type that simplifies the creation of other bridges.\n\nwarning: Warning\nT must be a AbstractFloat type because otherwise typemin and typemax would either be not implemented (e.g. BigInt), or would not give infinite value (e.g. Int). For this reason, this bridge is only added to MOI.Bridges.full_bridge_optimizer when T is a subtype of AbstractFloat.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.SetMapBridge","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.SetMapBridge","text":"abstract type SetMapBridge{T,S2,S1,F,G} <: AbstractBridge end\n\nConsider two type of sets, S1 and S2, and a linear mapping A such that the image of a set of type S1 under A is a set of type S2.\n\nA SetMapBridge{T,S2,S1,F,G} is a bridge that maps G-in-S1 constraints into F-in-S2 by mapping the function through A.\n\nThe linear map A is described by;\n\nMOI.Bridges.map_set\nMOI.Bridges.map_function.\n\nImplementing a method for these two functions is sufficient to bridge constraints. However, in order for the getters and setters of attributes such as dual solutions and starting values to work as well, a method for the following functions must be implemented:\n\nMOI.Bridges.inverse_map_set\nMOI.Bridges.inverse_map_function\nMOI.Bridges.adjoint_map_function\nMOI.Bridges.inverse_adjoint_map_function\n\nSee the docstrings of each function to see which feature would be missing if it was not implemented for a given bridge.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Constraint.conversion_cost","page":"API Reference","title":"MathOptInterface.Bridges.Constraint.conversion_cost","text":"conversion_cost(\n    F::Type{<:MOI.AbstractFunction},\n    G::Type{<:MOI.AbstractFunction},\n)::Float64\n\nReturn a Float64 returning the cost of converting any function of type G to a function of type F with convert.\n\nThis cost is used to compute MOI.Bridges.bridging_cost.\n\nThe default cost is Inf, which means that MOI.Bridges.Constraint.FunctionConversionBridge should not attempt the conversion.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#Objective-bridge-API","page":"API Reference","title":"Objective bridge API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.Objective.AbstractBridge\nBridges.Objective.supports_objective_function\nBridges.set_objective_function_type\nBridges.Objective.concrete_bridge_type\nBridges.Objective.bridge_objective\nBridges.Objective.SingleBridgeOptimizer\nBridges.Objective.add_all_bridges","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Objective.AbstractBridge","page":"API Reference","title":"MathOptInterface.Bridges.Objective.AbstractBridge","text":"abstract type AbstractBridge <: MOI.Bridges.AbstractBridge end\n\nSubtype of MOI.Bridges.AbstractBridge for objective bridges.\n\nIn addition to the required implementation described in MOI.Bridges.AbstractBridge, subtypes of AbstractBridge must additionally implement:\n\nsupports_objective_function\nconcrete_bridge_type\nbridge_objective\nMOI.Bridges.set_objective_function_type\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Objective.supports_objective_function","page":"API Reference","title":"MathOptInterface.Bridges.Objective.supports_objective_function","text":"supports_objective_function(\n    BT::Type{<:MOI.Bridges.Objective.AbstractBridge},\n    F::Type{<:MOI.AbstractFunction},\n)::Bool\n\nReturn a Bool indicating whether the bridges of type BT support bridging objective functions of type F.\n\nImplementation notes\n\nThis method depends only on the type of the inputs, not the runtime values.\nThere is a default fallback, so you need only implement this method For objective functions that the bridge implements.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.set_objective_function_type","page":"API Reference","title":"MathOptInterface.Bridges.set_objective_function_type","text":"set_objective_function_type(\n    BT::Type{<:Objective.AbstractBridge},\n)::Type{<:MOI.AbstractFunction}\n\nReturn the type of objective function that bridges of concrete type BT set.\n\nImplementation notes\n\nThis method depends only on the type of the bridge, not the runtime value.\n\nExample\n\njulia> MOI.Bridges.set_objective_function_type(\n           MOI.Bridges.Objective.FunctionizeBridge{Float64},\n       )\nMathOptInterface.ScalarAffineFunction{Float64}\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Objective.concrete_bridge_type","page":"API Reference","title":"MathOptInterface.Bridges.Objective.concrete_bridge_type","text":"concrete_bridge_type(\n    BT::Type{<:MOI.Bridges.Objective.AbstractBridge},\n    F::Type{<:MOI.AbstractFunction},\n)::Type\n\nReturn the concrete type of the bridge supporting objective functions of type F.\n\nThis function can only be called if MOI.supports_objective_function(BT, F) is true.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Objective.bridge_objective","page":"API Reference","title":"MathOptInterface.Bridges.Objective.bridge_objective","text":"bridge_objective(\n    BT::Type{<:MOI.Bridges.Objective.AbstractBridge},\n    model::MOI.ModelLike,\n    func::MOI.AbstractFunction,\n)::BT\n\nBridge the objective function func using bridge BT to model and returns a bridge object of type BT.\n\nImplementation notes\n\nThe bridge type BT must be a concrete type, that is, all the type parameters of the bridge must be set.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Objective.SingleBridgeOptimizer","page":"API Reference","title":"MathOptInterface.Bridges.Objective.SingleBridgeOptimizer","text":"SingleBridgeOptimizer{BT<:AbstractBridge}(model::MOI.ModelLike)\n\nReturn AbstractBridgeOptimizer that always bridges any objective function supported by the bridge BT.\n\nThis is in contrast with the MOI.Bridges.LazyBridgeOptimizer, which only bridges the objective function if it is supported by the bridge BT and unsupported by model.\n\nExample\n\njulia> struct MyNewBridge{T} <: MOI.Bridges.Objective.AbstractBridge end\n\njulia> bridge = MOI.Bridges.Objective.SingleBridgeOptimizer{MyNewBridge{Float64}}(\n           MOI.Utilities.Model{Float64}(),\n       )\nMOIB.Objective.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}\nwith 0 objective bridges\nwith inner model MOIU.Model{Float64}\n\nImplementation notes\n\nAll bridges should simplify the creation of SingleBridgeOptimizers by defining a constant that wraps the bridge in a SingleBridgeOptimizer.\n\njulia> const MyNewBridgeModel{T,OT<:MOI.ModelLike} =\n           MOI.Bridges.Objective.SingleBridgeOptimizer{MyNewBridge{T},OT};\n\nThis enables users to create bridged models as follows:\n\njulia> MyNewBridgeModel{Float64}(MOI.Utilities.Model{Float64}())\nMOIB.Objective.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}\nwith 0 objective bridges\nwith inner model MOIU.Model{Float64}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Objective.add_all_bridges","page":"API Reference","title":"MathOptInterface.Bridges.Objective.add_all_bridges","text":"add_all_bridges(model, ::Type{T}) where {T}\n\nAdd all bridges defined in the Bridges.Objective submodule to model.\n\nThe coefficient type used is T.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#ref_variable_bridges","page":"API Reference","title":"Variable bridge API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.Variable.AbstractBridge\nBridges.Variable.supports_constrained_variable\nBridges.Variable.concrete_bridge_type\nBridges.Variable.bridge_constrained_variable\nBridges.Variable.SingleBridgeOptimizer\nBridges.Variable.add_all_bridges\nBridges.Variable.FlipSignBridge\nBridges.Variable.SetMapBridge\nBridges.Variable.unbridged_map","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.AbstractBridge","page":"API Reference","title":"MathOptInterface.Bridges.Variable.AbstractBridge","text":"abstract type AbstractBridge <: MOI.Bridges.AbstractBridge end\n\nSubtype of MOI.Bridges.AbstractBridge for variable bridges.\n\nIn addition to the required implementation described in MOI.Bridges.AbstractBridge, subtypes of AbstractBridge must additionally implement:\n\nsupports_constrained_variable\nconcrete_bridge_type\nbridge_constrained_variable\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.supports_constrained_variable","page":"API Reference","title":"MathOptInterface.Bridges.Variable.supports_constrained_variable","text":"supports_constrained_variable(\n    BT::Type{<:AbstractBridge},\n    S::Type{<:MOI.AbstractSet},\n)::Bool\n\nReturn a Bool indicating whether the bridges of type BT support bridging constrained variables in S. That is, it returns true if the bridge of type BT converts constrained variables of type S into a form supported by the solver.\n\nImplementation notes\n\nThis method depends only on the type of the bridge and set, not the runtime values.\nThere is a default fallback, so you need only implement this method for sets that the bridge implements.\n\nExample\n\njulia> MOI.Bridges.Variable.supports_constrained_variable(\n           MOI.Bridges.Variable.NonposToNonnegBridge{Float64},\n           MOI.Nonpositives,\n       )\ntrue\n\njulia> MOI.Bridges.Variable.supports_constrained_variable(\n           MOI.Bridges.Variable.NonposToNonnegBridge{Float64},\n           MOI.Nonnegatives,\n       )\nfalse\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.concrete_bridge_type","page":"API Reference","title":"MathOptInterface.Bridges.Variable.concrete_bridge_type","text":"concrete_bridge_type(\n    BT::Type{<:AbstractBridge},\n    S::Type{<:MOI.AbstractSet},\n)::Type\n\nReturn the concrete type of the bridge supporting variables in S constraints.\n\nThis function can only be called if MOI.supports_constrained_variable(BT, S) is true.\n\nExamples\n\nAs a variable in MOI.GreaterThan is bridged into variables in MOI.Nonnegatives by the VectorizeBridge:\n\njulia> MOI.Bridges.Variable.concrete_bridge_type(\n           MOI.Bridges.Variable.VectorizeBridge{Float64},\n           MOI.GreaterThan{Float64},\n       )\nMathOptInterface.Bridges.Variable.VectorizeBridge{Float64, MathOptInterface.Nonnegatives}\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.bridge_constrained_variable","page":"API Reference","title":"MathOptInterface.Bridges.Variable.bridge_constrained_variable","text":"bridge_constrained_variable(\n    BT::Type{<:AbstractBridge},\n    model::MOI.ModelLike,\n    set::MOI.AbstractSet,\n)::BT\n\nBridge the constrained variable in set using bridge BT to model and returns a bridge object of type BT.\n\nImplementation notes\n\nThe bridge type BT must be a concrete type, that is, all the type parameters of the bridge must be set.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.SingleBridgeOptimizer","page":"API Reference","title":"MathOptInterface.Bridges.Variable.SingleBridgeOptimizer","text":"SingleBridgeOptimizer{BT<:AbstractBridge}(model::MOI.ModelLike)\n\nReturn MOI.Bridges.AbstractBridgeOptimizer that always bridges any variables constrained on creation supported by the bridge BT.\n\nThis is in contrast with the MOI.Bridges.LazyBridgeOptimizer, which only bridges the variables constrained on creation if they are supported by the bridge BT and unsupported by model.\n\nwarning: Warning\nTwo SingleBridgeOptimizers cannot be used together as both of them assume that the underlying model only returns variable indices with nonnegative values. Use MOI.Bridges.LazyBridgeOptimizer instead.\n\nExample\n\njulia> struct MyNewBridge{T} <: MOI.Bridges.Variable.AbstractBridge end\n\njulia> bridge = MOI.Bridges.Variable.SingleBridgeOptimizer{MyNewBridge{Float64}}(\n           MOI.Utilities.Model{Float64}(),\n       )\nMOIB.Variable.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}\nwith 0 variable bridges\nwith inner model MOIU.Model{Float64}\n\nImplementation notes\n\nAll bridges should simplify the creation of SingleBridgeOptimizers by defining a constant that wraps the bridge in a SingleBridgeOptimizer.\n\njulia> const MyNewBridgeModel{T,OT<:MOI.ModelLike} =\n           MOI.Bridges.Variable.SingleBridgeOptimizer{MyNewBridge{T},OT};\n\nThis enables users to create bridged models as follows:\n\njulia> MyNewBridgeModel{Float64}(MOI.Utilities.Model{Float64}())\nMOIB.Variable.SingleBridgeOptimizer{MyNewBridge{Float64}, MOIU.Model{Float64}}\nwith 0 variable bridges\nwith inner model MOIU.Model{Float64}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.add_all_bridges","page":"API Reference","title":"MathOptInterface.Bridges.Variable.add_all_bridges","text":"add_all_bridges(model, ::Type{T}) where {T}\n\nAdd all bridges defined in the Bridges.Variable submodule to model.\n\nThe coefficient type used is T.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.FlipSignBridge","page":"API Reference","title":"MathOptInterface.Bridges.Variable.FlipSignBridge","text":"abstract type FlipSignBridge{T,S1,S2} <: SetMapBridge{T,S2,S1} end\n\nAn abstract type that simplifies the creation of other bridges.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.SetMapBridge","page":"API Reference","title":"MathOptInterface.Bridges.Variable.SetMapBridge","text":"abstract type SetMapBridge{T,S1,S2} <: AbstractBridge end\n\nConsider two type of sets, S1 and S2, and a linear mapping A such that the image of a set of type S1 under A is a set of type S2.\n\nA SetMapBridge{T,S1,S2} is a bridge that substitutes constrained variables in S2 into the image through A of constrained variables in S1.\n\nThe linear map A is described by:\n\nMOI.Bridges.map_set\nMOI.Bridges.map_function\n\nImplementing a method for these two functions is sufficient to bridge constrained variables. However, in order for the getters and setters of attributes such as dual solutions and starting values to work as well, a method for the following functions must be implemented:\n\nMOI.Bridges.inverse_map_set\nMOI.Bridges.inverse_map_function\nMOI.Bridges.adjoint_map_function\nMOI.Bridges.inverse_adjoint_map_function.\n\nSee the docstrings of each function to see which feature would be missing if it was not implemented for a given bridge.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Variable.unbridged_map","page":"API Reference","title":"MathOptInterface.Bridges.Variable.unbridged_map","text":"unbridged_map(\n   bridge::MOI.Bridges.Variable.AbstractBridge,\n    vi::MOI.VariableIndex,\n)\n\nFor a bridged variable in a scalar set, return a tuple of pairs mapping the variables created by the bridge to an affine expression in terms of the bridged variable vi.\n\nunbridged_map(\n    bridge::MOI.Bridges.Variable.AbstractBridge,\n    vis::Vector{MOI.VariableIndex},\n)\n\nFor a bridged variable in a vector set, return a tuple of pairs mapping the variables created by the bridge to an affine expression in terms of the bridged variable vis. If this method is not implemented, it falls back to calling the following method for every variable of vis.\n\nunbridged_map(\n    bridge::MOI.Bridges.Variable.AbstractBridge,\n    vi::MOI.VariableIndex,\n    i::MOI.Bridges.IndexInVector,\n)\n\nFor a bridged variable in a vector set, return a tuple of pairs mapping the variables created by the bridge to an affine expression in terms of the bridged variable vi corresponding to the ith variable of the vector.\n\nIf there is no way to recover the expression in terms of the bridged variable(s) vi(s), return nothing. See ZerosBridge for an example of bridge returning nothing.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#AbstractBridgeOptimizer-API","page":"API Reference","title":"AbstractBridgeOptimizer API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.AbstractBridgeOptimizer\nBridges.bridged_variable_function\nBridges.unbridged_variable_function\nBridges.bridged_function\nBridges.supports_constraint_bridges\nBridges.recursive_model","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.AbstractBridgeOptimizer","page":"API Reference","title":"MathOptInterface.Bridges.AbstractBridgeOptimizer","text":"abstract type AbstractBridgeOptimizer <: MOI.AbstractOptimizer end\n\nAn abstract type that implements generic functions for bridges.\n\nImplementation notes\n\nBy convention, the inner optimizer should be stored in a model field. If not, the optimizer must implement MOI.optimize!.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.bridged_variable_function","page":"API Reference","title":"MathOptInterface.Bridges.bridged_variable_function","text":"bridged_variable_function(\n    b::AbstractBridgeOptimizer,\n    vi::MOI.VariableIndex,\n)\n\nReturn a MOI.AbstractScalarFunction of variables of b.model that equals vi. That is, if the variable vi is bridged, it returns its expression in terms of the variables of b.model. Otherwise, it returns vi.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.unbridged_variable_function","page":"API Reference","title":"MathOptInterface.Bridges.unbridged_variable_function","text":"unbridged_variable_function(\n    b::AbstractBridgeOptimizer,\n    vi::MOI.VariableIndex,\n)\n\nReturn a MOI.AbstractScalarFunction of variables of b that equals vi. That is, if the variable vi is an internal variable of b.model created by a bridge but not visible to the user, it returns its expression in terms of the variables of bridged variables. Otherwise, it returns vi.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.bridged_function","page":"API Reference","title":"MathOptInterface.Bridges.bridged_function","text":"bridged_function(b::AbstractBridgeOptimizer, value)::typeof(value)\n\nSubstitute any bridged MOI.VariableIndex in value by an equivalent expression in terms of variables of b.model.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.supports_constraint_bridges","page":"API Reference","title":"MathOptInterface.Bridges.supports_constraint_bridges","text":"supports_constraint_bridges(b::AbstractBridgeOptimizer)::Bool\n\nReturn a Bool indicating if b supports MOI.Bridges.Constraint.AbstractBridge.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.recursive_model","page":"API Reference","title":"MathOptInterface.Bridges.recursive_model","text":"recursive_model(b::AbstractBridgeOptimizer)\n\nIf a variable, constraint, or objective is bridged, return the context of the inner variables. For most optimizers, this should be b.model.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#LazyBridgeOptimizer-API","page":"API Reference","title":"LazyBridgeOptimizer API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.LazyBridgeOptimizer\nBridges.full_bridge_optimizer\nBridges.ListOfNonstandardBridges\nBridges.add_bridge\nBridges.remove_bridge\nBridges.has_bridge\nBridges.print_active_bridges\nBridges.print_graph\nBridges.debug_supports_constraint\nBridges.debug_supports","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.LazyBridgeOptimizer","page":"API Reference","title":"MathOptInterface.Bridges.LazyBridgeOptimizer","text":"LazyBridgeOptimizer(model::MOI.ModelLike)\n\nThe LazyBridgeOptimizer is a bridge optimizer that supports multiple bridges, and only bridges things which are not supported by the internal model.\n\nInternally, the LazyBridgeOptimizer solves a shortest hyper-path problem to determine which bridges to use.\n\nIn general, you should use full_bridge_optimizer instead of this constructor because full_bridge_optimizer automatically adds a large number of supported bridges.\n\nSee also: add_bridge, remove_bridge, has_bridge and full_bridge_optimizer.\n\nExample\n\njulia> model = MOI.Bridges.LazyBridgeOptimizer(MOI.Utilities.Model{Float64}())\nMOIB.LazyBridgeOptimizer{MOIU.Model{Float64}}\nwith 0 variable bridges\nwith 0 constraint bridges\nwith 0 objective bridges\nwith inner model MOIU.Model{Float64}\n\njulia> MOI.Bridges.add_bridge(model, MOI.Bridges.Variable.FreeBridge{Float64})\n\njulia> MOI.Bridges.has_bridge(model, MOI.Bridges.Variable.FreeBridge{Float64})\ntrue\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.full_bridge_optimizer","page":"API Reference","title":"MathOptInterface.Bridges.full_bridge_optimizer","text":"full_bridge_optimizer(model::MOI.ModelLike, ::Type{T}) where {T}\n\nReturns a LazyBridgeOptimizer bridging model for every bridge defined in this package (see below for the few exceptions) and for the coefficient type T, as well as the bridges in the list returned by the ListOfNonstandardBridges attribute.\n\nExample\n\njulia> model = MOI.Utilities.Model{Float64}();\n\njulia> bridged_model = MOI.Bridges.full_bridge_optimizer(model, Float64);\n\nExceptions\n\nThe following bridges are not added by full_bridge_optimizer, except if they are in the list returned by the ListOfNonstandardBridges attribute:\n\nConstraint.SOCtoNonConvexQuadBridge\nConstraint.RSOCtoNonConvexQuadBridge](@ref)\nConstraint.SOCtoPSDBridge\nIf T is not a subtype of AbstractFloat, subtypes of Constraint.AbstractToIntervalBridge\nConstraint.GreaterToIntervalBridge\nConstraint.LessToIntervalBridge)\n\nSee the docstring of the each bridge for the reason they are not added.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.ListOfNonstandardBridges","page":"API Reference","title":"MathOptInterface.Bridges.ListOfNonstandardBridges","text":"ListOfNonstandardBridges{T}() <: MOI.AbstractOptimizerAttribute\n\nAny optimizer can be wrapped in a LazyBridgeOptimizer using full_bridge_optimizer. However, by default LazyBridgeOptimizer uses a limited set of bridges that are:\n\nimplemented in MOI.Bridges\ngenerally applicable for all optimizers.\n\nFor some optimizers however, it is useful to add additional bridges, such as those that are implemented in external packages (e.g., within the solver package itself) or only apply in certain circumstances (e.g., Constraint.SOCtoNonConvexQuadBridge).\n\nSuch optimizers should implement the ListOfNonstandardBridges attribute to return a vector of bridge types that are added by full_bridge_optimizer in addition to the list of default bridges.\n\nNote that optimizers implementing ListOfNonstandardBridges may require package-specific functions or sets to be used if the non-standard bridges are not added. Therefore, you are recommended to use model = MOI.instantiate(Package.Optimizer; with_bridge_type = T) instead of model = MOI.instantiate(Package.Optimizer). See MOI.instantiate.\n\nExamples\n\nAn optimizer using a non-default bridge in MOI.Bridges\n\nSolvers supporting MOI.ScalarQuadraticFunction can support MOI.SecondOrderCone and MOI.RotatedSecondOrderCone by defining:\n\nfunction MOI.get(::MyQuadraticOptimizer, ::ListOfNonstandardBridges{Float64})\n    return Type[\n        MOI.Bridges.Constraint.SOCtoNonConvexQuadBridge{Float64},\n        MOI.Bridges.Constraint.RSOCtoNonConvexQuadBridge{Float64},\n    ]\nend\n\nAn optimizer defining an internal bridge\n\nSuppose an optimizer can exploit specific structure of a constraint, e.g., it can exploit the structure of the matrix A in the linear system of equations A * x = b.\n\nThe optimizer can define the function:\n\nstruct MatrixAffineFunction{T} <: MOI.AbstractVectorFunction\n    A::SomeStructuredMatrixType{T}\n    b::Vector{T}\nend\n\nand then a bridge\n\nstruct MatrixAffineFunctionBridge{T} <: MOI.Constraint.AbstractBridge\n    # ...\nend\n# ...\n\nfrom VectorAffineFunction{T} to the MatrixAffineFunction. Finally, it defines:\n\nfunction MOI.get(::Optimizer{T}, ::ListOfNonstandardBridges{T}) where {T}\n    return Type[MatrixAffineFunctionBridge{T}]\nend\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.add_bridge","page":"API Reference","title":"MathOptInterface.Bridges.add_bridge","text":"add_bridge(b::LazyBridgeOptimizer, BT::Type{<:AbstractBridge})\n\nEnable the use of the bridges of type BT by b.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.remove_bridge","page":"API Reference","title":"MathOptInterface.Bridges.remove_bridge","text":"remove_bridge(b::LazyBridgeOptimizer, BT::Type{<:AbstractBridge})\n\nDisable the use of the bridges of type BT by b.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.has_bridge","page":"API Reference","title":"MathOptInterface.Bridges.has_bridge","text":"has_bridge(b::LazyBridgeOptimizer, BT::Type{<:AbstractBridge})\n\nReturn a Bool indicating whether the bridges of type BT are used by b.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.print_active_bridges","page":"API Reference","title":"MathOptInterface.Bridges.print_active_bridges","text":"print_active_bridges([io::IO=stdout,] b::MOI.Bridges.LazyBridgeOptimizer)\n\nPrint the set of bridges that are active in the model b.\n\n\n\n\n\nprint_active_bridges(\n    [io::IO=stdout,]\n    b::MOI.Bridges.LazyBridgeOptimizer,\n    F::Type{<:MOI.AbstractFunction}\n)\n\nPrint the set of bridges required for an objective function of type F.\n\n\n\n\n\nprint_active_bridges(\n    [io::IO=stdout,]\n    b::MOI.Bridges.LazyBridgeOptimizer,\n    F::Type{<:MOI.AbstractFunction},\n    S::Type{<:MOI.AbstractSet},\n)\n\nPrint the set of bridges required for a constraint of type F-in-S.\n\n\n\n\n\nprint_active_bridges(\n    [io::IO=stdout,]\n    b::MOI.Bridges.LazyBridgeOptimizer,\n    S::Type{<:MOI.AbstractSet}\n)\n\nPrint the set of bridges required for a variable constrained to set S.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.print_graph","page":"API Reference","title":"MathOptInterface.Bridges.print_graph","text":"print_graph([io::IO = stdout,] b::LazyBridgeOptimizer)\n\nPrint the hyper-graph containing all variable, constraint, and objective types that could be obtained by bridging the variables, constraints, and objectives that are present in the model by all the bridges added to b.\n\nEach node in the hyper-graph corresponds to a variable, constraint, or objective type.\n\nVariable nodes are indicated by [ ]\nConstraint nodes are indicated by ( )\nObjective nodes are indicated by | |\n\nThe number inside each pair of brackets is an index of the node in the hyper-graph.\n\nNote that this hyper-graph is the full list of possible transformations. When the bridged model is created, we select the shortest hyper-path(s) from this graph, so many nodes may be un-used.\n\nTo see which nodes are used, call print_active_bridges.\n\nFor more information, see Legat, B., Dowson, O., Garcia, J., and Lubin, M. (2020).  \"MathOptInterface: a data structure for mathematical optimization problems.\" URL: https://arxiv.org/abs/2002.03447\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.debug_supports_constraint","page":"API Reference","title":"MathOptInterface.Bridges.debug_supports_constraint","text":"debug_supports_constraint(\n    b::LazyBridgeOptimizer,\n    F::Type{<:MOI.AbstractFunction},\n    S::Type{<:MOI.AbstractSet};\n    io::IO = Base.stdout,\n)\n\nPrints to io explanations for the value of MOI.supports_constraint with the same arguments.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.debug_supports","page":"API Reference","title":"MathOptInterface.Bridges.debug_supports","text":"debug_supports(\n    b::LazyBridgeOptimizer,\n    ::MOI.ObjectiveFunction{F};\n    io::IO = Base.stdout,\n) where F\n\nPrints to io explanations for the value of MOI.supports with the same arguments.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#constraint_set_map","page":"API Reference","title":"SetMap API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.map_set\nBridges.inverse_map_set\nBridges.map_function\nBridges.inverse_map_function\nBridges.adjoint_map_function\nBridges.inverse_adjoint_map_function","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.map_set","page":"API Reference","title":"MathOptInterface.Bridges.map_set","text":"map_set(::Type{BT}, set) where {BT}\n\nReturn the image of set through the linear map A defined in Variable.SetMapBridge and Constraint.SetMapBridge. This is used for bridging the constraint and setting the MOI.ConstraintSet.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.inverse_map_set","page":"API Reference","title":"MathOptInterface.Bridges.inverse_map_set","text":"inverse_map_set(::Type{BT}, set) where {BT}\n\nReturn the preimage of set through the linear map A defined in Variable.SetMapBridge and Constraint.SetMapBridge. This is used for getting the MOI.ConstraintSet.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.map_function","page":"API Reference","title":"MathOptInterface.Bridges.map_function","text":"map_function(::Type{BT}, func) where {BT}\n\nReturn the image of func through the linear map A defined in Variable.SetMapBridge and Constraint.SetMapBridge. This is used for getting the MOI.ConstraintPrimal of variable bridges. For constraint bridges, this is used for bridging the constraint, setting the MOI.ConstraintFunction and MOI.ConstraintPrimalStart and modifying the function with MOI.modify.\n\nmap_function(::Type{BT}, func, i::IndexInVector) where {BT}\n\nReturn the scalar function at the ith index of the vector function that would be returned by map_function(BT, func) except that it may compute the ith element. This is used by bridged_function and for getting the MOI.VariablePrimal and MOI.VariablePrimalStart of variable bridges.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.inverse_map_function","page":"API Reference","title":"MathOptInterface.Bridges.inverse_map_function","text":"inverse_map_function(::Type{BT}, func) where {BT}\n\nReturn the image of func through the inverse of the linear map A defined in Variable.SetMapBridge and Constraint.SetMapBridge. This is used by Variable.unbridged_map and for setting the MOI.VariablePrimalStart of variable bridges and for getting the MOI.ConstraintFunction, the MOI.ConstraintPrimal and the MOI.ConstraintPrimalStart of constraint bridges.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.adjoint_map_function","page":"API Reference","title":"MathOptInterface.Bridges.adjoint_map_function","text":"adjoint_map_function(::Type{BT}, func) where {BT}\n\nReturn the image of func through the adjoint of the linear map A defined in Variable.SetMapBridge and Constraint.SetMapBridge. This is used for getting the MOI.ConstraintDual and MOI.ConstraintDualStart of constraint bridges.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.inverse_adjoint_map_function","page":"API Reference","title":"MathOptInterface.Bridges.inverse_adjoint_map_function","text":"inverse_adjoint_map_function(::Type{BT}, func) where {BT}\n\nReturn the image of func through the inverse of the adjoint of the linear map A defined in Variable.SetMapBridge and Constraint.SetMapBridge. This is used for getting the MOI.ConstraintDual of variable bridges and setting the MOI.ConstraintDualStart of constraint bridges.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#Bridging-graph-API","page":"API Reference","title":"Bridging graph API","text":"","category":"section"},{"location":"moi/submodules/Bridges/reference/","page":"API Reference","title":"API Reference","text":"Bridges.Graph\nBridges.VariableNode\nBridges.ConstraintNode\nBridges.ObjectiveNode\nBridges.Edge\nBridges.ObjectiveEdge\nBridges.add_node\nBridges.add_edge\nBridges.set_variable_constraint_node\nBridges.bridge_index\nBridges.is_variable_edge_best","category":"page"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Graph","page":"API Reference","title":"MathOptInterface.Bridges.Graph","text":"Graph()\n\nA type-stable datastructure for computing the shortest hyperpath problem.\n\nNodes\n\nThere are three types of nodes in the graph:\n\nVariableNode\nConstraintNode\nObjectiveNode\n\nAdd nodes to the graph using add_node.\n\nEdges\n\nThere are two types of edges in the graph:\n\nEdge\nObjectiveEdge\n\nAdd edges to the graph using add_edge.\n\nFor the ability to add a variable constrained  on creation as a free variable followed by a constraint, use set_variable_constraint_node.\n\nOptimal hyper-edges\n\nUse bridge_index to compute the minimum-cost bridge leaving a node.\n\nNote that bridge_index lazy runs a Bellman-Ford algorithm to compute the set of minimum cost edges. Thus, the first call to bridge_index after adding new nodes or edges will take longer than subsequent calls.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.VariableNode","page":"API Reference","title":"MathOptInterface.Bridges.VariableNode","text":"VariableNode(index::Int)\n\nA node in Graph representing a variable constrained on creation.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.ConstraintNode","page":"API Reference","title":"MathOptInterface.Bridges.ConstraintNode","text":"ConstraintNode(index::Int)\n\nA node in Graph representing a constraint.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.ObjectiveNode","page":"API Reference","title":"MathOptInterface.Bridges.ObjectiveNode","text":"ObjectiveNode(index::Int)\n\nA node in Graph representing an objective function.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.Edge","page":"API Reference","title":"MathOptInterface.Bridges.Edge","text":"Edge(\n    bridge_index::Int,\n    added_variables::Vector{VariableNode},\n    added_constraints::Vector{ConstraintNode},\n    cost::Float64 = 1.0,\n)\n\nReturn a new datastructure representing an edge in Graph that starts at a VariableNode or a ConstraintNode.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.ObjectiveEdge","page":"API Reference","title":"MathOptInterface.Bridges.ObjectiveEdge","text":"ObjectiveEdge(\n    bridge_index::Int,\n    added_variables::Vector{VariableNode},\n    added_constraints::Vector{ConstraintNode},\n)\n\nReturn a new datastructure representing an edge in Graph that starts at an ObjectiveNode.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.add_node","page":"API Reference","title":"MathOptInterface.Bridges.add_node","text":"add_node(graph::Graph, ::Type{VariableNode})::VariableNode\nadd_node(graph::Graph, ::Type{ConstraintNode})::ConstraintNode\nadd_node(graph::Graph, ::Type{ObjectiveNode})::ObjectiveNode\n\nAdd a new node to graph.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.add_edge","page":"API Reference","title":"MathOptInterface.Bridges.add_edge","text":"add_edge(graph::Graph, node::VariableNode, edge::Edge)::Nothing\nadd_edge(graph::Graph, node::ConstraintNode, edge::Edge)::Nothing\nadd_edge(graph::Graph, node::ObjectiveNode, edge::ObjectiveEdge)::Nothing\n\nAdd edge to graph, where edge starts at node and connects to the nodes defined in edge.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.set_variable_constraint_node","page":"API Reference","title":"MathOptInterface.Bridges.set_variable_constraint_node","text":"set_variable_constraint_node(\n    graph::Graph,\n    variable_node::VariableNode,\n    constraint_node::ConstraintNode,\n    cost::Int,\n)\n\nAs an alternative to variable_node, add a virtual edge to graph that represents adding a free variable, followed by a constraint of type constraint_node, with bridging cost cost.\n\nWhy is this needed?\n\nVariables can either be added as a variable constrained on creation, or as a free variable which then has a constraint added to it.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.bridge_index","page":"API Reference","title":"MathOptInterface.Bridges.bridge_index","text":"bridge_index(graph::Graph, node::VariableNode)::Int\nbridge_index(graph::Graph, node::ConstraintNode)::Int\nbridge_index(graph::Graph, node::ObjectiveNode)::Int\n\nReturn the optimal index of the bridge to chose from node.\n\n\n\n\n\n","category":"function"},{"location":"moi/submodules/Bridges/reference/#MathOptInterface.Bridges.is_variable_edge_best","page":"API Reference","title":"MathOptInterface.Bridges.is_variable_edge_best","text":"is_variable_edge_best(graph::Graph, node::VariableNode)::Bool\n\nReturn a Bool indicating whether node should be added as a variable constrained on creation, or as a free variable followed by a constraint.\n\n\n\n\n\n","category":"function"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"EditURL = \"https://github.com/kocvara/Loraine.jl/blob/v0.1.2/README.md\"","category":"page"},{"location":"packages/Loraine/#Loraine.jl","page":"kocvara/Loraine.jl","title":"Loraine.jl","text":"","category":"section"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"Sweet Lor(r)aine, let the party carry on[1]...","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"[1]: https://www.youtube.com/watch?v=0D2wNf1lVrI","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"Loraine.jl is a Julia implementation of an interior point method algorithm for linear semidefinite optimization problems. ","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"The special feature of Loraine is the iterative solver for linear systems. This is to be used for problems with (very) low rank solution matrix.","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"Standard (non-low-rank) problems and linear programs can be solved using the direct solver; then the user gets a standard IP method akin SDPT3.","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"There is also a MATLAB version of the code at kocvara/Loraine.m.","category":"page"},{"location":"packages/Loraine/#License-and-Original-Contributors","page":"kocvara/Loraine.jl","title":"License and Original Contributors","text":"","category":"section"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"Loraine is licensed under the MIT License.","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"Loraine was developed by Soodeh Habibi and Michal Kočvara, University of Birmingham, and Michael Stingl, University of Erlangen, for H2020 ITN POEMA. ","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"The JuMP interface was provided by Benoît Legat. His help is greatly acknowledged.","category":"page"},{"location":"packages/Loraine/#Installation","page":"kocvara/Loraine.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"Install Loraine using Pkg.add:","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"import Pkg\nPkg.add(\"Loraine\")","category":"page"},{"location":"packages/Loraine/#Use-with-JuMP","page":"kocvara/Loraine.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"To use Loraine with JuMP, use Loraine.Optimizer:","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"using JuMP, Loraine\nmodel = Model(Loraine.Optimizer)\nset_attribute(model, \"maxit\", 100)","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"To solve an SDP problem stored in SDPA format, do","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"using JuMP, Loraine\nmodel = read_from_file(\"examples/data/theta1.dat-s\")\nset_optimizer(model, Loraine.Optimizer)\noptimize!(model)","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"For more examples, the folder examples includes a few examples of how to use Loraine via JuMP; in particular, solve_sdpa.jl reads an SDP in the SDPA input format and solves it by Loraine. A few sample problems can be found in folder examples/data.","category":"page"},{"location":"packages/Loraine/#Rank-one-data","page":"kocvara/Loraine.jl","title":"Rank-one data","text":"","category":"section"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"If the solution does not have low rank, it is recommended to use a direct  solver kit = 0. However, if you know that your data matrices are all rank-one,  use the option datarank = -1 to get a significant reduction in the complexity  (and CPU time). Examples of such problems are maxG11 and thetaG11 from the  SDPLIB collection.","category":"page"},{"location":"packages/Loraine/#Options","page":"kocvara/Loraine.jl","title":"Options","text":"","category":"section"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"The list of options:","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"kit             # kit = 0 for direct solver; kit = 1 for CG [0]\ntol_cg          # tolerance for CG solver [1.0e-2]\ntol_cg_up       # tolerance update [0.5]\ntol_cg_min      # minimal tolerance for CG solver [1.0e-6]\neDIMACS         # epsilon for DIMACS error stopping criterion [1.0e-5]\npreconditioner  # 0...no; 1...H_alpha; 2...H_beta; 4...hybrid [1]\nerank           # estimated rank [1]\naamat           # 0..A^TA; 1..diag(A^TA); 2..identity [2]\nverb            # 2..full output; 1..short output; 0..no output [1]\ndatarank        # 0..full rank matrices expected [0]\n                # -1..rank-1 matrices expected, converted to vectors, if possible\n                # (TBD) 1..vectors expected for low-rank data matrices\ninitpoint       # 0..Loraine heuristics, 1..SDPT3-like heuristics [0]\ntiming          # 1..yes, 0..no\nmaxit           # maximal number of global iterations [200]","category":"page"},{"location":"packages/Loraine/#Citing","page":"kocvara/Loraine.jl","title":"Citing","text":"","category":"section"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"If you find Loraine useful, please cite the following paper:","category":"page"},{"location":"packages/Loraine/","page":"kocvara/Loraine.jl","title":"kocvara/Loraine.jl","text":"@article{loraine2023,\n  title={Loraine-An interior-point solver for low-rank semidefinite programming},\n  author={Habibi, Soodeh and Ko{\\v{c}}vara, Michal and Stingl, Michael},\n  www={https://hal.science/hal-04076509/}\n  note={Preprint hal-04076509}\n  year={2023}\n}","category":"page"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"EditURL = nothing","category":"page"},{"location":"api/JuMP.Containers/#DocumenterReference_JuMP.Containers","page":"JuMP.Containers","title":"JuMP.Containers","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"This page lists the public API of JuMP.Containers.","category":"page"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"info: Info\nThis page is an unstructured list of the JuMP.Containers API. For a more structured overview, read the Manual or Tutorial parts of this documentation.","category":"page"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"Load all of the public the API into the current scope with:","category":"page"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"using JuMP.Containers","category":"page"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"Alternatively, load only the module with:","category":"page"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"import JuMP.Containers","category":"page"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"and then prefix all calls with JuMP.Containers. to create JuMP.Containers.<NAME>.","category":"page"},{"location":"api/JuMP.Containers/#DenseAxisArray","page":"JuMP.Containers","title":"DenseAxisArray","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.DenseAxisArray","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.DenseAxisArray","page":"JuMP.Containers","title":"JuMP.Containers.DenseAxisArray","text":"DenseAxisArray(data::Array{T, N}, axes...) where {T, N}\n\nConstruct a JuMP array with the underlying data specified by the data array and the given axes. Exactly N axes must be provided, and their lengths must match size(data) in the corresponding dimensions.\n\nExample\n\njulia> array = Containers.DenseAxisArray([1 2; 3 4], [:a, :b], 2:3)\n2-dimensional DenseAxisArray{Int64,2,...} with index sets:\n    Dimension 1, [:a, :b]\n    Dimension 2, 2:3\nAnd data, a 2×2 Matrix{Int64}:\n 1  2\n 3  4\n\njulia> array[:b, 3]\n4\n\n\n\n\n\nDenseAxisArray{T}(undef, axes...) where T\n\nConstruct an uninitialized DenseAxisArray with element-type T indexed over the given axes.\n\nExample\n\njulia> array = Containers.DenseAxisArray{Float64}(undef, [:a, :b], 1:2);\n\njulia> fill!(array, 1.0)\n2-dimensional DenseAxisArray{Float64,2,...} with index sets:\n    Dimension 1, [:a, :b]\n    Dimension 2, 1:2\nAnd data, a 2×2 Matrix{Float64}:\n 1.0  1.0\n 1.0  1.0\n\njulia> array[:a, 2] = 5.0\n5.0\n\njulia> array[:a, 2]\n5.0\n\njulia> array\n2-dimensional DenseAxisArray{Float64,2,...} with index sets:\n    Dimension 1, [:a, :b]\n    Dimension 2, 1:2\nAnd data, a 2×2 Matrix{Float64}:\n 1.0  5.0\n 1.0  1.0\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP.Containers/#SparseAxisArray","page":"JuMP.Containers","title":"SparseAxisArray","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.SparseAxisArray","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.SparseAxisArray","page":"JuMP.Containers","title":"JuMP.Containers.SparseAxisArray","text":"struct SparseAxisArray{T,N,K<:NTuple{N, Any}} <: AbstractArray{T,N}\n    data::Dict{K,T}\nend\n\nN-dimensional array with elements of type T where only a subset of the entries are defined. The entries with indices idx = (i1, i2, ..., iN) in keys(data) has value data[idx]. Note that as opposed to SparseArrays.AbstractSparseArray, the missing entries are not assumed to be zero(T), they are simply not part of the array. This means that the result of map(f, sa::SparseAxisArray) or f.(sa::SparseAxisArray) has the same sparsity structure than sa even if f(zero(T)) is not zero.\n\nExample\n\njulia> dict = Dict((:a, 2) => 1.0, (:a, 3) => 2.0, (:b, 3) => 3.0)\nDict{Tuple{Symbol, Int64}, Float64} with 3 entries:\n  (:a, 3) => 2.0\n  (:b, 3) => 3.0\n  (:a, 2) => 1.0\n\njulia> array = Containers.SparseAxisArray(dict)\nSparseAxisArray{Float64, 2, Tuple{Symbol, Int64}} with 3 entries:\n  [a, 2]  =  1.0\n  [a, 3]  =  2.0\n  [b, 3]  =  3.0\n\njulia> array[:b, 3]\n3.0\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP.Containers/#Containers.@container","page":"JuMP.Containers","title":"Containers.@container","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.@container","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.@container","page":"JuMP.Containers","title":"JuMP.Containers.@container","text":"@container([i=..., j=..., ...], expr[, container = :Auto])\n\nCreate a container with indices i, j, ... and values given by expr that may depend on the value of the indices.\n\n@container(ref[i=..., j=..., ...], expr[, container = :Auto])\n\nSame as above but the container is assigned to the variable of name ref.\n\nThe type of container can be controlled by the container keyword.\n\nnote: Note\nWhen the index set is explicitly given as 1:n for any expression n, it is transformed to Base.OneTo(n) before being given to container.\n\nExample\n\njulia> Containers.@container([i = 1:3, j = 1:3], i + j)\n3×3 Matrix{Int64}:\n 2  3  4\n 3  4  5\n 4  5  6\n\njulia> I = 1:3\n1:3\n\njulia> Containers.@container(x[i = I, j = I], i + j);\n\njulia> x\n2-dimensional DenseAxisArray{Int64,2,...} with index sets:\n    Dimension 1, 1:3\n    Dimension 2, 1:3\nAnd data, a 3×3 Matrix{Int64}:\n 2  3  4\n 3  4  5\n 4  5  6\n\njulia> Containers.@container([i = 2:3, j = 1:3], i + j)\n2-dimensional DenseAxisArray{Int64,2,...} with index sets:\n    Dimension 1, 2:3\n    Dimension 2, Base.OneTo(3)\nAnd data, a 2×3 Matrix{Int64}:\n 3  4  5\n 4  5  6\n\njulia> Containers.@container([i = 1:3, j = 1:3; i <= j], i + j)\nSparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 6 entries:\n  [1, 1]  =  2\n  [1, 2]  =  3\n  [1, 3]  =  4\n  [2, 2]  =  4\n  [2, 3]  =  5\n  [3, 3]  =  6\n\n\n\n\n\n","category":"macro"},{"location":"api/JuMP.Containers/#Containers.container","page":"JuMP.Containers","title":"Containers.container","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.container","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.container","page":"JuMP.Containers","title":"JuMP.Containers.container","text":"container(f::Function, indices[[, ::Type{C} = AutoContainerType], names])\n\nCreate a container of type C with index names names, indices indices and values at given indices given by f.\n\nIf the method with names is not specialized on Type{C}, it falls back to calling  container(f, indices, c) for backwards compatibility with containers not supporting index names.\n\nExample\n\njulia> Containers.container((i, j) -> i + j, Containers.vectorized_product(Base.OneTo(3), Base.OneTo(3)))\n3×3 Matrix{Int64}:\n 2  3  4\n 3  4  5\n 4  5  6\n\njulia> Containers.container((i, j) -> i + j, Containers.vectorized_product(1:3, 1:3))\n2-dimensional DenseAxisArray{Int64,2,...} with index sets:\n    Dimension 1, 1:3\n    Dimension 2, 1:3\nAnd data, a 3×3 Matrix{Int64}:\n 2  3  4\n 3  4  5\n 4  5  6\n\njulia> Containers.container((i, j) -> i + j, Containers.vectorized_product(2:3, Base.OneTo(3)))\n2-dimensional DenseAxisArray{Int64,2,...} with index sets:\n    Dimension 1, 2:3\n    Dimension 2, Base.OneTo(3)\nAnd data, a 2×3 Matrix{Int64}:\n 3  4  5\n 4  5  6\n\njulia> Containers.container((i, j) -> i + j, Containers.nested(() -> 1:3, i -> i:3, condition = (i, j) -> isodd(i) || isodd(j)))\nSparseAxisArray{Int64, 2, Tuple{Int64, Int64}} with 5 entries:\n  [1, 1]  =  2\n  [1, 2]  =  3\n  [1, 3]  =  4\n  [2, 3]  =  5\n  [3, 3]  =  6\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP.Containers/#Containers.rowtable","page":"JuMP.Containers","title":"Containers.rowtable","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.rowtable","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.rowtable","page":"JuMP.Containers","title":"JuMP.Containers.rowtable","text":"rowtable([f::Function=identity,] x; [header::Vector{Symbol} = Symbol[]])\n\nApplies the function f to all elements of the variable container x, returning the result as a Vector of NamedTuples, where header is a vector containing the corresponding axis names.\n\nIf x is an N-dimensional array, there must be N+1 names, so that the last name corresponds to the result of f(x[i]).\n\nIf header is left empty, then the default header is [:x1, :x2, ..., :xN, :y].\n\ninfo: Info\nA Vector of NamedTuples implements the Tables.jl interface, and so the result can be used as input for any function that consumes a 'Tables.jl' compatible source.\n\nExample\n\njulia> model = Model();\n\njulia> @variable(model, x[i=1:2, j=i:2] >= 0, start = i+j);\n\njulia> Containers.rowtable(start_value, x; header = [:i, :j, :start])\n3-element Vector{NamedTuple{(:i, :j, :start), Tuple{Int64, Int64, Float64}}}:\n (i = 1, j = 2, start = 3.0)\n (i = 1, j = 1, start = 2.0)\n (i = 2, j = 2, start = 4.0)\n\njulia> Containers.rowtable(x)\n3-element Vector{NamedTuple{(:x1, :x2, :y), Tuple{Int64, Int64, VariableRef}}}:\n (x1 = 1, x2 = 2, y = x[1,2])\n (x1 = 1, x2 = 1, y = x[1,1])\n (x1 = 2, x2 = 2, y = x[2,2])\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP.Containers/#Containers.default_container","page":"JuMP.Containers","title":"Containers.default_container","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.default_container","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.default_container","page":"JuMP.Containers","title":"JuMP.Containers.default_container","text":"default_container(indices)\n\nIf indices is a NestedIterator, return a SparseAxisArray. Otherwise, indices should be a VectorizedProductIterator and the function returns Array if all iterators of the product are Base.OneTo and retunrs DenseAxisArray otherwise.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP.Containers/#Containers.nested","page":"JuMP.Containers","title":"Containers.nested","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.nested","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.nested","page":"JuMP.Containers","title":"JuMP.Containers.nested","text":"nested(iterators...; condition = (args...) -> true)\n\nCreate a NestedIterator.\n\nExample\n\njulia> iterator = Containers.nested(\n           () -> 1:2,\n           (i,) -> [\"A\", \"B\"];\n           condition = (i, j) -> isodd(i) || j == \"B\",\n       );\n\njulia> collect(iterator)\n3-element Vector{Tuple{Int64, String}}:\n (1, \"A\")\n (1, \"B\")\n (2, \"B\")\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP.Containers/#Containers.vectorized_product","page":"JuMP.Containers","title":"Containers.vectorized_product","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.vectorized_product","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.vectorized_product","page":"JuMP.Containers","title":"JuMP.Containers.vectorized_product","text":"vectorized_product(iterators...)\n\nCreated a VectorizedProductIterator.\n\nExample\n\njulia> iterator = Containers.vectorized_product(1:2, [\"A\", \"B\"]);\n\njulia> collect(iterator)\n2×2 Matrix{Tuple{Int64, String}}:\n (1, \"A\")  (1, \"B\")\n (2, \"A\")  (2, \"B\")\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP.Containers/#Containers.build_ref_sets","page":"JuMP.Containers","title":"Containers.build_ref_sets","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.build_ref_sets","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.build_ref_sets","page":"JuMP.Containers","title":"JuMP.Containers.build_ref_sets","text":"build_ref_sets(_error::Function, expr)\n\nHelper function for macros to construct container objects.\n\nwarning: Warning\nThis function is for advanced users implementing JuMP extensions. See container_code for more details.\n\nArguments\n\n_error: a function that takes a String and throws an error, potentially annotating the input string with extra information such as from which macro it was thrown from. Use error if you do not want a modified error message.\nexpr: an Expr that specifies the container, e.g., :(x[i = 1:3, [:red, :blue], k = S; i + k <= 6])\n\nReturns\n\nindex_vars: a Vector{Any} of names for the index variables, e.g., [:i, gensym(), :k]. These may also be expressions, like :((i, j)) from a call like :(x[(i, j) in S]).\nindices: an iterator over the indices, for example, Containers.NestedIterator\n\nExample\n\nSee container_code for a worked example.\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP.Containers/#Containers.container_code","page":"JuMP.Containers","title":"Containers.container_code","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.container_code","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.container_code","page":"JuMP.Containers","title":"JuMP.Containers.container_code","text":"container_code(\n    index_vars::Vector{Any},\n    indices::Expr,\n    code,\n    requested_container::Union{Symbol,Expr},\n)\n\nUsed in macros to construct a call to container. This should be used in conjunction with build_ref_sets.\n\nArguments\n\nindex_vars::Vector{Any}: a vector of names for the indices of the container. These may also be expressions, like :((i, j)) from a call like :(x[(i, j) in S]).\nindices::Expr: an expression that evaluates to an iterator of the indices.\ncode: an expression or literal constant for the value to be stored in the container as a function of the named index_vars.\nrequested_container: passed to the third argument of container. For built-in JuMP types, choose one of :Array, :DenseAxisArray, :SparseAxisArray, or :Auto. For a user-defined container, this expression must evaluate to the correct type.\n\nwarning: Warning\nIn most cases, you should esc(code) before passing it to container_code.\n\nExample\n\njulia> macro foo(ref_sets, code)\n           index_vars, indices = Containers.build_ref_sets(error, ref_sets)\n           return Containers.container_code(\n               index_vars,\n               indices,\n               esc(code),\n               :Auto,\n            )\n       end\n@foo (macro with 1 method)\n\njulia> @foo(x[i=1:2, j=[\"A\", \"B\"]], j^i)\n2-dimensional DenseAxisArray{String,2,...} with index sets:\n    Dimension 1, Base.OneTo(2)\n    Dimension 2, [\"A\", \"B\"]\nAnd data, a 2×2 Matrix{String}:\n \"A\"   \"B\"\n \"AA\"  \"BB\"\n\n\n\n\n\n","category":"function"},{"location":"api/JuMP.Containers/#Containers.AutoContainerType","page":"JuMP.Containers","title":"Containers.AutoContainerType","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.AutoContainerType","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.AutoContainerType","page":"JuMP.Containers","title":"JuMP.Containers.AutoContainerType","text":"AutoContainerType\n\nPass AutoContainerType to container to let the container type be chosen based on the type of the indices using default_container.\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP.Containers/#Containers.NestedIterator","page":"JuMP.Containers","title":"Containers.NestedIterator","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.NestedIterator","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.NestedIterator","page":"JuMP.Containers","title":"JuMP.Containers.NestedIterator","text":"struct NestedIterator{T}\n    iterators::T # Tuple of functions\n    condition::Function\nend\n\nIterators over the tuples that are produced by a nested for loop.\n\nConstruct a NestedIterator using nested.\n\nExample\n\njulia> iterators = (() -> 1:2, (i,) -> [\"A\", \"B\"]);\n\njulia> condition = (i, j) -> isodd(i) || j == \"B\";\n\njulia> x = Containers.NestedIterator(iterators, condition);\n\njulia> for (i, j) in x\n           println((i, j))\n       end\n(1, \"A\")\n(1, \"B\")\n(2, \"B\")\n\nis the same as\n\njulia> for i in iterators[1]()\n           for j in iterators[2](i)\n               if condition(i, j)\n                   println((i, j))\n               end\n           end\n       end\n(1, \"A\")\n(1, \"B\")\n(2, \"B\")\n\n\n\n\n\n","category":"type"},{"location":"api/JuMP.Containers/#Containers.VectorizedProductIterator","page":"JuMP.Containers","title":"Containers.VectorizedProductIterator","text":"","category":"section"},{"location":"api/JuMP.Containers/","page":"JuMP.Containers","title":"JuMP.Containers","text":"JuMP.Containers.Containers.VectorizedProductIterator","category":"page"},{"location":"api/JuMP.Containers/#JuMP.Containers.VectorizedProductIterator","page":"JuMP.Containers","title":"JuMP.Containers.VectorizedProductIterator","text":"struct VectorizedProductIterator{T}\n    prod::Iterators.ProductIterator{T}\nend\n\nA wrapper type for Iterators.ProuctIterator that discards shape information and returns a Vector.\n\nConstruct a VectorizedProductIterator using vectorized_product.\n\n\n\n\n\n","category":"type"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"EditURL = \"multi_objective_knapsack.jl\"","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Multi-objective-knapsack","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"The purpose of this tutorial is to demonstrate how to create and solve a multi-objective linear program. In addition, it demonstrates how to work with solvers which return multiple solutions.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Required-packages","page":"Multi-objective knapsack","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"using JuMP\nimport HiGHS\nimport MultiObjectiveAlgorithms as MOA\nimport Plots\nimport Test  #hide","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"MultiObjectiveAlgorithms.jl is a package which implements a variety of algorithms for solving multi-objective optimization problems. Because it is a long package name, we import it instead as MOA.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Formulation","page":"Multi-objective knapsack","title":"Formulation","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"The knapsack problem is a classic problem in mixed-integer programming. Given a collection of items i in I, each of which has an associated weight, w_i, and profit, p_i, the knapsack problem determines which profit-maximizing subset of items to pack into a knapsack such that the total weight is less than a capacity c. The mathematical formulation is:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"beginaligned\nmax  sumlimits_i in I p_i x_i \ntextst   sumlimits_i in I w_i x_i le c\n x_i in 0 1  forall i in I\nendaligned","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"where x_i is 1 if we pack item i into the knapsack and 0 otherwise.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"For this tutorial, we extend the single-objective knapsack problem by adding another objective: given a desirability rating, r_i, we wish to maximize the total desirability of the items in our knapsack. Thus, our mathematical formulation is now:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"beginaligned\nmax  sumlimits_i in I p_i x_i \n      sumlimits_i in I r_i x_i \ntextst   sumlimits_i in I w_i x_i le c\n x_i in 0 1  forall i in I\nendaligned","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Data","page":"Multi-objective knapsack","title":"Data","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"The data for this example was taken from vOptGeneric, and the original author was @xgandibleux.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"profit = [77, 94, 71, 63, 96, 82, 85, 75, 72, 91, 99, 63, 84, 87, 79, 94, 90]\ndesire = [65, 90, 90, 77, 95, 84, 70, 94, 66, 92, 74, 97, 60, 60, 65, 97, 93]\nweight = [80, 87, 68, 72, 66, 77, 99, 85, 70, 93, 98, 72, 100, 89, 67, 86, 91]\ncapacity = 900\nN = length(profit)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Comparing the capacity to the total weight of all the items:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"capacity / sum(weight)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"shows that we can take approximately 64% of the items.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Plotting the items, we see that there are a range of items with different profits and desirability. Some items have a high profit and a high desirability, others have a low profit and a high desirability (and vice versa).","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Plots.scatter(\n    profit,\n    desire;\n    xlabel = \"Profit\",\n    ylabel = \"Desire\",\n    legend = false,\n)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"The goal of the bi-objective knapsack problem is to choose a subset which maximizes both objectives.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#JuMP-formulation","page":"Multi-objective knapsack","title":"JuMP formulation","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Our JuMP formulation is a direct translation of the mathematical formulation:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"model = Model()\n@variable(model, x[1:N], Bin)\n@constraint(model, sum(weight[i] * x[i] for i in 1:N) <= capacity)\n@expression(model, profit_expr, sum(profit[i] * x[i] for i in 1:N))\n@expression(model, desire_expr, sum(desire[i] * x[i] for i in 1:N))\n@objective(model, Max, [profit_expr, desire_expr])","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Note how we form a multi-objective program by passing a vector of scalar objective functions.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Solution","page":"Multi-objective knapsack","title":"Solution","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"To solve our model, we need an optimizer which supports multi-objective linear programs. One option is to use the MultiObjectiveAlgorithms.jl package.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"set_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))\nset_silent(model)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"MultiObjectiveAlgorithms.jl supports many different algorithms for solving multiobjective optimization problems. One option is the epsilon-constraint method:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Let's solve the problem and see the solution","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"There are 9 solutions available. We can also use result_count to see how many solutions are available:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"result_count(model)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Accessing-multiple-solutions","page":"Multi-objective knapsack","title":"Accessing multiple solutions","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Access the nine different solutions in the model using the result keyword to solution_summary, value, and objective_value:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"solution_summary(model; result = 5)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"objective_value(model; result = 5)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Note that because we set a vector of two objective functions, the objective value is a vector with two elements. We can also query the value of each objective separately:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"value(profit_expr; result = 5)","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Visualizing-objective-space","page":"Multi-objective knapsack","title":"Visualizing objective space","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Unlike single-objective optimization problems, multi-objective optimization problems do not have a single optimal solution. Instead, the solutions returned represent possible trade-offs that the decision maker can choose between the two objectives. A common way to visualize this is by plotting the objective values of each of the solutions:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"plot = Plots.scatter(\n    [value(profit_expr; result = i) for i in 1:result_count(model)],\n    [value(desire_expr; result = i) for i in 1:result_count(model)];\n    xlabel = \"Profit\",\n    ylabel = \"Desire\",\n    title = \"Objective space\",\n    label = \"\",\n    xlims = (915, 960),\n)\nfor i in 1:result_count(model)\n    y = objective_value(model; result = i)\n    Plots.annotate!(y[1] - 1, y[2], (i, 10))\nend\nideal_point = objective_bound(model)\nPlots.scatter!([ideal_point[1]], [ideal_point[2]]; label = \"Ideal point\")","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"Visualizing the objective space lets the decision maker choose a solution that suits their personal preferences. For example, result #7 is close to the maximum value of profit, but offers significantly higher desirability compared with solutions #8 and #9.","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"The set of items that are chosen in solution #7 are:","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"items_chosen = [i for i in 1:N if value(x[i]; result = 7) > 0.9]","category":"page"},{"location":"tutorials/linear/multi_objective_knapsack/#Next-steps","page":"Multi-objective knapsack","title":"Next steps","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_knapsack/","page":"Multi-objective knapsack","title":"Multi-objective knapsack","text":"MultiObjectiveAlgorithms.jl implements a number of different algorithms. Try solving the same problem using MOA.Dichotomy(). Does it find the same solution?","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"EditURL = \"https://github.com/chkwon/PATHSolver.jl/blob/v1.6.1/README.md\"","category":"page"},{"location":"packages/PATHSolver/#PATHSolver.jl","page":"chkwon/PATHSolver.jl","title":"PATHSolver.jl","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"PATHSolver.jl is a wrapper for the PATH solver.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"a thin wrapper around the C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"You can solve any complementarity problem using the wrapper around the C API, although you must manually provide the callback functions, including the Jacobian.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"The MathOptInterface wrapper is more limited, supporting only linear complementarity problems, but it enables PATHSolver to be used with JuMP.","category":"page"},{"location":"packages/PATHSolver/#Affiliation","page":"chkwon/PATHSolver.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"This wrapper is maintained by the JuMP community and is not an official wrapper of PATH. However, we are in close contact with the PATH developers, and they have given us permission to re-distribute the PATH binaries for automatic installation.","category":"page"},{"location":"packages/PATHSolver/#License","page":"chkwon/PATHSolver.jl","title":"License","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"PATHSolver.jl is licensed under the MIT License.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"The underlying solver, path is closed source and requires a license.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"Without a license, the PATH Solver can solve problem instances up to with up to 300 variables and 2000 non-zeros. For larger problems, this web page provides a temporary license that is valid for a year.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"You can either store the license in the PATH_LICENSE_STRING environment variable, or you can use the PATHSolver.c_api_License_SetString function immediately after importing the PATHSolver package:","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"import PATHSolver\nPATHSolver.c_api_License_SetString(\"<LICENSE STRING>\")","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"where <LICENSE STRING> is replaced by the current license string.","category":"page"},{"location":"packages/PATHSolver/#Installation","page":"chkwon/PATHSolver.jl","title":"Installation","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"Install PATHSolver.jl as follows:","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"import Pkg\nPkg.add(\"PATHSolver\")","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"By default, PATHSolver.jl will download a copy of the underlying PATH solver. To use a different version of PATH, see the Manual Installation section below.","category":"page"},{"location":"packages/PATHSolver/#Use-with-JuMP","page":"chkwon/PATHSolver.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"julia> using JuMP, PATHSolver\n\njulia> M = [\n           0  0 -1 -1\n           0  0  1 -2\n           1 -1  2 -2\n           1  2 -2  4\n       ]\n4×4 Array{Int64,2}:\n 0   0  -1  -1\n 0   0   1  -2\n 1  -1   2  -2\n 1   2  -2   4\n\njulia> q = [2, 2, -2, -6]\n4-element Array{Int64,1}:\n  2\n  2\n -2\n -6\n\njulia> model = Model(PATHSolver.Optimizer)\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: EMPTY_OPTIMIZER\nSolver name: Path 5.0.00\n\njulia> set_optimizer_attribute(model, \"output\", \"no\")\n\njulia> @variable(model, x[1:4] >= 0)\n4-element Array{VariableRef,1}:\n x[1]\n x[2]\n x[3]\n x[4]\n\njulia> @constraint(model, M * x .+ q ⟂ x)\n[-x[3] - x[4] + 2, x[3] - 2 x[4] + 2, x[1] - x[2] + 2 x[3] - 2 x[4] - 2, x[1] + 2 x[2] - 2 x[3] + 4 x[4] - 6, x[1], x[2], x[3], x[4]] ∈ MOI.Complements(4)\n\njulia> optimize!(model)\nReading options file /var/folders/bg/dzq_hhvx1dxgy6gb5510pxj80000gn/T/tmpiSsCRO\nRead of options file complete.\n\nPath 5.0.00 (Mon Aug 19 10:57:18 2019)\nWritten by Todd Munson, Steven Dirkse, Youngdae Kim, and Michael Ferris\n\njulia> value.(x)\n4-element Array{Float64,1}:\n 2.8\n 0.0\n 0.7999999999999998\n 1.2\n\njulia> termination_status(model)\nLOCALLY_SOLVED::TerminationStatusCode = 4","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"Note that options are set using JuMP.set_optimizer_attribute.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"The list of options supported by PATH can be found here: https://pages.cs.wisc.edu/~ferris/path/options.pdf","category":"page"},{"location":"packages/PATHSolver/#MathOptInterface-API","page":"chkwon/PATHSolver.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"The Path 5.0.03 optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"MOI.Reals","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"MOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VectorOfVariables in MOI.Complements\nMOI.VectorAffineFunction{Float64} in MOI.Complements\nMOI.VectorQuadraticFunction{Float64} in MOI.Complements\nMOI.VectorNonlinearFunction in MOI.Complements","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"MOI.Name()","category":"page"},{"location":"packages/PATHSolver/#Use-with-the-C-API","page":"chkwon/PATHSolver.jl","title":"Use with the C API","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"PATHSolver.jl wraps the PATH C API using PATHSolver.c_api_XXX for the C method XXX. However, using the C API directly from Julia can be challenging, particularly with respect to avoiding issues with Julia's garbage collector.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"Instead, we recommend that you use the PATHSolver.solve_mcp function, which wrappers the C API into a single call. See the docstring of PATHSolver.solve_mcp for a detailed description of the arguments.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"Here is the same example using PATHSolver.solve_mcp. Note that you must manually construct the sparse Jacobian callback.","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"julia> import PATHSolver\n\njulia> M = [\n           0  0 -1 -1\n           0  0  1 -2\n           1 -1  2 -2\n           1  2 -2  4\n       ]\n4×4 Matrix{Int64}:\n 0   0  -1  -1\n 0   0   1  -2\n 1  -1   2  -2\n 1   2  -2   4\n\njulia> q = [2, 2, -2, -6]\n4-element Vector{Int64}:\n  2\n  2\n -2\n -6\n\njulia> function F(n::Cint, x::Vector{Cdouble}, f::Vector{Cdouble})\n           @assert n == length(x) == length(f)\n           f .= M * x .+ q\n           return Cint(0)\n       end\nF (generic function with 1 method)\n\njulia> function J(\n           n::Cint,\n           nnz::Cint,\n           x::Vector{Cdouble},\n           col::Vector{Cint},\n           len::Vector{Cint},\n           row::Vector{Cint},\n           data::Vector{Cdouble},\n       )\n           @assert n == length(x) == length(col) == length(len) == 4\n           @assert nnz == length(row) == length(data)\n           i = 1\n           for c in 1:n\n               col[c], len[c] = i, 0\n               for r in 1:n\n                   if !iszero(M[r, c])\n                       row[i], data[i] = r, M[r, c]\n                       len[c] += 1\n                       i += 1\n                   end\n               end\n           end\n           return Cint(0)\n       end\nJ (generic function with 1 method)\n\njulia> status, z, info = PATHSolver.solve_mcp(\n           F,\n           J,\n           fill(0.0, 4),  # Lower bounds\n           fill(Inf, 4),  # Upper bounds\n           fill(0.0, 4);  # Starting point\n           nnz = 12,      # Number of nonzeros in the Jacobian\n           output = \"yes\",\n       )\nReading options file /var/folders/bg/dzq_hhvx1dxgy6gb5510pxj80000gn/T/jl_iftYBS\n > output yes\nRead of options file complete.\n\nPath 5.0.03 (Fri Jun 26 09:58:07 2020)\nWritten by Todd Munson, Steven Dirkse, Youngdae Kim, and Michael Ferris\n\nCrash Log\nmajor  func  diff  size  residual    step       prox   (label)\n    0     0             1.2649e+01             0.0e+00 (f[    4])\n    1     2     4     2 1.0535e+01  8.0e-01    0.0e+00 (f[    1])\n    2     3     2     4 8.4815e-01  1.0e+00    0.0e+00 (f[    4])\n    3     4     0     3 4.4409e-16  1.0e+00    0.0e+00 (f[    3])\npn_search terminated: no basis change.\n\nMajor Iteration Log\nmajor minor  func  grad  residual    step  type prox    inorm  (label)\n    0     0     5     4 4.4409e-16           I 0.0e+00 4.4e-16 (f[    3])\n\nMajor Iterations. . . . 0\nMinor Iterations. . . . 0\nRestarts. . . . . . . . 0\nCrash Iterations. . . . 3\nGradient Steps. . . . . 0\nFunction Evaluations. . 5\nGradient Evaluations. . 4\nBasis Time. . . . . . . 0.000016\nTotal Time. . . . . . . 0.044383\nResidual. . . . . . . . 4.440892e-16\n(PATHSolver.MCP_Solved, [2.8, 0.0, 0.8, 1.2], PATHSolver.Information(4.4408920985006247e-16, 0.0, 0.0, 0.044383, 1.6e-5, 0.0, 0, 0, 3, 5, 4, 0, 0, 0, 0, false, false, false, true, false, false, false))\n\njulia> status\nMCP_Solved::MCP_Termination = 1\n\njulia> z\n4-element Vector{Float64}:\n 2.8\n 0.0\n 0.8\n 1.2","category":"page"},{"location":"packages/PATHSolver/#Thread-safety","page":"chkwon/PATHSolver.jl","title":"Thread safety","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"PATH is not thread-safe and there are no known work-arounds. Do not run it in parallel using Threads.@threads. See issue #62 for more details.","category":"page"},{"location":"packages/PATHSolver/#Factorization-methods","page":"chkwon/PATHSolver.jl","title":"Factorization methods","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"By default, PATHSolver.jl will download the LUSOL shared library. To use LUSOL, set the following options:","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"model = Model(PATHSolver.Optimizer)\nset_optimizer_attribute(model, \"factorization_method\", \"blu_lusol\")\nset_optimizer_attribute(model, \"factorization_library_name\", PATHSolver.LUSOL_LIBRARY_PATH)","category":"page"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"To use factorization_method umfpack you will need the umfpack shared library that is available directly from the developers of that code for academic use.","category":"page"},{"location":"packages/PATHSolver/#Manual-installation","page":"chkwon/PATHSolver.jl","title":"Manual installation","text":"","category":"section"},{"location":"packages/PATHSolver/","page":"chkwon/PATHSolver.jl","title":"chkwon/PATHSolver.jl","text":"By default PATHSolver.jl will download a copy of the libpath library. If you already have one installed and want to use that, set the PATH_JL_LOCATION environment variable to point to the libpath50.xx library.","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"EditURL = \"multi_objective_examples.jl\"","category":"page"},{"location":"tutorials/linear/multi_objective_examples/#Simple-multi-objective-examples","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"This tutorial contains a number of examples of multi-objective programs from the literature.","category":"page"},{"location":"tutorials/linear/multi_objective_examples/#Required-packages","page":"Simple multi-objective examples","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"using JuMP\nimport HiGHS\nimport MultiObjectiveAlgorithms as MOA","category":"page"},{"location":"tutorials/linear/multi_objective_examples/#Bi-objective-linear-problem","page":"Simple multi-objective examples","title":"Bi-objective linear problem","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"This example is taken from Example 6.3 (from Steuer, 1985), page 154 of Ehrgott, M. (2005). Multicriteria Optimization. Springer, Berlin. The code was adapted from an example in vOptGeneric by @xgandibleux.","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"model = Model()\nset_silent(model)\n@variable(model, x1 >= 0)\n@variable(model, 0 <= x2 <= 3)\n@objective(model, Min, [3x1 + x2, -x1 - 2x2])\n@constraint(model, 3x1 - x2 <= 6)\nset_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))\nset_attribute(model, MOA.Algorithm(), MOA.Lexicographic())\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"for i in 1:result_count(model)\n    print(i, \": z = \", round.(Int, objective_value(model; result = i)), \" | \")\n    println(\"x = \", value.([x1, x2]; result = i))\nend","category":"page"},{"location":"tutorials/linear/multi_objective_examples/#Bi-objective-linear-assignment-problem","page":"Simple multi-objective examples","title":"Bi-objective linear assignment problem","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"This example is taken from Example 9.38 (from Ulungu and Teghem, 1994), page 255 of Ehrgott, M. (2005). Multicriteria Optimization. Springer, Berlin. The code was adapted from an example in vOptGeneric by @xgandibleux.","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"C1 = [5 1 4 7; 6 2 2 6; 2 8 4 4; 3 5 7 1]\nC2 = [3 6 4 2; 1 3 8 3; 5 2 2 3; 4 2 3 5]\nn = size(C2, 1)\nmodel = Model()\nset_silent(model)\n@variable(model, x[1:n, 1:n], Bin)\n@objective(model, Min, [sum(C1 .* x), sum(C2 .* x)])\n@constraint(model, [i = 1:n], sum(x[i, :]) == 1)\n@constraint(model, [j = 1:n], sum(x[:, j]) == 1)\nset_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))\nset_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"for i in 1:result_count(model)\n    print(i, \": z = \", round.(Int, objective_value(model; result = i)), \" | \")\n    println(\"x = \", round.(Int, value.(x; result = i)))\nend","category":"page"},{"location":"tutorials/linear/multi_objective_examples/#Bi-objective-shortest-path-problem","page":"Simple multi-objective examples","title":"Bi-objective shortest path problem","text":"","category":"section"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"This example is taken from Exercise 9.5 page 269 of Ehrgott, M. (2005). Multicriteria Optimization. Springer, Berlin. The code was adapted from an example in vOptGeneric by @xgandibleux.","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"M = 50\nC1 = [\n    M 4 5 M M M\n    M M 2 1 2 7\n    M M M 5 2 M\n    M M 5 M M 3\n    M M M M M 4\n    M M M M M M\n]\nC2 = [\n    M 3 1 M M M\n    M M 1 4 2 2\n    M M M 1 7 M\n    M M 1 M M 2\n    M M M M M 2\n    M M M M M M\n]\nn = size(C2, 1)\nmodel = Model()\nset_silent(model)\n@variable(model, x[1:n, 1:n], Bin)\n@objective(model, Min, [sum(C1 .* x), sum(C2 .* x)])\n@constraint(model, sum(x[1, :]) == 1)\n@constraint(model, sum(x[:, n]) == 1)\n@constraint(model, [i = 2:n-1], sum(x[i, :]) - sum(x[:, i]) == 0)\nset_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))\nset_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())\noptimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/linear/multi_objective_examples/","page":"Simple multi-objective examples","title":"Simple multi-objective examples","text":"for i in 1:result_count(model)\n    print(i, \": z = \", round.(Int, objective_value(model; result = i)), \" | \")\n    X = round.(Int, value.(x; result = i))\n    print(\"Path:\")\n    for ind in findall(val -> val ≈ 1, X)\n        i, j = ind.I\n        print(\" $i->$j\")\n    end\n    println()\nend","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"EditURL = \"network_flows.jl\"","category":"page"},{"location":"tutorials/linear/network_flows/#Network-flow-problems","page":"Network flow problems","title":"Network flow problems","text":"","category":"section"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"This tutorial was originally contributed by Arpit Bhatia.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"In graph theory, a flow network (also known as a transportation network) is a directed graph where each edge has a capacity and each edge receives a flow. The amount of flow on an edge cannot exceed the capacity of the edge.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"Often in operations research, a directed graph is called a network, the vertices are called nodes and the edges are called arcs.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"A flow must satisfy the restriction that the amount of flow into a node equals the amount of flow out of it,  unless it is a source, which has only outgoing flow, or sink, which has only incoming flow.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"A network can be used to model traffic in a computer network, circulation with demands, fluids in pipes,  currents in an electrical circuit, or anything similar in which something travels through a network of nodes.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/linear/network_flows/#The-shortest-path-problem","page":"Network flow problems","title":"The shortest path problem","text":"","category":"section"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"Suppose that each arc (i j) of a graph is assigned a scalar cost a_ij, and suppose that we define the cost of a forward path to be the sum of the costs of its arcs.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"Given a pair of nodes, the shortest path problem is to find a forward path that connects these nodes and has minimum cost.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"beginaligned\nmin  sum_forall e(ij) in E a_ij times x_ij \nst  sum_j x_ij - sum_k x_ki = b_i  forall i\n x_e in 01  forall e in E\nendaligned","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"where b_i is 1 if i is the starting node, -1 if i is the ending node, and 0 otherwise.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"G = [\n    0 100 30 0 0\n    0 0 20 0 0\n    0 0 0 10 60\n    0 15 0 0 50\n    0 0 0 0 0\n]\nn = size(G)[1]\nb = [1, -1, 0, 0, 0]\nshortest_path = Model(HiGHS.Optimizer)\nset_silent(shortest_path)\n@variable(shortest_path, x[1:n, 1:n], Bin)\n# Arcs with zero cost are not a part of the path as they do no exist\n@constraint(shortest_path, [i = 1:n, j = 1:n; G[i, j] == 0], x[i, j] == 0)\n# Flow conservation constraint\n@constraint(shortest_path, [i = 1:n], sum(x[i, :]) - sum(x[:, i]) == b[i],)\n@objective(shortest_path, Min, sum(G .* x))\noptimize!(shortest_path)\nobjective_value(shortest_path)","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"value.(x)","category":"page"},{"location":"tutorials/linear/network_flows/#The-assignment-problem","page":"Network flow problems","title":"The assignment problem","text":"","category":"section"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"Suppose that there are n persons and n objects that we have to match on a one-to-one basis. There is a benefit or value a_ij for matching person i with object j, and we want to assign persons to objects so as to maximize the total benefit.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"There is also a restriction that person i can be assigned to object j only if (i j) belongs to a given set of pairs A.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"Mathematically, we want to find a set of person-object pairs (1 j_1) (n j_n) from A such that the objects j_1j_n are all distinct, and the total benefit sum_i=1^y a_ij_i is maximized.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"beginaligned\nmax  sum_(ij) in A a_ij times y_ij \nst  sum_j(ij) in A y_ij = 1  forall i = 12n \n sum_i(ij) in A y_ij = 1  forall j = 12n \n y_ij in 01  forall (ij) in 12k\nendaligned","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"G = [\n    6 4 5 0\n    0 3 6 0\n    5 0 4 3\n    7 5 5 5\n]\nn = size(G)[1]\nassignment = Model(HiGHS.Optimizer)\nset_silent(assignment)\n@variable(assignment, y[1:n, 1:n], Bin)\n# One person can only be assigned to one object\n@constraint(assignment, [i = 1:n], sum(y[:, i]) == 1)\n# One object can only be assigned to one person\n@constraint(assignment, [j = 1:n], sum(y[j, :]) == 1)\n@objective(assignment, Max, sum(G .* y))\noptimize!(assignment)\nobjective_value(assignment)","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"value.(y)","category":"page"},{"location":"tutorials/linear/network_flows/#The-max-flow-problem","page":"Network flow problems","title":"The max-flow problem","text":"","category":"section"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"In the max-flow problem, we have a graph with two special nodes: the source, denoted by s, and the sink, denoted by t.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"The objective is to move as much flow as possible from s into t while observing the capacity constraints.","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"beginaligned\nmax  sum_v(sv) in E f(sv) \nst  sum_u(uv) in E f(uv)  = sum_w(vw) in E f(vw)  forall v in V - st \n f(uv) leq c(uv)  forall (uv) in E \n f(uv) geq 0  forall (uv) in E\nendaligned","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"G = [\n    0 3 2 2 0 0 0 0\n    0 0 0 0 5 1 0 0\n    0 0 0 0 1 3 1 0\n    0 0 0 0 0 1 0 0\n    0 0 0 0 0 0 0 4\n    0 0 0 0 0 0 0 2\n    0 0 0 0 0 0 0 4\n    0 0 0 0 0 0 0 0\n]\nn = size(G)[1]\nmax_flow = Model(HiGHS.Optimizer)\n@variable(max_flow, f[1:n, 1:n] >= 0)\n# Capacity constraints\n@constraint(max_flow, [i = 1:n, j = 1:n], f[i, j] <= G[i, j])\n# Flow conservation constraints\n@constraint(max_flow, [i = 1:n; i != 1 && i != 8], sum(f[i, :]) == sum(f[:, i]))\n@objective(max_flow, Max, sum(f[1, :]))\noptimize!(max_flow)\nobjective_value(max_flow)","category":"page"},{"location":"tutorials/linear/network_flows/","page":"Network flow problems","title":"Network flow problems","text":"value.(f)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"EditURL = \"optimal_power_flow.jl\"","category":"page"},{"location":"tutorials/applications/optimal_power_flow/#Optimal-power-flow","page":"Optimal power flow","title":"Optimal power flow","text":"","category":"section"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"This tutorial was originally contributed by James Foster (@jd-foster).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"This tutorial formulates and solves an alternating current optimal power flow (AC-OPF) problem, a much-studied nonlinear problem from the field of electrical engineering.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Once we've formulated and solved the nonlinear problem, we will turn our focus to obtaining a good estimate of the objective value at the global optimum through the use of semidefinite programming.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"One main purpose of this tutorial is to highlight JuMP's ability to directly formulate problems involving complex-valued decision variables and complex matrix cones such as the  HermitianPSDCone object.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"For another example of modeling with complex decision variables, see the Quantum state discrimination tutorial, and see the Complex number support section of the manual for more details.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"This tutorial takes a matrix-oriented approach focused on network nodes that simplifies the construction of semidefinite programs. Another approach is to formulate the problem focusing on network lines where it is easier to work with flow constraints. A general approach is provided by the Julia/JuMP package PowerModels.jl, an open-source framework to a broad range of power flow model formulations along with utilities for working with detailed network data.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/#Required-packages","page":"Optimal power flow","title":"Required packages","text":"","category":"section"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"using JuMP\nimport Clarabel\nimport DataFrames\nimport Ipopt\nimport LinearAlgebra\nimport SparseArrays","category":"page"},{"location":"tutorials/applications/optimal_power_flow/#Initial-formulation","page":"Optimal power flow","title":"Initial formulation","text":"","category":"section"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Optimal power flow problems for electrical transmission typically pose the following question: what is the most cost-effective operation of electricity generators while meeting constraints on the safe limits of network components?","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We'll use the 9-node network test case case9mod to explore this problem. The graph of the network, shown here, has three nodes (or buses) each for the different purposes of generation G (nodes 1, 2, and 3), trans-shipment (nodes 4, 6, and 8), and demand D (nodes 5, 7, and 9).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"(Image: Nine Nodes)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"This example is a modified version of the MATPOWER test case case9 (archive) created by Bukhsh et al. (2013) for their test case archive of optimal power flow problems with local optima. This test case is also extensively evaluated in Krasko and Rebennack (2017).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Here bus and network node are taken as analogous terms, as are branch and transmission line.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"For future reference, let's name the number of nodes in the network:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"N = 9;\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The network data can be summarised using a small number of arrays. Using the sparsevec function from the SparseArrays standard library package, we can give the indices and values of the non-zero data points:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"# Real generation: lower (`lb`) and upper (`ub`) bounds\nP_Gen_lb = SparseArrays.sparsevec([1, 2, 3], [10, 10, 10], N)\nP_Gen_ub = SparseArrays.sparsevec([1, 2, 3], [250, 300, 270], N)\n# Reactive generation: lower (`lb`) and upper (`ub`) bounds\nQ_Gen_lb = SparseArrays.sparsevec([1, 2, 3], [-5, -5, -5], N)\nQ_Gen_ub = SparseArrays.sparsevec([1, 2, 3], [300, 300, 300], N)\n# Power demand levels (real, reactive, and complex form)\nP_Demand = SparseArrays.sparsevec([5, 7, 9], [54, 60, 75], N)\nQ_Demand = SparseArrays.sparsevec([5, 7, 9], [18, 21, 30], N)\nS_Demand = P_Demand + im * Q_Demand","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The key decision variables are the real power injections P^G and reactive power injections Q^Gover the allowed range of the generators. All other buses must restrict their generation variables to 0. On the other hand, these non-generator nodes have a fixed  real and reactive power demand, denoted P^D and Q^D respectively (these are fixed at 0 in the case of trans-shipment and generator nodes).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The cost of operating each generator is modeled as a quadratic function of its real power output; in our specific test case, the objective function to minimize is:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"beginalign\n    min          011  (P^G_1)^2 +   5 P^G_1 + 150  \n               +  0085  (P^G_2)^2 + 12 P^G_2 + 600  \n               + 01225   (P^G_3)^2 +     P^G_3 + 335 \nendalign","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Let's create an initial JuMP model with some of this data:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"model = Model(Ipopt.Optimizer)\nset_silent(model)\n@variable(model, P_Gen_lb[i] <= P_G[i in 1:N] <= P_Gen_ub[i])\n@objective(\n    model,\n    Min,\n    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +\n    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +\n    (0.1225 * P_G[3]^2 + P_G[3] + 335),\n);\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Even before solving an optimization problem, we can estimate a lower bound on the best objective value by substituting the lower bound on each generator's real power range (all 10, as it turns out in this case):","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"basic_lower_bound = value(lower_bound, objective_function(model));\nprintln(\"Objective value (basic lower bound) : $basic_lower_bound\")","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"to see that we can do no better than an objective cost of 1188.75.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"(Direct substitution works because a quadratic function of a single variable x with positive coefficients is strictly increasing for all x geq 0.)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"In fact, we can get a quick but even better estimate from the direct observation that the real power generated must meet or exceed the real power demand:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"@constraint(model, sum(P_G) >= sum(P_Demand))\noptimize!(model)\nbetter_lower_bound = round(objective_value(model); digits = 2)\nprintln(\"Objective value (better lower bound): $better_lower_bound\")","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"However, there are additional power flow constraints that must be satisfied.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Power must flow from one or more generation nodes through the transmission lines and end up at a demand node. The state variables of our steady-state alternating current (AC) electrical network are complex-valued voltage variables V_1 ldots V_N. Voltages capture both a magnitude and phase of the node's electrical state in relation to the rest of the system. An AC power system also extends the notion of resistance in wires found in a direct current (DC) circuit to a  complex quantity, known as the impedance, of each transmission line. The reciprocal of impedance is known as admittance. Together, these complex quantities are used to express a complex version of Ohm's law: current flow through a line is proportional to the difference in voltages on each end of the line, multiplied by the admittance.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/#Network-data","page":"Optimal power flow","title":"Network data","text":"","category":"section"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Let's assemble the data we need for writing the complex power flow constraints. The data for the problem consists of a list of the real and imaginary parts of the line impedance. We obtain the following data table from the branch data section of the case9mod MATPOWER format file:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"branch_data = DataFrames.DataFrame([\n    (1, 4, 0.0, 0.0576, 0.0),\n    (4, 5, 0.017, 0.092, 0.158),\n    (6, 5, 0.039, 0.17, 0.358),\n    (3, 6, 0.0, 0.0586, 0.0),\n    (6, 7, 0.0119, 0.1008, 0.209),\n    (8, 7, 0.0085, 0.072, 0.149),\n    (2, 8, 0.0, 0.0625, 0.0),\n    (8, 9, 0.032, 0.161, 0.306),\n    (4, 9, 0.01, 0.085, 0.176),\n]);\nDataFrames.rename!(branch_data, [:F_BUS, :T_BUS, :BR_R, :BR_X, :BR_Bc])","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The first two columns describe the network, supplying the from and to connection points of the lines. The last three columns give the branch resistance, branch reactance and line-charging susceptance.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We will also need to reference the base_MVA number (used for re-scaling):","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"base_MVA = 100;\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"and the number of lines:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"M = size(branch_data, 1)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"From the first two columns of the branch data table, we can create a sparse incidence matrix that simplifies handling of the network layout:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"A =\n    SparseArrays.sparse(branch_data.F_BUS, 1:M, 1, N, M) +\n    SparseArrays.sparse(branch_data.T_BUS, 1:M, -1, N, M)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We form the network impedance vector from the next two columns","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"z = (branch_data.BR_R .+ im * branch_data.BR_X) / base_MVA;\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"and calculate it's corresponding bus admittance matrix as","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Y_0 = A * SparseArrays.spdiagm(1 ./ z) * A';\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"while the last column gives the branch line-charging susceptance","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"y_sh = 1 / 2 * (im * branch_data.BR_Bc) * base_MVA;\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"and leads to the shunt admittance matrix","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Y_sh = SparseArrays.spdiagm(\n    LinearAlgebra.diag(A * SparseArrays.spdiagm(y_sh) * A'),\n);\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"(The construction of the shunt admittance matrix Y_sh looks somewhat more complicated than Y_0 because we only want to add the diagonal elements in the calculation; the line-charging is used only in the nodal voltage terms and not the line voltage terms.)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The full bus admittance matrix Y is then defined as","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Y = Y_0 + Y_sh;\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/#JuMP-model","page":"Optimal power flow","title":"JuMP model","text":"","category":"section"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Now we're ready to write the complex power flow constraints we need to more accurately model the electricity system.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We'll introduce a number of constraints that model both the physics and operational requirements.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Let's start by initializing a new model:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"model = Model(Ipopt.Optimizer)\nset_silent(model)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Then we'll create the nodal power generation variables:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"@variable(\n    model,\n    S_G[i in 1:N] in ComplexPlane(),\n    lower_bound = P_Gen_lb[i] + Q_Gen_lb[i] * im,\n    upper_bound = P_Gen_ub[i] + Q_Gen_ub[i] * im,\n)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We need complex nodal voltages (the system state variables):","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"@variable(model, V[1:N] in ComplexPlane(), start = 1.0 + 0.0im)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"and operational constraints for maintaining voltage magnitude levels:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"@constraint(model, [i in 1:N], 0.9^2 <= real(V[i])^2 + imag(V[i])^2 <= 1.1^2)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We also need to fix an origin or reference angle from which all other complex voltage angles (arguments) are determined. Here we will use node 1 as the nominated reference bus. Fixing the imaginary component of a reference bus to zero sets its complex voltage angle to 0, while constraining the real part to be non-negative disallows equivalent solutions that are just a reflection by 180 degrees:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"@constraint(model, imag(V[1]) == 0);\n@constraint(model, real(V[1]) >= 0);\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The power flow equations express a conservation of energy (power) principle, where power generated less the power consumed must balance the power exchanged with the network:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"@constraint(model, S_G - S_Demand .== V .* conj(Y * V))","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"As above, the objective function is a quadratic cost of real power:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"P_G = real(S_G)\n@objective(\n    model,\n    Min,\n    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +\n    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +\n    (0.1225 * P_G[3]^2 + P_G[3] + 335),\n);\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We're finally ready to solve our nonlinear AC-OPF problem:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"objval_solution = round(objective_value(model); digits = 2)\nprintln(\"Objective value (feasible solution) : $(objval_solution)\")","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The solution's power generation (in rectangular form) and complex voltage values (in polar form using degrees) are:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"DataFrames.DataFrame(;\n    Bus = 1:N,\n    ComplexPowerGen = round.(value.(S_G); digits = 2),\n    VoltageMagnitude = round.(abs.(value.(V)); digits = 2),\n    VoltageAngle_Deg = round.(rad2deg.(angle.(value.(V))); digits = 2),\n)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/#Relaxations-and-better-objective-bounds","page":"Optimal power flow","title":"Relaxations and better objective bounds","text":"","category":"section"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The Ipopt solver uses an interior-point algorithm. It has local optimality guarantees, but is unable to certify whether the solution is globally optimal. The solution we found is indeed globally optimal. The work to verify this has been done in Bukhsh et al. (2013) and Krasko and Rebennack (2017), and different solvers (such as Gurobi, SCIP and GLOMIQO) are also able to verify this.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The techniques of convex relaxations can also be used to improve on our current best lower bound:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"better_lower_bound","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"To this end, observe that the nonlinear constraints in the AC-OPF formulation are quadratic equalities for power flow along with quadratic voltage inequalities.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Let's linearize these constraints by first making the substitution W = V V^*, where:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"W = V V^* quad iff  quad W_ii =  V_i ^2 quad W_ik = V_i  overlineV_k quad forall i  k in  1 ldots N ","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"and where V^* is the conjugate transpose of V.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"On the face of it, this turns a quadratic voltage bound constraint like:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"v_L leq V_i ^2  leq v_U quad i  in  1 ldots N ","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"for some real v_L and v_U into a simple two-sided bound:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"v_L leq W_ii  leq v_U","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"while each quadratic expression for the nodal power term:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"    S^Node_i = V_i overline(YV)_i","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"becomes the linear combination:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"    S^Node_i = (E_ii Y^T) bullet W","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Here A bullet B = operatornametr(A^* B) is the Frobenius inner product of two complex matrices, while E_kn denotes the matrix unit with a single nonzero entry of 1 in row k and column n.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"E(k, n) = SparseArrays.sparse([k], [n], 1, N, N);\nnothing #hide","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Of course, we've shifted the nonlinearity into the equality constraint W = V V^*: it is this constraint we will now relax using a semidefinite programming approach.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We will make use of complex voltages and relax W = V V^* to;","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"W succeq V V^*","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"where the relation succeq is the ordering in the Hermitian positive semidefinite cone.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"The above constraint is equivalent to:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"beginbmatrix 1  V^*  V  W  endbmatrix succeq 0","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"by the theory of the Schur complement. This matrix inequality implies a number of second-order cone constraints by taking certain 2 times 2 minors of the matrix for each i  in  1 ldots N :","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"beginbmatrix 1  V_i^*  V_i  W_ii  endbmatrix succeq 0","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"which is equivalent to the real second-order cone inequality:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"operatornamereal(W_ii) geq operatornamereal(V_i)^2 + operatornameimag(V_i)^2","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We include these implied constraints as well for demonstration purposes.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Putting it all together we get the following semidefinite relaxation of the AC-OPF problem:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"model = Model(Clarabel.Optimizer)\nset_attribute(model, \"tol_gap_rel\", 1e-3)\nset_attribute(model, \"tol_feas\", 1e-3)\nset_attribute(model, \"tol_ktratio\", 1e-3)\n@variable(\n    model,\n    S_G[i in 1:N] in ComplexPlane(),\n    lower_bound = P_Gen_lb[i] + Q_Gen_lb[i] * im,\n    upper_bound = P_Gen_ub[i] + Q_Gen_ub[i] * im,\n)\n@variable(model, W[1:N, 1:N] in HermitianPSDCone())\n@variable(model, V[1:N] in ComplexPlane(), start = 1.0 + 0.0im)\n@constraint(model, [i in 1:N], 0.9^2 <= real(W[i, i]) <= 1.1^2)\n@constraint(model, real(V[1]) >= 0)\n@constraint(model, imag(V[1]) == 0)\n@constraint(model, 0.9 <= real(V[1]) <= 1.1)\n@constraint(model, LinearAlgebra.Hermitian([1 V'; V W]) in HermitianPSDCone())\n# 2 x 2 minor inequalities:\n@constraint(\n    model,\n    [i in 1:N],\n    [0.5, real(W[i, i]), real(V[i]), imag(V[i])] in RotatedSecondOrderCone()\n)\n@constraint(\n    model,\n    [i in 1:N],\n    S_G[i] - S_Demand[i] == LinearAlgebra.tr((conj(Y) * E(i, i)) * W),\n)\nP_G = real(S_G)\n@objective(\n    model,\n    Min,\n    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +\n    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +\n    (0.1225 * P_G[3]^2 + P_G[3] + 335),\n)\noptimize!(model)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"sdp_relaxation_lower_bound = round(objective_value(model); digits = 2)\nprintln(\n    \"Objective value (W & V relax. lower bound): $sdp_relaxation_lower_bound\",\n)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"We can more easily see solution values by rounding out noisy data:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"W_1 = SparseArrays.sparse(round.(value.(W); digits = 2))","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"and recover an approximation to the voltage variables as:","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"DataFrames.DataFrame(;\n    Bus = 1:N,\n    Magnitude = round.(abs.(value.(V)); digits = 2),\n    AngleDeg = round.(rad2deg.(angle.(value.(V))); digits = 2),\n)","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"For further information on exploiting sparsity see Jabr (2012).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"This relaxation has the advantage that we can work directly with complex voltages to extend the formulation, strengthen the relaxation and gain additional approximate information about the voltage variables.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/#References-and-further-resources","page":"Optimal power flow","title":"References and further resources","text":"","category":"section"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Bukhsh, W. A., Grothey, A., McKinnon, K. I., & Trodden, P. A. Local solutions of the optimal power flow problem. IEEE Transactions on Power Systems, 28(4), 4780-4788 (2013).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Jabr, R. A. Exploiting sparsity in SDP relaxations of the OPF problem. IEEE Transactions on Power Systems, 27(2), 1138-1139 (2011).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Krasko, V., & S. Rebennack. Chapter 15: Global Optimization: Optimal Power Flow Problem. In Advances and Trends in Optimization with Engineering Applications, 187—205. MOS-SIAM Series on Optimization. Society for Industrial and Applied Mathematics, 2017.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"Test case case9mod: from the Test Case Archive of Optimal Power Flow (OPF) Problems with Local Optima.","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"R. D. Zimmerman, C. E. Murillo-Sanchez, & R. J. Thomas, MATPOWER: Steady-State Operations, Planning and Analysis Tools for Power Systems Research and Education, IEEE Transactions on Power Systems, 26(1), 12-19 (2011).","category":"page"},{"location":"tutorials/applications/optimal_power_flow/","page":"Optimal power flow","title":"Optimal power flow","text":"MATPOWER data format: MATPOWER manual; see especially Appendix B \"Data File Format\" and Table B-3.","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Bridges/list_of_bridges.md\"","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/#List-of-bridges","page":"List of bridges","title":"List of bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"This section describes the Bridges.AbstractBridges that are implemented in MathOptInterface.","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/#constraint_bridges_ref","page":"List of bridges","title":"Constraint bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"These bridges are subtypes of Bridges.Constraint.AbstractBridge.","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"Bridges.Constraint.GreaterToIntervalBridge\nBridges.Constraint.LessToIntervalBridge\nBridges.Constraint.GreaterToLessBridge\nBridges.Constraint.LessToGreaterBridge\nBridges.Constraint.NonnegToNonposBridge\nBridges.Constraint.NonposToNonnegBridge\nBridges.Constraint.VectorizeBridge\nBridges.Constraint.ScalarizeBridge\nBridges.Constraint.ScalarSlackBridge\nBridges.Constraint.VectorSlackBridge\nBridges.Constraint.ScalarFunctionizeBridge\nBridges.Constraint.VectorFunctionizeBridge\nBridges.Constraint.ToScalarQuadraticBridge\nBridges.Constraint.ToVectorQuadraticBridge\nBridges.Constraint.ToScalarNonlinearBridge\nBridges.Constraint.FunctionConversionBridge\nBridges.Constraint.SplitComplexEqualToBridge\nBridges.Constraint.SplitComplexZerosBridge\nBridges.Constraint.SplitHyperRectangleBridge\nBridges.Constraint.SplitIntervalBridge\nBridges.Constraint.SOCtoRSOCBridge\nBridges.Constraint.RSOCtoSOCBridge\nBridges.Constraint.SOCtoNonConvexQuadBridge\nBridges.Constraint.RSOCtoNonConvexQuadBridge\nBridges.Constraint.QuadtoSOCBridge\nBridges.Constraint.SOCtoPSDBridge\nBridges.Constraint.RSOCtoPSDBridge\nBridges.Constraint.NormInfinityBridge\nBridges.Constraint.NormOneBridge\nBridges.Constraint.NormToPowerBridge\nBridges.Constraint.NormOneConeToNormConeBridge\nBridges.Constraint.SecondOrderConeToNormConeBridge\nBridges.Constraint.NormInfinityConeToNormConeBridge\nBridges.Constraint.GeoMeantoRelEntrBridge\nBridges.Constraint.GeoMeanToPowerBridge\nBridges.Constraint.GeoMeanBridge\nBridges.Constraint.RelativeEntropyBridge\nBridges.Constraint.NormSpectralBridge\nBridges.Constraint.NormNuclearBridge\nBridges.Constraint.SquareBridge\nBridges.Constraint.HermitianToSymmetricPSDBridge\nBridges.Constraint.SetDotScalingBridge\nBridges.Constraint.SetDotInverseScalingBridge\nBridges.Constraint.RootDetBridge\nBridges.Constraint.LogDetBridge\nBridges.Constraint.IndicatorActiveOnFalseBridge\nBridges.Constraint.IndicatorGreaterToLessThanBridge\nBridges.Constraint.IndicatorLessToGreaterThanBridge\nBridges.Constraint.IndicatorSOS1Bridge\nBridges.Constraint.SemiToBinaryBridge\nBridges.Constraint.ZeroOneBridge\nBridges.Constraint.IntegerToZeroOneBridge\nBridges.Constraint.NumberConversionBridge\nBridges.Constraint.AllDifferentToCountDistinctBridge\nBridges.Constraint.ReifiedAllDifferentToCountDistinctBridge\nBridges.Constraint.BinPackingToMILPBridge\nBridges.Constraint.CircuitToMILPBridge\nBridges.Constraint.CountAtLeastToCountBelongsBridge\nBridges.Constraint.CountBelongsToMILPBridge\nBridges.Constraint.CountDistinctToMILPBridge\nBridges.Constraint.ReifiedCountDistinctToMILPBridge\nBridges.Constraint.CountGreaterThanToMILPBridge\nBridges.Constraint.TableToMILPBridge","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GreaterToIntervalBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.GreaterToIntervalBridge","text":"GreaterToIntervalBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nGreaterToIntervalBridge implements the following reformulations:\n\nf(x) ge l into f(x) in l infty)\n\nSource node\n\nGreaterToIntervalBridge supports:\n\nF in MOI.GreaterThan{T}\n\nTarget nodes\n\nGreaterToIntervalBridge creates:\n\nF in MOI.Interval{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.LessToIntervalBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.LessToIntervalBridge","text":"LessToIntervalBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nLessToIntervalBridge implements the following reformulations:\n\nf(x) le u into f(x) in (-infty u\n\nSource node\n\nLessToIntervalBridge supports:\n\nF in MOI.LessThan{T}\n\nTarget nodes\n\nLessToIntervalBridge creates:\n\nF in MOI.Interval{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GreaterToLessBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.GreaterToLessBridge","text":"GreaterToLessBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nGreaterToLessBridge implements the following reformulation:\n\nf(x) ge l into -f(x) le -l\n\nSource node\n\nGreaterToLessBridge supports:\n\nG in MOI.GreaterThan{T}\n\nTarget nodes\n\nGreaterToLessBridge creates:\n\nF in MOI.LessThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.LessToGreaterBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.LessToGreaterBridge","text":"LessToGreaterBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nLessToGreaterBridge implements the following reformulation:\n\nf(x) le u into -f(x) ge -u\n\nSource node\n\nLessToGreaterBridge supports:\n\nG in MOI.LessThan{T}\n\nTarget nodes\n\nLessToGreaterBridge creates:\n\nF in MOI.GreaterThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NonnegToNonposBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NonnegToNonposBridge","text":"NonnegToNonposBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nNonnegToNonposBridge implements the following reformulation:\n\nf(x) in mathbbR_+ into -f(x) in mathbbR_-\n\nSource node\n\nNonnegToNonposBridge supports:\n\nG in MOI.Nonnegatives\n\nTarget nodes\n\nNonnegToNonposBridge creates:\n\nF in MOI.Nonpositives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NonposToNonnegBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NonposToNonnegBridge","text":"NonposToNonnegBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nNonposToNonnegBridge implements the following reformulation:\n\nf(x) in mathbbR_- into -f(x) in mathbbR_+\n\nSource node\n\nNonposToNonnegBridge supports:\n\nG in MOI.Nonpositives\n\nTarget nodes\n\nNonposToNonnegBridge creates:\n\nF in MOI.Nonnegatives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.VectorizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.VectorizeBridge","text":"VectorizeBridge{T,F,S,G} <: Bridges.Constraint.AbstractBridge\n\nVectorizeBridge implements the following reformulations:\n\ng(x) ge a into g(x) - a in mathbbR_+\ng(x) le a into g(x) - a in mathbbR_-\ng(x) == a into g(x) - a in 0\n\nwhere T is the coefficient type of g(x) - a.\n\nSource node\n\nVectorizeBridge supports:\n\nG in MOI.GreaterThan{T}\nG in MOI.LessThan{T}\nG in MOI.EqualTo{T}\n\nTarget nodes\n\nVectorizeBridge creates:\n\nF in S, where S is one of MOI.Nonnegatives, MOI.Nonpositives, MOI.Zeros depending on the type of the input set.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ScalarizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ScalarizeBridge","text":"ScalarizeBridge{T,F,S}\n\nScalarizeBridge implements the following reformulations:\n\nf(x) - a in mathbbR_+ into f_i(x) ge a_i for all i\nf(x) - a in mathbbR_- into f_i(x) le a_i for all i\nf(x) - a in 0 into f_i(x) == a_i for all i\n\nSource node\n\nScalarizeBridge supports:\n\nG in MOI.Nonnegatives{T}\nG in MOI.Nonpositives{T}\nG in MOI.Zeros{T}\n\nTarget nodes\n\nScalarizeBridge creates:\n\nF in S, where S is one of MOI.GreaterThan{T}, MOI.LessThan{T}, and MOI.EqualTo{T}, depending on the type of the input set.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ScalarSlackBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ScalarSlackBridge","text":"ScalarSlackBridge{T,F,S} <: Bridges.Constraint.AbstractBridge\n\nScalarSlackBridge implements the following reformulation:\n\nf(x) in S into f(x) - y == 0 and y in S\n\nSource node\n\nScalarSlackBridge supports:\n\nG in S, where G is not MOI.VariableIndex and S is not MOI.EqualTo\n\nTarget nodes\n\nScalarSlackBridge creates:\n\nF in MOI.EqualTo{T}\nMOI.VariableIndex in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.VectorSlackBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.VectorSlackBridge","text":"VectorSlackBridge{T,F,S} <: Bridges.Constraint.AbstractBridge\n\nVectorSlackBridge implements the following reformulation:\n\nf(x) in S into f(x) - y in 0 and y in S\n\nSource node\n\nVectorSlackBridge supports:\n\nG in S, where G is not MOI.VectorOfVariables and S is not MOI.Zeros\n\nTarget nodes\n\nVectorSlackBridge creates:\n\nF in MOI.Zeros\nMOI.VectorOfVariables in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ScalarFunctionizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ScalarFunctionizeBridge","text":"ScalarFunctionizeBridge{T,S} = FunctionConversionBridge{T,MOI.ScalarAffineFunction{T},MOI.VariableIndex,S}\n\nScalarFunctionizeBridge implements the following reformulations:\n\nx in S into 1x + 0 in S\n\nSource node\n\nScalarFunctionizeBridge supports:\n\nMOI.VariableIndex in S\n\nTarget nodes\n\nScalarFunctionizeBridge creates:\n\nMOI.ScalarAffineFunction{T} in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.VectorFunctionizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.VectorFunctionizeBridge","text":"VectorFunctionizeBridge{T,S} = FunctionConversionBridge{T,MOI.VectorAffineFunction{T},S}\n\nVectorFunctionizeBridge implements the following reformulations:\n\nx in S into Ix + 0 in S\n\nSource node\n\nVectorFunctionizeBridge supports:\n\nMOI.VectorOfVariables in S\n\nTarget nodes\n\nVectorFunctionizeBridge creates:\n\nMOI.VectorAffineFunction{T} in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToScalarQuadraticBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ToScalarQuadraticBridge","text":"ToScalarQuadraticBridge{T,G,S} <: AbstractFunctionConversionBridge{G,S}\n\nToScalarQuadraticBridge implements the following reformulation:\n\ng(x) in S into f(x) in S\n\nwhere g is an abstract scalar function and f is a MOI.ScalarQuadraticFunction.\n\nSource node\n\nToScalarQuadraticBridge supports:\n\nG<:AbstractScalarFunction in S\n\nTarget nodes\n\nToScalarQuadraticBridge creates:\n\nMOI.ScalarQuadraticFunction in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToVectorQuadraticBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ToVectorQuadraticBridge","text":"ToVectorQuadraticBridge{T,G,S} <: AbstractFunctionConversionBridge{G,S}\n\nToVectorQuadraticBridge implements the following reformulation:\n\ng(x) in S into f(x) in S\n\nwhere g is an abstract vector function and f is a MOI.VectorQuadraticFunction.\n\nSource node\n\nToVectorQuadraticBridge supports:\n\nG<:AbstractVectorFunction in S\n\nTarget nodes\n\nToVectorQuadraticBridge creates:\n\nMOI.VectorQuadraticFunction in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ToScalarNonlinearBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ToScalarNonlinearBridge","text":"ToScalarNonlinearBridge{T,G,S} <: AbstractFunctionConversionBridge{G,S}\n\nToScalarNonlinearBridge implements the following reformulation:\n\ng(x) in S into f(x) in S\n\nwhere g is an abstract scalar function and f is a MOI.ScalarNonlinearFunction.\n\nSource node\n\nToScalarNonlinearBridge supports:\n\nG<:AbstractScalarFunction in S\n\nTarget nodes\n\nToScalarNonlinearBridge creates:\n\nMOI.ScalarNonlinearFunction in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.FunctionConversionBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.FunctionConversionBridge","text":"FunctionConversionBridge{T,F,G,S} <: AbstractFunctionConversionBridge{G,S}\n\nFunctionConversionBridge implements the following reformulations:\n\ng(x) in S into f(x) in S\n\nfor these pairs of functions:\n\nMOI.ScalarAffineFunctionto [MOI.ScalarQuadraticFunction`](@ref)\nMOI.ScalarQuadraticFunction  to MOI.ScalarNonlinearFunction\nMOI.VectorAffineFunction to MOI.VectorQuadraticFunction\n\nSource node\n\nFunctionConversionBridge supports:\n\nG in S\n\nTarget nodes\n\nFunctionConversionBridge creates:\n\nF in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitComplexEqualToBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SplitComplexEqualToBridge","text":"SplitComplexEqualToBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nSplitComplexEqualToBridge implements the following reformulation:\n\nf(x) + g(x) * im = a + b * im into f(x) = a and g(x) = b\n\nSource node\n\nSplitComplexEqualToBridge supports:\n\nG in MOI.EqualTo{Complex{T}}\n\nwhere G is a function with Complex coefficients.\n\nTarget nodes\n\nSplitComplexEqualToBridge creates:\n\nF in MOI.EqualTo{T}\n\nwhere F is the type of the real/imaginary part of G.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitComplexZerosBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SplitComplexZerosBridge","text":"SplitComplexZerosBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nSplitComplexZerosBridge implements the following reformulation:\n\nf(x) in 0^n into textRe(f(x)) in 0^n and textIm(f(x)) in 0^n\n\nSource node\n\nSplitComplexZerosBridge supports:\n\nG in MOI.Zeros\n\nwhere G is a function with Complex coefficients.\n\nTarget nodes\n\nSplitComplexZerosBridge creates:\n\nF in MOI.Zeros\n\nwhere F is the type of the real/imaginary part of G.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitHyperRectangleBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SplitHyperRectangleBridge","text":"SplitHyperRectangleBridge{T,G,F} <: Bridges.Constraint.AbstractBridge\n\nSplitHyperRectangleBridge implements the following reformulation:\n\nf(x) in textsfHyperRectangle(l u) to f(x) - l u - f(x) in mathbbR_+.\n\nSource node\n\nSplitHyperRectangleBridge supports:\n\nF in MOI.HyperRectangle\n\nTarget nodes\n\nSplitHyperRectangleBridge creates:\n\nG in MOI.Nonnegatives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SplitIntervalBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SplitIntervalBridge","text":"SplitIntervalBridge{T,F,S,LS,US} <: Bridges.Constraint.AbstractBridge\n\nSplitIntervalBridge implements the following reformulations:\n\nl le f(x) le u into f(x) ge l and f(x) le u\nf(x) = b into f(x) ge b and f(x) le b\nf(x) in 0 into f(x) in mathbbR_+ and f(x) in mathbbR_-\n\nSource node\n\nSplitIntervalBridge supports:\n\nF in MOI.Interval{T}\nF in MOI.EqualTo{T}\nF in MOI.Zeros\n\nTarget nodes\n\nSplitIntervalBridge creates:\n\nF in MOI.LessThan{T}\nF in MOI.GreaterThan{T}\n\nor\n\nF in MOI.Nonnegatives\nF in MOI.Nonpositives\n\nnote: Note\nIf T<:AbstractFloat and S is MOI.Interval{T} then no lower (resp. upper) bound constraint is created if the lower (resp. upper) bound is typemin(T) (resp. typemax(T)). Similarly, when MOI.ConstraintSet is set, a lower or upper bound constraint may be deleted or created accordingly.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoRSOCBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SOCtoRSOCBridge","text":"SOCtoRSOCBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nSOCtoRSOCBridge implements the following reformulation:\n\nx_2 le t into (t+x_1)(t-x_1)ge (x_2ldotsx_N)_2^2\n\nAssumptions\n\nSOCtoRSOCBridge assumes that the length of x is at least one.\n\nSource node\n\nSOCtoRSOCBridge supports:\n\nG in MOI.SecondOrderCone\n\nTarget node\n\nSOCtoRSOCBridge creates:\n\nF in MOI.RotatedSecondOrderCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.RSOCtoSOCBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.RSOCtoSOCBridge","text":"RSOCtoSOCBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nRSOCtoSOCBridge implements the following reformulation:\n\nx_2^2 le 2tu into fract - usqrt 2 x_2 le  fract + usqrt 2\n\nSource node\n\nRSOCtoSOCBridge supports:\n\nG in MOI.RotatedSecondOrderCone\n\nTarget node\n\nRSOCtoSOCBridge creates:\n\nF in MOI.SecondOrderCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SOCtoNonConvexQuadBridge","text":"SOCtoNonConvexQuadBridge{T} <: Bridges.Constraint.AbstractBridge\n\nSOCtoNonConvexQuadBridge implements the following reformulations:\n\nx_2 le t into sum x^2 - t^2 le 0 and 1t + 0 ge 0\n\nThe MOI.ScalarAffineFunction 1t + 0 is used in case the variable has other bound constraints.\n\nwarning: Warning\nThis transformation starts from a convex constraint and creates a non-convex constraint. Unless the solver has explicit support for detecting second-order cones in quadratic form, this may (wrongly) be interpreted by the solver as being non-convex. Therefore, this bridge is not added automatically by MOI.Bridges.full_bridge_optimizer. Care is recommended when adding this bridge to a optimizer.\n\nSource node\n\nSOCtoNonConvexQuadBridge supports:\n\nMOI.VectorOfVariables in MOI.SecondOrderCone\n\nTarget nodes\n\nSOCtoNonConvexQuadBridge creates:\n\nMOI.ScalarQuadraticFunction{T} in MOI.LessThan{T}\nMOI.ScalarAffineFunction{T} in MOI.GreaterThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.RSOCtoNonConvexQuadBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.RSOCtoNonConvexQuadBridge","text":"RSOCtoNonConvexQuadBridge{T} <: Bridges.Constraint.AbstractBridge\n\nRSOCtoNonConvexQuadBridge implements the following reformulations:\n\nx_2^2 le 2tu into sum x^2 - 2tu le 0, 1t + 0 ge 0, and 1u + 0 ge 0.\n\nThe MOI.ScalarAffineFunctions 1t + 0 and 1u + 0 are used in case the variables have other bound constraints.\n\nwarning: Warning\nThis transformation starts from a convex constraint and creates a non-convex constraint. Unless the solver has explicit support for detecting rotated second-order cones in quadratic form, this may (wrongly) be interpreted by the solver as being non-convex. Therefore, this bridge is not added automatically by MOI.Bridges.full_bridge_optimizer. Care is recommended when adding this bridge to a optimizer.\n\nSource node\n\nRSOCtoNonConvexQuadBridge supports:\n\nMOI.VectorOfVariables in MOI.RotatedSecondOrderCone\n\nTarget nodes\n\nRSOCtoNonConvexQuadBridge creates:\n\nMOI.ScalarQuadraticFunction{T} in MOI.LessThan{T}\nMOI.ScalarAffineFunction{T} in MOI.GreaterThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.QuadtoSOCBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.QuadtoSOCBridge","text":"QuadtoSOCBridge{T} <: Bridges.Constraint.AbstractBridge\n\nQuadtoSOCBridge converts quadratic inequalities\n\nfrac12x^T Q x + a^T x le ub\n\ninto MOI.RotatedSecondOrderCone constraints, but it only applies when Q is positive definite.\n\nThis is because, if Q is positive definite, there exists U such that Q = U^T U, and so the inequality can then be rewritten as;\n\nU x_2^2 le 2 (-a^T x + ub)\n\nTherefore, QuadtoSOCBridge implements the following reformulations:\n\nfrac12x^T Q x + a^T x le ub into (1 -a^T x + ub Ux) in RotatedSecondOrderCone where Q = U^T U\nfrac12x^T Q x + a^T x ge lb into (1 a^T x - lb Ux) in RotatedSecondOrderCone where -Q = U^T U\n\nSource node\n\nQuadtoSOCBridge supports:\n\nMOI.ScalarAffineFunction{T} in MOI.LessThan{T}\nMOI.ScalarAffineFunction{T} in MOI.GreaterThan{T}\n\nTarget nodes\n\nRelativeEntropyBridge creates:\n\nMOI.VectorAffineFunction{T} in MOI.RotatedSecondOrderCone\n\nErrors\n\nThis bridge errors if Q is not positive definite.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SOCtoPSDBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SOCtoPSDBridge","text":"SOCtoPSDBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nSOCtoPSDBridge implements the following reformulation:\n\nx_2 le t into leftbeginarrayc ct  x^top  x  t mathbfIendarrayrightsucceq 0\n\nwarning: Warning\nThis bridge is not added by default by MOI.Bridges.full_bridge_optimizer because bridging second order cone constraints to semidefinite constraints can be achieved by the SOCtoRSOCBridge followed by the RSOCtoPSDBridge, while creating a smaller semidefinite constraint.\n\nSource node\n\nSOCtoPSDBridge supports:\n\nG in MOI.SecondOrderCone\n\nTarget nodes\n\nSOCtoPSDBridge creates:\n\nF in MOI.PositiveSemidefiniteConeTriangle\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.RSOCtoPSDBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.RSOCtoPSDBridge","text":"RSOCtoPSDBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nRSOCtoPSDBridge implements the following reformulation:\n\nx_2^2 le 2tcdot u into leftbeginarrayc ct  x^top  x  2tu mathbfIendarrayrightsucceq 0\n\nSource node\n\nRSOCtoPSDBridge supports:\n\nG in MOI.RotatedSecondOrderCone\n\nTarget nodes\n\nRSOCtoPSDBridge creates:\n\nF in MOI.PositiveSemidefiniteConeTriangle\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NormInfinityBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NormInfinityBridge","text":"NormInfinityBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nNormInfinityBridge implements the following reformulation:\n\nx_infty le t into t - x_i t + x_i in mathbbR_+.\n\nSource node\n\nNormInfinityBridge supports:\n\nG in MOI.NormInfinityCone{T}\n\nTarget nodes\n\nNormInfinityBridge creates:\n\nF in MOI.Nonnegatives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NormOneBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NormOneBridge","text":"NormOneBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nNormOneBridge implements the following reformulation:\n\nsum x_i le t into t - sum y_i y_i - x_i y_i + x_i in mathbbR_+.\n\nSource node\n\nNormOneBridge supports:\n\nG in MOI.NormOneCone{T}\n\nTarget nodes\n\nNormOneBridge creates:\n\nF in MOI.Nonnegatives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NormToPowerBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NormToPowerBridge","text":"NormToPowerBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nNormToPowerBridge implements the following reformulation:\n\n(t x) in NormCone(p 1+d) into (r_i t x_i) in PowerCone(1  p) for all i, and sumlimits_i r_i == t.\n\nFor details, see Alizadeh, F., and Goldfarb, D. (2001). \"Second-order cone programming.\" Mathematical Programming, Series B, 95:3-51.\n\nSource node\n\nNormToPowerBridge supports:\n\nF in MOI.NormCone\n\nTarget nodes\n\nNormToPowerBridge creates:\n\nF in MOI.PowerCone{T}\nMOI.ScalarAffineFunction in MOI.EqualTo\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NormOneConeToNormConeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NormOneConeToNormConeBridge","text":"NormOneConeToNormConeBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nNormOneConeToNormConeBridge implements the following reformulations:\n\n(t x) in NormOneCone(d) into (t x) in NormCone(1 d)\n\nSource node\n\nNormOneConeToNormConeBridge supports:\n\nF in MOI.NormOneCone\n\nTarget nodes\n\nNormOneConeToNormConeBridge creates:\n\nF in MOI.NormCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SecondOrderConeToNormConeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SecondOrderConeToNormConeBridge","text":"SecondOrderConeToNormConeBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nSecondOrderConeToNormConeBridge implements the following reformulations:\n\n(t x) in SecondOrderCone(d) into (t x) in NormCone(2 d)\n\nSource node\n\nSecondOrderConeToNormConeBridge supports:\n\nF in MOI.SecondOrderCone\n\nTarget nodes\n\nSecondOrderConeToNormConeBridge creates:\n\nF in MOI.NormCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NormInfinityConeToNormConeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NormInfinityConeToNormConeBridge","text":"NormInfinityConeToNormConeBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nNormInfinityConeToNormConeBridge implements the following reformulations:\n\n(t x) in NormInfinityCone(d) into (t x) in NormCone(Inf d)\n\nSource node\n\nNormInfinityConeToNormConeBridge supports:\n\nF in MOI.NormInfinityCone\n\nTarget nodes\n\nNormInfinityConeToNormConeBridge creates:\n\nF in MOI.NormCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeantoRelEntrBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.GeoMeantoRelEntrBridge","text":"GeoMeantoRelEntrBridge{T,F,G,H} <: Bridges.Constraint.AbstractBridge\n\nGeoMeantoRelEntrBridge implements the following reformulation:\n\n(u w) in GeometricMeanCone into (0 w (u + y)mathbf1)in RelativeEntropyCone and y ge 0\n\nSource node\n\nGeoMeantoRelEntrBridge supports:\n\nH in MOI.GeometricMeanCone\n\nTarget nodes\n\nGeoMeantoRelEntrBridge creates:\n\nG in MOI.RelativeEntropyCone\nF in MOI.Nonnegatives\n\nDerivation\n\nThe derivation of the bridge is as follows:\n\nbeginaligned\n(u w) in GeometricMeanCone iff  u le left(prod_i=1^n w_iright)^1n \niff  0 le u + y le left(prod_i=1^n w_iright)^1n y ge 0 \niff  1 le fracleft(prod_i=1^n w_iright)^1nu + y y ge 0 \niff  1 le left(prod_i=1^n fracw_iu + yright)^1n y ge 0 \niff  0 le sum_i=1^n logleft(fracw_iu + yright) y ge 0 \niff  0 ge sum_i=1^n logleft(fracu + yw_iright) y ge 0 \niff  0 ge sum_i=1^n (u + y) logleft(fracu + yw_iright) y ge 0 \niff  (0 w (u + y)mathbf1) in RelativeEntropyCone y ge 0 \nendaligned\n\nThis derivation assumes that u + y  0, which is enforced by the relative entropy cone.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge","text":"GeoMeanToPowerBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nGeoMeanToPowerBridge implements the following reformulation:\n\n(y x) in GeometricMeanCone(1+d) into (x_1 t y) in PowerCone(1d) and (t x_2  x_d) in GeometricMeanCone(d), which is then recursively expanded into more PowerCone constraints.\n\nSource node\n\nGeoMeanToPowerBridge supports:\n\nF in MOI.GeometricMeanCone\n\nTarget nodes\n\nGeoMeanToPowerBridge creates:\n\nF in MOI.PowerCone{T}\nMOI.VectorOfVariables in MOI.Nonnegatives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeanBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.GeoMeanBridge","text":"GeoMeanBridge{T,F,G,H} <: Bridges.Constraint.AbstractBridge\n\nGeoMeanBridge implements a reformulation from MOI.GeometricMeanCone into MOI.RotatedSecondOrderCone.\n\nThe reformulation is best described in an example.\n\nConsider the cone of dimension 4:\n\nt le sqrt3x_1 x_2 x_3\n\nThis can be rewritten as exists y ge 0 such that:\n\nbeginalign*\n  t  le y\n  y^4  le x_1 x_2 x_3 y\nendalign*\n\nNote that we need to create y and not use t^4 directly because t is allowed to be negative.\n\nThis is equivalent to:\n\nbeginalign*\n    t  le fracy_1sqrt4\n    y_1^2  le 2y_2 y_3\n    y_2^2  le 2x_1 x_2 \n    y_3^2  le 2x_3(y_1sqrt4) \n    y      ge 0\nendalign*\n\nMore generally, you can show how the geometric mean code is recursively expanded into a set of new variables y in MOI.Nonnegatives, a set of MOI.RotatedSecondOrderCone constraints, and a MOI.LessThan constraint between t and y_1.\n\nSource node\n\nGeoMeanBridge supports:\n\nH in MOI.GeometricMeanCone\n\nTarget nodes\n\nGeoMeanBridge creates:\n\nF in MOI.LessThan{T}\nG in MOI.RotatedSecondOrderCone\nG in MOI.Nonnegatives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.RelativeEntropyBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.RelativeEntropyBridge","text":"RelativeEntropyBridge{T,F,G,H} <: Bridges.Constraint.AbstractBridge\n\nRelativeEntropyBridge implements the following reformulation that converts a MOI.RelativeEntropyCone into an MOI.ExponentialCone:\n\nu ge sum_i=1^n w_i log left(fracw_iv_iright) into y_i ge 0, u ge sum_i=1^n y_i, and (-y_i w_i v_i) in ExponentialCone.\n\nSource node\n\nRelativeEntropyBridge supports:\n\nH in MOI.RelativeEntropyCone\n\nTarget nodes\n\nRelativeEntropyBridge creates:\n\nF in MOI.GreaterThan{T}\nG in MOI.ExponentialCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NormSpectralBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NormSpectralBridge","text":"NormSpectralBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nNormSpectralBridge implements the following reformulation:\n\nt ge sigma_1(X) into leftbeginarrayc ctmathbfI  X^top  X  t mathbfIendarrayright succeq 0\n\nSource node\n\nNormSpectralBridge supports:\n\nG in MOI.NormSpectralCone\n\nTarget nodes\n\nNormSpectralBridge creates:\n\nF in MOI.PositiveSemidefiniteConeTriangle\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NormNuclearBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NormNuclearBridge","text":"NormNuclearBridge{T,F,G,H} <: Bridges.Constraint.AbstractBridge\n\nNormNuclearBridge implements the following reformulation:\n\nt ge sum_i sigma_i (X) into leftbeginarrayc cU  X^top  X  Vendarrayright succeq 0 and 2t ge tr(U) + tr(V).\n\nSource node\n\nNormNuclearBridge supports:\n\nH in MOI.NormNuclearCone\n\nTarget nodes\n\nNormNuclearBridge creates:\n\nF in MOI.GreaterThan{T}\nG in MOI.PositiveSemidefiniteConeTriangle\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SquareBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SquareBridge","text":"SquareBridge{T,F,G,TT,ST} <: Bridges.Constraint.AbstractBridge\n\nSquareBridge implements the following reformulations:\n\n(t u X) in LogDetConeSquare into (t u Y) in LogDetConeTriangle\n(t X) in RootDetConeSquare into (t Y) in RootDetConeTriangle\nX in AbstractSymmetricMatrixSetSquare into Y in AbstractSymmetricMatrixSetTriangle\n\nwhere Y is the upper triangluar component of X.\n\nIn addition, constraints are added as necessary to constrain the matrix X to be symmetric. For example, the constraint for the matrix:\n\nbeginpmatrix\n  1       1 + x  2 - 3x\n  1 +  x  2 + x  3 -  x\n  2 - 3x  2 + x      2x\nendpmatrix\n\ncan be broken down to the constraint of the symmetric matrix\n\nbeginpmatrix\n  1       1 + x  2 - 3x\n  cdot  2 + x  3 -  x\n  cdot  cdot     2x\nendpmatrix\n\nand the equality constraint between the off-diagonal entries (2, 3) and (3, 2) 3 - x == 2 + x. Note that no symmetrization constraint needs to be added between the off-diagonal entries (1, 2) and (2, 1) or between (1, 3) and (3, 1) because the expressions are the same.\n\nSource node\n\nSquareBridge supports:\n\nF in ST\n\nTarget nodes\n\nSquareBridge creates:\n\nG in TT\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.HermitianToSymmetricPSDBridge","text":"HermitianToSymmetricPSDBridge{T,F,G} <: Bridges.Constraint.AbstractBridge\n\nHermitianToSymmetricPSDBridge implements the following reformulation:\n\nHermitian positive semidefinite n x n complex matrix to a symmetric positive semidefinite 2n x 2n real matrix.\n\nSee also MOI.Bridges.Variable.HermitianToSymmetricPSDBridge.\n\nSource node\n\nHermitianToSymmetricPSDBridge supports:\n\nG in MOI.HermitianPositiveSemidefiniteConeTriangle\n\nTarget node\n\nHermitianToSymmetricPSDBridge creates:\n\nF in MOI.PositiveSemidefiniteConeTriangle\n\nReformulation\n\nThe reformulation is best described by example.\n\nThe Hermitian matrix:\n\nbeginbmatrix\n  x_11             x_12 + y_12im  x_13 + y_13im\n  x_12 - y_12im  x_22             x_23 + y_23im\n  x_13 - y_13im  x_23 - y_23im  x_33\nendbmatrix\n\nis positive semidefinite if and only if the symmetric matrix:\n\nbeginbmatrix\n    x_11  x_12  x_13  0        y_12   y_13 \n            x_22  x_23  -y_12  0        y_23 \n                    x_33  -y_13  -y_23  0      \n                            x_11   x_12   x_13 \n                                     x_22   x_23 \n                                              x_33\nendbmatrix\n\nis positive semidefinite.\n\nThe bridge achieves this reformulation by constraining the above matrix to belong to the MOI.PositiveSemidefiniteConeTriangle(6).\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SetDotScalingBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SetDotScalingBridge","text":"SetDotScalingBridge{T,S,F,G} <: Bridges.Constraint.AbstractBridge\n\nSetDotScalingBridge implements the reformulation from constraints in S to constraints in MOI.Scaled{S}.\n\nSource node\n\nSetDotScalingBridge supports:\n\nG in S\n\nTarget node\n\nSetDotScalingBridge creates:\n\nF in MOI.Scaled{S}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SetDotInverseScalingBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SetDotInverseScalingBridge","text":"SetDotInverseScalingBridge{T,S,F,G} <: Bridges.Constraint.AbstractBridge\n\nSetDotInverseScalingBridge implements the reformulation from constraints in the MOI.Scaled{S} to constraints in the S.\n\nSource node\n\nSetDotInverseScalingBridge supports:\n\nG in MOI.Scaled{S}\n\nTarget node\n\nSetDotInverseScalingBridge creates:\n\nF in S\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.RootDetBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.RootDetBridge","text":"RootDetBridge{T,F,G,H} <: Bridges.Constraint.AbstractBridge\n\nThe MOI.RootDetConeTriangle is representable by MOI.PositiveSemidefiniteConeTriangle and MOI.GeometricMeanCone constraints, see [1, p. 149].\n\nIndeed, t le det(X)^1n if and only if there exists a lower triangular matrix Δ such that:\n\nbeginalign*\n  beginpmatrix\n    X  Δ\n    Δ^top  mathrmDiag(Δ)\n  endpmatrix  succeq 0\n  (t mathrmDiag(Δ))  in GeometricMeanCone\nendalign*\n\nSource node\n\nRootDetBridge supports:\n\nI in MOI.RootDetConeTriangle\n\nTarget nodes\n\nRootDetBridge creates:\n\nF in MOI.PositiveSemidefiniteConeTriangle\nG in MOI.GeometricMeanCone\n\n[1] Ben-Tal, Aharon, and Arkadi Nemirovski. Lectures on modern convex     optimization: analysis, algorithms, and engineering applications. Society     for Industrial and Applied Mathematics, 2001.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.LogDetBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.LogDetBridge","text":"LogDetBridge{T,F,G,H,I} <: Bridges.Constraint.AbstractBridge\n\nThe MOI.LogDetConeTriangle is representable by MOI.PositiveSemidefiniteConeTriangle and MOI.ExponentialCone constraints.\n\nIndeed, logdet(X) = sumlimits_i=1^n log(delta_i) where delta_i are the eigenvalues of X.\n\nAdapting the method from [1, p. 149], we see that t le u log(det(Xu)) for u  0 if and only if there exists a lower triangular matrix Δ such that\n\nbeginalign*\n  beginpmatrix\n    X  Δ\n    Δ^top  mathrmDiag(Δ)\n  endpmatrix  succeq 0\n  t - sum_i=1^n u logleft(fracΔ_iiuright)  le 0\nendalign*\n\nWhich we reformulate further into\n\nbeginalign*\n  beginpmatrix\n    X  Δ\n    Δ^top  mathrmDiag(Δ)\n  endpmatrix  succeq 0\n  (l_i u  Δ_ii)  in ExponentialConequad forall i \n  t - sum_i=1^n l_i  le 0\nendalign*\n\nSource node\n\nLogDetBridge supports:\n\nI in MOI.LogDetConeTriangle\n\nTarget nodes\n\nLogDetBridge creates:\n\nF in MOI.PositiveSemidefiniteConeTriangle\nG in MOI.ExponentialCone\nH in MOI.LessThan{T}\n\n[1] Ben-Tal, Aharon, and Arkadi Nemirovski. Lectures on modern convex     optimization: analysis, algorithms, and engineering applications. Society     for Industrial and Applied Mathematics, 2001.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorActiveOnFalseBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.IndicatorActiveOnFalseBridge","text":"IndicatorActiveOnFalseBridge{T,F,S} <: Bridges.Constraint.AbstractBridge\n\nIndicatorActiveOnFalseBridge implements the following reformulation:\n\nneg z implies f(x) in S into y implies f(x) in S, z + y = 1, and y in 0 1\n\nSource node\n\nIndicatorActiveOnFalseBridge supports:\n\nMOI.VectorAffineFunction{T} in MOI.Indicator{MOI.ACTIVATE_ON_ZERO,S}\n\nTarget nodes\n\nIndicatorActiveOnFalseBridge creates:\n\nMOI.VectorAffineFunction{T} in MOI.Indicator{MOI.ACTIVATE_ON_ONE,S}\nMOI.ScalarAffineFunction{T} in MOI.EqualTo\nMOI.VariableIndex in MOI.ZeroOne\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorGreaterToLessThanBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.IndicatorGreaterToLessThanBridge","text":"IndicatorGreaterToLessThanBridge{T,A} <: Bridges.Constraint.AbstractBridge\n\nIndicatorGreaterToLessThanBridge implements the following reformulation:\n\nz implies f(x) ge l into z implies -f(x) le -l\n\nSource node\n\nIndicatorGreaterToLessThanBridge supports:\n\nMOI.VectorAffineFunction{T} in MOI.Indicator{A,MOI.GreaterThan{T}}\n\nTarget nodes\n\nIndicatorGreaterToLessThanBridge creates:\n\nMOI.VectorAffineFunction{T} in MOI.Indicator{A,MOI.LessThan{T}}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorLessToGreaterThanBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.IndicatorLessToGreaterThanBridge","text":"IndicatorLessToGreaterThanBridge{T,A} <: Bridges.Constraint.AbstractBridge\n\nIndicatorLessToGreaterThanBridge implements the following reformulations:\n\nz implies f(x) le u into z implies -f(x) ge -u\n\nSource node\n\nIndicatorLessToGreaterThanBridge supports:\n\nMOI.VectorAffineFunction{T} in MOI.Indicator{A,MOI.LessThan{T}}\n\nTarget nodes\n\nIndicatorLessToGreaterThanBridge creates:\n\nMOI.VectorAffineFunction{T} in MOI.Indicator{A,MOI.GreaterThan{T}}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IndicatorSOS1Bridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.IndicatorSOS1Bridge","text":"IndicatorSOS1Bridge{T,S} <: Bridges.Constraint.AbstractBridge\n\nIndicatorSOS1Bridge implements the following reformulation:\n\nz implies f(x) in S into f(x) + y in S, SOS1(y z)\n\nwarning: Warning\nThis bridge assumes that the solver supports MOI.SOS1{T} constraints in which one of the variables (y) is continuous.\n\nSource node\n\nIndicatorSOS1Bridge supports:\n\nMOI.VectorAffineFunction{T} in MOI.Indicator{MOI.ACTIVATE_ON_ONE,S}\n\nTarget nodes\n\nIndicatorSOS1Bridge creates:\n\nMOI.ScalarAffineFunction{T} in S\nMOI.VectorOfVariables in MOI.SOS1{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.SemiToBinaryBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.SemiToBinaryBridge","text":"SemiToBinaryBridge{T,S} <: Bridges.Constraint.AbstractBridge\n\nSemiToBinaryBridge implements the following reformulations:\n\nx in 0 cup l u into\nbeginaligned\nx leq z u \nx geq z l \nz in 0 1\nendaligned\nx in 0 cup l ldots u into\nbeginaligned\nx leq z u \nx geq z l \nz in 0 1 \nx in mathbbZ\nendaligned\n\nSource node\n\nSemiToBinaryBridge supports:\n\nMOI.VariableIndex in MOI.Semicontinuous{T}\nMOI.VariableIndex in MOI.Semiinteger{T}\n\nTarget nodes\n\nSemiToBinaryBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.LessThan{T}\nMOI.ScalarAffineFunction{T} in MOI.GreaterThan{T}\nMOI.VariableIndex{T} in MOI.Integer (if S is MOI.Semiinteger{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ZeroOneBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ZeroOneBridge","text":"ZeroOneBridge{T} <: Bridges.Constraint.AbstractBridge\n\nZeroOneBridge implements the following reformulation:\n\nx in 0 1 into x in mathbbZ, 1x in 0 1.\n\nnote: Note\nZeroOneBridge adds a linear constraint instead of adding variable bounds to avoid conflicting with bounds set by the user.\n\nSource node\n\nZeroOneBridge supports:\n\nMOI.VariableIndex in MOI.ZeroOne\n\nTarget nodes\n\nZeroOneBridge creates:\n\nMOI.VariableIndex in MOI.Integer\nMOI.ScalarAffineFunction{T} in MOI.Interval{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.IntegerToZeroOneBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.IntegerToZeroOneBridge","text":"IntegerToZeroOneBridge{T} <: Bridges.Constraint.AbstractBridge\n\nIntegerToZeroOneBridge implements the following reformulation:\n\nx in mathbfZ into y_i in 0 1, x == lb + sum 2^i-1 y_i.\n\nSource node\n\nIntegerToZeroOneBridge supports:\n\nVariableIndex in MOI.Integer\n\nTarget nodes\n\nIntegerToZeroOneBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\n\nDeveloper note\n\nThis bridge is implemented as a constraint bridge instead of a variable bridge because we don't want to substitute the linear combination of y for every instance of x. Doing so would be expensive and greatly reduce the sparsity of the constraints.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.NumberConversionBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.NumberConversionBridge","text":"NumberConversionBridge{T,F1,S1,F2,S2} <: Bridges.Constraint.AbstractBridge\n\nNumberConversionBridge implements the following reformulation:\n\nf1(x) in S1 to f2(x) in S2\n\nwhere f and S are the same functional form, but differ in their coefficient type.\n\nSource node\n\nNumberConversionBridge supports:\n\nF1 in S1\n\nTarget node\n\nNumberConversionBridge creates:\n\nF2 in S2\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.AllDifferentToCountDistinctBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.AllDifferentToCountDistinctBridge","text":"AllDifferentToCountDistinctBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nAllDifferentToCountDistinctBridge implements the following reformulations:\n\nx in textsfAllDifferent(d) to (n x) in textsfCountDistinct(1+d) and n = d\nf(x) in textsfAllDifferent(d) to (d f(x)) in textsfCountDistinct(1+d)\n\nSource node\n\nAllDifferentToCountDistinctBridge supports:\n\nF in MOI.AllDifferent\n\nwhere F is MOI.VectorOfVariables or MOI.VectorAffineFunction{T}.\n\nTarget nodes\n\nAllDifferentToCountDistinctBridge creates:\n\nF in MOI.CountDistinct\nMOI.VariableIndex in MOI.EqualTo{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ReifiedAllDifferentToCountDistinctBridge","text":"ReifiedAllDifferentToCountDistinctBridge{T,F} <:\nBridges.Constraint.AbstractBridge\n\nReifiedAllDifferentToCountDistinctBridge implements the following reformulations:\n\nr iff x in textsfAllDifferent(d) to r iff (n x) in textsfCountDistinct(1+d) and n = d\nr iff f(x) in textsfAllDifferent(d) to r iff (d f(x)) in textsfCountDistinct(1+d)\n\nSource node\n\nReifiedAllDifferentToCountDistinctBridge supports:\n\nF in MOI.Reified{MOI.AllDifferent}\n\nwhere F is MOI.VectorOfVariables or MOI.VectorAffineFunction{T}.\n\nTarget nodes\n\nReifiedAllDifferentToCountDistinctBridge creates:\n\nF in MOI.Reified{MOI.CountDistinct}\nMOI.VariableIndex in MOI.EqualTo{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.BinPackingToMILPBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.BinPackingToMILPBridge","text":"BinPackingToMILPBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nBinPackingToMILPBridge implements the following reformulation:\n\nx in BinPacking(c w) into a mixed-integer linear program.\n\nReformulation\n\nThe reformulation is non-trivial, and it depends on the finite domain of each variable x_i, which we as define S_i = l_ildotsu_i.\n\nFirst, we introduce new binary variables z_ij, which are 1 if variable x_i takes the value j in the optimal solution and 0 otherwise:\n\nbeginaligned\nz_ij in 0 1                                forall i in 1ldots d j in S_i  \nx_i - sumlimits_jin S_i j cdot z_ij = 0   forall i in 1ldots d              \nsumlimits_jin S_i z_ij = 1                  forall i in 1ldots d              \nendaligned\n\nThen, we add the capacity constraint for all possible bins j:\n\nsumlimits_i w_i z_ij le c forall j in bigcup_i=1ldotsd S_i\n\nSource node\n\nBinPackingToMILPBridge supports:\n\nF in MOI.BinPacking{T}\n\nTarget nodes\n\nBinPackingToMILPBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\nMOI.ScalarAffineFunction{T} in MOI.LessThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.CircuitToMILPBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.CircuitToMILPBridge","text":"CircuitToMILPBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nCircuitToMILPBridge implements the following reformulation:\n\nx in textsfCircuit(d) to the Miller-Tucker-Zemlin formulation of the Traveling Salesperson Problem.\n\nSource node\n\nCircuitToMILPBridge supports:\n\nF in MOI.Circuit\n\nwhere F is MOI.VectorOfVariables or MOI.VectorAffineFunction{T}.\n\nTarget nodes\n\nCircuitToMILPBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{T}\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\nMOI.ScalarAffineFunction{T} in MOI.LessThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.CountAtLeastToCountBelongsBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.CountAtLeastToCountBelongsBridge","text":"CountAtLeastToCountBelongsBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nCountAtLeastToCountBelongsBridge implements the following reformulation:\n\nx in textsfCountAtLeast(n d mathcalS) to (n_i x_d_i) in textsfCountBelongs(1+d mathcalS) and n_i ge n for all i.\n\nSource node\n\nCountAtLeastToCountBelongsBridge supports:\n\nF in MOI.CountAtLeast\n\nwhere F is MOI.VectorOfVariables or MOI.VectorAffineFunction{T}.\n\nTarget nodes\n\nCountAtLeastToCountBelongsBridge creates:\n\nF in MOI.CountBelongs\nMOI.VariableIndex in MOI.GreaterThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.CountBelongsToMILPBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.CountBelongsToMILPBridge","text":"CountBelongsToMILPBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nCountBelongsToMILPBridge implements the following reformulation:\n\n(n x) in textsfCountBelongs(1+d mathcalS) into a mixed-integer linear program.\n\nReformulation\n\nThe reformulation is non-trivial, and it depends on the finite domain of each variable x_i, which we as define S_i = l_ildotsu_i.\n\nFirst, we introduce new binary variables z_ij, which are 1 if variable x_i takes the value j in the optimal solution and 0 otherwise:\n\nbeginaligned\nz_ij in 0 1                                forall i in 1ldots d j in S_i  \nx_i - sumlimits_jin S_i j cdot z_ij = 0   forall i in 1ldots d              \nsumlimits_jin S_i z_ij = 1                  forall i in 1ldots d              \nendaligned\n\nFinally, n is constrained to be the number of z_ij elements that are in mathcalS:\n\nn - sumlimits_iin 1ldots d j in mathcalS z_ij = 0\n\nSource node\n\nCountBelongsToMILPBridge supports:\n\nF in MOI.CountBelongs\n\nwhere F is MOI.VectorOfVariables or MOI.VectorAffineFunction{T}.\n\nTarget nodes\n\nCountBelongsToMILPBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.CountDistinctToMILPBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.CountDistinctToMILPBridge","text":"CountDistinctToMILPBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nCountDistinctToMILPBridge implements the following reformulation:\n\n(n x) in textsfCountDistinct(1+d) into a mixed-integer linear program.\n\nReformulation\n\nThe reformulation is non-trivial, and it depends on the finite domain of each variable x_i, which we as define S_i = l_ildotsu_i.\n\nFirst, we introduce new binary variables z_ij, which are 1 if variable x_i takes the value j in the optimal solution and 0 otherwise:\n\nbeginaligned\nz_ij in 0 1                                forall i in 1ldots d j in S_i  \nx_i - sumlimits_jin S_i j cdot z_ij = 0   forall i in 1ldots d              \nsumlimits_jin S_i z_ij = 1                  forall i in 1ldots d              \nendaligned\n\nThen, we introduce new binary variables y_j, which are 1 if a variable takes the value j in the optimal solution and 0 otherwise.\n\nbeginaligned\ny_j in 0 1                          forall j in bigcup_i=1ldotsd S_i \ny_j le sumlimits_i in 1ldots d j in S_i z_ij le M y_j   forall j in bigcup_i=1ldotsd S_i\nendaligned\n\nFinally, n is constrained to be the number of y_j elements that are non-zero:\n\nn - sumlimits_j in bigcup_i=1ldotsd S_i y_j = 0\n\nSource node\n\nCountDistinctToMILPBridge supports:\n\nF in MOI.CountDistinct\n\nwhere F is MOI.VectorOfVariables or MOI.VectorAffineFunction{T}.\n\nTarget nodes\n\nCountDistinctToMILPBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\nMOI.ScalarAffineFunction{T} in MOI.LessThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.ReifiedCountDistinctToMILPBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.ReifiedCountDistinctToMILPBridge","text":"ReifiedCountDistinctToMILPBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nReifiedCountDistinctToMILPBridge implements the following reformulation:\n\nr iff (n x) in textsfCountDistinct(1+d) into a mixed-integer linear program.\n\nReformulation\n\nThe reformulation is non-trivial, and it depends on the finite domain of each variable x_i, which we as define S_i = l_ildotsu_i.\n\nFirst, we introduce new binary variables z_ij, which are 1 if variable x_i takes the value j in the optimal solution and 0 otherwise:\n\nbeginaligned\nz_ij in 0 1                                forall i in 1ldots d j in S_i  \nx_i - sumlimits_jin S_i j cdot z_ij = 0   forall i in 1ldots d              \nsumlimits_jin S_i z_ij = 1                  forall i in 1ldots d              \nendaligned\n\nThen, we introduce new binary variables y_j, which are 1 if a variable takes the value j in the optimal solution and 0 otherwise.\n\nbeginaligned\ny_j in 0 1                          forall j in bigcup_i=1ldotsd S_i \ny_j le sumlimits_i in 1ldots d j in S_i z_ij le M y_j   forall j in bigcup_i=1ldotsd S_i\nendaligned\n\nFinally, n is constrained to be the number of y_j elements that are non-zero, with some slack:\n\nn - sumlimits_j in bigcup_i=1ldotsd S_i y_j = delta^+ - delta^-\n\nAnd then the slack is constrained to respect the reif variable r:\n\nbeginaligned\nd_1 le delta^+ le M d_1 \nd_2 le delta^- le M d_s \nd_1 + d_2 + r = 1             \nd_1 d_2 in 0 1\nendaligned\n\nSource node\n\nReifiedCountDistinctToMILPBridge supports:\n\nF in MOI.Reified{MOI.CountDistinct}\n\nwhere F is MOI.VectorOfVariables or MOI.VectorAffineFunction{T}.\n\nTarget nodes\n\nReifiedCountDistinctToMILPBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\nMOI.ScalarAffineFunction{T} in MOI.LessThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.CountGreaterThanToMILPBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.CountGreaterThanToMILPBridge","text":"CountGreaterThanToMILPBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nCountGreaterThanToMILPBridge implements the following reformulation:\n\n(c y x) in CountGreaterThan() into a mixed-integer linear program.\n\nSource node\n\nCountGreaterThanToMILPBridge supports:\n\nF in MOI.CountGreaterThan\n\nTarget nodes\n\nCountGreaterThanToMILPBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\nMOI.ScalarAffineFunction{T} in MOI.GreaterThan{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.TableToMILPBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Constraint.TableToMILPBridge","text":"TableToMILPBridge{T,F} <: Bridges.Constraint.AbstractBridge\n\nTableToMILPBridge implements the following reformulation:\n\nx in Table(t) into\nbeginaligned\nz_j in 0 1                      quad forall i j \nsumlimits_j=1^n z_j = 1                           \nsumlimits_j=1^n t_ij z_j = x_i  quad forall i\nendaligned\n\nSource node\n\nTableToMILPBridge supports:\n\nF in MOI.Table{T}\n\nTarget nodes\n\nTableToMILPBridge creates:\n\nMOI.VariableIndex in MOI.ZeroOne\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#objective_bridges_ref","page":"List of bridges","title":"Objective bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"These bridges are subtypes of Bridges.Objective.AbstractBridge.","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"Bridges.Objective.FunctionizeBridge\nBridges.Objective.QuadratizeBridge\nBridges.Objective.SlackBridge\nBridges.Objective.VectorFunctionizeBridge\nBridges.Objective.VectorSlackBridge","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.FunctionizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Objective.FunctionizeBridge","text":"FunctionizeBridge{T}\n\nFunctionizeBridge implements the following reformulations:\n\nmin x into min1x + 0\nmax x into max1x + 0\n\nwhere T is the coefficient type of 1 and 0.\n\nSource node\n\nFunctionizeBridge supports:\n\nMOI.ObjectiveFunction{MOI.VariableIndex}\n\nTarget nodes\n\nFunctionizeBridge creates:\n\nOne objective node: MOI.ObjectiveFunction{MOI.ScalarAffineFunction{T}}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.QuadratizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Objective.QuadratizeBridge","text":"QuadratizeBridge{T}\n\nQuadratizeBridge implements the following reformulations:\n\nmin a^top x + b into minx^top mathbf0 x + a^top x + b\nmax a^top x + b into maxx^top mathbf0 x + a^top x + b\n\nwhere T is the coefficient type of 0.\n\nSource node\n\nQuadratizeBridge supports:\n\nMOI.ObjectiveFunction{MOI.ScalarAffineFunction{T}}\n\nTarget nodes\n\nQuadratizeBridge creates:\n\nOne objective node: MOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{T}}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.SlackBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Objective.SlackBridge","text":"SlackBridge{T,F,G}\n\nSlackBridge implements the following reformulations:\n\nminf(x) into miny f(x) - y le 0\nmaxf(x) into maxy f(x) - y ge 0\n\nwhere F is the type of f(x) - y, G is the type of f(x), and T is the coefficient type of f(x).\n\nSource node\n\nSlackBridge supports:\n\nMOI.ObjectiveFunction{G}\n\nTarget nodes\n\nSlackBridge creates:\n\nOne variable node: MOI.VariableIndex in MOI.Reals\nOne objective node: MOI.ObjectiveFunction{MOI.VariableIndex}\nOne constraint node, that depends on the MOI.ObjectiveSense:\nF-in-MOI.LessThan if MIN_SENSE\nF-in-MOI.GreaterThan if MAX_SENSE\n\nwarning: Warning\nWhen using this bridge, changing the optimization sense is not supported. Set the sense to MOI.FEASIBILITY_SENSE first to delete the bridge, then set MOI.ObjectiveSense and re-add the objective.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.VectorFunctionizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Objective.VectorFunctionizeBridge","text":"VectorFunctionizeBridge{T}\n\nVectorFunctionizeBridge implements the following reformulations:\n\nmin x into min1x + 0\nmax x into max1x + 0\n\nwhere T is the coefficient type of 1 and 0.\n\nSource node\n\nVectorFunctionizeBridge supports:\n\nMOI.ObjectiveFunction{MOI.VectorOfVariables}\n\nTarget nodes\n\nVectorFunctionizeBridge creates:\n\nOne objective node: MOI.ObjectiveFunction{MOI.VectorAffineFunction{T}}\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.VectorSlackBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Objective.VectorSlackBridge","text":"VectorSlackBridge{T,F,G}\n\nVectorSlackBridge implements the following reformulations:\n\nminf(x) into miny y - f(x) in mathbbR_+ \nmaxf(x) into maxy f(x) - y in mathbbR_+ \n\nwhere F is the type of f(x) - y, G is the type of f(x), and T is the coefficient type of f(x).\n\nSource node\n\nVectorSlackBridge supports:\n\nMOI.ObjectiveFunction{G}\n\nTarget nodes\n\nVectorSlackBridge creates:\n\nOne variable node: MOI.VectorOfVariables in MOI.Reals\nOne objective node: MOI.ObjectiveFunction{MOI.VectorOfVariables}\nOne constraint node: F-in-MOI.Nonnegatives\n\nwarning: Warning\nWhen using this bridge, changing the optimization sense is not supported. Set the sense to MOI.FEASIBILITY_SENSE first to delete the bridge, then set MOI.ObjectiveSense and re-add the objective.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#variable_bridges_ref","page":"List of bridges","title":"Variable bridges","text":"","category":"section"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"These bridges are subtypes of Bridges.Variable.AbstractBridge.","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/","page":"List of bridges","title":"List of bridges","text":"Bridges.Variable.FreeBridge\nBridges.Variable.NonposToNonnegBridge\nBridges.Variable.RSOCtoPSDBridge\nBridges.Variable.RSOCtoSOCBridge\nBridges.Variable.SOCtoRSOCBridge\nBridges.Variable.VectorizeBridge\nBridges.Variable.ZerosBridge\nBridges.Variable.HermitianToSymmetricPSDBridge","category":"page"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.FreeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.FreeBridge","text":"FreeBridge{T} <: Bridges.Variable.AbstractBridge\n\nFreeBridge implements the following reformulation:\n\nx in mathbbR into y z ge 0 with the substitution rule x = y - z,\n\nwhere T is the coefficient type of y - z.\n\nSource node\n\nFreeBridge supports:\n\nMOI.VectorOfVariables in MOI.Reals\n\nTarget nodes\n\nFreeBridge creates:\n\nOne variable node: MOI.VectorOfVariables in MOI.Nonnegatives\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.NonposToNonnegBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.NonposToNonnegBridge","text":"NonposToNonnegBridge{T} <: Bridges.Variable.AbstractBridge\n\nNonposToNonnegBridge implements the following reformulation:\n\nx in mathbbR_- into y in mathbbR_+ with the substitution rule x = -y,\n\nwhere T is the coefficient type of -y.\n\nSource node\n\nNonposToNonnegBridge supports:\n\nMOI.VectorOfVariables in MOI.Nonpositives\n\nTarget nodes\n\nNonposToNonnegBridge creates:\n\nOne variable node: MOI.VectorOfVariables in MOI.Nonnegatives,\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.RSOCtoPSDBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.RSOCtoPSDBridge","text":"RSOCtoPSDBridge{T} <: Bridges.Variable.AbstractBridge\n\nRSOCtoPSDBridge implements the following reformulation:\n\nx_2^2 le 2tu where t u ge 0 into Y succeq 0, with the substitution rule: Y = leftbeginarrayc ct  x^top  x  2u mathbfIendarrayright\n\nAdditional bounds are added to ensure the off-diagonals of the 2uI submatrix are 0, and linear constraints are added to ennsure the diagonal of 2uI takes the same values.\n\nAs a special case, if x = 0, then RSOCtoPSDBridge reformulates into (t u) in mathbbR_+.\n\nSource node\n\nRSOCtoPSDBridge supports:\n\nMOI.VectorOfVariables in MOI.RotatedSecondOrderCone\n\nTarget nodes\n\nRSOCtoPSDBridge creates:\n\nOne variable node that depends on the input dimension:\nMOI.VectorOfVariables in MOI.Nonnegatives if dimension is 1 or 2\nMOI.VectorOfVariables in\nMOI.PositiveSemidefiniteConeTriangle otherwise\nThe constraint node MOI.VariableIndex in MOI.EqualTo\nThe constrant node MOI.ScalarAffineFunction in MOI.EqualTo\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.RSOCtoSOCBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.RSOCtoSOCBridge","text":"RSOCtoSOCBridge{T} <: Bridges.Variable.AbstractBridge\n\nRSOCtoSOCBridge implements the following reformulation:\n\nx_2^2 le 2tu into v_2 le w, with the substitution rules t = fracwsqrt 2 + fracv_1sqrt 2, u = fracwsqrt 2 - fracv_1sqrt 2, and x = (v_2ldotsv_N).\n\nSource node\n\nRSOCtoSOCBridge supports:\n\nMOI.VectorOfVariables in MOI.RotatedSecondOrderCone\n\nTarget node\n\nRSOCtoSOCBridge creates:\n\nMOI.VectorOfVariables in MOI.SecondOrderCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.SOCtoRSOCBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.SOCtoRSOCBridge","text":"SOCtoRSOCBridge{T} <: Bridges.Variable.AbstractBridge\n\nSOCtoRSOCBridge implements the following reformulation:\n\nx_2 le t into 2uv ge w_2^2, with the substitution rules t = fracusqrt 2 + fracvsqrt 2, x = (fracusqrt 2 - fracvsqrt 2 w).\n\nAssumptions\n\nSOCtoRSOCBridge assumes that x ge 1.\n\nSource node\n\nSOCtoRSOCBridge supports:\n\nMOI.VectorOfVariables in MOI.SecondOrderCone\n\nTarget node\n\nSOCtoRSOCBridge creates:\n\nMOI.VectorOfVariables in MOI.RotatedSecondOrderCone\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.VectorizeBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.VectorizeBridge","text":"VectorizeBridge{T,S} <: Bridges.Variable.AbstractBridge\n\nVectorizeBridge implements the following reformulations:\n\nx ge a into y in mathbbR_+ with the substitution rule x = a + y\nx le a into y in mathbbR_- with the substitution rule x = a + y\nx == a into y in 0 with the substitution rule x = a + y\n\nwhere T is the coefficient type of a + y.\n\nSource node\n\nVectorizeBridge supports:\n\nMOI.VariableIndex in MOI.GreaterThan{T}\nMOI.VariableIndex in MOI.LessThan{T}\nMOI.VariableIndex in MOI.EqualTo{T}\n\nTarget nodes\n\nVectorizeBridge creates:\n\nOne variable node: MOI.VectorOfVariables in S, where S is one of MOI.Nonnegatives, MOI.Nonpositives, MOI.Zeros depending on the type of S.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.ZerosBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.ZerosBridge","text":"ZerosBridge{T} <: Bridges.Variable.AbstractBridge\n\nZerosBridge implements the following reformulation:\n\nx in 0 into the substitution rule x = 0,\n\nwhere T is the coefficient type of 0.\n\nSource node\n\nZerosBridge supports:\n\nMOI.VectorOfVariables in MOI.Zeros\n\nTarget nodes\n\nZerosBridge does not create target nodes. It replaces all instances of x with 0 via substitution. This means that no variables are created in the underlying model.\n\nCaveats\n\nThe bridged variables are similar to parameters with zero values. Parameters with non-zero values can be created with constrained variables in MOI.EqualTo by combining a VectorizeBridge and this bridge.\n\nHowever, functions modified by ZerosBridge cannot be unbridged. That is, for a given function, we cannot determine if the bridged variables were used.\n\nA related implication is that this bridge does not support MOI.ConstraintDual. However, if a MOI.Utilities.CachingOptimizer is used, the dual can be determined by the bridged optimizer using MOI.Utilities.get_fallback because the caching optimizer records the unbridged function.\n\n\n\n\n\n","category":"type"},{"location":"moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Variable.HermitianToSymmetricPSDBridge","page":"List of bridges","title":"MathOptInterface.Bridges.Variable.HermitianToSymmetricPSDBridge","text":"HermitianToSymmetricPSDBridge{T} <: Bridges.Variable.AbstractBridge\n\nHermitianToSymmetricPSDBridge implements the following reformulation:\n\nHermitian positive semidefinite n x n complex matrix to a symmetric positive semidefinite 2n x 2n real matrix satisfying equality constraints described below.\n\nSource node\n\nHermitianToSymmetricPSDBridge supports:\n\nMOI.VectorOfVariables in MOI.HermitianPositiveSemidefiniteConeTriangle\n\nTarget node\n\nHermitianToSymmetricPSDBridge creates:\n\nMOI.VectorOfVariables in MOI.PositiveSemidefiniteConeTriangle\nMOI.ScalarAffineFunction{T} in MOI.EqualTo{T}\n\nReformulation\n\nThe reformulation is best described by example.\n\nThe Hermitian matrix:\n\nbeginbmatrix\n  x_11             x_12 + y_12im  x_13 + y_13im\n  x_12 - y_12im  x_22             x_23 + y_23im\n  x_13 - y_13im  x_23 - y_23im  x_33\nendbmatrix\n\nis positive semidefinite if and only if the symmetric matrix:\n\nbeginbmatrix\n    x_11  x_12  x_13  0        y_12   y_13 \n            x_22  x_23  -y_12  0        y_23 \n                    x_33  -y_13  -y_23  0      \n                            x_11   x_12   x_13 \n                                     x_22   x_23 \n                                              x_33\nendbmatrix\n\nis positive semidefinite.\n\nThe bridge achieves this reformulation by adding a new set of variables in MOI.PositiveSemidefiniteConeTriangle(6), and then adding three groups of equality constraints to:\n\nconstrain the two x blocks to be equal\nforce the diagonal of the y blocks to be 0\nforce the lower triangular of the y block to be the negative of the upper triangle.\n\n\n\n\n\n","category":"type"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"EditURL = \"https://github.com/jump-dev/SeDuMi.jl/blob/v0.4.2/README.md\"","category":"page"},{"location":"packages/SeDuMi/#SeDuMi.jl","page":"jump-dev/SeDuMi.jl","title":"SeDuMi.jl","text":"","category":"section"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"SeDuMi.jl is wrapper for the SeDuMi solver.","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"an exported sedumi function that is a thin wrapper on top of the sedumi MATLAB function\nan interface to MathOptInterface","category":"page"},{"location":"packages/SeDuMi/#Affiliation","page":"jump-dev/SeDuMi.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"This wrapper is maintained by the JuMP community and is not an official wrapper of SeDuMi.","category":"page"},{"location":"packages/SeDuMi/#License","page":"jump-dev/SeDuMi.jl","title":"License","text":"","category":"section"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"SeDuMi.jl is licensed under the MIT License.","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"The underlying solver, sqlp/sedumi is licensed under the GPL v2 license.","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"In addition, SeDuMi requires an installation of MATLAB, which is a closed-source commercial product for which you must obtain a license.","category":"page"},{"location":"packages/SeDuMi/#Use-with-JuMP","page":"jump-dev/SeDuMi.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"To use SeDuMi with JuMP, do:","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"using JuMP, SeDuMi\nmodel = Model(SeDuMi.Optimizer)\nset_attribute(model, \"fid\", 0)","category":"page"},{"location":"packages/SeDuMi/#Installation","page":"jump-dev/SeDuMi.jl","title":"Installation","text":"","category":"section"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"First, make sure that you satisfy the requirements of the MATLAB.jl Julia package, and that the SeDuMi software is installed in your MATLAB™ installation.","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"Then, install SeDuMi.jl using Pkg.add:","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"import Pkg\nPkg.add(\"SeDuMi\")","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"If you get the error:","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"Undefined function or variable 'sedumi'.\n\nError using save\nVariable 'jx_sedumi_arg_out_1' not found.\n\nERROR: LoadError: MATLAB.MEngineError(\"failed to get variable jx_sedumi_arg_out_1 from MATLAB session\")","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"The error means that we couldn't find the sedumi function with one output argument using the MATLAB C API.","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"This most likely means that you did not add SeDuMi to the MATLAB's path, that is, the toolbox/local/pathdef.m file.","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"If modifying toolbox/local/pathdef.m does not work, the following should work, where /path/to/sedumi/ is the directory where the sedumi folder is located:","category":"page"},{"location":"packages/SeDuMi/","page":"jump-dev/SeDuMi.jl","title":"jump-dev/SeDuMi.jl","text":"julia> import MATLAB\n\njulia> cd(\"/path/to/sedumi/\") do\n           MATLAB.mat\"install_sedumi\"\n       end\n\njulia> MATLAB.mat\"savepath\"","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"EditURL = \"space_shuttle_reentry_trajectory.jl\"","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/#Optimal-control-for-a-Space-Shuttle-reentry-trajectory","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"","category":"section"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"This tutorial was originally contributed by Henrique Ferrolho.","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"This tutorial demonstrates how to compute a reentry trajectory for the Space Shuttle, by formulating and solving a nonlinear programming problem. The problem was drawn from Chapter 6 of \"Practical Methods for Optimal Control and Estimation Using Nonlinear Programming\", by John T. Betts.","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"tip: Tip\nThis tutorial is a more-complicated version of the Rocket Control example. If you are new to solving nonlinear programs in JuMP, you may want to start there instead.","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"The motion of the vehicle is defined by the following set of DAEs:","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"beginaligned\ndoth  = v sin gamma  \ndotphi  = fracvr cos gamma sin psi  cos theta  \ndottheta  = fracvr cos gamma cos psi  \ndotv  = -fracDm - g sin gamma  \ndotgamma  = fracLm v cos(beta) + cos gamma left ( fracvr - fracgv right )  \ndotpsi  = frac1m v cos gamma L sin(beta) + fracvr cos theta cos gamma sin psi sin theta  \nq  le q_U  \nendaligned","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"where the aerodynamic heating on the vehicle wing leading edge is q = q_a q_r and the dynamic variables are","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"beginaligned\n     h  quad textaltitude (ft)     qquad   gamma quad textflight path angle (rad) \n  phi  quad textlongitude (rad)   qquad   psi   quad textazimuth (rad)           \ntheta  quad textlatitude (rad)    qquad   alpha quad textangle of attack (rad)   \n     v  quad textvelocity (ftsec) qquad   beta  quad textbank angle (rad)\nendaligned","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"The aerodynamic and atmospheric forces on the vehicle are specified by the following quantities (English units):","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"beginaligned\n           D  = frac12 c_D S rho v^2                  qquad  a_0  = -020704 \n           L  = frac12 c_L S rho v^2                  qquad  a_1  =  0029244 \n           g  = mu  r^2                                   qquad  mu  =  014076539 times 10^17 \n           r  = R_e + h                                     qquad  b_0  =  007854 \n        rho  = rho_0 exp-hh_r                         qquad  b_1  = -061592  times 10^-2 \n      rho_0  = 0002378                                    qquad  b_2  =  0621408 times 10^-3 \n         h_r  = 23800                                       qquad  q_r  =  17700 sqrtrho (00001 v)^307 \n         c_L  = a_0 + a_1 hatalpha                      qquad  q_a  =  c_0 + c_1 hatalpha + c_2 hatalpha^2 + c_3 hatalpha^3 \n         c_D  = b_0 + b_1 hatalpha + b_2 hatalpha^2 qquad  c_0  =  10672181 \nhatalpha  = 180 alpha  pi                            qquad  c_1  = -019213774 times 10^-1 \n         R_e  = 20902900                                    qquad  c_2  =  021286289 times 10^-3 \n           S  = 2690                                        qquad  c_3  = -010117249 times 10^-5\nendaligned","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"The reentry trajectory begins at an altitude where the aerodynamic forces are quite small with a weight of w = 203000 (lb) and mass m = w  g_0 (slug), where g_0 = 32174 (ft/sec^2). The initial conditions are as follows:","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"beginaligned\n     h  = 260000 text ft  qquad  v       = 25600 text ftsec \n  phi  = 0      text deg qquad  gamma  = -1    text deg    \ntheta  = 0      text deg qquad  psi    = 90    text deg\nendaligned","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"The final point on the reentry trajectory occurs at the unknown (free) time t_F, at the so-called terminal area energy management (TAEM) interface, which is defined by the conditions","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"h = 80000 text ft qquad v = 2500 text ftsec qquad gamma = -5 text deg","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"As explained in the book, our goal is to maximize the final cross-range, which is equivalent to maximizing the final latitude of the vehicle, that is, J = theta(t_F).","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"(Image: Max cross-range shuttle reentry)","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/#Approach","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Approach","text":"","category":"section"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"We will use a discretized model of time, with a fixed number of discretized points, n. The decision variables at each point are going to be the state of the vehicle and the controls commanded to it. In addition, we will also make each time step size Delta t a decision variable; that way, we can either fix the time step size easily, or allow the solver to fine-tune the duration between each adjacent pair of points. Finally, in order to approximate the derivatives of the problem dynamics, we will use either rectangular or trapezoidal integration.","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"warning: Warning\nDo not try to actually land a Space Shuttle using this notebook. There's no mesh refinement going on, which can lead to unrealistic trajectories having position and velocity errors with orders of magnitude 10^4 ft and 10^2 ft/sec, respectively.","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"using JuMP\nimport Interpolations\nimport Ipopt\n\n# Global variables\nconst w = 203000.0  # weight (lb)\nconst g₀ = 32.174    # acceleration (ft/sec^2)\nconst m = w / g₀    # mass (slug)\n\n# Aerodynamic and atmospheric forces on the vehicle\nconst ρ₀ = 0.002378\nconst hᵣ = 23800.0\nconst Rₑ = 20902900.0\nconst μ = 0.14076539e17\nconst S = 2690.0\nconst a₀ = -0.20704\nconst a₁ = 0.029244\nconst b₀ = 0.07854\nconst b₁ = -0.61592e-2\nconst b₂ = 0.621408e-3\nconst c₀ = 1.0672181\nconst c₁ = -0.19213774e-1\nconst c₂ = 0.21286289e-3\nconst c₃ = -0.10117249e-5\n\n# Initial conditions\nconst h_s = 2.6          # altitude (ft) / 1e5\nconst ϕ_s = deg2rad(0)   # longitude (rad)\nconst θ_s = deg2rad(0)   # latitude (rad)\nconst v_s = 2.56         # velocity (ft/sec) / 1e4\nconst γ_s = deg2rad(-1)  # flight path angle (rad)\nconst ψ_s = deg2rad(90)  # azimuth (rad)\nconst α_s = deg2rad(0)   # angle of attack (rad)\nconst β_s = deg2rad(0)   # bank angle (rad)\nconst t_s = 1.00         # time step (sec)\n\n# Final conditions, the so-called Terminal Area Energy Management (TAEM)\nconst h_t = 0.8          # altitude (ft) / 1e5\nconst v_t = 0.25         # velocity (ft/sec) / 1e4\nconst γ_t = deg2rad(-5)  # flight path angle (rad)\n\n# Number of mesh points (knots) to be used\nconst n = 503\n\n# Integration scheme to be used for the dynamics\nconst integration_rule = \"rectangular\"\nnothing #hide","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"tip: Choose a good linear solver\nPicking a good linear solver is extremely important to maximize the performance of nonlinear solvers. For the best results, it is advised to experiment different linear solvers.For example, the linear solver MA27 is outdated and can be quite slow. MA57 is a much better alternative, especially for highly-sparse problems (such as trajectory optimization problems).","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"# Uncomment the lines below to pass user options to the solver\nuser_options = (\n# \"mu_strategy\" => \"monotone\",\n# \"linear_solver\" => \"ma27\",\n)\n\n# Create JuMP model, using Ipopt as the solver\nmodel = Model(optimizer_with_attributes(Ipopt.Optimizer, user_options...))\n\n@variables(model, begin\n    0 ≤ scaled_h[1:n]                # altitude (ft) / 1e5\n    ϕ[1:n]                # longitude (rad)\n    deg2rad(-89) ≤ θ[1:n] ≤ deg2rad(89)  # latitude (rad)\n    1e-4 ≤ scaled_v[1:n]                # velocity (ft/sec) / 1e4\n    deg2rad(-89) ≤ γ[1:n] ≤ deg2rad(89)  # flight path angle (rad)\n    ψ[1:n]                # azimuth (rad)\n    deg2rad(-90) ≤ α[1:n] ≤ deg2rad(90)  # angle of attack (rad)\n    deg2rad(-89) ≤ β[1:n] ≤ deg2rad(1)  # bank angle (rad)\n    #        3.5 ≤       Δt[1:n] ≤ 4.5          # time step (sec)\n    Δt[1:n] == 4.0         # time step (sec)\nend);\nnothing #hide","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"info: Info\nAbove you can find two alternatives for the Δt variables.The first one, 3.5 ≤ Δt[1:n] ≤ 4.5 (currently commented), allows some wiggle room for the solver to adjust the time step size between pairs of mesh points. This is neat because it allows the solver to figure out which parts of the flight require more dense discretization than others. (Remember, the number of discretized points is fixed, and this example does not implement mesh refinement.) However, this makes the problem more complex to solve, and therefore leads to a longer computation time.The second line, Δt[1:n] == 4.0, fixes the duration of every time step to exactly 4.0 seconds. This allows the problem to be solved faster. However, to do this we need to know beforehand that the close-to-optimal total duration of the flight is ~2009 seconds. Therefore, if we split the total duration in slices of 4.0 seconds, we know that we require n = 503 knots to discretize the whole trajectory.","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"# Fix initial conditions\nfix(scaled_h[1], h_s; force = true)\nfix(ϕ[1], ϕ_s; force = true)\nfix(θ[1], θ_s; force = true)\nfix(scaled_v[1], v_s; force = true)\nfix(γ[1], γ_s; force = true)\nfix(ψ[1], ψ_s; force = true)\n\n# Fix final conditions\nfix(scaled_h[n], h_t; force = true)\nfix(scaled_v[n], v_t; force = true)\nfix(γ[n], γ_t; force = true)\n\n# Initial guess: linear interpolation between boundary conditions\nx_s = [h_s, ϕ_s, θ_s, v_s, γ_s, ψ_s, α_s, β_s, t_s]\nx_t = [h_t, ϕ_s, θ_s, v_t, γ_t, ψ_s, α_s, β_s, t_s]\ninterp_linear = Interpolations.LinearInterpolation([1, n], [x_s, x_t])\ninitial_guess = mapreduce(transpose, vcat, interp_linear.(1:n))\nset_start_value.(all_variables(model), vec(initial_guess))\n\n# Functions to restore `h` and `v` to their true scale\n@expression(model, h[j = 1:n], scaled_h[j] * 1e5)\n@expression(model, v[j = 1:n], scaled_v[j] * 1e4)\n\n# Helper functions\n@expression(model, c_L[j = 1:n], a₀ + a₁ * rad2deg(α[j]))\n@expression(model, c_D[j = 1:n], b₀ + b₁ * rad2deg(α[j]) + b₂ * rad2deg(α[j])^2)\n@expression(model, ρ[j = 1:n], ρ₀ * exp(-h[j] / hᵣ))\n@expression(model, D[j = 1:n], 0.5 * c_D[j] * S * ρ[j] * v[j]^2)\n@expression(model, L[j = 1:n], 0.5 * c_L[j] * S * ρ[j] * v[j]^2)\n@expression(model, r[j = 1:n], Rₑ + h[j])\n@expression(model, g[j = 1:n], μ / r[j]^2)\n\n# Motion of the vehicle as a differential-algebraic system of equations (DAEs)\n@expression(model, δh[j = 1:n], v[j] * sin(γ[j]))\n@expression(\n    model,\n    δϕ[j = 1:n],\n    (v[j] / r[j]) * cos(γ[j]) * sin(ψ[j]) / cos(θ[j])\n)\n@expression(model, δθ[j = 1:n], (v[j] / r[j]) * cos(γ[j]) * cos(ψ[j]))\n@expression(model, δv[j = 1:n], -(D[j] / m) - g[j] * sin(γ[j]))\n@expression(\n    model,\n    δγ[j = 1:n],\n    (L[j] / (m * v[j])) * cos(β[j]) +\n    cos(γ[j]) * ((v[j] / r[j]) - (g[j] / v[j]))\n)\n@expression(\n    model,\n    δψ[j = 1:n],\n    (1 / (m * v[j] * cos(γ[j]))) * L[j] * sin(β[j]) +\n    (v[j] / (r[j] * cos(θ[j]))) * cos(γ[j]) * sin(ψ[j]) * sin(θ[j])\n)\n\n# System dynamics\nfor j in 2:n\n    i = j - 1  # index of previous knot\n\n    if integration_rule == \"rectangular\"\n        # Rectangular integration\n        @constraint(model, h[j] == h[i] + Δt[i] * δh[i])\n        @constraint(model, ϕ[j] == ϕ[i] + Δt[i] * δϕ[i])\n        @constraint(model, θ[j] == θ[i] + Δt[i] * δθ[i])\n        @constraint(model, v[j] == v[i] + Δt[i] * δv[i])\n        @constraint(model, γ[j] == γ[i] + Δt[i] * δγ[i])\n        @constraint(model, ψ[j] == ψ[i] + Δt[i] * δψ[i])\n    elseif integration_rule == \"trapezoidal\"\n        # Trapezoidal integration\n        @constraint(model, h[j] == h[i] + 0.5 * Δt[i] * (δh[j] + δh[i]))\n        @constraint(model, ϕ[j] == ϕ[i] + 0.5 * Δt[i] * (δϕ[j] + δϕ[i]))\n        @constraint(model, θ[j] == θ[i] + 0.5 * Δt[i] * (δθ[j] + δθ[i]))\n        @constraint(model, v[j] == v[i] + 0.5 * Δt[i] * (δv[j] + δv[i]))\n        @constraint(model, γ[j] == γ[i] + 0.5 * Δt[i] * (δγ[j] + δγ[i]))\n        @constraint(model, ψ[j] == ψ[i] + 0.5 * Δt[i] * (δψ[j] + δψ[i]))\n    else\n        @error \"Unexpected integration rule '$(integration_rule)'\"\n    end\nend\n\n# Objective: Maximize cross-range\n@objective(model, Max, θ[n])\n\nset_silent(model)  # Hide solver's verbose output\noptimize!(model)  # Solve for the control and state\n@assert termination_status(model) == LOCALLY_SOLVED\n\n# Show final cross-range of the solution\nprintln(\n    \"Final latitude θ = \",\n    round(objective_value(model) |> rad2deg; digits = 2),\n    \"°\",\n)","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/#Plotting-the-results","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Plotting the results","text":"","category":"section"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"Let's plot the results to visualize the optimal trajectory:","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"using Plots\nts = cumsum([0; value.(Δt)])[1:end-1]\nplt_altitude = plot(\n    ts,\n    value.(scaled_h);\n    legend = nothing,\n    title = \"Altitude (100,000 ft)\",\n)\nplt_longitude =\n    plot(ts, rad2deg.(value.(ϕ)); legend = nothing, title = \"Longitude (deg)\")\nplt_latitude =\n    plot(ts, rad2deg.(value.(θ)); legend = nothing, title = \"Latitude (deg)\")\nplt_velocity = plot(\n    ts,\n    value.(scaled_v);\n    legend = nothing,\n    title = \"Velocity (1000 ft/sec)\",\n)\nplt_flight_path =\n    plot(ts, rad2deg.(value.(γ)); legend = nothing, title = \"Flight Path (deg)\")\nplt_azimuth =\n    plot(ts, rad2deg.(value.(ψ)); legend = nothing, title = \"Azimuth (deg)\")\n\nplot(\n    plt_altitude,\n    plt_velocity,\n    plt_longitude,\n    plt_flight_path,\n    plt_latitude,\n    plt_azimuth;\n    layout = grid(3, 2),\n    linewidth = 2,\n    size = (700, 700),\n)","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"function q(h, v, a)\n    ρ(h) = ρ₀ * exp(-h / hᵣ)\n    qᵣ(h, v) = 17700 * √ρ(h) * (0.0001 * v)^3.07\n    qₐ(a) = c₀ + c₁ * rad2deg(a) + c₂ * rad2deg(a)^2 + c₃ * rad2deg(a)^3\n    # Aerodynamic heating on the vehicle wing leading edge\n    return qₐ(a) * qᵣ(h, v)\nend\n\nplt_attack_angle = plot(\n    ts[1:end-1],\n    rad2deg.(value.(α)[1:end-1]);\n    legend = nothing,\n    title = \"Angle of Attack (deg)\",\n)\nplt_bank_angle = plot(\n    ts[1:end-1],\n    rad2deg.(value.(β)[1:end-1]);\n    legend = nothing,\n    title = \"Bank Angle (deg)\",\n)\nplt_heating = plot(\n    ts,\n    q.(value.(scaled_h) * 1e5, value.(scaled_v) * 1e4, value.(α));\n    legend = nothing,\n    title = \"Heating (BTU/ft/ft/sec)\",\n)\n\nplot(\n    plt_attack_angle,\n    plt_bank_angle,\n    plt_heating;\n    layout = grid(3, 1),\n    linewidth = 2,\n    size = (700, 700),\n)","category":"page"},{"location":"tutorials/nonlinear/space_shuttle_reentry_trajectory/","page":"Optimal control for a Space Shuttle reentry trajectory","title":"Optimal control for a Space Shuttle reentry trajectory","text":"plot(\n    rad2deg.(value.(ϕ)),\n    rad2deg.(value.(θ)),\n    value.(scaled_h);\n    linewidth = 2,\n    legend = nothing,\n    title = \"Space Shuttle Reentry Trajectory\",\n    xlabel = \"Longitude (deg)\",\n    ylabel = \"Latitude (deg)\",\n    zlabel = \"Altitude (100,000 ft)\",\n)","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP, HiGHS\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"background/algebraic_modeling_languages/#algebraic-modeling-language","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"JuMP is an algebraic modeling language for mathematical optimization written in the Julia language. In this page, we explain what an algebraic modeling language actually is.","category":"page"},{"location":"background/algebraic_modeling_languages/#What-is-an-algebraic-modeling-language?","page":"Algebraic modeling languages","title":"What is an algebraic modeling language?","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"If you have taken a class in mixed-integer linear programming, you will have seen a formulation like:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"beginaligned\nmin       c^top x \ntextst  A x = b  \n             x ge 0  \n             x_i in mathbbZ quad forall i in mathcalI\nendaligned","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"where c, A, and b are appropriately sized vectors and matrices of data, and mathcalI denotes the set of variables that are integer.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"Solvers expect problems in a standard form like this because it limits the types of constraints that they need to consider. This makes writing a solver much easier.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"info: What is a solver?\nA solver is a software package that computes solutions to one or more classes of problems.For example, HiGHS is a solver for linear programming (LP) and mixed integer programming (MIP) problems. It incorporates algorithms such as the simplex method and the interior-point method.JuMP currently supports a number of open-source and commercial solvers, which can be viewed in the Supported-solvers table.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"Despite the textbook view of a linear program, you probably formulated problems algebraically like so:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"beginaligned\nmax       sumlimits_i = 1^n c_i x_i                   \ntextst  sumlimits_i = 1^n w_i x_i le b             \n             x_i ge 0 quad forall i = 1ldotsn          \n             x_i in mathbbZ quad forall i = 1ldotsn\nendaligned","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"info: Info\nDo you recognize this formulation? It's the knapsack problem.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"Users prefer to write problems in algebraic form because it is more convenient. For example, we used le b, even though the standard form only supported constraints of the form Ax = b.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"We could convert our knapsack problem into the standard form by adding a new slack variable x_0:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"beginaligned\nmax       sumlimits_i = 1^n c_i x_i            \ntextst  x_0 + sumlimits_i = 1^n w_i x_i = b  \n             x_i ge 0 quad forall i = 0ldotsn   \n             x_i in mathbbZ quad forall i = 1ldotsn\nendaligned","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"However, as models get more complicated, this manual conversion becomes more and more error-prone.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"An algebraic modeling language is a tool that simplifies the translation between the algebraic form of the modeler, and the standard form of the solver.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"Each algebraic modeling language has two main parts:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"A domain specific language for the user to write down problems in algebraic form.\nA converter from the algebraic form into a standard form supported by the solver (and back again).","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"Part 2 is less trivial than it might seem, because each solver has a unique application programming interface (API) and data structure for representing optimization models and obtaining results.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"JuMP uses the MathOptInterface.jl package to abstract these differences between solvers.","category":"page"},{"location":"background/algebraic_modeling_languages/#What-is-MathOptInterface?","page":"Algebraic modeling languages","title":"What is MathOptInterface?","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"MathOptInterface (MOI) is an abstraction layer designed to provide an interface to mathematical optimization solvers so that users do not need to understand multiple solver-specific APIs. MOI can be used directly, or through a higher-level modeling interface like JuMP.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"There are three main parts to MathOptInterface:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"A solver-independent API that abstracts concepts such as adding and deleting variables and constraints, setting and getting parameters, and querying results. For more information on the MathOptInterface API, read the documentation.\nAn automatic rewriting system based on equivalent formulations of a constraint. For more information on this rewriting system, read the LazyBridgeOptimizer section of the manual, and our paper on arXiv.\nUtilities for managing how and when models are copied to solvers. For more information on this, read the CachingOptimizer section of the manual.","category":"page"},{"location":"background/algebraic_modeling_languages/#From-user-to-solver","page":"Algebraic modeling languages","title":"From user to solver","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"This section provides a brief summary of the steps that happen in order to translate the model that the user writes into a model that the solver understands.","category":"page"},{"location":"background/algebraic_modeling_languages/#Step-I:-writing-in-algebraic-form","page":"Algebraic modeling languages","title":"Step I: writing in algebraic form","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"JuMP provides the first part of an algebraic modeling language using the @variable, @objective, and @constraint macros.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"For example, here's how we write the knapsack problem in JuMP:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"julia> using JuMP, HiGHS\n\njulia> function algebraic_knapsack(c, w, b)\n           n = length(c)\n           model = Model(HiGHS.Optimizer)\n           set_silent(model)\n           @variable(model, x[1:n] >= 0, Int)\n           @objective(model, Max, sum(c[i] * x[i] for i = 1:n))\n           @constraint(model, sum(w[i] * x[i] for i = 1:n) <= b)\n           optimize!(model)\n           return value.(x)\n       end\nalgebraic_knapsack (generic function with 1 method)\n\njulia> algebraic_knapsack([1, 2], [0.5, 0.5], 1.25)\n2-element Vector{Float64}:\n 0.0\n 2.0","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"This formulation is compact, and it closely matches the algebraic formulation of the model we wrote out above.","category":"page"},{"location":"background/algebraic_modeling_languages/#Step-II:-algebraic-to-functional","page":"Algebraic modeling languages","title":"Step II: algebraic to functional","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"For the next step, JuMP's macros re-write the variables and constraints into a functional form. Here's what the JuMP code looks like after this step:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"julia> using JuMP, HiGHS\n\njulia> function nonalgebraic_knapsack(c, w, b)\n           n = length(c)\n           model = Model(HiGHS.Optimizer)\n           set_silent(model)\n           x = [VariableRef(model) for i = 1:n]\n           for i = 1:n\n               set_lower_bound(x[i], 0)\n               set_integer(x[i])\n               set_name(x[i], \"x[$i]\")\n           end\n           obj = AffExpr(0.0)\n           for i = 1:n\n               add_to_expression!(obj, c[i], x[i])\n           end\n           set_objective(model, MAX_SENSE, obj)\n           lhs = AffExpr(0.0)\n           for i = 1:n\n               add_to_expression!(lhs, w[i], x[i])\n           end\n           con = build_constraint(error, lhs, MOI.LessThan(b))\n           add_constraint(model, con)\n           optimize!(model)\n           return value.(x)\n       end\nnonalgebraic_knapsack (generic function with 1 method)\n\njulia> nonalgebraic_knapsack([1, 2], [0.5, 0.5], 1.25)\n2-element Vector{Float64}:\n 0.0\n 2.0","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"Hopefully you agree that the macro version is much easier to read.","category":"page"},{"location":"background/algebraic_modeling_languages/#Part-III:-JuMP-to-MathOptInterface","page":"Algebraic modeling languages","title":"Part III: JuMP to MathOptInterface","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"In the third step, JuMP converts the functional form of the problem, that is, nonalgebraic_knapsack, into the MathOptInterface API:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"julia> import MathOptInterface as MOI\n\njulia> import HiGHS\n\njulia> function mathoptinterface_knapsack(optimizer, c, w, b)\n           n = length(c)\n           model = MOI.instantiate(optimizer)\n           MOI.set(model, MOI.Silent(), true)\n           x = MOI.add_variables(model, n)\n           for i in 1:n\n               MOI.add_constraint(model, x[i], MOI.GreaterThan(0.0))\n               MOI.add_constraint(model, x[i], MOI.Integer())\n               MOI.set(model, MOI.VariableName(), x[i], \"x[$i]\")\n           end\n           MOI.set(model, MOI.ObjectiveSense(), MOI.MAX_SENSE)\n           obj = MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(c, x), 0.0)\n           MOI.set(model, MOI.ObjectiveFunction{typeof(obj)}(), obj)\n           MOI.add_constraint(\n               model,\n               MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(w, x), 0.0),\n               MOI.LessThan(b),\n           )\n           MOI.optimize!(model)\n           return MOI.get.(model, MOI.VariablePrimal(), x)\n       end\nmathoptinterface_knapsack (generic function with 1 method)\n\njulia> mathoptinterface_knapsack(HiGHS.Optimizer, [1.0, 2.0], [0.5, 0.5], 1.25)\n2-element Vector{Float64}:\n 0.0\n 2.0","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"The code is becoming more verbose and looking less like the mathematical formulation that we started with.","category":"page"},{"location":"background/algebraic_modeling_languages/#Step-IV:-MathOptInterface-to-HiGHS","page":"Algebraic modeling languages","title":"Step IV: MathOptInterface to HiGHS","text":"","category":"section"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"As a final step, the HiGHS.jl package converts the MathOptInterface form, that is, mathoptinterface_knapsack, into a HiGHS-specific API:","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"julia> using HiGHS\n\njulia> function highs_knapsack(c, w, b)\n           n = length(c)\n           model = Highs_create()\n           Highs_setBoolOptionValue(model, \"output_flag\", false)\n           for i in 1:n\n               Highs_addCol(model, c[i], 0.0, Inf, 0, C_NULL, C_NULL)\n               Highs_changeColIntegrality(model, i-1, 1)\n           end\n           Highs_changeObjectiveSense(model, -1)\n           Highs_addRow(\n               model,\n               -Inf,\n               b,\n               Cint(length(w)),\n               collect(Cint(0):Cint(n-1)),\n               w,\n           )\n           Highs_run(model)\n           x = fill(NaN, 2)\n           Highs_getSolution(model, x, C_NULL, C_NULL, C_NULL)\n           Highs_destroy(model)\n           return x\n       end\nhighs_knapsack (generic function with 1 method)\n\njulia> highs_knapsack([1.0, 2.0], [0.5, 0.5], 1.25)\n2-element Vector{Float64}:\n 0.0\n 2.0","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"We've now gone from a algebraic model that looked identical to the mathematical model we started with, to a verbose function that uses HiGHS-specific functionality.","category":"page"},{"location":"background/algebraic_modeling_languages/","page":"Algebraic modeling languages","title":"Algebraic modeling languages","text":"The difference between algebraic_knapsack and highs_knapsack highlights the benefit that algebraic modeling languages provide to users. Moreover, if we used a different solver, the solver-specific function would be entirely different. A key benefit of an algebraic modeling language is that you can change the solver without needing to rewrite the model.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"EditURL = \"https://github.com/jump-dev/AmplNLWriter.jl/blob/v1.2.0/README.md\"","category":"page"},{"location":"packages/AmplNLWriter/#AmplNLWriter.jl","page":"jump-dev/AmplNLWriter.jl","title":"AmplNLWriter.jl","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"(Image: Build Status) (Image: MINLPTests) (Image: codecov)","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"AmplNLWriter.jl is an interface between MathOptInterface.jl and AMPL-enabled solvers.","category":"page"},{"location":"packages/AmplNLWriter/#Affiliation","page":"jump-dev/AmplNLWriter.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"This wrapper is maintained by the JuMP community and has no official connection with the AMPL modeling language or AMPL Optimization Inc.","category":"page"},{"location":"packages/AmplNLWriter/#Installation","page":"jump-dev/AmplNLWriter.jl","title":"Installation","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Install AmplNLWriter using Pkg.add:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"import Pkg\nPkg.add(\"AmplNLWriter\")","category":"page"},{"location":"packages/AmplNLWriter/#Use-with-JuMP","page":"jump-dev/AmplNLWriter.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"AmplNLWriter requires an AMPL compatible solver binary to function.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Pass a string pointing to any AMPL-compatible solver binary as the first positional argument to AmplNLWriter.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"For example, if the bonmin executable is on the system path, use:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"using JuMP, AmplNLWriter\nmodel = Model(() -> AmplNLWriter.Optimizer(\"bonmin\"))","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"If the solver is not on the system path, pass the full path to the solver:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"using JuMP, AmplNLWriter\nmodel = Model(() -> AmplNLWriter.Optimizer(\"/Users/Oscar/ampl.macos64/bonmin\"))","category":"page"},{"location":"packages/AmplNLWriter/#Precompiled-binaries","page":"jump-dev/AmplNLWriter.jl","title":"Precompiled binaries","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"To simplify the process of installing solver binaries, a number of Julia packages provide precompiled binaries that are compatible with AmplNLWriter. These are generally the name of the solver, followed by _jll. For example, bomin is provided by the Bonmin_jll package.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"To call Bonmin via AmplNLWriter.jl, install the Bonmin_jll package, then run:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"using JuMP, AmplNLWriter, Bonmin_jll\nmodel = Model(() -> AmplNLWriter.Optimizer(Bonmin_jll.amplexe))","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Supported packages include:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Solver Julia Package Executable\nBonmin Bonmin_jll.jl Bomin_jll.amplexe\nCouenne Couenne_jll.jl Couenne_jll.amplexe\nIpopt Ipopt_jll.jl Ipopt_jll.amplexe\nSHOT SHOT_jll.jl SHOT_jll.amplexe\nKNITRO KNITRO.jl KNITRO.amplexe","category":"page"},{"location":"packages/AmplNLWriter/#MathOptInterface-API","page":"jump-dev/AmplNLWriter.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"The AmplNLWriter optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}\nMOI.ObjectiveFunction{MOI.VariableIndex}","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"MOI.Reals","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.ZeroOne","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"MOI.NLPBlock()\nMOI.Name()\nMOI.ObjectiveSense()","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Note that some solver executables may not support the full list of constraint types. For example, Ipopt_jll does not support MOI.Integer or MOI.ZeroOne constraints.","category":"page"},{"location":"packages/AmplNLWriter/#Options","page":"jump-dev/AmplNLWriter.jl","title":"Options","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"A list of available options for each solver can be found here:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Bonmin (plus Ipopt options)\nCouenne (plus Ipopt and Bonmin options)\nIpopt\nSHOT","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Set an option using set_attribute. For example, to set the \"bonmin.nlp_log_level\" option to 0 in Bonmin, use:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"using JuMP\nimport AmplNLWriter\nimport Bonmin_jll\nmodel = Model(() -> AmplNLWriter.Optimizer(Bonmin_jll.amplexe))\nset_attribute(model, \"bonmin.nlp_log_level\", 0)","category":"page"},{"location":"packages/AmplNLWriter/#opt-files","page":"jump-dev/AmplNLWriter.jl","title":"opt files","text":"","category":"section"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"Some options need to be specified via an .opt file.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"This file must be located in the current working directory whenever the model is solved.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"The .opt file must be named after the name of the solver, for example, bonmin.opt, and each line must contain an option name and the desired value, separated by a space.","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"For example, to set the absolute and relative tolerances in Couenne to 1 and 0.05 respectively, the couenne.opt file should contain:","category":"page"},{"location":"packages/AmplNLWriter/","page":"jump-dev/AmplNLWriter.jl","title":"jump-dev/AmplNLWriter.jl","text":"allowable_gap 1\nallowable_fraction_gap 0.05","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"<img class=\"display-light-only\" src=\"assets/logo-with-text.svg\" alt=\"JuMP logo\"/>\n<img class=\"display-dark-only\" src=\"assets/logo-dark-with-text.svg\" alt=\"JuMP logo\"/>","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"# These comments do not display in the HTML output.\n# See https://github.com/JuliaDocs/Documenter.jl/issues/674.","category":"page"},{"location":"#Introduction","page":"Introduction","title":"Introduction","text":"","category":"section"},{"location":"","page":"Introduction","title":"Introduction","text":"\\begin{figure}[h]\n  \\includegraphics[width=0.5\\textwidth]{assets/logo-with-text.pdf}\n\\end{figure}","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"Welcome to the documentation for JuMP.","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"note: Note\nThis documentation is also available in PDF format: JuMP.pdf.","category":"page"},{"location":"#What-is-JuMP?","page":"Introduction","title":"What is JuMP?","text":"","category":"section"},{"location":"","page":"Introduction","title":"Introduction","text":"JuMP is a domain-specific modeling language for mathematical optimization embedded in Julia. It currently supports a number of open-source and commercial solvers for a variety of problem classes, including linear, mixed-integer, second-order conic, semidefinite, and nonlinear programming.","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"tip: Tip\nIf you aren't sure if you should use JuMP, read Should you use JuMP?.","category":"page"},{"location":"#Resources-for-getting-started","page":"Introduction","title":"Resources for getting started","text":"","category":"section"},{"location":"","page":"Introduction","title":"Introduction","text":"There are few ways to get started with JuMP:","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"Read the Installation Guide.\nRead the introductory tutorials Getting started with Julia and Getting started with JuMP.\nBrowse some of our modeling tutorials, including classics such as The diet problem, or the Maximum likelihood estimation problem using nonlinear programming.","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"tip: Tip\nNeed help? Join the community forum to search for answers to commonly asked questions.Before asking a question, make sure to read the post make it easier to help you, which contains a number of tips on how to ask a good question.","category":"page"},{"location":"#How-the-documentation-is-structured","page":"Introduction","title":"How the documentation is structured","text":"","category":"section"},{"location":"","page":"Introduction","title":"Introduction","text":"Having a high-level overview of how this documentation is structured will help you know where to look for certain things.","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"Tutorials contain worked examples of solving problems with JuMP. Start here if you are new to JuMP, or you have a particular problem class you want to model.\nThe Manual contains short code-snippets that explain how to achieve specific tasks in JuMP. Look here if you want to know how to achieve a particular task, such as how to Delete a variable or how to Modify an objective coefficient.\nThe API Reference contains a complete list of the functions you can use in JuMP. Look here if you want to know how to use a particular function.\nThe Background information section contains background reading material to provide context to JuMP. Look here if you want an understanding of what JuMP is and why we created it, rather than how to use it.\nThe Developer docs section contains information for people contributing to JuMP development or writing JuMP extensions. Don't worry about this section if you are using JuMP to formulate and solve problems as a user.\nThe MathOptInterface section is a self-contained copy of the documentation for MathOptInterface. Look here for functions and constants beginning with MOI., as well as for general information on how MathOptInterface works.","category":"page"},{"location":"#Citing-JuMP","page":"Introduction","title":"Citing JuMP","text":"","category":"section"},{"location":"","page":"Introduction","title":"Introduction","text":"If you find JuMP useful in your work, we kindly request that you cite the following paper (preprint):","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"@article{Lubin2023,\n    author = {Miles Lubin and Oscar Dowson and Joaquim {Dias Garcia} and Joey Huchette and Beno{\\^i}t Legat and Juan Pablo Vielma},\n    title = {{JuMP} 1.0: {R}ecent improvements to a modeling language for mathematical optimization},\n    journal = {Mathematical Programming Computation},\n    year = {2023},\n    doi = {10.1007/s12532-023-00239-3}\n}","category":"page"},{"location":"#NumFOCUS","page":"Introduction","title":"NumFOCUS","text":"","category":"section"},{"location":"","page":"Introduction","title":"Introduction","text":"(Image: NumFOCUS logo)","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"JuMP is a Sponsored Project of NumFOCUS, a 501(c)(3) nonprofit charity in the United States. NumFOCUS provides JuMP with fiscal, legal, and administrative support to help ensure the health and sustainability of the project. Visit numfocus.org for more information.","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"You can support JuMP by donating.","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"Donations to JuMP are managed by NumFOCUS. For donors in the United States, your gift is tax-deductible to the extent provided by law. As with any donation, you should consult with your tax adviser about your particular tax situation.","category":"page"},{"location":"","page":"Introduction","title":"Introduction","text":"JuMP's largest expense is the annual JuMP-dev workshop. Donations will help us provide travel support for JuMP-dev attendees and take advantage of other opportunities that arise to support JuMP development.","category":"page"},{"location":"#License","page":"Introduction","title":"License","text":"","category":"section"},{"location":"","page":"Introduction","title":"Introduction","text":"JuMP is licensed under the MPL-2.0 software license. Consult the license and the Mozilla FAQ for more information. In addition, JuMP is typically used in conjunction with solver packages and extensions which have their own licences. Consult their package repositories for the specific licenses that apply.","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"EditURL = \"https://github.com/infiniteopt/InfiniteOpt.jl/blob/v0.5.7/docs/jump/README.md\"","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"(Image: Logo)","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"InfiniteOpt.jl is a JuMP extension for expressing and solving infinite-dimensional optimization problems. Such areas include stochastic programming, dynamic programming, space-time optimization, and more. InfiniteOpt serves as an easy-to-use modeling interface for these advanced problem types that can be used by those with little background in these areas. It also it contains a wealth of capabilities making it a powerful and convenient tool for advanced users.  ","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"Current Version Documentation Build Status Citation\n(Image: ) (Image: ) (Image: Build Status) (Image: codecov.io) (Image: DOI)","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"InfiniteOpt builds upon JuMP to add support for many complex modeling objects which  include:","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"Infinite parameters (for example, time, space, and/or uncertainty)\nFinite parameters (similar to ParameterJuMP)\nInfinite variables (decision functions) (for example, y(t x))\nDerivatives (for example, fracpartial y(t x)partial t)\nMeasures (for example, int_t in Ty(t x)dt and mathbbEy(xi))\nFirst class nonlinear modeling","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"The unifying modeling abstraction behind InfiniteOpt captures a wide spectrum  of disciplines which include dynamic, PDE, stochastic, and semi-infinite  optimization. Moreover, we facilitate transferring techniques between these  to synthesize new optimization paradigms.","category":"page"},{"location":"packages/InfiniteOpt/#License","page":"infiniteopt/InfiniteOpt.jl","title":"License","text":"","category":"section"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"InfiniteOpt is licensed under the MIT \"Expat\" license.","category":"page"},{"location":"packages/InfiniteOpt/#Installation","page":"infiniteopt/InfiniteOpt.jl","title":"Installation","text":"","category":"section"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"InfiniteOpt.jl is a registered Julia package and  can be installed by entering the following in the REPL.","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"julia> import Pkg; Pkg.add(\"InfiniteOpt\")","category":"page"},{"location":"packages/InfiniteOpt/#Documentation","page":"infiniteopt/InfiniteOpt.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"Please visit our documentation pages  to learn more. These pages are quite extensive and feature overviews, guides, manuals, tutorials, examples, and more.","category":"page"},{"location":"packages/InfiniteOpt/#Questions","page":"infiniteopt/InfiniteOpt.jl","title":"Questions","text":"","category":"section"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"For additional help please visit and post in our  discussion forum.","category":"page"},{"location":"packages/InfiniteOpt/#Citing","page":"infiniteopt/InfiniteOpt.jl","title":"Citing","text":"","category":"section"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"(Image: DOI)  (Image: DOI)","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"If you use InfiniteOpt.jl in your research, we would greatly appreciate your  citing it.","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"@article{pulsipher2022unifying,\n      title = {A unifying modeling abstraction for infinite-dimensional optimization},\n      journal = {Computers & Chemical Engineering},\n      volume = {156},\n      year = {2022},\n      issn = {0098-1354},\n      doi = {https://doi.org/10.1016/j.compchemeng.2021.107567},\n      url = {https://www.sciencedirect.com/science/article/pii/S0098135421003458},\n      author = {Joshua L. Pulsipher and Weiqi Zhang and Tyler J. Hongisto and Victor M. Zavala},\n}","category":"page"},{"location":"packages/InfiniteOpt/","page":"infiniteopt/InfiniteOpt.jl","title":"infiniteopt/InfiniteOpt.jl","text":"A pre-print version is freely available though arXiv.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"EditURL = \"getting_started_with_sets_and_indexing.jl\"","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Getting-started-with-sets-and-indexing","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Most introductory courses to linear programming will teach you to identify sets over which the decision variables and constraints are indexed. Therefore, it is common to write variables such as x_i for all i in I.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"A common stumbling block for new users to JuMP is that JuMP does not provide specialized syntax for constructing and manipulating these sets.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"We made this decision because Julia already provides a wealth of data structures for working with sets.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"In contrast, because tools like AMPL are stand-alone software packages, they had to define their own syntax for set construction and manipulation. Indeed, the AMPL Book has two entire chapters devoted to sets and indexing (Chapter 5, \"Simple Sets and Indexing,\" and Chapter 6, \"Compound Sets and Indexing\").","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"The purpose of this tutorial is to demonstrate a variety of ways in which you can construct and manipulate sets for optimization models.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"If you haven't already, you should first read Getting started with JuMP.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"using JuMP","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Unordered-sets","page":"Getting started with sets and indexing","title":"Unordered sets","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Unordered sets are useful to describe non-numeric indices, such as the names of cities or types of products.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"The most common way to construct a set is by creating a vector:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"animals = [\"dog\", \"cat\", \"chicken\", \"cow\", \"pig\"]\nmodel = Model()\n@variable(model, x[animals])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"We can also use things like the keys of a dictionary:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"weight_of_animals = Dict(\n    \"dog\" => 20.0,\n    \"cat\" => 5.0,\n    \"chicken\" => 2.0,\n    \"cow\" => 720.0,\n    \"pig\" => 150.0,\n)\nanimal_keys = keys(weight_of_animals)\nmodel = Model()\n@variable(model, x[animal_keys])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"A third option is to use Julia's Set object.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"animal_set = Set()\nfor animal in keys(weight_of_animals)\n    push!(animal_set, animal)\nend\nanimal_set","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"The nice thing about Sets is that they automatically remove duplicates:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"push!(animal_set, \"dog\")\nanimal_set","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Note how dog does not appear twice.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\n@variable(model, x[animal_set])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Sets-of-numbers","page":"Getting started with sets and indexing","title":"Sets of numbers","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Sets of numbers are useful to describe sets that are ordered, such as years or elements in a vector. The easiest way to create sets of numbers is to use Julia's range syntax.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"These can start at 1:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\n@variable(model, x[1:4])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"but they don't have to:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\n@variable(model, x[2012:2021])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Ranges also have a start:step:stop syntax. So the Olympic years are:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\n@variable(model, x[1896:4:2020])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Sets-of-other-things","page":"Getting started with sets and indexing","title":"Sets of other things","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"An important observation is that you can have any Julia type as the element of a set. It doesn't have to be a String or a Number. For example, you can have tuples:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"sources = [\"A\", \"B\", \"C\"]\nsinks = [\"D\", \"E\"]\nS = [(source, sink) for source in sources, sink in sinks]\nmodel = Model()\n@variable(model, x[S])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"x[(\"A\", \"D\")]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"For multi-dimensional sets, you can use JuMP's syntax for constructing Containers:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\n@variable(model, x[sources, sinks])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"x[\"A\", \"D\"]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"info: Info\nNote how we indexed x[\"A\", \"D\"] instead of x[(\"A\", \"D\")] as above.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Sets-to-watch-out-for","page":"Getting started with sets and indexing","title":"Sets to watch out for","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"JuMP supports any sets which are iterable, that is, the set set supports a for-loop like: [i for i in set]. This causes a few common errors.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"First, if T = 3, you may pass the integer T by mistake instead of a range like 1:T:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\nT = 3\n@variable(model, x[T])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"This results in a single variable being created, instead of three as desired. Because this is a common error, a warning is printed, advising you to pass a Vector{Int} instead:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"@variable(model, x_fixed[[T]])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Second, because Strings are iterable, passing a \"index\" as a singleton index is the same as passing ['i', 'n', 'd', 'e', 'x']:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"@variable(model, y[\"index\"])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"This time, a warning is not printed, but the work-around is similar, pass a Vector{String} instead:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"@variable(model, y_fixed[[\"index\"]])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"tip: Tip\nAs a rule of thumb, if you want an index with one element, avoid confusion by passing [index] instead of index.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Set-operations","page":"Getting started with sets and indexing","title":"Set operations","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Julia has built-in support for set operations such as union, intersect, and setdiff.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Therefore, to create a set of all years in which the summer Olympics were held, we can use:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"baseline = 1896:4:2020\ncancelled = [1916, 1940, 1944, 2020]\noff_year = [2021]\nolympic_years = union(setdiff(baseline, cancelled), off_year)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"You can also find the number of elements (that is, the cardinality) in a set using length:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"length(olympic_years)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Set-membership-operations","page":"Getting started with sets and indexing","title":"Set membership operations","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"To compute membership of sets, use the in function.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"2000 in olympic_years","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"2001 in olympic_years","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Indexing-expressions","page":"Getting started with sets and indexing","title":"Indexing expressions","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Use Julia's generator syntax to compute new sets, such as the list of Olympic years that are divisible by 3:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"olympic_3_years = [year for year in olympic_years if mod(year, 3) == 0]\nmodel = Model()\n@variable(model, x[olympic_3_years])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Alternatively, use JuMP's syntax for constructing Containers:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\n@variable(model, x[year in olympic_years; mod(year, 3) == 0])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Compound-sets","page":"Getting started with sets and indexing","title":"Compound sets","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Consider a transportation problem in which we need to ship goods between cities. We have been provided a list of cities:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"cities = [\"Auckland\", \"Wellington\", \"Christchurch\", \"Dunedin\"]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"and a distance matrix which records the shipping distance between pairs of cities. If we can't ship between two cities, the distance is 0.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"distances = [0 643 1071 1426; 0 0 436 790; 0 0 0 360; 1426 0 0 0]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Let's have a look at ways we could write a model with an objective function to minimize the total shipping cost. For simplicity, we'll ignore all constraints.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Fix-unused-variables","page":"Getting started with sets and indexing","title":"Fix unused variables","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"One approach is to fix all variables that we can't use to zero. Most solvers are smart-enough to remove these during a presolve phase, so it has a very small impact on performance:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"N = length(cities)\nmodel = Model()\n@variable(model, x[1:N, 1:N] >= 0)\nfor i in 1:N, j in 1:N\n    if distances[i, j] == 0\n        fix(x[i, j], 0.0; force = true)\n    end\nend\n@objective(model, Min, sum(distances[i, j] * x[i, j] for i in 1:N, j in 1:N))","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Filtered-summation","page":"Getting started with sets and indexing","title":"Filtered summation","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Another approach is to define filters whenever we want to sum over our decision variables:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"N = length(cities)\nmodel = Model()\n@variable(model, x[1:N, 1:N] >= 0)\n@objective(\n    model,\n    Min,\n    sum(\n        distances[i, j] * x[i, j] for i in 1:N, j in 1:N if distances[i, j] > 0\n    ),\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Filtered-indexing","page":"Getting started with sets and indexing","title":"Filtered indexing","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"We could also use JuMP's support for Containers:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"N = length(cities)\nmodel = Model()\n@variable(model, x[i = 1:N, j = 1:N; distances[i, j] > 0])\n@objective(model, Min, sum(distances[i...] * x[i] for i in eachindex(x)))","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"note: Note\nThe i... is called  a \"splat.\" It converts a tuple like (1, 2) into two indices like distances[1, 2].","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Converting-to-a-different-data-structure","page":"Getting started with sets and indexing","title":"Converting to a different data structure","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Another approach, and one that is often the most readable, is to convert the data you have into something that is easier to work with. Originally, we had a vector of strings and a matrix of distances. What we really need is something that maps usable origin-destination pairs to distances. A dictionary is an obvious choice:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"routes = Dict(\n    (a, b) => distances[i, j] for\n    (i, a) in enumerate(cities), (j, b) in enumerate(cities) if\n    distances[i, j] > 0\n)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Then, we can create our model like so:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"model = Model()\n@variable(model, x[keys(routes)])\n@objective(model, Min, sum(v * x[k] for (k, v) in routes))","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"This has a number of benefits over the other approaches, including a compacter algebraic model and variables that are named in a more meaningful way.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"tip: Tip\nIf you're struggling to formulate a problem using the available syntax in JuMP, it's probably a sign that you should convert your data into a different form.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/#Next-steps","page":"Getting started with sets and indexing","title":"Next steps","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"The purpose of this tutorial was to show how JuMP does not have specialized syntax for set creation and manipulation. Instead, you should use the tools provided by Julia itself.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"This is both an opportunity and a challenge, because you are free to pick the syntax and data structures that best suit your problem, but for new users it can be daunting to decide which structure to use.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_sets_and_indexing/","page":"Getting started with sets and indexing","title":"Getting started with sets and indexing","text":"Read through some of the other JuMP tutorials to get inspiration and ideas for how you can use Julia's syntax and data structures to your advantage.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"EditURL = \"https://github.com/jump-dev/Gurobi.jl/blob/v1.0.3/README.md\"","category":"page"},{"location":"packages/Gurobi/#Gurobi.jl","page":"jump-dev/Gurobi.jl","title":"Gurobi.jl","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Gurobi.jl is a wrapper for the Gurobi Optimizer.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"It has two components:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/Gurobi/#Affiliation","page":"jump-dev/Gurobi.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"This wrapper is maintained by the JuMP community and is not officially supported by Gurobi. However, we thank Gurobi for providing us with a license to test Gurobi.jl on GitHub. If you are a commercial customer interested in official support for Gurobi in Julia, let them know.","category":"page"},{"location":"packages/Gurobi/#License","page":"jump-dev/Gurobi.jl","title":"License","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Gurobi.jl is licensed under the MIT License.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"The underlying solver is a closed-source commercial product for which you must obtain a license.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Free Gurobi licenses are available for academics and students.","category":"page"},{"location":"packages/Gurobi/#Installation","page":"jump-dev/Gurobi.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"First, obtain a license of Gurobi and install Gurobi solver.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Then, set the GUROBI_HOME environment variable as appropriate and run Pkg.add(\"Gurobi\"):","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"# On Windows, this might be\nENV[\"GUROBI_HOME\"] = \"C:\\\\Program Files\\\\gurobi1000\\\\win64\"\n# ... or perhaps ...\nENV[\"GUROBI_HOME\"] = \"C:\\\\gurobi1000\\\\win64\"\n# On Mac, this might be\nENV[\"GUROBI_HOME\"] = \"/Library/gurobi1000/mac64\"\n\nimport Pkg\nPkg.add(\"Gurobi\")","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Note: your path may differ. Check which folder you installed Gurobi in, and update the path accordingly.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"By default, building Gurobi.jl will fail if the Gurobi library is not found. This may not be desirable in certain cases, for example when part of a package's test suite uses Gurobi as an optional test dependency, but Gurobi cannot be installed on a CI server running the test suite. To support this use case, the GUROBI_JL_SKIP_LIB_CHECK environment variable may be set (to any value) to make Gurobi.jl installable (but not usable).","category":"page"},{"location":"packages/Gurobi/#Use-with-JuMP","page":"jump-dev/Gurobi.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"To use Gurobi with JuMP, use Gurobi.Optimizer:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"using JuMP, Gurobi\nmodel = Model(Gurobi.Optimizer)\nset_attribute(model, \"TimeLimit\", 100)\nset_attribute(model, \"Presolve\", 0)","category":"page"},{"location":"packages/Gurobi/#MathOptInterface-API","page":"jump-dev/Gurobi.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"The Gurobi optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}\nMOI.ObjectiveFunction{MOI.VariableIndex}\nMOI.ObjectiveFunction{MOI.VectorAffineFunction{Float64}}","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"MOI.Reals","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.Semicontinuous{Float64}\nMOI.VariableIndex in MOI.Semiinteger{Float64}\nMOI.VariableIndex in MOI.ZeroOne\nMOI.VectorOfVariables in MOI.SOS1{Float64}\nMOI.VectorOfVariables in MOI.SOS2{Float64}\nMOI.VectorOfVariables in MOI.SecondOrderCone","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"MOI.HeuristicCallback()\nMOI.LazyConstraintCallback()\nMOI.Name()\nMOI.ObjectiveSense()\nMOI.UserCutCallback()","category":"page"},{"location":"packages/Gurobi/#Options","page":"jump-dev/Gurobi.jl","title":"Options","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"See the Gurobi Documentation for a list and description of allowable parameters.","category":"page"},{"location":"packages/Gurobi/#C-API","page":"jump-dev/Gurobi.jl","title":"C API","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"The C API can be accessed via Gurobi.GRBxx functions, where the names and arguments are identical to the C API.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"See the Gurobi documentation for details.","category":"page"},{"location":"packages/Gurobi/#Reusing-the-same-Gurobi-environment-for-multiple-solves","page":"jump-dev/Gurobi.jl","title":"Reusing the same Gurobi environment for multiple solves","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"When using this package via other packages such as JuMP.jl, the default behavior is to obtain a new Gurobi license token every time a model is created. If you are using Gurobi in a setting where the number of concurrent Gurobi uses is limited (for example, \"Single-Use\" or \"Floating-Use\" licenses), you might instead prefer to obtain a single license token that is shared by all models that your program solves.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"You can do this by passing a Gurobi.Env() object as the first parameter to Gurobi.Optimizer. For example:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"using JuMP, Gurobi\nconst GRB_ENV = Gurobi.Env()\n\nmodel_1 = Model(() -> Gurobi.Optimizer(GRB_ENV))\n\n# The solvers can have different options too\nmodel_2 = direct_model(Gurobi.Optimizer(GRB_ENV))\nset_attribute(model_2, \"OutputFlag\", 0)","category":"page"},{"location":"packages/Gurobi/#Accessing-Gurobi-specific-attributes","page":"jump-dev/Gurobi.jl","title":"Accessing Gurobi-specific attributes","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Get and set Gurobi-specific variable, constraint, and model attributes as follows:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"using JuMP, Gurobi\nmodel = direct_model(Gurobi.Optimizer())\n@variable(model, x >= 0)\n@constraint(model, c, 2x >= 1)\n@objective(model, Min, x)\nMOI.set(model, Gurobi.ConstraintAttribute(\"Lazy\"), c, 2)\noptimize!(model)\nMOI.get(model, Gurobi.VariableAttribute(\"LB\"), x)  # Returns 0.0\nMOI.get(model, Gurobi.ModelAttribute(\"NumConstrs\")) # Returns 1","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"A complete list of supported Gurobi attributes can be found in their online documentation.","category":"page"},{"location":"packages/Gurobi/#Callbacks","page":"jump-dev/Gurobi.jl","title":"Callbacks","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Here is an example using Gurobi's solver-specific callbacks.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"using JuMP, Gurobi, Test\n\nmodel = direct_model(Gurobi.Optimizer())\n@variable(model, 0 <= x <= 2.5, Int)\n@variable(model, 0 <= y <= 2.5, Int)\n@objective(model, Max, y)\ncb_calls = Cint[]\nfunction my_callback_function(cb_data, cb_where::Cint)\n    # You can reference variables outside the function as normal\n    push!(cb_calls, cb_where)\n    # You can select where the callback is run\n    if cb_where != GRB_CB_MIPSOL && cb_where != GRB_CB_MIPNODE\n        return\n    end\n    # You can query a callback attribute using GRBcbget\n    if cb_where == GRB_CB_MIPNODE\n        resultP = Ref{Cint}()\n        GRBcbget(cb_data, cb_where, GRB_CB_MIPNODE_STATUS, resultP)\n        if resultP[] != GRB_OPTIMAL\n            return  # Solution is something other than optimal.\n        end\n    end\n    # Before querying `callback_value`, you must call:\n    Gurobi.load_callback_variable_primal(cb_data, cb_where)\n    x_val = callback_value(cb_data, x)\n    y_val = callback_value(cb_data, y)\n    # You can submit solver-independent MathOptInterface attributes such as\n    # lazy constraints, user-cuts, and heuristic solutions.\n    if y_val - x_val > 1 + 1e-6\n        con = @build_constraint(y - x <= 1)\n        MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n    elseif y_val + x_val > 3 + 1e-6\n        con = @build_constraint(y + x <= 3)\n        MOI.submit(model, MOI.LazyConstraint(cb_data), con)\n    end\n    if rand() < 0.1\n        # You can terminate the callback as follows:\n        GRBterminate(backend(model))\n    end\n    return\nend\n# You _must_ set this parameter if using lazy constraints.\nMOI.set(model, MOI.RawOptimizerAttribute(\"LazyConstraints\"), 1)\nMOI.set(model, Gurobi.CallbackFunction(), my_callback_function)\noptimize!(model)\n@test termination_status(model) == MOI.OPTIMAL\n@test primal_status(model) == MOI.FEASIBLE_POINT\n@test value(x) == 1\n@test value(y) == 2","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"See the Gurobi documentation for other information that can be queried with GRBcbget.","category":"page"},{"location":"packages/Gurobi/#Common-Performance-Pitfall-with-JuMP","page":"jump-dev/Gurobi.jl","title":"Common Performance Pitfall with JuMP","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Gurobi's API works differently than most solvers. Any changes to the model are not applied immediately, but instead go sit in a internal buffer (making any modifications appear to be instantaneous) waiting for a call to GRBupdatemodel (where the work is  done).","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"This leads to a common performance pitfall that has the following message as its main symptom:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Warning: excessive time spent in model updates. Consider calling update less frequently.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"This often means the JuMP program was structured in such a way that Gurobi.jl ends up calling GRBupdatemodel in each iteration of a loop.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Usually, it is possible (and easy) to restructure the JuMP program in a way it stays ssolver-agnostic and has a close-to-ideal performance with Gurobi.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"To guide such restructuring it is good to keep in mind the following bits of information:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"GRBupdatemodel is only called if changes were done since last GRBupdatemodel (that is, if the internal buffer is not empty).\nGRBupdatemodel is called when JuMP.optimize! is called, but this often is not the source of the problem.\nGRBupdatemodel may be called when any model attribute is queried, even if that specific attribute was not changed. This often the source of the problem.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"The worst-case scenario is, therefore, a loop of modify-query-modify-query, even if what is being modified and what is being queried are two completely distinct things.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"As an example, instead of:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"model = Model(Gurobi.Optimizer)\n@variable(model, x[1:100] >= 0)\nfor i in 1:100\n    set_upper_bound(x[i], i)\n    # `GRBupdatemodel` called on each iteration of this loop.\n    println(lower_bound(x[i]))\nend","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"do","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"model = Model(Gurobi.Optimizer)\n@variable(model, x[1:100] >= 0)\n# All modifications are done before any queries.\nfor i in 1:100\n    set_upper_bound(x[i], i)\nend\nfor i in 1:100\n    # Only the first `lower_bound` query may trigger an `GRBupdatemodel`.\n    println(lower_bound(x[i]))\nend","category":"page"},{"location":"packages/Gurobi/#Common-errors","page":"jump-dev/Gurobi.jl","title":"Common errors","text":"","category":"section"},{"location":"packages/Gurobi/#Using-Gurobi-v9.0-and-you-got-an-error-like-Q-not-PSD?","page":"jump-dev/Gurobi.jl","title":"Using Gurobi v9.0 and you got an error like Q not PSD?","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"You need to set the NonConvex parameter:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"model = Model(Gurobi.Optimizer)\nset_optimizer_attribute(model, \"NonConvex\", 2)","category":"page"},{"location":"packages/Gurobi/#Gurobi-Error-1009:-Version-number-is-XX.X,-license-is-for-version-XX.X","page":"jump-dev/Gurobi.jl","title":"Gurobi Error 1009: Version number is XX.X, license is for version XX.X","text":"","category":"section"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Make sure that your license is correct for your Gurobi version. See the Gurobi documentation for details.","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"Once you are sure that the license and Gurobi versions match, re-install Gurobi.jl by running:","category":"page"},{"location":"packages/Gurobi/","page":"jump-dev/Gurobi.jl","title":"jump-dev/Gurobi.jl","text":"import Pkg\nPkg.build(\"Gurobi\")","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/submodules/Benchmarks/overview.md\"","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/submodules/Benchmarks/overview/#The-Benchmarks-submodule","page":"Overview","title":"The Benchmarks submodule","text":"","category":"section"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"To aid the development of efficient solver wrappers, MathOptInterface provides benchmarking capability. Benchmarking a wrapper follows a two-step process.","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"First, prior to making changes, create a baseline for the benchmark results on a given benchmark suite as follows:","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"using SolverPackage  # Replace with your choice of solver.\nimport MathOptInterface as MOI\n\nsuite = MOI.Benchmarks.suite() do\n    SolverPackage.Optimizer()\nend\n\nMOI.Benchmarks.create_baseline(\n    suite, \"current\"; directory = \"/tmp\", verbose = true\n)","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"Use the exclude argument to Benchmarks.suite to exclude benchmarks that the solver doesn't support.","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"Second, after making changes to the package, re-run the benchmark suite and compare to the prior saved results:","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"using SolverPackage\nimport MathOptInterface as MOI\n\nsuite = MOI.Benchmarks.suite() do\n    SolverPackage.Optimizer()\nend\n\nMOI.Benchmarks.compare_against_baseline(\n    suite, \"current\"; directory = \"/tmp\", verbose = true\n)","category":"page"},{"location":"moi/submodules/Benchmarks/overview/","page":"Overview","title":"Overview","text":"This comparison will create a report detailing improvements and regressions.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/latency.md\"","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/tutorials/latency/#Latency","page":"Latency","title":"Latency","text":"","category":"section"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"MathOptInterface suffers the \"time-to-first-solve\" problem of start-up latency.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"This hurts both the user- and developer-experience of MathOptInterface. In the first case, because simple models have a multi-second delay before solving, and in the latter, because our tests take so long to run.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"This page contains some advice on profiling and fixing latency-related problems in the MathOptInterface.jl repository.","category":"page"},{"location":"moi/tutorials/latency/#Background","page":"Latency","title":"Background","text":"","category":"section"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Before reading this part of the documentation, you should familiarize yourself with the reasons for latency in Julia and how to fix them.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Read the blogposts on julialang.org on precompilation and SnoopCompile\nRead the SnoopCompile documentation.\nWatch Tim Holy's talk at JuliaCon 2021\nWatch the package development workshop at JuliaCon 2021","category":"page"},{"location":"moi/tutorials/latency/#Causes","page":"Latency","title":"Causes","text":"","category":"section"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"There are three main causes of latency in MathOptInterface:","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"A large number of types\nLack of method ownership\nType-instability in the bridge layer","category":"page"},{"location":"moi/tutorials/latency/#A-large-number-of-types","page":"Latency","title":"A large number of types","text":"","category":"section"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Julia is very good at specializing method calls based on the input type. Each specialization has a compilation cost, but the benefit of faster run-time performance.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"The best-case scenario is for a method to be called a large number of times with a single set of argument types. The worst-case scenario is for a method to be called a single time for a large set of argument types.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Because of MathOptInterface's function-in-set formulation, we fall into the worst-case situation.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"This is a fundamental limitation of Julia, so there isn't much we can do about it. However, if we can precompile MathOptInterface, much of the cost can be shifted from start-up latency to the time it takes to precompile a package on installation.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"However, there are two things which make MathOptInterface hard to precompile.","category":"page"},{"location":"moi/tutorials/latency/#Lack-of-method-ownership","page":"Latency","title":"Lack of method ownership","text":"","category":"section"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Lack of method ownership happens when a call is made using a mix of structs and methods from different modules. Because of this, no single module \"owns\" the method that is being dispatched, and so it cannot be precompiled.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"tip: Tip\nThis is a slightly simplified explanation. Read the precompilation tutorial for a more in-depth discussion on back-edges.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Unfortunately, the design of MOI means that this is a frequent occurrence: we have a bunch of types in MOI.Utilities that wrap types defined in external packages (for example, the Optimizers), which implement methods of functions defined in MOI (for example, add_variable, add_constraint).","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Here's a simple example of method-ownership in practice:","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"module MyMOI\nstruct Wrapper{T}\n    inner::T\nend\noptimize!(x::Wrapper) = optimize!(x.inner)\nend  # MyMOI\n\nmodule MyOptimizer\nusing ..MyMOI\nstruct Optimizer end\nMyMOI.optimize!(x::Optimizer) = 1\nend  # MyOptimizer\n\nusing SnoopCompile\nmodel = MyMOI.Wrapper(MyOptimizer.Optimizer())\n\njulia> tinf = @snoopi_deep MyMOI.optimize!(model)\nInferenceTimingNode: 0.008256/0.008543 on InferenceFrameInfo for Core.Compiler.Timings.ROOT() with 1 direct children","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"The result is that there was one method that required type inference. If we visualize tinf:","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"using ProfileView\nProfileView.view(flamegraph(tinf))","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"we see a flamegraph with a large red-bar indicating that the method MyMOI.optimize(MyMOI.Wrapper{MyOptimizer.Optimizer}) cannot be precompiled.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"To fix this, we need to designate a module to \"own\" that method (that is, create a back-edge). The easiest way to do this is for MyOptimizer to call MyMOI.optimize(MyMOI.Wrapper{MyOptimizer.Optimizer}) during using MyOptimizer. Let's see that in practice:","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"module MyMOI\nstruct Wrapper{T}\n    inner::T\nend\noptimize(x::Wrapper) = optimize(x.inner)\nend  # MyMOI\n\nmodule MyOptimizer\nusing ..MyMOI\nstruct Optimizer end\nMyMOI.optimize(x::Optimizer) = 1\n# The syntax of this let-while loop is very particular:\n#  * `let ... end` keeps everything local to avoid polluting the MyOptimizer\n#    namespace\n#  * `while true ... break end` runs the code once, and forces Julia to compile\n#    the inner loop, rather than interpret it.\nlet\n    while true\n        model = MyMOI.Wrapper(Optimizer())\n        MyMOI.optimize(model)\n        break\n    end\nend\nend  # MyOptimizer\n\nusing SnoopCompile\nmodel = MyMOI.Wrapper(MyOptimizer.Optimizer())\n\njulia> tinf = @snoopi_deep MyMOI.optimize(model)\nInferenceTimingNode: 0.006822/0.006822 on InferenceFrameInfo for Core.Compiler.Timings.ROOT() with 0 direct children","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"There are now 0 direct children that required type inference because the method was already stored in MyOptimizer!","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Unfortunately, this trick only works if the call-chain is fully inferrable. If there are breaks (due to type instability), then the benefit of doing this is reduced. And unfortunately for us, the design of MathOptInterface has a lot of type instabilities.","category":"page"},{"location":"moi/tutorials/latency/#Type-instability-in-the-bridge-layer","page":"Latency","title":"Type instability in the bridge layer","text":"","category":"section"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Most of MathOptInterface is pretty good at ensuring type-stability. However, a key component is not type stable, and that is the bridging layer.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"In particular, the bridging layer defines Bridges.LazyBridgeOptimizer, which has fields like:","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"struct LazyBridgeOptimizer\n    constraint_bridge_types::Vector{Any}\n    constraint_node::Dict{Tuple{Type,Type},ConstraintNode}\n    constraint_types::Vector{Tuple{Type,Type}}\nend","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"This is because the LazyBridgeOptimizer needs to be able to deal with any function-in-set type passed to it, and we also allow users to pass additional bridges that they defined in external packages.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"So to recap, MathOptInterface suffers package latency because:","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"there are a large number of types and functions\nand these are split between multiple modules, including external packages\nand there are type-instabilities like those in the bridging layer.","category":"page"},{"location":"moi/tutorials/latency/#Resolutions","page":"Latency","title":"Resolutions","text":"","category":"section"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"There are no magic solutions to reduce latency. Issue #1313 tracks progress on reducing latency in MathOptInterface.","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"A useful script is the following (replace GLPK as needed):","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"import GLPK\nimport MathOptInterface as MOI\n\nfunction example_diet(optimizer, bridge)\n    category_data = [\n        1800.0 2200.0;\n          91.0    Inf;\n           0.0   65.0;\n           0.0 1779.0\n    ]\n    cost = [2.49, 2.89, 1.50, 1.89, 2.09, 1.99, 2.49, 0.89, 1.59]\n    food_data = [\n        410 24 26 730;\n        420 32 10 1190;\n        560 20 32 1800;\n        380  4 19 270;\n        320 12 10 930;\n        320 15 12 820;\n        320 31 12 1230;\n        100  8 2.5 125;\n        330  8 10 180\n    ]\n    bridge_model = if bridge\n        MOI.instantiate(optimizer; with_bridge_type=Float64)\n    else\n        MOI.instantiate(optimizer)\n    end\n    model = MOI.Utilities.CachingOptimizer(\n        MOI.Utilities.UniversalFallback(MOI.Utilities.Model{Float64}()),\n        MOI.Utilities.AUTOMATIC,\n    )\n    MOI.Utilities.reset_optimizer(model, bridge_model)\n    MOI.set(model, MOI.Silent(), true)\n    nutrition = MOI.add_variables(model, size(category_data, 1))\n    for (i, v) in enumerate(nutrition)\n        MOI.add_constraint(model, v, MOI.GreaterThan(category_data[i, 1]))\n        MOI.add_constraint(model, v, MOI.LessThan(category_data[i, 2]))\n    end\n    buy = MOI.add_variables(model, size(food_data, 1))\n    MOI.add_constraint.(model, buy, MOI.GreaterThan(0.0))\n    MOI.set(model, MOI.ObjectiveSense(), MOI.MIN_SENSE)\n    f = MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(cost, buy), 0.0)\n    MOI.set(model, MOI.ObjectiveFunction{typeof(f)}(), f)\n    for (j, n) in enumerate(nutrition)\n        f = MOI.ScalarAffineFunction(\n            MOI.ScalarAffineTerm.(food_data[:, j], buy),\n            0.0,\n        )\n        push!(f.terms, MOI.ScalarAffineTerm(-1.0, n))\n        MOI.add_constraint(model, f, MOI.EqualTo(0.0))\n    end\n    MOI.optimize!(model)\n    term_status = MOI.get(model, MOI.TerminationStatus())\n    @assert term_status == MOI.OPTIMAL\n    MOI.add_constraint(\n        model,\n        MOI.ScalarAffineFunction(\n            MOI.ScalarAffineTerm.(1.0, [buy[end-1], buy[end]]),\n            0.0,\n        ),\n        MOI.LessThan(6.0),\n    )\n    MOI.optimize!(model)\n    @assert MOI.get(model, MOI.TerminationStatus()) == MOI.INFEASIBLE\n    return\nend\n\nif length(ARGS) > 0\n    bridge = get(ARGS, 2, \"\") != \"--no-bridge\"\n    println(\"Running: $(ARGS[1]) $(get(ARGS, 2, \"\"))\")\n    @time example_diet(GLPK.Optimizer, bridge)\n    @time example_diet(GLPK.Optimizer, bridge)\n    exit(0)\nend","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"You can create a flame-graph via","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"using SnoopComile\ntinf = @snoopi_deep example_diet(GLPK.Optimizer, true)\nusing ProfileView\nProfileView.view(flamegraph(tinf))","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"Here's how things looked in mid-August 2021: (Image: flamegraph)","category":"page"},{"location":"moi/tutorials/latency/","page":"Latency","title":"Latency","text":"There are a few opportunities for improvement (non-red flames, particularly on the right). But the main problem is a large red (non-precompilable due to method ownership) flame.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"EditURL = \"https://github.com/jump-dev/SCS.jl/blob/v1.3.1/README.md\"","category":"page"},{"location":"packages/SCS/#SCS.jl","page":"jump-dev/SCS.jl","title":"SCS.jl","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"SCS.jl is a wrapper for the SCS splitting cone solver.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"SCS can solve linear programs, second-order cone programs, semidefinite programs, exponential cone programs, and power cone programs.","category":"page"},{"location":"packages/SCS/#Affiliation","page":"jump-dev/SCS.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"This wrapper is maintained by the JuMP community and is not a project of the SCS developers.","category":"page"},{"location":"packages/SCS/#License","page":"jump-dev/SCS.jl","title":"License","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"SCS.jl is licensed under the MIT License.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"The underlying solver, cvxgrp/scs, is licensed under the MIT License.","category":"page"},{"location":"packages/SCS/#Installation","page":"jump-dev/SCS.jl","title":"Installation","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"Install SCS.jl using the Julia package manager:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"import Pkg\nPkg.add(\"SCS\")","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"In addition to installing the SCS.jl package, this will also download and install the SCS binaries. (You do not need to install SCS separately.)","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"To use a custom binary, read the Custom solver binaries section of the JuMP documentation.","category":"page"},{"location":"packages/SCS/#Use-with-Convex.jl","page":"jump-dev/SCS.jl","title":"Use with Convex.jl","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"This example shows how we can model a simple knapsack problem with Convex and use SCS to solve it.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"using Convex, SCS\nitems  = [:Gold, :Silver, :Bronze]\nvalues = [5.0, 3.0, 1.0]\nweights = [2.0, 1.5, 0.3]\n\n# Define a variable of size 3, each index representing an item\nx = Variable(3)\np = maximize(x' * values, 0 <= x, x <= 1, x' * weights <= 3)\nsolve!(p, SCS.Optimizer)\nprintln([items x.value])\n# [:Gold 0.9999971880377178\n#  :Silver 0.46667637765641057\n#  :Bronze 0.9999998036351865]","category":"page"},{"location":"packages/SCS/#Use-with-JuMP","page":"jump-dev/SCS.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"This example shows how we can model a simple knapsack problem with JuMP and use SCS to solve it.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"using JuMP, SCS\nitems  = [:Gold, :Silver, :Bronze]\nvalues = Dict(:Gold => 5.0,  :Silver => 3.0,  :Bronze => 1.0)\nweight = Dict(:Gold => 2.0,  :Silver => 1.5,  :Bronze => 0.3)\n\nmodel = Model(SCS.Optimizer)\n@variable(model, 0 <= take[items] <= 1)  # Define a variable for each item\n@objective(model, Max, sum(values[item] * take[item] for item in items))\n@constraint(model, sum(weight[item] * take[item] for item in items) <= 3)\noptimize!(model)\nprintln(value.(take))\n# 1-dimensional DenseAxisArray{Float64,1,...} with index sets:\n#     Dimension 1, Symbol[:Gold, :Silver, :Bronze]\n# And data, a 3-element Array{Float64,1}:\n#  1.0000002002226671\n#  0.4666659513182934\n#  1.0000007732744878","category":"page"},{"location":"packages/SCS/#MathOptInterface-API","page":"jump-dev/SCS.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"The SCS optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"MOI.Reals","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"MOI.VectorAffineFunction{Float64} in MOI.DualExponentialCone\nMOI.VectorAffineFunction{Float64} in MOI.DualPowerCone{Float64}\nMOI.VectorAffineFunction{Float64} in MOI.ExponentialCone\nMOI.VectorAffineFunction{Float64} in MOI.Nonnegatives\nMOI.VectorAffineFunction{Float64} in MOI.PowerCone{Float64}\nMOI.VectorAffineFunction{Float64} in MOI.SecondOrderCone\nMOI.VectorAffineFunction{Float64} in MOI.Zeros\nMOI.VectorAffineFunction{Float64} in SCS.ScaledPSDCone","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"MOI.ObjectiveSense()","category":"page"},{"location":"packages/SCS/#Options","page":"jump-dev/SCS.jl","title":"Options","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"All SCS solver options can be set through Convex.jl or MathOptInterface.jl.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"For example:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"model = Model(optimizer_with_attributes(SCS.Optimizer, \"max_iters\" => 10))\n\n# via MathOptInterface:\noptimizer = SCS.Optimizer()\nMOI.set(optimizer, MOI.RawOptimizerAttribute(\"max_iters\"), 10)\nMOI.set(optimizer, MOI.RawOptimizerAttribute(\"verbose\"), 0)","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"Common options are:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"max_iters: the maximum number of iterations to take\nverbose: turn printing on (1) or off (0)","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"See the glbopts.h header for other options.","category":"page"},{"location":"packages/SCS/#Linear-solvers","page":"jump-dev/SCS.jl","title":"Linear solvers","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"SCS uses a linear solver internally, see this section of SCS documentation. SCS.jl ships with","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"SCS.DirectSolver (sparse direct, the default) and\nSCS.LinearSolver (sparse indirect, by conjugate gradient)","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"enabled.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"The find currently available linear solvers one can inspect SCS.available_solvers:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"julia> using SCS\n\njulia> SCS.available_solvers\n2-element Vector{DataType}:\n SCS.DirectSolver\n SCS.IndirectSolver","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"To select the linear solver of choice:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"pass the linear_solver option to optimizer_with_attributes, or to MOI.OptimizerWithAttributes;\nspecify the solver as the first argument when using scs_solve directly (see the low-level wrapper section below).","category":"page"},{"location":"packages/SCS/#SCS-with-MKL-Pardiso-linear-solver","page":"jump-dev/SCS.jl","title":"SCS with MKL Pardiso linear solver","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"To enable the MKL Pardiso (direct sparse) solver one needs to load MKL_jll before SCS:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"julia> import Pkg\n\njulia> Pkg.add(Pkg.PackageSpec(name = \"MKL_jll\", version = \"2022.2\"))\n\njulia> using MKL_jll    # This must be called before `using SCS`.\n\njulia> using SCS\n\njulia> SCS.available_solvers\n3-element Vector{DataType}:\n SCS.DirectSolver\n SCS.IndirectSolver\n SCS.MKLDirectSolver","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"The MKLDirectSolver is available on Linux x86_64 platform only.","category":"page"},{"location":"packages/SCS/#SCS-with-Sparse-GPU-indirect-solver-(CUDA-only)","page":"jump-dev/SCS.jl","title":"SCS with Sparse GPU indirect solver (CUDA only)","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"To enable the indirect linear solver on GPU one needs to load CUDA_jll before SCS:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"julia> import Pkg\n\njulia> Pkg.add(Pkg.PackageSpec(name = \"CUDA_jll\", version = \"10.1\"))\n\njulia> using CUDA_jll  # This must be called before `using SCS`.\n\njulia> using SCS\n\njulia> SCS.available_solvers\n3-element Array{DataType,1}:\n SCS.DirectSolver\n SCS.IndirectSolver\n SCS.GpuIndirectSolver","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"The GpuIndirectSolver is available on Linux x86_64 platform only.","category":"page"},{"location":"packages/SCS/#Low-level-wrapper","page":"jump-dev/SCS.jl","title":"Low-level wrapper","text":"","category":"section"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"SCS.jl provides a low-level interface to solve a problem directly, without interfacing through MathOptInterface.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"This is an advanced interface with a risk of incorrect usage. For new users, we recommend that you use the JuMP or Convex interfaces instead.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"SCS solves a problem of the form:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"minimize        1/2 * x' * P * x + c' * x\nsubject to      A * x + s = b\n                s in K","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"where K is a product cone of:","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"zero cone\npositive orthant { x | x ≥ 0 }\nbox cone { (t,x) | t*l ≤ x ≤ t*u}\nsecond-order cone (SOC) { (t,x) | ||x||_2 ≤ t }\nsemi-definite cone (SDC) { X | X is psd }\nexponential cone { (x,y,z) | y e^(x/y) ≤ z, y>0 }\npower cone { (x,y,z) | x^a * y^(1-a) ≥ |z|, x ≥ 0, y ≥ 0 }\ndual power cone { (u,v,w) | (u/a)^a * (v/(1-a))^(1-a) ≥ |w|, u ≥ 0, v ≥ 0 }.","category":"page"},{"location":"packages/SCS/","page":"jump-dev/SCS.jl","title":"jump-dev/SCS.jl","text":"To solve this problem with SCS, call SCS.scs_solve; see the docstring for details.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"EditURL = \"lp_sensitivity.jl\"","category":"page"},{"location":"tutorials/linear/lp_sensitivity/#Sensitivity-analysis-of-a-linear-program","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"","category":"section"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"This tutorial explains how to use the lp_sensitivity_report function to create sensitivity reports like those that are produced by the Excel Solver. This is most often used in introductory classes to linear programming.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"In brief, sensitivity analysis of a linear program is about asking two questions:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"Given an optimal solution, how much can the objective coefficients change by before a different solution becomes optimal?\nGiven an optimal solution, how much can the right-hand side of a linear constraint change by before a different solution becomes optimal?","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"JuMP provides a function, lp_sensitivity_report, to help us compute these values, but this tutorial extends that to create more informative tables in the form of a DataFrame.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/#Setup","page":"Sensitivity analysis of a linear program","title":"Setup","text":"","category":"section"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"using JuMP\nimport HiGHS\nimport DataFrames","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"as well as this small linear program:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"model = Model(HiGHS.Optimizer)\n@variable(model, x >= 0)\n@variable(model, 0 <= y <= 3)\n@variable(model, z <= 1)\n@objective(model, Min, 12x + 20y - z)\n@constraint(model, c1, 6x + 8y >= 100)\n@constraint(model, c2, 7x + 12y >= 120)\n@constraint(model, c3, x + y <= 20)\noptimize!(model)\nsolution_summary(model; verbose = true)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"Can you identify:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"The objective coefficient of each variable?\nThe right-hand side of each constraint?\nThe optimal primal and dual solutions?","category":"page"},{"location":"tutorials/linear/lp_sensitivity/#Sensitivity-reports","page":"Sensitivity analysis of a linear program","title":"Sensitivity reports","text":"","category":"section"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"Now let's call lp_sensitivity_report:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"report = lp_sensitivity_report(model)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"It returns a SensitivityReport object, which maps:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"Every variable reference to a tuple (d_lo, d_hi)::Tuple{Float64,Float64}, explaining how much the objective coefficient of the corresponding variable can change by, such that the original basis remains optimal.\nEvery constraint reference to a tuple (d_lo, d_hi)::Tuple{Float64,Float64}, explaining how much the right-hand side of the corresponding constraint can change by, such that the basis remains optimal.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"Both tuples are relative, rather than absolute. So, given an objective coefficient of 1.0 and a tuple (-0.5, 0.5), the objective coefficient can range between 1.0 - 0.5 an 1.0 + 0.5.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"For example:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"report[x]","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"indicates that the objective coefficient on x, that is, 12, can decrease by -0.333 or increase by 3.0 and the primal solution (15, 1.25) will remain optimal. In addition:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"report[c1]","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"means that the right-hand side of the c1 constraint (100), can decrease by 4 units, or increase by 2.85 units, and the primal solution (15, 1.25) will remain optimal.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/#Variable-sensitivity","page":"Sensitivity analysis of a linear program","title":"Variable sensitivity","text":"","category":"section"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"By themselves, the tuples aren't informative. Let's put them in context by collating a range of other information about a variable:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"function variable_report(xi)\n    return (\n        name = name(xi),\n        lower_bound = has_lower_bound(xi) ? lower_bound(xi) : -Inf,\n        value = value(xi),\n        upper_bound = has_upper_bound(xi) ? upper_bound(xi) : Inf,\n        reduced_cost = reduced_cost(xi),\n        obj_coefficient = coefficient(objective_function(model), xi),\n        allowed_decrease = report[xi][1],\n        allowed_increase = report[xi][2],\n    )\nend","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"Calling our function on x:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"x_report = variable_report(x)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"That's a bit hard to read, so let's call this on every variable in the model and put things into a DataFrame:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"variable_df =\n    DataFrames.DataFrame(variable_report(xi) for xi in all_variables(model))","category":"page"},{"location":"tutorials/linear/lp_sensitivity/#Constraint-sensitivity","page":"Sensitivity analysis of a linear program","title":"Constraint sensitivity","text":"","category":"section"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"We can do something similar with constraints:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"function constraint_report(c::ConstraintRef)\n    return (\n        name = name(c),\n        value = value(c),\n        rhs = normalized_rhs(c),\n        slack = normalized_rhs(c) - value(c),\n        shadow_price = shadow_price(c),\n        allowed_decrease = report[c][1],\n        allowed_increase = report[c][2],\n    )\nend\n\nc1_report = constraint_report(c1)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"That's a bit hard to read, so let's call this on every variable in the model and put things into a DataFrame:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"constraint_df = DataFrames.DataFrame(\n    constraint_report(ci) for (F, S) in list_of_constraint_types(model) for\n    ci in all_constraints(model, F, S) if F == AffExpr\n)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/#Analysis-questions","page":"Sensitivity analysis of a linear program","title":"Analysis questions","text":"","category":"section"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"Now we can use these dataframes to ask questions of the solution.","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"For example, we can find basic variables by looking for variables with a reduced cost of 0:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"basic = filter(row -> iszero(row.reduced_cost), variable_df)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"and non-basic variables by looking for non-zero reduced costs:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"non_basic = filter(row -> !iszero(row.reduced_cost), variable_df)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"we can also find constraints that are binding by looking for zero slacks:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"binding = filter(row -> iszero(row.slack), constraint_df)","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"or non-zero shadow prices:","category":"page"},{"location":"tutorials/linear/lp_sensitivity/","page":"Sensitivity analysis of a linear program","title":"Sensitivity analysis of a linear program","text":"binding2 = filter(row -> !iszero(row.shadow_price), constraint_df)","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/tutorials/example.md\"","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/tutorials/example/#Solving-a-problem-using-MathOptInterface","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"In this tutorial we demonstrate how to use MathOptInterface to solve the binary-constrained knapsack problem:","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"beginaligned\nmax   c^top x       \nst   w^top x le C \n         x_i in 01quad forall i=1ldotsn\nendaligned","category":"page"},{"location":"moi/tutorials/example/#Required-packages","page":"Solving a problem using MathOptInterface","title":"Required packages","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"Load the MathOptInterface module and define the shorthand MOI:","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"import MathOptInterface as MOI","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"As an optimizer, we choose GLPK:","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"using GLPK\noptimizer = GLPK.Optimizer()","category":"page"},{"location":"moi/tutorials/example/#Define-the-data","page":"Solving a problem using MathOptInterface","title":"Define the data","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"We first define the constants of the problem:","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> c = [1.0, 2.0, 3.0]\n3-element Vector{Float64}:\n 1.0\n 2.0\n 3.0\n\njulia> w = [0.3, 0.5, 1.0]\n3-element Vector{Float64}:\n 0.3\n 0.5\n 1.0\n\njulia> C = 3.2\n3.2","category":"page"},{"location":"moi/tutorials/example/#Add-the-variables","page":"Solving a problem using MathOptInterface","title":"Add the variables","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> x = MOI.add_variables(optimizer, length(c));","category":"page"},{"location":"moi/tutorials/example/#set_objective_example","page":"Solving a problem using MathOptInterface","title":"Set the objective","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> MOI.set(\n           optimizer,\n           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),\n           MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(c, x), 0.0),\n       );\n\njulia> MOI.set(optimizer, MOI.ObjectiveSense(), MOI.MAX_SENSE)","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"tip: Tip\nMOI.ScalarAffineTerm.(c, x) is a shortcut for [MOI.ScalarAffineTerm(c[i], x[i]) for i = 1:3]. This is Julia's broadcast syntax in action, and is used quite often throughout MOI.","category":"page"},{"location":"moi/tutorials/example/#Add-the-constraints","page":"Solving a problem using MathOptInterface","title":"Add the constraints","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"We add the knapsack constraint and integrality constraints:","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> MOI.add_constraint(\n           optimizer,\n           MOI.ScalarAffineFunction(MOI.ScalarAffineTerm.(w, x), 0.0),\n           MOI.LessThan(C),\n       );","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"Add integrality constraints:","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> for x_i in x\n           MOI.add_constraint(optimizer, x_i, MOI.ZeroOne())\n       end","category":"page"},{"location":"moi/tutorials/example/#Optimize-the-model","page":"Solving a problem using MathOptInterface","title":"Optimize the model","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> MOI.optimize!(optimizer)","category":"page"},{"location":"moi/tutorials/example/#Understand-why-the-solver-stopped","page":"Solving a problem using MathOptInterface","title":"Understand why the solver stopped","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"The first thing to check after optimization is why the solver stopped, for example, did it stop because of a time limit or did it stop because it found the optimal solution?","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> MOI.get(optimizer, MOI.TerminationStatus())\nOPTIMAL::TerminationStatusCode = 1","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"Looks like we found an optimal solution.","category":"page"},{"location":"moi/tutorials/example/#Understand-what-solution-was-returned","page":"Solving a problem using MathOptInterface","title":"Understand what solution was returned","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> MOI.get(optimizer, MOI.ResultCount())\n1\n\njulia> MOI.get(optimizer, MOI.PrimalStatus())\nFEASIBLE_POINT::ResultStatusCode = 1\n\njulia> MOI.get(optimizer, MOI.DualStatus())\nNO_SOLUTION::ResultStatusCode = 0","category":"page"},{"location":"moi/tutorials/example/#Query-the-objective","page":"Solving a problem using MathOptInterface","title":"Query the objective","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"What is its objective value?","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> MOI.get(optimizer, MOI.ObjectiveValue())\n6.0","category":"page"},{"location":"moi/tutorials/example/#Query-the-primal-solution","page":"Solving a problem using MathOptInterface","title":"Query the primal solution","text":"","category":"section"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"And what is the value of the variables x?","category":"page"},{"location":"moi/tutorials/example/","page":"Solving a problem using MathOptInterface","title":"Solving a problem using MathOptInterface","text":"julia> MOI.get(optimizer, MOI.VariablePrimal(), x)\n3-element Vector{Float64}:\n 1.0\n 1.0\n 1.0","category":"page"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"EditURL = \"https://github.com/jump-dev/DiffOpt.jl/blob/v0.4.2/README.md\"","category":"page"},{"location":"packages/DiffOpt/#DiffOpt.jl","page":"jump-dev/DiffOpt.jl","title":"DiffOpt.jl","text":"","category":"section"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"(Image: stable docs) (Image: development docs) (Image: Build Status) (Image: Coverage)","category":"page"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"DiffOpt.jl is a package for differentiating convex optimization programs with respect to the program parameters. DiffOpt currently supports linear, quadratic, and conic programs.","category":"page"},{"location":"packages/DiffOpt/#License","page":"jump-dev/DiffOpt.jl","title":"License","text":"","category":"section"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"DiffOpt.jl is licensed under the MIT License.","category":"page"},{"location":"packages/DiffOpt/#Installation","page":"jump-dev/DiffOpt.jl","title":"Installation","text":"","category":"section"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"Install DiffOpt using Pkg.add:","category":"page"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"import Pkg\nPkg.add(\"DiffOpt\")","category":"page"},{"location":"packages/DiffOpt/#Documentation","page":"jump-dev/DiffOpt.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"The documentation for DiffOpt.jl includes a detailed description of the theory behind the package, along with examples, tutorials, and an API reference.","category":"page"},{"location":"packages/DiffOpt/#Use-with-JuMP","page":"jump-dev/DiffOpt.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"Use DiffOpt with JuMP by following this brief example:","category":"page"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"using JuMP, DiffOpt, HiGHS\n# Create a model using the wrapper\nmodel = Model(() -> DiffOpt.diff_optimizer(HiGHS.Optimizer))\n# Define your model and solve it\n@variable(model, x)\n@constraint(model, cons, x >= 3)\n@objective(model, Min, 2x)\noptimize!(model)\n# Choose the problem parameters to differentiate with respect to, and set their\n# perturbations.\nMOI.set(model, DiffOpt.ReverseVariablePrimal(), x, 1.0)\n# Differentiate the model\nDiffOpt.reverse_differentiate!(model)\n# fetch the gradients\ngrad_exp = MOI.get(model, DiffOpt.ReverseConstraintFunction(), cons)  # -3 x - 1\nconstant(grad_exp)        # -1\ncoefficient(grad_exp, x)  # -3","category":"page"},{"location":"packages/DiffOpt/#GSOC2020","page":"jump-dev/DiffOpt.jl","title":"GSOC2020","text":"","category":"section"},{"location":"packages/DiffOpt/","page":"jump-dev/DiffOpt.jl","title":"jump-dev/DiffOpt.jl","text":"DiffOpt began as a NumFOCUS sponsored Google Summer of Code (2020) project","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"EditURL = \"https://github.com/jump-dev/KNITRO.jl/blob/4c56de7684c42dd3c83c5fbc515ae1f424eed524/README.md\"","category":"page"},{"location":"packages/KNITRO/#KNITRO.jl","page":"jump-dev/KNITRO.jl","title":"KNITRO.jl","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"KNITRO.jl is a wrapper for the Artelys Knitro solver.","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"It has two components:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"a thin wrapper around the C API\nan interface to MathOptInterface.","category":"page"},{"location":"packages/KNITRO/#Affiliation","page":"jump-dev/KNITRO.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"This wrapper is maintained by the JuMP community with help from Artelys.","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"Contact Artelys support if you encounter any problem with this interface or the solver.","category":"page"},{"location":"packages/KNITRO/#License","page":"jump-dev/KNITRO.jl","title":"License","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"KNITRO.jl is licensed under the MIT License.","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"The underlying solver is a closed-source commercial product for which you must purchase a license.","category":"page"},{"location":"packages/KNITRO/#Installation","page":"jump-dev/KNITRO.jl","title":"Installation","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"First, obtain a license and install a copy of KNITRO from Artelys.","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"Then, install KNITRO.jl using the Julia package manager:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"import Pkg\nPkg.add(\"KNITRO\")","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"If you are having trouble installing KNITRO.jl, here are several things to try:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"Make sure that you have defined your global variables correctly, for example with export KNITRODIR=\"/path/to/knitro-vXXX-$OS-64\" and export LD_LIBRARY_PATH=\"$LD_LIBRARY_PATH:$KNITRODIR/lib\". You can check that KNITRO.jl sees your library with using KNITRO; KNITRO.has_knitro().\nIf KNITRO.has_knitro() returns false but you are confident that your paths are correct, try running Pkg.build(\"KNITRO\") and restarting Julia. In at least one user's experience, installing and using KNITRO in a temporary Julia environment (activated with ] activate --temp) does not work and the need to manually build is likely the reason why.","category":"page"},{"location":"packages/KNITRO/#Use-with-JuMP","page":"jump-dev/KNITRO.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"To use KNITRO with JuMP, use KNITRO.Optimizer:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"using JuMP, KNITRO\nmodel = Model(KNITRO.Optimizer)\nset_attribute(model, \"outlev\", 1)\nset_attribute(model, \"algorithm\", 4)","category":"page"},{"location":"packages/KNITRO/#Use-with-AMPL","page":"jump-dev/KNITRO.jl","title":"Use with AMPL","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"To use KNITRO with AmplNLWriter.jl, use KNITRO.amplexe:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"using JuMP\nimport AmplNLWriter\nimport KNITRO\nmodel = Model(() -> AmplNLWriter.Optimizer(KNITRO.amplexe, [\"outlev=3\"]))","category":"page"},{"location":"packages/KNITRO/#Use-with-other-packages","page":"jump-dev/KNITRO.jl","title":"Use with other packages","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"A variety of packages extend KNITRO.jl to support other optimization modeling systems. These include:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"NLPModelsKnitro\nOptimization.jl","category":"page"},{"location":"packages/KNITRO/#MathOptInterface-API","page":"jump-dev/KNITRO.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"The Knitro optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}\nMOI.ObjectiveFunction{MOI.ScalarQuadraticFunction{Float64}}\nMOI.ObjectiveFunction{MOI.VariableIndex}","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"MOI.Reals","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarAffineFunction{Float64} in MOI.LessThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.EqualTo{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.GreaterThan{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.Interval{Float64}\nMOI.ScalarQuadraticFunction{Float64} in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.EqualTo{Float64}\nMOI.VariableIndex in MOI.GreaterThan{Float64}\nMOI.VariableIndex in MOI.Integer\nMOI.VariableIndex in MOI.Interval{Float64}\nMOI.VariableIndex in MOI.LessThan{Float64}\nMOI.VariableIndex in MOI.ZeroOne\nMOI.VectorAffineFunction{Float64} in MOI.SecondOrderCone\nMOI.VectorOfVariables in MOI.Complements\nMOI.VectorOfVariables in MOI.SecondOrderCone","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"MOI.NLPBlock()\nMOI.NLPBlockDualStart()\nMOI.ObjectiveSense()","category":"page"},{"location":"packages/KNITRO/#Options","page":"jump-dev/KNITRO.jl","title":"Options","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"A list of available options is provided in the KNITRO reference manual.","category":"page"},{"location":"packages/KNITRO/#Low-level-wrapper","page":"jump-dev/KNITRO.jl","title":"Low-level wrapper","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"KNITRO.jl implements most of Knitro's functionalities. If you aim at using part of Knitro's API that are not implemented in the MathOptInterface/JuMP ecosystem, you can refer to the low-level API, which wraps Knitro's C API (whose templates are specified in the file knitro.h).","category":"page"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"Extensive examples using the C wrapper can be found in examples/.","category":"page"},{"location":"packages/KNITRO/#Multi-threading","page":"jump-dev/KNITRO.jl","title":"Multi-threading","text":"","category":"section"},{"location":"packages/KNITRO/","page":"jump-dev/KNITRO.jl","title":"jump-dev/KNITRO.jl","text":"Due to limitations in the interaction between Julia and C, KNITRO.jl disables multi-threading if the problem is nonlinear. This will override any options such as par_numthreads that you may have set. Read GitHub issue #93 for more details.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"EditURL = \"https://github.com/jump-dev/MathOptInterface.jl/blob/v1.20.0/docs/src/manual/modification.md\"","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"CurrentModule = MathOptInterface\nDocTestSetup = quote\n    import MathOptInterface as MOI\nend\nDocTestFilters = [r\"MathOptInterface|MOI\"]","category":"page"},{"location":"moi/manual/modification/#Problem-modification","page":"Problem modification","title":"Problem modification","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"In addition to adding and deleting constraints and variables, MathOptInterface supports modifying, in-place, coefficients in the constraints and the objective function of a model.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"These modifications can be grouped into two categories:","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"modifications which replace the set of function of a constraint with a new set or function\nmodifications which change, in-place, a component of a function","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"warning: Warning\nSome ModelLike objects do not support problem modification.","category":"page"},{"location":"moi/manual/modification/#Modify-the-set-of-a-constraint","page":"Problem modification","title":"Modify the set of a constraint","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"Use set and ConstraintSet to modify the set of a constraint by replacing it with a new instance of the same type.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> c = MOI.add_constraint(\n           model,\n           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),\n           MOI.EqualTo(1.0),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)\n\njulia> MOI.set(model, MOI.ConstraintSet(), c, MOI.EqualTo(2.0));\n\njulia> MOI.get(model, MOI.ConstraintSet(), c) == MOI.EqualTo(2.0)\ntrue","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"However, the following will fail as the new set is of a different type to the original set:","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> MOI.set(model, MOI.ConstraintSet(), c, MOI.GreaterThan(2.0))\nERROR: [...]","category":"page"},{"location":"moi/manual/modification/#Special-cases:-set-transforms","page":"Problem modification","title":"Special cases: set transforms","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"If our constraint is an affine inequality, then this corresponds to modifying the right-hand side of a constraint in linear programming.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"In some special cases, solvers may support efficiently changing the set of a constraint (for example, from LessThan to GreaterThan). For these cases, MathOptInterface provides the transform method.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"The transform function returns a new constraint index, and the old constraint index (that is, c) is no longer valid.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> c = MOI.add_constraint(\n           model,\n           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),\n           MOI.LessThan(1.0),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}(1)\n\njulia> new_c = MOI.transform(model, c, MOI.GreaterThan(2.0));\n\njulia> MOI.is_valid(model, c)\nfalse\n\njulia> MOI.is_valid(model, new_c)\ntrue","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"note: Note\ntransform cannot be called with a set of the same type. Use set instead.","category":"page"},{"location":"moi/manual/modification/#Modify-the-function-of-a-constraint","page":"Problem modification","title":"Modify the function of a constraint","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"Use set and ConstraintFunction to modify the function of a constraint by replacing it with a new instance of the same type.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> c = MOI.add_constraint(\n           model,\n           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),\n           MOI.EqualTo(1.0),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)\n\njulia> new_f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(2.0, x)], 1.0);\n\njulia> MOI.set(model, MOI.ConstraintFunction(), c, new_f);\n\njulia> MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f\ntrue","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"However, the following will fail as the new function is of a different type to the original function:","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> MOI.set(model, MOI.ConstraintFunction(), c, x)\nERROR: [...]","category":"page"},{"location":"moi/manual/modification/#Modify-constant-term-in-a-scalar-function","page":"Problem modification","title":"Modify constant term in a scalar function","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"Use modify and ScalarConstantChange to modify the constant term in a ScalarAffineFunction or ScalarQuadraticFunction.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> c = MOI.add_constraint(\n           model,\n           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),\n           MOI.EqualTo(1.0),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)\n\njulia> MOI.modify(model, c, MOI.ScalarConstantChange(1.0));\n\njulia> new_f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 1.0);\n\njulia> MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f\ntrue","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"ScalarConstantChange can also be used to modify the objective function by passing an instance of ObjectiveFunction:","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> MOI.set(\n           model,\n           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),\n           new_f,\n       );\n\njulia> MOI.modify(\n           model,\n           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),\n           MOI.ScalarConstantChange(-1.0)\n       );\n\njulia> MOI.get(\n           model,\n           MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}(),\n       ) ≈ MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], -1.0)\ntrue","category":"page"},{"location":"moi/manual/modification/#Modify-constant-terms-in-a-vector-function","page":"Problem modification","title":"Modify constant terms in a vector function","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"Use modify and VectorConstantChange to modify the constant vector in a VectorAffineFunction or VectorQuadraticFunction.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> c = MOI.add_constraint(\n           model,\n           MOI.VectorAffineFunction([\n                   MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, x)),\n                   MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(2.0, x)),\n               ],\n               [0.0, 0.0],\n           ),\n           MOI.Nonnegatives(2),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1)\n\njulia> MOI.modify(model, c, MOI.VectorConstantChange([3.0, 4.0]));\n\njulia> new_f = MOI.VectorAffineFunction(\n           [\n        MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, x)),\n        MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(2.0, x)),\n           ],\n           [3.0, 4.0],\n       );\n\njulia> MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f\ntrue","category":"page"},{"location":"moi/manual/modification/#Modify-affine-coefficients-in-a-scalar-function","page":"Problem modification","title":"Modify affine coefficients in a scalar function","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"Use modify and ScalarCoefficientChange to modify the affine coefficient of a ScalarAffineFunction or ScalarQuadraticFunction.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> c = MOI.add_constraint(\n           model,\n           MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(1.0, x)], 0.0),\n           MOI.EqualTo(1.0),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}(1)\n\njulia> MOI.modify(model, c, MOI.ScalarCoefficientChange(x, 2.0));\n\njulia> new_f = MOI.ScalarAffineFunction([MOI.ScalarAffineTerm(2.0, x)], 0.0);\n\njulia> MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f\ntrue","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"ScalarCoefficientChange can also be used to modify the objective function by passing an instance of ObjectiveFunction.","category":"page"},{"location":"moi/manual/modification/#Modify-affine-coefficients-in-a-vector-function","page":"Problem modification","title":"Modify affine coefficients in a vector function","text":"","category":"section"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"Use modify and MultirowChange to modify a vector of affine coefficients in a VectorAffineFunction or a VectorQuadraticFunction.","category":"page"},{"location":"moi/manual/modification/","page":"Problem modification","title":"Problem modification","text":"julia> c = MOI.add_constraint(\n           model,\n           MOI.VectorAffineFunction([\n                   MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(1.0, x)),\n                   MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(2.0, x)),\n               ],\n               [0.0, 0.0],\n           ),\n           MOI.Nonnegatives(2),\n       )\nMathOptInterface.ConstraintIndex{MathOptInterface.VectorAffineFunction{Float64}, MathOptInterface.Nonnegatives}(1)\n\njulia> MOI.modify(model, c, MOI.MultirowChange(x, [(1, 3.0), (2, 4.0)]));\n\njulia> new_f = MOI.VectorAffineFunction(\n           [\n        MOI.VectorAffineTerm(1, MOI.ScalarAffineTerm(3.0, x)),\n        MOI.VectorAffineTerm(2, MOI.ScalarAffineTerm(4.0, x)),\n           ],\n           [0.0, 0.0],\n       );\n\njulia> MOI.get(model, MOI.ConstraintFunction(), c) ≈ new_f\ntrue","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"EditURL = \"facility_location.jl\"","category":"page"},{"location":"tutorials/linear/facility_location/#The-facility-location-problem","page":"The facility location problem","title":"The facility location problem","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"This tutorial was originally contributed by Mathieu Tanneau and Alexis Montoison.","category":"page"},{"location":"tutorials/linear/facility_location/#Required-packages","page":"The facility location problem","title":"Required packages","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"This tutorial requires the following packages:","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"using JuMP\nimport HiGHS\nimport LinearAlgebra\nimport Plots\nimport Random","category":"page"},{"location":"tutorials/linear/facility_location/#Uncapacitated-facility-location","page":"The facility location problem","title":"Uncapacitated facility location","text":"","category":"section"},{"location":"tutorials/linear/facility_location/#Problem-description","page":"The facility location problem","title":"Problem description","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"We are given","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"A set M=1 dots m of clients\nA set N= 1 dots n of sites where a facility can be built","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Decision variables Decision variables are split into two categories:","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Binary variable y_j indicates whether facility j is built or not\nBinary variable x_i j indicates whether client i is assigned to facility j","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Objective The objective is to minimize the total cost of serving all clients. This costs breaks down into two components:","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Fixed cost of building a facility.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"In this example, this cost is f_j = 1  forall j.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Cost of serving clients from the assigned facility.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"In this example, the cost c_i j of serving client i from facility j is the Euclidean distance between the two.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Constraints","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Each customer must be served by exactly one facility\nA facility cannot serve any client unless it is open","category":"page"},{"location":"tutorials/linear/facility_location/#MILP-formulation","page":"The facility location problem","title":"MILP formulation","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"The problem can be formulated as the following MILP:","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"beginaligned\nmin_x y    \nsum_i j c_i j x_i j +\nsum_j f_j y_j \nst \nsum_j x_i j = 1  forall i in M \n x_i j leq y_j  forall i in M j in N \n x_i j y_j in 0 1  forall i in M j in N\nendaligned","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"where the first set of constraints ensures that each client is served exactly once, and the second set of constraints ensures that no client is served from an unopened facility.","category":"page"},{"location":"tutorials/linear/facility_location/#Problem-data","page":"The facility location problem","title":"Problem data","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"To ensure reproducibility, we set the random number seed:","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Random.seed!(314)","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Here's the data we need:","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"# Number of clients\nm = 12\n# Number of facility locations\nn = 5\n\n# Clients' locations\nx_c, y_c = rand(m), rand(m)\n\n# Facilities' potential locations\nx_f, y_f = rand(n), rand(n)\n\n# Fixed costs\nf = ones(n);\n\n# Distance\nc = zeros(m, n)\nfor i in 1:m\n    for j in 1:n\n        c[i, j] = LinearAlgebra.norm([x_c[i] - x_f[j], y_c[i] - y_f[j]], 2)\n    end\nend","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Display the data","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Plots.scatter(\n    x_c,\n    y_c;\n    label = \"Clients\",\n    markershape = :circle,\n    markercolor = :blue,\n)\nPlots.scatter!(\n    x_f,\n    y_f;\n    label = \"Facility\",\n    markershape = :square,\n    markercolor = :white,\n    markersize = 6,\n    markerstrokecolor = :red,\n    markerstrokewidth = 2,\n)","category":"page"},{"location":"tutorials/linear/facility_location/#JuMP-implementation","page":"The facility location problem","title":"JuMP implementation","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Create a JuMP model","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, y[1:n], Bin);\n@variable(model, x[1:m, 1:n], Bin);\n# Each client is served exactly once\n@constraint(model, client_service[i in 1:m], sum(x[i, j] for j in 1:n) == 1);\n# A facility must be open to serve a client\n@constraint(model, open_facility[i in 1:m, j in 1:n], x[i, j] <= y[j]);\n@objective(model, Min, f' * y + sum(c .* x));\nnothing #hide","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Solve the uncapacitated facility location problem with HiGHS","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"optimize!(model)\nprintln(\"Optimal value: \", objective_value(model))","category":"page"},{"location":"tutorials/linear/facility_location/#Visualizing-the-solution","page":"The facility location problem","title":"Visualizing the solution","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"The threshold 1e-5 ensure that edges between clients and facilities are drawn when x[i, j] ≈ 1.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"x_is_selected = isapprox.(value.(x), 1; atol = 1e-5);\ny_is_selected = isapprox.(value.(y), 1; atol = 1e-5);\n\np = Plots.scatter(\n    x_c,\n    y_c;\n    markershape = :circle,\n    markercolor = :blue,\n    label = nothing,\n)\n\nPlots.scatter!(\n    x_f,\n    y_f;\n    markershape = :square,\n    markercolor = [(y_is_selected[j] ? :red : :white) for j in 1:n],\n    markersize = 6,\n    markerstrokecolor = :red,\n    markerstrokewidth = 2,\n    label = nothing,\n)\n\nfor i in 1:m, j in 1:n\n    if x_is_selected[i, j]\n        Plots.plot!(\n            [x_c[i], x_f[j]],\n            [y_c[i], y_f[j]];\n            color = :black,\n            label = nothing,\n        )\n    end\nend\n\np","category":"page"},{"location":"tutorials/linear/facility_location/#Capacitated-facility-location","page":"The facility location problem","title":"Capacitated facility location","text":"","category":"section"},{"location":"tutorials/linear/facility_location/#Problem-formulation","page":"The facility location problem","title":"Problem formulation","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"The capacitated variant introduces a capacity constraint on each facility, that is, clients have a certain level of demand to be served, while each facility only has finite capacity which cannot be exceeded.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Specifically,","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"The demand of client i is denoted by a_i geq 0\nThe capacity of facility j is denoted by q_j geq 0","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"The capacity constraints then write","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"beginaligned\nsum_i a_i x_i j leq q_j y_j  forall j in N\nendaligned","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Note that, if y_j is set to 0, the capacity constraint above automatically forces x_i j to 0.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Thus, the capacitated facility location can be formulated as follows","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"beginaligned\nmin_x y    \nsum_i j c_i j x_i j +\nsum_j f_j y_j \nst \nsum_j x_i j = 1  forall i in M \n sum_i a_i x_i j leq q_j y_j  forall j in N \n x_i j y_j in 0 1  forall i in M j in N\nendaligned","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"For simplicity, we will assume that there is enough capacity to serve the demand, that is, there exists at least one feasible solution.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"We need some new data:","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"# Demands\na = rand(1:3, m);\n\n# Capacities\nq = rand(5:10, n);\nnothing #hide","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Display the data","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Plots.scatter(\n    x_c,\n    y_c;\n    label = nothing,\n    markershape = :circle,\n    markercolor = :blue,\n    markersize = 2 .* (2 .+ a),\n)\n\nPlots.scatter!(\n    x_f,\n    y_f;\n    label = nothing,\n    markershape = :rect,\n    markercolor = :white,\n    markersize = q,\n    markerstrokecolor = :red,\n    markerstrokewidth = 2,\n)","category":"page"},{"location":"tutorials/linear/facility_location/#JuMP-implementation-2","page":"The facility location problem","title":"JuMP implementation","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Create a JuMP model","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"model = Model(HiGHS.Optimizer)\nset_silent(model)\n@variable(model, y[1:n], Bin);\n@variable(model, x[1:m, 1:n], Bin);\n# Each client is served exactly once\n@constraint(model, client_service[i in 1:m], sum(x[i, :]) == 1);\n# Capacity constraint\n@constraint(model, capacity, x' * a .<= (q .* y));\n# Objective\n@objective(model, Min, f' * y + sum(c .* x));\nnothing #hide","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Solve the problem","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"optimize!(model)\nprintln(\"Optimal value: \", objective_value(model))","category":"page"},{"location":"tutorials/linear/facility_location/#Visualizing-the-solution-2","page":"The facility location problem","title":"Visualizing the solution","text":"","category":"section"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"The threshold 1e-5 ensure that edges between clients and facilities are drawn when x[i, j] ≈ 1.","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"x_is_selected = isapprox.(value.(x), 1; atol = 1e-5);\ny_is_selected = isapprox.(value.(y), 1; atol = 1e-5);\nnothing #hide","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"Display the solution","category":"page"},{"location":"tutorials/linear/facility_location/","page":"The facility location problem","title":"The facility location problem","text":"p = Plots.scatter(\n    x_c,\n    y_c;\n    label = nothing,\n    markershape = :circle,\n    markercolor = :blue,\n    markersize = 2 .* (2 .+ a),\n)\n\nPlots.scatter!(\n    x_f,\n    y_f;\n    label = nothing,\n    markershape = :rect,\n    markercolor = [(y_is_selected[j] ? :red : :white) for j in 1:n],\n    markersize = q,\n    markerstrokecolor = :red,\n    markerstrokewidth = 2,\n)\n\nfor i in 1:m, j in 1:n\n    if x_is_selected[i, j]\n        Plots.plot!(\n            [x_c[i], x_f[j]],\n            [y_c[i], y_f[j]];\n            color = :black,\n            label = nothing,\n        )\n    end\nend\n\np","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"EditURL = \"https://github.com/jump-dev/Xpress.jl/blob/v0.16.1/README.md\"","category":"page"},{"location":"packages/Xpress/#Xpress.jl","page":"jump-dev/Xpress.jl","title":"Xpress.jl","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"Xpress.jl is a wrapper for the FICO Xpress Solver.","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"It has two components:","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"a thin wrapper around the complete C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/Xpress/#Affiliation","page":"jump-dev/Xpress.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"The Xpress wrapper for Julia is community driven and not officially supported by FICO Xpress. If you are a commercial customer interested in official support for Julia from FICO Xpress, let them know.","category":"page"},{"location":"packages/Xpress/#License","page":"jump-dev/Xpress.jl","title":"License","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"Xpress.jl is licensed under the MIT License.","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"The underlying solver is a closed-source commercial product for which you must purchase a license.","category":"page"},{"location":"packages/Xpress/#Installation","page":"jump-dev/Xpress.jl","title":"Installation","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"First, obtain a license of Xpress and install Xpress solver, following the instructions on the FICO website.","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"Then, install this package using:","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"import Pkg\nPkg.add(\"Xpress\")","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"If you encounter an error, make sure that the XPRESSDIR environmental variable is set to the path of the Xpress directory. This should be part of a standard installation. The Xpress library will be searched for in XPRESSDIR/lib on Unix platforms and XPRESSDIR/bin on Windows.","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"For example, on macOS, you may need:","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"ENV[\"XPRESSDIR\"] = \"/Applications/FICO Xpress/xpressmp/\"\nimport Pkg\nPkg.add(\"Xpress\")","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"By default, building Xpress.jl will fail if the Xpress library is not found. This may not be desirable in certain cases, for example when part of a package's test suite uses Xpress as an optional test dependency, but Xpress cannot be installed on a CI server running the test suite. To support this use case, the XPRESS_JL_SKIP_LIB_CHECK environment variable may be set (to any value) to make Xpress.jl installable (but not usable).","category":"page"},{"location":"packages/Xpress/#Use-with-JuMP","page":"jump-dev/Xpress.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"To use Xpress with JuMP, use:","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"using JuMP, Xpress\nmodel = Model(Xpress.Optimizer)\nset_optimizer(model, \"PRESOLVE\", 0)","category":"page"},{"location":"packages/Xpress/#Options","page":"jump-dev/Xpress.jl","title":"Options","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"For other parameters use Xpress Optimizer manual or type julia -e \"using Xpress; println(keys(Xpress.XPRS_ATTRIBUTES))\".","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"If logfile is set to \"\", the log file is disabled and output is printed to the console (there might be issues with console output on windows (it is manually implemented with callbacks)). If logfile is set to a file's path, output is printed to that file. By default, logfile = \"\" (console).","category":"page"},{"location":"packages/Xpress/#Callbacks","page":"jump-dev/Xpress.jl","title":"Callbacks","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"Solver specific and solver independent callbacks are working in MathOptInterface and, consequently, in JuMP. However, the current implementation should be considered experimental.","category":"page"},{"location":"packages/Xpress/#Environment-variables","page":"jump-dev/Xpress.jl","title":"Environment variables","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"XPRESS_JL_SKIP_LIB_CHECK: Used to skip build lib check as previously described.\nXPRESS_JL_NO_INFO: Disable license info log.\nXPRESS_JL_NO_DEPS_ERROR: Disable error when do deps.jl file is found.\nXPRESS_JL_NO_AUTO_INIT: Disable automatic run of Xpress.initialize().","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"Specially useful for explicitly loading the dynamic library.","category":"page"},{"location":"packages/Xpress/#Skipping-Xpress.postsolve","page":"jump-dev/Xpress.jl","title":"Skipping Xpress.postsolve","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"In older versions of Xpress, the command XPRSpostsolve throws an error in infeasible models. In these older versions the post solve should not be executed. To do this, one can use the MOI.RawOptimizerAttribute(\"MOI_POST_SOLVE\") to skip this routine.","category":"page"},{"location":"packages/Xpress/#C-API","page":"jump-dev/Xpress.jl","title":"C API","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"The C API can be accessed via Xpress.Lib.XPRSxx functions, where the names and arguments are identical to the C API. ","category":"page"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"See the Xpress documentation for details.","category":"page"},{"location":"packages/Xpress/#Documentation","page":"jump-dev/Xpress.jl","title":"Documentation","text":"","category":"section"},{"location":"packages/Xpress/","page":"jump-dev/Xpress.jl","title":"jump-dev/Xpress.jl","text":"For more information, consult the FICO optimizer manual.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"EditURL = \"mip_duality.jl\"","category":"page"},{"location":"tutorials/linear/mip_duality/#Computing-the-duals-of-a-mixed-integer-program","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"","category":"section"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"This tutorial explains how to compute the duals of a mixed-integer linear program by fixing the discrete variables to their optimal solution and resolving as a linear program.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"using JuMP\nimport HiGHS","category":"page"},{"location":"tutorials/linear/mip_duality/#The-model","page":"Computing the duals of a mixed-integer program","title":"The model","text":"","category":"section"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"Our example model is the unit commitment example from Unit commitment. The details are unimportant, other than to note that there are two types of continuous variables, g and w, representing the quantity of generation from thermal and wind plants, and a discrete variable dispatch, which is 1 if plant i is operating, and 0 if not.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"We are interested in the \"dual\" of the power_balance constraint, because it represents the marginal price of electricity that consumers should pay for their consumption.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"generators = [\n    (min = 0.0, max = 1000.0, fixed_cost = 1000.0, variable_cost = 50.0),\n    (min = 300.0, max = 1000.0, fixed_cost = 0.0, variable_cost = 100.0),\n]\nN = length(generators)\nmodel = Model(HiGHS.Optimizer)\nset_silent(model)\n@variables(model, begin\n    generators[i].min <= g[i = 1:N] <= generators[i].max\n    0 <= w <= 200\n    dispatch[i = 1:N], Bin\nend)\n@constraints(model, begin\n    power_balance, sum(g[i] for i in 1:N) + w == 1500\n    [i = 1:N], g[i] <= generators[i].max * dispatch[i]\n    [i = 1:N], g[i] >= generators[i].min * dispatch[i]\nend)\n@objective(\n    model,\n    Min,\n    sum(\n        generators[i].fixed_cost * dispatch[i] +\n        generators[i].variable_cost * g[i] for i in 1:N\n    )\n)\nprint(model)","category":"page"},{"location":"tutorials/linear/mip_duality/#Manually-fix-the-variables","page":"Computing the duals of a mixed-integer program","title":"Manually fix the variables","text":"","category":"section"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"If we optimize this model, we obtain a dual_status of NO_SOLUTION:","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"optimize!(model)\ndual_status(model)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"This is because HiGHS cannot compute the duals of a mixed-integer program. We can work around this problem by fixing the integer variables to their optimal solution, relaxing integrality, and re-solving as a linear program.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"discrete_values = value.(dispatch)\nfix.(dispatch, discrete_values; force = true)\nunset_binary.(dispatch)\nprint(model)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"Now if we re-solve the problem, we obtain a FEASIBLE_POINT for the dual:","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"optimize!(model)\ndual_status(model)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"and a marginal price of electricity of $100/MWh:","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"dual(power_balance)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"To reset the problem back to a mixed-integer linear program, we need to unfix and call set_binary:","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"unfix.(dispatch)\nset_binary.(dispatch)\nprint(model)","category":"page"},{"location":"tutorials/linear/mip_duality/#Use-fix_discrete_variables","page":"Computing the duals of a mixed-integer program","title":"Use fix_discrete_variables","text":"","category":"section"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"Manually choosing the variables to relax and fix works for our small example, but it becomes more difficult in problems with a larger number of binary and integer variables. To automate the process we just did manually, JuMP provides the fix_discrete_variables function:","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"optimize!(model)\ndual_status(model)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"undo = fix_discrete_variables(model);\nnothing #hide","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"Here undo is a function that, when called with no arguments, returns the model to the original mixed-integer formulation.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"tip: Tip\nAfter calling fix_discrete_variables, you can set a new solver with set_optimizer if your mixed-integer solver does not support computing a dual solution.","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"print(model)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"optimize!(model)\ndual_status(model)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"dual(power_balance)","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"Finally, call undo to revert the reformulation","category":"page"},{"location":"tutorials/linear/mip_duality/","page":"Computing the duals of a mixed-integer program","title":"Computing the duals of a mixed-integer program","text":"undo()\nprint(model)","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"EditURL = \"min_ellipse.jl\"","category":"page"},{"location":"tutorials/conic/min_ellipse/#Minimal-ellipses","page":"Minimal ellipses","title":"Minimal ellipses","text":"","category":"section"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"This example comes from section 8.4.1 of the book Convex Optimization by Boyd and Vandenberghe (2004).","category":"page"},{"location":"tutorials/conic/min_ellipse/#Formulation","page":"Minimal ellipses","title":"Formulation","text":"","category":"section"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"Given a set of m ellipses of the form:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"E(A b c) =  x  x^top A x + 2 b^top x + c leq 0 ","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"the minimal ellipse problem finds an ellipse with the minimum area that encloses the given ellipses.","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"It is convenient to parameterize the minimal enclosing ellipse as","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":" x   Px + q  leq 1 ","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"Then the optimal P and q are given by the convex semidefinite program;","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"beginaligned\ntextmaximize   quad log(det(P))  \ntextsubject to   quad tau_i geq 0  i = 1 ldots m \n quadbeginbmatrix\n    P^2 - tau_i A_i          P q - tau_i b_i       0        \n    (P q - tau_i b_i)^top    -1 - tau_i c_i     (P q)^top \n                0                     (P q)       - P^2       \nendbmatrix\npreceq 0 text (PSD)    i=1 ldots m\nendaligned","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"with helper variables tau.","category":"page"},{"location":"tutorials/conic/min_ellipse/#Required-packages","page":"Minimal ellipses","title":"Required packages","text":"","category":"section"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"using JuMP\nimport LinearAlgebra\nimport Plots\nimport SCS","category":"page"},{"location":"tutorials/conic/min_ellipse/#Data","page":"Minimal ellipses","title":"Data","text":"","category":"section"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"First, define the m input ellipses (here m = 6), parameterized as x^T A_i x + 2 b_i^T x + c leq 0:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"struct Ellipse\n    A::Matrix{Float64}\n    b::Vector{Float64}\n    c::Float64\n    function Ellipse(A::Matrix{Float64}, b::Vector{Float64}, c::Float64)\n        @assert isreal(A) && LinearAlgebra.issymmetric(A)\n        return new(A, b, c)\n    end\nend\n\nellipses = [\n    Ellipse([1.2576 -0.3873; -0.3873 0.3467], [0.2722, 0.1969], 0.1831),\n    Ellipse([1.4125 -2.1777; -2.1777 6.7775], [-1.228, -0.0521], 0.3295),\n    Ellipse([1.7018 0.8141; 0.8141 1.7538], [-0.4049, 1.5713], 0.2077),\n    Ellipse([0.9742 -0.7202; -0.7202 1.5444], [0.0265, 0.5623], 0.2362),\n    Ellipse([0.6798 -0.1424; -0.1424 0.6871], [-0.4301, -1.0157], 0.3284),\n    Ellipse([0.1796 -0.1423; -0.1423 2.6181], [-0.3286, 0.557], 0.4931),\n];\nnothing #hide","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"We visualise the ellipses using the Plots package:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"function plot_ellipse(plot, ellipse::Ellipse)\n    A, b, c = ellipse.A, ellipse.b, ellipse.c\n    θ = range(0, 2pi + 0.05; step = 0.05)\n    # Some linear algebra to convert θ into (x,y) coordinates.\n    x_y = √A \\ (√(b' * (A \\ b) - c) .* hcat(cos.(θ), sin.(θ)) .- (√A \\ b)')'\n    Plots.plot!(plot, x_y[1, :], x_y[2, :]; label = nothing, c = :navy)\n    return\nend\n\nplot = Plots.plot(; size = (600, 600))\nfor ellipse in ellipses\n    plot_ellipse(plot, ellipse)\nend\nplot","category":"page"},{"location":"tutorials/conic/min_ellipse/#Build-the-model","page":"Minimal ellipses","title":"Build the model","text":"","category":"section"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"Now let's build the model, using the change-of-variables P² = P^2 and P_q = P q. We'll recover the true value of P and q after the solve.","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"model = Model(SCS.Optimizer)\n# We need to use a tighter tolerance for this example, otherwise the bounding\n# ellipse won't actually be bounding...\nset_attribute(model, \"eps_rel\", 1e-6)\nset_silent(model)\nm, n = length(ellipses), size(first(ellipses).A, 1)\n@variable(model, τ[1:m] >= 0)\n@variable(model, P²[1:n, 1:n], PSD)\n@variable(model, P_q[1:n])\n\nfor (i, ellipse) in enumerate(ellipses)\n    A, b, c = ellipse.A, ellipse.b, ellipse.c\n    X = [\n        #! format: off\n        (P² - τ[i] * A)   (P_q - τ[i] * b) zeros(n, n)\n        (P_q - τ[i] * b)' (-1 - τ[i] * c)  P_q'\n        zeros(n, n)       P_q              -P²\n        #! format: on\n    ]\n    @constraint(model, LinearAlgebra.Symmetric(X) <= 0, PSDCone())\nend","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"We cannot directly represent the objective log(det(P)), so we introduce the conic reformulation:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"@variable(model, log_det_P)\n@constraint(model, [log_det_P; 1; vec(P²)] in MOI.LogDetConeSquare(n))\n@objective(model, Max, log_det_P)","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"Now, solve the program:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"optimize!(model)\nsolution_summary(model)","category":"page"},{"location":"tutorials/conic/min_ellipse/#Results","page":"Minimal ellipses","title":"Results","text":"","category":"section"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"After solving the model to optimality we can recover the solution in terms of P and q:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"P = sqrt(value.(P²))\nq = P \\ value.(P_q)","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"Finally, overlaying the solution in the plot we see the minimal area enclosing ellipsoid:","category":"page"},{"location":"tutorials/conic/min_ellipse/","page":"Minimal ellipses","title":"Minimal ellipses","text":"Plots.plot!(\n    plot,\n    [tuple(P \\ [cos(θ) - q[1], sin(θ) - q[2]]...) for θ in 0:0.05:(2pi+0.05)];\n    c = :crimson,\n    label = nothing,\n)","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"EditURL = \"https://github.com/jump-dev/CSDP.jl/blob/3a2e4c7c48af7660a5ba258a4cb6549ab0970366/README.md\"","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"(Image: )","category":"page"},{"location":"packages/CSDP/#CSDP.jl","page":"jump-dev/CSDP.jl","title":"CSDP.jl","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"(Image: Build Status) (Image: codecov)","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"CSDP.jl is a wrapper for the COIN-OR SemiDefinite Programming solver.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"The wrapper has two components:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"a thin wrapper around the low-level C API\nan interface to MathOptInterface","category":"page"},{"location":"packages/CSDP/#Affiliation","page":"jump-dev/CSDP.jl","title":"Affiliation","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"This wrapper is maintained by the JuMP community and is not a COIN-OR project.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"The original algorithm is described by B. Borchers (1999). CSDP, A C Library for Semidefinite Programming. Optimization Methods and Software. 11(1), 613-623. [preprint]","category":"page"},{"location":"packages/CSDP/#License","page":"jump-dev/CSDP.jl","title":"License","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"CSDP.jl is licensed under the MIT License.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"The underlying solver, coin-or/Csdp, is licensed under the Eclipse public license.","category":"page"},{"location":"packages/CSDP/#Installation","page":"jump-dev/CSDP.jl","title":"Installation","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"Install CSDP using Pkg.add:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"import Pkg\nPkg.add(\"CSDP\")","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"In addition to installing the CSDP.jl package, this will also download and install the CSDP binaries. You do not need to install CSDP separately.","category":"page"},{"location":"packages/CSDP/#Use-with-JuMP","page":"jump-dev/CSDP.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"To use CSDP with JuMP, use CSDP.Optimizer:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"using JuMP, CSDP\nmodel = Model(CSDP.Optimizer)\nset_attribute(model, \"maxiter\", 1000)","category":"page"},{"location":"packages/CSDP/#MathOptInterface-API","page":"jump-dev/CSDP.jl","title":"MathOptInterface API","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"The CSDP optimizer supports the following constraints and attributes.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"List of supported objective functions:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"MOI.ObjectiveFunction{MOI.ScalarAffineFunction{Float64}}","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"List of supported variable types:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"MOI.Nonnegatives\nMOI.PositiveSemidefiniteConeTriangle","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"List of supported constraint types:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"MOI.ScalarAffineFunction{Float64} in MOI.EqualTo{Float64}","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"List of supported model attributes:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"MOI.ObjectiveSense()","category":"page"},{"location":"packages/CSDP/#Options","page":"jump-dev/CSDP.jl","title":"Options","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"The CSDP options are listed in the table below.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"Name Default Value Explanation\naxtol 1.0e-8 Tolerance for primal feasibility\natytol 1.0e-8 Tolerance for dual feasibility\nobjtol 1.0e-8 Tolerance for relative duality gap\npinftol 1.0e8 Tolerance for determining primal infeasibility\ndinftol 1.0e8 Tolerance for determining dual infeasibility\nmaxiter 100 Limit for the total number of iterations\nminstepfrac 0.90 The minstepfrac and maxstepfrac parameters determine how close to the edge of the feasible region CSDP will step\nmaxstepfrac 0.97 The minstepfrac and maxstepfrac parameters determine how close to the edge of the feasible region CSDP will step\nminstepp 1.0e-8 If the primal step is shorter than minstepp then CSDP declares a line search failure\nminstepd 1.0e-8 If the dual step is shorter than minstepd then CSDP declares a line search failure\nusexzgap 1 If usexzgap is 0 then CSDP will use the objective duality gap d - p instead of the XY duality gap ⟨Z, X⟩\ntweakgap 0 If tweakgap is set to 1, and usexzgap is set to 0, then CSDP will attempt to \"fix\" negative duality gaps\naffine 0 If affine is set to 1, then CSDP will take only primal-dual affine steps and not make use of the barrier term. This can be useful for some problems that do not have feasible solutions that are strictly in the interior of the cone of semidefinite matrices\nperturbobj 1 The perturbobj parameter determines whether the objective function will be perturbed to help deal with problems that have unbounded optimal solution sets. If perturbobj is 0, then the objective will not be perturbed. If perturbobj is 1, then the objective function will be perturbed by a default amount. Larger values of perturbobj (for example, 100) increase the size of the perturbation. This can be helpful in solving some difficult problems.\nfastmode 0 The fastmode parameter determines whether or not CSDP will skip certain time consuming operations that slightly improve the accuracy of the solutions. If fastmode is set to 1, then CSDP may be somewhat faster, but also somewhat less accurate\nprintlevel 1 The printlevel parameter determines how much debugging information is output. Use a printlevel of 0 for no output and a printlevel of 1 for normal output. Higher values of printlevel will generate more debugging output","category":"page"},{"location":"packages/CSDP/#Problem-representation","page":"jump-dev/CSDP.jl","title":"Problem representation","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"The primal is represented internally by CSDP as follows:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"max ⟨C, X⟩\n      A(X) = a\n         X ⪰ 0","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"where A(X) = [⟨A_1, X⟩, ..., ⟨A_m, X⟩]. The corresponding dual is:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"min ⟨a, y⟩\n     A'(y) - C = Z\n             Z ⪰ 0","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"where A'(y) = y_1A_1 + ... + y_mA_m","category":"page"},{"location":"packages/CSDP/#Termination-criteria","page":"jump-dev/CSDP.jl","title":"Termination criteria","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"CSDP will terminate successfully (or partially) in the following cases:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"If CSDP finds X, Z ⪰ 0 such that the following 3 tolerances are satisfied:\nprimal feasibility tolerance: ||A(x) - a||_2 / (1 + ||a||_2) < axtol\ndual feasibility tolerance: ||A'(y) - C - Z||_F / (1 + ||C||_F) < atytol\nrelative duality gap tolerance: gap / (1 + |⟨a, y⟩| + |⟨C, X⟩|) < objtol\nobjective duality gap: if usexygap is 0, gap = ⟨a, y⟩ - ⟨C, X⟩\nXY duality gap: if usexygap is 1, gap = ⟨Z, X⟩\nthen it returns 0.\nIf CSDP finds y and Z ⪰ 0 such that -⟨a, y⟩ / ||A'(y) - Z||_F > pinftol, it returns 1 with y such that ⟨a, y⟩ = -1.\nIf CSDP finds X ⪰ 0 such that ⟨C, X⟩ / ||A(X)||_2 > dinftol, it returns 2 with X such that ⟨C, X⟩ = 1.\nIf CSDP finds X, Z ⪰ 0 such that the following 3 tolerances are satisfied with 1000*axtol, 1000*atytol and 1000*objtol but at least one of them is not satisfied with axtol, atytol and objtol and cannot make progress, then it returns 3.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"In addition, if the printlevel option is at least 1, the following will be printed:","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"If the return code is 1, CSDP will print ⟨a, y⟩ and ||A'(y) - Z||_F\nIf the return code is 2, CSDP will print ⟨C, X⟩ and ||A(X)||_F\nOtherwise, CSDP will print\nthe primal/dual objective value,\nthe relative primal/dual infeasibility,\nthe objective duality gap ⟨a, y⟩ - ⟨C, X⟩ and objective relative duality gap (⟨a, y⟩ - ⟨C, X⟩) / (1 + |⟨a, y⟩| + |⟨C, X⟩|),\nthe XY duality gap ⟨Z, X⟩ and XY relative duality gap ⟨Z, X⟩ / (1 + |⟨a, y⟩| + |⟨C, X⟩|)\nand the DIMACS error measures.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"In theory, for feasible primal and dual solutions, ⟨a, y⟩ - ⟨C, X⟩ = ⟨Z, X⟩, so the objective and XY duality gap should be equivalent. However, in practice, there are sometimes solution which satisfy primal and dual feasibility tolerances but have objective duality gap which are not close to XY duality gap. In some cases, the objective duality gap may even become negative (hence the tweakgap option). This is the reason usexygap is 1 by default.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"CSDP considers that X ⪰ 0 (resp. Z ⪰ 0) is satisfied when the Cholesky factorizations can be computed. In practice, this is somewhat more conservative than simply requiring all eigenvalues to be nonnegative.","category":"page"},{"location":"packages/CSDP/#Status","page":"jump-dev/CSDP.jl","title":"Status","text":"","category":"section"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"The table below shows how the different CSDP statuses are converted to the MathOptInterface statuses.","category":"page"},{"location":"packages/CSDP/","page":"jump-dev/CSDP.jl","title":"jump-dev/CSDP.jl","text":"CSDP code State Description MOI status\n0 Success SDP solved MOI.OPTIMAL\n1 Success The problem is primal infeasible, and we have a certificate MOI.INFEASIBLE\n2 Success The problem is dual infeasible, and we have a certificate MOI.DUAL_INFEASIBLE\n3 Partial Success A solution has been found, but full accuracy was not achieved MOI.ALMOST_OPTIMAL\n4 Failure Maximum iterations reached MOI.ITERATION_LIMIT\n5 Failure Stuck at edge of primal feasibility MOI.SLOW_PROGRESS\n6 Failure Stuck at edge of dual infeasibility MOI.SLOW_PROGRESS\n7 Failure Lack of progress MOI.SLOW_PROGRESS\n8 Failure X, Z, or O was singular MOI.NUMERICAL_ERROR\n9 Failure Detected NaN or Inf values MOI.NUMERICAL_ERROR","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"EditURL = \"start_values.jl\"","category":"page"},{"location":"tutorials/conic/start_values/#Primal-and-dual-warm-starts","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"","category":"section"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"Some conic solvers have the ability to set warm-starts for the primal and dual solution. This can improve performance, particularly if you are repeatedly solving a sequence of related problems.","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"tip: Tip\nSee set_start_values for a generic implementation of this function that was added to JuMP after this tutorial was written.","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"In this tutorial, we demonstrate how to write a function that sets the primal and dual starts as the optimal solution stored in a model. It is intended to be a starting point for which you can modify if you want to do something similar in your own code.","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"warning: Warning\nThis tutorial does not set start values for nonlinear models.","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"using JuMP\nimport SCS","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"The main component of this tutorial is the following function. The most important observation is that we cache all of the solution values first, and then we modify the model second. (Alternating between querying a value and modifying the model is not allowed in JuMP.)","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"function set_optimal_start_values(model::Model)\n    # Store a mapping of the variable primal solution\n    variable_primal = Dict(x => value(x) for x in all_variables(model))\n    # In the following, we loop through every constraint and store a mapping\n    # from the constraint index to a tuple containing the primal and dual\n    # solutions.\n    constraint_solution = Dict()\n    for (F, S) in list_of_constraint_types(model)\n        # We add a try-catch here because some constraint types might not\n        # support getting the primal or dual solution.\n        try\n            for ci in all_constraints(model, F, S)\n                constraint_solution[ci] = (value(ci), dual(ci))\n            end\n        catch\n            @info(\"Something went wrong getting $F-in-$S. Skipping\")\n        end\n    end\n    # Now we can loop through our cached solutions and set the starting values.\n    for (x, primal_start) in variable_primal\n        set_start_value(x, primal_start)\n    end\n    for (ci, (primal_start, dual_start)) in constraint_solution\n        set_start_value(ci, primal_start)\n        set_dual_start_value(ci, dual_start)\n    end\n    return\nend","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"To test our function, we use the following linear program:","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"model = Model(SCS.Optimizer)\n@variable(model, x[1:3] >= 0)\n@constraint(model, sum(x) <= 1)\n@objective(model, Max, sum(i * x[i] for i in 1:3))\noptimize!(model)","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"By looking at the log, we can see that SCS took 75 iterations to find the optimal solution. Now we set the optimal solution as our starting point:","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"set_optimal_start_values(model)","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"and we re-optimize:","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"optimize!(model)","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"Now the optimization terminates after 0 iterations because our starting point is already optimal.","category":"page"},{"location":"tutorials/conic/start_values/","page":"Primal and dual warm-starts","title":"Primal and dual warm-starts","text":"Note that some solvers do not support setting some parts of the starting solution, for example, they may support only set_start_value for variables. If you encounter an UnsupportedSupported attribute error for MOI.VariablePrimalStart, MOI.ConstraintPrimalStart, or MOI.ConstraintDualStart, comment out the corresponding part of the set_optimal_start_values function.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP, HiGHS\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/solutions/#jump_solutions","page":"Solutions","title":"Solutions","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"This section of the manual describes how to access a solved solution to a problem. It uses the following model as an example:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> begin\n           model = Model(HiGHS.Optimizer)\n           set_silent(model)\n           @variable(model, x >= 0)\n           @variable(model, y[[:a, :b]] <= 1)\n           @objective(model, Max, -12x - 20y[:a])\n           @expression(model, my_expr, 6x + 8y[:a])\n           @constraint(model, my_expr >= 100)\n           @constraint(model, c1, 7x + 12y[:a] >= 120)\n           optimize!(model)\n           print(model)\n       end\nMax -12 x - 20 y[a]\nSubject to\n 6 x + 8 y[a] ≥ 100\n c1 : 7 x + 12 y[a] ≥ 120\n x ≥ 0\n y[a] ≤ 1\n y[b] ≤ 1","category":"page"},{"location":"manual/solutions/#Solutions-summary","page":"Solutions","title":"Solutions summary","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"solution_summary can be used for checking the summary of the optimization solutions.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> solution_summary(model)\n* Solver : HiGHS\n\n* Status\n  Result count       : 1\n  Termination status : OPTIMAL\n  Message from the solver:\n  \"kHighsModelStatusOptimal\"\n\n* Candidate solution (result #1)\n  Primal status      : FEASIBLE_POINT\n  Dual status        : FEASIBLE_POINT\n  Objective value    : -2.05143e+02\n  Objective bound    : -0.00000e+00\n  Relative gap       : Inf\n  Dual objective value : -2.05143e+02\n\n* Work counters\n  Solve time (sec)   : 6.70987e-04\n  Simplex iterations : 2\n  Barrier iterations : 0\n  Node count         : -1\n\njulia> solution_summary(model; verbose = true)\n* Solver : HiGHS\n\n* Status\n  Result count       : 1\n  Termination status : OPTIMAL\n  Message from the solver:\n  \"kHighsModelStatusOptimal\"\n\n* Candidate solution (result #1)\n  Primal status      : FEASIBLE_POINT\n  Dual status        : FEASIBLE_POINT\n  Objective value    : -2.05143e+02\n  Objective bound    : -0.00000e+00\n  Relative gap       : Inf\n  Dual objective value : -2.05143e+02\n  Primal solution :\n    x : 1.54286e+01\n    y[a] : 1.00000e+00\n    y[b] : 1.00000e+00\n  Dual solution :\n    c1 : 1.71429e+00\n\n* Work counters\n  Solve time (sec)   : 6.70987e-04\n  Simplex iterations : 2\n  Barrier iterations : 0\n  Node count         : -1","category":"page"},{"location":"manual/solutions/#Why-did-the-solver-stop?","page":"Solutions","title":"Why did the solver stop?","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Usetermination_status to understand why the solver stopped.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> termination_status(model)\nOPTIMAL::TerminationStatusCode = 1","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"The MOI.TerminationStatusCode enum describes the full list of statuses that could be returned.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Common return values include OPTIMAL, LOCALLY_SOLVED, INFEASIBLE, DUAL_INFEASIBLE, and TIME_LIMIT.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"info: Info\nA return status of OPTIMAL means the solver found (and proved) a globally optimal solution. A return status of LOCALLY_SOLVED means the solver found a locally optimal solution (which may also be globally optimal, but it could not prove so).","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"warning: Warning\nA return status of DUAL_INFEASIBLE does not guarantee that the primal is unbounded. When the dual is infeasible, the primal is unbounded if there exists a feasible primal solution.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Use raw_status to get a solver-specific string explaining why the optimization stopped:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> raw_status(model)\n\"kHighsModelStatusOptimal\"","category":"page"},{"location":"manual/solutions/#Primal-solutions","page":"Solutions","title":"Primal solutions","text":"","category":"section"},{"location":"manual/solutions/#Primal-solution-status","page":"Solutions","title":"Primal solution status","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Use primal_status to return an MOI.ResultStatusCode enum describing the status of the primal solution.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> primal_status(model)\nFEASIBLE_POINT::ResultStatusCode = 1","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Other common returns are NO_SOLUTION, and INFEASIBILITY_CERTIFICATE. The first means that the solver doesn't have a solution to return, and the second means that the primal solution is a certificate of dual infeasibility (a primal unbounded ray).","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"You can also use has_values, which returns true if there is a solution that can be queried, and false otherwise.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> has_values(model)\ntrue","category":"page"},{"location":"manual/solutions/#Objective-values","page":"Solutions","title":"Objective values","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"The objective value of a solved problem can be obtained via objective_value. The best known bound on the optimal objective value can be obtained via objective_bound. If the solver supports it, the value of the dual objective can be obtained via dual_objective_value.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> objective_value(model)\n-205.14285714285714\n\njulia> objective_bound(model)  # HiGHS only implements objective bound for MIPs\n-0.0\n\njulia> dual_objective_value(model)\n-205.1428571428571","category":"page"},{"location":"manual/solutions/#Primal-solution-values","page":"Solutions","title":"Primal solution values","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If the solver has a primal solution to return, use value to access it:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> value(x)\n15.428571428571429","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Broadcast value over containers:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> value.(y)\n1-dimensional DenseAxisArray{Float64,1,...} with index sets:\n    Dimension 1, [:a, :b]\nAnd data, a 2-element Vector{Float64}:\n 1.0\n 1.0","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"value also works on expressions:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> value(my_expr)\n100.57142857142857","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"and constraints:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> value(c1)\n120.0","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"info: Info\nCalling value on a constraint returns the constraint function evaluated at the solution.","category":"page"},{"location":"manual/solutions/#Dual-solutions","page":"Solutions","title":"Dual solutions","text":"","category":"section"},{"location":"manual/solutions/#Dual-solution-status","page":"Solutions","title":"Dual solution status","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Use dual_status to return an MOI.ResultStatusCode enum describing the status of the dual solution.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> dual_status(model)\nFEASIBLE_POINT::ResultStatusCode = 1","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Other common returns are NO_SOLUTION, and INFEASIBILITY_CERTIFICATE. The first means that the solver doesn't have a solution to return, and the second means that the dual solution is a certificate of primal infeasibility (a dual unbounded ray).","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"You can also use has_duals, which returns true if there is a solution that can be queried, and false otherwise.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> has_duals(model)\ntrue","category":"page"},{"location":"manual/solutions/#Dual-solution-values","page":"Solutions","title":"Dual solution values","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If the solver has a dual solution to return, use dual to access it:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> dual(c1)\n1.7142857142857142","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Query the duals of variable bounds using LowerBoundRef, UpperBoundRef, and FixRef:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> dual(LowerBoundRef(x))\n0.0\n\njulia> dual.(UpperBoundRef.(y))\n1-dimensional DenseAxisArray{Float64,1,...} with index sets:\n    Dimension 1, [:a, :b]\nAnd data, a 2-element Vector{Float64}:\n -0.5714285714285694\n  0.0","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"warning: Warning\nJuMP's definition of duality is independent of the objective sense. That is, the sign of feasible duals associated with a constraint depends on the direction of the constraint and not whether the problem is maximization or minimization. This is a different convention from linear programming duality in some common textbooks. If you have a linear program, and you want the textbook definition, you probably want to use shadow_price and reduced_cost instead.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> shadow_price(c1)\n1.7142857142857142\n\njulia> reduced_cost(x)\n-0.0\n\njulia> reduced_cost.(y)\n1-dimensional DenseAxisArray{Float64,1,...} with index sets:\n    Dimension 1, [:a, :b]\nAnd data, a 2-element Vector{Float64}:\n  0.5714285714285694\n -0.0","category":"page"},{"location":"manual/solutions/#Recommended-workflow","page":"Solutions","title":"Recommended workflow","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"The recommended workflow for solving a model and querying the solution is something like the following:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> begin\n           if termination_status(model) == OPTIMAL\n               println(\"Solution is optimal\")\n           elseif termination_status(model) == TIME_LIMIT && has_values(model)\n               println(\"Solution is suboptimal due to a time limit, but a primal solution is available\")\n           else\n               error(\"The model was not solved correctly.\")\n           end\n           println(\"  objective value = \", objective_value(model))\n           if primal_status(model) == FEASIBLE_POINT\n               println(\"  primal solution: x = \", value(x))\n           end\n           if dual_status(model) == FEASIBLE_POINT\n               println(\"  dual solution: c1 = \", dual(c1))\n           end\n       end\nSolution is optimal\n  objective value = -205.14285714285714\n  primal solution: x = 15.428571428571429\n  dual solution: c1 = 1.7142857142857142","category":"page"},{"location":"manual/solutions/#OptimizeNotCalled-errors","page":"Solutions","title":"OptimizeNotCalled errors","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Due to differences in how solvers cache solutions internally, modifying a model after calling optimize! will reset the model into the MOI.OPTIMIZE_NOT_CALLED state. If you then attempt to query solution information, an OptimizeNotCalled error will be thrown.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If you are iteratively querying solution information and modifying a model, query all the results first, then modify the problem.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"For example, instead of:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 0);\n\njulia> optimize!(model)\n\njulia> termination_status(model)\nOPTIMAL::TerminationStatusCode = 1\n\njulia> set_upper_bound(x, 1)\n\njulia> x_val = value(x)\n┌ Warning: The model has been modified since the last call to `optimize!` (or `optimize!` has not been called yet). If you are iteratively querying solution information and modifying a model, query all the results first, then modify the model.\n└ @ JuMP ~/work/JuMP.jl/JuMP.jl/src/optimizer_interface.jl:712\nERROR: OptimizeNotCalled()\nStacktrace:\n[...]\n\njulia> termination_status(model)\nOPTIMIZE_NOT_CALLED::TerminationStatusCode = 0","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"do","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 0);\n\njulia> optimize!(model);\n\njulia> x_val = value(x)\n0.0\n\njulia> termination_status(model)\nOPTIMAL::TerminationStatusCode = 1\n\njulia> set_upper_bound(x, 1)\n\njulia> set_lower_bound(x, x_val)\n\njulia> termination_status(model)\nOPTIMIZE_NOT_CALLED::TerminationStatusCode = 0","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If you know that your particular solver supports querying solution information after modifications, you can use direct_model to bypass the MOI.OPTIMIZE_NOT_CALLED state:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> model = direct_model(HiGHS.Optimizer());\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 0);\n\njulia> optimize!(model)\n\njulia> termination_status(model)\nOPTIMAL::TerminationStatusCode = 1\n\njulia> set_upper_bound(x, 1)\n\njulia> x_val = value(x)\n0.0\n\njulia> set_lower_bound(x, x_val)\n\njulia> termination_status(model)\nOPTIMAL::TerminationStatusCode = 1","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"warning: Warning\nBe careful doing this. If your particular solver does not support querying solution information after modification, it may silently return incorrect solutions or throw an error.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"# TODO: How to accurately measure the solve time.","category":"page"},{"location":"manual/solutions/#Accessing-attributes","page":"Solutions","title":"Accessing attributes","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"MathOptInterface defines many model attributes that can be queried. Some attributes can be directly accessed by getter functions. These include:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"solve_time\nrelative_gap\nsimplex_iterations\nbarrier_iterations\nnode_count","category":"page"},{"location":"manual/solutions/#Sensitivity-analysis-for-LP","page":"Solutions","title":"Sensitivity analysis for LP","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Given an LP problem and an optimal solution corresponding to a basis, we can question how much an objective coefficient or standard form right-hand side coefficient (c.f., normalized_rhs) can change without violating primal or dual feasibility of the basic solution.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Note that not all solvers compute the basis, and for sensitivity analysis, the solver interface must implement MOI.ConstraintBasisStatus.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"tip: Tip\nRead the Sensitivity analysis of a linear program for more information on sensitivity analysis.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"To give a simple example, we could analyze the sensitivity of the optimal solution to the following (non-degenerate) LP problem:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> begin\n           model = Model(HiGHS.Optimizer)\n           set_silent(model)\n           @variable(model, x[1:2])\n           set_lower_bound(x[2], -0.5)\n           set_upper_bound(x[2], 0.5)\n           @constraint(model, c1, x[1] + x[2] <= 1)\n           @constraint(model, c2, x[1] - x[2] <= 1)\n           @objective(model, Max, x[1])\n           print(model)\n       end\nMax x[1]\nSubject to\n c1 : x[1] + x[2] ≤ 1\n c2 : x[1] - x[2] ≤ 1\n x[2] ≥ -0.5\n x[2] ≤ 0.5","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"To analyze the sensitivity of the problem we could check the allowed perturbation ranges of, for example, the cost coefficients and the right-hand side coefficient of the constraint c1 as follows:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> optimize!(model)\n\njulia> value.(x)\n2-element Vector{Float64}:\n  1.0\n -0.0\n\njulia> report = lp_sensitivity_report(model);\n\njulia> x1_lo, x1_hi = report[x[1]]\n(-1.0, Inf)\n\njulia> println(\"The objective coefficient of x[1] could decrease by $(x1_lo) or increase by $(x1_hi).\")\nThe objective coefficient of x[1] could decrease by -1.0 or increase by Inf.\n\njulia> x2_lo, x2_hi = report[x[2]]\n(-1.0, 1.0)\n\njulia> println(\"The objective coefficient of x[2] could decrease by $(x2_lo) or increase by $(x2_hi).\")\nThe objective coefficient of x[2] could decrease by -1.0 or increase by 1.0.\n\njulia> c_lo, c_hi = report[c1]\n(-1.0, 1.0)\n\njulia> println(\"The RHS of c1 could decrease by $(c_lo) or increase by $(c_hi).\")\nThe RHS of c1 could decrease by -1.0 or increase by 1.0.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"The range associated with a variable is the range of the allowed perturbation of the corresponding objective coefficient. Note that the current primal solution remains optimal within this range; however the corresponding dual solution might change since a cost coefficient is perturbed. Similarly, the range associated with a constraint is the range of the allowed perturbation of the corresponding right-hand side coefficient. In this range the current dual solution remains optimal, but the optimal primal solution might change.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If the problem is degenerate, there are multiple optimal bases and hence these ranges might not be as intuitive and seem too narrow, for example, a larger cost coefficient perturbation might not invalidate the optimality of the current primal solution. Moreover, if a problem is degenerate, due to finite precision, it can happen that, for example, a perturbation seems to invalidate a basis even though it doesn't (again providing too narrow ranges). To prevent this, increase the atol keyword argument to lp_sensitivity_report. Note that this might make the ranges too wide for numerically challenging instances. Thus, do not blindly trust these ranges, especially not for highly degenerate or numerically unstable instances.","category":"page"},{"location":"manual/solutions/#Conflicts","page":"Solutions","title":"Conflicts","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"When the model you input is infeasible, some solvers can help you find the cause of this infeasibility by offering a conflict, that is, a subset of the constraints that create this infeasibility. Depending on the solver, this can also be called an IIS (irreducible inconsistent subsystem).","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If supported by the solver, use compute_conflict! to trigger the computation of a conflict. Once this process is finished, query the MOI.ConflictStatus attribute to check if a conflict was found.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If found, copy the IIS to a new model using copy_conflict, which you can then print or write to a file for easier debugging:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> using JuMP\n\njulia> import Gurobi\n\njulia> model = Model(Gurobi.Optimizer)\nA JuMP Model\nFeasibility problem with:\nVariables: 0\nModel mode: AUTOMATIC\nCachingOptimizer state: EMPTY_OPTIMIZER\nSolver name: Gurobi\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 0)\nx\n\njulia> @constraint(model, c1, x >= 2)\nc1 : x ≥ 2.0\n\njulia> @constraint(model, c2, x <= 1)\nc2 : x ≤ 1.0\n\njulia> optimize!(model)\n\njulia> compute_conflict!(model)\n\njulia> if get_attribute(model, MOI.ConflictStatus()) == MOI.CONFLICT_FOUND\n           iis_model, _ = copy_conflict(model)\n           print(iis_model)\n       end\nFeasibility\nSubject to\n c1 : x ≥ 2.0\n c2 : x ≤ 1.0","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If you need more control over the list of constraints that appear in the conflict, iterate over the list of constraints and query the MOI.ConstraintConflictStatus attribute:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> list_of_conflicting_constraints = ConstraintRef[]\nConstraintRef[]\n\njulia> for (F, S) in list_of_constraint_types(model)\n           for con in all_constraints(model, F, S)\n               if get_attribute(con, MOI.ConstraintConflictStatus()) == MOI.IN_CONFLICT\n                   push!(list_of_conflicting_constraints, con)\n               end\n           end\n       end\n\njulia> list_of_conflicting_constraints\n2-element Vector{ConstraintRef}:\n c1 : x ≥ 2.0\n c2 : x ≤ 1.0","category":"page"},{"location":"manual/solutions/#Multiple-solutions","page":"Solutions","title":"Multiple solutions","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Some solvers support returning multiple solutions. You can check how many solutions are available to query using result_count.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Functions for querying the solutions, for example, primal_status, dual_status, value, dual, and solution_summary all take an additional keyword argument result which can be used to specify which result to return.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"warning: Warning\nEven if termination_status is OPTIMAL, some of the returned solutions may be suboptimal. However, if the solver found at least one optimal solution, then result = 1 will always return an optimal solution. Use objective_value to assess the quality of the remaining solutions.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> using JuMP\n\njulia> import MultiObjectiveAlgorithms as MOA\n\njulia> import HiGHS\n\njulia> model = Model(() -> MOA.Optimizer(HiGHS.Optimizer));\n\njulia> set_attribute(model, MOA.Algorithm(), MOA.Dichotomy())\n\njulia> set_silent(model)\n\njulia> @variable(model, x1 >= 0)\nx1\n\njulia> @variable(model, 0 <= x2 <= 3)\nx2\n\njulia> @objective(model, Min, [3x1 + x2, -x1 - 2x2])\n2-element Vector{AffExpr}:\n 3 x1 + x2\n -x1 - 2 x2\n\njulia> @constraint(model, 3x1 - x2 <= 6)\n3 x1 - x2 ≤ 6\n\njulia> optimize!(model)\n\njulia> solution_summary(model; result = 1)\n* Solver : MOA[algorithm=MultiObjectiveAlgorithms.Dichotomy, optimizer=HiGHS]\n\n* Status\n  Result count       : 3\n  Termination status : OPTIMAL\n  Message from the solver:\n  \"Solve complete. Found 3 solution(s)\"\n\n* Candidate solution (result #1)\n  Primal status      : FEASIBLE_POINT\n  Dual status        : NO_SOLUTION\n  Objective value    : [0.00000e+00,0.00000e+00]\n  Objective bound    : [0.00000e+00,-9.00000e+00]\n  Relative gap       : Inf\n  Dual objective value : -9.00000e+00\n\n* Work counters\n  Solve time (sec)   : 1.82720e-03\n  Simplex iterations : 1\n  Barrier iterations : 0\n  Node count         : -1\n\njulia> for i in 1:result_count(model)\n           println(\"Solution $i\")\n           println(\"   x = \", value.([x1, x2]; result = i))\n           println(\" obj = \", objective_value(model; result = i))\n       end\nSolution 1\n   x = [0.0, 0.0]\n obj = [0.0, 0.0]\nSolution 2\n   x = [0.0, 3.0]\n obj = [3.0, -6.0]\nSolution 3\n   x = [3.0, 3.0]\n obj = [12.0, -9.0]","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"tip: Tip\nThe Multi-objective knapsack tutorial provides more examples of querying multiple solutions.","category":"page"},{"location":"manual/solutions/#Checking-feasibility-of-solutions","page":"Solutions","title":"Checking feasibility of solutions","text":"","category":"section"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"To check the feasibility of a primal solution, use primal_feasibility_report, which takes a model, a dictionary mapping each variable to a primal solution value (defaults to the last solved solution), and a tolerance atol (defaults to 0.0).","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"The function returns a dictionary which maps the infeasible constraint references to the distance between the primal value of the constraint and the nearest point in the corresponding set. A point is classed as infeasible if the distance is greater than the supplied tolerance atol.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"# Add a filter here because the output of the dictionary is not ordered, and\n# changes in printing order will cause the doctest to fail.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x >= 1, Int);\n\njulia> @variable(model, y);\n\njulia> @constraint(model, c1, x + y <= 1.95);\n\njulia> point = Dict(x => 1.9, y => 0.06);\n\njulia> primal_feasibility_report(model, point)\nDict{Any, Float64} with 2 entries:\n  x integer         => 0.1\n  c1 : x + y ≤ 1.95 => 0.01\n\njulia> primal_feasibility_report(model, point; atol = 0.02)\nDict{Any, Float64} with 1 entry:\n  x integer => 0.1","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"If the point is feasible, an empty dictionary is returned:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> primal_feasibility_report(model, Dict(x => 1.0, y => 0.0))\nDict{Any, Float64}()","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"To use the primal solution from a solve, omit the point argument:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> optimize!(model)\n\njulia> primal_feasibility_report(model; atol = 0.0)\nDict{Any, Float64}()","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Calling primal_feasibility_report without the point argument is useful when primal_status is FEASIBLE_POINT or NEARLY_FEASIBLE_POINT, and you want to assess the solution quality.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"warning: Warning\nTo apply primal_feasibility_report to infeasible models, you must also provide a candidate point (solvers generally do not provide one). To diagnose the source of infeasibility, see Conflicts.","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"Pass skip_mising = true to skip constraints which contain variables that are not in point:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> primal_feasibility_report(model, Dict(x => 2.1); skip_missing = true)\nDict{Any, Float64} with 1 entry:\n  x integer => 0.1","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"You can also use the functional form, where the first argument is a function that maps variables to their primal values:","category":"page"},{"location":"manual/solutions/","page":"Solutions","title":"Solutions","text":"julia> optimize!(model)\n\njulia> primal_feasibility_report(v -> value(v), model)\nDict{Any, Float64}()","category":"page"},{"location":"extensions/DimensionalData/#DimensionalData.jl","page":"rafaqz/DimensionalData.jl","title":"DimensionalData.jl","text":"","category":"section"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"DimensionalData.jl provides tools and abstractions for working with rectangular arrays that have named dimensions.","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"compat: Compat\nUsing the DimensionalData extension with JuMP requires Julia v1.9 or later.","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"The DimensionalData extension in JuMP lets you construct a DimensionalData.DimArray as an alternative to Containers.DenseAxisArray in the JuMP macros.","category":"page"},{"location":"extensions/DimensionalData/#License","page":"rafaqz/DimensionalData.jl","title":"License","text":"","category":"section"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"DimensionalData.jl is licensed under the MIT license.","category":"page"},{"location":"extensions/DimensionalData/#Installation","page":"rafaqz/DimensionalData.jl","title":"Installation","text":"","category":"section"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"Install DimensionalData using Pkg.add:","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"import Pkg\nPkg.add(\"DimensionalData\")","category":"page"},{"location":"extensions/DimensionalData/#Use-with-JuMP","page":"rafaqz/DimensionalData.jl","title":"Use with JuMP","text":"","category":"section"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"Activate the extension by loading both JuMP and DimensionalData:","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"julia> using JuMP, DimensionalData","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"Then, pass container = DimensionalData.DimArray in the @variable, @constraint, or @expression macros:","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"julia> model = Model();\n\njulia> @variable(\n           model,\n           x[i = 2:4, j = [\"a\", \"b\"]] >= i,\n           container = DimensionalData.DimArray,\n       )\n3×2 DimArray{VariableRef,2} with dimensions:\n  Dim{:i} Sampled{Int64} 2:4 ForwardOrdered Regular Points,\n  Dim{:j} Categorical{String} String[\"a\", \"b\"] ForwardOrdered\n    \"a\"     \"b\"\n 2  x[2,a]  x[2,b]\n 3  x[3,a]  x[3,b]\n 4  x[4,a]  x[4,b]","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"Here x is a DimensionalData.Dim array object, so indexing uses the DimensionalData syntax:","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"julia> x[At(2), At(\"a\")]\nx[2,a]\n\njulia> x[2, 2]\nx[3,b]","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"You can use container = DimensionalData.DimArray in the @expression macro:","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"julia> @expression(\n           model,\n           expr[j = [\"a\", \"b\"]],\n           sum(x[At(i), At(j)] for i in 2:4),\n           container = DimensionalData.DimArray,\n       )\n2-element DimArray{AffExpr,1} with dimensions:\n  Dim{:j} Categorical{String} String[\"a\", \"b\"] ForwardOrdered\n \"a\"  x[2,a] + x[3,a] + x[4,a]\n \"b\"  x[2,b] + x[3,b] + x[4,b]","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"and in @constraint:","category":"page"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"julia> @constraint(\n           model,\n           [j = [\"a\", \"b\"]],\n           expr[At(j)] <= 1,\n           container = DimensionalData.DimArray,\n       )\n2-element DimArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},1} with dimensions:\n  Dim{:j} Categorical{String} String[\"a\", \"b\"] ForwardOrdered\n \"a\"  x[2,a] + x[3,a] + x[4,a] ≤ 1\n \"b\"  x[2,b] + x[3,b] + x[4,b] ≤ 1","category":"page"},{"location":"extensions/DimensionalData/#Documentation","page":"rafaqz/DimensionalData.jl","title":"Documentation","text":"","category":"section"},{"location":"extensions/DimensionalData/","page":"rafaqz/DimensionalData.jl","title":"rafaqz/DimensionalData.jl","text":"See the DimensionalData.jl documentation for more details on the syntax and features of DimensionalData.DimArray.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"EditURL = \"getting_started_with_JuMP.jl\"","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Getting-started-with-JuMP","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"This tutorial is aimed at providing a quick introduction to writing and solving optimization models with JuMP.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"If you're new to Julia, start by reading Getting started with Julia.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#What-is-JuMP?","page":"Getting started with JuMP","title":"What is JuMP?","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"JuMP (\"Julia for Mathematical Programming\") is an open-source modeling language that is embedded in Julia. It allows users to formulate various classes of optimization problems (linear, mixed-integer, quadratic, conic quadratic, semidefinite, and nonlinear) with easy-to-read code. These problems can then be solved using state-of-the-art open-source and commercial solvers.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"JuMP also makes advanced optimization techniques easily accessible from a high-level language.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#What-is-a-solver?","page":"Getting started with JuMP","title":"What is a solver?","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"A solver is a software package that incorporates algorithms for finding solutions to one or more classes of problem.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"For example, HiGHS is a solver for linear programming (LP) and mixed integer programming (MIP) problems. It incorporates algorithms such as the simplex method and the interior-point method.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"The Supported-solvers table lists the open-source and commercial solvers that JuMP currently supports.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#What-is-MathOptInterface?","page":"Getting started with JuMP","title":"What is MathOptInterface?","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Each solver has its own concepts and data structures for representing optimization models and obtaining results.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"MathOptInterface (MOI) is an abstraction layer that JuMP uses to convert from the problem written in JuMP to the solver-specific data structures for each solver.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"MOI can be used directly, or through a higher-level modeling interface like JuMP.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Because JuMP is built on top of MOI, you'll often see the MathOptInterface. prefix displayed when JuMP types are printed. However, you'll only need to understand and interact with MOI to accomplish advanced tasks such as creating solver-independent callbacks.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Installation","page":"Getting started with JuMP","title":"Installation","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"JuMP is a package for Julia. From Julia, JuMP is installed by using the built-in package manager.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"import Pkg\nPkg.add(\"JuMP\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"You also need to include a Julia package which provides an appropriate solver. One such solver is HiGHS.Optimizer, which is provided by the HiGHS.jl package.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"import Pkg\nPkg.add(\"HiGHS\")","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"See Installation Guide for a list of other solvers you can use.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#An-example","page":"Getting started with JuMP","title":"An example","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Let's solve the following linear programming problem using JuMP and HiGHS. We will first look at the complete code to solve the problem and then go through it step by step.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Here's the problem:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"beginaligned\n min  12x + 20y \n textst  6x + 8y geq 100 \n  7x + 12y geq 120 \n  x geq 0 \n  y in 0 3 \nendaligned","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"And here's the code to solve this problem:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"using JuMP\nusing HiGHS\nmodel = Model(HiGHS.Optimizer)\n@variable(model, x >= 0)\n@variable(model, 0 <= y <= 3)\n@objective(model, Min, 12x + 20y)\n@constraint(model, c1, 6x + 8y >= 100)\n@constraint(model, c2, 7x + 12y >= 120)\nprint(model)\noptimize!(model)\ntermination_status(model)\nprimal_status(model)\ndual_status(model)\nobjective_value(model)\nvalue(x)\nvalue(y)\nshadow_price(c1)\nshadow_price(c2)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Step-by-step","page":"Getting started with JuMP","title":"Step-by-step","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Once JuMP is installed, to use JuMP in your programs write:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"using JuMP","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We also need to include a Julia package which provides an appropriate solver. We want to use HiGHS.Optimizer here which is provided by the HiGHS.jl package:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"using HiGHS","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"JuMP builds problems incrementally in a Model object. Create a model by passing an optimizer to the Model function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = Model(HiGHS.Optimizer)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Variables are modeled using @variable:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, x >= 0)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"info: Info\nThe macro creates a new Julia object, x, in the current scope. We could have made this more explicit by writing:x = @variable(model, x >= 0)but the macro does this automatically for us to save writing x twice.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Variables can have lower and upper bounds:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, 0 <= y <= 30)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"The objective is set using @objective:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@objective(model, Min, 12x + 20y)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Constraints are modeled using @constraint. Here, c1 and c2 are the names of our constraint:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@constraint(model, c1, 6x + 8y >= 100)\n@constraint(model, c2, 7x + 12y >= 120)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Call print to display the model:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"print(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"To solve the optimization problem, call the optimize! function:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"optimize!(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"info: Info\nThe ! after optimize is part of the name. It's nothing special. Julia has a convention that functions which mutate their arguments should end in !. A common example is push!.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Now let's see what information we can query about the solution.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"termination_status tells us why the solver stopped:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"termination_status(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"In this case, the solver found an optimal solution.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Check primal_status to see if the solver found a primal feasible point:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"primal_status(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"and dual_status to see if the solver found a dual feasible point:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"dual_status(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Now we know that our solver found an optimal solution, and that it has a primal and a dual solution to query.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Query the objective value using objective_value:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"objective_value(model)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"the primal solution using value:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"value(x)\nvalue(y)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"and the dual solution using shadow_price:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"shadow_price(c1)\nshadow_price(c2)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"That's it for our simple model. In the rest of this tutorial, we expand on some of the basic JuMP operations.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Model-basics","page":"Getting started with JuMP","title":"Model basics","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create a model by passing an optimizer:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = Model(HiGHS.Optimizer)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Alternatively, call set_optimizer at any point before calling optimize!:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = Model()\nset_optimizer(model, HiGHS.Optimizer)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"For some solvers, you can also use direct_model, which offers a more efficient connection to the underlying solver:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = direct_model(HiGHS.Optimizer())","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"warning: Warning\nSome solvers do not support direct_model!","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Solver-Options","page":"Getting started with JuMP","title":"Solver Options","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Pass options to solvers with optimizer_with_attributes:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model =\n    Model(optimizer_with_attributes(HiGHS.Optimizer, \"output_flag\" => false))","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"note: Note\nThese options are solver-specific. To find out the various options available, see the GitHub README of the individual solver packages. The link to each solver's GitHub page is in the Supported solvers table.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"You can also pass options with set_attribute:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = Model(HiGHS.Optimizer)\nset_attribute(model, \"output_flag\", false)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Solution-basics","page":"Getting started with JuMP","title":"Solution basics","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We saw above how to use termination_status and primal_status to understand the solution returned by the solver.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"However, only query solution attributes like value and objective_value if there is an available solution. Here's a recommended way to check:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"function solve_infeasible()\n    model = Model(HiGHS.Optimizer)\n    @variable(model, 0 <= x <= 1)\n    @variable(model, 0 <= y <= 1)\n    @constraint(model, x + y >= 3)\n    @objective(model, Max, x + 2y)\n    optimize!(model)\n    if termination_status(model) != OPTIMAL\n        @warn(\"The model was not solved correctly.\")\n        return\n    end\n    return value(x), value(y)\nend\n\nsolve_infeasible()","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Variable-basics","page":"Getting started with JuMP","title":"Variable basics","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Let's create a new empty model to explain some of the variable syntax:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = Model()","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Variable-bounds","page":"Getting started with JuMP","title":"Variable bounds","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"All of the variables we have created till now have had a bound. We can also create a free variable.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, free_x)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"While creating a variable, instead of using the <= and >= syntax, we can also use the lower_bound and upper_bound keyword arguments.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, keyword_x, lower_bound = 1, upper_bound = 2)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We can query whether a variable has a bound using the has_lower_bound and has_upper_bound functions. The values of the bound can be obtained using the lower_bound and upper_bound functions.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"has_upper_bound(keyword_x)\nupper_bound(keyword_x)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Note querying the value of a bound that does not exist will result in an error.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"try                         #hide\n    lower_bound(free_x)\ncatch err                   #hide\n    showerror(stderr, err)  #hide\nend                         #hide","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#tutorial_variable_container","page":"Getting started with JuMP","title":"Containers","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We have already seen how to add a single variable to a model using the @variable macro. Now let's look at ways to add multiple variables to a model.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"JuMP provides data structures for adding collections of variables to a model. These data structures are referred to as containers and are of three types: Arrays, DenseAxisArrays, and SparseAxisArrays.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Arrays","page":"Getting started with JuMP","title":"Arrays","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"JuMP arrays are created when you have integer indices that start at 1:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, a[1:2, 1:2])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Index elements in a as follows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"a[1, 1]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"a[2, :]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create an n-dimensional variable x in R^n with bounds l le x le u (l u in R^n) as follows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"n = 10\nl = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10];\nu = [10, 11, 12, 13, 14, 15, 16, 17, 18, 19];\n\n@variable(model, l[i] <= x[i = 1:n] <= u[i])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We can also create variable bounds that depend upon the indices:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, y[i = 1:2, j = 1:2] >= 2i + j)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#DenseAxisArrays","page":"Getting started with JuMP","title":"DenseAxisArrays","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"DenseAxisArrays are used when the indices are not one-based integer ranges. The syntax is similar except with an arbitrary vector as an index as opposed to a one-based range:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, z[i = 2:3, j = 1:2:3] >= 0)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Indices do not have to be integers. They can be any Julia type:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, w[1:5, [\"red\", \"blue\"]] <= 1)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Index elements in a DenseAxisArray as follows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"z[2, 1]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"w[2:3, [\"red\", \"blue\"]]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"See Forcing the container type for more details.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#SparseAxisArrays","page":"Getting started with JuMP","title":"SparseAxisArrays","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"SparseAxisArrays are created when the indices do not form a Cartesian product. For example, this applies when indices have a dependence upon previous indices (called triangular indexing):","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, u[i = 1:2, j = i:3])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We can also conditionally create variables by adding a comparison check that depends upon the named indices and is separated from the indices by a semi-colon ;:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, v[i = 1:9; mod(i, 3) == 0])","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Index elements in a DenseAxisArray as follows:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"u[1, 2]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"v[[3, 6]]","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Integrality","page":"Getting started with JuMP","title":"Integrality","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"JuMP can create binary and integer variables. Binary variables are constrained to the set  0 1, and integer variables are constrained to the set mathbbZ.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Integer-variables","page":"Getting started with JuMP","title":"Integer variables","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create an integer variable by passing  Int:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, integer_x, Int)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"or setting the integer keyword to true:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, integer_z, integer = true)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Binary-variables","page":"Getting started with JuMP","title":"Binary variables","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create a binary variable by passing Bin:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, binary_x, Bin)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"or setting the binary keyword to true:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@variable(model, binary_z, binary = true)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Constraint-basics","page":"Getting started with JuMP","title":"Constraint basics","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We'll need a new model to explain some of the constraint  basics:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = Model()\n@variable(model, x)\n@variable(model, y)\n@variable(model, z[1:10]);\nnothing #hide","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#tutorial_constraint_container","page":"Getting started with JuMP","title":"Containers","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Just as we had containers for variables, JuMP also provides Arrays, DenseAxisArrays, and SparseAxisArrays for storing collections of constraints. Examples for each container type are given below.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Arrays-2","page":"Getting started with JuMP","title":"Arrays","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create an Array of constraints:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@constraint(model, [i = 1:3], i * x <= i + 1)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#DenseAxisArrays-2","page":"Getting started with JuMP","title":"DenseAxisArrays","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create an DenseAxisArray of constraints:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@constraint(model, [i = 1:2, j = 2:3], i * x <= j + 1)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#SparseAxisArrays-2","page":"Getting started with JuMP","title":"SparseAxisArrays","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create an SparseAxisArray of constraints:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@constraint(model, [i = 1:2, j = 1:2; i != j], i * x <= j + 1)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Constraints-in-a-loop","page":"Getting started with JuMP","title":"Constraints in a loop","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We can add constraints using regular Julia loops:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"for i in 1:3\n    @constraint(model, 6x + 4y >= 5i)\nend","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"or use for each loops inside the @constraint macro:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@constraint(model, [i in 1:3], 6x + 4y >= 5i)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We can also create constraints such as sum _i = 1^10 z_i leq 1:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@constraint(model, sum(z[i] for i in 1:10) <= 1)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Objective-functions","page":"Getting started with JuMP","title":"Objective functions","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Set an objective function with @objective:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"model = Model(HiGHS.Optimizer)\n@variable(model, x >= 0)\n@variable(model, y >= 0)\n@objective(model, Min, 2x + y)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"Create a maximization objective using Max:","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"@objective(model, Max, 2x + y)","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"tip: Tip\nCalling @objective multiple times will over-write the previous objective. This can be useful when you want to solve the same problem with different objectives.","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/#Vectorized-syntax","page":"Getting started with JuMP","title":"Vectorized syntax","text":"","category":"section"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"We can also add constraints and an objective to JuMP using vectorized linear algebra. We'll illustrate this by solving an LP in standard form that is,","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"beginaligned\n min  c^T x \n textst  A x = b \n  x ge 0\nendaligned","category":"page"},{"location":"tutorials/getting_started/getting_started_with_JuMP/","page":"Getting started with JuMP","title":"Getting started with JuMP","text":"vector_model = Model(HiGHS.Optimizer)\nA = [1 1 9 5; 3 5 0 8; 2 0 6 13]\nb = [7, 3, 5]\nc = [1, 3, 5, 2]\n@variable(vector_model, x[1:4] >= 0)\n@constraint(vector_model, A * x .== b)\n@objective(vector_model, Min, c' * x)\noptimize!(vector_model)\nobjective_value(vector_model)","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"CurrentModule = JuMP\nDocTestSetup = quote\n    using JuMP\n    import HiGHS\nend\nDocTestFilters = [r\"≤|<=\", r\"≥|>=\", r\" == | = \", r\" ∈ | in \", r\"MathOptInterface|MOI\"]","category":"page"},{"location":"manual/complex/#Complex-number-support","page":"Complex number support","title":"Complex number support","text":"","category":"section"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"This page explains the complex-valued variables and constraints that JuMP supports. For a worked-example using these features, read the Quantum state discrimination tutorial.","category":"page"},{"location":"manual/complex/#Complex-valued-variables","page":"Complex number support","title":"Complex-valued variables","text":"","category":"section"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"Create a complex-valued variable using ComplexPlane:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model();\n\njulia> @variable(model, x in ComplexPlane())\nreal(x) + imag(x) im","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"Note that x is not a VariableRef; instead, it is an affine expression with Complex{Float64}-valued coefficients:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> typeof(x)\nGenericAffExpr{ComplexF64, VariableRef}","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"Behind the scenes, JuMP has created two real-valued variables, with names \"real(x)\" and \"imag(x)\":","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> all_variables(model)\n2-element Vector{VariableRef}:\n real(x)\n imag(x)\n\njulia> name.(all_variables(model))\n2-element Vector{String}:\n \"real(x)\"\n \"imag(x)\"","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"Use the real and imag functions on x to return a real-valued affine expression representing each variable:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> typeof(real(x))\nAffExpr (alias for GenericAffExpr{Float64, GenericVariableRef{Float64}})\n\njulia> typeof(imag(x))\nAffExpr (alias for GenericAffExpr{Float64, GenericVariableRef{Float64}})","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"To create an anonymous variable, use the set keyword argument:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model();\n\njulia> x = @variable(model, set = ComplexPlane())\n_[1] + _[2] im","category":"page"},{"location":"manual/complex/#Complex-valued-variable-bounds","page":"Complex number support","title":"Complex-valued variable bounds","text":"","category":"section"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"Because complex-valued variables lack a total ordering, the definition of a variable bound for a complex-valued variable is ambiguous. If you pass a real- or complex-valued argument to keywords such as lower_bound, upper_bound, and start_value, JuMP will apply the real and imaginary parts to the associated real-valued variables.","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model();\n\njulia> @variable(\n           model,\n           x in ComplexPlane(),\n           lower_bound = 1.0,\n           upper_bound = 2.0 + 3.0im,\n           start = 4im,\n       )\nreal(x) + imag(x) im\n\njulia> vars = all_variables(model)\n2-element Vector{VariableRef}:\n real(x)\n imag(x)\n\njulia> lower_bound.(vars)\n2-element Vector{Float64}:\n 1.0\n 0.0\n\njulia> upper_bound.(vars)\n2-element Vector{Float64}:\n 2.0\n 3.0\n\njulia> start_value.(vars)\n2-element Vector{Float64}:\n 0.0\n 4.0","category":"page"},{"location":"manual/complex/#Complex-valued-equality-constraints","page":"Complex number support","title":"Complex-valued equality constraints","text":"","category":"section"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"JuMP reformulates complex-valued equality constraints into two real-valued constraints: one representing the real part, and one representing the imaginary part. Thus, complex-valued equality constraints can be solved any solver that supports the real-valued constraint type.","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"For example:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x[1:2]);\n\njulia> @constraint(model, (1 + 2im) * x[1] + 3 * x[2] == 4 + 5im)\n(1 + 2im) x[1] + 3 x[2] = (4 + 5im)\n\njulia> optimize!(model)\n\njulia> value.(x)\n2-element Vector{Float64}:\n 2.5\n 0.5","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"is equivalent to","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x[1:2]);\n\njulia> @constraint(model, 1 * x[1] + 3 * x[2] == 4)  # real component\nx[1] + 3 x[2] = 4\n\njulia> @constraint(model, 2 * x[1] == 5)  # imag component\n2 x[1] = 5\n\njulia> optimize!(model)\n\njulia> value.(x)\n2-element Vector{Float64}:\n 2.5\n 0.5","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"This also applies if the variables are complex-valued:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x in ComplexPlane());\n\njulia> @constraint(model, (1 + 2im) * x + 3 * x == 4 + 5im)\n(4 + 2im) real(x) + (-2 + 4im) imag(x) = (4 + 5im)\n\njulia> optimize!(model)\n\njulia> value(x)\n1.3 + 0.6000000000000001im","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"which is equivalent to","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model(HiGHS.Optimizer);\n\njulia> set_silent(model)\n\njulia> @variable(model, x_real);\n\njulia> @variable(model, x_imag);\n\njulia> @constraint(model, x_real - 2 * x_imag + 3 * x_real == 4)\n4 x_real - 2 x_imag = 4\n\njulia> @constraint(model, x_imag + 2 * x_real + 3 * x_imag == 5)\n2 x_real + 4 x_imag = 5\n\njulia> optimize!(model)\n\njulia> value(x_real) + value(x_imag) * im\n1.3 + 0.6000000000000001im","category":"page"},{"location":"manual/complex/#Hermitian-PSD-Cones","page":"Complex number support","title":"Hermitian PSD Cones","text":"","category":"section"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"JuMP supports creating matrices where are Hermitian.","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model();\n\njulia> @variable(model, H[1:3, 1:3] in HermitianPSDCone())\n3×3 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n real(H[1,1])                    …  real(H[1,3]) + imag(H[1,3]) im\n real(H[1,2]) - imag(H[1,2]) im     real(H[2,3]) + imag(H[2,3]) im\n real(H[1,3]) - imag(H[1,3]) im     real(H[3,3])","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"Behind the scenes, JuMP has created nine real-valued decision variables:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> all_variables(model)\n9-element Vector{VariableRef}:\n real(H[1,1])\n real(H[1,2])\n real(H[2,2])\n real(H[1,3])\n real(H[2,3])\n real(H[3,3])\n imag(H[1,2])\n imag(H[1,3])\n imag(H[2,3])","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"and a Vector{VariableRef}-in-MOI.HermitianPositiveSemidefiniteConeTriangle constraint:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> num_constraints(model, Vector{VariableRef}, MOI.HermitianPositiveSemidefiniteConeTriangle)\n1","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"The MOI.HermitianPositiveSemidefiniteConeTriangle set can be efficiently bridged to MOI.PositiveSemidefiniteConeTriangle, so it can be solved by any solver that supports PSD constraints.","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"Each element of H is an affine expression with Complex{Float64}-valued coefficients:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> typeof(H[1, 1])\nGenericAffExpr{ComplexF64, VariableRef}\n\njulia> typeof(H[2, 1])\nGenericAffExpr{ComplexF64, VariableRef}","category":"page"},{"location":"manual/complex/#Hermitian-PSD-constraints","page":"Complex number support","title":"Hermitian PSD constraints","text":"","category":"section"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"The HermitianPSDCone can also be used in the @constraint macro:","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"julia> model = Model();\n\njulia> @variable(model, x[1:2])\n2-element Vector{VariableRef}:\n x[1]\n x[2]\n\njulia> import LinearAlgebra\n\njulia> H = LinearAlgebra.Hermitian([x[1] 1im; -1im -x[2]])\n2×2 LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}:\n x[1]  im\n -im   -x[2]\n\njulia> @constraint(model, H in HermitianPSDCone())\n[x[1]  im;\n -im   -x[2]] ∈ HermitianPSDCone()","category":"page"},{"location":"manual/complex/","page":"Complex number support","title":"Complex number support","text":"note: Note\nThe matrix H in H in HermitianPSDCone() must be a LinearAlgebra.Hermitian matrix type. A build_constraint error will be thrown if the matrix is a different matrix type.","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"EditURL = \"user_defined_hessians.jl\"","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/#User-defined-Hessians","page":"User-defined Hessians","title":"User-defined Hessians","text":"","category":"section"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"This tutorial was generated using Literate.jl. Download the source as a .jl file.","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"In this tutorial, we explain how to write a user-defined operator (see User-defined operators) with a Hessian matrix explicitly provided by the user.","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"For a more advanced example, see Nested optimization problems.","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"This tutorial uses the following packages:","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"using JuMP\nimport Ipopt","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/#Rosenbrock-example","page":"User-defined Hessians","title":"Rosenbrock example","text":"","category":"section"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"As a simple example, we consider the Rosenbrock function:","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"rosenbrock(x...) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"which has the gradient vector:","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"function ∇rosenbrock(g::AbstractVector, x...)\n    g[1] = 400 * x[1]^3 - 400 * x[1] * x[2] + 2 * x[1] - 2\n    g[2] = 200 * (x[2] - x[1]^2)\n    return\nend","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"and the Hessian matrix:","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"function ∇²rosenbrock(H::AbstractMatrix, x...)\n    H[1, 1] = 1200 * x[1]^2 - 400 * x[2] + 2\n    # H[1, 2] = -400 * x[1] <-- not needed because Hessian is symmetric\n    H[2, 1] = -400 * x[1]\n    H[2, 2] = 200.0\n    return\nend","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"You may assume the Hessian matrix H is initialized with zeros, and because it is symmetric you need only to fill in the non-zero of the lower-triangular terms.","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"The matrix type passed in as H depends on the automatic differentiation system, so make sure the first argument to the Hessian function supports an AbstractMatrix (it may be something other than Matrix{Float64}). However, you may assume only that H supports size(H) and setindex!.","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"Now that we have the function, its gradient, and its Hessian, we can construct a JuMP model, add the operator, and use it in a macro:","category":"page"},{"location":"tutorials/nonlinear/user_defined_hessians/","page":"User-defined Hessians","title":"User-defined Hessians","text":"model = Model(Ipopt.Optimizer)\n@variable(model, x[1:2])\n@operator(model, op_rosenbrock, 2, rosenbrock, ∇rosenbrock, ∇²rosenbrock)\n@objective(model, Min, op_rosenbrock(x[1], x[2]))\noptimize!(model)\nsolution_summary(model; verbose = true)","category":"page"}]
+}
diff --git a/previews/PR3525/should_i_use/index.html b/previews/PR3525/should_i_use/index.html
new file mode 100644
index 00000000000..4b05b5b86de
--- /dev/null
+++ b/previews/PR3525/should_i_use/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Should you use JuMP? · JuMP</title><meta name="title" content="Should you use JuMP? · JuMP"/><meta property="og:title" content="Should you use JuMP? · JuMP"/><meta property="twitter:title" content="Should you use JuMP? · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL=".."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../assets/documenter.js"></script><script src="../search_index.js"></script><script src="../siteinfo.js"></script><script src="../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../assets/themeswap.js"></script><link href="../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../"><img class="docs-light-only" src="../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox" checked/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../">Introduction</a></li><li class="is-active"><a class="tocitem" href>Should you use JuMP?</a><ul class="internal"><li><a class="tocitem" href="#When-should-you-use-JuMP?"><span>When should you use JuMP?</span></a></li><li><a class="tocitem" href="#When-should-you-not-use-JuMP?"><span>When should you not use JuMP?</span></a></li></ul></li><li><a class="tocitem" href="../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox"/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../tutorials/getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../tutorials/getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../tutorials/getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../tutorials/getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../tutorials/linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../tutorials/linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../tutorials/linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../tutorials/linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tutorials/linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../tutorials/linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../tutorials/linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../tutorials/linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../tutorials/linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../tutorials/linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../tutorials/linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../tutorials/linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../tutorials/linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../tutorials/linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../tutorials/linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../tutorials/linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../tutorials/nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../tutorials/nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../tutorials/nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tutorials/nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../tutorials/nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../tutorials/nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../tutorials/nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../tutorials/conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tutorials/conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../tutorials/conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../tutorials/conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../tutorials/conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../tutorials/conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../tutorials/conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../tutorials/conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../tutorials/conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../tutorials/conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../tutorials/algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../tutorials/algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../tutorials/algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../tutorials/applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../tutorials/applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../tutorials/applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../tutorials/applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../manual/models/">Models</a></li><li><a class="tocitem" href="../manual/variables/">Variables</a></li><li><a class="tocitem" href="../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../manual/containers/">Containers</a></li><li><a class="tocitem" href="../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/">Introduction</a></li><li><a class="tocitem" href="../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Introduction</a></li><li class="is-active"><a href>Should you use JuMP?</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Should you use JuMP?</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/should_i_use.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Should-you-use-JuMP?"><a class="docs-heading-anchor" href="#Should-you-use-JuMP?">Should you use JuMP?</a><a id="Should-you-use-JuMP?-1"></a><a class="docs-heading-anchor-permalink" href="#Should-you-use-JuMP?" title="Permalink"></a></h1><p>JuMP is an <a href="../background/algebraic_modeling_languages/#algebraic-modeling-language">algebraic modeling language</a> for mathematical optimization written in the <a href="https://julialang.org">Julia language</a>.</p><p>This page explains when you should consider using JuMP, and importantly, when you should <em>not</em> use JuMP.</p><h2 id="When-should-you-use-JuMP?"><a class="docs-heading-anchor" href="#When-should-you-use-JuMP?">When should you use JuMP?</a><a id="When-should-you-use-JuMP?-1"></a><a class="docs-heading-anchor-permalink" href="#When-should-you-use-JuMP?" title="Permalink"></a></h2><p>You should use JuMP if you have a constrained optimization problem for which you can formulate using the language of mathematical programming, that is:</p><ul><li>a set of decision variables</li><li>a scalar- or vector-valued objective function</li><li>a set of constraints.</li></ul><p>Key reasons to use JuMP include:</p><ul><li>User friendliness<ul><li>JuMP has syntax that mimics natural mathematical expressions. (See the section on <a href="../background/algebraic_modeling_languages/#algebraic-modeling-language">algebraic modeling languages</a>.)</li></ul></li><li>Solver independence<ul><li>JuMP uses a generic solver-independent interface provided by the <a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a> package, making it easy to change between a number of open-source and commercial optimization software packages (&quot;solvers&quot;). The <a href="../installation/#Supported-solvers">Supported solvers</a> section contains a table of the currently supported solvers.</li></ul></li><li>Ease of embedding<ul><li>JuMP itself is written purely in Julia. Solvers are the only binary dependencies.</li><li>JuMP provides automatic installation of many open-source solvers. This is different to modeling languages in Python which require you to download and install a solver yourself.</li><li>Because it is embedded in a general-purpose programming language, JuMP makes it easy to solve optimization problems as part of a larger workflow, for example, inside a simulation, behind a web server, or as a subproblem in a decomposition algorithm. As a trade-off, JuMP&#39;s syntax is constrained by the syntax and functionality available in Julia.</li><li>JuMP is <a href="https://www.mozilla.org/MPL/2.0/">MPL</a> licensed, meaning that it can be embedded in commercial software that complies with the terms of the license.</li></ul></li><li>Speed<ul><li>Benchmarking has shown that JuMP can create problems at similar speeds to special-purpose modeling languages such as <a href="https://ampl.com/">AMPL</a>.</li><li>JuMP communicates with most solvers in memory, avoiding the need to write intermediary files.</li></ul></li><li>Access to advanced algorithmic techniques<ul><li>JuMP supports efficient <em>in-memory</em> re-solves of linear programs, which previously required using solver-specific or low-level C++ libraries.</li><li>JuMP provides access to solver-independent and solver-dependent <a href="../manual/callbacks/#callbacks_manual">Callbacks</a>.</li></ul></li></ul><h2 id="When-should-you-not-use-JuMP?"><a class="docs-heading-anchor" href="#When-should-you-not-use-JuMP?">When should you not use JuMP?</a><a id="When-should-you-not-use-JuMP?-1"></a><a class="docs-heading-anchor-permalink" href="#When-should-you-not-use-JuMP?" title="Permalink"></a></h2><p>JuMP supports a broad range of optimization classes. However, there are still some that it doesn&#39;t support, or that are better supported by other software packages.</p><h3 id="You-want-to-optimize-a-complicated-Julia-function"><a class="docs-heading-anchor" href="#You-want-to-optimize-a-complicated-Julia-function">You want to optimize a complicated Julia function</a><a id="You-want-to-optimize-a-complicated-Julia-function-1"></a><a class="docs-heading-anchor-permalink" href="#You-want-to-optimize-a-complicated-Julia-function" title="Permalink"></a></h3><p>Packages in Julia compose well. It&#39;s common for people to pick two unrelated packages and use them in conjunction to create novel behavior. JuMP isn&#39;t one of those packages.</p><p>If you want to optimize an ordinary differential equation from <a href="https://github.com/SciML/DifferentialEquations.jl">DifferentialEquations.jl</a> or tune a neural network from <a href="https://github.com/FluxML/Flux.jl">Flux.jl</a>, consider using other packages such as:</p><ul><li><a href="https://github.com/JuliaNLSolvers/Optim.jl">Optim.jl</a></li><li><a href="https://github.com/SciML/Optimization.jl">Optimization.jl</a></li><li><a href="https://github.com/JuliaSmoothOptimizers/NLPModels.jl">NLPModels.jl</a></li><li><a href="https://github.com/JuliaNonconvex/Nonconvex.jl">Nonconvex.jl</a></li></ul><h3 id="Black-box,-derivative-free,-or-unconstrained-optimization"><a class="docs-heading-anchor" href="#Black-box,-derivative-free,-or-unconstrained-optimization">Black-box, derivative free, or unconstrained optimization</a><a id="Black-box,-derivative-free,-or-unconstrained-optimization-1"></a><a class="docs-heading-anchor-permalink" href="#Black-box,-derivative-free,-or-unconstrained-optimization" title="Permalink"></a></h3><p>JuMP does support nonlinear programs with constraints and objectives containing user-defined operators. However, the functions must be automatically differentiable, or need to provide explicit derivatives. (See <a href="../manual/nonlinear/#jump_user_defined_operators">User-defined operators</a> for more information.)</p><p>If your function is a black-box that is non-differentiable (for example, it is the output of a simulation written in C++), JuMP is not the right tool for the job. This also applies if you want to use a derivative free method.</p><p>Even if your problem is differentiable, if it is unconstrained there is limited benefit (and downsides in the form of more overhead) to using JuMP over tools which are only concerned with function minimization.</p><p>Alternatives to consider are:</p><ul><li><a href="https://github.com/JuliaNLSolvers/Optim.jl">Optim.jl</a></li><li><a href="https://github.com/SciML/Optimization.jl">Optimization.jl</a></li><li><a href="https://github.com/JuliaOpt/NLopt.jl">NLopt.jl</a></li></ul><h3 id="Optimal-control-problems"><a class="docs-heading-anchor" href="#Optimal-control-problems">Optimal control problems</a><a id="Optimal-control-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Optimal-control-problems" title="Permalink"></a></h3><p>JuMP supports formulating optimal control problems as large nonlinear programs (see, for example, <a href="../tutorials/nonlinear/space_shuttle_reentry_trajectory/#Optimal-control-for-a-Space-Shuttle-reentry-trajectory">Optimal control for a Space Shuttle reentry trajectory</a>). However, the nonlinear interface has a number of limitations (for example, the need to write out the dynamics in algebraic form) that mean JuMP might not be the right tool for the job.</p><p>Alternatives to consider are:</p><ul><li><a href="https://web.casadi.org">CasADi [MATLAB/Python]</a>, <a href="https://github.com/ichatzinikolaidis/CasADi.jl">CasADi.jl</a></li><li><a href="https://github.com/infiniteopt/InfiniteOpt.jl">InfiniteOpt.jl</a></li><li><a href="https://pyomo.readthedocs.io/en/stable/modeling_extensions/dae.html">pyomo.DAE [Python]</a></li></ul><h3 id="Disciplined-convex-programming"><a class="docs-heading-anchor" href="#Disciplined-convex-programming">Disciplined convex programming</a><a id="Disciplined-convex-programming-1"></a><a class="docs-heading-anchor-permalink" href="#Disciplined-convex-programming" title="Permalink"></a></h3><p>JuMP does not support <a href="https://dcp.stanford.edu">disciplined convex programming (DCP)</a>.</p><p>Alternatives to consider are:</p><ul><li><a href="https://github.com/jump-dev/Convex.jl">Convex.jl</a></li><li><a href="https://github.com/cvxpy/cvxpy">CVXPY [Python]</a></li><li><a href="https://yalmip.github.io">YALMIP [MATLAB]</a></li></ul><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p><code>Convex.jl</code> is also built on MathOptInterface, and shares the same set of underlying solvers. However, you input problems differently, and Convex.jl checks that the problem is DCP.</p></div></div><h3 id="Stochastic-programming"><a class="docs-heading-anchor" href="#Stochastic-programming">Stochastic programming</a><a id="Stochastic-programming-1"></a><a class="docs-heading-anchor-permalink" href="#Stochastic-programming" title="Permalink"></a></h3><p>JuMP requires deterministic input data.</p><p>If you have stochastic input data, consider using a JuMP extension such as:</p><ul><li><a href="https://github.com/infiniteopt/InfiniteOpt.jl">InfiniteOpt.jl</a></li><li><a href="https://github.com/martinbiel/StochasticPrograms.jl">StochasticPrograms.jl</a></li><li><a href="https://github.com/odow/SDDP.jl">SDDP.jl</a></li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../">« Introduction</a><a class="docs-footer-nextpage" href="../installation/">Installation Guide »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/siteinfo.js b/previews/PR3525/siteinfo.js
new file mode 100644
index 00000000000..532c214d768
--- /dev/null
+++ b/previews/PR3525/siteinfo.js
@@ -0,0 +1 @@
+var DOCUMENTER_CURRENT_VERSION = "previews/PR3525";
diff --git a/previews/PR3525/tutorials/algorithms/benders_decomposition.jl b/previews/PR3525/tutorials/algorithms/benders_decomposition.jl
new file mode 100644
index 00000000000..376a4fd19ff
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/benders_decomposition.jl
@@ -0,0 +1,379 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # [Benders decomposition](@id benders_decomposition_classical)
+
+# **This tutorial was originally contributed by Shuvomoy Das Gupta.**
+
+# This tutorial describes how to implement [Benders decomposition](https://en.wikipedia.org/wiki/Benders_decomposition)
+# in JuMP. It uses the following packages:
+
+using JuMP
+import GLPK
+import Printf
+import Test  #src
+
+# ## Theory
+
+# Benders decomposition is a useful algorithm for solving convex optimization
+# problems with a large number of variables. It works best when a larger problem
+# can be decomposed into two (or more) smaller problems that are individually
+# much easier to solve. This tutorial demonstrates Benders decomposition on the
+# following mixed-integer linear program:
+# ```math
+# \begin{aligned}
+# \text{min}        \ & c_1^\top x+c_2^\top y   \\
+# \text{subject to} \ & A_1 x+ A_2 y \le b      \\
+#                     & x \ge 0                 \\
+#                     & y \ge 0                 \\
+#                     & x \in \mathbb{Z}^n
+# \end{aligned}
+# ```
+# where $b \in \mathbb{R}^m$, $A_1 \in \mathbb{R}^{m \times n}$,
+# $A_2 \in \mathbb{R}^{m \times p}$ and $\mathbb{Z}$ is the set of integers.
+#
+# Any mixed integer programming problem can be written in the form above.
+
+# If there are relatively few integer variables, and many more continuous
+# variables, then it may be beneficial to decompose the problem into a small
+# problem containing only integer variables and a linear program containing
+# only continuous variables. Hopefully, the linear program will be much easier
+# to solve in isolation than in the full mixed-integer linear program.
+
+# For example, if we knew a feasible solution for ``x``, we could obtain a
+# solution for ``y`` by solving:
+# ```math
+# \begin{aligned}
+# V_2(x) = & \text{min}          \ & c_2^\top y                        \\
+#          & \text{subject to}   \ & A_2 y \le b - A_1 x & \quad [\pi] \\
+#          &                       & y \ge 0,
+# \end{aligned}
+# ```
+# where ``\pi`` is the dual variable associated with the constraints. Because
+# this is a linear program, it is easy to solve.
+
+# Replacing the ``c_2^\top y`` component of the objective in our original
+# problem with ``V_2`` yields:
+# ```math
+# \begin{aligned}
+# \text{min}        \ & c_1^\top x + V_2(x) \\
+# \text{subject to} \ & x \ge 0             \\
+#                     & x \in \mathbb{Z}^n
+# \end{aligned}
+# ```
+# This problem looks a lot simpler to solve, but we need to do something else
+# with ``V_2`` first.
+
+# Because ``x`` is a constant that appears on the right-hand side of the
+# constraints, ``V_2`` is a convex function with respect to ``x``, and the dual
+# variable ``\pi`` can be multiplied by ``-A_1`` to obtain a subgradient of
+# ``V_2(x)`` with respect to ``x``. Therefore, if we have a candidate solution
+# ``x_k``, then we can solve ``V_2(x_k)`` and obtain a feasible dual vector
+# ``\pi_k``. Using these values, we can construct a first-order Taylor-series
+# approximation of ``V_2`` about the point ``x_k``:
+# ```math
+# V_2(x) \ge V_2(x_k) + -\pi_k^\top A_1 (x - x_k).
+# ```
+# By convexity, we know that this inequality holds for all ``x``, and we call
+# these inequalities _cuts_.
+
+# Benders decomposition is an iterative technique that replaces ``V_2(x)`` with
+# a new decision variable ``\theta``, and approximates it from below using cuts:
+# ```math
+# \begin{aligned}
+# V_1^K = & \text{min}         \ & c_1^\top x + \theta  \\
+#         &  \text{subject to} \ & x \ge 0              \\
+#         &                    \ & x \in \mathbb{Z}^n   \\
+#         &                    \ & \theta \ge M         \\
+#         &                    \ & \theta \ge V_2(x_k) + \pi_k^\top(x - x_k) & \quad \forall k = 1,\ldots,K.
+# \end{aligned}
+# ```
+# This integer program is called the _first-stage_ subproblem.
+
+# To generate cuts, we solve ``V_1^K`` to obtain a candidate first-stage
+# solution ``x_k``, then we use that solution to solve ``V_2(x_k)``. Then, using
+# the optimal objective value and dual solution from ``V_2``, we add a new cut
+# to form ``V_1^{K+1}`` and repeat.
+
+# ### Bounds
+
+# Due to convexity, we know that ``V_2(x) \ge \theta`` for all ``x``. Therefore,
+# the optimal objective value of ``V_1^K`` provides a valid _lower_ bound on the
+# objective value of the full problem. In addition, if we take a feasible
+# solution for ``x`` from the first-stage problem, then ``c_1^\top x + V_2(x)``
+# is a valid _upper_ bound on the objective value of the full problem.
+
+# Benders decomposition uses the lower and upper bounds to determine when it has
+# found the global optimal solution.
+
+# ## Input data
+
+# As an example for this tutorial, we use the input data is from page 139 of
+# Garfinkel, R. & Nemhauser, G. L. Integer programming. (Wiley, 1972).
+
+c_1 = [1, 4]
+c_2 = [2, 3]
+dim_x = length(c_1)
+dim_y = length(c_2)
+b = [-2; -3]
+A_1 = [1 -3; -1 -3]
+A_2 = [1 -2; -1 -1]
+M = -1000;
+
+# ## [Iterative method](@id benders_iterative)
+
+# !!! warning
+#     This is a basic implementation for pedagogical purposes. We haven't
+#     discussed Benders feasibility cuts, or any of the computational tricks
+#     that are required to build a performant implementation for large-scale
+#     problems. See [In-place iterative method](@ref) for one improvement that
+#     helps computation time.
+
+# We start by formulating the first-stage subproblem:
+
+model = Model(GLPK.Optimizer)
+@variable(model, x[1:dim_x] >= 0, Int)
+@variable(model, θ >= M)
+@objective(model, Min, c_1' * x + θ)
+print(model)
+
+# For the next step, we need a function that takes a first-stage candidate
+# solution `x` and returns the optimal solution from the second-stage
+# subproblem:
+
+function solve_subproblem(x)
+    model = Model(GLPK.Optimizer)
+    @variable(model, y[1:dim_y] >= 0)
+    con = @constraint(model, A_2 * y .<= b - A_1 * x)
+    @objective(model, Min, c_2' * y)
+    optimize!(model)
+    @assert termination_status(model) == OPTIMAL
+    return (obj = objective_value(model), y = value.(y), π = dual.(con))
+end
+
+# Note that `solve_subproblem` returns a `NamedTuple` of the objective value,
+# the optimal primal solution for `y`, and the optimal dual solution for `π`.
+
+# We're almost ready for our optimization loop, but first, here's a helpful
+# function for logging:
+
+function print_iteration(k, args...)
+    f(x) = Printf.@sprintf("%12.4e", x)
+    println(lpad(k, 9), " ", join(f.(args), " "))
+    return
+end
+
+# We also need to put a limit on the number of iterations before termination:
+
+MAXIMUM_ITERATIONS = 100
+
+# And a way to check if the lower and upper bounds are close-enough to
+# terminate:
+
+ABSOLUTE_OPTIMALITY_GAP = 1e-6
+
+# Now we're ready to iterate Benders decomposition:
+
+println("Iteration  Lower Bound  Upper Bound          Gap")
+for k in 1:MAXIMUM_ITERATIONS
+    optimize!(model)
+    lower_bound = objective_value(model)
+    x_k = value.(x)
+    ret = solve_subproblem(x_k)
+    upper_bound = c_1' * x_k + ret.obj
+    gap = (upper_bound - lower_bound) / upper_bound
+    print_iteration(k, lower_bound, upper_bound, gap)
+    if gap < ABSOLUTE_OPTIMALITY_GAP
+        println("Terminating with the optimal solution")
+        break
+    end
+    cut = @constraint(model, θ >= ret.obj + -ret.π' * A_1 * (x .- x_k))
+    @info "Adding the cut $(cut)"
+end
+
+# Finally, we can obtain the optimal solution
+
+optimize!(model)
+Test.@test value.(x) == [0.0, 1.0]  #src
+x_optimal = value.(x)
+
+#-
+
+optimal_ret = solve_subproblem(x_optimal)
+Test.@test optimal_ret.y == [0.0, 0.0]  #src
+y_optimal = optimal_ret.y
+
+# ## Callback method
+
+# The [Iterative method](@ref benders_iterative) section implemented Benders
+# decomposition using a loop. In each iteration, we re-solved the first-stage
+# subproblem to generate a candidate solution. However, modern MILP solvers such
+# as CPLEX, Gurobi, and GLPK provide lazy constraint callbacks which allow us to
+# add new cuts _while the solver is running_. This can be more efficient than an
+# iterative method because we can avoid repeating work such as solving the root
+# node of the first-stage MILP at each iteration.
+
+# !!! tip
+#     For more information on callbacks, read the page
+#     [Solver-independent callbacks](@ref callbacks_manual).
+
+# As before, we construct the same first-stage subproblem:
+
+lazy_model = Model(GLPK.Optimizer)
+@variable(lazy_model, x[1:dim_x] >= 0, Int)
+@variable(lazy_model, θ >= M)
+@objective(lazy_model, Min, θ)
+print(lazy_model)
+
+# What differs is that we write a callback function instead of a loop:
+
+k = 0
+
+"""
+    my_callback(cb_data)
+
+A callback that implements Benders decomposition. Note how similar it is to the
+inner loop of the iterative method.
+"""
+function my_callback(cb_data)
+    global k += 1
+    x_k = callback_value.(cb_data, x)
+    θ_k = callback_value(cb_data, θ)
+    lower_bound = c_1' * x_k + θ_k
+    ret = solve_subproblem(x_k)
+    upper_bound = c_1' * x_k + c_2' * ret.y
+    gap = (upper_bound - lower_bound) / upper_bound
+    print_iteration(k, lower_bound, upper_bound, gap)
+    if gap < ABSOLUTE_OPTIMALITY_GAP
+        println("Terminating with the optimal solution")
+        return
+    end
+    cut = @build_constraint(θ >= ret.obj + -ret.π' * A_1 * (x .- x_k))
+    MOI.submit(model, MOI.LazyConstraint(cb_data), cut)
+    return
+end
+
+set_attribute(lazy_model, MOI.LazyConstraintCallback(), my_callback)
+
+# Now when we optimize!, our callback is run:
+
+optimize!(lazy_model)
+
+# Note how this problem also takes 4 iterations to converge, but the sequence
+# of bounds is different compared to the iterative method.
+
+# Finally, we can obtain the optimal solution:
+
+Test.@test value.(x) == [0.0, 1.0]  #src
+x_optimal = value.(x)
+
+#-
+
+optimal_ret = solve_subproblem(x_optimal)
+Test.@test optimal_ret.y == [0.0, 0.0]  #src
+y_optimal = optimal_ret.y
+
+# ## In-place iterative method
+
+# Our implementation of the iterative method has a problem: every time we need
+# to solve the subproblem, we must rebuild it from scratch. This is expensive,
+# and it can be the bottleneck in the solution process. We can improve our
+# implementation by using re-using the subproblem between solves.
+
+# First, we create our first-stage problem as usual:
+
+model = Model(GLPK.Optimizer)
+@variable(model, x[1:dim_x] >= 0, Int)
+@variable(model, θ >= M)
+@objective(model, Min, c_1' * x + θ)
+print(model)
+
+# Then, instead of building the subproblem in a function, we build it once here:
+
+subproblem = Model(GLPK.Optimizer)
+@variable(subproblem, x_copy[1:dim_x])
+@variable(subproblem, y[1:dim_y] >= 0)
+@constraint(subproblem, A_1 * x_copy + A_2 * y .<= b)
+@objective(subproblem, Min, c_2' * y)
+print(subproblem)
+
+# This formulation is slightly different. We have included a copy of the x
+# variables, `x_copy`, and used `x_copy` in the left-hand side of the
+# constraints.
+
+# Our function to solve the subproblem is also slightly different. First, we
+# need to fix the value of the `x_copy` variables to the value of `x` from the
+# first-stage problem, and second, we compute the dual using the
+# [`reduced_cost`](@ref) of `x_copy`, not the dual of the linear constraints:
+
+function solve_subproblem(model, x)
+    fix.(model[:x_copy], x)
+    optimize!(model)
+    @assert termination_status(model) == OPTIMAL
+    return (
+        obj = objective_value(model),
+        y = value.(model[:y]),
+        π = reduced_cost.(model[:x_copy]),
+    )
+end
+
+# Now we're ready to iterate our in-place Benders decomposition, but this time
+# the cut computation is slightly different. Because we used
+# [`reduced_cost`](@ref) on the `x_copy` variables, the value of `π` is a valid
+# subgradient on the objective of `subproblem` with respect to `x`. Therefore,
+# we don't need to multiply the duals by `-A_1`, and so our cut uses `ret.π'`
+# instead of `-ret.π' * A_1`:
+
+println("Iteration  Lower Bound  Upper Bound          Gap")
+for k in 1:MAXIMUM_ITERATIONS
+    optimize!(model)
+    lower_bound = objective_value(model)
+    x_k = value.(x)
+    ret = solve_subproblem(subproblem, x_k)
+    upper_bound = c_1' * x_k + ret.obj
+    gap = (upper_bound - lower_bound) / upper_bound
+    print_iteration(k, lower_bound, upper_bound, gap)
+    if gap < ABSOLUTE_OPTIMALITY_GAP
+        println("Terminating with the optimal solution")
+        break
+    end
+    cut = @constraint(model, θ >= ret.obj + ret.π' * (x .- x_k))
+    @info "Adding the cut $(cut)"
+end
+
+# Finally, we can obtain the optimal solution:
+
+optimize!(model)
+Test.@test value.(x) == [0.0, 1.0]  #src
+x_optimal = value.(x)
+
+#-
+
+optimal_ret = solve_subproblem(subproblem, x_optimal)
+Test.@test optimal_ret.y == [0.0, 0.0]  #src
+y_optimal = optimal_ret.y
+
+# For larger problems, the benefit of re-using the same subproblem and not
+# needing to multiply the duals by `A_1` in the cut coefficient usually
+# outweights the cost of needing a copy of the `x` variables in the subproblem.
+
+# As a secondary benefit, because we no longer need an explicit representation
+# of `A_1` in the cut, we can build the `model` and `subproblem` formulations
+# using arbitrary JuMP syntax; they do not need to be in matrix form.
diff --git a/previews/PR3525/tutorials/algorithms/benders_decomposition/index.html b/previews/PR3525/tutorials/algorithms/benders_decomposition/index.html
new file mode 100644
index 00000000000..9f5b7a48b0b
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/benders_decomposition/index.html
@@ -0,0 +1,189 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Benders decomposition · JuMP</title><meta name="title" content="Benders decomposition · JuMP"/><meta property="og:title" content="Benders decomposition · JuMP"/><meta property="twitter:title" content="Benders decomposition · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox" checked/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Benders decomposition</a><ul class="internal"><li><a class="tocitem" href="#Theory"><span>Theory</span></a></li><li><a class="tocitem" href="#Input-data"><span>Input data</span></a></li><li><a class="tocitem" href="#benders_iterative"><span>Iterative method</span></a></li><li><a class="tocitem" href="#Callback-method"><span>Callback method</span></a></li><li><a class="tocitem" href="#In-place-iterative-method"><span>In-place iterative method</span></a></li></ul></li><li><a class="tocitem" href="../cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Algorithms</a></li><li class="is-active"><a href>Benders decomposition</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Benders decomposition</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/algorithms/benders_decomposition.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="benders_decomposition_classical"><a class="docs-heading-anchor" href="#benders_decomposition_classical">Benders decomposition</a><a id="benders_decomposition_classical-1"></a><a class="docs-heading-anchor-permalink" href="#benders_decomposition_classical" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../benders_decomposition.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Shuvomoy Das Gupta.</strong></p><p>This tutorial describes how to implement <a href="https://en.wikipedia.org/wiki/Benders_decomposition">Benders decomposition</a> in JuMP. It uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import GLPK
+import Printf</code></pre><h2 id="Theory"><a class="docs-heading-anchor" href="#Theory">Theory</a><a id="Theory-1"></a><a class="docs-heading-anchor-permalink" href="#Theory" title="Permalink"></a></h2><p>Benders decomposition is a useful algorithm for solving convex optimization problems with a large number of variables. It works best when a larger problem can be decomposed into two (or more) smaller problems that are individually much easier to solve. This tutorial demonstrates Benders decomposition on the following mixed-integer linear program:</p><p class="math-container">\[\begin{aligned}
+\text{min}        \ &amp; c_1^\top x+c_2^\top y   \\
+\text{subject to} \ &amp; A_1 x+ A_2 y \le b      \\
+                    &amp; x \ge 0                 \\
+                    &amp; y \ge 0                 \\
+                    &amp; x \in \mathbb{Z}^n
+\end{aligned}\]</p><p>where <span>$b \in \mathbb{R}^m$</span>, <span>$A_1 \in \mathbb{R}^{m \times n}$</span>, <span>$A_2 \in \mathbb{R}^{m \times p}$</span> and <span>$\mathbb{Z}$</span> is the set of integers.</p><p>Any mixed integer programming problem can be written in the form above.</p><p>If there are relatively few integer variables, and many more continuous variables, then it may be beneficial to decompose the problem into a small problem containing only integer variables and a linear program containing only continuous variables. Hopefully, the linear program will be much easier to solve in isolation than in the full mixed-integer linear program.</p><p>For example, if we knew a feasible solution for <span>$x$</span>, we could obtain a solution for <span>$y$</span> by solving:</p><p class="math-container">\[\begin{aligned}
+V_2(x) = &amp; \text{min}          \ &amp; c_2^\top y                        \\
+         &amp; \text{subject to}   \ &amp; A_2 y \le b - A_1 x &amp; \quad [\pi] \\
+         &amp;                       &amp; y \ge 0,
+\end{aligned}\]</p><p>where <span>$\pi$</span> is the dual variable associated with the constraints. Because this is a linear program, it is easy to solve.</p><p>Replacing the <span>$c_2^\top y$</span> component of the objective in our original problem with <span>$V_2$</span> yields:</p><p class="math-container">\[\begin{aligned}
+\text{min}        \ &amp; c_1^\top x + V_2(x) \\
+\text{subject to} \ &amp; x \ge 0             \\
+                    &amp; x \in \mathbb{Z}^n
+\end{aligned}\]</p><p>This problem looks a lot simpler to solve, but we need to do something else with <span>$V_2$</span> first.</p><p>Because <span>$x$</span> is a constant that appears on the right-hand side of the constraints, <span>$V_2$</span> is a convex function with respect to <span>$x$</span>, and the dual variable <span>$\pi$</span> can be multiplied by <span>$-A_1$</span> to obtain a subgradient of <span>$V_2(x)$</span> with respect to <span>$x$</span>. Therefore, if we have a candidate solution <span>$x_k$</span>, then we can solve <span>$V_2(x_k)$</span> and obtain a feasible dual vector <span>$\pi_k$</span>. Using these values, we can construct a first-order Taylor-series approximation of <span>$V_2$</span> about the point <span>$x_k$</span>:</p><p class="math-container">\[V_2(x) \ge V_2(x_k) + -\pi_k^\top A_1 (x - x_k).\]</p><p>By convexity, we know that this inequality holds for all <span>$x$</span>, and we call these inequalities <em>cuts</em>.</p><p>Benders decomposition is an iterative technique that replaces <span>$V_2(x)$</span> with a new decision variable <span>$\theta$</span>, and approximates it from below using cuts:</p><p class="math-container">\[\begin{aligned}
+V_1^K = &amp; \text{min}         \ &amp; c_1^\top x + \theta  \\
+        &amp;  \text{subject to} \ &amp; x \ge 0              \\
+        &amp;                    \ &amp; x \in \mathbb{Z}^n   \\
+        &amp;                    \ &amp; \theta \ge M         \\
+        &amp;                    \ &amp; \theta \ge V_2(x_k) + \pi_k^\top(x - x_k) &amp; \quad \forall k = 1,\ldots,K.
+\end{aligned}\]</p><p>This integer program is called the <em>first-stage</em> subproblem.</p><p>To generate cuts, we solve <span>$V_1^K$</span> to obtain a candidate first-stage solution <span>$x_k$</span>, then we use that solution to solve <span>$V_2(x_k)$</span>. Then, using the optimal objective value and dual solution from <span>$V_2$</span>, we add a new cut to form <span>$V_1^{K+1}$</span> and repeat.</p><h3 id="Bounds"><a class="docs-heading-anchor" href="#Bounds">Bounds</a><a id="Bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Bounds" title="Permalink"></a></h3><p>Due to convexity, we know that <span>$V_2(x) \ge \theta$</span> for all <span>$x$</span>. Therefore, the optimal objective value of <span>$V_1^K$</span> provides a valid <em>lower</em> bound on the objective value of the full problem. In addition, if we take a feasible solution for <span>$x$</span> from the first-stage problem, then <span>$c_1^\top x + V_2(x)$</span> is a valid <em>upper</em> bound on the objective value of the full problem.</p><p>Benders decomposition uses the lower and upper bounds to determine when it has found the global optimal solution.</p><h2 id="Input-data"><a class="docs-heading-anchor" href="#Input-data">Input data</a><a id="Input-data-1"></a><a class="docs-heading-anchor-permalink" href="#Input-data" title="Permalink"></a></h2><p>As an example for this tutorial, we use the input data is from page 139 of Garfinkel, R. &amp; Nemhauser, G. L. Integer programming. (Wiley, 1972).</p><pre><code class="language-julia hljs">c_1 = [1, 4]
+c_2 = [2, 3]
+dim_x = length(c_1)
+dim_y = length(c_2)
+b = [-2; -3]
+A_1 = [1 -3; -1 -3]
+A_2 = [1 -2; -1 -1]
+M = -1000;</code></pre><h2 id="benders_iterative"><a class="docs-heading-anchor" href="#benders_iterative">Iterative method</a><a id="benders_iterative-1"></a><a class="docs-heading-anchor-permalink" href="#benders_iterative" title="Permalink"></a></h2><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This is a basic implementation for pedagogical purposes. We haven&#39;t discussed Benders feasibility cuts, or any of the computational tricks that are required to build a performant implementation for large-scale problems. See <a href="#In-place-iterative-method">In-place iterative method</a> for one improvement that helps computation time.</p></div></div><p>We start by formulating the first-stage subproblem:</p><pre><code class="language-julia hljs">model = Model(GLPK.Optimizer)
+@variable(model, x[1:dim_x] &gt;= 0, Int)
+@variable(model, θ &gt;= M)
+@objective(model, Min, c_1&#39; * x + θ)
+print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min x[1] + 4 x[2] + θ
+Subject to
+ x[1] ≥ 0
+ x[2] ≥ 0
+ θ ≥ -1000
+ x[1] integer
+ x[2] integer</code></pre><p>For the next step, we need a function that takes a first-stage candidate solution <code>x</code> and returns the optimal solution from the second-stage subproblem:</p><pre><code class="language-julia hljs">function solve_subproblem(x)
+    model = Model(GLPK.Optimizer)
+    @variable(model, y[1:dim_y] &gt;= 0)
+    con = @constraint(model, A_2 * y .&lt;= b - A_1 * x)
+    @objective(model, Min, c_2&#39; * y)
+    optimize!(model)
+    @assert termination_status(model) == OPTIMAL
+    return (obj = objective_value(model), y = value.(y), π = dual.(con))
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_subproblem (generic function with 1 method)</code></pre><p>Note that <code>solve_subproblem</code> returns a <code>NamedTuple</code> of the objective value, the optimal primal solution for <code>y</code>, and the optimal dual solution for <code>π</code>.</p><p>We&#39;re almost ready for our optimization loop, but first, here&#39;s a helpful function for logging:</p><pre><code class="language-julia hljs">function print_iteration(k, args...)
+    f(x) = Printf.@sprintf(&quot;%12.4e&quot;, x)
+    println(lpad(k, 9), &quot; &quot;, join(f.(args), &quot; &quot;))
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">print_iteration (generic function with 1 method)</code></pre><p>We also need to put a limit on the number of iterations before termination:</p><pre><code class="language-julia hljs">MAXIMUM_ITERATIONS = 100</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">100</code></pre><p>And a way to check if the lower and upper bounds are close-enough to terminate:</p><pre><code class="language-julia hljs">ABSOLUTE_OPTIMALITY_GAP = 1e-6</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1.0e-6</code></pre><p>Now we&#39;re ready to iterate Benders decomposition:</p><pre><code class="language-julia hljs">println(&quot;Iteration  Lower Bound  Upper Bound          Gap&quot;)
+for k in 1:MAXIMUM_ITERATIONS
+    optimize!(model)
+    lower_bound = objective_value(model)
+    x_k = value.(x)
+    ret = solve_subproblem(x_k)
+    upper_bound = c_1&#39; * x_k + ret.obj
+    gap = (upper_bound - lower_bound) / upper_bound
+    print_iteration(k, lower_bound, upper_bound, gap)
+    if gap &lt; ABSOLUTE_OPTIMALITY_GAP
+        println(&quot;Terminating with the optimal solution&quot;)
+        break
+    end
+    cut = @constraint(model, θ &gt;= ret.obj + -ret.π&#39; * A_1 * (x .- x_k))
+    @info &quot;Adding the cut $(cut)&quot;
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Iteration  Lower Bound  Upper Bound          Gap
+        1  -1.0000e+03   7.6667e+00   1.3143e+02
+<span class="sgr36"><span class="sgr1">[ Info: </span></span>Adding the cut 2 x[1] + 8 x[2] + θ ≥ 7.666666666666666
+        2  -4.9600e+02   1.2630e+03   1.3927e+00
+<span class="sgr36"><span class="sgr1">[ Info: </span></span>Adding the cut -1.5 x[1] + 4.5 x[2] + θ ≥ 3
+        3  -1.0800e+02   8.8800e+02   1.1216e+00
+<span class="sgr36"><span class="sgr1">[ Info: </span></span>Adding the cut θ ≥ 0
+        4   4.0000e+00   4.0000e+00   0.0000e+00
+Terminating with the optimal solution</code></pre><p>Finally, we can obtain the optimal solution</p><pre><code class="language-julia hljs">optimize!(model)
+x_optimal = value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 0.0
+ 1.0</code></pre><pre><code class="language-julia hljs">optimal_ret = solve_subproblem(x_optimal)
+y_optimal = optimal_ret.y</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 0.0
+ 0.0</code></pre><h2 id="Callback-method"><a class="docs-heading-anchor" href="#Callback-method">Callback method</a><a id="Callback-method-1"></a><a class="docs-heading-anchor-permalink" href="#Callback-method" title="Permalink"></a></h2><p>The <a href="#benders_iterative">Iterative method</a> section implemented Benders decomposition using a loop. In each iteration, we re-solved the first-stage subproblem to generate a candidate solution. However, modern MILP solvers such as CPLEX, Gurobi, and GLPK provide lazy constraint callbacks which allow us to add new cuts <em>while the solver is running</em>. This can be more efficient than an iterative method because we can avoid repeating work such as solving the root node of the first-stage MILP at each iteration.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>For more information on callbacks, read the page <a href="../../../manual/callbacks/#callbacks_manual">Solver-independent callbacks</a>.</p></div></div><p>As before, we construct the same first-stage subproblem:</p><pre><code class="language-julia hljs">lazy_model = Model(GLPK.Optimizer)
+@variable(lazy_model, x[1:dim_x] &gt;= 0, Int)
+@variable(lazy_model, θ &gt;= M)
+@objective(lazy_model, Min, θ)
+print(lazy_model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min θ
+Subject to
+ x[1] ≥ 0
+ x[2] ≥ 0
+ θ ≥ -1000
+ x[1] integer
+ x[2] integer</code></pre><p>What differs is that we write a callback function instead of a loop:</p><pre><code class="language-julia hljs">k = 0
+
+&quot;&quot;&quot;
+    my_callback(cb_data)
+
+A callback that implements Benders decomposition. Note how similar it is to the
+inner loop of the iterative method.
+&quot;&quot;&quot;
+function my_callback(cb_data)
+    global k += 1
+    x_k = callback_value.(cb_data, x)
+    θ_k = callback_value(cb_data, θ)
+    lower_bound = c_1&#39; * x_k + θ_k
+    ret = solve_subproblem(x_k)
+    upper_bound = c_1&#39; * x_k + c_2&#39; * ret.y
+    gap = (upper_bound - lower_bound) / upper_bound
+    print_iteration(k, lower_bound, upper_bound, gap)
+    if gap &lt; ABSOLUTE_OPTIMALITY_GAP
+        println(&quot;Terminating with the optimal solution&quot;)
+        return
+    end
+    cut = @build_constraint(θ &gt;= ret.obj + -ret.π&#39; * A_1 * (x .- x_k))
+    MOI.submit(model, MOI.LazyConstraint(cb_data), cut)
+    return
+end
+
+set_attribute(lazy_model, MOI.LazyConstraintCallback(), my_callback)</code></pre><p>Now when we optimize!, our callback is run:</p><pre><code class="language-julia hljs">optimize!(lazy_model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">        1  -1.0000e+03   7.6667e+00   1.3143e+02
+        2  -4.9617e+02   5.0383e+02   1.9848e+00
+        3   3.8333e+00   4.0833e+00   6.1224e-02
+        4   4.0000e+00   4.0000e+00   0.0000e+00
+Terminating with the optimal solution</code></pre><p>Note how this problem also takes 4 iterations to converge, but the sequence of bounds is different compared to the iterative method.</p><p>Finally, we can obtain the optimal solution:</p><pre><code class="language-julia hljs">x_optimal = value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 0.0
+ 1.0</code></pre><pre><code class="language-julia hljs">optimal_ret = solve_subproblem(x_optimal)
+y_optimal = optimal_ret.y</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 0.0
+ 0.0</code></pre><h2 id="In-place-iterative-method"><a class="docs-heading-anchor" href="#In-place-iterative-method">In-place iterative method</a><a id="In-place-iterative-method-1"></a><a class="docs-heading-anchor-permalink" href="#In-place-iterative-method" title="Permalink"></a></h2><p>Our implementation of the iterative method has a problem: every time we need to solve the subproblem, we must rebuild it from scratch. This is expensive, and it can be the bottleneck in the solution process. We can improve our implementation by using re-using the subproblem between solves.</p><p>First, we create our first-stage problem as usual:</p><pre><code class="language-julia hljs">model = Model(GLPK.Optimizer)
+@variable(model, x[1:dim_x] &gt;= 0, Int)
+@variable(model, θ &gt;= M)
+@objective(model, Min, c_1&#39; * x + θ)
+print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min x[1] + 4 x[2] + θ
+Subject to
+ x[1] ≥ 0
+ x[2] ≥ 0
+ θ ≥ -1000
+ x[1] integer
+ x[2] integer</code></pre><p>Then, instead of building the subproblem in a function, we build it once here:</p><pre><code class="language-julia hljs">subproblem = Model(GLPK.Optimizer)
+@variable(subproblem, x_copy[1:dim_x])
+@variable(subproblem, y[1:dim_y] &gt;= 0)
+@constraint(subproblem, A_1 * x_copy + A_2 * y .&lt;= b)
+@objective(subproblem, Min, c_2&#39; * y)
+print(subproblem)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 2 y[1] + 3 y[2]
+Subject to
+ x_copy[1] - 3 x_copy[2] + y[1] - 2 y[2] ≤ -2
+ -x_copy[1] - 3 x_copy[2] - y[1] - y[2] ≤ -3
+ y[1] ≥ 0
+ y[2] ≥ 0</code></pre><p>This formulation is slightly different. We have included a copy of the x variables, <code>x_copy</code>, and used <code>x_copy</code> in the left-hand side of the constraints.</p><p>Our function to solve the subproblem is also slightly different. First, we need to fix the value of the <code>x_copy</code> variables to the value of <code>x</code> from the first-stage problem, and second, we compute the dual using the <a href="../../../api/JuMP/#JuMP.reduced_cost"><code>reduced_cost</code></a> of <code>x_copy</code>, not the dual of the linear constraints:</p><pre><code class="language-julia hljs">function solve_subproblem(model, x)
+    fix.(model[:x_copy], x)
+    optimize!(model)
+    @assert termination_status(model) == OPTIMAL
+    return (
+        obj = objective_value(model),
+        y = value.(model[:y]),
+        π = reduced_cost.(model[:x_copy]),
+    )
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_subproblem (generic function with 2 methods)</code></pre><p>Now we&#39;re ready to iterate our in-place Benders decomposition, but this time the cut computation is slightly different. Because we used <a href="../../../api/JuMP/#JuMP.reduced_cost"><code>reduced_cost</code></a> on the <code>x_copy</code> variables, the value of <code>π</code> is a valid subgradient on the objective of <code>subproblem</code> with respect to <code>x</code>. Therefore, we don&#39;t need to multiply the duals by <code>-A_1</code>, and so our cut uses <code>ret.π&#39;</code> instead of <code>-ret.π&#39; * A_1</code>:</p><pre><code class="language-julia hljs">println(&quot;Iteration  Lower Bound  Upper Bound          Gap&quot;)
+for k in 1:MAXIMUM_ITERATIONS
+    optimize!(model)
+    lower_bound = objective_value(model)
+    x_k = value.(x)
+    ret = solve_subproblem(subproblem, x_k)
+    upper_bound = c_1&#39; * x_k + ret.obj
+    gap = (upper_bound - lower_bound) / upper_bound
+    print_iteration(k, lower_bound, upper_bound, gap)
+    if gap &lt; ABSOLUTE_OPTIMALITY_GAP
+        println(&quot;Terminating with the optimal solution&quot;)
+        break
+    end
+    cut = @constraint(model, θ &gt;= ret.obj + ret.π&#39; * (x .- x_k))
+    @info &quot;Adding the cut $(cut)&quot;
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Iteration  Lower Bound  Upper Bound          Gap
+        1  -1.0000e+03   7.6667e+00   1.3143e+02
+<span class="sgr36"><span class="sgr1">[ Info: </span></span>Adding the cut 2 x[1] + 8 x[2] + θ ≥ 7.666666666666666
+        2  -4.9600e+02   1.2630e+03   1.3927e+00
+<span class="sgr36"><span class="sgr1">[ Info: </span></span>Adding the cut -1.5 x[1] + 4.5 x[2] + θ ≥ 3
+        3  -1.0800e+02   8.8800e+02   1.1216e+00
+<span class="sgr36"><span class="sgr1">[ Info: </span></span>Adding the cut θ ≥ 0
+        4   4.0000e+00   4.0000e+00   0.0000e+00
+Terminating with the optimal solution</code></pre><p>Finally, we can obtain the optimal solution:</p><pre><code class="language-julia hljs">optimize!(model)
+x_optimal = value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 0.0
+ 1.0</code></pre><pre><code class="language-julia hljs">optimal_ret = solve_subproblem(subproblem, x_optimal)
+y_optimal = optimal_ret.y</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 0.0
+ 0.0</code></pre><p>For larger problems, the benefit of re-using the same subproblem and not needing to multiply the duals by <code>A_1</code> in the cut coefficient usually outweights the cost of needing a copy of the <code>x</code> variables in the subproblem.</p><p>As a secondary benefit, because we no longer need an explicit representation of <code>A_1</code> in the cut, we can build the <code>model</code> and <code>subproblem</code> formulations using arbitrary JuMP syntax; they do not need to be in matrix form.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../conic/quantum_discrimination/">« Quantum state discrimination</a><a class="docs-footer-nextpage" href="../cutting_stock_column_generation/">Column generation »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation.jl b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation.jl
new file mode 100644
index 00000000000..93e2623e2d1
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation.jl
@@ -0,0 +1,386 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Column generation
+
+# The purpose of this tutorial is to demonstrate the column generation
+# algorithm. As an example, it solves the [Cutting stock problem](https://en.wikipedia.org/wiki/Cutting_stock_problem).
+
+# This tutorial uses the following packages:
+
+using JuMP
+import DataFrames
+import HiGHS
+import Plots
+import SparseArrays
+import Test  #src
+
+# ## Background
+
+# The cutting stock problem is about cutting large rolls of paper into smaller
+# pieces.
+
+# We denote the set of possible sized pieces that a roll can be cut into by
+# ``i\in 1,\ldots,I``. Each piece ``i`` has a width, ``w_i``, and a demand,
+# ``d_i``. The width of the large roll is ``W``.
+
+# Our objective is to minimize the number of rolls needed to meet all demand.
+
+# Here's the data that we are going to use in this tutorial:
+
+struct Piece
+    w::Float64
+    d::Int
+end
+
+struct Data
+    pieces::Vector{Piece}
+    W::Float64
+end
+
+function Base.show(io::IO, d::Data)
+    println(io, "Data for the cutting stock problem:")
+    println(io, "  W = $(d.W)")
+    println(io, "with pieces:")
+    println(io, "   i   w_i d_i")
+    println(io, "  ------------")
+    for (i, p) in enumerate(d.pieces)
+        println(io, lpad(i, 4), " ", lpad(p.w, 5), " ", lpad(p.d, 3))
+    end
+    return
+end
+
+function get_data()
+    data = [
+        75.0 38
+        75.0 44
+        75.0 30
+        75.0 41
+        75.0 36
+        53.8 33
+        53.0 36
+        51.0 41
+        50.2 35
+        32.2 37
+        30.8 44
+        29.8 49
+        20.1 37
+        16.2 36
+        14.5 42
+        11.0 33
+        8.6 47
+        8.2 35
+        6.6 49
+        5.1 42
+    ]
+    return Data([Piece(data[i, 1], data[i, 2]) for i in axes(data, 1)], 100.0)
+end
+
+data = get_data()
+
+# ## Mathematical formulation
+
+# To formulate the cutting stock problem as a mixed-integer linear program, we
+# assume that there is a set of large rolls ``j=1,\ldots,J`` to use. Then, we
+# introduce two classes of decision variables:
+# * ``x_{ij} \ge 0,\; \text{integer}, \; \forall i=1,\ldots,I,\; j=1,\ldots,J``
+# * ``y_{j} \in \{0, 1\},\; \forall j=1,\ldots,J.``
+# ``y_j`` is a binary variable that indicates if we use roll ``j``, and
+# ``x_{ij}`` counts how many pieces of size ``i`` that we cut from roll ``j``.
+
+# Our mixed-integer linear program is therefore:
+# ```math
+# \begin{align}
+#     \min & \sum\limits_{j=1}^J y_j \\
+#     \;\;\text{s.t.} & \sum\limits_{i=1}^N w_i x_{ij} \le W y_j & \forall j=1,\ldots,J \\
+#          & \sum\limits_{j=1}^J x_{ij} \ge d_i & \forall i=1,\ldots,I \\
+#          & x_{ij} \ge 0 & \forall i=1,\ldots,N, j=1,\ldots,J \\
+#          & x_{ij} \in \mathbb{Z} & \forall i=1,\ldots,I, j=1,\ldots,J \\
+#          & y_{j} \in \{0, 1\} & \forall j=1,\ldots,J \\
+# \end{align}
+# ```
+# The objective is to minimize the number of rolls that we use, and the two
+# constraints ensure that we respect the total width of each large roll and that
+# we satisfy demand exactly.
+
+# The JuMP formulation of this model is:
+
+I = length(data.pieces)
+J = 1_000  # Some large number
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:I, 1:J] >= 0, Int)
+@variable(model, y[1:J], Bin)
+@objective(model, Min, sum(y))
+@constraint(model, [i in 1:I], sum(x[i, :]) >= data.pieces[i].d)
+@constraint(
+    model,
+    [j in 1:J],
+    sum(data.pieces[i].w * x[i, j] for i in 1:I) <= data.W * y[j],
+);
+
+# Unfortunately, we can't solve this formulation for realistic instances because
+# it takes a very long time to solve. (Try removing the time limit.)
+
+set_time_limit_sec(model, 5.0)
+optimize!(model)
+solution_summary(model)
+
+# However, there is a formulation that solves much faster, and that is to use a
+# column generation scheme.
+
+# ## Column generation theory
+
+# The key insight for column generation is to recognize that feasible columns
+# in the ``x`` matrix of variables encode _cutting patterns_.
+
+# For example, if we look only at the roll ``j=1``, then a feasible solution is:
+#
+#  * ``x_{1,1} = 1`` (1 unit of piece \#1)
+#  * ``x_{13,1} = 1`` (1 unit of piece \#13)
+#  * All other ``x_{i,1} = 0``
+#
+# Another solution is
+#
+#  * ``x_{20,1} = 19`` (19 unit of piece \#20)
+#  * All other ``x_{i,1} = 0``
+#
+# Cutting patterns like ``x_{1,1} = 1`` and ``x_{2,1} = 1`` are infeasible
+# because the combined length is greater than ``W``.
+
+# Since there are a finite number of ways that we could cut a roll into a
+# valid cutting pattern, we could create a set of all possible cutting patterns
+# ``p = 1,\ldots,P``, with data ``a_{i,p}`` indicating how many units of piece
+# ``i`` we cut in pattern ``p``. Then, we can formulate our mixed-integer linear
+# program as:
+# ```math
+# \begin{align}
+#     \min & \sum\limits_{p=1}^P x_p \\
+#     \;\;\text{s.t.} & \sum\limits_{p=1}^P a_{ip} x_p \ge d_i & \forall i=1,\ldots,I \\
+#          & x_{p} \ge 0 & \forall p=1,\ldots,P \\
+#          & x_{p} \in \mathbb{Z} & \forall p=1,\ldots,P
+# \end{align}
+# ```
+#
+# Unfortunately, there will be a very large number of these patterns, so it is
+# often intractable to enumerate all columns ``p=1,\ldots,P``.
+#
+# Column generation is an iterative algorithm that starts with a small set of
+# initial patterns, and then cleverly chooses new columns to add to the main
+# MILP so that we find the optimal solution without having to enumerate every
+# column.
+
+# ## Choosing the initial set of patterns
+
+# For the initial set of patterns, we create a trivial cutting pattern which
+# cuts as many units of piece ``i`` as will fit.
+
+patterns = map(1:I) do i
+    n_pieces = floor(Int, data.W / data.pieces[i].w)
+    return SparseArrays.sparsevec([i], [n_pieces], I)
+end
+
+# We can visualize the patterns as follows:
+
+"""
+    cutting_locations(data::Data, pattern::SparseArrays.SparseVector)
+
+A function which returns a vector of the locations along the roll at which to
+cut in order to produce pattern `pattern`.
+"""
+function cutting_locations(data::Data, pattern::SparseArrays.SparseVector)
+    locations = Float64[]
+    offset = 0.0
+    for (i, c) in zip(SparseArrays.findnz(pattern)...)
+        for _ in 1:c
+            offset += data.pieces[i].w
+            push!(locations, offset)
+        end
+    end
+    return locations
+end
+
+function plot_patterns(data::Data, patterns)
+    plot = Plots.bar(;
+        xlims = (0, length(patterns) + 1),
+        ylims = (0, data.W),
+        xlabel = "Pattern",
+        ylabel = "Roll length",
+    )
+    for (i, p) in enumerate(patterns)
+        locations = cutting_locations(data, p)
+        Plots.bar!(
+            plot,
+            fill(i, length(locations)),
+            reverse(locations);
+            bar_width = 0.6,
+            label = false,
+            color = "#90caf9",
+        )
+    end
+    return plot
+end
+
+plot_patterns(data, patterns)
+
+# ## The base problem
+
+# Using the initial set of patterns, we can create and optimize our base model:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:length(patterns)] >= 0, Int)
+@objective(model, Min, sum(x))
+@constraint(model, demand[i in 1:I], patterns[i]' * x >= data.pieces[i].d)
+optimize!(model)
+solution_summary(model)
+
+# This solution requires 421 rolls. This solution is sub-optimal because the
+# model does not contain the full set of possible patterns.
+
+# How do we find a new column that leads to an improved solution?
+
+# ## Choosing new columns
+
+# Column generation chooses a new column by relaxing the integrality constraint
+# on ``x`` and looking at the dual variable ``\pi_i`` associated with demand
+# constraint ``i``.
+
+# For example, the dual of `demand[13]` is:
+
+unset_integer.(x)
+optimize!(model)
+π_13 = dual(demand[13])
+
+# Using the economic interpretation of the dual variable, we can say that a one
+# unit increase in demand for piece ``i`` will cost an extra ``\pi_i`` rolls.
+# Alternatively, we can say that a one unit increase in the left-hand side
+# (for example, due to a new cutting pattern) will _save_ us ``\pi_i`` rolls.
+# Therefore, we want a new column that maximizes the savings associated with
+# the dual variables, while respecting the total width of the roll:
+# ```math
+# \begin{align}
+#     \max & \sum\limits_{i=1}^I \pi_i y_i \\
+#     \;\;\text{s.t.} & \sum\limits_{i=1}^I w_i y_{i} \le W \\
+#          & y_{i} \ge 0 & \forall i=1,\ldots,I \\
+#          & y_{i} \in \mathbb{Z} & \forall i=1,\ldots,I \\
+# \end{align}
+# ```
+# If this problem, called the _pricing problem_, has an objective value greater
+# than ``1``, then we estimate than adding `y` as the coefficients of a new
+# column will decrease the objective by more than the cost of an extra roll.
+
+# Here is code to solve the pricing problem:
+
+function solve_pricing(data::Data, π::Vector{Float64})
+    I = length(π)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, y[1:I] >= 0, Int)
+    @constraint(model, sum(data.pieces[i].w * y[i] for i in 1:I) <= data.W)
+    @objective(model, Max, sum(π[i] * y[i] for i in 1:I))
+    optimize!(model)
+    number_of_rolls_saved = objective_value(model)
+    if number_of_rolls_saved > 1 + 1e-8
+        ## Benefit of pattern is more than the cost of a new roll plus some
+        ## tolerance
+        return SparseArrays.sparse(round.(Int, value.(y)))
+    end
+    return nothing
+end
+
+# If we solve the pricing problem with an artificial dual vector:
+
+solve_pricing(data, [1.0 / i for i in 1:I])
+
+# the solution is a roll with 1 unit of piece \#1, 1 unit of piece \#17, and 3
+# units of piece \#20.
+
+# If we solve the pricing problem with a dual vector of zeros, then the benefit
+# of the new pattern is less than the cost of a roll, and so the function
+# returns `nothing`:
+
+solve_pricing(data, zeros(I))
+
+# ## Iterative algorithm
+
+# Now we can combine our base model with the pricing subproblem in an iterative
+# column generation scheme:
+
+while true
+    ## Solve the linear relaxation
+    optimize!(model)
+    ## Obtain a new dual vector
+    π = dual.(demand)
+    ## Solve the pricing problem
+    new_pattern = solve_pricing(data, π)
+    ## Stop iterating if there is no new pattern
+    if new_pattern === nothing
+        @info "No new patterns, terminating the algorithm."
+        break
+    end
+    push!(patterns, new_pattern)
+    ## Create a new column
+    push!(x, @variable(model, lower_bound = 0))
+    ## Update the objective coefficient of the new column
+    set_objective_coefficient(model, x[end], 1.0)
+    ## Update the non-zeros in the coefficient matrix
+    for (i, count) in zip(SparseArrays.findnz(new_pattern)...)
+        set_normalized_coefficient(demand[i], x[end], count)
+    end
+    println("Found new pattern. Total patterns = $(length(patterns))")
+end
+
+# We found lots of new patterns. Here's pattern 21:
+
+patterns[21]
+
+# Let's have a look at the patterns now:
+
+plot_patterns(data, patterns)
+
+# ## Looking at the solution
+
+# Let's see how many of each column we need:
+
+solution = DataFrames.DataFrame([
+    (pattern = p, rolls = value(x_p)) for (p, x_p) in enumerate(x)
+])
+filter!(row -> row.rolls > 0, solution)
+
+# Since we solved a linear program, some of our columns have fractional
+# solutions. We can create a integer feasible solution by rounding up the
+# orders. This requires 341 rolls:
+
+Test.@test sum(ceil.(Int, solution.rolls)) == 341  #src
+sum(ceil.(Int, solution.rolls))
+
+# Alternatively, we can re-introduce the integrality constraints and resolve the
+# problem:
+
+set_integer.(x)
+optimize!(model)
+solution = DataFrames.DataFrame([
+    (pattern = p, rolls = value(x_p)) for (p, x_p) in enumerate(x)
+])
+filter!(row -> row.rolls > 0, solution)
+
+# This now requires 334 rolls:
+
+Test.@test isapprox(sum(solution.rolls), 334; atol = 1e-6)  #src
+sum(solution.rolls)
+
+# Note that this may not be the global minimum because we are not adding new
+# columns during the solution of the mixed-integer problem `model` (an algorithm
+# known as [branch and price](https://en.wikipedia.org/wiki/Branch_and_price)).
+# Nevertheless, the column generation algorithm typically finds good integer
+# feasible solutions to an otherwise intractable optimization problem.
+
+# ## Next steps
+
+# * Our objective function is to minimize the total number of rolls. What is the
+#   total length of waste? How does that compare to the total demand?
+# * Writing the optimization algorithm is only part of the challenge. Can you
+#   develop a better way to communicate the solution to stakeholders?
diff --git a/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/703571d3.svg b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/703571d3.svg
new file mode 100644
index 00000000000..9733ccb96ba
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/703571d3.svg
@@ -0,0 +1,350 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip830">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip830)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip831">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip830)" d="M232.389 1423.18 L2352.76 1423.18 L2352.76 47.2441 L232.389 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip832">
+    <rect x="232" y="47" width="2121" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,1423.18 232.389,47.2441 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="535.299,1423.18 535.299,47.2441 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="838.208,1423.18 838.208,47.2441 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1141.12,1423.18 1141.12,47.2441 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1444.03,1423.18 1444.03,47.2441 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1746.94,1423.18 1746.94,47.2441 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2049.85,1423.18 2049.85,47.2441 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2352.76,1423.18 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 232.389,1404.28 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="535.299,1423.18 535.299,1404.28 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="838.208,1423.18 838.208,1404.28 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1141.12,1423.18 1141.12,1404.28 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1444.03,1423.18 1444.03,1404.28 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1746.94,1423.18 1746.94,1404.28 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2049.85,1423.18 2049.85,1404.28 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2352.76,1423.18 2352.76,1404.28 "/>
+<path clip-path="url(#clip830)" d="M232.389 1454.1 Q228.778 1454.1 226.95 1457.66 Q225.144 1461.2 225.144 1468.33 Q225.144 1475.44 226.95 1479.01 Q228.778 1482.55 232.389 1482.55 Q236.024 1482.55 237.829 1479.01 Q239.658 1475.44 239.658 1468.33 Q239.658 1461.2 237.829 1457.66 Q236.024 1454.1 232.389 1454.1 M232.389 1450.39 Q238.2 1450.39 241.255 1455 Q244.334 1459.58 244.334 1468.33 Q244.334 1477.06 241.255 1481.67 Q238.2 1486.25 232.389 1486.25 Q226.579 1486.25 223.501 1481.67 Q220.445 1477.06 220.445 1468.33 Q220.445 1459.58 223.501 1455 Q226.579 1450.39 232.389 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M539.547 1466.95 Q542.903 1467.66 544.778 1469.93 Q546.676 1472.2 546.676 1475.53 Q546.676 1480.65 543.158 1483.45 Q539.639 1486.25 533.158 1486.25 Q530.982 1486.25 528.667 1485.81 Q526.375 1485.39 523.922 1484.54 L523.922 1480.02 Q525.866 1481.16 528.181 1481.74 Q530.496 1482.32 533.019 1482.32 Q537.417 1482.32 539.709 1480.58 Q542.023 1478.84 542.023 1475.53 Q542.023 1472.48 539.871 1470.77 Q537.741 1469.03 533.922 1469.03 L529.894 1469.03 L529.894 1465.19 L534.107 1465.19 Q537.556 1465.19 539.384 1463.82 Q541.213 1462.43 541.213 1459.84 Q541.213 1457.18 539.315 1455.77 Q537.44 1454.33 533.922 1454.33 Q532 1454.33 529.801 1454.75 Q527.602 1455.16 524.963 1456.04 L524.963 1451.88 Q527.625 1451.14 529.94 1450.77 Q532.278 1450.39 534.338 1450.39 Q539.662 1450.39 542.764 1452.83 Q545.866 1455.23 545.866 1459.35 Q545.866 1462.22 544.222 1464.21 Q542.579 1466.18 539.547 1466.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M838.613 1466.44 Q835.465 1466.44 833.613 1468.59 Q831.785 1470.74 831.785 1474.49 Q831.785 1478.22 833.613 1480.39 Q835.465 1482.55 838.613 1482.55 Q841.762 1482.55 843.59 1480.39 Q845.442 1478.22 845.442 1474.49 Q845.442 1470.74 843.59 1468.59 Q841.762 1466.44 838.613 1466.44 M847.896 1451.78 L847.896 1456.04 Q846.137 1455.21 844.331 1454.77 Q842.549 1454.33 840.789 1454.33 Q836.16 1454.33 833.706 1457.45 Q831.276 1460.58 830.928 1466.9 Q832.294 1464.89 834.354 1463.82 Q836.414 1462.73 838.891 1462.73 Q844.1 1462.73 847.109 1465.9 Q850.141 1469.05 850.141 1474.49 Q850.141 1479.82 846.993 1483.03 Q843.845 1486.25 838.613 1486.25 Q832.618 1486.25 829.447 1481.67 Q826.276 1477.06 826.276 1468.33 Q826.276 1460.14 830.164 1455.28 Q834.053 1450.39 840.604 1450.39 Q842.363 1450.39 844.146 1450.74 Q845.951 1451.09 847.896 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1131.42 1484.86 L1131.42 1480.6 Q1133.18 1481.44 1134.98 1481.88 Q1136.79 1482.32 1138.53 1482.32 Q1143.15 1482.32 1145.59 1479.21 Q1148.04 1476.09 1148.39 1469.75 Q1147.04 1471.74 1144.98 1472.8 Q1142.92 1473.87 1140.42 1473.87 Q1135.24 1473.87 1132.21 1470.74 Q1129.2 1467.59 1129.2 1462.15 Q1129.2 1456.83 1132.34 1453.61 Q1135.49 1450.39 1140.72 1450.39 Q1146.72 1450.39 1149.87 1455 Q1153.04 1459.58 1153.04 1468.33 Q1153.04 1476.51 1149.15 1481.39 Q1145.28 1486.25 1138.73 1486.25 Q1136.97 1486.25 1135.17 1485.9 Q1133.36 1485.56 1131.42 1484.86 M1140.72 1470.21 Q1143.87 1470.21 1145.7 1468.06 Q1147.55 1465.9 1147.55 1462.15 Q1147.55 1458.43 1145.7 1456.27 Q1143.87 1454.1 1140.72 1454.1 Q1137.58 1454.1 1135.72 1456.27 Q1133.9 1458.43 1133.9 1462.15 Q1133.9 1465.9 1135.72 1468.06 Q1137.58 1470.21 1140.72 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1419.51 1481.64 L1427.15 1481.64 L1427.15 1455.28 L1418.84 1456.95 L1418.84 1452.69 L1427.11 1451.02 L1431.78 1451.02 L1431.78 1481.64 L1439.42 1481.64 L1439.42 1485.58 L1419.51 1485.58 L1419.51 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1452.89 1481.64 L1469.21 1481.64 L1469.21 1485.58 L1447.27 1485.58 L1447.27 1481.64 Q1449.93 1478.89 1454.51 1474.26 Q1459.12 1469.61 1460.3 1468.27 Q1462.55 1465.74 1463.43 1464.01 Q1464.33 1462.25 1464.33 1460.56 Q1464.33 1457.8 1462.38 1456.07 Q1460.46 1454.33 1457.36 1454.33 Q1455.16 1454.33 1452.71 1455.09 Q1450.28 1455.86 1447.5 1457.41 L1447.5 1452.69 Q1450.32 1451.55 1452.78 1450.97 Q1455.23 1450.39 1457.27 1450.39 Q1462.64 1450.39 1465.83 1453.08 Q1469.03 1455.77 1469.03 1460.26 Q1469.03 1462.39 1468.22 1464.31 Q1467.43 1466.2 1465.32 1468.8 Q1464.74 1469.47 1461.64 1472.69 Q1458.54 1475.88 1452.89 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1722.12 1481.64 L1729.76 1481.64 L1729.76 1455.28 L1721.45 1456.95 L1721.45 1452.69 L1729.71 1451.02 L1734.39 1451.02 L1734.39 1481.64 L1742.03 1481.64 L1742.03 1485.58 L1722.12 1485.58 L1722.12 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1751.52 1451.02 L1769.88 1451.02 L1769.88 1454.96 L1755.8 1454.96 L1755.8 1463.43 Q1756.82 1463.08 1757.84 1462.92 Q1758.86 1462.73 1759.88 1462.73 Q1765.66 1462.73 1769.04 1465.9 Q1772.42 1469.08 1772.42 1474.49 Q1772.42 1480.07 1768.95 1483.17 Q1765.48 1486.25 1759.16 1486.25 Q1756.98 1486.25 1754.71 1485.88 Q1752.47 1485.51 1750.06 1484.77 L1750.06 1480.07 Q1752.15 1481.2 1754.37 1481.76 Q1756.59 1482.32 1759.07 1482.32 Q1763.07 1482.32 1765.41 1480.21 Q1767.75 1478.1 1767.75 1474.49 Q1767.75 1470.88 1765.41 1468.77 Q1763.07 1466.67 1759.07 1466.67 Q1757.19 1466.67 1755.32 1467.08 Q1753.46 1467.5 1751.52 1468.38 L1751.52 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M2024.58 1481.64 L2032.22 1481.64 L2032.22 1455.28 L2023.91 1456.95 L2023.91 1452.69 L2032.17 1451.02 L2036.85 1451.02 L2036.85 1481.64 L2044.49 1481.64 L2044.49 1485.58 L2024.58 1485.58 L2024.58 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M2063.93 1469.17 Q2060.6 1469.17 2058.68 1470.95 Q2056.78 1472.73 2056.78 1475.86 Q2056.78 1478.98 2058.68 1480.77 Q2060.6 1482.55 2063.93 1482.55 Q2067.27 1482.55 2069.19 1480.77 Q2071.11 1478.96 2071.11 1475.86 Q2071.11 1472.73 2069.19 1470.95 Q2067.29 1469.17 2063.93 1469.17 M2059.26 1467.18 Q2056.25 1466.44 2054.56 1464.38 Q2052.89 1462.32 2052.89 1459.35 Q2052.89 1455.21 2055.83 1452.8 Q2058.79 1450.39 2063.93 1450.39 Q2069.09 1450.39 2072.03 1452.8 Q2074.97 1455.21 2074.97 1459.35 Q2074.97 1462.32 2073.28 1464.38 Q2071.62 1466.44 2068.63 1467.18 Q2072.01 1467.96 2073.89 1470.26 Q2075.78 1472.55 2075.78 1475.86 Q2075.78 1480.88 2072.71 1483.57 Q2069.65 1486.25 2063.93 1486.25 Q2058.21 1486.25 2055.14 1483.57 Q2052.08 1480.88 2052.08 1475.86 Q2052.08 1472.55 2053.98 1470.26 Q2055.88 1467.96 2059.26 1467.18 M2057.54 1459.79 Q2057.54 1462.48 2059.21 1463.98 Q2060.9 1465.49 2063.93 1465.49 Q2066.94 1465.49 2068.63 1463.98 Q2070.34 1462.48 2070.34 1459.79 Q2070.34 1457.11 2068.63 1455.6 Q2066.94 1454.1 2063.93 1454.1 Q2060.9 1454.1 2059.21 1455.6 Q2057.54 1457.11 2057.54 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M2332.14 1481.64 L2348.46 1481.64 L2348.46 1485.58 L2326.52 1485.58 L2326.52 1481.64 Q2329.18 1478.89 2333.76 1474.26 Q2338.37 1469.61 2339.55 1468.27 Q2341.8 1465.74 2342.67 1464.01 Q2343.58 1462.25 2343.58 1460.56 Q2343.58 1457.8 2341.63 1456.07 Q2339.71 1454.33 2336.61 1454.33 Q2334.41 1454.33 2331.96 1455.09 Q2329.53 1455.86 2326.75 1457.41 L2326.75 1452.69 Q2329.57 1451.55 2332.03 1450.97 Q2334.48 1450.39 2336.52 1450.39 Q2341.89 1450.39 2345.08 1453.08 Q2348.28 1455.77 2348.28 1460.26 Q2348.28 1462.39 2347.47 1464.31 Q2346.68 1466.2 2344.57 1468.8 Q2343.99 1469.47 2340.89 1472.69 Q2337.79 1475.88 2332.14 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M2359.09 1481.64 L2366.73 1481.64 L2366.73 1455.28 L2358.42 1456.95 L2358.42 1452.69 L2366.68 1451.02 L2371.36 1451.02 L2371.36 1481.64 L2378.99 1481.64 L2378.99 1485.58 L2359.09 1485.58 L2359.09 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1187.71 1525.81 L1187.71 1543.66 L1195.8 1543.66 Q1200.29 1543.66 1202.74 1541.34 Q1205.19 1539.02 1205.19 1534.72 Q1205.19 1530.45 1202.74 1528.13 Q1200.29 1525.81 1195.8 1525.81 L1187.71 1525.81 M1181.28 1520.52 L1195.8 1520.52 Q1203.79 1520.52 1207.86 1524.15 Q1211.97 1527.75 1211.97 1534.72 Q1211.97 1541.75 1207.86 1545.35 Q1203.79 1548.95 1195.8 1548.95 L1187.71 1548.95 L1187.71 1568.04 L1181.28 1568.04 L1181.28 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1233.61 1550.12 Q1226.51 1550.12 1223.78 1551.75 Q1221.04 1553.37 1221.04 1557.29 Q1221.04 1560.4 1223.08 1562.25 Q1225.14 1564.07 1228.68 1564.07 Q1233.55 1564.07 1236.47 1560.63 Q1239.43 1557.16 1239.43 1551.43 L1239.43 1550.12 L1233.61 1550.12 M1245.29 1547.71 L1245.29 1568.04 L1239.43 1568.04 L1239.43 1562.63 Q1237.43 1565.88 1234.44 1567.44 Q1231.45 1568.97 1227.12 1568.97 Q1221.64 1568.97 1218.4 1565.91 Q1215.18 1562.82 1215.18 1557.67 Q1215.18 1551.65 1219.19 1548.6 Q1223.23 1545.54 1231.22 1545.54 L1239.43 1545.54 L1239.43 1544.97 Q1239.43 1540.93 1236.76 1538.73 Q1234.12 1536.5 1229.31 1536.5 Q1226.26 1536.5 1223.36 1537.23 Q1220.47 1537.97 1217.79 1539.43 L1217.79 1534.02 Q1221.01 1532.78 1224.03 1532.17 Q1227.05 1531.54 1229.92 1531.54 Q1237.65 1531.54 1241.47 1535.55 Q1245.29 1539.56 1245.29 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1263.15 1522.27 L1263.15 1532.4 L1275.21 1532.4 L1275.21 1536.95 L1263.15 1536.95 L1263.15 1556.3 Q1263.15 1560.66 1264.32 1561.9 Q1265.53 1563.14 1269.19 1563.14 L1275.21 1563.14 L1275.21 1568.04 L1269.19 1568.04 Q1262.42 1568.04 1259.84 1565.53 Q1257.26 1562.98 1257.26 1556.3 L1257.26 1536.95 L1252.96 1536.95 L1252.96 1532.4 L1257.26 1532.4 L1257.26 1522.27 L1263.15 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1288.71 1522.27 L1288.71 1532.4 L1300.77 1532.4 L1300.77 1536.95 L1288.71 1536.95 L1288.71 1556.3 Q1288.71 1560.66 1289.88 1561.9 Q1291.09 1563.14 1294.75 1563.14 L1300.77 1563.14 L1300.77 1568.04 L1294.75 1568.04 Q1287.97 1568.04 1285.4 1565.53 Q1282.82 1562.98 1282.82 1556.3 L1282.82 1536.95 L1278.52 1536.95 L1278.52 1532.4 L1282.82 1532.4 L1282.82 1522.27 L1288.71 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1338.96 1548.76 L1338.96 1551.62 L1312.04 1551.62 Q1312.42 1557.67 1315.66 1560.85 Q1318.94 1564 1324.77 1564 Q1328.14 1564 1331.29 1563.17 Q1334.47 1562.35 1337.59 1560.69 L1337.59 1566.23 Q1334.44 1567.57 1331.13 1568.27 Q1327.82 1568.97 1324.42 1568.97 Q1315.89 1568.97 1310.89 1564 Q1305.92 1559.04 1305.92 1550.57 Q1305.92 1541.82 1310.64 1536.69 Q1315.38 1531.54 1323.4 1531.54 Q1330.59 1531.54 1334.76 1536.18 Q1338.96 1540.8 1338.96 1548.76 M1333.11 1547.04 Q1333.04 1542.23 1330.4 1539.37 Q1327.79 1536.5 1323.46 1536.5 Q1318.56 1536.5 1315.6 1539.27 Q1312.67 1542.04 1312.23 1547.07 L1333.11 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1369.23 1537.87 Q1368.24 1537.3 1367.07 1537.04 Q1365.92 1536.76 1364.52 1536.76 Q1359.56 1536.76 1356.88 1540 Q1354.24 1543.22 1354.24 1549.27 L1354.24 1568.04 L1348.35 1568.04 L1348.35 1532.4 L1354.24 1532.4 L1354.24 1537.93 Q1356.09 1534.69 1359.05 1533.13 Q1362.01 1531.54 1366.24 1531.54 Q1366.84 1531.54 1367.58 1531.63 Q1368.31 1531.7 1369.2 1531.85 L1369.23 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M1403.86 1546.53 L1403.86 1568.04 L1398 1568.04 L1398 1546.72 Q1398 1541.66 1396.03 1539.14 Q1394.06 1536.63 1390.11 1536.63 Q1385.37 1536.63 1382.63 1539.65 Q1379.89 1542.68 1379.89 1547.9 L1379.89 1568.04 L1374.01 1568.04 L1374.01 1532.4 L1379.89 1532.4 L1379.89 1537.93 Q1381.99 1534.72 1384.83 1533.13 Q1387.69 1531.54 1391.42 1531.54 Q1397.56 1531.54 1400.71 1535.36 Q1403.86 1539.14 1403.86 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,1147.99 2352.76,1147.99 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,872.806 2352.76,872.806 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,597.618 2352.76,597.618 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,322.431 2352.76,322.431 "/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,47.2441 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 232.389,47.2441 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 251.287,1423.18 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1147.99 251.287,1147.99 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,872.806 251.287,872.806 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,597.618 251.287,597.618 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,322.431 251.287,322.431 "/>
+<polyline clip-path="url(#clip830)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,47.2441 251.287,47.2441 "/>
+<path clip-path="url(#clip830)" d="M184.445 1408.98 Q180.834 1408.98 179.005 1412.54 Q177.2 1416.08 177.2 1423.21 Q177.2 1430.32 179.005 1433.89 Q180.834 1437.43 184.445 1437.43 Q188.079 1437.43 189.885 1433.89 Q191.713 1430.32 191.713 1423.21 Q191.713 1416.08 189.885 1412.54 Q188.079 1408.98 184.445 1408.98 M184.445 1405.27 Q190.255 1405.27 193.311 1409.88 Q196.389 1414.46 196.389 1423.21 Q196.389 1431.94 193.311 1436.55 Q190.255 1441.13 184.445 1441.13 Q178.635 1441.13 175.556 1436.55 Q172.501 1431.94 172.501 1423.21 Q172.501 1414.46 175.556 1409.88 Q178.635 1405.27 184.445 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M148.311 1161.34 L164.63 1161.34 L164.63 1165.27 L142.686 1165.27 L142.686 1161.34 Q145.348 1158.58 149.931 1153.95 Q154.538 1149.3 155.718 1147.96 Q157.964 1145.43 158.843 1143.7 Q159.746 1141.94 159.746 1140.25 Q159.746 1137.5 157.802 1135.76 Q155.88 1134.02 152.778 1134.02 Q150.579 1134.02 148.126 1134.79 Q145.695 1135.55 142.917 1137.1 L142.917 1132.38 Q145.741 1131.25 148.195 1130.67 Q150.649 1130.09 152.686 1130.09 Q158.056 1130.09 161.251 1132.77 Q164.445 1135.46 164.445 1139.95 Q164.445 1142.08 163.635 1144 Q162.848 1145.9 160.741 1148.49 Q160.163 1149.16 157.061 1152.38 Q153.959 1155.57 148.311 1161.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M184.445 1133.79 Q180.834 1133.79 179.005 1137.36 Q177.2 1140.9 177.2 1148.03 Q177.2 1155.13 179.005 1158.7 Q180.834 1162.24 184.445 1162.24 Q188.079 1162.24 189.885 1158.7 Q191.713 1155.13 191.713 1148.03 Q191.713 1140.9 189.885 1137.36 Q188.079 1133.79 184.445 1133.79 M184.445 1130.09 Q190.255 1130.09 193.311 1134.69 Q196.389 1139.28 196.389 1148.03 Q196.389 1156.75 193.311 1161.36 Q190.255 1165.94 184.445 1165.94 Q178.635 1165.94 175.556 1161.36 Q172.501 1156.75 172.501 1148.03 Q172.501 1139.28 175.556 1134.69 Q178.635 1130.09 184.445 1130.09 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M157.13 859.6 L145.325 878.049 L157.13 878.049 L157.13 859.6 M155.903 855.526 L161.783 855.526 L161.783 878.049 L166.714 878.049 L166.714 881.937 L161.783 881.937 L161.783 890.086 L157.13 890.086 L157.13 881.937 L141.529 881.937 L141.529 877.424 L155.903 855.526 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M184.445 858.604 Q180.834 858.604 179.005 862.169 Q177.2 865.711 177.2 872.84 Q177.2 879.947 179.005 883.512 Q180.834 887.053 184.445 887.053 Q188.079 887.053 189.885 883.512 Q191.713 879.947 191.713 872.84 Q191.713 865.711 189.885 862.169 Q188.079 858.604 184.445 858.604 M184.445 854.901 Q190.255 854.901 193.311 859.507 Q196.389 864.09 196.389 872.84 Q196.389 881.567 193.311 886.174 Q190.255 890.757 184.445 890.757 Q178.635 890.757 175.556 886.174 Q172.501 881.567 172.501 872.84 Q172.501 864.09 175.556 859.507 Q178.635 854.901 184.445 854.901 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M154.862 595.755 Q151.714 595.755 149.862 597.908 Q148.033 600.061 148.033 603.81 Q148.033 607.537 149.862 609.713 Q151.714 611.866 154.862 611.866 Q158.01 611.866 159.839 609.713 Q161.69 607.537 161.69 603.81 Q161.69 600.061 159.839 597.908 Q158.01 595.755 154.862 595.755 M164.144 581.102 L164.144 585.362 Q162.385 584.528 160.579 584.088 Q158.797 583.649 157.038 583.649 Q152.408 583.649 149.954 586.774 Q147.524 589.899 147.177 596.218 Q148.542 594.204 150.603 593.139 Q152.663 592.051 155.14 592.051 Q160.348 592.051 163.357 595.223 Q166.39 598.371 166.39 603.81 Q166.39 609.135 163.241 612.352 Q160.093 615.57 154.862 615.57 Q148.866 615.57 145.695 610.986 Q142.524 606.38 142.524 597.653 Q142.524 589.459 146.413 584.598 Q150.302 579.713 156.853 579.713 Q158.612 579.713 160.394 580.061 Q162.2 580.408 164.144 581.102 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M184.445 583.417 Q180.834 583.417 179.005 586.982 Q177.2 590.524 177.2 597.653 Q177.2 604.76 179.005 608.324 Q180.834 611.866 184.445 611.866 Q188.079 611.866 189.885 608.324 Q191.713 604.76 191.713 597.653 Q191.713 590.524 189.885 586.982 Q188.079 583.417 184.445 583.417 M184.445 579.713 Q190.255 579.713 193.311 584.32 Q196.389 588.903 196.389 597.653 Q196.389 606.38 193.311 610.986 Q190.255 615.57 184.445 615.57 Q178.635 615.57 175.556 610.986 Q172.501 606.38 172.501 597.653 Q172.501 588.903 175.556 584.32 Q178.635 579.713 184.445 579.713 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M154.283 323.299 Q150.95 323.299 149.028 325.082 Q147.13 326.864 147.13 329.989 Q147.13 333.114 149.028 334.896 Q150.95 336.679 154.283 336.679 Q157.616 336.679 159.538 334.896 Q161.459 333.091 161.459 329.989 Q161.459 326.864 159.538 325.082 Q157.64 323.299 154.283 323.299 M149.607 321.309 Q146.598 320.568 144.908 318.508 Q143.241 316.447 143.241 313.485 Q143.241 309.341 146.181 306.934 Q149.144 304.526 154.283 304.526 Q159.445 304.526 162.385 306.934 Q165.325 309.341 165.325 313.485 Q165.325 316.447 163.635 318.508 Q161.968 320.568 158.982 321.309 Q162.362 322.096 164.237 324.387 Q166.135 326.679 166.135 329.989 Q166.135 335.012 163.056 337.697 Q160.001 340.383 154.283 340.383 Q148.566 340.383 145.487 337.697 Q142.431 335.012 142.431 329.989 Q142.431 326.679 144.329 324.387 Q146.228 322.096 149.607 321.309 M147.894 313.924 Q147.894 316.61 149.561 318.114 Q151.251 319.619 154.283 319.619 Q157.292 319.619 158.982 318.114 Q160.695 316.61 160.695 313.924 Q160.695 311.239 158.982 309.735 Q157.292 308.23 154.283 308.23 Q151.251 308.23 149.561 309.735 Q147.894 311.239 147.894 313.924 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M184.445 308.23 Q180.834 308.23 179.005 311.795 Q177.2 315.336 177.2 322.466 Q177.2 329.572 179.005 333.137 Q180.834 336.679 184.445 336.679 Q188.079 336.679 189.885 333.137 Q191.713 329.572 191.713 322.466 Q191.713 315.336 189.885 311.795 Q188.079 308.23 184.445 308.23 M184.445 304.526 Q190.255 304.526 193.311 309.133 Q196.389 313.716 196.389 322.466 Q196.389 331.193 193.311 335.799 Q190.255 340.383 184.445 340.383 Q178.635 340.383 175.556 335.799 Q172.501 331.193 172.501 322.466 Q172.501 313.716 175.556 309.133 Q178.635 304.526 184.445 304.526 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M114.931 60.5889 L122.57 60.5889 L122.57 34.2233 L114.26 35.89 L114.26 31.6308 L122.524 29.9641 L127.2 29.9641 L127.2 60.5889 L134.839 60.5889 L134.839 64.5241 L114.931 64.5241 L114.931 60.5889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M154.283 33.0428 Q150.672 33.0428 148.843 36.6076 Q147.038 40.1492 147.038 47.2788 Q147.038 54.3853 148.843 57.9501 Q150.672 61.4917 154.283 61.4917 Q157.917 61.4917 159.723 57.9501 Q161.552 54.3853 161.552 47.2788 Q161.552 40.1492 159.723 36.6076 Q157.917 33.0428 154.283 33.0428 M154.283 29.3391 Q160.093 29.3391 163.149 33.9456 Q166.227 38.5289 166.227 47.2788 Q166.227 56.0056 163.149 60.6121 Q160.093 65.1954 154.283 65.1954 Q148.473 65.1954 145.394 60.6121 Q142.339 56.0056 142.339 47.2788 Q142.339 38.5289 145.394 33.9456 Q148.473 29.3391 154.283 29.3391 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M184.445 33.0428 Q180.834 33.0428 179.005 36.6076 Q177.2 40.1492 177.2 47.2788 Q177.2 54.3853 179.005 57.9501 Q180.834 61.4917 184.445 61.4917 Q188.079 61.4917 189.885 57.9501 Q191.713 54.3853 191.713 47.2788 Q191.713 40.1492 189.885 36.6076 Q188.079 33.0428 184.445 33.0428 M184.445 29.3391 Q190.255 29.3391 193.311 33.9456 Q196.389 38.5289 196.389 47.2788 Q196.389 56.0056 193.311 60.6121 Q190.255 65.1954 184.445 65.1954 Q178.635 65.1954 175.556 60.6121 Q172.501 56.0056 172.501 47.2788 Q172.501 38.5289 175.556 33.9456 Q178.635 29.3391 184.445 29.3391 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M41.7242 880.191 Q42.4244 878.122 44.7161 876.181 Q47.0077 874.207 51.0181 872.234 L64.0042 865.709 L64.0042 872.616 L51.8138 878.695 Q47.0395 881.05 45.48 883.278 Q43.9204 885.474 43.9204 889.294 L43.9204 896.296 L64.0042 896.296 L64.0042 902.726 L16.4842 902.726 L16.4842 888.212 Q16.4842 880.064 19.8898 876.053 Q23.2955 872.043 30.1704 872.043 Q34.6582 872.043 37.6183 874.144 Q40.5784 876.212 41.7242 880.191 M21.7677 896.296 L38.6368 896.296 L38.6368 888.212 Q38.6368 883.565 36.5043 881.209 Q34.34 878.822 30.1704 878.822 Q26.0009 878.822 23.9002 881.209 Q21.7677 883.565 21.7677 888.212 L21.7677 896.296 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M32.4621 846.803 Q32.4621 851.513 36.1542 854.251 Q39.8145 856.988 46.212 856.988 Q52.6095 856.988 56.3017 854.283 Q59.9619 851.545 59.9619 846.803 Q59.9619 842.124 56.2698 839.387 Q52.5777 836.65 46.212 836.65 Q39.8781 836.65 36.186 839.387 Q32.4621 842.124 32.4621 846.803 M27.4968 846.803 Q27.4968 839.164 32.4621 834.803 Q37.4273 830.443 46.212 830.443 Q54.9649 830.443 59.9619 834.803 Q64.9272 839.164 64.9272 846.803 Q64.9272 854.473 59.9619 858.834 Q54.9649 863.163 46.212 863.163 Q37.4273 863.163 32.4621 858.834 Q27.4968 854.473 27.4968 846.803 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M14.479 820.735 L14.479 814.879 L64.0042 814.879 L64.0042 820.735 L14.479 820.735 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M14.479 802.625 L14.479 796.768 L64.0042 796.768 L64.0042 802.625 L14.479 802.625 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M14.479 763.794 L14.479 757.938 L64.0042 757.938 L64.0042 763.794 L14.479 763.794 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M44.7161 715.192 L47.5806 715.192 L47.5806 742.119 Q53.6281 741.737 56.8109 738.49 Q59.9619 735.212 59.9619 729.387 Q59.9619 726.014 59.1344 722.862 Q58.3069 719.68 56.6518 716.56 L62.1899 716.56 Q63.5267 719.711 64.227 723.022 Q64.9272 726.332 64.9272 729.737 Q64.9272 738.268 59.9619 743.265 Q54.9967 748.23 46.5303 748.23 Q37.7774 748.23 32.6531 743.519 Q27.4968 738.777 27.4968 730.756 Q27.4968 723.563 32.1438 719.393 Q36.7589 715.192 44.7161 715.192 M42.9973 721.048 Q38.1912 721.112 35.3266 723.754 Q32.4621 726.364 32.4621 730.692 Q32.4621 735.594 35.2312 738.554 Q38.0002 741.482 43.0292 741.928 L42.9973 721.048 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M42.4881 675.947 L64.0042 675.947 L64.0042 681.804 L42.679 681.804 Q37.6183 681.804 35.1038 683.777 Q32.5894 685.75 32.5894 689.697 Q32.5894 694.44 35.6131 697.177 Q38.6368 699.914 43.8567 699.914 L64.0042 699.914 L64.0042 705.802 L28.3562 705.802 L28.3562 699.914 L33.8944 699.914 Q30.6797 697.813 29.0883 694.981 Q27.4968 692.116 27.4968 688.392 Q27.4968 682.249 31.3163 679.098 Q35.1038 675.947 42.4881 675.947 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M45.7664 640.809 Q39.4007 640.809 35.8996 643.45 Q32.3984 646.06 32.3984 650.803 Q32.3984 655.513 35.8996 658.155 Q39.4007 660.765 45.7664 660.765 Q52.1003 660.765 55.6014 658.155 Q59.1026 655.513 59.1026 650.803 Q59.1026 646.06 55.6014 643.45 Q52.1003 640.809 45.7664 640.809 M59.58 634.952 Q68.683 634.952 73.1071 638.994 Q77.5631 643.037 77.5631 651.376 Q77.5631 654.463 77.0857 657.2 Q76.6401 659.937 75.6852 662.516 L69.9879 662.516 Q71.3884 659.937 72.0568 657.423 Q72.7252 654.909 72.7252 652.299 Q72.7252 646.538 69.7015 643.673 Q66.7096 640.809 60.6303 640.809 L57.7339 640.809 Q60.885 642.623 62.4446 645.456 Q64.0042 648.288 64.0042 652.235 Q64.0042 658.792 59.0071 662.802 Q54.01 666.812 45.7664 666.812 Q37.491 666.812 32.4939 662.802 Q27.4968 658.792 27.4968 652.235 Q27.4968 648.288 29.0564 645.456 Q30.616 642.623 33.7671 640.809 L28.3562 640.809 L28.3562 634.952 L59.58 634.952 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M18.2347 617.096 L28.3562 617.096 L28.3562 605.033 L32.9077 605.033 L32.9077 617.096 L52.2594 617.096 Q56.6199 617.096 57.8613 615.919 Q59.1026 614.709 59.1026 611.049 L59.1026 605.033 L64.0042 605.033 L64.0042 611.049 Q64.0042 617.828 61.4897 620.406 Q58.9434 622.985 52.2594 622.985 L32.9077 622.985 L32.9077 627.281 L28.3562 627.281 L28.3562 622.985 L18.2347 622.985 L18.2347 617.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip830)" d="M42.4881 567.698 L64.0042 567.698 L64.0042 573.555 L42.679 573.555 Q37.6183 573.555 35.1038 575.528 Q32.5894 577.502 32.5894 581.448 Q32.5894 586.191 35.6131 588.928 Q38.6368 591.665 43.8567 591.665 L64.0042 591.665 L64.0042 597.554 L14.479 597.554 L14.479 591.665 L33.8944 591.665 Q30.6797 589.565 29.0883 586.732 Q27.4968 583.867 27.4968 580.143 Q27.4968 574 31.3163 570.849 Q35.1038 567.698 42.4881 567.698 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip832)" d="M303.068 391.228 L303.068 1423.18 L363.65 1423.18 L363.65 391.228 L303.068 391.228 L303.068 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="303.068,391.228 303.068,1423.18 363.65,1423.18 363.65,391.228 303.068,391.228 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="333.359" cy="391.228" r="2"/>
+<path clip-path="url(#clip832)" d="M404.038 391.228 L404.038 1423.18 L464.62 1423.18 L464.62 391.228 L404.038 391.228 L404.038 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="404.038,391.228 404.038,1423.18 464.62,1423.18 464.62,391.228 404.038,391.228 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="434.329" cy="391.228" r="2"/>
+<path clip-path="url(#clip832)" d="M505.008 391.228 L505.008 1423.18 L565.59 1423.18 L565.59 391.228 L505.008 391.228 L505.008 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="505.008,391.228 505.008,1423.18 565.59,1423.18 565.59,391.228 505.008,391.228 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="535.299" cy="391.228" r="2"/>
+<path clip-path="url(#clip832)" d="M605.978 391.228 L605.978 1423.18 L666.56 1423.18 L666.56 391.228 L605.978 391.228 L605.978 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="605.978,391.228 605.978,1423.18 666.56,1423.18 666.56,391.228 605.978,391.228 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="636.269" cy="391.228" r="2"/>
+<path clip-path="url(#clip832)" d="M706.948 391.228 L706.948 1423.18 L767.529 1423.18 L767.529 391.228 L706.948 391.228 L706.948 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="706.948,391.228 706.948,1423.18 767.529,1423.18 767.529,391.228 706.948,391.228 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="737.239" cy="391.228" r="2"/>
+<path clip-path="url(#clip832)" d="M807.917 682.926 L807.917 1423.18 L868.499 1423.18 L868.499 682.926 L807.917 682.926 L807.917 682.926  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="807.917,682.926 807.917,1423.18 868.499,1423.18 868.499,682.926 807.917,682.926 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="838.208" cy="682.926" r="2"/>
+<path clip-path="url(#clip832)" d="M908.887 693.934 L908.887 1423.18 L969.469 1423.18 L969.469 693.934 L908.887 693.934 L908.887 693.934  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="908.887,693.934 908.887,1423.18 969.469,1423.18 969.469,693.934 908.887,693.934 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="939.178" cy="693.934" r="2"/>
+<path clip-path="url(#clip832)" d="M1009.86 721.453 L1009.86 1423.18 L1070.44 1423.18 L1070.44 721.453 L1009.86 721.453 L1009.86 721.453  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1009.86,721.453 1009.86,1423.18 1070.44,1423.18 1070.44,721.453 1009.86,721.453 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1040.15" cy="721.453" r="2"/>
+<path clip-path="url(#clip832)" d="M1110.83 732.46 L1110.83 1423.18 L1171.41 1423.18 L1171.41 732.46 L1110.83 732.46 L1110.83 732.46  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1110.83,732.46 1110.83,1423.18 1171.41,1423.18 1171.41,732.46 1110.83,732.46 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1141.12" cy="732.46" r="2"/>
+<path clip-path="url(#clip832)" d="M1211.8 94.0259 L1211.8 1423.18 L1272.38 1423.18 L1272.38 94.0259 L1211.8 94.0259 L1211.8 94.0259  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1211.8,94.0259 1211.8,1423.18 1272.38,1423.18 1272.38,94.0259 1211.8,94.0259 "/>
+<path clip-path="url(#clip832)" d="M1211.8 537.077 L1211.8 1423.18 L1272.38 1423.18 L1272.38 537.077 L1211.8 537.077 L1211.8 537.077  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1211.8,537.077 1211.8,1423.18 1272.38,1423.18 1272.38,537.077 1211.8,537.077 "/>
+<path clip-path="url(#clip832)" d="M1211.8 980.129 L1211.8 1423.18 L1272.38 1423.18 L1272.38 980.129 L1211.8 980.129 L1211.8 980.129  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1211.8,980.129 1211.8,1423.18 1272.38,1423.18 1272.38,980.129 1211.8,980.129 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1242.09" cy="94.0259" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1242.09" cy="537.077" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1242.09" cy="980.129" r="2"/>
+<path clip-path="url(#clip832)" d="M1312.77 151.815 L1312.77 1423.18 L1373.35 1423.18 L1373.35 151.815 L1312.77 151.815 L1312.77 151.815  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1312.77,151.815 1312.77,1423.18 1373.35,1423.18 1373.35,151.815 1312.77,151.815 "/>
+<path clip-path="url(#clip832)" d="M1312.77 575.603 L1312.77 1423.18 L1373.35 1423.18 L1373.35 575.603 L1312.77 575.603 L1312.77 575.603  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1312.77,575.603 1312.77,1423.18 1373.35,1423.18 1373.35,575.603 1312.77,575.603 "/>
+<path clip-path="url(#clip832)" d="M1312.77 999.392 L1312.77 1423.18 L1373.35 1423.18 L1373.35 999.392 L1312.77 999.392 L1312.77 999.392  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1312.77,999.392 1312.77,1423.18 1373.35,1423.18 1373.35,999.392 1312.77,999.392 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1343.06" cy="151.815" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1343.06" cy="575.603" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1343.06" cy="999.392" r="2"/>
+<path clip-path="url(#clip832)" d="M1413.74 193.093 L1413.74 1423.18 L1474.32 1423.18 L1474.32 193.093 L1413.74 193.093 L1413.74 193.093  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1413.74,193.093 1413.74,1423.18 1474.32,1423.18 1474.32,193.093 1413.74,193.093 "/>
+<path clip-path="url(#clip832)" d="M1413.74 603.122 L1413.74 1423.18 L1474.32 1423.18 L1474.32 603.122 L1413.74 603.122 L1413.74 603.122  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1413.74,603.122 1413.74,1423.18 1474.32,1423.18 1474.32,603.122 1413.74,603.122 "/>
+<path clip-path="url(#clip832)" d="M1413.74 1013.15 L1413.74 1423.18 L1474.32 1423.18 L1474.32 1013.15 L1413.74 1013.15 L1413.74 1013.15  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1413.74,1013.15 1413.74,1423.18 1474.32,1423.18 1474.32,1013.15 1413.74,1013.15 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1444.03" cy="193.093" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1444.03" cy="603.122" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1444.03" cy="1013.15" r="2"/>
+<path clip-path="url(#clip832)" d="M1514.71 316.928 L1514.71 1423.18 L1575.29 1423.18 L1575.29 316.928 L1514.71 316.928 L1514.71 316.928  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1514.71,316.928 1514.71,1423.18 1575.29,1423.18 1575.29,316.928 1514.71,316.928 "/>
+<path clip-path="url(#clip832)" d="M1514.71 593.491 L1514.71 1423.18 L1575.29 1423.18 L1575.29 593.491 L1514.71 593.491 L1514.71 593.491  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1514.71,593.491 1514.71,1423.18 1575.29,1423.18 1575.29,593.491 1514.71,593.491 "/>
+<path clip-path="url(#clip832)" d="M1514.71 870.054 L1514.71 1423.18 L1575.29 1423.18 L1575.29 870.054 L1514.71 870.054 L1514.71 870.054  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1514.71,870.054 1514.71,1423.18 1575.29,1423.18 1575.29,870.054 1514.71,870.054 "/>
+<path clip-path="url(#clip832)" d="M1514.71 1146.62 L1514.71 1423.18 L1575.29 1423.18 L1575.29 1146.62 L1514.71 1146.62 L1514.71 1146.62  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1514.71,1146.62 1514.71,1423.18 1575.29,1423.18 1575.29,1146.62 1514.71,1146.62 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1545" cy="316.928" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1545" cy="593.491" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1545" cy="870.054" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1545" cy="1146.62" r="2"/>
+<path clip-path="url(#clip832)" d="M1615.68 85.7703 L1615.68 1423.18 L1676.26 1423.18 L1676.26 85.7703 L1615.68 85.7703 L1615.68 85.7703  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1615.68,85.7703 1615.68,1423.18 1676.26,1423.18 1676.26,85.7703 1615.68,85.7703 "/>
+<path clip-path="url(#clip832)" d="M1615.68 308.672 L1615.68 1423.18 L1676.26 1423.18 L1676.26 308.672 L1615.68 308.672 L1615.68 308.672  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1615.68,308.672 1615.68,1423.18 1676.26,1423.18 1676.26,308.672 1615.68,308.672 "/>
+<path clip-path="url(#clip832)" d="M1615.68 531.573 L1615.68 1423.18 L1676.26 1423.18 L1676.26 531.573 L1615.68 531.573 L1615.68 531.573  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1615.68,531.573 1615.68,1423.18 1676.26,1423.18 1676.26,531.573 1615.68,531.573 "/>
+<path clip-path="url(#clip832)" d="M1615.68 754.475 L1615.68 1423.18 L1676.26 1423.18 L1676.26 754.475 L1615.68 754.475 L1615.68 754.475  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1615.68,754.475 1615.68,1423.18 1676.26,1423.18 1676.26,754.475 1615.68,754.475 "/>
+<path clip-path="url(#clip832)" d="M1615.68 977.377 L1615.68 1423.18 L1676.26 1423.18 L1676.26 977.377 L1615.68 977.377 L1615.68 977.377  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1615.68,977.377 1615.68,1423.18 1676.26,1423.18 1676.26,977.377 1615.68,977.377 "/>
+<path clip-path="url(#clip832)" d="M1615.68 1200.28 L1615.68 1423.18 L1676.26 1423.18 L1676.26 1200.28 L1615.68 1200.28 L1615.68 1200.28  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1615.68,1200.28 1615.68,1423.18 1676.26,1423.18 1676.26,1200.28 1615.68,1200.28 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="85.7703" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="308.672" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="531.573" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="754.475" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="977.377" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="1200.28" r="2"/>
+<path clip-path="url(#clip832)" d="M1716.65 226.116 L1716.65 1423.18 L1777.23 1423.18 L1777.23 226.116 L1716.65 226.116 L1716.65 226.116  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1716.65,226.116 1716.65,1423.18 1777.23,1423.18 1777.23,226.116 1716.65,226.116 "/>
+<path clip-path="url(#clip832)" d="M1716.65 425.626 L1716.65 1423.18 L1777.23 1423.18 L1777.23 425.626 L1716.65 425.626 L1716.65 425.626  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1716.65,425.626 1716.65,1423.18 1777.23,1423.18 1777.23,425.626 1716.65,425.626 "/>
+<path clip-path="url(#clip832)" d="M1716.65 625.137 L1716.65 1423.18 L1777.23 1423.18 L1777.23 625.137 L1716.65 625.137 L1716.65 625.137  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1716.65,625.137 1716.65,1423.18 1777.23,1423.18 1777.23,625.137 1716.65,625.137 "/>
+<path clip-path="url(#clip832)" d="M1716.65 824.648 L1716.65 1423.18 L1777.23 1423.18 L1777.23 824.648 L1716.65 824.648 L1716.65 824.648  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1716.65,824.648 1716.65,1423.18 1777.23,1423.18 1777.23,824.648 1716.65,824.648 "/>
+<path clip-path="url(#clip832)" d="M1716.65 1024.16 L1716.65 1423.18 L1777.23 1423.18 L1777.23 1024.16 L1716.65 1024.16 L1716.65 1024.16  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1716.65,1024.16 1716.65,1423.18 1777.23,1423.18 1777.23,1024.16 1716.65,1024.16 "/>
+<path clip-path="url(#clip832)" d="M1716.65 1223.67 L1716.65 1423.18 L1777.23 1423.18 L1777.23 1223.67 L1716.65 1223.67 L1716.65 1223.67  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1716.65,1223.67 1716.65,1423.18 1777.23,1423.18 1777.23,1223.67 1716.65,1223.67 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1746.94" cy="226.116" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1746.94" cy="425.626" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1746.94" cy="625.137" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1746.94" cy="824.648" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1746.94" cy="1024.16" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1746.94" cy="1223.67" r="2"/>
+<path clip-path="url(#clip832)" d="M1817.62 61.0035 L1817.62 1423.18 L1878.2 1423.18 L1878.2 61.0035 L1817.62 61.0035 L1817.62 61.0035  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,61.0035 1817.62,1423.18 1878.2,1423.18 1878.2,61.0035 1817.62,61.0035 "/>
+<path clip-path="url(#clip832)" d="M1817.62 212.356 L1817.62 1423.18 L1878.2 1423.18 L1878.2 212.356 L1817.62 212.356 L1817.62 212.356  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,212.356 1817.62,1423.18 1878.2,1423.18 1878.2,212.356 1817.62,212.356 "/>
+<path clip-path="url(#clip832)" d="M1817.62 363.709 L1817.62 1423.18 L1878.2 1423.18 L1878.2 363.709 L1817.62 363.709 L1817.62 363.709  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,363.709 1817.62,1423.18 1878.2,1423.18 1878.2,363.709 1817.62,363.709 "/>
+<path clip-path="url(#clip832)" d="M1817.62 515.062 L1817.62 1423.18 L1878.2 1423.18 L1878.2 515.062 L1817.62 515.062 L1817.62 515.062  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,515.062 1817.62,1423.18 1878.2,1423.18 1878.2,515.062 1817.62,515.062 "/>
+<path clip-path="url(#clip832)" d="M1817.62 666.415 L1817.62 1423.18 L1878.2 1423.18 L1878.2 666.415 L1817.62 666.415 L1817.62 666.415  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,666.415 1817.62,1423.18 1878.2,1423.18 1878.2,666.415 1817.62,666.415 "/>
+<path clip-path="url(#clip832)" d="M1817.62 817.768 L1817.62 1423.18 L1878.2 1423.18 L1878.2 817.768 L1817.62 817.768 L1817.62 817.768  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,817.768 1817.62,1423.18 1878.2,1423.18 1878.2,817.768 1817.62,817.768 "/>
+<path clip-path="url(#clip832)" d="M1817.62 969.121 L1817.62 1423.18 L1878.2 1423.18 L1878.2 969.121 L1817.62 969.121 L1817.62 969.121  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,969.121 1817.62,1423.18 1878.2,1423.18 1878.2,969.121 1817.62,969.121 "/>
+<path clip-path="url(#clip832)" d="M1817.62 1120.47 L1817.62 1423.18 L1878.2 1423.18 L1878.2 1120.47 L1817.62 1120.47 L1817.62 1120.47  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,1120.47 1817.62,1423.18 1878.2,1423.18 1878.2,1120.47 1817.62,1120.47 "/>
+<path clip-path="url(#clip832)" d="M1817.62 1271.83 L1817.62 1423.18 L1878.2 1423.18 L1878.2 1271.83 L1817.62 1271.83 L1817.62 1271.83  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1817.62,1271.83 1817.62,1423.18 1878.2,1423.18 1878.2,1271.83 1817.62,1271.83 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="61.0035" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="212.356" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="363.709" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="515.062" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="666.415" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="817.768" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="969.121" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="1120.47" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1847.91" cy="1271.83" r="2"/>
+<path clip-path="url(#clip832)" d="M1918.59 121.545 L1918.59 1423.18 L1979.17 1423.18 L1979.17 121.545 L1918.59 121.545 L1918.59 121.545  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,121.545 1918.59,1423.18 1979.17,1423.18 1979.17,121.545 1918.59,121.545 "/>
+<path clip-path="url(#clip832)" d="M1918.59 239.875 L1918.59 1423.18 L1979.17 1423.18 L1979.17 239.875 L1918.59 239.875 L1918.59 239.875  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,239.875 1918.59,1423.18 1979.17,1423.18 1979.17,239.875 1918.59,239.875 "/>
+<path clip-path="url(#clip832)" d="M1918.59 358.206 L1918.59 1423.18 L1979.17 1423.18 L1979.17 358.206 L1918.59 358.206 L1918.59 358.206  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,358.206 1918.59,1423.18 1979.17,1423.18 1979.17,358.206 1918.59,358.206 "/>
+<path clip-path="url(#clip832)" d="M1918.59 476.536 L1918.59 1423.18 L1979.17 1423.18 L1979.17 476.536 L1918.59 476.536 L1918.59 476.536  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,476.536 1918.59,1423.18 1979.17,1423.18 1979.17,476.536 1918.59,476.536 "/>
+<path clip-path="url(#clip832)" d="M1918.59 594.867 L1918.59 1423.18 L1979.17 1423.18 L1979.17 594.867 L1918.59 594.867 L1918.59 594.867  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,594.867 1918.59,1423.18 1979.17,1423.18 1979.17,594.867 1918.59,594.867 "/>
+<path clip-path="url(#clip832)" d="M1918.59 713.197 L1918.59 1423.18 L1979.17 1423.18 L1979.17 713.197 L1918.59 713.197 L1918.59 713.197  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,713.197 1918.59,1423.18 1979.17,1423.18 1979.17,713.197 1918.59,713.197 "/>
+<path clip-path="url(#clip832)" d="M1918.59 831.527 L1918.59 1423.18 L1979.17 1423.18 L1979.17 831.527 L1918.59 831.527 L1918.59 831.527  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,831.527 1918.59,1423.18 1979.17,1423.18 1979.17,831.527 1918.59,831.527 "/>
+<path clip-path="url(#clip832)" d="M1918.59 949.858 L1918.59 1423.18 L1979.17 1423.18 L1979.17 949.858 L1918.59 949.858 L1918.59 949.858  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,949.858 1918.59,1423.18 1979.17,1423.18 1979.17,949.858 1918.59,949.858 "/>
+<path clip-path="url(#clip832)" d="M1918.59 1068.19 L1918.59 1423.18 L1979.17 1423.18 L1979.17 1068.19 L1918.59 1068.19 L1918.59 1068.19  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,1068.19 1918.59,1423.18 1979.17,1423.18 1979.17,1068.19 1918.59,1068.19 "/>
+<path clip-path="url(#clip832)" d="M1918.59 1186.52 L1918.59 1423.18 L1979.17 1423.18 L1979.17 1186.52 L1918.59 1186.52 L1918.59 1186.52  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,1186.52 1918.59,1423.18 1979.17,1423.18 1979.17,1186.52 1918.59,1186.52 "/>
+<path clip-path="url(#clip832)" d="M1918.59 1304.85 L1918.59 1423.18 L1979.17 1423.18 L1979.17 1304.85 L1918.59 1304.85 L1918.59 1304.85  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1918.59,1304.85 1918.59,1423.18 1979.17,1423.18 1979.17,1304.85 1918.59,1304.85 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="121.545" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="239.875" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="358.206" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="476.536" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="594.867" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="713.197" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="831.527" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="949.858" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="1068.19" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="1186.52" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1948.88" cy="1304.85" r="2"/>
+<path clip-path="url(#clip832)" d="M2019.56 69.2591 L2019.56 1423.18 L2080.14 1423.18 L2080.14 69.2591 L2019.56 69.2591 L2019.56 69.2591  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,69.2591 2019.56,1423.18 2080.14,1423.18 2080.14,69.2591 2019.56,69.2591 "/>
+<path clip-path="url(#clip832)" d="M2019.56 182.086 L2019.56 1423.18 L2080.14 1423.18 L2080.14 182.086 L2019.56 182.086 L2019.56 182.086  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,182.086 2019.56,1423.18 2080.14,1423.18 2080.14,182.086 2019.56,182.086 "/>
+<path clip-path="url(#clip832)" d="M2019.56 294.913 L2019.56 1423.18 L2080.14 1423.18 L2080.14 294.913 L2019.56 294.913 L2019.56 294.913  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,294.913 2019.56,1423.18 2080.14,1423.18 2080.14,294.913 2019.56,294.913 "/>
+<path clip-path="url(#clip832)" d="M2019.56 407.739 L2019.56 1423.18 L2080.14 1423.18 L2080.14 407.739 L2019.56 407.739 L2019.56 407.739  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,407.739 2019.56,1423.18 2080.14,1423.18 2080.14,407.739 2019.56,407.739 "/>
+<path clip-path="url(#clip832)" d="M2019.56 520.566 L2019.56 1423.18 L2080.14 1423.18 L2080.14 520.566 L2019.56 520.566 L2019.56 520.566  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,520.566 2019.56,1423.18 2080.14,1423.18 2080.14,520.566 2019.56,520.566 "/>
+<path clip-path="url(#clip832)" d="M2019.56 633.393 L2019.56 1423.18 L2080.14 1423.18 L2080.14 633.393 L2019.56 633.393 L2019.56 633.393  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,633.393 2019.56,1423.18 2080.14,1423.18 2080.14,633.393 2019.56,633.393 "/>
+<path clip-path="url(#clip832)" d="M2019.56 746.219 L2019.56 1423.18 L2080.14 1423.18 L2080.14 746.219 L2019.56 746.219 L2019.56 746.219  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,746.219 2019.56,1423.18 2080.14,1423.18 2080.14,746.219 2019.56,746.219 "/>
+<path clip-path="url(#clip832)" d="M2019.56 859.046 L2019.56 1423.18 L2080.14 1423.18 L2080.14 859.046 L2019.56 859.046 L2019.56 859.046  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,859.046 2019.56,1423.18 2080.14,1423.18 2080.14,859.046 2019.56,859.046 "/>
+<path clip-path="url(#clip832)" d="M2019.56 971.873 L2019.56 1423.18 L2080.14 1423.18 L2080.14 971.873 L2019.56 971.873 L2019.56 971.873  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,971.873 2019.56,1423.18 2080.14,1423.18 2080.14,971.873 2019.56,971.873 "/>
+<path clip-path="url(#clip832)" d="M2019.56 1084.7 L2019.56 1423.18 L2080.14 1423.18 L2080.14 1084.7 L2019.56 1084.7 L2019.56 1084.7  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,1084.7 2019.56,1423.18 2080.14,1423.18 2080.14,1084.7 2019.56,1084.7 "/>
+<path clip-path="url(#clip832)" d="M2019.56 1197.53 L2019.56 1423.18 L2080.14 1423.18 L2080.14 1197.53 L2019.56 1197.53 L2019.56 1197.53  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,1197.53 2019.56,1423.18 2080.14,1423.18 2080.14,1197.53 2019.56,1197.53 "/>
+<path clip-path="url(#clip832)" d="M2019.56 1310.35 L2019.56 1423.18 L2080.14 1423.18 L2080.14 1310.35 L2019.56 1310.35 L2019.56 1310.35  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2019.56,1310.35 2019.56,1423.18 2080.14,1423.18 2080.14,1310.35 2019.56,1310.35 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="69.2591" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="182.086" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="294.913" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="407.739" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="520.566" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="633.393" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="746.219" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="859.046" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="971.873" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="1084.7" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="1197.53" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2049.85" cy="1310.35" r="2"/>
+<path clip-path="url(#clip832)" d="M2120.53 61.0035 L2120.53 1423.18 L2181.11 1423.18 L2181.11 61.0035 L2120.53 61.0035 L2120.53 61.0035  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,61.0035 2120.53,1423.18 2181.11,1423.18 2181.11,61.0035 2120.53,61.0035 "/>
+<path clip-path="url(#clip832)" d="M2120.53 151.815 L2120.53 1423.18 L2181.11 1423.18 L2181.11 151.815 L2120.53 151.815 L2120.53 151.815  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,151.815 2120.53,1423.18 2181.11,1423.18 2181.11,151.815 2120.53,151.815 "/>
+<path clip-path="url(#clip832)" d="M2120.53 242.627 L2120.53 1423.18 L2181.11 1423.18 L2181.11 242.627 L2120.53 242.627 L2120.53 242.627  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,242.627 2120.53,1423.18 2181.11,1423.18 2181.11,242.627 2120.53,242.627 "/>
+<path clip-path="url(#clip832)" d="M2120.53 333.439 L2120.53 1423.18 L2181.11 1423.18 L2181.11 333.439 L2120.53 333.439 L2120.53 333.439  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,333.439 2120.53,1423.18 2181.11,1423.18 2181.11,333.439 2120.53,333.439 "/>
+<path clip-path="url(#clip832)" d="M2120.53 424.25 L2120.53 1423.18 L2181.11 1423.18 L2181.11 424.25 L2120.53 424.25 L2120.53 424.25  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,424.25 2120.53,1423.18 2181.11,1423.18 2181.11,424.25 2120.53,424.25 "/>
+<path clip-path="url(#clip832)" d="M2120.53 515.062 L2120.53 1423.18 L2181.11 1423.18 L2181.11 515.062 L2120.53 515.062 L2120.53 515.062  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,515.062 2120.53,1423.18 2181.11,1423.18 2181.11,515.062 2120.53,515.062 "/>
+<path clip-path="url(#clip832)" d="M2120.53 605.874 L2120.53 1423.18 L2181.11 1423.18 L2181.11 605.874 L2120.53 605.874 L2120.53 605.874  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,605.874 2120.53,1423.18 2181.11,1423.18 2181.11,605.874 2120.53,605.874 "/>
+<path clip-path="url(#clip832)" d="M2120.53 696.686 L2120.53 1423.18 L2181.11 1423.18 L2181.11 696.686 L2120.53 696.686 L2120.53 696.686  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,696.686 2120.53,1423.18 2181.11,1423.18 2181.11,696.686 2120.53,696.686 "/>
+<path clip-path="url(#clip832)" d="M2120.53 787.498 L2120.53 1423.18 L2181.11 1423.18 L2181.11 787.498 L2120.53 787.498 L2120.53 787.498  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,787.498 2120.53,1423.18 2181.11,1423.18 2181.11,787.498 2120.53,787.498 "/>
+<path clip-path="url(#clip832)" d="M2120.53 878.309 L2120.53 1423.18 L2181.11 1423.18 L2181.11 878.309 L2120.53 878.309 L2120.53 878.309  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,878.309 2120.53,1423.18 2181.11,1423.18 2181.11,878.309 2120.53,878.309 "/>
+<path clip-path="url(#clip832)" d="M2120.53 969.121 L2120.53 1423.18 L2181.11 1423.18 L2181.11 969.121 L2120.53 969.121 L2120.53 969.121  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,969.121 2120.53,1423.18 2181.11,1423.18 2181.11,969.121 2120.53,969.121 "/>
+<path clip-path="url(#clip832)" d="M2120.53 1059.93 L2120.53 1423.18 L2181.11 1423.18 L2181.11 1059.93 L2120.53 1059.93 L2120.53 1059.93  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,1059.93 2120.53,1423.18 2181.11,1423.18 2181.11,1059.93 2120.53,1059.93 "/>
+<path clip-path="url(#clip832)" d="M2120.53 1150.74 L2120.53 1423.18 L2181.11 1423.18 L2181.11 1150.74 L2120.53 1150.74 L2120.53 1150.74  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,1150.74 2120.53,1423.18 2181.11,1423.18 2181.11,1150.74 2120.53,1150.74 "/>
+<path clip-path="url(#clip832)" d="M2120.53 1241.56 L2120.53 1423.18 L2181.11 1423.18 L2181.11 1241.56 L2120.53 1241.56 L2120.53 1241.56  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,1241.56 2120.53,1423.18 2181.11,1423.18 2181.11,1241.56 2120.53,1241.56 "/>
+<path clip-path="url(#clip832)" d="M2120.53 1332.37 L2120.53 1423.18 L2181.11 1423.18 L2181.11 1332.37 L2120.53 1332.37 L2120.53 1332.37  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.53,1332.37 2120.53,1423.18 2181.11,1423.18 2181.11,1332.37 2120.53,1332.37 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="61.0035" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="151.815" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="242.627" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="333.439" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="424.25" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="515.062" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="605.874" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="696.686" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="787.498" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="878.309" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="969.121" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="1059.93" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="1150.74" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="1241.56" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2150.82" cy="1332.37" r="2"/>
+<path clip-path="url(#clip832)" d="M2221.5 89.8981 L2221.5 1423.18 L2282.08 1423.18 L2282.08 89.8981 L2221.5 89.8981 L2221.5 89.8981  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,89.8981 2221.5,1423.18 2282.08,1423.18 2282.08,89.8981 2221.5,89.8981 "/>
+<path clip-path="url(#clip832)" d="M2221.5 160.071 L2221.5 1423.18 L2282.08 1423.18 L2282.08 160.071 L2221.5 160.071 L2221.5 160.071  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,160.071 2221.5,1423.18 2282.08,1423.18 2282.08,160.071 2221.5,160.071 "/>
+<path clip-path="url(#clip832)" d="M2221.5 230.244 L2221.5 1423.18 L2282.08 1423.18 L2282.08 230.244 L2221.5 230.244 L2221.5 230.244  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,230.244 2221.5,1423.18 2282.08,1423.18 2282.08,230.244 2221.5,230.244 "/>
+<path clip-path="url(#clip832)" d="M2221.5 300.416 L2221.5 1423.18 L2282.08 1423.18 L2282.08 300.416 L2221.5 300.416 L2221.5 300.416  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,300.416 2221.5,1423.18 2282.08,1423.18 2282.08,300.416 2221.5,300.416 "/>
+<path clip-path="url(#clip832)" d="M2221.5 370.589 L2221.5 1423.18 L2282.08 1423.18 L2282.08 370.589 L2221.5 370.589 L2221.5 370.589  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,370.589 2221.5,1423.18 2282.08,1423.18 2282.08,370.589 2221.5,370.589 "/>
+<path clip-path="url(#clip832)" d="M2221.5 440.762 L2221.5 1423.18 L2282.08 1423.18 L2282.08 440.762 L2221.5 440.762 L2221.5 440.762  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,440.762 2221.5,1423.18 2282.08,1423.18 2282.08,440.762 2221.5,440.762 "/>
+<path clip-path="url(#clip832)" d="M2221.5 510.934 L2221.5 1423.18 L2282.08 1423.18 L2282.08 510.934 L2221.5 510.934 L2221.5 510.934  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,510.934 2221.5,1423.18 2282.08,1423.18 2282.08,510.934 2221.5,510.934 "/>
+<path clip-path="url(#clip832)" d="M2221.5 581.107 L2221.5 1423.18 L2282.08 1423.18 L2282.08 581.107 L2221.5 581.107 L2221.5 581.107  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,581.107 2221.5,1423.18 2282.08,1423.18 2282.08,581.107 2221.5,581.107 "/>
+<path clip-path="url(#clip832)" d="M2221.5 651.28 L2221.5 1423.18 L2282.08 1423.18 L2282.08 651.28 L2221.5 651.28 L2221.5 651.28  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,651.28 2221.5,1423.18 2282.08,1423.18 2282.08,651.28 2221.5,651.28 "/>
+<path clip-path="url(#clip832)" d="M2221.5 721.453 L2221.5 1423.18 L2282.08 1423.18 L2282.08 721.453 L2221.5 721.453 L2221.5 721.453  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,721.453 2221.5,1423.18 2282.08,1423.18 2282.08,721.453 2221.5,721.453 "/>
+<path clip-path="url(#clip832)" d="M2221.5 791.625 L2221.5 1423.18 L2282.08 1423.18 L2282.08 791.625 L2221.5 791.625 L2221.5 791.625  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,791.625 2221.5,1423.18 2282.08,1423.18 2282.08,791.625 2221.5,791.625 "/>
+<path clip-path="url(#clip832)" d="M2221.5 861.798 L2221.5 1423.18 L2282.08 1423.18 L2282.08 861.798 L2221.5 861.798 L2221.5 861.798  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,861.798 2221.5,1423.18 2282.08,1423.18 2282.08,861.798 2221.5,861.798 "/>
+<path clip-path="url(#clip832)" d="M2221.5 931.971 L2221.5 1423.18 L2282.08 1423.18 L2282.08 931.971 L2221.5 931.971 L2221.5 931.971  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,931.971 2221.5,1423.18 2282.08,1423.18 2282.08,931.971 2221.5,931.971 "/>
+<path clip-path="url(#clip832)" d="M2221.5 1002.14 L2221.5 1423.18 L2282.08 1423.18 L2282.08 1002.14 L2221.5 1002.14 L2221.5 1002.14  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,1002.14 2221.5,1423.18 2282.08,1423.18 2282.08,1002.14 2221.5,1002.14 "/>
+<path clip-path="url(#clip832)" d="M2221.5 1072.32 L2221.5 1423.18 L2282.08 1423.18 L2282.08 1072.32 L2221.5 1072.32 L2221.5 1072.32  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,1072.32 2221.5,1423.18 2282.08,1423.18 2282.08,1072.32 2221.5,1072.32 "/>
+<path clip-path="url(#clip832)" d="M2221.5 1142.49 L2221.5 1423.18 L2282.08 1423.18 L2282.08 1142.49 L2221.5 1142.49 L2221.5 1142.49  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,1142.49 2221.5,1423.18 2282.08,1423.18 2282.08,1142.49 2221.5,1142.49 "/>
+<path clip-path="url(#clip832)" d="M2221.5 1212.66 L2221.5 1423.18 L2282.08 1423.18 L2282.08 1212.66 L2221.5 1212.66 L2221.5 1212.66  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,1212.66 2221.5,1423.18 2282.08,1423.18 2282.08,1212.66 2221.5,1212.66 "/>
+<path clip-path="url(#clip832)" d="M2221.5 1282.83 L2221.5 1423.18 L2282.08 1423.18 L2282.08 1282.83 L2221.5 1282.83 L2221.5 1282.83  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,1282.83 2221.5,1423.18 2282.08,1423.18 2282.08,1282.83 2221.5,1282.83 "/>
+<path clip-path="url(#clip832)" d="M2221.5 1353.01 L2221.5 1423.18 L2282.08 1423.18 L2282.08 1353.01 L2221.5 1353.01 L2221.5 1353.01  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip832)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2221.5,1353.01 2221.5,1423.18 2282.08,1423.18 2282.08,1353.01 2221.5,1353.01 "/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="89.8981" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="160.071" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="230.244" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="300.416" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="370.589" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="440.762" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="510.934" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="581.107" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="651.28" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="721.453" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="791.625" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="861.798" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="931.971" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="1002.14" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="1072.32" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="1142.49" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="1212.66" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="1282.83" r="2"/>
+<circle clip-path="url(#clip832)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2251.79" cy="1353.01" r="2"/>
+</svg>
diff --git a/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/fa9bdba4.svg b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/fa9bdba4.svg
new file mode 100644
index 00000000000..d49c79c1d8f
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/fa9bdba4.svg
@@ -0,0 +1,516 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip890">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip890)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip891">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip890)" d="M232.389 1423.18 L2352.76 1423.18 L2352.76 47.2441 L232.389 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip892">
+    <rect x="232" y="47" width="2121" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,1423.18 232.389,47.2441 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="821.38,1423.18 821.38,47.2441 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1410.37,1423.18 1410.37,47.2441 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1999.36,1423.18 1999.36,47.2441 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 232.389,1404.28 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="821.38,1423.18 821.38,1404.28 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1410.37,1423.18 1410.37,1404.28 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1999.36,1423.18 1999.36,1404.28 "/>
+<path clip-path="url(#clip890)" d="M232.389 1454.1 Q228.778 1454.1 226.95 1457.66 Q225.144 1461.2 225.144 1468.33 Q225.144 1475.44 226.95 1479.01 Q228.778 1482.55 232.389 1482.55 Q236.024 1482.55 237.829 1479.01 Q239.658 1475.44 239.658 1468.33 Q239.658 1461.2 237.829 1457.66 Q236.024 1454.1 232.389 1454.1 M232.389 1450.39 Q238.2 1450.39 241.255 1455 Q244.334 1459.58 244.334 1468.33 Q244.334 1477.06 241.255 1481.67 Q238.2 1486.25 232.389 1486.25 Q226.579 1486.25 223.501 1481.67 Q220.445 1477.06 220.445 1468.33 Q220.445 1459.58 223.501 1455 Q226.579 1450.39 232.389 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M796.068 1481.64 L803.707 1481.64 L803.707 1455.28 L795.396 1456.95 L795.396 1452.69 L803.66 1451.02 L808.336 1451.02 L808.336 1481.64 L815.975 1481.64 L815.975 1485.58 L796.068 1485.58 L796.068 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M835.419 1454.1 Q831.808 1454.1 829.98 1457.66 Q828.174 1461.2 828.174 1468.33 Q828.174 1475.44 829.98 1479.01 Q831.808 1482.55 835.419 1482.55 Q839.054 1482.55 840.859 1479.01 Q842.688 1475.44 842.688 1468.33 Q842.688 1461.2 840.859 1457.66 Q839.054 1454.1 835.419 1454.1 M835.419 1450.39 Q841.229 1450.39 844.285 1455 Q847.364 1459.58 847.364 1468.33 Q847.364 1477.06 844.285 1481.67 Q841.229 1486.25 835.419 1486.25 Q829.609 1486.25 826.53 1481.67 Q823.475 1477.06 823.475 1468.33 Q823.475 1459.58 826.53 1455 Q829.609 1450.39 835.419 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1389.14 1481.64 L1405.46 1481.64 L1405.46 1485.58 L1383.52 1485.58 L1383.52 1481.64 Q1386.18 1478.89 1390.76 1474.26 Q1395.37 1469.61 1396.55 1468.27 Q1398.8 1465.74 1399.68 1464.01 Q1400.58 1462.25 1400.58 1460.56 Q1400.58 1457.8 1398.63 1456.07 Q1396.71 1454.33 1393.61 1454.33 Q1391.41 1454.33 1388.96 1455.09 Q1386.53 1455.86 1383.75 1457.41 L1383.75 1452.69 Q1386.57 1451.55 1389.03 1450.97 Q1391.48 1450.39 1393.52 1450.39 Q1398.89 1450.39 1402.08 1453.08 Q1405.28 1455.77 1405.28 1460.26 Q1405.28 1462.39 1404.47 1464.31 Q1403.68 1466.2 1401.57 1468.8 Q1401 1469.47 1397.89 1472.69 Q1394.79 1475.88 1389.14 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1425.28 1454.1 Q1421.67 1454.1 1419.84 1457.66 Q1418.03 1461.2 1418.03 1468.33 Q1418.03 1475.44 1419.84 1479.01 Q1421.67 1482.55 1425.28 1482.55 Q1428.91 1482.55 1430.72 1479.01 Q1432.55 1475.44 1432.55 1468.33 Q1432.55 1461.2 1430.72 1457.66 Q1428.91 1454.1 1425.28 1454.1 M1425.28 1450.39 Q1431.09 1450.39 1434.14 1455 Q1437.22 1459.58 1437.22 1468.33 Q1437.22 1477.06 1434.14 1481.67 Q1431.09 1486.25 1425.28 1486.25 Q1419.47 1486.25 1416.39 1481.67 Q1413.33 1477.06 1413.33 1468.33 Q1413.33 1459.58 1416.39 1455 Q1419.47 1450.39 1425.28 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1988.2 1466.95 Q1991.56 1467.66 1993.44 1469.93 Q1995.33 1472.2 1995.33 1475.53 Q1995.33 1480.65 1991.82 1483.45 Q1988.3 1486.25 1981.82 1486.25 Q1979.64 1486.25 1977.32 1485.81 Q1975.03 1485.39 1972.58 1484.54 L1972.58 1480.02 Q1974.52 1481.16 1976.84 1481.74 Q1979.15 1482.32 1981.68 1482.32 Q1986.07 1482.32 1988.37 1480.58 Q1990.68 1478.84 1990.68 1475.53 Q1990.68 1472.48 1988.53 1470.77 Q1986.4 1469.03 1982.58 1469.03 L1978.55 1469.03 L1978.55 1465.19 L1982.76 1465.19 Q1986.21 1465.19 1988.04 1463.82 Q1989.87 1462.43 1989.87 1459.84 Q1989.87 1457.18 1987.97 1455.77 Q1986.1 1454.33 1982.58 1454.33 Q1980.66 1454.33 1978.46 1454.75 Q1976.26 1455.16 1973.62 1456.04 L1973.62 1451.88 Q1976.28 1451.14 1978.6 1450.77 Q1980.94 1450.39 1983 1450.39 Q1988.32 1450.39 1991.42 1452.83 Q1994.52 1455.23 1994.52 1459.35 Q1994.52 1462.22 1992.88 1464.21 Q1991.24 1466.18 1988.2 1466.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M2014.2 1454.1 Q2010.59 1454.1 2008.76 1457.66 Q2006.95 1461.2 2006.95 1468.33 Q2006.95 1475.44 2008.76 1479.01 Q2010.59 1482.55 2014.2 1482.55 Q2017.83 1482.55 2019.64 1479.01 Q2021.47 1475.44 2021.47 1468.33 Q2021.47 1461.2 2019.64 1457.66 Q2017.83 1454.1 2014.2 1454.1 M2014.2 1450.39 Q2020.01 1450.39 2023.07 1455 Q2026.14 1459.58 2026.14 1468.33 Q2026.14 1477.06 2023.07 1481.67 Q2020.01 1486.25 2014.2 1486.25 Q2008.39 1486.25 2005.31 1481.67 Q2002.25 1477.06 2002.25 1468.33 Q2002.25 1459.58 2005.31 1455 Q2008.39 1450.39 2014.2 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1187.71 1525.81 L1187.71 1543.66 L1195.8 1543.66 Q1200.29 1543.66 1202.74 1541.34 Q1205.19 1539.02 1205.19 1534.72 Q1205.19 1530.45 1202.74 1528.13 Q1200.29 1525.81 1195.8 1525.81 L1187.71 1525.81 M1181.28 1520.52 L1195.8 1520.52 Q1203.79 1520.52 1207.86 1524.15 Q1211.97 1527.75 1211.97 1534.72 Q1211.97 1541.75 1207.86 1545.35 Q1203.79 1548.95 1195.8 1548.95 L1187.71 1548.95 L1187.71 1568.04 L1181.28 1568.04 L1181.28 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1233.61 1550.12 Q1226.51 1550.12 1223.78 1551.75 Q1221.04 1553.37 1221.04 1557.29 Q1221.04 1560.4 1223.08 1562.25 Q1225.14 1564.07 1228.68 1564.07 Q1233.55 1564.07 1236.47 1560.63 Q1239.43 1557.16 1239.43 1551.43 L1239.43 1550.12 L1233.61 1550.12 M1245.29 1547.71 L1245.29 1568.04 L1239.43 1568.04 L1239.43 1562.63 Q1237.43 1565.88 1234.44 1567.44 Q1231.45 1568.97 1227.12 1568.97 Q1221.64 1568.97 1218.4 1565.91 Q1215.18 1562.82 1215.18 1557.67 Q1215.18 1551.65 1219.19 1548.6 Q1223.23 1545.54 1231.22 1545.54 L1239.43 1545.54 L1239.43 1544.97 Q1239.43 1540.93 1236.76 1538.73 Q1234.12 1536.5 1229.31 1536.5 Q1226.26 1536.5 1223.36 1537.23 Q1220.47 1537.97 1217.79 1539.43 L1217.79 1534.02 Q1221.01 1532.78 1224.03 1532.17 Q1227.05 1531.54 1229.92 1531.54 Q1237.65 1531.54 1241.47 1535.55 Q1245.29 1539.56 1245.29 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1263.15 1522.27 L1263.15 1532.4 L1275.21 1532.4 L1275.21 1536.95 L1263.15 1536.95 L1263.15 1556.3 Q1263.15 1560.66 1264.32 1561.9 Q1265.53 1563.14 1269.19 1563.14 L1275.21 1563.14 L1275.21 1568.04 L1269.19 1568.04 Q1262.42 1568.04 1259.84 1565.53 Q1257.26 1562.98 1257.26 1556.3 L1257.26 1536.95 L1252.96 1536.95 L1252.96 1532.4 L1257.26 1532.4 L1257.26 1522.27 L1263.15 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1288.71 1522.27 L1288.71 1532.4 L1300.77 1532.4 L1300.77 1536.95 L1288.71 1536.95 L1288.71 1556.3 Q1288.71 1560.66 1289.88 1561.9 Q1291.09 1563.14 1294.75 1563.14 L1300.77 1563.14 L1300.77 1568.04 L1294.75 1568.04 Q1287.97 1568.04 1285.4 1565.53 Q1282.82 1562.98 1282.82 1556.3 L1282.82 1536.95 L1278.52 1536.95 L1278.52 1532.4 L1282.82 1532.4 L1282.82 1522.27 L1288.71 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1338.96 1548.76 L1338.96 1551.62 L1312.04 1551.62 Q1312.42 1557.67 1315.66 1560.85 Q1318.94 1564 1324.77 1564 Q1328.14 1564 1331.29 1563.17 Q1334.47 1562.35 1337.59 1560.69 L1337.59 1566.23 Q1334.44 1567.57 1331.13 1568.27 Q1327.82 1568.97 1324.42 1568.97 Q1315.89 1568.97 1310.89 1564 Q1305.92 1559.04 1305.92 1550.57 Q1305.92 1541.82 1310.64 1536.69 Q1315.38 1531.54 1323.4 1531.54 Q1330.59 1531.54 1334.76 1536.18 Q1338.96 1540.8 1338.96 1548.76 M1333.11 1547.04 Q1333.04 1542.23 1330.4 1539.37 Q1327.79 1536.5 1323.46 1536.5 Q1318.56 1536.5 1315.6 1539.27 Q1312.67 1542.04 1312.23 1547.07 L1333.11 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1369.23 1537.87 Q1368.24 1537.3 1367.07 1537.04 Q1365.92 1536.76 1364.52 1536.76 Q1359.56 1536.76 1356.88 1540 Q1354.24 1543.22 1354.24 1549.27 L1354.24 1568.04 L1348.35 1568.04 L1348.35 1532.4 L1354.24 1532.4 L1354.24 1537.93 Q1356.09 1534.69 1359.05 1533.13 Q1362.01 1531.54 1366.24 1531.54 Q1366.84 1531.54 1367.58 1531.63 Q1368.31 1531.7 1369.2 1531.85 L1369.23 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M1403.86 1546.53 L1403.86 1568.04 L1398 1568.04 L1398 1546.72 Q1398 1541.66 1396.03 1539.14 Q1394.06 1536.63 1390.11 1536.63 Q1385.37 1536.63 1382.63 1539.65 Q1379.89 1542.68 1379.89 1547.9 L1379.89 1568.04 L1374.01 1568.04 L1374.01 1532.4 L1379.89 1532.4 L1379.89 1537.93 Q1381.99 1534.72 1384.83 1533.13 Q1387.69 1531.54 1391.42 1531.54 Q1397.56 1531.54 1400.71 1535.36 Q1403.86 1539.14 1403.86 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,1147.99 2352.76,1147.99 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,872.806 2352.76,872.806 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,597.618 2352.76,597.618 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,322.431 2352.76,322.431 "/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="232.389,47.2441 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 232.389,47.2441 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1423.18 251.287,1423.18 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,1147.99 251.287,1147.99 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,872.806 251.287,872.806 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,597.618 251.287,597.618 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,322.431 251.287,322.431 "/>
+<polyline clip-path="url(#clip890)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="232.389,47.2441 251.287,47.2441 "/>
+<path clip-path="url(#clip890)" d="M184.445 1408.98 Q180.834 1408.98 179.005 1412.54 Q177.2 1416.08 177.2 1423.21 Q177.2 1430.32 179.005 1433.89 Q180.834 1437.43 184.445 1437.43 Q188.079 1437.43 189.885 1433.89 Q191.713 1430.32 191.713 1423.21 Q191.713 1416.08 189.885 1412.54 Q188.079 1408.98 184.445 1408.98 M184.445 1405.27 Q190.255 1405.27 193.311 1409.88 Q196.389 1414.46 196.389 1423.21 Q196.389 1431.94 193.311 1436.55 Q190.255 1441.13 184.445 1441.13 Q178.635 1441.13 175.556 1436.55 Q172.501 1431.94 172.501 1423.21 Q172.501 1414.46 175.556 1409.88 Q178.635 1405.27 184.445 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M148.311 1161.34 L164.63 1161.34 L164.63 1165.27 L142.686 1165.27 L142.686 1161.34 Q145.348 1158.58 149.931 1153.95 Q154.538 1149.3 155.718 1147.96 Q157.964 1145.43 158.843 1143.7 Q159.746 1141.94 159.746 1140.25 Q159.746 1137.5 157.802 1135.76 Q155.88 1134.02 152.778 1134.02 Q150.579 1134.02 148.126 1134.79 Q145.695 1135.55 142.917 1137.1 L142.917 1132.38 Q145.741 1131.25 148.195 1130.67 Q150.649 1130.09 152.686 1130.09 Q158.056 1130.09 161.251 1132.77 Q164.445 1135.46 164.445 1139.95 Q164.445 1142.08 163.635 1144 Q162.848 1145.9 160.741 1148.49 Q160.163 1149.16 157.061 1152.38 Q153.959 1155.57 148.311 1161.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M184.445 1133.79 Q180.834 1133.79 179.005 1137.36 Q177.2 1140.9 177.2 1148.03 Q177.2 1155.13 179.005 1158.7 Q180.834 1162.24 184.445 1162.24 Q188.079 1162.24 189.885 1158.7 Q191.713 1155.13 191.713 1148.03 Q191.713 1140.9 189.885 1137.36 Q188.079 1133.79 184.445 1133.79 M184.445 1130.09 Q190.255 1130.09 193.311 1134.69 Q196.389 1139.28 196.389 1148.03 Q196.389 1156.75 193.311 1161.36 Q190.255 1165.94 184.445 1165.94 Q178.635 1165.94 175.556 1161.36 Q172.501 1156.75 172.501 1148.03 Q172.501 1139.28 175.556 1134.69 Q178.635 1130.09 184.445 1130.09 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M157.13 859.6 L145.325 878.049 L157.13 878.049 L157.13 859.6 M155.903 855.526 L161.783 855.526 L161.783 878.049 L166.714 878.049 L166.714 881.937 L161.783 881.937 L161.783 890.086 L157.13 890.086 L157.13 881.937 L141.529 881.937 L141.529 877.424 L155.903 855.526 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M184.445 858.604 Q180.834 858.604 179.005 862.169 Q177.2 865.711 177.2 872.84 Q177.2 879.947 179.005 883.512 Q180.834 887.053 184.445 887.053 Q188.079 887.053 189.885 883.512 Q191.713 879.947 191.713 872.84 Q191.713 865.711 189.885 862.169 Q188.079 858.604 184.445 858.604 M184.445 854.901 Q190.255 854.901 193.311 859.507 Q196.389 864.09 196.389 872.84 Q196.389 881.567 193.311 886.174 Q190.255 890.757 184.445 890.757 Q178.635 890.757 175.556 886.174 Q172.501 881.567 172.501 872.84 Q172.501 864.09 175.556 859.507 Q178.635 854.901 184.445 854.901 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M154.862 595.755 Q151.714 595.755 149.862 597.908 Q148.033 600.061 148.033 603.81 Q148.033 607.537 149.862 609.713 Q151.714 611.866 154.862 611.866 Q158.01 611.866 159.839 609.713 Q161.69 607.537 161.69 603.81 Q161.69 600.061 159.839 597.908 Q158.01 595.755 154.862 595.755 M164.144 581.102 L164.144 585.362 Q162.385 584.528 160.579 584.088 Q158.797 583.649 157.038 583.649 Q152.408 583.649 149.954 586.774 Q147.524 589.899 147.177 596.218 Q148.542 594.204 150.603 593.139 Q152.663 592.051 155.14 592.051 Q160.348 592.051 163.357 595.223 Q166.39 598.371 166.39 603.81 Q166.39 609.135 163.241 612.352 Q160.093 615.57 154.862 615.57 Q148.866 615.57 145.695 610.986 Q142.524 606.38 142.524 597.653 Q142.524 589.459 146.413 584.598 Q150.302 579.713 156.853 579.713 Q158.612 579.713 160.394 580.061 Q162.2 580.408 164.144 581.102 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M184.445 583.417 Q180.834 583.417 179.005 586.982 Q177.2 590.524 177.2 597.653 Q177.2 604.76 179.005 608.324 Q180.834 611.866 184.445 611.866 Q188.079 611.866 189.885 608.324 Q191.713 604.76 191.713 597.653 Q191.713 590.524 189.885 586.982 Q188.079 583.417 184.445 583.417 M184.445 579.713 Q190.255 579.713 193.311 584.32 Q196.389 588.903 196.389 597.653 Q196.389 606.38 193.311 610.986 Q190.255 615.57 184.445 615.57 Q178.635 615.57 175.556 610.986 Q172.501 606.38 172.501 597.653 Q172.501 588.903 175.556 584.32 Q178.635 579.713 184.445 579.713 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M154.283 323.299 Q150.95 323.299 149.028 325.082 Q147.13 326.864 147.13 329.989 Q147.13 333.114 149.028 334.896 Q150.95 336.679 154.283 336.679 Q157.616 336.679 159.538 334.896 Q161.459 333.091 161.459 329.989 Q161.459 326.864 159.538 325.082 Q157.64 323.299 154.283 323.299 M149.607 321.309 Q146.598 320.568 144.908 318.508 Q143.241 316.447 143.241 313.485 Q143.241 309.341 146.181 306.934 Q149.144 304.526 154.283 304.526 Q159.445 304.526 162.385 306.934 Q165.325 309.341 165.325 313.485 Q165.325 316.447 163.635 318.508 Q161.968 320.568 158.982 321.309 Q162.362 322.096 164.237 324.387 Q166.135 326.679 166.135 329.989 Q166.135 335.012 163.056 337.697 Q160.001 340.383 154.283 340.383 Q148.566 340.383 145.487 337.697 Q142.431 335.012 142.431 329.989 Q142.431 326.679 144.329 324.387 Q146.228 322.096 149.607 321.309 M147.894 313.924 Q147.894 316.61 149.561 318.114 Q151.251 319.619 154.283 319.619 Q157.292 319.619 158.982 318.114 Q160.695 316.61 160.695 313.924 Q160.695 311.239 158.982 309.735 Q157.292 308.23 154.283 308.23 Q151.251 308.23 149.561 309.735 Q147.894 311.239 147.894 313.924 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M184.445 308.23 Q180.834 308.23 179.005 311.795 Q177.2 315.336 177.2 322.466 Q177.2 329.572 179.005 333.137 Q180.834 336.679 184.445 336.679 Q188.079 336.679 189.885 333.137 Q191.713 329.572 191.713 322.466 Q191.713 315.336 189.885 311.795 Q188.079 308.23 184.445 308.23 M184.445 304.526 Q190.255 304.526 193.311 309.133 Q196.389 313.716 196.389 322.466 Q196.389 331.193 193.311 335.799 Q190.255 340.383 184.445 340.383 Q178.635 340.383 175.556 335.799 Q172.501 331.193 172.501 322.466 Q172.501 313.716 175.556 309.133 Q178.635 304.526 184.445 304.526 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M114.931 60.5889 L122.57 60.5889 L122.57 34.2233 L114.26 35.89 L114.26 31.6308 L122.524 29.9641 L127.2 29.9641 L127.2 60.5889 L134.839 60.5889 L134.839 64.5241 L114.931 64.5241 L114.931 60.5889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M154.283 33.0428 Q150.672 33.0428 148.843 36.6076 Q147.038 40.1492 147.038 47.2788 Q147.038 54.3853 148.843 57.9501 Q150.672 61.4917 154.283 61.4917 Q157.917 61.4917 159.723 57.9501 Q161.552 54.3853 161.552 47.2788 Q161.552 40.1492 159.723 36.6076 Q157.917 33.0428 154.283 33.0428 M154.283 29.3391 Q160.093 29.3391 163.149 33.9456 Q166.227 38.5289 166.227 47.2788 Q166.227 56.0056 163.149 60.6121 Q160.093 65.1954 154.283 65.1954 Q148.473 65.1954 145.394 60.6121 Q142.339 56.0056 142.339 47.2788 Q142.339 38.5289 145.394 33.9456 Q148.473 29.3391 154.283 29.3391 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M184.445 33.0428 Q180.834 33.0428 179.005 36.6076 Q177.2 40.1492 177.2 47.2788 Q177.2 54.3853 179.005 57.9501 Q180.834 61.4917 184.445 61.4917 Q188.079 61.4917 189.885 57.9501 Q191.713 54.3853 191.713 47.2788 Q191.713 40.1492 189.885 36.6076 Q188.079 33.0428 184.445 33.0428 M184.445 29.3391 Q190.255 29.3391 193.311 33.9456 Q196.389 38.5289 196.389 47.2788 Q196.389 56.0056 193.311 60.6121 Q190.255 65.1954 184.445 65.1954 Q178.635 65.1954 175.556 60.6121 Q172.501 56.0056 172.501 47.2788 Q172.501 38.5289 175.556 33.9456 Q178.635 29.3391 184.445 29.3391 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M41.7242 880.191 Q42.4244 878.122 44.7161 876.181 Q47.0077 874.207 51.0181 872.234 L64.0042 865.709 L64.0042 872.616 L51.8138 878.695 Q47.0395 881.05 45.48 883.278 Q43.9204 885.474 43.9204 889.294 L43.9204 896.296 L64.0042 896.296 L64.0042 902.726 L16.4842 902.726 L16.4842 888.212 Q16.4842 880.064 19.8898 876.053 Q23.2955 872.043 30.1704 872.043 Q34.6582 872.043 37.6183 874.144 Q40.5784 876.212 41.7242 880.191 M21.7677 896.296 L38.6368 896.296 L38.6368 888.212 Q38.6368 883.565 36.5043 881.209 Q34.34 878.822 30.1704 878.822 Q26.0009 878.822 23.9002 881.209 Q21.7677 883.565 21.7677 888.212 L21.7677 896.296 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M32.4621 846.803 Q32.4621 851.513 36.1542 854.251 Q39.8145 856.988 46.212 856.988 Q52.6095 856.988 56.3017 854.283 Q59.9619 851.545 59.9619 846.803 Q59.9619 842.124 56.2698 839.387 Q52.5777 836.65 46.212 836.65 Q39.8781 836.65 36.186 839.387 Q32.4621 842.124 32.4621 846.803 M27.4968 846.803 Q27.4968 839.164 32.4621 834.803 Q37.4273 830.443 46.212 830.443 Q54.9649 830.443 59.9619 834.803 Q64.9272 839.164 64.9272 846.803 Q64.9272 854.473 59.9619 858.834 Q54.9649 863.163 46.212 863.163 Q37.4273 863.163 32.4621 858.834 Q27.4968 854.473 27.4968 846.803 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M14.479 820.735 L14.479 814.879 L64.0042 814.879 L64.0042 820.735 L14.479 820.735 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M14.479 802.625 L14.479 796.768 L64.0042 796.768 L64.0042 802.625 L14.479 802.625 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M14.479 763.794 L14.479 757.938 L64.0042 757.938 L64.0042 763.794 L14.479 763.794 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M44.7161 715.192 L47.5806 715.192 L47.5806 742.119 Q53.6281 741.737 56.8109 738.49 Q59.9619 735.212 59.9619 729.387 Q59.9619 726.014 59.1344 722.862 Q58.3069 719.68 56.6518 716.56 L62.1899 716.56 Q63.5267 719.711 64.227 723.022 Q64.9272 726.332 64.9272 729.737 Q64.9272 738.268 59.9619 743.265 Q54.9967 748.23 46.5303 748.23 Q37.7774 748.23 32.6531 743.519 Q27.4968 738.777 27.4968 730.756 Q27.4968 723.563 32.1438 719.393 Q36.7589 715.192 44.7161 715.192 M42.9973 721.048 Q38.1912 721.112 35.3266 723.754 Q32.4621 726.364 32.4621 730.692 Q32.4621 735.594 35.2312 738.554 Q38.0002 741.482 43.0292 741.928 L42.9973 721.048 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M42.4881 675.947 L64.0042 675.947 L64.0042 681.804 L42.679 681.804 Q37.6183 681.804 35.1038 683.777 Q32.5894 685.75 32.5894 689.697 Q32.5894 694.44 35.6131 697.177 Q38.6368 699.914 43.8567 699.914 L64.0042 699.914 L64.0042 705.802 L28.3562 705.802 L28.3562 699.914 L33.8944 699.914 Q30.6797 697.813 29.0883 694.981 Q27.4968 692.116 27.4968 688.392 Q27.4968 682.249 31.3163 679.098 Q35.1038 675.947 42.4881 675.947 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M45.7664 640.809 Q39.4007 640.809 35.8996 643.45 Q32.3984 646.06 32.3984 650.803 Q32.3984 655.513 35.8996 658.155 Q39.4007 660.765 45.7664 660.765 Q52.1003 660.765 55.6014 658.155 Q59.1026 655.513 59.1026 650.803 Q59.1026 646.06 55.6014 643.45 Q52.1003 640.809 45.7664 640.809 M59.58 634.952 Q68.683 634.952 73.1071 638.994 Q77.5631 643.037 77.5631 651.376 Q77.5631 654.463 77.0857 657.2 Q76.6401 659.937 75.6852 662.516 L69.9879 662.516 Q71.3884 659.937 72.0568 657.423 Q72.7252 654.909 72.7252 652.299 Q72.7252 646.538 69.7015 643.673 Q66.7096 640.809 60.6303 640.809 L57.7339 640.809 Q60.885 642.623 62.4446 645.456 Q64.0042 648.288 64.0042 652.235 Q64.0042 658.792 59.0071 662.802 Q54.01 666.812 45.7664 666.812 Q37.491 666.812 32.4939 662.802 Q27.4968 658.792 27.4968 652.235 Q27.4968 648.288 29.0564 645.456 Q30.616 642.623 33.7671 640.809 L28.3562 640.809 L28.3562 634.952 L59.58 634.952 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M18.2347 617.096 L28.3562 617.096 L28.3562 605.033 L32.9077 605.033 L32.9077 617.096 L52.2594 617.096 Q56.6199 617.096 57.8613 615.919 Q59.1026 614.709 59.1026 611.049 L59.1026 605.033 L64.0042 605.033 L64.0042 611.049 Q64.0042 617.828 61.4897 620.406 Q58.9434 622.985 52.2594 622.985 L32.9077 622.985 L32.9077 627.281 L28.3562 627.281 L28.3562 622.985 L18.2347 622.985 L18.2347 617.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip890)" d="M42.4881 567.698 L64.0042 567.698 L64.0042 573.555 L42.679 573.555 Q37.6183 573.555 35.1038 575.528 Q32.5894 577.502 32.5894 581.448 Q32.5894 586.191 35.6131 588.928 Q38.6368 591.665 43.8567 591.665 L64.0042 591.665 L64.0042 597.554 L14.479 597.554 L14.479 591.665 L33.8944 591.665 Q30.6797 589.565 29.0883 586.732 Q27.4968 583.867 27.4968 580.143 Q27.4968 574 31.3163 570.849 Q35.1038 567.698 42.4881 567.698 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip892)" d="M273.619 391.228 L273.619 1423.18 L308.958 1423.18 L308.958 391.228 L273.619 391.228 L273.619 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="273.619,391.228 273.619,1423.18 308.958,1423.18 308.958,391.228 273.619,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="291.288" cy="391.228" r="2"/>
+<path clip-path="url(#clip892)" d="M332.518 391.228 L332.518 1423.18 L367.857 1423.18 L367.857 391.228 L332.518 391.228 L332.518 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="332.518,391.228 332.518,1423.18 367.857,1423.18 367.857,391.228 332.518,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="350.187" cy="391.228" r="2"/>
+<path clip-path="url(#clip892)" d="M391.417 391.228 L391.417 1423.18 L426.756 1423.18 L426.756 391.228 L391.417 391.228 L391.417 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="391.417,391.228 391.417,1423.18 426.756,1423.18 426.756,391.228 391.417,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="409.087" cy="391.228" r="2"/>
+<path clip-path="url(#clip892)" d="M450.316 391.228 L450.316 1423.18 L485.655 1423.18 L485.655 391.228 L450.316 391.228 L450.316 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="450.316,391.228 450.316,1423.18 485.655,1423.18 485.655,391.228 450.316,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="467.986" cy="391.228" r="2"/>
+<path clip-path="url(#clip892)" d="M509.215 391.228 L509.215 1423.18 L544.554 1423.18 L544.554 391.228 L509.215 391.228 L509.215 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="509.215,391.228 509.215,1423.18 544.554,1423.18 544.554,391.228 509.215,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="526.885" cy="391.228" r="2"/>
+<path clip-path="url(#clip892)" d="M568.114 682.926 L568.114 1423.18 L603.453 1423.18 L603.453 682.926 L568.114 682.926 L568.114 682.926  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="568.114,682.926 568.114,1423.18 603.453,1423.18 603.453,682.926 568.114,682.926 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="585.784" cy="682.926" r="2"/>
+<path clip-path="url(#clip892)" d="M627.013 693.934 L627.013 1423.18 L662.353 1423.18 L662.353 693.934 L627.013 693.934 L627.013 693.934  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="627.013,693.934 627.013,1423.18 662.353,1423.18 662.353,693.934 627.013,693.934 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="644.683" cy="693.934" r="2"/>
+<path clip-path="url(#clip892)" d="M685.912 721.453 L685.912 1423.18 L721.252 1423.18 L721.252 721.453 L685.912 721.453 L685.912 721.453  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="685.912,721.453 685.912,1423.18 721.252,1423.18 721.252,721.453 685.912,721.453 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="703.582" cy="721.453" r="2"/>
+<path clip-path="url(#clip892)" d="M744.811 732.46 L744.811 1423.18 L780.151 1423.18 L780.151 732.46 L744.811 732.46 L744.811 732.46  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="744.811,732.46 744.811,1423.18 780.151,1423.18 780.151,732.46 744.811,732.46 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="762.481" cy="732.46" r="2"/>
+<path clip-path="url(#clip892)" d="M803.71 94.0259 L803.71 1423.18 L839.05 1423.18 L839.05 94.0259 L803.71 94.0259 L803.71 94.0259  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="803.71,94.0259 803.71,1423.18 839.05,1423.18 839.05,94.0259 803.71,94.0259 "/>
+<path clip-path="url(#clip892)" d="M803.71 537.077 L803.71 1423.18 L839.05 1423.18 L839.05 537.077 L803.71 537.077 L803.71 537.077  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="803.71,537.077 803.71,1423.18 839.05,1423.18 839.05,537.077 803.71,537.077 "/>
+<path clip-path="url(#clip892)" d="M803.71 980.129 L803.71 1423.18 L839.05 1423.18 L839.05 980.129 L803.71 980.129 L803.71 980.129  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="803.71,980.129 803.71,1423.18 839.05,1423.18 839.05,980.129 803.71,980.129 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="821.38" cy="94.0259" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="821.38" cy="537.077" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="821.38" cy="980.129" r="2"/>
+<path clip-path="url(#clip892)" d="M862.609 151.815 L862.609 1423.18 L897.949 1423.18 L897.949 151.815 L862.609 151.815 L862.609 151.815  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="862.609,151.815 862.609,1423.18 897.949,1423.18 897.949,151.815 862.609,151.815 "/>
+<path clip-path="url(#clip892)" d="M862.609 575.603 L862.609 1423.18 L897.949 1423.18 L897.949 575.603 L862.609 575.603 L862.609 575.603  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="862.609,575.603 862.609,1423.18 897.949,1423.18 897.949,575.603 862.609,575.603 "/>
+<path clip-path="url(#clip892)" d="M862.609 999.392 L862.609 1423.18 L897.949 1423.18 L897.949 999.392 L862.609 999.392 L862.609 999.392  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="862.609,999.392 862.609,1423.18 897.949,1423.18 897.949,999.392 862.609,999.392 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="880.279" cy="151.815" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="880.279" cy="575.603" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="880.279" cy="999.392" r="2"/>
+<path clip-path="url(#clip892)" d="M921.508 193.093 L921.508 1423.18 L956.848 1423.18 L956.848 193.093 L921.508 193.093 L921.508 193.093  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="921.508,193.093 921.508,1423.18 956.848,1423.18 956.848,193.093 921.508,193.093 "/>
+<path clip-path="url(#clip892)" d="M921.508 603.122 L921.508 1423.18 L956.848 1423.18 L956.848 603.122 L921.508 603.122 L921.508 603.122  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="921.508,603.122 921.508,1423.18 956.848,1423.18 956.848,603.122 921.508,603.122 "/>
+<path clip-path="url(#clip892)" d="M921.508 1013.15 L921.508 1423.18 L956.848 1423.18 L956.848 1013.15 L921.508 1013.15 L921.508 1013.15  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="921.508,1013.15 921.508,1423.18 956.848,1423.18 956.848,1013.15 921.508,1013.15 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="939.178" cy="193.093" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="939.178" cy="603.122" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="939.178" cy="1013.15" r="2"/>
+<path clip-path="url(#clip892)" d="M980.408 316.928 L980.408 1423.18 L1015.75 1423.18 L1015.75 316.928 L980.408 316.928 L980.408 316.928  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="980.408,316.928 980.408,1423.18 1015.75,1423.18 1015.75,316.928 980.408,316.928 "/>
+<path clip-path="url(#clip892)" d="M980.408 593.491 L980.408 1423.18 L1015.75 1423.18 L1015.75 593.491 L980.408 593.491 L980.408 593.491  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="980.408,593.491 980.408,1423.18 1015.75,1423.18 1015.75,593.491 980.408,593.491 "/>
+<path clip-path="url(#clip892)" d="M980.408 870.054 L980.408 1423.18 L1015.75 1423.18 L1015.75 870.054 L980.408 870.054 L980.408 870.054  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="980.408,870.054 980.408,1423.18 1015.75,1423.18 1015.75,870.054 980.408,870.054 "/>
+<path clip-path="url(#clip892)" d="M980.408 1146.62 L980.408 1423.18 L1015.75 1423.18 L1015.75 1146.62 L980.408 1146.62 L980.408 1146.62  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="980.408,1146.62 980.408,1423.18 1015.75,1423.18 1015.75,1146.62 980.408,1146.62 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="998.077" cy="316.928" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="998.077" cy="593.491" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="998.077" cy="870.054" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="998.077" cy="1146.62" r="2"/>
+<path clip-path="url(#clip892)" d="M1039.31 85.7703 L1039.31 1423.18 L1074.65 1423.18 L1074.65 85.7703 L1039.31 85.7703 L1039.31 85.7703  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1039.31,85.7703 1039.31,1423.18 1074.65,1423.18 1074.65,85.7703 1039.31,85.7703 "/>
+<path clip-path="url(#clip892)" d="M1039.31 308.672 L1039.31 1423.18 L1074.65 1423.18 L1074.65 308.672 L1039.31 308.672 L1039.31 308.672  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1039.31,308.672 1039.31,1423.18 1074.65,1423.18 1074.65,308.672 1039.31,308.672 "/>
+<path clip-path="url(#clip892)" d="M1039.31 531.573 L1039.31 1423.18 L1074.65 1423.18 L1074.65 531.573 L1039.31 531.573 L1039.31 531.573  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1039.31,531.573 1039.31,1423.18 1074.65,1423.18 1074.65,531.573 1039.31,531.573 "/>
+<path clip-path="url(#clip892)" d="M1039.31 754.475 L1039.31 1423.18 L1074.65 1423.18 L1074.65 754.475 L1039.31 754.475 L1039.31 754.475  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1039.31,754.475 1039.31,1423.18 1074.65,1423.18 1074.65,754.475 1039.31,754.475 "/>
+<path clip-path="url(#clip892)" d="M1039.31 977.377 L1039.31 1423.18 L1074.65 1423.18 L1074.65 977.377 L1039.31 977.377 L1039.31 977.377  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1039.31,977.377 1039.31,1423.18 1074.65,1423.18 1074.65,977.377 1039.31,977.377 "/>
+<path clip-path="url(#clip892)" d="M1039.31 1200.28 L1039.31 1423.18 L1074.65 1423.18 L1074.65 1200.28 L1039.31 1200.28 L1039.31 1200.28  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1039.31,1200.28 1039.31,1423.18 1074.65,1423.18 1074.65,1200.28 1039.31,1200.28 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1056.98" cy="85.7703" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1056.98" cy="308.672" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1056.98" cy="531.573" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1056.98" cy="754.475" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1056.98" cy="977.377" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1056.98" cy="1200.28" r="2"/>
+<path clip-path="url(#clip892)" d="M1098.21 226.116 L1098.21 1423.18 L1133.55 1423.18 L1133.55 226.116 L1098.21 226.116 L1098.21 226.116  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1098.21,226.116 1098.21,1423.18 1133.55,1423.18 1133.55,226.116 1098.21,226.116 "/>
+<path clip-path="url(#clip892)" d="M1098.21 425.626 L1098.21 1423.18 L1133.55 1423.18 L1133.55 425.626 L1098.21 425.626 L1098.21 425.626  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1098.21,425.626 1098.21,1423.18 1133.55,1423.18 1133.55,425.626 1098.21,425.626 "/>
+<path clip-path="url(#clip892)" d="M1098.21 625.137 L1098.21 1423.18 L1133.55 1423.18 L1133.55 625.137 L1098.21 625.137 L1098.21 625.137  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1098.21,625.137 1098.21,1423.18 1133.55,1423.18 1133.55,625.137 1098.21,625.137 "/>
+<path clip-path="url(#clip892)" d="M1098.21 824.648 L1098.21 1423.18 L1133.55 1423.18 L1133.55 824.648 L1098.21 824.648 L1098.21 824.648  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1098.21,824.648 1098.21,1423.18 1133.55,1423.18 1133.55,824.648 1098.21,824.648 "/>
+<path clip-path="url(#clip892)" d="M1098.21 1024.16 L1098.21 1423.18 L1133.55 1423.18 L1133.55 1024.16 L1098.21 1024.16 L1098.21 1024.16  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1098.21,1024.16 1098.21,1423.18 1133.55,1423.18 1133.55,1024.16 1098.21,1024.16 "/>
+<path clip-path="url(#clip892)" d="M1098.21 1223.67 L1098.21 1423.18 L1133.55 1423.18 L1133.55 1223.67 L1098.21 1223.67 L1098.21 1223.67  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1098.21,1223.67 1098.21,1423.18 1133.55,1423.18 1133.55,1223.67 1098.21,1223.67 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1115.88" cy="226.116" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1115.88" cy="425.626" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1115.88" cy="625.137" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1115.88" cy="824.648" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1115.88" cy="1024.16" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1115.88" cy="1223.67" r="2"/>
+<path clip-path="url(#clip892)" d="M1157.1 61.0035 L1157.1 1423.18 L1192.44 1423.18 L1192.44 61.0035 L1157.1 61.0035 L1157.1 61.0035  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,61.0035 1157.1,1423.18 1192.44,1423.18 1192.44,61.0035 1157.1,61.0035 "/>
+<path clip-path="url(#clip892)" d="M1157.1 212.356 L1157.1 1423.18 L1192.44 1423.18 L1192.44 212.356 L1157.1 212.356 L1157.1 212.356  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,212.356 1157.1,1423.18 1192.44,1423.18 1192.44,212.356 1157.1,212.356 "/>
+<path clip-path="url(#clip892)" d="M1157.1 363.709 L1157.1 1423.18 L1192.44 1423.18 L1192.44 363.709 L1157.1 363.709 L1157.1 363.709  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,363.709 1157.1,1423.18 1192.44,1423.18 1192.44,363.709 1157.1,363.709 "/>
+<path clip-path="url(#clip892)" d="M1157.1 515.062 L1157.1 1423.18 L1192.44 1423.18 L1192.44 515.062 L1157.1 515.062 L1157.1 515.062  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,515.062 1157.1,1423.18 1192.44,1423.18 1192.44,515.062 1157.1,515.062 "/>
+<path clip-path="url(#clip892)" d="M1157.1 666.415 L1157.1 1423.18 L1192.44 1423.18 L1192.44 666.415 L1157.1 666.415 L1157.1 666.415  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,666.415 1157.1,1423.18 1192.44,1423.18 1192.44,666.415 1157.1,666.415 "/>
+<path clip-path="url(#clip892)" d="M1157.1 817.768 L1157.1 1423.18 L1192.44 1423.18 L1192.44 817.768 L1157.1 817.768 L1157.1 817.768  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,817.768 1157.1,1423.18 1192.44,1423.18 1192.44,817.768 1157.1,817.768 "/>
+<path clip-path="url(#clip892)" d="M1157.1 969.121 L1157.1 1423.18 L1192.44 1423.18 L1192.44 969.121 L1157.1 969.121 L1157.1 969.121  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,969.121 1157.1,1423.18 1192.44,1423.18 1192.44,969.121 1157.1,969.121 "/>
+<path clip-path="url(#clip892)" d="M1157.1 1120.47 L1157.1 1423.18 L1192.44 1423.18 L1192.44 1120.47 L1157.1 1120.47 L1157.1 1120.47  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,1120.47 1157.1,1423.18 1192.44,1423.18 1192.44,1120.47 1157.1,1120.47 "/>
+<path clip-path="url(#clip892)" d="M1157.1 1271.83 L1157.1 1423.18 L1192.44 1423.18 L1192.44 1271.83 L1157.1 1271.83 L1157.1 1271.83  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1157.1,1271.83 1157.1,1423.18 1192.44,1423.18 1192.44,1271.83 1157.1,1271.83 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="61.0035" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="212.356" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="363.709" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="515.062" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="666.415" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="817.768" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="969.121" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="1120.47" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1174.77" cy="1271.83" r="2"/>
+<path clip-path="url(#clip892)" d="M1216 121.545 L1216 1423.18 L1251.34 1423.18 L1251.34 121.545 L1216 121.545 L1216 121.545  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,121.545 1216,1423.18 1251.34,1423.18 1251.34,121.545 1216,121.545 "/>
+<path clip-path="url(#clip892)" d="M1216 239.875 L1216 1423.18 L1251.34 1423.18 L1251.34 239.875 L1216 239.875 L1216 239.875  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,239.875 1216,1423.18 1251.34,1423.18 1251.34,239.875 1216,239.875 "/>
+<path clip-path="url(#clip892)" d="M1216 358.206 L1216 1423.18 L1251.34 1423.18 L1251.34 358.206 L1216 358.206 L1216 358.206  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,358.206 1216,1423.18 1251.34,1423.18 1251.34,358.206 1216,358.206 "/>
+<path clip-path="url(#clip892)" d="M1216 476.536 L1216 1423.18 L1251.34 1423.18 L1251.34 476.536 L1216 476.536 L1216 476.536  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,476.536 1216,1423.18 1251.34,1423.18 1251.34,476.536 1216,476.536 "/>
+<path clip-path="url(#clip892)" d="M1216 594.867 L1216 1423.18 L1251.34 1423.18 L1251.34 594.867 L1216 594.867 L1216 594.867  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,594.867 1216,1423.18 1251.34,1423.18 1251.34,594.867 1216,594.867 "/>
+<path clip-path="url(#clip892)" d="M1216 713.197 L1216 1423.18 L1251.34 1423.18 L1251.34 713.197 L1216 713.197 L1216 713.197  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,713.197 1216,1423.18 1251.34,1423.18 1251.34,713.197 1216,713.197 "/>
+<path clip-path="url(#clip892)" d="M1216 831.527 L1216 1423.18 L1251.34 1423.18 L1251.34 831.527 L1216 831.527 L1216 831.527  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,831.527 1216,1423.18 1251.34,1423.18 1251.34,831.527 1216,831.527 "/>
+<path clip-path="url(#clip892)" d="M1216 949.858 L1216 1423.18 L1251.34 1423.18 L1251.34 949.858 L1216 949.858 L1216 949.858  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,949.858 1216,1423.18 1251.34,1423.18 1251.34,949.858 1216,949.858 "/>
+<path clip-path="url(#clip892)" d="M1216 1068.19 L1216 1423.18 L1251.34 1423.18 L1251.34 1068.19 L1216 1068.19 L1216 1068.19  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,1068.19 1216,1423.18 1251.34,1423.18 1251.34,1068.19 1216,1068.19 "/>
+<path clip-path="url(#clip892)" d="M1216 1186.52 L1216 1423.18 L1251.34 1423.18 L1251.34 1186.52 L1216 1186.52 L1216 1186.52  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,1186.52 1216,1423.18 1251.34,1423.18 1251.34,1186.52 1216,1186.52 "/>
+<path clip-path="url(#clip892)" d="M1216 1304.85 L1216 1423.18 L1251.34 1423.18 L1251.34 1304.85 L1216 1304.85 L1216 1304.85  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1216,1304.85 1216,1423.18 1251.34,1423.18 1251.34,1304.85 1216,1304.85 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="121.545" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="239.875" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="358.206" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="476.536" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="594.867" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="713.197" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="831.527" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="949.858" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="1068.19" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="1186.52" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1233.67" cy="1304.85" r="2"/>
+<path clip-path="url(#clip892)" d="M1274.9 69.2591 L1274.9 1423.18 L1310.24 1423.18 L1310.24 69.2591 L1274.9 69.2591 L1274.9 69.2591  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,69.2591 1274.9,1423.18 1310.24,1423.18 1310.24,69.2591 1274.9,69.2591 "/>
+<path clip-path="url(#clip892)" d="M1274.9 182.086 L1274.9 1423.18 L1310.24 1423.18 L1310.24 182.086 L1274.9 182.086 L1274.9 182.086  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,182.086 1274.9,1423.18 1310.24,1423.18 1310.24,182.086 1274.9,182.086 "/>
+<path clip-path="url(#clip892)" d="M1274.9 294.913 L1274.9 1423.18 L1310.24 1423.18 L1310.24 294.913 L1274.9 294.913 L1274.9 294.913  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,294.913 1274.9,1423.18 1310.24,1423.18 1310.24,294.913 1274.9,294.913 "/>
+<path clip-path="url(#clip892)" d="M1274.9 407.739 L1274.9 1423.18 L1310.24 1423.18 L1310.24 407.739 L1274.9 407.739 L1274.9 407.739  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,407.739 1274.9,1423.18 1310.24,1423.18 1310.24,407.739 1274.9,407.739 "/>
+<path clip-path="url(#clip892)" d="M1274.9 520.566 L1274.9 1423.18 L1310.24 1423.18 L1310.24 520.566 L1274.9 520.566 L1274.9 520.566  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,520.566 1274.9,1423.18 1310.24,1423.18 1310.24,520.566 1274.9,520.566 "/>
+<path clip-path="url(#clip892)" d="M1274.9 633.393 L1274.9 1423.18 L1310.24 1423.18 L1310.24 633.393 L1274.9 633.393 L1274.9 633.393  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,633.393 1274.9,1423.18 1310.24,1423.18 1310.24,633.393 1274.9,633.393 "/>
+<path clip-path="url(#clip892)" d="M1274.9 746.219 L1274.9 1423.18 L1310.24 1423.18 L1310.24 746.219 L1274.9 746.219 L1274.9 746.219  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,746.219 1274.9,1423.18 1310.24,1423.18 1310.24,746.219 1274.9,746.219 "/>
+<path clip-path="url(#clip892)" d="M1274.9 859.046 L1274.9 1423.18 L1310.24 1423.18 L1310.24 859.046 L1274.9 859.046 L1274.9 859.046  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,859.046 1274.9,1423.18 1310.24,1423.18 1310.24,859.046 1274.9,859.046 "/>
+<path clip-path="url(#clip892)" d="M1274.9 971.873 L1274.9 1423.18 L1310.24 1423.18 L1310.24 971.873 L1274.9 971.873 L1274.9 971.873  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,971.873 1274.9,1423.18 1310.24,1423.18 1310.24,971.873 1274.9,971.873 "/>
+<path clip-path="url(#clip892)" d="M1274.9 1084.7 L1274.9 1423.18 L1310.24 1423.18 L1310.24 1084.7 L1274.9 1084.7 L1274.9 1084.7  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,1084.7 1274.9,1423.18 1310.24,1423.18 1310.24,1084.7 1274.9,1084.7 "/>
+<path clip-path="url(#clip892)" d="M1274.9 1197.53 L1274.9 1423.18 L1310.24 1423.18 L1310.24 1197.53 L1274.9 1197.53 L1274.9 1197.53  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,1197.53 1274.9,1423.18 1310.24,1423.18 1310.24,1197.53 1274.9,1197.53 "/>
+<path clip-path="url(#clip892)" d="M1274.9 1310.35 L1274.9 1423.18 L1310.24 1423.18 L1310.24 1310.35 L1274.9 1310.35 L1274.9 1310.35  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1274.9,1310.35 1274.9,1423.18 1310.24,1423.18 1310.24,1310.35 1274.9,1310.35 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="69.2591" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="182.086" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="294.913" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="407.739" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="520.566" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="633.393" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="746.219" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="859.046" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="971.873" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="1084.7" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="1197.53" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1292.57" cy="1310.35" r="2"/>
+<path clip-path="url(#clip892)" d="M1333.8 61.0035 L1333.8 1423.18 L1369.14 1423.18 L1369.14 61.0035 L1333.8 61.0035 L1333.8 61.0035  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,61.0035 1333.8,1423.18 1369.14,1423.18 1369.14,61.0035 1333.8,61.0035 "/>
+<path clip-path="url(#clip892)" d="M1333.8 151.815 L1333.8 1423.18 L1369.14 1423.18 L1369.14 151.815 L1333.8 151.815 L1333.8 151.815  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,151.815 1333.8,1423.18 1369.14,1423.18 1369.14,151.815 1333.8,151.815 "/>
+<path clip-path="url(#clip892)" d="M1333.8 242.627 L1333.8 1423.18 L1369.14 1423.18 L1369.14 242.627 L1333.8 242.627 L1333.8 242.627  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,242.627 1333.8,1423.18 1369.14,1423.18 1369.14,242.627 1333.8,242.627 "/>
+<path clip-path="url(#clip892)" d="M1333.8 333.439 L1333.8 1423.18 L1369.14 1423.18 L1369.14 333.439 L1333.8 333.439 L1333.8 333.439  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,333.439 1333.8,1423.18 1369.14,1423.18 1369.14,333.439 1333.8,333.439 "/>
+<path clip-path="url(#clip892)" d="M1333.8 424.25 L1333.8 1423.18 L1369.14 1423.18 L1369.14 424.25 L1333.8 424.25 L1333.8 424.25  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,424.25 1333.8,1423.18 1369.14,1423.18 1369.14,424.25 1333.8,424.25 "/>
+<path clip-path="url(#clip892)" d="M1333.8 515.062 L1333.8 1423.18 L1369.14 1423.18 L1369.14 515.062 L1333.8 515.062 L1333.8 515.062  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,515.062 1333.8,1423.18 1369.14,1423.18 1369.14,515.062 1333.8,515.062 "/>
+<path clip-path="url(#clip892)" d="M1333.8 605.874 L1333.8 1423.18 L1369.14 1423.18 L1369.14 605.874 L1333.8 605.874 L1333.8 605.874  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,605.874 1333.8,1423.18 1369.14,1423.18 1369.14,605.874 1333.8,605.874 "/>
+<path clip-path="url(#clip892)" d="M1333.8 696.686 L1333.8 1423.18 L1369.14 1423.18 L1369.14 696.686 L1333.8 696.686 L1333.8 696.686  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,696.686 1333.8,1423.18 1369.14,1423.18 1369.14,696.686 1333.8,696.686 "/>
+<path clip-path="url(#clip892)" d="M1333.8 787.498 L1333.8 1423.18 L1369.14 1423.18 L1369.14 787.498 L1333.8 787.498 L1333.8 787.498  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,787.498 1333.8,1423.18 1369.14,1423.18 1369.14,787.498 1333.8,787.498 "/>
+<path clip-path="url(#clip892)" d="M1333.8 878.309 L1333.8 1423.18 L1369.14 1423.18 L1369.14 878.309 L1333.8 878.309 L1333.8 878.309  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,878.309 1333.8,1423.18 1369.14,1423.18 1369.14,878.309 1333.8,878.309 "/>
+<path clip-path="url(#clip892)" d="M1333.8 969.121 L1333.8 1423.18 L1369.14 1423.18 L1369.14 969.121 L1333.8 969.121 L1333.8 969.121  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,969.121 1333.8,1423.18 1369.14,1423.18 1369.14,969.121 1333.8,969.121 "/>
+<path clip-path="url(#clip892)" d="M1333.8 1059.93 L1333.8 1423.18 L1369.14 1423.18 L1369.14 1059.93 L1333.8 1059.93 L1333.8 1059.93  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,1059.93 1333.8,1423.18 1369.14,1423.18 1369.14,1059.93 1333.8,1059.93 "/>
+<path clip-path="url(#clip892)" d="M1333.8 1150.74 L1333.8 1423.18 L1369.14 1423.18 L1369.14 1150.74 L1333.8 1150.74 L1333.8 1150.74  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,1150.74 1333.8,1423.18 1369.14,1423.18 1369.14,1150.74 1333.8,1150.74 "/>
+<path clip-path="url(#clip892)" d="M1333.8 1241.56 L1333.8 1423.18 L1369.14 1423.18 L1369.14 1241.56 L1333.8 1241.56 L1333.8 1241.56  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,1241.56 1333.8,1423.18 1369.14,1423.18 1369.14,1241.56 1333.8,1241.56 "/>
+<path clip-path="url(#clip892)" d="M1333.8 1332.37 L1333.8 1423.18 L1369.14 1423.18 L1369.14 1332.37 L1333.8 1332.37 L1333.8 1332.37  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1333.8,1332.37 1333.8,1423.18 1369.14,1423.18 1369.14,1332.37 1333.8,1332.37 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="61.0035" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="151.815" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="242.627" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="333.439" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="424.25" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="515.062" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="605.874" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="696.686" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="787.498" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="878.309" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="969.121" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="1059.93" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="1150.74" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="1241.56" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1351.47" cy="1332.37" r="2"/>
+<path clip-path="url(#clip892)" d="M1392.7 89.8981 L1392.7 1423.18 L1428.04 1423.18 L1428.04 89.8981 L1392.7 89.8981 L1392.7 89.8981  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,89.8981 1392.7,1423.18 1428.04,1423.18 1428.04,89.8981 1392.7,89.8981 "/>
+<path clip-path="url(#clip892)" d="M1392.7 160.071 L1392.7 1423.18 L1428.04 1423.18 L1428.04 160.071 L1392.7 160.071 L1392.7 160.071  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,160.071 1392.7,1423.18 1428.04,1423.18 1428.04,160.071 1392.7,160.071 "/>
+<path clip-path="url(#clip892)" d="M1392.7 230.244 L1392.7 1423.18 L1428.04 1423.18 L1428.04 230.244 L1392.7 230.244 L1392.7 230.244  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,230.244 1392.7,1423.18 1428.04,1423.18 1428.04,230.244 1392.7,230.244 "/>
+<path clip-path="url(#clip892)" d="M1392.7 300.416 L1392.7 1423.18 L1428.04 1423.18 L1428.04 300.416 L1392.7 300.416 L1392.7 300.416  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,300.416 1392.7,1423.18 1428.04,1423.18 1428.04,300.416 1392.7,300.416 "/>
+<path clip-path="url(#clip892)" d="M1392.7 370.589 L1392.7 1423.18 L1428.04 1423.18 L1428.04 370.589 L1392.7 370.589 L1392.7 370.589  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,370.589 1392.7,1423.18 1428.04,1423.18 1428.04,370.589 1392.7,370.589 "/>
+<path clip-path="url(#clip892)" d="M1392.7 440.762 L1392.7 1423.18 L1428.04 1423.18 L1428.04 440.762 L1392.7 440.762 L1392.7 440.762  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,440.762 1392.7,1423.18 1428.04,1423.18 1428.04,440.762 1392.7,440.762 "/>
+<path clip-path="url(#clip892)" d="M1392.7 510.934 L1392.7 1423.18 L1428.04 1423.18 L1428.04 510.934 L1392.7 510.934 L1392.7 510.934  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,510.934 1392.7,1423.18 1428.04,1423.18 1428.04,510.934 1392.7,510.934 "/>
+<path clip-path="url(#clip892)" d="M1392.7 581.107 L1392.7 1423.18 L1428.04 1423.18 L1428.04 581.107 L1392.7 581.107 L1392.7 581.107  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,581.107 1392.7,1423.18 1428.04,1423.18 1428.04,581.107 1392.7,581.107 "/>
+<path clip-path="url(#clip892)" d="M1392.7 651.28 L1392.7 1423.18 L1428.04 1423.18 L1428.04 651.28 L1392.7 651.28 L1392.7 651.28  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,651.28 1392.7,1423.18 1428.04,1423.18 1428.04,651.28 1392.7,651.28 "/>
+<path clip-path="url(#clip892)" d="M1392.7 721.453 L1392.7 1423.18 L1428.04 1423.18 L1428.04 721.453 L1392.7 721.453 L1392.7 721.453  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,721.453 1392.7,1423.18 1428.04,1423.18 1428.04,721.453 1392.7,721.453 "/>
+<path clip-path="url(#clip892)" d="M1392.7 791.625 L1392.7 1423.18 L1428.04 1423.18 L1428.04 791.625 L1392.7 791.625 L1392.7 791.625  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,791.625 1392.7,1423.18 1428.04,1423.18 1428.04,791.625 1392.7,791.625 "/>
+<path clip-path="url(#clip892)" d="M1392.7 861.798 L1392.7 1423.18 L1428.04 1423.18 L1428.04 861.798 L1392.7 861.798 L1392.7 861.798  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,861.798 1392.7,1423.18 1428.04,1423.18 1428.04,861.798 1392.7,861.798 "/>
+<path clip-path="url(#clip892)" d="M1392.7 931.971 L1392.7 1423.18 L1428.04 1423.18 L1428.04 931.971 L1392.7 931.971 L1392.7 931.971  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,931.971 1392.7,1423.18 1428.04,1423.18 1428.04,931.971 1392.7,931.971 "/>
+<path clip-path="url(#clip892)" d="M1392.7 1002.14 L1392.7 1423.18 L1428.04 1423.18 L1428.04 1002.14 L1392.7 1002.14 L1392.7 1002.14  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,1002.14 1392.7,1423.18 1428.04,1423.18 1428.04,1002.14 1392.7,1002.14 "/>
+<path clip-path="url(#clip892)" d="M1392.7 1072.32 L1392.7 1423.18 L1428.04 1423.18 L1428.04 1072.32 L1392.7 1072.32 L1392.7 1072.32  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,1072.32 1392.7,1423.18 1428.04,1423.18 1428.04,1072.32 1392.7,1072.32 "/>
+<path clip-path="url(#clip892)" d="M1392.7 1142.49 L1392.7 1423.18 L1428.04 1423.18 L1428.04 1142.49 L1392.7 1142.49 L1392.7 1142.49  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,1142.49 1392.7,1423.18 1428.04,1423.18 1428.04,1142.49 1392.7,1142.49 "/>
+<path clip-path="url(#clip892)" d="M1392.7 1212.66 L1392.7 1423.18 L1428.04 1423.18 L1428.04 1212.66 L1392.7 1212.66 L1392.7 1212.66  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,1212.66 1392.7,1423.18 1428.04,1423.18 1428.04,1212.66 1392.7,1212.66 "/>
+<path clip-path="url(#clip892)" d="M1392.7 1282.83 L1392.7 1423.18 L1428.04 1423.18 L1428.04 1282.83 L1392.7 1282.83 L1392.7 1282.83  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,1282.83 1392.7,1423.18 1428.04,1423.18 1428.04,1282.83 1392.7,1282.83 "/>
+<path clip-path="url(#clip892)" d="M1392.7 1353.01 L1392.7 1423.18 L1428.04 1423.18 L1428.04 1353.01 L1392.7 1353.01 L1392.7 1353.01  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1392.7,1353.01 1392.7,1423.18 1428.04,1423.18 1428.04,1353.01 1392.7,1353.01 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="89.8981" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="160.071" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="230.244" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="300.416" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="370.589" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="440.762" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="510.934" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="581.107" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="651.28" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="721.453" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="791.625" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="861.798" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="931.971" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="1002.14" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="1072.32" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="1142.49" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="1212.66" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="1282.83" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1410.37" cy="1353.01" r="2"/>
+<path clip-path="url(#clip892)" d="M1451.6 61.0035 L1451.6 1423.18 L1486.94 1423.18 L1486.94 61.0035 L1451.6 61.0035 L1451.6 61.0035  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1451.6,61.0035 1451.6,1423.18 1486.94,1423.18 1486.94,61.0035 1451.6,61.0035 "/>
+<path clip-path="url(#clip892)" d="M1451.6 179.334 L1451.6 1423.18 L1486.94 1423.18 L1486.94 179.334 L1451.6 179.334 L1451.6 179.334  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1451.6,179.334 1451.6,1423.18 1486.94,1423.18 1486.94,179.334 1451.6,179.334 "/>
+<path clip-path="url(#clip892)" d="M1451.6 455.897 L1451.6 1423.18 L1486.94 1423.18 L1486.94 455.897 L1451.6 455.897 L1451.6 455.897  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1451.6,455.897 1451.6,1423.18 1486.94,1423.18 1486.94,455.897 1451.6,455.897 "/>
+<path clip-path="url(#clip892)" d="M1451.6 732.46 L1451.6 1423.18 L1486.94 1423.18 L1486.94 732.46 L1451.6 732.46 L1451.6 732.46  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1451.6,732.46 1451.6,1423.18 1486.94,1423.18 1486.94,732.46 1451.6,732.46 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1469.27" cy="61.0035" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1469.27" cy="179.334" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1469.27" cy="455.897" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1469.27" cy="732.46" r="2"/>
+<path clip-path="url(#clip892)" d="M1510.5 52.7478 L1510.5 1423.18 L1545.84 1423.18 L1545.84 52.7478 L1510.5 52.7478 L1510.5 52.7478  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1510.5,52.7478 1510.5,1423.18 1545.84,1423.18 1545.84,52.7478 1510.5,52.7478 "/>
+<path clip-path="url(#clip892)" d="M1510.5 122.921 L1510.5 1423.18 L1545.84 1423.18 L1545.84 122.921 L1510.5 122.921 L1510.5 122.921  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1510.5,122.921 1510.5,1423.18 1545.84,1423.18 1545.84,122.921 1510.5,122.921 "/>
+<path clip-path="url(#clip892)" d="M1510.5 322.431 L1510.5 1423.18 L1545.84 1423.18 L1545.84 322.431 L1510.5 322.431 L1510.5 322.431  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1510.5,322.431 1510.5,1423.18 1545.84,1423.18 1545.84,322.431 1510.5,322.431 "/>
+<path clip-path="url(#clip892)" d="M1510.5 521.942 L1510.5 1423.18 L1545.84 1423.18 L1545.84 521.942 L1510.5 521.942 L1510.5 521.942  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1510.5,521.942 1510.5,1423.18 1545.84,1423.18 1545.84,521.942 1510.5,521.942 "/>
+<path clip-path="url(#clip892)" d="M1510.5 721.453 L1510.5 1423.18 L1545.84 1423.18 L1545.84 721.453 L1510.5 721.453 L1510.5 721.453  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1510.5,721.453 1510.5,1423.18 1545.84,1423.18 1545.84,721.453 1510.5,721.453 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1528.17" cy="52.7478" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1528.17" cy="122.921" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1528.17" cy="322.431" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1528.17" cy="521.942" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1528.17" cy="721.453" r="2"/>
+<path clip-path="url(#clip892)" d="M1569.4 47.2441 L1569.4 1423.18 L1604.74 1423.18 L1604.74 47.2441 L1569.4 47.2441 L1569.4 47.2441  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1569.4,47.2441 1569.4,1423.18 1604.74,1423.18 1604.74,47.2441 1569.4,47.2441 "/>
+<path clip-path="url(#clip892)" d="M1569.4 165.575 L1569.4 1423.18 L1604.74 1423.18 L1604.74 165.575 L1569.4 165.575 L1569.4 165.575  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1569.4,165.575 1569.4,1423.18 1604.74,1423.18 1604.74,165.575 1569.4,165.575 "/>
+<path clip-path="url(#clip892)" d="M1569.4 283.905 L1569.4 1423.18 L1604.74 1423.18 L1604.74 283.905 L1569.4 283.905 L1569.4 283.905  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1569.4,283.905 1569.4,1423.18 1604.74,1423.18 1604.74,283.905 1569.4,283.905 "/>
+<path clip-path="url(#clip892)" d="M1569.4 693.934 L1569.4 1423.18 L1604.74 1423.18 L1604.74 693.934 L1569.4 693.934 L1569.4 693.934  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1569.4,693.934 1569.4,1423.18 1604.74,1423.18 1604.74,693.934 1569.4,693.934 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1587.07" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1587.07" cy="165.575" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1587.07" cy="283.905" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1587.07" cy="693.934" r="2"/>
+<path clip-path="url(#clip892)" d="M1628.3 48.62 L1628.3 1423.18 L1663.64 1423.18 L1663.64 48.62 L1628.3 48.62 L1628.3 48.62  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1628.3,48.62 1628.3,1423.18 1663.64,1423.18 1663.64,48.62 1628.3,48.62 "/>
+<path clip-path="url(#clip892)" d="M1628.3 118.793 L1628.3 1423.18 L1663.64 1423.18 L1663.64 118.793 L1628.3 118.793 L1628.3 118.793  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1628.3,118.793 1628.3,1423.18 1663.64,1423.18 1663.64,118.793 1628.3,118.793 "/>
+<path clip-path="url(#clip892)" d="M1628.3 209.605 L1628.3 1423.18 L1663.64 1423.18 L1663.64 209.605 L1628.3 209.605 L1628.3 209.605  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1628.3,209.605 1628.3,1423.18 1663.64,1423.18 1663.64,209.605 1628.3,209.605 "/>
+<path clip-path="url(#clip892)" d="M1628.3 322.431 L1628.3 1423.18 L1663.64 1423.18 L1663.64 322.431 L1628.3 322.431 L1628.3 322.431  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1628.3,322.431 1628.3,1423.18 1663.64,1423.18 1663.64,322.431 1628.3,322.431 "/>
+<path clip-path="url(#clip892)" d="M1628.3 732.46 L1628.3 1423.18 L1663.64 1423.18 L1663.64 732.46 L1628.3 732.46 L1628.3 732.46  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1628.3,732.46 1628.3,1423.18 1663.64,1423.18 1663.64,732.46 1628.3,732.46 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="48.62" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="118.793" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="209.605" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="322.431" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1645.97" cy="732.46" r="2"/>
+<path clip-path="url(#clip892)" d="M1687.2 55.4997 L1687.2 1423.18 L1722.54 1423.18 L1722.54 55.4997 L1687.2 55.4997 L1687.2 55.4997  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1687.2,55.4997 1687.2,1423.18 1722.54,1423.18 1722.54,55.4997 1687.2,55.4997 "/>
+<path clip-path="url(#clip892)" d="M1687.2 146.311 L1687.2 1423.18 L1722.54 1423.18 L1722.54 146.311 L1687.2 146.311 L1687.2 146.311  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1687.2,146.311 1687.2,1423.18 1722.54,1423.18 1722.54,146.311 1687.2,146.311 "/>
+<path clip-path="url(#clip892)" d="M1687.2 297.664 L1687.2 1423.18 L1722.54 1423.18 L1722.54 297.664 L1687.2 297.664 L1687.2 297.664  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1687.2,297.664 1687.2,1423.18 1722.54,1423.18 1722.54,297.664 1687.2,297.664 "/>
+<path clip-path="url(#clip892)" d="M1687.2 721.453 L1687.2 1423.18 L1722.54 1423.18 L1722.54 721.453 L1687.2 721.453 L1687.2 721.453  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1687.2,721.453 1687.2,1423.18 1722.54,1423.18 1722.54,721.453 1687.2,721.453 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1704.87" cy="55.4997" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1704.87" cy="146.311" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1704.87" cy="297.664" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1704.87" cy="721.453" r="2"/>
+<path clip-path="url(#clip892)" d="M1746.1 47.2441 L1746.1 1423.18 L1781.43 1423.18 L1781.43 47.2441 L1746.1 47.2441 L1746.1 47.2441  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1746.1,47.2441 1746.1,1423.18 1781.43,1423.18 1781.43,47.2441 1746.1,47.2441 "/>
+<path clip-path="url(#clip892)" d="M1746.1 270.146 L1746.1 1423.18 L1781.43 1423.18 L1781.43 270.146 L1746.1 270.146 L1746.1 270.146  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1746.1,270.146 1746.1,1423.18 1781.43,1423.18 1781.43,270.146 1746.1,270.146 "/>
+<path clip-path="url(#clip892)" d="M1746.1 693.934 L1746.1 1423.18 L1781.43 1423.18 L1781.43 693.934 L1746.1 693.934 L1746.1 693.934  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1746.1,693.934 1746.1,1423.18 1781.43,1423.18 1781.43,693.934 1746.1,693.934 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1763.77" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1763.77" cy="270.146" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1763.77" cy="693.934" r="2"/>
+<path clip-path="url(#clip892)" d="M1804.99 59.6275 L1804.99 1423.18 L1840.33 1423.18 L1840.33 59.6275 L1804.99 59.6275 L1804.99 59.6275  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1804.99,59.6275 1804.99,1423.18 1840.33,1423.18 1840.33,59.6275 1804.99,59.6275 "/>
+<path clip-path="url(#clip892)" d="M1804.99 129.8 L1804.99 1423.18 L1840.33 1423.18 L1840.33 129.8 L1804.99 129.8 L1804.99 129.8  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1804.99,129.8 1804.99,1423.18 1840.33,1423.18 1840.33,129.8 1804.99,129.8 "/>
+<path clip-path="url(#clip892)" d="M1804.99 406.363 L1804.99 1423.18 L1840.33 1423.18 L1840.33 406.363 L1804.99 406.363 L1804.99 406.363  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1804.99,406.363 1804.99,1423.18 1840.33,1423.18 1840.33,406.363 1804.99,406.363 "/>
+<path clip-path="url(#clip892)" d="M1804.99 682.926 L1804.99 1423.18 L1840.33 1423.18 L1840.33 682.926 L1804.99 682.926 L1804.99 682.926  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1804.99,682.926 1804.99,1423.18 1840.33,1423.18 1840.33,682.926 1804.99,682.926 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1822.66" cy="59.6275" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1822.66" cy="129.8" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1822.66" cy="406.363" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1822.66" cy="682.926" r="2"/>
+<path clip-path="url(#clip892)" d="M1863.89 58.2516 L1863.89 1423.18 L1899.23 1423.18 L1899.23 58.2516 L1863.89 58.2516 L1863.89 58.2516  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1863.89,58.2516 1863.89,1423.18 1899.23,1423.18 1899.23,58.2516 1863.89,58.2516 "/>
+<path clip-path="url(#clip892)" d="M1863.89 176.582 L1863.89 1423.18 L1899.23 1423.18 L1899.23 176.582 L1863.89 176.582 L1863.89 176.582  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1863.89,176.582 1863.89,1423.18 1899.23,1423.18 1899.23,176.582 1863.89,176.582 "/>
+<path clip-path="url(#clip892)" d="M1863.89 294.913 L1863.89 1423.18 L1899.23 1423.18 L1899.23 294.913 L1863.89 294.913 L1863.89 294.913  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1863.89,294.913 1863.89,1423.18 1899.23,1423.18 1899.23,294.913 1863.89,294.913 "/>
+<path clip-path="url(#clip892)" d="M1863.89 413.243 L1863.89 1423.18 L1899.23 1423.18 L1899.23 413.243 L1863.89 413.243 L1863.89 413.243  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1863.89,413.243 1863.89,1423.18 1899.23,1423.18 1899.23,413.243 1863.89,413.243 "/>
+<path clip-path="url(#clip892)" d="M1863.89 531.573 L1863.89 1423.18 L1899.23 1423.18 L1899.23 531.573 L1863.89 531.573 L1863.89 531.573  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1863.89,531.573 1863.89,1423.18 1899.23,1423.18 1899.23,531.573 1863.89,531.573 "/>
+<path clip-path="url(#clip892)" d="M1863.89 682.926 L1863.89 1423.18 L1899.23 1423.18 L1899.23 682.926 L1863.89 682.926 L1863.89 682.926  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1863.89,682.926 1863.89,1423.18 1899.23,1423.18 1899.23,682.926 1863.89,682.926 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1881.56" cy="58.2516" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1881.56" cy="176.582" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1881.56" cy="294.913" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1881.56" cy="413.243" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1881.56" cy="531.573" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1881.56" cy="682.926" r="2"/>
+<path clip-path="url(#clip892)" d="M1922.79 55.4997 L1922.79 1423.18 L1958.13 1423.18 L1958.13 55.4997 L1922.79 55.4997 L1922.79 55.4997  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1922.79,55.4997 1922.79,1423.18 1958.13,1423.18 1958.13,55.4997 1922.79,55.4997 "/>
+<path clip-path="url(#clip892)" d="M1922.79 278.401 L1922.79 1423.18 L1958.13 1423.18 L1958.13 278.401 L1922.79 278.401 L1922.79 278.401  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1922.79,278.401 1922.79,1423.18 1958.13,1423.18 1958.13,278.401 1922.79,278.401 "/>
+<path clip-path="url(#clip892)" d="M1922.79 721.453 L1922.79 1423.18 L1958.13 1423.18 L1958.13 721.453 L1922.79 721.453 L1922.79 721.453  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1922.79,721.453 1922.79,1423.18 1958.13,1423.18 1958.13,721.453 1922.79,721.453 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1940.46" cy="55.4997" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1940.46" cy="278.401" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1940.46" cy="721.453" r="2"/>
+<path clip-path="url(#clip892)" d="M1981.69 47.2441 L1981.69 1423.18 L2017.03 1423.18 L2017.03 47.2441 L1981.69 47.2441 L1981.69 47.2441  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1981.69,47.2441 1981.69,1423.18 2017.03,1423.18 2017.03,47.2441 1981.69,47.2441 "/>
+<path clip-path="url(#clip892)" d="M1981.69 160.071 L1981.69 1423.18 L2017.03 1423.18 L2017.03 160.071 L1981.69 160.071 L1981.69 160.071  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1981.69,160.071 1981.69,1423.18 2017.03,1423.18 2017.03,160.071 1981.69,160.071 "/>
+<path clip-path="url(#clip892)" d="M1981.69 272.898 L1981.69 1423.18 L2017.03 1423.18 L2017.03 272.898 L1981.69 272.898 L1981.69 272.898  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1981.69,272.898 1981.69,1423.18 2017.03,1423.18 2017.03,272.898 1981.69,272.898 "/>
+<path clip-path="url(#clip892)" d="M1981.69 682.926 L1981.69 1423.18 L2017.03 1423.18 L2017.03 682.926 L1981.69 682.926 L1981.69 682.926  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1981.69,682.926 1981.69,1423.18 2017.03,1423.18 2017.03,682.926 1981.69,682.926 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1999.36" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1999.36" cy="160.071" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1999.36" cy="272.898" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="1999.36" cy="682.926" r="2"/>
+<path clip-path="url(#clip892)" d="M2040.59 88.5222 L2040.59 1423.18 L2075.93 1423.18 L2075.93 88.5222 L2040.59 88.5222 L2040.59 88.5222  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2040.59,88.5222 2040.59,1423.18 2075.93,1423.18 2075.93,88.5222 2040.59,88.5222 "/>
+<path clip-path="url(#clip892)" d="M2040.59 239.875 L2040.59 1423.18 L2075.93 1423.18 L2075.93 239.875 L2040.59 239.875 L2040.59 239.875  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2040.59,239.875 2040.59,1423.18 2075.93,1423.18 2075.93,239.875 2040.59,239.875 "/>
+<path clip-path="url(#clip892)" d="M2040.59 682.926 L2040.59 1423.18 L2075.93 1423.18 L2075.93 682.926 L2040.59 682.926 L2040.59 682.926  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2040.59,682.926 2040.59,1423.18 2075.93,1423.18 2075.93,682.926 2040.59,682.926 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2058.26" cy="88.5222" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2058.26" cy="239.875" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2058.26" cy="682.926" r="2"/>
+<path clip-path="url(#clip892)" d="M2099.49 49.996 L2099.49 1423.18 L2134.83 1423.18 L2134.83 49.996 L2099.49 49.996 L2099.49 49.996  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2099.49,49.996 2099.49,1423.18 2134.83,1423.18 2134.83,49.996 2099.49,49.996 "/>
+<path clip-path="url(#clip892)" d="M2099.49 168.326 L2099.49 1423.18 L2134.83 1423.18 L2134.83 168.326 L2099.49 168.326 L2099.49 168.326  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2099.49,168.326 2099.49,1423.18 2134.83,1423.18 2134.83,168.326 2099.49,168.326 "/>
+<path clip-path="url(#clip892)" d="M2099.49 391.228 L2099.49 1423.18 L2134.83 1423.18 L2134.83 391.228 L2099.49 391.228 L2099.49 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2099.49,391.228 2099.49,1423.18 2134.83,1423.18 2134.83,391.228 2099.49,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2117.16" cy="49.996" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2117.16" cy="168.326" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2117.16" cy="391.228" r="2"/>
+<path clip-path="url(#clip892)" d="M2158.39 114.665 L2158.39 1423.18 L2193.73 1423.18 L2193.73 114.665 L2158.39 114.665 L2158.39 114.665  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2158.39,114.665 2158.39,1423.18 2193.73,1423.18 2193.73,114.665 2158.39,114.665 "/>
+<path clip-path="url(#clip892)" d="M2158.39 391.228 L2158.39 1423.18 L2193.73 1423.18 L2193.73 391.228 L2158.39 391.228 L2158.39 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2158.39,391.228 2158.39,1423.18 2193.73,1423.18 2193.73,391.228 2158.39,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2176.06" cy="114.665" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2176.06" cy="391.228" r="2"/>
+<path clip-path="url(#clip892)" d="M2217.29 109.161 L2217.29 1423.18 L2252.63 1423.18 L2252.63 109.161 L2217.29 109.161 L2217.29 109.161  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2217.29,109.161 2217.29,1423.18 2252.63,1423.18 2252.63,109.161 2217.29,109.161 "/>
+<path clip-path="url(#clip892)" d="M2217.29 308.672 L2217.29 1423.18 L2252.63 1423.18 L2252.63 308.672 L2217.29 308.672 L2217.29 308.672  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2217.29,308.672 2217.29,1423.18 2252.63,1423.18 2252.63,308.672 2217.29,308.672 "/>
+<path clip-path="url(#clip892)" d="M2217.29 732.46 L2217.29 1423.18 L2252.63 1423.18 L2252.63 732.46 L2217.29 732.46 L2217.29 732.46  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2217.29,732.46 2217.29,1423.18 2252.63,1423.18 2252.63,732.46 2217.29,732.46 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2234.96" cy="109.161" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2234.96" cy="308.672" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2234.96" cy="732.46" r="2"/>
+<path clip-path="url(#clip892)" d="M2276.19 48.62 L2276.19 1423.18 L2311.53 1423.18 L2311.53 48.62 L2276.19 48.62 L2276.19 48.62  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2276.19,48.62 2276.19,1423.18 2311.53,1423.18 2311.53,48.62 2276.19,48.62 "/>
+<path clip-path="url(#clip892)" d="M2276.19 118.793 L2276.19 1423.18 L2311.53 1423.18 L2311.53 118.793 L2276.19 118.793 L2276.19 118.793  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2276.19,118.793 2276.19,1423.18 2311.53,1423.18 2311.53,118.793 2276.19,118.793 "/>
+<path clip-path="url(#clip892)" d="M2276.19 209.605 L2276.19 1423.18 L2311.53 1423.18 L2311.53 209.605 L2276.19 209.605 L2276.19 209.605  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2276.19,209.605 2276.19,1423.18 2311.53,1423.18 2311.53,209.605 2276.19,209.605 "/>
+<path clip-path="url(#clip892)" d="M2276.19 300.416 L2276.19 1423.18 L2311.53 1423.18 L2311.53 300.416 L2276.19 300.416 L2276.19 300.416  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2276.19,300.416 2276.19,1423.18 2311.53,1423.18 2311.53,300.416 2276.19,300.416 "/>
+<path clip-path="url(#clip892)" d="M2276.19 391.228 L2276.19 1423.18 L2311.53 1423.18 L2311.53 391.228 L2276.19 391.228 L2276.19 391.228  Z" fill="#90caf9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip892)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2276.19,391.228 2276.19,1423.18 2311.53,1423.18 2311.53,391.228 2276.19,391.228 "/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2293.86" cy="48.62" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2293.86" cy="118.793" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2293.86" cy="209.605" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2293.86" cy="300.416" r="2"/>
+<circle clip-path="url(#clip892)" style="fill:#90caf9; stroke:none; fill-opacity:0" cx="2293.86" cy="391.228" r="2"/>
+</svg>
diff --git a/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/index.html b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/index.html
new file mode 100644
index 00000000000..b9acb4ba5cd
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/cutting_stock_column_generation/index.html
@@ -0,0 +1,290 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Column generation · JuMP</title><meta name="title" content="Column generation · JuMP"/><meta property="og:title" content="Column generation · JuMP"/><meta property="twitter:title" content="Column generation · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox" checked/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../benders_decomposition/">Benders decomposition</a></li><li class="is-active"><a class="tocitem" href>Column generation</a><ul class="internal"><li><a class="tocitem" href="#Background"><span>Background</span></a></li><li><a class="tocitem" href="#Mathematical-formulation"><span>Mathematical formulation</span></a></li><li><a class="tocitem" href="#Column-generation-theory"><span>Column generation theory</span></a></li><li><a class="tocitem" href="#Choosing-the-initial-set-of-patterns"><span>Choosing the initial set of patterns</span></a></li><li><a class="tocitem" href="#The-base-problem"><span>The base problem</span></a></li><li><a class="tocitem" href="#Choosing-new-columns"><span>Choosing new columns</span></a></li><li><a class="tocitem" href="#Iterative-algorithm"><span>Iterative algorithm</span></a></li><li><a class="tocitem" href="#Looking-at-the-solution"><span>Looking at the solution</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Algorithms</a></li><li class="is-active"><a href>Column generation</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Column generation</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/algorithms/cutting_stock_column_generation.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Column-generation"><a class="docs-heading-anchor" href="#Column-generation">Column generation</a><a id="Column-generation-1"></a><a class="docs-heading-anchor-permalink" href="#Column-generation" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../cutting_stock_column_generation.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to demonstrate the column generation algorithm. As an example, it solves the <a href="https://en.wikipedia.org/wiki/Cutting_stock_problem">Cutting stock problem</a>.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import DataFrames
+import HiGHS
+import Plots
+import SparseArrays</code></pre><h2 id="Background"><a class="docs-heading-anchor" href="#Background">Background</a><a id="Background-1"></a><a class="docs-heading-anchor-permalink" href="#Background" title="Permalink"></a></h2><p>The cutting stock problem is about cutting large rolls of paper into smaller pieces.</p><p>We denote the set of possible sized pieces that a roll can be cut into by <span>$i\in 1,\ldots,I$</span>. Each piece <span>$i$</span> has a width, <span>$w_i$</span>, and a demand, <span>$d_i$</span>. The width of the large roll is <span>$W$</span>.</p><p>Our objective is to minimize the number of rolls needed to meet all demand.</p><p>Here&#39;s the data that we are going to use in this tutorial:</p><pre><code class="language-julia hljs">struct Piece
+    w::Float64
+    d::Int
+end
+
+struct Data
+    pieces::Vector{Piece}
+    W::Float64
+end
+
+function Base.show(io::IO, d::Data)
+    println(io, &quot;Data for the cutting stock problem:&quot;)
+    println(io, &quot;  W = $(d.W)&quot;)
+    println(io, &quot;with pieces:&quot;)
+    println(io, &quot;   i   w_i d_i&quot;)
+    println(io, &quot;  ------------&quot;)
+    for (i, p) in enumerate(d.pieces)
+        println(io, lpad(i, 4), &quot; &quot;, lpad(p.w, 5), &quot; &quot;, lpad(p.d, 3))
+    end
+    return
+end
+
+function get_data()
+    data = [
+        75.0 38
+        75.0 44
+        75.0 30
+        75.0 41
+        75.0 36
+        53.8 33
+        53.0 36
+        51.0 41
+        50.2 35
+        32.2 37
+        30.8 44
+        29.8 49
+        20.1 37
+        16.2 36
+        14.5 42
+        11.0 33
+        8.6 47
+        8.2 35
+        6.6 49
+        5.1 42
+    ]
+    return Data([Piece(data[i, 1], data[i, 2]) for i in axes(data, 1)], 100.0)
+end
+
+data = get_data()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Data for the cutting stock problem:
+  W = 100.0
+with pieces:
+   i   w_i d_i
+  ------------
+   1  75.0  38
+   2  75.0  44
+   3  75.0  30
+   4  75.0  41
+   5  75.0  36
+   6  53.8  33
+   7  53.0  36
+   8  51.0  41
+   9  50.2  35
+  10  32.2  37
+  11  30.8  44
+  12  29.8  49
+  13  20.1  37
+  14  16.2  36
+  15  14.5  42
+  16  11.0  33
+  17   8.6  47
+  18   8.2  35
+  19   6.6  49
+  20   5.1  42
+</code></pre><h2 id="Mathematical-formulation"><a class="docs-heading-anchor" href="#Mathematical-formulation">Mathematical formulation</a><a id="Mathematical-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Mathematical-formulation" title="Permalink"></a></h2><p>To formulate the cutting stock problem as a mixed-integer linear program, we assume that there is a set of large rolls <span>$j=1,\ldots,J$</span> to use. Then, we introduce two classes of decision variables:</p><ul><li><span>$x_{ij} \ge 0,\; \text{integer}, \; \forall i=1,\ldots,I,\; j=1,\ldots,J$</span></li><li><span>$y_{j} \in \{0, 1\},\; \forall j=1,\ldots,J.$</span></li></ul><p><span>$y_j$</span> is a binary variable that indicates if we use roll <span>$j$</span>, and <span>$x_{ij}$</span> counts how many pieces of size <span>$i$</span> that we cut from roll <span>$j$</span>.</p><p>Our mixed-integer linear program is therefore:</p><p class="math-container">\[\begin{align}
+    \min &amp; \sum\limits_{j=1}^J y_j \\
+    \;\;\text{s.t.} &amp; \sum\limits_{i=1}^N w_i x_{ij} \le W y_j &amp; \forall j=1,\ldots,J \\
+         &amp; \sum\limits_{j=1}^J x_{ij} \ge d_i &amp; \forall i=1,\ldots,I \\
+         &amp; x_{ij} \ge 0 &amp; \forall i=1,\ldots,N, j=1,\ldots,J \\
+         &amp; x_{ij} \in \mathbb{Z} &amp; \forall i=1,\ldots,I, j=1,\ldots,J \\
+         &amp; y_{j} \in \{0, 1\} &amp; \forall j=1,\ldots,J \\
+\end{align}\]</p><p>The objective is to minimize the number of rolls that we use, and the two constraints ensure that we respect the total width of each large roll and that we satisfy demand exactly.</p><p>The JuMP formulation of this model is:</p><pre><code class="language-julia hljs">I = length(data.pieces)
+J = 1_000  # Some large number
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:I, 1:J] &gt;= 0, Int)
+@variable(model, y[1:J], Bin)
+@objective(model, Min, sum(y))
+@constraint(model, [i in 1:I], sum(x[i, :]) &gt;= data.pieces[i].d)
+@constraint(
+    model,
+    [j in 1:J],
+    sum(data.pieces[i].w * x[i, j] for i in 1:I) &lt;= data.W * y[j],
+);</code></pre><p>Unfortunately, we can&#39;t solve this formulation for realistic instances because it takes a very long time to solve. (Try removing the time limit.)</p><pre><code class="language-julia hljs">set_time_limit_sec(model, 5.0)
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : TIME_LIMIT
+  Message from the solver:
+  &quot;kHighsModelStatusTimeLimit&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : 4.11000e+02
+  Objective bound    : 2.93000e+02
+  Relative gap       : 2.87105e-01
+
+* Work counters
+  Solve time (sec)   : 5.06210e+00
+  Simplex iterations : 16186
+  Barrier iterations : -1
+  Node count         : 0
+</code></pre><p>However, there is a formulation that solves much faster, and that is to use a column generation scheme.</p><h2 id="Column-generation-theory"><a class="docs-heading-anchor" href="#Column-generation-theory">Column generation theory</a><a id="Column-generation-theory-1"></a><a class="docs-heading-anchor-permalink" href="#Column-generation-theory" title="Permalink"></a></h2><p>The key insight for column generation is to recognize that feasible columns in the <span>$x$</span> matrix of variables encode <em>cutting patterns</em>.</p><p>For example, if we look only at the roll <span>$j=1$</span>, then a feasible solution is:</p><ul><li><span>$x_{1,1} = 1$</span> (1 unit of piece #1)</li><li><span>$x_{13,1} = 1$</span> (1 unit of piece #13)</li><li>All other <span>$x_{i,1} = 0$</span></li></ul><p>Another solution is</p><ul><li><span>$x_{20,1} = 19$</span> (19 unit of piece #20)</li><li>All other <span>$x_{i,1} = 0$</span></li></ul><p>Cutting patterns like <span>$x_{1,1} = 1$</span> and <span>$x_{2,1} = 1$</span> are infeasible because the combined length is greater than <span>$W$</span>.</p><p>Since there are a finite number of ways that we could cut a roll into a valid cutting pattern, we could create a set of all possible cutting patterns <span>$p = 1,\ldots,P$</span>, with data <span>$a_{i,p}$</span> indicating how many units of piece <span>$i$</span> we cut in pattern <span>$p$</span>. Then, we can formulate our mixed-integer linear program as:</p><p class="math-container">\[\begin{align}
+    \min &amp; \sum\limits_{p=1}^P x_p \\
+    \;\;\text{s.t.} &amp; \sum\limits_{p=1}^P a_{ip} x_p \ge d_i &amp; \forall i=1,\ldots,I \\
+         &amp; x_{p} \ge 0 &amp; \forall p=1,\ldots,P \\
+         &amp; x_{p} \in \mathbb{Z} &amp; \forall p=1,\ldots,P
+\end{align}\]</p><p>Unfortunately, there will be a very large number of these patterns, so it is often intractable to enumerate all columns <span>$p=1,\ldots,P$</span>.</p><p>Column generation is an iterative algorithm that starts with a small set of initial patterns, and then cleverly chooses new columns to add to the main MILP so that we find the optimal solution without having to enumerate every column.</p><h2 id="Choosing-the-initial-set-of-patterns"><a class="docs-heading-anchor" href="#Choosing-the-initial-set-of-patterns">Choosing the initial set of patterns</a><a id="Choosing-the-initial-set-of-patterns-1"></a><a class="docs-heading-anchor-permalink" href="#Choosing-the-initial-set-of-patterns" title="Permalink"></a></h2><p>For the initial set of patterns, we create a trivial cutting pattern which cuts as many units of piece <span>$i$</span> as will fit.</p><pre><code class="language-julia hljs">patterns = map(1:I) do i
+    n_pieces = floor(Int, data.W / data.pieces[i].w)
+    return SparseArrays.sparsevec([i], [n_pieces], I)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20-element Vector{SparseArrays.SparseVector{Int64, Int64}}:
+   [1]  =  1
+   [2]  =  1
+   [3]  =  1
+   [4]  =  1
+   [5]  =  1
+   [6]  =  1
+   [7]  =  1
+   [8]  =  1
+   [9]  =  1
+   [10]  =  3
+   [11]  =  3
+   [12]  =  3
+   [13]  =  4
+   [14]  =  6
+   [15]  =  6
+   [16]  =  9
+   [17]  =  11
+   [18]  =  12
+   [19]  =  15
+   [20]  =  19</code></pre><p>We can visualize the patterns as follows:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    cutting_locations(data::Data, pattern::SparseArrays.SparseVector)
+
+A function which returns a vector of the locations along the roll at which to
+cut in order to produce pattern `pattern`.
+&quot;&quot;&quot;
+function cutting_locations(data::Data, pattern::SparseArrays.SparseVector)
+    locations = Float64[]
+    offset = 0.0
+    for (i, c) in zip(SparseArrays.findnz(pattern)...)
+        for _ in 1:c
+            offset += data.pieces[i].w
+            push!(locations, offset)
+        end
+    end
+    return locations
+end
+
+function plot_patterns(data::Data, patterns)
+    plot = Plots.bar(;
+        xlims = (0, length(patterns) + 1),
+        ylims = (0, data.W),
+        xlabel = &quot;Pattern&quot;,
+        ylabel = &quot;Roll length&quot;,
+    )
+    for (i, p) in enumerate(patterns)
+        locations = cutting_locations(data, p)
+        Plots.bar!(
+            plot,
+            fill(i, length(locations)),
+            reverse(locations);
+            bar_width = 0.6,
+            label = false,
+            color = &quot;#90caf9&quot;,
+        )
+    end
+    return plot
+end
+
+plot_patterns(data, patterns)</code></pre><img src="703571d3.svg" alt="Example block output"/><h2 id="The-base-problem"><a class="docs-heading-anchor" href="#The-base-problem">The base problem</a><a id="The-base-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-base-problem" title="Permalink"></a></h2><p>Using the initial set of patterns, we can create and optimize our base model:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:length(patterns)] &gt;= 0, Int)
+@objective(model, Min, sum(x))
+@constraint(model, demand[i in 1:I], patterns[i]&#39; * x &gt;= data.pieces[i].d)
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : 4.21000e+02
+  Objective bound    : 4.21000e+02
+  Relative gap       : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 1.81437e-04
+  Simplex iterations : 0
+  Barrier iterations : -1
+  Node count         : 0
+</code></pre><p>This solution requires 421 rolls. This solution is sub-optimal because the model does not contain the full set of possible patterns.</p><p>How do we find a new column that leads to an improved solution?</p><h2 id="Choosing-new-columns"><a class="docs-heading-anchor" href="#Choosing-new-columns">Choosing new columns</a><a id="Choosing-new-columns-1"></a><a class="docs-heading-anchor-permalink" href="#Choosing-new-columns" title="Permalink"></a></h2><p>Column generation chooses a new column by relaxing the integrality constraint on <span>$x$</span> and looking at the dual variable <span>$\pi_i$</span> associated with demand constraint <span>$i$</span>.</p><p>For example, the dual of <code>demand[13]</code> is:</p><pre><code class="language-julia hljs">unset_integer.(x)
+optimize!(model)
+π_13 = dual(demand[13])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.25</code></pre><p>Using the economic interpretation of the dual variable, we can say that a one unit increase in demand for piece <span>$i$</span> will cost an extra <span>$\pi_i$</span> rolls. Alternatively, we can say that a one unit increase in the left-hand side (for example, due to a new cutting pattern) will <em>save</em> us <span>$\pi_i$</span> rolls. Therefore, we want a new column that maximizes the savings associated with the dual variables, while respecting the total width of the roll:</p><p class="math-container">\[\begin{align}
+    \max &amp; \sum\limits_{i=1}^I \pi_i y_i \\
+    \;\;\text{s.t.} &amp; \sum\limits_{i=1}^I w_i y_{i} \le W \\
+         &amp; y_{i} \ge 0 &amp; \forall i=1,\ldots,I \\
+         &amp; y_{i} \in \mathbb{Z} &amp; \forall i=1,\ldots,I \\
+\end{align}\]</p><p>If this problem, called the <em>pricing problem</em>, has an objective value greater than <span>$1$</span>, then we estimate than adding <code>y</code> as the coefficients of a new column will decrease the objective by more than the cost of an extra roll.</p><p>Here is code to solve the pricing problem:</p><pre><code class="language-julia hljs">function solve_pricing(data::Data, π::Vector{Float64})
+    I = length(π)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, y[1:I] &gt;= 0, Int)
+    @constraint(model, sum(data.pieces[i].w * y[i] for i in 1:I) &lt;= data.W)
+    @objective(model, Max, sum(π[i] * y[i] for i in 1:I))
+    optimize!(model)
+    number_of_rolls_saved = objective_value(model)
+    if number_of_rolls_saved &gt; 1 + 1e-8
+        # Benefit of pattern is more than the cost of a new roll plus some
+        # tolerance
+        return SparseArrays.sparse(round.(Int, value.(y)))
+    end
+    return nothing
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_pricing (generic function with 1 method)</code></pre><p>If we solve the pricing problem with an artificial dual vector:</p><pre><code class="language-julia hljs">solve_pricing(data, [1.0 / i for i in 1:I])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20-element SparseArrays.SparseVector{Int64, Int64} with 3 stored entries:
+  [1 ]  =  1
+  [17]  =  1
+  [20]  =  3</code></pre><p>the solution is a roll with 1 unit of piece #1, 1 unit of piece #17, and 3 units of piece #20.</p><p>If we solve the pricing problem with a dual vector of zeros, then the benefit of the new pattern is less than the cost of a roll, and so the function returns <code>nothing</code>:</p><pre><code class="language-julia hljs">solve_pricing(data, zeros(I))</code></pre><h2 id="Iterative-algorithm"><a class="docs-heading-anchor" href="#Iterative-algorithm">Iterative algorithm</a><a id="Iterative-algorithm-1"></a><a class="docs-heading-anchor-permalink" href="#Iterative-algorithm" title="Permalink"></a></h2><p>Now we can combine our base model with the pricing subproblem in an iterative column generation scheme:</p><pre><code class="language-julia hljs">while true
+    # Solve the linear relaxation
+    optimize!(model)
+    # Obtain a new dual vector
+    π = dual.(demand)
+    # Solve the pricing problem
+    new_pattern = solve_pricing(data, π)
+    # Stop iterating if there is no new pattern
+    if new_pattern === nothing
+        @info &quot;No new patterns, terminating the algorithm.&quot;
+        break
+    end
+    push!(patterns, new_pattern)
+    # Create a new column
+    push!(x, @variable(model, lower_bound = 0))
+    # Update the objective coefficient of the new column
+    set_objective_coefficient(model, x[end], 1.0)
+    # Update the non-zeros in the coefficient matrix
+    for (i, count) in zip(SparseArrays.findnz(new_pattern)...)
+        set_normalized_coefficient(demand[i], x[end], count)
+    end
+    println(&quot;Found new pattern. Total patterns = $(length(patterns))&quot;)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Found new pattern. Total patterns = 21
+Found new pattern. Total patterns = 22
+Found new pattern. Total patterns = 23
+Found new pattern. Total patterns = 24
+Found new pattern. Total patterns = 25
+Found new pattern. Total patterns = 26
+Found new pattern. Total patterns = 27
+Found new pattern. Total patterns = 28
+Found new pattern. Total patterns = 29
+Found new pattern. Total patterns = 30
+Found new pattern. Total patterns = 31
+Found new pattern. Total patterns = 32
+Found new pattern. Total patterns = 33
+Found new pattern. Total patterns = 34
+Found new pattern. Total patterns = 35
+<span class="sgr36"><span class="sgr1">[ Info: </span></span>No new patterns, terminating the algorithm.</code></pre><p>We found lots of new patterns. Here&#39;s pattern 21:</p><pre><code class="language-julia hljs">patterns[21]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20-element SparseArrays.SparseVector{Int64, Int64} with 3 stored entries:
+  [9 ]  =  1
+  [13]  =  2
+  [17]  =  1</code></pre><p>Let&#39;s have a look at the patterns now:</p><pre><code class="language-julia hljs">plot_patterns(data, patterns)</code></pre><img src="fa9bdba4.svg" alt="Example block output"/><h2 id="Looking-at-the-solution"><a class="docs-heading-anchor" href="#Looking-at-the-solution">Looking at the solution</a><a id="Looking-at-the-solution-1"></a><a class="docs-heading-anchor-permalink" href="#Looking-at-the-solution" title="Permalink"></a></h2><p>Let&#39;s see how many of each column we need:</p><pre><code class="language-julia hljs">solution = DataFrames.DataFrame([
+    (pattern = p, rolls = value(x_p)) for (p, x_p) in enumerate(x)
+])
+filter!(row -&gt; row.rolls &gt; 0, solution)</code></pre><div><div style = "float: left;"><span>16×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">pattern</th><th style = "text-align: left;">rolls</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">1</td><td style = "text-align: right;">38.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">2</td><td style = "text-align: right;">44.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">3</td><td style = "text-align: right;">30.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">21</td><td style = "text-align: right;">0.5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">22</td><td style = "text-align: right;">10.2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">23</td><td style = "text-align: right;">14.65</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">24</td><td style = "text-align: right;">23.1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">25</td><td style = "text-align: right;">11.25</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">26</td><td style = "text-align: right;">21.35</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">28</td><td style = "text-align: right;">4.3</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">29</td><td style = "text-align: right;">19.55</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">30</td><td style = "text-align: right;">11.25</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: right;">31</td><td style = "text-align: right;">17.45</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: right;">33</td><td style = "text-align: right;">36.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: right;">34</td><td style = "text-align: right;">11.4</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: right;">35</td><td style = "text-align: right;">41.0</td></tr></tbody></table></div><p>Since we solved a linear program, some of our columns have fractional solutions. We can create a integer feasible solution by rounding up the orders. This requires 341 rolls:</p><pre><code class="language-julia hljs">sum(ceil.(Int, solution.rolls))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">341</code></pre><p>Alternatively, we can re-introduce the integrality constraints and resolve the problem:</p><pre><code class="language-julia hljs">set_integer.(x)
+optimize!(model)
+solution = DataFrames.DataFrame([
+    (pattern = p, rolls = value(x_p)) for (p, x_p) in enumerate(x)
+])
+filter!(row -&gt; row.rolls &gt; 0, solution)</code></pre><div><div style = "float: left;"><span>17×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">pattern</th><th style = "text-align: left;">rolls</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">1</td><td style = "text-align: right;">38.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">2</td><td style = "text-align: right;">44.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">3</td><td style = "text-align: right;">30.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">5</td><td style = "text-align: right;">3.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">21</td><td style = "text-align: right;">5.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">22</td><td style = "text-align: right;">10.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">23</td><td style = "text-align: right;">18.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">24</td><td style = "text-align: right;">18.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">25</td><td style = "text-align: right;">14.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">26</td><td style = "text-align: right;">18.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">29</td><td style = "text-align: right;">17.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">30</td><td style = "text-align: right;">13.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: right;">31</td><td style = "text-align: right;">20.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: right;">32</td><td style = "text-align: right;">6.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: right;">33</td><td style = "text-align: right;">27.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: right;">34</td><td style = "text-align: right;">12.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: right;">35</td><td style = "text-align: right;">41.0</td></tr></tbody></table></div><p>This now requires 334 rolls:</p><pre><code class="language-julia hljs">sum(solution.rolls)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">334.0</code></pre><p>Note that this may not be the global minimum because we are not adding new columns during the solution of the mixed-integer problem <code>model</code> (an algorithm known as <a href="https://en.wikipedia.org/wiki/Branch_and_price">branch and price</a>). Nevertheless, the column generation algorithm typically finds good integer feasible solutions to an otherwise intractable optimization problem.</p><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><ul><li>Our objective function is to minimize the total number of rolls. What is the total length of waste? How does that compare to the total demand?</li><li>Writing the optimization algorithm is only part of the challenge. Can you develop a better way to communicate the solution to stakeholders?</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../benders_decomposition/">« Benders decomposition</a><a class="docs-footer-nextpage" href="../tsp_lazy_constraints/">Traveling Salesperson Problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/algorithms/parallelism/index.html b/previews/PR3525/tutorials/algorithms/parallelism/index.html
new file mode 100644
index 00000000000..0989a8306ea
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/parallelism/index.html
@@ -0,0 +1,129 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Parallelism · JuMP</title><meta name="title" content="Parallelism · JuMP"/><meta property="og:title" content="Parallelism · JuMP"/><meta property="twitter:title" content="Parallelism · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox" checked/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li class="is-active"><a class="tocitem" href>Parallelism</a><ul class="internal"><li><a class="tocitem" href="#Multi-threading-and-distributed-computing"><span>Multi-threading and distributed computing</span></a></li><li><a class="tocitem" href="#Using-parallelism-the-wrong-way"><span>Using parallelism the wrong way</span></a></li><li><a class="tocitem" href="#Using-parallelism-the-right-way"><span>Using parallelism the right way</span></a></li><li><a class="tocitem" href="#Other-types-of-parallelism"><span>Other types of parallelism</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Algorithms</a></li><li class="is-active"><a href>Parallelism</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Parallelism</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/algorithms/parallelism.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Parallelism"><a class="docs-heading-anchor" href="#Parallelism">Parallelism</a><a id="Parallelism-1"></a><a class="docs-heading-anchor-permalink" href="#Parallelism" title="Permalink"></a></h1><p>The purpose of this tutorial is to give a brief overview of parallelism in Julia as it pertains to JuMP, and to explain some of the things to be aware of when writing parallel algorithms involving JuMP models.</p><h2 id="Multi-threading-and-distributed-computing"><a class="docs-heading-anchor" href="#Multi-threading-and-distributed-computing">Multi-threading and distributed computing</a><a id="Multi-threading-and-distributed-computing-1"></a><a class="docs-heading-anchor-permalink" href="#Multi-threading-and-distributed-computing" title="Permalink"></a></h2><p>There are two main types of parallelism in Julia:</p><ol><li>Multi-threading</li><li>Distributed computing</li></ol><p>In multi-threading, multiple tasks are run in a single Julia process and share the same memory. In distributed computing, tasks are run in multiple Julia processes with independent memory spaces. This can include processes across multiple physical machines, such as in a high-performance computing cluster.</p><p>Choosing and understanding the type of parallelism you are using is important because the code you write for each type is different, and there are different limitations and benefits to each approach. However, the best choice is highly problem dependent, so you may want to experiment with both approaches to determine what works for your situation.</p><h3 id="Multi-threading"><a class="docs-heading-anchor" href="#Multi-threading">Multi-threading</a><a id="Multi-threading-1"></a><a class="docs-heading-anchor-permalink" href="#Multi-threading" title="Permalink"></a></h3><p>To use multi-threading with Julia, you must either start Julia with the command line flag <code>--threads=N</code>, or you must set the <code>JULIA_NUM_THREADS</code> environment variable before launching Julia. For this documentation, we set the environment variable to:</p><pre><code class="language-julia hljs">julia&gt; ENV[&quot;JULIA_NUM_THREADS&quot;]
+&quot;4&quot;</code></pre><p>You can check how many threads are available using:</p><pre><code class="language-julia hljs">julia&gt; Threads.nthreads()
+4</code></pre><p>The easiest way to use multi-threading in Julia is by placing the <code>Threads.@threads</code> macro in front of a <code>for</code>-loop:</p><pre><code class="language-julia hljs">julia&gt; @time begin
+           ids = Int[]
+           my_lock = Threads.ReentrantLock()
+           Threads.@threads for i in 1:Threads.nthreads()
+               global ids, my_lock
+               Threads.lock(my_lock) do
+                   push!(ids, Threads.threadid())
+               end
+               sleep(1.0)
+           end
+       end
+  1.037087 seconds (31.32 k allocations: 1.836 MiB, 2.02% compilation time)</code></pre><p>This for-loop sleeps for <code>1</code> second on each iteration. Thus, if it had executed sequentially, it should have taken the same number of seconds as there are threads available. Instead, it took only 1 second, showing that the iterations were executed simultaneously. We can verify this by checking the <code>Threads.threadid()</code> of the thread that executed each iteration:</p><pre><code class="language-julia hljs">julia&gt; ids
+4-element Vector{Int64}:
+ 2
+ 4
+ 1
+ 3</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>When working with threads, you need to avoid race conditions, in which two threads attempt to write to the same variable at the same time. In the above example we avoided a race condition by using <code>ReentrantLock</code>. See the <a href="https://docs.julialang.org/en/v1/manual/multi-threading/">Multi-threading</a> section of the Julia documentation for more details.</p></div></div><h3 id="Distributed-computing"><a class="docs-heading-anchor" href="#Distributed-computing">Distributed computing</a><a id="Distributed-computing-1"></a><a class="docs-heading-anchor-permalink" href="#Distributed-computing" title="Permalink"></a></h3><p>To use distributed computing with Julia, use the <code>Distributed</code> package:</p><pre><code class="language-julia hljs">julia&gt; import Distributed</code></pre><p>Like multi-threading, we need to tell Julia how many processes to add. We can do this either by starting Julia with the <code>-p N</code> command line argument, or by using <code>Distributed.addprocs</code>:</p><pre><code class="language-julia hljs">julia&gt; import Pkg
+
+julia&gt; project = Pkg.project();
+
+julia&gt; workers = Distributed.addprocs(4; exeflags = &quot;--project=$(project.path)&quot;)
+4-element Vector{Int64}:
+ 2
+ 3
+ 4
+ 5</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Not loading the parent environment with <code>--project</code> is a common mistake.</p></div></div><p>The added processes are &quot;worker&quot; processes that we can use to do computation with. They are orchestrated by the process with the id <code>1</code>. You can check what process the code is currently running on using <code>Distributed.myid()</code></p><pre><code class="language-julia hljs">julia&gt; Distributed.myid()
+1</code></pre><p>As a general rule, to get maximum performance you should add as many processes as you have logical cores available.</p><p>Unlike the <code>for</code>-loop approach of multi-threading, distributed computing extends the Julia <code>map</code> function to a &quot;parallel-map&quot; function <code>Distributed.pmap</code>. For each element in the list of arguments to map over, Julia will copy the element to an idle worker process and evaluate the function, passing the element as an input argument.</p><pre><code class="language-julia hljs">julia&gt; function hard_work(i::Int)
+           sleep(1.0)
+           return Distributed.myid()
+       end
+hard_work (generic function with 1 method)
+
+julia&gt; Distributed.pmap(hard_work, 1:4)
+ERROR: On worker 2:
+UndefVarError: #hard_work not defined
+Stacktrace:
+[...]</code></pre><p>Unfortunately, if you try this code directly, you will get an error message that says <code>On worker 2: UndefVarError: hard_work not defined</code>. The error is thrown because, although process <code>1</code> knows what the <code>hard_work</code> function is, the worker processes do not.</p><p>To fix the error, we need to use <code>Distributed.@everywhere</code>, which evaluates the code on every process:</p><pre><code class="language-julia hljs">julia&gt; Distributed.@everywhere begin
+           function hard_work(i::Int)
+               sleep(1.0)
+               return Distributed.myid()
+           end
+       end</code></pre><p>Now if we run <code>pmap</code>, we see that it took only 1 second instead of 4, and that it executed on each of the worker processes:</p><pre><code class="language-julia hljs">julia&gt; @time ids = Distributed.pmap(hard_work, 1:4)
+  1.202006 seconds (216.39 k allocations: 13.301 MiB, 4.07% compilation time)
+4-element Vector{Int64}:
+ 2
+ 3
+ 5
+ 4</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>For more information, read the Julia documentation <a href="https://docs.julialang.org/en/v1/manual/distributed-computing/">Distributed Computing</a>.</p></div></div><h2 id="Using-parallelism-the-wrong-way"><a class="docs-heading-anchor" href="#Using-parallelism-the-wrong-way">Using parallelism the wrong way</a><a id="Using-parallelism-the-wrong-way-1"></a><a class="docs-heading-anchor-permalink" href="#Using-parallelism-the-wrong-way" title="Permalink"></a></h2><h3 id="When-building-a-JuMP-model"><a class="docs-heading-anchor" href="#When-building-a-JuMP-model">When building a JuMP model</a><a id="When-building-a-JuMP-model-1"></a><a class="docs-heading-anchor-permalink" href="#When-building-a-JuMP-model" title="Permalink"></a></h3><p>For large problems, building the model in JuMP can be a bottleneck, and you may consider trying to write code that builds the model in parallel, for example, by wrapping a <code>for</code>-loop that adds constraints with <code>Threads.@threads</code>.</p><p>Unfortunately, you cannot build a JuMP model in parallel, and attempting to do so may error or produce incorrect results.</p><p>In most cases, we find that the reason for the bottleneck is not JuMP, but in how you are constructing the problem data, and that with changes, it is possible to build a model in a way that is not the bottleneck in the solution process.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Looking for help to make your code run faster? Ask for help on the <a href="https://jump.dev/forum">community forum</a>. Make sure to include a reproducible example of your code.</p></div></div><h3 id="With-a-single-JuMP-model"><a class="docs-heading-anchor" href="#With-a-single-JuMP-model">With a single JuMP model</a><a id="With-a-single-JuMP-model-1"></a><a class="docs-heading-anchor-permalink" href="#With-a-single-JuMP-model" title="Permalink"></a></h3><p>A common approach people try is to use parallelism with a single JuMP model. For example, to optimize a model over multiple right-hand side vectors, you may try:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x)
+@objective(model, Min, x)
+solutions = Pair{Int,Float64}[]
+my_lock = Threads.ReentrantLock()
+Threads.@threads for i in 1:10
+    set_lower_bound(x, i)
+    optimize!(model)
+    Threads.lock(my_lock) do
+        push!(solutions, i =&gt; objective_value(model))
+    end
+end</code></pre><p>This will not work, and attempting to do so may error, crash Julia or produce incorrect results.</p><h2 id="Using-parallelism-the-right-way"><a class="docs-heading-anchor" href="#Using-parallelism-the-right-way">Using parallelism the right way</a><a id="Using-parallelism-the-right-way-1"></a><a class="docs-heading-anchor-permalink" href="#Using-parallelism-the-right-way" title="Permalink"></a></h2><p>To use parallelism with JuMP, the simplest rule to remember is that each worker must have its own instance of a JuMP model.</p><h3 id="With-multi-threading"><a class="docs-heading-anchor" href="#With-multi-threading">With multi-threading</a><a id="With-multi-threading-1"></a><a class="docs-heading-anchor-permalink" href="#With-multi-threading" title="Permalink"></a></h3><p>With multi-threading, create a new instance of <code>model</code> in each iteration of the for-loop:</p><pre><code class="language-julia hljs">julia&gt; using JuMP
+
+julia&gt; import HiGHS
+
+julia&gt; solutions = Pair{Int,Float64}[]
+
+julia&gt; my_lock = Threads.ReentrantLock();
+
+julia&gt; Threads.@threads for i in 1:10
+           model = Model(HiGHS.Optimizer)
+           set_silent(model)
+           set_attribute(model, MOI.NumberOfThreads(), 1)
+           @variable(model, x)
+           @objective(model, Min, x)
+           set_lower_bound(x, i)
+           optimize!(model)
+           Threads.lock(my_lock) do
+               push!(solutions, i =&gt; objective_value(model))
+           end
+       end
+
+julia&gt; solutions
+10-element Vector{Pair{Int64, Float64}}:
+  7 =&gt; 7.0
+  4 =&gt; 4.0
+  1 =&gt; 1.0
+  9 =&gt; 9.0
+  5 =&gt; 5.0
+  8 =&gt; 8.0
+ 10 =&gt; 10.0
+  2 =&gt; 2.0
+  6 =&gt; 6.0
+  3 =&gt; 3.0</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>For some solvers, it may be necessary to limit the number of threads used internally by the solver to 1 by setting the <a href="../../../moi/reference/models/#MathOptInterface.NumberOfThreads"><code>MOI.NumberOfThreads</code></a> attribute.</p></div></div><h3 id="With-distributed-computing"><a class="docs-heading-anchor" href="#With-distributed-computing">With distributed computing</a><a id="With-distributed-computing-1"></a><a class="docs-heading-anchor-permalink" href="#With-distributed-computing" title="Permalink"></a></h3><p>With distributed computing, remember to evaluate all of the code on all of the processes using <code>Distributed.@everywhere</code>, and then write a function which creates a new instance of the model on every evaluation:</p><pre><code class="language-julia hljs">julia&gt; Distributed.@everywhere begin
+           using JuMP
+           import HiGHS
+       end
+
+julia&gt; Distributed.@everywhere begin
+           function solve_model_with_right_hand_side(i)
+               model = Model(HiGHS.Optimizer)
+               set_silent(model)
+               @variable(model, x)
+               @objective(model, Min, x)
+               set_lower_bound(x, i)
+               optimize!(model)
+               return objective_value(model)
+           end
+       end
+
+julia&gt; solutions = Distributed.pmap(solve_model_with_right_hand_side, 1:10)
+10-element Vector{Float64}:
+  1.0
+  2.0
+  3.0
+  4.0
+  5.0
+  6.0
+  7.0
+  8.0
+  9.0
+ 10.0</code></pre><h2 id="Other-types-of-parallelism"><a class="docs-heading-anchor" href="#Other-types-of-parallelism">Other types of parallelism</a><a id="Other-types-of-parallelism-1"></a><a class="docs-heading-anchor-permalink" href="#Other-types-of-parallelism" title="Permalink"></a></h2><h3 id="GPU"><a class="docs-heading-anchor" href="#GPU">GPU</a><a id="GPU-1"></a><a class="docs-heading-anchor-permalink" href="#GPU" title="Permalink"></a></h3><p>JuMP does not support GPU programming, and few solvers support execution on a GPU.</p><h3 id="Parallelism-within-the-solver"><a class="docs-heading-anchor" href="#Parallelism-within-the-solver">Parallelism within the solver</a><a id="Parallelism-within-the-solver-1"></a><a class="docs-heading-anchor-permalink" href="#Parallelism-within-the-solver" title="Permalink"></a></h3><p>Many solvers use parallelism internally. For example, commercial solvers like <a href="https://github.com/jump-dev/Gurobi.jl">Gurobi</a> and <a href="https://github.com/jump-dev/CPLEX.jl">CPLEX</a> both parallelize the search in branch-and-bound. Solvers supporting internal parallelism will typically support the <a href="../../../moi/reference/models/#MathOptInterface.NumberOfThreads"><code>MOI.NumberOfThreads</code></a> attribute, which you can set using <a href="../../../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../tsp_lazy_constraints/">« Traveling Salesperson Problem</a><a class="docs-footer-nextpage" href="../../applications/power_systems/">Power Systems »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints.jl b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints.jl
new file mode 100644
index 00000000000..62eb3d94978
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints.jl
@@ -0,0 +1,277 @@
+# Copyright (c) 2022 Daniel Schermer and contributors                            #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # [Traveling Salesperson Problem](@id tsp_lazy)
+
+# **This tutorial was originally contributed by Daniel Schermer.**
+
+# This tutorial describes how to implement the
+# [Traveling Salesperson Problem](https://en.wikipedia.org/wiki/Travelling_salesman_problem)
+# in JuMP using solver-independent lazy constraints that dynamically separate
+# subtours. To be more precise, we use lazy constraints to cut off infeasible
+# subtours only when necessary and not before needed.
+
+# It uses the following packages:
+
+using JuMP
+import GLPK
+import Random
+import Plots
+
+# ## [Mathematical Formulation](@id tsp_model)
+
+# Assume that we are given a complete graph $\mathcal{G}(V,E)$ where $V$ is the
+# set of vertices (or cities) and $E$ is the set of edges (or roads). For each
+# pair of vertices $i, j \in V, i \neq j$ the edge $(i,j) \in E$ is associated
+# with a weight (or distance) $d_{ij} \in \mathbb{R}^+$.
+
+# For this tutorial, we assume the problem to be symmetric, that is,
+# $d_{ij} = d_{ji} \, \forall i,j \in V$.
+
+# In the Traveling Salesperson Problem, we are tasked with finding a tour with
+# minimal length that visits every vertex exactly once and then returns to the
+# point of origin, that is, a *Hamiltonian cycle* with minimal weight.
+#
+# To model the problem, we introduce a binary variable,
+# $x_{ij} \in \{0,1\} \; \forall i, j \in V$, that indicates if edge $(i,j)$ is
+# part of the tour or not. Using these variables, the Traveling Salesperson
+# Problem can be modeled as the following integer linear program.
+
+# ### [Objective Function](@id tsp_objective)
+
+# The objective is to minimize the length of the tour (due to the assumed
+# symmetry, the second sum only contains $i<j$):
+# ```math
+# \text{min } \sum_{i \in V}  \sum_{j \in V, i < j} d_{ij} x_{ij}.
+# ```
+
+# Note that it is also possible to use the following objective function instead:
+# ```math
+# \text{min } \sum_{i \in V}  \sum_{j \in V} \dfrac{d_{ij} x_{ij}}{2}.
+# ```
+
+# ### [Constraints](@id tsp_constraints)
+
+# There are four classes of constraints in our formulation.
+
+# First, due to the presumed symmetry, the following constraints must hold:
+# ```math
+# x_{ij} = x_{ji} \quad \forall i,j \in V.
+# ```
+
+# Second, for each vertex $i$, exactly two edges must be selected that connect it
+# to other vertices $j$ in the graph $G$:
+# ```math
+# \sum_{j \in V} x_{ij} = 2 \quad \forall i \in V.
+# ```
+
+# Third, we do not permit loops to occur:
+# ```math
+# x_{ii} = 0 \quad \forall i \in V.
+# ```
+
+# The fourth constraint is more complicated. A major difficulty of the Traveling
+# Salesperson Problem arises from the fact that we need to prevent *subtours*,
+# that is, several distinct Hamiltonian cycles existing on subgraphs of $G$.
+
+# Note that the previous constraints *do not* guarantee that the solution will
+# be free of subtours. To this end, by $S$ we label a subset of vertices.
+# Then, for each proper subset $S \subset V$, the following constraints
+# guarantee that no subtour may occur:
+# ```math
+# \sum_{i \in S} \sum_{j \in S, i < j} x_{ij} \leq \vert S \vert - 1 \quad \forall S \subset V.
+# ```
+
+# Problematically, we require exponentially many of these constraints as
+# $\vert V \vert$ increases. Therefore, we will add these constraints only when
+# necessary.
+
+# ## [Implementation](@id tsp_implementation)
+
+# There are two ways we can eliminate subtours in JuMP, both of which will be
+# shown in what follows:
+# - iteratively solving a new model that incorporates previously identified
+#   subtours,
+# - or adding violated subtours as *lazy constraints*.
+
+# ### Data
+
+# The vertices are assumed to be randomly distributed in the Euclidean space;
+# thus, the weight (distance) of each edge is defined as follows.
+
+function generate_distance_matrix(n; random_seed = 1)
+    rng = Random.MersenneTwister(random_seed)
+    X = 100 * rand(rng, n)
+    Y = 100 * rand(rng, n)
+    d = [sqrt((X[i] - X[j])^2 + (Y[i] - Y[j])^2) for i in 1:n, j in 1:n]
+    return X, Y, d
+end
+
+n = 40
+X, Y, d = generate_distance_matrix(n)
+
+# For the JuMP model, we first initialize the model object. Then, we create the
+# binary decision variables and add the objective function and constraints. By
+# defining the `x` matrix as `Symmetric`, we do not need to add explicit
+# constraints that `x[i, j] == x[j, i]`.
+
+function build_tsp_model(d, n)
+    model = Model(GLPK.Optimizer)
+    @variable(model, x[1:n, 1:n], Bin, Symmetric)
+    @objective(model, Min, sum(d .* x) / 2)
+    @constraint(model, [i in 1:n], sum(x[i, :]) == 2)
+    @constraint(model, [i in 1:n], x[i, i] == 0)
+    return model
+end
+
+# To search for violated constraints, based on the edges that are currently in
+# the solution (that is, those that have value $x_{ij} = 1$), we identify the
+# shortest cycle through the function `subtour()`. Whenever a subtour has been
+# identified, a constraint corresponding to the form above can be added to the
+# model.
+
+function subtour(edges::Vector{Tuple{Int,Int}}, n)
+    shortest_subtour, unvisited = collect(1:n), Set(collect(1:n))
+    while !isempty(unvisited)
+        this_cycle, neighbors = Int[], unvisited
+        while !isempty(neighbors)
+            current = pop!(neighbors)
+            push!(this_cycle, current)
+            if length(this_cycle) > 1
+                pop!(unvisited, current)
+            end
+            neighbors =
+                [j for (i, j) in edges if i == current && j in unvisited]
+        end
+        if length(this_cycle) < length(shortest_subtour)
+            shortest_subtour = this_cycle
+        end
+    end
+    return shortest_subtour
+end
+
+# Let us declare a helper function `selected_edges()` that will be repeatedly
+# used in what follows.
+
+function selected_edges(x::Matrix{Float64}, n)
+    return Tuple{Int,Int}[(i, j) for i in 1:n, j in 1:n if x[i, j] > 0.5]
+end
+
+# Other helper functions for computing subtours:
+
+subtour(x::Matrix{Float64}) = subtour(selected_edges(x, size(x, 1)), size(x, 1))
+subtour(x::AbstractMatrix{VariableRef}) = subtour(value.(x))
+
+# ### Iterative method
+
+# An iterative way of eliminating subtours is the following.
+
+# However, it is reasonable to assume that this is not the most efficient way:
+# whenever a new subtour elimination constraint is added to the model, the
+# optimization has to start from the very beginning.
+
+# That way, the solver will repeatedly discard useful information encountered
+# during previous solves (for example, all cuts, the incumbent solution, or lower
+# bounds).
+
+# !!! info
+#     Note that, in principle, it would also be feasible to add all subtours
+#     (instead of just the shortest one) to the model. However, preventing just
+#     the shortest cycle is often sufficient for breaking other subtours and
+#     will keep the model size smaller.
+
+iterative_model = build_tsp_model(d, n)
+optimize!(iterative_model)
+time_iterated = solve_time(iterative_model)
+cycle = subtour(iterative_model[:x])
+while 1 < length(cycle) < n
+    println("Found cycle of length $(length(cycle))")
+    S = [(i, j) for (i, j) in Iterators.product(cycle, cycle) if i < j]
+    @constraint(
+        iterative_model,
+        sum(iterative_model[:x][i, j] for (i, j) in S) <= length(cycle) - 1,
+    )
+    optimize!(iterative_model)
+    global time_iterated += solve_time(iterative_model)
+    global cycle = subtour(iterative_model[:x])
+end
+
+objective_value(iterative_model)
+
+#-
+
+time_iterated
+
+# As a quick sanity check, we visualize the optimal tour to verify that no
+# subtour is present:
+
+function plot_tour(X, Y, x)
+    plot = Plots.plot()
+    for (i, j) in selected_edges(x, size(x, 1))
+        Plots.plot!([X[i], X[j]], [Y[i], Y[j]]; legend = false)
+    end
+    return plot
+end
+
+plot_tour(X, Y, value.(iterative_model[:x]))
+
+# ### Lazy constraint method
+
+# A more sophisticated approach makes use of _lazy constraints_. To be more
+# precise, we do this through the `subtour_elimination_callback()` below, which
+# is only run whenever we encounter a new integer-feasible solution.
+
+lazy_model = build_tsp_model(d, n)
+function subtour_elimination_callback(cb_data)
+    status = callback_node_status(cb_data, lazy_model)
+    if status != MOI.CALLBACK_NODE_STATUS_INTEGER
+        return  # Only run at integer solutions
+    end
+    cycle = subtour(callback_value.(cb_data, lazy_model[:x]))
+    if !(1 < length(cycle) < n)
+        return  # Only add a constraint if there is a cycle
+    end
+    println("Found cycle of length $(length(cycle))")
+    S = [(i, j) for (i, j) in Iterators.product(cycle, cycle) if i < j]
+    con = @build_constraint(
+        sum(lazy_model[:x][i, j] for (i, j) in S) <= length(cycle) - 1,
+    )
+    MOI.submit(lazy_model, MOI.LazyConstraint(cb_data), con)
+    return
+end
+set_attribute(
+    lazy_model,
+    MOI.LazyConstraintCallback(),
+    subtour_elimination_callback,
+)
+optimize!(lazy_model)
+objective_value(lazy_model)
+
+# This finds the same optimal tour:
+
+plot_tour(X, Y, value.(lazy_model[:x]))
+
+# Surprisingly, for this particular model with GLPK, the solution time is worse
+# than the iterative method:
+
+time_lazy = solve_time(lazy_model)
+
+# In most other cases and solvers, however, the lazy time should be faster than
+# the iterative method.
diff --git a/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/43baf431.svg b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/43baf431.svg
new file mode 100644
index 00000000000..1e7eaaa8d40
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/43baf431.svg
@@ -0,0 +1,122 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip010">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip010)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip011">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip010)" d="M169.121 1486.45 L2352.76 1486.45 L2352.76 47.2441 L169.121 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip012">
+    <rect x="169" y="47" width="2185" height="1440"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="214.497,1486.45 214.497,47.2441 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="733.661,1486.45 733.661,47.2441 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1252.83,1486.45 1252.83,47.2441 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1771.99,1486.45 1771.99,47.2441 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2291.15,1486.45 2291.15,47.2441 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="214.497,1486.45 214.497,1467.55 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="733.661,1486.45 733.661,1467.55 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1252.83,1486.45 1252.83,1467.55 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1771.99,1486.45 1771.99,1467.55 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2291.15,1486.45 2291.15,1467.55 "/>
+<path clip-path="url(#clip010)" d="M214.497 1517.37 Q210.886 1517.37 209.058 1520.93 Q207.252 1524.47 207.252 1531.6 Q207.252 1538.71 209.058 1542.27 Q210.886 1545.82 214.497 1545.82 Q218.132 1545.82 219.937 1542.27 Q221.766 1538.71 221.766 1531.6 Q221.766 1524.47 219.937 1520.93 Q218.132 1517.37 214.497 1517.37 M214.497 1513.66 Q220.308 1513.66 223.363 1518.27 Q226.442 1522.85 226.442 1531.6 Q226.442 1540.33 223.363 1544.94 Q220.308 1549.52 214.497 1549.52 Q208.687 1549.52 205.609 1544.94 Q202.553 1540.33 202.553 1531.6 Q202.553 1522.85 205.609 1518.27 Q208.687 1513.66 214.497 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M712.932 1544.91 L729.252 1544.91 L729.252 1548.85 L707.307 1548.85 L707.307 1544.91 Q709.969 1542.16 714.553 1537.53 Q719.159 1532.88 720.34 1531.53 Q722.585 1529.01 723.465 1527.27 Q724.367 1525.51 724.367 1523.82 Q724.367 1521.07 722.423 1519.33 Q720.502 1517.6 717.4 1517.6 Q715.201 1517.6 712.747 1518.36 Q710.317 1519.13 707.539 1520.68 L707.539 1515.95 Q710.363 1514.82 712.817 1514.24 Q715.27 1513.66 717.307 1513.66 Q722.678 1513.66 725.872 1516.35 Q729.066 1519.03 729.066 1523.52 Q729.066 1525.65 728.256 1527.57 Q727.469 1529.47 725.363 1532.07 Q724.784 1532.74 721.682 1535.95 Q718.58 1539.15 712.932 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M739.113 1514.29 L757.469 1514.29 L757.469 1518.22 L743.395 1518.22 L743.395 1526.7 Q744.414 1526.35 745.432 1526.19 Q746.451 1526 747.469 1526 Q753.256 1526 756.636 1529.17 Q760.015 1532.34 760.015 1537.76 Q760.015 1543.34 756.543 1546.44 Q753.071 1549.52 746.752 1549.52 Q744.576 1549.52 742.307 1549.15 Q740.062 1548.78 737.654 1548.04 L737.654 1543.34 Q739.738 1544.47 741.96 1545.03 Q744.182 1545.58 746.659 1545.58 Q750.664 1545.58 753.001 1543.48 Q755.339 1541.37 755.339 1537.76 Q755.339 1534.15 753.001 1532.04 Q750.664 1529.94 746.659 1529.94 Q744.784 1529.94 742.909 1530.35 Q741.057 1530.77 739.113 1531.65 L739.113 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M1227.52 1514.29 L1245.88 1514.29 L1245.88 1518.22 L1231.81 1518.22 L1231.81 1526.7 Q1232.83 1526.35 1233.84 1526.19 Q1234.86 1526 1235.88 1526 Q1241.67 1526 1245.05 1529.17 Q1248.43 1532.34 1248.43 1537.76 Q1248.43 1543.34 1244.95 1546.44 Q1241.48 1549.52 1235.16 1549.52 Q1232.99 1549.52 1230.72 1549.15 Q1228.47 1548.78 1226.07 1548.04 L1226.07 1543.34 Q1228.15 1544.47 1230.37 1545.03 Q1232.59 1545.58 1235.07 1545.58 Q1239.08 1545.58 1241.41 1543.48 Q1243.75 1541.37 1243.75 1537.76 Q1243.75 1534.15 1241.41 1532.04 Q1239.08 1529.94 1235.07 1529.94 Q1233.2 1529.94 1231.32 1530.35 Q1229.47 1530.77 1227.52 1531.65 L1227.52 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M1267.64 1517.37 Q1264.03 1517.37 1262.2 1520.93 Q1260.39 1524.47 1260.39 1531.6 Q1260.39 1538.71 1262.2 1542.27 Q1264.03 1545.82 1267.64 1545.82 Q1271.27 1545.82 1273.08 1542.27 Q1274.91 1538.71 1274.91 1531.6 Q1274.91 1524.47 1273.08 1520.93 Q1271.27 1517.37 1267.64 1517.37 M1267.64 1513.66 Q1273.45 1513.66 1276.51 1518.27 Q1279.58 1522.85 1279.58 1531.6 Q1279.58 1540.33 1276.51 1544.94 Q1273.45 1549.52 1267.64 1549.52 Q1261.83 1549.52 1258.75 1544.94 Q1255.7 1540.33 1255.7 1531.6 Q1255.7 1522.85 1258.75 1518.27 Q1261.83 1513.66 1267.64 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M1745.84 1514.29 L1768.07 1514.29 L1768.07 1516.28 L1755.52 1548.85 L1750.63 1548.85 L1762.44 1518.22 L1745.84 1518.22 L1745.84 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M1777.23 1514.29 L1795.59 1514.29 L1795.59 1518.22 L1781.51 1518.22 L1781.51 1526.7 Q1782.53 1526.35 1783.55 1526.19 Q1784.57 1526 1785.59 1526 Q1791.38 1526 1794.76 1529.17 Q1798.13 1532.34 1798.13 1537.76 Q1798.13 1543.34 1794.66 1546.44 Q1791.19 1549.52 1784.87 1549.52 Q1782.7 1549.52 1780.43 1549.15 Q1778.18 1548.78 1775.77 1548.04 L1775.77 1543.34 Q1777.86 1544.47 1780.08 1545.03 Q1782.3 1545.58 1784.78 1545.58 Q1788.78 1545.58 1791.12 1543.48 Q1793.46 1541.37 1793.46 1537.76 Q1793.46 1534.15 1791.12 1532.04 Q1788.78 1529.94 1784.78 1529.94 Q1782.9 1529.94 1781.03 1530.35 Q1779.18 1530.77 1777.23 1531.65 L1777.23 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M2250.76 1544.91 L2258.4 1544.91 L2258.4 1518.55 L2250.09 1520.21 L2250.09 1515.95 L2258.35 1514.29 L2263.03 1514.29 L2263.03 1544.91 L2270.67 1544.91 L2270.67 1548.85 L2250.76 1548.85 L2250.76 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M2290.11 1517.37 Q2286.5 1517.37 2284.67 1520.93 Q2282.87 1524.47 2282.87 1531.6 Q2282.87 1538.71 2284.67 1542.27 Q2286.5 1545.82 2290.11 1545.82 Q2293.75 1545.82 2295.55 1542.27 Q2297.38 1538.71 2297.38 1531.6 Q2297.38 1524.47 2295.55 1520.93 Q2293.75 1517.37 2290.11 1517.37 M2290.11 1513.66 Q2295.92 1513.66 2298.98 1518.27 Q2302.06 1522.85 2302.06 1531.6 Q2302.06 1540.33 2298.98 1544.94 Q2295.92 1549.52 2290.11 1549.52 Q2284.3 1549.52 2281.22 1544.94 Q2278.17 1540.33 2278.17 1531.6 Q2278.17 1522.85 2281.22 1518.27 Q2284.3 1513.66 2290.11 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M2320.27 1517.37 Q2316.66 1517.37 2314.83 1520.93 Q2313.03 1524.47 2313.03 1531.6 Q2313.03 1538.71 2314.83 1542.27 Q2316.66 1545.82 2320.27 1545.82 Q2323.91 1545.82 2325.71 1542.27 Q2327.54 1538.71 2327.54 1531.6 Q2327.54 1524.47 2325.71 1520.93 Q2323.91 1517.37 2320.27 1517.37 M2320.27 1513.66 Q2326.08 1513.66 2329.14 1518.27 Q2332.22 1522.85 2332.22 1531.6 Q2332.22 1540.33 2329.14 1544.94 Q2326.08 1549.52 2320.27 1549.52 Q2314.46 1549.52 2311.38 1544.94 Q2308.33 1540.33 2308.33 1531.6 Q2308.33 1522.85 2311.38 1518.27 Q2314.46 1513.66 2320.27 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,1464.39 2352.76,1464.39 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,1119.64 2352.76,1119.64 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,774.897 2352.76,774.897 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,430.151 2352.76,430.151 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,85.4059 2352.76,85.4059 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1486.45 169.121,47.2441 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1464.39 188.019,1464.39 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1119.64 188.019,1119.64 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,774.897 188.019,774.897 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,430.151 188.019,430.151 "/>
+<polyline clip-path="url(#clip010)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,85.4059 188.019,85.4059 "/>
+<path clip-path="url(#clip010)" d="M121.177 1450.19 Q117.566 1450.19 115.737 1453.75 Q113.932 1457.29 113.932 1464.42 Q113.932 1471.53 115.737 1475.09 Q117.566 1478.64 121.177 1478.64 Q124.811 1478.64 126.617 1475.09 Q128.445 1471.53 128.445 1464.42 Q128.445 1457.29 126.617 1453.75 Q124.811 1450.19 121.177 1450.19 M121.177 1446.48 Q126.987 1446.48 130.043 1451.09 Q133.121 1455.67 133.121 1464.42 Q133.121 1473.15 130.043 1477.76 Q126.987 1482.34 121.177 1482.34 Q115.367 1482.34 112.288 1477.76 Q109.233 1473.15 109.233 1464.42 Q109.233 1455.67 112.288 1451.09 Q115.367 1446.48 121.177 1446.48 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M86.0382 1132.99 L102.358 1132.99 L102.358 1136.92 L80.4133 1136.92 L80.4133 1132.99 Q83.0753 1130.23 87.6586 1125.6 Q92.2651 1120.95 93.4456 1119.61 Q95.691 1117.08 96.5706 1115.35 Q97.4734 1113.59 97.4734 1111.9 Q97.4734 1109.14 95.5289 1107.41 Q93.6076 1105.67 90.5058 1105.67 Q88.3067 1105.67 85.8531 1106.44 Q83.4225 1107.2 80.6447 1108.75 L80.6447 1104.03 Q83.4688 1102.89 85.9225 1102.32 Q88.3762 1101.74 90.4132 1101.74 Q95.7836 1101.74 98.978 1104.42 Q102.172 1107.11 102.172 1111.6 Q102.172 1113.73 101.362 1115.65 Q100.575 1117.55 98.4687 1120.14 Q97.89 1120.81 94.7882 1124.03 Q91.6864 1127.22 86.0382 1132.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M112.219 1102.36 L130.575 1102.36 L130.575 1106.3 L116.501 1106.3 L116.501 1114.77 Q117.52 1114.42 118.538 1114.26 Q119.557 1114.08 120.575 1114.08 Q126.362 1114.08 129.742 1117.25 Q133.121 1120.42 133.121 1125.83 Q133.121 1131.41 129.649 1134.51 Q126.177 1137.59 119.857 1137.59 Q117.682 1137.59 115.413 1137.22 Q113.168 1136.85 110.76 1136.11 L110.76 1131.41 Q112.844 1132.55 115.066 1133.1 Q117.288 1133.66 119.765 1133.66 Q123.77 1133.66 126.107 1131.55 Q128.445 1129.45 128.445 1125.83 Q128.445 1122.22 126.107 1120.12 Q123.77 1118.01 119.765 1118.01 Q117.89 1118.01 116.015 1118.43 Q114.163 1118.84 112.219 1119.72 L112.219 1102.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M81.0614 757.617 L99.4178 757.617 L99.4178 761.552 L85.3438 761.552 L85.3438 770.024 Q86.3623 769.677 87.3808 769.515 Q88.3993 769.33 89.4178 769.33 Q95.2049 769.33 98.5845 772.501 Q101.964 775.672 101.964 781.089 Q101.964 786.668 98.4919 789.769 Q95.0197 792.848 88.7003 792.848 Q86.5243 792.848 84.2558 792.478 Q82.0105 792.107 79.6031 791.367 L79.6031 786.668 Q81.6864 787.802 83.9086 788.357 Q86.1308 788.913 88.6077 788.913 Q92.6123 788.913 94.9502 786.806 Q97.2882 784.7 97.2882 781.089 Q97.2882 777.478 94.9502 775.371 Q92.6123 773.265 88.6077 773.265 Q86.7327 773.265 84.8577 773.681 Q83.0058 774.098 81.0614 774.978 L81.0614 757.617 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M121.177 760.695 Q117.566 760.695 115.737 764.26 Q113.932 767.802 113.932 774.931 Q113.932 782.038 115.737 785.603 Q117.566 789.144 121.177 789.144 Q124.811 789.144 126.617 785.603 Q128.445 782.038 128.445 774.931 Q128.445 767.802 126.617 764.26 Q124.811 760.695 121.177 760.695 M121.177 756.992 Q126.987 756.992 130.043 761.598 Q133.121 766.182 133.121 774.931 Q133.121 783.658 130.043 788.265 Q126.987 792.848 121.177 792.848 Q115.367 792.848 112.288 788.265 Q109.233 783.658 109.233 774.931 Q109.233 766.182 112.288 761.598 Q115.367 756.992 121.177 756.992 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M80.8299 412.871 L103.052 412.871 L103.052 414.862 L90.5058 447.431 L85.6216 447.431 L97.4271 416.807 L80.8299 416.807 L80.8299 412.871 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M112.219 412.871 L130.575 412.871 L130.575 416.807 L116.501 416.807 L116.501 425.279 Q117.52 424.931 118.538 424.769 Q119.557 424.584 120.575 424.584 Q126.362 424.584 129.742 427.756 Q133.121 430.927 133.121 436.343 Q133.121 441.922 129.649 445.024 Q126.177 448.103 119.857 448.103 Q117.682 448.103 115.413 447.732 Q113.168 447.362 110.76 446.621 L110.76 441.922 Q112.844 443.056 115.066 443.612 Q117.288 444.167 119.765 444.167 Q123.77 444.167 126.107 442.061 Q128.445 439.955 128.445 436.343 Q128.445 432.732 126.107 430.626 Q123.77 428.519 119.765 428.519 Q117.89 428.519 116.015 428.936 Q114.163 429.353 112.219 430.232 L112.219 412.871 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M51.6634 98.7508 L59.3023 98.7508 L59.3023 72.3852 L50.9921 74.0518 L50.9921 69.7926 L59.256 68.1259 L63.9319 68.1259 L63.9319 98.7508 L71.5707 98.7508 L71.5707 102.686 L51.6634 102.686 L51.6634 98.7508 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M91.0151 71.2046 Q87.404 71.2046 85.5753 74.7694 Q83.7697 78.3111 83.7697 85.4407 Q83.7697 92.5471 85.5753 96.1119 Q87.404 99.6535 91.0151 99.6535 Q94.6493 99.6535 96.4548 96.1119 Q98.2835 92.5471 98.2835 85.4407 Q98.2835 78.3111 96.4548 74.7694 Q94.6493 71.2046 91.0151 71.2046 M91.0151 67.5009 Q96.8252 67.5009 99.8808 72.1074 Q102.959 76.6907 102.959 85.4407 Q102.959 94.1675 99.8808 98.7739 Q96.8252 103.357 91.0151 103.357 Q85.2049 103.357 82.1262 98.7739 Q79.0707 94.1675 79.0707 85.4407 Q79.0707 76.6907 82.1262 72.1074 Q85.2049 67.5009 91.0151 67.5009 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip010)" d="M121.177 71.2046 Q117.566 71.2046 115.737 74.7694 Q113.932 78.3111 113.932 85.4407 Q113.932 92.5471 115.737 96.1119 Q117.566 99.6535 121.177 99.6535 Q124.811 99.6535 126.617 96.1119 Q128.445 92.5471 128.445 85.4407 Q128.445 78.3111 126.617 74.7694 Q124.811 71.2046 121.177 71.2046 M121.177 67.5009 Q126.987 67.5009 130.043 72.1074 Q133.121 76.6907 133.121 85.4407 Q133.121 94.1675 130.043 98.7739 Q126.987 103.357 121.177 103.357 Q115.367 103.357 112.288 98.7739 Q109.233 94.1675 109.233 85.4407 Q109.233 76.6907 112.288 72.1074 Q115.367 67.5009 121.177 67.5009 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip012)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="793.017,1101.73 704.657,952.826 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="447.043,922.181 704.657,952.826 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="392.332,151.788 934.094,231.349 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1354.55,87.9763 934.094,231.349 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="652.606,606.105 863.882,570.733 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="811.954,495.574 863.882,570.733 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="256.809,1046.01 230.922,889.298 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="321.683,705.487 230.922,889.298 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="798.433,1333.13 1229.18,1265.03 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1363.31,1066.22 1229.18,1265.03 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="863.882,570.733 652.606,606.105 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="584.653,754.401 652.606,606.105 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2263.46,244.214 2191.3,261.454 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1999.41,641.774 2191.3,261.454 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2028.55,707.83 2290.95,740.457 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1975.89,827.543 2290.95,740.457 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="793.017,1101.73 737.113,1131.24 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="600.384,1196.08 737.113,1131.24 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2191.3,261.454 2263.46,244.214 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2241.96,98.1744 2263.46,244.214 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1553.7,823.134 1368.6,856.398 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1331.8,663.5 1368.6,856.398 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1363.31,1066.22 1122.22,906.351 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="794.705,747.973 1122.22,906.351 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="811.954,495.574 1096.49,448.372 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1160.81,404.835 1096.49,448.372 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1544.38,261.436 1820.22,149.11 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2241.96,98.1744 1820.22,149.11 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1229.18,1265.03 798.433,1333.13 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="736.525,1445.72 798.433,1333.13 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="794.705,747.973 649.499,747.765 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="584.653,754.401 649.499,747.765 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="798.433,1333.13 736.525,1445.72 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="600.384,1196.08 736.525,1445.72 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="230.922,889.298 256.809,1046.01 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="376.116,1084.45 256.809,1046.01 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="863.882,570.733 811.954,495.574 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1096.49,448.372 811.954,495.574 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2191.3,261.454 1999.41,641.774 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2028.55,707.83 1999.41,641.774 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="392.332,151.788 374.298,583.308 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="321.683,705.487 374.298,583.308 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1820.22,149.11 1544.38,261.436 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1354.55,87.9763 1544.38,261.436 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2290.95,740.457 2028.55,707.83 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1999.41,641.774 2028.55,707.83 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="704.657,952.826 793.017,1101.73 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="737.113,1131.24 793.017,1101.73 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1553.7,823.134 1774.71,738.432 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1975.89,827.543 1774.71,738.432 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1368.6,856.398 1553.7,823.134 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1774.71,738.432 1553.7,823.134 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="256.809,1046.01 376.116,1084.45 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="447.043,922.181 376.116,1084.45 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2290.95,740.457 1975.89,827.543 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1774.71,738.432 1975.89,827.543 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="934.094,231.349 392.332,151.788 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="374.298,583.308 392.332,151.788 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1229.18,1265.03 1363.31,1066.22 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1122.22,906.351 1363.31,1066.22 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1331.8,663.5 1176.72,552.561 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1160.81,404.835 1176.72,552.561 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="737.113,1131.24 600.384,1196.08 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="736.525,1445.72 600.384,1196.08 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="704.657,952.826 447.043,922.181 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="376.116,1084.45 447.043,922.181 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2263.46,244.214 2241.96,98.1744 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1820.22,149.11 2241.96,98.1744 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="230.922,889.298 321.683,705.487 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="374.298,583.308 321.683,705.487 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1368.6,856.398 1331.8,663.5 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1176.72,552.561 1331.8,663.5 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1096.49,448.372 1160.81,404.835 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1176.72,552.561 1160.81,404.835 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1122.22,906.351 794.705,747.973 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="649.499,747.765 794.705,747.973 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="652.606,606.105 584.653,754.401 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="649.499,747.765 584.653,754.401 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="934.094,231.349 1354.55,87.9763 "/>
+<polyline clip-path="url(#clip012)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1544.38,261.436 1354.55,87.9763 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/8946bc6a.svg b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/8946bc6a.svg
new file mode 100644
index 00000000000..8a2f84641ca
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/8946bc6a.svg
@@ -0,0 +1,122 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip950">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip950)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip951">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip950)" d="M169.121 1486.45 L2352.76 1486.45 L2352.76 47.2441 L169.121 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip952">
+    <rect x="169" y="47" width="2185" height="1440"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="214.497,1486.45 214.497,47.2441 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="733.661,1486.45 733.661,47.2441 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1252.83,1486.45 1252.83,47.2441 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1771.99,1486.45 1771.99,47.2441 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2291.15,1486.45 2291.15,47.2441 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="214.497,1486.45 214.497,1467.55 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="733.661,1486.45 733.661,1467.55 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1252.83,1486.45 1252.83,1467.55 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1771.99,1486.45 1771.99,1467.55 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2291.15,1486.45 2291.15,1467.55 "/>
+<path clip-path="url(#clip950)" d="M214.497 1517.37 Q210.886 1517.37 209.058 1520.93 Q207.252 1524.47 207.252 1531.6 Q207.252 1538.71 209.058 1542.27 Q210.886 1545.82 214.497 1545.82 Q218.132 1545.82 219.937 1542.27 Q221.766 1538.71 221.766 1531.6 Q221.766 1524.47 219.937 1520.93 Q218.132 1517.37 214.497 1517.37 M214.497 1513.66 Q220.308 1513.66 223.363 1518.27 Q226.442 1522.85 226.442 1531.6 Q226.442 1540.33 223.363 1544.94 Q220.308 1549.52 214.497 1549.52 Q208.687 1549.52 205.609 1544.94 Q202.553 1540.33 202.553 1531.6 Q202.553 1522.85 205.609 1518.27 Q208.687 1513.66 214.497 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M712.932 1544.91 L729.252 1544.91 L729.252 1548.85 L707.307 1548.85 L707.307 1544.91 Q709.969 1542.16 714.553 1537.53 Q719.159 1532.88 720.34 1531.53 Q722.585 1529.01 723.465 1527.27 Q724.367 1525.51 724.367 1523.82 Q724.367 1521.07 722.423 1519.33 Q720.502 1517.6 717.4 1517.6 Q715.201 1517.6 712.747 1518.36 Q710.317 1519.13 707.539 1520.68 L707.539 1515.95 Q710.363 1514.82 712.817 1514.24 Q715.27 1513.66 717.307 1513.66 Q722.678 1513.66 725.872 1516.35 Q729.066 1519.03 729.066 1523.52 Q729.066 1525.65 728.256 1527.57 Q727.469 1529.47 725.363 1532.07 Q724.784 1532.74 721.682 1535.95 Q718.58 1539.15 712.932 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M739.113 1514.29 L757.469 1514.29 L757.469 1518.22 L743.395 1518.22 L743.395 1526.7 Q744.414 1526.35 745.432 1526.19 Q746.451 1526 747.469 1526 Q753.256 1526 756.636 1529.17 Q760.015 1532.34 760.015 1537.76 Q760.015 1543.34 756.543 1546.44 Q753.071 1549.52 746.752 1549.52 Q744.576 1549.52 742.307 1549.15 Q740.062 1548.78 737.654 1548.04 L737.654 1543.34 Q739.738 1544.47 741.96 1545.03 Q744.182 1545.58 746.659 1545.58 Q750.664 1545.58 753.001 1543.48 Q755.339 1541.37 755.339 1537.76 Q755.339 1534.15 753.001 1532.04 Q750.664 1529.94 746.659 1529.94 Q744.784 1529.94 742.909 1530.35 Q741.057 1530.77 739.113 1531.65 L739.113 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M1227.52 1514.29 L1245.88 1514.29 L1245.88 1518.22 L1231.81 1518.22 L1231.81 1526.7 Q1232.83 1526.35 1233.84 1526.19 Q1234.86 1526 1235.88 1526 Q1241.67 1526 1245.05 1529.17 Q1248.43 1532.34 1248.43 1537.76 Q1248.43 1543.34 1244.95 1546.44 Q1241.48 1549.52 1235.16 1549.52 Q1232.99 1549.52 1230.72 1549.15 Q1228.47 1548.78 1226.07 1548.04 L1226.07 1543.34 Q1228.15 1544.47 1230.37 1545.03 Q1232.59 1545.58 1235.07 1545.58 Q1239.08 1545.58 1241.41 1543.48 Q1243.75 1541.37 1243.75 1537.76 Q1243.75 1534.15 1241.41 1532.04 Q1239.08 1529.94 1235.07 1529.94 Q1233.2 1529.94 1231.32 1530.35 Q1229.47 1530.77 1227.52 1531.65 L1227.52 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M1267.64 1517.37 Q1264.03 1517.37 1262.2 1520.93 Q1260.39 1524.47 1260.39 1531.6 Q1260.39 1538.71 1262.2 1542.27 Q1264.03 1545.82 1267.64 1545.82 Q1271.27 1545.82 1273.08 1542.27 Q1274.91 1538.71 1274.91 1531.6 Q1274.91 1524.47 1273.08 1520.93 Q1271.27 1517.37 1267.64 1517.37 M1267.64 1513.66 Q1273.45 1513.66 1276.51 1518.27 Q1279.58 1522.85 1279.58 1531.6 Q1279.58 1540.33 1276.51 1544.94 Q1273.45 1549.52 1267.64 1549.52 Q1261.83 1549.52 1258.75 1544.94 Q1255.7 1540.33 1255.7 1531.6 Q1255.7 1522.85 1258.75 1518.27 Q1261.83 1513.66 1267.64 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M1745.84 1514.29 L1768.07 1514.29 L1768.07 1516.28 L1755.52 1548.85 L1750.63 1548.85 L1762.44 1518.22 L1745.84 1518.22 L1745.84 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M1777.23 1514.29 L1795.59 1514.29 L1795.59 1518.22 L1781.51 1518.22 L1781.51 1526.7 Q1782.53 1526.35 1783.55 1526.19 Q1784.57 1526 1785.59 1526 Q1791.38 1526 1794.76 1529.17 Q1798.13 1532.34 1798.13 1537.76 Q1798.13 1543.34 1794.66 1546.44 Q1791.19 1549.52 1784.87 1549.52 Q1782.7 1549.52 1780.43 1549.15 Q1778.18 1548.78 1775.77 1548.04 L1775.77 1543.34 Q1777.86 1544.47 1780.08 1545.03 Q1782.3 1545.58 1784.78 1545.58 Q1788.78 1545.58 1791.12 1543.48 Q1793.46 1541.37 1793.46 1537.76 Q1793.46 1534.15 1791.12 1532.04 Q1788.78 1529.94 1784.78 1529.94 Q1782.9 1529.94 1781.03 1530.35 Q1779.18 1530.77 1777.23 1531.65 L1777.23 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M2250.76 1544.91 L2258.4 1544.91 L2258.4 1518.55 L2250.09 1520.21 L2250.09 1515.95 L2258.35 1514.29 L2263.03 1514.29 L2263.03 1544.91 L2270.67 1544.91 L2270.67 1548.85 L2250.76 1548.85 L2250.76 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M2290.11 1517.37 Q2286.5 1517.37 2284.67 1520.93 Q2282.87 1524.47 2282.87 1531.6 Q2282.87 1538.71 2284.67 1542.27 Q2286.5 1545.82 2290.11 1545.82 Q2293.75 1545.82 2295.55 1542.27 Q2297.38 1538.71 2297.38 1531.6 Q2297.38 1524.47 2295.55 1520.93 Q2293.75 1517.37 2290.11 1517.37 M2290.11 1513.66 Q2295.92 1513.66 2298.98 1518.27 Q2302.06 1522.85 2302.06 1531.6 Q2302.06 1540.33 2298.98 1544.94 Q2295.92 1549.52 2290.11 1549.52 Q2284.3 1549.52 2281.22 1544.94 Q2278.17 1540.33 2278.17 1531.6 Q2278.17 1522.85 2281.22 1518.27 Q2284.3 1513.66 2290.11 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M2320.27 1517.37 Q2316.66 1517.37 2314.83 1520.93 Q2313.03 1524.47 2313.03 1531.6 Q2313.03 1538.71 2314.83 1542.27 Q2316.66 1545.82 2320.27 1545.82 Q2323.91 1545.82 2325.71 1542.27 Q2327.54 1538.71 2327.54 1531.6 Q2327.54 1524.47 2325.71 1520.93 Q2323.91 1517.37 2320.27 1517.37 M2320.27 1513.66 Q2326.08 1513.66 2329.14 1518.27 Q2332.22 1522.85 2332.22 1531.6 Q2332.22 1540.33 2329.14 1544.94 Q2326.08 1549.52 2320.27 1549.52 Q2314.46 1549.52 2311.38 1544.94 Q2308.33 1540.33 2308.33 1531.6 Q2308.33 1522.85 2311.38 1518.27 Q2314.46 1513.66 2320.27 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,1464.39 2352.76,1464.39 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,1119.64 2352.76,1119.64 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,774.897 2352.76,774.897 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,430.151 2352.76,430.151 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="169.121,85.4059 2352.76,85.4059 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1486.45 169.121,47.2441 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1464.39 188.019,1464.39 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,1119.64 188.019,1119.64 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,774.897 188.019,774.897 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,430.151 188.019,430.151 "/>
+<polyline clip-path="url(#clip950)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="169.121,85.4059 188.019,85.4059 "/>
+<path clip-path="url(#clip950)" d="M121.177 1450.19 Q117.566 1450.19 115.737 1453.75 Q113.932 1457.29 113.932 1464.42 Q113.932 1471.53 115.737 1475.09 Q117.566 1478.64 121.177 1478.64 Q124.811 1478.64 126.617 1475.09 Q128.445 1471.53 128.445 1464.42 Q128.445 1457.29 126.617 1453.75 Q124.811 1450.19 121.177 1450.19 M121.177 1446.48 Q126.987 1446.48 130.043 1451.09 Q133.121 1455.67 133.121 1464.42 Q133.121 1473.15 130.043 1477.76 Q126.987 1482.34 121.177 1482.34 Q115.367 1482.34 112.288 1477.76 Q109.233 1473.15 109.233 1464.42 Q109.233 1455.67 112.288 1451.09 Q115.367 1446.48 121.177 1446.48 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M86.0382 1132.99 L102.358 1132.99 L102.358 1136.92 L80.4133 1136.92 L80.4133 1132.99 Q83.0753 1130.23 87.6586 1125.6 Q92.2651 1120.95 93.4456 1119.61 Q95.691 1117.08 96.5706 1115.35 Q97.4734 1113.59 97.4734 1111.9 Q97.4734 1109.14 95.5289 1107.41 Q93.6076 1105.67 90.5058 1105.67 Q88.3067 1105.67 85.8531 1106.44 Q83.4225 1107.2 80.6447 1108.75 L80.6447 1104.03 Q83.4688 1102.89 85.9225 1102.32 Q88.3762 1101.74 90.4132 1101.74 Q95.7836 1101.74 98.978 1104.42 Q102.172 1107.11 102.172 1111.6 Q102.172 1113.73 101.362 1115.65 Q100.575 1117.55 98.4687 1120.14 Q97.89 1120.81 94.7882 1124.03 Q91.6864 1127.22 86.0382 1132.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M112.219 1102.36 L130.575 1102.36 L130.575 1106.3 L116.501 1106.3 L116.501 1114.77 Q117.52 1114.42 118.538 1114.26 Q119.557 1114.08 120.575 1114.08 Q126.362 1114.08 129.742 1117.25 Q133.121 1120.42 133.121 1125.83 Q133.121 1131.41 129.649 1134.51 Q126.177 1137.59 119.857 1137.59 Q117.682 1137.59 115.413 1137.22 Q113.168 1136.85 110.76 1136.11 L110.76 1131.41 Q112.844 1132.55 115.066 1133.1 Q117.288 1133.66 119.765 1133.66 Q123.77 1133.66 126.107 1131.55 Q128.445 1129.45 128.445 1125.83 Q128.445 1122.22 126.107 1120.12 Q123.77 1118.01 119.765 1118.01 Q117.89 1118.01 116.015 1118.43 Q114.163 1118.84 112.219 1119.72 L112.219 1102.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M81.0614 757.617 L99.4178 757.617 L99.4178 761.552 L85.3438 761.552 L85.3438 770.024 Q86.3623 769.677 87.3808 769.515 Q88.3993 769.33 89.4178 769.33 Q95.2049 769.33 98.5845 772.501 Q101.964 775.672 101.964 781.089 Q101.964 786.668 98.4919 789.769 Q95.0197 792.848 88.7003 792.848 Q86.5243 792.848 84.2558 792.478 Q82.0105 792.107 79.6031 791.367 L79.6031 786.668 Q81.6864 787.802 83.9086 788.357 Q86.1308 788.913 88.6077 788.913 Q92.6123 788.913 94.9502 786.806 Q97.2882 784.7 97.2882 781.089 Q97.2882 777.478 94.9502 775.371 Q92.6123 773.265 88.6077 773.265 Q86.7327 773.265 84.8577 773.681 Q83.0058 774.098 81.0614 774.978 L81.0614 757.617 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M121.177 760.695 Q117.566 760.695 115.737 764.26 Q113.932 767.802 113.932 774.931 Q113.932 782.038 115.737 785.603 Q117.566 789.144 121.177 789.144 Q124.811 789.144 126.617 785.603 Q128.445 782.038 128.445 774.931 Q128.445 767.802 126.617 764.26 Q124.811 760.695 121.177 760.695 M121.177 756.992 Q126.987 756.992 130.043 761.598 Q133.121 766.182 133.121 774.931 Q133.121 783.658 130.043 788.265 Q126.987 792.848 121.177 792.848 Q115.367 792.848 112.288 788.265 Q109.233 783.658 109.233 774.931 Q109.233 766.182 112.288 761.598 Q115.367 756.992 121.177 756.992 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M80.8299 412.871 L103.052 412.871 L103.052 414.862 L90.5058 447.431 L85.6216 447.431 L97.4271 416.807 L80.8299 416.807 L80.8299 412.871 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M112.219 412.871 L130.575 412.871 L130.575 416.807 L116.501 416.807 L116.501 425.279 Q117.52 424.931 118.538 424.769 Q119.557 424.584 120.575 424.584 Q126.362 424.584 129.742 427.756 Q133.121 430.927 133.121 436.343 Q133.121 441.922 129.649 445.024 Q126.177 448.103 119.857 448.103 Q117.682 448.103 115.413 447.732 Q113.168 447.362 110.76 446.621 L110.76 441.922 Q112.844 443.056 115.066 443.612 Q117.288 444.167 119.765 444.167 Q123.77 444.167 126.107 442.061 Q128.445 439.955 128.445 436.343 Q128.445 432.732 126.107 430.626 Q123.77 428.519 119.765 428.519 Q117.89 428.519 116.015 428.936 Q114.163 429.353 112.219 430.232 L112.219 412.871 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M51.6634 98.7508 L59.3023 98.7508 L59.3023 72.3852 L50.9921 74.0518 L50.9921 69.7926 L59.256 68.1259 L63.9319 68.1259 L63.9319 98.7508 L71.5707 98.7508 L71.5707 102.686 L51.6634 102.686 L51.6634 98.7508 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M91.0151 71.2046 Q87.404 71.2046 85.5753 74.7694 Q83.7697 78.3111 83.7697 85.4407 Q83.7697 92.5471 85.5753 96.1119 Q87.404 99.6535 91.0151 99.6535 Q94.6493 99.6535 96.4548 96.1119 Q98.2835 92.5471 98.2835 85.4407 Q98.2835 78.3111 96.4548 74.7694 Q94.6493 71.2046 91.0151 71.2046 M91.0151 67.5009 Q96.8252 67.5009 99.8808 72.1074 Q102.959 76.6907 102.959 85.4407 Q102.959 94.1675 99.8808 98.7739 Q96.8252 103.357 91.0151 103.357 Q85.2049 103.357 82.1262 98.7739 Q79.0707 94.1675 79.0707 85.4407 Q79.0707 76.6907 82.1262 72.1074 Q85.2049 67.5009 91.0151 67.5009 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip950)" d="M121.177 71.2046 Q117.566 71.2046 115.737 74.7694 Q113.932 78.3111 113.932 85.4407 Q113.932 92.5471 115.737 96.1119 Q117.566 99.6535 121.177 99.6535 Q124.811 99.6535 126.617 96.1119 Q128.445 92.5471 128.445 85.4407 Q128.445 78.3111 126.617 74.7694 Q124.811 71.2046 121.177 71.2046 M121.177 67.5009 Q126.987 67.5009 130.043 72.1074 Q133.121 76.6907 133.121 85.4407 Q133.121 94.1675 130.043 98.7739 Q126.987 103.357 121.177 103.357 Q115.367 103.357 112.288 98.7739 Q109.233 94.1675 109.233 85.4407 Q109.233 76.6907 112.288 72.1074 Q115.367 67.5009 121.177 67.5009 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip952)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="793.017,1101.73 704.657,952.826 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="447.043,922.181 704.657,952.826 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="392.332,151.788 934.094,231.349 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1354.55,87.9763 934.094,231.349 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="652.606,606.105 863.882,570.733 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="811.954,495.574 863.882,570.733 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="256.809,1046.01 230.922,889.298 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="321.683,705.487 230.922,889.298 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="798.433,1333.13 1229.18,1265.03 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1363.31,1066.22 1229.18,1265.03 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="863.882,570.733 652.606,606.105 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="584.653,754.401 652.606,606.105 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2263.46,244.214 2191.3,261.454 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1999.41,641.774 2191.3,261.454 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2028.55,707.83 2290.95,740.457 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1975.89,827.543 2290.95,740.457 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="793.017,1101.73 737.113,1131.24 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="600.384,1196.08 737.113,1131.24 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2191.3,261.454 2263.46,244.214 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2241.96,98.1744 2263.46,244.214 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1553.7,823.134 1368.6,856.398 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1331.8,663.5 1368.6,856.398 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1363.31,1066.22 1122.22,906.351 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="794.705,747.973 1122.22,906.351 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="811.954,495.574 1096.49,448.372 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1160.81,404.835 1096.49,448.372 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1544.38,261.436 1820.22,149.11 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2241.96,98.1744 1820.22,149.11 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1229.18,1265.03 798.433,1333.13 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="736.525,1445.72 798.433,1333.13 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="794.705,747.973 649.499,747.765 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="584.653,754.401 649.499,747.765 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="798.433,1333.13 736.525,1445.72 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="600.384,1196.08 736.525,1445.72 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="230.922,889.298 256.809,1046.01 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="376.116,1084.45 256.809,1046.01 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="863.882,570.733 811.954,495.574 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1096.49,448.372 811.954,495.574 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2191.3,261.454 1999.41,641.774 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2028.55,707.83 1999.41,641.774 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="392.332,151.788 374.298,583.308 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="321.683,705.487 374.298,583.308 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1820.22,149.11 1544.38,261.436 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1354.55,87.9763 1544.38,261.436 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2290.95,740.457 2028.55,707.83 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1999.41,641.774 2028.55,707.83 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="704.657,952.826 793.017,1101.73 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="737.113,1131.24 793.017,1101.73 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1553.7,823.134 1774.71,738.432 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1975.89,827.543 1774.71,738.432 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1368.6,856.398 1553.7,823.134 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1774.71,738.432 1553.7,823.134 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="256.809,1046.01 376.116,1084.45 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="447.043,922.181 376.116,1084.45 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2290.95,740.457 1975.89,827.543 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1774.71,738.432 1975.89,827.543 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="934.094,231.349 392.332,151.788 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="374.298,583.308 392.332,151.788 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1229.18,1265.03 1363.31,1066.22 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1122.22,906.351 1363.31,1066.22 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1331.8,663.5 1176.72,552.561 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1160.81,404.835 1176.72,552.561 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="737.113,1131.24 600.384,1196.08 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="736.525,1445.72 600.384,1196.08 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="704.657,952.826 447.043,922.181 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="376.116,1084.45 447.043,922.181 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2263.46,244.214 2241.96,98.1744 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1820.22,149.11 2241.96,98.1744 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ac8d18; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="230.922,889.298 321.683,705.487 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a9ad; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="374.298,583.308 321.683,705.487 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#ed5d92; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1368.6,856.398 1331.8,663.5 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#c68125; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1176.72,552.561 1331.8,663.5 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a98d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1096.49,448.372 1160.81,404.835 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#8e971d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1176.72,552.561 1160.81,404.835 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#00a8cb; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1122.22,906.351 794.705,747.973 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#9b7fe8; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="649.499,747.765 794.705,747.973 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#608cf6; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="652.606,606.105 584.653,754.401 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#f05f73; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="649.499,747.765 584.653,754.401 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#dd64b5; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="934.094,231.349 1354.55,87.9763 "/>
+<polyline clip-path="url(#clip952)" style="stroke:#6b9e32; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1544.38,261.436 1354.55,87.9763 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/index.html b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/index.html
new file mode 100644
index 00000000000..78e2f6daeab
--- /dev/null
+++ b/previews/PR3525/tutorials/algorithms/tsp_lazy_constraints/index.html
@@ -0,0 +1,94 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Traveling Salesperson Problem · JuMP</title><meta name="title" content="Traveling Salesperson Problem · JuMP"/><meta property="og:title" content="Traveling Salesperson Problem · JuMP"/><meta property="twitter:title" content="Traveling Salesperson Problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox" checked/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../cutting_stock_column_generation/">Column generation</a></li><li class="is-active"><a class="tocitem" href>Traveling Salesperson Problem</a><ul class="internal"><li><a class="tocitem" href="#tsp_model"><span>Mathematical Formulation</span></a></li><li><a class="tocitem" href="#tsp_implementation"><span>Implementation</span></a></li></ul></li><li><a class="tocitem" href="../parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Algorithms</a></li><li class="is-active"><a href>Traveling Salesperson Problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Traveling Salesperson Problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/algorithms/tsp_lazy_constraints.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="tsp_lazy"><a class="docs-heading-anchor" href="#tsp_lazy">Traveling Salesperson Problem</a><a id="tsp_lazy-1"></a><a class="docs-heading-anchor-permalink" href="#tsp_lazy" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../tsp_lazy_constraints.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Daniel Schermer.</strong></p><p>This tutorial describes how to implement the <a href="https://en.wikipedia.org/wiki/Travelling_salesman_problem">Traveling Salesperson Problem</a> in JuMP using solver-independent lazy constraints that dynamically separate subtours. To be more precise, we use lazy constraints to cut off infeasible subtours only when necessary and not before needed.</p><p>It uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import GLPK
+import Random
+import Plots</code></pre><h2 id="tsp_model"><a class="docs-heading-anchor" href="#tsp_model">Mathematical Formulation</a><a id="tsp_model-1"></a><a class="docs-heading-anchor-permalink" href="#tsp_model" title="Permalink"></a></h2><p>Assume that we are given a complete graph <span>$\mathcal{G}(V,E)$</span> where <span>$V$</span> is the set of vertices (or cities) and <span>$E$</span> is the set of edges (or roads). For each pair of vertices <span>$i, j \in V, i \neq j$</span> the edge <span>$(i,j) \in E$</span> is associated with a weight (or distance) <span>$d_{ij} \in \mathbb{R}^+$</span>.</p><p>For this tutorial, we assume the problem to be symmetric, that is, <span>$d_{ij} = d_{ji} \, \forall i,j \in V$</span>.</p><p>In the Traveling Salesperson Problem, we are tasked with finding a tour with minimal length that visits every vertex exactly once and then returns to the point of origin, that is, a <em>Hamiltonian cycle</em> with minimal weight.</p><p>To model the problem, we introduce a binary variable, <span>$x_{ij} \in \{0,1\} \; \forall i, j \in V$</span>, that indicates if edge <span>$(i,j)$</span> is part of the tour or not. Using these variables, the Traveling Salesperson Problem can be modeled as the following integer linear program.</p><h3 id="tsp_objective"><a class="docs-heading-anchor" href="#tsp_objective">Objective Function</a><a id="tsp_objective-1"></a><a class="docs-heading-anchor-permalink" href="#tsp_objective" title="Permalink"></a></h3><p>The objective is to minimize the length of the tour (due to the assumed symmetry, the second sum only contains <span>$i&lt;j$</span>):</p><p class="math-container">\[\text{min } \sum_{i \in V}  \sum_{j \in V, i &lt; j} d_{ij} x_{ij}.\]</p><p>Note that it is also possible to use the following objective function instead:</p><p class="math-container">\[\text{min } \sum_{i \in V}  \sum_{j \in V} \dfrac{d_{ij} x_{ij}}{2}.\]</p><h3 id="tsp_constraints"><a class="docs-heading-anchor" href="#tsp_constraints">Constraints</a><a id="tsp_constraints-1"></a><a class="docs-heading-anchor-permalink" href="#tsp_constraints" title="Permalink"></a></h3><p>There are four classes of constraints in our formulation.</p><p>First, due to the presumed symmetry, the following constraints must hold:</p><p class="math-container">\[x_{ij} = x_{ji} \quad \forall i,j \in V.\]</p><p>Second, for each vertex <span>$i$</span>, exactly two edges must be selected that connect it to other vertices <span>$j$</span> in the graph <span>$G$</span>:</p><p class="math-container">\[\sum_{j \in V} x_{ij} = 2 \quad \forall i \in V.\]</p><p>Third, we do not permit loops to occur:</p><p class="math-container">\[x_{ii} = 0 \quad \forall i \in V.\]</p><p>The fourth constraint is more complicated. A major difficulty of the Traveling Salesperson Problem arises from the fact that we need to prevent <em>subtours</em>, that is, several distinct Hamiltonian cycles existing on subgraphs of <span>$G$</span>.</p><p>Note that the previous constraints <em>do not</em> guarantee that the solution will be free of subtours. To this end, by <span>$S$</span> we label a subset of vertices. Then, for each proper subset <span>$S \subset V$</span>, the following constraints guarantee that no subtour may occur:</p><p class="math-container">\[\sum_{i \in S} \sum_{j \in S, i &lt; j} x_{ij} \leq \vert S \vert - 1 \quad \forall S \subset V.\]</p><p>Problematically, we require exponentially many of these constraints as <span>$\vert V \vert$</span> increases. Therefore, we will add these constraints only when necessary.</p><h2 id="tsp_implementation"><a class="docs-heading-anchor" href="#tsp_implementation">Implementation</a><a id="tsp_implementation-1"></a><a class="docs-heading-anchor-permalink" href="#tsp_implementation" title="Permalink"></a></h2><p>There are two ways we can eliminate subtours in JuMP, both of which will be shown in what follows:</p><ul><li>iteratively solving a new model that incorporates previously identified subtours,</li><li>or adding violated subtours as <em>lazy constraints</em>.</li></ul><h3 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h3><p>The vertices are assumed to be randomly distributed in the Euclidean space; thus, the weight (distance) of each edge is defined as follows.</p><pre><code class="language-julia hljs">function generate_distance_matrix(n; random_seed = 1)
+    rng = Random.MersenneTwister(random_seed)
+    X = 100 * rand(rng, n)
+    Y = 100 * rand(rng, n)
+    d = [sqrt((X[i] - X[j])^2 + (Y[i] - Y[j])^2) for i in 1:n, j in 1:n]
+    return X, Y, d
+end
+
+n = 40
+X, Y, d = generate_distance_matrix(n)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">([23.603334566204694, 34.651701419196044, 31.27069683360675, 0.790928339056074, 48.86128300795012, 21.096820215853597, 95.1916339835734, 99.99046588986135, 25.166218303197184, 98.66663668987997  …  46.33502459235987, 18.582130997265377, 11.198087695816717, 97.6311881619359, 5.161462067432709, 53.80295812064833, 45.56920516275036, 27.93951106725605, 17.824610354168602, 54.89828719625274], [37.097066286146884, 89.41659192657593, 64.80537482231894, 41.70393538841062, 14.456554241360564, 62.24031828206811, 87.23344353741976, 52.49746566167794, 24.159060827129643, 88.48369255734127  …  66.12321555087209, 19.45678064479248, 39.3193497656424, 99.07406554003964, 55.03342139580574, 58.07816346526631, 76.83586278313636, 51.952465724186084, 51.486297110544356, 99.81360570779374], [0.0 53.473350122820904 … 15.506244459460921 70.09092934998034; 53.473350122820904 0.0 … 41.49527995497558 22.760099542720535; … ; 15.506244459460921 41.49527995497558 … 0.0 60.9096566304971; 70.09092934998034 22.760099542720535 … 60.9096566304971 0.0])</code></pre><p>For the JuMP model, we first initialize the model object. Then, we create the binary decision variables and add the objective function and constraints. By defining the <code>x</code> matrix as <code>Symmetric</code>, we do not need to add explicit constraints that <code>x[i, j] == x[j, i]</code>.</p><pre><code class="language-julia hljs">function build_tsp_model(d, n)
+    model = Model(GLPK.Optimizer)
+    @variable(model, x[1:n, 1:n], Bin, Symmetric)
+    @objective(model, Min, sum(d .* x) / 2)
+    @constraint(model, [i in 1:n], sum(x[i, :]) == 2)
+    @constraint(model, [i in 1:n], x[i, i] == 0)
+    return model
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">build_tsp_model (generic function with 1 method)</code></pre><p>To search for violated constraints, based on the edges that are currently in the solution (that is, those that have value <span>$x_{ij} = 1$</span>), we identify the shortest cycle through the function <code>subtour()</code>. Whenever a subtour has been identified, a constraint corresponding to the form above can be added to the model.</p><pre><code class="language-julia hljs">function subtour(edges::Vector{Tuple{Int,Int}}, n)
+    shortest_subtour, unvisited = collect(1:n), Set(collect(1:n))
+    while !isempty(unvisited)
+        this_cycle, neighbors = Int[], unvisited
+        while !isempty(neighbors)
+            current = pop!(neighbors)
+            push!(this_cycle, current)
+            if length(this_cycle) &gt; 1
+                pop!(unvisited, current)
+            end
+            neighbors =
+                [j for (i, j) in edges if i == current &amp;&amp; j in unvisited]
+        end
+        if length(this_cycle) &lt; length(shortest_subtour)
+            shortest_subtour = this_cycle
+        end
+    end
+    return shortest_subtour
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">subtour (generic function with 1 method)</code></pre><p>Let us declare a helper function <code>selected_edges()</code> that will be repeatedly used in what follows.</p><pre><code class="language-julia hljs">function selected_edges(x::Matrix{Float64}, n)
+    return Tuple{Int,Int}[(i, j) for i in 1:n, j in 1:n if x[i, j] &gt; 0.5]
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">selected_edges (generic function with 1 method)</code></pre><p>Other helper functions for computing subtours:</p><pre><code class="language-julia hljs">subtour(x::Matrix{Float64}) = subtour(selected_edges(x, size(x, 1)), size(x, 1))
+subtour(x::AbstractMatrix{VariableRef}) = subtour(value.(x))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">subtour (generic function with 3 methods)</code></pre><h3 id="Iterative-method"><a class="docs-heading-anchor" href="#Iterative-method">Iterative method</a><a id="Iterative-method-1"></a><a class="docs-heading-anchor-permalink" href="#Iterative-method" title="Permalink"></a></h3><p>An iterative way of eliminating subtours is the following.</p><p>However, it is reasonable to assume that this is not the most efficient way: whenever a new subtour elimination constraint is added to the model, the optimization has to start from the very beginning.</p><p>That way, the solver will repeatedly discard useful information encountered during previous solves (for example, all cuts, the incumbent solution, or lower bounds).</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Note that, in principle, it would also be feasible to add all subtours (instead of just the shortest one) to the model. However, preventing just the shortest cycle is often sufficient for breaking other subtours and will keep the model size smaller.</p></div></div><pre><code class="language-julia hljs">iterative_model = build_tsp_model(d, n)
+optimize!(iterative_model)
+time_iterated = solve_time(iterative_model)
+cycle = subtour(iterative_model[:x])
+while 1 &lt; length(cycle) &lt; n
+    println(&quot;Found cycle of length $(length(cycle))&quot;)
+    S = [(i, j) for (i, j) in Iterators.product(cycle, cycle) if i &lt; j]
+    @constraint(
+        iterative_model,
+        sum(iterative_model[:x][i, j] for (i, j) in S) &lt;= length(cycle) - 1,
+    )
+    optimize!(iterative_model)
+    global time_iterated += solve_time(iterative_model)
+    global cycle = subtour(iterative_model[:x])
+end
+
+objective_value(iterative_model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">525.7039004442727</code></pre><pre><code class="language-julia hljs">time_iterated</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.03757762908935547</code></pre><p>As a quick sanity check, we visualize the optimal tour to verify that no subtour is present:</p><pre><code class="language-julia hljs">function plot_tour(X, Y, x)
+    plot = Plots.plot()
+    for (i, j) in selected_edges(x, size(x, 1))
+        Plots.plot!([X[i], X[j]], [Y[i], Y[j]]; legend = false)
+    end
+    return plot
+end
+
+plot_tour(X, Y, value.(iterative_model[:x]))</code></pre><img src="8946bc6a.svg" alt="Example block output"/><h3 id="Lazy-constraint-method"><a class="docs-heading-anchor" href="#Lazy-constraint-method">Lazy constraint method</a><a id="Lazy-constraint-method-1"></a><a class="docs-heading-anchor-permalink" href="#Lazy-constraint-method" title="Permalink"></a></h3><p>A more sophisticated approach makes use of <em>lazy constraints</em>. To be more precise, we do this through the <code>subtour_elimination_callback()</code> below, which is only run whenever we encounter a new integer-feasible solution.</p><pre><code class="language-julia hljs">lazy_model = build_tsp_model(d, n)
+function subtour_elimination_callback(cb_data)
+    status = callback_node_status(cb_data, lazy_model)
+    if status != MOI.CALLBACK_NODE_STATUS_INTEGER
+        return  # Only run at integer solutions
+    end
+    cycle = subtour(callback_value.(cb_data, lazy_model[:x]))
+    if !(1 &lt; length(cycle) &lt; n)
+        return  # Only add a constraint if there is a cycle
+    end
+    println(&quot;Found cycle of length $(length(cycle))&quot;)
+    S = [(i, j) for (i, j) in Iterators.product(cycle, cycle) if i &lt; j]
+    con = @build_constraint(
+        sum(lazy_model[:x][i, j] for (i, j) in S) &lt;= length(cycle) - 1,
+    )
+    MOI.submit(lazy_model, MOI.LazyConstraint(cb_data), con)
+    return
+end
+set_attribute(
+    lazy_model,
+    MOI.LazyConstraintCallback(),
+    subtour_elimination_callback,
+)
+optimize!(lazy_model)
+objective_value(lazy_model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">525.7039004442727</code></pre><p>This finds the same optimal tour:</p><pre><code class="language-julia hljs">plot_tour(X, Y, value.(lazy_model[:x]))</code></pre><img src="43baf431.svg" alt="Example block output"/><p>Surprisingly, for this particular model with GLPK, the solution time is worse than the iterative method:</p><pre><code class="language-julia hljs">time_lazy = solve_time(lazy_model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.17717790603637695</code></pre><p>In most other cases and solvers, however, the lazy time should be faster than the iterative method.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../cutting_stock_column_generation/">« Column generation</a><a class="docs-footer-nextpage" href="../parallelism/">Parallelism »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/applications/optimal_power_flow.jl b/previews/PR3525/tutorials/applications/optimal_power_flow.jl
new file mode 100644
index 00000000000..07ec2834bbe
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/optimal_power_flow.jl
@@ -0,0 +1,470 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Optimal power flow
+
+# *This tutorial was originally contributed by James Foster (@jd-foster).*
+
+# This tutorial formulates and solves an alternating current optimal power flow
+# (AC-OPF) problem, a much-studied nonlinear problem from the field of
+# electrical engineering.
+
+# Once we've formulated and solved the nonlinear problem, we will turn our focus
+# to obtaining a good estimate of the objective value at the global optimum
+# through the use of semidefinite programming.
+
+# One main purpose of this tutorial is to highlight JuMP's ability to directly
+# formulate problems involving complex-valued decision variables and complex
+# matrix cones such as the  [`HermitianPSDCone`](@ref) object.
+
+# For another example of modeling with complex decision variables, see the
+# [Quantum state discrimination](@ref) tutorial, and see the
+# [Complex number support](@ref) section of the manual for more details.
+
+# This tutorial takes a matrix-oriented approach focused on network nodes
+# that simplifies the construction of semidefinite programs.
+# Another approach is to formulate the problem focusing on network lines
+# where it is easier to work with flow constraints. A general approach is
+# provided by the Julia/JuMP package [PowerModels.jl](https://lanl-ansi.github.io/PowerModels.jl/stable/),
+# an open-source framework to a broad range of power flow model formulations
+# along with utilities for working with detailed network data.
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import Clarabel
+import DataFrames
+import Ipopt
+import LinearAlgebra
+import SparseArrays
+import Test #src
+
+# ## Initial formulation
+
+# Optimal power flow problems for electrical transmission typically pose the
+# following question: what is the most cost-effective operation of electricity
+# generators while meeting constraints on the safe limits of network components?
+
+# We'll use the 9-_node_ network test case `case9mod` to explore this problem.
+# The graph of the network, shown here, has three nodes (or _buses_) each for
+# the different purposes of generation ``G`` (nodes 1, 2, and 3), trans-shipment
+# (nodes 4, 6, and 8), and demand ``D`` (nodes 5, 7, and 9).
+
+# ![Nine Nodes](../../assets/case9mod.png)
+
+# This example is a modified version of the [MATPOWER](https://matpower.org/)
+# test case `case9` ([archive](https://github.com/MATPOWER/matpower/tree/master/data))
+# created by Bukhsh et al. (2013) for their test case archive of
+# optimal power flow problems with local optima. This test case is also
+# extensively evaluated in Krasko and Rebennack (2017).
+
+# Here *bus* and network *node* are taken as analogous terms, as are *branch*
+# and transmission *line*.
+
+# For future reference, let's name the number of nodes in the network:
+
+N = 9;
+
+# The network data can be summarised using a small number of arrays. Using the
+# `sparsevec` function from the `SparseArrays` standard library package, we can
+# give the indices and values of the non-zero data points:
+
+## Real generation: lower (`lb`) and upper (`ub`) bounds
+P_Gen_lb = SparseArrays.sparsevec([1, 2, 3], [10, 10, 10], N)
+P_Gen_ub = SparseArrays.sparsevec([1, 2, 3], [250, 300, 270], N)
+## Reactive generation: lower (`lb`) and upper (`ub`) bounds
+Q_Gen_lb = SparseArrays.sparsevec([1, 2, 3], [-5, -5, -5], N)
+Q_Gen_ub = SparseArrays.sparsevec([1, 2, 3], [300, 300, 300], N)
+## Power demand levels (real, reactive, and complex form)
+P_Demand = SparseArrays.sparsevec([5, 7, 9], [54, 60, 75], N)
+Q_Demand = SparseArrays.sparsevec([5, 7, 9], [18, 21, 30], N)
+S_Demand = P_Demand + im * Q_Demand
+
+# The key decision variables are the real power injections ``P^G`` and reactive
+# power injections ``Q^G``over the allowed range of the generators. All other
+# buses must restrict their generation variables to 0. On the other hand, these
+# non-generator nodes have a fixed  real and reactive power demand, denoted
+# ``P^D`` and ``Q^D`` respectively (these are fixed at 0 in the case of
+# trans-shipment and generator nodes).
+
+# The cost of operating each generator is modeled as a quadratic function of
+# its real power output; in our specific test case, the objective function to
+# minimize is:
+# ```math
+# \begin{align}
+#     \min      &    0.11 \;\; (P^G_1)^2 +   5 P^G_1 + 150  \\
+#               & +  0.085 \; (P^G_2)^2 + 1.2 P^G_2 + 600  \\
+#               & + 0.1225 \;  (P^G_3)^2 +     P^G_3 + 335 \\
+# \end{align}
+# ```
+# Let's create an initial JuMP model with some of this data:
+
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, P_Gen_lb[i] <= P_G[i in 1:N] <= P_Gen_ub[i])
+@objective(
+    model,
+    Min,
+    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +
+    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +
+    (0.1225 * P_G[3]^2 + P_G[3] + 335),
+);
+
+# Even before solving an optimization problem, we can estimate a lower bound on
+# the best objective value by substituting the lower bound on each generator's
+# real power range (all 10, as it turns out in this case):
+
+basic_lower_bound = value(lower_bound, objective_function(model));
+Test.@test isapprox(basic_lower_bound, 1188.75; atol = 1e-2)  #src
+println("Objective value (basic lower bound) : $basic_lower_bound")
+
+# to see that we can do no better than an objective cost of 1188.75.
+
+# (Direct substitution works because a
+# [quadratic function](https://en.wikipedia.org/wiki/Quadratic_function#Graph_of_the_univariate_function)
+# of a single variable ``x`` with positive coefficients is strictly increasing
+# for all ``x \geq 0``.)
+
+# In fact, we can get a quick but even better estimate from the direct
+# observation that the real power generated must meet or exceed the real power
+# demand:
+
+@constraint(model, sum(P_G) >= sum(P_Demand))
+optimize!(model)
+better_lower_bound = round(objective_value(model); digits = 2)
+println("Objective value (better lower bound): $better_lower_bound")
+
+# However, there are additional power flow constraints that must be satisfied.
+
+# Power must flow from one or more generation nodes through the transmission
+# lines and end up at a demand node. The state variables of our steady-state
+# alternating current (AC) electrical network are *complex-valued* voltage
+# variables ``V_1, \ldots, V_N``. Voltages capture both a magnitude and phase of
+# the node's electrical state in relation to the rest of the system. An AC power
+# system also extends the notion of resistance in wires found in a direct
+# current (DC) circuit to a  complex quantity, known as the *impedance*, of each
+# transmission line. The reciprocal of impedance is known as *admittance*.
+# Together, these complex quantities are used to express a complex version of
+# *Ohm's law*: current flow through a line is proportional to the difference in
+# voltages on each end of the line, multiplied by the admittance.
+
+# ## Network data
+
+# Let's assemble the data we need for writing the complex power flow constraints.
+# The data for the problem consists of a list of the real and imaginary parts of
+# the line impedance. We obtain the following data table from the `branch data`
+# section of the `case9mod` MATPOWER format file:
+
+branch_data = DataFrames.DataFrame([
+    (1, 4, 0.0, 0.0576, 0.0),
+    (4, 5, 0.017, 0.092, 0.158),
+    (6, 5, 0.039, 0.17, 0.358),
+    (3, 6, 0.0, 0.0586, 0.0),
+    (6, 7, 0.0119, 0.1008, 0.209),
+    (8, 7, 0.0085, 0.072, 0.149),
+    (2, 8, 0.0, 0.0625, 0.0),
+    (8, 9, 0.032, 0.161, 0.306),
+    (4, 9, 0.01, 0.085, 0.176),
+]);
+DataFrames.rename!(branch_data, [:F_BUS, :T_BUS, :BR_R, :BR_X, :BR_Bc])
+
+# The first two columns describe the network, supplying the *from* and *to*
+# connection points of the lines. The last three columns give the branch
+# resistance, branch reactance and *line-charging susceptance*.
+
+# We will also need to reference the `base_MVA` number (used for re-scaling):
+
+base_MVA = 100;
+
+# and the number of lines:
+
+M = size(branch_data, 1)
+
+# From the first two columns of the branch data table, we can create a sparse
+# [incidence matrix](https://en.wikipedia.org/wiki/Incidence_matrix) that
+# simplifies handling of the network layout:
+
+A =
+    SparseArrays.sparse(branch_data.F_BUS, 1:M, 1, N, M) +
+    SparseArrays.sparse(branch_data.T_BUS, 1:M, -1, N, M)
+
+# We form the network impedance vector from the next two columns
+
+z = (branch_data.BR_R .+ im * branch_data.BR_X) / base_MVA;
+
+# and calculate it's corresponding *bus admittance* matrix as
+
+Y_0 = A * SparseArrays.spdiagm(1 ./ z) * A';
+
+# while the last column gives the branch line-charging susceptance
+
+y_sh = 1 / 2 * (im * branch_data.BR_Bc) * base_MVA;
+
+# and leads to the shunt admittance matrix
+
+Y_sh = SparseArrays.spdiagm(
+    LinearAlgebra.diag(A * SparseArrays.spdiagm(y_sh) * A'),
+);
+
+# (The construction of the shunt admittance matrix `Y_sh` looks
+# somewhat more complicated than `Y_0` because we only want to add the
+# diagonal elements in the calculation; the line-charging is used only in the
+# nodal voltage terms and not the line voltage terms.)
+
+# The full *bus admittance* matrix ``Y`` is then defined as
+
+Y = Y_0 + Y_sh;
+
+# ## JuMP model
+
+# Now we're ready to write the complex power flow constraints we need to more
+# accurately model the electricity system.
+
+# We'll introduce a number of constraints that model both the physics and
+# operational requirements.
+
+# Let's start by initializing a new model:
+
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+
+# Then we'll create the nodal power generation variables:
+
+@variable(
+    model,
+    S_G[i in 1:N] in ComplexPlane(),
+    lower_bound = P_Gen_lb[i] + Q_Gen_lb[i] * im,
+    upper_bound = P_Gen_ub[i] + Q_Gen_ub[i] * im,
+)
+
+# We need complex nodal voltages (the system state variables):
+
+@variable(model, V[1:N] in ComplexPlane(), start = 1.0 + 0.0im)
+
+# and operational constraints for maintaining voltage magnitude levels:
+@constraint(model, [i in 1:N], 0.9^2 <= real(V[i])^2 + imag(V[i])^2 <= 1.1^2)
+
+# We also need to fix an origin or _reference angle_ from which all other
+# complex voltage angles (arguments) are determined.
+# Here we will use node 1 as the nominated _reference bus_.
+# Fixing the imaginary component of a reference bus to zero sets its complex
+# voltage angle to 0, while constraining the real part to be non-negative
+# disallows equivalent solutions that are just a reflection by 180 degrees:
+
+@constraint(model, imag(V[1]) == 0);
+@constraint(model, real(V[1]) >= 0);
+
+# The power flow equations express a conservation of energy (power) principle,
+# where power generated less the power consumed must balance the power exchanged
+# with the network:
+
+@constraint(model, S_G - S_Demand .== V .* conj(Y * V))
+
+# As above, the objective function is a quadratic cost of real power:
+
+P_G = real(S_G)
+@objective(
+    model,
+    Min,
+    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +
+    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +
+    (0.1225 * P_G[3]^2 + P_G[3] + 335),
+);
+
+# We're finally ready to solve our nonlinear AC-OPF problem:
+
+optimize!(model)
+Test.@test isapprox(objective_value(model), 3087.84; atol = 1e-2)  #src
+solution_summary(model)
+
+#-
+
+objval_solution = round(objective_value(model); digits = 2)
+println("Objective value (feasible solution) : $(objval_solution)")
+
+# The solution's power generation (in rectangular form)
+# and complex voltage values (in polar form using degrees) are:
+
+DataFrames.DataFrame(;
+    Bus = 1:N,
+    ComplexPowerGen = round.(value.(S_G); digits = 2),
+    VoltageMagnitude = round.(abs.(value.(V)); digits = 2),
+    VoltageAngle_Deg = round.(rad2deg.(angle.(value.(V))); digits = 2),
+)
+
+# ## Relaxations and better objective bounds
+
+# The Ipopt solver uses an interior-point algorithm. It has local optimality
+# guarantees, but is unable to certify whether the solution is globally optimal.
+# The solution we found is indeed globally optimal. The work to verify this has
+# been done in Bukhsh et al. (2013) and Krasko and Rebennack (2017), and
+# different solvers (such as Gurobi, SCIP and GLOMIQO) are also able to verify
+# this.
+
+# The techniques of *convex relaxations* can also be used to improve on our
+# current best lower bound:
+
+better_lower_bound
+
+# To this end, observe that the nonlinear constraints in the AC-OPF formulation
+# are quadratic equalities for power flow along with quadratic voltage
+# inequalities.
+
+# Let's linearize these constraints by first making the substitution
+# ``W = V V^*``, where:
+# ```math
+# W = V V^* \quad \iff  \quad W_{ii} = | V_i |^2, \quad W_{ik} = V_i \; \overline{V_k}, \quad \forall i, \, k \in \{ 1, \ldots, N \}
+# ```
+# and where ``V^*`` is the [conjugate transpose](https://en.wikipedia.org/wiki/Conjugate_transpose) of ``V``.
+
+# On the face of it, this turns a quadratic voltage bound constraint like:
+# ```math
+# v_L \leq |V_i |^2  \leq v_U, \quad i  \in \{ 1, \ldots, N \}
+# ```
+# for some real ``v_L`` and ``v_U`` into a simple two-sided bound:
+# ```math
+# v_L \leq W_{ii}  \leq v_U,
+# ```
+# while each quadratic expression for the nodal power term:
+# ```math
+#     S^{Node}_i = V_i \overline{(YV)_i}
+# ```
+# becomes the linear combination:
+# ```math
+#     S^{Node}_i = (E_{ii} Y^T) \bullet W.
+# ```
+# Here ``A \bullet B = \operatorname{tr}(A^* B)`` is the
+# [Frobenius inner product](https://en.wikipedia.org/wiki/Frobenius_inner_product)
+# of two complex matrices, while ``E_{kn}`` denotes the _matrix unit_ with a
+# single nonzero entry of 1 in row ``k`` and column ``n``.
+
+E(k, n) = SparseArrays.sparse([k], [n], 1, N, N);
+
+# Of course, we've shifted the nonlinearity into the equality constraint
+# ``W = V V^*``: it is this constraint we will now relax using a
+# semidefinite programming approach.
+
+# We will make use of complex voltages and relax
+# ``W = V V^*`` to;
+# ```math
+# W \succeq V V^*,
+# ```
+# where the relation ``\succeq`` is the ordering in the Hermitian positive
+# semidefinite cone.
+
+# The above constraint is equivalent to:
+# ```math
+# \begin{bmatrix} 1 & V^* \\ V & W \\ \end{bmatrix} \succeq 0
+# ```
+# by the theory of the [Schur complement](https://en.wikipedia.org/wiki/Schur_complement).
+# This matrix inequality implies a number of second-order cone constraints by
+# taking certain ``2 \times 2`` minors of the matrix for each
+# ``i  \in \{ 1, \ldots, N \}``:
+# ```math
+# \begin{bmatrix} 1 & V_i^* \\ V_i & W_{ii} \\ \end{bmatrix} \succeq 0,
+# ```
+# which is equivalent to the real second-order cone inequality:
+# ```math
+# \operatorname{real}(W_{ii}) \geq \operatorname{real}(V_i)^2 + \operatorname{imag}(V_i)^2.
+# ```
+# We include these implied constraints as well for demonstration purposes.
+
+# Putting it all together we get the following semidefinite relaxation of the
+# AC-OPF problem:
+
+model = Model(Clarabel.Optimizer)
+set_attribute(model, "tol_gap_rel", 1e-3)
+set_attribute(model, "tol_feas", 1e-3)
+set_attribute(model, "tol_ktratio", 1e-3)
+@variable(
+    model,
+    S_G[i in 1:N] in ComplexPlane(),
+    lower_bound = P_Gen_lb[i] + Q_Gen_lb[i] * im,
+    upper_bound = P_Gen_ub[i] + Q_Gen_ub[i] * im,
+)
+@variable(model, W[1:N, 1:N] in HermitianPSDCone())
+@variable(model, V[1:N] in ComplexPlane(), start = 1.0 + 0.0im)
+@constraint(model, [i in 1:N], 0.9^2 <= real(W[i, i]) <= 1.1^2)
+@constraint(model, real(V[1]) >= 0)
+@constraint(model, imag(V[1]) == 0)
+@constraint(model, 0.9 <= real(V[1]) <= 1.1)
+@constraint(model, LinearAlgebra.Hermitian([1 V'; V W]) in HermitianPSDCone())
+## 2 x 2 minor inequalities:
+@constraint(
+    model,
+    [i in 1:N],
+    [0.5, real(W[i, i]), real(V[i]), imag(V[i])] in RotatedSecondOrderCone()
+)
+@constraint(
+    model,
+    [i in 1:N],
+    S_G[i] - S_Demand[i] == LinearAlgebra.tr((conj(Y) * E(i, i)) * W),
+)
+P_G = real(S_G)
+@objective(
+    model,
+    Min,
+    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +
+    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +
+    (0.1225 * P_G[3]^2 + P_G[3] + 335),
+)
+optimize!(model)
+
+#-
+
+sdp_relaxation_lower_bound = round(objective_value(model); digits = 2)
+Test.@test termination_status(model) in (OPTIMAL, ALMOST_OPTIMAL)           #src
+Test.@test primal_status(model) in (FEASIBLE_POINT, NEARLY_FEASIBLE_POINT)  #src
+Test.@test isapprox(sdp_relaxation_lower_bound, 2753.04; rtol = 1e-3)     #src
+println(
+    "Objective value (W & V relax. lower bound): $sdp_relaxation_lower_bound",
+)
+
+# We can more easily see solution values by rounding out noisy data:
+
+W_1 = SparseArrays.sparse(round.(value.(W); digits = 2))
+
+# and recover an approximation to the voltage variables as:
+
+DataFrames.DataFrame(;
+    Bus = 1:N,
+    Magnitude = round.(abs.(value.(V)); digits = 2),
+    AngleDeg = round.(rad2deg.(angle.(value.(V))); digits = 2),
+)
+
+# For further information on exploiting sparsity see Jabr (2012).
+
+# This relaxation has the advantage that we can work directly with complex
+# voltages to extend the formulation, strengthen the relaxation and gain
+# additional approximate information about the voltage variables.
+
+# ## References and further resources
+
+# **Bukhsh**, W. A., Grothey, A., McKinnon, K. I., & Trodden, P. A.
+# [_Local solutions of the optimal power flow problem._](https://doi.org/10.1109/TPWRS.2013.2274577)
+# IEEE Transactions on Power Systems, 28(4), 4780-4788 (2013).
+
+# **Jabr**, R. A.
+# [_Exploiting sparsity in SDP relaxations of the OPF problem._](https://doi.org/10.1109/TPWRS.2011.2170772)
+# IEEE Transactions on Power Systems, 27(2), 1138-1139 (2011).
+
+# **Krasko**, V., & S. **Rebennack**.
+# [_Chapter 15: Global Optimization: Optimal Power Flow Problem._](https://doi.org/10.1137/1.9781611974683.ch15)
+# In Advances and Trends in Optimization with Engineering Applications, 187—205. MOS-SIAM Series on Optimization.
+# Society for Industrial and Applied Mathematics, 2017.
+
+# [**Test case `case9mod`**](https://www.maths.ed.ac.uk/optenergy/LocalOpt/9busnetwork.html):
+# from the
+# [Test Case Archive of Optimal Power Flow (OPF) Problems with Local Optima](https://www.maths.ed.ac.uk/optenergy/LocalOpt/).
+
+# R. D. Zimmerman, C. E. Murillo-Sanchez, & R. J. Thomas,
+# [_MATPOWER: Steady-State Operations, Planning and Analysis Tools for Power Systems
+# Research and Education,_](https://doi.org/10.1109/TPWRS.2010.2051168)
+# IEEE Transactions on Power Systems, 26(1), 12-19 (2011).
+
+# **MATPOWER data format**:
+# [MATPOWER manual](https://matpower.org/docs/MATPOWER-manual.pdf); see especially Appendix B "Data File Format" and Table B-3.
diff --git a/previews/PR3525/tutorials/applications/optimal_power_flow/index.html b/previews/PR3525/tutorials/applications/optimal_power_flow/index.html
new file mode 100644
index 00000000000..0cd090255ca
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/optimal_power_flow/index.html
@@ -0,0 +1,232 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Optimal power flow · JuMP</title><meta name="title" content="Optimal power flow · JuMP"/><meta property="og:title" content="Optimal power flow · JuMP"/><meta property="twitter:title" content="Optimal power flow · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox" checked/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../power_systems/">Power Systems</a></li><li class="is-active"><a class="tocitem" href>Optimal power flow</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Initial-formulation"><span>Initial formulation</span></a></li><li><a class="tocitem" href="#Network-data"><span>Network data</span></a></li><li><a class="tocitem" href="#JuMP-model"><span>JuMP model</span></a></li><li><a class="tocitem" href="#Relaxations-and-better-objective-bounds"><span>Relaxations and better objective bounds</span></a></li><li><a class="tocitem" href="#References-and-further-resources"><span>References and further resources</span></a></li></ul></li><li><a class="tocitem" href="../web_app/">Serving web apps</a></li><li><a class="tocitem" href="../two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Applications</a></li><li class="is-active"><a href>Optimal power flow</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Optimal power flow</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/applications/optimal_power_flow.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Optimal-power-flow"><a class="docs-heading-anchor" href="#Optimal-power-flow">Optimal power flow</a><a id="Optimal-power-flow-1"></a><a class="docs-heading-anchor-permalink" href="#Optimal-power-flow" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../optimal_power_flow.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><em>This tutorial was originally contributed by James Foster (@jd-foster).</em></p><p>This tutorial formulates and solves an alternating current optimal power flow (AC-OPF) problem, a much-studied nonlinear problem from the field of electrical engineering.</p><p>Once we&#39;ve formulated and solved the nonlinear problem, we will turn our focus to obtaining a good estimate of the objective value at the global optimum through the use of semidefinite programming.</p><p>One main purpose of this tutorial is to highlight JuMP&#39;s ability to directly formulate problems involving complex-valued decision variables and complex matrix cones such as the  <a href="../../../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a> object.</p><p>For another example of modeling with complex decision variables, see the <a href="../../conic/quantum_discrimination/#Quantum-state-discrimination">Quantum state discrimination</a> tutorial, and see the <a href="../../../manual/complex/#Complex-number-support">Complex number support</a> section of the manual for more details.</p><p>This tutorial takes a matrix-oriented approach focused on network nodes that simplifies the construction of semidefinite programs. Another approach is to formulate the problem focusing on network lines where it is easier to work with flow constraints. A general approach is provided by the Julia/JuMP package <a href="https://lanl-ansi.github.io/PowerModels.jl/stable/">PowerModels.jl</a>, an open-source framework to a broad range of power flow model formulations along with utilities for working with detailed network data.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import Clarabel
+import DataFrames
+import Ipopt
+import LinearAlgebra
+import SparseArrays</code></pre><h2 id="Initial-formulation"><a class="docs-heading-anchor" href="#Initial-formulation">Initial formulation</a><a id="Initial-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Initial-formulation" title="Permalink"></a></h2><p>Optimal power flow problems for electrical transmission typically pose the following question: what is the most cost-effective operation of electricity generators while meeting constraints on the safe limits of network components?</p><p>We&#39;ll use the 9-<em>node</em> network test case <code>case9mod</code> to explore this problem. The graph of the network, shown here, has three nodes (or <em>buses</em>) each for the different purposes of generation <span>$G$</span> (nodes 1, 2, and 3), trans-shipment (nodes 4, 6, and 8), and demand <span>$D$</span> (nodes 5, 7, and 9).</p><p><img src="../../../assets/case9mod.png" alt="Nine Nodes"/></p><p>This example is a modified version of the <a href="https://matpower.org/">MATPOWER</a> test case <code>case9</code> (<a href="https://github.com/MATPOWER/matpower/tree/master/data">archive</a>) created by Bukhsh et al. (2013) for their test case archive of optimal power flow problems with local optima. This test case is also extensively evaluated in Krasko and Rebennack (2017).</p><p>Here <em>bus</em> and network <em>node</em> are taken as analogous terms, as are <em>branch</em> and transmission <em>line</em>.</p><p>For future reference, let&#39;s name the number of nodes in the network:</p><pre><code class="language-julia hljs">N = 9;</code></pre><p>The network data can be summarised using a small number of arrays. Using the <code>sparsevec</code> function from the <code>SparseArrays</code> standard library package, we can give the indices and values of the non-zero data points:</p><pre><code class="language-julia hljs"># Real generation: lower (`lb`) and upper (`ub`) bounds
+P_Gen_lb = SparseArrays.sparsevec([1, 2, 3], [10, 10, 10], N)
+P_Gen_ub = SparseArrays.sparsevec([1, 2, 3], [250, 300, 270], N)
+# Reactive generation: lower (`lb`) and upper (`ub`) bounds
+Q_Gen_lb = SparseArrays.sparsevec([1, 2, 3], [-5, -5, -5], N)
+Q_Gen_ub = SparseArrays.sparsevec([1, 2, 3], [300, 300, 300], N)
+# Power demand levels (real, reactive, and complex form)
+P_Demand = SparseArrays.sparsevec([5, 7, 9], [54, 60, 75], N)
+Q_Demand = SparseArrays.sparsevec([5, 7, 9], [18, 21, 30], N)
+S_Demand = P_Demand + im * Q_Demand</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9-element SparseArrays.SparseVector{Complex{Int64}, Int64} with 3 stored entries:
+  [5]  =  54+18im
+  [7]  =  60+21im
+  [9]  =  75+30im</code></pre><p>The key decision variables are the real power injections <span>$P^G$</span> and reactive power injections <span>$Q^G$</span>over the allowed range of the generators. All other buses must restrict their generation variables to 0. On the other hand, these non-generator nodes have a fixed  real and reactive power demand, denoted <span>$P^D$</span> and <span>$Q^D$</span> respectively (these are fixed at 0 in the case of trans-shipment and generator nodes).</p><p>The cost of operating each generator is modeled as a quadratic function of its real power output; in our specific test case, the objective function to minimize is:</p><p class="math-container">\[\begin{align}
+    \min      &amp;    0.11 \;\; (P^G_1)^2 +   5 P^G_1 + 150  \\
+              &amp; +  0.085 \; (P^G_2)^2 + 1.2 P^G_2 + 600  \\
+              &amp; + 0.1225 \;  (P^G_3)^2 +     P^G_3 + 335 \\
+\end{align}\]</p><p>Let&#39;s create an initial JuMP model with some of this data:</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, P_Gen_lb[i] &lt;= P_G[i in 1:N] &lt;= P_Gen_ub[i])
+@objective(
+    model,
+    Min,
+    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +
+    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +
+    (0.1225 * P_G[3]^2 + P_G[3] + 335),
+);</code></pre><p>Even before solving an optimization problem, we can estimate a lower bound on the best objective value by substituting the lower bound on each generator&#39;s real power range (all 10, as it turns out in this case):</p><pre><code class="language-julia hljs">basic_lower_bound = value(lower_bound, objective_function(model));
+println(&quot;Objective value (basic lower bound) : $basic_lower_bound&quot;)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Objective value (basic lower bound) : 1188.75</code></pre><p>to see that we can do no better than an objective cost of 1188.75.</p><p>(Direct substitution works because a <a href="https://en.wikipedia.org/wiki/Quadratic_function#Graph_of_the_univariate_function">quadratic function</a> of a single variable <span>$x$</span> with positive coefficients is strictly increasing for all <span>$x \geq 0$</span>.)</p><p>In fact, we can get a quick but even better estimate from the direct observation that the real power generated must meet or exceed the real power demand:</p><pre><code class="language-julia hljs">@constraint(model, sum(P_G) &gt;= sum(P_Demand))
+optimize!(model)
+better_lower_bound = round(objective_value(model); digits = 2)
+println(&quot;Objective value (better lower bound): $better_lower_bound&quot;)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Objective value (better lower bound): 2733.55</code></pre><p>However, there are additional power flow constraints that must be satisfied.</p><p>Power must flow from one or more generation nodes through the transmission lines and end up at a demand node. The state variables of our steady-state alternating current (AC) electrical network are <em>complex-valued</em> voltage variables <span>$V_1, \ldots, V_N$</span>. Voltages capture both a magnitude and phase of the node&#39;s electrical state in relation to the rest of the system. An AC power system also extends the notion of resistance in wires found in a direct current (DC) circuit to a  complex quantity, known as the <em>impedance</em>, of each transmission line. The reciprocal of impedance is known as <em>admittance</em>. Together, these complex quantities are used to express a complex version of <em>Ohm&#39;s law</em>: current flow through a line is proportional to the difference in voltages on each end of the line, multiplied by the admittance.</p><h2 id="Network-data"><a class="docs-heading-anchor" href="#Network-data">Network data</a><a id="Network-data-1"></a><a class="docs-heading-anchor-permalink" href="#Network-data" title="Permalink"></a></h2><p>Let&#39;s assemble the data we need for writing the complex power flow constraints. The data for the problem consists of a list of the real and imaginary parts of the line impedance. We obtain the following data table from the <code>branch data</code> section of the <code>case9mod</code> MATPOWER format file:</p><pre><code class="language-julia hljs">branch_data = DataFrames.DataFrame([
+    (1, 4, 0.0, 0.0576, 0.0),
+    (4, 5, 0.017, 0.092, 0.158),
+    (6, 5, 0.039, 0.17, 0.358),
+    (3, 6, 0.0, 0.0586, 0.0),
+    (6, 7, 0.0119, 0.1008, 0.209),
+    (8, 7, 0.0085, 0.072, 0.149),
+    (2, 8, 0.0, 0.0625, 0.0),
+    (8, 9, 0.032, 0.161, 0.306),
+    (4, 9, 0.01, 0.085, 0.176),
+]);
+DataFrames.rename!(branch_data, [:F_BUS, :T_BUS, :BR_R, :BR_X, :BR_Bc])</code></pre><div><div style = "float: left;"><span>9×5 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">F_BUS</th><th style = "text-align: left;">T_BUS</th><th style = "text-align: left;">BR_R</th><th style = "text-align: left;">BR_X</th><th style = "text-align: left;">BR_Bc</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">1</td><td style = "text-align: right;">4</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">0.0576</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">4</td><td style = "text-align: right;">5</td><td style = "text-align: right;">0.017</td><td style = "text-align: right;">0.092</td><td style = "text-align: right;">0.158</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">6</td><td style = "text-align: right;">5</td><td style = "text-align: right;">0.039</td><td style = "text-align: right;">0.17</td><td style = "text-align: right;">0.358</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">3</td><td style = "text-align: right;">6</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">0.0586</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">6</td><td style = "text-align: right;">7</td><td style = "text-align: right;">0.0119</td><td style = "text-align: right;">0.1008</td><td style = "text-align: right;">0.209</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">8</td><td style = "text-align: right;">7</td><td style = "text-align: right;">0.0085</td><td style = "text-align: right;">0.072</td><td style = "text-align: right;">0.149</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">2</td><td style = "text-align: right;">8</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">0.0625</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">8</td><td style = "text-align: right;">9</td><td style = "text-align: right;">0.032</td><td style = "text-align: right;">0.161</td><td style = "text-align: right;">0.306</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">4</td><td style = "text-align: right;">9</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">0.085</td><td style = "text-align: right;">0.176</td></tr></tbody></table></div><p>The first two columns describe the network, supplying the <em>from</em> and <em>to</em> connection points of the lines. The last three columns give the branch resistance, branch reactance and <em>line-charging susceptance</em>.</p><p>We will also need to reference the <code>base_MVA</code> number (used for re-scaling):</p><pre><code class="language-julia hljs">base_MVA = 100;</code></pre><p>and the number of lines:</p><pre><code class="language-julia hljs">M = size(branch_data, 1)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9</code></pre><p>From the first two columns of the branch data table, we can create a sparse <a href="https://en.wikipedia.org/wiki/Incidence_matrix">incidence matrix</a> that simplifies handling of the network layout:</p><pre><code class="language-julia hljs">A =
+    SparseArrays.sparse(branch_data.F_BUS, 1:M, 1, N, M) +
+    SparseArrays.sparse(branch_data.T_BUS, 1:M, -1, N, M)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9×9 SparseArrays.SparseMatrixCSC{Int64, Int64} with 18 stored entries:
+  1   ⋅   ⋅   ⋅   ⋅   ⋅   ⋅   ⋅   ⋅
+  ⋅   ⋅   ⋅   ⋅   ⋅   ⋅   1   ⋅   ⋅
+  ⋅   ⋅   ⋅   1   ⋅   ⋅   ⋅   ⋅   ⋅
+ -1   1   ⋅   ⋅   ⋅   ⋅   ⋅   ⋅   1
+  ⋅  -1  -1   ⋅   ⋅   ⋅   ⋅   ⋅   ⋅
+  ⋅   ⋅   1  -1   1   ⋅   ⋅   ⋅   ⋅
+  ⋅   ⋅   ⋅   ⋅  -1  -1   ⋅   ⋅   ⋅
+  ⋅   ⋅   ⋅   ⋅   ⋅   1  -1   1   ⋅
+  ⋅   ⋅   ⋅   ⋅   ⋅   ⋅   ⋅  -1  -1</code></pre><p>We form the network impedance vector from the next two columns</p><pre><code class="language-julia hljs">z = (branch_data.BR_R .+ im * branch_data.BR_X) / base_MVA;</code></pre><p>and calculate it&#39;s corresponding <em>bus admittance</em> matrix as</p><pre><code class="language-julia hljs">Y_0 = A * SparseArrays.spdiagm(1 ./ z) * A&#39;;</code></pre><p>while the last column gives the branch line-charging susceptance</p><pre><code class="language-julia hljs">y_sh = 1 / 2 * (im * branch_data.BR_Bc) * base_MVA;</code></pre><p>and leads to the shunt admittance matrix</p><pre><code class="language-julia hljs">Y_sh = SparseArrays.spdiagm(
+    LinearAlgebra.diag(A * SparseArrays.spdiagm(y_sh) * A&#39;),
+);</code></pre><p>(The construction of the shunt admittance matrix <code>Y_sh</code> looks somewhat more complicated than <code>Y_0</code> because we only want to add the diagonal elements in the calculation; the line-charging is used only in the nodal voltage terms and not the line voltage terms.)</p><p>The full <em>bus admittance</em> matrix <span>$Y$</span> is then defined as</p><pre><code class="language-julia hljs">Y = Y_0 + Y_sh;</code></pre><h2 id="JuMP-model"><a class="docs-heading-anchor" href="#JuMP-model">JuMP model</a><a id="JuMP-model-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-model" title="Permalink"></a></h2><p>Now we&#39;re ready to write the complex power flow constraints we need to more accurately model the electricity system.</p><p>We&#39;ll introduce a number of constraints that model both the physics and operational requirements.</p><p>Let&#39;s start by initializing a new model:</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+set_silent(model)</code></pre><p>Then we&#39;ll create the nodal power generation variables:</p><pre><code class="language-julia hljs">@variable(
+    model,
+    S_G[i in 1:N] in ComplexPlane(),
+    lower_bound = P_Gen_lb[i] + Q_Gen_lb[i] * im,
+    upper_bound = P_Gen_ub[i] + Q_Gen_ub[i] * im,
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9-element Vector{GenericAffExpr{ComplexF64, VariableRef}}:
+ real(S_G[1]) + imag(S_G[1]) im
+ real(S_G[2]) + imag(S_G[2]) im
+ real(S_G[3]) + imag(S_G[3]) im
+ real(S_G[4]) + imag(S_G[4]) im
+ real(S_G[5]) + imag(S_G[5]) im
+ real(S_G[6]) + imag(S_G[6]) im
+ real(S_G[7]) + imag(S_G[7]) im
+ real(S_G[8]) + imag(S_G[8]) im
+ real(S_G[9]) + imag(S_G[9]) im</code></pre><p>We need complex nodal voltages (the system state variables):</p><pre><code class="language-julia hljs">@variable(model, V[1:N] in ComplexPlane(), start = 1.0 + 0.0im)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9-element Vector{GenericAffExpr{ComplexF64, VariableRef}}:
+ real(V[1]) + imag(V[1]) im
+ real(V[2]) + imag(V[2]) im
+ real(V[3]) + imag(V[3]) im
+ real(V[4]) + imag(V[4]) im
+ real(V[5]) + imag(V[5]) im
+ real(V[6]) + imag(V[6]) im
+ real(V[7]) + imag(V[7]) im
+ real(V[8]) + imag(V[8]) im
+ real(V[9]) + imag(V[9]) im</code></pre><p>and operational constraints for maintaining voltage magnitude levels:</p><pre><code class="language-julia hljs">@constraint(model, [i in 1:N], 0.9^2 &lt;= real(V[i])^2 + imag(V[i])^2 &lt;= 1.1^2)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarQuadraticFunction{Float64}, MathOptInterface.Interval{Float64}}, ScalarShape}}:
+ real(V[1])² + imag(V[1])² ∈ [0.81, 1.2100000000000002]
+ real(V[2])² + imag(V[2])² ∈ [0.81, 1.2100000000000002]
+ real(V[3])² + imag(V[3])² ∈ [0.81, 1.2100000000000002]
+ real(V[4])² + imag(V[4])² ∈ [0.81, 1.2100000000000002]
+ real(V[5])² + imag(V[5])² ∈ [0.81, 1.2100000000000002]
+ real(V[6])² + imag(V[6])² ∈ [0.81, 1.2100000000000002]
+ real(V[7])² + imag(V[7])² ∈ [0.81, 1.2100000000000002]
+ real(V[8])² + imag(V[8])² ∈ [0.81, 1.2100000000000002]
+ real(V[9])² + imag(V[9])² ∈ [0.81, 1.2100000000000002]</code></pre><p>We also need to fix an origin or <em>reference angle</em> from which all other complex voltage angles (arguments) are determined. Here we will use node 1 as the nominated <em>reference bus</em>. Fixing the imaginary component of a reference bus to zero sets its complex voltage angle to 0, while constraining the real part to be non-negative disallows equivalent solutions that are just a reflection by 180 degrees:</p><pre><code class="language-julia hljs">@constraint(model, imag(V[1]) == 0);
+@constraint(model, real(V[1]) &gt;= 0);</code></pre><p>The power flow equations express a conservation of energy (power) principle, where power generated less the power consumed must balance the power exchanged with the network:</p><pre><code class="language-julia hljs">@constraint(model, S_G - S_Demand .== V .* conj(Y * V))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarQuadraticFunction{ComplexF64}, MathOptInterface.EqualTo{ComplexF64}}, ScalarShape}}:
+ -1736.111111111111im real(V[1])² + 1736.111111111111im real(V[1])*real(V[4]) + 1736.111111111111 real(V[1])*imag(V[4]) - 1736.111111111111im imag(V[1])² - 1736.111111111111 imag(V[1])*real(V[4]) + 1736.111111111111im imag(V[1])*imag(V[4]) + real(S_G[1]) + imag(S_G[1]) im = 0
+ -1600im real(V[2])² + 1600im real(V[2])*real(V[8]) + 1600 real(V[2])*imag(V[8]) - 1600im imag(V[2])² - 1600 imag(V[2])*real(V[8]) + 1600im imag(V[2])*imag(V[8]) + real(S_G[2]) + imag(S_G[2]) im = 0
+ -1706.484641638225im real(V[3])² + 1706.484641638225im real(V[3])*real(V[6]) + 1706.484641638225 real(V[3])*imag(V[6]) - 1706.484641638225im imag(V[3])² - 1706.484641638225 imag(V[3])*real(V[6]) + 1706.484641638225im imag(V[3])*imag(V[6]) + real(S_G[3]) + imag(S_G[3]) im = 0
+ 1736.111111111111im real(V[4])*real(V[1]) - 1736.111111111111 imag(V[4])*real(V[1]) + 1736.111111111111 real(V[4])*imag(V[1]) + 1736.111111111111im imag(V[4])*imag(V[1]) + (-330.7378962025307 - 3930.8888726118976im) real(V[4])² + (194.21912487147264 + 1051.0682051867932im) real(V[4])*real(V[5]) + (1051.0682051867932 - 194.21912487147264im) real(V[4])*imag(V[5]) + (136.51877133105802 + 1160.409556313993im) real(V[4])*real(V[9]) + (1160.409556313993 - 136.51877133105802im) real(V[4])*imag(V[9]) + (-330.7378962025307 - 3930.8888726118976im) imag(V[4])² + (-1051.0682051867932 + 194.21912487147264im) imag(V[4])*real(V[5]) + (194.21912487147264 + 1051.0682051867932im) imag(V[4])*imag(V[5]) + (-1160.409556313993 + 136.51877133105802im) imag(V[4])*real(V[9]) + (136.51877133105802 + 1160.409556313993im) imag(V[4])*imag(V[9]) + real(S_G[4]) + imag(S_G[4]) im = 0
+ (194.21912487147264 + 1051.0682051867932im) real(V[5])*real(V[4]) + (-1051.0682051867932 + 194.21912487147264im) imag(V[5])*real(V[4]) + (1051.0682051867932 - 194.21912487147264im) real(V[5])*imag(V[4]) + (194.21912487147264 + 1051.0682051867932im) imag(V[5])*imag(V[4]) + (-322.4200387138841 - 1584.0927014229458im) real(V[5])² + (128.20091384241147 + 558.8244962361526im) real(V[5])*real(V[6]) + (558.8244962361526 - 128.20091384241147im) real(V[5])*imag(V[6]) + (-322.4200387138841 - 1584.0927014229458im) imag(V[5])² + (-558.8244962361526 + 128.20091384241147im) imag(V[5])*real(V[6]) + (128.20091384241147 + 558.8244962361526im) imag(V[5])*imag(V[6]) + real(S_G[5]) + imag(S_G[5]) im = (54 + 18im)
+ 1706.484641638225im real(V[6])*real(V[3]) - 1706.484641638225 imag(V[6])*real(V[3]) + 1706.484641638225 real(V[6])*imag(V[3]) + 1706.484641638225im imag(V[6])*imag(V[3]) + (128.20091384241147 + 558.8244962361526im) real(V[6])*real(V[5]) + (-558.8244962361526 + 128.20091384241147im) imag(V[6])*real(V[5]) + (558.8244962361526 - 128.20091384241147im) real(V[6])*imag(V[5]) + (128.20091384241147 + 558.8244962361526im) imag(V[6])*imag(V[5]) + (-243.70966193142118 - 3215.386180510695im) real(V[6])² + (115.5087480890097 + 978.4270426363173im) real(V[6])*real(V[7]) + (978.4270426363173 - 115.5087480890097im) real(V[6])*imag(V[7]) + (-243.70966193142118 - 3215.386180510695im) imag(V[6])² + (-978.4270426363173 + 115.5087480890097im) imag(V[6])*real(V[7]) + (115.5087480890097 + 978.4270426363173im) imag(V[6])*imag(V[7]) + real(S_G[6]) + imag(S_G[6]) im = 0
+ (115.5087480890097 + 978.4270426363173im) real(V[7])*real(V[6]) + (-978.4270426363173 + 115.5087480890097im) imag(V[7])*real(V[6]) + (978.4270426363173 - 115.5087480890097im) real(V[7])*imag(V[6]) + (115.5087480890097 + 978.4270426363173im) imag(V[7])*imag(V[6]) + (-277.22099541362326 - 2330.3249023271615im) real(V[7])² + (161.71224732461357 + 1369.7978596908442im) real(V[7])*real(V[8]) + (1369.7978596908442 - 161.71224732461357im) real(V[7])*imag(V[8]) + (-277.22099541362326 - 2330.3249023271615im) imag(V[7])² + (-1369.7978596908442 + 161.71224732461357im) imag(V[7])*real(V[8]) + (161.71224732461357 + 1369.7978596908442im) imag(V[7])*imag(V[8]) + real(S_G[7]) + imag(S_G[7]) im = (60 + 21im)
+ 1600im real(V[8])*real(V[2]) - 1600 imag(V[8])*real(V[2]) + 1600 real(V[8])*imag(V[2]) + 1600im imag(V[8])*imag(V[2]) + (161.71224732461357 + 1369.7978596908442im) real(V[8])*real(V[7]) + (-1369.7978596908442 + 161.71224732461357im) imag(V[8])*real(V[7]) + (1369.7978596908442 - 161.71224732461357im) real(V[8])*imag(V[7]) + (161.71224732461357 + 1369.7978596908442im) imag(V[8])*imag(V[7]) + (-280.4726852537284 - 3544.5613130217034im) real(V[8])² + (118.76043792911486 + 597.5134533308592im) real(V[8])*real(V[9]) + (597.5134533308592 - 118.76043792911486im) real(V[8])*imag(V[9]) + (-280.4726852537284 - 3544.5613130217034im) imag(V[8])² + (-597.5134533308592 + 118.76043792911486im) imag(V[8])*real(V[9]) + (118.76043792911486 + 597.5134533308592im) imag(V[8])*imag(V[9]) + real(S_G[8]) + imag(S_G[8]) im = 0
+ (136.51877133105802 + 1160.409556313993im) real(V[9])*real(V[4]) + (-1160.409556313993 + 136.51877133105802im) imag(V[9])*real(V[4]) + (1160.409556313993 - 136.51877133105802im) real(V[9])*imag(V[4]) + (136.51877133105802 + 1160.409556313993im) imag(V[9])*imag(V[4]) + (118.76043792911486 + 597.5134533308592im) real(V[9])*real(V[8]) + (-597.5134533308592 + 118.76043792911486im) imag(V[9])*real(V[8]) + (597.5134533308592 - 118.76043792911486im) real(V[9])*imag(V[8]) + (118.76043792911486 + 597.5134533308592im) imag(V[9])*imag(V[8]) + (-255.27920926017288 - 1733.8230096448524im) real(V[9])² + (-255.27920926017288 - 1733.8230096448524im) imag(V[9])² + real(S_G[9]) + imag(S_G[9]) im = (75 + 30im)</code></pre><p>As above, the objective function is a quadratic cost of real power:</p><pre><code class="language-julia hljs">P_G = real(S_G)
+@objective(
+    model,
+    Min,
+    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +
+    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +
+    (0.1225 * P_G[3]^2 + P_G[3] + 335),
+);</code></pre><p>We&#39;re finally ready to solve our nonlinear AC-OPF problem:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Ipopt
+
+* Status
+  Result count       : 1
+  Termination status : LOCALLY_SOLVED
+  Message from the solver:
+  &quot;Solve_Succeeded&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 3.08784e+03
+
+* Work counters
+  Solve time (sec)   : 9.59206e-03
+</code></pre><pre><code class="language-julia hljs">objval_solution = round(objective_value(model); digits = 2)
+println(&quot;Objective value (feasible solution) : $(objval_solution)&quot;)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Objective value (feasible solution) : 3087.84</code></pre><p>The solution&#39;s power generation (in rectangular form) and complex voltage values (in polar form using degrees) are:</p><pre><code class="language-julia hljs">DataFrames.DataFrame(;
+    Bus = 1:N,
+    ComplexPowerGen = round.(value.(S_G); digits = 2),
+    VoltageMagnitude = round.(abs.(value.(V)); digits = 2),
+    VoltageAngle_Deg = round.(rad2deg.(angle.(value.(V))); digits = 2),
+)</code></pre><div><div style = "float: left;"><span>9×4 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Bus</th><th style = "text-align: left;">ComplexPowerGen</th><th style = "text-align: left;">VoltageMagnitude</th><th style = "text-align: left;">VoltageAngle_Deg</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "ComplexF64" style = "text-align: left;">Complex…</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">1</td><td style = "text-align: right;">10.0-5.0im</td><td style = "text-align: right;">0.91</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">2</td><td style = "text-align: right;">125.37-5.0im</td><td style = "text-align: right;">0.92</td><td style = "text-align: right;">12.37</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">3</td><td style = "text-align: right;">57.03-5.0im</td><td style = "text-align: right;">0.94</td><td style = "text-align: right;">7.01</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">4</td><td style = "text-align: right;">0.0+0.0im</td><td style = "text-align: right;">0.91</td><td style = "text-align: right;">-0.4</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">5</td><td style = "text-align: right;">0.0+0.0im</td><td style = "text-align: right;">0.92</td><td style = "text-align: right;">-0.73</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">6</td><td style = "text-align: right;">0.0+0.0im</td><td style = "text-align: right;">0.94</td><td style = "text-align: right;">4.84</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">7</td><td style = "text-align: right;">0.0+0.0im</td><td style = "text-align: right;">0.93</td><td style = "text-align: right;">4.52</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">8</td><td style = "text-align: right;">0.0+0.0im</td><td style = "text-align: right;">0.93</td><td style = "text-align: right;">7.12</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">9</td><td style = "text-align: right;">0.0+0.0im</td><td style = "text-align: right;">0.9</td><td style = "text-align: right;">-0.63</td></tr></tbody></table></div><h2 id="Relaxations-and-better-objective-bounds"><a class="docs-heading-anchor" href="#Relaxations-and-better-objective-bounds">Relaxations and better objective bounds</a><a id="Relaxations-and-better-objective-bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Relaxations-and-better-objective-bounds" title="Permalink"></a></h2><p>The Ipopt solver uses an interior-point algorithm. It has local optimality guarantees, but is unable to certify whether the solution is globally optimal. The solution we found is indeed globally optimal. The work to verify this has been done in Bukhsh et al. (2013) and Krasko and Rebennack (2017), and different solvers (such as Gurobi, SCIP and GLOMIQO) are also able to verify this.</p><p>The techniques of <em>convex relaxations</em> can also be used to improve on our current best lower bound:</p><pre><code class="language-julia hljs">better_lower_bound</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2733.55</code></pre><p>To this end, observe that the nonlinear constraints in the AC-OPF formulation are quadratic equalities for power flow along with quadratic voltage inequalities.</p><p>Let&#39;s linearize these constraints by first making the substitution <span>$W = V V^*$</span>, where:</p><p class="math-container">\[W = V V^* \quad \iff  \quad W_{ii} = | V_i |^2, \quad W_{ik} = V_i \; \overline{V_k}, \quad \forall i, \, k \in \{ 1, \ldots, N \}\]</p><p>and where <span>$V^*$</span> is the <a href="https://en.wikipedia.org/wiki/Conjugate_transpose">conjugate transpose</a> of <span>$V$</span>.</p><p>On the face of it, this turns a quadratic voltage bound constraint like:</p><p class="math-container">\[v_L \leq |V_i |^2  \leq v_U, \quad i  \in \{ 1, \ldots, N \}\]</p><p>for some real <span>$v_L$</span> and <span>$v_U$</span> into a simple two-sided bound:</p><p class="math-container">\[v_L \leq W_{ii}  \leq v_U,\]</p><p>while each quadratic expression for the nodal power term:</p><p class="math-container">\[    S^{Node}_i = V_i \overline{(YV)_i}\]</p><p>becomes the linear combination:</p><p class="math-container">\[    S^{Node}_i = (E_{ii} Y^T) \bullet W.\]</p><p>Here <span>$A \bullet B = \operatorname{tr}(A^* B)$</span> is the <a href="https://en.wikipedia.org/wiki/Frobenius_inner_product">Frobenius inner product</a> of two complex matrices, while <span>$E_{kn}$</span> denotes the <em>matrix unit</em> with a single nonzero entry of 1 in row <span>$k$</span> and column <span>$n$</span>.</p><pre><code class="language-julia hljs">E(k, n) = SparseArrays.sparse([k], [n], 1, N, N);</code></pre><p>Of course, we&#39;ve shifted the nonlinearity into the equality constraint <span>$W = V V^*$</span>: it is this constraint we will now relax using a semidefinite programming approach.</p><p>We will make use of complex voltages and relax <span>$W = V V^*$</span> to;</p><p class="math-container">\[W \succeq V V^*,\]</p><p>where the relation <span>$\succeq$</span> is the ordering in the Hermitian positive semidefinite cone.</p><p>The above constraint is equivalent to:</p><p class="math-container">\[\begin{bmatrix} 1 &amp; V^* \\ V &amp; W \\ \end{bmatrix} \succeq 0\]</p><p>by the theory of the <a href="https://en.wikipedia.org/wiki/Schur_complement">Schur complement</a>. This matrix inequality implies a number of second-order cone constraints by taking certain <span>$2 \times 2$</span> minors of the matrix for each <span>$i  \in \{ 1, \ldots, N \}$</span>:</p><p class="math-container">\[\begin{bmatrix} 1 &amp; V_i^* \\ V_i &amp; W_{ii} \\ \end{bmatrix} \succeq 0,\]</p><p>which is equivalent to the real second-order cone inequality:</p><p class="math-container">\[\operatorname{real}(W_{ii}) \geq \operatorname{real}(V_i)^2 + \operatorname{imag}(V_i)^2.\]</p><p>We include these implied constraints as well for demonstration purposes.</p><p>Putting it all together we get the following semidefinite relaxation of the AC-OPF problem:</p><pre><code class="language-julia hljs">model = Model(Clarabel.Optimizer)
+set_attribute(model, &quot;tol_gap_rel&quot;, 1e-3)
+set_attribute(model, &quot;tol_feas&quot;, 1e-3)
+set_attribute(model, &quot;tol_ktratio&quot;, 1e-3)
+@variable(
+    model,
+    S_G[i in 1:N] in ComplexPlane(),
+    lower_bound = P_Gen_lb[i] + Q_Gen_lb[i] * im,
+    upper_bound = P_Gen_ub[i] + Q_Gen_ub[i] * im,
+)
+@variable(model, W[1:N, 1:N] in HermitianPSDCone())
+@variable(model, V[1:N] in ComplexPlane(), start = 1.0 + 0.0im)
+@constraint(model, [i in 1:N], 0.9^2 &lt;= real(W[i, i]) &lt;= 1.1^2)
+@constraint(model, real(V[1]) &gt;= 0)
+@constraint(model, imag(V[1]) == 0)
+@constraint(model, 0.9 &lt;= real(V[1]) &lt;= 1.1)
+@constraint(model, LinearAlgebra.Hermitian([1 V&#39;; V W]) in HermitianPSDCone())
+# 2 x 2 minor inequalities:
+@constraint(
+    model,
+    [i in 1:N],
+    [0.5, real(W[i, i]), real(V[i]), imag(V[i])] in RotatedSecondOrderCone()
+)
+@constraint(
+    model,
+    [i in 1:N],
+    S_G[i] - S_Demand[i] == LinearAlgebra.tr((conj(Y) * E(i, i)) * W),
+)
+P_G = real(S_G)
+@objective(
+    model,
+    Min,
+    (0.11 * P_G[1]^2 + 5 * P_G[1] + 150) +
+    (0.085 * P_G[2]^2 + 1.2 * P_G[2] + 600) +
+    (0.1225 * P_G[3]^2 + P_G[3] + 335),
+)
+optimize!(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">-------------------------------------------------------------
+           Clarabel.jl v0.5.1  -  Clever Acronym
+                   (c) Paul Goulart
+                University of Oxford, 2022
+-------------------------------------------------------------
+
+problem:
+  variables     = 117
+  constraints   = 493
+  nnz(P)        = 3
+  nnz(A)        = 547
+  cones (total) = 14
+    : Zero        = 1,  numel = 19
+    : Nonnegative = 2,  numel = (18,39)
+    : SecondOrder = 9,  numel = (4,4,4,4,...,4)
+    : PSDTriangle = 2,  numel = (171,210)
+
+settings:
+  linear algebra: direct / qdldl, precision: Float64
+  max iter = 200, time limit = Inf,  max step = 0.990
+  tol_feas = 1.0e-03, tol_gap_abs = 1.0e-08, tol_gap_rel = 1.0e-03,
+  static reg : on, ϵ1 = 1.0e-08, ϵ2 = 4.9e-32
+  dynamic reg: on, ϵ = 1.0e-13, δ = 2.0e-07
+  iter refine: on, reltol = 1.0e-13, abstol = 1.0e-12,
+               max iter = 10, stop ratio = 5.0
+  equilibrate: on, min_scale = 1.0e-04, max_scale = 1.0e+04
+               max iter = 10
+
+iter    pcost        dcost       gap       pres      dres      k/t        μ       step
+---------------------------------------------------------------------------------------------
+  0   6.6673e+03  -3.0980e+05  3.16e+05  1.65e-02  7.87e-01  1.00e+00  3.34e+03   ------
+  1   3.3968e+03  -6.4534e+04  6.79e+04  3.04e-03  2.44e-01  2.41e+02  9.42e+02  8.13e-01
+  2   1.9798e+03  -1.7012e+04  1.90e+04  8.25e-04  9.73e-02  1.51e+02  3.06e+02  9.90e-01
+  3   2.1414e+03  -2.3620e+03  4.50e+03  1.88e-04  2.03e-02  3.29e+01  7.63e+01  8.60e-01
+  4   2.0597e+03   1.1123e+03  8.52e-01  3.78e-05  4.35e-03  5.52e+00  1.93e+01  8.55e-01
+  5   1.6312e+03   1.4437e+03  1.30e-01  5.66e-06  6.85e-04  9.18e-01  3.20e+00  9.12e-01
+  6   1.5768e+03   1.5579e+03  1.21e-02  5.34e-07  6.55e-05  9.28e-02  3.09e-01  9.19e-01
+  7   1.5691e+03   1.5654e+03  2.35e-03  9.56e-08  1.22e-05  1.96e-02  5.73e-02  8.55e-01
+  8   1.5657e+03   1.5640e+03  1.08e-03  3.99e-08  5.32e-06  8.54e-03  2.47e-02  7.64e-01
+  9   1.5653e+03   1.5641e+03  7.65e-04  2.79e-08  3.75e-06  5.95e-03  1.73e-02  4.39e-01
+ 10   1.5645e+03   1.5641e+03  2.72e-04  9.71e-09  1.32e-06  2.11e-03  6.09e-03  6.69e-01
+ 11   1.5644e+03   1.5643e+03  1.04e-04  3.70e-09  5.02e-07  7.73e-04  2.30e-03  8.73e-01
+---------------------------------------------------------------------------------------------
+Terminated with status = solved
+solve time =  192ms</code></pre><pre><code class="language-julia hljs">sdp_relaxation_lower_bound = round(objective_value(model); digits = 2)
+println(
+    &quot;Objective value (W &amp; V relax. lower bound): $sdp_relaxation_lower_bound&quot;,
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Objective value (W &amp; V relax. lower bound): 2753.17</code></pre><p>We can more easily see solution values by rounding out noisy data:</p><pre><code class="language-julia hljs">W_1 = SparseArrays.sparse(round.(value.(W); digits = 2))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9×9 SparseArrays.SparseMatrixCSC{ComplexF64, Int64} with 81 stored entries:
+  1.2-0.0im   1.17-0.07im  1.16-0.05im  …  1.18-0.02im  1.17+0.05im
+ 1.17+0.07im   1.2-0.0im   1.17+0.02im     1.19+0.05im  1.17+0.12im
+ 1.16+0.05im  1.17-0.02im   1.2-0.0im      1.19+0.03im  1.17+0.1im
+ 1.17-0.03im  1.18-0.1im   1.17-0.08im      1.2-0.05im  1.18+0.03im
+ 1.18-0.05im  1.18-0.12im  1.18-0.1im       1.2-0.07im  1.19+0.01im
+ 1.18+0.02im  1.19-0.06im  1.19-0.03im  …  1.21-0.0im   1.19+0.07im
+ 1.18-0.01im  1.19-0.08im  1.18-0.06im      1.2-0.03im  1.19+0.05im
+ 1.18+0.02im  1.19-0.05im  1.19-0.03im     1.21-0.0im   1.19+0.07im
+ 1.17-0.05im  1.17-0.12im  1.17-0.1im      1.19-0.07im  1.18-0.0im</code></pre><p>and recover an approximation to the voltage variables as:</p><pre><code class="language-julia hljs">DataFrames.DataFrame(;
+    Bus = 1:N,
+    Magnitude = round.(abs.(value.(V)); digits = 2),
+    AngleDeg = round.(rad2deg.(angle.(value.(V))); digits = 2),
+)</code></pre><div><div style = "float: left;"><span>9×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Bus</th><th style = "text-align: left;">Magnitude</th><th style = "text-align: left;">AngleDeg</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">1</td><td style = "text-align: right;">0.94</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">2</td><td style = "text-align: right;">0.87</td><td style = "text-align: right;">3.63</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">3</td><td style = "text-align: right;">0.86</td><td style = "text-align: right;">2.56</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">4</td><td style = "text-align: right;">0.88</td><td style = "text-align: right;">-1.23</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">5</td><td style = "text-align: right;">0.88</td><td style = "text-align: right;">-2.1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">6</td><td style = "text-align: right;">0.89</td><td style = "text-align: right;">0.88</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">7</td><td style = "text-align: right;">0.89</td><td style = "text-align: right;">-0.2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">8</td><td style = "text-align: right;">0.89</td><td style = "text-align: right;">1.08</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">9</td><td style = "text-align: right;">0.88</td><td style = "text-align: right;">-2.44</td></tr></tbody></table></div><p>For further information on exploiting sparsity see Jabr (2012).</p><p>This relaxation has the advantage that we can work directly with complex voltages to extend the formulation, strengthen the relaxation and gain additional approximate information about the voltage variables.</p><h2 id="References-and-further-resources"><a class="docs-heading-anchor" href="#References-and-further-resources">References and further resources</a><a id="References-and-further-resources-1"></a><a class="docs-heading-anchor-permalink" href="#References-and-further-resources" title="Permalink"></a></h2><p><strong>Bukhsh</strong>, W. A., Grothey, A., McKinnon, K. I., &amp; Trodden, P. A. <a href="https://doi.org/10.1109/TPWRS.2013.2274577"><em>Local solutions of the optimal power flow problem.</em></a> IEEE Transactions on Power Systems, 28(4), 4780-4788 (2013).</p><p><strong>Jabr</strong>, R. A. <a href="https://doi.org/10.1109/TPWRS.2011.2170772"><em>Exploiting sparsity in SDP relaxations of the OPF problem.</em></a> IEEE Transactions on Power Systems, 27(2), 1138-1139 (2011).</p><p><strong>Krasko</strong>, V., &amp; S. <strong>Rebennack</strong>. <a href="https://doi.org/10.1137/1.9781611974683.ch15"><em>Chapter 15: Global Optimization: Optimal Power Flow Problem.</em></a> In Advances and Trends in Optimization with Engineering Applications, 187—205. MOS-SIAM Series on Optimization. Society for Industrial and Applied Mathematics, 2017.</p><p><a href="https://www.maths.ed.ac.uk/optenergy/LocalOpt/9busnetwork.html"><strong>Test case <code>case9mod</code></strong></a>: from the <a href="https://www.maths.ed.ac.uk/optenergy/LocalOpt/">Test Case Archive of Optimal Power Flow (OPF) Problems with Local Optima</a>.</p><p>R. D. Zimmerman, C. E. Murillo-Sanchez, &amp; R. J. Thomas, <a href="https://doi.org/10.1109/TPWRS.2010.2051168"><em>MATPOWER: Steady-State Operations, Planning and Analysis Tools for Power Systems Research and Education,</em></a> IEEE Transactions on Power Systems, 26(1), 12-19 (2011).</p><p><strong>MATPOWER data format</strong>: <a href="https://matpower.org/docs/MATPOWER-manual.pdf">MATPOWER manual</a>; see especially Appendix B &quot;Data File Format&quot; and Table B-3.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../power_systems/">« Power Systems</a><a class="docs-footer-nextpage" href="../web_app/">Serving web apps »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/applications/power_systems.jl b/previews/PR3525/tutorials/applications/power_systems.jl
new file mode 100644
index 00000000000..b5e9a5d289a
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/power_systems.jl
@@ -0,0 +1,559 @@
+# Copyright (c) 2019 Yury Dvorkin, Miles Lubin, and contributors                #src
+#                                                                               #src
+# This work is licensed under a Creative Commons Attribution-ShareAlike 4.0     #src
+# International License. URL: http://creativecommons.org/licenses/by-sa/4.0.    #src
+
+# # Power Systems
+
+# **This tutorial was originally contributed by Yury Dvorkin and Miles Lubin.**
+
+# This tutorial demonstrates how to formulate basic power systems engineering
+# models in JuMP.
+
+# We will consider basic "economic dispatch" and "unit commitment" models
+# without taking into account transmission constraints.
+
+# For this tutorial, we use the following packages:
+
+using JuMP
+import DataFrames
+import HiGHS
+import Plots
+import StatsPlots
+
+# ## Economic dispatch
+
+# Economic dispatch (ED) is an optimization problem that minimizes the cost of
+# supplying energy demand subject to operational constraints on power system
+# assets. In its simplest modification, ED is an LP problem solved for an
+# aggregated load and wind forecast and for a single infinitesimal moment.
+
+# Mathematically, the ED problem can be written as follows:
+
+# ```math
+# \min \sum_{i \in I} c^g_{i} \cdot g_{i} + c^w \cdot w,
+# ```
+
+# where $c_{i}$ and $g_{i}$ are the incremental cost (\$/MWh) and power
+# output (MW) of the $i^{th}$ generator, respectively, and $c^w$ and $w$ are the
+# incremental cost (\$/MWh) and wind power injection (MW), respectively.
+
+# Subject to the constraints:
+
+# * Minimum ($g^{\min}$) and maximum ($g^{\max}$) limits on power outputs of
+#   generators:
+#   $g^{\min}_{i} \leq g_{i} \leq g^{\max}_{i}.$
+# * Constraint on the wind power injection:
+#   $0 \leq w \leq w^f,$
+#   where $w$ and $w^f$ are the wind power injection and wind power forecast,
+#   respectively.
+# * Power balance constraint:
+#   $\sum_{i \in I} g_{i} + w = d^f,$
+#   where $d^f$ is the demand forecast.
+
+# Further reading on ED models can be found in A. J. Wood, B. F. Wollenberg, and
+# G. B. Sheblé, "Power Generation, Operation and Control," Wiley, 2013.
+
+# Define some input data about the test system.
+
+# We define some thermal generators:
+
+function ThermalGenerator(
+    min::Float64,
+    max::Float64,
+    fixed_cost::Float64,
+    variable_cost::Float64,
+)
+    return (
+        min = min,
+        max = max,
+        fixed_cost = fixed_cost,
+        variable_cost = variable_cost,
+    )
+end
+
+generators = [
+    ThermalGenerator(0.0, 1000.0, 1000.0, 50.0),
+    ThermalGenerator(300.0, 1000.0, 0.0, 100.0),
+]
+
+# A wind generator
+
+WindGenerator(variable_cost::Float64) = (variable_cost = variable_cost,)
+
+wind_generator = WindGenerator(50.0)
+
+# And a scenario
+
+function Scenario(demand::Float64, wind::Float64)
+    return (demand = demand, wind = wind)
+end
+
+scenario = Scenario(1500.0, 200.0)
+
+# Create a function `solve_economic_dispatch`, which solves the economic
+# dispatch problem for a given set of input parameters.
+
+function solve_economic_dispatch(generators::Vector, wind, scenario)
+    ## Define the economic dispatch (ED) model
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    ## Define decision variables
+    ## power output of generators
+    N = length(generators)
+    @variable(model, generators[i].min <= g[i = 1:N] <= generators[i].max)
+    ## wind power injection
+    @variable(model, 0 <= w <= scenario.wind)
+    ## Define the objective function
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * g[i] for i in 1:N) +
+        wind.variable_cost * w,
+    )
+    ## Define the power balance constraint
+    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)
+    ## Solve statement
+    optimize!(model)
+    ## return the optimal value of the objective function and its minimizers
+    return (
+        g = value.(g),
+        w = value(w),
+        wind_spill = scenario.wind - value(w),
+        total_cost = objective_value(model),
+    )
+end
+
+# Solve the economic dispatch problem
+
+solution = solve_economic_dispatch(generators, wind_generator, scenario);
+
+println("Dispatch of Generators: ", solution.g, " MW")
+println("Dispatch of Wind: ", solution.w, " MW")
+println("Wind spillage: ", solution.wind_spill, " MW")
+println("Total cost: \$", solution.total_cost)
+
+# ## Economic dispatch with adjustable incremental costs
+
+# In the following exercise we adjust the incremental cost of generator G1 and
+# observe its impact on the total cost.
+
+function scale_generator_cost(g, scale)
+    return ThermalGenerator(g.min, g.max, g.fixed_cost, scale * g.variable_cost)
+end
+
+start = time()
+c_g_scale_df = DataFrames.DataFrame(;
+    ## Scale factor
+    scale = Float64[],
+    ## Dispatch of Generator 1 [MW]
+    dispatch_G1 = Float64[],
+    ## Dispatch of Generator 2 [MW]
+    dispatch_G2 = Float64[],
+    ## Dispatch of Wind [MW]
+    dispatch_wind = Float64[],
+    ## Spillage of Wind [MW]
+    spillage_wind = Float64[],
+    ## Total cost [$]
+    total_cost = Float64[],
+)
+for c_g1_scale in 0.5:0.1:3.0
+    ## Update the incremental cost of the first generator at every iteration.
+    new_generators = scale_generator_cost.(generators, [c_g1_scale, 1.0])
+    ## Solve the economic-dispatch problem with the updated incremental cost
+    sol = solve_economic_dispatch(new_generators, wind_generator, scenario)
+    push!(
+        c_g_scale_df,
+        (c_g1_scale, sol.g[1], sol.g[2], sol.w, sol.wind_spill, sol.total_cost),
+    )
+end
+print(string("elapsed time: ", time() - start, " seconds"))
+
+#-
+
+c_g_scale_df
+
+# ## Modifying the JuMP model in-place
+
+# Note that in the previous exercise we entirely rebuilt the optimization model
+# at every iteration of the internal loop, which incurs an additional
+# computational burden. This burden can be alleviated if instead of re-building
+# the entire model, we modify the constraints or objective
+# function, as it shown in the example below.
+
+# Compare the computing time in case of the above and below models.
+
+function solve_economic_dispatch_inplace(
+    generators::Vector,
+    wind,
+    scenario,
+    scale::AbstractVector{Float64},
+)
+    obj_out = Float64[]
+    w_out = Float64[]
+    g1_out = Float64[]
+    g2_out = Float64[]
+    ## This function only works for two generators
+    @assert length(generators) == 2
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    N = length(generators)
+    @variable(model, generators[i].min <= g[i = 1:N] <= generators[i].max)
+    @variable(model, 0 <= w <= scenario.wind)
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * g[i] for i in 1:N) +
+        wind.variable_cost * w,
+    )
+    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)
+    for c_g1_scale in scale
+        @objective(
+            model,
+            Min,
+            c_g1_scale * generators[1].variable_cost * g[1] +
+            generators[2].variable_cost * g[2] +
+            wind.variable_cost * w,
+        )
+        optimize!(model)
+        push!(obj_out, objective_value(model))
+        push!(w_out, value(w))
+        push!(g1_out, value(g[1]))
+        push!(g2_out, value(g[2]))
+    end
+    df = DataFrames.DataFrame(;
+        scale = scale,
+        dispatch_G1 = g1_out,
+        dispatch_G2 = g2_out,
+        dispatch_wind = w_out,
+        spillage_wind = scenario.wind .- w_out,
+        total_cost = obj_out,
+    )
+    return df
+end
+
+start = time()
+inplace_df = solve_economic_dispatch_inplace(
+    generators,
+    wind_generator,
+    scenario,
+    0.5:0.1:3.0,
+)
+print(string("elapsed time: ", time() - start, " seconds"))
+
+# For small models, adjusting specific constraints or the objective function is
+# sometimes faster and sometimes slower than re-building the entire model.
+# However, as the problem size increases, updating the model in-place is usually
+# faster.
+
+inplace_df
+
+# ## Inefficient usage of wind generators
+
+# The economic dispatch problem does not perform commitment decisions and, thus,
+# assumes that all generators must be dispatched at least at their minimum power
+# output limit. This approach is not cost efficient and may lead to absurd
+# decisions. For example, if $d = \sum_{i \in I} g^{\min}_{i}$, the wind power
+# injection must be zero, that is, all available wind generation is spilled, to meet
+# the minimum power output constraints on generators.
+
+# In the following example, we adjust the total demand and observed how it
+# affects wind spillage.
+
+demand_scale_df = DataFrames.DataFrame(;
+    demand = Float64[],
+    dispatch_G1 = Float64[],
+    dispatch_G2 = Float64[],
+    dispatch_wind = Float64[],
+    spillage_wind = Float64[],
+    total_cost = Float64[],
+)
+
+function scale_demand(scenario, scale)
+    return Scenario(scale * scenario.demand, scenario.wind)
+end
+
+for demand_scale in 0.2:0.1:1.4
+    new_scenario = scale_demand(scenario, demand_scale)
+    sol = solve_economic_dispatch(generators, wind_generator, new_scenario)
+    push!(
+        demand_scale_df,
+        (
+            new_scenario.demand,
+            sol.g[1],
+            sol.g[2],
+            sol.w,
+            sol.wind_spill,
+            sol.total_cost,
+        ),
+    )
+end
+
+demand_scale_df
+
+#-
+
+dispatch_plot = StatsPlots.@df(
+    demand_scale_df,
+    Plots.plot(
+        :demand,
+        [:dispatch_G1, :dispatch_G2],
+        labels = ["G1" "G2"],
+        title = "Thermal Dispatch",
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = "Demand",
+        ylabel = "Dispatch [MW]",
+    ),
+)
+
+wind_plot = StatsPlots.@df(
+    demand_scale_df,
+    Plots.plot(
+        :demand,
+        [:dispatch_wind, :spillage_wind],
+        labels = ["Dispatch" "Spillage"],
+        title = "Wind",
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = "Demand [MW]",
+        ylabel = "Energy [MW]",
+    ),
+)
+
+Plots.plot(dispatch_plot, wind_plot)
+
+# This particular drawback can be overcome by introducing binary decisions on
+# the "on/off" status of generators. This model is called unit commitment and
+# considered later in these notes.
+
+# For further reading on the interplay between wind generation and the minimum
+# power output constraints of generators, we refer interested readers to R.
+# Baldick, "Wind and energy markets: a case study of Texas," IEEE Systems
+# Journal, vol. 6, pp. 27-34, 2012.
+
+# ## Unit commitment
+
+# The Unit Commitment (UC) model can be obtained from ED model by introducing
+# binary variable associated with each generator. This binary variable can
+# attain two values: if it is "1," the generator is synchronized and, thus, can
+# be dispatched, otherwise, that is, if the binary variable is "0," that generator
+# is not synchronized and its power output is set to 0.
+
+# To obtain the mathematical formulation of the UC model, we will modify the
+# constraints of the ED model as follows:
+
+# ```math
+# g^{\min}_{i} \cdot u_{t,i} \leq g_{i} \leq g^{\max}_{i} \cdot u_{t,i},
+# ```
+# where $u_{i} \in \{0,1\}.$ In this constraint, if $u_{i} = 0$, then
+# $g_{i}  = 0$. On the other hand, if $u_{i} = 1$, then
+# $g^{min}_{i} \leq g_{i} \leq g^{max}_{i}$.
+
+# For further reading on the UC problem we refer interested readers to G.
+# Morales-Espana, J. M. Latorre, and A. Ramos, "Tight and Compact MILP
+# Formulation for the Thermal Unit Commitment Problem," IEEE Transactions on
+# Power Systems, vol. 28, pp. 4897-4908, 2013.
+
+# In the following example we convert the ED model explained above to the UC
+# model.
+
+function solve_unit_commitment(generators::Vector, wind, scenario)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    N = length(generators)
+    @variable(model, 0 <= g[i = 1:N] <= generators[i].max)
+    @variable(model, 0 <= w <= scenario.wind)
+    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)
+    ## !!! New: add binary on-off variables for each generator
+    @variable(model, u[i = 1:N], Bin)
+    @constraint(model, [i = 1:N], g[i] <= generators[i].max * u[i])
+    @constraint(model, [i = 1:N], g[i] >= generators[i].min * u[i])
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * g[i] for i in 1:N) +
+        wind.variable_cost * w +
+        ## !!! new
+        sum(generators[i].fixed_cost * u[i] for i in 1:N)
+    )
+    optimize!(model)
+    status = termination_status(model)
+    if status != OPTIMAL
+        return (status = status,)
+    end
+    return (
+        status = status,
+        g = value.(g),
+        w = value(w),
+        wind_spill = scenario.wind - value(w),
+        u = value.(u),
+        total_cost = objective_value(model),
+    )
+end
+
+# Solve the unit commitment problem
+
+solution = solve_unit_commitment(generators, wind_generator, scenario)
+
+println("Dispatch of Generators: ", solution.g, " MW")
+println("Commitments of Generators: ", solution.u)
+println("Dispatch of Wind: ", solution.w, " MW")
+println("Wind spillage: ", solution.wind_spill, " MW")
+println("Total cost: \$", solution.total_cost)
+
+# ## Unit commitment as a function of demand
+
+# After implementing the unit commitment model, we can now assess the interplay
+# between the minimum power output constraints on generators and wind
+# generation.
+
+uc_df = DataFrames.DataFrame(;
+    demand = Float64[],
+    commitment_G1 = Float64[],
+    commitment_G2 = Float64[],
+    dispatch_G1 = Float64[],
+    dispatch_G2 = Float64[],
+    dispatch_wind = Float64[],
+    spillage_wind = Float64[],
+    total_cost = Float64[],
+)
+
+for demand_scale in 0.2:0.1:1.4
+    new_scenario = scale_demand(scenario, demand_scale)
+    sol = solve_unit_commitment(generators, wind_generator, new_scenario)
+    if sol.status == OPTIMAL
+        push!(
+            uc_df,
+            (
+                new_scenario.demand,
+                sol.u[1],
+                sol.u[2],
+                sol.g[1],
+                sol.g[2],
+                sol.w,
+                sol.wind_spill,
+                sol.total_cost,
+            ),
+        )
+    end
+    println("Status: $(sol.status) for demand_scale = $(demand_scale)")
+end
+
+#-
+
+uc_df
+
+#-
+
+commitment_plot = StatsPlots.@df(
+    uc_df,
+    Plots.plot(
+        :demand,
+        [:commitment_G1, :commitment_G2],
+        labels = ["G1" "G2"],
+        title = "Commitment",
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = "Demand [MW]",
+        ylabel = "Commitment decision {0, 1}",
+    ),
+)
+
+dispatch_plot = StatsPlots.@df(
+    uc_df,
+    Plots.plot(
+        :demand,
+        [:dispatch_G1, :dispatch_G2, :dispatch_wind],
+        labels = ["G1" "G2" "Wind"],
+        title = "Dispatch [MW]",
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = "Demand",
+        ylabel = "Dispatch [MW]",
+    ),
+)
+
+Plots.plot(commitment_plot, dispatch_plot)
+
+# ## Nonlinear economic dispatch
+
+# As a final example, we modify our economic dispatch problem in two ways:
+#
+#  * The thermal cost function is user-defined
+#  * The output of the wind is only the square-root of the dispatch
+
+import Ipopt
+
+"""
+    thermal_cost_function(g)
+
+A user-defined thermal cost function in pure-Julia! You can include
+nonlinearities, and even things like control flow.
+
+!!! warning
+    It's still up to you to make sure that the function has a meaningful
+    derivative.
+"""
+function thermal_cost_function(g)
+    if g <= 500
+        return g
+    else
+        return g + 1e-2 * (g - 500)^2
+    end
+end
+
+function solve_nonlinear_economic_dispatch(
+    generators::Vector,
+    wind,
+    scenario;
+    silent::Bool = false,
+)
+    model = Model(Ipopt.Optimizer)
+    if silent
+        set_silent(model)
+    end
+    @operator(model, op_tcf, 1, thermal_cost_function)
+    N = length(generators)
+    @variable(model, generators[i].min <= g[i = 1:N] <= generators[i].max)
+    @variable(model, 0 <= w <= scenario.wind)
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * op_tcf(g[i]) for i in 1:N) +
+        wind.variable_cost * w,
+    )
+    @constraint(model, sum(g[i] for i in 1:N) + sqrt(w) == scenario.demand)
+    optimize!(model)
+    return (
+        g = value.(g),
+        w = value(w),
+        wind_spill = scenario.wind - value(w),
+        total_cost = objective_value(model),
+    )
+end
+
+solution =
+    solve_nonlinear_economic_dispatch(generators, wind_generator, scenario)
+
+# Now let's see how the wind is dispatched as a function of the cost:
+
+wind_cost = 0.0:1:100
+wind_dispatch = Float64[]
+for c in wind_cost
+    sol = solve_nonlinear_economic_dispatch(
+        generators,
+        WindGenerator(c),
+        scenario;
+        silent = true,
+    )
+    push!(wind_dispatch, sol.w)
+end
+
+Plots.plot(
+    wind_cost,
+    wind_dispatch;
+    xlabel = "Cost",
+    ylabel = "Dispatch [MW]",
+    label = false,
+)
diff --git a/previews/PR3525/tutorials/applications/power_systems/6fee809a.svg b/previews/PR3525/tutorials/applications/power_systems/6fee809a.svg
new file mode 100644
index 00000000000..ea0b3e60abb
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/power_systems/6fee809a.svg
@@ -0,0 +1,43 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip190">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip190)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip191">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip190)" d="M234.125 1423.18 L2352.76 1423.18 L2352.76 47.2441 L234.125 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip192">
+    <rect x="234" y="47" width="2120" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="294.087,1423.18 294.087,47.2441 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="793.764,1423.18 793.764,47.2441 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1293.44,1423.18 1293.44,47.2441 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1793.12,1423.18 1793.12,47.2441 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2292.79,1423.18 2292.79,47.2441 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.125,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="294.087,1423.18 294.087,1404.28 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="793.764,1423.18 793.764,1404.28 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1293.44,1423.18 1293.44,1404.28 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1793.12,1423.18 1793.12,1404.28 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2292.79,1423.18 2292.79,1404.28 "/>
+<path clip-path="url(#clip190)" d="M294.087 1454.1 Q290.476 1454.1 288.647 1457.66 Q286.841 1461.2 286.841 1468.33 Q286.841 1475.44 288.647 1479.01 Q290.476 1482.55 294.087 1482.55 Q297.721 1482.55 299.526 1479.01 Q301.355 1475.44 301.355 1468.33 Q301.355 1461.2 299.526 1457.66 Q297.721 1454.1 294.087 1454.1 M294.087 1450.39 Q299.897 1450.39 302.952 1455 Q306.031 1459.58 306.031 1468.33 Q306.031 1477.06 302.952 1481.67 Q299.897 1486.25 294.087 1486.25 Q288.277 1486.25 285.198 1481.67 Q282.142 1477.06 282.142 1468.33 Q282.142 1459.58 285.198 1455 Q288.277 1450.39 294.087 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M773.035 1481.64 L789.354 1481.64 L789.354 1485.58 L767.41 1485.58 L767.41 1481.64 Q770.072 1478.89 774.655 1474.26 Q779.261 1469.61 780.442 1468.27 Q782.687 1465.74 783.567 1464.01 Q784.47 1462.25 784.47 1460.56 Q784.47 1457.8 782.525 1456.07 Q780.604 1454.33 777.502 1454.33 Q775.303 1454.33 772.849 1455.09 Q770.419 1455.86 767.641 1457.41 L767.641 1452.69 Q770.465 1451.55 772.919 1450.97 Q775.373 1450.39 777.41 1450.39 Q782.78 1450.39 785.974 1453.08 Q789.169 1455.77 789.169 1460.26 Q789.169 1462.39 788.359 1464.31 Q787.572 1466.2 785.465 1468.8 Q784.886 1469.47 781.785 1472.69 Q778.683 1475.88 773.035 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M799.215 1451.02 L817.571 1451.02 L817.571 1454.96 L803.497 1454.96 L803.497 1463.43 Q804.516 1463.08 805.534 1462.92 Q806.553 1462.73 807.571 1462.73 Q813.358 1462.73 816.738 1465.9 Q820.118 1469.08 820.118 1474.49 Q820.118 1480.07 816.646 1483.17 Q813.173 1486.25 806.854 1486.25 Q804.678 1486.25 802.409 1485.88 Q800.164 1485.51 797.757 1484.77 L797.757 1480.07 Q799.84 1481.2 802.062 1481.76 Q804.284 1482.32 806.761 1482.32 Q810.766 1482.32 813.104 1480.21 Q815.442 1478.1 815.442 1474.49 Q815.442 1470.88 813.104 1468.77 Q810.766 1466.67 806.761 1466.67 Q804.886 1466.67 803.011 1467.08 Q801.159 1467.5 799.215 1468.38 L799.215 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1268.14 1451.02 L1286.5 1451.02 L1286.5 1454.96 L1272.42 1454.96 L1272.42 1463.43 Q1273.44 1463.08 1274.46 1462.92 Q1275.48 1462.73 1276.5 1462.73 Q1282.28 1462.73 1285.66 1465.9 Q1289.04 1469.08 1289.04 1474.49 Q1289.04 1480.07 1285.57 1483.17 Q1282.1 1486.25 1275.78 1486.25 Q1273.6 1486.25 1271.33 1485.88 Q1269.09 1485.51 1266.68 1484.77 L1266.68 1480.07 Q1268.76 1481.2 1270.99 1481.76 Q1273.21 1482.32 1275.69 1482.32 Q1279.69 1482.32 1282.03 1480.21 Q1284.37 1478.1 1284.37 1474.49 Q1284.37 1470.88 1282.03 1468.77 Q1279.69 1466.67 1275.69 1466.67 Q1273.81 1466.67 1271.94 1467.08 Q1270.08 1467.5 1268.14 1468.38 L1268.14 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1308.26 1454.1 Q1304.64 1454.1 1302.82 1457.66 Q1301.01 1461.2 1301.01 1468.33 Q1301.01 1475.44 1302.82 1479.01 Q1304.64 1482.55 1308.26 1482.55 Q1311.89 1482.55 1313.7 1479.01 Q1315.52 1475.44 1315.52 1468.33 Q1315.52 1461.2 1313.7 1457.66 Q1311.89 1454.1 1308.26 1454.1 M1308.26 1450.39 Q1314.07 1450.39 1317.12 1455 Q1320.2 1459.58 1320.2 1468.33 Q1320.2 1477.06 1317.12 1481.67 Q1314.07 1486.25 1308.26 1486.25 Q1302.45 1486.25 1299.37 1481.67 Q1296.31 1477.06 1296.31 1468.33 Q1296.31 1459.58 1299.37 1455 Q1302.45 1450.39 1308.26 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1766.97 1451.02 L1789.19 1451.02 L1789.19 1453.01 L1776.65 1485.58 L1771.76 1485.58 L1783.57 1454.96 L1766.97 1454.96 L1766.97 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1798.36 1451.02 L1816.72 1451.02 L1816.72 1454.96 L1802.64 1454.96 L1802.64 1463.43 Q1803.66 1463.08 1804.68 1462.92 Q1805.7 1462.73 1806.72 1462.73 Q1812.5 1462.73 1815.88 1465.9 Q1819.26 1469.08 1819.26 1474.49 Q1819.26 1480.07 1815.79 1483.17 Q1812.32 1486.25 1806 1486.25 Q1803.82 1486.25 1801.56 1485.88 Q1799.31 1485.51 1796.9 1484.77 L1796.9 1480.07 Q1798.99 1481.2 1801.21 1481.76 Q1803.43 1482.32 1805.91 1482.32 Q1809.91 1482.32 1812.25 1480.21 Q1814.59 1478.1 1814.59 1474.49 Q1814.59 1470.88 1812.25 1468.77 Q1809.91 1466.67 1805.91 1466.67 Q1804.03 1466.67 1802.16 1467.08 Q1800.31 1467.5 1798.36 1468.38 L1798.36 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M2252.4 1481.64 L2260.04 1481.64 L2260.04 1455.28 L2251.73 1456.95 L2251.73 1452.69 L2259.99 1451.02 L2264.67 1451.02 L2264.67 1481.64 L2272.31 1481.64 L2272.31 1485.58 L2252.4 1485.58 L2252.4 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M2291.75 1454.1 Q2288.14 1454.1 2286.31 1457.66 Q2284.51 1461.2 2284.51 1468.33 Q2284.51 1475.44 2286.31 1479.01 Q2288.14 1482.55 2291.75 1482.55 Q2295.39 1482.55 2297.19 1479.01 Q2299.02 1475.44 2299.02 1468.33 Q2299.02 1461.2 2297.19 1457.66 Q2295.39 1454.1 2291.75 1454.1 M2291.75 1450.39 Q2297.56 1450.39 2300.62 1455 Q2303.7 1459.58 2303.7 1468.33 Q2303.7 1477.06 2300.62 1481.67 Q2297.56 1486.25 2291.75 1486.25 Q2285.94 1486.25 2282.86 1481.67 Q2279.81 1477.06 2279.81 1468.33 Q2279.81 1459.58 2282.86 1455 Q2285.94 1450.39 2291.75 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M2321.91 1454.1 Q2318.3 1454.1 2316.48 1457.66 Q2314.67 1461.2 2314.67 1468.33 Q2314.67 1475.44 2316.48 1479.01 Q2318.3 1482.55 2321.91 1482.55 Q2325.55 1482.55 2327.35 1479.01 Q2329.18 1475.44 2329.18 1468.33 Q2329.18 1461.2 2327.35 1457.66 Q2325.55 1454.1 2321.91 1454.1 M2321.91 1450.39 Q2327.73 1450.39 2330.78 1455 Q2333.86 1459.58 2333.86 1468.33 Q2333.86 1477.06 2330.78 1481.67 Q2327.73 1486.25 2321.91 1486.25 Q2316.1 1486.25 2313.03 1481.67 Q2309.97 1477.06 2309.97 1468.33 Q2309.97 1459.58 2313.03 1455 Q2316.1 1450.39 2321.91 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1261.91 1524.18 L1261.91 1530.96 Q1258.67 1527.94 1254.98 1526.44 Q1251.32 1524.95 1247.18 1524.95 Q1239.03 1524.95 1234.7 1529.95 Q1230.37 1534.91 1230.37 1544.33 Q1230.37 1553.72 1234.7 1558.72 Q1239.03 1563.68 1247.18 1563.68 Q1251.32 1563.68 1254.98 1562.19 Q1258.67 1560.69 1261.91 1557.67 L1261.91 1564.38 Q1258.54 1566.68 1254.75 1567.82 Q1251 1568.97 1246.8 1568.97 Q1236.01 1568.97 1229.8 1562.38 Q1223.59 1555.76 1223.59 1544.33 Q1223.59 1532.87 1229.8 1526.28 Q1236.01 1519.66 1246.8 1519.66 Q1251.06 1519.66 1254.82 1520.81 Q1258.6 1521.92 1261.91 1524.18 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1285.4 1536.5 Q1280.69 1536.5 1277.96 1540.19 Q1275.22 1543.85 1275.22 1550.25 Q1275.22 1556.65 1277.92 1560.34 Q1280.66 1564 1285.4 1564 Q1290.08 1564 1292.82 1560.31 Q1295.56 1556.62 1295.56 1550.25 Q1295.56 1543.92 1292.82 1540.23 Q1290.08 1536.5 1285.4 1536.5 M1285.4 1531.54 Q1293.04 1531.54 1297.4 1536.5 Q1301.76 1541.47 1301.76 1550.25 Q1301.76 1559 1297.4 1564 Q1293.04 1568.97 1285.4 1568.97 Q1277.73 1568.97 1273.37 1564 Q1269.04 1559 1269.04 1550.25 Q1269.04 1541.47 1273.37 1536.5 Q1277.73 1531.54 1285.4 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1334.2 1533.45 L1334.2 1538.98 Q1331.71 1537.71 1329.04 1537.07 Q1326.37 1536.44 1323.5 1536.44 Q1319.14 1536.44 1316.95 1537.77 Q1314.78 1539.11 1314.78 1541.79 Q1314.78 1543.82 1316.34 1545 Q1317.9 1546.15 1322.61 1547.2 L1324.62 1547.64 Q1330.86 1548.98 1333.47 1551.43 Q1336.11 1553.85 1336.11 1558.21 Q1336.11 1563.17 1332.16 1566.07 Q1328.25 1568.97 1321.37 1568.97 Q1318.51 1568.97 1315.39 1568.39 Q1312.3 1567.85 1308.86 1566.74 L1308.86 1560.69 Q1312.11 1562.38 1315.26 1563.24 Q1318.41 1564.07 1321.5 1564.07 Q1325.64 1564.07 1327.86 1562.66 Q1330.09 1561.23 1330.09 1558.65 Q1330.09 1556.27 1328.47 1554.99 Q1326.88 1553.72 1321.43 1552.54 L1319.4 1552.07 Q1313.95 1550.92 1311.54 1548.56 Q1309.12 1546.18 1309.12 1542.04 Q1309.12 1537.01 1312.68 1534.27 Q1316.25 1531.54 1322.8 1531.54 Q1326.05 1531.54 1328.91 1532.01 Q1331.78 1532.49 1334.2 1533.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M1351.23 1522.27 L1351.23 1532.4 L1363.29 1532.4 L1363.29 1536.95 L1351.23 1536.95 L1351.23 1556.3 Q1351.23 1560.66 1352.4 1561.9 Q1353.61 1563.14 1357.27 1563.14 L1363.29 1563.14 L1363.29 1568.04 L1357.27 1568.04 Q1350.49 1568.04 1347.92 1565.53 Q1345.34 1562.98 1345.34 1556.3 L1345.34 1536.95 L1341.04 1536.95 L1341.04 1532.4 L1345.34 1532.4 L1345.34 1522.27 L1351.23 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.125,1410.55 2352.76,1410.55 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.125,1079.46 2352.76,1079.46 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.125,748.368 2352.76,748.368 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.125,417.277 2352.76,417.277 "/>
+<polyline clip-path="url(#clip192)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.125,86.1857 2352.76,86.1857 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.125,1423.18 234.125,47.2441 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.125,1410.55 253.023,1410.55 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.125,1079.46 253.023,1079.46 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.125,748.368 253.023,748.368 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.125,417.277 253.023,417.277 "/>
+<polyline clip-path="url(#clip190)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.125,86.1857 253.023,86.1857 "/>
+<path clip-path="url(#clip190)" d="M186.181 1396.35 Q182.57 1396.35 180.741 1399.91 Q178.936 1403.45 178.936 1410.58 Q178.936 1417.69 180.741 1421.26 Q182.57 1424.8 186.181 1424.8 Q189.815 1424.8 191.621 1421.26 Q193.45 1417.69 193.45 1410.58 Q193.45 1403.45 191.621 1399.91 Q189.815 1396.35 186.181 1396.35 M186.181 1392.64 Q191.991 1392.64 195.047 1397.25 Q198.125 1401.83 198.125 1410.58 Q198.125 1419.31 195.047 1423.92 Q191.991 1428.5 186.181 1428.5 Q180.371 1428.5 177.292 1423.92 Q174.237 1419.31 174.237 1410.58 Q174.237 1401.83 177.292 1397.25 Q180.371 1392.64 186.181 1392.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M146.066 1062.18 L164.422 1062.18 L164.422 1066.11 L150.348 1066.11 L150.348 1074.59 Q151.366 1074.24 152.385 1074.08 Q153.403 1073.89 154.422 1073.89 Q160.209 1073.89 163.589 1077.06 Q166.968 1080.23 166.968 1085.65 Q166.968 1091.23 163.496 1094.33 Q160.024 1097.41 153.704 1097.41 Q151.528 1097.41 149.26 1097.04 Q147.015 1096.67 144.607 1095.93 L144.607 1091.23 Q146.691 1092.36 148.913 1092.92 Q151.135 1093.47 153.612 1093.47 Q157.616 1093.47 159.954 1091.37 Q162.292 1089.26 162.292 1085.65 Q162.292 1082.04 159.954 1079.93 Q157.616 1077.83 153.612 1077.83 Q151.737 1077.83 149.862 1078.24 Q148.01 1078.66 146.066 1079.54 L146.066 1062.18 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M186.181 1065.26 Q182.57 1065.26 180.741 1068.82 Q178.936 1072.36 178.936 1079.49 Q178.936 1086.6 180.741 1090.16 Q182.57 1093.71 186.181 1093.71 Q189.815 1093.71 191.621 1090.16 Q193.45 1086.6 193.45 1079.49 Q193.45 1072.36 191.621 1068.82 Q189.815 1065.26 186.181 1065.26 M186.181 1061.55 Q191.991 1061.55 195.047 1066.16 Q198.125 1070.74 198.125 1079.49 Q198.125 1088.22 195.047 1092.83 Q191.991 1097.41 186.181 1097.41 Q180.371 1097.41 177.292 1092.83 Q174.237 1088.22 174.237 1079.49 Q174.237 1070.74 177.292 1066.16 Q180.371 1061.55 186.181 1061.55 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M116.668 761.713 L124.306 761.713 L124.306 735.347 L115.996 737.014 L115.996 732.754 L124.26 731.088 L128.936 731.088 L128.936 761.713 L136.575 761.713 L136.575 765.648 L116.668 765.648 L116.668 761.713 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M156.019 734.166 Q152.408 734.166 150.579 737.731 Q148.774 741.273 148.774 748.402 Q148.774 755.509 150.579 759.074 Q152.408 762.615 156.019 762.615 Q159.653 762.615 161.459 759.074 Q163.288 755.509 163.288 748.402 Q163.288 741.273 161.459 737.731 Q159.653 734.166 156.019 734.166 M156.019 730.463 Q161.829 730.463 164.885 735.069 Q167.964 739.653 167.964 748.402 Q167.964 757.129 164.885 761.736 Q161.829 766.319 156.019 766.319 Q150.209 766.319 147.13 761.736 Q144.075 757.129 144.075 748.402 Q144.075 739.653 147.13 735.069 Q150.209 730.463 156.019 730.463 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M186.181 734.166 Q182.57 734.166 180.741 737.731 Q178.936 741.273 178.936 748.402 Q178.936 755.509 180.741 759.074 Q182.57 762.615 186.181 762.615 Q189.815 762.615 191.621 759.074 Q193.45 755.509 193.45 748.402 Q193.45 741.273 191.621 737.731 Q189.815 734.166 186.181 734.166 M186.181 730.463 Q191.991 730.463 195.047 735.069 Q198.125 739.653 198.125 748.402 Q198.125 757.129 195.047 761.736 Q191.991 766.319 186.181 766.319 Q180.371 766.319 177.292 761.736 Q174.237 757.129 174.237 748.402 Q174.237 739.653 177.292 735.069 Q180.371 730.463 186.181 730.463 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M116.668 430.622 L124.306 430.622 L124.306 404.256 L115.996 405.923 L115.996 401.663 L124.26 399.997 L128.936 399.997 L128.936 430.622 L136.575 430.622 L136.575 434.557 L116.668 434.557 L116.668 430.622 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M146.066 399.997 L164.422 399.997 L164.422 403.932 L150.348 403.932 L150.348 412.404 Q151.366 412.057 152.385 411.895 Q153.403 411.71 154.422 411.71 Q160.209 411.71 163.589 414.881 Q166.968 418.052 166.968 423.469 Q166.968 429.047 163.496 432.149 Q160.024 435.228 153.704 435.228 Q151.528 435.228 149.26 434.858 Q147.015 434.487 144.607 433.747 L144.607 429.047 Q146.691 430.182 148.913 430.737 Q151.135 431.293 153.612 431.293 Q157.616 431.293 159.954 429.186 Q162.292 427.08 162.292 423.469 Q162.292 419.858 159.954 417.751 Q157.616 415.645 153.612 415.645 Q151.737 415.645 149.862 416.061 Q148.01 416.478 146.066 417.358 L146.066 399.997 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M186.181 403.075 Q182.57 403.075 180.741 406.64 Q178.936 410.182 178.936 417.311 Q178.936 424.418 180.741 427.983 Q182.57 431.524 186.181 431.524 Q189.815 431.524 191.621 427.983 Q193.45 424.418 193.45 417.311 Q193.45 410.182 191.621 406.64 Q189.815 403.075 186.181 403.075 M186.181 399.372 Q191.991 399.372 195.047 403.978 Q198.125 408.561 198.125 417.311 Q198.125 426.038 195.047 430.645 Q191.991 435.228 186.181 435.228 Q180.371 435.228 177.292 430.645 Q174.237 426.038 174.237 417.311 Q174.237 408.561 177.292 403.978 Q180.371 399.372 186.181 399.372 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M119.885 99.5305 L136.204 99.5305 L136.204 103.466 L114.26 103.466 L114.26 99.5305 Q116.922 96.7759 121.505 92.1463 Q126.112 87.4935 127.292 86.151 Q129.538 83.6278 130.417 81.8917 Q131.32 80.1325 131.32 78.4427 Q131.32 75.6881 129.376 73.952 Q127.455 72.2159 124.353 72.2159 Q122.154 72.2159 119.7 72.9797 Q117.269 73.7436 114.492 75.2945 L114.492 70.5723 Q117.316 69.4381 119.769 68.8594 Q122.223 68.2807 124.26 68.2807 Q129.63 68.2807 132.825 70.9659 Q136.019 73.651 136.019 78.1417 Q136.019 80.2714 135.209 82.1927 Q134.422 84.0908 132.316 86.6834 Q131.737 87.3547 128.635 90.5722 Q125.533 93.7667 119.885 99.5305 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M156.019 71.9844 Q152.408 71.9844 150.579 75.5492 Q148.774 79.0908 148.774 86.2204 Q148.774 93.3269 150.579 96.8916 Q152.408 100.433 156.019 100.433 Q159.653 100.433 161.459 96.8916 Q163.288 93.3269 163.288 86.2204 Q163.288 79.0908 161.459 75.5492 Q159.653 71.9844 156.019 71.9844 M156.019 68.2807 Q161.829 68.2807 164.885 72.8871 Q167.964 77.4705 167.964 86.2204 Q167.964 94.9472 164.885 99.5537 Q161.829 104.137 156.019 104.137 Q150.209 104.137 147.13 99.5537 Q144.075 94.9472 144.075 86.2204 Q144.075 77.4705 147.13 72.8871 Q150.209 68.2807 156.019 68.2807 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M186.181 71.9844 Q182.57 71.9844 180.741 75.5492 Q178.936 79.0908 178.936 86.2204 Q178.936 93.3269 180.741 96.8916 Q182.57 100.433 186.181 100.433 Q189.815 100.433 191.621 96.8916 Q193.45 93.3269 193.45 86.2204 Q193.45 79.0908 191.621 75.5492 Q189.815 71.9844 186.181 71.9844 M186.181 68.2807 Q191.991 68.2807 195.047 72.8871 Q198.125 77.4705 198.125 86.2204 Q198.125 94.9472 195.047 99.5537 Q191.991 104.137 186.181 104.137 Q180.371 104.137 177.292 99.5537 Q174.237 94.9472 174.237 86.2204 Q174.237 77.4705 177.292 72.8871 Q180.371 68.2807 186.181 68.2807 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M21.7677 962.07 L58.7206 962.07 L58.7206 954.304 Q58.7206 944.469 54.2646 939.917 Q49.8086 935.334 40.1964 935.334 Q30.6479 935.334 26.2237 939.917 Q21.7677 944.469 21.7677 954.304 L21.7677 962.07 M16.4842 968.499 L16.4842 955.29 Q16.4842 941.477 22.2451 935.016 Q27.9743 928.554 40.1964 928.554 Q52.4822 928.554 58.2432 935.047 Q64.0042 941.54 64.0042 955.29 L64.0042 968.499 L16.4842 968.499 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M28.3562 918.56 L28.3562 912.704 L64.0042 912.704 L64.0042 918.56 L28.3562 918.56 M14.479 918.56 L14.479 912.704 L21.895 912.704 L21.895 918.56 L14.479 918.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M29.4065 877.724 L34.9447 877.724 Q33.6716 880.207 33.035 882.88 Q32.3984 885.554 32.3984 888.419 Q32.3984 892.779 33.7352 894.975 Q35.072 897.14 37.7456 897.14 Q39.7826 897.14 40.9603 895.58 Q42.1061 894.02 43.1565 889.31 L43.6021 887.305 Q44.9389 881.066 47.3897 878.456 Q49.8086 875.815 54.1691 875.815 Q59.1344 875.815 62.0308 879.761 Q64.9272 883.676 64.9272 890.551 Q64.9272 893.416 64.3543 896.535 Q63.8132 899.622 62.6992 903.06 L56.6518 903.06 Q58.3387 899.813 59.198 896.662 Q60.0256 893.511 60.0256 890.424 Q60.0256 886.286 58.6251 884.058 Q57.1929 881.83 54.6147 881.83 Q52.2276 881.83 50.9545 883.453 Q49.6813 885.045 48.5037 890.487 L48.0262 892.524 Q46.8804 897.967 44.5251 900.386 Q42.138 902.805 38.0002 902.805 Q32.9713 902.805 30.2341 899.24 Q27.4968 895.676 27.4968 889.119 Q27.4968 885.872 27.9743 883.008 Q28.4517 880.143 29.4065 877.724 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M58.657 860.823 L77.5631 860.823 L77.5631 866.712 L28.3562 866.712 L28.3562 860.823 L33.7671 860.823 Q30.5842 858.977 29.0564 856.176 Q27.4968 853.344 27.4968 849.429 Q27.4968 842.936 32.6531 838.893 Q37.8093 834.819 46.212 834.819 Q54.6147 834.819 59.771 838.893 Q64.9272 842.936 64.9272 849.429 Q64.9272 853.344 63.3994 856.176 Q61.8398 858.977 58.657 860.823 M46.212 840.899 Q39.7508 840.899 36.0905 843.572 Q32.3984 846.214 32.3984 850.861 Q32.3984 855.508 36.0905 858.182 Q39.7508 860.823 46.212 860.823 Q52.6732 860.823 56.3653 858.182 Q60.0256 855.508 60.0256 850.861 Q60.0256 846.214 56.3653 843.572 Q52.6732 840.899 46.212 840.899 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M46.0847 808.911 Q46.0847 816.009 47.7079 818.746 Q49.3312 821.483 53.2461 821.483 Q56.3653 821.483 58.2114 819.446 Q60.0256 817.377 60.0256 813.844 Q60.0256 808.975 56.5881 806.046 Q53.1188 803.086 47.3897 803.086 L46.0847 803.086 L46.0847 808.911 M43.6657 797.23 L64.0042 797.23 L64.0042 803.086 L58.5933 803.086 Q61.8398 805.092 63.3994 808.083 Q64.9272 811.075 64.9272 815.404 Q64.9272 820.878 61.8716 824.125 Q58.7843 827.34 53.6281 827.34 Q47.6125 827.34 44.5569 823.329 Q41.5014 819.287 41.5014 811.298 L41.5014 803.086 L40.9285 803.086 Q36.8862 803.086 34.6901 805.76 Q32.4621 808.402 32.4621 813.208 Q32.4621 816.263 33.1941 819.16 Q33.9262 822.056 35.3903 824.73 L29.9795 824.73 Q28.7381 821.515 28.1334 818.491 Q27.4968 815.468 27.4968 812.603 Q27.4968 804.869 31.5072 801.049 Q35.5176 797.23 43.6657 797.23 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M18.2347 779.374 L28.3562 779.374 L28.3562 767.311 L32.9077 767.311 L32.9077 779.374 L52.2594 779.374 Q56.6199 779.374 57.8613 778.196 Q59.1026 776.987 59.1026 773.327 L59.1026 767.311 L64.0042 767.311 L64.0042 773.327 Q64.0042 780.106 61.4897 782.684 Q58.9434 785.262 52.2594 785.262 L32.9077 785.262 L32.9077 789.559 L28.3562 789.559 L28.3562 785.262 L18.2347 785.262 L18.2347 779.374 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M29.7248 733.955 L35.1993 733.955 Q33.8307 736.437 33.1623 738.952 Q32.4621 741.434 32.4621 743.981 Q32.4621 749.678 36.0905 752.829 Q39.6872 755.98 46.212 755.98 Q52.7369 755.98 56.3653 752.829 Q59.9619 749.678 59.9619 743.981 Q59.9619 741.434 59.2935 738.952 Q58.5933 736.437 57.2247 733.955 L62.6355 733.955 Q63.7814 736.406 64.3543 739.047 Q64.9272 741.657 64.9272 744.617 Q64.9272 752.67 59.8664 757.412 Q54.8057 762.155 46.212 762.155 Q37.491 762.155 32.4939 757.381 Q27.4968 752.574 27.4968 744.235 Q27.4968 741.53 28.0697 738.952 Q28.6108 736.374 29.7248 733.955 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M42.4881 694.137 L64.0042 694.137 L64.0042 699.994 L42.679 699.994 Q37.6183 699.994 35.1038 701.967 Q32.5894 703.94 32.5894 707.887 Q32.5894 712.63 35.6131 715.367 Q38.6368 718.104 43.8567 718.104 L64.0042 718.104 L64.0042 723.992 L14.479 723.992 L14.479 718.104 L33.8944 718.104 Q30.6797 716.003 29.0883 713.171 Q27.4968 710.306 27.4968 706.582 Q27.4968 700.439 31.3163 697.288 Q35.1038 694.137 42.4881 694.137 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M14.479 662.277 L14.479 648.782 L19.0304 648.782 L19.0304 656.42 L68.0464 656.42 L68.0464 648.782 L72.5979 648.782 L72.5979 662.277 L14.479 662.277 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M16.4842 636.05 L16.4842 626.47 L48.8219 614.343 L16.4842 602.153 L16.4842 592.572 L64.0042 592.572 L64.0042 598.843 L22.277 598.843 L54.8694 611.097 L54.8694 617.558 L22.277 629.812 L64.0042 629.812 L64.0042 636.05 L16.4842 636.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M16.4842 584.042 L16.4842 577.549 L56.6518 567.555 L16.4842 557.593 L16.4842 550.368 L56.6518 540.374 L16.4842 530.411 L16.4842 523.886 L64.0042 535.822 L64.0042 543.907 L22.7544 553.933 L64.0042 564.054 L64.0042 572.138 L16.4842 584.042 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip190)" d="M14.479 501.925 L72.5979 501.925 L72.5979 515.42 L68.0464 515.42 L68.0464 507.813 L19.0304 507.813 L19.0304 515.42 L14.479 515.42 L14.479 501.925 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip192)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="294.087,86.1857 314.074,86.1857 334.061,86.1857 354.048,86.1857 374.035,86.1857 394.022,86.1857 414.009,86.1857 433.996,86.1857 453.983,86.1857 473.97,86.1857 493.957,86.1857 513.945,86.1857 533.932,86.1857 553.919,86.1857 573.906,121.282 593.893,283.964 613.88,417.69 633.867,528.946 653.854,622.498 673.841,701.913 693.828,769.901 713.815,828.554 733.802,879.506 753.79,924.049 773.777,963.213 793.764,997.832 813.751,1028.58 833.738,1056.02 853.725,1080.6 873.712,1102.72 893.699,1122.68 913.686,1140.76 933.673,1157.19 953.66,1172.16 973.647,1185.85 993.634,1198.39 1013.62,1209.91 1033.61,1220.51 1053.6,1230.3 1073.58,1239.35 1093.57,1247.73 1113.56,1255.51 1133.54,1262.75 1153.53,1269.49 1173.52,1275.78 1193.51,1281.66 1213.49,1287.17 1233.48,1292.33 1253.47,1297.17 1273.45,1301.72 1293.44,1306 1313.43,1310.03 1333.41,1313.84 1353.4,1317.43 1373.39,1320.83 1393.38,1324.04 1413.36,1327.08 1433.35,1329.97 1453.34,1332.71 1473.32,1335.31 1493.31,1337.78 1513.3,1340.14 1533.29,1342.38 1553.27,1344.52 1573.26,1346.55 1593.25,1348.5 1613.23,1350.35 1633.22,1352.13 1653.21,1353.83 1673.2,1355.45 1693.18,1357.01 1713.17,1358.5 1733.16,1359.93 1753.14,1361.3 1773.13,1362.61 1793.12,1363.88 1813.1,1365.09 1833.09,1366.26 1853.08,1367.38 1873.07,1368.46 1893.05,1369.51 1913.04,1370.51 1933.03,1371.48 1953.01,1372.41 1973,1373.31 1992.99,1374.18 2012.98,1375.01 2032.96,1375.82 2052.95,1376.6 2072.94,1377.36 2092.92,1378.09 2112.91,1378.8 2132.9,1379.48 2152.89,1380.14 2172.87,1380.78 2192.86,1381.41 2212.85,1382.01 2232.83,1382.59 2252.82,1383.16 2272.81,1383.71 2292.79,1384.24 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/applications/power_systems/9f136f96.svg b/previews/PR3525/tutorials/applications/power_systems/9f136f96.svg
new file mode 100644
index 00000000000..a06b7122099
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/power_systems/9f136f96.svg
@@ -0,0 +1,80 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip130">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip130)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip131">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip130)" d="M249.542 1423.18 L1146.25 1423.18 L1146.25 123.472 L249.542 123.472  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip132">
+    <rect x="249" y="123" width="898" height="1301"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="368.915,1423.18 368.915,123.472 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="603.902,1423.18 603.902,123.472 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="838.889,1423.18 838.889,123.472 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1073.88,1423.18 1073.88,123.472 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="249.542,1423.18 1146.25,1423.18 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="368.915,1423.18 368.915,1404.28 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="603.902,1423.18 603.902,1404.28 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="838.889,1423.18 838.889,1404.28 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1073.88,1423.18 1073.88,1404.28 "/>
+<path clip-path="url(#clip130)" d="M328.534 1451.02 L346.89 1451.02 L346.89 1454.96 L332.816 1454.96 L332.816 1463.43 Q333.834 1463.08 334.853 1462.92 Q335.871 1462.73 336.89 1462.73 Q342.677 1462.73 346.057 1465.9 Q349.436 1469.08 349.436 1474.49 Q349.436 1480.07 345.964 1483.17 Q342.492 1486.25 336.172 1486.25 Q333.996 1486.25 331.728 1485.88 Q329.483 1485.51 327.075 1484.77 L327.075 1480.07 Q329.159 1481.2 331.381 1481.76 Q333.603 1482.32 336.08 1482.32 Q340.084 1482.32 342.422 1480.21 Q344.76 1478.1 344.76 1474.49 Q344.76 1470.88 342.422 1468.77 Q340.084 1466.67 336.08 1466.67 Q334.205 1466.67 332.33 1467.08 Q330.478 1467.5 328.534 1468.38 L328.534 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M368.649 1454.1 Q365.038 1454.1 363.209 1457.66 Q361.404 1461.2 361.404 1468.33 Q361.404 1475.44 363.209 1479.01 Q365.038 1482.55 368.649 1482.55 Q372.283 1482.55 374.089 1479.01 Q375.918 1475.44 375.918 1468.33 Q375.918 1461.2 374.089 1457.66 Q372.283 1454.1 368.649 1454.1 M368.649 1450.39 Q374.459 1450.39 377.515 1455 Q380.593 1459.58 380.593 1468.33 Q380.593 1477.06 377.515 1481.67 Q374.459 1486.25 368.649 1486.25 Q362.839 1486.25 359.76 1481.67 Q356.705 1477.06 356.705 1468.33 Q356.705 1459.58 359.76 1455 Q362.839 1450.39 368.649 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M398.811 1454.1 Q395.2 1454.1 393.371 1457.66 Q391.566 1461.2 391.566 1468.33 Q391.566 1475.44 393.371 1479.01 Q395.2 1482.55 398.811 1482.55 Q402.445 1482.55 404.251 1479.01 Q406.079 1475.44 406.079 1468.33 Q406.079 1461.2 404.251 1457.66 Q402.445 1454.1 398.811 1454.1 M398.811 1450.39 Q404.621 1450.39 407.677 1455 Q410.755 1459.58 410.755 1468.33 Q410.755 1477.06 407.677 1481.67 Q404.621 1486.25 398.811 1486.25 Q393.001 1486.25 389.922 1481.67 Q386.867 1477.06 386.867 1468.33 Q386.867 1459.58 389.922 1455 Q393.001 1450.39 398.811 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M548.428 1481.64 L556.067 1481.64 L556.067 1455.28 L547.756 1456.95 L547.756 1452.69 L556.02 1451.02 L560.696 1451.02 L560.696 1481.64 L568.335 1481.64 L568.335 1485.58 L548.428 1485.58 L548.428 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M587.779 1454.1 Q584.168 1454.1 582.34 1457.66 Q580.534 1461.2 580.534 1468.33 Q580.534 1475.44 582.34 1479.01 Q584.168 1482.55 587.779 1482.55 Q591.414 1482.55 593.219 1479.01 Q595.048 1475.44 595.048 1468.33 Q595.048 1461.2 593.219 1457.66 Q591.414 1454.1 587.779 1454.1 M587.779 1450.39 Q593.59 1450.39 596.645 1455 Q599.724 1459.58 599.724 1468.33 Q599.724 1477.06 596.645 1481.67 Q593.59 1486.25 587.779 1486.25 Q581.969 1486.25 578.891 1481.67 Q575.835 1477.06 575.835 1468.33 Q575.835 1459.58 578.891 1455 Q581.969 1450.39 587.779 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M617.941 1454.1 Q614.33 1454.1 612.501 1457.66 Q610.696 1461.2 610.696 1468.33 Q610.696 1475.44 612.501 1479.01 Q614.33 1482.55 617.941 1482.55 Q621.576 1482.55 623.381 1479.01 Q625.21 1475.44 625.21 1468.33 Q625.21 1461.2 623.381 1457.66 Q621.576 1454.1 617.941 1454.1 M617.941 1450.39 Q623.751 1450.39 626.807 1455 Q629.886 1459.58 629.886 1468.33 Q629.886 1477.06 626.807 1481.67 Q623.751 1486.25 617.941 1486.25 Q612.131 1486.25 609.052 1481.67 Q605.997 1477.06 605.997 1468.33 Q605.997 1459.58 609.052 1455 Q612.131 1450.39 617.941 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M648.103 1454.1 Q644.492 1454.1 642.663 1457.66 Q640.858 1461.2 640.858 1468.33 Q640.858 1475.44 642.663 1479.01 Q644.492 1482.55 648.103 1482.55 Q651.737 1482.55 653.543 1479.01 Q655.372 1475.44 655.372 1468.33 Q655.372 1461.2 653.543 1457.66 Q651.737 1454.1 648.103 1454.1 M648.103 1450.39 Q653.913 1450.39 656.969 1455 Q660.048 1459.58 660.048 1468.33 Q660.048 1477.06 656.969 1481.67 Q653.913 1486.25 648.103 1486.25 Q642.293 1486.25 639.214 1481.67 Q636.159 1477.06 636.159 1468.33 Q636.159 1459.58 639.214 1455 Q642.293 1450.39 648.103 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M783.414 1481.64 L791.053 1481.64 L791.053 1455.28 L782.743 1456.95 L782.743 1452.69 L791.007 1451.02 L795.683 1451.02 L795.683 1481.64 L803.322 1481.64 L803.322 1485.58 L783.414 1485.58 L783.414 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M812.812 1451.02 L831.169 1451.02 L831.169 1454.96 L817.095 1454.96 L817.095 1463.43 Q818.113 1463.08 819.132 1462.92 Q820.15 1462.73 821.169 1462.73 Q826.956 1462.73 830.336 1465.9 Q833.715 1469.08 833.715 1474.49 Q833.715 1480.07 830.243 1483.17 Q826.771 1486.25 820.451 1486.25 Q818.275 1486.25 816.007 1485.88 Q813.762 1485.51 811.354 1484.77 L811.354 1480.07 Q813.437 1481.2 815.66 1481.76 Q817.882 1482.32 820.359 1482.32 Q824.363 1482.32 826.701 1480.21 Q829.039 1478.1 829.039 1474.49 Q829.039 1470.88 826.701 1468.77 Q824.363 1466.67 820.359 1466.67 Q818.484 1466.67 816.609 1467.08 Q814.757 1467.5 812.812 1468.38 L812.812 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M852.928 1454.1 Q849.317 1454.1 847.488 1457.66 Q845.683 1461.2 845.683 1468.33 Q845.683 1475.44 847.488 1479.01 Q849.317 1482.55 852.928 1482.55 Q856.562 1482.55 858.368 1479.01 Q860.196 1475.44 860.196 1468.33 Q860.196 1461.2 858.368 1457.66 Q856.562 1454.1 852.928 1454.1 M852.928 1450.39 Q858.738 1450.39 861.794 1455 Q864.872 1459.58 864.872 1468.33 Q864.872 1477.06 861.794 1481.67 Q858.738 1486.25 852.928 1486.25 Q847.118 1486.25 844.039 1481.67 Q840.984 1477.06 840.984 1468.33 Q840.984 1459.58 844.039 1455 Q847.118 1450.39 852.928 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M883.09 1454.1 Q879.479 1454.1 877.65 1457.66 Q875.845 1461.2 875.845 1468.33 Q875.845 1475.44 877.65 1479.01 Q879.479 1482.55 883.09 1482.55 Q886.724 1482.55 888.53 1479.01 Q890.358 1475.44 890.358 1468.33 Q890.358 1461.2 888.53 1457.66 Q886.724 1454.1 883.09 1454.1 M883.09 1450.39 Q888.9 1450.39 891.956 1455 Q895.034 1459.58 895.034 1468.33 Q895.034 1477.06 891.956 1481.67 Q888.9 1486.25 883.09 1486.25 Q877.28 1486.25 874.201 1481.67 Q871.145 1477.06 871.145 1468.33 Q871.145 1459.58 874.201 1455 Q877.28 1450.39 883.09 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1022.49 1481.64 L1038.81 1481.64 L1038.81 1485.58 L1016.86 1485.58 L1016.86 1481.64 Q1019.52 1478.89 1024.11 1474.26 Q1028.71 1469.61 1029.89 1468.27 Q1032.14 1465.74 1033.02 1464.01 Q1033.92 1462.25 1033.92 1460.56 Q1033.92 1457.8 1031.98 1456.07 Q1030.06 1454.33 1026.95 1454.33 Q1024.76 1454.33 1022.3 1455.09 Q1019.87 1455.86 1017.09 1457.41 L1017.09 1452.69 Q1019.92 1451.55 1022.37 1450.97 Q1024.82 1450.39 1026.86 1450.39 Q1032.23 1450.39 1035.43 1453.08 Q1038.62 1455.77 1038.62 1460.26 Q1038.62 1462.39 1037.81 1464.31 Q1037.02 1466.2 1034.92 1468.8 Q1034.34 1469.47 1031.24 1472.69 Q1028.13 1475.88 1022.49 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1058.62 1454.1 Q1055.01 1454.1 1053.18 1457.66 Q1051.38 1461.2 1051.38 1468.33 Q1051.38 1475.44 1053.18 1479.01 Q1055.01 1482.55 1058.62 1482.55 Q1062.26 1482.55 1064.06 1479.01 Q1065.89 1475.44 1065.89 1468.33 Q1065.89 1461.2 1064.06 1457.66 Q1062.26 1454.1 1058.62 1454.1 M1058.62 1450.39 Q1064.43 1450.39 1067.49 1455 Q1070.57 1459.58 1070.57 1468.33 Q1070.57 1477.06 1067.49 1481.67 Q1064.43 1486.25 1058.62 1486.25 Q1052.81 1486.25 1049.73 1481.67 Q1046.68 1477.06 1046.68 1468.33 Q1046.68 1459.58 1049.73 1455 Q1052.81 1450.39 1058.62 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1088.78 1454.1 Q1085.17 1454.1 1083.34 1457.66 Q1081.54 1461.2 1081.54 1468.33 Q1081.54 1475.44 1083.34 1479.01 Q1085.17 1482.55 1088.78 1482.55 Q1092.42 1482.55 1094.22 1479.01 Q1096.05 1475.44 1096.05 1468.33 Q1096.05 1461.2 1094.22 1457.66 Q1092.42 1454.1 1088.78 1454.1 M1088.78 1450.39 Q1094.59 1450.39 1097.65 1455 Q1100.73 1459.58 1100.73 1468.33 Q1100.73 1477.06 1097.65 1481.67 Q1094.59 1486.25 1088.78 1486.25 Q1082.97 1486.25 1079.89 1481.67 Q1076.84 1477.06 1076.84 1468.33 Q1076.84 1459.58 1079.89 1455 Q1082.97 1450.39 1088.78 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1118.94 1454.1 Q1115.33 1454.1 1113.5 1457.66 Q1111.7 1461.2 1111.7 1468.33 Q1111.7 1475.44 1113.5 1479.01 Q1115.33 1482.55 1118.94 1482.55 Q1122.58 1482.55 1124.38 1479.01 Q1126.21 1475.44 1126.21 1468.33 Q1126.21 1461.2 1124.38 1457.66 Q1122.58 1454.1 1118.94 1454.1 M1118.94 1450.39 Q1124.75 1450.39 1127.81 1455 Q1130.89 1459.58 1130.89 1468.33 Q1130.89 1477.06 1127.81 1481.67 Q1124.75 1486.25 1118.94 1486.25 Q1113.13 1486.25 1110.06 1481.67 Q1107 1477.06 1107 1468.33 Q1107 1459.58 1110.06 1455 Q1113.13 1450.39 1118.94 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M475.972 1525.81 L475.972 1562.76 L483.738 1562.76 Q493.573 1562.76 498.125 1558.3 Q502.708 1553.85 502.708 1544.24 Q502.708 1534.69 498.125 1530.26 Q493.573 1525.81 483.738 1525.81 L475.972 1525.81 M469.543 1520.52 L482.752 1520.52 Q496.565 1520.52 503.026 1526.28 Q509.488 1532.01 509.488 1544.24 Q509.488 1556.52 502.995 1562.28 Q496.502 1568.04 482.752 1568.04 L469.543 1568.04 L469.543 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M549.974 1548.76 L549.974 1551.62 L523.047 1551.62 Q523.429 1557.67 526.675 1560.85 Q529.953 1564 535.778 1564 Q539.152 1564 542.303 1563.17 Q545.486 1562.35 548.605 1560.69 L548.605 1566.23 Q545.454 1567.57 542.144 1568.27 Q538.834 1568.97 535.428 1568.97 Q526.898 1568.97 521.901 1564 Q516.936 1559.04 516.936 1550.57 Q516.936 1541.82 521.646 1536.69 Q526.389 1531.54 534.409 1531.54 Q541.603 1531.54 545.772 1536.18 Q549.974 1540.8 549.974 1548.76 M544.117 1547.04 Q544.053 1542.23 541.412 1539.37 Q538.802 1536.5 534.473 1536.5 Q529.571 1536.5 526.611 1539.27 Q523.683 1542.04 523.238 1547.07 L544.117 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M587.34 1539.24 Q589.536 1535.29 592.592 1533.41 Q595.648 1531.54 599.785 1531.54 Q605.355 1531.54 608.379 1535.45 Q611.403 1539.33 611.403 1546.53 L611.403 1568.04 L605.514 1568.04 L605.514 1546.72 Q605.514 1541.59 603.7 1539.11 Q601.886 1536.63 598.162 1536.63 Q593.61 1536.63 590.969 1539.65 Q588.327 1542.68 588.327 1547.9 L588.327 1568.04 L582.439 1568.04 L582.439 1546.72 Q582.439 1541.56 580.624 1539.11 Q578.81 1536.63 575.023 1536.63 Q570.535 1536.63 567.893 1539.68 Q565.251 1542.71 565.251 1547.9 L565.251 1568.04 L559.363 1568.04 L559.363 1532.4 L565.251 1532.4 L565.251 1537.93 Q567.256 1534.66 570.057 1533.1 Q572.858 1531.54 576.71 1531.54 Q580.593 1531.54 583.298 1533.51 Q586.035 1535.48 587.34 1539.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M639.284 1550.12 Q632.187 1550.12 629.449 1551.75 Q626.712 1553.37 626.712 1557.29 Q626.712 1560.4 628.749 1562.25 Q630.818 1564.07 634.351 1564.07 Q639.221 1564.07 642.149 1560.63 Q645.109 1557.16 645.109 1551.43 L645.109 1550.12 L639.284 1550.12 M650.965 1547.71 L650.965 1568.04 L645.109 1568.04 L645.109 1562.63 Q643.104 1565.88 640.112 1567.44 Q637.12 1568.97 632.791 1568.97 Q627.317 1568.97 624.07 1565.91 Q620.856 1562.82 620.856 1557.67 Q620.856 1551.65 624.866 1548.6 Q628.908 1545.54 636.897 1545.54 L645.109 1545.54 L645.109 1544.97 Q645.109 1540.93 642.435 1538.73 Q639.794 1536.5 634.988 1536.5 Q631.932 1536.5 629.036 1537.23 Q626.139 1537.97 623.466 1539.43 L623.466 1534.02 Q626.68 1532.78 629.704 1532.17 Q632.728 1531.54 635.592 1531.54 Q643.327 1531.54 647.146 1535.55 Q650.965 1539.56 650.965 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M692.661 1546.53 L692.661 1568.04 L686.804 1568.04 L686.804 1546.72 Q686.804 1541.66 684.831 1539.14 Q682.858 1536.63 678.911 1536.63 Q674.169 1536.63 671.431 1539.65 Q668.694 1542.68 668.694 1547.9 L668.694 1568.04 L662.806 1568.04 L662.806 1532.4 L668.694 1532.4 L668.694 1537.93 Q670.795 1534.72 673.627 1533.13 Q676.492 1531.54 680.216 1531.54 Q686.359 1531.54 689.51 1535.36 Q692.661 1539.14 692.661 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M727.8 1537.81 L727.8 1518.52 L733.656 1518.52 L733.656 1568.04 L727.8 1568.04 L727.8 1562.7 Q725.954 1565.88 723.121 1567.44 Q720.32 1568.97 716.373 1568.97 Q709.912 1568.97 705.838 1563.81 Q701.796 1558.65 701.796 1550.25 Q701.796 1541.85 705.838 1536.69 Q709.912 1531.54 716.373 1531.54 Q720.32 1531.54 723.121 1533.1 Q725.954 1534.62 727.8 1537.81 M707.843 1550.25 Q707.843 1556.71 710.485 1560.4 Q713.158 1564.07 717.805 1564.07 Q722.452 1564.07 725.126 1560.4 Q727.8 1556.71 727.8 1550.25 Q727.8 1543.79 725.126 1540.13 Q722.452 1536.44 717.805 1536.44 Q713.158 1536.44 710.485 1540.13 Q707.843 1543.79 707.843 1550.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M765.898 1518.52 L779.394 1518.52 L779.394 1523.07 L771.755 1523.07 L771.755 1572.09 L779.394 1572.09 L779.394 1576.64 L765.898 1576.64 L765.898 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M792.125 1520.52 L801.705 1520.52 L813.832 1552.86 L826.022 1520.52 L835.603 1520.52 L835.603 1568.04 L829.333 1568.04 L829.333 1526.32 L817.079 1558.91 L810.617 1558.91 L798.363 1526.32 L798.363 1568.04 L792.125 1568.04 L792.125 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M844.133 1520.52 L850.626 1520.52 L860.62 1560.69 L870.582 1520.52 L877.807 1520.52 L887.802 1560.69 L897.764 1520.52 L904.289 1520.52 L892.353 1568.04 L884.269 1568.04 L874.243 1526.79 L864.121 1568.04 L856.037 1568.04 L844.133 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M926.25 1518.52 L926.25 1576.64 L912.755 1576.64 L912.755 1572.09 L920.362 1572.09 L920.362 1523.07 L912.755 1523.07 L912.755 1518.52 L926.25 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="249.542,1386.4 1146.25,1386.4 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="249.542,1079.86 1146.25,1079.86 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="249.542,773.326 1146.25,773.326 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="249.542,466.791 1146.25,466.791 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="249.542,160.256 1146.25,160.256 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="249.542,1423.18 249.542,123.472 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="249.542,1386.4 268.44,1386.4 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="249.542,1079.86 268.44,1079.86 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="249.542,773.326 268.44,773.326 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="249.542,466.791 268.44,466.791 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="249.542,160.256 268.44,160.256 "/>
+<path clip-path="url(#clip130)" d="M126.205 1372.19 Q122.593 1372.19 120.765 1375.76 Q118.959 1379.3 118.959 1386.43 Q118.959 1393.54 120.765 1397.1 Q122.593 1400.64 126.205 1400.64 Q129.839 1400.64 131.644 1397.1 Q133.473 1393.54 133.473 1386.43 Q133.473 1379.3 131.644 1375.76 Q129.839 1372.19 126.205 1372.19 M126.205 1368.49 Q132.015 1368.49 135.07 1373.1 Q138.149 1377.68 138.149 1386.43 Q138.149 1395.16 135.07 1399.76 Q132.015 1404.35 126.205 1404.35 Q120.394 1404.35 117.316 1399.76 Q114.26 1395.16 114.26 1386.43 Q114.26 1377.68 117.316 1373.1 Q120.394 1368.49 126.205 1368.49 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M146.366 1397.8 L151.251 1397.8 L151.251 1403.68 L146.366 1403.68 L146.366 1397.8 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M171.436 1372.19 Q167.825 1372.19 165.996 1375.76 Q164.19 1379.3 164.19 1386.43 Q164.19 1393.54 165.996 1397.1 Q167.825 1400.64 171.436 1400.64 Q175.07 1400.64 176.876 1397.1 Q178.704 1393.54 178.704 1386.43 Q178.704 1379.3 176.876 1375.76 Q175.07 1372.19 171.436 1372.19 M171.436 1368.49 Q177.246 1368.49 180.301 1373.1 Q183.38 1377.68 183.38 1386.43 Q183.38 1395.16 180.301 1399.76 Q177.246 1404.35 171.436 1404.35 Q165.626 1404.35 162.547 1399.76 Q159.491 1395.16 159.491 1386.43 Q159.491 1377.68 162.547 1373.1 Q165.626 1368.49 171.436 1368.49 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M201.598 1372.19 Q197.987 1372.19 196.158 1375.76 Q194.352 1379.3 194.352 1386.43 Q194.352 1393.54 196.158 1397.1 Q197.987 1400.64 201.598 1400.64 Q205.232 1400.64 207.037 1397.1 Q208.866 1393.54 208.866 1386.43 Q208.866 1379.3 207.037 1375.76 Q205.232 1372.19 201.598 1372.19 M201.598 1368.49 Q207.408 1368.49 210.463 1373.1 Q213.542 1377.68 213.542 1386.43 Q213.542 1395.16 210.463 1399.76 Q207.408 1404.35 201.598 1404.35 Q195.787 1404.35 192.709 1399.76 Q189.653 1395.16 189.653 1386.43 Q189.653 1377.68 192.709 1373.1 Q195.787 1368.49 201.598 1368.49 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M127.2 1065.66 Q123.589 1065.66 121.76 1069.22 Q119.955 1072.77 119.955 1079.9 Q119.955 1087 121.76 1090.57 Q123.589 1094.11 127.2 1094.11 Q130.834 1094.11 132.64 1090.57 Q134.468 1087 134.468 1079.9 Q134.468 1072.77 132.64 1069.22 Q130.834 1065.66 127.2 1065.66 M127.2 1061.96 Q133.01 1061.96 136.066 1066.56 Q139.144 1071.15 139.144 1079.9 Q139.144 1088.62 136.066 1093.23 Q133.01 1097.81 127.2 1097.81 Q121.39 1097.81 118.311 1093.23 Q115.256 1088.62 115.256 1079.9 Q115.256 1071.15 118.311 1066.56 Q121.39 1061.96 127.2 1061.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M147.362 1091.26 L152.246 1091.26 L152.246 1097.14 L147.362 1097.14 L147.362 1091.26 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M166.459 1093.21 L182.778 1093.21 L182.778 1097.14 L160.834 1097.14 L160.834 1093.21 Q163.496 1090.45 168.079 1085.82 Q172.686 1081.17 173.866 1079.83 Q176.112 1077.3 176.991 1075.57 Q177.894 1073.81 177.894 1072.12 Q177.894 1069.36 175.95 1067.63 Q174.028 1065.89 170.927 1065.89 Q168.727 1065.89 166.274 1066.65 Q163.843 1067.42 161.065 1068.97 L161.065 1064.25 Q163.89 1063.11 166.343 1062.53 Q168.797 1061.96 170.834 1061.96 Q176.204 1061.96 179.399 1064.64 Q182.593 1067.33 182.593 1071.82 Q182.593 1073.95 181.783 1075.87 Q180.996 1077.77 178.889 1080.36 Q178.311 1081.03 175.209 1084.25 Q172.107 1087.44 166.459 1093.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M192.639 1062.58 L210.996 1062.58 L210.996 1066.52 L196.922 1066.52 L196.922 1074.99 Q197.94 1074.64 198.959 1074.48 Q199.977 1074.29 200.996 1074.29 Q206.783 1074.29 210.162 1077.47 Q213.542 1080.64 213.542 1086.05 Q213.542 1091.63 210.07 1094.73 Q206.598 1097.81 200.278 1097.81 Q198.102 1097.81 195.834 1097.44 Q193.588 1097.07 191.181 1096.33 L191.181 1091.63 Q193.264 1092.77 195.487 1093.32 Q197.709 1093.88 200.186 1093.88 Q204.19 1093.88 206.528 1091.77 Q208.866 1089.66 208.866 1086.05 Q208.866 1082.44 206.528 1080.34 Q204.19 1078.23 200.186 1078.23 Q198.311 1078.23 196.436 1078.65 Q194.584 1079.06 192.639 1079.94 L192.639 1062.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M126.205 759.125 Q122.593 759.125 120.765 762.689 Q118.959 766.231 118.959 773.361 Q118.959 780.467 120.765 784.032 Q122.593 787.574 126.205 787.574 Q129.839 787.574 131.644 784.032 Q133.473 780.467 133.473 773.361 Q133.473 766.231 131.644 762.689 Q129.839 759.125 126.205 759.125 M126.205 755.421 Q132.015 755.421 135.07 760.027 Q138.149 764.611 138.149 773.361 Q138.149 782.088 135.07 786.694 Q132.015 791.277 126.205 791.277 Q120.394 791.277 117.316 786.694 Q114.26 782.088 114.26 773.361 Q114.26 764.611 117.316 760.027 Q120.394 755.421 126.205 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M146.366 784.726 L151.251 784.726 L151.251 790.606 L146.366 790.606 L146.366 784.726 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M161.482 756.046 L179.839 756.046 L179.839 759.981 L165.765 759.981 L165.765 768.453 Q166.783 768.106 167.802 767.944 Q168.82 767.759 169.839 767.759 Q175.626 767.759 179.005 770.93 Q182.385 774.101 182.385 779.518 Q182.385 785.097 178.913 788.199 Q175.44 791.277 169.121 791.277 Q166.945 791.277 164.677 790.907 Q162.431 790.537 160.024 789.796 L160.024 785.097 Q162.107 786.231 164.329 786.787 Q166.552 787.342 169.028 787.342 Q173.033 787.342 175.371 785.236 Q177.709 783.129 177.709 779.518 Q177.709 775.907 175.371 773.801 Q173.033 771.694 169.028 771.694 Q167.153 771.694 165.278 772.111 Q163.427 772.527 161.482 773.407 L161.482 756.046 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M201.598 759.125 Q197.987 759.125 196.158 762.689 Q194.352 766.231 194.352 773.361 Q194.352 780.467 196.158 784.032 Q197.987 787.574 201.598 787.574 Q205.232 787.574 207.037 784.032 Q208.866 780.467 208.866 773.361 Q208.866 766.231 207.037 762.689 Q205.232 759.125 201.598 759.125 M201.598 755.421 Q207.408 755.421 210.463 760.027 Q213.542 764.611 213.542 773.361 Q213.542 782.088 210.463 786.694 Q207.408 791.277 201.598 791.277 Q195.787 791.277 192.709 786.694 Q189.653 782.088 189.653 773.361 Q189.653 764.611 192.709 760.027 Q195.787 755.421 201.598 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M127.2 452.59 Q123.589 452.59 121.76 456.155 Q119.955 459.696 119.955 466.826 Q119.955 473.932 121.76 477.497 Q123.589 481.039 127.2 481.039 Q130.834 481.039 132.64 477.497 Q134.468 473.932 134.468 466.826 Q134.468 459.696 132.64 456.155 Q130.834 452.59 127.2 452.59 M127.2 448.886 Q133.01 448.886 136.066 453.493 Q139.144 458.076 139.144 466.826 Q139.144 475.553 136.066 480.159 Q133.01 484.742 127.2 484.742 Q121.39 484.742 118.311 480.159 Q115.256 475.553 115.256 466.826 Q115.256 458.076 118.311 453.493 Q121.39 448.886 127.2 448.886 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M147.362 478.192 L152.246 478.192 L152.246 484.071 L147.362 484.071 L147.362 478.192 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M161.251 449.511 L183.473 449.511 L183.473 451.502 L170.927 484.071 L166.042 484.071 L177.848 453.446 L161.251 453.446 L161.251 449.511 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M192.639 449.511 L210.996 449.511 L210.996 453.446 L196.922 453.446 L196.922 461.918 Q197.94 461.571 198.959 461.409 Q199.977 461.224 200.996 461.224 Q206.783 461.224 210.162 464.395 Q213.542 467.567 213.542 472.983 Q213.542 478.562 210.07 481.664 Q206.598 484.742 200.278 484.742 Q198.102 484.742 195.834 484.372 Q193.588 484.002 191.181 483.261 L191.181 478.562 Q193.264 479.696 195.487 480.252 Q197.709 480.807 200.186 480.807 Q204.19 480.807 206.528 478.701 Q208.866 476.594 208.866 472.983 Q208.866 469.372 206.528 467.266 Q204.19 465.159 200.186 465.159 Q198.311 465.159 196.436 465.576 Q194.584 465.993 192.639 466.872 L192.639 449.511 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M117.015 173.601 L124.654 173.601 L124.654 147.236 L116.343 148.902 L116.343 144.643 L124.607 142.976 L129.283 142.976 L129.283 173.601 L136.922 173.601 L136.922 177.536 L117.015 177.536 L117.015 173.601 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M146.366 171.657 L151.251 171.657 L151.251 177.536 L146.366 177.536 L146.366 171.657 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M171.436 146.055 Q167.825 146.055 165.996 149.62 Q164.19 153.161 164.19 160.291 Q164.19 167.397 165.996 170.962 Q167.825 174.504 171.436 174.504 Q175.07 174.504 176.876 170.962 Q178.704 167.397 178.704 160.291 Q178.704 153.161 176.876 149.62 Q175.07 146.055 171.436 146.055 M171.436 142.351 Q177.246 142.351 180.301 146.958 Q183.38 151.541 183.38 160.291 Q183.38 169.018 180.301 173.624 Q177.246 178.208 171.436 178.208 Q165.626 178.208 162.547 173.624 Q159.491 169.018 159.491 160.291 Q159.491 151.541 162.547 146.958 Q165.626 142.351 171.436 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M201.598 146.055 Q197.987 146.055 196.158 149.62 Q194.352 153.161 194.352 160.291 Q194.352 167.397 196.158 170.962 Q197.987 174.504 201.598 174.504 Q205.232 174.504 207.037 170.962 Q208.866 167.397 208.866 160.291 Q208.866 153.161 207.037 149.62 Q205.232 146.055 201.598 146.055 M201.598 142.351 Q207.408 142.351 210.463 146.958 Q213.542 151.541 213.542 160.291 Q213.542 169.018 210.463 173.624 Q207.408 178.208 201.598 178.208 Q195.787 178.208 192.709 173.624 Q189.653 169.018 189.653 160.291 Q189.653 151.541 192.709 146.958 Q195.787 142.351 201.598 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M20.1444 1201.1 L26.9239 1201.1 Q23.9002 1204.35 22.4043 1208.04 Q20.9083 1211.7 20.9083 1215.84 Q20.9083 1223.99 25.9054 1228.31 Q30.8707 1232.64 40.2919 1232.64 Q49.6813 1232.64 54.6784 1228.31 Q59.6436 1223.99 59.6436 1215.84 Q59.6436 1211.7 58.1477 1208.04 Q56.6518 1204.35 53.6281 1201.1 L60.3439 1201.1 Q62.6355 1204.48 63.7814 1208.26 Q64.9272 1212.02 64.9272 1216.22 Q64.9272 1227.01 58.3387 1233.22 Q51.7183 1239.42 40.2919 1239.42 Q28.8336 1239.42 22.2451 1233.22 Q15.6248 1227.01 15.6248 1216.22 Q15.6248 1211.95 16.7706 1208.2 Q17.8846 1204.41 20.1444 1201.1 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M32.4621 1177.61 Q32.4621 1182.32 36.1542 1185.06 Q39.8145 1187.8 46.212 1187.8 Q52.6095 1187.8 56.3017 1185.09 Q59.9619 1182.35 59.9619 1177.61 Q59.9619 1172.93 56.2698 1170.2 Q52.5777 1167.46 46.212 1167.46 Q39.8781 1167.46 36.186 1170.2 Q32.4621 1172.93 32.4621 1177.61 M27.4968 1177.61 Q27.4968 1169.97 32.4621 1165.61 Q37.4273 1161.25 46.212 1161.25 Q54.9649 1161.25 59.9619 1165.61 Q64.9272 1169.97 64.9272 1177.61 Q64.9272 1185.28 59.9619 1189.64 Q54.9649 1193.97 46.212 1193.97 Q37.4273 1193.97 32.4621 1189.64 Q27.4968 1185.28 27.4968 1177.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M35.1993 1123.79 Q31.2526 1121.59 29.3747 1118.54 Q27.4968 1115.48 27.4968 1111.35 Q27.4968 1105.78 31.4117 1102.75 Q35.2948 1099.73 42.4881 1099.73 L64.0042 1099.73 L64.0042 1105.62 L42.679 1105.62 Q37.5546 1105.62 35.072 1107.43 Q32.5894 1109.24 32.5894 1112.97 Q32.5894 1117.52 35.6131 1120.16 Q38.6368 1122.8 43.8567 1122.8 L64.0042 1122.8 L64.0042 1128.69 L42.679 1128.69 Q37.5228 1128.69 35.072 1130.51 Q32.5894 1132.32 32.5894 1136.11 Q32.5894 1140.6 35.6449 1143.24 Q38.6686 1145.88 43.8567 1145.88 L64.0042 1145.88 L64.0042 1151.77 L28.3562 1151.77 L28.3562 1145.88 L33.8944 1145.88 Q30.616 1143.87 29.0564 1141.07 Q27.4968 1138.27 27.4968 1134.42 Q27.4968 1130.54 29.4702 1127.83 Q31.4436 1125.09 35.1993 1123.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M35.1993 1060.29 Q31.2526 1058.1 29.3747 1055.04 Q27.4968 1051.98 27.4968 1047.85 Q27.4968 1042.28 31.4117 1039.25 Q35.2948 1036.23 42.4881 1036.23 L64.0042 1036.23 L64.0042 1042.12 L42.679 1042.12 Q37.5546 1042.12 35.072 1043.93 Q32.5894 1045.75 32.5894 1049.47 Q32.5894 1054.02 35.6131 1056.66 Q38.6368 1059.31 43.8567 1059.31 L64.0042 1059.31 L64.0042 1065.19 L42.679 1065.19 Q37.5228 1065.19 35.072 1067.01 Q32.5894 1068.82 32.5894 1072.61 Q32.5894 1077.1 35.6449 1079.74 Q38.6686 1082.38 43.8567 1082.38 L64.0042 1082.38 L64.0042 1088.27 L28.3562 1088.27 L28.3562 1082.38 L33.8944 1082.38 Q30.616 1080.38 29.0564 1077.57 Q27.4968 1074.77 27.4968 1070.92 Q27.4968 1067.04 29.4702 1064.33 Q31.4436 1061.6 35.1993 1060.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M28.3562 1024.55 L28.3562 1018.69 L64.0042 1018.69 L64.0042 1024.55 L28.3562 1024.55 M14.479 1024.55 L14.479 1018.69 L21.895 1018.69 L21.895 1024.55 L14.479 1024.55 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M18.2347 1000.65 L28.3562 1000.65 L28.3562 988.582 L32.9077 988.582 L32.9077 1000.65 L52.2594 1000.65 Q56.6199 1000.65 57.8613 999.468 Q59.1026 998.258 59.1026 994.598 L59.1026 988.582 L64.0042 988.582 L64.0042 994.598 Q64.0042 1001.38 61.4897 1003.96 Q58.9434 1006.53 52.2594 1006.53 L32.9077 1006.53 L32.9077 1010.83 L28.3562 1010.83 L28.3562 1006.53 L18.2347 1006.53 L18.2347 1000.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M35.1993 953.125 Q31.2526 950.929 29.3747 947.874 Q27.4968 944.818 27.4968 940.68 Q27.4968 935.11 31.4117 932.087 Q35.2948 929.063 42.4881 929.063 L64.0042 929.063 L64.0042 934.951 L42.679 934.951 Q37.5546 934.951 35.072 936.766 Q32.5894 938.58 32.5894 942.304 Q32.5894 946.855 35.6131 949.497 Q38.6368 952.139 43.8567 952.139 L64.0042 952.139 L64.0042 958.027 L42.679 958.027 Q37.5228 958.027 35.072 959.841 Q32.5894 961.655 32.5894 965.443 Q32.5894 969.931 35.6449 972.573 Q38.6686 975.214 43.8567 975.214 L64.0042 975.214 L64.0042 981.103 L28.3562 981.103 L28.3562 975.214 L33.8944 975.214 Q30.616 973.209 29.0564 970.408 Q27.4968 967.607 27.4968 963.756 Q27.4968 959.873 29.4702 957.168 Q31.4436 954.43 35.1993 953.125 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M44.7161 886.89 L47.5806 886.89 L47.5806 913.817 Q53.6281 913.435 56.8109 910.189 Q59.9619 906.91 59.9619 901.086 Q59.9619 897.712 59.1344 894.561 Q58.3069 891.378 56.6518 888.259 L62.1899 888.259 Q63.5267 891.41 64.227 894.72 Q64.9272 898.03 64.9272 901.436 Q64.9272 909.966 59.9619 914.963 Q54.9967 919.928 46.5303 919.928 Q37.7774 919.928 32.6531 915.218 Q27.4968 910.475 27.4968 902.454 Q27.4968 895.261 32.1438 891.092 Q36.7589 886.89 44.7161 886.89 M42.9973 892.747 Q38.1912 892.81 35.3266 895.452 Q32.4621 898.062 32.4621 902.391 Q32.4621 907.292 35.2312 910.252 Q38.0002 913.181 43.0292 913.626 L42.9973 892.747 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M42.4881 847.646 L64.0042 847.646 L64.0042 853.502 L42.679 853.502 Q37.6183 853.502 35.1038 855.475 Q32.5894 857.449 32.5894 861.396 Q32.5894 866.138 35.6131 868.875 Q38.6368 871.612 43.8567 871.612 L64.0042 871.612 L64.0042 877.501 L28.3562 877.501 L28.3562 871.612 L33.8944 871.612 Q30.6797 869.512 29.0883 866.679 Q27.4968 863.815 27.4968 860.091 Q27.4968 853.948 31.3163 850.797 Q35.1038 847.646 42.4881 847.646 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M18.2347 830.172 L28.3562 830.172 L28.3562 818.109 L32.9077 818.109 L32.9077 830.172 L52.2594 830.172 Q56.6199 830.172 57.8613 828.994 Q59.1026 827.785 59.1026 824.124 L59.1026 818.109 L64.0042 818.109 L64.0042 824.124 Q64.0042 830.904 61.4897 833.482 Q58.9434 836.06 52.2594 836.06 L32.9077 836.06 L32.9077 840.357 L28.3562 840.357 L28.3562 836.06 L18.2347 836.06 L18.2347 830.172 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M33.7671 766.228 L14.479 766.228 L14.479 760.372 L64.0042 760.372 L64.0042 766.228 L58.657 766.228 Q61.8398 768.074 63.3994 770.907 Q64.9272 773.708 64.9272 777.655 Q64.9272 784.116 59.771 788.19 Q54.6147 792.232 46.212 792.232 Q37.8093 792.232 32.6531 788.19 Q27.4968 784.116 27.4968 777.655 Q27.4968 773.708 29.0564 770.907 Q30.5842 768.074 33.7671 766.228 M46.212 786.185 Q52.6732 786.185 56.3653 783.543 Q60.0256 780.869 60.0256 776.222 Q60.0256 771.575 56.3653 768.902 Q52.6732 766.228 46.212 766.228 Q39.7508 766.228 36.0905 768.902 Q32.3984 771.575 32.3984 776.222 Q32.3984 780.869 36.0905 783.543 Q39.7508 786.185 46.212 786.185 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M44.7161 717.817 L47.5806 717.817 L47.5806 744.744 Q53.6281 744.362 56.8109 741.116 Q59.9619 737.837 59.9619 732.013 Q59.9619 728.639 59.1344 725.488 Q58.3069 722.305 56.6518 719.186 L62.1899 719.186 Q63.5267 722.337 64.227 725.647 Q64.9272 728.957 64.9272 732.363 Q64.9272 740.893 59.9619 745.89 Q54.9967 750.855 46.5303 750.855 Q37.7774 750.855 32.6531 746.144 Q27.4968 741.402 27.4968 733.381 Q27.4968 726.188 32.1438 722.018 Q36.7589 717.817 44.7161 717.817 M42.9973 723.673 Q38.1912 723.737 35.3266 726.379 Q32.4621 728.989 32.4621 733.318 Q32.4621 738.219 35.2312 741.179 Q38.0002 744.107 43.0292 744.553 L42.9973 723.673 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M29.7248 682.551 L35.1993 682.551 Q33.8307 685.034 33.1623 687.548 Q32.4621 690.031 32.4621 692.577 Q32.4621 698.274 36.0905 701.425 Q39.6872 704.576 46.212 704.576 Q52.7369 704.576 56.3653 701.425 Q59.9619 698.274 59.9619 692.577 Q59.9619 690.031 59.2935 687.548 Q58.5933 685.034 57.2247 682.551 L62.6355 682.551 Q63.7814 685.002 64.3543 687.644 Q64.9272 690.254 64.9272 693.214 Q64.9272 701.266 59.8664 706.009 Q54.8057 710.751 46.212 710.751 Q37.491 710.751 32.4939 705.977 Q27.4968 701.171 27.4968 692.832 Q27.4968 690.126 28.0697 687.548 Q28.6108 684.97 29.7248 682.551 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M28.3562 672.366 L28.3562 666.509 L64.0042 666.509 L64.0042 672.366 L28.3562 672.366 M14.479 672.366 L14.479 666.509 L21.895 666.509 L21.895 672.366 L14.479 672.366 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M29.4065 631.53 L34.9447 631.53 Q33.6716 634.012 33.035 636.686 Q32.3984 639.36 32.3984 642.224 Q32.3984 646.585 33.7352 648.781 Q35.072 650.945 37.7456 650.945 Q39.7826 650.945 40.9603 649.386 Q42.1061 647.826 43.1565 643.115 L43.6021 641.11 Q44.9389 634.872 47.3897 632.262 Q49.8086 629.62 54.1691 629.62 Q59.1344 629.62 62.0308 633.567 Q64.9272 637.482 64.9272 644.357 Q64.9272 647.221 64.3543 650.341 Q63.8132 653.428 62.6992 656.865 L56.6518 656.865 Q58.3387 653.619 59.198 650.468 Q60.0256 647.317 60.0256 644.229 Q60.0256 640.092 58.6251 637.864 Q57.1929 635.636 54.6147 635.636 Q52.2276 635.636 50.9545 637.259 Q49.6813 638.85 48.5037 644.293 L48.0262 646.33 Q46.8804 651.773 44.5251 654.192 Q42.138 656.611 38.0002 656.611 Q32.9713 656.611 30.2341 653.046 Q27.4968 649.481 27.4968 642.924 Q27.4968 639.678 27.9743 636.813 Q28.4517 633.949 29.4065 631.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M28.3562 620.294 L28.3562 614.438 L64.0042 614.438 L64.0042 620.294 L28.3562 620.294 M14.479 620.294 L14.479 614.438 L21.895 614.438 L21.895 620.294 L14.479 620.294 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M32.4621 588.37 Q32.4621 593.081 36.1542 595.818 Q39.8145 598.556 46.212 598.556 Q52.6095 598.556 56.3017 595.85 Q59.9619 593.113 59.9619 588.37 Q59.9619 583.692 56.2698 580.954 Q52.5777 578.217 46.212 578.217 Q39.8781 578.217 36.186 580.954 Q32.4621 583.692 32.4621 588.37 M27.4968 588.37 Q27.4968 580.732 32.4621 576.371 Q37.4273 572.011 46.212 572.011 Q54.9649 572.011 59.9619 576.371 Q64.9272 580.732 64.9272 588.37 Q64.9272 596.041 59.9619 600.402 Q54.9649 604.73 46.212 604.73 Q37.4273 604.73 32.4621 600.402 Q27.4968 596.041 27.4968 588.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M42.4881 532.67 L64.0042 532.67 L64.0042 538.527 L42.679 538.527 Q37.6183 538.527 35.1038 540.5 Q32.5894 542.474 32.5894 546.42 Q32.5894 551.163 35.6131 553.9 Q38.6368 556.637 43.8567 556.637 L64.0042 556.637 L64.0042 562.526 L28.3562 562.526 L28.3562 556.637 L33.8944 556.637 Q30.6797 554.537 29.0883 551.704 Q27.4968 548.839 27.4968 545.115 Q27.4968 538.972 31.3163 535.821 Q35.1038 532.67 42.4881 532.67 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M70.0516 473.087 L74.6349 473.087 L74.6349 475.061 Q74.6349 482.986 72.2796 485.692 Q69.9243 488.365 62.8902 488.365 L55.2831 488.365 Q50.477 488.365 48.631 490.084 Q46.7849 491.803 46.7849 496.322 L46.7849 498.264 L42.2334 498.264 L42.2334 496.322 Q42.2334 491.771 40.4192 490.084 Q38.5732 488.365 33.8307 488.365 L26.1919 488.365 Q19.1578 488.365 16.8343 485.692 Q14.479 482.986 14.479 475.061 L14.479 473.087 L19.0304 473.087 L19.0304 475.252 Q19.0304 479.74 20.4309 481.108 Q21.8314 482.477 26.3192 482.477 L34.2127 482.477 Q39.2097 482.477 41.4696 483.941 Q43.7294 485.373 44.5251 488.874 Q45.3845 485.341 47.6443 483.909 Q49.9041 482.477 54.8694 482.477 L62.7628 482.477 Q67.2507 482.477 68.6511 481.108 Q70.0516 479.74 70.0516 475.252 L70.0516 473.087 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M20.7174 444.219 Q20.7174 449.184 25.6189 451.699 Q30.4887 454.181 40.2919 454.181 Q50.0633 454.181 54.9649 451.699 Q59.8346 449.184 59.8346 444.219 Q59.8346 439.222 54.9649 436.739 Q50.0633 434.225 40.2919 434.225 Q30.4887 434.225 25.6189 436.739 Q20.7174 439.222 20.7174 444.219 M15.6248 444.219 Q15.6248 436.23 21.9587 432.029 Q28.2607 427.795 40.2919 427.795 Q52.2913 427.795 58.6251 432.029 Q64.9272 436.23 64.9272 444.219 Q64.9272 452.208 58.6251 456.441 Q52.2913 460.642 40.2919 460.642 Q28.2607 460.642 21.9587 456.441 Q15.6248 452.208 15.6248 444.219 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M55.9197 415.828 L55.9197 409.112 L61.3942 409.112 L71.5793 414.332 L71.5793 418.438 L61.3942 415.828 L55.9197 415.828 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M58.5933 373.942 L58.5933 363.438 L22.3406 363.438 L24.6323 374.865 L18.7758 374.865 L16.4842 363.502 L16.4842 357.072 L58.5933 357.072 L58.5933 346.569 L64.0042 346.569 L64.0042 373.942 L58.5933 373.942 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M70.0516 332.405 L70.0516 330.177 Q70.0516 325.721 68.683 324.385 Q67.3143 323.016 62.7628 323.016 L54.8694 323.016 Q49.9041 323.016 47.6443 321.584 Q45.3845 320.151 44.5251 316.618 Q43.7294 320.151 41.4696 321.584 Q39.2097 323.016 34.2127 323.016 L26.3192 323.016 Q21.7995 323.016 20.4309 324.385 Q19.0304 325.721 19.0304 330.177 L19.0304 332.405 L14.479 332.405 L14.479 330.4 Q14.479 322.475 16.8343 319.833 Q19.1578 317.159 26.1919 317.159 L33.8307 317.159 Q38.5732 317.159 40.4192 315.441 Q42.2334 313.722 42.2334 309.202 L42.2334 307.229 L46.7849 307.229 L46.7849 309.202 Q46.7849 313.722 48.631 315.441 Q50.477 317.159 55.2831 317.159 L62.8902 317.159 Q69.9243 317.159 72.2796 319.833 Q74.6349 322.475 74.6349 330.4 L74.6349 332.405 L70.0516 332.405 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M478.56 16.7545 L478.56 25.383 Q474.429 21.5346 469.729 19.6307 Q465.071 17.7268 459.805 17.7268 Q449.434 17.7268 443.925 24.0867 Q438.416 30.4061 438.416 42.3968 Q438.416 54.3469 443.925 60.7069 Q449.434 67.0263 459.805 67.0263 Q465.071 67.0263 469.729 65.1223 Q474.429 63.2184 478.56 59.3701 L478.56 67.9175 Q474.267 70.8341 469.446 72.2924 Q464.666 73.7508 459.319 73.7508 Q445.586 73.7508 437.687 65.3654 Q429.788 56.9395 429.788 42.3968 Q429.788 27.8135 437.687 19.4281 Q445.586 11.0023 459.319 11.0023 Q464.747 11.0023 469.527 12.4606 Q474.348 13.8784 478.56 16.7545 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M508.456 32.4315 Q502.461 32.4315 498.977 37.1306 Q495.493 41.7891 495.493 49.9314 Q495.493 58.0738 498.937 62.7728 Q502.42 67.4314 508.456 67.4314 Q514.411 67.4314 517.895 62.7323 Q521.379 58.0333 521.379 49.9314 Q521.379 41.8701 517.895 37.1711 Q514.411 32.4315 508.456 32.4315 M508.456 26.1121 Q518.178 26.1121 523.728 32.4315 Q529.278 38.7509 529.278 49.9314 Q529.278 61.0714 523.728 67.4314 Q518.178 73.7508 508.456 73.7508 Q498.693 73.7508 493.144 67.4314 Q487.634 61.0714 487.634 49.9314 Q487.634 38.7509 493.144 32.4315 Q498.693 26.1121 508.456 26.1121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M576.957 35.9153 Q579.752 30.8922 583.641 28.5022 Q587.53 26.1121 592.796 26.1121 Q599.885 26.1121 603.733 31.0947 Q607.582 36.0368 607.582 45.1919 L607.582 72.576 L600.088 72.576 L600.088 45.4349 Q600.088 38.913 597.779 35.7533 Q595.47 32.5936 590.73 32.5936 Q584.937 32.5936 581.575 36.4419 Q578.213 40.2903 578.213 46.9338 L578.213 72.576 L570.719 72.576 L570.719 45.4349 Q570.719 38.8725 568.41 35.7533 Q566.101 32.5936 561.28 32.5936 Q555.568 32.5936 552.206 36.4824 Q548.844 40.3308 548.844 46.9338 L548.844 72.576 L541.349 72.576 L541.349 27.2059 L548.844 27.2059 L548.844 34.2544 Q551.396 30.082 554.961 28.0971 Q558.525 26.1121 563.427 26.1121 Q568.369 26.1121 571.812 28.6237 Q575.296 31.1352 576.957 35.9153 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M657.772 35.9153 Q660.568 30.8922 664.456 28.5022 Q668.345 26.1121 673.612 26.1121 Q680.701 26.1121 684.549 31.0947 Q688.397 36.0368 688.397 45.1919 L688.397 72.576 L680.903 72.576 L680.903 45.4349 Q680.903 38.913 678.594 35.7533 Q676.285 32.5936 671.546 32.5936 Q665.753 32.5936 662.39 36.4419 Q659.028 40.2903 659.028 46.9338 L659.028 72.576 L651.534 72.576 L651.534 45.4349 Q651.534 38.8725 649.225 35.7533 Q646.916 32.5936 642.095 32.5936 Q636.384 32.5936 633.021 36.4824 Q629.659 40.3308 629.659 46.9338 L629.659 72.576 L622.165 72.576 L622.165 27.2059 L629.659 27.2059 L629.659 34.2544 Q632.211 30.082 635.776 28.0971 Q639.341 26.1121 644.242 26.1121 Q649.185 26.1121 652.628 28.6237 Q656.112 31.1352 657.772 35.9153 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M703.264 27.2059 L710.718 27.2059 L710.718 72.576 L703.264 72.576 L703.264 27.2059 M703.264 9.54393 L710.718 9.54393 L710.718 18.9825 L703.264 18.9825 L703.264 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M733.686 14.324 L733.686 27.2059 L749.039 27.2059 L749.039 32.9987 L733.686 32.9987 L733.686 57.6282 Q733.686 63.1779 735.185 64.7578 Q736.725 66.3376 741.383 66.3376 L749.039 66.3376 L749.039 72.576 L741.383 72.576 Q732.755 72.576 729.473 69.3758 Q726.192 66.1351 726.192 57.6282 L726.192 32.9987 L720.724 32.9987 L720.724 27.2059 L726.192 27.2059 L726.192 14.324 L733.686 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M794.166 35.9153 Q796.962 30.8922 800.85 28.5022 Q804.739 26.1121 810.005 26.1121 Q817.095 26.1121 820.943 31.0947 Q824.791 36.0368 824.791 45.1919 L824.791 72.576 L817.297 72.576 L817.297 45.4349 Q817.297 38.913 814.988 35.7533 Q812.679 32.5936 807.939 32.5936 Q802.147 32.5936 798.784 36.4419 Q795.422 40.2903 795.422 46.9338 L795.422 72.576 L787.928 72.576 L787.928 45.4349 Q787.928 38.8725 785.619 35.7533 Q783.31 32.5936 778.489 32.5936 Q772.778 32.5936 769.415 36.4824 Q766.053 40.3308 766.053 46.9338 L766.053 72.576 L758.559 72.576 L758.559 27.2059 L766.053 27.2059 L766.053 34.2544 Q768.605 30.082 772.17 28.0971 Q775.735 26.1121 780.636 26.1121 Q785.578 26.1121 789.022 28.6237 Q792.506 31.1352 794.166 35.9153 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M878.466 48.0275 L878.466 51.6733 L844.195 51.6733 Q844.681 59.3701 848.813 63.421 Q852.986 67.4314 860.399 67.4314 Q864.693 67.4314 868.703 66.3781 Q872.754 65.3249 876.724 63.2184 L876.724 70.267 Q872.713 71.9684 868.501 72.8596 Q864.288 73.7508 859.953 73.7508 Q849.097 73.7508 842.737 67.4314 Q836.417 61.1119 836.417 50.3365 Q836.417 39.1965 842.413 32.6746 Q848.449 26.1121 858.657 26.1121 Q867.812 26.1121 873.119 32.0264 Q878.466 37.9003 878.466 48.0275 M871.012 45.84 Q870.931 39.7232 867.569 36.0774 Q864.247 32.4315 858.738 32.4315 Q852.499 32.4315 848.732 35.9558 Q845.005 39.4801 844.438 45.8805 L871.012 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M928.413 45.1919 L928.413 72.576 L920.96 72.576 L920.96 45.4349 Q920.96 38.994 918.448 35.7938 Q915.937 32.5936 910.913 32.5936 Q904.878 32.5936 901.394 36.4419 Q897.91 40.2903 897.91 46.9338 L897.91 72.576 L890.416 72.576 L890.416 27.2059 L897.91 27.2059 L897.91 34.2544 Q900.584 30.163 904.189 28.1376 Q907.835 26.1121 912.574 26.1121 Q920.393 26.1121 924.403 30.9732 Q928.413 35.7938 928.413 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M950.653 14.324 L950.653 27.2059 L966.006 27.2059 L966.006 32.9987 L950.653 32.9987 L950.653 57.6282 Q950.653 63.1779 952.152 64.7578 Q953.691 66.3376 958.35 66.3376 L966.006 66.3376 L966.006 72.576 L958.35 72.576 Q949.721 72.576 946.44 69.3758 Q943.159 66.1351 943.159 57.6282 L943.159 32.9987 L937.69 32.9987 L937.69 27.2059 L943.159 27.2059 L943.159 14.324 L950.653 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip132)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="274.921,160.256 345.417,160.256 415.913,160.256 486.409,160.256 556.905,160.256 627.401,160.256 697.897,160.256 768.393,160.256 838.889,160.256 909.385,160.256 979.881,160.256 1050.38,160.256 1120.87,160.256 "/>
+<polyline clip-path="url(#clip132)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="274.921,1386.4 345.417,1386.4 415.913,1386.4 486.409,1386.4 556.905,1386.4 627.401,1386.4 697.897,1386.4 768.393,160.256 838.889,160.256 909.385,160.256 979.881,160.256 1050.38,160.256 1120.87,160.256 "/>
+<path clip-path="url(#clip130)" d="M966.829 1379.86 L1116.36 1379.86 L1116.36 1224.34 L966.829 1224.34  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="966.829,1379.86 1116.36,1379.86 1116.36,1224.34 966.829,1224.34 966.829,1379.86 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="976.793,1276.18 1036.57,1276.18 "/>
+<path clip-path="url(#clip130)" d="M1072.09 1288.53 L1072.09 1279.24 L1064.45 1279.24 L1064.45 1275.4 L1076.72 1275.4 L1076.72 1290.24 Q1074.01 1292.16 1070.75 1293.16 Q1067.49 1294.13 1063.78 1294.13 Q1055.68 1294.13 1051.1 1289.41 Q1046.54 1284.66 1046.54 1276.21 Q1046.54 1267.74 1051.1 1263.02 Q1055.68 1258.27 1063.78 1258.27 Q1067.16 1258.27 1070.19 1259.1 Q1073.25 1259.94 1075.82 1261.56 L1075.82 1266.54 Q1073.23 1264.34 1070.31 1263.22 Q1067.39 1262.11 1064.18 1262.11 Q1057.83 1262.11 1054.64 1265.66 Q1051.47 1269.2 1051.47 1276.21 Q1051.47 1283.2 1054.64 1286.74 Q1057.83 1290.28 1064.18 1290.28 Q1066.65 1290.28 1068.6 1289.87 Q1070.54 1289.43 1072.09 1288.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1086.49 1289.52 L1094.13 1289.52 L1094.13 1263.16 L1085.82 1264.82 L1085.82 1260.56 L1094.08 1258.9 L1098.76 1258.9 L1098.76 1289.52 L1106.4 1289.52 L1106.4 1293.46 L1086.49 1293.46 L1086.49 1289.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip130)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="976.793,1328.02 1036.57,1328.02 "/>
+<path clip-path="url(#clip130)" d="M1072.09 1340.37 L1072.09 1331.08 L1064.45 1331.08 L1064.45 1327.24 L1076.72 1327.24 L1076.72 1342.08 Q1074.01 1344 1070.75 1345 Q1067.49 1345.97 1063.78 1345.97 Q1055.68 1345.97 1051.1 1341.25 Q1046.54 1336.5 1046.54 1328.05 Q1046.54 1319.58 1051.1 1314.86 Q1055.68 1310.11 1063.78 1310.11 Q1067.16 1310.11 1070.19 1310.94 Q1073.25 1311.78 1075.82 1313.4 L1075.82 1318.38 Q1073.23 1316.18 1070.31 1315.06 Q1067.39 1313.95 1064.18 1313.95 Q1057.83 1313.95 1054.64 1317.5 Q1051.47 1321.04 1051.47 1328.05 Q1051.47 1335.04 1054.64 1338.58 Q1057.83 1342.12 1064.18 1342.12 Q1066.65 1342.12 1068.6 1341.71 Q1070.54 1341.27 1072.09 1340.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1089.71 1341.36 L1106.03 1341.36 L1106.03 1345.3 L1084.08 1345.3 L1084.08 1341.36 Q1086.74 1338.61 1091.33 1333.98 Q1095.93 1329.32 1097.12 1327.98 Q1099.36 1325.46 1100.24 1323.72 Q1101.14 1321.96 1101.14 1320.27 Q1101.14 1317.52 1099.2 1315.78 Q1097.28 1314.05 1094.18 1314.05 Q1091.98 1314.05 1089.52 1314.81 Q1087.09 1315.57 1084.31 1317.13 L1084.31 1312.4 Q1087.14 1311.27 1089.59 1310.69 Q1092.05 1310.11 1094.08 1310.11 Q1099.45 1310.11 1102.65 1312.8 Q1105.84 1315.48 1105.84 1319.97 Q1105.84 1322.1 1105.03 1324.02 Q1104.24 1325.92 1102.14 1328.51 Q1101.56 1329.19 1098.46 1332.4 Q1095.36 1335.6 1089.71 1341.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1456.05 1423.18 L2352.76 1423.18 L2352.76 123.472 L1456.05 123.472  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip133">
+    <rect x="1456" y="123" width="898" height="1301"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1575.42,1423.18 1575.42,123.472 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1810.41,1423.18 1810.41,123.472 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2045.39,1423.18 2045.39,123.472 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2280.38,1423.18 2280.38,123.472 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1456.05,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1575.42,1423.18 1575.42,1404.28 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1810.41,1423.18 1810.41,1404.28 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2045.39,1423.18 2045.39,1404.28 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2280.38,1423.18 2280.38,1404.28 "/>
+<path clip-path="url(#clip130)" d="M1535.04 1451.02 L1553.39 1451.02 L1553.39 1454.96 L1539.32 1454.96 L1539.32 1463.43 Q1540.34 1463.08 1541.36 1462.92 Q1542.38 1462.73 1543.39 1462.73 Q1549.18 1462.73 1552.56 1465.9 Q1555.94 1469.08 1555.94 1474.49 Q1555.94 1480.07 1552.47 1483.17 Q1549 1486.25 1542.68 1486.25 Q1540.5 1486.25 1538.23 1485.88 Q1535.99 1485.51 1533.58 1484.77 L1533.58 1480.07 Q1535.66 1481.2 1537.89 1481.76 Q1540.11 1482.32 1542.58 1482.32 Q1546.59 1482.32 1548.93 1480.21 Q1551.26 1478.1 1551.26 1474.49 Q1551.26 1470.88 1548.93 1468.77 Q1546.59 1466.67 1542.58 1466.67 Q1540.71 1466.67 1538.83 1467.08 Q1536.98 1467.5 1535.04 1468.38 L1535.04 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1575.15 1454.1 Q1571.54 1454.1 1569.71 1457.66 Q1567.91 1461.2 1567.91 1468.33 Q1567.91 1475.44 1569.71 1479.01 Q1571.54 1482.55 1575.15 1482.55 Q1578.79 1482.55 1580.59 1479.01 Q1582.42 1475.44 1582.42 1468.33 Q1582.42 1461.2 1580.59 1457.66 Q1578.79 1454.1 1575.15 1454.1 M1575.15 1450.39 Q1580.96 1450.39 1584.02 1455 Q1587.1 1459.58 1587.1 1468.33 Q1587.1 1477.06 1584.02 1481.67 Q1580.96 1486.25 1575.15 1486.25 Q1569.34 1486.25 1566.26 1481.67 Q1563.21 1477.06 1563.21 1468.33 Q1563.21 1459.58 1566.26 1455 Q1569.34 1450.39 1575.15 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1605.32 1454.1 Q1601.7 1454.1 1599.88 1457.66 Q1598.07 1461.2 1598.07 1468.33 Q1598.07 1475.44 1599.88 1479.01 Q1601.7 1482.55 1605.32 1482.55 Q1608.95 1482.55 1610.76 1479.01 Q1612.58 1475.44 1612.58 1468.33 Q1612.58 1461.2 1610.76 1457.66 Q1608.95 1454.1 1605.32 1454.1 M1605.32 1450.39 Q1611.13 1450.39 1614.18 1455 Q1617.26 1459.58 1617.26 1468.33 Q1617.26 1477.06 1614.18 1481.67 Q1611.13 1486.25 1605.32 1486.25 Q1599.51 1486.25 1596.43 1481.67 Q1593.37 1477.06 1593.37 1468.33 Q1593.37 1459.58 1596.43 1455 Q1599.51 1450.39 1605.32 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1754.93 1481.64 L1762.57 1481.64 L1762.57 1455.28 L1754.26 1456.95 L1754.26 1452.69 L1762.52 1451.02 L1767.2 1451.02 L1767.2 1481.64 L1774.84 1481.64 L1774.84 1485.58 L1754.93 1485.58 L1754.93 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1794.28 1454.1 Q1790.67 1454.1 1788.84 1457.66 Q1787.04 1461.2 1787.04 1468.33 Q1787.04 1475.44 1788.84 1479.01 Q1790.67 1482.55 1794.28 1482.55 Q1797.92 1482.55 1799.72 1479.01 Q1801.55 1475.44 1801.55 1468.33 Q1801.55 1461.2 1799.72 1457.66 Q1797.92 1454.1 1794.28 1454.1 M1794.28 1450.39 Q1800.09 1450.39 1803.15 1455 Q1806.23 1459.58 1806.23 1468.33 Q1806.23 1477.06 1803.15 1481.67 Q1800.09 1486.25 1794.28 1486.25 Q1788.47 1486.25 1785.4 1481.67 Q1782.34 1477.06 1782.34 1468.33 Q1782.34 1459.58 1785.4 1455 Q1788.47 1450.39 1794.28 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1824.45 1454.1 Q1820.83 1454.1 1819.01 1457.66 Q1817.2 1461.2 1817.2 1468.33 Q1817.2 1475.44 1819.01 1479.01 Q1820.83 1482.55 1824.45 1482.55 Q1828.08 1482.55 1829.89 1479.01 Q1831.71 1475.44 1831.71 1468.33 Q1831.71 1461.2 1829.89 1457.66 Q1828.08 1454.1 1824.45 1454.1 M1824.45 1450.39 Q1830.26 1450.39 1833.31 1455 Q1836.39 1459.58 1836.39 1468.33 Q1836.39 1477.06 1833.31 1481.67 Q1830.26 1486.25 1824.45 1486.25 Q1818.64 1486.25 1815.56 1481.67 Q1812.5 1477.06 1812.5 1468.33 Q1812.5 1459.58 1815.56 1455 Q1818.64 1450.39 1824.45 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1854.61 1454.1 Q1851 1454.1 1849.17 1457.66 Q1847.36 1461.2 1847.36 1468.33 Q1847.36 1475.44 1849.17 1479.01 Q1851 1482.55 1854.61 1482.55 Q1858.24 1482.55 1860.05 1479.01 Q1861.88 1475.44 1861.88 1468.33 Q1861.88 1461.2 1860.05 1457.66 Q1858.24 1454.1 1854.61 1454.1 M1854.61 1450.39 Q1860.42 1450.39 1863.47 1455 Q1866.55 1459.58 1866.55 1468.33 Q1866.55 1477.06 1863.47 1481.67 Q1860.42 1486.25 1854.61 1486.25 Q1848.8 1486.25 1845.72 1481.67 Q1842.66 1477.06 1842.66 1468.33 Q1842.66 1459.58 1845.72 1455 Q1848.8 1450.39 1854.61 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1989.92 1481.64 L1997.56 1481.64 L1997.56 1455.28 L1989.25 1456.95 L1989.25 1452.69 L1997.51 1451.02 L2002.19 1451.02 L2002.19 1481.64 L2009.83 1481.64 L2009.83 1485.58 L1989.92 1485.58 L1989.92 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2019.32 1451.02 L2037.67 1451.02 L2037.67 1454.96 L2023.6 1454.96 L2023.6 1463.43 Q2024.62 1463.08 2025.64 1462.92 Q2026.65 1462.73 2027.67 1462.73 Q2033.46 1462.73 2036.84 1465.9 Q2040.22 1469.08 2040.22 1474.49 Q2040.22 1480.07 2036.75 1483.17 Q2033.28 1486.25 2026.96 1486.25 Q2024.78 1486.25 2022.51 1485.88 Q2020.27 1485.51 2017.86 1484.77 L2017.86 1480.07 Q2019.94 1481.2 2022.16 1481.76 Q2024.39 1482.32 2026.86 1482.32 Q2030.87 1482.32 2033.21 1480.21 Q2035.54 1478.1 2035.54 1474.49 Q2035.54 1470.88 2033.21 1468.77 Q2030.87 1466.67 2026.86 1466.67 Q2024.99 1466.67 2023.11 1467.08 Q2021.26 1467.5 2019.32 1468.38 L2019.32 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2059.43 1454.1 Q2055.82 1454.1 2053.99 1457.66 Q2052.19 1461.2 2052.19 1468.33 Q2052.19 1475.44 2053.99 1479.01 Q2055.82 1482.55 2059.43 1482.55 Q2063.07 1482.55 2064.87 1479.01 Q2066.7 1475.44 2066.7 1468.33 Q2066.7 1461.2 2064.87 1457.66 Q2063.07 1454.1 2059.43 1454.1 M2059.43 1450.39 Q2065.24 1450.39 2068.3 1455 Q2071.38 1459.58 2071.38 1468.33 Q2071.38 1477.06 2068.3 1481.67 Q2065.24 1486.25 2059.43 1486.25 Q2053.62 1486.25 2050.54 1481.67 Q2047.49 1477.06 2047.49 1468.33 Q2047.49 1459.58 2050.54 1455 Q2053.62 1450.39 2059.43 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2089.59 1454.1 Q2085.98 1454.1 2084.15 1457.66 Q2082.35 1461.2 2082.35 1468.33 Q2082.35 1475.44 2084.15 1479.01 Q2085.98 1482.55 2089.59 1482.55 Q2093.23 1482.55 2095.03 1479.01 Q2096.86 1475.44 2096.86 1468.33 Q2096.86 1461.2 2095.03 1457.66 Q2093.23 1454.1 2089.59 1454.1 M2089.59 1450.39 Q2095.4 1450.39 2098.46 1455 Q2101.54 1459.58 2101.54 1468.33 Q2101.54 1477.06 2098.46 1481.67 Q2095.4 1486.25 2089.59 1486.25 Q2083.78 1486.25 2080.71 1481.67 Q2077.65 1477.06 2077.65 1468.33 Q2077.65 1459.58 2080.71 1455 Q2083.78 1450.39 2089.59 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2228.99 1481.64 L2245.31 1481.64 L2245.31 1485.58 L2223.37 1485.58 L2223.37 1481.64 Q2226.03 1478.89 2230.61 1474.26 Q2235.22 1469.61 2236.4 1468.27 Q2238.64 1465.74 2239.52 1464.01 Q2240.43 1462.25 2240.43 1460.56 Q2240.43 1457.8 2238.48 1456.07 Q2236.56 1454.33 2233.46 1454.33 Q2231.26 1454.33 2228.81 1455.09 Q2226.38 1455.86 2223.6 1457.41 L2223.6 1452.69 Q2226.42 1451.55 2228.88 1450.97 Q2231.33 1450.39 2233.37 1450.39 Q2238.74 1450.39 2241.93 1453.08 Q2245.13 1455.77 2245.13 1460.26 Q2245.13 1462.39 2244.32 1464.31 Q2243.53 1466.2 2241.42 1468.8 Q2240.84 1469.47 2237.74 1472.69 Q2234.64 1475.88 2228.99 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2265.13 1454.1 Q2261.51 1454.1 2259.69 1457.66 Q2257.88 1461.2 2257.88 1468.33 Q2257.88 1475.44 2259.69 1479.01 Q2261.51 1482.55 2265.13 1482.55 Q2268.76 1482.55 2270.57 1479.01 Q2272.39 1475.44 2272.39 1468.33 Q2272.39 1461.2 2270.57 1457.66 Q2268.76 1454.1 2265.13 1454.1 M2265.13 1450.39 Q2270.94 1450.39 2273.99 1455 Q2277.07 1459.58 2277.07 1468.33 Q2277.07 1477.06 2273.99 1481.67 Q2270.94 1486.25 2265.13 1486.25 Q2259.32 1486.25 2256.24 1481.67 Q2253.18 1477.06 2253.18 1468.33 Q2253.18 1459.58 2256.24 1455 Q2259.32 1450.39 2265.13 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2295.29 1454.1 Q2291.68 1454.1 2289.85 1457.66 Q2288.04 1461.2 2288.04 1468.33 Q2288.04 1475.44 2289.85 1479.01 Q2291.68 1482.55 2295.29 1482.55 Q2298.92 1482.55 2300.73 1479.01 Q2302.56 1475.44 2302.56 1468.33 Q2302.56 1461.2 2300.73 1457.66 Q2298.92 1454.1 2295.29 1454.1 M2295.29 1450.39 Q2301.1 1450.39 2304.15 1455 Q2307.23 1459.58 2307.23 1468.33 Q2307.23 1477.06 2304.15 1481.67 Q2301.1 1486.25 2295.29 1486.25 Q2289.48 1486.25 2286.4 1481.67 Q2283.34 1477.06 2283.34 1468.33 Q2283.34 1459.58 2286.4 1455 Q2289.48 1450.39 2295.29 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2325.45 1454.1 Q2321.84 1454.1 2320.01 1457.66 Q2318.2 1461.2 2318.2 1468.33 Q2318.2 1475.44 2320.01 1479.01 Q2321.84 1482.55 2325.45 1482.55 Q2329.08 1482.55 2330.89 1479.01 Q2332.72 1475.44 2332.72 1468.33 Q2332.72 1461.2 2330.89 1457.66 Q2329.08 1454.1 2325.45 1454.1 M2325.45 1450.39 Q2331.26 1450.39 2334.31 1455 Q2337.39 1459.58 2337.39 1468.33 Q2337.39 1477.06 2334.31 1481.67 Q2331.26 1486.25 2325.45 1486.25 Q2319.64 1486.25 2316.56 1481.67 Q2313.5 1477.06 2313.5 1468.33 Q2313.5 1459.58 2316.56 1455 Q2319.64 1450.39 2325.45 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1778.77 1525.81 L1778.77 1562.76 L1786.54 1562.76 Q1796.38 1562.76 1800.93 1558.3 Q1805.51 1553.85 1805.51 1544.24 Q1805.51 1534.69 1800.93 1530.26 Q1796.38 1525.81 1786.54 1525.81 L1778.77 1525.81 M1772.34 1520.52 L1785.55 1520.52 Q1799.37 1520.52 1805.83 1526.28 Q1812.29 1532.01 1812.29 1544.24 Q1812.29 1556.52 1805.8 1562.28 Q1799.3 1568.04 1785.55 1568.04 L1772.34 1568.04 L1772.34 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1852.78 1548.76 L1852.78 1551.62 L1825.85 1551.62 Q1826.23 1557.67 1829.48 1560.85 Q1832.76 1564 1838.58 1564 Q1841.95 1564 1845.1 1563.17 Q1848.29 1562.35 1851.41 1560.69 L1851.41 1566.23 Q1848.26 1567.57 1844.95 1568.27 Q1841.64 1568.97 1838.23 1568.97 Q1829.7 1568.97 1824.7 1564 Q1819.74 1559.04 1819.74 1550.57 Q1819.74 1541.82 1824.45 1536.69 Q1829.19 1531.54 1837.21 1531.54 Q1844.4 1531.54 1848.57 1536.18 Q1852.78 1540.8 1852.78 1548.76 M1846.92 1547.04 Q1846.86 1542.23 1844.21 1539.37 Q1841.6 1536.5 1837.27 1536.5 Q1832.37 1536.5 1829.41 1539.27 Q1826.49 1542.04 1826.04 1547.07 L1846.92 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1890.14 1539.24 Q1892.34 1535.29 1895.39 1533.41 Q1898.45 1531.54 1902.59 1531.54 Q1908.16 1531.54 1911.18 1535.45 Q1914.2 1539.33 1914.2 1546.53 L1914.2 1568.04 L1908.32 1568.04 L1908.32 1546.72 Q1908.32 1541.59 1906.5 1539.11 Q1904.69 1536.63 1900.96 1536.63 Q1896.41 1536.63 1893.77 1539.65 Q1891.13 1542.68 1891.13 1547.9 L1891.13 1568.04 L1885.24 1568.04 L1885.24 1546.72 Q1885.24 1541.56 1883.43 1539.11 Q1881.61 1536.63 1877.82 1536.63 Q1873.34 1536.63 1870.69 1539.68 Q1868.05 1542.71 1868.05 1547.9 L1868.05 1568.04 L1862.16 1568.04 L1862.16 1532.4 L1868.05 1532.4 L1868.05 1537.93 Q1870.06 1534.66 1872.86 1533.1 Q1875.66 1531.54 1879.51 1531.54 Q1883.39 1531.54 1886.1 1533.51 Q1888.84 1535.48 1890.14 1539.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1942.09 1550.12 Q1934.99 1550.12 1932.25 1551.75 Q1929.51 1553.37 1929.51 1557.29 Q1929.51 1560.4 1931.55 1562.25 Q1933.62 1564.07 1937.15 1564.07 Q1942.02 1564.07 1944.95 1560.63 Q1947.91 1557.16 1947.91 1551.43 L1947.91 1550.12 L1942.09 1550.12 M1953.77 1547.71 L1953.77 1568.04 L1947.91 1568.04 L1947.91 1562.63 Q1945.91 1565.88 1942.91 1567.44 Q1939.92 1568.97 1935.59 1568.97 Q1930.12 1568.97 1926.87 1565.91 Q1923.66 1562.82 1923.66 1557.67 Q1923.66 1551.65 1927.67 1548.6 Q1931.71 1545.54 1939.7 1545.54 L1947.91 1545.54 L1947.91 1544.97 Q1947.91 1540.93 1945.24 1538.73 Q1942.6 1536.5 1937.79 1536.5 Q1934.73 1536.5 1931.84 1537.23 Q1928.94 1537.97 1926.27 1539.43 L1926.27 1534.02 Q1929.48 1532.78 1932.51 1532.17 Q1935.53 1531.54 1938.39 1531.54 Q1946.13 1531.54 1949.95 1535.55 Q1953.77 1539.56 1953.77 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1995.46 1546.53 L1995.46 1568.04 L1989.61 1568.04 L1989.61 1546.72 Q1989.61 1541.66 1987.63 1539.14 Q1985.66 1536.63 1981.71 1536.63 Q1976.97 1536.63 1974.23 1539.65 Q1971.5 1542.68 1971.5 1547.9 L1971.5 1568.04 L1965.61 1568.04 L1965.61 1532.4 L1971.5 1532.4 L1971.5 1537.93 Q1973.6 1534.72 1976.43 1533.13 Q1979.29 1531.54 1983.02 1531.54 Q1989.16 1531.54 1992.31 1535.36 Q1995.46 1539.14 1995.46 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2030.6 1537.81 L2030.6 1518.52 L2036.46 1518.52 L2036.46 1568.04 L2030.6 1568.04 L2030.6 1562.7 Q2028.76 1565.88 2025.92 1567.44 Q2023.12 1568.97 2019.17 1568.97 Q2012.71 1568.97 2008.64 1563.81 Q2004.6 1558.65 2004.6 1550.25 Q2004.6 1541.85 2008.64 1536.69 Q2012.71 1531.54 2019.17 1531.54 Q2023.12 1531.54 2025.92 1533.1 Q2028.76 1534.62 2030.6 1537.81 M2010.64 1550.25 Q2010.64 1556.71 2013.29 1560.4 Q2015.96 1564.07 2020.61 1564.07 Q2025.25 1564.07 2027.93 1560.4 Q2030.6 1556.71 2030.6 1550.25 Q2030.6 1543.79 2027.93 1540.13 Q2025.25 1536.44 2020.61 1536.44 Q2015.96 1536.44 2013.29 1540.13 Q2010.64 1543.79 2010.64 1550.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1456.05,1386.4 2352.76,1386.4 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1456.05,1079.86 2352.76,1079.86 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1456.05,773.326 2352.76,773.326 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1456.05,466.791 2352.76,466.791 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1456.05,160.256 2352.76,160.256 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1456.05,1423.18 1456.05,123.472 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1456.05,1386.4 1474.94,1386.4 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1456.05,1079.86 1474.94,1079.86 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1456.05,773.326 1474.94,773.326 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1456.05,466.791 1474.94,466.791 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1456.05,160.256 1474.94,160.256 "/>
+<path clip-path="url(#clip130)" d="M1408.1 1372.19 Q1404.49 1372.19 1402.66 1375.76 Q1400.86 1379.3 1400.86 1386.43 Q1400.86 1393.54 1402.66 1397.1 Q1404.49 1400.64 1408.1 1400.64 Q1411.74 1400.64 1413.54 1397.1 Q1415.37 1393.54 1415.37 1386.43 Q1415.37 1379.3 1413.54 1375.76 Q1411.74 1372.19 1408.1 1372.19 M1408.1 1368.49 Q1413.91 1368.49 1416.97 1373.1 Q1420.05 1377.68 1420.05 1386.43 Q1420.05 1395.16 1416.97 1399.76 Q1413.91 1404.35 1408.1 1404.35 Q1402.29 1404.35 1399.21 1399.76 Q1396.16 1395.16 1396.16 1386.43 Q1396.16 1377.68 1399.21 1373.1 Q1402.29 1368.49 1408.1 1368.49 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1341.81 1093.21 L1358.13 1093.21 L1358.13 1097.14 L1336.18 1097.14 L1336.18 1093.21 Q1338.84 1090.45 1343.43 1085.82 Q1348.03 1081.17 1349.21 1079.83 Q1351.46 1077.3 1352.34 1075.57 Q1353.24 1073.81 1353.24 1072.12 Q1353.24 1069.36 1351.3 1067.63 Q1349.38 1065.89 1346.27 1065.89 Q1344.07 1065.89 1341.62 1066.65 Q1339.19 1067.42 1336.41 1068.97 L1336.41 1064.25 Q1339.24 1063.11 1341.69 1062.53 Q1344.14 1061.96 1346.18 1061.96 Q1351.55 1061.96 1354.75 1064.64 Q1357.94 1067.33 1357.94 1071.82 Q1357.94 1073.95 1357.13 1075.87 Q1356.34 1077.77 1354.24 1080.36 Q1353.66 1081.03 1350.56 1084.25 Q1347.45 1087.44 1341.81 1093.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1367.99 1062.58 L1386.34 1062.58 L1386.34 1066.52 L1372.27 1066.52 L1372.27 1074.99 Q1373.29 1074.64 1374.31 1074.48 Q1375.32 1074.29 1376.34 1074.29 Q1382.13 1074.29 1385.51 1077.47 Q1388.89 1080.64 1388.89 1086.05 Q1388.89 1091.63 1385.42 1094.73 Q1381.94 1097.81 1375.63 1097.81 Q1373.45 1097.81 1371.18 1097.44 Q1368.94 1097.07 1366.53 1096.33 L1366.53 1091.63 Q1368.61 1092.77 1370.83 1093.32 Q1373.06 1093.88 1375.53 1093.88 Q1379.54 1093.88 1381.88 1091.77 Q1384.21 1089.66 1384.21 1086.05 Q1384.21 1082.44 1381.88 1080.34 Q1379.54 1078.23 1375.53 1078.23 Q1373.66 1078.23 1371.78 1078.65 Q1369.93 1079.06 1367.99 1079.94 L1367.99 1062.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1408.1 1065.66 Q1404.49 1065.66 1402.66 1069.22 Q1400.86 1072.77 1400.86 1079.9 Q1400.86 1087 1402.66 1090.57 Q1404.49 1094.11 1408.1 1094.11 Q1411.74 1094.11 1413.54 1090.57 Q1415.37 1087 1415.37 1079.9 Q1415.37 1072.77 1413.54 1069.22 Q1411.74 1065.66 1408.1 1065.66 M1408.1 1061.96 Q1413.91 1061.96 1416.97 1066.56 Q1420.05 1071.15 1420.05 1079.9 Q1420.05 1088.62 1416.97 1093.23 Q1413.91 1097.81 1408.1 1097.81 Q1402.29 1097.81 1399.21 1093.23 Q1396.16 1088.62 1396.16 1079.9 Q1396.16 1071.15 1399.21 1066.56 Q1402.29 1061.96 1408.1 1061.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1337.82 756.046 L1356.18 756.046 L1356.18 759.981 L1342.11 759.981 L1342.11 768.453 Q1343.13 768.106 1344.14 767.944 Q1345.16 767.759 1346.18 767.759 Q1351.97 767.759 1355.35 770.93 Q1358.73 774.101 1358.73 779.518 Q1358.73 785.097 1355.26 788.199 Q1351.78 791.277 1345.46 791.277 Q1343.29 791.277 1341.02 790.907 Q1338.77 790.537 1336.37 789.796 L1336.37 785.097 Q1338.45 786.231 1340.67 786.787 Q1342.89 787.342 1345.37 787.342 Q1349.38 787.342 1351.71 785.236 Q1354.05 783.129 1354.05 779.518 Q1354.05 775.907 1351.71 773.801 Q1349.38 771.694 1345.37 771.694 Q1343.5 771.694 1341.62 772.111 Q1339.77 772.527 1337.82 773.407 L1337.82 756.046 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1377.94 759.125 Q1374.33 759.125 1372.5 762.689 Q1370.7 766.231 1370.7 773.361 Q1370.7 780.467 1372.5 784.032 Q1374.33 787.574 1377.94 787.574 Q1381.57 787.574 1383.38 784.032 Q1385.21 780.467 1385.21 773.361 Q1385.21 766.231 1383.38 762.689 Q1381.57 759.125 1377.94 759.125 M1377.94 755.421 Q1383.75 755.421 1386.81 760.027 Q1389.88 764.611 1389.88 773.361 Q1389.88 782.088 1386.81 786.694 Q1383.75 791.277 1377.94 791.277 Q1372.13 791.277 1369.05 786.694 Q1366 782.088 1366 773.361 Q1366 764.611 1369.05 760.027 Q1372.13 755.421 1377.94 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1408.1 759.125 Q1404.49 759.125 1402.66 762.689 Q1400.86 766.231 1400.86 773.361 Q1400.86 780.467 1402.66 784.032 Q1404.49 787.574 1408.1 787.574 Q1411.74 787.574 1413.54 784.032 Q1415.37 780.467 1415.37 773.361 Q1415.37 766.231 1413.54 762.689 Q1411.74 759.125 1408.1 759.125 M1408.1 755.421 Q1413.91 755.421 1416.97 760.027 Q1420.05 764.611 1420.05 773.361 Q1420.05 782.088 1416.97 786.694 Q1413.91 791.277 1408.1 791.277 Q1402.29 791.277 1399.21 786.694 Q1396.16 782.088 1396.16 773.361 Q1396.16 764.611 1399.21 760.027 Q1402.29 755.421 1408.1 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1336.6 449.511 L1358.82 449.511 L1358.82 451.502 L1346.27 484.071 L1341.39 484.071 L1353.2 453.446 L1336.6 453.446 L1336.6 449.511 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1367.99 449.511 L1386.34 449.511 L1386.34 453.446 L1372.27 453.446 L1372.27 461.918 Q1373.29 461.571 1374.31 461.409 Q1375.32 461.224 1376.34 461.224 Q1382.13 461.224 1385.51 464.395 Q1388.89 467.567 1388.89 472.983 Q1388.89 478.562 1385.42 481.664 Q1381.94 484.742 1375.63 484.742 Q1373.45 484.742 1371.18 484.372 Q1368.94 484.002 1366.53 483.261 L1366.53 478.562 Q1368.61 479.696 1370.83 480.252 Q1373.06 480.807 1375.53 480.807 Q1379.54 480.807 1381.88 478.701 Q1384.21 476.594 1384.21 472.983 Q1384.21 469.372 1381.88 467.266 Q1379.54 465.159 1375.53 465.159 Q1373.66 465.159 1371.78 465.576 Q1369.93 465.993 1367.99 466.872 L1367.99 449.511 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1408.1 452.59 Q1404.49 452.59 1402.66 456.155 Q1400.86 459.696 1400.86 466.826 Q1400.86 473.932 1402.66 477.497 Q1404.49 481.039 1408.1 481.039 Q1411.74 481.039 1413.54 477.497 Q1415.37 473.932 1415.37 466.826 Q1415.37 459.696 1413.54 456.155 Q1411.74 452.59 1408.1 452.59 M1408.1 448.886 Q1413.91 448.886 1416.97 453.493 Q1420.05 458.076 1420.05 466.826 Q1420.05 475.553 1416.97 480.159 Q1413.91 484.742 1408.1 484.742 Q1402.29 484.742 1399.21 480.159 Q1396.16 475.553 1396.16 466.826 Q1396.16 458.076 1399.21 453.493 Q1402.29 448.886 1408.1 448.886 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1308.43 173.601 L1316.07 173.601 L1316.07 147.236 L1307.76 148.902 L1307.76 144.643 L1316.02 142.976 L1320.7 142.976 L1320.7 173.601 L1328.33 173.601 L1328.33 177.536 L1308.43 177.536 L1308.43 173.601 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1347.78 146.055 Q1344.17 146.055 1342.34 149.62 Q1340.53 153.161 1340.53 160.291 Q1340.53 167.397 1342.34 170.962 Q1344.17 174.504 1347.78 174.504 Q1351.41 174.504 1353.22 170.962 Q1355.05 167.397 1355.05 160.291 Q1355.05 153.161 1353.22 149.62 Q1351.41 146.055 1347.78 146.055 M1347.78 142.351 Q1353.59 142.351 1356.64 146.958 Q1359.72 151.541 1359.72 160.291 Q1359.72 169.018 1356.64 173.624 Q1353.59 178.208 1347.78 178.208 Q1341.97 178.208 1338.89 173.624 Q1335.83 169.018 1335.83 160.291 Q1335.83 151.541 1338.89 146.958 Q1341.97 142.351 1347.78 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1377.94 146.055 Q1374.33 146.055 1372.5 149.62 Q1370.7 153.161 1370.7 160.291 Q1370.7 167.397 1372.5 170.962 Q1374.33 174.504 1377.94 174.504 Q1381.57 174.504 1383.38 170.962 Q1385.21 167.397 1385.21 160.291 Q1385.21 153.161 1383.38 149.62 Q1381.57 146.055 1377.94 146.055 M1377.94 142.351 Q1383.75 142.351 1386.81 146.958 Q1389.88 151.541 1389.88 160.291 Q1389.88 169.018 1386.81 173.624 Q1383.75 178.208 1377.94 178.208 Q1372.13 178.208 1369.05 173.624 Q1366 169.018 1366 160.291 Q1366 151.541 1369.05 146.958 Q1372.13 142.351 1377.94 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1408.1 146.055 Q1404.49 146.055 1402.66 149.62 Q1400.86 153.161 1400.86 160.291 Q1400.86 167.397 1402.66 170.962 Q1404.49 174.504 1408.1 174.504 Q1411.74 174.504 1413.54 170.962 Q1415.37 167.397 1415.37 160.291 Q1415.37 153.161 1413.54 149.62 Q1411.74 146.055 1408.1 146.055 M1408.1 142.351 Q1413.91 142.351 1416.97 146.958 Q1420.05 151.541 1420.05 160.291 Q1420.05 169.018 1416.97 173.624 Q1413.91 178.208 1408.1 178.208 Q1402.29 178.208 1399.21 173.624 Q1396.16 169.018 1396.16 160.291 Q1396.16 151.541 1399.21 146.958 Q1402.29 142.351 1408.1 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1215.26 1000.18 L1252.22 1000.18 L1252.22 992.418 Q1252.22 982.583 1247.76 978.031 Q1243.3 973.448 1233.69 973.448 Q1224.14 973.448 1219.72 978.031 Q1215.26 982.583 1215.26 992.418 L1215.26 1000.18 M1209.98 1006.61 L1209.98 993.404 Q1209.98 979.591 1215.74 973.13 Q1221.47 966.668 1233.69 966.668 Q1245.98 966.668 1251.74 973.161 Q1257.5 979.654 1257.5 993.404 L1257.5 1006.61 L1209.98 1006.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1221.85 956.674 L1221.85 950.818 L1257.5 950.818 L1257.5 956.674 L1221.85 956.674 M1207.97 956.674 L1207.97 950.818 L1215.39 950.818 L1215.39 956.674 L1207.97 956.674 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1222.9 915.838 L1228.44 915.838 Q1227.17 918.321 1226.53 920.994 Q1225.89 923.668 1225.89 926.533 Q1225.89 930.893 1227.23 933.089 Q1228.57 935.254 1231.24 935.254 Q1233.28 935.254 1234.46 933.694 Q1235.6 932.134 1236.65 927.424 L1237.1 925.419 Q1238.43 919.18 1240.89 916.57 Q1243.3 913.929 1247.66 913.929 Q1252.63 913.929 1255.53 917.875 Q1258.42 921.79 1258.42 928.665 Q1258.42 931.53 1257.85 934.649 Q1257.31 937.736 1256.19 941.174 L1250.15 941.174 Q1251.83 937.927 1252.69 934.776 Q1253.52 931.625 1253.52 928.538 Q1253.52 924.4 1252.12 922.172 Q1250.69 919.944 1248.11 919.944 Q1245.72 919.944 1244.45 921.567 Q1243.18 923.159 1242 928.601 L1241.52 930.639 Q1240.38 936.081 1238.02 938.5 Q1235.63 940.919 1231.5 940.919 Q1226.47 940.919 1223.73 937.354 Q1220.99 933.79 1220.99 927.233 Q1220.99 923.986 1221.47 921.122 Q1221.95 918.257 1222.9 915.838 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1252.15 898.937 L1271.06 898.937 L1271.06 904.826 L1221.85 904.826 L1221.85 898.937 L1227.26 898.937 Q1224.08 897.091 1222.55 894.29 Q1220.99 891.458 1220.99 887.543 Q1220.99 881.05 1226.15 877.007 Q1231.3 872.933 1239.71 872.933 Q1248.11 872.933 1253.27 877.007 Q1258.42 881.05 1258.42 887.543 Q1258.42 891.458 1256.89 894.29 Q1255.34 897.091 1252.15 898.937 M1239.71 879.013 Q1233.25 879.013 1229.59 881.686 Q1225.89 884.328 1225.89 888.975 Q1225.89 893.622 1229.59 896.296 Q1233.25 898.937 1239.71 898.937 Q1246.17 898.937 1249.86 896.296 Q1253.52 893.622 1253.52 888.975 Q1253.52 884.328 1249.86 881.686 Q1246.17 879.013 1239.71 879.013 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1239.58 847.025 Q1239.58 854.123 1241.2 856.86 Q1242.83 859.597 1246.74 859.597 Q1249.86 859.597 1251.71 857.56 Q1253.52 855.491 1253.52 851.958 Q1253.52 847.089 1250.08 844.16 Q1246.61 841.2 1240.89 841.2 L1239.58 841.2 L1239.58 847.025 M1237.16 835.344 L1257.5 835.344 L1257.5 841.2 L1252.09 841.2 Q1255.34 843.206 1256.89 846.197 Q1258.42 849.189 1258.42 853.518 Q1258.42 858.992 1255.37 862.239 Q1252.28 865.454 1247.12 865.454 Q1241.11 865.454 1238.05 861.443 Q1235 857.401 1235 849.412 L1235 841.2 L1234.42 841.2 Q1230.38 841.2 1228.19 843.874 Q1225.96 846.516 1225.96 851.322 Q1225.96 854.377 1226.69 857.274 Q1227.42 860.17 1228.89 862.844 L1223.47 862.844 Q1222.23 859.629 1221.63 856.605 Q1220.99 853.582 1220.99 850.717 Q1220.99 842.983 1225 839.163 Q1229.01 835.344 1237.16 835.344 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1211.73 817.488 L1221.85 817.488 L1221.85 805.425 L1226.4 805.425 L1226.4 817.488 L1245.75 817.488 Q1250.12 817.488 1251.36 816.31 Q1252.6 815.101 1252.6 811.441 L1252.6 805.425 L1257.5 805.425 L1257.5 811.441 Q1257.5 818.22 1254.99 820.798 Q1252.44 823.376 1245.75 823.376 L1226.4 823.376 L1226.4 827.673 L1221.85 827.673 L1221.85 823.376 L1211.73 823.376 L1211.73 817.488 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1223.22 772.069 L1228.69 772.069 Q1227.33 774.551 1226.66 777.066 Q1225.96 779.548 1225.96 782.095 Q1225.96 787.792 1229.59 790.943 Q1233.18 794.094 1239.71 794.094 Q1246.23 794.094 1249.86 790.943 Q1253.46 787.792 1253.46 782.095 Q1253.46 779.548 1252.79 777.066 Q1252.09 774.551 1250.72 772.069 L1256.13 772.069 Q1257.28 774.52 1257.85 777.161 Q1258.42 779.771 1258.42 782.731 Q1258.42 790.784 1253.36 795.526 Q1248.3 800.269 1239.71 800.269 Q1230.99 800.269 1225.99 795.495 Q1220.99 790.688 1220.99 782.349 Q1220.99 779.644 1221.57 777.066 Q1222.11 774.488 1223.22 772.069 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1235.98 732.251 L1257.5 732.251 L1257.5 738.108 L1236.17 738.108 Q1231.11 738.108 1228.6 740.081 Q1226.08 742.054 1226.08 746.001 Q1226.08 750.744 1229.11 753.481 Q1232.13 756.218 1237.35 756.218 L1257.5 756.218 L1257.5 762.106 L1207.97 762.106 L1207.97 756.218 L1227.39 756.218 Q1224.18 754.117 1222.58 751.285 Q1220.99 748.42 1220.99 744.696 Q1220.99 738.553 1224.81 735.402 Q1228.6 732.251 1235.98 732.251 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1207.97 700.391 L1207.97 686.896 L1212.53 686.896 L1212.53 694.534 L1261.54 694.534 L1261.54 686.896 L1266.09 686.896 L1266.09 700.391 L1207.97 700.391 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1209.98 674.164 L1209.98 664.584 L1242.32 652.457 L1209.98 640.267 L1209.98 630.686 L1257.5 630.686 L1257.5 636.957 L1215.77 636.957 L1248.36 649.211 L1248.36 655.672 L1215.77 667.926 L1257.5 667.926 L1257.5 674.164 L1209.98 674.164 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1209.98 622.156 L1209.98 615.663 L1250.15 605.669 L1209.98 595.707 L1209.98 588.482 L1250.15 578.488 L1209.98 568.525 L1209.98 562 L1257.5 573.936 L1257.5 582.021 L1216.25 592.047 L1257.5 602.168 L1257.5 610.252 L1209.98 622.156 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1207.97 540.039 L1266.09 540.039 L1266.09 553.534 L1261.54 553.534 L1261.54 545.927 L1212.53 545.927 L1212.53 553.534 L1207.97 553.534 L1207.97 540.039 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1615.67 18.8205 L1615.67 65.8515 L1625.56 65.8515 Q1638.07 65.8515 1643.87 60.1802 Q1649.7 54.509 1649.7 42.2752 Q1649.7 30.1225 1643.87 24.4918 Q1638.07 18.8205 1625.56 18.8205 L1615.67 18.8205 M1607.49 12.096 L1624.3 12.096 Q1641.88 12.096 1650.11 19.4281 Q1658.33 26.7198 1658.33 42.2752 Q1658.33 57.9117 1650.07 65.2439 Q1641.8 72.576 1624.3 72.576 L1607.49 72.576 L1607.49 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1671.05 27.2059 L1678.5 27.2059 L1678.5 72.576 L1671.05 72.576 L1671.05 27.2059 M1671.05 9.54393 L1678.5 9.54393 L1678.5 18.9825 L1671.05 18.9825 L1671.05 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1723.02 28.5427 L1723.02 35.5912 Q1719.86 33.9709 1716.46 33.1607 Q1713.06 32.3505 1709.41 32.3505 Q1703.86 32.3505 1701.07 34.0519 Q1698.31 35.7533 1698.31 39.156 Q1698.31 41.7486 1700.3 43.2475 Q1702.28 44.7058 1708.28 46.0426 L1710.83 46.6097 Q1718.77 48.3111 1722.09 51.4303 Q1725.45 54.509 1725.45 60.0587 Q1725.45 66.3781 1720.43 70.0644 Q1715.45 73.7508 1706.7 73.7508 Q1703.05 73.7508 1699.08 73.0216 Q1695.15 72.3329 1690.78 70.9151 L1690.78 63.2184 Q1694.91 65.3654 1698.92 66.4591 Q1702.93 67.5124 1706.86 67.5124 Q1712.13 67.5124 1714.96 65.73 Q1717.8 63.9071 1717.8 60.6258 Q1717.8 57.5877 1715.73 55.9673 Q1713.7 54.3469 1706.78 52.8481 L1704.19 52.2405 Q1697.26 50.7821 1694.18 47.7845 Q1691.1 44.7463 1691.1 39.4801 Q1691.1 33.0797 1695.64 29.5959 Q1700.17 26.1121 1708.52 26.1121 Q1712.65 26.1121 1716.3 26.7198 Q1719.94 27.3274 1723.02 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1744.53 65.7705 L1744.53 89.8329 L1737.04 89.8329 L1737.04 27.2059 L1744.53 27.2059 L1744.53 34.0924 Q1746.88 30.0415 1750.45 28.0971 Q1754.05 26.1121 1759.03 26.1121 Q1767.3 26.1121 1772.44 32.6746 Q1777.63 39.2371 1777.63 49.9314 Q1777.63 60.6258 1772.44 67.1883 Q1767.3 73.7508 1759.03 73.7508 Q1754.05 73.7508 1750.45 71.8063 Q1746.88 69.8214 1744.53 65.7705 M1769.89 49.9314 Q1769.89 41.7081 1766.49 37.0496 Q1763.13 32.3505 1757.21 32.3505 Q1751.3 32.3505 1747.89 37.0496 Q1744.53 41.7081 1744.53 49.9314 Q1744.53 58.1548 1747.89 62.8538 Q1751.3 67.5124 1757.21 67.5124 Q1763.13 67.5124 1766.49 62.8538 Q1769.89 58.1548 1769.89 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1810.6 49.7694 Q1801.57 49.7694 1798.09 51.8354 Q1794.6 53.9013 1794.6 58.8839 Q1794.6 62.8538 1797.19 65.2034 Q1799.83 67.5124 1804.32 67.5124 Q1810.52 67.5124 1814.25 63.1374 Q1818.02 58.7219 1818.02 51.4303 L1818.02 49.7694 L1810.6 49.7694 M1825.47 46.6907 L1825.47 72.576 L1818.02 72.576 L1818.02 65.6895 Q1815.46 69.8214 1811.66 71.8063 Q1807.85 73.7508 1802.34 73.7508 Q1795.37 73.7508 1791.24 69.8619 Q1787.15 65.9325 1787.15 59.3701 Q1787.15 51.7138 1792.25 47.825 Q1797.4 43.9361 1807.56 43.9361 L1818.02 43.9361 L1818.02 43.2069 Q1818.02 38.0623 1814.61 35.2672 Q1811.25 32.4315 1805.13 32.4315 Q1801.24 32.4315 1797.56 33.3632 Q1793.87 34.295 1790.47 36.1584 L1790.47 29.2718 Q1794.56 27.692 1798.41 26.9223 Q1802.26 26.1121 1805.9 26.1121 Q1815.75 26.1121 1820.61 31.2163 Q1825.47 36.3204 1825.47 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1848.19 14.324 L1848.19 27.2059 L1863.55 27.2059 L1863.55 32.9987 L1848.19 32.9987 L1848.19 57.6282 Q1848.19 63.1779 1849.69 64.7578 Q1851.23 66.3376 1855.89 66.3376 L1863.55 66.3376 L1863.55 72.576 L1855.89 72.576 Q1847.26 72.576 1843.98 69.3758 Q1840.7 66.1351 1840.7 57.6282 L1840.7 32.9987 L1835.23 32.9987 L1835.23 27.2059 L1840.7 27.2059 L1840.7 14.324 L1848.19 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1906 28.9478 L1906 35.9153 Q1902.84 34.1734 1899.64 33.3227 Q1896.48 32.4315 1893.24 32.4315 Q1885.99 32.4315 1881.98 37.0496 Q1877.97 41.6271 1877.97 49.9314 Q1877.97 58.2358 1881.98 62.8538 Q1885.99 67.4314 1893.24 67.4314 Q1896.48 67.4314 1899.64 66.5807 Q1902.84 65.6895 1906 63.9476 L1906 70.8341 Q1902.88 72.2924 1899.52 73.0216 Q1896.2 73.7508 1892.43 73.7508 Q1882.18 73.7508 1876.15 67.3098 Q1870.11 60.8689 1870.11 49.9314 Q1870.11 38.832 1876.19 32.472 Q1882.3 26.1121 1892.92 26.1121 Q1896.36 26.1121 1899.64 26.8413 Q1902.92 27.5299 1906 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1956.68 45.1919 L1956.68 72.576 L1949.22 72.576 L1949.22 45.4349 Q1949.22 38.994 1946.71 35.7938 Q1944.2 32.5936 1939.18 32.5936 Q1933.14 32.5936 1929.66 36.4419 Q1926.17 40.2903 1926.17 46.9338 L1926.17 72.576 L1918.68 72.576 L1918.68 9.54393 L1926.17 9.54393 L1926.17 34.2544 Q1928.85 30.163 1932.45 28.1376 Q1936.1 26.1121 1940.84 26.1121 Q1948.66 26.1121 1952.67 30.9732 Q1956.68 35.7938 1956.68 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M1997.23 9.54393 L2014.4 9.54393 L2014.4 15.3367 L2004.68 15.3367 L2004.68 77.7206 L2014.4 77.7206 L2014.4 83.5134 L1997.23 83.5134 L1997.23 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2030.61 12.096 L2042.8 12.096 L2058.23 53.2532 L2073.75 12.096 L2085.94 12.096 L2085.94 72.576 L2077.96 72.576 L2077.96 19.4686 L2062.37 60.9499 L2054.14 60.9499 L2038.55 19.4686 L2038.55 72.576 L2030.61 72.576 L2030.61 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2096.8 12.096 L2105.06 12.096 L2117.78 63.2184 L2130.46 12.096 L2139.66 12.096 L2152.38 63.2184 L2165.06 12.096 L2173.36 12.096 L2158.17 72.576 L2147.88 72.576 L2135.12 20.0763 L2122.24 72.576 L2111.95 72.576 L2096.8 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2201.31 9.54393 L2201.31 83.5134 L2184.14 83.5134 L2184.14 77.7206 L2193.82 77.7206 L2193.82 15.3367 L2184.14 15.3367 L2184.14 9.54393 L2201.31 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip133)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1481.43,1263.78 1551.92,1079.86 1622.42,895.94 1692.91,712.019 1763.41,528.098 1833.91,344.177 1904.4,160.256 1974.9,344.177 2045.39,160.256 2115.89,160.256 2186.39,160.256 2256.88,160.256 2327.38,160.256 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1481.43,1386.4 1551.92,1386.4 1622.42,1386.4 1692.91,1386.4 1763.41,1386.4 1833.91,1386.4 1904.4,1386.4 1974.9,1018.55 2045.39,1018.55 2115.89,834.633 2186.39,650.712 2256.88,466.791 2327.38,282.87 "/>
+<polyline clip-path="url(#clip133)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1481.43,1141.17 1551.92,1141.17 1622.42,1141.17 1692.91,1141.17 1763.41,1141.17 1833.91,1141.17 1904.4,1141.17 1974.9,1141.17 2045.39,1141.17 2115.89,1141.17 2186.39,1141.17 2256.88,1141.17 2327.38,1141.17 "/>
+<path clip-path="url(#clip130)" d="M2119.93 1379.86 L2322.87 1379.86 L2322.87 1172.5 L2119.93 1172.5  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip130)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2119.93,1379.86 2322.87,1379.86 2322.87,1172.5 2119.93,1172.5 2119.93,1379.86 "/>
+<polyline clip-path="url(#clip130)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="2129.89,1224.34 2189.68,1224.34 "/>
+<path clip-path="url(#clip130)" d="M2225.19 1236.69 L2225.19 1227.4 L2217.56 1227.4 L2217.56 1223.56 L2229.82 1223.56 L2229.82 1238.4 Q2227.12 1240.32 2223.85 1241.32 Q2220.59 1242.29 2216.88 1242.29 Q2208.78 1242.29 2204.2 1237.57 Q2199.64 1232.82 2199.64 1224.37 Q2199.64 1215.9 2204.2 1211.18 Q2208.78 1206.43 2216.88 1206.43 Q2220.26 1206.43 2223.3 1207.26 Q2226.35 1208.1 2228.92 1209.72 L2228.92 1214.7 Q2226.33 1212.5 2223.41 1211.38 Q2220.5 1210.27 2217.28 1210.27 Q2210.94 1210.27 2207.74 1213.82 Q2204.57 1217.36 2204.57 1224.37 Q2204.57 1231.36 2207.74 1234.9 Q2210.94 1238.44 2217.28 1238.44 Q2219.75 1238.44 2221.7 1238.03 Q2223.64 1237.59 2225.19 1236.69 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2239.59 1237.68 L2247.23 1237.68 L2247.23 1211.32 L2238.92 1212.98 L2238.92 1208.72 L2247.18 1207.06 L2251.86 1207.06 L2251.86 1237.68 L2259.5 1237.68 L2259.5 1241.62 L2239.59 1241.62 L2239.59 1237.68 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip130)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="2129.89,1276.18 2189.68,1276.18 "/>
+<path clip-path="url(#clip130)" d="M2225.19 1288.53 L2225.19 1279.24 L2217.56 1279.24 L2217.56 1275.4 L2229.82 1275.4 L2229.82 1290.24 Q2227.12 1292.16 2223.85 1293.16 Q2220.59 1294.13 2216.88 1294.13 Q2208.78 1294.13 2204.2 1289.41 Q2199.64 1284.66 2199.64 1276.21 Q2199.64 1267.74 2204.2 1263.02 Q2208.78 1258.27 2216.88 1258.27 Q2220.26 1258.27 2223.3 1259.1 Q2226.35 1259.94 2228.92 1261.56 L2228.92 1266.54 Q2226.33 1264.34 2223.41 1263.22 Q2220.5 1262.11 2217.28 1262.11 Q2210.94 1262.11 2207.74 1265.66 Q2204.57 1269.2 2204.57 1276.21 Q2204.57 1283.2 2207.74 1286.74 Q2210.94 1290.28 2217.28 1290.28 Q2219.75 1290.28 2221.7 1289.87 Q2223.64 1289.43 2225.19 1288.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2242.81 1289.52 L2259.13 1289.52 L2259.13 1293.46 L2237.18 1293.46 L2237.18 1289.52 Q2239.85 1286.77 2244.43 1282.14 Q2249.04 1277.48 2250.22 1276.14 Q2252.46 1273.62 2253.34 1271.88 Q2254.25 1270.12 2254.25 1268.43 Q2254.25 1265.68 2252.3 1263.94 Q2250.38 1262.21 2247.28 1262.21 Q2245.08 1262.21 2242.62 1262.97 Q2240.19 1263.73 2237.42 1265.29 L2237.42 1260.56 Q2240.24 1259.43 2242.69 1258.85 Q2245.15 1258.27 2247.18 1258.27 Q2252.56 1258.27 2255.75 1260.96 Q2258.94 1263.64 2258.94 1268.13 Q2258.94 1270.26 2258.13 1272.18 Q2257.35 1274.08 2255.24 1276.67 Q2254.66 1277.35 2251.56 1280.56 Q2248.46 1283.76 2242.81 1289.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip130)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="2129.89,1328.02 2189.68,1328.02 "/>
+<path clip-path="url(#clip130)" d="M2199.64 1310.74 L2204.36 1310.74 L2211.63 1339.95 L2218.87 1310.74 L2224.13 1310.74 L2231.4 1339.95 L2238.64 1310.74 L2243.39 1310.74 L2234.71 1345.3 L2228.83 1345.3 L2221.54 1315.3 L2214.18 1345.3 L2208.3 1345.3 L2199.64 1310.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2248.37 1319.37 L2252.62 1319.37 L2252.62 1345.3 L2248.37 1345.3 L2248.37 1319.37 M2248.37 1309.28 L2252.62 1309.28 L2252.62 1314.67 L2248.37 1314.67 L2248.37 1309.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2283.09 1329.65 L2283.09 1345.3 L2278.83 1345.3 L2278.83 1329.79 Q2278.83 1326.11 2277.39 1324.28 Q2275.96 1322.45 2273.09 1322.45 Q2269.64 1322.45 2267.65 1324.65 Q2265.66 1326.85 2265.66 1330.64 L2265.66 1345.3 L2261.37 1345.3 L2261.37 1319.37 L2265.66 1319.37 L2265.66 1323.4 Q2267.18 1321.06 2269.24 1319.9 Q2271.33 1318.75 2274.04 1318.75 Q2278.5 1318.75 2280.8 1321.52 Q2283.09 1324.28 2283.09 1329.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip130)" d="M2308.64 1323.31 L2308.64 1309.28 L2312.9 1309.28 L2312.9 1345.3 L2308.64 1345.3 L2308.64 1341.41 Q2307.3 1343.72 2305.24 1344.86 Q2303.2 1345.97 2300.33 1345.97 Q2295.63 1345.97 2292.67 1342.22 Q2289.73 1338.47 2289.73 1332.36 Q2289.73 1326.25 2292.67 1322.5 Q2295.63 1318.75 2300.33 1318.75 Q2303.2 1318.75 2305.24 1319.88 Q2307.3 1320.99 2308.64 1323.31 M2294.13 1332.36 Q2294.13 1337.06 2296.05 1339.74 Q2297.99 1342.4 2301.37 1342.4 Q2304.75 1342.4 2306.7 1339.74 Q2308.64 1337.06 2308.64 1332.36 Q2308.64 1327.66 2306.7 1325 Q2304.75 1322.31 2301.37 1322.31 Q2297.99 1322.31 2296.05 1325 Q2294.13 1327.66 2294.13 1332.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/applications/power_systems/ab8a43ee.svg b/previews/PR3525/tutorials/applications/power_systems/ab8a43ee.svg
new file mode 100644
index 00000000000..b77c2dfe914
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/power_systems/ab8a43ee.svg
@@ -0,0 +1,78 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip070">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip070)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip071">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip070)" d="M262.551 1423.18 L1166.97 1423.18 L1166.97 123.472 L262.551 123.472  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip072">
+    <rect x="262" y="123" width="905" height="1301"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="382.951,1423.18 382.951,123.472 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="619.957,1423.18 619.957,123.472 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="856.964,1423.18 856.964,123.472 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1093.97,1423.18 1093.97,123.472 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="262.551,1423.18 1166.97,1423.18 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="382.951,1423.18 382.951,1404.28 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="619.957,1423.18 619.957,1404.28 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="856.964,1423.18 856.964,1404.28 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1093.97,1423.18 1093.97,1404.28 "/>
+<path clip-path="url(#clip070)" d="M342.569 1451.02 L360.925 1451.02 L360.925 1454.96 L346.851 1454.96 L346.851 1463.43 Q347.87 1463.08 348.888 1462.92 Q349.907 1462.73 350.925 1462.73 Q356.712 1462.73 360.092 1465.9 Q363.472 1469.08 363.472 1474.49 Q363.472 1480.07 359.999 1483.17 Q356.527 1486.25 350.208 1486.25 Q348.032 1486.25 345.763 1485.88 Q343.518 1485.51 341.111 1484.77 L341.111 1480.07 Q343.194 1481.2 345.416 1481.76 Q347.638 1482.32 350.115 1482.32 Q354.12 1482.32 356.458 1480.21 Q358.796 1478.1 358.796 1474.49 Q358.796 1470.88 356.458 1468.77 Q354.12 1466.67 350.115 1466.67 Q348.24 1466.67 346.365 1467.08 Q344.513 1467.5 342.569 1468.38 L342.569 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M382.684 1454.1 Q379.073 1454.1 377.245 1457.66 Q375.439 1461.2 375.439 1468.33 Q375.439 1475.44 377.245 1479.01 Q379.073 1482.55 382.684 1482.55 Q386.319 1482.55 388.124 1479.01 Q389.953 1475.44 389.953 1468.33 Q389.953 1461.2 388.124 1457.66 Q386.319 1454.1 382.684 1454.1 M382.684 1450.39 Q388.495 1450.39 391.55 1455 Q394.629 1459.58 394.629 1468.33 Q394.629 1477.06 391.55 1481.67 Q388.495 1486.25 382.684 1486.25 Q376.874 1486.25 373.796 1481.67 Q370.74 1477.06 370.74 1468.33 Q370.74 1459.58 373.796 1455 Q376.874 1450.39 382.684 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M412.846 1454.1 Q409.235 1454.1 407.407 1457.66 Q405.601 1461.2 405.601 1468.33 Q405.601 1475.44 407.407 1479.01 Q409.235 1482.55 412.846 1482.55 Q416.481 1482.55 418.286 1479.01 Q420.115 1475.44 420.115 1468.33 Q420.115 1461.2 418.286 1457.66 Q416.481 1454.1 412.846 1454.1 M412.846 1450.39 Q418.656 1450.39 421.712 1455 Q424.791 1459.58 424.791 1468.33 Q424.791 1477.06 421.712 1481.67 Q418.656 1486.25 412.846 1486.25 Q407.036 1486.25 403.957 1481.67 Q400.902 1477.06 400.902 1468.33 Q400.902 1459.58 403.957 1455 Q407.036 1450.39 412.846 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M564.483 1481.64 L572.122 1481.64 L572.122 1455.28 L563.812 1456.95 L563.812 1452.69 L572.076 1451.02 L576.752 1451.02 L576.752 1481.64 L584.39 1481.64 L584.39 1485.58 L564.483 1485.58 L564.483 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M603.835 1454.1 Q600.224 1454.1 598.395 1457.66 Q596.589 1461.2 596.589 1468.33 Q596.589 1475.44 598.395 1479.01 Q600.224 1482.55 603.835 1482.55 Q607.469 1482.55 609.275 1479.01 Q611.103 1475.44 611.103 1468.33 Q611.103 1461.2 609.275 1457.66 Q607.469 1454.1 603.835 1454.1 M603.835 1450.39 Q609.645 1450.39 612.7 1455 Q615.779 1459.58 615.779 1468.33 Q615.779 1477.06 612.7 1481.67 Q609.645 1486.25 603.835 1486.25 Q598.025 1486.25 594.946 1481.67 Q591.89 1477.06 591.89 1468.33 Q591.89 1459.58 594.946 1455 Q598.025 1450.39 603.835 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M633.997 1454.1 Q630.386 1454.1 628.557 1457.66 Q626.751 1461.2 626.751 1468.33 Q626.751 1475.44 628.557 1479.01 Q630.386 1482.55 633.997 1482.55 Q637.631 1482.55 639.436 1479.01 Q641.265 1475.44 641.265 1468.33 Q641.265 1461.2 639.436 1457.66 Q637.631 1454.1 633.997 1454.1 M633.997 1450.39 Q639.807 1450.39 642.862 1455 Q645.941 1459.58 645.941 1468.33 Q645.941 1477.06 642.862 1481.67 Q639.807 1486.25 633.997 1486.25 Q628.186 1486.25 625.108 1481.67 Q622.052 1477.06 622.052 1468.33 Q622.052 1459.58 625.108 1455 Q628.186 1450.39 633.997 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M664.158 1454.1 Q660.547 1454.1 658.719 1457.66 Q656.913 1461.2 656.913 1468.33 Q656.913 1475.44 658.719 1479.01 Q660.547 1482.55 664.158 1482.55 Q667.793 1482.55 669.598 1479.01 Q671.427 1475.44 671.427 1468.33 Q671.427 1461.2 669.598 1457.66 Q667.793 1454.1 664.158 1454.1 M664.158 1450.39 Q669.969 1450.39 673.024 1455 Q676.103 1459.58 676.103 1468.33 Q676.103 1477.06 673.024 1481.67 Q669.969 1486.25 664.158 1486.25 Q658.348 1486.25 655.27 1481.67 Q652.214 1477.06 652.214 1468.33 Q652.214 1459.58 655.27 1455 Q658.348 1450.39 664.158 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M801.49 1481.64 L809.129 1481.64 L809.129 1455.28 L800.818 1456.95 L800.818 1452.69 L809.082 1451.02 L813.758 1451.02 L813.758 1481.64 L821.397 1481.64 L821.397 1485.58 L801.49 1485.58 L801.49 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M830.888 1451.02 L849.244 1451.02 L849.244 1454.96 L835.17 1454.96 L835.17 1463.43 Q836.189 1463.08 837.207 1462.92 Q838.226 1462.73 839.244 1462.73 Q845.031 1462.73 848.411 1465.9 Q851.79 1469.08 851.79 1474.49 Q851.79 1480.07 848.318 1483.17 Q844.846 1486.25 838.527 1486.25 Q836.351 1486.25 834.082 1485.88 Q831.837 1485.51 829.429 1484.77 L829.429 1480.07 Q831.513 1481.2 833.735 1481.76 Q835.957 1482.32 838.434 1482.32 Q842.439 1482.32 844.777 1480.21 Q847.115 1478.1 847.115 1474.49 Q847.115 1470.88 844.777 1468.77 Q842.439 1466.67 838.434 1466.67 Q836.559 1466.67 834.684 1467.08 Q832.832 1467.5 830.888 1468.38 L830.888 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M871.003 1454.1 Q867.392 1454.1 865.564 1457.66 Q863.758 1461.2 863.758 1468.33 Q863.758 1475.44 865.564 1479.01 Q867.392 1482.55 871.003 1482.55 Q874.638 1482.55 876.443 1479.01 Q878.272 1475.44 878.272 1468.33 Q878.272 1461.2 876.443 1457.66 Q874.638 1454.1 871.003 1454.1 M871.003 1450.39 Q876.813 1450.39 879.869 1455 Q882.948 1459.58 882.948 1468.33 Q882.948 1477.06 879.869 1481.67 Q876.813 1486.25 871.003 1486.25 Q865.193 1486.25 862.114 1481.67 Q859.059 1477.06 859.059 1468.33 Q859.059 1459.58 862.114 1455 Q865.193 1450.39 871.003 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M901.165 1454.1 Q897.554 1454.1 895.725 1457.66 Q893.92 1461.2 893.92 1468.33 Q893.92 1475.44 895.725 1479.01 Q897.554 1482.55 901.165 1482.55 Q904.799 1482.55 906.605 1479.01 Q908.434 1475.44 908.434 1468.33 Q908.434 1461.2 906.605 1457.66 Q904.799 1454.1 901.165 1454.1 M901.165 1450.39 Q906.975 1450.39 910.031 1455 Q913.11 1459.58 913.11 1468.33 Q913.11 1477.06 910.031 1481.67 Q906.975 1486.25 901.165 1486.25 Q895.355 1486.25 892.276 1481.67 Q889.221 1477.06 889.221 1468.33 Q889.221 1459.58 892.276 1455 Q895.355 1450.39 901.165 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1042.58 1481.64 L1058.9 1481.64 L1058.9 1485.58 L1036.96 1485.58 L1036.96 1481.64 Q1039.62 1478.89 1044.2 1474.26 Q1048.81 1469.61 1049.99 1468.27 Q1052.23 1465.74 1053.11 1464.01 Q1054.02 1462.25 1054.02 1460.56 Q1054.02 1457.8 1052.07 1456.07 Q1050.15 1454.33 1047.05 1454.33 Q1044.85 1454.33 1042.4 1455.09 Q1039.97 1455.86 1037.19 1457.41 L1037.19 1452.69 Q1040.01 1451.55 1042.47 1450.97 Q1044.92 1450.39 1046.96 1450.39 Q1052.33 1450.39 1055.52 1453.08 Q1058.72 1455.77 1058.72 1460.26 Q1058.72 1462.39 1057.91 1464.31 Q1057.12 1466.2 1055.01 1468.8 Q1054.43 1469.47 1051.33 1472.69 Q1048.23 1475.88 1042.58 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1078.72 1454.1 Q1075.11 1454.1 1073.28 1457.66 Q1071.47 1461.2 1071.47 1468.33 Q1071.47 1475.44 1073.28 1479.01 Q1075.11 1482.55 1078.72 1482.55 Q1082.35 1482.55 1084.16 1479.01 Q1085.98 1475.44 1085.98 1468.33 Q1085.98 1461.2 1084.16 1457.66 Q1082.35 1454.1 1078.72 1454.1 M1078.72 1450.39 Q1084.53 1450.39 1087.58 1455 Q1090.66 1459.58 1090.66 1468.33 Q1090.66 1477.06 1087.58 1481.67 Q1084.53 1486.25 1078.72 1486.25 Q1072.91 1486.25 1069.83 1481.67 Q1066.77 1477.06 1066.77 1468.33 Q1066.77 1459.58 1069.83 1455 Q1072.91 1450.39 1078.72 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1108.88 1454.1 Q1105.27 1454.1 1103.44 1457.66 Q1101.63 1461.2 1101.63 1468.33 Q1101.63 1475.44 1103.44 1479.01 Q1105.27 1482.55 1108.88 1482.55 Q1112.51 1482.55 1114.32 1479.01 Q1116.15 1475.44 1116.15 1468.33 Q1116.15 1461.2 1114.32 1457.66 Q1112.51 1454.1 1108.88 1454.1 M1108.88 1450.39 Q1114.69 1450.39 1117.74 1455 Q1120.82 1459.58 1120.82 1468.33 Q1120.82 1477.06 1117.74 1481.67 Q1114.69 1486.25 1108.88 1486.25 Q1103.07 1486.25 1099.99 1481.67 Q1096.93 1477.06 1096.93 1468.33 Q1096.93 1459.58 1099.99 1455 Q1103.07 1450.39 1108.88 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1139.04 1454.1 Q1135.43 1454.1 1133.6 1457.66 Q1131.79 1461.2 1131.79 1468.33 Q1131.79 1475.44 1133.6 1479.01 Q1135.43 1482.55 1139.04 1482.55 Q1142.67 1482.55 1144.48 1479.01 Q1146.31 1475.44 1146.31 1468.33 Q1146.31 1461.2 1144.48 1457.66 Q1142.67 1454.1 1139.04 1454.1 M1139.04 1450.39 Q1144.85 1450.39 1147.91 1455 Q1150.98 1459.58 1150.98 1468.33 Q1150.98 1477.06 1147.91 1481.67 Q1144.85 1486.25 1139.04 1486.25 Q1133.23 1486.25 1130.15 1481.67 Q1127.1 1477.06 1127.1 1468.33 Q1127.1 1459.58 1130.15 1455 Q1133.23 1450.39 1139.04 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M589.133 1525.81 L589.133 1562.76 L596.899 1562.76 Q606.734 1562.76 611.285 1558.3 Q615.869 1553.85 615.869 1544.24 Q615.869 1534.69 611.285 1530.26 Q606.734 1525.81 596.899 1525.81 L589.133 1525.81 M582.703 1520.52 L595.912 1520.52 Q609.726 1520.52 616.187 1526.28 Q622.648 1532.01 622.648 1544.24 Q622.648 1556.52 616.155 1562.28 Q609.662 1568.04 595.912 1568.04 L582.703 1568.04 L582.703 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M663.134 1548.76 L663.134 1551.62 L636.207 1551.62 Q636.589 1557.67 639.836 1560.85 Q643.114 1564 648.939 1564 Q652.312 1564 655.463 1563.17 Q658.646 1562.35 661.765 1560.69 L661.765 1566.23 Q658.614 1567.57 655.304 1568.27 Q651.994 1568.97 648.588 1568.97 Q640.058 1568.97 635.061 1564 Q630.096 1559.04 630.096 1550.57 Q630.096 1541.82 634.807 1536.69 Q639.549 1531.54 647.57 1531.54 Q654.763 1531.54 658.933 1536.18 Q663.134 1540.8 663.134 1548.76 M657.278 1547.04 Q657.214 1542.23 654.572 1539.37 Q651.962 1536.5 647.634 1536.5 Q642.732 1536.5 639.772 1539.27 Q636.844 1542.04 636.398 1547.07 L657.278 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M700.501 1539.24 Q702.697 1535.29 705.753 1533.41 Q708.808 1531.54 712.946 1531.54 Q718.516 1531.54 721.539 1535.45 Q724.563 1539.33 724.563 1546.53 L724.563 1568.04 L718.675 1568.04 L718.675 1546.72 Q718.675 1541.59 716.861 1539.11 Q715.046 1536.63 711.323 1536.63 Q706.771 1536.63 704.129 1539.65 Q701.488 1542.68 701.488 1547.9 L701.488 1568.04 L695.599 1568.04 L695.599 1546.72 Q695.599 1541.56 693.785 1539.11 Q691.971 1536.63 688.183 1536.63 Q683.695 1536.63 681.054 1539.68 Q678.412 1542.71 678.412 1547.9 L678.412 1568.04 L672.524 1568.04 L672.524 1532.4 L678.412 1532.4 L678.412 1537.93 Q680.417 1534.66 683.218 1533.1 Q686.019 1531.54 689.87 1531.54 Q693.753 1531.54 696.459 1533.51 Q699.196 1535.48 700.501 1539.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M752.445 1550.12 Q745.347 1550.12 742.61 1551.75 Q739.873 1553.37 739.873 1557.29 Q739.873 1560.4 741.91 1562.25 Q743.979 1564.07 747.512 1564.07 Q752.381 1564.07 755.31 1560.63 Q758.27 1557.16 758.27 1551.43 L758.27 1550.12 L752.445 1550.12 M764.126 1547.71 L764.126 1568.04 L758.27 1568.04 L758.27 1562.63 Q756.264 1565.88 753.273 1567.44 Q750.281 1568.97 745.952 1568.97 Q740.477 1568.97 737.231 1565.91 Q734.016 1562.82 734.016 1557.67 Q734.016 1551.65 738.027 1548.6 Q742.069 1545.54 750.058 1545.54 L758.27 1545.54 L758.27 1544.97 Q758.27 1540.93 755.596 1538.73 Q752.954 1536.5 748.148 1536.5 Q745.093 1536.5 742.196 1537.23 Q739.3 1537.97 736.626 1539.43 L736.626 1534.02 Q739.841 1532.78 742.865 1532.17 Q745.888 1531.54 748.753 1531.54 Q756.487 1531.54 760.307 1535.55 Q764.126 1539.56 764.126 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M805.821 1546.53 L805.821 1568.04 L799.965 1568.04 L799.965 1546.72 Q799.965 1541.66 797.992 1539.14 Q796.018 1536.63 792.072 1536.63 Q787.329 1536.63 784.592 1539.65 Q781.855 1542.68 781.855 1547.9 L781.855 1568.04 L775.966 1568.04 L775.966 1532.4 L781.855 1532.4 L781.855 1537.93 Q783.955 1534.72 786.788 1533.13 Q789.653 1531.54 793.376 1531.54 Q799.519 1531.54 802.67 1535.36 Q805.821 1539.14 805.821 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M840.96 1537.81 L840.96 1518.52 L846.817 1518.52 L846.817 1568.04 L840.96 1568.04 L840.96 1562.7 Q839.114 1565.88 836.281 1567.44 Q833.48 1568.97 829.534 1568.97 Q823.073 1568.97 818.998 1563.81 Q814.956 1558.65 814.956 1550.25 Q814.956 1541.85 818.998 1536.69 Q823.073 1531.54 829.534 1531.54 Q833.48 1531.54 836.281 1533.1 Q839.114 1534.62 840.96 1537.81 M821.004 1550.25 Q821.004 1556.71 823.645 1560.4 Q826.319 1564.07 830.966 1564.07 Q835.613 1564.07 838.287 1560.4 Q840.96 1556.71 840.96 1550.25 Q840.96 1543.79 838.287 1540.13 Q835.613 1536.44 830.966 1536.44 Q826.319 1536.44 823.645 1540.13 Q821.004 1543.79 821.004 1550.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="262.551,1386.4 1166.97,1386.4 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="262.551,1079.86 1166.97,1079.86 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="262.551,773.326 1166.97,773.326 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="262.551,466.791 1166.97,466.791 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="262.551,160.256 1166.97,160.256 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="262.551,1423.18 262.551,123.472 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="262.551,1386.4 281.449,1386.4 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="262.551,1079.86 281.449,1079.86 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="262.551,773.326 281.449,773.326 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="262.551,466.791 281.449,466.791 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="262.551,160.256 281.449,160.256 "/>
+<path clip-path="url(#clip070)" d="M214.607 1372.19 Q210.996 1372.19 209.167 1375.76 Q207.362 1379.3 207.362 1386.43 Q207.362 1393.54 209.167 1397.1 Q210.996 1400.64 214.607 1400.64 Q218.241 1400.64 220.047 1397.1 Q221.875 1393.54 221.875 1386.43 Q221.875 1379.3 220.047 1375.76 Q218.241 1372.19 214.607 1372.19 M214.607 1368.49 Q220.417 1368.49 223.473 1373.1 Q226.551 1377.68 226.551 1386.43 Q226.551 1395.16 223.473 1399.76 Q220.417 1404.35 214.607 1404.35 Q208.797 1404.35 205.718 1399.76 Q202.662 1395.16 202.662 1386.43 Q202.662 1377.68 205.718 1373.1 Q208.797 1368.49 214.607 1368.49 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M148.311 1093.21 L164.63 1093.21 L164.63 1097.14 L142.686 1097.14 L142.686 1093.21 Q145.348 1090.45 149.931 1085.82 Q154.538 1081.17 155.718 1079.83 Q157.964 1077.3 158.843 1075.57 Q159.746 1073.81 159.746 1072.12 Q159.746 1069.36 157.802 1067.63 Q155.88 1065.89 152.778 1065.89 Q150.579 1065.89 148.126 1066.65 Q145.695 1067.42 142.917 1068.97 L142.917 1064.25 Q145.741 1063.11 148.195 1062.53 Q150.649 1061.96 152.686 1061.96 Q158.056 1061.96 161.251 1064.64 Q164.445 1067.33 164.445 1071.82 Q164.445 1073.95 163.635 1075.87 Q162.848 1077.77 160.741 1080.36 Q160.163 1081.03 157.061 1084.25 Q153.959 1087.44 148.311 1093.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M174.491 1062.58 L192.848 1062.58 L192.848 1066.52 L178.774 1066.52 L178.774 1074.99 Q179.792 1074.64 180.811 1074.48 Q181.829 1074.29 182.848 1074.29 Q188.635 1074.29 192.014 1077.47 Q195.394 1080.64 195.394 1086.05 Q195.394 1091.63 191.922 1094.73 Q188.45 1097.81 182.13 1097.81 Q179.954 1097.81 177.686 1097.44 Q175.44 1097.07 173.033 1096.33 L173.033 1091.63 Q175.116 1092.77 177.339 1093.32 Q179.561 1093.88 182.038 1093.88 Q186.042 1093.88 188.38 1091.77 Q190.718 1089.66 190.718 1086.05 Q190.718 1082.44 188.38 1080.34 Q186.042 1078.23 182.038 1078.23 Q180.163 1078.23 178.288 1078.65 Q176.436 1079.06 174.491 1079.94 L174.491 1062.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M214.607 1065.66 Q210.996 1065.66 209.167 1069.22 Q207.362 1072.77 207.362 1079.9 Q207.362 1087 209.167 1090.57 Q210.996 1094.11 214.607 1094.11 Q218.241 1094.11 220.047 1090.57 Q221.875 1087 221.875 1079.9 Q221.875 1072.77 220.047 1069.22 Q218.241 1065.66 214.607 1065.66 M214.607 1061.96 Q220.417 1061.96 223.473 1066.56 Q226.551 1071.15 226.551 1079.9 Q226.551 1088.62 223.473 1093.23 Q220.417 1097.81 214.607 1097.81 Q208.797 1097.81 205.718 1093.23 Q202.662 1088.62 202.662 1079.9 Q202.662 1071.15 205.718 1066.56 Q208.797 1061.96 214.607 1061.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M144.329 756.046 L162.686 756.046 L162.686 759.981 L148.612 759.981 L148.612 768.453 Q149.63 768.106 150.649 767.944 Q151.667 767.759 152.686 767.759 Q158.473 767.759 161.852 770.93 Q165.232 774.101 165.232 779.518 Q165.232 785.097 161.76 788.199 Q158.288 791.277 151.968 791.277 Q149.792 791.277 147.524 790.907 Q145.279 790.537 142.871 789.796 L142.871 785.097 Q144.954 786.231 147.177 786.787 Q149.399 787.342 151.876 787.342 Q155.88 787.342 158.218 785.236 Q160.556 783.129 160.556 779.518 Q160.556 775.907 158.218 773.801 Q155.88 771.694 151.876 771.694 Q150.001 771.694 148.126 772.111 Q146.274 772.527 144.329 773.407 L144.329 756.046 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M184.445 759.125 Q180.834 759.125 179.005 762.689 Q177.2 766.231 177.2 773.361 Q177.2 780.467 179.005 784.032 Q180.834 787.574 184.445 787.574 Q188.079 787.574 189.885 784.032 Q191.713 780.467 191.713 773.361 Q191.713 766.231 189.885 762.689 Q188.079 759.125 184.445 759.125 M184.445 755.421 Q190.255 755.421 193.311 760.027 Q196.389 764.611 196.389 773.361 Q196.389 782.088 193.311 786.694 Q190.255 791.277 184.445 791.277 Q178.635 791.277 175.556 786.694 Q172.501 782.088 172.501 773.361 Q172.501 764.611 175.556 760.027 Q178.635 755.421 184.445 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M214.607 759.125 Q210.996 759.125 209.167 762.689 Q207.362 766.231 207.362 773.361 Q207.362 780.467 209.167 784.032 Q210.996 787.574 214.607 787.574 Q218.241 787.574 220.047 784.032 Q221.875 780.467 221.875 773.361 Q221.875 766.231 220.047 762.689 Q218.241 759.125 214.607 759.125 M214.607 755.421 Q220.417 755.421 223.473 760.027 Q226.551 764.611 226.551 773.361 Q226.551 782.088 223.473 786.694 Q220.417 791.277 214.607 791.277 Q208.797 791.277 205.718 786.694 Q202.662 782.088 202.662 773.361 Q202.662 764.611 205.718 760.027 Q208.797 755.421 214.607 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M143.103 449.511 L165.325 449.511 L165.325 451.502 L152.778 484.071 L147.894 484.071 L159.7 453.446 L143.103 453.446 L143.103 449.511 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M174.491 449.511 L192.848 449.511 L192.848 453.446 L178.774 453.446 L178.774 461.918 Q179.792 461.571 180.811 461.409 Q181.829 461.224 182.848 461.224 Q188.635 461.224 192.014 464.395 Q195.394 467.567 195.394 472.983 Q195.394 478.562 191.922 481.664 Q188.45 484.742 182.13 484.742 Q179.954 484.742 177.686 484.372 Q175.44 484.002 173.033 483.261 L173.033 478.562 Q175.116 479.696 177.339 480.252 Q179.561 480.807 182.038 480.807 Q186.042 480.807 188.38 478.701 Q190.718 476.594 190.718 472.983 Q190.718 469.372 188.38 467.266 Q186.042 465.159 182.038 465.159 Q180.163 465.159 178.288 465.576 Q176.436 465.993 174.491 466.872 L174.491 449.511 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M214.607 452.59 Q210.996 452.59 209.167 456.155 Q207.362 459.696 207.362 466.826 Q207.362 473.932 209.167 477.497 Q210.996 481.039 214.607 481.039 Q218.241 481.039 220.047 477.497 Q221.875 473.932 221.875 466.826 Q221.875 459.696 220.047 456.155 Q218.241 452.59 214.607 452.59 M214.607 448.886 Q220.417 448.886 223.473 453.493 Q226.551 458.076 226.551 466.826 Q226.551 475.553 223.473 480.159 Q220.417 484.742 214.607 484.742 Q208.797 484.742 205.718 480.159 Q202.662 475.553 202.662 466.826 Q202.662 458.076 205.718 453.493 Q208.797 448.886 214.607 448.886 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M114.931 173.601 L122.57 173.601 L122.57 147.236 L114.26 148.902 L114.26 144.643 L122.524 142.976 L127.2 142.976 L127.2 173.601 L134.839 173.601 L134.839 177.536 L114.931 177.536 L114.931 173.601 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M154.283 146.055 Q150.672 146.055 148.843 149.62 Q147.038 153.161 147.038 160.291 Q147.038 167.397 148.843 170.962 Q150.672 174.504 154.283 174.504 Q157.917 174.504 159.723 170.962 Q161.552 167.397 161.552 160.291 Q161.552 153.161 159.723 149.62 Q157.917 146.055 154.283 146.055 M154.283 142.351 Q160.093 142.351 163.149 146.958 Q166.227 151.541 166.227 160.291 Q166.227 169.018 163.149 173.624 Q160.093 178.208 154.283 178.208 Q148.473 178.208 145.394 173.624 Q142.339 169.018 142.339 160.291 Q142.339 151.541 145.394 146.958 Q148.473 142.351 154.283 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M184.445 146.055 Q180.834 146.055 179.005 149.62 Q177.2 153.161 177.2 160.291 Q177.2 167.397 179.005 170.962 Q180.834 174.504 184.445 174.504 Q188.079 174.504 189.885 170.962 Q191.713 167.397 191.713 160.291 Q191.713 153.161 189.885 149.62 Q188.079 146.055 184.445 146.055 M184.445 142.351 Q190.255 142.351 193.311 146.958 Q196.389 151.541 196.389 160.291 Q196.389 169.018 193.311 173.624 Q190.255 178.208 184.445 178.208 Q178.635 178.208 175.556 173.624 Q172.501 169.018 172.501 160.291 Q172.501 151.541 175.556 146.958 Q178.635 142.351 184.445 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M214.607 146.055 Q210.996 146.055 209.167 149.62 Q207.362 153.161 207.362 160.291 Q207.362 167.397 209.167 170.962 Q210.996 174.504 214.607 174.504 Q218.241 174.504 220.047 170.962 Q221.875 167.397 221.875 160.291 Q221.875 153.161 220.047 149.62 Q218.241 146.055 214.607 146.055 M214.607 142.351 Q220.417 142.351 223.473 146.958 Q226.551 151.541 226.551 160.291 Q226.551 169.018 223.473 173.624 Q220.417 178.208 214.607 178.208 Q208.797 178.208 205.718 173.624 Q202.662 169.018 202.662 160.291 Q202.662 151.541 205.718 146.958 Q208.797 142.351 214.607 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M21.7677 1000.18 L58.7206 1000.18 L58.7206 992.418 Q58.7206 982.583 54.2646 978.031 Q49.8086 973.448 40.1964 973.448 Q30.6479 973.448 26.2237 978.031 Q21.7677 982.583 21.7677 992.418 L21.7677 1000.18 M16.4842 1006.61 L16.4842 993.404 Q16.4842 979.591 22.2451 973.13 Q27.9743 966.668 40.1964 966.668 Q52.4822 966.668 58.2432 973.161 Q64.0042 979.654 64.0042 993.404 L64.0042 1006.61 L16.4842 1006.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M28.3562 956.674 L28.3562 950.818 L64.0042 950.818 L64.0042 956.674 L28.3562 956.674 M14.479 956.674 L14.479 950.818 L21.895 950.818 L21.895 956.674 L14.479 956.674 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M29.4065 915.838 L34.9447 915.838 Q33.6716 918.321 33.035 920.994 Q32.3984 923.668 32.3984 926.533 Q32.3984 930.893 33.7352 933.089 Q35.072 935.254 37.7456 935.254 Q39.7826 935.254 40.9603 933.694 Q42.1061 932.134 43.1565 927.424 L43.6021 925.419 Q44.9389 919.18 47.3897 916.57 Q49.8086 913.929 54.1691 913.929 Q59.1344 913.929 62.0308 917.875 Q64.9272 921.79 64.9272 928.665 Q64.9272 931.53 64.3543 934.649 Q63.8132 937.736 62.6992 941.174 L56.6518 941.174 Q58.3387 937.927 59.198 934.776 Q60.0256 931.625 60.0256 928.538 Q60.0256 924.4 58.6251 922.172 Q57.1929 919.944 54.6147 919.944 Q52.2276 919.944 50.9545 921.567 Q49.6813 923.159 48.5037 928.601 L48.0262 930.639 Q46.8804 936.081 44.5251 938.5 Q42.138 940.919 38.0002 940.919 Q32.9713 940.919 30.2341 937.354 Q27.4968 933.79 27.4968 927.233 Q27.4968 923.986 27.9743 921.122 Q28.4517 918.257 29.4065 915.838 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M58.657 898.937 L77.5631 898.937 L77.5631 904.826 L28.3562 904.826 L28.3562 898.937 L33.7671 898.937 Q30.5842 897.091 29.0564 894.29 Q27.4968 891.458 27.4968 887.543 Q27.4968 881.05 32.6531 877.007 Q37.8093 872.933 46.212 872.933 Q54.6147 872.933 59.771 877.007 Q64.9272 881.05 64.9272 887.543 Q64.9272 891.458 63.3994 894.29 Q61.8398 897.091 58.657 898.937 M46.212 879.013 Q39.7508 879.013 36.0905 881.686 Q32.3984 884.328 32.3984 888.975 Q32.3984 893.622 36.0905 896.296 Q39.7508 898.937 46.212 898.937 Q52.6732 898.937 56.3653 896.296 Q60.0256 893.622 60.0256 888.975 Q60.0256 884.328 56.3653 881.686 Q52.6732 879.013 46.212 879.013 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M46.0847 847.025 Q46.0847 854.123 47.7079 856.86 Q49.3312 859.597 53.2461 859.597 Q56.3653 859.597 58.2114 857.56 Q60.0256 855.491 60.0256 851.958 Q60.0256 847.089 56.5881 844.16 Q53.1188 841.2 47.3897 841.2 L46.0847 841.2 L46.0847 847.025 M43.6657 835.344 L64.0042 835.344 L64.0042 841.2 L58.5933 841.2 Q61.8398 843.206 63.3994 846.197 Q64.9272 849.189 64.9272 853.518 Q64.9272 858.992 61.8716 862.239 Q58.7843 865.454 53.6281 865.454 Q47.6125 865.454 44.5569 861.443 Q41.5014 857.401 41.5014 849.412 L41.5014 841.2 L40.9285 841.2 Q36.8862 841.2 34.6901 843.874 Q32.4621 846.516 32.4621 851.322 Q32.4621 854.377 33.1941 857.274 Q33.9262 860.17 35.3903 862.844 L29.9795 862.844 Q28.7381 859.629 28.1334 856.605 Q27.4968 853.582 27.4968 850.717 Q27.4968 842.983 31.5072 839.163 Q35.5176 835.344 43.6657 835.344 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M18.2347 817.488 L28.3562 817.488 L28.3562 805.425 L32.9077 805.425 L32.9077 817.488 L52.2594 817.488 Q56.6199 817.488 57.8613 816.31 Q59.1026 815.101 59.1026 811.441 L59.1026 805.425 L64.0042 805.425 L64.0042 811.441 Q64.0042 818.22 61.4897 820.798 Q58.9434 823.376 52.2594 823.376 L32.9077 823.376 L32.9077 827.673 L28.3562 827.673 L28.3562 823.376 L18.2347 823.376 L18.2347 817.488 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M29.7248 772.069 L35.1993 772.069 Q33.8307 774.551 33.1623 777.066 Q32.4621 779.548 32.4621 782.095 Q32.4621 787.792 36.0905 790.943 Q39.6872 794.094 46.212 794.094 Q52.7369 794.094 56.3653 790.943 Q59.9619 787.792 59.9619 782.095 Q59.9619 779.548 59.2935 777.066 Q58.5933 774.551 57.2247 772.069 L62.6355 772.069 Q63.7814 774.52 64.3543 777.161 Q64.9272 779.771 64.9272 782.731 Q64.9272 790.784 59.8664 795.526 Q54.8057 800.269 46.212 800.269 Q37.491 800.269 32.4939 795.495 Q27.4968 790.688 27.4968 782.349 Q27.4968 779.644 28.0697 777.066 Q28.6108 774.488 29.7248 772.069 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M42.4881 732.251 L64.0042 732.251 L64.0042 738.108 L42.679 738.108 Q37.6183 738.108 35.1038 740.081 Q32.5894 742.054 32.5894 746.001 Q32.5894 750.744 35.6131 753.481 Q38.6368 756.218 43.8567 756.218 L64.0042 756.218 L64.0042 762.106 L14.479 762.106 L14.479 756.218 L33.8944 756.218 Q30.6797 754.117 29.0883 751.285 Q27.4968 748.42 27.4968 744.696 Q27.4968 738.553 31.3163 735.402 Q35.1038 732.251 42.4881 732.251 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M14.479 700.391 L14.479 686.896 L19.0304 686.896 L19.0304 694.534 L68.0464 694.534 L68.0464 686.896 L72.5979 686.896 L72.5979 700.391 L14.479 700.391 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M16.4842 674.164 L16.4842 664.584 L48.8219 652.457 L16.4842 640.267 L16.4842 630.686 L64.0042 630.686 L64.0042 636.957 L22.277 636.957 L54.8694 649.211 L54.8694 655.672 L22.277 667.926 L64.0042 667.926 L64.0042 674.164 L16.4842 674.164 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M16.4842 622.156 L16.4842 615.663 L56.6518 605.669 L16.4842 595.707 L16.4842 588.482 L56.6518 578.488 L16.4842 568.525 L16.4842 562 L64.0042 573.936 L64.0042 582.021 L22.7544 592.047 L64.0042 602.168 L64.0042 610.252 L16.4842 622.156 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M14.479 540.039 L72.5979 540.039 L72.5979 553.534 L68.0464 553.534 L68.0464 545.927 L19.0304 545.927 L19.0304 553.534 L14.479 553.534 L14.479 540.039 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M351.961 12.096 L403.124 12.096 L403.124 18.9825 L381.654 18.9825 L381.654 72.576 L373.431 72.576 L373.431 18.9825 L351.961 18.9825 L351.961 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M448.413 45.1919 L448.413 72.576 L440.959 72.576 L440.959 45.4349 Q440.959 38.994 438.448 35.7938 Q435.936 32.5936 430.913 32.5936 Q424.877 32.5936 421.394 36.4419 Q417.91 40.2903 417.91 46.9338 L417.91 72.576 L410.416 72.576 L410.416 9.54393 L417.91 9.54393 L417.91 34.2544 Q420.583 30.163 424.189 28.1376 Q427.834 26.1121 432.574 26.1121 Q440.392 26.1121 444.403 30.9732 Q448.413 35.7938 448.413 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M502.088 48.0275 L502.088 51.6733 L467.817 51.6733 Q468.303 59.3701 472.435 63.421 Q476.607 67.4314 484.021 67.4314 Q488.314 67.4314 492.325 66.3781 Q496.376 65.3249 500.346 63.2184 L500.346 70.267 Q496.335 71.9684 492.122 72.8596 Q487.909 73.7508 483.575 73.7508 Q472.718 73.7508 466.359 67.4314 Q460.039 61.1119 460.039 50.3365 Q460.039 39.1965 466.034 32.6746 Q472.07 26.1121 482.279 26.1121 Q491.434 26.1121 496.74 32.0264 Q502.088 37.9003 502.088 48.0275 M494.634 45.84 Q494.553 39.7232 491.191 36.0774 Q487.869 32.4315 482.36 32.4315 Q476.121 32.4315 472.354 35.9558 Q468.627 39.4801 468.06 45.8805 L494.634 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M540.612 34.1734 Q539.356 33.4443 537.857 33.1202 Q536.399 32.7556 534.616 32.7556 Q528.297 32.7556 524.894 36.8875 Q521.532 40.9789 521.532 48.6757 L521.532 72.576 L514.038 72.576 L514.038 27.2059 L521.532 27.2059 L521.532 34.2544 Q523.881 30.1225 527.649 28.1376 Q531.416 26.1121 536.804 26.1121 Q537.573 26.1121 538.505 26.2337 Q539.437 26.3147 540.571 26.5172 L540.612 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M582.295 35.9153 Q585.091 30.8922 588.979 28.5022 Q592.868 26.1121 598.134 26.1121 Q605.224 26.1121 609.072 31.0947 Q612.92 36.0368 612.92 45.1919 L612.92 72.576 L605.426 72.576 L605.426 45.4349 Q605.426 38.913 603.117 35.7533 Q600.808 32.5936 596.069 32.5936 Q590.276 32.5936 586.913 36.4419 Q583.551 40.2903 583.551 46.9338 L583.551 72.576 L576.057 72.576 L576.057 45.4349 Q576.057 38.8725 573.748 35.7533 Q571.439 32.5936 566.618 32.5936 Q560.907 32.5936 557.544 36.4824 Q554.182 40.3308 554.182 46.9338 L554.182 72.576 L546.688 72.576 L546.688 27.2059 L554.182 27.2059 L554.182 34.2544 Q556.734 30.082 560.299 28.0971 Q563.864 26.1121 568.765 26.1121 Q573.708 26.1121 577.151 28.6237 Q580.635 31.1352 582.295 35.9153 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M648.406 49.7694 Q639.373 49.7694 635.889 51.8354 Q632.405 53.9013 632.405 58.8839 Q632.405 62.8538 634.998 65.2034 Q637.631 67.5124 642.127 67.5124 Q648.325 67.5124 652.052 63.1374 Q655.819 58.7219 655.819 51.4303 L655.819 49.7694 L648.406 49.7694 M663.273 46.6907 L663.273 72.576 L655.819 72.576 L655.819 65.6895 Q653.267 69.8214 649.459 71.8063 Q645.652 73.7508 640.142 73.7508 Q633.175 73.7508 629.043 69.8619 Q624.951 65.9325 624.951 59.3701 Q624.951 51.7138 630.056 47.825 Q635.2 43.9361 645.368 43.9361 L655.819 43.9361 L655.819 43.2069 Q655.819 38.0623 652.417 35.2672 Q649.054 32.4315 642.937 32.4315 Q639.049 32.4315 635.362 33.3632 Q631.676 34.295 628.273 36.1584 L628.273 29.2718 Q632.365 27.692 636.213 26.9223 Q640.061 26.1121 643.707 26.1121 Q653.551 26.1121 658.412 31.2163 Q663.273 36.3204 663.273 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M678.626 9.54393 L686.08 9.54393 L686.08 72.576 L678.626 72.576 L678.626 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M736.554 18.8205 L736.554 65.8515 L746.438 65.8515 Q758.955 65.8515 764.748 60.1802 Q770.581 54.509 770.581 42.2752 Q770.581 30.1225 764.748 24.4918 Q758.955 18.8205 746.438 18.8205 L736.554 18.8205 M728.371 12.096 L745.182 12.096 Q762.763 12.096 770.987 19.4281 Q779.21 26.7198 779.21 42.2752 Q779.21 57.9117 770.946 65.2439 Q762.682 72.576 745.182 72.576 L728.371 72.576 L728.371 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M791.93 27.2059 L799.383 27.2059 L799.383 72.576 L791.93 72.576 L791.93 27.2059 M791.93 9.54393 L799.383 9.54393 L799.383 18.9825 L791.93 18.9825 L791.93 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M843.903 28.5427 L843.903 35.5912 Q840.743 33.9709 837.34 33.1607 Q833.938 32.3505 830.292 32.3505 Q824.742 32.3505 821.947 34.0519 Q819.192 35.7533 819.192 39.156 Q819.192 41.7486 821.177 43.2475 Q823.162 44.7058 829.158 46.0426 L831.71 46.6097 Q839.649 48.3111 842.971 51.4303 Q846.333 54.509 846.333 60.0587 Q846.333 66.3781 841.31 70.0644 Q836.328 73.7508 827.578 73.7508 Q823.932 73.7508 819.962 73.0216 Q816.033 72.3329 811.658 70.9151 L811.658 63.2184 Q815.79 65.3654 819.8 66.4591 Q823.81 67.5124 827.74 67.5124 Q833.006 67.5124 835.842 65.73 Q838.677 63.9071 838.677 60.6258 Q838.677 57.5877 836.611 55.9673 Q834.586 54.3469 827.659 52.8481 L825.066 52.2405 Q818.139 50.7821 815.06 47.7845 Q811.982 44.7463 811.982 39.4801 Q811.982 33.0797 816.519 29.5959 Q821.056 26.1121 829.401 26.1121 Q833.533 26.1121 837.178 26.7198 Q840.824 27.3274 843.903 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M865.413 65.7705 L865.413 89.8329 L857.919 89.8329 L857.919 27.2059 L865.413 27.2059 L865.413 34.0924 Q867.763 30.0415 871.327 28.0971 Q874.933 26.1121 879.915 26.1121 Q888.179 26.1121 893.324 32.6746 Q898.509 39.2371 898.509 49.9314 Q898.509 60.6258 893.324 67.1883 Q888.179 73.7508 879.915 73.7508 Q874.933 73.7508 871.327 71.8063 Q867.763 69.8214 865.413 65.7705 M890.772 49.9314 Q890.772 41.7081 887.369 37.0496 Q884.007 32.3505 878.092 32.3505 Q872.178 32.3505 868.775 37.0496 Q865.413 41.7081 865.413 49.9314 Q865.413 58.1548 868.775 62.8538 Q872.178 67.5124 878.092 67.5124 Q884.007 67.5124 887.369 62.8538 Q890.772 58.1548 890.772 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M931.483 49.7694 Q922.45 49.7694 918.966 51.8354 Q915.482 53.9013 915.482 58.8839 Q915.482 62.8538 918.075 65.2034 Q920.708 67.5124 925.204 67.5124 Q931.402 67.5124 935.129 63.1374 Q938.897 58.7219 938.897 51.4303 L938.897 49.7694 L931.483 49.7694 M946.35 46.6907 L946.35 72.576 L938.897 72.576 L938.897 65.6895 Q936.344 69.8214 932.537 71.8063 Q928.729 73.7508 923.22 73.7508 Q916.252 73.7508 912.12 69.8619 Q908.029 65.9325 908.029 59.3701 Q908.029 51.7138 913.133 47.825 Q918.277 43.9361 928.445 43.9361 L938.897 43.9361 L938.897 43.2069 Q938.897 38.0623 935.494 35.2672 Q932.132 32.4315 926.015 32.4315 Q922.126 32.4315 918.439 33.3632 Q914.753 34.295 911.35 36.1584 L911.35 29.2718 Q915.442 27.692 919.29 26.9223 Q923.139 26.1121 926.784 26.1121 Q936.628 26.1121 941.489 31.2163 Q946.35 36.3204 946.35 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M969.076 14.324 L969.076 27.2059 L984.429 27.2059 L984.429 32.9987 L969.076 32.9987 L969.076 57.6282 Q969.076 63.1779 970.575 64.7578 Q972.114 66.3376 976.772 66.3376 L984.429 66.3376 L984.429 72.576 L976.772 72.576 Q968.144 72.576 964.863 69.3758 Q961.582 66.1351 961.582 57.6282 L961.582 32.9987 L956.113 32.9987 L956.113 27.2059 L961.582 27.2059 L961.582 14.324 L969.076 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1026.88 28.9478 L1026.88 35.9153 Q1023.72 34.1734 1020.52 33.3227 Q1017.36 32.4315 1014.12 32.4315 Q1006.87 32.4315 1002.86 37.0496 Q998.85 41.6271 998.85 49.9314 Q998.85 58.2358 1002.86 62.8538 Q1006.87 67.4314 1014.12 67.4314 Q1017.36 67.4314 1020.52 66.5807 Q1023.72 65.6895 1026.88 63.9476 L1026.88 70.8341 Q1023.76 72.2924 1020.4 73.0216 Q1017.08 73.7508 1013.31 73.7508 Q1003.06 73.7508 997.027 67.3098 Q990.991 60.8689 990.991 49.9314 Q990.991 38.832 997.068 32.472 Q1003.18 26.1121 1013.8 26.1121 Q1017.24 26.1121 1020.52 26.8413 Q1023.8 27.5299 1026.88 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1077.56 45.1919 L1077.56 72.576 L1070.11 72.576 L1070.11 45.4349 Q1070.11 38.994 1067.59 35.7938 Q1065.08 32.5936 1060.06 32.5936 Q1054.02 32.5936 1050.54 36.4419 Q1047.06 40.2903 1047.06 46.9338 L1047.06 72.576 L1039.56 72.576 L1039.56 9.54393 L1047.06 9.54393 L1047.06 34.2544 Q1049.73 30.163 1053.33 28.1376 Q1056.98 26.1121 1061.72 26.1121 Q1069.54 26.1121 1073.55 30.9732 Q1077.56 35.7938 1077.56 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip072)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="288.148,1386.4 359.25,1202.47 430.352,1018.55 501.454,834.633 572.556,650.712 643.658,466.791 714.76,282.87 785.862,344.177 856.964,160.256 928.066,160.256 999.168,160.256 1070.27,160.256 1141.37,160.256 "/>
+<polyline clip-path="url(#clip072)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="288.148,1018.55 359.25,1018.55 430.352,1018.55 501.454,1018.55 572.556,1018.55 643.658,1018.55 714.76,1018.55 785.862,1018.55 856.964,1018.55 928.066,834.633 999.168,650.712 1070.27,466.791 1141.37,282.87 "/>
+<path clip-path="url(#clip070)" d="M986.519 1379.86 L1136.82 1379.86 L1136.82 1224.34 L986.519 1224.34  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="986.519,1379.86 1136.82,1379.86 1136.82,1224.34 986.519,1224.34 986.519,1379.86 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="996.568,1276.18 1056.86,1276.18 "/>
+<path clip-path="url(#clip070)" d="M1092.47 1288.53 L1092.47 1279.24 L1084.83 1279.24 L1084.83 1275.4 L1097.1 1275.4 L1097.1 1290.24 Q1094.39 1292.16 1091.12 1293.16 Q1087.86 1294.13 1084.16 1294.13 Q1076.06 1294.13 1071.47 1289.41 Q1066.91 1284.66 1066.91 1276.21 Q1066.91 1267.74 1071.47 1263.02 Q1076.06 1258.27 1084.16 1258.27 Q1087.54 1258.27 1090.57 1259.1 Q1093.62 1259.94 1096.19 1261.56 L1096.19 1266.54 Q1093.6 1264.34 1090.68 1263.22 Q1087.77 1262.11 1084.55 1262.11 Q1078.21 1262.11 1075.01 1265.66 Q1071.84 1269.2 1071.84 1276.21 Q1071.84 1283.2 1075.01 1286.74 Q1078.21 1290.28 1084.55 1290.28 Q1087.03 1290.28 1088.97 1289.87 Q1090.92 1289.43 1092.47 1288.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1106.87 1289.52 L1114.5 1289.52 L1114.5 1263.16 L1106.19 1264.82 L1106.19 1260.56 L1114.46 1258.9 L1119.13 1258.9 L1119.13 1289.52 L1126.77 1289.52 L1126.77 1293.46 L1106.87 1293.46 L1106.87 1289.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip070)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="996.568,1328.02 1056.86,1328.02 "/>
+<path clip-path="url(#clip070)" d="M1092.47 1340.37 L1092.47 1331.08 L1084.83 1331.08 L1084.83 1327.24 L1097.1 1327.24 L1097.1 1342.08 Q1094.39 1344 1091.12 1345 Q1087.86 1345.97 1084.16 1345.97 Q1076.06 1345.97 1071.47 1341.25 Q1066.91 1336.5 1066.91 1328.05 Q1066.91 1319.58 1071.47 1314.86 Q1076.06 1310.11 1084.16 1310.11 Q1087.54 1310.11 1090.57 1310.94 Q1093.62 1311.78 1096.19 1313.4 L1096.19 1318.38 Q1093.6 1316.18 1090.68 1315.06 Q1087.77 1313.95 1084.55 1313.95 Q1078.21 1313.95 1075.01 1317.5 Q1071.84 1321.04 1071.84 1328.05 Q1071.84 1335.04 1075.01 1338.58 Q1078.21 1342.12 1084.55 1342.12 Q1087.03 1342.12 1088.97 1341.71 Q1090.92 1341.27 1092.47 1340.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1110.08 1341.36 L1126.4 1341.36 L1126.4 1345.3 L1104.46 1345.3 L1104.46 1341.36 Q1107.12 1338.61 1111.7 1333.98 Q1116.31 1329.32 1117.49 1327.98 Q1119.74 1325.46 1120.62 1323.72 Q1121.52 1321.96 1121.52 1320.27 Q1121.52 1317.52 1119.57 1315.78 Q1117.65 1314.05 1114.55 1314.05 Q1112.35 1314.05 1109.9 1314.81 Q1107.47 1315.57 1104.69 1317.13 L1104.69 1312.4 Q1107.51 1311.27 1109.97 1310.69 Q1112.42 1310.11 1114.46 1310.11 Q1119.83 1310.11 1123.02 1312.8 Q1126.22 1315.48 1126.22 1319.97 Q1126.22 1322.1 1125.41 1324.02 Q1124.62 1325.92 1122.51 1328.51 Q1121.93 1329.19 1118.83 1332.4 Q1115.73 1335.6 1110.08 1341.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1448.34 1423.18 L2352.76 1423.18 L2352.76 123.472 L1448.34 123.472  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip073">
+    <rect x="1448" y="123" width="905" height="1301"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1568.74,1423.18 1568.74,123.472 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1805.74,1423.18 1805.74,123.472 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2042.75,1423.18 2042.75,123.472 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2279.76,1423.18 2279.76,123.472 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.34,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1568.74,1423.18 1568.74,1404.28 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1805.74,1423.18 1805.74,1404.28 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2042.75,1423.18 2042.75,1404.28 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2279.76,1423.18 2279.76,1404.28 "/>
+<path clip-path="url(#clip070)" d="M1528.36 1451.02 L1546.71 1451.02 L1546.71 1454.96 L1532.64 1454.96 L1532.64 1463.43 Q1533.66 1463.08 1534.68 1462.92 Q1535.69 1462.73 1536.71 1462.73 Q1542.5 1462.73 1545.88 1465.9 Q1549.26 1469.08 1549.26 1474.49 Q1549.26 1480.07 1545.79 1483.17 Q1542.31 1486.25 1535.99 1486.25 Q1533.82 1486.25 1531.55 1485.88 Q1529.31 1485.51 1526.9 1484.77 L1526.9 1480.07 Q1528.98 1481.2 1531.2 1481.76 Q1533.43 1482.32 1535.9 1482.32 Q1539.91 1482.32 1542.24 1480.21 Q1544.58 1478.1 1544.58 1474.49 Q1544.58 1470.88 1542.24 1468.77 Q1539.91 1466.67 1535.9 1466.67 Q1534.03 1466.67 1532.15 1467.08 Q1530.3 1467.5 1528.36 1468.38 L1528.36 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1568.47 1454.1 Q1564.86 1454.1 1563.03 1457.66 Q1561.23 1461.2 1561.23 1468.33 Q1561.23 1475.44 1563.03 1479.01 Q1564.86 1482.55 1568.47 1482.55 Q1572.11 1482.55 1573.91 1479.01 Q1575.74 1475.44 1575.74 1468.33 Q1575.74 1461.2 1573.91 1457.66 Q1572.11 1454.1 1568.47 1454.1 M1568.47 1450.39 Q1574.28 1450.39 1577.34 1455 Q1580.42 1459.58 1580.42 1468.33 Q1580.42 1477.06 1577.34 1481.67 Q1574.28 1486.25 1568.47 1486.25 Q1562.66 1486.25 1559.58 1481.67 Q1556.53 1477.06 1556.53 1468.33 Q1556.53 1459.58 1559.58 1455 Q1562.66 1450.39 1568.47 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1598.63 1454.1 Q1595.02 1454.1 1593.19 1457.66 Q1591.39 1461.2 1591.39 1468.33 Q1591.39 1475.44 1593.19 1479.01 Q1595.02 1482.55 1598.63 1482.55 Q1602.27 1482.55 1604.07 1479.01 Q1605.9 1475.44 1605.9 1468.33 Q1605.9 1461.2 1604.07 1457.66 Q1602.27 1454.1 1598.63 1454.1 M1598.63 1450.39 Q1604.44 1450.39 1607.5 1455 Q1610.58 1459.58 1610.58 1468.33 Q1610.58 1477.06 1607.5 1481.67 Q1604.44 1486.25 1598.63 1486.25 Q1592.82 1486.25 1589.74 1481.67 Q1586.69 1477.06 1586.69 1468.33 Q1586.69 1459.58 1589.74 1455 Q1592.82 1450.39 1598.63 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1750.27 1481.64 L1757.91 1481.64 L1757.91 1455.28 L1749.6 1456.95 L1749.6 1452.69 L1757.86 1451.02 L1762.54 1451.02 L1762.54 1481.64 L1770.18 1481.64 L1770.18 1485.58 L1750.27 1485.58 L1750.27 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1789.62 1454.1 Q1786.01 1454.1 1784.18 1457.66 Q1782.38 1461.2 1782.38 1468.33 Q1782.38 1475.44 1784.18 1479.01 Q1786.01 1482.55 1789.62 1482.55 Q1793.26 1482.55 1795.06 1479.01 Q1796.89 1475.44 1796.89 1468.33 Q1796.89 1461.2 1795.06 1457.66 Q1793.26 1454.1 1789.62 1454.1 M1789.62 1450.39 Q1795.43 1450.39 1798.49 1455 Q1801.57 1459.58 1801.57 1468.33 Q1801.57 1477.06 1798.49 1481.67 Q1795.43 1486.25 1789.62 1486.25 Q1783.81 1486.25 1780.73 1481.67 Q1777.68 1477.06 1777.68 1468.33 Q1777.68 1459.58 1780.73 1455 Q1783.81 1450.39 1789.62 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1819.78 1454.1 Q1816.17 1454.1 1814.34 1457.66 Q1812.54 1461.2 1812.54 1468.33 Q1812.54 1475.44 1814.34 1479.01 Q1816.17 1482.55 1819.78 1482.55 Q1823.42 1482.55 1825.22 1479.01 Q1827.05 1475.44 1827.05 1468.33 Q1827.05 1461.2 1825.22 1457.66 Q1823.42 1454.1 1819.78 1454.1 M1819.78 1450.39 Q1825.59 1450.39 1828.65 1455 Q1831.73 1459.58 1831.73 1468.33 Q1831.73 1477.06 1828.65 1481.67 Q1825.59 1486.25 1819.78 1486.25 Q1813.97 1486.25 1810.89 1481.67 Q1807.84 1477.06 1807.84 1468.33 Q1807.84 1459.58 1810.89 1455 Q1813.97 1450.39 1819.78 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1849.95 1454.1 Q1846.33 1454.1 1844.51 1457.66 Q1842.7 1461.2 1842.7 1468.33 Q1842.7 1475.44 1844.51 1479.01 Q1846.33 1482.55 1849.95 1482.55 Q1853.58 1482.55 1855.39 1479.01 Q1857.21 1475.44 1857.21 1468.33 Q1857.21 1461.2 1855.39 1457.66 Q1853.58 1454.1 1849.95 1454.1 M1849.95 1450.39 Q1855.76 1450.39 1858.81 1455 Q1861.89 1459.58 1861.89 1468.33 Q1861.89 1477.06 1858.81 1481.67 Q1855.76 1486.25 1849.95 1486.25 Q1844.14 1486.25 1841.06 1481.67 Q1838 1477.06 1838 1468.33 Q1838 1459.58 1841.06 1455 Q1844.14 1450.39 1849.95 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1987.28 1481.64 L1994.92 1481.64 L1994.92 1455.28 L1986.61 1456.95 L1986.61 1452.69 L1994.87 1451.02 L1999.55 1451.02 L1999.55 1481.64 L2007.18 1481.64 L2007.18 1485.58 L1987.28 1485.58 L1987.28 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2016.67 1451.02 L2035.03 1451.02 L2035.03 1454.96 L2020.96 1454.96 L2020.96 1463.43 Q2021.98 1463.08 2022.99 1462.92 Q2024.01 1462.73 2025.03 1462.73 Q2030.82 1462.73 2034.2 1465.9 Q2037.58 1469.08 2037.58 1474.49 Q2037.58 1480.07 2034.11 1483.17 Q2030.63 1486.25 2024.31 1486.25 Q2022.14 1486.25 2019.87 1485.88 Q2017.62 1485.51 2015.22 1484.77 L2015.22 1480.07 Q2017.3 1481.2 2019.52 1481.76 Q2021.74 1482.32 2024.22 1482.32 Q2028.23 1482.32 2030.56 1480.21 Q2032.9 1478.1 2032.9 1474.49 Q2032.9 1470.88 2030.56 1468.77 Q2028.23 1466.67 2024.22 1466.67 Q2022.35 1466.67 2020.47 1467.08 Q2018.62 1467.5 2016.67 1468.38 L2016.67 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2056.79 1454.1 Q2053.18 1454.1 2051.35 1457.66 Q2049.55 1461.2 2049.55 1468.33 Q2049.55 1475.44 2051.35 1479.01 Q2053.18 1482.55 2056.79 1482.55 Q2060.42 1482.55 2062.23 1479.01 Q2064.06 1475.44 2064.06 1468.33 Q2064.06 1461.2 2062.23 1457.66 Q2060.42 1454.1 2056.79 1454.1 M2056.79 1450.39 Q2062.6 1450.39 2065.66 1455 Q2068.73 1459.58 2068.73 1468.33 Q2068.73 1477.06 2065.66 1481.67 Q2062.6 1486.25 2056.79 1486.25 Q2050.98 1486.25 2047.9 1481.67 Q2044.85 1477.06 2044.85 1468.33 Q2044.85 1459.58 2047.9 1455 Q2050.98 1450.39 2056.79 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2086.95 1454.1 Q2083.34 1454.1 2081.51 1457.66 Q2079.71 1461.2 2079.71 1468.33 Q2079.71 1475.44 2081.51 1479.01 Q2083.34 1482.55 2086.95 1482.55 Q2090.59 1482.55 2092.39 1479.01 Q2094.22 1475.44 2094.22 1468.33 Q2094.22 1461.2 2092.39 1457.66 Q2090.59 1454.1 2086.95 1454.1 M2086.95 1450.39 Q2092.76 1450.39 2095.82 1455 Q2098.9 1459.58 2098.9 1468.33 Q2098.9 1477.06 2095.82 1481.67 Q2092.76 1486.25 2086.95 1486.25 Q2081.14 1486.25 2078.06 1481.67 Q2075.01 1477.06 2075.01 1468.33 Q2075.01 1459.58 2078.06 1455 Q2081.14 1450.39 2086.95 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2228.37 1481.64 L2244.69 1481.64 L2244.69 1485.58 L2222.74 1485.58 L2222.74 1481.64 Q2225.41 1478.89 2229.99 1474.26 Q2234.6 1469.61 2235.78 1468.27 Q2238.02 1465.74 2238.9 1464.01 Q2239.8 1462.25 2239.8 1460.56 Q2239.8 1457.8 2237.86 1456.07 Q2235.94 1454.33 2232.84 1454.33 Q2230.64 1454.33 2228.18 1455.09 Q2225.75 1455.86 2222.98 1457.41 L2222.98 1452.69 Q2225.8 1451.55 2228.25 1450.97 Q2230.71 1450.39 2232.74 1450.39 Q2238.11 1450.39 2241.31 1453.08 Q2244.5 1455.77 2244.5 1460.26 Q2244.5 1462.39 2243.69 1464.31 Q2242.91 1466.2 2240.8 1468.8 Q2240.22 1469.47 2237.12 1472.69 Q2234.02 1475.88 2228.37 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2264.5 1454.1 Q2260.89 1454.1 2259.06 1457.66 Q2257.26 1461.2 2257.26 1468.33 Q2257.26 1475.44 2259.06 1479.01 Q2260.89 1482.55 2264.5 1482.55 Q2268.14 1482.55 2269.94 1479.01 Q2271.77 1475.44 2271.77 1468.33 Q2271.77 1461.2 2269.94 1457.66 Q2268.14 1454.1 2264.5 1454.1 M2264.5 1450.39 Q2270.31 1450.39 2273.37 1455 Q2276.45 1459.58 2276.45 1468.33 Q2276.45 1477.06 2273.37 1481.67 Q2270.31 1486.25 2264.5 1486.25 Q2258.69 1486.25 2255.61 1481.67 Q2252.56 1477.06 2252.56 1468.33 Q2252.56 1459.58 2255.61 1455 Q2258.69 1450.39 2264.5 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2294.67 1454.1 Q2291.05 1454.1 2289.23 1457.66 Q2287.42 1461.2 2287.42 1468.33 Q2287.42 1475.44 2289.23 1479.01 Q2291.05 1482.55 2294.67 1482.55 Q2298.3 1482.55 2300.1 1479.01 Q2301.93 1475.44 2301.93 1468.33 Q2301.93 1461.2 2300.1 1457.66 Q2298.3 1454.1 2294.67 1454.1 M2294.67 1450.39 Q2300.48 1450.39 2303.53 1455 Q2306.61 1459.58 2306.61 1468.33 Q2306.61 1477.06 2303.53 1481.67 Q2300.48 1486.25 2294.67 1486.25 Q2288.86 1486.25 2285.78 1481.67 Q2282.72 1477.06 2282.72 1468.33 Q2282.72 1459.58 2285.78 1455 Q2288.86 1450.39 2294.67 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2324.83 1454.1 Q2321.22 1454.1 2319.39 1457.66 Q2317.58 1461.2 2317.58 1468.33 Q2317.58 1475.44 2319.39 1479.01 Q2321.22 1482.55 2324.83 1482.55 Q2328.46 1482.55 2330.27 1479.01 Q2332.1 1475.44 2332.1 1468.33 Q2332.1 1461.2 2330.27 1457.66 Q2328.46 1454.1 2324.83 1454.1 M2324.83 1450.39 Q2330.64 1450.39 2333.69 1455 Q2336.77 1459.58 2336.77 1468.33 Q2336.77 1477.06 2333.69 1481.67 Q2330.64 1486.25 2324.83 1486.25 Q2319.02 1486.25 2315.94 1481.67 Q2312.88 1477.06 2312.88 1468.33 Q2312.88 1459.58 2315.94 1455 Q2319.02 1450.39 2324.83 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1678.62 1525.81 L1678.62 1562.76 L1686.39 1562.76 Q1696.22 1562.76 1700.78 1558.3 Q1705.36 1553.85 1705.36 1544.24 Q1705.36 1534.69 1700.78 1530.26 Q1696.22 1525.81 1686.39 1525.81 L1678.62 1525.81 M1672.19 1520.52 L1685.4 1520.52 Q1699.22 1520.52 1705.68 1526.28 Q1712.14 1532.01 1712.14 1544.24 Q1712.14 1556.52 1705.65 1562.28 Q1699.15 1568.04 1685.4 1568.04 L1672.19 1568.04 L1672.19 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1752.62 1548.76 L1752.62 1551.62 L1725.7 1551.62 Q1726.08 1557.67 1729.33 1560.85 Q1732.6 1564 1738.43 1564 Q1741.8 1564 1744.95 1563.17 Q1748.14 1562.35 1751.26 1560.69 L1751.26 1566.23 Q1748.1 1567.57 1744.79 1568.27 Q1741.48 1568.97 1738.08 1568.97 Q1729.55 1568.97 1724.55 1564 Q1719.59 1559.04 1719.59 1550.57 Q1719.59 1541.82 1724.3 1536.69 Q1729.04 1531.54 1737.06 1531.54 Q1744.25 1531.54 1748.42 1536.18 Q1752.62 1540.8 1752.62 1548.76 M1746.77 1547.04 Q1746.7 1542.23 1744.06 1539.37 Q1741.45 1536.5 1737.12 1536.5 Q1732.22 1536.5 1729.26 1539.27 Q1726.33 1542.04 1725.89 1547.07 L1746.77 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1789.99 1539.24 Q1792.19 1535.29 1795.24 1533.41 Q1798.3 1531.54 1802.44 1531.54 Q1808.01 1531.54 1811.03 1535.45 Q1814.05 1539.33 1814.05 1546.53 L1814.05 1568.04 L1808.16 1568.04 L1808.16 1546.72 Q1808.16 1541.59 1806.35 1539.11 Q1804.54 1536.63 1800.81 1536.63 Q1796.26 1536.63 1793.62 1539.65 Q1790.98 1542.68 1790.98 1547.9 L1790.98 1568.04 L1785.09 1568.04 L1785.09 1546.72 Q1785.09 1541.56 1783.27 1539.11 Q1781.46 1536.63 1777.67 1536.63 Q1773.19 1536.63 1770.54 1539.68 Q1767.9 1542.71 1767.9 1547.9 L1767.9 1568.04 L1762.01 1568.04 L1762.01 1532.4 L1767.9 1532.4 L1767.9 1537.93 Q1769.91 1534.66 1772.71 1533.1 Q1775.51 1531.54 1779.36 1531.54 Q1783.24 1531.54 1785.95 1533.51 Q1788.69 1535.48 1789.99 1539.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1841.93 1550.12 Q1834.84 1550.12 1832.1 1551.75 Q1829.36 1553.37 1829.36 1557.29 Q1829.36 1560.4 1831.4 1562.25 Q1833.47 1564.07 1837 1564.07 Q1841.87 1564.07 1844.8 1560.63 Q1847.76 1557.16 1847.76 1551.43 L1847.76 1550.12 L1841.93 1550.12 M1853.62 1547.71 L1853.62 1568.04 L1847.76 1568.04 L1847.76 1562.63 Q1845.75 1565.88 1842.76 1567.44 Q1839.77 1568.97 1835.44 1568.97 Q1829.97 1568.97 1826.72 1565.91 Q1823.51 1562.82 1823.51 1557.67 Q1823.51 1551.65 1827.52 1548.6 Q1831.56 1545.54 1839.55 1545.54 L1847.76 1545.54 L1847.76 1544.97 Q1847.76 1540.93 1845.09 1538.73 Q1842.44 1536.5 1837.64 1536.5 Q1834.58 1536.5 1831.69 1537.23 Q1828.79 1537.97 1826.12 1539.43 L1826.12 1534.02 Q1829.33 1532.78 1832.35 1532.17 Q1835.38 1531.54 1838.24 1531.54 Q1845.98 1531.54 1849.8 1535.55 Q1853.62 1539.56 1853.62 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1895.31 1546.53 L1895.31 1568.04 L1889.45 1568.04 L1889.45 1546.72 Q1889.45 1541.66 1887.48 1539.14 Q1885.51 1536.63 1881.56 1536.63 Q1876.82 1536.63 1874.08 1539.65 Q1871.34 1542.68 1871.34 1547.9 L1871.34 1568.04 L1865.46 1568.04 L1865.46 1532.4 L1871.34 1532.4 L1871.34 1537.93 Q1873.45 1534.72 1876.28 1533.13 Q1879.14 1531.54 1882.87 1531.54 Q1889.01 1531.54 1892.16 1535.36 Q1895.31 1539.14 1895.31 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1930.45 1537.81 L1930.45 1518.52 L1936.31 1518.52 L1936.31 1568.04 L1930.45 1568.04 L1930.45 1562.7 Q1928.6 1565.88 1925.77 1567.44 Q1922.97 1568.97 1919.02 1568.97 Q1912.56 1568.97 1908.49 1563.81 Q1904.45 1558.65 1904.45 1550.25 Q1904.45 1541.85 1908.49 1536.69 Q1912.56 1531.54 1919.02 1531.54 Q1922.97 1531.54 1925.77 1533.1 Q1928.6 1534.62 1930.45 1537.81 M1910.49 1550.25 Q1910.49 1556.71 1913.14 1560.4 Q1915.81 1564.07 1920.46 1564.07 Q1925.1 1564.07 1927.78 1560.4 Q1930.45 1556.71 1930.45 1550.25 Q1930.45 1543.79 1927.78 1540.13 Q1925.1 1536.44 1920.46 1536.44 Q1915.81 1536.44 1913.14 1540.13 Q1910.49 1543.79 1910.49 1550.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1968.55 1518.52 L1982.04 1518.52 L1982.04 1523.07 L1974.41 1523.07 L1974.41 1572.09 L1982.04 1572.09 L1982.04 1576.64 L1968.55 1576.64 L1968.55 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1994.78 1520.52 L2004.36 1520.52 L2016.48 1552.86 L2028.67 1520.52 L2038.25 1520.52 L2038.25 1568.04 L2031.98 1568.04 L2031.98 1526.32 L2019.73 1558.91 L2013.27 1558.91 L2001.01 1526.32 L2001.01 1568.04 L1994.78 1568.04 L1994.78 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2046.78 1520.52 L2053.28 1520.52 L2063.27 1560.69 L2073.23 1520.52 L2080.46 1520.52 L2090.45 1560.69 L2100.41 1520.52 L2106.94 1520.52 L2095 1568.04 L2086.92 1568.04 L2076.89 1526.79 L2066.77 1568.04 L2058.69 1568.04 L2046.78 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2128.9 1518.52 L2128.9 1576.64 L2115.41 1576.64 L2115.41 1572.09 L2123.01 1572.09 L2123.01 1523.07 L2115.41 1523.07 L2115.41 1518.52 L2128.9 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.34,1386.4 2352.76,1386.4 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.34,1079.86 2352.76,1079.86 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.34,773.326 2352.76,773.326 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.34,466.791 2352.76,466.791 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.34,160.256 2352.76,160.256 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.34,1423.18 1448.34,123.472 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.34,1386.4 1467.24,1386.4 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.34,1079.86 1467.24,1079.86 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.34,773.326 1467.24,773.326 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.34,466.791 1467.24,466.791 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.34,160.256 1467.24,160.256 "/>
+<path clip-path="url(#clip070)" d="M1400.39 1372.19 Q1396.78 1372.19 1394.95 1375.76 Q1393.15 1379.3 1393.15 1386.43 Q1393.15 1393.54 1394.95 1397.1 Q1396.78 1400.64 1400.39 1400.64 Q1404.03 1400.64 1405.83 1397.1 Q1407.66 1393.54 1407.66 1386.43 Q1407.66 1379.3 1405.83 1375.76 Q1404.03 1372.19 1400.39 1372.19 M1400.39 1368.49 Q1406.2 1368.49 1409.26 1373.1 Q1412.34 1377.68 1412.34 1386.43 Q1412.34 1395.16 1409.26 1399.76 Q1406.2 1404.35 1400.39 1404.35 Q1394.58 1404.35 1391.51 1399.76 Q1388.45 1395.16 1388.45 1386.43 Q1388.45 1377.68 1391.51 1373.1 Q1394.58 1368.49 1400.39 1368.49 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1360.28 1062.58 L1378.63 1062.58 L1378.63 1066.52 L1364.56 1066.52 L1364.56 1074.99 Q1365.58 1074.64 1366.6 1074.48 Q1367.62 1074.29 1368.63 1074.29 Q1374.42 1074.29 1377.8 1077.47 Q1381.18 1080.64 1381.18 1086.05 Q1381.18 1091.63 1377.71 1094.73 Q1374.24 1097.81 1367.92 1097.81 Q1365.74 1097.81 1363.47 1097.44 Q1361.23 1097.07 1358.82 1096.33 L1358.82 1091.63 Q1360.9 1092.77 1363.13 1093.32 Q1365.35 1093.88 1367.82 1093.88 Q1371.83 1093.88 1374.17 1091.77 Q1376.51 1089.66 1376.51 1086.05 Q1376.51 1082.44 1374.17 1080.34 Q1371.83 1078.23 1367.82 1078.23 Q1365.95 1078.23 1364.07 1078.65 Q1362.22 1079.06 1360.28 1079.94 L1360.28 1062.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1400.39 1065.66 Q1396.78 1065.66 1394.95 1069.22 Q1393.15 1072.77 1393.15 1079.9 Q1393.15 1087 1394.95 1090.57 Q1396.78 1094.11 1400.39 1094.11 Q1404.03 1094.11 1405.83 1090.57 Q1407.66 1087 1407.66 1079.9 Q1407.66 1072.77 1405.83 1069.22 Q1404.03 1065.66 1400.39 1065.66 M1400.39 1061.96 Q1406.2 1061.96 1409.26 1066.56 Q1412.34 1071.15 1412.34 1079.9 Q1412.34 1088.62 1409.26 1093.23 Q1406.2 1097.81 1400.39 1097.81 Q1394.58 1097.81 1391.51 1093.23 Q1388.45 1088.62 1388.45 1079.9 Q1388.45 1071.15 1391.51 1066.56 Q1394.58 1061.96 1400.39 1061.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1330.88 786.671 L1338.52 786.671 L1338.52 760.305 L1330.21 761.972 L1330.21 757.713 L1338.47 756.046 L1343.15 756.046 L1343.15 786.671 L1350.79 786.671 L1350.79 790.606 L1330.88 790.606 L1330.88 786.671 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1370.23 759.125 Q1366.62 759.125 1364.79 762.689 Q1362.99 766.231 1362.99 773.361 Q1362.99 780.467 1364.79 784.032 Q1366.62 787.574 1370.23 787.574 Q1373.87 787.574 1375.67 784.032 Q1377.5 780.467 1377.5 773.361 Q1377.5 766.231 1375.67 762.689 Q1373.87 759.125 1370.23 759.125 M1370.23 755.421 Q1376.04 755.421 1379.1 760.027 Q1382.18 764.611 1382.18 773.361 Q1382.18 782.088 1379.1 786.694 Q1376.04 791.277 1370.23 791.277 Q1364.42 791.277 1361.34 786.694 Q1358.29 782.088 1358.29 773.361 Q1358.29 764.611 1361.34 760.027 Q1364.42 755.421 1370.23 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1400.39 759.125 Q1396.78 759.125 1394.95 762.689 Q1393.15 766.231 1393.15 773.361 Q1393.15 780.467 1394.95 784.032 Q1396.78 787.574 1400.39 787.574 Q1404.03 787.574 1405.83 784.032 Q1407.66 780.467 1407.66 773.361 Q1407.66 766.231 1405.83 762.689 Q1404.03 759.125 1400.39 759.125 M1400.39 755.421 Q1406.2 755.421 1409.26 760.027 Q1412.34 764.611 1412.34 773.361 Q1412.34 782.088 1409.26 786.694 Q1406.2 791.277 1400.39 791.277 Q1394.58 791.277 1391.51 786.694 Q1388.45 782.088 1388.45 773.361 Q1388.45 764.611 1391.51 760.027 Q1394.58 755.421 1400.39 755.421 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1330.88 480.136 L1338.52 480.136 L1338.52 453.77 L1330.21 455.437 L1330.21 451.178 L1338.47 449.511 L1343.15 449.511 L1343.15 480.136 L1350.79 480.136 L1350.79 484.071 L1330.88 484.071 L1330.88 480.136 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1360.28 449.511 L1378.63 449.511 L1378.63 453.446 L1364.56 453.446 L1364.56 461.918 Q1365.58 461.571 1366.6 461.409 Q1367.62 461.224 1368.63 461.224 Q1374.42 461.224 1377.8 464.395 Q1381.18 467.567 1381.18 472.983 Q1381.18 478.562 1377.71 481.664 Q1374.24 484.742 1367.92 484.742 Q1365.74 484.742 1363.47 484.372 Q1361.23 484.002 1358.82 483.261 L1358.82 478.562 Q1360.9 479.696 1363.13 480.252 Q1365.35 480.807 1367.82 480.807 Q1371.83 480.807 1374.17 478.701 Q1376.51 476.594 1376.51 472.983 Q1376.51 469.372 1374.17 467.266 Q1371.83 465.159 1367.82 465.159 Q1365.95 465.159 1364.07 465.576 Q1362.22 465.993 1360.28 466.872 L1360.28 449.511 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1400.39 452.59 Q1396.78 452.59 1394.95 456.155 Q1393.15 459.696 1393.15 466.826 Q1393.15 473.932 1394.95 477.497 Q1396.78 481.039 1400.39 481.039 Q1404.03 481.039 1405.83 477.497 Q1407.66 473.932 1407.66 466.826 Q1407.66 459.696 1405.83 456.155 Q1404.03 452.59 1400.39 452.59 M1400.39 448.886 Q1406.2 448.886 1409.26 453.493 Q1412.34 458.076 1412.34 466.826 Q1412.34 475.553 1409.26 480.159 Q1406.2 484.742 1400.39 484.742 Q1394.58 484.742 1391.51 480.159 Q1388.45 475.553 1388.45 466.826 Q1388.45 458.076 1391.51 453.493 Q1394.58 448.886 1400.39 448.886 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1334.1 173.601 L1350.42 173.601 L1350.42 177.536 L1328.47 177.536 L1328.47 173.601 Q1331.14 170.847 1335.72 166.217 Q1340.32 161.564 1341.51 160.222 Q1343.75 157.698 1344.63 155.962 Q1345.53 154.203 1345.53 152.513 Q1345.53 149.759 1343.59 148.023 Q1341.67 146.286 1338.57 146.286 Q1336.37 146.286 1333.91 147.05 Q1331.48 147.814 1328.7 149.365 L1328.7 144.643 Q1331.53 143.509 1333.98 142.93 Q1336.44 142.351 1338.47 142.351 Q1343.84 142.351 1347.04 145.036 Q1350.23 147.722 1350.23 152.212 Q1350.23 154.342 1349.42 156.263 Q1348.63 158.161 1346.53 160.754 Q1345.95 161.425 1342.85 164.643 Q1339.75 167.837 1334.1 173.601 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1370.23 146.055 Q1366.62 146.055 1364.79 149.62 Q1362.99 153.161 1362.99 160.291 Q1362.99 167.397 1364.79 170.962 Q1366.62 174.504 1370.23 174.504 Q1373.87 174.504 1375.67 170.962 Q1377.5 167.397 1377.5 160.291 Q1377.5 153.161 1375.67 149.62 Q1373.87 146.055 1370.23 146.055 M1370.23 142.351 Q1376.04 142.351 1379.1 146.958 Q1382.18 151.541 1382.18 160.291 Q1382.18 169.018 1379.1 173.624 Q1376.04 178.208 1370.23 178.208 Q1364.42 178.208 1361.34 173.624 Q1358.29 169.018 1358.29 160.291 Q1358.29 151.541 1361.34 146.958 Q1364.42 142.351 1370.23 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1400.39 146.055 Q1396.78 146.055 1394.95 149.62 Q1393.15 153.161 1393.15 160.291 Q1393.15 167.397 1394.95 170.962 Q1396.78 174.504 1400.39 174.504 Q1404.03 174.504 1405.83 170.962 Q1407.66 167.397 1407.66 160.291 Q1407.66 153.161 1405.83 149.62 Q1404.03 146.055 1400.39 146.055 M1400.39 142.351 Q1406.2 142.351 1409.26 146.958 Q1412.34 151.541 1412.34 160.291 Q1412.34 169.018 1409.26 173.624 Q1406.2 178.208 1400.39 178.208 Q1394.58 178.208 1391.51 173.624 Q1388.45 169.018 1388.45 160.291 Q1388.45 151.541 1391.51 146.958 Q1394.58 142.351 1400.39 142.351 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1230.7 977.57 L1230.7 947.524 L1236.11 947.524 L1236.11 971.14 L1250.18 971.14 L1250.18 948.51 L1255.59 948.51 L1255.59 971.14 L1272.81 971.14 L1272.81 946.951 L1278.22 946.951 L1278.22 977.57 L1230.7 977.57 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1256.7 907.006 L1278.22 907.006 L1278.22 912.862 L1256.89 912.862 Q1251.83 912.862 1249.32 914.836 Q1246.8 916.809 1246.8 920.756 Q1246.8 925.498 1249.83 928.235 Q1252.85 930.973 1258.07 930.973 L1278.22 930.973 L1278.22 936.861 L1242.57 936.861 L1242.57 930.973 L1248.11 930.973 Q1244.89 928.872 1243.3 926.039 Q1241.71 923.175 1241.71 919.451 Q1241.71 913.308 1245.53 910.157 Q1249.32 907.006 1256.7 907.006 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1258.93 864.833 L1261.79 864.833 L1261.79 891.76 Q1267.84 891.378 1271.02 888.132 Q1274.17 884.853 1274.17 879.029 Q1274.17 875.655 1273.35 872.504 Q1272.52 869.321 1270.86 866.202 L1276.4 866.202 Q1277.74 869.353 1278.44 872.663 Q1279.14 875.973 1279.14 879.379 Q1279.14 887.909 1274.17 892.906 Q1269.21 897.871 1260.74 897.871 Q1251.99 897.871 1246.87 893.16 Q1241.71 888.418 1241.71 880.397 Q1241.71 873.204 1246.36 869.034 Q1250.97 864.833 1258.93 864.833 M1257.21 870.689 Q1252.4 870.753 1249.54 873.395 Q1246.67 876.005 1246.67 880.334 Q1246.67 885.235 1249.44 888.195 Q1252.21 891.123 1257.24 891.569 L1257.21 870.689 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1248.04 834.564 Q1247.47 835.551 1247.22 836.728 Q1246.93 837.874 1246.93 839.275 Q1246.93 844.24 1250.18 846.914 Q1253.39 849.555 1259.44 849.555 L1278.22 849.555 L1278.22 855.444 L1242.57 855.444 L1242.57 849.555 L1248.11 849.555 Q1244.86 847.709 1243.3 844.749 Q1241.71 841.789 1241.71 837.556 Q1241.71 836.951 1241.81 836.219 Q1241.87 835.487 1242.03 834.596 L1248.04 834.564 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1259.98 806.109 Q1253.61 806.109 1250.11 808.751 Q1246.61 811.361 1246.61 816.104 Q1246.61 820.814 1250.11 823.456 Q1253.61 826.066 1259.98 826.066 Q1266.31 826.066 1269.81 823.456 Q1273.32 820.814 1273.32 816.104 Q1273.32 811.361 1269.81 808.751 Q1266.31 806.109 1259.98 806.109 M1273.79 800.253 Q1282.9 800.253 1287.32 804.295 Q1291.78 808.337 1291.78 816.676 Q1291.78 819.764 1291.3 822.501 Q1290.85 825.238 1289.9 827.816 L1284.2 827.816 Q1285.6 825.238 1286.27 822.724 Q1286.94 820.209 1286.94 817.599 Q1286.94 811.839 1283.91 808.974 Q1280.92 806.109 1274.84 806.109 L1271.95 806.109 Q1275.1 807.924 1276.66 810.756 Q1278.22 813.589 1278.22 817.536 Q1278.22 824.092 1273.22 828.103 Q1268.22 832.113 1259.98 832.113 Q1251.7 832.113 1246.71 828.103 Q1241.71 824.092 1241.71 817.536 Q1241.71 813.589 1243.27 810.756 Q1244.83 807.924 1247.98 806.109 L1242.57 806.109 L1242.57 800.253 L1273.79 800.253 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1281.53 773.358 Q1287.89 775.84 1289.83 778.196 Q1291.78 780.551 1291.78 784.498 L1291.78 789.177 L1286.87 789.177 L1286.87 785.739 Q1286.87 783.32 1285.73 781.983 Q1284.58 780.647 1280.32 779.023 L1277.64 777.973 L1242.57 792.391 L1242.57 786.185 L1270.45 775.045 L1242.57 763.905 L1242.57 757.698 L1281.53 773.358 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1228.69 729.434 L1228.69 715.939 L1233.24 715.939 L1233.24 723.578 L1282.26 723.578 L1282.26 715.939 L1286.81 715.939 L1286.81 729.434 L1228.69 729.434 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1230.7 703.208 L1230.7 693.627 L1263.03 681.501 L1230.7 669.31 L1230.7 659.73 L1278.22 659.73 L1278.22 666 L1236.49 666 L1269.08 678.254 L1269.08 684.715 L1236.49 696.969 L1278.22 696.969 L1278.22 703.208 L1230.7 703.208 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1230.7 651.2 L1230.7 644.707 L1270.86 634.713 L1230.7 624.75 L1230.7 617.525 L1270.86 607.531 L1230.7 597.569 L1230.7 591.044 L1278.22 602.98 L1278.22 611.064 L1236.97 621.09 L1278.22 631.212 L1278.22 639.296 L1230.7 651.2 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1228.69 569.082 L1286.81 569.082 L1286.81 582.578 L1282.26 582.578 L1282.26 574.971 L1233.24 574.971 L1233.24 582.578 L1228.69 582.578 L1228.69 569.082 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1801.44 12.096 L1809.71 12.096 L1822.43 63.2184 L1835.1 12.096 L1844.3 12.096 L1857.02 63.2184 L1869.7 12.096 L1878 12.096 L1862.81 72.576 L1852.52 72.576 L1839.76 20.0763 L1826.88 72.576 L1816.59 72.576 L1801.44 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1886.71 27.2059 L1894.17 27.2059 L1894.17 72.576 L1886.71 72.576 L1886.71 27.2059 M1886.71 9.54393 L1894.17 9.54393 L1894.17 18.9825 L1886.71 18.9825 L1886.71 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1947.48 45.1919 L1947.48 72.576 L1940.02 72.576 L1940.02 45.4349 Q1940.02 38.994 1937.51 35.7938 Q1935 32.5936 1929.98 32.5936 Q1923.94 32.5936 1920.46 36.4419 Q1916.97 40.2903 1916.97 46.9338 L1916.97 72.576 L1909.48 72.576 L1909.48 27.2059 L1916.97 27.2059 L1916.97 34.2544 Q1919.65 30.163 1923.25 28.1376 Q1926.9 26.1121 1931.64 26.1121 Q1939.46 26.1121 1943.47 30.9732 Q1947.48 35.7938 1947.48 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M1992.2 34.0924 L1992.2 9.54393 L1999.65 9.54393 L1999.65 72.576 L1992.2 72.576 L1992.2 65.7705 Q1989.85 69.8214 1986.24 71.8063 Q1982.68 73.7508 1977.66 73.7508 Q1969.43 73.7508 1964.25 67.1883 Q1959.1 60.6258 1959.1 49.9314 Q1959.1 39.2371 1964.25 32.6746 Q1969.43 26.1121 1977.66 26.1121 Q1982.68 26.1121 1986.24 28.0971 Q1989.85 30.0415 1992.2 34.0924 M1966.8 49.9314 Q1966.8 58.1548 1970.16 62.8538 Q1973.56 67.5124 1979.48 67.5124 Q1985.39 67.5124 1988.8 62.8538 Q1992.2 58.1548 1992.2 49.9314 Q1992.2 41.7081 1988.8 37.0496 Q1985.39 32.3505 1979.48 32.3505 Q1973.56 32.3505 1970.16 37.0496 Q1966.8 41.7081 1966.8 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip073)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1473.94,1386.4 1545.04,1386.4 1616.14,1386.4 1687.24,1386.4 1758.34,1386.4 1829.45,1386.4 1900.55,1386.4 1971.65,160.256 2042.75,160.256 2113.85,160.256 2184.96,160.256 2256.06,160.256 2327.16,160.256 "/>
+<polyline clip-path="url(#clip073)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1473.94,160.256 1545.04,160.256 1616.14,160.256 1687.24,160.256 1758.34,160.256 1829.45,160.256 1900.55,160.256 1971.65,1386.4 2042.75,1386.4 2113.85,1386.4 2184.96,1386.4 2256.06,1386.4 2327.16,1386.4 "/>
+<path clip-path="url(#clip070)" d="M2032.63 1379.86 L2322.61 1379.86 L2322.61 1224.34 L2032.63 1224.34  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip070)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2032.63,1379.86 2322.61,1379.86 2322.61,1224.34 2032.63,1224.34 2032.63,1379.86 "/>
+<polyline clip-path="url(#clip070)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="2042.68,1276.18 2102.97,1276.18 "/>
+<path clip-path="url(#clip070)" d="M2117.7 1262.74 L2117.7 1289.61 L2123.35 1289.61 Q2130.5 1289.61 2133.81 1286.37 Q2137.14 1283.13 2137.14 1276.14 Q2137.14 1269.2 2133.81 1265.98 Q2130.5 1262.74 2123.35 1262.74 L2117.7 1262.74 M2113.02 1258.9 L2122.63 1258.9 Q2132.68 1258.9 2137.38 1263.09 Q2142.07 1267.25 2142.07 1276.14 Q2142.07 1285.08 2137.35 1289.27 Q2132.63 1293.46 2122.63 1293.46 L2113.02 1293.46 L2113.02 1258.9 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2149.34 1267.53 L2153.6 1267.53 L2153.6 1293.46 L2149.34 1293.46 L2149.34 1267.53 M2149.34 1257.44 L2153.6 1257.44 L2153.6 1262.83 L2149.34 1262.83 L2149.34 1257.44 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2179.04 1268.29 L2179.04 1272.32 Q2177.24 1271.4 2175.29 1270.93 Q2173.35 1270.47 2171.26 1270.47 Q2168.09 1270.47 2166.5 1271.44 Q2164.92 1272.41 2164.92 1274.36 Q2164.92 1275.84 2166.06 1276.7 Q2167.19 1277.53 2170.62 1278.29 L2172.07 1278.62 Q2176.61 1279.59 2178.51 1281.37 Q2180.43 1283.13 2180.43 1286.3 Q2180.43 1289.91 2177.56 1292.02 Q2174.71 1294.13 2169.71 1294.13 Q2167.63 1294.13 2165.36 1293.71 Q2163.12 1293.32 2160.62 1292.51 L2160.62 1288.11 Q2162.98 1289.34 2165.27 1289.96 Q2167.56 1290.56 2169.81 1290.56 Q2172.81 1290.56 2174.44 1289.54 Q2176.06 1288.5 2176.06 1286.63 Q2176.06 1284.89 2174.88 1283.97 Q2173.72 1283.04 2169.76 1282.18 L2168.28 1281.84 Q2164.32 1281 2162.56 1279.29 Q2160.8 1277.55 2160.8 1274.54 Q2160.8 1270.89 2163.39 1268.9 Q2165.99 1266.91 2170.75 1266.91 Q2173.12 1266.91 2175.2 1267.25 Q2177.28 1267.6 2179.04 1268.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2191.33 1289.57 L2191.33 1303.32 L2187.05 1303.32 L2187.05 1267.53 L2191.33 1267.53 L2191.33 1271.47 Q2192.68 1269.15 2194.71 1268.04 Q2196.77 1266.91 2199.62 1266.91 Q2204.34 1266.91 2207.28 1270.66 Q2210.25 1274.41 2210.25 1280.52 Q2210.25 1286.63 2207.28 1290.38 Q2204.34 1294.13 2199.62 1294.13 Q2196.77 1294.13 2194.71 1293.02 Q2192.68 1291.88 2191.33 1289.57 M2205.82 1280.52 Q2205.82 1275.82 2203.88 1273.16 Q2201.96 1270.47 2198.58 1270.47 Q2195.2 1270.47 2193.25 1273.16 Q2191.33 1275.82 2191.33 1280.52 Q2191.33 1285.22 2193.25 1287.9 Q2195.2 1290.56 2198.58 1290.56 Q2201.96 1290.56 2203.88 1287.9 Q2205.82 1285.22 2205.82 1280.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2229.09 1280.42 Q2223.93 1280.42 2221.94 1281.6 Q2219.94 1282.79 2219.94 1285.63 Q2219.94 1287.9 2221.43 1289.24 Q2222.93 1290.56 2225.5 1290.56 Q2229.04 1290.56 2231.17 1288.06 Q2233.32 1285.54 2233.32 1281.37 L2233.32 1280.42 L2229.09 1280.42 M2237.58 1278.66 L2237.58 1293.46 L2233.32 1293.46 L2233.32 1289.52 Q2231.87 1291.88 2229.69 1293.02 Q2227.51 1294.13 2224.37 1294.13 Q2220.38 1294.13 2218.02 1291.91 Q2215.69 1289.66 2215.69 1285.91 Q2215.69 1281.54 2218.6 1279.31 Q2221.54 1277.09 2227.35 1277.09 L2233.32 1277.09 L2233.32 1276.67 Q2233.32 1273.73 2231.38 1272.14 Q2229.46 1270.52 2225.96 1270.52 Q2223.74 1270.52 2221.63 1271.05 Q2219.53 1271.58 2217.58 1272.65 L2217.58 1268.71 Q2219.92 1267.81 2222.12 1267.37 Q2224.32 1266.91 2226.4 1266.91 Q2232.03 1266.91 2234.81 1269.82 Q2237.58 1272.74 2237.58 1278.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2250.57 1260.17 L2250.57 1267.53 L2259.34 1267.53 L2259.34 1270.84 L2250.57 1270.84 L2250.57 1284.91 Q2250.57 1288.09 2251.43 1288.99 Q2252.31 1289.89 2254.97 1289.89 L2259.34 1289.89 L2259.34 1293.46 L2254.97 1293.46 Q2250.04 1293.46 2248.16 1291.63 Q2246.29 1289.78 2246.29 1284.91 L2246.29 1270.84 L2243.16 1270.84 L2243.16 1267.53 L2246.29 1267.53 L2246.29 1260.17 L2250.57 1260.17 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2283.6 1268.53 L2283.6 1272.51 Q2281.8 1271.51 2279.97 1271.03 Q2278.16 1270.52 2276.31 1270.52 Q2272.17 1270.52 2269.87 1273.16 Q2267.58 1275.77 2267.58 1280.52 Q2267.58 1285.26 2269.87 1287.9 Q2272.17 1290.52 2276.31 1290.52 Q2278.16 1290.52 2279.97 1290.03 Q2281.8 1289.52 2283.6 1288.53 L2283.6 1292.46 Q2281.82 1293.29 2279.9 1293.71 Q2278 1294.13 2275.85 1294.13 Q2269.99 1294.13 2266.54 1290.45 Q2263.09 1286.77 2263.09 1280.52 Q2263.09 1274.17 2266.56 1270.54 Q2270.06 1266.91 2276.12 1266.91 Q2278.09 1266.91 2279.97 1267.32 Q2281.84 1267.72 2283.6 1268.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2312.56 1277.81 L2312.56 1293.46 L2308.3 1293.46 L2308.3 1277.95 Q2308.3 1274.27 2306.87 1272.44 Q2305.43 1270.61 2302.56 1270.61 Q2299.11 1270.61 2297.12 1272.81 Q2295.13 1275.01 2295.13 1278.8 L2295.13 1293.46 L2290.85 1293.46 L2290.85 1257.44 L2295.13 1257.44 L2295.13 1271.56 Q2296.66 1269.22 2298.72 1268.06 Q2300.8 1266.91 2303.51 1266.91 Q2307.98 1266.91 2310.27 1269.68 Q2312.56 1272.44 2312.56 1277.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip070)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="2042.68,1328.02 2102.97,1328.02 "/>
+<path clip-path="url(#clip070)" d="M2135.27 1311.87 L2135.27 1316.43 Q2132.61 1315.16 2130.25 1314.53 Q2127.88 1313.91 2125.69 1313.91 Q2121.87 1313.91 2119.78 1315.39 Q2117.72 1316.87 2117.72 1319.6 Q2117.72 1321.89 2119.09 1323.07 Q2120.48 1324.23 2124.32 1324.95 L2127.14 1325.53 Q2132.38 1326.52 2134.85 1329.05 Q2137.35 1331.55 2137.35 1335.76 Q2137.35 1340.78 2133.97 1343.37 Q2130.62 1345.97 2124.11 1345.97 Q2121.66 1345.97 2118.88 1345.41 Q2116.13 1344.86 2113.16 1343.77 L2113.16 1338.95 Q2116.01 1340.55 2118.74 1341.36 Q2121.47 1342.17 2124.11 1342.17 Q2128.12 1342.17 2130.29 1340.6 Q2132.47 1339.02 2132.47 1336.11 Q2132.47 1333.56 2130.89 1332.13 Q2129.34 1330.69 2125.78 1329.97 L2122.93 1329.42 Q2117.7 1328.38 2115.36 1326.15 Q2113.02 1323.93 2113.02 1319.97 Q2113.02 1315.39 2116.24 1312.75 Q2119.48 1310.11 2125.15 1310.11 Q2127.58 1310.11 2130.11 1310.55 Q2132.63 1310.99 2135.27 1311.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2148.58 1341.41 L2148.58 1355.16 L2144.3 1355.16 L2144.3 1319.37 L2148.58 1319.37 L2148.58 1323.31 Q2149.92 1320.99 2151.96 1319.88 Q2154.02 1318.75 2156.87 1318.75 Q2161.59 1318.75 2164.53 1322.5 Q2167.49 1326.25 2167.49 1332.36 Q2167.49 1338.47 2164.53 1342.22 Q2161.59 1345.97 2156.87 1345.97 Q2154.02 1345.97 2151.96 1344.86 Q2149.92 1343.72 2148.58 1341.41 M2163.07 1332.36 Q2163.07 1327.66 2161.13 1325 Q2159.2 1322.31 2155.82 1322.31 Q2152.44 1322.31 2150.5 1325 Q2148.58 1327.66 2148.58 1332.36 Q2148.58 1337.06 2150.5 1339.74 Q2152.44 1342.4 2155.82 1342.4 Q2159.2 1342.4 2161.13 1339.74 Q2163.07 1337.06 2163.07 1332.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2174.55 1319.37 L2178.81 1319.37 L2178.81 1345.3 L2174.55 1345.3 L2174.55 1319.37 M2174.55 1309.28 L2178.81 1309.28 L2178.81 1314.67 L2174.55 1314.67 L2174.55 1309.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2187.72 1309.28 L2191.98 1309.28 L2191.98 1345.3 L2187.72 1345.3 L2187.72 1309.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2200.89 1309.28 L2205.15 1309.28 L2205.15 1345.3 L2200.89 1345.3 L2200.89 1309.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2225.85 1332.26 Q2220.69 1332.26 2218.69 1333.44 Q2216.7 1334.63 2216.7 1337.47 Q2216.7 1339.74 2218.19 1341.08 Q2219.69 1342.4 2222.26 1342.4 Q2225.8 1342.4 2227.93 1339.9 Q2230.08 1337.38 2230.08 1333.21 L2230.08 1332.26 L2225.85 1332.26 M2234.34 1330.5 L2234.34 1345.3 L2230.08 1345.3 L2230.08 1341.36 Q2228.62 1343.72 2226.45 1344.86 Q2224.27 1345.97 2221.12 1345.97 Q2217.14 1345.97 2214.78 1343.75 Q2212.44 1341.5 2212.44 1337.75 Q2212.44 1333.38 2215.36 1331.15 Q2218.3 1328.93 2224.11 1328.93 L2230.08 1328.93 L2230.08 1328.51 Q2230.08 1325.57 2228.14 1323.98 Q2226.22 1322.36 2222.72 1322.36 Q2220.5 1322.36 2218.39 1322.89 Q2216.29 1323.42 2214.34 1324.49 L2214.34 1320.55 Q2216.68 1319.65 2218.88 1319.21 Q2221.08 1318.75 2223.16 1318.75 Q2228.79 1318.75 2231.56 1321.66 Q2234.34 1324.58 2234.34 1330.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2260.18 1332.03 Q2260.18 1327.4 2258.25 1324.86 Q2256.36 1322.31 2252.91 1322.31 Q2249.48 1322.31 2247.56 1324.86 Q2245.66 1327.4 2245.66 1332.03 Q2245.66 1336.64 2247.56 1339.19 Q2249.48 1341.73 2252.91 1341.73 Q2256.36 1341.73 2258.25 1339.19 Q2260.18 1336.64 2260.18 1332.03 M2264.43 1342.08 Q2264.43 1348.7 2261.5 1351.92 Q2258.56 1355.16 2252.49 1355.16 Q2250.25 1355.16 2248.25 1354.81 Q2246.26 1354.49 2244.39 1353.79 L2244.39 1349.65 Q2246.26 1350.67 2248.09 1351.15 Q2249.92 1351.64 2251.82 1351.64 Q2256.01 1351.64 2258.09 1349.44 Q2260.18 1347.26 2260.18 1342.84 L2260.18 1340.74 Q2258.86 1343.03 2256.8 1344.16 Q2254.74 1345.3 2251.87 1345.3 Q2247.1 1345.3 2244.18 1341.66 Q2241.26 1338.03 2241.26 1332.03 Q2241.26 1326.01 2244.18 1322.38 Q2247.1 1318.75 2251.87 1318.75 Q2254.74 1318.75 2256.8 1319.88 Q2258.86 1321.01 2260.18 1323.31 L2260.18 1319.37 L2264.43 1319.37 L2264.43 1342.08 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip070)" d="M2295.38 1331.27 L2295.38 1333.35 L2275.8 1333.35 Q2276.08 1337.75 2278.44 1340.06 Q2280.82 1342.36 2285.06 1342.36 Q2287.51 1342.36 2289.81 1341.75 Q2292.12 1341.15 2294.39 1339.95 L2294.39 1343.98 Q2292.1 1344.95 2289.69 1345.46 Q2287.28 1345.97 2284.81 1345.97 Q2278.6 1345.97 2274.97 1342.36 Q2271.36 1338.75 2271.36 1332.59 Q2271.36 1326.22 2274.78 1322.5 Q2278.23 1318.75 2284.06 1318.75 Q2289.3 1318.75 2292.33 1322.13 Q2295.38 1325.48 2295.38 1331.27 M2291.12 1330.02 Q2291.08 1326.52 2289.16 1324.44 Q2287.26 1322.36 2284.11 1322.36 Q2280.55 1322.36 2278.39 1324.37 Q2276.26 1326.38 2275.94 1330.04 L2291.12 1330.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/applications/power_systems/index.html b/previews/PR3525/tutorials/applications/power_systems/index.html
new file mode 100644
index 00000000000..fe636125719
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/power_systems/index.html
@@ -0,0 +1,394 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Power Systems · JuMP</title><meta name="title" content="Power Systems · JuMP"/><meta property="og:title" content="Power Systems · JuMP"/><meta property="twitter:title" content="Power Systems · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox" checked/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Power Systems</a><ul class="internal"><li><a class="tocitem" href="#Economic-dispatch"><span>Economic dispatch</span></a></li><li><a class="tocitem" href="#Economic-dispatch-with-adjustable-incremental-costs"><span>Economic dispatch with adjustable incremental costs</span></a></li><li><a class="tocitem" href="#Modifying-the-JuMP-model-in-place"><span>Modifying the JuMP model in-place</span></a></li><li><a class="tocitem" href="#Inefficient-usage-of-wind-generators"><span>Inefficient usage of wind generators</span></a></li><li><a class="tocitem" href="#Unit-commitment"><span>Unit commitment</span></a></li><li><a class="tocitem" href="#Unit-commitment-as-a-function-of-demand"><span>Unit commitment as a function of demand</span></a></li><li><a class="tocitem" href="#Nonlinear-economic-dispatch"><span>Nonlinear economic dispatch</span></a></li></ul></li><li><a class="tocitem" href="../optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../web_app/">Serving web apps</a></li><li><a class="tocitem" href="../two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Applications</a></li><li class="is-active"><a href>Power Systems</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Power Systems</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/applications/power_systems.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Power-Systems"><a class="docs-heading-anchor" href="#Power-Systems">Power Systems</a><a id="Power-Systems-1"></a><a class="docs-heading-anchor-permalink" href="#Power-Systems" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../power_systems.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Yury Dvorkin and Miles Lubin.</strong></p><p>This tutorial demonstrates how to formulate basic power systems engineering models in JuMP.</p><p>We will consider basic &quot;economic dispatch&quot; and &quot;unit commitment&quot; models without taking into account transmission constraints.</p><p>For this tutorial, we use the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import DataFrames
+import HiGHS
+import Plots
+import StatsPlots</code></pre><h2 id="Economic-dispatch"><a class="docs-heading-anchor" href="#Economic-dispatch">Economic dispatch</a><a id="Economic-dispatch-1"></a><a class="docs-heading-anchor-permalink" href="#Economic-dispatch" title="Permalink"></a></h2><p>Economic dispatch (ED) is an optimization problem that minimizes the cost of supplying energy demand subject to operational constraints on power system assets. In its simplest modification, ED is an LP problem solved for an aggregated load and wind forecast and for a single infinitesimal moment.</p><p>Mathematically, the ED problem can be written as follows:</p><p class="math-container">\[\min \sum_{i \in I} c^g_{i} \cdot g_{i} + c^w \cdot w,\]</p><p>where <span>$c_{i}$</span> and <span>$g_{i}$</span> are the incremental cost (<span>$</span>/MWh) and power output (MW) of the <span>$i^{th}$</span> generator, respectively, and <span>$c^w$</span> and <span>$w$</span> are the incremental cost (<span>$</span>/MWh) and wind power injection (MW), respectively.</p><p>Subject to the constraints:</p><ul><li>Minimum (<span>$g^{\min}$</span>) and maximum (<span>$g^{\max}$</span>) limits on power outputs of generators: <span>$g^{\min}_{i} \leq g_{i} \leq g^{\max}_{i}.$</span></li><li>Constraint on the wind power injection: <span>$0 \leq w \leq w^f,$</span> where <span>$w$</span> and <span>$w^f$</span> are the wind power injection and wind power forecast, respectively.</li><li>Power balance constraint: <span>$\sum_{i \in I} g_{i} + w = d^f,$</span> where <span>$d^f$</span> is the demand forecast.</li></ul><p>Further reading on ED models can be found in A. J. Wood, B. F. Wollenberg, and G. B. Sheblé, &quot;Power Generation, Operation and Control,&quot; Wiley, 2013.</p><p>Define some input data about the test system.</p><p>We define some thermal generators:</p><pre><code class="language-julia hljs">function ThermalGenerator(
+    min::Float64,
+    max::Float64,
+    fixed_cost::Float64,
+    variable_cost::Float64,
+)
+    return (
+        min = min,
+        max = max,
+        fixed_cost = fixed_cost,
+        variable_cost = variable_cost,
+    )
+end
+
+generators = [
+    ThermalGenerator(0.0, 1000.0, 1000.0, 50.0),
+    ThermalGenerator(300.0, 1000.0, 0.0, 100.0),
+]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{NamedTuple{(:min, :max, :fixed_cost, :variable_cost), NTuple{4, Float64}}}:
+ (min = 0.0, max = 1000.0, fixed_cost = 1000.0, variable_cost = 50.0)
+ (min = 300.0, max = 1000.0, fixed_cost = 0.0, variable_cost = 100.0)</code></pre><p>A wind generator</p><pre><code class="language-julia hljs">WindGenerator(variable_cost::Float64) = (variable_cost = variable_cost,)
+
+wind_generator = WindGenerator(50.0)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(variable_cost = 50.0,)</code></pre><p>And a scenario</p><pre><code class="language-julia hljs">function Scenario(demand::Float64, wind::Float64)
+    return (demand = demand, wind = wind)
+end
+
+scenario = Scenario(1500.0, 200.0)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(demand = 1500.0, wind = 200.0)</code></pre><p>Create a function <code>solve_economic_dispatch</code>, which solves the economic dispatch problem for a given set of input parameters.</p><pre><code class="language-julia hljs">function solve_economic_dispatch(generators::Vector, wind, scenario)
+    # Define the economic dispatch (ED) model
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    # Define decision variables
+    # power output of generators
+    N = length(generators)
+    @variable(model, generators[i].min &lt;= g[i = 1:N] &lt;= generators[i].max)
+    # wind power injection
+    @variable(model, 0 &lt;= w &lt;= scenario.wind)
+    # Define the objective function
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * g[i] for i in 1:N) +
+        wind.variable_cost * w,
+    )
+    # Define the power balance constraint
+    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)
+    # Solve statement
+    optimize!(model)
+    # return the optimal value of the objective function and its minimizers
+    return (
+        g = value.(g),
+        w = value(w),
+        wind_spill = scenario.wind - value(w),
+        total_cost = objective_value(model),
+    )
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_economic_dispatch (generic function with 1 method)</code></pre><p>Solve the economic dispatch problem</p><pre><code class="language-julia hljs">solution = solve_economic_dispatch(generators, wind_generator, scenario);
+
+println(&quot;Dispatch of Generators: &quot;, solution.g, &quot; MW&quot;)
+println(&quot;Dispatch of Wind: &quot;, solution.w, &quot; MW&quot;)
+println(&quot;Wind spillage: &quot;, solution.wind_spill, &quot; MW&quot;)
+println(&quot;Total cost: \$&quot;, solution.total_cost)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dispatch of Generators: [1000.0, 300.0] MW
+Dispatch of Wind: 200.0 MW
+Wind spillage: 0.0 MW
+Total cost: $90000.0</code></pre><h2 id="Economic-dispatch-with-adjustable-incremental-costs"><a class="docs-heading-anchor" href="#Economic-dispatch-with-adjustable-incremental-costs">Economic dispatch with adjustable incremental costs</a><a id="Economic-dispatch-with-adjustable-incremental-costs-1"></a><a class="docs-heading-anchor-permalink" href="#Economic-dispatch-with-adjustable-incremental-costs" title="Permalink"></a></h2><p>In the following exercise we adjust the incremental cost of generator G1 and observe its impact on the total cost.</p><pre><code class="language-julia hljs">function scale_generator_cost(g, scale)
+    return ThermalGenerator(g.min, g.max, g.fixed_cost, scale * g.variable_cost)
+end
+
+start = time()
+c_g_scale_df = DataFrames.DataFrame(;
+    # Scale factor
+    scale = Float64[],
+    # Dispatch of Generator 1 [MW]
+    dispatch_G1 = Float64[],
+    # Dispatch of Generator 2 [MW]
+    dispatch_G2 = Float64[],
+    # Dispatch of Wind [MW]
+    dispatch_wind = Float64[],
+    # Spillage of Wind [MW]
+    spillage_wind = Float64[],
+    # Total cost [$]
+    total_cost = Float64[],
+)
+for c_g1_scale in 0.5:0.1:3.0
+    # Update the incremental cost of the first generator at every iteration.
+    new_generators = scale_generator_cost.(generators, [c_g1_scale, 1.0])
+    # Solve the economic-dispatch problem with the updated incremental cost
+    sol = solve_economic_dispatch(new_generators, wind_generator, scenario)
+    push!(
+        c_g_scale_df,
+        (c_g1_scale, sol.g[1], sol.g[2], sol.w, sol.wind_spill, sol.total_cost),
+    )
+end
+print(string(&quot;elapsed time: &quot;, time() - start, &quot; seconds&quot;))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">elapsed time: 0.13164806365966797 seconds</code></pre><pre><code class="language-julia hljs">c_g_scale_df</code></pre><div><div style = "float: left;"><span>26×6 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">scale</th><th style = "text-align: left;">dispatch_G1</th><th style = "text-align: left;">dispatch_G2</th><th style = "text-align: left;">dispatch_wind</th><th style = "text-align: left;">spillage_wind</th><th style = "text-align: left;">total_cost</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">0.5</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">65000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">0.6</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">70000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">0.7</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">75000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">0.8</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">80000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">0.9</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">85000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">90000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">1.1</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">95000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">1.2</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">100000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">1.3</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">105000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">1.4</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">110000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">1.5</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">115000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">1.6</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">120000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: right;">1.7</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">125000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: right;">1.8</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">130000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: right;">1.9</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">135000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: right;">2.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">140000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: right;">2.1</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">141500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: right;">2.2</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">143000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: right;">2.3</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">144500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: right;">2.4</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">146000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">21</td><td style = "text-align: right;">2.5</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">147500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">22</td><td style = "text-align: right;">2.6</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">149000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">23</td><td style = "text-align: right;">2.7</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">150500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">24</td><td style = "text-align: right;">2.8</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">152000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">25</td><td style = "text-align: right;">2.9</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">153500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">26</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">155000.0</td></tr></tbody></table></div><h2 id="Modifying-the-JuMP-model-in-place"><a class="docs-heading-anchor" href="#Modifying-the-JuMP-model-in-place">Modifying the JuMP model in-place</a><a id="Modifying-the-JuMP-model-in-place-1"></a><a class="docs-heading-anchor-permalink" href="#Modifying-the-JuMP-model-in-place" title="Permalink"></a></h2><p>Note that in the previous exercise we entirely rebuilt the optimization model at every iteration of the internal loop, which incurs an additional computational burden. This burden can be alleviated if instead of re-building the entire model, we modify the constraints or objective function, as it shown in the example below.</p><p>Compare the computing time in case of the above and below models.</p><pre><code class="language-julia hljs">function solve_economic_dispatch_inplace(
+    generators::Vector,
+    wind,
+    scenario,
+    scale::AbstractVector{Float64},
+)
+    obj_out = Float64[]
+    w_out = Float64[]
+    g1_out = Float64[]
+    g2_out = Float64[]
+    # This function only works for two generators
+    @assert length(generators) == 2
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    N = length(generators)
+    @variable(model, generators[i].min &lt;= g[i = 1:N] &lt;= generators[i].max)
+    @variable(model, 0 &lt;= w &lt;= scenario.wind)
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * g[i] for i in 1:N) +
+        wind.variable_cost * w,
+    )
+    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)
+    for c_g1_scale in scale
+        @objective(
+            model,
+            Min,
+            c_g1_scale * generators[1].variable_cost * g[1] +
+            generators[2].variable_cost * g[2] +
+            wind.variable_cost * w,
+        )
+        optimize!(model)
+        push!(obj_out, objective_value(model))
+        push!(w_out, value(w))
+        push!(g1_out, value(g[1]))
+        push!(g2_out, value(g[2]))
+    end
+    df = DataFrames.DataFrame(;
+        scale = scale,
+        dispatch_G1 = g1_out,
+        dispatch_G2 = g2_out,
+        dispatch_wind = w_out,
+        spillage_wind = scenario.wind .- w_out,
+        total_cost = obj_out,
+    )
+    return df
+end
+
+start = time()
+inplace_df = solve_economic_dispatch_inplace(
+    generators,
+    wind_generator,
+    scenario,
+    0.5:0.1:3.0,
+)
+print(string(&quot;elapsed time: &quot;, time() - start, &quot; seconds&quot;))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">elapsed time: 0.15134501457214355 seconds</code></pre><p>For small models, adjusting specific constraints or the objective function is sometimes faster and sometimes slower than re-building the entire model. However, as the problem size increases, updating the model in-place is usually faster.</p><pre><code class="language-julia hljs">inplace_df</code></pre><div><div style = "float: left;"><span>26×6 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">scale</th><th style = "text-align: left;">dispatch_G1</th><th style = "text-align: left;">dispatch_G2</th><th style = "text-align: left;">dispatch_wind</th><th style = "text-align: left;">spillage_wind</th><th style = "text-align: left;">total_cost</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">0.5</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">65000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">0.6</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">70000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">0.7</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">75000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">0.8</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">80000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">0.9</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">85000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">90000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">1.1</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">95000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">1.2</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">100000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">1.3</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">105000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">1.4</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">110000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">1.5</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">115000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">1.6</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">120000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: right;">1.7</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">125000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: right;">1.8</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">130000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: right;">1.9</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">135000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: right;">2.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">140000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: right;">2.1</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">141500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: right;">2.2</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">143000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: right;">2.3</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">144500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: right;">2.4</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">146000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">21</td><td style = "text-align: right;">2.5</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">147500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">22</td><td style = "text-align: right;">2.6</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">149000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">23</td><td style = "text-align: right;">2.7</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">150500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">24</td><td style = "text-align: right;">2.8</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">152000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">25</td><td style = "text-align: right;">2.9</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">153500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">26</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">155000.0</td></tr></tbody></table></div><h2 id="Inefficient-usage-of-wind-generators"><a class="docs-heading-anchor" href="#Inefficient-usage-of-wind-generators">Inefficient usage of wind generators</a><a id="Inefficient-usage-of-wind-generators-1"></a><a class="docs-heading-anchor-permalink" href="#Inefficient-usage-of-wind-generators" title="Permalink"></a></h2><p>The economic dispatch problem does not perform commitment decisions and, thus, assumes that all generators must be dispatched at least at their minimum power output limit. This approach is not cost efficient and may lead to absurd decisions. For example, if <span>$d = \sum_{i \in I} g^{\min}_{i}$</span>, the wind power injection must be zero, that is, all available wind generation is spilled, to meet the minimum power output constraints on generators.</p><p>In the following example, we adjust the total demand and observed how it affects wind spillage.</p><pre><code class="language-julia hljs">demand_scale_df = DataFrames.DataFrame(;
+    demand = Float64[],
+    dispatch_G1 = Float64[],
+    dispatch_G2 = Float64[],
+    dispatch_wind = Float64[],
+    spillage_wind = Float64[],
+    total_cost = Float64[],
+)
+
+function scale_demand(scenario, scale)
+    return Scenario(scale * scenario.demand, scenario.wind)
+end
+
+for demand_scale in 0.2:0.1:1.4
+    new_scenario = scale_demand(scenario, demand_scale)
+    sol = solve_economic_dispatch(generators, wind_generator, new_scenario)
+    push!(
+        demand_scale_df,
+        (
+            new_scenario.demand,
+            sol.g[1],
+            sol.g[2],
+            sol.w,
+            sol.wind_spill,
+            sol.total_cost,
+        ),
+    )
+end
+
+demand_scale_df</code></pre><div><div style = "float: left;"><span>13×6 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">demand</th><th style = "text-align: left;">dispatch_G1</th><th style = "text-align: left;">dispatch_G2</th><th style = "text-align: left;">dispatch_wind</th><th style = "text-align: left;">spillage_wind</th><th style = "text-align: left;">total_cost</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">30000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">450.0</td><td style = "text-align: right;">150.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">37500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">600.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">45000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">750.0</td><td style = "text-align: right;">450.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">52500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">900.0</td><td style = "text-align: right;">600.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">60000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">1050.0</td><td style = "text-align: right;">750.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">67500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">1200.0</td><td style = "text-align: right;">900.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">75000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">1350.0</td><td style = "text-align: right;">850.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">82500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">1500.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">90000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">1650.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">450.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">105000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">1800.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">600.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">120000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">1950.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">750.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">135000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: right;">2100.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">900.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">150000.0</td></tr></tbody></table></div><pre><code class="language-julia hljs">dispatch_plot = StatsPlots.@df(
+    demand_scale_df,
+    Plots.plot(
+        :demand,
+        [:dispatch_G1, :dispatch_G2],
+        labels = [&quot;G1&quot; &quot;G2&quot;],
+        title = &quot;Thermal Dispatch&quot;,
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = &quot;Demand&quot;,
+        ylabel = &quot;Dispatch [MW]&quot;,
+    ),
+)
+
+wind_plot = StatsPlots.@df(
+    demand_scale_df,
+    Plots.plot(
+        :demand,
+        [:dispatch_wind, :spillage_wind],
+        labels = [&quot;Dispatch&quot; &quot;Spillage&quot;],
+        title = &quot;Wind&quot;,
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = &quot;Demand [MW]&quot;,
+        ylabel = &quot;Energy [MW]&quot;,
+    ),
+)
+
+Plots.plot(dispatch_plot, wind_plot)</code></pre><img src="ab8a43ee.svg" alt="Example block output"/><p>This particular drawback can be overcome by introducing binary decisions on the &quot;on/off&quot; status of generators. This model is called unit commitment and considered later in these notes.</p><p>For further reading on the interplay between wind generation and the minimum power output constraints of generators, we refer interested readers to R. Baldick, &quot;Wind and energy markets: a case study of Texas,&quot; IEEE Systems Journal, vol. 6, pp. 27-34, 2012.</p><h2 id="Unit-commitment"><a class="docs-heading-anchor" href="#Unit-commitment">Unit commitment</a><a id="Unit-commitment-1"></a><a class="docs-heading-anchor-permalink" href="#Unit-commitment" title="Permalink"></a></h2><p>The Unit Commitment (UC) model can be obtained from ED model by introducing binary variable associated with each generator. This binary variable can attain two values: if it is &quot;1,&quot; the generator is synchronized and, thus, can be dispatched, otherwise, that is, if the binary variable is &quot;0,&quot; that generator is not synchronized and its power output is set to 0.</p><p>To obtain the mathematical formulation of the UC model, we will modify the constraints of the ED model as follows:</p><p class="math-container">\[g^{\min}_{i} \cdot u_{t,i} \leq g_{i} \leq g^{\max}_{i} \cdot u_{t,i},\]</p><p>where <span>$u_{i} \in \{0,1\}.$</span> In this constraint, if <span>$u_{i} = 0$</span>, then <span>$g_{i}  = 0$</span>. On the other hand, if <span>$u_{i} = 1$</span>, then <span>$g^{min}_{i} \leq g_{i} \leq g^{max}_{i}$</span>.</p><p>For further reading on the UC problem we refer interested readers to G. Morales-Espana, J. M. Latorre, and A. Ramos, &quot;Tight and Compact MILP Formulation for the Thermal Unit Commitment Problem,&quot; IEEE Transactions on Power Systems, vol. 28, pp. 4897-4908, 2013.</p><p>In the following example we convert the ED model explained above to the UC model.</p><pre><code class="language-julia hljs">function solve_unit_commitment(generators::Vector, wind, scenario)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    N = length(generators)
+    @variable(model, 0 &lt;= g[i = 1:N] &lt;= generators[i].max)
+    @variable(model, 0 &lt;= w &lt;= scenario.wind)
+    @constraint(model, sum(g[i] for i in 1:N) + w == scenario.demand)
+    # !!! New: add binary on-off variables for each generator
+    @variable(model, u[i = 1:N], Bin)
+    @constraint(model, [i = 1:N], g[i] &lt;= generators[i].max * u[i])
+    @constraint(model, [i = 1:N], g[i] &gt;= generators[i].min * u[i])
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * g[i] for i in 1:N) +
+        wind.variable_cost * w +
+        # !!! new
+        sum(generators[i].fixed_cost * u[i] for i in 1:N)
+    )
+    optimize!(model)
+    status = termination_status(model)
+    if status != OPTIMAL
+        return (status = status,)
+    end
+    return (
+        status = status,
+        g = value.(g),
+        w = value(w),
+        wind_spill = scenario.wind - value(w),
+        u = value.(u),
+        total_cost = objective_value(model),
+    )
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_unit_commitment (generic function with 1 method)</code></pre><p>Solve the unit commitment problem</p><pre><code class="language-julia hljs">solution = solve_unit_commitment(generators, wind_generator, scenario)
+
+println(&quot;Dispatch of Generators: &quot;, solution.g, &quot; MW&quot;)
+println(&quot;Commitments of Generators: &quot;, solution.u)
+println(&quot;Dispatch of Wind: &quot;, solution.w, &quot; MW&quot;)
+println(&quot;Wind spillage: &quot;, solution.wind_spill, &quot; MW&quot;)
+println(&quot;Total cost: \$&quot;, solution.total_cost)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dispatch of Generators: [1000.0, 300.0] MW
+Commitments of Generators: [1.0, 1.0]
+Dispatch of Wind: 200.0 MW
+Wind spillage: 0.0 MW
+Total cost: $91000.0</code></pre><h2 id="Unit-commitment-as-a-function-of-demand"><a class="docs-heading-anchor" href="#Unit-commitment-as-a-function-of-demand">Unit commitment as a function of demand</a><a id="Unit-commitment-as-a-function-of-demand-1"></a><a class="docs-heading-anchor-permalink" href="#Unit-commitment-as-a-function-of-demand" title="Permalink"></a></h2><p>After implementing the unit commitment model, we can now assess the interplay between the minimum power output constraints on generators and wind generation.</p><pre><code class="language-julia hljs">uc_df = DataFrames.DataFrame(;
+    demand = Float64[],
+    commitment_G1 = Float64[],
+    commitment_G2 = Float64[],
+    dispatch_G1 = Float64[],
+    dispatch_G2 = Float64[],
+    dispatch_wind = Float64[],
+    spillage_wind = Float64[],
+    total_cost = Float64[],
+)
+
+for demand_scale in 0.2:0.1:1.4
+    new_scenario = scale_demand(scenario, demand_scale)
+    sol = solve_unit_commitment(generators, wind_generator, new_scenario)
+    if sol.status == OPTIMAL
+        push!(
+            uc_df,
+            (
+                new_scenario.demand,
+                sol.u[1],
+                sol.u[2],
+                sol.g[1],
+                sol.g[2],
+                sol.w,
+                sol.wind_spill,
+                sol.total_cost,
+            ),
+        )
+    end
+    println(&quot;Status: $(sol.status) for demand_scale = $(demand_scale)&quot;)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Status: OPTIMAL for demand_scale = 0.2
+Status: OPTIMAL for demand_scale = 0.3
+Status: OPTIMAL for demand_scale = 0.4
+Status: OPTIMAL for demand_scale = 0.5
+Status: OPTIMAL for demand_scale = 0.6
+Status: OPTIMAL for demand_scale = 0.7
+Status: OPTIMAL for demand_scale = 0.8
+Status: OPTIMAL for demand_scale = 0.9
+Status: OPTIMAL for demand_scale = 1.0
+Status: OPTIMAL for demand_scale = 1.1
+Status: OPTIMAL for demand_scale = 1.2
+Status: OPTIMAL for demand_scale = 1.3
+Status: OPTIMAL for demand_scale = 1.4</code></pre><pre><code class="language-julia hljs">uc_df</code></pre><div><div style = "float: left;"><span>13×8 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">demand</th><th style = "text-align: left;">commitment_G1</th><th style = "text-align: left;">commitment_G2</th><th style = "text-align: left;">dispatch_G1</th><th style = "text-align: left;">dispatch_G2</th><th style = "text-align: left;">dispatch_wind</th><th style = "text-align: left;">spillage_wind</th><th style = "text-align: left;">total_cost</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">100.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">16000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">450.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">250.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">23500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">600.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">400.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">31000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">750.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">550.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">38500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">900.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">700.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">46000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">1050.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">850.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">53500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">1200.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">61000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">1350.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">850.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">83500.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">1500.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">300.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">91000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">1650.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">450.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">106000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">1800.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">600.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">121000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">1950.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">750.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">136000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: right;">2100.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1000.0</td><td style = "text-align: right;">900.0</td><td style = "text-align: right;">200.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">151000.0</td></tr></tbody></table></div><pre><code class="language-julia hljs">commitment_plot = StatsPlots.@df(
+    uc_df,
+    Plots.plot(
+        :demand,
+        [:commitment_G1, :commitment_G2],
+        labels = [&quot;G1&quot; &quot;G2&quot;],
+        title = &quot;Commitment&quot;,
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = &quot;Demand [MW]&quot;,
+        ylabel = &quot;Commitment decision {0, 1}&quot;,
+    ),
+)
+
+dispatch_plot = StatsPlots.@df(
+    uc_df,
+    Plots.plot(
+        :demand,
+        [:dispatch_G1, :dispatch_G2, :dispatch_wind],
+        labels = [&quot;G1&quot; &quot;G2&quot; &quot;Wind&quot;],
+        title = &quot;Dispatch [MW]&quot;,
+        legend = :bottomright,
+        linewidth = 3,
+        xlabel = &quot;Demand&quot;,
+        ylabel = &quot;Dispatch [MW]&quot;,
+    ),
+)
+
+Plots.plot(commitment_plot, dispatch_plot)</code></pre><img src="9f136f96.svg" alt="Example block output"/><h2 id="Nonlinear-economic-dispatch"><a class="docs-heading-anchor" href="#Nonlinear-economic-dispatch">Nonlinear economic dispatch</a><a id="Nonlinear-economic-dispatch-1"></a><a class="docs-heading-anchor-permalink" href="#Nonlinear-economic-dispatch" title="Permalink"></a></h2><p>As a final example, we modify our economic dispatch problem in two ways:</p><ul><li>The thermal cost function is user-defined</li><li>The output of the wind is only the square-root of the dispatch</li></ul><pre><code class="language-julia hljs">import Ipopt
+
+&quot;&quot;&quot;
+    thermal_cost_function(g)
+
+A user-defined thermal cost function in pure-Julia! You can include
+nonlinearities, and even things like control flow.
+
+!!! warning
+    It&#39;s still up to you to make sure that the function has a meaningful
+    derivative.
+&quot;&quot;&quot;
+function thermal_cost_function(g)
+    if g &lt;= 500
+        return g
+    else
+        return g + 1e-2 * (g - 500)^2
+    end
+end
+
+function solve_nonlinear_economic_dispatch(
+    generators::Vector,
+    wind,
+    scenario;
+    silent::Bool = false,
+)
+    model = Model(Ipopt.Optimizer)
+    if silent
+        set_silent(model)
+    end
+    @operator(model, op_tcf, 1, thermal_cost_function)
+    N = length(generators)
+    @variable(model, generators[i].min &lt;= g[i = 1:N] &lt;= generators[i].max)
+    @variable(model, 0 &lt;= w &lt;= scenario.wind)
+    @objective(
+        model,
+        Min,
+        sum(generators[i].variable_cost * op_tcf(g[i]) for i in 1:N) +
+        wind.variable_cost * w,
+    )
+    @constraint(model, sum(g[i] for i in 1:N) + sqrt(w) == scenario.demand)
+    optimize!(model)
+    return (
+        g = value.(g),
+        w = value(w),
+        wind_spill = scenario.wind - value(w),
+        total_cost = objective_value(model),
+    )
+end
+
+solution =
+    solve_nonlinear_economic_dispatch(generators, wind_generator, scenario)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(g = [847.3509933774712, 648.6754966887423], w = 15.78878119389903, wind_spill = 184.21121880610096, total_cost = 190455.298013245)</code></pre><p>Now let&#39;s see how the wind is dispatched as a function of the cost:</p><pre><code class="language-julia hljs">wind_cost = 0.0:1:100
+wind_dispatch = Float64[]
+for c in wind_cost
+    sol = solve_nonlinear_economic_dispatch(
+        generators,
+        WindGenerator(c),
+        scenario;
+        silent = true,
+    )
+    push!(wind_dispatch, sol.w)
+end
+
+Plots.plot(
+    wind_cost,
+    wind_dispatch;
+    xlabel = &quot;Cost&quot;,
+    ylabel = &quot;Dispatch [MW]&quot;,
+    label = false,
+)</code></pre><img src="6fee809a.svg" alt="Example block output"/></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../algorithms/parallelism/">« Parallelism</a><a class="docs-footer-nextpage" href="../optimal_power_flow/">Optimal power flow »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/applications/two_stage_stochastic.jl b/previews/PR3525/tutorials/applications/two_stage_stochastic.jl
new file mode 100644
index 00000000000..f724066d0ff
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/two_stage_stochastic.jl
@@ -0,0 +1,239 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Two-stage stochastic programs
+
+# The purpose of this tutorial is to demonstrate how to model and solve a
+# two-stage stochastic program.
+
+# This tutorial uses the following packages
+
+using JuMP
+import Distributions
+import HiGHS
+import Plots
+import StatsPlots
+import Statistics
+
+# ## Background
+
+# During the week, you are a busy practitioner of Operations Research. To escape
+# the drudgery of mathematics, you decide to open a side business selling creamy
+# mushroom pies with puff pastry. After a few weeks, it quickly becomes apparent
+# that operating a food business is not so easy.
+
+# The pies must be prepared in the morning, _before_ you open for the day and
+# can gauge the level of demand. If you bake too many, the unsold pies at the
+# end of the day must be discarded and you have wasted time and money on their
+# production. But if you bake too few, then there may be un-served customers and
+# you could have made more money by baking more pies.
+
+# After a few weeks of poor decision making, you decide to put your knowledge of
+# Operations Research to good use, starting with some data collection.
+
+# Each pie costs you \$2 to make, and you sell them at \$5 each. Disposal of an
+# unsold pie costs \$0.10. Based on three weeks of data collected, in which you
+# made 200 pies each week, you sold 150, 190, and 200 pies. Thus, as a guess,
+# you assume a triangular distribution of demand with a minimum of 150, a median
+# of 200, and a maximum of 250.
+
+# We can model this problem by a two-stage stochastic program. In the first
+# stage, we decide a quantity of pies to make ``x``. We make this decision
+# before we observe the demand ``d_\omega``. In the second stage, we sell
+# ``y_\omega`` pies, and incur any costs for unsold pies.
+
+# We can formulate this problem as follows:
+# ```math
+# \begin{aligned}
+# \max\limits_{x,y_\omega} \;\; & -2x + \mathbb{E}_\omega[5y_\omega - 0.1(x - y_\omega)] \\
+#   & y_\omega \le x              & \quad \forall \omega \in \Omega \\
+#   & 0 \le y_\omega \le d_\omega & \quad \forall \omega \in \Omega \\
+#   & x \ge 0.
+# \end{aligned}
+# ```
+
+# ## Sample Average approximation
+
+# If the distribution of demand is continuous, then our problem has an infinite
+# number of variables and constraints. To form a computationally tractable
+# problem, we instead use a finite set of samples drawn from the distribution.
+# This is called sample average approximation (SAA).
+
+D = Distributions.TriangularDist(150.0, 250.0, 200.0)
+N = 100
+d = sort!(rand(D, N));
+Ω = 1:N
+P = fill(1 / N, N);
+StatsPlots.histogram(d; bins = 20, label = "", xlabel = "Demand")
+
+# ## JuMP model
+
+# The implementation of our two-stage stochastic program in JuMP is:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x >= 0)
+@variable(model, 0 <= y[ω in Ω] <= d[ω])
+@constraint(model, [ω in Ω], y[ω] <= x)
+@expression(model, z[ω in Ω], 5y[ω] - 0.1 * (x - y[ω]))
+@objective(model, Max, -2x + sum(P[ω] * z[ω] for ω in Ω))
+optimize!(model)
+solution_summary(model)
+
+# The optimal number of pies to make is:
+
+value(x)
+
+# The distribution of total profit is:
+
+total_profit = [-2 * value(x) + value(z[ω]) for ω in Ω]
+
+# Let's plot it:
+
+"""
+    bin_distribution(x::Vector{Float64}, N::Int)
+
+A helper function that discretizes `x` into bins of width `N`.
+"""
+bin_distribution(x, N) = N * (floor(minimum(x) / N):ceil(maximum(x) / N))
+
+plot = StatsPlots.histogram(
+    total_profit;
+    bins = bin_distribution(total_profit, 25),
+    label = "",
+    xlabel = "Profit [\$]",
+    ylabel = "Number of outcomes",
+)
+μ = Statistics.mean(total_profit)
+Plots.vline!(
+    plot,
+    [μ];
+    label = "Expected profit (\$$(round(Int, μ)))",
+    linewidth = 3,
+)
+plot
+
+# ## Risk measures
+
+# A risk measure is a function which maps a random variable to a real number.
+# Common risk measures include the mean (expectation), median, mode, and
+# maximum. We need a risk measure to convert the distribution of second stage
+# costs into a single number that can be optimized.
+
+# Our model currently uses the expectation risk measure, but others are possible
+# too. One popular risk measure is the conditional value at risk (CVaR).
+
+# CVaR has a parameter $\gamma$, and it computes the expectation of the worst
+# $\gamma$ fraction of outcomes.
+
+# If we are maximizing, so that small outcomes are bad, the definition of CVaR
+# is:
+# ```math
+# CVaR_{\gamma}[Z] = \max\limits_{\xi} \;\; \xi - \frac{1}{\gamma}\mathbb{E}_\omega\left[(\xi - Z)_+\right]
+# ```
+# which can be formulated as the linear program:
+# ```math
+# \begin{aligned}
+# CVaR_{\gamma}[Z] = \max\limits_{\xi, z_\omega} \;\; & \xi - \frac{1}{\gamma}\sum P_\omega z_\omega\\
+#  & z_\omega \ge \xi - Z_\omega & \quad \forall \omega \\
+#  & z_\omega \ge 0 & \quad \forall \omega.
+# \end{aligned}
+# ```
+
+function CVaR(Z::Vector{Float64}, P::Vector{Float64}; γ::Float64)
+    @assert 0 < γ <= 1
+    N = length(Z)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, ξ)
+    @variable(model, z[1:N] >= 0)
+    @constraint(model, [i in 1:N], z[i] >= ξ - Z[i])
+    @objective(model, Max, ξ - 1 / γ * sum(P[i] * z[i] for i in 1:N))
+    optimize!(model)
+    return objective_value(model)
+end
+
+# When `γ` is `1.0`, we compute the mean of the profit:
+
+cvar_10 = CVaR(total_profit, P; γ = 1.0)
+
+#-
+
+Statistics.mean(total_profit)
+
+# As `γ` approaches `0.0`, we compute the worst-case (minimum) profit:
+
+cvar_00 = CVaR(total_profit, P; γ = 0.0001)
+
+#-
+
+minimum(total_profit)
+
+# By varying `γ` between `0` and `1` we can compute some trade-off of these two
+# extremes:
+
+cvar_05 = CVaR(total_profit, P; γ = 0.5)
+
+# Let's plot these outcomes on our distribution:
+
+plot = StatsPlots.histogram(
+    total_profit;
+    bins = bin_distribution(total_profit, 25),
+    label = "",
+    xlabel = "Profit [\$]",
+    ylabel = "Number of outcomes",
+)
+Plots.vline!(
+    plot,
+    [cvar_10 cvar_05 cvar_00];
+    label = ["γ = 1.0" "γ = 0.5" "γ = 0.0"],
+    linewidth = 3,
+)
+plot
+
+# ## Risk averse sample average approximation
+
+# Because CVaR can be formulated as a linear program, we can form a risk averse
+# sample average approximation model by combining the two formulations:
+
+γ = 0.4
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x >= 0)
+@variable(model, 0 <= y[ω in Ω] <= d[ω])
+@constraint(model, [ω in Ω], y[ω] <= x)
+@expression(model, Z[ω in Ω], 5 * y[ω] - 0.1(x - y[ω]))
+@variable(model, ξ)
+@variable(model, z[ω in Ω] >= 0)
+@constraint(model, [ω in Ω], z[ω] >= ξ - Z[ω])
+@objective(model, Max, -2x + ξ - 1 / γ * sum(P[ω] * z[ω] for ω in Ω))
+optimize!(model)
+
+# When ``\gamma = 0.4``, the optimal number of pies to bake is:
+
+value(x)
+
+# The distribution of total profit is:
+
+risk_averse_total_profit = [value(-2x + Z[ω]) for ω in Ω]
+bins = bin_distribution([total_profit; risk_averse_total_profit], 25)
+plot = StatsPlots.histogram(total_profit; label = "Expectation", bins = bins)
+StatsPlots.histogram!(
+    plot,
+    risk_averse_total_profit;
+    label = "CV@R",
+    bins = bins,
+    alpha = 0.5,
+)
+plot
+
+# ## Next steps
+#
+#  * Try solving this problem for different numbers of samples and different
+#    distributions.
+#  * Refactor the example to avoid hard-coding the costs. What happens to the
+#    solution if the cost of disposing unsold pies increases?
+#  * Plot the optimal number of pies to make for different values of the risk
+#    aversion parameter $\gamma$. What is the relationship?
diff --git a/previews/PR3525/tutorials/applications/two_stage_stochastic/01f8597b.svg b/previews/PR3525/tutorials/applications/two_stage_stochastic/01f8597b.svg
new file mode 100644
index 00000000000..4e9d872eafa
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/two_stage_stochastic/01f8597b.svg
@@ -0,0 +1,108 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip430">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip430)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip431">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip430)" d="M141.853 1486.45 L2352.76 1486.45 L2352.76 47.2441 L141.853 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip432">
+    <rect x="141" y="47" width="2212" height="1440"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="460.226,1486.45 460.226,47.2441 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="853.765,1486.45 853.765,47.2441 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1247.3,1486.45 1247.3,47.2441 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1640.84,1486.45 1640.84,47.2441 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2034.38,1486.45 2034.38,47.2441 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="141.853,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="460.226,1486.45 460.226,1467.55 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="853.765,1486.45 853.765,1467.55 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1247.3,1486.45 1247.3,1467.55 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1640.84,1486.45 1640.84,1467.55 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2034.38,1486.45 2034.38,1467.55 "/>
+<path clip-path="url(#clip430)" d="M433.317 1518.36 L421.511 1536.81 L433.317 1536.81 L433.317 1518.36 M432.09 1514.29 L437.969 1514.29 L437.969 1536.81 L442.9 1536.81 L442.9 1540.7 L437.969 1540.7 L437.969 1548.85 L433.317 1548.85 L433.317 1540.7 L417.715 1540.7 L417.715 1536.19 L432.09 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M460.631 1517.37 Q457.02 1517.37 455.191 1520.93 Q453.386 1524.47 453.386 1531.6 Q453.386 1538.71 455.191 1542.27 Q457.02 1545.82 460.631 1545.82 Q464.265 1545.82 466.071 1542.27 Q467.9 1538.71 467.9 1531.6 Q467.9 1524.47 466.071 1520.93 Q464.265 1517.37 460.631 1517.37 M460.631 1513.66 Q466.441 1513.66 469.497 1518.27 Q472.576 1522.85 472.576 1531.6 Q472.576 1540.33 469.497 1544.94 Q466.441 1549.52 460.631 1549.52 Q454.821 1549.52 451.742 1544.94 Q448.687 1540.33 448.687 1531.6 Q448.687 1522.85 451.742 1518.27 Q454.821 1513.66 460.631 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M490.793 1517.37 Q487.182 1517.37 485.353 1520.93 Q483.548 1524.47 483.548 1531.6 Q483.548 1538.71 485.353 1542.27 Q487.182 1545.82 490.793 1545.82 Q494.427 1545.82 496.233 1542.27 Q498.062 1538.71 498.062 1531.6 Q498.062 1524.47 496.233 1520.93 Q494.427 1517.37 490.793 1517.37 M490.793 1513.66 Q496.603 1513.66 499.659 1518.27 Q502.737 1522.85 502.737 1531.6 Q502.737 1540.33 499.659 1544.94 Q496.603 1549.52 490.793 1549.52 Q484.983 1549.52 481.904 1544.94 Q478.849 1540.33 478.849 1531.6 Q478.849 1522.85 481.904 1518.27 Q484.983 1513.66 490.793 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M826.856 1518.36 L815.05 1536.81 L826.856 1536.81 L826.856 1518.36 M825.629 1514.29 L831.508 1514.29 L831.508 1536.81 L836.439 1536.81 L836.439 1540.7 L831.508 1540.7 L831.508 1548.85 L826.856 1548.85 L826.856 1540.7 L811.254 1540.7 L811.254 1536.19 L825.629 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M844.217 1514.29 L862.573 1514.29 L862.573 1518.22 L848.499 1518.22 L848.499 1526.7 Q849.518 1526.35 850.536 1526.19 Q851.555 1526 852.573 1526 Q858.36 1526 861.74 1529.17 Q865.119 1532.34 865.119 1537.76 Q865.119 1543.34 861.647 1546.44 Q858.175 1549.52 851.856 1549.52 Q849.68 1549.52 847.411 1549.15 Q845.166 1548.78 842.758 1548.04 L842.758 1543.34 Q844.842 1544.47 847.064 1545.03 Q849.286 1545.58 851.763 1545.58 Q855.768 1545.58 858.106 1543.48 Q860.443 1541.37 860.443 1537.76 Q860.443 1534.15 858.106 1532.04 Q855.768 1529.94 851.763 1529.94 Q849.888 1529.94 848.013 1530.35 Q846.161 1530.77 844.217 1531.65 L844.217 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M884.332 1517.37 Q880.721 1517.37 878.892 1520.93 Q877.087 1524.47 877.087 1531.6 Q877.087 1538.71 878.892 1542.27 Q880.721 1545.82 884.332 1545.82 Q887.966 1545.82 889.772 1542.27 Q891.601 1538.71 891.601 1531.6 Q891.601 1524.47 889.772 1520.93 Q887.966 1517.37 884.332 1517.37 M884.332 1513.66 Q890.142 1513.66 893.198 1518.27 Q896.277 1522.85 896.277 1531.6 Q896.277 1540.33 893.198 1544.94 Q890.142 1549.52 884.332 1549.52 Q878.522 1549.52 875.443 1544.94 Q872.388 1540.33 872.388 1531.6 Q872.388 1522.85 875.443 1518.27 Q878.522 1513.66 884.332 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M1206.92 1514.29 L1225.28 1514.29 L1225.28 1518.22 L1211.21 1518.22 L1211.21 1526.7 Q1212.22 1526.35 1213.24 1526.19 Q1214.26 1526 1215.28 1526 Q1221.07 1526 1224.45 1529.17 Q1227.83 1532.34 1227.83 1537.76 Q1227.83 1543.34 1224.35 1546.44 Q1220.88 1549.52 1214.56 1549.52 Q1212.39 1549.52 1210.12 1549.15 Q1207.87 1548.78 1205.46 1548.04 L1205.46 1543.34 Q1207.55 1544.47 1209.77 1545.03 Q1211.99 1545.58 1214.47 1545.58 Q1218.47 1545.58 1220.81 1543.48 Q1223.15 1541.37 1223.15 1537.76 Q1223.15 1534.15 1220.81 1532.04 Q1218.47 1529.94 1214.47 1529.94 Q1212.59 1529.94 1210.72 1530.35 Q1208.87 1530.77 1206.92 1531.65 L1206.92 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M1247.04 1517.37 Q1243.43 1517.37 1241.6 1520.93 Q1239.79 1524.47 1239.79 1531.6 Q1239.79 1538.71 1241.6 1542.27 Q1243.43 1545.82 1247.04 1545.82 Q1250.67 1545.82 1252.48 1542.27 Q1254.31 1538.71 1254.31 1531.6 Q1254.31 1524.47 1252.48 1520.93 Q1250.67 1517.37 1247.04 1517.37 M1247.04 1513.66 Q1252.85 1513.66 1255.9 1518.27 Q1258.98 1522.85 1258.98 1531.6 Q1258.98 1540.33 1255.9 1544.94 Q1252.85 1549.52 1247.04 1549.52 Q1241.23 1549.52 1238.15 1544.94 Q1235.09 1540.33 1235.09 1531.6 Q1235.09 1522.85 1238.15 1518.27 Q1241.23 1513.66 1247.04 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M1277.2 1517.37 Q1273.59 1517.37 1271.76 1520.93 Q1269.95 1524.47 1269.95 1531.6 Q1269.95 1538.71 1271.76 1542.27 Q1273.59 1545.82 1277.2 1545.82 Q1280.83 1545.82 1282.64 1542.27 Q1284.47 1538.71 1284.47 1531.6 Q1284.47 1524.47 1282.64 1520.93 Q1280.83 1517.37 1277.2 1517.37 M1277.2 1513.66 Q1283.01 1513.66 1286.07 1518.27 Q1289.14 1522.85 1289.14 1531.6 Q1289.14 1540.33 1286.07 1544.94 Q1283.01 1549.52 1277.2 1549.52 Q1271.39 1549.52 1268.31 1544.94 Q1265.26 1540.33 1265.26 1531.6 Q1265.26 1522.85 1268.31 1518.27 Q1271.39 1513.66 1277.2 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M1600.46 1514.29 L1618.82 1514.29 L1618.82 1518.22 L1604.74 1518.22 L1604.74 1526.7 Q1605.76 1526.35 1606.78 1526.19 Q1607.8 1526 1608.82 1526 Q1614.61 1526 1617.98 1529.17 Q1621.36 1532.34 1621.36 1537.76 Q1621.36 1543.34 1617.89 1546.44 Q1614.42 1549.52 1608.1 1549.52 Q1605.92 1549.52 1603.66 1549.15 Q1601.41 1548.78 1599 1548.04 L1599 1543.34 Q1601.09 1544.47 1603.31 1545.03 Q1605.53 1545.58 1608.01 1545.58 Q1612.01 1545.58 1614.35 1543.48 Q1616.69 1541.37 1616.69 1537.76 Q1616.69 1534.15 1614.35 1532.04 Q1612.01 1529.94 1608.01 1529.94 Q1606.13 1529.94 1604.26 1530.35 Q1602.41 1530.77 1600.46 1531.65 L1600.46 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M1630.62 1514.29 L1648.98 1514.29 L1648.98 1518.22 L1634.91 1518.22 L1634.91 1526.7 Q1635.92 1526.35 1636.94 1526.19 Q1637.96 1526 1638.98 1526 Q1644.77 1526 1648.15 1529.17 Q1651.53 1532.34 1651.53 1537.76 Q1651.53 1543.34 1648.05 1546.44 Q1644.58 1549.52 1638.26 1549.52 Q1636.09 1549.52 1633.82 1549.15 Q1631.57 1548.78 1629.17 1548.04 L1629.17 1543.34 Q1631.25 1544.47 1633.47 1545.03 Q1635.69 1545.58 1638.17 1545.58 Q1642.17 1545.58 1644.51 1543.48 Q1646.85 1541.37 1646.85 1537.76 Q1646.85 1534.15 1644.51 1532.04 Q1642.17 1529.94 1638.17 1529.94 Q1636.29 1529.94 1634.42 1530.35 Q1632.57 1530.77 1630.62 1531.65 L1630.62 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M1670.74 1517.37 Q1667.13 1517.37 1665.3 1520.93 Q1663.49 1524.47 1663.49 1531.6 Q1663.49 1538.71 1665.3 1542.27 Q1667.13 1545.82 1670.74 1545.82 Q1674.37 1545.82 1676.18 1542.27 Q1678.01 1538.71 1678.01 1531.6 Q1678.01 1524.47 1676.18 1520.93 Q1674.37 1517.37 1670.74 1517.37 M1670.74 1513.66 Q1676.55 1513.66 1679.6 1518.27 Q1682.68 1522.85 1682.68 1531.6 Q1682.68 1540.33 1679.6 1544.94 Q1676.55 1549.52 1670.74 1549.52 Q1664.93 1549.52 1661.85 1544.94 Q1658.79 1540.33 1658.79 1531.6 Q1658.79 1522.85 1661.85 1518.27 Q1664.93 1513.66 1670.74 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2004.71 1529.7 Q2001.56 1529.7 1999.71 1531.86 Q1997.88 1534.01 1997.88 1537.76 Q1997.88 1541.49 1999.71 1543.66 Q2001.56 1545.82 2004.71 1545.82 Q2007.86 1545.82 2009.68 1543.66 Q2011.54 1541.49 2011.54 1537.76 Q2011.54 1534.01 2009.68 1531.86 Q2007.86 1529.7 2004.71 1529.7 M2013.99 1515.05 L2013.99 1519.31 Q2012.23 1518.48 2010.42 1518.04 Q2008.64 1517.6 2006.88 1517.6 Q2002.25 1517.6 1999.8 1520.72 Q1997.37 1523.85 1997.02 1530.17 Q1998.39 1528.15 2000.45 1527.09 Q2002.51 1526 2004.98 1526 Q2010.19 1526 2013.2 1529.17 Q2016.23 1532.32 2016.23 1537.76 Q2016.23 1543.08 2013.09 1546.3 Q2009.94 1549.52 2004.71 1549.52 Q1998.71 1549.52 1995.54 1544.94 Q1992.37 1540.33 1992.37 1531.6 Q1992.37 1523.41 1996.26 1518.55 Q2000.15 1513.66 2006.7 1513.66 Q2008.46 1513.66 2010.24 1514.01 Q2012.04 1514.36 2013.99 1515.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2034.29 1517.37 Q2030.68 1517.37 2028.85 1520.93 Q2027.04 1524.47 2027.04 1531.6 Q2027.04 1538.71 2028.85 1542.27 Q2030.68 1545.82 2034.29 1545.82 Q2037.92 1545.82 2039.73 1542.27 Q2041.56 1538.71 2041.56 1531.6 Q2041.56 1524.47 2039.73 1520.93 Q2037.92 1517.37 2034.29 1517.37 M2034.29 1513.66 Q2040.1 1513.66 2043.16 1518.27 Q2046.23 1522.85 2046.23 1531.6 Q2046.23 1540.33 2043.16 1544.94 Q2040.1 1549.52 2034.29 1549.52 Q2028.48 1549.52 2025.4 1544.94 Q2022.35 1540.33 2022.35 1531.6 Q2022.35 1522.85 2025.4 1518.27 Q2028.48 1513.66 2034.29 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2064.45 1517.37 Q2060.84 1517.37 2059.01 1520.93 Q2057.21 1524.47 2057.21 1531.6 Q2057.21 1538.71 2059.01 1542.27 Q2060.84 1545.82 2064.45 1545.82 Q2068.09 1545.82 2069.89 1542.27 Q2071.72 1538.71 2071.72 1531.6 Q2071.72 1524.47 2069.89 1520.93 Q2068.09 1517.37 2064.45 1517.37 M2064.45 1513.66 Q2070.26 1513.66 2073.32 1518.27 Q2076.4 1522.85 2076.4 1531.6 Q2076.4 1540.33 2073.32 1544.94 Q2070.26 1549.52 2064.45 1549.52 Q2058.64 1549.52 2055.56 1544.94 Q2052.51 1540.33 2052.51 1531.6 Q2052.51 1522.85 2055.56 1518.27 Q2058.64 1513.66 2064.45 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="141.853,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="141.853,1126.65 2352.76,1126.65 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="141.853,766.846 2352.76,766.846 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="141.853,407.045 2352.76,407.045 "/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="141.853,47.2441 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="141.853,1486.45 141.853,47.2441 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="141.853,1486.45 160.751,1486.45 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="141.853,1126.65 160.751,1126.65 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="141.853,766.846 160.751,766.846 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="141.853,407.045 160.751,407.045 "/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="141.853,47.2441 160.751,47.2441 "/>
+<path clip-path="url(#clip430)" d="M93.9086 1472.25 Q90.2975 1472.25 88.4688 1475.81 Q86.6632 1479.35 86.6632 1486.48 Q86.6632 1493.59 88.4688 1497.15 Q90.2975 1500.7 93.9086 1500.7 Q97.5428 1500.7 99.3483 1497.15 Q101.177 1493.59 101.177 1486.48 Q101.177 1479.35 99.3483 1475.81 Q97.5428 1472.25 93.9086 1472.25 M93.9086 1468.54 Q99.7187 1468.54 102.774 1473.15 Q105.853 1477.73 105.853 1486.48 Q105.853 1495.21 102.774 1499.82 Q99.7187 1504.4 93.9086 1504.4 Q88.0984 1504.4 85.0197 1499.82 Q81.9642 1495.21 81.9642 1486.48 Q81.9642 1477.73 85.0197 1473.15 Q88.0984 1468.54 93.9086 1468.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M57.7745 1139.99 L74.0939 1139.99 L74.0939 1143.93 L52.1495 1143.93 L52.1495 1139.99 Q54.8115 1137.24 59.3949 1132.61 Q64.0013 1127.95 65.1819 1126.61 Q67.4272 1124.09 68.3068 1122.35 Q69.2096 1120.59 69.2096 1118.9 Q69.2096 1116.15 67.2652 1114.41 Q65.3439 1112.68 62.2421 1112.68 Q60.043 1112.68 57.5893 1113.44 Q55.1588 1114.2 52.381 1115.76 L52.381 1111.03 Q55.2051 1109.9 57.6588 1109.32 Q60.1124 1108.74 62.1495 1108.74 Q67.5198 1108.74 70.7142 1111.43 Q73.9087 1114.11 73.9087 1118.6 Q73.9087 1120.73 73.0985 1122.65 Q72.3115 1124.55 70.205 1127.14 Q69.6263 1127.82 66.5245 1131.03 Q63.4226 1134.23 57.7745 1139.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M93.9086 1112.45 Q90.2975 1112.45 88.4688 1116.01 Q86.6632 1119.55 86.6632 1126.68 Q86.6632 1133.79 88.4688 1137.35 Q90.2975 1140.89 93.9086 1140.89 Q97.5428 1140.89 99.3483 1137.35 Q101.177 1133.79 101.177 1126.68 Q101.177 1119.55 99.3483 1116.01 Q97.5428 1112.45 93.9086 1112.45 M93.9086 1108.74 Q99.7187 1108.74 102.774 1113.35 Q105.853 1117.93 105.853 1126.68 Q105.853 1135.41 102.774 1140.01 Q99.7187 1144.6 93.9086 1144.6 Q88.0984 1144.6 85.0197 1140.01 Q81.9642 1135.41 81.9642 1126.68 Q81.9642 1117.93 85.0197 1113.35 Q88.0984 1108.74 93.9086 1108.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M66.5939 753.64 L54.7884 772.089 L66.5939 772.089 L66.5939 753.64 M65.367 749.566 L71.2466 749.566 L71.2466 772.089 L76.1772 772.089 L76.1772 775.978 L71.2466 775.978 L71.2466 784.126 L66.5939 784.126 L66.5939 775.978 L50.9921 775.978 L50.9921 771.464 L65.367 749.566 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M93.9086 752.645 Q90.2975 752.645 88.4688 756.209 Q86.6632 759.751 86.6632 766.881 Q86.6632 773.987 88.4688 777.552 Q90.2975 781.094 93.9086 781.094 Q97.5428 781.094 99.3483 777.552 Q101.177 773.987 101.177 766.881 Q101.177 759.751 99.3483 756.209 Q97.5428 752.645 93.9086 752.645 M93.9086 748.941 Q99.7187 748.941 102.774 753.547 Q105.853 758.131 105.853 766.881 Q105.853 775.608 102.774 780.214 Q99.7187 784.797 93.9086 784.797 Q88.0984 784.797 85.0197 780.214 Q81.9642 775.608 81.9642 766.881 Q81.9642 758.131 85.0197 753.547 Q88.0984 748.941 93.9086 748.941 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M64.3254 405.182 Q61.1773 405.182 59.3254 407.334 Q57.4967 409.487 57.4967 413.237 Q57.4967 416.964 59.3254 419.14 Q61.1773 421.293 64.3254 421.293 Q67.4735 421.293 69.3022 419.14 Q71.1541 416.964 71.1541 413.237 Q71.1541 409.487 69.3022 407.334 Q67.4735 405.182 64.3254 405.182 M73.6077 390.529 L73.6077 394.788 Q71.8485 393.955 70.0429 393.515 Q68.2606 393.075 66.5013 393.075 Q61.8717 393.075 59.418 396.2 Q56.9875 399.325 56.6402 405.645 Q58.006 403.631 60.0662 402.566 Q62.1263 401.478 64.6032 401.478 Q69.8115 401.478 72.8207 404.649 Q75.8531 407.797 75.8531 413.237 Q75.8531 418.561 72.705 421.779 Q69.5568 424.996 64.3254 424.996 Q58.33 424.996 55.1588 420.413 Q51.9875 415.807 51.9875 407.08 Q51.9875 398.885 55.8764 394.024 Q59.7652 389.14 66.3161 389.14 Q68.0754 389.14 69.8578 389.487 Q71.6633 389.834 73.6077 390.529 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M93.9086 392.844 Q90.2975 392.844 88.4688 396.409 Q86.6632 399.95 86.6632 407.08 Q86.6632 414.186 88.4688 417.751 Q90.2975 421.293 93.9086 421.293 Q97.5428 421.293 99.3483 417.751 Q101.177 414.186 101.177 407.08 Q101.177 399.95 99.3483 396.409 Q97.5428 392.844 93.9086 392.844 M93.9086 389.14 Q99.7187 389.14 102.774 393.746 Q105.853 398.33 105.853 407.08 Q105.853 415.807 102.774 420.413 Q99.7187 424.996 93.9086 424.996 Q88.0984 424.996 85.0197 420.413 Q81.9642 415.807 81.9642 407.08 Q81.9642 398.33 85.0197 393.746 Q88.0984 389.14 93.9086 389.14 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M63.7467 48.1121 Q60.4134 48.1121 58.4921 49.8945 Q56.5939 51.6769 56.5939 54.8019 Q56.5939 57.9269 58.4921 59.7093 Q60.4134 61.4917 63.7467 61.4917 Q67.08 61.4917 69.0013 59.7093 Q70.9226 57.9038 70.9226 54.8019 Q70.9226 51.6769 69.0013 49.8945 Q67.1032 48.1121 63.7467 48.1121 M59.0708 46.1214 Q56.0615 45.3807 54.3717 43.3205 Q52.7051 41.2603 52.7051 38.2974 Q52.7051 34.1539 55.6449 31.7465 Q58.6078 29.3391 63.7467 29.3391 Q68.9087 29.3391 71.8485 31.7465 Q74.7883 34.1539 74.7883 38.2974 Q74.7883 41.2603 73.0985 43.3205 Q71.4318 45.3807 68.4457 46.1214 Q71.8253 46.9084 73.7003 49.2001 Q75.5985 51.4918 75.5985 54.8019 Q75.5985 59.825 72.5198 62.5102 Q69.4642 65.1954 63.7467 65.1954 Q58.0291 65.1954 54.9504 62.5102 Q51.8949 59.825 51.8949 54.8019 Q51.8949 51.4918 53.793 49.2001 Q55.6912 46.9084 59.0708 46.1214 M57.3578 38.7372 Q57.3578 41.4224 59.0245 42.927 Q60.7143 44.4316 63.7467 44.4316 Q66.7559 44.4316 68.4457 42.927 Q70.1587 41.4224 70.1587 38.7372 Q70.1587 36.052 68.4457 34.5474 Q66.7559 33.0428 63.7467 33.0428 Q60.7143 33.0428 59.0245 34.5474 Q57.3578 36.052 57.3578 38.7372 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M93.9086 33.0428 Q90.2975 33.0428 88.4688 36.6076 Q86.6632 40.1492 86.6632 47.2788 Q86.6632 54.3853 88.4688 57.9501 Q90.2975 61.4917 93.9086 61.4917 Q97.5428 61.4917 99.3483 57.9501 Q101.177 54.3853 101.177 47.2788 Q101.177 40.1492 99.3483 36.6076 Q97.5428 33.0428 93.9086 33.0428 M93.9086 29.3391 Q99.7187 29.3391 102.774 33.9456 Q105.853 38.5289 105.853 47.2788 Q105.853 56.0056 102.774 60.6121 Q99.7187 65.1954 93.9086 65.1954 Q88.0984 65.1954 85.0197 60.6121 Q81.9642 56.0056 81.9642 47.2788 Q81.9642 38.5289 85.0197 33.9456 Q88.0984 29.3391 93.9086 29.3391 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip432)" d="M263.457 1450.47 L263.457 1486.45 L460.226 1486.45 L460.226 1450.47 L263.457 1450.47 L263.457 1450.47  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="263.457,1450.47 263.457,1486.45 460.226,1486.45 460.226,1450.47 263.457,1450.47 "/>
+<path clip-path="url(#clip432)" d="M460.226 1432.48 L460.226 1486.45 L656.996 1486.45 L656.996 1432.48 L460.226 1432.48 L460.226 1432.48  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="460.226,1432.48 460.226,1486.45 656.996,1486.45 656.996,1432.48 460.226,1432.48 "/>
+<path clip-path="url(#clip432)" d="M656.996 1360.52 L656.996 1486.45 L853.765 1486.45 L853.765 1360.52 L656.996 1360.52 L656.996 1360.52  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="656.996,1360.52 656.996,1486.45 853.765,1486.45 853.765,1360.52 656.996,1360.52 "/>
+<path clip-path="url(#clip432)" d="M853.765 1360.52 L853.765 1486.45 L1050.53 1486.45 L1050.53 1360.52 L853.765 1360.52 L853.765 1360.52  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="853.765,1360.52 853.765,1486.45 1050.53,1486.45 1050.53,1360.52 853.765,1360.52 "/>
+<path clip-path="url(#clip432)" d="M1050.53 1414.49 L1050.53 1486.45 L1247.3 1486.45 L1247.3 1414.49 L1050.53 1414.49 L1050.53 1414.49  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1050.53,1414.49 1050.53,1486.45 1247.3,1486.45 1247.3,1414.49 1050.53,1414.49 "/>
+<path clip-path="url(#clip432)" d="M1247.3 1450.47 L1247.3 1486.45 L1444.07 1486.45 L1444.07 1450.47 L1247.3 1450.47 L1247.3 1450.47  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1247.3,1450.47 1247.3,1486.45 1444.07,1486.45 1444.07,1450.47 1247.3,1450.47 "/>
+<path clip-path="url(#clip432)" d="M1444.07 1396.5 L1444.07 1486.45 L1640.84 1486.45 L1640.84 1396.5 L1444.07 1396.5 L1444.07 1396.5  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1444.07,1396.5 1444.07,1486.45 1640.84,1486.45 1640.84,1396.5 1444.07,1396.5 "/>
+<path clip-path="url(#clip432)" d="M1640.84 1288.56 L1640.84 1486.45 L1837.61 1486.45 L1837.61 1288.56 L1640.84 1288.56 L1640.84 1288.56  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1640.84,1288.56 1640.84,1486.45 1837.61,1486.45 1837.61,1288.56 1640.84,1288.56 "/>
+<path clip-path="url(#clip432)" d="M1837.61 1180.62 L1837.61 1486.45 L2034.38 1486.45 L2034.38 1180.62 L1837.61 1180.62 L1837.61 1180.62  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1837.61,1180.62 1837.61,1486.45 2034.38,1486.45 2034.38,1180.62 1837.61,1180.62 "/>
+<path clip-path="url(#clip432)" d="M2034.38 730.866 L2034.38 1486.45 L2231.15 1486.45 L2231.15 730.866 L2034.38 730.866 L2034.38 730.866  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2034.38,730.866 2034.38,1486.45 2231.15,1486.45 2231.15,730.866 2034.38,730.866 "/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="361.841" cy="1450.47" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="558.611" cy="1432.48" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="755.38" cy="1360.52" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="952.15" cy="1360.52" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1148.92" cy="1414.49" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1345.69" cy="1450.47" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1542.46" cy="1396.5" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1739.23" cy="1288.56" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1936" cy="1180.62" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2132.77" cy="730.866" r="2"/>
+<path clip-path="url(#clip432)" d="M263.457 1486.45 L263.457 1486.45 L460.226 1486.45 L460.226 1486.45 L263.457 1486.45 L263.457 1486.45  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="263.457,1486.45 263.457,1486.45 460.226,1486.45 263.457,1486.45 "/>
+<path clip-path="url(#clip432)" d="M460.226 1450.47 L460.226 1486.45 L656.996 1486.45 L656.996 1450.47 L460.226 1450.47 L460.226 1450.47  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="460.226,1450.47 460.226,1486.45 656.996,1486.45 656.996,1450.47 460.226,1450.47 "/>
+<path clip-path="url(#clip432)" d="M656.996 1486.45 L656.996 1486.45 L853.765 1486.45 L853.765 1486.45 L656.996 1486.45 L656.996 1486.45  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="656.996,1486.45 656.996,1486.45 853.765,1486.45 656.996,1486.45 "/>
+<path clip-path="url(#clip432)" d="M853.765 1360.52 L853.765 1486.45 L1050.53 1486.45 L1050.53 1360.52 L853.765 1360.52 L853.765 1360.52  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="853.765,1360.52 853.765,1486.45 1050.53,1486.45 1050.53,1360.52 853.765,1360.52 "/>
+<path clip-path="url(#clip432)" d="M1050.53 1342.53 L1050.53 1486.45 L1247.3 1486.45 L1247.3 1342.53 L1050.53 1342.53 L1050.53 1342.53  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="1050.53,1342.53 1050.53,1486.45 1247.3,1486.45 1247.3,1342.53 1050.53,1342.53 "/>
+<path clip-path="url(#clip432)" d="M1247.3 1432.48 L1247.3 1486.45 L1444.07 1486.45 L1444.07 1432.48 L1247.3 1432.48 L1247.3 1432.48  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="1247.3,1432.48 1247.3,1486.45 1444.07,1486.45 1444.07,1432.48 1247.3,1432.48 "/>
+<path clip-path="url(#clip432)" d="M1444.07 47.2441 L1444.07 1486.45 L1640.84 1486.45 L1640.84 47.2441 L1444.07 47.2441 L1444.07 47.2441  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="1444.07,47.2441 1444.07,1486.45 1640.84,1486.45 1640.84,47.2441 1444.07,47.2441 "/>
+<path clip-path="url(#clip432)" d="M1640.84 1486.45 L1640.84 1486.45 L1837.61 1486.45 L1837.61 1486.45 L1640.84 1486.45 L1640.84 1486.45  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="1640.84,1486.45 1640.84,1486.45 1837.61,1486.45 1640.84,1486.45 "/>
+<path clip-path="url(#clip432)" d="M1837.61 1486.45 L1837.61 1486.45 L2034.38 1486.45 L2034.38 1486.45 L1837.61 1486.45 L1837.61 1486.45  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="1837.61,1486.45 1837.61,1486.45 2034.38,1486.45 1837.61,1486.45 "/>
+<path clip-path="url(#clip432)" d="M2034.38 1486.45 L2034.38 1486.45 L2231.15 1486.45 L2231.15 1486.45 L2034.38 1486.45 L2034.38 1486.45  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip432)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:0.5; fill:none" points="2034.38,1486.45 2034.38,1486.45 2231.15,1486.45 2034.38,1486.45 "/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="361.841" cy="1486.45" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="558.611" cy="1450.47" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="755.38" cy="1486.45" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="952.15" cy="1360.52" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1148.92" cy="1342.53" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1345.69" cy="1432.48" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1542.46" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1739.23" cy="1486.45" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1936" cy="1486.45" r="2"/>
+<circle clip-path="url(#clip432)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2132.77" cy="1486.45" r="2"/>
+<path clip-path="url(#clip430)" d="M1784.87 250.738 L2279.06 250.738 L2279.06 95.2176 L1784.87 95.2176  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1784.87,250.738 2279.06,250.738 2279.06,95.2176 1784.87,95.2176 1784.87,250.738 "/>
+<path clip-path="url(#clip430)" d="M1809.43 167.794 L1956.83 167.794 L1956.83 126.322 L1809.43 126.322 L1809.43 167.794  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1809.43,167.794 1956.83,167.794 1956.83,126.322 1809.43,126.322 1809.43,167.794 "/>
+<path clip-path="url(#clip430)" d="M1981.39 129.778 L2003.24 129.778 L2003.24 133.713 L1986.07 133.713 L1986.07 143.944 L2002.53 143.944 L2002.53 147.879 L1986.07 147.879 L1986.07 160.402 L2003.66 160.402 L2003.66 164.338 L1981.39 164.338 L1981.39 129.778 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2032.71 138.412 L2023.34 151.027 L2033.2 164.338 L2028.18 164.338 L2020.63 154.152 L2013.08 164.338 L2008.06 164.338 L2018.13 150.773 L2008.92 138.412 L2013.94 138.412 L2020.81 147.648 L2027.69 138.412 L2032.71 138.412 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2043.34 160.449 L2043.34 174.199 L2039.05 174.199 L2039.05 138.412 L2043.34 138.412 L2043.34 142.347 Q2044.68 140.032 2046.72 138.921 Q2048.78 137.787 2051.62 137.787 Q2056.35 137.787 2059.29 141.537 Q2062.25 145.287 2062.25 151.398 Q2062.25 157.509 2059.29 161.259 Q2056.35 165.009 2051.62 165.009 Q2048.78 165.009 2046.72 163.898 Q2044.68 162.763 2043.34 160.449 M2057.83 151.398 Q2057.83 146.699 2055.88 144.037 Q2053.96 141.352 2050.58 141.352 Q2047.2 141.352 2045.26 144.037 Q2043.34 146.699 2043.34 151.398 Q2043.34 156.097 2045.26 158.782 Q2047.2 161.444 2050.58 161.444 Q2053.96 161.444 2055.88 158.782 Q2057.83 156.097 2057.83 151.398 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2091.49 150.31 L2091.49 152.393 L2071.9 152.393 Q2072.18 156.791 2074.54 159.106 Q2076.93 161.398 2081.16 161.398 Q2083.61 161.398 2085.91 160.796 Q2088.22 160.194 2090.49 158.99 L2090.49 163.018 Q2088.2 163.99 2085.79 164.5 Q2083.38 165.009 2080.91 165.009 Q2074.7 165.009 2071.07 161.398 Q2067.46 157.787 2067.46 151.629 Q2067.46 145.264 2070.88 141.537 Q2074.33 137.787 2080.17 137.787 Q2085.4 137.787 2088.43 141.166 Q2091.49 144.523 2091.49 150.31 M2087.23 149.06 Q2087.18 145.565 2085.26 143.481 Q2083.36 141.398 2080.21 141.398 Q2076.65 141.398 2074.49 143.412 Q2072.36 145.426 2072.04 149.083 L2087.23 149.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2117.13 139.407 L2117.13 143.389 Q2115.33 142.393 2113.5 141.907 Q2111.69 141.398 2109.84 141.398 Q2105.7 141.398 2103.41 144.037 Q2101.11 146.652 2101.11 151.398 Q2101.11 156.143 2103.41 158.782 Q2105.7 161.398 2109.84 161.398 Q2111.69 161.398 2113.5 160.912 Q2115.33 160.402 2117.13 159.407 L2117.13 163.342 Q2115.35 164.176 2113.43 164.592 Q2111.53 165.009 2109.38 165.009 Q2103.52 165.009 2100.07 161.328 Q2096.62 157.648 2096.62 151.398 Q2096.62 145.055 2100.1 141.421 Q2103.59 137.787 2109.66 137.787 Q2111.62 137.787 2113.5 138.203 Q2115.37 138.597 2117.13 139.407 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2128.75 131.051 L2128.75 138.412 L2137.53 138.412 L2137.53 141.722 L2128.75 141.722 L2128.75 155.796 Q2128.75 158.967 2129.61 159.87 Q2130.49 160.773 2133.15 160.773 L2137.53 160.773 L2137.53 164.338 L2133.15 164.338 Q2128.22 164.338 2126.35 162.509 Q2124.47 160.657 2124.47 155.796 L2124.47 141.722 L2121.35 141.722 L2121.35 138.412 L2124.47 138.412 L2124.47 131.051 L2128.75 131.051 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2154.91 151.305 Q2149.75 151.305 2147.76 152.486 Q2145.77 153.666 2145.77 156.514 Q2145.77 158.782 2147.25 160.125 Q2148.75 161.444 2151.32 161.444 Q2154.86 161.444 2156.99 158.944 Q2159.15 156.421 2159.15 152.254 L2159.15 151.305 L2154.91 151.305 M2163.41 149.546 L2163.41 164.338 L2159.15 164.338 L2159.15 160.402 Q2157.69 162.763 2155.51 163.898 Q2153.34 165.009 2150.19 165.009 Q2146.21 165.009 2143.85 162.787 Q2141.51 160.541 2141.51 156.791 Q2141.51 152.416 2144.42 150.194 Q2147.36 147.972 2153.17 147.972 L2159.15 147.972 L2159.15 147.555 Q2159.15 144.615 2157.2 143.018 Q2155.28 141.398 2151.79 141.398 Q2149.56 141.398 2147.46 141.93 Q2145.35 142.463 2143.41 143.527 L2143.41 139.592 Q2145.74 138.69 2147.94 138.25 Q2150.14 137.787 2152.23 137.787 Q2157.85 137.787 2160.63 140.703 Q2163.41 143.62 2163.41 149.546 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2176.39 131.051 L2176.39 138.412 L2185.17 138.412 L2185.17 141.722 L2176.39 141.722 L2176.39 155.796 Q2176.39 158.967 2177.25 159.87 Q2178.13 160.773 2180.79 160.773 L2185.17 160.773 L2185.17 164.338 L2180.79 164.338 Q2175.86 164.338 2173.98 162.509 Q2172.11 160.657 2172.11 155.796 L2172.11 141.722 L2168.98 141.722 L2168.98 138.412 L2172.11 138.412 L2172.11 131.051 L2176.39 131.051 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2190.77 138.412 L2195.03 138.412 L2195.03 164.338 L2190.77 164.338 L2190.77 138.412 M2190.77 128.319 L2195.03 128.319 L2195.03 133.713 L2190.77 133.713 L2190.77 128.319 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2213.98 141.398 Q2210.56 141.398 2208.57 144.083 Q2206.58 146.745 2206.58 151.398 Q2206.58 156.051 2208.54 158.736 Q2210.54 161.398 2213.98 161.398 Q2217.39 161.398 2219.38 158.713 Q2221.37 156.027 2221.37 151.398 Q2221.37 146.791 2219.38 144.106 Q2217.39 141.398 2213.98 141.398 M2213.98 137.787 Q2219.54 137.787 2222.71 141.398 Q2225.88 145.009 2225.88 151.398 Q2225.88 157.764 2222.71 161.398 Q2219.54 165.009 2213.98 165.009 Q2208.41 165.009 2205.23 161.398 Q2202.09 157.764 2202.09 151.398 Q2202.09 145.009 2205.23 141.398 Q2208.41 137.787 2213.98 137.787 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2254.49 148.689 L2254.49 164.338 L2250.23 164.338 L2250.23 148.828 Q2250.23 145.148 2248.8 143.319 Q2247.36 141.49 2244.49 141.49 Q2241.04 141.49 2239.05 143.69 Q2237.06 145.889 2237.06 149.685 L2237.06 164.338 L2232.78 164.338 L2232.78 138.412 L2237.06 138.412 L2237.06 142.44 Q2238.59 140.102 2240.65 138.944 Q2242.73 137.787 2245.44 137.787 Q2249.91 137.787 2252.2 140.565 Q2254.49 143.319 2254.49 148.689 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M1809.43 219.634 L1956.83 219.634 L1956.83 178.162 L1809.43 178.162 L1809.43 219.634  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="0.5"/>
+<polyline clip-path="url(#clip430)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1809.43,219.634 1956.83,219.634 1956.83,178.162 1809.43,178.162 1809.43,219.634 "/>
+<path clip-path="url(#clip430)" d="M2009.26 184.28 L2009.26 189.21 Q2006.9 187.011 2004.22 185.923 Q2001.56 184.835 1998.55 184.835 Q1992.62 184.835 1989.47 188.469 Q1986.32 192.08 1986.32 198.932 Q1986.32 205.761 1989.47 209.395 Q1992.62 213.006 1998.55 213.006 Q2001.56 213.006 2004.22 211.918 Q2006.9 210.83 2009.26 208.631 L2009.26 213.516 Q2006.81 215.182 2004.06 216.016 Q2001.32 216.849 1998.27 216.849 Q1990.42 216.849 1985.91 212.057 Q1981.39 207.242 1981.39 198.932 Q1981.39 190.599 1985.91 185.807 Q1990.42 180.993 1998.27 180.993 Q2001.37 180.993 2004.1 181.826 Q2006.86 182.636 2009.26 184.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2025.4 216.178 L2012.2 181.618 L2017.09 181.618 L2028.04 210.715 L2039.01 181.618 L2043.87 181.618 L2030.7 216.178 L2025.4 216.178 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2061.9 203.747 Q2061.9 207.057 2063.55 208.955 Q2065.19 210.83 2068.06 210.83 Q2070.91 210.83 2072.53 208.932 Q2074.17 207.034 2074.17 203.747 Q2074.17 200.506 2072.5 198.608 Q2070.84 196.687 2068.01 196.687 Q2065.21 196.687 2063.55 198.585 Q2061.9 200.483 2061.9 203.747 M2074.52 210.668 Q2073.13 212.451 2071.32 213.307 Q2069.54 214.141 2067.16 214.141 Q2063.18 214.141 2060.68 211.27 Q2058.2 208.377 2058.2 203.747 Q2058.2 199.117 2060.7 196.224 Q2063.2 193.33 2067.16 193.33 Q2069.54 193.33 2071.35 194.21 Q2073.15 195.067 2074.52 196.826 L2074.52 193.793 L2077.83 193.793 L2077.83 210.83 Q2081.21 210.321 2083.11 207.752 Q2085.03 205.159 2085.03 201.062 Q2085.03 198.585 2084.29 196.409 Q2083.57 194.233 2082.09 192.381 Q2079.68 189.349 2076.21 187.752 Q2072.76 186.131 2068.68 186.131 Q2065.84 186.131 2063.22 186.895 Q2060.61 187.636 2058.38 189.118 Q2054.75 191.479 2052.69 195.321 Q2050.65 199.141 2050.65 203.608 Q2050.65 207.289 2051.97 210.506 Q2053.31 213.724 2055.84 216.178 Q2058.27 218.585 2061.46 219.835 Q2064.66 221.108 2068.29 221.108 Q2071.28 221.108 2074.15 220.09 Q2077.04 219.094 2079.45 217.219 L2081.53 219.789 Q2078.64 222.034 2075.21 223.215 Q2071.81 224.418 2068.29 224.418 Q2064.01 224.418 2060.21 222.89 Q2056.42 221.386 2053.45 218.492 Q2050.49 215.599 2048.94 211.803 Q2047.39 207.983 2047.39 203.608 Q2047.39 199.395 2048.96 195.576 Q2050.54 191.756 2053.45 188.863 Q2056.44 185.923 2060.35 184.372 Q2064.26 182.798 2068.64 182.798 Q2073.55 182.798 2077.74 184.812 Q2081.95 186.826 2084.8 190.53 Q2086.53 192.798 2087.43 195.46 Q2088.36 198.122 2088.36 200.969 Q2088.36 207.057 2084.68 210.576 Q2081 214.094 2074.52 214.233 L2074.52 210.668 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip430)" d="M2112.71 199.974 Q2114.22 200.483 2115.63 202.15 Q2117.06 203.817 2118.5 206.733 L2123.24 216.178 L2118.22 216.178 L2113.8 207.312 Q2112.09 203.84 2110.47 202.705 Q2108.87 201.571 2106.09 201.571 L2101 201.571 L2101 216.178 L2096.32 216.178 L2096.32 181.618 L2106.88 181.618 Q2112.8 181.618 2115.72 184.094 Q2118.64 186.571 2118.64 191.571 Q2118.64 194.835 2117.11 196.988 Q2115.61 199.141 2112.71 199.974 M2101 185.46 L2101 197.729 L2106.88 197.729 Q2110.26 197.729 2111.97 196.178 Q2113.71 194.604 2113.71 191.571 Q2113.71 188.539 2111.97 187.011 Q2110.26 185.46 2106.88 185.46 L2101 185.46 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/applications/two_stage_stochastic/27244c58.svg b/previews/PR3525/tutorials/applications/two_stage_stochastic/27244c58.svg
new file mode 100644
index 00000000000..7fd994a0031
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/two_stage_stochastic/27244c58.svg
@@ -0,0 +1,100 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip250">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip250)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip251">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip250)" d="M138.867 1423.18 L2352.76 1423.18 L2352.76 47.2441 L138.867 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip252">
+    <rect x="138" y="47" width="2215" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="370.099,1423.18 370.099,47.2441 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="807.955,1423.18 807.955,47.2441 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1245.81,1423.18 1245.81,47.2441 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1683.67,1423.18 1683.67,47.2441 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2121.52,1423.18 2121.52,47.2441 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="370.099,1423.18 370.099,1404.28 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="807.955,1423.18 807.955,1404.28 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1245.81,1423.18 1245.81,1404.28 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1683.67,1423.18 1683.67,1404.28 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2121.52,1423.18 2121.52,1404.28 "/>
+<path clip-path="url(#clip250)" d="M329.705 1481.64 L337.344 1481.64 L337.344 1455.28 L329.034 1456.95 L329.034 1452.69 L337.298 1451.02 L341.974 1451.02 L341.974 1481.64 L349.613 1481.64 L349.613 1485.58 L329.705 1485.58 L329.705 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M369.636 1466.44 Q366.488 1466.44 364.636 1468.59 Q362.807 1470.74 362.807 1474.49 Q362.807 1478.22 364.636 1480.39 Q366.488 1482.55 369.636 1482.55 Q372.784 1482.55 374.613 1480.39 Q376.464 1478.22 376.464 1474.49 Q376.464 1470.74 374.613 1468.59 Q372.784 1466.44 369.636 1466.44 M378.918 1451.78 L378.918 1456.04 Q377.159 1455.21 375.353 1454.77 Q373.571 1454.33 371.812 1454.33 Q367.182 1454.33 364.728 1457.45 Q362.298 1460.58 361.951 1466.9 Q363.316 1464.89 365.377 1463.82 Q367.437 1462.73 369.914 1462.73 Q375.122 1462.73 378.131 1465.9 Q381.164 1469.05 381.164 1474.49 Q381.164 1479.82 378.015 1483.03 Q374.867 1486.25 369.636 1486.25 Q363.64 1486.25 360.469 1481.67 Q357.298 1477.06 357.298 1468.33 Q357.298 1460.14 361.187 1455.28 Q365.076 1450.39 371.627 1450.39 Q373.386 1450.39 375.168 1450.74 Q376.974 1451.09 378.918 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M399.219 1454.1 Q395.608 1454.1 393.779 1457.66 Q391.974 1461.2 391.974 1468.33 Q391.974 1475.44 393.779 1479.01 Q395.608 1482.55 399.219 1482.55 Q402.853 1482.55 404.659 1479.01 Q406.487 1475.44 406.487 1468.33 Q406.487 1461.2 404.659 1457.66 Q402.853 1454.1 399.219 1454.1 M399.219 1450.39 Q405.029 1450.39 408.085 1455 Q411.163 1459.58 411.163 1468.33 Q411.163 1477.06 408.085 1481.67 Q405.029 1486.25 399.219 1486.25 Q393.409 1486.25 390.33 1481.67 Q387.275 1477.06 387.275 1468.33 Q387.275 1459.58 390.33 1455 Q393.409 1450.39 399.219 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M767.562 1481.64 L775.201 1481.64 L775.201 1455.28 L766.89 1456.95 L766.89 1452.69 L775.154 1451.02 L779.83 1451.02 L779.83 1481.64 L787.469 1481.64 L787.469 1485.58 L767.562 1485.58 L767.562 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M806.913 1469.17 Q803.58 1469.17 801.659 1470.95 Q799.761 1472.73 799.761 1475.86 Q799.761 1478.98 801.659 1480.77 Q803.58 1482.55 806.913 1482.55 Q810.247 1482.55 812.168 1480.77 Q814.089 1478.96 814.089 1475.86 Q814.089 1472.73 812.168 1470.95 Q810.27 1469.17 806.913 1469.17 M802.238 1467.18 Q799.228 1466.44 797.538 1464.38 Q795.872 1462.32 795.872 1459.35 Q795.872 1455.21 798.812 1452.8 Q801.775 1450.39 806.913 1450.39 Q812.075 1450.39 815.015 1452.8 Q817.955 1455.21 817.955 1459.35 Q817.955 1462.32 816.265 1464.38 Q814.599 1466.44 811.612 1467.18 Q814.992 1467.96 816.867 1470.26 Q818.765 1472.55 818.765 1475.86 Q818.765 1480.88 815.687 1483.57 Q812.631 1486.25 806.913 1486.25 Q801.196 1486.25 798.117 1483.57 Q795.062 1480.88 795.062 1475.86 Q795.062 1472.55 796.96 1470.26 Q798.858 1467.96 802.238 1467.18 M800.525 1459.79 Q800.525 1462.48 802.191 1463.98 Q803.881 1465.49 806.913 1465.49 Q809.923 1465.49 811.612 1463.98 Q813.325 1462.48 813.325 1459.79 Q813.325 1457.11 811.612 1455.6 Q809.923 1454.1 806.913 1454.1 Q803.881 1454.1 802.191 1455.6 Q800.525 1457.11 800.525 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M837.075 1454.1 Q833.464 1454.1 831.635 1457.66 Q829.83 1461.2 829.83 1468.33 Q829.83 1475.44 831.635 1479.01 Q833.464 1482.55 837.075 1482.55 Q840.71 1482.55 842.515 1479.01 Q844.344 1475.44 844.344 1468.33 Q844.344 1461.2 842.515 1457.66 Q840.71 1454.1 837.075 1454.1 M837.075 1450.39 Q842.885 1450.39 845.941 1455 Q849.02 1459.58 849.02 1468.33 Q849.02 1477.06 845.941 1481.67 Q842.885 1486.25 837.075 1486.25 Q831.265 1486.25 828.186 1481.67 Q825.131 1477.06 825.131 1468.33 Q825.131 1459.58 828.186 1455 Q831.265 1450.39 837.075 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1209.5 1481.64 L1225.82 1481.64 L1225.82 1485.58 L1203.88 1485.58 L1203.88 1481.64 Q1206.54 1478.89 1211.12 1474.26 Q1215.73 1469.61 1216.91 1468.27 Q1219.16 1465.74 1220.04 1464.01 Q1220.94 1462.25 1220.94 1460.56 Q1220.94 1457.8 1218.99 1456.07 Q1217.07 1454.33 1213.97 1454.33 Q1211.77 1454.33 1209.32 1455.09 Q1206.89 1455.86 1204.11 1457.41 L1204.11 1452.69 Q1206.93 1451.55 1209.39 1450.97 Q1211.84 1450.39 1213.88 1450.39 Q1219.25 1450.39 1222.44 1453.08 Q1225.64 1455.77 1225.64 1460.26 Q1225.64 1462.39 1224.83 1464.31 Q1224.04 1466.2 1221.93 1468.8 Q1221.36 1469.47 1218.25 1472.69 Q1215.15 1475.88 1209.5 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1245.64 1454.1 Q1242.03 1454.1 1240.2 1457.66 Q1238.39 1461.2 1238.39 1468.33 Q1238.39 1475.44 1240.2 1479.01 Q1242.03 1482.55 1245.64 1482.55 Q1249.27 1482.55 1251.08 1479.01 Q1252.91 1475.44 1252.91 1468.33 Q1252.91 1461.2 1251.08 1457.66 Q1249.27 1454.1 1245.64 1454.1 M1245.64 1450.39 Q1251.45 1450.39 1254.5 1455 Q1257.58 1459.58 1257.58 1468.33 Q1257.58 1477.06 1254.5 1481.67 Q1251.45 1486.25 1245.64 1486.25 Q1239.83 1486.25 1236.75 1481.67 Q1233.69 1477.06 1233.69 1468.33 Q1233.69 1459.58 1236.75 1455 Q1239.83 1450.39 1245.64 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1275.8 1454.1 Q1272.19 1454.1 1270.36 1457.66 Q1268.55 1461.2 1268.55 1468.33 Q1268.55 1475.44 1270.36 1479.01 Q1272.19 1482.55 1275.8 1482.55 Q1279.43 1482.55 1281.24 1479.01 Q1283.07 1475.44 1283.07 1468.33 Q1283.07 1461.2 1281.24 1457.66 Q1279.43 1454.1 1275.8 1454.1 M1275.8 1450.39 Q1281.61 1450.39 1284.67 1455 Q1287.74 1459.58 1287.74 1468.33 Q1287.74 1477.06 1284.67 1481.67 Q1281.61 1486.25 1275.8 1486.25 Q1269.99 1486.25 1266.91 1481.67 Q1263.86 1477.06 1263.86 1468.33 Q1263.86 1459.58 1266.91 1455 Q1269.99 1450.39 1275.8 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1647.36 1481.64 L1663.68 1481.64 L1663.68 1485.58 L1641.74 1485.58 L1641.74 1481.64 Q1644.4 1478.89 1648.98 1474.26 Q1653.59 1469.61 1654.77 1468.27 Q1657.01 1465.74 1657.89 1464.01 Q1658.8 1462.25 1658.8 1460.56 Q1658.8 1457.8 1656.85 1456.07 Q1654.93 1454.33 1651.83 1454.33 Q1649.63 1454.33 1647.17 1455.09 Q1644.74 1455.86 1641.97 1457.41 L1641.97 1452.69 Q1644.79 1451.55 1647.24 1450.97 Q1649.7 1450.39 1651.73 1450.39 Q1657.11 1450.39 1660.3 1453.08 Q1663.49 1455.77 1663.49 1460.26 Q1663.49 1462.39 1662.68 1464.31 Q1661.9 1466.2 1659.79 1468.8 Q1659.21 1469.47 1656.11 1472.69 Q1653.01 1475.88 1647.36 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1677.52 1481.64 L1693.84 1481.64 L1693.84 1485.58 L1671.9 1485.58 L1671.9 1481.64 Q1674.56 1478.89 1679.14 1474.26 Q1683.75 1469.61 1684.93 1468.27 Q1687.17 1465.74 1688.05 1464.01 Q1688.96 1462.25 1688.96 1460.56 Q1688.96 1457.8 1687.01 1456.07 Q1685.09 1454.33 1681.99 1454.33 Q1679.79 1454.33 1677.34 1455.09 Q1674.91 1455.86 1672.13 1457.41 L1672.13 1452.69 Q1674.95 1451.55 1677.41 1450.97 Q1679.86 1450.39 1681.9 1450.39 Q1687.27 1450.39 1690.46 1453.08 Q1693.66 1455.77 1693.66 1460.26 Q1693.66 1462.39 1692.85 1464.31 Q1692.06 1466.2 1689.95 1468.8 Q1689.37 1469.47 1686.27 1472.69 Q1683.17 1475.88 1677.52 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1713.66 1454.1 Q1710.04 1454.1 1708.22 1457.66 Q1706.41 1461.2 1706.41 1468.33 Q1706.41 1475.44 1708.22 1479.01 Q1710.04 1482.55 1713.66 1482.55 Q1717.29 1482.55 1719.1 1479.01 Q1720.92 1475.44 1720.92 1468.33 Q1720.92 1461.2 1719.1 1457.66 Q1717.29 1454.1 1713.66 1454.1 M1713.66 1450.39 Q1719.47 1450.39 1722.52 1455 Q1725.6 1459.58 1725.6 1468.33 Q1725.6 1477.06 1722.52 1481.67 Q1719.47 1486.25 1713.66 1486.25 Q1707.85 1486.25 1704.77 1481.67 Q1701.71 1477.06 1701.71 1468.33 Q1701.71 1459.58 1704.77 1455 Q1707.85 1450.39 1713.66 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M2085.22 1481.64 L2101.54 1481.64 L2101.54 1485.58 L2079.59 1485.58 L2079.59 1481.64 Q2082.25 1478.89 2086.84 1474.26 Q2091.44 1469.61 2092.62 1468.27 Q2094.87 1465.74 2095.75 1464.01 Q2096.65 1462.25 2096.65 1460.56 Q2096.65 1457.8 2094.71 1456.07 Q2092.79 1454.33 2089.68 1454.33 Q2087.48 1454.33 2085.03 1455.09 Q2082.6 1455.86 2079.82 1457.41 L2079.82 1452.69 Q2082.65 1451.55 2085.1 1450.97 Q2087.55 1450.39 2089.59 1450.39 Q2094.96 1450.39 2098.16 1453.08 Q2101.35 1455.77 2101.35 1460.26 Q2101.35 1462.39 2100.54 1464.31 Q2099.75 1466.2 2097.65 1468.8 Q2097.07 1469.47 2093.97 1472.69 Q2090.86 1475.88 2085.22 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M2124.2 1455.09 L2112.39 1473.54 L2124.2 1473.54 L2124.2 1455.09 M2122.97 1451.02 L2128.85 1451.02 L2128.85 1473.54 L2133.78 1473.54 L2133.78 1477.43 L2128.85 1477.43 L2128.85 1485.58 L2124.2 1485.58 L2124.2 1477.43 L2108.6 1477.43 L2108.6 1472.92 L2122.97 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M2151.51 1454.1 Q2147.9 1454.1 2146.07 1457.66 Q2144.27 1461.2 2144.27 1468.33 Q2144.27 1475.44 2146.07 1479.01 Q2147.9 1482.55 2151.51 1482.55 Q2155.15 1482.55 2156.95 1479.01 Q2158.78 1475.44 2158.78 1468.33 Q2158.78 1461.2 2156.95 1457.66 Q2155.15 1454.1 2151.51 1454.1 M2151.51 1450.39 Q2157.32 1450.39 2160.38 1455 Q2163.46 1459.58 2163.46 1468.33 Q2163.46 1477.06 2160.38 1481.67 Q2157.32 1486.25 2151.51 1486.25 Q2145.7 1486.25 2142.62 1481.67 Q2139.57 1477.06 2139.57 1468.33 Q2139.57 1459.58 2142.62 1455 Q2145.7 1450.39 2151.51 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1120.18 1525.81 L1120.18 1562.76 L1127.95 1562.76 Q1137.79 1562.76 1142.34 1558.3 Q1146.92 1553.85 1146.92 1544.24 Q1146.92 1534.69 1142.34 1530.26 Q1137.79 1525.81 1127.95 1525.81 L1120.18 1525.81 M1113.75 1520.52 L1126.96 1520.52 Q1140.78 1520.52 1147.24 1526.28 Q1153.7 1532.01 1153.7 1544.24 Q1153.7 1556.52 1147.21 1562.28 Q1140.71 1568.04 1126.96 1568.04 L1113.75 1568.04 L1113.75 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1194.19 1548.76 L1194.19 1551.62 L1167.26 1551.62 Q1167.64 1557.67 1170.89 1560.85 Q1174.17 1564 1179.99 1564 Q1183.36 1564 1186.51 1563.17 Q1189.7 1562.35 1192.82 1560.69 L1192.82 1566.23 Q1189.67 1567.57 1186.36 1568.27 Q1183.05 1568.97 1179.64 1568.97 Q1171.11 1568.97 1166.11 1564 Q1161.15 1559.04 1161.15 1550.57 Q1161.15 1541.82 1165.86 1536.69 Q1170.6 1531.54 1178.62 1531.54 Q1185.81 1531.54 1189.98 1536.18 Q1194.19 1540.8 1194.19 1548.76 M1188.33 1547.04 Q1188.27 1542.23 1185.62 1539.37 Q1183.01 1536.5 1178.69 1536.5 Q1173.78 1536.5 1170.82 1539.27 Q1167.9 1542.04 1167.45 1547.07 L1188.33 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1231.55 1539.24 Q1233.75 1535.29 1236.8 1533.41 Q1239.86 1531.54 1244 1531.54 Q1249.57 1531.54 1252.59 1535.45 Q1255.61 1539.33 1255.61 1546.53 L1255.61 1568.04 L1249.73 1568.04 L1249.73 1546.72 Q1249.73 1541.59 1247.91 1539.11 Q1246.1 1536.63 1242.37 1536.63 Q1237.82 1536.63 1235.18 1539.65 Q1232.54 1542.68 1232.54 1547.9 L1232.54 1568.04 L1226.65 1568.04 L1226.65 1546.72 Q1226.65 1541.56 1224.84 1539.11 Q1223.02 1536.63 1219.23 1536.63 Q1214.75 1536.63 1212.1 1539.68 Q1209.46 1542.71 1209.46 1547.9 L1209.46 1568.04 L1203.57 1568.04 L1203.57 1532.4 L1209.46 1532.4 L1209.46 1537.93 Q1211.47 1534.66 1214.27 1533.1 Q1217.07 1531.54 1220.92 1531.54 Q1224.8 1531.54 1227.51 1533.51 Q1230.25 1535.48 1231.55 1539.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1283.5 1550.12 Q1276.4 1550.12 1273.66 1551.75 Q1270.92 1553.37 1270.92 1557.29 Q1270.92 1560.4 1272.96 1562.25 Q1275.03 1564.07 1278.56 1564.07 Q1283.43 1564.07 1286.36 1560.63 Q1289.32 1557.16 1289.32 1551.43 L1289.32 1550.12 L1283.5 1550.12 M1295.18 1547.71 L1295.18 1568.04 L1289.32 1568.04 L1289.32 1562.63 Q1287.32 1565.88 1284.32 1567.44 Q1281.33 1568.97 1277 1568.97 Q1271.53 1568.97 1268.28 1565.91 Q1265.07 1562.82 1265.07 1557.67 Q1265.07 1551.65 1269.08 1548.6 Q1273.12 1545.54 1281.11 1545.54 L1289.32 1545.54 L1289.32 1544.97 Q1289.32 1540.93 1286.65 1538.73 Q1284.01 1536.5 1279.2 1536.5 Q1276.14 1536.5 1273.25 1537.23 Q1270.35 1537.97 1267.68 1539.43 L1267.68 1534.02 Q1270.89 1532.78 1273.92 1532.17 Q1276.94 1531.54 1279.8 1531.54 Q1287.54 1531.54 1291.36 1535.55 Q1295.18 1539.56 1295.18 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1336.87 1546.53 L1336.87 1568.04 L1331.02 1568.04 L1331.02 1546.72 Q1331.02 1541.66 1329.04 1539.14 Q1327.07 1536.63 1323.12 1536.63 Q1318.38 1536.63 1315.64 1539.65 Q1312.91 1542.68 1312.91 1547.9 L1312.91 1568.04 L1307.02 1568.04 L1307.02 1532.4 L1312.91 1532.4 L1312.91 1537.93 Q1315.01 1534.72 1317.84 1533.13 Q1320.7 1531.54 1324.43 1531.54 Q1330.57 1531.54 1333.72 1535.36 Q1336.87 1539.14 1336.87 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M1372.01 1537.81 L1372.01 1518.52 L1377.87 1518.52 L1377.87 1568.04 L1372.01 1568.04 L1372.01 1562.7 Q1370.17 1565.88 1367.33 1567.44 Q1364.53 1568.97 1360.59 1568.97 Q1354.12 1568.97 1350.05 1563.81 Q1346.01 1558.65 1346.01 1550.25 Q1346.01 1541.85 1350.05 1536.69 Q1354.12 1531.54 1360.59 1531.54 Q1364.53 1531.54 1367.33 1533.1 Q1370.17 1534.62 1372.01 1537.81 M1352.06 1550.25 Q1352.06 1556.71 1354.7 1560.4 Q1357.37 1564.07 1362.02 1564.07 Q1366.66 1564.07 1369.34 1560.4 Q1372.01 1556.71 1372.01 1550.25 Q1372.01 1543.79 1369.34 1540.13 Q1366.66 1536.44 1362.02 1536.44 Q1357.37 1536.44 1354.7 1540.13 Q1352.06 1543.79 1352.06 1550.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="138.867,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="138.867,1193.86 2352.76,1193.86 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="138.867,964.535 2352.76,964.535 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="138.867,735.212 2352.76,735.212 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="138.867,505.889 2352.76,505.889 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="138.867,276.567 2352.76,276.567 "/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="138.867,47.2441 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,1423.18 138.867,47.2441 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,1423.18 157.764,1423.18 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,1193.86 157.764,1193.86 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,964.535 157.764,964.535 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,735.212 157.764,735.212 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,505.889 157.764,505.889 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,276.567 157.764,276.567 "/>
+<polyline clip-path="url(#clip250)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="138.867,47.2441 157.764,47.2441 "/>
+<path clip-path="url(#clip250)" d="M90.9225 1408.98 Q87.3114 1408.98 85.4827 1412.54 Q83.6771 1416.08 83.6771 1423.21 Q83.6771 1430.32 85.4827 1433.89 Q87.3114 1437.43 90.9225 1437.43 Q94.5567 1437.43 96.3623 1433.89 Q98.1909 1430.32 98.1909 1423.21 Q98.1909 1416.08 96.3623 1412.54 Q94.5567 1408.98 90.9225 1408.98 M90.9225 1405.27 Q96.7326 1405.27 99.7882 1409.88 Q102.867 1414.46 102.867 1423.21 Q102.867 1431.94 99.7882 1436.55 Q96.7326 1441.13 90.9225 1441.13 Q85.1123 1441.13 82.0336 1436.55 Q78.9781 1431.94 78.9781 1423.21 Q78.9781 1414.46 82.0336 1409.88 Q85.1123 1405.27 90.9225 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M95.7373 1192.5 Q99.0937 1193.22 100.969 1195.49 Q102.867 1197.76 102.867 1201.09 Q102.867 1206.21 99.3483 1209.01 Q95.8298 1211.81 89.3484 1211.81 Q87.1725 1211.81 84.8577 1211.37 Q82.566 1210.95 80.1123 1210.1 L80.1123 1205.58 Q82.0568 1206.72 84.3716 1207.29 Q86.6864 1207.87 89.2095 1207.87 Q93.6076 1207.87 95.8993 1206.14 Q98.2141 1204.4 98.2141 1201.09 Q98.2141 1198.04 96.0613 1196.32 Q93.9317 1194.59 90.1123 1194.59 L86.0845 1194.59 L86.0845 1190.74 L90.2975 1190.74 Q93.7465 1190.74 95.5752 1189.38 Q97.4039 1187.99 97.4039 1185.4 Q97.4039 1182.73 95.5058 1181.32 Q93.6308 1179.89 90.1123 1179.89 Q88.191 1179.89 85.9919 1180.3 Q83.7929 1180.72 81.154 1181.6 L81.154 1177.43 Q83.816 1176.69 86.1308 1176.32 Q88.4688 1175.95 90.529 1175.95 Q95.853 1175.95 98.9548 1178.38 Q102.057 1180.79 102.057 1184.91 Q102.057 1187.78 100.413 1189.77 Q98.7696 1191.74 95.7373 1192.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M91.3391 962.671 Q88.191 962.671 86.3392 964.824 Q84.5105 966.977 84.5105 970.727 Q84.5105 974.454 86.3392 976.629 Q88.191 978.782 91.3391 978.782 Q94.4873 978.782 96.316 976.629 Q98.1678 974.454 98.1678 970.727 Q98.1678 966.977 96.316 964.824 Q94.4873 962.671 91.3391 962.671 M100.621 948.018 L100.621 952.278 Q98.8622 951.444 97.0567 951.005 Q95.2743 950.565 93.515 950.565 Q88.8854 950.565 86.4318 953.69 Q84.0012 956.815 83.654 963.134 Q85.0197 961.12 87.0799 960.055 Q89.1401 958.967 91.6169 958.967 Q96.8252 958.967 99.8345 962.139 Q102.867 965.287 102.867 970.727 Q102.867 976.051 99.7187 979.268 Q96.5706 982.486 91.3391 982.486 Q85.3438 982.486 82.1725 977.903 Q79.0012 973.296 79.0012 964.569 Q79.0012 956.375 82.8901 951.514 Q86.779 946.63 93.3299 946.63 Q95.0891 946.63 96.8715 946.977 Q98.6771 947.324 100.621 948.018 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M81.2466 751.774 L81.2466 747.515 Q83.0058 748.348 84.8114 748.788 Q86.6169 749.228 88.353 749.228 Q92.9826 749.228 95.4132 746.126 Q97.8669 743.001 98.2141 736.659 Q96.8715 738.649 94.8113 739.714 Q92.7512 740.779 90.2512 740.779 Q85.066 740.779 82.0336 737.654 Q79.0244 734.506 79.0244 729.066 Q79.0244 723.742 82.1725 720.525 Q85.3206 717.307 90.5521 717.307 Q96.5474 717.307 99.6956 721.913 Q102.867 726.497 102.867 735.247 Q102.867 743.418 98.978 748.302 Q95.1123 753.163 88.5614 753.163 Q86.8021 753.163 84.9966 752.816 Q83.191 752.469 81.2466 751.774 M90.5521 737.122 Q93.7002 737.122 95.5289 734.969 Q97.3808 732.816 97.3808 729.066 Q97.3808 725.339 95.5289 723.187 Q93.7002 721.011 90.5521 721.011 Q87.404 721.011 85.5521 723.187 Q83.7234 725.339 83.7234 729.066 Q83.7234 732.816 85.5521 734.969 Q87.404 737.122 90.5521 737.122 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M53.168 519.234 L60.8069 519.234 L60.8069 492.869 L52.4967 494.535 L52.4967 490.276 L60.7606 488.609 L65.4365 488.609 L65.4365 519.234 L73.0753 519.234 L73.0753 523.169 L53.168 523.169 L53.168 519.234 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M86.5475 519.234 L102.867 519.234 L102.867 523.169 L80.9225 523.169 L80.9225 519.234 Q83.5845 516.48 88.1679 511.85 Q92.7743 507.197 93.9549 505.855 Q96.2002 503.331 97.0798 501.595 Q97.9826 499.836 97.9826 498.146 Q97.9826 495.392 96.0382 493.656 Q94.1169 491.92 91.0151 491.92 Q88.816 491.92 86.3623 492.683 Q83.9318 493.447 81.154 494.998 L81.154 490.276 Q83.9781 489.142 86.4318 488.563 Q88.8854 487.984 90.9225 487.984 Q96.2928 487.984 99.4872 490.67 Q102.682 493.355 102.682 497.845 Q102.682 499.975 101.871 501.896 Q101.084 503.794 98.978 506.387 Q98.3993 507.058 95.2974 510.276 Q92.1956 513.47 86.5475 519.234 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M52.5662 289.912 L60.205 289.912 L60.205 263.546 L51.8949 265.213 L51.8949 260.953 L60.1587 259.287 L64.8346 259.287 L64.8346 289.912 L72.4735 289.912 L72.4735 293.847 L52.5662 293.847 L52.5662 289.912 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M81.9642 259.287 L100.321 259.287 L100.321 263.222 L86.2466 263.222 L86.2466 271.694 Q87.2651 271.347 88.2836 271.185 Q89.3021 271 90.3206 271 Q96.1076 271 99.4872 274.171 Q102.867 277.342 102.867 282.759 Q102.867 288.337 99.3946 291.439 Q95.9224 294.518 89.603 294.518 Q87.4271 294.518 85.1586 294.148 Q82.9133 293.777 80.5059 293.037 L80.5059 288.337 Q82.5892 289.472 84.8114 290.027 Q87.0336 290.583 89.5104 290.583 Q93.515 290.583 95.853 288.476 Q98.1909 286.37 98.1909 282.759 Q98.1909 279.148 95.853 277.041 Q93.515 274.935 89.5104 274.935 Q87.6354 274.935 85.7605 275.351 Q83.9086 275.768 81.9642 276.648 L81.9642 259.287 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M51.6634 60.5889 L59.3023 60.5889 L59.3023 34.2233 L50.9921 35.89 L50.9921 31.6308 L59.256 29.9641 L63.9319 29.9641 L63.9319 60.5889 L71.5707 60.5889 L71.5707 64.5241 L51.6634 64.5241 L51.6634 60.5889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip250)" d="M91.0151 48.1121 Q87.6817 48.1121 85.7605 49.8945 Q83.8623 51.6769 83.8623 54.8019 Q83.8623 57.9269 85.7605 59.7093 Q87.6817 61.4917 91.0151 61.4917 Q94.3484 61.4917 96.2697 59.7093 Q98.1909 57.9038 98.1909 54.8019 Q98.1909 51.6769 96.2697 49.8945 Q94.3715 48.1121 91.0151 48.1121 M86.3392 46.1214 Q83.3299 45.3807 81.6401 43.3205 Q79.9735 41.2603 79.9735 38.2974 Q79.9735 34.1539 82.9133 31.7465 Q85.8762 29.3391 91.0151 29.3391 Q96.1771 29.3391 99.1169 31.7465 Q102.057 34.1539 102.057 38.2974 Q102.057 41.2603 100.367 43.3205 Q98.7002 45.3807 95.7141 46.1214 Q99.0937 46.9084 100.969 49.2001 Q102.867 51.4918 102.867 54.8019 Q102.867 59.825 99.7882 62.5102 Q96.7326 65.1954 91.0151 65.1954 Q85.2975 65.1954 82.2188 62.5102 Q79.1633 59.825 79.1633 54.8019 Q79.1633 51.4918 81.0614 49.2001 Q82.9595 46.9084 86.3392 46.1214 M84.6262 38.7372 Q84.6262 41.4224 86.2929 42.927 Q87.9827 44.4316 91.0151 44.4316 Q94.0243 44.4316 95.7141 42.927 Q97.4271 41.4224 97.4271 38.7372 Q97.4271 36.052 95.7141 34.5474 Q94.0243 33.0428 91.0151 33.0428 Q87.9827 33.0428 86.2929 34.5474 Q84.6262 36.052 84.6262 38.7372 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip252)" d="M260.635 1270.3 L260.635 1423.18 L370.099 1423.18 L370.099 1270.3 L260.635 1270.3 L260.635 1270.3  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="260.635,1270.3 260.635,1423.18 370.099,1423.18 370.099,1270.3 260.635,1270.3 "/>
+<path clip-path="url(#clip252)" d="M370.099 1270.3 L370.099 1423.18 L479.563 1423.18 L479.563 1270.3 L370.099 1270.3 L370.099 1270.3  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="370.099,1270.3 370.099,1423.18 479.563,1423.18 479.563,1270.3 370.099,1270.3 "/>
+<path clip-path="url(#clip252)" d="M479.563 888.094 L479.563 1423.18 L589.027 1423.18 L589.027 888.094 L479.563 888.094 L479.563 888.094  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="479.563,888.094 479.563,1423.18 589.027,1423.18 589.027,888.094 479.563,888.094 "/>
+<path clip-path="url(#clip252)" d="M589.027 811.653 L589.027 1423.18 L698.491 1423.18 L698.491 811.653 L589.027 811.653 L589.027 811.653  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="589.027,811.653 589.027,1423.18 698.491,1423.18 698.491,811.653 589.027,811.653 "/>
+<path clip-path="url(#clip252)" d="M698.491 1193.86 L698.491 1423.18 L807.955 1423.18 L807.955 1193.86 L698.491 1193.86 L698.491 1193.86  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="698.491,1193.86 698.491,1423.18 807.955,1423.18 807.955,1193.86 698.491,1193.86 "/>
+<path clip-path="url(#clip252)" d="M807.955 1193.86 L807.955 1423.18 L917.419 1423.18 L917.419 1193.86 L807.955 1193.86 L807.955 1193.86  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="807.955,1193.86 807.955,1423.18 917.419,1423.18 917.419,1193.86 807.955,1193.86 "/>
+<path clip-path="url(#clip252)" d="M917.419 1117.42 L917.419 1423.18 L1026.88 1423.18 L1026.88 1117.42 L917.419 1117.42 L917.419 1117.42  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="917.419,1117.42 917.419,1423.18 1026.88,1423.18 1026.88,1117.42 917.419,1117.42 "/>
+<path clip-path="url(#clip252)" d="M1026.88 735.212 L1026.88 1423.18 L1136.35 1423.18 L1136.35 735.212 L1026.88 735.212 L1026.88 735.212  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1026.88,735.212 1026.88,1423.18 1136.35,1423.18 1136.35,735.212 1026.88,735.212 "/>
+<path clip-path="url(#clip252)" d="M1136.35 47.2441 L1136.35 1423.18 L1245.81 1423.18 L1245.81 47.2441 L1136.35 47.2441 L1136.35 47.2441  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1136.35,47.2441 1136.35,1423.18 1245.81,1423.18 1245.81,47.2441 1136.35,47.2441 "/>
+<path clip-path="url(#clip252)" d="M1245.81 582.33 L1245.81 1423.18 L1355.28 1423.18 L1355.28 582.33 L1245.81 582.33 L1245.81 582.33  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1245.81,582.33 1245.81,1423.18 1355.28,1423.18 1355.28,582.33 1245.81,582.33 "/>
+<path clip-path="url(#clip252)" d="M1355.28 811.653 L1355.28 1423.18 L1464.74 1423.18 L1464.74 811.653 L1355.28 811.653 L1355.28 811.653  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1355.28,811.653 1355.28,1423.18 1464.74,1423.18 1464.74,811.653 1355.28,811.653 "/>
+<path clip-path="url(#clip252)" d="M1464.74 964.535 L1464.74 1423.18 L1574.2 1423.18 L1574.2 964.535 L1464.74 964.535 L1464.74 964.535  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1464.74,964.535 1464.74,1423.18 1574.2,1423.18 1574.2,964.535 1464.74,964.535 "/>
+<path clip-path="url(#clip252)" d="M1574.2 1040.98 L1574.2 1423.18 L1683.67 1423.18 L1683.67 1040.98 L1574.2 1040.98 L1574.2 1040.98  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1574.2,1040.98 1574.2,1423.18 1683.67,1423.18 1683.67,1040.98 1574.2,1040.98 "/>
+<path clip-path="url(#clip252)" d="M1683.67 964.535 L1683.67 1423.18 L1793.13 1423.18 L1793.13 964.535 L1683.67 964.535 L1683.67 964.535  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1683.67,964.535 1683.67,1423.18 1793.13,1423.18 1793.13,964.535 1683.67,964.535 "/>
+<path clip-path="url(#clip252)" d="M1793.13 1270.3 L1793.13 1423.18 L1902.6 1423.18 L1902.6 1270.3 L1793.13 1270.3 L1793.13 1270.3  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1793.13,1270.3 1793.13,1423.18 1902.6,1423.18 1902.6,1270.3 1793.13,1270.3 "/>
+<path clip-path="url(#clip252)" d="M1902.6 1193.86 L1902.6 1423.18 L2012.06 1423.18 L2012.06 1193.86 L1902.6 1193.86 L1902.6 1193.86  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1902.6,1193.86 1902.6,1423.18 2012.06,1423.18 2012.06,1193.86 1902.6,1193.86 "/>
+<path clip-path="url(#clip252)" d="M2012.06 1270.3 L2012.06 1423.18 L2121.52 1423.18 L2121.52 1270.3 L2012.06 1270.3 L2012.06 1270.3  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2012.06,1270.3 2012.06,1423.18 2121.52,1423.18 2121.52,1270.3 2012.06,1270.3 "/>
+<path clip-path="url(#clip252)" d="M2121.52 1346.74 L2121.52 1423.18 L2230.99 1423.18 L2230.99 1346.74 L2121.52 1346.74 L2121.52 1346.74  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip252)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2121.52,1346.74 2121.52,1423.18 2230.99,1423.18 2230.99,1346.74 2121.52,1346.74 "/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="315.367" cy="1270.3" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="424.831" cy="1270.3" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="534.295" cy="888.094" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="643.759" cy="811.653" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="753.223" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="862.687" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="972.151" cy="1117.42" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1081.62" cy="735.212" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1191.08" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1300.54" cy="582.33" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1410.01" cy="811.653" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1519.47" cy="964.535" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1628.94" cy="1040.98" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1738.4" cy="964.535" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1847.86" cy="1270.3" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1957.33" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2066.79" cy="1270.3" r="2"/>
+<circle clip-path="url(#clip252)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2176.26" cy="1346.74" r="2"/>
+</svg>
diff --git a/previews/PR3525/tutorials/applications/two_stage_stochastic/a5d2410e.svg b/previews/PR3525/tutorials/applications/two_stage_stochastic/a5d2410e.svg
new file mode 100644
index 00000000000..87f2fc1f9da
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/two_stage_stochastic/a5d2410e.svg
@@ -0,0 +1,76 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip310">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip310)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip311">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip310)" d="M205.121 1423.18 L2352.76 1423.18 L2352.76 47.2441 L205.121 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip312">
+    <rect x="205" y="47" width="2149" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="514.383,1423.18 514.383,47.2441 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="896.661,1423.18 896.661,47.2441 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1278.94,1423.18 1278.94,47.2441 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1661.22,1423.18 1661.22,47.2441 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2043.49,1423.18 2043.49,47.2441 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="514.383,1423.18 514.383,1404.28 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="896.661,1423.18 896.661,1404.28 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1278.94,1423.18 1278.94,1404.28 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1661.22,1423.18 1661.22,1404.28 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2043.49,1423.18 2043.49,1404.28 "/>
+<path clip-path="url(#clip310)" d="M487.474 1455.09 L475.668 1473.54 L487.474 1473.54 L487.474 1455.09 M486.247 1451.02 L492.127 1451.02 L492.127 1473.54 L497.057 1473.54 L497.057 1477.43 L492.127 1477.43 L492.127 1485.58 L487.474 1485.58 L487.474 1477.43 L471.872 1477.43 L471.872 1472.92 L486.247 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M514.789 1454.1 Q511.177 1454.1 509.349 1457.66 Q507.543 1461.2 507.543 1468.33 Q507.543 1475.44 509.349 1479.01 Q511.177 1482.55 514.789 1482.55 Q518.423 1482.55 520.228 1479.01 Q522.057 1475.44 522.057 1468.33 Q522.057 1461.2 520.228 1457.66 Q518.423 1454.1 514.789 1454.1 M514.789 1450.39 Q520.599 1450.39 523.654 1455 Q526.733 1459.58 526.733 1468.33 Q526.733 1477.06 523.654 1481.67 Q520.599 1486.25 514.789 1486.25 Q508.978 1486.25 505.9 1481.67 Q502.844 1477.06 502.844 1468.33 Q502.844 1459.58 505.9 1455 Q508.978 1450.39 514.789 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M544.95 1454.1 Q541.339 1454.1 539.511 1457.66 Q537.705 1461.2 537.705 1468.33 Q537.705 1475.44 539.511 1479.01 Q541.339 1482.55 544.95 1482.55 Q548.585 1482.55 550.39 1479.01 Q552.219 1475.44 552.219 1468.33 Q552.219 1461.2 550.39 1457.66 Q548.585 1454.1 544.95 1454.1 M544.95 1450.39 Q550.761 1450.39 553.816 1455 Q556.895 1459.58 556.895 1468.33 Q556.895 1477.06 553.816 1481.67 Q550.761 1486.25 544.95 1486.25 Q539.14 1486.25 536.062 1481.67 Q533.006 1477.06 533.006 1468.33 Q533.006 1459.58 536.062 1455 Q539.14 1450.39 544.95 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M869.751 1455.09 L857.946 1473.54 L869.751 1473.54 L869.751 1455.09 M868.525 1451.02 L874.404 1451.02 L874.404 1473.54 L879.335 1473.54 L879.335 1477.43 L874.404 1477.43 L874.404 1485.58 L869.751 1485.58 L869.751 1477.43 L854.15 1477.43 L854.15 1472.92 L868.525 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M887.112 1451.02 L905.469 1451.02 L905.469 1454.96 L891.395 1454.96 L891.395 1463.43 Q892.413 1463.08 893.432 1462.92 Q894.45 1462.73 895.469 1462.73 Q901.256 1462.73 904.635 1465.9 Q908.015 1469.08 908.015 1474.49 Q908.015 1480.07 904.543 1483.17 Q901.071 1486.25 894.751 1486.25 Q892.575 1486.25 890.307 1485.88 Q888.061 1485.51 885.654 1484.77 L885.654 1480.07 Q887.737 1481.2 889.96 1481.76 Q892.182 1482.32 894.659 1482.32 Q898.663 1482.32 901.001 1480.21 Q903.339 1478.1 903.339 1474.49 Q903.339 1470.88 901.001 1468.77 Q898.663 1466.67 894.659 1466.67 Q892.784 1466.67 890.909 1467.08 Q889.057 1467.5 887.112 1468.38 L887.112 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M927.228 1454.1 Q923.617 1454.1 921.788 1457.66 Q919.983 1461.2 919.983 1468.33 Q919.983 1475.44 921.788 1479.01 Q923.617 1482.55 927.228 1482.55 Q930.862 1482.55 932.668 1479.01 Q934.496 1475.44 934.496 1468.33 Q934.496 1461.2 932.668 1457.66 Q930.862 1454.1 927.228 1454.1 M927.228 1450.39 Q933.038 1450.39 936.094 1455 Q939.172 1459.58 939.172 1468.33 Q939.172 1477.06 936.094 1481.67 Q933.038 1486.25 927.228 1486.25 Q921.418 1486.25 918.339 1481.67 Q915.284 1477.06 915.284 1468.33 Q915.284 1459.58 918.339 1455 Q921.418 1450.39 927.228 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1238.56 1451.02 L1256.91 1451.02 L1256.91 1454.96 L1242.84 1454.96 L1242.84 1463.43 Q1243.86 1463.08 1244.88 1462.92 Q1245.89 1462.73 1246.91 1462.73 Q1252.7 1462.73 1256.08 1465.9 Q1259.46 1469.08 1259.46 1474.49 Q1259.46 1480.07 1255.99 1483.17 Q1252.51 1486.25 1246.2 1486.25 Q1244.02 1486.25 1241.75 1485.88 Q1239.51 1485.51 1237.1 1484.77 L1237.1 1480.07 Q1239.18 1481.2 1241.4 1481.76 Q1243.63 1482.32 1246.1 1482.32 Q1250.11 1482.32 1252.45 1480.21 Q1254.78 1478.1 1254.78 1474.49 Q1254.78 1470.88 1252.45 1468.77 Q1250.11 1466.67 1246.1 1466.67 Q1244.23 1466.67 1242.35 1467.08 Q1240.5 1467.5 1238.56 1468.38 L1238.56 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1278.67 1454.1 Q1275.06 1454.1 1273.23 1457.66 Q1271.43 1461.2 1271.43 1468.33 Q1271.43 1475.44 1273.23 1479.01 Q1275.06 1482.55 1278.67 1482.55 Q1282.31 1482.55 1284.11 1479.01 Q1285.94 1475.44 1285.94 1468.33 Q1285.94 1461.2 1284.11 1457.66 Q1282.31 1454.1 1278.67 1454.1 M1278.67 1450.39 Q1284.48 1450.39 1287.54 1455 Q1290.62 1459.58 1290.62 1468.33 Q1290.62 1477.06 1287.54 1481.67 Q1284.48 1486.25 1278.67 1486.25 Q1272.86 1486.25 1269.78 1481.67 Q1266.73 1477.06 1266.73 1468.33 Q1266.73 1459.58 1269.78 1455 Q1272.86 1450.39 1278.67 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1308.83 1454.1 Q1305.22 1454.1 1303.39 1457.66 Q1301.59 1461.2 1301.59 1468.33 Q1301.59 1475.44 1303.39 1479.01 Q1305.22 1482.55 1308.83 1482.55 Q1312.47 1482.55 1314.27 1479.01 Q1316.1 1475.44 1316.1 1468.33 Q1316.1 1461.2 1314.27 1457.66 Q1312.47 1454.1 1308.83 1454.1 M1308.83 1450.39 Q1314.64 1450.39 1317.7 1455 Q1320.78 1459.58 1320.78 1468.33 Q1320.78 1477.06 1317.7 1481.67 Q1314.64 1486.25 1308.83 1486.25 Q1303.02 1486.25 1299.95 1481.67 Q1296.89 1477.06 1296.89 1468.33 Q1296.89 1459.58 1299.95 1455 Q1303.02 1450.39 1308.83 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1620.83 1451.02 L1639.19 1451.02 L1639.19 1454.96 L1625.12 1454.96 L1625.12 1463.43 Q1626.14 1463.08 1627.15 1462.92 Q1628.17 1462.73 1629.19 1462.73 Q1634.98 1462.73 1638.36 1465.9 Q1641.74 1469.08 1641.74 1474.49 Q1641.74 1480.07 1638.26 1483.17 Q1634.79 1486.25 1628.47 1486.25 Q1626.3 1486.25 1624.03 1485.88 Q1621.78 1485.51 1619.38 1484.77 L1619.38 1480.07 Q1621.46 1481.2 1623.68 1481.76 Q1625.9 1482.32 1628.38 1482.32 Q1632.39 1482.32 1634.72 1480.21 Q1637.06 1478.1 1637.06 1474.49 Q1637.06 1470.88 1634.72 1468.77 Q1632.39 1466.67 1628.38 1466.67 Q1626.51 1466.67 1624.63 1467.08 Q1622.78 1467.5 1620.83 1468.38 L1620.83 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1651 1451.02 L1669.35 1451.02 L1669.35 1454.96 L1655.28 1454.96 L1655.28 1463.43 Q1656.3 1463.08 1657.32 1462.92 Q1658.33 1462.73 1659.35 1462.73 Q1665.14 1462.73 1668.52 1465.9 Q1671.9 1469.08 1671.9 1474.49 Q1671.9 1480.07 1668.43 1483.17 Q1664.95 1486.25 1658.63 1486.25 Q1656.46 1486.25 1654.19 1485.88 Q1651.95 1485.51 1649.54 1484.77 L1649.54 1480.07 Q1651.62 1481.2 1653.84 1481.76 Q1656.07 1482.32 1658.54 1482.32 Q1662.55 1482.32 1664.88 1480.21 Q1667.22 1478.1 1667.22 1474.49 Q1667.22 1470.88 1664.88 1468.77 Q1662.55 1466.67 1658.54 1466.67 Q1656.67 1466.67 1654.79 1467.08 Q1652.94 1467.5 1651 1468.38 L1651 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1691.11 1454.1 Q1687.5 1454.1 1685.67 1457.66 Q1683.87 1461.2 1683.87 1468.33 Q1683.87 1475.44 1685.67 1479.01 Q1687.5 1482.55 1691.11 1482.55 Q1694.75 1482.55 1696.55 1479.01 Q1698.38 1475.44 1698.38 1468.33 Q1698.38 1461.2 1696.55 1457.66 Q1694.75 1454.1 1691.11 1454.1 M1691.11 1450.39 Q1696.92 1450.39 1699.98 1455 Q1703.06 1459.58 1703.06 1468.33 Q1703.06 1477.06 1699.98 1481.67 Q1696.92 1486.25 1691.11 1486.25 Q1685.3 1486.25 1682.22 1481.67 Q1679.17 1477.06 1679.17 1468.33 Q1679.17 1459.58 1682.22 1455 Q1685.3 1450.39 1691.11 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M2013.82 1466.44 Q2010.67 1466.44 2008.82 1468.59 Q2006.99 1470.74 2006.99 1474.49 Q2006.99 1478.22 2008.82 1480.39 Q2010.67 1482.55 2013.82 1482.55 Q2016.97 1482.55 2018.79 1480.39 Q2020.65 1478.22 2020.65 1474.49 Q2020.65 1470.74 2018.79 1468.59 Q2016.97 1466.44 2013.82 1466.44 M2023.1 1451.78 L2023.1 1456.04 Q2021.34 1455.21 2019.54 1454.77 Q2017.75 1454.33 2015.99 1454.33 Q2011.36 1454.33 2008.91 1457.45 Q2006.48 1460.58 2006.13 1466.9 Q2007.5 1464.89 2009.56 1463.82 Q2011.62 1462.73 2014.1 1462.73 Q2019.3 1462.73 2022.31 1465.9 Q2025.35 1469.05 2025.35 1474.49 Q2025.35 1479.82 2022.2 1483.03 Q2019.05 1486.25 2013.82 1486.25 Q2007.82 1486.25 2004.65 1481.67 Q2001.48 1477.06 2001.48 1468.33 Q2001.48 1460.14 2005.37 1455.28 Q2009.26 1450.39 2015.81 1450.39 Q2017.57 1450.39 2019.35 1450.74 Q2021.16 1451.09 2023.1 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M2043.4 1454.1 Q2039.79 1454.1 2037.96 1457.66 Q2036.16 1461.2 2036.16 1468.33 Q2036.16 1475.44 2037.96 1479.01 Q2039.79 1482.55 2043.4 1482.55 Q2047.04 1482.55 2048.84 1479.01 Q2050.67 1475.44 2050.67 1468.33 Q2050.67 1461.2 2048.84 1457.66 Q2047.04 1454.1 2043.4 1454.1 M2043.4 1450.39 Q2049.21 1450.39 2052.27 1455 Q2055.35 1459.58 2055.35 1468.33 Q2055.35 1477.06 2052.27 1481.67 Q2049.21 1486.25 2043.4 1486.25 Q2037.59 1486.25 2034.51 1481.67 Q2031.46 1477.06 2031.46 1468.33 Q2031.46 1459.58 2034.51 1455 Q2037.59 1450.39 2043.4 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M2073.56 1454.1 Q2069.95 1454.1 2068.12 1457.66 Q2066.32 1461.2 2066.32 1468.33 Q2066.32 1475.44 2068.12 1479.01 Q2069.95 1482.55 2073.56 1482.55 Q2077.2 1482.55 2079 1479.01 Q2080.83 1475.44 2080.83 1468.33 Q2080.83 1461.2 2079 1457.66 Q2077.2 1454.1 2073.56 1454.1 M2073.56 1450.39 Q2079.37 1450.39 2082.43 1455 Q2085.51 1459.58 2085.51 1468.33 Q2085.51 1477.06 2082.43 1481.67 Q2079.37 1486.25 2073.56 1486.25 Q2067.75 1486.25 2064.67 1481.67 Q2061.62 1477.06 2061.62 1468.33 Q2061.62 1459.58 2064.67 1455 Q2067.75 1450.39 2073.56 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1149.83 1525.81 L1149.83 1543.66 L1157.91 1543.66 Q1162.4 1543.66 1164.85 1541.34 Q1167.3 1539.02 1167.3 1534.72 Q1167.3 1530.45 1164.85 1528.13 Q1162.4 1525.81 1157.91 1525.81 L1149.83 1525.81 M1143.4 1520.52 L1157.91 1520.52 Q1165.9 1520.52 1169.97 1524.15 Q1174.08 1527.75 1174.08 1534.72 Q1174.08 1541.75 1169.97 1545.35 Q1165.9 1548.95 1157.91 1548.95 L1149.83 1548.95 L1149.83 1568.04 L1143.4 1568.04 L1143.4 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1201.96 1537.87 Q1200.97 1537.3 1199.8 1537.04 Q1198.65 1536.76 1197.25 1536.76 Q1192.29 1536.76 1189.61 1540 Q1186.97 1543.22 1186.97 1549.27 L1186.97 1568.04 L1181.08 1568.04 L1181.08 1532.4 L1186.97 1532.4 L1186.97 1537.93 Q1188.82 1534.69 1191.78 1533.13 Q1194.74 1531.54 1198.97 1531.54 Q1199.57 1531.54 1200.31 1531.63 Q1201.04 1531.7 1201.93 1531.85 L1201.96 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1220.49 1536.5 Q1215.77 1536.5 1213.04 1540.19 Q1210.3 1543.85 1210.3 1550.25 Q1210.3 1556.65 1213.01 1560.34 Q1215.74 1564 1220.49 1564 Q1225.16 1564 1227.9 1560.31 Q1230.64 1556.62 1230.64 1550.25 Q1230.64 1543.92 1227.9 1540.23 Q1225.16 1536.5 1220.49 1536.5 M1220.49 1531.54 Q1228.12 1531.54 1232.48 1536.5 Q1236.85 1541.47 1236.85 1550.25 Q1236.85 1559 1232.48 1564 Q1228.12 1568.97 1220.49 1568.97 Q1212.81 1568.97 1208.45 1564 Q1204.13 1559 1204.13 1550.25 Q1204.13 1541.47 1208.45 1536.5 Q1212.81 1531.54 1220.49 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1264.6 1518.52 L1264.6 1523.39 L1259 1523.39 Q1255.85 1523.39 1254.61 1524.66 Q1253.4 1525.93 1253.4 1529.24 L1253.4 1532.4 L1263.04 1532.4 L1263.04 1536.95 L1253.4 1536.95 L1253.4 1568.04 L1247.51 1568.04 L1247.51 1536.95 L1241.91 1536.95 L1241.91 1532.4 L1247.51 1532.4 L1247.51 1529.91 Q1247.51 1523.96 1250.28 1521.26 Q1253.05 1518.52 1259.06 1518.52 L1264.6 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1269.5 1532.4 L1275.36 1532.4 L1275.36 1568.04 L1269.5 1568.04 L1269.5 1532.4 M1269.5 1518.52 L1275.36 1518.52 L1275.36 1525.93 L1269.5 1525.93 L1269.5 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1293.4 1522.27 L1293.4 1532.4 L1305.47 1532.4 L1305.47 1536.95 L1293.4 1536.95 L1293.4 1556.3 Q1293.4 1560.66 1294.58 1561.9 Q1295.79 1563.14 1299.45 1563.14 L1305.47 1563.14 L1305.47 1568.04 L1299.45 1568.04 Q1292.67 1568.04 1290.09 1565.53 Q1287.52 1562.98 1287.52 1556.3 L1287.52 1536.95 L1283.22 1536.95 L1283.22 1532.4 L1287.52 1532.4 L1287.52 1522.27 L1293.4 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1333.35 1518.52 L1346.84 1518.52 L1346.84 1523.07 L1339.21 1523.07 L1339.21 1572.09 L1346.84 1572.09 L1346.84 1576.64 L1333.35 1576.64 L1333.35 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1375.2 1577.62 L1372.02 1577.62 L1371.99 1568.04 Q1368.65 1567.98 1365.31 1567.28 Q1361.96 1566.55 1358.59 1565.12 L1358.59 1559.39 Q1361.84 1561.42 1365.15 1562.47 Q1368.49 1563.49 1372.02 1563.52 L1372.02 1549.01 Q1364.99 1547.86 1361.77 1545.13 Q1358.59 1542.39 1358.59 1537.62 Q1358.59 1532.43 1362.06 1529.44 Q1365.53 1526.44 1372.02 1526 L1372.02 1518.52 L1375.2 1518.52 L1375.2 1525.9 Q1378.16 1526.03 1380.93 1526.54 Q1383.7 1527.02 1386.34 1527.88 L1386.34 1533.45 Q1383.7 1532.11 1380.9 1531.38 Q1378.13 1530.65 1375.2 1530.52 L1375.2 1544.11 Q1382.43 1545.22 1385.83 1548.09 Q1389.24 1550.95 1389.24 1555.92 Q1389.24 1561.3 1385.61 1564.42 Q1382.02 1567.5 1375.2 1567.98 L1375.2 1577.62 M1372.02 1543.54 L1372.02 1530.49 Q1368.33 1530.9 1366.39 1532.59 Q1364.45 1534.27 1364.45 1537.07 Q1364.45 1539.81 1366.23 1541.34 Q1368.04 1542.87 1372.02 1543.54 M1375.2 1549.65 L1375.2 1563.43 Q1379.25 1562.89 1381.28 1561.14 Q1383.35 1559.39 1383.35 1556.52 Q1383.35 1553.72 1381.38 1552.07 Q1379.44 1550.41 1375.2 1549.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M1414.48 1518.52 L1414.48 1576.64 L1400.98 1576.64 L1400.98 1572.09 L1408.59 1572.09 L1408.59 1523.07 L1400.98 1523.07 L1400.98 1518.52 L1414.48 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,1095.58 2352.76,1095.58 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,767.972 2352.76,767.972 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,440.369 2352.76,440.369 "/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,112.765 2352.76,112.765 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1423.18 205.121,47.2441 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1423.18 224.019,1423.18 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1095.58 224.019,1095.58 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,767.972 224.019,767.972 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,440.369 224.019,440.369 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,112.765 224.019,112.765 "/>
+<path clip-path="url(#clip310)" d="M157.177 1408.98 Q153.566 1408.98 151.737 1412.54 Q149.931 1416.08 149.931 1423.21 Q149.931 1430.32 151.737 1433.89 Q153.566 1437.43 157.177 1437.43 Q160.811 1437.43 162.616 1433.89 Q164.445 1430.32 164.445 1423.21 Q164.445 1416.08 162.616 1412.54 Q160.811 1408.98 157.177 1408.98 M157.177 1405.27 Q162.987 1405.27 166.042 1409.88 Q169.121 1414.46 169.121 1423.21 Q169.121 1431.94 166.042 1436.55 Q162.987 1441.13 157.177 1441.13 Q151.366 1441.13 148.288 1436.55 Q145.232 1431.94 145.232 1423.21 Q145.232 1414.46 148.288 1409.88 Q151.366 1405.27 157.177 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M117.825 1108.92 L125.464 1108.92 L125.464 1082.56 L117.154 1084.22 L117.154 1079.96 L125.418 1078.3 L130.093 1078.3 L130.093 1108.92 L137.732 1108.92 L137.732 1112.86 L117.825 1112.86 L117.825 1108.92 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M157.177 1081.37 Q153.566 1081.37 151.737 1084.94 Q149.931 1088.48 149.931 1095.61 Q149.931 1102.72 151.737 1106.28 Q153.566 1109.82 157.177 1109.82 Q160.811 1109.82 162.616 1106.28 Q164.445 1102.72 164.445 1095.61 Q164.445 1088.48 162.616 1084.94 Q160.811 1081.37 157.177 1081.37 M157.177 1077.67 Q162.987 1077.67 166.042 1082.28 Q169.121 1086.86 169.121 1095.61 Q169.121 1104.34 166.042 1108.94 Q162.987 1113.53 157.177 1113.53 Q151.366 1113.53 148.288 1108.94 Q145.232 1104.34 145.232 1095.61 Q145.232 1086.86 148.288 1082.28 Q151.366 1077.67 157.177 1077.67 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M121.043 781.317 L137.362 781.317 L137.362 785.252 L115.418 785.252 L115.418 781.317 Q118.08 778.563 122.663 773.933 Q127.269 769.28 128.45 767.938 Q130.695 765.414 131.575 763.678 Q132.478 761.919 132.478 760.229 Q132.478 757.475 130.533 755.739 Q128.612 754.003 125.51 754.003 Q123.311 754.003 120.857 754.766 Q118.427 755.53 115.649 757.081 L115.649 752.359 Q118.473 751.225 120.927 750.646 Q123.38 750.067 125.418 750.067 Q130.788 750.067 133.982 752.753 Q137.177 755.438 137.177 759.928 Q137.177 762.058 136.367 763.979 Q135.579 765.877 133.473 768.47 Q132.894 769.141 129.792 772.359 Q126.691 775.553 121.043 781.317 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M157.177 753.771 Q153.566 753.771 151.737 757.336 Q149.931 760.877 149.931 768.007 Q149.931 775.114 151.737 778.678 Q153.566 782.22 157.177 782.22 Q160.811 782.22 162.616 778.678 Q164.445 775.114 164.445 768.007 Q164.445 760.877 162.616 757.336 Q160.811 753.771 157.177 753.771 M157.177 750.067 Q162.987 750.067 166.042 754.674 Q169.121 759.257 169.121 768.007 Q169.121 776.734 166.042 781.34 Q162.987 785.924 157.177 785.924 Q151.366 785.924 148.288 781.34 Q145.232 776.734 145.232 768.007 Q145.232 759.257 148.288 754.674 Q151.366 750.067 157.177 750.067 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M131.181 439.014 Q134.538 439.732 136.413 442.001 Q138.311 444.269 138.311 447.602 Q138.311 452.718 134.792 455.519 Q131.274 458.32 124.793 458.32 Q122.617 458.32 120.302 457.88 Q118.01 457.463 115.556 456.607 L115.556 452.093 Q117.501 453.227 119.816 453.806 Q122.13 454.385 124.654 454.385 Q129.052 454.385 131.343 452.649 Q133.658 450.913 133.658 447.602 Q133.658 444.547 131.505 442.834 Q129.376 441.098 125.556 441.098 L121.529 441.098 L121.529 437.255 L125.742 437.255 Q129.191 437.255 131.019 435.889 Q132.848 434.501 132.848 431.908 Q132.848 429.246 130.95 427.834 Q129.075 426.399 125.556 426.399 Q123.635 426.399 121.436 426.815 Q119.237 427.232 116.598 428.112 L116.598 423.945 Q119.26 423.204 121.575 422.834 Q123.913 422.464 125.973 422.464 Q131.297 422.464 134.399 424.894 Q137.501 427.302 137.501 431.422 Q137.501 434.292 135.857 436.283 Q134.214 438.251 131.181 439.014 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M157.177 426.167 Q153.566 426.167 151.737 429.732 Q149.931 433.274 149.931 440.403 Q149.931 447.51 151.737 451.075 Q153.566 454.616 157.177 454.616 Q160.811 454.616 162.616 451.075 Q164.445 447.51 164.445 440.403 Q164.445 433.274 162.616 429.732 Q160.811 426.167 157.177 426.167 M157.177 422.464 Q162.987 422.464 166.042 427.07 Q169.121 431.653 169.121 440.403 Q169.121 449.13 166.042 453.737 Q162.987 458.32 157.177 458.32 Q151.366 458.32 148.288 453.737 Q145.232 449.13 145.232 440.403 Q145.232 431.653 148.288 427.07 Q151.366 422.464 157.177 422.464 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M129.862 99.5589 L118.056 118.008 L129.862 118.008 L129.862 99.5589 M128.635 95.4848 L134.515 95.4848 L134.515 118.008 L139.445 118.008 L139.445 121.897 L134.515 121.897 L134.515 130.045 L129.862 130.045 L129.862 121.897 L114.26 121.897 L114.26 117.383 L128.635 95.4848 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M157.177 98.5635 Q153.566 98.5635 151.737 102.128 Q149.931 105.67 149.931 112.8 Q149.931 119.906 151.737 123.471 Q153.566 127.012 157.177 127.012 Q160.811 127.012 162.616 123.471 Q164.445 119.906 164.445 112.8 Q164.445 105.67 162.616 102.128 Q160.811 98.5635 157.177 98.5635 M157.177 94.8598 Q162.987 94.8598 166.042 99.4663 Q169.121 104.05 169.121 112.8 Q169.121 121.526 166.042 126.133 Q162.987 130.716 157.177 130.716 Q151.366 130.716 148.288 126.133 Q145.232 121.526 145.232 112.8 Q145.232 104.05 148.288 99.4663 Q151.366 94.8598 157.177 94.8598 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M16.4842 1073.5 L16.4842 1064.84 L56.238 1043.77 L16.4842 1043.77 L16.4842 1037.54 L64.0042 1037.54 L64.0042 1046.19 L24.2503 1067.26 L64.0042 1067.26 L64.0042 1073.5 L16.4842 1073.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M49.9359 1025.6 L28.3562 1025.6 L28.3562 1019.74 L49.7131 1019.74 Q54.7739 1019.74 57.3202 1017.77 Q59.8346 1015.8 59.8346 1011.85 Q59.8346 1007.11 56.8109 1004.37 Q53.7872 1001.6 48.5673 1001.6 L28.3562 1001.6 L28.3562 995.744 L64.0042 995.744 L64.0042 1001.6 L58.5296 1001.6 Q61.7762 1003.73 63.3676 1006.57 Q64.9272 1009.37 64.9272 1013.09 Q64.9272 1019.23 61.1078 1022.42 Q57.2883 1025.6 49.9359 1025.6 M27.4968 1010.86 L27.4968 1010.86 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M35.1993 955.927 Q31.2526 953.731 29.3747 950.675 Q27.4968 947.62 27.4968 943.482 Q27.4968 937.912 31.4117 934.888 Q35.2948 931.865 42.4881 931.865 L64.0042 931.865 L64.0042 937.753 L42.679 937.753 Q37.5546 937.753 35.072 939.567 Q32.5894 941.381 32.5894 945.105 Q32.5894 949.657 35.6131 952.298 Q38.6368 954.94 43.8567 954.94 L64.0042 954.94 L64.0042 960.829 L42.679 960.829 Q37.5228 960.829 35.072 962.643 Q32.5894 964.457 32.5894 968.245 Q32.5894 972.732 35.6449 975.374 Q38.6686 978.016 43.8567 978.016 L64.0042 978.016 L64.0042 983.904 L28.3562 983.904 L28.3562 978.016 L33.8944 978.016 Q30.616 976.011 29.0564 973.21 Q27.4968 970.409 27.4968 966.558 Q27.4968 962.675 29.4702 959.969 Q31.4436 957.232 35.1993 955.927 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M46.212 894.593 Q39.7508 894.593 36.0905 897.267 Q32.3984 899.909 32.3984 904.556 Q32.3984 909.203 36.0905 911.876 Q39.7508 914.518 46.212 914.518 Q52.6732 914.518 56.3653 911.876 Q60.0256 909.203 60.0256 904.556 Q60.0256 899.909 56.3653 897.267 Q52.6732 894.593 46.212 894.593 M33.7671 914.518 Q30.5842 912.672 29.0564 909.871 Q27.4968 907.038 27.4968 903.123 Q27.4968 896.63 32.6531 892.588 Q37.8093 888.514 46.212 888.514 Q54.6147 888.514 59.771 892.588 Q64.9272 896.63 64.9272 903.123 Q64.9272 907.038 63.3994 909.871 Q61.8398 912.672 58.657 914.518 L64.0042 914.518 L64.0042 920.406 L14.479 920.406 L14.479 914.518 L33.7671 914.518 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M44.7161 848.315 L47.5806 848.315 L47.5806 875.242 Q53.6281 874.86 56.8109 871.613 Q59.9619 868.335 59.9619 862.51 Q59.9619 859.136 59.1344 855.985 Q58.3069 852.802 56.6518 849.683 L62.1899 849.683 Q63.5267 852.834 64.227 856.144 Q64.9272 859.455 64.9272 862.86 Q64.9272 871.39 59.9619 876.387 Q54.9967 881.353 46.5303 881.353 Q37.7774 881.353 32.6531 876.642 Q27.4968 871.9 27.4968 863.879 Q27.4968 856.686 32.1438 852.516 Q36.7589 848.315 44.7161 848.315 M42.9973 854.171 Q38.1912 854.235 35.3266 856.877 Q32.4621 859.486 32.4621 863.815 Q32.4621 868.717 35.2312 871.677 Q38.0002 874.605 43.0292 875.051 L42.9973 854.171 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M33.8307 818.046 Q33.2578 819.032 33.0032 820.21 Q32.7167 821.356 32.7167 822.756 Q32.7167 827.722 35.9632 830.395 Q39.1779 833.037 45.2253 833.037 L64.0042 833.037 L64.0042 838.925 L28.3562 838.925 L28.3562 833.037 L33.8944 833.037 Q30.6479 831.191 29.0883 828.231 Q27.4968 825.271 27.4968 821.038 Q27.4968 820.433 27.5923 819.701 Q27.656 818.969 27.8151 818.078 L33.8307 818.046 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M32.4621 777.369 Q32.4621 782.079 36.1542 784.817 Q39.8145 787.554 46.212 787.554 Q52.6095 787.554 56.3017 784.849 Q59.9619 782.111 59.9619 777.369 Q59.9619 772.69 56.2698 769.953 Q52.5777 767.216 46.212 767.216 Q39.8781 767.216 36.186 769.953 Q32.4621 772.69 32.4621 777.369 M27.4968 777.369 Q27.4968 769.73 32.4621 765.37 Q37.4273 761.009 46.212 761.009 Q54.9649 761.009 59.9619 765.37 Q64.9272 769.73 64.9272 777.369 Q64.9272 785.04 59.9619 789.4 Q54.9649 793.729 46.212 793.729 Q37.4273 793.729 32.4621 789.4 Q27.4968 785.04 27.4968 777.369 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M14.479 733.255 L19.3487 733.255 L19.3487 738.856 Q19.3487 742.007 20.6219 743.249 Q21.895 744.458 25.2052 744.458 L28.3562 744.458 L28.3562 734.814 L32.9077 734.814 L32.9077 744.458 L64.0042 744.458 L64.0042 750.346 L32.9077 750.346 L32.9077 755.948 L28.3562 755.948 L28.3562 750.346 L25.8736 750.346 Q19.9216 750.346 17.2162 747.577 Q14.479 744.808 14.479 738.793 L14.479 733.255 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M32.4621 693.819 Q32.4621 698.53 36.1542 701.267 Q39.8145 704.004 46.212 704.004 Q52.6095 704.004 56.3017 701.299 Q59.9619 698.561 59.9619 693.819 Q59.9619 689.14 56.2698 686.403 Q52.5777 683.666 46.212 683.666 Q39.8781 683.666 36.186 686.403 Q32.4621 689.14 32.4621 693.819 M27.4968 693.819 Q27.4968 686.18 32.4621 681.82 Q37.4273 677.459 46.212 677.459 Q54.9649 677.459 59.9619 681.82 Q64.9272 686.18 64.9272 693.819 Q64.9272 701.49 59.9619 705.85 Q54.9649 710.179 46.212 710.179 Q37.4273 710.179 32.4621 705.85 Q27.4968 701.49 27.4968 693.819 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M49.9359 668.356 L28.3562 668.356 L28.3562 662.5 L49.7131 662.5 Q54.7739 662.5 57.3202 660.526 Q59.8346 658.553 59.8346 654.606 Q59.8346 649.864 56.8109 647.127 Q53.7872 644.357 48.5673 644.357 L28.3562 644.357 L28.3562 638.501 L64.0042 638.501 L64.0042 644.357 L58.5296 644.357 Q61.7762 646.49 63.3676 649.323 Q64.9272 652.124 64.9272 655.848 Q64.9272 661.99 61.1078 665.173 Q57.2883 668.356 49.9359 668.356 M27.4968 653.62 L27.4968 653.62 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M18.2347 620.645 L28.3562 620.645 L28.3562 608.582 L32.9077 608.582 L32.9077 620.645 L52.2594 620.645 Q56.6199 620.645 57.8613 619.468 Q59.1026 618.258 59.1026 614.598 L59.1026 608.582 L64.0042 608.582 L64.0042 614.598 Q64.0042 621.377 61.4897 623.955 Q58.9434 626.533 52.2594 626.533 L32.9077 626.533 L32.9077 630.83 L28.3562 630.83 L28.3562 626.533 L18.2347 626.533 L18.2347 620.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M29.7248 575.226 L35.1993 575.226 Q33.8307 577.708 33.1623 580.223 Q32.4621 582.706 32.4621 585.252 Q32.4621 590.949 36.0905 594.1 Q39.6872 597.251 46.212 597.251 Q52.7369 597.251 56.3653 594.1 Q59.9619 590.949 59.9619 585.252 Q59.9619 582.706 59.2935 580.223 Q58.5933 577.708 57.2247 575.226 L62.6355 575.226 Q63.7814 577.677 64.3543 580.318 Q64.9272 582.928 64.9272 585.888 Q64.9272 593.941 59.8664 598.683 Q54.8057 603.426 46.212 603.426 Q37.491 603.426 32.4939 598.652 Q27.4968 593.846 27.4968 585.506 Q27.4968 582.801 28.0697 580.223 Q28.6108 577.645 29.7248 575.226 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M32.4621 551.227 Q32.4621 555.938 36.1542 558.675 Q39.8145 561.412 46.212 561.412 Q52.6095 561.412 56.3017 558.707 Q59.9619 555.97 59.9619 551.227 Q59.9619 546.548 56.2698 543.811 Q52.5777 541.074 46.212 541.074 Q39.8781 541.074 36.186 543.811 Q32.4621 546.548 32.4621 551.227 M27.4968 551.227 Q27.4968 543.588 32.4621 539.228 Q37.4273 534.867 46.212 534.867 Q54.9649 534.867 59.9619 539.228 Q64.9272 543.588 64.9272 551.227 Q64.9272 558.898 59.9619 563.258 Q54.9649 567.587 46.212 567.587 Q37.4273 567.587 32.4621 563.258 Q27.4968 558.898 27.4968 551.227 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M35.1993 497.405 Q31.2526 495.209 29.3747 492.153 Q27.4968 489.098 27.4968 484.96 Q27.4968 479.39 31.4117 476.366 Q35.2948 473.343 42.4881 473.343 L64.0042 473.343 L64.0042 479.231 L42.679 479.231 Q37.5546 479.231 35.072 481.045 Q32.5894 482.859 32.5894 486.583 Q32.5894 491.135 35.6131 493.777 Q38.6368 496.418 43.8567 496.418 L64.0042 496.418 L64.0042 502.307 L42.679 502.307 Q37.5228 502.307 35.072 504.121 Q32.5894 505.935 32.5894 509.723 Q32.5894 514.211 35.6449 516.852 Q38.6686 519.494 43.8567 519.494 L64.0042 519.494 L64.0042 525.382 L28.3562 525.382 L28.3562 519.494 L33.8944 519.494 Q30.616 517.489 29.0564 514.688 Q27.4968 511.887 27.4968 508.036 Q27.4968 504.153 29.4702 501.447 Q31.4436 498.71 35.1993 497.405 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M44.7161 431.17 L47.5806 431.17 L47.5806 458.097 Q53.6281 457.715 56.8109 454.468 Q59.9619 451.19 59.9619 445.365 Q59.9619 441.992 59.1344 438.841 Q58.3069 435.658 56.6518 432.539 L62.1899 432.539 Q63.5267 435.69 64.227 439 Q64.9272 442.31 64.9272 445.716 Q64.9272 454.246 59.9619 459.243 Q54.9967 464.208 46.5303 464.208 Q37.7774 464.208 32.6531 459.497 Q27.4968 454.755 27.4968 446.734 Q27.4968 439.541 32.1438 435.371 Q36.7589 431.17 44.7161 431.17 M42.9973 437.026 Q38.1912 437.09 35.3266 439.732 Q32.4621 442.342 32.4621 446.67 Q32.4621 451.572 35.2312 454.532 Q38.0002 457.46 43.0292 457.906 L42.9973 437.026 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M29.4065 398.832 L34.9447 398.832 Q33.6716 401.315 33.035 403.988 Q32.3984 406.662 32.3984 409.527 Q32.3984 413.887 33.7352 416.083 Q35.072 418.248 37.7456 418.248 Q39.7826 418.248 40.9603 416.688 Q42.1061 415.128 43.1565 410.418 L43.6021 408.413 Q44.9389 402.174 47.3897 399.564 Q49.8086 396.922 54.1691 396.922 Q59.1344 396.922 62.0308 400.869 Q64.9272 404.784 64.9272 411.659 Q64.9272 414.524 64.3543 417.643 Q63.8132 420.73 62.6992 424.168 L56.6518 424.168 Q58.3387 420.921 59.198 417.77 Q60.0256 414.619 60.0256 411.532 Q60.0256 407.394 58.6251 405.166 Q57.1929 402.938 54.6147 402.938 Q52.2276 402.938 50.9545 404.561 Q49.6813 406.153 48.5037 411.595 L48.0262 413.632 Q46.8804 419.075 44.5251 421.494 Q42.138 423.913 38.0002 423.913 Q32.9713 423.913 30.2341 420.348 Q27.4968 416.783 27.4968 410.227 Q27.4968 406.98 27.9743 404.116 Q28.4517 401.251 29.4065 398.832 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip312)" d="M323.245 1357.66 L323.245 1423.18 L514.383 1423.18 L514.383 1357.66 L323.245 1357.66 L323.245 1357.66  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="323.245,1357.66 323.245,1423.18 514.383,1423.18 514.383,1357.66 323.245,1357.66 "/>
+<path clip-path="url(#clip312)" d="M514.383 1324.9 L514.383 1423.18 L705.522 1423.18 L705.522 1324.9 L514.383 1324.9 L514.383 1324.9  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="514.383,1324.9 514.383,1423.18 705.522,1423.18 705.522,1324.9 514.383,1324.9 "/>
+<path clip-path="url(#clip312)" d="M705.522 1193.86 L705.522 1423.18 L896.661 1423.18 L896.661 1193.86 L705.522 1193.86 L705.522 1193.86  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="705.522,1193.86 705.522,1423.18 896.661,1423.18 896.661,1193.86 705.522,1193.86 "/>
+<path clip-path="url(#clip312)" d="M896.661 1193.86 L896.661 1423.18 L1087.8 1423.18 L1087.8 1193.86 L896.661 1193.86 L896.661 1193.86  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="896.661,1193.86 896.661,1423.18 1087.8,1423.18 1087.8,1193.86 896.661,1193.86 "/>
+<path clip-path="url(#clip312)" d="M1087.8 1292.14 L1087.8 1423.18 L1278.94 1423.18 L1278.94 1292.14 L1087.8 1292.14 L1087.8 1292.14  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1087.8,1292.14 1087.8,1423.18 1278.94,1423.18 1278.94,1292.14 1087.8,1292.14 "/>
+<path clip-path="url(#clip312)" d="M1278.94 1357.66 L1278.94 1423.18 L1470.08 1423.18 L1470.08 1357.66 L1278.94 1357.66 L1278.94 1357.66  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1278.94,1357.66 1278.94,1423.18 1470.08,1423.18 1470.08,1357.66 1278.94,1357.66 "/>
+<path clip-path="url(#clip312)" d="M1470.08 1259.38 L1470.08 1423.18 L1661.22 1423.18 L1661.22 1259.38 L1470.08 1259.38 L1470.08 1259.38  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1470.08,1259.38 1470.08,1423.18 1661.22,1423.18 1661.22,1259.38 1470.08,1259.38 "/>
+<path clip-path="url(#clip312)" d="M1661.22 1062.82 L1661.22 1423.18 L1852.35 1423.18 L1852.35 1062.82 L1661.22 1062.82 L1661.22 1062.82  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1661.22,1062.82 1661.22,1423.18 1852.35,1423.18 1852.35,1062.82 1661.22,1062.82 "/>
+<path clip-path="url(#clip312)" d="M1852.35 866.253 L1852.35 1423.18 L2043.49 1423.18 L2043.49 866.253 L1852.35 866.253 L1852.35 866.253  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1852.35,866.253 1852.35,1423.18 2043.49,1423.18 2043.49,866.253 1852.35,866.253 "/>
+<path clip-path="url(#clip312)" d="M2043.49 47.2441 L2043.49 1423.18 L2234.63 1423.18 L2234.63 47.2441 L2043.49 47.2441 L2043.49 47.2441  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip312)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2043.49,47.2441 2043.49,1423.18 2234.63,1423.18 2234.63,47.2441 2043.49,47.2441 "/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="418.814" cy="1357.66" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="609.953" cy="1324.9" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="801.092" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="992.23" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1183.37" cy="1292.14" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1374.51" cy="1357.66" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1565.65" cy="1259.38" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1756.79" cy="1062.82" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1947.92" cy="866.253" r="2"/>
+<circle clip-path="url(#clip312)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2139.06" cy="47.2441" r="2"/>
+<polyline clip-path="url(#clip312)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1701.21,2799.12 1701.21,-1328.69 "/>
+<path clip-path="url(#clip310)" d="M276.709 196.789 L1017.7 196.789 L1017.7 93.1086 L276.709 93.1086  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip310)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="276.709,196.789 1017.7,196.789 1017.7,93.1086 276.709,93.1086 276.709,196.789 "/>
+<polyline clip-path="url(#clip310)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="300.571,144.949 443.747,144.949 "/>
+<path clip-path="url(#clip310)" d="M467.61 127.669 L489.461 127.669 L489.461 131.604 L472.286 131.604 L472.286 141.835 L488.744 141.835 L488.744 145.77 L472.286 145.77 L472.286 158.293 L489.878 158.293 L489.878 162.229 L467.61 162.229 L467.61 127.669 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M518.929 136.303 L509.554 148.919 L519.415 162.229 L514.392 162.229 L506.846 152.043 L499.299 162.229 L494.276 162.229 L504.346 148.664 L495.133 136.303 L500.156 136.303 L507.031 145.539 L513.906 136.303 L518.929 136.303 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M529.554 158.34 L529.554 172.09 L525.271 172.09 L525.271 136.303 L529.554 136.303 L529.554 140.238 Q530.896 137.923 532.933 136.812 Q534.994 135.678 537.841 135.678 Q542.563 135.678 545.503 139.428 Q548.466 143.178 548.466 149.289 Q548.466 155.4 545.503 159.15 Q542.563 162.9 537.841 162.9 Q534.994 162.9 532.933 161.789 Q530.896 160.655 529.554 158.34 M544.044 149.289 Q544.044 144.59 542.1 141.928 Q540.179 139.243 536.799 139.243 Q533.42 139.243 531.475 141.928 Q529.554 144.59 529.554 149.289 Q529.554 153.988 531.475 156.673 Q533.42 159.335 536.799 159.335 Q540.179 159.335 542.1 156.673 Q544.044 153.988 544.044 149.289 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M577.702 148.201 L577.702 150.284 L558.118 150.284 Q558.396 154.682 560.757 156.997 Q563.142 159.289 567.378 159.289 Q569.831 159.289 572.123 158.687 Q574.438 158.085 576.706 156.881 L576.706 160.909 Q574.415 161.881 572.007 162.391 Q569.6 162.9 567.123 162.9 Q560.919 162.9 557.285 159.289 Q553.674 155.678 553.674 149.52 Q553.674 143.155 557.1 139.428 Q560.549 135.678 566.382 135.678 Q571.614 135.678 574.646 139.057 Q577.702 142.414 577.702 148.201 M573.442 146.951 Q573.396 143.456 571.475 141.372 Q569.577 139.289 566.429 139.289 Q562.864 139.289 560.711 141.303 Q558.581 143.317 558.257 146.974 L573.442 146.951 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M603.35 137.298 L603.35 141.28 Q601.544 140.284 599.715 139.798 Q597.91 139.289 596.058 139.289 Q591.915 139.289 589.623 141.928 Q587.331 144.544 587.331 149.289 Q587.331 154.034 589.623 156.673 Q591.915 159.289 596.058 159.289 Q597.91 159.289 599.715 158.803 Q601.544 158.293 603.35 157.298 L603.35 161.233 Q601.567 162.067 599.646 162.483 Q597.748 162.9 595.595 162.9 Q589.739 162.9 586.29 159.219 Q582.841 155.539 582.841 149.289 Q582.841 142.946 586.313 139.312 Q589.808 135.678 595.873 135.678 Q597.84 135.678 599.715 136.095 Q601.59 136.488 603.35 137.298 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M614.97 128.942 L614.97 136.303 L623.743 136.303 L623.743 139.613 L614.97 139.613 L614.97 153.687 Q614.97 156.858 615.826 157.761 Q616.706 158.664 619.368 158.664 L623.743 158.664 L623.743 162.229 L619.368 162.229 Q614.438 162.229 612.563 160.4 Q610.688 158.548 610.688 153.687 L610.688 139.613 L607.563 139.613 L607.563 136.303 L610.688 136.303 L610.688 128.942 L614.97 128.942 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M651.521 148.201 L651.521 150.284 L631.938 150.284 Q632.215 154.682 634.576 156.997 Q636.961 159.289 641.197 159.289 Q643.65 159.289 645.942 158.687 Q648.257 158.085 650.525 156.881 L650.525 160.909 Q648.234 161.881 645.826 162.391 Q643.419 162.9 640.942 162.9 Q634.738 162.9 631.104 159.289 Q627.493 155.678 627.493 149.52 Q627.493 143.155 630.919 139.428 Q634.368 135.678 640.201 135.678 Q645.433 135.678 648.465 139.057 Q651.521 142.414 651.521 148.201 M647.262 146.951 Q647.215 143.456 645.294 141.372 Q643.396 139.289 640.248 139.289 Q636.683 139.289 634.53 141.303 Q632.4 143.317 632.076 146.974 L647.262 146.951 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M675.572 140.238 L675.572 126.21 L679.831 126.21 L679.831 162.229 L675.572 162.229 L675.572 158.34 Q674.229 160.655 672.169 161.789 Q670.132 162.9 667.261 162.9 Q662.562 162.9 659.599 159.15 Q656.66 155.4 656.66 149.289 Q656.66 143.178 659.599 139.428 Q662.562 135.678 667.261 135.678 Q670.132 135.678 672.169 136.812 Q674.229 137.923 675.572 140.238 M661.058 149.289 Q661.058 153.988 662.979 156.673 Q664.923 159.335 668.303 159.335 Q671.683 159.335 673.627 156.673 Q675.572 153.988 675.572 149.289 Q675.572 144.59 673.627 141.928 Q671.683 139.243 668.303 139.243 Q664.923 139.243 662.979 141.928 Q661.058 144.59 661.058 149.289 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M707.794 158.34 L707.794 172.09 L703.511 172.09 L703.511 136.303 L707.794 136.303 L707.794 140.238 Q709.136 137.923 711.173 136.812 Q713.233 135.678 716.081 135.678 Q720.803 135.678 723.743 139.428 Q726.706 143.178 726.706 149.289 Q726.706 155.4 723.743 159.15 Q720.803 162.9 716.081 162.9 Q713.233 162.9 711.173 161.789 Q709.136 160.655 707.794 158.34 M722.284 149.289 Q722.284 144.59 720.34 141.928 Q718.419 139.243 715.039 139.243 Q711.659 139.243 709.715 141.928 Q707.794 144.59 707.794 149.289 Q707.794 153.988 709.715 156.673 Q711.659 159.335 715.039 159.335 Q718.419 159.335 720.34 156.673 Q722.284 153.988 722.284 149.289 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M748.789 140.284 Q748.071 139.868 747.215 139.682 Q746.381 139.474 745.363 139.474 Q741.752 139.474 739.807 141.835 Q737.886 144.173 737.886 148.571 L737.886 162.229 L733.604 162.229 L733.604 136.303 L737.886 136.303 L737.886 140.331 Q739.229 137.969 741.381 136.835 Q743.534 135.678 746.613 135.678 Q747.053 135.678 747.585 135.747 Q748.117 135.794 748.766 135.909 L748.789 140.284 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M762.261 139.289 Q758.835 139.289 756.844 141.974 Q754.854 144.636 754.854 149.289 Q754.854 153.942 756.821 156.627 Q758.812 159.289 762.261 159.289 Q765.664 159.289 767.654 156.604 Q769.645 153.918 769.645 149.289 Q769.645 144.682 767.654 141.997 Q765.664 139.289 762.261 139.289 M762.261 135.678 Q767.816 135.678 770.988 139.289 Q774.159 142.9 774.159 149.289 Q774.159 155.655 770.988 159.289 Q767.816 162.9 762.261 162.9 Q756.682 162.9 753.511 159.289 Q750.363 155.655 750.363 149.289 Q750.363 142.9 753.511 139.289 Q756.682 135.678 762.261 135.678 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M794.344 126.21 L794.344 129.752 L790.27 129.752 Q787.978 129.752 787.076 130.678 Q786.196 131.604 786.196 134.011 L786.196 136.303 L793.21 136.303 L793.21 139.613 L786.196 139.613 L786.196 162.229 L781.914 162.229 L781.914 139.613 L777.84 139.613 L777.84 136.303 L781.914 136.303 L781.914 134.497 Q781.914 130.169 783.927 128.201 Q785.941 126.21 790.316 126.21 L794.344 126.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M797.909 136.303 L802.168 136.303 L802.168 162.229 L797.909 162.229 L797.909 136.303 M797.909 126.21 L802.168 126.21 L802.168 131.604 L797.909 131.604 L797.909 126.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M815.293 128.942 L815.293 136.303 L824.066 136.303 L824.066 139.613 L815.293 139.613 L815.293 153.687 Q815.293 156.858 816.149 157.761 Q817.029 158.664 819.691 158.664 L824.066 158.664 L824.066 162.229 L819.691 162.229 Q814.761 162.229 812.886 160.4 Q811.011 158.548 811.011 153.687 L811.011 139.613 L807.886 139.613 L807.886 136.303 L811.011 136.303 L811.011 128.942 L815.293 128.942 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M854.969 126.257 Q851.867 131.581 850.362 136.789 Q848.858 141.997 848.858 147.344 Q848.858 152.692 850.362 157.946 Q851.89 163.178 854.969 168.479 L851.265 168.479 Q847.793 163.039 846.057 157.784 Q844.344 152.53 844.344 147.344 Q844.344 142.182 846.057 136.951 Q847.77 131.72 851.265 126.257 L854.969 126.257 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M874.783 169.196 L872.469 169.196 L872.445 162.229 Q870.015 162.182 867.584 161.673 Q865.154 161.141 862.7 160.099 L862.7 155.932 Q865.061 157.414 867.469 158.178 Q869.899 158.918 872.469 158.942 L872.469 148.386 Q867.353 147.553 865.015 145.562 Q862.7 143.571 862.7 140.099 Q862.7 136.326 865.223 134.15 Q867.746 131.974 872.469 131.65 L872.469 126.21 L874.783 126.21 L874.783 131.581 Q876.936 131.673 878.95 132.044 Q880.964 132.391 882.885 133.016 L882.885 137.067 Q880.964 136.095 878.927 135.562 Q876.913 135.03 874.783 134.937 L874.783 144.821 Q880.038 145.631 882.515 147.715 Q884.992 149.798 884.992 153.409 Q884.992 157.321 882.353 159.59 Q879.737 161.835 874.783 162.182 L874.783 169.196 M872.469 144.405 L872.469 134.914 Q869.783 135.215 868.371 136.442 Q866.959 137.669 866.959 139.706 Q866.959 141.696 868.256 142.807 Q869.575 143.919 872.469 144.405 M874.783 148.849 L874.783 158.872 Q877.723 158.479 879.205 157.205 Q880.709 155.932 880.709 153.849 Q880.709 151.812 879.274 150.608 Q877.862 149.405 874.783 148.849 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M894.043 127.669 L912.399 127.669 L912.399 131.604 L898.325 131.604 L898.325 140.076 Q899.343 139.729 900.362 139.567 Q901.381 139.382 902.399 139.382 Q908.186 139.382 911.566 142.553 Q914.945 145.724 914.945 151.141 Q914.945 156.719 911.473 159.821 Q908.001 162.9 901.681 162.9 Q899.506 162.9 897.237 162.53 Q894.992 162.159 892.584 161.418 L892.584 156.719 Q894.668 157.854 896.89 158.409 Q899.112 158.965 901.589 158.965 Q905.593 158.965 907.931 156.858 Q910.269 154.752 910.269 151.141 Q910.269 147.53 907.931 145.423 Q905.593 143.317 901.589 143.317 Q899.714 143.317 897.839 143.733 Q895.987 144.15 894.043 145.03 L894.043 127.669 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M924.204 127.669 L942.561 127.669 L942.561 131.604 L928.487 131.604 L928.487 140.076 Q929.505 139.729 930.524 139.567 Q931.542 139.382 932.561 139.382 Q938.348 139.382 941.728 142.553 Q945.107 145.724 945.107 151.141 Q945.107 156.719 941.635 159.821 Q938.163 162.9 931.843 162.9 Q929.667 162.9 927.399 162.53 Q925.154 162.159 922.746 161.418 L922.746 156.719 Q924.829 157.854 927.052 158.409 Q929.274 158.965 931.751 158.965 Q935.755 158.965 938.093 156.858 Q940.431 154.752 940.431 151.141 Q940.431 147.53 938.093 145.423 Q935.755 143.317 931.751 143.317 Q929.876 143.317 928.001 143.733 Q926.149 144.15 924.204 145.03 L924.204 127.669 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M954.366 127.669 L972.723 127.669 L972.723 131.604 L958.649 131.604 L958.649 140.076 Q959.667 139.729 960.686 139.567 Q961.704 139.382 962.723 139.382 Q968.51 139.382 971.889 142.553 Q975.269 145.724 975.269 151.141 Q975.269 156.719 971.797 159.821 Q968.325 162.9 962.005 162.9 Q959.829 162.9 957.561 162.53 Q955.315 162.159 952.908 161.418 L952.908 156.719 Q954.991 157.854 957.214 158.409 Q959.436 158.965 961.913 158.965 Q965.917 158.965 968.255 156.858 Q970.593 154.752 970.593 151.141 Q970.593 147.53 968.255 145.423 Q965.917 143.317 961.913 143.317 Q960.038 143.317 958.163 143.733 Q956.311 144.15 954.366 145.03 L954.366 127.669 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip310)" d="M983.209 126.257 L986.912 126.257 Q990.385 131.72 992.098 136.951 Q993.834 142.182 993.834 147.344 Q993.834 152.53 992.098 157.784 Q990.385 163.039 986.912 168.479 L983.209 168.479 Q986.287 163.178 987.792 157.946 Q989.32 152.692 989.32 147.344 Q989.32 141.997 987.792 136.789 Q986.287 131.581 983.209 126.257 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/applications/two_stage_stochastic/ddf1a4d6.svg b/previews/PR3525/tutorials/applications/two_stage_stochastic/ddf1a4d6.svg
new file mode 100644
index 00000000000..5988f418a88
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/two_stage_stochastic/ddf1a4d6.svg
@@ -0,0 +1,80 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip370">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip370)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip371">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip370)" d="M205.121 1423.18 L2352.76 1423.18 L2352.76 47.2441 L205.121 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip372">
+    <rect x="205" y="47" width="2149" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="514.383,1423.18 514.383,47.2441 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="896.661,1423.18 896.661,47.2441 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1278.94,1423.18 1278.94,47.2441 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1661.22,1423.18 1661.22,47.2441 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2043.49,1423.18 2043.49,47.2441 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="514.383,1423.18 514.383,1404.28 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="896.661,1423.18 896.661,1404.28 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1278.94,1423.18 1278.94,1404.28 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1661.22,1423.18 1661.22,1404.28 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2043.49,1423.18 2043.49,1404.28 "/>
+<path clip-path="url(#clip370)" d="M487.474 1455.09 L475.668 1473.54 L487.474 1473.54 L487.474 1455.09 M486.247 1451.02 L492.127 1451.02 L492.127 1473.54 L497.057 1473.54 L497.057 1477.43 L492.127 1477.43 L492.127 1485.58 L487.474 1485.58 L487.474 1477.43 L471.872 1477.43 L471.872 1472.92 L486.247 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M514.789 1454.1 Q511.177 1454.1 509.349 1457.66 Q507.543 1461.2 507.543 1468.33 Q507.543 1475.44 509.349 1479.01 Q511.177 1482.55 514.789 1482.55 Q518.423 1482.55 520.228 1479.01 Q522.057 1475.44 522.057 1468.33 Q522.057 1461.2 520.228 1457.66 Q518.423 1454.1 514.789 1454.1 M514.789 1450.39 Q520.599 1450.39 523.654 1455 Q526.733 1459.58 526.733 1468.33 Q526.733 1477.06 523.654 1481.67 Q520.599 1486.25 514.789 1486.25 Q508.978 1486.25 505.9 1481.67 Q502.844 1477.06 502.844 1468.33 Q502.844 1459.58 505.9 1455 Q508.978 1450.39 514.789 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M544.95 1454.1 Q541.339 1454.1 539.511 1457.66 Q537.705 1461.2 537.705 1468.33 Q537.705 1475.44 539.511 1479.01 Q541.339 1482.55 544.95 1482.55 Q548.585 1482.55 550.39 1479.01 Q552.219 1475.44 552.219 1468.33 Q552.219 1461.2 550.39 1457.66 Q548.585 1454.1 544.95 1454.1 M544.95 1450.39 Q550.761 1450.39 553.816 1455 Q556.895 1459.58 556.895 1468.33 Q556.895 1477.06 553.816 1481.67 Q550.761 1486.25 544.95 1486.25 Q539.14 1486.25 536.062 1481.67 Q533.006 1477.06 533.006 1468.33 Q533.006 1459.58 536.062 1455 Q539.14 1450.39 544.95 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M869.751 1455.09 L857.946 1473.54 L869.751 1473.54 L869.751 1455.09 M868.525 1451.02 L874.404 1451.02 L874.404 1473.54 L879.335 1473.54 L879.335 1477.43 L874.404 1477.43 L874.404 1485.58 L869.751 1485.58 L869.751 1477.43 L854.15 1477.43 L854.15 1472.92 L868.525 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M887.112 1451.02 L905.469 1451.02 L905.469 1454.96 L891.395 1454.96 L891.395 1463.43 Q892.413 1463.08 893.432 1462.92 Q894.45 1462.73 895.469 1462.73 Q901.256 1462.73 904.635 1465.9 Q908.015 1469.08 908.015 1474.49 Q908.015 1480.07 904.543 1483.17 Q901.071 1486.25 894.751 1486.25 Q892.575 1486.25 890.307 1485.88 Q888.061 1485.51 885.654 1484.77 L885.654 1480.07 Q887.737 1481.2 889.96 1481.76 Q892.182 1482.32 894.659 1482.32 Q898.663 1482.32 901.001 1480.21 Q903.339 1478.1 903.339 1474.49 Q903.339 1470.88 901.001 1468.77 Q898.663 1466.67 894.659 1466.67 Q892.784 1466.67 890.909 1467.08 Q889.057 1467.5 887.112 1468.38 L887.112 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M927.228 1454.1 Q923.617 1454.1 921.788 1457.66 Q919.983 1461.2 919.983 1468.33 Q919.983 1475.44 921.788 1479.01 Q923.617 1482.55 927.228 1482.55 Q930.862 1482.55 932.668 1479.01 Q934.496 1475.44 934.496 1468.33 Q934.496 1461.2 932.668 1457.66 Q930.862 1454.1 927.228 1454.1 M927.228 1450.39 Q933.038 1450.39 936.094 1455 Q939.172 1459.58 939.172 1468.33 Q939.172 1477.06 936.094 1481.67 Q933.038 1486.25 927.228 1486.25 Q921.418 1486.25 918.339 1481.67 Q915.284 1477.06 915.284 1468.33 Q915.284 1459.58 918.339 1455 Q921.418 1450.39 927.228 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1238.56 1451.02 L1256.91 1451.02 L1256.91 1454.96 L1242.84 1454.96 L1242.84 1463.43 Q1243.86 1463.08 1244.88 1462.92 Q1245.89 1462.73 1246.91 1462.73 Q1252.7 1462.73 1256.08 1465.9 Q1259.46 1469.08 1259.46 1474.49 Q1259.46 1480.07 1255.99 1483.17 Q1252.51 1486.25 1246.2 1486.25 Q1244.02 1486.25 1241.75 1485.88 Q1239.51 1485.51 1237.1 1484.77 L1237.1 1480.07 Q1239.18 1481.2 1241.4 1481.76 Q1243.63 1482.32 1246.1 1482.32 Q1250.11 1482.32 1252.45 1480.21 Q1254.78 1478.1 1254.78 1474.49 Q1254.78 1470.88 1252.45 1468.77 Q1250.11 1466.67 1246.1 1466.67 Q1244.23 1466.67 1242.35 1467.08 Q1240.5 1467.5 1238.56 1468.38 L1238.56 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1278.67 1454.1 Q1275.06 1454.1 1273.23 1457.66 Q1271.43 1461.2 1271.43 1468.33 Q1271.43 1475.44 1273.23 1479.01 Q1275.06 1482.55 1278.67 1482.55 Q1282.31 1482.55 1284.11 1479.01 Q1285.94 1475.44 1285.94 1468.33 Q1285.94 1461.2 1284.11 1457.66 Q1282.31 1454.1 1278.67 1454.1 M1278.67 1450.39 Q1284.48 1450.39 1287.54 1455 Q1290.62 1459.58 1290.62 1468.33 Q1290.62 1477.06 1287.54 1481.67 Q1284.48 1486.25 1278.67 1486.25 Q1272.86 1486.25 1269.78 1481.67 Q1266.73 1477.06 1266.73 1468.33 Q1266.73 1459.58 1269.78 1455 Q1272.86 1450.39 1278.67 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1308.83 1454.1 Q1305.22 1454.1 1303.39 1457.66 Q1301.59 1461.2 1301.59 1468.33 Q1301.59 1475.44 1303.39 1479.01 Q1305.22 1482.55 1308.83 1482.55 Q1312.47 1482.55 1314.27 1479.01 Q1316.1 1475.44 1316.1 1468.33 Q1316.1 1461.2 1314.27 1457.66 Q1312.47 1454.1 1308.83 1454.1 M1308.83 1450.39 Q1314.64 1450.39 1317.7 1455 Q1320.78 1459.58 1320.78 1468.33 Q1320.78 1477.06 1317.7 1481.67 Q1314.64 1486.25 1308.83 1486.25 Q1303.02 1486.25 1299.95 1481.67 Q1296.89 1477.06 1296.89 1468.33 Q1296.89 1459.58 1299.95 1455 Q1303.02 1450.39 1308.83 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1620.83 1451.02 L1639.19 1451.02 L1639.19 1454.96 L1625.12 1454.96 L1625.12 1463.43 Q1626.14 1463.08 1627.15 1462.92 Q1628.17 1462.73 1629.19 1462.73 Q1634.98 1462.73 1638.36 1465.9 Q1641.74 1469.08 1641.74 1474.49 Q1641.74 1480.07 1638.26 1483.17 Q1634.79 1486.25 1628.47 1486.25 Q1626.3 1486.25 1624.03 1485.88 Q1621.78 1485.51 1619.38 1484.77 L1619.38 1480.07 Q1621.46 1481.2 1623.68 1481.76 Q1625.9 1482.32 1628.38 1482.32 Q1632.39 1482.32 1634.72 1480.21 Q1637.06 1478.1 1637.06 1474.49 Q1637.06 1470.88 1634.72 1468.77 Q1632.39 1466.67 1628.38 1466.67 Q1626.51 1466.67 1624.63 1467.08 Q1622.78 1467.5 1620.83 1468.38 L1620.83 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1651 1451.02 L1669.35 1451.02 L1669.35 1454.96 L1655.28 1454.96 L1655.28 1463.43 Q1656.3 1463.08 1657.32 1462.92 Q1658.33 1462.73 1659.35 1462.73 Q1665.14 1462.73 1668.52 1465.9 Q1671.9 1469.08 1671.9 1474.49 Q1671.9 1480.07 1668.43 1483.17 Q1664.95 1486.25 1658.63 1486.25 Q1656.46 1486.25 1654.19 1485.88 Q1651.95 1485.51 1649.54 1484.77 L1649.54 1480.07 Q1651.62 1481.2 1653.84 1481.76 Q1656.07 1482.32 1658.54 1482.32 Q1662.55 1482.32 1664.88 1480.21 Q1667.22 1478.1 1667.22 1474.49 Q1667.22 1470.88 1664.88 1468.77 Q1662.55 1466.67 1658.54 1466.67 Q1656.67 1466.67 1654.79 1467.08 Q1652.94 1467.5 1651 1468.38 L1651 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1691.11 1454.1 Q1687.5 1454.1 1685.67 1457.66 Q1683.87 1461.2 1683.87 1468.33 Q1683.87 1475.44 1685.67 1479.01 Q1687.5 1482.55 1691.11 1482.55 Q1694.75 1482.55 1696.55 1479.01 Q1698.38 1475.44 1698.38 1468.33 Q1698.38 1461.2 1696.55 1457.66 Q1694.75 1454.1 1691.11 1454.1 M1691.11 1450.39 Q1696.92 1450.39 1699.98 1455 Q1703.06 1459.58 1703.06 1468.33 Q1703.06 1477.06 1699.98 1481.67 Q1696.92 1486.25 1691.11 1486.25 Q1685.3 1486.25 1682.22 1481.67 Q1679.17 1477.06 1679.17 1468.33 Q1679.17 1459.58 1682.22 1455 Q1685.3 1450.39 1691.11 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M2013.82 1466.44 Q2010.67 1466.44 2008.82 1468.59 Q2006.99 1470.74 2006.99 1474.49 Q2006.99 1478.22 2008.82 1480.39 Q2010.67 1482.55 2013.82 1482.55 Q2016.97 1482.55 2018.79 1480.39 Q2020.65 1478.22 2020.65 1474.49 Q2020.65 1470.74 2018.79 1468.59 Q2016.97 1466.44 2013.82 1466.44 M2023.1 1451.78 L2023.1 1456.04 Q2021.34 1455.21 2019.54 1454.77 Q2017.75 1454.33 2015.99 1454.33 Q2011.36 1454.33 2008.91 1457.45 Q2006.48 1460.58 2006.13 1466.9 Q2007.5 1464.89 2009.56 1463.82 Q2011.62 1462.73 2014.1 1462.73 Q2019.3 1462.73 2022.31 1465.9 Q2025.35 1469.05 2025.35 1474.49 Q2025.35 1479.82 2022.2 1483.03 Q2019.05 1486.25 2013.82 1486.25 Q2007.82 1486.25 2004.65 1481.67 Q2001.48 1477.06 2001.48 1468.33 Q2001.48 1460.14 2005.37 1455.28 Q2009.26 1450.39 2015.81 1450.39 Q2017.57 1450.39 2019.35 1450.74 Q2021.16 1451.09 2023.1 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M2043.4 1454.1 Q2039.79 1454.1 2037.96 1457.66 Q2036.16 1461.2 2036.16 1468.33 Q2036.16 1475.44 2037.96 1479.01 Q2039.79 1482.55 2043.4 1482.55 Q2047.04 1482.55 2048.84 1479.01 Q2050.67 1475.44 2050.67 1468.33 Q2050.67 1461.2 2048.84 1457.66 Q2047.04 1454.1 2043.4 1454.1 M2043.4 1450.39 Q2049.21 1450.39 2052.27 1455 Q2055.35 1459.58 2055.35 1468.33 Q2055.35 1477.06 2052.27 1481.67 Q2049.21 1486.25 2043.4 1486.25 Q2037.59 1486.25 2034.51 1481.67 Q2031.46 1477.06 2031.46 1468.33 Q2031.46 1459.58 2034.51 1455 Q2037.59 1450.39 2043.4 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M2073.56 1454.1 Q2069.95 1454.1 2068.12 1457.66 Q2066.32 1461.2 2066.32 1468.33 Q2066.32 1475.44 2068.12 1479.01 Q2069.95 1482.55 2073.56 1482.55 Q2077.2 1482.55 2079 1479.01 Q2080.83 1475.44 2080.83 1468.33 Q2080.83 1461.2 2079 1457.66 Q2077.2 1454.1 2073.56 1454.1 M2073.56 1450.39 Q2079.37 1450.39 2082.43 1455 Q2085.51 1459.58 2085.51 1468.33 Q2085.51 1477.06 2082.43 1481.67 Q2079.37 1486.25 2073.56 1486.25 Q2067.75 1486.25 2064.67 1481.67 Q2061.62 1477.06 2061.62 1468.33 Q2061.62 1459.58 2064.67 1455 Q2067.75 1450.39 2073.56 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1149.83 1525.81 L1149.83 1543.66 L1157.91 1543.66 Q1162.4 1543.66 1164.85 1541.34 Q1167.3 1539.02 1167.3 1534.72 Q1167.3 1530.45 1164.85 1528.13 Q1162.4 1525.81 1157.91 1525.81 L1149.83 1525.81 M1143.4 1520.52 L1157.91 1520.52 Q1165.9 1520.52 1169.97 1524.15 Q1174.08 1527.75 1174.08 1534.72 Q1174.08 1541.75 1169.97 1545.35 Q1165.9 1548.95 1157.91 1548.95 L1149.83 1548.95 L1149.83 1568.04 L1143.4 1568.04 L1143.4 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1201.96 1537.87 Q1200.97 1537.3 1199.8 1537.04 Q1198.65 1536.76 1197.25 1536.76 Q1192.29 1536.76 1189.61 1540 Q1186.97 1543.22 1186.97 1549.27 L1186.97 1568.04 L1181.08 1568.04 L1181.08 1532.4 L1186.97 1532.4 L1186.97 1537.93 Q1188.82 1534.69 1191.78 1533.13 Q1194.74 1531.54 1198.97 1531.54 Q1199.57 1531.54 1200.31 1531.63 Q1201.04 1531.7 1201.93 1531.85 L1201.96 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1220.49 1536.5 Q1215.77 1536.5 1213.04 1540.19 Q1210.3 1543.85 1210.3 1550.25 Q1210.3 1556.65 1213.01 1560.34 Q1215.74 1564 1220.49 1564 Q1225.16 1564 1227.9 1560.31 Q1230.64 1556.62 1230.64 1550.25 Q1230.64 1543.92 1227.9 1540.23 Q1225.16 1536.5 1220.49 1536.5 M1220.49 1531.54 Q1228.12 1531.54 1232.48 1536.5 Q1236.85 1541.47 1236.85 1550.25 Q1236.85 1559 1232.48 1564 Q1228.12 1568.97 1220.49 1568.97 Q1212.81 1568.97 1208.45 1564 Q1204.13 1559 1204.13 1550.25 Q1204.13 1541.47 1208.45 1536.5 Q1212.81 1531.54 1220.49 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1264.6 1518.52 L1264.6 1523.39 L1259 1523.39 Q1255.85 1523.39 1254.61 1524.66 Q1253.4 1525.93 1253.4 1529.24 L1253.4 1532.4 L1263.04 1532.4 L1263.04 1536.95 L1253.4 1536.95 L1253.4 1568.04 L1247.51 1568.04 L1247.51 1536.95 L1241.91 1536.95 L1241.91 1532.4 L1247.51 1532.4 L1247.51 1529.91 Q1247.51 1523.96 1250.28 1521.26 Q1253.05 1518.52 1259.06 1518.52 L1264.6 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1269.5 1532.4 L1275.36 1532.4 L1275.36 1568.04 L1269.5 1568.04 L1269.5 1532.4 M1269.5 1518.52 L1275.36 1518.52 L1275.36 1525.93 L1269.5 1525.93 L1269.5 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1293.4 1522.27 L1293.4 1532.4 L1305.47 1532.4 L1305.47 1536.95 L1293.4 1536.95 L1293.4 1556.3 Q1293.4 1560.66 1294.58 1561.9 Q1295.79 1563.14 1299.45 1563.14 L1305.47 1563.14 L1305.47 1568.04 L1299.45 1568.04 Q1292.67 1568.04 1290.09 1565.53 Q1287.52 1562.98 1287.52 1556.3 L1287.52 1536.95 L1283.22 1536.95 L1283.22 1532.4 L1287.52 1532.4 L1287.52 1522.27 L1293.4 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1333.35 1518.52 L1346.84 1518.52 L1346.84 1523.07 L1339.21 1523.07 L1339.21 1572.09 L1346.84 1572.09 L1346.84 1576.64 L1333.35 1576.64 L1333.35 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1375.2 1577.62 L1372.02 1577.62 L1371.99 1568.04 Q1368.65 1567.98 1365.31 1567.28 Q1361.96 1566.55 1358.59 1565.12 L1358.59 1559.39 Q1361.84 1561.42 1365.15 1562.47 Q1368.49 1563.49 1372.02 1563.52 L1372.02 1549.01 Q1364.99 1547.86 1361.77 1545.13 Q1358.59 1542.39 1358.59 1537.62 Q1358.59 1532.43 1362.06 1529.44 Q1365.53 1526.44 1372.02 1526 L1372.02 1518.52 L1375.2 1518.52 L1375.2 1525.9 Q1378.16 1526.03 1380.93 1526.54 Q1383.7 1527.02 1386.34 1527.88 L1386.34 1533.45 Q1383.7 1532.11 1380.9 1531.38 Q1378.13 1530.65 1375.2 1530.52 L1375.2 1544.11 Q1382.43 1545.22 1385.83 1548.09 Q1389.24 1550.95 1389.24 1555.92 Q1389.24 1561.3 1385.61 1564.42 Q1382.02 1567.5 1375.2 1567.98 L1375.2 1577.62 M1372.02 1543.54 L1372.02 1530.49 Q1368.33 1530.9 1366.39 1532.59 Q1364.45 1534.27 1364.45 1537.07 Q1364.45 1539.81 1366.23 1541.34 Q1368.04 1542.87 1372.02 1543.54 M1375.2 1549.65 L1375.2 1563.43 Q1379.25 1562.89 1381.28 1561.14 Q1383.35 1559.39 1383.35 1556.52 Q1383.35 1553.72 1381.38 1552.07 Q1379.44 1550.41 1375.2 1549.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M1414.48 1518.52 L1414.48 1576.64 L1400.98 1576.64 L1400.98 1572.09 L1408.59 1572.09 L1408.59 1523.07 L1400.98 1523.07 L1400.98 1518.52 L1414.48 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,1095.58 2352.76,1095.58 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,767.972 2352.76,767.972 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,440.369 2352.76,440.369 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="205.121,112.765 2352.76,112.765 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1423.18 205.121,47.2441 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1423.18 224.019,1423.18 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,1095.58 224.019,1095.58 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,767.972 224.019,767.972 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,440.369 224.019,440.369 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="205.121,112.765 224.019,112.765 "/>
+<path clip-path="url(#clip370)" d="M157.177 1408.98 Q153.566 1408.98 151.737 1412.54 Q149.931 1416.08 149.931 1423.21 Q149.931 1430.32 151.737 1433.89 Q153.566 1437.43 157.177 1437.43 Q160.811 1437.43 162.616 1433.89 Q164.445 1430.32 164.445 1423.21 Q164.445 1416.08 162.616 1412.54 Q160.811 1408.98 157.177 1408.98 M157.177 1405.27 Q162.987 1405.27 166.042 1409.88 Q169.121 1414.46 169.121 1423.21 Q169.121 1431.94 166.042 1436.55 Q162.987 1441.13 157.177 1441.13 Q151.366 1441.13 148.288 1436.55 Q145.232 1431.94 145.232 1423.21 Q145.232 1414.46 148.288 1409.88 Q151.366 1405.27 157.177 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M117.825 1108.92 L125.464 1108.92 L125.464 1082.56 L117.154 1084.22 L117.154 1079.96 L125.418 1078.3 L130.093 1078.3 L130.093 1108.92 L137.732 1108.92 L137.732 1112.86 L117.825 1112.86 L117.825 1108.92 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M157.177 1081.37 Q153.566 1081.37 151.737 1084.94 Q149.931 1088.48 149.931 1095.61 Q149.931 1102.72 151.737 1106.28 Q153.566 1109.82 157.177 1109.82 Q160.811 1109.82 162.616 1106.28 Q164.445 1102.72 164.445 1095.61 Q164.445 1088.48 162.616 1084.94 Q160.811 1081.37 157.177 1081.37 M157.177 1077.67 Q162.987 1077.67 166.042 1082.28 Q169.121 1086.86 169.121 1095.61 Q169.121 1104.34 166.042 1108.94 Q162.987 1113.53 157.177 1113.53 Q151.366 1113.53 148.288 1108.94 Q145.232 1104.34 145.232 1095.61 Q145.232 1086.86 148.288 1082.28 Q151.366 1077.67 157.177 1077.67 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M121.043 781.317 L137.362 781.317 L137.362 785.252 L115.418 785.252 L115.418 781.317 Q118.08 778.563 122.663 773.933 Q127.269 769.28 128.45 767.938 Q130.695 765.414 131.575 763.678 Q132.478 761.919 132.478 760.229 Q132.478 757.475 130.533 755.739 Q128.612 754.003 125.51 754.003 Q123.311 754.003 120.857 754.766 Q118.427 755.53 115.649 757.081 L115.649 752.359 Q118.473 751.225 120.927 750.646 Q123.38 750.067 125.418 750.067 Q130.788 750.067 133.982 752.753 Q137.177 755.438 137.177 759.928 Q137.177 762.058 136.367 763.979 Q135.579 765.877 133.473 768.47 Q132.894 769.141 129.792 772.359 Q126.691 775.553 121.043 781.317 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M157.177 753.771 Q153.566 753.771 151.737 757.336 Q149.931 760.877 149.931 768.007 Q149.931 775.114 151.737 778.678 Q153.566 782.22 157.177 782.22 Q160.811 782.22 162.616 778.678 Q164.445 775.114 164.445 768.007 Q164.445 760.877 162.616 757.336 Q160.811 753.771 157.177 753.771 M157.177 750.067 Q162.987 750.067 166.042 754.674 Q169.121 759.257 169.121 768.007 Q169.121 776.734 166.042 781.34 Q162.987 785.924 157.177 785.924 Q151.366 785.924 148.288 781.34 Q145.232 776.734 145.232 768.007 Q145.232 759.257 148.288 754.674 Q151.366 750.067 157.177 750.067 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M131.181 439.014 Q134.538 439.732 136.413 442.001 Q138.311 444.269 138.311 447.602 Q138.311 452.718 134.792 455.519 Q131.274 458.32 124.793 458.32 Q122.617 458.32 120.302 457.88 Q118.01 457.463 115.556 456.607 L115.556 452.093 Q117.501 453.227 119.816 453.806 Q122.13 454.385 124.654 454.385 Q129.052 454.385 131.343 452.649 Q133.658 450.913 133.658 447.602 Q133.658 444.547 131.505 442.834 Q129.376 441.098 125.556 441.098 L121.529 441.098 L121.529 437.255 L125.742 437.255 Q129.191 437.255 131.019 435.889 Q132.848 434.501 132.848 431.908 Q132.848 429.246 130.95 427.834 Q129.075 426.399 125.556 426.399 Q123.635 426.399 121.436 426.815 Q119.237 427.232 116.598 428.112 L116.598 423.945 Q119.26 423.204 121.575 422.834 Q123.913 422.464 125.973 422.464 Q131.297 422.464 134.399 424.894 Q137.501 427.302 137.501 431.422 Q137.501 434.292 135.857 436.283 Q134.214 438.251 131.181 439.014 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M157.177 426.167 Q153.566 426.167 151.737 429.732 Q149.931 433.274 149.931 440.403 Q149.931 447.51 151.737 451.075 Q153.566 454.616 157.177 454.616 Q160.811 454.616 162.616 451.075 Q164.445 447.51 164.445 440.403 Q164.445 433.274 162.616 429.732 Q160.811 426.167 157.177 426.167 M157.177 422.464 Q162.987 422.464 166.042 427.07 Q169.121 431.653 169.121 440.403 Q169.121 449.13 166.042 453.737 Q162.987 458.32 157.177 458.32 Q151.366 458.32 148.288 453.737 Q145.232 449.13 145.232 440.403 Q145.232 431.653 148.288 427.07 Q151.366 422.464 157.177 422.464 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M129.862 99.5589 L118.056 118.008 L129.862 118.008 L129.862 99.5589 M128.635 95.4848 L134.515 95.4848 L134.515 118.008 L139.445 118.008 L139.445 121.897 L134.515 121.897 L134.515 130.045 L129.862 130.045 L129.862 121.897 L114.26 121.897 L114.26 117.383 L128.635 95.4848 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M157.177 98.5635 Q153.566 98.5635 151.737 102.128 Q149.931 105.67 149.931 112.8 Q149.931 119.906 151.737 123.471 Q153.566 127.012 157.177 127.012 Q160.811 127.012 162.616 123.471 Q164.445 119.906 164.445 112.8 Q164.445 105.67 162.616 102.128 Q160.811 98.5635 157.177 98.5635 M157.177 94.8598 Q162.987 94.8598 166.042 99.4663 Q169.121 104.05 169.121 112.8 Q169.121 121.526 166.042 126.133 Q162.987 130.716 157.177 130.716 Q151.366 130.716 148.288 126.133 Q145.232 121.526 145.232 112.8 Q145.232 104.05 148.288 99.4663 Q151.366 94.8598 157.177 94.8598 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M16.4842 1073.5 L16.4842 1064.84 L56.238 1043.77 L16.4842 1043.77 L16.4842 1037.54 L64.0042 1037.54 L64.0042 1046.19 L24.2503 1067.26 L64.0042 1067.26 L64.0042 1073.5 L16.4842 1073.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M49.9359 1025.6 L28.3562 1025.6 L28.3562 1019.74 L49.7131 1019.74 Q54.7739 1019.74 57.3202 1017.77 Q59.8346 1015.8 59.8346 1011.85 Q59.8346 1007.11 56.8109 1004.37 Q53.7872 1001.6 48.5673 1001.6 L28.3562 1001.6 L28.3562 995.744 L64.0042 995.744 L64.0042 1001.6 L58.5296 1001.6 Q61.7762 1003.73 63.3676 1006.57 Q64.9272 1009.37 64.9272 1013.09 Q64.9272 1019.23 61.1078 1022.42 Q57.2883 1025.6 49.9359 1025.6 M27.4968 1010.86 L27.4968 1010.86 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M35.1993 955.927 Q31.2526 953.731 29.3747 950.675 Q27.4968 947.62 27.4968 943.482 Q27.4968 937.912 31.4117 934.888 Q35.2948 931.865 42.4881 931.865 L64.0042 931.865 L64.0042 937.753 L42.679 937.753 Q37.5546 937.753 35.072 939.567 Q32.5894 941.381 32.5894 945.105 Q32.5894 949.657 35.6131 952.298 Q38.6368 954.94 43.8567 954.94 L64.0042 954.94 L64.0042 960.829 L42.679 960.829 Q37.5228 960.829 35.072 962.643 Q32.5894 964.457 32.5894 968.245 Q32.5894 972.732 35.6449 975.374 Q38.6686 978.016 43.8567 978.016 L64.0042 978.016 L64.0042 983.904 L28.3562 983.904 L28.3562 978.016 L33.8944 978.016 Q30.616 976.011 29.0564 973.21 Q27.4968 970.409 27.4968 966.558 Q27.4968 962.675 29.4702 959.969 Q31.4436 957.232 35.1993 955.927 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M46.212 894.593 Q39.7508 894.593 36.0905 897.267 Q32.3984 899.909 32.3984 904.556 Q32.3984 909.203 36.0905 911.876 Q39.7508 914.518 46.212 914.518 Q52.6732 914.518 56.3653 911.876 Q60.0256 909.203 60.0256 904.556 Q60.0256 899.909 56.3653 897.267 Q52.6732 894.593 46.212 894.593 M33.7671 914.518 Q30.5842 912.672 29.0564 909.871 Q27.4968 907.038 27.4968 903.123 Q27.4968 896.63 32.6531 892.588 Q37.8093 888.514 46.212 888.514 Q54.6147 888.514 59.771 892.588 Q64.9272 896.63 64.9272 903.123 Q64.9272 907.038 63.3994 909.871 Q61.8398 912.672 58.657 914.518 L64.0042 914.518 L64.0042 920.406 L14.479 920.406 L14.479 914.518 L33.7671 914.518 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M44.7161 848.315 L47.5806 848.315 L47.5806 875.242 Q53.6281 874.86 56.8109 871.613 Q59.9619 868.335 59.9619 862.51 Q59.9619 859.136 59.1344 855.985 Q58.3069 852.802 56.6518 849.683 L62.1899 849.683 Q63.5267 852.834 64.227 856.144 Q64.9272 859.455 64.9272 862.86 Q64.9272 871.39 59.9619 876.387 Q54.9967 881.353 46.5303 881.353 Q37.7774 881.353 32.6531 876.642 Q27.4968 871.9 27.4968 863.879 Q27.4968 856.686 32.1438 852.516 Q36.7589 848.315 44.7161 848.315 M42.9973 854.171 Q38.1912 854.235 35.3266 856.877 Q32.4621 859.486 32.4621 863.815 Q32.4621 868.717 35.2312 871.677 Q38.0002 874.605 43.0292 875.051 L42.9973 854.171 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M33.8307 818.046 Q33.2578 819.032 33.0032 820.21 Q32.7167 821.356 32.7167 822.756 Q32.7167 827.722 35.9632 830.395 Q39.1779 833.037 45.2253 833.037 L64.0042 833.037 L64.0042 838.925 L28.3562 838.925 L28.3562 833.037 L33.8944 833.037 Q30.6479 831.191 29.0883 828.231 Q27.4968 825.271 27.4968 821.038 Q27.4968 820.433 27.5923 819.701 Q27.656 818.969 27.8151 818.078 L33.8307 818.046 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M32.4621 777.369 Q32.4621 782.079 36.1542 784.817 Q39.8145 787.554 46.212 787.554 Q52.6095 787.554 56.3017 784.849 Q59.9619 782.111 59.9619 777.369 Q59.9619 772.69 56.2698 769.953 Q52.5777 767.216 46.212 767.216 Q39.8781 767.216 36.186 769.953 Q32.4621 772.69 32.4621 777.369 M27.4968 777.369 Q27.4968 769.73 32.4621 765.37 Q37.4273 761.009 46.212 761.009 Q54.9649 761.009 59.9619 765.37 Q64.9272 769.73 64.9272 777.369 Q64.9272 785.04 59.9619 789.4 Q54.9649 793.729 46.212 793.729 Q37.4273 793.729 32.4621 789.4 Q27.4968 785.04 27.4968 777.369 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M14.479 733.255 L19.3487 733.255 L19.3487 738.856 Q19.3487 742.007 20.6219 743.249 Q21.895 744.458 25.2052 744.458 L28.3562 744.458 L28.3562 734.814 L32.9077 734.814 L32.9077 744.458 L64.0042 744.458 L64.0042 750.346 L32.9077 750.346 L32.9077 755.948 L28.3562 755.948 L28.3562 750.346 L25.8736 750.346 Q19.9216 750.346 17.2162 747.577 Q14.479 744.808 14.479 738.793 L14.479 733.255 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M32.4621 693.819 Q32.4621 698.53 36.1542 701.267 Q39.8145 704.004 46.212 704.004 Q52.6095 704.004 56.3017 701.299 Q59.9619 698.561 59.9619 693.819 Q59.9619 689.14 56.2698 686.403 Q52.5777 683.666 46.212 683.666 Q39.8781 683.666 36.186 686.403 Q32.4621 689.14 32.4621 693.819 M27.4968 693.819 Q27.4968 686.18 32.4621 681.82 Q37.4273 677.459 46.212 677.459 Q54.9649 677.459 59.9619 681.82 Q64.9272 686.18 64.9272 693.819 Q64.9272 701.49 59.9619 705.85 Q54.9649 710.179 46.212 710.179 Q37.4273 710.179 32.4621 705.85 Q27.4968 701.49 27.4968 693.819 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M49.9359 668.356 L28.3562 668.356 L28.3562 662.5 L49.7131 662.5 Q54.7739 662.5 57.3202 660.526 Q59.8346 658.553 59.8346 654.606 Q59.8346 649.864 56.8109 647.127 Q53.7872 644.357 48.5673 644.357 L28.3562 644.357 L28.3562 638.501 L64.0042 638.501 L64.0042 644.357 L58.5296 644.357 Q61.7762 646.49 63.3676 649.323 Q64.9272 652.124 64.9272 655.848 Q64.9272 661.99 61.1078 665.173 Q57.2883 668.356 49.9359 668.356 M27.4968 653.62 L27.4968 653.62 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M18.2347 620.645 L28.3562 620.645 L28.3562 608.582 L32.9077 608.582 L32.9077 620.645 L52.2594 620.645 Q56.6199 620.645 57.8613 619.468 Q59.1026 618.258 59.1026 614.598 L59.1026 608.582 L64.0042 608.582 L64.0042 614.598 Q64.0042 621.377 61.4897 623.955 Q58.9434 626.533 52.2594 626.533 L32.9077 626.533 L32.9077 630.83 L28.3562 630.83 L28.3562 626.533 L18.2347 626.533 L18.2347 620.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M29.7248 575.226 L35.1993 575.226 Q33.8307 577.708 33.1623 580.223 Q32.4621 582.706 32.4621 585.252 Q32.4621 590.949 36.0905 594.1 Q39.6872 597.251 46.212 597.251 Q52.7369 597.251 56.3653 594.1 Q59.9619 590.949 59.9619 585.252 Q59.9619 582.706 59.2935 580.223 Q58.5933 577.708 57.2247 575.226 L62.6355 575.226 Q63.7814 577.677 64.3543 580.318 Q64.9272 582.928 64.9272 585.888 Q64.9272 593.941 59.8664 598.683 Q54.8057 603.426 46.212 603.426 Q37.491 603.426 32.4939 598.652 Q27.4968 593.846 27.4968 585.506 Q27.4968 582.801 28.0697 580.223 Q28.6108 577.645 29.7248 575.226 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M32.4621 551.227 Q32.4621 555.938 36.1542 558.675 Q39.8145 561.412 46.212 561.412 Q52.6095 561.412 56.3017 558.707 Q59.9619 555.97 59.9619 551.227 Q59.9619 546.548 56.2698 543.811 Q52.5777 541.074 46.212 541.074 Q39.8781 541.074 36.186 543.811 Q32.4621 546.548 32.4621 551.227 M27.4968 551.227 Q27.4968 543.588 32.4621 539.228 Q37.4273 534.867 46.212 534.867 Q54.9649 534.867 59.9619 539.228 Q64.9272 543.588 64.9272 551.227 Q64.9272 558.898 59.9619 563.258 Q54.9649 567.587 46.212 567.587 Q37.4273 567.587 32.4621 563.258 Q27.4968 558.898 27.4968 551.227 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M35.1993 497.405 Q31.2526 495.209 29.3747 492.153 Q27.4968 489.098 27.4968 484.96 Q27.4968 479.39 31.4117 476.366 Q35.2948 473.343 42.4881 473.343 L64.0042 473.343 L64.0042 479.231 L42.679 479.231 Q37.5546 479.231 35.072 481.045 Q32.5894 482.859 32.5894 486.583 Q32.5894 491.135 35.6131 493.777 Q38.6368 496.418 43.8567 496.418 L64.0042 496.418 L64.0042 502.307 L42.679 502.307 Q37.5228 502.307 35.072 504.121 Q32.5894 505.935 32.5894 509.723 Q32.5894 514.211 35.6449 516.852 Q38.6686 519.494 43.8567 519.494 L64.0042 519.494 L64.0042 525.382 L28.3562 525.382 L28.3562 519.494 L33.8944 519.494 Q30.616 517.489 29.0564 514.688 Q27.4968 511.887 27.4968 508.036 Q27.4968 504.153 29.4702 501.447 Q31.4436 498.71 35.1993 497.405 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M44.7161 431.17 L47.5806 431.17 L47.5806 458.097 Q53.6281 457.715 56.8109 454.468 Q59.9619 451.19 59.9619 445.365 Q59.9619 441.992 59.1344 438.841 Q58.3069 435.658 56.6518 432.539 L62.1899 432.539 Q63.5267 435.69 64.227 439 Q64.9272 442.31 64.9272 445.716 Q64.9272 454.246 59.9619 459.243 Q54.9967 464.208 46.5303 464.208 Q37.7774 464.208 32.6531 459.497 Q27.4968 454.755 27.4968 446.734 Q27.4968 439.541 32.1438 435.371 Q36.7589 431.17 44.7161 431.17 M42.9973 437.026 Q38.1912 437.09 35.3266 439.732 Q32.4621 442.342 32.4621 446.67 Q32.4621 451.572 35.2312 454.532 Q38.0002 457.46 43.0292 457.906 L42.9973 437.026 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M29.4065 398.832 L34.9447 398.832 Q33.6716 401.315 33.035 403.988 Q32.3984 406.662 32.3984 409.527 Q32.3984 413.887 33.7352 416.083 Q35.072 418.248 37.7456 418.248 Q39.7826 418.248 40.9603 416.688 Q42.1061 415.128 43.1565 410.418 L43.6021 408.413 Q44.9389 402.174 47.3897 399.564 Q49.8086 396.922 54.1691 396.922 Q59.1344 396.922 62.0308 400.869 Q64.9272 404.784 64.9272 411.659 Q64.9272 414.524 64.3543 417.643 Q63.8132 420.73 62.6992 424.168 L56.6518 424.168 Q58.3387 420.921 59.198 417.77 Q60.0256 414.619 60.0256 411.532 Q60.0256 407.394 58.6251 405.166 Q57.1929 402.938 54.6147 402.938 Q52.2276 402.938 50.9545 404.561 Q49.6813 406.153 48.5037 411.595 L48.0262 413.632 Q46.8804 419.075 44.5251 421.494 Q42.138 423.913 38.0002 423.913 Q32.9713 423.913 30.2341 420.348 Q27.4968 416.783 27.4968 410.227 Q27.4968 406.98 27.9743 404.116 Q28.4517 401.251 29.4065 398.832 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip372)" d="M323.245 1357.66 L323.245 1423.18 L514.383 1423.18 L514.383 1357.66 L323.245 1357.66 L323.245 1357.66  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="323.245,1357.66 323.245,1423.18 514.383,1423.18 514.383,1357.66 323.245,1357.66 "/>
+<path clip-path="url(#clip372)" d="M514.383 1324.9 L514.383 1423.18 L705.522 1423.18 L705.522 1324.9 L514.383 1324.9 L514.383 1324.9  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="514.383,1324.9 514.383,1423.18 705.522,1423.18 705.522,1324.9 514.383,1324.9 "/>
+<path clip-path="url(#clip372)" d="M705.522 1193.86 L705.522 1423.18 L896.661 1423.18 L896.661 1193.86 L705.522 1193.86 L705.522 1193.86  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="705.522,1193.86 705.522,1423.18 896.661,1423.18 896.661,1193.86 705.522,1193.86 "/>
+<path clip-path="url(#clip372)" d="M896.661 1193.86 L896.661 1423.18 L1087.8 1423.18 L1087.8 1193.86 L896.661 1193.86 L896.661 1193.86  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="896.661,1193.86 896.661,1423.18 1087.8,1423.18 1087.8,1193.86 896.661,1193.86 "/>
+<path clip-path="url(#clip372)" d="M1087.8 1292.14 L1087.8 1423.18 L1278.94 1423.18 L1278.94 1292.14 L1087.8 1292.14 L1087.8 1292.14  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1087.8,1292.14 1087.8,1423.18 1278.94,1423.18 1278.94,1292.14 1087.8,1292.14 "/>
+<path clip-path="url(#clip372)" d="M1278.94 1357.66 L1278.94 1423.18 L1470.08 1423.18 L1470.08 1357.66 L1278.94 1357.66 L1278.94 1357.66  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1278.94,1357.66 1278.94,1423.18 1470.08,1423.18 1470.08,1357.66 1278.94,1357.66 "/>
+<path clip-path="url(#clip372)" d="M1470.08 1259.38 L1470.08 1423.18 L1661.22 1423.18 L1661.22 1259.38 L1470.08 1259.38 L1470.08 1259.38  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1470.08,1259.38 1470.08,1423.18 1661.22,1423.18 1661.22,1259.38 1470.08,1259.38 "/>
+<path clip-path="url(#clip372)" d="M1661.22 1062.82 L1661.22 1423.18 L1852.35 1423.18 L1852.35 1062.82 L1661.22 1062.82 L1661.22 1062.82  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1661.22,1062.82 1661.22,1423.18 1852.35,1423.18 1852.35,1062.82 1661.22,1062.82 "/>
+<path clip-path="url(#clip372)" d="M1852.35 866.253 L1852.35 1423.18 L2043.49 1423.18 L2043.49 866.253 L1852.35 866.253 L1852.35 866.253  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1852.35,866.253 1852.35,1423.18 2043.49,1423.18 2043.49,866.253 1852.35,866.253 "/>
+<path clip-path="url(#clip372)" d="M2043.49 47.2441 L2043.49 1423.18 L2234.63 1423.18 L2234.63 47.2441 L2043.49 47.2441 L2043.49 47.2441  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip372)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2043.49,47.2441 2043.49,1423.18 2234.63,1423.18 2234.63,47.2441 2043.49,47.2441 "/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="418.814" cy="1357.66" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="609.953" cy="1324.9" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="801.092" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="992.23" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1183.37" cy="1292.14" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1374.51" cy="1357.66" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1565.65" cy="1259.38" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1756.79" cy="1062.82" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1947.92" cy="866.253" r="2"/>
+<circle clip-path="url(#clip372)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2139.06" cy="47.2441" r="2"/>
+<polyline clip-path="url(#clip372)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1701.21,2799.12 1701.21,-1328.69 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="1336.14,2799.12 1336.14,-1328.69 "/>
+<polyline clip-path="url(#clip372)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="378.498,2799.12 378.498,-1328.69 "/>
+<path clip-path="url(#clip370)" d="M276.709 300.469 L660.893 300.469 L660.893 93.1086 L276.709 93.1086  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip370)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="276.709,300.469 660.893,300.469 660.893,93.1086 276.709,93.1086 276.709,300.469 "/>
+<polyline clip-path="url(#clip370)" style="stroke:#e26f46; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="300.571,144.949 443.747,144.949 "/>
+<path clip-path="url(#clip370)" d="M475.225 140.377 L480.897 156.395 L488.998 136.303 L493.512 136.303 L483.026 162.229 L483.026 172.09 L478.767 172.09 L478.767 162.229 L471.962 142.877 Q470.943 139.96 468.744 139.96 L467.61 139.96 L467.61 136.303 L469.23 136.303 Q473.79 136.303 475.225 140.377 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M515.017 140.701 L544.693 140.701 L544.693 144.59 L515.017 144.59 L515.017 140.701 M515.017 150.145 L544.693 150.145 L544.693 154.081 L515.017 154.081 L515.017 150.145 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M570.665 158.293 L578.304 158.293 L578.304 131.928 L569.993 133.595 L569.993 129.335 L578.257 127.669 L582.933 127.669 L582.933 158.293 L590.572 158.293 L590.572 162.229 L570.665 162.229 L570.665 158.293 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M600.016 156.349 L604.901 156.349 L604.901 162.229 L600.016 162.229 L600.016 156.349 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M625.086 130.747 Q621.475 130.747 619.646 134.312 Q617.84 137.854 617.84 144.983 Q617.84 152.09 619.646 155.655 Q621.475 159.196 625.086 159.196 Q628.72 159.196 630.525 155.655 Q632.354 152.09 632.354 144.983 Q632.354 137.854 630.525 134.312 Q628.72 130.747 625.086 130.747 M625.086 127.044 Q630.896 127.044 633.951 131.65 Q637.03 136.233 637.03 144.983 Q637.03 153.71 633.951 158.317 Q630.896 162.9 625.086 162.9 Q619.276 162.9 616.197 158.317 Q613.141 153.71 613.141 144.983 Q613.141 136.233 616.197 131.65 Q619.276 127.044 625.086 127.044 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip370)" style="stroke:#3da44d; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="300.571,196.789 443.747,196.789 "/>
+<path clip-path="url(#clip370)" d="M475.225 192.217 L480.897 208.235 L488.998 188.143 L493.512 188.143 L483.026 214.069 L483.026 223.93 L478.767 223.93 L478.767 214.069 L471.962 194.717 Q470.943 191.8 468.744 191.8 L467.61 191.8 L467.61 188.143 L469.23 188.143 Q473.79 188.143 475.225 192.217 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M515.017 192.541 L544.693 192.541 L544.693 196.43 L515.017 196.43 L515.017 192.541 M515.017 201.985 L544.693 201.985 L544.693 205.921 L515.017 205.921 L515.017 201.985 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M579.854 182.587 Q576.243 182.587 574.415 186.152 Q572.609 189.694 572.609 196.823 Q572.609 203.93 574.415 207.495 Q576.243 211.036 579.854 211.036 Q583.489 211.036 585.294 207.495 Q587.123 203.93 587.123 196.823 Q587.123 189.694 585.294 186.152 Q583.489 182.587 579.854 182.587 M579.854 178.884 Q585.665 178.884 588.72 183.49 Q591.799 188.073 591.799 196.823 Q591.799 205.55 588.72 210.157 Q585.665 214.74 579.854 214.74 Q574.044 214.74 570.966 210.157 Q567.91 205.55 567.91 196.823 Q567.91 188.073 570.966 183.49 Q574.044 178.884 579.854 178.884 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M600.016 208.189 L604.901 208.189 L604.901 214.069 L600.016 214.069 L600.016 208.189 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M615.132 179.509 L633.488 179.509 L633.488 183.444 L619.414 183.444 L619.414 191.916 Q620.433 191.569 621.451 191.407 Q622.47 191.222 623.488 191.222 Q629.275 191.222 632.655 194.393 Q636.035 197.564 636.035 202.981 Q636.035 208.559 632.563 211.661 Q629.09 214.74 622.771 214.74 Q620.595 214.74 618.326 214.37 Q616.081 213.999 613.674 213.258 L613.674 208.559 Q615.757 209.694 617.979 210.249 Q620.201 210.805 622.678 210.805 Q626.683 210.805 629.021 208.698 Q631.359 206.592 631.359 202.981 Q631.359 199.37 629.021 197.263 Q626.683 195.157 622.678 195.157 Q620.803 195.157 618.928 195.573 Q617.076 195.99 615.132 196.87 L615.132 179.509 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip370)" style="stroke:#c271d2; stroke-linecap:round; stroke-linejoin:round; stroke-width:12; stroke-opacity:1; fill:none" points="300.571,248.629 443.747,248.629 "/>
+<path clip-path="url(#clip370)" d="M475.225 244.057 L480.897 260.075 L488.998 239.983 L493.512 239.983 L483.026 265.909 L483.026 275.77 L478.767 275.77 L478.767 265.909 L471.962 246.557 Q470.943 243.64 468.744 243.64 L467.61 243.64 L467.61 239.983 L469.23 239.983 Q473.79 239.983 475.225 244.057 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M515.017 244.381 L544.693 244.381 L544.693 248.27 L515.017 248.27 L515.017 244.381 M515.017 253.825 L544.693 253.825 L544.693 257.761 L515.017 257.761 L515.017 253.825 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M579.854 234.427 Q576.243 234.427 574.415 237.992 Q572.609 241.534 572.609 248.663 Q572.609 255.77 574.415 259.335 Q576.243 262.876 579.854 262.876 Q583.489 262.876 585.294 259.335 Q587.123 255.77 587.123 248.663 Q587.123 241.534 585.294 237.992 Q583.489 234.427 579.854 234.427 M579.854 230.724 Q585.665 230.724 588.72 235.33 Q591.799 239.913 591.799 248.663 Q591.799 257.39 588.72 261.997 Q585.665 266.58 579.854 266.58 Q574.044 266.58 570.966 261.997 Q567.91 257.39 567.91 248.663 Q567.91 239.913 570.966 235.33 Q574.044 230.724 579.854 230.724 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M600.016 260.029 L604.901 260.029 L604.901 265.909 L600.016 265.909 L600.016 260.029 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip370)" d="M625.086 234.427 Q621.475 234.427 619.646 237.992 Q617.84 241.534 617.84 248.663 Q617.84 255.77 619.646 259.335 Q621.475 262.876 625.086 262.876 Q628.72 262.876 630.525 259.335 Q632.354 255.77 632.354 248.663 Q632.354 241.534 630.525 237.992 Q628.72 234.427 625.086 234.427 M625.086 230.724 Q630.896 230.724 633.951 235.33 Q637.03 239.913 637.03 248.663 Q637.03 257.39 633.951 261.997 Q630.896 266.58 625.086 266.58 Q619.276 266.58 616.197 261.997 Q613.141 257.39 613.141 248.663 Q613.141 239.913 616.197 235.33 Q619.276 230.724 625.086 230.724 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/applications/two_stage_stochastic/index.html b/previews/PR3525/tutorials/applications/two_stage_stochastic/index.html
new file mode 100644
index 00000000000..b08ae59e133
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/two_stage_stochastic/index.html
@@ -0,0 +1,140 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Two-stage stochastic programs · JuMP</title><meta name="title" content="Two-stage stochastic programs · JuMP"/><meta property="og:title" content="Two-stage stochastic programs · JuMP"/><meta property="twitter:title" content="Two-stage stochastic programs · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox" checked/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../power_systems/">Power Systems</a></li><li><a class="tocitem" href="../optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../web_app/">Serving web apps</a></li><li class="is-active"><a class="tocitem" href>Two-stage stochastic programs</a><ul class="internal"><li><a class="tocitem" href="#Background"><span>Background</span></a></li><li><a class="tocitem" href="#Sample-Average-approximation"><span>Sample Average approximation</span></a></li><li><a class="tocitem" href="#JuMP-model"><span>JuMP model</span></a></li><li><a class="tocitem" href="#Risk-measures"><span>Risk measures</span></a></li><li><a class="tocitem" href="#Risk-averse-sample-average-approximation"><span>Risk averse sample average approximation</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Applications</a></li><li class="is-active"><a href>Two-stage stochastic programs</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Two-stage stochastic programs</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/applications/two_stage_stochastic.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Two-stage-stochastic-programs"><a class="docs-heading-anchor" href="#Two-stage-stochastic-programs">Two-stage stochastic programs</a><a id="Two-stage-stochastic-programs-1"></a><a class="docs-heading-anchor-permalink" href="#Two-stage-stochastic-programs" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../two_stage_stochastic.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to demonstrate how to model and solve a two-stage stochastic program.</p><p>This tutorial uses the following packages</p><pre><code class="language-julia hljs">using JuMP
+import Distributions
+import HiGHS
+import Plots
+import StatsPlots
+import Statistics</code></pre><h2 id="Background"><a class="docs-heading-anchor" href="#Background">Background</a><a id="Background-1"></a><a class="docs-heading-anchor-permalink" href="#Background" title="Permalink"></a></h2><p>During the week, you are a busy practitioner of Operations Research. To escape the drudgery of mathematics, you decide to open a side business selling creamy mushroom pies with puff pastry. After a few weeks, it quickly becomes apparent that operating a food business is not so easy.</p><p>The pies must be prepared in the morning, <em>before</em> you open for the day and can gauge the level of demand. If you bake too many, the unsold pies at the end of the day must be discarded and you have wasted time and money on their production. But if you bake too few, then there may be un-served customers and you could have made more money by baking more pies.</p><p>After a few weeks of poor decision making, you decide to put your knowledge of Operations Research to good use, starting with some data collection.</p><p>Each pie costs you <span>$</span>2 to make, and you sell them at <span>$</span>5 each. Disposal of an unsold pie costs <span>$</span>0.10. Based on three weeks of data collected, in which you made 200 pies each week, you sold 150, 190, and 200 pies. Thus, as a guess, you assume a triangular distribution of demand with a minimum of 150, a median of 200, and a maximum of 250.</p><p>We can model this problem by a two-stage stochastic program. In the first stage, we decide a quantity of pies to make <span>$x$</span>. We make this decision before we observe the demand <span>$d_\omega$</span>. In the second stage, we sell <span>$y_\omega$</span> pies, and incur any costs for unsold pies.</p><p>We can formulate this problem as follows:</p><p class="math-container">\[\begin{aligned}
+\max\limits_{x,y_\omega} \;\; &amp; -2x + \mathbb{E}_\omega[5y_\omega - 0.1(x - y_\omega)] \\
+  &amp; y_\omega \le x              &amp; \quad \forall \omega \in \Omega \\
+  &amp; 0 \le y_\omega \le d_\omega &amp; \quad \forall \omega \in \Omega \\
+  &amp; x \ge 0.
+\end{aligned}\]</p><h2 id="Sample-Average-approximation"><a class="docs-heading-anchor" href="#Sample-Average-approximation">Sample Average approximation</a><a id="Sample-Average-approximation-1"></a><a class="docs-heading-anchor-permalink" href="#Sample-Average-approximation" title="Permalink"></a></h2><p>If the distribution of demand is continuous, then our problem has an infinite number of variables and constraints. To form a computationally tractable problem, we instead use a finite set of samples drawn from the distribution. This is called sample average approximation (SAA).</p><pre><code class="language-julia hljs">D = Distributions.TriangularDist(150.0, 250.0, 200.0)
+N = 100
+d = sort!(rand(D, N));
+Ω = 1:N
+P = fill(1 / N, N);
+StatsPlots.histogram(d; bins = 20, label = &quot;&quot;, xlabel = &quot;Demand&quot;)</code></pre><img src="27244c58.svg" alt="Example block output"/><h2 id="JuMP-model"><a class="docs-heading-anchor" href="#JuMP-model">JuMP model</a><a id="JuMP-model-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-model" title="Permalink"></a></h2><p>The implementation of our two-stage stochastic program in JuMP is:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x &gt;= 0)
+@variable(model, 0 &lt;= y[ω in Ω] &lt;= d[ω])
+@constraint(model, [ω in Ω], y[ω] &lt;= x)
+@expression(model, z[ω in Ω], 5y[ω] - 0.1 * (x - y[ω]))
+@objective(model, Max, -2x + sum(P[ω] * z[ω] for ω in Ω))
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 5.55231e+02
+  Objective bound    : 5.55231e+02
+  Relative gap       : Inf
+  Dual objective value : 5.55231e+02
+
+* Work counters
+  Solve time (sec)   : 5.20229e-04
+  Simplex iterations : 42
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><p>The optimal number of pies to make is:</p><pre><code class="language-julia hljs">value(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">201.67200068573283</code></pre><p>The distribution of total profit is:</p><pre><code class="language-julia hljs">total_profit = [-2 * value(x) + value(z[ω]) for ω in Ω]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">100-element Vector{Float64}:
+ 382.2268618545604
+ 384.59851140069713
+ 414.79452566925033
+ 416.9310527907735
+ 421.76727530118046
+ 428.5588847547948
+ 428.9136936776533
+ 430.8190434664974
+ 435.64404827934845
+ 440.5326475948211
+   ⋮
+ 605.0160020571985
+ 605.0160020571985
+ 605.0160020571985
+ 605.0160020571985
+ 605.0160020571985
+ 605.0160020571985
+ 605.0160020571985
+ 605.0160020571985
+ 605.0160020571985</code></pre><p>Let&#39;s plot it:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    bin_distribution(x::Vector{Float64}, N::Int)
+
+A helper function that discretizes `x` into bins of width `N`.
+&quot;&quot;&quot;
+bin_distribution(x, N) = N * (floor(minimum(x) / N):ceil(maximum(x) / N))
+
+plot = StatsPlots.histogram(
+    total_profit;
+    bins = bin_distribution(total_profit, 25),
+    label = &quot;&quot;,
+    xlabel = &quot;Profit [\$]&quot;,
+    ylabel = &quot;Number of outcomes&quot;,
+)
+μ = Statistics.mean(total_profit)
+Plots.vline!(
+    plot,
+    [μ];
+    label = &quot;Expected profit (\$$(round(Int, μ)))&quot;,
+    linewidth = 3,
+)
+plot</code></pre><img src="a5d2410e.svg" alt="Example block output"/><h2 id="Risk-measures"><a class="docs-heading-anchor" href="#Risk-measures">Risk measures</a><a id="Risk-measures-1"></a><a class="docs-heading-anchor-permalink" href="#Risk-measures" title="Permalink"></a></h2><p>A risk measure is a function which maps a random variable to a real number. Common risk measures include the mean (expectation), median, mode, and maximum. We need a risk measure to convert the distribution of second stage costs into a single number that can be optimized.</p><p>Our model currently uses the expectation risk measure, but others are possible too. One popular risk measure is the conditional value at risk (CVaR).</p><p>CVaR has a parameter <span>$\gamma$</span>, and it computes the expectation of the worst <span>$\gamma$</span> fraction of outcomes.</p><p>If we are maximizing, so that small outcomes are bad, the definition of CVaR is:</p><p class="math-container">\[CVaR_{\gamma}[Z] = \max\limits_{\xi} \;\; \xi - \frac{1}{\gamma}\mathbb{E}_\omega\left[(\xi - Z)_+\right]\]</p><p>which can be formulated as the linear program:</p><p class="math-container">\[\begin{aligned}
+CVaR_{\gamma}[Z] = \max\limits_{\xi, z_\omega} \;\; &amp; \xi - \frac{1}{\gamma}\sum P_\omega z_\omega\\
+ &amp; z_\omega \ge \xi - Z_\omega &amp; \quad \forall \omega \\
+ &amp; z_\omega \ge 0 &amp; \quad \forall \omega.
+\end{aligned}\]</p><pre><code class="language-julia hljs">function CVaR(Z::Vector{Float64}, P::Vector{Float64}; γ::Float64)
+    @assert 0 &lt; γ &lt;= 1
+    N = length(Z)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, ξ)
+    @variable(model, z[1:N] &gt;= 0)
+    @constraint(model, [i in 1:N], z[i] &gt;= ξ - Z[i])
+    @objective(model, Max, ξ - 1 / γ * sum(P[i] * z[i] for i in 1:N))
+    optimize!(model)
+    return objective_value(model)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">CVaR (generic function with 1 method)</code></pre><p>When <code>γ</code> is <code>1.0</code>, we compute the mean of the profit:</p><pre><code class="language-julia hljs">cvar_10 = CVaR(total_profit, P; γ = 1.0)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">555.2312810525048</code></pre><pre><code class="language-julia hljs">Statistics.mean(total_profit)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">555.2312810525043</code></pre><p>As <code>γ</code> approaches <code>0.0</code>, we compute the worst-case (minimum) profit:</p><pre><code class="language-julia hljs">cvar_00 = CVaR(total_profit, P; γ = 0.0001)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">382.2268618545604</code></pre><pre><code class="language-julia hljs">minimum(total_profit)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">382.2268618545604</code></pre><p>By varying <code>γ</code> between <code>0</code> and <code>1</code> we can compute some trade-off of these two extremes:</p><pre><code class="language-julia hljs">cvar_05 = CVaR(total_profit, P; γ = 0.5)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">507.4811677603447</code></pre><p>Let&#39;s plot these outcomes on our distribution:</p><pre><code class="language-julia hljs">plot = StatsPlots.histogram(
+    total_profit;
+    bins = bin_distribution(total_profit, 25),
+    label = &quot;&quot;,
+    xlabel = &quot;Profit [\$]&quot;,
+    ylabel = &quot;Number of outcomes&quot;,
+)
+Plots.vline!(
+    plot,
+    [cvar_10 cvar_05 cvar_00];
+    label = [&quot;γ = 1.0&quot; &quot;γ = 0.5&quot; &quot;γ = 0.0&quot;],
+    linewidth = 3,
+)
+plot</code></pre><img src="ddf1a4d6.svg" alt="Example block output"/><h2 id="Risk-averse-sample-average-approximation"><a class="docs-heading-anchor" href="#Risk-averse-sample-average-approximation">Risk averse sample average approximation</a><a id="Risk-averse-sample-average-approximation-1"></a><a class="docs-heading-anchor-permalink" href="#Risk-averse-sample-average-approximation" title="Permalink"></a></h2><p>Because CVaR can be formulated as a linear program, we can form a risk averse sample average approximation model by combining the two formulations:</p><pre><code class="language-julia hljs">γ = 0.4
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x &gt;= 0)
+@variable(model, 0 &lt;= y[ω in Ω] &lt;= d[ω])
+@constraint(model, [ω in Ω], y[ω] &lt;= x)
+@expression(model, Z[ω in Ω], 5 * y[ω] - 0.1(x - y[ω]))
+@variable(model, ξ)
+@variable(model, z[ω in Ω] &gt;= 0)
+@constraint(model, [ω in Ω], z[ω] &gt;= ξ - Z[ω])
+@objective(model, Max, -2x + ξ - 1 / γ * sum(P[ω] * z[ω] for ω in Ω))
+optimize!(model)</code></pre><p>When <span>$\gamma = 0.4$</span>, the optimal number of pies to bake is:</p><pre><code class="language-julia hljs">value(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">183.15576067664117</code></pre><p>The distribution of total profit is:</p><pre><code class="language-julia hljs">risk_averse_total_profit = [value(-2x + Z[ω]) for ω in Ω]
+bins = bin_distribution([total_profit; risk_averse_total_profit], 25)
+plot = StatsPlots.histogram(total_profit; label = &quot;Expectation&quot;, bins = bins)
+StatsPlots.histogram!(
+    plot,
+    risk_averse_total_profit;
+    label = &quot;CV@R&quot;,
+    bins = bins,
+    alpha = 0.5,
+)
+plot</code></pre><img src="01f8597b.svg" alt="Example block output"/><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><ul><li>Try solving this problem for different numbers of samples and different distributions.</li><li>Refactor the example to avoid hard-coding the costs. What happens to the solution if the cost of disposing unsold pies increases?</li><li>Plot the optimal number of pies to make for different values of the risk aversion parameter <span>$\gamma$</span>. What is the relationship?</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../web_app/">« Serving web apps</a><a class="docs-footer-nextpage" href="../../../manual/models/">Models »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/applications/web_app.jl b/previews/PR3525/tutorials/applications/web_app.jl
new file mode 100644
index 00000000000..7481899871e
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/web_app.jl
@@ -0,0 +1,165 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Serving web apps
+
+# This tutorial demonstrates how to setup and serve JuMP models via a REST API.
+
+# In the example app we are building, we solve a trivial mixed-integer program,
+# which is parameterized by the lower bound of a variable. To call the service,
+# users send an HTTP POST request with JSON contents indicating the lower bound.
+# The returned value is the solution of the mixed-integer program as JSON.
+
+# First, we need JuMP and a solver:
+
+using JuMP
+import HiGHS
+
+# We also need [HTTP.jl](https://github.com/JuliaWeb/HTTP.jl) to act as our
+# REST server, and [JSON.jl](https://github.com/JuliaIO/JSON.jl) to marshal
+# data.
+
+import HTTP
+import JSON
+
+# ## The server side
+
+# The core components of our REST server are _endpoints_. These are functions
+# which accept a `Dict{String,Any}` of input parameters, and return a
+# `Dict{String,Any}` as output. The types are `Dict{String,Any}` because we're
+# going to read these to and from JSON.
+
+# Here's a very simple endpoint: it accepts `params` as input, formulates and
+# solves a trivial mixed-integer program, and then returns a dictionary with the
+# result.
+
+function endpoint_solve(params::Dict{String,Any})
+    if !haskey(params, "lower_bound")
+        return Dict{String,Any}(
+            "status" => "failure",
+            "reason" => "missing lower_bound param",
+        )
+    elseif !(params["lower_bound"] isa Real)
+        return Dict{String,Any}(
+            "status" => "failure",
+            "reason" => "lower_bound is not a number",
+        )
+    end
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, x >= params["lower_bound"], Int)
+    optimize!(model)
+    ret = Dict{String,Any}(
+        "status" => "okay",
+        "terminaton_status" => termination_status(model),
+        "primal_status" => primal_status(model),
+    )
+    ## Only include the `x` key if it has a value.
+    if has_values(model)
+        ret["x"] = value(x)
+    end
+    return ret
+end
+
+# When we call this, we get:
+
+endpoint_solve(Dict{String,Any}("lower_bound" => 1.2))
+
+#-
+
+endpoint_solve(Dict{String,Any}())
+
+# For a second function, we need a function that accepts an `HTTP.Request`
+# object and returns an `HTTP.Response` object.
+
+function serve_solve(request::HTTP.Request)
+    data = JSON.parse(String(request.body))
+    solution = endpoint_solve(data)
+    return HTTP.Response(200, JSON.json(solution))
+end
+
+# Finally, we need an HTTP server. There are a variety of ways you can do this
+# in HTTP.jl. We use an explicit `Sockets.listen` so we have manual control of
+# when we shutdown the server.
+
+function setup_server(host, port)
+    server = HTTP.Sockets.listen(host, port)
+    HTTP.serve!(host, port; server = server) do request
+        try
+            ## Extend the server by adding other endpoints here.
+            if request.target == "/api/solve"
+                return serve_solve(request)
+            else
+                return HTTP.Response(404, "target $(request.target) not found")
+            end
+        catch err
+            ## Log details about the exception server-side
+            @info "Unhandled exception: $err"
+            ## Return a response to the client
+            return HTTP.Response(500, "internal error")
+        end
+    end
+    return server
+end
+
+# !!! warning
+#     HTTP.jl does not serve requests on a separate thread. Therefore, a
+#     long-running job will block the main thread, preventing concurrent users from
+#     submitting requests. To work-around this, read [HTTP.jl issue 798](https://github.com/JuliaWeb/HTTP.jl/issues/798)
+#     or watch [Building Microservices and Applications in Julia](https://www.youtube.com/watch?v=uLhXgt_gKJc&t=9543s)
+#     from JuliaCon 2020.
+
+server = setup_server(HTTP.ip"127.0.0.1", 8080)
+
+# ## The client side
+
+# Now that we have a server, we can send it requests via this function:
+
+function send_request(data::Dict; endpoint::String = "solve")
+    ret = HTTP.request(
+        "POST",
+        ## This should match the URL and endpoint we defined for our server.
+        "http://127.0.0.1:8080/api/$endpoint",
+        ["Content-Type" => "application/json"],
+        JSON.json(data),
+    )
+    if ret.status != 200
+        ## This could happen if there are time-outs, network errors, etc.
+        return Dict(
+            "status" => "failure",
+            "code" => ret.status,
+            "body" => String(ret.body),
+        )
+    end
+    return JSON.parse(String(ret.body))
+end
+
+# Let's see what happens:
+
+send_request(Dict("lower_bound" => 0))
+
+#-
+
+send_request(Dict("lower_bound" => 1.2))
+
+# If we don't send a `lower_bound`, we get:
+
+send_request(Dict("invalid_param" => 1.2))
+
+# If we don't send a `lower_bound` that is a number, we get:
+
+send_request(Dict("lower_bound" => "1.2"))
+
+# Finally, we can shutdown our HTTP server:
+
+close(server)
+
+# ## Next steps
+
+# For more complicated examples relating to HTTP servers, consult the
+# [HTTP.jl documentation](https://juliaweb.github.io/HTTP.jl/stable/).
+
+# To see how you can integrate this with a larger JuMP model, read
+# [Design patterns for larger models](@ref).
diff --git a/previews/PR3525/tutorials/applications/web_app/index.html b/previews/PR3525/tutorials/applications/web_app/index.html
new file mode 100644
index 00000000000..f27d63bf93f
--- /dev/null
+++ b/previews/PR3525/tutorials/applications/web_app/index.html
@@ -0,0 +1,91 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Serving web apps · JuMP</title><meta name="title" content="Serving web apps · JuMP"/><meta property="og:title" content="Serving web apps · JuMP"/><meta property="twitter:title" content="Serving web apps · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox" checked/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../power_systems/">Power Systems</a></li><li><a class="tocitem" href="../optimal_power_flow/">Optimal power flow</a></li><li class="is-active"><a class="tocitem" href>Serving web apps</a><ul class="internal"><li><a class="tocitem" href="#The-server-side"><span>The server side</span></a></li><li><a class="tocitem" href="#The-client-side"><span>The client side</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Applications</a></li><li class="is-active"><a href>Serving web apps</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Serving web apps</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/applications/web_app.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Serving-web-apps"><a class="docs-heading-anchor" href="#Serving-web-apps">Serving web apps</a><a id="Serving-web-apps-1"></a><a class="docs-heading-anchor-permalink" href="#Serving-web-apps" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../web_app.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial demonstrates how to setup and serve JuMP models via a REST API.</p><p>In the example app we are building, we solve a trivial mixed-integer program, which is parameterized by the lower bound of a variable. To call the service, users send an HTTP POST request with JSON contents indicating the lower bound. The returned value is the solution of the mixed-integer program as JSON.</p><p>First, we need JuMP and a solver:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS</code></pre><p>We also need <a href="https://github.com/JuliaWeb/HTTP.jl">HTTP.jl</a> to act as our REST server, and <a href="https://github.com/JuliaIO/JSON.jl">JSON.jl</a> to marshal data.</p><pre><code class="language-julia hljs">import HTTP
+import JSON</code></pre><h2 id="The-server-side"><a class="docs-heading-anchor" href="#The-server-side">The server side</a><a id="The-server-side-1"></a><a class="docs-heading-anchor-permalink" href="#The-server-side" title="Permalink"></a></h2><p>The core components of our REST server are <em>endpoints</em>. These are functions which accept a <code>Dict{String,Any}</code> of input parameters, and return a <code>Dict{String,Any}</code> as output. The types are <code>Dict{String,Any}</code> because we&#39;re going to read these to and from JSON.</p><p>Here&#39;s a very simple endpoint: it accepts <code>params</code> as input, formulates and solves a trivial mixed-integer program, and then returns a dictionary with the result.</p><pre><code class="language-julia hljs">function endpoint_solve(params::Dict{String,Any})
+    if !haskey(params, &quot;lower_bound&quot;)
+        return Dict{String,Any}(
+            &quot;status&quot; =&gt; &quot;failure&quot;,
+            &quot;reason&quot; =&gt; &quot;missing lower_bound param&quot;,
+        )
+    elseif !(params[&quot;lower_bound&quot;] isa Real)
+        return Dict{String,Any}(
+            &quot;status&quot; =&gt; &quot;failure&quot;,
+            &quot;reason&quot; =&gt; &quot;lower_bound is not a number&quot;,
+        )
+    end
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, x &gt;= params[&quot;lower_bound&quot;], Int)
+    optimize!(model)
+    ret = Dict{String,Any}(
+        &quot;status&quot; =&gt; &quot;okay&quot;,
+        &quot;terminaton_status&quot; =&gt; termination_status(model),
+        &quot;primal_status&quot; =&gt; primal_status(model),
+    )
+    # Only include the `x` key if it has a value.
+    if has_values(model)
+        ret[&quot;x&quot;] = value(x)
+    end
+    return ret
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">endpoint_solve (generic function with 1 method)</code></pre><p>When we call this, we get:</p><pre><code class="language-julia hljs">endpoint_solve(Dict{String,Any}(&quot;lower_bound&quot; =&gt; 1.2))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{String, Any} with 4 entries:
+  &quot;status&quot;            =&gt; &quot;okay&quot;
+  &quot;x&quot;                 =&gt; 2.0
+  &quot;primal_status&quot;     =&gt; FEASIBLE_POINT
+  &quot;terminaton_status&quot; =&gt; OPTIMAL</code></pre><pre><code class="language-julia hljs">endpoint_solve(Dict{String,Any}())</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{String, Any} with 2 entries:
+  &quot;status&quot; =&gt; &quot;failure&quot;
+  &quot;reason&quot; =&gt; &quot;missing lower_bound param&quot;</code></pre><p>For a second function, we need a function that accepts an <code>HTTP.Request</code> object and returns an <code>HTTP.Response</code> object.</p><pre><code class="language-julia hljs">function serve_solve(request::HTTP.Request)
+    data = JSON.parse(String(request.body))
+    solution = endpoint_solve(data)
+    return HTTP.Response(200, JSON.json(solution))
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">serve_solve (generic function with 1 method)</code></pre><p>Finally, we need an HTTP server. There are a variety of ways you can do this in HTTP.jl. We use an explicit <code>Sockets.listen</code> so we have manual control of when we shutdown the server.</p><pre><code class="language-julia hljs">function setup_server(host, port)
+    server = HTTP.Sockets.listen(host, port)
+    HTTP.serve!(host, port; server = server) do request
+        try
+            # Extend the server by adding other endpoints here.
+            if request.target == &quot;/api/solve&quot;
+                return serve_solve(request)
+            else
+                return HTTP.Response(404, &quot;target $(request.target) not found&quot;)
+            end
+        catch err
+            # Log details about the exception server-side
+            @info &quot;Unhandled exception: $err&quot;
+            # Return a response to the client
+            return HTTP.Response(500, &quot;internal error&quot;)
+        end
+    end
+    return server
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">setup_server (generic function with 1 method)</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>HTTP.jl does not serve requests on a separate thread. Therefore, a long-running job will block the main thread, preventing concurrent users from submitting requests. To work-around this, read <a href="https://github.com/JuliaWeb/HTTP.jl/issues/798">HTTP.jl issue 798</a> or watch <a href="https://www.youtube.com/watch?v=uLhXgt_gKJc&amp;t=9543s">Building Microservices and Applications in Julia</a> from JuliaCon 2020.</p></div></div><pre><code class="language-julia hljs">server = setup_server(HTTP.ip&quot;127.0.0.1&quot;, 8080)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Sockets.TCPServer(RawFD(25) active)</code></pre><h2 id="The-client-side"><a class="docs-heading-anchor" href="#The-client-side">The client side</a><a id="The-client-side-1"></a><a class="docs-heading-anchor-permalink" href="#The-client-side" title="Permalink"></a></h2><p>Now that we have a server, we can send it requests via this function:</p><pre><code class="language-julia hljs">function send_request(data::Dict; endpoint::String = &quot;solve&quot;)
+    ret = HTTP.request(
+        &quot;POST&quot;,
+        # This should match the URL and endpoint we defined for our server.
+        &quot;http://127.0.0.1:8080/api/$endpoint&quot;,
+        [&quot;Content-Type&quot; =&gt; &quot;application/json&quot;],
+        JSON.json(data),
+    )
+    if ret.status != 200
+        # This could happen if there are time-outs, network errors, etc.
+        return Dict(
+            &quot;status&quot; =&gt; &quot;failure&quot;,
+            &quot;code&quot; =&gt; ret.status,
+            &quot;body&quot; =&gt; String(ret.body),
+        )
+    end
+    return JSON.parse(String(ret.body))
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">send_request (generic function with 1 method)</code></pre><p>Let&#39;s see what happens:</p><pre><code class="language-julia hljs">send_request(Dict(&quot;lower_bound&quot; =&gt; 0))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{String, Any} with 4 entries:
+  &quot;status&quot;            =&gt; &quot;okay&quot;
+  &quot;x&quot;                 =&gt; 0.0
+  &quot;primal_status&quot;     =&gt; &quot;FEASIBLE_POINT&quot;
+  &quot;terminaton_status&quot; =&gt; &quot;OPTIMAL&quot;</code></pre><pre><code class="language-julia hljs">send_request(Dict(&quot;lower_bound&quot; =&gt; 1.2))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{String, Any} with 4 entries:
+  &quot;status&quot;            =&gt; &quot;okay&quot;
+  &quot;x&quot;                 =&gt; 2.0
+  &quot;primal_status&quot;     =&gt; &quot;FEASIBLE_POINT&quot;
+  &quot;terminaton_status&quot; =&gt; &quot;OPTIMAL&quot;</code></pre><p>If we don&#39;t send a <code>lower_bound</code>, we get:</p><pre><code class="language-julia hljs">send_request(Dict(&quot;invalid_param&quot; =&gt; 1.2))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{String, Any} with 2 entries:
+  &quot;status&quot; =&gt; &quot;failure&quot;
+  &quot;reason&quot; =&gt; &quot;missing lower_bound param&quot;</code></pre><p>If we don&#39;t send a <code>lower_bound</code> that is a number, we get:</p><pre><code class="language-julia hljs">send_request(Dict(&quot;lower_bound&quot; =&gt; &quot;1.2&quot;))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{String, Any} with 2 entries:
+  &quot;status&quot; =&gt; &quot;failure&quot;
+  &quot;reason&quot; =&gt; &quot;lower_bound is not a number&quot;</code></pre><p>Finally, we can shutdown our HTTP server:</p><pre><code class="language-julia hljs">close(server)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi"><span class="sgr36"><span class="sgr1">[ Info: </span></span>Server on 127.0.0.1:8080 closing</code></pre><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><p>For more complicated examples relating to HTTP servers, consult the <a href="https://juliaweb.github.io/HTTP.jl/stable/">HTTP.jl documentation</a>.</p><p>To see how you can integrate this with a larger JuMP model, read <a href="../../getting_started/design_patterns_for_larger_models/#Design-patterns-for-larger-models">Design patterns for larger models</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../optimal_power_flow/">« Optimal power flow</a><a class="docs-footer-nextpage" href="../two_stage_stochastic/">Two-stage stochastic programs »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/arbitrary_precision.jl b/previews/PR3525/tutorials/conic/arbitrary_precision.jl
new file mode 100644
index 00000000000..c06a3def722
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/arbitrary_precision.jl
@@ -0,0 +1,156 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Arbitrary precision arithmetic
+
+# The purpose of this tutorial is to explain how to use a solver which supports
+# arithmetic using a number type other than `Float64`.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import CDDLib
+import Clarabel
+
+# ## Higher-precision arithmetic
+
+# To create a model with a number type other than `Float64`, use [`GenericModel`](@ref)
+# with an optimizer which supports the same number type:
+
+model = GenericModel{BigFloat}(Clarabel.Optimizer{BigFloat})
+
+# The syntax for adding decision variables is the same as a normal JuMP
+# model, except that values are converted to `BigFloat`:
+
+@variable(model, -1 <= x[1:2] <= sqrt(big"2"))
+
+# Note that each `x` is now a [`GenericVariableRef{BigFloat}`](@ref), which
+# means that the value of `x` in a solution will be a `BigFloat`.
+
+# The lower and upper bounds of the decision variables are also `BigFloat`:
+
+lower_bound(x[1])
+
+#-
+
+typeof(lower_bound(x[1]))
+
+#-
+
+upper_bound(x[2])
+
+#-
+
+typeof(upper_bound(x[2]))
+
+# The syntax for adding constraints is the same as a normal JuMP model, except
+# that coefficients are converted to `BigFloat`:
+
+@constraint(model, c, x[1] == big"2" * x[2])
+
+# The function is a [`GenericAffExpr`](@ref) with `BigFloat` for the
+# coefficient and variable types;
+
+constraint = constraint_object(c)
+typeof(constraint.func)
+
+# and the set is a [`MOI.EqualTo{BigFloat}`](@ref):
+
+typeof(constraint.set)
+
+# The syntax for adding and objective is the same as a normal JuMP model, except
+# that coefficients are converted to `BigFloat`:
+
+@objective(model, Min, 3x[1]^2 + 2x[2]^2 - x[1] - big"4" * x[2])
+
+#-
+
+typeof(objective_function(model))
+
+# Here's the model we have built:
+
+print(model)
+
+# Let's solve and inspect the solution:
+
+optimize!(model)
+solution_summary(model)
+
+# The value of each decision variable is a `BigFloat`:
+
+value.(x)
+
+# as well as other solution attributes like the objective value:
+
+objective_value(model)
+
+# and dual solution:
+
+dual(c)
+
+# This problem has an analytic solution of `x = [3//7, 3//14]`. Currently, our
+# solution has an error of approximately `1e-9`:
+
+value.(x) .- [3 // 7, 3 // 14]
+
+# But by reducing the tolerances, we can obtain a more accurate solution:
+
+set_attribute(model, "tol_gap_abs", 1e-32)
+set_attribute(model, "tol_gap_rel", 1e-32)
+optimize!(model)
+value.(x) .- [3 // 7, 3 // 14]
+
+# ## Rational arithmetic
+
+# In addition to higher-precision floating point number types like `BigFloat`,
+# JuMP also supports solvers with exact rational arithmetic. One example is
+# CDDLib.jl, which supports the `Rational{BigInt}` number type:
+
+model = GenericModel{Rational{BigInt}}(CDDLib.Optimizer{Rational{BigInt}})
+
+# As before, we can create variables using rational bounds:
+
+@variable(model, 1 // 7 <= x[1:2] <= 2 // 3)
+
+#-
+
+lower_bound(x[1])
+
+#-
+
+typeof(lower_bound(x[1]))
+
+# As well as constraints:
+
+@constraint(model, c1, (2 // 1) * x[1] + x[2] <= 1)
+
+#-
+
+@constraint(model, c2, x[1] + 3x[2] <= 9 // 4)
+
+# and objective functions:
+
+@objective(model, Max, sum(x))
+
+# Here's the model we have built:
+
+print(model)
+
+# Let's solve and inspect the solution:
+
+optimize!(model)
+solution_summary(model)
+
+# The optimal values are given in exact rational arithmetic:
+
+value.(x)
+
+#-
+
+objective_value(model)
+
+#-
+
+value(c2)
diff --git a/previews/PR3525/tutorials/conic/arbitrary_precision/index.html b/previews/PR3525/tutorials/conic/arbitrary_precision/index.html
new file mode 100644
index 00000000000..5691247581b
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/arbitrary_precision/index.html
@@ -0,0 +1,80 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Arbitrary precision arithmetic · JuMP</title><meta name="title" content="Arbitrary precision arithmetic · JuMP"/><meta property="og:title" content="Arbitrary precision arithmetic · JuMP"/><meta property="twitter:title" content="Arbitrary precision arithmetic · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li class="is-active"><a class="tocitem" href>Arbitrary precision arithmetic</a><ul class="internal"><li><a class="tocitem" href="#Higher-precision-arithmetic"><span>Higher-precision arithmetic</span></a></li><li><a class="tocitem" href="#Rational-arithmetic"><span>Rational arithmetic</span></a></li></ul></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Arbitrary precision arithmetic</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Arbitrary precision arithmetic</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/arbitrary_precision.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Arbitrary-precision-arithmetic"><a class="docs-heading-anchor" href="#Arbitrary-precision-arithmetic">Arbitrary precision arithmetic</a><a id="Arbitrary-precision-arithmetic-1"></a><a class="docs-heading-anchor-permalink" href="#Arbitrary-precision-arithmetic" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../arbitrary_precision.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to explain how to use a solver which supports arithmetic using a number type other than <code>Float64</code>.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import CDDLib
+import Clarabel</code></pre><h2 id="Higher-precision-arithmetic"><a class="docs-heading-anchor" href="#Higher-precision-arithmetic">Higher-precision arithmetic</a><a id="Higher-precision-arithmetic-1"></a><a class="docs-heading-anchor-permalink" href="#Higher-precision-arithmetic" title="Permalink"></a></h2><p>To create a model with a number type other than <code>Float64</code>, use <a href="../../../api/JuMP/#JuMP.GenericModel"><code>GenericModel</code></a> with an optimizer which supports the same number type:</p><pre><code class="language-julia hljs">model = GenericModel{BigFloat}(Clarabel.Optimizer{BigFloat})</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: Clarabel</code></pre><p>The syntax for adding decision variables is the same as a normal JuMP model, except that values are converted to <code>BigFloat</code>:</p><pre><code class="language-julia hljs">@variable(model, -1 &lt;= x[1:2] &lt;= sqrt(big&quot;2&quot;))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{GenericVariableRef{BigFloat}}:
+ x[1]
+ x[2]</code></pre><p>Note that each <code>x</code> is now a <a href="../../../api/JuMP/#JuMP.GenericVariableRef"><code>GenericVariableRef{BigFloat}</code></a>, which means that the value of <code>x</code> in a solution will be a <code>BigFloat</code>.</p><p>The lower and upper bounds of the decision variables are also <code>BigFloat</code>:</p><pre><code class="language-julia hljs">lower_bound(x[1])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">-1.0</code></pre><pre><code class="language-julia hljs">typeof(lower_bound(x[1]))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">BigFloat</code></pre><pre><code class="language-julia hljs">upper_bound(x[2])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1.414213562373095048801688724209698078569671875376948073176679737990732478462102</code></pre><pre><code class="language-julia hljs">typeof(upper_bound(x[2]))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">BigFloat</code></pre><p>The syntax for adding constraints is the same as a normal JuMP model, except that coefficients are converted to <code>BigFloat</code>:</p><pre><code class="language-julia hljs">@constraint(model, c, x[1] == big&quot;2&quot; * x[2])</code></pre><p class="math-container">\[ x_{1} - 2.0 x_{2} = 0.0 \]</p><p>The function is a <a href="../../../api/JuMP/#JuMP.GenericAffExpr"><code>GenericAffExpr</code></a> with <code>BigFloat</code> for the coefficient and variable types;</p><pre><code class="language-julia hljs">constraint = constraint_object(c)
+typeof(constraint.func)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">GenericAffExpr{BigFloat, GenericVariableRef{BigFloat}}</code></pre><p>and the set is a <a href="../../../moi/reference/standard_form/#MathOptInterface.EqualTo"><code>MOI.EqualTo{BigFloat}</code></a>:</p><pre><code class="language-julia hljs">typeof(constraint.set)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">MathOptInterface.EqualTo{BigFloat}</code></pre><p>The syntax for adding and objective is the same as a normal JuMP model, except that coefficients are converted to <code>BigFloat</code>:</p><pre><code class="language-julia hljs">@objective(model, Min, 3x[1]^2 + 2x[2]^2 - x[1] - big&quot;4&quot; * x[2])</code></pre><p>$ 3.0 x<em>{1}^2 + 2.0 x</em>{2}^2 - x<em>{1} - 4.0 x</em>{2} $</p><pre><code class="language-julia hljs">typeof(objective_function(model))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">GenericQuadExpr{BigFloat, GenericVariableRef{BigFloat}}</code></pre><p>Here&#39;s the model we have built:</p><pre><code class="language-julia hljs">print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 3.0 x[1]² + 2.0 x[2]² - x[1] - 4.0 x[2]
+Subject to
+ c : x[1] - 2.0 x[2] = 0.0
+ x[1] ≥ -1.0
+ x[2] ≥ -1.0
+ x[1] ≤ 1.414213562373095048801688724209698078569671875376948073176679737990732478462102
+ x[2] ≤ 1.414213562373095048801688724209698078569671875376948073176679737990732478462102</code></pre><p>Let&#39;s solve and inspect the solution:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Clarabel
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;SOLVED&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : -6.42857e-01
+  Dual objective value : -6.42857e-01
+
+* Work counters
+  Solve time (sec)   : 2.31577e-03
+  Barrier iterations : 5
+</code></pre><p>The value of each decision variable is a <code>BigFloat</code>:</p><pre><code class="language-julia hljs">value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{BigFloat}:
+ 0.4285714246558161076147072906813123533593766450416896337912086518811186790735016
+ 0.2142857123279078924828007272730108809297577877991360649674411645247653239673974</code></pre><p>as well as other solution attributes like the objective value:</p><pre><code class="language-julia hljs">objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">-0.6428571428571422964607590389935242587959291815638830868454759876473734138856053</code></pre><p>and dual solution:</p><pre><code class="language-julia hljs">dual(c)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1.571428571977140845343978069015092190548250919787945065022059071052557047888015</code></pre><p>This problem has an analytic solution of <code>x = [3//7, 3//14]</code>. Currently, our solution has an error of approximately <code>1e-9</code>:</p><pre><code class="language-julia hljs">value.(x) .- [3 // 7, 3 // 14]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{BigFloat}:
+ -3.915612463813864137890116218069194783529738937637362776690309892355071064813317e-09
+ -1.957806393231484987012703404784527926486578220746844549760948961746888943071481e-09</code></pre><p>But by reducing the tolerances, we can obtain a more accurate solution:</p><pre><code class="language-julia hljs">set_attribute(model, &quot;tol_gap_abs&quot;, 1e-32)
+set_attribute(model, &quot;tol_gap_rel&quot;, 1e-32)
+optimize!(model)
+value.(x) .- [3 // 7, 3 // 14]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{BigFloat}:
+ -4.120732596246374574619292889406407106157605546563218305172773512099467866195165e-32
+ -7.146646610782677659152301436088423235900252780211057986251367981130623553333357e-32</code></pre><h2 id="Rational-arithmetic"><a class="docs-heading-anchor" href="#Rational-arithmetic">Rational arithmetic</a><a id="Rational-arithmetic-1"></a><a class="docs-heading-anchor-permalink" href="#Rational-arithmetic" title="Permalink"></a></h2><p>In addition to higher-precision floating point number types like <code>BigFloat</code>, JuMP also supports solvers with exact rational arithmetic. One example is CDDLib.jl, which supports the <code>Rational{BigInt}</code> number type:</p><pre><code class="language-julia hljs">model = GenericModel{Rational{BigInt}}(CDDLib.Optimizer{Rational{BigInt}})</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: CDD</code></pre><p>As before, we can create variables using rational bounds:</p><pre><code class="language-julia hljs">@variable(model, 1 // 7 &lt;= x[1:2] &lt;= 2 // 3)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{GenericVariableRef{Rational{BigInt}}}:
+ x[1]
+ x[2]</code></pre><pre><code class="language-julia hljs">lower_bound(x[1])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1//7</code></pre><pre><code class="language-julia hljs">typeof(lower_bound(x[1]))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Rational{BigInt}</code></pre><p>As well as constraints:</p><pre><code class="language-julia hljs">@constraint(model, c1, (2 // 1) * x[1] + x[2] &lt;= 1)</code></pre><p class="math-container">\[ 2//1 x_{1} + x_{2} \leq 1//1 \]</p><pre><code class="language-julia hljs">@constraint(model, c2, x[1] + 3x[2] &lt;= 9 // 4)</code></pre><p class="math-container">\[ x_{1} + 3//1 x_{2} \leq 9//4 \]</p><p>and objective functions:</p><pre><code class="language-julia hljs">@objective(model, Max, sum(x))</code></pre><p>$ x<em>{1} + x</em>{2} $</p><p>Here&#39;s the model we have built:</p><pre><code class="language-julia hljs">print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Max x[1] + x[2]
+Subject to
+ c1 : 2//1 x[1] + x[2] ≤ 1//1
+ c2 : x[1] + 3//1 x[2] ≤ 9//4
+ x[1] ≥ 1//7
+ x[2] ≥ 1//7
+ x[1] ≤ 2//3
+ x[2] ≤ 2//3</code></pre><p>Let&#39;s solve and inspect the solution:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : CDD
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;Optimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : 5//6
+
+* Work counters
+</code></pre><p>The optimal values are given in exact rational arithmetic:</p><pre><code class="language-julia hljs">value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Rational{BigInt}}:
+ 1//6
+ 2//3</code></pre><pre><code class="language-julia hljs">objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5//6</code></pre><pre><code class="language-julia hljs">value(c2)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">13//6</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../dualization/">« Dualization</a><a class="docs-footer-nextpage" href="../logistic_regression/">Logistic regression »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/dualization.jl b/previews/PR3525/tutorials/conic/dualization.jl
new file mode 100644
index 00000000000..149c2d3f5f2
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/dualization.jl
@@ -0,0 +1,255 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Dualization
+
+# The purpose of this tutorial is to explain how to use [Dualization.jl](@ref) to
+# improve the performance of some conic optimization models. There are two
+# important takeaways:
+#
+#  1. JuMP reformulates problems to meet the input requirements of the
+#     solver, potentially increasing the problem size by adding slack variables
+#     and constraints.
+#  2. Solving the dual of a conic model can be more efficient than solving the
+#     primal.
+#
+# [Dualization.jl](@ref) is a package which fixes these problems, allowing you
+# to solve the dual instead of the primal with a one-line change to your code.
+
+# This tutorial uses the following packages
+
+using JuMP
+import Dualization
+import SCS
+
+# ## Background
+
+# Conic optimization solvers typically accept one of two input formulations.
+
+# The first is the _standard_ conic form:
+# ```math
+# \begin{align}
+#     \min_{x \in \mathbb{R}^n} \; & c^\top x \\
+#               \;\;\text{s.t.} \; & A x = b  \\
+#                               & x \in \mathcal{K}
+# \end{align}
+# ```
+# in which we have a set of linear equality constraints $Ax = b$ and the
+# variables belong to a cone $\mathcal{K}$.
+#
+# The second is the _geometric_ conic form:
+# ```math
+# \begin{align}
+#     \min_{x \in \mathbb{R}^n} \; & c^\top x \\
+#               \;\;\text{s.t.} \; & A x - b \in \mathcal{K}
+# \end{align}
+# ```
+# in which an affine function $Ax - b$ belongs to a cone $\mathcal{K}$ and the
+# variables are free.
+
+# It is trivial to convert between these two representations, for example, to go
+# from the geometric conic form to the standard conic form we introduce slack
+# variables $y$:
+# ```math
+# \begin{align}
+#     \min_{x \in \mathbb{R}^n} \; & c^\top x \\
+#               \;\;\text{s.t.} \; & \begin{bmatrix}A & -I\end{bmatrix} \begin{bmatrix}x\\y\end{bmatrix} = b \\
+#                               & \begin{bmatrix}x\\y\end{bmatrix} \in \mathbb{R}^n \times \mathcal{K}
+# \end{align}
+# ```
+# and to go from the standard conic form to the geometric conic form, we can
+# rewrite the equality constraint as a function belonging to the `{0}` cone:
+# ```math
+# \begin{align}
+#     \min_{x \in \mathbb{R}^n} \; & c^\top x \\
+#               \;\;\text{s.t.} & \begin{bmatrix}A\\I\end{bmatrix} x - \begin{bmatrix}b\\0\end{bmatrix} \in \{0\} \times \mathcal{K}
+# \end{align}
+# ```
+
+# From a theoretical perspective, the two formulations are equivalent, and if
+# you implement a model in the standard conic form and pass it to a geometric
+# conic form solver (or vice versa), then JuMP will automatically reformulate
+# the problem into the correct formulation.
+
+# From a practical perspective though, the reformulations are problematic because the
+# additional slack variables and constraints can make the problem much larger
+# and therefore harder to solve.
+
+# You should also note many problems contain a mix of conic constraints and
+# variables, and so they do not neatly fall into one of the two formulations. In
+# these cases, JuMP reformulates only the variables and constraints as necessary
+# to convert the problem into the desired form.
+
+# ## Primal and dual formulations
+
+# Duality plays a large role in conic optimization. For a detailed description
+# of conic duality, see [Duality](@ref).
+
+# A useful observation is that if the primal problem is in standard conic form,
+# then the dual problem is in geometric conic form, and vice versa. Moreover, the
+# primal and dual may have a different number of variables and constraints,
+# although which one is smaller depends on the problem. Therefore, instead of
+# reformulating the problem from one form to the other, it can be more
+# efficient to solve the dual instead of the primal.
+
+# To demonstrate, we use a variation of the [Maximum cut via SDP](@ref) example.
+
+# The primal formulation (in standard conic form) is:
+
+model_primal = Model()
+@variable(model_primal, X[1:2, 1:2], PSD)
+@objective(model_primal, Max, sum([1 -1; -1 1] .* X))
+@constraint(model_primal, primal_c[i = 1:2], 1 - X[i, i] == 0)
+print(model_primal)
+
+# This problem has three scalar decision variables (the matrix `X` is symmetric),
+# two scalar equality constraints, and a constraint that `X` is positive
+# semidefinite.
+
+# The dual of `model_primal` is:
+
+model_dual = Model()
+@variable(model_dual, y[1:2])
+@objective(model_dual, Min, sum(y))
+@constraint(model_dual, dual_c, [y[1]-1 1; 1 y[2]-1] in PSDCone())
+print(model_dual)
+
+# This problem has two scalar decision variables, and a 2x2 positive
+# semidefinite matrix constraint.
+
+# !!! tip
+#     If you haven't seen conic duality before, try deriving the dual problem
+#     based on the description in [Duality](@ref). You'll need to know that the
+#     dual cone of [`PSDCone`](@ref) is the [`PSDCone`](@ref).
+
+# When we solve `model_primal` with `SCS.Optimizer`, SCS reports three variables
+# (`variables n: 3`), five rows in the constraint matrix (`constraints m: 5`),
+# and five non-zeros in the matrix (`nnz(A): 5`):
+
+set_optimizer(model_primal, SCS.Optimizer)
+optimize!(model_primal)
+
+# (There are five rows in the constraint matrix because SCS expects problems in
+# geometric conic form, and so JuMP has reformulated the `X, PSD` variable
+# constraint into the affine constraint `X .+ 0 in PSDCone()`.)
+
+# The solution we obtain is:
+
+value.(X)
+
+#-
+
+dual.(primal_c)
+
+#-
+
+objective_value(model_primal)
+
+# When we solve `model_dual` with `SCS.Optimizer`, SCS reports two variables
+# (`variables n: 2`), three rows in the constraint matrix (`constraints m: 3`),
+# and two non-zeros in the matrix (`nnz(A): 2`):
+
+set_optimizer(model_dual, SCS.Optimizer)
+optimize!(model_dual)
+
+# and the solution we obtain is:
+
+dual.(dual_c)
+
+#-
+
+value.(y)
+
+#-
+
+objective_value(model_dual)
+
+# This particular problem is small enough that it isn't meaningful to compare
+# the solve times, but in general, we should expect `model_dual` to solve faster
+# than `model_primal` because it contains fewer variables and constraints. The
+# difference is particularly noticeable on large-scale optimization problems.
+
+# ## `dual_optimizer`
+
+# Manually deriving the conic dual is difficult and error-prone. The package
+# [Dualization.jl](@ref) provides the `Dualization.dual_optimizer` meta-solver,
+# which wraps any MathOptInterface-compatible solver in an interface that
+# automatically formulates and solves the dual of an input problem.
+
+# To demonstrate, we use `Dualization.dual_optimizer` to solve `model_primal`:
+
+set_optimizer(model_primal, Dualization.dual_optimizer(SCS.Optimizer))
+optimize!(model_primal)
+
+# The performance is the same as if we solved `model_dual`, and the correct
+# solution is returned to `X`:
+
+value.(X)
+
+#-
+
+dual.(primal_c)
+
+# Moreover, if we use `dual_optimizer` on `model_dual`, then we get the same
+# performance as if we had solved `model_primal`:
+
+set_optimizer(model_dual, Dualization.dual_optimizer(SCS.Optimizer))
+optimize!(model_dual)
+
+#-
+
+dual.(dual_c)
+
+#-
+
+value.(y)
+
+# ## A mixed example
+
+# The [Maximum cut via SDP](@ref) example is nicely defined because the primal
+# is in standard conic form and the dual is in geometric conic form. However,
+# many practical models contain a mix of the two formulations. One example is
+# [The minimum distortion problem](@ref):
+
+D = [0 1 1 1; 1 0 2 2; 1 2 0 2; 1 2 2 0]
+model = Model()
+@variable(model, c²)
+@variable(model, Q[1:4, 1:4], PSD)
+@objective(model, Min, c²)
+for i in 1:4, j in (i+1):4
+    @constraint(model, D[i, j]^2 <= Q[i, i] + Q[j, j] - 2 * Q[i, j])
+    @constraint(model, Q[i, i] + Q[j, j] - 2 * Q[i, j] <= c² * D[i, j]^2)
+end
+@constraint(model, Q[1, 1] == 0)
+@constraint(model, c² >= 1)
+
+# In this formulation, the `Q` variable is of the form $x\in\mathcal{K}$, but
+# there is also a free variable, `c²`, a linear equality constraint,
+# `Q[1, 1] == 0`, and some linear inequality constraints. Rather than attempting
+# to derive the formulation that JuMP would pass to SCS and its dual, the
+# simplest solution is to try solving the problem with and without
+# `dual_optimizer` to see which formulation is most efficient.
+
+set_optimizer(model, SCS.Optimizer)
+optimize!(model)
+
+#-
+set_optimizer(model, Dualization.dual_optimizer(SCS.Optimizer))
+optimize!(model)
+
+# For this problem, SCS reports that the primal has
+# `variables n: 11, constraints m: 24` and that the dual has
+# `variables n: 14, constraints m: 24`. Therefore, we should probably use the
+# primal formulation because it has fewer variables and the same number of
+# constraints.
+
+# ## When to use `dual_optimizer`
+
+# Because it can make the problem larger or smaller, depending on the problem
+# and the choice of solver, there is no definitive rule on when you should
+# use `dual_optimizer`. However, you should try `dual_optimizer` if your conic
+# optimization problem takes a long time to solve, or if you need to repeatedly
+# solve similarly structured problems with different data. In some cases solving
+# the dual instead of the primal can make a large difference.
diff --git a/previews/PR3525/tutorials/conic/dualization/index.html b/previews/PR3525/tutorials/conic/dualization/index.html
new file mode 100644
index 00000000000..7f4c4628c05
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/dualization/index.html
@@ -0,0 +1,213 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Dualization · JuMP</title><meta name="title" content="Dualization · JuMP"/><meta property="og:title" content="Dualization · JuMP"/><meta property="twitter:title" content="Dualization · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li class="is-active"><a class="tocitem" href>Dualization</a><ul class="internal"><li><a class="tocitem" href="#Background"><span>Background</span></a></li><li><a class="tocitem" href="#Primal-and-dual-formulations"><span>Primal and dual formulations</span></a></li><li><a class="tocitem" href="#dual_optimizer"><span><code>dual_optimizer</code></span></a></li><li><a class="tocitem" href="#A-mixed-example"><span>A mixed example</span></a></li><li><a class="tocitem" href="#When-to-use-dual_optimizer"><span>When to use <code>dual_optimizer</code></span></a></li></ul></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Dualization</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Dualization</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/dualization.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Dualization"><a class="docs-heading-anchor" href="#Dualization">Dualization</a><a id="Dualization-1"></a><a class="docs-heading-anchor-permalink" href="#Dualization" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../dualization.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to explain how to use <a href="../../../packages/Dualization/#Dualization.jl">Dualization.jl</a> to improve the performance of some conic optimization models. There are two important takeaways:</p><ol><li>JuMP reformulates problems to meet the input requirements of the solver, potentially increasing the problem size by adding slack variables and constraints.</li><li>Solving the dual of a conic model can be more efficient than solving the primal.</li></ol><p><a href="../../../packages/Dualization/#Dualization.jl">Dualization.jl</a> is a package which fixes these problems, allowing you to solve the dual instead of the primal with a one-line change to your code.</p><p>This tutorial uses the following packages</p><pre><code class="language-julia hljs">using JuMP
+import Dualization
+import SCS</code></pre><h2 id="Background"><a class="docs-heading-anchor" href="#Background">Background</a><a id="Background-1"></a><a class="docs-heading-anchor-permalink" href="#Background" title="Permalink"></a></h2><p>Conic optimization solvers typically accept one of two input formulations.</p><p>The first is the <em>standard</em> conic form:</p><p class="math-container">\[\begin{align}
+    \min_{x \in \mathbb{R}^n} \; &amp; c^\top x \\
+              \;\;\text{s.t.} \; &amp; A x = b  \\
+                              &amp; x \in \mathcal{K}
+\end{align}\]</p><p>in which we have a set of linear equality constraints <span>$Ax = b$</span> and the variables belong to a cone <span>$\mathcal{K}$</span>.</p><p>The second is the <em>geometric</em> conic form:</p><p class="math-container">\[\begin{align}
+    \min_{x \in \mathbb{R}^n} \; &amp; c^\top x \\
+              \;\;\text{s.t.} \; &amp; A x - b \in \mathcal{K}
+\end{align}\]</p><p>in which an affine function <span>$Ax - b$</span> belongs to a cone <span>$\mathcal{K}$</span> and the variables are free.</p><p>It is trivial to convert between these two representations, for example, to go from the geometric conic form to the standard conic form we introduce slack variables <span>$y$</span>:</p><p class="math-container">\[\begin{align}
+    \min_{x \in \mathbb{R}^n} \; &amp; c^\top x \\
+              \;\;\text{s.t.} \; &amp; \begin{bmatrix}A &amp; -I\end{bmatrix} \begin{bmatrix}x\\y\end{bmatrix} = b \\
+                              &amp; \begin{bmatrix}x\\y\end{bmatrix} \in \mathbb{R}^n \times \mathcal{K}
+\end{align}\]</p><p>and to go from the standard conic form to the geometric conic form, we can rewrite the equality constraint as a function belonging to the <code>{0}</code> cone:</p><p class="math-container">\[\begin{align}
+    \min_{x \in \mathbb{R}^n} \; &amp; c^\top x \\
+              \;\;\text{s.t.} &amp; \begin{bmatrix}A\\I\end{bmatrix} x - \begin{bmatrix}b\\0\end{bmatrix} \in \{0\} \times \mathcal{K}
+\end{align}\]</p><p>From a theoretical perspective, the two formulations are equivalent, and if you implement a model in the standard conic form and pass it to a geometric conic form solver (or vice versa), then JuMP will automatically reformulate the problem into the correct formulation.</p><p>From a practical perspective though, the reformulations are problematic because the additional slack variables and constraints can make the problem much larger and therefore harder to solve.</p><p>You should also note many problems contain a mix of conic constraints and variables, and so they do not neatly fall into one of the two formulations. In these cases, JuMP reformulates only the variables and constraints as necessary to convert the problem into the desired form.</p><h2 id="Primal-and-dual-formulations"><a class="docs-heading-anchor" href="#Primal-and-dual-formulations">Primal and dual formulations</a><a id="Primal-and-dual-formulations-1"></a><a class="docs-heading-anchor-permalink" href="#Primal-and-dual-formulations" title="Permalink"></a></h2><p>Duality plays a large role in conic optimization. For a detailed description of conic duality, see <a href="../../../moi/background/duality/#Duality">Duality</a>.</p><p>A useful observation is that if the primal problem is in standard conic form, then the dual problem is in geometric conic form, and vice versa. Moreover, the primal and dual may have a different number of variables and constraints, although which one is smaller depends on the problem. Therefore, instead of reformulating the problem from one form to the other, it can be more efficient to solve the dual instead of the primal.</p><p>To demonstrate, we use a variation of the <a href="../simple_examples/#Maximum-cut-via-SDP">Maximum cut via SDP</a> example.</p><p>The primal formulation (in standard conic form) is:</p><pre><code class="language-julia hljs">model_primal = Model()
+@variable(model_primal, X[1:2, 1:2], PSD)
+@objective(model_primal, Max, sum([1 -1; -1 1] .* X))
+@constraint(model_primal, primal_c[i = 1:2], 1 - X[i, i] == 0)
+print(model_primal)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Max X[1,1] - 2 X[1,2] + X[2,2]
+Subject to
+ primal_c[1] : -X[1,1] = -1
+ primal_c[2] : -X[2,2] = -1
+ [X[1,1], X[1,2], X[2,2]] ∈ MathOptInterface.PositiveSemidefiniteConeTriangle(2)</code></pre><p>This problem has three scalar decision variables (the matrix <code>X</code> is symmetric), two scalar equality constraints, and a constraint that <code>X</code> is positive semidefinite.</p><p>The dual of <code>model_primal</code> is:</p><pre><code class="language-julia hljs">model_dual = Model()
+@variable(model_dual, y[1:2])
+@objective(model_dual, Min, sum(y))
+@constraint(model_dual, dual_c, [y[1]-1 1; 1 y[2]-1] in PSDCone())
+print(model_dual)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min y[1] + y[2]
+Subject to
+ dual_c : [y[1] - 1  1;
+  1         y[2] - 1] ∈ PSDCone()</code></pre><p>This problem has two scalar decision variables, and a 2x2 positive semidefinite matrix constraint.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you haven&#39;t seen conic duality before, try deriving the dual problem based on the description in <a href="../../../moi/background/duality/#Duality">Duality</a>. You&#39;ll need to know that the dual cone of <a href="../../../api/JuMP/#JuMP.PSDCone"><code>PSDCone</code></a> is the <a href="../../../api/JuMP/#JuMP.PSDCone"><code>PSDCone</code></a>.</p></div></div><p>When we solve <code>model_primal</code> with <code>SCS.Optimizer</code>, SCS reports three variables (<code>variables n: 3</code>), five rows in the constraint matrix (<code>constraints m: 5</code>), and five non-zeros in the matrix (<code>nnz(A): 5</code>):</p><pre><code class="language-julia hljs">set_optimizer(model_primal, SCS.Optimizer)
+optimize!(model_primal)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 3, constraints m: 5
+cones: 	  z: primal zero / dual free vars: 2
+	  s: psd vars: 3, ssize: 1
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 5, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 1.65e+01  1.60e-01  5.09e+01 -2.91e+01  1.00e-01  8.22e-05
+    50| 1.74e-08  2.70e-10  4.88e-08 -4.00e+00  1.00e-01  1.47e-04
+------------------------------------------------------------------
+status:  solved
+timings: total: 1.48e-04s = setup: 5.28e-05s + solve: 9.53e-05s
+	 lin-sys: 1.03e-05s, cones: 5.05e-05s, accel: 2.50e-06s
+------------------------------------------------------------------
+objective = -4.000000
+------------------------------------------------------------------</code></pre><p>(There are five rows in the constraint matrix because SCS expects problems in geometric conic form, and so JuMP has reformulated the <code>X, PSD</code> variable constraint into the affine constraint <code>X .+ 0 in PSDCone()</code>.)</p><p>The solution we obtain is:</p><pre><code class="language-julia hljs">value.(X)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  1.0  -1.0
+ -1.0   1.0</code></pre><pre><code class="language-julia hljs">dual.(primal_c)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 1.9999999997299098
+ 1.9999999997299098</code></pre><pre><code class="language-julia hljs">objective_value(model_primal)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3.9999999506359964</code></pre><p>When we solve <code>model_dual</code> with <code>SCS.Optimizer</code>, SCS reports two variables (<code>variables n: 2</code>), three rows in the constraint matrix (<code>constraints m: 3</code>), and two non-zeros in the matrix (<code>nnz(A): 2</code>):</p><pre><code class="language-julia hljs">set_optimizer(model_dual, SCS.Optimizer)
+optimize!(model_dual)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 2, constraints m: 3
+cones: 	  s: psd vars: 3, ssize: 1
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 2, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 1.23e+01  1.00e+00  2.73e+01 -9.03e+00  1.00e-01  8.07e-05
+    50| 1.13e-07  1.05e-09  3.23e-07  4.00e+00  1.00e-01  1.49e-04
+------------------------------------------------------------------
+status:  solved
+timings: total: 1.50e-04s = setup: 5.59e-05s + solve: 9.37e-05s
+	 lin-sys: 8.00e-06s, cones: 5.11e-05s, accel: 2.90e-06s
+------------------------------------------------------------------
+objective = 4.000000
+------------------------------------------------------------------</code></pre><p>and the solution we obtain is:</p><pre><code class="language-julia hljs">dual.(dual_c)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  1.0  -1.0
+ -1.0   1.0</code></pre><pre><code class="language-julia hljs">value.(y)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 2.0000001592726786
+ 2.0000001592726786</code></pre><pre><code class="language-julia hljs">objective_value(model_dual)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4.000000318545357</code></pre><p>This particular problem is small enough that it isn&#39;t meaningful to compare the solve times, but in general, we should expect <code>model_dual</code> to solve faster than <code>model_primal</code> because it contains fewer variables and constraints. The difference is particularly noticeable on large-scale optimization problems.</p><h2 id="dual_optimizer"><a class="docs-heading-anchor" href="#dual_optimizer"><code>dual_optimizer</code></a><a id="dual_optimizer-1"></a><a class="docs-heading-anchor-permalink" href="#dual_optimizer" title="Permalink"></a></h2><p>Manually deriving the conic dual is difficult and error-prone. The package <a href="../../../packages/Dualization/#Dualization.jl">Dualization.jl</a> provides the <code>Dualization.dual_optimizer</code> meta-solver, which wraps any MathOptInterface-compatible solver in an interface that automatically formulates and solves the dual of an input problem.</p><p>To demonstrate, we use <code>Dualization.dual_optimizer</code> to solve <code>model_primal</code>:</p><pre><code class="language-julia hljs">set_optimizer(model_primal, Dualization.dual_optimizer(SCS.Optimizer))
+optimize!(model_primal)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 2, constraints m: 3
+cones: 	  s: psd vars: 3, ssize: 1
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 2, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 1.23e+01  1.00e+00  2.73e+01 -9.03e+00  1.00e-01  7.67e-05
+    50| 1.13e-07  1.05e-09  3.23e-07  4.00e+00  1.00e-01  1.44e-04
+------------------------------------------------------------------
+status:  solved
+timings: total: 1.44e-04s = setup: 5.45e-05s + solve: 9.00e-05s
+	 lin-sys: 8.30e-06s, cones: 4.90e-05s, accel: 2.90e-06s
+------------------------------------------------------------------
+objective = 4.000000
+------------------------------------------------------------------</code></pre><p>The performance is the same as if we solved <code>model_dual</code>, and the correct solution is returned to <code>X</code>:</p><pre><code class="language-julia hljs">value.(X)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  1.0  -1.0
+ -1.0   1.0</code></pre><pre><code class="language-julia hljs">dual.(primal_c)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 2.0000001592726786
+ 2.0000001592726786</code></pre><p>Moreover, if we use <code>dual_optimizer</code> on <code>model_dual</code>, then we get the same performance as if we had solved <code>model_primal</code>:</p><pre><code class="language-julia hljs">set_optimizer(model_dual, Dualization.dual_optimizer(SCS.Optimizer))
+optimize!(model_dual)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 3, constraints m: 5
+cones: 	  z: primal zero / dual free vars: 2
+	  s: psd vars: 3, ssize: 1
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 5, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 1.65e+01  1.60e-01  5.09e+01 -2.91e+01  1.00e-01  8.43e-05
+    50| 1.74e-08  2.70e-10  4.88e-08 -4.00e+00  1.00e-01  1.50e-04
+------------------------------------------------------------------
+status:  solved
+timings: total: 1.51e-04s = setup: 5.65e-05s + solve: 9.46e-05s
+	 lin-sys: 1.10e-05s, cones: 5.01e-05s, accel: 2.30e-06s
+------------------------------------------------------------------
+objective = -4.000000
+------------------------------------------------------------------</code></pre><pre><code class="language-julia hljs">dual.(dual_c)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  1.0  -1.0
+ -1.0   1.0</code></pre><pre><code class="language-julia hljs">value.(y)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 1.9999999997299098
+ 1.9999999997299098</code></pre><h2 id="A-mixed-example"><a class="docs-heading-anchor" href="#A-mixed-example">A mixed example</a><a id="A-mixed-example-1"></a><a class="docs-heading-anchor-permalink" href="#A-mixed-example" title="Permalink"></a></h2><p>The <a href="../simple_examples/#Maximum-cut-via-SDP">Maximum cut via SDP</a> example is nicely defined because the primal is in standard conic form and the dual is in geometric conic form. However, many practical models contain a mix of the two formulations. One example is <a href="../simple_examples/#The-minimum-distortion-problem">The minimum distortion problem</a>:</p><pre><code class="language-julia hljs">D = [0 1 1 1; 1 0 2 2; 1 2 0 2; 1 2 2 0]
+model = Model()
+@variable(model, c²)
+@variable(model, Q[1:4, 1:4], PSD)
+@objective(model, Min, c²)
+for i in 1:4, j in (i+1):4
+    @constraint(model, D[i, j]^2 &lt;= Q[i, i] + Q[j, j] - 2 * Q[i, j])
+    @constraint(model, Q[i, i] + Q[j, j] - 2 * Q[i, j] &lt;= c² * D[i, j]^2)
+end
+@constraint(model, Q[1, 1] == 0)
+@constraint(model, c² &gt;= 1)</code></pre><p class="math-container">\[ c² \geq 1 \]</p><p>In this formulation, the <code>Q</code> variable is of the form <span>$x\in\mathcal{K}$</span>, but there is also a free variable, <code>c²</code>, a linear equality constraint, <code>Q[1, 1] == 0</code>, and some linear inequality constraints. Rather than attempting to derive the formulation that JuMP would pass to SCS and its dual, the simplest solution is to try solving the problem with and without <code>dual_optimizer</code> to see which formulation is most efficient.</p><pre><code class="language-julia hljs">set_optimizer(model, SCS.Optimizer)
+optimize!(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 11, constraints m: 24
+cones: 	  z: primal zero / dual free vars: 1
+	  l: linear vars: 13
+	  s: psd vars: 10, ssize: 1
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 54, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 4.73e+00  1.00e+00  2.92e+00  1.23e+00  1.00e-01  1.29e-04
+   150| 1.01e-04  3.07e-05  6.08e-05  1.33e+00  1.00e-01  7.16e-04
+------------------------------------------------------------------
+status:  solved
+timings: total: 7.18e-04s = setup: 8.99e-05s + solve: 6.28e-04s
+	 lin-sys: 1.11e-04s, cones: 3.44e-04s, accel: 9.47e-05s
+------------------------------------------------------------------
+objective = 1.333363
+------------------------------------------------------------------</code></pre><pre><code class="language-julia hljs">set_optimizer(model, Dualization.dual_optimizer(SCS.Optimizer))
+optimize!(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 14, constraints m: 24
+cones: 	  z: primal zero / dual free vars: 1
+	  l: linear vars: 13
+	  s: psd vars: 10, ssize: 1
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 57, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 3.71e+01  1.48e+00  2.23e+02 -1.13e+02  1.00e-01  1.10e-04
+   150| 1.57e-04  2.28e-05  2.08e-04 -1.33e+00  1.00e-01  6.58e-04
+------------------------------------------------------------------
+status:  solved
+timings: total: 6.59e-04s = setup: 7.98e-05s + solve: 5.79e-04s
+	 lin-sys: 9.93e-05s, cones: 3.64e-04s, accel: 3.58e-05s
+------------------------------------------------------------------
+objective = -1.333460
+------------------------------------------------------------------</code></pre><p>For this problem, SCS reports that the primal has <code>variables n: 11, constraints m: 24</code> and that the dual has <code>variables n: 14, constraints m: 24</code>. Therefore, we should probably use the primal formulation because it has fewer variables and the same number of constraints.</p><h2 id="When-to-use-dual_optimizer"><a class="docs-heading-anchor" href="#When-to-use-dual_optimizer">When to use <code>dual_optimizer</code></a><a id="When-to-use-dual_optimizer-1"></a><a class="docs-heading-anchor-permalink" href="#When-to-use-dual_optimizer" title="Permalink"></a></h2><p>Because it can make the problem larger or smaller, depending on the problem and the choice of solver, there is no definitive rule on when you should use <code>dual_optimizer</code>. However, you should try <code>dual_optimizer</code> if your conic optimization problem takes a long time to solve, or if you need to repeatedly solve similarly structured problems with different data. In some cases solving the dual instead of the primal can make a large difference.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../simple_examples/">« Simple semidefinite programming examples</a><a class="docs-footer-nextpage" href="../arbitrary_precision/">Arbitrary precision arithmetic »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/ellipse_approx.jl b/previews/PR3525/tutorials/conic/ellipse_approx.jl
new file mode 100644
index 00000000000..6b6f1a2f2ed
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/ellipse_approx.jl
@@ -0,0 +1,262 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Ellipsoid approximation
+
+# This tutorial considers the problem of computing _extremal ellipsoids_:
+# finding ellipsoids that best approximate a given set. As an extension, we show
+# how to use JuMP to inspect the bridges that were used, and how to explore
+# alternative formulations.
+
+# The model comes from Section 4.9 "Applications VII: extremal ellipsoids"
+# of the book *Lectures on Modern Convex Optimization* by
+# [Ben-Tal and Nemirovski (2001)](http://epubs.siam.org/doi/book/10.1137/1.9780898718829).
+
+# For a related example, see also the [Minimal ellipses](@ref) tutorial.
+
+# ## Problem formulation
+
+# Suppose that we are given a set ``\mathcal{S}`` consisting of ``m`` points in
+# ``n``-dimensional space:
+# ```math
+# \mathcal{S} = \{ x_1, \ldots, x_m \} \subset \mathbb{R}^n
+# ```
+# Our goal is to determine an optimal vector ``c \in  \mathbb{R}^n`` and
+# an optimal ``n \times n`` real symmetric matrix ``D`` such that the ellipse:
+# ```math
+# E(D, c) = \{ x : (x - c)^\top D ( x - c) \leq 1 \},
+# ```
+# contains ``\mathcal{S}`` and has the smallest possible volume.
+
+# The optimal ``D`` and ``c`` are given by the optimization problem:
+# ```math
+# \begin{aligned}
+# \max        \quad & t                                                                    \\
+# \text{s.t.} \quad & Z \succeq 0 \\
+#                   & \begin{bmatrix} s & z^\top \\ z & Z \\ \end{bmatrix} \succeq 0 \\
+#                   & x_i^\top Z x_i - 2x_i^\top z + s \leq 1    & \quad i=1, \ldots, m \\
+#                   & t \le \sqrt[n]{\det(Z)},
+# \end{aligned}
+# ```
+# where ``D = Z_*`` and ``c = Z_*^{-1} z_*``.
+
+# ## Required packages
+
+# This tutorial uses the following packages:
+
+using JuMP
+import LinearAlgebra
+import Plots
+import Random
+import SCS
+import Test  #src
+
+# ## Data
+
+# We first need to generate some points to work with.
+
+function generate_point_cloud(
+    m;            # number of 2-dimensional points
+    a = 10,       # scaling in x direction
+    b = 2,        # scaling in y direction
+    rho = π / 6,  # rotation of points around origin
+    random_seed = 1,
+)
+    rng = Random.MersenneTwister(random_seed)
+    P = randn(rng, Float64, m, 2)
+    Phi = [a*cos(rho) a*sin(rho); -b*sin(rho) b*cos(rho)]
+    S = P * Phi
+    return S
+end
+
+# For the sake of this example, let's take ``m = 600``:
+S = generate_point_cloud(600);
+
+# We will visualise the points (and ellipse) using the Plots package:
+
+r = 1.1 * maximum(abs.(S))
+plot = Plots.scatter(
+    S[:, 1],
+    S[:, 2];
+    xlim = (-r, r),
+    ylim = (-r, r),
+    label = nothing,
+    c = :green,
+    shape = :x,
+    size = (600, 600),
+)
+
+# ## JuMP formulation
+
+# Now let's build and the JuMP model. We'll compute ``D`` and ``c`` after the
+# solve.
+
+model = Model(SCS.Optimizer)
+## We need to use a tighter tolerance for this example, otherwise the bounding
+## ellipse won't actually be bounding...
+set_attribute(model, "eps_rel", 1e-6)
+set_silent(model)
+m, n = size(S)
+@variable(model, z[1:n])
+@variable(model, Z[1:n, 1:n], PSD)
+@variable(model, s)
+@variable(model, t)
+@constraint(model, [s z'; z Z] >= 0, PSDCone())
+@constraint(
+    model,
+    [i in 1:m],
+    S[i, :]' * Z * S[i, :] - 2 * S[i, :]' * z + s <= 1,
+)
+@constraint(model, [t; vec(Z)] in MOI.RootDetConeSquare(n))
+@objective(model, Max, t)
+optimize!(model)
+Test.@test termination_status(model) == OPTIMAL    #src
+Test.@test primal_status(model) == FEASIBLE_POINT  #src
+solution_summary(model)
+
+# ## Results
+
+# After solving the model to optimality we can recover the solution in terms of
+# ``D`` and ``c``:
+
+D = value.(Z)
+
+#-
+
+c = D \ value.(z)
+
+# Finally, overlaying the solution in the plot we see the minimal volume approximating
+# ellipsoid:
+
+Test.@test isapprox(D, [0.00707 -0.0102; -0.0102173 0.0175624]; atol = 1e-2)  #src
+Test.@test isapprox(c, [-3.24802, -1.842825]; atol = 1e-2)                    #src
+
+P = sqrt(D)
+q = -P * c
+data = [tuple(P \ [cos(θ) - q[1], sin(θ) - q[2]]...) for θ in 0:0.05:(2pi+0.05)]
+Plots.plot!(plot, data; c = :crimson, label = nothing)
+
+# ## Alternative formulations
+
+# The formulation of `model` uses [`MOI.RootDetConeSquare`](@ref). However,
+# because SCS does not natively support this cone, JuMP automatically
+# reformulates the problem into an equivalent problem that SCS _does_ support.
+# You can see the reformulation that JuMP chose using [`print_active_bridges`](@ref):
+
+print_active_bridges(model)
+
+# There's a lot going on here, but the first bullet is:
+# ```raw
+# * Unsupported objective: MOI.VariableIndex
+# |  bridged by:
+# |   MOIB.Objective.FunctionizeBridge{Float64}
+# |  introduces:
+# |   * Supported objective: MOI.ScalarAffineFunction{Float64}
+# ```
+# This says that SCS does not support a `MOI.VariableIndex` objective function,
+# and that JuMP used a [`MOI.Bridges.Objective.FunctionizeBridge`](@ref) to
+# convert it into a `MOI.ScalarAffineFunction{Float64}` objective function.
+
+# We can leave JuMP to do the reformulation, or we can rewrite our model to
+# have an objective function that SCS natively supports:
+
+@objective(model, Max, 1.0 * t + 0.0);
+
+# Re-printing the active bridges:
+
+print_active_bridges(model)
+
+# we get `* Supported objective: MOI.ScalarAffineFunction{Float64}`.
+
+# We can manually implement some other reformulations to change our model to
+# something that SCS more closely supports by:
+#
+#  * Replacing the [`MOI.VectorOfVariables`](@ref) in [`MOI.PositiveSemidefiniteConeTriangle`](@ref)
+#    constraint `@variable(model, Z[1:n, 1:n], PSD)` with the
+#    [`MOI.VectorAffineFunction`](@ref) in [`MOI.PositiveSemidefiniteConeTriangle`](@ref)
+#    `@constraint(model, Z >= 0, PSDCone())`.
+#
+#  * Replacing the [`MOI.VectorOfVariables`](@ref) in [`MOI.PositiveSemidefiniteConeSquare`](@ref)
+#    constraint `[s z'; z Z] >= 0, PSDCone()` with the
+#    [`MOI.VectorAffineFunction`](@ref) in [`MOI.PositiveSemidefiniteConeTriangle`](@ref)
+#    `@constraint(model, LinearAlgebra.Symmetric([s z'; z Z]) >= 0, PSDCone())`.
+#
+#  * Replacing the [`MOI.ScalarAffineFunction`](@ref) in [`MOI.GreaterThan`](@ref)
+#    constraints with the vectorized equivalent of
+#    [`MOI.VectorAffineFunction`](@ref) in [`MOI.Nonnegatives`](@ref)
+#
+#  * Replacing the [`MOI.VectorOfVariables`](@ref) in [`MOI.RootDetConeSquare`](@ref)
+#    constraint with [`MOI.VectorAffineFunction`](@ref) in
+#    [`MOI.RootDetConeTriangle`](@ref).
+
+# Note that we still need to bridge [`MOI.PositiveSemidefiniteConeTriangle`](@ref)
+# constraints because SCS uses an internal `SCS.ScaledPSDCone` set instead.
+
+model = Model(SCS.Optimizer)
+set_attribute(model, "eps_rel", 1e-6)
+set_silent(model)
+@variable(model, z[1:n])
+@variable(model, s)
+@variable(model, t)
+## The former @variable(model, Z[1:n, 1:n], PSD)
+@variable(model, Z[1:n, 1:n], Symmetric)
+@constraint(model, Z >= 0, PSDCone())
+## The former [s z'; z Z] >= 0, PSDCone()
+@constraint(model, LinearAlgebra.Symmetric([s z'; z Z]) >= 0, PSDCone())
+## The former constraint S[i, :]' * Z * S[i, :] - 2 * S[i, :]' * z + s <= 1
+f = [1 - S[i, :]' * Z * S[i, :] + 2 * S[i, :]' * z - s for i in 1:m]
+@constraint(model, f in MOI.Nonnegatives(m))
+## The former constraint [t; vec(Z)] in MOI.RootDetConeSquare(n)
+@constraint(model, 1 * [t; triangle_vec(Z)] .+ 0 in MOI.RootDetConeTriangle(n))
+## The former @objective(model, Max, t)
+@objective(model, Max, 1 * t + 0)
+optimize!(model)
+Test.@test isapprox(D, value.(Z); atol = 1e-6)  #src
+solve_time_1 = solve_time(model)
+
+# This formulation gives the much smaller graph:
+
+print_active_bridges(model)
+
+# The last bullet shows how JuMP reformulated the [`MOI.RootDetConeTriangle`](@ref)
+# constraint by adding a mix of [`MOI.PositiveSemidefiniteConeTriangle`](@ref)
+# and [`MOI.GeometricMeanCone`](@ref) constraints.
+
+# Because SCS doesn't natively support the [`MOI.GeometricMeanCone`](@ref),
+# these constraints were further bridged using a [`MOI.Bridges.Constraint.GeoMeanToPowerBridge`](@ref)
+# to a series of [`MOI.PowerCone`](@ref) constraints.
+
+# However, there are many other ways that a [`MOI.GeometricMeanCone`](@ref) can
+# be reformulated into something that SCS supports. Let's see what happens if we
+# use [`remove_bridge`](@ref) to remove the
+# [`MOI.Bridges.Constraint.GeoMeanToPowerBridge`](@ref):
+
+remove_bridge(model, MOI.Bridges.Constraint.GeoMeanToPowerBridge)
+optimize!(model)
+
+# This time, the solve took:
+
+solve_time_2 = solve_time(model)
+
+# where previously it took
+
+solve_time_1
+
+# Why was the solve time different?
+
+print_active_bridges(model)
+
+# This time, JuMP used a [`MOI.Bridges.Constraint.GeoMeanBridge`](@ref) to
+# reformulate the constraint into a set of [`MOI.RotatedSecondOrderCone`](@ref)
+# constraints, which were further reformulated into a set of supported
+# [`MOI.SecondOrderCone`](@ref) constraints.
+
+# Since the two models are equivalent, we can conclude that for this particular
+# model, the [`MOI.SecondOrderCone`](@ref) formulation is more efficient.
+
+# In general though, the performance of a particular reformulation is problem-
+# and solver-specific. Therefore, JuMP chooses to minimize the number of bridges in
+# the default reformulation, leaving you to explore alternative formulations
+# using the tools and techniques shown in this tutorial.
diff --git a/previews/PR3525/tutorials/conic/ellipse_approx/07c030f6.svg b/previews/PR3525/tutorials/conic/ellipse_approx/07c030f6.svg
new file mode 100644
index 00000000000..df5ab89db96
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/ellipse_approx/07c030f6.svg
@@ -0,0 +1,2450 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="600" viewBox="0 0 2400 2400">
+<defs>
+  <clipPath id="clip490">
+    <rect x="0" y="0" width="2400" height="2400"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip490)" d="M0 2400 L2400 2400 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip491">
+    <rect x="480" y="240" width="1681" height="1681"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip490)" d="M178.867 2261.17 L2352.76 2261.17 L2352.76 47.2441 L178.867 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip492">
+    <rect x="178" y="47" width="2175" height="2215"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="327.236,2261.17 327.236,47.2441 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="640.095,2261.17 640.095,47.2441 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="952.953,2261.17 952.953,47.2441 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1265.81,2261.17 1265.81,47.2441 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1578.67,2261.17 1578.67,47.2441 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1891.53,2261.17 1891.53,47.2441 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2204.39,2261.17 2204.39,47.2441 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,2261.17 2352.76,2261.17 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="327.236,2261.17 327.236,2242.27 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="640.095,2261.17 640.095,2242.27 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="952.953,2261.17 952.953,2242.27 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1265.81,2261.17 1265.81,2242.27 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1578.67,2261.17 1578.67,2242.27 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1891.53,2261.17 1891.53,2242.27 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2204.39,2261.17 2204.39,2242.27 "/>
+<path clip-path="url(#clip490)" d="M281.299 2306.74 L310.975 2306.74 L310.975 2310.67 L281.299 2310.67 L281.299 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M335.234 2304.93 Q338.59 2305.65 340.465 2307.92 Q342.363 2310.19 342.363 2313.52 Q342.363 2318.64 338.845 2321.44 Q335.326 2324.24 328.845 2324.24 Q326.669 2324.24 324.354 2323.8 Q322.063 2323.38 319.609 2322.53 L319.609 2318.01 Q321.553 2319.15 323.868 2319.73 Q326.183 2320.3 328.706 2320.3 Q333.104 2320.3 335.396 2318.57 Q337.711 2316.83 337.711 2313.52 Q337.711 2310.47 335.558 2308.75 Q333.428 2307.02 329.609 2307.02 L325.581 2307.02 L325.581 2303.17 L329.794 2303.17 Q333.243 2303.17 335.072 2301.81 Q336.9 2300.42 336.9 2297.83 Q336.9 2295.17 335.002 2293.75 Q333.127 2292.32 329.609 2292.32 Q327.688 2292.32 325.488 2292.73 Q323.289 2293.15 320.651 2294.03 L320.651 2289.86 Q323.313 2289.12 325.627 2288.75 Q327.965 2288.38 330.025 2288.38 Q335.35 2288.38 338.451 2290.81 Q341.553 2293.22 341.553 2297.34 Q341.553 2300.21 339.91 2302.2 Q338.266 2304.17 335.234 2304.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M361.229 2292.09 Q357.618 2292.09 355.789 2295.65 Q353.984 2299.19 353.984 2306.32 Q353.984 2313.43 355.789 2316.99 Q357.618 2320.54 361.229 2320.54 Q364.863 2320.54 366.669 2316.99 Q368.497 2313.43 368.497 2306.32 Q368.497 2299.19 366.669 2295.65 Q364.863 2292.09 361.229 2292.09 M361.229 2288.38 Q367.039 2288.38 370.095 2292.99 Q373.173 2297.57 373.173 2306.32 Q373.173 2315.05 370.095 2319.66 Q367.039 2324.24 361.229 2324.24 Q355.419 2324.24 352.34 2319.66 Q349.285 2315.05 349.285 2306.32 Q349.285 2297.57 352.34 2292.99 Q355.419 2288.38 361.229 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M594.157 2306.74 L623.833 2306.74 L623.833 2310.67 L594.157 2310.67 L594.157 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M637.953 2319.63 L654.273 2319.63 L654.273 2323.57 L632.328 2323.57 L632.328 2319.63 Q634.99 2316.88 639.574 2312.25 Q644.18 2307.6 645.361 2306.25 Q647.606 2303.73 648.486 2301.99 Q649.388 2300.23 649.388 2298.54 Q649.388 2295.79 647.444 2294.05 Q645.523 2292.32 642.421 2292.32 Q640.222 2292.32 637.768 2293.08 Q635.338 2293.85 632.56 2295.4 L632.56 2290.67 Q635.384 2289.54 637.838 2288.96 Q640.291 2288.38 642.328 2288.38 Q647.699 2288.38 650.893 2291.07 Q654.087 2293.75 654.087 2298.24 Q654.087 2300.37 653.277 2302.29 Q652.49 2304.19 650.384 2306.79 Q649.805 2307.46 646.703 2310.67 Q643.601 2313.87 637.953 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M674.087 2292.09 Q670.476 2292.09 668.648 2295.65 Q666.842 2299.19 666.842 2306.32 Q666.842 2313.43 668.648 2316.99 Q670.476 2320.54 674.087 2320.54 Q677.722 2320.54 679.527 2316.99 Q681.356 2313.43 681.356 2306.32 Q681.356 2299.19 679.527 2295.65 Q677.722 2292.09 674.087 2292.09 M674.087 2288.38 Q679.898 2288.38 682.953 2292.99 Q686.032 2297.57 686.032 2306.32 Q686.032 2315.05 682.953 2319.66 Q679.898 2324.24 674.087 2324.24 Q668.277 2324.24 665.199 2319.66 Q662.143 2315.05 662.143 2306.32 Q662.143 2297.57 665.199 2292.99 Q668.277 2288.38 674.087 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M907.016 2306.74 L936.691 2306.74 L936.691 2310.67 L907.016 2310.67 L907.016 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M947.594 2319.63 L955.233 2319.63 L955.233 2293.27 L946.923 2294.93 L946.923 2290.67 L955.187 2289.01 L959.863 2289.01 L959.863 2319.63 L967.501 2319.63 L967.501 2323.57 L947.594 2323.57 L947.594 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M986.946 2292.09 Q983.335 2292.09 981.506 2295.65 Q979.7 2299.19 979.7 2306.32 Q979.7 2313.43 981.506 2316.99 Q983.335 2320.54 986.946 2320.54 Q990.58 2320.54 992.386 2316.99 Q994.214 2313.43 994.214 2306.32 Q994.214 2299.19 992.386 2295.65 Q990.58 2292.09 986.946 2292.09 M986.946 2288.38 Q992.756 2288.38 995.811 2292.99 Q998.89 2297.57 998.89 2306.32 Q998.89 2315.05 995.811 2319.66 Q992.756 2324.24 986.946 2324.24 Q981.136 2324.24 978.057 2319.66 Q975.001 2315.05 975.001 2306.32 Q975.001 2297.57 978.057 2292.99 Q981.136 2288.38 986.946 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M1265.81 2292.09 Q1262.2 2292.09 1260.37 2295.65 Q1258.57 2299.19 1258.57 2306.32 Q1258.57 2313.43 1260.37 2316.99 Q1262.2 2320.54 1265.81 2320.54 Q1269.45 2320.54 1271.25 2316.99 Q1273.08 2313.43 1273.08 2306.32 Q1273.08 2299.19 1271.25 2295.65 Q1269.45 2292.09 1265.81 2292.09 M1265.81 2288.38 Q1271.62 2288.38 1274.68 2292.99 Q1277.76 2297.57 1277.76 2306.32 Q1277.76 2315.05 1274.68 2319.66 Q1271.62 2324.24 1265.81 2324.24 Q1260 2324.24 1256.92 2319.66 Q1253.87 2315.05 1253.87 2306.32 Q1253.87 2297.57 1256.92 2292.99 Q1260 2288.38 1265.81 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M1553.36 2319.63 L1561 2319.63 L1561 2293.27 L1552.69 2294.93 L1552.69 2290.67 L1560.95 2289.01 L1565.63 2289.01 L1565.63 2319.63 L1573.26 2319.63 L1573.26 2323.57 L1553.36 2323.57 L1553.36 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M1592.71 2292.09 Q1589.1 2292.09 1587.27 2295.65 Q1585.46 2299.19 1585.46 2306.32 Q1585.46 2313.43 1587.27 2316.99 Q1589.1 2320.54 1592.71 2320.54 Q1596.34 2320.54 1598.15 2316.99 Q1599.98 2313.43 1599.98 2306.32 Q1599.98 2299.19 1598.15 2295.65 Q1596.34 2292.09 1592.71 2292.09 M1592.71 2288.38 Q1598.52 2288.38 1601.57 2292.99 Q1604.65 2297.57 1604.65 2306.32 Q1604.65 2315.05 1601.57 2319.66 Q1598.52 2324.24 1592.71 2324.24 Q1586.9 2324.24 1583.82 2319.66 Q1580.76 2315.05 1580.76 2306.32 Q1580.76 2297.57 1583.82 2292.99 Q1586.9 2288.38 1592.71 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M1870.3 2319.63 L1886.62 2319.63 L1886.62 2323.57 L1864.68 2323.57 L1864.68 2319.63 Q1867.34 2316.88 1871.92 2312.25 Q1876.53 2307.6 1877.71 2306.25 Q1879.95 2303.73 1880.83 2301.99 Q1881.74 2300.23 1881.74 2298.54 Q1881.74 2295.79 1879.79 2294.05 Q1877.87 2292.32 1874.77 2292.32 Q1872.57 2292.32 1870.12 2293.08 Q1867.69 2293.85 1864.91 2295.4 L1864.91 2290.67 Q1867.73 2289.54 1870.19 2288.96 Q1872.64 2288.38 1874.68 2288.38 Q1880.05 2288.38 1883.24 2291.07 Q1886.44 2293.75 1886.44 2298.24 Q1886.44 2300.37 1885.63 2302.29 Q1884.84 2304.19 1882.73 2306.79 Q1882.15 2307.46 1879.05 2310.67 Q1875.95 2313.87 1870.3 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M1906.44 2292.09 Q1902.82 2292.09 1901 2295.65 Q1899.19 2299.19 1899.19 2306.32 Q1899.19 2313.43 1901 2316.99 Q1902.82 2320.54 1906.44 2320.54 Q1910.07 2320.54 1911.88 2316.99 Q1913.7 2313.43 1913.7 2306.32 Q1913.7 2299.19 1911.88 2295.65 Q1910.07 2292.09 1906.44 2292.09 M1906.44 2288.38 Q1912.25 2288.38 1915.3 2292.99 Q1918.38 2297.57 1918.38 2306.32 Q1918.38 2315.05 1915.3 2319.66 Q1912.25 2324.24 1906.44 2324.24 Q1900.63 2324.24 1897.55 2319.66 Q1894.49 2315.05 1894.49 2306.32 Q1894.49 2297.57 1897.55 2292.99 Q1900.63 2288.38 1906.44 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M2193.23 2304.93 Q2196.59 2305.65 2198.46 2307.92 Q2200.36 2310.19 2200.36 2313.52 Q2200.36 2318.64 2196.84 2321.44 Q2193.32 2324.24 2186.84 2324.24 Q2184.66 2324.24 2182.35 2323.8 Q2180.06 2323.38 2177.6 2322.53 L2177.6 2318.01 Q2179.55 2319.15 2181.86 2319.73 Q2184.18 2320.3 2186.7 2320.3 Q2191.1 2320.3 2193.39 2318.57 Q2195.71 2316.83 2195.71 2313.52 Q2195.71 2310.47 2193.55 2308.75 Q2191.42 2307.02 2187.6 2307.02 L2183.58 2307.02 L2183.58 2303.17 L2187.79 2303.17 Q2191.24 2303.17 2193.07 2301.81 Q2194.9 2300.42 2194.9 2297.83 Q2194.9 2295.17 2193 2293.75 Q2191.12 2292.32 2187.6 2292.32 Q2185.68 2292.32 2183.48 2292.73 Q2181.28 2293.15 2178.65 2294.03 L2178.65 2289.86 Q2181.31 2289.12 2183.62 2288.75 Q2185.96 2288.38 2188.02 2288.38 Q2193.34 2288.38 2196.45 2290.81 Q2199.55 2293.22 2199.55 2297.34 Q2199.55 2300.21 2197.9 2302.2 Q2196.26 2304.17 2193.23 2304.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M2219.22 2292.09 Q2215.61 2292.09 2213.78 2295.65 Q2211.98 2299.19 2211.98 2306.32 Q2211.98 2313.43 2213.78 2316.99 Q2215.61 2320.54 2219.22 2320.54 Q2222.86 2320.54 2224.66 2316.99 Q2226.49 2313.43 2226.49 2306.32 Q2226.49 2299.19 2224.66 2295.65 Q2222.86 2292.09 2219.22 2292.09 M2219.22 2288.38 Q2225.03 2288.38 2228.09 2292.99 Q2231.17 2297.57 2231.17 2306.32 Q2231.17 2315.05 2228.09 2319.66 Q2225.03 2324.24 2219.22 2324.24 Q2213.41 2324.24 2210.34 2319.66 Q2207.28 2315.05 2207.28 2306.32 Q2207.28 2297.57 2210.34 2292.99 Q2213.41 2288.38 2219.22 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,2110.07 2352.76,2110.07 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,1791.45 2352.76,1791.45 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,1472.83 2352.76,1472.83 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,1154.21 2352.76,1154.21 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,835.586 2352.76,835.586 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,516.966 2352.76,516.966 "/>
+<polyline clip-path="url(#clip492)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,198.346 2352.76,198.346 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,2261.17 178.867,47.2441 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,2110.07 197.764,2110.07 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,1791.45 197.764,1791.45 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,1472.83 197.764,1472.83 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,1154.21 197.764,1154.21 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,835.586 197.764,835.586 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,516.966 197.764,516.966 "/>
+<polyline clip-path="url(#clip490)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,198.346 197.764,198.346 "/>
+<path clip-path="url(#clip490)" d="M50.9921 2110.52 L80.6679 2110.52 L80.6679 2114.45 L50.9921 2114.45 L50.9921 2110.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M104.927 2108.71 Q108.283 2109.43 110.158 2111.7 Q112.057 2113.97 112.057 2117.3 Q112.057 2122.42 108.538 2125.22 Q105.02 2128.02 98.5382 2128.02 Q96.3623 2128.02 94.0475 2127.58 Q91.7558 2127.16 89.3021 2126.3 L89.3021 2121.79 Q91.2465 2122.92 93.5613 2123.5 Q95.8761 2124.08 98.3993 2124.08 Q102.797 2124.08 105.089 2122.35 Q107.404 2120.61 107.404 2117.3 Q107.404 2114.24 105.251 2112.53 Q103.121 2110.8 99.3021 2110.8 L95.2743 2110.8 L95.2743 2106.95 L99.4872 2106.95 Q102.936 2106.95 104.765 2105.59 Q106.594 2104.2 106.594 2101.61 Q106.594 2098.94 104.696 2097.53 Q102.821 2096.1 99.3021 2096.1 Q97.3808 2096.1 95.1817 2096.51 Q92.9826 2096.93 90.3438 2097.81 L90.3438 2093.64 Q93.0058 2092.9 95.3206 2092.53 Q97.6585 2092.16 99.7187 2092.16 Q105.043 2092.16 108.145 2094.59 Q111.246 2097 111.246 2101.12 Q111.246 2103.99 109.603 2105.98 Q107.959 2107.95 104.927 2108.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M130.922 2095.86 Q127.311 2095.86 125.482 2099.43 Q123.677 2102.97 123.677 2110.1 Q123.677 2117.21 125.482 2120.77 Q127.311 2124.31 130.922 2124.31 Q134.556 2124.31 136.362 2120.77 Q138.191 2117.21 138.191 2110.1 Q138.191 2102.97 136.362 2099.43 Q134.556 2095.86 130.922 2095.86 M130.922 2092.16 Q136.732 2092.16 139.788 2096.77 Q142.867 2101.35 142.867 2110.1 Q142.867 2118.83 139.788 2123.43 Q136.732 2128.02 130.922 2128.02 Q125.112 2128.02 122.033 2123.43 Q118.978 2118.83 118.978 2110.1 Q118.978 2101.35 122.033 2096.77 Q125.112 2092.16 130.922 2092.16 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M50.9921 1791.9 L80.6679 1791.9 L80.6679 1795.83 L50.9921 1795.83 L50.9921 1791.9 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M94.7882 1804.79 L111.108 1804.79 L111.108 1808.73 L89.1632 1808.73 L89.1632 1804.79 Q91.8252 1802.04 96.4085 1797.41 Q101.015 1792.75 102.196 1791.41 Q104.441 1788.89 105.321 1787.15 Q106.223 1785.39 106.223 1783.7 Q106.223 1780.95 104.279 1779.21 Q102.358 1777.48 99.2558 1777.48 Q97.0567 1777.48 94.603 1778.24 Q92.1725 1779 89.3947 1780.55 L89.3947 1775.83 Q92.2188 1774.7 94.6724 1774.12 Q97.1261 1773.54 99.1632 1773.54 Q104.534 1773.54 107.728 1776.23 Q110.922 1778.91 110.922 1783.4 Q110.922 1785.53 110.112 1787.45 Q109.325 1789.35 107.219 1791.94 Q106.64 1792.61 103.538 1795.83 Q100.436 1799.03 94.7882 1804.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M130.922 1777.24 Q127.311 1777.24 125.482 1780.81 Q123.677 1784.35 123.677 1791.48 Q123.677 1798.59 125.482 1802.15 Q127.311 1805.69 130.922 1805.69 Q134.556 1805.69 136.362 1802.15 Q138.191 1798.59 138.191 1791.48 Q138.191 1784.35 136.362 1780.81 Q134.556 1777.24 130.922 1777.24 M130.922 1773.54 Q136.732 1773.54 139.788 1778.15 Q142.867 1782.73 142.867 1791.48 Q142.867 1800.21 139.788 1804.81 Q136.732 1809.4 130.922 1809.4 Q125.112 1809.4 122.033 1804.81 Q118.978 1800.21 118.978 1791.48 Q118.978 1782.73 122.033 1778.15 Q125.112 1773.54 130.922 1773.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M50.9921 1473.28 L80.6679 1473.28 L80.6679 1477.21 L50.9921 1477.21 L50.9921 1473.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M91.5706 1486.17 L99.2095 1486.17 L99.2095 1459.81 L90.8993 1461.47 L90.8993 1457.21 L99.1632 1455.55 L103.839 1455.55 L103.839 1486.17 L111.478 1486.17 L111.478 1490.11 L91.5706 1490.11 L91.5706 1486.17 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M130.922 1458.62 Q127.311 1458.62 125.482 1462.19 Q123.677 1465.73 123.677 1472.86 Q123.677 1479.97 125.482 1483.53 Q127.311 1487.07 130.922 1487.07 Q134.556 1487.07 136.362 1483.53 Q138.191 1479.97 138.191 1472.86 Q138.191 1465.73 136.362 1462.19 Q134.556 1458.62 130.922 1458.62 M130.922 1454.92 Q136.732 1454.92 139.788 1459.53 Q142.867 1464.11 142.867 1472.86 Q142.867 1481.59 139.788 1486.19 Q136.732 1490.78 130.922 1490.78 Q125.112 1490.78 122.033 1486.19 Q118.978 1481.59 118.978 1472.86 Q118.978 1464.11 122.033 1459.53 Q125.112 1454.92 130.922 1454.92 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M130.922 1140 Q127.311 1140 125.482 1143.57 Q123.677 1147.11 123.677 1154.24 Q123.677 1161.35 125.482 1164.91 Q127.311 1168.45 130.922 1168.45 Q134.556 1168.45 136.362 1164.91 Q138.191 1161.35 138.191 1154.24 Q138.191 1147.11 136.362 1143.57 Q134.556 1140 130.922 1140 M130.922 1136.3 Q136.732 1136.3 139.788 1140.91 Q142.867 1145.49 142.867 1154.24 Q142.867 1162.97 139.788 1167.57 Q136.732 1172.16 130.922 1172.16 Q125.112 1172.16 122.033 1167.57 Q118.978 1162.97 118.978 1154.24 Q118.978 1145.49 122.033 1140.91 Q125.112 1136.3 130.922 1136.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M91.5706 848.931 L99.2095 848.931 L99.2095 822.565 L90.8993 824.232 L90.8993 819.973 L99.1632 818.306 L103.839 818.306 L103.839 848.931 L111.478 848.931 L111.478 852.866 L91.5706 852.866 L91.5706 848.931 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M130.922 821.385 Q127.311 821.385 125.482 824.949 Q123.677 828.491 123.677 835.621 Q123.677 842.727 125.482 846.292 Q127.311 849.834 130.922 849.834 Q134.556 849.834 136.362 846.292 Q138.191 842.727 138.191 835.621 Q138.191 828.491 136.362 824.949 Q134.556 821.385 130.922 821.385 M130.922 817.681 Q136.732 817.681 139.788 822.287 Q142.867 826.871 142.867 835.621 Q142.867 844.348 139.788 848.954 Q136.732 853.537 130.922 853.537 Q125.112 853.537 122.033 848.954 Q118.978 844.348 118.978 835.621 Q118.978 826.871 122.033 822.287 Q125.112 817.681 130.922 817.681 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M94.7882 530.311 L111.108 530.311 L111.108 534.246 L89.1632 534.246 L89.1632 530.311 Q91.8252 527.556 96.4085 522.927 Q101.015 518.274 102.196 516.931 Q104.441 514.408 105.321 512.672 Q106.223 510.913 106.223 509.223 Q106.223 506.468 104.279 504.732 Q102.358 502.996 99.2558 502.996 Q97.0567 502.996 94.603 503.76 Q92.1725 504.524 89.3947 506.075 L89.3947 501.353 Q92.2188 500.218 94.6724 499.64 Q97.1261 499.061 99.1632 499.061 Q104.534 499.061 107.728 501.746 Q110.922 504.431 110.922 508.922 Q110.922 511.052 110.112 512.973 Q109.325 514.871 107.219 517.464 Q106.64 518.135 103.538 521.353 Q100.436 524.547 94.7882 530.311 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M130.922 502.765 Q127.311 502.765 125.482 506.329 Q123.677 509.871 123.677 517.001 Q123.677 524.107 125.482 527.672 Q127.311 531.214 130.922 531.214 Q134.556 531.214 136.362 527.672 Q138.191 524.107 138.191 517.001 Q138.191 509.871 136.362 506.329 Q134.556 502.765 130.922 502.765 M130.922 499.061 Q136.732 499.061 139.788 503.667 Q142.867 508.251 142.867 517.001 Q142.867 525.727 139.788 530.334 Q136.732 534.917 130.922 534.917 Q125.112 534.917 122.033 530.334 Q118.978 525.727 118.978 517.001 Q118.978 508.251 122.033 503.667 Q125.112 499.061 130.922 499.061 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M104.927 196.992 Q108.283 197.709 110.158 199.978 Q112.057 202.246 112.057 205.58 Q112.057 210.695 108.538 213.496 Q105.02 216.297 98.5382 216.297 Q96.3623 216.297 94.0475 215.857 Q91.7558 215.441 89.3021 214.584 L89.3021 210.07 Q91.2465 211.205 93.5613 211.783 Q95.8761 212.362 98.3993 212.362 Q102.797 212.362 105.089 210.626 Q107.404 208.89 107.404 205.58 Q107.404 202.524 105.251 200.811 Q103.121 199.075 99.3021 199.075 L95.2743 199.075 L95.2743 195.233 L99.4872 195.233 Q102.936 195.233 104.765 193.867 Q106.594 192.478 106.594 189.885 Q106.594 187.223 104.696 185.811 Q102.821 184.376 99.3021 184.376 Q97.3808 184.376 95.1817 184.793 Q92.9826 185.209 90.3438 186.089 L90.3438 181.922 Q93.0058 181.182 95.3206 180.811 Q97.6585 180.441 99.7187 180.441 Q105.043 180.441 108.145 182.872 Q111.246 185.279 111.246 189.399 Q111.246 192.27 109.603 194.26 Q107.959 196.228 104.927 196.992 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip490)" d="M130.922 184.145 Q127.311 184.145 125.482 187.709 Q123.677 191.251 123.677 198.381 Q123.677 205.487 125.482 209.052 Q127.311 212.594 130.922 212.594 Q134.556 212.594 136.362 209.052 Q138.191 205.487 138.191 198.381 Q138.191 191.251 136.362 187.709 Q134.556 184.145 130.922 184.145 M130.922 180.441 Q136.732 180.441 139.788 185.047 Q142.867 189.631 142.867 198.381 Q142.867 207.107 139.788 211.714 Q136.732 216.297 130.922 216.297 Q125.112 216.297 122.033 211.714 Q118.978 207.107 118.978 198.381 Q118.978 189.631 122.033 185.047 Q125.112 180.441 130.922 180.441 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><line clip-path="url(#clip492)" x1="1391.44" y1="1186.36" x2="1375.44" y2="1170.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1391.44" y1="1186.36" x2="1375.44" y2="1202.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1391.44" y1="1186.36" x2="1407.44" y2="1202.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1391.44" y1="1186.36" x2="1407.44" y2="1170.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1372.17" y1="1098.14" x2="1356.17" y2="1082.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1372.17" y1="1098.14" x2="1356.17" y2="1114.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1372.17" y1="1098.14" x2="1388.17" y2="1114.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1372.17" y1="1098.14" x2="1388.17" y2="1082.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.37" y1="1262.61" x2="1095.37" y2="1246.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.37" y1="1262.61" x2="1095.37" y2="1278.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.37" y1="1262.61" x2="1127.37" y2="1278.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.37" y1="1262.61" x2="1127.37" y2="1246.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1229.97" x2="1288.99" y2="1213.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1229.97" x2="1288.99" y2="1245.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1229.97" x2="1320.99" y2="1245.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1229.97" x2="1320.99" y2="1213.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1053.12" y1="1313.7" x2="1037.12" y2="1297.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1053.12" y1="1313.7" x2="1037.12" y2="1329.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1053.12" y1="1313.7" x2="1069.12" y2="1329.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1053.12" y1="1313.7" x2="1069.12" y2="1297.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1388.9" y1="1173.08" x2="1372.9" y2="1157.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1388.9" y1="1173.08" x2="1372.9" y2="1189.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1388.9" y1="1173.08" x2="1404.9" y2="1189.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1388.9" y1="1173.08" x2="1404.9" y2="1157.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1926.4" y1="856.923" x2="1910.4" y2="840.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1926.4" y1="856.923" x2="1910.4" y2="872.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1926.4" y1="856.923" x2="1942.4" y2="872.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1926.4" y1="856.923" x2="1942.4" y2="840.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="912.69" y1="1348.83" x2="896.69" y2="1332.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="912.69" y1="1348.83" x2="896.69" y2="1364.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="912.69" y1="1348.83" x2="928.69" y2="1364.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="912.69" y1="1348.83" x2="928.69" y2="1332.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.77" y1="1174.01" x2="1343.77" y2="1158.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.77" y1="1174.01" x2="1343.77" y2="1190.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.77" y1="1174.01" x2="1375.77" y2="1190.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.77" y1="1174.01" x2="1375.77" y2="1158.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="625.575" y1="1469.54" x2="609.575" y2="1453.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="625.575" y1="1469.54" x2="609.575" y2="1485.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="625.575" y1="1469.54" x2="641.575" y2="1485.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="625.575" y1="1469.54" x2="641.575" y2="1453.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.02" y1="1148.47" x2="1438.02" y2="1132.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.02" y1="1148.47" x2="1438.02" y2="1164.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.02" y1="1148.47" x2="1470.02" y2="1164.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.02" y1="1148.47" x2="1470.02" y2="1132.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.94" y1="1144.11" x2="1399.94" y2="1128.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.94" y1="1144.11" x2="1399.94" y2="1160.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.94" y1="1144.11" x2="1431.94" y2="1160.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.94" y1="1144.11" x2="1431.94" y2="1128.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.61" y1="1071.18" x2="1413.61" y2="1055.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.61" y1="1071.18" x2="1413.61" y2="1087.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.61" y1="1071.18" x2="1445.61" y2="1087.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.61" y1="1071.18" x2="1445.61" y2="1055.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.18" y1="1043.75" x2="1535.18" y2="1027.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.18" y1="1043.75" x2="1535.18" y2="1059.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.18" y1="1043.75" x2="1567.18" y2="1059.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.18" y1="1043.75" x2="1567.18" y2="1027.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1383.83" y1="1070.02" x2="1367.83" y2="1054.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1383.83" y1="1070.02" x2="1367.83" y2="1086.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1383.83" y1="1070.02" x2="1399.83" y2="1086.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1383.83" y1="1070.02" x2="1399.83" y2="1054.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.68" y1="1259.28" x2="1120.68" y2="1243.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.68" y1="1259.28" x2="1120.68" y2="1275.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.68" y1="1259.28" x2="1152.68" y2="1275.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.68" y1="1259.28" x2="1152.68" y2="1243.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.69" y1="1027.81" x2="1325.69" y2="1011.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.69" y1="1027.81" x2="1325.69" y2="1043.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.69" y1="1027.81" x2="1357.69" y2="1043.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.69" y1="1027.81" x2="1357.69" y2="1011.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1260.55" y1="1177.07" x2="1244.55" y2="1161.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1260.55" y1="1177.07" x2="1244.55" y2="1193.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1260.55" y1="1177.07" x2="1276.55" y2="1193.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1260.55" y1="1177.07" x2="1276.55" y2="1161.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.36" y1="1138.56" x2="990.356" y2="1122.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.36" y1="1138.56" x2="990.356" y2="1154.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.36" y1="1138.56" x2="1022.36" y2="1154.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.36" y1="1138.56" x2="1022.36" y2="1122.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="785.24" y1="1436.86" x2="769.24" y2="1420.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="785.24" y1="1436.86" x2="769.24" y2="1452.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="785.24" y1="1436.86" x2="801.24" y2="1452.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="785.24" y1="1436.86" x2="801.24" y2="1420.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1271.22" y1="1086.79" x2="1255.22" y2="1070.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1271.22" y1="1086.79" x2="1255.22" y2="1102.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1271.22" y1="1086.79" x2="1287.22" y2="1102.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1271.22" y1="1086.79" x2="1287.22" y2="1070.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1019.17" y1="1201.94" x2="1003.17" y2="1185.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1019.17" y1="1201.94" x2="1003.17" y2="1217.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1019.17" y1="1201.94" x2="1035.17" y2="1217.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1019.17" y1="1201.94" x2="1035.17" y2="1185.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.289" y1="1394.75" x2="761.289" y2="1378.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.289" y1="1394.75" x2="761.289" y2="1410.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.289" y1="1394.75" x2="793.289" y2="1410.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.289" y1="1394.75" x2="793.289" y2="1378.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1511.54" y1="1080.45" x2="1495.54" y2="1064.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1511.54" y1="1080.45" x2="1495.54" y2="1096.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1511.54" y1="1080.45" x2="1527.54" y2="1096.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1511.54" y1="1080.45" x2="1527.54" y2="1064.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1461.76" y1="1072.86" x2="1445.76" y2="1056.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1461.76" y1="1072.86" x2="1445.76" y2="1088.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1461.76" y1="1072.86" x2="1477.76" y2="1088.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1461.76" y1="1072.86" x2="1477.76" y2="1056.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.41" y1="1014.15" x2="1369.41" y2="998.153" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.41" y1="1014.15" x2="1369.41" y2="1030.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.41" y1="1014.15" x2="1401.41" y2="1030.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.41" y1="1014.15" x2="1401.41" y2="998.153" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.61" y1="1146.61" x2="1222.61" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.61" y1="1146.61" x2="1222.61" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.61" y1="1146.61" x2="1254.61" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.61" y1="1146.61" x2="1254.61" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1612.46" y1="913.718" x2="1596.46" y2="897.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1612.46" y1="913.718" x2="1596.46" y2="929.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1612.46" y1="913.718" x2="1628.46" y2="929.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1612.46" y1="913.718" x2="1628.46" y2="897.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.2" y1="1238.54" x2="1271.2" y2="1222.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.2" y1="1238.54" x2="1271.2" y2="1254.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.2" y1="1238.54" x2="1303.2" y2="1254.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.2" y1="1238.54" x2="1303.2" y2="1222.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.35" y1="1265.07" x2="1043.35" y2="1249.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.35" y1="1265.07" x2="1043.35" y2="1281.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.35" y1="1265.07" x2="1075.35" y2="1281.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.35" y1="1265.07" x2="1075.35" y2="1249.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="947.412" y1="1370.03" x2="931.412" y2="1354.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="947.412" y1="1370.03" x2="931.412" y2="1386.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="947.412" y1="1370.03" x2="963.412" y2="1386.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="947.412" y1="1370.03" x2="963.412" y2="1354.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1251.92" y1="1163.59" x2="1235.92" y2="1147.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1251.92" y1="1163.59" x2="1235.92" y2="1179.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1251.92" y1="1163.59" x2="1267.92" y2="1179.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1251.92" y1="1163.59" x2="1267.92" y2="1147.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.13" y1="1210.6" x2="1222.13" y2="1194.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.13" y1="1210.6" x2="1222.13" y2="1226.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.13" y1="1210.6" x2="1254.13" y2="1226.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.13" y1="1210.6" x2="1254.13" y2="1194.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.497" y1="1483.85" x2="672.497" y2="1467.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.497" y1="1483.85" x2="672.497" y2="1499.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.497" y1="1483.85" x2="704.497" y2="1499.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.497" y1="1483.85" x2="704.497" y2="1467.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1248.4" y1="1166.04" x2="1232.4" y2="1150.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1248.4" y1="1166.04" x2="1232.4" y2="1182.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1248.4" y1="1166.04" x2="1264.4" y2="1182.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1248.4" y1="1166.04" x2="1264.4" y2="1150.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.95" y1="869.353" x2="1529.95" y2="853.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.95" y1="869.353" x2="1529.95" y2="885.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.95" y1="869.353" x2="1561.95" y2="885.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.95" y1="869.353" x2="1561.95" y2="853.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.93" y1="1067.84" x2="1405.93" y2="1051.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.93" y1="1067.84" x2="1405.93" y2="1083.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.93" y1="1067.84" x2="1437.93" y2="1083.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.93" y1="1067.84" x2="1437.93" y2="1051.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.39" y1="1032.06" x2="1456.39" y2="1016.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.39" y1="1032.06" x2="1456.39" y2="1048.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.39" y1="1032.06" x2="1488.39" y2="1048.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.39" y1="1032.06" x2="1488.39" y2="1016.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.2" y1="1038.88" x2="1321.2" y2="1022.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.2" y1="1038.88" x2="1321.2" y2="1054.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.2" y1="1038.88" x2="1353.2" y2="1054.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.2" y1="1038.88" x2="1353.2" y2="1022.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.09" y1="1178.03" x2="1030.09" y2="1162.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.09" y1="1178.03" x2="1030.09" y2="1194.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.09" y1="1178.03" x2="1062.09" y2="1194.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.09" y1="1178.03" x2="1062.09" y2="1162.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.45" y1="1275.47" x2="1078.45" y2="1259.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.45" y1="1275.47" x2="1078.45" y2="1291.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.45" y1="1275.47" x2="1110.45" y2="1291.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.45" y1="1275.47" x2="1110.45" y2="1259.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="771.564" y1="1431.34" x2="755.564" y2="1415.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="771.564" y1="1431.34" x2="755.564" y2="1447.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="771.564" y1="1431.34" x2="787.564" y2="1447.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="771.564" y1="1431.34" x2="787.564" y2="1415.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="918.674" y1="1363.31" x2="902.674" y2="1347.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="918.674" y1="1363.31" x2="902.674" y2="1379.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="918.674" y1="1363.31" x2="934.674" y2="1379.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="918.674" y1="1363.31" x2="934.674" y2="1347.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.43" y1="1296.62" x2="1209.43" y2="1280.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.43" y1="1296.62" x2="1209.43" y2="1312.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.43" y1="1296.62" x2="1241.43" y2="1312.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.43" y1="1296.62" x2="1241.43" y2="1280.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.5" y1="1071.7" x2="1228.5" y2="1055.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.5" y1="1071.7" x2="1228.5" y2="1087.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.5" y1="1071.7" x2="1260.5" y2="1087.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.5" y1="1071.7" x2="1260.5" y2="1055.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.11" y1="1109.8" x2="1348.11" y2="1093.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.11" y1="1109.8" x2="1348.11" y2="1125.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.11" y1="1109.8" x2="1380.11" y2="1125.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.11" y1="1109.8" x2="1380.11" y2="1093.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1707.08" y1="826.86" x2="1691.08" y2="810.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1707.08" y1="826.86" x2="1691.08" y2="842.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1707.08" y1="826.86" x2="1723.08" y2="842.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1707.08" y1="826.86" x2="1723.08" y2="810.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.63" y1="875.887" x2="1561.63" y2="859.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.63" y1="875.887" x2="1561.63" y2="891.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.63" y1="875.887" x2="1593.63" y2="891.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.63" y1="875.887" x2="1593.63" y2="859.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1319.95" y1="1118.19" x2="1303.95" y2="1102.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1319.95" y1="1118.19" x2="1303.95" y2="1134.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1319.95" y1="1118.19" x2="1335.95" y2="1134.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1319.95" y1="1118.19" x2="1335.95" y2="1102.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="894.72" y1="1142.51" x2="878.72" y2="1126.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="894.72" y1="1142.51" x2="878.72" y2="1158.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="894.72" y1="1142.51" x2="910.72" y2="1158.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="894.72" y1="1142.51" x2="910.72" y2="1126.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.17" y1="1345.5" x2="1051.17" y2="1329.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.17" y1="1345.5" x2="1051.17" y2="1361.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.17" y1="1345.5" x2="1083.17" y2="1361.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.17" y1="1345.5" x2="1083.17" y2="1329.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1580.8" y1="990.127" x2="1564.8" y2="974.127" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1580.8" y1="990.127" x2="1564.8" y2="1006.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1580.8" y1="990.127" x2="1596.8" y2="1006.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1580.8" y1="990.127" x2="1596.8" y2="974.127" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="852.02" y1="1283.64" x2="836.02" y2="1267.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="852.02" y1="1283.64" x2="836.02" y2="1299.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="852.02" y1="1283.64" x2="868.02" y2="1299.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="852.02" y1="1283.64" x2="868.02" y2="1267.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.48" y1="962.993" x2="1589.48" y2="946.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.48" y1="962.993" x2="1589.48" y2="978.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.48" y1="962.993" x2="1621.48" y2="978.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.48" y1="962.993" x2="1621.48" y2="946.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1283.38" y1="1220.65" x2="1267.38" y2="1204.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1283.38" y1="1220.65" x2="1267.38" y2="1236.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1283.38" y1="1220.65" x2="1299.38" y2="1236.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1283.38" y1="1220.65" x2="1299.38" y2="1204.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.15" y1="1016.63" x2="1446.15" y2="1000.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.15" y1="1016.63" x2="1446.15" y2="1032.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.15" y1="1016.63" x2="1478.15" y2="1032.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.15" y1="1016.63" x2="1478.15" y2="1000.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1172.56" y1="1009.01" x2="1156.56" y2="993.009" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1172.56" y1="1009.01" x2="1156.56" y2="1025.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1172.56" y1="1009.01" x2="1188.56" y2="1025.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1172.56" y1="1009.01" x2="1188.56" y2="993.009" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.43" y1="1210.76" x2="1151.43" y2="1194.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.43" y1="1210.76" x2="1151.43" y2="1226.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.43" y1="1210.76" x2="1183.43" y2="1226.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.43" y1="1210.76" x2="1183.43" y2="1194.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.588" y1="1415.15" x2="672.588" y2="1399.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.588" y1="1415.15" x2="672.588" y2="1431.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.588" y1="1415.15" x2="704.588" y2="1431.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="688.588" y1="1415.15" x2="704.588" y2="1399.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.99" y1="1289.47" x2="1120.99" y2="1273.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.99" y1="1289.47" x2="1120.99" y2="1305.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.99" y1="1289.47" x2="1152.99" y2="1305.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1136.99" y1="1289.47" x2="1152.99" y2="1273.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="991.105" y1="1402.23" x2="975.105" y2="1386.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="991.105" y1="1402.23" x2="975.105" y2="1418.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="991.105" y1="1402.23" x2="1007.11" y2="1418.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="991.105" y1="1402.23" x2="1007.11" y2="1386.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1811.99" y1="919.23" x2="1795.99" y2="903.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1811.99" y1="919.23" x2="1795.99" y2="935.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1811.99" y1="919.23" x2="1827.99" y2="935.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1811.99" y1="919.23" x2="1827.99" y2="903.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.971" y1="1350.77" x2="829.971" y2="1334.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.971" y1="1350.77" x2="829.971" y2="1366.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.971" y1="1350.77" x2="861.971" y2="1366.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.971" y1="1350.77" x2="861.971" y2="1334.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1144.78" y1="1281.78" x2="1128.78" y2="1265.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1144.78" y1="1281.78" x2="1128.78" y2="1297.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1144.78" y1="1281.78" x2="1160.78" y2="1297.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1144.78" y1="1281.78" x2="1160.78" y2="1265.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1064.78" y1="1413.61" x2="1048.78" y2="1397.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1064.78" y1="1413.61" x2="1048.78" y2="1429.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1064.78" y1="1413.61" x2="1080.78" y2="1429.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1064.78" y1="1413.61" x2="1080.78" y2="1397.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="886.864" y1="1368.41" x2="870.864" y2="1352.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="886.864" y1="1368.41" x2="870.864" y2="1384.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="886.864" y1="1368.41" x2="902.864" y2="1384.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="886.864" y1="1368.41" x2="902.864" y2="1352.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.7" y1="1200.72" x2="1324.7" y2="1184.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.7" y1="1200.72" x2="1324.7" y2="1216.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.7" y1="1200.72" x2="1356.7" y2="1216.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.7" y1="1200.72" x2="1356.7" y2="1184.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.27" y1="1196.6" x2="1051.27" y2="1180.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.27" y1="1196.6" x2="1051.27" y2="1212.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.27" y1="1196.6" x2="1083.27" y2="1212.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.27" y1="1196.6" x2="1083.27" y2="1180.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.009" y1="1388.03" x2="761.009" y2="1372.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.009" y1="1388.03" x2="761.009" y2="1404.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.009" y1="1388.03" x2="793.009" y2="1404.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="777.009" y1="1388.03" x2="793.009" y2="1372.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.23" y1="998.156" x2="1277.23" y2="982.156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.23" y1="998.156" x2="1277.23" y2="1014.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.23" y1="998.156" x2="1309.23" y2="1014.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.23" y1="998.156" x2="1309.23" y2="982.156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="885.196" y1="1405.14" x2="869.196" y2="1389.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="885.196" y1="1405.14" x2="869.196" y2="1421.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="885.196" y1="1405.14" x2="901.196" y2="1421.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="885.196" y1="1405.14" x2="901.196" y2="1389.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1204.07" y1="1358.98" x2="1188.07" y2="1342.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1204.07" y1="1358.98" x2="1188.07" y2="1374.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1204.07" y1="1358.98" x2="1220.07" y2="1374.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1204.07" y1="1358.98" x2="1220.07" y2="1342.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1195.08" y1="1292.67" x2="1179.08" y2="1276.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1195.08" y1="1292.67" x2="1179.08" y2="1308.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1195.08" y1="1292.67" x2="1211.08" y2="1308.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1195.08" y1="1292.67" x2="1211.08" y2="1276.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.55" y1="853.956" x2="1441.55" y2="837.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.55" y1="853.956" x2="1441.55" y2="869.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.55" y1="853.956" x2="1473.55" y2="869.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.55" y1="853.956" x2="1473.55" y2="837.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.86" y1="1030.17" x2="1603.86" y2="1014.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.86" y1="1030.17" x2="1603.86" y2="1046.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.86" y1="1030.17" x2="1635.86" y2="1046.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.86" y1="1030.17" x2="1635.86" y2="1014.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="949.006" y1="1440.98" x2="933.006" y2="1424.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="949.006" y1="1440.98" x2="933.006" y2="1456.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="949.006" y1="1440.98" x2="965.006" y2="1456.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="949.006" y1="1440.98" x2="965.006" y2="1424.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.411" y1="1427.9" x2="829.411" y2="1411.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.411" y1="1427.9" x2="829.411" y2="1443.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.411" y1="1427.9" x2="861.411" y2="1443.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="845.411" y1="1427.9" x2="861.411" y2="1411.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.64" y1="1292.35" x2="1169.64" y2="1276.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.64" y1="1292.35" x2="1169.64" y2="1308.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.64" y1="1292.35" x2="1201.64" y2="1308.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.64" y1="1292.35" x2="1201.64" y2="1276.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.57" y1="868.321" x2="1530.57" y2="852.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.57" y1="868.321" x2="1530.57" y2="884.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.57" y1="868.321" x2="1562.57" y2="884.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.57" y1="868.321" x2="1562.57" y2="852.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.43" y1="1342.43" x2="834.43" y2="1326.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.43" y1="1342.43" x2="834.43" y2="1358.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.43" y1="1342.43" x2="866.43" y2="1358.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.43" y1="1342.43" x2="866.43" y2="1326.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.87" y1="1122.39" x2="1535.87" y2="1106.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.87" y1="1122.39" x2="1535.87" y2="1138.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.87" y1="1122.39" x2="1567.87" y2="1138.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1551.87" y1="1122.39" x2="1567.87" y2="1106.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.5" y1="829.178" x2="1603.5" y2="813.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.5" y1="829.178" x2="1603.5" y2="845.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.5" y1="829.178" x2="1635.5" y2="845.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1619.5" y1="829.178" x2="1635.5" y2="813.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1847.7" y1="804.547" x2="1831.7" y2="788.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1847.7" y1="804.547" x2="1831.7" y2="820.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1847.7" y1="804.547" x2="1863.7" y2="820.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1847.7" y1="804.547" x2="1863.7" y2="788.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="992.659" y1="1251.94" x2="976.659" y2="1235.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="992.659" y1="1251.94" x2="976.659" y2="1267.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="992.659" y1="1251.94" x2="1008.66" y2="1267.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="992.659" y1="1251.94" x2="1008.66" y2="1235.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.79" y1="1033.06" x2="1595.79" y2="1017.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.79" y1="1033.06" x2="1595.79" y2="1049.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.79" y1="1033.06" x2="1627.79" y2="1049.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.79" y1="1033.06" x2="1627.79" y2="1017.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1148.89" y1="1269.81" x2="1132.89" y2="1253.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1148.89" y1="1269.81" x2="1132.89" y2="1285.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1148.89" y1="1269.81" x2="1164.89" y2="1285.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1148.89" y1="1269.81" x2="1164.89" y2="1253.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1202.15" y1="1212.09" x2="1186.15" y2="1196.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1202.15" y1="1212.09" x2="1186.15" y2="1228.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1202.15" y1="1212.09" x2="1218.15" y2="1228.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1202.15" y1="1212.09" x2="1218.15" y2="1196.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1641.07" y1="860.861" x2="1625.07" y2="844.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1641.07" y1="860.861" x2="1625.07" y2="876.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1641.07" y1="860.861" x2="1657.07" y2="876.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1641.07" y1="860.861" x2="1657.07" y2="844.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.79" y1="1035.93" x2="1470.79" y2="1019.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.79" y1="1035.93" x2="1470.79" y2="1051.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.79" y1="1035.93" x2="1502.79" y2="1051.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.79" y1="1035.93" x2="1502.79" y2="1019.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1207.15" y1="1160.11" x2="1191.15" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1207.15" y1="1160.11" x2="1191.15" y2="1176.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1207.15" y1="1160.11" x2="1223.15" y2="1176.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1207.15" y1="1160.11" x2="1223.15" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1119.38" y1="1194.83" x2="1103.38" y2="1178.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1119.38" y1="1194.83" x2="1103.38" y2="1210.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1119.38" y1="1194.83" x2="1135.38" y2="1210.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1119.38" y1="1194.83" x2="1135.38" y2="1178.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1332.96" y1="1133.97" x2="1316.96" y2="1117.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1332.96" y1="1133.97" x2="1316.96" y2="1149.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1332.96" y1="1133.97" x2="1348.96" y2="1149.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1332.96" y1="1133.97" x2="1348.96" y2="1117.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1301.11" y1="987.604" x2="1285.11" y2="971.604" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1301.11" y1="987.604" x2="1285.11" y2="1003.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1301.11" y1="987.604" x2="1317.11" y2="1003.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1301.11" y1="987.604" x2="1317.11" y2="971.604" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.19" y1="1037.37" x2="1299.19" y2="1021.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.19" y1="1037.37" x2="1299.19" y2="1053.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.19" y1="1037.37" x2="1331.19" y2="1053.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.19" y1="1037.37" x2="1331.19" y2="1021.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1300.37" y1="1049.99" x2="1284.37" y2="1033.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1300.37" y1="1049.99" x2="1284.37" y2="1065.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1300.37" y1="1049.99" x2="1316.37" y2="1065.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1300.37" y1="1049.99" x2="1316.37" y2="1033.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.97" y1="1034.41" x2="1347.97" y2="1018.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.97" y1="1034.41" x2="1347.97" y2="1050.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.97" y1="1034.41" x2="1379.97" y2="1050.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.97" y1="1034.41" x2="1379.97" y2="1018.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.47" y1="1159.14" x2="1328.47" y2="1143.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.47" y1="1159.14" x2="1328.47" y2="1175.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.47" y1="1159.14" x2="1360.47" y2="1175.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.47" y1="1159.14" x2="1360.47" y2="1143.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1673.06" y1="901.944" x2="1657.06" y2="885.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1673.06" y1="901.944" x2="1657.06" y2="917.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1673.06" y1="901.944" x2="1689.06" y2="917.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1673.06" y1="901.944" x2="1689.06" y2="885.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.89" y1="961.872" x2="1595.89" y2="945.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.89" y1="961.872" x2="1595.89" y2="977.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.89" y1="961.872" x2="1627.89" y2="977.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.89" y1="961.872" x2="1627.89" y2="945.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.83" y1="1129.04" x2="1175.83" y2="1113.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.83" y1="1129.04" x2="1175.83" y2="1145.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.83" y1="1129.04" x2="1207.83" y2="1145.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.83" y1="1129.04" x2="1207.83" y2="1113.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.08" y1="1021.63" x2="1231.08" y2="1005.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.08" y1="1021.63" x2="1231.08" y2="1037.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.08" y1="1021.63" x2="1263.08" y2="1037.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.08" y1="1021.63" x2="1263.08" y2="1005.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1040.17" y1="1275.11" x2="1024.17" y2="1259.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1040.17" y1="1275.11" x2="1024.17" y2="1291.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1040.17" y1="1275.11" x2="1056.17" y2="1291.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1040.17" y1="1275.11" x2="1056.17" y2="1259.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="782.176" y1="1390.69" x2="766.176" y2="1374.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="782.176" y1="1390.69" x2="766.176" y2="1406.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="782.176" y1="1390.69" x2="798.176" y2="1406.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="782.176" y1="1390.69" x2="798.176" y2="1374.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.77" y1="921.537" x2="1659.77" y2="905.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.77" y1="921.537" x2="1659.77" y2="937.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.77" y1="921.537" x2="1691.77" y2="937.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.77" y1="921.537" x2="1691.77" y2="905.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.19" y1="1173.52" x2="1214.19" y2="1157.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.19" y1="1173.52" x2="1214.19" y2="1189.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.19" y1="1173.52" x2="1246.19" y2="1189.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.19" y1="1173.52" x2="1246.19" y2="1157.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1092.98" y1="1136.22" x2="1076.98" y2="1120.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1092.98" y1="1136.22" x2="1076.98" y2="1152.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1092.98" y1="1136.22" x2="1108.98" y2="1152.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1092.98" y1="1136.22" x2="1108.98" y2="1120.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="816.257" y1="1346.85" x2="800.257" y2="1330.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="816.257" y1="1346.85" x2="800.257" y2="1362.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="816.257" y1="1346.85" x2="832.257" y2="1362.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="816.257" y1="1346.85" x2="832.257" y2="1330.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="724.712" y1="1453.02" x2="708.712" y2="1437.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="724.712" y1="1453.02" x2="708.712" y2="1469.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="724.712" y1="1453.02" x2="740.712" y2="1469.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="724.712" y1="1453.02" x2="740.712" y2="1437.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.34" y1="968.47" x2="1659.34" y2="952.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.34" y1="968.47" x2="1659.34" y2="984.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.34" y1="968.47" x2="1691.34" y2="984.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1675.34" y1="968.47" x2="1691.34" y2="952.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.82" y1="1198.32" x2="1130.82" y2="1182.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.82" y1="1198.32" x2="1130.82" y2="1214.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.82" y1="1198.32" x2="1162.82" y2="1214.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.82" y1="1198.32" x2="1162.82" y2="1182.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.2" y1="1016.21" x2="1484.2" y2="1000.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.2" y1="1016.21" x2="1484.2" y2="1032.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.2" y1="1016.21" x2="1516.2" y2="1032.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.2" y1="1016.21" x2="1516.2" y2="1000.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.396" y1="1435.49" x2="858.396" y2="1419.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.396" y1="1435.49" x2="858.396" y2="1451.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.396" y1="1435.49" x2="890.396" y2="1451.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.396" y1="1435.49" x2="890.396" y2="1419.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.29" y1="1203.85" x2="1271.29" y2="1187.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.29" y1="1203.85" x2="1271.29" y2="1219.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.29" y1="1203.85" x2="1303.29" y2="1219.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1287.29" y1="1203.85" x2="1303.29" y2="1187.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1098.58" y1="1260.55" x2="1082.58" y2="1244.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1098.58" y1="1260.55" x2="1082.58" y2="1276.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1098.58" y1="1260.55" x2="1114.58" y2="1276.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1098.58" y1="1260.55" x2="1114.58" y2="1244.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1330.6" y1="1013.95" x2="1314.6" y2="997.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1330.6" y1="1013.95" x2="1314.6" y2="1029.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1330.6" y1="1013.95" x2="1346.6" y2="1029.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1330.6" y1="1013.95" x2="1346.6" y2="997.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.5" y1="1058.8" x2="1070.5" y2="1042.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.5" y1="1058.8" x2="1070.5" y2="1074.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.5" y1="1058.8" x2="1102.5" y2="1074.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.5" y1="1058.8" x2="1102.5" y2="1042.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1250.5" y1="1043.25" x2="1234.5" y2="1027.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1250.5" y1="1043.25" x2="1234.5" y2="1059.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1250.5" y1="1043.25" x2="1266.5" y2="1059.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1250.5" y1="1043.25" x2="1266.5" y2="1027.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1626.12" y1="969.424" x2="1610.12" y2="953.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1626.12" y1="969.424" x2="1610.12" y2="985.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1626.12" y1="969.424" x2="1642.12" y2="985.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1626.12" y1="969.424" x2="1642.12" y2="953.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.09" y1="1136.96" x2="1468.09" y2="1120.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.09" y1="1136.96" x2="1468.09" y2="1152.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.09" y1="1136.96" x2="1500.09" y2="1152.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.09" y1="1136.96" x2="1500.09" y2="1120.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="933.271" y1="1353.8" x2="917.271" y2="1337.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="933.271" y1="1353.8" x2="917.271" y2="1369.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="933.271" y1="1353.8" x2="949.271" y2="1369.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="933.271" y1="1353.8" x2="949.271" y2="1337.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1027.37" y1="1280.74" x2="1011.37" y2="1264.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1027.37" y1="1280.74" x2="1011.37" y2="1296.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1027.37" y1="1280.74" x2="1043.37" y2="1296.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1027.37" y1="1280.74" x2="1043.37" y2="1264.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1346.23" y1="1131.49" x2="1330.23" y2="1115.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1346.23" y1="1131.49" x2="1330.23" y2="1147.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1346.23" y1="1131.49" x2="1362.23" y2="1147.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1346.23" y1="1131.49" x2="1362.23" y2="1115.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="906.639" y1="1451.01" x2="890.639" y2="1435.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="906.639" y1="1451.01" x2="890.639" y2="1467.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="906.639" y1="1451.01" x2="922.639" y2="1467.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="906.639" y1="1451.01" x2="922.639" y2="1435.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.76" y1="1104.6" x2="1405.76" y2="1088.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.76" y1="1104.6" x2="1405.76" y2="1120.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.76" y1="1104.6" x2="1437.76" y2="1120.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.76" y1="1104.6" x2="1437.76" y2="1088.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="958.996" y1="1406.94" x2="942.996" y2="1390.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="958.996" y1="1406.94" x2="942.996" y2="1422.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="958.996" y1="1406.94" x2="974.996" y2="1422.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="958.996" y1="1406.94" x2="974.996" y2="1390.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.277" y1="1312.14" x2="858.277" y2="1296.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.277" y1="1312.14" x2="858.277" y2="1328.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.277" y1="1312.14" x2="890.277" y2="1328.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="874.277" y1="1312.14" x2="890.277" y2="1296.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.691" y1="1424.89" x2="911.691" y2="1408.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.691" y1="1424.89" x2="911.691" y2="1440.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.691" y1="1424.89" x2="943.691" y2="1440.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.691" y1="1424.89" x2="943.691" y2="1408.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1412.66" y1="1067.26" x2="1396.66" y2="1051.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1412.66" y1="1067.26" x2="1396.66" y2="1083.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1412.66" y1="1067.26" x2="1428.66" y2="1083.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1412.66" y1="1067.26" x2="1428.66" y2="1051.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="538.801" y1="1550.28" x2="522.801" y2="1534.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="538.801" y1="1550.28" x2="522.801" y2="1566.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="538.801" y1="1550.28" x2="554.801" y2="1566.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="538.801" y1="1550.28" x2="554.801" y2="1534.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1294.59" y1="1109.44" x2="1278.59" y2="1093.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1294.59" y1="1109.44" x2="1278.59" y2="1125.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1294.59" y1="1109.44" x2="1310.59" y2="1125.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1294.59" y1="1109.44" x2="1310.59" y2="1093.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="869.402" y1="1512" x2="853.402" y2="1496" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="869.402" y1="1512" x2="853.402" y2="1528" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="869.402" y1="1512" x2="885.402" y2="1528" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="869.402" y1="1512" x2="885.402" y2="1496" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.688" y1="1677.8" x2="553.688" y2="1661.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.688" y1="1677.8" x2="553.688" y2="1693.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.688" y1="1677.8" x2="585.688" y2="1693.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.688" y1="1677.8" x2="585.688" y2="1661.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1124.14" y1="1170.71" x2="1108.14" y2="1154.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1124.14" y1="1170.71" x2="1108.14" y2="1186.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1124.14" y1="1170.71" x2="1140.14" y2="1186.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1124.14" y1="1170.71" x2="1140.14" y2="1154.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="915.573" y1="1448.5" x2="899.573" y2="1432.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="915.573" y1="1448.5" x2="899.573" y2="1464.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="915.573" y1="1448.5" x2="931.573" y2="1464.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="915.573" y1="1448.5" x2="931.573" y2="1432.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1520.69" y1="868.39" x2="1504.69" y2="852.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1520.69" y1="868.39" x2="1504.69" y2="884.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1520.69" y1="868.39" x2="1536.69" y2="884.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1520.69" y1="868.39" x2="1536.69" y2="852.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1108.04" y1="1185.31" x2="1092.04" y2="1169.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1108.04" y1="1185.31" x2="1092.04" y2="1201.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1108.04" y1="1185.31" x2="1124.04" y2="1201.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1108.04" y1="1185.31" x2="1124.04" y2="1169.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.79" y1="1007.52" x2="1446.79" y2="991.518" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.79" y1="1007.52" x2="1446.79" y2="1023.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.79" y1="1007.52" x2="1478.79" y2="1023.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1462.79" y1="1007.52" x2="1478.79" y2="991.518" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.71" y1="986.213" x2="1427.71" y2="970.213" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.71" y1="986.213" x2="1427.71" y2="1002.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.71" y1="986.213" x2="1459.71" y2="1002.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.71" y1="986.213" x2="1459.71" y2="970.213" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="831.587" y1="1424.87" x2="815.587" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="831.587" y1="1424.87" x2="815.587" y2="1440.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="831.587" y1="1424.87" x2="847.587" y2="1440.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="831.587" y1="1424.87" x2="847.587" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.58" y1="1341.26" x2="1014.58" y2="1325.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.58" y1="1341.26" x2="1014.58" y2="1357.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.58" y1="1341.26" x2="1046.58" y2="1357.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.58" y1="1341.26" x2="1046.58" y2="1325.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.23" y1="1063.53" x2="1436.23" y2="1047.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.23" y1="1063.53" x2="1436.23" y2="1079.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.23" y1="1063.53" x2="1468.23" y2="1079.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.23" y1="1063.53" x2="1468.23" y2="1047.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.66" y1="1110.16" x2="1062.66" y2="1094.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.66" y1="1110.16" x2="1062.66" y2="1126.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.66" y1="1110.16" x2="1094.66" y2="1126.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.66" y1="1110.16" x2="1094.66" y2="1094.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.93" y1="1006.01" x2="1593.93" y2="990.013" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.93" y1="1006.01" x2="1593.93" y2="1022.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.93" y1="1006.01" x2="1625.93" y2="1022.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.93" y1="1006.01" x2="1625.93" y2="990.013" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1733.98" y1="723.632" x2="1717.98" y2="707.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1733.98" y1="723.632" x2="1717.98" y2="739.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1733.98" y1="723.632" x2="1749.98" y2="739.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1733.98" y1="723.632" x2="1749.98" y2="707.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.47" y1="1228.99" x2="1265.47" y2="1212.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.47" y1="1228.99" x2="1265.47" y2="1244.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.47" y1="1228.99" x2="1297.47" y2="1244.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.47" y1="1228.99" x2="1297.47" y2="1212.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.93" y1="1191.01" x2="1213.93" y2="1175.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.93" y1="1191.01" x2="1213.93" y2="1207.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.93" y1="1191.01" x2="1245.93" y2="1207.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.93" y1="1191.01" x2="1245.93" y2="1175.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.48" y1="1238.33" x2="1224.48" y2="1222.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.48" y1="1238.33" x2="1224.48" y2="1254.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.48" y1="1238.33" x2="1256.48" y2="1254.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.48" y1="1238.33" x2="1256.48" y2="1222.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.936" y1="1351.49" x2="834.936" y2="1335.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.936" y1="1351.49" x2="834.936" y2="1367.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.936" y1="1351.49" x2="866.936" y2="1367.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="850.936" y1="1351.49" x2="866.936" y2="1335.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.92" y1="896.324" x2="1752.92" y2="880.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.92" y1="896.324" x2="1752.92" y2="912.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.92" y1="896.324" x2="1784.92" y2="912.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.92" y1="896.324" x2="1784.92" y2="880.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.86" y1="1264.5" x2="984.861" y2="1248.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.86" y1="1264.5" x2="984.861" y2="1280.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.86" y1="1264.5" x2="1016.86" y2="1280.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.86" y1="1264.5" x2="1016.86" y2="1248.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1022.7" y1="1343.63" x2="1006.7" y2="1327.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1022.7" y1="1343.63" x2="1006.7" y2="1359.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1022.7" y1="1343.63" x2="1038.7" y2="1359.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1022.7" y1="1343.63" x2="1038.7" y2="1327.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.15" y1="1293.76" x2="1051.15" y2="1277.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.15" y1="1293.76" x2="1051.15" y2="1309.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.15" y1="1293.76" x2="1083.15" y2="1309.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1067.15" y1="1293.76" x2="1083.15" y2="1277.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1109.19" y1="1130.27" x2="1093.19" y2="1114.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1109.19" y1="1130.27" x2="1093.19" y2="1146.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1109.19" y1="1130.27" x2="1125.19" y2="1146.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1109.19" y1="1130.27" x2="1125.19" y2="1114.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.88" y1="1116.12" x2="1399.88" y2="1100.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.88" y1="1116.12" x2="1399.88" y2="1132.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.88" y1="1116.12" x2="1431.88" y2="1132.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.88" y1="1116.12" x2="1431.88" y2="1100.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.02" y1="1086.3" x2="1442.02" y2="1070.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.02" y1="1086.3" x2="1442.02" y2="1102.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.02" y1="1086.3" x2="1474.02" y2="1102.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.02" y1="1086.3" x2="1474.02" y2="1070.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1138.81" y1="1184.29" x2="1122.81" y2="1168.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1138.81" y1="1184.29" x2="1122.81" y2="1200.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1138.81" y1="1184.29" x2="1154.81" y2="1200.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1138.81" y1="1184.29" x2="1154.81" y2="1168.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.12" y1="1087.76" x2="1183.12" y2="1071.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.12" y1="1087.76" x2="1183.12" y2="1103.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.12" y1="1087.76" x2="1215.12" y2="1103.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.12" y1="1087.76" x2="1215.12" y2="1071.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.48" y1="973.842" x2="1509.48" y2="957.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.48" y1="973.842" x2="1509.48" y2="989.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.48" y1="973.842" x2="1541.48" y2="989.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.48" y1="973.842" x2="1541.48" y2="957.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.75" y1="1324.55" x2="1091.75" y2="1308.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.75" y1="1324.55" x2="1091.75" y2="1340.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.75" y1="1324.55" x2="1123.75" y2="1340.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.75" y1="1324.55" x2="1123.75" y2="1308.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="814.5" y1="1271.95" x2="798.5" y2="1255.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="814.5" y1="1271.95" x2="798.5" y2="1287.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="814.5" y1="1271.95" x2="830.5" y2="1287.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="814.5" y1="1271.95" x2="830.5" y2="1255.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1369.5" y1="1060.06" x2="1353.5" y2="1044.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1369.5" y1="1060.06" x2="1353.5" y2="1076.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1369.5" y1="1060.06" x2="1385.5" y2="1076.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1369.5" y1="1060.06" x2="1385.5" y2="1044.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.95" y1="1174.4" x2="1227.95" y2="1158.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.95" y1="1174.4" x2="1227.95" y2="1190.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.95" y1="1174.4" x2="1259.95" y2="1190.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.95" y1="1174.4" x2="1259.95" y2="1158.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1181.71" y1="1301.58" x2="1165.71" y2="1285.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1181.71" y1="1301.58" x2="1165.71" y2="1317.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1181.71" y1="1301.58" x2="1197.71" y2="1317.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1181.71" y1="1301.58" x2="1197.71" y2="1285.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1834.69" y1="763.508" x2="1818.69" y2="747.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1834.69" y1="763.508" x2="1818.69" y2="779.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1834.69" y1="763.508" x2="1850.69" y2="779.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1834.69" y1="763.508" x2="1850.69" y2="747.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.65" y1="1160.58" x2="1462.65" y2="1144.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.65" y1="1160.58" x2="1462.65" y2="1176.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.65" y1="1160.58" x2="1494.65" y2="1176.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.65" y1="1160.58" x2="1494.65" y2="1144.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.04" y1="1278.88" x2="990.041" y2="1262.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.04" y1="1278.88" x2="990.041" y2="1294.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.04" y1="1278.88" x2="1022.04" y2="1294.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1006.04" y1="1278.88" x2="1022.04" y2="1262.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1268.84" y1="1258.35" x2="1252.84" y2="1242.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1268.84" y1="1258.35" x2="1252.84" y2="1274.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1268.84" y1="1258.35" x2="1284.84" y2="1274.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1268.84" y1="1258.35" x2="1284.84" y2="1242.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.05" y1="1167.1" x2="1279.05" y2="1151.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.05" y1="1167.1" x2="1279.05" y2="1183.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.05" y1="1167.1" x2="1311.05" y2="1183.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.05" y1="1167.1" x2="1311.05" y2="1151.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239.88" y1="1281.07" x2="1223.88" y2="1265.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239.88" y1="1281.07" x2="1223.88" y2="1297.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239.88" y1="1281.07" x2="1255.88" y2="1297.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239.88" y1="1281.07" x2="1255.88" y2="1265.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.19" y1="1165.26" x2="1296.19" y2="1149.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.19" y1="1165.26" x2="1296.19" y2="1181.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.19" y1="1165.26" x2="1328.19" y2="1181.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.19" y1="1165.26" x2="1328.19" y2="1149.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1424" y1="1023.62" x2="1408" y2="1007.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1424" y1="1023.62" x2="1408" y2="1039.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1424" y1="1023.62" x2="1440" y2="1039.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1424" y1="1023.62" x2="1440" y2="1007.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1028.72" y1="1432.67" x2="1012.72" y2="1416.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1028.72" y1="1432.67" x2="1012.72" y2="1448.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1028.72" y1="1432.67" x2="1044.72" y2="1448.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1028.72" y1="1432.67" x2="1044.72" y2="1416.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.75" y1="1174.67" x2="1292.75" y2="1158.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.75" y1="1174.67" x2="1292.75" y2="1190.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.75" y1="1174.67" x2="1324.75" y2="1190.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.75" y1="1174.67" x2="1324.75" y2="1158.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.12" y1="1148.69" x2="1221.12" y2="1132.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.12" y1="1148.69" x2="1221.12" y2="1164.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.12" y1="1148.69" x2="1253.12" y2="1164.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.12" y1="1148.69" x2="1253.12" y2="1132.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1618.21" y1="979.458" x2="1602.21" y2="963.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1618.21" y1="979.458" x2="1602.21" y2="995.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1618.21" y1="979.458" x2="1634.21" y2="995.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1618.21" y1="979.458" x2="1634.21" y2="963.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1279.69" y1="1082.34" x2="1263.69" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1279.69" y1="1082.34" x2="1263.69" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1279.69" y1="1082.34" x2="1295.69" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1279.69" y1="1082.34" x2="1295.69" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1393.4" y1="1152.94" x2="1377.4" y2="1136.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1393.4" y1="1152.94" x2="1377.4" y2="1168.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1393.4" y1="1152.94" x2="1409.4" y2="1168.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1393.4" y1="1152.94" x2="1409.4" y2="1136.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1005.39" y1="1284.51" x2="989.392" y2="1268.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1005.39" y1="1284.51" x2="989.392" y2="1300.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1005.39" y1="1284.51" x2="1021.39" y2="1300.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1005.39" y1="1284.51" x2="1021.39" y2="1268.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.58" y1="1254.71" x2="1222.58" y2="1238.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.58" y1="1254.71" x2="1222.58" y2="1270.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.58" y1="1254.71" x2="1254.58" y2="1270.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1238.58" y1="1254.71" x2="1254.58" y2="1238.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.98" y1="894.258" x2="1593.98" y2="878.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.98" y1="894.258" x2="1593.98" y2="910.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.98" y1="894.258" x2="1625.98" y2="910.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1609.98" y1="894.258" x2="1625.98" y2="878.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1497.3" y1="1096.59" x2="1481.3" y2="1080.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1497.3" y1="1096.59" x2="1481.3" y2="1112.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1497.3" y1="1096.59" x2="1513.3" y2="1112.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1497.3" y1="1096.59" x2="1513.3" y2="1080.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1387.69" y1="1064.12" x2="1371.69" y2="1048.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1387.69" y1="1064.12" x2="1371.69" y2="1080.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1387.69" y1="1064.12" x2="1403.69" y2="1080.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1387.69" y1="1064.12" x2="1403.69" y2="1048.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1471.44" y1="1131.3" x2="1455.44" y2="1115.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1471.44" y1="1131.3" x2="1455.44" y2="1147.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1471.44" y1="1131.3" x2="1487.44" y2="1147.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1471.44" y1="1131.3" x2="1487.44" y2="1115.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1597.07" y1="924.794" x2="1581.07" y2="908.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1597.07" y1="924.794" x2="1581.07" y2="940.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1597.07" y1="924.794" x2="1613.07" y2="940.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1597.07" y1="924.794" x2="1613.07" y2="908.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="882.48" y1="1367.74" x2="866.48" y2="1351.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="882.48" y1="1367.74" x2="866.48" y2="1383.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="882.48" y1="1367.74" x2="898.48" y2="1383.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="882.48" y1="1367.74" x2="898.48" y2="1351.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1285.26" y1="1113.9" x2="1269.26" y2="1097.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1285.26" y1="1113.9" x2="1269.26" y2="1129.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1285.26" y1="1113.9" x2="1301.26" y2="1129.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1285.26" y1="1113.9" x2="1301.26" y2="1097.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.2" y1="1043.06" x2="1413.2" y2="1027.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.2" y1="1043.06" x2="1413.2" y2="1059.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.2" y1="1043.06" x2="1445.2" y2="1059.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1429.2" y1="1043.06" x2="1445.2" y2="1027.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1413.65" y1="1215.29" x2="1397.65" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1413.65" y1="1215.29" x2="1397.65" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1413.65" y1="1215.29" x2="1429.65" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1413.65" y1="1215.29" x2="1429.65" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1527.95" y1="988.132" x2="1511.95" y2="972.132" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1527.95" y1="988.132" x2="1511.95" y2="1004.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1527.95" y1="988.132" x2="1543.95" y2="1004.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1527.95" y1="988.132" x2="1543.95" y2="972.132" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.23" y1="1117.88" x2="1412.23" y2="1101.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.23" y1="1117.88" x2="1412.23" y2="1133.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.23" y1="1117.88" x2="1444.23" y2="1133.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.23" y1="1117.88" x2="1444.23" y2="1101.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.14" y1="1034.93" x2="1497.14" y2="1018.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.14" y1="1034.93" x2="1497.14" y2="1050.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.14" y1="1034.93" x2="1529.14" y2="1050.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.14" y1="1034.93" x2="1529.14" y2="1018.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1464.41" y1="1058.74" x2="1448.41" y2="1042.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1464.41" y1="1058.74" x2="1448.41" y2="1074.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1464.41" y1="1058.74" x2="1480.41" y2="1074.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1464.41" y1="1058.74" x2="1480.41" y2="1042.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1766.37" y1="756.534" x2="1750.37" y2="740.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1766.37" y1="756.534" x2="1750.37" y2="772.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1766.37" y1="756.534" x2="1782.37" y2="772.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1766.37" y1="756.534" x2="1782.37" y2="740.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.09" y1="1182.64" x2="1321.09" y2="1166.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.09" y1="1182.64" x2="1321.09" y2="1198.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.09" y1="1182.64" x2="1353.09" y2="1198.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.09" y1="1182.64" x2="1353.09" y2="1166.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.04" y1="1303.89" x2="1050.04" y2="1287.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.04" y1="1303.89" x2="1050.04" y2="1319.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.04" y1="1303.89" x2="1082.04" y2="1319.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.04" y1="1303.89" x2="1082.04" y2="1287.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.46" y1="1067.55" x2="1561.46" y2="1051.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.46" y1="1067.55" x2="1561.46" y2="1083.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.46" y1="1067.55" x2="1593.46" y2="1083.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1577.46" y1="1067.55" x2="1593.46" y2="1051.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.74" y1="1243.98" x2="1121.74" y2="1227.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.74" y1="1243.98" x2="1121.74" y2="1259.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.74" y1="1243.98" x2="1153.74" y2="1259.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.74" y1="1243.98" x2="1153.74" y2="1227.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.72" y1="1098.59" x2="1349.72" y2="1082.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.72" y1="1098.59" x2="1349.72" y2="1114.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.72" y1="1098.59" x2="1381.72" y2="1114.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.72" y1="1098.59" x2="1381.72" y2="1082.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1422.2" y1="999.716" x2="1406.2" y2="983.716" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1422.2" y1="999.716" x2="1406.2" y2="1015.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1422.2" y1="999.716" x2="1438.2" y2="1015.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1422.2" y1="999.716" x2="1438.2" y2="983.716" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1381.71" y1="975.726" x2="1365.71" y2="959.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1381.71" y1="975.726" x2="1365.71" y2="991.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1381.71" y1="975.726" x2="1397.71" y2="991.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1381.71" y1="975.726" x2="1397.71" y2="959.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.037" y1="1422.12" x2="859.037" y2="1406.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.037" y1="1422.12" x2="859.037" y2="1438.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.037" y1="1422.12" x2="891.037" y2="1438.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.037" y1="1422.12" x2="891.037" y2="1406.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.48" y1="1128.32" x2="1100.48" y2="1112.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.48" y1="1128.32" x2="1100.48" y2="1144.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.48" y1="1128.32" x2="1132.48" y2="1144.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.48" y1="1128.32" x2="1132.48" y2="1112.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="898.259" y1="1319.39" x2="882.259" y2="1303.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="898.259" y1="1319.39" x2="882.259" y2="1335.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="898.259" y1="1319.39" x2="914.259" y2="1335.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="898.259" y1="1319.39" x2="914.259" y2="1303.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.31" y1="1035.06" x2="1415.31" y2="1019.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.31" y1="1035.06" x2="1415.31" y2="1051.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.31" y1="1035.06" x2="1447.31" y2="1051.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.31" y1="1035.06" x2="1447.31" y2="1019.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.19" y1="815.289" x2="1746.19" y2="799.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.19" y1="815.289" x2="1746.19" y2="831.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.19" y1="815.289" x2="1778.19" y2="831.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.19" y1="815.289" x2="1778.19" y2="799.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="843.736" y1="1421.71" x2="827.736" y2="1405.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="843.736" y1="1421.71" x2="827.736" y2="1437.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="843.736" y1="1421.71" x2="859.736" y2="1437.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="843.736" y1="1421.71" x2="859.736" y2="1405.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1329.12" y1="1076.38" x2="1313.12" y2="1060.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1329.12" y1="1076.38" x2="1313.12" y2="1092.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1329.12" y1="1076.38" x2="1345.12" y2="1092.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1329.12" y1="1076.38" x2="1345.12" y2="1060.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.51" y1="1094.73" x2="1325.51" y2="1078.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.51" y1="1094.73" x2="1325.51" y2="1110.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.51" y1="1094.73" x2="1357.51" y2="1110.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1341.51" y1="1094.73" x2="1357.51" y2="1078.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.39" y1="1147.76" x2="1209.39" y2="1131.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.39" y1="1147.76" x2="1209.39" y2="1163.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.39" y1="1147.76" x2="1241.39" y2="1163.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.39" y1="1147.76" x2="1241.39" y2="1131.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.09" y1="1248.47" x2="1075.09" y2="1232.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.09" y1="1248.47" x2="1075.09" y2="1264.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.09" y1="1248.47" x2="1107.09" y2="1264.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.09" y1="1248.47" x2="1107.09" y2="1232.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1399.57" y1="992.596" x2="1383.57" y2="976.596" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1399.57" y1="992.596" x2="1383.57" y2="1008.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1399.57" y1="992.596" x2="1415.57" y2="1008.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1399.57" y1="992.596" x2="1415.57" y2="976.596" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="712.506" y1="1466.98" x2="696.506" y2="1450.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="712.506" y1="1466.98" x2="696.506" y2="1482.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="712.506" y1="1466.98" x2="728.506" y2="1482.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="712.506" y1="1466.98" x2="728.506" y2="1450.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1038.8" y1="1215.97" x2="1022.8" y2="1199.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1038.8" y1="1215.97" x2="1022.8" y2="1231.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1038.8" y1="1215.97" x2="1054.8" y2="1231.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1038.8" y1="1215.97" x2="1054.8" y2="1199.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="996.845" y1="1298.95" x2="980.845" y2="1282.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="996.845" y1="1298.95" x2="980.845" y2="1314.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="996.845" y1="1298.95" x2="1012.85" y2="1314.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="996.845" y1="1298.95" x2="1012.85" y2="1282.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="764.769" y1="1428.81" x2="748.769" y2="1412.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="764.769" y1="1428.81" x2="748.769" y2="1444.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="764.769" y1="1428.81" x2="780.769" y2="1444.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="764.769" y1="1428.81" x2="780.769" y2="1412.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1481.71" y1="979.429" x2="1465.71" y2="963.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1481.71" y1="979.429" x2="1465.71" y2="995.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1481.71" y1="979.429" x2="1497.71" y2="995.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1481.71" y1="979.429" x2="1497.71" y2="963.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="794.022" y1="1392.87" x2="778.022" y2="1376.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="794.022" y1="1392.87" x2="778.022" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="794.022" y1="1392.87" x2="810.022" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="794.022" y1="1392.87" x2="810.022" y2="1376.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.54" y1="1232.13" x2="1099.54" y2="1216.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.54" y1="1232.13" x2="1099.54" y2="1248.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.54" y1="1232.13" x2="1131.54" y2="1248.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.54" y1="1232.13" x2="1131.54" y2="1216.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="989.1" y1="1361.68" x2="973.1" y2="1345.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="989.1" y1="1361.68" x2="973.1" y2="1377.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="989.1" y1="1361.68" x2="1005.1" y2="1377.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="989.1" y1="1361.68" x2="1005.1" y2="1345.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="658.223" y1="1594.72" x2="642.223" y2="1578.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="658.223" y1="1594.72" x2="642.223" y2="1610.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="658.223" y1="1594.72" x2="674.223" y2="1610.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="658.223" y1="1594.72" x2="674.223" y2="1578.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.13" y1="922.6" x2="1595.13" y2="906.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.13" y1="922.6" x2="1595.13" y2="938.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.13" y1="922.6" x2="1627.13" y2="938.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1611.13" y1="922.6" x2="1627.13" y2="906.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1325.74" y1="1073.3" x2="1309.74" y2="1057.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1325.74" y1="1073.3" x2="1309.74" y2="1089.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1325.74" y1="1073.3" x2="1341.74" y2="1089.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1325.74" y1="1073.3" x2="1341.74" y2="1057.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.65" y1="1061.83" x2="1438.65" y2="1045.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.65" y1="1061.83" x2="1438.65" y2="1077.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.65" y1="1061.83" x2="1470.65" y2="1077.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.65" y1="1061.83" x2="1470.65" y2="1045.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1270.59" y1="1028.09" x2="1254.59" y2="1012.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1270.59" y1="1028.09" x2="1254.59" y2="1044.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1270.59" y1="1028.09" x2="1286.59" y2="1044.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1270.59" y1="1028.09" x2="1286.59" y2="1012.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.91" y1="1068.7" x2="1318.91" y2="1052.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.91" y1="1068.7" x2="1318.91" y2="1084.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.91" y1="1068.7" x2="1350.91" y2="1084.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.91" y1="1068.7" x2="1350.91" y2="1052.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.75" y1="1086.14" x2="1324.75" y2="1070.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.75" y1="1086.14" x2="1324.75" y2="1102.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.75" y1="1086.14" x2="1356.75" y2="1102.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1340.75" y1="1086.14" x2="1356.75" y2="1070.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="968.976" y1="1423.08" x2="952.976" y2="1407.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="968.976" y1="1423.08" x2="952.976" y2="1439.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="968.976" y1="1423.08" x2="984.976" y2="1439.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="968.976" y1="1423.08" x2="984.976" y2="1407.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1451.83" y1="926.379" x2="1435.83" y2="910.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1451.83" y1="926.379" x2="1435.83" y2="942.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1451.83" y1="926.379" x2="1467.83" y2="942.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1451.83" y1="926.379" x2="1467.83" y2="910.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1420.21" y1="1170.17" x2="1404.21" y2="1154.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1420.21" y1="1170.17" x2="1404.21" y2="1186.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1420.21" y1="1170.17" x2="1436.21" y2="1186.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1420.21" y1="1170.17" x2="1436.21" y2="1154.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.55" y1="1324.46" x2="1014.55" y2="1308.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.55" y1="1324.46" x2="1014.55" y2="1340.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.55" y1="1324.46" x2="1046.55" y2="1340.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1030.55" y1="1324.46" x2="1046.55" y2="1308.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.85" y1="1057.07" x2="1398.85" y2="1041.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.85" y1="1057.07" x2="1398.85" y2="1073.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.85" y1="1057.07" x2="1430.85" y2="1073.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.85" y1="1057.07" x2="1430.85" y2="1041.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1037.88" y1="1309.43" x2="1021.88" y2="1293.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1037.88" y1="1309.43" x2="1021.88" y2="1325.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1037.88" y1="1309.43" x2="1053.88" y2="1325.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1037.88" y1="1309.43" x2="1053.88" y2="1293.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1171.81" y1="1298.75" x2="1155.81" y2="1282.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1171.81" y1="1298.75" x2="1155.81" y2="1314.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1171.81" y1="1298.75" x2="1187.81" y2="1314.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1171.81" y1="1298.75" x2="1187.81" y2="1282.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.86" y1="1223.11" x2="1097.86" y2="1207.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.86" y1="1223.11" x2="1097.86" y2="1239.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.86" y1="1223.11" x2="1129.86" y2="1239.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.86" y1="1223.11" x2="1129.86" y2="1207.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.45" y1="1086.98" x2="1444.45" y2="1070.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.45" y1="1086.98" x2="1444.45" y2="1102.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.45" y1="1086.98" x2="1476.45" y2="1102.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.45" y1="1086.98" x2="1476.45" y2="1070.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1557.36" y1="1088.65" x2="1541.36" y2="1072.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1557.36" y1="1088.65" x2="1541.36" y2="1104.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1557.36" y1="1088.65" x2="1573.36" y2="1104.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1557.36" y1="1088.65" x2="1573.36" y2="1072.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="908.822" y1="1392.95" x2="892.822" y2="1376.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="908.822" y1="1392.95" x2="892.822" y2="1408.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="908.822" y1="1392.95" x2="924.822" y2="1408.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="908.822" y1="1392.95" x2="924.822" y2="1376.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.57" y1="1011.37" x2="1484.57" y2="995.374" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.57" y1="1011.37" x2="1484.57" y2="1027.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.57" y1="1011.37" x2="1516.57" y2="1027.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1500.57" y1="1011.37" x2="1516.57" y2="995.374" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1002.47" y1="1234.46" x2="986.471" y2="1218.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1002.47" y1="1234.46" x2="986.471" y2="1250.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1002.47" y1="1234.46" x2="1018.47" y2="1250.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1002.47" y1="1234.46" x2="1018.47" y2="1218.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.95" y1="1013.72" x2="1442.95" y2="997.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.95" y1="1013.72" x2="1442.95" y2="1029.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.95" y1="1013.72" x2="1474.95" y2="1029.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1458.95" y1="1013.72" x2="1474.95" y2="997.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.63" y1="1328.09" x2="1043.63" y2="1312.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.63" y1="1328.09" x2="1043.63" y2="1344.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.63" y1="1328.09" x2="1075.63" y2="1344.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1059.63" y1="1328.09" x2="1075.63" y2="1312.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1046.37" x2="1288.99" y2="1030.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1046.37" x2="1288.99" y2="1062.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1046.37" x2="1320.99" y2="1062.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1304.99" y1="1046.37" x2="1320.99" y2="1030.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1016.99" y1="1290.74" x2="1000.99" y2="1274.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1016.99" y1="1290.74" x2="1000.99" y2="1306.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1016.99" y1="1290.74" x2="1032.99" y2="1306.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1016.99" y1="1290.74" x2="1032.99" y2="1274.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="892.734" y1="1460.48" x2="876.734" y2="1444.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="892.734" y1="1460.48" x2="876.734" y2="1476.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="892.734" y1="1460.48" x2="908.734" y2="1476.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="892.734" y1="1460.48" x2="908.734" y2="1444.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.22" y1="1187.04" x2="1030.22" y2="1171.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.22" y1="1187.04" x2="1030.22" y2="1203.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.22" y1="1187.04" x2="1062.22" y2="1203.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.22" y1="1187.04" x2="1062.22" y2="1171.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.88" y1="1292.71" x2="1060.88" y2="1276.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.88" y1="1292.71" x2="1060.88" y2="1308.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.88" y1="1292.71" x2="1092.88" y2="1308.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.88" y1="1292.71" x2="1092.88" y2="1276.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.79" y1="924.706" x2="1529.79" y2="908.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.79" y1="924.706" x2="1529.79" y2="940.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.79" y1="924.706" x2="1561.79" y2="940.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.79" y1="924.706" x2="1561.79" y2="908.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1003.69" y1="1262.65" x2="987.692" y2="1246.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1003.69" y1="1262.65" x2="987.692" y2="1278.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1003.69" y1="1262.65" x2="1019.69" y2="1278.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1003.69" y1="1262.65" x2="1019.69" y2="1246.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1245.12" y1="1104.1" x2="1229.12" y2="1088.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1245.12" y1="1104.1" x2="1229.12" y2="1120.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1245.12" y1="1104.1" x2="1261.12" y2="1120.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1245.12" y1="1104.1" x2="1261.12" y2="1088.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.07" y1="1264.22" x2="1097.07" y2="1248.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.07" y1="1264.22" x2="1097.07" y2="1280.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.07" y1="1264.22" x2="1129.07" y2="1280.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1113.07" y1="1264.22" x2="1129.07" y2="1248.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="862.589" y1="1433.49" x2="846.589" y2="1417.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="862.589" y1="1433.49" x2="846.589" y2="1449.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="862.589" y1="1433.49" x2="878.589" y2="1449.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="862.589" y1="1433.49" x2="878.589" y2="1417.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1335.3" y1="1153.33" x2="1319.3" y2="1137.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1335.3" y1="1153.33" x2="1319.3" y2="1169.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1335.3" y1="1153.33" x2="1351.3" y2="1169.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1335.3" y1="1153.33" x2="1351.3" y2="1137.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1521.7" y1="1018.9" x2="1505.7" y2="1002.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1521.7" y1="1018.9" x2="1505.7" y2="1034.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1521.7" y1="1018.9" x2="1537.7" y2="1034.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1521.7" y1="1018.9" x2="1537.7" y2="1002.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1384.65" y1="1259.56" x2="1368.65" y2="1243.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1384.65" y1="1259.56" x2="1368.65" y2="1275.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1384.65" y1="1259.56" x2="1400.65" y2="1275.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1384.65" y1="1259.56" x2="1400.65" y2="1243.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1366.24" y1="1165.25" x2="1350.24" y2="1149.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1366.24" y1="1165.25" x2="1350.24" y2="1181.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1366.24" y1="1165.25" x2="1382.24" y2="1181.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1366.24" y1="1165.25" x2="1382.24" y2="1149.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="522.504" y1="1512.98" x2="506.504" y2="1496.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="522.504" y1="1512.98" x2="506.504" y2="1528.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="522.504" y1="1512.98" x2="538.504" y2="1528.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="522.504" y1="1512.98" x2="538.504" y2="1496.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.208" y1="1309.37" x2="956.208" y2="1293.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.208" y1="1309.37" x2="956.208" y2="1325.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.208" y1="1309.37" x2="988.208" y2="1325.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.208" y1="1309.37" x2="988.208" y2="1293.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.36" y1="1054" x2="1160.36" y2="1038" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.36" y1="1054" x2="1160.36" y2="1070" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.36" y1="1054" x2="1192.36" y2="1070" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.36" y1="1054" x2="1192.36" y2="1038" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.15" y1="1063.66" x2="1299.15" y2="1047.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.15" y1="1063.66" x2="1299.15" y2="1079.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.15" y1="1063.66" x2="1331.15" y2="1079.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1315.15" y1="1063.66" x2="1331.15" y2="1047.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1235.21" y1="1142.27" x2="1219.21" y2="1126.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1235.21" y1="1142.27" x2="1219.21" y2="1158.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1235.21" y1="1142.27" x2="1251.21" y2="1158.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1235.21" y1="1142.27" x2="1251.21" y2="1126.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.46" y1="1190.1" x2="1213.46" y2="1174.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.46" y1="1190.1" x2="1213.46" y2="1206.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.46" y1="1190.1" x2="1245.46" y2="1206.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1229.46" y1="1190.1" x2="1245.46" y2="1174.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.451" y1="1406.48" x2="728.451" y2="1390.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.451" y1="1406.48" x2="728.451" y2="1422.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.451" y1="1406.48" x2="760.451" y2="1422.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.451" y1="1406.48" x2="760.451" y2="1390.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.07" y1="1218.16" x2="1081.07" y2="1202.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.07" y1="1218.16" x2="1081.07" y2="1234.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.07" y1="1218.16" x2="1113.07" y2="1234.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.07" y1="1218.16" x2="1113.07" y2="1202.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1012.88" y1="1293.81" x2="996.879" y2="1277.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1012.88" y1="1293.81" x2="996.879" y2="1309.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1012.88" y1="1293.81" x2="1028.88" y2="1309.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1012.88" y1="1293.81" x2="1028.88" y2="1277.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="277.68" y1="1755.15" x2="261.68" y2="1739.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="277.68" y1="1755.15" x2="261.68" y2="1771.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="277.68" y1="1755.15" x2="293.68" y2="1771.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="277.68" y1="1755.15" x2="293.68" y2="1739.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1166.29" y1="1406.05" x2="1150.29" y2="1390.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1166.29" y1="1406.05" x2="1150.29" y2="1422.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1166.29" y1="1406.05" x2="1182.29" y2="1422.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1166.29" y1="1406.05" x2="1182.29" y2="1390.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1178.56" y1="1242.2" x2="1162.56" y2="1226.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1178.56" y1="1242.2" x2="1162.56" y2="1258.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1178.56" y1="1242.2" x2="1194.56" y2="1258.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1178.56" y1="1242.2" x2="1194.56" y2="1226.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1498.04" y1="1011.64" x2="1482.04" y2="995.641" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1498.04" y1="1011.64" x2="1482.04" y2="1027.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1498.04" y1="1011.64" x2="1514.04" y2="1027.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1498.04" y1="1011.64" x2="1514.04" y2="995.641" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1380.98" y1="1081.87" x2="1364.98" y2="1065.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1380.98" y1="1081.87" x2="1364.98" y2="1097.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1380.98" y1="1081.87" x2="1396.98" y2="1097.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1380.98" y1="1081.87" x2="1396.98" y2="1065.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1189.27" y1="1225.02" x2="1173.27" y2="1209.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1189.27" y1="1225.02" x2="1173.27" y2="1241.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1189.27" y1="1225.02" x2="1205.27" y2="1241.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1189.27" y1="1225.02" x2="1205.27" y2="1209.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.33" y1="1091.79" x2="1337.33" y2="1075.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.33" y1="1091.79" x2="1337.33" y2="1107.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.33" y1="1091.79" x2="1369.33" y2="1107.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.33" y1="1091.79" x2="1369.33" y2="1075.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.57" y1="1214.17" x2="1175.57" y2="1198.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.57" y1="1214.17" x2="1175.57" y2="1230.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.57" y1="1214.17" x2="1207.57" y2="1230.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1191.57" y1="1214.17" x2="1207.57" y2="1198.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.19" y1="842.118" x2="1764.19" y2="826.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.19" y1="842.118" x2="1764.19" y2="858.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.19" y1="842.118" x2="1796.19" y2="858.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.19" y1="842.118" x2="1796.19" y2="826.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="855.143" y1="1461.56" x2="839.143" y2="1445.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="855.143" y1="1461.56" x2="839.143" y2="1477.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="855.143" y1="1461.56" x2="871.143" y2="1477.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="855.143" y1="1461.56" x2="871.143" y2="1445.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1487.2" y1="1215.29" x2="1471.2" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1487.2" y1="1215.29" x2="1471.2" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1487.2" y1="1215.29" x2="1503.2" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1487.2" y1="1215.29" x2="1503.2" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="941.026" y1="1398.61" x2="925.026" y2="1382.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="941.026" y1="1398.61" x2="925.026" y2="1414.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="941.026" y1="1398.61" x2="957.026" y2="1414.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="941.026" y1="1398.61" x2="957.026" y2="1382.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1491.9" y1="862.19" x2="1475.9" y2="846.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1491.9" y1="862.19" x2="1475.9" y2="878.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1491.9" y1="862.19" x2="1507.9" y2="878.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1491.9" y1="862.19" x2="1507.9" y2="846.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1045.21" y1="1271.31" x2="1029.21" y2="1255.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1045.21" y1="1271.31" x2="1029.21" y2="1287.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1045.21" y1="1271.31" x2="1061.21" y2="1287.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1045.21" y1="1271.31" x2="1061.21" y2="1255.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.16" y1="1192.34" x2="1262.16" y2="1176.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.16" y1="1192.34" x2="1262.16" y2="1208.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.16" y1="1192.34" x2="1294.16" y2="1208.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.16" y1="1192.34" x2="1294.16" y2="1176.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.72" y1="1184.27" x2="1262.72" y2="1168.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.72" y1="1184.27" x2="1262.72" y2="1200.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.72" y1="1184.27" x2="1294.72" y2="1200.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1278.72" y1="1184.27" x2="1294.72" y2="1168.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1614.72" y1="834.262" x2="1598.72" y2="818.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1614.72" y1="834.262" x2="1598.72" y2="850.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1614.72" y1="834.262" x2="1630.72" y2="850.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1614.72" y1="834.262" x2="1630.72" y2="818.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.83" y1="933.105" x2="1599.83" y2="917.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.83" y1="933.105" x2="1599.83" y2="949.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.83" y1="933.105" x2="1631.83" y2="949.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.83" y1="933.105" x2="1631.83" y2="917.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1306.51" y1="1052.44" x2="1290.51" y2="1036.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1306.51" y1="1052.44" x2="1290.51" y2="1068.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1306.51" y1="1052.44" x2="1322.51" y2="1068.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1306.51" y1="1052.44" x2="1322.51" y2="1036.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.55" y1="1095.87" x2="1376.55" y2="1079.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.55" y1="1095.87" x2="1376.55" y2="1111.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.55" y1="1095.87" x2="1408.55" y2="1111.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.55" y1="1095.87" x2="1408.55" y2="1079.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.951" y1="1271.4" x2="883.951" y2="1255.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.951" y1="1271.4" x2="883.951" y2="1287.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.951" y1="1271.4" x2="915.951" y2="1287.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.951" y1="1271.4" x2="915.951" y2="1255.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.68" y1="1069.21" x2="1376.68" y2="1053.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.68" y1="1069.21" x2="1376.68" y2="1085.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.68" y1="1069.21" x2="1408.68" y2="1085.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1392.68" y1="1069.21" x2="1408.68" y2="1053.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1651.8" y1="933.534" x2="1635.8" y2="917.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1651.8" y1="933.534" x2="1635.8" y2="949.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1651.8" y1="933.534" x2="1667.8" y2="949.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1651.8" y1="933.534" x2="1667.8" y2="917.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.2" y1="982.773" x2="1530.2" y2="966.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.2" y1="982.773" x2="1530.2" y2="998.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.2" y1="982.773" x2="1562.2" y2="998.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1546.2" y1="982.773" x2="1562.2" y2="966.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1197.19" y1="1159.24" x2="1181.19" y2="1143.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1197.19" y1="1159.24" x2="1181.19" y2="1175.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1197.19" y1="1159.24" x2="1213.19" y2="1175.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1197.19" y1="1159.24" x2="1213.19" y2="1143.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1378.06" y1="1142.61" x2="1362.06" y2="1126.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1378.06" y1="1142.61" x2="1362.06" y2="1158.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1378.06" y1="1142.61" x2="1394.06" y2="1158.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1378.06" y1="1142.61" x2="1394.06" y2="1126.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="798.462" y1="1475.78" x2="782.462" y2="1459.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="798.462" y1="1475.78" x2="782.462" y2="1491.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="798.462" y1="1475.78" x2="814.462" y2="1491.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="798.462" y1="1475.78" x2="814.462" y2="1459.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1653.74" y1="920.543" x2="1637.74" y2="904.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1653.74" y1="920.543" x2="1637.74" y2="936.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1653.74" y1="920.543" x2="1669.74" y2="936.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1653.74" y1="920.543" x2="1669.74" y2="904.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.58" y1="1157.39" x2="1062.58" y2="1141.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.58" y1="1157.39" x2="1062.58" y2="1173.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.58" y1="1157.39" x2="1094.58" y2="1173.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1078.58" y1="1157.39" x2="1094.58" y2="1141.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.44" y1="1151.58" x2="1405.44" y2="1135.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.44" y1="1151.58" x2="1405.44" y2="1167.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.44" y1="1151.58" x2="1437.44" y2="1167.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.44" y1="1151.58" x2="1437.44" y2="1135.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.19" y1="1152.96" x2="1277.19" y2="1136.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.19" y1="1152.96" x2="1277.19" y2="1168.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.19" y1="1152.96" x2="1309.19" y2="1168.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.19" y1="1152.96" x2="1309.19" y2="1136.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1307.01" y1="1135.59" x2="1291.01" y2="1119.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1307.01" y1="1135.59" x2="1291.01" y2="1151.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1307.01" y1="1135.59" x2="1323.01" y2="1151.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1307.01" y1="1135.59" x2="1323.01" y2="1119.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1157.68" y1="1240.17" x2="1141.68" y2="1224.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1157.68" y1="1240.17" x2="1141.68" y2="1256.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1157.68" y1="1240.17" x2="1173.68" y2="1256.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1157.68" y1="1240.17" x2="1173.68" y2="1224.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1322.34" y1="1098.52" x2="1306.34" y2="1082.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1322.34" y1="1098.52" x2="1306.34" y2="1114.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1322.34" y1="1098.52" x2="1338.34" y2="1114.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1322.34" y1="1098.52" x2="1338.34" y2="1082.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.24" y1="1200.4" x2="1231.24" y2="1184.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.24" y1="1200.4" x2="1231.24" y2="1216.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.24" y1="1200.4" x2="1263.24" y2="1216.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1247.24" y1="1200.4" x2="1263.24" y2="1184.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.59" y1="1055.14" x2="1348.59" y2="1039.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.59" y1="1055.14" x2="1348.59" y2="1071.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.59" y1="1055.14" x2="1380.59" y2="1071.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.59" y1="1055.14" x2="1380.59" y2="1039.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1640.29" y1="1056.93" x2="1624.29" y2="1040.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1640.29" y1="1056.93" x2="1624.29" y2="1072.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1640.29" y1="1056.93" x2="1656.29" y2="1072.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1640.29" y1="1056.93" x2="1656.29" y2="1040.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1286.58" y1="1059.6" x2="1270.58" y2="1043.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1286.58" y1="1059.6" x2="1270.58" y2="1075.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1286.58" y1="1059.6" x2="1302.58" y2="1075.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1286.58" y1="1059.6" x2="1302.58" y2="1043.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1550.14" y1="1004.56" x2="1534.14" y2="988.562" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1550.14" y1="1004.56" x2="1534.14" y2="1020.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1550.14" y1="1004.56" x2="1566.14" y2="1020.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1550.14" y1="1004.56" x2="1566.14" y2="988.562" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="955.148" y1="1370.3" x2="939.148" y2="1354.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="955.148" y1="1370.3" x2="939.148" y2="1386.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="955.148" y1="1370.3" x2="971.148" y2="1386.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="955.148" y1="1370.3" x2="971.148" y2="1354.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1845.45" y1="659.705" x2="1829.45" y2="643.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1845.45" y1="659.705" x2="1829.45" y2="675.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1845.45" y1="659.705" x2="1861.45" y2="675.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1845.45" y1="659.705" x2="1861.45" y2="643.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1817.11" y1="731.36" x2="1801.11" y2="715.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1817.11" y1="731.36" x2="1801.11" y2="747.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1817.11" y1="731.36" x2="1833.11" y2="747.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1817.11" y1="731.36" x2="1833.11" y2="715.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1838.37" y1="806.858" x2="1822.37" y2="790.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1838.37" y1="806.858" x2="1822.37" y2="822.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1838.37" y1="806.858" x2="1854.37" y2="822.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1838.37" y1="806.858" x2="1854.37" y2="790.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1132.81" y1="1166.63" x2="1116.81" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1132.81" y1="1166.63" x2="1116.81" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1132.81" y1="1166.63" x2="1148.81" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1132.81" y1="1166.63" x2="1148.81" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244" y1="1263.65" x2="1228" y2="1247.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244" y1="1263.65" x2="1228" y2="1279.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244" y1="1263.65" x2="1260" y2="1279.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244" y1="1263.65" x2="1260" y2="1247.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.43" y1="1066.19" x2="1274.43" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.43" y1="1066.19" x2="1274.43" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.43" y1="1066.19" x2="1306.43" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.43" y1="1066.19" x2="1306.43" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1558.7" y1="965.028" x2="1542.7" y2="949.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1558.7" y1="965.028" x2="1542.7" y2="981.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1558.7" y1="965.028" x2="1574.7" y2="981.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1558.7" y1="965.028" x2="1574.7" y2="949.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.76" y1="970.455" x2="1458.76" y2="954.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.76" y1="970.455" x2="1458.76" y2="986.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.76" y1="970.455" x2="1490.76" y2="986.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.76" y1="970.455" x2="1490.76" y2="954.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1682.62" y1="820.381" x2="1666.62" y2="804.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1682.62" y1="820.381" x2="1666.62" y2="836.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1682.62" y1="820.381" x2="1698.62" y2="836.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1682.62" y1="820.381" x2="1698.62" y2="804.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="872.16" y1="1370.98" x2="856.16" y2="1354.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="872.16" y1="1370.98" x2="856.16" y2="1386.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="872.16" y1="1370.98" x2="888.16" y2="1386.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="872.16" y1="1370.98" x2="888.16" y2="1354.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1155.08" y1="1163.95" x2="1139.08" y2="1147.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1155.08" y1="1163.95" x2="1139.08" y2="1179.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1155.08" y1="1163.95" x2="1171.08" y2="1179.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1155.08" y1="1163.95" x2="1171.08" y2="1147.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1316.44" y1="1111.88" x2="1300.44" y2="1095.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1316.44" y1="1111.88" x2="1300.44" y2="1127.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1316.44" y1="1111.88" x2="1332.44" y2="1127.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1316.44" y1="1111.88" x2="1332.44" y2="1095.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="866.851" y1="1331.78" x2="850.851" y2="1315.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="866.851" y1="1331.78" x2="850.851" y2="1347.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="866.851" y1="1331.78" x2="882.851" y2="1347.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="866.851" y1="1331.78" x2="882.851" y2="1315.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.47" y1="1146.54" x2="1438.47" y2="1130.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.47" y1="1146.54" x2="1438.47" y2="1162.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.47" y1="1146.54" x2="1470.47" y2="1162.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1454.47" y1="1146.54" x2="1470.47" y2="1130.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1345.73" y1="1121.54" x2="1329.73" y2="1105.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1345.73" y1="1121.54" x2="1329.73" y2="1137.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1345.73" y1="1121.54" x2="1361.73" y2="1137.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1345.73" y1="1121.54" x2="1361.73" y2="1105.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1604.79" y1="991.27" x2="1588.79" y2="975.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1604.79" y1="991.27" x2="1588.79" y2="1007.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1604.79" y1="991.27" x2="1620.79" y2="1007.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1604.79" y1="991.27" x2="1620.79" y2="975.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1425.4" y1="1018.53" x2="1409.4" y2="1002.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1425.4" y1="1018.53" x2="1409.4" y2="1034.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1425.4" y1="1018.53" x2="1441.4" y2="1034.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1425.4" y1="1018.53" x2="1441.4" y2="1002.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1177.09" y1="1221.26" x2="1161.09" y2="1205.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1177.09" y1="1221.26" x2="1161.09" y2="1237.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1177.09" y1="1221.26" x2="1193.09" y2="1237.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1177.09" y1="1221.26" x2="1193.09" y2="1205.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="436.181" y1="1793.59" x2="420.181" y2="1777.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="436.181" y1="1793.59" x2="420.181" y2="1809.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="436.181" y1="1793.59" x2="452.181" y2="1809.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="436.181" y1="1793.59" x2="452.181" y2="1777.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="957.079" y1="1394.58" x2="941.079" y2="1378.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="957.079" y1="1394.58" x2="941.079" y2="1410.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="957.079" y1="1394.58" x2="973.079" y2="1410.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="957.079" y1="1394.58" x2="973.079" y2="1378.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1427.73" y1="1093.18" x2="1411.73" y2="1077.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1427.73" y1="1093.18" x2="1411.73" y2="1109.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1427.73" y1="1093.18" x2="1443.73" y2="1109.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1427.73" y1="1093.18" x2="1443.73" y2="1077.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1348.36" y1="1249.7" x2="1332.36" y2="1233.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1348.36" y1="1249.7" x2="1332.36" y2="1265.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1348.36" y1="1249.7" x2="1364.36" y2="1265.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1348.36" y1="1249.7" x2="1364.36" y2="1233.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.366" y1="1495.39" x2="775.366" y2="1479.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.366" y1="1495.39" x2="775.366" y2="1511.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.366" y1="1495.39" x2="807.366" y2="1511.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.366" y1="1495.39" x2="807.366" y2="1479.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1636.9" y1="890.133" x2="1620.9" y2="874.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1636.9" y1="890.133" x2="1620.9" y2="906.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1636.9" y1="890.133" x2="1652.9" y2="906.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1636.9" y1="890.133" x2="1652.9" y2="874.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1256" y1="1205.58" x2="1240" y2="1189.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1256" y1="1205.58" x2="1240" y2="1221.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1256" y1="1205.58" x2="1272" y2="1221.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1256" y1="1205.58" x2="1272" y2="1189.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.44" y1="1065.3" x2="1412.44" y2="1049.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.44" y1="1065.3" x2="1412.44" y2="1081.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.44" y1="1065.3" x2="1444.44" y2="1081.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.44" y1="1065.3" x2="1444.44" y2="1049.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.03" y1="1294.95" x2="1030.03" y2="1278.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.03" y1="1294.95" x2="1030.03" y2="1310.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.03" y1="1294.95" x2="1062.03" y2="1310.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1046.03" y1="1294.95" x2="1062.03" y2="1278.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="721.091" y1="1561.75" x2="705.091" y2="1545.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="721.091" y1="1561.75" x2="705.091" y2="1577.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="721.091" y1="1561.75" x2="737.091" y2="1577.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="721.091" y1="1561.75" x2="737.091" y2="1545.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1085.36" y1="1148.29" x2="1069.36" y2="1132.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1085.36" y1="1148.29" x2="1069.36" y2="1164.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1085.36" y1="1148.29" x2="1101.36" y2="1164.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1085.36" y1="1148.29" x2="1101.36" y2="1132.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="823.473" y1="1446.58" x2="807.473" y2="1430.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="823.473" y1="1446.58" x2="807.473" y2="1462.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="823.473" y1="1446.58" x2="839.473" y2="1462.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="823.473" y1="1446.58" x2="839.473" y2="1430.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="887.822" y1="1375.21" x2="871.822" y2="1359.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="887.822" y1="1375.21" x2="871.822" y2="1391.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="887.822" y1="1375.21" x2="903.822" y2="1391.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="887.822" y1="1375.21" x2="903.822" y2="1359.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.18" y1="1169.39" x2="1147.18" y2="1153.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.18" y1="1169.39" x2="1147.18" y2="1185.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.18" y1="1169.39" x2="1179.18" y2="1185.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.18" y1="1169.39" x2="1179.18" y2="1153.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.39" y1="1014.06" x2="1468.39" y2="998.061" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.39" y1="1014.06" x2="1468.39" y2="1030.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.39" y1="1014.06" x2="1500.39" y2="1030.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1484.39" y1="1014.06" x2="1500.39" y2="998.061" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1685.1" y1="924.047" x2="1669.1" y2="908.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1685.1" y1="924.047" x2="1669.1" y2="940.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1685.1" y1="924.047" x2="1701.1" y2="940.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1685.1" y1="924.047" x2="1701.1" y2="908.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1070.78" y1="1266.71" x2="1054.78" y2="1250.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1070.78" y1="1266.71" x2="1054.78" y2="1282.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1070.78" y1="1266.71" x2="1086.78" y2="1282.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1070.78" y1="1266.71" x2="1086.78" y2="1250.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1379.51" y1="1049.2" x2="1363.51" y2="1033.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1379.51" y1="1049.2" x2="1363.51" y2="1065.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1379.51" y1="1049.2" x2="1395.51" y2="1065.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1379.51" y1="1049.2" x2="1395.51" y2="1033.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="877.025" y1="1403.06" x2="861.025" y2="1387.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="877.025" y1="1403.06" x2="861.025" y2="1419.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="877.025" y1="1403.06" x2="893.025" y2="1419.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="877.025" y1="1403.06" x2="893.025" y2="1387.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.37" y1="1191.22" x2="1060.37" y2="1175.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.37" y1="1191.22" x2="1060.37" y2="1207.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.37" y1="1191.22" x2="1092.37" y2="1207.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1076.37" y1="1191.22" x2="1092.37" y2="1175.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1164.55" y1="1249.8" x2="1148.55" y2="1233.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1164.55" y1="1249.8" x2="1148.55" y2="1265.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1164.55" y1="1249.8" x2="1180.55" y2="1265.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1164.55" y1="1249.8" x2="1180.55" y2="1233.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1473.39" y1="1000.62" x2="1457.39" y2="984.624" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1473.39" y1="1000.62" x2="1457.39" y2="1016.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1473.39" y1="1000.62" x2="1489.39" y2="1016.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1473.39" y1="1000.62" x2="1489.39" y2="984.624" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.54" y1="1202.19" x2="1279.54" y2="1186.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.54" y1="1202.19" x2="1279.54" y2="1218.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.54" y1="1202.19" x2="1311.54" y2="1218.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1295.54" y1="1202.19" x2="1311.54" y2="1186.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.58" y1="1267.5" x2="1100.58" y2="1251.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.58" y1="1267.5" x2="1100.58" y2="1283.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.58" y1="1267.5" x2="1132.58" y2="1283.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1116.58" y1="1267.5" x2="1132.58" y2="1251.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1435.37" y1="1059.7" x2="1419.37" y2="1043.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1435.37" y1="1059.7" x2="1419.37" y2="1075.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1435.37" y1="1059.7" x2="1451.37" y2="1075.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1435.37" y1="1059.7" x2="1451.37" y2="1043.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="930.593" y1="1251.3" x2="914.593" y2="1235.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="930.593" y1="1251.3" x2="914.593" y2="1267.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="930.593" y1="1251.3" x2="946.593" y2="1267.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="930.593" y1="1251.3" x2="946.593" y2="1235.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1125.99" y1="1256.74" x2="1109.99" y2="1240.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1125.99" y1="1256.74" x2="1109.99" y2="1272.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1125.99" y1="1256.74" x2="1141.99" y2="1272.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1125.99" y1="1256.74" x2="1141.99" y2="1240.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.3" y1="1267.31" x2="1105.3" y2="1251.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.3" y1="1267.31" x2="1105.3" y2="1283.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.3" y1="1267.31" x2="1137.3" y2="1283.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.3" y1="1267.31" x2="1137.3" y2="1251.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.55" y1="972.554" x2="1599.55" y2="956.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.55" y1="972.554" x2="1599.55" y2="988.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.55" y1="972.554" x2="1631.55" y2="988.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1615.55" y1="972.554" x2="1631.55" y2="956.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1395.64" y1="960.36" x2="1379.64" y2="944.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1395.64" y1="960.36" x2="1379.64" y2="976.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1395.64" y1="960.36" x2="1411.64" y2="976.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1395.64" y1="960.36" x2="1411.64" y2="944.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.08" y1="1312.36" x2="984.075" y2="1296.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.08" y1="1312.36" x2="984.075" y2="1328.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.08" y1="1312.36" x2="1016.08" y2="1328.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1000.08" y1="1312.36" x2="1016.08" y2="1296.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.75" y1="1068.03" x2="1337.75" y2="1052.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.75" y1="1068.03" x2="1337.75" y2="1084.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.75" y1="1068.03" x2="1369.75" y2="1084.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1353.75" y1="1068.03" x2="1369.75" y2="1052.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1126.95" y1="1234.81" x2="1110.95" y2="1218.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1126.95" y1="1234.81" x2="1110.95" y2="1250.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1126.95" y1="1234.81" x2="1142.95" y2="1250.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1126.95" y1="1234.81" x2="1142.95" y2="1218.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.72" y1="1159.69" x2="1347.72" y2="1143.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.72" y1="1159.69" x2="1347.72" y2="1175.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.72" y1="1159.69" x2="1379.72" y2="1175.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.72" y1="1159.69" x2="1379.72" y2="1143.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="805.605" y1="1368.59" x2="789.605" y2="1352.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="805.605" y1="1368.59" x2="789.605" y2="1384.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="805.605" y1="1368.59" x2="821.605" y2="1384.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="805.605" y1="1368.59" x2="821.605" y2="1352.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="735.913" y1="1350.58" x2="719.913" y2="1334.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="735.913" y1="1350.58" x2="719.913" y2="1366.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="735.913" y1="1350.58" x2="751.913" y2="1366.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="735.913" y1="1350.58" x2="751.913" y2="1334.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1057.78" y1="1211.94" x2="1041.78" y2="1195.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1057.78" y1="1211.94" x2="1041.78" y2="1227.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1057.78" y1="1211.94" x2="1073.78" y2="1227.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1057.78" y1="1211.94" x2="1073.78" y2="1195.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1826.24" y1="921.284" x2="1810.24" y2="905.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1826.24" y1="921.284" x2="1810.24" y2="937.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1826.24" y1="921.284" x2="1842.24" y2="937.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1826.24" y1="921.284" x2="1842.24" y2="905.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.2" y1="1094.53" x2="1277.2" y2="1078.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.2" y1="1094.53" x2="1277.2" y2="1110.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.2" y1="1094.53" x2="1309.2" y2="1110.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1293.2" y1="1094.53" x2="1309.2" y2="1078.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.65" y1="1361.41" x2="1091.65" y2="1345.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.65" y1="1361.41" x2="1091.65" y2="1377.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.65" y1="1361.41" x2="1123.65" y2="1377.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.65" y1="1361.41" x2="1123.65" y2="1345.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.07" y1="1240.9" x2="1160.07" y2="1224.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.07" y1="1240.9" x2="1160.07" y2="1256.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.07" y1="1240.9" x2="1192.07" y2="1256.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1176.07" y1="1240.9" x2="1192.07" y2="1224.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1215.2" y1="1190.81" x2="1199.2" y2="1174.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1215.2" y1="1190.81" x2="1199.2" y2="1206.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1215.2" y1="1190.81" x2="1231.2" y2="1206.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1215.2" y1="1190.81" x2="1231.2" y2="1174.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1572.67" y1="995.733" x2="1556.67" y2="979.733" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1572.67" y1="995.733" x2="1556.67" y2="1011.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1572.67" y1="995.733" x2="1588.67" y2="1011.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1572.67" y1="995.733" x2="1588.67" y2="979.733" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.41" y1="1231.66" x2="1348.41" y2="1215.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.41" y1="1231.66" x2="1348.41" y2="1247.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.41" y1="1231.66" x2="1380.41" y2="1247.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1364.41" y1="1231.66" x2="1380.41" y2="1215.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1259.03" y1="1156.39" x2="1243.03" y2="1140.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1259.03" y1="1156.39" x2="1243.03" y2="1172.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1259.03" y1="1156.39" x2="1275.03" y2="1172.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1259.03" y1="1156.39" x2="1275.03" y2="1140.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1469" y1="974.22" x2="1453" y2="958.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1469" y1="974.22" x2="1453" y2="990.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1469" y1="974.22" x2="1485" y2="990.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1469" y1="974.22" x2="1485" y2="958.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1292.53" y1="1131.39" x2="1276.53" y2="1115.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1292.53" y1="1131.39" x2="1276.53" y2="1147.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1292.53" y1="1131.39" x2="1308.53" y2="1147.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1292.53" y1="1131.39" x2="1308.53" y2="1115.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.04" y1="1299.49" x2="1130.04" y2="1283.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.04" y1="1299.49" x2="1130.04" y2="1315.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.04" y1="1299.49" x2="1162.04" y2="1315.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1146.04" y1="1299.49" x2="1162.04" y2="1283.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.06" y1="1219.15" x2="1070.06" y2="1203.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.06" y1="1219.15" x2="1070.06" y2="1235.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.06" y1="1219.15" x2="1102.06" y2="1235.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1086.06" y1="1219.15" x2="1102.06" y2="1203.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.64" y1="1000.17" x2="1752.64" y2="984.169" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.64" y1="1000.17" x2="1752.64" y2="1016.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.64" y1="1000.17" x2="1784.64" y2="1016.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1768.64" y1="1000.17" x2="1784.64" y2="984.169" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.35" y1="1164.06" x2="1349.35" y2="1148.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.35" y1="1164.06" x2="1349.35" y2="1180.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.35" y1="1164.06" x2="1381.35" y2="1180.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.35" y1="1164.06" x2="1381.35" y2="1148.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.6" y1="944.573" x2="1470.6" y2="928.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.6" y1="944.573" x2="1470.6" y2="960.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.6" y1="944.573" x2="1502.6" y2="960.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1486.6" y1="944.573" x2="1502.6" y2="928.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.16" y1="1210.82" x2="1257.16" y2="1194.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.16" y1="1210.82" x2="1257.16" y2="1226.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.16" y1="1210.82" x2="1289.16" y2="1226.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.16" y1="1210.82" x2="1289.16" y2="1194.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.92" y1="1149.92" x2="1246.92" y2="1133.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.92" y1="1149.92" x2="1246.92" y2="1165.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.92" y1="1149.92" x2="1278.92" y2="1165.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.92" y1="1149.92" x2="1278.92" y2="1133.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1122.29" y1="1146.61" x2="1106.29" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1122.29" y1="1146.61" x2="1106.29" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1122.29" y1="1146.61" x2="1138.29" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1122.29" y1="1146.61" x2="1138.29" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.41" y1="1171.23" x2="1275.41" y2="1155.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.41" y1="1171.23" x2="1275.41" y2="1187.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.41" y1="1171.23" x2="1307.41" y2="1187.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.41" y1="1171.23" x2="1307.41" y2="1155.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1448.29" y1="1128.98" x2="1432.29" y2="1112.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1448.29" y1="1128.98" x2="1432.29" y2="1144.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1448.29" y1="1128.98" x2="1464.29" y2="1144.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1448.29" y1="1128.98" x2="1464.29" y2="1112.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1130.44" y1="1149.23" x2="1114.44" y2="1133.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1130.44" y1="1149.23" x2="1114.44" y2="1165.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1130.44" y1="1149.23" x2="1146.44" y2="1165.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1130.44" y1="1149.23" x2="1146.44" y2="1133.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1039.51" y1="1295.52" x2="1023.51" y2="1279.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1039.51" y1="1295.52" x2="1023.51" y2="1311.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1039.51" y1="1295.52" x2="1055.51" y2="1311.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1039.51" y1="1295.52" x2="1055.51" y2="1279.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.81" y1="1008.74" x2="1427.81" y2="992.741" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.81" y1="1008.74" x2="1427.81" y2="1024.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.81" y1="1008.74" x2="1459.81" y2="1024.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1443.81" y1="1008.74" x2="1459.81" y2="992.741" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.99" y1="1097.56" x2="1529.99" y2="1081.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.99" y1="1097.56" x2="1529.99" y2="1113.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.99" y1="1097.56" x2="1561.99" y2="1113.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1545.99" y1="1097.56" x2="1561.99" y2="1081.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.83" y1="1094.72" x2="1343.83" y2="1078.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.83" y1="1094.72" x2="1343.83" y2="1110.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.83" y1="1094.72" x2="1375.83" y2="1110.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1359.83" y1="1094.72" x2="1375.83" y2="1078.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.3" y1="1374.86" x2="1183.3" y2="1358.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.3" y1="1374.86" x2="1183.3" y2="1390.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.3" y1="1374.86" x2="1215.3" y2="1390.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1199.3" y1="1374.86" x2="1215.3" y2="1358.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1133.36" y1="1257.88" x2="1117.36" y2="1241.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1133.36" y1="1257.88" x2="1117.36" y2="1273.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1133.36" y1="1257.88" x2="1149.36" y2="1273.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1133.36" y1="1257.88" x2="1149.36" y2="1241.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1628.36" y1="926.189" x2="1612.36" y2="910.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1628.36" y1="926.189" x2="1612.36" y2="942.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1628.36" y1="926.189" x2="1644.36" y2="942.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1628.36" y1="926.189" x2="1644.36" y2="910.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1073.35" y1="1320.7" x2="1057.35" y2="1304.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1073.35" y1="1320.7" x2="1057.35" y2="1336.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1073.35" y1="1320.7" x2="1089.35" y2="1336.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1073.35" y1="1320.7" x2="1089.35" y2="1304.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.88" y1="1332.35" x2="1075.88" y2="1316.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.88" y1="1332.35" x2="1075.88" y2="1348.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.88" y1="1332.35" x2="1107.88" y2="1348.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1091.88" y1="1332.35" x2="1107.88" y2="1316.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1017.09" y1="1323.45" x2="1001.09" y2="1307.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1017.09" y1="1323.45" x2="1001.09" y2="1339.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1017.09" y1="1323.45" x2="1033.09" y2="1339.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1017.09" y1="1323.45" x2="1033.09" y2="1307.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.938" y1="1534.93" x2="553.938" y2="1518.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.938" y1="1534.93" x2="553.938" y2="1550.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.938" y1="1534.93" x2="585.938" y2="1550.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="569.938" y1="1534.93" x2="585.938" y2="1518.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1231.45" y1="1166.31" x2="1215.45" y2="1150.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1231.45" y1="1166.31" x2="1215.45" y2="1182.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1231.45" y1="1166.31" x2="1247.45" y2="1182.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1231.45" y1="1166.31" x2="1247.45" y2="1150.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.218" y1="1483.05" x2="728.218" y2="1467.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.218" y1="1483.05" x2="728.218" y2="1499.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.218" y1="1483.05" x2="760.218" y2="1499.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="744.218" y1="1483.05" x2="760.218" y2="1467.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.81" y1="1264.02" x2="1169.81" y2="1248.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.81" y1="1264.02" x2="1169.81" y2="1280.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.81" y1="1264.02" x2="1201.81" y2="1280.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1185.81" y1="1264.02" x2="1201.81" y2="1248.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1713.33" y1="929.721" x2="1697.33" y2="913.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1713.33" y1="929.721" x2="1697.33" y2="945.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1713.33" y1="929.721" x2="1729.33" y2="945.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1713.33" y1="929.721" x2="1729.33" y2="913.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.62" y1="1058.71" x2="1292.62" y2="1042.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.62" y1="1058.71" x2="1292.62" y2="1074.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.62" y1="1058.71" x2="1324.62" y2="1074.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1308.62" y1="1058.71" x2="1324.62" y2="1042.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="610.516" y1="1568.23" x2="594.516" y2="1552.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="610.516" y1="1568.23" x2="594.516" y2="1584.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="610.516" y1="1568.23" x2="626.516" y2="1584.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="610.516" y1="1568.23" x2="626.516" y2="1552.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1702.63" y1="949.089" x2="1686.63" y2="933.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1702.63" y1="949.089" x2="1686.63" y2="965.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1702.63" y1="949.089" x2="1718.63" y2="965.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1702.63" y1="949.089" x2="1718.63" y2="933.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1102.82" y1="1200.06" x2="1086.82" y2="1184.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1102.82" y1="1200.06" x2="1086.82" y2="1216.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1102.82" y1="1200.06" x2="1118.82" y2="1216.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1102.82" y1="1200.06" x2="1118.82" y2="1184.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.22" y1="1125.36" x2="1444.22" y2="1109.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.22" y1="1125.36" x2="1444.22" y2="1141.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.22" y1="1125.36" x2="1476.22" y2="1141.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1460.22" y1="1125.36" x2="1476.22" y2="1109.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1242.85" y1="1068.4" x2="1226.85" y2="1052.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1242.85" y1="1068.4" x2="1226.85" y2="1084.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1242.85" y1="1068.4" x2="1258.85" y2="1084.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1242.85" y1="1068.4" x2="1258.85" y2="1052.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1180.26" y1="1392.12" x2="1164.26" y2="1376.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1180.26" y1="1392.12" x2="1164.26" y2="1408.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1180.26" y1="1392.12" x2="1196.26" y2="1408.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1180.26" y1="1392.12" x2="1196.26" y2="1376.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1799.12" y1="921.26" x2="1783.12" y2="905.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1799.12" y1="921.26" x2="1783.12" y2="937.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1799.12" y1="921.26" x2="1815.12" y2="937.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1799.12" y1="921.26" x2="1815.12" y2="905.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1198.35" y1="1213.97" x2="1182.35" y2="1197.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1198.35" y1="1213.97" x2="1182.35" y2="1229.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1198.35" y1="1213.97" x2="1214.35" y2="1229.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1198.35" y1="1213.97" x2="1214.35" y2="1197.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1314.16" y1="1231.53" x2="1298.16" y2="1215.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1314.16" y1="1231.53" x2="1298.16" y2="1247.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1314.16" y1="1231.53" x2="1330.16" y2="1247.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1314.16" y1="1231.53" x2="1330.16" y2="1215.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1298.91" y1="1128.6" x2="1282.91" y2="1112.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1298.91" y1="1128.6" x2="1282.91" y2="1144.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1298.91" y1="1128.6" x2="1314.91" y2="1144.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1298.91" y1="1128.6" x2="1314.91" y2="1112.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1539.91" y1="992.445" x2="1523.91" y2="976.445" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1539.91" y1="992.445" x2="1523.91" y2="1008.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1539.91" y1="992.445" x2="1555.91" y2="1008.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1539.91" y1="992.445" x2="1555.91" y2="976.445" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.49" y1="1165.67" x2="1192.49" y2="1149.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.49" y1="1165.67" x2="1192.49" y2="1181.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.49" y1="1165.67" x2="1224.49" y2="1181.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.49" y1="1165.67" x2="1224.49" y2="1149.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.4" y1="1164.76" x2="1275.4" y2="1148.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.4" y1="1164.76" x2="1275.4" y2="1180.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.4" y1="1164.76" x2="1307.4" y2="1180.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1291.4" y1="1164.76" x2="1307.4" y2="1148.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1509.7" y1="1053.86" x2="1493.7" y2="1037.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1509.7" y1="1053.86" x2="1493.7" y2="1069.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1509.7" y1="1053.86" x2="1525.7" y2="1069.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1509.7" y1="1053.86" x2="1525.7" y2="1037.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1586.87" y1="1027.39" x2="1570.87" y2="1011.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1586.87" y1="1027.39" x2="1570.87" y2="1043.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1586.87" y1="1027.39" x2="1602.87" y2="1043.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1586.87" y1="1027.39" x2="1602.87" y2="1011.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.89" y1="1264.65" x2="1095.89" y2="1248.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.89" y1="1264.65" x2="1095.89" y2="1280.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.89" y1="1264.65" x2="1127.89" y2="1280.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.89" y1="1264.65" x2="1127.89" y2="1248.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.824" y1="1370.17" x2="947.824" y2="1354.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.824" y1="1370.17" x2="947.824" y2="1386.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.824" y1="1370.17" x2="979.824" y2="1386.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.824" y1="1370.17" x2="979.824" y2="1354.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1269.64" y1="1053.76" x2="1253.64" y2="1037.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1269.64" y1="1053.76" x2="1253.64" y2="1069.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1269.64" y1="1053.76" x2="1285.64" y2="1069.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1269.64" y1="1053.76" x2="1285.64" y2="1037.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="694.471" y1="1399.51" x2="678.471" y2="1383.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="694.471" y1="1399.51" x2="678.471" y2="1415.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="694.471" y1="1399.51" x2="710.471" y2="1415.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="694.471" y1="1399.51" x2="710.471" y2="1383.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1318.45" y1="1163.46" x2="1302.45" y2="1147.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1318.45" y1="1163.46" x2="1302.45" y2="1179.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1318.45" y1="1163.46" x2="1334.45" y2="1179.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1318.45" y1="1163.46" x2="1334.45" y2="1147.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1386.71" y1="1019.77" x2="1370.71" y2="1003.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1386.71" y1="1019.77" x2="1370.71" y2="1035.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1386.71" y1="1019.77" x2="1402.71" y2="1035.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1386.71" y1="1019.77" x2="1402.71" y2="1003.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1547.26" y1="1096.34" x2="1531.26" y2="1080.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1547.26" y1="1096.34" x2="1531.26" y2="1112.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1547.26" y1="1096.34" x2="1563.26" y2="1112.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1547.26" y1="1096.34" x2="1563.26" y2="1080.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1054.12" y1="1246.84" x2="1038.12" y2="1230.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1054.12" y1="1246.84" x2="1038.12" y2="1262.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1054.12" y1="1246.84" x2="1070.12" y2="1262.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1054.12" y1="1246.84" x2="1070.12" y2="1230.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.9" y1="1186.98" x2="1257.9" y2="1170.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.9" y1="1186.98" x2="1257.9" y2="1202.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.9" y1="1186.98" x2="1289.9" y2="1202.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.9" y1="1186.98" x2="1289.9" y2="1170.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.32" y1="993.037" x2="1462.32" y2="977.037" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.32" y1="993.037" x2="1462.32" y2="1009.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.32" y1="993.037" x2="1494.32" y2="1009.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1478.32" y1="993.037" x2="1494.32" y2="977.037" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.95" y1="1279" x2="1264.95" y2="1263" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.95" y1="1279" x2="1264.95" y2="1295" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.95" y1="1279" x2="1296.95" y2="1295" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.95" y1="1279" x2="1296.95" y2="1263" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.65" y1="1130.42" x2="1307.65" y2="1114.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.65" y1="1130.42" x2="1307.65" y2="1146.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.65" y1="1130.42" x2="1339.65" y2="1146.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.65" y1="1130.42" x2="1339.65" y2="1114.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.8" y1="1126.84" x2="1265.8" y2="1110.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.8" y1="1126.84" x2="1265.8" y2="1142.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.8" y1="1126.84" x2="1297.8" y2="1142.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1281.8" y1="1126.84" x2="1297.8" y2="1110.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.44" y1="1134.89" x2="1318.44" y2="1118.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.44" y1="1134.89" x2="1318.44" y2="1150.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.44" y1="1134.89" x2="1350.44" y2="1150.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1334.44" y1="1134.89" x2="1350.44" y2="1118.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1266.08" y1="1197.16" x2="1250.08" y2="1181.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1266.08" y1="1197.16" x2="1250.08" y2="1213.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1266.08" y1="1197.16" x2="1282.08" y2="1213.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1266.08" y1="1197.16" x2="1282.08" y2="1181.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1459.46" y1="1078.54" x2="1443.46" y2="1062.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1459.46" y1="1078.54" x2="1443.46" y2="1094.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1459.46" y1="1078.54" x2="1475.46" y2="1094.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1459.46" y1="1078.54" x2="1475.46" y2="1062.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1650.12" y1="918.653" x2="1634.12" y2="902.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1650.12" y1="918.653" x2="1634.12" y2="934.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1650.12" y1="918.653" x2="1666.12" y2="934.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1650.12" y1="918.653" x2="1666.12" y2="902.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1061.4" y1="1273.43" x2="1045.4" y2="1257.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1061.4" y1="1273.43" x2="1045.4" y2="1289.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1061.4" y1="1273.43" x2="1077.4" y2="1289.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1061.4" y1="1273.43" x2="1077.4" y2="1257.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.92" y1="1166.84" x2="1214.92" y2="1150.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.92" y1="1166.84" x2="1214.92" y2="1182.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.92" y1="1166.84" x2="1246.92" y2="1182.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1230.92" y1="1166.84" x2="1246.92" y2="1150.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1860.89" y1="778.201" x2="1844.89" y2="762.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1860.89" y1="778.201" x2="1844.89" y2="794.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1860.89" y1="778.201" x2="1876.89" y2="794.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1860.89" y1="778.201" x2="1876.89" y2="762.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.65" y1="1264.26" x2="1224.65" y2="1248.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.65" y1="1264.26" x2="1224.65" y2="1280.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.65" y1="1264.26" x2="1256.65" y2="1280.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.65" y1="1264.26" x2="1256.65" y2="1248.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.89" y1="1136.28" x2="1264.89" y2="1120.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.89" y1="1136.28" x2="1264.89" y2="1152.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.89" y1="1136.28" x2="1296.89" y2="1152.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.89" y1="1136.28" x2="1296.89" y2="1120.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1444.37" y1="1001.52" x2="1428.37" y2="985.521" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1444.37" y1="1001.52" x2="1428.37" y2="1017.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1444.37" y1="1001.52" x2="1460.37" y2="1017.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1444.37" y1="1001.52" x2="1460.37" y2="985.521" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1455.79" y1="1193.99" x2="1439.79" y2="1177.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1455.79" y1="1193.99" x2="1439.79" y2="1209.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1455.79" y1="1193.99" x2="1471.79" y2="1209.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1455.79" y1="1193.99" x2="1471.79" y2="1177.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.41" y1="1124.9" x2="1105.41" y2="1108.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.41" y1="1124.9" x2="1105.41" y2="1140.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.41" y1="1124.9" x2="1137.41" y2="1140.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1121.41" y1="1124.9" x2="1137.41" y2="1108.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.75" y1="1012.66" x2="1412.75" y2="996.663" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.75" y1="1012.66" x2="1412.75" y2="1028.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.75" y1="1012.66" x2="1444.75" y2="1028.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.75" y1="1012.66" x2="1444.75" y2="996.663" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.98" y1="1162.81" x2="1121.98" y2="1146.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.98" y1="1162.81" x2="1121.98" y2="1178.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.98" y1="1162.81" x2="1153.98" y2="1178.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1137.98" y1="1162.81" x2="1153.98" y2="1146.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1896.85" y1="631.005" x2="1880.85" y2="615.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1896.85" y1="631.005" x2="1880.85" y2="647.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1896.85" y1="631.005" x2="1912.85" y2="647.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1896.85" y1="631.005" x2="1912.85" y2="615.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1863.03" y1="689.163" x2="1847.03" y2="673.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1863.03" y1="689.163" x2="1847.03" y2="705.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1863.03" y1="689.163" x2="1879.03" y2="705.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1863.03" y1="689.163" x2="1879.03" y2="673.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1492.82" y1="1052.48" x2="1476.82" y2="1036.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1492.82" y1="1052.48" x2="1476.82" y2="1068.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1492.82" y1="1052.48" x2="1508.82" y2="1068.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1492.82" y1="1052.48" x2="1508.82" y2="1036.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.58" y1="1162.34" x2="1228.58" y2="1146.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.58" y1="1162.34" x2="1228.58" y2="1178.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.58" y1="1162.34" x2="1260.58" y2="1178.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1244.58" y1="1162.34" x2="1260.58" y2="1146.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.53" y1="1236.06" x2="1307.53" y2="1220.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.53" y1="1236.06" x2="1307.53" y2="1252.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.53" y1="1236.06" x2="1339.53" y2="1252.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1323.53" y1="1236.06" x2="1339.53" y2="1220.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.45" y1="1106.65" x2="1369.45" y2="1090.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.45" y1="1106.65" x2="1369.45" y2="1122.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.45" y1="1106.65" x2="1401.45" y2="1122.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.45" y1="1106.65" x2="1401.45" y2="1090.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="973.556" y1="1249.11" x2="957.556" y2="1233.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="973.556" y1="1249.11" x2="957.556" y2="1265.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="973.556" y1="1249.11" x2="989.556" y2="1265.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="973.556" y1="1249.11" x2="989.556" y2="1233.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.42" y1="974.507" x2="1399.42" y2="958.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.42" y1="974.507" x2="1399.42" y2="990.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.42" y1="974.507" x2="1431.42" y2="990.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1415.42" y1="974.507" x2="1431.42" y2="958.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.69" y1="1084.02" x2="1436.69" y2="1068.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.69" y1="1084.02" x2="1436.69" y2="1100.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.69" y1="1084.02" x2="1468.69" y2="1100.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1452.69" y1="1084.02" x2="1468.69" y2="1068.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1114.28" y1="1270.08" x2="1098.28" y2="1254.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1114.28" y1="1270.08" x2="1098.28" y2="1286.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1114.28" y1="1270.08" x2="1130.28" y2="1286.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1114.28" y1="1270.08" x2="1130.28" y2="1254.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.1" y1="1367.01" x2="1264.1" y2="1351.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.1" y1="1367.01" x2="1264.1" y2="1383.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.1" y1="1367.01" x2="1296.1" y2="1383.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1280.1" y1="1367.01" x2="1296.1" y2="1351.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="961.172" y1="1297.9" x2="945.172" y2="1281.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="961.172" y1="1297.9" x2="945.172" y2="1313.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="961.172" y1="1297.9" x2="977.172" y2="1313.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="961.172" y1="1297.9" x2="977.172" y2="1281.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="828.067" y1="1395.72" x2="812.067" y2="1379.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="828.067" y1="1395.72" x2="812.067" y2="1411.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="828.067" y1="1395.72" x2="844.067" y2="1411.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="828.067" y1="1395.72" x2="844.067" y2="1379.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="812.031" y1="1484.71" x2="796.031" y2="1468.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="812.031" y1="1484.71" x2="796.031" y2="1500.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="812.031" y1="1484.71" x2="828.031" y2="1500.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="812.031" y1="1484.71" x2="828.031" y2="1468.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="901.731" y1="1295.1" x2="885.731" y2="1279.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="901.731" y1="1295.1" x2="885.731" y2="1311.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="901.731" y1="1295.1" x2="917.731" y2="1311.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="901.731" y1="1295.1" x2="917.731" y2="1279.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1331.96" y1="1041.81" x2="1315.96" y2="1025.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1331.96" y1="1041.81" x2="1315.96" y2="1057.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1331.96" y1="1041.81" x2="1347.96" y2="1057.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1331.96" y1="1041.81" x2="1347.96" y2="1025.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.94" y1="1077.84" x2="1347.94" y2="1061.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.94" y1="1077.84" x2="1347.94" y2="1093.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.94" y1="1077.84" x2="1379.94" y2="1093.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1363.94" y1="1077.84" x2="1379.94" y2="1061.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.853" y1="1331.79" x2="956.853" y2="1315.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.853" y1="1331.79" x2="956.853" y2="1347.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.853" y1="1331.79" x2="988.853" y2="1347.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="972.853" y1="1331.79" x2="988.853" y2="1315.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.97" y1="912.119" x2="1722.97" y2="896.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.97" y1="912.119" x2="1722.97" y2="928.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.97" y1="912.119" x2="1754.97" y2="928.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.97" y1="912.119" x2="1754.97" y2="896.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1468.77" y1="1104.3" x2="1452.77" y2="1088.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1468.77" y1="1104.3" x2="1452.77" y2="1120.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1468.77" y1="1104.3" x2="1484.77" y2="1120.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1468.77" y1="1104.3" x2="1484.77" y2="1088.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.63" y1="1168.13" x2="1036.63" y2="1152.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.63" y1="1168.13" x2="1036.63" y2="1184.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.63" y1="1168.13" x2="1068.63" y2="1184.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.63" y1="1168.13" x2="1068.63" y2="1152.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.71" y1="1040.19" x2="1386.71" y2="1024.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.71" y1="1040.19" x2="1386.71" y2="1056.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.71" y1="1040.19" x2="1418.71" y2="1056.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.71" y1="1040.19" x2="1418.71" y2="1024.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1370.93" y1="1101.26" x2="1354.93" y2="1085.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1370.93" y1="1101.26" x2="1354.93" y2="1117.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1370.93" y1="1101.26" x2="1386.93" y2="1117.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1370.93" y1="1101.26" x2="1386.93" y2="1085.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1534.29" y1="956.26" x2="1518.29" y2="940.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1534.29" y1="956.26" x2="1518.29" y2="972.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1534.29" y1="956.26" x2="1550.29" y2="972.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1534.29" y1="956.26" x2="1550.29" y2="940.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1470.32" y1="1144.8" x2="1454.32" y2="1128.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1470.32" y1="1144.8" x2="1454.32" y2="1160.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1470.32" y1="1144.8" x2="1486.32" y2="1160.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1470.32" y1="1144.8" x2="1486.32" y2="1128.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.177" y1="1437.3" x2="863.177" y2="1421.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.177" y1="1437.3" x2="863.177" y2="1453.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.177" y1="1437.3" x2="895.177" y2="1453.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.177" y1="1437.3" x2="895.177" y2="1421.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.308" y1="1302.19" x2="775.308" y2="1286.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.308" y1="1302.19" x2="775.308" y2="1318.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.308" y1="1302.19" x2="807.308" y2="1318.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="791.308" y1="1302.19" x2="807.308" y2="1286.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1182.59" y1="1128.11" x2="1166.59" y2="1112.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1182.59" y1="1128.11" x2="1166.59" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1182.59" y1="1128.11" x2="1198.59" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1182.59" y1="1128.11" x2="1198.59" y2="1112.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.44" y1="890.561" x2="1574.44" y2="874.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.44" y1="890.561" x2="1574.44" y2="906.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.44" y1="890.561" x2="1606.44" y2="906.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.44" y1="890.561" x2="1606.44" y2="874.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1083.01" y1="1250.31" x2="1067.01" y2="1234.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1083.01" y1="1250.31" x2="1067.01" y2="1266.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1083.01" y1="1250.31" x2="1099.01" y2="1266.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1083.01" y1="1250.31" x2="1099.01" y2="1234.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1729.02" y1="853.811" x2="1713.02" y2="837.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1729.02" y1="853.811" x2="1713.02" y2="869.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1729.02" y1="853.811" x2="1745.02" y2="869.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1729.02" y1="853.811" x2="1745.02" y2="837.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="936.651" y1="1397.2" x2="920.651" y2="1381.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="936.651" y1="1397.2" x2="920.651" y2="1413.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="936.651" y1="1397.2" x2="952.651" y2="1413.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="936.651" y1="1397.2" x2="952.651" y2="1381.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1141.39" y1="1273.55" x2="1125.39" y2="1257.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1141.39" y1="1273.55" x2="1125.39" y2="1289.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1141.39" y1="1273.55" x2="1157.39" y2="1289.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1141.39" y1="1273.55" x2="1157.39" y2="1257.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.12" y1="1220.55" x2="1050.12" y2="1204.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.12" y1="1220.55" x2="1050.12" y2="1236.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.12" y1="1220.55" x2="1082.12" y2="1236.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1066.12" y1="1220.55" x2="1082.12" y2="1204.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="856.289" y1="1386.44" x2="840.289" y2="1370.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="856.289" y1="1386.44" x2="840.289" y2="1402.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="856.289" y1="1386.44" x2="872.289" y2="1402.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="856.289" y1="1386.44" x2="872.289" y2="1370.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1450.3" y1="1197.55" x2="1434.3" y2="1181.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1450.3" y1="1197.55" x2="1434.3" y2="1213.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1450.3" y1="1197.55" x2="1466.3" y2="1213.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1450.3" y1="1197.55" x2="1466.3" y2="1181.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1350.38" y1="930.437" x2="1334.38" y2="914.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1350.38" y1="930.437" x2="1334.38" y2="946.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1350.38" y1="930.437" x2="1366.38" y2="946.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1350.38" y1="930.437" x2="1366.38" y2="914.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1529.58" y1="996.113" x2="1513.58" y2="980.113" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1529.58" y1="996.113" x2="1513.58" y2="1012.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1529.58" y1="996.113" x2="1545.58" y2="1012.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1529.58" y1="996.113" x2="1545.58" y2="980.113" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.87" y1="1216.87" x2="1036.87" y2="1200.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.87" y1="1216.87" x2="1036.87" y2="1232.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.87" y1="1216.87" x2="1068.87" y2="1232.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1052.87" y1="1216.87" x2="1068.87" y2="1200.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.28" y1="1112.64" x2="1209.28" y2="1096.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.28" y1="1112.64" x2="1209.28" y2="1128.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.28" y1="1112.64" x2="1241.28" y2="1128.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1225.28" y1="1112.64" x2="1241.28" y2="1096.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.47" y1="955.973" x2="1497.47" y2="939.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.47" y1="955.973" x2="1497.47" y2="971.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.47" y1="955.973" x2="1529.47" y2="971.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.47" y1="955.973" x2="1529.47" y2="939.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.93" y1="1082.34" x2="1458.93" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.93" y1="1082.34" x2="1458.93" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.93" y1="1082.34" x2="1490.93" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1474.93" y1="1082.34" x2="1490.93" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1530.22" y1="1008.88" x2="1514.22" y2="992.884" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1530.22" y1="1008.88" x2="1514.22" y2="1024.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1530.22" y1="1008.88" x2="1546.22" y2="1024.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1530.22" y1="1008.88" x2="1546.22" y2="992.884" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.75" y1="828.151" x2="1574.75" y2="812.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.75" y1="828.151" x2="1574.75" y2="844.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.75" y1="828.151" x2="1606.75" y2="844.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1590.75" y1="828.151" x2="1606.75" y2="812.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.49" y1="1238.26" x2="1151.49" y2="1222.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.49" y1="1238.26" x2="1151.49" y2="1254.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.49" y1="1238.26" x2="1183.49" y2="1254.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1167.49" y1="1238.26" x2="1183.49" y2="1222.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.11" y1="1021.15" x2="1328.11" y2="1005.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.11" y1="1021.15" x2="1328.11" y2="1037.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.11" y1="1021.15" x2="1360.11" y2="1037.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1344.11" y1="1021.15" x2="1360.11" y2="1005.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1942.19" y1="883.328" x2="1926.19" y2="867.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1942.19" y1="883.328" x2="1926.19" y2="899.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1942.19" y1="883.328" x2="1958.19" y2="899.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1942.19" y1="883.328" x2="1958.19" y2="867.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="800.484" y1="1327.86" x2="784.484" y2="1311.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="800.484" y1="1327.86" x2="784.484" y2="1343.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="800.484" y1="1327.86" x2="816.484" y2="1343.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="800.484" y1="1327.86" x2="816.484" y2="1311.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.479" y1="1361.21" x2="863.479" y2="1345.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.479" y1="1361.21" x2="863.479" y2="1377.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.479" y1="1361.21" x2="895.479" y2="1377.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="879.479" y1="1361.21" x2="895.479" y2="1345.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1621.59" y1="888.584" x2="1605.59" y2="872.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1621.59" y1="888.584" x2="1605.59" y2="904.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1621.59" y1="888.584" x2="1637.59" y2="904.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1621.59" y1="888.584" x2="1637.59" y2="872.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1351.96" y1="1166.63" x2="1335.96" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1351.96" y1="1166.63" x2="1335.96" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1351.96" y1="1166.63" x2="1367.96" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1351.96" y1="1166.63" x2="1367.96" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.03" y1="917.837" x2="1589.03" y2="901.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.03" y1="917.837" x2="1589.03" y2="933.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.03" y1="917.837" x2="1621.03" y2="933.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1605.03" y1="917.837" x2="1621.03" y2="901.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.5" y1="1173.02" x2="1238.5" y2="1157.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.5" y1="1173.02" x2="1238.5" y2="1189.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.5" y1="1173.02" x2="1270.5" y2="1189.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.5" y1="1173.02" x2="1270.5" y2="1157.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.95" y1="1137.6" x2="1509.95" y2="1121.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.95" y1="1137.6" x2="1509.95" y2="1153.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.95" y1="1137.6" x2="1541.95" y2="1153.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1525.95" y1="1137.6" x2="1541.95" y2="1121.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="863.983" y1="1256.2" x2="847.983" y2="1240.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="863.983" y1="1256.2" x2="847.983" y2="1272.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="863.983" y1="1256.2" x2="879.983" y2="1272.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="863.983" y1="1256.2" x2="879.983" y2="1240.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1763.1" y1="789.55" x2="1747.1" y2="773.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1763.1" y1="789.55" x2="1747.1" y2="805.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1763.1" y1="789.55" x2="1779.1" y2="805.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1763.1" y1="789.55" x2="1779.1" y2="773.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.65" y1="1335.83" x2="1149.65" y2="1319.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.65" y1="1335.83" x2="1149.65" y2="1351.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.65" y1="1335.83" x2="1181.65" y2="1351.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.65" y1="1335.83" x2="1181.65" y2="1319.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.21" y1="1265.21" x2="1091.21" y2="1249.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.21" y1="1265.21" x2="1091.21" y2="1281.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.21" y1="1265.21" x2="1123.21" y2="1281.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1107.21" y1="1265.21" x2="1123.21" y2="1249.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1261.91" y1="1136.73" x2="1245.91" y2="1120.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1261.91" y1="1136.73" x2="1245.91" y2="1152.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1261.91" y1="1136.73" x2="1277.91" y2="1152.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1261.91" y1="1136.73" x2="1277.91" y2="1120.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1840.32" y1="839.458" x2="1824.32" y2="823.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1840.32" y1="839.458" x2="1824.32" y2="855.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1840.32" y1="839.458" x2="1856.32" y2="855.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1840.32" y1="839.458" x2="1856.32" y2="823.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1566.73" y1="937.472" x2="1550.73" y2="921.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1566.73" y1="937.472" x2="1550.73" y2="953.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1566.73" y1="937.472" x2="1582.73" y2="953.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1566.73" y1="937.472" x2="1582.73" y2="921.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.17" y1="997.851" x2="1441.17" y2="981.851" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.17" y1="997.851" x2="1441.17" y2="1013.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.17" y1="997.851" x2="1473.17" y2="1013.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.17" y1="997.851" x2="1473.17" y2="981.851" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1518.09" y1="1097.6" x2="1502.09" y2="1081.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1518.09" y1="1097.6" x2="1502.09" y2="1113.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1518.09" y1="1097.6" x2="1534.09" y2="1113.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1518.09" y1="1097.6" x2="1534.09" y2="1081.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.74" y1="1166.18" x2="1246.74" y2="1150.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.74" y1="1166.18" x2="1246.74" y2="1182.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.74" y1="1166.18" x2="1278.74" y2="1182.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1262.74" y1="1166.18" x2="1278.74" y2="1150.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="725.698" y1="1501.31" x2="709.698" y2="1485.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="725.698" y1="1501.31" x2="709.698" y2="1517.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="725.698" y1="1501.31" x2="741.698" y2="1517.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="725.698" y1="1501.31" x2="741.698" y2="1485.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="804.197" y1="1345.85" x2="788.197" y2="1329.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="804.197" y1="1345.85" x2="788.197" y2="1361.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="804.197" y1="1345.85" x2="820.197" y2="1361.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="804.197" y1="1345.85" x2="820.197" y2="1329.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1068.71" y1="1367.71" x2="1052.71" y2="1351.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1068.71" y1="1367.71" x2="1052.71" y2="1383.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1068.71" y1="1367.71" x2="1084.71" y2="1383.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1068.71" y1="1367.71" x2="1084.71" y2="1351.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.44" y1="832.688" x2="1722.44" y2="816.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.44" y1="832.688" x2="1722.44" y2="848.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.44" y1="832.688" x2="1754.44" y2="848.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1738.44" y1="832.688" x2="1754.44" y2="816.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1575.07" y1="927.362" x2="1559.07" y2="911.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1575.07" y1="927.362" x2="1559.07" y2="943.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1575.07" y1="927.362" x2="1591.07" y2="943.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1575.07" y1="927.362" x2="1591.07" y2="911.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.369" y1="1393.26" x2="947.369" y2="1377.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.369" y1="1393.26" x2="947.369" y2="1409.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.369" y1="1393.26" x2="979.369" y2="1409.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.369" y1="1393.26" x2="979.369" y2="1377.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1338.29" y1="1101.82" x2="1322.29" y2="1085.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1338.29" y1="1101.82" x2="1322.29" y2="1117.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1338.29" y1="1101.82" x2="1354.29" y2="1117.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1338.29" y1="1101.82" x2="1354.29" y2="1085.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.77" y1="1157.2" x2="1326.77" y2="1141.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.77" y1="1157.2" x2="1326.77" y2="1173.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.77" y1="1157.2" x2="1358.77" y2="1173.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.77" y1="1157.2" x2="1358.77" y2="1141.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1465.22" y1="1059.54" x2="1449.22" y2="1043.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1465.22" y1="1059.54" x2="1449.22" y2="1075.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1465.22" y1="1059.54" x2="1481.22" y2="1075.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1465.22" y1="1059.54" x2="1481.22" y2="1043.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1667.21" y1="967.122" x2="1651.21" y2="951.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1667.21" y1="967.122" x2="1651.21" y2="983.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1667.21" y1="967.122" x2="1683.21" y2="983.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1667.21" y1="967.122" x2="1683.21" y2="951.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1808.12" y1="803.64" x2="1792.12" y2="787.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1808.12" y1="803.64" x2="1792.12" y2="819.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1808.12" y1="803.64" x2="1824.12" y2="819.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1808.12" y1="803.64" x2="1824.12" y2="787.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.84" y1="1113.91" x2="1349.84" y2="1097.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.84" y1="1113.91" x2="1349.84" y2="1129.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.84" y1="1113.91" x2="1381.84" y2="1129.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1365.84" y1="1113.91" x2="1381.84" y2="1097.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.457" y1="1364.93" x2="911.457" y2="1348.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.457" y1="1364.93" x2="911.457" y2="1380.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.457" y1="1364.93" x2="943.457" y2="1380.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="927.457" y1="1364.93" x2="943.457" y2="1348.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.588" y1="1394.79" x2="938.588" y2="1378.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.588" y1="1394.79" x2="938.588" y2="1410.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.588" y1="1394.79" x2="970.588" y2="1410.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.588" y1="1394.79" x2="970.588" y2="1378.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1600.82" y1="848.267" x2="1584.82" y2="832.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1600.82" y1="848.267" x2="1584.82" y2="864.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1600.82" y1="848.267" x2="1616.82" y2="864.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1600.82" y1="848.267" x2="1616.82" y2="832.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1013.05" y1="1254.21" x2="997.055" y2="1238.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1013.05" y1="1254.21" x2="997.055" y2="1270.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1013.05" y1="1254.21" x2="1029.05" y2="1270.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1013.05" y1="1254.21" x2="1029.05" y2="1238.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.35" y1="1098.11" x2="1147.35" y2="1082.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.35" y1="1098.11" x2="1147.35" y2="1114.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.35" y1="1098.11" x2="1179.35" y2="1114.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1163.35" y1="1098.11" x2="1179.35" y2="1082.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1400.1" y1="1101.9" x2="1384.1" y2="1085.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1400.1" y1="1101.9" x2="1384.1" y2="1117.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1400.1" y1="1101.9" x2="1416.1" y2="1117.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1400.1" y1="1101.9" x2="1416.1" y2="1085.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="710.768" y1="1471.31" x2="694.768" y2="1455.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="710.768" y1="1471.31" x2="694.768" y2="1487.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="710.768" y1="1471.31" x2="726.768" y2="1487.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="710.768" y1="1471.31" x2="726.768" y2="1455.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.455" y1="1241.66" x2="938.455" y2="1225.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.455" y1="1241.66" x2="938.455" y2="1257.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.455" y1="1241.66" x2="970.455" y2="1257.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="954.455" y1="1241.66" x2="970.455" y2="1225.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.63" y1="1188.99" x2="1081.63" y2="1172.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.63" y1="1188.99" x2="1081.63" y2="1204.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.63" y1="1188.99" x2="1113.63" y2="1204.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1097.63" y1="1188.99" x2="1113.63" y2="1172.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.96" y1="1069.03" x2="1369.96" y2="1053.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.96" y1="1069.03" x2="1369.96" y2="1085.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.96" y1="1069.03" x2="1401.96" y2="1085.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1385.96" y1="1069.03" x2="1401.96" y2="1053.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.19" y1="997.572" x2="1412.19" y2="981.572" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.19" y1="997.572" x2="1412.19" y2="1013.57" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.19" y1="997.572" x2="1444.19" y2="1013.57" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1428.19" y1="997.572" x2="1444.19" y2="981.572" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1023.75" y1="1232.23" x2="1007.75" y2="1216.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1023.75" y1="1232.23" x2="1007.75" y2="1248.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1023.75" y1="1232.23" x2="1039.75" y2="1248.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1023.75" y1="1232.23" x2="1039.75" y2="1216.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1336.01" y1="1021.11" x2="1320.01" y2="1005.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1336.01" y1="1021.11" x2="1320.01" y2="1037.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1336.01" y1="1021.11" x2="1352.01" y2="1037.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1336.01" y1="1021.11" x2="1352.01" y2="1005.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.06" y1="1198.05" x2="1326.06" y2="1182.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.06" y1="1198.05" x2="1326.06" y2="1214.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.06" y1="1198.05" x2="1358.06" y2="1214.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1342.06" y1="1198.05" x2="1358.06" y2="1182.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1401.11" y1="1144.76" x2="1385.11" y2="1128.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1401.11" y1="1144.76" x2="1385.11" y2="1160.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1401.11" y1="1144.76" x2="1417.11" y2="1160.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1401.11" y1="1144.76" x2="1417.11" y2="1128.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.53" y1="1066.19" x2="1386.53" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.53" y1="1066.19" x2="1386.53" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.53" y1="1066.19" x2="1418.53" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1402.53" y1="1066.19" x2="1418.53" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1117.37" y1="1261.72" x2="1101.37" y2="1245.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1117.37" y1="1261.72" x2="1101.37" y2="1277.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1117.37" y1="1261.72" x2="1133.37" y2="1277.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1117.37" y1="1261.72" x2="1133.37" y2="1245.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.68" y1="1081.7" x2="1415.68" y2="1065.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.68" y1="1081.7" x2="1415.68" y2="1097.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.68" y1="1081.7" x2="1447.68" y2="1097.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1431.68" y1="1081.7" x2="1447.68" y2="1065.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="907.609" y1="1313.11" x2="891.609" y2="1297.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="907.609" y1="1313.11" x2="891.609" y2="1329.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="907.609" y1="1313.11" x2="923.609" y2="1329.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="907.609" y1="1313.11" x2="923.609" y2="1297.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1267.73" y1="1057.77" x2="1251.73" y2="1041.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1267.73" y1="1057.77" x2="1251.73" y2="1073.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1267.73" y1="1057.77" x2="1283.73" y2="1073.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1267.73" y1="1057.77" x2="1283.73" y2="1041.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="904.55" y1="1372.25" x2="888.55" y2="1356.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="904.55" y1="1372.25" x2="888.55" y2="1388.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="904.55" y1="1372.25" x2="920.55" y2="1388.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="904.55" y1="1372.25" x2="920.55" y2="1356.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.58" y1="1084.87" x2="1296.58" y2="1068.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.58" y1="1084.87" x2="1296.58" y2="1100.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.58" y1="1084.87" x2="1328.58" y2="1100.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1312.58" y1="1084.87" x2="1328.58" y2="1068.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1134.2" y1="1238.48" x2="1118.2" y2="1222.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1134.2" y1="1238.48" x2="1118.2" y2="1254.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1134.2" y1="1238.48" x2="1150.2" y2="1254.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1134.2" y1="1238.48" x2="1150.2" y2="1222.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1561.34" y1="914.945" x2="1545.34" y2="898.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1561.34" y1="914.945" x2="1545.34" y2="930.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1561.34" y1="914.945" x2="1577.34" y2="930.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1561.34" y1="914.945" x2="1577.34" y2="898.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1391.28" y1="1184.21" x2="1375.28" y2="1168.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1391.28" y1="1184.21" x2="1375.28" y2="1200.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1391.28" y1="1184.21" x2="1407.28" y2="1200.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1391.28" y1="1184.21" x2="1407.28" y2="1168.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1311.21" y1="973.579" x2="1295.21" y2="957.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1311.21" y1="973.579" x2="1295.21" y2="989.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1311.21" y1="973.579" x2="1327.21" y2="989.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1311.21" y1="973.579" x2="1327.21" y2="957.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.94" y1="1419.74" x2="1078.94" y2="1403.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.94" y1="1419.74" x2="1078.94" y2="1435.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.94" y1="1419.74" x2="1110.94" y2="1435.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1094.94" y1="1419.74" x2="1110.94" y2="1403.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1143.74" y1="1146.65" x2="1127.74" y2="1130.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1143.74" y1="1146.65" x2="1127.74" y2="1162.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1143.74" y1="1146.65" x2="1159.74" y2="1162.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1143.74" y1="1146.65" x2="1159.74" y2="1130.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="743.908" y1="1344.73" x2="727.908" y2="1328.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="743.908" y1="1344.73" x2="727.908" y2="1360.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="743.908" y1="1344.73" x2="759.908" y2="1360.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="743.908" y1="1344.73" x2="759.908" y2="1328.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.98" y1="1226.45" x2="1099.98" y2="1210.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.98" y1="1226.45" x2="1099.98" y2="1242.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.98" y1="1226.45" x2="1131.98" y2="1242.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1115.98" y1="1226.45" x2="1131.98" y2="1210.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1724.4" y1="1054.35" x2="1708.4" y2="1038.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1724.4" y1="1054.35" x2="1708.4" y2="1070.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1724.4" y1="1054.35" x2="1740.4" y2="1070.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1724.4" y1="1054.35" x2="1740.4" y2="1038.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.408" y1="1314.39" x2="947.408" y2="1298.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.408" y1="1314.39" x2="947.408" y2="1330.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.408" y1="1314.39" x2="979.408" y2="1330.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="963.408" y1="1314.39" x2="979.408" y2="1298.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.46" y1="1140.38" x2="1238.46" y2="1124.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.46" y1="1140.38" x2="1238.46" y2="1156.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.46" y1="1140.38" x2="1270.46" y2="1156.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1254.46" y1="1140.38" x2="1270.46" y2="1124.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="994.563" y1="1220.85" x2="978.563" y2="1204.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="994.563" y1="1220.85" x2="978.563" y2="1236.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="994.563" y1="1220.85" x2="1010.56" y2="1236.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="994.563" y1="1220.85" x2="1010.56" y2="1204.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1585.12" y1="1061.03" x2="1569.12" y2="1045.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1585.12" y1="1061.03" x2="1569.12" y2="1077.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1585.12" y1="1061.03" x2="1601.12" y2="1077.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1585.12" y1="1061.03" x2="1601.12" y2="1045.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.07" y1="993.346" x2="1764.07" y2="977.346" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.07" y1="993.346" x2="1764.07" y2="1009.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.07" y1="993.346" x2="1796.07" y2="1009.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1780.07" y1="993.346" x2="1796.07" y2="977.346" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.71" y1="1150.44" x2="1192.71" y2="1134.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.71" y1="1150.44" x2="1192.71" y2="1166.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.71" y1="1150.44" x2="1224.71" y2="1166.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.71" y1="1150.44" x2="1224.71" y2="1134.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.86" y1="1171.34" x2="1355.86" y2="1155.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.86" y1="1171.34" x2="1355.86" y2="1187.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.86" y1="1171.34" x2="1387.86" y2="1187.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.86" y1="1171.34" x2="1387.86" y2="1155.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.82" y1="1306.89" x2="1094.82" y2="1290.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.82" y1="1306.89" x2="1094.82" y2="1322.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.82" y1="1306.89" x2="1126.82" y2="1322.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.82" y1="1306.89" x2="1126.82" y2="1290.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.33" y1="1287.85" x2="1192.33" y2="1271.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.33" y1="1287.85" x2="1192.33" y2="1303.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.33" y1="1287.85" x2="1224.33" y2="1303.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1208.33" y1="1287.85" x2="1224.33" y2="1271.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.49" y1="1169.31" x2="1149.49" y2="1153.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.49" y1="1169.31" x2="1149.49" y2="1185.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.49" y1="1169.31" x2="1181.49" y2="1185.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165.49" y1="1169.31" x2="1181.49" y2="1153.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.71" y1="1125.59" x2="1257.71" y2="1109.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.71" y1="1125.59" x2="1257.71" y2="1141.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.71" y1="1125.59" x2="1289.71" y2="1141.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1273.71" y1="1125.59" x2="1289.71" y2="1109.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1079.45" y1="1149.7" x2="1063.45" y2="1133.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1079.45" y1="1149.7" x2="1063.45" y2="1165.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1079.45" y1="1149.7" x2="1095.45" y2="1165.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1079.45" y1="1149.7" x2="1095.45" y2="1133.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1184.12" y1="1234.64" x2="1168.12" y2="1218.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1184.12" y1="1234.64" x2="1168.12" y2="1250.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1184.12" y1="1234.64" x2="1200.12" y2="1250.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1184.12" y1="1234.64" x2="1200.12" y2="1218.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.3" y1="1156.32" x2="1355.3" y2="1140.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.3" y1="1156.32" x2="1355.3" y2="1172.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.3" y1="1156.32" x2="1387.3" y2="1172.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1371.3" y1="1156.32" x2="1387.3" y2="1140.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.95" y1="1258.36" x2="1221.95" y2="1242.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.95" y1="1258.36" x2="1221.95" y2="1274.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.95" y1="1258.36" x2="1253.95" y2="1274.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1237.95" y1="1258.36" x2="1253.95" y2="1242.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="983.358" y1="1291.75" x2="967.358" y2="1275.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="983.358" y1="1291.75" x2="967.358" y2="1307.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="983.358" y1="1291.75" x2="999.358" y2="1307.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="983.358" y1="1291.75" x2="999.358" y2="1275.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1156.87" y1="1192.88" x2="1140.87" y2="1176.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1156.87" y1="1192.88" x2="1140.87" y2="1208.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1156.87" y1="1192.88" x2="1172.87" y2="1208.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1156.87" y1="1192.88" x2="1172.87" y2="1176.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1496.43" y1="924.203" x2="1480.43" y2="908.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1496.43" y1="924.203" x2="1480.43" y2="940.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1496.43" y1="924.203" x2="1512.43" y2="940.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1496.43" y1="924.203" x2="1512.43" y2="908.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="938.137" y1="1395.53" x2="922.137" y2="1379.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="938.137" y1="1395.53" x2="922.137" y2="1411.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="938.137" y1="1395.53" x2="954.137" y2="1411.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="938.137" y1="1395.53" x2="954.137" y2="1379.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="745.679" y1="1455.46" x2="729.679" y2="1439.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="745.679" y1="1455.46" x2="729.679" y2="1471.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="745.679" y1="1455.46" x2="761.679" y2="1471.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="745.679" y1="1455.46" x2="761.679" y2="1439.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1495.35" y1="1033.28" x2="1479.35" y2="1017.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1495.35" y1="1033.28" x2="1479.35" y2="1049.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1495.35" y1="1033.28" x2="1511.35" y2="1049.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1495.35" y1="1033.28" x2="1511.35" y2="1017.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1226.7" y1="1172" x2="1210.7" y2="1156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1226.7" y1="1172" x2="1210.7" y2="1188" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1226.7" y1="1172" x2="1242.7" y2="1188" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1226.7" y1="1172" x2="1242.7" y2="1156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.77" y1="884.258" x2="1746.77" y2="868.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.77" y1="884.258" x2="1746.77" y2="900.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.77" y1="884.258" x2="1778.77" y2="900.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1762.77" y1="884.258" x2="1778.77" y2="868.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="977.334" y1="1349.1" x2="961.334" y2="1333.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="977.334" y1="1349.1" x2="961.334" y2="1365.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="977.334" y1="1349.1" x2="993.334" y2="1365.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="977.334" y1="1349.1" x2="993.334" y2="1333.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1274.34" y1="1100.11" x2="1258.34" y2="1084.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1274.34" y1="1100.11" x2="1258.34" y2="1116.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1274.34" y1="1100.11" x2="1290.34" y2="1116.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1274.34" y1="1100.11" x2="1290.34" y2="1084.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.89" y1="1051.48" x2="1497.89" y2="1035.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.89" y1="1051.48" x2="1497.89" y2="1067.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.89" y1="1051.48" x2="1529.89" y2="1067.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1513.89" y1="1051.48" x2="1529.89" y2="1035.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.49" y1="1113.18" x2="1441.49" y2="1097.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.49" y1="1113.18" x2="1441.49" y2="1129.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.49" y1="1113.18" x2="1473.49" y2="1129.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1457.49" y1="1113.18" x2="1473.49" y2="1097.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.94" y1="1156.93" x2="1405.94" y2="1140.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.94" y1="1156.93" x2="1405.94" y2="1172.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.94" y1="1156.93" x2="1437.94" y2="1172.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1421.94" y1="1156.93" x2="1437.94" y2="1140.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1638.84" y1="1106.16" x2="1622.84" y2="1090.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1638.84" y1="1106.16" x2="1622.84" y2="1122.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1638.84" y1="1106.16" x2="1654.84" y2="1122.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1638.84" y1="1106.16" x2="1654.84" y2="1090.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.292" y1="1482.58" x2="859.292" y2="1466.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.292" y1="1482.58" x2="859.292" y2="1498.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.292" y1="1482.58" x2="891.292" y2="1498.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="875.292" y1="1482.58" x2="891.292" y2="1466.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.87" y1="1032.79" x2="1274.87" y2="1016.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.87" y1="1032.79" x2="1274.87" y2="1048.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.87" y1="1032.79" x2="1306.87" y2="1048.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1290.87" y1="1032.79" x2="1306.87" y2="1016.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1405.31" y1="1103.12" x2="1389.31" y2="1087.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1405.31" y1="1103.12" x2="1389.31" y2="1119.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1405.31" y1="1103.12" x2="1421.31" y2="1119.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1405.31" y1="1103.12" x2="1421.31" y2="1087.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1142.23" y1="1185.99" x2="1126.23" y2="1169.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1142.23" y1="1185.99" x2="1126.23" y2="1201.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1142.23" y1="1185.99" x2="1158.23" y2="1201.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1142.23" y1="1185.99" x2="1158.23" y2="1169.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.433" y1="1329.29" x2="883.433" y2="1313.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.433" y1="1329.29" x2="883.433" y2="1345.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.433" y1="1329.29" x2="915.433" y2="1345.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="899.433" y1="1329.29" x2="915.433" y2="1313.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.35" y1="1160.97" x2="1224.35" y2="1144.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.35" y1="1160.97" x2="1224.35" y2="1176.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.35" y1="1160.97" x2="1256.35" y2="1176.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1240.35" y1="1160.97" x2="1256.35" y2="1144.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1058.78" y1="1213.58" x2="1042.78" y2="1197.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1058.78" y1="1213.58" x2="1042.78" y2="1229.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1058.78" y1="1213.58" x2="1074.78" y2="1229.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1058.78" y1="1213.58" x2="1074.78" y2="1197.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1093.83" y1="1252.83" x2="1077.83" y2="1236.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1093.83" y1="1252.83" x2="1077.83" y2="1268.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1093.83" y1="1252.83" x2="1109.83" y2="1268.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1093.83" y1="1252.83" x2="1109.83" y2="1236.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1423.88" y1="1021.7" x2="1407.88" y2="1005.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1423.88" y1="1021.7" x2="1407.88" y2="1037.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1423.88" y1="1021.7" x2="1439.88" y2="1037.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1423.88" y1="1021.7" x2="1439.88" y2="1005.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1552.62" y1="936.348" x2="1536.62" y2="920.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1552.62" y1="936.348" x2="1536.62" y2="952.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1552.62" y1="936.348" x2="1568.62" y2="952.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1552.62" y1="936.348" x2="1568.62" y2="920.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.54" y1="984.865" x2="1456.54" y2="968.865" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.54" y1="984.865" x2="1456.54" y2="1000.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.54" y1="984.865" x2="1488.54" y2="1000.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1472.54" y1="984.865" x2="1488.54" y2="968.865" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165" y1="1027.31" x2="1149" y2="1011.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165" y1="1027.31" x2="1149" y2="1043.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165" y1="1027.31" x2="1181" y2="1043.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1165" y1="1027.31" x2="1181" y2="1011.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.61" y1="1250.08" x2="1094.61" y2="1234.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.61" y1="1250.08" x2="1094.61" y2="1266.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.61" y1="1250.08" x2="1126.61" y2="1266.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1110.61" y1="1250.08" x2="1126.61" y2="1234.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1154.89" y1="1238.14" x2="1138.89" y2="1222.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1154.89" y1="1238.14" x2="1138.89" y2="1254.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1154.89" y1="1238.14" x2="1170.89" y2="1254.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1154.89" y1="1238.14" x2="1170.89" y2="1222.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.8" y1="1203.89" x2="1095.8" y2="1187.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.8" y1="1203.89" x2="1095.8" y2="1219.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.8" y1="1203.89" x2="1127.8" y2="1219.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1111.8" y1="1203.89" x2="1127.8" y2="1187.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239" y1="1106.12" x2="1223" y2="1090.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239" y1="1106.12" x2="1223" y2="1122.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239" y1="1106.12" x2="1255" y2="1122.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1239" y1="1106.12" x2="1255" y2="1090.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="839.379" y1="1287.74" x2="823.379" y2="1271.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="839.379" y1="1287.74" x2="823.379" y2="1303.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="839.379" y1="1287.74" x2="855.379" y2="1303.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="839.379" y1="1287.74" x2="855.379" y2="1271.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1100.71" y1="1170.24" x2="1084.71" y2="1154.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1100.71" y1="1170.24" x2="1084.71" y2="1186.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1100.71" y1="1170.24" x2="1116.71" y2="1186.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1100.71" y1="1170.24" x2="1116.71" y2="1154.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414" y1="1170.91" x2="1398" y2="1154.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414" y1="1170.91" x2="1398" y2="1186.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414" y1="1170.91" x2="1430" y2="1186.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414" y1="1170.91" x2="1430" y2="1154.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.86" y1="1217.88" x2="1398.86" y2="1201.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.86" y1="1217.88" x2="1398.86" y2="1233.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.86" y1="1217.88" x2="1430.86" y2="1233.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1414.86" y1="1217.88" x2="1430.86" y2="1201.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1194.86" y1="1244.81" x2="1178.86" y2="1228.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1194.86" y1="1244.81" x2="1178.86" y2="1260.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1194.86" y1="1244.81" x2="1210.86" y2="1260.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1194.86" y1="1244.81" x2="1210.86" y2="1228.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1257.94" y1="1026.73" x2="1241.94" y2="1010.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1257.94" y1="1026.73" x2="1241.94" y2="1042.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1257.94" y1="1026.73" x2="1273.94" y2="1042.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1257.94" y1="1026.73" x2="1273.94" y2="1010.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1296.85" y1="1204.15" x2="1280.85" y2="1188.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1296.85" y1="1204.15" x2="1280.85" y2="1220.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1296.85" y1="1204.15" x2="1312.85" y2="1220.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1296.85" y1="1204.15" x2="1312.85" y2="1188.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.11" y1="1134.69" x2="1227.11" y2="1118.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.11" y1="1134.69" x2="1227.11" y2="1150.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.11" y1="1134.69" x2="1259.11" y2="1150.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1243.11" y1="1134.69" x2="1259.11" y2="1118.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.52" y1="1257.08" x2="1321.52" y2="1241.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.52" y1="1257.08" x2="1321.52" y2="1273.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.52" y1="1257.08" x2="1353.52" y2="1273.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="1337.52" y1="1257.08" x2="1353.52" y2="1241.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="844.925" y1="1383.63" x2="828.925" y2="1367.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="844.925" y1="1383.63" x2="828.925" y2="1399.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="844.925" y1="1383.63" x2="860.925" y2="1399.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip492)" x1="844.925" y1="1383.63" x2="860.925" y2="1367.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+</svg>
diff --git a/previews/PR3525/tutorials/conic/ellipse_approx/e71dc9be.svg b/previews/PR3525/tutorials/conic/ellipse_approx/e71dc9be.svg
new file mode 100644
index 00000000000..002e3d79a68
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/ellipse_approx/e71dc9be.svg
@@ -0,0 +1,2451 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="600" viewBox="0 0 2400 2400">
+<defs>
+  <clipPath id="clip550">
+    <rect x="0" y="0" width="2400" height="2400"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip550)" d="M0 2400 L2400 2400 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip551">
+    <rect x="480" y="240" width="1681" height="1681"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip550)" d="M178.867 2261.17 L2352.76 2261.17 L2352.76 47.2441 L178.867 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip552">
+    <rect x="178" y="47" width="2175" height="2215"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="327.236,2261.17 327.236,47.2441 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="640.095,2261.17 640.095,47.2441 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="952.953,2261.17 952.953,47.2441 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1265.81,2261.17 1265.81,47.2441 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1578.67,2261.17 1578.67,47.2441 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1891.53,2261.17 1891.53,47.2441 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2204.39,2261.17 2204.39,47.2441 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,2261.17 2352.76,2261.17 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="327.236,2261.17 327.236,2242.27 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="640.095,2261.17 640.095,2242.27 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="952.953,2261.17 952.953,2242.27 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1265.81,2261.17 1265.81,2242.27 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1578.67,2261.17 1578.67,2242.27 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1891.53,2261.17 1891.53,2242.27 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2204.39,2261.17 2204.39,2242.27 "/>
+<path clip-path="url(#clip550)" d="M281.299 2306.74 L310.975 2306.74 L310.975 2310.67 L281.299 2310.67 L281.299 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M335.234 2304.93 Q338.59 2305.65 340.465 2307.92 Q342.363 2310.19 342.363 2313.52 Q342.363 2318.64 338.845 2321.44 Q335.326 2324.24 328.845 2324.24 Q326.669 2324.24 324.354 2323.8 Q322.063 2323.38 319.609 2322.53 L319.609 2318.01 Q321.553 2319.15 323.868 2319.73 Q326.183 2320.3 328.706 2320.3 Q333.104 2320.3 335.396 2318.57 Q337.711 2316.83 337.711 2313.52 Q337.711 2310.47 335.558 2308.75 Q333.428 2307.02 329.609 2307.02 L325.581 2307.02 L325.581 2303.17 L329.794 2303.17 Q333.243 2303.17 335.072 2301.81 Q336.9 2300.42 336.9 2297.83 Q336.9 2295.17 335.002 2293.75 Q333.127 2292.32 329.609 2292.32 Q327.688 2292.32 325.488 2292.73 Q323.289 2293.15 320.651 2294.03 L320.651 2289.86 Q323.313 2289.12 325.627 2288.75 Q327.965 2288.38 330.025 2288.38 Q335.35 2288.38 338.451 2290.81 Q341.553 2293.22 341.553 2297.34 Q341.553 2300.21 339.91 2302.2 Q338.266 2304.17 335.234 2304.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M361.229 2292.09 Q357.618 2292.09 355.789 2295.65 Q353.984 2299.19 353.984 2306.32 Q353.984 2313.43 355.789 2316.99 Q357.618 2320.54 361.229 2320.54 Q364.863 2320.54 366.669 2316.99 Q368.497 2313.43 368.497 2306.32 Q368.497 2299.19 366.669 2295.65 Q364.863 2292.09 361.229 2292.09 M361.229 2288.38 Q367.039 2288.38 370.095 2292.99 Q373.173 2297.57 373.173 2306.32 Q373.173 2315.05 370.095 2319.66 Q367.039 2324.24 361.229 2324.24 Q355.419 2324.24 352.34 2319.66 Q349.285 2315.05 349.285 2306.32 Q349.285 2297.57 352.34 2292.99 Q355.419 2288.38 361.229 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M594.157 2306.74 L623.833 2306.74 L623.833 2310.67 L594.157 2310.67 L594.157 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M637.953 2319.63 L654.273 2319.63 L654.273 2323.57 L632.328 2323.57 L632.328 2319.63 Q634.99 2316.88 639.574 2312.25 Q644.18 2307.6 645.361 2306.25 Q647.606 2303.73 648.486 2301.99 Q649.388 2300.23 649.388 2298.54 Q649.388 2295.79 647.444 2294.05 Q645.523 2292.32 642.421 2292.32 Q640.222 2292.32 637.768 2293.08 Q635.338 2293.85 632.56 2295.4 L632.56 2290.67 Q635.384 2289.54 637.838 2288.96 Q640.291 2288.38 642.328 2288.38 Q647.699 2288.38 650.893 2291.07 Q654.087 2293.75 654.087 2298.24 Q654.087 2300.37 653.277 2302.29 Q652.49 2304.19 650.384 2306.79 Q649.805 2307.46 646.703 2310.67 Q643.601 2313.87 637.953 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M674.087 2292.09 Q670.476 2292.09 668.648 2295.65 Q666.842 2299.19 666.842 2306.32 Q666.842 2313.43 668.648 2316.99 Q670.476 2320.54 674.087 2320.54 Q677.722 2320.54 679.527 2316.99 Q681.356 2313.43 681.356 2306.32 Q681.356 2299.19 679.527 2295.65 Q677.722 2292.09 674.087 2292.09 M674.087 2288.38 Q679.898 2288.38 682.953 2292.99 Q686.032 2297.57 686.032 2306.32 Q686.032 2315.05 682.953 2319.66 Q679.898 2324.24 674.087 2324.24 Q668.277 2324.24 665.199 2319.66 Q662.143 2315.05 662.143 2306.32 Q662.143 2297.57 665.199 2292.99 Q668.277 2288.38 674.087 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M907.016 2306.74 L936.691 2306.74 L936.691 2310.67 L907.016 2310.67 L907.016 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M947.594 2319.63 L955.233 2319.63 L955.233 2293.27 L946.923 2294.93 L946.923 2290.67 L955.187 2289.01 L959.863 2289.01 L959.863 2319.63 L967.501 2319.63 L967.501 2323.57 L947.594 2323.57 L947.594 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M986.946 2292.09 Q983.335 2292.09 981.506 2295.65 Q979.7 2299.19 979.7 2306.32 Q979.7 2313.43 981.506 2316.99 Q983.335 2320.54 986.946 2320.54 Q990.58 2320.54 992.386 2316.99 Q994.214 2313.43 994.214 2306.32 Q994.214 2299.19 992.386 2295.65 Q990.58 2292.09 986.946 2292.09 M986.946 2288.38 Q992.756 2288.38 995.811 2292.99 Q998.89 2297.57 998.89 2306.32 Q998.89 2315.05 995.811 2319.66 Q992.756 2324.24 986.946 2324.24 Q981.136 2324.24 978.057 2319.66 Q975.001 2315.05 975.001 2306.32 Q975.001 2297.57 978.057 2292.99 Q981.136 2288.38 986.946 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M1265.81 2292.09 Q1262.2 2292.09 1260.37 2295.65 Q1258.57 2299.19 1258.57 2306.32 Q1258.57 2313.43 1260.37 2316.99 Q1262.2 2320.54 1265.81 2320.54 Q1269.45 2320.54 1271.25 2316.99 Q1273.08 2313.43 1273.08 2306.32 Q1273.08 2299.19 1271.25 2295.65 Q1269.45 2292.09 1265.81 2292.09 M1265.81 2288.38 Q1271.62 2288.38 1274.68 2292.99 Q1277.76 2297.57 1277.76 2306.32 Q1277.76 2315.05 1274.68 2319.66 Q1271.62 2324.24 1265.81 2324.24 Q1260 2324.24 1256.92 2319.66 Q1253.87 2315.05 1253.87 2306.32 Q1253.87 2297.57 1256.92 2292.99 Q1260 2288.38 1265.81 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M1553.36 2319.63 L1561 2319.63 L1561 2293.27 L1552.69 2294.93 L1552.69 2290.67 L1560.95 2289.01 L1565.63 2289.01 L1565.63 2319.63 L1573.26 2319.63 L1573.26 2323.57 L1553.36 2323.57 L1553.36 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M1592.71 2292.09 Q1589.1 2292.09 1587.27 2295.65 Q1585.46 2299.19 1585.46 2306.32 Q1585.46 2313.43 1587.27 2316.99 Q1589.1 2320.54 1592.71 2320.54 Q1596.34 2320.54 1598.15 2316.99 Q1599.98 2313.43 1599.98 2306.32 Q1599.98 2299.19 1598.15 2295.65 Q1596.34 2292.09 1592.71 2292.09 M1592.71 2288.38 Q1598.52 2288.38 1601.57 2292.99 Q1604.65 2297.57 1604.65 2306.32 Q1604.65 2315.05 1601.57 2319.66 Q1598.52 2324.24 1592.71 2324.24 Q1586.9 2324.24 1583.82 2319.66 Q1580.76 2315.05 1580.76 2306.32 Q1580.76 2297.57 1583.82 2292.99 Q1586.9 2288.38 1592.71 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M1870.3 2319.63 L1886.62 2319.63 L1886.62 2323.57 L1864.68 2323.57 L1864.68 2319.63 Q1867.34 2316.88 1871.92 2312.25 Q1876.53 2307.6 1877.71 2306.25 Q1879.95 2303.73 1880.83 2301.99 Q1881.74 2300.23 1881.74 2298.54 Q1881.74 2295.79 1879.79 2294.05 Q1877.87 2292.32 1874.77 2292.32 Q1872.57 2292.32 1870.12 2293.08 Q1867.69 2293.85 1864.91 2295.4 L1864.91 2290.67 Q1867.73 2289.54 1870.19 2288.96 Q1872.64 2288.38 1874.68 2288.38 Q1880.05 2288.38 1883.24 2291.07 Q1886.44 2293.75 1886.44 2298.24 Q1886.44 2300.37 1885.63 2302.29 Q1884.84 2304.19 1882.73 2306.79 Q1882.15 2307.46 1879.05 2310.67 Q1875.95 2313.87 1870.3 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M1906.44 2292.09 Q1902.82 2292.09 1901 2295.65 Q1899.19 2299.19 1899.19 2306.32 Q1899.19 2313.43 1901 2316.99 Q1902.82 2320.54 1906.44 2320.54 Q1910.07 2320.54 1911.88 2316.99 Q1913.7 2313.43 1913.7 2306.32 Q1913.7 2299.19 1911.88 2295.65 Q1910.07 2292.09 1906.44 2292.09 M1906.44 2288.38 Q1912.25 2288.38 1915.3 2292.99 Q1918.38 2297.57 1918.38 2306.32 Q1918.38 2315.05 1915.3 2319.66 Q1912.25 2324.24 1906.44 2324.24 Q1900.63 2324.24 1897.55 2319.66 Q1894.49 2315.05 1894.49 2306.32 Q1894.49 2297.57 1897.55 2292.99 Q1900.63 2288.38 1906.44 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M2193.23 2304.93 Q2196.59 2305.65 2198.46 2307.92 Q2200.36 2310.19 2200.36 2313.52 Q2200.36 2318.64 2196.84 2321.44 Q2193.32 2324.24 2186.84 2324.24 Q2184.66 2324.24 2182.35 2323.8 Q2180.06 2323.38 2177.6 2322.53 L2177.6 2318.01 Q2179.55 2319.15 2181.86 2319.73 Q2184.18 2320.3 2186.7 2320.3 Q2191.1 2320.3 2193.39 2318.57 Q2195.71 2316.83 2195.71 2313.52 Q2195.71 2310.47 2193.55 2308.75 Q2191.42 2307.02 2187.6 2307.02 L2183.58 2307.02 L2183.58 2303.17 L2187.79 2303.17 Q2191.24 2303.17 2193.07 2301.81 Q2194.9 2300.42 2194.9 2297.83 Q2194.9 2295.17 2193 2293.75 Q2191.12 2292.32 2187.6 2292.32 Q2185.68 2292.32 2183.48 2292.73 Q2181.28 2293.15 2178.65 2294.03 L2178.65 2289.86 Q2181.31 2289.12 2183.62 2288.75 Q2185.96 2288.38 2188.02 2288.38 Q2193.34 2288.38 2196.45 2290.81 Q2199.55 2293.22 2199.55 2297.34 Q2199.55 2300.21 2197.9 2302.2 Q2196.26 2304.17 2193.23 2304.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M2219.22 2292.09 Q2215.61 2292.09 2213.78 2295.65 Q2211.98 2299.19 2211.98 2306.32 Q2211.98 2313.43 2213.78 2316.99 Q2215.61 2320.54 2219.22 2320.54 Q2222.86 2320.54 2224.66 2316.99 Q2226.49 2313.43 2226.49 2306.32 Q2226.49 2299.19 2224.66 2295.65 Q2222.86 2292.09 2219.22 2292.09 M2219.22 2288.38 Q2225.03 2288.38 2228.09 2292.99 Q2231.17 2297.57 2231.17 2306.32 Q2231.17 2315.05 2228.09 2319.66 Q2225.03 2324.24 2219.22 2324.24 Q2213.41 2324.24 2210.34 2319.66 Q2207.28 2315.05 2207.28 2306.32 Q2207.28 2297.57 2210.34 2292.99 Q2213.41 2288.38 2219.22 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,2110.07 2352.76,2110.07 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,1791.45 2352.76,1791.45 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,1472.83 2352.76,1472.83 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,1154.21 2352.76,1154.21 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,835.586 2352.76,835.586 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,516.966 2352.76,516.966 "/>
+<polyline clip-path="url(#clip552)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="178.867,198.346 2352.76,198.346 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,2261.17 178.867,47.2441 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,2110.07 197.764,2110.07 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,1791.45 197.764,1791.45 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,1472.83 197.764,1472.83 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,1154.21 197.764,1154.21 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,835.586 197.764,835.586 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,516.966 197.764,516.966 "/>
+<polyline clip-path="url(#clip550)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="178.867,198.346 197.764,198.346 "/>
+<path clip-path="url(#clip550)" d="M50.9921 2110.52 L80.6679 2110.52 L80.6679 2114.45 L50.9921 2114.45 L50.9921 2110.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M104.927 2108.71 Q108.283 2109.43 110.158 2111.7 Q112.057 2113.97 112.057 2117.3 Q112.057 2122.42 108.538 2125.22 Q105.02 2128.02 98.5382 2128.02 Q96.3623 2128.02 94.0475 2127.58 Q91.7558 2127.16 89.3021 2126.3 L89.3021 2121.79 Q91.2465 2122.92 93.5613 2123.5 Q95.8761 2124.08 98.3993 2124.08 Q102.797 2124.08 105.089 2122.35 Q107.404 2120.61 107.404 2117.3 Q107.404 2114.24 105.251 2112.53 Q103.121 2110.8 99.3021 2110.8 L95.2743 2110.8 L95.2743 2106.95 L99.4872 2106.95 Q102.936 2106.95 104.765 2105.59 Q106.594 2104.2 106.594 2101.61 Q106.594 2098.94 104.696 2097.53 Q102.821 2096.1 99.3021 2096.1 Q97.3808 2096.1 95.1817 2096.51 Q92.9826 2096.93 90.3438 2097.81 L90.3438 2093.64 Q93.0058 2092.9 95.3206 2092.53 Q97.6585 2092.16 99.7187 2092.16 Q105.043 2092.16 108.145 2094.59 Q111.246 2097 111.246 2101.12 Q111.246 2103.99 109.603 2105.98 Q107.959 2107.95 104.927 2108.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M130.922 2095.86 Q127.311 2095.86 125.482 2099.43 Q123.677 2102.97 123.677 2110.1 Q123.677 2117.21 125.482 2120.77 Q127.311 2124.31 130.922 2124.31 Q134.556 2124.31 136.362 2120.77 Q138.191 2117.21 138.191 2110.1 Q138.191 2102.97 136.362 2099.43 Q134.556 2095.86 130.922 2095.86 M130.922 2092.16 Q136.732 2092.16 139.788 2096.77 Q142.867 2101.35 142.867 2110.1 Q142.867 2118.83 139.788 2123.43 Q136.732 2128.02 130.922 2128.02 Q125.112 2128.02 122.033 2123.43 Q118.978 2118.83 118.978 2110.1 Q118.978 2101.35 122.033 2096.77 Q125.112 2092.16 130.922 2092.16 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M50.9921 1791.9 L80.6679 1791.9 L80.6679 1795.83 L50.9921 1795.83 L50.9921 1791.9 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M94.7882 1804.79 L111.108 1804.79 L111.108 1808.73 L89.1632 1808.73 L89.1632 1804.79 Q91.8252 1802.04 96.4085 1797.41 Q101.015 1792.75 102.196 1791.41 Q104.441 1788.89 105.321 1787.15 Q106.223 1785.39 106.223 1783.7 Q106.223 1780.95 104.279 1779.21 Q102.358 1777.48 99.2558 1777.48 Q97.0567 1777.48 94.603 1778.24 Q92.1725 1779 89.3947 1780.55 L89.3947 1775.83 Q92.2188 1774.7 94.6724 1774.12 Q97.1261 1773.54 99.1632 1773.54 Q104.534 1773.54 107.728 1776.23 Q110.922 1778.91 110.922 1783.4 Q110.922 1785.53 110.112 1787.45 Q109.325 1789.35 107.219 1791.94 Q106.64 1792.61 103.538 1795.83 Q100.436 1799.03 94.7882 1804.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M130.922 1777.24 Q127.311 1777.24 125.482 1780.81 Q123.677 1784.35 123.677 1791.48 Q123.677 1798.59 125.482 1802.15 Q127.311 1805.69 130.922 1805.69 Q134.556 1805.69 136.362 1802.15 Q138.191 1798.59 138.191 1791.48 Q138.191 1784.35 136.362 1780.81 Q134.556 1777.24 130.922 1777.24 M130.922 1773.54 Q136.732 1773.54 139.788 1778.15 Q142.867 1782.73 142.867 1791.48 Q142.867 1800.21 139.788 1804.81 Q136.732 1809.4 130.922 1809.4 Q125.112 1809.4 122.033 1804.81 Q118.978 1800.21 118.978 1791.48 Q118.978 1782.73 122.033 1778.15 Q125.112 1773.54 130.922 1773.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M50.9921 1473.28 L80.6679 1473.28 L80.6679 1477.21 L50.9921 1477.21 L50.9921 1473.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M91.5706 1486.17 L99.2095 1486.17 L99.2095 1459.81 L90.8993 1461.47 L90.8993 1457.21 L99.1632 1455.55 L103.839 1455.55 L103.839 1486.17 L111.478 1486.17 L111.478 1490.11 L91.5706 1490.11 L91.5706 1486.17 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M130.922 1458.62 Q127.311 1458.62 125.482 1462.19 Q123.677 1465.73 123.677 1472.86 Q123.677 1479.97 125.482 1483.53 Q127.311 1487.07 130.922 1487.07 Q134.556 1487.07 136.362 1483.53 Q138.191 1479.97 138.191 1472.86 Q138.191 1465.73 136.362 1462.19 Q134.556 1458.62 130.922 1458.62 M130.922 1454.92 Q136.732 1454.92 139.788 1459.53 Q142.867 1464.11 142.867 1472.86 Q142.867 1481.59 139.788 1486.19 Q136.732 1490.78 130.922 1490.78 Q125.112 1490.78 122.033 1486.19 Q118.978 1481.59 118.978 1472.86 Q118.978 1464.11 122.033 1459.53 Q125.112 1454.92 130.922 1454.92 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M130.922 1140 Q127.311 1140 125.482 1143.57 Q123.677 1147.11 123.677 1154.24 Q123.677 1161.35 125.482 1164.91 Q127.311 1168.45 130.922 1168.45 Q134.556 1168.45 136.362 1164.91 Q138.191 1161.35 138.191 1154.24 Q138.191 1147.11 136.362 1143.57 Q134.556 1140 130.922 1140 M130.922 1136.3 Q136.732 1136.3 139.788 1140.91 Q142.867 1145.49 142.867 1154.24 Q142.867 1162.97 139.788 1167.57 Q136.732 1172.16 130.922 1172.16 Q125.112 1172.16 122.033 1167.57 Q118.978 1162.97 118.978 1154.24 Q118.978 1145.49 122.033 1140.91 Q125.112 1136.3 130.922 1136.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M91.5706 848.931 L99.2095 848.931 L99.2095 822.565 L90.8993 824.232 L90.8993 819.973 L99.1632 818.306 L103.839 818.306 L103.839 848.931 L111.478 848.931 L111.478 852.866 L91.5706 852.866 L91.5706 848.931 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M130.922 821.385 Q127.311 821.385 125.482 824.949 Q123.677 828.491 123.677 835.621 Q123.677 842.727 125.482 846.292 Q127.311 849.834 130.922 849.834 Q134.556 849.834 136.362 846.292 Q138.191 842.727 138.191 835.621 Q138.191 828.491 136.362 824.949 Q134.556 821.385 130.922 821.385 M130.922 817.681 Q136.732 817.681 139.788 822.287 Q142.867 826.871 142.867 835.621 Q142.867 844.348 139.788 848.954 Q136.732 853.537 130.922 853.537 Q125.112 853.537 122.033 848.954 Q118.978 844.348 118.978 835.621 Q118.978 826.871 122.033 822.287 Q125.112 817.681 130.922 817.681 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M94.7882 530.311 L111.108 530.311 L111.108 534.246 L89.1632 534.246 L89.1632 530.311 Q91.8252 527.556 96.4085 522.927 Q101.015 518.274 102.196 516.931 Q104.441 514.408 105.321 512.672 Q106.223 510.913 106.223 509.223 Q106.223 506.468 104.279 504.732 Q102.358 502.996 99.2558 502.996 Q97.0567 502.996 94.603 503.76 Q92.1725 504.524 89.3947 506.075 L89.3947 501.353 Q92.2188 500.218 94.6724 499.64 Q97.1261 499.061 99.1632 499.061 Q104.534 499.061 107.728 501.746 Q110.922 504.431 110.922 508.922 Q110.922 511.052 110.112 512.973 Q109.325 514.871 107.219 517.464 Q106.64 518.135 103.538 521.353 Q100.436 524.547 94.7882 530.311 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M130.922 502.765 Q127.311 502.765 125.482 506.329 Q123.677 509.871 123.677 517.001 Q123.677 524.107 125.482 527.672 Q127.311 531.214 130.922 531.214 Q134.556 531.214 136.362 527.672 Q138.191 524.107 138.191 517.001 Q138.191 509.871 136.362 506.329 Q134.556 502.765 130.922 502.765 M130.922 499.061 Q136.732 499.061 139.788 503.667 Q142.867 508.251 142.867 517.001 Q142.867 525.727 139.788 530.334 Q136.732 534.917 130.922 534.917 Q125.112 534.917 122.033 530.334 Q118.978 525.727 118.978 517.001 Q118.978 508.251 122.033 503.667 Q125.112 499.061 130.922 499.061 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M104.927 196.992 Q108.283 197.709 110.158 199.978 Q112.057 202.246 112.057 205.58 Q112.057 210.695 108.538 213.496 Q105.02 216.297 98.5382 216.297 Q96.3623 216.297 94.0475 215.857 Q91.7558 215.441 89.3021 214.584 L89.3021 210.07 Q91.2465 211.205 93.5613 211.783 Q95.8761 212.362 98.3993 212.362 Q102.797 212.362 105.089 210.626 Q107.404 208.89 107.404 205.58 Q107.404 202.524 105.251 200.811 Q103.121 199.075 99.3021 199.075 L95.2743 199.075 L95.2743 195.233 L99.4872 195.233 Q102.936 195.233 104.765 193.867 Q106.594 192.478 106.594 189.885 Q106.594 187.223 104.696 185.811 Q102.821 184.376 99.3021 184.376 Q97.3808 184.376 95.1817 184.793 Q92.9826 185.209 90.3438 186.089 L90.3438 181.922 Q93.0058 181.182 95.3206 180.811 Q97.6585 180.441 99.7187 180.441 Q105.043 180.441 108.145 182.872 Q111.246 185.279 111.246 189.399 Q111.246 192.27 109.603 194.26 Q107.959 196.228 104.927 196.992 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip550)" d="M130.922 184.145 Q127.311 184.145 125.482 187.709 Q123.677 191.251 123.677 198.381 Q123.677 205.487 125.482 209.052 Q127.311 212.594 130.922 212.594 Q134.556 212.594 136.362 209.052 Q138.191 205.487 138.191 198.381 Q138.191 191.251 136.362 187.709 Q134.556 184.145 130.922 184.145 M130.922 180.441 Q136.732 180.441 139.788 185.047 Q142.867 189.631 142.867 198.381 Q142.867 207.107 139.788 211.714 Q136.732 216.297 130.922 216.297 Q125.112 216.297 122.033 211.714 Q118.978 207.107 118.978 198.381 Q118.978 189.631 122.033 185.047 Q125.112 180.441 130.922 180.441 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><line clip-path="url(#clip552)" x1="1391.44" y1="1186.36" x2="1375.44" y2="1170.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1391.44" y1="1186.36" x2="1375.44" y2="1202.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1391.44" y1="1186.36" x2="1407.44" y2="1202.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1391.44" y1="1186.36" x2="1407.44" y2="1170.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1372.17" y1="1098.14" x2="1356.17" y2="1082.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1372.17" y1="1098.14" x2="1356.17" y2="1114.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1372.17" y1="1098.14" x2="1388.17" y2="1114.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1372.17" y1="1098.14" x2="1388.17" y2="1082.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.37" y1="1262.61" x2="1095.37" y2="1246.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.37" y1="1262.61" x2="1095.37" y2="1278.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.37" y1="1262.61" x2="1127.37" y2="1278.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.37" y1="1262.61" x2="1127.37" y2="1246.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1229.97" x2="1288.99" y2="1213.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1229.97" x2="1288.99" y2="1245.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1229.97" x2="1320.99" y2="1245.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1229.97" x2="1320.99" y2="1213.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1053.12" y1="1313.7" x2="1037.12" y2="1297.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1053.12" y1="1313.7" x2="1037.12" y2="1329.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1053.12" y1="1313.7" x2="1069.12" y2="1329.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1053.12" y1="1313.7" x2="1069.12" y2="1297.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1388.9" y1="1173.08" x2="1372.9" y2="1157.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1388.9" y1="1173.08" x2="1372.9" y2="1189.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1388.9" y1="1173.08" x2="1404.9" y2="1189.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1388.9" y1="1173.08" x2="1404.9" y2="1157.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1926.4" y1="856.923" x2="1910.4" y2="840.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1926.4" y1="856.923" x2="1910.4" y2="872.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1926.4" y1="856.923" x2="1942.4" y2="872.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1926.4" y1="856.923" x2="1942.4" y2="840.923" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="912.69" y1="1348.83" x2="896.69" y2="1332.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="912.69" y1="1348.83" x2="896.69" y2="1364.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="912.69" y1="1348.83" x2="928.69" y2="1364.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="912.69" y1="1348.83" x2="928.69" y2="1332.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.77" y1="1174.01" x2="1343.77" y2="1158.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.77" y1="1174.01" x2="1343.77" y2="1190.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.77" y1="1174.01" x2="1375.77" y2="1190.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.77" y1="1174.01" x2="1375.77" y2="1158.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="625.575" y1="1469.54" x2="609.575" y2="1453.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="625.575" y1="1469.54" x2="609.575" y2="1485.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="625.575" y1="1469.54" x2="641.575" y2="1485.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="625.575" y1="1469.54" x2="641.575" y2="1453.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.02" y1="1148.47" x2="1438.02" y2="1132.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.02" y1="1148.47" x2="1438.02" y2="1164.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.02" y1="1148.47" x2="1470.02" y2="1164.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.02" y1="1148.47" x2="1470.02" y2="1132.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.94" y1="1144.11" x2="1399.94" y2="1128.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.94" y1="1144.11" x2="1399.94" y2="1160.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.94" y1="1144.11" x2="1431.94" y2="1160.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.94" y1="1144.11" x2="1431.94" y2="1128.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.61" y1="1071.18" x2="1413.61" y2="1055.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.61" y1="1071.18" x2="1413.61" y2="1087.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.61" y1="1071.18" x2="1445.61" y2="1087.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.61" y1="1071.18" x2="1445.61" y2="1055.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.18" y1="1043.75" x2="1535.18" y2="1027.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.18" y1="1043.75" x2="1535.18" y2="1059.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.18" y1="1043.75" x2="1567.18" y2="1059.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.18" y1="1043.75" x2="1567.18" y2="1027.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1383.83" y1="1070.02" x2="1367.83" y2="1054.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1383.83" y1="1070.02" x2="1367.83" y2="1086.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1383.83" y1="1070.02" x2="1399.83" y2="1086.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1383.83" y1="1070.02" x2="1399.83" y2="1054.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.68" y1="1259.28" x2="1120.68" y2="1243.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.68" y1="1259.28" x2="1120.68" y2="1275.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.68" y1="1259.28" x2="1152.68" y2="1275.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.68" y1="1259.28" x2="1152.68" y2="1243.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.69" y1="1027.81" x2="1325.69" y2="1011.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.69" y1="1027.81" x2="1325.69" y2="1043.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.69" y1="1027.81" x2="1357.69" y2="1043.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.69" y1="1027.81" x2="1357.69" y2="1011.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1260.55" y1="1177.07" x2="1244.55" y2="1161.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1260.55" y1="1177.07" x2="1244.55" y2="1193.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1260.55" y1="1177.07" x2="1276.55" y2="1193.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1260.55" y1="1177.07" x2="1276.55" y2="1161.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.36" y1="1138.56" x2="990.356" y2="1122.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.36" y1="1138.56" x2="990.356" y2="1154.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.36" y1="1138.56" x2="1022.36" y2="1154.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.36" y1="1138.56" x2="1022.36" y2="1122.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="785.24" y1="1436.86" x2="769.24" y2="1420.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="785.24" y1="1436.86" x2="769.24" y2="1452.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="785.24" y1="1436.86" x2="801.24" y2="1452.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="785.24" y1="1436.86" x2="801.24" y2="1420.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1271.22" y1="1086.79" x2="1255.22" y2="1070.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1271.22" y1="1086.79" x2="1255.22" y2="1102.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1271.22" y1="1086.79" x2="1287.22" y2="1102.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1271.22" y1="1086.79" x2="1287.22" y2="1070.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1019.17" y1="1201.94" x2="1003.17" y2="1185.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1019.17" y1="1201.94" x2="1003.17" y2="1217.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1019.17" y1="1201.94" x2="1035.17" y2="1217.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1019.17" y1="1201.94" x2="1035.17" y2="1185.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.289" y1="1394.75" x2="761.289" y2="1378.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.289" y1="1394.75" x2="761.289" y2="1410.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.289" y1="1394.75" x2="793.289" y2="1410.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.289" y1="1394.75" x2="793.289" y2="1378.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1511.54" y1="1080.45" x2="1495.54" y2="1064.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1511.54" y1="1080.45" x2="1495.54" y2="1096.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1511.54" y1="1080.45" x2="1527.54" y2="1096.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1511.54" y1="1080.45" x2="1527.54" y2="1064.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1461.76" y1="1072.86" x2="1445.76" y2="1056.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1461.76" y1="1072.86" x2="1445.76" y2="1088.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1461.76" y1="1072.86" x2="1477.76" y2="1088.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1461.76" y1="1072.86" x2="1477.76" y2="1056.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.41" y1="1014.15" x2="1369.41" y2="998.153" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.41" y1="1014.15" x2="1369.41" y2="1030.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.41" y1="1014.15" x2="1401.41" y2="1030.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.41" y1="1014.15" x2="1401.41" y2="998.153" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.61" y1="1146.61" x2="1222.61" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.61" y1="1146.61" x2="1222.61" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.61" y1="1146.61" x2="1254.61" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.61" y1="1146.61" x2="1254.61" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1612.46" y1="913.718" x2="1596.46" y2="897.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1612.46" y1="913.718" x2="1596.46" y2="929.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1612.46" y1="913.718" x2="1628.46" y2="929.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1612.46" y1="913.718" x2="1628.46" y2="897.718" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.2" y1="1238.54" x2="1271.2" y2="1222.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.2" y1="1238.54" x2="1271.2" y2="1254.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.2" y1="1238.54" x2="1303.2" y2="1254.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.2" y1="1238.54" x2="1303.2" y2="1222.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.35" y1="1265.07" x2="1043.35" y2="1249.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.35" y1="1265.07" x2="1043.35" y2="1281.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.35" y1="1265.07" x2="1075.35" y2="1281.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.35" y1="1265.07" x2="1075.35" y2="1249.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="947.412" y1="1370.03" x2="931.412" y2="1354.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="947.412" y1="1370.03" x2="931.412" y2="1386.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="947.412" y1="1370.03" x2="963.412" y2="1386.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="947.412" y1="1370.03" x2="963.412" y2="1354.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1251.92" y1="1163.59" x2="1235.92" y2="1147.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1251.92" y1="1163.59" x2="1235.92" y2="1179.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1251.92" y1="1163.59" x2="1267.92" y2="1179.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1251.92" y1="1163.59" x2="1267.92" y2="1147.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.13" y1="1210.6" x2="1222.13" y2="1194.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.13" y1="1210.6" x2="1222.13" y2="1226.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.13" y1="1210.6" x2="1254.13" y2="1226.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.13" y1="1210.6" x2="1254.13" y2="1194.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.497" y1="1483.85" x2="672.497" y2="1467.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.497" y1="1483.85" x2="672.497" y2="1499.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.497" y1="1483.85" x2="704.497" y2="1499.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.497" y1="1483.85" x2="704.497" y2="1467.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1248.4" y1="1166.04" x2="1232.4" y2="1150.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1248.4" y1="1166.04" x2="1232.4" y2="1182.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1248.4" y1="1166.04" x2="1264.4" y2="1182.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1248.4" y1="1166.04" x2="1264.4" y2="1150.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.95" y1="869.353" x2="1529.95" y2="853.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.95" y1="869.353" x2="1529.95" y2="885.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.95" y1="869.353" x2="1561.95" y2="885.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.95" y1="869.353" x2="1561.95" y2="853.353" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.93" y1="1067.84" x2="1405.93" y2="1051.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.93" y1="1067.84" x2="1405.93" y2="1083.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.93" y1="1067.84" x2="1437.93" y2="1083.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.93" y1="1067.84" x2="1437.93" y2="1051.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.39" y1="1032.06" x2="1456.39" y2="1016.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.39" y1="1032.06" x2="1456.39" y2="1048.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.39" y1="1032.06" x2="1488.39" y2="1048.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.39" y1="1032.06" x2="1488.39" y2="1016.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.2" y1="1038.88" x2="1321.2" y2="1022.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.2" y1="1038.88" x2="1321.2" y2="1054.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.2" y1="1038.88" x2="1353.2" y2="1054.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.2" y1="1038.88" x2="1353.2" y2="1022.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.09" y1="1178.03" x2="1030.09" y2="1162.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.09" y1="1178.03" x2="1030.09" y2="1194.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.09" y1="1178.03" x2="1062.09" y2="1194.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.09" y1="1178.03" x2="1062.09" y2="1162.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.45" y1="1275.47" x2="1078.45" y2="1259.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.45" y1="1275.47" x2="1078.45" y2="1291.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.45" y1="1275.47" x2="1110.45" y2="1291.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.45" y1="1275.47" x2="1110.45" y2="1259.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="771.564" y1="1431.34" x2="755.564" y2="1415.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="771.564" y1="1431.34" x2="755.564" y2="1447.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="771.564" y1="1431.34" x2="787.564" y2="1447.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="771.564" y1="1431.34" x2="787.564" y2="1415.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="918.674" y1="1363.31" x2="902.674" y2="1347.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="918.674" y1="1363.31" x2="902.674" y2="1379.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="918.674" y1="1363.31" x2="934.674" y2="1379.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="918.674" y1="1363.31" x2="934.674" y2="1347.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.43" y1="1296.62" x2="1209.43" y2="1280.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.43" y1="1296.62" x2="1209.43" y2="1312.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.43" y1="1296.62" x2="1241.43" y2="1312.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.43" y1="1296.62" x2="1241.43" y2="1280.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.5" y1="1071.7" x2="1228.5" y2="1055.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.5" y1="1071.7" x2="1228.5" y2="1087.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.5" y1="1071.7" x2="1260.5" y2="1087.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.5" y1="1071.7" x2="1260.5" y2="1055.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.11" y1="1109.8" x2="1348.11" y2="1093.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.11" y1="1109.8" x2="1348.11" y2="1125.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.11" y1="1109.8" x2="1380.11" y2="1125.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.11" y1="1109.8" x2="1380.11" y2="1093.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1707.08" y1="826.86" x2="1691.08" y2="810.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1707.08" y1="826.86" x2="1691.08" y2="842.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1707.08" y1="826.86" x2="1723.08" y2="842.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1707.08" y1="826.86" x2="1723.08" y2="810.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.63" y1="875.887" x2="1561.63" y2="859.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.63" y1="875.887" x2="1561.63" y2="891.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.63" y1="875.887" x2="1593.63" y2="891.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.63" y1="875.887" x2="1593.63" y2="859.887" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1319.95" y1="1118.19" x2="1303.95" y2="1102.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1319.95" y1="1118.19" x2="1303.95" y2="1134.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1319.95" y1="1118.19" x2="1335.95" y2="1134.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1319.95" y1="1118.19" x2="1335.95" y2="1102.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="894.72" y1="1142.51" x2="878.72" y2="1126.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="894.72" y1="1142.51" x2="878.72" y2="1158.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="894.72" y1="1142.51" x2="910.72" y2="1158.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="894.72" y1="1142.51" x2="910.72" y2="1126.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.17" y1="1345.5" x2="1051.17" y2="1329.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.17" y1="1345.5" x2="1051.17" y2="1361.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.17" y1="1345.5" x2="1083.17" y2="1361.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.17" y1="1345.5" x2="1083.17" y2="1329.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1580.8" y1="990.127" x2="1564.8" y2="974.127" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1580.8" y1="990.127" x2="1564.8" y2="1006.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1580.8" y1="990.127" x2="1596.8" y2="1006.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1580.8" y1="990.127" x2="1596.8" y2="974.127" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="852.02" y1="1283.64" x2="836.02" y2="1267.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="852.02" y1="1283.64" x2="836.02" y2="1299.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="852.02" y1="1283.64" x2="868.02" y2="1299.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="852.02" y1="1283.64" x2="868.02" y2="1267.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.48" y1="962.993" x2="1589.48" y2="946.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.48" y1="962.993" x2="1589.48" y2="978.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.48" y1="962.993" x2="1621.48" y2="978.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.48" y1="962.993" x2="1621.48" y2="946.993" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1283.38" y1="1220.65" x2="1267.38" y2="1204.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1283.38" y1="1220.65" x2="1267.38" y2="1236.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1283.38" y1="1220.65" x2="1299.38" y2="1236.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1283.38" y1="1220.65" x2="1299.38" y2="1204.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.15" y1="1016.63" x2="1446.15" y2="1000.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.15" y1="1016.63" x2="1446.15" y2="1032.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.15" y1="1016.63" x2="1478.15" y2="1032.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.15" y1="1016.63" x2="1478.15" y2="1000.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1172.56" y1="1009.01" x2="1156.56" y2="993.009" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1172.56" y1="1009.01" x2="1156.56" y2="1025.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1172.56" y1="1009.01" x2="1188.56" y2="1025.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1172.56" y1="1009.01" x2="1188.56" y2="993.009" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.43" y1="1210.76" x2="1151.43" y2="1194.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.43" y1="1210.76" x2="1151.43" y2="1226.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.43" y1="1210.76" x2="1183.43" y2="1226.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.43" y1="1210.76" x2="1183.43" y2="1194.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.588" y1="1415.15" x2="672.588" y2="1399.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.588" y1="1415.15" x2="672.588" y2="1431.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.588" y1="1415.15" x2="704.588" y2="1431.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="688.588" y1="1415.15" x2="704.588" y2="1399.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.99" y1="1289.47" x2="1120.99" y2="1273.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.99" y1="1289.47" x2="1120.99" y2="1305.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.99" y1="1289.47" x2="1152.99" y2="1305.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1136.99" y1="1289.47" x2="1152.99" y2="1273.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="991.105" y1="1402.23" x2="975.105" y2="1386.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="991.105" y1="1402.23" x2="975.105" y2="1418.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="991.105" y1="1402.23" x2="1007.11" y2="1418.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="991.105" y1="1402.23" x2="1007.11" y2="1386.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1811.99" y1="919.23" x2="1795.99" y2="903.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1811.99" y1="919.23" x2="1795.99" y2="935.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1811.99" y1="919.23" x2="1827.99" y2="935.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1811.99" y1="919.23" x2="1827.99" y2="903.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.971" y1="1350.77" x2="829.971" y2="1334.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.971" y1="1350.77" x2="829.971" y2="1366.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.971" y1="1350.77" x2="861.971" y2="1366.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.971" y1="1350.77" x2="861.971" y2="1334.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1144.78" y1="1281.78" x2="1128.78" y2="1265.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1144.78" y1="1281.78" x2="1128.78" y2="1297.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1144.78" y1="1281.78" x2="1160.78" y2="1297.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1144.78" y1="1281.78" x2="1160.78" y2="1265.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1064.78" y1="1413.61" x2="1048.78" y2="1397.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1064.78" y1="1413.61" x2="1048.78" y2="1429.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1064.78" y1="1413.61" x2="1080.78" y2="1429.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1064.78" y1="1413.61" x2="1080.78" y2="1397.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="886.864" y1="1368.41" x2="870.864" y2="1352.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="886.864" y1="1368.41" x2="870.864" y2="1384.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="886.864" y1="1368.41" x2="902.864" y2="1384.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="886.864" y1="1368.41" x2="902.864" y2="1352.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.7" y1="1200.72" x2="1324.7" y2="1184.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.7" y1="1200.72" x2="1324.7" y2="1216.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.7" y1="1200.72" x2="1356.7" y2="1216.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.7" y1="1200.72" x2="1356.7" y2="1184.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.27" y1="1196.6" x2="1051.27" y2="1180.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.27" y1="1196.6" x2="1051.27" y2="1212.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.27" y1="1196.6" x2="1083.27" y2="1212.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.27" y1="1196.6" x2="1083.27" y2="1180.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.009" y1="1388.03" x2="761.009" y2="1372.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.009" y1="1388.03" x2="761.009" y2="1404.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.009" y1="1388.03" x2="793.009" y2="1404.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="777.009" y1="1388.03" x2="793.009" y2="1372.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.23" y1="998.156" x2="1277.23" y2="982.156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.23" y1="998.156" x2="1277.23" y2="1014.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.23" y1="998.156" x2="1309.23" y2="1014.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.23" y1="998.156" x2="1309.23" y2="982.156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="885.196" y1="1405.14" x2="869.196" y2="1389.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="885.196" y1="1405.14" x2="869.196" y2="1421.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="885.196" y1="1405.14" x2="901.196" y2="1421.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="885.196" y1="1405.14" x2="901.196" y2="1389.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1204.07" y1="1358.98" x2="1188.07" y2="1342.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1204.07" y1="1358.98" x2="1188.07" y2="1374.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1204.07" y1="1358.98" x2="1220.07" y2="1374.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1204.07" y1="1358.98" x2="1220.07" y2="1342.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1195.08" y1="1292.67" x2="1179.08" y2="1276.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1195.08" y1="1292.67" x2="1179.08" y2="1308.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1195.08" y1="1292.67" x2="1211.08" y2="1308.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1195.08" y1="1292.67" x2="1211.08" y2="1276.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.55" y1="853.956" x2="1441.55" y2="837.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.55" y1="853.956" x2="1441.55" y2="869.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.55" y1="853.956" x2="1473.55" y2="869.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.55" y1="853.956" x2="1473.55" y2="837.956" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.86" y1="1030.17" x2="1603.86" y2="1014.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.86" y1="1030.17" x2="1603.86" y2="1046.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.86" y1="1030.17" x2="1635.86" y2="1046.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.86" y1="1030.17" x2="1635.86" y2="1014.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="949.006" y1="1440.98" x2="933.006" y2="1424.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="949.006" y1="1440.98" x2="933.006" y2="1456.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="949.006" y1="1440.98" x2="965.006" y2="1456.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="949.006" y1="1440.98" x2="965.006" y2="1424.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.411" y1="1427.9" x2="829.411" y2="1411.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.411" y1="1427.9" x2="829.411" y2="1443.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.411" y1="1427.9" x2="861.411" y2="1443.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="845.411" y1="1427.9" x2="861.411" y2="1411.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.64" y1="1292.35" x2="1169.64" y2="1276.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.64" y1="1292.35" x2="1169.64" y2="1308.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.64" y1="1292.35" x2="1201.64" y2="1308.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.64" y1="1292.35" x2="1201.64" y2="1276.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.57" y1="868.321" x2="1530.57" y2="852.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.57" y1="868.321" x2="1530.57" y2="884.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.57" y1="868.321" x2="1562.57" y2="884.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.57" y1="868.321" x2="1562.57" y2="852.321" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.43" y1="1342.43" x2="834.43" y2="1326.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.43" y1="1342.43" x2="834.43" y2="1358.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.43" y1="1342.43" x2="866.43" y2="1358.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.43" y1="1342.43" x2="866.43" y2="1326.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.87" y1="1122.39" x2="1535.87" y2="1106.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.87" y1="1122.39" x2="1535.87" y2="1138.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.87" y1="1122.39" x2="1567.87" y2="1138.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1551.87" y1="1122.39" x2="1567.87" y2="1106.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.5" y1="829.178" x2="1603.5" y2="813.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.5" y1="829.178" x2="1603.5" y2="845.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.5" y1="829.178" x2="1635.5" y2="845.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1619.5" y1="829.178" x2="1635.5" y2="813.178" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1847.7" y1="804.547" x2="1831.7" y2="788.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1847.7" y1="804.547" x2="1831.7" y2="820.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1847.7" y1="804.547" x2="1863.7" y2="820.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1847.7" y1="804.547" x2="1863.7" y2="788.547" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="992.659" y1="1251.94" x2="976.659" y2="1235.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="992.659" y1="1251.94" x2="976.659" y2="1267.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="992.659" y1="1251.94" x2="1008.66" y2="1267.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="992.659" y1="1251.94" x2="1008.66" y2="1235.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.79" y1="1033.06" x2="1595.79" y2="1017.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.79" y1="1033.06" x2="1595.79" y2="1049.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.79" y1="1033.06" x2="1627.79" y2="1049.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.79" y1="1033.06" x2="1627.79" y2="1017.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1148.89" y1="1269.81" x2="1132.89" y2="1253.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1148.89" y1="1269.81" x2="1132.89" y2="1285.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1148.89" y1="1269.81" x2="1164.89" y2="1285.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1148.89" y1="1269.81" x2="1164.89" y2="1253.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1202.15" y1="1212.09" x2="1186.15" y2="1196.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1202.15" y1="1212.09" x2="1186.15" y2="1228.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1202.15" y1="1212.09" x2="1218.15" y2="1228.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1202.15" y1="1212.09" x2="1218.15" y2="1196.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1641.07" y1="860.861" x2="1625.07" y2="844.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1641.07" y1="860.861" x2="1625.07" y2="876.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1641.07" y1="860.861" x2="1657.07" y2="876.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1641.07" y1="860.861" x2="1657.07" y2="844.861" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.79" y1="1035.93" x2="1470.79" y2="1019.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.79" y1="1035.93" x2="1470.79" y2="1051.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.79" y1="1035.93" x2="1502.79" y2="1051.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.79" y1="1035.93" x2="1502.79" y2="1019.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1207.15" y1="1160.11" x2="1191.15" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1207.15" y1="1160.11" x2="1191.15" y2="1176.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1207.15" y1="1160.11" x2="1223.15" y2="1176.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1207.15" y1="1160.11" x2="1223.15" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1119.38" y1="1194.83" x2="1103.38" y2="1178.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1119.38" y1="1194.83" x2="1103.38" y2="1210.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1119.38" y1="1194.83" x2="1135.38" y2="1210.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1119.38" y1="1194.83" x2="1135.38" y2="1178.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1332.96" y1="1133.97" x2="1316.96" y2="1117.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1332.96" y1="1133.97" x2="1316.96" y2="1149.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1332.96" y1="1133.97" x2="1348.96" y2="1149.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1332.96" y1="1133.97" x2="1348.96" y2="1117.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1301.11" y1="987.604" x2="1285.11" y2="971.604" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1301.11" y1="987.604" x2="1285.11" y2="1003.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1301.11" y1="987.604" x2="1317.11" y2="1003.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1301.11" y1="987.604" x2="1317.11" y2="971.604" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.19" y1="1037.37" x2="1299.19" y2="1021.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.19" y1="1037.37" x2="1299.19" y2="1053.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.19" y1="1037.37" x2="1331.19" y2="1053.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.19" y1="1037.37" x2="1331.19" y2="1021.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1300.37" y1="1049.99" x2="1284.37" y2="1033.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1300.37" y1="1049.99" x2="1284.37" y2="1065.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1300.37" y1="1049.99" x2="1316.37" y2="1065.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1300.37" y1="1049.99" x2="1316.37" y2="1033.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.97" y1="1034.41" x2="1347.97" y2="1018.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.97" y1="1034.41" x2="1347.97" y2="1050.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.97" y1="1034.41" x2="1379.97" y2="1050.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.97" y1="1034.41" x2="1379.97" y2="1018.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.47" y1="1159.14" x2="1328.47" y2="1143.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.47" y1="1159.14" x2="1328.47" y2="1175.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.47" y1="1159.14" x2="1360.47" y2="1175.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.47" y1="1159.14" x2="1360.47" y2="1143.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1673.06" y1="901.944" x2="1657.06" y2="885.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1673.06" y1="901.944" x2="1657.06" y2="917.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1673.06" y1="901.944" x2="1689.06" y2="917.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1673.06" y1="901.944" x2="1689.06" y2="885.944" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.89" y1="961.872" x2="1595.89" y2="945.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.89" y1="961.872" x2="1595.89" y2="977.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.89" y1="961.872" x2="1627.89" y2="977.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.89" y1="961.872" x2="1627.89" y2="945.872" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.83" y1="1129.04" x2="1175.83" y2="1113.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.83" y1="1129.04" x2="1175.83" y2="1145.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.83" y1="1129.04" x2="1207.83" y2="1145.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.83" y1="1129.04" x2="1207.83" y2="1113.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.08" y1="1021.63" x2="1231.08" y2="1005.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.08" y1="1021.63" x2="1231.08" y2="1037.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.08" y1="1021.63" x2="1263.08" y2="1037.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.08" y1="1021.63" x2="1263.08" y2="1005.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1040.17" y1="1275.11" x2="1024.17" y2="1259.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1040.17" y1="1275.11" x2="1024.17" y2="1291.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1040.17" y1="1275.11" x2="1056.17" y2="1291.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1040.17" y1="1275.11" x2="1056.17" y2="1259.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="782.176" y1="1390.69" x2="766.176" y2="1374.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="782.176" y1="1390.69" x2="766.176" y2="1406.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="782.176" y1="1390.69" x2="798.176" y2="1406.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="782.176" y1="1390.69" x2="798.176" y2="1374.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.77" y1="921.537" x2="1659.77" y2="905.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.77" y1="921.537" x2="1659.77" y2="937.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.77" y1="921.537" x2="1691.77" y2="937.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.77" y1="921.537" x2="1691.77" y2="905.537" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.19" y1="1173.52" x2="1214.19" y2="1157.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.19" y1="1173.52" x2="1214.19" y2="1189.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.19" y1="1173.52" x2="1246.19" y2="1189.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.19" y1="1173.52" x2="1246.19" y2="1157.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1092.98" y1="1136.22" x2="1076.98" y2="1120.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1092.98" y1="1136.22" x2="1076.98" y2="1152.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1092.98" y1="1136.22" x2="1108.98" y2="1152.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1092.98" y1="1136.22" x2="1108.98" y2="1120.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="816.257" y1="1346.85" x2="800.257" y2="1330.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="816.257" y1="1346.85" x2="800.257" y2="1362.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="816.257" y1="1346.85" x2="832.257" y2="1362.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="816.257" y1="1346.85" x2="832.257" y2="1330.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="724.712" y1="1453.02" x2="708.712" y2="1437.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="724.712" y1="1453.02" x2="708.712" y2="1469.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="724.712" y1="1453.02" x2="740.712" y2="1469.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="724.712" y1="1453.02" x2="740.712" y2="1437.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.34" y1="968.47" x2="1659.34" y2="952.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.34" y1="968.47" x2="1659.34" y2="984.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.34" y1="968.47" x2="1691.34" y2="984.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1675.34" y1="968.47" x2="1691.34" y2="952.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.82" y1="1198.32" x2="1130.82" y2="1182.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.82" y1="1198.32" x2="1130.82" y2="1214.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.82" y1="1198.32" x2="1162.82" y2="1214.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.82" y1="1198.32" x2="1162.82" y2="1182.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.2" y1="1016.21" x2="1484.2" y2="1000.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.2" y1="1016.21" x2="1484.2" y2="1032.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.2" y1="1016.21" x2="1516.2" y2="1032.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.2" y1="1016.21" x2="1516.2" y2="1000.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.396" y1="1435.49" x2="858.396" y2="1419.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.396" y1="1435.49" x2="858.396" y2="1451.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.396" y1="1435.49" x2="890.396" y2="1451.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.396" y1="1435.49" x2="890.396" y2="1419.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.29" y1="1203.85" x2="1271.29" y2="1187.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.29" y1="1203.85" x2="1271.29" y2="1219.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.29" y1="1203.85" x2="1303.29" y2="1219.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1287.29" y1="1203.85" x2="1303.29" y2="1187.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1098.58" y1="1260.55" x2="1082.58" y2="1244.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1098.58" y1="1260.55" x2="1082.58" y2="1276.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1098.58" y1="1260.55" x2="1114.58" y2="1276.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1098.58" y1="1260.55" x2="1114.58" y2="1244.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1330.6" y1="1013.95" x2="1314.6" y2="997.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1330.6" y1="1013.95" x2="1314.6" y2="1029.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1330.6" y1="1013.95" x2="1346.6" y2="1029.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1330.6" y1="1013.95" x2="1346.6" y2="997.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.5" y1="1058.8" x2="1070.5" y2="1042.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.5" y1="1058.8" x2="1070.5" y2="1074.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.5" y1="1058.8" x2="1102.5" y2="1074.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.5" y1="1058.8" x2="1102.5" y2="1042.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1250.5" y1="1043.25" x2="1234.5" y2="1027.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1250.5" y1="1043.25" x2="1234.5" y2="1059.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1250.5" y1="1043.25" x2="1266.5" y2="1059.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1250.5" y1="1043.25" x2="1266.5" y2="1027.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1626.12" y1="969.424" x2="1610.12" y2="953.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1626.12" y1="969.424" x2="1610.12" y2="985.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1626.12" y1="969.424" x2="1642.12" y2="985.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1626.12" y1="969.424" x2="1642.12" y2="953.424" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.09" y1="1136.96" x2="1468.09" y2="1120.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.09" y1="1136.96" x2="1468.09" y2="1152.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.09" y1="1136.96" x2="1500.09" y2="1152.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.09" y1="1136.96" x2="1500.09" y2="1120.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="933.271" y1="1353.8" x2="917.271" y2="1337.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="933.271" y1="1353.8" x2="917.271" y2="1369.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="933.271" y1="1353.8" x2="949.271" y2="1369.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="933.271" y1="1353.8" x2="949.271" y2="1337.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1027.37" y1="1280.74" x2="1011.37" y2="1264.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1027.37" y1="1280.74" x2="1011.37" y2="1296.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1027.37" y1="1280.74" x2="1043.37" y2="1296.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1027.37" y1="1280.74" x2="1043.37" y2="1264.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1346.23" y1="1131.49" x2="1330.23" y2="1115.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1346.23" y1="1131.49" x2="1330.23" y2="1147.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1346.23" y1="1131.49" x2="1362.23" y2="1147.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1346.23" y1="1131.49" x2="1362.23" y2="1115.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="906.639" y1="1451.01" x2="890.639" y2="1435.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="906.639" y1="1451.01" x2="890.639" y2="1467.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="906.639" y1="1451.01" x2="922.639" y2="1467.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="906.639" y1="1451.01" x2="922.639" y2="1435.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.76" y1="1104.6" x2="1405.76" y2="1088.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.76" y1="1104.6" x2="1405.76" y2="1120.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.76" y1="1104.6" x2="1437.76" y2="1120.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.76" y1="1104.6" x2="1437.76" y2="1088.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="958.996" y1="1406.94" x2="942.996" y2="1390.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="958.996" y1="1406.94" x2="942.996" y2="1422.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="958.996" y1="1406.94" x2="974.996" y2="1422.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="958.996" y1="1406.94" x2="974.996" y2="1390.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.277" y1="1312.14" x2="858.277" y2="1296.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.277" y1="1312.14" x2="858.277" y2="1328.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.277" y1="1312.14" x2="890.277" y2="1328.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="874.277" y1="1312.14" x2="890.277" y2="1296.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.691" y1="1424.89" x2="911.691" y2="1408.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.691" y1="1424.89" x2="911.691" y2="1440.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.691" y1="1424.89" x2="943.691" y2="1440.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.691" y1="1424.89" x2="943.691" y2="1408.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1412.66" y1="1067.26" x2="1396.66" y2="1051.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1412.66" y1="1067.26" x2="1396.66" y2="1083.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1412.66" y1="1067.26" x2="1428.66" y2="1083.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1412.66" y1="1067.26" x2="1428.66" y2="1051.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="538.801" y1="1550.28" x2="522.801" y2="1534.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="538.801" y1="1550.28" x2="522.801" y2="1566.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="538.801" y1="1550.28" x2="554.801" y2="1566.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="538.801" y1="1550.28" x2="554.801" y2="1534.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1294.59" y1="1109.44" x2="1278.59" y2="1093.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1294.59" y1="1109.44" x2="1278.59" y2="1125.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1294.59" y1="1109.44" x2="1310.59" y2="1125.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1294.59" y1="1109.44" x2="1310.59" y2="1093.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="869.402" y1="1512" x2="853.402" y2="1496" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="869.402" y1="1512" x2="853.402" y2="1528" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="869.402" y1="1512" x2="885.402" y2="1528" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="869.402" y1="1512" x2="885.402" y2="1496" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.688" y1="1677.8" x2="553.688" y2="1661.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.688" y1="1677.8" x2="553.688" y2="1693.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.688" y1="1677.8" x2="585.688" y2="1693.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.688" y1="1677.8" x2="585.688" y2="1661.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1124.14" y1="1170.71" x2="1108.14" y2="1154.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1124.14" y1="1170.71" x2="1108.14" y2="1186.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1124.14" y1="1170.71" x2="1140.14" y2="1186.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1124.14" y1="1170.71" x2="1140.14" y2="1154.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="915.573" y1="1448.5" x2="899.573" y2="1432.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="915.573" y1="1448.5" x2="899.573" y2="1464.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="915.573" y1="1448.5" x2="931.573" y2="1464.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="915.573" y1="1448.5" x2="931.573" y2="1432.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1520.69" y1="868.39" x2="1504.69" y2="852.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1520.69" y1="868.39" x2="1504.69" y2="884.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1520.69" y1="868.39" x2="1536.69" y2="884.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1520.69" y1="868.39" x2="1536.69" y2="852.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1108.04" y1="1185.31" x2="1092.04" y2="1169.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1108.04" y1="1185.31" x2="1092.04" y2="1201.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1108.04" y1="1185.31" x2="1124.04" y2="1201.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1108.04" y1="1185.31" x2="1124.04" y2="1169.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.79" y1="1007.52" x2="1446.79" y2="991.518" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.79" y1="1007.52" x2="1446.79" y2="1023.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.79" y1="1007.52" x2="1478.79" y2="1023.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1462.79" y1="1007.52" x2="1478.79" y2="991.518" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.71" y1="986.213" x2="1427.71" y2="970.213" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.71" y1="986.213" x2="1427.71" y2="1002.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.71" y1="986.213" x2="1459.71" y2="1002.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.71" y1="986.213" x2="1459.71" y2="970.213" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="831.587" y1="1424.87" x2="815.587" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="831.587" y1="1424.87" x2="815.587" y2="1440.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="831.587" y1="1424.87" x2="847.587" y2="1440.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="831.587" y1="1424.87" x2="847.587" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.58" y1="1341.26" x2="1014.58" y2="1325.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.58" y1="1341.26" x2="1014.58" y2="1357.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.58" y1="1341.26" x2="1046.58" y2="1357.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.58" y1="1341.26" x2="1046.58" y2="1325.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.23" y1="1063.53" x2="1436.23" y2="1047.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.23" y1="1063.53" x2="1436.23" y2="1079.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.23" y1="1063.53" x2="1468.23" y2="1079.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.23" y1="1063.53" x2="1468.23" y2="1047.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.66" y1="1110.16" x2="1062.66" y2="1094.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.66" y1="1110.16" x2="1062.66" y2="1126.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.66" y1="1110.16" x2="1094.66" y2="1126.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.66" y1="1110.16" x2="1094.66" y2="1094.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.93" y1="1006.01" x2="1593.93" y2="990.013" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.93" y1="1006.01" x2="1593.93" y2="1022.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.93" y1="1006.01" x2="1625.93" y2="1022.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.93" y1="1006.01" x2="1625.93" y2="990.013" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1733.98" y1="723.632" x2="1717.98" y2="707.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1733.98" y1="723.632" x2="1717.98" y2="739.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1733.98" y1="723.632" x2="1749.98" y2="739.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1733.98" y1="723.632" x2="1749.98" y2="707.632" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.47" y1="1228.99" x2="1265.47" y2="1212.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.47" y1="1228.99" x2="1265.47" y2="1244.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.47" y1="1228.99" x2="1297.47" y2="1244.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.47" y1="1228.99" x2="1297.47" y2="1212.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.93" y1="1191.01" x2="1213.93" y2="1175.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.93" y1="1191.01" x2="1213.93" y2="1207.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.93" y1="1191.01" x2="1245.93" y2="1207.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.93" y1="1191.01" x2="1245.93" y2="1175.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.48" y1="1238.33" x2="1224.48" y2="1222.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.48" y1="1238.33" x2="1224.48" y2="1254.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.48" y1="1238.33" x2="1256.48" y2="1254.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.48" y1="1238.33" x2="1256.48" y2="1222.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.936" y1="1351.49" x2="834.936" y2="1335.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.936" y1="1351.49" x2="834.936" y2="1367.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.936" y1="1351.49" x2="866.936" y2="1367.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="850.936" y1="1351.49" x2="866.936" y2="1335.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.92" y1="896.324" x2="1752.92" y2="880.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.92" y1="896.324" x2="1752.92" y2="912.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.92" y1="896.324" x2="1784.92" y2="912.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.92" y1="896.324" x2="1784.92" y2="880.324" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.86" y1="1264.5" x2="984.861" y2="1248.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.86" y1="1264.5" x2="984.861" y2="1280.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.86" y1="1264.5" x2="1016.86" y2="1280.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.86" y1="1264.5" x2="1016.86" y2="1248.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1022.7" y1="1343.63" x2="1006.7" y2="1327.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1022.7" y1="1343.63" x2="1006.7" y2="1359.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1022.7" y1="1343.63" x2="1038.7" y2="1359.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1022.7" y1="1343.63" x2="1038.7" y2="1327.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.15" y1="1293.76" x2="1051.15" y2="1277.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.15" y1="1293.76" x2="1051.15" y2="1309.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.15" y1="1293.76" x2="1083.15" y2="1309.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1067.15" y1="1293.76" x2="1083.15" y2="1277.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1109.19" y1="1130.27" x2="1093.19" y2="1114.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1109.19" y1="1130.27" x2="1093.19" y2="1146.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1109.19" y1="1130.27" x2="1125.19" y2="1146.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1109.19" y1="1130.27" x2="1125.19" y2="1114.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.88" y1="1116.12" x2="1399.88" y2="1100.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.88" y1="1116.12" x2="1399.88" y2="1132.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.88" y1="1116.12" x2="1431.88" y2="1132.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.88" y1="1116.12" x2="1431.88" y2="1100.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.02" y1="1086.3" x2="1442.02" y2="1070.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.02" y1="1086.3" x2="1442.02" y2="1102.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.02" y1="1086.3" x2="1474.02" y2="1102.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.02" y1="1086.3" x2="1474.02" y2="1070.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1138.81" y1="1184.29" x2="1122.81" y2="1168.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1138.81" y1="1184.29" x2="1122.81" y2="1200.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1138.81" y1="1184.29" x2="1154.81" y2="1200.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1138.81" y1="1184.29" x2="1154.81" y2="1168.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.12" y1="1087.76" x2="1183.12" y2="1071.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.12" y1="1087.76" x2="1183.12" y2="1103.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.12" y1="1087.76" x2="1215.12" y2="1103.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.12" y1="1087.76" x2="1215.12" y2="1071.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.48" y1="973.842" x2="1509.48" y2="957.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.48" y1="973.842" x2="1509.48" y2="989.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.48" y1="973.842" x2="1541.48" y2="989.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.48" y1="973.842" x2="1541.48" y2="957.842" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.75" y1="1324.55" x2="1091.75" y2="1308.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.75" y1="1324.55" x2="1091.75" y2="1340.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.75" y1="1324.55" x2="1123.75" y2="1340.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.75" y1="1324.55" x2="1123.75" y2="1308.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="814.5" y1="1271.95" x2="798.5" y2="1255.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="814.5" y1="1271.95" x2="798.5" y2="1287.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="814.5" y1="1271.95" x2="830.5" y2="1287.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="814.5" y1="1271.95" x2="830.5" y2="1255.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1369.5" y1="1060.06" x2="1353.5" y2="1044.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1369.5" y1="1060.06" x2="1353.5" y2="1076.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1369.5" y1="1060.06" x2="1385.5" y2="1076.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1369.5" y1="1060.06" x2="1385.5" y2="1044.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.95" y1="1174.4" x2="1227.95" y2="1158.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.95" y1="1174.4" x2="1227.95" y2="1190.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.95" y1="1174.4" x2="1259.95" y2="1190.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.95" y1="1174.4" x2="1259.95" y2="1158.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1181.71" y1="1301.58" x2="1165.71" y2="1285.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1181.71" y1="1301.58" x2="1165.71" y2="1317.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1181.71" y1="1301.58" x2="1197.71" y2="1317.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1181.71" y1="1301.58" x2="1197.71" y2="1285.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1834.69" y1="763.508" x2="1818.69" y2="747.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1834.69" y1="763.508" x2="1818.69" y2="779.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1834.69" y1="763.508" x2="1850.69" y2="779.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1834.69" y1="763.508" x2="1850.69" y2="747.508" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.65" y1="1160.58" x2="1462.65" y2="1144.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.65" y1="1160.58" x2="1462.65" y2="1176.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.65" y1="1160.58" x2="1494.65" y2="1176.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.65" y1="1160.58" x2="1494.65" y2="1144.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.04" y1="1278.88" x2="990.041" y2="1262.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.04" y1="1278.88" x2="990.041" y2="1294.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.04" y1="1278.88" x2="1022.04" y2="1294.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1006.04" y1="1278.88" x2="1022.04" y2="1262.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1268.84" y1="1258.35" x2="1252.84" y2="1242.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1268.84" y1="1258.35" x2="1252.84" y2="1274.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1268.84" y1="1258.35" x2="1284.84" y2="1274.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1268.84" y1="1258.35" x2="1284.84" y2="1242.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.05" y1="1167.1" x2="1279.05" y2="1151.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.05" y1="1167.1" x2="1279.05" y2="1183.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.05" y1="1167.1" x2="1311.05" y2="1183.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.05" y1="1167.1" x2="1311.05" y2="1151.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239.88" y1="1281.07" x2="1223.88" y2="1265.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239.88" y1="1281.07" x2="1223.88" y2="1297.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239.88" y1="1281.07" x2="1255.88" y2="1297.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239.88" y1="1281.07" x2="1255.88" y2="1265.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.19" y1="1165.26" x2="1296.19" y2="1149.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.19" y1="1165.26" x2="1296.19" y2="1181.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.19" y1="1165.26" x2="1328.19" y2="1181.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.19" y1="1165.26" x2="1328.19" y2="1149.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1424" y1="1023.62" x2="1408" y2="1007.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1424" y1="1023.62" x2="1408" y2="1039.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1424" y1="1023.62" x2="1440" y2="1039.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1424" y1="1023.62" x2="1440" y2="1007.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1028.72" y1="1432.67" x2="1012.72" y2="1416.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1028.72" y1="1432.67" x2="1012.72" y2="1448.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1028.72" y1="1432.67" x2="1044.72" y2="1448.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1028.72" y1="1432.67" x2="1044.72" y2="1416.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.75" y1="1174.67" x2="1292.75" y2="1158.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.75" y1="1174.67" x2="1292.75" y2="1190.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.75" y1="1174.67" x2="1324.75" y2="1190.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.75" y1="1174.67" x2="1324.75" y2="1158.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.12" y1="1148.69" x2="1221.12" y2="1132.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.12" y1="1148.69" x2="1221.12" y2="1164.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.12" y1="1148.69" x2="1253.12" y2="1164.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.12" y1="1148.69" x2="1253.12" y2="1132.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1618.21" y1="979.458" x2="1602.21" y2="963.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1618.21" y1="979.458" x2="1602.21" y2="995.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1618.21" y1="979.458" x2="1634.21" y2="995.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1618.21" y1="979.458" x2="1634.21" y2="963.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1279.69" y1="1082.34" x2="1263.69" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1279.69" y1="1082.34" x2="1263.69" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1279.69" y1="1082.34" x2="1295.69" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1279.69" y1="1082.34" x2="1295.69" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1393.4" y1="1152.94" x2="1377.4" y2="1136.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1393.4" y1="1152.94" x2="1377.4" y2="1168.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1393.4" y1="1152.94" x2="1409.4" y2="1168.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1393.4" y1="1152.94" x2="1409.4" y2="1136.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1005.39" y1="1284.51" x2="989.392" y2="1268.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1005.39" y1="1284.51" x2="989.392" y2="1300.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1005.39" y1="1284.51" x2="1021.39" y2="1300.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1005.39" y1="1284.51" x2="1021.39" y2="1268.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.58" y1="1254.71" x2="1222.58" y2="1238.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.58" y1="1254.71" x2="1222.58" y2="1270.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.58" y1="1254.71" x2="1254.58" y2="1270.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1238.58" y1="1254.71" x2="1254.58" y2="1238.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.98" y1="894.258" x2="1593.98" y2="878.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.98" y1="894.258" x2="1593.98" y2="910.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.98" y1="894.258" x2="1625.98" y2="910.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1609.98" y1="894.258" x2="1625.98" y2="878.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1497.3" y1="1096.59" x2="1481.3" y2="1080.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1497.3" y1="1096.59" x2="1481.3" y2="1112.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1497.3" y1="1096.59" x2="1513.3" y2="1112.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1497.3" y1="1096.59" x2="1513.3" y2="1080.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1387.69" y1="1064.12" x2="1371.69" y2="1048.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1387.69" y1="1064.12" x2="1371.69" y2="1080.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1387.69" y1="1064.12" x2="1403.69" y2="1080.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1387.69" y1="1064.12" x2="1403.69" y2="1048.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1471.44" y1="1131.3" x2="1455.44" y2="1115.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1471.44" y1="1131.3" x2="1455.44" y2="1147.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1471.44" y1="1131.3" x2="1487.44" y2="1147.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1471.44" y1="1131.3" x2="1487.44" y2="1115.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1597.07" y1="924.794" x2="1581.07" y2="908.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1597.07" y1="924.794" x2="1581.07" y2="940.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1597.07" y1="924.794" x2="1613.07" y2="940.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1597.07" y1="924.794" x2="1613.07" y2="908.794" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="882.48" y1="1367.74" x2="866.48" y2="1351.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="882.48" y1="1367.74" x2="866.48" y2="1383.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="882.48" y1="1367.74" x2="898.48" y2="1383.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="882.48" y1="1367.74" x2="898.48" y2="1351.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1285.26" y1="1113.9" x2="1269.26" y2="1097.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1285.26" y1="1113.9" x2="1269.26" y2="1129.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1285.26" y1="1113.9" x2="1301.26" y2="1129.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1285.26" y1="1113.9" x2="1301.26" y2="1097.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.2" y1="1043.06" x2="1413.2" y2="1027.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.2" y1="1043.06" x2="1413.2" y2="1059.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.2" y1="1043.06" x2="1445.2" y2="1059.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1429.2" y1="1043.06" x2="1445.2" y2="1027.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1413.65" y1="1215.29" x2="1397.65" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1413.65" y1="1215.29" x2="1397.65" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1413.65" y1="1215.29" x2="1429.65" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1413.65" y1="1215.29" x2="1429.65" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1527.95" y1="988.132" x2="1511.95" y2="972.132" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1527.95" y1="988.132" x2="1511.95" y2="1004.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1527.95" y1="988.132" x2="1543.95" y2="1004.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1527.95" y1="988.132" x2="1543.95" y2="972.132" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.23" y1="1117.88" x2="1412.23" y2="1101.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.23" y1="1117.88" x2="1412.23" y2="1133.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.23" y1="1117.88" x2="1444.23" y2="1133.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.23" y1="1117.88" x2="1444.23" y2="1101.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.14" y1="1034.93" x2="1497.14" y2="1018.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.14" y1="1034.93" x2="1497.14" y2="1050.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.14" y1="1034.93" x2="1529.14" y2="1050.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.14" y1="1034.93" x2="1529.14" y2="1018.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1464.41" y1="1058.74" x2="1448.41" y2="1042.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1464.41" y1="1058.74" x2="1448.41" y2="1074.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1464.41" y1="1058.74" x2="1480.41" y2="1074.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1464.41" y1="1058.74" x2="1480.41" y2="1042.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1766.37" y1="756.534" x2="1750.37" y2="740.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1766.37" y1="756.534" x2="1750.37" y2="772.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1766.37" y1="756.534" x2="1782.37" y2="772.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1766.37" y1="756.534" x2="1782.37" y2="740.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.09" y1="1182.64" x2="1321.09" y2="1166.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.09" y1="1182.64" x2="1321.09" y2="1198.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.09" y1="1182.64" x2="1353.09" y2="1198.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.09" y1="1182.64" x2="1353.09" y2="1166.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.04" y1="1303.89" x2="1050.04" y2="1287.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.04" y1="1303.89" x2="1050.04" y2="1319.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.04" y1="1303.89" x2="1082.04" y2="1319.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.04" y1="1303.89" x2="1082.04" y2="1287.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.46" y1="1067.55" x2="1561.46" y2="1051.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.46" y1="1067.55" x2="1561.46" y2="1083.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.46" y1="1067.55" x2="1593.46" y2="1083.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1577.46" y1="1067.55" x2="1593.46" y2="1051.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.74" y1="1243.98" x2="1121.74" y2="1227.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.74" y1="1243.98" x2="1121.74" y2="1259.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.74" y1="1243.98" x2="1153.74" y2="1259.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.74" y1="1243.98" x2="1153.74" y2="1227.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.72" y1="1098.59" x2="1349.72" y2="1082.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.72" y1="1098.59" x2="1349.72" y2="1114.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.72" y1="1098.59" x2="1381.72" y2="1114.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.72" y1="1098.59" x2="1381.72" y2="1082.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1422.2" y1="999.716" x2="1406.2" y2="983.716" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1422.2" y1="999.716" x2="1406.2" y2="1015.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1422.2" y1="999.716" x2="1438.2" y2="1015.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1422.2" y1="999.716" x2="1438.2" y2="983.716" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1381.71" y1="975.726" x2="1365.71" y2="959.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1381.71" y1="975.726" x2="1365.71" y2="991.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1381.71" y1="975.726" x2="1397.71" y2="991.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1381.71" y1="975.726" x2="1397.71" y2="959.726" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.037" y1="1422.12" x2="859.037" y2="1406.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.037" y1="1422.12" x2="859.037" y2="1438.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.037" y1="1422.12" x2="891.037" y2="1438.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.037" y1="1422.12" x2="891.037" y2="1406.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.48" y1="1128.32" x2="1100.48" y2="1112.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.48" y1="1128.32" x2="1100.48" y2="1144.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.48" y1="1128.32" x2="1132.48" y2="1144.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.48" y1="1128.32" x2="1132.48" y2="1112.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="898.259" y1="1319.39" x2="882.259" y2="1303.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="898.259" y1="1319.39" x2="882.259" y2="1335.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="898.259" y1="1319.39" x2="914.259" y2="1335.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="898.259" y1="1319.39" x2="914.259" y2="1303.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.31" y1="1035.06" x2="1415.31" y2="1019.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.31" y1="1035.06" x2="1415.31" y2="1051.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.31" y1="1035.06" x2="1447.31" y2="1051.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.31" y1="1035.06" x2="1447.31" y2="1019.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.19" y1="815.289" x2="1746.19" y2="799.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.19" y1="815.289" x2="1746.19" y2="831.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.19" y1="815.289" x2="1778.19" y2="831.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.19" y1="815.289" x2="1778.19" y2="799.289" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="843.736" y1="1421.71" x2="827.736" y2="1405.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="843.736" y1="1421.71" x2="827.736" y2="1437.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="843.736" y1="1421.71" x2="859.736" y2="1437.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="843.736" y1="1421.71" x2="859.736" y2="1405.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1329.12" y1="1076.38" x2="1313.12" y2="1060.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1329.12" y1="1076.38" x2="1313.12" y2="1092.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1329.12" y1="1076.38" x2="1345.12" y2="1092.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1329.12" y1="1076.38" x2="1345.12" y2="1060.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.51" y1="1094.73" x2="1325.51" y2="1078.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.51" y1="1094.73" x2="1325.51" y2="1110.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.51" y1="1094.73" x2="1357.51" y2="1110.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1341.51" y1="1094.73" x2="1357.51" y2="1078.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.39" y1="1147.76" x2="1209.39" y2="1131.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.39" y1="1147.76" x2="1209.39" y2="1163.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.39" y1="1147.76" x2="1241.39" y2="1163.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.39" y1="1147.76" x2="1241.39" y2="1131.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.09" y1="1248.47" x2="1075.09" y2="1232.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.09" y1="1248.47" x2="1075.09" y2="1264.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.09" y1="1248.47" x2="1107.09" y2="1264.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.09" y1="1248.47" x2="1107.09" y2="1232.47" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1399.57" y1="992.596" x2="1383.57" y2="976.596" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1399.57" y1="992.596" x2="1383.57" y2="1008.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1399.57" y1="992.596" x2="1415.57" y2="1008.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1399.57" y1="992.596" x2="1415.57" y2="976.596" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="712.506" y1="1466.98" x2="696.506" y2="1450.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="712.506" y1="1466.98" x2="696.506" y2="1482.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="712.506" y1="1466.98" x2="728.506" y2="1482.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="712.506" y1="1466.98" x2="728.506" y2="1450.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1038.8" y1="1215.97" x2="1022.8" y2="1199.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1038.8" y1="1215.97" x2="1022.8" y2="1231.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1038.8" y1="1215.97" x2="1054.8" y2="1231.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1038.8" y1="1215.97" x2="1054.8" y2="1199.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="996.845" y1="1298.95" x2="980.845" y2="1282.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="996.845" y1="1298.95" x2="980.845" y2="1314.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="996.845" y1="1298.95" x2="1012.85" y2="1314.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="996.845" y1="1298.95" x2="1012.85" y2="1282.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="764.769" y1="1428.81" x2="748.769" y2="1412.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="764.769" y1="1428.81" x2="748.769" y2="1444.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="764.769" y1="1428.81" x2="780.769" y2="1444.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="764.769" y1="1428.81" x2="780.769" y2="1412.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1481.71" y1="979.429" x2="1465.71" y2="963.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1481.71" y1="979.429" x2="1465.71" y2="995.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1481.71" y1="979.429" x2="1497.71" y2="995.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1481.71" y1="979.429" x2="1497.71" y2="963.429" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="794.022" y1="1392.87" x2="778.022" y2="1376.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="794.022" y1="1392.87" x2="778.022" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="794.022" y1="1392.87" x2="810.022" y2="1408.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="794.022" y1="1392.87" x2="810.022" y2="1376.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.54" y1="1232.13" x2="1099.54" y2="1216.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.54" y1="1232.13" x2="1099.54" y2="1248.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.54" y1="1232.13" x2="1131.54" y2="1248.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.54" y1="1232.13" x2="1131.54" y2="1216.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="989.1" y1="1361.68" x2="973.1" y2="1345.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="989.1" y1="1361.68" x2="973.1" y2="1377.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="989.1" y1="1361.68" x2="1005.1" y2="1377.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="989.1" y1="1361.68" x2="1005.1" y2="1345.68" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="658.223" y1="1594.72" x2="642.223" y2="1578.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="658.223" y1="1594.72" x2="642.223" y2="1610.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="658.223" y1="1594.72" x2="674.223" y2="1610.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="658.223" y1="1594.72" x2="674.223" y2="1578.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.13" y1="922.6" x2="1595.13" y2="906.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.13" y1="922.6" x2="1595.13" y2="938.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.13" y1="922.6" x2="1627.13" y2="938.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1611.13" y1="922.6" x2="1627.13" y2="906.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1325.74" y1="1073.3" x2="1309.74" y2="1057.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1325.74" y1="1073.3" x2="1309.74" y2="1089.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1325.74" y1="1073.3" x2="1341.74" y2="1089.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1325.74" y1="1073.3" x2="1341.74" y2="1057.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.65" y1="1061.83" x2="1438.65" y2="1045.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.65" y1="1061.83" x2="1438.65" y2="1077.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.65" y1="1061.83" x2="1470.65" y2="1077.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.65" y1="1061.83" x2="1470.65" y2="1045.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1270.59" y1="1028.09" x2="1254.59" y2="1012.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1270.59" y1="1028.09" x2="1254.59" y2="1044.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1270.59" y1="1028.09" x2="1286.59" y2="1044.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1270.59" y1="1028.09" x2="1286.59" y2="1012.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.91" y1="1068.7" x2="1318.91" y2="1052.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.91" y1="1068.7" x2="1318.91" y2="1084.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.91" y1="1068.7" x2="1350.91" y2="1084.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.91" y1="1068.7" x2="1350.91" y2="1052.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.75" y1="1086.14" x2="1324.75" y2="1070.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.75" y1="1086.14" x2="1324.75" y2="1102.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.75" y1="1086.14" x2="1356.75" y2="1102.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1340.75" y1="1086.14" x2="1356.75" y2="1070.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="968.976" y1="1423.08" x2="952.976" y2="1407.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="968.976" y1="1423.08" x2="952.976" y2="1439.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="968.976" y1="1423.08" x2="984.976" y2="1439.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="968.976" y1="1423.08" x2="984.976" y2="1407.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1451.83" y1="926.379" x2="1435.83" y2="910.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1451.83" y1="926.379" x2="1435.83" y2="942.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1451.83" y1="926.379" x2="1467.83" y2="942.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1451.83" y1="926.379" x2="1467.83" y2="910.379" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1420.21" y1="1170.17" x2="1404.21" y2="1154.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1420.21" y1="1170.17" x2="1404.21" y2="1186.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1420.21" y1="1170.17" x2="1436.21" y2="1186.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1420.21" y1="1170.17" x2="1436.21" y2="1154.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.55" y1="1324.46" x2="1014.55" y2="1308.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.55" y1="1324.46" x2="1014.55" y2="1340.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.55" y1="1324.46" x2="1046.55" y2="1340.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1030.55" y1="1324.46" x2="1046.55" y2="1308.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.85" y1="1057.07" x2="1398.85" y2="1041.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.85" y1="1057.07" x2="1398.85" y2="1073.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.85" y1="1057.07" x2="1430.85" y2="1073.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.85" y1="1057.07" x2="1430.85" y2="1041.07" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1037.88" y1="1309.43" x2="1021.88" y2="1293.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1037.88" y1="1309.43" x2="1021.88" y2="1325.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1037.88" y1="1309.43" x2="1053.88" y2="1325.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1037.88" y1="1309.43" x2="1053.88" y2="1293.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1171.81" y1="1298.75" x2="1155.81" y2="1282.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1171.81" y1="1298.75" x2="1155.81" y2="1314.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1171.81" y1="1298.75" x2="1187.81" y2="1314.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1171.81" y1="1298.75" x2="1187.81" y2="1282.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.86" y1="1223.11" x2="1097.86" y2="1207.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.86" y1="1223.11" x2="1097.86" y2="1239.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.86" y1="1223.11" x2="1129.86" y2="1239.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.86" y1="1223.11" x2="1129.86" y2="1207.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.45" y1="1086.98" x2="1444.45" y2="1070.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.45" y1="1086.98" x2="1444.45" y2="1102.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.45" y1="1086.98" x2="1476.45" y2="1102.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.45" y1="1086.98" x2="1476.45" y2="1070.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1557.36" y1="1088.65" x2="1541.36" y2="1072.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1557.36" y1="1088.65" x2="1541.36" y2="1104.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1557.36" y1="1088.65" x2="1573.36" y2="1104.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1557.36" y1="1088.65" x2="1573.36" y2="1072.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="908.822" y1="1392.95" x2="892.822" y2="1376.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="908.822" y1="1392.95" x2="892.822" y2="1408.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="908.822" y1="1392.95" x2="924.822" y2="1408.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="908.822" y1="1392.95" x2="924.822" y2="1376.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.57" y1="1011.37" x2="1484.57" y2="995.374" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.57" y1="1011.37" x2="1484.57" y2="1027.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.57" y1="1011.37" x2="1516.57" y2="1027.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1500.57" y1="1011.37" x2="1516.57" y2="995.374" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1002.47" y1="1234.46" x2="986.471" y2="1218.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1002.47" y1="1234.46" x2="986.471" y2="1250.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1002.47" y1="1234.46" x2="1018.47" y2="1250.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1002.47" y1="1234.46" x2="1018.47" y2="1218.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.95" y1="1013.72" x2="1442.95" y2="997.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.95" y1="1013.72" x2="1442.95" y2="1029.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.95" y1="1013.72" x2="1474.95" y2="1029.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1458.95" y1="1013.72" x2="1474.95" y2="997.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.63" y1="1328.09" x2="1043.63" y2="1312.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.63" y1="1328.09" x2="1043.63" y2="1344.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.63" y1="1328.09" x2="1075.63" y2="1344.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1059.63" y1="1328.09" x2="1075.63" y2="1312.09" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1046.37" x2="1288.99" y2="1030.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1046.37" x2="1288.99" y2="1062.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1046.37" x2="1320.99" y2="1062.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1304.99" y1="1046.37" x2="1320.99" y2="1030.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1016.99" y1="1290.74" x2="1000.99" y2="1274.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1016.99" y1="1290.74" x2="1000.99" y2="1306.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1016.99" y1="1290.74" x2="1032.99" y2="1306.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1016.99" y1="1290.74" x2="1032.99" y2="1274.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="892.734" y1="1460.48" x2="876.734" y2="1444.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="892.734" y1="1460.48" x2="876.734" y2="1476.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="892.734" y1="1460.48" x2="908.734" y2="1476.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="892.734" y1="1460.48" x2="908.734" y2="1444.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.22" y1="1187.04" x2="1030.22" y2="1171.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.22" y1="1187.04" x2="1030.22" y2="1203.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.22" y1="1187.04" x2="1062.22" y2="1203.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.22" y1="1187.04" x2="1062.22" y2="1171.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.88" y1="1292.71" x2="1060.88" y2="1276.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.88" y1="1292.71" x2="1060.88" y2="1308.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.88" y1="1292.71" x2="1092.88" y2="1308.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.88" y1="1292.71" x2="1092.88" y2="1276.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.79" y1="924.706" x2="1529.79" y2="908.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.79" y1="924.706" x2="1529.79" y2="940.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.79" y1="924.706" x2="1561.79" y2="940.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.79" y1="924.706" x2="1561.79" y2="908.706" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1003.69" y1="1262.65" x2="987.692" y2="1246.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1003.69" y1="1262.65" x2="987.692" y2="1278.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1003.69" y1="1262.65" x2="1019.69" y2="1278.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1003.69" y1="1262.65" x2="1019.69" y2="1246.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1245.12" y1="1104.1" x2="1229.12" y2="1088.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1245.12" y1="1104.1" x2="1229.12" y2="1120.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1245.12" y1="1104.1" x2="1261.12" y2="1120.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1245.12" y1="1104.1" x2="1261.12" y2="1088.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.07" y1="1264.22" x2="1097.07" y2="1248.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.07" y1="1264.22" x2="1097.07" y2="1280.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.07" y1="1264.22" x2="1129.07" y2="1280.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1113.07" y1="1264.22" x2="1129.07" y2="1248.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="862.589" y1="1433.49" x2="846.589" y2="1417.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="862.589" y1="1433.49" x2="846.589" y2="1449.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="862.589" y1="1433.49" x2="878.589" y2="1449.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="862.589" y1="1433.49" x2="878.589" y2="1417.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1335.3" y1="1153.33" x2="1319.3" y2="1137.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1335.3" y1="1153.33" x2="1319.3" y2="1169.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1335.3" y1="1153.33" x2="1351.3" y2="1169.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1335.3" y1="1153.33" x2="1351.3" y2="1137.33" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1521.7" y1="1018.9" x2="1505.7" y2="1002.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1521.7" y1="1018.9" x2="1505.7" y2="1034.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1521.7" y1="1018.9" x2="1537.7" y2="1034.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1521.7" y1="1018.9" x2="1537.7" y2="1002.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1384.65" y1="1259.56" x2="1368.65" y2="1243.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1384.65" y1="1259.56" x2="1368.65" y2="1275.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1384.65" y1="1259.56" x2="1400.65" y2="1275.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1384.65" y1="1259.56" x2="1400.65" y2="1243.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1366.24" y1="1165.25" x2="1350.24" y2="1149.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1366.24" y1="1165.25" x2="1350.24" y2="1181.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1366.24" y1="1165.25" x2="1382.24" y2="1181.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1366.24" y1="1165.25" x2="1382.24" y2="1149.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="522.504" y1="1512.98" x2="506.504" y2="1496.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="522.504" y1="1512.98" x2="506.504" y2="1528.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="522.504" y1="1512.98" x2="538.504" y2="1528.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="522.504" y1="1512.98" x2="538.504" y2="1496.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.208" y1="1309.37" x2="956.208" y2="1293.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.208" y1="1309.37" x2="956.208" y2="1325.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.208" y1="1309.37" x2="988.208" y2="1325.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.208" y1="1309.37" x2="988.208" y2="1293.37" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.36" y1="1054" x2="1160.36" y2="1038" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.36" y1="1054" x2="1160.36" y2="1070" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.36" y1="1054" x2="1192.36" y2="1070" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.36" y1="1054" x2="1192.36" y2="1038" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.15" y1="1063.66" x2="1299.15" y2="1047.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.15" y1="1063.66" x2="1299.15" y2="1079.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.15" y1="1063.66" x2="1331.15" y2="1079.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1315.15" y1="1063.66" x2="1331.15" y2="1047.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1235.21" y1="1142.27" x2="1219.21" y2="1126.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1235.21" y1="1142.27" x2="1219.21" y2="1158.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1235.21" y1="1142.27" x2="1251.21" y2="1158.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1235.21" y1="1142.27" x2="1251.21" y2="1126.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.46" y1="1190.1" x2="1213.46" y2="1174.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.46" y1="1190.1" x2="1213.46" y2="1206.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.46" y1="1190.1" x2="1245.46" y2="1206.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1229.46" y1="1190.1" x2="1245.46" y2="1174.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.451" y1="1406.48" x2="728.451" y2="1390.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.451" y1="1406.48" x2="728.451" y2="1422.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.451" y1="1406.48" x2="760.451" y2="1422.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.451" y1="1406.48" x2="760.451" y2="1390.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.07" y1="1218.16" x2="1081.07" y2="1202.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.07" y1="1218.16" x2="1081.07" y2="1234.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.07" y1="1218.16" x2="1113.07" y2="1234.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.07" y1="1218.16" x2="1113.07" y2="1202.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1012.88" y1="1293.81" x2="996.879" y2="1277.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1012.88" y1="1293.81" x2="996.879" y2="1309.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1012.88" y1="1293.81" x2="1028.88" y2="1309.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1012.88" y1="1293.81" x2="1028.88" y2="1277.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="277.68" y1="1755.15" x2="261.68" y2="1739.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="277.68" y1="1755.15" x2="261.68" y2="1771.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="277.68" y1="1755.15" x2="293.68" y2="1771.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="277.68" y1="1755.15" x2="293.68" y2="1739.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1166.29" y1="1406.05" x2="1150.29" y2="1390.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1166.29" y1="1406.05" x2="1150.29" y2="1422.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1166.29" y1="1406.05" x2="1182.29" y2="1422.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1166.29" y1="1406.05" x2="1182.29" y2="1390.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1178.56" y1="1242.2" x2="1162.56" y2="1226.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1178.56" y1="1242.2" x2="1162.56" y2="1258.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1178.56" y1="1242.2" x2="1194.56" y2="1258.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1178.56" y1="1242.2" x2="1194.56" y2="1226.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1498.04" y1="1011.64" x2="1482.04" y2="995.641" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1498.04" y1="1011.64" x2="1482.04" y2="1027.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1498.04" y1="1011.64" x2="1514.04" y2="1027.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1498.04" y1="1011.64" x2="1514.04" y2="995.641" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1380.98" y1="1081.87" x2="1364.98" y2="1065.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1380.98" y1="1081.87" x2="1364.98" y2="1097.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1380.98" y1="1081.87" x2="1396.98" y2="1097.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1380.98" y1="1081.87" x2="1396.98" y2="1065.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1189.27" y1="1225.02" x2="1173.27" y2="1209.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1189.27" y1="1225.02" x2="1173.27" y2="1241.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1189.27" y1="1225.02" x2="1205.27" y2="1241.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1189.27" y1="1225.02" x2="1205.27" y2="1209.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.33" y1="1091.79" x2="1337.33" y2="1075.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.33" y1="1091.79" x2="1337.33" y2="1107.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.33" y1="1091.79" x2="1369.33" y2="1107.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.33" y1="1091.79" x2="1369.33" y2="1075.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.57" y1="1214.17" x2="1175.57" y2="1198.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.57" y1="1214.17" x2="1175.57" y2="1230.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.57" y1="1214.17" x2="1207.57" y2="1230.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1191.57" y1="1214.17" x2="1207.57" y2="1198.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.19" y1="842.118" x2="1764.19" y2="826.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.19" y1="842.118" x2="1764.19" y2="858.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.19" y1="842.118" x2="1796.19" y2="858.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.19" y1="842.118" x2="1796.19" y2="826.118" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="855.143" y1="1461.56" x2="839.143" y2="1445.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="855.143" y1="1461.56" x2="839.143" y2="1477.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="855.143" y1="1461.56" x2="871.143" y2="1477.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="855.143" y1="1461.56" x2="871.143" y2="1445.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1487.2" y1="1215.29" x2="1471.2" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1487.2" y1="1215.29" x2="1471.2" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1487.2" y1="1215.29" x2="1503.2" y2="1231.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1487.2" y1="1215.29" x2="1503.2" y2="1199.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="941.026" y1="1398.61" x2="925.026" y2="1382.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="941.026" y1="1398.61" x2="925.026" y2="1414.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="941.026" y1="1398.61" x2="957.026" y2="1414.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="941.026" y1="1398.61" x2="957.026" y2="1382.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1491.9" y1="862.19" x2="1475.9" y2="846.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1491.9" y1="862.19" x2="1475.9" y2="878.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1491.9" y1="862.19" x2="1507.9" y2="878.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1491.9" y1="862.19" x2="1507.9" y2="846.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1045.21" y1="1271.31" x2="1029.21" y2="1255.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1045.21" y1="1271.31" x2="1029.21" y2="1287.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1045.21" y1="1271.31" x2="1061.21" y2="1287.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1045.21" y1="1271.31" x2="1061.21" y2="1255.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.16" y1="1192.34" x2="1262.16" y2="1176.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.16" y1="1192.34" x2="1262.16" y2="1208.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.16" y1="1192.34" x2="1294.16" y2="1208.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.16" y1="1192.34" x2="1294.16" y2="1176.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.72" y1="1184.27" x2="1262.72" y2="1168.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.72" y1="1184.27" x2="1262.72" y2="1200.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.72" y1="1184.27" x2="1294.72" y2="1200.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1278.72" y1="1184.27" x2="1294.72" y2="1168.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1614.72" y1="834.262" x2="1598.72" y2="818.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1614.72" y1="834.262" x2="1598.72" y2="850.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1614.72" y1="834.262" x2="1630.72" y2="850.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1614.72" y1="834.262" x2="1630.72" y2="818.262" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.83" y1="933.105" x2="1599.83" y2="917.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.83" y1="933.105" x2="1599.83" y2="949.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.83" y1="933.105" x2="1631.83" y2="949.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.83" y1="933.105" x2="1631.83" y2="917.105" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1306.51" y1="1052.44" x2="1290.51" y2="1036.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1306.51" y1="1052.44" x2="1290.51" y2="1068.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1306.51" y1="1052.44" x2="1322.51" y2="1068.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1306.51" y1="1052.44" x2="1322.51" y2="1036.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.55" y1="1095.87" x2="1376.55" y2="1079.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.55" y1="1095.87" x2="1376.55" y2="1111.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.55" y1="1095.87" x2="1408.55" y2="1111.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.55" y1="1095.87" x2="1408.55" y2="1079.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.951" y1="1271.4" x2="883.951" y2="1255.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.951" y1="1271.4" x2="883.951" y2="1287.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.951" y1="1271.4" x2="915.951" y2="1287.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.951" y1="1271.4" x2="915.951" y2="1255.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.68" y1="1069.21" x2="1376.68" y2="1053.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.68" y1="1069.21" x2="1376.68" y2="1085.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.68" y1="1069.21" x2="1408.68" y2="1085.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1392.68" y1="1069.21" x2="1408.68" y2="1053.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1651.8" y1="933.534" x2="1635.8" y2="917.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1651.8" y1="933.534" x2="1635.8" y2="949.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1651.8" y1="933.534" x2="1667.8" y2="949.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1651.8" y1="933.534" x2="1667.8" y2="917.534" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.2" y1="982.773" x2="1530.2" y2="966.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.2" y1="982.773" x2="1530.2" y2="998.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.2" y1="982.773" x2="1562.2" y2="998.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1546.2" y1="982.773" x2="1562.2" y2="966.773" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1197.19" y1="1159.24" x2="1181.19" y2="1143.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1197.19" y1="1159.24" x2="1181.19" y2="1175.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1197.19" y1="1159.24" x2="1213.19" y2="1175.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1197.19" y1="1159.24" x2="1213.19" y2="1143.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1378.06" y1="1142.61" x2="1362.06" y2="1126.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1378.06" y1="1142.61" x2="1362.06" y2="1158.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1378.06" y1="1142.61" x2="1394.06" y2="1158.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1378.06" y1="1142.61" x2="1394.06" y2="1126.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="798.462" y1="1475.78" x2="782.462" y2="1459.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="798.462" y1="1475.78" x2="782.462" y2="1491.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="798.462" y1="1475.78" x2="814.462" y2="1491.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="798.462" y1="1475.78" x2="814.462" y2="1459.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1653.74" y1="920.543" x2="1637.74" y2="904.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1653.74" y1="920.543" x2="1637.74" y2="936.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1653.74" y1="920.543" x2="1669.74" y2="936.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1653.74" y1="920.543" x2="1669.74" y2="904.543" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.58" y1="1157.39" x2="1062.58" y2="1141.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.58" y1="1157.39" x2="1062.58" y2="1173.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.58" y1="1157.39" x2="1094.58" y2="1173.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1078.58" y1="1157.39" x2="1094.58" y2="1141.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.44" y1="1151.58" x2="1405.44" y2="1135.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.44" y1="1151.58" x2="1405.44" y2="1167.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.44" y1="1151.58" x2="1437.44" y2="1167.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.44" y1="1151.58" x2="1437.44" y2="1135.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.19" y1="1152.96" x2="1277.19" y2="1136.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.19" y1="1152.96" x2="1277.19" y2="1168.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.19" y1="1152.96" x2="1309.19" y2="1168.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.19" y1="1152.96" x2="1309.19" y2="1136.96" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1307.01" y1="1135.59" x2="1291.01" y2="1119.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1307.01" y1="1135.59" x2="1291.01" y2="1151.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1307.01" y1="1135.59" x2="1323.01" y2="1151.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1307.01" y1="1135.59" x2="1323.01" y2="1119.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1157.68" y1="1240.17" x2="1141.68" y2="1224.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1157.68" y1="1240.17" x2="1141.68" y2="1256.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1157.68" y1="1240.17" x2="1173.68" y2="1256.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1157.68" y1="1240.17" x2="1173.68" y2="1224.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1322.34" y1="1098.52" x2="1306.34" y2="1082.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1322.34" y1="1098.52" x2="1306.34" y2="1114.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1322.34" y1="1098.52" x2="1338.34" y2="1114.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1322.34" y1="1098.52" x2="1338.34" y2="1082.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.24" y1="1200.4" x2="1231.24" y2="1184.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.24" y1="1200.4" x2="1231.24" y2="1216.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.24" y1="1200.4" x2="1263.24" y2="1216.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1247.24" y1="1200.4" x2="1263.24" y2="1184.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.59" y1="1055.14" x2="1348.59" y2="1039.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.59" y1="1055.14" x2="1348.59" y2="1071.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.59" y1="1055.14" x2="1380.59" y2="1071.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.59" y1="1055.14" x2="1380.59" y2="1039.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1640.29" y1="1056.93" x2="1624.29" y2="1040.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1640.29" y1="1056.93" x2="1624.29" y2="1072.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1640.29" y1="1056.93" x2="1656.29" y2="1072.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1640.29" y1="1056.93" x2="1656.29" y2="1040.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1286.58" y1="1059.6" x2="1270.58" y2="1043.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1286.58" y1="1059.6" x2="1270.58" y2="1075.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1286.58" y1="1059.6" x2="1302.58" y2="1075.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1286.58" y1="1059.6" x2="1302.58" y2="1043.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1550.14" y1="1004.56" x2="1534.14" y2="988.562" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1550.14" y1="1004.56" x2="1534.14" y2="1020.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1550.14" y1="1004.56" x2="1566.14" y2="1020.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1550.14" y1="1004.56" x2="1566.14" y2="988.562" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="955.148" y1="1370.3" x2="939.148" y2="1354.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="955.148" y1="1370.3" x2="939.148" y2="1386.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="955.148" y1="1370.3" x2="971.148" y2="1386.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="955.148" y1="1370.3" x2="971.148" y2="1354.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1845.45" y1="659.705" x2="1829.45" y2="643.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1845.45" y1="659.705" x2="1829.45" y2="675.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1845.45" y1="659.705" x2="1861.45" y2="675.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1845.45" y1="659.705" x2="1861.45" y2="643.705" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1817.11" y1="731.36" x2="1801.11" y2="715.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1817.11" y1="731.36" x2="1801.11" y2="747.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1817.11" y1="731.36" x2="1833.11" y2="747.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1817.11" y1="731.36" x2="1833.11" y2="715.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1838.37" y1="806.858" x2="1822.37" y2="790.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1838.37" y1="806.858" x2="1822.37" y2="822.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1838.37" y1="806.858" x2="1854.37" y2="822.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1838.37" y1="806.858" x2="1854.37" y2="790.858" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1132.81" y1="1166.63" x2="1116.81" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1132.81" y1="1166.63" x2="1116.81" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1132.81" y1="1166.63" x2="1148.81" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1132.81" y1="1166.63" x2="1148.81" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244" y1="1263.65" x2="1228" y2="1247.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244" y1="1263.65" x2="1228" y2="1279.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244" y1="1263.65" x2="1260" y2="1279.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244" y1="1263.65" x2="1260" y2="1247.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.43" y1="1066.19" x2="1274.43" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.43" y1="1066.19" x2="1274.43" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.43" y1="1066.19" x2="1306.43" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.43" y1="1066.19" x2="1306.43" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1558.7" y1="965.028" x2="1542.7" y2="949.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1558.7" y1="965.028" x2="1542.7" y2="981.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1558.7" y1="965.028" x2="1574.7" y2="981.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1558.7" y1="965.028" x2="1574.7" y2="949.028" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.76" y1="970.455" x2="1458.76" y2="954.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.76" y1="970.455" x2="1458.76" y2="986.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.76" y1="970.455" x2="1490.76" y2="986.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.76" y1="970.455" x2="1490.76" y2="954.455" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1682.62" y1="820.381" x2="1666.62" y2="804.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1682.62" y1="820.381" x2="1666.62" y2="836.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1682.62" y1="820.381" x2="1698.62" y2="836.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1682.62" y1="820.381" x2="1698.62" y2="804.381" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="872.16" y1="1370.98" x2="856.16" y2="1354.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="872.16" y1="1370.98" x2="856.16" y2="1386.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="872.16" y1="1370.98" x2="888.16" y2="1386.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="872.16" y1="1370.98" x2="888.16" y2="1354.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1155.08" y1="1163.95" x2="1139.08" y2="1147.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1155.08" y1="1163.95" x2="1139.08" y2="1179.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1155.08" y1="1163.95" x2="1171.08" y2="1179.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1155.08" y1="1163.95" x2="1171.08" y2="1147.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1316.44" y1="1111.88" x2="1300.44" y2="1095.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1316.44" y1="1111.88" x2="1300.44" y2="1127.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1316.44" y1="1111.88" x2="1332.44" y2="1127.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1316.44" y1="1111.88" x2="1332.44" y2="1095.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="866.851" y1="1331.78" x2="850.851" y2="1315.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="866.851" y1="1331.78" x2="850.851" y2="1347.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="866.851" y1="1331.78" x2="882.851" y2="1347.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="866.851" y1="1331.78" x2="882.851" y2="1315.78" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.47" y1="1146.54" x2="1438.47" y2="1130.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.47" y1="1146.54" x2="1438.47" y2="1162.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.47" y1="1146.54" x2="1470.47" y2="1162.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1454.47" y1="1146.54" x2="1470.47" y2="1130.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1345.73" y1="1121.54" x2="1329.73" y2="1105.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1345.73" y1="1121.54" x2="1329.73" y2="1137.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1345.73" y1="1121.54" x2="1361.73" y2="1137.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1345.73" y1="1121.54" x2="1361.73" y2="1105.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1604.79" y1="991.27" x2="1588.79" y2="975.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1604.79" y1="991.27" x2="1588.79" y2="1007.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1604.79" y1="991.27" x2="1620.79" y2="1007.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1604.79" y1="991.27" x2="1620.79" y2="975.27" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1425.4" y1="1018.53" x2="1409.4" y2="1002.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1425.4" y1="1018.53" x2="1409.4" y2="1034.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1425.4" y1="1018.53" x2="1441.4" y2="1034.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1425.4" y1="1018.53" x2="1441.4" y2="1002.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1177.09" y1="1221.26" x2="1161.09" y2="1205.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1177.09" y1="1221.26" x2="1161.09" y2="1237.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1177.09" y1="1221.26" x2="1193.09" y2="1237.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1177.09" y1="1221.26" x2="1193.09" y2="1205.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="436.181" y1="1793.59" x2="420.181" y2="1777.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="436.181" y1="1793.59" x2="420.181" y2="1809.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="436.181" y1="1793.59" x2="452.181" y2="1809.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="436.181" y1="1793.59" x2="452.181" y2="1777.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="957.079" y1="1394.58" x2="941.079" y2="1378.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="957.079" y1="1394.58" x2="941.079" y2="1410.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="957.079" y1="1394.58" x2="973.079" y2="1410.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="957.079" y1="1394.58" x2="973.079" y2="1378.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1427.73" y1="1093.18" x2="1411.73" y2="1077.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1427.73" y1="1093.18" x2="1411.73" y2="1109.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1427.73" y1="1093.18" x2="1443.73" y2="1109.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1427.73" y1="1093.18" x2="1443.73" y2="1077.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1348.36" y1="1249.7" x2="1332.36" y2="1233.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1348.36" y1="1249.7" x2="1332.36" y2="1265.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1348.36" y1="1249.7" x2="1364.36" y2="1265.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1348.36" y1="1249.7" x2="1364.36" y2="1233.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.366" y1="1495.39" x2="775.366" y2="1479.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.366" y1="1495.39" x2="775.366" y2="1511.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.366" y1="1495.39" x2="807.366" y2="1511.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.366" y1="1495.39" x2="807.366" y2="1479.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1636.9" y1="890.133" x2="1620.9" y2="874.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1636.9" y1="890.133" x2="1620.9" y2="906.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1636.9" y1="890.133" x2="1652.9" y2="906.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1636.9" y1="890.133" x2="1652.9" y2="874.133" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1256" y1="1205.58" x2="1240" y2="1189.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1256" y1="1205.58" x2="1240" y2="1221.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1256" y1="1205.58" x2="1272" y2="1221.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1256" y1="1205.58" x2="1272" y2="1189.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.44" y1="1065.3" x2="1412.44" y2="1049.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.44" y1="1065.3" x2="1412.44" y2="1081.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.44" y1="1065.3" x2="1444.44" y2="1081.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.44" y1="1065.3" x2="1444.44" y2="1049.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.03" y1="1294.95" x2="1030.03" y2="1278.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.03" y1="1294.95" x2="1030.03" y2="1310.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.03" y1="1294.95" x2="1062.03" y2="1310.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1046.03" y1="1294.95" x2="1062.03" y2="1278.95" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="721.091" y1="1561.75" x2="705.091" y2="1545.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="721.091" y1="1561.75" x2="705.091" y2="1577.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="721.091" y1="1561.75" x2="737.091" y2="1577.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="721.091" y1="1561.75" x2="737.091" y2="1545.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1085.36" y1="1148.29" x2="1069.36" y2="1132.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1085.36" y1="1148.29" x2="1069.36" y2="1164.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1085.36" y1="1148.29" x2="1101.36" y2="1164.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1085.36" y1="1148.29" x2="1101.36" y2="1132.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="823.473" y1="1446.58" x2="807.473" y2="1430.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="823.473" y1="1446.58" x2="807.473" y2="1462.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="823.473" y1="1446.58" x2="839.473" y2="1462.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="823.473" y1="1446.58" x2="839.473" y2="1430.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="887.822" y1="1375.21" x2="871.822" y2="1359.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="887.822" y1="1375.21" x2="871.822" y2="1391.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="887.822" y1="1375.21" x2="903.822" y2="1391.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="887.822" y1="1375.21" x2="903.822" y2="1359.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.18" y1="1169.39" x2="1147.18" y2="1153.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.18" y1="1169.39" x2="1147.18" y2="1185.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.18" y1="1169.39" x2="1179.18" y2="1185.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.18" y1="1169.39" x2="1179.18" y2="1153.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.39" y1="1014.06" x2="1468.39" y2="998.061" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.39" y1="1014.06" x2="1468.39" y2="1030.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.39" y1="1014.06" x2="1500.39" y2="1030.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1484.39" y1="1014.06" x2="1500.39" y2="998.061" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1685.1" y1="924.047" x2="1669.1" y2="908.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1685.1" y1="924.047" x2="1669.1" y2="940.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1685.1" y1="924.047" x2="1701.1" y2="940.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1685.1" y1="924.047" x2="1701.1" y2="908.047" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1070.78" y1="1266.71" x2="1054.78" y2="1250.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1070.78" y1="1266.71" x2="1054.78" y2="1282.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1070.78" y1="1266.71" x2="1086.78" y2="1282.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1070.78" y1="1266.71" x2="1086.78" y2="1250.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1379.51" y1="1049.2" x2="1363.51" y2="1033.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1379.51" y1="1049.2" x2="1363.51" y2="1065.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1379.51" y1="1049.2" x2="1395.51" y2="1065.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1379.51" y1="1049.2" x2="1395.51" y2="1033.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="877.025" y1="1403.06" x2="861.025" y2="1387.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="877.025" y1="1403.06" x2="861.025" y2="1419.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="877.025" y1="1403.06" x2="893.025" y2="1419.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="877.025" y1="1403.06" x2="893.025" y2="1387.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.37" y1="1191.22" x2="1060.37" y2="1175.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.37" y1="1191.22" x2="1060.37" y2="1207.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.37" y1="1191.22" x2="1092.37" y2="1207.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1076.37" y1="1191.22" x2="1092.37" y2="1175.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1164.55" y1="1249.8" x2="1148.55" y2="1233.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1164.55" y1="1249.8" x2="1148.55" y2="1265.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1164.55" y1="1249.8" x2="1180.55" y2="1265.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1164.55" y1="1249.8" x2="1180.55" y2="1233.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1473.39" y1="1000.62" x2="1457.39" y2="984.624" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1473.39" y1="1000.62" x2="1457.39" y2="1016.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1473.39" y1="1000.62" x2="1489.39" y2="1016.62" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1473.39" y1="1000.62" x2="1489.39" y2="984.624" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.54" y1="1202.19" x2="1279.54" y2="1186.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.54" y1="1202.19" x2="1279.54" y2="1218.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.54" y1="1202.19" x2="1311.54" y2="1218.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1295.54" y1="1202.19" x2="1311.54" y2="1186.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.58" y1="1267.5" x2="1100.58" y2="1251.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.58" y1="1267.5" x2="1100.58" y2="1283.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.58" y1="1267.5" x2="1132.58" y2="1283.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1116.58" y1="1267.5" x2="1132.58" y2="1251.5" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1435.37" y1="1059.7" x2="1419.37" y2="1043.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1435.37" y1="1059.7" x2="1419.37" y2="1075.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1435.37" y1="1059.7" x2="1451.37" y2="1075.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1435.37" y1="1059.7" x2="1451.37" y2="1043.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="930.593" y1="1251.3" x2="914.593" y2="1235.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="930.593" y1="1251.3" x2="914.593" y2="1267.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="930.593" y1="1251.3" x2="946.593" y2="1267.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="930.593" y1="1251.3" x2="946.593" y2="1235.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1125.99" y1="1256.74" x2="1109.99" y2="1240.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1125.99" y1="1256.74" x2="1109.99" y2="1272.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1125.99" y1="1256.74" x2="1141.99" y2="1272.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1125.99" y1="1256.74" x2="1141.99" y2="1240.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.3" y1="1267.31" x2="1105.3" y2="1251.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.3" y1="1267.31" x2="1105.3" y2="1283.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.3" y1="1267.31" x2="1137.3" y2="1283.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.3" y1="1267.31" x2="1137.3" y2="1251.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.55" y1="972.554" x2="1599.55" y2="956.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.55" y1="972.554" x2="1599.55" y2="988.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.55" y1="972.554" x2="1631.55" y2="988.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1615.55" y1="972.554" x2="1631.55" y2="956.554" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1395.64" y1="960.36" x2="1379.64" y2="944.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1395.64" y1="960.36" x2="1379.64" y2="976.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1395.64" y1="960.36" x2="1411.64" y2="976.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1395.64" y1="960.36" x2="1411.64" y2="944.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.08" y1="1312.36" x2="984.075" y2="1296.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.08" y1="1312.36" x2="984.075" y2="1328.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.08" y1="1312.36" x2="1016.08" y2="1328.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1000.08" y1="1312.36" x2="1016.08" y2="1296.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.75" y1="1068.03" x2="1337.75" y2="1052.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.75" y1="1068.03" x2="1337.75" y2="1084.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.75" y1="1068.03" x2="1369.75" y2="1084.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1353.75" y1="1068.03" x2="1369.75" y2="1052.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1126.95" y1="1234.81" x2="1110.95" y2="1218.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1126.95" y1="1234.81" x2="1110.95" y2="1250.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1126.95" y1="1234.81" x2="1142.95" y2="1250.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1126.95" y1="1234.81" x2="1142.95" y2="1218.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.72" y1="1159.69" x2="1347.72" y2="1143.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.72" y1="1159.69" x2="1347.72" y2="1175.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.72" y1="1159.69" x2="1379.72" y2="1175.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.72" y1="1159.69" x2="1379.72" y2="1143.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="805.605" y1="1368.59" x2="789.605" y2="1352.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="805.605" y1="1368.59" x2="789.605" y2="1384.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="805.605" y1="1368.59" x2="821.605" y2="1384.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="805.605" y1="1368.59" x2="821.605" y2="1352.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="735.913" y1="1350.58" x2="719.913" y2="1334.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="735.913" y1="1350.58" x2="719.913" y2="1366.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="735.913" y1="1350.58" x2="751.913" y2="1366.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="735.913" y1="1350.58" x2="751.913" y2="1334.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1057.78" y1="1211.94" x2="1041.78" y2="1195.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1057.78" y1="1211.94" x2="1041.78" y2="1227.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1057.78" y1="1211.94" x2="1073.78" y2="1227.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1057.78" y1="1211.94" x2="1073.78" y2="1195.94" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1826.24" y1="921.284" x2="1810.24" y2="905.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1826.24" y1="921.284" x2="1810.24" y2="937.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1826.24" y1="921.284" x2="1842.24" y2="937.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1826.24" y1="921.284" x2="1842.24" y2="905.284" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.2" y1="1094.53" x2="1277.2" y2="1078.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.2" y1="1094.53" x2="1277.2" y2="1110.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.2" y1="1094.53" x2="1309.2" y2="1110.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1293.2" y1="1094.53" x2="1309.2" y2="1078.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.65" y1="1361.41" x2="1091.65" y2="1345.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.65" y1="1361.41" x2="1091.65" y2="1377.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.65" y1="1361.41" x2="1123.65" y2="1377.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.65" y1="1361.41" x2="1123.65" y2="1345.41" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.07" y1="1240.9" x2="1160.07" y2="1224.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.07" y1="1240.9" x2="1160.07" y2="1256.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.07" y1="1240.9" x2="1192.07" y2="1256.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1176.07" y1="1240.9" x2="1192.07" y2="1224.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1215.2" y1="1190.81" x2="1199.2" y2="1174.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1215.2" y1="1190.81" x2="1199.2" y2="1206.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1215.2" y1="1190.81" x2="1231.2" y2="1206.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1215.2" y1="1190.81" x2="1231.2" y2="1174.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1572.67" y1="995.733" x2="1556.67" y2="979.733" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1572.67" y1="995.733" x2="1556.67" y2="1011.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1572.67" y1="995.733" x2="1588.67" y2="1011.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1572.67" y1="995.733" x2="1588.67" y2="979.733" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.41" y1="1231.66" x2="1348.41" y2="1215.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.41" y1="1231.66" x2="1348.41" y2="1247.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.41" y1="1231.66" x2="1380.41" y2="1247.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1364.41" y1="1231.66" x2="1380.41" y2="1215.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1259.03" y1="1156.39" x2="1243.03" y2="1140.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1259.03" y1="1156.39" x2="1243.03" y2="1172.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1259.03" y1="1156.39" x2="1275.03" y2="1172.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1259.03" y1="1156.39" x2="1275.03" y2="1140.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1469" y1="974.22" x2="1453" y2="958.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1469" y1="974.22" x2="1453" y2="990.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1469" y1="974.22" x2="1485" y2="990.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1469" y1="974.22" x2="1485" y2="958.22" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1292.53" y1="1131.39" x2="1276.53" y2="1115.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1292.53" y1="1131.39" x2="1276.53" y2="1147.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1292.53" y1="1131.39" x2="1308.53" y2="1147.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1292.53" y1="1131.39" x2="1308.53" y2="1115.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.04" y1="1299.49" x2="1130.04" y2="1283.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.04" y1="1299.49" x2="1130.04" y2="1315.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.04" y1="1299.49" x2="1162.04" y2="1315.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1146.04" y1="1299.49" x2="1162.04" y2="1283.49" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.06" y1="1219.15" x2="1070.06" y2="1203.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.06" y1="1219.15" x2="1070.06" y2="1235.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.06" y1="1219.15" x2="1102.06" y2="1235.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1086.06" y1="1219.15" x2="1102.06" y2="1203.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.64" y1="1000.17" x2="1752.64" y2="984.169" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.64" y1="1000.17" x2="1752.64" y2="1016.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.64" y1="1000.17" x2="1784.64" y2="1016.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1768.64" y1="1000.17" x2="1784.64" y2="984.169" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.35" y1="1164.06" x2="1349.35" y2="1148.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.35" y1="1164.06" x2="1349.35" y2="1180.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.35" y1="1164.06" x2="1381.35" y2="1180.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.35" y1="1164.06" x2="1381.35" y2="1148.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.6" y1="944.573" x2="1470.6" y2="928.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.6" y1="944.573" x2="1470.6" y2="960.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.6" y1="944.573" x2="1502.6" y2="960.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1486.6" y1="944.573" x2="1502.6" y2="928.573" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.16" y1="1210.82" x2="1257.16" y2="1194.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.16" y1="1210.82" x2="1257.16" y2="1226.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.16" y1="1210.82" x2="1289.16" y2="1226.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.16" y1="1210.82" x2="1289.16" y2="1194.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.92" y1="1149.92" x2="1246.92" y2="1133.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.92" y1="1149.92" x2="1246.92" y2="1165.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.92" y1="1149.92" x2="1278.92" y2="1165.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.92" y1="1149.92" x2="1278.92" y2="1133.92" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1122.29" y1="1146.61" x2="1106.29" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1122.29" y1="1146.61" x2="1106.29" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1122.29" y1="1146.61" x2="1138.29" y2="1162.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1122.29" y1="1146.61" x2="1138.29" y2="1130.61" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.41" y1="1171.23" x2="1275.41" y2="1155.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.41" y1="1171.23" x2="1275.41" y2="1187.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.41" y1="1171.23" x2="1307.41" y2="1187.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.41" y1="1171.23" x2="1307.41" y2="1155.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1448.29" y1="1128.98" x2="1432.29" y2="1112.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1448.29" y1="1128.98" x2="1432.29" y2="1144.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1448.29" y1="1128.98" x2="1464.29" y2="1144.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1448.29" y1="1128.98" x2="1464.29" y2="1112.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1130.44" y1="1149.23" x2="1114.44" y2="1133.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1130.44" y1="1149.23" x2="1114.44" y2="1165.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1130.44" y1="1149.23" x2="1146.44" y2="1165.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1130.44" y1="1149.23" x2="1146.44" y2="1133.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1039.51" y1="1295.52" x2="1023.51" y2="1279.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1039.51" y1="1295.52" x2="1023.51" y2="1311.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1039.51" y1="1295.52" x2="1055.51" y2="1311.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1039.51" y1="1295.52" x2="1055.51" y2="1279.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.81" y1="1008.74" x2="1427.81" y2="992.741" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.81" y1="1008.74" x2="1427.81" y2="1024.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.81" y1="1008.74" x2="1459.81" y2="1024.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1443.81" y1="1008.74" x2="1459.81" y2="992.741" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.99" y1="1097.56" x2="1529.99" y2="1081.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.99" y1="1097.56" x2="1529.99" y2="1113.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.99" y1="1097.56" x2="1561.99" y2="1113.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1545.99" y1="1097.56" x2="1561.99" y2="1081.56" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.83" y1="1094.72" x2="1343.83" y2="1078.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.83" y1="1094.72" x2="1343.83" y2="1110.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.83" y1="1094.72" x2="1375.83" y2="1110.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1359.83" y1="1094.72" x2="1375.83" y2="1078.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.3" y1="1374.86" x2="1183.3" y2="1358.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.3" y1="1374.86" x2="1183.3" y2="1390.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.3" y1="1374.86" x2="1215.3" y2="1390.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1199.3" y1="1374.86" x2="1215.3" y2="1358.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1133.36" y1="1257.88" x2="1117.36" y2="1241.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1133.36" y1="1257.88" x2="1117.36" y2="1273.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1133.36" y1="1257.88" x2="1149.36" y2="1273.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1133.36" y1="1257.88" x2="1149.36" y2="1241.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1628.36" y1="926.189" x2="1612.36" y2="910.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1628.36" y1="926.189" x2="1612.36" y2="942.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1628.36" y1="926.189" x2="1644.36" y2="942.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1628.36" y1="926.189" x2="1644.36" y2="910.189" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1073.35" y1="1320.7" x2="1057.35" y2="1304.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1073.35" y1="1320.7" x2="1057.35" y2="1336.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1073.35" y1="1320.7" x2="1089.35" y2="1336.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1073.35" y1="1320.7" x2="1089.35" y2="1304.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.88" y1="1332.35" x2="1075.88" y2="1316.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.88" y1="1332.35" x2="1075.88" y2="1348.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.88" y1="1332.35" x2="1107.88" y2="1348.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1091.88" y1="1332.35" x2="1107.88" y2="1316.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1017.09" y1="1323.45" x2="1001.09" y2="1307.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1017.09" y1="1323.45" x2="1001.09" y2="1339.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1017.09" y1="1323.45" x2="1033.09" y2="1339.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1017.09" y1="1323.45" x2="1033.09" y2="1307.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.938" y1="1534.93" x2="553.938" y2="1518.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.938" y1="1534.93" x2="553.938" y2="1550.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.938" y1="1534.93" x2="585.938" y2="1550.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="569.938" y1="1534.93" x2="585.938" y2="1518.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1231.45" y1="1166.31" x2="1215.45" y2="1150.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1231.45" y1="1166.31" x2="1215.45" y2="1182.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1231.45" y1="1166.31" x2="1247.45" y2="1182.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1231.45" y1="1166.31" x2="1247.45" y2="1150.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.218" y1="1483.05" x2="728.218" y2="1467.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.218" y1="1483.05" x2="728.218" y2="1499.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.218" y1="1483.05" x2="760.218" y2="1499.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="744.218" y1="1483.05" x2="760.218" y2="1467.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.81" y1="1264.02" x2="1169.81" y2="1248.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.81" y1="1264.02" x2="1169.81" y2="1280.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.81" y1="1264.02" x2="1201.81" y2="1280.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1185.81" y1="1264.02" x2="1201.81" y2="1248.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1713.33" y1="929.721" x2="1697.33" y2="913.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1713.33" y1="929.721" x2="1697.33" y2="945.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1713.33" y1="929.721" x2="1729.33" y2="945.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1713.33" y1="929.721" x2="1729.33" y2="913.721" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.62" y1="1058.71" x2="1292.62" y2="1042.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.62" y1="1058.71" x2="1292.62" y2="1074.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.62" y1="1058.71" x2="1324.62" y2="1074.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1308.62" y1="1058.71" x2="1324.62" y2="1042.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="610.516" y1="1568.23" x2="594.516" y2="1552.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="610.516" y1="1568.23" x2="594.516" y2="1584.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="610.516" y1="1568.23" x2="626.516" y2="1584.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="610.516" y1="1568.23" x2="626.516" y2="1552.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1702.63" y1="949.089" x2="1686.63" y2="933.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1702.63" y1="949.089" x2="1686.63" y2="965.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1702.63" y1="949.089" x2="1718.63" y2="965.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1702.63" y1="949.089" x2="1718.63" y2="933.089" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1102.82" y1="1200.06" x2="1086.82" y2="1184.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1102.82" y1="1200.06" x2="1086.82" y2="1216.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1102.82" y1="1200.06" x2="1118.82" y2="1216.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1102.82" y1="1200.06" x2="1118.82" y2="1184.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.22" y1="1125.36" x2="1444.22" y2="1109.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.22" y1="1125.36" x2="1444.22" y2="1141.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.22" y1="1125.36" x2="1476.22" y2="1141.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1460.22" y1="1125.36" x2="1476.22" y2="1109.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1242.85" y1="1068.4" x2="1226.85" y2="1052.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1242.85" y1="1068.4" x2="1226.85" y2="1084.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1242.85" y1="1068.4" x2="1258.85" y2="1084.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1242.85" y1="1068.4" x2="1258.85" y2="1052.4" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1180.26" y1="1392.12" x2="1164.26" y2="1376.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1180.26" y1="1392.12" x2="1164.26" y2="1408.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1180.26" y1="1392.12" x2="1196.26" y2="1408.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1180.26" y1="1392.12" x2="1196.26" y2="1376.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1799.12" y1="921.26" x2="1783.12" y2="905.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1799.12" y1="921.26" x2="1783.12" y2="937.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1799.12" y1="921.26" x2="1815.12" y2="937.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1799.12" y1="921.26" x2="1815.12" y2="905.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1198.35" y1="1213.97" x2="1182.35" y2="1197.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1198.35" y1="1213.97" x2="1182.35" y2="1229.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1198.35" y1="1213.97" x2="1214.35" y2="1229.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1198.35" y1="1213.97" x2="1214.35" y2="1197.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1314.16" y1="1231.53" x2="1298.16" y2="1215.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1314.16" y1="1231.53" x2="1298.16" y2="1247.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1314.16" y1="1231.53" x2="1330.16" y2="1247.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1314.16" y1="1231.53" x2="1330.16" y2="1215.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1298.91" y1="1128.6" x2="1282.91" y2="1112.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1298.91" y1="1128.6" x2="1282.91" y2="1144.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1298.91" y1="1128.6" x2="1314.91" y2="1144.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1298.91" y1="1128.6" x2="1314.91" y2="1112.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1539.91" y1="992.445" x2="1523.91" y2="976.445" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1539.91" y1="992.445" x2="1523.91" y2="1008.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1539.91" y1="992.445" x2="1555.91" y2="1008.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1539.91" y1="992.445" x2="1555.91" y2="976.445" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.49" y1="1165.67" x2="1192.49" y2="1149.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.49" y1="1165.67" x2="1192.49" y2="1181.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.49" y1="1165.67" x2="1224.49" y2="1181.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.49" y1="1165.67" x2="1224.49" y2="1149.67" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.4" y1="1164.76" x2="1275.4" y2="1148.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.4" y1="1164.76" x2="1275.4" y2="1180.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.4" y1="1164.76" x2="1307.4" y2="1180.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1291.4" y1="1164.76" x2="1307.4" y2="1148.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1509.7" y1="1053.86" x2="1493.7" y2="1037.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1509.7" y1="1053.86" x2="1493.7" y2="1069.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1509.7" y1="1053.86" x2="1525.7" y2="1069.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1509.7" y1="1053.86" x2="1525.7" y2="1037.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1586.87" y1="1027.39" x2="1570.87" y2="1011.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1586.87" y1="1027.39" x2="1570.87" y2="1043.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1586.87" y1="1027.39" x2="1602.87" y2="1043.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1586.87" y1="1027.39" x2="1602.87" y2="1011.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.89" y1="1264.65" x2="1095.89" y2="1248.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.89" y1="1264.65" x2="1095.89" y2="1280.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.89" y1="1264.65" x2="1127.89" y2="1280.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.89" y1="1264.65" x2="1127.89" y2="1248.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.824" y1="1370.17" x2="947.824" y2="1354.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.824" y1="1370.17" x2="947.824" y2="1386.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.824" y1="1370.17" x2="979.824" y2="1386.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.824" y1="1370.17" x2="979.824" y2="1354.17" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1269.64" y1="1053.76" x2="1253.64" y2="1037.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1269.64" y1="1053.76" x2="1253.64" y2="1069.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1269.64" y1="1053.76" x2="1285.64" y2="1069.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1269.64" y1="1053.76" x2="1285.64" y2="1037.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="694.471" y1="1399.51" x2="678.471" y2="1383.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="694.471" y1="1399.51" x2="678.471" y2="1415.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="694.471" y1="1399.51" x2="710.471" y2="1415.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="694.471" y1="1399.51" x2="710.471" y2="1383.51" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1318.45" y1="1163.46" x2="1302.45" y2="1147.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1318.45" y1="1163.46" x2="1302.45" y2="1179.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1318.45" y1="1163.46" x2="1334.45" y2="1179.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1318.45" y1="1163.46" x2="1334.45" y2="1147.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1386.71" y1="1019.77" x2="1370.71" y2="1003.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1386.71" y1="1019.77" x2="1370.71" y2="1035.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1386.71" y1="1019.77" x2="1402.71" y2="1035.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1386.71" y1="1019.77" x2="1402.71" y2="1003.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1547.26" y1="1096.34" x2="1531.26" y2="1080.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1547.26" y1="1096.34" x2="1531.26" y2="1112.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1547.26" y1="1096.34" x2="1563.26" y2="1112.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1547.26" y1="1096.34" x2="1563.26" y2="1080.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1054.12" y1="1246.84" x2="1038.12" y2="1230.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1054.12" y1="1246.84" x2="1038.12" y2="1262.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1054.12" y1="1246.84" x2="1070.12" y2="1262.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1054.12" y1="1246.84" x2="1070.12" y2="1230.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.9" y1="1186.98" x2="1257.9" y2="1170.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.9" y1="1186.98" x2="1257.9" y2="1202.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.9" y1="1186.98" x2="1289.9" y2="1202.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.9" y1="1186.98" x2="1289.9" y2="1170.98" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.32" y1="993.037" x2="1462.32" y2="977.037" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.32" y1="993.037" x2="1462.32" y2="1009.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.32" y1="993.037" x2="1494.32" y2="1009.04" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1478.32" y1="993.037" x2="1494.32" y2="977.037" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.95" y1="1279" x2="1264.95" y2="1263" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.95" y1="1279" x2="1264.95" y2="1295" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.95" y1="1279" x2="1296.95" y2="1295" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.95" y1="1279" x2="1296.95" y2="1263" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.65" y1="1130.42" x2="1307.65" y2="1114.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.65" y1="1130.42" x2="1307.65" y2="1146.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.65" y1="1130.42" x2="1339.65" y2="1146.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.65" y1="1130.42" x2="1339.65" y2="1114.42" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.8" y1="1126.84" x2="1265.8" y2="1110.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.8" y1="1126.84" x2="1265.8" y2="1142.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.8" y1="1126.84" x2="1297.8" y2="1142.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1281.8" y1="1126.84" x2="1297.8" y2="1110.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.44" y1="1134.89" x2="1318.44" y2="1118.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.44" y1="1134.89" x2="1318.44" y2="1150.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.44" y1="1134.89" x2="1350.44" y2="1150.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1334.44" y1="1134.89" x2="1350.44" y2="1118.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1266.08" y1="1197.16" x2="1250.08" y2="1181.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1266.08" y1="1197.16" x2="1250.08" y2="1213.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1266.08" y1="1197.16" x2="1282.08" y2="1213.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1266.08" y1="1197.16" x2="1282.08" y2="1181.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1459.46" y1="1078.54" x2="1443.46" y2="1062.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1459.46" y1="1078.54" x2="1443.46" y2="1094.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1459.46" y1="1078.54" x2="1475.46" y2="1094.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1459.46" y1="1078.54" x2="1475.46" y2="1062.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1650.12" y1="918.653" x2="1634.12" y2="902.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1650.12" y1="918.653" x2="1634.12" y2="934.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1650.12" y1="918.653" x2="1666.12" y2="934.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1650.12" y1="918.653" x2="1666.12" y2="902.653" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1061.4" y1="1273.43" x2="1045.4" y2="1257.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1061.4" y1="1273.43" x2="1045.4" y2="1289.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1061.4" y1="1273.43" x2="1077.4" y2="1289.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1061.4" y1="1273.43" x2="1077.4" y2="1257.43" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.92" y1="1166.84" x2="1214.92" y2="1150.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.92" y1="1166.84" x2="1214.92" y2="1182.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.92" y1="1166.84" x2="1246.92" y2="1182.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1230.92" y1="1166.84" x2="1246.92" y2="1150.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1860.89" y1="778.201" x2="1844.89" y2="762.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1860.89" y1="778.201" x2="1844.89" y2="794.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1860.89" y1="778.201" x2="1876.89" y2="794.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1860.89" y1="778.201" x2="1876.89" y2="762.201" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.65" y1="1264.26" x2="1224.65" y2="1248.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.65" y1="1264.26" x2="1224.65" y2="1280.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.65" y1="1264.26" x2="1256.65" y2="1280.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.65" y1="1264.26" x2="1256.65" y2="1248.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.89" y1="1136.28" x2="1264.89" y2="1120.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.89" y1="1136.28" x2="1264.89" y2="1152.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.89" y1="1136.28" x2="1296.89" y2="1152.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.89" y1="1136.28" x2="1296.89" y2="1120.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1444.37" y1="1001.52" x2="1428.37" y2="985.521" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1444.37" y1="1001.52" x2="1428.37" y2="1017.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1444.37" y1="1001.52" x2="1460.37" y2="1017.52" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1444.37" y1="1001.52" x2="1460.37" y2="985.521" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1455.79" y1="1193.99" x2="1439.79" y2="1177.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1455.79" y1="1193.99" x2="1439.79" y2="1209.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1455.79" y1="1193.99" x2="1471.79" y2="1209.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1455.79" y1="1193.99" x2="1471.79" y2="1177.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.41" y1="1124.9" x2="1105.41" y2="1108.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.41" y1="1124.9" x2="1105.41" y2="1140.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.41" y1="1124.9" x2="1137.41" y2="1140.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1121.41" y1="1124.9" x2="1137.41" y2="1108.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.75" y1="1012.66" x2="1412.75" y2="996.663" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.75" y1="1012.66" x2="1412.75" y2="1028.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.75" y1="1012.66" x2="1444.75" y2="1028.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.75" y1="1012.66" x2="1444.75" y2="996.663" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.98" y1="1162.81" x2="1121.98" y2="1146.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.98" y1="1162.81" x2="1121.98" y2="1178.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.98" y1="1162.81" x2="1153.98" y2="1178.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1137.98" y1="1162.81" x2="1153.98" y2="1146.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1896.85" y1="631.005" x2="1880.85" y2="615.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1896.85" y1="631.005" x2="1880.85" y2="647.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1896.85" y1="631.005" x2="1912.85" y2="647.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1896.85" y1="631.005" x2="1912.85" y2="615.005" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1863.03" y1="689.163" x2="1847.03" y2="673.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1863.03" y1="689.163" x2="1847.03" y2="705.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1863.03" y1="689.163" x2="1879.03" y2="705.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1863.03" y1="689.163" x2="1879.03" y2="673.163" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1492.82" y1="1052.48" x2="1476.82" y2="1036.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1492.82" y1="1052.48" x2="1476.82" y2="1068.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1492.82" y1="1052.48" x2="1508.82" y2="1068.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1492.82" y1="1052.48" x2="1508.82" y2="1036.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.58" y1="1162.34" x2="1228.58" y2="1146.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.58" y1="1162.34" x2="1228.58" y2="1178.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.58" y1="1162.34" x2="1260.58" y2="1178.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1244.58" y1="1162.34" x2="1260.58" y2="1146.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.53" y1="1236.06" x2="1307.53" y2="1220.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.53" y1="1236.06" x2="1307.53" y2="1252.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.53" y1="1236.06" x2="1339.53" y2="1252.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1323.53" y1="1236.06" x2="1339.53" y2="1220.06" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.45" y1="1106.65" x2="1369.45" y2="1090.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.45" y1="1106.65" x2="1369.45" y2="1122.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.45" y1="1106.65" x2="1401.45" y2="1122.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.45" y1="1106.65" x2="1401.45" y2="1090.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="973.556" y1="1249.11" x2="957.556" y2="1233.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="973.556" y1="1249.11" x2="957.556" y2="1265.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="973.556" y1="1249.11" x2="989.556" y2="1265.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="973.556" y1="1249.11" x2="989.556" y2="1233.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.42" y1="974.507" x2="1399.42" y2="958.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.42" y1="974.507" x2="1399.42" y2="990.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.42" y1="974.507" x2="1431.42" y2="990.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1415.42" y1="974.507" x2="1431.42" y2="958.507" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.69" y1="1084.02" x2="1436.69" y2="1068.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.69" y1="1084.02" x2="1436.69" y2="1100.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.69" y1="1084.02" x2="1468.69" y2="1100.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1452.69" y1="1084.02" x2="1468.69" y2="1068.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1114.28" y1="1270.08" x2="1098.28" y2="1254.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1114.28" y1="1270.08" x2="1098.28" y2="1286.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1114.28" y1="1270.08" x2="1130.28" y2="1286.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1114.28" y1="1270.08" x2="1130.28" y2="1254.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.1" y1="1367.01" x2="1264.1" y2="1351.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.1" y1="1367.01" x2="1264.1" y2="1383.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.1" y1="1367.01" x2="1296.1" y2="1383.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1280.1" y1="1367.01" x2="1296.1" y2="1351.01" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="961.172" y1="1297.9" x2="945.172" y2="1281.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="961.172" y1="1297.9" x2="945.172" y2="1313.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="961.172" y1="1297.9" x2="977.172" y2="1313.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="961.172" y1="1297.9" x2="977.172" y2="1281.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="828.067" y1="1395.72" x2="812.067" y2="1379.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="828.067" y1="1395.72" x2="812.067" y2="1411.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="828.067" y1="1395.72" x2="844.067" y2="1411.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="828.067" y1="1395.72" x2="844.067" y2="1379.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="812.031" y1="1484.71" x2="796.031" y2="1468.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="812.031" y1="1484.71" x2="796.031" y2="1500.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="812.031" y1="1484.71" x2="828.031" y2="1500.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="812.031" y1="1484.71" x2="828.031" y2="1468.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="901.731" y1="1295.1" x2="885.731" y2="1279.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="901.731" y1="1295.1" x2="885.731" y2="1311.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="901.731" y1="1295.1" x2="917.731" y2="1311.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="901.731" y1="1295.1" x2="917.731" y2="1279.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1331.96" y1="1041.81" x2="1315.96" y2="1025.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1331.96" y1="1041.81" x2="1315.96" y2="1057.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1331.96" y1="1041.81" x2="1347.96" y2="1057.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1331.96" y1="1041.81" x2="1347.96" y2="1025.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.94" y1="1077.84" x2="1347.94" y2="1061.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.94" y1="1077.84" x2="1347.94" y2="1093.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.94" y1="1077.84" x2="1379.94" y2="1093.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1363.94" y1="1077.84" x2="1379.94" y2="1061.84" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.853" y1="1331.79" x2="956.853" y2="1315.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.853" y1="1331.79" x2="956.853" y2="1347.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.853" y1="1331.79" x2="988.853" y2="1347.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="972.853" y1="1331.79" x2="988.853" y2="1315.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.97" y1="912.119" x2="1722.97" y2="896.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.97" y1="912.119" x2="1722.97" y2="928.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.97" y1="912.119" x2="1754.97" y2="928.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.97" y1="912.119" x2="1754.97" y2="896.119" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1468.77" y1="1104.3" x2="1452.77" y2="1088.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1468.77" y1="1104.3" x2="1452.77" y2="1120.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1468.77" y1="1104.3" x2="1484.77" y2="1120.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1468.77" y1="1104.3" x2="1484.77" y2="1088.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.63" y1="1168.13" x2="1036.63" y2="1152.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.63" y1="1168.13" x2="1036.63" y2="1184.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.63" y1="1168.13" x2="1068.63" y2="1184.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.63" y1="1168.13" x2="1068.63" y2="1152.13" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.71" y1="1040.19" x2="1386.71" y2="1024.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.71" y1="1040.19" x2="1386.71" y2="1056.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.71" y1="1040.19" x2="1418.71" y2="1056.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.71" y1="1040.19" x2="1418.71" y2="1024.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1370.93" y1="1101.26" x2="1354.93" y2="1085.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1370.93" y1="1101.26" x2="1354.93" y2="1117.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1370.93" y1="1101.26" x2="1386.93" y2="1117.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1370.93" y1="1101.26" x2="1386.93" y2="1085.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1534.29" y1="956.26" x2="1518.29" y2="940.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1534.29" y1="956.26" x2="1518.29" y2="972.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1534.29" y1="956.26" x2="1550.29" y2="972.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1534.29" y1="956.26" x2="1550.29" y2="940.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1470.32" y1="1144.8" x2="1454.32" y2="1128.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1470.32" y1="1144.8" x2="1454.32" y2="1160.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1470.32" y1="1144.8" x2="1486.32" y2="1160.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1470.32" y1="1144.8" x2="1486.32" y2="1128.8" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.177" y1="1437.3" x2="863.177" y2="1421.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.177" y1="1437.3" x2="863.177" y2="1453.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.177" y1="1437.3" x2="895.177" y2="1453.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.177" y1="1437.3" x2="895.177" y2="1421.3" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.308" y1="1302.19" x2="775.308" y2="1286.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.308" y1="1302.19" x2="775.308" y2="1318.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.308" y1="1302.19" x2="807.308" y2="1318.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="791.308" y1="1302.19" x2="807.308" y2="1286.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1182.59" y1="1128.11" x2="1166.59" y2="1112.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1182.59" y1="1128.11" x2="1166.59" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1182.59" y1="1128.11" x2="1198.59" y2="1144.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1182.59" y1="1128.11" x2="1198.59" y2="1112.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.44" y1="890.561" x2="1574.44" y2="874.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.44" y1="890.561" x2="1574.44" y2="906.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.44" y1="890.561" x2="1606.44" y2="906.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.44" y1="890.561" x2="1606.44" y2="874.561" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1083.01" y1="1250.31" x2="1067.01" y2="1234.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1083.01" y1="1250.31" x2="1067.01" y2="1266.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1083.01" y1="1250.31" x2="1099.01" y2="1266.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1083.01" y1="1250.31" x2="1099.01" y2="1234.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1729.02" y1="853.811" x2="1713.02" y2="837.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1729.02" y1="853.811" x2="1713.02" y2="869.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1729.02" y1="853.811" x2="1745.02" y2="869.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1729.02" y1="853.811" x2="1745.02" y2="837.811" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="936.651" y1="1397.2" x2="920.651" y2="1381.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="936.651" y1="1397.2" x2="920.651" y2="1413.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="936.651" y1="1397.2" x2="952.651" y2="1413.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="936.651" y1="1397.2" x2="952.651" y2="1381.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1141.39" y1="1273.55" x2="1125.39" y2="1257.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1141.39" y1="1273.55" x2="1125.39" y2="1289.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1141.39" y1="1273.55" x2="1157.39" y2="1289.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1141.39" y1="1273.55" x2="1157.39" y2="1257.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.12" y1="1220.55" x2="1050.12" y2="1204.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.12" y1="1220.55" x2="1050.12" y2="1236.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.12" y1="1220.55" x2="1082.12" y2="1236.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1066.12" y1="1220.55" x2="1082.12" y2="1204.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="856.289" y1="1386.44" x2="840.289" y2="1370.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="856.289" y1="1386.44" x2="840.289" y2="1402.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="856.289" y1="1386.44" x2="872.289" y2="1402.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="856.289" y1="1386.44" x2="872.289" y2="1370.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1450.3" y1="1197.55" x2="1434.3" y2="1181.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1450.3" y1="1197.55" x2="1434.3" y2="1213.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1450.3" y1="1197.55" x2="1466.3" y2="1213.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1450.3" y1="1197.55" x2="1466.3" y2="1181.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1350.38" y1="930.437" x2="1334.38" y2="914.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1350.38" y1="930.437" x2="1334.38" y2="946.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1350.38" y1="930.437" x2="1366.38" y2="946.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1350.38" y1="930.437" x2="1366.38" y2="914.437" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1529.58" y1="996.113" x2="1513.58" y2="980.113" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1529.58" y1="996.113" x2="1513.58" y2="1012.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1529.58" y1="996.113" x2="1545.58" y2="1012.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1529.58" y1="996.113" x2="1545.58" y2="980.113" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.87" y1="1216.87" x2="1036.87" y2="1200.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.87" y1="1216.87" x2="1036.87" y2="1232.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.87" y1="1216.87" x2="1068.87" y2="1232.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1052.87" y1="1216.87" x2="1068.87" y2="1200.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.28" y1="1112.64" x2="1209.28" y2="1096.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.28" y1="1112.64" x2="1209.28" y2="1128.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.28" y1="1112.64" x2="1241.28" y2="1128.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1225.28" y1="1112.64" x2="1241.28" y2="1096.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.47" y1="955.973" x2="1497.47" y2="939.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.47" y1="955.973" x2="1497.47" y2="971.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.47" y1="955.973" x2="1529.47" y2="971.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.47" y1="955.973" x2="1529.47" y2="939.973" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.93" y1="1082.34" x2="1458.93" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.93" y1="1082.34" x2="1458.93" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.93" y1="1082.34" x2="1490.93" y2="1098.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1474.93" y1="1082.34" x2="1490.93" y2="1066.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1530.22" y1="1008.88" x2="1514.22" y2="992.884" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1530.22" y1="1008.88" x2="1514.22" y2="1024.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1530.22" y1="1008.88" x2="1546.22" y2="1024.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1530.22" y1="1008.88" x2="1546.22" y2="992.884" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.75" y1="828.151" x2="1574.75" y2="812.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.75" y1="828.151" x2="1574.75" y2="844.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.75" y1="828.151" x2="1606.75" y2="844.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1590.75" y1="828.151" x2="1606.75" y2="812.151" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.49" y1="1238.26" x2="1151.49" y2="1222.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.49" y1="1238.26" x2="1151.49" y2="1254.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.49" y1="1238.26" x2="1183.49" y2="1254.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1167.49" y1="1238.26" x2="1183.49" y2="1222.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.11" y1="1021.15" x2="1328.11" y2="1005.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.11" y1="1021.15" x2="1328.11" y2="1037.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.11" y1="1021.15" x2="1360.11" y2="1037.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1344.11" y1="1021.15" x2="1360.11" y2="1005.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1942.19" y1="883.328" x2="1926.19" y2="867.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1942.19" y1="883.328" x2="1926.19" y2="899.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1942.19" y1="883.328" x2="1958.19" y2="899.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1942.19" y1="883.328" x2="1958.19" y2="867.328" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="800.484" y1="1327.86" x2="784.484" y2="1311.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="800.484" y1="1327.86" x2="784.484" y2="1343.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="800.484" y1="1327.86" x2="816.484" y2="1343.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="800.484" y1="1327.86" x2="816.484" y2="1311.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.479" y1="1361.21" x2="863.479" y2="1345.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.479" y1="1361.21" x2="863.479" y2="1377.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.479" y1="1361.21" x2="895.479" y2="1377.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="879.479" y1="1361.21" x2="895.479" y2="1345.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1621.59" y1="888.584" x2="1605.59" y2="872.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1621.59" y1="888.584" x2="1605.59" y2="904.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1621.59" y1="888.584" x2="1637.59" y2="904.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1621.59" y1="888.584" x2="1637.59" y2="872.584" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1351.96" y1="1166.63" x2="1335.96" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1351.96" y1="1166.63" x2="1335.96" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1351.96" y1="1166.63" x2="1367.96" y2="1182.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1351.96" y1="1166.63" x2="1367.96" y2="1150.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.03" y1="917.837" x2="1589.03" y2="901.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.03" y1="917.837" x2="1589.03" y2="933.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.03" y1="917.837" x2="1621.03" y2="933.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1605.03" y1="917.837" x2="1621.03" y2="901.837" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.5" y1="1173.02" x2="1238.5" y2="1157.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.5" y1="1173.02" x2="1238.5" y2="1189.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.5" y1="1173.02" x2="1270.5" y2="1189.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.5" y1="1173.02" x2="1270.5" y2="1157.02" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.95" y1="1137.6" x2="1509.95" y2="1121.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.95" y1="1137.6" x2="1509.95" y2="1153.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.95" y1="1137.6" x2="1541.95" y2="1153.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1525.95" y1="1137.6" x2="1541.95" y2="1121.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="863.983" y1="1256.2" x2="847.983" y2="1240.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="863.983" y1="1256.2" x2="847.983" y2="1272.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="863.983" y1="1256.2" x2="879.983" y2="1272.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="863.983" y1="1256.2" x2="879.983" y2="1240.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1763.1" y1="789.55" x2="1747.1" y2="773.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1763.1" y1="789.55" x2="1747.1" y2="805.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1763.1" y1="789.55" x2="1779.1" y2="805.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1763.1" y1="789.55" x2="1779.1" y2="773.55" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.65" y1="1335.83" x2="1149.65" y2="1319.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.65" y1="1335.83" x2="1149.65" y2="1351.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.65" y1="1335.83" x2="1181.65" y2="1351.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.65" y1="1335.83" x2="1181.65" y2="1319.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.21" y1="1265.21" x2="1091.21" y2="1249.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.21" y1="1265.21" x2="1091.21" y2="1281.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.21" y1="1265.21" x2="1123.21" y2="1281.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1107.21" y1="1265.21" x2="1123.21" y2="1249.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1261.91" y1="1136.73" x2="1245.91" y2="1120.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1261.91" y1="1136.73" x2="1245.91" y2="1152.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1261.91" y1="1136.73" x2="1277.91" y2="1152.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1261.91" y1="1136.73" x2="1277.91" y2="1120.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1840.32" y1="839.458" x2="1824.32" y2="823.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1840.32" y1="839.458" x2="1824.32" y2="855.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1840.32" y1="839.458" x2="1856.32" y2="855.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1840.32" y1="839.458" x2="1856.32" y2="823.458" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1566.73" y1="937.472" x2="1550.73" y2="921.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1566.73" y1="937.472" x2="1550.73" y2="953.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1566.73" y1="937.472" x2="1582.73" y2="953.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1566.73" y1="937.472" x2="1582.73" y2="921.472" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.17" y1="997.851" x2="1441.17" y2="981.851" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.17" y1="997.851" x2="1441.17" y2="1013.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.17" y1="997.851" x2="1473.17" y2="1013.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.17" y1="997.851" x2="1473.17" y2="981.851" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1518.09" y1="1097.6" x2="1502.09" y2="1081.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1518.09" y1="1097.6" x2="1502.09" y2="1113.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1518.09" y1="1097.6" x2="1534.09" y2="1113.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1518.09" y1="1097.6" x2="1534.09" y2="1081.6" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.74" y1="1166.18" x2="1246.74" y2="1150.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.74" y1="1166.18" x2="1246.74" y2="1182.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.74" y1="1166.18" x2="1278.74" y2="1182.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1262.74" y1="1166.18" x2="1278.74" y2="1150.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="725.698" y1="1501.31" x2="709.698" y2="1485.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="725.698" y1="1501.31" x2="709.698" y2="1517.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="725.698" y1="1501.31" x2="741.698" y2="1517.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="725.698" y1="1501.31" x2="741.698" y2="1485.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="804.197" y1="1345.85" x2="788.197" y2="1329.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="804.197" y1="1345.85" x2="788.197" y2="1361.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="804.197" y1="1345.85" x2="820.197" y2="1361.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="804.197" y1="1345.85" x2="820.197" y2="1329.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1068.71" y1="1367.71" x2="1052.71" y2="1351.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1068.71" y1="1367.71" x2="1052.71" y2="1383.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1068.71" y1="1367.71" x2="1084.71" y2="1383.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1068.71" y1="1367.71" x2="1084.71" y2="1351.71" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.44" y1="832.688" x2="1722.44" y2="816.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.44" y1="832.688" x2="1722.44" y2="848.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.44" y1="832.688" x2="1754.44" y2="848.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1738.44" y1="832.688" x2="1754.44" y2="816.688" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1575.07" y1="927.362" x2="1559.07" y2="911.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1575.07" y1="927.362" x2="1559.07" y2="943.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1575.07" y1="927.362" x2="1591.07" y2="943.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1575.07" y1="927.362" x2="1591.07" y2="911.362" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.369" y1="1393.26" x2="947.369" y2="1377.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.369" y1="1393.26" x2="947.369" y2="1409.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.369" y1="1393.26" x2="979.369" y2="1409.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.369" y1="1393.26" x2="979.369" y2="1377.26" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1338.29" y1="1101.82" x2="1322.29" y2="1085.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1338.29" y1="1101.82" x2="1322.29" y2="1117.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1338.29" y1="1101.82" x2="1354.29" y2="1117.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1338.29" y1="1101.82" x2="1354.29" y2="1085.82" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.77" y1="1157.2" x2="1326.77" y2="1141.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.77" y1="1157.2" x2="1326.77" y2="1173.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.77" y1="1157.2" x2="1358.77" y2="1173.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.77" y1="1157.2" x2="1358.77" y2="1141.2" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1465.22" y1="1059.54" x2="1449.22" y2="1043.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1465.22" y1="1059.54" x2="1449.22" y2="1075.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1465.22" y1="1059.54" x2="1481.22" y2="1075.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1465.22" y1="1059.54" x2="1481.22" y2="1043.54" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1667.21" y1="967.122" x2="1651.21" y2="951.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1667.21" y1="967.122" x2="1651.21" y2="983.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1667.21" y1="967.122" x2="1683.21" y2="983.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1667.21" y1="967.122" x2="1683.21" y2="951.122" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1808.12" y1="803.64" x2="1792.12" y2="787.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1808.12" y1="803.64" x2="1792.12" y2="819.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1808.12" y1="803.64" x2="1824.12" y2="819.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1808.12" y1="803.64" x2="1824.12" y2="787.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.84" y1="1113.91" x2="1349.84" y2="1097.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.84" y1="1113.91" x2="1349.84" y2="1129.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.84" y1="1113.91" x2="1381.84" y2="1129.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1365.84" y1="1113.91" x2="1381.84" y2="1097.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.457" y1="1364.93" x2="911.457" y2="1348.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.457" y1="1364.93" x2="911.457" y2="1380.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.457" y1="1364.93" x2="943.457" y2="1380.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="927.457" y1="1364.93" x2="943.457" y2="1348.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.588" y1="1394.79" x2="938.588" y2="1378.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.588" y1="1394.79" x2="938.588" y2="1410.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.588" y1="1394.79" x2="970.588" y2="1410.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.588" y1="1394.79" x2="970.588" y2="1378.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1600.82" y1="848.267" x2="1584.82" y2="832.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1600.82" y1="848.267" x2="1584.82" y2="864.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1600.82" y1="848.267" x2="1616.82" y2="864.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1600.82" y1="848.267" x2="1616.82" y2="832.267" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1013.05" y1="1254.21" x2="997.055" y2="1238.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1013.05" y1="1254.21" x2="997.055" y2="1270.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1013.05" y1="1254.21" x2="1029.05" y2="1270.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1013.05" y1="1254.21" x2="1029.05" y2="1238.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.35" y1="1098.11" x2="1147.35" y2="1082.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.35" y1="1098.11" x2="1147.35" y2="1114.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.35" y1="1098.11" x2="1179.35" y2="1114.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1163.35" y1="1098.11" x2="1179.35" y2="1082.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1400.1" y1="1101.9" x2="1384.1" y2="1085.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1400.1" y1="1101.9" x2="1384.1" y2="1117.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1400.1" y1="1101.9" x2="1416.1" y2="1117.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1400.1" y1="1101.9" x2="1416.1" y2="1085.9" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="710.768" y1="1471.31" x2="694.768" y2="1455.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="710.768" y1="1471.31" x2="694.768" y2="1487.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="710.768" y1="1471.31" x2="726.768" y2="1487.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="710.768" y1="1471.31" x2="726.768" y2="1455.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.455" y1="1241.66" x2="938.455" y2="1225.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.455" y1="1241.66" x2="938.455" y2="1257.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.455" y1="1241.66" x2="970.455" y2="1257.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="954.455" y1="1241.66" x2="970.455" y2="1225.66" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.63" y1="1188.99" x2="1081.63" y2="1172.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.63" y1="1188.99" x2="1081.63" y2="1204.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.63" y1="1188.99" x2="1113.63" y2="1204.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1097.63" y1="1188.99" x2="1113.63" y2="1172.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.96" y1="1069.03" x2="1369.96" y2="1053.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.96" y1="1069.03" x2="1369.96" y2="1085.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.96" y1="1069.03" x2="1401.96" y2="1085.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1385.96" y1="1069.03" x2="1401.96" y2="1053.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.19" y1="997.572" x2="1412.19" y2="981.572" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.19" y1="997.572" x2="1412.19" y2="1013.57" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.19" y1="997.572" x2="1444.19" y2="1013.57" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1428.19" y1="997.572" x2="1444.19" y2="981.572" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1023.75" y1="1232.23" x2="1007.75" y2="1216.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1023.75" y1="1232.23" x2="1007.75" y2="1248.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1023.75" y1="1232.23" x2="1039.75" y2="1248.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1023.75" y1="1232.23" x2="1039.75" y2="1216.23" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1336.01" y1="1021.11" x2="1320.01" y2="1005.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1336.01" y1="1021.11" x2="1320.01" y2="1037.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1336.01" y1="1021.11" x2="1352.01" y2="1037.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1336.01" y1="1021.11" x2="1352.01" y2="1005.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.06" y1="1198.05" x2="1326.06" y2="1182.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.06" y1="1198.05" x2="1326.06" y2="1214.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.06" y1="1198.05" x2="1358.06" y2="1214.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1342.06" y1="1198.05" x2="1358.06" y2="1182.05" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1401.11" y1="1144.76" x2="1385.11" y2="1128.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1401.11" y1="1144.76" x2="1385.11" y2="1160.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1401.11" y1="1144.76" x2="1417.11" y2="1160.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1401.11" y1="1144.76" x2="1417.11" y2="1128.76" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.53" y1="1066.19" x2="1386.53" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.53" y1="1066.19" x2="1386.53" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.53" y1="1066.19" x2="1418.53" y2="1082.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1402.53" y1="1066.19" x2="1418.53" y2="1050.19" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1117.37" y1="1261.72" x2="1101.37" y2="1245.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1117.37" y1="1261.72" x2="1101.37" y2="1277.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1117.37" y1="1261.72" x2="1133.37" y2="1277.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1117.37" y1="1261.72" x2="1133.37" y2="1245.72" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.68" y1="1081.7" x2="1415.68" y2="1065.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.68" y1="1081.7" x2="1415.68" y2="1097.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.68" y1="1081.7" x2="1447.68" y2="1097.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1431.68" y1="1081.7" x2="1447.68" y2="1065.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="907.609" y1="1313.11" x2="891.609" y2="1297.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="907.609" y1="1313.11" x2="891.609" y2="1329.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="907.609" y1="1313.11" x2="923.609" y2="1329.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="907.609" y1="1313.11" x2="923.609" y2="1297.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1267.73" y1="1057.77" x2="1251.73" y2="1041.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1267.73" y1="1057.77" x2="1251.73" y2="1073.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1267.73" y1="1057.77" x2="1283.73" y2="1073.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1267.73" y1="1057.77" x2="1283.73" y2="1041.77" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="904.55" y1="1372.25" x2="888.55" y2="1356.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="904.55" y1="1372.25" x2="888.55" y2="1388.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="904.55" y1="1372.25" x2="920.55" y2="1388.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="904.55" y1="1372.25" x2="920.55" y2="1356.25" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.58" y1="1084.87" x2="1296.58" y2="1068.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.58" y1="1084.87" x2="1296.58" y2="1100.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.58" y1="1084.87" x2="1328.58" y2="1100.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1312.58" y1="1084.87" x2="1328.58" y2="1068.87" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1134.2" y1="1238.48" x2="1118.2" y2="1222.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1134.2" y1="1238.48" x2="1118.2" y2="1254.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1134.2" y1="1238.48" x2="1150.2" y2="1254.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1134.2" y1="1238.48" x2="1150.2" y2="1222.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1561.34" y1="914.945" x2="1545.34" y2="898.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1561.34" y1="914.945" x2="1545.34" y2="930.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1561.34" y1="914.945" x2="1577.34" y2="930.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1561.34" y1="914.945" x2="1577.34" y2="898.945" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1391.28" y1="1184.21" x2="1375.28" y2="1168.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1391.28" y1="1184.21" x2="1375.28" y2="1200.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1391.28" y1="1184.21" x2="1407.28" y2="1200.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1391.28" y1="1184.21" x2="1407.28" y2="1168.21" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1311.21" y1="973.579" x2="1295.21" y2="957.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1311.21" y1="973.579" x2="1295.21" y2="989.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1311.21" y1="973.579" x2="1327.21" y2="989.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1311.21" y1="973.579" x2="1327.21" y2="957.579" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.94" y1="1419.74" x2="1078.94" y2="1403.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.94" y1="1419.74" x2="1078.94" y2="1435.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.94" y1="1419.74" x2="1110.94" y2="1435.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1094.94" y1="1419.74" x2="1110.94" y2="1403.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1143.74" y1="1146.65" x2="1127.74" y2="1130.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1143.74" y1="1146.65" x2="1127.74" y2="1162.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1143.74" y1="1146.65" x2="1159.74" y2="1162.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1143.74" y1="1146.65" x2="1159.74" y2="1130.65" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="743.908" y1="1344.73" x2="727.908" y2="1328.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="743.908" y1="1344.73" x2="727.908" y2="1360.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="743.908" y1="1344.73" x2="759.908" y2="1360.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="743.908" y1="1344.73" x2="759.908" y2="1328.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.98" y1="1226.45" x2="1099.98" y2="1210.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.98" y1="1226.45" x2="1099.98" y2="1242.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.98" y1="1226.45" x2="1131.98" y2="1242.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1115.98" y1="1226.45" x2="1131.98" y2="1210.45" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1724.4" y1="1054.35" x2="1708.4" y2="1038.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1724.4" y1="1054.35" x2="1708.4" y2="1070.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1724.4" y1="1054.35" x2="1740.4" y2="1070.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1724.4" y1="1054.35" x2="1740.4" y2="1038.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.408" y1="1314.39" x2="947.408" y2="1298.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.408" y1="1314.39" x2="947.408" y2="1330.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.408" y1="1314.39" x2="979.408" y2="1330.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="963.408" y1="1314.39" x2="979.408" y2="1298.39" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.46" y1="1140.38" x2="1238.46" y2="1124.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.46" y1="1140.38" x2="1238.46" y2="1156.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.46" y1="1140.38" x2="1270.46" y2="1156.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1254.46" y1="1140.38" x2="1270.46" y2="1124.38" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="994.563" y1="1220.85" x2="978.563" y2="1204.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="994.563" y1="1220.85" x2="978.563" y2="1236.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="994.563" y1="1220.85" x2="1010.56" y2="1236.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="994.563" y1="1220.85" x2="1010.56" y2="1204.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1585.12" y1="1061.03" x2="1569.12" y2="1045.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1585.12" y1="1061.03" x2="1569.12" y2="1077.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1585.12" y1="1061.03" x2="1601.12" y2="1077.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1585.12" y1="1061.03" x2="1601.12" y2="1045.03" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.07" y1="993.346" x2="1764.07" y2="977.346" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.07" y1="993.346" x2="1764.07" y2="1009.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.07" y1="993.346" x2="1796.07" y2="1009.35" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1780.07" y1="993.346" x2="1796.07" y2="977.346" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.71" y1="1150.44" x2="1192.71" y2="1134.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.71" y1="1150.44" x2="1192.71" y2="1166.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.71" y1="1150.44" x2="1224.71" y2="1166.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.71" y1="1150.44" x2="1224.71" y2="1134.44" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.86" y1="1171.34" x2="1355.86" y2="1155.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.86" y1="1171.34" x2="1355.86" y2="1187.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.86" y1="1171.34" x2="1387.86" y2="1187.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.86" y1="1171.34" x2="1387.86" y2="1155.34" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.82" y1="1306.89" x2="1094.82" y2="1290.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.82" y1="1306.89" x2="1094.82" y2="1322.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.82" y1="1306.89" x2="1126.82" y2="1322.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.82" y1="1306.89" x2="1126.82" y2="1290.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.33" y1="1287.85" x2="1192.33" y2="1271.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.33" y1="1287.85" x2="1192.33" y2="1303.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.33" y1="1287.85" x2="1224.33" y2="1303.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1208.33" y1="1287.85" x2="1224.33" y2="1271.85" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.49" y1="1169.31" x2="1149.49" y2="1153.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.49" y1="1169.31" x2="1149.49" y2="1185.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.49" y1="1169.31" x2="1181.49" y2="1185.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165.49" y1="1169.31" x2="1181.49" y2="1153.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.71" y1="1125.59" x2="1257.71" y2="1109.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.71" y1="1125.59" x2="1257.71" y2="1141.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.71" y1="1125.59" x2="1289.71" y2="1141.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1273.71" y1="1125.59" x2="1289.71" y2="1109.59" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1079.45" y1="1149.7" x2="1063.45" y2="1133.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1079.45" y1="1149.7" x2="1063.45" y2="1165.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1079.45" y1="1149.7" x2="1095.45" y2="1165.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1079.45" y1="1149.7" x2="1095.45" y2="1133.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1184.12" y1="1234.64" x2="1168.12" y2="1218.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1184.12" y1="1234.64" x2="1168.12" y2="1250.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1184.12" y1="1234.64" x2="1200.12" y2="1250.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1184.12" y1="1234.64" x2="1200.12" y2="1218.64" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.3" y1="1156.32" x2="1355.3" y2="1140.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.3" y1="1156.32" x2="1355.3" y2="1172.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.3" y1="1156.32" x2="1387.3" y2="1172.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1371.3" y1="1156.32" x2="1387.3" y2="1140.32" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.95" y1="1258.36" x2="1221.95" y2="1242.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.95" y1="1258.36" x2="1221.95" y2="1274.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.95" y1="1258.36" x2="1253.95" y2="1274.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1237.95" y1="1258.36" x2="1253.95" y2="1242.36" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="983.358" y1="1291.75" x2="967.358" y2="1275.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="983.358" y1="1291.75" x2="967.358" y2="1307.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="983.358" y1="1291.75" x2="999.358" y2="1307.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="983.358" y1="1291.75" x2="999.358" y2="1275.75" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1156.87" y1="1192.88" x2="1140.87" y2="1176.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1156.87" y1="1192.88" x2="1140.87" y2="1208.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1156.87" y1="1192.88" x2="1172.87" y2="1208.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1156.87" y1="1192.88" x2="1172.87" y2="1176.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1496.43" y1="924.203" x2="1480.43" y2="908.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1496.43" y1="924.203" x2="1480.43" y2="940.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1496.43" y1="924.203" x2="1512.43" y2="940.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1496.43" y1="924.203" x2="1512.43" y2="908.203" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="938.137" y1="1395.53" x2="922.137" y2="1379.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="938.137" y1="1395.53" x2="922.137" y2="1411.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="938.137" y1="1395.53" x2="954.137" y2="1411.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="938.137" y1="1395.53" x2="954.137" y2="1379.53" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="745.679" y1="1455.46" x2="729.679" y2="1439.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="745.679" y1="1455.46" x2="729.679" y2="1471.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="745.679" y1="1455.46" x2="761.679" y2="1471.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="745.679" y1="1455.46" x2="761.679" y2="1439.46" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1495.35" y1="1033.28" x2="1479.35" y2="1017.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1495.35" y1="1033.28" x2="1479.35" y2="1049.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1495.35" y1="1033.28" x2="1511.35" y2="1049.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1495.35" y1="1033.28" x2="1511.35" y2="1017.28" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1226.7" y1="1172" x2="1210.7" y2="1156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1226.7" y1="1172" x2="1210.7" y2="1188" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1226.7" y1="1172" x2="1242.7" y2="1188" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1226.7" y1="1172" x2="1242.7" y2="1156" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.77" y1="884.258" x2="1746.77" y2="868.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.77" y1="884.258" x2="1746.77" y2="900.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.77" y1="884.258" x2="1778.77" y2="900.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1762.77" y1="884.258" x2="1778.77" y2="868.258" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="977.334" y1="1349.1" x2="961.334" y2="1333.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="977.334" y1="1349.1" x2="961.334" y2="1365.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="977.334" y1="1349.1" x2="993.334" y2="1365.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="977.334" y1="1349.1" x2="993.334" y2="1333.1" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1274.34" y1="1100.11" x2="1258.34" y2="1084.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1274.34" y1="1100.11" x2="1258.34" y2="1116.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1274.34" y1="1100.11" x2="1290.34" y2="1116.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1274.34" y1="1100.11" x2="1290.34" y2="1084.11" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.89" y1="1051.48" x2="1497.89" y2="1035.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.89" y1="1051.48" x2="1497.89" y2="1067.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.89" y1="1051.48" x2="1529.89" y2="1067.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1513.89" y1="1051.48" x2="1529.89" y2="1035.48" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.49" y1="1113.18" x2="1441.49" y2="1097.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.49" y1="1113.18" x2="1441.49" y2="1129.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.49" y1="1113.18" x2="1473.49" y2="1129.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1457.49" y1="1113.18" x2="1473.49" y2="1097.18" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.94" y1="1156.93" x2="1405.94" y2="1140.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.94" y1="1156.93" x2="1405.94" y2="1172.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.94" y1="1156.93" x2="1437.94" y2="1172.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1421.94" y1="1156.93" x2="1437.94" y2="1140.93" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1638.84" y1="1106.16" x2="1622.84" y2="1090.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1638.84" y1="1106.16" x2="1622.84" y2="1122.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1638.84" y1="1106.16" x2="1654.84" y2="1122.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1638.84" y1="1106.16" x2="1654.84" y2="1090.16" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.292" y1="1482.58" x2="859.292" y2="1466.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.292" y1="1482.58" x2="859.292" y2="1498.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.292" y1="1482.58" x2="891.292" y2="1498.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="875.292" y1="1482.58" x2="891.292" y2="1466.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.87" y1="1032.79" x2="1274.87" y2="1016.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.87" y1="1032.79" x2="1274.87" y2="1048.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.87" y1="1032.79" x2="1306.87" y2="1048.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1290.87" y1="1032.79" x2="1306.87" y2="1016.79" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1405.31" y1="1103.12" x2="1389.31" y2="1087.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1405.31" y1="1103.12" x2="1389.31" y2="1119.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1405.31" y1="1103.12" x2="1421.31" y2="1119.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1405.31" y1="1103.12" x2="1421.31" y2="1087.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1142.23" y1="1185.99" x2="1126.23" y2="1169.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1142.23" y1="1185.99" x2="1126.23" y2="1201.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1142.23" y1="1185.99" x2="1158.23" y2="1201.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1142.23" y1="1185.99" x2="1158.23" y2="1169.99" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.433" y1="1329.29" x2="883.433" y2="1313.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.433" y1="1329.29" x2="883.433" y2="1345.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.433" y1="1329.29" x2="915.433" y2="1345.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="899.433" y1="1329.29" x2="915.433" y2="1313.29" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.35" y1="1160.97" x2="1224.35" y2="1144.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.35" y1="1160.97" x2="1224.35" y2="1176.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.35" y1="1160.97" x2="1256.35" y2="1176.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1240.35" y1="1160.97" x2="1256.35" y2="1144.97" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1058.78" y1="1213.58" x2="1042.78" y2="1197.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1058.78" y1="1213.58" x2="1042.78" y2="1229.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1058.78" y1="1213.58" x2="1074.78" y2="1229.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1058.78" y1="1213.58" x2="1074.78" y2="1197.58" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1093.83" y1="1252.83" x2="1077.83" y2="1236.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1093.83" y1="1252.83" x2="1077.83" y2="1268.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1093.83" y1="1252.83" x2="1109.83" y2="1268.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1093.83" y1="1252.83" x2="1109.83" y2="1236.83" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1423.88" y1="1021.7" x2="1407.88" y2="1005.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1423.88" y1="1021.7" x2="1407.88" y2="1037.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1423.88" y1="1021.7" x2="1439.88" y2="1037.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1423.88" y1="1021.7" x2="1439.88" y2="1005.7" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1552.62" y1="936.348" x2="1536.62" y2="920.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1552.62" y1="936.348" x2="1536.62" y2="952.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1552.62" y1="936.348" x2="1568.62" y2="952.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1552.62" y1="936.348" x2="1568.62" y2="920.348" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.54" y1="984.865" x2="1456.54" y2="968.865" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.54" y1="984.865" x2="1456.54" y2="1000.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.54" y1="984.865" x2="1488.54" y2="1000.86" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1472.54" y1="984.865" x2="1488.54" y2="968.865" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165" y1="1027.31" x2="1149" y2="1011.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165" y1="1027.31" x2="1149" y2="1043.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165" y1="1027.31" x2="1181" y2="1043.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1165" y1="1027.31" x2="1181" y2="1011.31" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.61" y1="1250.08" x2="1094.61" y2="1234.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.61" y1="1250.08" x2="1094.61" y2="1266.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.61" y1="1250.08" x2="1126.61" y2="1266.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1110.61" y1="1250.08" x2="1126.61" y2="1234.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1154.89" y1="1238.14" x2="1138.89" y2="1222.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1154.89" y1="1238.14" x2="1138.89" y2="1254.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1154.89" y1="1238.14" x2="1170.89" y2="1254.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1154.89" y1="1238.14" x2="1170.89" y2="1222.14" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.8" y1="1203.89" x2="1095.8" y2="1187.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.8" y1="1203.89" x2="1095.8" y2="1219.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.8" y1="1203.89" x2="1127.8" y2="1219.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1111.8" y1="1203.89" x2="1127.8" y2="1187.89" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239" y1="1106.12" x2="1223" y2="1090.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239" y1="1106.12" x2="1223" y2="1122.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239" y1="1106.12" x2="1255" y2="1122.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1239" y1="1106.12" x2="1255" y2="1090.12" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="839.379" y1="1287.74" x2="823.379" y2="1271.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="839.379" y1="1287.74" x2="823.379" y2="1303.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="839.379" y1="1287.74" x2="855.379" y2="1303.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="839.379" y1="1287.74" x2="855.379" y2="1271.74" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1100.71" y1="1170.24" x2="1084.71" y2="1154.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1100.71" y1="1170.24" x2="1084.71" y2="1186.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1100.71" y1="1170.24" x2="1116.71" y2="1186.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1100.71" y1="1170.24" x2="1116.71" y2="1154.24" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414" y1="1170.91" x2="1398" y2="1154.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414" y1="1170.91" x2="1398" y2="1186.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414" y1="1170.91" x2="1430" y2="1186.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414" y1="1170.91" x2="1430" y2="1154.91" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.86" y1="1217.88" x2="1398.86" y2="1201.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.86" y1="1217.88" x2="1398.86" y2="1233.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.86" y1="1217.88" x2="1430.86" y2="1233.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1414.86" y1="1217.88" x2="1430.86" y2="1201.88" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1194.86" y1="1244.81" x2="1178.86" y2="1228.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1194.86" y1="1244.81" x2="1178.86" y2="1260.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1194.86" y1="1244.81" x2="1210.86" y2="1260.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1194.86" y1="1244.81" x2="1210.86" y2="1228.81" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1257.94" y1="1026.73" x2="1241.94" y2="1010.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1257.94" y1="1026.73" x2="1241.94" y2="1042.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1257.94" y1="1026.73" x2="1273.94" y2="1042.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1257.94" y1="1026.73" x2="1273.94" y2="1010.73" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1296.85" y1="1204.15" x2="1280.85" y2="1188.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1296.85" y1="1204.15" x2="1280.85" y2="1220.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1296.85" y1="1204.15" x2="1312.85" y2="1220.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1296.85" y1="1204.15" x2="1312.85" y2="1188.15" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.11" y1="1134.69" x2="1227.11" y2="1118.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.11" y1="1134.69" x2="1227.11" y2="1150.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.11" y1="1134.69" x2="1259.11" y2="1150.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1243.11" y1="1134.69" x2="1259.11" y2="1118.69" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.52" y1="1257.08" x2="1321.52" y2="1241.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.52" y1="1257.08" x2="1321.52" y2="1273.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.52" y1="1257.08" x2="1353.52" y2="1273.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="1337.52" y1="1257.08" x2="1353.52" y2="1241.08" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="844.925" y1="1383.63" x2="828.925" y2="1367.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="844.925" y1="1383.63" x2="828.925" y2="1399.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="844.925" y1="1383.63" x2="860.925" y2="1399.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<line clip-path="url(#clip552)" x1="844.925" y1="1383.63" x2="860.925" y2="1367.63" style="stroke:#008000; stroke-width:4.8; stroke-opacity:1"/>
+<polyline clip-path="url(#clip552)" style="stroke:#dc143c; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2006.95,814.774 2025.44,792.795 2041.77,771.865 2055.91,752.038 2067.82,733.363 2077.48,715.886 2084.85,699.652 2089.91,684.701 2092.67,671.07 2093.1,658.793 2091.21,647.901 2087.01,638.422 2080.5,630.379 2071.7,623.791 2060.63,618.676 2047.32,615.047 2031.8,612.912 2014.11,612.276 1994.3,613.142 1972.42,615.507 1948.51,619.365 1922.64,624.707 1894.88,631.519 1865.29,639.784 1833.95,649.482 1800.94,660.588 1766.33,673.075 1730.22,686.911 1692.69,702.061 1653.85,718.489 1613.78,736.152 1572.58,755.007 1530.37,775.007 1487.24,796.101 1443.3,818.237 1398.66,841.36 1353.44,865.411 1307.75,890.331 1261.69,916.057 1215.4,942.525 1168.97,969.669 1122.53,997.421 1076.2,1025.71 1030.09,1054.47 984.309,1083.63 938.981,1113.1 894.216,1142.83 850.125,1172.73 806.82,1202.74 764.408,1232.76 722.995,1262.74 682.684,1292.6 643.578,1322.25 605.772,1351.63 569.362,1380.67 534.439,1409.29 501.091,1437.41 469.399,1464.97 439.444,1491.91 411.301,1518.15 385.04,1543.62 360.726,1568.27 338.42,1592.03 318.178,1614.84 300.051,1636.64 284.084,1657.39 270.316,1677.03 258.783,1695.51 249.513,1712.78 242.529,1728.8 237.849,1743.53 235.484,1756.93 235.44,1768.98 237.718,1779.64 242.312,1788.88 249.21,1796.68 258.394,1803.02 269.843,1807.88 283.527,1811.26 299.413,1813.14 317.46,1813.53 337.623,1812.41 359.852,1809.79 384.092,1805.68 410.282,1800.09 438.356,1793.04 468.244,1784.53 499.871,1774.59 533.16,1763.25 568.025,1750.54 604.381,1736.48 642.135,1721.11 681.195,1704.47 721.462,1686.6 762.835,1667.55 805.212,1647.37 848.486,1626.09 892.549,1603.79 937.291,1580.5 982.6,1556.3 1028.36,1531.24 1074.47,1505.39 1120.79,1478.8 1167.23,1451.55 1213.66,1423.7 1259.96,1395.33 1306.02,1366.5 1351.73,1337.28 1396.98,1307.76 1441.64,1278 1485.6,1248.07 1528.77,1218.06 1571.02,1188.03 1612.25,1158.07 1652.37,1128.24 1691.26,1098.63 1728.84,1069.3 1765,1040.33 1799.67,1011.79 1832.74,983.758 1864.15,956.295 1893.8,929.473 1921.64,903.361 1947.57,878.021 1971.56,853.519 1993.52,829.915 2013.41,807.269 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/conic/ellipse_approx/index.html b/previews/PR3525/tutorials/conic/ellipse_approx/index.html
new file mode 100644
index 00000000000..a1f5ff109a5
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/ellipse_approx/index.html
@@ -0,0 +1,341 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Ellipsoid approximation · JuMP</title><meta name="title" content="Ellipsoid approximation · JuMP"/><meta property="og:title" content="Ellipsoid approximation · JuMP"/><meta property="twitter:title" content="Ellipsoid approximation · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li class="is-active"><a class="tocitem" href>Ellipsoid approximation</a><ul class="internal"><li><a class="tocitem" href="#Problem-formulation"><span>Problem formulation</span></a></li><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Results"><span>Results</span></a></li><li><a class="tocitem" href="#Alternative-formulations"><span>Alternative formulations</span></a></li></ul></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Ellipsoid approximation</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Ellipsoid approximation</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/ellipse_approx.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Ellipsoid-approximation"><a class="docs-heading-anchor" href="#Ellipsoid-approximation">Ellipsoid approximation</a><a id="Ellipsoid-approximation-1"></a><a class="docs-heading-anchor-permalink" href="#Ellipsoid-approximation" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../ellipse_approx.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial considers the problem of computing <em>extremal ellipsoids</em>: finding ellipsoids that best approximate a given set. As an extension, we show how to use JuMP to inspect the bridges that were used, and how to explore alternative formulations.</p><p>The model comes from Section 4.9 &quot;Applications VII: extremal ellipsoids&quot; of the book <em>Lectures on Modern Convex Optimization</em> by <a href="http://epubs.siam.org/doi/book/10.1137/1.9780898718829">Ben-Tal and Nemirovski (2001)</a>.</p><p>For a related example, see also the <a href="../min_ellipse/#Minimal-ellipses">Minimal ellipses</a> tutorial.</p><h2 id="Problem-formulation"><a class="docs-heading-anchor" href="#Problem-formulation">Problem formulation</a><a id="Problem-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-formulation" title="Permalink"></a></h2><p>Suppose that we are given a set <span>$\mathcal{S}$</span> consisting of <span>$m$</span> points in <span>$n$</span>-dimensional space:</p><p class="math-container">\[\mathcal{S} = \{ x_1, \ldots, x_m \} \subset \mathbb{R}^n\]</p><p>Our goal is to determine an optimal vector <span>$c \in  \mathbb{R}^n$</span> and an optimal <span>$n \times n$</span> real symmetric matrix <span>$D$</span> such that the ellipse:</p><p class="math-container">\[E(D, c) = \{ x : (x - c)^\top D ( x - c) \leq 1 \},\]</p><p>contains <span>$\mathcal{S}$</span> and has the smallest possible volume.</p><p>The optimal <span>$D$</span> and <span>$c$</span> are given by the optimization problem:</p><p class="math-container">\[\begin{aligned}
+\max        \quad &amp; t                                                                    \\
+\text{s.t.} \quad &amp; Z \succeq 0 \\
+                  &amp; \begin{bmatrix} s &amp; z^\top \\ z &amp; Z \\ \end{bmatrix} \succeq 0 \\
+                  &amp; x_i^\top Z x_i - 2x_i^\top z + s \leq 1    &amp; \quad i=1, \ldots, m \\
+                  &amp; t \le \sqrt[n]{\det(Z)},
+\end{aligned}\]</p><p>where <span>$D = Z_*$</span> and <span>$c = Z_*^{-1} z_*$</span>.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import LinearAlgebra
+import Plots
+import Random
+import SCS</code></pre><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>We first need to generate some points to work with.</p><pre><code class="language-julia hljs">function generate_point_cloud(
+    m;            # number of 2-dimensional points
+    a = 10,       # scaling in x direction
+    b = 2,        # scaling in y direction
+    rho = π / 6,  # rotation of points around origin
+    random_seed = 1,
+)
+    rng = Random.MersenneTwister(random_seed)
+    P = randn(rng, Float64, m, 2)
+    Phi = [a*cos(rho) a*sin(rho); -b*sin(rho) b*cos(rho)]
+    S = P * Phi
+    return S
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">generate_point_cloud (generic function with 1 method)</code></pre><p>For the sake of this example, let&#39;s take <span>$m = 600$</span>:</p><pre><code class="language-julia hljs">S = generate_point_cloud(600);</code></pre><p>We will visualise the points (and ellipse) using the Plots package:</p><pre><code class="language-julia hljs">r = 1.1 * maximum(abs.(S))
+plot = Plots.scatter(
+    S[:, 1],
+    S[:, 2];
+    xlim = (-r, r),
+    ylim = (-r, r),
+    label = nothing,
+    c = :green,
+    shape = :x,
+    size = (600, 600),
+)</code></pre><img src="07c030f6.svg" alt="Example block output"/><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>Now let&#39;s build and the JuMP model. We&#39;ll compute <span>$D$</span> and <span>$c$</span> after the solve.</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+# We need to use a tighter tolerance for this example, otherwise the bounding
+# ellipse won&#39;t actually be bounding...
+set_attribute(model, &quot;eps_rel&quot;, 1e-6)
+set_silent(model)
+m, n = size(S)
+@variable(model, z[1:n])
+@variable(model, Z[1:n, 1:n], PSD)
+@variable(model, s)
+@variable(model, t)
+@constraint(model, [s z&#39;; z Z] &gt;= 0, PSDCone())
+@constraint(
+    model,
+    [i in 1:m],
+    S[i, :]&#39; * Z * S[i, :] - 2 * S[i, :]&#39; * z + s &lt;= 1,
+)
+@constraint(model, [t; vec(Z)] in MOI.RootDetConeSquare(n))
+@objective(model, Max, t)
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : SCS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;solved&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 4.50777e-03
+  Dual objective value : 4.47966e-03
+
+* Work counters
+  Solve time (sec)   : 7.60457e-01
+</code></pre><h2 id="Results"><a class="docs-heading-anchor" href="#Results">Results</a><a id="Results-1"></a><a class="docs-heading-anchor-permalink" href="#Results" title="Permalink"></a></h2><p>After solving the model to optimality we can recover the solution in terms of <span>$D$</span> and <span>$c$</span>:</p><pre><code class="language-julia hljs">D = value.(Z)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  0.00707822  -0.0102173
+ -0.0102173    0.0175624</code></pre><pre><code class="language-julia hljs">c = D \ value.(z)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ -3.2480287747852175
+ -1.8428256866083088</code></pre><p>Finally, overlaying the solution in the plot we see the minimal volume approximating ellipsoid:</p><pre><code class="language-julia hljs">P = sqrt(D)
+q = -P * c
+data = [tuple(P \ [cos(θ) - q[1], sin(θ) - q[2]]...) for θ in 0:0.05:(2pi+0.05)]
+Plots.plot!(plot, data; c = :crimson, label = nothing)</code></pre><img src="e71dc9be.svg" alt="Example block output"/><h2 id="Alternative-formulations"><a class="docs-heading-anchor" href="#Alternative-formulations">Alternative formulations</a><a id="Alternative-formulations-1"></a><a class="docs-heading-anchor-permalink" href="#Alternative-formulations" title="Permalink"></a></h2><p>The formulation of <code>model</code> uses <a href="../../../moi/reference/standard_form/#MathOptInterface.RootDetConeSquare"><code>MOI.RootDetConeSquare</code></a>. However, because SCS does not natively support this cone, JuMP automatically reformulates the problem into an equivalent problem that SCS <em>does</em> support. You can see the reformulation that JuMP chose using <a href="../../../api/JuMP/#JuMP.print_active_bridges"><code>print_active_bridges</code></a>:</p><pre><code class="language-julia hljs">print_active_bridges(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi"> * <span class="sgr31">Unsupported objective: MOI.VariableIndex</span>
+ |  bridged by:
+ |   MOIB.Objective.FunctionizeBridge{Float64}
+ |  may introduce:
+ |   * <span class="sgr32">Supported objective: MOI.ScalarAffineFunction{Float64}</span>
+ * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.LessThan{Float64}</span>
+ |  bridged by:
+ |   MOIB.Constraint.LessToGreaterBridge{Float64, MOI.ScalarAffineFunction{Float64}, MOI.ScalarAffineFunction{Float64}}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.GreaterThan{Float64}</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.VectorizeBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.Nonnegatives, MOI.ScalarAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeSquare</span>
+ |  bridged by:
+ |   MOIB.Constraint.SquareBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.ScalarAffineFunction{Float64}, MOI.PositiveSemidefiniteConeTriangle, MOI.PositiveSemidefiniteConeSquare}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |   |  bridged by:
+ |   |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.EqualTo{Float64}</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.VectorizeBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.Zeros, MOI.ScalarAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Zeros</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |  bridged by:
+ |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |  bridged by:
+ |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.RootDetConeSquare</span>
+ |  bridged by:
+ |   MOIB.Constraint.SquareBridge{Float64, MOI.VectorOfVariables, MOI.ScalarAffineFunction{Float64}, MOI.RootDetConeTriangle, MOI.RootDetConeSquare}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.RootDetConeTriangle</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.RootDetBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables, MOI.VectorOfVariables}
+ |   |  may introduce:
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |   |   |  bridged by:
+ |   |   |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |   |   |  bridged by:
+ |   |   |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |   |  may introduce:
+ |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.GeometricMeanCone</span>
+ |   |   |  bridged by:
+ |   |   |   MOIB.Constraint.FunctionConversionBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables, MOI.GeometricMeanCone}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.GeometricMeanCone</span>
+ |   |   |   |  bridged by:
+ |   |   |   |   MOIB.Constraint.GeoMeanToPowerBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |   |  may introduce:
+ |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PowerCone{Float64}</span>
+ |   |   |   |   * <span class="sgr31">Unsupported variable: MOI.Nonnegatives</span>
+ |   |   |   |   |  adding as constraint:
+ |   |   |   |   |   * <span class="sgr32">Supported variable: MOI.Reals</span>
+ |   |   |   |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.Nonnegatives</span>
+ |   |   |   |   |   |  bridged by:
+ |   |   |   |   |   |   MOIB.Constraint.FunctionConversionBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables, MOI.Nonnegatives}
+ |   |   |   |   |   |  may introduce:
+ |   |   |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ |   |   * <span class="sgr32">Supported variable: MOI.Reals</span>
+ |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.EqualTo{Float64}</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.VectorizeBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.Zeros, MOI.ScalarAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Zeros</span></code></pre><p>There&#39;s a lot going on here, but the first bullet is:</p><pre><code class="language-raw hljs">* Unsupported objective: MOI.VariableIndex
+|  bridged by:
+|   MOIB.Objective.FunctionizeBridge{Float64}
+|  introduces:
+|   * Supported objective: MOI.ScalarAffineFunction{Float64}</code></pre><p>This says that SCS does not support a <code>MOI.VariableIndex</code> objective function, and that JuMP used a <a href="../../../moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Objective.FunctionizeBridge"><code>MOI.Bridges.Objective.FunctionizeBridge</code></a> to convert it into a <code>MOI.ScalarAffineFunction{Float64}</code> objective function.</p><p>We can leave JuMP to do the reformulation, or we can rewrite our model to have an objective function that SCS natively supports:</p><pre><code class="language-julia hljs">@objective(model, Max, 1.0 * t + 0.0);</code></pre><p>Re-printing the active bridges:</p><pre><code class="language-julia hljs">print_active_bridges(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi"> * <span class="sgr32">Supported objective: MOI.ScalarAffineFunction{Float64}</span>
+ * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.LessThan{Float64}</span>
+ |  bridged by:
+ |   MOIB.Constraint.LessToGreaterBridge{Float64, MOI.ScalarAffineFunction{Float64}, MOI.ScalarAffineFunction{Float64}}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.GreaterThan{Float64}</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.VectorizeBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.Nonnegatives, MOI.ScalarAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeSquare</span>
+ |  bridged by:
+ |   MOIB.Constraint.SquareBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.ScalarAffineFunction{Float64}, MOI.PositiveSemidefiniteConeTriangle, MOI.PositiveSemidefiniteConeSquare}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |   |  bridged by:
+ |   |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.EqualTo{Float64}</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.VectorizeBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.Zeros, MOI.ScalarAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Zeros</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |  bridged by:
+ |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |  bridged by:
+ |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.RootDetConeSquare</span>
+ |  bridged by:
+ |   MOIB.Constraint.SquareBridge{Float64, MOI.VectorOfVariables, MOI.ScalarAffineFunction{Float64}, MOI.RootDetConeTriangle, MOI.RootDetConeSquare}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.RootDetConeTriangle</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.RootDetBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables, MOI.VectorOfVariables}
+ |   |  may introduce:
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |   |   |  bridged by:
+ |   |   |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |   |   |  bridged by:
+ |   |   |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |   |  may introduce:
+ |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.GeometricMeanCone</span>
+ |   |   |  bridged by:
+ |   |   |   MOIB.Constraint.FunctionConversionBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables, MOI.GeometricMeanCone}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.GeometricMeanCone</span>
+ |   |   |   |  bridged by:
+ |   |   |   |   MOIB.Constraint.GeoMeanToPowerBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |   |  may introduce:
+ |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PowerCone{Float64}</span>
+ |   |   |   |   * <span class="sgr31">Unsupported variable: MOI.Nonnegatives</span>
+ |   |   |   |   |  adding as constraint:
+ |   |   |   |   |   * <span class="sgr32">Supported variable: MOI.Reals</span>
+ |   |   |   |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.Nonnegatives</span>
+ |   |   |   |   |   |  bridged by:
+ |   |   |   |   |   |   MOIB.Constraint.FunctionConversionBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables, MOI.Nonnegatives}
+ |   |   |   |   |   |  may introduce:
+ |   |   |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ |   |   * <span class="sgr32">Supported variable: MOI.Reals</span>
+ |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.EqualTo{Float64}</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.VectorizeBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.Zeros, MOI.ScalarAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Zeros</span></code></pre><p>we get <code>* Supported objective: MOI.ScalarAffineFunction{Float64}</code>.</p><p>We can manually implement some other reformulations to change our model to something that SCS more closely supports by:</p><ul><li><p>Replacing the <a href="../../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> constraint <code>@variable(model, Z[1:n, 1:n], PSD)</code> with the <a href="../../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> <code>@constraint(model, Z &gt;= 0, PSDCone())</code>.</p></li><li><p>Replacing the <a href="../../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeSquare"><code>MOI.PositiveSemidefiniteConeSquare</code></a> constraint <code>[s z&#39;; z Z] &gt;= 0, PSDCone()</code> with the <a href="../../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> <code>@constraint(model, LinearAlgebra.Symmetric([s z&#39;; z Z]) &gt;= 0, PSDCone())</code>.</p></li><li><p>Replacing the <a href="../../../moi/reference/standard_form/#MathOptInterface.ScalarAffineFunction"><code>MOI.ScalarAffineFunction</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.GreaterThan"><code>MOI.GreaterThan</code></a> constraints with the vectorized equivalent of <a href="../../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.Nonnegatives"><code>MOI.Nonnegatives</code></a></p></li><li><p>Replacing the <a href="../../../moi/reference/standard_form/#MathOptInterface.VectorOfVariables"><code>MOI.VectorOfVariables</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.RootDetConeSquare"><code>MOI.RootDetConeSquare</code></a> constraint with <a href="../../../moi/reference/standard_form/#MathOptInterface.VectorAffineFunction"><code>MOI.VectorAffineFunction</code></a> in <a href="../../../moi/reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>MOI.RootDetConeTriangle</code></a>.</p></li></ul><p>Note that we still need to bridge <a href="../../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> constraints because SCS uses an internal <code>SCS.ScaledPSDCone</code> set instead.</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_attribute(model, &quot;eps_rel&quot;, 1e-6)
+set_silent(model)
+@variable(model, z[1:n])
+@variable(model, s)
+@variable(model, t)
+# The former @variable(model, Z[1:n, 1:n], PSD)
+@variable(model, Z[1:n, 1:n], Symmetric)
+@constraint(model, Z &gt;= 0, PSDCone())
+# The former [s z&#39;; z Z] &gt;= 0, PSDCone()
+@constraint(model, LinearAlgebra.Symmetric([s z&#39;; z Z]) &gt;= 0, PSDCone())
+# The former constraint S[i, :]&#39; * Z * S[i, :] - 2 * S[i, :]&#39; * z + s &lt;= 1
+f = [1 - S[i, :]&#39; * Z * S[i, :] + 2 * S[i, :]&#39; * z - s for i in 1:m]
+@constraint(model, f in MOI.Nonnegatives(m))
+# The former constraint [t; vec(Z)] in MOI.RootDetConeSquare(n)
+@constraint(model, 1 * [t; triangle_vec(Z)] .+ 0 in MOI.RootDetConeTriangle(n))
+# The former @objective(model, Max, t)
+@objective(model, Max, 1 * t + 0)
+optimize!(model)
+solve_time_1 = solve_time(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.7580219220000001</code></pre><p>This formulation gives the much smaller graph:</p><pre><code class="language-julia hljs">print_active_bridges(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi"> * <span class="sgr32">Supported objective: MOI.ScalarAffineFunction{Float64}</span>
+ * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |  bridged by:
+ |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |  bridged by:
+ |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.RootDetConeTriangle</span>
+ |  bridged by:
+ |   MOIB.Constraint.RootDetBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |   |  bridged by:
+ |   |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.GeometricMeanCone</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.GeoMeanToPowerBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PowerCone{Float64}</span>
+ |   |   * <span class="sgr31">Unsupported variable: MOI.Nonnegatives</span>
+ |   |   |  adding as constraint:
+ |   |   |   * <span class="sgr32">Supported variable: MOI.Reals</span>
+ |   |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorOfVariables-in-MOI.Nonnegatives</span>
+ |   |   |   |  bridged by:
+ |   |   |   |   MOIB.Constraint.FunctionConversionBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorOfVariables, MOI.Nonnegatives}
+ |   |   |   |  may introduce:
+ |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ |   * <span class="sgr32">Supported variable: MOI.Reals</span></code></pre><p>The last bullet shows how JuMP reformulated the <a href="../../../moi/reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>MOI.RootDetConeTriangle</code></a> constraint by adding a mix of <a href="../../../moi/reference/standard_form/#MathOptInterface.PositiveSemidefiniteConeTriangle"><code>MOI.PositiveSemidefiniteConeTriangle</code></a> and <a href="../../../moi/reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a> constraints.</p><p>Because SCS doesn&#39;t natively support the <a href="../../../moi/reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a>, these constraints were further bridged using a <a href="../../../moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge"><code>MOI.Bridges.Constraint.GeoMeanToPowerBridge</code></a> to a series of <a href="../../../moi/reference/standard_form/#MathOptInterface.PowerCone"><code>MOI.PowerCone</code></a> constraints.</p><p>However, there are many other ways that a <a href="../../../moi/reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a> can be reformulated into something that SCS supports. Let&#39;s see what happens if we use <a href="../../../api/JuMP/#JuMP.remove_bridge"><code>remove_bridge</code></a> to remove the <a href="../../../moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeanToPowerBridge"><code>MOI.Bridges.Constraint.GeoMeanToPowerBridge</code></a>:</p><pre><code class="language-julia hljs">remove_bridge(model, MOI.Bridges.Constraint.GeoMeanToPowerBridge)
+optimize!(model)</code></pre><p>This time, the solve took:</p><pre><code class="language-julia hljs">solve_time_2 = solve_time(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.458393269</code></pre><p>where previously it took</p><pre><code class="language-julia hljs">solve_time_1</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.7580219220000001</code></pre><p>Why was the solve time different?</p><pre><code class="language-julia hljs">print_active_bridges(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi"> * <span class="sgr32">Supported objective: MOI.ScalarAffineFunction{Float64}</span>
+ * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |  bridged by:
+ |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |  bridged by:
+ |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.RootDetConeTriangle</span>
+ |  bridged by:
+ |   MOIB.Constraint.RootDetBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |  may introduce:
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.PositiveSemidefiniteConeTriangle</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.SetDotScalingBridge{Float64, MOI.PositiveSemidefiniteConeTriangle, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Scaled{MOI.PositiveSemidefiniteConeTriangle}</span>
+ |   |   |  bridged by:
+ |   |   |   SCS.ScaledPSDConeBridge{Float64, MOI.VectorAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-SCS.ScaledPSDCone</span>
+ |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.GeometricMeanCone</span>
+ |   |  bridged by:
+ |   |   MOIB.Constraint.GeoMeanBridge{Float64, MOI.ScalarAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |  may introduce:
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.LessThan{Float64}</span>
+ |   |   |  bridged by:
+ |   |   |   MOIB.Constraint.LessToGreaterBridge{Float64, MOI.ScalarAffineFunction{Float64}, MOI.ScalarAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr31">Unsupported constraint: MOI.ScalarAffineFunction{Float64}-in-MOI.GreaterThan{Float64}</span>
+ |   |   |   |  bridged by:
+ |   |   |   |   MOIB.Constraint.VectorizeBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.Nonnegatives, MOI.ScalarAffineFunction{Float64}}
+ |   |   |   |  may introduce:
+ |   |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ |   |   * <span class="sgr31">Unsupported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.RotatedSecondOrderCone</span>
+ |   |   |  bridged by:
+ |   |   |   MOIB.Constraint.RSOCtoSOCBridge{Float64, MOI.VectorAffineFunction{Float64}, MOI.VectorAffineFunction{Float64}}
+ |   |   |  may introduce:
+ |   |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.SecondOrderCone</span>
+ |   |   * <span class="sgr32">Supported constraint: MOI.VectorAffineFunction{Float64}-in-MOI.Nonnegatives</span>
+ |   |   * <span class="sgr32">Supported variable: MOI.Reals</span>
+ |   * <span class="sgr32">Supported variable: MOI.Reals</span></code></pre><p>This time, JuMP used a <a href="../../../moi/submodules/Bridges/list_of_bridges/#MathOptInterface.Bridges.Constraint.GeoMeanBridge"><code>MOI.Bridges.Constraint.GeoMeanBridge</code></a> to reformulate the constraint into a set of <a href="../../../moi/reference/standard_form/#MathOptInterface.RotatedSecondOrderCone"><code>MOI.RotatedSecondOrderCone</code></a> constraints, which were further reformulated into a set of supported <a href="../../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a> constraints.</p><p>Since the two models are equivalent, we can conclude that for this particular model, the <a href="../../../moi/reference/standard_form/#MathOptInterface.SecondOrderCone"><code>MOI.SecondOrderCone</code></a> formulation is more efficient.</p><p>In general though, the performance of a particular reformulation is problem- and solver-specific. Therefore, JuMP chooses to minimize the number of bridges in the default reformulation, leaving you to explore alternative formulations using the tools and techniques shown in this tutorial.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../min_ellipse/">« Minimal ellipses</a><a class="docs-footer-nextpage" href="../quantum_discrimination/">Quantum state discrimination »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/experiment_design.jl b/previews/PR3525/tutorials/conic/experiment_design.jl
new file mode 100644
index 00000000000..f8763b6b76f
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/experiment_design.jl
@@ -0,0 +1,218 @@
+# Copyright (c) 2019 Arpit Bhatia, Chris Coey, and contributors                  #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Experiment design
+
+# **This tutorial was originally contributed by Arpit Bhatia and Chris Coey.**
+
+# This tutorial covers experiment design examples (D-optimal, A-optimal, and
+# E-optimal) from section 7.5 of the book [Convex Optimization](https://doi.org/10.1017/CBO9780511804441)
+# by Boyd and Vandenberghe.
+
+# The tutorial uses the following packages
+using JuMP
+import SCS
+import LinearAlgebra
+import Random
+
+# !!! info
+#     This tutorial uses sets from [MathOptInterface](@ref moi_documentation).
+#     By default, JuMP exports the `MOI` symbol as an alias for the
+#     MathOptInterface.jl package. We recommend making this more explicit in
+#     your code by adding the following lines:
+#     ```julia
+#     import MathOptInterface as MOI
+#     ```
+
+# We set a seed so the random numbers are repeatable:
+Random.seed!(1234)
+
+# ## The relaxed experiment design problem
+
+# The basic experiment design problem is as follows.
+
+# Given the menu of possible choices for experiments, $v_{1}, \ldots, v_{p}$,
+# and the total number $m$ of experiments to be carried out, choose the numbers
+# of each type of experiment, that is, $m_{1}, \ldots, m_{p}$ to make the error
+# covariance $E$ small (in  some sense).
+
+# The variables $m_{1}, \ldots, m_{p}$ must, of course, be integers and sum to
+# $m$ the given total number of experiments. This leads to the optimization
+# problem:
+
+# ```math
+# \begin{aligned}
+# \min\left(\mathrm{w.r.t.} \mathbf{S}_{+}^{n}\right) & E=\left(\sum_{j=1}^{p} m_{j} v_{j} v_{j}^{T}\right)^{-1} \\
+# \text{subject to} & m_{i} \geq 0 \\
+# & \sum\limits_{i=1}^p m_{i} = m \\
+# & m_{i} \in \mathbb{Z},\quad i=1,\ldots,p
+# \end{aligned}
+# ```
+
+# The basic experiment design problem can be a hard combinatorial problem when
+# $m$, the total number of experiments, is comparable to $n$, since in this case
+# the $m_{i}$ are all small integers.
+
+# In the case when $m$ is large compared to $n$, however, a good approximate
+# solution can be found by ignoring, or relaxing, the constraint that the
+# $m_{i}$ are integers.
+
+# Let $\lambda_{i}=m_{i} / m,$ which is the fraction of the total number of
+# experiments for which $a_{j}=v_{i},$ or the relative frequency of experiment
+# $i$. We can express the error covariance in terms of $\lambda_{i}$ as:
+
+# ```math
+# E=\frac{1}{m}\left(\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}\right)^{-1}
+# ```
+
+# The vector $\lambda \in \mathbf{R}^{p}$ satisfies
+# $\lambda \succeq 0, \mathbf{1}^{T} \lambda=1,$ and also, each
+# $\lambda_{i}$ is an integer multiple of $1 / m$. By ignoring this last
+# constraint, we arrive at the problem:
+
+# ```math
+# \begin{aligned}
+# \min\left(\mathrm{w.r.t.} \mathbf{S}_{+}^{n}\right) & E=(1 / m)\left(\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}\right)^{-1} \\
+# \text{subject to:}\quad & \lambda \succeq 0 \\
+# & \mathbf{1}^{T} \lambda=1
+# \end{aligned}
+# ```
+
+# Several scalarizations have been proposed for the experiment design problem,
+# which is a vector optimization problem over the positive semidefinite cone.
+
+q = 4 # dimension of estimate space
+p = 8 # number of experimental vectors
+n_max = 3 # upper bound on lambda
+n = 12
+
+V = randn(q, p)
+
+eye = Matrix{Float64}(LinearAlgebra.I, q, q);
+
+# ## A-optimal design
+
+# In A-optimal experiment design, we minimize tr $E$, the trace of the
+# covariance matrix. This objective is simply the mean of the norm of the error
+# squared:
+
+# ```math
+# \mathbf{E}\|e\|_{2}^{2}=\mathbf{E} \operatorname{tr}\left(e e^{T}\right)=\operatorname{tr} E
+# ```
+
+# The A-optimal experiment design problem in SDP form is
+
+# ```math
+# \begin{aligned}
+# \min & \mathbf{1}^{T} u \\
+# \text{subject to} & \left[\begin{aligned}{\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}} & {e_{k}} \\ {e_{k}^{T}} & {u_{k}}\end{aligned}\right] \succeq 0, \quad k=1, \ldots, n \\
+# & \lambda \succeq 0 \\
+# &  \mathbf{1}^{T} \lambda=1
+# \end{aligned}
+# ```
+
+aOpt = Model(SCS.Optimizer)
+set_silent(aOpt)
+@variable(aOpt, np[1:p], lower_bound = 0, upper_bound = n_max)
+@variable(aOpt, u[1:q], lower_bound = 0)
+@constraint(aOpt, sum(np) <= n)
+for i in 1:q
+    matrix = [
+        V*LinearAlgebra.diagm(0 => np ./ n)*V' eye[:, i]
+        eye[i, :]' u[i]
+    ]
+    @constraint(aOpt, matrix >= 0, PSDCone())
+end
+@objective(aOpt, Min, sum(u))
+optimize!(aOpt)
+objective_value(aOpt)
+
+#-
+
+value.(np)
+
+# ## E-optimal design
+
+# In $E$ -optimal design, we minimize the norm of the error covariance matrix,
+# that is, the maximum eigenvalue of $E$.
+
+# Since the diameter (twice the longest semi-axis) of the confidence ellipsoid
+# $\mathcal{E}$ is proportional to $\|E\|_{2}^{1 / 2}$, minimizing $\|E\|_{2}$
+# can be interpreted geometrically as minimizing the diameter of the confidence
+# ellipsoid.
+
+# E-optimal design can also be interpreted as minimizing the maximum variance of
+# $q^{T} e$, over all $q$ with $\|q\|_{2}=1$. The E-optimal experiment design
+# problem in SDP form is:
+
+# ```math
+# \begin{aligned}
+# \min & t \\
+# \text{subject to} & \sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T} \succeq t I \\
+# & \lambda \succeq 0 \\
+# & \mathbf{1}^{T} \lambda=1
+# \end{aligned}
+# ```
+
+eOpt = Model(SCS.Optimizer)
+set_silent(eOpt)
+@variable(eOpt, 0 <= np[1:p] <= n_max)
+@variable(eOpt, t)
+@constraint(
+    eOpt,
+    V * LinearAlgebra.diagm(0 => np ./ n) * V' - (t .* eye) >= 0,
+    PSDCone(),
+)
+@constraint(eOpt, sum(np) <= n)
+@objective(eOpt, Max, t)
+optimize!(eOpt)
+objective_value(eOpt)
+#-
+value.(np)
+
+# ## D-optimal design
+
+# The most widely used scalarization is called $D$ -optimal design,
+# in which we minimize the determinant of the error covariance matrix $E$.
+# This corresponds to designing the experiment to minimize the volume of the
+# resulting confidence ellipsoid (for a fixed confidence level). Ignoring the
+# constant factor $1 / m$ in $E$, and taking the logarithm of the objective, we
+# can pose this problem as convex optimization problem:
+
+# ```math
+# \begin{aligned}
+# \min & \log \operatorname{det}\left(\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}\right)^{-1} \\
+# \text{subject to} & \lambda \succeq 0 \\
+# & \mathbf{1}^{T} \lambda=1
+# \end{aligned}
+# ```
+
+dOpt = Model(SCS.Optimizer)
+set_silent(dOpt)
+@variable(dOpt, np[1:p], lower_bound = 0, upper_bound = n_max)
+@variable(dOpt, t)
+@objective(dOpt, Max, t)
+@constraint(dOpt, sum(np) <= n)
+E = V * LinearAlgebra.diagm(0 => np ./ n) * V'
+@constraint(dOpt, [t; 1; triangle_vec(E)] in MOI.LogDetConeTriangle(q))
+optimize!(dOpt)
+objective_value(dOpt)
+#-
+value.(np)
diff --git a/previews/PR3525/tutorials/conic/experiment_design/index.html b/previews/PR3525/tutorials/conic/experiment_design/index.html
new file mode 100644
index 00000000000..ed63250aa30
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/experiment_design/index.html
@@ -0,0 +1,98 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Experiment design · JuMP</title><meta name="title" content="Experiment design · JuMP"/><meta property="og:title" content="Experiment design · JuMP"/><meta property="twitter:title" content="Experiment design · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li class="is-active"><a class="tocitem" href>Experiment design</a><ul class="internal"><li><a class="tocitem" href="#The-relaxed-experiment-design-problem"><span>The relaxed experiment design problem</span></a></li><li><a class="tocitem" href="#A-optimal-design"><span>A-optimal design</span></a></li><li><a class="tocitem" href="#E-optimal-design"><span>E-optimal design</span></a></li><li><a class="tocitem" href="#D-optimal-design"><span>D-optimal design</span></a></li></ul></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Experiment design</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Experiment design</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/experiment_design.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Experiment-design"><a class="docs-heading-anchor" href="#Experiment-design">Experiment design</a><a id="Experiment-design-1"></a><a class="docs-heading-anchor-permalink" href="#Experiment-design" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../experiment_design.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Arpit Bhatia and Chris Coey.</strong></p><p>This tutorial covers experiment design examples (D-optimal, A-optimal, and E-optimal) from section 7.5 of the book <a href="https://doi.org/10.1017/CBO9780511804441">Convex Optimization</a> by Boyd and Vandenberghe.</p><p>The tutorial uses the following packages</p><pre><code class="language-julia hljs">using JuMP
+import SCS
+import LinearAlgebra
+import Random</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This tutorial uses sets from <a href="../../../moi/#moi_documentation">MathOptInterface</a>. By default, JuMP exports the <code>MOI</code> symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre></div></div><p>We set a seed so the random numbers are repeatable:</p><pre><code class="language-julia hljs">Random.seed!(1234)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Random.TaskLocalRNG()</code></pre><h2 id="The-relaxed-experiment-design-problem"><a class="docs-heading-anchor" href="#The-relaxed-experiment-design-problem">The relaxed experiment design problem</a><a id="The-relaxed-experiment-design-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-relaxed-experiment-design-problem" title="Permalink"></a></h2><p>The basic experiment design problem is as follows.</p><p>Given the menu of possible choices for experiments, <span>$v_{1}, \ldots, v_{p}$</span>, and the total number <span>$m$</span> of experiments to be carried out, choose the numbers of each type of experiment, that is, <span>$m_{1}, \ldots, m_{p}$</span> to make the error covariance <span>$E$</span> small (in  some sense).</p><p>The variables <span>$m_{1}, \ldots, m_{p}$</span> must, of course, be integers and sum to <span>$m$</span> the given total number of experiments. This leads to the optimization problem:</p><p class="math-container">\[\begin{aligned}
+\min\left(\mathrm{w.r.t.} \mathbf{S}_{+}^{n}\right) &amp; E=\left(\sum_{j=1}^{p} m_{j} v_{j} v_{j}^{T}\right)^{-1} \\
+\text{subject to} &amp; m_{i} \geq 0 \\
+&amp; \sum\limits_{i=1}^p m_{i} = m \\
+&amp; m_{i} \in \mathbb{Z},\quad i=1,\ldots,p
+\end{aligned}\]</p><p>The basic experiment design problem can be a hard combinatorial problem when <span>$m$</span>, the total number of experiments, is comparable to <span>$n$</span>, since in this case the <span>$m_{i}$</span> are all small integers.</p><p>In the case when <span>$m$</span> is large compared to <span>$n$</span>, however, a good approximate solution can be found by ignoring, or relaxing, the constraint that the <span>$m_{i}$</span> are integers.</p><p>Let <span>$\lambda_{i}=m_{i} / m,$</span> which is the fraction of the total number of experiments for which <span>$a_{j}=v_{i},$</span> or the relative frequency of experiment <span>$i$</span>. We can express the error covariance in terms of <span>$\lambda_{i}$</span> as:</p><p class="math-container">\[E=\frac{1}{m}\left(\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}\right)^{-1}\]</p><p>The vector <span>$\lambda \in \mathbf{R}^{p}$</span> satisfies <span>$\lambda \succeq 0, \mathbf{1}^{T} \lambda=1,$</span> and also, each <span>$\lambda_{i}$</span> is an integer multiple of <span>$1 / m$</span>. By ignoring this last constraint, we arrive at the problem:</p><p class="math-container">\[\begin{aligned}
+\min\left(\mathrm{w.r.t.} \mathbf{S}_{+}^{n}\right) &amp; E=(1 / m)\left(\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}\right)^{-1} \\
+\text{subject to:}\quad &amp; \lambda \succeq 0 \\
+&amp; \mathbf{1}^{T} \lambda=1
+\end{aligned}\]</p><p>Several scalarizations have been proposed for the experiment design problem, which is a vector optimization problem over the positive semidefinite cone.</p><pre><code class="language-julia hljs">q = 4 # dimension of estimate space
+p = 8 # number of experimental vectors
+n_max = 3 # upper bound on lambda
+n = 12
+
+V = randn(q, p)
+
+eye = Matrix{Float64}(LinearAlgebra.I, q, q);</code></pre><h2 id="A-optimal-design"><a class="docs-heading-anchor" href="#A-optimal-design">A-optimal design</a><a id="A-optimal-design-1"></a><a class="docs-heading-anchor-permalink" href="#A-optimal-design" title="Permalink"></a></h2><p>In A-optimal experiment design, we minimize tr <span>$E$</span>, the trace of the covariance matrix. This objective is simply the mean of the norm of the error squared:</p><p class="math-container">\[\mathbf{E}\|e\|_{2}^{2}=\mathbf{E} \operatorname{tr}\left(e e^{T}\right)=\operatorname{tr} E\]</p><p>The A-optimal experiment design problem in SDP form is</p><p class="math-container">\[\begin{aligned}
+\min &amp; \mathbf{1}^{T} u \\
+\text{subject to} &amp; \left[\begin{aligned}{\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}} &amp; {e_{k}} \\ {e_{k}^{T}} &amp; {u_{k}}\end{aligned}\right] \succeq 0, \quad k=1, \ldots, n \\
+&amp; \lambda \succeq 0 \\
+&amp;  \mathbf{1}^{T} \lambda=1
+\end{aligned}\]</p><pre><code class="language-julia hljs">aOpt = Model(SCS.Optimizer)
+set_silent(aOpt)
+@variable(aOpt, np[1:p], lower_bound = 0, upper_bound = n_max)
+@variable(aOpt, u[1:q], lower_bound = 0)
+@constraint(aOpt, sum(np) &lt;= n)
+for i in 1:q
+    matrix = [
+        V*LinearAlgebra.diagm(0 =&gt; np ./ n)*V&#39; eye[:, i]
+        eye[i, :]&#39; u[i]
+    ]
+    @constraint(aOpt, matrix &gt;= 0, PSDCone())
+end
+@objective(aOpt, Min, sum(u))
+optimize!(aOpt)
+objective_value(aOpt)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5.103223362935134</code></pre><pre><code class="language-julia hljs">value.(np)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">8-element Vector{Float64}:
+ 2.949521135024003
+ 1.7790921964770645
+ 8.10487134967608e-6
+ 4.706163786495679e-6
+ 2.1082669841842883
+ 1.6983463798099796
+ 1.2635010842548617
+ 2.201229233017432</code></pre><h2 id="E-optimal-design"><a class="docs-heading-anchor" href="#E-optimal-design">E-optimal design</a><a id="E-optimal-design-1"></a><a class="docs-heading-anchor-permalink" href="#E-optimal-design" title="Permalink"></a></h2><p>In <span>$E$</span> -optimal design, we minimize the norm of the error covariance matrix, that is, the maximum eigenvalue of <span>$E$</span>.</p><p>Since the diameter (twice the longest semi-axis) of the confidence ellipsoid <span>$\mathcal{E}$</span> is proportional to <span>$\|E\|_{2}^{1 / 2}$</span>, minimizing <span>$\|E\|_{2}$</span> can be interpreted geometrically as minimizing the diameter of the confidence ellipsoid.</p><p>E-optimal design can also be interpreted as minimizing the maximum variance of <span>$q^{T} e$</span>, over all <span>$q$</span> with <span>$\|q\|_{2}=1$</span>. The E-optimal experiment design problem in SDP form is:</p><p class="math-container">\[\begin{aligned}
+\min &amp; t \\
+\text{subject to} &amp; \sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T} \succeq t I \\
+&amp; \lambda \succeq 0 \\
+&amp; \mathbf{1}^{T} \lambda=1
+\end{aligned}\]</p><pre><code class="language-julia hljs">eOpt = Model(SCS.Optimizer)
+set_silent(eOpt)
+@variable(eOpt, 0 &lt;= np[1:p] &lt;= n_max)
+@variable(eOpt, t)
+@constraint(
+    eOpt,
+    V * LinearAlgebra.diagm(0 =&gt; np ./ n) * V&#39; - (t .* eye) &gt;= 0,
+    PSDCone(),
+)
+@constraint(eOpt, sum(np) &lt;= n)
+@objective(eOpt, Max, t)
+optimize!(eOpt)
+objective_value(eOpt)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.4353863856688645</code></pre><pre><code class="language-julia hljs">value.(np)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">8-element Vector{Float64}:
+  2.9999986784996944
+  2.752810163779928
+ -3.6095474331105857e-7
+ -1.4872411636950634e-6
+  2.1818462791954287
+  2.325263546428314
+  0.21896764666380808
+  1.5211163050754828</code></pre><h2 id="D-optimal-design"><a class="docs-heading-anchor" href="#D-optimal-design">D-optimal design</a><a id="D-optimal-design-1"></a><a class="docs-heading-anchor-permalink" href="#D-optimal-design" title="Permalink"></a></h2><p>The most widely used scalarization is called <span>$D$</span> -optimal design, in which we minimize the determinant of the error covariance matrix <span>$E$</span>. This corresponds to designing the experiment to minimize the volume of the resulting confidence ellipsoid (for a fixed confidence level). Ignoring the constant factor <span>$1 / m$</span> in <span>$E$</span>, and taking the logarithm of the objective, we can pose this problem as convex optimization problem:</p><p class="math-container">\[\begin{aligned}
+\min &amp; \log \operatorname{det}\left(\sum_{i=1}^{p} \lambda_{i} v_{i} v_{i}^{T}\right)^{-1} \\
+\text{subject to} &amp; \lambda \succeq 0 \\
+&amp; \mathbf{1}^{T} \lambda=1
+\end{aligned}\]</p><pre><code class="language-julia hljs">dOpt = Model(SCS.Optimizer)
+set_silent(dOpt)
+@variable(dOpt, np[1:p], lower_bound = 0, upper_bound = n_max)
+@variable(dOpt, t)
+@objective(dOpt, Max, t)
+@constraint(dOpt, sum(np) &lt;= n)
+E = V * LinearAlgebra.diagm(0 =&gt; np ./ n) * V&#39;
+@constraint(dOpt, [t; 1; triangle_vec(E)] in MOI.LogDetConeTriangle(q))
+optimize!(dOpt)
+objective_value(dOpt)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.3084539346723429</code></pre><pre><code class="language-julia hljs">value.(np)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">8-element Vector{Float64}:
+ 0.4275824266221196
+ 2.910016316893973
+ 2.9168724688641994e-6
+ 2.629197986378904e-6
+ 2.9157806299834754
+ 2.6733756645925353
+ 2.7353950122195503
+ 0.3378388086259769</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../logistic_regression/">« Logistic regression</a><a class="docs-footer-nextpage" href="../min_ellipse/">Minimal ellipses »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/introduction/index.html b/previews/PR3525/tutorials/conic/introduction/index.html
new file mode 100644
index 00000000000..2038281ac54
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/introduction/index.html
@@ -0,0 +1,9 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#How-to-choose-a-solver"><span>How to choose a solver</span></a></li><li><a class="tocitem" href="#How-these-tutorials-are-structured"><span>How these tutorials are structured</span></a></li></ul></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/introduction.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Introduction"><a class="docs-heading-anchor" href="#Introduction">Introduction</a><a id="Introduction-1"></a><a class="docs-heading-anchor-permalink" href="#Introduction" title="Permalink"></a></h1><p><a href="https://en.wikipedia.org/wiki/Conic_optimization">Conic programs</a> are a class of convex nonlinear optimization problems which use cones to represent the nonlinearities. They have the form:</p><p class="math-container">\[\begin{align}
+    &amp; \min_{x \in \mathbb{R}^n} &amp; f_0(x) \\
+    &amp; \;\;\text{s.t.} &amp; f_j(x) \in \mathcal{S}_j &amp; \;\; j = 1 \ldots m
+\end{align}\]</p><p>Mixed-integer conic programs (MICPs) are extensions of conic programs in which some (or all) of the decision variables take discrete values.</p><h2 id="How-to-choose-a-solver"><a class="docs-heading-anchor" href="#How-to-choose-a-solver">How to choose a solver</a><a id="How-to-choose-a-solver-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-choose-a-solver" title="Permalink"></a></h2><p>JuMP supports a range of conic solvers, although support differs on what types of cones each solver supports. In the list of <a href="../../../installation/#Supported-solvers">Supported solvers</a>, &quot;SOCP&quot; denotes solvers supporting second-order cones and &quot;SDP&quot; denotes solvers supporting semidefinite cones. In addition, solvers such as SCS and Mosek have support for the exponential cone. Moreover, due to the bridging system in MathOptInterface, many of these solvers support a much wider range of exotic cones than they natively support. Solvers supporting discrete variables start with &quot;(MI)&quot; in the list of <a href="../../../installation/#Supported-solvers">Supported solvers</a>.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Duality plays a large role in solving conic optimization models. Depending on the solver, it can be more efficient to solve the dual instead of the primal. If performance is an issue, see the <a href="../dualization/#Dualization">Dualization</a> tutorial for more details.</p></div></div><h2 id="How-these-tutorials-are-structured"><a class="docs-heading-anchor" href="#How-these-tutorials-are-structured">How these tutorials are structured</a><a id="How-these-tutorials-are-structured-1"></a><a class="docs-heading-anchor-permalink" href="#How-these-tutorials-are-structured" title="Permalink"></a></h2><p>Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.</p><ul><li>The following tutorials are worked examples that present a problem in words, then formulate it in mathematics, and then solve it in JuMP. This usually involves some sort of visualization of the solution. Start here if you are new to JuMP.<ul><li><a href="../experiment_design/#Experiment-design">Experiment design</a></li><li><a href="../logistic_regression/#Logistic-regression">Logistic regression</a></li></ul></li><li>The <a href="../tips_and_tricks/#conic_tips_and_tricks">Tips and tricks</a> tutorial contains a number of helpful reformulations and tricks you can use when modeling conic programs. Look here if you are stuck trying to formulate a problem as a conic program.</li><li>The remaining tutorials are less verbose and styled in the form of short code examples. These tutorials have less explanation, but may contain useful code snippets, particularly if they are similar to a problem you are trying to solve.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../nonlinear/complementarity/">« Mixed complementarity problems</a><a class="docs-footer-nextpage" href="../start_values/">Primal and dual warm-starts »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/logistic_regression.jl b/previews/PR3525/tutorials/conic/logistic_regression.jl
new file mode 100644
index 00000000000..cdfa33fe74a
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/logistic_regression.jl
@@ -0,0 +1,275 @@
+# MIT License                                                                    #src
+#                                                                                #src
+# Copyright (c) 2020 François Pacaud                                             #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Logistic regression
+
+# **This tutorial was originally contributed by François Pacaud.**
+
+# This tutorial shows how to solve a logistic regression problem
+# with JuMP. Logistic regression is a well known method in machine learning,
+# useful when we want to classify binary variables with the help of
+# a given set of features. To this goal,
+# we find the optimal combination of features maximizing
+# the (log)-likelihood onto a training set. From a modern optimization glance,
+# the resulting problem is convex and differentiable. On a modern optimization
+# glance, it is even conic representable.
+#
+# ## Formulating the logistic regression problem
+#
+# Suppose we have a set of training data-point $i = 1, \cdots, n$, where
+# for each $i$ we have a vector of features $x_i \in \mathbb{R}^p$ and a
+# categorical observation $y_i \in \{-1, 1\}$.
+#
+# The log-likelihood is given by
+#
+# ```math
+# l(\theta) = \sum_{i=1}^n \log(\dfrac{1}{1 + \exp(-y_i \theta^\top x_i)})
+# ```
+#
+# and the optimal $\theta$ minimizes the logistic loss function:
+#
+# ```math
+# \min_{\theta}\; \sum_{i=1}^n \log(1 + \exp(-y_i \theta^\top x_i)).
+# ```
+#
+# Most of the time, instead of solving directly the previous optimization problem, we
+# prefer to add a regularization term:
+#
+# ```math
+# \min_{\theta}\; \sum_{i=1}^n \log(1 + \exp(-y_i \theta^\top x_i)) + \lambda \| \theta \|
+# ```
+#
+# with $\lambda \in \mathbb{R}_+$ a penalty and $\|.\|$ a norm function. By adding
+# such a regularization term, we avoid overfitting on the training set and usually
+# achieve a greater score in cross-validation.
+
+# ## Reformulation as a conic optimization problem
+#
+# By introducing auxiliary variables $t_1, \cdots, t_n$ and $r$,
+# the optimization problem is equivalent to
+#
+# ```math
+# \begin{aligned}
+# \min_{t, r, \theta} \;& \sum_{i=1}^n t_i + \lambda r \\
+# \text{subject to } & \quad t_i \geq \log(1 + \exp(- y_i \theta^\top x_i)) \\
+#                    & \quad r \geq \|\theta\|
+# \end{aligned}
+# ```
+#
+# Now, the trick is to reformulate the constraints $t_i \geq \log(1 + \exp(- y_i \theta^\top x_i))$
+# with the help of the *exponential cone*
+#
+# ```math
+# K_{exp} = \{ (x, y, z) \in \mathbb{R}^3 : \; y \exp(x / y) \leq z \} .
+# ```
+#
+# Indeed, by passing to the exponential, we
+# see that for all $i=1, \cdots, n$, the constraint $t_i \geq \log(1 + \exp(- y_i \theta^\top x_i))$
+# is equivalent to
+#
+# ```math
+# \exp(-t_i) + \exp(u_i - t_i) \leq 1
+# ```
+#
+# with $u_i = -y_i \theta^\top x_i$. Then, by adding two auxiliary variables
+# $z_{i1}$ and $z_{i2}$ such that $z_{i1} \geq \exp(u_i-t_i)$ and $z_{i2} \geq \exp(-t_i)$, we get
+# the equivalent formulation
+#
+# ```math
+# \left\{
+# \begin{aligned}
+# (u_i -t_i , 1, z_{i1}) & \in  K_{exp}  \\
+# (-t_i , 1, z_{i2}) & \in  K_{exp}  \\
+# z_{i1} + z_{i2} & \leq  1
+# \end{aligned}
+# \right.
+# ```
+#
+# In this setting, the conic version of the logistic regression problems writes out
+#
+# ```math
+# \begin{aligned}
+# \min_{t, z, r, \theta}&  \; \sum_{i=1}^n t_i + \lambda r \\
+# \text{subject to } & \quad  (u_i -t_i , 1, z_{i1})  \in  K_{exp}  \\
+#                    & \quad  (-t_i , 1, z_{i2})  \in  K_{exp}  \\
+#                    & \quad  z_{i1} + z_{i2}  \leq  1 \\
+#                    & \quad u_i = -y_i x_i^\top \theta \\
+#                    & \quad r \geq \|\theta\|
+# \end{aligned}
+# ```
+#
+# and thus encompasses $3n + p + 1$ variables and $3n + 1$ constraints ($u_i = -y_i \theta^\top x_i$
+# is only a virtual constraint used to clarify the notation).
+# Thus, if $n \gg 1$, we get a large number of variables and constraints.
+
+# ## Fitting logistic regression with a conic solver
+#
+# It is now time to pass to the implementation. We choose SCS as a conic solver.
+using JuMP
+import Random
+import SCS
+
+# !!! info
+#     This tutorial uses sets from [MathOptInterface](@ref moi_documentation).
+#     By default, JuMP exports the `MOI` symbol as an alias for the
+#     MathOptInterface.jl package. We recommend making this more explicit in
+#     your code by adding the following lines:
+#     ```julia
+#     import MathOptInterface as MOI
+#     ```
+
+Random.seed!(2713);
+
+# We start by implementing a function to generate a fake dataset, and where
+# we could tune the correlation between the feature variables. The function
+# is a direct transcription of the one used in [this blog post](http://fa.bianp.net/blog/2013/numerical-optimizers-for-logistic-regression/).
+function generate_dataset(n_samples = 100, n_features = 10; shift = 0.0)
+    X = randn(n_samples, n_features)
+    w = randn(n_features)
+    y = sign.(X * w)
+    X .+= 0.8 * randn(n_samples, n_features) # add noise
+    X .+= shift # shift the points in the feature space
+    X = hcat(X, ones(n_samples, 1))
+    return X, y
+end
+
+# We write a `softplus` function to formulate each constraint
+# $t \geq \log(1 + \exp(u))$ with two exponential cones.
+function softplus(model, t, u)
+    z = @variable(model, [1:2], lower_bound = 0.0)
+    @constraint(model, sum(z) <= 1.0)
+    @constraint(model, [u - t, 1, z[1]] in MOI.ExponentialCone())
+    @constraint(model, [-t, 1, z[2]] in MOI.ExponentialCone())
+end
+
+# ### $\ell_2$ regularized logistic regression
+# Then, with the help of the `softplus` function, we could write our
+# optimization model. In the $\ell_2$ regularization case, the constraint
+# $r \geq \|\theta\|_2$ rewrites as a second order cone constraint.
+function build_logit_model(X, y, λ)
+    n, p = size(X)
+    model = Model()
+    @variable(model, θ[1:p])
+    @variable(model, t[1:n])
+    for i in 1:n
+        u = -(X[i, :]' * θ) * y[i]
+        softplus(model, t[i], u)
+    end
+    ## Add ℓ2 regularization
+    @variable(model, 0.0 <= reg)
+    @constraint(model, [reg; θ] in SecondOrderCone())
+    ## Define objective
+    @objective(model, Min, sum(t) + λ * reg)
+    return model
+end
+
+# We generate the dataset.
+#
+# !!! warning
+#     Be careful here, for large n and p SCS could fail to converge.
+#
+n, p = 200, 10
+X, y = generate_dataset(n, p; shift = 10.0);
+
+## We could now solve the logistic regression problem
+λ = 10.0
+model = build_logit_model(X, y, λ)
+set_optimizer(model, SCS.Optimizer)
+set_silent(model)
+JuMP.optimize!(model)
+
+#-
+
+θ♯ = JuMP.value.(model[:θ])
+
+# It appears that the speed of convergence is not that impacted by the correlation
+# of the dataset, nor by the penalty $\lambda$.
+
+# ### $\ell_1$ regularized logistic regression
+#
+# We now formulate the logistic problem with a $\ell_1$ regularization term.
+# The $\ell_1$ regularization ensures sparsity in the optimal
+# solution of the resulting optimization problem. Luckily, the $\ell_1$ norm
+# is implemented as a set in `MathOptInterface`. Thus, we could formulate
+# the sparse logistic regression problem with the help of a `MOI.NormOneCone`
+# set.
+function build_sparse_logit_model(X, y, λ)
+    n, p = size(X)
+    model = Model()
+    @variable(model, θ[1:p])
+    @variable(model, t[1:n])
+    for i in 1:n
+        u = -(X[i, :]' * θ) * y[i]
+        softplus(model, t[i], u)
+    end
+    ## Add ℓ1 regularization
+    @variable(model, 0.0 <= reg)
+    @constraint(model, [reg; θ] in MOI.NormOneCone(p + 1))
+    ## Define objective
+    @objective(model, Min, sum(t) + λ * reg)
+    return model
+end
+
+## Auxiliary function to count non-null components:
+count_nonzero(v::Vector; tol = 1e-6) = sum(abs.(v) .>= tol)
+
+## We solve the sparse logistic regression problem on the same dataset as before.
+λ = 10.0
+sparse_model = build_sparse_logit_model(X, y, λ)
+set_optimizer(sparse_model, SCS.Optimizer)
+set_silent(sparse_model)
+JuMP.optimize!(sparse_model)
+
+#-
+
+θ♯ = JuMP.value.(sparse_model[:θ])
+println(
+    "Number of non-zero components: ",
+    count_nonzero(θ♯),
+    " (out of ",
+    p,
+    " features)",
+)
+
+# ### Extensions
+# A direct extension would be to consider the sparse logistic regression with
+# *hard* thresholding, which, on contrary to the *soft* version using a $\ell_1$ regularization,
+# adds an explicit cardinality constraint in its formulation:
+#
+# ```math
+# \begin{aligned}
+# \min_{\theta} & \; \sum_{i=1}^n \log(1 + \exp(-y_i \theta^\top x_i)) + \lambda \| \theta \|_2^2 \\
+# \text{subject to } & \quad \| \theta \|_0 <= k
+# \end{aligned}
+# ```
+#
+# where $k$ is the maximum number of non-zero components in the vector $\theta$,
+# and $\|.\|_0$ is the $\ell_0$ pseudo-norm:
+#
+# ```math
+# \| x\|_0 = \#\{i : \; x_i \neq 0\}
+# ```
+#
+# The cardinality constraint $\|\theta\|_0 \leq k$ could be reformulated with
+# binary variables. Thus the hard sparse regression problem could be solved
+# by any solver supporting mixed integer conic problems.
+#
diff --git a/previews/PR3525/tutorials/conic/logistic_regression/index.html b/previews/PR3525/tutorials/conic/logistic_regression/index.html
new file mode 100644
index 00000000000..d408f79c761
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/logistic_regression/index.html
@@ -0,0 +1,107 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Logistic regression · JuMP</title><meta name="title" content="Logistic regression · JuMP"/><meta property="og:title" content="Logistic regression · JuMP"/><meta property="twitter:title" content="Logistic regression · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li class="is-active"><a class="tocitem" href>Logistic regression</a><ul class="internal"><li><a class="tocitem" href="#Formulating-the-logistic-regression-problem"><span>Formulating the logistic regression problem</span></a></li><li><a class="tocitem" href="#Reformulation-as-a-conic-optimization-problem"><span>Reformulation as a conic optimization problem</span></a></li><li><a class="tocitem" href="#Fitting-logistic-regression-with-a-conic-solver"><span>Fitting logistic regression with a conic solver</span></a></li></ul></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Logistic regression</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Logistic regression</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/logistic_regression.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Logistic-regression"><a class="docs-heading-anchor" href="#Logistic-regression">Logistic regression</a><a id="Logistic-regression-1"></a><a class="docs-heading-anchor-permalink" href="#Logistic-regression" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../logistic_regression.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by François Pacaud.</strong></p><p>This tutorial shows how to solve a logistic regression problem with JuMP. Logistic regression is a well known method in machine learning, useful when we want to classify binary variables with the help of a given set of features. To this goal, we find the optimal combination of features maximizing the (log)-likelihood onto a training set. From a modern optimization glance, the resulting problem is convex and differentiable. On a modern optimization glance, it is even conic representable.</p><h2 id="Formulating-the-logistic-regression-problem"><a class="docs-heading-anchor" href="#Formulating-the-logistic-regression-problem">Formulating the logistic regression problem</a><a id="Formulating-the-logistic-regression-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Formulating-the-logistic-regression-problem" title="Permalink"></a></h2><p>Suppose we have a set of training data-point <span>$i = 1, \cdots, n$</span>, where for each <span>$i$</span> we have a vector of features <span>$x_i \in \mathbb{R}^p$</span> and a categorical observation <span>$y_i \in \{-1, 1\}$</span>.</p><p>The log-likelihood is given by</p><p class="math-container">\[l(\theta) = \sum_{i=1}^n \log(\dfrac{1}{1 + \exp(-y_i \theta^\top x_i)})\]</p><p>and the optimal <span>$\theta$</span> minimizes the logistic loss function:</p><p class="math-container">\[\min_{\theta}\; \sum_{i=1}^n \log(1 + \exp(-y_i \theta^\top x_i)).\]</p><p>Most of the time, instead of solving directly the previous optimization problem, we prefer to add a regularization term:</p><p class="math-container">\[\min_{\theta}\; \sum_{i=1}^n \log(1 + \exp(-y_i \theta^\top x_i)) + \lambda \| \theta \|\]</p><p>with <span>$\lambda \in \mathbb{R}_+$</span> a penalty and <span>$\|.\|$</span> a norm function. By adding such a regularization term, we avoid overfitting on the training set and usually achieve a greater score in cross-validation.</p><h2 id="Reformulation-as-a-conic-optimization-problem"><a class="docs-heading-anchor" href="#Reformulation-as-a-conic-optimization-problem">Reformulation as a conic optimization problem</a><a id="Reformulation-as-a-conic-optimization-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Reformulation-as-a-conic-optimization-problem" title="Permalink"></a></h2><p>By introducing auxiliary variables <span>$t_1, \cdots, t_n$</span> and <span>$r$</span>, the optimization problem is equivalent to</p><p class="math-container">\[\begin{aligned}
+\min_{t, r, \theta} \;&amp; \sum_{i=1}^n t_i + \lambda r \\
+\text{subject to } &amp; \quad t_i \geq \log(1 + \exp(- y_i \theta^\top x_i)) \\
+                   &amp; \quad r \geq \|\theta\|
+\end{aligned}\]</p><p>Now, the trick is to reformulate the constraints <span>$t_i \geq \log(1 + \exp(- y_i \theta^\top x_i))$</span> with the help of the <em>exponential cone</em></p><p class="math-container">\[K_{exp} = \{ (x, y, z) \in \mathbb{R}^3 : \; y \exp(x / y) \leq z \} .\]</p><p>Indeed, by passing to the exponential, we see that for all <span>$i=1, \cdots, n$</span>, the constraint <span>$t_i \geq \log(1 + \exp(- y_i \theta^\top x_i))$</span> is equivalent to</p><p class="math-container">\[\exp(-t_i) + \exp(u_i - t_i) \leq 1\]</p><p>with <span>$u_i = -y_i \theta^\top x_i$</span>. Then, by adding two auxiliary variables <span>$z_{i1}$</span> and <span>$z_{i2}$</span> such that <span>$z_{i1} \geq \exp(u_i-t_i)$</span> and <span>$z_{i2} \geq \exp(-t_i)$</span>, we get the equivalent formulation</p><p class="math-container">\[\left\{
+\begin{aligned}
+(u_i -t_i , 1, z_{i1}) &amp; \in  K_{exp}  \\
+(-t_i , 1, z_{i2}) &amp; \in  K_{exp}  \\
+z_{i1} + z_{i2} &amp; \leq  1
+\end{aligned}
+\right.\]</p><p>In this setting, the conic version of the logistic regression problems writes out</p><p class="math-container">\[\begin{aligned}
+\min_{t, z, r, \theta}&amp;  \; \sum_{i=1}^n t_i + \lambda r \\
+\text{subject to } &amp; \quad  (u_i -t_i , 1, z_{i1})  \in  K_{exp}  \\
+                   &amp; \quad  (-t_i , 1, z_{i2})  \in  K_{exp}  \\
+                   &amp; \quad  z_{i1} + z_{i2}  \leq  1 \\
+                   &amp; \quad u_i = -y_i x_i^\top \theta \\
+                   &amp; \quad r \geq \|\theta\|
+\end{aligned}\]</p><p>and thus encompasses <span>$3n + p + 1$</span> variables and <span>$3n + 1$</span> constraints (<span>$u_i = -y_i \theta^\top x_i$</span> is only a virtual constraint used to clarify the notation). Thus, if <span>$n \gg 1$</span>, we get a large number of variables and constraints.</p><h2 id="Fitting-logistic-regression-with-a-conic-solver"><a class="docs-heading-anchor" href="#Fitting-logistic-regression-with-a-conic-solver">Fitting logistic regression with a conic solver</a><a id="Fitting-logistic-regression-with-a-conic-solver-1"></a><a class="docs-heading-anchor-permalink" href="#Fitting-logistic-regression-with-a-conic-solver" title="Permalink"></a></h2><p>It is now time to pass to the implementation. We choose SCS as a conic solver.</p><pre><code class="language-julia hljs">using JuMP
+import Random
+import SCS</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This tutorial uses sets from <a href="../../../moi/#moi_documentation">MathOptInterface</a>. By default, JuMP exports the <code>MOI</code> symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre></div></div><pre><code class="language-julia hljs">Random.seed!(2713);</code></pre><p>We start by implementing a function to generate a fake dataset, and where we could tune the correlation between the feature variables. The function is a direct transcription of the one used in <a href="http://fa.bianp.net/blog/2013/numerical-optimizers-for-logistic-regression/">this blog post</a>.</p><pre><code class="language-julia hljs">function generate_dataset(n_samples = 100, n_features = 10; shift = 0.0)
+    X = randn(n_samples, n_features)
+    w = randn(n_features)
+    y = sign.(X * w)
+    X .+= 0.8 * randn(n_samples, n_features) # add noise
+    X .+= shift # shift the points in the feature space
+    X = hcat(X, ones(n_samples, 1))
+    return X, y
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">generate_dataset (generic function with 3 methods)</code></pre><p>We write a <code>softplus</code> function to formulate each constraint <span>$t \geq \log(1 + \exp(u))$</span> with two exponential cones.</p><pre><code class="language-julia hljs">function softplus(model, t, u)
+    z = @variable(model, [1:2], lower_bound = 0.0)
+    @constraint(model, sum(z) &lt;= 1.0)
+    @constraint(model, [u - t, 1, z[1]] in MOI.ExponentialCone())
+    @constraint(model, [-t, 1, z[2]] in MOI.ExponentialCone())
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">softplus (generic function with 1 method)</code></pre><h3 id="\\ell_2-regularized-logistic-regression"><a class="docs-heading-anchor" href="#\\ell_2-regularized-logistic-regression"><span>$\ell_2$</span> regularized logistic regression</a><a id="\\ell_2-regularized-logistic-regression-1"></a><a class="docs-heading-anchor-permalink" href="#\\ell_2-regularized-logistic-regression" title="Permalink"></a></h3><p>Then, with the help of the <code>softplus</code> function, we could write our optimization model. In the <span>$\ell_2$</span> regularization case, the constraint <span>$r \geq \|\theta\|_2$</span> rewrites as a second order cone constraint.</p><pre><code class="language-julia hljs">function build_logit_model(X, y, λ)
+    n, p = size(X)
+    model = Model()
+    @variable(model, θ[1:p])
+    @variable(model, t[1:n])
+    for i in 1:n
+        u = -(X[i, :]&#39; * θ) * y[i]
+        softplus(model, t[i], u)
+    end
+    # Add ℓ2 regularization
+    @variable(model, 0.0 &lt;= reg)
+    @constraint(model, [reg; θ] in SecondOrderCone())
+    # Define objective
+    @objective(model, Min, sum(t) + λ * reg)
+    return model
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">build_logit_model (generic function with 1 method)</code></pre><p>We generate the dataset.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Be careful here, for large n and p SCS could fail to converge.</p></div></div><pre><code class="language-julia hljs">n, p = 200, 10
+X, y = generate_dataset(n, p; shift = 10.0);
+
+# We could now solve the logistic regression problem
+λ = 10.0
+model = build_logit_model(X, y, λ)
+set_optimizer(model, SCS.Optimizer)
+set_silent(model)
+JuMP.optimize!(model)</code></pre><pre><code class="language-julia hljs">θ♯ = JuMP.value.(model[:θ])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">11-element Vector{Float64}:
+  0.02041320313399188
+  0.16139903246568316
+  0.3570077183843268
+ -0.3078808218759439
+ -0.3939184218209453
+ -0.05914040150982515
+  0.3471736591163877
+ -0.8812123235444688
+  0.20125660912000326
+  0.5409398851901871
+  0.08090419699796532</code></pre><p>It appears that the speed of convergence is not that impacted by the correlation of the dataset, nor by the penalty <span>$\lambda$</span>.</p><h3 id="\\ell_1-regularized-logistic-regression"><a class="docs-heading-anchor" href="#\\ell_1-regularized-logistic-regression"><span>$\ell_1$</span> regularized logistic regression</a><a id="\\ell_1-regularized-logistic-regression-1"></a><a class="docs-heading-anchor-permalink" href="#\\ell_1-regularized-logistic-regression" title="Permalink"></a></h3><p>We now formulate the logistic problem with a <span>$\ell_1$</span> regularization term. The <span>$\ell_1$</span> regularization ensures sparsity in the optimal solution of the resulting optimization problem. Luckily, the <span>$\ell_1$</span> norm is implemented as a set in <code>MathOptInterface</code>. Thus, we could formulate the sparse logistic regression problem with the help of a <code>MOI.NormOneCone</code> set.</p><pre><code class="language-julia hljs">function build_sparse_logit_model(X, y, λ)
+    n, p = size(X)
+    model = Model()
+    @variable(model, θ[1:p])
+    @variable(model, t[1:n])
+    for i in 1:n
+        u = -(X[i, :]&#39; * θ) * y[i]
+        softplus(model, t[i], u)
+    end
+    # Add ℓ1 regularization
+    @variable(model, 0.0 &lt;= reg)
+    @constraint(model, [reg; θ] in MOI.NormOneCone(p + 1))
+    # Define objective
+    @objective(model, Min, sum(t) + λ * reg)
+    return model
+end
+
+# Auxiliary function to count non-null components:
+count_nonzero(v::Vector; tol = 1e-6) = sum(abs.(v) .&gt;= tol)
+
+# We solve the sparse logistic regression problem on the same dataset as before.
+λ = 10.0
+sparse_model = build_sparse_logit_model(X, y, λ)
+set_optimizer(sparse_model, SCS.Optimizer)
+set_silent(sparse_model)
+JuMP.optimize!(sparse_model)</code></pre><pre><code class="language-julia hljs">θ♯ = JuMP.value.(sparse_model[:θ])
+println(
+    &quot;Number of non-zero components: &quot;,
+    count_nonzero(θ♯),
+    &quot; (out of &quot;,
+    p,
+    &quot; features)&quot;,
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Number of non-zero components: 8 (out of 10 features)</code></pre><h3 id="Extensions"><a class="docs-heading-anchor" href="#Extensions">Extensions</a><a id="Extensions-1"></a><a class="docs-heading-anchor-permalink" href="#Extensions" title="Permalink"></a></h3><p>A direct extension would be to consider the sparse logistic regression with <em>hard</em> thresholding, which, on contrary to the <em>soft</em> version using a <span>$\ell_1$</span> regularization, adds an explicit cardinality constraint in its formulation:</p><p class="math-container">\[\begin{aligned}
+\min_{\theta} &amp; \; \sum_{i=1}^n \log(1 + \exp(-y_i \theta^\top x_i)) + \lambda \| \theta \|_2^2 \\
+\text{subject to } &amp; \quad \| \theta \|_0 &lt;= k
+\end{aligned}\]</p><p>where <span>$k$</span> is the maximum number of non-zero components in the vector <span>$\theta$</span>, and <span>$\|.\|_0$</span> is the <span>$\ell_0$</span> pseudo-norm:</p><p class="math-container">\[\| x\|_0 = \#\{i : \; x_i \neq 0\}\]</p><p>The cardinality constraint <span>$\|\theta\|_0 \leq k$</span> could be reformulated with binary variables. Thus the hard sparse regression problem could be solved by any solver supporting mixed integer conic problems.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../arbitrary_precision/">« Arbitrary precision arithmetic</a><a class="docs-footer-nextpage" href="../experiment_design/">Experiment design »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/min_ellipse.jl b/previews/PR3525/tutorials/conic/min_ellipse.jl
new file mode 100644
index 00000000000..1fc33027a85
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/min_ellipse.jl
@@ -0,0 +1,151 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Minimal ellipses
+
+# This example comes from section 8.4.1 of the book *Convex Optimization* by
+# [Boyd and Vandenberghe (2004)](https://web.stanford.edu/~boyd/cvxbook/).
+
+# ## Formulation
+
+# Given a set of ``m`` ellipses of the form:
+# ```math
+# E(A, b, c) = \{ x : x^\top A x + 2 b^\top x + c \leq 0 \},
+# ```
+# the minimal ellipse problem finds an ellipse with the minimum area that
+# encloses the given ellipses.
+
+# It is convenient to parameterize the minimal enclosing ellipse as
+# ```math
+# \{ x : || Px + q || \leq 1 \}.
+# ```
+
+# Then the optimal ``P`` and ``q`` are given by the convex semidefinite program;
+# ```math
+# \begin{aligned}
+# \text{maximize } & \quad \log(\det(P))  \\
+# \text{subject to } & \quad \tau_i \geq 0, & i = 1, \ldots, m \\
+# & \quad\begin{bmatrix}
+#     P^2 - \tau_i A_i        &  P q - \tau_i b_i   &    0        \\
+#     (P q - \tau_i b_i)^\top &   -1 - \tau_i c_i   &  (P q)^\top \\
+#                 0           &          (P q)      & - P^2       \\
+# \end{bmatrix}
+# \preceq 0 \text{ (PSD) } &  i=1, \ldots, m
+# \end{aligned}
+# ```
+# with helper variables ``\tau``.
+
+# ## Required packages
+
+# This tutorial uses the following packages:
+
+using JuMP
+import LinearAlgebra
+import Plots
+import SCS
+import Test  #src
+
+# ## Data
+
+# First, define the ``m`` input ellipses (here ``m = 6``), parameterized as
+# ``x^T A_i x + 2 b_i^T x + c \leq 0``:
+
+struct Ellipse
+    A::Matrix{Float64}
+    b::Vector{Float64}
+    c::Float64
+    function Ellipse(A::Matrix{Float64}, b::Vector{Float64}, c::Float64)
+        @assert isreal(A) && LinearAlgebra.issymmetric(A)
+        return new(A, b, c)
+    end
+end
+
+ellipses = [
+    Ellipse([1.2576 -0.3873; -0.3873 0.3467], [0.2722, 0.1969], 0.1831),
+    Ellipse([1.4125 -2.1777; -2.1777 6.7775], [-1.228, -0.0521], 0.3295),
+    Ellipse([1.7018 0.8141; 0.8141 1.7538], [-0.4049, 1.5713], 0.2077),
+    Ellipse([0.9742 -0.7202; -0.7202 1.5444], [0.0265, 0.5623], 0.2362),
+    Ellipse([0.6798 -0.1424; -0.1424 0.6871], [-0.4301, -1.0157], 0.3284),
+    Ellipse([0.1796 -0.1423; -0.1423 2.6181], [-0.3286, 0.557], 0.4931),
+];
+
+# We visualise the ellipses using the Plots package:
+
+function plot_ellipse(plot, ellipse::Ellipse)
+    A, b, c = ellipse.A, ellipse.b, ellipse.c
+    θ = range(0, 2pi + 0.05; step = 0.05)
+    ## Some linear algebra to convert θ into (x,y) coordinates.
+    x_y = √A \ (√(b' * (A \ b) - c) .* hcat(cos.(θ), sin.(θ)) .- (√A \ b)')'
+    Plots.plot!(plot, x_y[1, :], x_y[2, :]; label = nothing, c = :navy)
+    return
+end
+
+plot = Plots.plot(; size = (600, 600))
+for ellipse in ellipses
+    plot_ellipse(plot, ellipse)
+end
+plot
+
+# ## Build the model
+
+# Now let's build the model, using the change-of-variables `P²` = ``P^2`` and
+# `P_q` = ``P q``. We'll recover the true value of `P` and `q` after the solve.
+
+model = Model(SCS.Optimizer)
+## We need to use a tighter tolerance for this example, otherwise the bounding
+## ellipse won't actually be bounding...
+set_attribute(model, "eps_rel", 1e-6)
+set_silent(model)
+m, n = length(ellipses), size(first(ellipses).A, 1)
+@variable(model, τ[1:m] >= 0)
+@variable(model, P²[1:n, 1:n], PSD)
+@variable(model, P_q[1:n])
+
+for (i, ellipse) in enumerate(ellipses)
+    A, b, c = ellipse.A, ellipse.b, ellipse.c
+    X = [
+        #! format: off
+        (P² - τ[i] * A)   (P_q - τ[i] * b) zeros(n, n)
+        (P_q - τ[i] * b)' (-1 - τ[i] * c)  P_q'
+        zeros(n, n)       P_q              -P²
+        #! format: on
+    ]
+    @constraint(model, LinearAlgebra.Symmetric(X) <= 0, PSDCone())
+end
+
+# We cannot directly represent the objective ``\log(\det(P))``, so we introduce
+# the conic reformulation:
+
+@variable(model, log_det_P)
+@constraint(model, [log_det_P; 1; vec(P²)] in MOI.LogDetConeSquare(n))
+@objective(model, Max, log_det_P)
+
+# Now, solve the program:
+
+optimize!(model)
+Test.@test termination_status(model) == OPTIMAL    #src
+Test.@test primal_status(model) == FEASIBLE_POINT  #src
+solution_summary(model)
+
+# ## Results
+
+# After solving the model to optimality we can recover the solution in terms of
+# ``P`` and ``q``:
+
+P = sqrt(value.(P²))
+q = P \ value.(P_q)
+
+# Finally, overlaying the solution in the plot we see the minimal area enclosing
+# ellipsoid:
+
+Test.@test isapprox(P, [0.4237 -0.0396; -0.0396 0.3163]; atol = 1e-2)  #src
+Test.@test isapprox(q, [-0.3960, -0.0214]; atol = 1e-2)                #src
+
+Plots.plot!(
+    plot,
+    [tuple(P \ [cos(θ) - q[1], sin(θ) - q[2]]...) for θ in 0:0.05:(2pi+0.05)];
+    c = :crimson,
+    label = nothing,
+)
diff --git a/previews/PR3525/tutorials/conic/min_ellipse/4146ff43.svg b/previews/PR3525/tutorials/conic/min_ellipse/4146ff43.svg
new file mode 100644
index 00000000000..69fc5d8ef11
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/min_ellipse/4146ff43.svg
@@ -0,0 +1,53 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="600" viewBox="0 0 2400 2400">
+<defs>
+  <clipPath id="clip670">
+    <rect x="0" y="0" width="2400" height="2400"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip670)" d="M0 2400 L2400 2400 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip671">
+    <rect x="480" y="240" width="1681" height="1681"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip670)" d="M148.057 2261.17 L2352.76 2261.17 L2352.76 47.2441 L148.057 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip672">
+    <rect x="148" y="47" width="2206" height="2215"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="406.43,2261.17 406.43,47.2441 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="838.578,2261.17 838.578,47.2441 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1270.73,2261.17 1270.73,47.2441 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1702.88,2261.17 1702.88,47.2441 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2135.02,2261.17 2135.02,47.2441 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,2261.17 2352.76,2261.17 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="406.43,2261.17 406.43,2242.27 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="838.578,2261.17 838.578,2242.27 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1270.73,2261.17 1270.73,2242.27 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1702.88,2261.17 1702.88,2242.27 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2135.02,2261.17 2135.02,2242.27 "/>
+<path clip-path="url(#clip670)" d="M376.187 2306.74 L405.862 2306.74 L405.862 2310.67 L376.187 2310.67 L376.187 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M416.765 2319.63 L424.404 2319.63 L424.404 2293.27 L416.094 2294.93 L416.094 2290.67 L424.358 2289.01 L429.034 2289.01 L429.034 2319.63 L436.672 2319.63 L436.672 2323.57 L416.765 2323.57 L416.765 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M838.578 2292.09 Q834.967 2292.09 833.138 2295.65 Q831.333 2299.19 831.333 2306.32 Q831.333 2313.43 833.138 2316.99 Q834.967 2320.54 838.578 2320.54 Q842.212 2320.54 844.018 2316.99 Q845.847 2313.43 845.847 2306.32 Q845.847 2299.19 844.018 2295.65 Q842.212 2292.09 838.578 2292.09 M838.578 2288.38 Q844.388 2288.38 847.444 2292.99 Q850.523 2297.57 850.523 2306.32 Q850.523 2315.05 847.444 2319.66 Q844.388 2324.24 838.578 2324.24 Q832.768 2324.24 829.689 2319.66 Q826.634 2315.05 826.634 2306.32 Q826.634 2297.57 829.689 2292.99 Q832.768 2288.38 838.578 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M1261.11 2319.63 L1268.75 2319.63 L1268.75 2293.27 L1260.44 2294.93 L1260.44 2290.67 L1268.7 2289.01 L1273.38 2289.01 L1273.38 2319.63 L1281.02 2319.63 L1281.02 2323.57 L1261.11 2323.57 L1261.11 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M1697.53 2319.63 L1713.85 2319.63 L1713.85 2323.57 L1691.9 2323.57 L1691.9 2319.63 Q1694.57 2316.88 1699.15 2312.25 Q1703.75 2307.6 1704.94 2306.25 Q1707.18 2303.73 1708.06 2301.99 Q1708.96 2300.23 1708.96 2298.54 Q1708.96 2295.79 1707.02 2294.05 Q1705.1 2292.32 1702 2292.32 Q1699.8 2292.32 1697.34 2293.08 Q1694.91 2293.85 1692.13 2295.4 L1692.13 2290.67 Q1694.96 2289.54 1697.41 2288.96 Q1699.87 2288.38 1701.9 2288.38 Q1707.27 2288.38 1710.47 2291.07 Q1713.66 2293.75 1713.66 2298.24 Q1713.66 2300.37 1712.85 2302.29 Q1712.07 2304.19 1709.96 2306.79 Q1709.38 2307.46 1706.28 2310.67 Q1703.18 2313.87 1697.53 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M2139.27 2304.93 Q2142.63 2305.65 2144.5 2307.92 Q2146.4 2310.19 2146.4 2313.52 Q2146.4 2318.64 2142.88 2321.44 Q2139.36 2324.24 2132.88 2324.24 Q2130.71 2324.24 2128.39 2323.8 Q2126.1 2323.38 2123.65 2322.53 L2123.65 2318.01 Q2125.59 2319.15 2127.91 2319.73 Q2130.22 2320.3 2132.74 2320.3 Q2137.14 2320.3 2139.43 2318.57 Q2141.75 2316.83 2141.75 2313.52 Q2141.75 2310.47 2139.6 2308.75 Q2137.47 2307.02 2133.65 2307.02 L2129.62 2307.02 L2129.62 2303.17 L2133.83 2303.17 Q2137.28 2303.17 2139.11 2301.81 Q2140.94 2300.42 2140.94 2297.83 Q2140.94 2295.17 2139.04 2293.75 Q2137.17 2292.32 2133.65 2292.32 Q2131.73 2292.32 2129.53 2292.73 Q2127.33 2293.15 2124.69 2294.03 L2124.69 2289.86 Q2127.35 2289.12 2129.67 2288.75 Q2132 2288.38 2134.06 2288.38 Q2139.39 2288.38 2142.49 2290.81 Q2145.59 2293.22 2145.59 2297.34 Q2145.59 2300.21 2143.95 2302.2 Q2142.3 2304.17 2139.27 2304.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="148.057,2190 2352.76,2190 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="148.057,1864.99 2352.76,1864.99 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="148.057,1539.98 2352.76,1539.98 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="148.057,1214.97 2352.76,1214.97 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="148.057,889.962 2352.76,889.962 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="148.057,564.952 2352.76,564.952 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="148.057,239.942 2352.76,239.942 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,2261.17 148.057,47.2441 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,2190 166.954,2190 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,1864.99 166.954,1864.99 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,1539.98 166.954,1539.98 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,1214.97 166.954,1214.97 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,889.962 166.954,889.962 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,564.952 166.954,564.952 "/>
+<polyline clip-path="url(#clip670)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="148.057,239.942 166.954,239.942 "/>
+<path clip-path="url(#clip670)" d="M50.9921 2190.45 L80.6679 2190.45 L80.6679 2194.39 L50.9921 2194.39 L50.9921 2190.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M104.927 2188.65 Q108.283 2189.36 110.158 2191.63 Q112.057 2193.9 112.057 2197.23 Q112.057 2202.35 108.538 2205.15 Q105.02 2207.95 98.5382 2207.95 Q96.3623 2207.95 94.0475 2207.51 Q91.7558 2207.1 89.3021 2206.24 L89.3021 2201.72 Q91.2465 2202.86 93.5613 2203.44 Q95.8761 2204.02 98.3993 2204.02 Q102.797 2204.02 105.089 2202.28 Q107.404 2200.54 107.404 2197.23 Q107.404 2194.18 105.251 2192.47 Q103.121 2190.73 99.3021 2190.73 L95.2743 2190.73 L95.2743 2186.89 L99.4872 2186.89 Q102.936 2186.89 104.765 2185.52 Q106.594 2184.13 106.594 2181.54 Q106.594 2178.88 104.696 2177.47 Q102.821 2176.03 99.3021 2176.03 Q97.3808 2176.03 95.1817 2176.45 Q92.9826 2176.86 90.3438 2177.74 L90.3438 2173.58 Q93.0058 2172.84 95.3206 2172.47 Q97.6585 2172.1 99.7187 2172.1 Q105.043 2172.1 108.145 2174.53 Q111.246 2176.93 111.246 2181.05 Q111.246 2183.92 109.603 2185.91 Q107.959 2187.88 104.927 2188.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M51.9412 1865.44 L81.617 1865.44 L81.617 1869.38 L51.9412 1869.38 L51.9412 1865.44 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M95.7373 1878.34 L112.057 1878.34 L112.057 1882.27 L90.1123 1882.27 L90.1123 1878.34 Q92.7743 1875.58 97.3576 1870.95 Q101.964 1866.3 103.145 1864.96 Q105.39 1862.43 106.27 1860.7 Q107.172 1858.94 107.172 1857.25 Q107.172 1854.49 105.228 1852.76 Q103.307 1851.02 100.205 1851.02 Q98.0058 1851.02 95.5521 1851.78 Q93.1215 1852.55 90.3438 1854.1 L90.3438 1849.38 Q93.1678 1848.24 95.6215 1847.66 Q98.0752 1847.09 100.112 1847.09 Q105.483 1847.09 108.677 1849.77 Q111.871 1852.46 111.871 1856.95 Q111.871 1859.08 111.061 1861 Q110.274 1862.9 108.168 1865.49 Q107.589 1866.16 104.487 1869.38 Q101.385 1872.57 95.7373 1878.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M51.5708 1540.43 L81.2466 1540.43 L81.2466 1544.37 L51.5708 1544.37 L51.5708 1540.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M92.1493 1553.33 L99.7882 1553.33 L99.7882 1526.96 L91.478 1528.63 L91.478 1524.37 L99.7419 1522.7 L104.418 1522.7 L104.418 1553.33 L112.057 1553.33 L112.057 1557.26 L92.1493 1557.26 L92.1493 1553.33 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M100.112 1200.77 Q96.5011 1200.77 94.6724 1204.33 Q92.8669 1207.88 92.8669 1215.01 Q92.8669 1222.11 94.6724 1225.68 Q96.5011 1229.22 100.112 1229.22 Q103.746 1229.22 105.552 1225.68 Q107.381 1222.11 107.381 1215.01 Q107.381 1207.88 105.552 1204.33 Q103.746 1200.77 100.112 1200.77 M100.112 1197.07 Q105.922 1197.07 108.978 1201.67 Q112.057 1206.26 112.057 1215.01 Q112.057 1223.73 108.978 1228.34 Q105.922 1232.92 100.112 1232.92 Q94.3021 1232.92 91.2234 1228.34 Q88.1679 1223.73 88.1679 1215.01 Q88.1679 1206.26 91.2234 1201.67 Q94.3021 1197.07 100.112 1197.07 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M92.1493 903.307 L99.7882 903.307 L99.7882 876.941 L91.478 878.608 L91.478 874.348 L99.7419 872.682 L104.418 872.682 L104.418 903.307 L112.057 903.307 L112.057 907.242 L92.1493 907.242 L92.1493 903.307 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M95.7373 578.297 L112.057 578.297 L112.057 582.232 L90.1123 582.232 L90.1123 578.297 Q92.7743 575.542 97.3576 570.913 Q101.964 566.26 103.145 564.917 Q105.39 562.394 106.27 560.658 Q107.172 558.899 107.172 557.209 Q107.172 554.454 105.228 552.718 Q103.307 550.982 100.205 550.982 Q98.0058 550.982 95.5521 551.746 Q93.1215 552.51 90.3438 554.061 L90.3438 549.339 Q93.1678 548.204 95.6215 547.626 Q98.0752 547.047 100.112 547.047 Q105.483 547.047 108.677 549.732 Q111.871 552.417 111.871 556.908 Q111.871 559.038 111.061 560.959 Q110.274 562.857 108.168 565.45 Q107.589 566.121 104.487 569.339 Q101.385 572.533 95.7373 578.297 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip670)" d="M104.927 238.588 Q108.283 239.306 110.158 241.574 Q112.057 243.843 112.057 247.176 Q112.057 252.292 108.538 255.093 Q105.02 257.894 98.5382 257.894 Q96.3623 257.894 94.0475 257.454 Q91.7558 257.037 89.3021 256.181 L89.3021 251.667 Q91.2465 252.801 93.5613 253.38 Q95.8761 253.959 98.3993 253.959 Q102.797 253.959 105.089 252.222 Q107.404 250.486 107.404 247.176 Q107.404 244.121 105.251 242.408 Q103.121 240.672 99.3021 240.672 L95.2743 240.672 L95.2743 236.829 L99.4872 236.829 Q102.936 236.829 104.765 235.463 Q106.594 234.074 106.594 231.482 Q106.594 228.82 104.696 227.408 Q102.821 225.973 99.3021 225.973 Q97.3808 225.973 95.1817 226.389 Q92.9826 226.806 90.3438 227.686 L90.3438 223.519 Q93.0058 222.778 95.3206 222.408 Q97.6585 222.037 99.7187 222.037 Q105.043 222.037 108.145 224.468 Q111.246 226.875 111.246 230.996 Q111.246 233.866 109.603 235.857 Q107.959 237.824 104.927 238.588 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip672)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="786.163,1548.29 790.416,1532.67 794.146,1517.27 797.343,1502.11 799.998,1487.23 802.105,1472.68 803.658,1458.49 804.655,1444.69 805.092,1431.32 804.968,1418.42 804.284,1406 803.042,1394.11 801.243,1382.78 798.894,1372.03 795.999,1361.89 792.567,1352.39 788.605,1343.54 784.123,1335.38 779.133,1327.92 773.648,1321.18 767.68,1315.17 761.245,1309.92 754.359,1305.43 747.039,1301.73 739.303,1298.8 731.171,1296.67 722.663,1295.34 713.801,1294.81 704.606,1295.08 695.101,1296.16 685.311,1298.03 675.259,1300.71 664.971,1304.16 654.473,1308.4 643.79,1313.41 632.95,1319.18 621.979,1325.68 610.906,1332.92 599.757,1340.86 588.56,1349.49 577.344,1358.78 566.137,1368.72 554.966,1379.28 543.86,1390.43 532.846,1402.14 521.951,1414.39 511.204,1427.15 500.63,1440.37 490.257,1454.04 480.11,1468.11 470.214,1482.55 460.595,1497.32 451.276,1512.39 442.281,1527.72 433.632,1543.28 425.351,1559.01 417.458,1574.89 409.973,1590.87 402.916,1606.91 396.302,1622.98 390.15,1639.02 384.474,1655.02 379.289,1670.91 374.607,1686.67 370.441,1702.25 366.8,1717.62 363.695,1732.73 361.131,1747.55 359.117,1762.04 357.657,1776.17 356.754,1789.9 356.411,1803.2 356.629,1816.02 357.407,1828.35 358.744,1840.15 360.635,1851.38 363.076,1862.03 366.062,1872.07 369.584,1881.46 373.635,1890.19 378.202,1898.24 383.276,1905.58 388.844,1912.2 394.892,1918.08 401.404,1923.2 408.364,1927.56 415.755,1931.14 423.559,1933.93 431.755,1935.92 440.324,1937.12 449.244,1937.52 458.493,1937.11 468.047,1935.9 477.883,1933.89 487.977,1931.08 498.302,1927.49 508.833,1923.12 519.544,1917.99 530.408,1912.1 541.398,1905.47 552.486,1898.11 563.645,1890.05 574.847,1881.31 586.063,1871.91 597.266,1861.86 608.428,1851.2 619.521,1839.95 630.517,1828.15 641.389,1815.81 652.109,1802.98 662.65,1789.68 672.987,1775.94 683.094,1761.81 692.945,1747.31 702.515,1732.48 711.781,1717.36 720.719,1701.99 729.308,1686.41 737.526,1670.65 745.351,1654.75 752.765,1638.76 759.75,1622.71 766.286,1606.64 772.36,1590.6 777.954,1574.62 783.055,1558.74 787.651,1543.01 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2268.85,906.823 2276.2,896.762 2281.84,887.035 2285.76,877.664 2287.95,868.675 2288.4,860.089 2287.12,851.929 2284.1,844.213 2279.35,836.963 2272.89,830.195 2264.73,823.926 2254.89,818.173 2243.4,812.95 2230.29,808.27 2215.58,804.144 2199.31,800.582 2181.53,797.595 2162.29,795.188 2141.61,793.369 2119.57,792.142 2096.21,791.509 2071.6,791.472 2045.79,792.033 2018.85,793.188 1990.84,794.935 1961.84,797.27 1931.92,800.187 1901.15,803.679 1869.62,807.737 1837.39,812.35 1804.55,817.508 1771.18,823.197 1737.37,829.403 1703.2,836.111 1668.76,843.304 1634.12,850.964 1599.39,859.071 1564.63,867.606 1529.95,876.547 1495.43,885.872 1461.15,895.558 1427.21,905.58 1393.67,915.913 1360.64,926.531 1328.19,937.408 1296.4,948.517 1265.35,959.83 1235.12,971.319 1205.79,982.954 1177.43,994.708 1150.1,1006.55 1123.88,1018.45 1098.84,1030.38 1075.03,1042.31 1052.52,1054.21 1031.35,1066.05 1011.6,1077.79 993.294,1089.42 976.49,1100.91 961.228,1112.21 947.545,1123.31 935.476,1134.17 925.052,1144.78 916.297,1155.1 909.235,1165.11 903.882,1174.78 900.252,1184.08 898.355,1193.01 898.194,1201.52 899.771,1209.61 903.081,1217.25 908.116,1224.42 914.863,1231.1 923.306,1237.29 933.424,1242.95 945.191,1248.08 958.578,1252.67 973.551,1256.7 990.073,1260.17 1008.1,1263.06 1027.59,1265.37 1048.5,1267.09 1070.77,1268.22 1094.34,1268.75 1119.16,1268.68 1145.17,1268.02 1172.29,1266.77 1200.47,1264.92 1229.63,1262.49 1259.7,1259.47 1290.6,1255.89 1322.26,1251.73 1354.59,1247.03 1387.53,1241.78 1420.97,1236 1454.85,1229.71 1489.07,1222.92 1523.56,1215.65 1558.21,1207.91 1592.96,1199.73 1627.71,1191.13 1662.37,1182.12 1696.85,1172.73 1731.08,1162.99 1764.96,1152.91 1798.42,1142.53 1831.36,1131.87 1863.71,1120.95 1895.38,1109.8 1926.29,1098.46 1956.38,1086.94 1985.55,1075.29 2013.75,1063.52 2040.89,1051.66 2066.91,1039.75 2091.75,1027.82 2115.35,1015.9 2137.64,1004.01 2158.57,992.18 2178.08,980.449 2196.14,968.843 2212.69,957.389 2227.69,946.118 2241.1,935.056 2252.89,924.232 2263.04,913.673 2271.51,903.404 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1746.02,1735.77 1738.68,1715.77 1730.02,1695.57 1720.05,1675.23 1708.8,1654.78 1696.31,1634.28 1682.6,1613.79 1667.7,1593.35 1651.66,1573.02 1634.5,1552.84 1616.29,1532.87 1597.06,1513.15 1576.85,1493.74 1555.73,1474.69 1533.74,1456.03 1510.94,1437.83 1487.38,1420.11 1463.12,1402.94 1438.23,1386.35 1412.77,1370.38 1386.8,1355.07 1360.38,1340.46 1333.59,1326.6 1306.48,1313.5 1279.12,1301.2 1251.6,1289.75 1223.96,1279.15 1196.29,1269.45 1168.65,1260.66 1141.11,1252.81 1113.74,1245.92 1086.6,1239.99 1059.77,1235.06 1033.32,1231.13 1007.3,1228.21 981.787,1226.3 956.841,1225.42 932.525,1225.56 908.9,1226.73 886.024,1228.92 863.955,1232.12 842.748,1236.34 822.456,1241.55 803.13,1247.74 784.818,1254.91 767.566,1263.02 751.417,1272.07 736.411,1282.02 722.585,1292.86 709.976,1304.55 698.613,1317.07 688.525,1330.39 679.738,1344.47 672.273,1359.27 666.149,1374.77 661.382,1390.92 657.983,1407.68 655.961,1425 655.321,1442.86 656.065,1461.19 658.19,1479.96 661.691,1499.12 666.56,1518.62 672.784,1538.42 680.348,1558.45 689.233,1578.68 699.417,1599.05 710.874,1619.51 723.576,1640 737.49,1660.49 752.583,1680.92 768.816,1701.23 786.149,1721.38 804.538,1741.31 823.938,1760.97 844.3,1780.33 865.573,1799.32 887.704,1817.9 910.638,1836.03 934.317,1853.65 958.682,1870.73 983.673,1887.22 1009.23,1903.08 1035.28,1918.27 1061.76,1932.76 1088.62,1946.5 1115.77,1959.47 1143.16,1971.62 1170.71,1982.93 1198.35,1993.38 1226.03,2002.93 1253.66,2011.56 1281.17,2019.26 1308.51,2025.99 1335.6,2031.74 1362.37,2036.51 1388.75,2040.27 1414.69,2043.02 1440.11,2044.76 1464.96,2045.47 1489.16,2045.15 1512.66,2043.81 1535.41,2041.45 1557.34,2038.08 1578.39,2033.7 1598.53,2028.32 1617.69,2021.96 1635.82,2014.63 1652.89,2006.36 1668.85,1997.16 1683.66,1987.06 1697.28,1976.07 1709.69,1964.24 1720.84,1951.58 1730.71,1938.13 1739.27,1923.93 1746.52,1909.01 1752.41,1893.4 1756.95,1877.14 1760.12,1860.29 1761.91,1842.87 1762.31,1824.93 1761.34,1806.52 1758.98,1787.68 1755.25,1768.46 1750.15,1748.91 1743.7,1729.07 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="806.932,1366.88 809.051,1362.15 810.756,1357.52 812.041,1353 812.904,1348.61 813.343,1344.34 813.356,1340.22 812.944,1336.25 812.108,1332.45 810.849,1328.82 809.17,1325.37 807.077,1322.12 804.574,1319.06 801.668,1316.21 798.365,1313.58 794.674,1311.16 790.604,1308.97 786.166,1307.02 781.37,1305.3 776.229,1303.82 770.755,1302.59 764.962,1301.6 758.864,1300.87 752.477,1300.39 745.817,1300.16 738.9,1300.19 731.743,1300.47 724.366,1301 716.785,1301.79 709.02,1302.82 701.09,1304.11 693.015,1305.64 684.816,1307.4 676.513,1309.41 668.126,1311.64 659.676,1314.1 651.185,1316.78 642.674,1319.68 634.164,1322.78 625.677,1326.07 617.233,1329.56 608.854,1333.22 600.56,1337.06 592.373,1341.06 584.313,1345.21 576.4,1349.5 568.653,1353.93 561.093,1358.47 553.737,1363.12 546.605,1367.87 539.715,1372.71 533.083,1377.61 526.725,1382.58 520.659,1387.6 514.899,1392.65 509.459,1397.72 504.354,1402.81 499.595,1407.89 495.195,1412.96 491.165,1417.99 487.514,1422.99 484.253,1427.94 481.389,1432.82 478.928,1437.62 476.879,1442.34 475.244,1446.95 474.03,1451.45 473.238,1455.83 472.871,1460.07 472.929,1464.16 473.412,1468.1 474.32,1471.88 475.65,1475.48 477.398,1478.89 479.561,1482.12 482.132,1485.14 485.106,1487.95 488.474,1490.55 492.23,1492.93 496.362,1495.08 500.861,1496.99 505.716,1498.67 510.914,1500.11 516.443,1501.3 522.288,1502.24 528.436,1502.93 534.87,1503.37 541.574,1503.56 548.533,1503.49 555.728,1503.17 563.141,1502.59 570.754,1501.76 578.548,1500.68 586.504,1499.36 594.601,1497.79 602.819,1495.98 611.138,1493.94 619.537,1491.67 627.995,1489.17 636.491,1486.45 645.003,1483.52 653.51,1480.39 661.992,1477.06 670.427,1473.55 678.793,1469.85 687.07,1465.98 695.237,1461.96 703.274,1457.78 711.161,1453.47 718.877,1449.02 726.404,1444.46 733.724,1439.79 740.816,1435.03 747.665,1430.18 754.252,1425.26 760.561,1420.29 766.577,1415.26 772.284,1410.21 777.668,1405.13 782.717,1400.05 787.416,1394.96 791.754,1389.9 795.721,1384.87 799.307,1379.88 802.502,1374.94 805.299,1370.07 807.691,1365.28 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1980.15,611.617 1983.01,584.925 1984.08,558.44 1983.35,532.23 1980.82,506.36 1976.5,480.894 1970.4,455.897 1962.54,431.43 1952.93,407.554 1941.59,384.33 1928.57,361.816 1913.89,340.067 1897.58,319.138 1879.69,299.082 1860.26,279.948 1839.34,261.785 1816.98,244.638 1793.24,228.549 1768.18,213.559 1741.86,199.705 1714.35,187.022 1685.71,175.542 1656.02,165.293 1625.35,156.301 1593.78,148.588 1561.38,142.174 1528.24,137.075 1494.44,133.303 1460.07,130.869 1425.21,129.777 1389.94,130.031 1354.36,131.63 1318.56,134.57 1282.62,138.844 1246.63,144.441 1210.69,151.347 1174.88,159.544 1139.29,169.013 1104.02,179.73 1069.15,191.668 1034.77,204.796 1000.96,219.083 967.805,234.492 935.394,250.985 903.805,268.521 873.117,287.056 843.407,306.544 814.749,326.935 787.214,348.18 760.872,370.224 735.788,393.013 712.026,416.489 689.643,440.595 668.697,465.269 649.24,490.451 631.319,516.077 614.981,542.083 600.266,568.404 587.211,594.975 575.848,621.728 566.205,648.598 558.308,675.517 552.175,702.417 547.822,729.232 545.26,755.894 544.495,782.336 545.529,808.494 548.36,834.301 552.98,859.692 559.378,884.605 567.538,908.977 577.439,932.747 589.057,955.855 602.363,978.245 617.324,999.859 633.901,1020.64 652.054,1040.55 671.738,1059.52 692.902,1077.52 715.495,1094.49 739.459,1110.4 764.735,1125.2 791.26,1138.86 818.967,1151.34 847.788,1162.62 877.649,1172.66 908.476,1181.43 940.193,1188.93 972.72,1195.12 1005.98,1200 1039.88,1203.55 1074.34,1205.76 1109.28,1206.62 1144.6,1206.14 1180.23,1204.32 1216.06,1201.15 1252.01,1196.66 1288,1190.84 1323.92,1183.71 1359.7,1175.3 1395.24,1165.62 1430.45,1154.7 1465.25,1142.56 1499.54,1129.23 1533.25,1114.75 1566.28,1099.16 1598.56,1082.49 1630,1064.78 1660.53,1046.08 1690.07,1026.44 1718.54,1005.9 1745.88,984.517 1772.01,962.344 1796.88,939.435 1820.42,915.848 1842.56,891.643 1863.26,866.879 1882.46,841.618 1900.12,815.924 1916.19,789.86 1930.63,763.492 1943.4,736.886 1954.47,710.108 1963.82,683.226 1971.43,656.306 1977.26,629.415 1981.31,602.621 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1982.85,1240.63 1983.21,1236.81 1982.59,1233.04 1980.98,1229.32 1978.39,1225.66 1974.82,1222.06 1970.29,1218.55 1964.81,1215.12 1958.38,1211.79 1951.03,1208.57 1942.78,1205.45 1933.64,1202.46 1923.64,1199.59 1912.8,1196.86 1901.15,1194.27 1888.72,1191.82 1875.55,1189.54 1861.65,1187.41 1847.07,1185.44 1831.85,1183.65 1816.01,1182.03 1799.62,1180.59 1782.69,1179.33 1765.28,1178.25 1747.42,1177.37 1729.17,1176.67 1710.57,1176.17 1691.67,1175.86 1672.5,1175.74 1653.13,1175.82 1633.6,1176.09 1613.95,1176.56 1594.24,1177.21 1574.52,1178.06 1554.83,1179.1 1535.23,1180.32 1515.76,1181.73 1496.47,1183.31 1477.42,1185.07 1458.64,1187 1440.19,1189.1 1422.11,1191.35 1404.44,1193.77 1387.24,1196.33 1370.54,1199.03 1354.38,1201.87 1338.81,1204.84 1323.86,1207.93 1309.58,1211.14 1295.99,1214.45 1283.14,1217.86 1271.05,1221.36 1259.75,1224.93 1249.28,1228.58 1239.65,1232.29 1230.9,1236.06 1223.04,1239.86 1216.1,1243.7 1210.08,1247.57 1205.01,1251.45 1200.91,1255.33 1197.77,1259.22 1195.61,1263.08 1194.43,1266.92 1194.23,1270.73 1195.02,1274.5 1196.8,1278.21 1199.56,1281.86 1203.28,1285.44 1207.98,1288.94 1213.62,1292.35 1220.2,1295.66 1227.7,1298.87 1236.11,1301.97 1245.39,1304.94 1255.53,1307.79 1266.51,1310.5 1278.29,1313.06 1290.85,1315.48 1304.15,1317.74 1318.17,1319.84 1332.85,1321.78 1348.18,1323.54 1364.11,1325.13 1380.61,1326.55 1397.62,1327.77 1415.11,1328.82 1433.03,1329.67 1451.34,1330.33 1470,1330.8 1488.95,1331.08 1508.15,1331.17 1527.55,1331.06 1547.11,1330.75 1566.77,1330.25 1586.48,1329.56 1606.2,1328.68 1625.88,1327.62 1645.46,1326.36 1664.91,1324.93 1684.16,1323.31 1703.17,1321.53 1721.9,1319.57 1740.29,1317.44 1758.3,1315.16 1775.9,1312.72 1793.02,1310.13 1809.63,1307.41 1825.69,1304.54 1841.16,1301.55 1855.99,1298.44 1870.16,1295.22 1883.63,1291.89 1896.36,1288.47 1908.32,1284.96 1919.48,1281.37 1929.81,1277.71 1939.29,1273.99 1947.89,1270.21 1955.6,1266.4 1962.39,1262.56 1968.25,1258.69 1973.15,1254.81 1977.1,1250.92 1980.07,1247.04 1982.07,1243.18 1983.09,1239.34 "/>
+<polyline clip-path="url(#clip672)" style="stroke:#dc143c; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2282.61,1056.92 2287.78,1005.07 2290.36,953.582 2290.34,902.599 2287.72,852.246 2282.51,802.647 2274.72,753.926 2264.37,706.206 2251.49,659.606 2236.1,614.242 2218.25,570.227 2197.98,527.672 2175.34,486.683 2150.4,447.362 2123.2,409.808 2093.82,374.115 2062.33,340.371 2028.82,308.661 1993.35,279.064 1956.04,251.655 1916.95,226.502 1876.21,203.667 1833.9,183.208 1790.13,165.176 1745.01,149.616 1698.65,136.567 1651.18,126.062 1602.7,118.126 1553.34,112.779 1503.23,110.036 1452.49,109.902 1401.24,112.379 1349.61,117.459 1297.73,125.131 1245.74,135.375 1193.76,148.165 1141.92,163.469 1090.35,181.25 1039.19,201.463 988.548,224.057 938.564,248.976 889.359,276.157 841.057,305.533 793.778,337.03 747.641,370.569 702.761,406.068 659.249,443.435 617.216,482.58 576.765,523.403 537.999,565.802 501.013,609.673 465.901,654.904 432.749,701.383 401.642,748.994 372.656,797.617 345.865,847.132 321.334,897.414 299.126,948.338 279.296,999.776 261.893,1051.6 246.962,1103.68 234.538,1155.89 224.654,1208.09 217.335,1260.16 212.597,1311.96 210.454,1363.37 210.91,1414.26 213.965,1464.49 219.61,1513.95 227.832,1562.51 238.611,1610.05 251.918,1656.45 267.721,1701.6 285.981,1745.37 306.652,1787.67 329.681,1828.39 355.013,1867.42 382.582,1904.67 412.321,1940.04 444.155,1973.45 478.004,2004.81 513.784,2034.04 551.406,2061.07 590.775,2085.84 631.792,2108.28 674.357,2128.34 718.361,2145.95 763.695,2161.09 810.246,2173.72 857.898,2183.79 906.531,2191.3 956.023,2196.2 1006.25,2198.51 1057.09,2198.2 1108.41,2195.29 1160.09,2189.77 1211.99,2181.67 1263.99,2170.99 1315.96,2157.78 1367.76,2142.05 1419.27,2123.86 1470.36,2103.24 1520.9,2080.25 1570.76,2054.95 1619.82,2027.4 1667.96,1997.66 1715.05,1965.81 1760.99,1931.94 1805.65,1896.12 1848.92,1858.44 1890.69,1819.01 1930.86,1777.92 1969.34,1735.26 2006.02,1691.15 2040.8,1645.71 2073.62,1599.03 2104.37,1551.24 2133,1502.46 2159.41,1452.8 2183.56,1402.41 2205.37,1351.39 2224.79,1299.87 2241.78,1248 2256.29,1195.89 2268.29,1143.67 2277.75,1091.48 2284.63,1039.45 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/conic/min_ellipse/804a7acd.svg b/previews/PR3525/tutorials/conic/min_ellipse/804a7acd.svg
new file mode 100644
index 00000000000..d783158add5
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/min_ellipse/804a7acd.svg
@@ -0,0 +1,50 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="600" viewBox="0 0 2400 2400">
+<defs>
+  <clipPath id="clip610">
+    <rect x="0" y="0" width="2400" height="2400"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip610)" d="M0 2400 L2400 2400 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip611">
+    <rect x="480" y="240" width="1681" height="1681"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip610)" d="M147.478 2261.17 L2352.76 2261.17 L2352.76 47.2441 L147.478 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip612">
+    <rect x="147" y="47" width="2206" height="2215"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="263.754,2261.17 263.754,47.2441 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="729.11,2261.17 729.11,47.2441 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1194.47,2261.17 1194.47,47.2441 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1659.82,2261.17 1659.82,47.2441 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2125.18,2261.17 2125.18,47.2441 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,2261.17 2352.76,2261.17 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="263.754,2261.17 263.754,2242.27 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="729.11,2261.17 729.11,2242.27 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1194.47,2261.17 1194.47,2242.27 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1659.82,2261.17 1659.82,2242.27 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2125.18,2261.17 2125.18,2242.27 "/>
+<path clip-path="url(#clip610)" d="M233.511 2306.74 L263.186 2306.74 L263.186 2310.67 L233.511 2310.67 L233.511 2306.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M274.089 2319.63 L281.728 2319.63 L281.728 2293.27 L273.418 2294.93 L273.418 2290.67 L281.682 2289.01 L286.358 2289.01 L286.358 2319.63 L293.996 2319.63 L293.996 2323.57 L274.089 2323.57 L274.089 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M729.11 2292.09 Q725.499 2292.09 723.67 2295.65 Q721.864 2299.19 721.864 2306.32 Q721.864 2313.43 723.67 2316.99 Q725.499 2320.54 729.11 2320.54 Q732.744 2320.54 734.55 2316.99 Q736.378 2313.43 736.378 2306.32 Q736.378 2299.19 734.55 2295.65 Q732.744 2292.09 729.11 2292.09 M729.11 2288.38 Q734.92 2288.38 737.975 2292.99 Q741.054 2297.57 741.054 2306.32 Q741.054 2315.05 737.975 2319.66 Q734.92 2324.24 729.11 2324.24 Q723.3 2324.24 720.221 2319.66 Q717.165 2315.05 717.165 2306.32 Q717.165 2297.57 720.221 2292.99 Q723.3 2288.38 729.11 2288.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M1184.85 2319.63 L1192.49 2319.63 L1192.49 2293.27 L1184.18 2294.93 L1184.18 2290.67 L1192.44 2289.01 L1197.12 2289.01 L1197.12 2319.63 L1204.76 2319.63 L1204.76 2323.57 L1184.85 2323.57 L1184.85 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M1654.47 2319.63 L1670.79 2319.63 L1670.79 2323.57 L1648.85 2323.57 L1648.85 2319.63 Q1651.51 2316.88 1656.1 2312.25 Q1660.7 2307.6 1661.88 2306.25 Q1664.13 2303.73 1665.01 2301.99 Q1665.91 2300.23 1665.91 2298.54 Q1665.91 2295.79 1663.97 2294.05 Q1662.04 2292.32 1658.94 2292.32 Q1656.74 2292.32 1654.29 2293.08 Q1651.86 2293.85 1649.08 2295.4 L1649.08 2290.67 Q1651.91 2289.54 1654.36 2288.96 Q1656.81 2288.38 1658.85 2288.38 Q1664.22 2288.38 1667.41 2291.07 Q1670.61 2293.75 1670.61 2298.24 Q1670.61 2300.37 1669.8 2302.29 Q1669.01 2304.19 1666.91 2306.79 Q1666.33 2307.46 1663.22 2310.67 Q1660.12 2313.87 1654.47 2319.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M2129.43 2304.93 Q2132.78 2305.65 2134.66 2307.92 Q2136.56 2310.19 2136.56 2313.52 Q2136.56 2318.64 2133.04 2321.44 Q2129.52 2324.24 2123.04 2324.24 Q2120.86 2324.24 2118.55 2323.8 Q2116.25 2323.38 2113.8 2322.53 L2113.8 2318.01 Q2115.75 2319.15 2118.06 2319.73 Q2120.37 2320.3 2122.9 2320.3 Q2127.3 2320.3 2129.59 2318.57 Q2131.9 2316.83 2131.9 2313.52 Q2131.9 2310.47 2129.75 2308.75 Q2127.62 2307.02 2123.8 2307.02 L2119.77 2307.02 L2119.77 2303.17 L2123.99 2303.17 Q2127.44 2303.17 2129.26 2301.81 Q2131.09 2300.42 2131.09 2297.83 Q2131.09 2295.17 2129.19 2293.75 Q2127.32 2292.32 2123.8 2292.32 Q2121.88 2292.32 2119.68 2292.73 Q2117.48 2293.15 2114.84 2294.03 L2114.84 2289.86 Q2117.5 2289.12 2119.82 2288.75 Q2122.16 2288.38 2124.22 2288.38 Q2129.54 2288.38 2132.64 2290.81 Q2135.75 2293.22 2135.75 2297.34 Q2135.75 2300.21 2134.1 2302.2 Q2132.46 2304.17 2129.43 2304.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="147.478,2001.74 2352.76,2001.74 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="147.478,1647.4 2352.76,1647.4 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="147.478,1293.05 2352.76,1293.05 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="147.478,938.704 2352.76,938.704 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="147.478,584.358 2352.76,584.358 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="147.478,230.012 2352.76,230.012 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,2261.17 147.478,47.2441 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,2001.74 166.376,2001.74 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,1647.4 166.376,1647.4 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,1293.05 166.376,1293.05 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,938.704 166.376,938.704 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,584.358 166.376,584.358 "/>
+<polyline clip-path="url(#clip610)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="147.478,230.012 166.376,230.012 "/>
+<path clip-path="url(#clip610)" d="M51.3625 2002.19 L81.0383 2002.19 L81.0383 2006.13 L51.3625 2006.13 L51.3625 2002.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M95.1586 2015.09 L111.478 2015.09 L111.478 2019.02 L89.5336 2019.02 L89.5336 2015.09 Q92.1956 2012.33 96.7789 2007.7 Q101.385 2003.05 102.566 2001.71 Q104.811 1999.18 105.691 1997.45 Q106.594 1995.69 106.594 1994 Q106.594 1991.25 104.649 1989.51 Q102.728 1987.77 99.6261 1987.77 Q97.4271 1987.77 94.9734 1988.54 Q92.5428 1989.3 89.7651 1990.85 L89.7651 1986.13 Q92.5891 1985 95.0428 1984.42 Q97.4965 1983.84 99.5335 1983.84 Q104.904 1983.84 108.098 1986.52 Q111.293 1989.21 111.293 1993.7 Q111.293 1995.83 110.483 1997.75 Q109.696 1999.65 107.589 2002.24 Q107.01 2002.91 103.909 2006.13 Q100.807 2009.32 95.1586 2015.09 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M50.9921 1647.85 L80.6679 1647.85 L80.6679 1651.78 L50.9921 1651.78 L50.9921 1647.85 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M91.5706 1660.74 L99.2095 1660.74 L99.2095 1634.38 L90.8993 1636.04 L90.8993 1631.78 L99.1632 1630.12 L103.839 1630.12 L103.839 1660.74 L111.478 1660.74 L111.478 1664.68 L91.5706 1664.68 L91.5706 1660.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M99.5335 1278.85 Q95.9224 1278.85 94.0937 1282.41 Q92.2882 1285.96 92.2882 1293.09 Q92.2882 1300.19 94.0937 1303.76 Q95.9224 1307.3 99.5335 1307.3 Q103.168 1307.3 104.973 1303.76 Q106.802 1300.19 106.802 1293.09 Q106.802 1285.96 104.973 1282.41 Q103.168 1278.85 99.5335 1278.85 M99.5335 1275.15 Q105.344 1275.15 108.399 1279.75 Q111.478 1284.34 111.478 1293.09 Q111.478 1301.81 108.399 1306.42 Q105.344 1311 99.5335 1311 Q93.7234 1311 90.6447 1306.42 Q87.5892 1301.81 87.5892 1293.09 Q87.5892 1284.34 90.6447 1279.75 Q93.7234 1275.15 99.5335 1275.15 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M91.5706 952.049 L99.2095 952.049 L99.2095 925.683 L90.8993 927.35 L90.8993 923.091 L99.1632 921.424 L103.839 921.424 L103.839 952.049 L111.478 952.049 L111.478 955.984 L91.5706 955.984 L91.5706 952.049 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M95.1586 597.703 L111.478 597.703 L111.478 601.638 L89.5336 601.638 L89.5336 597.703 Q92.1956 594.948 96.7789 590.319 Q101.385 585.666 102.566 584.323 Q104.811 581.8 105.691 580.064 Q106.594 578.305 106.594 576.615 Q106.594 573.86 104.649 572.124 Q102.728 570.388 99.6261 570.388 Q97.4271 570.388 94.9734 571.152 Q92.5428 571.916 89.7651 573.467 L89.7651 568.745 Q92.5891 567.61 95.0428 567.032 Q97.4965 566.453 99.5335 566.453 Q104.904 566.453 108.098 569.138 Q111.293 571.823 111.293 576.314 Q111.293 578.444 110.483 580.365 Q109.696 582.263 107.589 584.856 Q107.01 585.527 103.909 588.745 Q100.807 591.939 95.1586 597.703 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip610)" d="M104.348 228.658 Q107.705 229.375 109.58 231.644 Q111.478 233.912 111.478 237.246 Q111.478 242.361 107.959 245.162 Q104.441 247.963 97.9595 247.963 Q95.7836 247.963 93.4688 247.523 Q91.1771 247.107 88.7234 246.25 L88.7234 241.736 Q90.6678 242.87 92.9826 243.449 Q95.2974 244.028 97.8206 244.028 Q102.219 244.028 104.51 242.292 Q106.825 240.556 106.825 237.246 Q106.825 234.19 104.672 232.477 Q102.543 230.741 98.7234 230.741 L94.6956 230.741 L94.6956 226.898 L98.9085 226.898 Q102.358 226.898 104.186 225.533 Q106.015 224.144 106.015 221.551 Q106.015 218.889 104.117 217.477 Q102.242 216.042 98.7234 216.042 Q96.8021 216.042 94.603 216.459 Q92.4039 216.875 89.7651 217.755 L89.7651 213.588 Q92.4271 212.847 94.7419 212.477 Q97.0798 212.107 99.14 212.107 Q104.464 212.107 107.566 214.537 Q110.668 216.945 110.668 221.065 Q110.668 223.935 109.024 225.926 Q107.381 227.894 104.348 228.658 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip612)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="672.666,1656.45 677.247,1639.43 681.263,1622.63 684.705,1606.1 687.564,1589.89 689.833,1574.02 691.507,1558.55 692.58,1543.51 693.051,1528.93 692.917,1514.86 692.181,1501.32 690.843,1488.36 688.906,1476.01 686.376,1464.29 683.259,1453.23 679.563,1442.87 675.296,1433.23 670.47,1424.33 665.097,1416.19 659.19,1408.84 652.763,1402.3 645.834,1396.57 638.419,1391.68 630.536,1387.64 622.206,1384.45 613.449,1382.13 604.288,1380.67 594.744,1380.1 584.843,1380.39 574.608,1381.57 564.065,1383.61 553.241,1386.52 542.162,1390.29 530.857,1394.92 519.354,1400.38 507.681,1406.66 495.867,1413.76 483.942,1421.64 471.937,1430.3 459.88,1439.71 447.802,1449.84 435.733,1460.68 423.704,1472.19 411.744,1484.35 399.884,1497.12 388.152,1510.47 376.579,1524.38 365.193,1538.8 354.023,1553.7 343.096,1569.04 332.44,1584.78 322.082,1600.89 312.047,1617.32 302.36,1634.03 293.047,1650.99 284.129,1668.14 275.63,1685.45 267.57,1702.87 259.97,1720.37 252.848,1737.88 246.223,1755.38 240.111,1772.82 234.527,1790.14 229.486,1807.33 225,1824.31 221.079,1841.07 217.735,1857.54 214.974,1873.7 212.805,1889.5 211.233,1904.91 210.261,1919.88 209.891,1934.37 210.126,1948.36 210.964,1961.8 212.403,1974.66 214.44,1986.91 217.069,1998.52 220.284,2009.46 224.077,2019.7 228.438,2029.22 233.357,2037.99 238.821,2046 244.817,2053.21 251.329,2059.62 258.341,2065.21 265.836,2069.96 273.796,2073.86 282.199,2076.9 291.025,2079.08 300.253,2080.38 309.858,2080.82 319.818,2080.37 330.106,2079.05 340.698,2076.86 351.567,2073.8 362.685,2069.89 374.026,2065.12 385.56,2059.52 397.259,2053.1 409.093,2045.87 421.033,2037.85 433.05,2029.07 445.112,2019.54 457.191,2009.28 469.255,1998.33 481.274,1986.71 493.22,1974.45 505.061,1961.57 516.767,1948.13 528.311,1934.13 539.663,1919.63 550.794,1904.65 561.678,1889.24 572.285,1873.44 582.591,1857.27 592.569,1840.79 602.194,1824.03 611.443,1807.04 620.292,1789.86 628.719,1772.52 636.703,1755.09 644.224,1737.59 651.263,1720.07 657.803,1702.58 663.827,1685.16 669.32,1667.85 674.269,1650.7 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2269.29,957.088 2277.2,946.119 2283.28,935.513 2287.5,925.297 2289.85,915.496 2290.34,906.136 2288.96,897.238 2285.71,888.826 2280.6,880.921 2273.64,873.542 2264.85,866.708 2254.26,860.436 2241.89,854.741 2227.76,849.638 2211.92,845.14 2194.41,841.257 2175.26,838 2154.53,835.376 2132.27,833.393 2108.54,832.054 2083.38,831.365 2056.88,831.325 2029.08,831.935 2000.07,833.195 1969.91,835.1 1938.69,837.646 1906.47,840.826 1873.34,844.633 1839.38,849.057 1804.67,854.087 1769.31,859.71 1733.38,865.913 1696.97,872.679 1660.18,879.993 1623.08,887.835 1585.79,896.186 1548.38,905.025 1510.96,914.331 1473.61,924.079 1436.44,934.246 1399.53,944.805 1362.97,955.732 1326.86,966.997 1291.29,978.574 1256.34,990.433 1222.11,1002.55 1188.68,1014.88 1156.13,1027.4 1124.54,1040.09 1094,1052.9 1064.57,1065.82 1036.34,1078.79 1009.37,1091.8 983.731,1104.8 959.488,1117.78 936.7,1130.68 915.424,1143.49 895.714,1156.17 877.619,1168.69 861.184,1181.01 846.45,1193.11 833.454,1204.96 822.228,1216.52 812.801,1227.77 805.196,1238.68 799.432,1249.23 795.523,1259.37 793.48,1269.1 793.307,1278.39 795.005,1287.2 798.569,1295.53 803.991,1303.35 811.257,1310.64 820.349,1317.38 831.244,1323.56 843.915,1329.15 858.33,1334.15 874.454,1338.55 892.245,1342.33 911.66,1345.48 932.65,1348 955.163,1349.87 979.142,1351.1 1004.53,1351.68 1031.25,1351.61 1059.26,1350.89 1088.47,1349.52 1118.81,1347.51 1150.21,1344.86 1182.59,1341.57 1215.87,1337.66 1249.96,1333.13 1284.78,1328 1320.24,1322.28 1356.26,1315.98 1392.74,1309.12 1429.59,1301.72 1466.72,1293.79 1504.05,1285.35 1541.46,1276.43 1578.88,1267.05 1616.2,1257.23 1653.34,1247 1690.19,1236.38 1726.68,1225.39 1762.71,1214.07 1798.18,1202.44 1833.01,1190.54 1867.12,1178.39 1900.41,1166.02 1932.8,1153.47 1964.22,1140.76 1994.58,1127.92 2023.81,1115 2051.83,1102.02 2078.58,1089.01 2103.99,1076.01 2127.99,1063.04 2150.53,1050.15 2171.55,1037.36 2190.99,1024.71 2208.81,1012.22 2224.96,999.929 2239.41,987.869 2252.11,976.068 2263.03,964.555 2272.15,953.36 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1706.28,1860.86 1698.38,1839.06 1689.05,1817.03 1678.32,1794.85 1666.21,1772.56 1652.75,1750.21 1637.99,1727.87 1621.94,1705.58 1604.67,1683.41 1586.2,1661.41 1566.58,1639.64 1545.87,1618.14 1524.12,1596.98 1501.37,1576.21 1477.69,1555.87 1453.13,1536.02 1427.77,1516.71 1401.65,1497.99 1374.85,1479.9 1347.43,1462.49 1319.46,1445.8 1291.01,1429.87 1262.15,1414.75 1232.96,1400.47 1203.51,1387.07 1173.87,1374.58 1144.11,1363.03 1114.31,1352.45 1084.54,1342.87 1054.89,1334.31 1025.41,1326.79 996.192,1320.33 967.302,1314.95 938.813,1310.67 910.797,1307.48 883.323,1305.4 856.46,1304.44 830.276,1304.6 804.835,1305.87 780.201,1308.26 756.437,1311.75 733.6,1316.34 711.749,1322.03 690.938,1328.78 671.219,1336.59 652.641,1345.44 635.25,1355.3 619.091,1366.15 604.204,1377.97 590.625,1390.72 578.389,1404.37 567.526,1418.89 558.063,1434.24 550.025,1450.38 543.431,1467.27 538.297,1484.88 534.637,1503.15 532.46,1522.04 531.771,1541.51 532.571,1561.5 534.859,1581.96 538.63,1602.85 543.873,1624.11 550.575,1645.69 558.721,1667.53 568.289,1689.59 579.255,1711.79 591.592,1734.1 605.27,1756.45 620.254,1778.79 636.506,1801.05 653.987,1823.2 672.652,1845.16 692.454,1866.9 713.345,1888.34 735.271,1909.44 758.179,1930.14 782.011,1950.4 806.707,1970.16 832.205,1989.38 858.443,2008 885.354,2025.98 912.87,2043.27 940.924,2059.84 969.446,2075.63 998.363,2090.61 1027.6,2104.75 1057.09,2118 1086.76,2130.33 1116.53,2141.72 1146.33,2152.13 1176.08,2161.55 1205.72,2169.93 1235.15,2177.27 1264.32,2183.55 1293.15,2188.74 1321.56,2192.85 1349.49,2195.85 1376.87,2197.74 1403.62,2198.51 1429.69,2198.17 1455,2196.71 1479.49,2194.13 1503.1,2190.45 1525.77,2185.68 1547.45,2179.81 1568.09,2172.88 1587.62,2164.89 1606,2155.87 1623.18,2145.84 1639.13,2134.83 1653.8,2122.85 1667.16,2109.95 1679.16,2096.15 1689.79,2081.49 1699.02,2066 1706.82,2049.73 1713.16,2032.71 1718.05,2014.99 1721.46,1996.61 1723.39,1977.62 1723.83,1958.07 1722.78,1937.99 1720.24,1917.45 1716.22,1896.5 1710.73,1875.18 1703.79,1853.55 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="695.032,1458.68 697.314,1453.52 699.149,1448.47 700.533,1443.54 701.463,1438.75 701.935,1434.1 701.95,1429.61 701.506,1425.28 700.605,1421.13 699.249,1417.18 697.442,1413.42 695.188,1409.87 692.493,1406.54 689.363,1403.43 685.806,1400.56 681.832,1397.92 677.449,1395.54 672.67,1393.41 667.506,1391.53 661.969,1389.92 656.074,1388.58 649.836,1387.5 643.27,1386.7 636.392,1386.18 629.22,1385.93 621.772,1385.96 614.066,1386.26 606.121,1386.85 597.957,1387.7 589.596,1388.83 581.057,1390.23 572.361,1391.9 563.532,1393.83 554.591,1396.01 545.559,1398.45 536.46,1401.13 527.317,1404.05 518.152,1407.21 508.988,1410.59 499.848,1414.18 490.756,1417.98 481.733,1421.98 472.802,1426.16 463.986,1430.52 455.306,1435.05 446.785,1439.73 438.443,1444.55 430.301,1449.5 422.381,1454.58 414.701,1459.75 407.281,1465.02 400.139,1470.37 393.293,1475.79 386.761,1481.26 380.558,1486.76 374.7,1492.3 369.202,1497.84 364.078,1503.38 359.34,1508.91 355,1514.4 351.069,1519.85 347.557,1525.24 344.473,1530.56 341.823,1535.8 339.616,1540.94 337.856,1545.97 336.548,1550.88 335.696,1555.65 335.3,1560.27 335.363,1564.74 335.883,1569.03 336.861,1573.15 338.293,1577.07 340.176,1580.79 342.504,1584.31 345.273,1587.6 348.475,1590.67 352.103,1593.5 356.147,1596.1 360.597,1598.44 365.442,1600.53 370.67,1602.36 376.267,1603.93 382.221,1605.22 388.515,1606.25 395.135,1607.01 402.064,1607.48 409.283,1607.69 416.776,1607.61 424.524,1607.26 432.507,1606.63 440.705,1605.73 449.098,1604.55 457.665,1603.11 466.384,1601.4 475.234,1599.43 484.192,1597.2 493.237,1594.72 502.345,1592 511.493,1589.03 520.659,1585.84 529.821,1582.43 538.954,1578.8 548.037,1574.96 557.046,1570.94 565.959,1566.72 574.754,1562.33 583.408,1557.78 591.901,1553.07 600.211,1548.23 608.316,1543.25 616.198,1538.16 623.835,1532.97 631.21,1527.68 638.303,1522.32 645.097,1516.9 651.576,1511.42 657.721,1505.91 663.52,1500.37 668.956,1494.83 674.016,1489.29 678.688,1483.77 682.96,1478.28 686.821,1472.84 690.262,1467.46 693.273,1462.15 695.849,1456.93 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1958.4,635.235 1961.49,606.133 1962.64,577.258 1961.85,548.683 1959.13,520.477 1954.47,492.713 1947.91,465.459 1939.44,438.783 1929.09,412.753 1916.89,387.433 1902.86,362.886 1887.05,339.174 1869.49,316.356 1850.22,294.49 1829.3,273.629 1806.77,253.826 1782.7,235.131 1757.13,217.59 1730.15,201.246 1701.81,186.142 1672.18,172.314 1641.34,159.798 1609.37,148.624 1576.34,138.82 1542.34,130.412 1507.46,123.419 1471.77,117.859 1435.37,113.747 1398.36,111.093 1360.82,109.902 1322.84,110.179 1284.53,111.922 1245.97,115.128 1207.27,119.788 1168.52,125.89 1129.81,133.419 1091.25,142.357 1052.93,152.68 1014.95,164.364 977.4,177.379 940.374,191.693 903.966,207.269 868.267,224.07 833.365,242.051 799.349,261.17 766.303,281.378 734.31,302.625 703.449,324.857 673.799,348.019 645.433,372.053 618.421,396.899 592.833,422.494 568.73,448.776 546.174,475.678 525.222,503.132 505.925,531.071 488.331,559.425 472.485,588.122 458.427,617.091 446.19,646.259 435.807,675.554 427.303,704.903 420.699,734.231 416.011,763.466 413.252,792.535 412.428,821.364 413.542,849.883 416.59,878.019 421.565,905.702 428.455,932.864 437.242,959.435 447.904,985.351 460.415,1010.55 474.744,1034.96 490.854,1058.52 508.705,1081.18 528.253,1102.88 549.449,1123.57 572.24,1143.19 596.568,1161.7 622.374,1179.04 649.593,1195.18 678.156,1210.07 707.992,1223.68 739.027,1235.97 771.183,1246.92 804.379,1256.49 838.533,1264.66 873.56,1271.41 909.371,1276.73 945.878,1280.6 982.988,1283 1020.61,1283.95 1058.65,1283.42 1097.01,1281.44 1135.6,1277.98 1174.32,1273.08 1213.07,1266.74 1251.75,1258.97 1290.28,1249.8 1328.55,1239.24 1366.46,1227.33 1403.93,1214.1 1440.86,1199.57 1477.16,1183.78 1512.73,1166.78 1547.49,1148.61 1581.35,1129.3 1614.22,1108.92 1646.03,1087.5 1676.69,1065.11 1706.13,1041.79 1734.27,1017.62 1761.05,992.643 1786.39,966.928 1810.24,940.537 1832.53,913.538 1853.21,885.997 1872.22,857.983 1889.53,829.567 1905.07,800.819 1918.83,771.811 1930.75,742.617 1940.82,713.307 1949.01,683.957 1955.29,654.64 1959.66,625.427 "/>
+<polyline clip-path="url(#clip612)" style="stroke:#000080; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1961.32,1321.02 1961.7,1316.86 1961.03,1312.75 1959.29,1308.69 1956.5,1304.7 1952.66,1300.78 1947.79,1296.95 1941.88,1293.22 1934.96,1289.59 1927.05,1286.07 1918.16,1282.67 1908.32,1279.41 1897.55,1276.28 1885.88,1273.3 1873.34,1270.48 1859.95,1267.81 1845.76,1265.32 1830.8,1263 1815.1,1260.86 1798.7,1258.9 1781.66,1257.13 1764,1255.56 1745.77,1254.19 1727.02,1253.02 1707.79,1252.05 1688.14,1251.3 1668.11,1250.75 1647.75,1250.41 1627.12,1250.28 1606.25,1250.37 1585.22,1250.66 1564.06,1251.17 1542.84,1251.89 1521.6,1252.81 1500.4,1253.94 1479.29,1255.27 1458.33,1256.8 1437.56,1258.53 1417.04,1260.45 1396.82,1262.55 1376.95,1264.84 1357.48,1267.3 1338.46,1269.93 1319.93,1272.72 1301.94,1275.67 1284.55,1278.77 1267.78,1282.01 1251.69,1285.38 1236.31,1288.87 1221.68,1292.48 1207.83,1296.2 1194.81,1300.01 1182.65,1303.91 1171.37,1307.89 1161.01,1311.93 1151.58,1316.04 1143.12,1320.19 1135.64,1324.38 1129.16,1328.59 1123.7,1332.82 1119.28,1337.06 1115.9,1341.29 1113.57,1345.5 1112.3,1349.69 1112.09,1353.84 1112.95,1357.95 1114.86,1362 1117.83,1365.98 1121.84,1369.88 1126.89,1373.7 1132.97,1377.41 1140.06,1381.03 1148.13,1384.53 1157.18,1387.9 1167.18,1391.14 1178.11,1394.24 1189.93,1397.2 1202.61,1400 1216.14,1402.63 1230.46,1405.1 1245.55,1407.39 1261.37,1409.5 1277.87,1411.42 1295.03,1413.16 1312.79,1414.7 1331.11,1416.03 1349.94,1417.17 1369.24,1418.1 1388.96,1418.82 1409.05,1419.34 1429.46,1419.64 1450.13,1419.73 1471.03,1419.61 1492.09,1419.28 1513.26,1418.74 1534.49,1417.99 1555.72,1417.03 1576.91,1415.86 1598,1414.5 1618.94,1412.93 1639.67,1411.17 1660.14,1409.22 1680.3,1407.09 1700.11,1404.77 1719.51,1402.28 1738.45,1399.62 1756.89,1396.8 1774.78,1393.83 1792.07,1390.71 1808.73,1387.45 1824.71,1384.06 1839.97,1380.54 1854.46,1376.91 1868.17,1373.18 1881.05,1369.35 1893.07,1365.44 1904.19,1361.45 1914.4,1357.39 1923.67,1353.28 1931.97,1349.12 1939.28,1344.93 1945.58,1340.71 1950.87,1336.48 1955.12,1332.24 1958.32,1328.01 1960.47,1323.8 1961.56,1319.62 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/conic/min_ellipse/index.html b/previews/PR3525/tutorials/conic/min_ellipse/index.html
new file mode 100644
index 00000000000..921052baf53
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/min_ellipse/index.html
@@ -0,0 +1,95 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Minimal ellipses · JuMP</title><meta name="title" content="Minimal ellipses · JuMP"/><meta property="og:title" content="Minimal ellipses · JuMP"/><meta property="twitter:title" content="Minimal ellipses · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li class="is-active"><a class="tocitem" href>Minimal ellipses</a><ul class="internal"><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#Build-the-model"><span>Build the model</span></a></li><li><a class="tocitem" href="#Results"><span>Results</span></a></li></ul></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Minimal ellipses</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Minimal ellipses</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/min_ellipse.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Minimal-ellipses"><a class="docs-heading-anchor" href="#Minimal-ellipses">Minimal ellipses</a><a id="Minimal-ellipses-1"></a><a class="docs-heading-anchor-permalink" href="#Minimal-ellipses" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../min_ellipse.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This example comes from section 8.4.1 of the book <em>Convex Optimization</em> by <a href="https://web.stanford.edu/~boyd/cvxbook/">Boyd and Vandenberghe (2004)</a>.</p><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>Given a set of <span>$m$</span> ellipses of the form:</p><p class="math-container">\[E(A, b, c) = \{ x : x^\top A x + 2 b^\top x + c \leq 0 \},\]</p><p>the minimal ellipse problem finds an ellipse with the minimum area that encloses the given ellipses.</p><p>It is convenient to parameterize the minimal enclosing ellipse as</p><p class="math-container">\[\{ x : || Px + q || \leq 1 \}.\]</p><p>Then the optimal <span>$P$</span> and <span>$q$</span> are given by the convex semidefinite program;</p><p class="math-container">\[\begin{aligned}
+\text{maximize } &amp; \quad \log(\det(P))  \\
+\text{subject to } &amp; \quad \tau_i \geq 0, &amp; i = 1, \ldots, m \\
+&amp; \quad\begin{bmatrix}
+    P^2 - \tau_i A_i        &amp;  P q - \tau_i b_i   &amp;    0        \\
+    (P q - \tau_i b_i)^\top &amp;   -1 - \tau_i c_i   &amp;  (P q)^\top \\
+                0           &amp;          (P q)      &amp; - P^2       \\
+\end{bmatrix}
+\preceq 0 \text{ (PSD) } &amp;  i=1, \ldots, m
+\end{aligned}\]</p><p>with helper variables <span>$\tau$</span>.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import LinearAlgebra
+import Plots
+import SCS</code></pre><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>First, define the <span>$m$</span> input ellipses (here <span>$m = 6$</span>), parameterized as <span>$x^T A_i x + 2 b_i^T x + c \leq 0$</span>:</p><pre><code class="language-julia hljs">struct Ellipse
+    A::Matrix{Float64}
+    b::Vector{Float64}
+    c::Float64
+    function Ellipse(A::Matrix{Float64}, b::Vector{Float64}, c::Float64)
+        @assert isreal(A) &amp;&amp; LinearAlgebra.issymmetric(A)
+        return new(A, b, c)
+    end
+end
+
+ellipses = [
+    Ellipse([1.2576 -0.3873; -0.3873 0.3467], [0.2722, 0.1969], 0.1831),
+    Ellipse([1.4125 -2.1777; -2.1777 6.7775], [-1.228, -0.0521], 0.3295),
+    Ellipse([1.7018 0.8141; 0.8141 1.7538], [-0.4049, 1.5713], 0.2077),
+    Ellipse([0.9742 -0.7202; -0.7202 1.5444], [0.0265, 0.5623], 0.2362),
+    Ellipse([0.6798 -0.1424; -0.1424 0.6871], [-0.4301, -1.0157], 0.3284),
+    Ellipse([0.1796 -0.1423; -0.1423 2.6181], [-0.3286, 0.557], 0.4931),
+];</code></pre><p>We visualise the ellipses using the Plots package:</p><pre><code class="language-julia hljs">function plot_ellipse(plot, ellipse::Ellipse)
+    A, b, c = ellipse.A, ellipse.b, ellipse.c
+    θ = range(0, 2pi + 0.05; step = 0.05)
+    # Some linear algebra to convert θ into (x,y) coordinates.
+    x_y = √A \ (√(b&#39; * (A \ b) - c) .* hcat(cos.(θ), sin.(θ)) .- (√A \ b)&#39;)&#39;
+    Plots.plot!(plot, x_y[1, :], x_y[2, :]; label = nothing, c = :navy)
+    return
+end
+
+plot = Plots.plot(; size = (600, 600))
+for ellipse in ellipses
+    plot_ellipse(plot, ellipse)
+end
+plot</code></pre><img src="804a7acd.svg" alt="Example block output"/><h2 id="Build-the-model"><a class="docs-heading-anchor" href="#Build-the-model">Build the model</a><a id="Build-the-model-1"></a><a class="docs-heading-anchor-permalink" href="#Build-the-model" title="Permalink"></a></h2><p>Now let&#39;s build the model, using the change-of-variables <code>P²</code> = <span>$P^2$</span> and <code>P_q</code> = <span>$P q$</span>. We&#39;ll recover the true value of <code>P</code> and <code>q</code> after the solve.</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+# We need to use a tighter tolerance for this example, otherwise the bounding
+# ellipse won&#39;t actually be bounding...
+set_attribute(model, &quot;eps_rel&quot;, 1e-6)
+set_silent(model)
+m, n = length(ellipses), size(first(ellipses).A, 1)
+@variable(model, τ[1:m] &gt;= 0)
+@variable(model, P²[1:n, 1:n], PSD)
+@variable(model, P_q[1:n])
+
+for (i, ellipse) in enumerate(ellipses)
+    A, b, c = ellipse.A, ellipse.b, ellipse.c
+    X = [
+        #! format: off
+        (P² - τ[i] * A)   (P_q - τ[i] * b) zeros(n, n)
+        (P_q - τ[i] * b)&#39; (-1 - τ[i] * c)  P_q&#39;
+        zeros(n, n)       P_q              -P²
+        #! format: on
+    ]
+    @constraint(model, LinearAlgebra.Symmetric(X) &lt;= 0, PSDCone())
+end</code></pre><p>We cannot directly represent the objective <span>$\log(\det(P))$</span>, so we introduce the conic reformulation:</p><pre><code class="language-julia hljs">@variable(model, log_det_P)
+@constraint(model, [log_det_P; 1; vec(P²)] in MOI.LogDetConeSquare(n))
+@objective(model, Max, log_det_P)</code></pre><p>$ log_det_P $</p><p>Now, solve the program:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : SCS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;solved&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : -4.04358e+00
+  Dual objective value : -4.04364e+00
+
+* Work counters
+  Solve time (sec)   : 2.01622e-01
+</code></pre><h2 id="Results"><a class="docs-heading-anchor" href="#Results">Results</a><a id="Results-1"></a><a class="docs-heading-anchor-permalink" href="#Results" title="Permalink"></a></h2><p>After solving the model to optimality we can recover the solution in terms of <span>$P$</span> and <span>$q$</span>:</p><pre><code class="language-julia hljs">P = sqrt(value.(P²))
+q = P \ value.(P_q)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ -0.39642176943764484
+ -0.021394178202894645</code></pre><p>Finally, overlaying the solution in the plot we see the minimal area enclosing ellipsoid:</p><pre><code class="language-julia hljs">Plots.plot!(
+    plot,
+    [tuple(P \ [cos(θ) - q[1], sin(θ) - q[2]]...) for θ in 0:0.05:(2pi+0.05)];
+    c = :crimson,
+    label = nothing,
+)</code></pre><img src="4146ff43.svg" alt="Example block output"/></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../experiment_design/">« Experiment design</a><a class="docs-footer-nextpage" href="../ellipse_approx/">Ellipsoid approximation »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/quantum_discrimination.jl b/previews/PR3525/tutorials/conic/quantum_discrimination.jl
new file mode 100644
index 00000000000..d743cc0b005
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/quantum_discrimination.jl
@@ -0,0 +1,147 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Quantum state discrimination
+
+# This tutorial solves the problem of [quantum state discrimination](https://en.wikipedia.org/wiki/Quantum_state_discrimination).
+
+# The purpose of this tutorial to demonstrate how to solve problems involving
+# complex-valued decision variables and the [`HermitianPSDCone`](@ref). See
+# [Complex number support](@ref) for more details.
+
+# ## Required packages
+
+# This tutorial makes use of the following packages:
+
+using JuMP
+import LinearAlgebra
+import SCS
+
+# ## Formulation
+
+# A `d`-dimensional quantum state, ``\rho``, can be defined by a complex-valued
+# Hermitian matrix with a trace of `1`. Assume we have `N` `d`-dimensional
+# quantum states, ``\{\rho_i\}_{i=1}^N``, each of which is equally likely.
+
+# The goal of the quantum state discrimination problem is to choose a
+# positive operator-valued measure ([POVM](https://en.wikipedia.org/wiki/POVM)), ``\{ E_i \}_{i=1}^N``, such that if we observe
+# ``E_i`` then the most probable state that we are in is ``\rho_i``.
+
+# Each POVM element, ``E_i``, is a complex-valued Hermitian matrix, and there is a
+# requirement that ``\sum\limits_{i=1}^N E_i = \mathbf{I}``.
+
+# To choose a POVM, we want to maximize the probability that we guess
+# the quantum state correctly. This can be formulated as the following
+# optimization problem:
+
+# ```math
+# \begin{aligned}
+# \max\limits_{E} \;\; & \frac{1}{N} \sum\limits_{i=1}^N \operatorname{tr}(\rho_i  E_i) \\
+# \text{s.t.}     \;\; & \sum\limits_{i=1}^N E_i = \mathbf{I} \\
+#                      & E_i \succeq 0 \; \forall i = 1,\ldots,N.
+# \end{aligned}
+# ```
+
+# ## Data
+
+# To setup our problem, we need `N` `d-`dimensional quantum states. To keep the
+# problem simple, we use `N = 2` and `d = 2`.
+
+N, d = 2, 2
+
+# We then generated `N` random `d`-dimensional quantum states:
+
+function random_state(d)
+    x = randn(ComplexF64, (d, d))
+    y = x * x'
+    return LinearAlgebra.Hermitian(y / LinearAlgebra.tr(y))
+end
+
+ρ = [random_state(d) for i in 1:N]
+
+# ## JuMP formulation
+
+# To model the problem in JuMP, we need a solver that supports positive
+# semidefinite matrices:
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+
+# Then, we construct our set of `E` variables:
+
+E = [@variable(model, [1:d, 1:d] in HermitianPSDCone()) for i in 1:N]
+
+# Here we have created a vector of matrices. This is different to other modeling
+# languages such as YALMIP, which allow you to create a multi-dimensional array
+# in which 2-dimensional slices of the array are Hermitian matrices.
+
+# We also need to enforce the constraint that
+# ``\sum\limits_{i=1}^N E_i = \mathbf{I}``:
+
+@constraint(model, sum(E) == LinearAlgebra.I)
+
+# This constraint is a complex-valued equality constraint. In the solver, it
+# will be decomposed onto two types of equality constraints: one to enforce
+# equality of the real components, and one to enforce equality of the imaginary
+# components.
+
+# Our objective is to maximize the expected probability of guessing correctly:
+
+@objective(
+    model,
+    Max,
+    sum(real(LinearAlgebra.tr(ρ[i] * E[i])) for i in 1:N) / N,
+)
+
+# Now we optimize:
+
+optimize!(model)
+solution_summary(model)
+
+# The probability of guessing correctly is:
+
+objective_value(model)
+
+# When `N = 2`, there is a known analytical solution of:
+
+0.5 + 0.25 * sum(LinearAlgebra.svdvals(ρ[1] - ρ[2]))
+
+# proving that we found the optimal solution.
+
+# Finally, the optimal POVM is:
+
+solution = [value.(e) for e in E]
+
+# !!! tip
+#     Duality plays a large role in solving conic optimization models. Depending
+#     on the solver, it can be more efficient to solve the dual of this problem
+#     instead of the primal. If performance is an issue, see the [Dualization](@ref)
+#     tutorial for more details.
+
+# ## Alternative formulation
+
+# The formulation above includes `N` Hermitian matrices and a set of linear
+# equality constraints. We can simplify the problem by replacing ``E_N`` with
+# ``E_N = I - \sum\limits_{i=1}^{N-1} E_i``. This results in:
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+E = [@variable(model, [1:d, 1:d] in HermitianPSDCone()) for i in 1:N-1]
+E_N = LinearAlgebra.Hermitian(LinearAlgebra.I - sum(E))
+@constraint(model, E_N in HermitianPSDCone())
+push!(E, E_N)
+
+# The objective can also be simplified, by observing that it is equivalent to:
+
+@objective(model, Max, real(LinearAlgebra.dot(ρ, E)) / N)
+
+# Then we can check that we get the same solution:
+
+optimize!(model)
+solution_summary(model)
+
+#-
+
+objective_value(model)
diff --git a/previews/PR3525/tutorials/conic/quantum_discrimination/index.html b/previews/PR3525/tutorials/conic/quantum_discrimination/index.html
new file mode 100644
index 00000000000..36e50cba6dd
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/quantum_discrimination/index.html
@@ -0,0 +1,73 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Quantum state discrimination · JuMP</title><meta name="title" content="Quantum state discrimination · JuMP"/><meta property="og:title" content="Quantum state discrimination · JuMP"/><meta property="twitter:title" content="Quantum state discrimination · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li class="is-active"><a class="tocitem" href>Quantum state discrimination</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Alternative-formulation"><span>Alternative formulation</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Quantum state discrimination</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Quantum state discrimination</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/quantum_discrimination.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Quantum-state-discrimination"><a class="docs-heading-anchor" href="#Quantum-state-discrimination">Quantum state discrimination</a><a id="Quantum-state-discrimination-1"></a><a class="docs-heading-anchor-permalink" href="#Quantum-state-discrimination" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../quantum_discrimination.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial solves the problem of <a href="https://en.wikipedia.org/wiki/Quantum_state_discrimination">quantum state discrimination</a>.</p><p>The purpose of this tutorial to demonstrate how to solve problems involving complex-valued decision variables and the <a href="../../../api/JuMP/#JuMP.HermitianPSDCone"><code>HermitianPSDCone</code></a>. See <a href="../../../manual/complex/#Complex-number-support">Complex number support</a> for more details.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial makes use of the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import LinearAlgebra
+import SCS</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>A <code>d</code>-dimensional quantum state, <span>$\rho$</span>, can be defined by a complex-valued Hermitian matrix with a trace of <code>1</code>. Assume we have <code>N</code> <code>d</code>-dimensional quantum states, <span>$\{\rho_i\}_{i=1}^N$</span>, each of which is equally likely.</p><p>The goal of the quantum state discrimination problem is to choose a positive operator-valued measure (<a href="https://en.wikipedia.org/wiki/POVM">POVM</a>), <span>$\{ E_i \}_{i=1}^N$</span>, such that if we observe <span>$E_i$</span> then the most probable state that we are in is <span>$\rho_i$</span>.</p><p>Each POVM element, <span>$E_i$</span>, is a complex-valued Hermitian matrix, and there is a requirement that <span>$\sum\limits_{i=1}^N E_i = \mathbf{I}$</span>.</p><p>To choose a POVM, we want to maximize the probability that we guess the quantum state correctly. This can be formulated as the following optimization problem:</p><p class="math-container">\[\begin{aligned}
+\max\limits_{E} \;\; &amp; \frac{1}{N} \sum\limits_{i=1}^N \operatorname{tr}(\rho_i  E_i) \\
+\text{s.t.}     \;\; &amp; \sum\limits_{i=1}^N E_i = \mathbf{I} \\
+                     &amp; E_i \succeq 0 \; \forall i = 1,\ldots,N.
+\end{aligned}\]</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>To setup our problem, we need <code>N</code> <code>d-</code>dimensional quantum states. To keep the problem simple, we use <code>N = 2</code> and <code>d = 2</code>.</p><pre><code class="language-julia hljs">N, d = 2, 2</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(2, 2)</code></pre><p>We then generated <code>N</code> random <code>d</code>-dimensional quantum states:</p><pre><code class="language-julia hljs">function random_state(d)
+    x = randn(ComplexF64, (d, d))
+    y = x * x&#39;
+    return LinearAlgebra.Hermitian(y / LinearAlgebra.tr(y))
+end
+
+ρ = [random_state(d) for i in 1:N]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{LinearAlgebra.Hermitian{ComplexF64, Matrix{ComplexF64}}}:
+ [0.9049983143615443 + 0.0im -0.11984167294191471 + 0.2224268161763913im; -0.11984167294191471 - 0.2224268161763913im 0.09500168563845571 + 0.0im]
+ [0.2503060704439256 + 0.0im -0.16997258971668044 - 0.09226624832975384im; -0.16997258971668044 + 0.09226624832975384im 0.7496939295560743 + 0.0im]</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>To model the problem in JuMP, we need a solver that supports positive semidefinite matrices:</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)</code></pre><p>Then, we construct our set of <code>E</code> variables:</p><pre><code class="language-julia hljs">E = [@variable(model, [1:d, 1:d] in HermitianPSDCone()) for i in 1:N]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}}:
+ [_[1] _[2] + _[4] im; _[2] - _[4] im _[3]]
+ [_[5] _[6] + _[8] im; _[6] - _[8] im _[7]]</code></pre><p>Here we have created a vector of matrices. This is different to other modeling languages such as YALMIP, which allow you to create a multi-dimensional array in which 2-dimensional slices of the array are Hermitian matrices.</p><p>We also need to enforce the constraint that <span>$\sum\limits_{i=1}^N E_i = \mathbf{I}$</span>:</p><pre><code class="language-julia hljs">@constraint(model, sum(E) == LinearAlgebra.I)</code></pre><p class="math-container">\[ \begin{bmatrix}
+{\_}_{1} + {\_}_{5} - 1 &amp; {\_}_{2} + {\_}_{6} + {\_}_{4} im + {\_}_{8} im\\
+{\_}_{2} + {\_}_{6} - {\_}_{4} im - {\_}_{8} im &amp; {\_}_{3} + {\_}_{7} - 1\\
+\end{bmatrix} \in \text{Zeros()} \]</p><p>This constraint is a complex-valued equality constraint. In the solver, it will be decomposed onto two types of equality constraints: one to enforce equality of the real components, and one to enforce equality of the imaginary components.</p><p>Our objective is to maximize the expected probability of guessing correctly:</p><pre><code class="language-julia hljs">@objective(
+    model,
+    Max,
+    sum(real(LinearAlgebra.tr(ρ[i] * E[i])) for i in 1:N) / N,
+)</code></pre><p>$ 0.45249915718077216 {_}<em>{1} - 0.11984167294191471 {\</em>}<em>{2} + 0.2224268161763913 {\</em>}<em>{4} + 0.047500842819227854 {\</em>}<em>{3} + 0.1251530352219628 {\</em>}<em>{5} - 0.16997258971668044 {\</em>}<em>{6} - 0.09226624832975384 {\</em>}<em>{8} + 0.37484696477803714 {\</em>}_{7} $</p><p>Now we optimize:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : SCS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;solved&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 8.64061e-01
+  Dual objective value : 8.64062e-01
+
+* Work counters
+  Solve time (sec)   : 3.78496e-04
+</code></pre><p>The probability of guessing correctly is:</p><pre><code class="language-julia hljs">objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.8640614507314223</code></pre><p>When <code>N = 2</code>, there is a known analytical solution of:</p><pre><code class="language-julia hljs">0.5 + 0.25 * sum(LinearAlgebra.svdvals(ρ[1] - ρ[2]))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.8640627582954737</code></pre><p>proving that we found the optimal solution.</p><p>Finally, the optimal POVM is:</p><pre><code class="language-julia hljs">solution = [value.(e) for e in E]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Matrix{ComplexF64}}:
+ [0.9495721399750018 - 0.0im 0.03442451603977107 + 0.216097313711905im; 0.03442451603977107 - 0.216097313711905im 0.050427855129854156 - 0.0im]
+ [0.050427855176020456 - 0.0im -0.034424516053125256 - 0.21609731370614849im; -0.034424516053125256 + 0.21609731370614849im 0.9495721400119368 - 0.0im]</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Duality plays a large role in solving conic optimization models. Depending on the solver, it can be more efficient to solve the dual of this problem instead of the primal. If performance is an issue, see the <a href="../dualization/#Dualization">Dualization</a> tutorial for more details.</p></div></div><h2 id="Alternative-formulation"><a class="docs-heading-anchor" href="#Alternative-formulation">Alternative formulation</a><a id="Alternative-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Alternative-formulation" title="Permalink"></a></h2><p>The formulation above includes <code>N</code> Hermitian matrices and a set of linear equality constraints. We can simplify the problem by replacing <span>$E_N$</span> with <span>$E_N = I - \sum\limits_{i=1}^{N-1} E_i$</span>. This results in:</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+E = [@variable(model, [1:d, 1:d] in HermitianPSDCone()) for i in 1:N-1]
+E_N = LinearAlgebra.Hermitian(LinearAlgebra.I - sum(E))
+@constraint(model, E_N in HermitianPSDCone())
+push!(E, E_N)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{LinearAlgebra.Hermitian{GenericAffExpr{ComplexF64, VariableRef}, Matrix{GenericAffExpr{ComplexF64, VariableRef}}}}:
+ [_[1] _[2] + _[4] im; _[2] - _[4] im _[3]]
+ [-_[1] + 1 -_[2] - _[4] im; -_[2] + _[4] im -_[3] + 1]</code></pre><p>The objective can also be simplified, by observing that it is equivalent to:</p><pre><code class="language-julia hljs">@objective(model, Max, real(LinearAlgebra.dot(ρ, E)) / N)</code></pre><p>$ 0.32734612195880936 {_}<em>{1} + 0.050130916774765735 {\</em>}<em>{2} + 0.31469306450614515 {\</em>}<em>{4} - 0.3273461219588093 {\</em>}_{3} + 0.49999999999999994 $</p><p>Then we can check that we get the same solution:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : SCS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;solved&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 8.64060e-01
+  Dual objective value : 8.64062e-01
+
+* Work counters
+  Solve time (sec)   : 3.62695e-04
+</code></pre><pre><code class="language-julia hljs">objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.8640596603179971</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../ellipse_approx/">« Ellipsoid approximation</a><a class="docs-footer-nextpage" href="../../algorithms/benders_decomposition/">Benders decomposition »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/simple_examples.jl b/previews/PR3525/tutorials/conic/simple_examples.jl
new file mode 100644
index 00000000000..d3015150d61
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/simple_examples.jl
@@ -0,0 +1,422 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Simple semidefinite programming examples
+
+# This tutorial is a collection of examples of small conic programs from the field of
+# [semidefinite programming](https://en.wikipedia.org/wiki/Semidefinite_programming) (SDP).
+#
+# This tutorial makes use of the following packages:
+
+using JuMP
+import LinearAlgebra
+import Plots
+import Random
+import SCS
+import Test
+
+# ## Maximum cut via SDP
+
+# The [maximum cut problem](https://en.wikipedia.org/wiki/Maximum_cut) is a
+# classical example in graph theory, where we seek to partition a graph into
+# two complementary sets, such that the weight of edges between the two sets is
+# maximized. This problem is NP-hard, but it is possible to obtain an
+# approximate solution using the semidefinite programming relaxation:
+# ```math
+#     \text{max}  \quad   0.25 L•X \\
+#     \text{    s.t.} \quad   \mathrm{diag}(X) = e \\
+#                  \qquad X \succeq 0
+# ```
+# where ``L`` is the weighted graph Laplacian and ``e`` is a vector of ones.
+# For more details, see:
+#
+# Goemans, M. X., & Williamson, D. P. (1995). 
+# [_Improved approximation algorithms for maximum cut and satisfiability problems
+# using semidefinite programming._](https://doi.org/10.1145/227683.227684)
+# Journal of the ACM (JACM), 42(6), 1115-1145.
+
+"""
+    svd_cholesky(X::AbstractMatrix, rtol)
+
+Return the matrix `U` of the Cholesky decomposition of `X` as `U' * U`.
+Note that we do not use the `LinearAlgebra.cholesky` function because it
+requires the matrix to be positive definite while `X` may be only
+positive *semi*definite.
+
+We use the convention `U' * U` instead of `U * U'` to be consistent with
+`LinearAlgebra.cholesky`.
+"""
+function svd_cholesky(X::AbstractMatrix)
+    F = LinearAlgebra.svd(X)
+    ## We now have `X ≈ `F.U * D² * F.U'` where:
+    D = LinearAlgebra.Diagonal(sqrt.(F.S))
+    ## So `X ≈ U' * U` where `U` is:
+    return (F.U * D)'
+end
+
+function solve_max_cut_sdp(weights)
+    N = size(weights, 1)
+    ## Calculate the (weighted) Laplacian of the graph: L = D - W.
+    L = LinearAlgebra.diagm(0 => weights * ones(N)) - weights
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, X[1:N, 1:N], PSD)
+    for i in 1:N
+        set_start_value(X[i, i], 1.0)
+    end
+    @objective(model, Max, 0.25 * LinearAlgebra.dot(L, X))
+    @constraint(model, LinearAlgebra.diag(X) .== 1)
+    optimize!(model)
+    V = svd_cholesky(value(X))
+    Random.seed!(N)
+    r = rand(N)
+    r /= LinearAlgebra.norm(r)
+    cut = [LinearAlgebra.dot(r, V[:, i]) > 0 for i in 1:N]
+    S = findall(cut)
+    T = findall(.!cut)
+    println("Solution:")
+    println(" (S, T) = ({", join(S, ", "), "}, {", join(T, ", "), "})")
+    return S, T
+end
+
+# Given the graph
+# ```raw
+# [1] --- 5 --- [2]
+# ```
+# The solution is `(S, T)  = ({1}, {2})`
+
+S, T = solve_max_cut_sdp([0 5; 5 0])
+
+# Given the graph
+# ```raw
+# [1] --- 5 --- [2]
+#  |  \          |
+#  |    \        |
+#  7      6      1
+#  |        \    |
+#  |          \  |
+# [3] --- 1 --- [4]
+# ```
+# The solution is `(S, T)  = ({1}, {2, 3, 4})`
+
+S, T = solve_max_cut_sdp([0 5 7 6; 5 0 0 1; 7 0 0 1; 6 1 1 0])
+
+# Given the graph
+# ```raw
+# [1] --- 1 --- [2]
+#  |             |
+#  |             |
+#  5             9
+#  |             |
+#  |             |
+# [3] --- 2 --- [4]
+# ```
+# The solution is `(S, T)  = ({1, 4}, {2, 3})`
+
+S, T = solve_max_cut_sdp([0 1 5 0; 1 0 0 9; 5 0 0 2; 0 9 2 0])
+
+# ## K-means clustering via SDP
+
+# Given a set of points ``a_1, \ldots, a_m``  in ``\mathbb{R}^n``, allocate them to ``k`` clusters.
+#
+# For more details, see:
+# 
+# Peng, J., & Wei, Y. (2007).
+# [_Approximating k-means-type clustering via semidefinite programming_](https://doi.org/10.1137/050641983).
+# SIAM Journal on Optimization, 18(1), 186-205.
+
+function example_k_means_clustering()
+    a = [[2.0, 2.0], [2.5, 2.1], [7.0, 7.0], [2.2, 2.3], [6.8, 7.0], [7.2, 7.5]]
+    m = length(a)
+    num_clusters = 2
+    W = zeros(m, m)
+    for i in 1:m, j in i+1:m
+        W[i, j] = W[j, i] = exp(-LinearAlgebra.norm(a[i] - a[j]) / 1.0)
+    end
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, Z[1:m, 1:m] >= 0, PSD)
+    @objective(model, Min, LinearAlgebra.tr(W * (LinearAlgebra.I - Z)))
+    @constraint(model, [i = 1:m], sum(Z[i, :]) .== 1)
+    @constraint(model, LinearAlgebra.tr(Z) == num_clusters)
+    optimize!(model)
+    Z_val = value.(Z)
+    current_cluster, visited = 0, Set{Int}()
+    solution = [1, 1, 2, 1, 2, 2]  #src
+    for i in 1:m
+        if !(i in visited)
+            current_cluster += 1
+            println("Cluster $current_cluster")
+            for j in i:m
+                if isapprox(Z_val[i, i], Z_val[i, j]; atol = 1e-3)
+                    println(a[j])
+                    push!(visited, j)
+                    Test.@test solution[j] == current_cluster  #src
+                end
+            end
+        end
+    end
+    return
+end
+
+example_k_means_clustering()
+
+# ## The correlation problem
+
+# Given three random variables ``A``, ``B``, and ``C``, and given bounds on two of the three
+# correlation coefficients:
+# ```math
+#     -0.2 \leq ρ_{AB} \leq -0.1 \\
+#     0.4  \leq ρ_{BC} \leq  0.5
+# ```
+# our problem is to determine upper and lower bounds on other correlation coefficient ``ρ_{AC}``.
+#
+# We solve an SDP to make use of the following positive semidefinite property
+# of the correlation matrix:
+# ```math
+# \begin{bmatrix}
+#       1   &  ρ_{AB} &  ρ_{AC} \\
+#      ρ_{AB} &  1    &  ρ_{BC}  \\
+#      ρ_{AC} &  ρ_{BC} &  1   
+# \end{bmatrix} \succeq 0
+# ```
+
+function example_correlation_problem()
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, X[1:3, 1:3], PSD)
+    S = ["A", "B", "C"]
+    ρ = Containers.DenseAxisArray(X, S, S)
+    @constraint(model, [i in S], ρ[i, i] == 1)
+    @constraint(model, -0.2 <= ρ["A", "B"] <= -0.1)
+    @constraint(model, 0.4 <= ρ["B", "C"] <= 0.5)
+    @objective(model, Max, ρ["A", "C"])
+    optimize!(model)
+    println("An upper bound for ρ_AC is $(value(ρ["A", "C"]))")
+    Test.@test value(ρ["A", "C"]) ≈ 0.87195 atol = 1e-4  #src
+    @objective(model, Min, ρ["A", "C"])
+    optimize!(model)
+    println("A lower bound for ρ_AC is $(value(ρ["A", "C"]))")
+    Test.@test value(ρ["A", "C"]) ≈ -0.978 atol = 1e-3  #src
+    return
+end
+
+example_correlation_problem()
+
+# ## The minimum distortion problem
+
+# This example arises from computational geometry, in particular the problem of
+# embedding a general finite metric space into a Euclidean space.
+#
+# It is known that the 4-point metric space defined by the star graph
+# ```raw
+#   [1]
+#     \
+#      1
+#       \
+#       [0] —- 1 -- [2]
+#       /
+#      1
+#     /
+#   [3]
+# ```
+# cannot be exactly embedded into a Euclidean space of any dimension,
+# where distances are computed by length of the shortest path between vertices.
+# A distance-preserving embedding would require the three leaf nodes to form
+# an equilateral triangle of side length 2, with the centre node (0) mapped to
+# an equidistant point at distance 1; this is impossible since the
+# [triangle inequality](https://en.wikipedia.org/wiki/Triangle_inequality)
+# in Euclidean space implies all points would need to be simultaneously
+# [collinear](https://en.wikipedia.org/wiki/Collinearity).
+#
+# Here we will formulate and solve an SDP to compute the best possible embedding,
+# that is, the embedding ``f`` assigning each vertex ``v`` to a vector ``f(v)``
+# that minimizes the distortion ``c`` such that
+# ```math
+#     D[a, b] \leq ||f(a) - f(b)|| \leq c \; D[a, b]
+# ```
+# for all edges ``(a, b)`` in the graph, where ``D[a, b]`` is the distance in the graph metric space.
+#
+# Any embedding ``f`` can be characterized by a Gram matrix ``Q``, which is PSD and
+# such that
+# ```math
+#     ||f(a) - f(b)||^2 = Q[a, a] + Q[b, b] - 2 Q[a, b]
+# ```
+# The matrix entry ``Q[a,b]`` represents the inner product of ``f(a)`` with ``f(b)``.
+#
+# We therefore impose the constraint
+# ```math
+#     D[a, b]^2 \leq Q[a, a] + Q[b, b] - 2 Q[a, b] \leq c^2 \; D[a, b]^2
+# ```
+# for all edges ``(a, b)`` in the graph and minimize ``c^2``, 
+# which gives us the SDP formulation below.
+# Since we may choose any point to be the origin, we fix the first vertex at 0.
+#
+# For more details, see:
+#
+# J. Matoušek (2002), [_Lectures on discrete geometry_](https://doi.org/10.1007/978-1-4613-0039-7),
+# Springer, pp. 378-379
+# 
+#  N. Linial (2002), 
+# _[Finite metric spaces--combinatorics, geometry and algorithms](https://arxiv.org/abs/math/0304466)_,
+# Proceedings of the ICM, Vol. 3, 573-586
+
+function example_minimum_distortion()
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    D = [
+        0.0 1.0 1.0 1.0
+        1.0 0.0 2.0 2.0
+        1.0 2.0 0.0 2.0
+        1.0 2.0 2.0 0.0
+    ]
+    @variable(model, c² >= 1.0)
+    @variable(model, Q[1:4, 1:4], PSD)
+    for i in 1:4, j in (i+1):4
+        @constraint(model, D[i, j]^2 <= Q[i, i] + Q[j, j] - 2 * Q[i, j])
+        @constraint(model, Q[i, i] + Q[j, j] - 2 * Q[i, j] <= c² * D[i, j]^2)
+    end
+    fix(Q[1, 1], 0)
+    @objective(model, Min, c²)
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ 4 / 3 atol = 1e-4
+    ## Recover the minimal distorted embedding:
+    X = [zeros(3) sqrt(value.(Q)[2:end, 2:end])]
+    return Plots.plot(
+        X[1, :],
+        X[2, :],
+        X[3, :];
+        seriestype = :mesh3d,
+        connections = ([0, 0, 0, 1], [1, 2, 3, 2], [2, 3, 1, 3]),
+        legend = false,
+        fillalpha = 0.1,
+        lw = 3,
+        ratio = :equal,
+        xlim = (-1.1, 1.1),
+        ylim = (-1.1, 1.1),
+        zlim = (-1.5, 1.0),
+        zticks = -1:1,
+        camera = (60, 30),
+    )
+end
+
+example_minimum_distortion()
+
+# ## Lovász numbers
+
+# The [Lovász number](https://en.wikipedia.org/wiki/Lov%C3%A1sz_number)
+# of a graph, also known as Lovász's theta-function, is a number that
+# lies between two important and related numbers that are computationally
+# hard to determine, namely the chromatic and clique numbers of the graph.
+# It is possible however to efficient compute the Lovász number as the
+# optimal value of a semidefinite program.
+
+# Consider the pentagon graph:
+# ```raw
+#      [5]
+#     /   \    
+#    /     \     
+#  [1]     [4]
+#   |       |
+#   |       |  
+#  [2] --- [3]
+# ```
+# with five vertices and edges. Its Lovász number is known to be precisely
+# ``\sqrt{5} \approx 2.236``, lying between 2 (the largest clique size) and
+# 3 (the smallest number needed for a vertex coloring).
+
+# Let ``i, \, j`` be integers such that ``1 \leq i < j \leq 5``.
+# We define ``A^{ij}`` to be the ``5 \times 5`` symmetric
+# matrix with entries ``(i,j)`` and ``(j,i)`` equal to 1, with all other entries 0.
+# Let  ``E`` be the graph's edge set; in this example, ``E`` contains
+# (1,2), (2,3), (3,4), (4,5), (5,1)
+# and their transposes. The Lovász number can be computed from the program 
+# ```math
+# \begin{align}
+#     \text{max}      & \quad    J • X & \\
+#     \text{    s.t.} & \quad A^{ij} • X = 0 \text{ for all } (i,j) \notin E   \\
+#                     & \quad I • X = 1 \\
+#                     & \quad X \succeq 0
+# \end{align}
+# ```
+# where ``J`` is the matrix filled with ones, and ``I`` is the identity matrix.
+#
+# For more details, see:
+#
+# Barvinok, A. (2002).
+# A course in convexity,
+# American Mathematical Society (Vol. 54), pp. 182-185.
+#
+# Knuth, D. E. (1994), 
+# [_The sandwich theorem_](https://doi.org/10.37236%2F1193), 
+# Electronic Journal of Combinatorics, Volume 1, Issue 1, A1.
+
+function example_theta_problem()
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    E = [(1, 2), (2, 3), (3, 4), (4, 5), (5, 1)]
+    @variable(model, X[1:5, 1:5], PSD)
+    for i in 1:5
+        for j in (i+1):5
+            if !((i, j) in E || (j, i) in E)
+                A = zeros(Int, 5, 5)
+                A[i, j] = 1
+                A[j, i] = 1
+                @constraint(model, LinearAlgebra.dot(A, X) == 0)
+            end
+        end
+    end
+    @constraint(model, LinearAlgebra.tr(LinearAlgebra.I * X) == 1)
+    J = ones(Int, 5, 5)
+    @objective(model, Max, LinearAlgebra.dot(J, X))
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ sqrt(5) rtol = 1e-4
+    println("The Lovász number is: $(objective_value(model))")
+    return
+end
+
+example_theta_problem()
+
+# ## Robust uncertainty sets
+
+# This example computes the Value at Risk for a data-driven uncertainty set.
+# Closed-form expressions for the optimal value are available.
+# For more details, see:
+
+# Bertsimas, D., Gupta, V., & Kallus, N. (2018). 
+# [_Data-driven robust optimization._](https://doi.org/10.1007/s10107-017-1125-8)
+# Mathematical Programming, 167, 235-292.
+
+function example_robust_uncertainty_sets()
+    R, d, 𝛿, ɛ = 1, 3, 0.05, 0.05
+    N = ceil((2 + 2 * log(2 / 𝛿))^2) + 1
+    c, μhat, M = randn(d), rand(d), rand(d, d)
+    Σhat = 1 / (d - 1) * (M - ones(d) * μhat')' * (M - ones(d) * μhat')
+    Γ1(𝛿, N) = R / sqrt(N) * (2 + sqrt(2 * log(1 / 𝛿)))
+    Γ2(𝛿, N) = 2 * R^2 / sqrt(N) * (2 + sqrt(2 * log(2 / 𝛿)))
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, Σ[1:d, 1:d], PSD)
+    @variable(model, u[1:d])
+    @variable(model, μ[1:d])
+    @constraint(model, [Γ1(𝛿 / 2, N); μ - μhat] in SecondOrderCone())
+    @constraint(model, [Γ2(𝛿 / 2, N); vec(Σ - Σhat)] in SecondOrderCone())
+    @constraint(model, [((1-ɛ)/ɛ) (u - μ)'; (u-μ) Σ] >= 0, PSDCone())
+    @objective(model, Max, c' * u)
+    optimize!(model)
+    exact =
+        μhat' * c +
+        Γ1(𝛿 / 2, N) * LinearAlgebra.norm(c) +
+        sqrt((1 - ɛ) / ɛ) *
+        sqrt(c' * (Σhat + Γ2(𝛿 / 2, N) * LinearAlgebra.I) * c)
+    Test.@test objective_value(model) ≈ exact atol = 1e-2
+    return
+end
+
+example_robust_uncertainty_sets()
diff --git a/previews/PR3525/tutorials/conic/simple_examples/ee174d1a.svg b/previews/PR3525/tutorials/conic/simple_examples/ee174d1a.svg
new file mode 100644
index 00000000000..01aad96495a
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/simple_examples/ee174d1a.svg
@@ -0,0 +1,49 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip730">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip730)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip731">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<defs>
+  <clipPath id="clip732">
+    <rect x="530" y="47" width="1440" height="1440"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip732)" d="M734.179 1024.82 L734.179 370.632 L1388.36 181.785 L1766.06 508.877 L1766.06 1163.06 L1111.87 1351.91 L734.179 1024.82  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="701.469,1066.97 1421.07,859.235 1421.07,139.634 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="795.893,1148.74 1515.49,941.008 1515.49,221.407 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="890.316,1230.51 1609.92,1022.78 1609.92,303.179 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="984.739,1312.29 1704.34,1104.55 1704.34,384.952 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1079.16,1394.06 1798.76,1186.33 1798.76,466.725 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="682.585,1050.61 1098.05,1410.41 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="701.469,1066.97 710.105,1064.47 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="795.893,1148.74 804.528,1146.25 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="890.316,1230.51 898.951,1228.02 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="984.739,1312.29 993.374,1309.79 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1079.16,1394.06 1087.8,1391.57 "/>
+<path clip-path="url(#clip730)" d="M575.672 1089.19 L605.347 1089.19 L605.347 1093.12 L575.672 1093.12 L575.672 1089.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M616.25 1102.08 L623.889 1102.08 L623.889 1075.72 L615.579 1077.38 L615.579 1073.12 L623.843 1071.46 L628.518 1071.46 L628.518 1102.08 L636.157 1102.08 L636.157 1106.02 L616.25 1106.02 L616.25 1102.08 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M645.602 1100.14 L650.486 1100.14 L650.486 1106.02 L645.602 1106.02 L645.602 1100.14 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M670.671 1074.54 Q667.06 1074.54 665.231 1078.1 Q663.426 1081.64 663.426 1088.77 Q663.426 1095.88 665.231 1099.44 Q667.06 1102.98 670.671 1102.98 Q674.305 1102.98 676.111 1099.44 Q677.94 1095.88 677.94 1088.77 Q677.94 1081.64 676.111 1078.1 Q674.305 1074.54 670.671 1074.54 M670.671 1070.83 Q676.481 1070.83 679.537 1075.44 Q682.615 1080.02 682.615 1088.77 Q682.615 1097.5 679.537 1102.11 Q676.481 1106.69 670.671 1106.69 Q664.861 1106.69 661.782 1102.11 Q658.727 1097.5 658.727 1088.77 Q658.727 1080.02 661.782 1075.44 Q664.861 1070.83 670.671 1070.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M671.09 1170.96 L700.766 1170.96 L700.766 1174.9 L671.09 1174.9 L671.09 1170.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M720.858 1156.31 Q717.247 1156.31 715.419 1159.87 Q713.613 1163.42 713.613 1170.55 Q713.613 1177.65 715.419 1181.22 Q717.247 1184.76 720.858 1184.76 Q724.493 1184.76 726.298 1181.22 Q728.127 1177.65 728.127 1170.55 Q728.127 1163.42 726.298 1159.87 Q724.493 1156.31 720.858 1156.31 M720.858 1152.61 Q726.669 1152.61 729.724 1157.21 Q732.803 1161.8 732.803 1170.55 Q732.803 1179.27 729.724 1183.88 Q726.669 1188.46 720.858 1188.46 Q715.048 1188.46 711.97 1183.88 Q708.914 1179.27 708.914 1170.55 Q708.914 1161.8 711.97 1157.21 Q715.048 1152.61 720.858 1152.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M741.02 1181.91 L745.905 1181.91 L745.905 1187.79 L741.02 1187.79 L741.02 1181.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M756.136 1153.23 L774.492 1153.23 L774.492 1157.17 L760.418 1157.17 L760.418 1165.64 Q761.437 1165.29 762.455 1165.13 Q763.474 1164.94 764.492 1164.94 Q770.279 1164.94 773.659 1168.11 Q777.039 1171.29 777.039 1176.7 Q777.039 1182.28 773.566 1185.38 Q770.094 1188.46 763.775 1188.46 Q761.599 1188.46 759.33 1188.09 Q757.085 1187.72 754.678 1186.98 L754.678 1182.28 Q756.761 1183.42 758.983 1183.97 Q761.205 1184.53 763.682 1184.53 Q767.687 1184.53 770.025 1182.42 Q772.363 1180.31 772.363 1176.7 Q772.363 1173.09 770.025 1170.98 Q767.687 1168.88 763.682 1168.88 Q761.807 1168.88 759.932 1169.3 Q758.08 1169.71 756.136 1170.59 L756.136 1153.23 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M814.286 1238.08 Q810.675 1238.08 808.847 1241.65 Q807.041 1245.19 807.041 1252.32 Q807.041 1259.42 808.847 1262.99 Q810.675 1266.53 814.286 1266.53 Q817.921 1266.53 819.726 1262.99 Q821.555 1259.42 821.555 1252.32 Q821.555 1245.19 819.726 1241.65 Q817.921 1238.08 814.286 1238.08 M814.286 1234.38 Q820.096 1234.38 823.152 1238.98 Q826.231 1243.57 826.231 1252.32 Q826.231 1261.04 823.152 1265.65 Q820.096 1270.23 814.286 1270.23 Q808.476 1270.23 805.397 1265.65 Q802.342 1261.04 802.342 1252.32 Q802.342 1243.57 805.397 1238.98 Q808.476 1234.38 814.286 1234.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M834.448 1263.68 L839.332 1263.68 L839.332 1269.56 L834.448 1269.56 L834.448 1263.68 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M859.518 1238.08 Q855.906 1238.08 854.078 1241.65 Q852.272 1245.19 852.272 1252.32 Q852.272 1259.42 854.078 1262.99 Q855.906 1266.53 859.518 1266.53 Q863.152 1266.53 864.957 1262.99 Q866.786 1259.42 866.786 1252.32 Q866.786 1245.19 864.957 1241.65 Q863.152 1238.08 859.518 1238.08 M859.518 1234.38 Q865.328 1234.38 868.383 1238.98 Q871.462 1243.57 871.462 1252.32 Q871.462 1261.04 868.383 1265.65 Q865.328 1270.23 859.518 1270.23 Q853.707 1270.23 850.629 1265.65 Q847.573 1261.04 847.573 1252.32 Q847.573 1243.57 850.629 1238.98 Q853.707 1234.38 859.518 1234.38 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M909.705 1319.85 Q906.094 1319.85 904.265 1323.42 Q902.46 1326.96 902.46 1334.09 Q902.46 1341.2 904.265 1344.76 Q906.094 1348.3 909.705 1348.3 Q913.339 1348.3 915.145 1344.76 Q916.973 1341.2 916.973 1334.09 Q916.973 1326.96 915.145 1323.42 Q913.339 1319.85 909.705 1319.85 M909.705 1316.15 Q915.515 1316.15 918.571 1320.76 Q921.649 1325.34 921.649 1334.09 Q921.649 1342.82 918.571 1347.42 Q915.515 1352.01 909.705 1352.01 Q903.895 1352.01 900.816 1347.42 Q897.761 1342.82 897.761 1334.09 Q897.761 1325.34 900.816 1320.76 Q903.895 1316.15 909.705 1316.15 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M929.867 1345.46 L934.751 1345.46 L934.751 1351.34 L929.867 1351.34 L929.867 1345.46 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M944.983 1316.78 L963.339 1316.78 L963.339 1320.71 L949.265 1320.71 L949.265 1329.18 Q950.283 1328.84 951.302 1328.67 Q952.32 1328.49 953.339 1328.49 Q959.126 1328.49 962.506 1331.66 Q965.885 1334.83 965.885 1340.25 Q965.885 1345.83 962.413 1348.93 Q958.941 1352.01 952.621 1352.01 Q950.445 1352.01 948.177 1351.64 Q945.932 1351.27 943.524 1350.53 L943.524 1345.83 Q945.608 1346.96 947.83 1347.52 Q950.052 1348.07 952.529 1348.07 Q956.533 1348.07 958.871 1345.97 Q961.209 1343.86 961.209 1340.25 Q961.209 1336.64 958.871 1334.53 Q956.533 1332.42 952.529 1332.42 Q950.654 1332.42 948.779 1332.84 Q946.927 1333.26 944.983 1334.14 L944.983 1316.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M993.943 1429.17 L1001.58 1429.17 L1001.58 1402.81 L993.272 1404.48 L993.272 1400.22 L1001.54 1398.55 L1006.21 1398.55 L1006.21 1429.17 L1013.85 1429.17 L1013.85 1433.11 L993.943 1433.11 L993.943 1429.17 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1023.29 1427.23 L1028.18 1427.23 L1028.18 1433.11 L1023.29 1433.11 L1023.29 1427.23 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1048.36 1401.63 Q1044.75 1401.63 1042.92 1405.19 Q1041.12 1408.73 1041.12 1415.86 Q1041.12 1422.97 1042.92 1426.54 Q1044.75 1430.08 1048.36 1430.08 Q1052 1430.08 1053.8 1426.54 Q1055.63 1422.97 1055.63 1415.86 Q1055.63 1408.73 1053.8 1405.19 Q1052 1401.63 1048.36 1401.63 M1048.36 1397.92 Q1054.17 1397.92 1057.23 1402.53 Q1060.31 1407.11 1060.31 1415.86 Q1060.31 1424.59 1057.23 1429.2 Q1054.17 1433.78 1048.36 1433.78 Q1042.55 1433.78 1039.48 1429.2 Q1036.42 1424.59 1036.42 1415.86 Q1036.42 1407.11 1039.48 1402.53 Q1042.55 1397.92 1048.36 1397.92 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1130.76,1400.97 715.294,1041.17 715.294,321.568 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1294.3,1353.76 878.84,993.958 878.84,274.356 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1457.85,1306.55 1042.39,946.746 1042.39,227.145 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1621.39,1259.34 1205.93,899.535 1205.93,179.933 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1784.94,1212.12 1369.48,852.323 1369.48,132.721 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1098.05,1410.41 1817.65,1202.68 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1130.76,1400.97 1125.77,1396.65 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1294.3,1353.76 1289.32,1349.44 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1457.85,1306.55 1452.86,1302.23 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1621.39,1259.34 1616.41,1255.02 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1784.94,1212.12 1779.95,1207.81 "/>
+<path clip-path="url(#clip730)" d="M1138.74 1429.09 L1168.42 1429.09 L1168.42 1433.03 L1138.74 1433.03 L1138.74 1429.09 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1179.32 1441.99 L1186.96 1441.99 L1186.96 1415.62 L1178.65 1417.29 L1178.65 1413.03 L1186.92 1411.36 L1191.59 1411.36 L1191.59 1441.99 L1199.23 1441.99 L1199.23 1445.92 L1179.32 1445.92 L1179.32 1441.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1208.67 1440.04 L1213.56 1440.04 L1213.56 1445.92 L1208.67 1445.92 L1208.67 1440.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1233.74 1414.44 Q1230.13 1414.44 1228.3 1418 Q1226.5 1421.55 1226.5 1428.68 Q1226.5 1435.78 1228.3 1439.35 Q1230.13 1442.89 1233.74 1442.89 Q1237.38 1442.89 1239.18 1439.35 Q1241.01 1435.78 1241.01 1428.68 Q1241.01 1421.55 1239.18 1418 Q1237.38 1414.44 1233.74 1414.44 M1233.74 1410.74 Q1239.55 1410.74 1242.61 1415.34 Q1245.69 1419.93 1245.69 1428.68 Q1245.69 1437.4 1242.61 1442.01 Q1239.55 1446.59 1233.74 1446.59 Q1227.93 1446.59 1224.85 1442.01 Q1221.8 1437.4 1221.8 1428.68 Q1221.8 1419.93 1224.85 1415.34 Q1227.93 1410.74 1233.74 1410.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1302.29 1381.88 L1331.97 1381.88 L1331.97 1385.82 L1302.29 1385.82 L1302.29 1381.88 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1352.06 1367.23 Q1348.45 1367.23 1346.62 1370.79 Q1344.81 1374.33 1344.81 1381.46 Q1344.81 1388.57 1346.62 1392.14 Q1348.45 1395.68 1352.06 1395.68 Q1355.69 1395.68 1357.5 1392.14 Q1359.33 1388.57 1359.33 1381.46 Q1359.33 1374.33 1357.5 1370.79 Q1355.69 1367.23 1352.06 1367.23 M1352.06 1363.52 Q1357.87 1363.52 1360.92 1368.13 Q1364 1372.71 1364 1381.46 Q1364 1390.19 1360.92 1394.8 Q1357.87 1399.38 1352.06 1399.38 Q1346.25 1399.38 1343.17 1394.8 Q1340.11 1390.19 1340.11 1381.46 Q1340.11 1372.71 1343.17 1368.13 Q1346.25 1363.52 1352.06 1363.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1372.22 1392.83 L1377.1 1392.83 L1377.1 1398.71 L1372.22 1398.71 L1372.22 1392.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1387.34 1364.15 L1405.69 1364.15 L1405.69 1368.08 L1391.62 1368.08 L1391.62 1376.56 Q1392.64 1376.21 1393.66 1376.05 Q1394.67 1375.86 1395.69 1375.86 Q1401.48 1375.86 1404.86 1379.03 Q1408.24 1382.2 1408.24 1387.62 Q1408.24 1393.2 1404.77 1396.3 Q1401.29 1399.38 1394.97 1399.38 Q1392.8 1399.38 1390.53 1399.01 Q1388.28 1398.64 1385.88 1397.9 L1385.88 1393.2 Q1387.96 1394.33 1390.18 1394.89 Q1392.41 1395.45 1394.88 1395.45 Q1398.89 1395.45 1401.22 1393.34 Q1403.56 1391.23 1403.56 1387.62 Q1403.56 1384.01 1401.22 1381.9 Q1398.89 1379.8 1394.88 1379.8 Q1393.01 1379.8 1391.13 1380.21 Q1389.28 1380.63 1387.34 1381.51 L1387.34 1364.15 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1477.78 1320.02 Q1474.17 1320.02 1472.34 1323.58 Q1470.53 1327.12 1470.53 1334.25 Q1470.53 1341.36 1472.34 1344.92 Q1474.17 1348.47 1477.78 1348.47 Q1481.41 1348.47 1483.22 1344.92 Q1485.05 1341.36 1485.05 1334.25 Q1485.05 1327.12 1483.22 1323.58 Q1481.41 1320.02 1477.78 1320.02 M1477.78 1316.31 Q1483.59 1316.31 1486.65 1320.92 Q1489.72 1325.5 1489.72 1334.25 Q1489.72 1342.98 1486.65 1347.59 Q1483.59 1352.17 1477.78 1352.17 Q1471.97 1352.17 1468.89 1347.59 Q1465.84 1342.98 1465.84 1334.25 Q1465.84 1325.5 1468.89 1320.92 Q1471.97 1316.31 1477.78 1316.31 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1497.94 1345.62 L1502.83 1345.62 L1502.83 1351.5 L1497.94 1351.5 L1497.94 1345.62 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1523.01 1320.02 Q1519.4 1320.02 1517.57 1323.58 Q1515.77 1327.12 1515.77 1334.25 Q1515.77 1341.36 1517.57 1344.92 Q1519.4 1348.47 1523.01 1348.47 Q1526.65 1348.47 1528.45 1344.92 Q1530.28 1341.36 1530.28 1334.25 Q1530.28 1327.12 1528.45 1323.58 Q1526.65 1320.02 1523.01 1320.02 M1523.01 1316.31 Q1528.82 1316.31 1531.88 1320.92 Q1534.96 1325.5 1534.96 1334.25 Q1534.96 1342.98 1531.88 1347.59 Q1528.82 1352.17 1523.01 1352.17 Q1517.2 1352.17 1514.12 1347.59 Q1511.07 1342.98 1511.07 1334.25 Q1511.07 1325.5 1514.12 1320.92 Q1517.2 1316.31 1523.01 1316.31 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1641.33 1272.8 Q1637.71 1272.8 1635.89 1276.37 Q1634.08 1279.91 1634.08 1287.04 Q1634.08 1294.15 1635.89 1297.71 Q1637.71 1301.25 1641.33 1301.25 Q1644.96 1301.25 1646.77 1297.71 Q1648.59 1294.15 1648.59 1287.04 Q1648.59 1279.91 1646.77 1276.37 Q1644.96 1272.8 1641.33 1272.8 M1641.33 1269.1 Q1647.14 1269.1 1650.19 1273.71 Q1653.27 1278.29 1653.27 1287.04 Q1653.27 1295.77 1650.19 1300.37 Q1647.14 1304.96 1641.33 1304.96 Q1635.52 1304.96 1632.44 1300.37 Q1629.38 1295.77 1629.38 1287.04 Q1629.38 1278.29 1632.44 1273.71 Q1635.52 1269.1 1641.33 1269.1 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1661.49 1298.41 L1666.37 1298.41 L1666.37 1304.29 L1661.49 1304.29 L1661.49 1298.41 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1676.6 1269.73 L1694.96 1269.73 L1694.96 1273.66 L1680.89 1273.66 L1680.89 1282.13 Q1681.9 1281.79 1682.92 1281.62 Q1683.94 1281.44 1684.96 1281.44 Q1690.75 1281.44 1694.13 1284.61 Q1697.51 1287.78 1697.51 1293.2 Q1697.51 1298.78 1694.03 1301.88 Q1690.56 1304.96 1684.24 1304.96 Q1682.07 1304.96 1679.8 1304.59 Q1677.55 1304.22 1675.15 1303.48 L1675.15 1298.78 Q1677.23 1299.91 1679.45 1300.47 Q1681.67 1301.02 1684.15 1301.02 Q1688.15 1301.02 1690.49 1298.92 Q1692.83 1296.81 1692.83 1293.2 Q1692.83 1289.59 1690.49 1287.48 Q1688.15 1285.37 1684.15 1285.37 Q1682.27 1285.37 1680.4 1285.79 Q1678.55 1286.21 1676.6 1287.09 L1676.6 1269.73 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1793.6 1253.14 L1801.24 1253.14 L1801.24 1226.77 L1792.93 1228.44 L1792.93 1224.18 L1801.19 1222.51 L1805.87 1222.51 L1805.87 1253.14 L1813.51 1253.14 L1813.51 1257.07 L1793.6 1257.07 L1793.6 1253.14 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1822.95 1251.19 L1827.83 1251.19 L1827.83 1257.07 L1822.95 1257.07 L1822.95 1251.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M1848.02 1225.59 Q1844.41 1225.59 1842.58 1229.16 Q1840.77 1232.7 1840.77 1239.83 Q1840.77 1246.94 1842.58 1250.5 Q1844.41 1254.04 1848.02 1254.04 Q1851.65 1254.04 1853.46 1250.5 Q1855.29 1246.94 1855.29 1239.83 Q1855.29 1232.7 1853.46 1229.16 Q1851.65 1225.59 1848.02 1225.59 M1848.02 1221.89 Q1853.83 1221.89 1856.89 1226.5 Q1859.96 1231.08 1859.96 1239.83 Q1859.96 1248.56 1856.89 1253.16 Q1853.83 1257.75 1848.02 1257.75 Q1842.21 1257.75 1839.13 1253.16 Q1836.08 1248.56 1836.08 1239.83 Q1836.08 1231.08 1839.13 1226.5 Q1842.21 1221.89 1848.02 1221.89 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="682.585,906.692 1402.19,698.96 1817.65,1058.76 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="682.585,618.851 1402.19,411.12 1817.65,770.921 "/>
+<polyline clip-path="url(#clip732)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="682.585,331.01 1402.19,123.279 1817.65,483.08 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="682.585,1050.61 682.585,331.01 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="682.585,906.692 691.22,904.199 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="682.585,618.851 691.22,616.358 "/>
+<polyline clip-path="url(#clip730)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="682.585,331.01 691.22,328.517 "/>
+<path clip-path="url(#clip730)" d="M593.299 907.143 L622.975 907.143 L622.975 911.078 L593.299 911.078 L593.299 907.143 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M633.877 920.037 L641.516 920.037 L641.516 893.671 L633.206 895.338 L633.206 891.078 L641.47 889.412 L646.146 889.412 L646.146 920.037 L653.785 920.037 L653.785 923.972 L633.877 923.972 L633.877 920.037 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M641.84 604.65 Q638.229 604.65 636.401 608.214 Q634.595 611.756 634.595 618.886 Q634.595 625.992 636.401 629.557 Q638.229 633.099 641.84 633.099 Q645.475 633.099 647.28 629.557 Q649.109 625.992 649.109 618.886 Q649.109 611.756 647.28 608.214 Q645.475 604.65 641.84 604.65 M641.84 600.946 Q647.651 600.946 650.706 605.552 Q653.785 610.136 653.785 618.886 Q653.785 627.612 650.706 632.219 Q647.651 636.802 641.84 636.802 Q636.03 636.802 632.952 632.219 Q629.896 627.612 629.896 618.886 Q629.896 610.136 632.952 605.552 Q636.03 600.946 641.84 600.946 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip730)" d="M633.877 344.355 L641.516 344.355 L641.516 317.989 L633.206 319.656 L633.206 315.397 L641.47 313.73 L646.146 313.73 L646.146 344.355 L653.785 344.355 L653.785 348.29 L633.877 348.29 L633.877 344.355 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip732)" d="M1250.12 694.886 L1008.51 390.241 L1275.58 1028.6 Z" fill="#009afa" fill-rule="evenodd" fill-opacity="0.101961" stroke="#009af9" stroke-opacity="0.101961" stroke-linejoin="round" stroke-width="9.6" /><path clip-path="url(#clip732)" d="M1275.58 1028.6 L1471.09 663.774 L1008.51 390.241 Z" fill="#009afa" fill-rule="evenodd" fill-opacity="0.101961" stroke="#009af9" stroke-opacity="0.101961" stroke-linejoin="round" stroke-width="9.6" /><path clip-path="url(#clip732)" d="M1250.12 694.886 L1275.58 1028.6 L1471.09 663.774 Z" fill="#009afa" fill-rule="evenodd" fill-opacity="0.101961" stroke="#009af9" stroke-opacity="0.101961" stroke-linejoin="round" stroke-width="9.6" /><path clip-path="url(#clip732)" d="M1250.12 694.886 L1471.09 663.774 L1008.51 390.241 Z" fill="#009afa" fill-rule="evenodd" fill-opacity="0.101961" stroke="#009af9" stroke-opacity="0.101961" stroke-linejoin="round" stroke-width="9.6" /></svg>
diff --git a/previews/PR3525/tutorials/conic/simple_examples/index.html b/previews/PR3525/tutorials/conic/simple_examples/index.html
new file mode 100644
index 00000000000..79f79c9d98f
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/simple_examples/index.html
@@ -0,0 +1,242 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Simple semidefinite programming examples · JuMP</title><meta name="title" content="Simple semidefinite programming examples · JuMP"/><meta property="og:title" content="Simple semidefinite programming examples · JuMP"/><meta property="twitter:title" content="Simple semidefinite programming examples · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li class="is-active"><a class="tocitem" href>Simple semidefinite programming examples</a><ul class="internal"><li><a class="tocitem" href="#Maximum-cut-via-SDP"><span>Maximum cut via SDP</span></a></li><li><a class="tocitem" href="#K-means-clustering-via-SDP"><span>K-means clustering via SDP</span></a></li><li><a class="tocitem" href="#The-correlation-problem"><span>The correlation problem</span></a></li><li><a class="tocitem" href="#The-minimum-distortion-problem"><span>The minimum distortion problem</span></a></li><li><a class="tocitem" href="#Lovász-numbers"><span>Lovász numbers</span></a></li><li><a class="tocitem" href="#Robust-uncertainty-sets"><span>Robust uncertainty sets</span></a></li></ul></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Simple semidefinite programming examples</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Simple semidefinite programming examples</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/simple_examples.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Simple-semidefinite-programming-examples"><a class="docs-heading-anchor" href="#Simple-semidefinite-programming-examples">Simple semidefinite programming examples</a><a id="Simple-semidefinite-programming-examples-1"></a><a class="docs-heading-anchor-permalink" href="#Simple-semidefinite-programming-examples" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../simple_examples.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial is a collection of examples of small conic programs from the field of <a href="https://en.wikipedia.org/wiki/Semidefinite_programming">semidefinite programming</a> (SDP).</p><p>This tutorial makes use of the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import LinearAlgebra
+import Plots
+import Random
+import SCS
+import Test</code></pre><h2 id="Maximum-cut-via-SDP"><a class="docs-heading-anchor" href="#Maximum-cut-via-SDP">Maximum cut via SDP</a><a id="Maximum-cut-via-SDP-1"></a><a class="docs-heading-anchor-permalink" href="#Maximum-cut-via-SDP" title="Permalink"></a></h2><p>The <a href="https://en.wikipedia.org/wiki/Maximum_cut">maximum cut problem</a> is a classical example in graph theory, where we seek to partition a graph into two complementary sets, such that the weight of edges between the two sets is maximized. This problem is NP-hard, but it is possible to obtain an approximate solution using the semidefinite programming relaxation:</p><p class="math-container">\[    \text{max}  \quad   0.25 L•X \\
+    \text{    s.t.} \quad   \mathrm{diag}(X) = e \\
+                 \qquad X \succeq 0\]</p><p>where <span>$L$</span> is the weighted graph Laplacian and <span>$e$</span> is a vector of ones. For more details, see:</p><p>Goemans, M. X., &amp; Williamson, D. P. (1995). <a href="https://doi.org/10.1145/227683.227684"><em>Improved approximation algorithms for maximum cut and satisfiability problems using semidefinite programming.</em></a> Journal of the ACM (JACM), 42(6), 1115-1145.</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    svd_cholesky(X::AbstractMatrix, rtol)
+
+Return the matrix `U` of the Cholesky decomposition of `X` as `U&#39; * U`.
+Note that we do not use the `LinearAlgebra.cholesky` function because it
+requires the matrix to be positive definite while `X` may be only
+positive *semi*definite.
+
+We use the convention `U&#39; * U` instead of `U * U&#39;` to be consistent with
+`LinearAlgebra.cholesky`.
+&quot;&quot;&quot;
+function svd_cholesky(X::AbstractMatrix)
+    F = LinearAlgebra.svd(X)
+    # We now have `X ≈ `F.U * D² * F.U&#39;` where:
+    D = LinearAlgebra.Diagonal(sqrt.(F.S))
+    # So `X ≈ U&#39; * U` where `U` is:
+    return (F.U * D)&#39;
+end
+
+function solve_max_cut_sdp(weights)
+    N = size(weights, 1)
+    # Calculate the (weighted) Laplacian of the graph: L = D - W.
+    L = LinearAlgebra.diagm(0 =&gt; weights * ones(N)) - weights
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, X[1:N, 1:N], PSD)
+    for i in 1:N
+        set_start_value(X[i, i], 1.0)
+    end
+    @objective(model, Max, 0.25 * LinearAlgebra.dot(L, X))
+    @constraint(model, LinearAlgebra.diag(X) .== 1)
+    optimize!(model)
+    V = svd_cholesky(value(X))
+    Random.seed!(N)
+    r = rand(N)
+    r /= LinearAlgebra.norm(r)
+    cut = [LinearAlgebra.dot(r, V[:, i]) &gt; 0 for i in 1:N]
+    S = findall(cut)
+    T = findall(.!cut)
+    println(&quot;Solution:&quot;)
+    println(&quot; (S, T) = ({&quot;, join(S, &quot;, &quot;), &quot;}, {&quot;, join(T, &quot;, &quot;), &quot;})&quot;)
+    return S, T
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_max_cut_sdp (generic function with 1 method)</code></pre><p>Given the graph</p><pre><code class="language-raw hljs">[1] --- 5 --- [2]</code></pre><p>The solution is <code>(S, T)  = ({1}, {2})</code></p><pre><code class="language-julia hljs">S, T = solve_max_cut_sdp([0 5; 5 0])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">([2], [1])</code></pre><p>Given the graph</p><pre><code class="language-raw hljs">[1] --- 5 --- [2]
+ |  \          |
+ |    \        |
+ 7      6      1
+ |        \    |
+ |          \  |
+[3] --- 1 --- [4]</code></pre><p>The solution is <code>(S, T)  = ({1}, {2, 3, 4})</code></p><pre><code class="language-julia hljs">S, T = solve_max_cut_sdp([0 5 7 6; 5 0 0 1; 7 0 0 1; 6 1 1 0])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">([1], [2, 3, 4])</code></pre><p>Given the graph</p><pre><code class="language-raw hljs">[1] --- 1 --- [2]
+ |             |
+ |             |
+ 5             9
+ |             |
+ |             |
+[3] --- 2 --- [4]</code></pre><p>The solution is <code>(S, T)  = ({1, 4}, {2, 3})</code></p><pre><code class="language-julia hljs">S, T = solve_max_cut_sdp([0 1 5 0; 1 0 0 9; 5 0 0 2; 0 9 2 0])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">([1, 4], [2, 3])</code></pre><h2 id="K-means-clustering-via-SDP"><a class="docs-heading-anchor" href="#K-means-clustering-via-SDP">K-means clustering via SDP</a><a id="K-means-clustering-via-SDP-1"></a><a class="docs-heading-anchor-permalink" href="#K-means-clustering-via-SDP" title="Permalink"></a></h2><p>Given a set of points <span>$a_1, \ldots, a_m$</span>  in <span>$\mathbb{R}^n$</span>, allocate them to <span>$k$</span> clusters.</p><p>For more details, see:</p><p>Peng, J., &amp; Wei, Y. (2007). <a href="https://doi.org/10.1137/050641983"><em>Approximating k-means-type clustering via semidefinite programming</em></a>. SIAM Journal on Optimization, 18(1), 186-205.</p><pre><code class="language-julia hljs">function example_k_means_clustering()
+    a = [[2.0, 2.0], [2.5, 2.1], [7.0, 7.0], [2.2, 2.3], [6.8, 7.0], [7.2, 7.5]]
+    m = length(a)
+    num_clusters = 2
+    W = zeros(m, m)
+    for i in 1:m, j in i+1:m
+        W[i, j] = W[j, i] = exp(-LinearAlgebra.norm(a[i] - a[j]) / 1.0)
+    end
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, Z[1:m, 1:m] &gt;= 0, PSD)
+    @objective(model, Min, LinearAlgebra.tr(W * (LinearAlgebra.I - Z)))
+    @constraint(model, [i = 1:m], sum(Z[i, :]) .== 1)
+    @constraint(model, LinearAlgebra.tr(Z) == num_clusters)
+    optimize!(model)
+    Z_val = value.(Z)
+    current_cluster, visited = 0, Set{Int}()
+    for i in 1:m
+        if !(i in visited)
+            current_cluster += 1
+            println(&quot;Cluster $current_cluster&quot;)
+            for j in i:m
+                if isapprox(Z_val[i, i], Z_val[i, j]; atol = 1e-3)
+                    println(a[j])
+                    push!(visited, j)
+                end
+            end
+        end
+    end
+    return
+end
+
+example_k_means_clustering()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Cluster 1
+[2.0, 2.0]
+[2.5, 2.1]
+[2.2, 2.3]
+Cluster 2
+[7.0, 7.0]
+[6.8, 7.0]
+[7.2, 7.5]</code></pre><h2 id="The-correlation-problem"><a class="docs-heading-anchor" href="#The-correlation-problem">The correlation problem</a><a id="The-correlation-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-correlation-problem" title="Permalink"></a></h2><p>Given three random variables <span>$A$</span>, <span>$B$</span>, and <span>$C$</span>, and given bounds on two of the three correlation coefficients:</p><p class="math-container">\[    -0.2 \leq ρ_{AB} \leq -0.1 \\
+    0.4  \leq ρ_{BC} \leq  0.5\]</p><p>our problem is to determine upper and lower bounds on other correlation coefficient <span>$ρ_{AC}$</span>.</p><p>We solve an SDP to make use of the following positive semidefinite property of the correlation matrix:</p><p class="math-container">\[\begin{bmatrix}
+      1   &amp;  ρ_{AB} &amp;  ρ_{AC} \\
+     ρ_{AB} &amp;  1    &amp;  ρ_{BC}  \\
+     ρ_{AC} &amp;  ρ_{BC} &amp;  1
+\end{bmatrix} \succeq 0\]</p><pre><code class="language-julia hljs">function example_correlation_problem()
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, X[1:3, 1:3], PSD)
+    S = [&quot;A&quot;, &quot;B&quot;, &quot;C&quot;]
+    ρ = Containers.DenseAxisArray(X, S, S)
+    @constraint(model, [i in S], ρ[i, i] == 1)
+    @constraint(model, -0.2 &lt;= ρ[&quot;A&quot;, &quot;B&quot;] &lt;= -0.1)
+    @constraint(model, 0.4 &lt;= ρ[&quot;B&quot;, &quot;C&quot;] &lt;= 0.5)
+    @objective(model, Max, ρ[&quot;A&quot;, &quot;C&quot;])
+    optimize!(model)
+    println(&quot;An upper bound for ρ_AC is $(value(ρ[&quot;A&quot;, &quot;C&quot;]))&quot;)
+    @objective(model, Min, ρ[&quot;A&quot;, &quot;C&quot;])
+    optimize!(model)
+    println(&quot;A lower bound for ρ_AC is $(value(ρ[&quot;A&quot;, &quot;C&quot;]))&quot;)
+    return
+end
+
+example_correlation_problem()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">An upper bound for ρ_AC is 0.8719220302963583
+A lower bound for ρ_AC is -0.9779989594132403</code></pre><h2 id="The-minimum-distortion-problem"><a class="docs-heading-anchor" href="#The-minimum-distortion-problem">The minimum distortion problem</a><a id="The-minimum-distortion-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-minimum-distortion-problem" title="Permalink"></a></h2><p>This example arises from computational geometry, in particular the problem of embedding a general finite metric space into a Euclidean space.</p><p>It is known that the 4-point metric space defined by the star graph</p><pre><code class="language-raw hljs">  [1]
+    \
+     1
+      \
+      [0] —- 1 -- [2]
+      /
+     1
+    /
+  [3]</code></pre><p>cannot be exactly embedded into a Euclidean space of any dimension, where distances are computed by length of the shortest path between vertices. A distance-preserving embedding would require the three leaf nodes to form an equilateral triangle of side length 2, with the centre node (0) mapped to an equidistant point at distance 1; this is impossible since the <a href="https://en.wikipedia.org/wiki/Triangle_inequality">triangle inequality</a> in Euclidean space implies all points would need to be simultaneously <a href="https://en.wikipedia.org/wiki/Collinearity">collinear</a>.</p><p>Here we will formulate and solve an SDP to compute the best possible embedding, that is, the embedding <span>$f$</span> assigning each vertex <span>$v$</span> to a vector <span>$f(v)$</span> that minimizes the distortion <span>$c$</span> such that</p><p class="math-container">\[    D[a, b] \leq ||f(a) - f(b)|| \leq c \; D[a, b]\]</p><p>for all edges <span>$(a, b)$</span> in the graph, where <span>$D[a, b]$</span> is the distance in the graph metric space.</p><p>Any embedding <span>$f$</span> can be characterized by a Gram matrix <span>$Q$</span>, which is PSD and such that</p><p class="math-container">\[    ||f(a) - f(b)||^2 = Q[a, a] + Q[b, b] - 2 Q[a, b]\]</p><p>The matrix entry <span>$Q[a,b]$</span> represents the inner product of <span>$f(a)$</span> with <span>$f(b)$</span>.</p><p>We therefore impose the constraint</p><p class="math-container">\[    D[a, b]^2 \leq Q[a, a] + Q[b, b] - 2 Q[a, b] \leq c^2 \; D[a, b]^2\]</p><p>for all edges <span>$(a, b)$</span> in the graph and minimize <span>$c^2$</span>, which gives us the SDP formulation below. Since we may choose any point to be the origin, we fix the first vertex at 0.</p><p>For more details, see:</p><p>J. Matoušek (2002), <a href="https://doi.org/10.1007/978-1-4613-0039-7"><em>Lectures on discrete geometry</em></a>, Springer, pp. 378-379</p><p>N. Linial (2002), <em><a href="https://arxiv.org/abs/math/0304466">Finite metric spaces–combinatorics, geometry and algorithms</a></em>, Proceedings of the ICM, Vol. 3, 573-586</p><pre><code class="language-julia hljs">function example_minimum_distortion()
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    D = [
+        0.0 1.0 1.0 1.0
+        1.0 0.0 2.0 2.0
+        1.0 2.0 0.0 2.0
+        1.0 2.0 2.0 0.0
+    ]
+    @variable(model, c² &gt;= 1.0)
+    @variable(model, Q[1:4, 1:4], PSD)
+    for i in 1:4, j in (i+1):4
+        @constraint(model, D[i, j]^2 &lt;= Q[i, i] + Q[j, j] - 2 * Q[i, j])
+        @constraint(model, Q[i, i] + Q[j, j] - 2 * Q[i, j] &lt;= c² * D[i, j]^2)
+    end
+    fix(Q[1, 1], 0)
+    @objective(model, Min, c²)
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ 4 / 3 atol = 1e-4
+    # Recover the minimal distorted embedding:
+    X = [zeros(3) sqrt(value.(Q)[2:end, 2:end])]
+    return Plots.plot(
+        X[1, :],
+        X[2, :],
+        X[3, :];
+        seriestype = :mesh3d,
+        connections = ([0, 0, 0, 1], [1, 2, 3, 2], [2, 3, 1, 3]),
+        legend = false,
+        fillalpha = 0.1,
+        lw = 3,
+        ratio = :equal,
+        xlim = (-1.1, 1.1),
+        ylim = (-1.1, 1.1),
+        zlim = (-1.5, 1.0),
+        zticks = -1:1,
+        camera = (60, 30),
+    )
+end
+
+example_minimum_distortion()</code></pre><img src="ee174d1a.svg" alt="Example block output"/><h2 id="Lovász-numbers"><a class="docs-heading-anchor" href="#Lovász-numbers">Lovász numbers</a><a id="Lovász-numbers-1"></a><a class="docs-heading-anchor-permalink" href="#Lovász-numbers" title="Permalink"></a></h2><p>The <a href="https://en.wikipedia.org/wiki/Lov%C3%A1sz_number">Lovász number</a> of a graph, also known as Lovász&#39;s theta-function, is a number that lies between two important and related numbers that are computationally hard to determine, namely the chromatic and clique numbers of the graph. It is possible however to efficient compute the Lovász number as the optimal value of a semidefinite program.</p><p>Consider the pentagon graph:</p><pre><code class="language-raw hljs">     [5]
+    /   \
+   /     \
+ [1]     [4]
+  |       |
+  |       |
+ [2] --- [3]</code></pre><p>with five vertices and edges. Its Lovász number is known to be precisely <span>$\sqrt{5} \approx 2.236$</span>, lying between 2 (the largest clique size) and 3 (the smallest number needed for a vertex coloring).</p><p>Let <span>$i, \, j$</span> be integers such that <span>$1 \leq i &lt; j \leq 5$</span>. We define <span>$A^{ij}$</span> to be the <span>$5 \times 5$</span> symmetric matrix with entries <span>$(i,j)$</span> and <span>$(j,i)$</span> equal to 1, with all other entries 0. Let  <span>$E$</span> be the graph&#39;s edge set; in this example, <span>$E$</span> contains (1,2), (2,3), (3,4), (4,5), (5,1) and their transposes. The Lovász number can be computed from the program</p><p class="math-container">\[\begin{align}
+    \text{max}      &amp; \quad    J • X &amp; \\
+    \text{    s.t.} &amp; \quad A^{ij} • X = 0 \text{ for all } (i,j) \notin E   \\
+                    &amp; \quad I • X = 1 \\
+                    &amp; \quad X \succeq 0
+\end{align}\]</p><p>where <span>$J$</span> is the matrix filled with ones, and <span>$I$</span> is the identity matrix.</p><p>For more details, see:</p><p>Barvinok, A. (2002). A course in convexity, American Mathematical Society (Vol. 54), pp. 182-185.</p><p>Knuth, D. E. (1994), <a href="https://doi.org/10.37236%2F1193"><em>The sandwich theorem</em></a>, Electronic Journal of Combinatorics, Volume 1, Issue 1, A1.</p><pre><code class="language-julia hljs">function example_theta_problem()
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    E = [(1, 2), (2, 3), (3, 4), (4, 5), (5, 1)]
+    @variable(model, X[1:5, 1:5], PSD)
+    for i in 1:5
+        for j in (i+1):5
+            if !((i, j) in E || (j, i) in E)
+                A = zeros(Int, 5, 5)
+                A[i, j] = 1
+                A[j, i] = 1
+                @constraint(model, LinearAlgebra.dot(A, X) == 0)
+            end
+        end
+    end
+    @constraint(model, LinearAlgebra.tr(LinearAlgebra.I * X) == 1)
+    J = ones(Int, 5, 5)
+    @objective(model, Max, LinearAlgebra.dot(J, X))
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ sqrt(5) rtol = 1e-4
+    println(&quot;The Lovász number is: $(objective_value(model))&quot;)
+    return
+end
+
+example_theta_problem()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">The Lovász number is: 2.236067861794848</code></pre><h2 id="Robust-uncertainty-sets"><a class="docs-heading-anchor" href="#Robust-uncertainty-sets">Robust uncertainty sets</a><a id="Robust-uncertainty-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Robust-uncertainty-sets" title="Permalink"></a></h2><p>This example computes the Value at Risk for a data-driven uncertainty set. Closed-form expressions for the optimal value are available. For more details, see:</p><p>Bertsimas, D., Gupta, V., &amp; Kallus, N. (2018). <a href="https://doi.org/10.1007/s10107-017-1125-8"><em>Data-driven robust optimization.</em></a> Mathematical Programming, 167, 235-292.</p><pre><code class="language-julia hljs">function example_robust_uncertainty_sets()
+    R, d, 𝛿, ɛ = 1, 3, 0.05, 0.05
+    N = ceil((2 + 2 * log(2 / 𝛿))^2) + 1
+    c, μhat, M = randn(d), rand(d), rand(d, d)
+    Σhat = 1 / (d - 1) * (M - ones(d) * μhat&#39;)&#39; * (M - ones(d) * μhat&#39;)
+    Γ1(𝛿, N) = R / sqrt(N) * (2 + sqrt(2 * log(1 / 𝛿)))
+    Γ2(𝛿, N) = 2 * R^2 / sqrt(N) * (2 + sqrt(2 * log(2 / 𝛿)))
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, Σ[1:d, 1:d], PSD)
+    @variable(model, u[1:d])
+    @variable(model, μ[1:d])
+    @constraint(model, [Γ1(𝛿 / 2, N); μ - μhat] in SecondOrderCone())
+    @constraint(model, [Γ2(𝛿 / 2, N); vec(Σ - Σhat)] in SecondOrderCone())
+    @constraint(model, [((1-ɛ)/ɛ) (u - μ)&#39;; (u-μ) Σ] &gt;= 0, PSDCone())
+    @objective(model, Max, c&#39; * u)
+    optimize!(model)
+    exact =
+        μhat&#39; * c +
+        Γ1(𝛿 / 2, N) * LinearAlgebra.norm(c) +
+        sqrt((1 - ɛ) / ɛ) *
+        sqrt(c&#39; * (Σhat + Γ2(𝛿 / 2, N) * LinearAlgebra.I) * c)
+    Test.@test objective_value(model) ≈ exact atol = 1e-2
+    return
+end
+
+example_robust_uncertainty_sets()</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../tips_and_tricks/">« Tips and Tricks</a><a class="docs-footer-nextpage" href="../dualization/">Dualization »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/start_values.jl b/previews/PR3525/tutorials/conic/start_values.jl
new file mode 100644
index 00000000000..d632fc62045
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/start_values.jl
@@ -0,0 +1,88 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Primal and dual warm-starts
+
+# Some conic solvers have the ability to set warm-starts for the primal and dual
+# solution. This can improve performance, particularly if you are repeatedly
+# solving a sequence of related problems.
+
+# !!! tip
+#     See [`set_start_values`](@ref) for a generic implementation of this
+#     function that was added to JuMP after this tutorial was written.
+
+# In this tutorial, we demonstrate how to write a function that sets the primal
+# and dual starts as the optimal solution stored in a model. It is intended to
+# be a starting point for which you can modify if you want to do something
+# similar in your own code.
+
+# !!! warning
+#     This tutorial does not set start values for nonlinear models.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import SCS
+
+# The main component of this tutorial is the following function. The most
+# important observation is that we cache all of the solution values first, and
+# then we modify the model second. (Alternating between querying a value and
+# modifying the model is not allowed in JuMP.)
+
+function set_optimal_start_values(model::Model)
+    ## Store a mapping of the variable primal solution
+    variable_primal = Dict(x => value(x) for x in all_variables(model))
+    ## In the following, we loop through every constraint and store a mapping
+    ## from the constraint index to a tuple containing the primal and dual
+    ## solutions.
+    constraint_solution = Dict()
+    for (F, S) in list_of_constraint_types(model)
+        ## We add a try-catch here because some constraint types might not
+        ## support getting the primal or dual solution.
+        try
+            for ci in all_constraints(model, F, S)
+                constraint_solution[ci] = (value(ci), dual(ci))
+            end
+        catch
+            @info("Something went wrong getting $F-in-$S. Skipping")
+        end
+    end
+    ## Now we can loop through our cached solutions and set the starting values.
+    for (x, primal_start) in variable_primal
+        set_start_value(x, primal_start)
+    end
+    for (ci, (primal_start, dual_start)) in constraint_solution
+        set_start_value(ci, primal_start)
+        set_dual_start_value(ci, dual_start)
+    end
+    return
+end
+
+# To test our function, we use the following linear program:
+
+model = Model(SCS.Optimizer)
+@variable(model, x[1:3] >= 0)
+@constraint(model, sum(x) <= 1)
+@objective(model, Max, sum(i * x[i] for i in 1:3))
+optimize!(model)
+
+# By looking at the log, we can see that SCS took 75 iterations to find the optimal
+# solution. Now we set the optimal solution as our starting point:
+
+set_optimal_start_values(model)
+
+# and we re-optimize:
+
+optimize!(model)
+
+# Now the optimization terminates after 0 iterations because our starting point
+# is already optimal.
+
+# Note that some solvers do not support setting some parts of the starting
+# solution, for example, they may support only `set_start_value` for variables.
+# If you encounter an `UnsupportedSupported` attribute error for
+# [`MOI.VariablePrimalStart`](@ref), [`MOI.ConstraintPrimalStart`](@ref), or
+# [`MOI.ConstraintDualStart`](@ref), comment out the corresponding part of the
+# `set_optimal_start_values` function.
diff --git a/previews/PR3525/tutorials/conic/start_values/index.html b/previews/PR3525/tutorials/conic/start_values/index.html
new file mode 100644
index 00000000000..13ec6792b93
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/start_values/index.html
@@ -0,0 +1,83 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Primal and dual warm-starts · JuMP</title><meta name="title" content="Primal and dual warm-starts · JuMP"/><meta property="og:title" content="Primal and dual warm-starts · JuMP"/><meta property="twitter:title" content="Primal and dual warm-starts · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li class="is-active"><a class="tocitem" href>Primal and dual warm-starts</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Primal and dual warm-starts</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Primal and dual warm-starts</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/start_values.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Primal-and-dual-warm-starts"><a class="docs-heading-anchor" href="#Primal-and-dual-warm-starts">Primal and dual warm-starts</a><a id="Primal-and-dual-warm-starts-1"></a><a class="docs-heading-anchor-permalink" href="#Primal-and-dual-warm-starts" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../start_values.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>Some conic solvers have the ability to set warm-starts for the primal and dual solution. This can improve performance, particularly if you are repeatedly solving a sequence of related problems.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>See <a href="../../../api/JuMP/#JuMP.set_start_values"><code>set_start_values</code></a> for a generic implementation of this function that was added to JuMP after this tutorial was written.</p></div></div><p>In this tutorial, we demonstrate how to write a function that sets the primal and dual starts as the optimal solution stored in a model. It is intended to be a starting point for which you can modify if you want to do something similar in your own code.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This tutorial does not set start values for nonlinear models.</p></div></div><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import SCS</code></pre><p>The main component of this tutorial is the following function. The most important observation is that we cache all of the solution values first, and then we modify the model second. (Alternating between querying a value and modifying the model is not allowed in JuMP.)</p><pre><code class="language-julia hljs">function set_optimal_start_values(model::Model)
+    # Store a mapping of the variable primal solution
+    variable_primal = Dict(x =&gt; value(x) for x in all_variables(model))
+    # In the following, we loop through every constraint and store a mapping
+    # from the constraint index to a tuple containing the primal and dual
+    # solutions.
+    constraint_solution = Dict()
+    for (F, S) in list_of_constraint_types(model)
+        # We add a try-catch here because some constraint types might not
+        # support getting the primal or dual solution.
+        try
+            for ci in all_constraints(model, F, S)
+                constraint_solution[ci] = (value(ci), dual(ci))
+            end
+        catch
+            @info(&quot;Something went wrong getting $F-in-$S. Skipping&quot;)
+        end
+    end
+    # Now we can loop through our cached solutions and set the starting values.
+    for (x, primal_start) in variable_primal
+        set_start_value(x, primal_start)
+    end
+    for (ci, (primal_start, dual_start)) in constraint_solution
+        set_start_value(ci, primal_start)
+        set_dual_start_value(ci, dual_start)
+    end
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">set_optimal_start_values (generic function with 1 method)</code></pre><p>To test our function, we use the following linear program:</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+@variable(model, x[1:3] &gt;= 0)
+@constraint(model, sum(x) &lt;= 1)
+@objective(model, Max, sum(i * x[i] for i in 1:3))
+optimize!(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 3, constraints m: 4
+cones: 	  l: linear vars: 4
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 6, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 4.42e+01  1.00e+00  1.28e+02 -6.64e+01  1.00e-01  6.54e-05
+    75| 5.30e-07  2.63e-06  3.15e-07 -3.00e+00  1.00e-01  1.09e-04
+------------------------------------------------------------------
+status:  solved
+timings: total: 1.10e-04s = setup: 4.96e-05s + solve: 6.08e-05s
+	 lin-sys: 1.38e-05s, cones: 6.80e-06s, accel: 4.10e-06s
+------------------------------------------------------------------
+objective = -2.999998
+------------------------------------------------------------------</code></pre><p>By looking at the log, we can see that SCS took 75 iterations to find the optimal solution. Now we set the optimal solution as our starting point:</p><pre><code class="language-julia hljs">set_optimal_start_values(model)</code></pre><p>and we re-optimize:</p><pre><code class="language-julia hljs">optimize!(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">------------------------------------------------------------------
+	       SCS v3.2.3 - Splitting Conic Solver
+	(c) Brendan O&#39;Donoghue, Stanford University, 2012
+------------------------------------------------------------------
+problem:  variables n: 3, constraints m: 4
+cones: 	  l: linear vars: 4
+settings: eps_abs: 1.0e-04, eps_rel: 1.0e-04, eps_infeas: 1.0e-07
+	  alpha: 1.50, scale: 1.00e-01, adaptive_scale: 1
+	  max_iters: 100000, normalize: 1, rho_x: 1.00e-06
+	  acceleration_lookback: 10, acceleration_interval: 10
+lin-sys:  sparse-direct-amd-qdldl
+	  nnz(A): 6, nnz(P): 0
+------------------------------------------------------------------
+ iter | pri res | dua res |   gap   |   obj   |  scale  | time (s)
+------------------------------------------------------------------
+     0| 1.90e-05  1.56e-06  9.14e-05 -3.00e+00  1.00e-01  6.62e-05
+------------------------------------------------------------------
+status:  solved
+timings: total: 6.71e-05s = setup: 5.04e-05s + solve: 1.67e-05s
+	 lin-sys: 1.60e-06s, cones: 1.40e-06s, accel: 1.00e-07s
+------------------------------------------------------------------
+objective = -3.000044
+------------------------------------------------------------------</code></pre><p>Now the optimization terminates after 0 iterations because our starting point is already optimal.</p><p>Note that some solvers do not support setting some parts of the starting solution, for example, they may support only <code>set_start_value</code> for variables. If you encounter an <code>UnsupportedSupported</code> attribute error for <a href="../../../moi/reference/variables/#MathOptInterface.VariablePrimalStart"><code>MOI.VariablePrimalStart</code></a>, <a href="../../../moi/reference/constraints/#MathOptInterface.ConstraintPrimalStart"><code>MOI.ConstraintPrimalStart</code></a>, or <a href="../../../moi/reference/constraints/#MathOptInterface.ConstraintDualStart"><code>MOI.ConstraintDualStart</code></a>, comment out the corresponding part of the <code>set_optimal_start_values</code> function.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../introduction/">« Introduction</a><a class="docs-footer-nextpage" href="../tips_and_tricks/">Tips and Tricks »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/conic/tips_and_tricks.jl b/previews/PR3525/tutorials/conic/tips_and_tricks.jl
new file mode 100644
index 00000000000..6d558a2116c
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/tips_and_tricks.jl
@@ -0,0 +1,418 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # [Tips and Tricks](@id conic_tips_and_tricks)
+
+# **This tutorial was originally contributed by Arpit Bhatia.**
+
+# This tutorial is aimed at providing a simplistic introduction to conic
+# programming using JuMP.
+
+# It uses the following packages:
+
+using JuMP
+import SCS
+import LinearAlgebra
+
+# !!! info
+#     This tutorial uses sets from [MathOptInterface](@ref moi_documentation).
+#     By default, JuMP exports the `MOI` symbol as an alias for the
+#     MathOptInterface.jl package. We recommend making this more explicit in
+#     your code by adding the following lines:
+#     ```julia
+#     import MathOptInterface as MOI
+#     ```
+
+import Random      # hide
+Random.seed!(1234) # hide
+nothing            # hide
+
+# !!! tip
+#     A good resource for learning more about functions which can be modeled
+#     using cones is the [MOSEK Modeling Cookbook](https://docs.mosek.com/modeling-cookbook/index.html).
+
+# ## Background theory
+
+# A subset $C$ of a vector space $V$ is a cone if $\forall x \in C$ and positive
+# scalars $\lambda > 0$, the product $\lambda x \in C$.
+
+# A cone $C$ is a convex cone if $\lambda x + (1 - \lambda) y \in C$, for any
+# $\lambda \in [0, 1]$, and any $x, y \in C$.
+
+# Conic programming problems are convex optimization problems in which a convex
+# function is minimized over the intersection of an affine subspace and a convex
+# cone. An example of a conic-form minimization problems, in the primal form is:
+
+# ```math
+# \begin{aligned}
+# & \min_{x \in \mathbb{R}^n} & a_0^T x + b_0 \\
+# & \;\;\text{s.t.} & A_i x + b_i & \in \mathcal{C}_i & i = 1 \ldots m
+# \end{aligned}
+# ```
+
+# The corresponding dual problem is:
+
+# ```math
+# \begin{aligned}
+# & \max_{y_1, \ldots, y_m} & -\sum_{i=1}^m b_i^T y_i + b_0 \\
+# & \;\;\text{s.t.} & a_0 - \sum_{i=1}^m A_i^T y_i & = 0 \\
+# & & y_i & \in \mathcal{C}_i^* & i = 1 \ldots m
+# \end{aligned}
+# ```
+#
+# where each $\mathcal{C}_i$ is a closed convex cone and $\mathcal{C}_i^*$ is
+# its dual cone.
+
+# ## Second-Order Cone
+
+# The [`SecondOrderCone`](@ref) (or Lorentz Cone) of dimension $n$ is a cone of
+# the form:
+# ```math
+# K_{soc} = \{ (t, x) \in \mathbb{R}^n : t \ge ||x||_2 \}
+# ```
+
+# It is most commonly used to represent the L2-norm of the vector $x$:
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x[1:3])
+@variable(model, t)
+@constraint(model, sum(x) == 1)
+@constraint(model, [t; x] in SecondOrderCone())
+@objective(model, Min, t)
+optimize!(model)
+value(t), value.(x)
+
+# ## Rotated Second-Order Cone
+
+# A Second-Order Cone rotated by $\pi/4$ in the $(x_1,x_2)$ plane is called a
+# [`RotatedSecondOrderCone`](@ref). It is a cone of the form:
+# ```math
+# K_{rsoc} = \{ (t,u,x) \in \mathbb{R}^n : 2tu \ge ||x||_2^2, t,u \ge 0 \}
+# ```
+
+# When `u = 0.5`, it represents the sum of squares of a vector $x$:
+
+data = [1.0, 2.0, 3.0, 4.0]
+target = [0.45, 1.04, 1.51, 1.97]
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, θ)
+@variable(model, t)
+@expression(model, residuals, θ * data .- target)
+@constraint(model, [t; 0.5; residuals] in RotatedSecondOrderCone())
+@objective(model, Min, t)
+optimize!(model)
+value(θ), value(t)
+
+# ## Exponential Cone
+
+# The [`MOI.ExponentialCone`](@ref) is a set of the form:
+
+# ```math
+# K_{exp} = \{ (x,y,z) \in \mathbb{R}^3 : y \exp (x/y) \le z, y > 0 \}
+# ```
+# It can be used to model problems involving `log` and `exp`.
+
+# ### Exponential
+
+# To model ``\exp(x) \le z``, use `(x, 1, z)`:
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x == 1.5)
+@variable(model, z)
+@objective(model, Min, z)
+@constraint(model, [x, 1, z] in MOI.ExponentialCone())
+optimize!(model)
+value(z), exp(1.5)
+
+# ### Logarithm
+
+# To model ``x \le \log(z)``, use `(x, 1, z)`:
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x)
+@variable(model, z == 1.5)
+@objective(model, Max, x)
+@constraint(model, [x, 1, z] in MOI.ExponentialCone())
+optimize!(model)
+value(x), log(1.5)
+
+# ### Log-sum-exp
+
+# To model ``t \ge \log\left(\sum e^{x_i}\right)``, use:
+
+N = 3
+x0 = rand(N)
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x[i = 1:N] == x0[i])
+@variable(model, t)
+@objective(model, Min, t)
+@variable(model, u[1:N])
+@constraint(model, sum(u) <= 1)
+@constraint(model, [i = 1:N], [x[i] - t, 1, u[i]] in MOI.ExponentialCone())
+optimize!(model)
+value(t), log(sum(exp.(x0)))
+
+# ### Entropy
+
+# The entropy maximization problem consists of maximizing the entropy function,
+# $H(x) = -x\log{x}$ subject to linear inequality constraints.
+
+# ```math
+# \begin{aligned}
+# & \max & - \sum_{i=1}^n x_i \log x_i \\
+# & \;\;\text{s.t.} & \mathbf{1}^\top x = 1 \\
+# & & Ax \leq b
+# \end{aligned}
+# ```
+
+# We can model this problem using an exponential cone by using the following
+# transformation:
+
+# ```math
+# t\leq -x\log{x} \iff t\leq x\log(1/x)  \iff (t, x, 1) \in K_{exp}
+# ```
+
+# Thus, our problem becomes,
+
+# ```math
+# \begin{aligned}
+# & \max & 1^Tt \\
+# & \;\;\text{s.t.} & Ax \leq b \\
+# & & 1^T x = 1 \\
+# & & (t_i, x_i, 1) \in K_{exp} && \forall i = 1 \ldots n \\
+# \end{aligned}
+# ```
+
+m, n = 10, 15
+A, b = randn(m, n), rand(m, 1)
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t[1:n])
+@variable(model, x[1:n])
+@objective(model, Max, sum(t))
+@constraint(model, sum(x) == 1)
+@constraint(model, A * x .<= b)
+@constraint(model, [i = 1:n], [t[i], x[i], 1] in MOI.ExponentialCone())
+optimize!(model)
+objective_value(model)
+
+# The [`MOI.ExponentialCone`](@ref) has a dual, the [`MOI.DualExponentialCone`](@ref),
+# that offers an alternative formulation that can be more efficient for some
+# formulations.
+
+# There is also the [`MOI.RelativeEntropyCone`](@ref) for explicitly encoding
+# the relative entropy function
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, x[1:n])
+@objective(model, Max, -t)
+@constraint(model, sum(x) == 1)
+@constraint(model, A * x .<= b)
+@constraint(model, [t; ones(n); x] in MOI.RelativeEntropyCone(2n + 1))
+optimize!(model)
+objective_value(model)
+
+# ## PowerCone
+
+# The [`MOI.PowerCone`](@ref) is a three-dimensional set parameterized by a
+# scalar value `α`. It has the form:
+# ```math
+# K_{p} = \{ (x,y,z) \in \mathbb{R}^3 : x^{\alpha} y^{1-\alpha} \ge |z|, x \ge 0, y \ge 0 \}
+# ```
+
+# The power cone permits a number of reformulations. For example, when ``p > 1``,
+# we can model ``t \ge x^p`` using the power cone ``(t, 1, x)`` with
+# ``\alpha = 1 / p``. Thus, to model ``t \ge x^3`` with ``x \ge 0``
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, x >= 1.5)
+@constraint(model, [t, 1, x] in MOI.PowerCone(1 / 3))
+@objective(model, Min, t)
+optimize!(model)
+value(t), value(x)
+
+# The [`MOI.PowerCone`](@ref) has a dual, the [`MOI.DualPowerCone`](@ref),
+# that offers an alternative formulation that can be more efficient for some
+# formulations.
+
+# ## P-Norm
+
+# The p-norm ``||x||_p = \left(\sum\limits_{i} |x_i|^p\right)^{\frac{1}{p}}``
+# can be modeled using [`MOI.PowerCone`](@ref)s. See the [Mosek Modeling Cookbook](https://docs.mosek.com/modeling-cookbook/powo.html#p-norm-cones)
+# for the derivation.
+
+function p_norm(x::Vector, p)
+    N = length(x)
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, r[1:N])
+    @variable(model, t)
+    @constraint(model, [i = 1:N], [r[i], t, x[i]] in MOI.PowerCone(1 / p))
+    @constraint(model, sum(r) == t)
+    @objective(model, Min, t)
+    optimize!(model)
+    return value(t)
+end
+
+x = rand(5);
+LinearAlgebra.norm(x, 4), p_norm(x, 4)
+
+# ## Positive Semidefinite Cone
+
+# The set of positive semidefinite matrices (PSD) of dimension $n$ form a cone
+# in $\mathbb{R}^n$. We write this set mathematically as:
+
+# ```math
+# \mathcal{S}_{+}^n = \{ X \in \mathcal{S}^n \mid z^T X z \geq 0, \: \forall z\in \mathbb{R}^n \}.
+# ```
+
+# A PSD cone is represented in JuMP using the MOI sets
+# `PositiveSemidefiniteConeTriangle` (for upper triangle of a PSD matrix) and
+# `PositiveSemidefiniteConeSquare` (for a complete PSD matrix). However, it is
+# preferable to use the `PSDCone` shortcut as illustrated below.
+
+# #### Example: largest eigenvalue of a symmetric matrix
+
+# Suppose $A$ has eigenvalues $\lambda_{1} \geq \lambda_{2} \ldots \geq \lambda_{n}$.
+# Then the matrix $t I-A$ has eigenvalues $t-\lambda_{1}, t-\lambda_{2}, \ldots, t-\lambda_{n}$.
+# Note that $t I-A$ is PSD exactly when all these eigenvalues are non-negative,
+# and this happens for values $t \geq \lambda_{1}$. Thus, we can model the
+# problem of finding the largest eigenvalue of a symmetric matrix as:
+
+# ```math
+# \begin{aligned}
+# \lambda_{1} = \min t \\
+# \text { s.t. } t I-A \succeq 0
+# \end{aligned}
+# ```
+
+A = [3 2 4; 2 0 2; 4 2 3]
+I = Matrix{Float64}(LinearAlgebra.I, 3, 3)
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@objective(model, Min, t)
+@constraint(model, t .* I - A in PSDCone())
+optimize!(model)
+objective_value(model)
+
+# ## GeometricMeanCone
+
+# The [`MOI.GeometricMeanCone`](@ref) is a cone of the form:
+# ```math
+# K_{geo} = \{ (t, x) \in \mathbb{R}^n : x \ge 0, t \le \sqrt[n-1]{x_1 x_2 \cdots x_{n-1}} \}
+# ```
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x[1:4])
+@variable(model, t)
+@constraint(model, sum(x) == 1)
+@constraint(model, [t; x] in MOI.GeometricMeanCone(5))
+optimize!(model)
+value(t), value.(x)
+
+# ## RootDetCone
+
+# The [`MOI.RootDetConeSquare`](@ref) is a cone of the form:
+# ```math
+# K = \{ (t, X) \in \mathbb{R}^{1+d^2} : t \le \det(X)^{\frac{1}{d}} \}
+# ```
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, X[1:2, 1:2])
+@objective(model, Max, t)
+@constraint(model, [t; vec(X)] in MOI.RootDetConeSquare(2))
+@constraint(model, X .== [2 1; 1 3])
+optimize!(model)
+value(t), sqrt(LinearAlgebra.det(value.(X)))
+
+# If `X` is symmetric, then you can use [`MOI.RootDetConeTriangle`](@ref)
+# instead. This can be more efficient because the solver does not need to add
+# additional constraints to ensure `X` is symmetric.
+
+# When forming the function, use [`triangle_vec`](@ref) to obtain the
+# column-wise upper triangle of the matrix as a vector in the order that JuMP
+# requires.
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, X[1:2, 1:2], Symmetric)
+@objective(model, Max, t)
+@constraint(model, [t; triangle_vec(X)] in MOI.RootDetConeTriangle(2))
+@constraint(model, X .== [2 1; 1 3])
+optimize!(model)
+value(t), sqrt(LinearAlgebra.det(value.(X)))
+
+# ## LogDetCone
+
+# The [`MOI.LogDetConeSquare`](@ref) is a cone of the form:
+# ```math
+# K = \{ (t, u, X) \in \mathbb{R}^{2+d^2} : t \le u \log(\det(X / u)) \}
+# ```
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, u)
+@variable(model, X[1:2, 1:2])
+@objective(model, Max, t)
+@constraint(model, [t; u; vec(X)] in MOI.LogDetConeSquare(2))
+@constraint(model, X .== [2 1; 1 3])
+@constraint(model, u == 0.5)
+optimize!(model)
+value(t), 0.5 * log(LinearAlgebra.det(value.(X) ./ 0.5))
+
+# If `X` is symmetric, then you can use [`MOI.LogDetConeTriangle`](@ref)
+# instead. This can be more efficient because the solver does not need to add
+# additional constraints to ensure `X` is symmetric.
+
+# When forming the function, use [`triangle_vec`](@ref) to obtain the
+# column-wise upper triangle of the matrix as a vector in the order that JuMP
+# requires.
+
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, u)
+@variable(model, X[1:2, 1:2], Symmetric)
+@objective(model, Max, t)
+@constraint(model, [t; u; triangle_vec(X)] in MOI.LogDetConeTriangle(2))
+@constraint(model, X .== [2 1; 1 3])
+@constraint(model, u == 0.5)
+optimize!(model)
+value(t), 0.5 * log(LinearAlgebra.det(value.(X) ./ 0.5))
+
+# ## Other Cones and Functions
+
+# For other cones supported by JuMP, check out the
+# [MathOptInterface Manual](https://jump.dev/MathOptInterface.jl/stable).
diff --git a/previews/PR3525/tutorials/conic/tips_and_tricks/index.html b/previews/PR3525/tutorials/conic/tips_and_tricks/index.html
new file mode 100644
index 00000000000..7903be71efb
--- /dev/null
+++ b/previews/PR3525/tutorials/conic/tips_and_tricks/index.html
@@ -0,0 +1,162 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Tips and Tricks · JuMP</title><meta name="title" content="Tips and Tricks · JuMP"/><meta property="og:title" content="Tips and Tricks · JuMP"/><meta property="twitter:title" content="Tips and Tricks · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox" checked/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../start_values/">Primal and dual warm-starts</a></li><li class="is-active"><a class="tocitem" href>Tips and Tricks</a><ul class="internal"><li><a class="tocitem" href="#Background-theory"><span>Background theory</span></a></li><li><a class="tocitem" href="#Second-Order-Cone"><span>Second-Order Cone</span></a></li><li><a class="tocitem" href="#Rotated-Second-Order-Cone"><span>Rotated Second-Order Cone</span></a></li><li><a class="tocitem" href="#Exponential-Cone"><span>Exponential Cone</span></a></li><li><a class="tocitem" href="#PowerCone"><span>PowerCone</span></a></li><li><a class="tocitem" href="#P-Norm"><span>P-Norm</span></a></li><li><a class="tocitem" href="#Positive-Semidefinite-Cone"><span>Positive Semidefinite Cone</span></a></li><li><a class="tocitem" href="#GeometricMeanCone"><span>GeometricMeanCone</span></a></li><li><a class="tocitem" href="#RootDetCone"><span>RootDetCone</span></a></li><li><a class="tocitem" href="#LogDetCone"><span>LogDetCone</span></a></li><li><a class="tocitem" href="#Other-Cones-and-Functions"><span>Other Cones and Functions</span></a></li></ul></li><li><a class="tocitem" href="../simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../dualization/">Dualization</a></li><li><a class="tocitem" href="../arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Conic programs</a></li><li class="is-active"><a href>Tips and Tricks</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Tips and Tricks</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/conic/tips_and_tricks.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="conic_tips_and_tricks"><a class="docs-heading-anchor" href="#conic_tips_and_tricks">Tips and Tricks</a><a id="conic_tips_and_tricks-1"></a><a class="docs-heading-anchor-permalink" href="#conic_tips_and_tricks" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../tips_and_tricks.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Arpit Bhatia.</strong></p><p>This tutorial is aimed at providing a simplistic introduction to conic programming using JuMP.</p><p>It uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import SCS
+import LinearAlgebra</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This tutorial uses sets from <a href="../../../moi/#moi_documentation">MathOptInterface</a>. By default, JuMP exports the <code>MOI</code> symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre></div></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>A good resource for learning more about functions which can be modeled using cones is the <a href="https://docs.mosek.com/modeling-cookbook/index.html">MOSEK Modeling Cookbook</a>.</p></div></div><h2 id="Background-theory"><a class="docs-heading-anchor" href="#Background-theory">Background theory</a><a id="Background-theory-1"></a><a class="docs-heading-anchor-permalink" href="#Background-theory" title="Permalink"></a></h2><p>A subset <span>$C$</span> of a vector space <span>$V$</span> is a cone if <span>$\forall x \in C$</span> and positive scalars <span>$\lambda &gt; 0$</span>, the product <span>$\lambda x \in C$</span>.</p><p>A cone <span>$C$</span> is a convex cone if <span>$\lambda x + (1 - \lambda) y \in C$</span>, for any <span>$\lambda \in [0, 1]$</span>, and any <span>$x, y \in C$</span>.</p><p>Conic programming problems are convex optimization problems in which a convex function is minimized over the intersection of an affine subspace and a convex cone. An example of a conic-form minimization problems, in the primal form is:</p><p class="math-container">\[\begin{aligned}
+&amp; \min_{x \in \mathbb{R}^n} &amp; a_0^T x + b_0 \\
+&amp; \;\;\text{s.t.} &amp; A_i x + b_i &amp; \in \mathcal{C}_i &amp; i = 1 \ldots m
+\end{aligned}\]</p><p>The corresponding dual problem is:</p><p class="math-container">\[\begin{aligned}
+&amp; \max_{y_1, \ldots, y_m} &amp; -\sum_{i=1}^m b_i^T y_i + b_0 \\
+&amp; \;\;\text{s.t.} &amp; a_0 - \sum_{i=1}^m A_i^T y_i &amp; = 0 \\
+&amp; &amp; y_i &amp; \in \mathcal{C}_i^* &amp; i = 1 \ldots m
+\end{aligned}\]</p><p>where each <span>$\mathcal{C}_i$</span> is a closed convex cone and <span>$\mathcal{C}_i^*$</span> is its dual cone.</p><h2 id="Second-Order-Cone"><a class="docs-heading-anchor" href="#Second-Order-Cone">Second-Order Cone</a><a id="Second-Order-Cone-1"></a><a class="docs-heading-anchor-permalink" href="#Second-Order-Cone" title="Permalink"></a></h2><p>The <a href="../../../api/JuMP/#JuMP.SecondOrderCone"><code>SecondOrderCone</code></a> (or Lorentz Cone) of dimension <span>$n$</span> is a cone of the form:</p><p class="math-container">\[K_{soc} = \{ (t, x) \in \mathbb{R}^n : t \ge ||x||_2 \}\]</p><p>It is most commonly used to represent the L2-norm of the vector <span>$x$</span>:</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x[1:3])
+@variable(model, t)
+@constraint(model, sum(x) == 1)
+@constraint(model, [t; x] in SecondOrderCone())
+@objective(model, Min, t)
+optimize!(model)
+value(t), value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(0.5773503148525505, [0.3333333333638094, 0.33333333336381227, 0.33333333336381227])</code></pre><h2 id="Rotated-Second-Order-Cone"><a class="docs-heading-anchor" href="#Rotated-Second-Order-Cone">Rotated Second-Order Cone</a><a id="Rotated-Second-Order-Cone-1"></a><a class="docs-heading-anchor-permalink" href="#Rotated-Second-Order-Cone" title="Permalink"></a></h2><p>A Second-Order Cone rotated by <span>$\pi/4$</span> in the <span>$(x_1,x_2)$</span> plane is called a <a href="../../../api/JuMP/#JuMP.RotatedSecondOrderCone"><code>RotatedSecondOrderCone</code></a>. It is a cone of the form:</p><p class="math-container">\[K_{rsoc} = \{ (t,u,x) \in \mathbb{R}^n : 2tu \ge ||x||_2^2, t,u \ge 0 \}\]</p><p>When <code>u = 0.5</code>, it represents the sum of squares of a vector <span>$x$</span>:</p><pre><code class="language-julia hljs">data = [1.0, 2.0, 3.0, 4.0]
+target = [0.45, 1.04, 1.51, 1.97]
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, θ)
+@variable(model, t)
+@expression(model, residuals, θ * data .- target)
+@constraint(model, [t; 0.5; residuals] in RotatedSecondOrderCone())
+@objective(model, Min, t)
+optimize!(model)
+value(θ), value(t)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(0.4980000000000055, 0.004979422159620655)</code></pre><h2 id="Exponential-Cone"><a class="docs-heading-anchor" href="#Exponential-Cone">Exponential Cone</a><a id="Exponential-Cone-1"></a><a class="docs-heading-anchor-permalink" href="#Exponential-Cone" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.ExponentialCone"><code>MOI.ExponentialCone</code></a> is a set of the form:</p><p class="math-container">\[K_{exp} = \{ (x,y,z) \in \mathbb{R}^3 : y \exp (x/y) \le z, y &gt; 0 \}\]</p><p>It can be used to model problems involving <code>log</code> and <code>exp</code>.</p><h3 id="Exponential"><a class="docs-heading-anchor" href="#Exponential">Exponential</a><a id="Exponential-1"></a><a class="docs-heading-anchor-permalink" href="#Exponential" title="Permalink"></a></h3><p>To model <span>$\exp(x) \le z$</span>, use <code>(x, 1, z)</code>:</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x == 1.5)
+@variable(model, z)
+@objective(model, Min, z)
+@constraint(model, [x, 1, z] in MOI.ExponentialCone())
+optimize!(model)
+value(z), exp(1.5)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(4.481654619603649, 4.4816890703380645)</code></pre><h3 id="Logarithm"><a class="docs-heading-anchor" href="#Logarithm">Logarithm</a><a id="Logarithm-1"></a><a class="docs-heading-anchor-permalink" href="#Logarithm" title="Permalink"></a></h3><p>To model <span>$x \le \log(z)$</span>, use <code>(x, 1, z)</code>:</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x)
+@variable(model, z == 1.5)
+@objective(model, Max, x)
+@constraint(model, [x, 1, z] in MOI.ExponentialCone())
+optimize!(model)
+value(x), log(1.5)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(0.4055956586655662, 0.4054651081081644)</code></pre><h3 id="Log-sum-exp"><a class="docs-heading-anchor" href="#Log-sum-exp">Log-sum-exp</a><a id="Log-sum-exp-1"></a><a class="docs-heading-anchor-permalink" href="#Log-sum-exp" title="Permalink"></a></h3><p>To model <span>$t \ge \log\left(\sum e^{x_i}\right)$</span>, use:</p><pre><code class="language-julia hljs">N = 3
+x0 = rand(N)
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x[i = 1:N] == x0[i])
+@variable(model, t)
+@objective(model, Min, t)
+@variable(model, u[1:N])
+@constraint(model, sum(u) &lt;= 1)
+@constraint(model, [i = 1:N], [x[i] - t, 1, u[i]] in MOI.ExponentialCone())
+optimize!(model)
+value(t), log(sum(exp.(x0)))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(1.4726849472138828, 1.472772274699489)</code></pre><h3 id="Entropy"><a class="docs-heading-anchor" href="#Entropy">Entropy</a><a id="Entropy-1"></a><a class="docs-heading-anchor-permalink" href="#Entropy" title="Permalink"></a></h3><p>The entropy maximization problem consists of maximizing the entropy function, <span>$H(x) = -x\log{x}$</span> subject to linear inequality constraints.</p><p class="math-container">\[\begin{aligned}
+&amp; \max &amp; - \sum_{i=1}^n x_i \log x_i \\
+&amp; \;\;\text{s.t.} &amp; \mathbf{1}^\top x = 1 \\
+&amp; &amp; Ax \leq b
+\end{aligned}\]</p><p>We can model this problem using an exponential cone by using the following transformation:</p><p class="math-container">\[t\leq -x\log{x} \iff t\leq x\log(1/x)  \iff (t, x, 1) \in K_{exp}\]</p><p>Thus, our problem becomes,</p><p class="math-container">\[\begin{aligned}
+&amp; \max &amp; 1^Tt \\
+&amp; \;\;\text{s.t.} &amp; Ax \leq b \\
+&amp; &amp; 1^T x = 1 \\
+&amp; &amp; (t_i, x_i, 1) \in K_{exp} &amp;&amp; \forall i = 1 \ldots n \\
+\end{aligned}\]</p><pre><code class="language-julia hljs">m, n = 10, 15
+A, b = randn(m, n), rand(m, 1)
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t[1:n])
+@variable(model, x[1:n])
+@objective(model, Max, sum(t))
+@constraint(model, sum(x) == 1)
+@constraint(model, A * x .&lt;= b)
+@constraint(model, [i = 1:n], [t[i], x[i], 1] in MOI.ExponentialCone())
+optimize!(model)
+objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2.663206172203935</code></pre><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.ExponentialCone"><code>MOI.ExponentialCone</code></a> has a dual, the <a href="../../../moi/reference/standard_form/#MathOptInterface.DualExponentialCone"><code>MOI.DualExponentialCone</code></a>, that offers an alternative formulation that can be more efficient for some formulations.</p><p>There is also the <a href="../../../moi/reference/standard_form/#MathOptInterface.RelativeEntropyCone"><code>MOI.RelativeEntropyCone</code></a> for explicitly encoding the relative entropy function</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, x[1:n])
+@objective(model, Max, -t)
+@constraint(model, sum(x) == 1)
+@constraint(model, A * x .&lt;= b)
+@constraint(model, [t; ones(n); x] in MOI.RelativeEntropyCone(2n + 1))
+optimize!(model)
+objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2.6631736431749626</code></pre><h2 id="PowerCone"><a class="docs-heading-anchor" href="#PowerCone">PowerCone</a><a id="PowerCone-1"></a><a class="docs-heading-anchor-permalink" href="#PowerCone" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.PowerCone"><code>MOI.PowerCone</code></a> is a three-dimensional set parameterized by a scalar value <code>α</code>. It has the form:</p><p class="math-container">\[K_{p} = \{ (x,y,z) \in \mathbb{R}^3 : x^{\alpha} y^{1-\alpha} \ge |z|, x \ge 0, y \ge 0 \}\]</p><p>The power cone permits a number of reformulations. For example, when <span>$p &gt; 1$</span>, we can model <span>$t \ge x^p$</span> using the power cone <span>$(t, 1, x)$</span> with <span>$\alpha = 1 / p$</span>. Thus, to model <span>$t \ge x^3$</span> with <span>$x \ge 0$</span></p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, x &gt;= 1.5)
+@constraint(model, [t, 1, x] in MOI.PowerCone(1 / 3))
+@objective(model, Min, t)
+optimize!(model)
+value(t), value(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(3.374754835774048, 1.4999552263153844)</code></pre><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.PowerCone"><code>MOI.PowerCone</code></a> has a dual, the <a href="../../../moi/reference/standard_form/#MathOptInterface.DualPowerCone"><code>MOI.DualPowerCone</code></a>, that offers an alternative formulation that can be more efficient for some formulations.</p><h2 id="P-Norm"><a class="docs-heading-anchor" href="#P-Norm">P-Norm</a><a id="P-Norm-1"></a><a class="docs-heading-anchor-permalink" href="#P-Norm" title="Permalink"></a></h2><p>The p-norm <span>$||x||_p = \left(\sum\limits_{i} |x_i|^p\right)^{\frac{1}{p}}$</span> can be modeled using <a href="../../../moi/reference/standard_form/#MathOptInterface.PowerCone"><code>MOI.PowerCone</code></a>s. See the <a href="https://docs.mosek.com/modeling-cookbook/powo.html#p-norm-cones">Mosek Modeling Cookbook</a> for the derivation.</p><pre><code class="language-julia hljs">function p_norm(x::Vector, p)
+    N = length(x)
+    model = Model(SCS.Optimizer)
+    set_silent(model)
+    @variable(model, r[1:N])
+    @variable(model, t)
+    @constraint(model, [i = 1:N], [r[i], t, x[i]] in MOI.PowerCone(1 / p))
+    @constraint(model, sum(r) == t)
+    @objective(model, Min, t)
+    optimize!(model)
+    return value(t)
+end
+
+x = rand(5);
+LinearAlgebra.norm(x, 4), p_norm(x, 4)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(0.9316922467512209, 0.931687599495156)</code></pre><h2 id="Positive-Semidefinite-Cone"><a class="docs-heading-anchor" href="#Positive-Semidefinite-Cone">Positive Semidefinite Cone</a><a id="Positive-Semidefinite-Cone-1"></a><a class="docs-heading-anchor-permalink" href="#Positive-Semidefinite-Cone" title="Permalink"></a></h2><p>The set of positive semidefinite matrices (PSD) of dimension <span>$n$</span> form a cone in <span>$\mathbb{R}^n$</span>. We write this set mathematically as:</p><p class="math-container">\[\mathcal{S}_{+}^n = \{ X \in \mathcal{S}^n \mid z^T X z \geq 0, \: \forall z\in \mathbb{R}^n \}.\]</p><p>A PSD cone is represented in JuMP using the MOI sets <code>PositiveSemidefiniteConeTriangle</code> (for upper triangle of a PSD matrix) and <code>PositiveSemidefiniteConeSquare</code> (for a complete PSD matrix). However, it is preferable to use the <code>PSDCone</code> shortcut as illustrated below.</p><h4 id="Example:-largest-eigenvalue-of-a-symmetric-matrix"><a class="docs-heading-anchor" href="#Example:-largest-eigenvalue-of-a-symmetric-matrix">Example: largest eigenvalue of a symmetric matrix</a><a id="Example:-largest-eigenvalue-of-a-symmetric-matrix-1"></a><a class="docs-heading-anchor-permalink" href="#Example:-largest-eigenvalue-of-a-symmetric-matrix" title="Permalink"></a></h4><p>Suppose <span>$A$</span> has eigenvalues <span>$\lambda_{1} \geq \lambda_{2} \ldots \geq \lambda_{n}$</span>. Then the matrix <span>$t I-A$</span> has eigenvalues <span>$t-\lambda_{1}, t-\lambda_{2}, \ldots, t-\lambda_{n}$</span>. Note that <span>$t I-A$</span> is PSD exactly when all these eigenvalues are non-negative, and this happens for values <span>$t \geq \lambda_{1}$</span>. Thus, we can model the problem of finding the largest eigenvalue of a symmetric matrix as:</p><p class="math-container">\[\begin{aligned}
+\lambda_{1} = \min t \\
+\text { s.t. } t I-A \succeq 0
+\end{aligned}\]</p><pre><code class="language-julia hljs">A = [3 2 4; 2 0 2; 4 2 3]
+I = Matrix{Float64}(LinearAlgebra.I, 3, 3)
+model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@objective(model, Min, t)
+@constraint(model, t .* I - A in PSDCone())
+optimize!(model)
+objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">8.000003377698652</code></pre><h2 id="GeometricMeanCone"><a class="docs-heading-anchor" href="#GeometricMeanCone">GeometricMeanCone</a><a id="GeometricMeanCone-1"></a><a class="docs-heading-anchor-permalink" href="#GeometricMeanCone" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.GeometricMeanCone"><code>MOI.GeometricMeanCone</code></a> is a cone of the form:</p><p class="math-container">\[K_{geo} = \{ (t, x) \in \mathbb{R}^n : x \ge 0, t \le \sqrt[n-1]{x_1 x_2 \cdots x_{n-1}} \}\]</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, x[1:4])
+@variable(model, t)
+@constraint(model, sum(x) == 1)
+@constraint(model, [t; x] in MOI.GeometricMeanCone(5))
+optimize!(model)
+value(t), value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(0.0, [0.25000000110304854, 0.2500000011030519, 0.2500000011030504, 0.25000000110305026])</code></pre><h2 id="RootDetCone"><a class="docs-heading-anchor" href="#RootDetCone">RootDetCone</a><a id="RootDetCone-1"></a><a class="docs-heading-anchor-permalink" href="#RootDetCone" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.RootDetConeSquare"><code>MOI.RootDetConeSquare</code></a> is a cone of the form:</p><p class="math-container">\[K = \{ (t, X) \in \mathbb{R}^{1+d^2} : t \le \det(X)^{\frac{1}{d}} \}\]</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, X[1:2, 1:2])
+@objective(model, Max, t)
+@constraint(model, [t; vec(X)] in MOI.RootDetConeSquare(2))
+@constraint(model, X .== [2 1; 1 3])
+optimize!(model)
+value(t), sqrt(LinearAlgebra.det(value.(X)))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(2.2361163647438964, 2.236067957402722)</code></pre><p>If <code>X</code> is symmetric, then you can use <a href="../../../moi/reference/standard_form/#MathOptInterface.RootDetConeTriangle"><code>MOI.RootDetConeTriangle</code></a> instead. This can be more efficient because the solver does not need to add additional constraints to ensure <code>X</code> is symmetric.</p><p>When forming the function, use <a href="../../../api/JuMP/#JuMP.triangle_vec"><code>triangle_vec</code></a> to obtain the column-wise upper triangle of the matrix as a vector in the order that JuMP requires.</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, X[1:2, 1:2], Symmetric)
+@objective(model, Max, t)
+@constraint(model, [t; triangle_vec(X)] in MOI.RootDetConeTriangle(2))
+@constraint(model, X .== [2 1; 1 3])
+optimize!(model)
+value(t), sqrt(LinearAlgebra.det(value.(X)))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(2.2361792479821303, 2.2360679863316997)</code></pre><h2 id="LogDetCone"><a class="docs-heading-anchor" href="#LogDetCone">LogDetCone</a><a id="LogDetCone-1"></a><a class="docs-heading-anchor-permalink" href="#LogDetCone" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.LogDetConeSquare"><code>MOI.LogDetConeSquare</code></a> is a cone of the form:</p><p class="math-container">\[K = \{ (t, u, X) \in \mathbb{R}^{2+d^2} : t \le u \log(\det(X / u)) \}\]</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, u)
+@variable(model, X[1:2, 1:2])
+@objective(model, Max, t)
+@constraint(model, [t; u; vec(X)] in MOI.LogDetConeSquare(2))
+@constraint(model, X .== [2 1; 1 3])
+@constraint(model, u == 0.5)
+optimize!(model)
+value(t), 0.5 * log(LinearAlgebra.det(value.(X) ./ 0.5))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(1.4979204458220523, 1.4978661006407135)</code></pre><p>If <code>X</code> is symmetric, then you can use <a href="../../../moi/reference/standard_form/#MathOptInterface.LogDetConeTriangle"><code>MOI.LogDetConeTriangle</code></a> instead. This can be more efficient because the solver does not need to add additional constraints to ensure <code>X</code> is symmetric.</p><p>When forming the function, use <a href="../../../api/JuMP/#JuMP.triangle_vec"><code>triangle_vec</code></a> to obtain the column-wise upper triangle of the matrix as a vector in the order that JuMP requires.</p><pre><code class="language-julia hljs">model = Model(SCS.Optimizer)
+set_silent(model)
+@variable(model, t)
+@variable(model, u)
+@variable(model, X[1:2, 1:2], Symmetric)
+@objective(model, Max, t)
+@constraint(model, [t; u; triangle_vec(X)] in MOI.LogDetConeTriangle(2))
+@constraint(model, X .== [2 1; 1 3])
+@constraint(model, u == 0.5)
+optimize!(model)
+value(t), 0.5 * log(LinearAlgebra.det(value.(X) ./ 0.5))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(1.4979204481049286, 1.497866101098124)</code></pre><h2 id="Other-Cones-and-Functions"><a class="docs-heading-anchor" href="#Other-Cones-and-Functions">Other Cones and Functions</a><a id="Other-Cones-and-Functions-1"></a><a class="docs-heading-anchor-permalink" href="#Other-Cones-and-Functions" title="Permalink"></a></h2><p>For other cones supported by JuMP, check out the <a href="https://jump.dev/MathOptInterface.jl/stable">MathOptInterface Manual</a>.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../start_values/">« Primal and dual warm-starts</a><a class="docs-footer-nextpage" href="../simple_examples/">Simple semidefinite programming examples »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/data/Cereal.txt b/previews/PR3525/tutorials/getting_started/data/Cereal.txt
new file mode 100644
index 00000000000..077f0a411b3
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/data/Cereal.txt
@@ -0,0 +1,24 @@
+Name Cups Calories Carbs Fat Fiber Potassium Protein Sodium Sugars
+CapnCrunch 0.75 120 12 2 0 35 1 220 12
+CocoaPuffs 1 110 12 1 0 55 1 180 13
+Trix 1 110 13 1 0 25 1 140 12
+AppleJacks 1 110 11 0 1 30 2 125 14
+CornChex 1 110 22 0 0 25 2 280 3
+CornFlakes 1 100 21 0 1 35 2 290 2
+Nut&Honey 0.67 120 15 1 0 40 2 190 9
+Smacks 0.75 110 9 1 1 40 2 70 15
+MultiGrain 1 100 15 1 2 90 2 220 6
+CracklinOat 0.5 110 10 3 4 160 3 140 7
+GrapeNuts 0.25 110 17 0 3 90 3 179 3
+HoneyNutCheerios 0.75 110 11.5 1 1.5 90 3 250 10
+NutriGrain 0.67 140 21 2 3 130 3 220 7
+Product19 1 100 20 0 1 45 3 320 3
+TotalRaisinBran 1 140 15 1 4 230 3 190 14
+WheatChex 0.67 100 17 1 3 115 3 230 3
+Oatmeal 0.5 130 13.5 2 1.5 120 3 170 10
+Life 0.67 100 12 2 2 95 4 150 6
+Maypo 1 100 16 1 0 95 4 0 3
+QuakerOats 0.5 100 14 1 2 110 4 135 6
+Muesli 1 150 16 3 3 170 4 150 11
+Cheerios 1.25 110 17 2 2 105 6 290 1
+SpecialK 1 110 16 0 1 55 6 230 3
diff --git a/previews/PR3525/tutorials/getting_started/data/SalesData.xlsx b/previews/PR3525/tutorials/getting_started/data/SalesData.xlsx
new file mode 100644
index 0000000000000000000000000000000000000000..b5d5606c032046f519c11d40b26edde4d4e62e32
GIT binary patch
literal 7463
zcmaKR1z6PI@;==l-O?eTG)PNJEuGS_bS@nWk}loS4bsvb(ji^a-2#HNlz;Gh|5vW}
zcb$En-Tgez&g_|Y=FFLSkD@Fr94-_R5)xFIP>(#+BSF3Yt`7oQJF+l8yqCtb%XhM(
z1s!^ZFkWyi3L|BGXp)m^qU0oNqp>Ba3C`*AKRy@7!A1+O?2K^m_PI04TavSVzQjr=
zs~4=M6%5ltXAI8o8n|<CXW<^o_U^&rd97?G#t4d3kU1S(75SX&=@cb{vU38E9fRIB
z-Zy~~6n+QNRdZ#(Gl6UcXzhn#JD90w&-lPK2LNoWxNN1#Nj-!()-<SMDP8(x<Z`_7
zd781&fk7zlOF(eL!17@n1`!&o6aXuMa_d_k{zP2_<!#)+Zek)%W5X86-sluzs2Mwe
z)y&BgKn<3X`<8+CltW)8^iBHBz!r|8EFw~1GH3F)dk}_Upr920Crmi^zc8^gRs`AE
zJF*x%J386fFuU1U*GvNJ{3!?#kFJFRS8XM<=jRFl#WpT8hSK`q(7w~EP`uS0syq}*
ztpPk~vOYr)y?f;bb-UTX7><KgX}}3}js)YcI2hw+3eB@$w?#^*OqrY+qlEo7NU9ty
z<3^*M2dju;d=yy~fkGpM<+<!ys;X4PDeDw90}h^YBNKaQo`tXUpuH|u(R0L(!a%D+
zP2jN=>CQ1e{n>P1z^aS$alKvrFzT(Wn{jSDV+@p6k1cL5vlGB<UE%}!LCtP>OA_)n
zCZ84koJa8x0jnGTEHr{d{L+t*6Jrhfr|!z3F+NhXqMm$y#z^43^Zq)I?-}Ryb=2w(
z;wk(*Myk<OMex!7E3stLQ;S$Sp`yC@igF#cpBlfj4j(j_fi1J2?=*#h#5rP|c1Vmj
z18*&a(9+M3WUa_9)4QdjeaEySzp*k(=jI?FY-_N-&IYr>dg>h->JEklUpF=J8Y|zq
zSe4FyqP;<Q2vd_&51T186jUbM-@=6XI8282_75>ii&9XyPwDPMuMoz$8o(P7E`S9c
z;Tk702E0SD=vP~jA#&Pplir>%5F0bw+KjWR<@Nd5uvui$$zkr?=Y;)jjMG-0=6zPz
zab?o@N^^P<!=hsN5ECZbOIdp|7)yv2)yzy3b@s9%@f22Am_&??G6F$Z*JKc^b7v-=
z03R1oxfD~6B?^b%LDtAEk(76E3eFwD9RJy^zLyn2hMQlV!Ob%)xcr1l=V!g@B$&^`
zN6<fAcu&e0$%yZty?(o$fphxtS!iQ{x~ENkgQxzqE;vQ^g;ReA{j*$J4(l`4-HH(=
z`7#Hu?XUzFEPgXKtqd4(`JX|lddhhEzmVf#DZaVKCHWq`|DE7ikC8LB0|6hP1IH@a
z%~Rk8E?+QV9%8sOOFoq;sLqRLES<>x@U{A7h}~25HYBOEn``KZ#cE5YAMIPI*A5Mc
zEG*}5vdrxDTucLtwSut-$GFC;nY`*67>C)8eN%#phucud+tyLffeg>_()^b{R;~EK
zMN;*-G47+-iIvA;veafaWJtUVUNhe2eDZywrVI4FOPk`PQU?>=_oaY+%-g^;$R)(O
zE1;S`pFI-`4%eKuJ=Rh^n!pY!|F#KUi`T)D2m^nqCSq-4$&;7px;5j>xwhwy>3Dl<
z+eXNe5<naeHTb?)vcSF#mh&0RPWfWN!t|>zDNKV@>sCtXc)*&%VT+nfv(Qhn#~*~C
z-(3t>a0&24e9@^v$z7&9$)(kY)#I@c3yWI>+2B!(sD%|c0_2FhX-0=(V-VnvC?&(0
zFJN$G7)<6EC?#&;-U;D*p09=W$&p#g3l_q61DZmSIas^T$_7*4V=Fq4+pfLvj2rcO
zRgUJ?#2q0_J}s1Hcvkw1;QOp9SK4$nvw|rP%<(WrI<RryI{jsukjk|8#nBUS!gt)H
zE`k_9)mV<c%oW(>OyJd?e!wl_L(Ey0vCeAl<82H3x0n;($K1`D<)8BJY6r41va_>#
zi2FZBK{UYX*HLhbCwb8=jc6Rk(-J09c6$lkqqH{K#U1md=BGC^i8cywe>8S<`lMCr
z>?>#ZS4fu_8bVPvTwr^*3sQ^k?YDzfBK4u{r4@79$Z~S*;%1zgw*=BGZ>mmsw3v-L
zqWE=l`(@+{r~4cI50qF2j-@A3p@;BPO0{9DZJ89C96NrN8&9KS=;wH5u=b%N4}r*{
zImVPDjaHag*bz@{eryVEzM^_ML;bnTKSP<@6?sbEl<d$0)ed>SRMC6?hUVnBE=i)a
zY-!O8<^dUnqaPmQ_kPp<wcJ1X2bss1gsaShQdlv;hZ?LNbqzBPHjxfJ-KH@~wYyyw
z$IlJeiUPZ&_O$tLFPJ-WcdMU(p)6QZ1<hQw7SgSDpSEjesSS1*wh1UUFbB+a@9(Mb
zC9;oDX}~AE5gBAQm_0aNT+1C2n(#OryZGEwB5J+WYIlNSY5T4bS|VCGC7C+ObbG!B
zip3iKZOv&8MdBqFL5MSw<H;++#^(|kR%zsxiZi~1QKZ2wrQ*i3ASD#!ku-|PmqTWH
zI0pd(sXqoZ=Jdl7zI*wYD}RpA>d%A6ezQ9vc8(6Bk<gLZ!Xu?jg<luie1a4nCk^bI
z<x7W8-9L1*;_KqK>2fWN8nCggj$;#ZtXHSbes=65QEQWv?q=$_{X?QaG{Z=ey0N(-
zW8q8!S91)CK4-W7V$pX-jj+msCp7V$fTuMHUe&C1SF<u<=tqdyYVHP-YFUj`)KAIi
z6fGZ*&s<Iev@?h*iaFY?xb-Xr&D!|6mr9z^(%Cj!elT0D5K`+%iDXA&KVjCmj2(X}
zGDBfGw73i>5lkl0rXGEIr<!SsXq*W{OgO_(RI?lNv9xJ$O!7CMgE-tIU)#_HQMN+^
z@j>tA4DnbKzF>qgMpE<+dq<8-R^nx}aX_;2b&NB+tj{=zcam37x_3D{>(wUK=Yv25
zuZY>LRDT^_d7(UKyVpa91jav5AJRu?<R}~FK6RY`y0LKoNgpReBWs`|%dZc%hka!e
zy(r%yj2U<c4jEb)z*kiqf1gMk_A;45piSnTOYE(M3qo|AD^ua<=m78bd{vd%M0X_D
z8VRh$Av<9tHs(%k>*UAJ232}XLVgtLA60w!2Uj3)p*M8EGoWHuX9RmeQc6Qyb?F!*
z@ec%DCl?{jZU?mZT)IrgLP-;=4I1eW<BSKwJM2p+9SkI$(>CU*GJtTU-ivfX`-Jur
zw>v&>7^PlLG$x;CH%s63;XKuYhLu5-;4gP8Zx|G&(r=YXdPZdjRoq>UCr(i8&c8&@
zq{EnvPFxrJ1k{g&8o{{(0DP8#YWwa;4gAzYQ9jE*?801-+;-lEPsoXNoyK4W(vxMI
z{>&JAKaG7{8D=}^1O0&MbA~k^?E7NH_t!0n{tu>(PVV=ZK7e_kH6}mLiVI%W(A<Lm
zAQJQO1?xC2!q@6|0CQXmbGbemd8^f#=ylP7>5+(7PKG@M31MM(kQWI2tv~qSirdzQ
zSjj$ILX>Au#B3l4iX#ItAl&1~iDzcPmm(Sh4*>;v+6UpxTr4#Nrz&}|_h`a&x5bc=
z$`bihmUNM?BhhLL^vEqDH$b9Pb=fLoG!1ZRJQT%C-`lvBB4Hs2CIGTUXmD0ekYfdB
zm?;dB%Y4Vb*oYbIV7O=IdbOjWif5fougG)F=wW{;4FXq1lvGu<+|rKCPwj-BT^Jm6
z&Ly`1=LtwW8$DaoVYi9OXAko)&RMuQi{k~0w}VpE$hBb8Qx^QJ#MQXz@qDSQuiimo
zWJNd$vmv8NzS_h`y;3w^)e?Ozh{bA!Ei@ROz_H!MBOlD6DUjIUW>vEm!K1G<u*G1b
zaw|TK(tTo9;qgLlNL4Jh=a$}sW^~ZPSq64oa<gTFN|A|rWWS9R23mxA7*jCZlU1Cv
zz5H#&CHjEKNFGmIh-TE7W?fA0d%EbD*I%i~b5uyS)<p9z*z)K-l5Wwq73`SCE!<qS
z%ByqK%f6)R5`QT1UrQWD(212|YbQG5MK2Uaj0@8YAb^R4Pnb#AWsdGh4?PxU-SS&t
zw7i9RJzF%H5<=4&G<YIP@m=+91)B&zZ&xzYt)Q(#yu**PqA?*1-cUnKh@#sDsG`#c
z6?d({(<c!jof-6=s??%>B~I|7Uu!B~ZaZ5ESii3-gsvBvs7`GfoESf?PW=IE0%Fm{
zX{3%7`?N7baySN7{^J&{Jdf*BbAD$&rt0|`<%y!$<V6B}<9JyF72!Bb#9pl$m&72~
zyaaeb<!dV@8ZzT~$7#LRkoN3~67Kgvv5e&^`a5K6y17sEZ@wv@Ki!+0V7pz`dU0E{
zi)M9GRHgr9_M+*Gvt+pG?lwf+IBCMYICdYMP`y~%6iBPOYb>WE?`((+e^sUa#Cw_?
z(Za$)PaY{g04V4aQ*gCe)h^S3Z-L>Cn^pOPHaYT1F7&7UWq7qZ(J3(0lsMAhjaGZL
zuA$|t)WU45{Ek|%Pp$Fu)BR@<7bfwvP4N?Dtri`tJFfX_*oSSm{ZS)F;J%DulmD$p
zp#FUlI+_E4PLBU-7k)kbDr&!)h49D$lnz!*i3@Lu#FnHlfe_9E$pG6tZ(+H7iwgHB
z@v%=YGf`2Qb%w#ItLj6YgZuK7^1~?4+@ZrtK({6}ZyjBA4l@firh-kAKr|7tk<VSN
zaaliXTtJk<O-we1V~2)U{Gu5$L~;_oGIFhccvef;N~Fk=SaJX#P|M=$tf26-e34Zy
z1VeriZT55>p3Xgq0TwtK2WNk$E+pB{fyo-#+1+>r|KRtVhAZ{_d#`Z++7A={m*2lV
zZb{a4%;CljI=p5gWcJVqyENUapO|rI<gloku%1s^!lLdSG!hRawL72Iz&2QS+^DHb
zGi0eG)QgB9dm5mpJ&_?~C%YGa{J~%<96&l<YhGjIuGF$H<Kv}s7d4|!UCB5iqdrD*
z9nO8OP<8mhv*#5TiA*3Ztm;sB3s$!G%OuWX^hML~PRlp_OjbR0Nm1lK2678|hd|(o
z+EllXB!F5irCFN**PCQQ>lS8pr*`$G7p3AP0Tb*v?6;;Wli`9XU-5<Pt#sr3j`g*e
zC-lsKVw*L5o>>a2H%fbZYt+{i3a1J5RlB)O?%YGDvwqn|4vno9K6HoQq>E_Yd^Z!h
z7+RT`BTySYC+ER&qUDrrcU7&NFTYiGG0-ptEJp~rkA&W273}QJF3)-of7Yj==cVZ+
z^idhcpm+OT8Qt(|w1#KlMdddAo6kKKLd}N`^fI%bC+5$S;@?5>>Y6nv7m%7bhFxsy
z12t^-WuRGrQtHKYTh-6;pyuB&9C#AFLb2XMHJ;B`Q<w4Taz@PXnW*cmLOB&X?{P`i
za8*+q<1o&cWWbxA5>r*$J5kZa7fmZ)Y7~RPIksksEzt*>t0j%dA=Ro)J9ndSm862R
z;>kF}CBsciyLEGgIx0eYM{I=DPZVC>ePHe6R-kUsN8Kco1)O=gEf7p86SS0!+ZLD@
z-iS_>%VMU-2DtdBu%rSucQ~)gi1XQtNz{;aovlLz@~2kUxvxv^yAYY5%aE6X5g_pp
z9qSD&hqW2>g#htZg1}ap%d0*;f>{o6d6Q!{<mPX7r#8{I%$0VKRK8%Dsc^_PyvUV}
zuRpWB0TL;H|4#_1OLV80i5(`dwK1JVYEa>YO>_=3$b?vUT3Tn#W0BB%9jZrnV+O+O
z)cKQCYiR~n8l(MI#NJMmjQA|~g|+TMsGLa^gPCXOUaESNq|n;~4tr}2K^jWSPgCa#
zO0F?AC+o1_ygY9yrI#j$s>YTeb0=_L%9PcX^Oih)8DopQ!#rb$uU{(TmgZ_2BYr6r
z@v|G1Zrx%CV#tbWMHgF#O1Gf8IAXW>xFE&3q{RMiHFlc%l*<skEgWkdBfFFkLSh9K
zbfO>}1$?4*;LAQ*c-(|inx|5bSU;W%UFL6!@wpKs$-dV)Fc*L}K}Cl;p$=)`-4S$b
z#cNb@O8U&pL$59YIP@3bL4od3Fq$A>vkPcS7WP?ymF>xj4z`Y|ZPeRnZ8(Af0pULI
zQS%Gf0D-nHSh1wuLL7M0jn#raF`Q$}LE^eq;06M6q89-rvm>kBLLj685u_EQ5$_Jk
zdguMMtCh%ulkv0<P7tBqmRb&Xa(RIQx*mV5n=kTc5@t_8bWku%fPK;(m%lxUFY>hn
z8A+5U<xYo<*Z13CyrWwp{q#X&aRI+r_b>>3qMJ4IF-Kdhk}o~o3Y$(jO#(f>KOB_;
z^A|V2gy&Iipdp=fgSGK9o-R3or}1u<ROuafvK|#(6@E}?p{Jnmd^b8orT4pBp<R=T
z>lHW|lpmdpiOHwi1VUE!s$#)4g8*X9?KIQAMOcS#d_(<C7#iDYTLI16aG2nX@hEVd
zxV(pHg_NXAS{K|a-?B0JbLUG0be&wBeKa;NzZFEX0EMH0N!LvO#zPf8_q0aaom{h(
z$i3IW{ZT#4zDE_y;-Q4G2p^4*T=K~7A6d>QJ@;i>2%}$Kj(72EoUAu08&&49r9QLm
z$7`L;^<{GLOJne=!rqr#4M`(REreZ0xaiuDdwMbGmln1s7u=uR<Biln5=PJH>W>i5
z?~>ogh_$b<dTc3OF2dDp6x^Q`1cD4dr|GZ9k53e^B=v}DY4@c%*IdskBSPoKW41*)
zFC67y>35clCro9ew8ZoIc-(td)McW+j>JcARGw|#(@uODIOnHxUJOT~&g1X`*`?$I
zL$8SzPOsoeI1ABqC3dV?IF5N1Qf+QxvxtA6zM&4X%SZdPYMPlw;xJq`_vtq2;ZoZk
zuClHTJsbiL22dcF%L3EG+JMc&KTtv~Qn19~Gh4l8OxNy@4)W<f?ATbqiqZ9Z!Z60{
zNG`X}4IM++{%HUT@V*%!(2lj*vN2My&71vFt!ztzmdvUpzH@F=?m)x>!$g17JGe86
zp79DDjie`(r371xH$Ohp>&R+@X){}M6*bmHqpys}hry&XZ%K8P&tc}EYUdS_x>e7C
z-QAUa&V|J@5h<y)pm#pCHKuY-30L?g$xU-bRmsQsYqHlVmK-y}70hnblib0}F7syK
zvc+#oKZ>uV>#p5q5Dl~$_o=6OH>~x+3!`6^&Q-opvY5QAa-o+Qs$bEfw6>ZaeG<aH
z8S0|l9=+w5R4R0<v*}}l;xL0{yeUAk&!#n2nOapk=P^-WLxnd*nZWw}uALXHC!5*U
zTHm>nJa9b;uNbvxKm%kM%*-Xg#PY?8S2=zZoz(HG!mF(xi+kULQvq7(b=qbohUgKS
zh9nN>;M!6jFU^Ad*eTAs^yQKUxhQXZetD8Ta1E&1q;>f)ZR&f7z|vRLPkWzGpSFhb
zrlG~A10>E%T$=31&>~*keT>5EUEXd?iOn(x&B$_H6Fi*rcy@7Lvk{@7B5D8D-~4M-
zb2K*u0Zmk#Ko+)Uzov`uIKTsI7<70O^33(UjVMm?oP7EB7jQ-?Jg-Z&sEPBt^bmiN
z!6u|dMM|e%)}&1{;Ad2Dz?z^S%kmS2mqM4Q;Mi--9c*2lE)W|<h96>!T6}uDpUuYr
zndwu@9j_g{UEnR!hNMdtBF4qm#pRf3ZKdT1`P&u1ugo;;x(PUrAy$8o_uLpCru(wb
z=CyD5qmE6mN=+n-(OY6`^7>6REs6+Po4P`SZ}Tx);Chk$tBqHusQbBmv~$Jv{6PgU
zb%cqNxmt8acrDWV{CMu0@sx*~zRrb0p%Q-bVB|inq+Kz>%!#Z7=x*r7N-zAEd#NCt
zxT|q!ybF}iTV*O_Fm!m=&p0-nuj+OwHmzjYtww=OFdoyh)qo26%_Cid*j0NI5&BPZ
zXXp=LU+>kHG2J&*Ie(>RA6Rs4F*{o)pska>s=GbVQTJhzYl%~^f?>n#KJ*TuovtJ~
z$cn2hM$GC}FAz*f470UAc_&wWb-=GxR~qLR5}FVY<$m}w36X)Gf5Jo?sUWR%rhu>#
zt;9)YUafm+d*laoN*Lm%*C8}VQH^E=pXARj-p|BgZ5{6APE}6r5UJksHNCzXP3TYL
z807U%+<xKia%9#zt=P=w@UwCOmsDt~VLou?Q3Mvi*;5Ls!4Tv_k{1%Wcb8v^UR5_!
z(*m{qYtG$X!-xG;P?jI|JL%g*$l;J+Y=P{&mwn-D%^gJ3-A;G0flO)1^F%(ZxWAk_
z55}26yAdj(^Alp)q#3p=AYm^bbuZ9yPrOB<2Tn**j|+GT9ne8cp?q?nN0@bh?2Q|W
za8w@dVVlh**xp-D@4Z}R2N+58VrpZ)J!frnIM+q|v`^N1wC=;`{{~Lcj&0Oz&J(wH
zNtfI|rtatI<u9TF!-bSW@jf75;2*Ld8U`2Y_x;b~E02f!pZ}Ccw?Tg@J|^KGV)<L>
z@BixmxF7j11CO}q2d@6NJY2v0+rYn=`#-fGvziZl^>0DCPrH9>|9|%SPxZ&t-vjme
zTW;<_`rkz8pLQN|Ko3OWZ^6GWH}`fP(}sU4KPEdKYR_*`L;j`wU!>8W79P)S52VO%
zDMa~Y;WtI{r|RR;=V7q>Efg64Uisg~ygv;-uCWgj({CZXN9JEp{LA3)`RPw9kBi5D
uVgg0*k4Jx2=RYkx&i4OIOq=^<{U0SqQ5GKIVX<iUFP!@!_>Sn|>Hh%`$)+j*

literal 0
HcmV?d00001

diff --git a/previews/PR3525/tutorials/getting_started/data/Soccer.txt b/previews/PR3525/tutorials/getting_started/data/Soccer.txt
new file mode 100644
index 00000000000..c19056509de
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/data/Soccer.txt
@@ -0,0 +1,21 @@
+"Team"::"Played"::"Wins"::"Draws"::"Losses"::"Goals_for"::"Goals_against"
+"Barcelona"::38::30::4::4::"110 goals"::"21 goals"
+"Real Madrid"::38::30::2::6::"118 goals"::"38 goals"
+"Atletico Madrid"::38::23::9::6::"67 goals"::"29 goals"
+"Valencia"::38::22::11::5::"70 goals"::"32 goals"
+"Seville"::38::23::7::8::"71 goals"::"45 goals"
+"Villarreal"::38::16::12::10::"48 goals"::"37 goals"
+"Athletic Bilbao"::38::15::10::13::"42 goals"::"41 goals"
+"Celta Vigo"::38::13::12::13::"47 goals"::"44 goals"
+"Malaga"::38::14::8::16::"42 goals"::"48 goals"
+"Espanyol"::38::13::10::15::"47 goals"::"51 goals"
+"Rayo Vallecano"::38::15::4::19::"46 goals"::"68 goals"
+"Real Sociedad"::38::11::13::14::"44 goals"::"51 goals"
+"Elche"::38::11::8::19::"35 goals"::"62 goals"
+"Levante"::38::9::10::19::"34 goals"::"67 goals"
+"Getafe"::38::10::7::21::"33 goals"::"64 goals"
+"Deportivo La Coruna"::38::7::14::17::"35 goals"::"60 goals"
+"Granada"::38::7::14::17::"29 goals"::"64 goals"
+"Eibar"::38::9::8::21::"34 goals"::"55 goals"
+"Almeria"::38::8::8::22::"35 goals"::"64 goals"
+"Cordoba"::38::3::11::24::"22 goals"::"68 goals"
diff --git a/previews/PR3525/tutorials/getting_started/data/StarWars.csv b/previews/PR3525/tutorials/getting_started/data/StarWars.csv
new file mode 100644
index 00000000000..bf3394390dc
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/data/StarWars.csv
@@ -0,0 +1,21 @@
+Name,Gender,Height,Weight,Eyecolor,Haircolor,Skincolor,Homeland,Born,Died,Jedi,Species,Weapon
+Anakin Skywalker,male,1.88,84,blue,blond,fair,Tatooine,41.9BBY,4ABY,jedi,human,lightsaber
+Padme Amidala,female,1.65,45,brown,brown,light,Naboo,46BBY,19BBY,no_jedi,human,unarmed
+Luke Skywalker,male,1.72,77,blue,blond,fair,Tatooine,19BBY,unk_died,jedi,human,lightsaber
+Leia Skywalker,female,1.5,49,brown,brown,light,Alderaan,19BBY,unk_died,no_jedi,human,blaster
+Qui-Gon Jinn,male,1.93,88.5,blue,brown,light,unk_planet,92BBY,32BBY,jedi,human,lightsaber
+Obi-Wan Kenobi,male,1.82,77,bluegray,auburn,fair,Stewjon,57BBY,0BBY,jedi,human,lightsaber
+Han Solo,male,1.8,80,brown,brown,light,Corellia,29BBY,unk_died,no_jedi,human,blaster
+Sheev Palpatine,male,1.73,75,blue,red,pale,Naboo,82BBY,10ABY,no_jedi,human,force-lightning
+R2-D2,male,0.96,32,NA,NA,NA,Naboo,33BBY,unk_died,no_jedi,droid,unarmed
+C-3PO,male,1.67,75,NA,NA,NA,Tatooine,112BBY,3ABY,no_jedi,droid,unarmed
+Yoda,male,0.66,17,brown,brown,green,unk_planet,896BBY,4ABY,jedi,yoda,lightsaber
+Darth Maul,male,1.75,80,yellow,none,red,Dathomir,54BBY,unk_died,no_jedi,dathomirian,lightsaber
+Dooku,male,1.93,86,brown,brown,light,Serenno,102BBY,19BBY,jedi,human,lightsaber
+Chewbacca,male,2.28,112,blue,brown,NA,Kashyyyk,200BBY,25ABY,no_jedi,wookiee,bowcaster
+Jabba,male,3.9,NA,yellow,none,tan-green,Tatooine,unk_born,4ABY,no_jedi,hutt,unarmed
+Lando Calrissian,male,1.78,79,brown,blank,dark,Socorro,31BBY,unk_died,no_jedi,human,blaster
+Boba Fett,male,1.83,78,brown,black,brown,Kamino,31.5BBY,unk_died,no_jedi,human,blaster
+Jango Fett,male,1.83,79,brown,black,brown,ConcordDawn,66BBY,22BBY,no_jedi,human,blaster
+Grievous,male,2.16,159,gold,black,orange,Kalee,unk_born,19BBY,no_jedi,kaleesh,slugthrower
+Chief Chirpa,male,1,50,black,gray,brown,Endor,unk_born,4ABY,no_jedi,ewok,spear
diff --git a/previews/PR3525/tutorials/getting_started/data/passport-index-matrix.csv b/previews/PR3525/tutorials/getting_started/data/passport-index-matrix.csv
new file mode 100644
index 00000000000..633e868fac7
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/data/passport-index-matrix.csv
@@ -0,0 +1,200 @@
+Passport,"Afghanistan","Albania","Algeria","Andorra","Angola","Antigua and Barbuda","Argentina","Armenia","Australia","Austria","Azerbaijan","Bahamas","Bahrain","Bangladesh","Barbados","Belarus","Belgium","Belize","Benin","Bhutan","Bolivia","Bosnia and Herzegovina","Botswana","Brazil","Brunei","Bulgaria","Burkina Faso","Burundi","Cambodia","Cameroon","Canada","Cape Verde","Central African Republic","Chad","Chile","China","Colombia","Comoros","Congo","DR Congo","Costa Rica","Ivory Coast","Croatia","Cuba","Cyprus","Czech Republic","Denmark","Djibouti","Dominica","Dominican Republic","Ecuador","Egypt","El Salvador","Equatorial Guinea","Eritrea","Estonia","Swaziland","Ethiopia","Fiji","Finland","France","Gabon","Gambia","Georgia","Germany","Ghana","Greece","Grenada","Guatemala","Guinea","Guinea-Bissau","Guyana","Haiti","Honduras","Hong Kong","Hungary","Iceland","India","Indonesia","Iran","Iraq","Ireland","Israel","Italy","Jamaica","Japan","Jordan","Kazakhstan","Kenya","Kiribati","Kosovo","Kuwait","Kyrgyzstan","Laos","Latvia","Lebanon","Lesotho","Liberia","Libya","Liechtenstein","Lithuania","Luxembourg","Macao","Madagascar","Malawi","Malaysia","Maldives","Mali","Malta","Marshall Islands","Mauritania","Mauritius","Mexico","Micronesia","Moldova","Monaco","Mongolia","Montenegro","Morocco","Mozambique","Myanmar","Namibia","Nauru","Nepal","Netherlands","New Zealand","Nicaragua","Niger","Nigeria","North Korea","North Macedonia","Norway","Oman","Pakistan","Palau","Palestine","Panama","Papua New Guinea","Paraguay","Peru","Philippines","Poland","Portugal","Qatar","Romania","Russian Federation","Rwanda","Saint Kitts and Nevis","Saint Lucia","Samoa","San Marino","Sao Tome and Principe","Saudi Arabia","Senegal","Serbia","Seychelles","Sierra Leone","Singapore","Slovakia","Slovenia","Solomon Islands","Somalia","South Africa","South Korea","South Sudan","Spain","Sri Lanka","Saint Vincent and the Grenadines","Sudan","Suriname","Sweden","Switzerland","Syria","Taiwan","Tajikistan","Tanzania","Thailand","Timor-Leste","Togo","Tonga","Trinidad and Tobago","Tunisia","Turkey","Turkmenistan","Tuvalu","Uganda","Ukraine","United Arab Emirates","United Kingdom","United States","Uruguay","Uzbekistan","Vanuatu","Vatican","Venezuela","Viet Nam","Yemen","Zambia","Zimbabwe"
+"Afghanistan",-1,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Albania",0,-1,0,3,0,3,0,3,0,3,1,0,0,1,3,3,3,0,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,0,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,0,3,3,1,3,3,3,0,3,0,0,0,1,0,3,0,3,3,3,2,3,1,0,0,3,3,1,0,0,0,1,0,3,0,1,1,3,0,1,0,0,3,3,3,3,1,1,3,1,0,3,0,1,1,0,3,3,3,0,3,0,1,1,0,0,1,3,0,0,0,0,0,3,3,1,0,1,3,0,0,0,0,0,3,3,1,3,0,1,0,0,1,3,0,0,0,3,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,0,0,1,0,1,1,0,3,0,3,0,1,1,3,0,0,0,0,1,0,3,0,0,0,1,1
+"Algeria",0,0,-1,0,1,0,0,0,0,0,1,0,0,0,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,3,1,0,3,0,3,0,0,0,3,1,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,3,1,3,0,0,3,1,0,3,0,0,0,0,3,1,1,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,1,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,0,0,0,3,0,1,1,0,1,1,0,0,3,1,0,1,1,0,0,0,0,0,1,0,0,0,0,1,0,1
+"Andorra",0,3,0,-1,0,3,3,3,2,3,1,3,1,1,0,3,3,3,0,0,3,3,0,3,0,3,0,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,0,3,3,1,3,3,3,0,3,0,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,1,3,1,0,1,0,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,1,1,1,3,3,3,3,0,3,3,1,0,0,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,3,3,3,3,3,0,1,0,3,1,3,3,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,0,0,3,3,0,1,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Angola",0,0,0,0,-1,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,3,0,0,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,0,3,0,1,0,0,1,0,0,0,0,0,0,1,1,0,3,0,0,0,3,0,0,1,0,0,1,0,0,1,0,3,0,0,0,1,0,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,1,0,0,3,0,0,1,3,0,0,0,0,0,1,0,0,0,0,0,3,3
+"Antigua and Barbuda",0,3,0,3,0,-1,0,1,0,3,0,3,0,1,3,3,3,3,0,0,1,3,3,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,3,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,0,3,3,0,1,0,1,3,3,0,0,1,3,1,3,0,0,3,3,3,1,1,3,3,1,0,3,0,1,3,0,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,3,3,0,0,1,3,3,0,0,3,3,3,3,3,3,0,1,3,3,1,3,0,0,0,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,0,0,3,0,1,1,0,3,3,1,0,1,3,3,0,3,0,0,1,3,3,3,0,0,3,3
+"Argentina",0,3,0,3,1,3,-1,3,0,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,0,3,1,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,0,3,0,3,1,3,1,1,0,0,3,3,3,1,1,1,3,1,0,3,1,1,3,3,3,3,3,3,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,3,3,3,3,3,3,3,3,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,0,1,1,3,1,1,0,3,3,3,0,1,1,3,3,3,0,3,3,3,3,3,1,0,1,1
+"Armenia",0,3,0,0,0,3,3,-1,0,0,0,3,0,1,3,3,0,0,0,0,1,0,0,3,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,3,0,0,2,3,3,0,0,0,0,1,0,1,3,0,0,0,0,3,1,0,1,1,0,0,0,0,0,3,1,1,3,1,0,0,0,1,1,0,3,3,0,0,3,0,1,0,3,0,1,0,0,0,0,0,0,0,0,0,0,1,3,3,0,0,0,0,0,0,1,0,3,1,0,0,1,0,0,0,1,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,3,1,0,1,1,0,0,0,1,0,1,1,3,0,0,0,3,3,0,0,0,1,0,0,1
+"Australia",0,3,0,3,1,3,3,3,-1,3,1,3,1,1,3,3,3,3,0,0,3,3,3,0,1,3,1,0,1,0,2,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,1,3,3,3,3,3,3,0,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,1,1,3,1,0,1,1,0,3,3,2,3,3,3,3,3,1,0,1,1
+"Austria",0,3,0,3,1,3,3,3,2,-1,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,1,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Azerbaijan",0,3,0,0,0,3,0,0,0,0,-1,3,0,1,3,3,0,0,0,0,1,3,0,0,0,0,0,0,1,0,0,1,0,0,0,0,3,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,0,0,0,2,3,3,0,0,0,0,1,0,1,3,0,0,0,0,3,1,0,1,1,0,0,0,0,0,1,1,1,3,1,0,0,1,1,1,0,3,3,0,0,3,0,1,0,3,0,1,0,0,1,0,0,0,3,0,1,1,1,3,0,0,0,0,0,0,0,3,0,3,1,0,0,1,0,0,0,0,3,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,3,0,0,1,1,0,0,0,3,0,1,1,3,0,0,0,0,3,3,0,0,1,0,0,1
+"Bahamas",0,3,0,3,0,3,0,1,0,3,1,-1,0,1,3,0,3,3,0,0,1,3,3,3,0,3,0,0,1,0,2,1,0,0,3,3,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,3,3,3,0,0,1,3,1,3,0,0,3,3,3,1,1,3,3,1,0,3,0,1,3,3,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,3,3,0,1,1,3,3,0,0,3,3,3,3,3,3,0,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,0,0,3,0,1,1,1,3,0,1,0,1,3,0,3,3,3,3,1,3,3,0,0,0,3,3
+"Bahrain",0,3,0,0,0,0,0,1,0,0,1,3,-1,1,0,3,0,0,0,0,1,3,3,0,1,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,3,3,3,3,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,3,0,1,0,3,3,3,0,0,0,3,1,0,0,0,0,0,0,3,0,1,0,3,3,3,1,0,3,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,3,0,3,0,0,0,0,3,1,0,0,0,1,0,3,3,0,0,0,0,0,3,1,1,3,0,0,0,0,3,0,0,3,0,0,1,3,0,1,0,0,3,1,3,1,0,3,0,0,0,1,0,3,0,0,2,3,0,1,0,0,3,1,1,1,3,1,1,0,0,3,3,0,1,1,0,3,0,0,0,1,3,0,0,0,1,0,1
+"Bangladesh",0,0,0,0,0,0,0,0,0,0,0,3,0,-1,3,0,0,0,0,3,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,1,3,0,0,1,3,0,0,0,0,3,0,0,1,0,3,0,0,0,0,0,3,0,0,0,0,0,3,0,0,0,1,0,0,0,0,0,0,0,3,0,0,0,0,0,0,1,0,1,1,0,0,0,1,0,0,3,0,0,0,0,0,1,1,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,1,3,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,0,0,1,1,0,3,0,0,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0
+"Barbados",0,3,0,3,0,3,3,1,0,3,1,3,0,3,-1,3,3,3,0,0,1,3,3,3,0,3,0,0,1,0,2,1,0,0,3,3,3,1,0,0,3,2,3,3,3,3,3,1,3,3,3,0,3,3,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,3,3,3,0,0,1,3,1,3,0,0,3,3,3,1,1,3,3,1,0,3,1,1,3,3,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,3,3,0,1,1,3,3,0,0,3,3,3,3,1,3,0,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,0,0,3,0,1,1,1,3,3,1,0,1,3,0,1,3,0,3,1,3,3,3,0,0,3,3
+"Belarus",0,3,0,0,0,3,3,3,0,0,3,0,0,1,3,-1,0,0,0,0,1,0,0,3,0,0,0,0,1,0,0,1,0,0,0,3,0,1,0,0,0,2,0,3,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,3,0,0,0,0,0,0,1,0,3,0,3,0,0,0,3,1,0,0,3,0,1,0,1,3,1,0,0,0,3,1,0,1,1,0,0,0,0,0,3,1,1,3,1,0,0,0,1,1,0,3,3,0,3,3,0,1,1,3,0,1,0,0,3,0,0,0,0,0,1,0,1,3,3,0,0,3,0,0,0,3,0,3,1,0,0,1,0,0,0,0,3,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,3,1,0,1,1,0,0,0,3,0,1,1,3,0,0,0,0,3,3,0,3,3,0,1,1
+"Belgium",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,-1,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,1,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Belize",0,0,0,0,0,3,0,1,0,0,0,3,0,1,3,0,0,-1,0,0,1,0,3,3,0,0,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,0,3,0,0,0,3,1,3,0,0,1,3,3,0,0,0,3,3,0,1,3,3,3,3,0,0,2,3,1,0,3,3,0,3,0,0,0,3,3,3,0,0,1,0,1,3,0,0,0,0,0,1,1,3,3,1,0,0,1,1,3,3,3,0,0,0,0,0,1,0,0,0,1,0,0,3,0,0,0,0,0,0,0,1,3,3,0,0,3,3,0,0,1,0,0,1,3,3,1,0,0,0,0,0,1,0,3,0,0,1,1,3,0,0,0,2,3,0,3,0,0,0,3,0,3,0,1,1,0,3,0,3,0,1,3,0,0,3,0,3,1,3,0,3,0,0,3,3
+"Benin",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,-1,0,1,0,0,0,0,0,3,0,1,0,0,3,3,1,0,0,0,1,1,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,3,0,0,0,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,3,0,0,0,0,1,1,1,3,1,3,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,3,0,1,1,0,0,0,3,0,1,3,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Bhutan",0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,-1,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,3,1,0,0,0,2,0,1,0,0,0,1,3,0,3,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,0,3,0,3,0,0,3,3,1,0,0,0,0,0,0,1,0,1,0,0,1,0,1,0,1,1,0,0,0,0,0,1,1,1,1,1,0,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,3,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,1,1,1,1,0,0,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,0,1,1
+"Bolivia",0,0,0,0,0,0,3,1,0,0,1,3,1,1,0,0,0,3,0,0,-1,0,0,3,0,0,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,1,3,0,3,0,0,2,3,3,0,3,0,0,3,0,1,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,0,0,0,0,0,1,0,1,3,3,0,3,3,3,0,0,3,0,3,1,3,1,1,0,0,0,0,3,1,0,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,1,0,0,0,3,0,1,1,0,0,0,0,3,1,0,0,3,0,0,1,0
+"Bosnia and Herzegovina",0,3,0,3,0,0,0,1,0,3,1,3,0,1,0,3,3,0,0,0,1,-1,0,3,0,3,0,0,1,0,0,1,0,0,3,3,3,1,0,0,0,2,3,3,3,3,3,1,3,3,3,0,0,0,0,3,3,1,0,3,3,1,3,3,3,0,3,0,0,0,1,0,3,0,3,3,3,2,3,1,0,0,0,3,1,0,1,0,1,0,0,0,3,1,3,0,1,0,0,3,3,3,3,1,1,3,1,0,3,0,1,1,0,3,3,3,0,3,0,1,1,0,0,1,3,0,0,0,0,0,3,3,1,0,1,3,3,0,0,0,0,3,3,1,3,3,1,0,3,1,3,0,0,0,3,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,0,0,1,0,1,1,0,3,3,3,0,1,1,3,0,0,0,0,3,0,3,0,1,0,1,1
+"Botswana",0,0,0,0,3,3,0,0,0,0,0,3,0,3,3,0,0,3,3,0,1,0,-1,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,3,3,0,0,0,0,0,3,1,3,0,0,1,3,3,0,1,0,3,0,0,1,3,3,0,3,0,0,2,3,1,0,3,3,0,3,0,0,0,3,0,3,0,0,1,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,3,0,3,0,1,0,0,0,0,0,0,3,0,0,1,1,0,3,0,0,0,3,0,0,1,0,0,1,3,3,1,0,0,0,1,0,1,0,3,0,0,0,1,3,3,1,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,3,0,0,0,1,1,0,0,3,0,0,0,3,0,0,0,0,3,3
+"Brazil",0,3,0,3,1,3,3,3,0,3,1,3,1,0,3,3,3,3,0,0,3,3,3,-1,0,3,1,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,0,1,3,1,0,3,0,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,3,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,3,3,3,3,3,3,1,3,3,1,3,3,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,0,1,1,3,1,1,1,3,3,3,0,1,1,3,3,3,0,3,3,3,3,3,1,0,1,1
+"Brunei",0,3,0,3,0,3,0,1,2,3,1,3,1,1,3,0,3,3,0,0,1,3,3,0,-1,3,0,0,3,0,2,1,0,0,0,3,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,0,3,3,3,1,0,3,0,3,1,3,3,3,0,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,0,3,3,3,0,3,0,1,3,0,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,0,3,3,3,3,3,3,3,1,3,3,1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,3,1,3,3,1,1,1,3,3,3,0,1,1,3,3,3,2,0,3,3,3,0,3,0,1,1
+"Bulgaria",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,3,3,-1,0,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,1,3,1,3,1,3,3,1,1,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,1,3,1,3,3,3,3,0,3,1,3,3,3,0,3,3,3,3,3,1,0,1,1
+"Burkina Faso",0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,0,0,3,0,1,0,0,0,0,0,-1,0,1,0,0,3,3,3,0,0,0,1,1,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,3,0,0,0,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,3,0,0,0,0,1,1,0,0,1,3,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,1,0,1,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Burundi",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,-1,1,0,0,1,3,0,0,0,0,1,0,3,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,3,0,0,0,0,0,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,3,0,1,1,0,0,0,1,3,1,0,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,0,0,0,1,1
+"Cambodia",0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,3,0,0,0,-1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,1,0,3,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,1,1,1,0,3,0,0,0,0,0,1,3,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,3,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,3,1,1,0,0,0,1,0,1,1,0,0,0,0,0,1,0,0,0,3,0,1,0
+"Cameroon",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,-1,0,1,3,3,0,0,0,1,3,0,0,2,0,0,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,3,0,0,1,0,3,0,0,0,0,2,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,3,0,0,0,0,0,1,1,0,0,1,3,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,0,0,0,0,0,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,1,0,1,0,3,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0
+"Canada",0,3,0,3,1,3,3,1,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,0,3,3,1,0,1,0,-1,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,0,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,3,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,1,3,3,3,3,3,3,0,1,3,3,1,3,3,0,3,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,1,3,3,1,0,1,1,3,3,3,3,3,3,3,3,3,1,0,1,1
+"Cape Verde",0,0,0,0,1,0,0,0,0,0,0,3,0,1,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,-1,0,0,0,0,0,1,0,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,3,0,0,2,3,1,0,0,0,0,0,0,1,0,1,0,0,0,0,1,0,0,1,3,0,0,0,0,3,1,1,3,1,3,0,1,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,3,3,0,0,0,3,0,0,1,0,0,1,0,1,1,0,3,0,3,0,1,3,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,1,0,0,0,0,0,1,1
+"Central African Republic",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,3,0,1,-1,3,0,0,0,1,3,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,3,0,0,1,0,0,0,3,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Chad",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,3,0,1,3,-1,0,0,0,1,3,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,3,0,0,0,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,3,1,3,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,3,0,0,1,0,0,0,3,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Chile",0,3,0,3,1,3,3,1,0,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,0,3,0,0,1,0,2,1,0,0,-1,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,0,3,0,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,3,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,3,3,3,3,3,3,1,3,3,1,3,0,0,0,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,0,3,3,3,0,1,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"China",0,3,0,0,1,3,0,1,0,0,1,3,1,1,3,3,0,0,0,0,1,3,0,0,1,0,0,0,1,0,0,1,0,0,0,-1,0,1,0,0,0,2,0,0,0,0,0,1,3,0,3,3,0,3,0,0,0,1,3,0,0,1,0,0,0,0,0,3,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,3,0,1,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,1,1,1,0,0,0,1,3,0,3,0,0,0,0,3,1,1,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,3,0,0,1,3,3,1,0,1,0,1,3,1,0,0,0,0,0,1,0,0,0,0,2,0,0,1,0,0,0,0,0,1,1,1,1,1,0,3,1,0,1,1,0,3,0,0,0,1,3,0,0,1,0,0,0
+"Colombia",0,3,0,3,0,3,3,1,0,3,1,3,1,1,3,0,3,3,0,0,3,3,0,3,0,3,1,0,1,0,0,1,0,0,3,0,-1,1,0,0,0,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,0,3,3,0,3,0,3,0,1,3,0,3,3,3,3,2,0,0,0,0,3,3,3,0,0,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,1,1,0,0,1,0,3,1,1,1,3,3,3,3,0,3,0,1,1,0,0,1,3,0,1,0,0,0,3,3,1,0,1,3,3,0,3,3,3,3,3,3,3,3,1,3,0,1,3,0,0,0,3,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,1,0,1,0,1,1,0,3,0,3,0,1,1,0,0,0,0,3,1,0,3,3,1,0,1,0
+"Comoros",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,0,0,1,0,0,0,0,0,-1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,3,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,0,3,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,1,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,1,0,0,3,0,0,1,3,0,0,0,0,0,0,0,0,0,0,0,1,1
+"Congo",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,3,0,1,3,3,0,0,0,1,-1,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,1,1,1,0,1,0,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,1,0,1,0,3,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"DR Congo",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,1,3,1,0,0,1,3,0,0,0,0,1,0,-1,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,3,0,0,1,0,0,0,3,0,1,0,3,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,1,3
+"Costa Rica",0,3,0,3,0,0,3,1,0,3,1,3,0,1,3,0,3,3,0,0,3,3,3,3,3,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,-1,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,0,3,3,1,0,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,1,0,3,0,0,1,3,1,1,0,0,3,3,3,1,1,1,3,1,0,3,1,1,1,3,3,3,3,0,3,0,1,1,0,0,1,3,0,3,0,0,0,3,3,0,1,1,3,3,0,3,3,3,3,3,3,3,0,1,3,3,1,3,0,0,1,3,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,0,0,1,0,1,1,0,3,3,3,0,1,1,0,1,3,0,3,1,0,3,3,0,0,1,1
+"Ivory Coast",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,3,3,0,0,0,1,1,0,0,-1,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,1,3,0,0,0,0,1,1,1,0,1,3,0,0,3,1,0,3,0,0,0,0,3,1,1,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,1,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,1,0,0,0,0,0,0,0
+"Croatia",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,0,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,-1,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,0,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,1,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,0,3,1,3,3,3,1,0,3,1,3,3,3,0,3,3,3,3,3,1,0,1,1
+"Cuba",0,0,0,0,1,3,0,1,0,0,1,0,0,1,3,3,0,0,0,0,1,0,3,0,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,-1,0,0,0,1,3,0,0,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,3,0,3,1,3,0,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,3,1,0,0,1,0,0,0,0,0,1,1,0,3,1,0,0,1,1,1,0,3,0,0,3,3,0,1,0,3,0,1,0,0,0,0,0,0,3,0,0,0,1,3,0,0,0,0,0,0,0,3,0,3,1,3,3,1,0,0,0,0,3,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,1,1,0,1,1,0,3,0,0,0,1,1,0,0,0,0,0,1,3,0,0,1,0,1,1
+"Cyprus",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,3,3,3,0,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,-1,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,3,3,3,1,1,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,3,1,3,1,3,3,0,1,0,3,3,3,3,3,0,3,3,3,3,3,1,0,3,3
+"Czech Republic",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,-1,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Denmark",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,-1,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Djibouti",0,0,0,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,-1,3,0,3,0,0,0,0,0,0,3,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,3,0,0,2,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,3,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,1,0
+"Dominica",0,0,0,3,0,3,0,1,0,3,0,3,0,1,3,3,3,3,0,0,1,3,3,3,0,3,0,0,1,0,0,1,0,0,0,0,3,1,0,0,3,2,3,3,3,3,3,1,-1,3,3,1,0,0,0,3,3,1,3,3,3,1,3,0,3,0,3,3,0,0,1,3,3,0,3,3,3,2,3,1,0,3,3,3,3,0,1,0,3,0,3,0,0,1,3,0,3,0,0,3,3,3,3,1,3,3,1,0,3,0,1,3,0,3,3,3,0,3,0,1,0,0,0,1,3,0,1,0,0,0,0,3,0,0,1,3,3,0,0,3,3,3,3,1,3,3,1,3,3,1,3,0,0,0,3,1,0,3,3,3,1,1,0,3,0,3,2,3,0,3,3,3,0,1,0,3,0,1,1,1,3,0,1,0,1,1,0,0,3,0,3,1,3,3,3,0,0,3,0
+"Dominican Republic",0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,3,0,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,3,1,0,0,0,2,0,1,0,0,0,1,0,-1,3,1,3,0,0,0,0,1,0,0,0,1,3,3,0,0,0,3,3,0,1,3,0,0,3,0,0,2,3,1,0,0,3,0,0,3,1,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,3,3,3,0,0,3,0,0,1,0,0,1,0,0,0,1,0,1,0,3,0,0,1,1,0,3,0,0,2,0,0,1,0,0,0,3,0,1,0,1,1,0,3,0,1,0,1,1,0,0,0,0,0,1,0,0,0,0,0,1,1
+"Ecuador",0,0,0,0,0,0,3,1,0,0,1,3,1,1,0,3,0,3,0,0,3,0,0,3,0,0,0,0,1,0,0,1,0,0,3,3,3,1,0,0,0,2,0,1,0,0,0,1,3,3,-1,1,3,0,0,0,0,1,0,0,0,1,3,3,0,0,0,0,0,0,1,3,3,3,3,0,0,2,3,1,0,0,3,0,3,0,1,3,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,1,3,3,3,0,0,3,0,1,1,0,0,1,0,0,3,0,0,0,0,0,1,0,1,3,3,1,3,3,3,0,0,3,0,3,1,3,1,1,0,0,0,0,0,1,0,3,0,0,0,1,3,3,0,0,2,3,0,1,0,0,0,1,0,1,0,1,1,0,3,0,3,0,1,1,0,0,0,0,3,1,0,0,3,0,0,1,1
+"Egypt",0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,-1,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,3,1,0,3,0,3,0,0,0,3,3,0,0,0,0,0,0,3,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,3,1,0,3,1,0,0,1,1,3,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,3,0,1,0,0,0,1,0,1,0,0,0,0,0,1,0,0,0,0,0,3,0,1,0,0,0,0,1,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,1,0,0,0,0,1,0,1
+"El Salvador",0,3,0,3,0,0,3,1,0,3,0,3,0,1,3,3,3,3,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,0,-1,0,0,3,0,1,0,3,3,1,0,3,3,0,3,0,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,0,1,1,0,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,3,3,0,0,1,3,3,0,3,0,3,3,3,1,3,3,1,3,1,1,3,0,0,1,0,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,0,0,1,0,1,1,0,3,0,3,0,1,1,0,1,3,0,3,1,0,3,0,0,0,1,1
+"Equatorial Guinea",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,3,0,1,3,3,0,0,0,1,3,0,0,2,0,1,0,0,0,1,3,0,3,0,0,-1,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,3,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,1,0,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,3,0,1,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,0,0,1,1,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,1
+"Eritrea",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,0,0,0,0,-1,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,2,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,0,0,0,1,0
+"Estonia",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,-1,3,1,3,3,3,1,0,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Swaziland",0,0,0,0,1,3,0,0,0,0,0,3,0,1,3,0,0,3,3,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,-1,1,3,0,0,1,3,0,0,3,0,3,0,0,1,3,3,0,3,0,0,2,3,1,0,3,3,0,3,0,1,0,3,0,3,1,0,0,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,3,0,3,0,0,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,1,3,3,1,0,0,0,1,0,1,0,3,0,0,0,1,3,3,0,0,2,3,0,1,0,0,0,1,0,3,0,1,1,0,3,0,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,3,3
+"Ethiopia",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,0,0,0,0,1,0,0,-1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,3,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,1,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,0,1,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,1
+"Fiji",0,0,0,0,0,3,3,1,0,0,0,3,0,3,3,0,0,3,0,0,1,0,0,3,0,0,0,0,1,0,0,1,0,0,3,3,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,1,3,0,0,0,0,1,-1,0,0,1,3,0,0,0,0,0,3,0,1,0,3,0,3,0,0,2,3,1,0,3,3,0,3,0,1,0,3,3,3,0,0,1,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,1,1,3,0,3,0,0,0,0,0,1,1,0,1,1,0,0,3,0,0,0,0,0,0,0,1,3,3,1,0,3,3,0,0,1,0,3,1,3,3,1,0,0,0,1,0,1,0,3,0,0,1,1,0,3,0,0,2,3,0,1,0,0,0,0,0,1,1,1,1,1,0,3,1,0,1,3,0,0,0,0,0,1,3,0,0,1,0,3,3
+"Finland",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,-1,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"France",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,-1,1,1,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Gabon",0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,3,0,1,3,3,0,0,0,1,3,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,-1,0,0,0,1,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,3,0,3,0,0,0,0,3,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,0,3,0,0,0,3,0,0,1,0,0,1,0,1,1,0,3,0,1,0,1,0,3,0,0,0,1,3,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,1,0,0,3,0,0,1,1,0,0,0,0,0,1,0,0,0,0,0,0,0
+"Gambia",0,0,0,0,0,0,0,0,0,0,0,3,0,3,3,0,0,3,3,0,1,0,3,0,0,0,3,0,1,0,0,3,0,0,0,0,0,1,0,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,-1,0,0,3,0,3,0,3,3,0,3,0,0,0,0,2,3,1,0,0,0,0,3,0,0,0,3,0,0,0,0,1,0,0,3,3,0,0,0,0,1,1,3,3,1,3,0,0,3,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,3,1,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,3,0,1,3,0,3,3,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,0,0
+"Georgia",0,3,0,3,0,3,3,3,0,3,3,3,1,1,3,3,3,0,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,0,2,3,1,3,3,3,1,3,0,3,1,0,0,0,3,0,1,0,3,3,1,0,-1,3,0,3,0,0,0,1,0,3,0,0,3,3,2,3,3,0,0,3,3,1,0,1,3,1,0,0,1,3,1,3,1,1,0,0,3,3,3,1,1,1,3,1,0,3,1,1,3,0,3,3,3,0,3,0,1,1,0,0,1,3,0,1,0,0,0,0,3,1,0,1,3,3,0,0,0,0,3,3,3,3,0,1,0,3,1,3,0,0,0,3,1,0,0,3,3,0,1,0,0,0,3,2,3,0,1,3,3,0,0,3,0,1,1,1,0,0,0,3,3,1,1,3,0,0,0,3,3,0,3,0,1,0,1,1
+"Germany",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,-1,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,3,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Ghana",0,0,0,0,0,0,0,0,0,0,0,0,0,3,3,0,0,3,3,0,1,0,0,0,0,0,3,0,1,0,0,3,0,0,0,0,0,1,0,0,0,3,0,0,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,3,0,0,-1,0,3,0,3,3,0,3,0,0,0,0,2,3,1,0,0,0,0,3,0,1,0,3,0,0,0,0,0,0,1,1,3,0,0,0,0,1,1,0,0,1,3,0,0,1,3,0,3,0,0,0,0,0,1,1,0,0,0,0,0,1,3,3,0,0,0,0,1,1,0,0,0,0,0,3,0,0,1,0,0,3,3,1,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,3,0,3,0,0,0,1,3,0,0,0,0,0,1,3,0,0,0,0,1,3
+"Greece",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,-1,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Grenada",0,0,0,3,0,3,3,1,0,3,0,3,0,3,3,0,3,3,0,0,1,3,3,3,0,3,0,0,1,0,0,1,0,0,3,3,3,1,0,0,3,2,3,3,3,3,3,1,3,3,3,1,0,0,0,3,3,1,3,3,3,1,3,0,3,0,3,-1,0,0,1,3,3,0,3,3,3,2,3,1,0,3,3,3,3,0,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,3,1,3,3,1,0,3,0,1,3,0,3,3,3,0,3,0,1,0,0,0,1,3,0,1,0,0,0,0,3,0,0,1,3,3,0,0,3,3,3,3,1,3,3,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,0,3,0,3,2,3,0,3,3,3,0,0,0,3,0,1,1,0,3,0,1,0,1,3,0,0,3,0,3,1,3,3,3,0,0,3,3
+"Guatemala",0,3,0,3,0,0,3,1,0,3,1,3,0,1,0,0,3,3,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,0,3,0,0,3,0,1,0,3,3,1,0,0,3,0,3,0,-1,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,1,1,1,0,3,3,3,0,3,0,1,1,0,0,1,3,0,3,0,0,0,3,3,0,0,1,3,3,0,3,3,3,3,3,1,3,3,1,3,1,1,3,0,0,1,0,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,3,0,1,0,1,1,0,3,0,3,0,1,1,0,0,3,0,3,1,0,3,3,0,0,1,1
+"Guinea",0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,0,0,0,0,0,1,0,0,0,3,0,0,0,0,0,1,3,0,3,1,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,-1,3,0,3,0,3,0,0,2,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,3,0,0,0,0,1,1,1,3,1,3,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,3,0,1,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,1,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Guinea-Bissau",0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,0,0,0,0,0,1,0,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,-1,0,3,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,1,1,1,0,1,3,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,3,0,3,3,1,3,3,0,0,0,1,0,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Guyana",0,0,0,0,0,3,3,1,0,0,0,3,1,1,3,0,0,3,0,0,1,0,3,3,0,0,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,1,0,0,0,0,3,1,3,0,0,1,3,0,0,0,0,3,0,0,1,-1,3,0,3,0,0,2,3,1,0,3,0,0,3,0,1,0,1,0,3,0,0,1,0,0,3,0,0,0,0,0,1,1,1,3,1,0,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,3,0,0,3,3,0,0,3,0,3,1,3,3,1,0,0,0,0,0,1,0,3,0,0,1,1,3,3,0,0,2,3,0,3,0,0,0,0,0,3,0,1,1,0,3,0,1,0,1,1,0,0,0,0,3,1,3,0,0,0,0,1,1
+"Haiti",0,0,0,0,0,0,0,1,0,0,0,0,0,1,3,0,0,0,3,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,3,-1,0,3,0,0,2,3,1,0,0,3,0,0,0,1,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,3,0,0,1,0,0,0,1,0,1,0,3,0,0,0,1,0,3,0,0,2,3,0,0,0,0,0,3,0,1,0,1,1,0,0,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,0,1,1
+"Honduras",0,3,0,3,0,0,3,1,0,3,1,3,0,1,0,0,3,3,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,0,3,0,0,3,0,1,0,3,3,1,3,3,3,0,3,0,3,0,1,0,3,-1,3,3,3,2,3,1,0,3,3,3,0,3,1,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,0,1,1,0,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,3,3,0,0,1,3,3,0,3,3,3,3,3,1,3,3,1,3,1,1,3,0,0,1,0,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,3,0,1,0,1,1,0,3,3,3,0,1,1,0,3,3,0,3,1,0,3,0,0,0,1,1
+"Hong Kong",0,3,0,3,0,3,3,3,2,3,1,3,1,0,3,3,3,3,3,0,1,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,0,2,3,1,3,3,3,1,3,3,3,3,0,0,0,3,0,1,3,3,3,1,0,0,3,0,3,3,0,0,1,3,3,0,-1,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,0,1,3,1,3,0,0,3,3,3,3,1,3,3,1,0,3,0,1,3,3,3,3,3,3,3,3,1,3,3,0,1,3,3,3,3,0,0,3,3,1,0,1,3,3,1,0,3,3,3,3,3,3,3,1,3,3,1,3,0,0,0,3,1,0,3,3,3,0,1,3,3,0,3,2,3,0,3,3,3,0,1,0,3,3,1,1,1,3,3,3,0,1,3,3,3,3,0,3,1,3,3,3,1,3,3,3
+"Hungary",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,-1,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Iceland",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,-1,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,1,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"India",0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,0,0,0,3,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,3,0,3,0,0,0,0,1,3,0,0,1,0,0,0,0,0,3,0,0,1,0,3,0,2,0,0,-1,3,1,0,0,0,0,3,0,1,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,3,1,0,1,1,0,0,1,1,3,0,3,0,0,0,0,0,1,1,0,0,3,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,3,0,0,1,3,1,1,0,0,0,3,3,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,1,1,1,0,3,3,0,0,1,1,0,0,0,0,0,1,3,0,0,1,0,0,1
+"Indonesia",0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,3,0,0,0,0,1,0,0,3,3,0,0,0,3,0,0,1,0,0,3,0,3,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,3,0,0,1,3,0,0,0,0,0,0,0,1,3,3,0,3,0,0,2,-1,1,0,0,0,0,0,3,1,0,1,0,0,0,1,3,0,0,1,0,0,0,0,0,3,1,1,3,1,3,0,1,1,1,0,3,0,0,0,0,3,1,3,3,0,1,0,0,1,0,0,0,0,0,1,1,1,3,0,1,0,3,3,0,0,3,0,0,3,3,0,1,0,0,0,0,3,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,1,1,3,1,1,0,0,0,1,0,1,1,0,0,0,0,0,3,0,0,0,3,0,0,1
+"Iran",0,0,0,0,0,0,0,3,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,0,0,0,2,0,-1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,0,0,1,0,0,0,0,0,0,0,1,1,0,0,0,3,0,1,1,0,0,0,0,0,0,0,0,3,0,0,0,1
+"Iraq",0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,0,0,0,0,0,0,-1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,0,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Ireland",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,-1,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,3,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,1,3,3,1,0,1,1,3,3,3,2,3,3,3,3,3,1,0,3,1
+"Israel",0,3,0,3,1,0,3,1,0,3,1,3,0,1,3,3,3,3,0,0,1,3,3,3,0,3,0,0,1,0,2,1,3,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,0,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,0,0,0,3,-1,3,3,3,1,3,1,0,3,0,1,1,3,0,3,0,0,3,3,3,3,1,3,0,1,0,3,1,1,3,3,3,3,3,3,3,0,1,1,0,1,1,3,3,3,0,0,0,3,3,0,0,3,3,3,1,3,3,3,3,3,1,3,3,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,3,1,1,3,1,1,1,3,0,3,0,1,1,3,0,3,0,3,3,3,3,0,0,0,1,1
+"Italy",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,-1,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Jamaica",0,0,0,0,0,3,3,1,0,0,1,3,0,3,3,0,0,3,0,0,1,0,3,3,0,0,0,0,1,0,0,1,0,0,3,0,3,1,0,0,0,2,0,1,0,0,0,1,3,3,3,1,0,0,0,0,3,1,3,0,0,1,3,0,0,0,0,3,0,0,1,3,3,0,3,0,0,2,3,1,0,0,3,0,-1,0,1,0,3,3,0,0,0,1,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,3,3,0,0,0,0,0,1,1,3,0,1,0,0,1,0,0,0,0,0,0,0,1,3,3,0,0,3,3,0,0,1,0,3,1,3,3,1,0,0,0,1,3,1,0,3,0,0,0,1,3,3,0,0,2,3,0,3,0,0,0,0,0,3,0,1,1,0,3,0,1,0,1,3,0,0,0,0,3,1,3,0,3,0,0,3,3
+"Japan",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,0,3,3,0,0,1,0,2,1,0,0,3,3,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,-1,1,3,1,3,3,1,3,3,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,3,3,3,1,3,3,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,3,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,3,1,1,3,1,1,1,3,3,3,0,1,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Jordan",0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,3,3,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,3,0,0,2,3,0,0,0,0,0,0,0,-1,0,1,0,3,0,0,0,0,3,1,0,3,0,0,0,1,1,1,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,3,0,1,0,0,0,0,0,1,0,0,0,0,0,1,3,0,0,0,2,0,0,1,0,0,3,0,1,1,0,1,1,0,0,3,3,0,1,1,0,0,0,0,0,1,0,0,0,0,1,0,0
+"Kazakhstan",0,3,0,0,0,3,3,3,0,0,3,0,1,1,3,3,0,0,0,0,1,0,0,3,0,0,0,0,1,0,0,1,0,0,0,0,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,1,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,1,0,1,-1,1,0,0,0,3,1,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,1,1,1,0,3,3,0,3,3,0,1,1,3,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,3,0,3,1,0,0,1,0,0,0,0,3,1,0,0,0,0,0,1,0,3,0,0,2,3,0,1,0,0,0,0,3,0,1,1,1,0,0,0,3,0,1,1,3,1,0,0,0,3,0,0,0,1,0,1,1
+"Kenya",0,0,0,0,0,3,0,0,0,0,0,3,0,0,3,0,0,0,3,0,1,0,3,0,0,0,1,3,1,0,0,1,0,0,0,0,0,1,0,1,0,2,0,3,0,0,0,1,3,0,0,0,0,0,0,0,3,3,3,0,0,1,3,0,0,3,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,3,0,1,0,-1,3,0,0,0,1,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,1,1,3,0,1,0,0,0,0,1,0,0,0,0,0,1,0,3,0,0,0,3,0,0,1,0,0,3,3,1,1,0,0,0,1,0,1,1,3,0,0,0,1,0,0,1,0,0,3,1,1,0,0,0,0,0,3,0,1,1,0,3,0,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,3,3
+"Kiribati",0,0,0,3,0,3,0,1,0,3,0,3,0,1,3,0,3,3,0,0,1,3,3,0,0,3,0,0,1,0,0,1,0,0,0,0,0,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,0,0,0,3,0,1,3,3,3,1,3,0,3,0,3,3,0,0,1,0,3,0,3,3,3,2,3,1,0,3,0,3,3,0,1,0,3,-1,3,0,0,1,3,0,3,0,0,3,3,3,3,1,3,3,1,0,3,1,1,3,0,3,3,3,0,0,0,1,0,0,0,1,3,0,1,0,0,0,0,3,0,0,1,3,3,1,0,3,3,3,3,1,3,0,1,3,3,1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,1,0,3,0,1,1,1,3,3,1,0,1,1,0,0,3,0,0,1,3,3,0,0,0,3,3
+"Kosovo",0,3,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,0,3,0,0,0,0,0,0,1,0,0,0,0,1,0,1,0,0,0,-1,0,0,1,0,0,1,0,0,0,0,0,1,1,0,0,1,0,0,1,1,0,0,3,0,0,0,3,0,1,0,0,0,1,0,0,0,0,0,0,3,0,1,0,1,3,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,3,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,1,0,0,0,3,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Kuwait",0,3,0,0,0,0,0,1,0,0,1,3,3,1,0,3,0,0,0,0,1,3,3,0,1,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,3,3,3,3,0,0,0,3,1,0,0,0,1,0,3,0,0,0,0,3,0,1,0,3,3,3,0,0,0,3,1,0,0,0,0,0,0,3,0,1,0,3,-1,3,1,0,3,1,0,0,0,0,0,1,1,1,3,1,0,0,1,1,3,0,3,0,0,0,3,3,1,1,0,0,1,0,3,3,0,0,0,0,0,3,1,1,3,3,0,0,0,3,0,0,3,0,0,1,3,3,1,0,3,3,1,3,1,0,3,0,0,1,1,0,3,0,0,2,3,3,1,0,0,3,1,1,1,3,1,1,0,0,3,3,0,1,1,0,3,2,0,0,1,3,0,0,0,1,0,1
+"Kyrgyzstan",0,0,0,0,0,3,0,3,0,0,3,0,0,1,3,3,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,0,0,0,0,3,1,0,0,0,0,1,0,1,3,1,0,0,0,-1,1,0,1,1,0,0,0,0,0,1,1,1,3,1,0,0,1,1,1,0,3,3,0,3,0,0,1,1,3,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,3,1,0,1,1,0,0,0,0,3,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,3,0,0,1,1,0,0,0,3,0,1,1,3,0,0,0,0,3,0,0,0,3,0,0,1
+"Laos",0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,3,0,0,0,3,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,1,0,-1,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,0,0,3,0,0,3,0,0,1,3,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,3,1,0,0,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,3,1,1,0,0,0,0,0,1,1,0,0,0,0,0,1,0,0,0,3,0,1,0
+"Latvia",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,-1,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,0,0,1,1
+"Lebanon",0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,0,0,0,0,3,3,1,0,0,0,0,0,3,0,0,0,0,0,0,0,0,-1,1,0,0,0,0,0,3,1,0,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,1,3,0,0,0,0,0,0,0,3,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,0,0,1,0,0,3,0,1,0,0,1,1,0,0,0,3,0,1,1,0,0,0,0,0,1,0,0,0,0,0,0,0
+"Lesotho",0,0,0,0,1,3,0,0,0,0,0,3,0,3,3,0,0,3,3,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,3,0,0,1,0,3,0,0,1,3,3,0,3,0,0,2,3,1,0,3,3,0,3,3,1,0,3,3,3,0,0,0,0,0,-1,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,1,0,3,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,1,3,3,1,0,0,0,1,0,1,0,3,0,0,0,1,3,3,0,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,3,0,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,3,3
+"Liberia",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,3,0,0,0,0,1,0,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,0,0,0,2,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,-1,0,0,0,0,1,1,0,0,1,3,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,0,0,0,0,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,3,0,1,3,3,0,0,0,1,0,3,0,0,0,3,0,0,0,0,0,0,0,1,0,1,3,0,0,0,0,0,1,1,0,3,0,0,0,0,0,0,0,0,0,0,0
+"Libya",0,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,-1,0,0,0,1,1,0,3,1,0,0,0,3,0,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,1,0,1,1,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Liechtenstein",0,3,0,3,0,3,3,3,2,3,1,3,1,1,3,3,3,0,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,1,3,3,1,3,1,3,1,1,0,0,-1,3,3,3,1,1,3,1,0,3,0,1,3,3,3,3,3,0,3,3,1,0,3,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,1,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Lithuania",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,-1,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Luxembourg",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,3,3,1,3,0,0,3,3,-1,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Macao",0,3,0,3,0,3,0,3,0,3,1,0,1,1,3,3,3,0,0,0,1,3,0,3,3,3,0,0,1,0,0,3,0,0,3,3,0,1,0,0,0,2,3,1,3,3,3,1,3,3,3,3,0,0,0,3,0,1,3,3,3,1,0,0,3,0,3,3,0,0,1,0,3,0,3,3,3,2,3,1,0,3,3,3,0,3,1,0,1,3,3,0,0,1,3,1,1,0,0,3,3,3,-1,1,1,3,1,3,3,0,1,3,3,3,3,3,3,3,3,1,3,3,0,1,3,3,3,0,0,0,3,3,1,0,1,3,0,1,0,0,3,3,3,1,3,3,1,3,1,3,3,0,0,0,3,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,1,0,3,3,1,1,1,0,0,3,0,1,1,0,0,3,0,3,0,3,3,0,1,0,0,1
+"Madagascar",0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,3,0,0,0,3,1,0,0,0,1,0,0,0,1,0,0,3,0,1,0,3,3,3,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,-1,1,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,3,0,0,0,0,0,0,0,1,0,3,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,1,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,0,0,0,1,3
+"Malawi",0,0,0,0,1,3,0,0,0,0,0,3,0,3,3,0,0,3,3,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,3,0,0,1,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,0,3,0,3,0,1,0,3,3,3,0,0,1,0,0,3,0,0,0,0,0,1,1,-1,3,1,0,0,0,1,3,0,3,0,0,0,0,0,3,0,3,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,1,3,3,1,0,0,0,1,0,1,0,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,3,0,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,3,3
+"Malaysia",0,3,3,3,0,3,3,1,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,3,3,3,0,0,3,0,0,1,0,0,3,0,0,1,0,0,3,2,3,3,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,3,0,3,0,3,3,3,1,3,3,3,3,1,3,3,3,1,3,0,0,3,3,3,3,1,3,-1,1,0,3,1,1,3,3,3,3,3,3,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,0,3,3,3,3,3,3,0,1,3,1,1,3,0,0,1,0,1,0,3,3,3,1,1,3,3,0,3,2,3,1,3,3,3,0,3,1,3,3,1,1,1,3,3,3,0,1,1,0,3,3,0,3,3,3,3,3,3,1,3,3
+"Maldives",0,0,0,0,0,3,0,1,0,0,1,3,0,3,3,0,0,3,0,3,1,0,3,0,3,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,3,2,0,1,0,0,0,1,3,0,3,1,0,0,0,0,3,1,3,0,0,1,3,0,0,0,0,3,0,0,1,0,3,0,3,0,0,3,3,1,0,3,0,0,3,0,1,0,3,0,3,0,0,1,0,0,3,0,0,0,0,0,1,1,1,3,-1,0,0,1,1,3,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,0,0,0,0,0,0,3,1,3,3,1,0,0,3,0,0,3,0,0,1,3,3,1,0,0,0,0,0,1,0,3,0,0,1,1,3,0,0,0,1,3,0,1,0,0,0,0,0,1,3,1,1,0,3,3,1,0,1,1,0,1,3,0,0,1,3,0,0,0,0,3,3
+"Mali",0,0,3,0,0,0,0,0,0,0,0,3,0,1,3,0,0,0,3,0,1,0,0,0,0,0,3,0,1,3,0,3,0,3,0,0,0,1,0,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,3,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,3,0,0,0,0,3,1,0,0,1,-1,0,0,3,0,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,3,0,1,3,0,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Malta",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,-1,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,3,1,3,1,3,3,3,1,0,3,3,3,3,3,2,3,3,3,3,3,1,0,3,3
+"Marshall Islands",0,0,0,3,0,3,0,1,0,3,0,3,0,1,0,0,3,3,0,0,1,3,0,0,0,3,0,0,1,0,0,1,0,0,0,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,0,0,3,0,3,0,3,0,1,0,3,3,3,3,3,2,3,1,0,0,0,3,0,0,1,0,1,3,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,-1,1,1,3,3,3,3,0,0,0,1,0,0,1,1,3,0,3,0,0,0,0,3,0,0,3,3,3,1,0,3,3,3,3,1,3,0,1,0,3,1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,3,0,1,0,1,1,1,0,0,1,0,1,1,0,0,3,3,0,1,3,3,0,1,0,3,1
+"Mauritania",0,0,3,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,0,3,0,0,0,1,1,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,1,0,0,0,3,1,0,3,0,3,0,0,0,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,3,1,3,0,0,-1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,3,0,0,0,0,1,0,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Mauritius",0,3,0,3,3,3,0,0,0,3,1,3,0,0,3,0,3,3,3,0,1,3,3,0,0,3,0,0,1,0,0,1,0,0,3,3,0,1,0,1,3,2,3,1,3,3,3,1,3,3,3,0,0,0,0,3,3,1,3,3,3,3,3,3,3,3,3,3,0,0,1,0,3,0,3,3,3,2,3,1,0,0,3,3,3,3,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,3,1,3,3,1,0,3,1,1,-1,0,3,3,3,0,3,0,3,1,3,0,1,3,3,1,0,0,0,3,3,0,0,1,3,3,0,0,0,3,3,3,1,3,3,3,3,3,1,3,0,0,3,0,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,1,0,3,3,1,1,0,3,3,1,0,1,3,0,1,3,0,0,1,3,3,0,0,0,3,3
+"Mexico",0,3,0,3,0,3,3,1,0,3,1,3,1,0,3,3,3,3,0,0,3,3,3,3,0,3,0,0,1,0,2,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,0,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,0,3,0,1,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,-1,3,3,3,0,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,0,0,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,0,0,0,3,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,0,1,1,1,1,1,0,3,3,1,0,1,1,0,3,3,0,3,1,3,3,3,1,0,1,1
+"Micronesia",0,0,0,3,0,0,0,1,0,3,0,0,0,1,0,3,3,3,0,0,1,3,0,0,0,3,0,0,1,0,0,1,0,0,0,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,0,0,0,3,0,1,3,3,3,1,3,0,3,0,3,0,0,0,1,0,3,0,3,3,3,2,0,1,0,0,3,3,0,0,0,0,1,3,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,0,1,0,3,3,1,1,3,-1,3,3,0,0,0,1,0,0,1,1,3,0,1,0,0,0,0,3,0,0,3,3,3,1,0,3,3,3,3,1,3,0,1,0,3,1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,0,0,1,0,1,1,1,0,0,1,0,1,1,0,0,3,3,0,1,3,3,0,1,0,1,1
+"Moldova",0,3,0,3,0,3,0,3,0,3,3,3,0,1,3,3,3,0,0,0,1,3,0,0,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,0,2,3,1,3,3,3,1,3,0,3,0,0,0,0,3,0,1,3,3,3,1,0,3,3,0,3,0,0,0,1,0,3,0,0,3,3,2,3,1,0,0,3,3,1,0,0,3,1,0,0,0,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,0,1,1,0,3,-1,3,0,3,0,1,0,3,0,1,3,0,1,0,0,0,3,3,1,0,1,3,3,0,0,3,0,3,3,3,3,3,1,3,0,1,3,0,0,0,3,1,0,0,3,3,0,1,0,0,0,3,2,3,0,1,3,3,0,0,3,1,0,1,1,0,0,3,3,0,1,1,3,0,0,0,0,3,0,3,0,1,0,1,1
+"Monaco",0,3,0,3,1,3,3,3,2,3,1,3,1,1,0,3,3,3,0,0,3,3,3,3,0,3,0,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,0,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,-1,0,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,1,0,3,3,0,1,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Mongolia",0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,3,0,0,0,0,1,0,0,3,0,0,0,0,1,0,0,1,0,0,3,3,0,1,0,0,0,2,0,3,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,3,0,0,0,0,3,1,0,0,0,3,3,0,0,1,0,0,0,0,0,3,1,1,3,1,0,0,1,1,1,0,3,0,0,-1,0,0,1,1,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,3,0,0,0,3,0,0,1,0,3,1,0,1,1,0,0,0,0,3,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,1,1,3,1,1,0,0,0,3,0,1,1,3,0,0,0,3,3,0,0,0,1,0,1,0
+"Montenegro",0,3,0,3,0,0,3,3,0,3,0,0,0,1,0,3,3,0,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,3,3,3,3,1,3,0,3,1,0,0,0,3,3,1,0,3,3,1,3,3,3,0,3,0,0,0,1,0,3,0,3,3,3,2,0,1,0,0,3,3,1,0,1,0,1,0,3,0,3,1,3,1,1,0,0,3,3,3,3,1,1,1,1,0,3,0,1,1,0,3,3,3,0,-1,0,1,0,0,0,1,3,0,1,0,0,0,3,3,1,0,1,3,3,0,0,3,0,3,3,1,3,3,1,0,3,1,3,0,0,0,3,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,1,0,1,0,1,1,0,3,3,3,0,1,1,3,3,0,0,3,3,0,3,0,1,0,3,0
+"Morocco",0,0,3,0,1,0,0,0,0,0,1,0,0,0,0,0,0,3,3,0,1,0,0,3,0,0,0,0,1,0,0,0,0,0,0,0,0,1,1,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,3,3,0,0,0,0,3,0,3,1,0,3,0,3,0,0,0,3,1,0,0,0,0,0,0,1,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,3,1,0,3,1,3,0,1,1,1,0,3,0,0,0,0,-1,1,1,0,0,1,0,0,1,3,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,3,0,3,0,1,0,0,0,0,0,1,0,3,0,0,2,3,0,1,0,0,3,0,1,1,0,1,1,0,0,3,3,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0
+"Mozambique",0,0,0,0,3,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,0,0,0,1,0,0,0,1,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,3,0,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,1,1,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,-1,0,3,0,1,0,0,1,0,0,0,0,0,0,1,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,3,0,0,0,1,0,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,3,0,0,0,0,3,3
+"Myanmar",0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,3,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,1,0,1,0,0,0,0,3,0,0,1,0,0,0,0,0,1,1,1,1,1,0,0,0,1,1,0,3,0,0,0,0,0,1,-1,0,0,1,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,1,3,1,1,0,0,0,1,0,1,1,0,0,0,0,0,0,0,0,0,3,0,1,0
+"Namibia",0,0,0,0,3,3,0,0,0,0,0,3,0,1,0,0,0,3,3,0,1,0,3,3,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,3,0,0,0,1,3,3,3,0,0,0,0,0,3,1,0,0,0,1,3,0,0,1,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,3,0,0,0,3,0,3,0,0,1,0,0,3,0,0,0,0,0,3,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,1,0,-1,0,1,0,0,1,0,0,0,0,0,0,0,1,0,3,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,1,0,1,0,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,3,3,1,0,1,1,0,0,3,0,0,0,3,0,0,0,0,3,3
+"Nauru",0,0,0,0,0,3,0,1,0,0,0,3,0,1,0,0,0,0,0,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,3,2,0,1,0,0,0,1,3,3,3,1,0,0,0,0,3,1,3,0,0,1,3,0,0,0,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,3,3,0,3,0,1,0,3,3,3,0,0,0,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,1,1,3,0,3,0,0,0,0,0,1,0,0,-1,1,0,0,1,0,0,0,0,0,0,0,1,3,3,1,0,3,0,0,0,1,0,3,1,3,3,1,0,0,0,0,0,1,0,3,0,0,1,1,0,3,0,0,2,3,0,0,0,0,0,3,0,3,0,1,1,1,3,0,1,0,1,1,0,0,3,0,0,1,3,0,0,1,0,1,3
+"Nepal",0,0,0,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,0,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,0,3,0,0,0,0,3,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,0,1,0,1,1,0,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,-1,0,0,0,0,0,0,0,0,0,3,1,3,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,1,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,1,0,0,0,0,0,0,0
+"Netherlands",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,3,0,1,-1,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,1,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"New Zealand",0,3,0,3,1,3,3,3,3,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,-1,3,0,0,0,3,3,3,1,1,3,3,1,1,3,3,3,3,3,3,0,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,1,1,3,3,0,1,1,0,3,3,2,3,3,3,3,3,1,0,1,1
+"Nicaragua",0,3,0,3,0,0,3,1,0,3,0,3,0,1,3,3,3,3,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,0,1,0,0,0,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,0,3,3,1,0,0,3,0,3,0,3,0,1,0,3,3,0,3,3,2,3,1,0,3,0,3,0,0,1,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,0,1,1,0,3,3,3,0,3,0,1,0,0,0,1,3,0,-1,0,0,0,3,3,0,0,1,3,3,0,3,0,3,3,3,1,3,3,1,3,1,1,3,0,0,0,0,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,3,0,1,0,1,1,0,3,0,3,0,1,1,0,0,3,0,3,1,0,3,0,0,0,1,1
+"Niger",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,3,3,0,0,0,1,1,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,3,0,0,2,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,1,1,0,0,1,3,0,0,3,1,0,3,0,0,0,0,3,1,0,0,0,1,0,0,1,-1,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,0,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Nigeria",0,0,0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,0,3,0,1,0,0,0,0,0,3,0,1,3,0,3,0,3,0,0,0,1,0,0,0,3,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,1,3,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,1,3,0,0,0,0,0,1,0,0,1,3,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,0,0,0,0,3,-1,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,1,3,0,1,0,0,0,3,0,1,3,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,1,1,0,0,0,0,0,0,3,0,0,0,0,0,0
+"North Korea",0,0,0,0,0,0,0,1,0,0,0,0,0,1,0,3,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,0,0,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,0,0,0,1,3,3,0,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,0,3,1,0,0,1,0,0,0,0,0,1,1,1,0,1,0,0,0,1,0,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,0,0,-1,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"North Macedonia",0,3,0,3,0,3,3,1,0,3,1,3,0,1,0,3,3,0,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,0,2,3,3,3,3,3,1,3,3,3,1,3,0,0,3,0,1,0,3,3,1,3,0,3,0,3,0,3,0,1,0,3,3,3,3,3,2,3,1,0,0,3,3,1,3,1,0,1,0,3,0,1,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,0,1,1,0,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,-1,3,1,0,1,3,3,0,0,3,0,3,3,3,3,0,1,0,0,1,3,0,0,0,3,1,0,3,3,3,0,1,0,0,0,3,2,3,0,1,3,3,0,3,0,1,0,1,1,0,0,3,3,0,1,1,3,0,0,0,0,1,0,3,0,1,0,1,1
+"Norway",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,-1,1,0,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,1,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,1,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Oman",0,0,0,0,0,0,0,1,0,0,1,3,3,1,0,3,0,0,0,0,1,3,3,0,3,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,3,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,0,0,3,0,1,0,3,3,3,1,0,3,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,3,0,3,0,0,0,0,3,1,0,0,0,1,0,3,1,0,0,0,0,0,-1,1,1,3,0,0,1,0,3,0,0,3,0,0,1,0,0,1,0,0,3,1,3,1,0,3,0,0,0,1,0,3,0,0,2,3,0,1,0,0,3,1,1,1,3,1,1,0,0,3,1,0,1,1,0,3,2,0,0,1,3,0,0,0,1,0,0
+"Pakistan",0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,0,1,1,0,0,0,1,0,0,3,0,0,0,0,0,1,1,0,0,1,0,0,0,0,0,0,0,0,0,-1,1,0,0,0,0,0,0,0,0,3,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,1,0,3,0,1,0,1,1,0,0,0,0,0,0,3,0,0,0,0,0,0
+"Palau",0,0,0,3,0,0,0,1,0,3,0,0,0,1,0,0,3,3,0,0,1,3,0,0,0,3,0,0,1,0,0,1,0,0,0,0,3,1,0,0,3,2,3,1,3,3,3,1,3,0,3,1,0,0,0,3,0,1,3,3,3,1,3,0,3,0,3,0,0,0,1,0,3,0,3,3,3,2,3,1,0,0,3,3,0,0,1,0,1,3,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,3,1,1,3,3,3,3,0,0,0,1,0,0,1,1,3,0,1,0,0,0,0,3,0,0,-1,3,3,1,0,3,3,3,3,1,3,3,1,0,1,1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,3,0,1,0,1,1,1,0,0,1,0,1,1,0,0,3,3,0,1,3,3,0,1,0,0,1
+"Palestine",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,3,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,2,3,1,0,0,0,0,0,0,3,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,0,0,0,3,0,0,0,0,0,0,0,1,-1,0,0,0,0,0,0,0,1,0,0,1,0,1,1,0,0,0,1,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,3,0,0,0,1
+"Panama",0,3,0,3,0,3,3,1,0,3,1,3,0,1,3,3,3,3,0,0,3,3,0,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,0,3,3,1,0,3,3,0,3,0,3,0,1,3,0,3,3,3,3,2,3,1,0,3,3,3,3,0,1,0,1,0,3,0,0,1,3,1,1,0,0,3,3,3,1,1,1,3,1,0,3,0,1,1,3,3,3,3,0,3,0,1,1,0,0,1,3,0,3,0,0,0,3,3,0,0,1,3,-1,0,3,3,3,3,3,3,3,3,1,3,3,1,3,0,0,0,0,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,1,0,1,0,1,1,0,3,0,3,0,1,1,3,0,3,0,3,1,0,3,0,1,0,1,1
+"Papua New Guinea",0,0,0,0,0,3,0,1,0,0,0,3,0,3,0,0,0,3,0,0,1,0,3,0,0,0,0,0,1,0,2,1,0,0,0,0,3,1,0,0,0,2,0,1,0,0,0,1,3,3,3,1,0,0,0,0,3,1,3,0,0,1,3,0,0,0,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,0,3,0,3,0,0,0,3,0,3,0,0,1,0,0,3,0,0,0,0,0,1,1,1,3,1,0,0,1,1,3,0,3,0,0,0,0,0,1,0,0,1,1,0,0,1,0,0,0,0,0,0,0,1,3,3,-1,0,3,3,0,0,1,0,0,1,3,1,1,0,0,0,1,0,1,0,3,0,0,1,1,0,0,0,0,2,3,0,1,0,0,0,0,0,3,1,1,1,1,0,0,1,0,1,1,0,0,3,0,0,0,3,0,0,1,0,0,1
+"Paraguay",0,3,0,3,0,0,3,1,0,3,1,3,1,1,0,0,3,3,0,0,3,3,3,3,0,3,1,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,0,0,3,0,3,0,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,0,1,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,0,1,3,3,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,3,3,1,1,1,3,3,0,-1,3,3,3,3,3,3,3,1,3,1,1,3,0,0,0,3,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,3,0,1,0,1,1,0,0,0,3,0,1,1,3,3,3,0,3,1,0,3,3,0,0,1,1
+"Peru",0,3,0,3,0,3,3,1,0,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,0,0,0,3,0,1,3,3,3,1,0,0,3,0,3,0,3,0,1,3,3,3,3,3,3,2,3,1,0,0,3,3,3,0,1,0,1,0,0,0,0,1,3,1,1,0,0,3,3,3,1,1,1,3,1,0,3,1,1,1,3,3,3,3,0,3,3,1,1,0,0,1,3,0,1,0,0,0,3,3,0,0,1,3,3,1,3,-1,3,3,3,3,3,3,1,3,1,1,3,0,0,0,3,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,1,0,1,3,1,1,0,3,0,3,0,1,1,0,0,0,0,3,1,3,3,3,1,0,1,1
+"Philippines",0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,3,0,0,3,3,0,0,0,3,0,0,1,0,0,0,0,3,1,0,0,3,3,0,0,0,0,0,1,3,0,3,0,0,0,0,0,0,1,3,0,0,1,3,0,0,0,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,3,0,0,0,0,0,1,0,0,0,1,3,0,0,1,0,0,0,0,0,3,1,1,3,1,0,0,1,1,1,0,3,0,0,3,0,3,1,3,0,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,1,0,3,-1,0,0,1,0,0,3,0,1,1,0,0,0,1,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,3,0,0,0,3,1,1,3,1,1,0,1,0,0,0,1,1,0,0,0,0,0,1,3,0,0,3,0,0,0
+"Poland",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,-1,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,1,0,3,1,3,3,3,0,3,3,3,3,3,1,0,1,1
+"Portugal",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,1,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,-1,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,1,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Qatar",0,3,0,0,0,0,3,3,0,0,1,3,0,1,0,3,0,0,0,0,1,3,3,0,1,0,0,0,1,0,0,1,0,0,0,3,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,0,3,0,0,0,3,1,0,0,0,1,0,3,0,0,0,0,3,0,1,0,3,3,3,0,0,2,3,1,0,0,0,0,0,0,3,0,1,0,3,3,0,1,0,3,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,3,0,3,0,0,0,3,3,1,1,0,0,1,0,3,3,0,0,0,0,0,3,1,1,3,3,0,1,0,3,0,0,-1,0,0,1,3,0,1,0,3,0,0,3,1,0,3,0,0,0,1,0,3,0,0,2,3,3,1,0,0,3,1,1,1,3,1,1,0,0,3,3,0,1,1,3,0,2,0,0,1,3,0,0,1,1,0,0
+"Romania",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,3,3,3,0,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,1,3,1,3,1,3,3,1,1,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,-1,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,3,1,3,1,3,3,3,3,0,3,1,3,3,3,0,3,3,3,3,3,1,0,3,1
+"Russian Federation",0,3,0,0,1,3,3,3,0,0,3,3,1,1,3,3,0,0,0,0,3,3,3,3,3,0,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,0,3,0,0,0,1,3,3,3,1,3,0,0,0,3,1,3,0,0,1,3,3,0,0,0,3,3,0,1,3,3,3,3,0,0,2,3,1,0,0,3,0,3,0,1,3,1,0,0,0,3,3,0,1,1,0,0,0,0,0,3,1,0,3,1,0,0,1,1,3,2,3,3,0,3,3,3,1,1,3,1,1,0,0,3,0,0,0,3,0,1,0,3,3,3,0,3,3,3,0,0,3,0,-1,1,3,3,3,0,3,0,1,3,3,0,0,0,0,0,1,3,3,0,0,2,3,0,0,0,0,0,3,3,1,3,1,1,1,3,3,3,0,1,1,3,3,0,0,3,3,3,0,3,3,0,1,1
+"Rwanda",0,0,0,0,3,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,3,1,0,0,1,3,0,0,0,0,1,1,3,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,3,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,1,0,3,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,0,1,1,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,0,1,1,3,0,0,0,0,3,0,0,1,0,0,-1,0,0,1,0,3,0,1,0,1,0,3,0,0,0,1,0,0,1,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,1,1
+"Saint Kitts and Nevis",0,3,0,3,0,3,3,1,0,3,0,3,0,3,3,0,3,3,0,0,1,3,3,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,3,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,3,0,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,0,1,0,3,3,3,0,0,1,3,1,3,0,0,3,3,3,1,1,3,3,1,0,3,0,1,3,0,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,3,3,0,1,1,3,3,0,0,3,3,3,3,1,3,3,1,-1,3,1,3,0,0,1,3,1,0,3,3,3,1,1,0,3,0,3,2,3,0,3,3,3,0,3,0,3,0,1,1,1,3,3,3,0,1,1,3,0,3,0,3,1,3,3,3,0,0,3,3
+"Saint Lucia",0,0,0,3,0,3,3,1,0,3,0,3,0,1,3,0,3,3,0,0,1,3,3,0,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,3,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,0,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,0,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,1,1,3,3,1,0,3,0,1,3,0,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,3,0,1,1,3,3,0,0,3,3,3,3,1,3,0,1,3,-1,1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,3,3,3,0,3,0,3,0,1,1,1,3,3,1,0,1,1,0,0,3,0,0,1,3,3,3,0,0,3,3
+"Samoa",0,0,0,3,0,3,0,1,0,3,0,3,0,3,3,0,3,3,0,0,1,3,3,0,0,3,0,0,1,0,2,1,0,0,0,0,3,1,0,0,3,2,3,1,3,3,3,1,3,0,3,1,0,0,0,3,3,1,3,3,3,1,3,0,3,0,3,3,0,0,1,0,3,0,3,3,3,2,3,1,0,3,0,3,3,0,1,0,3,3,3,0,0,1,3,1,3,0,0,3,3,3,3,1,3,3,1,0,3,1,1,3,0,3,3,3,0,3,0,1,0,0,1,1,3,0,1,0,0,0,0,3,0,3,1,3,3,1,0,3,3,3,3,1,3,3,1,0,3,-1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,0,0,3,0,1,1,1,3,0,1,0,1,1,0,0,3,0,0,1,3,3,0,0,0,3,3
+"San Marino",0,3,0,3,0,3,3,3,2,3,1,3,1,1,3,3,3,0,0,0,1,3,3,3,0,3,0,0,1,0,2,3,0,0,3,3,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,0,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,3,0,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,0,1,3,3,3,3,3,0,3,3,1,0,0,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,0,3,3,3,3,3,3,0,1,0,3,1,-1,3,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,0,1,3,1,1,0,0,3,3,0,1,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Sao Tome and Principe",0,0,0,0,1,0,0,0,0,0,0,3,0,1,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,3,2,0,1,0,0,0,1,3,0,3,1,3,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,3,0,1,0,3,3,3,0,0,0,3,1,0,0,0,0,0,0,1,0,1,0,3,0,0,1,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,3,0,0,0,0,0,0,0,1,3,3,0,0,0,3,0,0,1,0,0,3,0,1,1,0,-1,0,0,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,1,1
+"Saudi Arabia",0,3,0,0,0,0,0,0,0,0,1,0,3,1,0,3,0,0,0,0,1,0,3,0,1,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,3,3,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,3,0,1,0,3,3,3,0,0,0,3,1,0,0,0,0,0,0,3,0,1,0,3,3,3,1,0,3,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,3,0,3,0,0,0,0,3,1,1,0,0,1,0,3,3,0,0,0,0,0,3,1,1,3,3,0,0,0,3,0,0,3,0,0,1,3,0,1,0,0,-1,0,0,1,0,0,0,0,0,1,0,3,0,0,2,3,0,0,0,0,3,1,1,1,1,1,1,0,0,3,1,0,1,1,0,3,0,0,0,1,3,0,0,0,1,0,0
+"Senegal",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,3,3,0,0,0,1,1,0,0,3,0,1,0,0,0,1,3,0,0,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,1,1,0,3,1,3,0,0,3,1,0,3,0,0,0,0,3,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,3,0,1,1,0,0,0,-1,0,1,3,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,0,0,1,3,0,0,3,0,0,1,1,0,0,0,0,0,1,0,0,0,0,0,0,1
+"Serbia",0,3,0,3,0,0,3,1,0,3,1,0,0,1,0,3,3,0,0,0,1,3,0,3,0,3,0,0,1,0,0,1,0,0,3,3,3,1,0,0,3,2,3,3,3,3,3,1,3,3,3,1,0,0,0,3,3,1,3,3,3,1,3,3,3,0,3,0,0,0,1,0,3,0,3,3,3,2,3,1,0,0,3,3,1,3,1,3,1,0,3,1,3,1,3,1,1,0,0,3,3,3,3,1,0,1,1,0,3,0,1,1,0,3,3,3,3,3,0,1,1,0,0,1,3,0,1,0,0,0,3,3,1,0,1,3,3,0,0,3,0,3,3,1,3,3,1,0,0,1,3,0,0,0,-1,1,0,3,3,3,0,1,0,3,0,3,2,3,0,3,3,3,0,0,0,1,0,1,1,0,3,3,3,0,1,1,3,0,0,0,3,3,3,3,0,1,0,3,1
+"Seychelles",0,3,3,3,3,3,0,0,0,3,1,3,0,3,3,3,3,3,3,0,1,3,3,3,0,3,0,0,1,0,0,1,0,0,0,3,0,1,0,0,3,3,3,1,3,3,3,1,3,3,3,1,0,0,0,3,3,1,3,3,3,1,3,3,3,1,3,3,0,0,1,0,3,0,3,3,3,2,3,1,0,3,0,3,3,0,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,3,1,3,3,1,0,3,1,1,3,0,3,3,3,0,3,0,1,0,3,0,1,3,3,1,0,3,0,3,3,1,0,1,3,3,0,0,0,3,3,3,3,3,3,3,3,3,1,3,0,0,1,3,-1,0,3,3,3,0,1,3,3,0,3,1,3,0,1,3,3,0,0,0,3,0,1,1,1,3,3,3,0,1,3,0,3,3,0,3,1,3,3,0,0,0,3,3
+"Sierra Leone",0,0,0,0,0,0,0,0,0,0,0,3,0,3,3,0,0,3,3,0,1,0,3,0,0,0,3,0,1,0,0,3,0,0,0,0,0,1,0,0,0,3,0,0,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,3,0,0,3,0,3,0,3,3,0,3,0,0,0,0,2,0,1,0,0,0,0,3,0,0,0,3,3,0,0,0,0,0,0,3,3,0,0,0,0,1,1,1,3,1,3,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,3,3,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,1,3,1,1,0,0,0,3,0,1,-1,3,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,1,3,0,3,0,0,0,1,3,0,3,0,0,0,0,3,0,0,0,0,0,0
+"Singapore",0,3,0,3,3,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,3,3,3,3,0,3,0,2,3,0,0,3,3,3,1,0,0,3,3,3,3,3,3,3,3,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,3,3,3,3,3,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,3,3,3,1,3,3,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,3,3,3,1,3,3,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,3,3,3,3,3,3,3,0,3,3,3,1,3,0,0,3,3,1,0,-1,3,3,1,1,3,3,0,3,1,3,0,1,3,3,0,3,1,3,3,1,1,1,3,3,3,0,1,3,0,3,3,2,3,3,3,3,0,3,0,3,3
+"Slovakia",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,-1,3,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Slovenia",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,-1,3,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Solomon Islands",0,0,0,3,0,3,0,1,0,3,0,3,0,1,3,0,3,3,0,0,1,3,3,0,0,3,0,0,1,0,2,1,0,0,0,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,0,3,0,3,3,3,0,1,0,3,3,0,3,3,2,3,1,0,3,3,3,3,0,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,1,1,1,3,1,0,3,1,1,3,0,3,3,3,0,0,0,1,0,0,1,1,3,0,3,0,0,0,0,3,0,0,1,3,3,1,0,3,3,3,3,1,3,0,1,3,3,1,3,0,0,0,0,1,0,3,3,3,-1,1,0,3,0,3,2,3,0,1,3,3,0,1,0,3,0,1,1,1,3,0,0,0,1,3,0,0,3,0,0,1,3,3,0,1,0,3,3
+"Somalia",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,0,0,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,-1,0,0,0,0,2,3,0,1,0,0,0,0,0,0,0,1,1,0,0,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"South Africa",0,0,0,0,3,3,3,1,0,0,1,3,0,0,3,0,0,3,3,0,1,0,3,3,0,0,0,0,1,0,0,1,0,0,3,0,0,1,0,0,3,2,0,1,0,0,0,1,3,3,3,0,3,0,0,0,3,1,3,0,0,1,0,3,0,1,0,3,3,0,1,3,3,3,3,0,0,2,3,1,0,3,3,0,3,0,1,0,3,0,3,0,1,1,0,0,3,0,0,0,0,0,3,1,3,3,1,0,0,1,1,3,0,3,0,0,0,0,0,3,1,3,0,1,0,0,3,0,0,0,0,0,1,0,1,3,3,0,3,3,3,0,0,3,0,3,1,3,3,1,0,0,0,1,0,1,0,3,0,0,0,1,-1,3,0,0,2,3,0,1,0,0,0,0,1,3,3,1,1,0,3,3,1,0,1,1,0,0,0,0,3,0,3,0,3,0,0,3,3
+"South Korea",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,3,3,3,0,0,1,0,2,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,3,3,1,3,3,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,0,3,3,1,3,0,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,3,3,3,3,3,3,3,3,1,3,3,1,3,3,0,3,3,1,0,3,3,3,1,1,3,-1,0,3,2,3,0,3,3,3,0,3,1,1,3,1,1,1,3,3,3,0,1,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"South Sudan",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,3,0,0,0,0,3,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,1,0,0,0,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,3,0,0,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,-1,0,2,3,0,0,0,0,0,0,0,1,0,1,1,0,0,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Spain",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,1,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,1,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,-1,2,3,0,1,3,3,0,3,1,1,3,3,1,3,3,3,1,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Sri Lanka",0,0,0,0,0,0,0,0,0,0,1,3,0,0,3,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,3,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,3,0,0,0,0,0,0,1,0,1,1,0,0,0,1,1,0,3,0,0,0,0,0,1,1,0,0,1,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,1,0,0,1,3,0,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,0,0,-1,3,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0
+"Saint Vincent and the Grenadines",0,0,0,3,0,3,3,1,0,3,0,3,0,1,3,3,3,3,0,0,1,3,3,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,3,3,3,3,1,3,3,3,1,3,0,0,3,0,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,0,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,1,1,3,3,1,0,3,1,1,3,0,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,3,0,0,1,3,3,0,0,3,3,3,3,1,3,3,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,-1,0,3,3,3,0,3,0,3,0,1,1,1,3,0,1,0,1,3,0,3,3,0,3,1,3,3,3,0,0,3,3
+"Sudan",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,0,0,0,0,1,3,0,3,0,0,0,1,0,0,1,0,0,0,1,0,0,0,1,0,0,0,0,1,0,3,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,1,0,0,1,0,1,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,3,-1,0,0,0,3,0,0,1,0,1,1,0,0,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Suriname",0,0,0,0,0,3,3,1,0,0,0,3,1,1,3,0,0,3,0,0,1,0,0,3,0,0,0,0,1,0,0,1,0,0,0,0,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,1,0,0,0,0,0,1,0,0,0,1,3,0,0,0,0,3,0,0,1,3,3,0,3,0,0,2,3,1,0,0,3,0,3,3,1,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,1,1,1,3,1,0,0,0,1,3,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,0,0,0,0,0,1,0,1,3,0,0,0,3,3,0,0,3,0,3,1,3,3,1,0,0,0,1,3,1,0,3,0,0,0,1,0,3,0,0,2,3,0,-1,0,0,0,0,0,1,0,1,1,0,3,0,1,0,1,1,0,0,0,0,0,1,0,0,0,0,0,1,1
+"Sweden",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,3,3,3,1,3,3,3,3,3,3,3,0,1,3,3,3,3,3,0,3,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,-1,3,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,3,0,1,1
+"Switzerland",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,3,1,3,3,1,3,3,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,1,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,1,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,1,3,1,0,3,3,3,3,1,3,3,0,3,2,3,0,1,3,-1,0,3,1,1,3,3,1,3,3,3,3,0,3,1,3,3,3,2,3,3,3,3,3,1,0,1,1
+"Syria",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,0,0,0,0,3,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,0,0,3,0,0,0,-1,0,0,0,0,1,1,0,0,0,1,0,1,1,0,0,0,0,0,0,0,0,0,0,1,0,0
+"Taiwan",0,3,0,3,0,3,0,1,2,3,0,0,0,1,3,0,3,3,0,0,1,3,0,0,1,3,1,0,1,0,2,1,0,0,3,1,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,0,0,0,3,3,1,3,3,3,1,3,0,3,0,3,0,3,0,1,0,3,3,0,3,3,2,3,1,0,3,3,3,1,3,1,0,1,3,3,0,0,1,3,0,1,0,0,3,3,3,3,1,1,3,1,0,3,3,1,1,0,3,0,3,0,3,0,1,1,0,1,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,2,3,3,1,3,0,1,3,3,1,3,0,0,0,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,-1,0,1,1,1,1,0,0,0,1,0,3,1,0,0,3,2,3,0,3,3,0,0,0,0,0
+"Tajikistan",0,0,0,0,0,3,0,3,0,0,3,0,0,1,3,3,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,1,0,1,3,0,0,0,0,3,1,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,1,1,1,0,3,3,0,0,0,0,1,0,3,0,1,0,0,1,0,0,0,0,0,0,1,1,3,0,0,0,0,3,0,0,1,0,3,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,-1,0,0,1,1,0,0,0,3,0,1,1,3,0,0,0,0,3,0,0,0,0,0,0,1
+"Tanzania",0,0,0,0,0,3,0,0,0,0,0,3,0,3,3,0,0,3,3,0,1,0,3,0,0,0,0,3,1,0,0,1,0,0,0,0,0,1,0,1,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,3,0,0,1,0,3,0,3,1,0,3,0,3,0,0,2,3,1,0,0,0,0,3,0,0,0,3,0,0,0,0,1,0,0,3,0,0,0,0,0,3,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,3,0,3,0,1,0,0,1,0,0,0,0,0,0,1,1,0,0,0,0,0,3,0,0,1,0,0,3,3,3,1,0,0,0,1,0,1,0,3,0,0,0,1,3,0,1,0,2,3,0,1,0,0,0,0,0,-1,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,3,3
+"Thailand",0,3,0,0,0,0,3,1,0,0,1,0,1,0,0,0,0,0,0,0,1,0,0,3,3,0,0,0,3,0,0,1,0,0,3,0,0,1,0,0,0,2,0,1,0,0,0,1,3,3,3,0,0,0,0,0,0,1,3,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,0,3,1,0,1,0,0,0,1,3,0,0,1,0,0,0,0,0,3,1,1,3,1,0,0,1,1,1,0,3,0,0,3,0,0,1,3,0,0,1,0,0,1,0,0,0,0,0,1,0,1,3,3,1,0,3,3,0,0,3,0,3,1,0,1,1,0,0,0,1,0,1,0,3,0,0,1,1,3,3,0,0,2,3,0,1,0,0,0,3,1,1,-1,1,1,0,0,0,3,0,1,1,0,0,0,0,0,1,3,0,0,3,0,1,0
+"Timor-Leste",0,0,0,3,1,0,0,1,0,3,0,0,0,1,3,0,3,0,0,0,1,3,0,0,0,3,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,3,1,3,3,3,1,3,0,3,1,0,0,0,3,3,1,0,3,3,1,3,0,3,0,3,0,0,0,1,0,3,0,3,3,3,2,3,1,0,0,0,3,0,0,1,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,0,1,1,0,3,3,3,0,3,0,1,0,0,0,1,3,0,0,0,0,0,0,3,0,0,1,3,0,0,0,0,0,3,3,1,3,0,1,0,1,1,3,3,0,0,0,1,0,3,3,3,0,1,0,0,0,3,2,3,0,1,3,3,0,0,0,1,0,-1,1,0,0,0,1,0,1,1,0,0,3,0,0,0,0,3,0,1,0,0,0
+"Togo",0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,3,0,1,0,0,0,0,0,3,0,1,0,0,3,0,3,0,0,0,1,1,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,0,0,3,0,0,0,3,3,0,3,0,0,0,0,0,3,1,0,0,0,0,0,0,0,0,1,0,0,0,0,1,0,0,1,3,0,0,0,0,1,1,0,3,1,3,0,0,1,1,0,3,0,0,0,0,0,1,0,0,0,1,0,0,1,3,3,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,3,0,1,3,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,0,1,0,1,-1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,0,0,0
+"Tonga",0,0,0,3,0,0,0,1,0,3,0,3,0,3,3,0,3,3,0,0,1,3,3,0,0,3,0,0,1,0,0,1,0,0,3,3,0,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,0,0,0,3,3,1,3,3,3,1,3,0,3,0,3,3,0,0,1,0,3,0,3,3,3,2,3,1,0,3,3,3,3,0,1,0,3,3,3,0,0,0,3,0,3,0,0,3,3,3,1,1,1,3,1,0,3,1,1,3,0,3,3,3,0,0,0,1,0,0,1,1,3,0,1,0,0,0,0,3,0,3,1,3,3,1,0,3,0,3,3,1,3,0,1,3,3,1,3,0,0,0,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,0,0,3,0,1,1,-1,3,0,0,0,1,3,0,0,3,0,0,1,3,3,0,0,0,3,3
+"Trinidad and Tobago",0,3,0,3,0,3,3,1,0,3,1,3,0,1,3,0,3,3,0,0,1,3,3,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,0,3,3,3,3,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,0,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,1,1,3,3,1,0,3,0,1,3,3,3,3,3,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,3,0,3,1,3,3,0,0,3,3,3,3,1,3,0,1,3,3,1,3,0,0,1,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,3,3,3,0,0,0,3,0,1,1,0,-1,0,3,0,1,1,0,0,3,0,3,1,3,3,3,0,0,3,3
+"Tunisia",0,0,3,0,0,0,0,0,0,0,0,0,0,1,3,0,0,3,3,0,1,0,0,3,0,0,1,0,1,0,0,1,0,0,0,0,0,1,0,0,0,3,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,3,0,0,3,0,0,0,1,0,0,0,3,1,0,3,0,3,0,0,0,3,1,0,0,0,0,0,3,1,0,1,3,0,0,0,1,0,1,1,0,3,0,0,0,1,1,0,3,1,3,0,0,3,3,0,3,0,0,0,0,3,1,0,0,0,1,0,0,1,3,0,0,0,0,0,1,1,3,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,3,3,1,0,0,0,0,0,1,0,3,0,0,2,3,0,1,0,0,3,0,1,1,0,1,1,0,0,-1,3,0,1,1,0,0,0,0,0,1,3,0,0,0,0,0,0
+"Turkey",0,3,0,0,0,3,3,1,0,0,1,3,1,1,3,3,0,3,0,0,3,3,3,3,3,0,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,0,1,0,0,0,1,3,3,3,0,3,0,0,0,3,1,3,0,0,1,3,3,0,0,0,0,3,0,1,0,3,3,3,0,0,0,3,3,3,0,0,0,3,3,3,3,1,0,3,1,3,1,0,1,1,0,1,0,0,0,3,1,0,3,1,0,0,1,1,3,2,3,3,0,3,3,3,1,1,0,0,1,0,0,3,0,0,0,3,0,1,1,1,3,3,0,3,3,3,0,0,3,0,0,1,3,3,1,0,3,0,1,3,1,0,3,0,0,0,1,3,3,0,0,2,3,1,1,0,0,0,1,1,1,3,1,1,1,3,3,-1,0,1,1,3,0,0,0,3,3,3,0,3,0,0,1,1
+"Turkmenistan",0,0,0,0,0,3,0,1,0,0,1,0,0,1,3,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,3,3,0,0,0,0,0,0,1,0,3,0,0,0,0,0,3,1,0,0,0,0,1,0,1,0,1,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,1,1,1,0,3,0,0,0,0,0,1,0,3,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,0,0,0,0,0,1,0,0,1,1,0,0,0,3,-1,1,1,0,0,0,0,0,0,0,0,0,0,0,0,1
+"Tuvalu",0,0,0,3,0,3,0,1,0,3,0,3,0,1,3,0,3,3,0,0,1,3,3,0,0,3,0,0,1,0,0,1,0,0,0,0,0,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,0,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,0,3,3,0,1,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,1,1,3,3,1,0,3,1,1,3,0,3,0,3,0,0,0,1,0,0,1,1,3,0,3,0,0,0,0,3,0,0,1,3,3,1,0,3,1,3,3,1,3,0,1,3,3,1,3,0,0,1,0,1,0,3,3,3,1,1,0,3,0,3,2,3,0,1,3,3,0,3,0,3,0,1,1,1,3,0,1,0,-1,1,0,0,3,0,0,0,3,3,0,0,0,3,3
+"Uganda",0,0,0,0,0,3,0,0,0,0,0,3,0,1,3,0,0,3,3,0,1,0,3,0,0,0,0,3,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,3,0,3,1,3,0,0,1,3,0,0,1,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,3,0,0,0,3,0,0,0,0,1,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,1,1,0,0,1,0,0,1,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,1,0,0,3,3,1,1,0,0,0,1,0,1,1,3,0,0,0,1,0,0,1,0,0,3,0,1,0,0,0,0,0,3,0,1,1,0,0,0,0,0,1,-1,0,0,0,0,0,0,3,0,0,0,0,3,3
+"Ukraine",0,3,0,3,0,3,3,3,0,3,3,0,0,1,3,3,3,0,0,0,1,3,0,3,3,3,0,0,1,0,0,1,0,0,3,0,0,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,0,3,0,1,0,3,3,3,3,3,2,3,1,0,0,3,3,3,0,1,3,1,0,0,1,3,1,3,1,1,0,0,3,3,3,1,1,0,3,1,0,3,1,1,3,2,3,3,3,3,3,0,1,1,3,0,1,3,0,3,0,0,0,3,3,1,0,1,3,3,0,3,3,0,3,3,3,3,3,1,3,0,1,3,0,0,1,3,1,0,0,3,3,0,1,0,0,0,3,2,3,0,1,3,3,0,0,3,1,3,1,1,1,0,0,3,0,1,1,-1,3,0,0,3,3,3,3,0,0,0,1,1
+"United Arab Emirates",0,3,0,3,1,3,3,3,1,3,1,3,3,1,3,3,3,0,0,0,1,3,3,3,3,3,3,1,1,0,2,1,0,3,3,3,3,1,1,1,3,1,3,1,3,3,3,1,3,3,3,3,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,1,1,1,3,3,3,3,3,2,3,1,0,3,0,3,0,3,3,3,1,0,3,3,3,1,3,3,1,1,0,3,3,3,1,1,1,3,1,3,3,1,1,3,3,3,3,3,3,3,3,1,0,0,3,1,3,3,3,1,0,0,3,3,3,1,1,3,3,0,1,0,3,3,3,3,3,3,1,3,3,1,3,3,3,1,3,3,1,3,3,3,3,1,0,3,3,3,1,3,3,1,3,3,3,1,1,1,3,1,1,3,0,3,1,0,1,1,3,-1,2,0,3,3,3,3,0,1,1,1,1
+"United Kingdom",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,3,3,1,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,0,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,0,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,3,3,3,3,3,0,1,3,3,1,3,3,0,3,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,1,3,3,1,0,1,1,3,3,-1,2,3,3,3,3,3,3,0,1,1
+"United States",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,1,3,3,0,3,3,1,0,1,0,3,1,3,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,3,0,3,3,1,3,3,3,1,1,3,3,0,3,3,3,0,1,3,3,3,3,3,3,2,3,0,0,3,3,3,3,3,1,3,1,3,3,1,3,1,3,1,3,0,0,3,3,3,3,1,1,3,1,0,3,3,1,3,3,3,3,3,3,3,3,1,1,3,0,1,3,3,3,0,0,0,3,3,1,0,3,3,3,1,1,3,3,3,3,3,3,0,1,3,3,1,3,3,0,3,3,1,0,3,3,3,1,1,3,3,0,3,2,3,0,1,3,3,0,3,1,1,3,1,1,1,3,3,1,0,1,1,3,3,3,-1,3,1,3,3,0,1,0,1,1
+"Uruguay",0,3,0,3,1,0,3,3,0,3,0,3,1,1,3,3,3,3,0,0,3,3,3,3,0,3,0,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,0,3,3,0,3,0,3,0,1,3,3,3,3,3,3,2,3,1,0,3,3,3,3,3,1,0,1,0,3,0,0,1,3,0,1,0,0,3,3,3,3,1,1,3,1,0,3,1,1,1,3,3,3,3,3,3,0,1,1,0,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,3,3,3,3,3,3,3,3,1,3,3,1,3,0,0,0,3,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,0,0,1,0,1,1,0,3,0,3,0,1,1,3,3,3,0,-1,1,3,3,3,1,0,1,1
+"Uzbekistan",0,0,0,0,0,3,0,3,0,0,3,0,0,0,3,3,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,3,0,0,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,1,0,1,3,1,0,0,0,3,1,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,1,0,3,3,0,0,0,0,1,1,3,0,1,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,3,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,3,0,1,1,1,0,0,0,3,0,1,1,3,0,0,0,0,-1,0,0,0,0,0,0,1
+"Vanuatu",0,0,0,3,0,3,0,1,0,3,0,3,0,1,3,3,3,3,0,0,1,3,3,0,0,3,0,0,1,0,0,1,0,0,0,0,0,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,0,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,3,0,0,0,3,3,3,0,0,1,3,0,3,0,0,3,3,3,1,1,1,3,1,0,3,1,1,3,0,3,3,3,0,3,0,1,0,0,1,1,3,0,3,0,0,0,0,3,0,0,1,3,3,1,0,3,3,3,3,1,3,3,1,3,3,1,3,0,0,0,0,1,0,3,3,3,1,1,0,0,0,3,2,3,0,1,3,3,0,0,0,3,1,1,1,1,3,0,1,0,1,3,0,0,3,0,0,1,-1,3,0,1,0,3,3
+"Vatican",0,3,0,3,1,3,3,3,2,3,1,3,1,1,3,3,3,3,0,0,3,3,3,3,0,3,0,0,1,0,2,3,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,3,0,0,3,3,1,3,3,3,1,3,3,3,0,3,3,3,0,1,0,3,3,3,3,3,2,3,1,0,3,3,3,0,0,1,0,1,0,3,1,3,1,3,0,1,0,0,3,3,3,1,1,1,3,1,0,3,1,1,3,0,3,3,3,0,3,0,1,0,0,0,1,3,3,3,0,0,0,3,3,1,0,1,3,3,1,0,3,3,3,3,3,3,0,1,3,0,1,3,3,0,1,3,1,0,3,3,3,0,1,0,3,0,3,2,3,0,1,3,3,0,3,1,1,0,1,1,0,0,3,3,0,1,1,3,3,3,0,3,3,3,-1,0,0,0,1,1
+"Venezuela",0,3,0,3,1,3,3,1,0,3,0,3,1,1,3,3,3,3,0,0,3,3,3,3,0,3,1,0,1,0,0,1,0,0,3,0,3,1,0,0,3,2,3,1,3,3,3,1,3,3,3,1,0,0,0,3,0,1,3,3,3,1,3,0,3,0,3,3,0,0,1,0,3,0,3,3,3,2,3,3,0,0,0,3,3,0,1,0,1,0,3,0,1,1,3,1,1,0,0,3,3,3,1,1,1,3,1,0,3,0,1,1,3,3,3,3,0,3,0,1,1,0,0,1,3,0,3,0,0,0,3,3,1,0,1,3,0,0,3,3,3,3,3,3,3,3,1,3,0,1,3,0,0,0,0,1,0,3,3,3,0,1,3,3,0,3,2,3,0,1,3,3,0,0,1,1,0,1,1,0,3,0,3,0,1,1,0,0,0,0,3,1,0,3,-1,1,0,1,1
+"Viet Nam",0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,0,0,0,3,0,0,0,3,0,0,1,0,0,3,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,0,0,0,2,3,1,0,0,0,0,0,0,0,0,1,0,0,1,3,3,0,0,1,0,0,0,0,0,0,1,1,3,1,0,0,1,1,1,0,3,0,0,0,0,0,1,3,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,3,0,0,0,3,0,0,1,0,0,1,0,1,1,0,0,0,0,0,1,0,3,0,0,0,1,0,0,0,0,2,3,0,1,0,0,0,0,1,1,3,1,1,0,0,0,0,0,1,1,0,0,0,0,0,1,0,0,0,-1,0,1,0
+"Yemen",0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,1,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,1,1,0,0,0,0,0,1,1,0,3,1,0,0,0,1,0,0,3,0,0,0,0,0,1,0,0,0,1,0,0,0,0,0,0,0,0,0,0,1,3,0,0,0,0,0,0,0,1,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,0,0,1,0,0,0,0,2,3,3,0,0,0,3,0,1,0,0,1,1,0,0,0,0,0,1,1,0,0,0,0,0,0,0,0,0,0,-1,0,0
+"Zambia",0,0,0,0,3,3,0,0,0,0,0,3,0,3,3,0,0,3,3,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,0,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,3,0,0,1,0,3,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,3,0,0,0,3,3,0,0,0,0,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,3,0,3,0,1,0,0,1,0,0,0,0,0,0,1,1,0,0,0,0,0,3,0,0,1,0,0,1,3,3,1,0,0,0,1,0,1,0,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,3,0,1,0,1,3,0,0,0,0,0,0,3,0,0,0,0,-1,3
+"Zimbabwe",0,0,0,0,3,0,0,0,0,0,0,3,0,1,3,0,0,3,3,0,1,0,3,0,0,0,0,0,1,0,0,1,0,0,0,0,0,1,0,3,0,2,0,1,0,0,0,1,3,0,3,0,0,0,0,0,3,1,3,0,0,1,3,0,0,3,0,0,0,0,1,0,3,0,3,0,0,2,3,1,0,0,0,0,0,0,1,0,3,3,0,0,0,0,0,0,3,0,0,0,0,0,1,1,3,3,1,0,0,0,1,3,0,3,0,0,0,0,0,3,0,3,0,0,0,0,1,0,0,0,0,0,0,0,1,3,0,0,0,0,3,0,0,1,0,0,1,3,1,1,0,0,0,1,0,1,0,3,0,0,0,1,3,0,0,0,2,3,0,1,0,0,0,0,0,3,0,1,1,0,3,0,0,0,1,3,0,0,0,0,0,0,3,0,0,0,0,3,-1
diff --git a/previews/PR3525/tutorials/getting_started/debugging.jl b/previews/PR3525/tutorials/getting_started/debugging.jl
new file mode 100644
index 00000000000..a76a09570d8
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/debugging.jl
@@ -0,0 +1,355 @@
+# Copyright (c) 2021 Oscar Dowson and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Debugging
+
+# Dealing with bugs is an unavoidable part of coding optimization models in
+# any framework, including JuMP. Sources of bugs include not only generic coding
+# errors (method errors, typos, off-by-one issues), but also semantic mistakes
+# in the formulation of an optimization problem and the incorrect use of a
+# solver.
+
+# This tutorial explains some common sources of bugs and modeling issues that
+# you might encounter when writing models in JuMP, and it suggests a variety of
+# strategies to deal with them.
+
+# !!! tip
+#     This tutorial is more advanced than the other "Getting started" tutorials.
+#     It's in the "Getting started" section to give you an early preview of how
+#     to debug JuMP models. However, if you are new to JuMP, you may want to
+#     briefly skim the tutorial, and come back to it once you have written a few
+#     JuMP models.
+
+using JuMP
+import HiGHS
+
+# ## Getting help
+
+# Debugging can be a frustrating part of modeling, particularly if you're new to
+# optimization and programming. If you're stuck, join the
+# [community forum](https://jump.dev/forum) to search for answers to commonly
+# asked questions.
+
+# Before asking a new question, make sure to read the post
+# [Make it easier to help you](https://discourse.julialang.org/t/please-read-make-it-easier-to-help-you/14757),
+# which contains a number of tips on how to ask a good question.
+
+# Above all else, take time to simplify your code as much as possible. The
+# fewer lines of code you can post that reproduces the same issue, the faster
+# someone can answer your question.
+
+# ## Debugging Julia code
+
+# Read the [Debugging chapter](https://benlauwens.github.io/ThinkJulia.jl/latest/book.html#chap21)
+# in the book [ThinkJulia.jl](https://benlauwens.github.io/ThinkJulia.jl/latest/book.html).
+# It has a number of great tips and tricks for debugging Julia code.
+
+# ## Solve failures
+
+# When a solver experiences an issue that prevents it from finding an optimal
+# solution (or proving that one does not exist), JuMP may return one of a number
+# of [`termination_status`](@ref)es.
+
+# For example, if the solver found a solution, but experienced numerical
+# imprecision, it may return a status such as `ALMOST_OPTIMAL` or
+# `ALMOST_LOCALLY_SOLVED` indicating that the problem was solved to a relaxed
+# set of tolerances. Alternatively, the solver may return a problematic status
+# such as `NUMERICAL_ERROR`, `SLOW_PROGRESS`, or `OTHER_ERROR`, indicating that
+# it could not find a solution to the problem.
+
+# Most solvers can experience numerical imprecision because they use
+# [floating-point arithmetic](https://en.wikipedia.org/wiki/Floating-point_arithmetic)
+# to perform operations such as addition, subtraction, and multiplication. These
+# operations aren't exact, and small errors can accrue between the theoretical
+# value and the value that the computer computes. For example:
+
+0.1 * 3 == 0.3
+
+# !!! tip
+#     Read the [Guidelines for numerical issues](https://www.gurobi.com/documentation/9.5/refman/guidelines_for_numerical_i.html)
+#     section of the Gurobi documentation, along with the
+#     [Debugging numerical problems](https://yalmip.github.io/inside/debuggingnumerics/)
+#     section of the YALMIP documentation.
+
+# ### Common sources
+
+# Common sources of solve failures are:
+#
+#  * Very large numbers and very small numbers as problem coefficients. Exactly
+#    what "large" is depends on the solver and the problem, but in general,
+#    values above 1e6 or smaller than 1e-6 cause problems.
+#  * Nonlinear problems with functions that are not defined in parts of their
+#    domain. For example, minimizing `log(x)` where `x >= 0` is undefined when
+#    `x = 0` (a common starting value).
+
+# ### Strategies
+
+# Strategies to debug sources of solve failures include:
+#
+#  * Rescale variables in the problem and their associated coefficients to
+#    make the magnitudes of all coefficients in the 1e-4 to 1e4 range. For
+#    example, that might mean rescaling a variable from measuring distance in
+#    centimeters to kilometers.
+#  * Try a different solver. Some solvers might be more robust than others for a
+#    particular problem.
+#  * Read the documentation of your solver, and try settings that encourage
+#    numerical robustness.
+#  * Set bounds or add constraints so that all nonlinear functions are defined
+#    across all of the feasible region. This particularly applies for functions
+#    like `1 / x` and `log(x)` which are not defined for `x = 0`.
+
+# ## Incorrect results
+
+# Sometimes, you might find that the solver returns an "optimal" solution that
+# is incorrect according to the model you are trying to solve (perhaps the
+# solution is suboptimal, or it doesn't satisfy some of the constraints).
+
+# Incorrect results can be hard to detect and debug, because the solver gives no
+# hints that there is a problem. Indeed, the [`termination_status`](@ref) will
+# likely be `OPTIMAL` and a solution will be available.
+
+# ### Common sources
+
+# Common sources of incorrect results are:
+#
+#  * A modeling error, so that your JuMP model does not match the formulation
+#    you have on paper
+#  * Not accounting for the tolerances that solvers use (for example, if `x` is
+#    binary, a value like `x = 1.0000001` may still be considered feasible)
+#  * A bug in JuMP or the solver.
+
+# The probability of the issue being a bug in JuMP or the solver is much smaller
+# than a modeling error. When in doubt, first assume there is a bug in your code
+# before assuming that there is a bug in JuMP.
+
+# ### Strategies
+
+# Strategies to debug sources of incorrect results include:
+
+#  * Print your JuMP model to see if it matches the formulation you have on
+#    paper. Look out for incorrect signs `+` instead of `-`, and off-by-one
+#    errors such as `x[t]` instead of `x[t-1]`.
+#  * Check that you are not using exact comparisons like `value(x) == 1.0`;
+#    always use `isapprox(value(x), 1.0; atol = 1e-6)` where you manually
+#    specify the comparison tolerance.
+#  * Try a different solver. If one solver succeeds where another doesn't this
+#    is a sign that the problem is a numerical issue or a bug in the solver.
+
+# ## Debugging an infeasible model
+
+# A model is infeasible if there is no primal solution that satisfies all of
+# the constraints. In general, an infeasible model means one of two things:
+#
+#  * Your problem really has no feasible solution
+#  * There is a mistake in your model.
+
+# ### Example
+
+# A simple example of an infeasible model is:
+
+model = Model(HiGHS.Optimizer);
+set_silent(model)
+@variable(model, x >= 0)
+@objective(model, Max, 2x + 1)
+@constraint(model, con, 2x - 1 <= -2)
+
+# because the bound says that `x >= 0`, but we can rewrite the constraint to be
+# `x <= -1/2`. When the problem is infeasible, JuMP may return one of a number
+# of statuses. The most common is `INFEASIBLE`:
+
+optimize!(model)
+termination_status(model)
+
+# Depending on the solver, you may also receive `INFEASIBLE_OR_UNBOUNDED` or
+# `LOCALLY_INFEASIBLE`.
+
+# A termination status of `INFEASIBLE_OR_UNBOUNDED` means that the solver could
+# not prove if the solver was infeasible or unbounded, only that the model does
+# not have a finite feasible optimal solution.
+
+# Nonlinear optimizers such as Ipopt may return the status `LOCALLY_INFEASIBLE`.
+# This does not mean that the solver _proved_ no feasible solution exists, only
+# that it could not find one. If you know a primal feasible point, try providing
+# it as a starting point using [`set_start_value`](@ref) and re-optimize.
+
+# ### Common sources
+
+# Common sources of infeasibility are:
+#
+#  * Incorrect units, for example, using a lower bound of megawatts and an upper
+#    bound of kilowatts
+#  * Using `+` instead of `-` in a constraint
+#  * Off-by-one and related errors, for example, using `x[t]` instead of
+#    `x[t-1]` in part of a constraint
+#  * Otherwise invalid mathematical formulations
+
+# ### Strategies
+
+# Strategies to debug sources of infeasibility include:
+
+#  * Iteratively comment out a constraint (or block of constraints) and re-solve
+#    the problem. When you find a constraint that makes the problem infeasible
+#    when added, check the constraint carefully for errors.
+#  * If the problem is still infeasible with all constraints commented out,
+#    check all variable bounds. Do they use the right data?
+#  * If you have a known feasible solution, use [`primal_feasibility_report`](@ref)
+#    to evaluate the constraints and check for violations. You'll probably find
+#    that you have a typo in one of the constraints.
+#  * Try a different solver. Sometimes, solvers have bugs, and they can
+#    incorrectly report a problem as infeasible when it isn't. If you find such
+#    a case where one solver reports the problem is infeasible and another can
+#    find an optimal solution, please report it by opening an issue on the
+#    GitHub repository of the solver that reports infeasibility.
+
+# !!! tip
+#     Some solvers also have specialized support for debugging sources of
+#     infeasibility via an [irreducible infeasible subsystem](@ref Conflicts).
+#     To see if your solver has support, try calling [`compute_conflict!`](@ref):
+#     ```julia
+#     julia> compute_conflict!(model)
+#     ERROR: ArgumentError: The optimizer HiGHS.Optimizer does not support `compute_conflict!`
+#     ```
+#     In this case, HiGHS does not support computing conflicts, but other
+#     solvers such as Gurobi and CPLEX do. If the solver does support computing
+#     conflicts, read [Conflicts](@ref) for more details.
+
+# ### Penalty relaxation
+
+# Another strategy to debug sources of infeasibility is the
+# [`relax_with_penalty!`](@ref) function.
+#
+# The penalty relaxation modifies constraints of the form ``f(x) \in S`` into
+# ``f(x) + y - z \in S``, where ``y, z \ge 0``, and then it introduces a
+# penalty term into the objective of ``a \times (y + z)`` (if minimizing, else
+# ``-a``), where ``a`` is a penalty.
+
+map = relax_with_penalty!(model)
+
+# Here `map` is a dictionary which maps constraint indices to an affine
+# expression representing ``(y + z)``.
+
+# If we optimize the relaxed model, this time we get a feasible solution:
+
+optimize!(model)
+termination_status(model)
+
+# Iterate over the contents of `map` to see which constraints are violated:
+
+for (con, penalty) in map
+    violation = value(penalty)
+    if violation > 0
+        println("Constraint `$(name(con))` is violated by $violation")
+    end
+end
+
+# Once you find a violated constraint in the relaxed problem, take a look to see
+# if there is a typo or other common mistake in that particular constraint.
+
+# Consult the docstring [`relax_with_penalty!`](@ref) for information on how to
+# modify the penalty cost term `a`, either for every constraint in the model or
+# a particular subset of the constraints.
+
+# When using [`relax_with_penalty!`](@ref), you should be aware that:
+#
+#  * Variable bounds and integrality restrictions are not relaxed. If the
+#    problem is still infeasible after calling [`relax_with_penalty!`](@ref),
+#    check the variable bounds.
+#  * You cannot undo the penalty relaxation. If you need an unmodified model,
+#    rebuild the problem, or call [`copy_model`](@ref) before calling
+#    [`relax_with_penalty!`](@ref).
+
+# ## Debugging an unbounded model
+
+# A model is unbounded if there is no limit on how good the objective value can
+# get. Most often, an unbounded model means that you have an error in your
+# modeling, because all physical systems have limits. (You cannot make an
+# infinite amount of profit.)
+
+# ### Example
+
+# A simple example of an unbounded model is:
+
+model = Model(HiGHS.Optimizer);
+set_silent(model)
+@variable(model, x >= 0)
+@objective(model, Max, 2x + 1)
+
+# because we can increase `x` without limit, and the objective value `2x + 1`
+# gets better as `x` increases.
+
+# When the problem is unbounded, JuMP may return one of a number of statuses.
+# The most common is `DUAL_INFEASIBLE`:
+
+optimize!(model)
+termination_status(model)
+
+# Depending on the solver, you may also receive `INFEASIBLE_OR_UNBOUNDED` or
+# an error code like `NORM_LIMIT`.
+
+# ### Common sources
+
+# Common sources of unboundedness are:
+#
+#  * Using `Max` instead of `Min`
+#  * Omitting variable bounds, such as `0 <= x <= 1`
+#  * Using `+` instead of `-` in a term of the objective function.
+
+# ### Strategies
+
+# Strategies to debug sources of unboundedness include:
+
+#  * Double check whether you intended `Min` or `Max` in the [`@objective`](@ref)
+#    line.
+#  * Print the objective function with `print(objective_function(model))` and
+#    verify that the value and sign of each coefficient is as you expect.
+#  * Add large bounds to all variables that are free or have one-sided bounds,
+#    then re-solve the problem. Because all variables are now bounded, the
+#    problem will have a finite optimal solution. Look at the value of each
+#    variable in the optimal solution to see if it is at one of the new bounds.
+#    If it is, you either need to specify a better bound for that variable, or
+#    there might be a mistake in the objective function associated with that
+#    variable (for example, a `+` instead of a `-`).
+
+# If there are too many variables to add bounds to, or there are too many terms
+# to examine by hand, another strategy is to create a new variable with a large
+# upper bound (if maximizing, lower bound if minimizing) and a constraint that
+# the variable must be less-than or equal to the expression of the objective
+# function. For example:
+
+model = Model(HiGHS.Optimizer);
+set_silent(model)
+@variable(model, x >= 0)
+## @objective(model, Max, 2x + 1)
+@variable(model, objective <= 10_000)
+@constraint(model, objective <= 2x + 1)
+@objective(model, Max, objective)
+
+# This new model has a finite optimal solution, so we can solve it and then look
+# for variables with large positive or negative values in the optimal solution.
+
+optimize!(model)
+for var in all_variables(model)
+    if var == objective
+        continue
+    end
+    if abs(value(var)) > 1e3
+        println("Variable `$(name(var))` may be unbounded")
+    end
+end
diff --git a/previews/PR3525/tutorials/getting_started/debugging/index.html b/previews/PR3525/tutorials/getting_started/debugging/index.html
new file mode 100644
index 00000000000..33466917ea2
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/debugging/index.html
@@ -0,0 +1,23 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Debugging · JuMP</title><meta name="title" content="Debugging · JuMP"/><meta property="og:title" content="Debugging · JuMP"/><meta property="twitter:title" content="Debugging · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li class="is-active"><a class="tocitem" href>Debugging</a><ul class="internal"><li><a class="tocitem" href="#Getting-help"><span>Getting help</span></a></li><li><a class="tocitem" href="#Debugging-Julia-code"><span>Debugging Julia code</span></a></li><li><a class="tocitem" href="#Solve-failures"><span>Solve failures</span></a></li><li><a class="tocitem" href="#Incorrect-results"><span>Incorrect results</span></a></li><li><a class="tocitem" href="#Debugging-an-infeasible-model"><span>Debugging an infeasible model</span></a></li><li><a class="tocitem" href="#Debugging-an-unbounded-model"><span>Debugging an unbounded model</span></a></li></ul></li><li><a class="tocitem" href="../design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Debugging</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Debugging</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/debugging.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Debugging"><a class="docs-heading-anchor" href="#Debugging">Debugging</a><a id="Debugging-1"></a><a class="docs-heading-anchor-permalink" href="#Debugging" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../debugging.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>Dealing with bugs is an unavoidable part of coding optimization models in any framework, including JuMP. Sources of bugs include not only generic coding errors (method errors, typos, off-by-one issues), but also semantic mistakes in the formulation of an optimization problem and the incorrect use of a solver.</p><p>This tutorial explains some common sources of bugs and modeling issues that you might encounter when writing models in JuMP, and it suggests a variety of strategies to deal with them.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This tutorial is more advanced than the other &quot;Getting started&quot; tutorials. It&#39;s in the &quot;Getting started&quot; section to give you an early preview of how to debug JuMP models. However, if you are new to JuMP, you may want to briefly skim the tutorial, and come back to it once you have written a few JuMP models.</p></div></div><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; import HiGHS</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><h2 id="Getting-help"><a class="docs-heading-anchor" href="#Getting-help">Getting help</a><a id="Getting-help-1"></a><a class="docs-heading-anchor-permalink" href="#Getting-help" title="Permalink"></a></h2><p>Debugging can be a frustrating part of modeling, particularly if you&#39;re new to optimization and programming. If you&#39;re stuck, join the <a href="https://jump.dev/forum">community forum</a> to search for answers to commonly asked questions.</p><p>Before asking a new question, make sure to read the post <a href="https://discourse.julialang.org/t/please-read-make-it-easier-to-help-you/14757">Make it easier to help you</a>, which contains a number of tips on how to ask a good question.</p><p>Above all else, take time to simplify your code as much as possible. The fewer lines of code you can post that reproduces the same issue, the faster someone can answer your question.</p><h2 id="Debugging-Julia-code"><a class="docs-heading-anchor" href="#Debugging-Julia-code">Debugging Julia code</a><a id="Debugging-Julia-code-1"></a><a class="docs-heading-anchor-permalink" href="#Debugging-Julia-code" title="Permalink"></a></h2><p>Read the <a href="https://benlauwens.github.io/ThinkJulia.jl/latest/book.html#chap21">Debugging chapter</a> in the book <a href="https://benlauwens.github.io/ThinkJulia.jl/latest/book.html">ThinkJulia.jl</a>. It has a number of great tips and tricks for debugging Julia code.</p><h2 id="Solve-failures"><a class="docs-heading-anchor" href="#Solve-failures">Solve failures</a><a id="Solve-failures-1"></a><a class="docs-heading-anchor-permalink" href="#Solve-failures" title="Permalink"></a></h2><p>When a solver experiences an issue that prevents it from finding an optimal solution (or proving that one does not exist), JuMP may return one of a number of <a href="../../../api/JuMP/#JuMP.termination_status"><code>termination_status</code></a>es.</p><p>For example, if the solver found a solution, but experienced numerical imprecision, it may return a status such as <code>ALMOST_OPTIMAL</code> or <code>ALMOST_LOCALLY_SOLVED</code> indicating that the problem was solved to a relaxed set of tolerances. Alternatively, the solver may return a problematic status such as <code>NUMERICAL_ERROR</code>, <code>SLOW_PROGRESS</code>, or <code>OTHER_ERROR</code>, indicating that it could not find a solution to the problem.</p><p>Most solvers can experience numerical imprecision because they use <a href="https://en.wikipedia.org/wiki/Floating-point_arithmetic">floating-point arithmetic</a> to perform operations such as addition, subtraction, and multiplication. These operations aren&#39;t exact, and small errors can accrue between the theoretical value and the value that the computer computes. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 0.1 * 3 == 0.3</code><code class="nohighlight hljs ansi" style="display:block;">false</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Read the <a href="https://www.gurobi.com/documentation/9.5/refman/guidelines_for_numerical_i.html">Guidelines for numerical issues</a> section of the Gurobi documentation, along with the <a href="https://yalmip.github.io/inside/debuggingnumerics/">Debugging numerical problems</a> section of the YALMIP documentation.</p></div></div><h3 id="Common-sources"><a class="docs-heading-anchor" href="#Common-sources">Common sources</a><a id="Common-sources-1"></a><a class="docs-heading-anchor-permalink" href="#Common-sources" title="Permalink"></a></h3><p>Common sources of solve failures are:</p><ul><li>Very large numbers and very small numbers as problem coefficients. Exactly what &quot;large&quot; is depends on the solver and the problem, but in general, values above 1e6 or smaller than 1e-6 cause problems.</li><li>Nonlinear problems with functions that are not defined in parts of their domain. For example, minimizing <code>log(x)</code> where <code>x &gt;= 0</code> is undefined when <code>x = 0</code> (a common starting value).</li></ul><h3 id="Strategies"><a class="docs-heading-anchor" href="#Strategies">Strategies</a><a id="Strategies-1"></a><a class="docs-heading-anchor-permalink" href="#Strategies" title="Permalink"></a></h3><p>Strategies to debug sources of solve failures include:</p><ul><li>Rescale variables in the problem and their associated coefficients to make the magnitudes of all coefficients in the 1e-4 to 1e4 range. For example, that might mean rescaling a variable from measuring distance in centimeters to kilometers.</li><li>Try a different solver. Some solvers might be more robust than others for a particular problem.</li><li>Read the documentation of your solver, and try settings that encourage numerical robustness.</li><li>Set bounds or add constraints so that all nonlinear functions are defined across all of the feasible region. This particularly applies for functions like <code>1 / x</code> and <code>log(x)</code> which are not defined for <code>x = 0</code>.</li></ul><h2 id="Incorrect-results"><a class="docs-heading-anchor" href="#Incorrect-results">Incorrect results</a><a id="Incorrect-results-1"></a><a class="docs-heading-anchor-permalink" href="#Incorrect-results" title="Permalink"></a></h2><p>Sometimes, you might find that the solver returns an &quot;optimal&quot; solution that is incorrect according to the model you are trying to solve (perhaps the solution is suboptimal, or it doesn&#39;t satisfy some of the constraints).</p><p>Incorrect results can be hard to detect and debug, because the solver gives no hints that there is a problem. Indeed, the <a href="../../../api/JuMP/#JuMP.termination_status"><code>termination_status</code></a> will likely be <code>OPTIMAL</code> and a solution will be available.</p><h3 id="Common-sources-2"><a class="docs-heading-anchor" href="#Common-sources-2">Common sources</a><a class="docs-heading-anchor-permalink" href="#Common-sources-2" title="Permalink"></a></h3><p>Common sources of incorrect results are:</p><ul><li>A modeling error, so that your JuMP model does not match the formulation you have on paper</li><li>Not accounting for the tolerances that solvers use (for example, if <code>x</code> is binary, a value like <code>x = 1.0000001</code> may still be considered feasible)</li><li>A bug in JuMP or the solver.</li></ul><p>The probability of the issue being a bug in JuMP or the solver is much smaller than a modeling error. When in doubt, first assume there is a bug in your code before assuming that there is a bug in JuMP.</p><h3 id="Strategies-2"><a class="docs-heading-anchor" href="#Strategies-2">Strategies</a><a class="docs-heading-anchor-permalink" href="#Strategies-2" title="Permalink"></a></h3><p>Strategies to debug sources of incorrect results include:</p><ul><li>Print your JuMP model to see if it matches the formulation you have on paper. Look out for incorrect signs <code>+</code> instead of <code>-</code>, and off-by-one errors such as <code>x[t]</code> instead of <code>x[t-1]</code>.</li><li>Check that you are not using exact comparisons like <code>value(x) == 1.0</code>; always use <code>isapprox(value(x), 1.0; atol = 1e-6)</code> where you manually specify the comparison tolerance.</li><li>Try a different solver. If one solver succeeds where another doesn&#39;t this is a sign that the problem is a numerical issue or a bug in the solver.</li></ul><h2 id="Debugging-an-infeasible-model"><a class="docs-heading-anchor" href="#Debugging-an-infeasible-model">Debugging an infeasible model</a><a id="Debugging-an-infeasible-model-1"></a><a class="docs-heading-anchor-permalink" href="#Debugging-an-infeasible-model" title="Permalink"></a></h2><p>A model is infeasible if there is no primal solution that satisfies all of the constraints. In general, an infeasible model means one of two things:</p><ul><li>Your problem really has no feasible solution</li><li>There is a mistake in your model.</li></ul><h3 id="Example"><a class="docs-heading-anchor" href="#Example">Example</a><a id="Example-1"></a><a class="docs-heading-anchor-permalink" href="#Example" title="Permalink"></a></h3><p>A simple example of an infeasible model is:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer);</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_silent(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Max, 2x + 1)</code><code class="nohighlight hljs ansi" style="display:block;">2 x + 1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, con, 2x - 1 &lt;= -2)</code><code class="nohighlight hljs ansi" style="display:block;">con : 2 x ≤ -1</code></pre><p>because the bound says that <code>x &gt;= 0</code>, but we can rewrite the constraint to be <code>x &lt;= -1/2</code>. When the problem is infeasible, JuMP may return one of a number of statuses. The most common is <code>INFEASIBLE</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; termination_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">INFEASIBLE::TerminationStatusCode = 2</code></pre><p>Depending on the solver, you may also receive <code>INFEASIBLE_OR_UNBOUNDED</code> or <code>LOCALLY_INFEASIBLE</code>.</p><p>A termination status of <code>INFEASIBLE_OR_UNBOUNDED</code> means that the solver could not prove if the solver was infeasible or unbounded, only that the model does not have a finite feasible optimal solution.</p><p>Nonlinear optimizers such as Ipopt may return the status <code>LOCALLY_INFEASIBLE</code>. This does not mean that the solver <em>proved</em> no feasible solution exists, only that it could not find one. If you know a primal feasible point, try providing it as a starting point using <a href="../../../api/JuMP/#JuMP.set_start_value"><code>set_start_value</code></a> and re-optimize.</p><h3 id="Common-sources-3"><a class="docs-heading-anchor" href="#Common-sources-3">Common sources</a><a class="docs-heading-anchor-permalink" href="#Common-sources-3" title="Permalink"></a></h3><p>Common sources of infeasibility are:</p><ul><li>Incorrect units, for example, using a lower bound of megawatts and an upper bound of kilowatts</li><li>Using <code>+</code> instead of <code>-</code> in a constraint</li><li>Off-by-one and related errors, for example, using <code>x[t]</code> instead of <code>x[t-1]</code> in part of a constraint</li><li>Otherwise invalid mathematical formulations</li></ul><h3 id="Strategies-3"><a class="docs-heading-anchor" href="#Strategies-3">Strategies</a><a class="docs-heading-anchor-permalink" href="#Strategies-3" title="Permalink"></a></h3><p>Strategies to debug sources of infeasibility include:</p><ul><li>Iteratively comment out a constraint (or block of constraints) and re-solve the problem. When you find a constraint that makes the problem infeasible when added, check the constraint carefully for errors.</li><li>If the problem is still infeasible with all constraints commented out, check all variable bounds. Do they use the right data?</li><li>If you have a known feasible solution, use <a href="../../../api/JuMP/#JuMP.primal_feasibility_report"><code>primal_feasibility_report</code></a> to evaluate the constraints and check for violations. You&#39;ll probably find that you have a typo in one of the constraints.</li><li>Try a different solver. Sometimes, solvers have bugs, and they can incorrectly report a problem as infeasible when it isn&#39;t. If you find such a case where one solver reports the problem is infeasible and another can find an optimal solution, please report it by opening an issue on the GitHub repository of the solver that reports infeasibility.</li></ul><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Some solvers also have specialized support for debugging sources of infeasibility via an <a href="../../../manual/solutions/#Conflicts">irreducible infeasible subsystem</a>. To see if your solver has support, try calling <a href="../../../api/JuMP/#JuMP.compute_conflict!"><code>compute_conflict!</code></a>:</p><pre><code class="language-julia hljs">julia&gt; compute_conflict!(model)
+ERROR: ArgumentError: The optimizer HiGHS.Optimizer does not support `compute_conflict!`</code></pre><p>In this case, HiGHS does not support computing conflicts, but other solvers such as Gurobi and CPLEX do. If the solver does support computing conflicts, read <a href="../../../manual/solutions/#Conflicts">Conflicts</a> for more details.</p></div></div><h3 id="Penalty-relaxation"><a class="docs-heading-anchor" href="#Penalty-relaxation">Penalty relaxation</a><a id="Penalty-relaxation-1"></a><a class="docs-heading-anchor-permalink" href="#Penalty-relaxation" title="Permalink"></a></h3><p>Another strategy to debug sources of infeasibility is the <a href="../../../api/JuMP/#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a> function.</p><p>The penalty relaxation modifies constraints of the form <span>$f(x) \in S$</span> into <span>$f(x) + y - z \in S$</span>, where <span>$y, z \ge 0$</span>, and then it introduces a penalty term into the objective of <span>$a \times (y + z)$</span> (if minimizing, else <span>$-a$</span>), where <span>$a$</span> is a penalty.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; map = relax_with_penalty!(model)</code><code class="nohighlight hljs ansi" style="display:block;"><span class="sgr33"><span class="sgr1">┌ Warning: </span></span>Skipping PenaltyRelaxation for ConstraintIndex{MathOptInterface.VariableIndex,MathOptInterface.GreaterThan{Float64}}
+<span class="sgr33"><span class="sgr1">└ </span></span><span class="sgr90">@ MathOptInterface.Utilities ~/.julia/dev/MathOptInterface/src/Utilities/penalty_relaxation.jl:289</span>
+Dict{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, AffExpr} with 1 entry:
+  con : 2 x - _[2] ≤ -1 =&gt; _[2]</code></pre><p>Here <code>map</code> is a dictionary which maps constraint indices to an affine expression representing <span>$(y + z)$</span>.</p><p>If we optimize the relaxed model, this time we get a feasible solution:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; termination_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">OPTIMAL::TerminationStatusCode = 1</code></pre><p>Iterate over the contents of <code>map</code> to see which constraints are violated:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; for (con, penalty) in map
+           violation = value(penalty)
+           if violation &gt; 0
+               println(&quot;Constraint `$(name(con))` is violated by $violation&quot;)
+           end
+       end</code><code class="nohighlight hljs ansi" style="display:block;">Constraint `con` is violated by 1.0</code></pre><p>Once you find a violated constraint in the relaxed problem, take a look to see if there is a typo or other common mistake in that particular constraint.</p><p>Consult the docstring <a href="../../../api/JuMP/#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a> for information on how to modify the penalty cost term <code>a</code>, either for every constraint in the model or a particular subset of the constraints.</p><p>When using <a href="../../../api/JuMP/#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a>, you should be aware that:</p><ul><li>Variable bounds and integrality restrictions are not relaxed. If the problem is still infeasible after calling <a href="../../../api/JuMP/#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a>, check the variable bounds.</li><li>You cannot undo the penalty relaxation. If you need an unmodified model, rebuild the problem, or call <a href="../../../api/JuMP/#JuMP.copy_model"><code>copy_model</code></a> before calling <a href="../../../api/JuMP/#JuMP.relax_with_penalty!"><code>relax_with_penalty!</code></a>.</li></ul><h2 id="Debugging-an-unbounded-model"><a class="docs-heading-anchor" href="#Debugging-an-unbounded-model">Debugging an unbounded model</a><a id="Debugging-an-unbounded-model-1"></a><a class="docs-heading-anchor-permalink" href="#Debugging-an-unbounded-model" title="Permalink"></a></h2><p>A model is unbounded if there is no limit on how good the objective value can get. Most often, an unbounded model means that you have an error in your modeling, because all physical systems have limits. (You cannot make an infinite amount of profit.)</p><h3 id="Example-2"><a class="docs-heading-anchor" href="#Example-2">Example</a><a class="docs-heading-anchor-permalink" href="#Example-2" title="Permalink"></a></h3><p>A simple example of an unbounded model is:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer);</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_silent(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Max, 2x + 1)</code><code class="nohighlight hljs ansi" style="display:block;">2 x + 1</code></pre><p>because we can increase <code>x</code> without limit, and the objective value <code>2x + 1</code> gets better as <code>x</code> increases.</p><p>When the problem is unbounded, JuMP may return one of a number of statuses. The most common is <code>DUAL_INFEASIBLE</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; termination_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">DUAL_INFEASIBLE::TerminationStatusCode = 3</code></pre><p>Depending on the solver, you may also receive <code>INFEASIBLE_OR_UNBOUNDED</code> or an error code like <code>NORM_LIMIT</code>.</p><h3 id="Common-sources-4"><a class="docs-heading-anchor" href="#Common-sources-4">Common sources</a><a class="docs-heading-anchor-permalink" href="#Common-sources-4" title="Permalink"></a></h3><p>Common sources of unboundedness are:</p><ul><li>Using <code>Max</code> instead of <code>Min</code></li><li>Omitting variable bounds, such as <code>0 &lt;= x &lt;= 1</code></li><li>Using <code>+</code> instead of <code>-</code> in a term of the objective function.</li></ul><h3 id="Strategies-4"><a class="docs-heading-anchor" href="#Strategies-4">Strategies</a><a class="docs-heading-anchor-permalink" href="#Strategies-4" title="Permalink"></a></h3><p>Strategies to debug sources of unboundedness include:</p><ul><li>Double check whether you intended <code>Min</code> or <code>Max</code> in the <a href="../../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> line.</li><li>Print the objective function with <code>print(objective_function(model))</code> and verify that the value and sign of each coefficient is as you expect.</li><li>Add large bounds to all variables that are free or have one-sided bounds, then re-solve the problem. Because all variables are now bounded, the problem will have a finite optimal solution. Look at the value of each variable in the optimal solution to see if it is at one of the new bounds. If it is, you either need to specify a better bound for that variable, or there might be a mistake in the objective function associated with that variable (for example, a <code>+</code> instead of a <code>-</code>).</li></ul><p>If there are too many variables to add bounds to, or there are too many terms to examine by hand, another strategy is to create a new variable with a large upper bound (if maximizing, lower bound if minimizing) and a constraint that the variable must be less-than or equal to the expression of the objective function. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer);</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_silent(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; # @objective(model, Max, 2x + 1)
+       @variable(model, objective &lt;= 10_000)</code><code class="nohighlight hljs ansi" style="display:block;">objective</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, objective &lt;= 2x + 1)</code><code class="nohighlight hljs ansi" style="display:block;">-2 x + objective ≤ 1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Max, objective)</code><code class="nohighlight hljs ansi" style="display:block;">objective</code></pre><p>This new model has a finite optimal solution, so we can solve it and then look for variables with large positive or negative values in the optimal solution.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; for var in all_variables(model)
+           if var == objective
+               continue
+           end
+           if abs(value(var)) &gt; 1e3
+               println(&quot;Variable `$(name(var))` may be unbounded&quot;)
+           end
+       end</code><code class="nohighlight hljs ansi" style="display:block;">Variable `x` may be unbounded</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../getting_started_with_data_and_plotting/">« Getting started with data and plotting</a><a class="docs-footer-nextpage" href="../design_patterns_for_larger_models/">Design patterns for larger models »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models.jl b/previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models.jl
new file mode 100644
index 00000000000..ea9530beccb
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models.jl
@@ -0,0 +1,688 @@
+# Copyright (c) 2021 Oscar Dowson and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Design patterns for larger models
+
+# JuMP makes it easy to build and solve optimization models. However, once you
+# start to construct larger models, and especially ones that interact with
+# external data sources or have customizable sets of variables and constraints
+# based on client choices, you may find that your scripts become unwieldy. This
+# tutorial demonstrates a variety of ways in which you can structure larger JuMP
+# models to improve their readability and maintainability.
+
+# !!! tip
+#     This tutorial is more advanced than the other "Getting started" tutorials.
+#     It's in the "Getting started" section to give you an early preview of how
+#     JuMP makes it easy to structure larger models. However, if you are new to
+#     JuMP you may want to briefly skim the tutorial, and come back to it once
+#     you have written a few JuMP models.
+
+# ## Overview
+
+# This tutorial uses explanation-by-example. We're going to start with a simple
+# [knapsack model](https://en.wikipedia.org/wiki/Knapsack_problem), and then
+# expand it to add various features and structure.
+
+# ## A simple script
+
+# Your first prototype of a JuMP model is probably a script that uses a small
+# set of hard-coded data.
+
+using JuMP, HiGHS
+profit = [5, 3, 2, 7, 4]
+weight = [2, 8, 4, 2, 5]
+capacity = 10
+N = 5
+model = Model(HiGHS.Optimizer)
+@variable(model, x[1:N], Bin)
+@objective(model, Max, sum(profit[i] * x[i] for i in 1:N))
+@constraint(model, sum(weight[i] * x[i] for i in 1:N) <= capacity)
+optimize!(model)
+value.(x)
+
+# The benefits of this approach are:
+#  * it is quick to code
+#  * it is quick to make changes.
+
+# The downsides include:
+#  * all variables are global (read [Performance tips](https://docs.julialang.org/en/v1/manual/performance-tips/))
+#  * it is easy to introduce errors, for example, having `profit` and `weight` be
+#    vectors of different lengths, or not match `N`
+#  * the solution, `x[i]`, is hard to interpret without knowing the order in
+#    which we provided the data.
+
+# ## Wrap the model in a function
+
+# A good next step is to wrap your model in a function. This is useful for a few
+# reasons:
+#  * it removes global variables
+#  * it encapsulates the JuMP model and forces you to clarify your inputs and
+#    outputs
+#  * we can add some error checking.
+
+function solve_knapsack_1(profit::Vector, weight::Vector, capacity::Real)
+    if length(profit) != length(weight)
+        throw(DimensionMismatch("profit and weight are different sizes"))
+    end
+    N = length(weight)
+    model = Model(HiGHS.Optimizer)
+    @variable(model, x[1:N], Bin)
+    @objective(model, Max, sum(profit[i] * x[i] for i in 1:N))
+    @constraint(model, sum(weight[i] * x[i] for i in 1:N) <= capacity)
+    optimize!(model)
+    return value.(x)
+end
+
+solve_knapsack_1([5, 3, 2, 7, 4], [2, 8, 4, 2, 5], 10)
+
+# ## Create better data structures
+
+# Although we can check for errors like mis-matched vector lengths, if you start
+# to develop models with a lot of data, keeping track of vectors and lengths and
+# indices is fragile and a common source of bugs. A good solution is to use
+# Julia's type system to create an abstraction over your data.
+
+# For example, we can create a `struct` that represents a single object, with a
+# constructor that lets us validate assumptions on the input data:
+
+struct KnapsackObject
+    profit::Float64
+    weight::Float64
+    function KnapsackObject(profit::Float64, weight::Float64)
+        if weight < 0
+            throw(DomainError("Weight of object cannot be negative"))
+        end
+        return new(profit, weight)
+    end
+end
+
+# as well as a `struct` that holds a dictionary of objects and the knapsack's
+# capacity:
+
+struct KnapsackData
+    objects::Dict{String,KnapsackObject}
+    capacity::Float64
+end
+
+# Here's what our data might look like now:
+
+objects = Dict(
+    "apple" => KnapsackObject(5.0, 2.0),
+    "banana" => KnapsackObject(3.0, 8.0),
+    "cherry" => KnapsackObject(2.0, 4.0),
+    "date" => KnapsackObject(7.0, 2.0),
+    "eggplant" => KnapsackObject(4.0, 5.0),
+)
+data = KnapsackData(objects, 10.0)
+
+# If you want, you can add custom printing to make it easier to visualize:
+
+function Base.show(io::IO, data::KnapsackData)
+    println(io, "A knapsack with capacity $(data.capacity) and possible items:")
+    for (k, v) in data.objects
+        println(
+            io,
+            "  $(rpad(k, 8)) : profit = $(v.profit), weight = $(v.weight)",
+        )
+    end
+    return
+end
+
+data
+
+# Then, we can re-write our `solve_knapsack` function to take our `KnapsackData`
+# as input:
+
+function solve_knapsack_2(data::KnapsackData)
+    model = Model(HiGHS.Optimizer)
+    @variable(model, x[keys(data.objects)], Bin)
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    @constraint(
+        model,
+        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,
+    )
+    optimize!(model)
+    return value.(x)
+end
+
+solve_knapsack_2(data)
+
+# ## Read in data from files
+
+# Having a data structure is a good step. But it is still annoying that we have
+# to hard-code the data into Julia. A good next step is to separate the data
+# into an external file format; JSON is a common choice.
+
+json_data = """
+{
+    "objects": {
+        "apple": {"profit": 5.0, "weight": 2.0},
+        "banana": {"profit": 3.0, "weight": 8.0},
+        "cherry": {"profit": 2.0, "weight": 4.0},
+        "date": {"profit": 7.0, "weight": 2.0},
+        "eggplant": {"profit": 4.0, "weight": 5.0}
+    },
+    "capacity": 10.0
+}
+"""
+temp_dir = mktempdir()
+knapsack_json_filename = joinpath(temp_dir, "knapsack.json")
+## Instead of writing a new file here you could replace `knapsack_json_filename`
+## with the path to a local file.
+write(knapsack_json_filename, json_data);
+
+# Now let's write a function that reads this file and builds a `KnapsackData`
+# object:
+
+import JSON
+
+function read_data(filename)
+    d = JSON.parsefile(filename)
+    return KnapsackData(
+        Dict(
+            k => KnapsackObject(v["profit"], v["weight"]) for
+            (k, v) in d["objects"]
+        ),
+        d["capacity"],
+    )
+end
+
+data = read_data(knapsack_json_filename)
+
+# ## Add options via if-else
+
+# At this point, we have data in a file format which we can load and solve a
+# single problem. For many users, this might be sufficient. However, at some
+# point you may be asked to add features like "but what if we want to take more
+# than one of a particular item?"
+
+# If this is the first time that you've been asked to add a feature, adding
+# options via `if-else` statements is a good approach. For example, we might
+# write:
+
+function solve_knapsack_3(data::KnapsackData; binary_knapsack::Bool)
+    model = Model(HiGHS.Optimizer)
+    if binary_knapsack
+        @variable(model, x[keys(data.objects)], Bin)
+    else
+        @variable(model, x[keys(data.objects)] >= 0, Int)
+    end
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    @constraint(
+        model,
+        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,
+    )
+    optimize!(model)
+    return value.(x)
+end
+
+# Now we can solve the binary knapsack:
+
+solve_knapsack_3(data; binary_knapsack = true)
+
+# And an integer knapsack where we can take more than one copy of each item:
+
+solve_knapsack_3(data; binary_knapsack = false)
+
+# ## Add configuration options via dispatch
+
+# If you get repeated requests to add different options, you'll quickly find
+# yourself in a mess of different flags and `if-else` statements. It's hard to
+# write, hard to read, and hard to ensure you haven't introduced any bugs.
+# A good solution is to use Julia's type dispatch to control the configuration
+# of the model. The easiest way to explain this is by example.
+
+# First, start by defining a new abstract type, as well as new subtypes for each
+# of our options. These types are going to control the configuration of the
+# knapsack model.
+
+abstract type AbstractConfiguration end
+
+struct BinaryKnapsackConfig <: AbstractConfiguration end
+
+struct IntegerKnapsackConfig <: AbstractConfiguration end
+
+# Then, we rewrite our `solve_knapsack` function to take a `config` argument,
+# and we introduce an `add_knapsack_variables` function to abstract the creation
+# of our variables.
+
+function solve_knapsack_4(data::KnapsackData, config::AbstractConfiguration)
+    model = Model(HiGHS.Optimizer)
+    x = add_knapsack_variables(model, data, config)
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    @constraint(
+        model,
+        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,
+    )
+    optimize!(model)
+    return value.(x)
+end
+
+# For the binary knapsack problem, `add_knapsack_variables` looks like this:
+
+function add_knapsack_variables(
+    model::Model,
+    data::KnapsackData,
+    ::BinaryKnapsackConfig,
+)
+    return @variable(model, x[keys(data.objects)], Bin)
+end
+
+# For the integer knapsack problem, `add_knapsack_variables` looks like this:
+
+function add_knapsack_variables(
+    model::Model,
+    data::KnapsackData,
+    ::IntegerKnapsackConfig,
+)
+    return @variable(model, x[keys(data.objects)] >= 0, Int)
+end
+
+# Now we can solve the binary knapsack:
+
+solve_knapsack_4(data, BinaryKnapsackConfig())
+
+# and the integer knapsack problem:
+
+solve_knapsack_4(data, IntegerKnapsackConfig())
+
+# The main benefit of the dispatch approach is that you can quickly add new
+# options without needing to modify the existing code. For example:
+
+struct UpperBoundedKnapsackConfig <: AbstractConfiguration
+    limit::Int
+end
+
+function add_knapsack_variables(
+    model::Model,
+    data::KnapsackData,
+    config::UpperBoundedKnapsackConfig,
+)
+    return @variable(model, 0 <= x[keys(data.objects)] <= config.limit, Int)
+end
+
+solve_knapsack_4(data, UpperBoundedKnapsackConfig(3))
+
+# ## Generalize constraints and objectives
+
+# It's easy to extend the dispatch approach to constraints and objectives as
+# well. The key points to notice in the next two functions are that:
+#  * we can access registered variables via `model[:x]`
+#  * we can define generic functions which accept any `AbstractConfiguration` as a
+#    configuration argument. That means we can implement a single method and
+#    have it apply to multiple configuration types.
+
+function add_knapsack_constraints(
+    model::Model,
+    data::KnapsackData,
+    ::AbstractConfiguration,
+)
+    x = model[:x]
+    @constraint(
+        model,
+        capacity_constraint,
+        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,
+    )
+    return
+end
+
+function add_knapsack_objective(
+    model::Model,
+    data::KnapsackData,
+    ::AbstractConfiguration,
+)
+    x = model[:x]
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    return
+end
+
+function solve_knapsack_5(data::KnapsackData, config::AbstractConfiguration)
+    model = Model(HiGHS.Optimizer)
+    add_knapsack_variables(model, data, config)
+    add_knapsack_constraints(model, data, config)
+    add_knapsack_objective(model, data, config)
+    optimize!(model)
+    return value.(model[:x])
+end
+
+solve_knapsack_5(data, BinaryKnapsackConfig())
+
+# ## Remove solver dependence, add error checks
+
+# Compared to where we started, our knapsack model is now significantly
+# different. We've wrapped it in a function, defined some data types, and
+# introduced configuration options to control the variables and constraints that
+# get added. There are a few other steps we can do to further improve things:
+#  * remove the dependence on `HiGHS`
+#  * add checks that we found an optimal solution
+#  * add a helper function to avoid the need to explicitly construct the data.
+
+function solve_knapsack_6(
+    optimizer,
+    data::KnapsackData,
+    config::AbstractConfiguration,
+)
+    model = Model(optimizer)
+    add_knapsack_variables(model, data, config)
+    add_knapsack_constraints(model, data, config)
+    add_knapsack_objective(model, data, config)
+    optimize!(model)
+    if termination_status(model) != OPTIMAL
+        @warn("Model not solved to optimality")
+        return nothing
+    end
+    return value.(model[:x])
+end
+
+function solve_knapsack_6(
+    optimizer,
+    data::String,
+    config::AbstractConfiguration,
+)
+    return solve_knapsack_6(optimizer, read_data(data), config)
+end
+
+solution = solve_knapsack_6(
+    HiGHS.Optimizer,
+    knapsack_json_filename,
+    BinaryKnapsackConfig(),
+)
+
+# ## Create a module
+
+# Now we're ready to expose our model to the wider world. That might be as part
+# of a larger Julia project that we're contributing to, or as a stand-alone
+# script that we can run on-demand. In either case, it's good practice to wrap
+# everything in a module. This further encapsulates our code into a single
+# namespace, and we can add documentation in the form of
+# [docstrings](https://docs.julialang.org/en/v1/manual/documentation/).
+
+# Some good rules to follow when creating a module are:
+# * use `import` in a module instead of `using` to make it clear which functions
+#   are from which packages
+# * use `_` to start function and type names that are considered private
+# * add docstrings to all public variables and functions.
+
+module KnapsackModel
+
+import JuMP
+import JSON
+
+struct _KnapsackObject
+    profit::Float64
+    weight::Float64
+    function _KnapsackObject(profit::Float64, weight::Float64)
+        if weight < 0
+            throw(DomainError("Weight of object cannot be negative"))
+        end
+        return new(profit, weight)
+    end
+end
+
+struct _KnapsackData
+    objects::Dict{String,_KnapsackObject}
+    capacity::Float64
+end
+
+function _read_data(filename)
+    d = JSON.parsefile(filename)
+    return _KnapsackData(
+        Dict(
+            k => _KnapsackObject(v["profit"], v["weight"]) for
+            (k, v) in d["objects"]
+        ),
+        d["capacity"],
+    )
+end
+
+abstract type _AbstractConfiguration end
+
+"""
+    BinaryKnapsackConfig()
+
+Create a binary knapsack problem where each object can be taken 0 or 1 times.
+"""
+struct BinaryKnapsackConfig <: _AbstractConfiguration end
+
+"""
+    IntegerKnapsackConfig()
+
+Create an integer knapsack problem where each object can be taken any number of
+times.
+"""
+struct IntegerKnapsackConfig <: _AbstractConfiguration end
+
+function _add_knapsack_variables(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::BinaryKnapsackConfig,
+)
+    return JuMP.@variable(model, x[keys(data.objects)], Bin)
+end
+
+function _add_knapsack_variables(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::IntegerKnapsackConfig,
+)
+    return JuMP.@variable(model, x[keys(data.objects)] >= 0, Int)
+end
+
+function _add_knapsack_constraints(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::_AbstractConfiguration,
+)
+    x = model[:x]
+    JuMP.@constraint(
+        model,
+        capacity_constraint,
+        sum(v.weight * x[k] for (k, v) in data.objects) <= data.capacity,
+    )
+    return
+end
+
+function _add_knapsack_objective(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::_AbstractConfiguration,
+)
+    x = model[:x]
+    JuMP.@objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    return
+end
+
+function _solve_knapsack(
+    optimizer,
+    data::_KnapsackData,
+    config::_AbstractConfiguration,
+)
+    model = JuMP.Model(optimizer)
+    _add_knapsack_variables(model, data, config)
+    _add_knapsack_constraints(model, data, config)
+    _add_knapsack_objective(model, data, config)
+    JuMP.optimize!(model)
+    if JuMP.termination_status(model) != JuMP.OPTIMAL
+        @warn("Model not solved to optimality")
+        return nothing
+    end
+    return JuMP.value.(model[:x])
+end
+
+"""
+    solve_knapsack(
+        optimizer,
+        knapsack_json_filename::String,
+        config::_AbstractConfiguration,
+    )
+
+Solve the knapsack problem and return the optimal primal solution
+
+## Arguments
+
+ * `optimizer` : an object that can be passed to `JuMP.Model` to construct a new
+   JuMP model.
+ * `knapsack_json_filename` : the filename of a JSON file containing the data for the
+   problem.
+ * `config` : an object to control the type of knapsack model constructed.
+   Valid options are:
+    * `BinaryKnapsackConfig()`
+    * `IntegerKnapsackConfig()`
+
+## Returns
+
+ * If an optimal solution exists: a `JuMP.DenseAxisArray` that maps the `String`
+   name of each object to the number of objects to pack into the knapsack.
+ * Otherwise, `nothing`, indicating that the problem does not have an optimal
+   solution.
+
+## Examples
+
+```julia
+solution = solve_knapsack(
+    HiGHS.Optimizer,
+    "path/to/data.json",
+    BinaryKnapsackConfig(),
+)
+```
+
+```julia
+solution = solve_knapsack(
+    MOI.OptimizerWithAttributes(HiGHS.Optimizer, "output_flag" => false),
+    "path/to/data.json",
+    IntegerKnapsackConfig(),
+)
+```
+"""
+function solve_knapsack(
+    optimizer,
+    knapsack_json_filename::String,
+    config::_AbstractConfiguration,
+)
+    data = _read_data(knapsack_json_filename)
+    return _solve_knapsack(optimizer, data, config)
+end
+
+end
+
+# Finally, you can call your model:
+
+import .KnapsackModel
+
+KnapsackModel.solve_knapsack(
+    HiGHS.Optimizer,
+    knapsack_json_filename,
+    KnapsackModel.BinaryKnapsackConfig(),
+)
+
+# !!! note
+#     The `.` in `.KnapsackModel` denotes that it is a submodule and not a
+#     separate package that we installed with `Pkg.add`. If you put the
+#     `KnapsackModel` in a separate file, load it with:
+#     ```julia
+#     include("path/to/KnapsackModel.jl")
+#     import .KnapsackModel
+#     ```
+
+# ## Add tests
+
+# As a final step, you should add tests for your model. This often means testing
+# on a small problem for which you can work out the optimal solution by hand.
+# The Julia standard library `Test` has good unit-testing functionality.
+
+import .KnapsackModel
+using Test
+
+@testset "KnapsackModel" begin
+    @testset "feasible_binary_knapsack" begin
+        x = KnapsackModel.solve_knapsack(
+            HiGHS.Optimizer,
+            knapsack_json_filename,
+            KnapsackModel.BinaryKnapsackConfig(),
+        )
+        @test isapprox(x["apple"], 1, atol = 1e-5)
+        @test isapprox(x["banana"], 0, atol = 1e-5)
+        @test isapprox(x["cherry"], 0, atol = 1e-5)
+        @test isapprox(x["date"], 1, atol = 1e-5)
+        @test isapprox(x["eggplant"], 1, atol = 1e-5)
+    end
+    @testset "feasible_integer_knapsack" begin
+        x = KnapsackModel.solve_knapsack(
+            HiGHS.Optimizer,
+            knapsack_json_filename,
+            KnapsackModel.IntegerKnapsackConfig(),
+        )
+        @test isapprox(x["apple"], 0, atol = 1e-5)
+        @test isapprox(x["banana"], 0, atol = 1e-5)
+        @test isapprox(x["cherry"], 0, atol = 1e-5)
+        @test isapprox(x["date"], 5, atol = 1e-5)
+        @test isapprox(x["eggplant"], 0, atol = 1e-5)
+    end
+    @testset "infeasible_binary_knapsack" begin
+        dir = mktempdir()
+        infeasible_filename = joinpath(dir, "infeasible.json")
+        write(
+            infeasible_filename,
+            """{
+                "objects": {
+                    "apple": {"profit": 5.0, "weight": 2.0},
+                    "banana": {"profit": 3.0, "weight": 8.0},
+                    "cherry": {"profit": 2.0, "weight": 4.0},
+                    "date": {"profit": 7.0, "weight": 2.0},
+                    "eggplant": {"profit": 4.0, "weight": 5.0}
+                },
+                "capacity": -10.0
+            }""",
+        )
+        x = KnapsackModel.solve_knapsack(
+            HiGHS.Optimizer,
+            infeasible_filename,
+            KnapsackModel.BinaryKnapsackConfig(),
+        )
+        @test x === nothing
+    end
+end
+
+# !!! tip
+#     Place these tests in a separate file `test_knapsack_model.jl` so that you
+#     can run the tests by adding `include("test_knapsack_model.jl")` to any
+#     file where needed.
+
+# ## Next steps
+
+# We've only briefly scratched the surface of ways to create and structure large
+# JuMP models, so consider this tutorial a starting point, rather than a
+# comprehensive list of all the possible ways to structure JuMP models.  If you
+# are embarking on a large project that uses JuMP, a good next step is to
+# look at ways people have written large JuMP projects "in the wild."
+
+# Here are some good examples (all co-incidentally related to energy):
+# * AnyMOD.jl
+#   * [JuMP-dev 2021 talk](https://www.youtube.com/watch?v=QE_tNDER0F4)
+#   * [source code](https://github.com/leonardgoeke/AnyMOD.jl)
+# * PowerModels.jl
+#   * [JuMP-dev 2021 talk](https://www.youtube.com/watch?v=POOt1FCA8LI)
+#   * [source code](https://github.com/lanl-ansi/PowerModels.jl)
+# * PowerSimulations.jl
+#    * [JuliaCon 2021 talk](https://www.youtube.com/watch?v=-ZoO3npjwYU)
+#    * [source code](https://github.com/NREL-SIIP/PowerSimulations.jl)
+# * UnitCommitment.jl
+#   * [JuMP-dev 2021 talk](https://www.youtube.com/watch?v=rYUZK9kYeIY)
+#   * [source code](https://github.com/ANL-CEEESA/UnitCommitment.jl)
diff --git a/previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models/index.html b/previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models/index.html
new file mode 100644
index 00000000000..4b375a968cc
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/design_patterns_for_larger_models/index.html
@@ -0,0 +1,517 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Design patterns for larger models · JuMP</title><meta name="title" content="Design patterns for larger models · JuMP"/><meta property="og:title" content="Design patterns for larger models · JuMP"/><meta property="twitter:title" content="Design patterns for larger models · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../debugging/">Debugging</a></li><li class="is-active"><a class="tocitem" href>Design patterns for larger models</a><ul class="internal"><li><a class="tocitem" href="#Overview"><span>Overview</span></a></li><li><a class="tocitem" href="#A-simple-script"><span>A simple script</span></a></li><li><a class="tocitem" href="#Wrap-the-model-in-a-function"><span>Wrap the model in a function</span></a></li><li><a class="tocitem" href="#Create-better-data-structures"><span>Create better data structures</span></a></li><li><a class="tocitem" href="#Read-in-data-from-files"><span>Read in data from files</span></a></li><li><a class="tocitem" href="#Add-options-via-if-else"><span>Add options via if-else</span></a></li><li><a class="tocitem" href="#Add-configuration-options-via-dispatch"><span>Add configuration options via dispatch</span></a></li><li><a class="tocitem" href="#Generalize-constraints-and-objectives"><span>Generalize constraints and objectives</span></a></li><li><a class="tocitem" href="#Remove-solver-dependence,-add-error-checks"><span>Remove solver dependence, add error checks</span></a></li><li><a class="tocitem" href="#Create-a-module"><span>Create a module</span></a></li><li><a class="tocitem" href="#Add-tests"><span>Add tests</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Design patterns for larger models</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Design patterns for larger models</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/design_patterns_for_larger_models.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Design-patterns-for-larger-models"><a class="docs-heading-anchor" href="#Design-patterns-for-larger-models">Design patterns for larger models</a><a id="Design-patterns-for-larger-models-1"></a><a class="docs-heading-anchor-permalink" href="#Design-patterns-for-larger-models" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../design_patterns_for_larger_models.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>JuMP makes it easy to build and solve optimization models. However, once you start to construct larger models, and especially ones that interact with external data sources or have customizable sets of variables and constraints based on client choices, you may find that your scripts become unwieldy. This tutorial demonstrates a variety of ways in which you can structure larger JuMP models to improve their readability and maintainability.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This tutorial is more advanced than the other &quot;Getting started&quot; tutorials. It&#39;s in the &quot;Getting started&quot; section to give you an early preview of how JuMP makes it easy to structure larger models. However, if you are new to JuMP you may want to briefly skim the tutorial, and come back to it once you have written a few JuMP models.</p></div></div><h2 id="Overview"><a class="docs-heading-anchor" href="#Overview">Overview</a><a id="Overview-1"></a><a class="docs-heading-anchor-permalink" href="#Overview" title="Permalink"></a></h2><p>This tutorial uses explanation-by-example. We&#39;re going to start with a simple <a href="https://en.wikipedia.org/wiki/Knapsack_problem">knapsack model</a>, and then expand it to add various features and structure.</p><h2 id="A-simple-script"><a class="docs-heading-anchor" href="#A-simple-script">A simple script</a><a id="A-simple-script-1"></a><a class="docs-heading-anchor-permalink" href="#A-simple-script" title="Permalink"></a></h2><p>Your first prototype of a JuMP model is probably a script that uses a small set of hard-coded data.</p><pre><code class="language-julia hljs">using JuMP, HiGHS
+profit = [5, 3, 2, 7, 4]
+weight = [2, 8, 4, 2, 5]
+capacity = 10
+N = 5
+model = Model(HiGHS.Optimizer)
+@variable(model, x[1:N], Bin)
+@objective(model, Max, sum(profit[i] * x[i] for i in 1:N))
+@constraint(model, sum(weight[i] * x[i] for i in 1:N) &lt;= capacity)
+optimize!(model)
+value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5-element Vector{Float64}:
+  1.0
+  0.0
+ -0.0
+  1.0
+  1.0</code></pre><p>The benefits of this approach are:</p><ul><li>it is quick to code</li><li>it is quick to make changes.</li></ul><p>The downsides include:</p><ul><li>all variables are global (read <a href="https://docs.julialang.org/en/v1/manual/performance-tips/">Performance tips</a>)</li><li>it is easy to introduce errors, for example, having <code>profit</code> and <code>weight</code> be vectors of different lengths, or not match <code>N</code></li><li>the solution, <code>x[i]</code>, is hard to interpret without knowing the order in which we provided the data.</li></ul><h2 id="Wrap-the-model-in-a-function"><a class="docs-heading-anchor" href="#Wrap-the-model-in-a-function">Wrap the model in a function</a><a id="Wrap-the-model-in-a-function-1"></a><a class="docs-heading-anchor-permalink" href="#Wrap-the-model-in-a-function" title="Permalink"></a></h2><p>A good next step is to wrap your model in a function. This is useful for a few reasons:</p><ul><li>it removes global variables</li><li>it encapsulates the JuMP model and forces you to clarify your inputs and outputs</li><li>we can add some error checking.</li></ul><pre><code class="language-julia hljs">function solve_knapsack_1(profit::Vector, weight::Vector, capacity::Real)
+    if length(profit) != length(weight)
+        throw(DimensionMismatch(&quot;profit and weight are different sizes&quot;))
+    end
+    N = length(weight)
+    model = Model(HiGHS.Optimizer)
+    @variable(model, x[1:N], Bin)
+    @objective(model, Max, sum(profit[i] * x[i] for i in 1:N))
+    @constraint(model, sum(weight[i] * x[i] for i in 1:N) &lt;= capacity)
+    optimize!(model)
+    return value.(x)
+end
+
+solve_knapsack_1([5, 3, 2, 7, 4], [2, 8, 4, 2, 5], 10)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5-element Vector{Float64}:
+  1.0
+  0.0
+ -0.0
+  1.0
+  1.0</code></pre><h2 id="Create-better-data-structures"><a class="docs-heading-anchor" href="#Create-better-data-structures">Create better data structures</a><a id="Create-better-data-structures-1"></a><a class="docs-heading-anchor-permalink" href="#Create-better-data-structures" title="Permalink"></a></h2><p>Although we can check for errors like mis-matched vector lengths, if you start to develop models with a lot of data, keeping track of vectors and lengths and indices is fragile and a common source of bugs. A good solution is to use Julia&#39;s type system to create an abstraction over your data.</p><p>For example, we can create a <code>struct</code> that represents a single object, with a constructor that lets us validate assumptions on the input data:</p><pre><code class="language-julia hljs">struct KnapsackObject
+    profit::Float64
+    weight::Float64
+    function KnapsackObject(profit::Float64, weight::Float64)
+        if weight &lt; 0
+            throw(DomainError(&quot;Weight of object cannot be negative&quot;))
+        end
+        return new(profit, weight)
+    end
+end</code></pre><p>as well as a <code>struct</code> that holds a dictionary of objects and the knapsack&#39;s capacity:</p><pre><code class="language-julia hljs">struct KnapsackData
+    objects::Dict{String,KnapsackObject}
+    capacity::Float64
+end</code></pre><p>Here&#39;s what our data might look like now:</p><pre><code class="language-julia hljs">objects = Dict(
+    &quot;apple&quot; =&gt; KnapsackObject(5.0, 2.0),
+    &quot;banana&quot; =&gt; KnapsackObject(3.0, 8.0),
+    &quot;cherry&quot; =&gt; KnapsackObject(2.0, 4.0),
+    &quot;date&quot; =&gt; KnapsackObject(7.0, 2.0),
+    &quot;eggplant&quot; =&gt; KnapsackObject(4.0, 5.0),
+)
+data = KnapsackData(objects, 10.0)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Main.KnapsackData(Dict{String, Main.KnapsackObject}(&quot;cherry&quot; =&gt; Main.KnapsackObject(2.0, 4.0), &quot;banana&quot; =&gt; Main.KnapsackObject(3.0, 8.0), &quot;date&quot; =&gt; Main.KnapsackObject(7.0, 2.0), &quot;eggplant&quot; =&gt; Main.KnapsackObject(4.0, 5.0), &quot;apple&quot; =&gt; Main.KnapsackObject(5.0, 2.0)), 10.0)</code></pre><p>If you want, you can add custom printing to make it easier to visualize:</p><pre><code class="language-julia hljs">function Base.show(io::IO, data::KnapsackData)
+    println(io, &quot;A knapsack with capacity $(data.capacity) and possible items:&quot;)
+    for (k, v) in data.objects
+        println(
+            io,
+            &quot;  $(rpad(k, 8)) : profit = $(v.profit), weight = $(v.weight)&quot;,
+        )
+    end
+    return
+end
+
+data</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A knapsack with capacity 10.0 and possible items:
+  cherry   : profit = 2.0, weight = 4.0
+  banana   : profit = 3.0, weight = 8.0
+  date     : profit = 7.0, weight = 2.0
+  eggplant : profit = 4.0, weight = 5.0
+  apple    : profit = 5.0, weight = 2.0
+</code></pre><p>Then, we can re-write our <code>solve_knapsack</code> function to take our <code>KnapsackData</code> as input:</p><pre><code class="language-julia hljs">function solve_knapsack_2(data::KnapsackData)
+    model = Model(HiGHS.Optimizer)
+    @variable(model, x[keys(data.objects)], Bin)
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    @constraint(
+        model,
+        sum(v.weight * x[k] for (k, v) in data.objects) &lt;= data.capacity,
+    )
+    optimize!(model)
+    return value.(x)
+end
+
+solve_knapsack_2(data)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ -0.0
+  0.0
+  1.0
+  1.0
+  1.0</code></pre><h2 id="Read-in-data-from-files"><a class="docs-heading-anchor" href="#Read-in-data-from-files">Read in data from files</a><a id="Read-in-data-from-files-1"></a><a class="docs-heading-anchor-permalink" href="#Read-in-data-from-files" title="Permalink"></a></h2><p>Having a data structure is a good step. But it is still annoying that we have to hard-code the data into Julia. A good next step is to separate the data into an external file format; JSON is a common choice.</p><pre><code class="language-julia hljs">json_data = &quot;&quot;&quot;
+{
+    &quot;objects&quot;: {
+        &quot;apple&quot;: {&quot;profit&quot;: 5.0, &quot;weight&quot;: 2.0},
+        &quot;banana&quot;: {&quot;profit&quot;: 3.0, &quot;weight&quot;: 8.0},
+        &quot;cherry&quot;: {&quot;profit&quot;: 2.0, &quot;weight&quot;: 4.0},
+        &quot;date&quot;: {&quot;profit&quot;: 7.0, &quot;weight&quot;: 2.0},
+        &quot;eggplant&quot;: {&quot;profit&quot;: 4.0, &quot;weight&quot;: 5.0}
+    },
+    &quot;capacity&quot;: 10.0
+}
+&quot;&quot;&quot;
+temp_dir = mktempdir()
+knapsack_json_filename = joinpath(temp_dir, &quot;knapsack.json&quot;)
+# Instead of writing a new file here you could replace `knapsack_json_filename`
+# with the path to a local file.
+write(knapsack_json_filename, json_data);</code></pre><p>Now let&#39;s write a function that reads this file and builds a <code>KnapsackData</code> object:</p><pre><code class="language-julia hljs">import JSON
+
+function read_data(filename)
+    d = JSON.parsefile(filename)
+    return KnapsackData(
+        Dict(
+            k =&gt; KnapsackObject(v[&quot;profit&quot;], v[&quot;weight&quot;]) for
+            (k, v) in d[&quot;objects&quot;]
+        ),
+        d[&quot;capacity&quot;],
+    )
+end
+
+data = read_data(knapsack_json_filename)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A knapsack with capacity 10.0 and possible items:
+  cherry   : profit = 2.0, weight = 4.0
+  banana   : profit = 3.0, weight = 8.0
+  date     : profit = 7.0, weight = 2.0
+  eggplant : profit = 4.0, weight = 5.0
+  apple    : profit = 5.0, weight = 2.0
+</code></pre><h2 id="Add-options-via-if-else"><a class="docs-heading-anchor" href="#Add-options-via-if-else">Add options via if-else</a><a id="Add-options-via-if-else-1"></a><a class="docs-heading-anchor-permalink" href="#Add-options-via-if-else" title="Permalink"></a></h2><p>At this point, we have data in a file format which we can load and solve a single problem. For many users, this might be sufficient. However, at some point you may be asked to add features like &quot;but what if we want to take more than one of a particular item?&quot;</p><p>If this is the first time that you&#39;ve been asked to add a feature, adding options via <code>if-else</code> statements is a good approach. For example, we might write:</p><pre><code class="language-julia hljs">function solve_knapsack_3(data::KnapsackData; binary_knapsack::Bool)
+    model = Model(HiGHS.Optimizer)
+    if binary_knapsack
+        @variable(model, x[keys(data.objects)], Bin)
+    else
+        @variable(model, x[keys(data.objects)] &gt;= 0, Int)
+    end
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    @constraint(
+        model,
+        sum(v.weight * x[k] for (k, v) in data.objects) &lt;= data.capacity,
+    )
+    optimize!(model)
+    return value.(x)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_knapsack_3 (generic function with 1 method)</code></pre><p>Now we can solve the binary knapsack:</p><pre><code class="language-julia hljs">solve_knapsack_3(data; binary_knapsack = true)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ -0.0
+  0.0
+  1.0
+  1.0
+  1.0</code></pre><p>And an integer knapsack where we can take more than one copy of each item:</p><pre><code class="language-julia hljs">solve_knapsack_3(data; binary_knapsack = false)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ 0.0
+ 0.0
+ 5.0
+ 0.0
+ 0.0</code></pre><h2 id="Add-configuration-options-via-dispatch"><a class="docs-heading-anchor" href="#Add-configuration-options-via-dispatch">Add configuration options via dispatch</a><a id="Add-configuration-options-via-dispatch-1"></a><a class="docs-heading-anchor-permalink" href="#Add-configuration-options-via-dispatch" title="Permalink"></a></h2><p>If you get repeated requests to add different options, you&#39;ll quickly find yourself in a mess of different flags and <code>if-else</code> statements. It&#39;s hard to write, hard to read, and hard to ensure you haven&#39;t introduced any bugs. A good solution is to use Julia&#39;s type dispatch to control the configuration of the model. The easiest way to explain this is by example.</p><p>First, start by defining a new abstract type, as well as new subtypes for each of our options. These types are going to control the configuration of the knapsack model.</p><pre><code class="language-julia hljs">abstract type AbstractConfiguration end
+
+struct BinaryKnapsackConfig &lt;: AbstractConfiguration end
+
+struct IntegerKnapsackConfig &lt;: AbstractConfiguration end</code></pre><p>Then, we rewrite our <code>solve_knapsack</code> function to take a <code>config</code> argument, and we introduce an <code>add_knapsack_variables</code> function to abstract the creation of our variables.</p><pre><code class="language-julia hljs">function solve_knapsack_4(data::KnapsackData, config::AbstractConfiguration)
+    model = Model(HiGHS.Optimizer)
+    x = add_knapsack_variables(model, data, config)
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    @constraint(
+        model,
+        sum(v.weight * x[k] for (k, v) in data.objects) &lt;= data.capacity,
+    )
+    optimize!(model)
+    return value.(x)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_knapsack_4 (generic function with 1 method)</code></pre><p>For the binary knapsack problem, <code>add_knapsack_variables</code> looks like this:</p><pre><code class="language-julia hljs">function add_knapsack_variables(
+    model::Model,
+    data::KnapsackData,
+    ::BinaryKnapsackConfig,
+)
+    return @variable(model, x[keys(data.objects)], Bin)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">add_knapsack_variables (generic function with 1 method)</code></pre><p>For the integer knapsack problem, <code>add_knapsack_variables</code> looks like this:</p><pre><code class="language-julia hljs">function add_knapsack_variables(
+    model::Model,
+    data::KnapsackData,
+    ::IntegerKnapsackConfig,
+)
+    return @variable(model, x[keys(data.objects)] &gt;= 0, Int)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">add_knapsack_variables (generic function with 2 methods)</code></pre><p>Now we can solve the binary knapsack:</p><pre><code class="language-julia hljs">solve_knapsack_4(data, BinaryKnapsackConfig())</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ -0.0
+  0.0
+  1.0
+  1.0
+  1.0</code></pre><p>and the integer knapsack problem:</p><pre><code class="language-julia hljs">solve_knapsack_4(data, IntegerKnapsackConfig())</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ 0.0
+ 0.0
+ 5.0
+ 0.0
+ 0.0</code></pre><p>The main benefit of the dispatch approach is that you can quickly add new options without needing to modify the existing code. For example:</p><pre><code class="language-julia hljs">struct UpperBoundedKnapsackConfig &lt;: AbstractConfiguration
+    limit::Int
+end
+
+function add_knapsack_variables(
+    model::Model,
+    data::KnapsackData,
+    config::UpperBoundedKnapsackConfig,
+)
+    return @variable(model, 0 &lt;= x[keys(data.objects)] &lt;= config.limit, Int)
+end
+
+solve_knapsack_4(data, UpperBoundedKnapsackConfig(3))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ 0.0
+ 0.0
+ 3.0
+ 0.0
+ 2.0</code></pre><h2 id="Generalize-constraints-and-objectives"><a class="docs-heading-anchor" href="#Generalize-constraints-and-objectives">Generalize constraints and objectives</a><a id="Generalize-constraints-and-objectives-1"></a><a class="docs-heading-anchor-permalink" href="#Generalize-constraints-and-objectives" title="Permalink"></a></h2><p>It&#39;s easy to extend the dispatch approach to constraints and objectives as well. The key points to notice in the next two functions are that:</p><ul><li>we can access registered variables via <code>model[:x]</code></li><li>we can define generic functions which accept any <code>AbstractConfiguration</code> as a configuration argument. That means we can implement a single method and have it apply to multiple configuration types.</li></ul><pre><code class="language-julia hljs">function add_knapsack_constraints(
+    model::Model,
+    data::KnapsackData,
+    ::AbstractConfiguration,
+)
+    x = model[:x]
+    @constraint(
+        model,
+        capacity_constraint,
+        sum(v.weight * x[k] for (k, v) in data.objects) &lt;= data.capacity,
+    )
+    return
+end
+
+function add_knapsack_objective(
+    model::Model,
+    data::KnapsackData,
+    ::AbstractConfiguration,
+)
+    x = model[:x]
+    @objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    return
+end
+
+function solve_knapsack_5(data::KnapsackData, config::AbstractConfiguration)
+    model = Model(HiGHS.Optimizer)
+    add_knapsack_variables(model, data, config)
+    add_knapsack_constraints(model, data, config)
+    add_knapsack_objective(model, data, config)
+    optimize!(model)
+    return value.(model[:x])
+end
+
+solve_knapsack_5(data, BinaryKnapsackConfig())</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ -0.0
+  0.0
+  1.0
+  1.0
+  1.0</code></pre><h2 id="Remove-solver-dependence,-add-error-checks"><a class="docs-heading-anchor" href="#Remove-solver-dependence,-add-error-checks">Remove solver dependence, add error checks</a><a id="Remove-solver-dependence,-add-error-checks-1"></a><a class="docs-heading-anchor-permalink" href="#Remove-solver-dependence,-add-error-checks" title="Permalink"></a></h2><p>Compared to where we started, our knapsack model is now significantly different. We&#39;ve wrapped it in a function, defined some data types, and introduced configuration options to control the variables and constraints that get added. There are a few other steps we can do to further improve things:</p><ul><li>remove the dependence on <code>HiGHS</code></li><li>add checks that we found an optimal solution</li><li>add a helper function to avoid the need to explicitly construct the data.</li></ul><pre><code class="language-julia hljs">function solve_knapsack_6(
+    optimizer,
+    data::KnapsackData,
+    config::AbstractConfiguration,
+)
+    model = Model(optimizer)
+    add_knapsack_variables(model, data, config)
+    add_knapsack_constraints(model, data, config)
+    add_knapsack_objective(model, data, config)
+    optimize!(model)
+    if termination_status(model) != OPTIMAL
+        @warn(&quot;Model not solved to optimality&quot;)
+        return nothing
+    end
+    return value.(model[:x])
+end
+
+function solve_knapsack_6(
+    optimizer,
+    data::String,
+    config::AbstractConfiguration,
+)
+    return solve_knapsack_6(optimizer, read_data(data), config)
+end
+
+solution = solve_knapsack_6(
+    HiGHS.Optimizer,
+    knapsack_json_filename,
+    BinaryKnapsackConfig(),
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ -0.0
+  0.0
+  1.0
+  1.0
+  1.0</code></pre><h2 id="Create-a-module"><a class="docs-heading-anchor" href="#Create-a-module">Create a module</a><a id="Create-a-module-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-module" title="Permalink"></a></h2><p>Now we&#39;re ready to expose our model to the wider world. That might be as part of a larger Julia project that we&#39;re contributing to, or as a stand-alone script that we can run on-demand. In either case, it&#39;s good practice to wrap everything in a module. This further encapsulates our code into a single namespace, and we can add documentation in the form of <a href="https://docs.julialang.org/en/v1/manual/documentation/">docstrings</a>.</p><p>Some good rules to follow when creating a module are:</p><ul><li>use <code>import</code> in a module instead of <code>using</code> to make it clear which functions are from which packages</li><li>use <code>_</code> to start function and type names that are considered private</li><li>add docstrings to all public variables and functions.</li></ul><pre><code class="language-julia hljs">module KnapsackModel
+
+import JuMP
+import JSON
+
+struct _KnapsackObject
+    profit::Float64
+    weight::Float64
+    function _KnapsackObject(profit::Float64, weight::Float64)
+        if weight &lt; 0
+            throw(DomainError(&quot;Weight of object cannot be negative&quot;))
+        end
+        return new(profit, weight)
+    end
+end
+
+struct _KnapsackData
+    objects::Dict{String,_KnapsackObject}
+    capacity::Float64
+end
+
+function _read_data(filename)
+    d = JSON.parsefile(filename)
+    return _KnapsackData(
+        Dict(
+            k =&gt; _KnapsackObject(v[&quot;profit&quot;], v[&quot;weight&quot;]) for
+            (k, v) in d[&quot;objects&quot;]
+        ),
+        d[&quot;capacity&quot;],
+    )
+end
+
+abstract type _AbstractConfiguration end
+
+&quot;&quot;&quot;
+    BinaryKnapsackConfig()
+
+Create a binary knapsack problem where each object can be taken 0 or 1 times.
+&quot;&quot;&quot;
+struct BinaryKnapsackConfig &lt;: _AbstractConfiguration end
+
+&quot;&quot;&quot;
+    IntegerKnapsackConfig()
+
+Create an integer knapsack problem where each object can be taken any number of
+times.
+&quot;&quot;&quot;
+struct IntegerKnapsackConfig &lt;: _AbstractConfiguration end
+
+function _add_knapsack_variables(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::BinaryKnapsackConfig,
+)
+    return JuMP.@variable(model, x[keys(data.objects)], Bin)
+end
+
+function _add_knapsack_variables(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::IntegerKnapsackConfig,
+)
+    return JuMP.@variable(model, x[keys(data.objects)] &gt;= 0, Int)
+end
+
+function _add_knapsack_constraints(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::_AbstractConfiguration,
+)
+    x = model[:x]
+    JuMP.@constraint(
+        model,
+        capacity_constraint,
+        sum(v.weight * x[k] for (k, v) in data.objects) &lt;= data.capacity,
+    )
+    return
+end
+
+function _add_knapsack_objective(
+    model::JuMP.Model,
+    data::_KnapsackData,
+    ::_AbstractConfiguration,
+)
+    x = model[:x]
+    JuMP.@objective(model, Max, sum(v.profit * x[k] for (k, v) in data.objects))
+    return
+end
+
+function _solve_knapsack(
+    optimizer,
+    data::_KnapsackData,
+    config::_AbstractConfiguration,
+)
+    model = JuMP.Model(optimizer)
+    _add_knapsack_variables(model, data, config)
+    _add_knapsack_constraints(model, data, config)
+    _add_knapsack_objective(model, data, config)
+    JuMP.optimize!(model)
+    if JuMP.termination_status(model) != JuMP.OPTIMAL
+        @warn(&quot;Model not solved to optimality&quot;)
+        return nothing
+    end
+    return JuMP.value.(model[:x])
+end
+
+&quot;&quot;&quot;
+    solve_knapsack(
+        optimizer,
+        knapsack_json_filename::String,
+        config::_AbstractConfiguration,
+    )
+
+Solve the knapsack problem and return the optimal primal solution
+
+# Arguments
+
+ * `optimizer` : an object that can be passed to `JuMP.Model` to construct a new
+   JuMP model.
+ * `knapsack_json_filename` : the filename of a JSON file containing the data for the
+   problem.
+ * `config` : an object to control the type of knapsack model constructed.
+   Valid options are:
+    * `BinaryKnapsackConfig()`
+    * `IntegerKnapsackConfig()`
+
+# Returns
+
+ * If an optimal solution exists: a `JuMP.DenseAxisArray` that maps the `String`
+   name of each object to the number of objects to pack into the knapsack.
+ * Otherwise, `nothing`, indicating that the problem does not have an optimal
+   solution.
+
+# Examples
+
+```julia
+solution = solve_knapsack(
+    HiGHS.Optimizer,
+    &quot;path/to/data.json&quot;,
+    BinaryKnapsackConfig(),
+)
+```
+
+```julia
+solution = solve_knapsack(
+    MOI.OptimizerWithAttributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; false),
+    &quot;path/to/data.json&quot;,
+    IntegerKnapsackConfig(),
+)
+```
+&quot;&quot;&quot;
+function solve_knapsack(
+    optimizer,
+    knapsack_json_filename::String,
+    config::_AbstractConfiguration,
+)
+    data = _read_data(knapsack_json_filename)
+    return _solve_knapsack(optimizer, data, config)
+end
+
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Main.KnapsackModel</code></pre><p>Finally, you can call your model:</p><pre><code class="language-julia hljs">import .KnapsackModel
+
+KnapsackModel.solve_knapsack(
+    HiGHS.Optimizer,
+    knapsack_json_filename,
+    KnapsackModel.BinaryKnapsackConfig(),
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;cherry&quot;, &quot;banana&quot;, &quot;date&quot;, &quot;eggplant&quot;, &quot;apple&quot;]
+And data, a 5-element Vector{Float64}:
+ -0.0
+  0.0
+  1.0
+  1.0
+  1.0</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The <code>.</code> in <code>.KnapsackModel</code> denotes that it is a submodule and not a separate package that we installed with <code>Pkg.add</code>. If you put the <code>KnapsackModel</code> in a separate file, load it with:</p><pre><code class="language-julia hljs">include(&quot;path/to/KnapsackModel.jl&quot;)
+import .KnapsackModel</code></pre></div></div><h2 id="Add-tests"><a class="docs-heading-anchor" href="#Add-tests">Add tests</a><a id="Add-tests-1"></a><a class="docs-heading-anchor-permalink" href="#Add-tests" title="Permalink"></a></h2><p>As a final step, you should add tests for your model. This often means testing on a small problem for which you can work out the optimal solution by hand. The Julia standard library <code>Test</code> has good unit-testing functionality.</p><pre><code class="language-julia hljs">import .KnapsackModel
+using Test
+
+@testset &quot;KnapsackModel&quot; begin
+    @testset &quot;feasible_binary_knapsack&quot; begin
+        x = KnapsackModel.solve_knapsack(
+            HiGHS.Optimizer,
+            knapsack_json_filename,
+            KnapsackModel.BinaryKnapsackConfig(),
+        )
+        @test isapprox(x[&quot;apple&quot;], 1, atol = 1e-5)
+        @test isapprox(x[&quot;banana&quot;], 0, atol = 1e-5)
+        @test isapprox(x[&quot;cherry&quot;], 0, atol = 1e-5)
+        @test isapprox(x[&quot;date&quot;], 1, atol = 1e-5)
+        @test isapprox(x[&quot;eggplant&quot;], 1, atol = 1e-5)
+    end
+    @testset &quot;feasible_integer_knapsack&quot; begin
+        x = KnapsackModel.solve_knapsack(
+            HiGHS.Optimizer,
+            knapsack_json_filename,
+            KnapsackModel.IntegerKnapsackConfig(),
+        )
+        @test isapprox(x[&quot;apple&quot;], 0, atol = 1e-5)
+        @test isapprox(x[&quot;banana&quot;], 0, atol = 1e-5)
+        @test isapprox(x[&quot;cherry&quot;], 0, atol = 1e-5)
+        @test isapprox(x[&quot;date&quot;], 5, atol = 1e-5)
+        @test isapprox(x[&quot;eggplant&quot;], 0, atol = 1e-5)
+    end
+    @testset &quot;infeasible_binary_knapsack&quot; begin
+        dir = mktempdir()
+        infeasible_filename = joinpath(dir, &quot;infeasible.json&quot;)
+        write(
+            infeasible_filename,
+            &quot;&quot;&quot;{
+                &quot;objects&quot;: {
+                    &quot;apple&quot;: {&quot;profit&quot;: 5.0, &quot;weight&quot;: 2.0},
+                    &quot;banana&quot;: {&quot;profit&quot;: 3.0, &quot;weight&quot;: 8.0},
+                    &quot;cherry&quot;: {&quot;profit&quot;: 2.0, &quot;weight&quot;: 4.0},
+                    &quot;date&quot;: {&quot;profit&quot;: 7.0, &quot;weight&quot;: 2.0},
+                    &quot;eggplant&quot;: {&quot;profit&quot;: 4.0, &quot;weight&quot;: 5.0}
+                },
+                &quot;capacity&quot;: -10.0
+            }&quot;&quot;&quot;,
+        )
+        x = KnapsackModel.solve_knapsack(
+            HiGHS.Optimizer,
+            infeasible_filename,
+            KnapsackModel.BinaryKnapsackConfig(),
+        )
+        @test x === nothing
+    end
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Test.DefaultTestSet(&quot;KnapsackModel&quot;, Any[Test.DefaultTestSet(&quot;feasible_binary_knapsack&quot;, Any[], 5, false, false, true, 1.695444463671342e9, 1.695444463678041e9, false), Test.DefaultTestSet(&quot;feasible_integer_knapsack&quot;, Any[], 5, false, false, true, 1.695444463678091e9, 1.695444463815511e9, false), Test.DefaultTestSet(&quot;infeasible_binary_knapsack&quot;, Any[], 1, false, false, true, 1.695444463815553e9, 1.69544446381692e9, false)], 0, false, false, true, 1.6954444636713e9, 1.695444463816928e9, false)</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Place these tests in a separate file <code>test_knapsack_model.jl</code> so that you can run the tests by adding <code>include(&quot;test_knapsack_model.jl&quot;)</code> to any file where needed.</p></div></div><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><p>We&#39;ve only briefly scratched the surface of ways to create and structure large JuMP models, so consider this tutorial a starting point, rather than a comprehensive list of all the possible ways to structure JuMP models.  If you are embarking on a large project that uses JuMP, a good next step is to look at ways people have written large JuMP projects &quot;in the wild.&quot;</p><p>Here are some good examples (all co-incidentally related to energy):</p><ul><li>AnyMOD.jl<ul><li><a href="https://www.youtube.com/watch?v=QE_tNDER0F4">JuMP-dev 2021 talk</a></li><li><a href="https://github.com/leonardgoeke/AnyMOD.jl">source code</a></li></ul></li><li>PowerModels.jl<ul><li><a href="https://www.youtube.com/watch?v=POOt1FCA8LI">JuMP-dev 2021 talk</a></li><li><a href="https://github.com/lanl-ansi/PowerModels.jl">source code</a></li></ul></li><li>PowerSimulations.jl<ul><li><a href="https://www.youtube.com/watch?v=-ZoO3npjwYU">JuliaCon 2021 talk</a></li><li><a href="https://github.com/NREL-SIIP/PowerSimulations.jl">source code</a></li></ul></li><li>UnitCommitment.jl<ul><li><a href="https://www.youtube.com/watch?v=rYUZK9kYeIY">JuMP-dev 2021 talk</a></li><li><a href="https://github.com/ANL-CEEESA/UnitCommitment.jl">source code</a></li></ul></li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../debugging/">« Debugging</a><a class="docs-footer-nextpage" href="../performance_tips/">Performance tips »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_JuMP.jl b/previews/PR3525/tutorials/getting_started/getting_started_with_JuMP.jl
new file mode 100644
index 00000000000..8d04bfebab1
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_JuMP.jl
@@ -0,0 +1,509 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Getting started with JuMP
+
+# This tutorial is aimed at providing a quick introduction to writing and
+# solving optimization models with JuMP.
+
+# If you're new to Julia, start by reading [Getting started with Julia](@ref).
+
+# ## What is JuMP?
+
+# JuMP ("Julia for Mathematical Programming") is an open-source modeling
+# language that is embedded in Julia. It allows users to formulate various
+# classes of optimization problems (linear, mixed-integer, quadratic, conic
+# quadratic, semidefinite, and nonlinear) with easy-to-read code. These problems
+# can then be solved using state-of-the-art open-source and commercial solvers.
+
+# JuMP also makes advanced optimization techniques easily accessible from a
+# high-level language.
+
+# ## What is a solver?
+
+# A solver is a software package that incorporates algorithms for finding
+# solutions to one or more classes of problem.
+
+# For example, HiGHS is a solver for linear programming (LP) and mixed integer
+# programming (MIP) problems. It incorporates algorithms such as the simplex
+# method and the interior-point method.
+
+# The [Supported-solvers](@ref) table lists the open-source and commercial
+# solvers that JuMP currently supports.
+
+# ## What is MathOptInterface?
+
+# Each solver has its own concepts and data structures for representing
+# optimization models and obtaining results.
+
+# [MathOptInterface](https://github.com/jump-dev/MathOptInterface.jl) (MOI) is
+# an abstraction layer that JuMP uses to convert from the problem written in
+# JuMP to the solver-specific data structures for each solver.
+
+# MOI can be used directly, or through a higher-level modeling interface like
+# JuMP.
+
+# Because JuMP is built on top of MOI, you'll often see the `MathOptInterface.`
+# prefix displayed when JuMP types are printed. However, you'll only need to
+# understand and interact with MOI to accomplish advanced tasks such as creating
+# [solver-independent callbacks](@ref callbacks_manual).
+
+# ## Installation
+
+# JuMP is a package for Julia. From Julia, JuMP is installed by using the
+# built-in package manager.
+
+# ```julia
+# import Pkg
+# Pkg.add("JuMP")
+# ```
+
+# You also need to include a Julia package which provides an appropriate solver.
+# One such solver is `HiGHS.Optimizer`, which is provided by the
+# [HiGHS.jl package](https://github.com/jump-dev/HiGHS.jl).
+# ```julia
+# import Pkg
+# Pkg.add("HiGHS")
+# ```
+# See [Installation Guide](@ref) for a list of other solvers you can use.
+
+# ## An example
+
+# Let's solve the following linear programming problem using JuMP and HiGHS.
+# We will first look at the complete code to solve the problem and then go
+# through it step by step.
+
+# Here's the problem:
+# ```math
+# \begin{aligned}
+# & \min & 12x + 20y \\
+# & \;\;\text{s.t.} & 6x + 8y \geq 100 \\
+# & & 7x + 12y \geq 120 \\
+# & & x \geq 0 \\
+# & & y \in [0, 3] \\
+# \end{aligned}
+# ```
+
+# And here's the code to solve this problem:
+
+using JuMP
+using HiGHS
+model = Model(HiGHS.Optimizer)
+@variable(model, x >= 0)
+@variable(model, 0 <= y <= 3)
+@objective(model, Min, 12x + 20y)
+@constraint(model, c1, 6x + 8y >= 100)
+@constraint(model, c2, 7x + 12y >= 120)
+print(model)
+optimize!(model)
+termination_status(model)
+primal_status(model)
+dual_status(model)
+objective_value(model)
+value(x)
+value(y)
+shadow_price(c1)
+shadow_price(c2)
+
+# ## Step-by-step
+
+# Once JuMP is installed, to use JuMP in your programs write:
+
+using JuMP
+
+# We also need to include a Julia package which provides an appropriate solver.
+# We want to use `HiGHS.Optimizer` here which is provided by the `HiGHS.jl`
+# package:
+
+using HiGHS
+
+# JuMP builds problems incrementally in a `Model` object. Create a model by
+# passing an optimizer to the [`Model`](@ref) function:
+
+model = Model(HiGHS.Optimizer)
+
+# Variables are modeled using [`@variable`](@ref):
+
+@variable(model, x >= 0)
+
+# !!! info
+#     The macro creates a new Julia object, `x`, in the current scope. We could
+#     have made this more explicit by writing:
+#     ```julia
+#     x = @variable(model, x >= 0)
+#     ```
+#     but the macro does this automatically for us to save writing `x` twice.
+
+# Variables can have lower and upper bounds:
+
+@variable(model, 0 <= y <= 30)
+
+# The objective is set using [`@objective`](@ref):
+
+@objective(model, Min, 12x + 20y)
+
+# Constraints are modeled using [`@constraint`](@ref). Here, `c1` and `c2` are
+# the names of our constraint:
+
+@constraint(model, c1, 6x + 8y >= 100)
+@constraint(model, c2, 7x + 12y >= 120)
+
+# Call `print` to display the model:
+
+print(model)
+
+# To solve the optimization problem, call the [`optimize!`](@ref) function:
+
+optimize!(model)
+
+# !!! info
+#     The `!` after optimize is part of the name. It's nothing special.
+#     Julia has a convention that functions which mutate their arguments should
+#     end in `!`. A common example is `push!`.
+
+# Now let's see what information we can query about the solution.
+
+# [`termination_status`](@ref) tells us why the solver stopped:
+
+termination_status(model)
+
+# In this case, the solver found an optimal solution.
+
+# Check [`primal_status`](@ref) to see if the solver found a primal feasible
+# point:
+
+primal_status(model)
+
+# and [`dual_status`](@ref) to see if the solver found a dual feasible point:
+
+dual_status(model)
+
+# Now we know that our solver found an optimal solution, and that it has a
+# primal and a dual solution to query.
+
+# Query the objective value using [`objective_value`](@ref):
+
+objective_value(model)
+
+# the primal solution using [`value`](@ref):
+
+value(x)
+value(y)
+
+# and the dual solution using [`shadow_price`](@ref):
+
+shadow_price(c1)
+shadow_price(c2)
+
+# That's it for our simple model. In the rest of this tutorial, we expand on
+# some of the basic JuMP operations.
+
+# ## Model basics
+
+# Create a model by passing an optimizer:
+
+model = Model(HiGHS.Optimizer)
+
+# Alternatively, call [`set_optimizer`](@ref) at any point before calling
+# [`optimize!`](@ref):
+
+model = Model()
+set_optimizer(model, HiGHS.Optimizer)
+
+# For some solvers, you can also use [`direct_model`](@ref), which offers a more
+# efficient connection to the underlying solver:
+
+model = direct_model(HiGHS.Optimizer())
+
+# !!! warning
+#     Some solvers do not support [`direct_model`](@ref)!
+
+# ### Solver Options
+
+# Pass options to solvers with [`optimizer_with_attributes`](@ref):
+
+model =
+    Model(optimizer_with_attributes(HiGHS.Optimizer, "output_flag" => false))
+
+# !!! note
+#     These options are solver-specific. To find out the various options
+#     available, see the GitHub README of the individual solver packages. The
+#     link to each solver's GitHub page is in the [Supported solvers](@ref)
+#     table.
+
+# You can also pass options with [`set_attribute`](@ref):
+
+model = Model(HiGHS.Optimizer)
+set_attribute(model, "output_flag", false)
+
+# ## Solution basics
+
+# We saw above how to use [`termination_status`](@ref) and
+# [`primal_status`](@ref) to understand the solution returned by the solver.
+
+# However, only query solution attributes like [`value`](@ref) and
+# [`objective_value`](@ref) if there is an available solution. Here's a
+# recommended way to check:
+
+function solve_infeasible()
+    model = Model(HiGHS.Optimizer)
+    @variable(model, 0 <= x <= 1)
+    @variable(model, 0 <= y <= 1)
+    @constraint(model, x + y >= 3)
+    @objective(model, Max, x + 2y)
+    optimize!(model)
+    if termination_status(model) != OPTIMAL
+        @warn("The model was not solved correctly.")
+        return
+    end
+    return value(x), value(y)
+end
+
+solve_infeasible()
+
+# ## Variable basics
+
+# Let's create a new empty model to explain some of the variable syntax:
+
+model = Model()
+
+# ### Variable bounds
+
+# All of the variables we have created till now have had a bound. We can also
+# create a free variable.
+
+@variable(model, free_x)
+
+# While creating a variable, instead of using the <= and >= syntax, we can also
+# use the `lower_bound` and `upper_bound` keyword arguments.
+
+@variable(model, keyword_x, lower_bound = 1, upper_bound = 2)
+
+# We can query whether a variable has a bound using the `has_lower_bound` and
+# `has_upper_bound` functions. The values of the bound can be obtained using the
+# `lower_bound` and `upper_bound` functions.
+
+has_upper_bound(keyword_x)
+upper_bound(keyword_x)
+
+# Note querying the value of a bound that does not exist will result in an error.
+
+try                         #hide
+    lower_bound(free_x)
+catch err                   #hide
+    showerror(stderr, err)  #hide
+end                         #hide
+
+# ### [Containers](@id tutorial_variable_container)
+
+# We have already seen how to add a single variable to a model using the
+# [`@variable`](@ref) macro. Now let's look at ways to add multiple variables to
+# a model.
+
+# JuMP provides data structures for adding collections of variables to a model.
+# These data structures are referred to as _containers_ and are of three types:
+# `Arrays`, `DenseAxisArrays`, and `SparseAxisArrays`.
+
+# #### Arrays
+
+# JuMP arrays are created when you have integer indices that start at `1`:
+
+@variable(model, a[1:2, 1:2])
+
+# Index elements in `a` as follows:
+
+a[1, 1]
+
+#-
+
+a[2, :]
+
+# Create an n-dimensional variable $x \in {R}^n$ with bounds $l \le x \le u$
+# ($l, u \in {R}^n$) as follows:
+
+n = 10
+l = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10];
+u = [10, 11, 12, 13, 14, 15, 16, 17, 18, 19];
+
+@variable(model, l[i] <= x[i = 1:n] <= u[i])
+
+# We can also create variable bounds that depend upon the indices:
+
+@variable(model, y[i = 1:2, j = 1:2] >= 2i + j)
+
+# #### DenseAxisArrays
+
+# `DenseAxisArrays` are used when the indices are not one-based integer ranges.
+# The syntax is similar except with an arbitrary vector as an index as opposed
+# to a one-based range:
+
+@variable(model, z[i = 2:3, j = 1:2:3] >= 0)
+
+# Indices do not have to be integers. They can be any Julia type:
+
+@variable(model, w[1:5, ["red", "blue"]] <= 1)
+
+# Index elements in a `DenseAxisArray` as follows:
+
+z[2, 1]
+
+#-
+
+w[2:3, ["red", "blue"]]
+
+# See [Forcing the container type](@ref variable_forcing) for more details.
+
+# #### SparseAxisArrays
+
+# `SparseAxisArrays` are created when the indices do not form a Cartesian product.
+# For example, this applies when indices have a dependence upon previous indices
+# (called triangular indexing):
+
+@variable(model, u[i = 1:2, j = i:3])
+
+# We can also conditionally create variables by adding a comparison check that
+# depends upon the named indices and is separated from the indices by a
+# semi-colon `;`:
+
+@variable(model, v[i = 1:9; mod(i, 3) == 0])
+
+# Index elements in a `DenseAxisArray` as follows:
+
+u[1, 2]
+
+#-
+
+v[[3, 6]]
+
+# ### Integrality
+
+# JuMP can create binary and integer variables. Binary variables are constrained
+# to the set  ``\{0, 1\}``, and integer variables are constrained to the set
+# ``\mathbb{Z}``.
+
+# #### Integer variables
+
+# Create an integer variable by passing  `Int`:
+
+@variable(model, integer_x, Int)
+
+# or setting the `integer` keyword to `true`:
+
+@variable(model, integer_z, integer = true)
+
+# #### Binary variables
+
+# Create a binary variable by passing `Bin`:
+
+@variable(model, binary_x, Bin)
+
+# or setting the `binary` keyword to `true`:
+
+@variable(model, binary_z, binary = true)
+
+# ## Constraint basics
+
+# We'll need a new model to explain some of the constraint  basics:
+
+model = Model()
+@variable(model, x)
+@variable(model, y)
+@variable(model, z[1:10]);
+
+# ### [Containers](@id tutorial_constraint_container)
+
+# Just as we had containers for variables, JuMP also provides `Arrays`,
+# `DenseAxisArrays`, and `SparseAxisArrays` for storing collections of
+# constraints. Examples for each container type are given below.
+
+# #### Arrays
+
+# Create an `Array` of constraints:
+
+@constraint(model, [i = 1:3], i * x <= i + 1)
+
+# #### DenseAxisArrays
+
+# Create an `DenseAxisArray` of constraints:
+
+@constraint(model, [i = 1:2, j = 2:3], i * x <= j + 1)
+
+# #### SparseAxisArrays
+
+# Create an `SparseAxisArray` of constraints:
+
+@constraint(model, [i = 1:2, j = 1:2; i != j], i * x <= j + 1)
+
+# ### Constraints in a loop
+
+# We can add constraints using regular Julia loops:
+
+for i in 1:3
+    @constraint(model, 6x + 4y >= 5i)
+end
+
+# or use for each loops inside the `@constraint` macro:
+
+@constraint(model, [i in 1:3], 6x + 4y >= 5i)
+
+# We can also create constraints such as $\sum _{i = 1}^{10} z_i \leq 1$:
+
+@constraint(model, sum(z[i] for i in 1:10) <= 1)
+
+# ## Objective functions
+
+# Set an objective function with [`@objective`](@ref):
+
+model = Model(HiGHS.Optimizer)
+@variable(model, x >= 0)
+@variable(model, y >= 0)
+@objective(model, Min, 2x + y)
+
+# Create a maximization objective using `Max`:
+
+@objective(model, Max, 2x + y)
+
+# !!! tip
+#     Calling [`@objective`](@ref) multiple times will over-write the previous
+#     objective. This can be useful when you want to solve the same problem with
+#     different objectives.
+
+# ## Vectorized syntax
+
+# We can also add constraints and an objective to JuMP using vectorized linear
+# algebra. We'll illustrate this by solving an LP in standard form that is,
+
+# ```math
+# \begin{aligned}
+# & \min & c^T x \\
+# & \;\;\text{s.t.} & A x = b \\
+# & & x \ge 0
+# \end{aligned}
+# ```
+
+vector_model = Model(HiGHS.Optimizer)
+A = [1 1 9 5; 3 5 0 8; 2 0 6 13]
+b = [7, 3, 5]
+c = [1, 3, 5, 2]
+@variable(vector_model, x[1:4] >= 0)
+@constraint(vector_model, A * x .== b)
+@objective(vector_model, Min, c' * x)
+optimize!(vector_model)
+objective_value(vector_model)
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_JuMP/index.html b/previews/PR3525/tutorials/getting_started/getting_started_with_JuMP/index.html
new file mode 100644
index 00000000000..b067c3758b0
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_JuMP/index.html
@@ -0,0 +1,224 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Getting started with JuMP · JuMP</title><meta name="title" content="Getting started with JuMP · JuMP"/><meta property="og:title" content="Getting started with JuMP · JuMP"/><meta property="twitter:title" content="Getting started with JuMP · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../getting_started_with_julia/">Getting started with Julia</a></li><li class="is-active"><a class="tocitem" href>Getting started with JuMP</a><ul class="internal"><li><a class="tocitem" href="#What-is-JuMP?"><span>What is JuMP?</span></a></li><li><a class="tocitem" href="#What-is-a-solver?"><span>What is a solver?</span></a></li><li><a class="tocitem" href="#What-is-MathOptInterface?"><span>What is MathOptInterface?</span></a></li><li><a class="tocitem" href="#Installation"><span>Installation</span></a></li><li><a class="tocitem" href="#An-example"><span>An example</span></a></li><li><a class="tocitem" href="#Step-by-step"><span>Step-by-step</span></a></li><li><a class="tocitem" href="#Model-basics"><span>Model basics</span></a></li><li><a class="tocitem" href="#Solution-basics"><span>Solution basics</span></a></li><li><a class="tocitem" href="#Variable-basics"><span>Variable basics</span></a></li><li><a class="tocitem" href="#Constraint-basics"><span>Constraint basics</span></a></li><li><a class="tocitem" href="#Objective-functions"><span>Objective functions</span></a></li><li><a class="tocitem" href="#Vectorized-syntax"><span>Vectorized syntax</span></a></li></ul></li><li><a class="tocitem" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../debugging/">Debugging</a></li><li><a class="tocitem" href="../design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Getting started with JuMP</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Getting started with JuMP</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/getting_started_with_JuMP.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Getting-started-with-JuMP"><a class="docs-heading-anchor" href="#Getting-started-with-JuMP">Getting started with JuMP</a><a id="Getting-started-with-JuMP-1"></a><a class="docs-heading-anchor-permalink" href="#Getting-started-with-JuMP" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../getting_started_with_JuMP.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial is aimed at providing a quick introduction to writing and solving optimization models with JuMP.</p><p>If you&#39;re new to Julia, start by reading <a href="../getting_started_with_julia/#Getting-started-with-Julia">Getting started with Julia</a>.</p><h2 id="What-is-JuMP?"><a class="docs-heading-anchor" href="#What-is-JuMP?">What is JuMP?</a><a id="What-is-JuMP?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-JuMP?" title="Permalink"></a></h2><p>JuMP (&quot;Julia for Mathematical Programming&quot;) is an open-source modeling language that is embedded in Julia. It allows users to formulate various classes of optimization problems (linear, mixed-integer, quadratic, conic quadratic, semidefinite, and nonlinear) with easy-to-read code. These problems can then be solved using state-of-the-art open-source and commercial solvers.</p><p>JuMP also makes advanced optimization techniques easily accessible from a high-level language.</p><h2 id="What-is-a-solver?"><a class="docs-heading-anchor" href="#What-is-a-solver?">What is a solver?</a><a id="What-is-a-solver?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-a-solver?" title="Permalink"></a></h2><p>A solver is a software package that incorporates algorithms for finding solutions to one or more classes of problem.</p><p>For example, HiGHS is a solver for linear programming (LP) and mixed integer programming (MIP) problems. It incorporates algorithms such as the simplex method and the interior-point method.</p><p>The <a href="../../../installation/#Supported-solvers">Supported-solvers</a> table lists the open-source and commercial solvers that JuMP currently supports.</p><h2 id="What-is-MathOptInterface?"><a class="docs-heading-anchor" href="#What-is-MathOptInterface?">What is MathOptInterface?</a><a id="What-is-MathOptInterface?-1"></a><a class="docs-heading-anchor-permalink" href="#What-is-MathOptInterface?" title="Permalink"></a></h2><p>Each solver has its own concepts and data structures for representing optimization models and obtaining results.</p><p><a href="https://github.com/jump-dev/MathOptInterface.jl">MathOptInterface</a> (MOI) is an abstraction layer that JuMP uses to convert from the problem written in JuMP to the solver-specific data structures for each solver.</p><p>MOI can be used directly, or through a higher-level modeling interface like JuMP.</p><p>Because JuMP is built on top of MOI, you&#39;ll often see the <code>MathOptInterface.</code> prefix displayed when JuMP types are printed. However, you&#39;ll only need to understand and interact with MOI to accomplish advanced tasks such as creating <a href="../../../manual/callbacks/#callbacks_manual">solver-independent callbacks</a>.</p><h2 id="Installation"><a class="docs-heading-anchor" href="#Installation">Installation</a><a id="Installation-1"></a><a class="docs-heading-anchor-permalink" href="#Installation" title="Permalink"></a></h2><p>JuMP is a package for Julia. From Julia, JuMP is installed by using the built-in package manager.</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;JuMP&quot;)</code></pre><p>You also need to include a Julia package which provides an appropriate solver. One such solver is <code>HiGHS.Optimizer</code>, which is provided by the <a href="https://github.com/jump-dev/HiGHS.jl">HiGHS.jl package</a>.</p><pre><code class="language-julia hljs">import Pkg
+Pkg.add(&quot;HiGHS&quot;)</code></pre><p>See <a href="../../../installation/#Installation-Guide">Installation Guide</a> for a list of other solvers you can use.</p><h2 id="An-example"><a class="docs-heading-anchor" href="#An-example">An example</a><a id="An-example-1"></a><a class="docs-heading-anchor-permalink" href="#An-example" title="Permalink"></a></h2><p>Let&#39;s solve the following linear programming problem using JuMP and HiGHS. We will first look at the complete code to solve the problem and then go through it step by step.</p><p>Here&#39;s the problem:</p><p class="math-container">\[\begin{aligned}
+&amp; \min &amp; 12x + 20y \\
+&amp; \;\;\text{s.t.} &amp; 6x + 8y \geq 100 \\
+&amp; &amp; 7x + 12y \geq 120 \\
+&amp; &amp; x \geq 0 \\
+&amp; &amp; y \in [0, 3] \\
+\end{aligned}\]</p><p>And here&#39;s the code to solve this problem:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; using HiGHS</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer)</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, 0 &lt;= y &lt;= 3)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, 12x + 20y)</code><code class="nohighlight hljs ansi" style="display:block;">12 x + 20 y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, c1, 6x + 8y &gt;= 100)</code><code class="nohighlight hljs ansi" style="display:block;">c1 : 6 x + 8 y ≥ 100</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, c2, 7x + 12y &gt;= 120)</code><code class="nohighlight hljs ansi" style="display:block;">c2 : 7 x + 12 y ≥ 120</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; print(model)</code><code class="nohighlight hljs ansi" style="display:block;">Min 12 x + 20 y
+Subject to
+ c1 : 6 x + 8 y ≥ 100
+ c2 : 7 x + 12 y ≥ 120
+ x ≥ 0
+ y ≥ 0
+ y ≤ 3</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;">Running HiGHS 1.5.3 [date: 1970-01-01, git hash: 45a127b78]
+Copyright (c) 2023 HiGHS under MIT licence terms
+Presolving model
+2 rows, 2 cols, 4 nonzeros
+2 rows, 2 cols, 4 nonzeros
+Presolve : Reductions: rows 2(-0); columns 2(-0); elements 4(-0) - Not reduced
+Problem not reduced by presolve: solving the LP
+Using EKK dual simplex solver - serial
+  Iteration        Objective     Infeasibilities num(sum)
+          0     0.0000000000e+00 Pr: 2(220) 0s
+          2     2.0500000000e+02 Pr: 0(0) 0s
+Model   status      : Optimal
+Simplex   iterations: 2
+Objective value     :  2.0500000000e+02
+HiGHS run time      :          0.00</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; termination_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">OPTIMAL::TerminationStatusCode = 1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; primal_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">FEASIBLE_POINT::ResultStatusCode = 1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; dual_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">FEASIBLE_POINT::ResultStatusCode = 1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; objective_value(model)</code><code class="nohighlight hljs ansi" style="display:block;">204.99999999999997</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; value(x)</code><code class="nohighlight hljs ansi" style="display:block;">15.000000000000005</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; value(y)</code><code class="nohighlight hljs ansi" style="display:block;">1.249999999999996</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; shadow_price(c1)</code><code class="nohighlight hljs ansi" style="display:block;">-0.24999999999999922</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; shadow_price(c2)</code><code class="nohighlight hljs ansi" style="display:block;">-1.5000000000000007</code></pre><h2 id="Step-by-step"><a class="docs-heading-anchor" href="#Step-by-step">Step-by-step</a><a id="Step-by-step-1"></a><a class="docs-heading-anchor-permalink" href="#Step-by-step" title="Permalink"></a></h2><p>Once JuMP is installed, to use JuMP in your programs write:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><p>We also need to include a Julia package which provides an appropriate solver. We want to use <code>HiGHS.Optimizer</code> here which is provided by the <code>HiGHS.jl</code> package:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using HiGHS</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><p>JuMP builds problems incrementally in a <code>Model</code> object. Create a model by passing an optimizer to the <a href="../../../api/JuMP/#JuMP.Model"><code>Model</code></a> function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer)</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><p>Variables are modeled using <a href="../../../api/JuMP/#JuMP.@variable"><code>@variable</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">x</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The macro creates a new Julia object, <code>x</code>, in the current scope. We could have made this more explicit by writing:</p><pre><code class="language-julia hljs">x = @variable(model, x &gt;= 0)</code></pre><p>but the macro does this automatically for us to save writing <code>x</code> twice.</p></div></div><p>Variables can have lower and upper bounds:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, 0 &lt;= y &lt;= 30)</code><code class="nohighlight hljs ansi" style="display:block;">y</code></pre><p>The objective is set using <a href="../../../api/JuMP/#JuMP.@objective"><code>@objective</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, 12x + 20y)</code><code class="nohighlight hljs ansi" style="display:block;">12 x + 20 y</code></pre><p>Constraints are modeled using <a href="../../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a>. Here, <code>c1</code> and <code>c2</code> are the names of our constraint:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, c1, 6x + 8y &gt;= 100)</code><code class="nohighlight hljs ansi" style="display:block;">c1 : 6 x + 8 y ≥ 100</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, c2, 7x + 12y &gt;= 120)</code><code class="nohighlight hljs ansi" style="display:block;">c2 : 7 x + 12 y ≥ 120</code></pre><p>Call <code>print</code> to display the model:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; print(model)</code><code class="nohighlight hljs ansi" style="display:block;">Min 12 x + 20 y
+Subject to
+ c1 : 6 x + 8 y ≥ 100
+ c2 : 7 x + 12 y ≥ 120
+ x ≥ 0
+ y ≥ 0
+ y ≤ 30</code></pre><p>To solve the optimization problem, call the <a href="../../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a> function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(model)</code><code class="nohighlight hljs ansi" style="display:block;">Running HiGHS 1.5.3 [date: 1970-01-01, git hash: 45a127b78]
+Copyright (c) 2023 HiGHS under MIT licence terms
+Presolving model
+2 rows, 2 cols, 4 nonzeros
+2 rows, 2 cols, 4 nonzeros
+Presolve : Reductions: rows 2(-0); columns 2(-0); elements 4(-0) - Not reduced
+Problem not reduced by presolve: solving the LP
+Using EKK dual simplex solver - serial
+  Iteration        Objective     Infeasibilities num(sum)
+          0     0.0000000000e+00 Pr: 2(220) 0s
+          2     2.0500000000e+02 Pr: 0(0) 0s
+Model   status      : Optimal
+Simplex   iterations: 2
+Objective value     :  2.0500000000e+02
+HiGHS run time      :          0.00</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The <code>!</code> after optimize is part of the name. It&#39;s nothing special. Julia has a convention that functions which mutate their arguments should end in <code>!</code>. A common example is <code>push!</code>.</p></div></div><p>Now let&#39;s see what information we can query about the solution.</p><p><a href="../../../api/JuMP/#JuMP.termination_status"><code>termination_status</code></a> tells us why the solver stopped:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; termination_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">OPTIMAL::TerminationStatusCode = 1</code></pre><p>In this case, the solver found an optimal solution.</p><p>Check <a href="../../../api/JuMP/#JuMP.primal_status"><code>primal_status</code></a> to see if the solver found a primal feasible point:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; primal_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">FEASIBLE_POINT::ResultStatusCode = 1</code></pre><p>and <a href="../../../api/JuMP/#JuMP.dual_status"><code>dual_status</code></a> to see if the solver found a dual feasible point:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; dual_status(model)</code><code class="nohighlight hljs ansi" style="display:block;">FEASIBLE_POINT::ResultStatusCode = 1</code></pre><p>Now we know that our solver found an optimal solution, and that it has a primal and a dual solution to query.</p><p>Query the objective value using <a href="../../../api/JuMP/#JuMP.objective_value"><code>objective_value</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; objective_value(model)</code><code class="nohighlight hljs ansi" style="display:block;">205.0</code></pre><p>the primal solution using <a href="../../../api/JuMP/#JuMP.value"><code>value</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; value(x)</code><code class="nohighlight hljs ansi" style="display:block;">15.000000000000004</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; value(y)</code><code class="nohighlight hljs ansi" style="display:block;">1.2499999999999976</code></pre><p>and the dual solution using <a href="../../../api/JuMP/#JuMP.shadow_price"><code>shadow_price</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; shadow_price(c1)</code><code class="nohighlight hljs ansi" style="display:block;">-0.24999999999999917</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; shadow_price(c2)</code><code class="nohighlight hljs ansi" style="display:block;">-1.5000000000000007</code></pre><p>That&#39;s it for our simple model. In the rest of this tutorial, we expand on some of the basic JuMP operations.</p><h2 id="Model-basics"><a class="docs-heading-anchor" href="#Model-basics">Model basics</a><a id="Model-basics-1"></a><a class="docs-heading-anchor-permalink" href="#Model-basics" title="Permalink"></a></h2><p>Create a model by passing an optimizer:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer)</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><p>Alternatively, call <a href="../../../api/JuMP/#JuMP.set_optimizer"><code>set_optimizer</code></a> at any point before calling <a href="../../../api/JuMP/#JuMP.optimize!"><code>optimize!</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model()</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_optimizer(model, HiGHS.Optimizer)</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><p>For some solvers, you can also use <a href="../../../api/JuMP/#JuMP.direct_model"><code>direct_model</code></a>, which offers a more efficient connection to the underlying solver:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = direct_model(HiGHS.Optimizer())</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: DIRECT
+Solver name: HiGHS</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Some solvers do not support <a href="../../../api/JuMP/#JuMP.direct_model"><code>direct_model</code></a>!</p></div></div><h3 id="Solver-Options"><a class="docs-heading-anchor" href="#Solver-Options">Solver Options</a><a id="Solver-Options-1"></a><a class="docs-heading-anchor-permalink" href="#Solver-Options" title="Permalink"></a></h3><p>Pass options to solvers with <a href="../../../api/JuMP/#JuMP.optimizer_with_attributes"><code>optimizer_with_attributes</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model =
+           Model(optimizer_with_attributes(HiGHS.Optimizer, &quot;output_flag&quot; =&gt; false))</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>These options are solver-specific. To find out the various options available, see the GitHub README of the individual solver packages. The link to each solver&#39;s GitHub page is in the <a href="../../../installation/#Supported-solvers">Supported solvers</a> table.</p></div></div><p>You can also pass options with <a href="../../../api/JuMP/#JuMP.set_attribute"><code>set_attribute</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer)</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_attribute(model, &quot;output_flag&quot;, false)</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><h2 id="Solution-basics"><a class="docs-heading-anchor" href="#Solution-basics">Solution basics</a><a id="Solution-basics-1"></a><a class="docs-heading-anchor-permalink" href="#Solution-basics" title="Permalink"></a></h2><p>We saw above how to use <a href="../../../api/JuMP/#JuMP.termination_status"><code>termination_status</code></a> and <a href="../../../api/JuMP/#JuMP.primal_status"><code>primal_status</code></a> to understand the solution returned by the solver.</p><p>However, only query solution attributes like <a href="../../../api/JuMP/#JuMP.value"><code>value</code></a> and <a href="../../../api/JuMP/#JuMP.objective_value"><code>objective_value</code></a> if there is an available solution. Here&#39;s a recommended way to check:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; function solve_infeasible()
+           model = Model(HiGHS.Optimizer)
+           @variable(model, 0 &lt;= x &lt;= 1)
+           @variable(model, 0 &lt;= y &lt;= 1)
+           @constraint(model, x + y &gt;= 3)
+           @objective(model, Max, x + 2y)
+           optimize!(model)
+           if termination_status(model) != OPTIMAL
+               @warn(&quot;The model was not solved correctly.&quot;)
+               return
+           end
+           return value(x), value(y)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">solve_infeasible (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; solve_infeasible()</code><code class="nohighlight hljs ansi" style="display:block;">Running HiGHS 1.5.3 [date: 1970-01-01, git hash: 45a127b78]
+Copyright (c) 2023 HiGHS under MIT licence terms
+Presolving model
+Problem status detected on presolve: Infeasible
+Model   status      : Infeasible
+Objective value     :  0.0000000000e+00
+HiGHS run time      :          0.00
+ERROR:   No invertible representation for getDualRay
+<span class="sgr33"><span class="sgr1">┌ Warning: </span></span>The model was not solved correctly.
+<span class="sgr33"><span class="sgr1">└ </span></span><span class="sgr90">@ Main REPL[1]:9</span></code></pre><h2 id="Variable-basics"><a class="docs-heading-anchor" href="#Variable-basics">Variable basics</a><a id="Variable-basics-1"></a><a class="docs-heading-anchor-permalink" href="#Variable-basics" title="Permalink"></a></h2><p>Let&#39;s create a new empty model to explain some of the variable syntax:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model()</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.</code></pre><h3 id="Variable-bounds"><a class="docs-heading-anchor" href="#Variable-bounds">Variable bounds</a><a id="Variable-bounds-1"></a><a class="docs-heading-anchor-permalink" href="#Variable-bounds" title="Permalink"></a></h3><p>All of the variables we have created till now have had a bound. We can also create a free variable.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, free_x)</code><code class="nohighlight hljs ansi" style="display:block;">free_x</code></pre><p>While creating a variable, instead of using the &lt;= and &gt;= syntax, we can also use the <code>lower_bound</code> and <code>upper_bound</code> keyword arguments.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, keyword_x, lower_bound = 1, upper_bound = 2)</code><code class="nohighlight hljs ansi" style="display:block;">keyword_x</code></pre><p>We can query whether a variable has a bound using the <code>has_lower_bound</code> and <code>has_upper_bound</code> functions. The values of the bound can be obtained using the <code>lower_bound</code> and <code>upper_bound</code> functions.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; has_upper_bound(keyword_x)</code><code class="nohighlight hljs ansi" style="display:block;">true</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; upper_bound(keyword_x)</code><code class="nohighlight hljs ansi" style="display:block;">2.0</code></pre><p>Note querying the value of a bound that does not exist will result in an error.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt;     lower_bound(free_x)</code><code class="nohighlight hljs ansi" style="display:block;">Variable free_x does not have a lower bound.</code></pre><h3 id="tutorial_variable_container"><a class="docs-heading-anchor" href="#tutorial_variable_container">Containers</a><a id="tutorial_variable_container-1"></a><a class="docs-heading-anchor-permalink" href="#tutorial_variable_container" title="Permalink"></a></h3><p>We have already seen how to add a single variable to a model using the <a href="../../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> macro. Now let&#39;s look at ways to add multiple variables to a model.</p><p>JuMP provides data structures for adding collections of variables to a model. These data structures are referred to as <em>containers</em> and are of three types: <code>Arrays</code>, <code>DenseAxisArrays</code>, and <code>SparseAxisArrays</code>.</p><h4 id="Arrays"><a class="docs-heading-anchor" href="#Arrays">Arrays</a><a id="Arrays-1"></a><a class="docs-heading-anchor-permalink" href="#Arrays" title="Permalink"></a></h4><p>JuMP arrays are created when you have integer indices that start at <code>1</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, a[1:2, 1:2])</code><code class="nohighlight hljs ansi" style="display:block;">2×2 Matrix{VariableRef}:
+ a[1,1]  a[1,2]
+ a[2,1]  a[2,2]</code></pre><p>Index elements in <code>a</code> as follows:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; a[1, 1]</code><code class="nohighlight hljs ansi" style="display:block;">a[1,1]</code></pre><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; a[2, :]</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ a[2,1]
+ a[2,2]</code></pre><p>Create an n-dimensional variable <span>$x \in {R}^n$</span> with bounds <span>$l \le x \le u$</span> (<span>$l, u \in {R}^n$</span>) as follows:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; n = 10</code><code class="nohighlight hljs ansi" style="display:block;">10</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; l = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10];</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; u = [10, 11, 12, 13, 14, 15, 16, 17, 18, 19];</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, l[i] &lt;= x[i = 1:n] &lt;= u[i])</code><code class="nohighlight hljs ansi" style="display:block;">10-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+ x[4]
+ x[5]
+ x[6]
+ x[7]
+ x[8]
+ x[9]
+ x[10]</code></pre><p>We can also create variable bounds that depend upon the indices:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, y[i = 1:2, j = 1:2] &gt;= 2i + j)</code><code class="nohighlight hljs ansi" style="display:block;">2×2 Matrix{VariableRef}:
+ y[1,1]  y[1,2]
+ y[2,1]  y[2,2]</code></pre><h4 id="DenseAxisArrays"><a class="docs-heading-anchor" href="#DenseAxisArrays">DenseAxisArrays</a><a id="DenseAxisArrays-1"></a><a class="docs-heading-anchor-permalink" href="#DenseAxisArrays" title="Permalink"></a></h4><p><code>DenseAxisArrays</code> are used when the indices are not one-based integer ranges. The syntax is similar except with an arbitrary vector as an index as opposed to a one-based range:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, z[i = 2:3, j = 1:2:3] &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:
+    Dimension 1, 2:3
+    Dimension 2, 1:2:3
+And data, a 2×2 Matrix{VariableRef}:
+ z[2,1]  z[2,3]
+ z[3,1]  z[3,3]</code></pre><p>Indices do not have to be integers. They can be any Julia type:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, w[1:5, [&quot;red&quot;, &quot;blue&quot;]] &lt;= 1)</code><code class="nohighlight hljs ansi" style="display:block;">2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:
+    Dimension 1, Base.OneTo(5)
+    Dimension 2, [&quot;red&quot;, &quot;blue&quot;]
+And data, a 5×2 Matrix{VariableRef}:
+ w[1,red]  w[1,blue]
+ w[2,red]  w[2,blue]
+ w[3,red]  w[3,blue]
+ w[4,red]  w[4,blue]
+ w[5,red]  w[5,blue]</code></pre><p>Index elements in a <code>DenseAxisArray</code> as follows:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; z[2, 1]</code><code class="nohighlight hljs ansi" style="display:block;">z[2,1]</code></pre><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; w[2:3, [&quot;red&quot;, &quot;blue&quot;]]</code><code class="nohighlight hljs ansi" style="display:block;">2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:
+    Dimension 1, [2, 3]
+    Dimension 2, [&quot;red&quot;, &quot;blue&quot;]
+And data, a 2×2 Matrix{VariableRef}:
+ w[2,red]  w[2,blue]
+ w[3,red]  w[3,blue]</code></pre><p>See <a href="../../../manual/variables/#variable_forcing">Forcing the container type</a> for more details.</p><h4 id="SparseAxisArrays"><a class="docs-heading-anchor" href="#SparseAxisArrays">SparseAxisArrays</a><a id="SparseAxisArrays-1"></a><a class="docs-heading-anchor-permalink" href="#SparseAxisArrays" title="Permalink"></a></h4><p><code>SparseAxisArrays</code> are created when the indices do not form a Cartesian product. For example, this applies when indices have a dependence upon previous indices (called triangular indexing):</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, u[i = 1:2, j = i:3])</code><code class="nohighlight hljs ansi" style="display:block;">SparseAxisArray{VariableRef, 2, Tuple{Int64, Int64}} with 5 entries:
+  [1, 1]  =  u[1,1]
+  [1, 2]  =  u[1,2]
+  [1, 3]  =  u[1,3]
+  [2, 2]  =  u[2,2]
+  [2, 3]  =  u[2,3]</code></pre><p>We can also conditionally create variables by adding a comparison check that depends upon the named indices and is separated from the indices by a semi-colon <code>;</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, v[i = 1:9; mod(i, 3) == 0])</code><code class="nohighlight hljs ansi" style="display:block;">SparseAxisArray{VariableRef, 1, Tuple{Int64}} with 3 entries:
+  [3]  =  v[3]
+  [6]  =  v[6]
+  [9]  =  v[9]</code></pre><p>Index elements in a <code>DenseAxisArray</code> as follows:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; u[1, 2]</code><code class="nohighlight hljs ansi" style="display:block;">u[1,2]</code></pre><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; v[[3, 6]]</code><code class="nohighlight hljs ansi" style="display:block;">SparseAxisArray{VariableRef, 1, Tuple{Int64}} with 2 entries:
+  [3]  =  v[3]
+  [6]  =  v[6]</code></pre><h3 id="Integrality"><a class="docs-heading-anchor" href="#Integrality">Integrality</a><a id="Integrality-1"></a><a class="docs-heading-anchor-permalink" href="#Integrality" title="Permalink"></a></h3><p>JuMP can create binary and integer variables. Binary variables are constrained to the set  <span>$\{0, 1\}$</span>, and integer variables are constrained to the set <span>$\mathbb{Z}$</span>.</p><h4 id="Integer-variables"><a class="docs-heading-anchor" href="#Integer-variables">Integer variables</a><a id="Integer-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Integer-variables" title="Permalink"></a></h4><p>Create an integer variable by passing  <code>Int</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, integer_x, Int)</code><code class="nohighlight hljs ansi" style="display:block;">integer_x</code></pre><p>or setting the <code>integer</code> keyword to <code>true</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, integer_z, integer = true)</code><code class="nohighlight hljs ansi" style="display:block;">integer_z</code></pre><h4 id="Binary-variables"><a class="docs-heading-anchor" href="#Binary-variables">Binary variables</a><a id="Binary-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Binary-variables" title="Permalink"></a></h4><p>Create a binary variable by passing <code>Bin</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, binary_x, Bin)</code><code class="nohighlight hljs ansi" style="display:block;">binary_x</code></pre><p>or setting the <code>binary</code> keyword to <code>true</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, binary_z, binary = true)</code><code class="nohighlight hljs ansi" style="display:block;">binary_z</code></pre><h2 id="Constraint-basics"><a class="docs-heading-anchor" href="#Constraint-basics">Constraint basics</a><a id="Constraint-basics-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-basics" title="Permalink"></a></h2><p>We&#39;ll need a new model to explain some of the constraint  basics:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model()</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, y)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, z[1:10]);</code><code class="nohighlight hljs ansi" style="display:block;"></code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><h3 id="tutorial_constraint_container"><a class="docs-heading-anchor" href="#tutorial_constraint_container">Containers</a><a id="tutorial_constraint_container-1"></a><a class="docs-heading-anchor-permalink" href="#tutorial_constraint_container" title="Permalink"></a></h3><p>Just as we had containers for variables, JuMP also provides <code>Arrays</code>, <code>DenseAxisArrays</code>, and <code>SparseAxisArrays</code> for storing collections of constraints. Examples for each container type are given below.</p><h4 id="Arrays-2"><a class="docs-heading-anchor" href="#Arrays-2">Arrays</a><a class="docs-heading-anchor-permalink" href="#Arrays-2" title="Permalink"></a></h4><p>Create an <code>Array</code> of constraints:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, [i = 1:3], i * x &lt;= i + 1)</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ x ≤ 2
+ 2 x ≤ 3
+ 3 x ≤ 4</code></pre><h4 id="DenseAxisArrays-2"><a class="docs-heading-anchor" href="#DenseAxisArrays-2">DenseAxisArrays</a><a class="docs-heading-anchor-permalink" href="#DenseAxisArrays-2" title="Permalink"></a></h4><p>Create an <code>DenseAxisArray</code> of constraints:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, [i = 1:2, j = 2:3], i * x &lt;= j + 1)</code><code class="nohighlight hljs ansi" style="display:block;">2-dimensional DenseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},2,...} with index sets:
+    Dimension 1, Base.OneTo(2)
+    Dimension 2, 2:3
+And data, a 2×2 Matrix{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ x ≤ 3    x ≤ 4
+ 2 x ≤ 3  2 x ≤ 4</code></pre><h4 id="SparseAxisArrays-2"><a class="docs-heading-anchor" href="#SparseAxisArrays-2">SparseAxisArrays</a><a class="docs-heading-anchor-permalink" href="#SparseAxisArrays-2" title="Permalink"></a></h4><p>Create an <code>SparseAxisArray</code> of constraints:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, [i = 1:2, j = 1:2; i != j], i * x &lt;= j + 1)</code><code class="nohighlight hljs ansi" style="display:block;">SparseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}, 2, Tuple{Int64, Int64}} with 2 entries:
+  [1, 2]  =  x ≤ 3
+  [2, 1]  =  2 x ≤ 2</code></pre><h3 id="Constraints-in-a-loop"><a class="docs-heading-anchor" href="#Constraints-in-a-loop">Constraints in a loop</a><a id="Constraints-in-a-loop-1"></a><a class="docs-heading-anchor-permalink" href="#Constraints-in-a-loop" title="Permalink"></a></h3><p>We can add constraints using regular Julia loops:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; for i in 1:3
+           @constraint(model, 6x + 4y &gt;= 5i)
+       end</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><p>or use for each loops inside the <code>@constraint</code> macro:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, [i in 1:3], 6x + 4y &gt;= 5i)</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:
+ 6 x + 4 y ≥ 5
+ 6 x + 4 y ≥ 10
+ 6 x + 4 y ≥ 15</code></pre><p>We can also create constraints such as <span>$\sum _{i = 1}^{10} z_i \leq 1$</span>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, sum(z[i] for i in 1:10) &lt;= 1)</code><code class="nohighlight hljs ansi" style="display:block;">z[1] + z[2] + z[3] + z[4] + z[5] + z[6] + z[7] + z[8] + z[9] + z[10] ≤ 1</code></pre><h2 id="Objective-functions"><a class="docs-heading-anchor" href="#Objective-functions">Objective functions</a><a id="Objective-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Objective-functions" title="Permalink"></a></h2><p>Set an objective function with <a href="../../../api/JuMP/#JuMP.@objective"><code>@objective</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer)</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, y &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, 2x + y)</code><code class="nohighlight hljs ansi" style="display:block;">2 x + y</code></pre><p>Create a maximization objective using <code>Max</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Max, 2x + y)</code><code class="nohighlight hljs ansi" style="display:block;">2 x + y</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Calling <a href="../../../api/JuMP/#JuMP.@objective"><code>@objective</code></a> multiple times will over-write the previous objective. This can be useful when you want to solve the same problem with different objectives.</p></div></div><h2 id="Vectorized-syntax"><a class="docs-heading-anchor" href="#Vectorized-syntax">Vectorized syntax</a><a id="Vectorized-syntax-1"></a><a class="docs-heading-anchor-permalink" href="#Vectorized-syntax" title="Permalink"></a></h2><p>We can also add constraints and an objective to JuMP using vectorized linear algebra. We&#39;ll illustrate this by solving an LP in standard form that is,</p><p class="math-container">\[\begin{aligned}
+&amp; \min &amp; c^T x \\
+&amp; \;\;\text{s.t.} &amp; A x = b \\
+&amp; &amp; x \ge 0
+\end{aligned}\]</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; vector_model = Model(HiGHS.Optimizer)</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; A = [1 1 9 5; 3 5 0 8; 2 0 6 13]</code><code class="nohighlight hljs ansi" style="display:block;">3×4 Matrix{Int64}:
+ 1  1  9   5
+ 3  5  0   8
+ 2  0  6  13</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; b = [7, 3, 5]</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{Int64}:
+ 7
+ 3
+ 5</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; c = [1, 3, 5, 2]</code><code class="nohighlight hljs ansi" style="display:block;">4-element Vector{Int64}:
+ 1
+ 3
+ 5
+ 2</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(vector_model, x[1:4] &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">4-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+ x[4]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(vector_model, A * x .== b)</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}, ScalarShape}}:
+ x[1] + x[2] + 9 x[3] + 5 x[4] = 7
+ 3 x[1] + 5 x[2] + 8 x[4] = 3
+ 2 x[1] + 6 x[3] + 13 x[4] = 5</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(vector_model, Min, c&#39; * x)</code><code class="nohighlight hljs ansi" style="display:block;">x[1] + 3 x[2] + 5 x[3] + 2 x[4]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; optimize!(vector_model)</code><code class="nohighlight hljs ansi" style="display:block;">Running HiGHS 1.5.3 [date: 1970-01-01, git hash: 45a127b78]
+Copyright (c) 2023 HiGHS under MIT licence terms
+Presolving model
+3 rows, 4 cols, 10 nonzeros
+3 rows, 4 cols, 10 nonzeros
+Presolve : Reductions: rows 3(-0); columns 4(-0); elements 10(-0) - Not reduced
+Problem not reduced by presolve: solving the LP
+Using EKK dual simplex solver - serial
+  Iteration        Objective     Infeasibilities num(sum)
+          0     0.0000000000e+00 Pr: 3(13.5) 0s
+          4     4.9230769231e+00 Pr: 0(0) 0s
+Model   status      : Optimal
+Simplex   iterations: 4
+Objective value     :  4.9230769231e+00
+HiGHS run time      :          0.00</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; objective_value(vector_model)</code><code class="nohighlight hljs ansi" style="display:block;">4.923076923076922</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../getting_started_with_julia/">« Getting started with Julia</a><a class="docs-footer-nextpage" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting.jl b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting.jl
new file mode 100644
index 00000000000..c977d57c5b9
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting.jl
@@ -0,0 +1,389 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Getting started with data and plotting
+
+# In this tutorial we will learn how to read tabular data into Julia, and some
+# of the basics of plotting.
+
+# If you're new to Julia, start by reading [Getting started with Julia](@ref)
+# and [Getting started with JuMP](@ref) first.
+
+# !!! note
+#     There are multiple ways to read the same kind of data into Julia. This
+#     tutorial focuses on DataFrames.jl because it provides the ecosystem to
+#     work with most of the required file types in a straightforward manner.
+
+# Before we get started, we need this constant to point to where the data files
+# are.
+
+import JuMP
+const DATA_DIR = joinpath(
+    dirname(pathof(JuMP)),
+    joinpath("..", "docs", "src", "tutorials", "getting_started", "data"),
+);
+
+# ## Where to get help
+
+# Read the documentation
+# * Plots.jl: [http://docs.juliaplots.org/latest/](http://docs.juliaplots.org/latest/)
+# * CSV.jl: [http://csv.juliadata.org/stable](http://csv.juliadata.org/stable)
+# * DataFrames.jl: [https://dataframes.juliadata.org/stable/](https://dataframes.juliadata.org/stable/)
+
+# ## Preliminaries
+
+# To get started, we need to install some packages.
+
+# ### DataFrames.jl
+
+# The `DataFrames` package provides a set of tools for working with tabular
+# data. It is available through the Julia package manager.
+# ```julia
+# using Pkg
+# Pkg.add("DataFrames")
+# ```
+
+import DataFrames
+
+# !!! info "What is a DataFrame?"
+#     A DataFrame is a data structure like a table or spreadsheet. You can use
+#     it for storing and exploring a set of related data values. Think of it as
+#     a smarter array for holding tabular data.
+
+# ### Plots.jl
+
+# The `Plots` package provides a set of tools for plotting. It is available
+# through the Julia package manager.
+# ```julia
+# using Pkg
+# Pkg.add("Plots")
+# ```
+
+import Plots
+
+# ### CSV .jl
+
+# CSV and other delimited text files can be read by the CSV.jl package.
+
+# ```julia
+# Pkg.add("CSV")
+# ```
+
+import CSV
+
+# ## DataFrame basics
+
+# To read a CSV file into a DataFrame, we use the `CSV.read` function.
+
+csv_df = CSV.read(joinpath(DATA_DIR, "StarWars.csv"), DataFrames.DataFrame)
+
+# Let's try plotting some of this data
+
+Plots.scatter(
+    csv_df.Weight,
+    csv_df.Height;
+    xlabel = "Weight",
+    ylabel = "Height",
+)
+
+# That doesn't look right. What happened? If you look at the dataframe above, it
+# read `Weight` in as a `String` column because there are "NA" fields. Let's
+# correct that, by telling CSV to consider "NA" as `missing`.
+
+csv_df = CSV.read(
+    joinpath(DATA_DIR, "StarWars.csv"),
+    DataFrames.DataFrame;
+    missingstring = "NA",
+)
+
+# Then let's re-plot our data
+
+Plots.scatter(
+    csv_df.Weight,
+    csv_df.Height;
+    title = "Height vs Weight of StarWars characters",
+    xlabel = "Weight",
+    ylabel = "Height",
+    label = false,
+    ylims = (0, 3),
+)
+
+# That looks better.
+
+# !!! tip
+#     Read the [CSV documentation](https://csv.juliadata.org/stable/) for other
+#     parsing options.
+
+# DataFrames.jl supports manipulation using functions similar to pandas. For
+# example, split the dataframe into groups based on eye-color:
+
+by_eyecolor = DataFrames.groupby(csv_df, :Eyecolor)
+
+# Then recombine into a single dataframe based on a function operating over the
+# split dataframes:
+
+eyecolor_count = DataFrames.combine(by_eyecolor) do df
+    return DataFrames.nrow(df)
+end
+
+# We can rename columns:
+
+DataFrames.rename!(eyecolor_count, :x1 => :count)
+
+# Drop some missing rows:
+
+DataFrames.dropmissing!(eyecolor_count, :Eyecolor)
+
+# Then we can visualize the data:
+
+sort!(eyecolor_count, :count; rev = true)
+Plots.bar(
+    eyecolor_count.Eyecolor,
+    eyecolor_count.count;
+    xlabel = "Eye color",
+    ylabel = "Number of characters",
+    label = false,
+)
+
+# ### Other Delimited Files
+
+# We can also use the `CSV.jl` package to read any other delimited text file
+# format.
+
+# By default, CSV.File will try to detect a file's delimiter from the first 10
+# lines of the file.
+
+# Candidate delimiters include `','`, `'\t'`, `' '`, `'|'`, `';'`, and `':'`. If
+# it can't auto-detect the delimiter, it will assume `','`.
+
+# Let's take the example of space separated data.
+
+ss_df = CSV.read(joinpath(DATA_DIR, "Cereal.txt"), DataFrames.DataFrame)
+
+# We can also specify the delimiter as follows:
+
+delim_df = CSV.read(
+    joinpath(DATA_DIR, "Soccer.txt"),
+    DataFrames.DataFrame;
+    delim = "::",
+)
+
+# ## Working with DataFrames
+
+# Now that we have read the required data into a DataFrame, let us look at some
+# basic operations we can perform on it.
+
+# ### Querying Basic Information
+
+# The `size` function gets us the dimensions of the DataFrame:
+
+DataFrames.size(ss_df)
+
+# We can also use the `nrow` and `ncol` functions to get the number of rows and
+# columns respectively:
+
+DataFrames.nrow(ss_df), DataFrames.ncol(ss_df)
+
+# The `describe` function gives basic summary statistics of data in a DataFrame:
+
+DataFrames.describe(ss_df)
+
+# Names of every column can be obtained by the `names` function:
+
+DataFrames.names(ss_df)
+
+# Corresponding data types are obtained using the broadcasted `eltype` function:
+
+eltype.(ss_df)
+
+# ### Accessing the Data
+
+# Similar to regular arrays, we use numerical indexing to access elements of a
+# DataFrame:
+
+csv_df[1, 1]
+
+# The following are different ways to access a column:
+
+csv_df[!, 1]
+
+#-
+
+csv_df[!, :Name]
+
+#-
+
+csv_df.Name
+
+#-
+
+csv_df[:, 1] # Note that this creates a copy.
+
+# The following are different ways to access a row:
+
+csv_df[1:1, :]
+
+#-
+
+csv_df[1, :] # This produces a DataFrameRow.
+
+# We can change the values just as we normally assign values.
+
+# Assign a range to scalar:
+
+csv_df[1:3, :Height] .= 1.83
+
+# Assign a vector:
+
+csv_df[4:6, :Height] = [1.8, 1.6, 1.8]
+
+#-
+
+csv_df
+
+# !!! tip
+#     There are a lot more things which can be done with a DataFrame. Read the
+#     [docs](https://juliadata.github.io/DataFrames.jl/stable/) for more
+#     information.
+
+# For information on dplyr-type syntax:
+# * Read the [DataFrames.jl documentation](https://dataframes.juliadata.org/stable/man/querying_frameworks/)
+# * Check out [DataFramesMeta.jl](https://github.com/JuliaData/DataFramesMeta.jl)
+
+# ## Example: the passport problem
+
+# Let's now apply what we have learned to solve a real problem.
+
+# ### Data manipulation
+
+# The [Passport Index Dataset](https://github.com/ilyankou/passport-index-dataset)
+# lists travel visa requirements for 199 countries, in `.csv` format. Our task
+# is to find the minimum number of passports required to visit all countries.
+
+passport_data = CSV.read(
+    joinpath(DATA_DIR, "passport-index-matrix.csv"),
+    DataFrames.DataFrame,
+);
+
+# In this dataset, the first column represents a passport (=from) and each
+# remaining column represents a foreign country (=to).
+
+# The values in each cell are as follows:
+# * 3 = visa-free travel
+# * 2 = eTA is required
+# * 1 = visa can be obtained on arrival
+# * 0 = visa is required
+# * -1 is for all instances where passport and destination are the same
+
+# Our task is to find out the minimum number of passports needed to visit every
+# country without requiring a visa.
+
+# The values we are interested in are -1 and 3. Let's modify the dataframe so
+# that the -1 and 3 are `1` (true), and all others are `0` (false):
+
+function modifier(x)
+    if x == -1 || x == 3
+        return 1
+    else
+        return 0
+    end
+end
+
+for country in passport_data.Passport
+    passport_data[!, country] = modifier.(passport_data[!, country])
+end
+
+# The values in the cells now represent:
+# * 1 = no visa required for travel
+# * 0 = visa required for travel
+
+# ### JuMP Modeling
+
+# To model the problem as a mixed-integer linear program, we need a binary
+# decision variable $x_c$ for each country $c$. $x_c$ is $1$ if we select
+# passport $c$ and $0$ otherwise. Our objective is to minimize the sum
+# $\sum x_c$ over all countries.
+
+# Since we wish to visit all the countries, for every country, we must own at
+# least one passport that lets us travel to that country visa free. For one
+# destination, this can be mathematically represented as
+# $\sum_{c \in C} a_{c,d} \cdot x_{d} \geq 1$, where $a$ is the `passport_data`
+# dataframe.
+
+# Thus, we can represent this problem using the following model:
+
+# ```math
+# \begin{aligned}
+# \min && \sum_{c \in C} x_c \\
+# \text{s.t.} && \sum_{c \in C} a_{c,d} x_c \geq 1 && \forall d \in C \\
+# && x_c \in \{0,1\} && \forall c \in C.
+# \end{aligned}
+# ```
+
+# We'll now solve the problem using JuMP:
+
+using JuMP
+import HiGHS
+
+# First, create the set of countries:
+
+C = passport_data.Passport
+
+# Then, create the model and initialize the decision variables:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[C], Bin)
+@objective(model, Min, sum(x))
+@constraint(model, [d in C], passport_data[!, d]' * x >= 1)
+model
+
+# Now optimize:
+
+optimize!(model)
+
+# We can use the [`solution_summary`](@ref) function to get an overview of the
+# solution:
+
+solution_summary(model)
+
+# ### Solution
+
+# Let's  have a look at the solution in more detail:
+
+println("Minimum number of passports needed: ", objective_value(model))
+
+#-
+
+println("Optimal passports:")
+for c in C
+    if value(x[c]) > 0.5
+        println(" * ", c)
+    end
+end
+
+# We need some passports, like New Zealand and the United States, which have
+# widespread access to a large number of countries. However, we also need
+# passports like North Korea which only have visa-free access to a very
+# limited number of countries.
+
+# !!! note
+#     We use `value(x[c]) > 0.5` rather than `value(x[c]) == 1` to avoid
+#     excluding solutions like `x[c] = 0.99999` that are "1" to some tolerance.
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/0c9607c5.svg b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/0c9607c5.svg
new file mode 100644
index 00000000000..2003bea014d
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/0c9607c5.svg
@@ -0,0 +1,59 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip850">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip850)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip851">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip850)" d="M174.149 1423.18 L2352.76 1423.18 L2352.76 123.472 L174.149 123.472  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip852">
+    <rect x="174" y="123" width="2180" height="1301"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="423.968,1423.18 423.968,123.472 "/>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="858.184,1423.18 858.184,123.472 "/>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1292.4,1423.18 1292.4,123.472 "/>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1726.62,1423.18 1726.62,123.472 "/>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2160.83,1423.18 2160.83,123.472 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="174.149,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="423.968,1423.18 423.968,1404.28 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="858.184,1423.18 858.184,1404.28 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1292.4,1423.18 1292.4,1404.28 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1726.62,1423.18 1726.62,1404.28 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.83,1423.18 2160.83,1404.28 "/>
+<path clip-path="url(#clip850)" d="M412.811 1466.95 Q416.167 1467.66 418.042 1469.93 Q419.94 1472.2 419.94 1475.53 Q419.94 1480.65 416.422 1483.45 Q412.903 1486.25 406.422 1486.25 Q404.246 1486.25 401.931 1485.81 Q399.639 1485.39 397.186 1484.54 L397.186 1480.02 Q399.13 1481.16 401.445 1481.74 Q403.76 1482.32 406.283 1482.32 Q410.681 1482.32 412.973 1480.58 Q415.287 1478.84 415.287 1475.53 Q415.287 1472.48 413.135 1470.77 Q411.005 1469.03 407.186 1469.03 L403.158 1469.03 L403.158 1465.19 L407.371 1465.19 Q410.82 1465.19 412.649 1463.82 Q414.477 1462.43 414.477 1459.84 Q414.477 1457.18 412.579 1455.77 Q410.704 1454.33 407.186 1454.33 Q405.264 1454.33 403.065 1454.75 Q400.866 1455.16 398.227 1456.04 L398.227 1451.88 Q400.889 1451.14 403.204 1450.77 Q405.542 1450.39 407.602 1450.39 Q412.926 1450.39 416.028 1452.83 Q419.13 1455.23 419.13 1459.35 Q419.13 1462.22 417.486 1464.21 Q415.843 1466.18 412.811 1466.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M438.806 1454.1 Q435.195 1454.1 433.366 1457.66 Q431.56 1461.2 431.56 1468.33 Q431.56 1475.44 433.366 1479.01 Q435.195 1482.55 438.806 1482.55 Q442.44 1482.55 444.246 1479.01 Q446.074 1475.44 446.074 1468.33 Q446.074 1461.2 444.246 1457.66 Q442.44 1454.1 438.806 1454.1 M438.806 1450.39 Q444.616 1450.39 447.671 1455 Q450.75 1459.58 450.75 1468.33 Q450.75 1477.06 447.671 1481.67 Q444.616 1486.25 438.806 1486.25 Q432.996 1486.25 429.917 1481.67 Q426.861 1477.06 426.861 1468.33 Q426.861 1459.58 429.917 1455 Q432.996 1450.39 438.806 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M843.589 1466.44 Q840.441 1466.44 838.589 1468.59 Q836.761 1470.74 836.761 1474.49 Q836.761 1478.22 838.589 1480.39 Q840.441 1482.55 843.589 1482.55 Q846.737 1482.55 848.566 1480.39 Q850.418 1478.22 850.418 1474.49 Q850.418 1470.74 848.566 1468.59 Q846.737 1466.44 843.589 1466.44 M852.872 1451.78 L852.872 1456.04 Q851.112 1455.21 849.307 1454.77 Q847.524 1454.33 845.765 1454.33 Q841.136 1454.33 838.682 1457.45 Q836.251 1460.58 835.904 1466.9 Q837.27 1464.89 839.33 1463.82 Q841.39 1462.73 843.867 1462.73 Q849.075 1462.73 852.085 1465.9 Q855.117 1469.05 855.117 1474.49 Q855.117 1479.82 851.969 1483.03 Q848.821 1486.25 843.589 1486.25 Q837.594 1486.25 834.423 1481.67 Q831.251 1477.06 831.251 1468.33 Q831.251 1460.14 835.14 1455.28 Q839.029 1450.39 845.58 1450.39 Q847.339 1450.39 849.122 1450.74 Q850.927 1451.09 852.872 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M873.172 1454.1 Q869.561 1454.1 867.733 1457.66 Q865.927 1461.2 865.927 1468.33 Q865.927 1475.44 867.733 1479.01 Q869.561 1482.55 873.172 1482.55 Q876.807 1482.55 878.612 1479.01 Q880.441 1475.44 880.441 1468.33 Q880.441 1461.2 878.612 1457.66 Q876.807 1454.1 873.172 1454.1 M873.172 1450.39 Q878.983 1450.39 882.038 1455 Q885.117 1459.58 885.117 1468.33 Q885.117 1477.06 882.038 1481.67 Q878.983 1486.25 873.172 1486.25 Q867.362 1486.25 864.284 1481.67 Q861.228 1477.06 861.228 1468.33 Q861.228 1459.58 864.284 1455 Q867.362 1450.39 873.172 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1267.53 1484.86 L1267.53 1480.6 Q1269.29 1481.44 1271.09 1481.88 Q1272.9 1482.32 1274.63 1482.32 Q1279.26 1482.32 1281.69 1479.21 Q1284.15 1476.09 1284.5 1469.75 Q1283.15 1471.74 1281.09 1472.8 Q1279.03 1473.87 1276.53 1473.87 Q1271.35 1473.87 1268.31 1470.74 Q1265.31 1467.59 1265.31 1462.15 Q1265.31 1456.83 1268.45 1453.61 Q1271.6 1450.39 1276.83 1450.39 Q1282.83 1450.39 1285.98 1455 Q1289.15 1459.58 1289.15 1468.33 Q1289.15 1476.51 1285.26 1481.39 Q1281.39 1486.25 1274.84 1486.25 Q1273.08 1486.25 1271.28 1485.9 Q1269.47 1485.56 1267.53 1484.86 M1276.83 1470.21 Q1279.98 1470.21 1281.81 1468.06 Q1283.66 1465.9 1283.66 1462.15 Q1283.66 1458.43 1281.81 1456.27 Q1279.98 1454.1 1276.83 1454.1 Q1273.68 1454.1 1271.83 1456.27 Q1270 1458.43 1270 1462.15 Q1270 1465.9 1271.83 1468.06 Q1273.68 1470.21 1276.83 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1307.55 1454.1 Q1303.94 1454.1 1302.11 1457.66 Q1300.31 1461.2 1300.31 1468.33 Q1300.31 1475.44 1302.11 1479.01 Q1303.94 1482.55 1307.55 1482.55 Q1311.18 1482.55 1312.99 1479.01 Q1314.82 1475.44 1314.82 1468.33 Q1314.82 1461.2 1312.99 1457.66 Q1311.18 1454.1 1307.55 1454.1 M1307.55 1450.39 Q1313.36 1450.39 1316.42 1455 Q1319.49 1459.58 1319.49 1468.33 Q1319.49 1477.06 1316.42 1481.67 Q1313.36 1486.25 1307.55 1486.25 Q1301.74 1486.25 1298.66 1481.67 Q1295.61 1477.06 1295.61 1468.33 Q1295.61 1459.58 1298.66 1455 Q1301.74 1450.39 1307.55 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1686.22 1481.64 L1693.86 1481.64 L1693.86 1455.28 L1685.55 1456.95 L1685.55 1452.69 L1693.82 1451.02 L1698.49 1451.02 L1698.49 1481.64 L1706.13 1481.64 L1706.13 1485.58 L1686.22 1485.58 L1686.22 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1719.6 1481.64 L1735.92 1481.64 L1735.92 1485.58 L1713.98 1485.58 L1713.98 1481.64 Q1716.64 1478.89 1721.22 1474.26 Q1725.83 1469.61 1727.01 1468.27 Q1729.26 1465.74 1730.13 1464.01 Q1731.04 1462.25 1731.04 1460.56 Q1731.04 1457.8 1729.09 1456.07 Q1727.17 1454.33 1724.07 1454.33 Q1721.87 1454.33 1719.42 1455.09 Q1716.99 1455.86 1714.21 1457.41 L1714.21 1452.69 Q1717.03 1451.55 1719.49 1450.97 Q1721.94 1450.39 1723.98 1450.39 Q1729.35 1450.39 1732.54 1453.08 Q1735.74 1455.77 1735.74 1460.26 Q1735.74 1462.39 1734.93 1464.31 Q1734.14 1466.2 1732.03 1468.8 Q1731.45 1469.47 1728.35 1472.69 Q1725.25 1475.88 1719.6 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1755.74 1454.1 Q1752.13 1454.1 1750.3 1457.66 Q1748.49 1461.2 1748.49 1468.33 Q1748.49 1475.44 1750.3 1479.01 Q1752.13 1482.55 1755.74 1482.55 Q1759.37 1482.55 1761.18 1479.01 Q1763 1475.44 1763 1468.33 Q1763 1461.2 1761.18 1457.66 Q1759.37 1454.1 1755.74 1454.1 M1755.74 1450.39 Q1761.55 1450.39 1764.6 1455 Q1767.68 1459.58 1767.68 1468.33 Q1767.68 1477.06 1764.6 1481.67 Q1761.55 1486.25 1755.74 1486.25 Q1749.93 1486.25 1746.85 1481.67 Q1743.79 1477.06 1743.79 1468.33 Q1743.79 1459.58 1746.85 1455 Q1749.93 1450.39 1755.74 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M2120.44 1481.64 L2128.08 1481.64 L2128.08 1455.28 L2119.77 1456.95 L2119.77 1452.69 L2128.03 1451.02 L2132.71 1451.02 L2132.71 1481.64 L2140.35 1481.64 L2140.35 1485.58 L2120.44 1485.58 L2120.44 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M2149.84 1451.02 L2168.19 1451.02 L2168.19 1454.96 L2154.12 1454.96 L2154.12 1463.43 Q2155.14 1463.08 2156.16 1462.92 Q2157.17 1462.73 2158.19 1462.73 Q2163.98 1462.73 2167.36 1465.9 Q2170.74 1469.08 2170.74 1474.49 Q2170.74 1480.07 2167.27 1483.17 Q2163.8 1486.25 2157.48 1486.25 Q2155.3 1486.25 2153.03 1485.88 Q2150.79 1485.51 2148.38 1484.77 L2148.38 1480.07 Q2150.46 1481.2 2152.68 1481.76 Q2154.91 1482.32 2157.38 1482.32 Q2161.39 1482.32 2163.73 1480.21 Q2166.06 1478.1 2166.06 1474.49 Q2166.06 1470.88 2163.73 1468.77 Q2161.39 1466.67 2157.38 1466.67 Q2155.51 1466.67 2153.63 1467.08 Q2151.78 1467.5 2149.84 1468.38 L2149.84 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M2189.95 1454.1 Q2186.34 1454.1 2184.51 1457.66 Q2182.71 1461.2 2182.71 1468.33 Q2182.71 1475.44 2184.51 1479.01 Q2186.34 1482.55 2189.95 1482.55 Q2193.59 1482.55 2195.39 1479.01 Q2197.22 1475.44 2197.22 1468.33 Q2197.22 1461.2 2195.39 1457.66 Q2193.59 1454.1 2189.95 1454.1 M2189.95 1450.39 Q2195.76 1450.39 2198.82 1455 Q2201.9 1459.58 2201.9 1468.33 Q2201.9 1477.06 2198.82 1481.67 Q2195.76 1486.25 2189.95 1486.25 Q2184.14 1486.25 2181.06 1481.67 Q2178.01 1477.06 2178.01 1468.33 Q2178.01 1459.58 2181.06 1455 Q2184.14 1450.39 2189.95 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1151.77 1520.52 L1158.26 1520.52 L1168.25 1560.69 L1178.22 1520.52 L1185.44 1520.52 L1195.43 1560.69 L1205.4 1520.52 L1211.92 1520.52 L1199.99 1568.04 L1191.9 1568.04 L1181.88 1526.79 L1171.75 1568.04 L1163.67 1568.04 L1151.77 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1246.87 1548.76 L1246.87 1551.62 L1219.94 1551.62 Q1220.32 1557.67 1223.57 1560.85 Q1226.85 1564 1232.67 1564 Q1236.05 1564 1239.2 1563.17 Q1242.38 1562.35 1245.5 1560.69 L1245.5 1566.23 Q1242.35 1567.57 1239.04 1568.27 Q1235.73 1568.97 1232.32 1568.97 Q1223.79 1568.97 1218.8 1564 Q1213.83 1559.04 1213.83 1550.57 Q1213.83 1541.82 1218.54 1536.69 Q1223.28 1531.54 1231.31 1531.54 Q1238.5 1531.54 1242.67 1536.18 Q1246.87 1540.8 1246.87 1548.76 M1241.01 1547.04 Q1240.95 1542.23 1238.31 1539.37 Q1235.7 1536.5 1231.37 1536.5 Q1226.47 1536.5 1223.51 1539.27 Q1220.58 1542.04 1220.13 1547.07 L1241.01 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1256.48 1532.4 L1262.34 1532.4 L1262.34 1568.04 L1256.48 1568.04 L1256.48 1532.4 M1256.48 1518.52 L1262.34 1518.52 L1262.34 1525.93 L1256.48 1525.93 L1256.48 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1298.05 1549.81 Q1298.05 1543.44 1295.41 1539.94 Q1292.8 1536.44 1288.06 1536.44 Q1283.35 1536.44 1280.7 1539.94 Q1278.09 1543.44 1278.09 1549.81 Q1278.09 1556.14 1280.7 1559.64 Q1283.35 1563.14 1288.06 1563.14 Q1292.8 1563.14 1295.41 1559.64 Q1298.05 1556.14 1298.05 1549.81 M1303.91 1563.62 Q1303.91 1572.72 1299.86 1577.15 Q1295.82 1581.6 1287.48 1581.6 Q1284.4 1581.6 1281.66 1581.13 Q1278.92 1580.68 1276.34 1579.72 L1276.34 1574.03 Q1278.92 1575.43 1281.44 1576.1 Q1283.95 1576.76 1286.56 1576.76 Q1292.32 1576.76 1295.19 1573.74 Q1298.05 1570.75 1298.05 1564.67 L1298.05 1561.77 Q1296.24 1564.92 1293.4 1566.48 Q1290.57 1568.04 1286.62 1568.04 Q1280.07 1568.04 1276.06 1563.05 Q1272.05 1558.05 1272.05 1549.81 Q1272.05 1541.53 1276.06 1536.53 Q1280.07 1531.54 1286.62 1531.54 Q1290.57 1531.54 1293.4 1533.1 Q1296.24 1534.66 1298.05 1537.81 L1298.05 1532.4 L1303.91 1532.4 L1303.91 1563.62 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1345.6 1546.53 L1345.6 1568.04 L1339.75 1568.04 L1339.75 1546.72 Q1339.75 1541.66 1337.77 1539.14 Q1335.8 1536.63 1331.85 1536.63 Q1327.11 1536.63 1324.37 1539.65 Q1321.63 1542.68 1321.63 1547.9 L1321.63 1568.04 L1315.75 1568.04 L1315.75 1518.52 L1321.63 1518.52 L1321.63 1537.93 Q1323.74 1534.72 1326.57 1533.13 Q1329.43 1531.54 1333.16 1531.54 Q1339.3 1531.54 1342.45 1535.36 Q1345.6 1539.14 1345.6 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1363.08 1522.27 L1363.08 1532.4 L1375.14 1532.4 L1375.14 1536.95 L1363.08 1536.95 L1363.08 1556.3 Q1363.08 1560.66 1364.25 1561.9 Q1365.46 1563.14 1369.12 1563.14 L1375.14 1563.14 L1375.14 1568.04 L1369.12 1568.04 Q1362.34 1568.04 1359.77 1565.53 Q1357.19 1562.98 1357.19 1556.3 L1357.19 1536.95 L1352.89 1536.95 L1352.89 1532.4 L1357.19 1532.4 L1357.19 1522.27 L1363.08 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="174.149,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="174.149,989.944 2352.76,989.944 "/>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="174.149,556.708 2352.76,556.708 "/>
+<polyline clip-path="url(#clip852)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="174.149,123.472 2352.76,123.472 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="174.149,1423.18 174.149,123.472 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="174.149,1423.18 193.047,1423.18 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="174.149,989.944 193.047,989.944 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="174.149,556.708 193.047,556.708 "/>
+<polyline clip-path="url(#clip850)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="174.149,123.472 193.047,123.472 "/>
+<path clip-path="url(#clip850)" d="M126.205 1408.98 Q122.593 1408.98 120.765 1412.54 Q118.959 1416.08 118.959 1423.21 Q118.959 1430.32 120.765 1433.89 Q122.593 1437.43 126.205 1437.43 Q129.839 1437.43 131.644 1433.89 Q133.473 1430.32 133.473 1423.21 Q133.473 1416.08 131.644 1412.54 Q129.839 1408.98 126.205 1408.98 M126.205 1405.27 Q132.015 1405.27 135.07 1409.88 Q138.149 1414.46 138.149 1423.21 Q138.149 1431.94 135.07 1436.55 Q132.015 1441.13 126.205 1441.13 Q120.394 1441.13 117.316 1436.55 Q114.26 1431.94 114.26 1423.21 Q114.26 1414.46 117.316 1409.88 Q120.394 1405.27 126.205 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M118.242 1003.29 L125.88 1003.29 L125.88 976.923 L117.57 978.59 L117.57 974.331 L125.834 972.664 L130.51 972.664 L130.51 1003.29 L138.149 1003.29 L138.149 1007.22 L118.242 1007.22 L118.242 1003.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M121.83 570.053 L138.149 570.053 L138.149 573.988 L116.205 573.988 L116.205 570.053 Q118.867 567.298 123.45 562.669 Q128.056 558.016 129.237 556.673 Q131.482 554.15 132.362 552.414 Q133.265 550.655 133.265 548.965 Q133.265 546.21 131.32 544.474 Q129.399 542.738 126.297 542.738 Q124.098 542.738 121.644 543.502 Q119.214 544.266 116.436 545.817 L116.436 541.095 Q119.26 539.96 121.714 539.382 Q124.168 538.803 126.205 538.803 Q131.575 538.803 134.769 541.488 Q137.964 544.173 137.964 548.664 Q137.964 550.794 137.154 552.715 Q136.367 554.613 134.26 557.206 Q133.681 557.877 130.58 561.095 Q127.478 564.289 121.83 570.053 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M131.019 122.118 Q134.376 122.836 136.251 125.104 Q138.149 127.373 138.149 130.706 Q138.149 135.822 134.63 138.623 Q131.112 141.423 124.63 141.423 Q122.455 141.423 120.14 140.984 Q117.848 140.567 115.394 139.71 L115.394 135.197 Q117.339 136.331 119.654 136.91 Q121.968 137.488 124.492 137.488 Q128.89 137.488 131.181 135.752 Q133.496 134.016 133.496 130.706 Q133.496 127.65 131.343 125.937 Q129.214 124.201 125.394 124.201 L121.367 124.201 L121.367 120.359 L125.58 120.359 Q129.029 120.359 130.857 118.993 Q132.686 117.604 132.686 115.012 Q132.686 112.35 130.788 110.937 Q128.913 109.502 125.394 109.502 Q123.473 109.502 121.274 109.919 Q119.075 110.336 116.436 111.215 L116.436 107.049 Q119.098 106.308 121.413 105.937 Q123.751 105.567 125.811 105.567 Q131.135 105.567 134.237 107.998 Q137.339 110.405 137.339 114.525 Q137.339 117.396 135.695 119.386 Q134.052 121.354 131.019 122.118 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M16.4842 877.087 L16.4842 870.658 L35.9632 870.658 L35.9632 847.295 L16.4842 847.295 L16.4842 840.866 L64.0042 840.866 L64.0042 847.295 L41.3741 847.295 L41.3741 870.658 L64.0042 870.658 L64.0042 877.087 L16.4842 877.087 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M44.7161 797.834 L47.5806 797.834 L47.5806 824.761 Q53.6281 824.379 56.8109 821.132 Q59.9619 817.854 59.9619 812.029 Q59.9619 808.656 59.1344 805.505 Q58.3069 802.322 56.6518 799.203 L62.1899 799.203 Q63.5267 802.354 64.227 805.664 Q64.9272 808.974 64.9272 812.38 Q64.9272 820.91 59.9619 825.907 Q54.9967 830.872 46.5303 830.872 Q37.7774 830.872 32.6531 826.161 Q27.4968 821.419 27.4968 813.398 Q27.4968 806.205 32.1438 802.035 Q36.7589 797.834 44.7161 797.834 M42.9973 803.69 Q38.1912 803.754 35.3266 806.396 Q32.4621 809.006 32.4621 813.334 Q32.4621 818.236 35.2312 821.196 Q38.0002 824.124 43.0292 824.57 L42.9973 803.69 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M28.3562 788.222 L28.3562 782.365 L64.0042 782.365 L64.0042 788.222 L28.3562 788.222 M14.479 788.222 L14.479 782.365 L21.895 782.365 L21.895 788.222 L14.479 788.222 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M45.7664 746.654 Q39.4007 746.654 35.8996 749.295 Q32.3984 751.905 32.3984 756.648 Q32.3984 761.358 35.8996 764 Q39.4007 766.61 45.7664 766.61 Q52.1003 766.61 55.6014 764 Q59.1026 761.358 59.1026 756.648 Q59.1026 751.905 55.6014 749.295 Q52.1003 746.654 45.7664 746.654 M59.58 740.797 Q68.683 740.797 73.1071 744.839 Q77.5631 748.882 77.5631 757.221 Q77.5631 760.308 77.0857 763.045 Q76.6401 765.783 75.6852 768.361 L69.9879 768.361 Q71.3884 765.783 72.0568 763.268 Q72.7252 760.754 72.7252 758.144 Q72.7252 752.383 69.7015 749.518 Q66.7096 746.654 60.6303 746.654 L57.7339 746.654 Q60.885 748.468 62.4446 751.301 Q64.0042 754.133 64.0042 758.08 Q64.0042 764.637 59.0071 768.647 Q54.01 772.658 45.7664 772.658 Q37.491 772.658 32.4939 768.647 Q27.4968 764.637 27.4968 758.08 Q27.4968 754.133 29.0564 751.301 Q30.616 748.468 33.7671 746.654 L28.3562 746.654 L28.3562 740.797 L59.58 740.797 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M42.4881 699.102 L64.0042 699.102 L64.0042 704.958 L42.679 704.958 Q37.6183 704.958 35.1038 706.932 Q32.5894 708.905 32.5894 712.852 Q32.5894 717.594 35.6131 720.331 Q38.6368 723.069 43.8567 723.069 L64.0042 723.069 L64.0042 728.957 L14.479 728.957 L14.479 723.069 L33.8944 723.069 Q30.6797 720.968 29.0883 718.135 Q27.4968 715.271 27.4968 711.547 Q27.4968 705.404 31.3163 702.253 Q35.1038 699.102 42.4881 699.102 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M18.2347 681.628 L28.3562 681.628 L28.3562 669.565 L32.9077 669.565 L32.9077 681.628 L52.2594 681.628 Q56.6199 681.628 57.8613 680.45 Q59.1026 679.241 59.1026 675.581 L59.1026 669.565 L64.0042 669.565 L64.0042 675.581 Q64.0042 682.36 61.4897 684.938 Q58.9434 687.516 52.2594 687.516 L32.9077 687.516 L32.9077 691.813 L28.3562 691.813 L28.3562 687.516 L18.2347 687.516 L18.2347 681.628 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M428.075 12.096 L436.258 12.096 L436.258 36.8875 L465.991 36.8875 L465.991 12.096 L474.174 12.096 L474.174 72.576 L465.991 72.576 L465.991 43.7741 L436.258 43.7741 L436.258 72.576 L428.075 72.576 L428.075 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M528.942 48.0275 L528.942 51.6733 L494.672 51.6733 Q495.158 59.3701 499.29 63.421 Q503.462 67.4314 510.875 67.4314 Q515.169 67.4314 519.18 66.3781 Q523.231 65.3249 527.201 63.2184 L527.201 70.267 Q523.19 71.9684 518.977 72.8596 Q514.764 73.7508 510.43 73.7508 Q499.573 73.7508 493.213 67.4314 Q486.894 61.1119 486.894 50.3365 Q486.894 39.1965 492.889 32.6746 Q498.925 26.1121 509.134 26.1121 Q518.289 26.1121 523.595 32.0264 Q528.942 37.9003 528.942 48.0275 M521.489 45.84 Q521.408 39.7232 518.046 36.0774 Q514.724 32.4315 509.215 32.4315 Q502.976 32.4315 499.209 35.9558 Q495.482 39.4801 494.915 45.8805 L521.489 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M541.176 27.2059 L548.63 27.2059 L548.63 72.576 L541.176 72.576 L541.176 27.2059 M541.176 9.54393 L548.63 9.54393 L548.63 18.9825 L541.176 18.9825 L541.176 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M594.081 49.3643 Q594.081 41.2625 590.719 36.8065 Q587.397 32.3505 581.361 32.3505 Q575.366 32.3505 572.004 36.8065 Q568.682 41.2625 568.682 49.3643 Q568.682 57.4256 572.004 61.8816 Q575.366 66.3376 581.361 66.3376 Q587.397 66.3376 590.719 61.8816 Q594.081 57.4256 594.081 49.3643 M601.535 66.9452 Q601.535 78.5308 596.39 84.1616 Q591.245 89.8329 580.632 89.8329 Q576.703 89.8329 573.219 89.2252 Q569.735 88.6581 566.454 87.4428 L566.454 80.1917 Q569.735 81.9741 572.935 82.8248 Q576.135 83.6755 579.457 83.6755 Q586.789 83.6755 590.435 79.8271 Q594.081 76.0193 594.081 68.282 L594.081 64.5957 Q591.772 68.6061 588.167 70.5911 Q584.561 72.576 579.538 72.576 Q571.193 72.576 566.089 66.2161 Q560.985 59.8562 560.985 49.3643 Q560.985 38.832 566.089 32.472 Q571.193 26.1121 579.538 26.1121 Q584.561 26.1121 588.167 28.0971 Q591.772 30.082 594.081 34.0924 L594.081 27.2059 L601.535 27.2059 L601.535 66.9452 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M654.602 45.1919 L654.602 72.576 L647.148 72.576 L647.148 45.4349 Q647.148 38.994 644.636 35.7938 Q642.125 32.5936 637.102 32.5936 Q631.066 32.5936 627.582 36.4419 Q624.098 40.2903 624.098 46.9338 L624.098 72.576 L616.604 72.576 L616.604 9.54393 L624.098 9.54393 L624.098 34.2544 Q626.772 30.163 630.377 28.1376 Q634.023 26.1121 638.762 26.1121 Q646.581 26.1121 650.591 30.9732 Q654.602 35.7938 654.602 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M676.841 14.324 L676.841 27.2059 L692.194 27.2059 L692.194 32.9987 L676.841 32.9987 L676.841 57.6282 Q676.841 63.1779 678.34 64.7578 Q679.879 66.3376 684.538 66.3376 L692.194 66.3376 L692.194 72.576 L684.538 72.576 Q675.909 72.576 672.628 69.3758 Q669.347 66.1351 669.347 57.6282 L669.347 32.9987 L663.878 32.9987 L663.878 27.2059 L669.347 27.2059 L669.347 14.324 L676.841 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M723.021 27.2059 L730.921 27.2059 L745.099 65.2844 L759.277 27.2059 L767.176 27.2059 L750.162 72.576 L740.035 72.576 L723.021 27.2059 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M806.389 28.5427 L806.389 35.5912 Q803.229 33.9709 799.826 33.1607 Q796.424 32.3505 792.778 32.3505 Q787.228 32.3505 784.433 34.0519 Q781.678 35.7533 781.678 39.156 Q781.678 41.7486 783.663 43.2475 Q785.648 44.7058 791.644 46.0426 L794.196 46.6097 Q802.135 48.3111 805.457 51.4303 Q808.819 54.509 808.819 60.0587 Q808.819 66.3781 803.796 70.0644 Q798.814 73.7508 790.064 73.7508 Q786.418 73.7508 782.448 73.0216 Q778.519 72.3329 774.144 70.9151 L774.144 63.2184 Q778.276 65.3654 782.286 66.4591 Q786.296 67.5124 790.226 67.5124 Q795.492 67.5124 798.328 65.73 Q801.163 63.9071 801.163 60.6258 Q801.163 57.5877 799.097 55.9673 Q797.072 54.3469 790.145 52.8481 L787.552 52.2405 Q780.625 50.7821 777.546 47.7845 Q774.468 44.7463 774.468 39.4801 Q774.468 33.0797 779.005 29.5959 Q783.542 26.1121 791.887 26.1121 Q796.019 26.1121 799.664 26.7198 Q803.31 27.3274 806.389 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M841.996 12.096 L850.26 12.096 L862.98 63.2184 L875.659 12.096 L884.855 12.096 L897.575 63.2184 L910.254 12.096 L918.558 12.096 L903.368 72.576 L893.078 72.576 L880.318 20.0763 L867.436 72.576 L857.147 72.576 L841.996 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M963.037 48.0275 L963.037 51.6733 L928.767 51.6733 Q929.253 59.3701 933.385 63.421 Q937.557 67.4314 944.97 67.4314 Q949.264 67.4314 953.275 66.3781 Q957.326 65.3249 961.295 63.2184 L961.295 70.267 Q957.285 71.9684 953.072 72.8596 Q948.859 73.7508 944.525 73.7508 Q933.668 73.7508 927.308 67.4314 Q920.989 61.1119 920.989 50.3365 Q920.989 39.1965 926.984 32.6746 Q933.02 26.1121 943.228 26.1121 Q952.383 26.1121 957.69 32.0264 Q963.037 37.9003 963.037 48.0275 M955.584 45.84 Q955.503 39.7232 952.14 36.0774 Q948.819 32.4315 943.309 32.4315 Q937.071 32.4315 933.304 35.9558 Q929.577 39.4801 929.01 45.8805 L955.584 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M975.271 27.2059 L982.725 27.2059 L982.725 72.576 L975.271 72.576 L975.271 27.2059 M975.271 9.54393 L982.725 9.54393 L982.725 18.9825 L975.271 18.9825 L975.271 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1028.18 49.3643 Q1028.18 41.2625 1024.81 36.8065 Q1021.49 32.3505 1015.46 32.3505 Q1009.46 32.3505 1006.1 36.8065 Q1002.78 41.2625 1002.78 49.3643 Q1002.78 57.4256 1006.1 61.8816 Q1009.46 66.3376 1015.46 66.3376 Q1021.49 66.3376 1024.81 61.8816 Q1028.18 57.4256 1028.18 49.3643 M1035.63 66.9452 Q1035.63 78.5308 1030.48 84.1616 Q1025.34 89.8329 1014.73 89.8329 Q1010.8 89.8329 1007.31 89.2252 Q1003.83 88.6581 1000.55 87.4428 L1000.55 80.1917 Q1003.83 81.9741 1007.03 82.8248 Q1010.23 83.6755 1013.55 83.6755 Q1020.88 83.6755 1024.53 79.8271 Q1028.18 76.0193 1028.18 68.282 L1028.18 64.5957 Q1025.87 68.6061 1022.26 70.5911 Q1018.66 72.576 1013.63 72.576 Q1005.29 72.576 1000.18 66.2161 Q995.08 59.8562 995.08 49.3643 Q995.08 38.832 1000.18 32.472 Q1005.29 26.1121 1013.63 26.1121 Q1018.66 26.1121 1022.26 28.0971 Q1025.87 30.082 1028.18 34.0924 L1028.18 27.2059 L1035.63 27.2059 L1035.63 66.9452 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1088.7 45.1919 L1088.7 72.576 L1081.24 72.576 L1081.24 45.4349 Q1081.24 38.994 1078.73 35.7938 Q1076.22 32.5936 1071.2 32.5936 Q1065.16 32.5936 1061.68 36.4419 Q1058.19 40.2903 1058.19 46.9338 L1058.19 72.576 L1050.7 72.576 L1050.7 9.54393 L1058.19 9.54393 L1058.19 34.2544 Q1060.87 30.163 1064.47 28.1376 Q1068.12 26.1121 1072.86 26.1121 Q1080.68 26.1121 1084.69 30.9732 Q1088.7 35.7938 1088.7 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1110.94 14.324 L1110.94 27.2059 L1126.29 27.2059 L1126.29 32.9987 L1110.94 32.9987 L1110.94 57.6282 Q1110.94 63.1779 1112.43 64.7578 Q1113.97 66.3376 1118.63 66.3376 L1126.29 66.3376 L1126.29 72.576 L1118.63 72.576 Q1110 72.576 1106.72 69.3758 Q1103.44 66.1351 1103.44 57.6282 L1103.44 32.9987 L1097.97 32.9987 L1097.97 27.2059 L1103.44 27.2059 L1103.44 14.324 L1110.94 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1180.04 32.4315 Q1174.05 32.4315 1170.57 37.1306 Q1167.08 41.7891 1167.08 49.9314 Q1167.08 58.0738 1170.52 62.7728 Q1174.01 67.4314 1180.04 67.4314 Q1186 67.4314 1189.48 62.7323 Q1192.97 58.0333 1192.97 49.9314 Q1192.97 41.8701 1189.48 37.1711 Q1186 32.4315 1180.04 32.4315 M1180.04 26.1121 Q1189.77 26.1121 1195.32 32.4315 Q1200.87 38.7509 1200.87 49.9314 Q1200.87 61.0714 1195.32 67.4314 Q1189.77 73.7508 1180.04 73.7508 Q1170.28 73.7508 1164.73 67.4314 Q1159.22 61.0714 1159.22 49.9314 Q1159.22 38.7509 1164.73 32.4315 Q1170.28 26.1121 1180.04 26.1121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1236.19 9.54393 L1236.19 15.7418 L1229.06 15.7418 Q1225.05 15.7418 1223.47 17.3622 Q1221.93 18.9825 1221.93 23.1955 L1221.93 27.2059 L1234.2 27.2059 L1234.2 32.9987 L1221.93 32.9987 L1221.93 72.576 L1214.44 72.576 L1214.44 32.9987 L1207.31 32.9987 L1207.31 27.2059 L1214.44 27.2059 L1214.44 24.0462 Q1214.44 16.471 1217.96 13.0277 Q1221.49 9.54393 1229.14 9.54393 L1236.19 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1305.38 14.0809 L1305.38 22.0612 Q1300.72 19.8332 1296.59 18.7395 Q1292.46 17.6457 1288.61 17.6457 Q1281.92 17.6457 1278.28 20.2383 Q1274.67 22.8309 1274.67 27.611 Q1274.67 31.6214 1277.06 33.6873 Q1279.49 35.7128 1286.22 36.9686 L1291.16 37.9813 Q1300.32 39.7232 1304.65 44.1387 Q1309.03 48.5136 1309.03 55.8863 Q1309.03 64.6767 1303.11 69.2137 Q1297.24 73.7508 1285.85 73.7508 Q1281.56 73.7508 1276.7 72.7785 Q1271.88 71.8063 1266.69 69.9024 L1266.69 61.4765 Q1271.68 64.2716 1276.46 65.6895 Q1281.24 67.1073 1285.85 67.1073 Q1292.86 67.1073 1296.67 64.3527 Q1300.48 61.598 1300.48 56.4939 Q1300.48 52.0379 1297.72 49.5264 Q1295.01 47.0148 1288.77 45.759 L1283.79 44.7868 Q1274.63 42.9639 1270.54 39.075 Q1266.45 35.1862 1266.45 28.2591 Q1266.45 20.2383 1272.08 15.6203 Q1277.75 11.0023 1287.68 11.0023 Q1291.93 11.0023 1296.35 11.7719 Q1300.76 12.5416 1305.38 14.0809 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1328.83 14.324 L1328.83 27.2059 L1344.19 27.2059 L1344.19 32.9987 L1328.83 32.9987 L1328.83 57.6282 Q1328.83 63.1779 1330.33 64.7578 Q1331.87 66.3376 1336.53 66.3376 L1344.19 66.3376 L1344.19 72.576 L1336.53 72.576 Q1327.9 72.576 1324.62 69.3758 Q1321.34 66.1351 1321.34 57.6282 L1321.34 32.9987 L1315.87 32.9987 L1315.87 27.2059 L1321.34 27.2059 L1321.34 14.324 L1328.83 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1374.61 49.7694 Q1365.58 49.7694 1362.09 51.8354 Q1358.61 53.9013 1358.61 58.8839 Q1358.61 62.8538 1361.2 65.2034 Q1363.83 67.5124 1368.33 67.5124 Q1374.53 67.5124 1378.26 63.1374 Q1382.02 58.7219 1382.02 51.4303 L1382.02 49.7694 L1374.61 49.7694 M1389.48 46.6907 L1389.48 72.576 L1382.02 72.576 L1382.02 65.6895 Q1379.47 69.8214 1375.66 71.8063 Q1371.85 73.7508 1366.35 73.7508 Q1359.38 73.7508 1355.25 69.8619 Q1351.15 65.9325 1351.15 59.3701 Q1351.15 51.7138 1356.26 47.825 Q1361.4 43.9361 1371.57 43.9361 L1382.02 43.9361 L1382.02 43.2069 Q1382.02 38.0623 1378.62 35.2672 Q1375.26 32.4315 1369.14 32.4315 Q1365.25 32.4315 1361.57 33.3632 Q1357.88 34.295 1354.48 36.1584 L1354.48 29.2718 Q1358.57 27.692 1362.42 26.9223 Q1366.26 26.1121 1369.91 26.1121 Q1379.75 26.1121 1384.61 31.2163 Q1389.48 36.3204 1389.48 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1431.12 34.1734 Q1429.86 33.4443 1428.36 33.1202 Q1426.91 32.7556 1425.12 32.7556 Q1418.8 32.7556 1415.4 36.8875 Q1412.04 40.9789 1412.04 48.6757 L1412.04 72.576 L1404.55 72.576 L1404.55 27.2059 L1412.04 27.2059 L1412.04 34.2544 Q1414.39 30.1225 1418.16 28.1376 Q1421.92 26.1121 1427.31 26.1121 Q1428.08 26.1121 1429.01 26.2337 Q1429.94 26.3147 1431.08 26.5172 L1431.12 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1433.87 12.096 L1442.14 12.096 L1454.86 63.2184 L1467.54 12.096 L1476.73 12.096 L1489.45 63.2184 L1502.13 12.096 L1510.44 12.096 L1495.25 72.576 L1484.96 72.576 L1472.2 20.0763 L1459.31 72.576 L1449.02 72.576 L1433.87 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1536.28 49.7694 Q1527.25 49.7694 1523.76 51.8354 Q1520.28 53.9013 1520.28 58.8839 Q1520.28 62.8538 1522.87 65.2034 Q1525.51 67.5124 1530 67.5124 Q1536.2 67.5124 1539.93 63.1374 Q1543.69 58.7219 1543.69 51.4303 L1543.69 49.7694 L1536.28 49.7694 M1551.15 46.6907 L1551.15 72.576 L1543.69 72.576 L1543.69 65.6895 Q1541.14 69.8214 1537.33 71.8063 Q1533.53 73.7508 1528.02 73.7508 Q1521.05 73.7508 1516.92 69.8619 Q1512.83 65.9325 1512.83 59.3701 Q1512.83 51.7138 1517.93 47.825 Q1523.07 43.9361 1533.24 43.9361 L1543.69 43.9361 L1543.69 43.2069 Q1543.69 38.0623 1540.29 35.2672 Q1536.93 32.4315 1530.81 32.4315 Q1526.92 32.4315 1523.24 33.3632 Q1519.55 34.295 1516.15 36.1584 L1516.15 29.2718 Q1520.24 27.692 1524.09 26.9223 Q1527.94 26.1121 1531.58 26.1121 Q1541.43 26.1121 1546.29 31.2163 Q1551.15 36.3204 1551.15 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1592.79 34.1734 Q1591.54 33.4443 1590.04 33.1202 Q1588.58 32.7556 1586.8 32.7556 Q1580.48 32.7556 1577.07 36.8875 Q1573.71 40.9789 1573.71 48.6757 L1573.71 72.576 L1566.22 72.576 L1566.22 27.2059 L1573.71 27.2059 L1573.71 34.2544 Q1576.06 30.1225 1579.83 28.1376 Q1583.6 26.1121 1588.98 26.1121 Q1589.75 26.1121 1590.68 26.2337 Q1591.62 26.3147 1592.75 26.5172 L1592.79 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1629.53 28.5427 L1629.53 35.5912 Q1626.37 33.9709 1622.97 33.1607 Q1619.57 32.3505 1615.92 32.3505 Q1610.37 32.3505 1607.58 34.0519 Q1604.82 35.7533 1604.82 39.156 Q1604.82 41.7486 1606.81 43.2475 Q1608.79 44.7058 1614.79 46.0426 L1617.34 46.6097 Q1625.28 48.3111 1628.6 51.4303 Q1631.96 54.509 1631.96 60.0587 Q1631.96 66.3781 1626.94 70.0644 Q1621.96 73.7508 1613.21 73.7508 Q1609.56 73.7508 1605.59 73.0216 Q1601.66 72.3329 1597.29 70.9151 L1597.29 63.2184 Q1601.42 65.3654 1605.43 66.4591 Q1609.44 67.5124 1613.37 67.5124 Q1618.64 67.5124 1621.47 65.73 Q1624.31 63.9071 1624.31 60.6258 Q1624.31 57.5877 1622.24 55.9673 Q1620.22 54.3469 1613.29 52.8481 L1610.7 52.2405 Q1603.77 50.7821 1600.69 47.7845 Q1597.61 44.7463 1597.61 39.4801 Q1597.61 33.0797 1602.15 29.5959 Q1606.69 26.1121 1615.03 26.1121 Q1619.16 26.1121 1622.81 26.7198 Q1626.45 27.3274 1629.53 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1702.85 28.9478 L1702.85 35.9153 Q1699.69 34.1734 1696.49 33.3227 Q1693.33 32.4315 1690.09 32.4315 Q1682.84 32.4315 1678.83 37.0496 Q1674.82 41.6271 1674.82 49.9314 Q1674.82 58.2358 1678.83 62.8538 Q1682.84 67.4314 1690.09 67.4314 Q1693.33 67.4314 1696.49 66.5807 Q1699.69 65.6895 1702.85 63.9476 L1702.85 70.8341 Q1699.73 72.2924 1696.37 73.0216 Q1693.05 73.7508 1689.28 73.7508 Q1679.03 73.7508 1673 67.3098 Q1666.96 60.8689 1666.96 49.9314 Q1666.96 38.832 1673.04 32.472 Q1679.16 26.1121 1689.77 26.1121 Q1693.21 26.1121 1696.49 26.8413 Q1699.78 27.5299 1702.85 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1753.53 45.1919 L1753.53 72.576 L1746.08 72.576 L1746.08 45.4349 Q1746.08 38.994 1743.57 35.7938 Q1741.05 32.5936 1736.03 32.5936 Q1730 32.5936 1726.51 36.4419 Q1723.03 40.2903 1723.03 46.9338 L1723.03 72.576 L1715.53 72.576 L1715.53 9.54393 L1723.03 9.54393 L1723.03 34.2544 Q1725.7 30.163 1729.31 28.1376 Q1732.95 26.1121 1737.69 26.1121 Q1745.51 26.1121 1749.52 30.9732 Q1753.53 35.7938 1753.53 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1789.02 49.7694 Q1779.98 49.7694 1776.5 51.8354 Q1773.02 53.9013 1773.02 58.8839 Q1773.02 62.8538 1775.61 65.2034 Q1778.24 67.5124 1782.74 67.5124 Q1788.94 67.5124 1792.66 63.1374 Q1796.43 58.7219 1796.43 51.4303 L1796.43 49.7694 L1789.02 49.7694 M1803.88 46.6907 L1803.88 72.576 L1796.43 72.576 L1796.43 65.6895 Q1793.88 69.8214 1790.07 71.8063 Q1786.26 73.7508 1780.75 73.7508 Q1773.79 73.7508 1769.65 69.8619 Q1765.56 65.9325 1765.56 59.3701 Q1765.56 51.7138 1770.67 47.825 Q1775.81 43.9361 1785.98 43.9361 L1796.43 43.9361 L1796.43 43.2069 Q1796.43 38.0623 1793.03 35.2672 Q1789.66 32.4315 1783.55 32.4315 Q1779.66 32.4315 1775.97 33.3632 Q1772.29 34.295 1768.88 36.1584 L1768.88 29.2718 Q1772.98 27.692 1776.82 26.9223 Q1780.67 26.1121 1784.32 26.1121 Q1794.16 26.1121 1799.02 31.2163 Q1803.88 36.3204 1803.88 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1845.53 34.1734 Q1844.27 33.4443 1842.77 33.1202 Q1841.31 32.7556 1839.53 32.7556 Q1833.21 32.7556 1829.81 36.8875 Q1826.45 40.9789 1826.45 48.6757 L1826.45 72.576 L1818.95 72.576 L1818.95 27.2059 L1826.45 27.2059 L1826.45 34.2544 Q1828.8 30.1225 1832.56 28.1376 Q1836.33 26.1121 1841.72 26.1121 Q1842.49 26.1121 1843.42 26.2337 Q1844.35 26.3147 1845.49 26.5172 L1845.53 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1873.96 49.7694 Q1864.93 49.7694 1861.45 51.8354 Q1857.96 53.9013 1857.96 58.8839 Q1857.96 62.8538 1860.56 65.2034 Q1863.19 67.5124 1867.69 67.5124 Q1873.88 67.5124 1877.61 63.1374 Q1881.38 58.7219 1881.38 51.4303 L1881.38 49.7694 L1873.96 49.7694 M1888.83 46.6907 L1888.83 72.576 L1881.38 72.576 L1881.38 65.6895 Q1878.83 69.8214 1875.02 71.8063 Q1871.21 73.7508 1865.7 73.7508 Q1858.73 73.7508 1854.6 69.8619 Q1850.51 65.9325 1850.51 59.3701 Q1850.51 51.7138 1855.61 47.825 Q1860.76 43.9361 1870.93 43.9361 L1881.38 43.9361 L1881.38 43.2069 Q1881.38 38.0623 1877.97 35.2672 Q1874.61 32.4315 1868.5 32.4315 Q1864.61 32.4315 1860.92 33.3632 Q1857.23 34.295 1853.83 36.1584 L1853.83 29.2718 Q1857.92 27.692 1861.77 26.9223 Q1865.62 26.1121 1869.27 26.1121 Q1879.11 26.1121 1883.97 31.2163 Q1888.83 36.3204 1888.83 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1936.83 28.9478 L1936.83 35.9153 Q1933.67 34.1734 1930.47 33.3227 Q1927.31 32.4315 1924.07 32.4315 Q1916.82 32.4315 1912.81 37.0496 Q1908.8 41.6271 1908.8 49.9314 Q1908.8 58.2358 1912.81 62.8538 Q1916.82 67.4314 1924.07 67.4314 Q1927.31 67.4314 1930.47 66.5807 Q1933.67 65.6895 1936.83 63.9476 L1936.83 70.8341 Q1933.72 72.2924 1930.35 73.0216 Q1927.03 73.7508 1923.26 73.7508 Q1913.01 73.7508 1906.98 67.3098 Q1900.94 60.8689 1900.94 49.9314 Q1900.94 38.832 1907.02 32.472 Q1913.14 26.1121 1923.75 26.1121 Q1927.19 26.1121 1930.47 26.8413 Q1933.76 27.5299 1936.83 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M1957.17 14.324 L1957.17 27.2059 L1972.52 27.2059 L1972.52 32.9987 L1957.17 32.9987 L1957.17 57.6282 Q1957.17 63.1779 1958.67 64.7578 Q1960.21 66.3376 1964.87 66.3376 L1972.52 66.3376 L1972.52 72.576 L1964.87 72.576 Q1956.24 72.576 1952.96 69.3758 Q1949.68 66.1351 1949.68 57.6282 L1949.68 32.9987 L1944.21 32.9987 L1944.21 27.2059 L1949.68 27.2059 L1949.68 14.324 L1957.17 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M2021.13 48.0275 L2021.13 51.6733 L1986.86 51.6733 Q1987.35 59.3701 1991.48 63.421 Q1995.65 67.4314 2003.07 67.4314 Q2007.36 67.4314 2011.37 66.3781 Q2015.42 65.3249 2019.39 63.2184 L2019.39 70.267 Q2015.38 71.9684 2011.17 72.8596 Q2006.96 73.7508 2002.62 73.7508 Q1991.76 73.7508 1985.4 67.4314 Q1979.09 61.1119 1979.09 50.3365 Q1979.09 39.1965 1985.08 32.6746 Q1991.12 26.1121 2001.32 26.1121 Q2010.48 26.1121 2015.79 32.0264 Q2021.13 37.9003 2021.13 48.0275 M2013.68 45.84 Q2013.6 39.7232 2010.24 36.0774 Q2006.91 32.4315 2001.41 32.4315 Q1995.17 32.4315 1991.4 35.9558 Q1987.67 39.4801 1987.11 45.8805 L2013.68 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M2059.66 34.1734 Q2058.4 33.4443 2056.9 33.1202 Q2055.44 32.7556 2053.66 32.7556 Q2047.34 32.7556 2043.94 36.8875 Q2040.58 40.9789 2040.58 48.6757 L2040.58 72.576 L2033.08 72.576 L2033.08 27.2059 L2040.58 27.2059 L2040.58 34.2544 Q2042.93 30.1225 2046.69 28.1376 Q2050.46 26.1121 2055.85 26.1121 Q2056.62 26.1121 2057.55 26.2337 Q2058.48 26.3147 2059.62 26.5172 L2059.66 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip850)" d="M2096.4 28.5427 L2096.4 35.5912 Q2093.24 33.9709 2089.84 33.1607 Q2086.43 32.3505 2082.79 32.3505 Q2077.24 32.3505 2074.44 34.0519 Q2071.69 35.7533 2071.69 39.156 Q2071.69 41.7486 2073.67 43.2475 Q2075.66 44.7058 2081.65 46.0426 L2084.21 46.6097 Q2092.15 48.3111 2095.47 51.4303 Q2098.83 54.509 2098.83 60.0587 Q2098.83 66.3781 2093.81 70.0644 Q2088.82 73.7508 2080.07 73.7508 Q2076.43 73.7508 2072.46 73.0216 Q2068.53 72.3329 2064.15 70.9151 L2064.15 63.2184 Q2068.29 65.3654 2072.3 66.4591 Q2076.31 67.5124 2080.24 67.5124 Q2085.5 67.5124 2088.34 65.73 Q2091.17 63.9071 2091.17 60.6258 Q2091.17 57.5877 2089.11 55.9673 Q2087.08 54.3469 2080.16 52.8481 L2077.56 52.2405 Q2070.64 50.7821 2067.56 47.7845 Q2064.48 44.7463 2064.48 39.4801 Q2064.48 33.0797 2069.02 29.5959 Q2073.55 26.1121 2081.9 26.1121 Q2086.03 26.1121 2089.67 26.7198 Q2093.32 27.3274 2096.4 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip852)" cx="1205.56" cy="608.696" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="641.076" cy="708.341" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1104.24" cy="678.014" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="698.971" cy="773.326" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1270.69" cy="587.035" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1104.24" cy="634.69" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1147.66" cy="643.355" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1075.29" cy="673.682" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="452.916" cy="1007.27" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1075.29" cy="699.676" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="235.808" cy="1137.24" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1147.66" cy="665.017" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1234.5" cy="587.035" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1610.83" cy="435.402" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1133.19" cy="652.02" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1118.71" cy="630.358" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="1133.19" cy="630.358" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="2291.1" cy="487.39" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip852)" cx="713.445" cy="989.944" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+</svg>
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/2af4f5d4.svg b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/2af4f5d4.svg
new file mode 100644
index 00000000000..189a7662164
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/2af4f5d4.svg
@@ -0,0 +1,73 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip790">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip790)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip791">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip790)" d="M219.033 1423.18 L2352.76 1423.18 L2352.76 47.2441 L219.033 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip792">
+    <rect x="219" y="47" width="2135" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="413.618,1423.18 413.618,47.2441 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="682.011,1423.18 682.011,47.2441 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="950.403,1423.18 950.403,47.2441 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1218.8,1423.18 1218.8,47.2441 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1487.19,1423.18 1487.19,47.2441 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1755.58,1423.18 1755.58,47.2441 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2023.97,1423.18 2023.97,47.2441 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2292.37,1423.18 2292.37,47.2441 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="413.618,1423.18 413.618,1404.28 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="682.011,1423.18 682.011,1404.28 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="950.403,1423.18 950.403,1404.28 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1218.8,1423.18 1218.8,1404.28 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1487.19,1423.18 1487.19,1404.28 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1755.58,1423.18 1755.58,1404.28 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2023.97,1423.18 2023.97,1404.28 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2292.37,1423.18 2292.37,1404.28 "/>
+<path clip-path="url(#clip790)" d="M402.287 1455.09 L390.481 1473.54 L402.287 1473.54 L402.287 1455.09 M401.06 1451.02 L406.94 1451.02 L406.94 1473.54 L411.87 1473.54 L411.87 1477.43 L406.94 1477.43 L406.94 1485.58 L402.287 1485.58 L402.287 1477.43 L386.685 1477.43 L386.685 1472.92 L401.06 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M419.648 1451.02 L438.004 1451.02 L438.004 1454.96 L423.93 1454.96 L423.93 1463.43 Q424.949 1463.08 425.967 1462.92 Q426.986 1462.73 428.004 1462.73 Q433.791 1462.73 437.171 1465.9 Q440.55 1469.08 440.55 1474.49 Q440.55 1480.07 437.078 1483.17 Q433.606 1486.25 427.287 1486.25 Q425.111 1486.25 422.842 1485.88 Q420.597 1485.51 418.189 1484.77 L418.189 1480.07 Q420.273 1481.2 422.495 1481.76 Q424.717 1482.32 427.194 1482.32 Q431.199 1482.32 433.537 1480.21 Q435.875 1478.1 435.875 1474.49 Q435.875 1470.88 433.537 1468.77 Q431.199 1466.67 427.194 1466.67 Q425.319 1466.67 423.444 1467.08 Q421.592 1467.5 419.648 1468.38 L419.648 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M670.275 1455.09 L658.469 1473.54 L670.275 1473.54 L670.275 1455.09 M669.048 1451.02 L674.927 1451.02 L674.927 1473.54 L679.858 1473.54 L679.858 1477.43 L674.927 1477.43 L674.927 1485.58 L670.275 1485.58 L670.275 1477.43 L654.673 1477.43 L654.673 1472.92 L669.048 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M687.728 1484.86 L687.728 1480.6 Q689.487 1481.44 691.293 1481.88 Q693.098 1482.32 694.835 1482.32 Q699.464 1482.32 701.895 1479.21 Q704.348 1476.09 704.696 1469.75 Q703.353 1471.74 701.293 1472.8 Q699.233 1473.87 696.733 1473.87 Q691.548 1473.87 688.515 1470.74 Q685.506 1467.59 685.506 1462.15 Q685.506 1456.83 688.654 1453.61 Q691.802 1450.39 697.034 1450.39 Q703.029 1450.39 706.177 1455 Q709.348 1459.58 709.348 1468.33 Q709.348 1476.51 705.46 1481.39 Q701.594 1486.25 695.043 1486.25 Q693.284 1486.25 691.478 1485.9 Q689.673 1485.56 687.728 1484.86 M697.034 1470.21 Q700.182 1470.21 702.01 1468.06 Q703.862 1465.9 703.862 1462.15 Q703.862 1458.43 702.01 1456.27 Q700.182 1454.1 697.034 1454.1 Q693.885 1454.1 692.034 1456.27 Q690.205 1458.43 690.205 1462.15 Q690.205 1465.9 692.034 1468.06 Q693.885 1470.21 697.034 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M935.276 1469.17 Q931.943 1469.17 930.022 1470.95 Q928.123 1472.73 928.123 1475.86 Q928.123 1478.98 930.022 1480.77 Q931.943 1482.55 935.276 1482.55 Q938.609 1482.55 940.531 1480.77 Q942.452 1478.96 942.452 1475.86 Q942.452 1472.73 940.531 1470.95 Q938.633 1469.17 935.276 1469.17 M930.6 1467.18 Q927.591 1466.44 925.901 1464.38 Q924.235 1462.32 924.235 1459.35 Q924.235 1455.21 927.174 1452.8 Q930.137 1450.39 935.276 1450.39 Q940.438 1450.39 943.378 1452.8 Q946.318 1455.21 946.318 1459.35 Q946.318 1462.32 944.628 1464.38 Q942.961 1466.44 939.975 1467.18 Q943.355 1467.96 945.23 1470.26 Q947.128 1472.55 947.128 1475.86 Q947.128 1480.88 944.049 1483.57 Q940.994 1486.25 935.276 1486.25 Q929.559 1486.25 926.48 1483.57 Q923.424 1480.88 923.424 1475.86 Q923.424 1472.55 925.323 1470.26 Q927.221 1467.96 930.6 1467.18 M928.887 1459.79 Q928.887 1462.48 930.554 1463.98 Q932.244 1465.49 935.276 1465.49 Q938.285 1465.49 939.975 1463.98 Q941.688 1462.48 941.688 1459.79 Q941.688 1457.11 939.975 1455.6 Q938.285 1454.1 935.276 1454.1 Q932.244 1454.1 930.554 1455.6 Q928.887 1457.11 928.887 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M965.438 1454.1 Q961.827 1454.1 959.998 1457.66 Q958.193 1461.2 958.193 1468.33 Q958.193 1475.44 959.998 1479.01 Q961.827 1482.55 965.438 1482.55 Q969.072 1482.55 970.878 1479.01 Q972.707 1475.44 972.707 1468.33 Q972.707 1461.2 970.878 1457.66 Q969.072 1454.1 965.438 1454.1 M965.438 1450.39 Q971.248 1450.39 974.304 1455 Q977.382 1459.58 977.382 1468.33 Q977.382 1477.06 974.304 1481.67 Q971.248 1486.25 965.438 1486.25 Q959.628 1486.25 956.549 1481.67 Q953.494 1477.06 953.494 1468.33 Q953.494 1459.58 956.549 1455 Q959.628 1450.39 965.438 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1208.44 1466.95 Q1211.79 1467.66 1213.67 1469.93 Q1215.57 1472.2 1215.57 1475.53 Q1215.57 1480.65 1212.05 1483.45 Q1208.53 1486.25 1202.05 1486.25 Q1199.87 1486.25 1197.56 1485.81 Q1195.27 1485.39 1192.81 1484.54 L1192.81 1480.02 Q1194.76 1481.16 1197.07 1481.74 Q1199.39 1482.32 1201.91 1482.32 Q1206.31 1482.32 1208.6 1480.58 Q1210.91 1478.84 1210.91 1475.53 Q1210.91 1472.48 1208.76 1470.77 Q1206.63 1469.03 1202.81 1469.03 L1198.78 1469.03 L1198.78 1465.19 L1203 1465.19 Q1206.45 1465.19 1208.28 1463.82 Q1210.1 1462.43 1210.1 1459.84 Q1210.1 1457.18 1208.21 1455.77 Q1206.33 1454.33 1202.81 1454.33 Q1200.89 1454.33 1198.69 1454.75 Q1196.49 1455.16 1193.85 1456.04 L1193.85 1451.88 Q1196.52 1451.14 1198.83 1450.77 Q1201.17 1450.39 1203.23 1450.39 Q1208.55 1450.39 1211.66 1452.83 Q1214.76 1455.23 1214.76 1459.35 Q1214.76 1462.22 1213.11 1464.21 Q1211.47 1466.18 1208.44 1466.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1228.46 1481.64 L1244.78 1481.64 L1244.78 1485.58 L1222.84 1485.58 L1222.84 1481.64 Q1225.5 1478.89 1230.08 1474.26 Q1234.69 1469.61 1235.87 1468.27 Q1238.11 1465.74 1238.99 1464.01 Q1239.9 1462.25 1239.9 1460.56 Q1239.9 1457.8 1237.95 1456.07 Q1236.03 1454.33 1232.93 1454.33 Q1230.73 1454.33 1228.28 1455.09 Q1225.84 1455.86 1223.07 1457.41 L1223.07 1452.69 Q1225.89 1451.55 1228.34 1450.97 Q1230.8 1450.39 1232.84 1450.39 Q1238.21 1450.39 1241.4 1453.08 Q1244.59 1455.77 1244.59 1460.26 Q1244.59 1462.39 1243.78 1464.31 Q1243 1466.2 1240.89 1468.8 Q1240.31 1469.47 1237.21 1472.69 Q1234.11 1475.88 1228.46 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1471.98 1469.17 Q1468.65 1469.17 1466.73 1470.95 Q1464.83 1472.73 1464.83 1475.86 Q1464.83 1478.98 1466.73 1480.77 Q1468.65 1482.55 1471.98 1482.55 Q1475.31 1482.55 1477.24 1480.77 Q1479.16 1478.96 1479.16 1475.86 Q1479.16 1472.73 1477.24 1470.95 Q1475.34 1469.17 1471.98 1469.17 M1467.3 1467.18 Q1464.3 1466.44 1462.61 1464.38 Q1460.94 1462.32 1460.94 1459.35 Q1460.94 1455.21 1463.88 1452.8 Q1466.84 1450.39 1471.98 1450.39 Q1477.14 1450.39 1480.08 1452.8 Q1483.02 1455.21 1483.02 1459.35 Q1483.02 1462.32 1481.33 1464.38 Q1479.67 1466.44 1476.68 1467.18 Q1480.06 1467.96 1481.93 1470.26 Q1483.83 1472.55 1483.83 1475.86 Q1483.83 1480.88 1480.75 1483.57 Q1477.7 1486.25 1471.98 1486.25 Q1466.26 1486.25 1463.18 1483.57 Q1460.13 1480.88 1460.13 1475.86 Q1460.13 1472.55 1462.03 1470.26 Q1463.93 1467.96 1467.3 1467.18 M1465.59 1459.79 Q1465.59 1462.48 1467.26 1463.98 Q1468.95 1465.49 1471.98 1465.49 Q1474.99 1465.49 1476.68 1463.98 Q1478.39 1462.48 1478.39 1459.79 Q1478.39 1457.11 1476.68 1455.6 Q1474.99 1454.1 1471.98 1454.1 Q1468.95 1454.1 1467.26 1455.6 Q1465.59 1457.11 1465.59 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1502.72 1466.44 Q1499.57 1466.44 1497.72 1468.59 Q1495.89 1470.74 1495.89 1474.49 Q1495.89 1478.22 1497.72 1480.39 Q1499.57 1482.55 1502.72 1482.55 Q1505.87 1482.55 1507.7 1480.39 Q1509.55 1478.22 1509.55 1474.49 Q1509.55 1470.74 1507.7 1468.59 Q1505.87 1466.44 1502.72 1466.44 M1512 1451.78 L1512 1456.04 Q1510.24 1455.21 1508.44 1454.77 Q1506.66 1454.33 1504.9 1454.33 Q1500.27 1454.33 1497.81 1457.45 Q1495.38 1460.58 1495.04 1466.9 Q1496.4 1464.89 1498.46 1463.82 Q1500.52 1462.73 1503 1462.73 Q1508.21 1462.73 1511.22 1465.9 Q1514.25 1469.05 1514.25 1474.49 Q1514.25 1479.82 1511.1 1483.03 Q1507.95 1486.25 1502.72 1486.25 Q1496.73 1486.25 1493.55 1481.67 Q1490.38 1477.06 1490.38 1468.33 Q1490.38 1460.14 1494.27 1455.28 Q1498.16 1450.39 1504.71 1450.39 Q1506.47 1450.39 1508.25 1450.74 Q1510.06 1451.09 1512 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1724.16 1451.02 L1730.45 1451.02 L1745.78 1479.93 L1745.78 1451.02 L1750.32 1451.02 L1750.32 1485.58 L1744.02 1485.58 L1728.7 1456.67 L1728.7 1485.58 L1724.16 1485.58 L1724.16 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1771.17 1455.63 L1764.83 1472.83 L1777.54 1472.83 L1771.17 1455.63 M1768.53 1451.02 L1773.83 1451.02 L1787.01 1485.58 L1782.14 1485.58 L1779 1476.71 L1763.42 1476.71 L1760.27 1485.58 L1755.34 1485.58 L1768.53 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1997.38 1451.02 L2019.6 1451.02 L2019.6 1453.01 L2007.05 1485.58 L2002.17 1485.58 L2013.97 1454.96 L1997.38 1454.96 L1997.38 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M2038.72 1469.17 Q2035.39 1469.17 2033.47 1470.95 Q2031.57 1472.73 2031.57 1475.86 Q2031.57 1478.98 2033.47 1480.77 Q2035.39 1482.55 2038.72 1482.55 Q2042.05 1482.55 2043.97 1480.77 Q2045.9 1478.96 2045.9 1475.86 Q2045.9 1472.73 2043.97 1470.95 Q2042.08 1469.17 2038.72 1469.17 M2034.04 1467.18 Q2031.03 1466.44 2029.35 1464.38 Q2027.68 1462.32 2027.68 1459.35 Q2027.68 1455.21 2030.62 1452.8 Q2033.58 1450.39 2038.72 1450.39 Q2043.88 1450.39 2046.82 1452.8 Q2049.76 1455.21 2049.76 1459.35 Q2049.76 1462.32 2048.07 1464.38 Q2046.41 1466.44 2043.42 1467.18 Q2046.8 1467.96 2048.67 1470.26 Q2050.57 1472.55 2050.57 1475.86 Q2050.57 1480.88 2047.49 1483.57 Q2044.44 1486.25 2038.72 1486.25 Q2033 1486.25 2029.92 1483.57 Q2026.87 1480.88 2026.87 1475.86 Q2026.87 1472.55 2028.77 1470.26 Q2030.66 1467.96 2034.04 1467.18 M2032.33 1459.79 Q2032.33 1462.48 2034 1463.98 Q2035.69 1465.49 2038.72 1465.49 Q2041.73 1465.49 2043.42 1463.98 Q2045.13 1462.48 2045.13 1459.79 Q2045.13 1457.11 2043.42 1455.6 Q2041.73 1454.1 2038.72 1454.1 Q2035.69 1454.1 2034 1455.6 Q2032.33 1457.11 2032.33 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M2267.07 1451.02 L2285.42 1451.02 L2285.42 1454.96 L2271.35 1454.96 L2271.35 1463.43 Q2272.37 1463.08 2273.39 1462.92 Q2274.4 1462.73 2275.42 1462.73 Q2281.21 1462.73 2284.59 1465.9 Q2287.97 1469.08 2287.97 1474.49 Q2287.97 1480.07 2284.5 1483.17 Q2281.02 1486.25 2274.71 1486.25 Q2272.53 1486.25 2270.26 1485.88 Q2268.02 1485.51 2265.61 1484.77 L2265.61 1480.07 Q2267.69 1481.2 2269.91 1481.76 Q2272.14 1482.32 2274.61 1482.32 Q2278.62 1482.32 2280.96 1480.21 Q2283.29 1478.1 2283.29 1474.49 Q2283.29 1470.88 2280.96 1468.77 Q2278.62 1466.67 2274.61 1466.67 Q2272.74 1466.67 2270.86 1467.08 Q2269.01 1467.5 2267.07 1468.38 L2267.07 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M2307.18 1454.1 Q2303.57 1454.1 2301.74 1457.66 Q2299.94 1461.2 2299.94 1468.33 Q2299.94 1475.44 2301.74 1479.01 Q2303.57 1482.55 2307.18 1482.55 Q2310.82 1482.55 2312.62 1479.01 Q2314.45 1475.44 2314.45 1468.33 Q2314.45 1461.2 2312.62 1457.66 Q2310.82 1454.1 2307.18 1454.1 M2307.18 1450.39 Q2312.99 1450.39 2316.05 1455 Q2319.13 1459.58 2319.13 1468.33 Q2319.13 1477.06 2316.05 1481.67 Q2312.99 1486.25 2307.18 1486.25 Q2301.37 1486.25 2298.29 1481.67 Q2295.24 1477.06 2295.24 1468.33 Q2295.24 1459.58 2298.29 1455 Q2301.37 1450.39 2307.18 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1174.21 1520.52 L1180.7 1520.52 L1190.7 1560.69 L1200.66 1520.52 L1207.88 1520.52 L1217.88 1560.69 L1227.84 1520.52 L1234.36 1520.52 L1222.43 1568.04 L1214.34 1568.04 L1204.32 1526.79 L1194.2 1568.04 L1186.11 1568.04 L1174.21 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1269.31 1548.76 L1269.31 1551.62 L1242.38 1551.62 Q1242.77 1557.67 1246.01 1560.85 Q1249.29 1564 1255.12 1564 Q1258.49 1564 1261.64 1563.17 Q1264.82 1562.35 1267.94 1560.69 L1267.94 1566.23 Q1264.79 1567.57 1261.48 1568.27 Q1258.17 1568.97 1254.77 1568.97 Q1246.24 1568.97 1241.24 1564 Q1236.27 1559.04 1236.27 1550.57 Q1236.27 1541.82 1240.98 1536.69 Q1245.73 1531.54 1253.75 1531.54 Q1260.94 1531.54 1265.11 1536.18 Q1269.31 1540.8 1269.31 1548.76 M1263.46 1547.04 Q1263.39 1542.23 1260.75 1539.37 Q1258.14 1536.5 1253.81 1536.5 Q1248.91 1536.5 1245.95 1539.27 Q1243.02 1542.04 1242.58 1547.07 L1263.46 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1278.92 1532.4 L1284.78 1532.4 L1284.78 1568.04 L1278.92 1568.04 L1278.92 1532.4 M1278.92 1518.52 L1284.78 1518.52 L1284.78 1525.93 L1278.92 1525.93 L1278.92 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1320.49 1549.81 Q1320.49 1543.44 1317.85 1539.94 Q1315.24 1536.44 1310.5 1536.44 Q1305.79 1536.44 1303.15 1539.94 Q1300.54 1543.44 1300.54 1549.81 Q1300.54 1556.14 1303.15 1559.64 Q1305.79 1563.14 1310.5 1563.14 Q1315.24 1563.14 1317.85 1559.64 Q1320.49 1556.14 1320.49 1549.81 M1326.35 1563.62 Q1326.35 1572.72 1322.31 1577.15 Q1318.26 1581.6 1309.92 1581.6 Q1306.84 1581.6 1304.1 1581.13 Q1301.36 1580.68 1298.78 1579.72 L1298.78 1574.03 Q1301.36 1575.43 1303.88 1576.1 Q1306.39 1576.76 1309 1576.76 Q1314.76 1576.76 1317.63 1573.74 Q1320.49 1570.75 1320.49 1564.67 L1320.49 1561.77 Q1318.68 1564.92 1315.85 1566.48 Q1313.01 1568.04 1309.07 1568.04 Q1302.51 1568.04 1298.5 1563.05 Q1294.49 1558.05 1294.49 1549.81 Q1294.49 1541.53 1298.5 1536.53 Q1302.51 1531.54 1309.07 1531.54 Q1313.01 1531.54 1315.85 1533.1 Q1318.68 1534.66 1320.49 1537.81 L1320.49 1532.4 L1326.35 1532.4 L1326.35 1563.62 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1368.04 1546.53 L1368.04 1568.04 L1362.19 1568.04 L1362.19 1546.72 Q1362.19 1541.66 1360.21 1539.14 Q1358.24 1536.63 1354.29 1536.63 Q1349.55 1536.63 1346.81 1539.65 Q1344.08 1542.68 1344.08 1547.9 L1344.08 1568.04 L1338.19 1568.04 L1338.19 1518.52 L1344.08 1518.52 L1344.08 1537.93 Q1346.18 1534.72 1349.01 1533.13 Q1351.87 1531.54 1355.6 1531.54 Q1361.74 1531.54 1364.89 1535.36 Q1368.04 1539.14 1368.04 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M1385.52 1522.27 L1385.52 1532.4 L1397.58 1532.4 L1397.58 1536.95 L1385.52 1536.95 L1385.52 1556.3 Q1385.52 1560.66 1386.7 1561.9 Q1387.9 1563.14 1391.57 1563.14 L1397.58 1563.14 L1397.58 1568.04 L1391.57 1568.04 Q1384.79 1568.04 1382.21 1565.53 Q1379.63 1562.98 1379.63 1556.3 L1379.63 1536.95 L1375.33 1536.95 L1375.33 1532.4 L1379.63 1532.4 L1379.63 1522.27 L1385.52 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.033,1248.02 2352.76,1248.02 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.033,1047.71 2352.76,1047.71 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.033,847.389 2352.76,847.389 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.033,647.073 2352.76,647.073 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.033,446.756 2352.76,446.756 "/>
+<polyline clip-path="url(#clip792)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.033,246.439 2352.76,246.439 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,1423.18 219.033,47.2441 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,1248.02 237.931,1248.02 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,1047.71 237.931,1047.71 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,847.389 237.931,847.389 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,647.073 237.931,647.073 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,446.756 237.931,446.756 "/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.033,246.439 237.931,246.439 "/>
+<path clip-path="url(#clip790)" d="M116.668 1261.37 L124.306 1261.37 L124.306 1235 L115.996 1236.67 L115.996 1232.41 L124.26 1230.74 L128.936 1230.74 L128.936 1261.37 L136.575 1261.37 L136.575 1265.3 L116.668 1265.3 L116.668 1261.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M146.019 1259.42 L150.903 1259.42 L150.903 1265.3 L146.019 1265.3 L146.019 1259.42 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M171.089 1233.82 Q167.477 1233.82 165.649 1237.39 Q163.843 1240.93 163.843 1248.06 Q163.843 1255.16 165.649 1258.73 Q167.477 1262.27 171.089 1262.27 Q174.723 1262.27 176.528 1258.73 Q178.357 1255.16 178.357 1248.06 Q178.357 1240.93 176.528 1237.39 Q174.723 1233.82 171.089 1233.82 M171.089 1230.12 Q176.899 1230.12 179.954 1234.72 Q183.033 1239.31 183.033 1248.06 Q183.033 1256.78 179.954 1261.39 Q176.899 1265.97 171.089 1265.97 Q165.278 1265.97 162.2 1261.39 Q159.144 1256.78 159.144 1248.06 Q159.144 1239.31 162.2 1234.72 Q165.278 1230.12 171.089 1230.12 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M117.663 1061.05 L125.302 1061.05 L125.302 1034.69 L116.992 1036.35 L116.992 1032.09 L125.255 1030.43 L129.931 1030.43 L129.931 1061.05 L137.57 1061.05 L137.57 1064.99 L117.663 1064.99 L117.663 1061.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M147.015 1059.11 L151.899 1059.11 L151.899 1064.99 L147.015 1064.99 L147.015 1059.11 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M162.13 1030.43 L180.487 1030.43 L180.487 1034.36 L166.413 1034.36 L166.413 1042.83 Q167.431 1042.49 168.45 1042.32 Q169.468 1042.14 170.487 1042.14 Q176.274 1042.14 179.653 1045.31 Q183.033 1048.48 183.033 1053.9 Q183.033 1059.48 179.561 1062.58 Q176.089 1065.66 169.769 1065.66 Q167.593 1065.66 165.325 1065.29 Q163.079 1064.92 160.672 1064.18 L160.672 1059.48 Q162.755 1060.61 164.977 1061.17 Q167.2 1061.72 169.677 1061.72 Q173.681 1061.72 176.019 1059.62 Q178.357 1057.51 178.357 1053.9 Q178.357 1050.29 176.019 1048.18 Q173.681 1046.07 169.677 1046.07 Q167.802 1046.07 165.927 1046.49 Q164.075 1046.91 162.13 1047.79 L162.13 1030.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M119.885 860.734 L136.204 860.734 L136.204 864.669 L114.26 864.669 L114.26 860.734 Q116.922 857.98 121.505 853.35 Q126.112 848.697 127.292 847.355 Q129.538 844.831 130.417 843.095 Q131.32 841.336 131.32 839.646 Q131.32 836.892 129.376 835.156 Q127.455 833.42 124.353 833.42 Q122.154 833.42 119.7 834.183 Q117.269 834.947 114.492 836.498 L114.492 831.776 Q117.316 830.642 119.769 830.063 Q122.223 829.484 124.26 829.484 Q129.63 829.484 132.825 832.17 Q136.019 834.855 136.019 839.345 Q136.019 841.475 135.209 843.396 Q134.422 845.294 132.316 847.887 Q131.737 848.558 128.635 851.776 Q125.533 854.97 119.885 860.734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M146.019 858.79 L150.903 858.79 L150.903 864.669 L146.019 864.669 L146.019 858.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M171.089 833.188 Q167.477 833.188 165.649 836.753 Q163.843 840.294 163.843 847.424 Q163.843 854.531 165.649 858.095 Q167.477 861.637 171.089 861.637 Q174.723 861.637 176.528 858.095 Q178.357 854.531 178.357 847.424 Q178.357 840.294 176.528 836.753 Q174.723 833.188 171.089 833.188 M171.089 829.484 Q176.899 829.484 179.954 834.091 Q183.033 838.674 183.033 847.424 Q183.033 856.151 179.954 860.757 Q176.899 865.341 171.089 865.341 Q165.278 865.341 162.2 860.757 Q159.144 856.151 159.144 847.424 Q159.144 838.674 162.2 834.091 Q165.278 829.484 171.089 829.484 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M120.88 660.417 L137.2 660.417 L137.2 664.353 L115.256 664.353 L115.256 660.417 Q117.918 657.663 122.501 653.033 Q127.107 648.38 128.288 647.038 Q130.533 644.515 131.413 642.779 Q132.316 641.019 132.316 639.33 Q132.316 636.575 130.371 634.839 Q128.45 633.103 125.348 633.103 Q123.149 633.103 120.695 633.867 Q118.265 634.631 115.487 636.181 L115.487 631.459 Q118.311 630.325 120.765 629.746 Q123.218 629.168 125.255 629.168 Q130.626 629.168 133.82 631.853 Q137.015 634.538 137.015 639.029 Q137.015 641.158 136.204 643.08 Q135.417 644.978 133.311 647.57 Q132.732 648.242 129.63 651.459 Q126.529 654.654 120.88 660.417 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M147.015 658.473 L151.899 658.473 L151.899 664.353 L147.015 664.353 L147.015 658.473 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M162.13 629.793 L180.487 629.793 L180.487 633.728 L166.413 633.728 L166.413 642.2 Q167.431 641.853 168.45 641.691 Q169.468 641.505 170.487 641.505 Q176.274 641.505 179.653 644.677 Q183.033 647.848 183.033 653.265 Q183.033 658.843 179.561 661.945 Q176.089 665.024 169.769 665.024 Q167.593 665.024 165.325 664.654 Q163.079 664.283 160.672 663.542 L160.672 658.843 Q162.755 659.978 164.977 660.533 Q167.2 661.089 169.677 661.089 Q173.681 661.089 176.019 658.982 Q178.357 656.876 178.357 653.265 Q178.357 649.654 176.019 647.547 Q173.681 645.441 169.677 645.441 Q167.802 645.441 165.927 645.857 Q164.075 646.274 162.13 647.154 L162.13 629.793 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M130.024 445.402 Q133.38 446.119 135.255 448.388 Q137.154 450.656 137.154 453.99 Q137.154 459.105 133.635 461.906 Q130.117 464.707 123.635 464.707 Q121.459 464.707 119.144 464.267 Q116.853 463.851 114.399 462.994 L114.399 458.48 Q116.343 459.615 118.658 460.193 Q120.973 460.772 123.496 460.772 Q127.894 460.772 130.186 459.036 Q132.501 457.3 132.501 453.99 Q132.501 450.934 130.348 449.221 Q128.218 447.485 124.399 447.485 L120.371 447.485 L120.371 443.642 L124.584 443.642 Q128.033 443.642 129.862 442.277 Q131.691 440.888 131.691 438.295 Q131.691 435.633 129.792 434.221 Q127.917 432.786 124.399 432.786 Q122.478 432.786 120.279 433.203 Q118.08 433.619 115.441 434.499 L115.441 430.332 Q118.103 429.592 120.418 429.221 Q122.755 428.851 124.816 428.851 Q130.14 428.851 133.242 431.281 Q136.343 433.689 136.343 437.809 Q136.343 440.679 134.7 442.67 Q133.056 444.638 130.024 445.402 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M146.019 458.156 L150.903 458.156 L150.903 464.036 L146.019 464.036 L146.019 458.156 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M171.089 432.555 Q167.477 432.555 165.649 436.119 Q163.843 439.661 163.843 446.791 Q163.843 453.897 165.649 457.462 Q167.477 461.003 171.089 461.003 Q174.723 461.003 176.528 457.462 Q178.357 453.897 178.357 446.791 Q178.357 439.661 176.528 436.119 Q174.723 432.555 171.089 432.555 M171.089 428.851 Q176.899 428.851 179.954 433.457 Q183.033 438.041 183.033 446.791 Q183.033 455.517 179.954 460.124 Q176.899 464.707 171.089 464.707 Q165.278 464.707 162.2 460.124 Q159.144 455.517 159.144 446.791 Q159.144 438.041 162.2 433.457 Q165.278 428.851 171.089 428.851 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M131.019 245.085 Q134.376 245.803 136.251 248.071 Q138.149 250.34 138.149 253.673 Q138.149 258.789 134.63 261.589 Q131.112 264.39 124.63 264.39 Q122.455 264.39 120.14 263.951 Q117.848 263.534 115.394 262.677 L115.394 258.164 Q117.339 259.298 119.654 259.877 Q121.968 260.455 124.492 260.455 Q128.89 260.455 131.181 258.719 Q133.496 256.983 133.496 253.673 Q133.496 250.617 131.343 248.904 Q129.214 247.168 125.394 247.168 L121.367 247.168 L121.367 243.326 L125.58 243.326 Q129.029 243.326 130.857 241.96 Q132.686 240.571 132.686 237.978 Q132.686 235.316 130.788 233.904 Q128.913 232.469 125.394 232.469 Q123.473 232.469 121.274 232.886 Q119.075 233.303 116.436 234.182 L116.436 230.016 Q119.098 229.275 121.413 228.904 Q123.751 228.534 125.811 228.534 Q131.135 228.534 134.237 230.965 Q137.339 233.372 137.339 237.492 Q137.339 240.363 135.695 242.353 Q134.052 244.321 131.019 245.085 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M147.015 257.839 L151.899 257.839 L151.899 263.719 L147.015 263.719 L147.015 257.839 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M162.13 229.159 L180.487 229.159 L180.487 233.094 L166.413 233.094 L166.413 241.566 Q167.431 241.219 168.45 241.057 Q169.468 240.872 170.487 240.872 Q176.274 240.872 179.653 244.043 Q183.033 247.215 183.033 252.631 Q183.033 258.21 179.561 261.312 Q176.089 264.39 169.769 264.39 Q167.593 264.39 165.325 264.02 Q163.079 263.65 160.672 262.909 L160.672 258.21 Q162.755 259.344 164.977 259.9 Q167.2 260.455 169.677 260.455 Q173.681 260.455 176.019 258.349 Q178.357 256.242 178.357 252.631 Q178.357 249.02 176.019 246.914 Q173.681 244.807 169.677 244.807 Q167.802 244.807 165.927 245.224 Q164.075 245.64 162.13 246.52 L162.13 229.159 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M16.4842 838.973 L16.4842 832.544 L35.9632 832.544 L35.9632 809.181 L16.4842 809.181 L16.4842 802.752 L64.0042 802.752 L64.0042 809.181 L41.3741 809.181 L41.3741 832.544 L64.0042 832.544 L64.0042 838.973 L16.4842 838.973 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M44.7161 759.72 L47.5806 759.72 L47.5806 786.647 Q53.6281 786.265 56.8109 783.018 Q59.9619 779.74 59.9619 773.915 Q59.9619 770.542 59.1344 767.391 Q58.3069 764.208 56.6518 761.089 L62.1899 761.089 Q63.5267 764.24 64.227 767.55 Q64.9272 770.86 64.9272 774.266 Q64.9272 782.796 59.9619 787.793 Q54.9967 792.758 46.5303 792.758 Q37.7774 792.758 32.6531 788.047 Q27.4968 783.305 27.4968 775.284 Q27.4968 768.091 32.1438 763.921 Q36.7589 759.72 44.7161 759.72 M42.9973 765.576 Q38.1912 765.64 35.3266 768.282 Q32.4621 770.892 32.4621 775.22 Q32.4621 780.122 35.2312 783.082 Q38.0002 786.01 43.0292 786.456 L42.9973 765.576 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M28.3562 750.108 L28.3562 744.251 L64.0042 744.251 L64.0042 750.108 L28.3562 750.108 M14.479 750.108 L14.479 744.251 L21.895 744.251 L21.895 750.108 L14.479 750.108 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M45.7664 708.54 Q39.4007 708.54 35.8996 711.181 Q32.3984 713.791 32.3984 718.534 Q32.3984 723.244 35.8996 725.886 Q39.4007 728.496 45.7664 728.496 Q52.1003 728.496 55.6014 725.886 Q59.1026 723.244 59.1026 718.534 Q59.1026 713.791 55.6014 711.181 Q52.1003 708.54 45.7664 708.54 M59.58 702.683 Q68.683 702.683 73.1071 706.725 Q77.5631 710.768 77.5631 719.107 Q77.5631 722.194 77.0857 724.931 Q76.6401 727.669 75.6852 730.247 L69.9879 730.247 Q71.3884 727.669 72.0568 725.154 Q72.7252 722.64 72.7252 720.03 Q72.7252 714.269 69.7015 711.404 Q66.7096 708.54 60.6303 708.54 L57.7339 708.54 Q60.885 710.354 62.4446 713.187 Q64.0042 716.019 64.0042 719.966 Q64.0042 726.523 59.0071 730.533 Q54.01 734.544 45.7664 734.544 Q37.491 734.544 32.4939 730.533 Q27.4968 726.523 27.4968 719.966 Q27.4968 716.019 29.0564 713.187 Q30.616 710.354 33.7671 708.54 L28.3562 708.54 L28.3562 702.683 L59.58 702.683 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M42.4881 660.988 L64.0042 660.988 L64.0042 666.844 L42.679 666.844 Q37.6183 666.844 35.1038 668.818 Q32.5894 670.791 32.5894 674.738 Q32.5894 679.48 35.6131 682.217 Q38.6368 684.955 43.8567 684.955 L64.0042 684.955 L64.0042 690.843 L14.479 690.843 L14.479 684.955 L33.8944 684.955 Q30.6797 682.854 29.0883 680.021 Q27.4968 677.157 27.4968 673.433 Q27.4968 667.29 31.3163 664.139 Q35.1038 660.988 42.4881 660.988 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M18.2347 643.514 L28.3562 643.514 L28.3562 631.451 L32.9077 631.451 L32.9077 643.514 L52.2594 643.514 Q56.6199 643.514 57.8613 642.336 Q59.1026 641.127 59.1026 637.467 L59.1026 631.451 L64.0042 631.451 L64.0042 637.467 Q64.0042 644.246 61.4897 646.824 Q58.9434 649.402 52.2594 649.402 L32.9077 649.402 L32.9077 653.699 L28.3562 653.699 L28.3562 649.402 L18.2347 649.402 L18.2347 643.514 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip792)" cx="279.421" cy="895.465" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="413.618" cy="987.611" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="547.814" cy="959.567" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="682.011" cy="1047.71" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="816.207" cy="875.434" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="547.814" cy="919.503" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="950.403" cy="927.516" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1084.6" cy="955.56" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1218.8" cy="1264.05" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1084.6" cy="979.598" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1352.99" cy="1384.24" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="950.403" cy="947.548" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1487.19" cy="875.434" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1621.39" cy="735.212" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1755.58" cy="86.1857" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1889.78" cy="935.529" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="2023.97" cy="915.497" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="1889.78" cy="915.497" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="2158.17" cy="783.288" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip792)" cx="2292.37" cy="1248.02" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<path clip-path="url(#clip790)" d="M2015.83 196.789 L2281.63 196.789 L2281.63 93.1086 L2015.83 93.1086  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip790)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2015.83,196.789 2281.63,196.789 2281.63,93.1086 2015.83,93.1086 2015.83,196.789 "/>
+<circle clip-path="url(#clip790)" cx="2110.66" cy="144.949" r="20.48" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.55111"/>
+<path clip-path="url(#clip790)" d="M2219.34 164.636 Q2217.53 169.266 2215.82 170.678 Q2214.1 172.09 2211.23 172.09 L2207.83 172.09 L2207.83 168.525 L2210.33 168.525 Q2212.09 168.525 2213.06 167.692 Q2214.04 166.858 2215.22 163.756 L2215.98 161.812 L2205.49 136.303 L2210.01 136.303 L2218.11 156.581 L2226.21 136.303 L2230.72 136.303 L2219.34 164.636 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip790)" d="M2238.02 158.293 L2245.66 158.293 L2245.66 131.928 L2237.35 133.595 L2237.35 129.335 L2245.61 127.669 L2250.28 127.669 L2250.28 158.293 L2257.92 158.293 L2257.92 162.229 L2238.02 162.229 L2238.02 158.293 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/f8d9d8f6.svg b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/f8d9d8f6.svg
new file mode 100644
index 00000000000..cf0e51dada9
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/f8d9d8f6.svg
@@ -0,0 +1,62 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip910">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip910)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip911">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip910)" d="M175.445 1423.18 L2352.76 1423.18 L2352.76 47.2441 L175.445 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip912">
+    <rect x="175" y="47" width="2178" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="428.842,1423.18 428.842,47.2441 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="762.946,1423.18 762.946,47.2441 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1097.05,1423.18 1097.05,47.2441 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1431.15,1423.18 1431.15,47.2441 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1765.26,1423.18 1765.26,47.2441 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2099.36,1423.18 2099.36,47.2441 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="175.445,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="428.842,1423.18 428.842,1404.28 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="762.946,1423.18 762.946,1404.28 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1097.05,1423.18 1097.05,1404.28 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1431.15,1423.18 1431.15,1404.28 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1765.26,1423.18 1765.26,1404.28 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2099.36,1423.18 2099.36,1404.28 "/>
+<path clip-path="url(#clip910)" d="M378.6 1472.64 Q378.6 1467.94 376.655 1465.28 Q374.734 1462.59 371.354 1462.59 Q367.975 1462.59 366.03 1465.28 Q364.109 1467.94 364.109 1472.64 Q364.109 1477.34 366.03 1480.02 Q367.975 1482.69 371.354 1482.69 Q374.734 1482.69 376.655 1480.02 Q378.6 1477.34 378.6 1472.64 M364.109 1463.59 Q365.452 1461.27 367.489 1460.16 Q369.549 1459.03 372.396 1459.03 Q377.118 1459.03 380.058 1462.78 Q383.021 1466.53 383.021 1472.64 Q383.021 1478.75 380.058 1482.5 Q377.118 1486.25 372.396 1486.25 Q369.549 1486.25 367.489 1485.14 Q365.452 1484.01 364.109 1481.69 L364.109 1485.58 L359.827 1485.58 L359.827 1449.56 L364.109 1449.56 L364.109 1463.59 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M405.104 1463.64 Q404.387 1463.22 403.53 1463.03 Q402.697 1462.83 401.678 1462.83 Q398.067 1462.83 396.123 1465.19 Q394.201 1467.52 394.201 1471.92 L394.201 1485.58 L389.919 1485.58 L389.919 1459.65 L394.201 1459.65 L394.201 1463.68 Q395.544 1461.32 397.697 1460.19 Q399.85 1459.03 402.928 1459.03 Q403.368 1459.03 403.9 1459.1 Q404.433 1459.14 405.081 1459.26 L405.104 1463.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M418.576 1462.64 Q415.15 1462.64 413.16 1465.33 Q411.169 1467.99 411.169 1472.64 Q411.169 1477.29 413.137 1479.98 Q415.127 1482.64 418.576 1482.64 Q421.979 1482.64 423.97 1479.95 Q425.961 1477.27 425.961 1472.64 Q425.961 1468.03 423.97 1465.35 Q421.979 1462.64 418.576 1462.64 M418.576 1459.03 Q424.132 1459.03 427.303 1462.64 Q430.474 1466.25 430.474 1472.64 Q430.474 1479.01 427.303 1482.64 Q424.132 1486.25 418.576 1486.25 Q412.998 1486.25 409.826 1482.64 Q406.678 1479.01 406.678 1472.64 Q406.678 1466.25 409.826 1462.64 Q412.998 1459.03 418.576 1459.03 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M435.058 1459.65 L439.317 1459.65 L444.641 1479.89 L449.942 1459.65 L454.965 1459.65 L460.289 1479.89 L465.59 1459.65 L469.849 1459.65 L463.067 1485.58 L458.044 1485.58 L452.465 1464.33 L446.863 1485.58 L441.84 1485.58 L435.058 1459.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M497.858 1469.93 L497.858 1485.58 L493.599 1485.58 L493.599 1470.07 Q493.599 1466.39 492.164 1464.56 Q490.729 1462.73 487.858 1462.73 Q484.409 1462.73 482.419 1464.93 Q480.428 1467.13 480.428 1470.93 L480.428 1485.58 L476.145 1485.58 L476.145 1459.65 L480.428 1459.65 L480.428 1463.68 Q481.956 1461.34 484.016 1460.19 Q486.099 1459.03 488.807 1459.03 Q493.275 1459.03 495.567 1461.81 Q497.858 1464.56 497.858 1469.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M733.895 1472.64 Q733.895 1467.94 731.95 1465.28 Q730.029 1462.59 726.65 1462.59 Q723.27 1462.59 721.326 1465.28 Q719.404 1467.94 719.404 1472.64 Q719.404 1477.34 721.326 1480.02 Q723.27 1482.69 726.65 1482.69 Q730.029 1482.69 731.95 1480.02 Q733.895 1477.34 733.895 1472.64 M719.404 1463.59 Q720.747 1461.27 722.784 1460.16 Q724.844 1459.03 727.691 1459.03 Q732.413 1459.03 735.353 1462.78 Q738.316 1466.53 738.316 1472.64 Q738.316 1478.75 735.353 1482.5 Q732.413 1486.25 727.691 1486.25 Q724.844 1486.25 722.784 1485.14 Q720.747 1484.01 719.404 1481.69 L719.404 1485.58 L715.122 1485.58 L715.122 1449.56 L719.404 1449.56 L719.404 1463.59 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M745.376 1449.56 L749.636 1449.56 L749.636 1485.58 L745.376 1485.58 L745.376 1449.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M758.108 1475.35 L758.108 1459.65 L762.367 1459.65 L762.367 1475.19 Q762.367 1478.87 763.802 1480.72 Q765.237 1482.55 768.108 1482.55 Q771.557 1482.55 773.547 1480.35 Q775.561 1478.15 775.561 1474.35 L775.561 1459.65 L779.821 1459.65 L779.821 1485.58 L775.561 1485.58 L775.561 1481.6 Q774.01 1483.96 771.95 1485.12 Q769.913 1486.25 767.205 1486.25 Q762.737 1486.25 760.423 1483.47 Q758.108 1480.7 758.108 1475.35 M768.825 1459.03 L768.825 1459.03 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M810.77 1471.55 L810.77 1473.64 L791.186 1473.64 Q791.464 1478.03 793.825 1480.35 Q796.209 1482.64 800.446 1482.64 Q802.899 1482.64 805.191 1482.04 Q807.506 1481.44 809.774 1480.23 L809.774 1484.26 Q807.483 1485.23 805.075 1485.74 Q802.668 1486.25 800.191 1486.25 Q793.987 1486.25 790.353 1482.64 Q786.742 1479.03 786.742 1472.87 Q786.742 1466.51 790.168 1462.78 Q793.617 1459.03 799.45 1459.03 Q804.682 1459.03 807.714 1462.41 Q810.77 1465.77 810.77 1471.55 M806.51 1470.3 Q806.464 1466.81 804.543 1464.72 Q802.645 1462.64 799.496 1462.64 Q795.932 1462.64 793.779 1464.65 Q791.649 1466.67 791.325 1470.33 L806.51 1470.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1036.92 1487.99 Q1035.12 1492.62 1033.4 1494.03 Q1031.69 1495.44 1028.82 1495.44 L1025.42 1495.44 L1025.42 1491.88 L1027.92 1491.88 Q1029.68 1491.88 1030.65 1491.04 Q1031.62 1490.21 1032.8 1487.11 L1033.57 1485.16 L1023.08 1459.65 L1027.59 1459.65 L1035.7 1479.93 L1043.8 1459.65 L1048.31 1459.65 L1036.92 1487.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1076.37 1471.55 L1076.37 1473.64 L1056.78 1473.64 Q1057.06 1478.03 1059.42 1480.35 Q1061.81 1482.64 1066.04 1482.64 Q1068.5 1482.64 1070.79 1482.04 Q1073.1 1481.44 1075.37 1480.23 L1075.37 1484.26 Q1073.08 1485.23 1070.67 1485.74 Q1068.26 1486.25 1065.79 1486.25 Q1059.58 1486.25 1055.95 1482.64 Q1052.34 1479.03 1052.34 1472.87 Q1052.34 1466.51 1055.76 1462.78 Q1059.21 1459.03 1065.05 1459.03 Q1070.28 1459.03 1073.31 1462.41 Q1076.37 1465.77 1076.37 1471.55 M1072.11 1470.3 Q1072.06 1466.81 1070.14 1464.72 Q1068.24 1462.64 1065.09 1462.64 Q1061.53 1462.64 1059.38 1464.65 Q1057.25 1466.67 1056.92 1470.33 L1072.11 1470.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1083.36 1449.56 L1087.62 1449.56 L1087.62 1485.58 L1083.36 1485.58 L1083.36 1449.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1096.53 1449.56 L1100.79 1449.56 L1100.79 1485.58 L1096.53 1485.58 L1096.53 1449.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1119.75 1462.64 Q1116.32 1462.64 1114.33 1465.33 Q1112.34 1467.99 1112.34 1472.64 Q1112.34 1477.29 1114.31 1479.98 Q1116.3 1482.64 1119.75 1482.64 Q1123.15 1482.64 1125.14 1479.95 Q1127.13 1477.27 1127.13 1472.64 Q1127.13 1468.03 1125.14 1465.35 Q1123.15 1462.64 1119.75 1462.64 M1119.75 1459.03 Q1125.3 1459.03 1128.47 1462.64 Q1131.64 1466.25 1131.64 1472.64 Q1131.64 1479.01 1128.47 1482.64 Q1125.3 1486.25 1119.75 1486.25 Q1114.17 1486.25 1111 1482.64 Q1107.85 1479.01 1107.85 1472.64 Q1107.85 1466.25 1111 1462.64 Q1114.17 1459.03 1119.75 1459.03 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1136.23 1459.65 L1140.49 1459.65 L1145.81 1479.89 L1151.11 1459.65 L1156.13 1459.65 L1161.46 1479.89 L1166.76 1459.65 L1171.02 1459.65 L1164.24 1485.58 L1159.21 1485.58 L1153.63 1464.33 L1148.03 1485.58 L1143.01 1485.58 L1136.23 1459.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1348.2 1472.64 Q1348.2 1467.94 1346.26 1465.28 Q1344.34 1462.59 1340.96 1462.59 Q1337.58 1462.59 1335.63 1465.28 Q1333.71 1467.94 1333.71 1472.64 Q1333.71 1477.34 1335.63 1480.02 Q1337.58 1482.69 1340.96 1482.69 Q1344.34 1482.69 1346.26 1480.02 Q1348.2 1477.34 1348.2 1472.64 M1333.71 1463.59 Q1335.05 1461.27 1337.09 1460.16 Q1339.15 1459.03 1342 1459.03 Q1346.72 1459.03 1349.66 1462.78 Q1352.62 1466.53 1352.62 1472.64 Q1352.62 1478.75 1349.66 1482.5 Q1346.72 1486.25 1342 1486.25 Q1339.15 1486.25 1337.09 1485.14 Q1335.05 1484.01 1333.71 1481.69 L1333.71 1485.58 L1329.43 1485.58 L1329.43 1449.56 L1333.71 1449.56 L1333.71 1463.59 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1359.68 1449.56 L1363.94 1449.56 L1363.94 1485.58 L1359.68 1485.58 L1359.68 1449.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1372.41 1475.35 L1372.41 1459.65 L1376.67 1459.65 L1376.67 1475.19 Q1376.67 1478.87 1378.11 1480.72 Q1379.54 1482.55 1382.41 1482.55 Q1385.86 1482.55 1387.85 1480.35 Q1389.87 1478.15 1389.87 1474.35 L1389.87 1459.65 L1394.13 1459.65 L1394.13 1485.58 L1389.87 1485.58 L1389.87 1481.6 Q1388.32 1483.96 1386.26 1485.12 Q1384.22 1486.25 1381.51 1486.25 Q1377.04 1486.25 1374.73 1483.47 Q1372.41 1480.7 1372.41 1475.35 M1383.13 1459.03 L1383.13 1459.03 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1425.08 1471.55 L1425.08 1473.64 L1405.49 1473.64 Q1405.77 1478.03 1408.13 1480.35 Q1410.52 1482.64 1414.75 1482.64 Q1417.21 1482.64 1419.5 1482.04 Q1421.81 1481.44 1424.08 1480.23 L1424.08 1484.26 Q1421.79 1485.23 1419.38 1485.74 Q1416.97 1486.25 1414.5 1486.25 Q1408.29 1486.25 1404.66 1482.64 Q1401.05 1479.03 1401.05 1472.87 Q1401.05 1466.51 1404.47 1462.78 Q1407.92 1459.03 1413.76 1459.03 Q1418.99 1459.03 1422.02 1462.41 Q1425.08 1465.77 1425.08 1471.55 M1420.82 1470.3 Q1420.77 1466.81 1418.85 1464.72 Q1416.95 1462.64 1413.8 1462.64 Q1410.24 1462.64 1408.09 1464.65 Q1405.96 1466.67 1405.63 1470.33 L1420.82 1470.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1449.13 1472.32 Q1449.13 1467.69 1447.21 1465.14 Q1445.31 1462.59 1441.86 1462.59 Q1438.43 1462.59 1436.51 1465.14 Q1434.61 1467.69 1434.61 1472.32 Q1434.61 1476.92 1436.51 1479.47 Q1438.43 1482.02 1441.86 1482.02 Q1445.31 1482.02 1447.21 1479.47 Q1449.13 1476.92 1449.13 1472.32 M1453.39 1482.36 Q1453.39 1488.98 1450.45 1492.2 Q1447.51 1495.44 1441.44 1495.44 Q1439.2 1495.44 1437.21 1495.09 Q1435.21 1494.77 1433.34 1494.08 L1433.34 1489.93 Q1435.21 1490.95 1437.04 1491.44 Q1438.87 1491.92 1440.77 1491.92 Q1444.96 1491.92 1447.04 1489.72 Q1449.13 1487.55 1449.13 1483.13 L1449.13 1481.02 Q1447.81 1483.31 1445.75 1484.45 Q1443.69 1485.58 1440.82 1485.58 Q1436.05 1485.58 1433.13 1481.95 Q1430.21 1478.31 1430.21 1472.32 Q1430.21 1466.3 1433.13 1462.66 Q1436.05 1459.03 1440.82 1459.03 Q1443.69 1459.03 1445.75 1460.16 Q1447.81 1461.3 1449.13 1463.59 L1449.13 1459.65 L1453.39 1459.65 L1453.39 1482.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1477.18 1463.64 Q1476.46 1463.22 1475.61 1463.03 Q1474.77 1462.83 1473.76 1462.83 Q1470.15 1462.83 1468.2 1465.19 Q1466.28 1467.52 1466.28 1471.92 L1466.28 1485.58 L1462 1485.58 L1462 1459.65 L1466.28 1459.65 L1466.28 1463.68 Q1467.62 1461.32 1469.77 1460.19 Q1471.93 1459.03 1475.01 1459.03 Q1475.45 1459.03 1475.98 1459.1 Q1476.51 1459.14 1477.16 1459.26 L1477.18 1463.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1493.43 1472.55 Q1488.27 1472.55 1486.28 1473.73 Q1484.29 1474.91 1484.29 1477.76 Q1484.29 1480.02 1485.77 1481.37 Q1487.27 1482.69 1489.84 1482.69 Q1493.39 1482.69 1495.52 1480.19 Q1497.67 1477.66 1497.67 1473.5 L1497.67 1472.55 L1493.43 1472.55 M1501.93 1470.79 L1501.93 1485.58 L1497.67 1485.58 L1497.67 1481.64 Q1496.21 1484.01 1494.03 1485.14 Q1491.86 1486.25 1488.71 1486.25 Q1484.73 1486.25 1482.37 1484.03 Q1480.03 1481.78 1480.03 1478.03 Q1480.03 1473.66 1482.95 1471.44 Q1485.89 1469.21 1491.7 1469.21 L1497.67 1469.21 L1497.67 1468.8 Q1497.67 1465.86 1495.72 1464.26 Q1493.8 1462.64 1490.31 1462.64 Q1488.08 1462.64 1485.98 1463.17 Q1483.87 1463.7 1481.93 1464.77 L1481.93 1460.83 Q1484.27 1459.93 1486.46 1459.49 Q1488.66 1459.03 1490.75 1459.03 Q1496.37 1459.03 1499.15 1461.95 Q1501.93 1464.86 1501.93 1470.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1521.49 1487.99 Q1519.68 1492.62 1517.97 1494.03 Q1516.26 1495.44 1513.39 1495.44 L1509.98 1495.44 L1509.98 1491.88 L1512.48 1491.88 Q1514.24 1491.88 1515.21 1491.04 Q1516.19 1490.21 1517.37 1487.11 L1518.13 1485.16 L1507.64 1459.65 L1512.16 1459.65 L1520.26 1479.93 L1528.36 1459.65 L1532.88 1459.65 L1521.49 1487.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1736.45 1472.32 Q1736.45 1467.69 1734.53 1465.14 Q1732.63 1462.59 1729.18 1462.59 Q1725.75 1462.59 1723.83 1465.14 Q1721.93 1467.69 1721.93 1472.32 Q1721.93 1476.92 1723.83 1479.47 Q1725.75 1482.02 1729.18 1482.02 Q1732.63 1482.02 1734.53 1479.47 Q1736.45 1476.92 1736.45 1472.32 M1740.71 1482.36 Q1740.71 1488.98 1737.77 1492.2 Q1734.83 1495.44 1728.76 1495.44 Q1726.52 1495.44 1724.53 1495.09 Q1722.54 1494.77 1720.66 1494.08 L1720.66 1489.93 Q1722.54 1490.95 1724.36 1491.44 Q1726.19 1491.92 1728.09 1491.92 Q1732.28 1491.92 1734.36 1489.72 Q1736.45 1487.55 1736.45 1483.13 L1736.45 1481.02 Q1735.13 1483.31 1733.07 1484.45 Q1731.01 1485.58 1728.14 1485.58 Q1723.37 1485.58 1720.45 1481.95 Q1717.54 1478.31 1717.54 1472.32 Q1717.54 1466.3 1720.45 1462.66 Q1723.37 1459.03 1728.14 1459.03 Q1731.01 1459.03 1733.07 1460.16 Q1735.13 1461.3 1736.45 1463.59 L1736.45 1459.65 L1740.71 1459.65 L1740.71 1482.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1759.53 1462.64 Q1756.1 1462.64 1754.11 1465.33 Q1752.12 1467.99 1752.12 1472.64 Q1752.12 1477.29 1754.09 1479.98 Q1756.08 1482.64 1759.53 1482.64 Q1762.93 1482.64 1764.92 1479.95 Q1766.91 1477.27 1766.91 1472.64 Q1766.91 1468.03 1764.92 1465.35 Q1762.93 1462.64 1759.53 1462.64 M1759.53 1459.03 Q1765.08 1459.03 1768.25 1462.64 Q1771.42 1466.25 1771.42 1472.64 Q1771.42 1479.01 1768.25 1482.64 Q1765.08 1486.25 1759.53 1486.25 Q1753.95 1486.25 1750.78 1482.64 Q1747.63 1479.01 1747.63 1472.64 Q1747.63 1466.25 1750.78 1462.64 Q1753.95 1459.03 1759.53 1459.03 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1778.48 1449.56 L1782.74 1449.56 L1782.74 1485.58 L1778.48 1485.58 L1778.48 1449.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1808.72 1463.59 L1808.72 1449.56 L1812.98 1449.56 L1812.98 1485.58 L1808.72 1485.58 L1808.72 1481.69 Q1807.37 1484.01 1805.31 1485.14 Q1803.28 1486.25 1800.41 1486.25 Q1795.71 1486.25 1792.74 1482.5 Q1789.8 1478.75 1789.8 1472.64 Q1789.8 1466.53 1792.74 1462.78 Q1795.71 1459.03 1800.41 1459.03 Q1803.28 1459.03 1805.31 1460.16 Q1807.37 1461.27 1808.72 1463.59 M1794.2 1472.64 Q1794.2 1477.34 1796.12 1480.02 Q1798.07 1482.69 1801.45 1482.69 Q1804.83 1482.69 1806.77 1480.02 Q1808.72 1477.34 1808.72 1472.64 Q1808.72 1467.94 1806.77 1465.28 Q1804.83 1462.59 1801.45 1462.59 Q1798.07 1462.59 1796.12 1465.28 Q1794.2 1467.94 1794.2 1472.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M2057.44 1472.64 Q2057.44 1467.94 2055.49 1465.28 Q2053.57 1462.59 2050.19 1462.59 Q2046.81 1462.59 2044.87 1465.28 Q2042.95 1467.94 2042.95 1472.64 Q2042.95 1477.34 2044.87 1480.02 Q2046.81 1482.69 2050.19 1482.69 Q2053.57 1482.69 2055.49 1480.02 Q2057.44 1477.34 2057.44 1472.64 M2042.95 1463.59 Q2044.29 1461.27 2046.33 1460.16 Q2048.39 1459.03 2051.23 1459.03 Q2055.96 1459.03 2058.9 1462.78 Q2061.86 1466.53 2061.86 1472.64 Q2061.86 1478.75 2058.9 1482.5 Q2055.96 1486.25 2051.23 1486.25 Q2048.39 1486.25 2046.33 1485.14 Q2044.29 1484.01 2042.95 1481.69 L2042.95 1485.58 L2038.66 1485.58 L2038.66 1449.56 L2042.95 1449.56 L2042.95 1463.59 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M2068.92 1449.56 L2073.18 1449.56 L2073.18 1485.58 L2068.92 1485.58 L2068.92 1449.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M2093.87 1472.55 Q2088.71 1472.55 2086.72 1473.73 Q2084.73 1474.91 2084.73 1477.76 Q2084.73 1480.02 2086.21 1481.37 Q2087.72 1482.69 2090.28 1482.69 Q2093.83 1482.69 2095.96 1480.19 Q2098.11 1477.66 2098.11 1473.5 L2098.11 1472.55 L2093.87 1472.55 M2102.37 1470.79 L2102.37 1485.58 L2098.11 1485.58 L2098.11 1481.64 Q2096.65 1484.01 2094.47 1485.14 Q2092.3 1486.25 2089.15 1486.25 Q2085.17 1486.25 2082.81 1484.03 Q2080.47 1481.78 2080.47 1478.03 Q2080.47 1473.66 2083.39 1471.44 Q2086.33 1469.21 2092.14 1469.21 L2098.11 1469.21 L2098.11 1468.8 Q2098.11 1465.86 2096.16 1464.26 Q2094.24 1462.64 2090.75 1462.64 Q2088.53 1462.64 2086.42 1463.17 Q2084.31 1463.7 2082.37 1464.77 L2082.37 1460.83 Q2084.71 1459.93 2086.91 1459.49 Q2089.1 1459.03 2091.19 1459.03 Q2096.81 1459.03 2099.59 1461.95 Q2102.37 1464.86 2102.37 1470.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M2129.8 1460.65 L2129.8 1464.63 Q2127.99 1463.64 2126.16 1463.15 Q2124.36 1462.64 2122.51 1462.64 Q2118.36 1462.64 2116.07 1465.28 Q2113.78 1467.89 2113.78 1472.64 Q2113.78 1477.39 2116.07 1480.02 Q2118.36 1482.64 2122.51 1482.64 Q2124.36 1482.64 2126.16 1482.15 Q2127.99 1481.64 2129.8 1480.65 L2129.8 1484.58 Q2128.02 1485.42 2126.09 1485.83 Q2124.2 1486.25 2122.04 1486.25 Q2116.19 1486.25 2112.74 1482.57 Q2109.29 1478.89 2109.29 1472.64 Q2109.29 1466.3 2112.76 1462.66 Q2116.26 1459.03 2122.32 1459.03 Q2124.29 1459.03 2126.16 1459.45 Q2128.04 1459.84 2129.8 1460.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M2137.04 1449.56 L2141.33 1449.56 L2141.33 1470.83 L2154.03 1459.65 L2159.47 1459.65 L2145.72 1471.78 L2160.05 1485.58 L2154.5 1485.58 L2141.33 1472.92 L2141.33 1485.58 L2137.04 1485.58 L2137.04 1449.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1116.75 1520.52 L1146.8 1520.52 L1146.8 1525.93 L1123.18 1525.93 L1123.18 1540 L1145.81 1540 L1145.81 1545.41 L1123.18 1545.41 L1123.18 1562.63 L1147.37 1562.63 L1147.37 1568.04 L1116.75 1568.04 L1116.75 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1172.51 1571.35 Q1170.03 1577.72 1167.68 1579.66 Q1165.32 1581.6 1161.37 1581.6 L1156.7 1581.6 L1156.7 1576.7 L1160.13 1576.7 Q1162.55 1576.7 1163.89 1575.56 Q1165.23 1574.41 1166.85 1570.14 L1167.9 1567.47 L1153.48 1532.4 L1159.69 1532.4 L1170.83 1560.28 L1181.97 1532.4 L1188.17 1532.4 L1172.51 1571.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1226.75 1548.76 L1226.75 1551.62 L1199.82 1551.62 Q1200.2 1557.67 1203.45 1560.85 Q1206.73 1564 1212.55 1564 Q1215.93 1564 1219.08 1563.17 Q1222.26 1562.35 1225.38 1560.69 L1225.38 1566.23 Q1222.23 1567.57 1218.92 1568.27 Q1215.61 1568.97 1212.2 1568.97 Q1203.67 1568.97 1198.68 1564 Q1193.71 1559.04 1193.71 1550.57 Q1193.71 1541.82 1198.42 1536.69 Q1203.16 1531.54 1211.19 1531.54 Q1218.38 1531.54 1222.55 1536.18 Q1226.75 1540.8 1226.75 1548.76 M1220.89 1547.04 Q1220.83 1542.23 1218.19 1539.37 Q1215.58 1536.5 1211.25 1536.5 Q1206.35 1536.5 1203.39 1539.27 Q1200.46 1542.04 1200.01 1547.07 L1220.89 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1282.74 1533.76 L1282.74 1539.24 Q1280.25 1537.87 1277.74 1537.2 Q1275.26 1536.5 1272.71 1536.5 Q1267.01 1536.5 1263.86 1540.13 Q1260.71 1543.73 1260.71 1550.25 Q1260.71 1556.78 1263.86 1560.4 Q1267.01 1564 1272.71 1564 Q1275.26 1564 1277.74 1563.33 Q1280.25 1562.63 1282.74 1561.26 L1282.74 1566.68 Q1280.29 1567.82 1277.64 1568.39 Q1275.03 1568.97 1272.07 1568.97 Q1264.02 1568.97 1259.28 1563.91 Q1254.54 1558.85 1254.54 1550.25 Q1254.54 1541.53 1259.31 1536.53 Q1264.12 1531.54 1272.46 1531.54 Q1275.16 1531.54 1277.74 1532.11 Q1280.32 1532.65 1282.74 1533.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1306.73 1536.5 Q1302.02 1536.5 1299.29 1540.19 Q1296.55 1543.85 1296.55 1550.25 Q1296.55 1556.65 1299.26 1560.34 Q1301.99 1564 1306.73 1564 Q1311.41 1564 1314.15 1560.31 Q1316.89 1556.62 1316.89 1550.25 Q1316.89 1543.92 1314.15 1540.23 Q1311.41 1536.5 1306.73 1536.5 M1306.73 1531.54 Q1314.37 1531.54 1318.73 1536.5 Q1323.09 1541.47 1323.09 1550.25 Q1323.09 1559 1318.73 1564 Q1314.37 1568.97 1306.73 1568.97 Q1299.06 1568.97 1294.7 1564 Q1290.38 1559 1290.38 1550.25 Q1290.38 1541.47 1294.7 1536.5 Q1299.06 1531.54 1306.73 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1332.8 1518.52 L1338.66 1518.52 L1338.66 1568.04 L1332.8 1568.04 L1332.8 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1364.73 1536.5 Q1360.02 1536.5 1357.28 1540.19 Q1354.54 1543.85 1354.54 1550.25 Q1354.54 1556.65 1357.25 1560.34 Q1359.98 1564 1364.73 1564 Q1369.41 1564 1372.14 1560.31 Q1374.88 1556.62 1374.88 1550.25 Q1374.88 1543.92 1372.14 1540.23 Q1369.41 1536.5 1364.73 1536.5 M1364.73 1531.54 Q1372.37 1531.54 1376.73 1536.5 Q1381.09 1541.47 1381.09 1550.25 Q1381.09 1559 1376.73 1564 Q1372.37 1568.97 1364.73 1568.97 Q1357.06 1568.97 1352.7 1564 Q1348.37 1559 1348.37 1550.25 Q1348.37 1541.47 1352.7 1536.5 Q1357.06 1531.54 1364.73 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M1411.45 1537.87 Q1410.46 1537.3 1409.29 1537.04 Q1408.14 1536.76 1406.74 1536.76 Q1401.77 1536.76 1399.1 1540 Q1396.46 1543.22 1396.46 1549.27 L1396.46 1568.04 L1390.57 1568.04 L1390.57 1532.4 L1396.46 1532.4 L1396.46 1537.93 Q1398.31 1534.69 1401.27 1533.13 Q1404.23 1531.54 1408.46 1531.54 Q1409.06 1531.54 1409.8 1531.63 Q1410.53 1531.7 1411.42 1531.85 L1411.45 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="175.445,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="175.445,1079.2 2352.76,1079.2 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="175.445,735.212 2352.76,735.212 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="175.445,391.228 2352.76,391.228 "/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="175.445,47.2441 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="175.445,1423.18 175.445,47.2441 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="175.445,1423.18 194.343,1423.18 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="175.445,1079.2 194.343,1079.2 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="175.445,735.212 194.343,735.212 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="175.445,391.228 194.343,391.228 "/>
+<polyline clip-path="url(#clip910)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="175.445,47.2441 194.343,47.2441 "/>
+<path clip-path="url(#clip910)" d="M127.501 1408.98 Q123.89 1408.98 122.061 1412.54 Q120.255 1416.08 120.255 1423.21 Q120.255 1430.32 122.061 1433.89 Q123.89 1437.43 127.501 1437.43 Q131.135 1437.43 132.941 1433.89 Q134.769 1430.32 134.769 1423.21 Q134.769 1416.08 132.941 1412.54 Q131.135 1408.98 127.501 1408.98 M127.501 1405.27 Q133.311 1405.27 136.367 1409.88 Q139.445 1414.46 139.445 1423.21 Q139.445 1431.94 136.367 1436.55 Q133.311 1441.13 127.501 1441.13 Q121.691 1441.13 118.612 1436.55 Q115.556 1431.94 115.556 1423.21 Q115.556 1414.46 118.612 1409.88 Q121.691 1405.27 127.501 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M123.126 1092.54 L139.445 1092.54 L139.445 1096.48 L117.501 1096.48 L117.501 1092.54 Q120.163 1089.79 124.746 1085.16 Q129.353 1080.5 130.533 1079.16 Q132.779 1076.64 133.658 1074.9 Q134.561 1073.14 134.561 1071.45 Q134.561 1068.7 132.617 1066.96 Q130.695 1065.23 127.593 1065.23 Q125.394 1065.23 122.941 1065.99 Q120.51 1066.75 117.732 1068.3 L117.732 1063.58 Q120.556 1062.45 123.01 1061.87 Q125.464 1061.29 127.501 1061.29 Q132.871 1061.29 136.066 1063.98 Q139.26 1066.66 139.26 1071.15 Q139.26 1073.28 138.45 1075.2 Q137.663 1077.1 135.556 1079.69 Q134.978 1080.36 131.876 1083.58 Q128.774 1086.78 123.126 1092.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M129.862 722.006 L118.056 740.455 L129.862 740.455 L129.862 722.006 M128.635 717.932 L134.515 717.932 L134.515 740.455 L139.445 740.455 L139.445 744.344 L134.515 744.344 L134.515 752.492 L129.862 752.492 L129.862 744.344 L114.26 744.344 L114.26 739.83 L128.635 717.932 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M127.917 389.365 Q124.769 389.365 122.918 391.517 Q121.089 393.67 121.089 397.42 Q121.089 401.147 122.918 403.323 Q124.769 405.476 127.917 405.476 Q131.066 405.476 132.894 403.323 Q134.746 401.147 134.746 397.42 Q134.746 393.67 132.894 391.517 Q131.066 389.365 127.917 389.365 M137.2 374.712 L137.2 378.971 Q135.441 378.138 133.635 377.698 Q131.853 377.258 130.093 377.258 Q125.464 377.258 123.01 380.383 Q120.58 383.508 120.232 389.828 Q121.598 387.814 123.658 386.749 Q125.718 385.661 128.195 385.661 Q133.404 385.661 136.413 388.832 Q139.445 391.98 139.445 397.42 Q139.445 402.744 136.297 405.962 Q133.149 409.179 127.917 409.179 Q121.922 409.179 118.751 404.596 Q115.58 399.99 115.58 391.263 Q115.58 383.068 119.468 378.207 Q123.357 373.323 129.908 373.323 Q131.667 373.323 133.45 373.67 Q135.255 374.017 137.2 374.712 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M127.593 48.1121 Q124.26 48.1121 122.339 49.8945 Q120.441 51.6769 120.441 54.8019 Q120.441 57.9269 122.339 59.7093 Q124.26 61.4917 127.593 61.4917 Q130.927 61.4917 132.848 59.7093 Q134.769 57.9038 134.769 54.8019 Q134.769 51.6769 132.848 49.8945 Q130.95 48.1121 127.593 48.1121 M122.918 46.1214 Q119.908 45.3807 118.218 43.3205 Q116.552 41.2603 116.552 38.2974 Q116.552 34.1539 119.492 31.7465 Q122.455 29.3391 127.593 29.3391 Q132.755 29.3391 135.695 31.7465 Q138.635 34.1539 138.635 38.2974 Q138.635 41.2603 136.945 43.3205 Q135.279 45.3807 132.292 46.1214 Q135.672 46.9084 137.547 49.2001 Q139.445 51.4918 139.445 54.8019 Q139.445 59.825 136.367 62.5102 Q133.311 65.1954 127.593 65.1954 Q121.876 65.1954 118.797 62.5102 Q115.742 59.825 115.742 54.8019 Q115.742 51.4918 117.64 49.2001 Q119.538 46.9084 122.918 46.1214 M121.205 38.7372 Q121.205 41.4224 122.871 42.927 Q124.561 44.4316 127.593 44.4316 Q130.603 44.4316 132.292 42.927 Q134.005 41.4224 134.005 38.7372 Q134.005 36.052 132.292 34.5474 Q130.603 33.0428 127.593 33.0428 Q124.561 33.0428 122.871 34.5474 Q121.205 36.052 121.205 38.7372 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M16.4842 1086.54 L16.4842 1077.88 L56.238 1056.81 L16.4842 1056.81 L16.4842 1050.57 L64.0042 1050.57 L64.0042 1059.23 L24.2503 1080.3 L64.0042 1080.3 L64.0042 1086.54 L16.4842 1086.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M49.9359 1038.63 L28.3562 1038.63 L28.3562 1032.78 L49.7131 1032.78 Q54.7739 1032.78 57.3202 1030.8 Q59.8346 1028.83 59.8346 1024.88 Q59.8346 1020.14 56.8109 1017.4 Q53.7872 1014.63 48.5673 1014.63 L28.3562 1014.63 L28.3562 1008.78 L64.0042 1008.78 L64.0042 1014.63 L58.5296 1014.63 Q61.7762 1016.77 63.3676 1019.6 Q64.9272 1022.4 64.9272 1026.12 Q64.9272 1032.27 61.1078 1035.45 Q57.2883 1038.63 49.9359 1038.63 M27.4968 1023.9 L27.4968 1023.9 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M35.1993 968.961 Q31.2526 966.765 29.3747 963.709 Q27.4968 960.653 27.4968 956.516 Q27.4968 950.946 31.4117 947.922 Q35.2948 944.898 42.4881 944.898 L64.0042 944.898 L64.0042 950.787 L42.679 950.787 Q37.5546 950.787 35.072 952.601 Q32.5894 954.415 32.5894 958.139 Q32.5894 962.69 35.6131 965.332 Q38.6368 967.974 43.8567 967.974 L64.0042 967.974 L64.0042 973.862 L42.679 973.862 Q37.5228 973.862 35.072 975.677 Q32.5894 977.491 32.5894 981.278 Q32.5894 985.766 35.6449 988.408 Q38.6686 991.05 43.8567 991.05 L64.0042 991.05 L64.0042 996.938 L28.3562 996.938 L28.3562 991.05 L33.8944 991.05 Q30.616 989.045 29.0564 986.244 Q27.4968 983.443 27.4968 979.591 Q27.4968 975.708 29.4702 973.003 Q31.4436 970.266 35.1993 968.961 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M46.212 907.627 Q39.7508 907.627 36.0905 910.301 Q32.3984 912.942 32.3984 917.589 Q32.3984 922.236 36.0905 924.91 Q39.7508 927.552 46.212 927.552 Q52.6732 927.552 56.3653 924.91 Q60.0256 922.236 60.0256 917.589 Q60.0256 912.942 56.3653 910.301 Q52.6732 907.627 46.212 907.627 M33.7671 927.552 Q30.5842 925.706 29.0564 922.905 Q27.4968 920.072 27.4968 916.157 Q27.4968 909.664 32.6531 905.622 Q37.8093 901.548 46.212 901.548 Q54.6147 901.548 59.771 905.622 Q64.9272 909.664 64.9272 916.157 Q64.9272 920.072 63.3994 922.905 Q61.8398 925.706 58.657 927.552 L64.0042 927.552 L64.0042 933.44 L14.479 933.44 L14.479 927.552 L33.7671 927.552 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M44.7161 861.348 L47.5806 861.348 L47.5806 888.275 Q53.6281 887.893 56.8109 884.647 Q59.9619 881.369 59.9619 875.544 Q59.9619 872.17 59.1344 869.019 Q58.3069 865.836 56.6518 862.717 L62.1899 862.717 Q63.5267 865.868 64.227 869.178 Q64.9272 872.488 64.9272 875.894 Q64.9272 884.424 59.9619 889.421 Q54.9967 894.386 46.5303 894.386 Q37.7774 894.386 32.6531 889.676 Q27.4968 884.933 27.4968 876.913 Q27.4968 869.719 32.1438 865.55 Q36.7589 861.348 44.7161 861.348 M42.9973 867.205 Q38.1912 867.269 35.3266 869.91 Q32.4621 872.52 32.4621 876.849 Q32.4621 881.751 35.2312 884.711 Q38.0002 887.639 43.0292 888.084 L42.9973 867.205 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M33.8307 831.08 Q33.2578 832.066 33.0032 833.244 Q32.7167 834.39 32.7167 835.79 Q32.7167 840.755 35.9632 843.429 Q39.1779 846.071 45.2253 846.071 L64.0042 846.071 L64.0042 851.959 L28.3562 851.959 L28.3562 846.071 L33.8944 846.071 Q30.6479 844.225 29.0883 841.265 Q27.4968 838.305 27.4968 834.071 Q27.4968 833.467 27.5923 832.735 Q27.656 832.003 27.8151 831.111 L33.8307 831.08 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M32.4621 790.403 Q32.4621 795.113 36.1542 797.851 Q39.8145 800.588 46.212 800.588 Q52.6095 800.588 56.3017 797.882 Q59.9619 795.145 59.9619 790.403 Q59.9619 785.724 56.2698 782.987 Q52.5777 780.249 46.212 780.249 Q39.8781 780.249 36.186 782.987 Q32.4621 785.724 32.4621 790.403 M27.4968 790.403 Q27.4968 782.764 32.4621 778.403 Q37.4273 774.043 46.212 774.043 Q54.9649 774.043 59.9619 778.403 Q64.9272 782.764 64.9272 790.403 Q64.9272 798.073 59.9619 802.434 Q54.9649 806.763 46.212 806.763 Q37.4273 806.763 32.4621 802.434 Q27.4968 798.073 27.4968 790.403 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M14.479 746.288 L19.3487 746.288 L19.3487 751.89 Q19.3487 755.041 20.6219 756.282 Q21.895 757.492 25.2052 757.492 L28.3562 757.492 L28.3562 747.848 L32.9077 747.848 L32.9077 757.492 L64.0042 757.492 L64.0042 763.38 L32.9077 763.38 L32.9077 768.982 L28.3562 768.982 L28.3562 763.38 L25.8736 763.38 Q19.9216 763.38 17.2162 760.611 Q14.479 757.842 14.479 751.826 L14.479 746.288 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M29.7248 695.013 L35.1993 695.013 Q33.8307 697.495 33.1623 700.01 Q32.4621 702.492 32.4621 705.039 Q32.4621 710.736 36.0905 713.887 Q39.6872 717.038 46.212 717.038 Q52.7369 717.038 56.3653 713.887 Q59.9619 710.736 59.9619 705.039 Q59.9619 702.492 59.2935 700.01 Q58.5933 697.495 57.2247 695.013 L62.6355 695.013 Q63.7814 697.463 64.3543 700.105 Q64.9272 702.715 64.9272 705.675 Q64.9272 713.728 59.8664 718.47 Q54.8057 723.213 46.212 723.213 Q37.491 723.213 32.4939 718.438 Q27.4968 713.632 27.4968 705.293 Q27.4968 702.588 28.0697 700.01 Q28.6108 697.431 29.7248 695.013 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M42.4881 655.195 L64.0042 655.195 L64.0042 661.051 L42.679 661.051 Q37.6183 661.051 35.1038 663.025 Q32.5894 664.998 32.5894 668.945 Q32.5894 673.687 35.6131 676.425 Q38.6368 679.162 43.8567 679.162 L64.0042 679.162 L64.0042 685.05 L14.479 685.05 L14.479 679.162 L33.8944 679.162 Q30.6797 677.061 29.0883 674.228 Q27.4968 671.364 27.4968 667.64 Q27.4968 661.497 31.3163 658.346 Q35.1038 655.195 42.4881 655.195 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M46.0847 627.313 Q46.0847 634.411 47.7079 637.148 Q49.3312 639.886 53.2461 639.886 Q56.3653 639.886 58.2114 637.848 Q60.0256 635.78 60.0256 632.247 Q60.0256 627.377 56.5881 624.449 Q53.1188 621.489 47.3897 621.489 L46.0847 621.489 L46.0847 627.313 M43.6657 615.632 L64.0042 615.632 L64.0042 621.489 L58.5933 621.489 Q61.8398 623.494 63.3994 626.486 Q64.9272 629.478 64.9272 633.806 Q64.9272 639.281 61.8716 642.527 Q58.7843 645.742 53.6281 645.742 Q47.6125 645.742 44.5569 641.732 Q41.5014 637.689 41.5014 629.7 L41.5014 621.489 L40.9285 621.489 Q36.8862 621.489 34.6901 624.162 Q32.4621 626.804 32.4621 631.61 Q32.4621 634.666 33.1941 637.562 Q33.9262 640.458 35.3903 643.132 L29.9795 643.132 Q28.7381 639.917 28.1334 636.894 Q27.4968 633.87 27.4968 631.005 Q27.4968 623.271 31.5072 619.452 Q35.5176 615.632 43.6657 615.632 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M33.8307 582.912 Q33.2578 583.899 33.0032 585.077 Q32.7167 586.223 32.7167 587.623 Q32.7167 592.588 35.9632 595.262 Q39.1779 597.904 45.2253 597.904 L64.0042 597.904 L64.0042 603.792 L28.3562 603.792 L28.3562 597.904 L33.8944 597.904 Q30.6479 596.058 29.0883 593.098 Q27.4968 590.138 27.4968 585.904 Q27.4968 585.3 27.5923 584.568 Q27.656 583.835 27.8151 582.944 L33.8307 582.912 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M46.0847 560.569 Q46.0847 567.667 47.7079 570.404 Q49.3312 573.141 53.2461 573.141 Q56.3653 573.141 58.2114 571.104 Q60.0256 569.035 60.0256 565.502 Q60.0256 560.632 56.5881 557.704 Q53.1188 554.744 47.3897 554.744 L46.0847 554.744 L46.0847 560.569 M43.6657 548.888 L64.0042 548.888 L64.0042 554.744 L58.5933 554.744 Q61.8398 556.749 63.3994 559.741 Q64.9272 562.733 64.9272 567.062 Q64.9272 572.536 61.8716 575.783 Q58.7843 578.998 53.6281 578.998 Q47.6125 578.998 44.5569 574.987 Q41.5014 570.945 41.5014 562.956 L41.5014 554.744 L40.9285 554.744 Q36.8862 554.744 34.6901 557.418 Q32.4621 560.06 32.4621 564.866 Q32.4621 567.921 33.1941 570.818 Q33.9262 573.714 35.3903 576.388 L29.9795 576.388 Q28.7381 573.173 28.1334 570.149 Q27.4968 567.125 27.4968 564.261 Q27.4968 556.527 31.5072 552.707 Q35.5176 548.888 43.6657 548.888 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M29.7248 511.171 L35.1993 511.171 Q33.8307 513.654 33.1623 516.168 Q32.4621 518.651 32.4621 521.197 Q32.4621 526.894 36.0905 530.045 Q39.6872 533.196 46.212 533.196 Q52.7369 533.196 56.3653 530.045 Q59.9619 526.894 59.9619 521.197 Q59.9619 518.651 59.2935 516.168 Q58.5933 513.654 57.2247 511.171 L62.6355 511.171 Q63.7814 513.622 64.3543 516.263 Q64.9272 518.873 64.9272 521.833 Q64.9272 529.886 59.8664 534.629 Q54.8057 539.371 46.212 539.371 Q37.491 539.371 32.4939 534.597 Q27.4968 529.791 27.4968 521.452 Q27.4968 518.746 28.0697 516.168 Q28.6108 513.59 29.7248 511.171 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M18.2347 495.193 L28.3562 495.193 L28.3562 483.13 L32.9077 483.13 L32.9077 495.193 L52.2594 495.193 Q56.6199 495.193 57.8613 494.015 Q59.1026 492.806 59.1026 489.146 L59.1026 483.13 L64.0042 483.13 L64.0042 489.146 Q64.0042 495.925 61.4897 498.503 Q58.9434 501.081 52.2594 501.081 L32.9077 501.081 L32.9077 505.378 L28.3562 505.378 L28.3562 501.081 L18.2347 501.081 L18.2347 495.193 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M44.7161 444.936 L47.5806 444.936 L47.5806 471.863 Q53.6281 471.481 56.8109 468.234 Q59.9619 464.956 59.9619 459.131 Q59.9619 455.757 59.1344 452.606 Q58.3069 449.424 56.6518 446.304 L62.1899 446.304 Q63.5267 449.455 64.227 452.766 Q64.9272 456.076 64.9272 459.481 Q64.9272 468.011 59.9619 473.009 Q54.9967 477.974 46.5303 477.974 Q37.7774 477.974 32.6531 473.263 Q27.4968 468.521 27.4968 460.5 Q27.4968 453.307 32.1438 449.137 Q36.7589 444.936 44.7161 444.936 M42.9973 450.792 Q38.1912 450.856 35.3266 453.498 Q32.4621 456.108 32.4621 460.436 Q32.4621 465.338 35.2312 468.298 Q38.0002 471.226 43.0292 471.672 L42.9973 450.792 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M33.8307 414.667 Q33.2578 415.653 33.0032 416.831 Q32.7167 417.977 32.7167 419.377 Q32.7167 424.343 35.9632 427.016 Q39.1779 429.658 45.2253 429.658 L64.0042 429.658 L64.0042 435.546 L28.3562 435.546 L28.3562 429.658 L33.8944 429.658 Q30.6479 427.812 29.0883 424.852 Q27.4968 421.892 27.4968 417.659 Q27.4968 417.054 27.5923 416.322 Q27.656 415.59 27.8151 414.699 L33.8307 414.667 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip910)" d="M29.4065 385.798 L34.9447 385.798 Q33.6716 388.281 33.035 390.955 Q32.3984 393.628 32.3984 396.493 Q32.3984 400.853 33.7352 403.049 Q35.072 405.214 37.7456 405.214 Q39.7826 405.214 40.9603 403.654 Q42.1061 402.095 43.1565 397.384 L43.6021 395.379 Q44.9389 389.14 47.3897 386.53 Q49.8086 383.889 54.1691 383.889 Q59.1344 383.889 62.0308 387.835 Q64.9272 391.75 64.9272 398.625 Q64.9272 401.49 64.3543 404.609 Q63.8132 407.696 62.6992 411.134 L56.6518 411.134 Q58.3387 407.887 59.198 404.736 Q60.0256 401.585 60.0256 398.498 Q60.0256 394.36 58.6251 392.132 Q57.1929 389.904 54.6147 389.904 Q52.2276 389.904 50.9545 391.527 Q49.6813 393.119 48.5037 398.562 L48.0262 400.599 Q46.8804 406.041 44.5251 408.46 Q42.138 410.879 38.0002 410.879 Q32.9713 410.879 30.2341 407.314 Q27.4968 403.75 27.4968 397.193 Q27.4968 393.946 27.9743 391.082 Q28.4517 388.217 29.4065 385.798 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip912)" d="M295.201 47.2441 L295.201 1423.18 L562.484 1423.18 L562.484 47.2441 L295.201 47.2441 L295.201 47.2441  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="295.201,47.2441 295.201,1423.18 562.484,1423.18 562.484,47.2441 295.201,47.2441 "/>
+<path clip-path="url(#clip912)" d="M629.304 563.22 L629.304 1423.18 L896.587 1423.18 L896.587 563.22 L629.304 563.22 L629.304 563.22  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="629.304,563.22 629.304,1423.18 896.587,1423.18 896.587,563.22 629.304,563.22 "/>
+<path clip-path="url(#clip912)" d="M963.408 1079.2 L963.408 1423.18 L1230.69 1423.18 L1230.69 1079.2 L963.408 1079.2 L963.408 1079.2  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="963.408,1079.2 963.408,1423.18 1230.69,1423.18 1230.69,1079.2 963.408,1079.2 "/>
+<path clip-path="url(#clip912)" d="M1297.51 1251.19 L1297.51 1423.18 L1564.79 1423.18 L1564.79 1251.19 L1297.51 1251.19 L1297.51 1251.19  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1297.51,1251.19 1297.51,1423.18 1564.79,1423.18 1564.79,1251.19 1297.51,1251.19 "/>
+<path clip-path="url(#clip912)" d="M1631.61 1251.19 L1631.61 1423.18 L1898.9 1423.18 L1898.9 1251.19 L1631.61 1251.19 L1631.61 1251.19  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1631.61,1251.19 1631.61,1423.18 1898.9,1423.18 1898.9,1251.19 1631.61,1251.19 "/>
+<path clip-path="url(#clip912)" d="M1965.72 1251.19 L1965.72 1423.18 L2233 1423.18 L2233 1251.19 L1965.72 1251.19 L1965.72 1251.19  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip912)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1965.72,1251.19 1965.72,1423.18 2233,1423.18 2233,1251.19 1965.72,1251.19 "/>
+<circle clip-path="url(#clip912)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="428.842" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip912)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="762.946" cy="563.22" r="2"/>
+<circle clip-path="url(#clip912)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1097.05" cy="1079.2" r="2"/>
+<circle clip-path="url(#clip912)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1431.15" cy="1251.19" r="2"/>
+<circle clip-path="url(#clip912)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1765.26" cy="1251.19" r="2"/>
+<circle clip-path="url(#clip912)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2099.36" cy="1251.19" r="2"/>
+</svg>
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/index.html b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/index.html
new file mode 100644
index 00000000000..938da741dcc
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_data_and_plotting/index.html
@@ -0,0 +1,238 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Getting started with data and plotting · JuMP</title><meta name="title" content="Getting started with data and plotting · JuMP"/><meta property="og:title" content="Getting started with data and plotting · JuMP"/><meta property="twitter:title" content="Getting started with data and plotting · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li class="is-active"><a class="tocitem" href>Getting started with data and plotting</a><ul class="internal"><li><a class="tocitem" href="#Where-to-get-help"><span>Where to get help</span></a></li><li><a class="tocitem" href="#Preliminaries"><span>Preliminaries</span></a></li><li><a class="tocitem" href="#DataFrame-basics"><span>DataFrame basics</span></a></li><li><a class="tocitem" href="#Working-with-DataFrames"><span>Working with DataFrames</span></a></li><li><a class="tocitem" href="#Example:-the-passport-problem"><span>Example: the passport problem</span></a></li></ul></li><li><a class="tocitem" href="../debugging/">Debugging</a></li><li><a class="tocitem" href="../design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Getting started with data and plotting</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Getting started with data and plotting</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/getting_started_with_data_and_plotting.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Getting-started-with-data-and-plotting"><a class="docs-heading-anchor" href="#Getting-started-with-data-and-plotting">Getting started with data and plotting</a><a id="Getting-started-with-data-and-plotting-1"></a><a class="docs-heading-anchor-permalink" href="#Getting-started-with-data-and-plotting" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../getting_started_with_data_and_plotting.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>In this tutorial we will learn how to read tabular data into Julia, and some of the basics of plotting.</p><p>If you&#39;re new to Julia, start by reading <a href="../getting_started_with_julia/#Getting-started-with-Julia">Getting started with Julia</a> and <a href="../getting_started_with_JuMP/#Getting-started-with-JuMP">Getting started with JuMP</a> first.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>There are multiple ways to read the same kind of data into Julia. This tutorial focuses on DataFrames.jl because it provides the ecosystem to work with most of the required file types in a straightforward manner.</p></div></div><p>Before we get started, we need this constant to point to where the data files are.</p><pre><code class="language-julia hljs">import JuMP
+const DATA_DIR = joinpath(
+    dirname(pathof(JuMP)),
+    joinpath(&quot;..&quot;, &quot;docs&quot;, &quot;src&quot;, &quot;tutorials&quot;, &quot;getting_started&quot;, &quot;data&quot;),
+);</code></pre><h2 id="Where-to-get-help"><a class="docs-heading-anchor" href="#Where-to-get-help">Where to get help</a><a id="Where-to-get-help-1"></a><a class="docs-heading-anchor-permalink" href="#Where-to-get-help" title="Permalink"></a></h2><p>Read the documentation</p><ul><li>Plots.jl: <a href="http://docs.juliaplots.org/latest/">http://docs.juliaplots.org/latest/</a></li><li>CSV.jl: <a href="http://csv.juliadata.org/stable">http://csv.juliadata.org/stable</a></li><li>DataFrames.jl: <a href="https://dataframes.juliadata.org/stable/">https://dataframes.juliadata.org/stable/</a></li></ul><h2 id="Preliminaries"><a class="docs-heading-anchor" href="#Preliminaries">Preliminaries</a><a id="Preliminaries-1"></a><a class="docs-heading-anchor-permalink" href="#Preliminaries" title="Permalink"></a></h2><p>To get started, we need to install some packages.</p><h3 id="DataFrames.jl"><a class="docs-heading-anchor" href="#DataFrames.jl">DataFrames.jl</a><a id="DataFrames.jl-1"></a><a class="docs-heading-anchor-permalink" href="#DataFrames.jl" title="Permalink"></a></h3><p>The <code>DataFrames</code> package provides a set of tools for working with tabular data. It is available through the Julia package manager.</p><pre><code class="language-julia hljs">using Pkg
+Pkg.add(&quot;DataFrames&quot;)</code></pre><pre><code class="language-julia hljs">import DataFrames</code></pre><div class="admonition is-info"><header class="admonition-header">What is a DataFrame?</header><div class="admonition-body"><p>A DataFrame is a data structure like a table or spreadsheet. You can use it for storing and exploring a set of related data values. Think of it as a smarter array for holding tabular data.</p></div></div><h3 id="Plots.jl"><a class="docs-heading-anchor" href="#Plots.jl">Plots.jl</a><a id="Plots.jl-1"></a><a class="docs-heading-anchor-permalink" href="#Plots.jl" title="Permalink"></a></h3><p>The <code>Plots</code> package provides a set of tools for plotting. It is available through the Julia package manager.</p><pre><code class="language-julia hljs">using Pkg
+Pkg.add(&quot;Plots&quot;)</code></pre><pre><code class="language-julia hljs">import Plots</code></pre><h3 id="CSV-.jl"><a class="docs-heading-anchor" href="#CSV-.jl">CSV .jl</a><a id="CSV-.jl-1"></a><a class="docs-heading-anchor-permalink" href="#CSV-.jl" title="Permalink"></a></h3><p>CSV and other delimited text files can be read by the CSV.jl package.</p><pre><code class="language-julia hljs">Pkg.add(&quot;CSV&quot;)</code></pre><pre><code class="language-julia hljs">import CSV</code></pre><h2 id="DataFrame-basics"><a class="docs-heading-anchor" href="#DataFrame-basics">DataFrame basics</a><a id="DataFrame-basics-1"></a><a class="docs-heading-anchor-permalink" href="#DataFrame-basics" title="Permalink"></a></h2><p>To read a CSV file into a DataFrame, we use the <code>CSV.read</code> function.</p><pre><code class="language-julia hljs">csv_df = CSV.read(joinpath(DATA_DIR, &quot;StarWars.csv&quot;), DataFrames.DataFrame)</code></pre><div><div style = "float: left;"><span>20×13 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Gender</th><th style = "text-align: left;">Height</th><th style = "text-align: left;">Weight</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">Haircolor</th><th style = "text-align: left;">Skincolor</th><th style = "text-align: left;">Homeland</th><th style = "text-align: left;">Born</th><th style = "text-align: left;">Died</th><th style = "text-align: left;">Jedi</th><th style = "text-align: left;">Species</th><th style = "text-align: left;">Weapon</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Anakin Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.88</td><td style = "text-align: left;">84</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">41.9BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Padme Amidala</td><td style = "text-align: left;">female</td><td style = "text-align: right;">1.65</td><td style = "text-align: left;">45</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">46BBY</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Luke Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.72</td><td style = "text-align: left;">77</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Leia Skywalker</td><td style = "text-align: left;">female</td><td style = "text-align: right;">1.5</td><td style = "text-align: left;">49</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Alderaan</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Qui-Gon Jinn</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.93</td><td style = "text-align: left;">88.5</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">unk_planet</td><td style = "text-align: left;">92BBY</td><td style = "text-align: left;">32BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">Obi-Wan Kenobi</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.82</td><td style = "text-align: left;">77</td><td style = "text-align: left;">bluegray</td><td style = "text-align: left;">auburn</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Stewjon</td><td style = "text-align: left;">57BBY</td><td style = "text-align: left;">0BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">Han Solo</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.8</td><td style = "text-align: left;">80</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Corellia</td><td style = "text-align: left;">29BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">Sheev Palpatine</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.73</td><td style = "text-align: left;">75</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">red</td><td style = "text-align: left;">pale</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">82BBY</td><td style = "text-align: left;">10ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">force-lightning</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">R2-D2</td><td style = "text-align: left;">male</td><td style = "text-align: right;">0.96</td><td style = "text-align: left;">32</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">33BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">droid</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">C-3PO</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.67</td><td style = "text-align: left;">75</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">112BBY</td><td style = "text-align: left;">3ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">droid</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">Yoda</td><td style = "text-align: left;">male</td><td style = "text-align: right;">0.66</td><td style = "text-align: left;">17</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">green</td><td style = "text-align: left;">unk_planet</td><td style = "text-align: left;">896BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">yoda</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">Darth Maul</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.75</td><td style = "text-align: left;">80</td><td style = "text-align: left;">yellow</td><td style = "text-align: left;">none</td><td style = "text-align: left;">red</td><td style = "text-align: left;">Dathomir</td><td style = "text-align: left;">54BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">dathomirian</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">Dooku</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.93</td><td style = "text-align: left;">86</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Serenno</td><td style = "text-align: left;">102BBY</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">Chewbacca</td><td style = "text-align: left;">male</td><td style = "text-align: right;">2.28</td><td style = "text-align: left;">112</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">Kashyyyk</td><td style = "text-align: left;">200BBY</td><td style = "text-align: left;">25ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">wookiee</td><td style = "text-align: left;">bowcaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">Jabba</td><td style = "text-align: left;">male</td><td style = "text-align: right;">3.9</td><td style = "text-align: left;">NA</td><td style = "text-align: left;">yellow</td><td style = "text-align: left;">none</td><td style = "text-align: left;">tan-green</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">hutt</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">Lando Calrissian</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.78</td><td style = "text-align: left;">79</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">blank</td><td style = "text-align: left;">dark</td><td style = "text-align: left;">Socorro</td><td style = "text-align: left;">31BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">Boba Fett</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: left;">78</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">black</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">Kamino</td><td style = "text-align: left;">31.5BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">Jango Fett</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: left;">79</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">black</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">ConcordDawn</td><td style = "text-align: left;">66BBY</td><td style = "text-align: left;">22BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">Grievous</td><td style = "text-align: left;">male</td><td style = "text-align: right;">2.16</td><td style = "text-align: left;">159</td><td style = "text-align: left;">gold</td><td style = "text-align: left;">black</td><td style = "text-align: left;">orange</td><td style = "text-align: left;">Kalee</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">kaleesh</td><td style = "text-align: left;">slugthrower</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">Chief Chirpa</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.0</td><td style = "text-align: left;">50</td><td style = "text-align: left;">black</td><td style = "text-align: left;">gray</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">Endor</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">ewok</td><td style = "text-align: left;">spear</td></tr></tbody></table></div><p>Let&#39;s try plotting some of this data</p><pre><code class="language-julia hljs">Plots.scatter(
+    csv_df.Weight,
+    csv_df.Height;
+    xlabel = &quot;Weight&quot;,
+    ylabel = &quot;Height&quot;,
+)</code></pre><img src="2af4f5d4.svg" alt="Example block output"/><p>That doesn&#39;t look right. What happened? If you look at the dataframe above, it read <code>Weight</code> in as a <code>String</code> column because there are &quot;NA&quot; fields. Let&#39;s correct that, by telling CSV to consider &quot;NA&quot; as <code>missing</code>.</p><pre><code class="language-julia hljs">csv_df = CSV.read(
+    joinpath(DATA_DIR, &quot;StarWars.csv&quot;),
+    DataFrames.DataFrame;
+    missingstring = &quot;NA&quot;,
+)</code></pre><div><div style = "float: left;"><span>20×13 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Gender</th><th style = "text-align: left;">Height</th><th style = "text-align: left;">Weight</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">Haircolor</th><th style = "text-align: left;">Skincolor</th><th style = "text-align: left;">Homeland</th><th style = "text-align: left;">Born</th><th style = "text-align: left;">Died</th><th style = "text-align: left;">Jedi</th><th style = "text-align: left;">Species</th><th style = "text-align: left;">Weapon</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Union{Missing, InlineStrings.String7}" style = "text-align: left;">String7</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Anakin Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.88</td><td style = "text-align: right;">84.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">41.9BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Padme Amidala</td><td style = "text-align: left;">female</td><td style = "text-align: right;">1.65</td><td style = "text-align: right;">45.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">46BBY</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Luke Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.72</td><td style = "text-align: right;">77.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Leia Skywalker</td><td style = "text-align: left;">female</td><td style = "text-align: right;">1.5</td><td style = "text-align: right;">49.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Alderaan</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Qui-Gon Jinn</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.93</td><td style = "text-align: right;">88.5</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">unk_planet</td><td style = "text-align: left;">92BBY</td><td style = "text-align: left;">32BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">Obi-Wan Kenobi</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.82</td><td style = "text-align: right;">77.0</td><td style = "text-align: left;">bluegray</td><td style = "text-align: left;">auburn</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Stewjon</td><td style = "text-align: left;">57BBY</td><td style = "text-align: left;">0BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">Han Solo</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.8</td><td style = "text-align: right;">80.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Corellia</td><td style = "text-align: left;">29BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">Sheev Palpatine</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.73</td><td style = "text-align: right;">75.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">red</td><td style = "text-align: left;">pale</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">82BBY</td><td style = "text-align: left;">10ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">force-lightning</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">R2-D2</td><td style = "text-align: left;">male</td><td style = "text-align: right;">0.96</td><td style = "text-align: right;">32.0</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">33BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">droid</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">C-3PO</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.67</td><td style = "text-align: right;">75.0</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">112BBY</td><td style = "text-align: left;">3ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">droid</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">Yoda</td><td style = "text-align: left;">male</td><td style = "text-align: right;">0.66</td><td style = "text-align: right;">17.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">green</td><td style = "text-align: left;">unk_planet</td><td style = "text-align: left;">896BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">yoda</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">Darth Maul</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.75</td><td style = "text-align: right;">80.0</td><td style = "text-align: left;">yellow</td><td style = "text-align: left;">none</td><td style = "text-align: left;">red</td><td style = "text-align: left;">Dathomir</td><td style = "text-align: left;">54BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">dathomirian</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">Dooku</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.93</td><td style = "text-align: right;">86.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Serenno</td><td style = "text-align: left;">102BBY</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">Chewbacca</td><td style = "text-align: left;">male</td><td style = "text-align: right;">2.28</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">Kashyyyk</td><td style = "text-align: left;">200BBY</td><td style = "text-align: left;">25ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">wookiee</td><td style = "text-align: left;">bowcaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">Jabba</td><td style = "text-align: left;">male</td><td style = "text-align: right;">3.9</td><td style = "font-style: italic; text-align: right;">missing</td><td style = "text-align: left;">yellow</td><td style = "text-align: left;">none</td><td style = "text-align: left;">tan-green</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">hutt</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">Lando Calrissian</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.78</td><td style = "text-align: right;">79.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">blank</td><td style = "text-align: left;">dark</td><td style = "text-align: left;">Socorro</td><td style = "text-align: left;">31BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">Boba Fett</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: right;">78.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">black</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">Kamino</td><td style = "text-align: left;">31.5BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">Jango Fett</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: right;">79.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">black</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">ConcordDawn</td><td style = "text-align: left;">66BBY</td><td style = "text-align: left;">22BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">Grievous</td><td style = "text-align: left;">male</td><td style = "text-align: right;">2.16</td><td style = "text-align: right;">159.0</td><td style = "text-align: left;">gold</td><td style = "text-align: left;">black</td><td style = "text-align: left;">orange</td><td style = "text-align: left;">Kalee</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">kaleesh</td><td style = "text-align: left;">slugthrower</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">Chief Chirpa</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">50.0</td><td style = "text-align: left;">black</td><td style = "text-align: left;">gray</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">Endor</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">ewok</td><td style = "text-align: left;">spear</td></tr></tbody></table></div><p>Then let&#39;s re-plot our data</p><pre><code class="language-julia hljs">Plots.scatter(
+    csv_df.Weight,
+    csv_df.Height;
+    title = &quot;Height vs Weight of StarWars characters&quot;,
+    xlabel = &quot;Weight&quot;,
+    ylabel = &quot;Height&quot;,
+    label = false,
+    ylims = (0, 3),
+)</code></pre><img src="0c9607c5.svg" alt="Example block output"/><p>That looks better.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Read the <a href="https://csv.juliadata.org/stable/">CSV documentation</a> for other parsing options.</p></div></div><p>DataFrames.jl supports manipulation using functions similar to pandas. For example, split the dataframe into groups based on eye-color:</p><pre><code class="language-julia hljs">by_eyecolor = DataFrames.groupby(csv_df, :Eyecolor)</code></pre><p><b>GroupedDataFrame with 7 groups based on key: Eyecolor</b></p><div><div style = "float: left;"><span>First Group (5 rows): Eyecolor = &quot;blue&quot;</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Gender</th><th style = "text-align: left;">Height</th><th style = "text-align: left;">Weight</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">Haircolor</th><th style = "text-align: left;">Skincolor</th><th style = "text-align: left;">Homeland</th><th style = "text-align: left;">Born</th><th style = "text-align: left;">Died</th><th style = "text-align: left;">Jedi</th><th style = "text-align: left;">Species</th><th style = "text-align: left;">Weapon</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Union{Missing, InlineStrings.String7}" style = "text-align: left;">String7</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Anakin Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.88</td><td style = "text-align: right;">84.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">41.9BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Luke Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.72</td><td style = "text-align: right;">77.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Qui-Gon Jinn</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.93</td><td style = "text-align: right;">88.5</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">unk_planet</td><td style = "text-align: left;">92BBY</td><td style = "text-align: left;">32BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Sheev Palpatine</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.73</td><td style = "text-align: right;">75.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">red</td><td style = "text-align: left;">pale</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">82BBY</td><td style = "text-align: left;">10ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">force-lightning</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Chewbacca</td><td style = "text-align: left;">male</td><td style = "text-align: right;">2.28</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">Kashyyyk</td><td style = "text-align: left;">200BBY</td><td style = "text-align: left;">25ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">wookiee</td><td style = "text-align: left;">bowcaster</td></tr></tbody></table></div><p>&vellip;</p><div><div style = "float: left;"><span>Last Group (1 row): Eyecolor = &quot;black&quot;</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Gender</th><th style = "text-align: left;">Height</th><th style = "text-align: left;">Weight</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">Haircolor</th><th style = "text-align: left;">Skincolor</th><th style = "text-align: left;">Homeland</th><th style = "text-align: left;">Born</th><th style = "text-align: left;">Died</th><th style = "text-align: left;">Jedi</th><th style = "text-align: left;">Species</th><th style = "text-align: left;">Weapon</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Union{Missing, InlineStrings.String7}" style = "text-align: left;">String7</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Chief Chirpa</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">50.0</td><td style = "text-align: left;">black</td><td style = "text-align: left;">gray</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">Endor</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">ewok</td><td style = "text-align: left;">spear</td></tr></tbody></table></div><p>Then recombine into a single dataframe based on a function operating over the split dataframes:</p><pre><code class="language-julia hljs">eyecolor_count = DataFrames.combine(by_eyecolor) do df
+    return DataFrames.nrow(df)
+end</code></pre><div><div style = "float: left;"><span>7×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">x1</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Int64" style = "text-align: left;">Int64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">blue</td><td style = "text-align: right;">5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">brown</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">bluegray</td><td style = "text-align: right;">1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: right;">2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">yellow</td><td style = "text-align: right;">2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">gold</td><td style = "text-align: right;">1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">black</td><td style = "text-align: right;">1</td></tr></tbody></table></div><p>We can rename columns:</p><pre><code class="language-julia hljs">DataFrames.rename!(eyecolor_count, :x1 =&gt; :count)</code></pre><div><div style = "float: left;"><span>7×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">count</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Int64" style = "text-align: left;">Int64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">blue</td><td style = "text-align: right;">5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">brown</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">bluegray</td><td style = "text-align: right;">1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: right;">2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">yellow</td><td style = "text-align: right;">2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">gold</td><td style = "text-align: right;">1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">black</td><td style = "text-align: right;">1</td></tr></tbody></table></div><p>Drop some missing rows:</p><pre><code class="language-julia hljs">DataFrames.dropmissing!(eyecolor_count, :Eyecolor)</code></pre><div><div style = "float: left;"><span>6×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">count</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "Int64" style = "text-align: left;">Int64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">blue</td><td style = "text-align: right;">5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">brown</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">bluegray</td><td style = "text-align: right;">1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">yellow</td><td style = "text-align: right;">2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">gold</td><td style = "text-align: right;">1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">black</td><td style = "text-align: right;">1</td></tr></tbody></table></div><p>Then we can visualize the data:</p><pre><code class="language-julia hljs">sort!(eyecolor_count, :count; rev = true)
+Plots.bar(
+    eyecolor_count.Eyecolor,
+    eyecolor_count.count;
+    xlabel = &quot;Eye color&quot;,
+    ylabel = &quot;Number of characters&quot;,
+    label = false,
+)</code></pre><img src="f8d9d8f6.svg" alt="Example block output"/><h3 id="Other-Delimited-Files"><a class="docs-heading-anchor" href="#Other-Delimited-Files">Other Delimited Files</a><a id="Other-Delimited-Files-1"></a><a class="docs-heading-anchor-permalink" href="#Other-Delimited-Files" title="Permalink"></a></h3><p>We can also use the <code>CSV.jl</code> package to read any other delimited text file format.</p><p>By default, CSV.File will try to detect a file&#39;s delimiter from the first 10 lines of the file.</p><p>Candidate delimiters include <code>&#39;,&#39;</code>, <code>&#39;\t&#39;</code>, <code>&#39; &#39;</code>, <code>&#39;|&#39;</code>, <code>&#39;;&#39;</code>, and <code>&#39;:&#39;</code>. If it can&#39;t auto-detect the delimiter, it will assume <code>&#39;,&#39;</code>.</p><p>Let&#39;s take the example of space separated data.</p><pre><code class="language-julia hljs">ss_df = CSV.read(joinpath(DATA_DIR, &quot;Cereal.txt&quot;), DataFrames.DataFrame)</code></pre><div><div style = "float: left;"><span>23×10 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Cups</th><th style = "text-align: left;">Calories</th><th style = "text-align: left;">Carbs</th><th style = "text-align: left;">Fat</th><th style = "text-align: left;">Fiber</th><th style = "text-align: left;">Potassium</th><th style = "text-align: left;">Protein</th><th style = "text-align: left;">Sodium</th><th style = "text-align: left;">Sugars</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">CapnCrunch</td><td style = "text-align: right;">0.75</td><td style = "text-align: right;">120</td><td style = "text-align: right;">12.0</td><td style = "text-align: right;">2</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">35</td><td style = "text-align: right;">1</td><td style = "text-align: right;">220</td><td style = "text-align: right;">12</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">CocoaPuffs</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">110</td><td style = "text-align: right;">12.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">55</td><td style = "text-align: right;">1</td><td style = "text-align: right;">180</td><td style = "text-align: right;">13</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Trix</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">110</td><td style = "text-align: right;">13.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">25</td><td style = "text-align: right;">1</td><td style = "text-align: right;">140</td><td style = "text-align: right;">12</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">AppleJacks</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">110</td><td style = "text-align: right;">11.0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">30</td><td style = "text-align: right;">2</td><td style = "text-align: right;">125</td><td style = "text-align: right;">14</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">CornChex</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">110</td><td style = "text-align: right;">22.0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">25</td><td style = "text-align: right;">2</td><td style = "text-align: right;">280</td><td style = "text-align: right;">3</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">CornFlakes</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">100</td><td style = "text-align: right;">21.0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">35</td><td style = "text-align: right;">2</td><td style = "text-align: right;">290</td><td style = "text-align: right;">2</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">Nut&amp;Honey</td><td style = "text-align: right;">0.67</td><td style = "text-align: right;">120</td><td style = "text-align: right;">15.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">40</td><td style = "text-align: right;">2</td><td style = "text-align: right;">190</td><td style = "text-align: right;">9</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">Smacks</td><td style = "text-align: right;">0.75</td><td style = "text-align: right;">110</td><td style = "text-align: right;">9.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">40</td><td style = "text-align: right;">2</td><td style = "text-align: right;">70</td><td style = "text-align: right;">15</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">MultiGrain</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">100</td><td style = "text-align: right;">15.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">2.0</td><td style = "text-align: right;">90</td><td style = "text-align: right;">2</td><td style = "text-align: right;">220</td><td style = "text-align: right;">6</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">CracklinOat</td><td style = "text-align: right;">0.5</td><td style = "text-align: right;">110</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">3</td><td style = "text-align: right;">4.0</td><td style = "text-align: right;">160</td><td style = "text-align: right;">3</td><td style = "text-align: right;">140</td><td style = "text-align: right;">7</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">GrapeNuts</td><td style = "text-align: right;">0.25</td><td style = "text-align: right;">110</td><td style = "text-align: right;">17.0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">90</td><td style = "text-align: right;">3</td><td style = "text-align: right;">179</td><td style = "text-align: right;">3</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">HoneyNutCheerios</td><td style = "text-align: right;">0.75</td><td style = "text-align: right;">110</td><td style = "text-align: right;">11.5</td><td style = "text-align: right;">1</td><td style = "text-align: right;">1.5</td><td style = "text-align: right;">90</td><td style = "text-align: right;">3</td><td style = "text-align: right;">250</td><td style = "text-align: right;">10</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">NutriGrain</td><td style = "text-align: right;">0.67</td><td style = "text-align: right;">140</td><td style = "text-align: right;">21.0</td><td style = "text-align: right;">2</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">130</td><td style = "text-align: right;">3</td><td style = "text-align: right;">220</td><td style = "text-align: right;">7</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">Product19</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">100</td><td style = "text-align: right;">20.0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">45</td><td style = "text-align: right;">3</td><td style = "text-align: right;">320</td><td style = "text-align: right;">3</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">TotalRaisinBran</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">140</td><td style = "text-align: right;">15.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">4.0</td><td style = "text-align: right;">230</td><td style = "text-align: right;">3</td><td style = "text-align: right;">190</td><td style = "text-align: right;">14</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">WheatChex</td><td style = "text-align: right;">0.67</td><td style = "text-align: right;">100</td><td style = "text-align: right;">17.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">115</td><td style = "text-align: right;">3</td><td style = "text-align: right;">230</td><td style = "text-align: right;">3</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">Oatmeal</td><td style = "text-align: right;">0.5</td><td style = "text-align: right;">130</td><td style = "text-align: right;">13.5</td><td style = "text-align: right;">2</td><td style = "text-align: right;">1.5</td><td style = "text-align: right;">120</td><td style = "text-align: right;">3</td><td style = "text-align: right;">170</td><td style = "text-align: right;">10</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">Life</td><td style = "text-align: right;">0.67</td><td style = "text-align: right;">100</td><td style = "text-align: right;">12.0</td><td style = "text-align: right;">2</td><td style = "text-align: right;">2.0</td><td style = "text-align: right;">95</td><td style = "text-align: right;">4</td><td style = "text-align: right;">150</td><td style = "text-align: right;">6</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">Maypo</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">100</td><td style = "text-align: right;">16.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">95</td><td style = "text-align: right;">4</td><td style = "text-align: right;">0</td><td style = "text-align: right;">3</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">QuakerOats</td><td style = "text-align: right;">0.5</td><td style = "text-align: right;">100</td><td style = "text-align: right;">14.0</td><td style = "text-align: right;">1</td><td style = "text-align: right;">2.0</td><td style = "text-align: right;">110</td><td style = "text-align: right;">4</td><td style = "text-align: right;">135</td><td style = "text-align: right;">6</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">21</td><td style = "text-align: left;">Muesli</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">150</td><td style = "text-align: right;">16.0</td><td style = "text-align: right;">3</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">170</td><td style = "text-align: right;">4</td><td style = "text-align: right;">150</td><td style = "text-align: right;">11</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">22</td><td style = "text-align: left;">Cheerios</td><td style = "text-align: right;">1.25</td><td style = "text-align: right;">110</td><td style = "text-align: right;">17.0</td><td style = "text-align: right;">2</td><td style = "text-align: right;">2.0</td><td style = "text-align: right;">105</td><td style = "text-align: right;">6</td><td style = "text-align: right;">290</td><td style = "text-align: right;">1</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">23</td><td style = "text-align: left;">SpecialK</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">110</td><td style = "text-align: right;">16.0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">55</td><td style = "text-align: right;">6</td><td style = "text-align: right;">230</td><td style = "text-align: right;">3</td></tr></tbody></table></div><p>We can also specify the delimiter as follows:</p><pre><code class="language-julia hljs">delim_df = CSV.read(
+    joinpath(DATA_DIR, &quot;Soccer.txt&quot;),
+    DataFrames.DataFrame;
+    delim = &quot;::&quot;,
+)</code></pre><div><div style = "float: left;"><span>20×7 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Team</th><th style = "text-align: left;">Played</th><th style = "text-align: left;">Wins</th><th style = "text-align: left;">Draws</th><th style = "text-align: left;">Losses</th><th style = "text-align: left;">Goals_for</th><th style = "text-align: left;">Goals_against</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Barcelona</td><td style = "text-align: right;">38</td><td style = "text-align: right;">30</td><td style = "text-align: right;">4</td><td style = "text-align: right;">4</td><td style = "text-align: left;">110 goals</td><td style = "text-align: left;">21 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Real Madrid</td><td style = "text-align: right;">38</td><td style = "text-align: right;">30</td><td style = "text-align: right;">2</td><td style = "text-align: right;">6</td><td style = "text-align: left;">118 goals</td><td style = "text-align: left;">38 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Atletico Madrid</td><td style = "text-align: right;">38</td><td style = "text-align: right;">23</td><td style = "text-align: right;">9</td><td style = "text-align: right;">6</td><td style = "text-align: left;">67 goals</td><td style = "text-align: left;">29 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Valencia</td><td style = "text-align: right;">38</td><td style = "text-align: right;">22</td><td style = "text-align: right;">11</td><td style = "text-align: right;">5</td><td style = "text-align: left;">70 goals</td><td style = "text-align: left;">32 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Seville</td><td style = "text-align: right;">38</td><td style = "text-align: right;">23</td><td style = "text-align: right;">7</td><td style = "text-align: right;">8</td><td style = "text-align: left;">71 goals</td><td style = "text-align: left;">45 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">Villarreal</td><td style = "text-align: right;">38</td><td style = "text-align: right;">16</td><td style = "text-align: right;">12</td><td style = "text-align: right;">10</td><td style = "text-align: left;">48 goals</td><td style = "text-align: left;">37 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">Athletic Bilbao</td><td style = "text-align: right;">38</td><td style = "text-align: right;">15</td><td style = "text-align: right;">10</td><td style = "text-align: right;">13</td><td style = "text-align: left;">42 goals</td><td style = "text-align: left;">41 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">Celta Vigo</td><td style = "text-align: right;">38</td><td style = "text-align: right;">13</td><td style = "text-align: right;">12</td><td style = "text-align: right;">13</td><td style = "text-align: left;">47 goals</td><td style = "text-align: left;">44 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">Malaga</td><td style = "text-align: right;">38</td><td style = "text-align: right;">14</td><td style = "text-align: right;">8</td><td style = "text-align: right;">16</td><td style = "text-align: left;">42 goals</td><td style = "text-align: left;">48 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">Espanyol</td><td style = "text-align: right;">38</td><td style = "text-align: right;">13</td><td style = "text-align: right;">10</td><td style = "text-align: right;">15</td><td style = "text-align: left;">47 goals</td><td style = "text-align: left;">51 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">Rayo Vallecano</td><td style = "text-align: right;">38</td><td style = "text-align: right;">15</td><td style = "text-align: right;">4</td><td style = "text-align: right;">19</td><td style = "text-align: left;">46 goals</td><td style = "text-align: left;">68 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">Real Sociedad</td><td style = "text-align: right;">38</td><td style = "text-align: right;">11</td><td style = "text-align: right;">13</td><td style = "text-align: right;">14</td><td style = "text-align: left;">44 goals</td><td style = "text-align: left;">51 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">Elche</td><td style = "text-align: right;">38</td><td style = "text-align: right;">11</td><td style = "text-align: right;">8</td><td style = "text-align: right;">19</td><td style = "text-align: left;">35 goals</td><td style = "text-align: left;">62 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">Levante</td><td style = "text-align: right;">38</td><td style = "text-align: right;">9</td><td style = "text-align: right;">10</td><td style = "text-align: right;">19</td><td style = "text-align: left;">34 goals</td><td style = "text-align: left;">67 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">Getafe</td><td style = "text-align: right;">38</td><td style = "text-align: right;">10</td><td style = "text-align: right;">7</td><td style = "text-align: right;">21</td><td style = "text-align: left;">33 goals</td><td style = "text-align: left;">64 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">Deportivo La Coruna</td><td style = "text-align: right;">38</td><td style = "text-align: right;">7</td><td style = "text-align: right;">14</td><td style = "text-align: right;">17</td><td style = "text-align: left;">35 goals</td><td style = "text-align: left;">60 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">Granada</td><td style = "text-align: right;">38</td><td style = "text-align: right;">7</td><td style = "text-align: right;">14</td><td style = "text-align: right;">17</td><td style = "text-align: left;">29 goals</td><td style = "text-align: left;">64 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">Eibar</td><td style = "text-align: right;">38</td><td style = "text-align: right;">9</td><td style = "text-align: right;">8</td><td style = "text-align: right;">21</td><td style = "text-align: left;">34 goals</td><td style = "text-align: left;">55 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">Almeria</td><td style = "text-align: right;">38</td><td style = "text-align: right;">8</td><td style = "text-align: right;">8</td><td style = "text-align: right;">22</td><td style = "text-align: left;">35 goals</td><td style = "text-align: left;">64 goals</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">Cordoba</td><td style = "text-align: right;">38</td><td style = "text-align: right;">3</td><td style = "text-align: right;">11</td><td style = "text-align: right;">24</td><td style = "text-align: left;">22 goals</td><td style = "text-align: left;">68 goals</td></tr></tbody></table></div><h2 id="Working-with-DataFrames"><a class="docs-heading-anchor" href="#Working-with-DataFrames">Working with DataFrames</a><a id="Working-with-DataFrames-1"></a><a class="docs-heading-anchor-permalink" href="#Working-with-DataFrames" title="Permalink"></a></h2><p>Now that we have read the required data into a DataFrame, let us look at some basic operations we can perform on it.</p><h3 id="Querying-Basic-Information"><a class="docs-heading-anchor" href="#Querying-Basic-Information">Querying Basic Information</a><a id="Querying-Basic-Information-1"></a><a class="docs-heading-anchor-permalink" href="#Querying-Basic-Information" title="Permalink"></a></h3><p>The <code>size</code> function gets us the dimensions of the DataFrame:</p><pre><code class="language-julia hljs">DataFrames.size(ss_df)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(23, 10)</code></pre><p>We can also use the <code>nrow</code> and <code>ncol</code> functions to get the number of rows and columns respectively:</p><pre><code class="language-julia hljs">DataFrames.nrow(ss_df), DataFrames.ncol(ss_df)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(23, 10)</code></pre><p>The <code>describe</code> function gives basic summary statistics of data in a DataFrame:</p><pre><code class="language-julia hljs">DataFrames.describe(ss_df)</code></pre><div><div style = "float: left;"><span>10×7 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">variable</th><th style = "text-align: left;">mean</th><th style = "text-align: left;">min</th><th style = "text-align: left;">median</th><th style = "text-align: left;">max</th><th style = "text-align: left;">nmissing</th><th style = "text-align: left;">eltype</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Symbol" style = "text-align: left;">Symbol</th><th title = "Union{Nothing, Float64}" style = "text-align: left;">Union…</th><th title = "Any" style = "text-align: left;">Any</th><th title = "Union{Nothing, Float64}" style = "text-align: left;">Union…</th><th title = "Any" style = "text-align: left;">Any</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "DataType" style = "text-align: left;">DataType</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Name</td><td style = "font-style: italic; text-align: left;"></td><td style = "text-align: left;">AppleJacks</td><td style = "font-style: italic; text-align: left;"></td><td style = "text-align: left;">WheatChex</td><td style = "text-align: right;">0</td><td style = "text-align: left;">String31</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Cups</td><td style = "text-align: left;">0.823043</td><td style = "text-align: left;">0.25</td><td style = "text-align: left;">1.0</td><td style = "text-align: left;">1.25</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Float64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Calories</td><td style = "text-align: left;">113.043</td><td style = "text-align: left;">100</td><td style = "text-align: left;">110.0</td><td style = "text-align: left;">150</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Carbs</td><td style = "text-align: left;">15.0435</td><td style = "text-align: left;">9.0</td><td style = "text-align: left;">15.0</td><td style = "text-align: left;">22.0</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Float64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Fat</td><td style = "text-align: left;">1.13043</td><td style = "text-align: left;">0</td><td style = "text-align: left;">1.0</td><td style = "text-align: left;">3</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">Fiber</td><td style = "text-align: left;">1.56522</td><td style = "text-align: left;">0.0</td><td style = "text-align: left;">1.5</td><td style = "text-align: left;">4.0</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Float64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">Potassium</td><td style = "text-align: left;">86.3043</td><td style = "text-align: left;">25</td><td style = "text-align: left;">90.0</td><td style = "text-align: left;">230</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">Protein</td><td style = "text-align: left;">2.91304</td><td style = "text-align: left;">1</td><td style = "text-align: left;">3.0</td><td style = "text-align: left;">6</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">Sodium</td><td style = "text-align: left;">189.957</td><td style = "text-align: left;">0</td><td style = "text-align: left;">190.0</td><td style = "text-align: left;">320</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">Sugars</td><td style = "text-align: left;">7.52174</td><td style = "text-align: left;">1</td><td style = "text-align: left;">7.0</td><td style = "text-align: left;">15</td><td style = "text-align: right;">0</td><td style = "text-align: left;">Int64</td></tr></tbody></table></div><p>Names of every column can be obtained by the <code>names</code> function:</p><pre><code class="language-julia hljs">DataFrames.names(ss_df)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">10-element Vector{String}:
+ &quot;Name&quot;
+ &quot;Cups&quot;
+ &quot;Calories&quot;
+ &quot;Carbs&quot;
+ &quot;Fat&quot;
+ &quot;Fiber&quot;
+ &quot;Potassium&quot;
+ &quot;Protein&quot;
+ &quot;Sodium&quot;
+ &quot;Sugars&quot;</code></pre><p>Corresponding data types are obtained using the broadcasted <code>eltype</code> function:</p><pre><code class="language-julia hljs">eltype.(ss_df)</code></pre><div><div style = "float: left;"><span>23×10 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Cups</th><th style = "text-align: left;">Calories</th><th style = "text-align: left;">Carbs</th><th style = "text-align: left;">Fat</th><th style = "text-align: left;">Fiber</th><th style = "text-align: left;">Potassium</th><th style = "text-align: left;">Protein</th><th style = "text-align: left;">Sodium</th><th style = "text-align: left;">Sugars</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th><th title = "DataType" style = "text-align: left;">DataType</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">21</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">22</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">23</td><td style = "text-align: left;">Char</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Float64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td><td style = "text-align: left;">Int64</td></tr></tbody></table></div><h3 id="Accessing-the-Data"><a class="docs-heading-anchor" href="#Accessing-the-Data">Accessing the Data</a><a id="Accessing-the-Data-1"></a><a class="docs-heading-anchor-permalink" href="#Accessing-the-Data" title="Permalink"></a></h3><p>Similar to regular arrays, we use numerical indexing to access elements of a DataFrame:</p><pre><code class="language-julia hljs">csv_df[1, 1]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">&quot;Anakin Skywalker&quot;</code></pre><p>The following are different ways to access a column:</p><pre><code class="language-julia hljs">csv_df[!, 1]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20-element Vector{InlineStrings.String31}:
+ &quot;Anakin Skywalker&quot;
+ &quot;Padme Amidala&quot;
+ &quot;Luke Skywalker&quot;
+ &quot;Leia Skywalker&quot;
+ &quot;Qui-Gon Jinn&quot;
+ &quot;Obi-Wan Kenobi&quot;
+ &quot;Han Solo&quot;
+ &quot;Sheev Palpatine&quot;
+ &quot;R2-D2&quot;
+ &quot;C-3PO&quot;
+ &quot;Yoda&quot;
+ &quot;Darth Maul&quot;
+ &quot;Dooku&quot;
+ &quot;Chewbacca&quot;
+ &quot;Jabba&quot;
+ &quot;Lando Calrissian&quot;
+ &quot;Boba Fett&quot;
+ &quot;Jango Fett&quot;
+ &quot;Grievous&quot;
+ &quot;Chief Chirpa&quot;</code></pre><pre><code class="language-julia hljs">csv_df[!, :Name]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20-element Vector{InlineStrings.String31}:
+ &quot;Anakin Skywalker&quot;
+ &quot;Padme Amidala&quot;
+ &quot;Luke Skywalker&quot;
+ &quot;Leia Skywalker&quot;
+ &quot;Qui-Gon Jinn&quot;
+ &quot;Obi-Wan Kenobi&quot;
+ &quot;Han Solo&quot;
+ &quot;Sheev Palpatine&quot;
+ &quot;R2-D2&quot;
+ &quot;C-3PO&quot;
+ &quot;Yoda&quot;
+ &quot;Darth Maul&quot;
+ &quot;Dooku&quot;
+ &quot;Chewbacca&quot;
+ &quot;Jabba&quot;
+ &quot;Lando Calrissian&quot;
+ &quot;Boba Fett&quot;
+ &quot;Jango Fett&quot;
+ &quot;Grievous&quot;
+ &quot;Chief Chirpa&quot;</code></pre><pre><code class="language-julia hljs">csv_df.Name</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20-element Vector{InlineStrings.String31}:
+ &quot;Anakin Skywalker&quot;
+ &quot;Padme Amidala&quot;
+ &quot;Luke Skywalker&quot;
+ &quot;Leia Skywalker&quot;
+ &quot;Qui-Gon Jinn&quot;
+ &quot;Obi-Wan Kenobi&quot;
+ &quot;Han Solo&quot;
+ &quot;Sheev Palpatine&quot;
+ &quot;R2-D2&quot;
+ &quot;C-3PO&quot;
+ &quot;Yoda&quot;
+ &quot;Darth Maul&quot;
+ &quot;Dooku&quot;
+ &quot;Chewbacca&quot;
+ &quot;Jabba&quot;
+ &quot;Lando Calrissian&quot;
+ &quot;Boba Fett&quot;
+ &quot;Jango Fett&quot;
+ &quot;Grievous&quot;
+ &quot;Chief Chirpa&quot;</code></pre><pre><code class="language-julia hljs">csv_df[:, 1] # Note that this creates a copy.</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20-element Vector{InlineStrings.String31}:
+ &quot;Anakin Skywalker&quot;
+ &quot;Padme Amidala&quot;
+ &quot;Luke Skywalker&quot;
+ &quot;Leia Skywalker&quot;
+ &quot;Qui-Gon Jinn&quot;
+ &quot;Obi-Wan Kenobi&quot;
+ &quot;Han Solo&quot;
+ &quot;Sheev Palpatine&quot;
+ &quot;R2-D2&quot;
+ &quot;C-3PO&quot;
+ &quot;Yoda&quot;
+ &quot;Darth Maul&quot;
+ &quot;Dooku&quot;
+ &quot;Chewbacca&quot;
+ &quot;Jabba&quot;
+ &quot;Lando Calrissian&quot;
+ &quot;Boba Fett&quot;
+ &quot;Jango Fett&quot;
+ &quot;Grievous&quot;
+ &quot;Chief Chirpa&quot;</code></pre><p>The following are different ways to access a row:</p><pre><code class="language-julia hljs">csv_df[1:1, :]</code></pre><div><div style = "float: left;"><span>1×13 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Gender</th><th style = "text-align: left;">Height</th><th style = "text-align: left;">Weight</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">Haircolor</th><th style = "text-align: left;">Skincolor</th><th style = "text-align: left;">Homeland</th><th style = "text-align: left;">Born</th><th style = "text-align: left;">Died</th><th style = "text-align: left;">Jedi</th><th style = "text-align: left;">Species</th><th style = "text-align: left;">Weapon</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Union{Missing, InlineStrings.String7}" style = "text-align: left;">String7</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Anakin Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.88</td><td style = "text-align: right;">84.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">41.9BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr></tbody></table></div><pre><code class="language-julia hljs">csv_df[1, :] # This produces a DataFrameRow.</code></pre><div><div style = "float: left;"><span>DataFrameRow (13 columns)</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowLabel" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Gender</th><th style = "text-align: left;">Height</th><th style = "text-align: left;">Weight</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">Haircolor</th><th style = "text-align: left;">Skincolor</th><th style = "text-align: left;">Homeland</th><th style = "text-align: left;">Born</th><th style = "text-align: left;">Died</th><th style = "text-align: left;">Jedi</th><th style = "text-align: left;">Species</th><th style = "text-align: left;">Weapon</th></tr><tr class = "subheader headerLastRow"><th class = "rowLabel" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Union{Missing, InlineStrings.String7}" style = "text-align: left;">String7</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowLabel" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Anakin Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.88</td><td style = "text-align: right;">84.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">41.9BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr></tbody></table></div><p>We can change the values just as we normally assign values.</p><p>Assign a range to scalar:</p><pre><code class="language-julia hljs">csv_df[1:3, :Height] .= 1.83</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element view(::Vector{Float64}, 1:3) with eltype Float64:
+ 1.83
+ 1.83
+ 1.83</code></pre><p>Assign a vector:</p><pre><code class="language-julia hljs">csv_df[4:6, :Height] = [1.8, 1.6, 1.8]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{Float64}:
+ 1.8
+ 1.6
+ 1.8</code></pre><pre><code class="language-julia hljs">csv_df</code></pre><div><div style = "float: left;"><span>20×13 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">Name</th><th style = "text-align: left;">Gender</th><th style = "text-align: left;">Height</th><th style = "text-align: left;">Weight</th><th style = "text-align: left;">Eyecolor</th><th style = "text-align: left;">Haircolor</th><th style = "text-align: left;">Skincolor</th><th style = "text-align: left;">Homeland</th><th style = "text-align: left;">Born</th><th style = "text-align: left;">Died</th><th style = "text-align: left;">Jedi</th><th style = "text-align: left;">Species</th><th style = "text-align: left;">Weapon</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String31" style = "text-align: left;">String31</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "Union{Missing, InlineStrings.String7}" style = "text-align: left;">String7</th><th title = "Union{Missing, InlineStrings.String15}" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String7" style = "text-align: left;">String7</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">Anakin Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: right;">84.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">41.9BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Padme Amidala</td><td style = "text-align: left;">female</td><td style = "text-align: right;">1.83</td><td style = "text-align: right;">45.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">46BBY</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Luke Skywalker</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: right;">77.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">blond</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Leia Skywalker</td><td style = "text-align: left;">female</td><td style = "text-align: right;">1.8</td><td style = "text-align: right;">49.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Alderaan</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Qui-Gon Jinn</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.6</td><td style = "text-align: right;">88.5</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">unk_planet</td><td style = "text-align: left;">92BBY</td><td style = "text-align: left;">32BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">Obi-Wan Kenobi</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.8</td><td style = "text-align: right;">77.0</td><td style = "text-align: left;">bluegray</td><td style = "text-align: left;">auburn</td><td style = "text-align: left;">fair</td><td style = "text-align: left;">Stewjon</td><td style = "text-align: left;">57BBY</td><td style = "text-align: left;">0BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">Han Solo</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.8</td><td style = "text-align: right;">80.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Corellia</td><td style = "text-align: left;">29BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">Sheev Palpatine</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.73</td><td style = "text-align: right;">75.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">red</td><td style = "text-align: left;">pale</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">82BBY</td><td style = "text-align: left;">10ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">force-lightning</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">R2-D2</td><td style = "text-align: left;">male</td><td style = "text-align: right;">0.96</td><td style = "text-align: right;">32.0</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">Naboo</td><td style = "text-align: left;">33BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">droid</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">C-3PO</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.67</td><td style = "text-align: right;">75.0</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">112BBY</td><td style = "text-align: left;">3ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">droid</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">Yoda</td><td style = "text-align: left;">male</td><td style = "text-align: right;">0.66</td><td style = "text-align: right;">17.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">green</td><td style = "text-align: left;">unk_planet</td><td style = "text-align: left;">896BBY</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">yoda</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">Darth Maul</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.75</td><td style = "text-align: right;">80.0</td><td style = "text-align: left;">yellow</td><td style = "text-align: left;">none</td><td style = "text-align: left;">red</td><td style = "text-align: left;">Dathomir</td><td style = "text-align: left;">54BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">dathomirian</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">Dooku</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.93</td><td style = "text-align: right;">86.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">light</td><td style = "text-align: left;">Serenno</td><td style = "text-align: left;">102BBY</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">lightsaber</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">Chewbacca</td><td style = "text-align: left;">male</td><td style = "text-align: right;">2.28</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">blue</td><td style = "text-align: left;">brown</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">Kashyyyk</td><td style = "text-align: left;">200BBY</td><td style = "text-align: left;">25ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">wookiee</td><td style = "text-align: left;">bowcaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">Jabba</td><td style = "text-align: left;">male</td><td style = "text-align: right;">3.9</td><td style = "font-style: italic; text-align: right;">missing</td><td style = "text-align: left;">yellow</td><td style = "text-align: left;">none</td><td style = "text-align: left;">tan-green</td><td style = "text-align: left;">Tatooine</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">hutt</td><td style = "text-align: left;">unarmed</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">Lando Calrissian</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.78</td><td style = "text-align: right;">79.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">blank</td><td style = "text-align: left;">dark</td><td style = "text-align: left;">Socorro</td><td style = "text-align: left;">31BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">Boba Fett</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: right;">78.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">black</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">Kamino</td><td style = "text-align: left;">31.5BBY</td><td style = "text-align: left;">unk_died</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">Jango Fett</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.83</td><td style = "text-align: right;">79.0</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">black</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">ConcordDawn</td><td style = "text-align: left;">66BBY</td><td style = "text-align: left;">22BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">human</td><td style = "text-align: left;">blaster</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">Grievous</td><td style = "text-align: left;">male</td><td style = "text-align: right;">2.16</td><td style = "text-align: right;">159.0</td><td style = "text-align: left;">gold</td><td style = "text-align: left;">black</td><td style = "text-align: left;">orange</td><td style = "text-align: left;">Kalee</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">19BBY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">kaleesh</td><td style = "text-align: left;">slugthrower</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">Chief Chirpa</td><td style = "text-align: left;">male</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">50.0</td><td style = "text-align: left;">black</td><td style = "text-align: left;">gray</td><td style = "text-align: left;">brown</td><td style = "text-align: left;">Endor</td><td style = "text-align: left;">unk_born</td><td style = "text-align: left;">4ABY</td><td style = "text-align: left;">no_jedi</td><td style = "text-align: left;">ewok</td><td style = "text-align: left;">spear</td></tr></tbody></table></div><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>There are a lot more things which can be done with a DataFrame. Read the <a href="https://juliadata.github.io/DataFrames.jl/stable/">docs</a> for more information.</p></div></div><p>For information on dplyr-type syntax:</p><ul><li>Read the <a href="https://dataframes.juliadata.org/stable/man/querying_frameworks/">DataFrames.jl documentation</a></li><li>Check out <a href="https://github.com/JuliaData/DataFramesMeta.jl">DataFramesMeta.jl</a></li></ul><h2 id="Example:-the-passport-problem"><a class="docs-heading-anchor" href="#Example:-the-passport-problem">Example: the passport problem</a><a id="Example:-the-passport-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Example:-the-passport-problem" title="Permalink"></a></h2><p>Let&#39;s now apply what we have learned to solve a real problem.</p><h3 id="Data-manipulation"><a class="docs-heading-anchor" href="#Data-manipulation">Data manipulation</a><a id="Data-manipulation-1"></a><a class="docs-heading-anchor-permalink" href="#Data-manipulation" title="Permalink"></a></h3><p>The <a href="https://github.com/ilyankou/passport-index-dataset">Passport Index Dataset</a> lists travel visa requirements for 199 countries, in <code>.csv</code> format. Our task is to find the minimum number of passports required to visit all countries.</p><pre><code class="language-julia hljs">passport_data = CSV.read(
+    joinpath(DATA_DIR, &quot;passport-index-matrix.csv&quot;),
+    DataFrames.DataFrame,
+);</code></pre><p>In this dataset, the first column represents a passport (=from) and each remaining column represents a foreign country (=to).</p><p>The values in each cell are as follows:</p><ul><li>3 = visa-free travel</li><li>2 = eTA is required</li><li>1 = visa can be obtained on arrival</li><li>0 = visa is required</li><li>-1 is for all instances where passport and destination are the same</li></ul><p>Our task is to find out the minimum number of passports needed to visit every country without requiring a visa.</p><p>The values we are interested in are -1 and 3. Let&#39;s modify the dataframe so that the -1 and 3 are <code>1</code> (true), and all others are <code>0</code> (false):</p><pre><code class="language-julia hljs">function modifier(x)
+    if x == -1 || x == 3
+        return 1
+    else
+        return 0
+    end
+end
+
+for country in passport_data.Passport
+    passport_data[!, country] = modifier.(passport_data[!, country])
+end</code></pre><p>The values in the cells now represent:</p><ul><li>1 = no visa required for travel</li><li>0 = visa required for travel</li></ul><h3 id="JuMP-Modeling"><a class="docs-heading-anchor" href="#JuMP-Modeling">JuMP Modeling</a><a id="JuMP-Modeling-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-Modeling" title="Permalink"></a></h3><p>To model the problem as a mixed-integer linear program, we need a binary decision variable <span>$x_c$</span> for each country <span>$c$</span>. <span>$x_c$</span> is <span>$1$</span> if we select passport <span>$c$</span> and <span>$0$</span> otherwise. Our objective is to minimize the sum <span>$\sum x_c$</span> over all countries.</p><p>Since we wish to visit all the countries, for every country, we must own at least one passport that lets us travel to that country visa free. For one destination, this can be mathematically represented as <span>$\sum_{c \in C} a_{c,d} \cdot x_{d} \geq 1$</span>, where <span>$a$</span> is the <code>passport_data</code> dataframe.</p><p>Thus, we can represent this problem using the following model:</p><p class="math-container">\[\begin{aligned}
+\min &amp;&amp; \sum_{c \in C} x_c \\
+\text{s.t.} &amp;&amp; \sum_{c \in C} a_{c,d} x_c \geq 1 &amp;&amp; \forall d \in C \\
+&amp;&amp; x_c \in \{0,1\} &amp;&amp; \forall c \in C.
+\end{aligned}\]</p><p>We&#39;ll now solve the problem using JuMP:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS</code></pre><p>First, create the set of countries:</p><pre><code class="language-julia hljs">C = passport_data.Passport</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">199-element Vector{String}:
+ &quot;Afghanistan&quot;
+ &quot;Albania&quot;
+ &quot;Algeria&quot;
+ &quot;Andorra&quot;
+ &quot;Angola&quot;
+ &quot;Antigua and Barbuda&quot;
+ &quot;Argentina&quot;
+ &quot;Armenia&quot;
+ &quot;Australia&quot;
+ &quot;Austria&quot;
+ ⋮
+ &quot;Uruguay&quot;
+ &quot;Uzbekistan&quot;
+ &quot;Vanuatu&quot;
+ &quot;Vatican&quot;
+ &quot;Venezuela&quot;
+ &quot;Viet Nam&quot;
+ &quot;Yemen&quot;
+ &quot;Zambia&quot;
+ &quot;Zimbabwe&quot;</code></pre><p>Then, create the model and initialize the decision variables:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[C], Bin)
+@objective(model, Min, sum(x))
+@constraint(model, [d in C], passport_data[!, d]&#39; * x &gt;= 1)
+model</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A JuMP Model
+Minimization problem with:
+Variables: 199
+Objective function type: AffExpr
+`AffExpr`-in-`MathOptInterface.GreaterThan{Float64}`: 199 constraints
+`VariableRef`-in-`MathOptInterface.ZeroOne`: 199 constraints
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS
+Names registered in the model: x</code></pre><p>Now optimize:</p><pre><code class="language-julia hljs">optimize!(model)</code></pre><p>We can use the <a href="../../../api/JuMP/#JuMP.solution_summary"><code>solution_summary</code></a> function to get an overview of the solution:</p><pre><code class="language-julia hljs">solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : 2.30000e+01
+  Objective bound    : 2.30000e+01
+  Relative gap       : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 7.45583e-03
+  Simplex iterations : 26
+  Barrier iterations : -1
+  Node count         : 1
+</code></pre><h3 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h3><p>Let&#39;s  have a look at the solution in more detail:</p><pre><code class="language-julia hljs">println(&quot;Minimum number of passports needed: &quot;, objective_value(model))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Minimum number of passports needed: 23.0</code></pre><pre><code class="language-julia hljs">println(&quot;Optimal passports:&quot;)
+for c in C
+    if value(x[c]) &gt; 0.5
+        println(&quot; * &quot;, c)
+    end
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Optimal passports:
+ * Afghanistan
+ * Chad
+ * Comoros
+ * Djibouti
+ * Georgia
+ * Hong Kong
+ * India
+ * Luxembourg
+ * Madagascar
+ * Maldives
+ * Mali
+ * New Zealand
+ * North Korea
+ * Papua New Guinea
+ * Singapore
+ * Somalia
+ * Sri Lanka
+ * Tunisia
+ * Turkey
+ * Uganda
+ * United Arab Emirates
+ * United States
+ * Zimbabwe</code></pre><p>We need some passports, like New Zealand and the United States, which have widespread access to a large number of countries. However, we also need passports like North Korea which only have visa-free access to a very limited number of countries.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>We use <code>value(x[c]) &gt; 0.5</code> rather than <code>value(x[c]) == 1</code> to avoid excluding solutions like <code>x[c] = 0.99999</code> that are &quot;1&quot; to some tolerance.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../getting_started_with_sets_and_indexing/">« Getting started with sets and indexing</a><a class="docs-footer-nextpage" href="../debugging/">Debugging »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_julia.jl b/previews/PR3525/tutorials/getting_started/getting_started_with_julia.jl
new file mode 100644
index 00000000000..e6b3a2f0f22
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_julia.jl
@@ -0,0 +1,722 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Getting started with Julia
+
+# Because JuMP is embedded in Julia, knowing some basic Julia is important
+# before you start learning JuMP.
+
+# !!! tip
+#     This tutorial is designed to provide a minimalist crash course in the
+#     basics of Julia. You can find resources that provide a more comprehensive
+#     introduction to Julia [here](https://julialang.org/learning/).
+
+# ## Installing Julia
+
+# To install Julia, [download the latest stable release](https://julialang.org/downloads/),
+# then follow the [platform specific install instructions](https://julialang.org/downloads/platform/).
+
+# !!! tip
+#     Unless you know otherwise, you probably want the 64-bit version.
+
+# Next, you need an IDE to develop in. VS Code is a popular choice, so follow
+# [these install instructions](https://www.julia-vscode.org/docs/stable/gettingstarted/).
+
+# Julia can also be used with [Jupyter notebooks](https://github.com/JuliaLang/IJulia.jl)
+# or the reactive notebooks of [Pluto.jl](https://github.com/fonsp/Pluto.jl).
+
+# ## The Julia REPL
+
+# The main way of interacting with Julia is via its REPL (Read Evaluate Print
+# Loop). To access the REPL, start the Julia executable to arrive at the
+# `julia>` prompt, and then start coding:
+
+1 + 1
+
+# As your programs become larger, write a script as a text file, and then run
+# that file using:
+# ```julia
+# julia> include("path/to/file.jl")
+# ```
+
+# !!! warning
+#     Because of Julia's startup latency, running scripts from the command line
+#     like the following is slow:
+#     ```
+#     $ julia path/to/file.jl
+#     ```
+#     Use the REPL or a notebook instead, and read [The "time-to-first-solve" issue](@ref)
+#     for more information.
+
+# ### Code blocks in this documentation
+
+# In this documentation you'll see a mix of code examples with and without the
+# `julia>`.
+
+# The Julia prompt is mostly used to demonstrate short code snippets, and the
+# output is exactly what you will see if run from the REPL.
+
+# Blocks without the `julia>` can be copy-pasted into the REPL, but they are
+# used because they enable richer output like plots or LaTeX to be displayed in
+# the online and [PDF](https://jump.dev/JuMP.jl/stable/JuMP.pdf) versions of the
+# documentation. If you run them from the REPL you may see different output.
+
+# ## Where to get help
+
+#  * Read the documentation
+#    * JuMP [https://jump.dev/JuMP.jl/stable/](https://jump.dev/JuMP.jl/stable/)
+#    * Julia [https://docs.julialang.org/en/v1/](https://docs.julialang.org/en/v1/)
+#  * Ask (or browse) the Julia community forum: [https://discourse.julialang.org](https://discourse.julialang.org)
+#    * If the question is JuMP-related, ask in the [Optimization (Mathematical)](https://discourse.julialang.org/c/domain/opt/13)
+#      section, or tag your question with "jump"
+
+# To access the built-in help at the REPL, type `?` to enter help-mode, followed
+# by the name of the function to lookup:
+# ```julia
+# help?> print
+# search: print println printstyled sprint isprint prevind parentindices precision escape_string
+#
+#   print([io::IO], xs...)
+#
+#   Write to io (or to the default output stream stdout if io is not given) a canonical
+#   (un-decorated) text representation. The representation used by print includes minimal formatting
+#   and tries to avoid Julia-specific details.
+#
+#   print falls back to calling show, so most types should just define show. Define print if your
+#   type has a separate "plain" representation. For example, show displays strings with quotes, and
+#   print displays strings without quotes.
+#
+#   string returns the output of print as a string.
+#
+#   Examples
+#   ≡≡≡≡≡≡≡≡≡≡
+#
+#   julia> print("Hello World!")
+#   Hello World!
+#   julia> io = IOBuffer();
+#
+#   julia> print(io, "Hello", ' ', :World!)
+#
+#   julia> String(take!(io))
+#   "Hello World!"
+# ```
+
+# ## Numbers and arithmetic
+
+# Since we want to solve optimization problems, we're going to be using a lot of
+# math. Luckily, Julia is great for math, with all the usual operators:
+
+1 + 1
+1 - 2
+2 * 2
+4 / 5
+3^2
+
+# Did you notice how Julia didn't print `.0` after some of the numbers? Julia is
+# a dynamic language, which means you never have to explicitly declare the type
+# of a variable. However, in the background, Julia is giving each variable a
+# type. Check the type of something using the `typeof` function:
+
+typeof(1)
+typeof(1.0)
+
+# Here `1` is an `Int64`, which is an integer with 64 bits of precision, and
+# `1.0` is a `Float64`, which is a floating point number with 64-bits of
+# precision.
+
+# !!! tip
+#     If you aren't familiar with floating point numbers, make sure to read
+#     the [Floating point numbers](@ref) section.
+
+# We create complex numbers using `im`:
+
+x = 2 + 1im
+real(x)
+imag(x)
+typeof(x)
+x * (1 - 2im)
+
+# !!! info
+#     The curly brackets surround what we call the _parameters_ of a type. You
+#     can read `Complex{Int64}`  as "a complex number, where the real and
+#     imaginary parts are represented by `Int64`." If we call
+#     `typeof(1.0 + 2.0im)` it will be `Complex{Float64}`, which a complex
+#     number with the parts represented by `Float64`.
+
+# There are also some cool things like an irrational representation of π.
+
+π
+
+# !!! tip
+#     To make π (and most other Greek letters), type `\pi` and then press
+#     `[TAB]`.
+
+typeof(π)
+
+# However, if we do math with irrational numbers, they get converted to
+# `Float64`:
+
+typeof(2π / 3)
+
+# ### Floating point numbers
+
+# !!! warning
+#     If you aren't familiar with floating point numbers, make sure to read this
+#     section carefully.
+
+# A `Float64` is a [floating point](https://en.wikipedia.org/wiki/Floating-point_arithmetic)
+# approximation of a real number using 64-bits of information.
+
+# Because it is an approximation, things we know hold true in mathematics don't
+# hold true in a computer. For example:
+
+0.1 * 3 == 0.3
+
+# A more complicated example is:
+
+sin(2π / 3) == √3 / 2
+
+# !!! tip
+#     Get `√` by typing `\sqrt` then press `[TAB]`.
+
+# Let's see what the differences are:
+
+0.1 * 3 - 0.3
+sin(2π / 3) - √3 / 2
+
+# They are small, but not zero.
+
+# One way of explaining this difference is to consider how we would write
+# `1 / 3` and `2 / 3` using only four digits after the decimal point. We would
+# write `1 / 3` as `0.3333`, and `2 / 3` as `0.6667`. So, despite the fact that
+# `2 * (1 / 3) == 2 / 3`, `2 * 0.3333 == 0.6666 != 0.6667`.
+
+# Let's try that again using ≈ (`\approx + [TAB]`) instead of `==`:
+
+0.1 * 3 ≈ 0.3
+sin(2π / 3) ≈ √3 / 2
+
+# `≈` is a clever way of calling the `isapprox` function:
+
+isapprox(sin(2π / 3), √3 / 2; atol = 1e-8)
+
+# !!! warning
+#     Floating point is the reason solvers use tolerances when they solve
+#     optimization models. A common mistake you're likely to make is checking
+#     whether a binary variable is 0 using `value(z) == 0`. Always remember to
+#     use something like `isapprox` when comparing floating point numbers.
+
+# Note that `isapprox` will always return `false` if one of the number being
+# compared is `0` and `atol` is zero (its default value).
+
+1e-300 ≈ 0.0
+
+# so always set a nonzero value of `atol` if one of the arguments can be zero.
+
+isapprox(1e-9, 0.0; atol = 1e-8)
+
+# !!! tip
+#     Gurobi has a [good series of articles](https://www.gurobi.com/documentation/9.0/refman/num_grb_guidelines_for_num.html)
+#     on the implications of floating point in optimization if you want to read
+#     more.
+
+# If you aren't careful, floating point arithmetic can throw up all manner of
+# issues. For example:
+
+1 + 1e-16 == 1
+
+# It even turns out that floating point numbers aren't associative:
+
+(1 + 1e-16) - 1e-16 == 1 + (1e-16 - 1e-16)
+
+# It's important to note that this issue isn't Julia-specific. It happens in
+# every programming language (try it out in Python).
+
+# ## Vectors, matrices and arrays
+
+# Similar to MATLAB, Julia has native support for vectors, matrices and tensors;
+# all of which are represented by arrays of different dimensions. Vectors are
+# constructed by comma-separated elements surrounded by square brackets:
+
+b = [5, 6]
+
+# Matrices can be constructed with spaces separating the columns, and semicolons
+# separating the rows:
+
+A = [1.0 2.0; 3.0 4.0]
+
+# We can do linear algebra:
+
+x = A \ b
+
+# !!! info
+#     Here is floating point at work again; `x` is approximately `[-4, 4.5]`.
+
+A * x
+A * x ≈ b
+
+# Note that when multiplying vectors and matrices, dimensions matter. For
+# example, you can't multiply a vector by a vector:
+
+try                         #hide
+    b * b
+catch err                   #hide
+    showerror(stderr, err)  #hide
+end                         #hide
+
+# But multiplying transposes works:
+
+b' * b
+b * b'
+
+# ## Other common types
+
+# ### Comments
+
+# Although not technically a type, code comments begin with the `#` character:
+
+1 + 1  # This is a comment
+
+# Multiline comments begin with `#=` and end with `=#`:
+# ```julia
+# #=
+# Here is a
+# multiline comment
+# =#
+# ```
+
+# Comments can even be nested inside expressions. This is sometimes helpful when
+# documenting inputs to functions:
+
+isapprox(
+    sin(π),
+    0.0;
+    #= We need an explicit atol here because we are comparing with 0 =#
+    atol = 0.001,
+)
+
+# ### Strings
+
+# Double quotes are used for strings:
+
+typeof("This is Julia")
+
+# Unicode is fine in strings:
+
+typeof("π is about 3.1415")
+
+# Use `println` to print a string:
+
+println("Hello, World!")
+
+# Use `$()` to interpolate values into a string:
+
+x = 123
+println("The value of x is: $(x)")
+
+# Use triple-quotes for multiline strings:
+
+s = """
+Here is
+a
+multiline string
+"""
+
+println(s)
+
+# ### Symbols
+
+# Julia `Symbol`s are a data structure from the compiler that represent Julia
+# identifiers (that is, variable names).
+
+println("The value of x is: $(eval(:x))")
+
+# !!! warning
+#     We used `eval` here to demonstrate how Julia links `Symbol`s to variables.
+#     However, avoid calling `eval` in your code. It is usually a sign that your
+#     code is doing something that could be more easily achieved a different
+#     way. The [Community Forum](https://jump.dev/forum) is a good place to ask
+#     for advice on alternative approaches.
+
+typeof(:x)
+
+# You can think of a `Symbol` as a `String` that takes up less memory, and that
+# can't be modified.
+
+# Convert between `String` and `Symbol` using their constructors:
+
+String(:abc)
+Symbol("abc")
+
+# !!! tip
+#     `Symbol`s are often (ab)used to stand in for a `String` or an `Enum`, when
+#     one of the latter is likely a better choice. The JuMP [Style guide](@ref)
+#     recommends reserving `Symbol`s for identifiers. See [@enum vs. Symbol](@ref)
+#     for more.
+
+# ### Tuples
+
+# Julia makes extensive use of a simple data structure called Tuples. Tuples are
+# immutable collections of values. For example:
+
+t = ("hello", 1.2, :foo)
+typeof(t)
+
+# Tuples can be accessed by index, similar to arrays:
+
+t[2]
+
+# And they can be "unpacked" like so:
+
+a, b, c = t
+b
+
+# The values can also be given names, which is a convenient way of making
+# light-weight data structures.
+
+t = (word = "hello", num = 1.2, sym = :foo)
+
+# Values can be accessed using dot syntax:
+
+t.word
+
+# ## Dictionaries
+
+# Similar to Python, Julia has native support for dictionaries. Dictionaries
+# provide a very generic way of mapping keys to values.  For example, a map of
+# integers to strings:
+
+d1 = Dict(1 => "A", 2 => "B", 4 => "D")
+
+# !!! info
+#     Type-stuff again: `Dict{Int64,String}` is a dictionary with `Int64` keys
+#     and `String` values.
+
+# Looking up a value uses the bracket syntax:
+
+d1[2]
+
+# Dictionaries support non-integer keys and can mix data types:
+
+Dict("A" => 1, "B" => 2.5, "D" => 2 - 3im)
+
+# !!! info
+#     Julia types form a hierarchy. Here the value type of the dictionary is
+#     `Number`, which is a generalization of `Int64`, `Float64`, and
+#     `Complex{Int}`. Leaf nodes in this hierarchy are called "concrete" types,
+#     and all others are called "Abstract." In general, having variables with
+#     abstract types like `Number` can lead to slower code, so you should try to
+#     make sure every element in a dictionary or vector is the same type. For
+#     example, in this case we could represent every element as a
+#     `Complex{Float64}`:
+
+Dict("A" => 1.0 + 0.0im, "B" => 2.5 + 0.0im, "D" => 2.0 - 3.0im)
+
+# Dictionaries can be nested:
+
+d2 = Dict("A" => 1, "B" => 2, "D" => Dict(:foo => 3, :bar => 4))
+d2["B"]
+d2["D"][:foo]
+
+# ## Structs
+
+# You can define custom datastructures with `struct`:
+
+struct MyStruct
+    x::Int
+    y::String
+    z::Dict{Int,Int}
+end
+
+a = MyStruct(1, "a", Dict(2 => 3))
+a.x
+
+# By default, these are not mutable
+
+try                         #hide
+    a.x = 2
+catch err                   #hide
+    showerror(stderr, err)  #hide
+end                         #hide
+
+# However, you can declare a `mutable struct` which is mutable:
+
+mutable struct MyStructMutable
+    x::Int
+    y::String
+    z::Dict{Int,Int}
+end
+
+a = MyStructMutable(1, "a", Dict(2 => 3))
+a.x
+a.x = 2
+a
+
+# ## Loops
+
+# Julia has native support for for-each style loops with the syntax
+# `for <value> in <collection> end`:
+
+for i in 1:5
+    println(i)
+end
+
+# !!! info
+#     Ranges are constructed as `start:stop`, or `start:step:stop`.
+
+for i in 1.2:1.1:5.6
+    println(i)
+end
+
+# This for-each loop also works with dictionaries:
+
+for (key, value) in Dict("A" => 1, "B" => 2.5, "D" => 2 - 3im)
+    println("$(key): $(value)")
+end
+
+# Note that in contrast to vector languages like MATLAB and R, loops do not
+# result in a significant performance degradation in Julia.
+
+# ## Control flow
+
+# Julia control flow is similar to MATLAB, using the keywords
+# `if-elseif-else-end`, and the logical operators `||` and `&&` for **or** and
+# **and** respectively:
+
+for i in 0:5:15
+    if i < 5
+        println("$(i) is less than 5")
+    elseif i < 10
+        println("$(i) is less than 10")
+    else
+        if i == 10
+            println("the value is 10")
+        else
+            println("$(i) is bigger than 10")
+        end
+    end
+end
+
+# ## Comprehensions
+
+# Similar to languages like Haskell and Python, Julia supports the use of simple
+# loops in the construction of arrays and dictionaries, called comprehensions.
+#
+# A list of increasing integers:
+
+[i for i in 1:5]
+
+# Matrices can be built by including multiple indices:
+
+[i * j for i in 1:5, j in 5:10]
+
+# Conditional statements can be used to filter out some values:
+
+[i for i in 1:10 if i % 2 == 1]
+
+# A similar syntax can be used for building dictionaries:
+
+Dict("$(i)" => i for i in 1:10 if i % 2 == 1)
+
+# ## Functions
+
+# A simple function is defined as follows:
+
+function print_hello()
+    return println("hello")
+end
+print_hello()
+
+# Arguments can be added to a function:
+
+function print_it(x)
+    return println(x)
+end
+print_it("hello")
+print_it(1.234)
+print_it(:my_id)
+
+# Optional keyword arguments are also possible:
+
+function print_it(x; prefix = "value:")
+    return println("$(prefix) $(x)")
+end
+print_it(1.234)
+print_it(1.234; prefix = "val:")
+
+# The keyword `return` is used to specify the return values of a function:
+
+function mult(x; y = 2.0)
+    return x * y
+end
+
+mult(4.0)
+mult(4.0; y = 5.0)
+
+# ### Anonymous functions
+
+# The syntax `input -> output` creates an anonymous function. These are most
+# useful when passed to other functions. For example:
+
+f = x -> x^2
+f(2)
+map(x -> x^2, 1:4)
+
+# ### Type parameters
+
+# We can constrain the inputs to a function using type parameters, which are
+# `::` followed by the type of the input we want. For example:
+
+function foo(x::Int)
+    return x^2
+end
+
+function foo(x::Float64)
+    return exp(x)
+end
+
+function foo(x::Number)
+    return x + 1
+end
+
+foo(2)
+foo(2.0)
+foo(1 + 1im)
+
+# But what happens if we call `foo` with something we haven't defined it for?
+
+try                         #hide
+    foo([1, 2, 3])
+catch err                   #hide
+    showerror(stderr, err)  #hide
+end                         #hide
+
+# A `MethodError` means that you passed a
+# function something that didn't match the type that it was expecting. In this
+# case, the error message says that it doesn't know how to handle an
+# `Vector{Int64}`, but it does know how to handle `Float64`, `Int64`, and
+# `Number`.
+#
+# !!! tip
+#     Read the "Closest candidates" part of the error message carefully to get a
+#     hint as to what was expected.
+
+# ### Broadcasting
+
+# In the example above, we didn't define what to do if `f` was passed a
+# `Vector`. Luckily, Julia provides a convenient syntax for mapping `f`
+# element-wise over arrays. Just add a `.` between the name of the function and
+# the opening `(`. This works for _any_ function, including functions with
+# multiple arguments. For example:
+
+f.([1, 2, 3])
+
+# !!! tip
+#     Get a `MethodError` when calling a function that takes a `Vector`,
+#     `Matrix`, or `Array`? Try broadcasting.
+
+# ## Mutable vs immutable objects
+
+# Some types in Julia are *mutable*, which means you can change the values
+# inside them. A good example is an array. You can modify the contents of an
+# array without having to make a new array.
+
+# In contrast, types like `Float64` are *immutable*. You cannot modify the
+# contents of a `Float64`.
+
+# This is something to be aware of when passing types into functions. For
+# example:
+
+function mutability_example(mutable_type::Vector{Int}, immutable_type::Int)
+    mutable_type[1] += 1
+    immutable_type += 1
+    return
+end
+
+mutable_type = [1, 2, 3]
+immutable_type = 1
+
+mutability_example(mutable_type, immutable_type)
+
+println("mutable_type: $(mutable_type)")
+println("immutable_type: $(immutable_type)")
+
+# Because `Vector{Int}` is a mutable type, modifying the variable inside the
+# function changed the value outside of the function. In contrast, the change
+# to `immutable_type` didn't modify the value outside the function.
+
+# You can check mutability with the `isimmutable` function:
+
+isimmutable([1, 2, 3])
+isimmutable(1)
+
+# ## The package manager
+
+# ### Installing packages
+
+# No matter how wonderful Julia's base language is, at some point you will want
+# to use an extension package.  Some of these are built-in, for example random
+# number generation is available in the `Random` package in the standard
+# library. These packages are loaded with the commands `using` and `import`.
+
+using Random  # The equivalent of Python's `from Random import *`
+import Random  # The equivalent of Python's `import Random`
+
+Random.seed!(33)
+
+[rand() for i in 1:10]
+
+# The Package Manager is used to install packages that are not part of Julia's
+# standard library.
+
+# For example the following can be used to install JuMP,
+# ```julia
+# using Pkg
+# Pkg.add("JuMP")
+# ```
+
+# For a complete list of registered Julia packages see the package listing at
+# [JuliaHub](https://juliahub.com).
+
+# From time to you may wish to use a Julia package that is not registered.  In
+# this case a git repository URL can be used to install the package.
+# ```julia
+# using Pkg
+# Pkg.add("https://github.com/user-name/MyPackage.jl.git")
+# ```
+
+# ### Package environments
+
+# By default, `Pkg.add` will add packages to Julia's global environment.
+# However, Julia also has built-in support for virtual environments.
+
+# Activate a virtual environment with:
+# ```julia
+# import Pkg; Pkg.activate("/path/to/environment")
+# ```
+
+# You can see what packages are installed in the current environment with
+# `Pkg.status()`.
+
+# !!! tip
+#     We _strongly_ recommend you create a Pkg environment for each project
+#     that you create in Julia, and add only the packages that you need, instead
+#     of adding lots of packages to the global environment. The [Pkg manager documentation](https://julialang.github.io/Pkg.jl/v1/environments/)
+#     has more information on this topic.
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_julia/index.html b/previews/PR3525/tutorials/getting_started/getting_started_with_julia/index.html
new file mode 100644
index 00000000000..d86de264979
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_julia/index.html
@@ -0,0 +1,183 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Getting started with Julia · JuMP</title><meta name="title" content="Getting started with Julia · JuMP"/><meta property="og:title" content="Getting started with Julia · JuMP"/><meta property="twitter:title" content="Getting started with Julia · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li class="is-active"><a class="tocitem" href>Getting started with Julia</a><ul class="internal"><li><a class="tocitem" href="#Installing-Julia"><span>Installing Julia</span></a></li><li><a class="tocitem" href="#The-Julia-REPL"><span>The Julia REPL</span></a></li><li><a class="tocitem" href="#Where-to-get-help"><span>Where to get help</span></a></li><li><a class="tocitem" href="#Numbers-and-arithmetic"><span>Numbers and arithmetic</span></a></li><li><a class="tocitem" href="#Vectors,-matrices-and-arrays"><span>Vectors, matrices and arrays</span></a></li><li><a class="tocitem" href="#Other-common-types"><span>Other common types</span></a></li><li><a class="tocitem" href="#Dictionaries"><span>Dictionaries</span></a></li><li><a class="tocitem" href="#Structs"><span>Structs</span></a></li><li><a class="tocitem" href="#Loops"><span>Loops</span></a></li><li><a class="tocitem" href="#Control-flow"><span>Control flow</span></a></li><li><a class="tocitem" href="#Comprehensions"><span>Comprehensions</span></a></li><li><a class="tocitem" href="#Functions"><span>Functions</span></a></li><li><a class="tocitem" href="#Mutable-vs-immutable-objects"><span>Mutable vs immutable objects</span></a></li><li><a class="tocitem" href="#The-package-manager"><span>The package manager</span></a></li></ul></li><li><a class="tocitem" href="../getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../debugging/">Debugging</a></li><li><a class="tocitem" href="../design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Getting started with Julia</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Getting started with Julia</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/getting_started_with_julia.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Getting-started-with-Julia"><a class="docs-heading-anchor" href="#Getting-started-with-Julia">Getting started with Julia</a><a id="Getting-started-with-Julia-1"></a><a class="docs-heading-anchor-permalink" href="#Getting-started-with-Julia" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../getting_started_with_julia.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>Because JuMP is embedded in Julia, knowing some basic Julia is important before you start learning JuMP.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This tutorial is designed to provide a minimalist crash course in the basics of Julia. You can find resources that provide a more comprehensive introduction to Julia <a href="https://julialang.org/learning/">here</a>.</p></div></div><h2 id="Installing-Julia"><a class="docs-heading-anchor" href="#Installing-Julia">Installing Julia</a><a id="Installing-Julia-1"></a><a class="docs-heading-anchor-permalink" href="#Installing-Julia" title="Permalink"></a></h2><p>To install Julia, <a href="https://julialang.org/downloads/">download the latest stable release</a>, then follow the <a href="https://julialang.org/downloads/platform/">platform specific install instructions</a>.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Unless you know otherwise, you probably want the 64-bit version.</p></div></div><p>Next, you need an IDE to develop in. VS Code is a popular choice, so follow <a href="https://www.julia-vscode.org/docs/stable/gettingstarted/">these install instructions</a>.</p><p>Julia can also be used with <a href="https://github.com/JuliaLang/IJulia.jl">Jupyter notebooks</a> or the reactive notebooks of <a href="https://github.com/fonsp/Pluto.jl">Pluto.jl</a>.</p><h2 id="The-Julia-REPL"><a class="docs-heading-anchor" href="#The-Julia-REPL">The Julia REPL</a><a id="The-Julia-REPL-1"></a><a class="docs-heading-anchor-permalink" href="#The-Julia-REPL" title="Permalink"></a></h2><p>The main way of interacting with Julia is via its REPL (Read Evaluate Print Loop). To access the REPL, start the Julia executable to arrive at the <code>julia&gt;</code> prompt, and then start coding:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 1 + 1</code><code class="nohighlight hljs ansi" style="display:block;">2</code></pre><p>As your programs become larger, write a script as a text file, and then run that file using:</p><pre><code class="language-julia hljs">julia&gt; include(&quot;path/to/file.jl&quot;)</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Because of Julia&#39;s startup latency, running scripts from the command line like the following is slow:</p><pre><code class="nohighlight hljs">$ julia path/to/file.jl</code></pre><p>Use the REPL or a notebook instead, and read <a href="../performance_tips/#The-&quot;time-to-first-solve&quot;-issue">The &quot;time-to-first-solve&quot; issue</a> for more information.</p></div></div><h3 id="Code-blocks-in-this-documentation"><a class="docs-heading-anchor" href="#Code-blocks-in-this-documentation">Code blocks in this documentation</a><a id="Code-blocks-in-this-documentation-1"></a><a class="docs-heading-anchor-permalink" href="#Code-blocks-in-this-documentation" title="Permalink"></a></h3><p>In this documentation you&#39;ll see a mix of code examples with and without the <code>julia&gt;</code>.</p><p>The Julia prompt is mostly used to demonstrate short code snippets, and the output is exactly what you will see if run from the REPL.</p><p>Blocks without the <code>julia&gt;</code> can be copy-pasted into the REPL, but they are used because they enable richer output like plots or LaTeX to be displayed in the online and <a href="https://jump.dev/JuMP.jl/stable/JuMP.pdf">PDF</a> versions of the documentation. If you run them from the REPL you may see different output.</p><h2 id="Where-to-get-help"><a class="docs-heading-anchor" href="#Where-to-get-help">Where to get help</a><a id="Where-to-get-help-1"></a><a class="docs-heading-anchor-permalink" href="#Where-to-get-help" title="Permalink"></a></h2><ul><li>Read the documentation<ul><li>JuMP <a href="https://jump.dev/JuMP.jl/stable/">https://jump.dev/JuMP.jl/stable/</a></li><li>Julia <a href="https://docs.julialang.org/en/v1/">https://docs.julialang.org/en/v1/</a></li></ul></li><li>Ask (or browse) the Julia community forum: <a href="https://discourse.julialang.org">https://discourse.julialang.org</a><ul><li>If the question is JuMP-related, ask in the <a href="https://discourse.julialang.org/c/domain/opt/13">Optimization (Mathematical)</a> section, or tag your question with &quot;jump&quot;</li></ul></li></ul><p>To access the built-in help at the REPL, type <code>?</code> to enter help-mode, followed by the name of the function to lookup:</p><pre><code class="language-julia hljs">help?&gt; print
+search: print println printstyled sprint isprint prevind parentindices precision escape_string
+
+  print([io::IO], xs...)
+
+  Write to io (or to the default output stream stdout if io is not given) a canonical
+  (un-decorated) text representation. The representation used by print includes minimal formatting
+  and tries to avoid Julia-specific details.
+
+  print falls back to calling show, so most types should just define show. Define print if your
+  type has a separate &quot;plain&quot; representation. For example, show displays strings with quotes, and
+  print displays strings without quotes.
+
+  string returns the output of print as a string.
+
+  Examples
+  ≡≡≡≡≡≡≡≡≡≡
+
+  julia&gt; print(&quot;Hello World!&quot;)
+  Hello World!
+  julia&gt; io = IOBuffer();
+
+  julia&gt; print(io, &quot;Hello&quot;, &#39; &#39;, :World!)
+
+  julia&gt; String(take!(io))
+  &quot;Hello World!&quot;</code></pre><h2 id="Numbers-and-arithmetic"><a class="docs-heading-anchor" href="#Numbers-and-arithmetic">Numbers and arithmetic</a><a id="Numbers-and-arithmetic-1"></a><a class="docs-heading-anchor-permalink" href="#Numbers-and-arithmetic" title="Permalink"></a></h2><p>Since we want to solve optimization problems, we&#39;re going to be using a lot of math. Luckily, Julia is great for math, with all the usual operators:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 1 + 1</code><code class="nohighlight hljs ansi" style="display:block;">2</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; 1 - 2</code><code class="nohighlight hljs ansi" style="display:block;">-1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; 2 * 2</code><code class="nohighlight hljs ansi" style="display:block;">4</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; 4 / 5</code><code class="nohighlight hljs ansi" style="display:block;">0.8</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; 3^2</code><code class="nohighlight hljs ansi" style="display:block;">9</code></pre><p>Did you notice how Julia didn&#39;t print <code>.0</code> after some of the numbers? Julia is a dynamic language, which means you never have to explicitly declare the type of a variable. However, in the background, Julia is giving each variable a type. Check the type of something using the <code>typeof</code> function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(1)</code><code class="nohighlight hljs ansi" style="display:block;">Int64</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(1.0)</code><code class="nohighlight hljs ansi" style="display:block;">Float64</code></pre><p>Here <code>1</code> is an <code>Int64</code>, which is an integer with 64 bits of precision, and <code>1.0</code> is a <code>Float64</code>, which is a floating point number with 64-bits of precision.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you aren&#39;t familiar with floating point numbers, make sure to read the <a href="#Floating-point-numbers">Floating point numbers</a> section.</p></div></div><p>We create complex numbers using <code>im</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; x = 2 + 1im</code><code class="nohighlight hljs ansi" style="display:block;">2 + 1im</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; real(x)</code><code class="nohighlight hljs ansi" style="display:block;">2</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; imag(x)</code><code class="nohighlight hljs ansi" style="display:block;">1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(x)</code><code class="nohighlight hljs ansi" style="display:block;">Complex{Int64}</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; x * (1 - 2im)</code><code class="nohighlight hljs ansi" style="display:block;">4 - 3im</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The curly brackets surround what we call the <em>parameters</em> of a type. You can read <code>Complex{Int64}</code>  as &quot;a complex number, where the real and imaginary parts are represented by <code>Int64</code>.&quot; If we call <code>typeof(1.0 + 2.0im)</code> it will be <code>Complex{Float64}</code>, which a complex number with the parts represented by <code>Float64</code>.</p></div></div><p>There are also some cool things like an irrational representation of π.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; π</code><code class="nohighlight hljs ansi" style="display:block;">π = 3.1415926535897...</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>To make π (and most other Greek letters), type <code>\pi</code> and then press <code>[TAB]</code>.</p></div></div><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(π)</code><code class="nohighlight hljs ansi" style="display:block;">Irrational{:π}</code></pre><p>However, if we do math with irrational numbers, they get converted to <code>Float64</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(2π / 3)</code><code class="nohighlight hljs ansi" style="display:block;">Float64</code></pre><h3 id="Floating-point-numbers"><a class="docs-heading-anchor" href="#Floating-point-numbers">Floating point numbers</a><a id="Floating-point-numbers-1"></a><a class="docs-heading-anchor-permalink" href="#Floating-point-numbers" title="Permalink"></a></h3><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>If you aren&#39;t familiar with floating point numbers, make sure to read this section carefully.</p></div></div><p>A <code>Float64</code> is a <a href="https://en.wikipedia.org/wiki/Floating-point_arithmetic">floating point</a> approximation of a real number using 64-bits of information.</p><p>Because it is an approximation, things we know hold true in mathematics don&#39;t hold true in a computer. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 0.1 * 3 == 0.3</code><code class="nohighlight hljs ansi" style="display:block;">false</code></pre><p>A more complicated example is:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; sin(2π / 3) == √3 / 2</code><code class="nohighlight hljs ansi" style="display:block;">false</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Get <code>√</code> by typing <code>\sqrt</code> then press <code>[TAB]</code>.</p></div></div><p>Let&#39;s see what the differences are:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 0.1 * 3 - 0.3</code><code class="nohighlight hljs ansi" style="display:block;">5.551115123125783e-17</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; sin(2π / 3) - √3 / 2</code><code class="nohighlight hljs ansi" style="display:block;">1.1102230246251565e-16</code></pre><p>They are small, but not zero.</p><p>One way of explaining this difference is to consider how we would write <code>1 / 3</code> and <code>2 / 3</code> using only four digits after the decimal point. We would write <code>1 / 3</code> as <code>0.3333</code>, and <code>2 / 3</code> as <code>0.6667</code>. So, despite the fact that <code>2 * (1 / 3) == 2 / 3</code>, <code>2 * 0.3333 == 0.6666 != 0.6667</code>.</p><p>Let&#39;s try that again using ≈ (<code>\approx + [TAB]</code>) instead of <code>==</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 0.1 * 3 ≈ 0.3</code><code class="nohighlight hljs ansi" style="display:block;">true</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; sin(2π / 3) ≈ √3 / 2</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><p><code>≈</code> is a clever way of calling the <code>isapprox</code> function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; isapprox(sin(2π / 3), √3 / 2; atol = 1e-8)</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Floating point is the reason solvers use tolerances when they solve optimization models. A common mistake you&#39;re likely to make is checking whether a binary variable is 0 using <code>value(z) == 0</code>. Always remember to use something like <code>isapprox</code> when comparing floating point numbers.</p></div></div><p>Note that <code>isapprox</code> will always return <code>false</code> if one of the number being compared is <code>0</code> and <code>atol</code> is zero (its default value).</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 1e-300 ≈ 0.0</code><code class="nohighlight hljs ansi" style="display:block;">false</code></pre><p>so always set a nonzero value of <code>atol</code> if one of the arguments can be zero.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; isapprox(1e-9, 0.0; atol = 1e-8)</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Gurobi has a <a href="https://www.gurobi.com/documentation/9.0/refman/num_grb_guidelines_for_num.html">good series of articles</a> on the implications of floating point in optimization if you want to read more.</p></div></div><p>If you aren&#39;t careful, floating point arithmetic can throw up all manner of issues. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 1 + 1e-16 == 1</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><p>It even turns out that floating point numbers aren&#39;t associative:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; (1 + 1e-16) - 1e-16 == 1 + (1e-16 - 1e-16)</code><code class="nohighlight hljs ansi" style="display:block;">false</code></pre><p>It&#39;s important to note that this issue isn&#39;t Julia-specific. It happens in every programming language (try it out in Python).</p><h2 id="Vectors,-matrices-and-arrays"><a class="docs-heading-anchor" href="#Vectors,-matrices-and-arrays">Vectors, matrices and arrays</a><a id="Vectors,-matrices-and-arrays-1"></a><a class="docs-heading-anchor-permalink" href="#Vectors,-matrices-and-arrays" title="Permalink"></a></h2><p>Similar to MATLAB, Julia has native support for vectors, matrices and tensors; all of which are represented by arrays of different dimensions. Vectors are constructed by comma-separated elements surrounded by square brackets:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; b = [5, 6]</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{Int64}:
+ 5
+ 6</code></pre><p>Matrices can be constructed with spaces separating the columns, and semicolons separating the rows:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; A = [1.0 2.0; 3.0 4.0]</code><code class="nohighlight hljs ansi" style="display:block;">2×2 Matrix{Float64}:
+ 1.0  2.0
+ 3.0  4.0</code></pre><p>We can do linear algebra:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; x = A \ b</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{Float64}:
+ -3.9999999999999987
+  4.499999999999999</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Here is floating point at work again; <code>x</code> is approximately <code>[-4, 4.5]</code>.</p></div></div><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; A * x</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{Float64}:
+ 5.0
+ 6.0</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; A * x ≈ b</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><p>Note that when multiplying vectors and matrices, dimensions matter. For example, you can&#39;t multiply a vector by a vector:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt;     b * b</code><code class="nohighlight hljs ansi" style="display:block;">MethodError: no method matching *(::Vector{Int64}, ::Vector{Int64})
+
+Closest candidates are:
+  *(::Any, ::Any, <span class="sgr91">::Any</span>, <span class="sgr91">::Any...</span>)
+<span class="sgr90">   @</span> <span class="sgr90">Base</span> <span class="sgr90"><span class="sgr4">operators.jl:578</span></span>
+  *(<span class="sgr91">::StridedMatrix{T}</span>, ::StridedVector{S}) where {T&lt;:Union{Float32, Float64, ComplexF64, ComplexF32}, S&lt;:Real}
+<span class="sgr90">   @</span> <span class="sgr35">LinearAlgebra</span> <span class="sgr90">/opt/hostedtoolcache/julia/1.9.3/x64/share/julia/stdlib/v1.9/LinearAlgebra/src/<span class="sgr4">matmul.jl:49</span></span>
+  *(::StridedVecOrMat, <span class="sgr91">::LinearAlgebra.Adjoint{&lt;:Any, &lt;:LinearAlgebra.LQPackedQ}</span>)
+<span class="sgr90">   @</span> <span class="sgr35">LinearAlgebra</span> <span class="sgr90">/opt/hostedtoolcache/julia/1.9.3/x64/share/julia/stdlib/v1.9/LinearAlgebra/src/<span class="sgr4">lq.jl:269</span></span>
+  ...</code></pre><p>But multiplying transposes works:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; b&#39; * b</code><code class="nohighlight hljs ansi" style="display:block;">61</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; b * b&#39;</code><code class="nohighlight hljs ansi" style="display:block;">2×2 Matrix{Int64}:
+ 25  30
+ 30  36</code></pre><h2 id="Other-common-types"><a class="docs-heading-anchor" href="#Other-common-types">Other common types</a><a id="Other-common-types-1"></a><a class="docs-heading-anchor-permalink" href="#Other-common-types" title="Permalink"></a></h2><h3 id="Comments"><a class="docs-heading-anchor" href="#Comments">Comments</a><a id="Comments-1"></a><a class="docs-heading-anchor-permalink" href="#Comments" title="Permalink"></a></h3><p>Although not technically a type, code comments begin with the <code>#</code> character:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; 1 + 1  # This is a comment</code><code class="nohighlight hljs ansi" style="display:block;">2</code></pre><p>Multiline comments begin with <code>#=</code> and end with <code>=#</code>:</p><pre><code class="language-julia hljs">#=
+Here is a
+multiline comment
+=#</code></pre><p>Comments can even be nested inside expressions. This is sometimes helpful when documenting inputs to functions:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; isapprox(
+           sin(π),
+           0.0;
+           #= We need an explicit atol here because we are comparing with 0 =#
+           atol = 0.001,
+       )</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><h3 id="Strings"><a class="docs-heading-anchor" href="#Strings">Strings</a><a id="Strings-1"></a><a class="docs-heading-anchor-permalink" href="#Strings" title="Permalink"></a></h3><p>Double quotes are used for strings:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(&quot;This is Julia&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">String</code></pre><p>Unicode is fine in strings:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(&quot;π is about 3.1415&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">String</code></pre><p>Use <code>println</code> to print a string:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; println(&quot;Hello, World!&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">Hello, World!</code></pre><p>Use <code>$()</code> to interpolate values into a string:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; x = 123</code><code class="nohighlight hljs ansi" style="display:block;">123</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; println(&quot;The value of x is: $(x)&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">The value of x is: 123</code></pre><p>Use triple-quotes for multiline strings:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; s = &quot;&quot;&quot;
+       Here is
+       a
+       multiline string
+       &quot;&quot;&quot;</code><code class="nohighlight hljs ansi" style="display:block;">&quot;Here is\na\nmultiline string\n&quot;</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; println(s)</code><code class="nohighlight hljs ansi" style="display:block;">Here is
+a
+multiline string</code></pre><h3 id="Symbols"><a class="docs-heading-anchor" href="#Symbols">Symbols</a><a id="Symbols-1"></a><a class="docs-heading-anchor-permalink" href="#Symbols" title="Permalink"></a></h3><p>Julia <code>Symbol</code>s are a data structure from the compiler that represent Julia identifiers (that is, variable names).</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; println(&quot;The value of x is: $(eval(:x))&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">The value of x is: 123</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>We used <code>eval</code> here to demonstrate how Julia links <code>Symbol</code>s to variables. However, avoid calling <code>eval</code> in your code. It is usually a sign that your code is doing something that could be more easily achieved a different way. The <a href="https://jump.dev/forum">Community Forum</a> is a good place to ask for advice on alternative approaches.</p></div></div><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(:x)</code><code class="nohighlight hljs ansi" style="display:block;">Symbol</code></pre><p>You can think of a <code>Symbol</code> as a <code>String</code> that takes up less memory, and that can&#39;t be modified.</p><p>Convert between <code>String</code> and <code>Symbol</code> using their constructors:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; String(:abc)</code><code class="nohighlight hljs ansi" style="display:block;">&quot;abc&quot;</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; Symbol(&quot;abc&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">:abc</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p><code>Symbol</code>s are often (ab)used to stand in for a <code>String</code> or an <code>Enum</code>, when one of the latter is likely a better choice. The JuMP <a href="../../../developers/style/#Style-guide">Style guide</a> recommends reserving <code>Symbol</code>s for identifiers. See <a href="../../../developers/style/#@enum-vs.-Symbol">@enum vs. Symbol</a> for more.</p></div></div><h3 id="Tuples"><a class="docs-heading-anchor" href="#Tuples">Tuples</a><a id="Tuples-1"></a><a class="docs-heading-anchor-permalink" href="#Tuples" title="Permalink"></a></h3><p>Julia makes extensive use of a simple data structure called Tuples. Tuples are immutable collections of values. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; t = (&quot;hello&quot;, 1.2, :foo)</code><code class="nohighlight hljs ansi" style="display:block;">(&quot;hello&quot;, 1.2, :foo)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; typeof(t)</code><code class="nohighlight hljs ansi" style="display:block;">Tuple{String, Float64, Symbol}</code></pre><p>Tuples can be accessed by index, similar to arrays:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; t[2]</code><code class="nohighlight hljs ansi" style="display:block;">1.2</code></pre><p>And they can be &quot;unpacked&quot; like so:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; a, b, c = t</code><code class="nohighlight hljs ansi" style="display:block;">(&quot;hello&quot;, 1.2, :foo)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; b</code><code class="nohighlight hljs ansi" style="display:block;">1.2</code></pre><p>The values can also be given names, which is a convenient way of making light-weight data structures.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; t = (word = &quot;hello&quot;, num = 1.2, sym = :foo)</code><code class="nohighlight hljs ansi" style="display:block;">(word = &quot;hello&quot;, num = 1.2, sym = :foo)</code></pre><p>Values can be accessed using dot syntax:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; t.word</code><code class="nohighlight hljs ansi" style="display:block;">&quot;hello&quot;</code></pre><h2 id="Dictionaries"><a class="docs-heading-anchor" href="#Dictionaries">Dictionaries</a><a id="Dictionaries-1"></a><a class="docs-heading-anchor-permalink" href="#Dictionaries" title="Permalink"></a></h2><p>Similar to Python, Julia has native support for dictionaries. Dictionaries provide a very generic way of mapping keys to values.  For example, a map of integers to strings:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; d1 = Dict(1 =&gt; &quot;A&quot;, 2 =&gt; &quot;B&quot;, 4 =&gt; &quot;D&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">Dict{Int64, String} with 3 entries:
+  4 =&gt; &quot;D&quot;
+  2 =&gt; &quot;B&quot;
+  1 =&gt; &quot;A&quot;</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Type-stuff again: <code>Dict{Int64,String}</code> is a dictionary with <code>Int64</code> keys and <code>String</code> values.</p></div></div><p>Looking up a value uses the bracket syntax:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; d1[2]</code><code class="nohighlight hljs ansi" style="display:block;">&quot;B&quot;</code></pre><p>Dictionaries support non-integer keys and can mix data types:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; Dict(&quot;A&quot; =&gt; 1, &quot;B&quot; =&gt; 2.5, &quot;D&quot; =&gt; 2 - 3im)</code><code class="nohighlight hljs ansi" style="display:block;">Dict{String, Number} with 3 entries:
+  &quot;B&quot; =&gt; 2.5
+  &quot;A&quot; =&gt; 1
+  &quot;D&quot; =&gt; 2-3im</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Julia types form a hierarchy. Here the value type of the dictionary is <code>Number</code>, which is a generalization of <code>Int64</code>, <code>Float64</code>, and <code>Complex{Int}</code>. Leaf nodes in this hierarchy are called &quot;concrete&quot; types, and all others are called &quot;Abstract.&quot; In general, having variables with abstract types like <code>Number</code> can lead to slower code, so you should try to make sure every element in a dictionary or vector is the same type. For example, in this case we could represent every element as a <code>Complex{Float64}</code>:</p></div></div><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; Dict(&quot;A&quot; =&gt; 1.0 + 0.0im, &quot;B&quot; =&gt; 2.5 + 0.0im, &quot;D&quot; =&gt; 2.0 - 3.0im)</code><code class="nohighlight hljs ansi" style="display:block;">Dict{String, ComplexF64} with 3 entries:
+  &quot;B&quot; =&gt; 2.5+0.0im
+  &quot;A&quot; =&gt; 1.0+0.0im
+  &quot;D&quot; =&gt; 2.0-3.0im</code></pre><p>Dictionaries can be nested:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; d2 = Dict(&quot;A&quot; =&gt; 1, &quot;B&quot; =&gt; 2, &quot;D&quot; =&gt; Dict(:foo =&gt; 3, :bar =&gt; 4))</code><code class="nohighlight hljs ansi" style="display:block;">Dict{String, Any} with 3 entries:
+  &quot;B&quot; =&gt; 2
+  &quot;A&quot; =&gt; 1
+  &quot;D&quot; =&gt; Dict(:bar=&gt;4, :foo=&gt;3)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; d2[&quot;B&quot;]</code><code class="nohighlight hljs ansi" style="display:block;">2</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; d2[&quot;D&quot;][:foo]</code><code class="nohighlight hljs ansi" style="display:block;">3</code></pre><h2 id="Structs"><a class="docs-heading-anchor" href="#Structs">Structs</a><a id="Structs-1"></a><a class="docs-heading-anchor-permalink" href="#Structs" title="Permalink"></a></h2><p>You can define custom datastructures with <code>struct</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; struct MyStruct
+           x::Int
+           y::String
+           z::Dict{Int,Int}
+       end</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; a = MyStruct(1, &quot;a&quot;, Dict(2 =&gt; 3))</code><code class="nohighlight hljs ansi" style="display:block;">Main.MyStruct(1, &quot;a&quot;, Dict(2 =&gt; 3))</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; a.x</code><code class="nohighlight hljs ansi" style="display:block;">1</code></pre><p>By default, these are not mutable</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt;     a.x = 2</code><code class="nohighlight hljs ansi" style="display:block;">setfield!: immutable struct of type MyStruct cannot be changed</code></pre><p>However, you can declare a <code>mutable struct</code> which is mutable:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; mutable struct MyStructMutable
+           x::Int
+           y::String
+           z::Dict{Int,Int}
+       end</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; a = MyStructMutable(1, &quot;a&quot;, Dict(2 =&gt; 3))</code><code class="nohighlight hljs ansi" style="display:block;">Main.MyStructMutable(1, &quot;a&quot;, Dict(2 =&gt; 3))</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; a.x</code><code class="nohighlight hljs ansi" style="display:block;">1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; a.x = 2</code><code class="nohighlight hljs ansi" style="display:block;">2</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; a</code><code class="nohighlight hljs ansi" style="display:block;">Main.MyStructMutable(2, &quot;a&quot;, Dict(2 =&gt; 3))</code></pre><h2 id="Loops"><a class="docs-heading-anchor" href="#Loops">Loops</a><a id="Loops-1"></a><a class="docs-heading-anchor-permalink" href="#Loops" title="Permalink"></a></h2><p>Julia has native support for for-each style loops with the syntax <code>for &lt;value&gt; in &lt;collection&gt; end</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; for i in 1:5
+           println(i)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">1
+2
+3
+4
+5</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Ranges are constructed as <code>start:stop</code>, or <code>start:step:stop</code>.</p></div></div><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; for i in 1.2:1.1:5.6
+           println(i)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">1.2
+2.3
+3.4
+4.5
+5.6</code></pre><p>This for-each loop also works with dictionaries:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; for (key, value) in Dict(&quot;A&quot; =&gt; 1, &quot;B&quot; =&gt; 2.5, &quot;D&quot; =&gt; 2 - 3im)
+           println(&quot;$(key): $(value)&quot;)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">B: 2.5
+A: 1
+D: 2 - 3im</code></pre><p>Note that in contrast to vector languages like MATLAB and R, loops do not result in a significant performance degradation in Julia.</p><h2 id="Control-flow"><a class="docs-heading-anchor" href="#Control-flow">Control flow</a><a id="Control-flow-1"></a><a class="docs-heading-anchor-permalink" href="#Control-flow" title="Permalink"></a></h2><p>Julia control flow is similar to MATLAB, using the keywords <code>if-elseif-else-end</code>, and the logical operators <code>||</code> and <code>&amp;&amp;</code> for <strong>or</strong> and <strong>and</strong> respectively:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; for i in 0:5:15
+           if i &lt; 5
+               println(&quot;$(i) is less than 5&quot;)
+           elseif i &lt; 10
+               println(&quot;$(i) is less than 10&quot;)
+           else
+               if i == 10
+                   println(&quot;the value is 10&quot;)
+               else
+                   println(&quot;$(i) is bigger than 10&quot;)
+               end
+           end
+       end</code><code class="nohighlight hljs ansi" style="display:block;">0 is less than 5
+5 is less than 10
+the value is 10
+15 is bigger than 10</code></pre><h2 id="Comprehensions"><a class="docs-heading-anchor" href="#Comprehensions">Comprehensions</a><a id="Comprehensions-1"></a><a class="docs-heading-anchor-permalink" href="#Comprehensions" title="Permalink"></a></h2><p>Similar to languages like Haskell and Python, Julia supports the use of simple loops in the construction of arrays and dictionaries, called comprehensions.</p><p>A list of increasing integers:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; [i for i in 1:5]</code><code class="nohighlight hljs ansi" style="display:block;">5-element Vector{Int64}:
+ 1
+ 2
+ 3
+ 4
+ 5</code></pre><p>Matrices can be built by including multiple indices:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; [i * j for i in 1:5, j in 5:10]</code><code class="nohighlight hljs ansi" style="display:block;">5×6 Matrix{Int64}:
+  5   6   7   8   9  10
+ 10  12  14  16  18  20
+ 15  18  21  24  27  30
+ 20  24  28  32  36  40
+ 25  30  35  40  45  50</code></pre><p>Conditional statements can be used to filter out some values:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; [i for i in 1:10 if i % 2 == 1]</code><code class="nohighlight hljs ansi" style="display:block;">5-element Vector{Int64}:
+ 1
+ 3
+ 5
+ 7
+ 9</code></pre><p>A similar syntax can be used for building dictionaries:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; Dict(&quot;$(i)&quot; =&gt; i for i in 1:10 if i % 2 == 1)</code><code class="nohighlight hljs ansi" style="display:block;">Dict{String, Int64} with 5 entries:
+  &quot;1&quot; =&gt; 1
+  &quot;5&quot; =&gt; 5
+  &quot;7&quot; =&gt; 7
+  &quot;9&quot; =&gt; 9
+  &quot;3&quot; =&gt; 3</code></pre><h2 id="Functions"><a class="docs-heading-anchor" href="#Functions">Functions</a><a id="Functions-1"></a><a class="docs-heading-anchor-permalink" href="#Functions" title="Permalink"></a></h2><p>A simple function is defined as follows:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; function print_hello()
+           return println(&quot;hello&quot;)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">print_hello (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; print_hello()</code><code class="nohighlight hljs ansi" style="display:block;">hello</code></pre><p>Arguments can be added to a function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; function print_it(x)
+           return println(x)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">print_it (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; print_it(&quot;hello&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">hello</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; print_it(1.234)</code><code class="nohighlight hljs ansi" style="display:block;">1.234</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; print_it(:my_id)</code><code class="nohighlight hljs ansi" style="display:block;">my_id</code></pre><p>Optional keyword arguments are also possible:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; function print_it(x; prefix = &quot;value:&quot;)
+           return println(&quot;$(prefix) $(x)&quot;)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">print_it (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; print_it(1.234)</code><code class="nohighlight hljs ansi" style="display:block;">value: 1.234</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; print_it(1.234; prefix = &quot;val:&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">val: 1.234</code></pre><p>The keyword <code>return</code> is used to specify the return values of a function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; function mult(x; y = 2.0)
+           return x * y
+       end</code><code class="nohighlight hljs ansi" style="display:block;">mult (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; mult(4.0)</code><code class="nohighlight hljs ansi" style="display:block;">8.0</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; mult(4.0; y = 5.0)</code><code class="nohighlight hljs ansi" style="display:block;">20.0</code></pre><h3 id="Anonymous-functions"><a class="docs-heading-anchor" href="#Anonymous-functions">Anonymous functions</a><a id="Anonymous-functions-1"></a><a class="docs-heading-anchor-permalink" href="#Anonymous-functions" title="Permalink"></a></h3><p>The syntax <code>input -&gt; output</code> creates an anonymous function. These are most useful when passed to other functions. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; f = x -&gt; x^2</code><code class="nohighlight hljs ansi" style="display:block;">#11 (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; f(2)</code><code class="nohighlight hljs ansi" style="display:block;">4</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; map(x -&gt; x^2, 1:4)</code><code class="nohighlight hljs ansi" style="display:block;">4-element Vector{Int64}:
+  1
+  4
+  9
+ 16</code></pre><h3 id="Type-parameters"><a class="docs-heading-anchor" href="#Type-parameters">Type parameters</a><a id="Type-parameters-1"></a><a class="docs-heading-anchor-permalink" href="#Type-parameters" title="Permalink"></a></h3><p>We can constrain the inputs to a function using type parameters, which are <code>::</code> followed by the type of the input we want. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; function foo(x::Int)
+           return x^2
+       end</code><code class="nohighlight hljs ansi" style="display:block;">foo (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; function foo(x::Float64)
+           return exp(x)
+       end</code><code class="nohighlight hljs ansi" style="display:block;">foo (generic function with 2 methods)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; function foo(x::Number)
+           return x + 1
+       end</code><code class="nohighlight hljs ansi" style="display:block;">foo (generic function with 3 methods)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; foo(2)</code><code class="nohighlight hljs ansi" style="display:block;">4</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; foo(2.0)</code><code class="nohighlight hljs ansi" style="display:block;">7.38905609893065</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; foo(1 + 1im)</code><code class="nohighlight hljs ansi" style="display:block;">2 + 1im</code></pre><p>But what happens if we call <code>foo</code> with something we haven&#39;t defined it for?</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt;     foo([1, 2, 3])</code><code class="nohighlight hljs ansi" style="display:block;">MethodError: no method matching foo(::Vector{Int64})
+
+Closest candidates are:
+  foo(<span class="sgr91">::Int64</span>)
+<span class="sgr90">   @</span> <span class="sgr36">Main</span> <span class="sgr90"><span class="sgr4">REPL[1]:1</span></span>
+  foo(<span class="sgr91">::Float64</span>)
+<span class="sgr90">   @</span> <span class="sgr36">Main</span> <span class="sgr90"><span class="sgr4">REPL[2]:1</span></span>
+  foo(<span class="sgr91">::Number</span>)
+<span class="sgr90">   @</span> <span class="sgr36">Main</span> <span class="sgr90"><span class="sgr4">REPL[3]:1</span></span></code></pre><p>A <code>MethodError</code> means that you passed a function something that didn&#39;t match the type that it was expecting. In this case, the error message says that it doesn&#39;t know how to handle an <code>Vector{Int64}</code>, but it does know how to handle <code>Float64</code>, <code>Int64</code>, and <code>Number</code>.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Read the &quot;Closest candidates&quot; part of the error message carefully to get a hint as to what was expected.</p></div></div><h3 id="Broadcasting"><a class="docs-heading-anchor" href="#Broadcasting">Broadcasting</a><a id="Broadcasting-1"></a><a class="docs-heading-anchor-permalink" href="#Broadcasting" title="Permalink"></a></h3><p>In the example above, we didn&#39;t define what to do if <code>f</code> was passed a <code>Vector</code>. Luckily, Julia provides a convenient syntax for mapping <code>f</code> element-wise over arrays. Just add a <code>.</code> between the name of the function and the opening <code>(</code>. This works for <em>any</em> function, including functions with multiple arguments. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; f.([1, 2, 3])</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{Int64}:
+ 1
+ 4
+ 9</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>Get a <code>MethodError</code> when calling a function that takes a <code>Vector</code>, <code>Matrix</code>, or <code>Array</code>? Try broadcasting.</p></div></div><h2 id="Mutable-vs-immutable-objects"><a class="docs-heading-anchor" href="#Mutable-vs-immutable-objects">Mutable vs immutable objects</a><a id="Mutable-vs-immutable-objects-1"></a><a class="docs-heading-anchor-permalink" href="#Mutable-vs-immutable-objects" title="Permalink"></a></h2><p>Some types in Julia are <em>mutable</em>, which means you can change the values inside them. A good example is an array. You can modify the contents of an array without having to make a new array.</p><p>In contrast, types like <code>Float64</code> are <em>immutable</em>. You cannot modify the contents of a <code>Float64</code>.</p><p>This is something to be aware of when passing types into functions. For example:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; function mutability_example(mutable_type::Vector{Int}, immutable_type::Int)
+           mutable_type[1] += 1
+           immutable_type += 1
+           return
+       end</code><code class="nohighlight hljs ansi" style="display:block;">mutability_example (generic function with 1 method)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; mutable_type = [1, 2, 3]</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{Int64}:
+ 1
+ 2
+ 3</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; immutable_type = 1</code><code class="nohighlight hljs ansi" style="display:block;">1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; mutability_example(mutable_type, immutable_type)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; println(&quot;mutable_type: $(mutable_type)&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">mutable_type: [2, 2, 3]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; println(&quot;immutable_type: $(immutable_type)&quot;)</code><code class="nohighlight hljs ansi" style="display:block;">immutable_type: 1</code></pre><p>Because <code>Vector{Int}</code> is a mutable type, modifying the variable inside the function changed the value outside of the function. In contrast, the change to <code>immutable_type</code> didn&#39;t modify the value outside the function.</p><p>You can check mutability with the <code>isimmutable</code> function:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; isimmutable([1, 2, 3])</code><code class="nohighlight hljs ansi" style="display:block;">false</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; isimmutable(1)</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><h2 id="The-package-manager"><a class="docs-heading-anchor" href="#The-package-manager">The package manager</a><a id="The-package-manager-1"></a><a class="docs-heading-anchor-permalink" href="#The-package-manager" title="Permalink"></a></h2><h3 id="Installing-packages"><a class="docs-heading-anchor" href="#Installing-packages">Installing packages</a><a id="Installing-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Installing-packages" title="Permalink"></a></h3><p>No matter how wonderful Julia&#39;s base language is, at some point you will want to use an extension package.  Some of these are built-in, for example random number generation is available in the <code>Random</code> package in the standard library. These packages are loaded with the commands <code>using</code> and <code>import</code>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using Random  # The equivalent of Python&#39;s `from Random import *`</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; import Random  # The equivalent of Python&#39;s `import Random`</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; Random.seed!(33)</code><code class="nohighlight hljs ansi" style="display:block;">Random.TaskLocalRNG()</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; [rand() for i in 1:10]</code><code class="nohighlight hljs ansi" style="display:block;">10-element Vector{Float64}:
+ 0.4745319377345316
+ 0.9650392357070774
+ 0.8194019096093067
+ 0.9297749959069098
+ 0.3127122336048005
+ 0.9684448191382753
+ 0.9063743823581542
+ 0.8386731983150535
+ 0.5103924401614957
+ 0.9296414851080324</code></pre><p>The Package Manager is used to install packages that are not part of Julia&#39;s standard library.</p><p>For example the following can be used to install JuMP,</p><pre><code class="language-julia hljs">using Pkg
+Pkg.add(&quot;JuMP&quot;)</code></pre><p>For a complete list of registered Julia packages see the package listing at <a href="https://juliahub.com">JuliaHub</a>.</p><p>From time to you may wish to use a Julia package that is not registered.  In this case a git repository URL can be used to install the package.</p><pre><code class="language-julia hljs">using Pkg
+Pkg.add(&quot;https://github.com/user-name/MyPackage.jl.git&quot;)</code></pre><h3 id="Package-environments"><a class="docs-heading-anchor" href="#Package-environments">Package environments</a><a id="Package-environments-1"></a><a class="docs-heading-anchor-permalink" href="#Package-environments" title="Permalink"></a></h3><p>By default, <code>Pkg.add</code> will add packages to Julia&#39;s global environment. However, Julia also has built-in support for virtual environments.</p><p>Activate a virtual environment with:</p><pre><code class="language-julia hljs">import Pkg; Pkg.activate(&quot;/path/to/environment&quot;)</code></pre><p>You can see what packages are installed in the current environment with <code>Pkg.status()</code>.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>We <em>strongly</em> recommend you create a Pkg environment for each project that you create in Julia, and add only the packages that you need, instead of adding lots of packages to the global environment. The <a href="https://julialang.github.io/Pkg.jl/v1/environments/">Pkg manager documentation</a> has more information on this topic.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../introduction/">« Introduction</a><a class="docs-footer-nextpage" href="../getting_started_with_JuMP/">Getting started with JuMP »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing.jl b/previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing.jl
new file mode 100644
index 00000000000..8b1e550d603
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing.jl
@@ -0,0 +1,293 @@
+# # Getting started with sets and indexing
+
+# Most introductory courses to linear programming will teach you to identify
+# sets over which the decision variables and constraints are indexed. Therefore,
+# it is common to write variables such as ``x_i`` for all ``i \in I``.
+
+# A common stumbling block for new users to JuMP is that _JuMP does not provide
+# specialized syntax for constructing and manipulating these sets_.
+
+# We made this decision because Julia already provides a wealth of data
+# structures for working with sets.
+
+# In contrast, because tools like AMPL are stand-alone software packages, they
+# had to define their own syntax for set construction and manipulation. Indeed,
+# the [AMPL Book](https://ampl.com/resources/the-ampl-book/chapter-downloads/)
+# has two entire chapters devoted to sets and indexing (Chapter 5, "Simple Sets
+# and Indexing," and Chapter 6, "Compound Sets and Indexing").
+
+# The purpose of this tutorial is to demonstrate a variety of ways in which you
+# can construct and manipulate sets for optimization models.
+
+# If you haven't already, you should first read
+# [Getting started with JuMP](@ref).
+
+using JuMP
+
+# ## Unordered sets
+
+# Unordered sets are useful to describe non-numeric indices, such as the names
+# of cities or types of products.
+
+# The most common way to construct a set is by creating a vector:
+
+animals = ["dog", "cat", "chicken", "cow", "pig"]
+model = Model()
+@variable(model, x[animals])
+
+# We can also use things like the `keys` of a dictionary:
+
+weight_of_animals = Dict(
+    "dog" => 20.0,
+    "cat" => 5.0,
+    "chicken" => 2.0,
+    "cow" => 720.0,
+    "pig" => 150.0,
+)
+animal_keys = keys(weight_of_animals)
+model = Model()
+@variable(model, x[animal_keys])
+
+# A third option is to use Julia's `Set` object.
+
+animal_set = Set()
+for animal in keys(weight_of_animals)
+    push!(animal_set, animal)
+end
+animal_set
+
+# The nice thing about `Set`s is that they automatically remove duplicates:
+
+push!(animal_set, "dog")
+animal_set
+
+# Note how `dog` does not appear twice.
+
+model = Model()
+@variable(model, x[animal_set])
+
+# ## Sets of numbers
+
+# Sets of numbers are useful to describe sets that are ordered, such as years or
+# elements in a vector. The easiest way to create sets of numbers is to use
+# Julia's `range` syntax.
+
+# These can start at `1`:
+
+model = Model()
+@variable(model, x[1:4])
+
+# but they don't have to:
+
+model = Model()
+@variable(model, x[2012:2021])
+
+# Ranges also have a `start:step:stop` syntax. So the Olympic years are:
+
+model = Model()
+@variable(model, x[1896:4:2020])
+
+# ## Sets of other things
+
+# An important observation is that you can have _any_ Julia type as the element
+# of a set. It doesn't have to be a `String` or a `Number`. For example, you
+# can have tuples:
+
+sources = ["A", "B", "C"]
+sinks = ["D", "E"]
+S = [(source, sink) for source in sources, sink in sinks]
+model = Model()
+@variable(model, x[S])
+
+#-
+
+x[("A", "D")]
+
+# For multi-dimensional sets, you can use JuMP's syntax for constructing
+# [Containers](@ref):
+
+model = Model()
+@variable(model, x[sources, sinks])
+
+#-
+
+x["A", "D"]
+
+# !!! info
+#     Note how we indexed `x["A", "D"]` instead of `x[("A", "D")]` as above.
+
+# ## Sets to watch out for
+
+# JuMP supports any sets which are [iterable] (https://docs.julialang.org/en/v1/base/collections/),
+# that is, the set `set` supports a for-loop like: `[i for i in set]`. This
+# causes a few common errors.
+
+# First, if `T = 3`, you may pass the integer `T` by mistake instead of a range
+# like `1:T`:
+model = Model()
+T = 3
+@variable(model, x[T])
+
+# This results in a single variable being created, instead of three as desired.
+# Because this is a common error, a warning is printed, advising you to pass a
+# `Vector{Int}` instead:
+
+@variable(model, x_fixed[[T]])
+
+# Second, because `String`s are iterable, passing a `"index"` as a singleton
+# index is the same as passing `['i', 'n', 'd', 'e', 'x']`:
+
+@variable(model, y["index"])
+
+# This time, a warning is not printed, but the work-around is similar, pass a
+# `Vector{String}` instead:
+
+@variable(model, y_fixed[["index"]])
+
+# !!! tip
+#     As a rule of thumb, if you want an index with one element, avoid confusion
+#     by passing `[index]` instead of `index`.
+
+# ## Set operations
+
+# Julia has built-in support for set operations such as `union`, `intersect`,
+# and `setdiff`.
+
+# Therefore, to create a set of all years in which the summer Olympics were
+# held, we can use:
+
+baseline = 1896:4:2020
+cancelled = [1916, 1940, 1944, 2020]
+off_year = [2021]
+olympic_years = union(setdiff(baseline, cancelled), off_year)
+
+# You can also find the number of elements (that is, the cardinality) in a set
+# using `length`:
+
+length(olympic_years)
+
+# ## Set membership operations
+
+# To compute membership of sets, use the `in` function.
+
+2000 in olympic_years
+
+#-
+
+2001 in olympic_years
+
+# ## Indexing expressions
+
+# Use Julia's generator syntax to compute new sets, such as the list of
+# Olympic years that are divisible by 3:
+
+olympic_3_years = [year for year in olympic_years if mod(year, 3) == 0]
+model = Model()
+@variable(model, x[olympic_3_years])
+
+# Alternatively, use JuMP's syntax for constructing [Containers](@ref):
+
+model = Model()
+@variable(model, x[year in olympic_years; mod(year, 3) == 0])
+
+# ## Compound sets
+
+# Consider a transportation problem in which we need to ship goods between
+# cities. We have been provided a list of cities:
+
+cities = ["Auckland", "Wellington", "Christchurch", "Dunedin"]
+
+# and a distance matrix which records the shipping distance between pairs of
+# cities. If we can't ship between two cities, the distance is `0`.
+
+distances = [0 643 1071 1426; 0 0 436 790; 0 0 0 360; 1426 0 0 0]
+
+# Let's have a look at ways we could write a model with an objective function to
+# minimize the total shipping cost. For simplicity, we'll ignore all
+# constraints.
+
+# ### Fix unused variables
+
+# One approach is to fix all variables that we can't use to zero. Most solvers
+# are smart-enough to remove these during a presolve phase, so it has a very
+# small impact on performance:
+
+N = length(cities)
+model = Model()
+@variable(model, x[1:N, 1:N] >= 0)
+for i in 1:N, j in 1:N
+    if distances[i, j] == 0
+        fix(x[i, j], 0.0; force = true)
+    end
+end
+@objective(model, Min, sum(distances[i, j] * x[i, j] for i in 1:N, j in 1:N))
+
+# ### Filtered summation
+
+# Another approach is to define filters whenever we want to sum over our
+# decision variables:
+
+N = length(cities)
+model = Model()
+@variable(model, x[1:N, 1:N] >= 0)
+@objective(
+    model,
+    Min,
+    sum(
+        distances[i, j] * x[i, j] for i in 1:N, j in 1:N if distances[i, j] > 0
+    ),
+)
+
+# ### Filtered indexing
+
+# We could also use JuMP's support for [Containers](@ref):
+
+N = length(cities)
+model = Model()
+@variable(model, x[i = 1:N, j = 1:N; distances[i, j] > 0])
+@objective(model, Min, sum(distances[i...] * x[i] for i in eachindex(x)))
+
+# !!! note
+#     The `i...` is called  a "splat." It converts a tuple like `(1, 2)` into
+#     two indices like `distances[1, 2]`.
+
+# ### Converting to a different data structure
+
+# Another approach, and one that is often the most readable, is to convert the
+# data you have into something that is easier to work with. Originally, we had
+# a vector of strings and a matrix of distances. What we really need is
+# something that maps usable origin-destination pairs to distances. A dictionary
+# is an obvious choice:
+
+routes = Dict(
+    (a, b) => distances[i, j] for
+    (i, a) in enumerate(cities), (j, b) in enumerate(cities) if
+    distances[i, j] > 0
+)
+
+# Then, we can create our model like so:
+
+model = Model()
+@variable(model, x[keys(routes)])
+@objective(model, Min, sum(v * x[k] for (k, v) in routes))
+
+# This has a number of benefits over the other approaches, including a compacter
+# algebraic model and variables that are named in a more meaningful way.
+
+# !!! tip
+#     If you're struggling to formulate a problem using the available syntax in
+#     JuMP, it's probably a sign that you should convert your data into a
+#     different form.
+
+# ## Next steps
+
+# The purpose of this tutorial was to show how JuMP does not have specialized
+# syntax for set creation and manipulation. Instead, you should use the tools
+# provided by Julia itself.
+
+# This is both an opportunity and a challenge, because you are free to pick the
+# syntax and data structures that best suit your problem, but for new users it
+# can be daunting to decide which structure to use.
+
+# Read through some of the other JuMP tutorials to get inspiration and ideas for
+# how you can use Julia's syntax and data structures to your advantage.
diff --git a/previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing/index.html b/previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing/index.html
new file mode 100644
index 00000000000..89e2fe95e1b
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/getting_started_with_sets_and_indexing/index.html
@@ -0,0 +1,222 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Getting started with sets and indexing · JuMP</title><meta name="title" content="Getting started with sets and indexing · JuMP"/><meta property="og:title" content="Getting started with sets and indexing · JuMP"/><meta property="twitter:title" content="Getting started with sets and indexing · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../getting_started_with_JuMP/">Getting started with JuMP</a></li><li class="is-active"><a class="tocitem" href>Getting started with sets and indexing</a><ul class="internal"><li><a class="tocitem" href="#Unordered-sets"><span>Unordered sets</span></a></li><li><a class="tocitem" href="#Sets-of-numbers"><span>Sets of numbers</span></a></li><li><a class="tocitem" href="#Sets-of-other-things"><span>Sets of other things</span></a></li><li><a class="tocitem" href="#Sets-to-watch-out-for"><span>Sets to watch out for</span></a></li><li><a class="tocitem" href="#Set-operations"><span>Set operations</span></a></li><li><a class="tocitem" href="#Set-membership-operations"><span>Set membership operations</span></a></li><li><a class="tocitem" href="#Indexing-expressions"><span>Indexing expressions</span></a></li><li><a class="tocitem" href="#Compound-sets"><span>Compound sets</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../debugging/">Debugging</a></li><li><a class="tocitem" href="../design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Getting started with sets and indexing</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Getting started with sets and indexing</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/getting_started_with_sets_and_indexing.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Getting-started-with-sets-and-indexing"><a class="docs-heading-anchor" href="#Getting-started-with-sets-and-indexing">Getting started with sets and indexing</a><a id="Getting-started-with-sets-and-indexing-1"></a><a class="docs-heading-anchor-permalink" href="#Getting-started-with-sets-and-indexing" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../getting_started_with_sets_and_indexing.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>Most introductory courses to linear programming will teach you to identify sets over which the decision variables and constraints are indexed. Therefore, it is common to write variables such as <span>$x_i$</span> for all <span>$i \in I$</span>.</p><p>A common stumbling block for new users to JuMP is that <em>JuMP does not provide specialized syntax for constructing and manipulating these sets</em>.</p><p>We made this decision because Julia already provides a wealth of data structures for working with sets.</p><p>In contrast, because tools like AMPL are stand-alone software packages, they had to define their own syntax for set construction and manipulation. Indeed, the <a href="https://ampl.com/resources/the-ampl-book/chapter-downloads/">AMPL Book</a> has two entire chapters devoted to sets and indexing (Chapter 5, &quot;Simple Sets and Indexing,&quot; and Chapter 6, &quot;Compound Sets and Indexing&quot;).</p><p>The purpose of this tutorial is to demonstrate a variety of ways in which you can construct and manipulate sets for optimization models.</p><p>If you haven&#39;t already, you should first read <a href="../getting_started_with_JuMP/#Getting-started-with-JuMP">Getting started with JuMP</a>.</p><pre><code class="language-julia hljs">using JuMP</code></pre><h2 id="Unordered-sets"><a class="docs-heading-anchor" href="#Unordered-sets">Unordered sets</a><a id="Unordered-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Unordered-sets" title="Permalink"></a></h2><p>Unordered sets are useful to describe non-numeric indices, such as the names of cities or types of products.</p><p>The most common way to construct a set is by creating a vector:</p><pre><code class="language-julia hljs">animals = [&quot;dog&quot;, &quot;cat&quot;, &quot;chicken&quot;, &quot;cow&quot;, &quot;pig&quot;]
+model = Model()
+@variable(model, x[animals])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [&quot;dog&quot;, &quot;cat&quot;, &quot;chicken&quot;, &quot;cow&quot;, &quot;pig&quot;]
+And data, a 5-element Vector{VariableRef}:
+ x[dog]
+ x[cat]
+ x[chicken]
+ x[cow]
+ x[pig]</code></pre><p>We can also use things like the <code>keys</code> of a dictionary:</p><pre><code class="language-julia hljs">weight_of_animals = Dict(
+    &quot;dog&quot; =&gt; 20.0,
+    &quot;cat&quot; =&gt; 5.0,
+    &quot;chicken&quot; =&gt; 2.0,
+    &quot;cow&quot; =&gt; 720.0,
+    &quot;pig&quot; =&gt; 150.0,
+)
+animal_keys = keys(weight_of_animals)
+model = Model()
+@variable(model, x[animal_keys])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [&quot;cow&quot;, &quot;chicken&quot;, &quot;cat&quot;, &quot;pig&quot;, &quot;dog&quot;]
+And data, a 5-element Vector{VariableRef}:
+ x[cow]
+ x[chicken]
+ x[cat]
+ x[pig]
+ x[dog]</code></pre><p>A third option is to use Julia&#39;s <code>Set</code> object.</p><pre><code class="language-julia hljs">animal_set = Set()
+for animal in keys(weight_of_animals)
+    push!(animal_set, animal)
+end
+animal_set</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Set{Any} with 5 elements:
+  &quot;cow&quot;
+  &quot;chicken&quot;
+  &quot;cat&quot;
+  &quot;pig&quot;
+  &quot;dog&quot;</code></pre><p>The nice thing about <code>Set</code>s is that they automatically remove duplicates:</p><pre><code class="language-julia hljs">push!(animal_set, &quot;dog&quot;)
+animal_set</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Set{Any} with 5 elements:
+  &quot;cow&quot;
+  &quot;chicken&quot;
+  &quot;cat&quot;
+  &quot;pig&quot;
+  &quot;dog&quot;</code></pre><p>Note how <code>dog</code> does not appear twice.</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[animal_set])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [&quot;cow&quot;, &quot;chicken&quot;, &quot;cat&quot;, &quot;pig&quot;, &quot;dog&quot;]
+And data, a 5-element Vector{VariableRef}:
+ x[cow]
+ x[chicken]
+ x[cat]
+ x[pig]
+ x[dog]</code></pre><h2 id="Sets-of-numbers"><a class="docs-heading-anchor" href="#Sets-of-numbers">Sets of numbers</a><a id="Sets-of-numbers-1"></a><a class="docs-heading-anchor-permalink" href="#Sets-of-numbers" title="Permalink"></a></h2><p>Sets of numbers are useful to describe sets that are ordered, such as years or elements in a vector. The easiest way to create sets of numbers is to use Julia&#39;s <code>range</code> syntax.</p><p>These can start at <code>1</code>:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[1:4])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+ x[4]</code></pre><p>but they don&#39;t have to:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[2012:2021])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, 2012:2021
+And data, a 10-element Vector{VariableRef}:
+ x[2012]
+ x[2013]
+ x[2014]
+ x[2015]
+ x[2016]
+ x[2017]
+ x[2018]
+ x[2019]
+ x[2020]
+ x[2021]</code></pre><p>Ranges also have a <code>start:step:stop</code> syntax. So the Olympic years are:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[1896:4:2020])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, 1896:4:2020
+And data, a 32-element Vector{VariableRef}:
+ x[1896]
+ x[1900]
+ x[1904]
+ x[1908]
+ x[1912]
+ x[1916]
+ x[1920]
+ x[1924]
+ x[1928]
+ x[1932]
+ ⋮
+ x[1988]
+ x[1992]
+ x[1996]
+ x[2000]
+ x[2004]
+ x[2008]
+ x[2012]
+ x[2016]
+ x[2020]</code></pre><h2 id="Sets-of-other-things"><a class="docs-heading-anchor" href="#Sets-of-other-things">Sets of other things</a><a id="Sets-of-other-things-1"></a><a class="docs-heading-anchor-permalink" href="#Sets-of-other-things" title="Permalink"></a></h2><p>An important observation is that you can have <em>any</em> Julia type as the element of a set. It doesn&#39;t have to be a <code>String</code> or a <code>Number</code>. For example, you can have tuples:</p><pre><code class="language-julia hljs">sources = [&quot;A&quot;, &quot;B&quot;, &quot;C&quot;]
+sinks = [&quot;D&quot;, &quot;E&quot;]
+S = [(source, sink) for source in sources, sink in sinks]
+model = Model()
+@variable(model, x[S])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [(&quot;A&quot;, &quot;D&quot;), (&quot;B&quot;, &quot;D&quot;), (&quot;C&quot;, &quot;D&quot;), (&quot;A&quot;, &quot;E&quot;), (&quot;B&quot;, &quot;E&quot;), (&quot;C&quot;, &quot;E&quot;)]
+And data, a 6-element Vector{VariableRef}:
+ x[(&quot;A&quot;, &quot;D&quot;)]
+ x[(&quot;B&quot;, &quot;D&quot;)]
+ x[(&quot;C&quot;, &quot;D&quot;)]
+ x[(&quot;A&quot;, &quot;E&quot;)]
+ x[(&quot;B&quot;, &quot;E&quot;)]
+ x[(&quot;C&quot;, &quot;E&quot;)]</code></pre><pre><code class="language-julia hljs">x[(&quot;A&quot;, &quot;D&quot;)]</code></pre><p>$ x_{(&quot;A&quot;, &quot;D&quot;)} $</p><p>For multi-dimensional sets, you can use JuMP&#39;s syntax for constructing <a href="../../../manual/containers/#Containers">Containers</a>:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[sources, sinks])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:
+    Dimension 1, [&quot;A&quot;, &quot;B&quot;, &quot;C&quot;]
+    Dimension 2, [&quot;D&quot;, &quot;E&quot;]
+And data, a 3×2 Matrix{VariableRef}:
+ x[A,D]  x[A,E]
+ x[B,D]  x[B,E]
+ x[C,D]  x[C,E]</code></pre><pre><code class="language-julia hljs">x[&quot;A&quot;, &quot;D&quot;]</code></pre><p>$ x_{A,D} $</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Note how we indexed <code>x[&quot;A&quot;, &quot;D&quot;]</code> instead of <code>x[(&quot;A&quot;, &quot;D&quot;)]</code> as above.</p></div></div><h2 id="Sets-to-watch-out-for"><a class="docs-heading-anchor" href="#Sets-to-watch-out-for">Sets to watch out for</a><a id="Sets-to-watch-out-for-1"></a><a class="docs-heading-anchor-permalink" href="#Sets-to-watch-out-for" title="Permalink"></a></h2><p>JuMP supports any sets which are <a href="https://docs.julialang.org/en/v1/base/collections/">iterable</a>, that is, the set <code>set</code> supports a for-loop like: <code>[i for i in set]</code>. This causes a few common errors.</p><p>First, if <code>T = 3</code>, you may pass the integer <code>T</code> by mistake instead of a range like <code>1:T</code>:</p><pre><code class="language-julia hljs">model = Model()
+T = 3
+@variable(model, x[T])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [3]
+And data, a 1-element Vector{VariableRef}:
+ x[3]</code></pre><p>This results in a single variable being created, instead of three as desired. Because this is a common error, a warning is printed, advising you to pass a <code>Vector{Int}</code> instead:</p><pre><code class="language-julia hljs">@variable(model, x_fixed[[T]])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [3]
+And data, a 1-element Vector{VariableRef}:
+ x_fixed[3]</code></pre><p>Second, because <code>String</code>s are iterable, passing a <code>&quot;index&quot;</code> as a singleton index is the same as passing <code>[&#39;i&#39;, &#39;n&#39;, &#39;d&#39;, &#39;e&#39;, &#39;x&#39;]</code>:</p><pre><code class="language-julia hljs">@variable(model, y[&quot;index&quot;])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [&#39;i&#39;, &#39;n&#39;, &#39;d&#39;, &#39;e&#39;, &#39;x&#39;]
+And data, a 5-element Vector{VariableRef}:
+ y[i]
+ y[n]
+ y[d]
+ y[e]
+ y[x]</code></pre><p>This time, a warning is not printed, but the work-around is similar, pass a <code>Vector{String}</code> instead:</p><pre><code class="language-julia hljs">@variable(model, y_fixed[[&quot;index&quot;]])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [&quot;index&quot;]
+And data, a 1-element Vector{VariableRef}:
+ y_fixed[index]</code></pre><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>As a rule of thumb, if you want an index with one element, avoid confusion by passing <code>[index]</code> instead of <code>index</code>.</p></div></div><h2 id="Set-operations"><a class="docs-heading-anchor" href="#Set-operations">Set operations</a><a id="Set-operations-1"></a><a class="docs-heading-anchor-permalink" href="#Set-operations" title="Permalink"></a></h2><p>Julia has built-in support for set operations such as <code>union</code>, <code>intersect</code>, and <code>setdiff</code>.</p><p>Therefore, to create a set of all years in which the summer Olympics were held, we can use:</p><pre><code class="language-julia hljs">baseline = 1896:4:2020
+cancelled = [1916, 1940, 1944, 2020]
+off_year = [2021]
+olympic_years = union(setdiff(baseline, cancelled), off_year)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">29-element Vector{Int64}:
+ 1896
+ 1900
+ 1904
+ 1908
+ 1912
+ 1920
+ 1924
+ 1928
+ 1932
+ 1936
+    ⋮
+ 1988
+ 1992
+ 1996
+ 2000
+ 2004
+ 2008
+ 2012
+ 2016
+ 2021</code></pre><p>You can also find the number of elements (that is, the cardinality) in a set using <code>length</code>:</p><pre><code class="language-julia hljs">length(olympic_years)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">29</code></pre><h2 id="Set-membership-operations"><a class="docs-heading-anchor" href="#Set-membership-operations">Set membership operations</a><a id="Set-membership-operations-1"></a><a class="docs-heading-anchor-permalink" href="#Set-membership-operations" title="Permalink"></a></h2><p>To compute membership of sets, use the <code>in</code> function.</p><pre><code class="language-julia hljs">2000 in olympic_years</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">true</code></pre><pre><code class="language-julia hljs">2001 in olympic_years</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">false</code></pre><h2 id="Indexing-expressions"><a class="docs-heading-anchor" href="#Indexing-expressions">Indexing expressions</a><a id="Indexing-expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Indexing-expressions" title="Permalink"></a></h2><p>Use Julia&#39;s generator syntax to compute new sets, such as the list of Olympic years that are divisible by 3:</p><pre><code class="language-julia hljs">olympic_3_years = [year for year in olympic_years if mod(year, 3) == 0]
+model = Model()
+@variable(model, x[olympic_3_years])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, [1896, 1908, 1920, 1932, 1956, 1968, 1980, 1992, 2004, 2016]
+And data, a 10-element Vector{VariableRef}:
+ x[1896]
+ x[1908]
+ x[1920]
+ x[1932]
+ x[1956]
+ x[1968]
+ x[1980]
+ x[1992]
+ x[2004]
+ x[2016]</code></pre><p>Alternatively, use JuMP&#39;s syntax for constructing <a href="../../../manual/containers/#Containers">Containers</a>:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[year in olympic_years; mod(year, 3) == 0])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">SparseAxisArray{VariableRef, 1, Tuple{Int64}} with 10 entries:
+  [1896]  =  x[1896]
+  [1908]  =  x[1908]
+  [1920]  =  x[1920]
+  [1932]  =  x[1932]
+  [1956]  =  x[1956]
+  [1968]  =  x[1968]
+  [1980]  =  x[1980]
+  [1992]  =  x[1992]
+          ⋮
+  [2004]  =  x[2004]
+  [2016]  =  x[2016]</code></pre><h2 id="Compound-sets"><a class="docs-heading-anchor" href="#Compound-sets">Compound sets</a><a id="Compound-sets-1"></a><a class="docs-heading-anchor-permalink" href="#Compound-sets" title="Permalink"></a></h2><p>Consider a transportation problem in which we need to ship goods between cities. We have been provided a list of cities:</p><pre><code class="language-julia hljs">cities = [&quot;Auckland&quot;, &quot;Wellington&quot;, &quot;Christchurch&quot;, &quot;Dunedin&quot;]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{String}:
+ &quot;Auckland&quot;
+ &quot;Wellington&quot;
+ &quot;Christchurch&quot;
+ &quot;Dunedin&quot;</code></pre><p>and a distance matrix which records the shipping distance between pairs of cities. If we can&#39;t ship between two cities, the distance is <code>0</code>.</p><pre><code class="language-julia hljs">distances = [0 643 1071 1426; 0 0 436 790; 0 0 0 360; 1426 0 0 0]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4×4 Matrix{Int64}:
+    0  643  1071  1426
+    0    0   436   790
+    0    0     0   360
+ 1426    0     0     0</code></pre><p>Let&#39;s have a look at ways we could write a model with an objective function to minimize the total shipping cost. For simplicity, we&#39;ll ignore all constraints.</p><h3 id="Fix-unused-variables"><a class="docs-heading-anchor" href="#Fix-unused-variables">Fix unused variables</a><a id="Fix-unused-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Fix-unused-variables" title="Permalink"></a></h3><p>One approach is to fix all variables that we can&#39;t use to zero. Most solvers are smart-enough to remove these during a presolve phase, so it has a very small impact on performance:</p><pre><code class="language-julia hljs">N = length(cities)
+model = Model()
+@variable(model, x[1:N, 1:N] &gt;= 0)
+for i in 1:N, j in 1:N
+    if distances[i, j] == 0
+        fix(x[i, j], 0.0; force = true)
+    end
+end
+@objective(model, Min, sum(distances[i, j] * x[i, j] for i in 1:N, j in 1:N))</code></pre><p>$ 643 x<em>{1,2} + 1071 x</em>{1,3} + 1426 x<em>{1,4} + 436 x</em>{2,3} + 790 x<em>{2,4} + 360 x</em>{3,4} + 1426 x_{4,1} $</p><h3 id="Filtered-summation"><a class="docs-heading-anchor" href="#Filtered-summation">Filtered summation</a><a id="Filtered-summation-1"></a><a class="docs-heading-anchor-permalink" href="#Filtered-summation" title="Permalink"></a></h3><p>Another approach is to define filters whenever we want to sum over our decision variables:</p><pre><code class="language-julia hljs">N = length(cities)
+model = Model()
+@variable(model, x[1:N, 1:N] &gt;= 0)
+@objective(
+    model,
+    Min,
+    sum(
+        distances[i, j] * x[i, j] for i in 1:N, j in 1:N if distances[i, j] &gt; 0
+    ),
+)</code></pre><p>$ 643 x<em>{1,2} + 1071 x</em>{1,3} + 1426 x<em>{1,4} + 436 x</em>{2,3} + 790 x<em>{2,4} + 360 x</em>{3,4} + 1426 x_{4,1} $</p><h3 id="Filtered-indexing"><a class="docs-heading-anchor" href="#Filtered-indexing">Filtered indexing</a><a id="Filtered-indexing-1"></a><a class="docs-heading-anchor-permalink" href="#Filtered-indexing" title="Permalink"></a></h3><p>We could also use JuMP&#39;s support for <a href="../../../manual/containers/#Containers">Containers</a>:</p><pre><code class="language-julia hljs">N = length(cities)
+model = Model()
+@variable(model, x[i = 1:N, j = 1:N; distances[i, j] &gt; 0])
+@objective(model, Min, sum(distances[i...] * x[i] for i in eachindex(x)))</code></pre><p>$ 790 x<em>{2,4} + 643 x</em>{1,2} + 1071 x<em>{1,3} + 1426 x</em>{4,1} + 360 x<em>{3,4} + 1426 x</em>{1,4} + 436 x_{2,3} $</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p>The <code>i...</code> is called  a &quot;splat.&quot; It converts a tuple like <code>(1, 2)</code> into two indices like <code>distances[1, 2]</code>.</p></div></div><h3 id="Converting-to-a-different-data-structure"><a class="docs-heading-anchor" href="#Converting-to-a-different-data-structure">Converting to a different data structure</a><a id="Converting-to-a-different-data-structure-1"></a><a class="docs-heading-anchor-permalink" href="#Converting-to-a-different-data-structure" title="Permalink"></a></h3><p>Another approach, and one that is often the most readable, is to convert the data you have into something that is easier to work with. Originally, we had a vector of strings and a matrix of distances. What we really need is something that maps usable origin-destination pairs to distances. A dictionary is an obvious choice:</p><pre><code class="language-julia hljs">routes = Dict(
+    (a, b) =&gt; distances[i, j] for
+    (i, a) in enumerate(cities), (j, b) in enumerate(cities) if
+    distances[i, j] &gt; 0
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{Tuple{String, String}, Int64} with 7 entries:
+  (&quot;Auckland&quot;, &quot;Wellington&quot;)     =&gt; 643
+  (&quot;Wellington&quot;, &quot;Christchurch&quot;) =&gt; 436
+  (&quot;Wellington&quot;, &quot;Dunedin&quot;)      =&gt; 790
+  (&quot;Christchurch&quot;, &quot;Dunedin&quot;)    =&gt; 360
+  (&quot;Auckland&quot;, &quot;Dunedin&quot;)        =&gt; 1426
+  (&quot;Dunedin&quot;, &quot;Auckland&quot;)        =&gt; 1426
+  (&quot;Auckland&quot;, &quot;Christchurch&quot;)   =&gt; 1071</code></pre><p>Then, we can create our model like so:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[keys(routes)])
+@objective(model, Min, sum(v * x[k] for (k, v) in routes))</code></pre><p>$ 643 x<em>{(&quot;Auckland&quot;, &quot;Wellington&quot;)} + 436 x</em>{(&quot;Wellington&quot;, &quot;Christchurch&quot;)} + 790 x<em>{(&quot;Wellington&quot;, &quot;Dunedin&quot;)} + 360 x</em>{(&quot;Christchurch&quot;, &quot;Dunedin&quot;)} + 1426 x<em>{(&quot;Auckland&quot;, &quot;Dunedin&quot;)} + 1426 x</em>{(&quot;Dunedin&quot;, &quot;Auckland&quot;)} + 1071 x_{(&quot;Auckland&quot;, &quot;Christchurch&quot;)} $</p><p>This has a number of benefits over the other approaches, including a compacter algebraic model and variables that are named in a more meaningful way.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>If you&#39;re struggling to formulate a problem using the available syntax in JuMP, it&#39;s probably a sign that you should convert your data into a different form.</p></div></div><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><p>The purpose of this tutorial was to show how JuMP does not have specialized syntax for set creation and manipulation. Instead, you should use the tools provided by Julia itself.</p><p>This is both an opportunity and a challenge, because you are free to pick the syntax and data structures that best suit your problem, but for new users it can be daunting to decide which structure to use.</p><p>Read through some of the other JuMP tutorials to get inspiration and ideas for how you can use Julia&#39;s syntax and data structures to your advantage.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../getting_started_with_JuMP/">« Getting started with JuMP</a><a class="docs-footer-nextpage" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/introduction/index.html b/previews/PR3525/tutorials/getting_started/introduction/index.html
new file mode 100644
index 00000000000..ebd5949e8e4
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/introduction/index.html
@@ -0,0 +1,6 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#How-these-tutorials-are-structured"><span>How these tutorials are structured</span></a></li></ul></li><li><a class="tocitem" href="../getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../debugging/">Debugging</a></li><li><a class="tocitem" href="../design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/introduction.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Introduction"><a class="docs-heading-anchor" href="#Introduction">Introduction</a><a id="Introduction-1"></a><a class="docs-heading-anchor-permalink" href="#Introduction" title="Permalink"></a></h1><p>The purpose of these &quot;Getting started&quot; tutorials is to teach new users the basics of Julia and JuMP.</p><h2 id="How-these-tutorials-are-structured"><a class="docs-heading-anchor" href="#How-these-tutorials-are-structured">How these tutorials are structured</a><a id="How-these-tutorials-are-structured-1"></a><a class="docs-heading-anchor-permalink" href="#How-these-tutorials-are-structured" title="Permalink"></a></h2><p>Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.</p><ul><li>The &quot;Getting started with&quot; tutorials are basic introductions to different aspects of JuMP and Julia. If you are new to JuMP and Julia, start by reading them in the following order:<ul><li><a href="../getting_started_with_julia/#Getting-started-with-Julia">Getting started with Julia</a></li><li><a href="../getting_started_with_JuMP/#Getting-started-with-JuMP">Getting started with JuMP</a></li><li><a href="../getting_started_with_sets_and_indexing/#Getting-started-with-sets-and-indexing">Getting started with sets and indexing</a></li><li><a href="../getting_started_with_data_and_plotting/#Getting-started-with-data-and-plotting">Getting started with data and plotting</a></li></ul></li><li>Julia has a reputation for being &quot;fast.&quot; Unfortunately, it is also easy to write <em>slow</em> Julia code. <a href="../performance_tips/#Performance-tips">Performance tips</a> contains a number of important tips on how to improve the performance of models you write in JuMP.</li><li><a href="../design_patterns_for_larger_models/#Design-patterns-for-larger-models">Design patterns for larger models</a> is a more advanced tutorial that is aimed at users writing large JuMP models. It&#39;s in the &quot;Getting started&quot; section to give you an early preview of how JuMP makes it easy to structure larger models. If you are new to JuMP you may want to skip or briefly skim this tutorial, and come back to it once you have written a few JuMP models.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../../installation/">« Installation Guide</a><a class="docs-footer-nextpage" href="../getting_started_with_julia/">Getting started with Julia »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/getting_started/performance_tips.jl b/previews/PR3525/tutorials/getting_started/performance_tips.jl
new file mode 100644
index 00000000000..71b97f4050d
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/performance_tips.jl
@@ -0,0 +1,193 @@
+# Copyright (c) 2021 Oscar Dowson and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Performance tips
+
+# By now you should have read the other "getting started" tutorials. You're
+# almost ready to write your own models, but before you do so there are some
+# important things to be aware of.
+
+using JuMP
+import HiGHS
+
+# ## Read the Julia performance tips
+
+# The first thing to do is read the [Performance tips](https://docs.julialang.org/en/v1/manual/performance-tips/index.html)
+# section of the Julia manual. The most important rule is to avoid global
+# variables. This is particularly important if you're learning JuMP after using
+# a language like MATLAB.
+
+# ## The "time-to-first-solve" issue
+
+# Similar to the infamous [time-to-first-plot](https://discourse.julialang.org/t/roadmap-for-a-faster-time-to-first-plot/22956)
+# plotting problem, JuMP suffers from time-to-first-solve latency. This latency
+# occurs because the first time you call JuMP code in each session, Julia needs
+# to compile a lot of code specific to your problem. This issue is actively being
+# worked on, but there are a few things you can do to improve things.
+
+# ### Suggestion 1: don't call JuMP from the command line
+
+# In other languages, you might be used to a workflow like:
+# ```
+# $ julia my_script.jl
+# ```
+# This doesn't work for JuMP, because we have to pay the compilation latency
+# every time you run the script. Instead, use one of the [suggested workflows](https://docs.julialang.org/en/v1/manual/workflow-tips/)
+# from the Julia documentation.
+
+# ### Suggestion 2: disable bridges if none are being used
+
+# At present, the majority of the latency problems are caused by JuMP's bridging
+# mechanism. If you only use constraints that are natively supported by the
+# solver, you can disable bridges by passing `add_bridges = false` to
+# [`Model`](@ref).
+
+model = Model(HiGHS.Optimizer; add_bridges = false)
+
+# ### Suggestion 3: use PackageCompiler
+
+# As an example of compilation latency, consider the following linear program
+# with two variables and two constraints:
+
+# ```julia
+# using JuMP, HiGHS
+# model = Model(HiGHS.Optimizer)
+# set_silent(model)
+# @variable(model, x >= 0)
+# @variable(model, 0 <= y <= 3)
+# @objective(model, Min, 12x + 20y)
+# @constraint(model, c1, 6x + 8y >= 100)
+# @constraint(model, c2, 7x + 12y >= 120)
+# optimize!(model)
+# open("model.log", "w") do io
+#     print(io, solution_summary(model; verbose = true))
+#     return
+# end
+# ```
+
+# Saving the problem in `model.jl` and calling from the command line results in:
+# ```
+# $ time julia model.jl
+# 15.78s user 0.48s system 100% cpu 16.173 total
+# ```
+# Clearly, 16 seconds is a large overhead to pay for solving this trivial model.
+# However, the compilation latency is independent on the problem size, and so 16
+# seconds of additional overhead may be tolerable for larger models that take
+# minutes or hours to solve.
+
+# In cases where the compilation latency is intolerable, JuMP is compatible with
+# the [PackageCompiler.jl](https://julialang.github.io/PackageCompiler.jl/dev/)
+# package, which makes it easy to generate a custom _sysimage_ (a binary
+# extension to Julia that caches compiled code) that dramatically reduces the
+# compilation latency. A custom image for our problem can be created as follows:
+# ```julia
+# using PackageCompiler, Libdl
+# PackageCompiler.create_sysimage(
+#     ["JuMP", "HiGHS"],
+#     sysimage_path = "customimage." * Libdl.dlext,
+#     precompile_execution_file = "model.jl",
+# )
+# ```
+# When Julia is run with the custom image, the run time is now 0.7 seconds
+# instead of 16:
+# ```
+# $ time julia --sysimage customimage model.jl
+# 0.68s user 0.22s system 153% cpu 0.587 total
+# ```
+# Other performance tweaks, such as disabling bridges or using direct mode can
+# reduce this time further.
+
+# !!! note
+#     `create_sysimage` only needs to be run once, and the same sysimage can be
+#     used--to a slight detriment of performance--even if we modify
+#     `model.jl` or run a different file.
+
+# ## Use macros to build expressions
+
+# Use JuMP's macros (or [`add_to_expression!`](@ref)) to build expressions.
+# Avoid constructing expressions outside the macros.
+
+# Constructing an expression outside the macro results in intermediate copies of
+# the expression. For example,
+# ```julia
+# x[1] + x[2] + x[3]
+# ```
+# is equivalent to
+# ```julia
+# a = x[1]
+# b = a + x[2]
+# c = b + x[3]
+# ```
+# Since we only care about `c`, the `a` and `b` expressions are not needed and
+# constructing them slows the program down.
+
+# JuMP's macros rewrite the expressions to operate in-place and avoid these
+# extra copies. Because they allocate less memory, they are faster, particularly
+# for large expressions.
+
+# Here's an example.
+
+model = Model()
+@variable(model, x[1:3])
+
+# Here's what happens if we construct the expression outside the macro:
+
+@allocated x[1] + x[2] + x[3]
+
+# !!! info
+#     The `@allocated` measures how many bytes were allocated during the
+#     evaluation of an expression. Fewer is better.
+
+# If we use the [`@expression`](@ref) macro, we get many fewer allocations:
+
+@allocated @expression(model, x[1] + x[2] + x[3])
+
+# ## Disable string names
+
+# By default, JuMP creates `String` names for variables and constraints and
+# passes these to the solver. The benefit of passing names is that it improves
+# the readability of log messages from the solver (for example, "variable x has
+# invalid bounds" instead of "variable v1203 has invalid bounds"), but for
+# larger models the overhead of passing names can be non-trivial.
+
+# Disable the creation of `String` names by setting `set_string_name = false` in
+# the [`@variable`](@ref) and [`@constraint`](@ref) macros, or by calling
+# [`set_string_names_on_creation`](@ref) to disable all names for a particular
+# model:
+
+model = Model();
+set_string_names_on_creation(model, false)
+@variable(model, x)
+@constraint(model, c, 2x <= 1)
+
+# Note that this doesn't change how symbolic names and bindings are stored:
+
+x
+model[:x]
+x === model[:x]
+
+# But you can no longer look up the variable by the string name:
+
+variable_by_name(model, "x") === nothing
+
+# !!! info
+#     For more information on the difference between string names, symbolic
+#     names, and bindings, see
+#     [String names, symbolic names, and bindings](@ref variable_names_and_bindings).
diff --git a/previews/PR3525/tutorials/getting_started/performance_tips/index.html b/previews/PR3525/tutorials/getting_started/performance_tips/index.html
new file mode 100644
index 00000000000..9dce4fd6022
--- /dev/null
+++ b/previews/PR3525/tutorials/getting_started/performance_tips/index.html
@@ -0,0 +1,40 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Performance tips · JuMP</title><meta name="title" content="Performance tips · JuMP"/><meta property="og:title" content="Performance tips · JuMP"/><meta property="twitter:title" content="Performance tips · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox" checked/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../debugging/">Debugging</a></li><li><a class="tocitem" href="../design_patterns_for_larger_models/">Design patterns for larger models</a></li><li class="is-active"><a class="tocitem" href>Performance tips</a><ul class="internal"><li><a class="tocitem" href="#Read-the-Julia-performance-tips"><span>Read the Julia performance tips</span></a></li><li><a class="tocitem" href="#The-&quot;time-to-first-solve&quot;-issue"><span>The &quot;time-to-first-solve&quot; issue</span></a></li><li><a class="tocitem" href="#Use-macros-to-build-expressions"><span>Use macros to build expressions</span></a></li><li><a class="tocitem" href="#Disable-string-names"><span>Disable string names</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Getting started</a></li><li class="is-active"><a href>Performance tips</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Performance tips</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/getting_started/performance_tips.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Performance-tips"><a class="docs-heading-anchor" href="#Performance-tips">Performance tips</a><a id="Performance-tips-1"></a><a class="docs-heading-anchor-permalink" href="#Performance-tips" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../performance_tips.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>By now you should have read the other &quot;getting started&quot; tutorials. You&#39;re almost ready to write your own models, but before you do so there are some important things to be aware of.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; import HiGHS</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><h2 id="Read-the-Julia-performance-tips"><a class="docs-heading-anchor" href="#Read-the-Julia-performance-tips">Read the Julia performance tips</a><a id="Read-the-Julia-performance-tips-1"></a><a class="docs-heading-anchor-permalink" href="#Read-the-Julia-performance-tips" title="Permalink"></a></h2><p>The first thing to do is read the <a href="https://docs.julialang.org/en/v1/manual/performance-tips/index.html">Performance tips</a> section of the Julia manual. The most important rule is to avoid global variables. This is particularly important if you&#39;re learning JuMP after using a language like MATLAB.</p><h2 id="The-&quot;time-to-first-solve&quot;-issue"><a class="docs-heading-anchor" href="#The-&quot;time-to-first-solve&quot;-issue">The &quot;time-to-first-solve&quot; issue</a><a id="The-&quot;time-to-first-solve&quot;-issue-1"></a><a class="docs-heading-anchor-permalink" href="#The-&quot;time-to-first-solve&quot;-issue" title="Permalink"></a></h2><p>Similar to the infamous <a href="https://discourse.julialang.org/t/roadmap-for-a-faster-time-to-first-plot/22956">time-to-first-plot</a> plotting problem, JuMP suffers from time-to-first-solve latency. This latency occurs because the first time you call JuMP code in each session, Julia needs to compile a lot of code specific to your problem. This issue is actively being worked on, but there are a few things you can do to improve things.</p><h3 id="Suggestion-1:-don&#39;t-call-JuMP-from-the-command-line"><a class="docs-heading-anchor" href="#Suggestion-1:-don&#39;t-call-JuMP-from-the-command-line">Suggestion 1: don&#39;t call JuMP from the command line</a><a id="Suggestion-1:-don&#39;t-call-JuMP-from-the-command-line-1"></a><a class="docs-heading-anchor-permalink" href="#Suggestion-1:-don&#39;t-call-JuMP-from-the-command-line" title="Permalink"></a></h3><p>In other languages, you might be used to a workflow like:</p><pre><code class="nohighlight hljs">$ julia my_script.jl</code></pre><p>This doesn&#39;t work for JuMP, because we have to pay the compilation latency every time you run the script. Instead, use one of the <a href="https://docs.julialang.org/en/v1/manual/workflow-tips/">suggested workflows</a> from the Julia documentation.</p><h3 id="Suggestion-2:-disable-bridges-if-none-are-being-used"><a class="docs-heading-anchor" href="#Suggestion-2:-disable-bridges-if-none-are-being-used">Suggestion 2: disable bridges if none are being used</a><a id="Suggestion-2:-disable-bridges-if-none-are-being-used-1"></a><a class="docs-heading-anchor-permalink" href="#Suggestion-2:-disable-bridges-if-none-are-being-used" title="Permalink"></a></h3><p>At present, the majority of the latency problems are caused by JuMP&#39;s bridging mechanism. If you only use constraints that are natively supported by the solver, you can disable bridges by passing <code>add_bridges = false</code> to <a href="../../../api/JuMP/#JuMP.Model"><code>Model</code></a>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model(HiGHS.Optimizer; add_bridges = false)</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><h3 id="Suggestion-3:-use-PackageCompiler"><a class="docs-heading-anchor" href="#Suggestion-3:-use-PackageCompiler">Suggestion 3: use PackageCompiler</a><a id="Suggestion-3:-use-PackageCompiler-1"></a><a class="docs-heading-anchor-permalink" href="#Suggestion-3:-use-PackageCompiler" title="Permalink"></a></h3><p>As an example of compilation latency, consider the following linear program with two variables and two constraints:</p><pre><code class="language-julia hljs">using JuMP, HiGHS
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x &gt;= 0)
+@variable(model, 0 &lt;= y &lt;= 3)
+@objective(model, Min, 12x + 20y)
+@constraint(model, c1, 6x + 8y &gt;= 100)
+@constraint(model, c2, 7x + 12y &gt;= 120)
+optimize!(model)
+open(&quot;model.log&quot;, &quot;w&quot;) do io
+    print(io, solution_summary(model; verbose = true))
+    return
+end</code></pre><p>Saving the problem in <code>model.jl</code> and calling from the command line results in:</p><pre><code class="nohighlight hljs">$ time julia model.jl
+15.78s user 0.48s system 100% cpu 16.173 total</code></pre><p>Clearly, 16 seconds is a large overhead to pay for solving this trivial model. However, the compilation latency is independent on the problem size, and so 16 seconds of additional overhead may be tolerable for larger models that take minutes or hours to solve.</p><p>In cases where the compilation latency is intolerable, JuMP is compatible with the <a href="https://julialang.github.io/PackageCompiler.jl/dev/">PackageCompiler.jl</a> package, which makes it easy to generate a custom <em>sysimage</em> (a binary extension to Julia that caches compiled code) that dramatically reduces the compilation latency. A custom image for our problem can be created as follows:</p><pre><code class="language-julia hljs">using PackageCompiler, Libdl
+PackageCompiler.create_sysimage(
+    [&quot;JuMP&quot;, &quot;HiGHS&quot;],
+    sysimage_path = &quot;customimage.&quot; * Libdl.dlext,
+    precompile_execution_file = &quot;model.jl&quot;,
+)</code></pre><p>When Julia is run with the custom image, the run time is now 0.7 seconds instead of 16:</p><pre><code class="nohighlight hljs">$ time julia --sysimage customimage model.jl
+0.68s user 0.22s system 153% cpu 0.587 total</code></pre><p>Other performance tweaks, such as disabling bridges or using direct mode can reduce this time further.</p><div class="admonition is-info"><header class="admonition-header">Note</header><div class="admonition-body"><p><code>create_sysimage</code> only needs to be run once, and the same sysimage can be used–to a slight detriment of performance–even if we modify <code>model.jl</code> or run a different file.</p></div></div><h2 id="Use-macros-to-build-expressions"><a class="docs-heading-anchor" href="#Use-macros-to-build-expressions">Use macros to build expressions</a><a id="Use-macros-to-build-expressions-1"></a><a class="docs-heading-anchor-permalink" href="#Use-macros-to-build-expressions" title="Permalink"></a></h2><p>Use JuMP&#39;s macros (or <a href="../../../api/JuMP/#JuMP.add_to_expression!"><code>add_to_expression!</code></a>) to build expressions. Avoid constructing expressions outside the macros.</p><p>Constructing an expression outside the macro results in intermediate copies of the expression. For example,</p><pre><code class="language-julia hljs">x[1] + x[2] + x[3]</code></pre><p>is equivalent to</p><pre><code class="language-julia hljs">a = x[1]
+b = a + x[2]
+c = b + x[3]</code></pre><p>Since we only care about <code>c</code>, the <code>a</code> and <code>b</code> expressions are not needed and constructing them slows the program down.</p><p>JuMP&#39;s macros rewrite the expressions to operate in-place and avoid these extra copies. Because they allocate less memory, they are faster, particularly for large expressions.</p><p>Here&#39;s an example.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model()</code><code class="nohighlight hljs ansi" style="display:block;">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: NO_OPTIMIZER
+Solver name: No optimizer attached.</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:3])</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code></pre><p>Here&#39;s what happens if we construct the expression outside the macro:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @allocated x[1] + x[2] + x[3]</code><code class="nohighlight hljs ansi" style="display:block;">1424</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The <code>@allocated</code> measures how many bytes were allocated during the evaluation of an expression. Fewer is better.</p></div></div><p>If we use the <a href="../../../api/JuMP/#JuMP.@expression"><code>@expression</code></a> macro, we get many fewer allocations:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @allocated @expression(model, x[1] + x[2] + x[3])</code><code class="nohighlight hljs ansi" style="display:block;">800</code></pre><h2 id="Disable-string-names"><a class="docs-heading-anchor" href="#Disable-string-names">Disable string names</a><a id="Disable-string-names-1"></a><a class="docs-heading-anchor-permalink" href="#Disable-string-names" title="Permalink"></a></h2><p>By default, JuMP creates <code>String</code> names for variables and constraints and passes these to the solver. The benefit of passing names is that it improves the readability of log messages from the solver (for example, &quot;variable x has invalid bounds&quot; instead of &quot;variable v1203 has invalid bounds&quot;), but for larger models the overhead of passing names can be non-trivial.</p><p>Disable the creation of <code>String</code> names by setting <code>set_string_name = false</code> in the <a href="../../../api/JuMP/#JuMP.@variable"><code>@variable</code></a> and <a href="../../../api/JuMP/#JuMP.@constraint"><code>@constraint</code></a> macros, or by calling <a href="../../../api/JuMP/#JuMP.set_string_names_on_creation"><code>set_string_names_on_creation</code></a> to disable all names for a particular model:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; set_string_names_on_creation(model, false)</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">_[1]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, c, 2x &lt;= 1)</code><code class="nohighlight hljs ansi" style="display:block;">2 _[1] ≤ 1</code></pre><p>Note that this doesn&#39;t change how symbolic names and bindings are stored:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; x</code><code class="nohighlight hljs ansi" style="display:block;">_[1]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model[:x]</code><code class="nohighlight hljs ansi" style="display:block;">_[1]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; x === model[:x]</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><p>But you can no longer look up the variable by the string name:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; variable_by_name(model, &quot;x&quot;) === nothing</code><code class="nohighlight hljs ansi" style="display:block;">true</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>For more information on the difference between string names, symbolic names, and bindings, see <a href="../../../manual/variables/#variable_names_and_bindings">String names, symbolic names, and bindings</a>.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../design_patterns_for_larger_models/">« Design patterns for larger models</a><a class="docs-footer-nextpage" href="../../linear/introduction/">Introduction »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/callbacks.jl b/previews/PR3525/tutorials/linear/callbacks.jl
new file mode 100644
index 00000000000..f4517c8f374
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/callbacks.jl
@@ -0,0 +1,185 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # [Callbacks](@id callbacks_tutorial)
+
+# The purpose of the tutorial is to demonstrate the various solver-independent
+# and solver-dependent callbacks that are supported by JuMP.
+
+# The tutorial uses the following packages:
+
+using JuMP
+import GLPK
+import Random
+import Test  #src
+
+# !!! info
+#     This tutorial uses the [MathOptInterface](@ref moi_documentation) API.
+#     By default, JuMP exports the `MOI` symbol as an alias for the
+#     MathOptInterface.jl package. We recommend making this more explicit in
+#     your code by adding the following lines:
+#     ```julia
+#     import MathOptInterface as MOI
+#     ```
+
+# ## Lazy constraints
+
+# An example using a lazy constraint callback.
+
+function example_lazy_constraint()
+    model = Model(GLPK.Optimizer)
+    @variable(model, 0 <= x <= 2.5, Int)
+    @variable(model, 0 <= y <= 2.5, Int)
+    @objective(model, Max, y)
+    lazy_called = false
+    function my_callback_function(cb_data)
+        lazy_called = true
+        x_val = callback_value(cb_data, x)
+        y_val = callback_value(cb_data, y)
+        println("Called from (x, y) = ($x_val, $y_val)")
+        status = callback_node_status(cb_data, model)
+        if status == MOI.CALLBACK_NODE_STATUS_FRACTIONAL
+            println(" - Solution is integer infeasible!")
+        elseif status == MOI.CALLBACK_NODE_STATUS_INTEGER
+            println(" - Solution is integer feasible!")
+        else
+            @assert status == MOI.CALLBACK_NODE_STATUS_UNKNOWN
+            println(" - I don't know if the solution is integer feasible :(")
+        end
+        if y_val - x_val > 1 + 1e-6
+            con = @build_constraint(y - x <= 1)
+            println("Adding $(con)")
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        elseif y_val + x_val > 3 + 1e-6
+            con = @build_constraint(y + x <= 3)
+            println("Adding $(con)")
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        end
+    end
+    set_attribute(model, MOI.LazyConstraintCallback(), my_callback_function)
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL    #src
+    Test.@test primal_status(model) == FEASIBLE_POINT  #src
+    Test.@test lazy_called    #src
+    Test.@test value(x) == 1  #src
+    Test.@test value(y) == 2  #src
+    println("Optimal solution (x, y) = ($(value(x)), $(value(y)))")
+    return
+end
+
+example_lazy_constraint()
+
+# ## User-cuts
+
+# An example using a user-cut callback.
+
+function example_user_cut_constraint()
+    Random.seed!(1)
+    N = 30
+    item_weights, item_values = rand(N), rand(N)
+    model = Model(GLPK.Optimizer)
+    @variable(model, x[1:N], Bin)
+    @constraint(model, sum(item_weights[i] * x[i] for i in 1:N) <= 10)
+    @objective(model, Max, sum(item_values[i] * x[i] for i in 1:N))
+    callback_called = false
+    function my_callback_function(cb_data)
+        callback_called = true
+        x_vals = callback_value.(Ref(cb_data), x)
+        accumulated = sum(item_weights[i] for i in 1:N if x_vals[i] > 1e-4)
+        println("Called with accumulated = $(accumulated)")
+        n_terms = sum(1 for i in 1:N if x_vals[i] > 1e-4)
+        if accumulated > 10
+            con = @build_constraint(
+                sum(x[i] for i in 1:N if x_vals[i] > 0.5) <= n_terms - 1
+            )
+            println("Adding $(con)")
+            MOI.submit(model, MOI.UserCut(cb_data), con)
+        end
+    end
+    set_attribute(model, MOI.UserCutCallback(), my_callback_function)
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL  #src
+    Test.@test primal_status(model) == FEASIBLE_POINT  #src
+    Test.@test callback_called  #src
+    @show callback_called
+    return
+end
+
+example_user_cut_constraint()
+
+# ## Heuristic solutions
+
+# An example using a heuristic solution callback.
+
+function example_heuristic_solution()
+    Random.seed!(1)
+    N = 30
+    item_weights, item_values = rand(N), rand(N)
+    model = Model(GLPK.Optimizer)
+    @variable(model, x[1:N], Bin)
+    @constraint(model, sum(item_weights[i] * x[i] for i in 1:N) <= 10)
+    @objective(model, Max, sum(item_values[i] * x[i] for i in 1:N))
+    callback_called = false
+    function my_callback_function(cb_data)
+        callback_called = true
+        x_vals = callback_value.(Ref(cb_data), x)
+        ret =
+            MOI.submit(model, MOI.HeuristicSolution(cb_data), x, floor.(x_vals))
+        println("Heuristic solution status = $(ret)")
+        Test.@test ret in (                     #src
+            MOI.HEURISTIC_SOLUTION_ACCEPTED,    #src
+            MOI.HEURISTIC_SOLUTION_REJECTED,    #src
+        )                                       #src
+    end
+    set_attribute(model, MOI.HeuristicCallback(), my_callback_function)
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL  #src
+    Test.@test primal_status(model) == FEASIBLE_POINT  #src
+    Test.@test callback_called  #src
+    return
+end
+
+example_heuristic_solution()
+
+# ## GLPK solver-dependent callback
+
+# An example using GLPK's solver-dependent callback.
+
+function example_solver_dependent_callback()
+    model = Model(GLPK.Optimizer)
+    @variable(model, 0 <= x <= 2.5, Int)
+    @variable(model, 0 <= y <= 2.5, Int)
+    @objective(model, Max, y)
+    lazy_called = false
+    function my_callback_function(cb_data)
+        lazy_called = true
+        reason = GLPK.glp_ios_reason(cb_data.tree)
+        println("Called from reason = $(reason)")
+        if reason != GLPK.GLP_IROWGEN
+            return
+        end
+        x_val = callback_value(cb_data, x)
+        y_val = callback_value(cb_data, y)
+        if y_val - x_val > 1 + 1e-6
+            con = @build_constraint(y - x <= 1)
+            println("Adding $(con)")
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        elseif y_val + x_val > 3 + 1e-6
+            con = @build_constraint(y - x <= 1)
+            println("Adding $(con)")
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        end
+    end
+    set_attribute(model, GLPK.CallbackFunction(), my_callback_function)
+    optimize!(model)
+    Test.@test termination_status(model) == OPTIMAL  #src
+    Test.@test primal_status(model) == FEASIBLE_POINT  #src
+    Test.@test lazy_called  #src
+    Test.@test value(x) == 1  #src
+    Test.@test value(y) == 2  #src
+    return
+end
+
+example_solver_dependent_callback()
diff --git a/previews/PR3525/tutorials/linear/callbacks/index.html b/previews/PR3525/tutorials/linear/callbacks/index.html
new file mode 100644
index 00000000000..c6c73d8245a
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/callbacks/index.html
@@ -0,0 +1,152 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Callbacks · JuMP</title><meta name="title" content="Callbacks · JuMP"/><meta property="og:title" content="Callbacks · JuMP"/><meta property="twitter:title" content="Callbacks · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li class="is-active"><a class="tocitem" href>Callbacks</a><ul class="internal"><li><a class="tocitem" href="#Lazy-constraints"><span>Lazy constraints</span></a></li><li><a class="tocitem" href="#User-cuts"><span>User-cuts</span></a></li><li><a class="tocitem" href="#Heuristic-solutions"><span>Heuristic solutions</span></a></li><li><a class="tocitem" href="#GLPK-solver-dependent-callback"><span>GLPK solver-dependent callback</span></a></li></ul></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Callbacks</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Callbacks</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/callbacks.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="callbacks_tutorial"><a class="docs-heading-anchor" href="#callbacks_tutorial">Callbacks</a><a id="callbacks_tutorial-1"></a><a class="docs-heading-anchor-permalink" href="#callbacks_tutorial" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../callbacks.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of the tutorial is to demonstrate the various solver-independent and solver-dependent callbacks that are supported by JuMP.</p><p>The tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import GLPK
+import Random</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This tutorial uses the <a href="../../../moi/#moi_documentation">MathOptInterface</a> API. By default, JuMP exports the <code>MOI</code> symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre></div></div><h2 id="Lazy-constraints"><a class="docs-heading-anchor" href="#Lazy-constraints">Lazy constraints</a><a id="Lazy-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Lazy-constraints" title="Permalink"></a></h2><p>An example using a lazy constraint callback.</p><pre><code class="language-julia hljs">function example_lazy_constraint()
+    model = Model(GLPK.Optimizer)
+    @variable(model, 0 &lt;= x &lt;= 2.5, Int)
+    @variable(model, 0 &lt;= y &lt;= 2.5, Int)
+    @objective(model, Max, y)
+    lazy_called = false
+    function my_callback_function(cb_data)
+        lazy_called = true
+        x_val = callback_value(cb_data, x)
+        y_val = callback_value(cb_data, y)
+        println(&quot;Called from (x, y) = ($x_val, $y_val)&quot;)
+        status = callback_node_status(cb_data, model)
+        if status == MOI.CALLBACK_NODE_STATUS_FRACTIONAL
+            println(&quot; - Solution is integer infeasible!&quot;)
+        elseif status == MOI.CALLBACK_NODE_STATUS_INTEGER
+            println(&quot; - Solution is integer feasible!&quot;)
+        else
+            @assert status == MOI.CALLBACK_NODE_STATUS_UNKNOWN
+            println(&quot; - I don&#39;t know if the solution is integer feasible :(&quot;)
+        end
+        if y_val - x_val &gt; 1 + 1e-6
+            con = @build_constraint(y - x &lt;= 1)
+            println(&quot;Adding $(con)&quot;)
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        elseif y_val + x_val &gt; 3 + 1e-6
+            con = @build_constraint(y + x &lt;= 3)
+            println(&quot;Adding $(con)&quot;)
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        end
+    end
+    set_attribute(model, MOI.LazyConstraintCallback(), my_callback_function)
+    optimize!(model)
+    println(&quot;Optimal solution (x, y) = ($(value(x)), $(value(y)))&quot;)
+    return
+end
+
+example_lazy_constraint()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Called from (x, y) = (0.0, 2.0)
+ - Solution is integer feasible!
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(y - x, MathOptInterface.LessThan{Float64}(1.0))
+Called from (x, y) = (1.0, 2.0)
+ - Solution is integer feasible!
+Optimal solution (x, y) = (1.0, 2.0)</code></pre><h2 id="User-cuts"><a class="docs-heading-anchor" href="#User-cuts">User-cuts</a><a id="User-cuts-1"></a><a class="docs-heading-anchor-permalink" href="#User-cuts" title="Permalink"></a></h2><p>An example using a user-cut callback.</p><pre><code class="language-julia hljs">function example_user_cut_constraint()
+    Random.seed!(1)
+    N = 30
+    item_weights, item_values = rand(N), rand(N)
+    model = Model(GLPK.Optimizer)
+    @variable(model, x[1:N], Bin)
+    @constraint(model, sum(item_weights[i] * x[i] for i in 1:N) &lt;= 10)
+    @objective(model, Max, sum(item_values[i] * x[i] for i in 1:N))
+    callback_called = false
+    function my_callback_function(cb_data)
+        callback_called = true
+        x_vals = callback_value.(Ref(cb_data), x)
+        accumulated = sum(item_weights[i] for i in 1:N if x_vals[i] &gt; 1e-4)
+        println(&quot;Called with accumulated = $(accumulated)&quot;)
+        n_terms = sum(1 for i in 1:N if x_vals[i] &gt; 1e-4)
+        if accumulated &gt; 10
+            con = @build_constraint(
+                sum(x[i] for i in 1:N if x_vals[i] &gt; 0.5) &lt;= n_terms - 1
+            )
+            println(&quot;Adding $(con)&quot;)
+            MOI.submit(model, MOI.UserCut(cb_data), con)
+        end
+    end
+    set_attribute(model, MOI.UserCutCallback(), my_callback_function)
+    optimize!(model)
+    @show callback_called
+    return
+end
+
+example_user_cut_constraint()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Called with accumulated = 10.37975831721494
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(x[1] + x[2] + x[3] + x[4] + x[5] + x[7] + x[8] + x[9] + x[10] + x[11] + x[12] + x[13] + x[14] + x[16] + x[17] + x[18] + x[20] + x[22] + x[23] + x[25] + x[26] + x[28] + x[29] + x[30], MathOptInterface.LessThan{Float64}(23.0))
+Called with accumulated = 10.37975831721494
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(x[1] + x[2] + x[3] + x[4] + x[5] + x[7] + x[8] + x[9] + x[10] + x[12] + x[13] + x[14] + x[16] + x[17] + x[18] + x[20] + x[22] + x[23] + x[25] + x[26] + x[28] + x[29] + x[30], MathOptInterface.LessThan{Float64}(23.0))
+Called with accumulated = 10.585271197221452
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(x[1] + x[2] + x[3] + x[4] + x[5] + x[7] + x[8] + x[9] + x[10] + x[11] + x[12] + x[13] + x[14] + x[16] + x[17] + x[18] + x[20] + x[22] + x[23] + x[25] + x[26] + x[29] + x[30], MathOptInterface.LessThan{Float64}(23.0))
+Called with accumulated = 10.574173763699463
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(x[1] + x[2] + x[3] + x[4] + x[5] + x[7] + x[8] + x[9] + x[10] + x[11] + x[12] + x[13] + x[14] + x[16] + x[17] + x[18] + x[20] + x[22] + x[23] + x[25] + x[26] + x[29] + x[30], MathOptInterface.LessThan{Float64}(23.0))
+Called with accumulated = 10.193650358656257
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(x[1] + x[2] + x[3] + x[4] + x[5] + x[7] + x[8] + x[9] + x[10] + x[11] + x[12] + x[13] + x[14] + x[15] + x[16] + x[17] + x[18] + x[20] + x[22] + x[23] + x[25] + x[26] + x[29] + x[30], MathOptInterface.LessThan{Float64}(23.0))
+Called with accumulated = 10.54116374481702
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(x[1] + x[2] + x[3] + x[4] + x[5] + x[7] + x[8] + x[9] + x[10] + x[11] + x[12] + x[13] + x[14] + x[16] + x[17] + x[18] + x[20] + x[22] + x[23] + x[25] + x[26] + x[27] + x[29] + x[30], MathOptInterface.LessThan{Float64}(24.0))
+Called with accumulated = 10.37975831721494
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(x[1] + x[2] + x[3] + x[4] + x[5] + x[7] + x[8] + x[9] + x[10] + x[12] + x[13] + x[14] + x[16] + x[17] + x[18] + x[20] + x[22] + x[23] + x[25] + x[26] + x[28] + x[29] + x[30], MathOptInterface.LessThan{Float64}(23.0))
+callback_called = true</code></pre><h2 id="Heuristic-solutions"><a class="docs-heading-anchor" href="#Heuristic-solutions">Heuristic solutions</a><a id="Heuristic-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Heuristic-solutions" title="Permalink"></a></h2><p>An example using a heuristic solution callback.</p><pre><code class="language-julia hljs">function example_heuristic_solution()
+    Random.seed!(1)
+    N = 30
+    item_weights, item_values = rand(N), rand(N)
+    model = Model(GLPK.Optimizer)
+    @variable(model, x[1:N], Bin)
+    @constraint(model, sum(item_weights[i] * x[i] for i in 1:N) &lt;= 10)
+    @objective(model, Max, sum(item_values[i] * x[i] for i in 1:N))
+    callback_called = false
+    function my_callback_function(cb_data)
+        callback_called = true
+        x_vals = callback_value.(Ref(cb_data), x)
+        ret =
+            MOI.submit(model, MOI.HeuristicSolution(cb_data), x, floor.(x_vals))
+        println(&quot;Heuristic solution status = $(ret)&quot;)
+    end
+    set_attribute(model, MOI.HeuristicCallback(), my_callback_function)
+    optimize!(model)
+    return
+end
+
+example_heuristic_solution()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Heuristic solution status = HEURISTIC_SOLUTION_ACCEPTED
+Heuristic solution status = HEURISTIC_SOLUTION_REJECTED
+Heuristic solution status = HEURISTIC_SOLUTION_REJECTED
+Heuristic solution status = HEURISTIC_SOLUTION_REJECTED</code></pre><h2 id="GLPK-solver-dependent-callback"><a class="docs-heading-anchor" href="#GLPK-solver-dependent-callback">GLPK solver-dependent callback</a><a id="GLPK-solver-dependent-callback-1"></a><a class="docs-heading-anchor-permalink" href="#GLPK-solver-dependent-callback" title="Permalink"></a></h2><p>An example using GLPK&#39;s solver-dependent callback.</p><pre><code class="language-julia hljs">function example_solver_dependent_callback()
+    model = Model(GLPK.Optimizer)
+    @variable(model, 0 &lt;= x &lt;= 2.5, Int)
+    @variable(model, 0 &lt;= y &lt;= 2.5, Int)
+    @objective(model, Max, y)
+    lazy_called = false
+    function my_callback_function(cb_data)
+        lazy_called = true
+        reason = GLPK.glp_ios_reason(cb_data.tree)
+        println(&quot;Called from reason = $(reason)&quot;)
+        if reason != GLPK.GLP_IROWGEN
+            return
+        end
+        x_val = callback_value(cb_data, x)
+        y_val = callback_value(cb_data, y)
+        if y_val - x_val &gt; 1 + 1e-6
+            con = @build_constraint(y - x &lt;= 1)
+            println(&quot;Adding $(con)&quot;)
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        elseif y_val + x_val &gt; 3 + 1e-6
+            con = @build_constraint(y - x &lt;= 1)
+            println(&quot;Adding $(con)&quot;)
+            MOI.submit(model, MOI.LazyConstraint(cb_data), con)
+        end
+    end
+    set_attribute(model, GLPK.CallbackFunction(), my_callback_function)
+    optimize!(model)
+    return
+end
+
+example_solver_dependent_callback()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Called from reason = 6
+Called from reason = 7
+Called from reason = 1
+Adding ScalarConstraint{AffExpr, MathOptInterface.LessThan{Float64}}(y - x, MathOptInterface.LessThan{Float64}(1.0))
+Called from reason = 7
+Called from reason = 1
+Called from reason = 2</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../constraint_programming/">« Constraint programming</a><a class="docs-footer-nextpage" href="../lp_sensitivity/">Sensitivity analysis of a linear program »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/cannery.jl b/previews/PR3525/tutorials/linear/cannery.jl
new file mode 100644
index 00000000000..aaf3c3b344e
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/cannery.jl
@@ -0,0 +1,130 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The cannery problem
+
+# **This tutorial was originally contributed by Louis Luangkesorn.**
+
+# This tutorial solves the cannery problem from Dantzig,
+# _Linear Programming and Extensions_, Princeton University Press, Princeton,
+# NJ, 1963. This class of problem is known as a transshipment problem.
+
+# The purpose of this tutorial is to demonstrate how to use JSON data in the
+# formulation of a JuMP model.
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import HiGHS
+import JSON
+import Test  #src
+
+# ## Formulation
+
+# The cannery problem assumes we are optimizing the shipment of cases of 
+# cans from production plants ``p \in P`` to markets ``m \in M``.
+
+# Each production plant ``p`` has a capacity ``c_p``, and each market ``m``
+# has a demand ``d_m``. The shipping cost per case of cans from plant ``p`` 
+# to market ``m`` is ``d_{p,m}``.
+
+# We wish to find the distribution plan ``x_{p,m}``, the number of cases of cans
+# to ship from plant ``p`` to market ``m``, for ``p \in P`` and ``m \in M``  
+# that minimizes the shipping costs. We can formulate our problem as the
+# following linear program:
+# ```math
+# \begin{aligned}
+# \min & \sum\limits_{p \in P}\sum\limits_{m \in M} d_{p,m} x_{p,m} \\
+# \text{s.t.} & \sum\limits_{m \in M} x_{p,m} \le c_p, && \forall p \in P \\
+#             & \sum\limits_{p \in P} x_{p,m} \ge d_m, && \forall m \in M \\
+#             & x_{p,m} \ge 0, && \forall p \in P, m \in M
+# \end{aligned}
+# ```
+
+# ## Data
+
+# A key feature of the tutorial is to demonstrate how to load data from JSON.
+
+# For simplicity, we've hard-coded it below. But if the data was available as a
+# `.json` file, we could use `data = JSON.parsefile(filename)` to read in the
+# data.
+
+data = JSON.parse("""
+{
+    "plants": {
+        "Seattle": {"capacity": 350},
+        "San-Diego": {"capacity": 600}
+    },
+    "markets": {
+        "New-York": {"demand": 300},
+        "Chicago": {"demand": 300},
+        "Topeka": {"demand": 300}
+    },
+    "distances": {
+        "Seattle => New-York": 2.5,
+        "Seattle => Chicago": 1.7,
+        "Seattle => Topeka": 1.8,
+        "San-Diego => New-York": 2.5,
+        "San-Diego => Chicago": 1.8,
+        "San-Diego => Topeka": 1.4
+    }
+}
+""")
+
+# Create the set of plants:
+
+P = keys(data["plants"])
+
+# Create the set of markets:
+
+M = keys(data["markets"])
+
+# We also need a function to compute the distance from plant to market:
+
+distance(p::String, m::String) = data["distances"]["$(p) => $(m)"]
+
+# ## JuMP formulation
+
+# Now we're ready to convert our mathematical formulation into a JuMP model.
+
+# First, create a new JuMP model. Since we have a linear program, we'll use
+# HiGHS as our optimizer:
+
+model = Model(HiGHS.Optimizer)
+
+# Our decision variables are indexed over the set of plants and markets:
+
+@variable(model, x[P, M] >= 0)
+
+# We need a constraint that each plant can ship no more than its capacity:
+
+@constraint(model, [p in P], sum(x[p, :]) <= data["plants"][p]["capacity"])
+
+# and each market must receive at least its demand:
+
+@constraint(model, [m in M], sum(x[:, m]) >= data["markets"][m]["demand"])
+
+# Finally, our objective is to minimize the transportation distance:
+
+@objective(model, Min, sum(distance(p, m) * x[p, m] for p in P, m in M));
+
+# ## Solution
+
+# Let's optimize and look at the solution:
+
+optimize!(model)
+solution_summary(model)
+
+# What's the optimal shipment?
+
+for p in P, m in M
+    println(p, " => ", m, ": ", value(x[p, m]))
+end
+
+Test.@test termination_status(model) == OPTIMAL    #src
+Test.@test primal_status(model) == FEASIBLE_POINT  #src
+Test.@test objective_value(model) == 1_680.0           #src
diff --git a/previews/PR3525/tutorials/linear/cannery/index.html b/previews/PR3525/tutorials/linear/cannery/index.html
new file mode 100644
index 00000000000..69eb42f271f
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/cannery/index.html
@@ -0,0 +1,89 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The cannery problem · JuMP</title><meta name="title" content="The cannery problem · JuMP"/><meta property="og:title" content="The cannery problem · JuMP"/><meta property="twitter:title" content="The cannery problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li class="is-active"><a class="tocitem" href>The cannery problem</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Solution"><span>Solution</span></a></li></ul></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The cannery problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The cannery problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/cannery.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-cannery-problem"><a class="docs-heading-anchor" href="#The-cannery-problem">The cannery problem</a><a id="The-cannery-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-cannery-problem" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../cannery.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Louis Luangkesorn.</strong></p><p>This tutorial solves the cannery problem from Dantzig, <em>Linear Programming and Extensions</em>, Princeton University Press, Princeton, NJ, 1963. This class of problem is known as a transshipment problem.</p><p>The purpose of this tutorial is to demonstrate how to use JSON data in the formulation of a JuMP model.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+import JSON</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>The cannery problem assumes we are optimizing the shipment of cases of cans from production plants <span>$p \in P$</span> to markets <span>$m \in M$</span>.</p><p>Each production plant <span>$p$</span> has a capacity <span>$c_p$</span>, and each market <span>$m$</span> has a demand <span>$d_m$</span>. The shipping cost per case of cans from plant <span>$p$</span> to market <span>$m$</span> is <span>$d_{p,m}$</span>.</p><p>We wish to find the distribution plan <span>$x_{p,m}$</span>, the number of cases of cans to ship from plant <span>$p$</span> to market <span>$m$</span>, for <span>$p \in P$</span> and <span>$m \in M$</span> that minimizes the shipping costs. We can formulate our problem as the following linear program:</p><p class="math-container">\[\begin{aligned}
+\min &amp; \sum\limits_{p \in P}\sum\limits_{m \in M} d_{p,m} x_{p,m} \\
+\text{s.t.} &amp; \sum\limits_{m \in M} x_{p,m} \le c_p, &amp;&amp; \forall p \in P \\
+            &amp; \sum\limits_{p \in P} x_{p,m} \ge d_m, &amp;&amp; \forall m \in M \\
+            &amp; x_{p,m} \ge 0, &amp;&amp; \forall p \in P, m \in M
+\end{aligned}\]</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>A key feature of the tutorial is to demonstrate how to load data from JSON.</p><p>For simplicity, we&#39;ve hard-coded it below. But if the data was available as a <code>.json</code> file, we could use <code>data = JSON.parsefile(filename)</code> to read in the data.</p><pre><code class="language-julia hljs">data = JSON.parse(&quot;&quot;&quot;
+{
+    &quot;plants&quot;: {
+        &quot;Seattle&quot;: {&quot;capacity&quot;: 350},
+        &quot;San-Diego&quot;: {&quot;capacity&quot;: 600}
+    },
+    &quot;markets&quot;: {
+        &quot;New-York&quot;: {&quot;demand&quot;: 300},
+        &quot;Chicago&quot;: {&quot;demand&quot;: 300},
+        &quot;Topeka&quot;: {&quot;demand&quot;: 300}
+    },
+    &quot;distances&quot;: {
+        &quot;Seattle =&gt; New-York&quot;: 2.5,
+        &quot;Seattle =&gt; Chicago&quot;: 1.7,
+        &quot;Seattle =&gt; Topeka&quot;: 1.8,
+        &quot;San-Diego =&gt; New-York&quot;: 2.5,
+        &quot;San-Diego =&gt; Chicago&quot;: 1.8,
+        &quot;San-Diego =&gt; Topeka&quot;: 1.4
+    }
+}
+&quot;&quot;&quot;)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Dict{String, Any} with 3 entries:
+  &quot;plants&quot;    =&gt; Dict{String, Any}(&quot;Seattle&quot;=&gt;Dict{String, Any}(&quot;capacity&quot;=&gt;350…
+  &quot;distances&quot; =&gt; Dict{String, Any}(&quot;San-Diego =&gt; New-York&quot;=&gt;2.5, &quot;Seattle =&gt; To…
+  &quot;markets&quot;   =&gt; Dict{String, Any}(&quot;Chicago&quot;=&gt;Dict{String, Any}(&quot;demand&quot;=&gt;300),…</code></pre><p>Create the set of plants:</p><pre><code class="language-julia hljs">P = keys(data[&quot;plants&quot;])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">KeySet for a Dict{String, Any} with 2 entries. Keys:
+  &quot;Seattle&quot;
+  &quot;San-Diego&quot;</code></pre><p>Create the set of markets:</p><pre><code class="language-julia hljs">M = keys(data[&quot;markets&quot;])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">KeySet for a Dict{String, Any} with 3 entries. Keys:
+  &quot;Chicago&quot;
+  &quot;Topeka&quot;
+  &quot;New-York&quot;</code></pre><p>We also need a function to compute the distance from plant to market:</p><pre><code class="language-julia hljs">distance(p::String, m::String) = data[&quot;distances&quot;][&quot;$(p) =&gt; $(m)&quot;]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">distance (generic function with 1 method)</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>Now we&#39;re ready to convert our mathematical formulation into a JuMP model.</p><p>First, create a new JuMP model. Since we have a linear program, we&#39;ll use HiGHS as our optimizer:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><p>Our decision variables are indexed over the set of plants and markets:</p><pre><code class="language-julia hljs">@variable(model, x[P, M] &gt;= 0)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-dimensional DenseAxisArray{VariableRef,2,...} with index sets:
+    Dimension 1, [&quot;Seattle&quot;, &quot;San-Diego&quot;]
+    Dimension 2, [&quot;Chicago&quot;, &quot;Topeka&quot;, &quot;New-York&quot;]
+And data, a 2×3 Matrix{VariableRef}:
+ x[Seattle,Chicago]    x[Seattle,Topeka]    x[Seattle,New-York]
+ x[San-Diego,Chicago]  x[San-Diego,Topeka]  x[San-Diego,New-York]</code></pre><p>We need a constraint that each plant can ship no more than its capacity:</p><pre><code class="language-julia hljs">@constraint(model, [p in P], sum(x[p, :]) &lt;= data[&quot;plants&quot;][p][&quot;capacity&quot;])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape},1,...} with index sets:
+    Dimension 1, [&quot;Seattle&quot;, &quot;San-Diego&quot;]
+And data, a 2-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.LessThan{Float64}}, ScalarShape}}:
+ x[Seattle,Chicago] + x[Seattle,Topeka] + x[Seattle,New-York] ≤ 350
+ x[San-Diego,Chicago] + x[San-Diego,Topeka] + x[San-Diego,New-York] ≤ 600</code></pre><p>and each market must receive at least its demand:</p><pre><code class="language-julia hljs">@constraint(model, [m in M], sum(x[:, m]) &gt;= data[&quot;markets&quot;][m][&quot;demand&quot;])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.GreaterThan{Float64}}, ScalarShape},1,...} with index sets:
+    Dimension 1, [&quot;Chicago&quot;, &quot;Topeka&quot;, &quot;New-York&quot;]
+And data, a 3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.GreaterThan{Float64}}, ScalarShape}}:
+ x[Seattle,Chicago] + x[San-Diego,Chicago] ≥ 300
+ x[Seattle,Topeka] + x[San-Diego,Topeka] ≥ 300
+ x[Seattle,New-York] + x[San-Diego,New-York] ≥ 300</code></pre><p>Finally, our objective is to minimize the transportation distance:</p><pre><code class="language-julia hljs">@objective(model, Min, sum(distance(p, m) * x[p, m] for p in P, m in M));</code></pre><h2 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h2><p>Let&#39;s optimize and look at the solution:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 1.68000e+03
+  Objective bound    : 0.00000e+00
+  Relative gap       : Inf
+  Dual objective value : 1.68000e+03
+
+* Work counters
+  Solve time (sec)   : 2.44141e-04
+  Simplex iterations : 3
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><p>What&#39;s the optimal shipment?</p><pre><code class="language-julia hljs">for p in P, m in M
+    println(p, &quot; =&gt; &quot;, m, &quot;: &quot;, value(x[p, m]))
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Seattle =&gt; Chicago: 300.0
+Seattle =&gt; Topeka: 0.0
+Seattle =&gt; New-York: 0.0
+San-Diego =&gt; Chicago: 0.0
+San-Diego =&gt; Topeka: 300.0
+San-Diego =&gt; New-York: 300.0</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../diet/">« The diet problem</a><a class="docs-footer-nextpage" href="../factory_schedule/">The factory schedule example »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/commodity_nz.db b/previews/PR3525/tutorials/linear/commodity_nz.db
new file mode 100644
index 0000000000000000000000000000000000000000..77ba15ddcffbdb85da151b006a4442fe7d6784e0
GIT binary patch
literal 45056
zcmeI5U1%It6vt;~XFsz$XY;YyY_^--Bn?g4nivWq_z=?)YEAl)G=3B*%XYJE)@-ty
z-3hVwK@vs!Ac#~!@xdo43RMa|_$1;(Rq&w-f)x>iltQqmZz4W;@666*HZ!|vQz$h3
z&oY_GJ!kH{=QnqToOAD<oSdA<tGU$KVtGEJrVa>UK@^22Qz=1Ur-IPLb}!o{w)eAL
zX1lS8#z})?SQs3CD-=8;_yRYDrZ1GI12==WLvIGZr0L+LU?o^-a4ke200JNY0w4ea
zAOHgQOdxnfydE7I5?^grGiM69%3|S6TFuRu(rSkNjrI!{t!A|I>(Sw1adph5#;!^a
zKWe>DFOd#Sjg3x^rKU#@PK>2`><T@p{(!${CRb7O3mG+ET<Gynk3Bz~I(D4x&rD40
zXVOx+IJ1;h9V9ck`OLyhkDoO%VZG$U)RCj3Q>RknW2gFSD&KEbHV_zCdA8LXO(ewC
zS9E)3^QBUL;asJ9bZzQ;wOH?Qv7A4bck1l6P03f(%tAJoo}1soo@&$k8yMzm)(l(8
zYKB$$Rg7bx)-a=3H>^JFY_X!IOS$saJv8cIj6+?*(;DMtN5<}rt{m#{M4uZLh5W)y
z?j<%*nE#Q^EU889_jF}(ACr&7>dEzykRH~Lo_+0&lq*Z6QsI1hSU>uP+C0%iiF!2~
z!-1(+M}S|sAS^~>F>!TLH%hBcuWa*x*F;+k83Pv|`76gFUhR1s*Q?ea#g)fdDQ~b^
zYH83yP<z7ew!zv*Et@H2vU&BqeF*mO+ix$fdD6_0X;{Ptz)(vgxK+hUde}JQuH}up
zxf{NH!AAGGc3?24RotVIJfQK_rTR~KV=Qu#df_(SPb;HQulAaJ#Ci+oODhAcoZm#f
z+}iY}QM2u)EicV#9ou|0hS**3(NhB5rXT4m`k3CK3pB$Duz>&wfB*=900@8p2!H?x
zfB*=900`I!bo%7Jq?ny8v*=!ScB!15oy#xh&z6_+svPmkeUTd3e7-OzcX;IPq_~_b
z6j&@$EiRZBNUd(UJ7SS=b*WvJ6G>6cER{10=Q110%`Q0+G0SyTiK}LovU3F%H8oYk
z%_3Jd%DHNo`~P9ZFVIc;hAz_^v`8oEF^VhyDnBcqDQ_!_%5i0gDPRKu5C8!X009sH
z0T2KI5C8!X*ck+NhuwXNq-6DA4M)**Amza>awp;&Ug&ZrKbKL9rovoGaVJ<UhL>uq
zL0Ib!x#QhQsd}Y@_bXoP3cBOWf32$6>L~4Qawqzdk|D5m+3X4HYC!9%p|!cIrKF!3
z9C0*QbI%83K32hoXI-s;%jV)<W@f}@CYK3oyFARoh}}Xi6)(oz%sA%U>*NejTCFkJ
zDl$6GMWv*RnH+I6nakzDxTN*SxX4<suoe?FOAMjKk`nj-`AUCn{{KB%p#n|O6O`ni
zoiVUs1qgru2!H?xfB*=900@8p2!H?xY)POq&|q5M?>w#Vt2?dlb(+@q)S1?I+ottd
zZwJh<h#BeNU(IhZ0}4DEAhqA!{J%`^^7;So>2r314Fo^{1V8`;KmY_l00ck)1V8`;
zK;XU+=#t&+I|E+p+6M002S$EBplwmMARsCFjd^}E8qnFo76L>z1pFcooBwAE8N8Nx
z|G>y^x+wv6@X@3|Khd8w$!7mQrE9cG@835=(GUoL00@8p2!H?xfB*=900@8p2t0@c
zI_oYN@I{=~3wS%~EEn*!+E)v>+Z`7R$jzJA3b-^yW5K@}9gxCI(Qg#<*gzFA&*d-E
zUK*o|^Z~s<D*rr)_|az&009sH0T2KI5C8!X009sH0T8%10`(pHfTLsYujknN92|SE
zvt#eEIreU5#2zDifa%xsf6LcwlybNJj{Sc=|1Z;3fv(cm^eg>Ew?ZE(e<%Su7+R-2
z%8PV~&MSNAbm&`Z;-7nCh^s*W1V8`;KmY_l00ck)1V8`;8WQN`srhOW$Qx6*rU%%j
zj0-i{<g1IlO@-F3hMD_Xge1@0XXcT&HPD>mdHW0>(;;tPSh*r3O$FP+bn0rD)KBT<
zIsMGs^wxFnrU@bDex^2bY)u4p6OFEKCeYNy6aG13)Rxw>Td>QTb>FgF-XE)%3eYEX
z^2B>KZoEcsoF@XbBjYkp%$n=ZFojFpH4?)ZE|HmieU7mP**SdRvV{*MlH(D-#WL>y
zhm~GIxkgv$J7t}2DZS)Z&eB2UF!d{!HUIy0I^A%5LI?sN00JNY0w4eaAOHd&00JPe
z6A0Wv^da0hxIkOP7L=`G3!%orgWw$m519AQ|8s%iPj3zm1R4hi8Uz6DxBqXNV2zRR
z@PJG2iD98NrXY3j_=UOf-{SxC{QnZ&5ZJf>Z|sD@3L8KG1V8`;KmY_l00ck)1V8`;
zK;WSw&?dQ(UUTWh$ZIdJ9sZ>40*S7Omj??*PMkUV!`~nM$LIegS{G=Y{&}eQz*rCf
m0T2KI5C8!X009sH0T2KI5ZI9fLawA}#seg;E8^#Ifd2s$yJ}hh

literal 0
HcmV?d00001

diff --git a/previews/PR3525/tutorials/linear/constraint_programming.jl b/previews/PR3525/tutorials/linear/constraint_programming.jl
new file mode 100644
index 00000000000..ed47265c03f
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/constraint_programming.jl
@@ -0,0 +1,190 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Constraint programming
+
+# JuMP supports a range of constraint-programming type constraints via the
+# corresponding sets in MathOptInterface. For most constraints, there are
+# reformulations built-in that convert the constraint programming constraint
+# into a mixed-integer programming equivalent.
+
+# Because of this reformulation, all variables must be integer, and they must
+# typically have finite bounds. An error will be thrown if the reformulation
+# requires finiteness and you have a variable with non-finite bounds.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import HiGHS
+
+# ## AllDifferent
+
+# The [`MOI.AllDifferent`](@ref) set ensures that every element in a list takes
+# a different integer value.
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 1 <= x[1:4] <= 4, Int)
+@constraint(model, x in MOI.AllDifferent(4))
+optimize!(model)
+value.(x)
+
+# ## BinPacking
+
+# The [`MOI.BinPacking`](@ref) set can be used to divide up a set of items into
+# different groups, such that the sum of their weights does not exceed the
+# capacity of a bin.
+
+weights, capacity = Float64[1, 1, 2, 2, 3], 3.0;
+number_of_bins = 3
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 1 <= x[1:length(weights)] <= number_of_bins, Int)
+@constraint(model, x in MOI.BinPacking(capacity, weights))
+optimize!(model)
+value.(x)
+
+# Here, the value of `x[i]` is the bin that item `i` was placed into.
+
+# ## Circuit
+
+# The [`MOI.Circuit`](@ref) set is used to construct a tour of a list of `N`
+# variables. They will each be assigned an integer from `1` to `N`, that
+# describes the successor to each variable in the list:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:4], Int)
+@constraint(model, x in MOI.Circuit(4))
+optimize!(model)
+
+# Let's see what tour was found, starting at node number `1`:
+y = round.(Int, value.(x))
+tour = Int[1]
+while length(tour) < length(y)
+    push!(tour, y[tour[end]])
+end
+tour
+
+# ## CountAtLeast
+
+# The [`MOI.CountAtLeast`](@ref) set is used to ensure that at least `n`
+# elements in a set of variables belong to a set of values.
+
+# For example, here is a model with three variables, constrained between 0 and
+# 5:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 <= x[1:3] <= 5, Int)
+
+# If we want to ensure that at least one element of each set `{x[1], x[2]}` and
+# `{x[2], x[3]}` is in the set `{3}`, then we create a list of variables by
+# concatenating the sets together:
+
+variables = [x[1], x[2], x[2], x[3]]
+
+# Then we need a partition list that contains the number of elements in each
+# set of variables:
+
+partitions = [2, 2]
+
+# Finally, we need a set of values that the elements must be a part of:
+
+values = Set([3])
+
+# And the number of elements that must be part of the set `values`:
+
+n = 1
+
+# The constraint is:
+
+@constraint(model, variables in MOI.CountAtLeast(n, partitions, values))
+
+# To ensure the uniqueness of the solution, we'll add a constraint that `x[2]`
+# must be `<= 2`. This ensures that the only feasible solution is for `x[1]` and
+# `x[3]` to be `3`:
+
+@constraint(model, x[2] <= 2)
+
+# Let's check that we found a valid solution:
+
+optimize!(model)
+value.(x)
+
+# ## CountBelongs
+
+# The [`MOI.CountBelongs`](@ref) set is used to count how many elements in a set
+# of variables belong to a set of values.
+
+# For example, to count how many elements in a set of 4 variables belong to the
+# set `{2, 3}`, do:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 <= x[i = 1:4] <= i, Int)
+@variable(model, n, Int)
+@objective(model, Max, sum(x))
+set = Set([2, 3])
+@constraint(model, [n; x] in MOI.CountBelongs(1 + length(x), set))
+optimize!(model)
+value(n), value.(x)
+
+# ## CountDistinct
+
+# The [`MOI.CountDistinct`](@ref) set is used to count the number of distinct
+# elements in a set of variables.
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 <= x[i = 1:4] <= i, Int)
+@variable(model, n, Int)
+@objective(model, Max, sum(x))
+@constraint(model, [n; x] in MOI.CountDistinct(1 + length(x)))
+optimize!(model)
+value(n), value.(x)
+
+# ## CountGreaterThan
+
+# The [`MOI.CountGreaterThan`](@ref) set is used to strictly upper-bound the
+# number of distinct elements in a set of variables that have a value equal to
+# another variable.
+
+# For example, to count the number `n` of times that `y` appears in the vector
+# `x`, use:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 <= x[i = 1:4] <= i, Int)
+@variable(model, n, Int)
+@variable(model, 3 <= y <= 4, Int)
+@objective(model, Max, sum(x))
+@constraint(model, [n; y; x] in MOI.CountGreaterThan(1 + 1 + length(x)))
+optimize!(model)
+value(n), value(y), value.(x)
+
+# Here `n` is strictly greater than the count, and there is no limit on how
+# large `n` could be. For example, `n = 100` is also a feasible solution. The
+# only constraint is that `n` cannot be equal to or smaller than the number of
+# times that `y` appears.
+
+# ## Table
+
+# The [`MOI.Table`](@ref) set is used to select a single row from a matrix of
+# values.
+
+# For example, given a matrix:
+
+table = Float64[1 1 0; 0 1 1; 1 0 1; 1 1 1]
+
+# we can constraint a 3-element vector `x` to equal one of the rows in `table`
+# via:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[i = 1:3], Int)
+@constraint(model, x in MOI.Table(table))
+optimize!(model)
+value.(x)
diff --git a/previews/PR3525/tutorials/linear/constraint_programming/index.html b/previews/PR3525/tutorials/linear/constraint_programming/index.html
new file mode 100644
index 00000000000..5962f4bd399
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/constraint_programming/index.html
@@ -0,0 +1,92 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Constraint programming · JuMP</title><meta name="title" content="Constraint programming · JuMP"/><meta property="og:title" content="Constraint programming · JuMP"/><meta property="twitter:title" content="Constraint programming · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li class="is-active"><a class="tocitem" href>Constraint programming</a><ul class="internal"><li><a class="tocitem" href="#AllDifferent"><span>AllDifferent</span></a></li><li><a class="tocitem" href="#BinPacking"><span>BinPacking</span></a></li><li><a class="tocitem" href="#Circuit"><span>Circuit</span></a></li><li><a class="tocitem" href="#CountAtLeast"><span>CountAtLeast</span></a></li><li><a class="tocitem" href="#CountBelongs"><span>CountBelongs</span></a></li><li><a class="tocitem" href="#CountDistinct"><span>CountDistinct</span></a></li><li><a class="tocitem" href="#CountGreaterThan"><span>CountGreaterThan</span></a></li><li><a class="tocitem" href="#Table"><span>Table</span></a></li></ul></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Constraint programming</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Constraint programming</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/constraint_programming.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Constraint-programming"><a class="docs-heading-anchor" href="#Constraint-programming">Constraint programming</a><a id="Constraint-programming-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-programming" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../constraint_programming.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>JuMP supports a range of constraint-programming type constraints via the corresponding sets in MathOptInterface. For most constraints, there are reformulations built-in that convert the constraint programming constraint into a mixed-integer programming equivalent.</p><p>Because of this reformulation, all variables must be integer, and they must typically have finite bounds. An error will be thrown if the reformulation requires finiteness and you have a variable with non-finite bounds.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS</code></pre><h2 id="AllDifferent"><a class="docs-heading-anchor" href="#AllDifferent">AllDifferent</a><a id="AllDifferent-1"></a><a class="docs-heading-anchor-permalink" href="#AllDifferent" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.AllDifferent"><code>MOI.AllDifferent</code></a> set ensures that every element in a list takes a different integer value.</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 1 &lt;= x[1:4] &lt;= 4, Int)
+@constraint(model, x in MOI.AllDifferent(4))
+optimize!(model)
+value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{Float64}:
+ 1.0
+ 4.0
+ 3.0000000000000004
+ 1.9999999999999996</code></pre><h2 id="BinPacking"><a class="docs-heading-anchor" href="#BinPacking">BinPacking</a><a id="BinPacking-1"></a><a class="docs-heading-anchor-permalink" href="#BinPacking" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.BinPacking"><code>MOI.BinPacking</code></a> set can be used to divide up a set of items into different groups, such that the sum of their weights does not exceed the capacity of a bin.</p><pre><code class="language-julia hljs">weights, capacity = Float64[1, 1, 2, 2, 3], 3.0;
+number_of_bins = 3
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 1 &lt;= x[1:length(weights)] &lt;= number_of_bins, Int)
+@constraint(model, x in MOI.BinPacking(capacity, weights))
+optimize!(model)
+value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5-element Vector{Float64}:
+ 2.0
+ 1.0
+ 2.0
+ 1.0
+ 3.0</code></pre><p>Here, the value of <code>x[i]</code> is the bin that item <code>i</code> was placed into.</p><h2 id="Circuit"><a class="docs-heading-anchor" href="#Circuit">Circuit</a><a id="Circuit-1"></a><a class="docs-heading-anchor-permalink" href="#Circuit" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.Circuit"><code>MOI.Circuit</code></a> set is used to construct a tour of a list of <code>N</code> variables. They will each be assigned an integer from <code>1</code> to <code>N</code>, that describes the successor to each variable in the list:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:4], Int)
+@constraint(model, x in MOI.Circuit(4))
+optimize!(model)</code></pre><p>Let&#39;s see what tour was found, starting at node number <code>1</code>:</p><pre><code class="language-julia hljs">y = round.(Int, value.(x))
+tour = Int[1]
+while length(tour) &lt; length(y)
+    push!(tour, y[tour[end]])
+end
+tour</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{Int64}:
+ 1
+ 4
+ 3
+ 2</code></pre><h2 id="CountAtLeast"><a class="docs-heading-anchor" href="#CountAtLeast">CountAtLeast</a><a id="CountAtLeast-1"></a><a class="docs-heading-anchor-permalink" href="#CountAtLeast" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.CountAtLeast"><code>MOI.CountAtLeast</code></a> set is used to ensure that at least <code>n</code> elements in a set of variables belong to a set of values.</p><p>For example, here is a model with three variables, constrained between 0 and 5:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 &lt;= x[1:3] &lt;= 5, Int)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code></pre><p>If we want to ensure that at least one element of each set <code>{x[1], x[2]}</code> and <code>{x[2], x[3]}</code> is in the set <code>{3}</code>, then we create a list of variables by concatenating the sets together:</p><pre><code class="language-julia hljs">variables = [x[1], x[2], x[2], x[3]]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[2]
+ x[3]</code></pre><p>Then we need a partition list that contains the number of elements in each set of variables:</p><pre><code class="language-julia hljs">partitions = [2, 2]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Int64}:
+ 2
+ 2</code></pre><p>Finally, we need a set of values that the elements must be a part of:</p><pre><code class="language-julia hljs">values = Set([3])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Set{Int64} with 1 element:
+  3</code></pre><p>And the number of elements that must be part of the set <code>values</code>:</p><pre><code class="language-julia hljs">n = 1</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1</code></pre><p>The constraint is:</p><pre><code class="language-julia hljs">@constraint(model, variables in MOI.CountAtLeast(n, partitions, values))</code></pre><p class="math-container">\[ [x_{1}, x_{2}, x_{2}, x_{3}] \in \text{MathOptInterface.CountAtLeast(1, [2, 2], Set([3]))} \]</p><p>To ensure the uniqueness of the solution, we&#39;ll add a constraint that <code>x[2]</code> must be <code>&lt;= 2</code>. This ensures that the only feasible solution is for <code>x[1]</code> and <code>x[3]</code> to be <code>3</code>:</p><pre><code class="language-julia hljs">@constraint(model, x[2] &lt;= 2)</code></pre><p class="math-container">\[ x_{2} \leq 2 \]</p><p>Let&#39;s check that we found a valid solution:</p><pre><code class="language-julia hljs">optimize!(model)
+value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{Float64}:
+ 3.0
+ 0.0
+ 3.0</code></pre><h2 id="CountBelongs"><a class="docs-heading-anchor" href="#CountBelongs">CountBelongs</a><a id="CountBelongs-1"></a><a class="docs-heading-anchor-permalink" href="#CountBelongs" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.CountBelongs"><code>MOI.CountBelongs</code></a> set is used to count how many elements in a set of variables belong to a set of values.</p><p>For example, to count how many elements in a set of 4 variables belong to the set <code>{2, 3}</code>, do:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 &lt;= x[i = 1:4] &lt;= i, Int)
+@variable(model, n, Int)
+@objective(model, Max, sum(x))
+set = Set([2, 3])
+@constraint(model, [n; x] in MOI.CountBelongs(1 + length(x), set))
+optimize!(model)
+value(n), value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(2.0, [1.0, 2.0, 3.0, 4.0])</code></pre><h2 id="CountDistinct"><a class="docs-heading-anchor" href="#CountDistinct">CountDistinct</a><a id="CountDistinct-1"></a><a class="docs-heading-anchor-permalink" href="#CountDistinct" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.CountDistinct"><code>MOI.CountDistinct</code></a> set is used to count the number of distinct elements in a set of variables.</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 &lt;= x[i = 1:4] &lt;= i, Int)
+@variable(model, n, Int)
+@objective(model, Max, sum(x))
+@constraint(model, [n; x] in MOI.CountDistinct(1 + length(x)))
+optimize!(model)
+value(n), value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(4.0, [1.0, 2.0, 3.0, 4.0])</code></pre><h2 id="CountGreaterThan"><a class="docs-heading-anchor" href="#CountGreaterThan">CountGreaterThan</a><a id="CountGreaterThan-1"></a><a class="docs-heading-anchor-permalink" href="#CountGreaterThan" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.CountGreaterThan"><code>MOI.CountGreaterThan</code></a> set is used to strictly upper-bound the number of distinct elements in a set of variables that have a value equal to another variable.</p><p>For example, to count the number <code>n</code> of times that <code>y</code> appears in the vector <code>x</code>, use:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, 0 &lt;= x[i = 1:4] &lt;= i, Int)
+@variable(model, n, Int)
+@variable(model, 3 &lt;= y &lt;= 4, Int)
+@objective(model, Max, sum(x))
+@constraint(model, [n; y; x] in MOI.CountGreaterThan(1 + 1 + length(x)))
+optimize!(model)
+value(n), value(y), value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(2.0, 3.0, [1.0, 2.0, 3.0, 4.0])</code></pre><p>Here <code>n</code> is strictly greater than the count, and there is no limit on how large <code>n</code> could be. For example, <code>n = 100</code> is also a feasible solution. The only constraint is that <code>n</code> cannot be equal to or smaller than the number of times that <code>y</code> appears.</p><h2 id="Table"><a class="docs-heading-anchor" href="#Table">Table</a><a id="Table-1"></a><a class="docs-heading-anchor-permalink" href="#Table" title="Permalink"></a></h2><p>The <a href="../../../moi/reference/standard_form/#MathOptInterface.Table"><code>MOI.Table</code></a> set is used to select a single row from a matrix of values.</p><p>For example, given a matrix:</p><pre><code class="language-julia hljs">table = Float64[1 1 0; 0 1 1; 1 0 1; 1 1 1]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4×3 Matrix{Float64}:
+ 1.0  1.0  0.0
+ 0.0  1.0  1.0
+ 1.0  0.0  1.0
+ 1.0  1.0  1.0</code></pre><p>we can constraint a 3-element vector <code>x</code> to equal one of the rows in <code>table</code> via:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[i = 1:3], Int)
+@constraint(model, x in MOI.Table(table))
+optimize!(model)
+value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{Float64}:
+ 1.0
+ 1.0
+ 0.0</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../n-queens/">« N-Queens</a><a class="docs-footer-nextpage" href="../callbacks/">Callbacks »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/diet.jl b/previews/PR3525/tutorials/linear/diet.jl
new file mode 100644
index 00000000000..6e9ef81915a
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/diet.jl
@@ -0,0 +1,194 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The diet problem
+
+# The purpose of this tutorial is to demonstrate how to incorporate DataFrames
+# into a JuMP model. As an example, we use classic [Stigler diet problem](https://en.wikipedia.org/wiki/Stigler_diet).
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import CSV
+import DataFrames
+import HiGHS
+import Test  #hide
+
+# ## Formulation
+
+# We wish to cook a nutritionally balanced meal by choosing the quantity of each
+# food ``f`` to eat from a set of foods ``F`` in our kitchen.
+
+# Each food ``f`` has a cost, ``c_f``, as well as a macro-nutrient profile
+# ``a_{m,f}`` for each macro-nutrient ``m \in M``.
+
+# Because we care about a nutritionally balanced meal, we set some minimum and
+# maximum limits for each nutrient, which we denote ``l_m`` and ``u_m``
+# respectively.
+
+# Furthermore, because we are optimizers, we seek the minimum cost solution.
+
+# With a little effort, we can formulate our dinner problem as the following
+# linear program:
+# ```math
+# \begin{aligned}
+# \min & \sum\limits_{f \in F} c_f x_f \\
+# \text{s.t.}\ \ & l_m \le \sum\limits_{f \in F} a_{m,f} x_f \le u_m, && \forall m \in M \\
+# & x_f \ge 0, && \forall f \in F.
+# \end{aligned}
+# ```
+
+# In the rest of this tutorial, we will create and solve this problem in JuMP,
+# and learn what we should cook for dinner.
+
+# ## Data
+
+# First, we need some data for the problem. For this tutorial, we'll write CSV
+# files to a temporary directory from Julia. If you have existing files, you
+# could change the filenames to point to them instead.
+
+dir = mktempdir()
+
+# The first file is a list of foods with their macro-nutrient profile:
+
+food_csv_filename = joinpath(dir, "diet_foods.csv")
+open(food_csv_filename, "w") do io
+    write(
+        io,
+        """
+        name,cost,calories,protein,fat,sodium
+        hamburger,2.49,410,24,26,730
+        chicken,2.89,420,32,10,1190
+        hot dog,1.50,560,20,32,1800
+        fries,1.89,380,4,19,270
+        macaroni,2.09,320,12,10,930
+        pizza,1.99,320,15,12,820
+        salad,2.49,320,31,12,1230
+        milk,0.89,100,8,2.5,125
+        ice cream,1.59,330,8,10,180
+        """,
+    )
+    return
+end
+foods = CSV.read(food_csv_filename, DataFrames.DataFrame)
+
+# Here, ``F`` is `foods.name` and ``c_f`` is `foods.cost`. (We're also playing
+# a bit loose the term "macro-nutrient" by including calories and sodium.)
+
+# We also need our minimum and maximum limits:
+
+nutrient_csv_filename = joinpath(dir, "diet_nutrient.csv")
+open(nutrient_csv_filename, "w") do io
+    write(
+        io,
+        """
+        nutrient,min,max
+        calories,1800,2200
+        protein,91,
+        fat,0,65
+        sodium,0,1779
+        """,
+    )
+    return
+end
+limits = CSV.read(nutrient_csv_filename, DataFrames.DataFrame)
+
+# Protein is missing data for the maximum. Let's fix that using `coalesce`:
+
+limits.max = coalesce.(limits.max, Inf)
+limits
+
+# ## JuMP formulation
+
+# Now we're ready to convert our mathematical formulation into a JuMP model.
+
+# First, create a new JuMP model. Since we have a linear program, we'll use
+# HiGHS as our optimizer:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+
+# Next, we create a set of decision variables `x`, with one element for each row
+# in the DataFrame, and each `x` has a lower bound of `0`:
+
+@variable(model, x[foods.name] >= 0)
+
+# To simplify things later on, we store the vector as a new column `x` in the
+# DataFrame `foods`. Since `x` is a `DenseAxisArray`, we first need to convert
+# it to an `Array`:
+
+foods.x = Array(x)
+
+# Our objective is to minimize the total cost of purchasing food:
+
+@objective(model, Min, sum(foods.cost .* foods.x));
+
+# For the next component, we need to add a constraint that our total intake of
+# each component is within the limits contained in the `limits` DataFrame:
+
+@constraint(
+    model,
+    [row in eachrow(limits)],
+    row.min <= sum(foods[!, row.nutrient] .* foods.x) <= row.max,
+);
+
+# What does our model look like?
+
+print(model)
+
+# ## Solution
+
+# Let's optimize and take a look at the solution:
+
+optimize!(model)
+Test.@test primal_status(model) == FEASIBLE_POINT        #hide
+Test.@test objective_value(model) ≈ 11.8288 atol = 1e-4  #hide
+solution_summary(model)
+
+# We found an optimal solution. Let's see what the optimal solution is:
+
+for row in eachrow(foods)
+    println(row.name, " = ", value(row.x))
+end
+
+# That's a lot of milk and ice cream, and sadly, we only get `0.6` of a
+# hamburger.
+
+# We can also use the function [`Containers.rowtable`](@ref) to easily convert
+# the result into a DataFrame:
+
+table = Containers.rowtable(value, x; header = [:food, :quantity])
+solution = DataFrames.DataFrame(table)
+
+# This makes it easy to perform analyses our solution:
+
+filter!(row -> row.quantity > 0.0, solution)
+
+# ## Problem modification
+
+# JuMP makes it easy to take an existing model and modify it by adding extra
+# constraints. Let's see what happens if we add a constraint that we can buy at
+# most 6 units of milk or ice cream combined.
+
+dairy_foods = ["milk", "ice cream"]
+is_dairy = map(name -> name in dairy_foods, foods.name)
+dairy_constraint = @constraint(model, sum(foods[is_dairy, :x]) <= 6)
+optimize!(model)
+Test.@test termination_status(model) == INFEASIBLE  #hide
+Test.@test primal_status(model) == NO_SOLUTION      #hide
+solution_summary(model)
+
+# There exists no feasible solution to our problem. Looks like we're stuck
+# eating ice cream for dinner.
+
+# ## Next steps
+
+# * You can delete a constraint using `delete(model, dairy_constraint)`. Can you
+#   add a different constraint to provide a diet with less dairy?
+# * Some food items (like hamburgers) are discrete. You can use `set_integer`
+#   to force a variable to take integer values. What happens to the solution if
+#   you do?
diff --git a/previews/PR3525/tutorials/linear/diet/index.html b/previews/PR3525/tutorials/linear/diet/index.html
new file mode 100644
index 00000000000..f9d46f1a8e7
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/diet/index.html
@@ -0,0 +1,144 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The diet problem · JuMP</title><meta name="title" content="The diet problem · JuMP"/><meta property="og:title" content="The diet problem · JuMP"/><meta property="twitter:title" content="The diet problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li class="is-active"><a class="tocitem" href>The diet problem</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Solution"><span>Solution</span></a></li><li><a class="tocitem" href="#Problem-modification"><span>Problem modification</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The diet problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The diet problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/diet.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-diet-problem"><a class="docs-heading-anchor" href="#The-diet-problem">The diet problem</a><a id="The-diet-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-diet-problem" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../diet.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to demonstrate how to incorporate DataFrames into a JuMP model. As an example, we use classic <a href="https://en.wikipedia.org/wiki/Stigler_diet">Stigler diet problem</a>.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import CSV
+import DataFrames
+import HiGHS</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>We wish to cook a nutritionally balanced meal by choosing the quantity of each food <span>$f$</span> to eat from a set of foods <span>$F$</span> in our kitchen.</p><p>Each food <span>$f$</span> has a cost, <span>$c_f$</span>, as well as a macro-nutrient profile <span>$a_{m,f}$</span> for each macro-nutrient <span>$m \in M$</span>.</p><p>Because we care about a nutritionally balanced meal, we set some minimum and maximum limits for each nutrient, which we denote <span>$l_m$</span> and <span>$u_m$</span> respectively.</p><p>Furthermore, because we are optimizers, we seek the minimum cost solution.</p><p>With a little effort, we can formulate our dinner problem as the following linear program:</p><p class="math-container">\[\begin{aligned}
+\min &amp; \sum\limits_{f \in F} c_f x_f \\
+\text{s.t.}\ \ &amp; l_m \le \sum\limits_{f \in F} a_{m,f} x_f \le u_m, &amp;&amp; \forall m \in M \\
+&amp; x_f \ge 0, &amp;&amp; \forall f \in F.
+\end{aligned}\]</p><p>In the rest of this tutorial, we will create and solve this problem in JuMP, and learn what we should cook for dinner.</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>First, we need some data for the problem. For this tutorial, we&#39;ll write CSV files to a temporary directory from Julia. If you have existing files, you could change the filenames to point to them instead.</p><pre><code class="language-julia hljs">dir = mktempdir()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">&quot;/tmp/jl_cbwSL8&quot;</code></pre><p>The first file is a list of foods with their macro-nutrient profile:</p><pre><code class="language-julia hljs">food_csv_filename = joinpath(dir, &quot;diet_foods.csv&quot;)
+open(food_csv_filename, &quot;w&quot;) do io
+    write(
+        io,
+        &quot;&quot;&quot;
+        name,cost,calories,protein,fat,sodium
+        hamburger,2.49,410,24,26,730
+        chicken,2.89,420,32,10,1190
+        hot dog,1.50,560,20,32,1800
+        fries,1.89,380,4,19,270
+        macaroni,2.09,320,12,10,930
+        pizza,1.99,320,15,12,820
+        salad,2.49,320,31,12,1230
+        milk,0.89,100,8,2.5,125
+        ice cream,1.59,330,8,10,180
+        &quot;&quot;&quot;,
+    )
+    return
+end
+foods = CSV.read(food_csv_filename, DataFrames.DataFrame)</code></pre><div><div style = "float: left;"><span>9×6 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">name</th><th style = "text-align: left;">cost</th><th style = "text-align: left;">calories</th><th style = "text-align: left;">protein</th><th style = "text-align: left;">fat</th><th style = "text-align: left;">sodium</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Int64" style = "text-align: left;">Int64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">hamburger</td><td style = "text-align: right;">2.49</td><td style = "text-align: right;">410</td><td style = "text-align: right;">24</td><td style = "text-align: right;">26.0</td><td style = "text-align: right;">730</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">chicken</td><td style = "text-align: right;">2.89</td><td style = "text-align: right;">420</td><td style = "text-align: right;">32</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">1190</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">hot dog</td><td style = "text-align: right;">1.5</td><td style = "text-align: right;">560</td><td style = "text-align: right;">20</td><td style = "text-align: right;">32.0</td><td style = "text-align: right;">1800</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">fries</td><td style = "text-align: right;">1.89</td><td style = "text-align: right;">380</td><td style = "text-align: right;">4</td><td style = "text-align: right;">19.0</td><td style = "text-align: right;">270</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">macaroni</td><td style = "text-align: right;">2.09</td><td style = "text-align: right;">320</td><td style = "text-align: right;">12</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">930</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">pizza</td><td style = "text-align: right;">1.99</td><td style = "text-align: right;">320</td><td style = "text-align: right;">15</td><td style = "text-align: right;">12.0</td><td style = "text-align: right;">820</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">salad</td><td style = "text-align: right;">2.49</td><td style = "text-align: right;">320</td><td style = "text-align: right;">31</td><td style = "text-align: right;">12.0</td><td style = "text-align: right;">1230</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">0.89</td><td style = "text-align: right;">100</td><td style = "text-align: right;">8</td><td style = "text-align: right;">2.5</td><td style = "text-align: right;">125</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">ice cream</td><td style = "text-align: right;">1.59</td><td style = "text-align: right;">330</td><td style = "text-align: right;">8</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">180</td></tr></tbody></table></div><p>Here, <span>$F$</span> is <code>foods.name</code> and <span>$c_f$</span> is <code>foods.cost</code>. (We&#39;re also playing a bit loose the term &quot;macro-nutrient&quot; by including calories and sodium.)</p><p>We also need our minimum and maximum limits:</p><pre><code class="language-julia hljs">nutrient_csv_filename = joinpath(dir, &quot;diet_nutrient.csv&quot;)
+open(nutrient_csv_filename, &quot;w&quot;) do io
+    write(
+        io,
+        &quot;&quot;&quot;
+        nutrient,min,max
+        calories,1800,2200
+        protein,91,
+        fat,0,65
+        sodium,0,1779
+        &quot;&quot;&quot;,
+    )
+    return
+end
+limits = CSV.read(nutrient_csv_filename, DataFrames.DataFrame)</code></pre><div><div style = "float: left;"><span>4×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">nutrient</th><th style = "text-align: left;">min</th><th style = "text-align: left;">max</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Union{Missing, Int64}" style = "text-align: left;">Int64?</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">calories</td><td style = "text-align: right;">1800</td><td style = "text-align: right;">2200</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">protein</td><td style = "text-align: right;">91</td><td style = "font-style: italic; text-align: right;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">fat</td><td style = "text-align: right;">0</td><td style = "text-align: right;">65</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">sodium</td><td style = "text-align: right;">0</td><td style = "text-align: right;">1779</td></tr></tbody></table></div><p>Protein is missing data for the maximum. Let&#39;s fix that using <code>coalesce</code>:</p><pre><code class="language-julia hljs">limits.max = coalesce.(limits.max, Inf)
+limits</code></pre><div><div style = "float: left;"><span>4×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">nutrient</th><th style = "text-align: left;">min</th><th style = "text-align: left;">max</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Real" style = "text-align: left;">Real</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">calories</td><td style = "text-align: right;">1800</td><td style = "text-align: right;">2200</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">protein</td><td style = "text-align: right;">91</td><td style = "text-align: right;">Inf</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">fat</td><td style = "text-align: right;">0</td><td style = "text-align: right;">65</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">sodium</td><td style = "text-align: right;">0</td><td style = "text-align: right;">1779</td></tr></tbody></table></div><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>Now we&#39;re ready to convert our mathematical formulation into a JuMP model.</p><p>First, create a new JuMP model. Since we have a linear program, we&#39;ll use HiGHS as our optimizer:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)</code></pre><p>Next, we create a set of decision variables <code>x</code>, with one element for each row in the DataFrame, and each <code>x</code> has a lower bound of <code>0</code>:</p><pre><code class="language-julia hljs">@variable(model, x[foods.name] &gt;= 0)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{VariableRef,1,...} with index sets:
+    Dimension 1, InlineStrings.String15[&quot;hamburger&quot;, &quot;chicken&quot;, &quot;hot dog&quot;, &quot;fries&quot;, &quot;macaroni&quot;, &quot;pizza&quot;, &quot;salad&quot;, &quot;milk&quot;, &quot;ice cream&quot;]
+And data, a 9-element Vector{VariableRef}:
+ x[hamburger]
+ x[chicken]
+ x[hot dog]
+ x[fries]
+ x[macaroni]
+ x[pizza]
+ x[salad]
+ x[milk]
+ x[ice cream]</code></pre><p>To simplify things later on, we store the vector as a new column <code>x</code> in the DataFrame <code>foods</code>. Since <code>x</code> is a <code>DenseAxisArray</code>, we first need to convert it to an <code>Array</code>:</p><pre><code class="language-julia hljs">foods.x = Array(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9-element Vector{VariableRef}:
+ x[hamburger]
+ x[chicken]
+ x[hot dog]
+ x[fries]
+ x[macaroni]
+ x[pizza]
+ x[salad]
+ x[milk]
+ x[ice cream]</code></pre><p>Our objective is to minimize the total cost of purchasing food:</p><pre><code class="language-julia hljs">@objective(model, Min, sum(foods.cost .* foods.x));</code></pre><p>For the next component, we need to add a constraint that our total intake of each component is within the limits contained in the <code>limits</code> DataFrame:</p><pre><code class="language-julia hljs">@constraint(
+    model,
+    [row in eachrow(limits)],
+    row.min &lt;= sum(foods[!, row.nutrient] .* foods.x) &lt;= row.max,
+);</code></pre><p>What does our model look like?</p><pre><code class="language-julia hljs">print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 2.49 x[hamburger] + 2.89 x[chicken] + 1.5 x[hot dog] + 1.89 x[fries] + 2.09 x[macaroni] + 1.99 x[pizza] + 2.49 x[salad] + 0.89 x[milk] + 1.59 x[ice cream]
+Subject to
+ 410 x[hamburger] + 420 x[chicken] + 560 x[hot dog] + 380 x[fries] + 320 x[macaroni] + 320 x[pizza] + 320 x[salad] + 100 x[milk] + 330 x[ice cream] ∈ [1800, 2200]
+ 24 x[hamburger] + 32 x[chicken] + 20 x[hot dog] + 4 x[fries] + 12 x[macaroni] + 15 x[pizza] + 31 x[salad] + 8 x[milk] + 8 x[ice cream] ∈ [91, Inf]
+ 26 x[hamburger] + 10 x[chicken] + 32 x[hot dog] + 19 x[fries] + 10 x[macaroni] + 12 x[pizza] + 12 x[salad] + 2.5 x[milk] + 10 x[ice cream] ∈ [0, 65]
+ 730 x[hamburger] + 1190 x[chicken] + 1800 x[hot dog] + 270 x[fries] + 930 x[macaroni] + 820 x[pizza] + 1230 x[salad] + 125 x[milk] + 180 x[ice cream] ∈ [0, 1779]
+ x[hamburger] ≥ 0
+ x[chicken] ≥ 0
+ x[hot dog] ≥ 0
+ x[fries] ≥ 0
+ x[macaroni] ≥ 0
+ x[pizza] ≥ 0
+ x[salad] ≥ 0
+ x[milk] ≥ 0
+ x[ice cream] ≥ 0</code></pre><h2 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h2><p>Let&#39;s optimize and take a look at the solution:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 1.18289e+01
+  Objective bound    : 0.00000e+00
+  Relative gap       : Inf
+  Dual objective value : 1.18289e+01
+
+* Work counters
+  Solve time (sec)   : 2.48909e-04
+  Simplex iterations : 6
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><p>We found an optimal solution. Let&#39;s see what the optimal solution is:</p><pre><code class="language-julia hljs">for row in eachrow(foods)
+    println(row.name, &quot; = &quot;, value(row.x))
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">hamburger = 0.6045138888888871
+chicken = 0.0
+hot dog = 0.0
+fries = 0.0
+macaroni = 0.0
+pizza = 0.0
+salad = 0.0
+milk = 6.9701388888888935
+ice cream = 2.5913194444444447</code></pre><p>That&#39;s a lot of milk and ice cream, and sadly, we only get <code>0.6</code> of a hamburger.</p><p>We can also use the function <a href="../../../api/JuMP.Containers/#JuMP.Containers.rowtable"><code>Containers.rowtable</code></a> to easily convert the result into a DataFrame:</p><pre><code class="language-julia hljs">table = Containers.rowtable(value, x; header = [:food, :quantity])
+solution = DataFrames.DataFrame(table)</code></pre><div><div style = "float: left;"><span>9×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">food</th><th style = "text-align: left;">quantity</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">hamburger</td><td style = "text-align: right;">0.604514</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">chicken</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">hot dog</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">fries</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">macaroni</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">pizza</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">salad</td><td style = "text-align: right;">0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">6.97014</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">ice cream</td><td style = "text-align: right;">2.59132</td></tr></tbody></table></div><p>This makes it easy to perform analyses our solution:</p><pre><code class="language-julia hljs">filter!(row -&gt; row.quantity &gt; 0.0, solution)</code></pre><div><div style = "float: left;"><span>3×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">food</th><th style = "text-align: left;">quantity</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String15" style = "text-align: left;">String15</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">hamburger</td><td style = "text-align: right;">0.604514</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">6.97014</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">ice cream</td><td style = "text-align: right;">2.59132</td></tr></tbody></table></div><h2 id="Problem-modification"><a class="docs-heading-anchor" href="#Problem-modification">Problem modification</a><a id="Problem-modification-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-modification" title="Permalink"></a></h2><p>JuMP makes it easy to take an existing model and modify it by adding extra constraints. Let&#39;s see what happens if we add a constraint that we can buy at most 6 units of milk or ice cream combined.</p><pre><code class="language-julia hljs">dairy_foods = [&quot;milk&quot;, &quot;ice cream&quot;]
+is_dairy = map(name -&gt; name in dairy_foods, foods.name)
+dairy_constraint = @constraint(model, sum(foods[is_dairy, :x]) &lt;= 6)
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : INFEASIBLE
+  Message from the solver:
+  &quot;kHighsModelStatusInfeasible&quot;
+
+* Candidate solution (result #1)
+  Primal status      : NO_SOLUTION
+  Dual status        : INFEASIBILITY_CERTIFICATE
+  Objective value    : 1.18289e+01
+  Objective bound    : 0.00000e+00
+  Relative gap       : Inf
+  Dual objective value : 3.56146e+00
+
+* Work counters
+  Solve time (sec)   : 4.21524e-04
+  Simplex iterations : 0
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><p>There exists no feasible solution to our problem. Looks like we&#39;re stuck eating ice cream for dinner.</p><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><ul><li>You can delete a constraint using <code>delete(model, dairy_constraint)</code>. Can you add a different constraint to provide a diet with less dairy?</li><li>Some food items (like hamburgers) are discrete. You can use <code>set_integer</code> to force a variable to take integer values. What happens to the solution if you do?</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../knapsack/">« The knapsack problem example</a><a class="docs-footer-nextpage" href="../cannery/">The cannery problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/facility_location.jl b/previews/PR3525/tutorials/linear/facility_location.jl
new file mode 100644
index 00000000000..b435f45f0fc
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/facility_location.jl
@@ -0,0 +1,300 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The facility location problem
+
+# **This tutorial was originally contributed by Mathieu Tanneau and Alexis Montoison.**
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import HiGHS
+import LinearAlgebra
+import Plots
+import Random
+
+# ## Uncapacitated facility location
+
+# ### Problem description
+#
+# We are given
+# * A set $M=\{1, \dots, m\}$ of clients
+# * A set $N=\{ 1, \dots, n\}$ of sites where a facility can be built
+#
+# **Decision variables**
+# Decision variables are split into two categories:
+# * Binary variable $y_{j}$ indicates whether facility $j$ is built or not
+# * Binary variable $x_{i, j}$ indicates whether client $i$ is assigned to facility $j$
+#
+# **Objective**
+# The objective is to minimize the total cost of serving all clients.
+# This costs breaks down into two components:
+# * Fixed cost of building a facility.
+# In this example, this cost is $f_{j} = 1, \ \forall j$.
+# * Cost of serving clients from the assigned facility.
+# In this example, the cost $c_{i, j}$
+# of serving client $i$ from facility $j$
+# is the Euclidean distance between the two.
+#
+# **Constraints**
+# * Each customer must be served by exactly one facility
+# * A facility cannot serve any client unless it is open
+
+# ### MILP formulation
+#
+# The problem can be formulated as the following MILP:
+#
+# ```math
+# \begin{aligned}
+# \min_{x, y} \ \ \ &
+# \sum_{i, j} c_{i, j} x_{i, j} +
+# \sum_{j} f_{j} y_{j} \\
+# s.t. &
+# \sum_{j} x_{i, j} = 1, && \forall i \in M \\
+# & x_{i, j} \leq y_{j}, && \forall i \in M, j \in N \\
+# & x_{i, j}, y_{j} \in \{0, 1\}, && \forall i \in M, j \in N
+# \end{aligned}
+# ```
+#
+# where the first set of constraints ensures
+# that each client is served exactly once,
+# and the second set of constraints ensures
+# that no client is served from an unopened facility.
+
+# ### Problem data
+
+# To ensure reproducibility, we set the random number seed:
+
+Random.seed!(314)
+
+# Here's the data we need:
+
+## Number of clients
+m = 12
+## Number of facility locations
+n = 5
+
+## Clients' locations
+x_c, y_c = rand(m), rand(m)
+
+## Facilities' potential locations
+x_f, y_f = rand(n), rand(n)
+
+## Fixed costs
+f = ones(n);
+
+## Distance
+c = zeros(m, n)
+for i in 1:m
+    for j in 1:n
+        c[i, j] = LinearAlgebra.norm([x_c[i] - x_f[j], y_c[i] - y_f[j]], 2)
+    end
+end
+
+# Display the data
+Plots.scatter(
+    x_c,
+    y_c;
+    label = "Clients",
+    markershape = :circle,
+    markercolor = :blue,
+)
+Plots.scatter!(
+    x_f,
+    y_f;
+    label = "Facility",
+    markershape = :square,
+    markercolor = :white,
+    markersize = 6,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+)
+
+# ### JuMP implementation
+
+# Create a JuMP model
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, y[1:n], Bin);
+@variable(model, x[1:m, 1:n], Bin);
+## Each client is served exactly once
+@constraint(model, client_service[i in 1:m], sum(x[i, j] for j in 1:n) == 1);
+## A facility must be open to serve a client
+@constraint(model, open_facility[i in 1:m, j in 1:n], x[i, j] <= y[j]);
+@objective(model, Min, f' * y + sum(c .* x));
+
+# Solve the uncapacitated facility location problem with HiGHS
+
+optimize!(model)
+println("Optimal value: ", objective_value(model))
+
+# ### Visualizing the solution
+
+# The threshold 1e-5 ensure that edges between clients and facilities are drawn
+# when `x[i, j] ≈ 1`.
+x_is_selected = isapprox.(value.(x), 1; atol = 1e-5);
+y_is_selected = isapprox.(value.(y), 1; atol = 1e-5);
+
+p = Plots.scatter(
+    x_c,
+    y_c;
+    markershape = :circle,
+    markercolor = :blue,
+    label = nothing,
+)
+
+Plots.scatter!(
+    x_f,
+    y_f;
+    markershape = :square,
+    markercolor = [(y_is_selected[j] ? :red : :white) for j in 1:n],
+    markersize = 6,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+    label = nothing,
+)
+
+for i in 1:m, j in 1:n
+    if x_is_selected[i, j]
+        Plots.plot!(
+            [x_c[i], x_f[j]],
+            [y_c[i], y_f[j]];
+            color = :black,
+            label = nothing,
+        )
+    end
+end
+
+p
+
+# ## Capacitated facility location
+
+# ### Problem formulation
+#
+# The capacitated variant introduces a capacity constraint on each facility,
+# that is, clients have a certain level of demand to be served, while each facility
+# only has finite capacity which cannot be exceeded.
+#
+# Specifically,
+# * The demand of client $i$ is denoted by $a_{i} \geq 0$
+# * The capacity of facility $j$ is denoted by $q_{j} \geq 0$
+#
+# The capacity constraints then write
+# ```math
+# \begin{aligned}
+# \sum_{i} a_{i} x_{i, j} &\leq q_{j} y_{j} && \forall j \in N
+# \end{aligned}
+# ```
+#
+# Note that, if $y_{j}$ is set to $0$, the capacity constraint above
+# automatically forces $x_{i, j}$ to $0$.
+
+# Thus, the capacitated facility location can be formulated as follows
+#
+# ```math
+# \begin{aligned}
+# \min_{x, y} \ \ \ &
+# \sum_{i, j} c_{i, j} x_{i, j} +
+# \sum_{j} f_{j} y_{j} \\
+# s.t. &
+# \sum_{j} x_{i, j} = 1, && \forall i \in M \\
+# & \sum_{i} a_{i} x_{i, j} \leq q_{j} y_{j}, && \forall j \in N \\
+# & x_{i, j}, y_{j} \in \{0, 1\}, && \forall i \in M, j \in N
+# \end{aligned}
+# ```
+#
+# For simplicity, we will assume that there is enough capacity to serve the
+# demand, that is, there exists at least one feasible solution.
+
+# We need some new data:
+
+## Demands
+a = rand(1:3, m);
+
+## Capacities
+q = rand(5:10, n);
+
+# Display the data
+Plots.scatter(
+    x_c,
+    y_c;
+    label = nothing,
+    markershape = :circle,
+    markercolor = :blue,
+    markersize = 2 .* (2 .+ a),
+)
+
+Plots.scatter!(
+    x_f,
+    y_f;
+    label = nothing,
+    markershape = :rect,
+    markercolor = :white,
+    markersize = q,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+)
+
+# ### JuMP implementation
+
+# Create a JuMP model
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, y[1:n], Bin);
+@variable(model, x[1:m, 1:n], Bin);
+## Each client is served exactly once
+@constraint(model, client_service[i in 1:m], sum(x[i, :]) == 1);
+## Capacity constraint
+@constraint(model, capacity, x' * a .<= (q .* y));
+## Objective
+@objective(model, Min, f' * y + sum(c .* x));
+
+# Solve the problem
+
+optimize!(model)
+println("Optimal value: ", objective_value(model))
+
+# ### Visualizing the solution
+
+# The threshold 1e-5 ensure that edges between clients and facilities are drawn
+# when `x[i, j] ≈ 1`.
+x_is_selected = isapprox.(value.(x), 1; atol = 1e-5);
+y_is_selected = isapprox.(value.(y), 1; atol = 1e-5);
+
+# Display the solution
+p = Plots.scatter(
+    x_c,
+    y_c;
+    label = nothing,
+    markershape = :circle,
+    markercolor = :blue,
+    markersize = 2 .* (2 .+ a),
+)
+
+Plots.scatter!(
+    x_f,
+    y_f;
+    label = nothing,
+    markershape = :rect,
+    markercolor = [(y_is_selected[j] ? :red : :white) for j in 1:n],
+    markersize = q,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+)
+
+for i in 1:m, j in 1:n
+    if x_is_selected[i, j]
+        Plots.plot!(
+            [x_c[i], x_f[j]],
+            [y_c[i], y_f[j]];
+            color = :black,
+            label = nothing,
+        )
+    end
+end
+
+p
diff --git a/previews/PR3525/tutorials/linear/facility_location/49a24bc3.svg b/previews/PR3525/tutorials/linear/facility_location/49a24bc3.svg
new file mode 100644
index 00000000000..f4217b2d697
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/facility_location/49a24bc3.svg
@@ -0,0 +1,69 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip150">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip150)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip151">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip150)" d="M156.598 1486.45 L2352.76 1486.45 L2352.76 47.2441 L156.598 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip152">
+    <rect x="156" y="47" width="2197" height="1440"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="496.242,1486.45 496.242,47.2441 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="953.502,1486.45 953.502,47.2441 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1410.76,1486.45 1410.76,47.2441 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1868.02,1486.45 1868.02,47.2441 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2325.28,1486.45 2325.28,47.2441 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="496.242,1486.45 496.242,1467.55 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="953.502,1486.45 953.502,1467.55 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1410.76,1486.45 1410.76,1467.55 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1868.02,1486.45 1868.02,1467.55 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2325.28,1486.45 2325.28,1467.55 "/>
+<path clip-path="url(#clip150)" d="M474.425 1517.37 Q470.814 1517.37 468.985 1520.93 Q467.18 1524.47 467.18 1531.6 Q467.18 1538.71 468.985 1542.27 Q470.814 1545.82 474.425 1545.82 Q478.059 1545.82 479.865 1542.27 Q481.694 1538.71 481.694 1531.6 Q481.694 1524.47 479.865 1520.93 Q478.059 1517.37 474.425 1517.37 M474.425 1513.66 Q480.235 1513.66 483.291 1518.27 Q486.37 1522.85 486.37 1531.6 Q486.37 1540.33 483.291 1544.94 Q480.235 1549.52 474.425 1549.52 Q468.615 1549.52 465.536 1544.94 Q462.481 1540.33 462.481 1531.6 Q462.481 1522.85 465.536 1518.27 Q468.615 1513.66 474.425 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M494.587 1542.97 L499.471 1542.97 L499.471 1548.85 L494.587 1548.85 L494.587 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M513.684 1544.91 L530.004 1544.91 L530.004 1548.85 L508.059 1548.85 L508.059 1544.91 Q510.721 1542.16 515.305 1537.53 Q519.911 1532.88 521.092 1531.53 Q523.337 1529.01 524.217 1527.27 Q525.119 1525.51 525.119 1523.82 Q525.119 1521.07 523.175 1519.33 Q521.254 1517.6 518.152 1517.6 Q515.953 1517.6 513.499 1518.36 Q511.069 1519.13 508.291 1520.68 L508.291 1515.95 Q511.115 1514.82 513.569 1514.24 Q516.022 1513.66 518.059 1513.66 Q523.43 1513.66 526.624 1516.35 Q529.818 1519.03 529.818 1523.52 Q529.818 1525.65 529.008 1527.57 Q528.221 1529.47 526.115 1532.07 Q525.536 1532.74 522.434 1535.95 Q519.332 1539.15 513.684 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M930.644 1517.37 Q927.032 1517.37 925.204 1520.93 Q923.398 1524.47 923.398 1531.6 Q923.398 1538.71 925.204 1542.27 Q927.032 1545.82 930.644 1545.82 Q934.278 1545.82 936.083 1542.27 Q937.912 1538.71 937.912 1531.6 Q937.912 1524.47 936.083 1520.93 Q934.278 1517.37 930.644 1517.37 M930.644 1513.66 Q936.454 1513.66 939.509 1518.27 Q942.588 1522.85 942.588 1531.6 Q942.588 1540.33 939.509 1544.94 Q936.454 1549.52 930.644 1549.52 Q924.833 1549.52 921.755 1544.94 Q918.699 1540.33 918.699 1531.6 Q918.699 1522.85 921.755 1518.27 Q924.833 1513.66 930.644 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M950.805 1542.97 L955.69 1542.97 L955.69 1548.85 L950.805 1548.85 L950.805 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M978.722 1518.36 L966.917 1536.81 L978.722 1536.81 L978.722 1518.36 M977.495 1514.29 L983.375 1514.29 L983.375 1536.81 L988.305 1536.81 L988.305 1540.7 L983.375 1540.7 L983.375 1548.85 L978.722 1548.85 L978.722 1540.7 L963.12 1540.7 L963.12 1536.19 L977.495 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M1388.07 1517.37 Q1384.45 1517.37 1382.63 1520.93 Q1380.82 1524.47 1380.82 1531.6 Q1380.82 1538.71 1382.63 1542.27 Q1384.45 1545.82 1388.07 1545.82 Q1391.7 1545.82 1393.51 1542.27 Q1395.33 1538.71 1395.33 1531.6 Q1395.33 1524.47 1393.51 1520.93 Q1391.7 1517.37 1388.07 1517.37 M1388.07 1513.66 Q1393.88 1513.66 1396.93 1518.27 Q1400.01 1522.85 1400.01 1531.6 Q1400.01 1540.33 1396.93 1544.94 Q1393.88 1549.52 1388.07 1549.52 Q1382.26 1549.52 1379.18 1544.94 Q1376.12 1540.33 1376.12 1531.6 Q1376.12 1522.85 1379.18 1518.27 Q1382.26 1513.66 1388.07 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M1408.23 1542.97 L1413.11 1542.97 L1413.11 1548.85 L1408.23 1548.85 L1408.23 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M1433.88 1529.7 Q1430.73 1529.7 1428.88 1531.86 Q1427.05 1534.01 1427.05 1537.76 Q1427.05 1541.49 1428.88 1543.66 Q1430.73 1545.82 1433.88 1545.82 Q1437.02 1545.82 1438.85 1543.66 Q1440.7 1541.49 1440.7 1537.76 Q1440.7 1534.01 1438.85 1531.86 Q1437.02 1529.7 1433.88 1529.7 M1443.16 1515.05 L1443.16 1519.31 Q1441.4 1518.48 1439.59 1518.04 Q1437.81 1517.6 1436.05 1517.6 Q1431.42 1517.6 1428.97 1520.72 Q1426.54 1523.85 1426.19 1530.17 Q1427.56 1528.15 1429.62 1527.09 Q1431.68 1526 1434.15 1526 Q1439.36 1526 1442.37 1529.17 Q1445.4 1532.32 1445.4 1537.76 Q1445.4 1543.08 1442.26 1546.3 Q1439.11 1549.52 1433.88 1549.52 Q1427.88 1549.52 1424.71 1544.94 Q1421.54 1540.33 1421.54 1531.6 Q1421.54 1523.41 1425.43 1518.55 Q1429.32 1513.66 1435.87 1513.66 Q1437.63 1513.66 1439.41 1514.01 Q1441.21 1514.36 1443.16 1515.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M1845.45 1517.37 Q1841.84 1517.37 1840.01 1520.93 Q1838.21 1524.47 1838.21 1531.6 Q1838.21 1538.71 1840.01 1542.27 Q1841.84 1545.82 1845.45 1545.82 Q1849.09 1545.82 1850.89 1542.27 Q1852.72 1538.71 1852.72 1531.6 Q1852.72 1524.47 1850.89 1520.93 Q1849.09 1517.37 1845.45 1517.37 M1845.45 1513.66 Q1851.26 1513.66 1854.32 1518.27 Q1857.4 1522.85 1857.4 1531.6 Q1857.4 1540.33 1854.32 1544.94 Q1851.26 1549.52 1845.45 1549.52 Q1839.64 1549.52 1836.56 1544.94 Q1833.51 1540.33 1833.51 1531.6 Q1833.51 1522.85 1836.56 1518.27 Q1839.64 1513.66 1845.45 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M1865.61 1542.97 L1870.5 1542.97 L1870.5 1548.85 L1865.61 1548.85 L1865.61 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M1890.68 1532.44 Q1887.35 1532.44 1885.43 1534.22 Q1883.53 1536 1883.53 1539.13 Q1883.53 1542.25 1885.43 1544.03 Q1887.35 1545.82 1890.68 1545.82 Q1894.02 1545.82 1895.94 1544.03 Q1897.86 1542.23 1897.86 1539.13 Q1897.86 1536 1895.94 1534.22 Q1894.04 1532.44 1890.68 1532.44 M1886.01 1530.45 Q1883 1529.7 1881.31 1527.64 Q1879.64 1525.58 1879.64 1522.62 Q1879.64 1518.48 1882.58 1516.07 Q1885.55 1513.66 1890.68 1513.66 Q1895.85 1513.66 1898.79 1516.07 Q1901.73 1518.48 1901.73 1522.62 Q1901.73 1525.58 1900.04 1527.64 Q1898.37 1529.7 1895.38 1530.45 Q1898.76 1531.23 1900.64 1533.52 Q1902.54 1535.82 1902.54 1539.13 Q1902.54 1544.15 1899.46 1546.83 Q1896.4 1549.52 1890.68 1549.52 Q1884.97 1549.52 1881.89 1546.83 Q1878.83 1544.15 1878.83 1539.13 Q1878.83 1535.82 1880.73 1533.52 Q1882.63 1531.23 1886.01 1530.45 M1884.3 1523.06 Q1884.3 1525.75 1885.96 1527.25 Q1887.65 1528.76 1890.68 1528.76 Q1893.69 1528.76 1895.38 1527.25 Q1897.1 1525.75 1897.1 1523.06 Q1897.1 1520.38 1895.38 1518.87 Q1893.69 1517.37 1890.68 1517.37 Q1887.65 1517.37 1885.96 1518.87 Q1884.3 1520.38 1884.3 1523.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M2292.44 1544.91 L2300.07 1544.91 L2300.07 1518.55 L2291.76 1520.21 L2291.76 1515.95 L2300.03 1514.29 L2304.7 1514.29 L2304.7 1544.91 L2312.34 1544.91 L2312.34 1548.85 L2292.44 1548.85 L2292.44 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M2321.79 1542.97 L2326.67 1542.97 L2326.67 1548.85 L2321.79 1548.85 L2321.79 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M2346.86 1517.37 Q2343.25 1517.37 2341.42 1520.93 Q2339.61 1524.47 2339.61 1531.6 Q2339.61 1538.71 2341.42 1542.27 Q2343.25 1545.82 2346.86 1545.82 Q2350.49 1545.82 2352.3 1542.27 Q2354.12 1538.71 2354.12 1531.6 Q2354.12 1524.47 2352.3 1520.93 Q2350.49 1517.37 2346.86 1517.37 M2346.86 1513.66 Q2352.67 1513.66 2355.72 1518.27 Q2358.8 1522.85 2358.8 1531.6 Q2358.8 1540.33 2355.72 1544.94 Q2352.67 1549.52 2346.86 1549.52 Q2341.05 1549.52 2337.97 1544.94 Q2334.91 1540.33 2334.91 1531.6 Q2334.91 1522.85 2337.97 1518.27 Q2341.05 1513.66 2346.86 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,1203.47 2352.76,1203.47 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,883.412 2352.76,883.412 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,563.356 2352.76,563.356 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,243.3 2352.76,243.3 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 156.598,47.2441 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1203.47 175.496,1203.47 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,883.412 175.496,883.412 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,563.356 175.496,563.356 "/>
+<polyline clip-path="url(#clip150)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,243.3 175.496,243.3 "/>
+<path clip-path="url(#clip150)" d="M65.0198 1189.27 Q61.4087 1189.27 59.58 1192.83 Q57.7745 1196.37 57.7745 1203.5 Q57.7745 1210.61 59.58 1214.17 Q61.4087 1217.72 65.0198 1217.72 Q68.6541 1217.72 70.4596 1214.17 Q72.2883 1210.61 72.2883 1203.5 Q72.2883 1196.37 70.4596 1192.83 Q68.6541 1189.27 65.0198 1189.27 M65.0198 1185.56 Q70.83 1185.56 73.8855 1190.17 Q76.9642 1194.75 76.9642 1203.5 Q76.9642 1212.23 73.8855 1216.84 Q70.83 1221.42 65.0198 1221.42 Q59.2097 1221.42 56.131 1216.84 Q53.0754 1212.23 53.0754 1203.5 Q53.0754 1194.75 56.131 1190.17 Q59.2097 1185.56 65.0198 1185.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M85.1818 1214.87 L90.066 1214.87 L90.066 1220.75 L85.1818 1220.75 L85.1818 1214.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M104.279 1216.81 L120.598 1216.81 L120.598 1220.75 L98.6539 1220.75 L98.6539 1216.81 Q101.316 1214.06 105.899 1209.43 Q110.506 1204.78 111.686 1203.43 Q113.932 1200.91 114.811 1199.17 Q115.714 1197.41 115.714 1195.73 Q115.714 1192.97 113.77 1191.23 Q111.848 1189.5 108.746 1189.5 Q106.547 1189.5 104.094 1190.26 Q101.663 1191.03 98.8854 1192.58 L98.8854 1187.85 Q101.709 1186.72 104.163 1186.14 Q106.617 1185.56 108.654 1185.56 Q114.024 1185.56 117.219 1188.25 Q120.413 1190.93 120.413 1195.42 Q120.413 1197.55 119.603 1199.48 Q118.816 1201.37 116.709 1203.97 Q116.131 1204.64 113.029 1207.85 Q109.927 1211.05 104.279 1216.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M62.9365 869.211 Q59.3254 869.211 57.4967 872.775 Q55.6912 876.317 55.6912 883.447 Q55.6912 890.553 57.4967 894.118 Q59.3254 897.66 62.9365 897.66 Q66.5707 897.66 68.3763 894.118 Q70.205 890.553 70.205 883.447 Q70.205 876.317 68.3763 872.775 Q66.5707 869.211 62.9365 869.211 M62.9365 865.507 Q68.7467 865.507 71.8022 870.113 Q74.8809 874.697 74.8809 883.447 Q74.8809 892.174 71.8022 896.78 Q68.7467 901.363 62.9365 901.363 Q57.1264 901.363 54.0477 896.78 Q50.9921 892.174 50.9921 883.447 Q50.9921 874.697 54.0477 870.113 Q57.1264 865.507 62.9365 865.507 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M83.0984 894.812 L87.9827 894.812 L87.9827 900.692 L83.0984 900.692 L83.0984 894.812 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M111.015 870.206 L99.2095 888.655 L111.015 888.655 L111.015 870.206 M109.788 866.132 L115.668 866.132 L115.668 888.655 L120.598 888.655 L120.598 892.544 L115.668 892.544 L115.668 900.692 L111.015 900.692 L111.015 892.544 L95.4132 892.544 L95.4132 888.03 L109.788 866.132 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M63.2606 549.155 Q59.6495 549.155 57.8208 552.719 Q56.0152 556.261 56.0152 563.391 Q56.0152 570.497 57.8208 574.062 Q59.6495 577.604 63.2606 577.604 Q66.8948 577.604 68.7004 574.062 Q70.5291 570.497 70.5291 563.391 Q70.5291 556.261 68.7004 552.719 Q66.8948 549.155 63.2606 549.155 M63.2606 545.451 Q69.0707 545.451 72.1263 550.057 Q75.205 554.641 75.205 563.391 Q75.205 572.117 72.1263 576.724 Q69.0707 581.307 63.2606 581.307 Q57.4504 581.307 54.3717 576.724 Q51.3162 572.117 51.3162 563.391 Q51.3162 554.641 54.3717 550.057 Q57.4504 545.451 63.2606 545.451 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M83.4225 574.756 L88.3067 574.756 L88.3067 580.636 L83.4225 580.636 L83.4225 574.756 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M109.071 561.493 Q105.922 561.493 104.071 563.645 Q102.242 565.798 102.242 569.548 Q102.242 573.275 104.071 575.451 Q105.922 577.604 109.071 577.604 Q112.219 577.604 114.047 575.451 Q115.899 573.275 115.899 569.548 Q115.899 565.798 114.047 563.645 Q112.219 561.493 109.071 561.493 M118.353 546.84 L118.353 551.099 Q116.594 550.266 114.788 549.826 Q113.006 549.386 111.246 549.386 Q106.617 549.386 104.163 552.511 Q101.733 555.636 101.385 561.955 Q102.751 559.942 104.811 558.877 Q106.871 557.789 109.348 557.789 Q114.557 557.789 117.566 560.96 Q120.598 564.108 120.598 569.548 Q120.598 574.872 117.45 578.09 Q114.302 581.307 109.071 581.307 Q103.075 581.307 99.9039 576.724 Q96.7326 572.117 96.7326 563.391 Q96.7326 555.196 100.621 550.335 Q104.51 545.451 111.061 545.451 Q112.82 545.451 114.603 545.798 Q116.408 546.145 118.353 546.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M63.5152 229.099 Q59.9041 229.099 58.0754 232.663 Q56.2699 236.205 56.2699 243.335 Q56.2699 250.441 58.0754 254.006 Q59.9041 257.547 63.5152 257.547 Q67.1494 257.547 68.955 254.006 Q70.7837 250.441 70.7837 243.335 Q70.7837 236.205 68.955 232.663 Q67.1494 229.099 63.5152 229.099 M63.5152 225.395 Q69.3254 225.395 72.3809 230.001 Q75.4596 234.585 75.4596 243.335 Q75.4596 252.061 72.3809 256.668 Q69.3254 261.251 63.5152 261.251 Q57.7051 261.251 54.6264 256.668 Q51.5708 252.061 51.5708 243.335 Q51.5708 234.585 54.6264 230.001 Q57.7051 225.395 63.5152 225.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M83.6771 254.7 L88.5614 254.7 L88.5614 260.58 L83.6771 260.58 L83.6771 254.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip150)" d="M108.746 244.168 Q105.413 244.168 103.492 245.95 Q101.594 247.733 101.594 250.858 Q101.594 253.983 103.492 255.765 Q105.413 257.547 108.746 257.547 Q112.08 257.547 114.001 255.765 Q115.922 253.96 115.922 250.858 Q115.922 247.733 114.001 245.95 Q112.103 244.168 108.746 244.168 M104.071 242.177 Q101.061 241.436 99.3715 239.376 Q97.7048 237.316 97.7048 234.353 Q97.7048 230.21 100.645 227.802 Q103.608 225.395 108.746 225.395 Q113.908 225.395 116.848 227.802 Q119.788 230.21 119.788 234.353 Q119.788 237.316 118.098 239.376 Q116.432 241.436 113.445 242.177 Q116.825 242.964 118.7 245.256 Q120.598 247.548 120.598 250.858 Q120.598 255.881 117.52 258.566 Q114.464 261.251 108.746 261.251 Q103.029 261.251 99.9502 258.566 Q96.8947 255.881 96.8947 250.858 Q96.8947 247.548 98.7928 245.256 Q100.691 242.964 104.071 242.177 M102.358 234.793 Q102.358 237.478 104.024 238.983 Q105.714 240.487 108.746 240.487 Q111.756 240.487 113.445 238.983 Q115.158 237.478 115.158 234.793 Q115.158 232.108 113.445 230.603 Q111.756 229.099 108.746 229.099 Q105.714 229.099 104.024 230.603 Q102.358 232.108 102.358 234.793 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip152)" cx="900.754" cy="117.213" r="21.6" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="1337.66" cy="998.336" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="2288.96" cy="1030.31" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="1840.64" cy="87.9763" r="21.6" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="218.754" cy="1410.69" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="768.03" cy="1026.75" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="2017.02" cy="967.646" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="2283.71" cy="1087.25" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="1955.74" cy="1445.72" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="257.146" cy="888.562" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="1413.3" cy="296.531" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip152)" cx="611.082" cy="118.991" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<path clip-path="url(#clip152)" d="M602.682 154.046 L602.682 226.046 L674.682 226.046 L674.682 154.046 L602.682 154.046 Z" fill="#ff0000" fill-rule="evenodd" fill-opacity="1" stroke="none"/>
+<path clip-path="url(#clip152)" d="M1946.4 748.137 L1946.4 820.137 L2018.4 820.137 L2018.4 748.137 L1946.4 748.137 Z" fill="#ff0000" fill-rule="evenodd" fill-opacity="1" stroke="none"/>
+<path clip-path="url(#clip152)" d="M998.777 723.237 L998.777 763.237 L1038.78 763.237 L1038.78 723.237 L998.777 723.237 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip152)" d="M798.323 1246.75 L798.323 1310.75 L862.323 1310.75 L862.323 1246.75 L798.323 1246.75 Z" fill="#ff0000" fill-rule="evenodd" fill-opacity="1" stroke="none"/>
+<path clip-path="url(#clip152)" d="M2262.6 872.347 L2262.6 928.347 L2318.6 928.347 L2318.6 872.347 L2262.6 872.347 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="900.754,117.213 638.682,190.046 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1337.66,998.336 830.323,1278.75 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2288.96,1030.31 1982.4,784.137 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1840.64,87.9763 638.682,190.046 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="218.754,1410.69 830.323,1278.75 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="768.03,1026.75 830.323,1278.75 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2017.02,967.646 1982.4,784.137 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2283.71,1087.25 1982.4,784.137 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1955.74,1445.72 1982.4,784.137 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="257.146,888.562 638.682,190.046 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1413.3,296.531 638.682,190.046 "/>
+<polyline clip-path="url(#clip152)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="611.082,118.991 638.682,190.046 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/linear/facility_location/a3695eca.svg b/previews/PR3525/tutorials/linear/facility_location/a3695eca.svg
new file mode 100644
index 00000000000..1338a4a896f
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/facility_location/a3695eca.svg
@@ -0,0 +1,69 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip030">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip030)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip031">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip030)" d="M156.598 1486.45 L2352.76 1486.45 L2352.76 47.2441 L156.598 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip032">
+    <rect x="156" y="47" width="2197" height="1440"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="496.242,1486.45 496.242,47.2441 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="953.502,1486.45 953.502,47.2441 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1410.76,1486.45 1410.76,47.2441 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1868.02,1486.45 1868.02,47.2441 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2325.28,1486.45 2325.28,47.2441 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="496.242,1486.45 496.242,1467.55 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="953.502,1486.45 953.502,1467.55 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1410.76,1486.45 1410.76,1467.55 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1868.02,1486.45 1868.02,1467.55 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2325.28,1486.45 2325.28,1467.55 "/>
+<path clip-path="url(#clip030)" d="M474.425 1517.37 Q470.814 1517.37 468.985 1520.93 Q467.18 1524.47 467.18 1531.6 Q467.18 1538.71 468.985 1542.27 Q470.814 1545.82 474.425 1545.82 Q478.059 1545.82 479.865 1542.27 Q481.694 1538.71 481.694 1531.6 Q481.694 1524.47 479.865 1520.93 Q478.059 1517.37 474.425 1517.37 M474.425 1513.66 Q480.235 1513.66 483.291 1518.27 Q486.37 1522.85 486.37 1531.6 Q486.37 1540.33 483.291 1544.94 Q480.235 1549.52 474.425 1549.52 Q468.615 1549.52 465.536 1544.94 Q462.481 1540.33 462.481 1531.6 Q462.481 1522.85 465.536 1518.27 Q468.615 1513.66 474.425 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M494.587 1542.97 L499.471 1542.97 L499.471 1548.85 L494.587 1548.85 L494.587 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M513.684 1544.91 L530.004 1544.91 L530.004 1548.85 L508.059 1548.85 L508.059 1544.91 Q510.721 1542.16 515.305 1537.53 Q519.911 1532.88 521.092 1531.53 Q523.337 1529.01 524.217 1527.27 Q525.119 1525.51 525.119 1523.82 Q525.119 1521.07 523.175 1519.33 Q521.254 1517.6 518.152 1517.6 Q515.953 1517.6 513.499 1518.36 Q511.069 1519.13 508.291 1520.68 L508.291 1515.95 Q511.115 1514.82 513.569 1514.24 Q516.022 1513.66 518.059 1513.66 Q523.43 1513.66 526.624 1516.35 Q529.818 1519.03 529.818 1523.52 Q529.818 1525.65 529.008 1527.57 Q528.221 1529.47 526.115 1532.07 Q525.536 1532.74 522.434 1535.95 Q519.332 1539.15 513.684 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M930.644 1517.37 Q927.032 1517.37 925.204 1520.93 Q923.398 1524.47 923.398 1531.6 Q923.398 1538.71 925.204 1542.27 Q927.032 1545.82 930.644 1545.82 Q934.278 1545.82 936.083 1542.27 Q937.912 1538.71 937.912 1531.6 Q937.912 1524.47 936.083 1520.93 Q934.278 1517.37 930.644 1517.37 M930.644 1513.66 Q936.454 1513.66 939.509 1518.27 Q942.588 1522.85 942.588 1531.6 Q942.588 1540.33 939.509 1544.94 Q936.454 1549.52 930.644 1549.52 Q924.833 1549.52 921.755 1544.94 Q918.699 1540.33 918.699 1531.6 Q918.699 1522.85 921.755 1518.27 Q924.833 1513.66 930.644 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M950.805 1542.97 L955.69 1542.97 L955.69 1548.85 L950.805 1548.85 L950.805 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M978.722 1518.36 L966.917 1536.81 L978.722 1536.81 L978.722 1518.36 M977.495 1514.29 L983.375 1514.29 L983.375 1536.81 L988.305 1536.81 L988.305 1540.7 L983.375 1540.7 L983.375 1548.85 L978.722 1548.85 L978.722 1540.7 L963.12 1540.7 L963.12 1536.19 L977.495 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M1388.07 1517.37 Q1384.45 1517.37 1382.63 1520.93 Q1380.82 1524.47 1380.82 1531.6 Q1380.82 1538.71 1382.63 1542.27 Q1384.45 1545.82 1388.07 1545.82 Q1391.7 1545.82 1393.51 1542.27 Q1395.33 1538.71 1395.33 1531.6 Q1395.33 1524.47 1393.51 1520.93 Q1391.7 1517.37 1388.07 1517.37 M1388.07 1513.66 Q1393.88 1513.66 1396.93 1518.27 Q1400.01 1522.85 1400.01 1531.6 Q1400.01 1540.33 1396.93 1544.94 Q1393.88 1549.52 1388.07 1549.52 Q1382.26 1549.52 1379.18 1544.94 Q1376.12 1540.33 1376.12 1531.6 Q1376.12 1522.85 1379.18 1518.27 Q1382.26 1513.66 1388.07 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M1408.23 1542.97 L1413.11 1542.97 L1413.11 1548.85 L1408.23 1548.85 L1408.23 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M1433.88 1529.7 Q1430.73 1529.7 1428.88 1531.86 Q1427.05 1534.01 1427.05 1537.76 Q1427.05 1541.49 1428.88 1543.66 Q1430.73 1545.82 1433.88 1545.82 Q1437.02 1545.82 1438.85 1543.66 Q1440.7 1541.49 1440.7 1537.76 Q1440.7 1534.01 1438.85 1531.86 Q1437.02 1529.7 1433.88 1529.7 M1443.16 1515.05 L1443.16 1519.31 Q1441.4 1518.48 1439.59 1518.04 Q1437.81 1517.6 1436.05 1517.6 Q1431.42 1517.6 1428.97 1520.72 Q1426.54 1523.85 1426.19 1530.17 Q1427.56 1528.15 1429.62 1527.09 Q1431.68 1526 1434.15 1526 Q1439.36 1526 1442.37 1529.17 Q1445.4 1532.32 1445.4 1537.76 Q1445.4 1543.08 1442.26 1546.3 Q1439.11 1549.52 1433.88 1549.52 Q1427.88 1549.52 1424.71 1544.94 Q1421.54 1540.33 1421.54 1531.6 Q1421.54 1523.41 1425.43 1518.55 Q1429.32 1513.66 1435.87 1513.66 Q1437.63 1513.66 1439.41 1514.01 Q1441.21 1514.36 1443.16 1515.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M1845.45 1517.37 Q1841.84 1517.37 1840.01 1520.93 Q1838.21 1524.47 1838.21 1531.6 Q1838.21 1538.71 1840.01 1542.27 Q1841.84 1545.82 1845.45 1545.82 Q1849.09 1545.82 1850.89 1542.27 Q1852.72 1538.71 1852.72 1531.6 Q1852.72 1524.47 1850.89 1520.93 Q1849.09 1517.37 1845.45 1517.37 M1845.45 1513.66 Q1851.26 1513.66 1854.32 1518.27 Q1857.4 1522.85 1857.4 1531.6 Q1857.4 1540.33 1854.32 1544.94 Q1851.26 1549.52 1845.45 1549.52 Q1839.64 1549.52 1836.56 1544.94 Q1833.51 1540.33 1833.51 1531.6 Q1833.51 1522.85 1836.56 1518.27 Q1839.64 1513.66 1845.45 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M1865.61 1542.97 L1870.5 1542.97 L1870.5 1548.85 L1865.61 1548.85 L1865.61 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M1890.68 1532.44 Q1887.35 1532.44 1885.43 1534.22 Q1883.53 1536 1883.53 1539.13 Q1883.53 1542.25 1885.43 1544.03 Q1887.35 1545.82 1890.68 1545.82 Q1894.02 1545.82 1895.94 1544.03 Q1897.86 1542.23 1897.86 1539.13 Q1897.86 1536 1895.94 1534.22 Q1894.04 1532.44 1890.68 1532.44 M1886.01 1530.45 Q1883 1529.7 1881.31 1527.64 Q1879.64 1525.58 1879.64 1522.62 Q1879.64 1518.48 1882.58 1516.07 Q1885.55 1513.66 1890.68 1513.66 Q1895.85 1513.66 1898.79 1516.07 Q1901.73 1518.48 1901.73 1522.62 Q1901.73 1525.58 1900.04 1527.64 Q1898.37 1529.7 1895.38 1530.45 Q1898.76 1531.23 1900.64 1533.52 Q1902.54 1535.82 1902.54 1539.13 Q1902.54 1544.15 1899.46 1546.83 Q1896.4 1549.52 1890.68 1549.52 Q1884.97 1549.52 1881.89 1546.83 Q1878.83 1544.15 1878.83 1539.13 Q1878.83 1535.82 1880.73 1533.52 Q1882.63 1531.23 1886.01 1530.45 M1884.3 1523.06 Q1884.3 1525.75 1885.96 1527.25 Q1887.65 1528.76 1890.68 1528.76 Q1893.69 1528.76 1895.38 1527.25 Q1897.1 1525.75 1897.1 1523.06 Q1897.1 1520.38 1895.38 1518.87 Q1893.69 1517.37 1890.68 1517.37 Q1887.65 1517.37 1885.96 1518.87 Q1884.3 1520.38 1884.3 1523.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M2292.44 1544.91 L2300.07 1544.91 L2300.07 1518.55 L2291.76 1520.21 L2291.76 1515.95 L2300.03 1514.29 L2304.7 1514.29 L2304.7 1544.91 L2312.34 1544.91 L2312.34 1548.85 L2292.44 1548.85 L2292.44 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M2321.79 1542.97 L2326.67 1542.97 L2326.67 1548.85 L2321.79 1548.85 L2321.79 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M2346.86 1517.37 Q2343.25 1517.37 2341.42 1520.93 Q2339.61 1524.47 2339.61 1531.6 Q2339.61 1538.71 2341.42 1542.27 Q2343.25 1545.82 2346.86 1545.82 Q2350.49 1545.82 2352.3 1542.27 Q2354.12 1538.71 2354.12 1531.6 Q2354.12 1524.47 2352.3 1520.93 Q2350.49 1517.37 2346.86 1517.37 M2346.86 1513.66 Q2352.67 1513.66 2355.72 1518.27 Q2358.8 1522.85 2358.8 1531.6 Q2358.8 1540.33 2355.72 1544.94 Q2352.67 1549.52 2346.86 1549.52 Q2341.05 1549.52 2337.97 1544.94 Q2334.91 1540.33 2334.91 1531.6 Q2334.91 1522.85 2337.97 1518.27 Q2341.05 1513.66 2346.86 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,1203.47 2352.76,1203.47 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,883.412 2352.76,883.412 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,563.356 2352.76,563.356 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,243.3 2352.76,243.3 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 156.598,47.2441 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1203.47 175.496,1203.47 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,883.412 175.496,883.412 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,563.356 175.496,563.356 "/>
+<polyline clip-path="url(#clip030)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,243.3 175.496,243.3 "/>
+<path clip-path="url(#clip030)" d="M65.0198 1189.27 Q61.4087 1189.27 59.58 1192.83 Q57.7745 1196.37 57.7745 1203.5 Q57.7745 1210.61 59.58 1214.17 Q61.4087 1217.72 65.0198 1217.72 Q68.6541 1217.72 70.4596 1214.17 Q72.2883 1210.61 72.2883 1203.5 Q72.2883 1196.37 70.4596 1192.83 Q68.6541 1189.27 65.0198 1189.27 M65.0198 1185.56 Q70.83 1185.56 73.8855 1190.17 Q76.9642 1194.75 76.9642 1203.5 Q76.9642 1212.23 73.8855 1216.84 Q70.83 1221.42 65.0198 1221.42 Q59.2097 1221.42 56.131 1216.84 Q53.0754 1212.23 53.0754 1203.5 Q53.0754 1194.75 56.131 1190.17 Q59.2097 1185.56 65.0198 1185.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M85.1818 1214.87 L90.066 1214.87 L90.066 1220.75 L85.1818 1220.75 L85.1818 1214.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M104.279 1216.81 L120.598 1216.81 L120.598 1220.75 L98.6539 1220.75 L98.6539 1216.81 Q101.316 1214.06 105.899 1209.43 Q110.506 1204.78 111.686 1203.43 Q113.932 1200.91 114.811 1199.17 Q115.714 1197.41 115.714 1195.73 Q115.714 1192.97 113.77 1191.23 Q111.848 1189.5 108.746 1189.5 Q106.547 1189.5 104.094 1190.26 Q101.663 1191.03 98.8854 1192.58 L98.8854 1187.85 Q101.709 1186.72 104.163 1186.14 Q106.617 1185.56 108.654 1185.56 Q114.024 1185.56 117.219 1188.25 Q120.413 1190.93 120.413 1195.42 Q120.413 1197.55 119.603 1199.48 Q118.816 1201.37 116.709 1203.97 Q116.131 1204.64 113.029 1207.85 Q109.927 1211.05 104.279 1216.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M62.9365 869.211 Q59.3254 869.211 57.4967 872.775 Q55.6912 876.317 55.6912 883.447 Q55.6912 890.553 57.4967 894.118 Q59.3254 897.66 62.9365 897.66 Q66.5707 897.66 68.3763 894.118 Q70.205 890.553 70.205 883.447 Q70.205 876.317 68.3763 872.775 Q66.5707 869.211 62.9365 869.211 M62.9365 865.507 Q68.7467 865.507 71.8022 870.113 Q74.8809 874.697 74.8809 883.447 Q74.8809 892.174 71.8022 896.78 Q68.7467 901.363 62.9365 901.363 Q57.1264 901.363 54.0477 896.78 Q50.9921 892.174 50.9921 883.447 Q50.9921 874.697 54.0477 870.113 Q57.1264 865.507 62.9365 865.507 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M83.0984 894.812 L87.9827 894.812 L87.9827 900.692 L83.0984 900.692 L83.0984 894.812 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M111.015 870.206 L99.2095 888.655 L111.015 888.655 L111.015 870.206 M109.788 866.132 L115.668 866.132 L115.668 888.655 L120.598 888.655 L120.598 892.544 L115.668 892.544 L115.668 900.692 L111.015 900.692 L111.015 892.544 L95.4132 892.544 L95.4132 888.03 L109.788 866.132 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M63.2606 549.155 Q59.6495 549.155 57.8208 552.719 Q56.0152 556.261 56.0152 563.391 Q56.0152 570.497 57.8208 574.062 Q59.6495 577.604 63.2606 577.604 Q66.8948 577.604 68.7004 574.062 Q70.5291 570.497 70.5291 563.391 Q70.5291 556.261 68.7004 552.719 Q66.8948 549.155 63.2606 549.155 M63.2606 545.451 Q69.0707 545.451 72.1263 550.057 Q75.205 554.641 75.205 563.391 Q75.205 572.117 72.1263 576.724 Q69.0707 581.307 63.2606 581.307 Q57.4504 581.307 54.3717 576.724 Q51.3162 572.117 51.3162 563.391 Q51.3162 554.641 54.3717 550.057 Q57.4504 545.451 63.2606 545.451 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M83.4225 574.756 L88.3067 574.756 L88.3067 580.636 L83.4225 580.636 L83.4225 574.756 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M109.071 561.493 Q105.922 561.493 104.071 563.645 Q102.242 565.798 102.242 569.548 Q102.242 573.275 104.071 575.451 Q105.922 577.604 109.071 577.604 Q112.219 577.604 114.047 575.451 Q115.899 573.275 115.899 569.548 Q115.899 565.798 114.047 563.645 Q112.219 561.493 109.071 561.493 M118.353 546.84 L118.353 551.099 Q116.594 550.266 114.788 549.826 Q113.006 549.386 111.246 549.386 Q106.617 549.386 104.163 552.511 Q101.733 555.636 101.385 561.955 Q102.751 559.942 104.811 558.877 Q106.871 557.789 109.348 557.789 Q114.557 557.789 117.566 560.96 Q120.598 564.108 120.598 569.548 Q120.598 574.872 117.45 578.09 Q114.302 581.307 109.071 581.307 Q103.075 581.307 99.9039 576.724 Q96.7326 572.117 96.7326 563.391 Q96.7326 555.196 100.621 550.335 Q104.51 545.451 111.061 545.451 Q112.82 545.451 114.603 545.798 Q116.408 546.145 118.353 546.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M63.5152 229.099 Q59.9041 229.099 58.0754 232.663 Q56.2699 236.205 56.2699 243.335 Q56.2699 250.441 58.0754 254.006 Q59.9041 257.547 63.5152 257.547 Q67.1494 257.547 68.955 254.006 Q70.7837 250.441 70.7837 243.335 Q70.7837 236.205 68.955 232.663 Q67.1494 229.099 63.5152 229.099 M63.5152 225.395 Q69.3254 225.395 72.3809 230.001 Q75.4596 234.585 75.4596 243.335 Q75.4596 252.061 72.3809 256.668 Q69.3254 261.251 63.5152 261.251 Q57.7051 261.251 54.6264 256.668 Q51.5708 252.061 51.5708 243.335 Q51.5708 234.585 54.6264 230.001 Q57.7051 225.395 63.5152 225.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M83.6771 254.7 L88.5614 254.7 L88.5614 260.58 L83.6771 260.58 L83.6771 254.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip030)" d="M108.746 244.168 Q105.413 244.168 103.492 245.95 Q101.594 247.733 101.594 250.858 Q101.594 253.983 103.492 255.765 Q105.413 257.547 108.746 257.547 Q112.08 257.547 114.001 255.765 Q115.922 253.96 115.922 250.858 Q115.922 247.733 114.001 245.95 Q112.103 244.168 108.746 244.168 M104.071 242.177 Q101.061 241.436 99.3715 239.376 Q97.7048 237.316 97.7048 234.353 Q97.7048 230.21 100.645 227.802 Q103.608 225.395 108.746 225.395 Q113.908 225.395 116.848 227.802 Q119.788 230.21 119.788 234.353 Q119.788 237.316 118.098 239.376 Q116.432 241.436 113.445 242.177 Q116.825 242.964 118.7 245.256 Q120.598 247.548 120.598 250.858 Q120.598 255.881 117.52 258.566 Q114.464 261.251 108.746 261.251 Q103.029 261.251 99.9502 258.566 Q96.8947 255.881 96.8947 250.858 Q96.8947 247.548 98.7928 245.256 Q100.691 242.964 104.071 242.177 M102.358 234.793 Q102.358 237.478 104.024 238.983 Q105.714 240.487 108.746 240.487 Q111.756 240.487 113.445 238.983 Q115.158 237.478 115.158 234.793 Q115.158 232.108 113.445 230.603 Q111.756 229.099 108.746 229.099 Q105.714 229.099 104.024 230.603 Q102.358 232.108 102.358 234.793 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip032)" cx="900.754" cy="117.213" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="1337.66" cy="998.336" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="2288.96" cy="1030.31" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="1840.64" cy="87.9763" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="218.754" cy="1410.69" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="768.03" cy="1026.75" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="2017.02" cy="967.646" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="2283.71" cy="1087.25" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="1955.74" cy="1445.72" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="257.146" cy="888.562" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="1413.3" cy="296.531" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip032)" cx="611.082" cy="118.991" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<path clip-path="url(#clip032)" d="M614.682 166.046 L614.682 214.046 L662.682 214.046 L662.682 166.046 L614.682 166.046 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip032)" d="M1958.4 760.137 L1958.4 808.137 L2006.4 808.137 L2006.4 760.137 L1958.4 760.137 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip032)" d="M994.777 719.237 L994.777 767.237 L1042.78 767.237 L1042.78 719.237 L994.777 719.237 Z" fill="#ff0000" fill-rule="evenodd" fill-opacity="1" stroke="none"/>
+<path clip-path="url(#clip032)" d="M806.323 1254.75 L806.323 1302.75 L854.323 1302.75 L854.323 1254.75 L806.323 1254.75 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip032)" d="M2266.6 876.347 L2266.6 924.347 L2314.6 924.347 L2314.6 876.347 L2266.6 876.347 Z" fill="#ff0000" fill-rule="evenodd" fill-opacity="1" stroke="none"/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="900.754,117.213 1018.78,743.237 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1337.66,998.336 1018.78,743.237 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2288.96,1030.31 2290.6,900.347 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1840.64,87.9763 2290.6,900.347 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="218.754,1410.69 1018.78,743.237 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="768.03,1026.75 1018.78,743.237 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2017.02,967.646 2290.6,900.347 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2283.71,1087.25 2290.6,900.347 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1955.74,1445.72 2290.6,900.347 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="257.146,888.562 1018.78,743.237 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1413.3,296.531 1018.78,743.237 "/>
+<polyline clip-path="url(#clip032)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="611.082,118.991 1018.78,743.237 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/linear/facility_location/a9295f9b.svg b/previews/PR3525/tutorials/linear/facility_location/a9295f9b.svg
new file mode 100644
index 00000000000..2d2f5501ac3
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/facility_location/a9295f9b.svg
@@ -0,0 +1,57 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip090">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip090)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip091">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip090)" d="M156.598 1486.45 L2352.76 1486.45 L2352.76 47.2441 L156.598 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip092">
+    <rect x="156" y="47" width="2197" height="1440"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="496.242,1486.45 496.242,47.2441 "/>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="953.502,1486.45 953.502,47.2441 "/>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1410.76,1486.45 1410.76,47.2441 "/>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1868.02,1486.45 1868.02,47.2441 "/>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2325.28,1486.45 2325.28,47.2441 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="496.242,1486.45 496.242,1467.55 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="953.502,1486.45 953.502,1467.55 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1410.76,1486.45 1410.76,1467.55 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1868.02,1486.45 1868.02,1467.55 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2325.28,1486.45 2325.28,1467.55 "/>
+<path clip-path="url(#clip090)" d="M474.425 1517.37 Q470.814 1517.37 468.985 1520.93 Q467.18 1524.47 467.18 1531.6 Q467.18 1538.71 468.985 1542.27 Q470.814 1545.82 474.425 1545.82 Q478.059 1545.82 479.865 1542.27 Q481.694 1538.71 481.694 1531.6 Q481.694 1524.47 479.865 1520.93 Q478.059 1517.37 474.425 1517.37 M474.425 1513.66 Q480.235 1513.66 483.291 1518.27 Q486.37 1522.85 486.37 1531.6 Q486.37 1540.33 483.291 1544.94 Q480.235 1549.52 474.425 1549.52 Q468.615 1549.52 465.536 1544.94 Q462.481 1540.33 462.481 1531.6 Q462.481 1522.85 465.536 1518.27 Q468.615 1513.66 474.425 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M494.587 1542.97 L499.471 1542.97 L499.471 1548.85 L494.587 1548.85 L494.587 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M513.684 1544.91 L530.004 1544.91 L530.004 1548.85 L508.059 1548.85 L508.059 1544.91 Q510.721 1542.16 515.305 1537.53 Q519.911 1532.88 521.092 1531.53 Q523.337 1529.01 524.217 1527.27 Q525.119 1525.51 525.119 1523.82 Q525.119 1521.07 523.175 1519.33 Q521.254 1517.6 518.152 1517.6 Q515.953 1517.6 513.499 1518.36 Q511.069 1519.13 508.291 1520.68 L508.291 1515.95 Q511.115 1514.82 513.569 1514.24 Q516.022 1513.66 518.059 1513.66 Q523.43 1513.66 526.624 1516.35 Q529.818 1519.03 529.818 1523.52 Q529.818 1525.65 529.008 1527.57 Q528.221 1529.47 526.115 1532.07 Q525.536 1532.74 522.434 1535.95 Q519.332 1539.15 513.684 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M930.644 1517.37 Q927.032 1517.37 925.204 1520.93 Q923.398 1524.47 923.398 1531.6 Q923.398 1538.71 925.204 1542.27 Q927.032 1545.82 930.644 1545.82 Q934.278 1545.82 936.083 1542.27 Q937.912 1538.71 937.912 1531.6 Q937.912 1524.47 936.083 1520.93 Q934.278 1517.37 930.644 1517.37 M930.644 1513.66 Q936.454 1513.66 939.509 1518.27 Q942.588 1522.85 942.588 1531.6 Q942.588 1540.33 939.509 1544.94 Q936.454 1549.52 930.644 1549.52 Q924.833 1549.52 921.755 1544.94 Q918.699 1540.33 918.699 1531.6 Q918.699 1522.85 921.755 1518.27 Q924.833 1513.66 930.644 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M950.805 1542.97 L955.69 1542.97 L955.69 1548.85 L950.805 1548.85 L950.805 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M978.722 1518.36 L966.917 1536.81 L978.722 1536.81 L978.722 1518.36 M977.495 1514.29 L983.375 1514.29 L983.375 1536.81 L988.305 1536.81 L988.305 1540.7 L983.375 1540.7 L983.375 1548.85 L978.722 1548.85 L978.722 1540.7 L963.12 1540.7 L963.12 1536.19 L977.495 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M1388.07 1517.37 Q1384.45 1517.37 1382.63 1520.93 Q1380.82 1524.47 1380.82 1531.6 Q1380.82 1538.71 1382.63 1542.27 Q1384.45 1545.82 1388.07 1545.82 Q1391.7 1545.82 1393.51 1542.27 Q1395.33 1538.71 1395.33 1531.6 Q1395.33 1524.47 1393.51 1520.93 Q1391.7 1517.37 1388.07 1517.37 M1388.07 1513.66 Q1393.88 1513.66 1396.93 1518.27 Q1400.01 1522.85 1400.01 1531.6 Q1400.01 1540.33 1396.93 1544.94 Q1393.88 1549.52 1388.07 1549.52 Q1382.26 1549.52 1379.18 1544.94 Q1376.12 1540.33 1376.12 1531.6 Q1376.12 1522.85 1379.18 1518.27 Q1382.26 1513.66 1388.07 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M1408.23 1542.97 L1413.11 1542.97 L1413.11 1548.85 L1408.23 1548.85 L1408.23 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M1433.88 1529.7 Q1430.73 1529.7 1428.88 1531.86 Q1427.05 1534.01 1427.05 1537.76 Q1427.05 1541.49 1428.88 1543.66 Q1430.73 1545.82 1433.88 1545.82 Q1437.02 1545.82 1438.85 1543.66 Q1440.7 1541.49 1440.7 1537.76 Q1440.7 1534.01 1438.85 1531.86 Q1437.02 1529.7 1433.88 1529.7 M1443.16 1515.05 L1443.16 1519.31 Q1441.4 1518.48 1439.59 1518.04 Q1437.81 1517.6 1436.05 1517.6 Q1431.42 1517.6 1428.97 1520.72 Q1426.54 1523.85 1426.19 1530.17 Q1427.56 1528.15 1429.62 1527.09 Q1431.68 1526 1434.15 1526 Q1439.36 1526 1442.37 1529.17 Q1445.4 1532.32 1445.4 1537.76 Q1445.4 1543.08 1442.26 1546.3 Q1439.11 1549.52 1433.88 1549.52 Q1427.88 1549.52 1424.71 1544.94 Q1421.54 1540.33 1421.54 1531.6 Q1421.54 1523.41 1425.43 1518.55 Q1429.32 1513.66 1435.87 1513.66 Q1437.63 1513.66 1439.41 1514.01 Q1441.21 1514.36 1443.16 1515.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M1845.45 1517.37 Q1841.84 1517.37 1840.01 1520.93 Q1838.21 1524.47 1838.21 1531.6 Q1838.21 1538.71 1840.01 1542.27 Q1841.84 1545.82 1845.45 1545.82 Q1849.09 1545.82 1850.89 1542.27 Q1852.72 1538.71 1852.72 1531.6 Q1852.72 1524.47 1850.89 1520.93 Q1849.09 1517.37 1845.45 1517.37 M1845.45 1513.66 Q1851.26 1513.66 1854.32 1518.27 Q1857.4 1522.85 1857.4 1531.6 Q1857.4 1540.33 1854.32 1544.94 Q1851.26 1549.52 1845.45 1549.52 Q1839.64 1549.52 1836.56 1544.94 Q1833.51 1540.33 1833.51 1531.6 Q1833.51 1522.85 1836.56 1518.27 Q1839.64 1513.66 1845.45 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M1865.61 1542.97 L1870.5 1542.97 L1870.5 1548.85 L1865.61 1548.85 L1865.61 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M1890.68 1532.44 Q1887.35 1532.44 1885.43 1534.22 Q1883.53 1536 1883.53 1539.13 Q1883.53 1542.25 1885.43 1544.03 Q1887.35 1545.82 1890.68 1545.82 Q1894.02 1545.82 1895.94 1544.03 Q1897.86 1542.23 1897.86 1539.13 Q1897.86 1536 1895.94 1534.22 Q1894.04 1532.44 1890.68 1532.44 M1886.01 1530.45 Q1883 1529.7 1881.31 1527.64 Q1879.64 1525.58 1879.64 1522.62 Q1879.64 1518.48 1882.58 1516.07 Q1885.55 1513.66 1890.68 1513.66 Q1895.85 1513.66 1898.79 1516.07 Q1901.73 1518.48 1901.73 1522.62 Q1901.73 1525.58 1900.04 1527.64 Q1898.37 1529.7 1895.38 1530.45 Q1898.76 1531.23 1900.64 1533.52 Q1902.54 1535.82 1902.54 1539.13 Q1902.54 1544.15 1899.46 1546.83 Q1896.4 1549.52 1890.68 1549.52 Q1884.97 1549.52 1881.89 1546.83 Q1878.83 1544.15 1878.83 1539.13 Q1878.83 1535.82 1880.73 1533.52 Q1882.63 1531.23 1886.01 1530.45 M1884.3 1523.06 Q1884.3 1525.75 1885.96 1527.25 Q1887.65 1528.76 1890.68 1528.76 Q1893.69 1528.76 1895.38 1527.25 Q1897.1 1525.75 1897.1 1523.06 Q1897.1 1520.38 1895.38 1518.87 Q1893.69 1517.37 1890.68 1517.37 Q1887.65 1517.37 1885.96 1518.87 Q1884.3 1520.38 1884.3 1523.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M2292.44 1544.91 L2300.07 1544.91 L2300.07 1518.55 L2291.76 1520.21 L2291.76 1515.95 L2300.03 1514.29 L2304.7 1514.29 L2304.7 1544.91 L2312.34 1544.91 L2312.34 1548.85 L2292.44 1548.85 L2292.44 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M2321.79 1542.97 L2326.67 1542.97 L2326.67 1548.85 L2321.79 1548.85 L2321.79 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M2346.86 1517.37 Q2343.25 1517.37 2341.42 1520.93 Q2339.61 1524.47 2339.61 1531.6 Q2339.61 1538.71 2341.42 1542.27 Q2343.25 1545.82 2346.86 1545.82 Q2350.49 1545.82 2352.3 1542.27 Q2354.12 1538.71 2354.12 1531.6 Q2354.12 1524.47 2352.3 1520.93 Q2350.49 1517.37 2346.86 1517.37 M2346.86 1513.66 Q2352.67 1513.66 2355.72 1518.27 Q2358.8 1522.85 2358.8 1531.6 Q2358.8 1540.33 2355.72 1544.94 Q2352.67 1549.52 2346.86 1549.52 Q2341.05 1549.52 2337.97 1544.94 Q2334.91 1540.33 2334.91 1531.6 Q2334.91 1522.85 2337.97 1518.27 Q2341.05 1513.66 2346.86 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,1203.47 2352.76,1203.47 "/>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,883.412 2352.76,883.412 "/>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,563.356 2352.76,563.356 "/>
+<polyline clip-path="url(#clip092)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,243.3 2352.76,243.3 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 156.598,47.2441 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1203.47 175.496,1203.47 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,883.412 175.496,883.412 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,563.356 175.496,563.356 "/>
+<polyline clip-path="url(#clip090)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,243.3 175.496,243.3 "/>
+<path clip-path="url(#clip090)" d="M65.0198 1189.27 Q61.4087 1189.27 59.58 1192.83 Q57.7745 1196.37 57.7745 1203.5 Q57.7745 1210.61 59.58 1214.17 Q61.4087 1217.72 65.0198 1217.72 Q68.6541 1217.72 70.4596 1214.17 Q72.2883 1210.61 72.2883 1203.5 Q72.2883 1196.37 70.4596 1192.83 Q68.6541 1189.27 65.0198 1189.27 M65.0198 1185.56 Q70.83 1185.56 73.8855 1190.17 Q76.9642 1194.75 76.9642 1203.5 Q76.9642 1212.23 73.8855 1216.84 Q70.83 1221.42 65.0198 1221.42 Q59.2097 1221.42 56.131 1216.84 Q53.0754 1212.23 53.0754 1203.5 Q53.0754 1194.75 56.131 1190.17 Q59.2097 1185.56 65.0198 1185.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M85.1818 1214.87 L90.066 1214.87 L90.066 1220.75 L85.1818 1220.75 L85.1818 1214.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M104.279 1216.81 L120.598 1216.81 L120.598 1220.75 L98.6539 1220.75 L98.6539 1216.81 Q101.316 1214.06 105.899 1209.43 Q110.506 1204.78 111.686 1203.43 Q113.932 1200.91 114.811 1199.17 Q115.714 1197.41 115.714 1195.73 Q115.714 1192.97 113.77 1191.23 Q111.848 1189.5 108.746 1189.5 Q106.547 1189.5 104.094 1190.26 Q101.663 1191.03 98.8854 1192.58 L98.8854 1187.85 Q101.709 1186.72 104.163 1186.14 Q106.617 1185.56 108.654 1185.56 Q114.024 1185.56 117.219 1188.25 Q120.413 1190.93 120.413 1195.42 Q120.413 1197.55 119.603 1199.48 Q118.816 1201.37 116.709 1203.97 Q116.131 1204.64 113.029 1207.85 Q109.927 1211.05 104.279 1216.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M62.9365 869.211 Q59.3254 869.211 57.4967 872.775 Q55.6912 876.317 55.6912 883.447 Q55.6912 890.553 57.4967 894.118 Q59.3254 897.66 62.9365 897.66 Q66.5707 897.66 68.3763 894.118 Q70.205 890.553 70.205 883.447 Q70.205 876.317 68.3763 872.775 Q66.5707 869.211 62.9365 869.211 M62.9365 865.507 Q68.7467 865.507 71.8022 870.113 Q74.8809 874.697 74.8809 883.447 Q74.8809 892.174 71.8022 896.78 Q68.7467 901.363 62.9365 901.363 Q57.1264 901.363 54.0477 896.78 Q50.9921 892.174 50.9921 883.447 Q50.9921 874.697 54.0477 870.113 Q57.1264 865.507 62.9365 865.507 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M83.0984 894.812 L87.9827 894.812 L87.9827 900.692 L83.0984 900.692 L83.0984 894.812 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M111.015 870.206 L99.2095 888.655 L111.015 888.655 L111.015 870.206 M109.788 866.132 L115.668 866.132 L115.668 888.655 L120.598 888.655 L120.598 892.544 L115.668 892.544 L115.668 900.692 L111.015 900.692 L111.015 892.544 L95.4132 892.544 L95.4132 888.03 L109.788 866.132 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M63.2606 549.155 Q59.6495 549.155 57.8208 552.719 Q56.0152 556.261 56.0152 563.391 Q56.0152 570.497 57.8208 574.062 Q59.6495 577.604 63.2606 577.604 Q66.8948 577.604 68.7004 574.062 Q70.5291 570.497 70.5291 563.391 Q70.5291 556.261 68.7004 552.719 Q66.8948 549.155 63.2606 549.155 M63.2606 545.451 Q69.0707 545.451 72.1263 550.057 Q75.205 554.641 75.205 563.391 Q75.205 572.117 72.1263 576.724 Q69.0707 581.307 63.2606 581.307 Q57.4504 581.307 54.3717 576.724 Q51.3162 572.117 51.3162 563.391 Q51.3162 554.641 54.3717 550.057 Q57.4504 545.451 63.2606 545.451 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M83.4225 574.756 L88.3067 574.756 L88.3067 580.636 L83.4225 580.636 L83.4225 574.756 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M109.071 561.493 Q105.922 561.493 104.071 563.645 Q102.242 565.798 102.242 569.548 Q102.242 573.275 104.071 575.451 Q105.922 577.604 109.071 577.604 Q112.219 577.604 114.047 575.451 Q115.899 573.275 115.899 569.548 Q115.899 565.798 114.047 563.645 Q112.219 561.493 109.071 561.493 M118.353 546.84 L118.353 551.099 Q116.594 550.266 114.788 549.826 Q113.006 549.386 111.246 549.386 Q106.617 549.386 104.163 552.511 Q101.733 555.636 101.385 561.955 Q102.751 559.942 104.811 558.877 Q106.871 557.789 109.348 557.789 Q114.557 557.789 117.566 560.96 Q120.598 564.108 120.598 569.548 Q120.598 574.872 117.45 578.09 Q114.302 581.307 109.071 581.307 Q103.075 581.307 99.9039 576.724 Q96.7326 572.117 96.7326 563.391 Q96.7326 555.196 100.621 550.335 Q104.51 545.451 111.061 545.451 Q112.82 545.451 114.603 545.798 Q116.408 546.145 118.353 546.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M63.5152 229.099 Q59.9041 229.099 58.0754 232.663 Q56.2699 236.205 56.2699 243.335 Q56.2699 250.441 58.0754 254.006 Q59.9041 257.547 63.5152 257.547 Q67.1494 257.547 68.955 254.006 Q70.7837 250.441 70.7837 243.335 Q70.7837 236.205 68.955 232.663 Q67.1494 229.099 63.5152 229.099 M63.5152 225.395 Q69.3254 225.395 72.3809 230.001 Q75.4596 234.585 75.4596 243.335 Q75.4596 252.061 72.3809 256.668 Q69.3254 261.251 63.5152 261.251 Q57.7051 261.251 54.6264 256.668 Q51.5708 252.061 51.5708 243.335 Q51.5708 234.585 54.6264 230.001 Q57.7051 225.395 63.5152 225.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M83.6771 254.7 L88.5614 254.7 L88.5614 260.58 L83.6771 260.58 L83.6771 254.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip090)" d="M108.746 244.168 Q105.413 244.168 103.492 245.95 Q101.594 247.733 101.594 250.858 Q101.594 253.983 103.492 255.765 Q105.413 257.547 108.746 257.547 Q112.08 257.547 114.001 255.765 Q115.922 253.96 115.922 250.858 Q115.922 247.733 114.001 245.95 Q112.103 244.168 108.746 244.168 M104.071 242.177 Q101.061 241.436 99.3715 239.376 Q97.7048 237.316 97.7048 234.353 Q97.7048 230.21 100.645 227.802 Q103.608 225.395 108.746 225.395 Q113.908 225.395 116.848 227.802 Q119.788 230.21 119.788 234.353 Q119.788 237.316 118.098 239.376 Q116.432 241.436 113.445 242.177 Q116.825 242.964 118.7 245.256 Q120.598 247.548 120.598 250.858 Q120.598 255.881 117.52 258.566 Q114.464 261.251 108.746 261.251 Q103.029 261.251 99.9502 258.566 Q96.8947 255.881 96.8947 250.858 Q96.8947 247.548 98.7928 245.256 Q100.691 242.964 104.071 242.177 M102.358 234.793 Q102.358 237.478 104.024 238.983 Q105.714 240.487 108.746 240.487 Q111.756 240.487 113.445 238.983 Q115.158 237.478 115.158 234.793 Q115.158 232.108 113.445 230.603 Q111.756 229.099 108.746 229.099 Q105.714 229.099 104.024 230.603 Q102.358 232.108 102.358 234.793 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip092)" cx="900.754" cy="117.213" r="21.6" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="1337.66" cy="998.336" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="2288.96" cy="1030.31" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="1840.64" cy="87.9763" r="21.6" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="218.754" cy="1410.69" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="768.03" cy="1026.75" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="2017.02" cy="967.646" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="2283.71" cy="1087.25" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="1955.74" cy="1445.72" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="257.146" cy="888.562" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="1413.3" cy="296.531" r="36" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip092)" cx="611.082" cy="118.991" r="28.8" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<path clip-path="url(#clip092)" d="M602.682 154.046 L602.682 226.046 L674.682 226.046 L674.682 154.046 L602.682 154.046 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip092)" d="M1946.4 748.137 L1946.4 820.137 L2018.4 820.137 L2018.4 748.137 L1946.4 748.137 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip092)" d="M998.777 723.237 L998.777 763.237 L1038.78 763.237 L1038.78 723.237 L998.777 723.237 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip092)" d="M798.323 1246.75 L798.323 1310.75 L862.323 1310.75 L862.323 1246.75 L798.323 1246.75 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip092)" d="M2262.6 872.347 L2262.6 928.347 L2318.6 928.347 L2318.6 872.347 L2262.6 872.347 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+</svg>
diff --git a/previews/PR3525/tutorials/linear/facility_location/fb5e71cf.svg b/previews/PR3525/tutorials/linear/facility_location/fb5e71cf.svg
new file mode 100644
index 00000000000..0f6aab11dfd
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/facility_location/fb5e71cf.svg
@@ -0,0 +1,61 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip970">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip970)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip971">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip970)" d="M156.598 1486.45 L2352.76 1486.45 L2352.76 47.2441 L156.598 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip972">
+    <rect x="156" y="47" width="2197" height="1440"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="496.242,1486.45 496.242,47.2441 "/>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="953.502,1486.45 953.502,47.2441 "/>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1410.76,1486.45 1410.76,47.2441 "/>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1868.02,1486.45 1868.02,47.2441 "/>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2325.28,1486.45 2325.28,47.2441 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 2352.76,1486.45 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="496.242,1486.45 496.242,1467.55 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="953.502,1486.45 953.502,1467.55 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1410.76,1486.45 1410.76,1467.55 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1868.02,1486.45 1868.02,1467.55 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2325.28,1486.45 2325.28,1467.55 "/>
+<path clip-path="url(#clip970)" d="M474.425 1517.37 Q470.814 1517.37 468.985 1520.93 Q467.18 1524.47 467.18 1531.6 Q467.18 1538.71 468.985 1542.27 Q470.814 1545.82 474.425 1545.82 Q478.059 1545.82 479.865 1542.27 Q481.694 1538.71 481.694 1531.6 Q481.694 1524.47 479.865 1520.93 Q478.059 1517.37 474.425 1517.37 M474.425 1513.66 Q480.235 1513.66 483.291 1518.27 Q486.37 1522.85 486.37 1531.6 Q486.37 1540.33 483.291 1544.94 Q480.235 1549.52 474.425 1549.52 Q468.615 1549.52 465.536 1544.94 Q462.481 1540.33 462.481 1531.6 Q462.481 1522.85 465.536 1518.27 Q468.615 1513.66 474.425 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M494.587 1542.97 L499.471 1542.97 L499.471 1548.85 L494.587 1548.85 L494.587 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M513.684 1544.91 L530.004 1544.91 L530.004 1548.85 L508.059 1548.85 L508.059 1544.91 Q510.721 1542.16 515.305 1537.53 Q519.911 1532.88 521.092 1531.53 Q523.337 1529.01 524.217 1527.27 Q525.119 1525.51 525.119 1523.82 Q525.119 1521.07 523.175 1519.33 Q521.254 1517.6 518.152 1517.6 Q515.953 1517.6 513.499 1518.36 Q511.069 1519.13 508.291 1520.68 L508.291 1515.95 Q511.115 1514.82 513.569 1514.24 Q516.022 1513.66 518.059 1513.66 Q523.43 1513.66 526.624 1516.35 Q529.818 1519.03 529.818 1523.52 Q529.818 1525.65 529.008 1527.57 Q528.221 1529.47 526.115 1532.07 Q525.536 1532.74 522.434 1535.95 Q519.332 1539.15 513.684 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M930.644 1517.37 Q927.032 1517.37 925.204 1520.93 Q923.398 1524.47 923.398 1531.6 Q923.398 1538.71 925.204 1542.27 Q927.032 1545.82 930.644 1545.82 Q934.278 1545.82 936.083 1542.27 Q937.912 1538.71 937.912 1531.6 Q937.912 1524.47 936.083 1520.93 Q934.278 1517.37 930.644 1517.37 M930.644 1513.66 Q936.454 1513.66 939.509 1518.27 Q942.588 1522.85 942.588 1531.6 Q942.588 1540.33 939.509 1544.94 Q936.454 1549.52 930.644 1549.52 Q924.833 1549.52 921.755 1544.94 Q918.699 1540.33 918.699 1531.6 Q918.699 1522.85 921.755 1518.27 Q924.833 1513.66 930.644 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M950.805 1542.97 L955.69 1542.97 L955.69 1548.85 L950.805 1548.85 L950.805 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M978.722 1518.36 L966.917 1536.81 L978.722 1536.81 L978.722 1518.36 M977.495 1514.29 L983.375 1514.29 L983.375 1536.81 L988.305 1536.81 L988.305 1540.7 L983.375 1540.7 L983.375 1548.85 L978.722 1548.85 L978.722 1540.7 L963.12 1540.7 L963.12 1536.19 L977.495 1514.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M1388.07 1517.37 Q1384.45 1517.37 1382.63 1520.93 Q1380.82 1524.47 1380.82 1531.6 Q1380.82 1538.71 1382.63 1542.27 Q1384.45 1545.82 1388.07 1545.82 Q1391.7 1545.82 1393.51 1542.27 Q1395.33 1538.71 1395.33 1531.6 Q1395.33 1524.47 1393.51 1520.93 Q1391.7 1517.37 1388.07 1517.37 M1388.07 1513.66 Q1393.88 1513.66 1396.93 1518.27 Q1400.01 1522.85 1400.01 1531.6 Q1400.01 1540.33 1396.93 1544.94 Q1393.88 1549.52 1388.07 1549.52 Q1382.26 1549.52 1379.18 1544.94 Q1376.12 1540.33 1376.12 1531.6 Q1376.12 1522.85 1379.18 1518.27 Q1382.26 1513.66 1388.07 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M1408.23 1542.97 L1413.11 1542.97 L1413.11 1548.85 L1408.23 1548.85 L1408.23 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M1433.88 1529.7 Q1430.73 1529.7 1428.88 1531.86 Q1427.05 1534.01 1427.05 1537.76 Q1427.05 1541.49 1428.88 1543.66 Q1430.73 1545.82 1433.88 1545.82 Q1437.02 1545.82 1438.85 1543.66 Q1440.7 1541.49 1440.7 1537.76 Q1440.7 1534.01 1438.85 1531.86 Q1437.02 1529.7 1433.88 1529.7 M1443.16 1515.05 L1443.16 1519.31 Q1441.4 1518.48 1439.59 1518.04 Q1437.81 1517.6 1436.05 1517.6 Q1431.42 1517.6 1428.97 1520.72 Q1426.54 1523.85 1426.19 1530.17 Q1427.56 1528.15 1429.62 1527.09 Q1431.68 1526 1434.15 1526 Q1439.36 1526 1442.37 1529.17 Q1445.4 1532.32 1445.4 1537.76 Q1445.4 1543.08 1442.26 1546.3 Q1439.11 1549.52 1433.88 1549.52 Q1427.88 1549.52 1424.71 1544.94 Q1421.54 1540.33 1421.54 1531.6 Q1421.54 1523.41 1425.43 1518.55 Q1429.32 1513.66 1435.87 1513.66 Q1437.63 1513.66 1439.41 1514.01 Q1441.21 1514.36 1443.16 1515.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M1845.45 1517.37 Q1841.84 1517.37 1840.01 1520.93 Q1838.21 1524.47 1838.21 1531.6 Q1838.21 1538.71 1840.01 1542.27 Q1841.84 1545.82 1845.45 1545.82 Q1849.09 1545.82 1850.89 1542.27 Q1852.72 1538.71 1852.72 1531.6 Q1852.72 1524.47 1850.89 1520.93 Q1849.09 1517.37 1845.45 1517.37 M1845.45 1513.66 Q1851.26 1513.66 1854.32 1518.27 Q1857.4 1522.85 1857.4 1531.6 Q1857.4 1540.33 1854.32 1544.94 Q1851.26 1549.52 1845.45 1549.52 Q1839.64 1549.52 1836.56 1544.94 Q1833.51 1540.33 1833.51 1531.6 Q1833.51 1522.85 1836.56 1518.27 Q1839.64 1513.66 1845.45 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M1865.61 1542.97 L1870.5 1542.97 L1870.5 1548.85 L1865.61 1548.85 L1865.61 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M1890.68 1532.44 Q1887.35 1532.44 1885.43 1534.22 Q1883.53 1536 1883.53 1539.13 Q1883.53 1542.25 1885.43 1544.03 Q1887.35 1545.82 1890.68 1545.82 Q1894.02 1545.82 1895.94 1544.03 Q1897.86 1542.23 1897.86 1539.13 Q1897.86 1536 1895.94 1534.22 Q1894.04 1532.44 1890.68 1532.44 M1886.01 1530.45 Q1883 1529.7 1881.31 1527.64 Q1879.64 1525.58 1879.64 1522.62 Q1879.64 1518.48 1882.58 1516.07 Q1885.55 1513.66 1890.68 1513.66 Q1895.85 1513.66 1898.79 1516.07 Q1901.73 1518.48 1901.73 1522.62 Q1901.73 1525.58 1900.04 1527.64 Q1898.37 1529.7 1895.38 1530.45 Q1898.76 1531.23 1900.64 1533.52 Q1902.54 1535.82 1902.54 1539.13 Q1902.54 1544.15 1899.46 1546.83 Q1896.4 1549.52 1890.68 1549.52 Q1884.97 1549.52 1881.89 1546.83 Q1878.83 1544.15 1878.83 1539.13 Q1878.83 1535.82 1880.73 1533.52 Q1882.63 1531.23 1886.01 1530.45 M1884.3 1523.06 Q1884.3 1525.75 1885.96 1527.25 Q1887.65 1528.76 1890.68 1528.76 Q1893.69 1528.76 1895.38 1527.25 Q1897.1 1525.75 1897.1 1523.06 Q1897.1 1520.38 1895.38 1518.87 Q1893.69 1517.37 1890.68 1517.37 Q1887.65 1517.37 1885.96 1518.87 Q1884.3 1520.38 1884.3 1523.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2292.44 1544.91 L2300.07 1544.91 L2300.07 1518.55 L2291.76 1520.21 L2291.76 1515.95 L2300.03 1514.29 L2304.7 1514.29 L2304.7 1544.91 L2312.34 1544.91 L2312.34 1548.85 L2292.44 1548.85 L2292.44 1544.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2321.79 1542.97 L2326.67 1542.97 L2326.67 1548.85 L2321.79 1548.85 L2321.79 1542.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2346.86 1517.37 Q2343.25 1517.37 2341.42 1520.93 Q2339.61 1524.47 2339.61 1531.6 Q2339.61 1538.71 2341.42 1542.27 Q2343.25 1545.82 2346.86 1545.82 Q2350.49 1545.82 2352.3 1542.27 Q2354.12 1538.71 2354.12 1531.6 Q2354.12 1524.47 2352.3 1520.93 Q2350.49 1517.37 2346.86 1517.37 M2346.86 1513.66 Q2352.67 1513.66 2355.72 1518.27 Q2358.8 1522.85 2358.8 1531.6 Q2358.8 1540.33 2355.72 1544.94 Q2352.67 1549.52 2346.86 1549.52 Q2341.05 1549.52 2337.97 1544.94 Q2334.91 1540.33 2334.91 1531.6 Q2334.91 1522.85 2337.97 1518.27 Q2341.05 1513.66 2346.86 1513.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,1203.47 2352.76,1203.47 "/>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,883.412 2352.76,883.412 "/>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,563.356 2352.76,563.356 "/>
+<polyline clip-path="url(#clip972)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="156.598,243.3 2352.76,243.3 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1486.45 156.598,47.2441 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,1203.47 175.496,1203.47 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,883.412 175.496,883.412 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,563.356 175.496,563.356 "/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="156.598,243.3 175.496,243.3 "/>
+<path clip-path="url(#clip970)" d="M65.0198 1189.27 Q61.4087 1189.27 59.58 1192.83 Q57.7745 1196.37 57.7745 1203.5 Q57.7745 1210.61 59.58 1214.17 Q61.4087 1217.72 65.0198 1217.72 Q68.6541 1217.72 70.4596 1214.17 Q72.2883 1210.61 72.2883 1203.5 Q72.2883 1196.37 70.4596 1192.83 Q68.6541 1189.27 65.0198 1189.27 M65.0198 1185.56 Q70.83 1185.56 73.8855 1190.17 Q76.9642 1194.75 76.9642 1203.5 Q76.9642 1212.23 73.8855 1216.84 Q70.83 1221.42 65.0198 1221.42 Q59.2097 1221.42 56.131 1216.84 Q53.0754 1212.23 53.0754 1203.5 Q53.0754 1194.75 56.131 1190.17 Q59.2097 1185.56 65.0198 1185.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M85.1818 1214.87 L90.066 1214.87 L90.066 1220.75 L85.1818 1220.75 L85.1818 1214.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M104.279 1216.81 L120.598 1216.81 L120.598 1220.75 L98.6539 1220.75 L98.6539 1216.81 Q101.316 1214.06 105.899 1209.43 Q110.506 1204.78 111.686 1203.43 Q113.932 1200.91 114.811 1199.17 Q115.714 1197.41 115.714 1195.73 Q115.714 1192.97 113.77 1191.23 Q111.848 1189.5 108.746 1189.5 Q106.547 1189.5 104.094 1190.26 Q101.663 1191.03 98.8854 1192.58 L98.8854 1187.85 Q101.709 1186.72 104.163 1186.14 Q106.617 1185.56 108.654 1185.56 Q114.024 1185.56 117.219 1188.25 Q120.413 1190.93 120.413 1195.42 Q120.413 1197.55 119.603 1199.48 Q118.816 1201.37 116.709 1203.97 Q116.131 1204.64 113.029 1207.85 Q109.927 1211.05 104.279 1216.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M62.9365 869.211 Q59.3254 869.211 57.4967 872.775 Q55.6912 876.317 55.6912 883.447 Q55.6912 890.553 57.4967 894.118 Q59.3254 897.66 62.9365 897.66 Q66.5707 897.66 68.3763 894.118 Q70.205 890.553 70.205 883.447 Q70.205 876.317 68.3763 872.775 Q66.5707 869.211 62.9365 869.211 M62.9365 865.507 Q68.7467 865.507 71.8022 870.113 Q74.8809 874.697 74.8809 883.447 Q74.8809 892.174 71.8022 896.78 Q68.7467 901.363 62.9365 901.363 Q57.1264 901.363 54.0477 896.78 Q50.9921 892.174 50.9921 883.447 Q50.9921 874.697 54.0477 870.113 Q57.1264 865.507 62.9365 865.507 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M83.0984 894.812 L87.9827 894.812 L87.9827 900.692 L83.0984 900.692 L83.0984 894.812 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M111.015 870.206 L99.2095 888.655 L111.015 888.655 L111.015 870.206 M109.788 866.132 L115.668 866.132 L115.668 888.655 L120.598 888.655 L120.598 892.544 L115.668 892.544 L115.668 900.692 L111.015 900.692 L111.015 892.544 L95.4132 892.544 L95.4132 888.03 L109.788 866.132 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M63.2606 549.155 Q59.6495 549.155 57.8208 552.719 Q56.0152 556.261 56.0152 563.391 Q56.0152 570.497 57.8208 574.062 Q59.6495 577.604 63.2606 577.604 Q66.8948 577.604 68.7004 574.062 Q70.5291 570.497 70.5291 563.391 Q70.5291 556.261 68.7004 552.719 Q66.8948 549.155 63.2606 549.155 M63.2606 545.451 Q69.0707 545.451 72.1263 550.057 Q75.205 554.641 75.205 563.391 Q75.205 572.117 72.1263 576.724 Q69.0707 581.307 63.2606 581.307 Q57.4504 581.307 54.3717 576.724 Q51.3162 572.117 51.3162 563.391 Q51.3162 554.641 54.3717 550.057 Q57.4504 545.451 63.2606 545.451 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M83.4225 574.756 L88.3067 574.756 L88.3067 580.636 L83.4225 580.636 L83.4225 574.756 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M109.071 561.493 Q105.922 561.493 104.071 563.645 Q102.242 565.798 102.242 569.548 Q102.242 573.275 104.071 575.451 Q105.922 577.604 109.071 577.604 Q112.219 577.604 114.047 575.451 Q115.899 573.275 115.899 569.548 Q115.899 565.798 114.047 563.645 Q112.219 561.493 109.071 561.493 M118.353 546.84 L118.353 551.099 Q116.594 550.266 114.788 549.826 Q113.006 549.386 111.246 549.386 Q106.617 549.386 104.163 552.511 Q101.733 555.636 101.385 561.955 Q102.751 559.942 104.811 558.877 Q106.871 557.789 109.348 557.789 Q114.557 557.789 117.566 560.96 Q120.598 564.108 120.598 569.548 Q120.598 574.872 117.45 578.09 Q114.302 581.307 109.071 581.307 Q103.075 581.307 99.9039 576.724 Q96.7326 572.117 96.7326 563.391 Q96.7326 555.196 100.621 550.335 Q104.51 545.451 111.061 545.451 Q112.82 545.451 114.603 545.798 Q116.408 546.145 118.353 546.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M63.5152 229.099 Q59.9041 229.099 58.0754 232.663 Q56.2699 236.205 56.2699 243.335 Q56.2699 250.441 58.0754 254.006 Q59.9041 257.547 63.5152 257.547 Q67.1494 257.547 68.955 254.006 Q70.7837 250.441 70.7837 243.335 Q70.7837 236.205 68.955 232.663 Q67.1494 229.099 63.5152 229.099 M63.5152 225.395 Q69.3254 225.395 72.3809 230.001 Q75.4596 234.585 75.4596 243.335 Q75.4596 252.061 72.3809 256.668 Q69.3254 261.251 63.5152 261.251 Q57.7051 261.251 54.6264 256.668 Q51.5708 252.061 51.5708 243.335 Q51.5708 234.585 54.6264 230.001 Q57.7051 225.395 63.5152 225.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M83.6771 254.7 L88.5614 254.7 L88.5614 260.58 L83.6771 260.58 L83.6771 254.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M108.746 244.168 Q105.413 244.168 103.492 245.95 Q101.594 247.733 101.594 250.858 Q101.594 253.983 103.492 255.765 Q105.413 257.547 108.746 257.547 Q112.08 257.547 114.001 255.765 Q115.922 253.96 115.922 250.858 Q115.922 247.733 114.001 245.95 Q112.103 244.168 108.746 244.168 M104.071 242.177 Q101.061 241.436 99.3715 239.376 Q97.7048 237.316 97.7048 234.353 Q97.7048 230.21 100.645 227.802 Q103.608 225.395 108.746 225.395 Q113.908 225.395 116.848 227.802 Q119.788 230.21 119.788 234.353 Q119.788 237.316 118.098 239.376 Q116.432 241.436 113.445 242.177 Q116.825 242.964 118.7 245.256 Q120.598 247.548 120.598 250.858 Q120.598 255.881 117.52 258.566 Q114.464 261.251 108.746 261.251 Q103.029 261.251 99.9502 258.566 Q96.8947 255.881 96.8947 250.858 Q96.8947 247.548 98.7928 245.256 Q100.691 242.964 104.071 242.177 M102.358 234.793 Q102.358 237.478 104.024 238.983 Q105.714 240.487 108.746 240.487 Q111.756 240.487 113.445 238.983 Q115.158 237.478 115.158 234.793 Q115.158 232.108 113.445 230.603 Q111.756 229.099 108.746 229.099 Q105.714 229.099 104.024 230.603 Q102.358 232.108 102.358 234.793 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip972)" cx="900.754" cy="117.213" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="1337.66" cy="998.336" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="2288.96" cy="1030.31" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="1840.64" cy="87.9763" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="218.754" cy="1410.69" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="768.03" cy="1026.75" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="2017.02" cy="967.646" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="2283.71" cy="1087.25" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="1955.74" cy="1445.72" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="257.146" cy="888.562" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="1413.3" cy="296.531" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip972)" cx="611.082" cy="118.991" r="14.4" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<path clip-path="url(#clip972)" d="M614.682 166.046 L614.682 214.046 L662.682 214.046 L662.682 166.046 L614.682 166.046 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip972)" d="M1958.4 760.137 L1958.4 808.137 L2006.4 808.137 L2006.4 760.137 L1958.4 760.137 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip972)" d="M994.777 719.237 L994.777 767.237 L1042.78 767.237 L1042.78 719.237 L994.777 719.237 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip972)" d="M806.323 1254.75 L806.323 1302.75 L854.323 1302.75 L854.323 1254.75 L806.323 1254.75 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip972)" d="M2266.6 876.347 L2266.6 924.347 L2314.6 924.347 L2314.6 876.347 L2266.6 876.347 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="6.4"/>
+<path clip-path="url(#clip970)" d="M1901.81 250.738 L2279.55 250.738 L2279.55 95.2176 L1901.81 95.2176  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip970)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1901.81,250.738 2279.55,250.738 2279.55,95.2176 1901.81,95.2176 1901.81,250.738 "/>
+<circle clip-path="url(#clip970)" cx="1999.42" cy="147.058" r="20.48" fill="#0000ff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.55111"/>
+<path clip-path="url(#clip970)" d="M2124.89 132.44 L2124.89 137.37 Q2122.53 135.171 2119.85 134.083 Q2117.19 132.995 2114.18 132.995 Q2108.25 132.995 2105.1 136.629 Q2101.96 140.24 2101.96 147.092 Q2101.96 153.921 2105.1 157.555 Q2108.25 161.166 2114.18 161.166 Q2117.19 161.166 2119.85 160.078 Q2122.53 158.99 2124.89 156.791 L2124.89 161.676 Q2122.44 163.342 2119.69 164.176 Q2116.96 165.009 2113.9 165.009 Q2106.05 165.009 2101.54 160.217 Q2097.02 155.402 2097.02 147.092 Q2097.02 138.759 2101.54 133.967 Q2106.05 129.153 2113.9 129.153 Q2117 129.153 2119.73 129.986 Q2122.49 130.796 2124.89 132.44 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2131.93 128.319 L2136.19 128.319 L2136.19 164.338 L2131.93 164.338 L2131.93 128.319 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2145.1 138.412 L2149.36 138.412 L2149.36 164.338 L2145.1 164.338 L2145.1 138.412 M2145.1 128.319 L2149.36 128.319 L2149.36 133.713 L2145.1 133.713 L2145.1 128.319 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2180.45 150.31 L2180.45 152.393 L2160.87 152.393 Q2161.14 156.791 2163.51 159.106 Q2165.89 161.398 2170.13 161.398 Q2172.58 161.398 2174.87 160.796 Q2177.19 160.194 2179.45 158.99 L2179.45 163.018 Q2177.16 163.99 2174.76 164.5 Q2172.35 165.009 2169.87 165.009 Q2163.67 165.009 2160.03 161.398 Q2156.42 157.787 2156.42 151.629 Q2156.42 145.264 2159.85 141.537 Q2163.3 137.787 2169.13 137.787 Q2174.36 137.787 2177.39 141.166 Q2180.45 144.523 2180.45 150.31 M2176.19 149.06 Q2176.14 145.565 2174.22 143.481 Q2172.33 141.398 2169.18 141.398 Q2165.61 141.398 2163.46 143.412 Q2161.33 145.426 2161.01 149.083 L2176.19 149.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2208.99 148.689 L2208.99 164.338 L2204.73 164.338 L2204.73 148.828 Q2204.73 145.148 2203.3 143.319 Q2201.86 141.49 2198.99 141.49 Q2195.54 141.49 2193.55 143.69 Q2191.56 145.889 2191.56 149.685 L2191.56 164.338 L2187.28 164.338 L2187.28 138.412 L2191.56 138.412 L2191.56 142.44 Q2193.09 140.102 2195.15 138.944 Q2197.23 137.787 2199.94 137.787 Q2204.41 137.787 2206.7 140.565 Q2208.99 143.319 2208.99 148.689 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2221.7 131.051 L2221.7 138.412 L2230.47 138.412 L2230.47 141.722 L2221.7 141.722 L2221.7 155.796 Q2221.7 158.967 2222.56 159.87 Q2223.44 160.773 2226.1 160.773 L2230.47 160.773 L2230.47 164.338 L2226.1 164.338 Q2221.17 164.338 2219.29 162.509 Q2217.42 160.657 2217.42 155.796 L2217.42 141.722 L2214.29 141.722 L2214.29 138.412 L2217.42 138.412 L2217.42 131.051 L2221.7 131.051 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2252.6 139.176 L2252.6 143.203 Q2250.8 142.277 2248.85 141.815 Q2246.91 141.352 2244.82 141.352 Q2241.65 141.352 2240.06 142.324 Q2238.48 143.296 2238.48 145.24 Q2238.48 146.722 2239.62 147.578 Q2240.75 148.412 2244.18 149.176 L2245.64 149.5 Q2250.17 150.472 2252.07 152.254 Q2253.99 154.014 2253.99 157.185 Q2253.99 160.796 2251.12 162.902 Q2248.27 165.009 2243.27 165.009 Q2241.19 165.009 2238.92 164.592 Q2236.68 164.199 2234.18 163.388 L2234.18 158.99 Q2236.54 160.217 2238.83 160.842 Q2241.12 161.444 2243.37 161.444 Q2246.38 161.444 2248 160.426 Q2249.62 159.384 2249.62 157.509 Q2249.62 155.773 2248.44 154.847 Q2247.28 153.921 2243.32 153.064 L2241.84 152.717 Q2237.88 151.884 2236.12 150.171 Q2234.36 148.435 2234.36 145.426 Q2234.36 141.768 2236.95 139.778 Q2239.55 137.787 2244.32 137.787 Q2246.68 137.787 2248.76 138.134 Q2250.84 138.481 2252.6 139.176 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M1977.47 176.955 L1977.47 220.84 L2021.36 220.84 L2021.36 176.955 L1977.47 176.955 Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#ff0000" stroke-opacity="1" stroke-width="5.85143"/>
+<path clip-path="url(#clip970)" d="M2097.02 181.618 L2116.89 181.618 L2116.89 185.553 L2101.7 185.553 L2101.7 195.738 L2115.4 195.738 L2115.4 199.673 L2101.7 199.673 L2101.7 216.178 L2097.02 216.178 L2097.02 181.618 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2131.54 203.145 Q2126.38 203.145 2124.39 204.326 Q2122.39 205.506 2122.39 208.354 Q2122.39 210.622 2123.88 211.965 Q2125.38 213.284 2127.95 213.284 Q2131.49 213.284 2133.62 210.784 Q2135.77 208.261 2135.77 204.094 L2135.77 203.145 L2131.54 203.145 M2140.03 201.386 L2140.03 216.178 L2135.77 216.178 L2135.77 212.242 Q2134.32 214.603 2132.14 215.738 Q2129.96 216.849 2126.82 216.849 Q2122.83 216.849 2120.47 214.627 Q2118.14 212.381 2118.14 208.631 Q2118.14 204.256 2121.05 202.034 Q2123.99 199.812 2129.8 199.812 L2135.77 199.812 L2135.77 199.395 Q2135.77 196.455 2133.83 194.858 Q2131.91 193.238 2128.41 193.238 Q2126.19 193.238 2124.08 193.77 Q2121.98 194.303 2120.03 195.367 L2120.03 191.432 Q2122.37 190.53 2124.57 190.09 Q2126.77 189.627 2128.85 189.627 Q2134.48 189.627 2137.26 192.543 Q2140.03 195.46 2140.03 201.386 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2167.46 191.247 L2167.46 195.229 Q2165.66 194.233 2163.83 193.747 Q2162.02 193.238 2160.17 193.238 Q2156.03 193.238 2153.74 195.877 Q2151.45 198.492 2151.45 203.238 Q2151.45 207.983 2153.74 210.622 Q2156.03 213.238 2160.17 213.238 Q2162.02 213.238 2163.83 212.752 Q2165.66 212.242 2167.46 211.247 L2167.46 215.182 Q2165.68 216.016 2163.76 216.432 Q2161.86 216.849 2159.71 216.849 Q2153.85 216.849 2150.4 213.168 Q2146.96 209.488 2146.96 203.238 Q2146.96 196.895 2150.43 193.261 Q2153.92 189.627 2159.99 189.627 Q2161.95 189.627 2163.83 190.043 Q2165.7 190.437 2167.46 191.247 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2174.87 190.252 L2179.13 190.252 L2179.13 216.178 L2174.87 216.178 L2174.87 190.252 M2174.87 180.159 L2179.13 180.159 L2179.13 185.553 L2174.87 185.553 L2174.87 180.159 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2188.04 180.159 L2192.3 180.159 L2192.3 216.178 L2188.04 216.178 L2188.04 180.159 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2201.21 190.252 L2205.47 190.252 L2205.47 216.178 L2201.21 216.178 L2201.21 190.252 M2201.21 180.159 L2205.47 180.159 L2205.47 185.553 L2201.21 185.553 L2201.21 180.159 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2218.6 182.891 L2218.6 190.252 L2227.37 190.252 L2227.37 193.562 L2218.6 193.562 L2218.6 207.636 Q2218.6 210.807 2219.45 211.71 Q2220.33 212.613 2223 212.613 L2227.37 212.613 L2227.37 216.178 L2223 216.178 Q2218.07 216.178 2216.19 214.349 Q2214.32 212.497 2214.32 207.636 L2214.32 193.562 L2211.19 193.562 L2211.19 190.252 L2214.32 190.252 L2214.32 182.891 L2218.6 182.891 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip970)" d="M2243.76 218.585 Q2241.95 223.215 2240.24 224.627 Q2238.53 226.039 2235.66 226.039 L2232.26 226.039 L2232.26 222.474 L2234.76 222.474 Q2236.51 222.474 2237.49 221.64 Q2238.46 220.807 2239.64 217.705 L2240.4 215.761 L2229.92 190.252 L2234.43 190.252 L2242.53 210.529 L2250.64 190.252 L2255.15 190.252 L2243.76 218.585 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/linear/facility_location/index.html b/previews/PR3525/tutorials/linear/facility_location/index.html
new file mode 100644
index 00000000000..addf1f73794
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/facility_location/index.html
@@ -0,0 +1,170 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The facility location problem · JuMP</title><meta name="title" content="The facility location problem · JuMP"/><meta property="og:title" content="The facility location problem · JuMP"/><meta property="twitter:title" content="The facility location problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li class="is-active"><a class="tocitem" href>The facility location problem</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Uncapacitated-facility-location"><span>Uncapacitated facility location</span></a></li><li><a class="tocitem" href="#Capacitated-facility-location"><span>Capacitated facility location</span></a></li></ul></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The facility location problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The facility location problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/facility_location.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-facility-location-problem"><a class="docs-heading-anchor" href="#The-facility-location-problem">The facility location problem</a><a id="The-facility-location-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-facility-location-problem" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../facility_location.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Mathieu Tanneau and Alexis Montoison.</strong></p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+import LinearAlgebra
+import Plots
+import Random</code></pre><h2 id="Uncapacitated-facility-location"><a class="docs-heading-anchor" href="#Uncapacitated-facility-location">Uncapacitated facility location</a><a id="Uncapacitated-facility-location-1"></a><a class="docs-heading-anchor-permalink" href="#Uncapacitated-facility-location" title="Permalink"></a></h2><h3 id="Problem-description"><a class="docs-heading-anchor" href="#Problem-description">Problem description</a><a id="Problem-description-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-description" title="Permalink"></a></h3><p>We are given</p><ul><li>A set <span>$M=\{1, \dots, m\}$</span> of clients</li><li>A set <span>$N=\{ 1, \dots, n\}$</span> of sites where a facility can be built</li></ul><p><strong>Decision variables</strong> Decision variables are split into two categories:</p><ul><li>Binary variable <span>$y_{j}$</span> indicates whether facility <span>$j$</span> is built or not</li><li>Binary variable <span>$x_{i, j}$</span> indicates whether client <span>$i$</span> is assigned to facility <span>$j$</span></li></ul><p><strong>Objective</strong> The objective is to minimize the total cost of serving all clients. This costs breaks down into two components:</p><ul><li>Fixed cost of building a facility.</li></ul><p>In this example, this cost is <span>$f_{j} = 1, \ \forall j$</span>.</p><ul><li>Cost of serving clients from the assigned facility.</li></ul><p>In this example, the cost <span>$c_{i, j}$</span> of serving client <span>$i$</span> from facility <span>$j$</span> is the Euclidean distance between the two.</p><p><strong>Constraints</strong></p><ul><li>Each customer must be served by exactly one facility</li><li>A facility cannot serve any client unless it is open</li></ul><h3 id="MILP-formulation"><a class="docs-heading-anchor" href="#MILP-formulation">MILP formulation</a><a id="MILP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#MILP-formulation" title="Permalink"></a></h3><p>The problem can be formulated as the following MILP:</p><p class="math-container">\[\begin{aligned}
+\min_{x, y} \ \ \ &amp;
+\sum_{i, j} c_{i, j} x_{i, j} +
+\sum_{j} f_{j} y_{j} \\
+s.t. &amp;
+\sum_{j} x_{i, j} = 1, &amp;&amp; \forall i \in M \\
+&amp; x_{i, j} \leq y_{j}, &amp;&amp; \forall i \in M, j \in N \\
+&amp; x_{i, j}, y_{j} \in \{0, 1\}, &amp;&amp; \forall i \in M, j \in N
+\end{aligned}\]</p><p>where the first set of constraints ensures that each client is served exactly once, and the second set of constraints ensures that no client is served from an unopened facility.</p><h3 id="Problem-data"><a class="docs-heading-anchor" href="#Problem-data">Problem data</a><a id="Problem-data-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-data" title="Permalink"></a></h3><p>To ensure reproducibility, we set the random number seed:</p><pre><code class="language-julia hljs">Random.seed!(314)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Random.TaskLocalRNG()</code></pre><p>Here&#39;s the data we need:</p><pre><code class="language-julia hljs"># Number of clients
+m = 12
+# Number of facility locations
+n = 5
+
+# Clients&#39; locations
+x_c, y_c = rand(m), rand(m)
+
+# Facilities&#39; potential locations
+x_f, y_f = rand(n), rand(n)
+
+# Fixed costs
+f = ones(n);
+
+# Distance
+c = zeros(m, n)
+for i in 1:m
+    for j in 1:n
+        c[i, j] = LinearAlgebra.norm([x_c[i] - x_f[j], y_c[i] - y_f[j]], 2)
+    end
+end</code></pre><p>Display the data</p><pre><code class="language-julia hljs">Plots.scatter(
+    x_c,
+    y_c;
+    label = &quot;Clients&quot;,
+    markershape = :circle,
+    markercolor = :blue,
+)
+Plots.scatter!(
+    x_f,
+    y_f;
+    label = &quot;Facility&quot;,
+    markershape = :square,
+    markercolor = :white,
+    markersize = 6,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+)</code></pre><img src="fb5e71cf.svg" alt="Example block output"/><h3 id="JuMP-implementation"><a class="docs-heading-anchor" href="#JuMP-implementation">JuMP implementation</a><a id="JuMP-implementation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-implementation" title="Permalink"></a></h3><p>Create a JuMP model</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, y[1:n], Bin);
+@variable(model, x[1:m, 1:n], Bin);
+# Each client is served exactly once
+@constraint(model, client_service[i in 1:m], sum(x[i, j] for j in 1:n) == 1);
+# A facility must be open to serve a client
+@constraint(model, open_facility[i in 1:m, j in 1:n], x[i, j] &lt;= y[j]);
+@objective(model, Min, f&#39; * y + sum(c .* x));</code></pre><p>Solve the uncapacitated facility location problem with HiGHS</p><pre><code class="language-julia hljs">optimize!(model)
+println(&quot;Optimal value: &quot;, objective_value(model))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Optimal value: 5.7018394545724185</code></pre><h3 id="Visualizing-the-solution"><a class="docs-heading-anchor" href="#Visualizing-the-solution">Visualizing the solution</a><a id="Visualizing-the-solution-1"></a><a class="docs-heading-anchor-permalink" href="#Visualizing-the-solution" title="Permalink"></a></h3><p>The threshold 1e-5 ensure that edges between clients and facilities are drawn when <code>x[i, j] ≈ 1</code>.</p><pre><code class="language-julia hljs">x_is_selected = isapprox.(value.(x), 1; atol = 1e-5);
+y_is_selected = isapprox.(value.(y), 1; atol = 1e-5);
+
+p = Plots.scatter(
+    x_c,
+    y_c;
+    markershape = :circle,
+    markercolor = :blue,
+    label = nothing,
+)
+
+Plots.scatter!(
+    x_f,
+    y_f;
+    markershape = :square,
+    markercolor = [(y_is_selected[j] ? :red : :white) for j in 1:n],
+    markersize = 6,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+    label = nothing,
+)
+
+for i in 1:m, j in 1:n
+    if x_is_selected[i, j]
+        Plots.plot!(
+            [x_c[i], x_f[j]],
+            [y_c[i], y_f[j]];
+            color = :black,
+            label = nothing,
+        )
+    end
+end
+
+p</code></pre><img src="a3695eca.svg" alt="Example block output"/><h2 id="Capacitated-facility-location"><a class="docs-heading-anchor" href="#Capacitated-facility-location">Capacitated facility location</a><a id="Capacitated-facility-location-1"></a><a class="docs-heading-anchor-permalink" href="#Capacitated-facility-location" title="Permalink"></a></h2><h3 id="Problem-formulation"><a class="docs-heading-anchor" href="#Problem-formulation">Problem formulation</a><a id="Problem-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Problem-formulation" title="Permalink"></a></h3><p>The capacitated variant introduces a capacity constraint on each facility, that is, clients have a certain level of demand to be served, while each facility only has finite capacity which cannot be exceeded.</p><p>Specifically,</p><ul><li>The demand of client <span>$i$</span> is denoted by <span>$a_{i} \geq 0$</span></li><li>The capacity of facility <span>$j$</span> is denoted by <span>$q_{j} \geq 0$</span></li></ul><p>The capacity constraints then write</p><p class="math-container">\[\begin{aligned}
+\sum_{i} a_{i} x_{i, j} &amp;\leq q_{j} y_{j} &amp;&amp; \forall j \in N
+\end{aligned}\]</p><p>Note that, if <span>$y_{j}$</span> is set to <span>$0$</span>, the capacity constraint above automatically forces <span>$x_{i, j}$</span> to <span>$0$</span>.</p><p>Thus, the capacitated facility location can be formulated as follows</p><p class="math-container">\[\begin{aligned}
+\min_{x, y} \ \ \ &amp;
+\sum_{i, j} c_{i, j} x_{i, j} +
+\sum_{j} f_{j} y_{j} \\
+s.t. &amp;
+\sum_{j} x_{i, j} = 1, &amp;&amp; \forall i \in M \\
+&amp; \sum_{i} a_{i} x_{i, j} \leq q_{j} y_{j}, &amp;&amp; \forall j \in N \\
+&amp; x_{i, j}, y_{j} \in \{0, 1\}, &amp;&amp; \forall i \in M, j \in N
+\end{aligned}\]</p><p>For simplicity, we will assume that there is enough capacity to serve the demand, that is, there exists at least one feasible solution.</p><p>We need some new data:</p><pre><code class="language-julia hljs"># Demands
+a = rand(1:3, m);
+
+# Capacities
+q = rand(5:10, n);</code></pre><p>Display the data</p><pre><code class="language-julia hljs">Plots.scatter(
+    x_c,
+    y_c;
+    label = nothing,
+    markershape = :circle,
+    markercolor = :blue,
+    markersize = 2 .* (2 .+ a),
+)
+
+Plots.scatter!(
+    x_f,
+    y_f;
+    label = nothing,
+    markershape = :rect,
+    markercolor = :white,
+    markersize = q,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+)</code></pre><img src="a9295f9b.svg" alt="Example block output"/><h3 id="JuMP-implementation-2"><a class="docs-heading-anchor" href="#JuMP-implementation-2">JuMP implementation</a><a class="docs-heading-anchor-permalink" href="#JuMP-implementation-2" title="Permalink"></a></h3><p>Create a JuMP model</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, y[1:n], Bin);
+@variable(model, x[1:m, 1:n], Bin);
+# Each client is served exactly once
+@constraint(model, client_service[i in 1:m], sum(x[i, :]) == 1);
+# Capacity constraint
+@constraint(model, capacity, x&#39; * a .&lt;= (q .* y));
+# Objective
+@objective(model, Min, f&#39; * y + sum(c .* x));</code></pre><p>Solve the problem</p><pre><code class="language-julia hljs">optimize!(model)
+println(&quot;Optimal value: &quot;, objective_value(model))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Optimal value: 6.1980444155009975</code></pre><h3 id="Visualizing-the-solution-2"><a class="docs-heading-anchor" href="#Visualizing-the-solution-2">Visualizing the solution</a><a class="docs-heading-anchor-permalink" href="#Visualizing-the-solution-2" title="Permalink"></a></h3><p>The threshold 1e-5 ensure that edges between clients and facilities are drawn when <code>x[i, j] ≈ 1</code>.</p><pre><code class="language-julia hljs">x_is_selected = isapprox.(value.(x), 1; atol = 1e-5);
+y_is_selected = isapprox.(value.(y), 1; atol = 1e-5);</code></pre><p>Display the solution</p><pre><code class="language-julia hljs">p = Plots.scatter(
+    x_c,
+    y_c;
+    label = nothing,
+    markershape = :circle,
+    markercolor = :blue,
+    markersize = 2 .* (2 .+ a),
+)
+
+Plots.scatter!(
+    x_f,
+    y_f;
+    label = nothing,
+    markershape = :rect,
+    markercolor = [(y_is_selected[j] ? :red : :white) for j in 1:n],
+    markersize = q,
+    markerstrokecolor = :red,
+    markerstrokewidth = 2,
+)
+
+for i in 1:m, j in 1:n
+    if x_is_selected[i, j]
+        Plots.plot!(
+            [x_c[i], x_f[j]],
+            [y_c[i], y_f[j]];
+            color = :black,
+            label = nothing,
+        )
+    end
+end
+
+p</code></pre><img src="49a24bc3.svg" alt="Example block output"/></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../tips_and_tricks/">« Tips and tricks</a><a class="docs-footer-nextpage" href="../finance/">Financial modeling problems »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/factory_schedule.jl b/previews/PR3525/tutorials/linear/factory_schedule.jl
new file mode 100644
index 00000000000..f034a93532b
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/factory_schedule.jl
@@ -0,0 +1,327 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The factory schedule example
+
+# **This tutorial was originally contributed by `@Crghilardi`.**
+
+# This tutorial is a Julia translation of Part 5 from [Introduction to Linear
+# Programming with Python](https://github.com/benalexkeen/Introduction-to-linear-programming).
+
+# The purpose of this tutorial is to demonstrate how to use DataFrames and
+# delimited files, and to structure your code that is robust to infeasibilities
+# and permits running with different datasets.
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import CSV
+import DataFrames
+import HiGHS
+import StatsPlots
+import Test  #src
+
+# ## Formulation
+
+# The Factory Scheduling Problem assumes we are optimizing the production of a
+# good from factories ``f \in F`` over the course of 12 months ``m \in M``.
+
+# If a factory ``f`` runs during a month ``m``, a fixed cost of ``a_f`` is
+# incurred, the factory must produce ``x_{m,f}`` units that is within some
+# minimum and maximum production levels ``l_f`` and ``u_f`` respectively, and
+# each unit of production incurs a variable cost ``c_f``. Otherwise, the factory
+# can be shut for the month with zero production and no fixed-cost is incurred.
+# We denote the run/not-run decision by ``z_{m,f} \in \{0, 1\}``, where
+# ``z_{m,f}`` is ``1`` if factory ``f`` runs in month ``m``. The factory must
+# produce enough units to satisfy demand ``d_m``.
+
+# With a little effort, we can formulate our problem as the following linear
+# program:
+# ```math
+# \begin{aligned}
+# \min & \sum\limits_{f \in F, m \in M} a_f z_{m,f} + c_f x_{m,f} \\
+# \text{s.t.} & x_{m,f} \le u_f z_{m,f} && \forall f \in F, m \in M \\
+#             & x_{m,f} \ge l_f z_{m,f} && \forall f \in F, m \in M \\
+#             & \sum\limits_{f\in F} x_{m,f} = d_m && \forall f \in F, m \in M \\
+#             & z_{m,f} \in \{0, 1\} && \forall f \in F, m \in M.
+# \end{aligned}
+# ```
+
+# However, this formulation has a problem: if demand is too high, we may be
+# unable to satisfy the demand constraint, and the problem will be infeasible.
+
+# !!! tip
+#     When modeling, consider ways to formulate your model such that it always
+#     has a feasible solution. This greatly simplifies debugging data errors
+#     that would otherwise result in an infeasible solution. In practice, most
+#     practical decisions have a feasible solution. In our case, we could
+#     satisfy demand (at a high cost) by buying replacement items for the buyer,
+#     or running the factories in overtime to make up the difference.
+
+# We can improve our model by adding a new variable, ``\delta_m``, which
+# represents the quantity of unmet demand in each month ``m``. We penalize
+# ``\delta_m`` by an arbitrarily large value of \$10,000/unit in the objective.
+
+# ```math
+# \begin{aligned}
+# \min & \sum\limits_{f \in F, m \in M} a_f z_{m,f} + c_f x_{m,f} + \sum\limits_{m \in M}10000 \delta_m \\
+# \text{s.t.} & x_{m,f} \le u_f z_{m,f} && \forall f \in F, m \in M \\
+#             & x_{m,f} \ge l_f z_{m,f} && \forall f \in F, m \in M \\
+#             & \sum\limits_{f\in F} x_{m,f} - \delta_m = d_m && \forall f \in F, m \in M \\
+#             & z_{m,f} \in \{0, 1\} && \forall f \in F, m \in M \\
+#             & \delta_m \ge 0 && \forall m \in M.
+# \end{aligned}
+# ```
+
+# ## Data
+
+# The JuMP GitHub repository contains two text files with the data we need for
+# this tutorial.
+
+# The first file contains a dataset of our factories, `A` and `B`, with their
+# production and cost levels for each month. For the documentation, the file is
+# located at:
+
+factories_filename = joinpath(@__DIR__, "factory_schedule_factories.txt");
+
+# To run locally, download [`factory_schedule_factories.txt`](factory_schedule_factories.txt)
+# and update `factories_filename` appropriately.
+
+# The file has the following contents:
+
+print(read(factories_filename, String))
+
+# We use the `CSV` and `DataFrames` packages to read it into Julia:
+
+factory_df = CSV.read(
+    factories_filename,
+    DataFrames.DataFrame;
+    delim = ' ',
+    ignorerepeated = true,
+)
+
+# The second file contains the demand data by month:
+
+demand_filename = joinpath(@__DIR__, "factory_schedule_demand.txt");
+
+# To run locally, download [`factory_schedule_demand.txt`](factory_schedule_demand.txt)
+# and update `demand_filename` appropriately.
+
+demand_df = CSV.read(
+    demand_filename,
+    DataFrames.DataFrame;
+    delim = ' ',
+    ignorerepeated = true,
+)
+
+# ### Data validation
+
+# Before moving on, it's always good practice to validate the data you read from
+# external sources. The more effort you spend here, the fewer issues you will
+# have later. The following function contains a few simple checks, but we could
+# add more. For example, you might want to check that none of the values are
+# too large (or too small), which might indicate a typo or a unit conversion
+# issue (perhaps the variable costs are in \$/1000 units instead of \$/unit).
+
+function valiate_data(
+    demand_df::DataFrames.DataFrame,
+    factory_df::DataFrames.DataFrame,
+)
+    ## Minimum production must not exceed maximum production.
+    @assert all(factory_df.min_production .<= factory_df.max_production)
+    ## Demand, minimum production, fixed costs, and variable costs must all be
+    ## non-negative.
+    @assert all(demand_df.demand .>= 0)
+    @assert all(factory_df.min_production .>= 0)
+    @assert all(factory_df.fixed_cost .>= 0)
+    @assert all(factory_df.variable_cost .>= 0)
+    return
+end
+
+valiate_data(demand_df, factory_df)
+
+# ## JuMP formulation
+
+# Next, we need to code our JuMP formulation. As shown in
+# [Design patterns for larger models](@ref), it's always good practice to code
+# your model in a function that accepts well-defined input and returns
+# well-defined output.
+
+function solve_factory_scheduling(
+    demand_df::DataFrames.DataFrame,
+    factory_df::DataFrames.DataFrame,
+)
+    ## Even though we validated the data above, it's good practice to do it here
+    ## too.
+    valiate_data(demand_df, factory_df)
+    months, factories = unique(factory_df.month), unique(factory_df.factory)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, status[months, factories], Bin)
+    @variable(model, production[months, factories], Int)
+    @variable(model, unmet_demand[months] >= 0)
+    ## We use `eachrow` to loop through the rows of the dataframe and add the
+    ## relevant constraints.
+    for r in eachrow(factory_df)
+        m, f = r.month, r.factory
+        @constraint(model, production[m, f] <= r.max_production * status[m, f])
+        @constraint(model, production[m, f] >= r.min_production * status[m, f])
+    end
+    @constraint(
+        model,
+        [r in eachrow(demand_df)],
+        sum(production[r.month, :]) + unmet_demand[r.month] == r.demand,
+    )
+    @objective(
+        model,
+        Min,
+        10_000 * sum(unmet_demand) + sum(
+            r.fixed_cost * status[r.month, r.factory] +
+            r.variable_cost * production[r.month, r.factory] for
+            r in eachrow(factory_df)
+        )
+    )
+    optimize!(model)
+    schedules = Dict{Symbol,Vector{Float64}}(
+        Symbol(f) => value.(production[:, f]) for f in factories
+    )
+    schedules[:unmet_demand] = value.(unmet_demand)
+    return (
+        termination_status = termination_status(model),
+        cost = objective_value(model),
+        ## This `select` statement re-orders the columns in the DataFrame.
+        schedules = DataFrames.select(
+            DataFrames.DataFrame(schedules),
+            [:unmet_demand, :A, :B],
+        ),
+    )
+end
+
+# ## Solution
+
+# Now we can call our `solve_factory_scheduling` function using the data we read
+# in above.
+
+solution = solve_factory_scheduling(demand_df, factory_df);
+
+Test.@test solution.termination_status == OPTIMAL  #src
+Test.@test solution.cost == 12_906_400.0  #src
+
+# Let's see what `solution` contains:
+
+solution.termination_status
+
+#-
+
+solution.cost
+
+#-
+
+solution.schedules
+
+# These schedules will be easier to visualize as a graph:
+
+StatsPlots.groupedbar(
+    Matrix(solution.schedules);
+    bar_position = :stack,
+    labels = ["unmet demand" "A" "B"],
+    xlabel = "Month",
+    ylabel = "Production",
+    legend = :topleft,
+    color = ["#20326c" "#4063d8" "#a0b1ec"],
+)
+
+# Note that we don't have any unmet demand.
+
+# ## What happens if demand increases?
+
+# Let's run an experiment by increasing the demand by 50% in all time periods:
+
+demand_df.demand .*= 1.5
+
+# Now we resolve the problem:
+
+high_demand_solution = solve_factory_scheduling(demand_df, factory_df);
+
+# and visualize the solution:
+
+StatsPlots.groupedbar(
+    Matrix(high_demand_solution.schedules);
+    bar_position = :stack,
+    labels = ["unmet demand" "A" "B"],
+    xlabel = "Month",
+    ylabel = "Production",
+    legend = :topleft,
+    color = ["#20326c" "#4063d8" "#a0b1ec"],
+)
+
+# Uh oh, we can't satisfy all of the demand.
+
+# ## How sensitive is the solution to changes in variable cost?
+
+# Let's run another experiment, this time seeing how the optimal objective
+# value changes as we vary the variable costs of each factory.
+
+# First though, let's reset the demand to it's original level:
+
+demand_df.demand ./= 1.5;
+
+# For our experiment, we're going to scale the variable costs of both factories
+# by a set of values from `0.0` to `1.5`:
+
+scale_factors = 0:0.1:1.5
+
+# At a high level, we're going to loop over the scale factors for A, then the
+# scale factors for B, rescale the input data, call our
+# `solve_factory_scheduling` example, and then store the optimal objective
+# value in the following `cost` matrix:
+
+cost = zeros(length(scale_factors), length(scale_factors));
+
+# Because we're modifying `factory_df` in-place, we need to store the original
+# variable costs in a new column:
+
+factory_df[!, :old_variable_cost] = copy(factory_df.variable_cost);
+
+# Then, we need a function to scale the `:variable_cost` column for a particular
+# factory by a value `scale`:
+
+function scale_variable_cost(df, factory, scale)
+    rows = df.factory .== factory
+    df[rows, :variable_cost] .=
+        round.(Int, df[rows, :old_variable_cost] .* scale)
+    return
+end
+
+# Our experiment is just a nested for-loop, modifying A and B and storing the
+# cost:
+
+for (j, a) in enumerate(scale_factors)
+    scale_variable_cost(factory_df, "A", a)
+    for (i, b) in enumerate(scale_factors)
+        scale_variable_cost(factory_df, "B", b)
+        cost[i, j] = solve_factory_scheduling(demand_df, factory_df).cost
+    end
+end
+
+# Let's visualize the cost matrix:
+
+StatsPlots.contour(
+    scale_factors,
+    scale_factors,
+    cost;
+    xlabel = "Scale of factory A",
+    ylabel = "Scale of factory B",
+)
+
+# What can you infer from the solution?
+
+# !!! info
+#     The [Power Systems](@ref) tutorial explains a number of other ways you can
+#     structure a problem to perform a parametric analysis of the solution. In
+#     particular, you can use in-place modification to reduce the time it takes
+#     to build and solve the resulting models.
diff --git a/previews/PR3525/tutorials/linear/factory_schedule/34a51857.svg b/previews/PR3525/tutorials/linear/factory_schedule/34a51857.svg
new file mode 100644
index 00000000000..69f1bc88b00
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/factory_schedule/34a51857.svg
@@ -0,0 +1,160 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip270">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip270)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip271">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip270)" d="M337.247 1423.18 L2352.76 1423.18 L2352.76 47.2441 L337.247 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip272">
+    <rect x="337" y="47" width="2017" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="356.894,1423.18 356.894,47.2441 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="660.927,1423.18 660.927,47.2441 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="964.96,1423.18 964.96,47.2441 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1268.99,1423.18 1268.99,47.2441 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1573.03,1423.18 1573.03,47.2441 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1877.06,1423.18 1877.06,47.2441 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2181.09,1423.18 2181.09,47.2441 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="337.247,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="356.894,1423.18 356.894,1404.28 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="660.927,1423.18 660.927,1404.28 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="964.96,1423.18 964.96,1404.28 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1268.99,1423.18 1268.99,1404.28 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1573.03,1423.18 1573.03,1404.28 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1877.06,1423.18 1877.06,1404.28 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2181.09,1423.18 2181.09,1404.28 "/>
+<path clip-path="url(#clip270)" d="M356.894 1454.1 Q353.283 1454.1 351.454 1457.66 Q349.648 1461.2 349.648 1468.33 Q349.648 1475.44 351.454 1479.01 Q353.283 1482.55 356.894 1482.55 Q360.528 1482.55 362.333 1479.01 Q364.162 1475.44 364.162 1468.33 Q364.162 1461.2 362.333 1457.66 Q360.528 1454.1 356.894 1454.1 M356.894 1450.39 Q362.704 1450.39 365.759 1455 Q368.838 1459.58 368.838 1468.33 Q368.838 1477.06 365.759 1481.67 Q362.704 1486.25 356.894 1486.25 Q351.083 1486.25 348.005 1481.67 Q344.949 1477.06 344.949 1468.33 Q344.949 1459.58 348.005 1455 Q351.083 1450.39 356.894 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M655.58 1481.64 L671.899 1481.64 L671.899 1485.58 L649.955 1485.58 L649.955 1481.64 Q652.617 1478.89 657.2 1474.26 Q661.806 1469.61 662.987 1468.27 Q665.232 1465.74 666.112 1464.01 Q667.015 1462.25 667.015 1460.56 Q667.015 1457.8 665.07 1456.07 Q663.149 1454.33 660.047 1454.33 Q657.848 1454.33 655.394 1455.09 Q652.964 1455.86 650.186 1457.41 L650.186 1452.69 Q653.01 1451.55 655.464 1450.97 Q657.918 1450.39 659.955 1450.39 Q665.325 1450.39 668.519 1453.08 Q671.714 1455.77 671.714 1460.26 Q671.714 1462.39 670.904 1464.31 Q670.117 1466.2 668.01 1468.8 Q667.431 1469.47 664.33 1472.69 Q661.228 1475.88 655.58 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M967.969 1455.09 L956.164 1473.54 L967.969 1473.54 L967.969 1455.09 M966.742 1451.02 L972.622 1451.02 L972.622 1473.54 L977.553 1473.54 L977.553 1477.43 L972.622 1477.43 L972.622 1485.58 L967.969 1485.58 L967.969 1477.43 L952.367 1477.43 L952.367 1472.92 L966.742 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1269.4 1466.44 Q1266.25 1466.44 1264.4 1468.59 Q1262.57 1470.74 1262.57 1474.49 Q1262.57 1478.22 1264.4 1480.39 Q1266.25 1482.55 1269.4 1482.55 Q1272.55 1482.55 1274.38 1480.39 Q1276.23 1478.22 1276.23 1474.49 Q1276.23 1470.74 1274.38 1468.59 Q1272.55 1466.44 1269.4 1466.44 M1278.68 1451.78 L1278.68 1456.04 Q1276.92 1455.21 1275.12 1454.77 Q1273.33 1454.33 1271.57 1454.33 Q1266.94 1454.33 1264.49 1457.45 Q1262.06 1460.58 1261.71 1466.9 Q1263.08 1464.89 1265.14 1463.82 Q1267.2 1462.73 1269.68 1462.73 Q1274.88 1462.73 1277.89 1465.9 Q1280.93 1469.05 1280.93 1474.49 Q1280.93 1479.82 1277.78 1483.03 Q1274.63 1486.25 1269.4 1486.25 Q1263.4 1486.25 1260.23 1481.67 Q1257.06 1477.06 1257.06 1468.33 Q1257.06 1460.14 1260.95 1455.28 Q1264.84 1450.39 1271.39 1450.39 Q1273.15 1450.39 1274.93 1450.74 Q1276.74 1451.09 1278.68 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1573.03 1469.17 Q1569.69 1469.17 1567.77 1470.95 Q1565.87 1472.73 1565.87 1475.86 Q1565.87 1478.98 1567.77 1480.77 Q1569.69 1482.55 1573.03 1482.55 Q1576.36 1482.55 1578.28 1480.77 Q1580.2 1478.96 1580.2 1475.86 Q1580.2 1472.73 1578.28 1470.95 Q1576.38 1469.17 1573.03 1469.17 M1568.35 1467.18 Q1565.34 1466.44 1563.65 1464.38 Q1561.98 1462.32 1561.98 1459.35 Q1561.98 1455.21 1564.92 1452.8 Q1567.89 1450.39 1573.03 1450.39 Q1578.19 1450.39 1581.13 1452.8 Q1584.07 1455.21 1584.07 1459.35 Q1584.07 1462.32 1582.38 1464.38 Q1580.71 1466.44 1577.73 1467.18 Q1581.1 1467.96 1582.98 1470.26 Q1584.88 1472.55 1584.88 1475.86 Q1584.88 1480.88 1581.8 1483.57 Q1578.74 1486.25 1573.03 1486.25 Q1567.31 1486.25 1564.23 1483.57 Q1561.17 1480.88 1561.17 1475.86 Q1561.17 1472.55 1563.07 1470.26 Q1564.97 1467.96 1568.35 1467.18 M1566.64 1459.79 Q1566.64 1462.48 1568.3 1463.98 Q1569.99 1465.49 1573.03 1465.49 Q1576.04 1465.49 1577.73 1463.98 Q1579.44 1462.48 1579.44 1459.79 Q1579.44 1457.11 1577.73 1455.6 Q1576.04 1454.1 1573.03 1454.1 Q1569.99 1454.1 1568.3 1455.6 Q1566.64 1457.11 1566.64 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1851.75 1481.64 L1859.39 1481.64 L1859.39 1455.28 L1851.08 1456.95 L1851.08 1452.69 L1859.34 1451.02 L1864.02 1451.02 L1864.02 1481.64 L1871.65 1481.64 L1871.65 1485.58 L1851.75 1485.58 L1851.75 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1891.1 1454.1 Q1887.49 1454.1 1885.66 1457.66 Q1883.85 1461.2 1883.85 1468.33 Q1883.85 1475.44 1885.66 1479.01 Q1887.49 1482.55 1891.1 1482.55 Q1894.73 1482.55 1896.54 1479.01 Q1898.37 1475.44 1898.37 1468.33 Q1898.37 1461.2 1896.54 1457.66 Q1894.73 1454.1 1891.1 1454.1 M1891.1 1450.39 Q1896.91 1450.39 1899.96 1455 Q1903.04 1459.58 1903.04 1468.33 Q1903.04 1477.06 1899.96 1481.67 Q1896.91 1486.25 1891.1 1486.25 Q1885.29 1486.25 1882.21 1481.67 Q1879.15 1477.06 1879.15 1468.33 Q1879.15 1459.58 1882.21 1455 Q1885.29 1450.39 1891.1 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M2156.58 1481.64 L2164.22 1481.64 L2164.22 1455.28 L2155.91 1456.95 L2155.91 1452.69 L2164.17 1451.02 L2168.85 1451.02 L2168.85 1481.64 L2176.49 1481.64 L2176.49 1485.58 L2156.58 1485.58 L2156.58 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M2189.96 1481.64 L2206.28 1481.64 L2206.28 1485.58 L2184.33 1485.58 L2184.33 1481.64 Q2187 1478.89 2191.58 1474.26 Q2196.19 1469.61 2197.37 1468.27 Q2199.61 1465.74 2200.49 1464.01 Q2201.39 1462.25 2201.39 1460.56 Q2201.39 1457.8 2199.45 1456.07 Q2197.53 1454.33 2194.43 1454.33 Q2192.23 1454.33 2189.77 1455.09 Q2187.34 1455.86 2184.56 1457.41 L2184.56 1452.69 Q2187.39 1451.55 2189.84 1450.97 Q2192.3 1450.39 2194.33 1450.39 Q2199.7 1450.39 2202.9 1453.08 Q2206.09 1455.77 2206.09 1460.26 Q2206.09 1462.39 2205.28 1464.31 Q2204.5 1466.2 2202.39 1468.8 Q2201.81 1469.47 2198.71 1472.69 Q2195.61 1475.88 2189.96 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1248.82 1520.52 L1258.4 1520.52 L1270.52 1552.86 L1282.71 1520.52 L1292.29 1520.52 L1292.29 1568.04 L1286.02 1568.04 L1286.02 1526.32 L1273.77 1558.91 L1267.31 1558.91 L1255.05 1526.32 L1255.05 1568.04 L1248.82 1568.04 L1248.82 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1318.62 1536.5 Q1313.9 1536.5 1311.17 1540.19 Q1308.43 1543.85 1308.43 1550.25 Q1308.43 1556.65 1311.14 1560.34 Q1313.87 1564 1318.62 1564 Q1323.29 1564 1326.03 1560.31 Q1328.77 1556.62 1328.77 1550.25 Q1328.77 1543.92 1326.03 1540.23 Q1323.29 1536.5 1318.62 1536.5 M1318.62 1531.54 Q1326.25 1531.54 1330.61 1536.5 Q1334.98 1541.47 1334.98 1550.25 Q1334.98 1559 1330.61 1564 Q1326.25 1568.97 1318.62 1568.97 Q1310.94 1568.97 1306.58 1564 Q1302.26 1559 1302.26 1550.25 Q1302.26 1541.47 1306.58 1536.5 Q1310.94 1531.54 1318.62 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1374.32 1546.53 L1374.32 1568.04 L1368.46 1568.04 L1368.46 1546.72 Q1368.46 1541.66 1366.49 1539.14 Q1364.51 1536.63 1360.57 1536.63 Q1355.82 1536.63 1353.09 1539.65 Q1350.35 1542.68 1350.35 1547.9 L1350.35 1568.04 L1344.46 1568.04 L1344.46 1532.4 L1350.35 1532.4 L1350.35 1537.93 Q1352.45 1534.72 1355.28 1533.13 Q1358.15 1531.54 1361.87 1531.54 Q1368.01 1531.54 1371.16 1535.36 Q1374.32 1539.14 1374.32 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1391.79 1522.27 L1391.79 1532.4 L1403.85 1532.4 L1403.85 1536.95 L1391.79 1536.95 L1391.79 1556.3 Q1391.79 1560.66 1392.97 1561.9 Q1394.18 1563.14 1397.84 1563.14 L1403.85 1563.14 L1403.85 1568.04 L1397.84 1568.04 Q1391.06 1568.04 1388.48 1565.53 Q1385.9 1562.98 1385.9 1556.3 L1385.9 1536.95 L1381.6 1536.95 L1381.6 1532.4 L1385.9 1532.4 L1385.9 1522.27 L1391.79 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M1441.19 1546.53 L1441.19 1568.04 L1435.33 1568.04 L1435.33 1546.72 Q1435.33 1541.66 1433.36 1539.14 Q1431.38 1536.63 1427.44 1536.63 Q1422.69 1536.63 1419.96 1539.65 Q1417.22 1542.68 1417.22 1547.9 L1417.22 1568.04 L1411.33 1568.04 L1411.33 1518.52 L1417.22 1518.52 L1417.22 1537.93 Q1419.32 1534.72 1422.15 1533.13 Q1425.02 1531.54 1428.74 1531.54 Q1434.89 1531.54 1438.04 1535.36 Q1441.19 1539.14 1441.19 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="337.247,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="337.247,964.535 2352.76,964.535 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="337.247,505.889 2352.76,505.889 "/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="337.247,47.2441 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="337.247,1423.18 337.247,47.2441 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="337.247,1423.18 356.145,1423.18 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="337.247,964.535 356.145,964.535 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="337.247,505.889 356.145,505.889 "/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="337.247,47.2441 356.145,47.2441 "/>
+<path clip-path="url(#clip270)" d="M289.303 1408.98 Q285.692 1408.98 283.863 1412.54 Q282.057 1416.08 282.057 1423.21 Q282.057 1430.32 283.863 1433.89 Q285.692 1437.43 289.303 1437.43 Q292.937 1437.43 294.742 1433.89 Q296.571 1430.32 296.571 1423.21 Q296.571 1416.08 294.742 1412.54 Q292.937 1408.98 289.303 1408.98 M289.303 1405.27 Q295.113 1405.27 298.168 1409.88 Q301.247 1414.46 301.247 1423.21 Q301.247 1431.94 298.168 1436.55 Q295.113 1441.13 289.303 1441.13 Q283.492 1441.13 280.414 1436.55 Q277.358 1431.94 277.358 1423.21 Q277.358 1414.46 280.414 1409.88 Q283.492 1405.27 289.303 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M116.668 984.327 L124.306 984.327 L124.306 957.961 L115.996 959.628 L115.996 955.369 L124.26 953.702 L128.936 953.702 L128.936 984.327 L136.575 984.327 L136.575 988.262 L116.668 988.262 L116.668 984.327 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M146.019 982.383 L150.903 982.383 L150.903 988.262 L146.019 988.262 L146.019 982.383 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M171.089 956.781 Q167.477 956.781 165.649 960.346 Q163.843 963.887 163.843 971.017 Q163.843 978.123 165.649 981.688 Q167.477 985.23 171.089 985.23 Q174.723 985.23 176.528 981.688 Q178.357 978.123 178.357 971.017 Q178.357 963.887 176.528 960.346 Q174.723 956.781 171.089 956.781 M171.089 953.077 Q176.899 953.077 179.954 957.684 Q183.033 962.267 183.033 971.017 Q183.033 979.744 179.954 984.35 Q176.899 988.933 171.089 988.933 Q165.278 988.933 162.2 984.35 Q159.144 979.744 159.144 971.017 Q159.144 962.267 162.2 957.684 Q165.278 953.077 171.089 953.077 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M219.422 962.799 L208.843 973.424 L219.422 984.003 L216.667 986.804 L206.042 976.179 L195.417 986.804 L192.686 984.003 L203.241 973.424 L192.686 962.799 L195.417 959.998 L206.042 970.623 L216.667 959.998 L219.422 962.799 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M231.783 984.327 L239.422 984.327 L239.422 957.961 L231.111 959.628 L231.111 955.369 L239.375 953.702 L244.051 953.702 L244.051 984.327 L251.69 984.327 L251.69 988.262 L231.783 988.262 L231.783 984.327 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M271.134 956.781 Q267.523 956.781 265.695 960.346 Q263.889 963.887 263.889 971.017 Q263.889 978.123 265.695 981.688 Q267.523 985.23 271.134 985.23 Q274.769 985.23 276.574 981.688 Q278.403 978.123 278.403 971.017 Q278.403 963.887 276.574 960.346 Q274.769 956.781 271.134 956.781 M271.134 953.077 Q276.944 953.077 280 957.684 Q283.079 962.267 283.079 971.017 Q283.079 979.744 280 984.35 Q276.944 988.933 271.134 988.933 Q265.324 988.933 262.245 984.35 Q259.19 979.744 259.19 971.017 Q259.19 962.267 262.245 957.684 Q265.324 953.077 271.134 953.077 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M284.264 932.772 L299.178 932.772 L299.178 935.969 L287.743 935.969 L287.743 942.853 Q288.571 942.571 289.398 942.439 Q290.226 942.289 291.053 942.289 Q295.755 942.289 298.501 944.865 Q301.247 947.442 301.247 951.843 Q301.247 956.376 298.426 958.896 Q295.605 961.397 290.47 961.397 Q288.702 961.397 286.859 961.096 Q285.035 960.795 283.079 960.194 L283.079 956.376 Q284.771 957.297 286.577 957.749 Q288.382 958.2 290.395 958.2 Q293.649 958.2 295.548 956.488 Q297.448 954.777 297.448 951.843 Q297.448 948.909 295.548 947.197 Q293.649 945.486 290.395 945.486 Q288.871 945.486 287.348 945.824 Q285.843 946.163 284.264 946.878 L284.264 932.772 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M119.885 525.682 L136.204 525.682 L136.204 529.617 L114.26 529.617 L114.26 525.682 Q116.922 522.927 121.505 518.298 Q126.112 513.645 127.292 512.302 Q129.538 509.779 130.417 508.043 Q131.32 506.284 131.32 504.594 Q131.32 501.839 129.376 500.103 Q127.455 498.367 124.353 498.367 Q122.154 498.367 119.7 499.131 Q117.269 499.895 114.492 501.446 L114.492 496.724 Q117.316 495.589 119.769 495.011 Q122.223 494.432 124.26 494.432 Q129.63 494.432 132.825 497.117 Q136.019 499.802 136.019 504.293 Q136.019 506.423 135.209 508.344 Q134.422 510.242 132.316 512.835 Q131.737 513.506 128.635 516.723 Q125.533 519.918 119.885 525.682 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M146.019 523.737 L150.903 523.737 L150.903 529.617 L146.019 529.617 L146.019 523.737 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M171.089 498.136 Q167.477 498.136 165.649 501.7 Q163.843 505.242 163.843 512.372 Q163.843 519.478 165.649 523.043 Q167.477 526.585 171.089 526.585 Q174.723 526.585 176.528 523.043 Q178.357 519.478 178.357 512.372 Q178.357 505.242 176.528 501.7 Q174.723 498.136 171.089 498.136 M171.089 494.432 Q176.899 494.432 179.954 499.038 Q183.033 503.622 183.033 512.372 Q183.033 521.098 179.954 525.705 Q176.899 530.288 171.089 530.288 Q165.278 530.288 162.2 525.705 Q159.144 521.098 159.144 512.372 Q159.144 503.622 162.2 499.038 Q165.278 494.432 171.089 494.432 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M219.422 504.154 L208.843 514.779 L219.422 525.358 L216.667 528.159 L206.042 517.534 L195.417 528.159 L192.686 525.358 L203.241 514.779 L192.686 504.154 L195.417 501.353 L206.042 511.978 L216.667 501.353 L219.422 504.154 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M231.783 525.682 L239.422 525.682 L239.422 499.316 L231.111 500.983 L231.111 496.724 L239.375 495.057 L244.051 495.057 L244.051 525.682 L251.69 525.682 L251.69 529.617 L231.783 529.617 L231.783 525.682 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M271.134 498.136 Q267.523 498.136 265.695 501.7 Q263.889 505.242 263.889 512.372 Q263.889 519.478 265.695 523.043 Q267.523 526.585 271.134 526.585 Q274.769 526.585 276.574 523.043 Q278.403 519.478 278.403 512.372 Q278.403 505.242 276.574 501.7 Q274.769 498.136 271.134 498.136 M271.134 494.432 Q276.944 494.432 280 499.038 Q283.079 503.622 283.079 512.372 Q283.079 521.098 280 525.705 Q276.944 530.288 271.134 530.288 Q265.324 530.288 262.245 525.705 Q259.19 521.098 259.19 512.372 Q259.19 503.622 262.245 499.038 Q265.324 494.432 271.134 494.432 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M284.264 474.127 L299.178 474.127 L299.178 477.324 L287.743 477.324 L287.743 484.208 Q288.571 483.925 289.398 483.794 Q290.226 483.643 291.053 483.643 Q295.755 483.643 298.501 486.22 Q301.247 488.797 301.247 493.198 Q301.247 497.73 298.426 500.251 Q295.605 502.752 290.47 502.752 Q288.702 502.752 286.859 502.451 Q285.035 502.15 283.079 501.548 L283.079 497.73 Q284.771 498.652 286.577 499.103 Q288.382 499.555 290.395 499.555 Q293.649 499.555 295.548 497.843 Q297.448 496.132 297.448 493.198 Q297.448 490.264 295.548 488.552 Q293.649 486.841 290.395 486.841 Q288.871 486.841 287.348 487.179 Q285.843 487.518 284.264 488.232 L284.264 474.127 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M130.024 52.3375 Q133.38 53.0551 135.255 55.3236 Q137.154 57.5921 137.154 60.9255 Q137.154 66.0412 133.635 68.8421 Q130.117 71.643 123.635 71.643 Q121.459 71.643 119.144 71.2032 Q116.853 70.7865 114.399 69.93 L114.399 65.4162 Q116.343 66.5504 118.658 67.1291 Q120.973 67.7078 123.496 67.7078 Q127.894 67.7078 130.186 65.9717 Q132.501 64.2356 132.501 60.9255 Q132.501 57.8699 130.348 56.157 Q128.218 54.4209 124.399 54.4209 L120.371 54.4209 L120.371 50.5783 L124.584 50.5783 Q128.033 50.5783 129.862 49.2126 Q131.691 47.8237 131.691 45.2311 Q131.691 42.5691 129.792 41.157 Q127.917 39.7219 124.399 39.7219 Q122.478 39.7219 120.279 40.1385 Q118.08 40.5552 115.441 41.4348 L115.441 37.2682 Q118.103 36.5274 120.418 36.1571 Q122.755 35.7867 124.816 35.7867 Q130.14 35.7867 133.242 38.2172 Q136.343 40.6246 136.343 44.745 Q136.343 47.6153 134.7 49.6061 Q133.056 51.5737 130.024 52.3375 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M146.019 65.0921 L150.903 65.0921 L150.903 70.9717 L146.019 70.9717 L146.019 65.0921 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M171.089 39.4904 Q167.477 39.4904 165.649 43.0552 Q163.843 46.5968 163.843 53.7264 Q163.843 60.8329 165.649 64.3977 Q167.477 67.9393 171.089 67.9393 Q174.723 67.9393 176.528 64.3977 Q178.357 60.8329 178.357 53.7264 Q178.357 46.5968 176.528 43.0552 Q174.723 39.4904 171.089 39.4904 M171.089 35.7867 Q176.899 35.7867 179.954 40.3932 Q183.033 44.9765 183.033 53.7264 Q183.033 62.4532 179.954 67.0597 Q176.899 71.643 171.089 71.643 Q165.278 71.643 162.2 67.0597 Q159.144 62.4532 159.144 53.7264 Q159.144 44.9765 162.2 40.3932 Q165.278 35.7867 171.089 35.7867 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M219.422 45.5089 L208.843 56.1338 L219.422 66.7125 L216.667 69.5134 L206.042 58.8884 L195.417 69.5134 L192.686 66.7125 L203.241 56.1338 L192.686 45.5089 L195.417 42.708 L206.042 53.3329 L216.667 42.708 L219.422 45.5089 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M231.783 67.0365 L239.422 67.0365 L239.422 40.6709 L231.111 42.3376 L231.111 38.0784 L239.375 36.4117 L244.051 36.4117 L244.051 67.0365 L251.69 67.0365 L251.69 70.9717 L231.783 70.9717 L231.783 67.0365 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M271.134 39.4904 Q267.523 39.4904 265.695 43.0552 Q263.889 46.5968 263.889 53.7264 Q263.889 60.8329 265.695 64.3977 Q267.523 67.9393 271.134 67.9393 Q274.769 67.9393 276.574 64.3977 Q278.403 60.8329 278.403 53.7264 Q278.403 46.5968 276.574 43.0552 Q274.769 39.4904 271.134 39.4904 M271.134 35.7867 Q276.944 35.7867 280 40.3932 Q283.079 44.9765 283.079 53.7264 Q283.079 62.4532 280 67.0597 Q276.944 71.643 271.134 71.643 Q265.324 71.643 262.245 67.0597 Q259.19 62.4532 259.19 53.7264 Q259.19 44.9765 262.245 40.3932 Q265.324 35.7867 271.134 35.7867 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M284.264 15.4813 L299.178 15.4813 L299.178 18.6786 L287.743 18.6786 L287.743 25.5623 Q288.571 25.2801 289.398 25.1485 Q290.226 24.998 291.053 24.998 Q295.755 24.998 298.501 27.5747 Q301.247 30.1514 301.247 34.5524 Q301.247 39.085 298.426 41.6053 Q295.605 44.1067 290.47 44.1067 Q288.702 44.1067 286.859 43.8058 Q285.035 43.5049 283.079 42.903 L283.079 39.085 Q284.771 40.0066 286.577 40.458 Q288.382 40.9094 290.395 40.9094 Q293.649 40.9094 295.548 39.1979 Q297.448 37.4864 297.448 34.5524 Q297.448 31.6184 295.548 29.9069 Q293.649 28.1953 290.395 28.1953 Q288.871 28.1953 287.348 28.5339 Q285.843 28.8724 284.264 29.5871 L284.264 15.4813 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M21.7677 896.217 L39.6235 896.217 L39.6235 888.132 Q39.6235 883.644 37.3 881.194 Q34.9765 878.743 30.6797 878.743 Q26.4147 878.743 24.0912 881.194 Q21.7677 883.644 21.7677 888.132 L21.7677 896.217 M16.4842 902.646 L16.4842 888.132 Q16.4842 880.143 20.1126 876.069 Q23.7092 871.963 30.6797 871.963 Q37.7138 871.963 41.3104 876.069 Q44.907 880.143 44.907 888.132 L44.907 896.217 L64.0042 896.217 L64.0042 902.646 L16.4842 902.646 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M33.8307 844.081 Q33.2578 845.068 33.0032 846.246 Q32.7167 847.392 32.7167 848.792 Q32.7167 853.757 35.9632 856.431 Q39.1779 859.073 45.2253 859.073 L64.0042 859.073 L64.0042 864.961 L28.3562 864.961 L28.3562 859.073 L33.8944 859.073 Q30.6479 857.227 29.0883 854.267 Q27.4968 851.307 27.4968 847.073 Q27.4968 846.469 27.5923 845.737 Q27.656 845.004 27.8151 844.113 L33.8307 844.081 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M32.4621 825.557 Q32.4621 830.268 36.1542 833.005 Q39.8145 835.742 46.212 835.742 Q52.6095 835.742 56.3017 833.037 Q59.9619 830.3 59.9619 825.557 Q59.9619 820.878 56.2698 818.141 Q52.5777 815.404 46.212 815.404 Q39.8781 815.404 36.186 818.141 Q32.4621 820.878 32.4621 825.557 M27.4968 825.557 Q27.4968 817.918 32.4621 813.558 Q37.4273 809.197 46.212 809.197 Q54.9649 809.197 59.9619 813.558 Q64.9272 817.918 64.9272 825.557 Q64.9272 833.228 59.9619 837.588 Q54.9649 841.917 46.212 841.917 Q37.4273 841.917 32.4621 837.588 Q27.4968 833.228 27.4968 825.557 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M33.7671 776.032 L14.479 776.032 L14.479 770.176 L64.0042 770.176 L64.0042 776.032 L58.657 776.032 Q61.8398 777.878 63.3994 780.711 Q64.9272 783.512 64.9272 787.459 Q64.9272 793.92 59.771 797.994 Q54.6147 802.036 46.212 802.036 Q37.8093 802.036 32.6531 797.994 Q27.4968 793.92 27.4968 787.459 Q27.4968 783.512 29.0564 780.711 Q30.5842 777.878 33.7671 776.032 M46.212 795.989 Q52.6732 795.989 56.3653 793.347 Q60.0256 790.673 60.0256 786.026 Q60.0256 781.379 56.3653 778.706 Q52.6732 776.032 46.212 776.032 Q39.7508 776.032 36.0905 778.706 Q32.3984 781.379 32.3984 786.026 Q32.3984 790.673 36.0905 793.347 Q39.7508 795.989 46.212 795.989 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M49.9359 758.717 L28.3562 758.717 L28.3562 752.861 L49.7131 752.861 Q54.7739 752.861 57.3202 750.888 Q59.8346 748.914 59.8346 744.967 Q59.8346 740.225 56.8109 737.488 Q53.7872 734.719 48.5673 734.719 L28.3562 734.719 L28.3562 728.862 L64.0042 728.862 L64.0042 734.719 L58.5296 734.719 Q61.7762 736.851 63.3676 739.684 Q64.9272 742.485 64.9272 746.209 Q64.9272 752.352 61.1078 755.534 Q57.2883 758.717 49.9359 758.717 M27.4968 743.981 L27.4968 743.981 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M29.7248 691.145 L35.1993 691.145 Q33.8307 693.628 33.1623 696.142 Q32.4621 698.625 32.4621 701.171 Q32.4621 706.869 36.0905 710.02 Q39.6872 713.171 46.212 713.171 Q52.7369 713.171 56.3653 710.02 Q59.9619 706.869 59.9619 701.171 Q59.9619 698.625 59.2935 696.142 Q58.5933 693.628 57.2247 691.145 L62.6355 691.145 Q63.7814 693.596 64.3543 696.238 Q64.9272 698.848 64.9272 701.808 Q64.9272 709.861 59.8664 714.603 Q54.8057 719.345 46.212 719.345 Q37.491 719.345 32.4939 714.571 Q27.4968 709.765 27.4968 701.426 Q27.4968 698.721 28.0697 696.142 Q28.6108 693.564 29.7248 691.145 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M18.2347 675.167 L28.3562 675.167 L28.3562 663.104 L32.9077 663.104 L32.9077 675.167 L52.2594 675.167 Q56.6199 675.167 57.8613 673.99 Q59.1026 672.78 59.1026 669.12 L59.1026 663.104 L64.0042 663.104 L64.0042 669.12 Q64.0042 675.899 61.4897 678.478 Q58.9434 681.056 52.2594 681.056 L32.9077 681.056 L32.9077 685.353 L28.3562 685.353 L28.3562 681.056 L18.2347 681.056 L18.2347 675.167 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M28.3562 655.402 L28.3562 649.545 L64.0042 649.545 L64.0042 655.402 L28.3562 655.402 M14.479 655.402 L14.479 649.545 L21.895 649.545 L21.895 655.402 L14.479 655.402 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M32.4621 623.478 Q32.4621 628.189 36.1542 630.926 Q39.8145 633.663 46.212 633.663 Q52.6095 633.663 56.3017 630.958 Q59.9619 628.22 59.9619 623.478 Q59.9619 618.799 56.2698 616.062 Q52.5777 613.325 46.212 613.325 Q39.8781 613.325 36.186 616.062 Q32.4621 618.799 32.4621 623.478 M27.4968 623.478 Q27.4968 615.839 32.4621 611.479 Q37.4273 607.118 46.212 607.118 Q54.9649 607.118 59.9619 611.479 Q64.9272 615.839 64.9272 623.478 Q64.9272 631.149 59.9619 635.509 Q54.9649 639.838 46.212 639.838 Q37.4273 639.838 32.4621 635.509 Q27.4968 631.149 27.4968 623.478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M42.4881 567.778 L64.0042 567.778 L64.0042 573.634 L42.679 573.634 Q37.6183 573.634 35.1038 575.608 Q32.5894 577.581 32.5894 581.528 Q32.5894 586.27 35.6131 589.008 Q38.6368 591.745 43.8567 591.745 L64.0042 591.745 L64.0042 597.633 L28.3562 597.633 L28.3562 591.745 L33.8944 591.745 Q30.6797 589.644 29.0883 586.811 Q27.4968 583.947 27.4968 580.223 Q27.4968 574.08 31.3163 570.929 Q35.1038 567.778 42.4881 567.778 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip272)" d="M448.104 597.618 L448.104 735.212 L569.717 735.212 L569.717 597.618 L448.104 597.618 L448.104 597.618  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="448.104,597.618 448.104,735.212 569.717,735.212 569.717,597.618 448.104,597.618 "/>
+<path clip-path="url(#clip272)" d="M600.12 735.212 L600.12 735.212 L721.733 735.212 L721.733 735.212 L600.12 735.212 L600.12 735.212  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="600.12,735.212 600.12,735.212 721.733,735.212 600.12,735.212 "/>
+<path clip-path="url(#clip272)" d="M752.137 528.822 L752.137 597.618 L873.75 597.618 L873.75 528.822 L752.137 528.822 L752.137 528.822  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="752.137,528.822 752.137,597.618 873.75,597.618 873.75,528.822 752.137,528.822 "/>
+<path clip-path="url(#clip272)" d="M904.153 528.822 L904.153 528.822 L1025.77 528.822 L1025.77 528.822 L904.153 528.822 L904.153 528.822  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="904.153,528.822 904.153,528.822 1025.77,528.822 904.153,528.822 "/>
+<path clip-path="url(#clip272)" d="M1056.17 460.025 L1056.17 689.347 L1177.78 689.347 L1177.78 460.025 L1056.17 460.025 L1056.17 460.025  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1056.17,460.025 1056.17,689.347 1177.78,689.347 1177.78,460.025 1056.17,460.025 "/>
+<path clip-path="url(#clip272)" d="M1208.19 528.822 L1208.19 528.822 L1329.8 528.822 L1329.8 528.822 L1208.19 528.822 L1208.19 528.822  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1208.19,528.822 1208.19,528.822 1329.8,528.822 1208.19,528.822 "/>
+<path clip-path="url(#clip272)" d="M1360.2 391.228 L1360.2 437.093 L1481.82 437.093 L1481.82 391.228 L1360.2 391.228 L1360.2 391.228  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1360.2,391.228 1360.2,437.093 1481.82,437.093 1481.82,391.228 1360.2,391.228 "/>
+<path clip-path="url(#clip272)" d="M1512.22 253.634 L1512.22 253.634 L1633.83 253.634 L1633.83 253.634 L1512.22 253.634 L1512.22 253.634  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1512.22,253.634 1512.22,253.634 1633.83,253.634 1512.22,253.634 "/>
+<path clip-path="url(#clip272)" d="M1664.24 47.2441 L1664.24 47.2441 L1785.85 47.2441 L1785.85 47.2441 L1664.24 47.2441 L1664.24 47.2441  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1664.24,47.2441 1664.24,47.2441 1785.85,47.2441 1664.24,47.2441 "/>
+<path clip-path="url(#clip272)" d="M1816.25 116.041 L1816.25 116.041 L1937.87 116.041 L1937.87 116.041 L1816.25 116.041 L1816.25 116.041  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1816.25,116.041 1816.25,116.041 1937.87,116.041 1816.25,116.041 "/>
+<path clip-path="url(#clip272)" d="M1968.27 460.025 L1968.27 505.889 L2089.88 505.889 L2089.88 460.025 L1968.27 460.025 L1968.27 460.025  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1968.27,460.025 1968.27,505.889 2089.88,505.889 2089.88,460.025 1968.27,460.025 "/>
+<path clip-path="url(#clip272)" d="M2120.29 735.212 L2120.29 735.212 L2241.9 735.212 L2241.9 735.212 L2120.29 735.212 L2120.29 735.212  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.29,735.212 2120.29,735.212 2241.9,735.212 2120.29,735.212 "/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="508.91" cy="597.618" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="660.927" cy="735.212" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="812.943" cy="528.822" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="964.96" cy="528.822" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1116.98" cy="460.025" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1268.99" cy="528.822" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1421.01" cy="391.228" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1573.03" cy="253.634" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1725.04" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1877.06" cy="116.041" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="2029.08" cy="460.025" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="2181.09" cy="735.212" r="2"/>
+<path clip-path="url(#clip272)" d="M448.104 735.212 L448.104 1193.86 L569.717 1193.86 L569.717 735.212 L448.104 735.212 L448.104 735.212  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="448.104,735.212 448.104,1193.86 569.717,1193.86 569.717,735.212 448.104,735.212 "/>
+<path clip-path="url(#clip272)" d="M600.12 735.212 L600.12 1170.92 L721.733 1170.92 L721.733 735.212 L600.12 735.212 L600.12 735.212  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="600.12,735.212 600.12,1170.92 721.733,1170.92 721.733,735.212 600.12,735.212 "/>
+<path clip-path="url(#clip272)" d="M752.137 597.618 L752.137 1147.99 L873.75 1147.99 L873.75 597.618 L752.137 597.618 L752.137 597.618  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="752.137,597.618 752.137,1147.99 873.75,1147.99 873.75,597.618 752.137,597.618 "/>
+<path clip-path="url(#clip272)" d="M904.153 528.822 L904.153 964.535 L1025.77 964.535 L1025.77 528.822 L904.153 528.822 L904.153 528.822  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="904.153,528.822 904.153,964.535 1025.77,964.535 1025.77,528.822 904.153,528.822 "/>
+<path clip-path="url(#clip272)" d="M1056.17 689.347 L1056.17 1423.18 L1177.78 1423.18 L1177.78 689.347 L1056.17 689.347 L1056.17 689.347  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1056.17,689.347 1056.17,1423.18 1177.78,1423.18 1177.78,689.347 1056.17,689.347 "/>
+<path clip-path="url(#clip272)" d="M1208.19 528.822 L1208.19 1102.13 L1329.8 1102.13 L1329.8 528.822 L1208.19 528.822 L1208.19 528.822  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1208.19,528.822 1208.19,1102.13 1329.8,1102.13 1329.8,528.822 1208.19,528.822 "/>
+<path clip-path="url(#clip272)" d="M1360.2 437.093 L1360.2 1147.99 L1481.82 1147.99 L1481.82 437.093 L1360.2 437.093 L1360.2 437.093  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1360.2,437.093 1360.2,1147.99 1481.82,1147.99 1481.82,437.093 1360.2,437.093 "/>
+<path clip-path="url(#clip272)" d="M1512.22 253.634 L1512.22 964.535 L1633.83 964.535 L1633.83 253.634 L1512.22 253.634 L1512.22 253.634  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1512.22,253.634 1512.22,964.535 1633.83,964.535 1633.83,253.634 1512.22,253.634 "/>
+<path clip-path="url(#clip272)" d="M1664.24 47.2441 L1664.24 964.535 L1785.85 964.535 L1785.85 47.2441 L1664.24 47.2441 L1664.24 47.2441  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1664.24,47.2441 1664.24,964.535 1785.85,964.535 1785.85,47.2441 1664.24,47.2441 "/>
+<path clip-path="url(#clip272)" d="M1816.25 116.041 L1816.25 1019.57 L1937.87 1019.57 L1937.87 116.041 L1816.25 116.041 L1816.25 116.041  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1816.25,116.041 1816.25,1019.57 1937.87,1019.57 1937.87,116.041 1816.25,116.041 "/>
+<path clip-path="url(#clip272)" d="M1968.27 505.889 L1968.27 872.806 L2089.88 872.806 L2089.88 505.889 L1968.27 505.889 L1968.27 505.889  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1968.27,505.889 1968.27,872.806 2089.88,872.806 2089.88,505.889 1968.27,505.889 "/>
+<path clip-path="url(#clip272)" d="M2120.29 735.212 L2120.29 1423.18 L2241.9 1423.18 L2241.9 735.212 L2120.29 735.212 L2120.29 735.212  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.29,735.212 2120.29,1423.18 2241.9,1423.18 2241.9,735.212 2120.29,735.212 "/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="508.91" cy="735.212" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="660.927" cy="735.212" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="812.943" cy="597.618" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="964.96" cy="528.822" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1116.98" cy="689.347" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1268.99" cy="528.822" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1421.01" cy="437.093" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1573.03" cy="253.634" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1725.04" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1877.06" cy="116.041" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="2029.08" cy="505.889" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="2181.09" cy="735.212" r="2"/>
+<path clip-path="url(#clip272)" d="M448.104 1193.86 L448.104 1423.18 L569.717 1423.18 L569.717 1193.86 L448.104 1193.86 L448.104 1193.86  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="448.104,1193.86 448.104,1423.18 569.717,1423.18 569.717,1193.86 448.104,1193.86 "/>
+<path clip-path="url(#clip272)" d="M600.12 1170.92 L600.12 1423.18 L721.733 1423.18 L721.733 1170.92 L600.12 1170.92 L600.12 1170.92  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="600.12,1170.92 600.12,1423.18 721.733,1423.18 721.733,1170.92 600.12,1170.92 "/>
+<path clip-path="url(#clip272)" d="M752.137 1147.99 L752.137 1423.18 L873.75 1423.18 L873.75 1147.99 L752.137 1147.99 L752.137 1147.99  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="752.137,1147.99 752.137,1423.18 873.75,1423.18 873.75,1147.99 752.137,1147.99 "/>
+<path clip-path="url(#clip272)" d="M904.153 964.535 L904.153 1423.18 L1025.77 1423.18 L1025.77 964.535 L904.153 964.535 L904.153 964.535  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="904.153,964.535 904.153,1423.18 1025.77,1423.18 1025.77,964.535 904.153,964.535 "/>
+<path clip-path="url(#clip272)" d="M1056.17 1423.18 L1056.17 1423.18 L1177.78 1423.18 L1177.78 1423.18 L1056.17 1423.18 L1056.17 1423.18  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1056.17,1423.18 1056.17,1423.18 1177.78,1423.18 1056.17,1423.18 "/>
+<path clip-path="url(#clip272)" d="M1208.19 1102.13 L1208.19 1423.18 L1329.8 1423.18 L1329.8 1102.13 L1208.19 1102.13 L1208.19 1102.13  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1208.19,1102.13 1208.19,1423.18 1329.8,1423.18 1329.8,1102.13 1208.19,1102.13 "/>
+<path clip-path="url(#clip272)" d="M1360.2 1147.99 L1360.2 1423.18 L1481.82 1423.18 L1481.82 1147.99 L1360.2 1147.99 L1360.2 1147.99  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1360.2,1147.99 1360.2,1423.18 1481.82,1423.18 1481.82,1147.99 1360.2,1147.99 "/>
+<path clip-path="url(#clip272)" d="M1512.22 964.535 L1512.22 1423.18 L1633.83 1423.18 L1633.83 964.535 L1512.22 964.535 L1512.22 964.535  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1512.22,964.535 1512.22,1423.18 1633.83,1423.18 1633.83,964.535 1512.22,964.535 "/>
+<path clip-path="url(#clip272)" d="M1664.24 964.535 L1664.24 1423.18 L1785.85 1423.18 L1785.85 964.535 L1664.24 964.535 L1664.24 964.535  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1664.24,964.535 1664.24,1423.18 1785.85,1423.18 1785.85,964.535 1664.24,964.535 "/>
+<path clip-path="url(#clip272)" d="M1816.25 1019.57 L1816.25 1423.18 L1937.87 1423.18 L1937.87 1019.57 L1816.25 1019.57 L1816.25 1019.57  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1816.25,1019.57 1816.25,1423.18 1937.87,1423.18 1937.87,1019.57 1816.25,1019.57 "/>
+<path clip-path="url(#clip272)" d="M1968.27 872.806 L1968.27 1423.18 L2089.88 1423.18 L2089.88 872.806 L1968.27 872.806 L1968.27 872.806  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1968.27,872.806 1968.27,1423.18 2089.88,1423.18 2089.88,872.806 1968.27,872.806 "/>
+<path clip-path="url(#clip272)" d="M2120.29 1423.18 L2120.29 1423.18 L2241.9 1423.18 L2241.9 1423.18 L2120.29 1423.18 L2120.29 1423.18  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip272)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2120.29,1423.18 2120.29,1423.18 2241.9,1423.18 2120.29,1423.18 "/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="508.91" cy="1193.86" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="660.927" cy="1170.92" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="812.943" cy="1147.99" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="964.96" cy="964.535" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1116.98" cy="1423.18" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1268.99" cy="1102.13" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1421.01" cy="1147.99" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1573.03" cy="964.535" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1725.04" cy="964.535" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1877.06" cy="1019.57" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="2029.08" cy="872.806" r="2"/>
+<circle clip-path="url(#clip272)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="2181.09" cy="1423.18" r="2"/>
+<path clip-path="url(#clip270)" d="M404.431 300.469 L961.373 300.469 L961.373 93.1086 L404.431 93.1086  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="404.431,300.469 961.373,300.469 961.373,93.1086 404.431,93.1086 404.431,300.469 "/>
+<path clip-path="url(#clip270)" d="M426.825 165.685 L561.192 165.685 L561.192 124.213 L426.825 124.213 L426.825 165.685  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="426.825,165.685 561.192,165.685 561.192,124.213 426.825,124.213 426.825,165.685 "/>
+<path clip-path="url(#clip270)" d="M583.587 151.997 L583.587 136.303 L587.846 136.303 L587.846 151.835 Q587.846 155.516 589.281 157.368 Q590.717 159.196 593.587 159.196 Q597.036 159.196 599.027 156.997 Q601.041 154.798 601.041 151.002 L601.041 136.303 L605.3 136.303 L605.3 162.229 L601.041 162.229 L601.041 158.247 Q599.49 160.608 597.429 161.766 Q595.392 162.9 592.684 162.9 Q588.217 162.9 585.902 160.122 Q583.587 157.344 583.587 151.997 M594.305 135.678 L594.305 135.678 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M635.624 146.581 L635.624 162.229 L631.365 162.229 L631.365 146.719 Q631.365 143.039 629.929 141.21 Q628.494 139.382 625.624 139.382 Q622.175 139.382 620.184 141.581 Q618.193 143.78 618.193 147.576 L618.193 162.229 L613.911 162.229 L613.911 136.303 L618.193 136.303 L618.193 140.331 Q619.721 137.993 621.781 136.835 Q623.865 135.678 626.573 135.678 Q631.04 135.678 633.332 138.456 Q635.624 141.21 635.624 146.581 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M664.304 141.28 Q665.901 138.409 668.124 137.044 Q670.346 135.678 673.355 135.678 Q677.406 135.678 679.605 138.525 Q681.804 141.349 681.804 146.581 L681.804 162.229 L677.522 162.229 L677.522 146.719 Q677.522 142.993 676.202 141.187 Q674.883 139.382 672.174 139.382 Q668.864 139.382 666.943 141.581 Q665.022 143.78 665.022 147.576 L665.022 162.229 L660.739 162.229 L660.739 146.719 Q660.739 142.969 659.42 141.187 Q658.1 139.382 655.346 139.382 Q652.082 139.382 650.161 141.604 Q648.239 143.803 648.239 147.576 L648.239 162.229 L643.957 162.229 L643.957 136.303 L648.239 136.303 L648.239 140.331 Q649.698 137.946 651.735 136.812 Q653.772 135.678 656.573 135.678 Q659.397 135.678 661.364 137.113 Q663.355 138.548 664.304 141.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M712.475 148.201 L712.475 150.284 L692.892 150.284 Q693.17 154.682 695.531 156.997 Q697.915 159.289 702.151 159.289 Q704.605 159.289 706.897 158.687 Q709.211 158.085 711.48 156.881 L711.48 160.909 Q709.188 161.881 706.781 162.391 Q704.373 162.9 701.897 162.9 Q695.693 162.9 692.059 159.289 Q688.448 155.678 688.448 149.52 Q688.448 143.155 691.873 139.428 Q695.322 135.678 701.156 135.678 Q706.387 135.678 709.42 139.057 Q712.475 142.414 712.475 148.201 M708.216 146.951 Q708.17 143.456 706.248 141.372 Q704.35 139.289 701.202 139.289 Q697.637 139.289 695.485 141.303 Q693.355 143.317 693.031 146.974 L708.216 146.951 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M723.679 128.942 L723.679 136.303 L732.452 136.303 L732.452 139.613 L723.679 139.613 L723.679 153.687 Q723.679 156.858 724.535 157.761 Q725.415 158.664 728.077 158.664 L732.452 158.664 L732.452 162.229 L728.077 162.229 Q723.146 162.229 721.271 160.4 Q719.396 158.548 719.396 153.687 L719.396 139.613 L716.271 139.613 L716.271 136.303 L719.396 136.303 L719.396 128.942 L723.679 128.942 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M770.183 140.238 L770.183 126.21 L774.442 126.21 L774.442 162.229 L770.183 162.229 L770.183 158.34 Q768.841 160.655 766.78 161.789 Q764.743 162.9 761.873 162.9 Q757.174 162.9 754.211 159.15 Q751.271 155.4 751.271 149.289 Q751.271 143.178 754.211 139.428 Q757.174 135.678 761.873 135.678 Q764.743 135.678 766.78 136.812 Q768.841 137.923 770.183 140.238 M755.669 149.289 Q755.669 153.988 757.591 156.673 Q759.535 159.335 762.915 159.335 Q766.294 159.335 768.239 156.673 Q770.183 153.988 770.183 149.289 Q770.183 144.59 768.239 141.928 Q766.294 139.243 762.915 139.243 Q759.535 139.243 757.591 141.928 Q755.669 144.59 755.669 149.289 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M805.391 148.201 L805.391 150.284 L785.808 150.284 Q786.086 154.682 788.447 156.997 Q790.831 159.289 795.067 159.289 Q797.521 159.289 799.813 158.687 Q802.127 158.085 804.396 156.881 L804.396 160.909 Q802.104 161.881 799.697 162.391 Q797.29 162.9 794.813 162.9 Q788.609 162.9 784.975 159.289 Q781.364 155.678 781.364 149.52 Q781.364 143.155 784.79 139.428 Q788.239 135.678 794.072 135.678 Q799.303 135.678 802.336 139.057 Q805.391 142.414 805.391 148.201 M801.132 146.951 Q801.086 143.456 799.165 141.372 Q797.266 139.289 794.118 139.289 Q790.553 139.289 788.401 141.303 Q786.271 143.317 785.947 146.974 L801.132 146.951 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M832.567 141.28 Q834.164 138.409 836.387 137.044 Q838.609 135.678 841.618 135.678 Q845.669 135.678 847.868 138.525 Q850.067 141.349 850.067 146.581 L850.067 162.229 L845.785 162.229 L845.785 146.719 Q845.785 142.993 844.465 141.187 Q843.146 139.382 840.437 139.382 Q837.127 139.382 835.206 141.581 Q833.285 143.78 833.285 147.576 L833.285 162.229 L829.002 162.229 L829.002 146.719 Q829.002 142.969 827.683 141.187 Q826.363 139.382 823.609 139.382 Q820.345 139.382 818.424 141.604 Q816.502 143.803 816.502 147.576 L816.502 162.229 L812.22 162.229 L812.22 136.303 L816.502 136.303 L816.502 140.331 Q817.961 137.946 819.998 136.812 Q822.035 135.678 824.836 135.678 Q827.66 135.678 829.627 137.113 Q831.618 138.548 832.567 141.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M870.345 149.196 Q865.183 149.196 863.192 150.377 Q861.201 151.557 861.201 154.405 Q861.201 156.673 862.683 158.016 Q864.187 159.335 866.757 159.335 Q870.298 159.335 872.428 156.835 Q874.581 154.312 874.581 150.145 L874.581 149.196 L870.345 149.196 M878.84 147.437 L878.84 162.229 L874.581 162.229 L874.581 158.293 Q873.122 160.655 870.947 161.789 Q868.771 162.9 865.623 162.9 Q861.641 162.9 859.28 160.678 Q856.942 158.432 856.942 154.682 Q856.942 150.307 859.859 148.085 Q862.798 145.863 868.609 145.863 L874.581 145.863 L874.581 145.446 Q874.581 142.507 872.636 140.909 Q870.715 139.289 867.22 139.289 Q864.998 139.289 862.891 139.821 Q860.785 140.354 858.84 141.419 L858.84 137.483 Q861.178 136.581 863.377 136.141 Q865.576 135.678 867.66 135.678 Q873.285 135.678 876.062 138.594 Q878.84 141.511 878.84 147.437 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M909.164 146.581 L909.164 162.229 L904.905 162.229 L904.905 146.719 Q904.905 143.039 903.47 141.21 Q902.034 139.382 899.164 139.382 Q895.715 139.382 893.724 141.581 Q891.733 143.78 891.733 147.576 L891.733 162.229 L887.451 162.229 L887.451 136.303 L891.733 136.303 L891.733 140.331 Q893.261 137.993 895.321 136.835 Q897.405 135.678 900.113 135.678 Q904.581 135.678 906.872 138.456 Q909.164 141.21 909.164 146.581 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M934.719 140.238 L934.719 126.21 L938.979 126.21 L938.979 162.229 L934.719 162.229 L934.719 158.34 Q933.377 160.655 931.317 161.789 Q929.28 162.9 926.409 162.9 Q921.71 162.9 918.747 159.15 Q915.807 155.4 915.807 149.289 Q915.807 143.178 918.747 139.428 Q921.71 135.678 926.409 135.678 Q929.28 135.678 931.317 136.812 Q933.377 137.923 934.719 140.238 M920.206 149.289 Q920.206 153.988 922.127 156.673 Q924.071 159.335 927.451 159.335 Q930.83 159.335 932.775 156.673 Q934.719 153.988 934.719 149.289 Q934.719 144.59 932.775 141.928 Q930.83 139.243 927.451 139.243 Q924.071 139.243 922.127 141.928 Q920.206 144.59 920.206 149.289 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M426.825 217.525 L561.192 217.525 L561.192 176.053 L426.825 176.053 L426.825 217.525  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="426.825,217.525 561.192,217.525 561.192,176.053 426.825,176.053 426.825,217.525 "/>
+<path clip-path="url(#clip270)" d="M599.42 184.115 L593.078 201.314 L605.786 201.314 L599.42 184.115 M596.781 179.509 L602.082 179.509 L615.253 214.069 L610.392 214.069 L607.244 205.203 L591.666 205.203 L588.518 214.069 L583.587 214.069 L596.781 179.509 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip270)" d="M426.825 269.365 L561.192 269.365 L561.192 227.893 L426.825 227.893 L426.825 269.365  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip270)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="426.825,269.365 561.192,269.365 561.192,227.893 426.825,227.893 426.825,269.365 "/>
+<path clip-path="url(#clip270)" d="M588.263 249.404 L588.263 262.066 L595.763 262.066 Q599.536 262.066 601.342 260.515 Q603.17 258.941 603.17 255.723 Q603.17 252.483 601.342 250.955 Q599.536 249.404 595.763 249.404 L588.263 249.404 M588.263 235.191 L588.263 245.608 L595.184 245.608 Q598.61 245.608 600.277 244.335 Q601.967 243.038 601.967 240.399 Q601.967 237.784 600.277 236.487 Q598.61 235.191 595.184 235.191 L588.263 235.191 M583.587 231.349 L595.531 231.349 Q600.879 231.349 603.772 233.571 Q606.666 235.793 606.666 239.89 Q606.666 243.062 605.184 244.937 Q603.703 246.811 600.832 247.274 Q604.281 248.015 606.179 250.376 Q608.101 252.714 608.101 256.233 Q608.101 260.862 604.953 263.385 Q601.804 265.909 595.994 265.909 L583.587 265.909 L583.587 231.349 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/linear/factory_schedule/4024f2dd.svg b/previews/PR3525/tutorials/linear/factory_schedule/4024f2dd.svg
new file mode 100644
index 00000000000..4840b59ad01
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/factory_schedule/4024f2dd.svg
@@ -0,0 +1,74 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip330">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip330)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip331">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip330)" d="M219.38 1423.18 L2112.76 1423.18 L2112.76 47.2441 L219.38 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip332">
+    <rect x="219" y="47" width="1894" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.38,1423.18 219.38,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="850.505,1423.18 850.505,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1481.63,1423.18 1481.63,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2112.76,1423.18 2112.76,47.2441 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.38,1423.18 2112.76,1423.18 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.38,1423.18 219.38,1404.28 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="850.505,1423.18 850.505,1404.28 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1481.63,1423.18 1481.63,1404.28 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2112.76,1423.18 2112.76,1404.28 "/>
+<path clip-path="url(#clip330)" d="M196.765 1454.1 Q193.153 1454.1 191.325 1457.66 Q189.519 1461.2 189.519 1468.33 Q189.519 1475.44 191.325 1479.01 Q193.153 1482.55 196.765 1482.55 Q200.399 1482.55 202.204 1479.01 Q204.033 1475.44 204.033 1468.33 Q204.033 1461.2 202.204 1457.66 Q200.399 1454.1 196.765 1454.1 M196.765 1450.39 Q202.575 1450.39 205.63 1455 Q208.709 1459.58 208.709 1468.33 Q208.709 1477.06 205.63 1481.67 Q202.575 1486.25 196.765 1486.25 Q190.954 1486.25 187.876 1481.67 Q184.82 1477.06 184.82 1468.33 Q184.82 1459.58 187.876 1455 Q190.954 1450.39 196.765 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M216.926 1479.7 L221.811 1479.7 L221.811 1485.58 L216.926 1485.58 L216.926 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M241.996 1454.1 Q238.385 1454.1 236.556 1457.66 Q234.75 1461.2 234.75 1468.33 Q234.75 1475.44 236.556 1479.01 Q238.385 1482.55 241.996 1482.55 Q245.63 1482.55 247.436 1479.01 Q249.264 1475.44 249.264 1468.33 Q249.264 1461.2 247.436 1457.66 Q245.63 1454.1 241.996 1454.1 M241.996 1450.39 Q247.806 1450.39 250.861 1455 Q253.94 1459.58 253.94 1468.33 Q253.94 1477.06 250.861 1481.67 Q247.806 1486.25 241.996 1486.25 Q236.186 1486.25 233.107 1481.67 Q230.051 1477.06 230.051 1468.33 Q230.051 1459.58 233.107 1455 Q236.186 1450.39 241.996 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M828.387 1454.1 Q824.776 1454.1 822.948 1457.66 Q821.142 1461.2 821.142 1468.33 Q821.142 1475.44 822.948 1479.01 Q824.776 1482.55 828.387 1482.55 Q832.022 1482.55 833.827 1479.01 Q835.656 1475.44 835.656 1468.33 Q835.656 1461.2 833.827 1457.66 Q832.022 1454.1 828.387 1454.1 M828.387 1450.39 Q834.198 1450.39 837.253 1455 Q840.332 1459.58 840.332 1468.33 Q840.332 1477.06 837.253 1481.67 Q834.198 1486.25 828.387 1486.25 Q822.577 1486.25 819.499 1481.67 Q816.443 1477.06 816.443 1468.33 Q816.443 1459.58 819.499 1455 Q822.577 1450.39 828.387 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M848.549 1479.7 L853.434 1479.7 L853.434 1485.58 L848.549 1485.58 L848.549 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M863.665 1451.02 L882.021 1451.02 L882.021 1454.96 L867.947 1454.96 L867.947 1463.43 Q868.966 1463.08 869.984 1462.92 Q871.003 1462.73 872.021 1462.73 Q877.809 1462.73 881.188 1465.9 Q884.568 1469.08 884.568 1474.49 Q884.568 1480.07 881.096 1483.17 Q877.623 1486.25 871.304 1486.25 Q869.128 1486.25 866.859 1485.88 Q864.614 1485.51 862.207 1484.77 L862.207 1480.07 Q864.29 1481.2 866.512 1481.76 Q868.734 1482.32 871.211 1482.32 Q875.216 1482.32 877.554 1480.21 Q879.892 1478.1 879.892 1474.49 Q879.892 1470.88 877.554 1468.77 Q875.216 1466.67 871.211 1466.67 Q869.336 1466.67 867.461 1467.08 Q865.609 1467.5 863.665 1468.38 L863.665 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1448.78 1481.64 L1456.42 1481.64 L1456.42 1455.28 L1448.11 1456.95 L1448.11 1452.69 L1456.38 1451.02 L1461.05 1451.02 L1461.05 1481.64 L1468.69 1481.64 L1468.69 1485.58 L1448.78 1485.58 L1448.78 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1478.14 1479.7 L1483.02 1479.7 L1483.02 1485.58 L1478.14 1485.58 L1478.14 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1503.2 1454.1 Q1499.59 1454.1 1497.76 1457.66 Q1495.96 1461.2 1495.96 1468.33 Q1495.96 1475.44 1497.76 1479.01 Q1499.59 1482.55 1503.2 1482.55 Q1506.84 1482.55 1508.64 1479.01 Q1510.47 1475.44 1510.47 1468.33 Q1510.47 1461.2 1508.64 1457.66 Q1506.84 1454.1 1503.2 1454.1 M1503.2 1450.39 Q1509.01 1450.39 1512.07 1455 Q1515.15 1459.58 1515.15 1468.33 Q1515.15 1477.06 1512.07 1481.67 Q1509.01 1486.25 1503.2 1486.25 Q1497.39 1486.25 1494.32 1481.67 Q1491.26 1477.06 1491.26 1468.33 Q1491.26 1459.58 1494.32 1455 Q1497.39 1450.39 1503.2 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2080.41 1481.64 L2088.05 1481.64 L2088.05 1455.28 L2079.74 1456.95 L2079.74 1452.69 L2088 1451.02 L2092.67 1451.02 L2092.67 1481.64 L2100.31 1481.64 L2100.31 1485.58 L2080.41 1485.58 L2080.41 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2109.76 1479.7 L2114.64 1479.7 L2114.64 1485.58 L2109.76 1485.58 L2109.76 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2124.87 1451.02 L2143.23 1451.02 L2143.23 1454.96 L2129.16 1454.96 L2129.16 1463.43 Q2130.17 1463.08 2131.19 1462.92 Q2132.21 1462.73 2133.23 1462.73 Q2139.02 1462.73 2142.4 1465.9 Q2145.78 1469.08 2145.78 1474.49 Q2145.78 1480.07 2142.3 1483.17 Q2138.83 1486.25 2132.51 1486.25 Q2130.34 1486.25 2128.07 1485.88 Q2125.82 1485.51 2123.42 1484.77 L2123.42 1480.07 Q2125.5 1481.2 2127.72 1481.76 Q2129.94 1482.32 2132.42 1482.32 Q2136.42 1482.32 2138.76 1480.21 Q2141.1 1478.1 2141.1 1474.49 Q2141.1 1470.88 2138.76 1468.77 Q2136.42 1466.67 2132.42 1466.67 Q2130.55 1466.67 2128.67 1467.08 Q2126.82 1467.5 2124.87 1468.38 L2124.87 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M911.822 1522.08 L911.822 1528.35 Q908.161 1526.6 904.915 1525.74 Q901.668 1524.88 898.645 1524.88 Q893.393 1524.88 890.528 1526.92 Q887.696 1528.96 887.696 1532.71 Q887.696 1535.87 889.574 1537.49 Q891.483 1539.08 896.767 1540.07 L900.65 1540.86 Q907.843 1542.23 911.249 1545.7 Q914.686 1549.14 914.686 1554.93 Q914.686 1561.84 910.039 1565.4 Q905.424 1568.97 896.48 1568.97 Q893.107 1568.97 889.287 1568.2 Q885.5 1567.44 881.425 1565.94 L881.425 1559.32 Q885.34 1561.52 889.096 1562.63 Q892.852 1563.75 896.48 1563.75 Q901.987 1563.75 904.979 1561.58 Q907.97 1559.42 907.97 1555.41 Q907.97 1551.91 905.806 1549.93 Q903.674 1547.96 898.772 1546.97 L894.857 1546.21 Q887.664 1544.78 884.449 1541.72 Q881.234 1538.67 881.234 1533.22 Q881.234 1526.92 885.659 1523.29 Q890.115 1519.66 897.913 1519.66 Q901.255 1519.66 904.724 1520.27 Q908.193 1520.87 911.822 1522.08 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M950.111 1533.76 L950.111 1539.24 Q947.629 1537.87 945.114 1537.2 Q942.632 1536.5 940.085 1536.5 Q934.388 1536.5 931.237 1540.13 Q928.086 1543.73 928.086 1550.25 Q928.086 1556.78 931.237 1560.4 Q934.388 1564 940.085 1564 Q942.632 1564 945.114 1563.33 Q947.629 1562.63 950.111 1561.26 L950.111 1566.68 Q947.661 1567.82 945.019 1568.39 Q942.409 1568.97 939.449 1568.97 Q931.396 1568.97 926.654 1563.91 Q921.911 1558.85 921.911 1550.25 Q921.911 1541.53 926.686 1536.53 Q931.492 1531.54 939.831 1531.54 Q942.536 1531.54 945.114 1532.11 Q947.692 1532.65 950.111 1533.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M976.497 1550.12 Q969.4 1550.12 966.662 1551.75 Q963.925 1553.37 963.925 1557.29 Q963.925 1560.4 965.962 1562.25 Q968.031 1564.07 971.564 1564.07 Q976.434 1564.07 979.362 1560.63 Q982.322 1557.16 982.322 1551.43 L982.322 1550.12 L976.497 1550.12 M988.178 1547.71 L988.178 1568.04 L982.322 1568.04 L982.322 1562.63 Q980.317 1565.88 977.325 1567.44 Q974.333 1568.97 970.004 1568.97 Q964.53 1568.97 961.283 1565.91 Q958.069 1562.82 958.069 1557.67 Q958.069 1551.65 962.079 1548.6 Q966.121 1545.54 974.11 1545.54 L982.322 1545.54 L982.322 1544.97 Q982.322 1540.93 979.648 1538.73 Q977.007 1536.5 972.2 1536.5 Q969.145 1536.5 966.249 1537.23 Q963.352 1537.97 960.679 1539.43 L960.679 1534.02 Q963.893 1532.78 966.917 1532.17 Q969.941 1531.54 972.805 1531.54 Q980.54 1531.54 984.359 1535.55 Q988.178 1539.56 988.178 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1000.24 1518.52 L1006.1 1518.52 L1006.1 1568.04 L1000.24 1568.04 L1000.24 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1048.84 1548.76 L1048.84 1551.62 L1021.92 1551.62 Q1022.3 1557.67 1025.55 1560.85 Q1028.82 1564 1034.65 1564 Q1038.02 1564 1041.17 1563.17 Q1044.36 1562.35 1047.47 1560.69 L1047.47 1566.23 Q1044.32 1567.57 1041.01 1568.27 Q1037.7 1568.97 1034.3 1568.97 Q1025.77 1568.97 1020.77 1564 Q1015.81 1559.04 1015.81 1550.57 Q1015.81 1541.82 1020.52 1536.69 Q1025.26 1531.54 1033.28 1531.54 Q1040.47 1531.54 1044.64 1536.18 Q1048.84 1540.8 1048.84 1548.76 M1042.99 1547.04 Q1042.92 1542.23 1040.28 1539.37 Q1037.67 1536.5 1033.34 1536.5 Q1028.44 1536.5 1025.48 1539.27 Q1022.55 1542.04 1022.11 1547.07 L1042.99 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1092.99 1536.5 Q1088.28 1536.5 1085.54 1540.19 Q1082.8 1543.85 1082.8 1550.25 Q1082.8 1556.65 1085.51 1560.34 Q1088.25 1564 1092.99 1564 Q1097.67 1564 1100.41 1560.31 Q1103.14 1556.62 1103.14 1550.25 Q1103.14 1543.92 1100.41 1540.23 Q1097.67 1536.5 1092.99 1536.5 M1092.99 1531.54 Q1100.63 1531.54 1104.99 1536.5 Q1109.35 1541.47 1109.35 1550.25 Q1109.35 1559 1104.99 1564 Q1100.63 1568.97 1092.99 1568.97 Q1085.32 1568.97 1080.96 1564 Q1076.63 1559 1076.63 1550.25 Q1076.63 1541.47 1080.96 1536.5 Q1085.32 1531.54 1092.99 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1137.1 1518.52 L1137.1 1523.39 L1131.5 1523.39 Q1128.35 1523.39 1127.11 1524.66 Q1125.9 1525.93 1125.9 1529.24 L1125.9 1532.4 L1135.54 1532.4 L1135.54 1536.95 L1125.9 1536.95 L1125.9 1568.04 L1120.01 1568.04 L1120.01 1536.95 L1114.41 1536.95 L1114.41 1532.4 L1120.01 1532.4 L1120.01 1529.91 Q1120.01 1523.96 1122.78 1521.26 Q1125.55 1518.52 1131.57 1518.52 L1137.1 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1180.77 1518.52 L1180.77 1523.39 L1175.17 1523.39 Q1172.02 1523.39 1170.78 1524.66 Q1169.57 1525.93 1169.57 1529.24 L1169.57 1532.4 L1179.21 1532.4 L1179.21 1536.95 L1169.57 1536.95 L1169.57 1568.04 L1163.68 1568.04 L1163.68 1536.95 L1158.08 1536.95 L1158.08 1532.4 L1163.68 1532.4 L1163.68 1529.91 Q1163.68 1523.96 1166.45 1521.26 Q1169.22 1518.52 1175.23 1518.52 L1180.77 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1201.88 1550.12 Q1194.78 1550.12 1192.04 1551.75 Q1189.3 1553.37 1189.3 1557.29 Q1189.3 1560.4 1191.34 1562.25 Q1193.41 1564.07 1196.94 1564.07 Q1201.81 1564.07 1204.74 1560.63 Q1207.7 1557.16 1207.7 1551.43 L1207.7 1550.12 L1201.88 1550.12 M1213.56 1547.71 L1213.56 1568.04 L1207.7 1568.04 L1207.7 1562.63 Q1205.69 1565.88 1202.7 1567.44 Q1199.71 1568.97 1195.38 1568.97 Q1189.91 1568.97 1186.66 1565.91 Q1183.45 1562.82 1183.45 1557.67 Q1183.45 1551.65 1187.46 1548.6 Q1191.5 1545.54 1199.49 1545.54 L1207.7 1545.54 L1207.7 1544.97 Q1207.7 1540.93 1205.03 1538.73 Q1202.38 1536.5 1197.58 1536.5 Q1194.52 1536.5 1191.63 1537.23 Q1188.73 1537.97 1186.06 1539.43 L1186.06 1534.02 Q1189.27 1532.78 1192.29 1532.17 Q1195.32 1531.54 1198.18 1531.54 Q1205.92 1531.54 1209.74 1535.55 Q1213.56 1539.56 1213.56 1547.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1251.27 1533.76 L1251.27 1539.24 Q1248.79 1537.87 1246.28 1537.2 Q1243.79 1536.5 1241.25 1536.5 Q1235.55 1536.5 1232.4 1540.13 Q1229.25 1543.73 1229.25 1550.25 Q1229.25 1556.78 1232.4 1560.4 Q1235.55 1564 1241.25 1564 Q1243.79 1564 1246.28 1563.33 Q1248.79 1562.63 1251.27 1561.26 L1251.27 1566.68 Q1248.82 1567.82 1246.18 1568.39 Q1243.57 1568.97 1240.61 1568.97 Q1232.56 1568.97 1227.82 1563.91 Q1223.07 1558.85 1223.07 1550.25 Q1223.07 1541.53 1227.85 1536.53 Q1232.65 1531.54 1240.99 1531.54 Q1243.7 1531.54 1246.28 1532.11 Q1248.85 1532.65 1251.27 1533.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1267.25 1522.27 L1267.25 1532.4 L1279.31 1532.4 L1279.31 1536.95 L1267.25 1536.95 L1267.25 1556.3 Q1267.25 1560.66 1268.43 1561.9 Q1269.64 1563.14 1273.3 1563.14 L1279.31 1563.14 L1279.31 1568.04 L1273.3 1568.04 Q1266.52 1568.04 1263.94 1565.53 Q1261.36 1562.98 1261.36 1556.3 L1261.36 1536.95 L1257.07 1536.95 L1257.07 1532.4 L1261.36 1532.4 L1261.36 1522.27 L1267.25 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1300.83 1536.5 Q1296.12 1536.5 1293.38 1540.19 Q1290.64 1543.85 1290.64 1550.25 Q1290.64 1556.65 1293.35 1560.34 Q1296.09 1564 1300.83 1564 Q1305.51 1564 1308.25 1560.31 Q1310.98 1556.62 1310.98 1550.25 Q1310.98 1543.92 1308.25 1540.23 Q1305.51 1536.5 1300.83 1536.5 M1300.83 1531.54 Q1308.47 1531.54 1312.83 1536.5 Q1317.19 1541.47 1317.19 1550.25 Q1317.19 1559 1312.83 1564 Q1308.47 1568.97 1300.83 1568.97 Q1293.16 1568.97 1288.8 1564 Q1284.47 1559 1284.47 1550.25 Q1284.47 1541.47 1288.8 1536.5 Q1293.16 1531.54 1300.83 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1347.55 1537.87 Q1346.57 1537.3 1345.39 1537.04 Q1344.24 1536.76 1342.84 1536.76 Q1337.88 1536.76 1335.2 1540 Q1332.56 1543.22 1332.56 1549.27 L1332.56 1568.04 L1326.67 1568.04 L1326.67 1532.4 L1332.56 1532.4 L1332.56 1537.93 Q1334.41 1534.69 1337.37 1533.13 Q1340.33 1531.54 1344.56 1531.54 Q1345.17 1531.54 1345.9 1531.63 Q1346.63 1531.7 1347.52 1531.85 L1347.55 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1368.53 1571.35 Q1366.05 1577.72 1363.69 1579.66 Q1361.34 1581.6 1357.39 1581.6 L1352.71 1581.6 L1352.71 1576.7 L1356.15 1576.7 Q1358.57 1576.7 1359.9 1575.56 Q1361.24 1574.41 1362.86 1570.14 L1363.91 1567.47 L1349.5 1532.4 L1355.7 1532.4 L1366.84 1560.28 L1377.98 1532.4 L1384.19 1532.4 L1368.53 1571.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M1429.13 1526.86 L1420.41 1550.51 L1437.88 1550.51 L1429.13 1526.86 M1425.5 1520.52 L1432.79 1520.52 L1450.9 1568.04 L1444.22 1568.04 L1439.89 1555.85 L1418.47 1555.85 L1414.14 1568.04 L1407.36 1568.04 L1425.5 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.38,1423.18 2112.76,1423.18 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.38,964.535 2112.76,964.535 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.38,505.889 2112.76,505.889 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="219.38,47.2441 2112.76,47.2441 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.38,1423.18 219.38,47.2441 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.38,1423.18 236.152,1423.18 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.38,964.535 236.152,964.535 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.38,505.889 236.152,505.889 "/>
+<polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="219.38,47.2441 236.152,47.2441 "/>
+<path clip-path="url(#clip330)" d="M126.205 1408.98 Q122.593 1408.98 120.765 1412.54 Q118.959 1416.08 118.959 1423.21 Q118.959 1430.32 120.765 1433.89 Q122.593 1437.43 126.205 1437.43 Q129.839 1437.43 131.644 1433.89 Q133.473 1430.32 133.473 1423.21 Q133.473 1416.08 131.644 1412.54 Q129.839 1408.98 126.205 1408.98 M126.205 1405.27 Q132.015 1405.27 135.07 1409.88 Q138.149 1414.46 138.149 1423.21 Q138.149 1431.94 135.07 1436.55 Q132.015 1441.13 126.205 1441.13 Q120.394 1441.13 117.316 1436.55 Q114.26 1431.94 114.26 1423.21 Q114.26 1414.46 117.316 1409.88 Q120.394 1405.27 126.205 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M146.366 1434.58 L151.251 1434.58 L151.251 1440.46 L146.366 1440.46 L146.366 1434.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M171.436 1408.98 Q167.825 1408.98 165.996 1412.54 Q164.19 1416.08 164.19 1423.21 Q164.19 1430.32 165.996 1433.89 Q167.825 1437.43 171.436 1437.43 Q175.07 1437.43 176.876 1433.89 Q178.704 1430.32 178.704 1423.21 Q178.704 1416.08 176.876 1412.54 Q175.07 1408.98 171.436 1408.98 M171.436 1405.27 Q177.246 1405.27 180.301 1409.88 Q183.38 1414.46 183.38 1423.21 Q183.38 1431.94 180.301 1436.55 Q177.246 1441.13 171.436 1441.13 Q165.626 1441.13 162.547 1436.55 Q159.491 1431.94 159.491 1423.21 Q159.491 1414.46 162.547 1409.88 Q165.626 1405.27 171.436 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M127.2 950.333 Q123.589 950.333 121.76 953.898 Q119.955 957.44 119.955 964.569 Q119.955 971.676 121.76 975.241 Q123.589 978.782 127.2 978.782 Q130.834 978.782 132.64 975.241 Q134.468 971.676 134.468 964.569 Q134.468 957.44 132.64 953.898 Q130.834 950.333 127.2 950.333 M127.2 946.63 Q133.01 946.63 136.066 951.236 Q139.144 955.819 139.144 964.569 Q139.144 973.296 136.066 977.903 Q133.01 982.486 127.2 982.486 Q121.39 982.486 118.311 977.903 Q115.256 973.296 115.256 964.569 Q115.256 955.819 118.311 951.236 Q121.39 946.63 127.2 946.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M147.362 975.935 L152.246 975.935 L152.246 981.815 L147.362 981.815 L147.362 975.935 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M162.477 947.255 L180.834 947.255 L180.834 951.19 L166.76 951.19 L166.76 959.662 Q167.778 959.315 168.797 959.153 Q169.815 958.967 170.834 958.967 Q176.621 958.967 180.001 962.139 Q183.38 965.31 183.38 970.727 Q183.38 976.305 179.908 979.407 Q176.436 982.486 170.116 982.486 Q167.94 982.486 165.672 982.116 Q163.427 981.745 161.019 981.004 L161.019 976.305 Q163.102 977.44 165.325 977.995 Q167.547 978.551 170.024 978.551 Q174.028 978.551 176.366 976.444 Q178.704 974.338 178.704 970.727 Q178.704 967.116 176.366 965.009 Q174.028 962.903 170.024 962.903 Q168.149 962.903 166.274 963.319 Q164.422 963.736 162.477 964.616 L162.477 947.255 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M117.015 519.234 L124.654 519.234 L124.654 492.869 L116.343 494.535 L116.343 490.276 L124.607 488.609 L129.283 488.609 L129.283 519.234 L136.922 519.234 L136.922 523.169 L117.015 523.169 L117.015 519.234 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M146.366 517.29 L151.251 517.29 L151.251 523.169 L146.366 523.169 L146.366 517.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M171.436 491.688 Q167.825 491.688 165.996 495.253 Q164.19 498.794 164.19 505.924 Q164.19 513.031 165.996 516.595 Q167.825 520.137 171.436 520.137 Q175.07 520.137 176.876 516.595 Q178.704 513.031 178.704 505.924 Q178.704 498.794 176.876 495.253 Q175.07 491.688 171.436 491.688 M171.436 487.984 Q177.246 487.984 180.301 492.591 Q183.38 497.174 183.38 505.924 Q183.38 514.651 180.301 519.257 Q177.246 523.841 171.436 523.841 Q165.626 523.841 162.547 519.257 Q159.491 514.651 159.491 505.924 Q159.491 497.174 162.547 492.591 Q165.626 487.984 171.436 487.984 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M118.01 60.5889 L125.649 60.5889 L125.649 34.2233 L117.339 35.89 L117.339 31.6308 L125.603 29.9641 L130.279 29.9641 L130.279 60.5889 L137.917 60.5889 L137.917 64.5241 L118.01 64.5241 L118.01 60.5889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M147.362 58.6445 L152.246 58.6445 L152.246 64.5241 L147.362 64.5241 L147.362 58.6445 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M162.477 29.9641 L180.834 29.9641 L180.834 33.8993 L166.76 33.8993 L166.76 42.3714 Q167.778 42.0242 168.797 41.8622 Q169.815 41.677 170.834 41.677 Q176.621 41.677 180.001 44.8483 Q183.38 48.0196 183.38 53.4362 Q183.38 59.0149 179.908 62.1167 Q176.436 65.1954 170.116 65.1954 Q167.94 65.1954 165.672 64.825 Q163.427 64.4547 161.019 63.7139 L161.019 59.0149 Q163.102 60.1491 165.325 60.7047 Q167.547 61.2602 170.024 61.2602 Q174.028 61.2602 176.366 59.1538 Q178.704 57.0473 178.704 53.4362 Q178.704 49.8251 176.366 47.7186 Q174.028 45.6122 170.024 45.6122 Q168.149 45.6122 166.274 46.0288 Q164.422 46.4455 162.477 47.3251 L162.477 29.9641 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M18.0438 987.485 L24.314 987.485 Q22.5634 991.145 21.704 994.392 Q20.8447 997.638 20.8447 1000.66 Q20.8447 1005.91 22.8817 1008.78 Q24.9187 1011.61 28.6745 1011.61 Q31.8255 1011.61 33.4488 1009.73 Q35.0402 1007.82 36.0269 1002.54 L36.8226 998.657 Q38.1912 991.463 41.6605 988.058 Q45.098 984.62 50.8908 984.62 Q57.7976 984.62 61.3624 989.267 Q64.9272 993.882 64.9272 1002.83 Q64.9272 1006.2 64.1633 1010.02 Q63.3994 1013.81 61.9035 1017.88 L55.2831 1017.88 Q57.4793 1013.97 58.5933 1010.21 Q59.7073 1006.45 59.7073 1002.83 Q59.7073 997.32 57.543 994.328 Q55.3786 991.336 51.3682 991.336 Q47.8671 991.336 45.8937 993.501 Q43.9204 995.633 42.9337 1000.53 L42.1698 1004.45 Q40.7375 1011.64 37.682 1014.86 Q34.6264 1018.07 29.1837 1018.07 Q22.8817 1018.07 19.2532 1013.65 Q15.6248 1009.19 15.6248 1001.39 Q15.6248 998.052 16.2295 994.583 Q16.8343 991.113 18.0438 987.485 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M29.7248 949.195 L35.1993 949.195 Q33.8307 951.678 33.1623 954.192 Q32.4621 956.675 32.4621 959.221 Q32.4621 964.918 36.0905 968.07 Q39.6872 971.221 46.212 971.221 Q52.7369 971.221 56.3653 968.07 Q59.9619 964.918 59.9619 959.221 Q59.9619 956.675 59.2935 954.192 Q58.5933 951.678 57.2247 949.195 L62.6355 949.195 Q63.7814 951.646 64.3543 954.288 Q64.9272 956.898 64.9272 959.858 Q64.9272 967.91 59.8664 972.653 Q54.8057 977.395 46.212 977.395 Q37.491 977.395 32.4939 972.621 Q27.4968 967.815 27.4968 959.476 Q27.4968 956.77 28.0697 954.192 Q28.6108 951.614 29.7248 949.195 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M46.0847 922.809 Q46.0847 929.907 47.7079 932.644 Q49.3312 935.382 53.2461 935.382 Q56.3653 935.382 58.2114 933.345 Q60.0256 931.276 60.0256 927.743 Q60.0256 922.873 56.5881 919.945 Q53.1188 916.985 47.3897 916.985 L46.0847 916.985 L46.0847 922.809 M43.6657 911.128 L64.0042 911.128 L64.0042 916.985 L58.5933 916.985 Q61.8398 918.99 63.3994 921.982 Q64.9272 924.974 64.9272 929.302 Q64.9272 934.777 61.8716 938.023 Q58.7843 941.238 53.6281 941.238 Q47.6125 941.238 44.5569 937.228 Q41.5014 933.185 41.5014 925.196 L41.5014 916.985 L40.9285 916.985 Q36.8862 916.985 34.6901 919.658 Q32.4621 922.3 32.4621 927.106 Q32.4621 930.162 33.1941 933.058 Q33.9262 935.955 35.3903 938.628 L29.9795 938.628 Q28.7381 935.413 28.1334 932.39 Q27.4968 929.366 27.4968 926.501 Q27.4968 918.767 31.5072 914.948 Q35.5176 911.128 43.6657 911.128 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M14.479 899.065 L14.479 893.209 L64.0042 893.209 L64.0042 899.065 L14.479 899.065 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M44.7161 850.463 L47.5806 850.463 L47.5806 877.39 Q53.6281 877.008 56.8109 873.762 Q59.9619 870.483 59.9619 864.659 Q59.9619 861.285 59.1344 858.134 Q58.3069 854.951 56.6518 851.832 L62.1899 851.832 Q63.5267 854.983 64.227 858.293 Q64.9272 861.603 64.9272 865.009 Q64.9272 873.539 59.9619 878.536 Q54.9967 883.501 46.5303 883.501 Q37.7774 883.501 32.6531 878.79 Q27.4968 874.048 27.4968 866.027 Q27.4968 858.834 32.1438 854.664 Q36.7589 850.463 44.7161 850.463 M42.9973 856.32 Q38.1912 856.383 35.3266 859.025 Q32.4621 861.635 32.4621 865.964 Q32.4621 870.865 35.2312 873.825 Q38.0002 876.753 43.0292 877.199 L42.9973 856.32 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M32.4621 806.317 Q32.4621 811.028 36.1542 813.765 Q39.8145 816.502 46.212 816.502 Q52.6095 816.502 56.3017 813.797 Q59.9619 811.059 59.9619 806.317 Q59.9619 801.638 56.2698 798.901 Q52.5777 796.164 46.212 796.164 Q39.8781 796.164 36.186 798.901 Q32.4621 801.638 32.4621 806.317 M27.4968 806.317 Q27.4968 798.678 32.4621 794.318 Q37.4273 789.957 46.212 789.957 Q54.9649 789.957 59.9619 794.318 Q64.9272 798.678 64.9272 806.317 Q64.9272 813.988 59.9619 818.348 Q54.9649 822.677 46.212 822.677 Q37.4273 822.677 32.4621 818.348 Q27.4968 813.988 27.4968 806.317 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M14.479 762.203 L19.3487 762.203 L19.3487 767.804 Q19.3487 770.955 20.6219 772.197 Q21.895 773.406 25.2052 773.406 L28.3562 773.406 L28.3562 763.762 L32.9077 763.762 L32.9077 773.406 L64.0042 773.406 L64.0042 779.294 L32.9077 779.294 L32.9077 784.896 L28.3562 784.896 L28.3562 779.294 L25.8736 779.294 Q19.9216 779.294 17.2162 776.525 Q14.479 773.756 14.479 767.741 L14.479 762.203 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M14.479 718.534 L19.3487 718.534 L19.3487 724.136 Q19.3487 727.287 20.6219 728.528 Q21.895 729.737 25.2052 729.737 L28.3562 729.737 L28.3562 720.093 L32.9077 720.093 L32.9077 729.737 L64.0042 729.737 L64.0042 735.626 L32.9077 735.626 L32.9077 741.228 L28.3562 741.228 L28.3562 735.626 L25.8736 735.626 Q19.9216 735.626 17.2162 732.857 Q14.479 730.088 14.479 724.072 L14.479 718.534 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M46.0847 697.431 Q46.0847 704.529 47.7079 707.267 Q49.3312 710.004 53.2461 710.004 Q56.3653 710.004 58.2114 707.967 Q60.0256 705.898 60.0256 702.365 Q60.0256 697.495 56.5881 694.567 Q53.1188 691.607 47.3897 691.607 L46.0847 691.607 L46.0847 697.431 M43.6657 685.75 L64.0042 685.75 L64.0042 691.607 L58.5933 691.607 Q61.8398 693.612 63.3994 696.604 Q64.9272 699.596 64.9272 703.925 Q64.9272 709.399 61.8716 712.646 Q58.7843 715.86 53.6281 715.86 Q47.6125 715.86 44.5569 711.85 Q41.5014 707.808 41.5014 699.819 L41.5014 691.607 L40.9285 691.607 Q36.8862 691.607 34.6901 694.28 Q32.4621 696.922 32.4621 701.728 Q32.4621 704.784 33.1941 707.68 Q33.9262 710.577 35.3903 713.25 L29.9795 713.25 Q28.7381 710.036 28.1334 707.012 Q27.4968 703.988 27.4968 701.124 Q27.4968 693.389 31.5072 689.57 Q35.5176 685.75 43.6657 685.75 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M29.7248 648.034 L35.1993 648.034 Q33.8307 650.516 33.1623 653.031 Q32.4621 655.513 32.4621 658.06 Q32.4621 663.757 36.0905 666.908 Q39.6872 670.059 46.212 670.059 Q52.7369 670.059 56.3653 666.908 Q59.9619 663.757 59.9619 658.06 Q59.9619 655.513 59.2935 653.031 Q58.5933 650.516 57.2247 648.034 L62.6355 648.034 Q63.7814 650.484 64.3543 653.126 Q64.9272 655.736 64.9272 658.696 Q64.9272 666.749 59.8664 671.491 Q54.8057 676.234 46.212 676.234 Q37.491 676.234 32.4939 671.459 Q27.4968 666.653 27.4968 658.314 Q27.4968 655.609 28.0697 653.031 Q28.6108 650.453 29.7248 648.034 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M18.2347 632.056 L28.3562 632.056 L28.3562 619.993 L32.9077 619.993 L32.9077 632.056 L52.2594 632.056 Q56.6199 632.056 57.8613 630.878 Q59.1026 629.669 59.1026 626.008 L59.1026 619.993 L64.0042 619.993 L64.0042 626.008 Q64.0042 632.788 61.4897 635.366 Q58.9434 637.944 52.2594 637.944 L32.9077 637.944 L32.9077 642.241 L28.3562 642.241 L28.3562 637.944 L18.2347 637.944 L18.2347 632.056 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M32.4621 598.477 Q32.4621 603.187 36.1542 605.924 Q39.8145 608.662 46.212 608.662 Q52.6095 608.662 56.3017 605.956 Q59.9619 603.219 59.9619 598.477 Q59.9619 593.798 56.2698 591.061 Q52.5777 588.323 46.212 588.323 Q39.8781 588.323 36.186 591.061 Q32.4621 593.798 32.4621 598.477 M27.4968 598.477 Q27.4968 590.838 32.4621 586.477 Q37.4273 582.117 46.212 582.117 Q54.9649 582.117 59.9619 586.477 Q64.9272 590.838 64.9272 598.477 Q64.9272 606.147 59.9619 610.508 Q54.9649 614.836 46.212 614.836 Q37.4273 614.836 32.4621 610.508 Q27.4968 606.147 27.4968 598.477 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M33.8307 551.752 Q33.2578 552.739 33.0032 553.917 Q32.7167 555.062 32.7167 556.463 Q32.7167 561.428 35.9632 564.102 Q39.1779 566.744 45.2253 566.744 L64.0042 566.744 L64.0042 572.632 L28.3562 572.632 L28.3562 566.744 L33.8944 566.744 Q30.6479 564.897 29.0883 561.937 Q27.4968 558.977 27.4968 554.744 Q27.4968 554.139 27.5923 553.407 Q27.656 552.675 27.8151 551.784 L33.8307 551.752 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M67.3143 530.777 Q73.68 533.26 75.6216 535.615 Q77.5631 537.971 77.5631 541.917 L77.5631 546.596 L72.6615 546.596 L72.6615 543.159 Q72.6615 540.74 71.5157 539.403 Q70.3699 538.066 66.1048 536.443 L63.4312 535.392 L28.3562 549.811 L28.3562 543.604 L56.238 532.464 L28.3562 521.324 L28.3562 515.118 L67.3143 530.777 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M41.3104 479.629 L58.7206 479.629 L58.7206 469.316 Q58.7206 464.128 56.5881 461.646 Q54.4238 459.131 49.9996 459.131 Q45.5436 459.131 43.4429 461.646 Q41.3104 464.128 41.3104 469.316 L41.3104 479.629 M21.7677 479.629 L36.0905 479.629 L36.0905 470.112 Q36.0905 465.401 34.34 463.11 Q32.5576 460.786 28.9291 460.786 Q25.3325 460.786 23.5501 463.11 Q21.7677 465.401 21.7677 470.112 L21.7677 479.629 M16.4842 486.058 L16.4842 469.635 Q16.4842 462.282 19.5397 458.304 Q22.5952 454.325 28.2289 454.325 Q32.5894 454.325 35.1675 456.362 Q37.7456 458.399 38.3822 462.346 Q39.4007 457.604 42.6472 454.994 Q45.8619 452.352 50.6998 452.352 Q57.0655 452.352 60.5348 456.68 Q64.0042 461.009 64.0042 468.998 L64.0042 486.058 L16.4842 486.058 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip332)" style="stroke:#090621; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="436.715,1423.18 345.605,1342.53 338.299,1331.45 311.239,1239.72 304.275,1147.99 299.052,1056.26 292.088,964.535 285.124,872.806 278.16,781.076 272.937,689.347 265.972,597.618 259.008,505.889 252.044,414.16 245.08,322.431 239.857,230.702 232.893,138.973 225.929,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#1f0c47; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="331.716,47.2441 338.68,138.973 345.605,230.186 345.645,230.702 350.941,322.431 358.003,414.16 365.064,505.889 372.126,597.618 379.187,689.347 384.483,781.076 391.545,872.806 398.606,964.535 407.396,1056.26 415.584,1147.99 436.704,1239.72 471.83,1278.77 556.579,1331.45 598.055,1363.92 667.279,1423.18 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#3b0964; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="901.944,1423.18 850.505,1377.88 788.712,1331.45 724.28,1293.11 645.537,1239.72 598.055,1189.57 565.364,1147.99 535.903,1056.26 520.698,964.535 506.109,872.806 495.971,781.076 489.553,689.347 480.694,597.618 472.417,505.889 471.83,499.382 465.27,414.16 458.208,322.431 452.912,230.702 445.85,138.973 438.789,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#550f6d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="556.07,47.2441 564.085,138.973 572.099,230.702 578.11,322.431 589.871,414.16 598.055,504.857 598.136,505.889 605.298,597.618 616.762,689.347 622.349,781.076 638.767,872.806 660.364,964.535 693.661,1056.26 724.28,1113.22 749.472,1147.99 850.505,1218.45 880.799,1239.72 976.73,1306.63 1009.15,1331.45 1102.96,1422.15 1104.48,1423.18 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#6e186e; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1363.74,1423.18 1355.41,1415.12 1246.72,1331.45 1229.18,1320.76 1102.96,1253.66 1084.76,1239.72 976.73,1156.44 964.697,1147.99 883.397,1056.26 850.505,1016.38 810.477,964.535 770.999,872.806 738.909,781.076 731.626,689.347 724.28,657.513 714.086,597.618 706.924,505.889 699.763,414.16 689.02,322.431 683.649,230.702 676.487,138.973 669.325,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#88216a; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="788.607,47.2441 797.165,138.973 809.658,230.702 816.387,322.431 835.267,414.16 844.699,505.889 850.505,546.959 856.866,597.618 875.188,689.347 883.361,781.076 916.204,872.806 971.914,964.535 976.73,969.796 1075.15,1056.26 1102.96,1085.92 1186.81,1147.99 1229.18,1170.67 1341.68,1239.72 1355.41,1248.09 1464.27,1331.45 1481.63,1348.23 1559.14,1423.18 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#a12b61; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1800.62,1423.18 1734.08,1378.52 1672.94,1331.45 1607.86,1291.77 1539.88,1239.72 1481.63,1194.81 1420.89,1147.99 1355.41,1078.16 1329.61,1056.26 1229.18,995.436 1170.67,964.535 1102.96,922.59 1061.09,872.806 1018,781.076 1001.58,689.347 983.534,597.618 976.73,566.79 963.283,505.889 953.185,414.16 936.639,322.431 930.966,230.702 918.228,138.973 910.667,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#ba3655; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1020.02,47.2441 1026.73,138.973 1039.3,230.702 1044.44,322.431 1059.41,414.16 1075.48,505.889 1101.02,597.618 1102.96,605.445 1144.84,689.347 1189.73,781.076 1229.18,816.376 1285.35,872.806 1355.41,922.693 1409.67,964.535 1481.63,1019.27 1530.27,1056.26 1607.86,1138.99 1618.46,1147.99 1734.08,1218.96 1772.87,1239.72 1860.31,1286.2 1919.4,1331.45 1986.53,1396.36 2016.98,1423.18 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#d04544; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2112.76,1321.27 1995.84,1239.72 1986.53,1233.19 1876.02,1147.99 1860.31,1131.24 1759.03,1056.26 1734.08,1043.09 1607.86,966.647 1605.08,964.535 1486.09,872.806 1481.63,869.092 1387.9,781.076 1355.41,717.633 1340.92,689.347 1293.27,597.618 1238.25,505.889 1229.18,474.953 1208.41,414.16 1182.28,322.431 1173.33,230.702 1143.89,138.973 1132.27,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#e35832; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1280.18,47.2441 1298.13,138.973 1329.81,230.702 1336.99,322.431 1355.41,403.127 1357.54,414.16 1391.63,505.889 1449.48,597.618 1481.63,653.862 1505.6,689.347 1571.35,781.076 1607.86,811.481 1700.34,872.806 1734.08,893.513 1860.31,961.07 1864.8,964.535 1985.4,1056.26 1986.53,1057.47 2079.73,1147.99 2112.76,1171.18 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#f0701e; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2112.76,995.946 2067.42,964.535 1986.53,907.74 1941.21,872.806 1860.31,805.425 1818.19,781.076 1734.08,724.603 1694.61,689.347 1608.43,597.618 1607.86,597.1 1538.27,505.889 1481.63,416.062 1480.63,414.16 1458.01,322.431 1447.32,230.702 1423,138.973 1403.75,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#f98c09; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1515.28,47.2441 1537.37,138.973 1566.86,230.702 1588.55,322.431 1607.86,370.518 1630.55,414.16 1710.54,505.889 1734.08,526.887 1848.38,597.618 1860.31,607.162 1931.61,689.347 1986.53,742.457 2030.38,781.076 2112.76,843.559 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#fbaa0e; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2112.76,664.893 2035.73,597.618 1986.53,557.67 1935,505.889 1860.31,423.06 1847.29,414.16 1764.28,322.431 1734.08,260.722 1723.38,230.702 1670.15,138.973 1634.33,47.2441 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#f8c931; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1799.26,47.2441 1860.31,125.534 1868.77,138.973 1920.04,230.702 1957.02,322.431 1986.53,369.044 2019.67,414.16 2112.76,500.307 "/>
+<polyline clip-path="url(#clip332)" style="stroke:#f1e864; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2112.76,289.305 2082.36,230.702 2017.57,138.973 1986.53,93.1516 1957.56,47.2441 "/>
+<defs>
+  <clipPath id="clip333">
+    <rect x="2160" y="47" width="73" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip333)" style="stroke:#0a0723; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,1337.18 2232.76,1337.18 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#200c4a; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,1251.19 2232.76,1251.19 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#3c0965; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,1165.19 2232.76,1165.19 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#570f6d; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,1079.2 2232.76,1079.2 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#70196e; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,993.2 2232.76,993.2 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#892269; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,907.204 2232.76,907.204 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#a32b61; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,821.208 2232.76,821.208 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#bb3754; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,735.212 2232.76,735.212 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#d04544; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,649.216 2232.76,649.216 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#e35832; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,563.22 2232.76,563.22 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#f0701e; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,477.224 2232.76,477.224 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#f98c09; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,391.228 2232.76,391.228 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#fbaa0e; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,305.232 2232.76,305.232 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#f8c931; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,219.236 2232.76,219.236 "/>
+<polyline clip-path="url(#clip333)" style="stroke:#f1e864; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2160.76,133.24 2232.76,133.24 "/>
+<path clip-path="url(#clip330)" d="M2274.38 1256.53 L2290.7 1256.53 L2290.7 1260.47 L2268.76 1260.47 L2268.76 1256.53 Q2271.42 1253.78 2276 1249.15 Q2280.61 1244.5 2281.79 1243.15 Q2284.03 1240.63 2284.91 1238.9 Q2285.82 1237.14 2285.82 1235.45 Q2285.82 1232.69 2283.87 1230.96 Q2281.95 1229.22 2278.85 1229.22 Q2276.65 1229.22 2274.2 1229.98 Q2271.77 1230.75 2268.99 1232.3 L2268.99 1227.58 Q2271.81 1226.44 2274.27 1225.86 Q2276.72 1225.28 2278.76 1225.28 Q2284.13 1225.28 2287.32 1227.97 Q2290.52 1230.65 2290.52 1235.15 Q2290.52 1237.27 2289.7 1239.2 Q2288.92 1241.09 2286.81 1243.69 Q2286.23 1244.36 2283.13 1247.58 Q2280.03 1250.77 2274.38 1256.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2300.56 1225.91 L2318.92 1225.91 L2318.92 1229.84 L2304.84 1229.84 L2304.84 1238.32 Q2305.86 1237.97 2306.88 1237.81 Q2307.9 1237.62 2308.92 1237.62 Q2314.7 1237.62 2318.08 1240.79 Q2321.46 1243.96 2321.46 1249.38 Q2321.46 1254.96 2317.99 1258.06 Q2314.52 1261.14 2308.2 1261.14 Q2306.02 1261.14 2303.76 1260.77 Q2301.51 1260.4 2299.1 1259.66 L2299.1 1254.96 Q2301.19 1256.09 2303.41 1256.65 Q2305.63 1257.21 2308.11 1257.21 Q2312.11 1257.21 2314.45 1255.1 Q2316.79 1252.99 2316.79 1249.38 Q2316.79 1245.77 2314.45 1243.66 Q2312.11 1241.56 2308.11 1241.56 Q2306.23 1241.56 2304.36 1241.97 Q2302.51 1242.39 2300.56 1243.27 L2300.56 1225.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2340.68 1228.99 Q2337.07 1228.99 2335.24 1232.55 Q2333.43 1236.09 2333.43 1243.22 Q2333.43 1250.33 2335.24 1253.9 Q2337.07 1257.44 2340.68 1257.44 Q2344.31 1257.44 2346.12 1253.9 Q2347.95 1250.33 2347.95 1243.22 Q2347.95 1236.09 2346.12 1232.55 Q2344.31 1228.99 2340.68 1228.99 M2340.68 1225.28 Q2346.49 1225.28 2349.54 1229.89 Q2352.62 1234.47 2352.62 1243.22 Q2352.62 1251.95 2349.54 1256.56 Q2346.49 1261.14 2340.68 1261.14 Q2334.87 1261.14 2331.79 1256.56 Q2328.73 1251.95 2328.73 1243.22 Q2328.73 1234.47 2331.79 1229.89 Q2334.87 1225.28 2340.68 1225.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2370.84 1228.99 Q2367.23 1228.99 2365.4 1232.55 Q2363.59 1236.09 2363.59 1243.22 Q2363.59 1250.33 2365.4 1253.9 Q2367.23 1257.44 2370.84 1257.44 Q2374.47 1257.44 2376.28 1253.9 Q2378.11 1250.33 2378.11 1243.22 Q2378.11 1236.09 2376.28 1232.55 Q2374.47 1228.99 2370.84 1228.99 M2370.84 1225.28 Q2376.65 1225.28 2379.7 1229.89 Q2382.78 1234.47 2382.78 1243.22 Q2382.78 1251.95 2379.7 1256.56 Q2376.65 1261.14 2370.84 1261.14 Q2365.03 1261.14 2361.95 1256.56 Q2358.89 1251.95 2358.89 1243.22 Q2358.89 1234.47 2361.95 1229.89 Q2365.03 1225.28 2370.84 1225.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2401 1228.99 Q2397.39 1228.99 2395.56 1232.55 Q2393.76 1236.09 2393.76 1243.22 Q2393.76 1250.33 2395.56 1253.9 Q2397.39 1257.44 2401 1257.44 Q2404.63 1257.44 2406.44 1253.9 Q2408.27 1250.33 2408.27 1243.22 Q2408.27 1236.09 2406.44 1232.55 Q2404.63 1228.99 2401 1228.99 M2401 1225.28 Q2406.81 1225.28 2409.87 1229.89 Q2412.94 1234.47 2412.94 1243.22 Q2412.94 1251.95 2409.87 1256.56 Q2406.81 1261.14 2401 1261.14 Q2395.19 1261.14 2392.11 1256.56 Q2389.06 1251.95 2389.06 1243.22 Q2389.06 1234.47 2392.11 1229.89 Q2395.19 1225.28 2401 1225.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2431.16 1228.99 Q2427.55 1228.99 2425.72 1232.55 Q2423.92 1236.09 2423.92 1243.22 Q2423.92 1250.33 2425.72 1253.9 Q2427.55 1257.44 2431.16 1257.44 Q2434.8 1257.44 2436.6 1253.9 Q2438.43 1250.33 2438.43 1243.22 Q2438.43 1236.09 2436.6 1232.55 Q2434.8 1228.99 2431.16 1228.99 M2431.16 1225.28 Q2436.97 1225.28 2440.03 1229.89 Q2443.11 1234.47 2443.11 1243.22 Q2443.11 1251.95 2440.03 1256.56 Q2436.97 1261.14 2431.16 1261.14 Q2425.35 1261.14 2422.27 1256.56 Q2419.22 1251.95 2419.22 1243.22 Q2419.22 1234.47 2422.27 1229.89 Q2425.35 1225.28 2431.16 1225.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2461.32 1228.99 Q2457.71 1228.99 2455.88 1232.55 Q2454.08 1236.09 2454.08 1243.22 Q2454.08 1250.33 2455.88 1253.9 Q2457.71 1257.44 2461.32 1257.44 Q2464.96 1257.44 2466.76 1253.9 Q2468.59 1250.33 2468.59 1243.22 Q2468.59 1236.09 2466.76 1232.55 Q2464.96 1228.99 2461.32 1228.99 M2461.32 1225.28 Q2467.13 1225.28 2470.19 1229.89 Q2473.27 1234.47 2473.27 1243.22 Q2473.27 1251.95 2470.19 1256.56 Q2467.13 1261.14 2461.32 1261.14 Q2455.51 1261.14 2452.44 1256.56 Q2449.38 1251.95 2449.38 1243.22 Q2449.38 1234.47 2452.44 1229.89 Q2455.51 1225.28 2461.32 1225.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2270.21 1049 L2288.57 1049 L2288.57 1052.93 L2274.5 1052.93 L2274.5 1061.4 Q2275.52 1061.06 2276.53 1060.89 Q2277.55 1060.71 2278.57 1060.71 Q2284.36 1060.71 2287.74 1063.88 Q2291.12 1067.05 2291.12 1072.47 Q2291.12 1078.05 2287.64 1081.15 Q2284.17 1084.23 2277.85 1084.23 Q2275.68 1084.23 2273.41 1083.86 Q2271.16 1083.49 2268.76 1082.75 L2268.76 1078.05 Q2270.84 1079.18 2273.06 1079.74 Q2275.28 1080.29 2277.76 1080.29 Q2281.77 1080.29 2284.1 1078.19 Q2286.44 1076.08 2286.44 1072.47 Q2286.44 1068.86 2284.1 1066.75 Q2281.77 1064.64 2277.76 1064.64 Q2275.89 1064.64 2274.01 1065.06 Q2272.16 1065.48 2270.21 1066.36 L2270.21 1049 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2310.33 1052.07 Q2306.72 1052.07 2304.89 1055.64 Q2303.08 1059.18 2303.08 1066.31 Q2303.08 1073.42 2304.89 1076.98 Q2306.72 1080.52 2310.33 1080.52 Q2313.96 1080.52 2315.77 1076.98 Q2317.6 1073.42 2317.6 1066.31 Q2317.6 1059.18 2315.77 1055.64 Q2313.96 1052.07 2310.33 1052.07 M2310.33 1048.37 Q2316.14 1048.37 2319.2 1052.98 Q2322.27 1057.56 2322.27 1066.31 Q2322.27 1075.04 2319.2 1079.64 Q2316.14 1084.23 2310.33 1084.23 Q2304.52 1084.23 2301.44 1079.64 Q2298.39 1075.04 2298.39 1066.31 Q2298.39 1057.56 2301.44 1052.98 Q2304.52 1048.37 2310.33 1048.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2340.49 1052.07 Q2336.88 1052.07 2335.05 1055.64 Q2333.25 1059.18 2333.25 1066.31 Q2333.25 1073.42 2335.05 1076.98 Q2336.88 1080.52 2340.49 1080.52 Q2344.13 1080.52 2345.93 1076.98 Q2347.76 1073.42 2347.76 1066.31 Q2347.76 1059.18 2345.93 1055.64 Q2344.13 1052.07 2340.49 1052.07 M2340.49 1048.37 Q2346.3 1048.37 2349.36 1052.98 Q2352.44 1057.56 2352.44 1066.31 Q2352.44 1075.04 2349.36 1079.64 Q2346.3 1084.23 2340.49 1084.23 Q2334.68 1084.23 2331.6 1079.64 Q2328.55 1075.04 2328.55 1066.31 Q2328.55 1057.56 2331.6 1052.98 Q2334.68 1048.37 2340.49 1048.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2370.65 1052.07 Q2367.04 1052.07 2365.21 1055.64 Q2363.41 1059.18 2363.41 1066.31 Q2363.41 1073.42 2365.21 1076.98 Q2367.04 1080.52 2370.65 1080.52 Q2374.29 1080.52 2376.09 1076.98 Q2377.92 1073.42 2377.92 1066.31 Q2377.92 1059.18 2376.09 1055.64 Q2374.29 1052.07 2370.65 1052.07 M2370.65 1048.37 Q2376.46 1048.37 2379.52 1052.98 Q2382.6 1057.56 2382.6 1066.31 Q2382.6 1075.04 2379.52 1079.64 Q2376.46 1084.23 2370.65 1084.23 Q2364.84 1084.23 2361.76 1079.64 Q2358.71 1075.04 2358.71 1066.31 Q2358.71 1057.56 2361.76 1052.98 Q2364.84 1048.37 2370.65 1048.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2400.82 1052.07 Q2397.2 1052.07 2395.38 1055.64 Q2393.57 1059.18 2393.57 1066.31 Q2393.57 1073.42 2395.38 1076.98 Q2397.2 1080.52 2400.82 1080.52 Q2404.45 1080.52 2406.26 1076.98 Q2408.08 1073.42 2408.08 1066.31 Q2408.08 1059.18 2406.26 1055.64 Q2404.45 1052.07 2400.82 1052.07 M2400.82 1048.37 Q2406.63 1048.37 2409.68 1052.98 Q2412.76 1057.56 2412.76 1066.31 Q2412.76 1075.04 2409.68 1079.64 Q2406.63 1084.23 2400.82 1084.23 Q2395.01 1084.23 2391.93 1079.64 Q2388.87 1075.04 2388.87 1066.31 Q2388.87 1057.56 2391.93 1052.98 Q2395.01 1048.37 2400.82 1048.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2430.98 1052.07 Q2427.37 1052.07 2425.54 1055.64 Q2423.73 1059.18 2423.73 1066.31 Q2423.73 1073.42 2425.54 1076.98 Q2427.37 1080.52 2430.98 1080.52 Q2434.61 1080.52 2436.42 1076.98 Q2438.25 1073.42 2438.25 1066.31 Q2438.25 1059.18 2436.42 1055.64 Q2434.61 1052.07 2430.98 1052.07 M2430.98 1048.37 Q2436.79 1048.37 2439.84 1052.98 Q2442.92 1057.56 2442.92 1066.31 Q2442.92 1075.04 2439.84 1079.64 Q2436.79 1084.23 2430.98 1084.23 Q2425.17 1084.23 2422.09 1079.64 Q2419.03 1075.04 2419.03 1066.31 Q2419.03 1057.56 2422.09 1052.98 Q2425.17 1048.37 2430.98 1048.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2461.14 1052.07 Q2457.53 1052.07 2455.7 1055.64 Q2453.89 1059.18 2453.89 1066.31 Q2453.89 1073.42 2455.7 1076.98 Q2457.53 1080.52 2461.14 1080.52 Q2464.77 1080.52 2466.58 1076.98 Q2468.41 1073.42 2468.41 1066.31 Q2468.41 1059.18 2466.58 1055.64 Q2464.77 1052.07 2461.14 1052.07 M2461.14 1048.37 Q2466.95 1048.37 2470 1052.98 Q2473.08 1057.56 2473.08 1066.31 Q2473.08 1075.04 2470 1079.64 Q2466.95 1084.23 2461.14 1084.23 Q2455.33 1084.23 2452.25 1079.64 Q2449.19 1075.04 2449.19 1066.31 Q2449.19 1057.56 2452.25 1052.98 Q2455.33 1048.37 2461.14 1048.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2268.76 872.082 L2290.98 872.082 L2290.98 874.073 L2278.43 906.642 L2273.55 906.642 L2285.35 876.017 L2268.76 876.017 L2268.76 872.082 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2300.14 872.082 L2318.5 872.082 L2318.5 876.017 L2304.43 876.017 L2304.43 884.489 Q2305.45 884.142 2306.46 883.98 Q2307.48 883.795 2308.5 883.795 Q2314.29 883.795 2317.67 886.966 Q2321.05 890.137 2321.05 895.554 Q2321.05 901.133 2317.58 904.234 Q2314.1 907.313 2307.78 907.313 Q2305.61 907.313 2303.34 906.943 Q2301.09 906.572 2298.69 905.832 L2298.69 901.133 Q2300.77 902.267 2302.99 902.822 Q2305.21 903.378 2307.69 903.378 Q2311.7 903.378 2314.03 901.272 Q2316.37 899.165 2316.37 895.554 Q2316.37 891.943 2314.03 889.836 Q2311.7 887.73 2307.69 887.73 Q2305.82 887.73 2303.94 888.147 Q2302.09 888.563 2300.14 889.443 L2300.14 872.082 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2340.26 875.161 Q2336.65 875.161 2334.82 878.725 Q2333.01 882.267 2333.01 889.397 Q2333.01 896.503 2334.82 900.068 Q2336.65 903.609 2340.26 903.609 Q2343.89 903.609 2345.7 900.068 Q2347.53 896.503 2347.53 889.397 Q2347.53 882.267 2345.7 878.725 Q2343.89 875.161 2340.26 875.161 M2340.26 871.457 Q2346.07 871.457 2349.13 876.063 Q2352.2 880.647 2352.2 889.397 Q2352.2 898.123 2349.13 902.73 Q2346.07 907.313 2340.26 907.313 Q2334.45 907.313 2331.37 902.73 Q2328.32 898.123 2328.32 889.397 Q2328.32 880.647 2331.37 876.063 Q2334.45 871.457 2340.26 871.457 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2370.42 875.161 Q2366.81 875.161 2364.98 878.725 Q2363.18 882.267 2363.18 889.397 Q2363.18 896.503 2364.98 900.068 Q2366.81 903.609 2370.42 903.609 Q2374.06 903.609 2375.86 900.068 Q2377.69 896.503 2377.69 889.397 Q2377.69 882.267 2375.86 878.725 Q2374.06 875.161 2370.42 875.161 M2370.42 871.457 Q2376.23 871.457 2379.29 876.063 Q2382.37 880.647 2382.37 889.397 Q2382.37 898.123 2379.29 902.73 Q2376.23 907.313 2370.42 907.313 Q2364.61 907.313 2361.53 902.73 Q2358.48 898.123 2358.48 889.397 Q2358.48 880.647 2361.53 876.063 Q2364.61 871.457 2370.42 871.457 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2400.58 875.161 Q2396.97 875.161 2395.14 878.725 Q2393.34 882.267 2393.34 889.397 Q2393.34 896.503 2395.14 900.068 Q2396.97 903.609 2400.58 903.609 Q2404.22 903.609 2406.02 900.068 Q2407.85 896.503 2407.85 889.397 Q2407.85 882.267 2406.02 878.725 Q2404.22 875.161 2400.58 875.161 M2400.58 871.457 Q2406.39 871.457 2409.45 876.063 Q2412.53 880.647 2412.53 889.397 Q2412.53 898.123 2409.45 902.73 Q2406.39 907.313 2400.58 907.313 Q2394.77 907.313 2391.7 902.73 Q2388.64 898.123 2388.64 889.397 Q2388.64 880.647 2391.7 876.063 Q2394.77 871.457 2400.58 871.457 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2430.75 875.161 Q2427.13 875.161 2425.31 878.725 Q2423.5 882.267 2423.5 889.397 Q2423.5 896.503 2425.31 900.068 Q2427.13 903.609 2430.75 903.609 Q2434.38 903.609 2436.19 900.068 Q2438.01 896.503 2438.01 889.397 Q2438.01 882.267 2436.19 878.725 Q2434.38 875.161 2430.75 875.161 M2430.75 871.457 Q2436.56 871.457 2439.61 876.063 Q2442.69 880.647 2442.69 889.397 Q2442.69 898.123 2439.61 902.73 Q2436.56 907.313 2430.75 907.313 Q2424.94 907.313 2421.86 902.73 Q2418.8 898.123 2418.8 889.397 Q2418.8 880.647 2421.86 876.063 Q2424.94 871.457 2430.75 871.457 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2460.91 875.161 Q2457.3 875.161 2455.47 878.725 Q2453.66 882.267 2453.66 889.397 Q2453.66 896.503 2455.47 900.068 Q2457.3 903.609 2460.91 903.609 Q2464.54 903.609 2466.35 900.068 Q2468.18 896.503 2468.18 889.397 Q2468.18 882.267 2466.35 878.725 Q2464.54 875.161 2460.91 875.161 M2460.91 871.457 Q2466.72 871.457 2469.77 876.063 Q2472.85 880.647 2472.85 889.397 Q2472.85 898.123 2469.77 902.73 Q2466.72 907.313 2460.91 907.313 Q2455.1 907.313 2452.02 902.73 Q2448.96 898.123 2448.96 889.397 Q2448.96 880.647 2452.02 876.063 Q2455.1 871.457 2460.91 871.457 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2269.43 725.793 L2277.07 725.793 L2277.07 699.427 L2268.76 701.094 L2268.76 696.835 L2277.02 695.168 L2281.7 695.168 L2281.7 725.793 L2289.33 725.793 L2289.33 729.728 L2269.43 729.728 L2269.43 725.793 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2308.78 698.247 Q2305.17 698.247 2303.34 701.812 Q2301.53 705.353 2301.53 712.483 Q2301.53 719.589 2303.34 723.154 Q2305.17 726.696 2308.78 726.696 Q2312.41 726.696 2314.22 723.154 Q2316.05 719.589 2316.05 712.483 Q2316.05 705.353 2314.22 701.812 Q2312.41 698.247 2308.78 698.247 M2308.78 694.543 Q2314.59 694.543 2317.64 699.15 Q2320.72 703.733 2320.72 712.483 Q2320.72 721.21 2317.64 725.816 Q2314.59 730.399 2308.78 730.399 Q2302.97 730.399 2299.89 725.816 Q2296.83 721.21 2296.83 712.483 Q2296.83 703.733 2299.89 699.15 Q2302.97 694.543 2308.78 694.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2338.94 698.247 Q2335.33 698.247 2333.5 701.812 Q2331.7 705.353 2331.7 712.483 Q2331.7 719.589 2333.5 723.154 Q2335.33 726.696 2338.94 726.696 Q2342.57 726.696 2344.38 723.154 Q2346.21 719.589 2346.21 712.483 Q2346.21 705.353 2344.38 701.812 Q2342.57 698.247 2338.94 698.247 M2338.94 694.543 Q2344.75 694.543 2347.81 699.15 Q2350.89 703.733 2350.89 712.483 Q2350.89 721.21 2347.81 725.816 Q2344.75 730.399 2338.94 730.399 Q2333.13 730.399 2330.05 725.816 Q2327 721.21 2327 712.483 Q2327 703.733 2330.05 699.15 Q2333.13 694.543 2338.94 694.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2369.1 698.247 Q2365.49 698.247 2363.66 701.812 Q2361.86 705.353 2361.86 712.483 Q2361.86 719.589 2363.66 723.154 Q2365.49 726.696 2369.1 726.696 Q2372.74 726.696 2374.54 723.154 Q2376.37 719.589 2376.37 712.483 Q2376.37 705.353 2374.54 701.812 Q2372.74 698.247 2369.1 698.247 M2369.1 694.543 Q2374.91 694.543 2377.97 699.15 Q2381.05 703.733 2381.05 712.483 Q2381.05 721.21 2377.97 725.816 Q2374.91 730.399 2369.1 730.399 Q2363.29 730.399 2360.21 725.816 Q2357.16 721.21 2357.16 712.483 Q2357.16 703.733 2360.21 699.15 Q2363.29 694.543 2369.1 694.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2399.26 698.247 Q2395.65 698.247 2393.82 701.812 Q2392.02 705.353 2392.02 712.483 Q2392.02 719.589 2393.82 723.154 Q2395.65 726.696 2399.26 726.696 Q2402.9 726.696 2404.7 723.154 Q2406.53 719.589 2406.53 712.483 Q2406.53 705.353 2404.7 701.812 Q2402.9 698.247 2399.26 698.247 M2399.26 694.543 Q2405.07 694.543 2408.13 699.15 Q2411.21 703.733 2411.21 712.483 Q2411.21 721.21 2408.13 725.816 Q2405.07 730.399 2399.26 730.399 Q2393.45 730.399 2390.38 725.816 Q2387.32 721.21 2387.32 712.483 Q2387.32 703.733 2390.38 699.15 Q2393.45 694.543 2399.26 694.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2429.43 698.247 Q2425.82 698.247 2423.99 701.812 Q2422.18 705.353 2422.18 712.483 Q2422.18 719.589 2423.99 723.154 Q2425.82 726.696 2429.43 726.696 Q2433.06 726.696 2434.87 723.154 Q2436.69 719.589 2436.69 712.483 Q2436.69 705.353 2434.87 701.812 Q2433.06 698.247 2429.43 698.247 M2429.43 694.543 Q2435.24 694.543 2438.29 699.15 Q2441.37 703.733 2441.37 712.483 Q2441.37 721.21 2438.29 725.816 Q2435.24 730.399 2429.43 730.399 Q2423.62 730.399 2420.54 725.816 Q2417.48 721.21 2417.48 712.483 Q2417.48 703.733 2420.54 699.15 Q2423.62 694.543 2429.43 694.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2459.59 698.247 Q2455.98 698.247 2454.15 701.812 Q2452.34 705.353 2452.34 712.483 Q2452.34 719.589 2454.15 723.154 Q2455.98 726.696 2459.59 726.696 Q2463.22 726.696 2465.03 723.154 Q2466.86 719.589 2466.86 712.483 Q2466.86 705.353 2465.03 701.812 Q2463.22 698.247 2459.59 698.247 M2459.59 694.543 Q2465.4 694.543 2468.45 699.15 Q2471.53 703.733 2471.53 712.483 Q2471.53 721.21 2468.45 725.816 Q2465.4 730.399 2459.59 730.399 Q2453.78 730.399 2450.7 725.816 Q2447.64 721.21 2447.64 712.483 Q2447.64 703.733 2450.7 699.15 Q2453.78 694.543 2459.59 694.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2489.75 698.247 Q2486.14 698.247 2484.31 701.812 Q2482.5 705.353 2482.5 712.483 Q2482.5 719.589 2484.31 723.154 Q2486.14 726.696 2489.75 726.696 Q2493.38 726.696 2495.19 723.154 Q2497.02 719.589 2497.02 712.483 Q2497.02 705.353 2495.19 701.812 Q2493.38 698.247 2489.75 698.247 M2489.75 694.543 Q2495.56 694.543 2498.62 699.15 Q2501.69 703.733 2501.69 712.483 Q2501.69 721.21 2498.62 725.816 Q2495.56 730.399 2489.75 730.399 Q2483.94 730.399 2480.86 725.816 Q2477.81 721.21 2477.81 712.483 Q2477.81 703.733 2480.86 699.15 Q2483.94 694.543 2489.75 694.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2269.43 548.879 L2277.07 548.879 L2277.07 522.514 L2268.76 524.18 L2268.76 519.921 L2277.02 518.254 L2281.7 518.254 L2281.7 548.879 L2289.33 548.879 L2289.33 552.814 L2269.43 552.814 L2269.43 548.879 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2302.81 548.879 L2319.13 548.879 L2319.13 552.814 L2297.18 552.814 L2297.18 548.879 Q2299.84 546.125 2304.43 541.495 Q2309.03 536.842 2310.21 535.5 Q2312.46 532.977 2313.34 531.241 Q2314.24 529.481 2314.24 527.791 Q2314.24 525.037 2312.3 523.301 Q2310.38 521.565 2307.27 521.565 Q2305.08 521.565 2302.62 522.329 Q2300.19 523.092 2297.41 524.643 L2297.41 519.921 Q2300.24 518.787 2302.69 518.208 Q2305.14 517.629 2307.18 517.629 Q2312.55 517.629 2315.75 520.315 Q2318.94 523 2318.94 527.491 Q2318.94 529.62 2318.13 531.541 Q2317.34 533.44 2315.24 536.032 Q2314.66 536.703 2311.56 539.921 Q2308.45 543.115 2302.81 548.879 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2328.99 518.254 L2347.34 518.254 L2347.34 522.19 L2333.27 522.19 L2333.27 530.662 Q2334.29 530.315 2335.31 530.153 Q2336.32 529.967 2337.34 529.967 Q2343.13 529.967 2346.51 533.139 Q2349.89 536.31 2349.89 541.727 Q2349.89 547.305 2346.42 550.407 Q2342.95 553.486 2336.63 553.486 Q2334.45 553.486 2332.18 553.115 Q2329.94 552.745 2327.53 552.004 L2327.53 547.305 Q2329.61 548.439 2331.83 548.995 Q2334.06 549.551 2336.53 549.551 Q2340.54 549.551 2342.88 547.444 Q2345.21 545.338 2345.21 541.727 Q2345.21 538.115 2342.88 536.009 Q2340.54 533.903 2336.53 533.903 Q2334.66 533.903 2332.78 534.319 Q2330.93 534.736 2328.99 535.615 L2328.99 518.254 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2369.1 521.333 Q2365.49 521.333 2363.66 524.898 Q2361.86 528.44 2361.86 535.569 Q2361.86 542.676 2363.66 546.24 Q2365.49 549.782 2369.1 549.782 Q2372.74 549.782 2374.54 546.24 Q2376.37 542.676 2376.37 535.569 Q2376.37 528.44 2374.54 524.898 Q2372.74 521.333 2369.1 521.333 M2369.1 517.629 Q2374.91 517.629 2377.97 522.236 Q2381.05 526.819 2381.05 535.569 Q2381.05 544.296 2377.97 548.902 Q2374.91 553.486 2369.1 553.486 Q2363.29 553.486 2360.21 548.902 Q2357.16 544.296 2357.16 535.569 Q2357.16 526.819 2360.21 522.236 Q2363.29 517.629 2369.1 517.629 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2399.26 521.333 Q2395.65 521.333 2393.82 524.898 Q2392.02 528.44 2392.02 535.569 Q2392.02 542.676 2393.82 546.24 Q2395.65 549.782 2399.26 549.782 Q2402.9 549.782 2404.7 546.24 Q2406.53 542.676 2406.53 535.569 Q2406.53 528.44 2404.7 524.898 Q2402.9 521.333 2399.26 521.333 M2399.26 517.629 Q2405.07 517.629 2408.13 522.236 Q2411.21 526.819 2411.21 535.569 Q2411.21 544.296 2408.13 548.902 Q2405.07 553.486 2399.26 553.486 Q2393.45 553.486 2390.38 548.902 Q2387.32 544.296 2387.32 535.569 Q2387.32 526.819 2390.38 522.236 Q2393.45 517.629 2399.26 517.629 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2429.43 521.333 Q2425.82 521.333 2423.99 524.898 Q2422.18 528.44 2422.18 535.569 Q2422.18 542.676 2423.99 546.24 Q2425.82 549.782 2429.43 549.782 Q2433.06 549.782 2434.87 546.24 Q2436.69 542.676 2436.69 535.569 Q2436.69 528.44 2434.87 524.898 Q2433.06 521.333 2429.43 521.333 M2429.43 517.629 Q2435.24 517.629 2438.29 522.236 Q2441.37 526.819 2441.37 535.569 Q2441.37 544.296 2438.29 548.902 Q2435.24 553.486 2429.43 553.486 Q2423.62 553.486 2420.54 548.902 Q2417.48 544.296 2417.48 535.569 Q2417.48 526.819 2420.54 522.236 Q2423.62 517.629 2429.43 517.629 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2459.59 521.333 Q2455.98 521.333 2454.15 524.898 Q2452.34 528.44 2452.34 535.569 Q2452.34 542.676 2454.15 546.24 Q2455.98 549.782 2459.59 549.782 Q2463.22 549.782 2465.03 546.24 Q2466.86 542.676 2466.86 535.569 Q2466.86 528.44 2465.03 524.898 Q2463.22 521.333 2459.59 521.333 M2459.59 517.629 Q2465.4 517.629 2468.45 522.236 Q2471.53 526.819 2471.53 535.569 Q2471.53 544.296 2468.45 548.902 Q2465.4 553.486 2459.59 553.486 Q2453.78 553.486 2450.7 548.902 Q2447.64 544.296 2447.64 535.569 Q2447.64 526.819 2450.7 522.236 Q2453.78 517.629 2459.59 517.629 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2489.75 521.333 Q2486.14 521.333 2484.31 524.898 Q2482.5 528.44 2482.5 535.569 Q2482.5 542.676 2484.31 546.24 Q2486.14 549.782 2489.75 549.782 Q2493.38 549.782 2495.19 546.24 Q2497.02 542.676 2497.02 535.569 Q2497.02 528.44 2495.19 524.898 Q2493.38 521.333 2489.75 521.333 M2489.75 517.629 Q2495.56 517.629 2498.62 522.236 Q2501.69 526.819 2501.69 535.569 Q2501.69 544.296 2498.62 548.902 Q2495.56 553.486 2489.75 553.486 Q2483.94 553.486 2480.86 548.902 Q2477.81 544.296 2477.81 535.569 Q2477.81 526.819 2480.86 522.236 Q2483.94 517.629 2489.75 517.629 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2269.43 371.966 L2277.07 371.966 L2277.07 345.6 L2268.76 347.267 L2268.76 343.007 L2277.02 341.341 L2281.7 341.341 L2281.7 371.966 L2289.33 371.966 L2289.33 375.901 L2269.43 375.901 L2269.43 371.966 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2298.83 341.341 L2317.18 341.341 L2317.18 345.276 L2303.11 345.276 L2303.11 353.748 Q2304.13 353.401 2305.14 353.239 Q2306.16 353.054 2307.18 353.054 Q2312.97 353.054 2316.35 356.225 Q2319.73 359.396 2319.73 364.813 Q2319.73 370.392 2316.26 373.493 Q2312.78 376.572 2306.46 376.572 Q2304.29 376.572 2302.02 376.202 Q2299.77 375.831 2297.37 375.091 L2297.37 370.392 Q2299.45 371.526 2301.67 372.081 Q2303.89 372.637 2306.37 372.637 Q2310.38 372.637 2312.71 370.53 Q2315.05 368.424 2315.05 364.813 Q2315.05 361.202 2312.71 359.095 Q2310.38 356.989 2306.37 356.989 Q2304.5 356.989 2302.62 357.405 Q2300.77 357.822 2298.83 358.702 L2298.83 341.341 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2338.94 344.419 Q2335.33 344.419 2333.5 347.984 Q2331.7 351.526 2331.7 358.655 Q2331.7 365.762 2333.5 369.327 Q2335.33 372.868 2338.94 372.868 Q2342.57 372.868 2344.38 369.327 Q2346.21 365.762 2346.21 358.655 Q2346.21 351.526 2344.38 347.984 Q2342.57 344.419 2338.94 344.419 M2338.94 340.716 Q2344.75 340.716 2347.81 345.322 Q2350.89 349.906 2350.89 358.655 Q2350.89 367.382 2347.81 371.989 Q2344.75 376.572 2338.94 376.572 Q2333.13 376.572 2330.05 371.989 Q2327 367.382 2327 358.655 Q2327 349.906 2330.05 345.322 Q2333.13 340.716 2338.94 340.716 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2369.1 344.419 Q2365.49 344.419 2363.66 347.984 Q2361.86 351.526 2361.86 358.655 Q2361.86 365.762 2363.66 369.327 Q2365.49 372.868 2369.1 372.868 Q2372.74 372.868 2374.54 369.327 Q2376.37 365.762 2376.37 358.655 Q2376.37 351.526 2374.54 347.984 Q2372.74 344.419 2369.1 344.419 M2369.1 340.716 Q2374.91 340.716 2377.97 345.322 Q2381.05 349.906 2381.05 358.655 Q2381.05 367.382 2377.97 371.989 Q2374.91 376.572 2369.1 376.572 Q2363.29 376.572 2360.21 371.989 Q2357.16 367.382 2357.16 358.655 Q2357.16 349.906 2360.21 345.322 Q2363.29 340.716 2369.1 340.716 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2399.26 344.419 Q2395.65 344.419 2393.82 347.984 Q2392.02 351.526 2392.02 358.655 Q2392.02 365.762 2393.82 369.327 Q2395.65 372.868 2399.26 372.868 Q2402.9 372.868 2404.7 369.327 Q2406.53 365.762 2406.53 358.655 Q2406.53 351.526 2404.7 347.984 Q2402.9 344.419 2399.26 344.419 M2399.26 340.716 Q2405.07 340.716 2408.13 345.322 Q2411.21 349.906 2411.21 358.655 Q2411.21 367.382 2408.13 371.989 Q2405.07 376.572 2399.26 376.572 Q2393.45 376.572 2390.38 371.989 Q2387.32 367.382 2387.32 358.655 Q2387.32 349.906 2390.38 345.322 Q2393.45 340.716 2399.26 340.716 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2429.43 344.419 Q2425.82 344.419 2423.99 347.984 Q2422.18 351.526 2422.18 358.655 Q2422.18 365.762 2423.99 369.327 Q2425.82 372.868 2429.43 372.868 Q2433.06 372.868 2434.87 369.327 Q2436.69 365.762 2436.69 358.655 Q2436.69 351.526 2434.87 347.984 Q2433.06 344.419 2429.43 344.419 M2429.43 340.716 Q2435.24 340.716 2438.29 345.322 Q2441.37 349.906 2441.37 358.655 Q2441.37 367.382 2438.29 371.989 Q2435.24 376.572 2429.43 376.572 Q2423.62 376.572 2420.54 371.989 Q2417.48 367.382 2417.48 358.655 Q2417.48 349.906 2420.54 345.322 Q2423.62 340.716 2429.43 340.716 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2459.59 344.419 Q2455.98 344.419 2454.15 347.984 Q2452.34 351.526 2452.34 358.655 Q2452.34 365.762 2454.15 369.327 Q2455.98 372.868 2459.59 372.868 Q2463.22 372.868 2465.03 369.327 Q2466.86 365.762 2466.86 358.655 Q2466.86 351.526 2465.03 347.984 Q2463.22 344.419 2459.59 344.419 M2459.59 340.716 Q2465.4 340.716 2468.45 345.322 Q2471.53 349.906 2471.53 358.655 Q2471.53 367.382 2468.45 371.989 Q2465.4 376.572 2459.59 376.572 Q2453.78 376.572 2450.7 371.989 Q2447.64 367.382 2447.64 358.655 Q2447.64 349.906 2450.7 345.322 Q2453.78 340.716 2459.59 340.716 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2489.75 344.419 Q2486.14 344.419 2484.31 347.984 Q2482.5 351.526 2482.5 358.655 Q2482.5 365.762 2484.31 369.327 Q2486.14 372.868 2489.75 372.868 Q2493.38 372.868 2495.19 369.327 Q2497.02 365.762 2497.02 358.655 Q2497.02 351.526 2495.19 347.984 Q2493.38 344.419 2489.75 344.419 M2489.75 340.716 Q2495.56 340.716 2498.62 345.322 Q2501.69 349.906 2501.69 358.655 Q2501.69 367.382 2498.62 371.989 Q2495.56 376.572 2489.75 376.572 Q2483.94 376.572 2480.86 371.989 Q2477.81 367.382 2477.81 358.655 Q2477.81 349.906 2480.86 345.322 Q2483.94 340.716 2489.75 340.716 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2269.43 195.052 L2277.07 195.052 L2277.07 168.686 L2268.76 170.353 L2268.76 166.094 L2277.02 164.427 L2281.7 164.427 L2281.7 195.052 L2289.33 195.052 L2289.33 198.987 L2269.43 198.987 L2269.43 195.052 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2297.6 164.427 L2319.82 164.427 L2319.82 166.418 L2307.27 198.987 L2302.39 198.987 L2314.2 168.362 L2297.6 168.362 L2297.6 164.427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2328.99 164.427 L2347.34 164.427 L2347.34 168.362 L2333.27 168.362 L2333.27 176.834 Q2334.29 176.487 2335.31 176.325 Q2336.32 176.14 2337.34 176.14 Q2343.13 176.14 2346.51 179.311 Q2349.89 182.483 2349.89 187.899 Q2349.89 193.478 2346.42 196.58 Q2342.95 199.658 2336.63 199.658 Q2334.45 199.658 2332.18 199.288 Q2329.94 198.918 2327.53 198.177 L2327.53 193.478 Q2329.61 194.612 2331.83 195.168 Q2334.06 195.723 2336.53 195.723 Q2340.54 195.723 2342.88 193.617 Q2345.21 191.51 2345.21 187.899 Q2345.21 184.288 2342.88 182.182 Q2340.54 180.075 2336.53 180.075 Q2334.66 180.075 2332.78 180.492 Q2330.93 180.908 2328.99 181.788 L2328.99 164.427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2369.1 167.506 Q2365.49 167.506 2363.66 171.071 Q2361.86 174.612 2361.86 181.742 Q2361.86 188.848 2363.66 192.413 Q2365.49 195.955 2369.1 195.955 Q2372.74 195.955 2374.54 192.413 Q2376.37 188.848 2376.37 181.742 Q2376.37 174.612 2374.54 171.071 Q2372.74 167.506 2369.1 167.506 M2369.1 163.802 Q2374.91 163.802 2377.97 168.409 Q2381.05 172.992 2381.05 181.742 Q2381.05 190.469 2377.97 195.075 Q2374.91 199.658 2369.1 199.658 Q2363.29 199.658 2360.21 195.075 Q2357.16 190.469 2357.16 181.742 Q2357.16 172.992 2360.21 168.409 Q2363.29 163.802 2369.1 163.802 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2399.26 167.506 Q2395.65 167.506 2393.82 171.071 Q2392.02 174.612 2392.02 181.742 Q2392.02 188.848 2393.82 192.413 Q2395.65 195.955 2399.26 195.955 Q2402.9 195.955 2404.7 192.413 Q2406.53 188.848 2406.53 181.742 Q2406.53 174.612 2404.7 171.071 Q2402.9 167.506 2399.26 167.506 M2399.26 163.802 Q2405.07 163.802 2408.13 168.409 Q2411.21 172.992 2411.21 181.742 Q2411.21 190.469 2408.13 195.075 Q2405.07 199.658 2399.26 199.658 Q2393.45 199.658 2390.38 195.075 Q2387.32 190.469 2387.32 181.742 Q2387.32 172.992 2390.38 168.409 Q2393.45 163.802 2399.26 163.802 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2429.43 167.506 Q2425.82 167.506 2423.99 171.071 Q2422.18 174.612 2422.18 181.742 Q2422.18 188.848 2423.99 192.413 Q2425.82 195.955 2429.43 195.955 Q2433.06 195.955 2434.87 192.413 Q2436.69 188.848 2436.69 181.742 Q2436.69 174.612 2434.87 171.071 Q2433.06 167.506 2429.43 167.506 M2429.43 163.802 Q2435.24 163.802 2438.29 168.409 Q2441.37 172.992 2441.37 181.742 Q2441.37 190.469 2438.29 195.075 Q2435.24 199.658 2429.43 199.658 Q2423.62 199.658 2420.54 195.075 Q2417.48 190.469 2417.48 181.742 Q2417.48 172.992 2420.54 168.409 Q2423.62 163.802 2429.43 163.802 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2459.59 167.506 Q2455.98 167.506 2454.15 171.071 Q2452.34 174.612 2452.34 181.742 Q2452.34 188.848 2454.15 192.413 Q2455.98 195.955 2459.59 195.955 Q2463.22 195.955 2465.03 192.413 Q2466.86 188.848 2466.86 181.742 Q2466.86 174.612 2465.03 171.071 Q2463.22 167.506 2459.59 167.506 M2459.59 163.802 Q2465.4 163.802 2468.45 168.409 Q2471.53 172.992 2471.53 181.742 Q2471.53 190.469 2468.45 195.075 Q2465.4 199.658 2459.59 199.658 Q2453.78 199.658 2450.7 195.075 Q2447.64 190.469 2447.64 181.742 Q2447.64 172.992 2450.7 168.409 Q2453.78 163.802 2459.59 163.802 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip330)" d="M2489.75 167.506 Q2486.14 167.506 2484.31 171.071 Q2482.5 174.612 2482.5 181.742 Q2482.5 188.848 2484.31 192.413 Q2486.14 195.955 2489.75 195.955 Q2493.38 195.955 2495.19 192.413 Q2497.02 188.848 2497.02 181.742 Q2497.02 174.612 2495.19 171.071 Q2493.38 167.506 2489.75 167.506 M2489.75 163.802 Q2495.56 163.802 2498.62 168.409 Q2501.69 172.992 2501.69 181.742 Q2501.69 190.469 2498.62 195.075 Q2495.56 199.658 2489.75 199.658 Q2483.94 199.658 2480.86 195.075 Q2477.81 190.469 2477.81 181.742 Q2477.81 172.992 2480.86 168.409 Q2483.94 163.802 2489.75 163.802 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip330)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2232.76,1423.18 2232.76,1246.82 2256.76,1246.82 2232.76,1246.82 2232.76,1069.9 2256.76,1069.9 2232.76,1069.9 2232.76,892.991 2256.76,892.991 2232.76,892.991 2232.76,716.077 2256.76,716.077 2232.76,716.077 2232.76,539.163 2256.76,539.163 2232.76,539.163 2232.76,362.25 2256.76,362.25 2232.76,362.25 2232.76,185.336 2256.76,185.336 2232.76,185.336 2232.76,47.2441 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/linear/factory_schedule/c5e9cd8e.svg b/previews/PR3525/tutorials/linear/factory_schedule/c5e9cd8e.svg
new file mode 100644
index 00000000000..9d174e492f6
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/factory_schedule/c5e9cd8e.svg
@@ -0,0 +1,162 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip210">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip210)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip211">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip210)" d="M369.518 1423.18 L2352.76 1423.18 L2352.76 47.2441 L369.518 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip212">
+    <rect x="369" y="47" width="1984" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="388.85,1423.18 388.85,47.2441 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="688.015,1423.18 688.015,47.2441 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="987.181,1423.18 987.181,47.2441 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1286.35,1423.18 1286.35,47.2441 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1585.51,1423.18 1585.51,47.2441 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1884.68,1423.18 1884.68,47.2441 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2183.84,1423.18 2183.84,47.2441 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="369.518,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="388.85,1423.18 388.85,1404.28 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="688.015,1423.18 688.015,1404.28 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="987.181,1423.18 987.181,1404.28 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1286.35,1423.18 1286.35,1404.28 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1585.51,1423.18 1585.51,1404.28 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1884.68,1423.18 1884.68,1404.28 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2183.84,1423.18 2183.84,1404.28 "/>
+<path clip-path="url(#clip210)" d="M388.85 1454.1 Q385.239 1454.1 383.411 1457.66 Q381.605 1461.2 381.605 1468.33 Q381.605 1475.44 383.411 1479.01 Q385.239 1482.55 388.85 1482.55 Q392.485 1482.55 394.29 1479.01 Q396.119 1475.44 396.119 1468.33 Q396.119 1461.2 394.29 1457.66 Q392.485 1454.1 388.85 1454.1 M388.85 1450.39 Q394.66 1450.39 397.716 1455 Q400.795 1459.58 400.795 1468.33 Q400.795 1477.06 397.716 1481.67 Q394.66 1486.25 388.85 1486.25 Q383.04 1486.25 379.961 1481.67 Q376.906 1477.06 376.906 1468.33 Q376.906 1459.58 379.961 1455 Q383.04 1450.39 388.85 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M682.668 1481.64 L698.988 1481.64 L698.988 1485.58 L677.043 1485.58 L677.043 1481.64 Q679.705 1478.89 684.289 1474.26 Q688.895 1469.61 690.076 1468.27 Q692.321 1465.74 693.201 1464.01 Q694.103 1462.25 694.103 1460.56 Q694.103 1457.8 692.159 1456.07 Q690.238 1454.33 687.136 1454.33 Q684.937 1454.33 682.483 1455.09 Q680.053 1455.86 677.275 1457.41 L677.275 1452.69 Q680.099 1451.55 682.553 1450.97 Q685.006 1450.39 687.043 1450.39 Q692.414 1450.39 695.608 1453.08 Q698.802 1455.77 698.802 1460.26 Q698.802 1462.39 697.992 1464.31 Q697.205 1466.2 695.099 1468.8 Q694.52 1469.47 691.418 1472.69 Q688.316 1475.88 682.668 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M990.19 1455.09 L978.384 1473.54 L990.19 1473.54 L990.19 1455.09 M988.963 1451.02 L994.843 1451.02 L994.843 1473.54 L999.773 1473.54 L999.773 1477.43 L994.843 1477.43 L994.843 1485.58 L990.19 1485.58 L990.19 1477.43 L974.588 1477.43 L974.588 1472.92 L988.963 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1286.75 1466.44 Q1283.6 1466.44 1281.75 1468.59 Q1279.92 1470.74 1279.92 1474.49 Q1279.92 1478.22 1281.75 1480.39 Q1283.6 1482.55 1286.75 1482.55 Q1289.9 1482.55 1291.73 1480.39 Q1293.58 1478.22 1293.58 1474.49 Q1293.58 1470.74 1291.73 1468.59 Q1289.9 1466.44 1286.75 1466.44 M1296.03 1451.78 L1296.03 1456.04 Q1294.27 1455.21 1292.47 1454.77 Q1290.69 1454.33 1288.93 1454.33 Q1284.3 1454.33 1281.84 1457.45 Q1279.41 1460.58 1279.07 1466.9 Q1280.43 1464.89 1282.49 1463.82 Q1284.55 1462.73 1287.03 1462.73 Q1292.24 1462.73 1295.25 1465.9 Q1298.28 1469.05 1298.28 1474.49 Q1298.28 1479.82 1295.13 1483.03 Q1291.98 1486.25 1286.75 1486.25 Q1280.76 1486.25 1277.58 1481.67 Q1274.41 1477.06 1274.41 1468.33 Q1274.41 1460.14 1278.3 1455.28 Q1282.19 1450.39 1288.74 1450.39 Q1290.5 1450.39 1292.28 1450.74 Q1294.09 1451.09 1296.03 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1585.51 1469.17 Q1582.18 1469.17 1580.26 1470.95 Q1578.36 1472.73 1578.36 1475.86 Q1578.36 1478.98 1580.26 1480.77 Q1582.18 1482.55 1585.51 1482.55 Q1588.84 1482.55 1590.77 1480.77 Q1592.69 1478.96 1592.69 1475.86 Q1592.69 1472.73 1590.77 1470.95 Q1588.87 1469.17 1585.51 1469.17 M1580.84 1467.18 Q1577.83 1466.44 1576.14 1464.38 Q1574.47 1462.32 1574.47 1459.35 Q1574.47 1455.21 1577.41 1452.8 Q1580.37 1450.39 1585.51 1450.39 Q1590.67 1450.39 1593.61 1452.8 Q1596.55 1455.21 1596.55 1459.35 Q1596.55 1462.32 1594.86 1464.38 Q1593.2 1466.44 1590.21 1467.18 Q1593.59 1467.96 1595.46 1470.26 Q1597.36 1472.55 1597.36 1475.86 Q1597.36 1480.88 1594.28 1483.57 Q1591.23 1486.25 1585.51 1486.25 Q1579.79 1486.25 1576.71 1483.57 Q1573.66 1480.88 1573.66 1475.86 Q1573.66 1472.55 1575.56 1470.26 Q1577.46 1467.96 1580.84 1467.18 M1579.12 1459.79 Q1579.12 1462.48 1580.79 1463.98 Q1582.48 1465.49 1585.51 1465.49 Q1588.52 1465.49 1590.21 1463.98 Q1591.92 1462.48 1591.92 1459.79 Q1591.92 1457.11 1590.21 1455.6 Q1588.52 1454.1 1585.51 1454.1 Q1582.48 1454.1 1580.79 1455.6 Q1579.12 1457.11 1579.12 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1859.36 1481.64 L1867 1481.64 L1867 1455.28 L1858.69 1456.95 L1858.69 1452.69 L1866.96 1451.02 L1871.63 1451.02 L1871.63 1481.64 L1879.27 1481.64 L1879.27 1485.58 L1859.36 1485.58 L1859.36 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1898.72 1454.1 Q1895.1 1454.1 1893.28 1457.66 Q1891.47 1461.2 1891.47 1468.33 Q1891.47 1475.44 1893.28 1479.01 Q1895.1 1482.55 1898.72 1482.55 Q1902.35 1482.55 1904.16 1479.01 Q1905.98 1475.44 1905.98 1468.33 Q1905.98 1461.2 1904.16 1457.66 Q1902.35 1454.1 1898.72 1454.1 M1898.72 1450.39 Q1904.53 1450.39 1907.58 1455 Q1910.66 1459.58 1910.66 1468.33 Q1910.66 1477.06 1907.58 1481.67 Q1904.53 1486.25 1898.72 1486.25 Q1892.91 1486.25 1889.83 1481.67 Q1886.77 1477.06 1886.77 1468.33 Q1886.77 1459.58 1889.83 1455 Q1892.91 1450.39 1898.72 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M2159.33 1481.64 L2166.97 1481.64 L2166.97 1455.28 L2158.66 1456.95 L2158.66 1452.69 L2166.92 1451.02 L2171.6 1451.02 L2171.6 1481.64 L2179.23 1481.64 L2179.23 1485.58 L2159.33 1485.58 L2159.33 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M2192.71 1481.64 L2209.03 1481.64 L2209.03 1485.58 L2187.08 1485.58 L2187.08 1481.64 Q2189.74 1478.89 2194.33 1474.26 Q2198.93 1469.61 2200.11 1468.27 Q2202.36 1465.74 2203.24 1464.01 Q2204.14 1462.25 2204.14 1460.56 Q2204.14 1457.8 2202.2 1456.07 Q2200.28 1454.33 2197.17 1454.33 Q2194.98 1454.33 2192.52 1455.09 Q2190.09 1455.86 2187.31 1457.41 L2187.31 1452.69 Q2190.14 1451.55 2192.59 1450.97 Q2195.04 1450.39 2197.08 1450.39 Q2202.45 1450.39 2205.65 1453.08 Q2208.84 1455.77 2208.84 1460.26 Q2208.84 1462.39 2208.03 1464.31 Q2207.24 1466.2 2205.14 1468.8 Q2204.56 1469.47 2201.46 1472.69 Q2198.36 1475.88 2192.71 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1264.95 1520.52 L1274.53 1520.52 L1286.66 1552.86 L1298.85 1520.52 L1308.43 1520.52 L1308.43 1568.04 L1302.16 1568.04 L1302.16 1526.32 L1289.9 1558.91 L1283.44 1558.91 L1271.19 1526.32 L1271.19 1568.04 L1264.95 1568.04 L1264.95 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1334.75 1536.5 Q1330.04 1536.5 1327.3 1540.19 Q1324.57 1543.85 1324.57 1550.25 Q1324.57 1556.65 1327.27 1560.34 Q1330.01 1564 1334.75 1564 Q1339.43 1564 1342.17 1560.31 Q1344.9 1556.62 1344.9 1550.25 Q1344.9 1543.92 1342.17 1540.23 Q1339.43 1536.5 1334.75 1536.5 M1334.75 1531.54 Q1342.39 1531.54 1346.75 1536.5 Q1351.11 1541.47 1351.11 1550.25 Q1351.11 1559 1346.75 1564 Q1342.39 1568.97 1334.75 1568.97 Q1327.08 1568.97 1322.72 1564 Q1318.39 1559 1318.39 1550.25 Q1318.39 1541.47 1322.72 1536.5 Q1327.08 1531.54 1334.75 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1390.45 1546.53 L1390.45 1568.04 L1384.59 1568.04 L1384.59 1546.72 Q1384.59 1541.66 1382.62 1539.14 Q1380.65 1536.63 1376.7 1536.63 Q1371.96 1536.63 1369.22 1539.65 Q1366.48 1542.68 1366.48 1547.9 L1366.48 1568.04 L1360.6 1568.04 L1360.6 1532.4 L1366.48 1532.4 L1366.48 1537.93 Q1368.58 1534.72 1371.42 1533.13 Q1374.28 1531.54 1378.01 1531.54 Q1384.15 1531.54 1387.3 1535.36 Q1390.45 1539.14 1390.45 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1407.93 1522.27 L1407.93 1532.4 L1419.99 1532.4 L1419.99 1536.95 L1407.93 1536.95 L1407.93 1556.3 Q1407.93 1560.66 1409.1 1561.9 Q1410.31 1563.14 1413.97 1563.14 L1419.99 1563.14 L1419.99 1568.04 L1413.97 1568.04 Q1407.19 1568.04 1404.61 1565.53 Q1402.04 1562.98 1402.04 1556.3 L1402.04 1536.95 L1397.74 1536.95 L1397.74 1532.4 L1402.04 1532.4 L1402.04 1522.27 L1407.93 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M1457.32 1546.53 L1457.32 1568.04 L1451.47 1568.04 L1451.47 1546.72 Q1451.47 1541.66 1449.49 1539.14 Q1447.52 1536.63 1443.57 1536.63 Q1438.83 1536.63 1436.09 1539.65 Q1433.36 1542.68 1433.36 1547.9 L1433.36 1568.04 L1427.47 1568.04 L1427.47 1518.52 L1433.36 1518.52 L1433.36 1537.93 Q1435.46 1534.72 1438.29 1533.13 Q1441.15 1531.54 1444.88 1531.54 Q1451.02 1531.54 1454.17 1535.36 Q1457.32 1539.14 1457.32 1546.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="369.518,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="369.518,1079.2 2352.76,1079.2 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="369.518,735.212 2352.76,735.212 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="369.518,391.228 2352.76,391.228 "/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="369.518,47.2441 2352.76,47.2441 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="369.518,1423.18 369.518,47.2441 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="369.518,1423.18 388.416,1423.18 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="369.518,1079.2 388.416,1079.2 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="369.518,735.212 388.416,735.212 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="369.518,391.228 388.416,391.228 "/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="369.518,47.2441 388.416,47.2441 "/>
+<path clip-path="url(#clip210)" d="M321.574 1408.98 Q317.963 1408.98 316.134 1412.54 Q314.329 1416.08 314.329 1423.21 Q314.329 1430.32 316.134 1433.89 Q317.963 1437.43 321.574 1437.43 Q325.208 1437.43 327.014 1433.89 Q328.842 1430.32 328.842 1423.21 Q328.842 1416.08 327.014 1412.54 Q325.208 1408.98 321.574 1408.98 M321.574 1405.27 Q327.384 1405.27 330.44 1409.88 Q333.518 1414.46 333.518 1423.21 Q333.518 1431.94 330.44 1436.55 Q327.384 1441.13 321.574 1441.13 Q315.764 1441.13 312.685 1436.55 Q309.629 1431.94 309.629 1423.21 Q309.629 1414.46 312.685 1409.88 Q315.764 1405.27 321.574 1405.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M115.718 1068.36 L134.075 1068.36 L134.075 1072.3 L120.001 1072.3 L120.001 1080.77 Q121.019 1080.42 122.038 1080.26 Q123.056 1080.08 124.075 1080.08 Q129.862 1080.08 133.242 1083.25 Q136.621 1086.42 136.621 1091.84 Q136.621 1097.41 133.149 1100.52 Q129.677 1103.59 123.357 1103.59 Q121.181 1103.59 118.913 1103.22 Q116.668 1102.85 114.26 1102.11 L114.26 1097.41 Q116.343 1098.55 118.566 1099.1 Q120.788 1099.66 123.265 1099.66 Q127.269 1099.66 129.607 1097.55 Q131.945 1095.45 131.945 1091.84 Q131.945 1088.22 129.607 1086.12 Q127.269 1084.01 123.265 1084.01 Q121.39 1084.01 119.515 1084.43 Q117.663 1084.84 115.718 1085.72 L115.718 1068.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M145.834 1097.04 L150.718 1097.04 L150.718 1102.92 L145.834 1102.92 L145.834 1097.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M170.903 1071.44 Q167.292 1071.44 165.464 1075.01 Q163.658 1078.55 163.658 1085.68 Q163.658 1092.78 165.464 1096.35 Q167.292 1099.89 170.903 1099.89 Q174.538 1099.89 176.343 1096.35 Q178.172 1092.78 178.172 1085.68 Q178.172 1078.55 176.343 1075.01 Q174.538 1071.44 170.903 1071.44 M170.903 1067.74 Q176.714 1067.74 179.769 1072.34 Q182.848 1076.93 182.848 1085.68 Q182.848 1094.41 179.769 1099.01 Q176.714 1103.59 170.903 1103.59 Q165.093 1103.59 162.015 1099.01 Q158.959 1094.41 158.959 1085.68 Q158.959 1076.93 162.015 1072.34 Q165.093 1067.74 170.903 1067.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M201.065 1071.44 Q197.454 1071.44 195.625 1075.01 Q193.82 1078.55 193.82 1085.68 Q193.82 1092.78 195.625 1096.35 Q197.454 1099.89 201.065 1099.89 Q204.699 1099.89 206.505 1096.35 Q208.334 1092.78 208.334 1085.68 Q208.334 1078.55 206.505 1075.01 Q204.699 1071.44 201.065 1071.44 M201.065 1067.74 Q206.875 1067.74 209.931 1072.34 Q213.01 1076.93 213.01 1085.68 Q213.01 1094.41 209.931 1099.01 Q206.875 1103.59 201.065 1103.59 Q195.255 1103.59 192.176 1099.01 Q189.121 1094.41 189.121 1085.68 Q189.121 1076.93 192.176 1072.34 Q195.255 1067.74 201.065 1067.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M249.398 1077.46 L238.82 1088.09 L249.398 1098.66 L246.644 1101.47 L236.019 1090.84 L225.394 1101.47 L222.662 1098.66 L233.218 1088.09 L222.662 1077.46 L225.394 1074.66 L236.019 1085.28 L246.644 1074.66 L249.398 1077.46 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M261.759 1098.99 L269.398 1098.99 L269.398 1072.62 L261.088 1074.29 L261.088 1070.03 L269.352 1068.36 L274.028 1068.36 L274.028 1098.99 L281.667 1098.99 L281.667 1102.92 L261.759 1102.92 L261.759 1098.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M301.111 1071.44 Q297.5 1071.44 295.671 1075.01 Q293.866 1078.55 293.866 1085.68 Q293.866 1092.78 295.671 1096.35 Q297.5 1099.89 301.111 1099.89 Q304.745 1099.89 306.551 1096.35 Q308.379 1092.78 308.379 1085.68 Q308.379 1078.55 306.551 1075.01 Q304.745 1071.44 301.111 1071.44 M301.111 1067.74 Q306.921 1067.74 309.977 1072.34 Q313.055 1076.93 313.055 1085.68 Q313.055 1094.41 309.977 1099.01 Q306.921 1103.59 301.111 1103.59 Q295.301 1103.59 292.222 1099.01 Q289.167 1094.41 289.167 1085.68 Q289.167 1076.93 292.222 1072.34 Q295.301 1067.74 301.111 1067.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M325.732 1050.74 L316.14 1065.73 L325.732 1065.73 L325.732 1050.74 M324.735 1047.43 L329.512 1047.43 L329.512 1065.73 L333.518 1065.73 L333.518 1068.89 L329.512 1068.89 L329.512 1075.51 L325.732 1075.51 L325.732 1068.89 L313.055 1068.89 L313.055 1065.23 L324.735 1047.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M118.777 755.004 L126.416 755.004 L126.416 728.639 L118.106 730.305 L118.106 726.046 L126.369 724.38 L131.045 724.38 L131.045 755.004 L138.684 755.004 L138.684 758.94 L118.777 758.94 L118.777 755.004 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M148.129 753.06 L153.013 753.06 L153.013 758.94 L148.129 758.94 L148.129 753.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M173.198 727.458 Q169.587 727.458 167.758 731.023 Q165.953 734.565 165.953 741.694 Q165.953 748.801 167.758 752.366 Q169.587 755.907 173.198 755.907 Q176.832 755.907 178.638 752.366 Q180.466 748.801 180.466 741.694 Q180.466 734.565 178.638 731.023 Q176.832 727.458 173.198 727.458 M173.198 723.755 Q179.008 723.755 182.064 728.361 Q185.142 732.944 185.142 741.694 Q185.142 750.421 182.064 755.028 Q179.008 759.611 173.198 759.611 Q167.388 759.611 164.309 755.028 Q161.254 750.421 161.254 741.694 Q161.254 732.944 164.309 728.361 Q167.388 723.755 173.198 723.755 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M203.36 727.458 Q199.749 727.458 197.92 731.023 Q196.114 734.565 196.114 741.694 Q196.114 748.801 197.92 752.366 Q199.749 755.907 203.36 755.907 Q206.994 755.907 208.8 752.366 Q210.628 748.801 210.628 741.694 Q210.628 734.565 208.8 731.023 Q206.994 727.458 203.36 727.458 M203.36 723.755 Q209.17 723.755 212.225 728.361 Q215.304 732.944 215.304 741.694 Q215.304 750.421 212.225 755.028 Q209.17 759.611 203.36 759.611 Q197.55 759.611 194.471 755.028 Q191.415 750.421 191.415 741.694 Q191.415 732.944 194.471 728.361 Q197.55 723.755 203.36 723.755 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M251.693 733.477 L241.114 744.102 L251.693 754.68 L248.938 757.481 L238.313 746.856 L227.688 757.481 L224.957 754.68 L235.512 744.102 L224.957 733.477 L227.688 730.676 L238.313 741.301 L248.938 730.676 L251.693 733.477 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M264.054 755.004 L271.693 755.004 L271.693 728.639 L263.383 730.305 L263.383 726.046 L271.646 724.38 L276.322 724.38 L276.322 755.004 L283.961 755.004 L283.961 758.94 L264.054 758.94 L264.054 755.004 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M303.406 727.458 Q299.794 727.458 297.966 731.023 Q296.16 734.565 296.16 741.694 Q296.16 748.801 297.966 752.366 Q299.794 755.907 303.406 755.907 Q307.04 755.907 308.845 752.366 Q310.674 748.801 310.674 741.694 Q310.674 734.565 308.845 731.023 Q307.04 727.458 303.406 727.458 M303.406 723.755 Q309.216 723.755 312.271 728.361 Q315.35 732.944 315.35 741.694 Q315.35 750.421 312.271 755.028 Q309.216 759.611 303.406 759.611 Q297.595 759.611 294.517 755.028 Q291.461 750.421 291.461 741.694 Q291.461 732.944 294.517 728.361 Q297.595 723.755 303.406 723.755 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M316.535 703.449 L331.449 703.449 L331.449 706.646 L320.014 706.646 L320.014 713.53 Q320.842 713.248 321.669 713.116 Q322.497 712.966 323.324 712.966 Q328.026 712.966 330.772 715.543 Q333.518 718.119 333.518 722.52 Q333.518 727.053 330.697 729.573 Q327.876 732.075 322.741 732.075 Q320.973 732.075 319.13 731.774 Q317.306 731.473 315.35 730.871 L315.35 727.053 Q317.043 727.975 318.848 728.426 Q320.654 728.877 322.666 728.877 Q325.92 728.877 327.819 727.166 Q329.719 725.454 329.719 722.52 Q329.719 719.586 327.819 717.875 Q325.92 716.163 322.666 716.163 Q321.143 716.163 319.619 716.502 Q318.115 716.84 316.535 717.555 L316.535 703.449 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M118.777 411.02 L126.416 411.02 L126.416 384.655 L118.106 386.322 L118.106 382.062 L126.369 380.396 L131.045 380.396 L131.045 411.02 L138.684 411.02 L138.684 414.956 L118.777 414.956 L118.777 411.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M148.129 409.076 L153.013 409.076 L153.013 414.956 L148.129 414.956 L148.129 409.076 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M163.244 380.396 L181.601 380.396 L181.601 384.331 L167.527 384.331 L167.527 392.803 Q168.545 392.456 169.564 392.294 Q170.582 392.109 171.601 392.109 Q177.388 392.109 180.767 395.28 Q184.147 398.451 184.147 403.868 Q184.147 409.446 180.675 412.548 Q177.203 415.627 170.883 415.627 Q168.707 415.627 166.439 415.257 Q164.193 414.886 161.786 414.145 L161.786 409.446 Q163.869 410.581 166.091 411.136 Q168.314 411.692 170.791 411.692 Q174.795 411.692 177.133 409.585 Q179.471 407.479 179.471 403.868 Q179.471 400.257 177.133 398.15 Q174.795 396.044 170.791 396.044 Q168.916 396.044 167.041 396.46 Q165.189 396.877 163.244 397.757 L163.244 380.396 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M203.36 383.474 Q199.749 383.474 197.92 387.039 Q196.114 390.581 196.114 397.71 Q196.114 404.817 197.92 408.382 Q199.749 411.923 203.36 411.923 Q206.994 411.923 208.8 408.382 Q210.628 404.817 210.628 397.71 Q210.628 390.581 208.8 387.039 Q206.994 383.474 203.36 383.474 M203.36 379.771 Q209.17 379.771 212.225 384.377 Q215.304 388.96 215.304 397.71 Q215.304 406.437 212.225 411.044 Q209.17 415.627 203.36 415.627 Q197.55 415.627 194.471 411.044 Q191.415 406.437 191.415 397.71 Q191.415 388.96 194.471 384.377 Q197.55 379.771 203.36 379.771 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M251.693 389.493 L241.114 400.118 L251.693 410.696 L248.938 413.497 L238.313 402.872 L227.688 413.497 L224.957 410.696 L235.512 400.118 L224.957 389.493 L227.688 386.692 L238.313 397.317 L248.938 386.692 L251.693 389.493 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M264.054 411.02 L271.693 411.02 L271.693 384.655 L263.383 386.322 L263.383 382.062 L271.646 380.396 L276.322 380.396 L276.322 411.02 L283.961 411.02 L283.961 414.956 L264.054 414.956 L264.054 411.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M303.406 383.474 Q299.794 383.474 297.966 387.039 Q296.16 390.581 296.16 397.71 Q296.16 404.817 297.966 408.382 Q299.794 411.923 303.406 411.923 Q307.04 411.923 308.845 408.382 Q310.674 404.817 310.674 397.71 Q310.674 390.581 308.845 387.039 Q307.04 383.474 303.406 383.474 M303.406 379.771 Q309.216 379.771 312.271 384.377 Q315.35 388.96 315.35 397.71 Q315.35 406.437 312.271 411.044 Q309.216 415.627 303.406 415.627 Q297.595 415.627 294.517 411.044 Q291.461 406.437 291.461 397.71 Q291.461 388.96 294.517 384.377 Q297.595 379.771 303.406 379.771 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M316.535 359.465 L331.449 359.465 L331.449 362.663 L320.014 362.663 L320.014 369.546 Q320.842 369.264 321.669 369.132 Q322.497 368.982 323.324 368.982 Q328.026 368.982 330.772 371.559 Q333.518 374.135 333.518 378.536 Q333.518 383.069 330.697 385.589 Q327.876 388.091 322.741 388.091 Q320.973 388.091 319.13 387.79 Q317.306 387.489 315.35 386.887 L315.35 383.069 Q317.043 383.991 318.848 384.442 Q320.654 384.893 322.666 384.893 Q325.92 384.893 327.819 383.182 Q329.719 381.47 329.719 378.536 Q329.719 375.602 327.819 373.891 Q325.92 372.179 322.666 372.179 Q321.143 372.179 319.619 372.518 Q318.115 372.856 316.535 373.571 L316.535 359.465 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M121.994 67.0365 L138.314 67.0365 L138.314 70.9717 L116.37 70.9717 L116.37 67.0365 Q119.032 64.2819 123.615 59.6523 Q128.221 54.9996 129.402 53.657 Q131.647 51.1338 132.527 49.3977 Q133.43 47.6385 133.43 45.9487 Q133.43 43.1941 131.485 41.458 Q129.564 39.7219 126.462 39.7219 Q124.263 39.7219 121.809 40.4857 Q119.379 41.2496 116.601 42.8005 L116.601 38.0784 Q119.425 36.9441 121.879 36.3654 Q124.332 35.7867 126.369 35.7867 Q131.74 35.7867 134.934 38.4719 Q138.129 41.157 138.129 45.6478 Q138.129 47.7774 137.318 49.6987 Q136.531 51.5968 134.425 54.1894 Q133.846 54.8607 130.744 58.0782 Q127.643 61.2727 121.994 67.0365 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M148.129 65.0921 L153.013 65.0921 L153.013 70.9717 L148.129 70.9717 L148.129 65.0921 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M173.198 39.4904 Q169.587 39.4904 167.758 43.0552 Q165.953 46.5968 165.953 53.7264 Q165.953 60.8329 167.758 64.3977 Q169.587 67.9393 173.198 67.9393 Q176.832 67.9393 178.638 64.3977 Q180.466 60.8329 180.466 53.7264 Q180.466 46.5968 178.638 43.0552 Q176.832 39.4904 173.198 39.4904 M173.198 35.7867 Q179.008 35.7867 182.064 40.3932 Q185.142 44.9765 185.142 53.7264 Q185.142 62.4532 182.064 67.0597 Q179.008 71.643 173.198 71.643 Q167.388 71.643 164.309 67.0597 Q161.254 62.4532 161.254 53.7264 Q161.254 44.9765 164.309 40.3932 Q167.388 35.7867 173.198 35.7867 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M203.36 39.4904 Q199.749 39.4904 197.92 43.0552 Q196.114 46.5968 196.114 53.7264 Q196.114 60.8329 197.92 64.3977 Q199.749 67.9393 203.36 67.9393 Q206.994 67.9393 208.8 64.3977 Q210.628 60.8329 210.628 53.7264 Q210.628 46.5968 208.8 43.0552 Q206.994 39.4904 203.36 39.4904 M203.36 35.7867 Q209.17 35.7867 212.225 40.3932 Q215.304 44.9765 215.304 53.7264 Q215.304 62.4532 212.225 67.0597 Q209.17 71.643 203.36 71.643 Q197.55 71.643 194.471 67.0597 Q191.415 62.4532 191.415 53.7264 Q191.415 44.9765 194.471 40.3932 Q197.55 35.7867 203.36 35.7867 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M251.693 45.5089 L241.114 56.1338 L251.693 66.7125 L248.938 69.5134 L238.313 58.8884 L227.688 69.5134 L224.957 66.7125 L235.512 56.1338 L224.957 45.5089 L227.688 42.708 L238.313 53.3329 L248.938 42.708 L251.693 45.5089 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M264.054 67.0365 L271.693 67.0365 L271.693 40.6709 L263.383 42.3376 L263.383 38.0784 L271.646 36.4117 L276.322 36.4117 L276.322 67.0365 L283.961 67.0365 L283.961 70.9717 L264.054 70.9717 L264.054 67.0365 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M303.406 39.4904 Q299.794 39.4904 297.966 43.0552 Q296.16 46.5968 296.16 53.7264 Q296.16 60.8329 297.966 64.3977 Q299.794 67.9393 303.406 67.9393 Q307.04 67.9393 308.845 64.3977 Q310.674 60.8329 310.674 53.7264 Q310.674 46.5968 308.845 43.0552 Q307.04 39.4904 303.406 39.4904 M303.406 35.7867 Q309.216 35.7867 312.271 40.3932 Q315.35 44.9765 315.35 53.7264 Q315.35 62.4532 312.271 67.0597 Q309.216 71.643 303.406 71.643 Q297.595 71.643 294.517 67.0597 Q291.461 62.4532 291.461 53.7264 Q291.461 44.9765 294.517 40.3932 Q297.595 35.7867 303.406 35.7867 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M316.535 15.4813 L331.449 15.4813 L331.449 18.6786 L320.014 18.6786 L320.014 25.5623 Q320.842 25.2801 321.669 25.1485 Q322.497 24.998 323.324 24.998 Q328.026 24.998 330.772 27.5747 Q333.518 30.1514 333.518 34.5524 Q333.518 39.085 330.697 41.6053 Q327.876 44.1067 322.741 44.1067 Q320.973 44.1067 319.13 43.8058 Q317.306 43.5049 315.35 42.903 L315.35 39.085 Q317.043 40.0066 318.848 40.458 Q320.654 40.9094 322.666 40.9094 Q325.92 40.9094 327.819 39.1979 Q329.719 37.4864 329.719 34.5524 Q329.719 31.6184 327.819 29.9069 Q325.92 28.1953 322.666 28.1953 Q321.143 28.1953 319.619 28.5339 Q318.115 28.8724 316.535 29.5871 L316.535 15.4813 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M21.7677 896.217 L39.6235 896.217 L39.6235 888.132 Q39.6235 883.644 37.3 881.194 Q34.9765 878.743 30.6797 878.743 Q26.4147 878.743 24.0912 881.194 Q21.7677 883.644 21.7677 888.132 L21.7677 896.217 M16.4842 902.646 L16.4842 888.132 Q16.4842 880.143 20.1126 876.069 Q23.7092 871.963 30.6797 871.963 Q37.7138 871.963 41.3104 876.069 Q44.907 880.143 44.907 888.132 L44.907 896.217 L64.0042 896.217 L64.0042 902.646 L16.4842 902.646 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M33.8307 844.081 Q33.2578 845.068 33.0032 846.246 Q32.7167 847.392 32.7167 848.792 Q32.7167 853.757 35.9632 856.431 Q39.1779 859.073 45.2253 859.073 L64.0042 859.073 L64.0042 864.961 L28.3562 864.961 L28.3562 859.073 L33.8944 859.073 Q30.6479 857.227 29.0883 854.267 Q27.4968 851.307 27.4968 847.073 Q27.4968 846.469 27.5923 845.737 Q27.656 845.004 27.8151 844.113 L33.8307 844.081 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M32.4621 825.557 Q32.4621 830.268 36.1542 833.005 Q39.8145 835.742 46.212 835.742 Q52.6095 835.742 56.3017 833.037 Q59.9619 830.3 59.9619 825.557 Q59.9619 820.878 56.2698 818.141 Q52.5777 815.404 46.212 815.404 Q39.8781 815.404 36.186 818.141 Q32.4621 820.878 32.4621 825.557 M27.4968 825.557 Q27.4968 817.918 32.4621 813.558 Q37.4273 809.197 46.212 809.197 Q54.9649 809.197 59.9619 813.558 Q64.9272 817.918 64.9272 825.557 Q64.9272 833.228 59.9619 837.588 Q54.9649 841.917 46.212 841.917 Q37.4273 841.917 32.4621 837.588 Q27.4968 833.228 27.4968 825.557 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M33.7671 776.032 L14.479 776.032 L14.479 770.176 L64.0042 770.176 L64.0042 776.032 L58.657 776.032 Q61.8398 777.878 63.3994 780.711 Q64.9272 783.512 64.9272 787.459 Q64.9272 793.92 59.771 797.994 Q54.6147 802.036 46.212 802.036 Q37.8093 802.036 32.6531 797.994 Q27.4968 793.92 27.4968 787.459 Q27.4968 783.512 29.0564 780.711 Q30.5842 777.878 33.7671 776.032 M46.212 795.989 Q52.6732 795.989 56.3653 793.347 Q60.0256 790.673 60.0256 786.026 Q60.0256 781.379 56.3653 778.706 Q52.6732 776.032 46.212 776.032 Q39.7508 776.032 36.0905 778.706 Q32.3984 781.379 32.3984 786.026 Q32.3984 790.673 36.0905 793.347 Q39.7508 795.989 46.212 795.989 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M49.9359 758.717 L28.3562 758.717 L28.3562 752.861 L49.7131 752.861 Q54.7739 752.861 57.3202 750.888 Q59.8346 748.914 59.8346 744.967 Q59.8346 740.225 56.8109 737.488 Q53.7872 734.719 48.5673 734.719 L28.3562 734.719 L28.3562 728.862 L64.0042 728.862 L64.0042 734.719 L58.5296 734.719 Q61.7762 736.851 63.3676 739.684 Q64.9272 742.485 64.9272 746.209 Q64.9272 752.352 61.1078 755.534 Q57.2883 758.717 49.9359 758.717 M27.4968 743.981 L27.4968 743.981 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M29.7248 691.145 L35.1993 691.145 Q33.8307 693.628 33.1623 696.142 Q32.4621 698.625 32.4621 701.171 Q32.4621 706.869 36.0905 710.02 Q39.6872 713.171 46.212 713.171 Q52.7369 713.171 56.3653 710.02 Q59.9619 706.869 59.9619 701.171 Q59.9619 698.625 59.2935 696.142 Q58.5933 693.628 57.2247 691.145 L62.6355 691.145 Q63.7814 693.596 64.3543 696.238 Q64.9272 698.848 64.9272 701.808 Q64.9272 709.861 59.8664 714.603 Q54.8057 719.345 46.212 719.345 Q37.491 719.345 32.4939 714.571 Q27.4968 709.765 27.4968 701.426 Q27.4968 698.721 28.0697 696.142 Q28.6108 693.564 29.7248 691.145 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M18.2347 675.167 L28.3562 675.167 L28.3562 663.104 L32.9077 663.104 L32.9077 675.167 L52.2594 675.167 Q56.6199 675.167 57.8613 673.99 Q59.1026 672.78 59.1026 669.12 L59.1026 663.104 L64.0042 663.104 L64.0042 669.12 Q64.0042 675.899 61.4897 678.478 Q58.9434 681.056 52.2594 681.056 L32.9077 681.056 L32.9077 685.353 L28.3562 685.353 L28.3562 681.056 L18.2347 681.056 L18.2347 675.167 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M28.3562 655.402 L28.3562 649.545 L64.0042 649.545 L64.0042 655.402 L28.3562 655.402 M14.479 655.402 L14.479 649.545 L21.895 649.545 L21.895 655.402 L14.479 655.402 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M32.4621 623.478 Q32.4621 628.189 36.1542 630.926 Q39.8145 633.663 46.212 633.663 Q52.6095 633.663 56.3017 630.958 Q59.9619 628.22 59.9619 623.478 Q59.9619 618.799 56.2698 616.062 Q52.5777 613.325 46.212 613.325 Q39.8781 613.325 36.186 616.062 Q32.4621 618.799 32.4621 623.478 M27.4968 623.478 Q27.4968 615.839 32.4621 611.479 Q37.4273 607.118 46.212 607.118 Q54.9649 607.118 59.9619 611.479 Q64.9272 615.839 64.9272 623.478 Q64.9272 631.149 59.9619 635.509 Q54.9649 639.838 46.212 639.838 Q37.4273 639.838 32.4621 635.509 Q27.4968 631.149 27.4968 623.478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M42.4881 567.778 L64.0042 567.778 L64.0042 573.634 L42.679 573.634 Q37.6183 573.634 35.1038 575.608 Q32.5894 577.581 32.5894 581.528 Q32.5894 586.27 35.6131 589.008 Q38.6368 591.745 43.8567 591.745 L64.0042 591.745 L64.0042 597.633 L28.3562 597.633 L28.3562 591.745 L33.8944 591.745 Q30.6797 589.644 29.0883 586.811 Q27.4968 583.947 27.4968 580.223 Q27.4968 574.08 31.3163 570.929 Q35.1038 567.778 42.4881 567.778 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip212)" d="M478.6 597.618 L478.6 597.618 L598.266 597.618 L598.266 597.618 L478.6 597.618 L478.6 597.618  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="478.6,597.618 478.6,597.618 598.266,597.618 478.6,597.618 "/>
+<path clip-path="url(#clip212)" d="M628.182 735.212 L628.182 735.212 L747.848 735.212 L747.848 735.212 L628.182 735.212 L628.182 735.212  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="628.182,735.212 628.182,735.212 747.848,735.212 628.182,735.212 "/>
+<path clip-path="url(#clip212)" d="M777.765 528.822 L777.765 528.822 L897.431 528.822 L897.431 528.822 L777.765 528.822 L777.765 528.822  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="777.765,528.822 777.765,528.822 897.431,528.822 777.765,528.822 "/>
+<path clip-path="url(#clip212)" d="M927.348 528.822 L927.348 528.822 L1047.01 528.822 L1047.01 528.822 L927.348 528.822 L927.348 528.822  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="927.348,528.822 927.348,528.822 1047.01,528.822 927.348,528.822 "/>
+<path clip-path="url(#clip212)" d="M1076.93 460.025 L1076.93 460.025 L1196.6 460.025 L1196.6 460.025 L1076.93 460.025 L1076.93 460.025  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1076.93,460.025 1076.93,460.025 1196.6,460.025 1076.93,460.025 "/>
+<path clip-path="url(#clip212)" d="M1226.51 528.822 L1226.51 528.822 L1346.18 528.822 L1346.18 528.822 L1226.51 528.822 L1226.51 528.822  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1226.51,528.822 1226.51,528.822 1346.18,528.822 1226.51,528.822 "/>
+<path clip-path="url(#clip212)" d="M1376.1 391.228 L1376.1 391.228 L1495.76 391.228 L1495.76 391.228 L1376.1 391.228 L1376.1 391.228  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1376.1,391.228 1376.1,391.228 1495.76,391.228 1376.1,391.228 "/>
+<path clip-path="url(#clip212)" d="M1525.68 253.634 L1525.68 253.634 L1645.34 253.634 L1645.34 253.634 L1525.68 253.634 L1525.68 253.634  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1525.68,253.634 1525.68,253.634 1645.34,253.634 1525.68,253.634 "/>
+<path clip-path="url(#clip212)" d="M1675.26 47.2441 L1675.26 47.2441 L1794.93 47.2441 L1794.93 47.2441 L1675.26 47.2441 L1675.26 47.2441  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1675.26,47.2441 1675.26,47.2441 1794.93,47.2441 1675.26,47.2441 "/>
+<path clip-path="url(#clip212)" d="M1824.84 116.041 L1824.84 116.041 L1944.51 116.041 L1944.51 116.041 L1824.84 116.041 L1824.84 116.041  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1824.84,116.041 1824.84,116.041 1944.51,116.041 1824.84,116.041 "/>
+<path clip-path="url(#clip212)" d="M1974.43 460.025 L1974.43 460.025 L2094.09 460.025 L2094.09 460.025 L1974.43 460.025 L1974.43 460.025  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1974.43,460.025 1974.43,460.025 2094.09,460.025 1974.43,460.025 "/>
+<path clip-path="url(#clip212)" d="M2124.01 735.212 L2124.01 735.212 L2243.67 735.212 L2243.67 735.212 L2124.01 735.212 L2124.01 735.212  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2124.01,735.212 2124.01,735.212 2243.67,735.212 2124.01,735.212 "/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="538.433" cy="597.618" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="688.015" cy="735.212" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="837.598" cy="528.822" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="987.181" cy="528.822" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1136.76" cy="460.025" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1286.35" cy="528.822" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1435.93" cy="391.228" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1585.51" cy="253.634" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1735.09" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="1884.68" cy="116.041" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="2034.26" cy="460.025" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#20326c; stroke:none; fill-opacity:0" cx="2183.84" cy="735.212" r="2"/>
+<path clip-path="url(#clip212)" d="M478.6 597.618 L478.6 1079.2 L598.266 1079.2 L598.266 597.618 L478.6 597.618 L478.6 597.618  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="478.6,597.618 478.6,1079.2 598.266,1079.2 598.266,597.618 478.6,597.618 "/>
+<path clip-path="url(#clip212)" d="M628.182 735.212 L628.182 1044.8 L747.848 1044.8 L747.848 735.212 L628.182 735.212 L628.182 735.212  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="628.182,735.212 628.182,1044.8 747.848,1044.8 747.848,735.212 628.182,735.212 "/>
+<path clip-path="url(#clip212)" d="M777.765 528.822 L777.765 1010.4 L897.431 1010.4 L897.431 528.822 L777.765 528.822 L777.765 528.822  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="777.765,528.822 777.765,1010.4 897.431,1010.4 897.431,528.822 777.765,528.822 "/>
+<path clip-path="url(#clip212)" d="M927.348 528.822 L927.348 735.212 L1047.01 735.212 L1047.01 528.822 L927.348 528.822 L927.348 528.822  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="927.348,528.822 927.348,735.212 1047.01,735.212 1047.01,528.822 927.348,528.822 "/>
+<path clip-path="url(#clip212)" d="M1076.93 460.025 L1076.93 1423.18 L1196.6 1423.18 L1196.6 460.025 L1076.93 460.025 L1076.93 460.025  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1076.93,460.025 1076.93,1423.18 1196.6,1423.18 1196.6,460.025 1076.93,460.025 "/>
+<path clip-path="url(#clip212)" d="M1226.51 528.822 L1226.51 941.602 L1346.18 941.602 L1346.18 528.822 L1226.51 528.822 L1226.51 528.822  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1226.51,528.822 1226.51,941.602 1346.18,941.602 1346.18,528.822 1226.51,528.822 "/>
+<path clip-path="url(#clip212)" d="M1376.1 391.228 L1376.1 1010.4 L1495.76 1010.4 L1495.76 391.228 L1376.1 391.228 L1376.1 391.228  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1376.1,391.228 1376.1,1010.4 1495.76,1010.4 1495.76,391.228 1376.1,391.228 "/>
+<path clip-path="url(#clip212)" d="M1525.68 253.634 L1525.68 735.212 L1645.34 735.212 L1645.34 253.634 L1525.68 253.634 L1525.68 253.634  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1525.68,253.634 1525.68,735.212 1645.34,735.212 1645.34,253.634 1525.68,253.634 "/>
+<path clip-path="url(#clip212)" d="M1675.26 47.2441 L1675.26 735.212 L1794.93 735.212 L1794.93 47.2441 L1675.26 47.2441 L1675.26 47.2441  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1675.26,47.2441 1675.26,735.212 1794.93,735.212 1794.93,47.2441 1675.26,47.2441 "/>
+<path clip-path="url(#clip212)" d="M1824.84 116.041 L1824.84 1423.18 L1944.51 1423.18 L1944.51 116.041 L1824.84 116.041 L1824.84 116.041  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1824.84,116.041 1824.84,1423.18 1944.51,1423.18 1944.51,116.041 1824.84,116.041 "/>
+<path clip-path="url(#clip212)" d="M1974.43 460.025 L1974.43 1010.4 L2094.09 1010.4 L2094.09 460.025 L1974.43 460.025 L1974.43 460.025  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1974.43,460.025 1974.43,1010.4 2094.09,1010.4 2094.09,460.025 1974.43,460.025 "/>
+<path clip-path="url(#clip212)" d="M2124.01 735.212 L2124.01 1423.18 L2243.67 1423.18 L2243.67 735.212 L2124.01 735.212 L2124.01 735.212  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2124.01,735.212 2124.01,1423.18 2243.67,1423.18 2243.67,735.212 2124.01,735.212 "/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="538.433" cy="597.618" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="688.015" cy="735.212" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="837.598" cy="528.822" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="987.181" cy="528.822" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1136.76" cy="460.025" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1286.35" cy="528.822" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1435.93" cy="391.228" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1585.51" cy="253.634" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1735.09" cy="47.2441" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="1884.68" cy="116.041" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="2034.26" cy="460.025" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#4063d8; stroke:none; fill-opacity:0" cx="2183.84" cy="735.212" r="2"/>
+<path clip-path="url(#clip212)" d="M478.6 1079.2 L478.6 1423.18 L598.266 1423.18 L598.266 1079.2 L478.6 1079.2 L478.6 1079.2  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="478.6,1079.2 478.6,1423.18 598.266,1423.18 598.266,1079.2 478.6,1079.2 "/>
+<path clip-path="url(#clip212)" d="M628.182 1044.8 L628.182 1423.18 L747.848 1423.18 L747.848 1044.8 L628.182 1044.8 L628.182 1044.8  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="628.182,1044.8 628.182,1423.18 747.848,1423.18 747.848,1044.8 628.182,1044.8 "/>
+<path clip-path="url(#clip212)" d="M777.765 1010.4 L777.765 1423.18 L897.431 1423.18 L897.431 1010.4 L777.765 1010.4 L777.765 1010.4  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="777.765,1010.4 777.765,1423.18 897.431,1423.18 897.431,1010.4 777.765,1010.4 "/>
+<path clip-path="url(#clip212)" d="M927.348 735.212 L927.348 1423.18 L1047.01 1423.18 L1047.01 735.212 L927.348 735.212 L927.348 735.212  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="927.348,735.212 927.348,1423.18 1047.01,1423.18 1047.01,735.212 927.348,735.212 "/>
+<path clip-path="url(#clip212)" d="M1076.93 1423.18 L1076.93 1423.18 L1196.6 1423.18 L1196.6 1423.18 L1076.93 1423.18 L1076.93 1423.18  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1076.93,1423.18 1076.93,1423.18 1196.6,1423.18 1076.93,1423.18 "/>
+<path clip-path="url(#clip212)" d="M1226.51 941.602 L1226.51 1423.18 L1346.18 1423.18 L1346.18 941.602 L1226.51 941.602 L1226.51 941.602  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1226.51,941.602 1226.51,1423.18 1346.18,1423.18 1346.18,941.602 1226.51,941.602 "/>
+<path clip-path="url(#clip212)" d="M1376.1 1010.4 L1376.1 1423.18 L1495.76 1423.18 L1495.76 1010.4 L1376.1 1010.4 L1376.1 1010.4  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1376.1,1010.4 1376.1,1423.18 1495.76,1423.18 1495.76,1010.4 1376.1,1010.4 "/>
+<path clip-path="url(#clip212)" d="M1525.68 735.212 L1525.68 1423.18 L1645.34 1423.18 L1645.34 735.212 L1525.68 735.212 L1525.68 735.212  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1525.68,735.212 1525.68,1423.18 1645.34,1423.18 1645.34,735.212 1525.68,735.212 "/>
+<path clip-path="url(#clip212)" d="M1675.26 735.212 L1675.26 1423.18 L1794.93 1423.18 L1794.93 735.212 L1675.26 735.212 L1675.26 735.212  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1675.26,735.212 1675.26,1423.18 1794.93,1423.18 1794.93,735.212 1675.26,735.212 "/>
+<path clip-path="url(#clip212)" d="M1824.84 1423.18 L1824.84 1423.18 L1944.51 1423.18 L1944.51 1423.18 L1824.84 1423.18 L1824.84 1423.18  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1824.84,1423.18 1824.84,1423.18 1944.51,1423.18 1824.84,1423.18 "/>
+<path clip-path="url(#clip212)" d="M1974.43 1010.4 L1974.43 1423.18 L2094.09 1423.18 L2094.09 1010.4 L1974.43 1010.4 L1974.43 1010.4  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1974.43,1010.4 1974.43,1423.18 2094.09,1423.18 2094.09,1010.4 1974.43,1010.4 "/>
+<path clip-path="url(#clip212)" d="M2124.01 1423.18 L2124.01 1423.18 L2243.67 1423.18 L2243.67 1423.18 L2124.01 1423.18 L2124.01 1423.18  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip212)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2124.01,1423.18 2124.01,1423.18 2243.67,1423.18 2124.01,1423.18 "/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="538.433" cy="1079.2" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="688.015" cy="1044.8" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="837.598" cy="1010.4" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="987.181" cy="735.212" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1136.76" cy="1423.18" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1286.35" cy="941.602" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1435.93" cy="1010.4" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1585.51" cy="735.212" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1735.09" cy="735.212" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="1884.68" cy="1423.18" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="2034.26" cy="1010.4" r="2"/>
+<circle clip-path="url(#clip212)" style="fill:#a0b1ec; stroke:none; fill-opacity:0" cx="2183.84" cy="1423.18" r="2"/>
+<path clip-path="url(#clip210)" d="M435.626 300.469 L989.342 300.469 L989.342 93.1086 L435.626 93.1086  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="435.626,300.469 989.342,300.469 989.342,93.1086 435.626,93.1086 435.626,300.469 "/>
+<path clip-path="url(#clip210)" d="M457.662 165.685 L589.878 165.685 L589.878 124.213 L457.662 124.213 L457.662 165.685  Z" fill="#20326c" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="457.662,165.685 589.878,165.685 589.878,124.213 457.662,124.213 457.662,165.685 "/>
+<path clip-path="url(#clip210)" d="M611.914 151.997 L611.914 136.303 L616.173 136.303 L616.173 151.835 Q616.173 155.516 617.608 157.368 Q619.044 159.196 621.914 159.196 Q625.363 159.196 627.354 156.997 Q629.368 154.798 629.368 151.002 L629.368 136.303 L633.627 136.303 L633.627 162.229 L629.368 162.229 L629.368 158.247 Q627.817 160.608 625.756 161.766 Q623.719 162.9 621.011 162.9 Q616.544 162.9 614.229 160.122 Q611.914 157.344 611.914 151.997 M622.631 135.678 L622.631 135.678 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M663.951 146.581 L663.951 162.229 L659.691 162.229 L659.691 146.719 Q659.691 143.039 658.256 141.21 Q656.821 139.382 653.951 139.382 Q650.502 139.382 648.511 141.581 Q646.52 143.78 646.52 147.576 L646.52 162.229 L642.238 162.229 L642.238 136.303 L646.52 136.303 L646.52 140.331 Q648.048 137.993 650.108 136.835 Q652.192 135.678 654.9 135.678 Q659.367 135.678 661.659 138.456 Q663.951 141.21 663.951 146.581 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M692.631 141.28 Q694.228 138.409 696.451 137.044 Q698.673 135.678 701.682 135.678 Q705.733 135.678 707.932 138.525 Q710.131 141.349 710.131 146.581 L710.131 162.229 L705.849 162.229 L705.849 146.719 Q705.849 142.993 704.529 141.187 Q703.21 139.382 700.501 139.382 Q697.191 139.382 695.27 141.581 Q693.349 143.78 693.349 147.576 L693.349 162.229 L689.066 162.229 L689.066 146.719 Q689.066 142.969 687.747 141.187 Q686.427 139.382 683.673 139.382 Q680.409 139.382 678.488 141.604 Q676.566 143.803 676.566 147.576 L676.566 162.229 L672.284 162.229 L672.284 136.303 L676.566 136.303 L676.566 140.331 Q678.025 137.946 680.062 136.812 Q682.099 135.678 684.9 135.678 Q687.724 135.678 689.691 137.113 Q691.682 138.548 692.631 141.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M740.802 148.201 L740.802 150.284 L721.219 150.284 Q721.497 154.682 723.858 156.997 Q726.242 159.289 730.478 159.289 Q732.932 159.289 735.223 158.687 Q737.538 158.085 739.807 156.881 L739.807 160.909 Q737.515 161.881 735.108 162.391 Q732.7 162.9 730.224 162.9 Q724.02 162.9 720.386 159.289 Q716.775 155.678 716.775 149.52 Q716.775 143.155 720.2 139.428 Q723.649 135.678 729.483 135.678 Q734.714 135.678 737.747 139.057 Q740.802 142.414 740.802 148.201 M736.543 146.951 Q736.497 143.456 734.575 141.372 Q732.677 139.289 729.529 139.289 Q725.964 139.289 723.812 141.303 Q721.682 143.317 721.358 146.974 L736.543 146.951 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M752.006 128.942 L752.006 136.303 L760.779 136.303 L760.779 139.613 L752.006 139.613 L752.006 153.687 Q752.006 156.858 752.862 157.761 Q753.742 158.664 756.404 158.664 L760.779 158.664 L760.779 162.229 L756.404 162.229 Q751.473 162.229 749.598 160.4 Q747.723 158.548 747.723 153.687 L747.723 139.613 L744.598 139.613 L744.598 136.303 L747.723 136.303 L747.723 128.942 L752.006 128.942 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M798.51 140.238 L798.51 126.21 L802.769 126.21 L802.769 162.229 L798.51 162.229 L798.51 158.34 Q797.168 160.655 795.107 161.789 Q793.07 162.9 790.2 162.9 Q785.501 162.9 782.538 159.15 Q779.598 155.4 779.598 149.289 Q779.598 143.178 782.538 139.428 Q785.501 135.678 790.2 135.678 Q793.07 135.678 795.107 136.812 Q797.168 137.923 798.51 140.238 M783.996 149.289 Q783.996 153.988 785.918 156.673 Q787.862 159.335 791.242 159.335 Q794.621 159.335 796.566 156.673 Q798.51 153.988 798.51 149.289 Q798.51 144.59 796.566 141.928 Q794.621 139.243 791.242 139.243 Q787.862 139.243 785.918 141.928 Q783.996 144.59 783.996 149.289 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M833.718 148.201 L833.718 150.284 L814.135 150.284 Q814.413 154.682 816.774 156.997 Q819.158 159.289 823.394 159.289 Q825.848 159.289 828.14 158.687 Q830.454 158.085 832.723 156.881 L832.723 160.909 Q830.431 161.881 828.024 162.391 Q825.617 162.9 823.14 162.9 Q816.936 162.9 813.302 159.289 Q809.691 155.678 809.691 149.52 Q809.691 143.155 813.117 139.428 Q816.566 135.678 822.399 135.678 Q827.63 135.678 830.663 139.057 Q833.718 142.414 833.718 148.201 M829.459 146.951 Q829.413 143.456 827.492 141.372 Q825.593 139.289 822.445 139.289 Q818.88 139.289 816.728 141.303 Q814.598 143.317 814.274 146.974 L829.459 146.951 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M860.894 141.28 Q862.491 138.409 864.714 137.044 Q866.936 135.678 869.945 135.678 Q873.996 135.678 876.195 138.525 Q878.394 141.349 878.394 146.581 L878.394 162.229 L874.112 162.229 L874.112 146.719 Q874.112 142.993 872.792 141.187 Q871.473 139.382 868.764 139.382 Q865.454 139.382 863.533 141.581 Q861.612 143.78 861.612 147.576 L861.612 162.229 L857.329 162.229 L857.329 146.719 Q857.329 142.969 856.01 141.187 Q854.69 139.382 851.936 139.382 Q848.672 139.382 846.751 141.604 Q844.829 143.803 844.829 147.576 L844.829 162.229 L840.547 162.229 L840.547 136.303 L844.829 136.303 L844.829 140.331 Q846.288 137.946 848.325 136.812 Q850.362 135.678 853.163 135.678 Q855.987 135.678 857.954 137.113 Q859.945 138.548 860.894 141.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M898.672 149.196 Q893.51 149.196 891.519 150.377 Q889.528 151.557 889.528 154.405 Q889.528 156.673 891.01 158.016 Q892.514 159.335 895.084 159.335 Q898.625 159.335 900.755 156.835 Q902.908 154.312 902.908 150.145 L902.908 149.196 L898.672 149.196 M907.167 147.437 L907.167 162.229 L902.908 162.229 L902.908 158.293 Q901.449 160.655 899.274 161.789 Q897.098 162.9 893.949 162.9 Q889.968 162.9 887.607 160.678 Q885.269 158.432 885.269 154.682 Q885.269 150.307 888.186 148.085 Q891.125 145.863 896.936 145.863 L902.908 145.863 L902.908 145.446 Q902.908 142.507 900.963 140.909 Q899.042 139.289 895.547 139.289 Q893.324 139.289 891.218 139.821 Q889.112 140.354 887.167 141.419 L887.167 137.483 Q889.505 136.581 891.704 136.141 Q893.903 135.678 895.987 135.678 Q901.611 135.678 904.389 138.594 Q907.167 141.511 907.167 147.437 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M937.491 146.581 L937.491 162.229 L933.232 162.229 L933.232 146.719 Q933.232 143.039 931.797 141.21 Q930.361 139.382 927.491 139.382 Q924.042 139.382 922.051 141.581 Q920.06 143.78 920.06 147.576 L920.06 162.229 L915.778 162.229 L915.778 136.303 L920.06 136.303 L920.06 140.331 Q921.588 137.993 923.648 136.835 Q925.732 135.678 928.44 135.678 Q932.908 135.678 935.199 138.456 Q937.491 141.21 937.491 146.581 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M963.046 140.238 L963.046 126.21 L967.306 126.21 L967.306 162.229 L963.046 162.229 L963.046 158.34 Q961.704 160.655 959.644 161.789 Q957.607 162.9 954.736 162.9 Q950.037 162.9 947.074 159.15 Q944.134 155.4 944.134 149.289 Q944.134 143.178 947.074 139.428 Q950.037 135.678 954.736 135.678 Q957.607 135.678 959.644 136.812 Q961.704 137.923 963.046 140.238 M948.533 149.289 Q948.533 153.988 950.454 156.673 Q952.398 159.335 955.778 159.335 Q959.157 159.335 961.102 156.673 Q963.046 153.988 963.046 149.289 Q963.046 144.59 961.102 141.928 Q959.157 139.243 955.778 139.243 Q952.398 139.243 950.454 141.928 Q948.533 144.59 948.533 149.289 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M457.662 217.525 L589.878 217.525 L589.878 176.053 L457.662 176.053 L457.662 217.525  Z" fill="#4063d8" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="457.662,217.525 589.878,217.525 589.878,176.053 457.662,176.053 457.662,217.525 "/>
+<path clip-path="url(#clip210)" d="M627.747 184.115 L621.405 201.314 L634.113 201.314 L627.747 184.115 M625.108 179.509 L630.409 179.509 L643.58 214.069 L638.719 214.069 L635.571 205.203 L619.993 205.203 L616.844 214.069 L611.914 214.069 L625.108 179.509 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip210)" d="M457.662 269.365 L589.878 269.365 L589.878 227.893 L457.662 227.893 L457.662 269.365  Z" fill="#a0b1ec" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip210)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="457.662,269.365 589.878,269.365 589.878,227.893 457.662,227.893 457.662,269.365 "/>
+<path clip-path="url(#clip210)" d="M616.59 249.404 L616.59 262.066 L624.09 262.066 Q627.863 262.066 629.668 260.515 Q631.497 258.941 631.497 255.723 Q631.497 252.483 629.668 250.955 Q627.863 249.404 624.09 249.404 L616.59 249.404 M616.59 235.191 L616.59 245.608 L623.511 245.608 Q626.937 245.608 628.604 244.335 Q630.293 243.038 630.293 240.399 Q630.293 237.784 628.604 236.487 Q626.937 235.191 623.511 235.191 L616.59 235.191 M611.914 231.349 L623.858 231.349 Q629.206 231.349 632.099 233.571 Q634.993 235.793 634.993 239.89 Q634.993 243.062 633.511 244.937 Q632.03 246.811 629.159 247.274 Q632.608 248.015 634.506 250.376 Q636.428 252.714 636.428 256.233 Q636.428 260.862 633.28 263.385 Q630.131 265.909 624.321 265.909 L611.914 265.909 L611.914 231.349 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/linear/factory_schedule/index.html b/previews/PR3525/tutorials/linear/factory_schedule/index.html
new file mode 100644
index 00000000000..ce2a83f201e
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/factory_schedule/index.html
@@ -0,0 +1,165 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The factory schedule example · JuMP</title><meta name="title" content="The factory schedule example · JuMP"/><meta property="og:title" content="The factory schedule example · JuMP"/><meta property="twitter:title" content="The factory schedule example · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li class="is-active"><a class="tocitem" href>The factory schedule example</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Solution"><span>Solution</span></a></li><li><a class="tocitem" href="#What-happens-if-demand-increases?"><span>What happens if demand increases?</span></a></li><li><a class="tocitem" href="#How-sensitive-is-the-solution-to-changes-in-variable-cost?"><span>How sensitive is the solution to changes in variable cost?</span></a></li></ul></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The factory schedule example</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The factory schedule example</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/factory_schedule.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-factory-schedule-example"><a class="docs-heading-anchor" href="#The-factory-schedule-example">The factory schedule example</a><a id="The-factory-schedule-example-1"></a><a class="docs-heading-anchor-permalink" href="#The-factory-schedule-example" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../factory_schedule.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by <code>@Crghilardi</code>.</strong></p><p>This tutorial is a Julia translation of Part 5 from <a href="https://github.com/benalexkeen/Introduction-to-linear-programming">Introduction to Linear Programming with Python</a>.</p><p>The purpose of this tutorial is to demonstrate how to use DataFrames and delimited files, and to structure your code that is robust to infeasibilities and permits running with different datasets.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import CSV
+import DataFrames
+import HiGHS
+import StatsPlots</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>The Factory Scheduling Problem assumes we are optimizing the production of a good from factories <span>$f \in F$</span> over the course of 12 months <span>$m \in M$</span>.</p><p>If a factory <span>$f$</span> runs during a month <span>$m$</span>, a fixed cost of <span>$a_f$</span> is incurred, the factory must produce <span>$x_{m,f}$</span> units that is within some minimum and maximum production levels <span>$l_f$</span> and <span>$u_f$</span> respectively, and each unit of production incurs a variable cost <span>$c_f$</span>. Otherwise, the factory can be shut for the month with zero production and no fixed-cost is incurred. We denote the run/not-run decision by <span>$z_{m,f} \in \{0, 1\}$</span>, where <span>$z_{m,f}$</span> is <span>$1$</span> if factory <span>$f$</span> runs in month <span>$m$</span>. The factory must produce enough units to satisfy demand <span>$d_m$</span>.</p><p>With a little effort, we can formulate our problem as the following linear program:</p><p class="math-container">\[\begin{aligned}
+\min &amp; \sum\limits_{f \in F, m \in M} a_f z_{m,f} + c_f x_{m,f} \\
+\text{s.t.} &amp; x_{m,f} \le u_f z_{m,f} &amp;&amp; \forall f \in F, m \in M \\
+            &amp; x_{m,f} \ge l_f z_{m,f} &amp;&amp; \forall f \in F, m \in M \\
+            &amp; \sum\limits_{f\in F} x_{m,f} = d_m &amp;&amp; \forall f \in F, m \in M \\
+            &amp; z_{m,f} \in \{0, 1\} &amp;&amp; \forall f \in F, m \in M.
+\end{aligned}\]</p><p>However, this formulation has a problem: if demand is too high, we may be unable to satisfy the demand constraint, and the problem will be infeasible.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>When modeling, consider ways to formulate your model such that it always has a feasible solution. This greatly simplifies debugging data errors that would otherwise result in an infeasible solution. In practice, most practical decisions have a feasible solution. In our case, we could satisfy demand (at a high cost) by buying replacement items for the buyer, or running the factories in overtime to make up the difference.</p></div></div><p>We can improve our model by adding a new variable, <span>$\delta_m$</span>, which represents the quantity of unmet demand in each month <span>$m$</span>. We penalize <span>$\delta_m$</span> by an arbitrarily large value of <span>$</span>10,000/unit in the objective.</p><p class="math-container">\[\begin{aligned}
+\min &amp; \sum\limits_{f \in F, m \in M} a_f z_{m,f} + c_f x_{m,f} + \sum\limits_{m \in M}10000 \delta_m \\
+\text{s.t.} &amp; x_{m,f} \le u_f z_{m,f} &amp;&amp; \forall f \in F, m \in M \\
+            &amp; x_{m,f} \ge l_f z_{m,f} &amp;&amp; \forall f \in F, m \in M \\
+            &amp; \sum\limits_{f\in F} x_{m,f} - \delta_m = d_m &amp;&amp; \forall f \in F, m \in M \\
+            &amp; z_{m,f} \in \{0, 1\} &amp;&amp; \forall f \in F, m \in M \\
+            &amp; \delta_m \ge 0 &amp;&amp; \forall m \in M.
+\end{aligned}\]</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>The JuMP GitHub repository contains two text files with the data we need for this tutorial.</p><p>The first file contains a dataset of our factories, <code>A</code> and <code>B</code>, with their production and cost levels for each month. For the documentation, the file is located at:</p><pre><code class="language-julia hljs">factories_filename = joinpath(@__DIR__, &quot;factory_schedule_factories.txt&quot;);</code></pre><p>To run locally, download <a href="../factory_schedule_factories.txt"><code>factory_schedule_factories.txt</code></a> and update <code>factories_filename</code> appropriately.</p><p>The file has the following contents:</p><pre><code class="language-julia hljs">print(read(factories_filename, String))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">factory month min_production max_production fixed_cost variable_cost
+A       1     20000          100000         500        10
+A       2     20000          110000         500        11
+A       3     20000          120000         500        12
+A       4     20000          145000         500        9
+A       5     20000          160000         500        8
+A       6     20000          140000         500        8
+A       7     20000          155000         500        5
+A       8     20000          200000         500        7
+A       9     20000          210000         500        9
+A       10    20000          197000         500        10
+A       11    20000          80000          500        8
+A       12    20000          150000         500        8
+B       1     20000          50000          600        5
+B       2     20000          55000          600        4
+B       3     20000          60000          600        3
+B       4     20000          100000         600        5
+B       5     0              0              0          0
+B       6     20000          70000          600        6
+B       7     20000          60000          600        4
+B       8     20000          100000         600        6
+B       9     20000          100000         600        8
+B       10    20000          100000         600        11
+B       11    20000          120000         600        10
+B       12    20000          150000         600        12</code></pre><p>We use the <code>CSV</code> and <code>DataFrames</code> packages to read it into Julia:</p><pre><code class="language-julia hljs">factory_df = CSV.read(
+    factories_filename,
+    DataFrames.DataFrame;
+    delim = &#39; &#39;,
+    ignorerepeated = true,
+)</code></pre><div><div style = "float: left;"><span>24×6 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">factory</th><th style = "text-align: left;">month</th><th style = "text-align: left;">min_production</th><th style = "text-align: left;">max_production</th><th style = "text-align: left;">fixed_cost</th><th style = "text-align: left;">variable_cost</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "InlineStrings.String1" style = "text-align: left;">String1</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">A</td><td style = "text-align: right;">1</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">100000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">10</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">A</td><td style = "text-align: right;">2</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">110000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">11</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">A</td><td style = "text-align: right;">3</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">120000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">12</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">A</td><td style = "text-align: right;">4</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">145000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">9</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">A</td><td style = "text-align: right;">5</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">160000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">A</td><td style = "text-align: right;">6</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">140000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">A</td><td style = "text-align: right;">7</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">155000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">A</td><td style = "text-align: right;">8</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">200000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">7</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">A</td><td style = "text-align: right;">9</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">210000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">9</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">A</td><td style = "text-align: right;">10</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">197000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">10</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">A</td><td style = "text-align: right;">11</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">80000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">A</td><td style = "text-align: right;">12</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">150000</td><td style = "text-align: right;">500</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">B</td><td style = "text-align: right;">1</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">50000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">B</td><td style = "text-align: right;">2</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">55000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">4</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">B</td><td style = "text-align: right;">3</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">60000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">3</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">B</td><td style = "text-align: right;">4</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">100000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">B</td><td style = "text-align: right;">5</td><td style = "text-align: right;">0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">0</td><td style = "text-align: right;">0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">B</td><td style = "text-align: right;">6</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">70000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">6</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">B</td><td style = "text-align: right;">7</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">60000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">4</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">B</td><td style = "text-align: right;">8</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">100000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">6</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">21</td><td style = "text-align: left;">B</td><td style = "text-align: right;">9</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">100000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">8</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">22</td><td style = "text-align: left;">B</td><td style = "text-align: right;">10</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">100000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">11</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">23</td><td style = "text-align: left;">B</td><td style = "text-align: right;">11</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">120000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">10</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">24</td><td style = "text-align: left;">B</td><td style = "text-align: right;">12</td><td style = "text-align: right;">20000</td><td style = "text-align: right;">150000</td><td style = "text-align: right;">600</td><td style = "text-align: right;">12</td></tr></tbody></table></div><p>The second file contains the demand data by month:</p><pre><code class="language-julia hljs">demand_filename = joinpath(@__DIR__, &quot;factory_schedule_demand.txt&quot;);</code></pre><p>To run locally, download <a href="../factory_schedule_demand.txt"><code>factory_schedule_demand.txt</code></a> and update <code>demand_filename</code> appropriately.</p><pre><code class="language-julia hljs">demand_df = CSV.read(
+    demand_filename,
+    DataFrames.DataFrame;
+    delim = &#39; &#39;,
+    ignorerepeated = true,
+)</code></pre><div><div style = "float: left;"><span>12×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">month</th><th style = "text-align: left;">demand</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Int64" style = "text-align: left;">Int64</th><th title = "Int64" style = "text-align: left;">Int64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">1</td><td style = "text-align: right;">120000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">2</td><td style = "text-align: right;">100000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">3</td><td style = "text-align: right;">130000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">4</td><td style = "text-align: right;">130000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">5</td><td style = "text-align: right;">140000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">6</td><td style = "text-align: right;">130000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">7</td><td style = "text-align: right;">150000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">8</td><td style = "text-align: right;">170000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">9</td><td style = "text-align: right;">200000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">10</td><td style = "text-align: right;">190000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">11</td><td style = "text-align: right;">140000</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">12</td><td style = "text-align: right;">100000</td></tr></tbody></table></div><h3 id="Data-validation"><a class="docs-heading-anchor" href="#Data-validation">Data validation</a><a id="Data-validation-1"></a><a class="docs-heading-anchor-permalink" href="#Data-validation" title="Permalink"></a></h3><p>Before moving on, it&#39;s always good practice to validate the data you read from external sources. The more effort you spend here, the fewer issues you will have later. The following function contains a few simple checks, but we could add more. For example, you might want to check that none of the values are too large (or too small), which might indicate a typo or a unit conversion issue (perhaps the variable costs are in <span>$</span>/1000 units instead of <span>$</span>/unit).</p><pre><code class="language-julia hljs">function valiate_data(
+    demand_df::DataFrames.DataFrame,
+    factory_df::DataFrames.DataFrame,
+)
+    # Minimum production must not exceed maximum production.
+    @assert all(factory_df.min_production .&lt;= factory_df.max_production)
+    # Demand, minimum production, fixed costs, and variable costs must all be
+    # non-negative.
+    @assert all(demand_df.demand .&gt;= 0)
+    @assert all(factory_df.min_production .&gt;= 0)
+    @assert all(factory_df.fixed_cost .&gt;= 0)
+    @assert all(factory_df.variable_cost .&gt;= 0)
+    return
+end
+
+valiate_data(demand_df, factory_df)</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>Next, we need to code our JuMP formulation. As shown in <a href="../../getting_started/design_patterns_for_larger_models/#Design-patterns-for-larger-models">Design patterns for larger models</a>, it&#39;s always good practice to code your model in a function that accepts well-defined input and returns well-defined output.</p><pre><code class="language-julia hljs">function solve_factory_scheduling(
+    demand_df::DataFrames.DataFrame,
+    factory_df::DataFrames.DataFrame,
+)
+    # Even though we validated the data above, it&#39;s good practice to do it here
+    # too.
+    valiate_data(demand_df, factory_df)
+    months, factories = unique(factory_df.month), unique(factory_df.factory)
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, status[months, factories], Bin)
+    @variable(model, production[months, factories], Int)
+    @variable(model, unmet_demand[months] &gt;= 0)
+    # We use `eachrow` to loop through the rows of the dataframe and add the
+    # relevant constraints.
+    for r in eachrow(factory_df)
+        m, f = r.month, r.factory
+        @constraint(model, production[m, f] &lt;= r.max_production * status[m, f])
+        @constraint(model, production[m, f] &gt;= r.min_production * status[m, f])
+    end
+    @constraint(
+        model,
+        [r in eachrow(demand_df)],
+        sum(production[r.month, :]) + unmet_demand[r.month] == r.demand,
+    )
+    @objective(
+        model,
+        Min,
+        10_000 * sum(unmet_demand) + sum(
+            r.fixed_cost * status[r.month, r.factory] +
+            r.variable_cost * production[r.month, r.factory] for
+            r in eachrow(factory_df)
+        )
+    )
+    optimize!(model)
+    schedules = Dict{Symbol,Vector{Float64}}(
+        Symbol(f) =&gt; value.(production[:, f]) for f in factories
+    )
+    schedules[:unmet_demand] = value.(unmet_demand)
+    return (
+        termination_status = termination_status(model),
+        cost = objective_value(model),
+        # This `select` statement re-orders the columns in the DataFrame.
+        schedules = DataFrames.select(
+            DataFrames.DataFrame(schedules),
+            [:unmet_demand, :A, :B],
+        ),
+    )
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_factory_scheduling (generic function with 1 method)</code></pre><h2 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h2><p>Now we can call our <code>solve_factory_scheduling</code> function using the data we read in above.</p><pre><code class="language-julia hljs">solution = solve_factory_scheduling(demand_df, factory_df);</code></pre><p>Let&#39;s see what <code>solution</code> contains:</p><pre><code class="language-julia hljs">solution.termination_status</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">OPTIMAL::TerminationStatusCode = 1</code></pre><pre><code class="language-julia hljs">solution.cost</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1.29064e7</code></pre><pre><code class="language-julia hljs">solution.schedules</code></pre><div><div style = "float: left;"><span>12×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">unmet_demand</th><th style = "text-align: left;">A</th><th style = "text-align: left;">B</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">70000.0</td><td style = "text-align: right;">50000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">45000.0</td><td style = "text-align: right;">55000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">70000.0</td><td style = "text-align: right;">60000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">30000.0</td><td style = "text-align: right;">100000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">140000.0</td><td style = "text-align: right;">-0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">60000.0</td><td style = "text-align: right;">70000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">90000.0</td><td style = "text-align: right;">60000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">70000.0</td><td style = "text-align: right;">100000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">100000.0</td><td style = "text-align: right;">100000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">190000.0</td><td style = "text-align: right;">-0.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">80000.0</td><td style = "text-align: right;">60000.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">100000.0</td><td style = "text-align: right;">-0.0</td></tr></tbody></table></div><p>These schedules will be easier to visualize as a graph:</p><pre><code class="language-julia hljs">StatsPlots.groupedbar(
+    Matrix(solution.schedules);
+    bar_position = :stack,
+    labels = [&quot;unmet demand&quot; &quot;A&quot; &quot;B&quot;],
+    xlabel = &quot;Month&quot;,
+    ylabel = &quot;Production&quot;,
+    legend = :topleft,
+    color = [&quot;#20326c&quot; &quot;#4063d8&quot; &quot;#a0b1ec&quot;],
+)</code></pre><img src="c5e9cd8e.svg" alt="Example block output"/><p>Note that we don&#39;t have any unmet demand.</p><h2 id="What-happens-if-demand-increases?"><a class="docs-heading-anchor" href="#What-happens-if-demand-increases?">What happens if demand increases?</a><a id="What-happens-if-demand-increases?-1"></a><a class="docs-heading-anchor-permalink" href="#What-happens-if-demand-increases?" title="Permalink"></a></h2><p>Let&#39;s run an experiment by increasing the demand by 50% in all time periods:</p><pre><code class="language-julia hljs">demand_df.demand .*= 1.5</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">12-element Vector{Float64}:
+ 180000.0
+ 150000.0
+ 195000.0
+ 195000.0
+ 210000.0
+ 195000.0
+ 225000.0
+ 255000.0
+ 300000.0
+ 285000.0
+ 210000.0
+ 150000.0</code></pre><p>Now we resolve the problem:</p><pre><code class="language-julia hljs">high_demand_solution = solve_factory_scheduling(demand_df, factory_df);</code></pre><p>and visualize the solution:</p><pre><code class="language-julia hljs">StatsPlots.groupedbar(
+    Matrix(high_demand_solution.schedules);
+    bar_position = :stack,
+    labels = [&quot;unmet demand&quot; &quot;A&quot; &quot;B&quot;],
+    xlabel = &quot;Month&quot;,
+    ylabel = &quot;Production&quot;,
+    legend = :topleft,
+    color = [&quot;#20326c&quot; &quot;#4063d8&quot; &quot;#a0b1ec&quot;],
+)</code></pre><img src="34a51857.svg" alt="Example block output"/><p>Uh oh, we can&#39;t satisfy all of the demand.</p><h2 id="How-sensitive-is-the-solution-to-changes-in-variable-cost?"><a class="docs-heading-anchor" href="#How-sensitive-is-the-solution-to-changes-in-variable-cost?">How sensitive is the solution to changes in variable cost?</a><a id="How-sensitive-is-the-solution-to-changes-in-variable-cost?-1"></a><a class="docs-heading-anchor-permalink" href="#How-sensitive-is-the-solution-to-changes-in-variable-cost?" title="Permalink"></a></h2><p>Let&#39;s run another experiment, this time seeing how the optimal objective value changes as we vary the variable costs of each factory.</p><p>First though, let&#39;s reset the demand to it&#39;s original level:</p><pre><code class="language-julia hljs">demand_df.demand ./= 1.5;</code></pre><p>For our experiment, we&#39;re going to scale the variable costs of both factories by a set of values from <code>0.0</code> to <code>1.5</code>:</p><pre><code class="language-julia hljs">scale_factors = 0:0.1:1.5</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.0:0.1:1.5</code></pre><p>At a high level, we&#39;re going to loop over the scale factors for A, then the scale factors for B, rescale the input data, call our <code>solve_factory_scheduling</code> example, and then store the optimal objective value in the following <code>cost</code> matrix:</p><pre><code class="language-julia hljs">cost = zeros(length(scale_factors), length(scale_factors));</code></pre><p>Because we&#39;re modifying <code>factory_df</code> in-place, we need to store the original variable costs in a new column:</p><pre><code class="language-julia hljs">factory_df[!, :old_variable_cost] = copy(factory_df.variable_cost);</code></pre><p>Then, we need a function to scale the <code>:variable_cost</code> column for a particular factory by a value <code>scale</code>:</p><pre><code class="language-julia hljs">function scale_variable_cost(df, factory, scale)
+    rows = df.factory .== factory
+    df[rows, :variable_cost] .=
+        round.(Int, df[rows, :old_variable_cost] .* scale)
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">scale_variable_cost (generic function with 1 method)</code></pre><p>Our experiment is just a nested for-loop, modifying A and B and storing the cost:</p><pre><code class="language-julia hljs">for (j, a) in enumerate(scale_factors)
+    scale_variable_cost(factory_df, &quot;A&quot;, a)
+    for (i, b) in enumerate(scale_factors)
+        scale_variable_cost(factory_df, &quot;B&quot;, b)
+        cost[i, j] = solve_factory_scheduling(demand_df, factory_df).cost
+    end
+end</code></pre><p>Let&#39;s visualize the cost matrix:</p><pre><code class="language-julia hljs">StatsPlots.contour(
+    scale_factors,
+    scale_factors,
+    cost;
+    xlabel = &quot;Scale of factory A&quot;,
+    ylabel = &quot;Scale of factory B&quot;,
+)</code></pre><img src="4024f2dd.svg" alt="Example block output"/><p>What can you infer from the solution?</p><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>The <a href="../../applications/power_systems/#Power-Systems">Power Systems</a> tutorial explains a number of other ways you can structure a problem to perform a parametric analysis of the solution. In particular, you can use in-place modification to reduce the time it takes to build and solve the resulting models.</p></div></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../cannery/">« The cannery problem</a><a class="docs-footer-nextpage" href="../multi/">The multi-commodity flow problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/factory_schedule_demand.txt b/previews/PR3525/tutorials/linear/factory_schedule_demand.txt
new file mode 100644
index 00000000000..daf4eaf2eb1
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/factory_schedule_demand.txt
@@ -0,0 +1,13 @@
+month demand
+1     120000
+2     100000
+3     130000
+4     130000
+5     140000
+6     130000
+7     150000
+8     170000
+9     200000
+10    190000
+11    140000
+12    100000
diff --git a/previews/PR3525/tutorials/linear/factory_schedule_factories.txt b/previews/PR3525/tutorials/linear/factory_schedule_factories.txt
new file mode 100644
index 00000000000..c525f224301
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/factory_schedule_factories.txt
@@ -0,0 +1,25 @@
+factory month min_production max_production fixed_cost variable_cost
+A       1     20000          100000         500        10
+A       2     20000          110000         500        11
+A       3     20000          120000         500        12
+A       4     20000          145000         500        9
+A       5     20000          160000         500        8
+A       6     20000          140000         500        8
+A       7     20000          155000         500        5
+A       8     20000          200000         500        7
+A       9     20000          210000         500        9
+A       10    20000          197000         500        10
+A       11    20000          80000          500        8
+A       12    20000          150000         500        8
+B       1     20000          50000          600        5
+B       2     20000          55000          600        4
+B       3     20000          60000          600        3
+B       4     20000          100000         600        5
+B       5     0              0              0          0
+B       6     20000          70000          600        6
+B       7     20000          60000          600        4
+B       8     20000          100000         600        6
+B       9     20000          100000         600        8
+B       10    20000          100000         600        11
+B       11    20000          120000         600        10
+B       12    20000          150000         600        12
diff --git a/previews/PR3525/tutorials/linear/finance.jl b/previews/PR3525/tutorials/linear/finance.jl
new file mode 100644
index 00000000000..e58157c3c9b
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/finance.jl
@@ -0,0 +1,146 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Financial modeling problems
+
+# **This tutorial was originally contributed by Arpit Bhatia.**
+
+# Optimization models play an increasingly important role in financial
+# decisions. Many computational finance problems can be solved efficiently using
+# modern optimization techniques.
+#
+# In this tutorial we will discuss two such examples taken from the book
+# [Optimization Methods in Finance](https://doi.org/10.1017/9781107297340).
+
+# This tutorial uses the following packages
+using JuMP
+import HiGHS
+
+# ## Short-term financing
+
+# Corporations routinely face the problem of financing short term cash
+# commitments such as the following:
+
+# **Month**        |Jan    |Feb    |Mar    |Apr    |May    |Jun
+# :-----:          |:-----:|:-----:|:-----:|:-----:|:-----:|:-----:
+# **Net Cash Flow**|-150   |-100   |200    |-200   |50     |300
+
+# Net cash flow requirements are given in thousands of dollars. The company has
+# the following sources of funds:
+
+# - A line of credit of up to \$100K at an interest rate of 1% per month,
+# - In any one of the first three months, it can issue 90-day commercial paper
+#   bearing a total interest of 2% for the 3-month period,
+# - Excess funds can be invested at an interest rate of 0.3% per month.
+
+# Our task is to find out the most economical way to use these 3 sources such
+# that we end up with the most amount of money at the end of June.
+#
+# We model this problem in the following manner:
+
+# We will use the following decision variables:
+# - the amount $u_{i}$ drawn from the line of credit in month $i$
+# - the amount $v_{i}$ of commercial paper issued in month $i$
+# - the excess funds $w_{i}$ in month $i$
+
+# Here we have three types of constraints:
+# 1. for every month, cash inflow = cash outflow for each month
+# 2. upper bounds on $u_{i}$
+# 3. nonnegativity of the decision variables $u_{i}$, $v_{i}$ and $w_{i}$.
+
+# Our objective will be to simply maximize the company's wealth in June, which
+# say we represent with the variable $m$.
+
+financing = Model(HiGHS.Optimizer)
+
+@variables(financing, begin
+    0 <= u[1:5] <= 100
+    0 <= v[1:3]
+    0 <= w[1:5]
+    m
+end)
+
+@objective(financing, Max, m)
+
+@constraints(
+    financing,
+    begin
+        u[1] + v[1] - w[1] == 150 # January
+        u[2] + v[2] - w[2] - 1.01u[1] + 1.003w[1] == 100 # February
+        u[3] + v[3] - w[3] - 1.01u[2] + 1.003w[2] == -200 # March
+        u[4] - w[4] - 1.02v[1] - 1.01u[3] + 1.003w[3] == 200 # April
+        u[5] - w[5] - 1.02v[2] - 1.01u[4] + 1.003w[4] == -50 # May
+        -m - 1.02v[3] - 1.01u[5] + 1.003w[5] == -300 # June
+    end
+)
+
+optimize!(financing)
+
+objective_value(financing)
+
+# ## Combinatorial auctions
+
+# In many auctions, the value that a bidder has for a set of items may not be
+# the sum of the values that he has for individual items.
+#
+# Examples are equity trading, electricity markets, pollution right auctions and
+# auctions for airport landing slots.
+#
+# To take this into account, combinatorial auctions allow the bidders to submit
+# bids on combinations of items.
+
+# Let $M=\{1,2, \ldots, m\}$ be the set of items that the auctioneer has to
+# sell. A bid is a pair $B_{j}=\left(S_{j}, p_{j}\right)$ where
+# $S_{j} \subseteq M$ is a nonempty set of items and $p_{j}$ is the price offer
+# for this set.
+#
+# Suppose that the auctioneer has received $n$ bids
+# $B_{1}, B_{2}, \ldots, B_{n}.$ The goal of this problem is to help an
+# auctioneer determine the winners in order to maximize his revenue.
+
+# We model this problem by taking a decision variable $y_{j}$ for every bid. We
+# add a constraint that each item $i$ is sold at most once. This gives us the
+# following model:
+
+# ```math
+# \begin{aligned}
+# \max && \sum_{i=1}^{n} p_{j} y_{j} \\
+# \text { s.t. }  && \sum_{j : i \in S_{j}} y_{j} \leq 1 && \forall i=\{1,2 \ldots m\} \\
+# && y_{j} \in\{0,1\} && \forall j \in\{1,2 \ldots n\}
+# \end{aligned}
+# ```
+
+bid_values = [6 3 12 12 8 16]
+bid_items = [[1], [2], [3 4], [1 3], [2 4], [1 3 4]]
+
+auction = Model(HiGHS.Optimizer)
+@variable(auction, y[1:6], Bin)
+@objective(auction, Max, sum(y' .* bid_values))
+for i in 1:6
+    @constraint(auction, sum(y[j] for j in 1:6 if i in bid_items[j]) <= 1)
+end
+
+optimize!(auction)
+
+objective_value(auction)
+
+#-
+
+value.(y)
diff --git a/previews/PR3525/tutorials/linear/finance/index.html b/previews/PR3525/tutorials/linear/finance/index.html
new file mode 100644
index 00000000000..7c7976d92b7
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/finance/index.html
@@ -0,0 +1,54 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Financial modeling problems · JuMP</title><meta name="title" content="Financial modeling problems · JuMP"/><meta property="og:title" content="Financial modeling problems · JuMP"/><meta property="twitter:title" content="Financial modeling problems · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li class="is-active"><a class="tocitem" href>Financial modeling problems</a><ul class="internal"><li><a class="tocitem" href="#Short-term-financing"><span>Short-term financing</span></a></li><li><a class="tocitem" href="#Combinatorial-auctions"><span>Combinatorial auctions</span></a></li></ul></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Financial modeling problems</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Financial modeling problems</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/finance.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Financial-modeling-problems"><a class="docs-heading-anchor" href="#Financial-modeling-problems">Financial modeling problems</a><a id="Financial-modeling-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Financial-modeling-problems" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../finance.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Arpit Bhatia.</strong></p><p>Optimization models play an increasingly important role in financial decisions. Many computational finance problems can be solved efficiently using modern optimization techniques.</p><p>In this tutorial we will discuss two such examples taken from the book <a href="https://doi.org/10.1017/9781107297340">Optimization Methods in Finance</a>.</p><p>This tutorial uses the following packages</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS</code></pre><h2 id="Short-term-financing"><a class="docs-heading-anchor" href="#Short-term-financing">Short-term financing</a><a id="Short-term-financing-1"></a><a class="docs-heading-anchor-permalink" href="#Short-term-financing" title="Permalink"></a></h2><p>Corporations routinely face the problem of financing short term cash commitments such as the following:</p><table><tr><th style="text-align: center"><strong>Month</strong></th><th style="text-align: center">Jan</th><th style="text-align: center">Feb</th><th style="text-align: center">Mar</th><th style="text-align: center">Apr</th><th style="text-align: center">May</th><th style="text-align: center">Jun</th></tr><tr><td style="text-align: center"><strong>Net Cash Flow</strong></td><td style="text-align: center">-150</td><td style="text-align: center">-100</td><td style="text-align: center">200</td><td style="text-align: center">-200</td><td style="text-align: center">50</td><td style="text-align: center">300</td></tr></table><p>Net cash flow requirements are given in thousands of dollars. The company has the following sources of funds:</p><ul><li>A line of credit of up to <span>$</span>100K at an interest rate of 1% per month,</li><li>In any one of the first three months, it can issue 90-day commercial paper bearing a total interest of 2% for the 3-month period,</li><li>Excess funds can be invested at an interest rate of 0.3% per month.</li></ul><p>Our task is to find out the most economical way to use these 3 sources such that we end up with the most amount of money at the end of June.</p><p>We model this problem in the following manner:</p><p>We will use the following decision variables:</p><ul><li>the amount <span>$u_{i}$</span> drawn from the line of credit in month <span>$i$</span></li><li>the amount <span>$v_{i}$</span> of commercial paper issued in month <span>$i$</span></li><li>the excess funds <span>$w_{i}$</span> in month <span>$i$</span></li></ul><p>Here we have three types of constraints:</p><ol><li>for every month, cash inflow = cash outflow for each month</li><li>upper bounds on <span>$u_{i}$</span></li><li>nonnegativity of the decision variables <span>$u_{i}$</span>, <span>$v_{i}$</span> and <span>$w_{i}$</span>.</li></ol><p>Our objective will be to simply maximize the company&#39;s wealth in June, which say we represent with the variable <span>$m$</span>.</p><pre><code class="language-julia hljs">financing = Model(HiGHS.Optimizer)
+
+@variables(financing, begin
+    0 &lt;= u[1:5] &lt;= 100
+    0 &lt;= v[1:3]
+    0 &lt;= w[1:5]
+    m
+end)
+
+@objective(financing, Max, m)
+
+@constraints(
+    financing,
+    begin
+        u[1] + v[1] - w[1] == 150 # January
+        u[2] + v[2] - w[2] - 1.01u[1] + 1.003w[1] == 100 # February
+        u[3] + v[3] - w[3] - 1.01u[2] + 1.003w[2] == -200 # March
+        u[4] - w[4] - 1.02v[1] - 1.01u[3] + 1.003w[3] == 200 # April
+        u[5] - w[5] - 1.02v[2] - 1.01u[4] + 1.003w[4] == -50 # May
+        -m - 1.02v[3] - 1.01u[5] + 1.003w[5] == -300 # June
+    end
+)
+
+optimize!(financing)
+
+objective_value(financing)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">92.49694915254233</code></pre><h2 id="Combinatorial-auctions"><a class="docs-heading-anchor" href="#Combinatorial-auctions">Combinatorial auctions</a><a id="Combinatorial-auctions-1"></a><a class="docs-heading-anchor-permalink" href="#Combinatorial-auctions" title="Permalink"></a></h2><p>In many auctions, the value that a bidder has for a set of items may not be the sum of the values that he has for individual items.</p><p>Examples are equity trading, electricity markets, pollution right auctions and auctions for airport landing slots.</p><p>To take this into account, combinatorial auctions allow the bidders to submit bids on combinations of items.</p><p>Let <span>$M=\{1,2, \ldots, m\}$</span> be the set of items that the auctioneer has to sell. A bid is a pair <span>$B_{j}=\left(S_{j}, p_{j}\right)$</span> where <span>$S_{j} \subseteq M$</span> is a nonempty set of items and <span>$p_{j}$</span> is the price offer for this set.</p><p>Suppose that the auctioneer has received <span>$n$</span> bids <span>$B_{1}, B_{2}, \ldots, B_{n}.$</span> The goal of this problem is to help an auctioneer determine the winners in order to maximize his revenue.</p><p>We model this problem by taking a decision variable <span>$y_{j}$</span> for every bid. We add a constraint that each item <span>$i$</span> is sold at most once. This gives us the following model:</p><p class="math-container">\[\begin{aligned}
+\max &amp;&amp; \sum_{i=1}^{n} p_{j} y_{j} \\
+\text { s.t. }  &amp;&amp; \sum_{j : i \in S_{j}} y_{j} \leq 1 &amp;&amp; \forall i=\{1,2 \ldots m\} \\
+&amp;&amp; y_{j} \in\{0,1\} &amp;&amp; \forall j \in\{1,2 \ldots n\}
+\end{aligned}\]</p><pre><code class="language-julia hljs">bid_values = [6 3 12 12 8 16]
+bid_items = [[1], [2], [3 4], [1 3], [2 4], [1 3 4]]
+
+auction = Model(HiGHS.Optimizer)
+@variable(auction, y[1:6], Bin)
+@objective(auction, Max, sum(y&#39; .* bid_values))
+for i in 1:6
+    @constraint(auction, sum(y[j] for j in 1:6 if i in bid_items[j]) &lt;= 1)
+end
+
+optimize!(auction)
+
+objective_value(auction)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">21.0</code></pre><pre><code class="language-julia hljs">value.(y)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">6-element Vector{Float64}:
+  1.0
+  1.0
+  1.0
+ -0.0
+ -0.0
+  0.0</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../facility_location/">« The facility location problem</a><a class="docs-footer-nextpage" href="../geographic_clustering/">Geographical clustering »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/geographic_clustering.jl b/previews/PR3525/tutorials/linear/geographic_clustering.jl
new file mode 100644
index 00000000000..448ae2e53df
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/geographic_clustering.jl
@@ -0,0 +1,176 @@
+# Copyright (c) 2019 Matthew Help, Mathieu Tanneau and contributors              #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Geographical clustering
+
+# **This tutorial was originally contributed by Matthew Helm and Mathieu Tanneau.**
+
+# The goal of this exercise is to cluster $n$ cities into $k$ groups, minimizing
+# the total pairwise distance between cities *and* ensuring that the variance in
+# the total populations of each group is relatively small.
+
+# This tutorial uses the following packages:
+using JuMP
+import DataFrames
+import HiGHS
+import LinearAlgebra
+
+# For this example, we'll use the 20 most populous cities in the United States.
+cities = DataFrames.DataFrame(
+    Union{String,Float64}[
+        "New York, NY" 8.405 40.7127 -74.0059
+        "Los Angeles, CA" 3.884 34.0522 -118.2436
+        "Chicago, IL" 2.718 41.8781 -87.6297
+        "Houston, TX" 2.195 29.7604 -95.3698
+        "Philadelphia, PA" 1.553 39.9525 -75.1652
+        "Phoenix, AZ" 1.513 33.4483 -112.0740
+        "San Antonio, TX" 1.409 29.4241 -98.4936
+        "San Diego, CA" 1.355 32.7157 -117.1610
+        "Dallas, TX" 1.257 32.7766 -96.7969
+        "San Jose, CA" 0.998 37.3382 -121.8863
+        "Austin, TX" 0.885 30.2671 -97.7430
+        "Indianapolis, IN" 0.843 39.7684 -86.1580
+        "Jacksonville, FL" 0.842 30.3321 -81.6556
+        "San Francisco, CA" 0.837 37.7749 -122.4194
+        "Columbus, OH" 0.822 39.9611 -82.9987
+        "Charlotte, NC" 0.792 35.2270 -80.8431
+        "Fort Worth, TX" 0.792 32.7554 -97.3307
+        "Detroit, MI" 0.688 42.3314 -83.0457
+        "El Paso, TX" 0.674 31.7775 -106.4424
+        "Memphis, TN" 0.653 35.1495 -90.0489
+    ],
+    ["city", "population", "lat", "lon"],
+)
+
+# ### Model Specifics
+
+# We will cluster these 20 cities into 3 different groups and we will assume
+# that the ideal or target population $P$ for a group is simply the total
+# population of the 20 cities divided by 3:
+
+n = size(cities, 1)
+k = 3
+P = sum(cities.population) / k
+
+# ### Obtaining the distances between each city
+
+# Let's compute the pairwise Haversine distance between each of the cities in
+# our data set and store the result in a variable we'll call `dm`:
+
+"""
+    haversine(lat1, long1, lat2, long2, r = 6372.8)
+
+Compute the haversine distance between two points on a sphere of radius `r`,
+where the points are given by the latitude/longitude pairs `lat1/long1` and
+`lat2/long2` (in degrees).
+"""
+function haversine(lat1, long1, lat2, long2, r = 6372.8)
+    lat1, long1 = deg2rad(lat1), deg2rad(long1)
+    lat2, long2 = deg2rad(lat2), deg2rad(long2)
+    hav(a, b) = sin((b - a) / 2)^2
+    inner_term = hav(lat1, lat2) + cos(lat1) * cos(lat2) * hav(long1, long2)
+    d = 2 * r * asin(sqrt(inner_term))
+    ## Round distance to nearest kilometer.
+    return round(Int, d)
+end
+
+# Our distance matrix is symmetric so we'll convert it to a `LowerTriangular`
+# matrix so that we can better interpret the objective value of our model:
+
+dm = LinearAlgebra.LowerTriangular([
+    haversine(cities.lat[i], cities.lon[i], cities.lat[j], cities.lon[j])
+    for i in 1:n, j in 1:n
+])
+
+# ### Build the model
+
+# Now that we have the basics taken  care of, we can set up our model, create
+# decision variables, add constraints, and then solve.
+
+# First, we'll set up a model that leverages the Cbc solver. Next, we'll set up
+# a binary variable $x_{i,k}$ that takes the value $1$ if city $i$ is in group
+# $k$ and $0$ otherwise. Each city must be in a group, so we'll add the
+# constraint $\sum_k x_{i,k} = 1$ for every $i$.
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:n, 1:k], Bin)
+@constraint(model, [i = 1:n], sum(x[i, :]) == 1);
+## To reduce symmetry, we fix the first city to belong to the first group.
+fix(x[1, 1], 1; force = true)
+
+# The total population of a group $k$ is $Q_k = \sum_ix_{i,k}q_i$ where $q_i$ is
+# simply the $i$-th value from the `population` column in our `cities` DataFrame.
+# Let's add constraints so that $\alpha \leq (Q_k - P) \leq \beta$. We'll set
+# $\alpha$ equal to $-3$ million and $\beta$ equal to $3$. By adjusting
+# these thresholds you'll find that there is a tradeoff between having
+# relatively even populations between groups and having geographically close
+# cities within each group. In other words, the larger the absolute values of
+# $\alpha$ and $\beta$, the closer together the cities in a group will be but
+# the variance between the group populations will be higher.
+
+@variable(model, -3 <= population_diff[1:k] <= 3)
+@constraint(model, population_diff .== x' * cities.population .- P)
+
+# Now we need to add one last binary variable $z_{i,j}$ to our model that we'll
+# use to compute the total distance between the cities in our groups, defined as
+# $\sum_{i,j}d_{i,j}z_{i,j}$. Variable $z_{i,j}$ will equal $1$ if cities $i$
+# and $j$ are in the same group, and $0$ if they are not in the same group.
+
+# To ensure that $z_{i,j} = 1$ if and only if cities $i$ and $j$ are in the same
+# group, we add the constraints $z_{i,j} \geq x_{i,k} + x_{j,k} - 1$ for every
+# pair $i,j$ and every $k$:
+
+@variable(model, z[i = 1:n, j = 1:i], Bin)
+for k in 1:k, i in 1:n, j in 1:i
+    @constraint(model, z[i, j] >= x[i, k] + x[j, k] - 1)
+end
+
+# We can now add an objective to our model which will simply be to minimize the
+# dot product of $z$ and our distance matrix, `dm`.
+
+@objective(model, Min, sum(dm[i, j] * z[i, j] for i in 1:n, j in 1:i));
+
+# We can then call `optimize!` and review the results.
+
+optimize!(model)
+
+# ### Reviewing the Results
+
+# Now that we have results, we can add a column to our `cities` DataFrame for
+# the group and then loop through our $x$ variable to assign each city to its
+# group. Once we have that, we can look at the total population for each group
+# and also look at the cities in each group to verify that they are grouped by
+# geographic proximity.
+
+cities.group = zeros(n)
+
+for i in 1:n, j in 1:k
+    if round(Int, value(x[i, j])) == 1
+        cities.group[i] = j
+    end
+end
+
+for group in DataFrames.groupby(cities, :group)
+    @show group
+    println("")
+    @show sum(group.population)
+    println("")
+end
diff --git a/previews/PR3525/tutorials/linear/geographic_clustering/index.html b/previews/PR3525/tutorials/linear/geographic_clustering/index.html
new file mode 100644
index 00000000000..c886a42fa90
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/geographic_clustering/index.html
@@ -0,0 +1,137 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Geographical clustering · JuMP</title><meta name="title" content="Geographical clustering · JuMP"/><meta property="og:title" content="Geographical clustering · JuMP"/><meta property="twitter:title" content="Geographical clustering · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li class="is-active"><a class="tocitem" href>Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Geographical clustering</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Geographical clustering</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/geographic_clustering.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Geographical-clustering"><a class="docs-heading-anchor" href="#Geographical-clustering">Geographical clustering</a><a id="Geographical-clustering-1"></a><a class="docs-heading-anchor-permalink" href="#Geographical-clustering" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../geographic_clustering.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Matthew Helm and Mathieu Tanneau.</strong></p><p>The goal of this exercise is to cluster <span>$n$</span> cities into <span>$k$</span> groups, minimizing the total pairwise distance between cities <em>and</em> ensuring that the variance in the total populations of each group is relatively small.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import DataFrames
+import HiGHS
+import LinearAlgebra</code></pre><p>For this example, we&#39;ll use the 20 most populous cities in the United States.</p><pre><code class="language-julia hljs">cities = DataFrames.DataFrame(
+    Union{String,Float64}[
+        &quot;New York, NY&quot; 8.405 40.7127 -74.0059
+        &quot;Los Angeles, CA&quot; 3.884 34.0522 -118.2436
+        &quot;Chicago, IL&quot; 2.718 41.8781 -87.6297
+        &quot;Houston, TX&quot; 2.195 29.7604 -95.3698
+        &quot;Philadelphia, PA&quot; 1.553 39.9525 -75.1652
+        &quot;Phoenix, AZ&quot; 1.513 33.4483 -112.0740
+        &quot;San Antonio, TX&quot; 1.409 29.4241 -98.4936
+        &quot;San Diego, CA&quot; 1.355 32.7157 -117.1610
+        &quot;Dallas, TX&quot; 1.257 32.7766 -96.7969
+        &quot;San Jose, CA&quot; 0.998 37.3382 -121.8863
+        &quot;Austin, TX&quot; 0.885 30.2671 -97.7430
+        &quot;Indianapolis, IN&quot; 0.843 39.7684 -86.1580
+        &quot;Jacksonville, FL&quot; 0.842 30.3321 -81.6556
+        &quot;San Francisco, CA&quot; 0.837 37.7749 -122.4194
+        &quot;Columbus, OH&quot; 0.822 39.9611 -82.9987
+        &quot;Charlotte, NC&quot; 0.792 35.2270 -80.8431
+        &quot;Fort Worth, TX&quot; 0.792 32.7554 -97.3307
+        &quot;Detroit, MI&quot; 0.688 42.3314 -83.0457
+        &quot;El Paso, TX&quot; 0.674 31.7775 -106.4424
+        &quot;Memphis, TN&quot; 0.653 35.1495 -90.0489
+    ],
+    [&quot;city&quot;, &quot;population&quot;, &quot;lat&quot;, &quot;lon&quot;],
+)</code></pre><div><div style = "float: left;"><span>20×4 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">city</th><th style = "text-align: left;">population</th><th style = "text-align: left;">lat</th><th style = "text-align: left;">lon</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Union{Float64, String}" style = "text-align: left;">Union…</th><th title = "Union{Float64, String}" style = "text-align: left;">Union…</th><th title = "Union{Float64, String}" style = "text-align: left;">Union…</th><th title = "Union{Float64, String}" style = "text-align: left;">Union…</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">New York, NY</td><td style = "text-align: left;">8.405</td><td style = "text-align: left;">40.7127</td><td style = "text-align: left;">-74.0059</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">Los Angeles, CA</td><td style = "text-align: left;">3.884</td><td style = "text-align: left;">34.0522</td><td style = "text-align: left;">-118.244</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">Chicago, IL</td><td style = "text-align: left;">2.718</td><td style = "text-align: left;">41.8781</td><td style = "text-align: left;">-87.6297</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">Houston, TX</td><td style = "text-align: left;">2.195</td><td style = "text-align: left;">29.7604</td><td style = "text-align: left;">-95.3698</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">Philadelphia, PA</td><td style = "text-align: left;">1.553</td><td style = "text-align: left;">39.9525</td><td style = "text-align: left;">-75.1652</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">Phoenix, AZ</td><td style = "text-align: left;">1.513</td><td style = "text-align: left;">33.4483</td><td style = "text-align: left;">-112.074</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">San Antonio, TX</td><td style = "text-align: left;">1.409</td><td style = "text-align: left;">29.4241</td><td style = "text-align: left;">-98.4936</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">San Diego, CA</td><td style = "text-align: left;">1.355</td><td style = "text-align: left;">32.7157</td><td style = "text-align: left;">-117.161</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">Dallas, TX</td><td style = "text-align: left;">1.257</td><td style = "text-align: left;">32.7766</td><td style = "text-align: left;">-96.7969</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">San Jose, CA</td><td style = "text-align: left;">0.998</td><td style = "text-align: left;">37.3382</td><td style = "text-align: left;">-121.886</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">Austin, TX</td><td style = "text-align: left;">0.885</td><td style = "text-align: left;">30.2671</td><td style = "text-align: left;">-97.743</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">Indianapolis, IN</td><td style = "text-align: left;">0.843</td><td style = "text-align: left;">39.7684</td><td style = "text-align: left;">-86.158</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">Jacksonville, FL</td><td style = "text-align: left;">0.842</td><td style = "text-align: left;">30.3321</td><td style = "text-align: left;">-81.6556</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">San Francisco, CA</td><td style = "text-align: left;">0.837</td><td style = "text-align: left;">37.7749</td><td style = "text-align: left;">-122.419</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">Columbus, OH</td><td style = "text-align: left;">0.822</td><td style = "text-align: left;">39.9611</td><td style = "text-align: left;">-82.9987</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">Charlotte, NC</td><td style = "text-align: left;">0.792</td><td style = "text-align: left;">35.227</td><td style = "text-align: left;">-80.8431</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">17</td><td style = "text-align: left;">Fort Worth, TX</td><td style = "text-align: left;">0.792</td><td style = "text-align: left;">32.7554</td><td style = "text-align: left;">-97.3307</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">18</td><td style = "text-align: left;">Detroit, MI</td><td style = "text-align: left;">0.688</td><td style = "text-align: left;">42.3314</td><td style = "text-align: left;">-83.0457</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">19</td><td style = "text-align: left;">El Paso, TX</td><td style = "text-align: left;">0.674</td><td style = "text-align: left;">31.7775</td><td style = "text-align: left;">-106.442</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">20</td><td style = "text-align: left;">Memphis, TN</td><td style = "text-align: left;">0.653</td><td style = "text-align: left;">35.1495</td><td style = "text-align: left;">-90.0489</td></tr></tbody></table></div><h3 id="Model-Specifics"><a class="docs-heading-anchor" href="#Model-Specifics">Model Specifics</a><a id="Model-Specifics-1"></a><a class="docs-heading-anchor-permalink" href="#Model-Specifics" title="Permalink"></a></h3><p>We will cluster these 20 cities into 3 different groups and we will assume that the ideal or target population <span>$P$</span> for a group is simply the total population of the 20 cities divided by 3:</p><pre><code class="language-julia hljs">n = size(cities, 1)
+k = 3
+P = sum(cities.population) / k</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">11.038333333333334</code></pre><h3 id="Obtaining-the-distances-between-each-city"><a class="docs-heading-anchor" href="#Obtaining-the-distances-between-each-city">Obtaining the distances between each city</a><a id="Obtaining-the-distances-between-each-city-1"></a><a class="docs-heading-anchor-permalink" href="#Obtaining-the-distances-between-each-city" title="Permalink"></a></h3><p>Let&#39;s compute the pairwise Haversine distance between each of the cities in our data set and store the result in a variable we&#39;ll call <code>dm</code>:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    haversine(lat1, long1, lat2, long2, r = 6372.8)
+
+Compute the haversine distance between two points on a sphere of radius `r`,
+where the points are given by the latitude/longitude pairs `lat1/long1` and
+`lat2/long2` (in degrees).
+&quot;&quot;&quot;
+function haversine(lat1, long1, lat2, long2, r = 6372.8)
+    lat1, long1 = deg2rad(lat1), deg2rad(long1)
+    lat2, long2 = deg2rad(lat2), deg2rad(long2)
+    hav(a, b) = sin((b - a) / 2)^2
+    inner_term = hav(lat1, lat2) + cos(lat1) * cos(lat2) * hav(long1, long2)
+    d = 2 * r * asin(sqrt(inner_term))
+    # Round distance to nearest kilometer.
+    return round(Int, d)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Main.haversine</code></pre><p>Our distance matrix is symmetric so we&#39;ll convert it to a <code>LowerTriangular</code> matrix so that we can better interpret the objective value of our model:</p><pre><code class="language-julia hljs">dm = LinearAlgebra.LowerTriangular([
+    haversine(cities.lat[i], cities.lon[i], cities.lat[j], cities.lon[j])
+    for i in 1:n, j in 1:n
+])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20×20 LinearAlgebra.LowerTriangular{Int64, Matrix{Int64}}:
+    0     ⋅     ⋅     ⋅     ⋅     ⋅     ⋅  …     ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 3937     0     ⋅     ⋅     ⋅     ⋅     ⋅        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 1145  2805     0     ⋅     ⋅     ⋅     ⋅        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 2282  2207  1516     0     ⋅     ⋅     ⋅        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+  130  3845  1068  2157     0     ⋅     ⋅        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 3445   574  2337  1633  3345     0     ⋅  …     ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 2546  1934  1695   304  2423  1363     0        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 3908   179  2787  2094  3812   481  1813        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 2206  1993  1295   362  2089  1424   406        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 4103   492  2958  2588  4023   989  2336        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 2432  1972  1577   235  2310  1398   118  …     ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 1036  2907   265  1394   938  2409  1609        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 1345  3450  1391  1321  1221  2883  1626        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+ 4130   559  2986  2644  4052  1051  2394        ⋅     ⋅     ⋅     ⋅     ⋅  ⋅
+  767  3177   444  1598   668  2679  1834        0     ⋅     ⋅     ⋅     ⋅  ⋅
+  855  3405   946  1490   725  2863  1777  …   560     0     ⋅     ⋅     ⋅  ⋅
+ 2251  1944  1327   382  2134  1375   387     1511  1543     0     ⋅     ⋅  ⋅
+  774  3186   382  1780   711  2716  1994      264   813  1646     0     ⋅  ⋅
+ 3054  1130  2010  1081  2945   559   804     2292  2398   864  2374     0  ⋅
+ 1534  2576   777   780  1415  2028  1017      820   837   722  1003  1565  0</code></pre><h3 id="Build-the-model"><a class="docs-heading-anchor" href="#Build-the-model">Build the model</a><a id="Build-the-model-1"></a><a class="docs-heading-anchor-permalink" href="#Build-the-model" title="Permalink"></a></h3><p>Now that we have the basics taken  care of, we can set up our model, create decision variables, add constraints, and then solve.</p><p>First, we&#39;ll set up a model that leverages the Cbc solver. Next, we&#39;ll set up a binary variable <span>$x_{i,k}$</span> that takes the value <span>$1$</span> if city <span>$i$</span> is in group <span>$k$</span> and <span>$0$</span> otherwise. Each city must be in a group, so we&#39;ll add the constraint <span>$\sum_k x_{i,k} = 1$</span> for every <span>$i$</span>.</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[1:n, 1:k], Bin)
+@constraint(model, [i = 1:n], sum(x[i, :]) == 1);
+# To reduce symmetry, we fix the first city to belong to the first group.
+fix(x[1, 1], 1; force = true)</code></pre><p>The total population of a group <span>$k$</span> is <span>$Q_k = \sum_ix_{i,k}q_i$</span> where <span>$q_i$</span> is simply the <span>$i$</span>-th value from the <code>population</code> column in our <code>cities</code> DataFrame. Let&#39;s add constraints so that <span>$\alpha \leq (Q_k - P) \leq \beta$</span>. We&#39;ll set <span>$\alpha$</span> equal to <span>$-3$</span> million and <span>$\beta$</span> equal to <span>$3$</span>. By adjusting these thresholds you&#39;ll find that there is a tradeoff between having relatively even populations between groups and having geographically close cities within each group. In other words, the larger the absolute values of <span>$\alpha$</span> and <span>$\beta$</span>, the closer together the cities in a group will be but the variance between the group populations will be higher.</p><pre><code class="language-julia hljs">@variable(model, -3 &lt;= population_diff[1:k] &lt;= 3)
+@constraint(model, population_diff .== x&#39; * cities.population .- P)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{ConstraintRef{Model, MathOptInterface.ConstraintIndex{MathOptInterface.ScalarAffineFunction{Float64}, MathOptInterface.EqualTo{Float64}}, ScalarShape}}:
+ -8.405 x[1,1] - 3.884 x[2,1] - 2.718 x[3,1] - 2.195 x[4,1] - 1.553 x[5,1] - 1.513 x[6,1] - 1.409 x[7,1] - 1.355 x[8,1] - 1.257 x[9,1] - 0.998 x[10,1] - 0.885 x[11,1] - 0.843 x[12,1] - 0.842 x[13,1] - 0.837 x[14,1] - 0.822 x[15,1] - 0.792 x[16,1] - 0.792 x[17,1] - 0.688 x[18,1] - 0.674 x[19,1] - 0.653 x[20,1] + population_diff[1] = -11.038333333333334
+ -8.405 x[1,2] - 3.884 x[2,2] - 2.718 x[3,2] - 2.195 x[4,2] - 1.553 x[5,2] - 1.513 x[6,2] - 1.409 x[7,2] - 1.355 x[8,2] - 1.257 x[9,2] - 0.998 x[10,2] - 0.885 x[11,2] - 0.843 x[12,2] - 0.842 x[13,2] - 0.837 x[14,2] - 0.822 x[15,2] - 0.792 x[16,2] - 0.792 x[17,2] - 0.688 x[18,2] - 0.674 x[19,2] - 0.653 x[20,2] + population_diff[2] = -11.038333333333334
+ -8.405 x[1,3] - 3.884 x[2,3] - 2.718 x[3,3] - 2.195 x[4,3] - 1.553 x[5,3] - 1.513 x[6,3] - 1.409 x[7,3] - 1.355 x[8,3] - 1.257 x[9,3] - 0.998 x[10,3] - 0.885 x[11,3] - 0.843 x[12,3] - 0.842 x[13,3] - 0.837 x[14,3] - 0.822 x[15,3] - 0.792 x[16,3] - 0.792 x[17,3] - 0.688 x[18,3] - 0.674 x[19,3] - 0.653 x[20,3] + population_diff[3] = -11.038333333333334</code></pre><p>Now we need to add one last binary variable <span>$z_{i,j}$</span> to our model that we&#39;ll use to compute the total distance between the cities in our groups, defined as <span>$\sum_{i,j}d_{i,j}z_{i,j}$</span>. Variable <span>$z_{i,j}$</span> will equal <span>$1$</span> if cities <span>$i$</span> and <span>$j$</span> are in the same group, and <span>$0$</span> if they are not in the same group.</p><p>To ensure that <span>$z_{i,j} = 1$</span> if and only if cities <span>$i$</span> and <span>$j$</span> are in the same group, we add the constraints <span>$z_{i,j} \geq x_{i,k} + x_{j,k} - 1$</span> for every pair <span>$i,j$</span> and every <span>$k$</span>:</p><pre><code class="language-julia hljs">@variable(model, z[i = 1:n, j = 1:i], Bin)
+for k in 1:k, i in 1:n, j in 1:i
+    @constraint(model, z[i, j] &gt;= x[i, k] + x[j, k] - 1)
+end</code></pre><p>We can now add an objective to our model which will simply be to minimize the dot product of <span>$z$</span> and our distance matrix, <code>dm</code>.</p><pre><code class="language-julia hljs">@objective(model, Min, sum(dm[i, j] * z[i, j] for i in 1:n, j in 1:i));</code></pre><p>We can then call <code>optimize!</code> and review the results.</p><pre><code class="language-julia hljs">optimize!(model)</code></pre><h3 id="Reviewing-the-Results"><a class="docs-heading-anchor" href="#Reviewing-the-Results">Reviewing the Results</a><a id="Reviewing-the-Results-1"></a><a class="docs-heading-anchor-permalink" href="#Reviewing-the-Results" title="Permalink"></a></h3><p>Now that we have results, we can add a column to our <code>cities</code> DataFrame for the group and then loop through our <span>$x$</span> variable to assign each city to its group. Once we have that, we can look at the total population for each group and also look at the cities in each group to verify that they are grouped by geographic proximity.</p><pre><code class="language-julia hljs">cities.group = zeros(n)
+
+for i in 1:n, j in 1:k
+    if round(Int, value(x[i, j])) == 1
+        cities.group[i] = j
+    end
+end
+
+for group in DataFrames.groupby(cities, :group)
+    @show group
+    println(&quot;&quot;)
+    @show sum(group.population)
+    println(&quot;&quot;)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">group = 7×5 SubDataFrame
+ Row │ city              population  lat      lon       group
+     │ Union…            Union…      Union…   Union…    Float64
+─────┼──────────────────────────────────────────────────────────
+   1 │ New York, NY      8.405       40.7127  -74.0059      1.0
+   2 │ Philadelphia, PA  1.553       39.9525  -75.1652      1.0
+   3 │ Indianapolis, IN  0.843       39.7684  -86.158       1.0
+   4 │ Jacksonville, FL  0.842       30.3321  -81.6556      1.0
+   5 │ Columbus, OH      0.822       39.9611  -82.9987      1.0
+   6 │ Charlotte, NC     0.792       35.227   -80.8431      1.0
+   7 │ Detroit, MI       0.688       42.3314  -83.0457      1.0
+
+sum(group.population) = 13.944999999999999
+
+group = 7×5 SubDataFrame
+ Row │ city             population  lat      lon       group
+     │ Union…           Union…      Union…   Union…    Float64
+─────┼─────────────────────────────────────────────────────────
+   1 │ Chicago, IL      2.718       41.8781  -87.6297      2.0
+   2 │ Houston, TX      2.195       29.7604  -95.3698      2.0
+   3 │ San Antonio, TX  1.409       29.4241  -98.4936      2.0
+   4 │ Dallas, TX       1.257       32.7766  -96.7969      2.0
+   5 │ Austin, TX       0.885       30.2671  -97.743       2.0
+   6 │ Fort Worth, TX   0.792       32.7554  -97.3307      2.0
+   7 │ Memphis, TN      0.653       35.1495  -90.0489      2.0
+
+sum(group.population) = 9.909
+
+group = 6×5 SubDataFrame
+ Row │ city               population  lat      lon       group
+     │ Union…             Union…      Union…   Union…    Float64
+─────┼───────────────────────────────────────────────────────────
+   1 │ Los Angeles, CA    3.884       34.0522  -118.244      3.0
+   2 │ Phoenix, AZ        1.513       33.4483  -112.074      3.0
+   3 │ San Diego, CA      1.355       32.7157  -117.161      3.0
+   4 │ San Jose, CA       0.998       37.3382  -121.886      3.0
+   5 │ San Francisco, CA  0.837       37.7749  -122.419      3.0
+   6 │ El Paso, TX        0.674       31.7775  -106.442      3.0
+
+sum(group.population) = 9.261000000000001</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../finance/">« Financial modeling problems</a><a class="docs-footer-nextpage" href="../network_flows/">Network flow problems »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/introduction/index.html b/previews/PR3525/tutorials/linear/introduction/index.html
new file mode 100644
index 00000000000..ef72f7eca47
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/introduction/index.html
@@ -0,0 +1,10 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#How-to-choose-a-solver"><span>How to choose a solver</span></a></li><li><a class="tocitem" href="#How-these-tutorials-are-structured"><span>How these tutorials are structured</span></a></li></ul></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/introduction.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Introduction"><a class="docs-heading-anchor" href="#Introduction">Introduction</a><a id="Introduction-1"></a><a class="docs-heading-anchor-permalink" href="#Introduction" title="Permalink"></a></h1><p><a href="https://en.wikipedia.org/wiki/Linear_programming">Linear programs (LPs)</a> are a fundamental class of optimization problems of the form:</p><p class="math-container">\[\begin{align}
+    \min_{x \in \mathbb{R}^n} &amp; \sum\limits_{i=1}^n c_i x_i \\
+    \;\;\text{s.t.} &amp; l_j \le \sum\limits_{i=1}^n a_{ij} x_i \le u_j &amp; j = 1 \ldots m \\
+    &amp; l_i \le x_i \le u_i &amp; i = 1 \ldots n.
+\end{align}\]</p><p>The most important thing to note is that all terms are of the form <code>coefficient * variable</code>, and that there are no nonlinear terms or multiplications between variables.</p><p>Mixed-integer linear programs (MILPs) are extensions of linear programs in which some (or all) of the decision variables take discrete values.</p><h2 id="How-to-choose-a-solver"><a class="docs-heading-anchor" href="#How-to-choose-a-solver">How to choose a solver</a><a id="How-to-choose-a-solver-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-choose-a-solver" title="Permalink"></a></h2><p>Almost all solvers support linear programs; look for &quot;LP&quot; in the list of <a href="../../../installation/#Supported-solvers">Supported solvers</a>. However, fewer solvers support mixed-integer linear programs. Solvers supporting discrete variables start with &quot;(MI)&quot; in the list of <a href="../../../installation/#Supported-solvers">Supported solvers</a>.</p><h2 id="How-these-tutorials-are-structured"><a class="docs-heading-anchor" href="#How-these-tutorials-are-structured">How these tutorials are structured</a><a id="How-these-tutorials-are-structured-1"></a><a class="docs-heading-anchor-permalink" href="#How-these-tutorials-are-structured" title="Permalink"></a></h2><p>Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.</p><ul><li>The following tutorials are worked examples that present a problem in words, then formulate it in mathematics, and then solve it in JuMP. This usually involves some sort of visualization of the solution. Start here if you are new to JuMP.<ul><li><a href="../diet/#The-diet-problem">The diet problem</a></li><li><a href="../cannery/#The-cannery-problem">The cannery problem</a></li><li><a href="../facility_location/#The-facility-location-problem">The facility location problem</a></li><li><a href="../finance/#Financial-modeling-problems">Financial modeling problems</a></li><li><a href="../network_flows/#Network-flow-problems">Network flow problems</a></li><li><a href="../n-queens/#N-Queens">N-Queens</a></li><li><a href="../sudoku/#Sudoku">Sudoku</a></li></ul></li><li>The <a href="../tips_and_tricks/#linear_tips_and_tricks">Tips and tricks</a> tutorial contains a number of helpful reformulations and tricks you can use when modeling linear programs. Look here if you are stuck trying to formulate a problem as a linear program.</li><li>The <a href="../lp_sensitivity/#Sensitivity-analysis-of-a-linear-program">Sensitivity analysis of a linear program</a> tutorial explains how to create sensitivity reports like those produced by the Excel Solver.</li><li>The <a href="../callbacks/#callbacks_tutorial">Callbacks</a> tutorial explains how to write a variety of solver-independent callbacks. Look here if you want to write a callback.</li><li>The remaining tutorials are less verbose and styled in the form of short code examples. These tutorials have less explanation, but may contain useful code snippets, particularly if they are similar to a problem you are trying to solve.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../getting_started/performance_tips/">« Performance tips</a><a class="docs-footer-nextpage" href="../knapsack/">The knapsack problem example »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/knapsack.jl b/previews/PR3525/tutorials/linear/knapsack.jl
new file mode 100644
index 00000000000..6e85075a29a
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/knapsack.jl
@@ -0,0 +1,154 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The knapsack problem example
+
+# The purpose of this tutorial is to demonstrate how to formulate and solve a
+# simple optimization problem.
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import HiGHS
+
+# ## Formulation
+
+# The [knapsack problem](https://en.wikipedia.org/wiki/Knapsack_problem)
+# is a classical optimization problem: given a set of items and a container with
+# a fixed capacity, choose a subset of items having the greatest combined
+# value that will fit within the container without exceeding the capacity.
+
+# The name of the problem suggests its analogy to packing for a trip,
+# where the baggage weight limit is the capacity and the goal is to pack the
+# most profitable combination of belongings.
+
+# We can formulate the knapsack problem as the integer linear program:
+# ```math
+# \begin{aligned}
+# \max \; & \sum_{i=1}^n c_i x_i      \\
+# s.t. \; & \sum_{i=1}^n w_i x_i \le C, \\
+#         & x_i \in \{0,1\},\quad \forall i=1,\ldots,n,
+# \end{aligned}
+# ```
+# where ``C`` is the capacity, and there is a choice between ``n`` items, with
+# item ``i`` having weight ``w_i``, profit ``c_i``. Decision variable ``x_i`` is
+# equal to 1 if the item is chosen and 0 if not.
+
+# This formulation can be written more compactly as:
+# ```math
+# \begin{aligned}
+# \max \; & c^\top x       \\
+# s.t. \; & w^\top x \le C \\
+#         & x \text{ binary }.
+# \end{aligned}
+# ```
+
+# ## Data
+
+# The data for the problem consists of two vectors (one for the profits and one
+# for the weights) along with a capacity.
+
+# There are five objects:
+
+n = 5;
+
+# For our example, we use a capacity of 10 units:
+
+capacity = 10.0;
+
+# and the profit and cost data:
+
+profit = [5.0, 3.0, 2.0, 7.0, 4.0];
+weight = [2.0, 8.0, 4.0, 2.0, 5.0];
+
+# ## JuMP formulation
+
+# Let's begin constructing the JuMP model for our knapsack problem.
+
+# First, we'll create a `Model` object for holding model elements as we
+# construct each part. We'll also set the solver that will ultimately be called
+# to solve the model, once it's constructed.
+
+model = Model(HiGHS.Optimizer)
+
+# Next we need the decision variables representing which items are chosen:
+
+@variable(model, x[1:n], Bin)
+
+# We now want to constrain those variables so that their combined
+# weight is less than or equal to the given capacity:
+
+@constraint(model, sum(weight[i] * x[i] for i in 1:n) <= capacity)
+
+# Finally, our objective is to maximize the combined profit of the chosen items:
+
+@objective(model, Max, sum(profit[i] * x[i] for i in 1:n))
+
+# Let's print a human-readable description of the model and check that the model
+# looks as expected:
+
+print(model)
+
+# We can now solve the optimization problem and inspect the results.
+
+optimize!(model)
+solution_summary(model)
+
+# The items chosen are
+
+items_chosen = [i for i in 1:n if value(x[i]) > 0.5]
+
+# ## Writing a function
+
+# After working interactively, it is good practice to implement your model in a
+# function.
+
+# The function can be used to ensure that the model is given well-defined input
+# data with validation checks, and that the solution process went as expected.
+
+function solve_knapsack_problem(;
+    profit::Vector{Float64},
+    weight::Vector{Float64},
+    capacity::Float64,
+)
+    n = length(weight)
+    ## The profit and weight vectors must be of equal length.
+    @assert length(profit) == n
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, x[1:n], Bin)
+    @objective(model, Max, profit' * x)
+    @constraint(model, weight' * x <= capacity)
+    optimize!(model)
+    @assert termination_status(model) == OPTIMAL
+    @assert primal_status(model) == FEASIBLE_POINT
+    println("Objective is: ", objective_value(model))
+    println("Solution is:")
+    for i in 1:n
+        print("x[$i] = ", round(Int, value(x[i])))
+        println(", c[$i] / w[$i] = ", profit[i] / weight[i])
+    end
+    chosen_items = [i for i in 1:n if value(x[i]) > 0.5]
+    return return chosen_items
+end
+
+solve_knapsack_problem(; profit = profit, weight = weight, capacity = capacity)
+
+# We observe that the chosen items (1, 4, and 5) have the best
+# profit to weight ratio in this particular example.
+
+# ## Next steps
+
+# Here are some things to try next:
+#
+# * Call the function with different data. What happens as the capacity
+#   increases?
+# * What happens if the profit and weight vectors are different lengths?
+# * Instead of creating a binary variable with `Bin`, we could have written
+#   `@variable(model, 0 <= x[1:n] <= 1, Int)`. Verify that this formulation
+#   finds the same solution. What happens if we are allowed to take more than
+#   one of each item?
diff --git a/previews/PR3525/tutorials/linear/knapsack/index.html b/previews/PR3525/tutorials/linear/knapsack/index.html
new file mode 100644
index 00000000000..a9ac73eeaac
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/knapsack/index.html
@@ -0,0 +1,86 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The knapsack problem example · JuMP</title><meta name="title" content="The knapsack problem example · JuMP"/><meta property="og:title" content="The knapsack problem example · JuMP"/><meta property="twitter:title" content="The knapsack problem example · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li class="is-active"><a class="tocitem" href>The knapsack problem example</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Writing-a-function"><span>Writing a function</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The knapsack problem example</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The knapsack problem example</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/knapsack.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-knapsack-problem-example"><a class="docs-heading-anchor" href="#The-knapsack-problem-example">The knapsack problem example</a><a id="The-knapsack-problem-example-1"></a><a class="docs-heading-anchor-permalink" href="#The-knapsack-problem-example" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../knapsack.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to demonstrate how to formulate and solve a simple optimization problem.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>The <a href="https://en.wikipedia.org/wiki/Knapsack_problem">knapsack problem</a> is a classical optimization problem: given a set of items and a container with a fixed capacity, choose a subset of items having the greatest combined value that will fit within the container without exceeding the capacity.</p><p>The name of the problem suggests its analogy to packing for a trip, where the baggage weight limit is the capacity and the goal is to pack the most profitable combination of belongings.</p><p>We can formulate the knapsack problem as the integer linear program:</p><p class="math-container">\[\begin{aligned}
+\max \; &amp; \sum_{i=1}^n c_i x_i      \\
+s.t. \; &amp; \sum_{i=1}^n w_i x_i \le C, \\
+        &amp; x_i \in \{0,1\},\quad \forall i=1,\ldots,n,
+\end{aligned}\]</p><p>where <span>$C$</span> is the capacity, and there is a choice between <span>$n$</span> items, with item <span>$i$</span> having weight <span>$w_i$</span>, profit <span>$c_i$</span>. Decision variable <span>$x_i$</span> is equal to 1 if the item is chosen and 0 if not.</p><p>This formulation can be written more compactly as:</p><p class="math-container">\[\begin{aligned}
+\max \; &amp; c^\top x       \\
+s.t. \; &amp; w^\top x \le C \\
+        &amp; x \text{ binary }.
+\end{aligned}\]</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>The data for the problem consists of two vectors (one for the profits and one for the weights) along with a capacity.</p><p>There are five objects:</p><pre><code class="language-julia hljs">n = 5;</code></pre><p>For our example, we use a capacity of 10 units:</p><pre><code class="language-julia hljs">capacity = 10.0;</code></pre><p>and the profit and cost data:</p><pre><code class="language-julia hljs">profit = [5.0, 3.0, 2.0, 7.0, 4.0];
+weight = [2.0, 8.0, 4.0, 2.0, 5.0];</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>Let&#39;s begin constructing the JuMP model for our knapsack problem.</p><p>First, we&#39;ll create a <code>Model</code> object for holding model elements as we construct each part. We&#39;ll also set the solver that will ultimately be called to solve the model, once it&#39;s constructed.</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A JuMP Model
+Feasibility problem with:
+Variables: 0
+Model mode: AUTOMATIC
+CachingOptimizer state: EMPTY_OPTIMIZER
+Solver name: HiGHS</code></pre><p>Next we need the decision variables representing which items are chosen:</p><pre><code class="language-julia hljs">@variable(model, x[1:n], Bin)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]
+ x[4]
+ x[5]</code></pre><p>We now want to constrain those variables so that their combined weight is less than or equal to the given capacity:</p><pre><code class="language-julia hljs">@constraint(model, sum(weight[i] * x[i] for i in 1:n) &lt;= capacity)</code></pre><p class="math-container">\[ 2 x_{1} + 8 x_{2} + 4 x_{3} + 2 x_{4} + 5 x_{5} \leq 10 \]</p><p>Finally, our objective is to maximize the combined profit of the chosen items:</p><pre><code class="language-julia hljs">@objective(model, Max, sum(profit[i] * x[i] for i in 1:n))</code></pre><p>$ 5 x<em>{1} + 3 x</em>{2} + 2 x<em>{3} + 7 x</em>{4} + 4 x_{5} $</p><p>Let&#39;s print a human-readable description of the model and check that the model looks as expected:</p><pre><code class="language-julia hljs">print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Max 5 x[1] + 3 x[2] + 2 x[3] + 7 x[4] + 4 x[5]
+Subject to
+ 2 x[1] + 8 x[2] + 4 x[3] + 2 x[4] + 5 x[5] ≤ 10
+ x[1] binary
+ x[2] binary
+ x[3] binary
+ x[4] binary
+ x[5] binary</code></pre><p>We can now solve the optimization problem and inspect the results.</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : 1.60000e+01
+  Objective bound    : 1.60000e+01
+  Relative gap       : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 5.28336e-04
+  Simplex iterations : 1
+  Barrier iterations : -1
+  Node count         : 1
+</code></pre><p>The items chosen are</p><pre><code class="language-julia hljs">items_chosen = [i for i in 1:n if value(x[i]) &gt; 0.5]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{Int64}:
+ 1
+ 4
+ 5</code></pre><h2 id="Writing-a-function"><a class="docs-heading-anchor" href="#Writing-a-function">Writing a function</a><a id="Writing-a-function-1"></a><a class="docs-heading-anchor-permalink" href="#Writing-a-function" title="Permalink"></a></h2><p>After working interactively, it is good practice to implement your model in a function.</p><p>The function can be used to ensure that the model is given well-defined input data with validation checks, and that the solution process went as expected.</p><pre><code class="language-julia hljs">function solve_knapsack_problem(;
+    profit::Vector{Float64},
+    weight::Vector{Float64},
+    capacity::Float64,
+)
+    n = length(weight)
+    # The profit and weight vectors must be of equal length.
+    @assert length(profit) == n
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, x[1:n], Bin)
+    @objective(model, Max, profit&#39; * x)
+    @constraint(model, weight&#39; * x &lt;= capacity)
+    optimize!(model)
+    @assert termination_status(model) == OPTIMAL
+    @assert primal_status(model) == FEASIBLE_POINT
+    println(&quot;Objective is: &quot;, objective_value(model))
+    println(&quot;Solution is:&quot;)
+    for i in 1:n
+        print(&quot;x[$i] = &quot;, round(Int, value(x[i])))
+        println(&quot;, c[$i] / w[$i] = &quot;, profit[i] / weight[i])
+    end
+    chosen_items = [i for i in 1:n if value(x[i]) &gt; 0.5]
+    return return chosen_items
+end
+
+solve_knapsack_problem(; profit = profit, weight = weight, capacity = capacity)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{Int64}:
+ 1
+ 4
+ 5</code></pre><p>We observe that the chosen items (1, 4, and 5) have the best profit to weight ratio in this particular example.</p><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><p>Here are some things to try next:</p><ul><li>Call the function with different data. What happens as the capacity increases?</li><li>What happens if the profit and weight vectors are different lengths?</li><li>Instead of creating a binary variable with <code>Bin</code>, we could have written <code>@variable(model, 0 &lt;= x[1:n] &lt;= 1, Int)</code>. Verify that this formulation finds the same solution. What happens if we are allowed to take more than one of each item?</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../introduction/">« Introduction</a><a class="docs-footer-nextpage" href="../diet/">The diet problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/lp_sensitivity.jl b/previews/PR3525/tutorials/linear/lp_sensitivity.jl
new file mode 100644
index 00000000000..5d9e3deecf1
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/lp_sensitivity.jl
@@ -0,0 +1,155 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Sensitivity analysis of a linear program
+
+# This tutorial explains how to use the [`lp_sensitivity_report`](@ref) function
+# to create sensitivity reports like those that are produced by the Excel
+# Solver. This is most often used in introductory classes to linear programming.
+
+# In brief, sensitivity analysis of a linear program is about asking two
+# questions:
+#  1) Given an optimal solution, how much can the objective coefficients change
+#     by before a different solution becomes optimal?
+#  2) Given an optimal solution, how much can the right-hand side of a linear
+#     constraint change by before a different solution becomes optimal?
+
+# JuMP provides a function, [`lp_sensitivity_report`](@ref), to help us compute
+# these values, but this tutorial extends that to create more informative
+# tables in the form of a `DataFrame`.
+
+# ## Setup
+
+# This tutorial uses the following packages:
+
+using JuMP
+import HiGHS
+import DataFrames
+
+# as well as this small linear program:
+
+model = Model(HiGHS.Optimizer)
+@variable(model, x >= 0)
+@variable(model, 0 <= y <= 3)
+@variable(model, z <= 1)
+@objective(model, Min, 12x + 20y - z)
+@constraint(model, c1, 6x + 8y >= 100)
+@constraint(model, c2, 7x + 12y >= 120)
+@constraint(model, c3, x + y <= 20)
+optimize!(model)
+solution_summary(model; verbose = true)
+
+# Can you identify:
+#  * The objective coefficient of each variable?
+#  * The right-hand side of each constraint?
+#  * The optimal primal and dual solutions?
+
+# ## Sensitivity reports
+
+# Now let's call [`lp_sensitivity_report`](@ref):
+
+report = lp_sensitivity_report(model)
+
+# It returns a [`SensitivityReport`](@ref) object, which maps:
+#
+#  - Every variable reference to a tuple `(d_lo, d_hi)::Tuple{Float64,Float64}`,
+#    explaining how much the objective coefficient of the corresponding variable
+#    can change by, such that the original basis remains optimal.
+#  - Every constraint reference to a tuple `(d_lo, d_hi)::Tuple{Float64,Float64}`,
+#    explaining how much the right-hand side of the corresponding constraint can
+#    change by, such that the basis remains optimal.
+
+# Both tuples are relative, rather than absolute. So, given an objective
+# coefficient of `1.0` and a tuple `(-0.5, 0.5)`, the objective coefficient can
+# range between `1.0 - 0.5` an `1.0 + 0.5`.
+
+# For example:
+
+report[x]
+
+# indicates that the objective coefficient on `x`, that is, `12`, can
+# decrease by `-0.333` or increase by `3.0` and the primal solution
+# `(15, 1.25)` will remain optimal. In addition:
+
+report[c1]
+
+# means that the right-hand side of the `c1` constraint (100), can decrease
+# by 4 units, or increase by 2.85 units, and the primal solution `(15, 1.25)`
+# will remain optimal.
+
+# ## Variable sensitivity
+
+# By themselves, the tuples aren't informative. Let's put them in context by
+# collating a range of other information about a variable:
+
+function variable_report(xi)
+    return (
+        name = name(xi),
+        lower_bound = has_lower_bound(xi) ? lower_bound(xi) : -Inf,
+        value = value(xi),
+        upper_bound = has_upper_bound(xi) ? upper_bound(xi) : Inf,
+        reduced_cost = reduced_cost(xi),
+        obj_coefficient = coefficient(objective_function(model), xi),
+        allowed_decrease = report[xi][1],
+        allowed_increase = report[xi][2],
+    )
+end
+
+# Calling our function on `x`:
+
+x_report = variable_report(x)
+
+# That's a bit hard to read, so let's call this on every variable in the model
+# and put things into a DataFrame:
+
+variable_df =
+    DataFrames.DataFrame(variable_report(xi) for xi in all_variables(model))
+
+# ## Constraint sensitivity
+
+# We can do something similar with constraints:
+
+function constraint_report(c::ConstraintRef)
+    return (
+        name = name(c),
+        value = value(c),
+        rhs = normalized_rhs(c),
+        slack = normalized_rhs(c) - value(c),
+        shadow_price = shadow_price(c),
+        allowed_decrease = report[c][1],
+        allowed_increase = report[c][2],
+    )
+end
+
+c1_report = constraint_report(c1)
+
+# That's a bit hard to read, so let's call this on every variable in the model
+# and put things into a DataFrame:
+
+constraint_df = DataFrames.DataFrame(
+    constraint_report(ci) for (F, S) in list_of_constraint_types(model) for
+    ci in all_constraints(model, F, S) if F == AffExpr
+)
+
+# ## Analysis questions
+
+# Now we can use these dataframes to ask questions of the solution.
+
+# For example, we can find basic variables by looking for variables with a
+# reduced cost of 0:
+
+basic = filter(row -> iszero(row.reduced_cost), variable_df)
+
+# and non-basic variables by looking for non-zero reduced costs:
+
+non_basic = filter(row -> !iszero(row.reduced_cost), variable_df)
+
+# we can also find constraints that are binding by looking for zero slacks:
+
+binding = filter(row -> iszero(row.slack), constraint_df)
+
+# or non-zero shadow prices:
+
+binding2 = filter(row -> !iszero(row.shadow_price), constraint_df)
diff --git a/previews/PR3525/tutorials/linear/lp_sensitivity/index.html b/previews/PR3525/tutorials/linear/lp_sensitivity/index.html
new file mode 100644
index 00000000000..b35cf3fb611
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/lp_sensitivity/index.html
@@ -0,0 +1,73 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Sensitivity analysis of a linear program · JuMP</title><meta name="title" content="Sensitivity analysis of a linear program · JuMP"/><meta property="og:title" content="Sensitivity analysis of a linear program · JuMP"/><meta property="twitter:title" content="Sensitivity analysis of a linear program · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li class="is-active"><a class="tocitem" href>Sensitivity analysis of a linear program</a><ul class="internal"><li><a class="tocitem" href="#Setup"><span>Setup</span></a></li><li><a class="tocitem" href="#Sensitivity-reports"><span>Sensitivity reports</span></a></li><li><a class="tocitem" href="#Variable-sensitivity"><span>Variable sensitivity</span></a></li><li><a class="tocitem" href="#Constraint-sensitivity"><span>Constraint sensitivity</span></a></li><li><a class="tocitem" href="#Analysis-questions"><span>Analysis questions</span></a></li></ul></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Sensitivity analysis of a linear program</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Sensitivity analysis of a linear program</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/lp_sensitivity.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Sensitivity-analysis-of-a-linear-program"><a class="docs-heading-anchor" href="#Sensitivity-analysis-of-a-linear-program">Sensitivity analysis of a linear program</a><a id="Sensitivity-analysis-of-a-linear-program-1"></a><a class="docs-heading-anchor-permalink" href="#Sensitivity-analysis-of-a-linear-program" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../lp_sensitivity.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial explains how to use the <a href="../../../api/JuMP/#JuMP.lp_sensitivity_report"><code>lp_sensitivity_report</code></a> function to create sensitivity reports like those that are produced by the Excel Solver. This is most often used in introductory classes to linear programming.</p><p>In brief, sensitivity analysis of a linear program is about asking two questions:</p><ol><li>Given an optimal solution, how much can the objective coefficients change by before a different solution becomes optimal?</li><li>Given an optimal solution, how much can the right-hand side of a linear constraint change by before a different solution becomes optimal?</li></ol><p>JuMP provides a function, <a href="../../../api/JuMP/#JuMP.lp_sensitivity_report"><code>lp_sensitivity_report</code></a>, to help us compute these values, but this tutorial extends that to create more informative tables in the form of a <code>DataFrame</code>.</p><h2 id="Setup"><a class="docs-heading-anchor" href="#Setup">Setup</a><a id="Setup-1"></a><a class="docs-heading-anchor-permalink" href="#Setup" title="Permalink"></a></h2><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+import DataFrames</code></pre><p>as well as this small linear program:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+@variable(model, x &gt;= 0)
+@variable(model, 0 &lt;= y &lt;= 3)
+@variable(model, z &lt;= 1)
+@objective(model, Min, 12x + 20y - z)
+@constraint(model, c1, 6x + 8y &gt;= 100)
+@constraint(model, c2, 7x + 12y &gt;= 120)
+@constraint(model, c3, x + y &lt;= 20)
+optimize!(model)
+solution_summary(model; verbose = true)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 2.04000e+02
+  Objective bound    : 0.00000e+00
+  Relative gap       : Inf
+  Dual objective value : 2.04000e+02
+  Primal solution :
+    x : 1.50000e+01
+    y : 1.25000e+00
+    z : 1.00000e+00
+  Dual solution :
+    c1 : 2.50000e-01
+    c2 : 1.50000e+00
+    c3 : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 3.46899e-04
+  Simplex iterations : 2
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><p>Can you identify:</p><ul><li>The objective coefficient of each variable?</li><li>The right-hand side of each constraint?</li><li>The optimal primal and dual solutions?</li></ul><h2 id="Sensitivity-reports"><a class="docs-heading-anchor" href="#Sensitivity-reports">Sensitivity reports</a><a id="Sensitivity-reports-1"></a><a class="docs-heading-anchor-permalink" href="#Sensitivity-reports" title="Permalink"></a></h2><p>Now let&#39;s call <a href="../../../api/JuMP/#JuMP.lp_sensitivity_report"><code>lp_sensitivity_report</code></a>:</p><pre><code class="language-julia hljs">report = lp_sensitivity_report(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">SensitivityReport{Float64}(Dict{ConstraintRef, Tuple{Float64, Float64}}(y ≥ 0 =&gt; (-Inf, 1.25), c3 : x + y ≤ 20 =&gt; (-3.75, Inf), y ≤ 3 =&gt; (-1.75, Inf), x ≥ 0 =&gt; (-Inf, 15.0), c2 : 7 x + 12 y ≥ 120 =&gt; (-3.3333333333333335, 4.666666666666667), c1 : 6 x + 8 y ≥ 100 =&gt; (-4.0, 2.857142857142857), z ≤ 1 =&gt; (-Inf, Inf)), Dict{VariableRef, Tuple{Float64, Float64}}(y =&gt; (-4.0, 0.5714285714285714), x =&gt; (-0.3333333333333333, 3.0), z =&gt; (-Inf, 1.0)))</code></pre><p>It returns a <a href="../../../api/JuMP/#JuMP.SensitivityReport"><code>SensitivityReport</code></a> object, which maps:</p><ul><li>Every variable reference to a tuple <code>(d_lo, d_hi)::Tuple{Float64,Float64}</code>, explaining how much the objective coefficient of the corresponding variable can change by, such that the original basis remains optimal.</li><li>Every constraint reference to a tuple <code>(d_lo, d_hi)::Tuple{Float64,Float64}</code>, explaining how much the right-hand side of the corresponding constraint can change by, such that the basis remains optimal.</li></ul><p>Both tuples are relative, rather than absolute. So, given an objective coefficient of <code>1.0</code> and a tuple <code>(-0.5, 0.5)</code>, the objective coefficient can range between <code>1.0 - 0.5</code> an <code>1.0 + 0.5</code>.</p><p>For example:</p><pre><code class="language-julia hljs">report[x]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(-0.3333333333333333, 3.0)</code></pre><p>indicates that the objective coefficient on <code>x</code>, that is, <code>12</code>, can decrease by <code>-0.333</code> or increase by <code>3.0</code> and the primal solution <code>(15, 1.25)</code> will remain optimal. In addition:</p><pre><code class="language-julia hljs">report[c1]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(-4.0, 2.857142857142857)</code></pre><p>means that the right-hand side of the <code>c1</code> constraint (100), can decrease by 4 units, or increase by 2.85 units, and the primal solution <code>(15, 1.25)</code> will remain optimal.</p><h2 id="Variable-sensitivity"><a class="docs-heading-anchor" href="#Variable-sensitivity">Variable sensitivity</a><a id="Variable-sensitivity-1"></a><a class="docs-heading-anchor-permalink" href="#Variable-sensitivity" title="Permalink"></a></h2><p>By themselves, the tuples aren&#39;t informative. Let&#39;s put them in context by collating a range of other information about a variable:</p><pre><code class="language-julia hljs">function variable_report(xi)
+    return (
+        name = name(xi),
+        lower_bound = has_lower_bound(xi) ? lower_bound(xi) : -Inf,
+        value = value(xi),
+        upper_bound = has_upper_bound(xi) ? upper_bound(xi) : Inf,
+        reduced_cost = reduced_cost(xi),
+        obj_coefficient = coefficient(objective_function(model), xi),
+        allowed_decrease = report[xi][1],
+        allowed_increase = report[xi][2],
+    )
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">variable_report (generic function with 1 method)</code></pre><p>Calling our function on <code>x</code>:</p><pre><code class="language-julia hljs">x_report = variable_report(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(name = &quot;x&quot;, lower_bound = 0.0, value = 15.0, upper_bound = Inf, reduced_cost = 0.0, obj_coefficient = 12.0, allowed_decrease = -0.3333333333333333, allowed_increase = 3.0)</code></pre><p>That&#39;s a bit hard to read, so let&#39;s call this on every variable in the model and put things into a DataFrame:</p><pre><code class="language-julia hljs">variable_df =
+    DataFrames.DataFrame(variable_report(xi) for xi in all_variables(model))</code></pre><div><div style = "float: left;"><span>3×8 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">name</th><th style = "text-align: left;">lower_bound</th><th style = "text-align: left;">value</th><th style = "text-align: left;">upper_bound</th><th style = "text-align: left;">reduced_cost</th><th style = "text-align: left;">obj_coefficient</th><th style = "text-align: left;">allowed_decrease</th><th style = "text-align: left;">allowed_increase</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">x</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">15.0</td><td style = "text-align: right;">Inf</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">12.0</td><td style = "text-align: right;">-0.333333</td><td style = "text-align: right;">3.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">y</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">1.25</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">20.0</td><td style = "text-align: right;">-4.0</td><td style = "text-align: right;">0.571429</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">z</td><td style = "text-align: right;">-Inf</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">-1.0</td><td style = "text-align: right;">-1.0</td><td style = "text-align: right;">-Inf</td><td style = "text-align: right;">1.0</td></tr></tbody></table></div><h2 id="Constraint-sensitivity"><a class="docs-heading-anchor" href="#Constraint-sensitivity">Constraint sensitivity</a><a id="Constraint-sensitivity-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-sensitivity" title="Permalink"></a></h2><p>We can do something similar with constraints:</p><pre><code class="language-julia hljs">function constraint_report(c::ConstraintRef)
+    return (
+        name = name(c),
+        value = value(c),
+        rhs = normalized_rhs(c),
+        slack = normalized_rhs(c) - value(c),
+        shadow_price = shadow_price(c),
+        allowed_decrease = report[c][1],
+        allowed_increase = report[c][2],
+    )
+end
+
+c1_report = constraint_report(c1)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">(name = &quot;c1&quot;, value = 100.0, rhs = 100.0, slack = 0.0, shadow_price = -0.25, allowed_decrease = -4.0, allowed_increase = 2.857142857142857)</code></pre><p>That&#39;s a bit hard to read, so let&#39;s call this on every variable in the model and put things into a DataFrame:</p><pre><code class="language-julia hljs">constraint_df = DataFrames.DataFrame(
+    constraint_report(ci) for (F, S) in list_of_constraint_types(model) for
+    ci in all_constraints(model, F, S) if F == AffExpr
+)</code></pre><div><div style = "float: left;"><span>3×7 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">name</th><th style = "text-align: left;">value</th><th style = "text-align: left;">rhs</th><th style = "text-align: left;">slack</th><th style = "text-align: left;">shadow_price</th><th style = "text-align: left;">allowed_decrease</th><th style = "text-align: left;">allowed_increase</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">c1</td><td style = "text-align: right;">100.0</td><td style = "text-align: right;">100.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">-0.25</td><td style = "text-align: right;">-4.0</td><td style = "text-align: right;">2.85714</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">c2</td><td style = "text-align: right;">120.0</td><td style = "text-align: right;">120.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">-1.5</td><td style = "text-align: right;">-3.33333</td><td style = "text-align: right;">4.66667</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">c3</td><td style = "text-align: right;">16.25</td><td style = "text-align: right;">20.0</td><td style = "text-align: right;">3.75</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">-3.75</td><td style = "text-align: right;">Inf</td></tr></tbody></table></div><h2 id="Analysis-questions"><a class="docs-heading-anchor" href="#Analysis-questions">Analysis questions</a><a id="Analysis-questions-1"></a><a class="docs-heading-anchor-permalink" href="#Analysis-questions" title="Permalink"></a></h2><p>Now we can use these dataframes to ask questions of the solution.</p><p>For example, we can find basic variables by looking for variables with a reduced cost of 0:</p><pre><code class="language-julia hljs">basic = filter(row -&gt; iszero(row.reduced_cost), variable_df)</code></pre><div><div style = "float: left;"><span>2×8 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">name</th><th style = "text-align: left;">lower_bound</th><th style = "text-align: left;">value</th><th style = "text-align: left;">upper_bound</th><th style = "text-align: left;">reduced_cost</th><th style = "text-align: left;">obj_coefficient</th><th style = "text-align: left;">allowed_decrease</th><th style = "text-align: left;">allowed_increase</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">x</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">15.0</td><td style = "text-align: right;">Inf</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">12.0</td><td style = "text-align: right;">-0.333333</td><td style = "text-align: right;">3.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">y</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">1.25</td><td style = "text-align: right;">3.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">20.0</td><td style = "text-align: right;">-4.0</td><td style = "text-align: right;">0.571429</td></tr></tbody></table></div><p>and non-basic variables by looking for non-zero reduced costs:</p><pre><code class="language-julia hljs">non_basic = filter(row -&gt; !iszero(row.reduced_cost), variable_df)</code></pre><div><div style = "float: left;"><span>1×8 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">name</th><th style = "text-align: left;">lower_bound</th><th style = "text-align: left;">value</th><th style = "text-align: left;">upper_bound</th><th style = "text-align: left;">reduced_cost</th><th style = "text-align: left;">obj_coefficient</th><th style = "text-align: left;">allowed_decrease</th><th style = "text-align: left;">allowed_increase</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">z</td><td style = "text-align: right;">-Inf</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: right;">-1.0</td><td style = "text-align: right;">-1.0</td><td style = "text-align: right;">-Inf</td><td style = "text-align: right;">1.0</td></tr></tbody></table></div><p>we can also find constraints that are binding by looking for zero slacks:</p><pre><code class="language-julia hljs">binding = filter(row -&gt; iszero(row.slack), constraint_df)</code></pre><div><div style = "float: left;"><span>2×7 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">name</th><th style = "text-align: left;">value</th><th style = "text-align: left;">rhs</th><th style = "text-align: left;">slack</th><th style = "text-align: left;">shadow_price</th><th style = "text-align: left;">allowed_decrease</th><th style = "text-align: left;">allowed_increase</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">c1</td><td style = "text-align: right;">100.0</td><td style = "text-align: right;">100.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">-0.25</td><td style = "text-align: right;">-4.0</td><td style = "text-align: right;">2.85714</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">c2</td><td style = "text-align: right;">120.0</td><td style = "text-align: right;">120.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">-1.5</td><td style = "text-align: right;">-3.33333</td><td style = "text-align: right;">4.66667</td></tr></tbody></table></div><p>or non-zero shadow prices:</p><pre><code class="language-julia hljs">binding2 = filter(row -&gt; !iszero(row.shadow_price), constraint_df)</code></pre><div><div style = "float: left;"><span>2×7 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">name</th><th style = "text-align: left;">value</th><th style = "text-align: left;">rhs</th><th style = "text-align: left;">slack</th><th style = "text-align: left;">shadow_price</th><th style = "text-align: left;">allowed_decrease</th><th style = "text-align: left;">allowed_increase</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">c1</td><td style = "text-align: right;">100.0</td><td style = "text-align: right;">100.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">-0.25</td><td style = "text-align: right;">-4.0</td><td style = "text-align: right;">2.85714</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">c2</td><td style = "text-align: right;">120.0</td><td style = "text-align: right;">120.0</td><td style = "text-align: right;">0.0</td><td style = "text-align: right;">-1.5</td><td style = "text-align: right;">-3.33333</td><td style = "text-align: right;">4.66667</td></tr></tbody></table></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../callbacks/">« Callbacks</a><a class="docs-footer-nextpage" href="../mip_duality/">Computing the duals of a mixed-integer program »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/mip_duality.jl b/previews/PR3525/tutorials/linear/mip_duality.jl
new file mode 100644
index 00000000000..34bcf4d2ddc
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/mip_duality.jl
@@ -0,0 +1,124 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Computing the duals of a mixed-integer program
+
+# This tutorial explains how to compute the duals of a mixed-integer linear
+# program by fixing the discrete variables to their optimal solution and
+# resolving as a linear program.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import HiGHS
+
+# ## The model
+
+# Our example model is the unit commitment example from [Unit commitment](@ref).
+# The details are unimportant, other than to note that there are two types of
+# continuous variables, `g` and `w`, representing the quantity of generation
+# from thermal and wind plants, and a discrete variable `dispatch`, which is
+# `1` if plant `i` is operating, and `0` if not.
+
+# We are interested in the "dual" of the `power_balance` constraint, because it
+# represents the marginal price of electricity that consumers should pay for
+# their consumption.
+
+generators = [
+    (min = 0.0, max = 1000.0, fixed_cost = 1000.0, variable_cost = 50.0),
+    (min = 300.0, max = 1000.0, fixed_cost = 0.0, variable_cost = 100.0),
+]
+N = length(generators)
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variables(model, begin
+    generators[i].min <= g[i = 1:N] <= generators[i].max
+    0 <= w <= 200
+    dispatch[i = 1:N], Bin
+end)
+@constraints(model, begin
+    power_balance, sum(g[i] for i in 1:N) + w == 1500
+    [i = 1:N], g[i] <= generators[i].max * dispatch[i]
+    [i = 1:N], g[i] >= generators[i].min * dispatch[i]
+end)
+@objective(
+    model,
+    Min,
+    sum(
+        generators[i].fixed_cost * dispatch[i] +
+        generators[i].variable_cost * g[i] for i in 1:N
+    )
+)
+print(model)
+
+# ## Manually fix the variables
+
+# If we optimize this model, we obtain a [`dual_status`](@ref) of `NO_SOLUTION`:
+
+optimize!(model)
+dual_status(model)
+
+# This is because HiGHS cannot compute the duals of a mixed-integer program. We
+# can work around this problem by fixing the integer variables to their optimal
+# solution, relaxing integrality, and re-solving as a linear program.
+
+discrete_values = value.(dispatch)
+fix.(dispatch, discrete_values; force = true)
+unset_binary.(dispatch)
+print(model)
+
+# Now if we re-solve the problem, we obtain a `FEASIBLE_POINT` for the dual:
+
+optimize!(model)
+dual_status(model)
+
+# and a marginal price of electricity of \$100/MWh:
+
+dual(power_balance)
+
+# To reset the problem back to a mixed-integer linear program, we need to
+# [`unfix`](@ref) and call [`set_binary`](@ref):
+
+unfix.(dispatch)
+set_binary.(dispatch)
+print(model)
+
+# ## Use `fix_discrete_variables`
+
+# Manually choosing the variables to relax and fix works for our small example,
+# but it becomes more difficult in problems with a larger number of binary and
+# integer variables. To automate the process we just did manually, JuMP provides
+# the [`fix_discrete_variables`](@ref) function:
+
+optimize!(model)
+dual_status(model)
+
+#-
+
+undo = fix_discrete_variables(model);
+
+# Here `undo` is a function that, when called with no arguments, returns the
+# model to the original mixed-integer formulation.
+
+# !!! tip
+#     After calling [`fix_discrete_variables`](@ref), you can set a new solver
+#     with [`set_optimizer`](@ref) if your mixed-integer solver does not support
+#     computing a dual solution.
+
+print(model)
+
+#-
+
+optimize!(model)
+dual_status(model)
+
+#-
+
+dual(power_balance)
+
+# Finally, call `undo` to revert the reformulation
+
+undo()
+print(model)
diff --git a/previews/PR3525/tutorials/linear/mip_duality/index.html b/previews/PR3525/tutorials/linear/mip_duality/index.html
new file mode 100644
index 00000000000..4977a0e0a25
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/mip_duality/index.html
@@ -0,0 +1,112 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Computing the duals of a mixed-integer program · JuMP</title><meta name="title" content="Computing the duals of a mixed-integer program · JuMP"/><meta property="og:title" content="Computing the duals of a mixed-integer program · JuMP"/><meta property="twitter:title" content="Computing the duals of a mixed-integer program · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li class="is-active"><a class="tocitem" href>Computing the duals of a mixed-integer program</a><ul class="internal"><li><a class="tocitem" href="#The-model"><span>The model</span></a></li><li><a class="tocitem" href="#Manually-fix-the-variables"><span>Manually fix the variables</span></a></li><li><a class="tocitem" href="#Use-fix_discrete_variables"><span>Use <code>fix_discrete_variables</code></span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Computing the duals of a mixed-integer program</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Computing the duals of a mixed-integer program</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/mip_duality.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Computing-the-duals-of-a-mixed-integer-program"><a class="docs-heading-anchor" href="#Computing-the-duals-of-a-mixed-integer-program">Computing the duals of a mixed-integer program</a><a id="Computing-the-duals-of-a-mixed-integer-program-1"></a><a class="docs-heading-anchor-permalink" href="#Computing-the-duals-of-a-mixed-integer-program" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../mip_duality.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial explains how to compute the duals of a mixed-integer linear program by fixing the discrete variables to their optimal solution and resolving as a linear program.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS</code></pre><h2 id="The-model"><a class="docs-heading-anchor" href="#The-model">The model</a><a id="The-model-1"></a><a class="docs-heading-anchor-permalink" href="#The-model" title="Permalink"></a></h2><p>Our example model is the unit commitment example from <a href="../../applications/power_systems/#Unit-commitment">Unit commitment</a>. The details are unimportant, other than to note that there are two types of continuous variables, <code>g</code> and <code>w</code>, representing the quantity of generation from thermal and wind plants, and a discrete variable <code>dispatch</code>, which is <code>1</code> if plant <code>i</code> is operating, and <code>0</code> if not.</p><p>We are interested in the &quot;dual&quot; of the <code>power_balance</code> constraint, because it represents the marginal price of electricity that consumers should pay for their consumption.</p><pre><code class="language-julia hljs">generators = [
+    (min = 0.0, max = 1000.0, fixed_cost = 1000.0, variable_cost = 50.0),
+    (min = 300.0, max = 1000.0, fixed_cost = 0.0, variable_cost = 100.0),
+]
+N = length(generators)
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variables(model, begin
+    generators[i].min &lt;= g[i = 1:N] &lt;= generators[i].max
+    0 &lt;= w &lt;= 200
+    dispatch[i = 1:N], Bin
+end)
+@constraints(model, begin
+    power_balance, sum(g[i] for i in 1:N) + w == 1500
+    [i = 1:N], g[i] &lt;= generators[i].max * dispatch[i]
+    [i = 1:N], g[i] &gt;= generators[i].min * dispatch[i]
+end)
+@objective(
+    model,
+    Min,
+    sum(
+        generators[i].fixed_cost * dispatch[i] +
+        generators[i].variable_cost * g[i] for i in 1:N
+    )
+)
+print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 1000 dispatch[1] + 50 g[1] + 100 g[2]
+Subject to
+ power_balance : g[1] + g[2] + w = 1500
+ g[1] ≥ 0
+ g[2] - 300 dispatch[2] ≥ 0
+ g[1] - 1000 dispatch[1] ≤ 0
+ g[2] - 1000 dispatch[2] ≤ 0
+ g[1] ≥ 0
+ g[2] ≥ 300
+ w ≥ 0
+ g[1] ≤ 1000
+ g[2] ≤ 1000
+ w ≤ 200
+ dispatch[1] binary
+ dispatch[2] binary</code></pre><h2 id="Manually-fix-the-variables"><a class="docs-heading-anchor" href="#Manually-fix-the-variables">Manually fix the variables</a><a id="Manually-fix-the-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Manually-fix-the-variables" title="Permalink"></a></h2><p>If we optimize this model, we obtain a <a href="../../../api/JuMP/#JuMP.dual_status"><code>dual_status</code></a> of <code>NO_SOLUTION</code>:</p><pre><code class="language-julia hljs">optimize!(model)
+dual_status(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">NO_SOLUTION::ResultStatusCode = 0</code></pre><p>This is because HiGHS cannot compute the duals of a mixed-integer program. We can work around this problem by fixing the integer variables to their optimal solution, relaxing integrality, and re-solving as a linear program.</p><pre><code class="language-julia hljs">discrete_values = value.(dispatch)
+fix.(dispatch, discrete_values; force = true)
+unset_binary.(dispatch)
+print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 1000 dispatch[1] + 50 g[1] + 100 g[2]
+Subject to
+ power_balance : g[1] + g[2] + w = 1500
+ g[1] ≥ 0
+ g[2] - 300 dispatch[2] ≥ 0
+ g[1] - 1000 dispatch[1] ≤ 0
+ g[2] - 1000 dispatch[2] ≤ 0
+ dispatch[1] = 1
+ dispatch[2] = 1
+ g[1] ≥ 0
+ g[2] ≥ 300
+ w ≥ 0
+ g[1] ≤ 1000
+ g[2] ≤ 1000
+ w ≤ 200</code></pre><p>Now if we re-solve the problem, we obtain a <code>FEASIBLE_POINT</code> for the dual:</p><pre><code class="language-julia hljs">optimize!(model)
+dual_status(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">FEASIBLE_POINT::ResultStatusCode = 1</code></pre><p>and a marginal price of electricity of <span>$</span>100/MWh:</p><pre><code class="language-julia hljs">dual(power_balance)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">100.0</code></pre><p>To reset the problem back to a mixed-integer linear program, we need to <a href="../../../api/JuMP/#JuMP.unfix"><code>unfix</code></a> and call <a href="../../../api/JuMP/#JuMP.set_binary"><code>set_binary</code></a>:</p><pre><code class="language-julia hljs">unfix.(dispatch)
+set_binary.(dispatch)
+print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 1000 dispatch[1] + 50 g[1] + 100 g[2]
+Subject to
+ power_balance : g[1] + g[2] + w = 1500
+ g[1] ≥ 0
+ g[2] - 300 dispatch[2] ≥ 0
+ g[1] - 1000 dispatch[1] ≤ 0
+ g[2] - 1000 dispatch[2] ≤ 0
+ g[1] ≥ 0
+ g[2] ≥ 300
+ w ≥ 0
+ g[1] ≤ 1000
+ g[2] ≤ 1000
+ w ≤ 200
+ dispatch[1] binary
+ dispatch[2] binary</code></pre><h2 id="Use-fix_discrete_variables"><a class="docs-heading-anchor" href="#Use-fix_discrete_variables">Use <code>fix_discrete_variables</code></a><a id="Use-fix_discrete_variables-1"></a><a class="docs-heading-anchor-permalink" href="#Use-fix_discrete_variables" title="Permalink"></a></h2><p>Manually choosing the variables to relax and fix works for our small example, but it becomes more difficult in problems with a larger number of binary and integer variables. To automate the process we just did manually, JuMP provides the <a href="../../../api/JuMP/#JuMP.fix_discrete_variables"><code>fix_discrete_variables</code></a> function:</p><pre><code class="language-julia hljs">optimize!(model)
+dual_status(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">NO_SOLUTION::ResultStatusCode = 0</code></pre><pre><code class="language-julia hljs">undo = fix_discrete_variables(model);</code></pre><p>Here <code>undo</code> is a function that, when called with no arguments, returns the model to the original mixed-integer formulation.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>After calling <a href="../../../api/JuMP/#JuMP.fix_discrete_variables"><code>fix_discrete_variables</code></a>, you can set a new solver with <a href="../../../api/JuMP/#JuMP.set_optimizer"><code>set_optimizer</code></a> if your mixed-integer solver does not support computing a dual solution.</p></div></div><pre><code class="language-julia hljs">print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 1000 dispatch[1] + 50 g[1] + 100 g[2]
+Subject to
+ power_balance : g[1] + g[2] + w = 1500
+ g[1] ≥ 0
+ g[2] - 300 dispatch[2] ≥ 0
+ g[1] - 1000 dispatch[1] ≤ 0
+ g[2] - 1000 dispatch[2] ≤ 0
+ dispatch[1] = 1
+ dispatch[2] = 1
+ g[1] ≥ 0
+ g[2] ≥ 300
+ w ≥ 0
+ g[1] ≤ 1000
+ g[2] ≤ 1000
+ w ≤ 200</code></pre><pre><code class="language-julia hljs">optimize!(model)
+dual_status(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">FEASIBLE_POINT::ResultStatusCode = 1</code></pre><pre><code class="language-julia hljs">dual(power_balance)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">100.0</code></pre><p>Finally, call <code>undo</code> to revert the reformulation</p><pre><code class="language-julia hljs">undo()
+print(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Min 1000 dispatch[1] + 50 g[1] + 100 g[2]
+Subject to
+ power_balance : g[1] + g[2] + w = 1500
+ g[1] ≥ 0
+ g[2] - 300 dispatch[2] ≥ 0
+ g[1] - 1000 dispatch[1] ≤ 0
+ g[2] - 1000 dispatch[2] ≤ 0
+ g[1] ≥ 0
+ g[2] ≥ 300
+ w ≥ 0
+ g[1] ≤ 1000
+ g[2] ≤ 1000
+ w ≤ 200
+ dispatch[1] binary
+ dispatch[2] binary</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../lp_sensitivity/">« Sensitivity analysis of a linear program</a><a class="docs-footer-nextpage" href="../../nonlinear/introduction/">Introduction »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/multi.jl b/previews/PR3525/tutorials/linear/multi.jl
new file mode 100644
index 00000000000..91cef976613
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi.jl
@@ -0,0 +1,193 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The multi-commodity flow problem
+
+# **This tutorial was originally contributed by Louis Luangkesorn.**
+
+# This tutorial is a JuMP implementation of the multi-commodity transportation
+# model described in
+# [_AMPL: A Modeling Language for Mathematical Programming_](https://ampl.com/resources/the-ampl-book/),
+# by R. Fourer, D.M. Gay and B.W. Kernighan.
+
+# The purpose of this tutorial is to demonstrate creating a JuMP model from an
+# SQLite database.
+
+# ## Required packages
+
+# This tutorial uses the following packages
+
+using JuMP
+import DataFrames
+import HiGHS
+import SQLite
+import Tables
+import Test  #src
+
+const DBInterface = SQLite.DBInterface
+
+# ## Formulation
+
+# The multi-commondity flow problem is a simple extension of
+# [The transportation problem](@ref) to multiple types of products. Briefly, we
+# start with the formulation of the transportation problem:
+# ```math
+# \begin{aligned}
+# \min  && \sum_{i \in O, j \in D} c_{i,j} x_{i,j} \\
+# s.t.  && \sum_{j \in D} x_{i, j} \le s_i && \forall i \in O \\
+#       && \sum_{i \in O} x_{i, j} = d_j && \forall j \in D \\
+#       && x_{i, j} \ge 0 && \forall i \in O, j \in D
+# \end{aligned}
+# ```
+# but introduce a set of products ``P``, resulting in:
+# ```math
+# \begin{aligned}
+# \min  && \sum_{i \in O, j \in D, k \in P} c_{i,j,k} x_{i,j,k} \\
+# s.t.  && \sum_{j \in D} x_{i, j, k} \le s_{i,k} && \forall i \in O, k \in P \\
+#       && \sum_{i \in O} x_{i, j, k} = d_{j,k} && \forall j \in D, k \in P \\
+#       && x_{i, j,k} \ge 0 && \forall i \in O, j \in D, k \in P \\
+#       && \sum_{k \in P} x_{i, j, k} \le u_{i,j} && \forall i \in O, j \in D
+# \end{aligned}
+# ```
+# Note that the last constraint is new; it says that there is a maximum quantity
+# of goods (of any type) that can be transported from origin ``i`` to
+# destination ``j``.
+
+# ## Data
+
+# For the purpose of this tutorial, the JuMP repository contains an example
+# database called `multi.sqlite`.
+
+filename = joinpath(@__DIR__, "multi.sqlite");
+
+# To run locally, download [`multi.sqlite`](multi.sqlite) and update `filename`
+# appropriately.
+
+# Load the database using `SQLite.DB`:
+
+db = SQLite.DB(filename)
+
+# A quick way to see the schema of the database is via `SQLite.tables`:
+
+SQLite.tables(db)
+
+# We interact with the database by executing queries, and then piping the
+# results to an appropriate table. One example is a `DataFrame`:
+
+DBInterface.execute(db, "SELECT * FROM locations") |> DataFrames.DataFrame
+
+# But other table types are supported, such as `Tables.rowtable`:
+
+DBInterface.execute(db, "SELECT * FROM locations") |> Tables.rowtable
+
+# A `rowtable` is a `Vector` of `NamedTuple`s.
+
+# You can construct more complicated SQL queries:
+
+origins =
+    DBInterface.execute(
+        db,
+        "SELECT location FROM locations WHERE type = \"origin\"",
+    ) |> Tables.rowtable
+
+# But for our purpose, we just want the list of strings:
+
+origins = map(y -> y.location, origins)
+
+# We can compose these two operations to get a list of destinations:
+
+destinations =
+    DBInterface.execute(
+        db,
+        "SELECT location FROM locations WHERE type = \"destination\"",
+    ) |>
+    Tables.rowtable |>
+    x -> map(y -> y.location, x)
+
+# And a list of products from our `products` table:
+
+products =
+    DBInterface.execute(db, "SELECT product FROM products") |>
+    Tables.rowtable |>
+    x -> map(y -> y.product, x)
+
+# ## JuMP formulation
+
+# We start by creating a model and our decision variables:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[origins, destinations, products] >= 0)
+
+# One approach when working with databases is to extract all of the data into a
+# Julia datastructure. For example, let's pull the cost table into a DataFrame
+# and then construct our objective by iterating over the rows of the DataFrame:
+
+cost = DBInterface.execute(db, "SELECT * FROM cost") |> DataFrames.DataFrame
+@objective(
+    model,
+    Max,
+    sum(r.cost * x[r.origin, r.destination, r.product] for r in eachrow(cost)),
+);
+
+# If we don't want to use a DataFrame, we can use a `Tables.rowtable` instead:
+
+supply = DBInterface.execute(db, "SELECT * FROM supply") |> Tables.rowtable
+for r in supply
+    @constraint(model, sum(x[r.origin, :, r.product]) <= r.supply)
+end
+
+# Another approach is to execute the query, and then to iterate through the rows
+# of the query using `Tables.rows`:
+
+demand = DBInterface.execute(db, "SELECT * FROM demand")
+for r in Tables.rows(demand)
+    @constraint(model, sum(x[:, r.destination, r.product]) == r.demand)
+end
+
+# !!! warning
+#     Iterating through the rows of a query result works by incrementing a
+#     cursor inside the database. As a consequence, you cannot call
+#     `Tables.rows` twice on the same query result.
+
+# The SQLite queries can be arbitrarily complex. For example, here's a query
+# which builds every possible origin-destination pair:
+
+od_pairs = DBInterface.execute(
+    db,
+    """
+    SELECT a.location as 'origin',
+           b.location as 'destination'
+    FROM locations a
+    INNER JOIN locations b
+    ON a.type = 'origin' AND b.type = 'destination'
+    """,
+)
+
+# With a constraint that we cannot send more than 625 units between each pair:
+
+for r in Tables.rows(od_pairs)
+    @constraint(model, sum(x[r.origin, r.destination, :]) <= 625)
+end
+
+# ## Solution
+
+# Finally, we can optimize the model:
+
+optimize!(model)
+Test.@test termination_status(model) == OPTIMAL     #src
+Test.@test primal_status(model) == FEASIBLE_POINT   #src
+Test.@test objective_value(model) == 225_700.0      #src
+solution_summary(model)
+
+# and print the solution:
+
+begin
+    println("         ", join(products, ' '))
+    for o in origins, d in destinations
+        v = lpad.([round(Int, value(x[o, d, p])) for p in products], 5)
+        println(o, " ", d, " ", join(replace.(v, "   0" => "  . "), " "))
+    end
+end
diff --git a/previews/PR3525/tutorials/linear/multi.sqlite b/previews/PR3525/tutorials/linear/multi.sqlite
new file mode 100644
index 0000000000000000000000000000000000000000..7750c3ef93a39913100b087e98f24188c17a7bec
GIT binary patch
literal 45056
zcmeI5eQ2Fm9mk)S`~E!V{E~BXdz-XrZr7J2Yt}Svx3+7$H0y3_X`D5^))mBSmSk-p
zZL;PTx^91%Iq_X9n>a-91;LjoD9%x^lMS3!P}p$bv?80`Kw+y>iyK4GbAHcrp648w
zh(p+z{w^f<{yz77pO?>RLw@JE_vY}sN2gm;wL4}{ofvP`wm5B$>pFXCHOFyE@+**E
z`j(XsGP;oWuJL%Ohb3oYvx~v2PAMFBU>iKa&qIz~;h$g|+2i3jQ`@8)86X2>fDDiU
zGC&5%02v?y|0@O-Ru}WtO2xf*U2FXInW>4{`Ih=B?H#L+H0rg+$jzhmnj*EnQcd1w
zPfZ`2K3QwjZ)*%7n4Fq#O`jZZP0yNi?$qq$>4}yx4!5yy^j#xy?$FqQcaDr5tsSf%
z?Ta=#P&2kQP>VOwU+V8I<*GX?t}}gda_Y?doinl{YL1_7&0^hD8*UD%@4-S_K40zZ
zbQkW%lbM`4F@AC~`d+Y3C}Qy`8GAL3_w{EGhn6@8V=LydZz$xd!<|bVd$h^sP;_5h
z=*r^p-wSyB^QY(LX6}l<=d9z8SbY48PK$!CKLthOtZ^EOE;b#>45$J98*=~dOrote
zhobxZsGqBLce^L>JmT%l>-&s#E}GW*;zu|M<^lB&W^&cN-AhT-8)y#cN7;qI%~yMR
z+=VGTfSK6|{0BOpJn*bTNU&j4Gvl9x($KnVZt7o<sZOT`a9}_F59uHBa@BgzQirc2
z^Ua~;p)Y^D6x5FxCcjE>s{?<A-@`BAG58LA0X_mP7?+T4WPl8i0Wv@a$N(8217v^<
zkO4A42FSo72HH!0dClm^K4Yz>y(rPXvAT&05*==gnkX;Pdk@@dq8x6^LRpD!sW;3~
z8Hx6djTngeex>u!fkq=*^<thZ_m0+&#HB0C{c>R}D%C~)64V`d4!#MG!cn;nP>0XK
zQ*Z(n7CAveWPl8i0Wv@a$N(8217v^<kO4A42L68ylr!ZuuHE*RklB5I5t-fi7m(SV
ze?H!x+4|>@*}Z=jnce(nAXDjd)9&BLI&J?wE}t)^{lB{Y@52iYyaHF`9o@(P86X2>
zfDDiUGC&5%02v?yWPl8ifw!xHyzf=#X2x4ndC#j(%udhD=UuOQyL|7;yt@AH!)4X~
zza;PIMh3_L86X2>fDDiUGC&5%02v?yWPl93Jq;B7s*ByfuO4Ch?-l%N5&M62{l5su
z9QYIb4t@a_;34_`|Bt~LiReZK$N(8217v^<kO4A42FL&zAOmE847@P|P|Q@So_bk6
zcH-^;6f}sg#}^>40rus74RQ*oSMKXbRs)RW_CQ7hjCdEoS3teEA0y61@HD`P+^DMo
zM%;7i`hOA5I`9NM0$+!}z-4$E9+WEpXW!Tf(r7Y32FL&zAOmE843GgbKnBPF86X2>
zU@-&ojrdi3qdxZP#iYbey^xgHr{}Ru)&g=#iG#CAiGwp(CaVB`QsQ8b$^GI70N`MC
z{lCJ`JMc0*D?jz`arhp574C-*!D+Yyj=+BTH~V^_3VHqt{{w%9Uz9`XMh3_L86X2>
zfDDiUGC&5%02v?yWZ+FRz$=+@)svsO7hUOojd#Q~Ho5LqE<YO&pT{oOd5sBKyYln+
zaQq~1S5M0lhx=o^Ev}>C&v8(-{HVV8LOSPh9gTm4v$)3b-p5%eYgc}ZACBM51NF2l
zak$)jDSjG<9}J7KER2psqj^CMmQ{4<=U841mNgEZWI0vJPyUOKJgY{_5=VPicqXoK
zv^&6jHC%ouU_9JYPs=je_*L%4bv#^M|F7^%4*U#$DF61~U*QG#6|9CbWM#+yM}7&u
z1fP|w0S93>49YcuV{r6Mn_}7#86X2>fDDiUGC&5%02v?yWPl8if&avSYSGoF@?#@h
z(G~W=9lF9UxQvxv8xXp}k?p#|k!@J%WdYC?j>IJaS3ZX$89t{M1VUZm$Us*(vV@gh
z3@GXfM;3I2BlB44m4KYCaAa0jI5LBkUIg%Ug(E$+R8f@AMaL;k4hJ0$ya><0kKw!U
zMffPR-~%uU+hHxV@xROK{!jDA__z6k{60R%-^bs<xAGbf*{ke%_9VN&zQG<~A7nG^
zHg*f!#8$FGcr|=B{CRjj{91S}yeB*!Ho`sOK)5{21TP1_4K4-`2VV|88Jr0wgG0ft
zpil0eZe)NAkO4A42FL&zAOmFJEi)j7vlZDosCVn?_Mt=EWHS6kx<l+X8U8vAVpqa+
zLvdW}G#O?m#EmAy>|wDZVY<cW5!+3M*_C3O$uQd~Zb+DJI+(cLWSC`QtI0515?d0c
z+mLp#*<_e)7egk)Y$!G*OgAQ644MqHTx>KMW&?3u!gQ<BB?e4}+2vw`$uL_M*CJCd
ziPm!@M8C=6W9~B<W|`<sn4Y5vvEF2uZHjd!!|a$?n=m~`E5#a<VYW+LV=~NE#Oj3U
zIVy^p$uL_KJto6!UaU%(o+Bo@O@`S}tTY*BOJYUB^c)pMm&xM&cA6~S@A8D{IqDWw
zlVNs+SY|TJE)x}G>XqnvjtZi~WSA|8vdJ);6Cz=Hj*f_SlVNsDw3!UE`voLS&r!SJ
zCc|u-U?#(CAi{*{IcgJu$uP@B$z+%<iDJU^9EGA_GRy`dZ!*jlL@r@^jzDBh7VkG>
zvUtCK!t@;V3(sVj?G>)cFuO`!{~rz-4qwm9+~G|=%6G7v*`Vy>i{anG%j_lg8}?In
zH#^SW!yaZ|VV{N{$ae^QM*g+HOK_1t#ec*fk}Crr=BMBun3nGtD9P6iJP&(e6Lg24
z5AO@_44(`i4Zjt9U-tZ;3eE<{f=2lM@L;$-cro~O@RM*|D1v{yWk2d^$7FyEkO4A4
z2FL&zAOmFJzcvuHF6Q#Vu!Zoq@tZ7!zl`rrk?whRSqOtWErh`vQ=}W79Tvjib_-!}
zTZ(kIbAyF2c)f)%xHUz()!AYp3~sg%28U9l`<zV{!r-8VFt{;Ay2-iDLKqyd5C%6y
z$XqG7)<PKUw-5&VQlzJ&*FqRvZy^k>OOc+AwHCtQ8Vh0YniT2jSZyH;)+~g<o)qcn
zSY;s$c3TL8D^sMWV}*q<*kvIMcBV*A$8rl{uxcR;E{l-4C{VEw20JW-!E%c9bl8go
z7;LvjVX!SldOBoltycKdK}du_ZXpb^6zS;*Erh|qLKrNiNKZ%6LKrMq2!r_)>FLN>
c2!mM*VK9>-JsrM<FzDd|3cmSYj>6zS0d@CBs{jB1

literal 0
HcmV?d00001

diff --git a/previews/PR3525/tutorials/linear/multi/index.html b/previews/PR3525/tutorials/linear/multi/index.html
new file mode 100644
index 00000000000..32d7ecdf949
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi/index.html
@@ -0,0 +1,171 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The multi-commodity flow problem · JuMP</title><meta name="title" content="The multi-commodity flow problem · JuMP"/><meta property="og:title" content="The multi-commodity flow problem · JuMP"/><meta property="twitter:title" content="The multi-commodity flow problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li class="is-active"><a class="tocitem" href>The multi-commodity flow problem</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Solution"><span>Solution</span></a></li></ul></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The multi-commodity flow problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The multi-commodity flow problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/multi.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-multi-commodity-flow-problem"><a class="docs-heading-anchor" href="#The-multi-commodity-flow-problem">The multi-commodity flow problem</a><a id="The-multi-commodity-flow-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-multi-commodity-flow-problem" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../multi.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Louis Luangkesorn.</strong></p><p>This tutorial is a JuMP implementation of the multi-commodity transportation model described in <a href="https://ampl.com/resources/the-ampl-book/"><em>AMPL: A Modeling Language for Mathematical Programming</em></a>, by R. Fourer, D.M. Gay and B.W. Kernighan.</p><p>The purpose of this tutorial is to demonstrate creating a JuMP model from an SQLite database.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial uses the following packages</p><pre><code class="language-julia hljs">using JuMP
+import DataFrames
+import HiGHS
+import SQLite
+import Tables
+
+const DBInterface = SQLite.DBInterface</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">DBInterface</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>The multi-commondity flow problem is a simple extension of <a href="../transp/#The-transportation-problem">The transportation problem</a> to multiple types of products. Briefly, we start with the formulation of the transportation problem:</p><p class="math-container">\[\begin{aligned}
+\min  &amp;&amp; \sum_{i \in O, j \in D} c_{i,j} x_{i,j} \\
+s.t.  &amp;&amp; \sum_{j \in D} x_{i, j} \le s_i &amp;&amp; \forall i \in O \\
+      &amp;&amp; \sum_{i \in O} x_{i, j} = d_j &amp;&amp; \forall j \in D \\
+      &amp;&amp; x_{i, j} \ge 0 &amp;&amp; \forall i \in O, j \in D
+\end{aligned}\]</p><p>but introduce a set of products <span>$P$</span>, resulting in:</p><p class="math-container">\[\begin{aligned}
+\min  &amp;&amp; \sum_{i \in O, j \in D, k \in P} c_{i,j,k} x_{i,j,k} \\
+s.t.  &amp;&amp; \sum_{j \in D} x_{i, j, k} \le s_{i,k} &amp;&amp; \forall i \in O, k \in P \\
+      &amp;&amp; \sum_{i \in O} x_{i, j, k} = d_{j,k} &amp;&amp; \forall j \in D, k \in P \\
+      &amp;&amp; x_{i, j,k} \ge 0 &amp;&amp; \forall i \in O, j \in D, k \in P \\
+      &amp;&amp; \sum_{k \in P} x_{i, j, k} \le u_{i,j} &amp;&amp; \forall i \in O, j \in D
+\end{aligned}\]</p><p>Note that the last constraint is new; it says that there is a maximum quantity of goods (of any type) that can be transported from origin <span>$i$</span> to destination <span>$j$</span>.</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>For the purpose of this tutorial, the JuMP repository contains an example database called <code>multi.sqlite</code>.</p><pre><code class="language-julia hljs">filename = joinpath(@__DIR__, &quot;multi.sqlite&quot;);</code></pre><p>To run locally, download <a href="../multi.sqlite"><code>multi.sqlite</code></a> and update <code>filename</code> appropriately.</p><p>Load the database using <code>SQLite.DB</code>:</p><pre><code class="language-julia hljs">db = SQLite.DB(filename)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">SQLite.DB(&quot;/home/runner/work/JuMP.jl/JuMP.jl/docs/build/tutorials/linear/multi.sqlite&quot;)</code></pre><p>A quick way to see the schema of the database is via <code>SQLite.tables</code>:</p><pre><code class="language-julia hljs">SQLite.tables(db)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5-element Vector{SQLite.DBTable}:
+ SQLite.DBTable(&quot;locations&quot;, Tables.Schema:
+ :location  Union{Missing, String}
+ :type      Union{Missing, String})
+ SQLite.DBTable(&quot;products&quot;, Tables.Schema:
+ :product  Union{Missing, String})
+ SQLite.DBTable(&quot;supply&quot;, Tables.Schema:
+ :origin   Union{Missing, String}
+ :product  Union{Missing, String}
+ :supply   Union{Missing, Float64})
+ SQLite.DBTable(&quot;demand&quot;, Tables.Schema:
+ :destination  Union{Missing, String}
+ :product      Union{Missing, String}
+ :demand       Union{Missing, Float64})
+ SQLite.DBTable(&quot;cost&quot;, Tables.Schema:
+ :origin       Union{Missing, String}
+ :destination  Union{Missing, String}
+ :product      Union{Missing, String}
+ :cost         Union{Missing, Float64})</code></pre><p>We interact with the database by executing queries, and then piping the results to an appropriate table. One example is a <code>DataFrame</code>:</p><pre><code class="language-julia hljs">DBInterface.execute(db, &quot;SELECT * FROM locations&quot;) |&gt; DataFrames.DataFrame</code></pre><div><div style = "float: left;"><span>10×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">location</th><th style = "text-align: left;">type</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">GARY</td><td style = "text-align: left;">origin</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">CLEV</td><td style = "text-align: left;">origin</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">PITT</td><td style = "text-align: left;">origin</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">FRA</td><td style = "text-align: left;">destination</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">DET</td><td style = "text-align: left;">destination</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">LAN</td><td style = "text-align: left;">destination</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">WIN</td><td style = "text-align: left;">destination</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">STL</td><td style = "text-align: left;">destination</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">FRE</td><td style = "text-align: left;">destination</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">LAF</td><td style = "text-align: left;">destination</td></tr></tbody></table></div><p>But other table types are supported, such as <code>Tables.rowtable</code>:</p><pre><code class="language-julia hljs">DBInterface.execute(db, &quot;SELECT * FROM locations&quot;) |&gt; Tables.rowtable</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">10-element Vector{NamedTuple{(:location, :type), Tuple{String, String}}}:
+ (location = &quot;GARY&quot;, type = &quot;origin&quot;)
+ (location = &quot;CLEV&quot;, type = &quot;origin&quot;)
+ (location = &quot;PITT&quot;, type = &quot;origin&quot;)
+ (location = &quot;FRA&quot;, type = &quot;destination&quot;)
+ (location = &quot;DET&quot;, type = &quot;destination&quot;)
+ (location = &quot;LAN&quot;, type = &quot;destination&quot;)
+ (location = &quot;WIN&quot;, type = &quot;destination&quot;)
+ (location = &quot;STL&quot;, type = &quot;destination&quot;)
+ (location = &quot;FRE&quot;, type = &quot;destination&quot;)
+ (location = &quot;LAF&quot;, type = &quot;destination&quot;)</code></pre><p>A <code>rowtable</code> is a <code>Vector</code> of <code>NamedTuple</code>s.</p><p>You can construct more complicated SQL queries:</p><pre><code class="language-julia hljs">origins =
+    DBInterface.execute(
+        db,
+        &quot;SELECT location FROM locations WHERE type = \&quot;origin\&quot;&quot;,
+    ) |&gt; Tables.rowtable</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{NamedTuple{(:location,), Tuple{String}}}:
+ (location = &quot;GARY&quot;,)
+ (location = &quot;CLEV&quot;,)
+ (location = &quot;PITT&quot;,)</code></pre><p>But for our purpose, we just want the list of strings:</p><pre><code class="language-julia hljs">origins = map(y -&gt; y.location, origins)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{String}:
+ &quot;GARY&quot;
+ &quot;CLEV&quot;
+ &quot;PITT&quot;</code></pre><p>We can compose these two operations to get a list of destinations:</p><pre><code class="language-julia hljs">destinations =
+    DBInterface.execute(
+        db,
+        &quot;SELECT location FROM locations WHERE type = \&quot;destination\&quot;&quot;,
+    ) |&gt;
+    Tables.rowtable |&gt;
+    x -&gt; map(y -&gt; y.location, x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">7-element Vector{String}:
+ &quot;FRA&quot;
+ &quot;DET&quot;
+ &quot;LAN&quot;
+ &quot;WIN&quot;
+ &quot;STL&quot;
+ &quot;FRE&quot;
+ &quot;LAF&quot;</code></pre><p>And a list of products from our <code>products</code> table:</p><pre><code class="language-julia hljs">products =
+    DBInterface.execute(db, &quot;SELECT product FROM products&quot;) |&gt;
+    Tables.rowtable |&gt;
+    x -&gt; map(y -&gt; y.product, x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{String}:
+ &quot;bands&quot;
+ &quot;coils&quot;
+ &quot;plate&quot;</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>We start by creating a model and our decision variables:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+@variable(model, x[origins, destinations, products] &gt;= 0)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-dimensional DenseAxisArray{VariableRef,3,...} with index sets:
+    Dimension 1, [&quot;GARY&quot;, &quot;CLEV&quot;, &quot;PITT&quot;]
+    Dimension 2, [&quot;FRA&quot;, &quot;DET&quot;, &quot;LAN&quot;, &quot;WIN&quot;, &quot;STL&quot;, &quot;FRE&quot;, &quot;LAF&quot;]
+    Dimension 3, [&quot;bands&quot;, &quot;coils&quot;, &quot;plate&quot;]
+And data, a 3×7×3 Array{VariableRef, 3}:
+[:, :, &quot;bands&quot;] =
+ x[GARY,FRA,bands]  x[GARY,DET,bands]  …  x[GARY,LAF,bands]
+ x[CLEV,FRA,bands]  x[CLEV,DET,bands]     x[CLEV,LAF,bands]
+ x[PITT,FRA,bands]  x[PITT,DET,bands]     x[PITT,LAF,bands]
+
+[:, :, &quot;coils&quot;] =
+ x[GARY,FRA,coils]  x[GARY,DET,coils]  …  x[GARY,LAF,coils]
+ x[CLEV,FRA,coils]  x[CLEV,DET,coils]     x[CLEV,LAF,coils]
+ x[PITT,FRA,coils]  x[PITT,DET,coils]     x[PITT,LAF,coils]
+
+[:, :, &quot;plate&quot;] =
+ x[GARY,FRA,plate]  x[GARY,DET,plate]  …  x[GARY,LAF,plate]
+ x[CLEV,FRA,plate]  x[CLEV,DET,plate]     x[CLEV,LAF,plate]
+ x[PITT,FRA,plate]  x[PITT,DET,plate]     x[PITT,LAF,plate]</code></pre><p>One approach when working with databases is to extract all of the data into a Julia datastructure. For example, let&#39;s pull the cost table into a DataFrame and then construct our objective by iterating over the rows of the DataFrame:</p><pre><code class="language-julia hljs">cost = DBInterface.execute(db, &quot;SELECT * FROM cost&quot;) |&gt; DataFrames.DataFrame
+@objective(
+    model,
+    Max,
+    sum(r.cost * x[r.origin, r.destination, r.product] for r in eachrow(cost)),
+);</code></pre><p>If we don&#39;t want to use a DataFrame, we can use a <code>Tables.rowtable</code> instead:</p><pre><code class="language-julia hljs">supply = DBInterface.execute(db, &quot;SELECT * FROM supply&quot;) |&gt; Tables.rowtable
+for r in supply
+    @constraint(model, sum(x[r.origin, :, r.product]) &lt;= r.supply)
+end</code></pre><p>Another approach is to execute the query, and then to iterate through the rows of the query using <code>Tables.rows</code>:</p><pre><code class="language-julia hljs">demand = DBInterface.execute(db, &quot;SELECT * FROM demand&quot;)
+for r in Tables.rows(demand)
+    @constraint(model, sum(x[:, r.destination, r.product]) == r.demand)
+end</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Iterating through the rows of a query result works by incrementing a cursor inside the database. As a consequence, you cannot call <code>Tables.rows</code> twice on the same query result.</p></div></div><p>The SQLite queries can be arbitrarily complex. For example, here&#39;s a query which builds every possible origin-destination pair:</p><pre><code class="language-julia hljs">od_pairs = DBInterface.execute(
+    db,
+    &quot;&quot;&quot;
+    SELECT a.location as &#39;origin&#39;,
+           b.location as &#39;destination&#39;
+    FROM locations a
+    INNER JOIN locations b
+    ON a.type = &#39;origin&#39; AND b.type = &#39;destination&#39;
+    &quot;&quot;&quot;,
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">SQLite.Query{false}(SQLite.Stmt(SQLite.DB(&quot;/home/runner/work/JuMP.jl/JuMP.jl/docs/build/tutorials/linear/multi.sqlite&quot;), Base.RefValue{Ptr{SQLite.C.sqlite3_stmt}}(Ptr{SQLite.C.sqlite3_stmt} @0x000000003dce34e8), Dict{Int64, Any}()), Base.RefValue{Int32}(100), [:origin, :destination], Type[Union{Missing, String}, Union{Missing, String}], Dict(:origin =&gt; 1, :destination =&gt; 2), Base.RefValue{Int64}(0))</code></pre><p>With a constraint that we cannot send more than 625 units between each pair:</p><pre><code class="language-julia hljs">for r in Tables.rows(od_pairs)
+    @constraint(model, sum(x[r.origin, r.destination, :]) &lt;= 625)
+end</code></pre><h2 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h2><p>Finally, we can optimize the model:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 2.25700e+05
+  Objective bound    : 2.25700e+05
+  Relative gap       : Inf
+  Dual objective value : 2.25700e+05
+
+* Work counters
+  Solve time (sec)   : 8.72850e-04
+  Simplex iterations : 54
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><p>and print the solution:</p><pre><code class="language-julia hljs">begin
+    println(&quot;         &quot;, join(products, &#39; &#39;))
+    for o in origins, d in destinations
+        v = lpad.([round(Int, value(x[o, d, p])) for p in products], 5)
+        println(o, &quot; &quot;, d, &quot; &quot;, join(replace.(v, &quot;   0&quot; =&gt; &quot;  . &quot;), &quot; &quot;))
+    end
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">         bands coils plate
+GARY FRA    25   500   100
+GARY DET   125    .     50
+GARY LAN    .     .     .
+GARY WIN    .     .     50
+GARY STL   250   300    .
+GARY FRE    .     .     .
+GARY LAF    .     .     .
+CLEV FRA   275    .     .
+CLEV DET   100   200    50
+CLEV LAN   100    .     .
+CLEV WIN    .     .     .
+CLEV STL    .    625    .
+CLEV FRE   225   400    .
+CLEV LAF    .    375   250
+PITT FRA    .     .     .
+PITT DET    75   550    .
+PITT LAN    .    400    .
+PITT WIN    75   250    .
+PITT STL   400    25   200
+PITT FRE    .    450   100
+PITT LAF   250   125    .</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../factory_schedule/">« The factory schedule example</a><a class="docs-footer-nextpage" href="../multi_commodity_network/">The network multi-commodity flow problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/multi_commodity_network.jl b/previews/PR3525/tutorials/linear/multi_commodity_network.jl
new file mode 100644
index 00000000000..3f6f234a76b
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_commodity_network.jl
@@ -0,0 +1,218 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The network multi-commodity flow problem
+
+# This tutorial is a variation of [The multi-commodity flow problem](@ref) where
+# the graph is a network instead of a bipartite graph.
+
+# The purpose of this tutorial is to demonstrate a style of modeling that
+# uses relational algebra.
+
+# ## Required packages
+
+# This tutorial uses the following packages:
+
+using JuMP
+import DataFrames
+import HiGHS
+import SQLite
+import SQLite.DBInterface
+import Test  #src
+
+# ## Formulation
+
+# The network multi-commondity flow problem is an extension of the
+# [The multi-commodity flow problem](@ref), where instead of having a bipartite
+# graph of supply and demand nodes, the graph can contains a set of nodes,
+# $i \in \mathcal{N}$ , which each have a (potentially zero) supply capacity,
+# $u^s_{i,p}$, and (potentially zero) a demand, $d_{i,p}$ for each commodity
+# $p \in P$. The nodes are connected by a set of edges $(i, j) \in \mathcal{E}$,
+# which have a shipment cost $c^x_{i,j,p}$ and a total flow capacity of
+# $u^x_{i,j}$.
+
+# Our take is to choose an optimal supply for each node $s_{i,p}$, as well as
+# the optimal transshipment $x_{i,j,p}$ that minimizes the total cost.
+
+# The mathematical formulation is:
+# ```math
+# \begin{aligned}
+# \min \;\; & \sum_{(i,j)\in\mathcal{E}, p \in P} c^x_{i,j,p} x_{i,j,p} + \sum_{i\in\mathcal{N}, p \in P} c^s_{i,p} s_{i,p} \\
+# s.t. \;\; & s_{i,p} + \sum_{(j, i) \in \mathcal{E}} x_{j,i,p} - \sum_{(i,j) \in \mathcal{E}} x_{i,j,p} = d_{i,p} & \forall i \in \mathcal{N}, p \in P \\
+#           & x_{i,j,p} \ge 0           & \forall (i, j) \in \mathcal{E}, p \in P \\
+#           & \sum_{p \in P} x_{i,j,p} \le u^x_{i,j}           & \forall (i, j) \in \mathcal{E} \\
+#           & 0 \le s_{i,p} \le u^s_{i,p} & \forall i \in \mathcal{N}, p \in P.
+# \end{aligned}
+# ```
+#
+# The purpose of this tutorial is to demonstrate how this model can be built
+# using relational algebra instead of a direct math-to-code translation of the
+# summations.
+
+# ## Data
+
+# For the purpose of this tutorial, the JuMP repository contains an example
+# database called `commodity_nz.db`:
+
+filename = joinpath(@__DIR__, "commodity_nz.db");
+
+# To run locally, download [`commodity_nz.db`](commodity_nz.db) and update
+# `filename` appropriately.
+
+# Load the database using `SQLite.DB`:
+
+db = SQLite.DB(filename)
+
+# A quick way to see the schema of the database is via `SQLite.tables`:
+
+SQLite.tables(db)
+
+# We interact with the database by executing queries and then loading the
+# results into a DataFrame:
+
+function get_table(db, table)
+    query = DBInterface.execute(db, "SELECT * FROM $table")
+    return DataFrames.DataFrame(query)
+end
+
+# The `shipping` table contains the set of arcs and their distances:
+
+df_shipping = get_table(db, "shipping")
+
+# The `products` table contains the shipping cost per kilometer of each product:
+
+df_products = get_table(db, "products")
+
+# The `supply` table contains the supply capacity of each node, as well as the
+# cost:
+
+df_supply = get_table(db, "supply")
+
+# The `demand` table contains the demand of each node:
+
+df_demand = get_table(db, "demand")
+
+# ## JuMP formulation
+
+# We start by creating a model and our decision variables:
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+
+# For the shipping decisions, we create a new column in `df_shipping` called
+# `x_flow`, which has one non-negative decision variable for each row:
+
+df_shipping.x_flow = @variable(model, x[1:size(df_shipping, 1)] >= 0)
+df_shipping
+
+# For the supply, we add a variable to each row, and then set the upper bound to
+# the capacity of each supply node:
+
+df_supply.x_supply = @variable(model, s[1:size(df_supply, 1)] >= 0)
+set_upper_bound.(df_supply.x_supply, df_supply.capacity)
+df_supply
+
+# Our objective is to minimize the shipping cost plus the supply cost. To
+# compute the flow cost, we need to join the shipping table, which contains
+# `distance_km` with the products table, which contains `cost_per_km`:
+
+df_cost = DataFrames.leftjoin(df_shipping, df_products; on = [:product])
+df_cost.flow_cost = df_cost.cost_per_km .* df_cost.distance_km
+df_cost
+
+# Then we can use linear algebra to compute the inner product between two
+# columns:
+
+@objective(
+    model,
+    Min,
+    df_cost.flow_cost' * df_shipping.x_flow +
+    df_supply.cost' * df_supply.x_supply
+);
+
+# For the flow capacities on each arc, we use `DataFrames.groupby` to partition
+# the flow variables based on `:origin` and `:destination`, and then we
+# constrain their sum to be less than a fixed capacity.
+
+capacity = 30
+for df in DataFrames.groupby(df_shipping, [:origin, :destination])
+    @constraint(model, sum(df.x_flow) <= capacity)
+end
+
+# For each node in the graph, we need to compute a mass balance constraint
+# which says that for each product, the supply, plus the flow into the node, and
+# less the flow out of the node is equal to the demand.
+
+# We can compute an expression for the flow out of each node using
+# `DataFrames.groupby` on the `origin` and `product` columns of the
+# `df_shipping` table:
+
+df_flow_out = DataFrames.DataFrame(
+    (node = i.origin, product = i.product, x_flow_out = sum(df.x_flow)) for
+    (i, df) in pairs(DataFrames.groupby(df_shipping, [:origin, :product]))
+)
+
+# We can compute an expression for the flow into each node using
+# `DataFrames.groupby` on the `destination` and `product` columns of the
+# `df_shipping` table:
+
+df_flow_in = DataFrames.DataFrame(
+    (node = i.destination, product = i.product, x_flow_in = sum(df.x_flow))
+    for (i, df) in
+    pairs(DataFrames.groupby(df_shipping, [:destination, :product]))
+)
+
+# We can join the two tables together using `DataFrames.outerjoin`. We need to
+# use `outerjoin` here because there might be missing rows.
+
+df = DataFrames.outerjoin(df_flow_in, df_flow_out; on = [:node, :product])
+
+# Next, we need to join the supply column:
+
+df = DataFrames.leftjoin(
+    df,
+    DataFrames.select(df_supply, [:origin, :product, :x_supply]);
+    on = [:node => :origin, :product],
+)
+
+# and then the demand column
+
+df = DataFrames.leftjoin(
+    df,
+    DataFrames.select(df_demand, [:destination, :product, :demand]);
+    on = [:node => :destination, :product],
+)
+
+# Now we're ready to add our mass balance constraint. Because some rows contain
+# `missing` values, we need to use `coalesce` to convert any `missing` into a
+# numeric value:
+
+@constraint(
+    model,
+    [r in eachrow(df)],
+    coalesce(r.x_supply, 0.0) + coalesce(r.x_flow_in, 0.0) -
+    coalesce(r.x_flow_out, 0.0) == coalesce(r.demand, 0.0),
+);
+
+# ## Solution
+
+# Finally, we can optimize the model:
+
+optimize!(model)
+Test.@test termination_status(model) == OPTIMAL     #src
+Test.@test primal_status(model) == FEASIBLE_POINT   #src
+solution_summary(model)
+
+# update the solution in the DataFrames:
+
+df_shipping.x_flow = value.(df_shipping.x_flow)
+df_supply.x_supply = value.(df_supply.x_supply);
+
+# and display the optimal solution for flows:
+
+DataFrames.select(
+    filter!(row -> row.x_flow > 0.0, df_shipping),
+    [:origin, :destination, :product, :x_flow],
+)
diff --git a/previews/PR3525/tutorials/linear/multi_commodity_network/index.html b/previews/PR3525/tutorials/linear/multi_commodity_network/index.html
new file mode 100644
index 00000000000..a49b78563f3
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_commodity_network/index.html
@@ -0,0 +1,96 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The network multi-commodity flow problem · JuMP</title><meta name="title" content="The network multi-commodity flow problem · JuMP"/><meta property="og:title" content="The network multi-commodity flow problem · JuMP"/><meta property="twitter:title" content="The network multi-commodity flow problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li class="is-active"><a class="tocitem" href>The network multi-commodity flow problem</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Solution"><span>Solution</span></a></li></ul></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The network multi-commodity flow problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The network multi-commodity flow problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/multi_commodity_network.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-network-multi-commodity-flow-problem"><a class="docs-heading-anchor" href="#The-network-multi-commodity-flow-problem">The network multi-commodity flow problem</a><a id="The-network-multi-commodity-flow-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-network-multi-commodity-flow-problem" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../multi_commodity_network.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial is a variation of <a href="../multi/#The-multi-commodity-flow-problem">The multi-commodity flow problem</a> where the graph is a network instead of a bipartite graph.</p><p>The purpose of this tutorial is to demonstrate a style of modeling that uses relational algebra.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import DataFrames
+import HiGHS
+import SQLite
+import SQLite.DBInterface</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>The network multi-commondity flow problem is an extension of the <a href="../multi/#The-multi-commodity-flow-problem">The multi-commodity flow problem</a>, where instead of having a bipartite graph of supply and demand nodes, the graph can contains a set of nodes, <span>$i \in \mathcal{N}$</span> , which each have a (potentially zero) supply capacity, <span>$u^s_{i,p}$</span>, and (potentially zero) a demand, <span>$d_{i,p}$</span> for each commodity <span>$p \in P$</span>. The nodes are connected by a set of edges <span>$(i, j) \in \mathcal{E}$</span>, which have a shipment cost <span>$c^x_{i,j,p}$</span> and a total flow capacity of <span>$u^x_{i,j}$</span>.</p><p>Our take is to choose an optimal supply for each node <span>$s_{i,p}$</span>, as well as the optimal transshipment <span>$x_{i,j,p}$</span> that minimizes the total cost.</p><p>The mathematical formulation is:</p><p class="math-container">\[\begin{aligned}
+\min \;\; &amp; \sum_{(i,j)\in\mathcal{E}, p \in P} c^x_{i,j,p} x_{i,j,p} + \sum_{i\in\mathcal{N}, p \in P} c^s_{i,p} s_{i,p} \\
+s.t. \;\; &amp; s_{i,p} + \sum_{(j, i) \in \mathcal{E}} x_{j,i,p} - \sum_{(i,j) \in \mathcal{E}} x_{i,j,p} = d_{i,p} &amp; \forall i \in \mathcal{N}, p \in P \\
+          &amp; x_{i,j,p} \ge 0           &amp; \forall (i, j) \in \mathcal{E}, p \in P \\
+          &amp; \sum_{p \in P} x_{i,j,p} \le u^x_{i,j}           &amp; \forall (i, j) \in \mathcal{E} \\
+          &amp; 0 \le s_{i,p} \le u^s_{i,p} &amp; \forall i \in \mathcal{N}, p \in P.
+\end{aligned}\]</p><p>The purpose of this tutorial is to demonstrate how this model can be built using relational algebra instead of a direct math-to-code translation of the summations.</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>For the purpose of this tutorial, the JuMP repository contains an example database called <code>commodity_nz.db</code>:</p><pre><code class="language-julia hljs">filename = joinpath(@__DIR__, &quot;commodity_nz.db&quot;);</code></pre><p>To run locally, download <a href="../commodity_nz.db"><code>commodity_nz.db</code></a> and update <code>filename</code> appropriately.</p><p>Load the database using <code>SQLite.DB</code>:</p><pre><code class="language-julia hljs">db = SQLite.DB(filename)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">SQLite.DB(&quot;/home/runner/work/JuMP.jl/JuMP.jl/docs/build/tutorials/linear/commodity_nz.db&quot;)</code></pre><p>A quick way to see the schema of the database is via <code>SQLite.tables</code>:</p><pre><code class="language-julia hljs">SQLite.tables(db)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{SQLite.DBTable}:
+ SQLite.DBTable(&quot;products&quot;, Tables.Schema:
+ :product      Union{Missing, String}
+ :cost_per_km  Union{Missing, Float64})
+ SQLite.DBTable(&quot;shipping&quot;, Tables.Schema:
+ :origin       Union{Missing, String}
+ :destination  Union{Missing, String}
+ :product      Union{Missing, String}
+ :distance_km  Union{Missing, Float64})
+ SQLite.DBTable(&quot;supply&quot;, Tables.Schema:
+ :origin    Union{Missing, String}
+ :product   Union{Missing, String}
+ :capacity  Union{Missing, Float64}
+ :cost      Union{Missing, Float64})
+ SQLite.DBTable(&quot;demand&quot;, Tables.Schema:
+ :destination  Union{Missing, String}
+ :product      Union{Missing, String}
+ :demand       Union{Missing, Float64})</code></pre><p>We interact with the database by executing queries and then loading the results into a DataFrame:</p><pre><code class="language-julia hljs">function get_table(db, table)
+    query = DBInterface.execute(db, &quot;SELECT * FROM $table&quot;)
+    return DataFrames.DataFrame(query)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">get_table (generic function with 1 method)</code></pre><p>The <code>shipping</code> table contains the set of arcs and their distances:</p><pre><code class="language-julia hljs">df_shipping = get_table(db, &quot;shipping&quot;)</code></pre><div><div style = "float: left;"><span>16×4 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">origin</th><th style = "text-align: left;">destination</th><th style = "text-align: left;">product</th><th style = "text-align: left;">distance_km</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">112.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">225.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">1070.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">112.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">107.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">392.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">225.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">107.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">1070.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">112.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">1070.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">112.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">392.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">225.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">107.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">1070.0</td></tr></tbody></table></div><p>The <code>products</code> table contains the shipping cost per kilometer of each product:</p><pre><code class="language-julia hljs">df_products = get_table(db, &quot;products&quot;)</code></pre><div><div style = "float: left;"><span>2×2 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">product</th><th style = "text-align: left;">cost_per_km</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">0.001</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">0.01</td></tr></tbody></table></div><p>The <code>supply</code> table contains the supply capacity of each node, as well as the cost:</p><pre><code class="language-julia hljs">df_supply = get_table(db, &quot;supply&quot;)</code></pre><div><div style = "float: left;"><span>4×4 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">origin</th><th style = "text-align: left;">product</th><th style = "text-align: left;">capacity</th><th style = "text-align: left;">cost</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">0.5</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">6.0</td><td style = "text-align: right;">1.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">26.0</td><td style = "text-align: right;">1.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">0.6</td></tr></tbody></table></div><p>The <code>demand</code> table contains the demand of each node:</p><pre><code class="language-julia hljs">df_demand = get_table(db, &quot;demand&quot;)</code></pre><div><div style = "float: left;"><span>8×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">destination</th><th style = "text-align: left;">product</th><th style = "text-align: left;">demand</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">16.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">16.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">4.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">4.0</td></tr></tbody></table></div><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>We start by creating a model and our decision variables:</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)</code></pre><p>For the shipping decisions, we create a new column in <code>df_shipping</code> called <code>x_flow</code>, which has one non-negative decision variable for each row:</p><pre><code class="language-julia hljs">df_shipping.x_flow = @variable(model, x[1:size(df_shipping, 1)] &gt;= 0)
+df_shipping</code></pre><div><div style = "float: left;"><span>16×5 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">origin</th><th style = "text-align: left;">destination</th><th style = "text-align: left;">product</th><th style = "text-align: left;">distance_km</th><th style = "text-align: left;">x_flow</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "VariableRef" style = "text-align: left;">GenericV…</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[1]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">225.0</td><td style = "text-align: left;">x[2]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[3]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[4]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">107.0</td><td style = "text-align: left;">x[5]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">392.0</td><td style = "text-align: left;">x[6]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">225.0</td><td style = "text-align: left;">x[7]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">107.0</td><td style = "text-align: left;">x[8]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[9]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[10]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[11]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[12]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">392.0</td><td style = "text-align: left;">x[13]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">225.0</td><td style = "text-align: left;">x[14]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">107.0</td><td style = "text-align: left;">x[15]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[16]</td></tr></tbody></table></div><p>For the supply, we add a variable to each row, and then set the upper bound to the capacity of each supply node:</p><pre><code class="language-julia hljs">df_supply.x_supply = @variable(model, s[1:size(df_supply, 1)] &gt;= 0)
+set_upper_bound.(df_supply.x_supply, df_supply.capacity)
+df_supply</code></pre><div><div style = "float: left;"><span>4×5 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">origin</th><th style = "text-align: left;">product</th><th style = "text-align: left;">capacity</th><th style = "text-align: left;">cost</th><th style = "text-align: left;">x_supply</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "VariableRef" style = "text-align: left;">GenericV…</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">0.5</td><td style = "text-align: left;">s[1]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">6.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: left;">s[2]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">26.0</td><td style = "text-align: right;">1.0</td><td style = "text-align: left;">s[3]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">10.0</td><td style = "text-align: right;">0.6</td><td style = "text-align: left;">s[4]</td></tr></tbody></table></div><p>Our objective is to minimize the shipping cost plus the supply cost. To compute the flow cost, we need to join the shipping table, which contains <code>distance_km</code> with the products table, which contains <code>cost_per_km</code>:</p><pre><code class="language-julia hljs">df_cost = DataFrames.leftjoin(df_shipping, df_products; on = [:product])
+df_cost.flow_cost = df_cost.cost_per_km .* df_cost.distance_km
+df_cost</code></pre><div><div style = "float: left;"><span>16×7 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">origin</th><th style = "text-align: left;">destination</th><th style = "text-align: left;">product</th><th style = "text-align: left;">distance_km</th><th style = "text-align: left;">x_flow</th><th style = "text-align: left;">cost_per_km</th><th style = "text-align: left;">flow_cost</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "VariableRef" style = "text-align: left;">GenericV…</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[1]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">0.112</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">225.0</td><td style = "text-align: left;">x[2]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">0.225</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[3]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">1.07</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[4]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">0.112</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">107.0</td><td style = "text-align: left;">x[5]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">0.107</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">392.0</td><td style = "text-align: left;">x[6]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">0.392</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">225.0</td><td style = "text-align: left;">x[7]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">0.225</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">107.0</td><td style = "text-align: left;">x[8]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">0.107</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[9]</td><td style = "text-align: right;">0.001</td><td style = "text-align: right;">1.07</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[10]</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">1.12</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[11]</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">10.7</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">112.0</td><td style = "text-align: left;">x[12]</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">1.12</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">392.0</td><td style = "text-align: left;">x[13]</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">3.92</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">14</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">225.0</td><td style = "text-align: left;">x[14]</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">2.25</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">15</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">107.0</td><td style = "text-align: left;">x[15]</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">1.07</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">16</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">1070.0</td><td style = "text-align: left;">x[16]</td><td style = "text-align: right;">0.01</td><td style = "text-align: right;">10.7</td></tr></tbody></table></div><p>Then we can use linear algebra to compute the inner product between two columns:</p><pre><code class="language-julia hljs">@objective(
+    model,
+    Min,
+    df_cost.flow_cost&#39; * df_shipping.x_flow +
+    df_supply.cost&#39; * df_supply.x_supply
+);</code></pre><p>For the flow capacities on each arc, we use <code>DataFrames.groupby</code> to partition the flow variables based on <code>:origin</code> and <code>:destination</code>, and then we constrain their sum to be less than a fixed capacity.</p><pre><code class="language-julia hljs">capacity = 30
+for df in DataFrames.groupby(df_shipping, [:origin, :destination])
+    @constraint(model, sum(df.x_flow) &lt;= capacity)
+end</code></pre><p>For each node in the graph, we need to compute a mass balance constraint which says that for each product, the supply, plus the flow into the node, and less the flow out of the node is equal to the demand.</p><p>We can compute an expression for the flow out of each node using <code>DataFrames.groupby</code> on the <code>origin</code> and <code>product</code> columns of the <code>df_shipping</code> table:</p><pre><code class="language-julia hljs">df_flow_out = DataFrames.DataFrame(
+    (node = i.origin, product = i.product, x_flow_out = sum(df.x_flow)) for
+    (i, df) in pairs(DataFrames.groupby(df_shipping, [:origin, :product]))
+)</code></pre><div><div style = "float: left;"><span>8×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">node</th><th style = "text-align: left;">product</th><th style = "text-align: left;">x_flow_out</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "AffExpr" style = "text-align: left;">AffExpr</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[1] + x[2] + x[3]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[4] + x[5] + x[6]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[7] + x[8]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[9]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[10] + x[11]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[12] + x[13]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[14] + x[15]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[16]</td></tr></tbody></table></div><p>We can compute an expression for the flow into each node using <code>DataFrames.groupby</code> on the <code>destination</code> and <code>product</code> columns of the <code>df_shipping</code> table:</p><pre><code class="language-julia hljs">df_flow_in = DataFrames.DataFrame(
+    (node = i.destination, product = i.product, x_flow_in = sum(df.x_flow))
+    for (i, df) in
+    pairs(DataFrames.groupby(df_shipping, [:destination, :product]))
+)</code></pre><div><div style = "float: left;"><span>9×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">node</th><th style = "text-align: left;">product</th><th style = "text-align: left;">x_flow_in</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "AffExpr" style = "text-align: left;">AffExpr</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[1] + x[8]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[2] + x[5]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[3]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[4] + x[7] + x[9]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[6]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[10] + x[15]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[11]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[12] + x[14] + x[16]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[13]</td></tr></tbody></table></div><p>We can join the two tables together using <code>DataFrames.outerjoin</code>. We need to use <code>outerjoin</code> here because there might be missing rows.</p><pre><code class="language-julia hljs">df = DataFrames.outerjoin(df_flow_in, df_flow_out; on = [:node, :product])</code></pre><div><div style = "float: left;"><span>10×4 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">node</th><th style = "text-align: left;">product</th><th style = "text-align: left;">x_flow_in</th><th style = "text-align: left;">x_flow_out</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Union{Missing, AffExpr}" style = "text-align: left;">AffExpr?</th><th title = "Union{Missing, AffExpr}" style = "text-align: left;">AffExpr?</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[1] + x[8]</td><td style = "text-align: left;">x[4] + x[5] + x[6]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[2] + x[5]</td><td style = "text-align: left;">x[7] + x[8]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[3]</td><td style = "text-align: left;">x[9]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[4] + x[7] + x[9]</td><td style = "text-align: left;">x[1] + x[2] + x[3]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[10] + x[15]</td><td style = "text-align: left;">x[12] + x[13]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[11]</td><td style = "text-align: left;">x[16]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[12] + x[14] + x[16]</td><td style = "text-align: left;">x[10] + x[11]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[6]</td><td style = "font-style: italic; text-align: left;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[13]</td><td style = "font-style: italic; text-align: left;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">kiwifruit</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">x[14] + x[15]</td></tr></tbody></table></div><p>Next, we need to join the supply column:</p><pre><code class="language-julia hljs">df = DataFrames.leftjoin(
+    df,
+    DataFrames.select(df_supply, [:origin, :product, :x_supply]);
+    on = [:node =&gt; :origin, :product],
+)</code></pre><div><div style = "float: left;"><span>10×5 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">node</th><th style = "text-align: left;">product</th><th style = "text-align: left;">x_flow_in</th><th style = "text-align: left;">x_flow_out</th><th style = "text-align: left;">x_supply</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Union{Missing, AffExpr}" style = "text-align: left;">AffExpr?</th><th title = "Union{Missing, AffExpr}" style = "text-align: left;">AffExpr?</th><th title = "Union{Missing, VariableRef}" style = "text-align: left;">GenericV…?</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[1] + x[8]</td><td style = "text-align: left;">x[4] + x[5] + x[6]</td><td style = "text-align: left;">s[1]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[2] + x[5]</td><td style = "text-align: left;">x[7] + x[8]</td><td style = "text-align: left;">s[2]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[3]</td><td style = "text-align: left;">x[9]</td><td style = "text-align: left;">s[4]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">kiwifruit</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">x[14] + x[15]</td><td style = "text-align: left;">s[3]</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[4] + x[7] + x[9]</td><td style = "text-align: left;">x[1] + x[2] + x[3]</td><td style = "font-style: italic; text-align: left;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[10] + x[15]</td><td style = "text-align: left;">x[12] + x[13]</td><td style = "font-style: italic; text-align: left;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[11]</td><td style = "text-align: left;">x[16]</td><td style = "font-style: italic; text-align: left;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[12] + x[14] + x[16]</td><td style = "text-align: left;">x[10] + x[11]</td><td style = "font-style: italic; text-align: left;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[6]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[13]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td></tr></tbody></table></div><p>and then the demand column</p><pre><code class="language-julia hljs">df = DataFrames.leftjoin(
+    df,
+    DataFrames.select(df_demand, [:destination, :product, :demand]);
+    on = [:node =&gt; :destination, :product],
+)</code></pre><div><div style = "float: left;"><span>10×6 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">node</th><th style = "text-align: left;">product</th><th style = "text-align: left;">x_flow_in</th><th style = "text-align: left;">x_flow_out</th><th style = "text-align: left;">x_supply</th><th style = "text-align: left;">demand</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Union{Missing, AffExpr}" style = "text-align: left;">AffExpr?</th><th title = "Union{Missing, AffExpr}" style = "text-align: left;">AffExpr?</th><th title = "Union{Missing, VariableRef}" style = "text-align: left;">GenericV…?</th><th title = "Union{Missing, Float64}" style = "text-align: left;">Float64?</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[2] + x[5]</td><td style = "text-align: left;">x[7] + x[8]</td><td style = "text-align: left;">s[2]</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[3]</td><td style = "text-align: left;">x[9]</td><td style = "text-align: left;">s[4]</td><td style = "text-align: right;">4.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">kiwifruit</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: left;">x[14] + x[15]</td><td style = "text-align: left;">s[3]</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[4] + x[7] + x[9]</td><td style = "text-align: left;">x[1] + x[2] + x[3]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: right;">16.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[11]</td><td style = "text-align: left;">x[16]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: right;">4.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[12] + x[14] + x[16]</td><td style = "text-align: left;">x[10] + x[11]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: right;">16.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[6]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[13]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: left;">x[1] + x[8]</td><td style = "text-align: left;">x[4] + x[5] + x[6]</td><td style = "text-align: left;">s[1]</td><td style = "font-style: italic; text-align: right;">missing</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: left;">x[10] + x[15]</td><td style = "text-align: left;">x[12] + x[13]</td><td style = "font-style: italic; text-align: left;">missing</td><td style = "font-style: italic; text-align: right;">missing</td></tr></tbody></table></div><p>Now we&#39;re ready to add our mass balance constraint. Because some rows contain <code>missing</code> values, we need to use <code>coalesce</code> to convert any <code>missing</code> into a numeric value:</p><pre><code class="language-julia hljs">@constraint(
+    model,
+    [r in eachrow(df)],
+    coalesce(r.x_supply, 0.0) + coalesce(r.x_flow_in, 0.0) -
+    coalesce(r.x_flow_out, 0.0) == coalesce(r.demand, 0.0),
+);</code></pre><h2 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h2><p>Finally, we can optimize the model:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : HiGHS
+
+* Status
+  Result count       : 1
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;kHighsModelStatusOptimal&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 1.43228e+02
+  Objective bound    : 0.00000e+00
+  Relative gap       : Inf
+  Dual objective value : 1.43228e+02
+
+* Work counters
+  Solve time (sec)   : 4.01735e-04
+  Simplex iterations : 8
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><p>update the solution in the DataFrames:</p><pre><code class="language-julia hljs">df_shipping.x_flow = value.(df_shipping.x_flow)
+df_supply.x_supply = value.(df_supply.x_supply);</code></pre><p>and display the optimal solution for flows:</p><pre><code class="language-julia hljs">DataFrames.select(
+    filter!(row -&gt; row.x_flow &gt; 0.0, df_shipping),
+    [:origin, :destination, :product, :x_flow],
+)</code></pre><div><div style = "float: left;"><span>9×4 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">origin</th><th style = "text-align: left;">destination</th><th style = "text-align: left;">product</th><th style = "text-align: left;">x_flow</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "String" style = "text-align: left;">String</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">10.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">milk</td><td style = "text-align: right;">4.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">christchurch</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">4.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">auckland</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">20.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">wellington</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">2.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: left;">tauranga</td><td style = "text-align: left;">waikato</td><td style = "text-align: left;">kiwifruit</td><td style = "text-align: right;">22.0</td></tr></tbody></table></div></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../multi/">« The multi-commodity flow problem</a><a class="docs-footer-nextpage" href="../tips_and_tricks/">Tips and tricks »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/multi_objective_examples.jl b/previews/PR3525/tutorials/linear/multi_objective_examples.jl
new file mode 100644
index 00000000000..b4d9c6fa170
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_objective_examples.jl
@@ -0,0 +1,120 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Simple multi-objective examples
+
+# This tutorial contains a number of examples of multi-objective programs from
+# the literature.
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import HiGHS
+import MultiObjectiveAlgorithms as MOA
+
+# ## Bi-objective linear problem
+
+# This example is taken from Example 6.3 (from Steuer, 1985), page 154 of
+# Ehrgott, M. (2005). _Multicriteria Optimization_. Springer, Berlin. The code
+# was adapted from an example in [vOptGeneric](https://github.com/vOptSolver/vOptGeneric.jl)
+# by [@xgandibleux](https://github.com/xgandibleux).
+
+model = Model()
+set_silent(model)
+@variable(model, x1 >= 0)
+@variable(model, 0 <= x2 <= 3)
+@objective(model, Min, [3x1 + x2, -x1 - 2x2])
+@constraint(model, 3x1 - x2 <= 6)
+set_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))
+set_attribute(model, MOA.Algorithm(), MOA.Lexicographic())
+optimize!(model)
+solution_summary(model)
+
+#-
+
+for i in 1:result_count(model)
+    print(i, ": z = ", round.(Int, objective_value(model; result = i)), " | ")
+    println("x = ", value.([x1, x2]; result = i))
+end
+
+# ## Bi-objective linear assignment problem
+
+# This example is taken from Example 9.38 (from Ulungu and Teghem, 1994),
+# page 255 of Ehrgott, M. (2005). _Multicriteria Optimization_. Springer, Berlin.
+# The code was adapted from an example in [vOptGeneric](https://github.com/vOptSolver/vOptGeneric.jl)
+# by [@xgandibleux](https://github.com/xgandibleux).
+
+C1 = [5 1 4 7; 6 2 2 6; 2 8 4 4; 3 5 7 1]
+C2 = [3 6 4 2; 1 3 8 3; 5 2 2 3; 4 2 3 5]
+n = size(C2, 1)
+model = Model()
+set_silent(model)
+@variable(model, x[1:n, 1:n], Bin)
+@objective(model, Min, [sum(C1 .* x), sum(C2 .* x)])
+@constraint(model, [i = 1:n], sum(x[i, :]) == 1)
+@constraint(model, [j = 1:n], sum(x[:, j]) == 1)
+set_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))
+set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
+optimize!(model)
+solution_summary(model)
+
+#-
+
+for i in 1:result_count(model)
+    print(i, ": z = ", round.(Int, objective_value(model; result = i)), " | ")
+    println("x = ", round.(Int, value.(x; result = i)))
+end
+
+# ## Bi-objective shortest path problem
+
+# This example is taken from Exercise 9.5 page 269 of Ehrgott, M. (2005).
+# _Multicriteria Optimization_. Springer, Berlin. The code was adapted from an
+# example in [vOptGeneric](https://github.com/vOptSolver/vOptGeneric.jl) by
+# [@xgandibleux](https://github.com/xgandibleux).
+
+M = 50
+C1 = [
+    M 4 5 M M M
+    M M 2 1 2 7
+    M M M 5 2 M
+    M M 5 M M 3
+    M M M M M 4
+    M M M M M M
+]
+C2 = [
+    M 3 1 M M M
+    M M 1 4 2 2
+    M M M 1 7 M
+    M M 1 M M 2
+    M M M M M 2
+    M M M M M M
+]
+n = size(C2, 1)
+model = Model()
+set_silent(model)
+@variable(model, x[1:n, 1:n], Bin)
+@objective(model, Min, [sum(C1 .* x), sum(C2 .* x)])
+@constraint(model, sum(x[1, :]) == 1)
+@constraint(model, sum(x[:, n]) == 1)
+@constraint(model, [i = 2:n-1], sum(x[i, :]) - sum(x[:, i]) == 0)
+set_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))
+set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
+optimize!(model)
+solution_summary(model)
+
+#-
+
+for i in 1:result_count(model)
+    print(i, ": z = ", round.(Int, objective_value(model; result = i)), " | ")
+    X = round.(Int, value.(x; result = i))
+    print("Path:")
+    for ind in findall(val -> val ≈ 1, X)
+        i, j = ind.I
+        print(" $i->$j")
+    end
+    println()
+end
diff --git a/previews/PR3525/tutorials/linear/multi_objective_examples/index.html b/previews/PR3525/tutorials/linear/multi_objective_examples/index.html
new file mode 100644
index 00000000000..c3d39464147
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_objective_examples/index.html
@@ -0,0 +1,142 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Simple multi-objective examples · JuMP</title><meta name="title" content="Simple multi-objective examples · JuMP"/><meta property="og:title" content="Simple multi-objective examples · JuMP"/><meta property="twitter:title" content="Simple multi-objective examples · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li class="is-active"><a class="tocitem" href>Simple multi-objective examples</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Bi-objective-linear-problem"><span>Bi-objective linear problem</span></a></li><li><a class="tocitem" href="#Bi-objective-linear-assignment-problem"><span>Bi-objective linear assignment problem</span></a></li><li><a class="tocitem" href="#Bi-objective-shortest-path-problem"><span>Bi-objective shortest path problem</span></a></li></ul></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Simple multi-objective examples</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Simple multi-objective examples</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/multi_objective_examples.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Simple-multi-objective-examples"><a class="docs-heading-anchor" href="#Simple-multi-objective-examples">Simple multi-objective examples</a><a id="Simple-multi-objective-examples-1"></a><a class="docs-heading-anchor-permalink" href="#Simple-multi-objective-examples" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../multi_objective_examples.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial contains a number of examples of multi-objective programs from the literature.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+import MultiObjectiveAlgorithms as MOA</code></pre><h2 id="Bi-objective-linear-problem"><a class="docs-heading-anchor" href="#Bi-objective-linear-problem">Bi-objective linear problem</a><a id="Bi-objective-linear-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Bi-objective-linear-problem" title="Permalink"></a></h2><p>This example is taken from Example 6.3 (from Steuer, 1985), page 154 of Ehrgott, M. (2005). <em>Multicriteria Optimization</em>. Springer, Berlin. The code was adapted from an example in <a href="https://github.com/vOptSolver/vOptGeneric.jl">vOptGeneric</a> by <a href="https://github.com/xgandibleux">@xgandibleux</a>.</p><pre><code class="language-julia hljs">model = Model()
+set_silent(model)
+@variable(model, x1 &gt;= 0)
+@variable(model, 0 &lt;= x2 &lt;= 3)
+@objective(model, Min, [3x1 + x2, -x1 - 2x2])
+@constraint(model, 3x1 - x2 &lt;= 6)
+set_optimizer(model, () -&gt; MOA.Optimizer(HiGHS.Optimizer))
+set_attribute(model, MOA.Algorithm(), MOA.Lexicographic())
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : MOA[algorithm=MultiObjectiveAlgorithms.Lexicographic, optimizer=HiGHS]
+
+* Status
+  Result count       : 2
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;Solve complete. Found 2 solution(s)&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : [0.00000e+00,0.00000e+00]
+  Objective bound    : [0.00000e+00,-9.00000e+00]
+  Relative gap       : Inf
+  Dual objective value : -9.00000e+00
+
+* Work counters
+  Solve time (sec)   : 1.08314e-03
+  Simplex iterations : 1
+  Barrier iterations : 0
+  Node count         : -1
+</code></pre><pre><code class="language-julia hljs">for i in 1:result_count(model)
+    print(i, &quot;: z = &quot;, round.(Int, objective_value(model; result = i)), &quot; | &quot;)
+    println(&quot;x = &quot;, value.([x1, x2]; result = i))
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1: z = [0, 0] | x = [0.0, -0.0]
+2: z = [12, -9] | x = [3.0, 3.0]</code></pre><h2 id="Bi-objective-linear-assignment-problem"><a class="docs-heading-anchor" href="#Bi-objective-linear-assignment-problem">Bi-objective linear assignment problem</a><a id="Bi-objective-linear-assignment-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Bi-objective-linear-assignment-problem" title="Permalink"></a></h2><p>This example is taken from Example 9.38 (from Ulungu and Teghem, 1994), page 255 of Ehrgott, M. (2005). <em>Multicriteria Optimization</em>. Springer, Berlin. The code was adapted from an example in <a href="https://github.com/vOptSolver/vOptGeneric.jl">vOptGeneric</a> by <a href="https://github.com/xgandibleux">@xgandibleux</a>.</p><pre><code class="language-julia hljs">C1 = [5 1 4 7; 6 2 2 6; 2 8 4 4; 3 5 7 1]
+C2 = [3 6 4 2; 1 3 8 3; 5 2 2 3; 4 2 3 5]
+n = size(C2, 1)
+model = Model()
+set_silent(model)
+@variable(model, x[1:n, 1:n], Bin)
+@objective(model, Min, [sum(C1 .* x), sum(C2 .* x)])
+@constraint(model, [i = 1:n], sum(x[i, :]) == 1)
+@constraint(model, [j = 1:n], sum(x[:, j]) == 1)
+set_optimizer(model, () -&gt; MOA.Optimizer(HiGHS.Optimizer))
+set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : MOA[algorithm=MultiObjectiveAlgorithms.EpsilonConstraint, optimizer=HiGHS]
+
+* Status
+  Result count       : 6
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;Solve complete. Found 6 solution(s)&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : [6.00000e+00,2.40000e+01]
+  Objective bound    : [6.00000e+00,7.00000e+00]
+  Relative gap       : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 8.38804e-03
+  Simplex iterations : 5
+  Barrier iterations : -1
+  Node count         : 1
+</code></pre><pre><code class="language-julia hljs">for i in 1:result_count(model)
+    print(i, &quot;: z = &quot;, round.(Int, objective_value(model; result = i)), &quot; | &quot;)
+    println(&quot;x = &quot;, round.(Int, value.(x; result = i)))
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1: z = [6, 24] | x = [0 1 0 0; 0 0 1 0; 1 0 0 0; 0 0 0 1]
+2: z = [9, 17] | x = [0 0 1 0; 0 1 0 0; 1 0 0 0; 0 0 0 1]
+3: z = [12, 13] | x = [1 0 0 0; 0 1 0 0; 0 0 1 0; 0 0 0 1]
+4: z = [16, 11] | x = [0 0 0 1; 0 1 0 0; 0 0 1 0; 1 0 0 0]
+5: z = [19, 10] | x = [0 0 1 0; 1 0 0 0; 0 0 0 1; 0 1 0 0]
+6: z = [22, 7] | x = [0 0 0 1; 1 0 0 0; 0 0 1 0; 0 1 0 0]</code></pre><h2 id="Bi-objective-shortest-path-problem"><a class="docs-heading-anchor" href="#Bi-objective-shortest-path-problem">Bi-objective shortest path problem</a><a id="Bi-objective-shortest-path-problem-1"></a><a class="docs-heading-anchor-permalink" href="#Bi-objective-shortest-path-problem" title="Permalink"></a></h2><p>This example is taken from Exercise 9.5 page 269 of Ehrgott, M. (2005). <em>Multicriteria Optimization</em>. Springer, Berlin. The code was adapted from an example in <a href="https://github.com/vOptSolver/vOptGeneric.jl">vOptGeneric</a> by <a href="https://github.com/xgandibleux">@xgandibleux</a>.</p><pre><code class="language-julia hljs">M = 50
+C1 = [
+    M 4 5 M M M
+    M M 2 1 2 7
+    M M M 5 2 M
+    M M 5 M M 3
+    M M M M M 4
+    M M M M M M
+]
+C2 = [
+    M 3 1 M M M
+    M M 1 4 2 2
+    M M M 1 7 M
+    M M 1 M M 2
+    M M M M M 2
+    M M M M M M
+]
+n = size(C2, 1)
+model = Model()
+set_silent(model)
+@variable(model, x[1:n, 1:n], Bin)
+@objective(model, Min, [sum(C1 .* x), sum(C2 .* x)])
+@constraint(model, sum(x[1, :]) == 1)
+@constraint(model, sum(x[:, n]) == 1)
+@constraint(model, [i = 2:n-1], sum(x[i, :]) - sum(x[:, i]) == 0)
+set_optimizer(model, () -&gt; MOA.Optimizer(HiGHS.Optimizer))
+set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : MOA[algorithm=MultiObjectiveAlgorithms.EpsilonConstraint, optimizer=HiGHS]
+
+* Status
+  Result count       : 4
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;Solve complete. Found 4 solution(s)&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : [8.00000e+00,9.00000e+00]
+  Objective bound    : [8.00000e+00,4.00000e+00]
+  Relative gap       : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 4.42290e-03
+  Simplex iterations : 4
+  Barrier iterations : -1
+  Node count         : 1
+</code></pre><pre><code class="language-julia hljs">for i in 1:result_count(model)
+    print(i, &quot;: z = &quot;, round.(Int, objective_value(model; result = i)), &quot; | &quot;)
+    X = round.(Int, value.(x; result = i))
+    print(&quot;Path:&quot;)
+    for ind in findall(val -&gt; val ≈ 1, X)
+        i, j = ind.I
+        print(&quot; $i-&gt;$j&quot;)
+    end
+    println()
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1: z = [8, 9] | Path: 1-&gt;2 2-&gt;4 4-&gt;6
+2: z = [10, 7] | Path: 1-&gt;2 2-&gt;5 5-&gt;6
+3: z = [11, 5] | Path: 1-&gt;2 2-&gt;6
+4: z = [13, 4] | Path: 1-&gt;3 3-&gt;4 4-&gt;6</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../multi_objective_knapsack/">« Multi-objective knapsack</a><a class="docs-footer-nextpage" href="../sudoku/">Sudoku »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/multi_objective_knapsack.jl b/previews/PR3525/tutorials/linear/multi_objective_knapsack.jl
new file mode 100644
index 00000000000..e8f81f92528
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_objective_knapsack.jl
@@ -0,0 +1,185 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Multi-objective knapsack
+
+# The purpose of this tutorial is to demonstrate how to create and solve a
+# multi-objective linear program. In addition, it demonstrates how to work with
+# solvers which return multiple solutions.
+
+# ## Required packages
+
+# This tutorial requires the following packages:
+
+using JuMP
+import HiGHS
+import MultiObjectiveAlgorithms as MOA
+import Plots
+import Test  #hide
+
+# [MultiObjectiveAlgorithms.jl](@ref) is a package which implements a variety of
+# algorithms for solving multi-objective optimization problems. Because it is a
+# long package name, we import it instead as `MOA`.
+
+# ## Formulation
+
+# The [knapsack problem](https://en.wikipedia.org/wiki/Knapsack_problem) is a
+# classic problem in mixed-integer programming. Given a collection of items
+# ``i \in I``, each of which has an associated weight, ``w_i``, and profit,
+# ``p_i``, the knapsack problem determines which profit-maximizing subset of
+# items to pack into a knapsack such that the total weight is less than a
+# capacity ``c``. The mathematical formulation is:
+
+# ```math
+# \begin{aligned}
+# \max & \sum\limits_{i \in I} p_i x_i \\
+# \text{s.t.}\ \ & \sum\limits_{i \in I} w_i x_i \le c\\
+# & x_i \in \{0, 1\} && \forall i \in I
+# \end{aligned}
+# ```
+# where ``x_i`` is ``1`` if we pack item ``i`` into the knapsack and ``0``
+# otherwise.
+
+# For this tutorial, we extend the single-objective knapsack problem by adding
+# another objective: given a desirability rating, ``r_i``, we wish to maximize
+# the total desirability of the items in our knapsack. Thus, our mathematical
+# formulation is now:
+
+# ```math
+# \begin{aligned}
+# \max & \sum\limits_{i \in I} p_i x_i \\
+#      & \sum\limits_{i \in I} r_i x_i \\
+# \text{s.t.}\ \ & \sum\limits_{i \in I} w_i x_i \le c\\
+# & x_i \in \{0, 1\} && \forall i \in I
+# \end{aligned}
+# ```
+
+# ## Data
+
+# The data for this example was taken from [vOptGeneric](https://github.com/vOptSolver/vOptGeneric.jl),
+# and the original author was [@xgandibleux](https://github.com/xgandibleux).
+
+profit = [77, 94, 71, 63, 96, 82, 85, 75, 72, 91, 99, 63, 84, 87, 79, 94, 90]
+desire = [65, 90, 90, 77, 95, 84, 70, 94, 66, 92, 74, 97, 60, 60, 65, 97, 93]
+weight = [80, 87, 68, 72, 66, 77, 99, 85, 70, 93, 98, 72, 100, 89, 67, 86, 91]
+capacity = 900
+N = length(profit)
+
+# Comparing the capacity to the total weight of all the items:
+
+capacity / sum(weight)
+
+# shows that we can take approximately 64% of the items.
+
+# Plotting the items, we see that there are a range of items with different
+# profits and desirability. Some items have a high profit and a high
+# desirability, others have a low profit and a high desirability (and vice
+# versa).
+
+Plots.scatter(
+    profit,
+    desire;
+    xlabel = "Profit",
+    ylabel = "Desire",
+    legend = false,
+)
+
+# The goal of the bi-objective knapsack problem is to choose a subset which
+# maximizes both objectives.
+
+# ## JuMP formulation
+
+# Our JuMP formulation is a direct translation of the mathematical formulation:
+
+model = Model()
+@variable(model, x[1:N], Bin)
+@constraint(model, sum(weight[i] * x[i] for i in 1:N) <= capacity)
+@expression(model, profit_expr, sum(profit[i] * x[i] for i in 1:N))
+@expression(model, desire_expr, sum(desire[i] * x[i] for i in 1:N))
+@objective(model, Max, [profit_expr, desire_expr])
+
+# Note how we form a multi-objective program by passing a vector of scalar
+# objective functions.
+
+# ## Solution
+
+# To solve our model, we need an optimizer which supports multi-objective linear
+# programs. One option is to use the [MultiObjectiveAlgorithms.jl](@ref)
+# package.
+
+set_optimizer(model, () -> MOA.Optimizer(HiGHS.Optimizer))
+set_silent(model)
+
+# MultiObjectiveAlgorithms.jl supports many different algorithms for solving
+# multiobjective optimization problems. One option is the epsilon-constraint
+# method:
+
+set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
+
+# Let's solve the problem and see the solution
+
+optimize!(model)
+solution_summary(model)
+
+# There are 9 solutions available. We can also use [`result_count`](@ref) to see
+# how many solutions are available:
+
+result_count(model)
+
+# ## Accessing multiple solutions
+
+# Access the nine different solutions in the model using the `result` keyword to
+# [`solution_summary`](@ref), [`value`](@ref), and [`objective_value`](@ref):
+
+solution_summary(model; result = 5)
+
+#-
+
+objective_value(model; result = 5)
+
+# Note that because we set a vector of two objective functions, the objective
+# value is a vector with two elements. We can also query the value of each
+# objective separately:
+
+value(profit_expr; result = 5)
+
+# ## Visualizing objective space
+
+# Unlike single-objective optimization problems, multi-objective optimization
+# problems do not have a single optimal solution. Instead, the solutions
+# returned represent possible trade-offs that the decision maker can choose
+# between the two objectives. A common way to visualize this is by plotting
+# the objective values of each of the solutions:
+
+plot = Plots.scatter(
+    [value(profit_expr; result = i) for i in 1:result_count(model)],
+    [value(desire_expr; result = i) for i in 1:result_count(model)];
+    xlabel = "Profit",
+    ylabel = "Desire",
+    title = "Objective space",
+    label = "",
+    xlims = (915, 960),
+)
+for i in 1:result_count(model)
+    y = objective_value(model; result = i)
+    Plots.annotate!(y[1] - 1, y[2], (i, 10))
+end
+ideal_point = objective_bound(model)
+Plots.scatter!([ideal_point[1]], [ideal_point[2]]; label = "Ideal point")
+
+# Visualizing the objective space lets the decision maker choose a solution that
+# suits their personal preferences. For example, result `#7` is close to the
+# maximum value of profit, but offers significantly higher desirability compared
+# with solutions `#8` and `#9`.
+
+# The set of items that are chosen in solution `#7` are:
+
+items_chosen = [i for i in 1:N if value(x[i]; result = 7) > 0.9]
+
+# ## Next steps
+
+# [MultiObjectiveAlgorithms.jl](@ref) implements a number of different
+# algorithms. Try solving the same problem using `MOA.Dichotomy()`. Does it find
+# the same solution?
diff --git a/previews/PR3525/tutorials/linear/multi_objective_knapsack/0a06bb1f.svg b/previews/PR3525/tutorials/linear/multi_objective_knapsack/0a06bb1f.svg
new file mode 100644
index 00000000000..e5f4f44b28b
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_objective_knapsack/0a06bb1f.svg
@@ -0,0 +1,55 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip390">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip390)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip391">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip390)" d="M204.45 1423.18 L2352.76 1423.18 L2352.76 47.2441 L204.45 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip392">
+    <rect x="204" y="47" width="2149" height="1377"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="659.332,1423.18 659.332,47.2441 "/>
+<polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1222.31,1423.18 1222.31,47.2441 "/>
+<polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1785.28,1423.18 1785.28,47.2441 "/>
+<polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2348.25,1423.18 2348.25,47.2441 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="204.45,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="659.332,1423.18 659.332,1404.28 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1222.31,1423.18 1222.31,1404.28 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1785.28,1423.18 1785.28,1404.28 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2348.25,1423.18 2348.25,1404.28 "/>
+<path clip-path="url(#clip390)" d="M632.689 1451.02 L654.911 1451.02 L654.911 1453.01 L642.365 1485.58 L637.48 1485.58 L649.286 1454.96 L632.689 1454.96 L632.689 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M674.031 1454.1 Q670.42 1454.1 668.591 1457.66 Q666.786 1461.2 666.786 1468.33 Q666.786 1475.44 668.591 1479.01 Q670.42 1482.55 674.031 1482.55 Q677.665 1482.55 679.471 1479.01 Q681.3 1475.44 681.3 1468.33 Q681.3 1461.2 679.471 1457.66 Q677.665 1454.1 674.031 1454.1 M674.031 1450.39 Q679.841 1450.39 682.897 1455 Q685.976 1459.58 685.976 1468.33 Q685.976 1477.06 682.897 1481.67 Q679.841 1486.25 674.031 1486.25 Q668.221 1486.25 665.142 1481.67 Q662.087 1477.06 662.087 1468.33 Q662.087 1459.58 665.142 1455 Q668.221 1450.39 674.031 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1207.18 1469.17 Q1203.84 1469.17 1201.92 1470.95 Q1200.03 1472.73 1200.03 1475.86 Q1200.03 1478.98 1201.92 1480.77 Q1203.84 1482.55 1207.18 1482.55 Q1210.51 1482.55 1212.43 1480.77 Q1214.35 1478.96 1214.35 1475.86 Q1214.35 1472.73 1212.43 1470.95 Q1210.53 1469.17 1207.18 1469.17 M1202.5 1467.18 Q1199.49 1466.44 1197.8 1464.38 Q1196.14 1462.32 1196.14 1459.35 Q1196.14 1455.21 1199.08 1452.8 Q1202.04 1450.39 1207.18 1450.39 Q1212.34 1450.39 1215.28 1452.8 Q1218.22 1455.21 1218.22 1459.35 Q1218.22 1462.32 1216.53 1464.38 Q1214.86 1466.44 1211.88 1467.18 Q1215.26 1467.96 1217.13 1470.26 Q1219.03 1472.55 1219.03 1475.86 Q1219.03 1480.88 1215.95 1483.57 Q1212.9 1486.25 1207.18 1486.25 Q1201.46 1486.25 1198.38 1483.57 Q1195.33 1480.88 1195.33 1475.86 Q1195.33 1472.55 1197.22 1470.26 Q1199.12 1467.96 1202.5 1467.18 M1200.79 1459.79 Q1200.79 1462.48 1202.46 1463.98 Q1204.15 1465.49 1207.18 1465.49 Q1210.19 1465.49 1211.88 1463.98 Q1213.59 1462.48 1213.59 1459.79 Q1213.59 1457.11 1211.88 1455.6 Q1210.19 1454.1 1207.18 1454.1 Q1204.15 1454.1 1202.46 1455.6 Q1200.79 1457.11 1200.79 1459.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1237.34 1454.1 Q1233.73 1454.1 1231.9 1457.66 Q1230.09 1461.2 1230.09 1468.33 Q1230.09 1475.44 1231.9 1479.01 Q1233.73 1482.55 1237.34 1482.55 Q1240.97 1482.55 1242.78 1479.01 Q1244.61 1475.44 1244.61 1468.33 Q1244.61 1461.2 1242.78 1457.66 Q1240.97 1454.1 1237.34 1454.1 M1237.34 1450.39 Q1243.15 1450.39 1246.21 1455 Q1249.28 1459.58 1249.28 1468.33 Q1249.28 1477.06 1246.21 1481.67 Q1243.15 1486.25 1237.34 1486.25 Q1231.53 1486.25 1228.45 1481.67 Q1225.4 1477.06 1225.4 1468.33 Q1225.4 1459.58 1228.45 1455 Q1231.53 1450.39 1237.34 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1760.41 1484.86 L1760.41 1480.6 Q1762.17 1481.44 1763.97 1481.88 Q1765.78 1482.32 1767.51 1482.32 Q1772.14 1482.32 1774.57 1479.21 Q1777.03 1476.09 1777.37 1469.75 Q1776.03 1471.74 1773.97 1472.8 Q1771.91 1473.87 1769.41 1473.87 Q1764.23 1473.87 1761.19 1470.74 Q1758.18 1467.59 1758.18 1462.15 Q1758.18 1456.83 1761.33 1453.61 Q1764.48 1450.39 1769.71 1450.39 Q1775.71 1450.39 1778.86 1455 Q1782.03 1459.58 1782.03 1468.33 Q1782.03 1476.51 1778.14 1481.39 Q1774.27 1486.25 1767.72 1486.25 Q1765.96 1486.25 1764.16 1485.9 Q1762.35 1485.56 1760.41 1484.86 M1769.71 1470.21 Q1772.86 1470.21 1774.69 1468.06 Q1776.54 1465.9 1776.54 1462.15 Q1776.54 1458.43 1774.69 1456.27 Q1772.86 1454.1 1769.71 1454.1 Q1766.56 1454.1 1764.71 1456.27 Q1762.88 1458.43 1762.88 1462.15 Q1762.88 1465.9 1764.71 1468.06 Q1766.56 1470.21 1769.71 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1800.43 1454.1 Q1796.82 1454.1 1794.99 1457.66 Q1793.18 1461.2 1793.18 1468.33 Q1793.18 1475.44 1794.99 1479.01 Q1796.82 1482.55 1800.43 1482.55 Q1804.06 1482.55 1805.87 1479.01 Q1807.7 1475.44 1807.7 1468.33 Q1807.7 1461.2 1805.87 1457.66 Q1804.06 1454.1 1800.43 1454.1 M1800.43 1450.39 Q1806.24 1450.39 1809.29 1455 Q1812.37 1459.58 1812.37 1468.33 Q1812.37 1477.06 1809.29 1481.67 Q1806.24 1486.25 1800.43 1486.25 Q1794.62 1486.25 1791.54 1481.67 Q1788.48 1477.06 1788.48 1468.33 Q1788.48 1459.58 1791.54 1455 Q1794.62 1450.39 1800.43 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M2307.86 1481.64 L2315.5 1481.64 L2315.5 1455.28 L2307.19 1456.95 L2307.19 1452.69 L2315.45 1451.02 L2320.13 1451.02 L2320.13 1481.64 L2327.77 1481.64 L2327.77 1485.58 L2307.86 1485.58 L2307.86 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M2347.21 1454.1 Q2343.6 1454.1 2341.77 1457.66 Q2339.97 1461.2 2339.97 1468.33 Q2339.97 1475.44 2341.77 1479.01 Q2343.6 1482.55 2347.21 1482.55 Q2350.84 1482.55 2352.65 1479.01 Q2354.48 1475.44 2354.48 1468.33 Q2354.48 1461.2 2352.65 1457.66 Q2350.84 1454.1 2347.21 1454.1 M2347.21 1450.39 Q2353.02 1450.39 2356.08 1455 Q2359.15 1459.58 2359.15 1468.33 Q2359.15 1477.06 2356.08 1481.67 Q2353.02 1486.25 2347.21 1486.25 Q2341.4 1486.25 2338.32 1481.67 Q2335.27 1477.06 2335.27 1468.33 Q2335.27 1459.58 2338.32 1455 Q2341.4 1450.39 2347.21 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M2377.37 1454.1 Q2373.76 1454.1 2371.93 1457.66 Q2370.13 1461.2 2370.13 1468.33 Q2370.13 1475.44 2371.93 1479.01 Q2373.76 1482.55 2377.37 1482.55 Q2381.01 1482.55 2382.81 1479.01 Q2384.64 1475.44 2384.64 1468.33 Q2384.64 1461.2 2382.81 1457.66 Q2381.01 1454.1 2377.37 1454.1 M2377.37 1450.39 Q2383.18 1450.39 2386.24 1455 Q2389.32 1459.58 2389.32 1468.33 Q2389.32 1477.06 2386.24 1481.67 Q2383.18 1486.25 2377.37 1486.25 Q2371.56 1486.25 2368.48 1481.67 Q2365.43 1477.06 2365.43 1468.33 Q2365.43 1459.58 2368.48 1455 Q2371.56 1450.39 2377.37 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1204 1525.81 L1204 1543.66 L1212.08 1543.66 Q1216.57 1543.66 1219.02 1541.34 Q1221.47 1539.02 1221.47 1534.72 Q1221.47 1530.45 1219.02 1528.13 Q1216.57 1525.81 1212.08 1525.81 L1204 1525.81 M1197.57 1520.52 L1212.08 1520.52 Q1220.07 1520.52 1224.14 1524.15 Q1228.25 1527.75 1228.25 1534.72 Q1228.25 1541.75 1224.14 1545.35 Q1220.07 1548.95 1212.08 1548.95 L1204 1548.95 L1204 1568.04 L1197.57 1568.04 L1197.57 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1256.13 1537.87 Q1255.15 1537.3 1253.97 1537.04 Q1252.82 1536.76 1251.42 1536.76 Q1246.46 1536.76 1243.78 1540 Q1241.14 1543.22 1241.14 1549.27 L1241.14 1568.04 L1235.25 1568.04 L1235.25 1532.4 L1241.14 1532.4 L1241.14 1537.93 Q1242.99 1534.69 1245.95 1533.13 Q1248.91 1531.54 1253.14 1531.54 Q1253.74 1531.54 1254.48 1531.63 Q1255.21 1531.7 1256.1 1531.85 L1256.13 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1274.66 1536.5 Q1269.95 1536.5 1267.21 1540.19 Q1264.47 1543.85 1264.47 1550.25 Q1264.47 1556.65 1267.18 1560.34 Q1269.91 1564 1274.66 1564 Q1279.33 1564 1282.07 1560.31 Q1284.81 1556.62 1284.81 1550.25 Q1284.81 1543.92 1282.07 1540.23 Q1279.33 1536.5 1274.66 1536.5 M1274.66 1531.54 Q1282.29 1531.54 1286.66 1536.5 Q1291.02 1541.47 1291.02 1550.25 Q1291.02 1559 1286.66 1564 Q1282.29 1568.97 1274.66 1568.97 Q1266.99 1568.97 1262.62 1564 Q1258.3 1559 1258.3 1550.25 Q1258.3 1541.47 1262.62 1536.5 Q1266.99 1531.54 1274.66 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1318.77 1518.52 L1318.77 1523.39 L1313.17 1523.39 Q1310.02 1523.39 1308.78 1524.66 Q1307.57 1525.93 1307.57 1529.24 L1307.57 1532.4 L1317.21 1532.4 L1317.21 1536.95 L1307.57 1536.95 L1307.57 1568.04 L1301.68 1568.04 L1301.68 1536.95 L1296.08 1536.95 L1296.08 1532.4 L1301.68 1532.4 L1301.68 1529.91 Q1301.68 1523.96 1304.45 1521.26 Q1307.22 1518.52 1313.23 1518.52 L1318.77 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1323.67 1532.4 L1329.53 1532.4 L1329.53 1568.04 L1323.67 1568.04 L1323.67 1532.4 M1323.67 1518.52 L1329.53 1518.52 L1329.53 1525.93 L1323.67 1525.93 L1323.67 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M1347.58 1522.27 L1347.58 1532.4 L1359.64 1532.4 L1359.64 1536.95 L1347.58 1536.95 L1347.58 1556.3 Q1347.58 1560.66 1348.75 1561.9 Q1349.96 1563.14 1353.62 1563.14 L1359.64 1563.14 L1359.64 1568.04 L1353.62 1568.04 Q1346.84 1568.04 1344.27 1565.53 Q1341.69 1562.98 1341.69 1556.3 L1341.69 1536.95 L1337.39 1536.95 L1337.39 1532.4 L1341.69 1532.4 L1341.69 1522.27 L1347.58 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="204.45,1384.24 2352.76,1384.24 "/>
+<polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="204.45,1033.41 2352.76,1033.41 "/>
+<polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="204.45,682.588 2352.76,682.588 "/>
+<polyline clip-path="url(#clip392)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="204.45,331.763 2352.76,331.763 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="204.45,1423.18 204.45,47.2441 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="204.45,1384.24 223.347,1384.24 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="204.45,1033.41 223.347,1033.41 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="204.45,682.588 223.347,682.588 "/>
+<polyline clip-path="url(#clip390)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="204.45,331.763 223.347,331.763 "/>
+<path clip-path="url(#clip390)" d="M126.922 1382.37 Q123.774 1382.37 121.922 1384.53 Q120.093 1386.68 120.093 1390.43 Q120.093 1394.16 121.922 1396.33 Q123.774 1398.49 126.922 1398.49 Q130.07 1398.49 131.899 1396.33 Q133.751 1394.16 133.751 1390.43 Q133.751 1386.68 131.899 1384.53 Q130.07 1382.37 126.922 1382.37 M136.204 1367.72 L136.204 1371.98 Q134.445 1371.15 132.64 1370.71 Q130.857 1370.27 129.098 1370.27 Q124.468 1370.27 122.015 1373.39 Q119.584 1376.52 119.237 1382.84 Q120.603 1380.82 122.663 1379.76 Q124.723 1378.67 127.2 1378.67 Q132.408 1378.67 135.417 1381.84 Q138.45 1384.99 138.45 1390.43 Q138.45 1395.75 135.302 1398.97 Q132.154 1402.19 126.922 1402.19 Q120.927 1402.19 117.756 1397.61 Q114.584 1393 114.584 1384.27 Q114.584 1376.08 118.473 1371.22 Q122.362 1366.33 128.913 1366.33 Q130.672 1366.33 132.455 1366.68 Q134.26 1367.03 136.204 1367.72 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M156.505 1370.04 Q152.894 1370.04 151.066 1373.6 Q149.26 1377.14 149.26 1384.27 Q149.26 1391.38 151.066 1394.94 Q152.894 1398.49 156.505 1398.49 Q160.14 1398.49 161.945 1394.94 Q163.774 1391.38 163.774 1384.27 Q163.774 1377.14 161.945 1373.6 Q160.14 1370.04 156.505 1370.04 M156.505 1366.33 Q162.315 1366.33 165.371 1370.94 Q168.45 1375.52 168.45 1384.27 Q168.45 1393 165.371 1397.61 Q162.315 1402.19 156.505 1402.19 Q150.695 1402.19 147.616 1397.61 Q144.561 1393 144.561 1384.27 Q144.561 1375.52 147.616 1370.94 Q150.695 1366.33 156.505 1366.33 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M115.163 1016.13 L137.385 1016.13 L137.385 1018.12 L124.839 1050.69 L119.955 1050.69 L131.76 1020.07 L115.163 1020.07 L115.163 1016.13 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M156.505 1019.21 Q152.894 1019.21 151.066 1022.78 Q149.26 1026.32 149.26 1033.45 Q149.26 1040.55 151.066 1044.12 Q152.894 1047.66 156.505 1047.66 Q160.14 1047.66 161.945 1044.12 Q163.774 1040.55 163.774 1033.45 Q163.774 1026.32 161.945 1022.78 Q160.14 1019.21 156.505 1019.21 M156.505 1015.51 Q162.315 1015.51 165.371 1020.11 Q168.45 1024.7 168.45 1033.45 Q168.45 1042.17 165.371 1046.78 Q162.315 1051.36 156.505 1051.36 Q150.695 1051.36 147.616 1046.78 Q144.561 1042.17 144.561 1033.45 Q144.561 1024.7 147.616 1020.11 Q150.695 1015.51 156.505 1015.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M126.343 683.456 Q123.01 683.456 121.089 685.239 Q119.191 687.021 119.191 690.146 Q119.191 693.271 121.089 695.053 Q123.01 696.836 126.343 696.836 Q129.677 696.836 131.598 695.053 Q133.519 693.248 133.519 690.146 Q133.519 687.021 131.598 685.239 Q129.7 683.456 126.343 683.456 M121.668 681.466 Q118.658 680.725 116.968 678.665 Q115.302 676.604 115.302 673.642 Q115.302 669.498 118.242 667.091 Q121.205 664.683 126.343 664.683 Q131.505 664.683 134.445 667.091 Q137.385 669.498 137.385 673.642 Q137.385 676.604 135.695 678.665 Q134.029 680.725 131.042 681.466 Q134.422 682.253 136.297 684.544 Q138.195 686.836 138.195 690.146 Q138.195 695.169 135.117 697.854 Q132.061 700.54 126.343 700.54 Q120.626 700.54 117.547 697.854 Q114.492 695.169 114.492 690.146 Q114.492 686.836 116.39 684.544 Q118.288 682.253 121.668 681.466 M119.955 674.081 Q119.955 676.766 121.621 678.271 Q123.311 679.776 126.343 679.776 Q129.353 679.776 131.042 678.271 Q132.755 676.766 132.755 674.081 Q132.755 671.396 131.042 669.892 Q129.353 668.387 126.343 668.387 Q123.311 668.387 121.621 669.892 Q119.955 671.396 119.955 674.081 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M156.505 668.387 Q152.894 668.387 151.066 671.952 Q149.26 675.493 149.26 682.623 Q149.26 689.729 151.066 693.294 Q152.894 696.836 156.505 696.836 Q160.14 696.836 161.945 693.294 Q163.774 689.729 163.774 682.623 Q163.774 675.493 161.945 671.952 Q160.14 668.387 156.505 668.387 M156.505 664.683 Q162.315 664.683 165.371 669.29 Q168.45 673.873 168.45 682.623 Q168.45 691.35 165.371 695.956 Q162.315 700.54 156.505 700.54 Q150.695 700.54 147.616 695.956 Q144.561 691.35 144.561 682.623 Q144.561 673.873 147.616 669.29 Q150.695 664.683 156.505 664.683 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M116.482 348.326 L116.482 344.066 Q118.242 344.9 120.047 345.34 Q121.853 345.779 123.589 345.779 Q128.218 345.779 130.649 342.677 Q133.103 339.552 133.45 333.21 Q132.107 335.201 130.047 336.265 Q127.987 337.33 125.487 337.33 Q120.302 337.33 117.269 334.205 Q114.26 331.057 114.26 325.617 Q114.26 320.293 117.408 317.076 Q120.556 313.858 125.788 313.858 Q131.783 313.858 134.931 318.465 Q138.103 323.048 138.103 331.798 Q138.103 339.969 134.214 344.853 Q130.348 349.714 123.797 349.714 Q122.038 349.714 120.232 349.367 Q118.427 349.02 116.482 348.326 M125.788 333.673 Q128.936 333.673 130.765 331.52 Q132.617 329.367 132.617 325.617 Q132.617 321.891 130.765 319.738 Q128.936 317.562 125.788 317.562 Q122.64 317.562 120.788 319.738 Q118.959 321.891 118.959 325.617 Q118.959 329.367 120.788 331.52 Q122.64 333.673 125.788 333.673 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M156.505 317.562 Q152.894 317.562 151.066 321.127 Q149.26 324.668 149.26 331.798 Q149.26 338.904 151.066 342.469 Q152.894 346.011 156.505 346.011 Q160.14 346.011 161.945 342.469 Q163.774 338.904 163.774 331.798 Q163.774 324.668 161.945 321.127 Q160.14 317.562 156.505 317.562 M156.505 313.858 Q162.315 313.858 165.371 318.465 Q168.45 323.048 168.45 331.798 Q168.45 340.525 165.371 345.131 Q162.315 349.714 156.505 349.714 Q150.695 349.714 147.616 345.131 Q144.561 340.525 144.561 331.798 Q144.561 323.048 147.616 318.465 Q150.695 313.858 156.505 313.858 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M21.7677 827.769 L58.7206 827.769 L58.7206 820.003 Q58.7206 810.168 54.2646 805.617 Q49.8086 801.033 40.1964 801.033 Q30.6479 801.033 26.2237 805.617 Q21.7677 810.168 21.7677 820.003 L21.7677 827.769 M16.4842 834.199 L16.4842 820.99 Q16.4842 807.176 22.2451 800.715 Q27.9743 794.254 40.1964 794.254 Q52.4822 794.254 58.2432 800.747 Q64.0042 807.24 64.0042 820.99 L64.0042 834.199 L16.4842 834.199 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M44.7161 753.768 L47.5806 753.768 L47.5806 780.695 Q53.6281 780.313 56.8109 777.066 Q59.9619 773.788 59.9619 767.964 Q59.9619 764.59 59.1344 761.439 Q58.3069 758.256 56.6518 755.137 L62.1899 755.137 Q63.5267 758.288 64.227 761.598 Q64.9272 764.908 64.9272 768.314 Q64.9272 776.844 59.9619 781.841 Q54.9967 786.806 46.5303 786.806 Q37.7774 786.806 32.6531 782.095 Q27.4968 777.353 27.4968 769.332 Q27.4968 762.139 32.1438 757.969 Q36.7589 753.768 44.7161 753.768 M42.9973 759.624 Q38.1912 759.688 35.3266 762.33 Q32.4621 764.94 32.4621 769.268 Q32.4621 774.17 35.2312 777.13 Q38.0002 780.058 43.0292 780.504 L42.9973 759.624 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M29.4065 721.43 L34.9447 721.43 Q33.6716 723.913 33.035 726.586 Q32.3984 729.26 32.3984 732.125 Q32.3984 736.485 33.7352 738.681 Q35.072 740.846 37.7456 740.846 Q39.7826 740.846 40.9603 739.286 Q42.1061 737.726 43.1565 733.016 L43.6021 731.011 Q44.9389 724.772 47.3897 722.162 Q49.8086 719.521 54.1691 719.521 Q59.1344 719.521 62.0308 723.467 Q64.9272 727.382 64.9272 734.257 Q64.9272 737.122 64.3543 740.241 Q63.8132 743.328 62.6992 746.766 L56.6518 746.766 Q58.3387 743.519 59.198 740.368 Q60.0256 737.217 60.0256 734.13 Q60.0256 729.992 58.6251 727.764 Q57.1929 725.536 54.6147 725.536 Q52.2276 725.536 50.9545 727.159 Q49.6813 728.751 48.5037 734.193 L48.0262 736.23 Q46.8804 741.673 44.5251 744.092 Q42.138 746.511 38.0002 746.511 Q32.9713 746.511 30.2341 742.946 Q27.4968 739.382 27.4968 732.825 Q27.4968 729.578 27.9743 726.714 Q28.4517 723.849 29.4065 721.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M28.3562 710.195 L28.3562 704.338 L64.0042 704.338 L64.0042 710.195 L28.3562 710.195 M14.479 710.195 L14.479 704.338 L21.895 704.338 L21.895 710.195 L14.479 710.195 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M33.8307 671.428 Q33.2578 672.414 33.0032 673.592 Q32.7167 674.738 32.7167 676.138 Q32.7167 681.103 35.9632 683.777 Q39.1779 686.419 45.2253 686.419 L64.0042 686.419 L64.0042 692.307 L28.3562 692.307 L28.3562 686.419 L33.8944 686.419 Q30.6479 684.573 29.0883 681.613 Q27.4968 678.653 27.4968 674.419 Q27.4968 673.815 27.5923 673.083 Q27.656 672.351 27.8151 671.459 L33.8307 671.428 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip390)" d="M44.7161 636.225 L47.5806 636.225 L47.5806 663.152 Q53.6281 662.77 56.8109 659.524 Q59.9619 656.245 59.9619 650.421 Q59.9619 647.047 59.1344 643.896 Q58.3069 640.713 56.6518 637.594 L62.1899 637.594 Q63.5267 640.745 64.227 644.055 Q64.9272 647.365 64.9272 650.771 Q64.9272 659.301 59.9619 664.298 Q54.9967 669.263 46.5303 669.263 Q37.7774 669.263 32.6531 664.553 Q27.4968 659.81 27.4968 651.789 Q27.4968 644.596 32.1438 640.427 Q36.7589 636.225 44.7161 636.225 M42.9973 642.082 Q38.1912 642.145 35.3266 644.787 Q32.4621 647.397 32.4621 651.726 Q32.4621 656.627 35.2312 659.587 Q38.0002 662.516 43.0292 662.961 L42.9973 642.082 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip392)" cx="1053.41" cy="1208.83" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="2010.47" cy="331.763" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="715.629" cy="331.763" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="265.251" cy="787.836" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="2123.06" cy="156.351" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="1334.9" cy="542.258" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="1503.79" cy="1033.41" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="940.819" cy="191.433" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="771.927" cy="1173.74" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="1841.58" cy="261.598" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="2291.95" cy="893.083" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="265.251" cy="86.1857" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="1447.49" cy="1384.24" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="1616.39" cy="1384.24" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="1166.01" cy="1208.83" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="2010.47" cy="86.1857" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip392)" cx="1785.28" cy="226.516" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+</svg>
diff --git a/previews/PR3525/tutorials/linear/multi_objective_knapsack/144638a3.svg b/previews/PR3525/tutorials/linear/multi_objective_knapsack/144638a3.svg
new file mode 100644
index 00000000000..ccb5f3922b6
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_objective_knapsack/144638a3.svg
@@ -0,0 +1,53 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip450">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip450)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip451">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip450)" d="M234.612 1423.18 L2352.76 1423.18 L2352.76 123.472 L234.612 123.472  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip452">
+    <rect x="234" y="123" width="2119" height="1301"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="469.961,1423.18 469.961,123.472 "/>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="940.66,1423.18 940.66,123.472 "/>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1411.36,1423.18 1411.36,123.472 "/>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1882.06,1423.18 1882.06,123.472 "/>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2352.76,1423.18 2352.76,123.472 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.612,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="469.961,1423.18 469.961,1404.28 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="940.66,1423.18 940.66,1404.28 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1411.36,1423.18 1411.36,1404.28 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1882.06,1423.18 1882.06,1404.28 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2352.76,1423.18 2352.76,1404.28 "/>
+<path clip-path="url(#clip450)" d="M430.007 1484.86 L430.007 1480.6 Q431.767 1481.44 433.572 1481.88 Q435.378 1482.32 437.114 1482.32 Q441.743 1482.32 444.174 1479.21 Q446.628 1476.09 446.975 1469.75 Q445.632 1471.74 443.572 1472.8 Q441.512 1473.87 439.012 1473.87 Q433.827 1473.87 430.794 1470.74 Q427.785 1467.59 427.785 1462.15 Q427.785 1456.83 430.933 1453.61 Q434.081 1450.39 439.313 1450.39 Q445.308 1450.39 448.456 1455 Q451.628 1459.58 451.628 1468.33 Q451.628 1476.51 447.739 1481.39 Q443.873 1486.25 437.322 1486.25 Q435.563 1486.25 433.757 1485.9 Q431.952 1485.56 430.007 1484.86 M439.313 1470.21 Q442.461 1470.21 444.29 1468.06 Q446.142 1465.9 446.142 1462.15 Q446.142 1458.43 444.29 1456.27 Q442.461 1454.1 439.313 1454.1 Q436.165 1454.1 434.313 1456.27 Q432.484 1458.43 432.484 1462.15 Q432.484 1465.9 434.313 1468.06 Q436.165 1470.21 439.313 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M464.058 1481.64 L480.378 1481.64 L480.378 1485.58 L458.433 1485.58 L458.433 1481.64 Q461.095 1478.89 465.679 1474.26 Q470.285 1469.61 471.466 1468.27 Q473.711 1465.74 474.591 1464.01 Q475.493 1462.25 475.493 1460.56 Q475.493 1457.8 473.549 1456.07 Q471.628 1454.33 468.526 1454.33 Q466.327 1454.33 463.873 1455.09 Q461.442 1455.86 458.665 1457.41 L458.665 1452.69 Q461.489 1451.55 463.942 1450.97 Q466.396 1450.39 468.433 1450.39 Q473.804 1450.39 476.998 1453.08 Q480.192 1455.77 480.192 1460.26 Q480.192 1462.39 479.382 1464.31 Q478.595 1466.2 476.489 1468.8 Q475.91 1469.47 472.808 1472.69 Q469.706 1475.88 464.058 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M500.192 1454.1 Q496.581 1454.1 494.752 1457.66 Q492.947 1461.2 492.947 1468.33 Q492.947 1475.44 494.752 1479.01 Q496.581 1482.55 500.192 1482.55 Q503.826 1482.55 505.632 1479.01 Q507.461 1475.44 507.461 1468.33 Q507.461 1461.2 505.632 1457.66 Q503.826 1454.1 500.192 1454.1 M500.192 1450.39 Q506.002 1450.39 509.058 1455 Q512.137 1459.58 512.137 1468.33 Q512.137 1477.06 509.058 1481.67 Q506.002 1486.25 500.192 1486.25 Q494.382 1486.25 491.303 1481.67 Q488.248 1477.06 488.248 1468.33 Q488.248 1459.58 491.303 1455 Q494.382 1450.39 500.192 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M900.706 1484.86 L900.706 1480.6 Q902.465 1481.44 904.271 1481.88 Q906.077 1482.32 907.813 1482.32 Q912.442 1482.32 914.873 1479.21 Q917.326 1476.09 917.674 1469.75 Q916.331 1471.74 914.271 1472.8 Q912.211 1473.87 909.711 1473.87 Q904.526 1473.87 901.493 1470.74 Q898.484 1467.59 898.484 1462.15 Q898.484 1456.83 901.632 1453.61 Q904.78 1450.39 910.012 1450.39 Q916.007 1450.39 919.155 1455 Q922.326 1459.58 922.326 1468.33 Q922.326 1476.51 918.438 1481.39 Q914.572 1486.25 908.021 1486.25 Q906.262 1486.25 904.456 1485.9 Q902.651 1485.56 900.706 1484.86 M910.012 1470.21 Q913.16 1470.21 914.989 1468.06 Q916.84 1465.9 916.84 1462.15 Q916.84 1458.43 914.989 1456.27 Q913.16 1454.1 910.012 1454.1 Q906.864 1454.1 905.012 1456.27 Q903.183 1458.43 903.183 1462.15 Q903.183 1465.9 905.012 1468.06 Q906.864 1470.21 910.012 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M944.896 1466.95 Q948.252 1467.66 950.127 1469.93 Q952.025 1472.2 952.025 1475.53 Q952.025 1480.65 948.507 1483.45 Q944.988 1486.25 938.507 1486.25 Q936.331 1486.25 934.016 1485.81 Q931.725 1485.39 929.271 1484.54 L929.271 1480.02 Q931.215 1481.16 933.53 1481.74 Q935.845 1482.32 938.368 1482.32 Q942.766 1482.32 945.058 1480.58 Q947.373 1478.84 947.373 1475.53 Q947.373 1472.48 945.22 1470.77 Q943.09 1469.03 939.271 1469.03 L935.243 1469.03 L935.243 1465.19 L939.456 1465.19 Q942.905 1465.19 944.734 1463.82 Q946.562 1462.43 946.562 1459.84 Q946.562 1457.18 944.664 1455.77 Q942.789 1454.33 939.271 1454.33 Q937.35 1454.33 935.15 1454.75 Q932.951 1455.16 930.313 1456.04 L930.313 1451.88 Q932.975 1451.14 935.289 1450.77 Q937.627 1450.39 939.687 1450.39 Q945.012 1450.39 948.113 1452.83 Q951.215 1455.23 951.215 1459.35 Q951.215 1462.22 949.572 1464.21 Q947.928 1466.18 944.896 1466.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M970.891 1454.1 Q967.28 1454.1 965.451 1457.66 Q963.646 1461.2 963.646 1468.33 Q963.646 1475.44 965.451 1479.01 Q967.28 1482.55 970.891 1482.55 Q974.525 1482.55 976.331 1479.01 Q978.159 1475.44 978.159 1468.33 Q978.159 1461.2 976.331 1457.66 Q974.525 1454.1 970.891 1454.1 M970.891 1450.39 Q976.701 1450.39 979.757 1455 Q982.835 1459.58 982.835 1468.33 Q982.835 1477.06 979.757 1481.67 Q976.701 1486.25 970.891 1486.25 Q965.081 1486.25 962.002 1481.67 Q958.947 1477.06 958.947 1468.33 Q958.947 1459.58 962.002 1455 Q965.081 1450.39 970.891 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1371.4 1484.86 L1371.4 1480.6 Q1373.16 1481.44 1374.97 1481.88 Q1376.78 1482.32 1378.51 1482.32 Q1383.14 1482.32 1385.57 1479.21 Q1388.03 1476.09 1388.37 1469.75 Q1387.03 1471.74 1384.97 1472.8 Q1382.91 1473.87 1380.41 1473.87 Q1375.22 1473.87 1372.19 1470.74 Q1369.18 1467.59 1369.18 1462.15 Q1369.18 1456.83 1372.33 1453.61 Q1375.48 1450.39 1380.71 1450.39 Q1386.71 1450.39 1389.85 1455 Q1393.03 1459.58 1393.03 1468.33 Q1393.03 1476.51 1389.14 1481.39 Q1385.27 1486.25 1378.72 1486.25 Q1376.96 1486.25 1375.15 1485.9 Q1373.35 1485.56 1371.4 1484.86 M1380.71 1470.21 Q1383.86 1470.21 1385.69 1468.06 Q1387.54 1465.9 1387.54 1462.15 Q1387.54 1458.43 1385.69 1456.27 Q1383.86 1454.1 1380.71 1454.1 Q1377.56 1454.1 1375.71 1456.27 Q1373.88 1458.43 1373.88 1462.15 Q1373.88 1465.9 1375.71 1468.06 Q1377.56 1470.21 1380.71 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1414.28 1455.09 L1402.47 1473.54 L1414.28 1473.54 L1414.28 1455.09 M1413.05 1451.02 L1418.93 1451.02 L1418.93 1473.54 L1423.86 1473.54 L1423.86 1477.43 L1418.93 1477.43 L1418.93 1485.58 L1414.28 1485.58 L1414.28 1477.43 L1398.67 1477.43 L1398.67 1472.92 L1413.05 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1441.59 1454.1 Q1437.98 1454.1 1436.15 1457.66 Q1434.34 1461.2 1434.34 1468.33 Q1434.34 1475.44 1436.15 1479.01 Q1437.98 1482.55 1441.59 1482.55 Q1445.22 1482.55 1447.03 1479.01 Q1448.86 1475.44 1448.86 1468.33 Q1448.86 1461.2 1447.03 1457.66 Q1445.22 1454.1 1441.59 1454.1 M1441.59 1450.39 Q1447.4 1450.39 1450.46 1455 Q1453.53 1459.58 1453.53 1468.33 Q1453.53 1477.06 1450.46 1481.67 Q1447.4 1486.25 1441.59 1486.25 Q1435.78 1486.25 1432.7 1481.67 Q1429.65 1477.06 1429.65 1468.33 Q1429.65 1459.58 1432.7 1455 Q1435.78 1450.39 1441.59 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1842.1 1484.86 L1842.1 1480.6 Q1843.86 1481.44 1845.67 1481.88 Q1847.47 1482.32 1849.21 1482.32 Q1853.84 1482.32 1856.27 1479.21 Q1858.72 1476.09 1859.07 1469.75 Q1857.73 1471.74 1855.67 1472.8 Q1853.61 1473.87 1851.11 1473.87 Q1845.92 1473.87 1842.89 1470.74 Q1839.88 1467.59 1839.88 1462.15 Q1839.88 1456.83 1843.03 1453.61 Q1846.18 1450.39 1851.41 1450.39 Q1857.4 1450.39 1860.55 1455 Q1863.72 1459.58 1863.72 1468.33 Q1863.72 1476.51 1859.84 1481.39 Q1855.97 1486.25 1849.42 1486.25 Q1847.66 1486.25 1845.85 1485.9 Q1844.05 1485.56 1842.1 1484.86 M1851.41 1470.21 Q1854.56 1470.21 1856.39 1468.06 Q1858.24 1465.9 1858.24 1462.15 Q1858.24 1458.43 1856.39 1456.27 Q1854.56 1454.1 1851.41 1454.1 Q1848.26 1454.1 1846.41 1456.27 Q1844.58 1458.43 1844.58 1462.15 Q1844.58 1465.9 1846.41 1468.06 Q1848.26 1470.21 1851.41 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1872.17 1451.02 L1890.53 1451.02 L1890.53 1454.96 L1876.46 1454.96 L1876.46 1463.43 Q1877.47 1463.08 1878.49 1462.92 Q1879.51 1462.73 1880.53 1462.73 Q1886.32 1462.73 1889.7 1465.9 Q1893.08 1469.08 1893.08 1474.49 Q1893.08 1480.07 1889.6 1483.17 Q1886.13 1486.25 1879.81 1486.25 Q1877.64 1486.25 1875.37 1485.88 Q1873.12 1485.51 1870.71 1484.77 L1870.71 1480.07 Q1872.8 1481.2 1875.02 1481.76 Q1877.24 1482.32 1879.72 1482.32 Q1883.72 1482.32 1886.06 1480.21 Q1888.4 1478.1 1888.4 1474.49 Q1888.4 1470.88 1886.06 1468.77 Q1883.72 1466.67 1879.72 1466.67 Q1877.84 1466.67 1875.97 1467.08 Q1874.12 1467.5 1872.17 1468.38 L1872.17 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1912.29 1454.1 Q1908.68 1454.1 1906.85 1457.66 Q1905.04 1461.2 1905.04 1468.33 Q1905.04 1475.44 1906.85 1479.01 Q1908.68 1482.55 1912.29 1482.55 Q1915.92 1482.55 1917.73 1479.01 Q1919.56 1475.44 1919.56 1468.33 Q1919.56 1461.2 1917.73 1457.66 Q1915.92 1454.1 1912.29 1454.1 M1912.29 1450.39 Q1918.1 1450.39 1921.15 1455 Q1924.23 1459.58 1924.23 1468.33 Q1924.23 1477.06 1921.15 1481.67 Q1918.1 1486.25 1912.29 1486.25 Q1906.48 1486.25 1903.4 1481.67 Q1900.34 1477.06 1900.34 1468.33 Q1900.34 1459.58 1903.4 1455 Q1906.48 1450.39 1912.29 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M2312.8 1484.86 L2312.8 1480.6 Q2314.56 1481.44 2316.37 1481.88 Q2318.17 1482.32 2319.91 1482.32 Q2324.54 1482.32 2326.97 1479.21 Q2329.42 1476.09 2329.77 1469.75 Q2328.43 1471.74 2326.37 1472.8 Q2324.31 1473.87 2321.81 1473.87 Q2316.62 1473.87 2313.59 1470.74 Q2310.58 1467.59 2310.58 1462.15 Q2310.58 1456.83 2313.73 1453.61 Q2316.88 1450.39 2322.11 1450.39 Q2328.1 1450.39 2331.25 1455 Q2334.42 1459.58 2334.42 1468.33 Q2334.42 1476.51 2330.53 1481.39 Q2326.67 1486.25 2320.12 1486.25 Q2318.36 1486.25 2316.55 1485.9 Q2314.75 1485.56 2312.8 1484.86 M2322.11 1470.21 Q2325.26 1470.21 2327.08 1468.06 Q2328.94 1465.9 2328.94 1462.15 Q2328.94 1458.43 2327.08 1456.27 Q2325.26 1454.1 2322.11 1454.1 Q2318.96 1454.1 2317.11 1456.27 Q2315.28 1458.43 2315.28 1462.15 Q2315.28 1465.9 2317.11 1468.06 Q2318.96 1470.21 2322.11 1470.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M2353.4 1466.44 Q2350.26 1466.44 2348.4 1468.59 Q2346.58 1470.74 2346.58 1474.49 Q2346.58 1478.22 2348.4 1480.39 Q2350.26 1482.55 2353.4 1482.55 Q2356.55 1482.55 2358.38 1480.39 Q2360.23 1478.22 2360.23 1474.49 Q2360.23 1470.74 2358.38 1468.59 Q2356.55 1466.44 2353.4 1466.44 M2362.69 1451.78 L2362.69 1456.04 Q2360.93 1455.21 2359.12 1454.77 Q2357.34 1454.33 2355.58 1454.33 Q2350.95 1454.33 2348.5 1457.45 Q2346.07 1460.58 2345.72 1466.9 Q2347.08 1464.89 2349.14 1463.82 Q2351.2 1462.73 2353.68 1462.73 Q2358.89 1462.73 2361.9 1465.9 Q2364.93 1469.05 2364.93 1474.49 Q2364.93 1479.82 2361.78 1483.03 Q2358.64 1486.25 2353.4 1486.25 Q2347.41 1486.25 2344.24 1481.67 Q2341.07 1477.06 2341.07 1468.33 Q2341.07 1460.14 2344.96 1455.28 Q2348.84 1450.39 2355.39 1450.39 Q2357.15 1450.39 2358.94 1450.74 Q2360.74 1451.09 2362.69 1451.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M2382.99 1454.1 Q2379.38 1454.1 2377.55 1457.66 Q2375.74 1461.2 2375.74 1468.33 Q2375.74 1475.44 2377.55 1479.01 Q2379.38 1482.55 2382.99 1482.55 Q2386.62 1482.55 2388.43 1479.01 Q2390.26 1475.44 2390.26 1468.33 Q2390.26 1461.2 2388.43 1457.66 Q2386.62 1454.1 2382.99 1454.1 M2382.99 1450.39 Q2388.8 1450.39 2391.85 1455 Q2394.93 1459.58 2394.93 1468.33 Q2394.93 1477.06 2391.85 1481.67 Q2388.8 1486.25 2382.99 1486.25 Q2377.18 1486.25 2374.1 1481.67 Q2371.04 1477.06 2371.04 1468.33 Q2371.04 1459.58 2374.1 1455 Q2377.18 1450.39 2382.99 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1219.08 1525.81 L1219.08 1543.66 L1227.16 1543.66 Q1231.65 1543.66 1234.1 1541.34 Q1236.55 1539.02 1236.55 1534.72 Q1236.55 1530.45 1234.1 1528.13 Q1231.65 1525.81 1227.16 1525.81 L1219.08 1525.81 M1212.65 1520.52 L1227.16 1520.52 Q1235.15 1520.52 1239.23 1524.15 Q1243.33 1527.75 1243.33 1534.72 Q1243.33 1541.75 1239.23 1545.35 Q1235.15 1548.95 1227.16 1548.95 L1219.08 1548.95 L1219.08 1568.04 L1212.65 1568.04 L1212.65 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1271.21 1537.87 Q1270.23 1537.3 1269.05 1537.04 Q1267.9 1536.76 1266.5 1536.76 Q1261.54 1536.76 1258.86 1540 Q1256.22 1543.22 1256.22 1549.27 L1256.22 1568.04 L1250.33 1568.04 L1250.33 1532.4 L1256.22 1532.4 L1256.22 1537.93 Q1258.07 1534.69 1261.03 1533.13 Q1263.99 1531.54 1268.22 1531.54 Q1268.83 1531.54 1269.56 1531.63 Q1270.29 1531.7 1271.18 1531.85 L1271.21 1537.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1289.74 1536.5 Q1285.03 1536.5 1282.29 1540.19 Q1279.55 1543.85 1279.55 1550.25 Q1279.55 1556.65 1282.26 1560.34 Q1284.99 1564 1289.74 1564 Q1294.42 1564 1297.15 1560.31 Q1299.89 1556.62 1299.89 1550.25 Q1299.89 1543.92 1297.15 1540.23 Q1294.42 1536.5 1289.74 1536.5 M1289.74 1531.54 Q1297.38 1531.54 1301.74 1536.5 Q1306.1 1541.47 1306.1 1550.25 Q1306.1 1559 1301.74 1564 Q1297.38 1568.97 1289.74 1568.97 Q1282.07 1568.97 1277.71 1564 Q1273.38 1559 1273.38 1550.25 Q1273.38 1541.47 1277.71 1536.5 Q1282.07 1531.54 1289.74 1531.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1333.85 1518.52 L1333.85 1523.39 L1328.25 1523.39 Q1325.1 1523.39 1323.86 1524.66 Q1322.65 1525.93 1322.65 1529.24 L1322.65 1532.4 L1332.29 1532.4 L1332.29 1536.95 L1322.65 1536.95 L1322.65 1568.04 L1316.76 1568.04 L1316.76 1536.95 L1311.16 1536.95 L1311.16 1532.4 L1316.76 1532.4 L1316.76 1529.91 Q1316.76 1523.96 1319.53 1521.26 Q1322.3 1518.52 1328.31 1518.52 L1333.85 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1338.75 1532.4 L1344.61 1532.4 L1344.61 1568.04 L1338.75 1568.04 L1338.75 1532.4 M1338.75 1518.52 L1344.61 1518.52 L1344.61 1525.93 L1338.75 1525.93 L1338.75 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1362.66 1522.27 L1362.66 1532.4 L1374.72 1532.4 L1374.72 1536.95 L1362.66 1536.95 L1362.66 1556.3 Q1362.66 1560.66 1363.83 1561.9 Q1365.04 1563.14 1368.7 1563.14 L1374.72 1563.14 L1374.72 1568.04 L1368.7 1568.04 Q1361.92 1568.04 1359.35 1565.53 Q1356.77 1562.98 1356.77 1556.3 L1356.77 1536.95 L1352.47 1536.95 L1352.47 1532.4 L1356.77 1532.4 L1356.77 1522.27 L1362.66 1522.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.612,1163.46 2352.76,1163.46 "/>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.612,844.983 2352.76,844.983 "/>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.612,526.506 2352.76,526.506 "/>
+<polyline clip-path="url(#clip452)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="234.612,208.028 2352.76,208.028 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.612,1423.18 234.612,123.472 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.612,1163.46 253.509,1163.46 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.612,844.983 253.509,844.983 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.612,526.506 253.509,526.506 "/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="234.612,208.028 253.509,208.028 "/>
+<path clip-path="url(#clip450)" d="M116.482 1180.02 L116.482 1175.76 Q118.242 1176.6 120.047 1177.04 Q121.853 1177.48 123.589 1177.48 Q128.218 1177.48 130.649 1174.38 Q133.103 1171.25 133.45 1164.91 Q132.107 1166.9 130.047 1167.96 Q127.987 1169.03 125.487 1169.03 Q120.302 1169.03 117.269 1165.9 Q114.26 1162.76 114.26 1157.32 Q114.26 1151.99 117.408 1148.77 Q120.556 1145.56 125.788 1145.56 Q131.783 1145.56 134.931 1150.16 Q138.103 1154.75 138.103 1163.5 Q138.103 1171.67 134.214 1176.55 Q130.348 1181.41 123.797 1181.41 Q122.038 1181.41 120.232 1181.07 Q118.427 1180.72 116.482 1180.02 M125.788 1165.37 Q128.936 1165.37 130.765 1163.22 Q132.617 1161.07 132.617 1157.32 Q132.617 1153.59 130.765 1151.44 Q128.936 1149.26 125.788 1149.26 Q122.64 1149.26 120.788 1151.44 Q118.959 1153.59 118.959 1157.32 Q118.959 1161.07 120.788 1163.22 Q122.64 1165.37 125.788 1165.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M150.533 1176.81 L166.852 1176.81 L166.852 1180.74 L144.908 1180.74 L144.908 1176.81 Q147.57 1174.05 152.153 1169.42 Q156.76 1164.77 157.94 1163.43 Q160.186 1160.9 161.065 1159.17 Q161.968 1157.41 161.968 1155.72 Q161.968 1152.96 160.024 1151.23 Q158.103 1149.49 155.001 1149.49 Q152.802 1149.49 150.348 1150.26 Q147.917 1151.02 145.14 1152.57 L145.14 1147.85 Q147.964 1146.71 150.417 1146.13 Q152.871 1145.56 154.908 1145.56 Q160.278 1145.56 163.473 1148.24 Q166.667 1150.93 166.667 1155.42 Q166.667 1157.55 165.857 1159.47 Q165.07 1161.37 162.964 1163.96 Q162.385 1164.63 159.283 1167.85 Q156.181 1171.04 150.533 1176.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M186.667 1149.26 Q183.056 1149.26 181.227 1152.82 Q179.422 1156.37 179.422 1163.5 Q179.422 1170.6 181.227 1174.17 Q183.056 1177.71 186.667 1177.71 Q190.301 1177.71 192.107 1174.17 Q193.936 1170.6 193.936 1163.5 Q193.936 1156.37 192.107 1152.82 Q190.301 1149.26 186.667 1149.26 M186.667 1145.56 Q192.477 1145.56 195.533 1150.16 Q198.612 1154.75 198.612 1163.5 Q198.612 1172.22 195.533 1176.83 Q192.477 1181.41 186.667 1181.41 Q180.857 1181.41 177.778 1176.83 Q174.723 1172.22 174.723 1163.5 Q174.723 1154.75 177.778 1150.16 Q180.857 1145.56 186.667 1145.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M116.482 861.546 L116.482 857.287 Q118.242 858.12 120.047 858.56 Q121.853 859 123.589 859 Q128.218 859 130.649 855.898 Q133.103 852.773 133.45 846.43 Q132.107 848.421 130.047 849.486 Q127.987 850.551 125.487 850.551 Q120.302 850.551 117.269 847.426 Q114.26 844.277 114.26 838.838 Q114.26 833.514 117.408 830.296 Q120.556 827.078 125.788 827.078 Q131.783 827.078 134.931 831.685 Q138.103 836.268 138.103 845.018 Q138.103 853.189 134.214 858.074 Q130.348 862.935 123.797 862.935 Q122.038 862.935 120.232 862.588 Q118.427 862.24 116.482 861.546 M125.788 846.893 Q128.936 846.893 130.765 844.74 Q132.617 842.588 132.617 838.838 Q132.617 835.111 130.765 832.958 Q128.936 830.782 125.788 830.782 Q122.64 830.782 120.788 832.958 Q118.959 835.111 118.959 838.838 Q118.959 842.588 120.788 844.74 Q122.64 846.893 125.788 846.893 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M159.353 831.778 L147.547 850.227 L159.353 850.227 L159.353 831.778 M158.126 827.703 L164.005 827.703 L164.005 850.227 L168.936 850.227 L168.936 854.115 L164.005 854.115 L164.005 862.263 L159.353 862.263 L159.353 854.115 L143.751 854.115 L143.751 849.602 L158.126 827.703 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M186.667 830.782 Q183.056 830.782 181.227 834.347 Q179.422 837.889 179.422 845.018 Q179.422 852.125 181.227 855.689 Q183.056 859.231 186.667 859.231 Q190.301 859.231 192.107 855.689 Q193.936 852.125 193.936 845.018 Q193.936 837.889 192.107 834.347 Q190.301 830.782 186.667 830.782 M186.667 827.078 Q192.477 827.078 195.533 831.685 Q198.612 836.268 198.612 845.018 Q198.612 853.745 195.533 858.351 Q192.477 862.935 186.667 862.935 Q180.857 862.935 177.778 858.351 Q174.723 853.745 174.723 845.018 Q174.723 836.268 177.778 831.685 Q180.857 827.078 186.667 827.078 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M116.482 543.068 L116.482 538.809 Q118.242 539.642 120.047 540.082 Q121.853 540.522 123.589 540.522 Q128.218 540.522 130.649 537.42 Q133.103 534.295 133.45 527.952 Q132.107 529.943 130.047 531.008 Q127.987 532.073 125.487 532.073 Q120.302 532.073 117.269 528.948 Q114.26 525.8 114.26 520.36 Q114.26 515.036 117.408 511.818 Q120.556 508.601 125.788 508.601 Q131.783 508.601 134.931 513.207 Q138.103 517.79 138.103 526.54 Q138.103 534.712 134.214 539.596 Q130.348 544.457 123.797 544.457 Q122.038 544.457 120.232 544.11 Q118.427 543.763 116.482 543.068 M125.788 528.415 Q128.936 528.415 130.765 526.263 Q132.617 524.11 132.617 520.36 Q132.617 516.633 130.765 514.48 Q128.936 512.304 125.788 512.304 Q122.64 512.304 120.788 514.48 Q118.959 516.633 118.959 520.36 Q118.959 524.11 120.788 526.263 Q122.64 528.415 125.788 528.415 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M157.084 524.642 Q153.936 524.642 152.084 526.795 Q150.255 528.948 150.255 532.698 Q150.255 536.425 152.084 538.601 Q153.936 540.753 157.084 540.753 Q160.232 540.753 162.061 538.601 Q163.913 536.425 163.913 532.698 Q163.913 528.948 162.061 526.795 Q160.232 524.642 157.084 524.642 M166.366 509.99 L166.366 514.249 Q164.607 513.416 162.802 512.976 Q161.019 512.536 159.26 512.536 Q154.63 512.536 152.177 515.661 Q149.746 518.786 149.399 525.105 Q150.765 523.091 152.825 522.027 Q154.885 520.939 157.362 520.939 Q162.57 520.939 165.579 524.11 Q168.612 527.258 168.612 532.698 Q168.612 538.022 165.464 541.239 Q162.315 544.457 157.084 544.457 Q151.089 544.457 147.917 539.874 Q144.746 535.267 144.746 526.54 Q144.746 518.346 148.635 513.485 Q152.524 508.601 159.075 508.601 Q160.834 508.601 162.616 508.948 Q164.422 509.295 166.366 509.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M186.667 512.304 Q183.056 512.304 181.227 515.869 Q179.422 519.411 179.422 526.54 Q179.422 533.647 181.227 537.212 Q183.056 540.753 186.667 540.753 Q190.301 540.753 192.107 537.212 Q193.936 533.647 193.936 526.54 Q193.936 519.411 192.107 515.869 Q190.301 512.304 186.667 512.304 M186.667 508.601 Q192.477 508.601 195.533 513.207 Q198.612 517.79 198.612 526.54 Q198.612 535.267 195.533 539.874 Q192.477 544.457 186.667 544.457 Q180.857 544.457 177.778 539.874 Q174.723 535.267 174.723 526.54 Q174.723 517.79 177.778 513.207 Q180.857 508.601 186.667 508.601 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M116.482 224.59 L116.482 220.331 Q118.242 221.164 120.047 221.604 Q121.853 222.044 123.589 222.044 Q128.218 222.044 130.649 218.942 Q133.103 215.817 133.45 209.475 Q132.107 211.465 130.047 212.53 Q127.987 213.595 125.487 213.595 Q120.302 213.595 117.269 210.47 Q114.26 207.322 114.26 201.882 Q114.26 196.558 117.408 193.341 Q120.556 190.123 125.788 190.123 Q131.783 190.123 134.931 194.729 Q138.103 199.313 138.103 208.063 Q138.103 216.234 134.214 221.118 Q130.348 225.979 123.797 225.979 Q122.038 225.979 120.232 225.632 Q118.427 225.285 116.482 224.59 M125.788 209.938 Q128.936 209.938 130.765 207.785 Q132.617 205.632 132.617 201.882 Q132.617 198.155 130.765 196.003 Q128.936 193.827 125.788 193.827 Q122.64 193.827 120.788 196.003 Q118.959 198.155 118.959 201.882 Q118.959 205.632 120.788 207.785 Q122.64 209.938 125.788 209.938 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M156.505 208.896 Q153.172 208.896 151.251 210.678 Q149.353 212.461 149.353 215.586 Q149.353 218.711 151.251 220.493 Q153.172 222.276 156.505 222.276 Q159.839 222.276 161.76 220.493 Q163.681 218.688 163.681 215.586 Q163.681 212.461 161.76 210.678 Q159.862 208.896 156.505 208.896 M151.829 206.905 Q148.82 206.165 147.13 204.104 Q145.464 202.044 145.464 199.081 Q145.464 194.938 148.403 192.53 Q151.366 190.123 156.505 190.123 Q161.667 190.123 164.607 192.53 Q167.547 194.938 167.547 199.081 Q167.547 202.044 165.857 204.104 Q164.19 206.165 161.204 206.905 Q164.584 207.692 166.459 209.984 Q168.357 212.276 168.357 215.586 Q168.357 220.609 165.278 223.294 Q162.223 225.979 156.505 225.979 Q150.788 225.979 147.709 223.294 Q144.654 220.609 144.654 215.586 Q144.654 212.276 146.552 209.984 Q148.45 207.692 151.829 206.905 M150.116 199.521 Q150.116 202.206 151.783 203.711 Q153.473 205.215 156.505 205.215 Q159.515 205.215 161.204 203.711 Q162.917 202.206 162.917 199.521 Q162.917 196.836 161.204 195.331 Q159.515 193.827 156.505 193.827 Q153.473 193.827 151.783 195.331 Q150.116 196.836 150.116 199.521 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M186.667 193.827 Q183.056 193.827 181.227 197.391 Q179.422 200.933 179.422 208.063 Q179.422 215.169 181.227 218.734 Q183.056 222.276 186.667 222.276 Q190.301 222.276 192.107 218.734 Q193.936 215.169 193.936 208.063 Q193.936 200.933 192.107 197.391 Q190.301 193.827 186.667 193.827 M186.667 190.123 Q192.477 190.123 195.533 194.729 Q198.612 199.313 198.612 208.063 Q198.612 216.789 195.533 221.396 Q192.477 225.979 186.667 225.979 Q180.857 225.979 177.778 221.396 Q174.723 216.789 174.723 208.063 Q174.723 199.313 177.778 194.729 Q180.857 190.123 186.667 190.123 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M21.7677 865.883 L58.7206 865.883 L58.7206 858.117 Q58.7206 848.282 54.2646 843.731 Q49.8086 839.147 40.1964 839.147 Q30.6479 839.147 26.2237 843.731 Q21.7677 848.282 21.7677 858.117 L21.7677 865.883 M16.4842 872.313 L16.4842 859.104 Q16.4842 845.29 22.2451 838.829 Q27.9743 832.368 40.1964 832.368 Q52.4822 832.368 58.2432 838.861 Q64.0042 845.354 64.0042 859.104 L64.0042 872.313 L16.4842 872.313 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M44.7161 791.882 L47.5806 791.882 L47.5806 818.809 Q53.6281 818.427 56.8109 815.181 Q59.9619 811.902 59.9619 806.078 Q59.9619 802.704 59.1344 799.553 Q58.3069 796.37 56.6518 793.251 L62.1899 793.251 Q63.5267 796.402 64.227 799.712 Q64.9272 803.022 64.9272 806.428 Q64.9272 814.958 59.9619 819.955 Q54.9967 824.92 46.5303 824.92 Q37.7774 824.92 32.6531 820.209 Q27.4968 815.467 27.4968 807.446 Q27.4968 800.253 32.1438 796.083 Q36.7589 791.882 44.7161 791.882 M42.9973 797.738 Q38.1912 797.802 35.3266 800.444 Q32.4621 803.054 32.4621 807.383 Q32.4621 812.284 35.2312 815.244 Q38.0002 818.172 43.0292 818.618 L42.9973 797.738 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M29.4065 759.544 L34.9447 759.544 Q33.6716 762.027 33.035 764.7 Q32.3984 767.374 32.3984 770.239 Q32.3984 774.599 33.7352 776.795 Q35.072 778.96 37.7456 778.96 Q39.7826 778.96 40.9603 777.4 Q42.1061 775.84 43.1565 771.13 L43.6021 769.125 Q44.9389 762.886 47.3897 760.276 Q49.8086 757.635 54.1691 757.635 Q59.1344 757.635 62.0308 761.581 Q64.9272 765.496 64.9272 772.371 Q64.9272 775.236 64.3543 778.355 Q63.8132 781.442 62.6992 784.88 L56.6518 784.88 Q58.3387 781.633 59.198 778.482 Q60.0256 775.331 60.0256 772.244 Q60.0256 768.106 58.6251 765.878 Q57.1929 763.65 54.6147 763.65 Q52.2276 763.65 50.9545 765.273 Q49.6813 766.865 48.5037 772.307 L48.0262 774.344 Q46.8804 779.787 44.5251 782.206 Q42.138 784.625 38.0002 784.625 Q32.9713 784.625 30.2341 781.06 Q27.4968 777.496 27.4968 770.939 Q27.4968 767.692 27.9743 764.828 Q28.4517 761.963 29.4065 759.544 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M28.3562 748.309 L28.3562 742.452 L64.0042 742.452 L64.0042 748.309 L28.3562 748.309 M14.479 748.309 L14.479 742.452 L21.895 742.452 L21.895 748.309 L14.479 748.309 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M33.8307 709.542 Q33.2578 710.528 33.0032 711.706 Q32.7167 712.852 32.7167 714.252 Q32.7167 719.217 35.9632 721.891 Q39.1779 724.533 45.2253 724.533 L64.0042 724.533 L64.0042 730.421 L28.3562 730.421 L28.3562 724.533 L33.8944 724.533 Q30.6479 722.687 29.0883 719.727 Q27.4968 716.767 27.4968 712.533 Q27.4968 711.929 27.5923 711.197 Q27.656 710.465 27.8151 709.573 L33.8307 709.542 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M44.7161 674.339 L47.5806 674.339 L47.5806 701.266 Q53.6281 700.884 56.8109 697.638 Q59.9619 694.359 59.9619 688.535 Q59.9619 685.161 59.1344 682.01 Q58.3069 678.827 56.6518 675.708 L62.1899 675.708 Q63.5267 678.859 64.227 682.169 Q64.9272 685.479 64.9272 688.885 Q64.9272 697.415 59.9619 702.412 Q54.9967 707.377 46.5303 707.377 Q37.7774 707.377 32.6531 702.667 Q27.4968 697.924 27.4968 689.903 Q27.4968 682.71 32.1438 678.541 Q36.7589 674.339 44.7161 674.339 M42.9973 680.196 Q38.1912 680.259 35.3266 682.901 Q32.4621 685.511 32.4621 689.84 Q32.4621 694.741 35.2312 697.701 Q38.0002 700.63 43.0292 701.075 L42.9973 680.196 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M994.686 17.6457 Q985.775 17.6457 980.508 24.2892 Q975.283 30.9327 975.283 42.3968 Q975.283 53.8203 980.508 60.4638 Q985.775 67.1073 994.686 67.1073 Q1003.6 67.1073 1008.78 60.4638 Q1014.01 53.8203 1014.01 42.3968 Q1014.01 30.9327 1008.78 24.2892 Q1003.6 17.6457 994.686 17.6457 M994.686 11.0023 Q1007.41 11.0023 1015.02 19.5497 Q1022.64 28.0566 1022.64 42.3968 Q1022.64 56.6965 1015.02 65.2439 Q1007.41 73.7508 994.686 73.7508 Q981.926 73.7508 974.27 65.2439 Q966.654 56.737 966.654 42.3968 Q966.654 28.0566 974.27 19.5497 Q981.926 11.0023 994.686 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1067.68 49.9314 Q1067.68 41.7081 1064.28 37.0496 Q1060.92 32.3505 1055 32.3505 Q1049.09 32.3505 1045.69 37.0496 Q1042.33 41.7081 1042.33 49.9314 Q1042.33 58.1548 1045.69 62.8538 Q1049.09 67.5124 1055 67.5124 Q1060.92 67.5124 1064.28 62.8538 Q1067.68 58.1548 1067.68 49.9314 M1042.33 34.0924 Q1044.67 30.0415 1048.24 28.0971 Q1051.84 26.1121 1056.83 26.1121 Q1065.09 26.1121 1070.24 32.6746 Q1075.42 39.2371 1075.42 49.9314 Q1075.42 60.6258 1070.24 67.1883 Q1065.09 73.7508 1056.83 73.7508 Q1051.84 73.7508 1048.24 71.8063 Q1044.67 69.8214 1042.33 65.7705 L1042.33 72.576 L1034.83 72.576 L1034.83 9.54393 L1042.33 9.54393 L1042.33 34.0924 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1087.78 27.2059 L1095.23 27.2059 L1095.23 73.3862 Q1095.23 82.0551 1091.91 85.944 Q1088.63 89.8329 1081.29 89.8329 L1078.46 89.8329 L1078.46 83.5134 L1080.44 83.5134 Q1084.7 83.5134 1086.24 81.5285 Q1087.78 79.5841 1087.78 73.3862 L1087.78 27.2059 M1087.78 9.54393 L1095.23 9.54393 L1095.23 18.9825 L1087.78 18.9825 L1087.78 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1149.63 48.0275 L1149.63 51.6733 L1115.36 51.6733 Q1115.85 59.3701 1119.98 63.421 Q1124.15 67.4314 1131.57 67.4314 Q1135.86 67.4314 1139.87 66.3781 Q1143.92 65.3249 1147.89 63.2184 L1147.89 70.267 Q1143.88 71.9684 1139.67 72.8596 Q1135.46 73.7508 1131.12 73.7508 Q1120.26 73.7508 1113.9 67.4314 Q1107.59 61.1119 1107.59 50.3365 Q1107.59 39.1965 1113.58 32.6746 Q1119.62 26.1121 1129.82 26.1121 Q1138.98 26.1121 1144.29 32.0264 Q1149.63 37.9003 1149.63 48.0275 M1142.18 45.84 Q1142.1 39.7232 1138.74 36.0774 Q1135.41 32.4315 1129.91 32.4315 Q1123.67 32.4315 1119.9 35.9558 Q1116.17 39.4801 1115.61 45.8805 L1142.18 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1194.52 28.9478 L1194.52 35.9153 Q1191.36 34.1734 1188.16 33.3227 Q1185 32.4315 1181.76 32.4315 Q1174.51 32.4315 1170.5 37.0496 Q1166.49 41.6271 1166.49 49.9314 Q1166.49 58.2358 1170.5 62.8538 Q1174.51 67.4314 1181.76 67.4314 Q1185 67.4314 1188.16 66.5807 Q1191.36 65.6895 1194.52 63.9476 L1194.52 70.8341 Q1191.4 72.2924 1188.04 73.0216 Q1184.71 73.7508 1180.95 73.7508 Q1170.7 73.7508 1164.66 67.3098 Q1158.63 60.8689 1158.63 49.9314 Q1158.63 38.832 1164.7 32.472 Q1170.82 26.1121 1181.43 26.1121 Q1184.88 26.1121 1188.16 26.8413 Q1191.44 27.5299 1194.52 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1214.85 14.324 L1214.85 27.2059 L1230.21 27.2059 L1230.21 32.9987 L1214.85 32.9987 L1214.85 57.6282 Q1214.85 63.1779 1216.35 64.7578 Q1217.89 66.3376 1222.55 66.3376 L1230.21 66.3376 L1230.21 72.576 L1222.55 72.576 Q1213.92 72.576 1210.64 69.3758 Q1207.36 66.1351 1207.36 57.6282 L1207.36 32.9987 L1201.89 32.9987 L1201.89 27.2059 L1207.36 27.2059 L1207.36 14.324 L1214.85 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1240.01 27.2059 L1247.46 27.2059 L1247.46 72.576 L1240.01 72.576 L1240.01 27.2059 M1240.01 9.54393 L1247.46 9.54393 L1247.46 18.9825 L1240.01 18.9825 L1240.01 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1257.71 27.2059 L1265.61 27.2059 L1279.79 65.2844 L1293.97 27.2059 L1301.87 27.2059 L1284.85 72.576 L1274.73 72.576 L1257.71 27.2059 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1350.96 48.0275 L1350.96 51.6733 L1316.69 51.6733 Q1317.18 59.3701 1321.31 63.421 Q1325.48 67.4314 1332.9 67.4314 Q1337.19 67.4314 1341.2 66.3781 Q1345.25 65.3249 1349.22 63.2184 L1349.22 70.267 Q1345.21 71.9684 1341 72.8596 Q1336.79 73.7508 1332.45 73.7508 Q1321.59 73.7508 1315.23 67.4314 Q1308.92 61.1119 1308.92 50.3365 Q1308.92 39.1965 1314.91 32.6746 Q1320.95 26.1121 1331.15 26.1121 Q1340.31 26.1121 1345.62 32.0264 Q1350.96 37.9003 1350.96 48.0275 M1343.51 45.84 Q1343.43 39.7232 1340.07 36.0774 Q1336.74 32.4315 1331.24 32.4315 Q1325 32.4315 1321.23 35.9558 Q1317.5 39.4801 1316.94 45.8805 L1343.51 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1418.49 28.5427 L1418.49 35.5912 Q1415.33 33.9709 1411.93 33.1607 Q1408.53 32.3505 1404.88 32.3505 Q1399.33 32.3505 1396.54 34.0519 Q1393.78 35.7533 1393.78 39.156 Q1393.78 41.7486 1395.77 43.2475 Q1397.75 44.7058 1403.75 46.0426 L1406.3 46.6097 Q1414.24 48.3111 1417.56 51.4303 Q1420.92 54.509 1420.92 60.0587 Q1420.92 66.3781 1415.9 70.0644 Q1410.92 73.7508 1402.17 73.7508 Q1398.52 73.7508 1394.55 73.0216 Q1390.62 72.3329 1386.25 70.9151 L1386.25 63.2184 Q1390.38 65.3654 1394.39 66.4591 Q1398.4 67.5124 1402.33 67.5124 Q1407.6 67.5124 1410.43 65.73 Q1413.27 63.9071 1413.27 60.6258 Q1413.27 57.5877 1411.2 55.9673 Q1409.18 54.3469 1402.25 52.8481 L1399.66 52.2405 Q1392.73 50.7821 1389.65 47.7845 Q1386.57 44.7463 1386.57 39.4801 Q1386.57 33.0797 1391.11 29.5959 Q1395.64 26.1121 1403.99 26.1121 Q1408.12 26.1121 1411.77 26.7198 Q1415.41 27.3274 1418.49 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1440 65.7705 L1440 89.8329 L1432.51 89.8329 L1432.51 27.2059 L1440 27.2059 L1440 34.0924 Q1442.35 30.0415 1445.92 28.0971 Q1449.52 26.1121 1454.5 26.1121 Q1462.77 26.1121 1467.91 32.6746 Q1473.1 39.2371 1473.1 49.9314 Q1473.1 60.6258 1467.91 67.1883 Q1462.77 73.7508 1454.5 73.7508 Q1449.52 73.7508 1445.92 71.8063 Q1442.35 69.8214 1440 65.7705 M1465.36 49.9314 Q1465.36 41.7081 1461.96 37.0496 Q1458.6 32.3505 1452.68 32.3505 Q1446.77 32.3505 1443.36 37.0496 Q1440 41.7081 1440 49.9314 Q1440 58.1548 1443.36 62.8538 Q1446.77 67.5124 1452.68 67.5124 Q1458.6 67.5124 1461.96 62.8538 Q1465.36 58.1548 1465.36 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1506.07 49.7694 Q1497.04 49.7694 1493.56 51.8354 Q1490.07 53.9013 1490.07 58.8839 Q1490.07 62.8538 1492.66 65.2034 Q1495.3 67.5124 1499.79 67.5124 Q1505.99 67.5124 1509.72 63.1374 Q1513.49 58.7219 1513.49 51.4303 L1513.49 49.7694 L1506.07 49.7694 M1520.94 46.6907 L1520.94 72.576 L1513.49 72.576 L1513.49 65.6895 Q1510.93 69.8214 1507.13 71.8063 Q1503.32 73.7508 1497.81 73.7508 Q1490.84 73.7508 1486.71 69.8619 Q1482.62 65.9325 1482.62 59.3701 Q1482.62 51.7138 1487.72 47.825 Q1492.87 43.9361 1503.03 43.9361 L1513.49 43.9361 L1513.49 43.2069 Q1513.49 38.0623 1510.08 35.2672 Q1506.72 32.4315 1500.6 32.4315 Q1496.72 32.4315 1493.03 33.3632 Q1489.34 34.295 1485.94 36.1584 L1485.94 29.2718 Q1490.03 27.692 1493.88 26.9223 Q1497.73 26.1121 1501.37 26.1121 Q1511.22 26.1121 1516.08 31.2163 Q1520.94 36.3204 1520.94 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1568.94 28.9478 L1568.94 35.9153 Q1565.78 34.1734 1562.58 33.3227 Q1559.42 32.4315 1556.18 32.4315 Q1548.93 32.4315 1544.92 37.0496 Q1540.91 41.6271 1540.91 49.9314 Q1540.91 58.2358 1544.92 62.8538 Q1548.93 67.4314 1556.18 67.4314 Q1559.42 67.4314 1562.58 66.5807 Q1565.78 65.6895 1568.94 63.9476 L1568.94 70.8341 Q1565.82 72.2924 1562.46 73.0216 Q1559.14 73.7508 1555.37 73.7508 Q1545.12 73.7508 1539.09 67.3098 Q1533.05 60.8689 1533.05 49.9314 Q1533.05 38.832 1539.13 32.472 Q1545.24 26.1121 1555.86 26.1121 Q1559.3 26.1121 1562.58 26.8413 Q1565.86 27.5299 1568.94 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1620.71 48.0275 L1620.71 51.6733 L1586.44 51.6733 Q1586.93 59.3701 1591.06 63.421 Q1595.23 67.4314 1602.65 67.4314 Q1606.94 67.4314 1610.95 66.3781 Q1615 65.3249 1618.97 63.2184 L1618.97 70.267 Q1614.96 71.9684 1610.75 72.8596 Q1606.54 73.7508 1602.2 73.7508 Q1591.34 73.7508 1584.98 67.4314 Q1578.66 61.1119 1578.66 50.3365 Q1578.66 39.1965 1584.66 32.6746 Q1590.7 26.1121 1600.9 26.1121 Q1610.06 26.1121 1615.37 32.0264 Q1620.71 37.9003 1620.71 48.0275 M1613.26 45.84 Q1613.18 39.7232 1609.82 36.0774 Q1606.49 32.4315 1600.99 32.4315 Q1594.75 32.4315 1590.98 35.9558 Q1587.25 39.4801 1586.69 45.8805 L1613.26 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><circle clip-path="url(#clip452)" cx="375.821" cy="160.256" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="799.45" cy="335.419" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="1128.94" cy="351.343" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="1176.01" cy="733.516" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="1223.08" cy="813.136" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="1552.57" cy="844.983" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="1787.92" cy="860.907" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="1834.99" cy="1243.08" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="2117.41" cy="1386.4" r="14.4" fill="#009af9" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<circle clip-path="url(#clip452)" cx="2117.41" cy="160.256" r="14.4" fill="#e26f46" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="3.2"/>
+<path clip-path="url(#clip450)" d="M305.216 1379.86 L762.678 1379.86 L762.678 1276.18 L305.216 1276.18  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip450)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="305.216,1379.86 762.678,1379.86 762.678,1276.18 305.216,1276.18 305.216,1379.86 "/>
+<circle clip-path="url(#clip450)" cx="399.356" cy="1328.02" r="20.48" fill="#e26f46" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.55111"/>
+<path clip-path="url(#clip450)" d="M493.496 1310.74 L498.172 1310.74 L498.172 1345.3 L493.496 1345.3 L493.496 1310.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M524.352 1323.31 L524.352 1309.28 L528.611 1309.28 L528.611 1345.3 L524.352 1345.3 L524.352 1341.41 Q523.01 1343.72 520.949 1344.86 Q518.912 1345.97 516.042 1345.97 Q511.343 1345.97 508.38 1342.22 Q505.44 1338.47 505.44 1332.36 Q505.44 1326.25 508.38 1322.5 Q511.343 1318.75 516.042 1318.75 Q518.912 1318.75 520.949 1319.88 Q523.01 1320.99 524.352 1323.31 M509.838 1332.36 Q509.838 1337.06 511.76 1339.74 Q513.704 1342.4 517.084 1342.4 Q520.463 1342.4 522.408 1339.74 Q524.352 1337.06 524.352 1332.36 Q524.352 1327.66 522.408 1325 Q520.463 1322.31 517.084 1322.31 Q513.704 1322.31 511.76 1325 Q509.838 1327.66 509.838 1332.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M559.56 1331.27 L559.56 1333.35 L539.977 1333.35 Q540.255 1337.75 542.616 1340.06 Q545 1342.36 549.236 1342.36 Q551.69 1342.36 553.982 1341.75 Q556.296 1341.15 558.565 1339.95 L558.565 1343.98 Q556.273 1344.95 553.866 1345.46 Q551.459 1345.97 548.982 1345.97 Q542.778 1345.97 539.144 1342.36 Q535.533 1338.75 535.533 1332.59 Q535.533 1326.22 538.959 1322.5 Q542.408 1318.75 548.241 1318.75 Q553.472 1318.75 556.505 1322.13 Q559.56 1325.48 559.56 1331.27 M555.301 1330.02 Q555.255 1326.52 553.334 1324.44 Q551.435 1322.36 548.287 1322.36 Q544.722 1322.36 542.57 1324.37 Q540.44 1326.38 540.116 1330.04 L555.301 1330.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M578.333 1332.26 Q573.171 1332.26 571.181 1333.44 Q569.19 1334.63 569.19 1337.47 Q569.19 1339.74 570.671 1341.08 Q572.176 1342.4 574.745 1342.4 Q578.287 1342.4 580.417 1339.9 Q582.569 1337.38 582.569 1333.21 L582.569 1332.26 L578.333 1332.26 M586.829 1330.5 L586.829 1345.3 L582.569 1345.3 L582.569 1341.36 Q581.111 1343.72 578.935 1344.86 Q576.759 1345.97 573.611 1345.97 Q569.63 1345.97 567.269 1343.75 Q564.931 1341.5 564.931 1337.75 Q564.931 1333.38 567.847 1331.15 Q570.787 1328.93 576.597 1328.93 L582.569 1328.93 L582.569 1328.51 Q582.569 1325.57 580.625 1323.98 Q578.704 1322.36 575.208 1322.36 Q572.986 1322.36 570.88 1322.89 Q568.773 1323.42 566.829 1324.49 L566.829 1320.55 Q569.167 1319.65 571.366 1319.21 Q573.565 1318.75 575.648 1318.75 Q581.273 1318.75 584.051 1321.66 Q586.829 1324.58 586.829 1330.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M595.602 1309.28 L599.861 1309.28 L599.861 1345.3 L595.602 1345.3 L595.602 1309.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M627.963 1341.41 L627.963 1355.16 L623.68 1355.16 L623.68 1319.37 L627.963 1319.37 L627.963 1323.31 Q629.305 1320.99 631.342 1319.88 Q633.403 1318.75 636.25 1318.75 Q640.972 1318.75 643.912 1322.5 Q646.875 1326.25 646.875 1332.36 Q646.875 1338.47 643.912 1342.22 Q640.972 1345.97 636.25 1345.97 Q633.403 1345.97 631.342 1344.86 Q629.305 1343.72 627.963 1341.41 M642.453 1332.36 Q642.453 1327.66 640.509 1325 Q638.588 1322.31 635.208 1322.31 Q631.828 1322.31 629.884 1325 Q627.963 1327.66 627.963 1332.36 Q627.963 1337.06 629.884 1339.74 Q631.828 1342.4 635.208 1342.4 Q638.588 1342.4 640.509 1339.74 Q642.453 1337.06 642.453 1332.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M663.981 1322.36 Q660.555 1322.36 658.564 1325.04 Q656.574 1327.7 656.574 1332.36 Q656.574 1337.01 658.541 1339.69 Q660.532 1342.36 663.981 1342.36 Q667.384 1342.36 669.375 1339.67 Q671.365 1336.99 671.365 1332.36 Q671.365 1327.75 669.375 1325.06 Q667.384 1322.36 663.981 1322.36 M663.981 1318.75 Q669.537 1318.75 672.708 1322.36 Q675.879 1325.97 675.879 1332.36 Q675.879 1338.72 672.708 1342.36 Q669.537 1345.97 663.981 1345.97 Q658.402 1345.97 655.231 1342.36 Q652.083 1338.72 652.083 1332.36 Q652.083 1325.97 655.231 1322.36 Q658.402 1318.75 663.981 1318.75 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M682.939 1319.37 L687.199 1319.37 L687.199 1345.3 L682.939 1345.3 L682.939 1319.37 M682.939 1309.28 L687.199 1309.28 L687.199 1314.67 L682.939 1314.67 L682.939 1309.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M717.661 1329.65 L717.661 1345.3 L713.402 1345.3 L713.402 1329.79 Q713.402 1326.11 711.967 1324.28 Q710.532 1322.45 707.661 1322.45 Q704.212 1322.45 702.222 1324.65 Q700.231 1326.85 700.231 1330.64 L700.231 1345.3 L695.948 1345.3 L695.948 1319.37 L700.231 1319.37 L700.231 1323.4 Q701.759 1321.06 703.819 1319.9 Q705.902 1318.75 708.61 1318.75 Q713.078 1318.75 715.37 1321.52 Q717.661 1324.28 717.661 1329.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M730.37 1312.01 L730.37 1319.37 L739.143 1319.37 L739.143 1322.68 L730.37 1322.68 L730.37 1336.75 Q730.37 1339.93 731.226 1340.83 Q732.106 1341.73 734.768 1341.73 L739.143 1341.73 L739.143 1345.3 L734.768 1345.3 Q729.837 1345.3 727.962 1343.47 Q726.087 1341.62 726.087 1336.75 L726.087 1322.68 L722.962 1322.68 L722.962 1319.37 L726.087 1319.37 L726.087 1312.01 L730.37 1312.01 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M316.729 176.937 L326.277 176.937 L326.277 143.98 L315.89 146.064 L315.89 140.74 L326.219 138.656 L332.064 138.656 L332.064 176.937 L341.613 176.937 L341.613 181.856 L316.729 181.856 L316.729 176.937 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M745.696 352.1 L766.095 352.1 L766.095 357.019 L738.665 357.019 L738.665 352.1 Q741.992 348.657 747.722 342.87 Q753.48 337.054 754.955 335.376 Q757.762 332.222 758.862 330.052 Q759.99 327.853 759.99 325.74 Q759.99 322.297 757.56 320.127 Q755.158 317.957 751.281 317.957 Q748.532 317.957 745.465 318.912 Q742.427 319.866 738.954 321.805 L738.954 315.902 Q742.484 314.485 745.552 313.761 Q748.619 313.038 751.165 313.038 Q757.878 313.038 761.871 316.394 Q765.864 319.751 765.864 325.364 Q765.864 328.026 764.851 330.428 Q763.867 332.8 761.234 336.041 Q760.511 336.88 756.634 340.902 Q752.756 344.895 745.696 352.1 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1087.18 349.65 Q1091.37 350.547 1093.72 353.383 Q1096.09 356.219 1096.09 360.385 Q1096.09 366.78 1091.69 370.281 Q1087.29 373.782 1079.19 373.782 Q1076.47 373.782 1073.58 373.232 Q1070.71 372.711 1067.65 371.641 L1067.65 365.999 Q1070.08 367.416 1072.97 368.14 Q1075.87 368.863 1079.02 368.863 Q1084.52 368.863 1087.38 366.693 Q1090.27 364.523 1090.27 360.385 Q1090.27 356.566 1087.58 354.425 Q1084.92 352.254 1080.15 352.254 L1075.11 352.254 L1075.11 347.451 L1080.38 347.451 Q1084.69 347.451 1086.98 345.744 Q1089.26 344.008 1089.26 340.767 Q1089.26 337.44 1086.89 335.675 Q1084.55 333.881 1080.15 333.881 Q1077.75 333.881 1075 334.401 Q1072.25 334.922 1068.95 336.022 L1068.95 330.814 Q1072.28 329.888 1075.17 329.425 Q1078.09 328.962 1080.67 328.962 Q1087.32 328.962 1091.2 332 Q1095.08 335.009 1095.08 340.16 Q1095.08 343.748 1093.02 346.236 Q1090.97 348.695 1087.18 349.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1132.7 717.009 L1117.94 740.07 L1132.7 740.07 L1132.7 717.009 M1131.17 711.916 L1138.52 711.916 L1138.52 740.07 L1144.68 740.07 L1144.68 744.931 L1138.52 744.931 L1138.52 755.116 L1132.7 755.116 L1132.7 744.931 L1113.2 744.931 L1113.2 739.289 L1131.17 711.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1163.86 791.536 L1186.8 791.536 L1186.8 796.455 L1169.21 796.455 L1169.21 807.045 Q1170.48 806.611 1171.76 806.408 Q1173.03 806.177 1174.3 806.177 Q1181.54 806.177 1185.76 810.141 Q1189.98 814.105 1189.98 820.876 Q1189.98 827.849 1185.64 831.726 Q1181.3 835.575 1173.4 835.575 Q1170.68 835.575 1167.85 835.112 Q1165.04 834.649 1162.03 833.723 L1162.03 827.849 Q1164.64 829.267 1167.42 829.961 Q1170.19 830.656 1173.29 830.656 Q1178.29 830.656 1181.22 828.023 Q1184.14 825.39 1184.14 820.876 Q1184.14 816.362 1181.22 813.729 Q1178.29 811.096 1173.29 811.096 Q1170.95 811.096 1168.6 811.617 Q1166.29 812.137 1163.86 813.237 L1163.86 791.536 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1506 842.654 Q1502.07 842.654 1499.75 845.345 Q1497.47 848.036 1497.47 852.724 Q1497.47 857.382 1499.75 860.102 Q1502.07 862.793 1506 862.793 Q1509.94 862.793 1512.23 860.102 Q1514.54 857.382 1514.54 852.724 Q1514.54 848.036 1512.23 845.345 Q1509.94 842.654 1506 842.654 M1517.61 824.338 L1517.61 829.662 Q1515.41 828.621 1513.15 828.071 Q1510.92 827.521 1508.72 827.521 Q1502.94 827.521 1499.87 831.427 Q1496.83 835.334 1496.4 843.233 Q1498.11 840.716 1500.68 839.385 Q1503.26 838.025 1506.35 838.025 Q1512.86 838.025 1516.62 841.989 Q1520.41 845.924 1520.41 852.724 Q1520.41 859.379 1516.48 863.401 Q1512.54 867.423 1506 867.423 Q1498.51 867.423 1494.55 861.693 Q1490.58 855.935 1490.58 845.027 Q1490.58 834.784 1495.44 828.708 Q1500.3 822.602 1508.49 822.602 Q1510.69 822.602 1512.92 823.036 Q1515.18 823.47 1517.61 824.338 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1726.96 839.307 L1754.74 839.307 L1754.74 841.796 L1739.05 882.507 L1732.95 882.507 L1747.71 844.226 L1726.96 844.226 L1726.96 839.307 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M1787.92 1244.17 Q1783.75 1244.17 1781.35 1246.39 Q1778.98 1248.62 1778.98 1252.53 Q1778.98 1256.43 1781.35 1258.66 Q1783.75 1260.89 1787.92 1260.89 Q1792.08 1260.89 1794.49 1258.66 Q1796.89 1256.41 1796.89 1252.53 Q1796.89 1248.62 1794.49 1246.39 Q1792.11 1244.17 1787.92 1244.17 M1782.07 1241.68 Q1778.31 1240.75 1776.2 1238.18 Q1774.12 1235.6 1774.12 1231.9 Q1774.12 1226.72 1777.79 1223.71 Q1781.49 1220.7 1787.92 1220.7 Q1794.37 1220.7 1798.04 1223.71 Q1801.72 1226.72 1801.72 1231.9 Q1801.72 1235.6 1799.61 1238.18 Q1797.52 1240.75 1793.79 1241.68 Q1798.02 1242.66 1800.36 1245.53 Q1802.73 1248.39 1802.73 1252.53 Q1802.73 1258.81 1798.88 1262.16 Q1795.06 1265.52 1787.92 1265.52 Q1780.77 1265.52 1776.92 1262.16 Q1773.1 1258.81 1773.1 1252.53 Q1773.1 1248.39 1775.48 1245.53 Q1777.85 1242.66 1782.07 1241.68 M1779.93 1232.45 Q1779.93 1235.8 1782.01 1237.68 Q1784.13 1239.57 1787.92 1239.57 Q1791.68 1239.57 1793.79 1237.68 Q1795.93 1235.8 1795.93 1232.45 Q1795.93 1229.09 1793.79 1227.21 Q1791.68 1225.33 1787.92 1225.33 Q1784.13 1225.33 1782.01 1227.21 Q1779.93 1229.09 1779.93 1232.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip450)" d="M2058.21 1407.1 L2058.21 1401.77 Q2060.41 1402.82 2062.67 1403.37 Q2064.93 1403.92 2067.1 1403.92 Q2072.88 1403.92 2075.92 1400.04 Q2078.99 1396.13 2079.42 1388.2 Q2077.74 1390.69 2075.17 1392.02 Q2072.59 1393.35 2069.47 1393.35 Q2062.99 1393.35 2059.2 1389.45 Q2055.44 1385.51 2055.44 1378.71 Q2055.44 1372.06 2059.37 1368.04 Q2063.31 1364.01 2069.84 1364.01 Q2077.34 1364.01 2081.27 1369.77 Q2085.24 1375.5 2085.24 1386.44 Q2085.24 1396.65 2080.38 1402.76 Q2075.54 1408.83 2067.36 1408.83 Q2065.16 1408.83 2062.9 1408.4 Q2060.64 1407.97 2058.21 1407.1 M2069.84 1388.78 Q2073.78 1388.78 2076.07 1386.09 Q2078.38 1383.4 2078.38 1378.71 Q2078.38 1374.05 2076.07 1371.36 Q2073.78 1368.64 2069.84 1368.64 Q2065.91 1368.64 2063.59 1371.36 Q2061.31 1374.05 2061.31 1378.71 Q2061.31 1383.4 2063.59 1386.09 Q2065.91 1388.78 2069.84 1388.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/linear/multi_objective_knapsack/index.html b/previews/PR3525/tutorials/linear/multi_objective_knapsack/index.html
new file mode 100644
index 00000000000..14489cc0613
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/multi_objective_knapsack/index.html
@@ -0,0 +1,94 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Multi-objective knapsack · JuMP</title><meta name="title" content="Multi-objective knapsack · JuMP"/><meta property="og:title" content="Multi-objective knapsack · JuMP"/><meta property="twitter:title" content="Multi-objective knapsack · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li class="is-active"><a class="tocitem" href>Multi-objective knapsack</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Solution"><span>Solution</span></a></li><li><a class="tocitem" href="#Accessing-multiple-solutions"><span>Accessing multiple solutions</span></a></li><li><a class="tocitem" href="#Visualizing-objective-space"><span>Visualizing objective space</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Multi-objective knapsack</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Multi-objective knapsack</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/multi_objective_knapsack.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Multi-objective-knapsack"><a class="docs-heading-anchor" href="#Multi-objective-knapsack">Multi-objective knapsack</a><a id="Multi-objective-knapsack-1"></a><a class="docs-heading-anchor-permalink" href="#Multi-objective-knapsack" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../multi_objective_knapsack.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to demonstrate how to create and solve a multi-objective linear program. In addition, it demonstrates how to work with solvers which return multiple solutions.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+import MultiObjectiveAlgorithms as MOA
+import Plots</code></pre><p><a href="../../../packages/MultiObjectiveAlgorithms/#MultiObjectiveAlgorithms.jl">MultiObjectiveAlgorithms.jl</a> is a package which implements a variety of algorithms for solving multi-objective optimization problems. Because it is a long package name, we import it instead as <code>MOA</code>.</p><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>The <a href="https://en.wikipedia.org/wiki/Knapsack_problem">knapsack problem</a> is a classic problem in mixed-integer programming. Given a collection of items <span>$i \in I$</span>, each of which has an associated weight, <span>$w_i$</span>, and profit, <span>$p_i$</span>, the knapsack problem determines which profit-maximizing subset of items to pack into a knapsack such that the total weight is less than a capacity <span>$c$</span>. The mathematical formulation is:</p><p class="math-container">\[\begin{aligned}
+\max &amp; \sum\limits_{i \in I} p_i x_i \\
+\text{s.t.}\ \ &amp; \sum\limits_{i \in I} w_i x_i \le c\\
+&amp; x_i \in \{0, 1\} &amp;&amp; \forall i \in I
+\end{aligned}\]</p><p>where <span>$x_i$</span> is <span>$1$</span> if we pack item <span>$i$</span> into the knapsack and <span>$0$</span> otherwise.</p><p>For this tutorial, we extend the single-objective knapsack problem by adding another objective: given a desirability rating, <span>$r_i$</span>, we wish to maximize the total desirability of the items in our knapsack. Thus, our mathematical formulation is now:</p><p class="math-container">\[\begin{aligned}
+\max &amp; \sum\limits_{i \in I} p_i x_i \\
+     &amp; \sum\limits_{i \in I} r_i x_i \\
+\text{s.t.}\ \ &amp; \sum\limits_{i \in I} w_i x_i \le c\\
+&amp; x_i \in \{0, 1\} &amp;&amp; \forall i \in I
+\end{aligned}\]</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>The data for this example was taken from <a href="https://github.com/vOptSolver/vOptGeneric.jl">vOptGeneric</a>, and the original author was <a href="https://github.com/xgandibleux">@xgandibleux</a>.</p><pre><code class="language-julia hljs">profit = [77, 94, 71, 63, 96, 82, 85, 75, 72, 91, 99, 63, 84, 87, 79, 94, 90]
+desire = [65, 90, 90, 77, 95, 84, 70, 94, 66, 92, 74, 97, 60, 60, 65, 97, 93]
+weight = [80, 87, 68, 72, 66, 77, 99, 85, 70, 93, 98, 72, 100, 89, 67, 86, 91]
+capacity = 900
+N = length(profit)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">17</code></pre><p>Comparing the capacity to the total weight of all the items:</p><pre><code class="language-julia hljs">capacity / sum(weight)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.6428571428571429</code></pre><p>shows that we can take approximately 64% of the items.</p><p>Plotting the items, we see that there are a range of items with different profits and desirability. Some items have a high profit and a high desirability, others have a low profit and a high desirability (and vice versa).</p><pre><code class="language-julia hljs">Plots.scatter(
+    profit,
+    desire;
+    xlabel = &quot;Profit&quot;,
+    ylabel = &quot;Desire&quot;,
+    legend = false,
+)</code></pre><img src="0a06bb1f.svg" alt="Example block output"/><p>The goal of the bi-objective knapsack problem is to choose a subset which maximizes both objectives.</p><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>Our JuMP formulation is a direct translation of the mathematical formulation:</p><pre><code class="language-julia hljs">model = Model()
+@variable(model, x[1:N], Bin)
+@constraint(model, sum(weight[i] * x[i] for i in 1:N) &lt;= capacity)
+@expression(model, profit_expr, sum(profit[i] * x[i] for i in 1:N))
+@expression(model, desire_expr, sum(desire[i] * x[i] for i in 1:N))
+@objective(model, Max, [profit_expr, desire_expr])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{AffExpr}:
+ 77 x[1] + 94 x[2] + 71 x[3] + 63 x[4] + 96 x[5] + 82 x[6] + 85 x[7] + 75 x[8] + 72 x[9] + 91 x[10] + 99 x[11] + 63 x[12] + 84 x[13] + 87 x[14] + 79 x[15] + 94 x[16] + 90 x[17]
+ 65 x[1] + 90 x[2] + 90 x[3] + 77 x[4] + 95 x[5] + 84 x[6] + 70 x[7] + 94 x[8] + 66 x[9] + 92 x[10] + 74 x[11] + 97 x[12] + 60 x[13] + 60 x[14] + 65 x[15] + 97 x[16] + 93 x[17]</code></pre><p>Note how we form a multi-objective program by passing a vector of scalar objective functions.</p><h2 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h2><p>To solve our model, we need an optimizer which supports multi-objective linear programs. One option is to use the <a href="../../../packages/MultiObjectiveAlgorithms/#MultiObjectiveAlgorithms.jl">MultiObjectiveAlgorithms.jl</a> package.</p><pre><code class="language-julia hljs">set_optimizer(model, () -&gt; MOA.Optimizer(HiGHS.Optimizer))
+set_silent(model)</code></pre><p>MultiObjectiveAlgorithms.jl supports many different algorithms for solving multiobjective optimization problems. One option is the epsilon-constraint method:</p><pre><code class="language-julia hljs">set_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())</code></pre><p>Let&#39;s solve the problem and see the solution</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : MOA[algorithm=MultiObjectiveAlgorithms.EpsilonConstraint, optimizer=HiGHS]
+
+* Status
+  Result count       : 9
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;Solve complete. Found 9 solution(s)&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : [9.18000e+02,9.83000e+02]
+  Objective bound    : [9.55000e+02,9.83000e+02]
+  Relative gap       : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 1.25902e-01
+  Simplex iterations : 6
+  Barrier iterations : -1
+  Node count         : 1
+</code></pre><p>There are 9 solutions available. We can also use <a href="../../../api/JuMP/#JuMP.result_count"><code>result_count</code></a> to see how many solutions are available:</p><pre><code class="language-julia hljs">result_count(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9</code></pre><h2 id="Accessing-multiple-solutions"><a class="docs-heading-anchor" href="#Accessing-multiple-solutions">Accessing multiple solutions</a><a id="Accessing-multiple-solutions-1"></a><a class="docs-heading-anchor-permalink" href="#Accessing-multiple-solutions" title="Permalink"></a></h2><p>Access the nine different solutions in the model using the <code>result</code> keyword to <a href="../../../api/JuMP/#JuMP.solution_summary"><code>solution_summary</code></a>, <a href="../../../api/JuMP/#JuMP.value"><code>value</code></a>, and <a href="../../../api/JuMP/#JuMP.objective_value"><code>objective_value</code></a>:</p><pre><code class="language-julia hljs">solution_summary(model; result = 5)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : MOA[algorithm=MultiObjectiveAlgorithms.EpsilonConstraint, optimizer=HiGHS]
+
+* Status
+  Result count       : 9
+  Termination status : OPTIMAL
+
+* Candidate solution (result #5)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : [9.36000e+02,9.42000e+02]
+</code></pre><pre><code class="language-julia hljs">objective_value(model; result = 5)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 936.0
+ 942.0</code></pre><p>Note that because we set a vector of two objective functions, the objective value is a vector with two elements. We can also query the value of each objective separately:</p><pre><code class="language-julia hljs">value(profit_expr; result = 5)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">936.0</code></pre><h2 id="Visualizing-objective-space"><a class="docs-heading-anchor" href="#Visualizing-objective-space">Visualizing objective space</a><a id="Visualizing-objective-space-1"></a><a class="docs-heading-anchor-permalink" href="#Visualizing-objective-space" title="Permalink"></a></h2><p>Unlike single-objective optimization problems, multi-objective optimization problems do not have a single optimal solution. Instead, the solutions returned represent possible trade-offs that the decision maker can choose between the two objectives. A common way to visualize this is by plotting the objective values of each of the solutions:</p><pre><code class="language-julia hljs">plot = Plots.scatter(
+    [value(profit_expr; result = i) for i in 1:result_count(model)],
+    [value(desire_expr; result = i) for i in 1:result_count(model)];
+    xlabel = &quot;Profit&quot;,
+    ylabel = &quot;Desire&quot;,
+    title = &quot;Objective space&quot;,
+    label = &quot;&quot;,
+    xlims = (915, 960),
+)
+for i in 1:result_count(model)
+    y = objective_value(model; result = i)
+    Plots.annotate!(y[1] - 1, y[2], (i, 10))
+end
+ideal_point = objective_bound(model)
+Plots.scatter!([ideal_point[1]], [ideal_point[2]]; label = &quot;Ideal point&quot;)</code></pre><img src="144638a3.svg" alt="Example block output"/><p>Visualizing the objective space lets the decision maker choose a solution that suits their personal preferences. For example, result <code>#7</code> is close to the maximum value of profit, but offers significantly higher desirability compared with solutions <code>#8</code> and <code>#9</code>.</p><p>The set of items that are chosen in solution <code>#7</code> are:</p><pre><code class="language-julia hljs">items_chosen = [i for i in 1:N if value(x[i]; result = 7) &gt; 0.9]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">11-element Vector{Int64}:
+  1
+  2
+  3
+  5
+  6
+  8
+ 10
+ 11
+ 15
+ 16
+ 17</code></pre><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><p><a href="../../../packages/MultiObjectiveAlgorithms/#MultiObjectiveAlgorithms.jl">MultiObjectiveAlgorithms.jl</a> implements a number of different algorithms. Try solving the same problem using <code>MOA.Dichotomy()</code>. Does it find the same solution?</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../transp/">« The transportation problem</a><a class="docs-footer-nextpage" href="../multi_objective_examples/">Simple multi-objective examples »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/n-queens.jl b/previews/PR3525/tutorials/linear/n-queens.jl
new file mode 100644
index 00000000000..36fba85372f
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/n-queens.jl
@@ -0,0 +1,72 @@
+# Copyright (c) 2019 Matthew Help and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # N-Queens
+
+# **This tutorial was originally contributed by Matthew Helm and Mathieu Tanneau.**
+
+# The N-Queens problem involves placing N queens on an N x N chessboard such
+# that none of the queens attacks another. In chess, a queen can move
+# vertically, horizontally, and diagonally so there cannot be more than one
+# queen on any given row, column, or diagonal.
+
+# ![Four Queens](../../assets/n_queens4.png)
+
+# *Note that none of the queens above are able to attack any other as a result
+# of their careful placement.*
+
+using JuMP
+import HiGHS
+import LinearAlgebra
+
+# N-Queens
+N = 8
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+
+# Next, let's create an N x N chessboard of binary values. 0 will represent an
+# empty space on the board and 1 will represent a space occupied by one of our
+# queens:
+
+@variable(model, x[1:N, 1:N], Bin);
+
+# Now we can add our constraints:
+
+# There must be exactly one queen in a given row/column
+for i in 1:N
+    @constraint(model, sum(x[i, :]) == 1)
+    @constraint(model, sum(x[:, i]) == 1)
+end
+
+# There can only be one queen on any given diagonal
+for i in -(N - 1):(N-1)
+    @constraint(model, sum(LinearAlgebra.diag(x, i)) <= 1)
+    @constraint(model, sum(LinearAlgebra.diag(reverse(x; dims = 1), i)) <= 1)
+end
+
+# We are ready to put our model to work and see if it is able to find
+# a feasible solution:
+
+optimize!(model)
+
+# We can now review the solution that our model found:
+
+solution = round.(Int, value.(x))
diff --git a/previews/PR3525/tutorials/linear/n-queens/index.html b/previews/PR3525/tutorials/linear/n-queens/index.html
new file mode 100644
index 00000000000..026291308d5
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/n-queens/index.html
@@ -0,0 +1,25 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>N-Queens · JuMP</title><meta name="title" content="N-Queens · JuMP"/><meta property="og:title" content="N-Queens · JuMP"/><meta property="twitter:title" content="N-Queens · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li class="is-active"><a class="tocitem" href>N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>N-Queens</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>N-Queens</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/n-queens.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="N-Queens"><a class="docs-heading-anchor" href="#N-Queens">N-Queens</a><a id="N-Queens-1"></a><a class="docs-heading-anchor-permalink" href="#N-Queens" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../n-queens.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Matthew Helm and Mathieu Tanneau.</strong></p><p>The N-Queens problem involves placing N queens on an N x N chessboard such that none of the queens attacks another. In chess, a queen can move vertically, horizontally, and diagonally so there cannot be more than one queen on any given row, column, or diagonal.</p><p><img src="../../../assets/n_queens4.png" alt="Four Queens"/></p><p><em>Note that none of the queens above are able to attack any other as a result of their careful placement.</em></p><pre><code class="language-julia hljs">using JuMP
+import HiGHS
+import LinearAlgebra</code></pre><p>N-Queens</p><pre><code class="language-julia hljs">N = 8
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)</code></pre><p>Next, let&#39;s create an N x N chessboard of binary values. 0 will represent an empty space on the board and 1 will represent a space occupied by one of our queens:</p><pre><code class="language-julia hljs">@variable(model, x[1:N, 1:N], Bin);</code></pre><p>Now we can add our constraints:</p><p>There must be exactly one queen in a given row/column</p><pre><code class="language-julia hljs">for i in 1:N
+    @constraint(model, sum(x[i, :]) == 1)
+    @constraint(model, sum(x[:, i]) == 1)
+end</code></pre><p>There can only be one queen on any given diagonal</p><pre><code class="language-julia hljs">for i in -(N - 1):(N-1)
+    @constraint(model, sum(LinearAlgebra.diag(x, i)) &lt;= 1)
+    @constraint(model, sum(LinearAlgebra.diag(reverse(x; dims = 1), i)) &lt;= 1)
+end</code></pre><p>We are ready to put our model to work and see if it is able to find a feasible solution:</p><pre><code class="language-julia hljs">optimize!(model)</code></pre><p>We can now review the solution that our model found:</p><pre><code class="language-julia hljs">solution = round.(Int, value.(x))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">8×8 Matrix{Int64}:
+ 0  0  0  0  1  0  0  0
+ 0  0  1  0  0  0  0  0
+ 0  0  0  0  0  0  0  1
+ 0  0  0  1  0  0  0  0
+ 0  0  0  0  0  0  1  0
+ 1  0  0  0  0  0  0  0
+ 0  0  0  0  0  1  0  0
+ 0  1  0  0  0  0  0  0</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../sudoku/">« Sudoku</a><a class="docs-footer-nextpage" href="../constraint_programming/">Constraint programming »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/network_flows.jl b/previews/PR3525/tutorials/linear/network_flows.jl
new file mode 100644
index 00000000000..b87030a028e
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/network_flows.jl
@@ -0,0 +1,168 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Network flow problems
+
+# **This tutorial was originally contributed by Arpit Bhatia.**
+
+# In graph theory, a flow network (also known as a transportation network) is a
+# directed graph where each edge has a capacity and each edge receives a flow.
+# The amount of flow on an edge cannot exceed the capacity of the edge.
+
+# Often in operations research, a directed graph is called a network, the
+# vertices are called nodes and the edges are called arcs.
+
+# A flow must satisfy the restriction that the amount of flow into a node equals
+# the amount of flow out of it,  unless it is a source, which has only outgoing
+# flow, or sink, which has only incoming flow.
+
+# A network can be used to model traffic in a computer network, circulation with
+# demands, fluids in pipes,  currents in an electrical circuit, or anything
+# similar in which something travels through a network of nodes.
+
+# This tutorial requires the following packages:
+
+using JuMP
+import HiGHS
+
+# ## The shortest path problem
+
+# Suppose that each arc $(i, j)$ of a graph is assigned a scalar cost $a_{i,j}$,
+# and suppose that we define the cost of a forward path to be the sum of the
+# costs of its arcs.
+
+# Given a pair of nodes, the shortest path problem is to find a forward path
+# that connects these nodes and has minimum cost.
+
+# ```math
+# \begin{aligned}
+# \min && \sum_{\forall e(i,j) \in E} a_{i,j} \times x_{i,j} \\
+# s.t. && \sum_j x_{ij} - \sum_k x_{ki} = b_i && \forall i\\
+# && x_{e} \in \{0,1\} && \forall e \in E
+# \end{aligned}
+# ```
+# where ``b_i`` is ``1`` if ``i`` is the starting node, ``-1`` if ``i`` is the
+# ending node, and ``0`` otherwise.
+
+G = [
+    0 100 30 0 0
+    0 0 20 0 0
+    0 0 0 10 60
+    0 15 0 0 50
+    0 0 0 0 0
+]
+n = size(G)[1]
+b = [1, -1, 0, 0, 0]
+shortest_path = Model(HiGHS.Optimizer)
+set_silent(shortest_path)
+@variable(shortest_path, x[1:n, 1:n], Bin)
+## Arcs with zero cost are not a part of the path as they do no exist
+@constraint(shortest_path, [i = 1:n, j = 1:n; G[i, j] == 0], x[i, j] == 0)
+## Flow conservation constraint
+@constraint(shortest_path, [i = 1:n], sum(x[i, :]) - sum(x[:, i]) == b[i],)
+@objective(shortest_path, Min, sum(G .* x))
+optimize!(shortest_path)
+objective_value(shortest_path)
+#-
+value.(x)
+
+# ## The assignment problem
+
+# Suppose that there are $n$ persons and $n$ objects that we have to match on a
+# one-to-one basis. There is a benefit or value $a_{i,j}$ for matching person
+# $i$ with object $j$, and we want to assign persons to objects so as to
+# maximize the total benefit.
+
+# There is also a restriction that person $i$ can be assigned to object $j$ only
+# if $(i, j)$ belongs to a given set of pairs $A$.
+
+# Mathematically, we want to find a set of person-object pairs
+# $(1, j_{1}),..., (n, j_{n})$ from $A$ such that the objects $j_{1},...,j_{n}$
+# are all distinct, and the total benefit $\sum_{i=1}^{y} a_{ij_{i}}$ is
+# maximized.
+
+# ```math
+# \begin{aligned}
+# \max && \sum_{(i,j) \in A} a_{i,j} \times y_{i,j} \\
+# s.t. && \sum_{\{j|(i,j) \in A\}} y_{i,j} = 1 && \forall i = \{1,2....n\} \\
+# && \sum_{\{i|(i,j) \in A\}} y_{i,j} = 1 && \forall j = \{1,2....n\} \\
+# && y_{i,j} \in \{0,1\} && \forall (i,j) \in \{1,2...k\}
+# \end{aligned}
+# ```
+
+G = [
+    6 4 5 0
+    0 3 6 0
+    5 0 4 3
+    7 5 5 5
+]
+n = size(G)[1]
+assignment = Model(HiGHS.Optimizer)
+set_silent(assignment)
+@variable(assignment, y[1:n, 1:n], Bin)
+## One person can only be assigned to one object
+@constraint(assignment, [i = 1:n], sum(y[:, i]) == 1)
+## One object can only be assigned to one person
+@constraint(assignment, [j = 1:n], sum(y[j, :]) == 1)
+@objective(assignment, Max, sum(G .* y))
+optimize!(assignment)
+objective_value(assignment)
+#-
+value.(y)
+
+# ## The max-flow problem
+
+# In the max-flow problem, we have a graph with two special nodes: the $source$,
+# denoted by $s$, and the $sink$, denoted by $t$.
+
+# The objective is to move as much flow as possible from $s$ into $t$ while
+# observing the capacity constraints.
+
+# ```math
+# \begin{aligned}
+# \max && \sum_{v:(s,v) \in E} f(s,v) \\
+# s.t. && \sum_{u:(u,v) \in E} f(u,v)  = \sum_{w:(v,w) \in E} f(v,w) && \forall v \in V - \{s,t\} \\
+# && f(u,v) \leq c(u,v) && \forall (u,v) \in E \\
+# && f(u,v) \geq 0 && \forall (u,v) \in E
+# \end{aligned}
+# ```
+
+G = [
+    0 3 2 2 0 0 0 0
+    0 0 0 0 5 1 0 0
+    0 0 0 0 1 3 1 0
+    0 0 0 0 0 1 0 0
+    0 0 0 0 0 0 0 4
+    0 0 0 0 0 0 0 2
+    0 0 0 0 0 0 0 4
+    0 0 0 0 0 0 0 0
+]
+n = size(G)[1]
+max_flow = Model(HiGHS.Optimizer)
+@variable(max_flow, f[1:n, 1:n] >= 0)
+## Capacity constraints
+@constraint(max_flow, [i = 1:n, j = 1:n], f[i, j] <= G[i, j])
+## Flow conservation constraints
+@constraint(max_flow, [i = 1:n; i != 1 && i != 8], sum(f[i, :]) == sum(f[:, i]))
+@objective(max_flow, Max, sum(f[1, :]))
+optimize!(max_flow)
+objective_value(max_flow)
+#-
+value.(f)
diff --git a/previews/PR3525/tutorials/linear/network_flows/index.html b/previews/PR3525/tutorials/linear/network_flows/index.html
new file mode 100644
index 00000000000..49928887221
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/network_flows/index.html
@@ -0,0 +1,91 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Network flow problems · JuMP</title><meta name="title" content="Network flow problems · JuMP"/><meta property="og:title" content="Network flow problems · JuMP"/><meta property="twitter:title" content="Network flow problems · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li class="is-active"><a class="tocitem" href>Network flow problems</a><ul class="internal"><li><a class="tocitem" href="#The-shortest-path-problem"><span>The shortest path problem</span></a></li><li><a class="tocitem" href="#The-assignment-problem"><span>The assignment problem</span></a></li><li><a class="tocitem" href="#The-max-flow-problem"><span>The max-flow problem</span></a></li></ul></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Network flow problems</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Network flow problems</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/network_flows.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Network-flow-problems"><a class="docs-heading-anchor" href="#Network-flow-problems">Network flow problems</a><a id="Network-flow-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Network-flow-problems" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../network_flows.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Arpit Bhatia.</strong></p><p>In graph theory, a flow network (also known as a transportation network) is a directed graph where each edge has a capacity and each edge receives a flow. The amount of flow on an edge cannot exceed the capacity of the edge.</p><p>Often in operations research, a directed graph is called a network, the vertices are called nodes and the edges are called arcs.</p><p>A flow must satisfy the restriction that the amount of flow into a node equals the amount of flow out of it,  unless it is a source, which has only outgoing flow, or sink, which has only incoming flow.</p><p>A network can be used to model traffic in a computer network, circulation with demands, fluids in pipes,  currents in an electrical circuit, or anything similar in which something travels through a network of nodes.</p><p>This tutorial requires the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import HiGHS</code></pre><h2 id="The-shortest-path-problem"><a class="docs-heading-anchor" href="#The-shortest-path-problem">The shortest path problem</a><a id="The-shortest-path-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-shortest-path-problem" title="Permalink"></a></h2><p>Suppose that each arc <span>$(i, j)$</span> of a graph is assigned a scalar cost <span>$a_{i,j}$</span>, and suppose that we define the cost of a forward path to be the sum of the costs of its arcs.</p><p>Given a pair of nodes, the shortest path problem is to find a forward path that connects these nodes and has minimum cost.</p><p class="math-container">\[\begin{aligned}
+\min &amp;&amp; \sum_{\forall e(i,j) \in E} a_{i,j} \times x_{i,j} \\
+s.t. &amp;&amp; \sum_j x_{ij} - \sum_k x_{ki} = b_i &amp;&amp; \forall i\\
+&amp;&amp; x_{e} \in \{0,1\} &amp;&amp; \forall e \in E
+\end{aligned}\]</p><p>where <span>$b_i$</span> is <span>$1$</span> if <span>$i$</span> is the starting node, <span>$-1$</span> if <span>$i$</span> is the ending node, and <span>$0$</span> otherwise.</p><pre><code class="language-julia hljs">G = [
+    0 100 30 0 0
+    0 0 20 0 0
+    0 0 0 10 60
+    0 15 0 0 50
+    0 0 0 0 0
+]
+n = size(G)[1]
+b = [1, -1, 0, 0, 0]
+shortest_path = Model(HiGHS.Optimizer)
+set_silent(shortest_path)
+@variable(shortest_path, x[1:n, 1:n], Bin)
+# Arcs with zero cost are not a part of the path as they do no exist
+@constraint(shortest_path, [i = 1:n, j = 1:n; G[i, j] == 0], x[i, j] == 0)
+# Flow conservation constraint
+@constraint(shortest_path, [i = 1:n], sum(x[i, :]) - sum(x[:, i]) == b[i],)
+@objective(shortest_path, Min, sum(G .* x))
+optimize!(shortest_path)
+objective_value(shortest_path)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">55.0</code></pre><pre><code class="language-julia hljs">value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5×5 Matrix{Float64}:
+ 0.0  0.0  1.0  0.0  0.0
+ 0.0  0.0  0.0  0.0  0.0
+ 0.0  0.0  0.0  1.0  0.0
+ 0.0  1.0  0.0  0.0  0.0
+ 0.0  0.0  0.0  0.0  0.0</code></pre><h2 id="The-assignment-problem"><a class="docs-heading-anchor" href="#The-assignment-problem">The assignment problem</a><a id="The-assignment-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-assignment-problem" title="Permalink"></a></h2><p>Suppose that there are <span>$n$</span> persons and <span>$n$</span> objects that we have to match on a one-to-one basis. There is a benefit or value <span>$a_{i,j}$</span> for matching person <span>$i$</span> with object <span>$j$</span>, and we want to assign persons to objects so as to maximize the total benefit.</p><p>There is also a restriction that person <span>$i$</span> can be assigned to object <span>$j$</span> only if <span>$(i, j)$</span> belongs to a given set of pairs <span>$A$</span>.</p><p>Mathematically, we want to find a set of person-object pairs <span>$(1, j_{1}),..., (n, j_{n})$</span> from <span>$A$</span> such that the objects <span>$j_{1},...,j_{n}$</span> are all distinct, and the total benefit <span>$\sum_{i=1}^{y} a_{ij_{i}}$</span> is maximized.</p><p class="math-container">\[\begin{aligned}
+\max &amp;&amp; \sum_{(i,j) \in A} a_{i,j} \times y_{i,j} \\
+s.t. &amp;&amp; \sum_{\{j|(i,j) \in A\}} y_{i,j} = 1 &amp;&amp; \forall i = \{1,2....n\} \\
+&amp;&amp; \sum_{\{i|(i,j) \in A\}} y_{i,j} = 1 &amp;&amp; \forall j = \{1,2....n\} \\
+&amp;&amp; y_{i,j} \in \{0,1\} &amp;&amp; \forall (i,j) \in \{1,2...k\}
+\end{aligned}\]</p><pre><code class="language-julia hljs">G = [
+    6 4 5 0
+    0 3 6 0
+    5 0 4 3
+    7 5 5 5
+]
+n = size(G)[1]
+assignment = Model(HiGHS.Optimizer)
+set_silent(assignment)
+@variable(assignment, y[1:n, 1:n], Bin)
+# One person can only be assigned to one object
+@constraint(assignment, [i = 1:n], sum(y[:, i]) == 1)
+# One object can only be assigned to one person
+@constraint(assignment, [j = 1:n], sum(y[j, :]) == 1)
+@objective(assignment, Max, sum(G .* y))
+optimize!(assignment)
+objective_value(assignment)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">20.0</code></pre><pre><code class="language-julia hljs">value.(y)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4×4 Matrix{Float64}:
+ -0.0  1.0  -0.0   0.0
+  0.0  0.0   1.0   0.0
+  1.0  0.0   0.0  -0.0
+ -0.0  0.0   0.0   1.0</code></pre><h2 id="The-max-flow-problem"><a class="docs-heading-anchor" href="#The-max-flow-problem">The max-flow problem</a><a id="The-max-flow-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-max-flow-problem" title="Permalink"></a></h2><p>In the max-flow problem, we have a graph with two special nodes: the <span>$source$</span>, denoted by <span>$s$</span>, and the <span>$sink$</span>, denoted by <span>$t$</span>.</p><p>The objective is to move as much flow as possible from <span>$s$</span> into <span>$t$</span> while observing the capacity constraints.</p><p class="math-container">\[\begin{aligned}
+\max &amp;&amp; \sum_{v:(s,v) \in E} f(s,v) \\
+s.t. &amp;&amp; \sum_{u:(u,v) \in E} f(u,v)  = \sum_{w:(v,w) \in E} f(v,w) &amp;&amp; \forall v \in V - \{s,t\} \\
+&amp;&amp; f(u,v) \leq c(u,v) &amp;&amp; \forall (u,v) \in E \\
+&amp;&amp; f(u,v) \geq 0 &amp;&amp; \forall (u,v) \in E
+\end{aligned}\]</p><pre><code class="language-julia hljs">G = [
+    0 3 2 2 0 0 0 0
+    0 0 0 0 5 1 0 0
+    0 0 0 0 1 3 1 0
+    0 0 0 0 0 1 0 0
+    0 0 0 0 0 0 0 4
+    0 0 0 0 0 0 0 2
+    0 0 0 0 0 0 0 4
+    0 0 0 0 0 0 0 0
+]
+n = size(G)[1]
+max_flow = Model(HiGHS.Optimizer)
+@variable(max_flow, f[1:n, 1:n] &gt;= 0)
+# Capacity constraints
+@constraint(max_flow, [i = 1:n, j = 1:n], f[i, j] &lt;= G[i, j])
+# Flow conservation constraints
+@constraint(max_flow, [i = 1:n; i != 1 &amp;&amp; i != 8], sum(f[i, :]) == sum(f[:, i]))
+@objective(max_flow, Max, sum(f[1, :]))
+optimize!(max_flow)
+objective_value(max_flow)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">6.0</code></pre><pre><code class="language-julia hljs">value.(f)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">8×8 Matrix{Float64}:
+ -0.0   3.0   2.0  1.0  -0.0  -0.0  -0.0  -0.0
+  0.0   0.0   0.0  0.0   2.0   1.0   0.0   0.0
+  0.0   0.0   0.0  0.0   1.0  -0.0   1.0   0.0
+ -0.0  -0.0  -0.0  0.0  -0.0   1.0  -0.0  -0.0
+  0.0   0.0   0.0  0.0   0.0   0.0   0.0   3.0
+  0.0   0.0   0.0  0.0   0.0   0.0   0.0   2.0
+  0.0   0.0   0.0  0.0   0.0   0.0   0.0   1.0
+  0.0   0.0   0.0  0.0   0.0   0.0   0.0   0.0</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../geographic_clustering/">« Geographical clustering</a><a class="docs-footer-nextpage" href="../transp/">The transportation problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/sudoku.jl b/previews/PR3525/tutorials/linear/sudoku.jl
new file mode 100644
index 00000000000..8e2c207188c
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/sudoku.jl
@@ -0,0 +1,212 @@
+# Copyright (c) 2019 Iain Dunning and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Sudoku
+
+# **This tutorial was originally contributed by Iain Dunning.**
+
+# [Sudoku](http://en.wikipedia.org/wiki/Sudoku) is a popular number puzzle. The
+# goal is to place the digits 1 to 9 on a nine-by-nine grid, with some of the
+# digits already filled in. Your solution must satisfy the following rules:
+#
+# * The numbers 1 to 9 must appear in each 3x3 square
+# * The numbers 1 to 9 must appear in each row
+# * The numbers 1 to 9 must appear in each column
+
+# Here is a partially solved Sudoku problem:
+
+# ![Partially solved Sudoku](../../assets/partial_sudoku.png)
+
+# Solving a Sudoku isn't an optimization problem with an objective; its actually
+# a *feasibility* problem: we wish to find a feasible solution that satisfies
+# these rules. You can think of it as an optimization problem with an objective
+# of 0.
+
+# ## Mixed-integer linear programming formulation
+
+# We can model this problem using 0-1 integer programming: a problem where all
+# the decision variables are binary. We'll use JuMP to create the model, and
+# then we can solve it with any integer programming solver.
+
+using JuMP
+using HiGHS
+
+# We will define a binary variable (a variable that is either 0 or 1) for each
+# possible number in each possible cell. The meaning of each variable is as
+# follows:
+# `x[i,j,k] = 1  if and only if cell (i,j) has number k`,
+# where `i` is the row and `j` is the column.
+
+# Create a model
+
+sudoku = Model(HiGHS.Optimizer)
+set_silent(sudoku)
+
+# Create our variables
+@variable(sudoku, x[i = 1:9, j = 1:9, k = 1:9], Bin);
+
+# Now we can begin to add our constraints. We'll actually start with something
+# obvious to us as humans, but what we need to enforce: that there can be only
+# one number per cell.
+
+for i in 1:9  # For each row
+    for j in 1:9  # and each column
+        ## Sum across all the possible digits. One and only one of the digits
+        ## can be in this cell, so the sum must be equal to one.
+        @constraint(sudoku, sum(x[i, j, k] for k in 1:9) == 1)
+    end
+end
+
+# Next we'll add the constraints for the rows and the columns. These constraints
+# are all very similar, so much so that we can actually add them at the same
+# time.
+
+for ind in 1:9  # Each row, OR each column
+    for k in 1:9  # Each digit
+        ## Sum across columns (j) - row constraint
+        @constraint(sudoku, sum(x[ind, j, k] for j in 1:9) == 1)
+        ## Sum across rows (i) - column constraint
+        @constraint(sudoku, sum(x[i, ind, k] for i in 1:9) == 1)
+    end
+end
+
+# Finally, we have the to enforce the constraint that each digit appears once in
+# each of the nine 3x3 sub-grids. Our strategy will be to index over the
+# top-left corners of each 3x3 square with `for` loops, then sum over the
+# squares.
+
+for i in 1:3:7
+    for j in 1:3:7
+        for k in 1:9
+            ## i is the top left row, j is the top left column.
+            ## We'll sum from i to i+2, for example, i=4, r=4, 5, 6.
+            @constraint(
+                sudoku,
+                sum(x[r, c, k] for r in i:(i+2), c in j:(j+2)) == 1
+            )
+        end
+    end
+end
+
+# The final step is to add the initial solution as a set of constraints. We'll
+# solve the problem that is in the picture at the start of the tutorial. We'll
+# put a `0` if there is no digit in that location.
+
+# The given digits
+init_sol = [
+    5 3 0 0 7 0 0 0 0
+    6 0 0 1 9 5 0 0 0
+    0 9 8 0 0 0 0 6 0
+    8 0 0 0 6 0 0 0 3
+    4 0 0 8 0 3 0 0 1
+    7 0 0 0 2 0 0 0 6
+    0 6 0 0 0 0 2 8 0
+    0 0 0 4 1 9 0 0 5
+    0 0 0 0 8 0 0 7 9
+]
+for i in 1:9
+    for j in 1:9
+        ## If the space isn't empty
+        if init_sol[i, j] != 0
+            ## Then the corresponding variable for that digit and location must
+            ## be 1.
+            fix(x[i, j, init_sol[i, j]], 1; force = true)
+        end
+    end
+end
+
+# solve problem
+optimize!(sudoku)
+
+# Extract the values of x
+x_val = value.(x);
+# Create a matrix to store the solution
+sol = zeros(Int, 9, 9)  # 9x9 matrix of integers
+for i in 1:9
+    for j in 1:9
+        for k in 1:9
+            ## Integer programs are solved as a series of linear programs so the
+            ## values might not be precisely 0 and 1. We can round them to
+            ## the nearest integer to make it easier.
+            if round(Int, x_val[i, j, k]) == 1
+                sol[i, j] = k
+            end
+        end
+    end
+end
+
+# Display the solution
+sol
+
+# Which is the correct solution:
+
+# ![Solved Sudoku](../../assets/full_sudoku.png)
+
+# ## Constraint programming formulation
+
+# We can also model this problem using constraint programming and the
+# all-different constraint, which says that no two elements of a vector can take
+# the same value.
+
+# Because of the reformulation system in MathOptInterface, we can still solve
+# this problem using HiGHS.
+
+model = Model(HiGHS.Optimizer)
+set_silent(model)
+## HiGHS v1.2 has a bug in presolve which causes the problem to be classified as
+## infeasible.
+set_attribute(model, "presolve", "off")
+
+# Instead of the binary variables, we directly define a 9x9 grid of integer
+# values between 1 and 9:
+
+@variable(model, 1 <= x[1:9, 1:9] <= 9, Int);
+
+# Then, we enforce that the values in each row must be all-different:
+
+@constraint(model, [i = 1:9], x[i, :] in MOI.AllDifferent(9));
+
+# That the values in each column must be all-different:
+
+@constraint(model, [j = 1:9], x[:, j] in MOI.AllDifferent(9));
+
+# And that the values in each 3x3 sub-grid must be all-different:
+
+for i in (0, 3, 6), j in (0, 3, 6)
+    @constraint(model, vec(x[i.+(1:3), j.+(1:3)]) in MOI.AllDifferent(9))
+end
+
+# Finally, as before we set the initial solution and optimize:
+
+for i in 1:9, j in 1:9
+    if init_sol[i, j] != 0
+        fix(x[i, j], init_sol[i, j]; force = true)
+    end
+end
+
+optimize!(model)
+
+# Display the solution
+
+csp_sol = round.(Int, value.(x))
+
+# Which is the same as we found before:
+
+sol == csp_sol
diff --git a/previews/PR3525/tutorials/linear/sudoku/index.html b/previews/PR3525/tutorials/linear/sudoku/index.html
new file mode 100644
index 00000000000..0d9542bd493
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/sudoku/index.html
@@ -0,0 +1,94 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Sudoku · JuMP</title><meta name="title" content="Sudoku · JuMP"/><meta property="og:title" content="Sudoku · JuMP"/><meta property="twitter:title" content="Sudoku · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li class="is-active"><a class="tocitem" href>Sudoku</a><ul class="internal"><li><a class="tocitem" href="#Mixed-integer-linear-programming-formulation"><span>Mixed-integer linear programming formulation</span></a></li><li><a class="tocitem" href="#Constraint-programming-formulation"><span>Constraint programming formulation</span></a></li></ul></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Sudoku</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Sudoku</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/sudoku.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Sudoku"><a class="docs-heading-anchor" href="#Sudoku">Sudoku</a><a id="Sudoku-1"></a><a class="docs-heading-anchor-permalink" href="#Sudoku" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../sudoku.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Iain Dunning.</strong></p><p><a href="http://en.wikipedia.org/wiki/Sudoku">Sudoku</a> is a popular number puzzle. The goal is to place the digits 1 to 9 on a nine-by-nine grid, with some of the digits already filled in. Your solution must satisfy the following rules:</p><ul><li>The numbers 1 to 9 must appear in each 3x3 square</li><li>The numbers 1 to 9 must appear in each row</li><li>The numbers 1 to 9 must appear in each column</li></ul><p>Here is a partially solved Sudoku problem:</p><p><img src="../../../assets/partial_sudoku.png" alt="Partially solved Sudoku"/></p><p>Solving a Sudoku isn&#39;t an optimization problem with an objective; its actually a <em>feasibility</em> problem: we wish to find a feasible solution that satisfies these rules. You can think of it as an optimization problem with an objective of 0.</p><h2 id="Mixed-integer-linear-programming-formulation"><a class="docs-heading-anchor" href="#Mixed-integer-linear-programming-formulation">Mixed-integer linear programming formulation</a><a id="Mixed-integer-linear-programming-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Mixed-integer-linear-programming-formulation" title="Permalink"></a></h2><p>We can model this problem using 0-1 integer programming: a problem where all the decision variables are binary. We&#39;ll use JuMP to create the model, and then we can solve it with any integer programming solver.</p><pre><code class="language-julia hljs">using JuMP
+using HiGHS</code></pre><p>We will define a binary variable (a variable that is either 0 or 1) for each possible number in each possible cell. The meaning of each variable is as follows: <code>x[i,j,k] = 1  if and only if cell (i,j) has number k</code>, where <code>i</code> is the row and <code>j</code> is the column.</p><p>Create a model</p><pre><code class="language-julia hljs">sudoku = Model(HiGHS.Optimizer)
+set_silent(sudoku)</code></pre><p>Create our variables</p><pre><code class="language-julia hljs">@variable(sudoku, x[i = 1:9, j = 1:9, k = 1:9], Bin);</code></pre><p>Now we can begin to add our constraints. We&#39;ll actually start with something obvious to us as humans, but what we need to enforce: that there can be only one number per cell.</p><pre><code class="language-julia hljs">for i in 1:9  # For each row
+    for j in 1:9  # and each column
+        # Sum across all the possible digits. One and only one of the digits
+        # can be in this cell, so the sum must be equal to one.
+        @constraint(sudoku, sum(x[i, j, k] for k in 1:9) == 1)
+    end
+end</code></pre><p>Next we&#39;ll add the constraints for the rows and the columns. These constraints are all very similar, so much so that we can actually add them at the same time.</p><pre><code class="language-julia hljs">for ind in 1:9  # Each row, OR each column
+    for k in 1:9  # Each digit
+        # Sum across columns (j) - row constraint
+        @constraint(sudoku, sum(x[ind, j, k] for j in 1:9) == 1)
+        # Sum across rows (i) - column constraint
+        @constraint(sudoku, sum(x[i, ind, k] for i in 1:9) == 1)
+    end
+end</code></pre><p>Finally, we have the to enforce the constraint that each digit appears once in each of the nine 3x3 sub-grids. Our strategy will be to index over the top-left corners of each 3x3 square with <code>for</code> loops, then sum over the squares.</p><pre><code class="language-julia hljs">for i in 1:3:7
+    for j in 1:3:7
+        for k in 1:9
+            # i is the top left row, j is the top left column.
+            # We&#39;ll sum from i to i+2, for example, i=4, r=4, 5, 6.
+            @constraint(
+                sudoku,
+                sum(x[r, c, k] for r in i:(i+2), c in j:(j+2)) == 1
+            )
+        end
+    end
+end</code></pre><p>The final step is to add the initial solution as a set of constraints. We&#39;ll solve the problem that is in the picture at the start of the tutorial. We&#39;ll put a <code>0</code> if there is no digit in that location.</p><p>The given digits</p><pre><code class="language-julia hljs">init_sol = [
+    5 3 0 0 7 0 0 0 0
+    6 0 0 1 9 5 0 0 0
+    0 9 8 0 0 0 0 6 0
+    8 0 0 0 6 0 0 0 3
+    4 0 0 8 0 3 0 0 1
+    7 0 0 0 2 0 0 0 6
+    0 6 0 0 0 0 2 8 0
+    0 0 0 4 1 9 0 0 5
+    0 0 0 0 8 0 0 7 9
+]
+for i in 1:9
+    for j in 1:9
+        # If the space isn&#39;t empty
+        if init_sol[i, j] != 0
+            # Then the corresponding variable for that digit and location must
+            # be 1.
+            fix(x[i, j, init_sol[i, j]], 1; force = true)
+        end
+    end
+end</code></pre><p>solve problem</p><pre><code class="language-julia hljs">optimize!(sudoku)</code></pre><p>Extract the values of x</p><pre><code class="language-julia hljs">x_val = value.(x);</code></pre><p>Create a matrix to store the solution</p><pre><code class="language-julia hljs">sol = zeros(Int, 9, 9)  # 9x9 matrix of integers
+for i in 1:9
+    for j in 1:9
+        for k in 1:9
+            # Integer programs are solved as a series of linear programs so the
+            # values might not be precisely 0 and 1. We can round them to
+            # the nearest integer to make it easier.
+            if round(Int, x_val[i, j, k]) == 1
+                sol[i, j] = k
+            end
+        end
+    end
+end</code></pre><p>Display the solution</p><pre><code class="language-julia hljs">sol</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9×9 Matrix{Int64}:
+ 5  3  4  6  7  8  9  1  2
+ 6  7  2  1  9  5  3  4  8
+ 1  9  8  3  4  2  5  6  7
+ 8  5  9  7  6  1  4  2  3
+ 4  2  6  8  5  3  7  9  1
+ 7  1  3  9  2  4  8  5  6
+ 9  6  1  5  3  7  2  8  4
+ 2  8  7  4  1  9  6  3  5
+ 3  4  5  2  8  6  1  7  9</code></pre><p>Which is the correct solution:</p><p><img src="../../../assets/full_sudoku.png" alt="Solved Sudoku"/></p><h2 id="Constraint-programming-formulation"><a class="docs-heading-anchor" href="#Constraint-programming-formulation">Constraint programming formulation</a><a id="Constraint-programming-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Constraint-programming-formulation" title="Permalink"></a></h2><p>We can also model this problem using constraint programming and the all-different constraint, which says that no two elements of a vector can take the same value.</p><p>Because of the reformulation system in MathOptInterface, we can still solve this problem using HiGHS.</p><pre><code class="language-julia hljs">model = Model(HiGHS.Optimizer)
+set_silent(model)
+# HiGHS v1.2 has a bug in presolve which causes the problem to be classified as
+# infeasible.
+set_attribute(model, &quot;presolve&quot;, &quot;off&quot;)</code></pre><p>Instead of the binary variables, we directly define a 9x9 grid of integer values between 1 and 9:</p><pre><code class="language-julia hljs">@variable(model, 1 &lt;= x[1:9, 1:9] &lt;= 9, Int);</code></pre><p>Then, we enforce that the values in each row must be all-different:</p><pre><code class="language-julia hljs">@constraint(model, [i = 1:9], x[i, :] in MOI.AllDifferent(9));</code></pre><p>That the values in each column must be all-different:</p><pre><code class="language-julia hljs">@constraint(model, [j = 1:9], x[:, j] in MOI.AllDifferent(9));</code></pre><p>And that the values in each 3x3 sub-grid must be all-different:</p><pre><code class="language-julia hljs">for i in (0, 3, 6), j in (0, 3, 6)
+    @constraint(model, vec(x[i.+(1:3), j.+(1:3)]) in MOI.AllDifferent(9))
+end</code></pre><p>Finally, as before we set the initial solution and optimize:</p><pre><code class="language-julia hljs">for i in 1:9, j in 1:9
+    if init_sol[i, j] != 0
+        fix(x[i, j], init_sol[i, j]; force = true)
+    end
+end
+
+optimize!(model)</code></pre><p>Display the solution</p><pre><code class="language-julia hljs">csp_sol = round.(Int, value.(x))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">9×9 Matrix{Int64}:
+ 5  3  4  6  7  8  9  1  2
+ 6  7  2  1  9  5  3  4  8
+ 1  9  8  3  4  2  5  6  7
+ 8  5  9  7  6  1  4  2  3
+ 4  2  6  8  5  3  7  9  1
+ 7  1  3  9  2  4  8  5  6
+ 9  6  1  5  3  7  2  8  4
+ 2  8  7  4  1  9  6  3  5
+ 3  4  5  2  8  6  1  7  9</code></pre><p>Which is the same as we found before:</p><pre><code class="language-julia hljs">sol == csp_sol</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">true</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../multi_objective_examples/">« Simple multi-objective examples</a><a class="docs-footer-nextpage" href="../n-queens/">N-Queens »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/tips_and_tricks.jl b/previews/PR3525/tutorials/linear/tips_and_tricks.jl
new file mode 100644
index 00000000000..ab2b5a7e9d2
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/tips_and_tricks.jl
@@ -0,0 +1,330 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # [Tips and tricks](@id linear_tips_and_tricks)
+
+# **This tutorial was originally contributed by Arpit Bhatia.**
+
+# !!! tip
+#     A good source of tips is the [Mosek Modeling Cookbook](https://docs.mosek.com/modeling-cookbook/mio.html).
+
+# This tutorial collates some tips and tricks you can use when formulating
+# mixed-integer programs. It uses the following packages:
+
+using JuMP
+
+# ## Absolute value
+
+# To model the absolute value function ``t \ge |x|``, there are a few options.
+# In all cases, these reformulations only work if you are minimizing ``t``
+# "down" into ``|x|``. They do not work if you are trying to maximize ``|x|``.
+
+# ### Option 1
+
+# This option adds two linear inequality constraints:
+
+model = Model();
+@variable(model, x)
+@variable(model, t)
+@constraint(model, t >= x)
+@constraint(model, t >= -x)
+
+# ### Option 2
+
+# This option uses two non-negative variables and forms expressions for ``x``
+# and ``t``:
+
+model = Model();
+@variable(model, z[1:2] >= 0)
+@expression(model, t, z[1] + z[2])
+@expression(model, x, z[1] - z[2])
+
+# ### Option 3
+
+# This option uses [`MOI.NormOneCone`](@ref) and lets JuMP choose the
+# reformulation:
+
+model = Model();
+@variable(model, x)
+@variable(model, t)
+@constraint(model, [t; x] in MOI.NormOneCone(2))
+
+# ## L1-norm
+
+# To model ``\min ||x||_1``, that is, ``\min \sum\limits_i |x_i|``, use the
+# [`MOI.NormOneCone`](@ref):
+
+model = Model();
+@variable(model, x[1:3])
+@variable(model, t)
+@constraint(model, [t; x] in MOI.NormOneCone(1 + length(x)))
+@objective(model, Min, t)
+
+# ## Infinity-norm
+
+# To model ``\min ||x||_\infty``, that is, ``\min \max\limits_i |x_i|``, use the
+# [`MOI.NormInfinityCone`](@ref):
+
+model = Model();
+@variable(model, x[1:3])
+@variable(model, t)
+@constraint(model, [t; x] in MOI.NormInfinityCone(1 + length(x)))
+@objective(model, Min, t)
+
+# ## Max
+
+# To model ``t \ge \max\{x, y\}``, do:
+
+model = Model();
+@variable(model, t)
+@variable(model, x)
+@variable(model, y)
+@constraint(model, t >= x)
+@constraint(model, t >= y)
+
+# This reformulation does not work for ``t \ge \min\{x, y\}``.
+
+# ## Min
+
+# To model ``t \le \min\{x, y\}``, do:
+
+model = Model();
+@variable(model, t)
+@variable(model, x)
+@variable(model, y)
+@constraint(model, t <= x)
+@constraint(model, t <= y)
+
+# This reformulation does not work for ``t \le \max\{x, y\}``.
+
+# ## Boolean operators
+
+# Binary variables can be used to construct logical operators. Here are some
+# example.
+
+# ### Or
+
+# $$x_3 = x_1 \lor x_2$$
+
+model = Model();
+@variable(model, x[1:3], Bin)
+@constraints(model, begin
+    x[1] <= x[3]
+    x[2] <= x[3]
+    x[3] <= x[1] + x[2]
+end)
+
+# ### And
+
+# $$x_3 = x_1 \land x_2$$
+
+model = Model();
+@variable(model, x[1:3], Bin)
+@constraints(model, begin
+    x[3] <= x[1]
+    x[3] <= x[2]
+    x[3] >= x[1] + x[2] - 1
+end)
+
+# ### Not
+
+# $$x_1 \neg x_2$$
+
+model = Model();
+@variable(model, x[1:2], Bin)
+@constraint(model, x[1] == 1 - x[2])
+
+# ### Implies
+
+# $$x_1 \implies x_2$$
+
+model = Model();
+@variable(model, x[1:2], Bin)
+@constraint(model, x[1] <= x[2])
+
+# ## Disjunctions
+
+# ### Problem
+
+# Suppose that we have two constraints $a^\top x \leq b$ and
+# $c^\top x \leq d$, and we want at least one to hold.
+
+# ### Trick
+
+# Introduce a "big-M" multiplied by a binary variable to relax one of
+# the constraints.
+
+# **Example** Either $x_1 \leq 1$ or $x_2 \leq 2$.
+
+model = Model();
+@variable(model, x[1:2])
+@variable(model, y, Bin)
+M = 100
+@constraint(model, x[1] <= 1 + M * y)
+@constraint(model, x[2] <= 2 + M * (1 - y))
+
+# !!! warning
+#     If `M` is too small, the solution may be suboptimal. If `M` is too big,
+#     the solver may encounter numerical issues. Try to use domain knowledge to
+#     choose an `M` that is just right. Gurobi has a [good documentation section](https://www.gurobi.com/documentation/9.1/refman/dealing_with_big_m_constra.html)
+#     on this topic.
+
+# ## Indicator constraints
+
+# ### Problem
+
+# Suppose we want to model that a certain linear inequality must be satisfied
+# when some other event occurs, that is, for a binary variable $z$, we want to
+# model the implication:
+
+# $$z = 1 \implies a^\top x \leq b$$
+
+# ### Trick 1
+
+# Some solvers have native support for indicator constraints.
+
+# **Example** $x_1 + x_2 \leq 1$ if $z = 1$.
+
+model = Model();
+@variable(model, x[1:2])
+@variable(model, z, Bin)
+@constraint(model, z => {sum(x) <= 1})
+
+# **Example** $x_1 + x_2 \leq 1$ if $z = 0$.
+
+model = Model();
+@variable(model, x[1:2])
+@variable(model, z, Bin)
+@constraint(model, !z => {sum(x) <= 1})
+
+# ### Trick 2
+
+# If the solver doesn't support indicator constraints, you an use the big-M
+# trick.
+
+# **Example** $x_1 + x_2 \leq 1$ if $z = 1$.
+
+model = Model();
+@variable(model, x[1:2])
+@variable(model, z, Bin)
+M = 100
+@constraint(model, sum(x) <= 1 + M * (1 - z))
+
+# **Example** $x_1 + x_2 \leq 1$ if $z = 0$.
+
+model = Model();
+@variable(model, x[1:2])
+@variable(model, z, Bin)
+M = 100
+@constraint(model, sum(x) <= 1 + M * z)
+
+# ## Semi-continuous variables
+
+# !!! info
+#     This section uses sets from [MathOptInterface](@ref moi_documentation).
+#     By default, JuMP exports the `MOI` symbol as an alias for the
+#     MathOptInterface.jl package. We recommend making this more explicit in
+#     your code by adding the following lines:
+#     ```julia
+#     import MathOptInterface as MOI
+#     ```
+
+# A semi-continuous variable is a continuous variable between bounds $[l,u]$
+# that also can assume the value zero, that is:
+# $$x \in \{0\} \cup [l,u].$$
+
+# **Example** $$x \in \{0\}\cup [1, 2]$$
+
+model = Model();
+@variable(model, x in Semicontinuous(1.0, 2.0))
+
+# ## Semi-integer variables
+
+# A semi-integer variable is a variable which assumes integer values between
+# bounds $[l,u]$ and can also assume the value zero:
+# $$x \in \{0\} \cup [l, u] \cap \mathbb{Z}.$$
+
+model = Model();
+@variable(model, x in Semiinteger(5.0, 10.0))
+
+# ## Special Ordered Sets of Type 1
+
+# A Special Ordered Set of Type 1 is a set of variables, at most one of which
+# can take a non-zero value, all others being at 0.
+#
+# They most frequently apply where a set of variables are actually binary
+# variables. In other words, we have to choose at most one from a set of
+# possibilities.
+
+model = Model();
+@variable(model, x[1:3], Bin)
+@constraint(model, x in SOS1())
+
+# You can optionally pass `SOS1` a weight vector like
+
+@constraint(model, x in SOS1([0.2, 0.5, 0.3]))
+
+# If the decision variables are related and have a physical ordering, then the
+# weight vector, although not used directly in the constraint, can help the
+# solver make a better decision in the solution process.
+
+# ## [Special Ordered Sets of Type 2](@id tip_sos2)
+
+# A Special Ordered Set of type 2 is a set of non-negative variables, of which
+# at most two can be non-zero, and if two are non-zero these must be consecutive
+# in their ordering.
+
+model = Model();
+@variable(model, x[1:3])
+@constraint(model, x in SOS2([3.0, 1.0, 2.0]))
+
+# The ordering provided by the weight vector is more important in this case as
+# the variables need to be consecutive according to the ordering.
+# For example, in the above constraint, the possible pairs are:
+# * Consecutive
+#   * (`x[1]` and `x[3]`) as they correspond to 3 and 2 resp. and thus can be non-zero
+#   * (`x[2]` and `x[3]`) as they correspond to 1 and 2 resp. and thus can be non-zero
+# * Non-consecutive
+#   * (`x[1]` and `x[2]`) as they correspond to 3 and 1 resp. and thus cannot be non-zero
+
+# ## Piecewise linear approximations
+
+# [SOSII constraints](@ref tip_sos2) are most often used to form piecewise
+# linear approximations of a function.
+
+# Given a set of points for `x`:
+x̂ = -1:0.5:2
+# and a set of corresponding points for `y`:
+ŷ = x̂ .^ 2
+# the piecewise linear approximation is constructed by representing `x` and `y`
+# as convex combinations of `x̂` and `ŷ`.
+
+N = length(x̂)
+model = Model();
+@variable(model, -1 <= x <= 2)
+@variable(model, y)
+@variable(model, 0 <= λ[1:N] <= 1)
+@objective(model, Max, y)
+@constraints(model, begin
+    x == sum(x̂[i] * λ[i] for i in 1:N)
+    y == sum(ŷ[i] * λ[i] for i in 1:N)
+    sum(λ) == 1
+    λ in SOS2()
+end)
diff --git a/previews/PR3525/tutorials/linear/tips_and_tricks/index.html b/previews/PR3525/tutorials/linear/tips_and_tricks/index.html
new file mode 100644
index 00000000000..12cec92d4f0
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/tips_and_tricks/index.html
@@ -0,0 +1,67 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Tips and tricks · JuMP</title><meta name="title" content="Tips and tricks · JuMP"/><meta property="og:title" content="Tips and tricks · JuMP"/><meta property="twitter:title" content="Tips and tricks · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li class="is-active"><a class="tocitem" href>Tips and tricks</a><ul class="internal"><li><a class="tocitem" href="#Absolute-value"><span>Absolute value</span></a></li><li><a class="tocitem" href="#L1-norm"><span>L1-norm</span></a></li><li><a class="tocitem" href="#Infinity-norm"><span>Infinity-norm</span></a></li><li><a class="tocitem" href="#Max"><span>Max</span></a></li><li><a class="tocitem" href="#Min"><span>Min</span></a></li><li><a class="tocitem" href="#Boolean-operators"><span>Boolean operators</span></a></li><li><a class="tocitem" href="#Disjunctions"><span>Disjunctions</span></a></li><li><a class="tocitem" href="#Indicator-constraints"><span>Indicator constraints</span></a></li><li><a class="tocitem" href="#Semi-continuous-variables"><span>Semi-continuous variables</span></a></li><li><a class="tocitem" href="#Semi-integer-variables"><span>Semi-integer variables</span></a></li><li><a class="tocitem" href="#Special-Ordered-Sets-of-Type-1"><span>Special Ordered Sets of Type 1</span></a></li><li><a class="tocitem" href="#tip_sos2"><span>Special Ordered Sets of Type 2</span></a></li><li><a class="tocitem" href="#Piecewise-linear-approximations"><span>Piecewise linear approximations</span></a></li></ul></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../transp/">The transportation problem</a></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>Tips and tricks</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Tips and tricks</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/tips_and_tricks.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="linear_tips_and_tricks"><a class="docs-heading-anchor" href="#linear_tips_and_tricks">Tips and tricks</a><a id="linear_tips_and_tricks-1"></a><a class="docs-heading-anchor-permalink" href="#linear_tips_and_tricks" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../tips_and_tricks.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Arpit Bhatia.</strong></p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>A good source of tips is the <a href="https://docs.mosek.com/modeling-cookbook/mio.html">Mosek Modeling Cookbook</a>.</p></div></div><p>This tutorial collates some tips and tricks you can use when formulating mixed-integer programs. It uses the following packages:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; using JuMP</code><code class="nohighlight hljs ansi" style="display:block;"></code></pre><h2 id="Absolute-value"><a class="docs-heading-anchor" href="#Absolute-value">Absolute value</a><a id="Absolute-value-1"></a><a class="docs-heading-anchor-permalink" href="#Absolute-value" title="Permalink"></a></h2><p>To model the absolute value function <span>$t \ge |x|$</span>, there are a few options. In all cases, these reformulations only work if you are minimizing <span>$t$</span> &quot;down&quot; into <span>$|x|$</span>. They do not work if you are trying to maximize <span>$|x|$</span>.</p><h3 id="Option-1"><a class="docs-heading-anchor" href="#Option-1">Option 1</a><a id="Option-1-1"></a><a class="docs-heading-anchor-permalink" href="#Option-1" title="Permalink"></a></h3><p>This option adds two linear inequality constraints:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, t &gt;= x)</code><code class="nohighlight hljs ansi" style="display:block;">-x + t ≥ 0</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, t &gt;= -x)</code><code class="nohighlight hljs ansi" style="display:block;">x + t ≥ 0</code></pre><h3 id="Option-2"><a class="docs-heading-anchor" href="#Option-2">Option 2</a><a id="Option-2-1"></a><a class="docs-heading-anchor-permalink" href="#Option-2" title="Permalink"></a></h3><p>This option uses two non-negative variables and forms expressions for <span>$x$</span> and <span>$t$</span>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, z[1:2] &gt;= 0)</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ z[1]
+ z[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @expression(model, t, z[1] + z[2])</code><code class="nohighlight hljs ansi" style="display:block;">z[1] + z[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @expression(model, x, z[1] - z[2])</code><code class="nohighlight hljs ansi" style="display:block;">z[1] - z[2]</code></pre><h3 id="Option-3"><a class="docs-heading-anchor" href="#Option-3">Option 3</a><a id="Option-3-1"></a><a class="docs-heading-anchor-permalink" href="#Option-3" title="Permalink"></a></h3><p>This option uses <a href="../../../moi/reference/standard_form/#MathOptInterface.NormOneCone"><code>MOI.NormOneCone</code></a> and lets JuMP choose the reformulation:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, [t; x] in MOI.NormOneCone(2))</code><code class="nohighlight hljs ansi" style="display:block;">[t, x] ∈ MathOptInterface.NormOneCone(2)</code></pre><h2 id="L1-norm"><a class="docs-heading-anchor" href="#L1-norm">L1-norm</a><a id="L1-norm-1"></a><a class="docs-heading-anchor-permalink" href="#L1-norm" title="Permalink"></a></h2><p>To model <span>$\min ||x||_1$</span>, that is, <span>$\min \sum\limits_i |x_i|$</span>, use the <a href="../../../moi/reference/standard_form/#MathOptInterface.NormOneCone"><code>MOI.NormOneCone</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:3])</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, [t; x] in MOI.NormOneCone(1 + length(x)))</code><code class="nohighlight hljs ansi" style="display:block;">[t, x[1], x[2], x[3]] ∈ MathOptInterface.NormOneCone(4)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code></pre><h2 id="Infinity-norm"><a class="docs-heading-anchor" href="#Infinity-norm">Infinity-norm</a><a id="Infinity-norm-1"></a><a class="docs-heading-anchor-permalink" href="#Infinity-norm" title="Permalink"></a></h2><p>To model <span>$\min ||x||_\infty$</span>, that is, <span>$\min \max\limits_i |x_i|$</span>, use the <a href="../../../moi/reference/standard_form/#MathOptInterface.NormInfinityCone"><code>MOI.NormInfinityCone</code></a>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:3])</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, [t; x] in MOI.NormInfinityCone(1 + length(x)))</code><code class="nohighlight hljs ansi" style="display:block;">[t, x[1], x[2], x[3]] ∈ MathOptInterface.NormInfinityCone(4)</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Min, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code></pre><h2 id="Max"><a class="docs-heading-anchor" href="#Max">Max</a><a id="Max-1"></a><a class="docs-heading-anchor-permalink" href="#Max" title="Permalink"></a></h2><p>To model <span>$t \ge \max\{x, y\}$</span>, do:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, y)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, t &gt;= x)</code><code class="nohighlight hljs ansi" style="display:block;">t - x ≥ 0</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, t &gt;= y)</code><code class="nohighlight hljs ansi" style="display:block;">t - y ≥ 0</code></pre><p>This reformulation does not work for <span>$t \ge \min\{x, y\}$</span>.</p><h2 id="Min"><a class="docs-heading-anchor" href="#Min">Min</a><a id="Min-1"></a><a class="docs-heading-anchor-permalink" href="#Min" title="Permalink"></a></h2><p>To model <span>$t \le \min\{x, y\}$</span>, do:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, t)</code><code class="nohighlight hljs ansi" style="display:block;">t</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, y)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, t &lt;= x)</code><code class="nohighlight hljs ansi" style="display:block;">t - x ≤ 0</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, t &lt;= y)</code><code class="nohighlight hljs ansi" style="display:block;">t - y ≤ 0</code></pre><p>This reformulation does not work for <span>$t \le \max\{x, y\}$</span>.</p><h2 id="Boolean-operators"><a class="docs-heading-anchor" href="#Boolean-operators">Boolean operators</a><a id="Boolean-operators-1"></a><a class="docs-heading-anchor-permalink" href="#Boolean-operators" title="Permalink"></a></h2><p>Binary variables can be used to construct logical operators. Here are some example.</p><h3 id="Or"><a class="docs-heading-anchor" href="#Or">Or</a><a id="Or-1"></a><a class="docs-heading-anchor-permalink" href="#Or" title="Permalink"></a></h3><p class="math-container">\[x_3 = x_1 \lor x_2\]</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:3], Bin)</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraints(model, begin
+           x[1] &lt;= x[3]
+           x[2] &lt;= x[3]
+           x[3] &lt;= x[1] + x[2]
+       end)</code><code class="nohighlight hljs ansi" style="display:block;">(x[1] - x[3] ≤ 0, x[2] - x[3] ≤ 0, -x[1] - x[2] + x[3] ≤ 0)</code></pre><h3 id="And"><a class="docs-heading-anchor" href="#And">And</a><a id="And-1"></a><a class="docs-heading-anchor-permalink" href="#And" title="Permalink"></a></h3><p class="math-container">\[x_3 = x_1 \land x_2\]</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:3], Bin)</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraints(model, begin
+           x[3] &lt;= x[1]
+           x[3] &lt;= x[2]
+           x[3] &gt;= x[1] + x[2] - 1
+       end)</code><code class="nohighlight hljs ansi" style="display:block;">(-x[1] + x[3] ≤ 0, -x[2] + x[3] ≤ 0, -x[1] - x[2] + x[3] ≥ -1)</code></pre><h3 id="Not"><a class="docs-heading-anchor" href="#Not">Not</a><a id="Not-1"></a><a class="docs-heading-anchor-permalink" href="#Not" title="Permalink"></a></h3><p class="math-container">\[x_1 \neg x_2\]</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2], Bin)</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, x[1] == 1 - x[2])</code><code class="nohighlight hljs ansi" style="display:block;">x[1] + x[2] = 1</code></pre><h3 id="Implies"><a class="docs-heading-anchor" href="#Implies">Implies</a><a id="Implies-1"></a><a class="docs-heading-anchor-permalink" href="#Implies" title="Permalink"></a></h3><p class="math-container">\[x_1 \implies x_2\]</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2], Bin)</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, x[1] &lt;= x[2])</code><code class="nohighlight hljs ansi" style="display:block;">x[1] - x[2] ≤ 0</code></pre><h2 id="Disjunctions"><a class="docs-heading-anchor" href="#Disjunctions">Disjunctions</a><a id="Disjunctions-1"></a><a class="docs-heading-anchor-permalink" href="#Disjunctions" title="Permalink"></a></h2><h3 id="Problem"><a class="docs-heading-anchor" href="#Problem">Problem</a><a id="Problem-1"></a><a class="docs-heading-anchor-permalink" href="#Problem" title="Permalink"></a></h3><p>Suppose that we have two constraints <span>$a^\top x \leq b$</span> and <span>$c^\top x \leq d$</span>, and we want at least one to hold.</p><h3 id="Trick"><a class="docs-heading-anchor" href="#Trick">Trick</a><a id="Trick-1"></a><a class="docs-heading-anchor-permalink" href="#Trick" title="Permalink"></a></h3><p>Introduce a &quot;big-M&quot; multiplied by a binary variable to relax one of the constraints.</p><p><strong>Example</strong> Either <span>$x_1 \leq 1$</span> or <span>$x_2 \leq 2$</span>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2])</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, y, Bin)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; M = 100</code><code class="nohighlight hljs ansi" style="display:block;">100</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, x[1] &lt;= 1 + M * y)</code><code class="nohighlight hljs ansi" style="display:block;">x[1] - 100 y ≤ 1</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, x[2] &lt;= 2 + M * (1 - y))</code><code class="nohighlight hljs ansi" style="display:block;">x[2] + 100 y ≤ 102</code></pre><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>If <code>M</code> is too small, the solution may be suboptimal. If <code>M</code> is too big, the solver may encounter numerical issues. Try to use domain knowledge to choose an <code>M</code> that is just right. Gurobi has a <a href="https://www.gurobi.com/documentation/9.1/refman/dealing_with_big_m_constra.html">good documentation section</a> on this topic.</p></div></div><h2 id="Indicator-constraints"><a class="docs-heading-anchor" href="#Indicator-constraints">Indicator constraints</a><a id="Indicator-constraints-1"></a><a class="docs-heading-anchor-permalink" href="#Indicator-constraints" title="Permalink"></a></h2><h3 id="Problem-2"><a class="docs-heading-anchor" href="#Problem-2">Problem</a><a class="docs-heading-anchor-permalink" href="#Problem-2" title="Permalink"></a></h3><p>Suppose we want to model that a certain linear inequality must be satisfied when some other event occurs, that is, for a binary variable <span>$z$</span>, we want to model the implication:</p><p class="math-container">\[z = 1 \implies a^\top x \leq b\]</p><h3 id="Trick-1"><a class="docs-heading-anchor" href="#Trick-1">Trick 1</a><a id="Trick-1-1"></a><a class="docs-heading-anchor-permalink" href="#Trick-1" title="Permalink"></a></h3><p>Some solvers have native support for indicator constraints.</p><p><strong>Example</strong> <span>$x_1 + x_2 \leq 1$</span> if <span>$z = 1$</span>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2])</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, z, Bin)</code><code class="nohighlight hljs ansi" style="display:block;">z</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, z =&gt; {sum(x) &lt;= 1})</code><code class="nohighlight hljs ansi" style="display:block;">z =&gt; {x[1] + x[2] ≤ 1}</code></pre><p><strong>Example</strong> <span>$x_1 + x_2 \leq 1$</span> if <span>$z = 0$</span>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2])</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, z, Bin)</code><code class="nohighlight hljs ansi" style="display:block;">z</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, !z =&gt; {sum(x) &lt;= 1})</code><code class="nohighlight hljs ansi" style="display:block;">!z =&gt; {x[1] + x[2] ≤ 1}</code></pre><h3 id="Trick-2"><a class="docs-heading-anchor" href="#Trick-2">Trick 2</a><a id="Trick-2-1"></a><a class="docs-heading-anchor-permalink" href="#Trick-2" title="Permalink"></a></h3><p>If the solver doesn&#39;t support indicator constraints, you an use the big-M trick.</p><p><strong>Example</strong> <span>$x_1 + x_2 \leq 1$</span> if <span>$z = 1$</span>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2])</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, z, Bin)</code><code class="nohighlight hljs ansi" style="display:block;">z</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; M = 100</code><code class="nohighlight hljs ansi" style="display:block;">100</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, sum(x) &lt;= 1 + M * (1 - z))</code><code class="nohighlight hljs ansi" style="display:block;">x[1] + x[2] + 100 z ≤ 101</code></pre><p><strong>Example</strong> <span>$x_1 + x_2 \leq 1$</span> if <span>$z = 0$</span>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:2])</code><code class="nohighlight hljs ansi" style="display:block;">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, z, Bin)</code><code class="nohighlight hljs ansi" style="display:block;">z</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; M = 100</code><code class="nohighlight hljs ansi" style="display:block;">100</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, sum(x) &lt;= 1 + M * z)</code><code class="nohighlight hljs ansi" style="display:block;">x[1] + x[2] - 100 z ≤ 1</code></pre><h2 id="Semi-continuous-variables"><a class="docs-heading-anchor" href="#Semi-continuous-variables">Semi-continuous variables</a><a id="Semi-continuous-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Semi-continuous-variables" title="Permalink"></a></h2><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>This section uses sets from <a href="../../../moi/#moi_documentation">MathOptInterface</a>. By default, JuMP exports the <code>MOI</code> symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre></div></div><p>A semi-continuous variable is a continuous variable between bounds <span>$[l,u]$</span> that also can assume the value zero, that is: <span>$x \in \{0\} \cup [l,u].$</span></p><p><strong>Example</strong> <span>$x \in \{0\}\cup [1, 2]$</span></p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x in Semicontinuous(1.0, 2.0))</code><code class="nohighlight hljs ansi" style="display:block;">x</code></pre><h2 id="Semi-integer-variables"><a class="docs-heading-anchor" href="#Semi-integer-variables">Semi-integer variables</a><a id="Semi-integer-variables-1"></a><a class="docs-heading-anchor-permalink" href="#Semi-integer-variables" title="Permalink"></a></h2><p>A semi-integer variable is a variable which assumes integer values between bounds <span>$[l,u]$</span> and can also assume the value zero: <span>$x \in \{0\} \cup [l, u] \cap \mathbb{Z}.$</span></p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x in Semiinteger(5.0, 10.0))</code><code class="nohighlight hljs ansi" style="display:block;">x</code></pre><h2 id="Special-Ordered-Sets-of-Type-1"><a class="docs-heading-anchor" href="#Special-Ordered-Sets-of-Type-1">Special Ordered Sets of Type 1</a><a id="Special-Ordered-Sets-of-Type-1-1"></a><a class="docs-heading-anchor-permalink" href="#Special-Ordered-Sets-of-Type-1" title="Permalink"></a></h2><p>A Special Ordered Set of Type 1 is a set of variables, at most one of which can take a non-zero value, all others being at 0.</p><p>They most frequently apply where a set of variables are actually binary variables. In other words, we have to choose at most one from a set of possibilities.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:3], Bin)</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, x in SOS1())</code><code class="nohighlight hljs ansi" style="display:block;">[x[1], x[2], x[3]] ∈ MathOptInterface.SOS1{Float64}([1.0, 2.0, 3.0])</code></pre><p>You can optionally pass <code>SOS1</code> a weight vector like</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, x in SOS1([0.2, 0.5, 0.3]))</code><code class="nohighlight hljs ansi" style="display:block;">[x[1], x[2], x[3]] ∈ MathOptInterface.SOS1{Float64}([0.2, 0.5, 0.3])</code></pre><p>If the decision variables are related and have a physical ordering, then the weight vector, although not used directly in the constraint, can help the solver make a better decision in the solution process.</p><h2 id="tip_sos2"><a class="docs-heading-anchor" href="#tip_sos2">Special Ordered Sets of Type 2</a><a id="tip_sos2-1"></a><a class="docs-heading-anchor-permalink" href="#tip_sos2" title="Permalink"></a></h2><p>A Special Ordered Set of type 2 is a set of non-negative variables, of which at most two can be non-zero, and if two are non-zero these must be consecutive in their ordering.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, x[1:3])</code><code class="nohighlight hljs ansi" style="display:block;">3-element Vector{VariableRef}:
+ x[1]
+ x[2]
+ x[3]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraint(model, x in SOS2([3.0, 1.0, 2.0]))</code><code class="nohighlight hljs ansi" style="display:block;">[x[1], x[2], x[3]] ∈ MathOptInterface.SOS2{Float64}([3.0, 1.0, 2.0])</code></pre><p>The ordering provided by the weight vector is more important in this case as the variables need to be consecutive according to the ordering. For example, in the above constraint, the possible pairs are:</p><ul><li>Consecutive<ul><li>(<code>x[1]</code> and <code>x[3]</code>) as they correspond to 3 and 2 resp. and thus can be non-zero</li><li>(<code>x[2]</code> and <code>x[3]</code>) as they correspond to 1 and 2 resp. and thus can be non-zero</li></ul></li><li>Non-consecutive<ul><li>(<code>x[1]</code> and <code>x[2]</code>) as they correspond to 3 and 1 resp. and thus cannot be non-zero</li></ul></li></ul><h2 id="Piecewise-linear-approximations"><a class="docs-heading-anchor" href="#Piecewise-linear-approximations">Piecewise linear approximations</a><a id="Piecewise-linear-approximations-1"></a><a class="docs-heading-anchor-permalink" href="#Piecewise-linear-approximations" title="Permalink"></a></h2><p><a href="#tip_sos2">SOSII constraints</a> are most often used to form piecewise linear approximations of a function.</p><p>Given a set of points for <code>x</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; x̂ = -1:0.5:2</code><code class="nohighlight hljs ansi" style="display:block;">-1.0:0.5:2.0</code></pre><p>and a set of corresponding points for <code>y</code>:</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; ŷ = x̂ .^ 2</code><code class="nohighlight hljs ansi" style="display:block;">7-element Vector{Float64}:
+ 1.0
+ 0.25
+ 0.0
+ 0.25
+ 1.0
+ 2.25
+ 4.0</code></pre><p>the piecewise linear approximation is constructed by representing <code>x</code> and <code>y</code> as convex combinations of <code>x̂</code> and <code>ŷ</code>.</p><pre><code class="language-julia-repl hljs" style="display:block;">julia&gt; N = length(x̂)</code><code class="nohighlight hljs ansi" style="display:block;">7</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; model = Model();</code><code class="nohighlight hljs ansi" style="display:block;"></code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, -1 &lt;= x &lt;= 2)</code><code class="nohighlight hljs ansi" style="display:block;">x</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, y)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @variable(model, 0 &lt;= λ[1:N] &lt;= 1)</code><code class="nohighlight hljs ansi" style="display:block;">7-element Vector{VariableRef}:
+ λ[1]
+ λ[2]
+ λ[3]
+ λ[4]
+ λ[5]
+ λ[6]
+ λ[7]</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @objective(model, Max, y)</code><code class="nohighlight hljs ansi" style="display:block;">y</code><br/><code class="language-julia-repl hljs" style="display:block;">julia&gt; @constraints(model, begin
+           x == sum(x̂[i] * λ[i] for i in 1:N)
+           y == sum(ŷ[i] * λ[i] for i in 1:N)
+           sum(λ) == 1
+           λ in SOS2()
+       end)</code><code class="nohighlight hljs ansi" style="display:block;">(x + λ[1] + 0.5 λ[2] - 0.5 λ[4] - λ[5] - 1.5 λ[6] - 2 λ[7] = 0, y - λ[1] - 0.25 λ[2] - 0.25 λ[4] - λ[5] - 2.25 λ[6] - 4 λ[7] = 0, λ[1] + λ[2] + λ[3] + λ[4] + λ[5] + λ[6] + λ[7] = 1, [λ[1], λ[2], λ[3], λ[4], λ[5], λ[6], λ[7]] ∈ MathOptInterface.SOS2{Float64}([1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0]))</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../multi_commodity_network/">« The network multi-commodity flow problem</a><a class="docs-footer-nextpage" href="../facility_location/">The facility location problem »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/linear/transp.jl b/previews/PR3525/tutorials/linear/transp.jl
new file mode 100644
index 00000000000..dd5ab0e253e
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/transp.jl
@@ -0,0 +1,142 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # The transportation problem
+
+# **This tutorial was originally contributed by Louis Luangkesorn.**
+
+# This tutorial is an adaptation of the transportation problem described in
+# [_AMPL: A Modeling Language for Mathematical Programming_](https://ampl.com/resources/the-ampl-book/),
+# by R. Fourer, D.M. Gay and B.W. Kernighan.
+
+# The purpose of this tutorial is to demonstrate how to create a JuMP model from
+# an ad-hoc structured text file.
+
+# ## Required packages
+
+# This tutorial uses the following packages:
+
+using JuMP
+import DelimitedFiles
+import HiGHS
+
+# ## Formulation
+
+# Suppose that we have a set of factories that produce [pogo sticks](https://en.wikipedia.org/wiki/Pogo_stick),
+# and a set of retail stores in which to sell them. Each factory has a maximum
+# number of pogo sticks that it can produce, and each retail store has a demand
+# of pogo sticks that it can sell.
+
+# In the transportation problem, we want to choose the number of pogo sticks to
+# make and ship from each factory to each retail store that minimizes the total
+# shipping cost.
+
+# Mathematically, we represent our set of factories by a set of origins
+# ``i \in O`` and our retail stores by a set of destinations ``j \in D``. The
+# maximum supply at each factory is ``s_i`` and the demand from each retail
+# store is ``d_j``. The cost of shipping one pogo stick from ``i`` to ``j`` is
+# ``c_{i,j}``.
+
+# With a little effort, we can model the transportation problem as the following
+# linear program:
+# ```math
+# \begin{aligned}
+# \min  && \sum_{i \in O, j \in D} c_{i,j} x_{i,j} \\
+# s.t.  && \sum_{j \in D} x_{i, j} \le s_i && \forall i \in O \\
+#       && \sum_{i \in O} x_{i, j} = d_j && \forall j \in D \\
+#       && x_{i, j} \ge 0 && \forall i \in O, j \in D
+# \end{aligned}
+# ```
+
+# ## Data
+
+# We assume our data is in the form of a text file that has the following form.
+# In practice, we would obtain this text file from the user as input, but for
+# the purpose of this tutorial we're going to create it from Julia.
+
+open(joinpath(@__DIR__, "transp.txt"), "w") do io
+    print(
+        io,
+        """
+             . FRA  DET LAN WIN  STL  FRE  LAF SUPPLY
+          GARY  39   14  11  14   16   82    8   1400
+          CLEV  27    .  12   .   26   95   17   2600
+          PITT  24   14  17  13   28   99   20   2900
+        DEMAND 900 1200 600 400 1700 1100 1000      0
+        """,
+    )
+    return
+end
+
+# Here the rows are the origins, the columns are the destinations, and the
+# values are the cost of shipping one pogo stick from the origin to the
+# destination. If pogo stick cannot be transported from a source to a
+# destination, then the value is `.`. The final row and final column are the
+# demand and supply of each location respectively.
+
+# We didn't account for arcs which do not exist in our formulation, but we can
+# make a small change and fix ``x_{i,j} = 0`` if ``c_{i,j} = "."``.
+
+# Our first step is to convert this text format into an appropriate Julia
+# datastructure that we can work with. Since our data is tabular with named
+# rows and columns, one option is JuMP's [`Containers.DenseAxisArray`](@ref)
+# object:
+
+function read_data(filename::String)
+    data = DelimitedFiles.readdlm(filename)
+    rows, columns = data[2:end, 1], data[1, 2:end]
+    return Containers.DenseAxisArray(data[2:end, 2:end], rows, columns)
+end
+
+data = read_data(joinpath(@__DIR__, "transp.txt"))
+
+# ## JuMP formulation
+
+# Following [Design patterns for larger models](@ref), we code our JuMP model
+# as a function which takes in an input. In this example, we print the output to
+# `stdout`:
+
+function solve_transportation_problem(data::Containers.DenseAxisArray)
+    ## Get the set of supplies and demands
+    O, D = axes(data)
+    ## Drop the SUPPLY and DEMAND nodes from our sets
+    O, D = setdiff(O, ["DEMAND"]), setdiff(D, ["SUPPLY"])
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, x[o in O, d in D] >= 0)
+    ## Remove arcs with "." cost by fixing them to 0.0.
+    for o in O, d in D
+        if data[o, d] == "."
+            fix(x[o, d], 0.0; force = true)
+        end
+    end
+    @objective(
+        model,
+        Min,
+        sum(data[o, d] * x[o, d] for o in O, d in D if data[o, d] != "."),
+    )
+    @constraint(model, [o in O], sum(x[o, :]) <= data[o, "SUPPLY"])
+    @constraint(model, [d in D], sum(x[:, d]) == data["DEMAND", d])
+    optimize!(model)
+    ## Pretty print the solution in the format of the input
+    print("    ", join(lpad.(D, 7, ' ')))
+    for o in O
+        print("\n", o)
+        for d in D
+            if isapprox(value(x[o, d]), 0.0; atol = 1e-6)
+                print("      .")
+            else
+                print(" ", lpad(value(x[o, d]), 6, ' '))
+            end
+        end
+    end
+    return
+end
+
+# ## Solution
+
+# Let's solve and view the solution:
+
+solve_transportation_problem(data)
diff --git a/previews/PR3525/tutorials/linear/transp.txt b/previews/PR3525/tutorials/linear/transp.txt
new file mode 100644
index 00000000000..0b03b4258ac
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/transp.txt
@@ -0,0 +1,5 @@
+     . FRA  DET LAN WIN  STL  FRE  LAF SUPPLY
+  GARY  39   14  11  14   16   82    8   1400
+  CLEV  27    .  12   .   26   95   17   2600
+  PITT  24   14  17  13   28   99   20   2900
+DEMAND 900 1200 600 400 1700 1100 1000      0
diff --git a/previews/PR3525/tutorials/linear/transp/index.html b/previews/PR3525/tutorials/linear/transp/index.html
new file mode 100644
index 00000000000..dbacd955678
--- /dev/null
+++ b/previews/PR3525/tutorials/linear/transp/index.html
@@ -0,0 +1,76 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>The transportation problem · JuMP</title><meta name="title" content="The transportation problem · JuMP"/><meta property="og:title" content="The transportation problem · JuMP"/><meta property="twitter:title" content="The transportation problem · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../diet/">The diet problem</a></li><li><a class="tocitem" href="../cannery/">The cannery problem</a></li><li><a class="tocitem" href="../factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../network_flows/">Network flow problems</a></li><li class="is-active"><a class="tocitem" href>The transportation problem</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Solution"><span>Solution</span></a></li></ul></li><li><a class="tocitem" href="../multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../sudoku/">Sudoku</a></li><li><a class="tocitem" href="../n-queens/">N-Queens</a></li><li><a class="tocitem" href="../constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../callbacks/">Callbacks</a></li><li><a class="tocitem" href="../lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox"/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../nonlinear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../nonlinear/simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../../nonlinear/portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../../nonlinear/rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../../nonlinear/space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../../nonlinear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../nonlinear/user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../../nonlinear/nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../../nonlinear/querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../../nonlinear/complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Linear programs</a></li><li class="is-active"><a href>The transportation problem</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>The transportation problem</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/linear/transp.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="The-transportation-problem"><a class="docs-heading-anchor" href="#The-transportation-problem">The transportation problem</a><a id="The-transportation-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-transportation-problem" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../transp.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Louis Luangkesorn.</strong></p><p>This tutorial is an adaptation of the transportation problem described in <a href="https://ampl.com/resources/the-ampl-book/"><em>AMPL: A Modeling Language for Mathematical Programming</em></a>, by R. Fourer, D.M. Gay and B.W. Kernighan.</p><p>The purpose of this tutorial is to demonstrate how to create a JuMP model from an ad-hoc structured text file.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import DelimitedFiles
+import HiGHS</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>Suppose that we have a set of factories that produce <a href="https://en.wikipedia.org/wiki/Pogo_stick">pogo sticks</a>, and a set of retail stores in which to sell them. Each factory has a maximum number of pogo sticks that it can produce, and each retail store has a demand of pogo sticks that it can sell.</p><p>In the transportation problem, we want to choose the number of pogo sticks to make and ship from each factory to each retail store that minimizes the total shipping cost.</p><p>Mathematically, we represent our set of factories by a set of origins <span>$i \in O$</span> and our retail stores by a set of destinations <span>$j \in D$</span>. The maximum supply at each factory is <span>$s_i$</span> and the demand from each retail store is <span>$d_j$</span>. The cost of shipping one pogo stick from <span>$i$</span> to <span>$j$</span> is <span>$c_{i,j}$</span>.</p><p>With a little effort, we can model the transportation problem as the following linear program:</p><p class="math-container">\[\begin{aligned}
+\min  &amp;&amp; \sum_{i \in O, j \in D} c_{i,j} x_{i,j} \\
+s.t.  &amp;&amp; \sum_{j \in D} x_{i, j} \le s_i &amp;&amp; \forall i \in O \\
+      &amp;&amp; \sum_{i \in O} x_{i, j} = d_j &amp;&amp; \forall j \in D \\
+      &amp;&amp; x_{i, j} \ge 0 &amp;&amp; \forall i \in O, j \in D
+\end{aligned}\]</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>We assume our data is in the form of a text file that has the following form. In practice, we would obtain this text file from the user as input, but for the purpose of this tutorial we&#39;re going to create it from Julia.</p><pre><code class="language-julia hljs">open(joinpath(@__DIR__, &quot;transp.txt&quot;), &quot;w&quot;) do io
+    print(
+        io,
+        &quot;&quot;&quot;
+             . FRA  DET LAN WIN  STL  FRE  LAF SUPPLY
+          GARY  39   14  11  14   16   82    8   1400
+          CLEV  27    .  12   .   26   95   17   2600
+          PITT  24   14  17  13   28   99   20   2900
+        DEMAND 900 1200 600 400 1700 1100 1000      0
+        &quot;&quot;&quot;,
+    )
+    return
+end</code></pre><p>Here the rows are the origins, the columns are the destinations, and the values are the cost of shipping one pogo stick from the origin to the destination. If pogo stick cannot be transported from a source to a destination, then the value is <code>.</code>. The final row and final column are the demand and supply of each location respectively.</p><p>We didn&#39;t account for arcs which do not exist in our formulation, but we can make a small change and fix <span>$x_{i,j} = 0$</span> if <span>$c_{i,j} = &quot;.&quot;$</span>.</p><p>Our first step is to convert this text format into an appropriate Julia datastructure that we can work with. Since our data is tabular with named rows and columns, one option is JuMP&#39;s <a href="../../../api/JuMP.Containers/#JuMP.Containers.DenseAxisArray"><code>Containers.DenseAxisArray</code></a> object:</p><pre><code class="language-julia hljs">function read_data(filename::String)
+    data = DelimitedFiles.readdlm(filename)
+    rows, columns = data[2:end, 1], data[1, 2:end]
+    return Containers.DenseAxisArray(data[2:end, 2:end], rows, columns)
+end
+
+data = read_data(joinpath(@__DIR__, &quot;transp.txt&quot;))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-dimensional DenseAxisArray{Any,2,...} with index sets:
+    Dimension 1, Any[&quot;GARY&quot;, &quot;CLEV&quot;, &quot;PITT&quot;, &quot;DEMAND&quot;]
+    Dimension 2, Any[&quot;FRA&quot;, &quot;DET&quot;, &quot;LAN&quot;, &quot;WIN&quot;, &quot;STL&quot;, &quot;FRE&quot;, &quot;LAF&quot;, &quot;SUPPLY&quot;]
+And data, a 4×8 Matrix{Any}:
+  39    14      11   14       16    82     8  1400
+  27      &quot;.&quot;   12     &quot;.&quot;    26    95    17  2600
+  24    14      17   13       28    99    20  2900
+ 900  1200     600  400     1700  1100  1000     0</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>Following <a href="../../getting_started/design_patterns_for_larger_models/#Design-patterns-for-larger-models">Design patterns for larger models</a>, we code our JuMP model as a function which takes in an input. In this example, we print the output to <code>stdout</code>:</p><pre><code class="language-julia hljs">function solve_transportation_problem(data::Containers.DenseAxisArray)
+    # Get the set of supplies and demands
+    O, D = axes(data)
+    # Drop the SUPPLY and DEMAND nodes from our sets
+    O, D = setdiff(O, [&quot;DEMAND&quot;]), setdiff(D, [&quot;SUPPLY&quot;])
+    model = Model(HiGHS.Optimizer)
+    set_silent(model)
+    @variable(model, x[o in O, d in D] &gt;= 0)
+    # Remove arcs with &quot;.&quot; cost by fixing them to 0.0.
+    for o in O, d in D
+        if data[o, d] == &quot;.&quot;
+            fix(x[o, d], 0.0; force = true)
+        end
+    end
+    @objective(
+        model,
+        Min,
+        sum(data[o, d] * x[o, d] for o in O, d in D if data[o, d] != &quot;.&quot;),
+    )
+    @constraint(model, [o in O], sum(x[o, :]) &lt;= data[o, &quot;SUPPLY&quot;])
+    @constraint(model, [d in D], sum(x[:, d]) == data[&quot;DEMAND&quot;, d])
+    optimize!(model)
+    # Pretty print the solution in the format of the input
+    print(&quot;    &quot;, join(lpad.(D, 7, &#39; &#39;)))
+    for o in O
+        print(&quot;\n&quot;, o)
+        for d in D
+            if isapprox(value(x[o, d]), 0.0; atol = 1e-6)
+                print(&quot;      .&quot;)
+            else
+                print(&quot; &quot;, lpad(value(x[o, d]), 6, &#39; &#39;))
+            end
+        end
+    end
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_transportation_problem (generic function with 1 method)</code></pre><h2 id="Solution"><a class="docs-heading-anchor" href="#Solution">Solution</a><a id="Solution-1"></a><a class="docs-heading-anchor-permalink" href="#Solution" title="Permalink"></a></h2><p>Let&#39;s solve and view the solution:</p><pre><code class="language-julia hljs">solve_transportation_problem(data)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">        FRA    DET    LAN    WIN    STL    FRE    LAF
+GARY      .      .      .      .  300.0 1100.0      .
+CLEV      .      .  600.0      . 1000.0      . 1000.0
+PITT  900.0 1200.0      .  400.0  400.0      .      .</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../network_flows/">« Network flow problems</a><a class="docs-footer-nextpage" href="../multi_objective_knapsack/">Multi-objective knapsack »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/complementarity.jl b/previews/PR3525/tutorials/nonlinear/complementarity.jl
new file mode 100644
index 00000000000..a301aab933f
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/complementarity.jl
@@ -0,0 +1,208 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Mixed complementarity problems
+
+# This tutorial is a collection of mixed complementarity programs.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import PATHSolver
+import Test  #src
+
+# ## Background
+
+# A mixed complementarity problem has the form:
+# ```math
+# \begin{align}
+#     F_i(x) \perp x_i & i = 1 \ldots n \\
+#     l_i \le x_i \le u_i & i = 1 \ldots n.
+# \end{align}
+# ```
+# where the ``\perp`` constraint enforces the following relations:
+#
+#  - If ``l_i < x_i < u_i``, then ``F_i(x) = 0``
+#  - If ``l_i = x_i``, then ``F_i(x) \ge 0``
+#  - If ``x_i = u_i``, then ``F_i(x) \le 0``
+
+# You may have seen a complementarity problem written as
+# ``0 \le F(x) \perp x \ge 0``. This is a special case of a mixed
+# complementarity problem in which ``l_i = 0`` and ``u_i = \infty``.
+
+# Importantly, a mixed complementarity problem does not have an objective, and
+# no other constraint types are present.
+
+# ## Linear complementarity
+
+# Form a mixed complementarity problem using the perp symbol `⟂` (type
+# `\perp<tab>` in the REPL).
+
+M = [0 0 -1 -1; 0 0 1 -2; 1 -1 2 -2; 1 2 -2 4]
+q = [2, 2, -2, -6]
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, 0 <= x[1:4] <= 10, start = 0)
+@constraint(model, M * x + q ⟂ x)
+optimize!(model)
+Test.@test value.(x) ≈ [2.8, 0.0, 0.8, 1.2]  #src
+value.(x)
+
+# ## Other ways of writing linear complementarity problems
+
+# You do not need to use a single vector of variables, and the complementarity
+# constraints can be given in any order. In addition, you can use the perp
+# symbol, the `complements(F, x)` syntax, or the [`MOI.Complements`](@ref) set.
+
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, 0 <= w <= 10, start = 0)
+@variable(model, 0 <= x <= 10, start = 0)
+@variable(model, 0 <= y <= 10, start = 0)
+@variable(model, 0 <= z <= 10, start = 0)
+@constraint(model, complements(y - 2z + 2, x))
+@constraint(model, [-y - z + 2, w] in MOI.Complements(2))
+@constraint(model, w + 2x - 2y + 4z - 6 ⟂ z)
+@constraint(model, w - x + 2y - 2z - 2 ⟂ y)
+optimize!(model)
+Test.@test value.([w, x, y, z]) ≈ [2.8, 0.0, 0.8, 1.2]  #src
+value.([w, x, y, z])
+
+# ## Transportation
+
+# This example is a reformulation of the transportation problem from Chapter
+# 3.3 of Dantzig, G.B. (1963). _Linear Programming and Extensions_. Princeton
+# University Press, Princeton, New Jersey. It is based on the GAMS model
+# [`gamslib_transmcp`](https://www.gams.com/latest/gamslib_ml/libhtml/gamslib_transmcp.html).
+
+capacity = Dict("seattle" => 350, "san-diego" => 600)
+demand = Dict("new-york" => 325, "chicago" => 300, "topeka" => 275)
+cost = Dict(
+    ("seattle" => "new-york") => 90 * 2.5 / 1_000,
+    ("seattle" => "chicago") => 90 * 1.7 / 1_000,
+    ("seattle" => "topeka") => 90 * 1.8 / 1_000,
+    ("san-diego" => "new-york") => 90 * 2.5 / 1_000,
+    ("san-diego" => "chicago") => 90 * 1.8 / 1_000,
+    ("san-diego" => "topeka") => 90 * 1.4 / 1_000,
+)
+plants, markets = keys(capacity), keys(demand)
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, w[i in plants] >= 0)
+@variable(model, p[j in markets] >= 0)
+@variable(model, x[i in plants, j in markets] >= 0)
+@constraints(
+    model,
+    begin
+        [i in plants, j in markets], w[i] + cost[i=>j] - p[j] ⟂ x[i, j]
+        [i in plants], capacity[i] - sum(x[i, :]) ⟂ w[i]
+        [j in markets], sum(x[:, j]) - demand[j] ⟂ p[j]
+    end
+)
+optimize!(model)
+Test.@test isapprox(value(p["new-york"]), 0.225; atol = 1e-3)  #src
+value.(p)
+
+# ## Expected utility of insurance
+
+# This example is taken from a lecture of the course AAE706, given by Thomas F.
+# Rutherford at the University of Wisconsin, Madison. It models the expected
+# coverage of insurance `K` that a rational actor would obtain to insure a risk
+# that occurs with probability `pi` and results in a loss of `L`.
+
+pi = 0.01  # Probability of a bad outcome
+L = 0.5    # Loss with a bad outcome
+γ = 0.02   # Premium for coverage
+σ = 0.5    # Elasticity
+ρ = -1     # Risk exponent
+U(C) = C^ρ / ρ
+MU(C) = C^(ρ - 1)
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, EU, start = 1)   # Expected utilitiy
+@variable(model, EV, start = 1)   # Equivalent variation in income
+@variable(model, C_G, start = 1)  # Consumption on a good day
+@variable(model, C_B, start = 1)  # Consumption on a bad day
+@variable(model, K, start = 1)    # Coverage
+@constraints(
+    model,
+    begin
+        (1 - pi) * U(C_G) + pi * U(C_B) - EU ⟂ EU
+        100 * (((1 - pi) * C_G^ρ + pi * C_B^ρ)^(1 / ρ) - 1) - EV ⟂ EV
+        1 - γ * K - C_G ⟂ C_G
+        1 - L + (1 - γ) * K - C_B ⟂ C_B
+        γ * ((1 - pi) * MU(C_G) + pi * MU(C_B)) - pi * MU(C_B) ⟂ K
+    end
+)
+optimize!(model)
+Test.@test isapprox(value(C_G), 0.996; atol = 1e-3)  #src
+value(K)
+
+# ## Electricity consumption
+
+# This example is a mixed complementarity formulation of example 3.3.1 from
+# D’Aertrycke, G., Ehrenmann, A., Ralph, D., & Smeers, Y. (2017). [Risk trading
+# in capacity equilibrium models](https://doi.org/10.17863/CAM.17552).
+
+# This example models a risk neutral competitive equilibrium between a producer
+# and a consumer of electricity.
+
+# In our example, we assume a producer is looking to invest in a new power
+# plant with capacity ``x`` [MW]. This plant has an annualized capital cost of
+# ``I`` [€/MW] and an operating cost of ``C`` [€/MWh]. There are 8760 hours in a
+# year.
+#
+# After making the capital investment, there are five possible consumption
+# scenarios, ``\omega``, which occur with probability ``\theta_\omega``. In each
+# scenario , the producer makes ``Y_ω`` MW of electricity.
+#
+# There is one consumer in the model, who has a quadratic utility function,
+# ``U(Q_ω) = A_ω Q_ω + \frac{B_ω Q_ω^2}{2}``.
+#
+# We now build and solve the mixed complementarity problem with a few brief
+# comments. The economic justification for the model would require a larger
+# tutorial than the space available here. Consult the [original text](https://doi.org/10.17863/CAM.17552)
+# for details.
+
+I = 90_000                     # Annualized capital cost
+C = 60                         # Operation cost per MWh
+τ = 8_760                      # Hours per year
+θ = [0.2, 0.2, 0.2, 0.2, 0.2]  # Scenario probabilities
+A = [300, 350, 400, 450, 500]  # Utility function coefficients
+B = 1                          # Utility function coefficients
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, x >= 0, start = 1)           # Installed capacity
+@variable(model, Q[ω = 1:5] >= 0, start = 1)  # Consumption
+@variable(model, Y[ω = 1:5] >= 0, start = 1)  # Production
+@variable(model, P[ω = 1:5], start = 1)       # Electricity price
+@variable(model, μ[ω = 1:5] >= 0, start = 1)  # Capital scarcity margin
+## Unit investment cost equals annualized scarcity profit or investment is 0
+@constraint(model, I - τ * θ' * μ ⟂ x)
+## Difference between price and scarcity margin is equal to operation cost
+@constraint(model, [ω = 1:5], C - (P[ω] - μ[ω]) ⟂ Y[ω])
+## Price is equal to consumer's marginal utility
+@constraint(model, [ω = 1:5], P[ω] - (A[ω] - B * Q[ω]) ⟂ Q[ω])
+## Production is equal to consumption
+@constraint(model, [ω = 1:5], Y[ω] - Q[ω] ⟂ P[ω])
+## Production does not exceed capacity
+@constraint(model, [ω = 1:5], x - Y[ω] ⟂ μ[ω])
+optimize!(model)
+solution_summary(model)
+
+# An equilibrium solution is to build 389 MW:
+
+Test.@test isapprox(value(x), 389; atol = 1)  #src
+value(x)
+
+# The production in each scenario is:
+
+Test.@test isapprox(value.(Q), [240, 290, 340, 389, 389]; atol = 1)  #src
+value.(Q)
+
+# The price in each scenario is:
+
+Test.@test isapprox(value.(P), [60, 60, 60, 61, 111]; atol = 1)  #src
+value.(P)
diff --git a/previews/PR3525/tutorials/nonlinear/complementarity/index.html b/previews/PR3525/tutorials/nonlinear/complementarity/index.html
new file mode 100644
index 00000000000..6232d6e6a57
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/complementarity/index.html
@@ -0,0 +1,140 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Mixed complementarity problems · JuMP</title><meta name="title" content="Mixed complementarity problems · JuMP"/><meta property="og:title" content="Mixed complementarity problems · JuMP"/><meta property="twitter:title" content="Mixed complementarity problems · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li class="is-active"><a class="tocitem" href>Mixed complementarity problems</a><ul class="internal"><li><a class="tocitem" href="#Background"><span>Background</span></a></li><li><a class="tocitem" href="#Linear-complementarity"><span>Linear complementarity</span></a></li><li><a class="tocitem" href="#Other-ways-of-writing-linear-complementarity-problems"><span>Other ways of writing linear complementarity problems</span></a></li><li><a class="tocitem" href="#Transportation"><span>Transportation</span></a></li><li><a class="tocitem" href="#Expected-utility-of-insurance"><span>Expected utility of insurance</span></a></li><li><a class="tocitem" href="#Electricity-consumption"><span>Electricity consumption</span></a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Mixed complementarity problems</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Mixed complementarity problems</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/complementarity.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Mixed-complementarity-problems"><a class="docs-heading-anchor" href="#Mixed-complementarity-problems">Mixed complementarity problems</a><a id="Mixed-complementarity-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Mixed-complementarity-problems" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../complementarity.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial is a collection of mixed complementarity programs.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import PATHSolver</code></pre><h2 id="Background"><a class="docs-heading-anchor" href="#Background">Background</a><a id="Background-1"></a><a class="docs-heading-anchor-permalink" href="#Background" title="Permalink"></a></h2><p>A mixed complementarity problem has the form:</p><p class="math-container">\[\begin{align}
+    F_i(x) \perp x_i &amp; i = 1 \ldots n \\
+    l_i \le x_i \le u_i &amp; i = 1 \ldots n.
+\end{align}\]</p><p>where the <span>$\perp$</span> constraint enforces the following relations:</p><ul><li>If <span>$l_i &lt; x_i &lt; u_i$</span>, then <span>$F_i(x) = 0$</span></li><li>If <span>$l_i = x_i$</span>, then <span>$F_i(x) \ge 0$</span></li><li>If <span>$x_i = u_i$</span>, then <span>$F_i(x) \le 0$</span></li></ul><p>You may have seen a complementarity problem written as <span>$0 \le F(x) \perp x \ge 0$</span>. This is a special case of a mixed complementarity problem in which <span>$l_i = 0$</span> and <span>$u_i = \infty$</span>.</p><p>Importantly, a mixed complementarity problem does not have an objective, and no other constraint types are present.</p><h2 id="Linear-complementarity"><a class="docs-heading-anchor" href="#Linear-complementarity">Linear complementarity</a><a id="Linear-complementarity-1"></a><a class="docs-heading-anchor-permalink" href="#Linear-complementarity" title="Permalink"></a></h2><p>Form a mixed complementarity problem using the perp symbol <code>⟂</code> (type <code>\perp&lt;tab&gt;</code> in the REPL).</p><pre><code class="language-julia hljs">M = [0 0 -1 -1; 0 0 1 -2; 1 -1 2 -2; 1 2 -2 4]
+q = [2, 2, -2, -6]
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, 0 &lt;= x[1:4] &lt;= 10, start = 0)
+@constraint(model, M * x + q ⟂ x)
+optimize!(model)
+value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{Float64}:
+ 2.8
+ 0.0
+ 0.8
+ 1.2</code></pre><h2 id="Other-ways-of-writing-linear-complementarity-problems"><a class="docs-heading-anchor" href="#Other-ways-of-writing-linear-complementarity-problems">Other ways of writing linear complementarity problems</a><a id="Other-ways-of-writing-linear-complementarity-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Other-ways-of-writing-linear-complementarity-problems" title="Permalink"></a></h2><p>You do not need to use a single vector of variables, and the complementarity constraints can be given in any order. In addition, you can use the perp symbol, the <code>complements(F, x)</code> syntax, or the <a href="../../../moi/reference/standard_form/#MathOptInterface.Complements"><code>MOI.Complements</code></a> set.</p><pre><code class="language-julia hljs">model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, 0 &lt;= w &lt;= 10, start = 0)
+@variable(model, 0 &lt;= x &lt;= 10, start = 0)
+@variable(model, 0 &lt;= y &lt;= 10, start = 0)
+@variable(model, 0 &lt;= z &lt;= 10, start = 0)
+@constraint(model, complements(y - 2z + 2, x))
+@constraint(model, [-y - z + 2, w] in MOI.Complements(2))
+@constraint(model, w + 2x - 2y + 4z - 6 ⟂ z)
+@constraint(model, w - x + 2y - 2z - 2 ⟂ y)
+optimize!(model)
+value.([w, x, y, z])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">4-element Vector{Float64}:
+ 2.8
+ 0.0
+ 0.8
+ 1.2</code></pre><h2 id="Transportation"><a class="docs-heading-anchor" href="#Transportation">Transportation</a><a id="Transportation-1"></a><a class="docs-heading-anchor-permalink" href="#Transportation" title="Permalink"></a></h2><p>This example is a reformulation of the transportation problem from Chapter 3.3 of Dantzig, G.B. (1963). <em>Linear Programming and Extensions</em>. Princeton University Press, Princeton, New Jersey. It is based on the GAMS model <a href="https://www.gams.com/latest/gamslib_ml/libhtml/gamslib_transmcp.html"><code>gamslib_transmcp</code></a>.</p><pre><code class="language-julia hljs">capacity = Dict(&quot;seattle&quot; =&gt; 350, &quot;san-diego&quot; =&gt; 600)
+demand = Dict(&quot;new-york&quot; =&gt; 325, &quot;chicago&quot; =&gt; 300, &quot;topeka&quot; =&gt; 275)
+cost = Dict(
+    (&quot;seattle&quot; =&gt; &quot;new-york&quot;) =&gt; 90 * 2.5 / 1_000,
+    (&quot;seattle&quot; =&gt; &quot;chicago&quot;) =&gt; 90 * 1.7 / 1_000,
+    (&quot;seattle&quot; =&gt; &quot;topeka&quot;) =&gt; 90 * 1.8 / 1_000,
+    (&quot;san-diego&quot; =&gt; &quot;new-york&quot;) =&gt; 90 * 2.5 / 1_000,
+    (&quot;san-diego&quot; =&gt; &quot;chicago&quot;) =&gt; 90 * 1.8 / 1_000,
+    (&quot;san-diego&quot; =&gt; &quot;topeka&quot;) =&gt; 90 * 1.4 / 1_000,
+)
+plants, markets = keys(capacity), keys(demand)
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, w[i in plants] &gt;= 0)
+@variable(model, p[j in markets] &gt;= 0)
+@variable(model, x[i in plants, j in markets] &gt;= 0)
+@constraints(
+    model,
+    begin
+        [i in plants, j in markets], w[i] + cost[i=&gt;j] - p[j] ⟂ x[i, j]
+        [i in plants], capacity[i] - sum(x[i, :]) ⟂ w[i]
+        [j in markets], sum(x[:, j]) - demand[j] ⟂ p[j]
+    end
+)
+optimize!(model)
+value.(p)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">1-dimensional DenseAxisArray{Float64,1,...} with index sets:
+    Dimension 1, [&quot;new-york&quot;, &quot;chicago&quot;, &quot;topeka&quot;]
+And data, a 3-element Vector{Float64}:
+ 0.22500000000033224
+ 0.15299999994933886
+ 0.126</code></pre><h2 id="Expected-utility-of-insurance"><a class="docs-heading-anchor" href="#Expected-utility-of-insurance">Expected utility of insurance</a><a id="Expected-utility-of-insurance-1"></a><a class="docs-heading-anchor-permalink" href="#Expected-utility-of-insurance" title="Permalink"></a></h2><p>This example is taken from a lecture of the course AAE706, given by Thomas F. Rutherford at the University of Wisconsin, Madison. It models the expected coverage of insurance <code>K</code> that a rational actor would obtain to insure a risk that occurs with probability <code>pi</code> and results in a loss of <code>L</code>.</p><pre><code class="language-julia hljs">pi = 0.01  # Probability of a bad outcome
+L = 0.5    # Loss with a bad outcome
+γ = 0.02   # Premium for coverage
+σ = 0.5    # Elasticity
+ρ = -1     # Risk exponent
+U(C) = C^ρ / ρ
+MU(C) = C^(ρ - 1)
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, EU, start = 1)   # Expected utilitiy
+@variable(model, EV, start = 1)   # Equivalent variation in income
+@variable(model, C_G, start = 1)  # Consumption on a good day
+@variable(model, C_B, start = 1)  # Consumption on a bad day
+@variable(model, K, start = 1)    # Coverage
+@constraints(
+    model,
+    begin
+        (1 - pi) * U(C_G) + pi * U(C_B) - EU ⟂ EU
+        100 * (((1 - pi) * C_G^ρ + pi * C_B^ρ)^(1 / ρ) - 1) - EV ⟂ EV
+        1 - γ * K - C_G ⟂ C_G
+        1 - L + (1 - γ) * K - C_B ⟂ C_B
+        γ * ((1 - pi) * MU(C_G) + pi * MU(C_B)) - pi * MU(C_B) ⟂ K
+    end
+)
+optimize!(model)
+value(K)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">0.20474003534537774</code></pre><h2 id="Electricity-consumption"><a class="docs-heading-anchor" href="#Electricity-consumption">Electricity consumption</a><a id="Electricity-consumption-1"></a><a class="docs-heading-anchor-permalink" href="#Electricity-consumption" title="Permalink"></a></h2><p>This example is a mixed complementarity formulation of example 3.3.1 from D’Aertrycke, G., Ehrenmann, A., Ralph, D., &amp; Smeers, Y. (2017). <a href="https://doi.org/10.17863/CAM.17552">Risk trading in capacity equilibrium models</a>.</p><p>This example models a risk neutral competitive equilibrium between a producer and a consumer of electricity.</p><p>In our example, we assume a producer is looking to invest in a new power plant with capacity <span>$x$</span> [MW]. This plant has an annualized capital cost of <span>$I$</span> [€/MW] and an operating cost of <span>$C$</span> [€/MWh]. There are 8760 hours in a year.</p><p>After making the capital investment, there are five possible consumption scenarios, <span>$\omega$</span>, which occur with probability <span>$\theta_\omega$</span>. In each scenario , the producer makes <span>$Y_ω$</span> MW of electricity.</p><p>There is one consumer in the model, who has a quadratic utility function, <span>$U(Q_ω) = A_ω Q_ω + \frac{B_ω Q_ω^2}{2}$</span>.</p><p>We now build and solve the mixed complementarity problem with a few brief comments. The economic justification for the model would require a larger tutorial than the space available here. Consult the <a href="https://doi.org/10.17863/CAM.17552">original text</a> for details.</p><pre><code class="language-julia hljs">I = 90_000                     # Annualized capital cost
+C = 60                         # Operation cost per MWh
+τ = 8_760                      # Hours per year
+θ = [0.2, 0.2, 0.2, 0.2, 0.2]  # Scenario probabilities
+A = [300, 350, 400, 450, 500]  # Utility function coefficients
+B = 1                          # Utility function coefficients
+model = Model(PATHSolver.Optimizer)
+set_silent(model)
+@variable(model, x &gt;= 0, start = 1)           # Installed capacity
+@variable(model, Q[ω = 1:5] &gt;= 0, start = 1)  # Consumption
+@variable(model, Y[ω = 1:5] &gt;= 0, start = 1)  # Production
+@variable(model, P[ω = 1:5], start = 1)       # Electricity price
+@variable(model, μ[ω = 1:5] &gt;= 0, start = 1)  # Capital scarcity margin
+# Unit investment cost equals annualized scarcity profit or investment is 0
+@constraint(model, I - τ * θ&#39; * μ ⟂ x)
+# Difference between price and scarcity margin is equal to operation cost
+@constraint(model, [ω = 1:5], C - (P[ω] - μ[ω]) ⟂ Y[ω])
+# Price is equal to consumer&#39;s marginal utility
+@constraint(model, [ω = 1:5], P[ω] - (A[ω] - B * Q[ω]) ⟂ Q[ω])
+# Production is equal to consumption
+@constraint(model, [ω = 1:5], Y[ω] - Q[ω] ⟂ P[ω])
+# Production does not exceed capacity
+@constraint(model, [ω = 1:5], x - Y[ω] ⟂ μ[ω])
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Path 5.0.03
+
+* Status
+  Result count       : 1
+  Termination status : LOCALLY_SOLVED
+  Message from the solver:
+  &quot;The problem was solved&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 2.07000e-04
+</code></pre><p>An equilibrium solution is to build 389 MW:</p><pre><code class="language-julia hljs">value(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">389.31506849315065</code></pre><p>The production in each scenario is:</p><pre><code class="language-julia hljs">value.(Q)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5-element Vector{Float64}:
+ 240.00000000000014
+ 289.99999999999994
+ 340.0
+ 389.31506849315065
+ 389.31506849315065</code></pre><p>The price in each scenario is:</p><pre><code class="language-julia hljs">value.(P)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">5-element Vector{Float64}:
+  60.00000000000001
+  60.0
+  59.99999999999999
+  60.68493150684928
+ 110.68493150684935</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../querying_hessians/">« Computing Hessians</a><a class="docs-footer-nextpage" href="../../conic/introduction/">Introduction »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/introduction/index.html b/previews/PR3525/tutorials/nonlinear/introduction/index.html
new file mode 100644
index 00000000000..5afdb1ed3c3
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/introduction/index.html
@@ -0,0 +1,10 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Introduction · JuMP</title><meta name="title" content="Introduction · JuMP"/><meta property="og:title" content="Introduction · JuMP"/><meta property="twitter:title" content="Introduction · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li class="is-active"><a class="tocitem" href>Introduction</a><ul class="internal"><li><a class="tocitem" href="#How-to-choose-a-solver"><span>How to choose a solver</span></a></li><li><a class="tocitem" href="#How-these-tutorials-are-structured"><span>How these tutorials are structured</span></a></li></ul></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Introduction</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Introduction</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/introduction.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Introduction"><a class="docs-heading-anchor" href="#Introduction">Introduction</a><a id="Introduction-1"></a><a class="docs-heading-anchor-permalink" href="#Introduction" title="Permalink"></a></h1><p><a href="https://en.wikipedia.org/wiki/Nonlinear_programming">Nonlinear programs (NLPs)</a> are a class of optimization problems in which some of the constraints or the objective function are nonlinear:</p><p class="math-container">\[\begin{align}
+    \min_{x \in \mathbb{R}^n} &amp; f_0(x) \\
+    \;\;\text{s.t.} &amp; l_j \le f_j(x) \le u_j &amp; j = 1 \ldots m \\
+    &amp; l_i \le x_i \le u_i &amp; i = 1 \ldots n.
+\end{align}\]</p><p>Mixed-integer nonlinear linear programs (MINLPs) are extensions of nonlinear programs in which some (or all) of the decision variables take discrete values.</p><h2 id="How-to-choose-a-solver"><a class="docs-heading-anchor" href="#How-to-choose-a-solver">How to choose a solver</a><a id="How-to-choose-a-solver-1"></a><a class="docs-heading-anchor-permalink" href="#How-to-choose-a-solver" title="Permalink"></a></h2><p>JuMP supports a range of nonlinear solvers; look for &quot;NLP&quot; in the list of <a href="../../../installation/#Supported-solvers">Supported solvers</a>. However, very few solvers support mixed-integer nonlinear linear programs. Solvers supporting discrete variables start with &quot;(MI)&quot; in the list of <a href="../../../installation/#Supported-solvers">Supported solvers</a>.</p><p>If the only nonlinearities in your model are quadratic terms (that is, multiplication between two decision variables), you can also use second-order cone solvers, which are indicated by &quot;SOCP.&quot; In most cases, these solvers are restricted to convex quadratic problems and will error if you pass a nonconvex quadratic function; however, Gurobi has the ability to solve nonconvex quadratic terms.</p><h2 id="How-these-tutorials-are-structured"><a class="docs-heading-anchor" href="#How-these-tutorials-are-structured">How these tutorials are structured</a><a id="How-these-tutorials-are-structured-1"></a><a class="docs-heading-anchor-permalink" href="#How-these-tutorials-are-structured" title="Permalink"></a></h2><p>Having a high-level overview of how this part of the documentation is structured will help you know where to look for certain things.</p><ul><li>The following tutorials are worked examples that present a problem in words, then formulate it in mathematics, and then solve it in JuMP. This usually involves some sort of visualization of the solution. Start here if you are new to JuMP.<ul><li><a href="../rocket_control/#Rocket-Control">Rocket Control</a></li><li><a href="../space_shuttle_reentry_trajectory/#Optimal-control-for-a-Space-Shuttle-reentry-trajectory">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a href="../portfolio/#Portfolio-optimization">Portfolio optimization</a></li></ul></li><li>The <a href="../tips_and_tricks/#nonlinear_tips_and_tricks">Tips and tricks</a> tutorial contains a number of helpful reformulations and tricks you can use when modeling nonlinear programs. Look here if you are stuck trying to formulate a problem as a nonlinear program.</li><li>The <a href="../querying_hessians/#Computing-Hessians">Computing Hessians</a> is an advanced tutorial which explains how to compute the Hessian of the Lagrangian of a nonlinear program. This is useful only in particular cases.</li><li>The remaining tutorials are less verbose and styled in the form of short code examples. These tutorials have less explanation, but may contain useful code snippets, particularly if they are similar to a problem you are trying to solve.</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../../linear/mip_duality/">« Computing the duals of a mixed-integer program</a><a class="docs-footer-nextpage" href="../simple_examples/">Simple examples »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/nested_problems.jl b/previews/PR3525/tutorials/nonlinear/nested_problems.jl
new file mode 100644
index 00000000000..54865d05c13
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/nested_problems.jl
@@ -0,0 +1,235 @@
+# Copyright (c) 2022 Oscar Dowson and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Nested optimization problems
+
+# In this tutorial we explain how to write a *nested* optimization problem,
+# where an *upper* problem uses the results from the optimization of a *lower*
+# subproblem.
+#
+# To model the problem, we define a user-defined operator to handle the
+# decomposition of the lower problem inside the upper one. Finally, we show how
+# to improve the performance by using a cache that avoids resolving the lower
+# problem.
+#
+# For a simpler example of writing a user-defined operator,
+# see the [User-defined Hessians](@ref) tutorial.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import Ipopt
+
+# In the rest of this tutorial, our goal is to solve the bilevel
+# optimization problem:
+
+# ```math
+# \begin{array}{r l}
+# \min\limits_{x,z} & x_1^2 + x_2^2 + z \\
+# s.t.            & \begin{array}{r l}
+#                       z = \max\limits_{y} & x_1^2 y_1 + x_2^2 y_2  - x_1 y_1^4 - 2 x_2 y_2^4 \\
+#                       s.t.                  & (y_1 - 10)^2 + (y_2 - 10)^2 \le 25
+#                   \end{array} \\
+#                 & x \ge 0.
+# \end{array}
+# ```
+
+# This bilevel optimization problem is composed of two nested optimization
+# problems. An _upper_ level, involving variables ``x``, and a _lower_ level,
+# involving variables ``y``. From the perspective of the lower-level problem,
+# the values of ``x`` are fixed parameters, and so the model optimizes ``y``
+# given those fixed parameters. Simultaneously, the upper-level problem
+# optimizes ``x`` and ``z`` given the response of ``y``.
+
+# ## Decomposition
+
+# There are a few ways to solve this problem, but we are going to use a
+# nonlinear decomposition method. The first step is to write a function to
+# compute the lower-level problem:
+
+# ```math
+# \begin{array}{r l}
+#   V(x_1, x_2) = \max\limits_{y} & x_1^2 y_1 + x_2^2 y_2  - x_1 y_1^4 - 2 x_2 y_2^4 \\
+#                            s.t. & (y_1 - 10)^2 + (y_2 - 10)^2 \le 25
+# \end{array}
+# ```
+
+function solve_lower_level(x...)
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, y[1:2])
+    @objective(
+        model,
+        Max,
+        x[1]^2 * y[1] + x[2]^2 * y[2] - x[1] * y[1]^4 - 2 * x[2] * y[2]^4,
+    )
+    @constraint(model, (y[1] - 10)^2 + (y[2] - 10)^2 <= 25)
+    optimize!(model)
+    @assert termination_status(model) == LOCALLY_SOLVED
+    return objective_value(model), value.(y)
+end
+
+# The next function takes a value of ``x`` and returns the optimal lower-level
+# objective-value and the optimal response ``y``. The reason why we need both
+# the objective and the optimal ``y`` will be made clear shortly, but for now
+# let us define:
+
+function V(x...)
+    f, _ = solve_lower_level(x...)
+    return f
+end
+
+# Then, we can substitute ``V`` into our full problem to create:
+
+# ```math
+# \begin{array}{r l}
+# \min\limits_{x} & x_1^2 + x_2^2 + V(x_1, x_2) \\
+# s.t.            & x \ge 0.
+# \end{array}
+# ```
+
+# This looks like a nonlinear optimization problem with a user-defined operator
+# ``V``! However, because ``V`` solves an optimization problem internally, we
+# can't use automatic differentiation to compute the first and second
+# derivatives. Instead, we can use JuMP's ability to pass callback functions
+# for the gradient and Hessian instead.
+
+# First up, we need to define the gradient of ``V`` with respect to ``x``. In
+# general, this may be difficult to compute, but because ``x`` appears only in
+# the objective, we can just differentiate the objective function with respect
+# to ``x``, giving:
+
+function ∇V(g::AbstractVector, x...)
+    _, y = solve_lower_level(x...)
+    g[1] = 2 * x[1] * y[1] - y[1]^4
+    g[2] = 2 * x[2] * y[2] - 2 * y[2]^4
+    return
+end
+
+# Second, we need to define the Hessian of ``V`` with respect to ``x``. This is
+# a symmetric matrix, but in our example only the diagonal elements are
+# non-zero:
+
+function ∇²V(H::AbstractMatrix, x...)
+    _, y = solve_lower_level(x...)
+    H[1, 1] = 2 * y[1]
+    H[2, 2] = 2 * y[2]
+    return
+end
+
+# !!! info
+#     Providing an explicit Hessian function is optional
+#     if first derivatives are already available.
+
+# We now have enough to define our bilevel optimization problem:
+
+model = Model(Ipopt.Optimizer)
+@variable(model, x[1:2] >= 0)
+@operator(model, op_V, 2, V, ∇V, ∇²V)
+@objective(model, Min, x[1]^2 + x[2]^2 + op_V(x[1], x[2]))
+optimize!(model)
+solution_summary(model)
+
+# The optimal objective value is:
+
+objective_value(model)
+
+# and the optimal upper-level decision variables ``x`` are:
+
+value.(x)
+
+# To compute the optimal lower-level decision variables, we need to call
+# `solve_lower_level` with the optimal upper-level decision variables:
+
+_, y = solve_lower_level(value.(x)...)
+y
+
+# ## Improving performance
+
+# Our solution approach works, but it has a performance problem: every time
+# we need to compute the value, gradient, or Hessian of ``V``, we have to
+# re-solve the lower-level optimization problem. This is wasteful, because we
+# will often call the gradient and Hessian at the same point, and so solving the
+# problem twice with the same input repeats work unnecessarily.
+
+# We can work around this by using a cache:
+
+mutable struct Cache
+    x::Any
+    f::Float64
+    y::Vector{Float64}
+end
+
+# with a function to update the cache if needed:
+
+function _update_if_needed(cache::Cache, x...)
+    if cache.x !== x
+        cache.f, cache.y = solve_lower_level(x...)
+        cache.x = x
+    end
+    return
+end
+
+# Then, we define cached versions of out three functions which call
+# `_updated_if_needed` and return values from the cache.
+
+function cached_f(cache::Cache, x...)
+    _update_if_needed(cache, x...)
+    return cache.f
+end
+
+function cached_∇f(cache::Cache, g::AbstractVector, x...)
+    _update_if_needed(cache, x...)
+    g[1] = 2 * x[1] * cache.y[1] - cache.y[1]^4
+    g[2] = 2 * x[2] * cache.y[2] - 2 * cache.y[2]^4
+    return
+end
+
+function cached_∇²f(cache::Cache, H::AbstractMatrix, x...)
+    _update_if_needed(cache, x...)
+    H[1, 1] = 2 * cache.y[1]
+    H[2, 2] = 2 * cache.y[2]
+    return
+end
+
+# Now we're ready to setup and solve the upper level optimization problem:
+
+model = Model(Ipopt.Optimizer)
+@variable(model, x[1:2] >= 0)
+cache = Cache(Float64[], NaN, Float64[])
+@operator(
+    model,
+    op_cached_f,
+    2,
+    (x...) -> cached_f(cache, x...),
+    (g, x...) -> cached_∇f(cache, g, x...),
+    (H, x...) -> cached_∇²f(cache, H, x...),
+)
+@objective(model, Min, x[1]^2 + x[2]^2 + op_cached_f(x[1], x[2]))
+optimize!(model)
+solution_summary(model)
+
+# an we can check we get the same objective value:
+
+objective_value(model)
+
+# and upper-level decision variable ``x``:
+
+value.(x)
diff --git a/previews/PR3525/tutorials/nonlinear/nested_problems/index.html b/previews/PR3525/tutorials/nonlinear/nested_problems/index.html
new file mode 100644
index 00000000000..cc60b5febe3
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/nested_problems/index.html
@@ -0,0 +1,130 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Nested optimization problems · JuMP</title><meta name="title" content="Nested optimization problems · JuMP"/><meta property="og:title" content="Nested optimization problems · JuMP"/><meta property="twitter:title" content="Nested optimization problems · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li class="is-active"><a class="tocitem" href>Nested optimization problems</a><ul class="internal"><li><a class="tocitem" href="#Decomposition"><span>Decomposition</span></a></li><li><a class="tocitem" href="#Improving-performance"><span>Improving performance</span></a></li></ul></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Nested optimization problems</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Nested optimization problems</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/nested_problems.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Nested-optimization-problems"><a class="docs-heading-anchor" href="#Nested-optimization-problems">Nested optimization problems</a><a id="Nested-optimization-problems-1"></a><a class="docs-heading-anchor-permalink" href="#Nested-optimization-problems" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../nested_problems.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>In this tutorial we explain how to write a <em>nested</em> optimization problem, where an <em>upper</em> problem uses the results from the optimization of a <em>lower</em> subproblem.</p><p>To model the problem, we define a user-defined operator to handle the decomposition of the lower problem inside the upper one. Finally, we show how to improve the performance by using a cache that avoids resolving the lower problem.</p><p>For a simpler example of writing a user-defined operator, see the <a href="../user_defined_hessians/#User-defined-Hessians">User-defined Hessians</a> tutorial.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import Ipopt</code></pre><p>In the rest of this tutorial, our goal is to solve the bilevel optimization problem:</p><p class="math-container">\[\begin{array}{r l}
+\min\limits_{x,z} &amp; x_1^2 + x_2^2 + z \\
+s.t.            &amp; \begin{array}{r l}
+                      z = \max\limits_{y} &amp; x_1^2 y_1 + x_2^2 y_2  - x_1 y_1^4 - 2 x_2 y_2^4 \\
+                      s.t.                  &amp; (y_1 - 10)^2 + (y_2 - 10)^2 \le 25
+                  \end{array} \\
+                &amp; x \ge 0.
+\end{array}\]</p><p>This bilevel optimization problem is composed of two nested optimization problems. An <em>upper</em> level, involving variables <span>$x$</span>, and a <em>lower</em> level, involving variables <span>$y$</span>. From the perspective of the lower-level problem, the values of <span>$x$</span> are fixed parameters, and so the model optimizes <span>$y$</span> given those fixed parameters. Simultaneously, the upper-level problem optimizes <span>$x$</span> and <span>$z$</span> given the response of <span>$y$</span>.</p><h2 id="Decomposition"><a class="docs-heading-anchor" href="#Decomposition">Decomposition</a><a id="Decomposition-1"></a><a class="docs-heading-anchor-permalink" href="#Decomposition" title="Permalink"></a></h2><p>There are a few ways to solve this problem, but we are going to use a nonlinear decomposition method. The first step is to write a function to compute the lower-level problem:</p><p class="math-container">\[\begin{array}{r l}
+  V(x_1, x_2) = \max\limits_{y} &amp; x_1^2 y_1 + x_2^2 y_2  - x_1 y_1^4 - 2 x_2 y_2^4 \\
+                           s.t. &amp; (y_1 - 10)^2 + (y_2 - 10)^2 \le 25
+\end{array}\]</p><pre><code class="language-julia hljs">function solve_lower_level(x...)
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, y[1:2])
+    @objective(
+        model,
+        Max,
+        x[1]^2 * y[1] + x[2]^2 * y[2] - x[1] * y[1]^4 - 2 * x[2] * y[2]^4,
+    )
+    @constraint(model, (y[1] - 10)^2 + (y[2] - 10)^2 &lt;= 25)
+    optimize!(model)
+    @assert termination_status(model) == LOCALLY_SOLVED
+    return objective_value(model), value.(y)
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">solve_lower_level (generic function with 1 method)</code></pre><p>The next function takes a value of <span>$x$</span> and returns the optimal lower-level objective-value and the optimal response <span>$y$</span>. The reason why we need both the objective and the optimal <span>$y$</span> will be made clear shortly, but for now let us define:</p><pre><code class="language-julia hljs">function V(x...)
+    f, _ = solve_lower_level(x...)
+    return f
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">V (generic function with 1 method)</code></pre><p>Then, we can substitute <span>$V$</span> into our full problem to create:</p><p class="math-container">\[\begin{array}{r l}
+\min\limits_{x} &amp; x_1^2 + x_2^2 + V(x_1, x_2) \\
+s.t.            &amp; x \ge 0.
+\end{array}\]</p><p>This looks like a nonlinear optimization problem with a user-defined operator <span>$V$</span>! However, because <span>$V$</span> solves an optimization problem internally, we can&#39;t use automatic differentiation to compute the first and second derivatives. Instead, we can use JuMP&#39;s ability to pass callback functions for the gradient and Hessian instead.</p><p>First up, we need to define the gradient of <span>$V$</span> with respect to <span>$x$</span>. In general, this may be difficult to compute, but because <span>$x$</span> appears only in the objective, we can just differentiate the objective function with respect to <span>$x$</span>, giving:</p><pre><code class="language-julia hljs">function ∇V(g::AbstractVector, x...)
+    _, y = solve_lower_level(x...)
+    g[1] = 2 * x[1] * y[1] - y[1]^4
+    g[2] = 2 * x[2] * y[2] - 2 * y[2]^4
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">∇V (generic function with 1 method)</code></pre><p>Second, we need to define the Hessian of <span>$V$</span> with respect to <span>$x$</span>. This is a symmetric matrix, but in our example only the diagonal elements are non-zero:</p><pre><code class="language-julia hljs">function ∇²V(H::AbstractMatrix, x...)
+    _, y = solve_lower_level(x...)
+    H[1, 1] = 2 * y[1]
+    H[2, 2] = 2 * y[2]
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">∇²V (generic function with 1 method)</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Providing an explicit Hessian function is optional if first derivatives are already available.</p></div></div><p>We now have enough to define our bilevel optimization problem:</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+@variable(model, x[1:2] &gt;= 0)
+@operator(model, op_V, 2, V, ∇V, ∇²V)
+@objective(model, Min, x[1]^2 + x[2]^2 + op_V(x[1], x[2]))
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Ipopt
+
+* Status
+  Result count       : 1
+  Termination status : LOCALLY_SOLVED
+  Message from the solver:
+  &quot;Solve_Succeeded&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : -4.18983e+05
+  Dual objective value : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 7.90128e-01
+</code></pre><p>The optimal objective value is:</p><pre><code class="language-julia hljs">objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">-418983.4868064087</code></pre><p>and the optimal upper-level decision variables <span>$x$</span> are:</p><pre><code class="language-julia hljs">value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 154.97862339279123
+ 180.00961428934832</code></pre><p>To compute the optimal lower-level decision variables, we need to call <code>solve_lower_level</code> with the optimal upper-level decision variables:</p><pre><code class="language-julia hljs">_, y = solve_lower_level(value.(x)...)
+y</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 7.0725939609898285
+ 5.946569892949626</code></pre><h2 id="Improving-performance"><a class="docs-heading-anchor" href="#Improving-performance">Improving performance</a><a id="Improving-performance-1"></a><a class="docs-heading-anchor-permalink" href="#Improving-performance" title="Permalink"></a></h2><p>Our solution approach works, but it has a performance problem: every time we need to compute the value, gradient, or Hessian of <span>$V$</span>, we have to re-solve the lower-level optimization problem. This is wasteful, because we will often call the gradient and Hessian at the same point, and so solving the problem twice with the same input repeats work unnecessarily.</p><p>We can work around this by using a cache:</p><pre><code class="language-julia hljs">mutable struct Cache
+    x::Any
+    f::Float64
+    y::Vector{Float64}
+end</code></pre><p>with a function to update the cache if needed:</p><pre><code class="language-julia hljs">function _update_if_needed(cache::Cache, x...)
+    if cache.x !== x
+        cache.f, cache.y = solve_lower_level(x...)
+        cache.x = x
+    end
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">_update_if_needed (generic function with 1 method)</code></pre><p>Then, we define cached versions of out three functions which call <code>_updated_if_needed</code> and return values from the cache.</p><pre><code class="language-julia hljs">function cached_f(cache::Cache, x...)
+    _update_if_needed(cache, x...)
+    return cache.f
+end
+
+function cached_∇f(cache::Cache, g::AbstractVector, x...)
+    _update_if_needed(cache, x...)
+    g[1] = 2 * x[1] * cache.y[1] - cache.y[1]^4
+    g[2] = 2 * x[2] * cache.y[2] - 2 * cache.y[2]^4
+    return
+end
+
+function cached_∇²f(cache::Cache, H::AbstractMatrix, x...)
+    _update_if_needed(cache, x...)
+    H[1, 1] = 2 * cache.y[1]
+    H[2, 2] = 2 * cache.y[2]
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">cached_∇²f (generic function with 1 method)</code></pre><p>Now we&#39;re ready to setup and solve the upper level optimization problem:</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+@variable(model, x[1:2] &gt;= 0)
+cache = Cache(Float64[], NaN, Float64[])
+@operator(
+    model,
+    op_cached_f,
+    2,
+    (x...) -&gt; cached_f(cache, x...),
+    (g, x...) -&gt; cached_∇f(cache, g, x...),
+    (H, x...) -&gt; cached_∇²f(cache, H, x...),
+)
+@objective(model, Min, x[1]^2 + x[2]^2 + op_cached_f(x[1], x[2]))
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Ipopt
+
+* Status
+  Result count       : 1
+  Termination status : LOCALLY_SOLVED
+  Message from the solver:
+  &quot;Solve_Succeeded&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : -4.18983e+05
+  Dual objective value : 0.00000e+00
+
+* Work counters
+  Solve time (sec)   : 3.12727e-01
+</code></pre><p>an we can check we get the same objective value:</p><pre><code class="language-julia hljs">objective_value(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">-418983.4868064087</code></pre><p>and upper-level decision variable <span>$x$</span>:</p><pre><code class="language-julia hljs">value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 154.97862339279123
+ 180.00961428934832</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../user_defined_hessians/">« User-defined Hessians</a><a class="docs-footer-nextpage" href="../querying_hessians/">Computing Hessians »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/portfolio.jl b/previews/PR3525/tutorials/nonlinear/portfolio.jl
new file mode 100644
index 00000000000..666d92b8ab9
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/portfolio.jl
@@ -0,0 +1,255 @@
+# Copyright (c) 2019 Arpit Bhatia and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Portfolio optimization
+
+# **This tutorial was originally contributed by Arpit Bhatia.**
+
+# Optimization models play an increasingly important role in financial
+# decisions. Many computational finance problems can be solved efficiently using
+# modern optimization techniques.
+#
+# This tutorial solves the famous Markowitz Portfolio Optimization problem with
+# data from [lecture notes from a course taught at Georgia Tech by Shabbir Ahmed](https://www2.isye.gatech.edu/~sahmed/isye6669/).
+
+# ## Required packages
+
+# This tutorial uses the following packages
+
+using JuMP
+import DataFrames
+import Ipopt
+import MultiObjectiveAlgorithms as MOA
+import Plots
+import Statistics
+import StatsPlots
+
+# ## Formulation
+
+# Suppose we are considering investing 1000 dollars in three non-dividend paying
+# stocks, IBM (IBM), Walmart (WMT), and Southern Electric (SEHI), for a
+# one-month period.
+
+# We will use the initial money to buy shares of the three stocks at the
+# current market prices, hold these for one month, and sell the shares off at
+# the prevailing market prices at the end of the month.
+#
+# As a rational investor, we hope to make some profit out of this endeavor,
+# that is, the return on our investment should be positive.
+#
+# Suppose we bought a stock at $p$ dollars per share in the beginning of the
+# month, and sold it off at $s$ dollars per share at the end of the month. Then
+# the one-month return on a share of the stock is $\frac{s-p}{p}$.
+
+# Since the stock prices are quite uncertain, so is the end-of-month return on
+# our investment. Our goal is to invest in such a way that the expected
+# end-of-month return is at least \$50 or 5%. Furthermore, we want to make
+# sure that the “risk” of not achieving our desired return is minimum.
+
+# Note that we are solving the problem under the following assumptions:
+# 1. We can trade any continuum of shares.
+# 2. No short-selling is allowed.
+# 3. There are no transaction costs.
+
+# We model this problem by taking decision variables $x_{i}, i=1,2,3,$ denoting
+# the dollars invested in each of the 3 stocks.
+
+# Let us denote by $\tilde{r}_{i}$ the random variable corresponding to the
+# monthly return (increase in the stock price) per dollar for stock $i$.
+
+# Then, the return (or profit) on $x_{i}$ dollars invested in stock $i$ is
+# $\tilde{r}_{i} x_{i},$ and the total (random) return on our investment is
+# $\sum_{i=1}^{3} \tilde{r}_{i} x_{i}.$ The expected return on our investment is
+# then $\mathbb{E}\left[\sum_{i=1}^{3} \tilde{r}_{i} x_{i}\right]=\sum_{i=1}^{3} \overline{r}_{i} x_{i},$
+# where $\overline{r}_{i}$ is the expected value of the $\tilde{r}_{i}.$
+
+# Now we need to quantify the notion of “risk” in our investment.
+
+# Markowitz, in his Nobel prize winning work, showed that a rational investor’s
+# notion of minimizing risk can be closely approximated by minimizing the
+# variance of the return of the investment portfolio. This variance is given by:
+
+# ```math
+# \operatorname{Var}\left[\sum_{i=1}^{3} \tilde{r}_{i} x_{i}\right] = \sum_{i=1}^{3} \sum_{j=1}^{3} x_{i} x_{j} \sigma_{i j}
+# ```
+
+# where $\sigma_{i j}$ is the covariance of the return of stock $i$ with stock $j$.
+
+# Note that the right hand side of the equation is the most reduced form of the
+# expression and we have not shown the intermediate steps involved in getting to
+# this form. We can also write this equation as:
+
+# ```math
+# \operatorname{Var}\left[\sum_{i=1}^{3} \tilde{r}_{i} x_{i}\right] =x^\top Q x
+# ```
+
+# Where $Q$ is the covariance matrix for the random vector $\tilde{r}$.
+
+# Finally, we can write the model as:
+
+# ```math
+# \begin{aligned}
+# \min x^\top Q x \\
+# \text { s.t. } \sum_{i=1}^{3} x_{i} \leq 1000 \\
+# \overline{r}^\top x \geq 50 \\
+# x \geq 0
+# \end{aligned}
+# ```
+
+# ## Data
+
+# For the data in our problem, we use the stock prices given below, in monthly
+# values from November 2000, through November 2001.
+
+df = DataFrames.DataFrame(
+    [
+        93.043 51.826 1.063
+        84.585 52.823 0.938
+        111.453 56.477 1.000
+        99.525 49.805 0.938
+        95.819 50.287 1.438
+        114.708 51.521 1.700
+        111.515 51.531 2.540
+        113.211 48.664 2.390
+        104.942 55.744 3.120
+        99.827 47.916 2.980
+        91.607 49.438 1.900
+        107.937 51.336 1.750
+        115.590 55.081 1.800
+    ],
+    [:IBM, :WMT, :SEHI],
+)
+
+# Next, we compute the percentage return for the stock in each month:
+
+returns = diff(Matrix(df); dims = 1) ./ Matrix(df[1:end-1, :])
+
+# The expected monthly return is:
+
+r = vec(Statistics.mean(returns; dims = 1))
+
+# and the covariance matrix is:
+
+Q = Statistics.cov(returns)
+
+# ## JuMP formulation
+
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[1:3] >= 0)
+@objective(model, Min, x' * Q * x)
+@constraint(model, sum(x) <= 1000)
+@constraint(model, r' * x >= 50)
+optimize!(model)
+solution_summary(model)
+
+# The optimal allocation of our assets is:
+
+value.(x)
+
+# So we spend \$497 on IBM, and \$503 on SEHI. This results in a variance of:
+
+scalar_variance = value(x' * Q * x)
+
+# and an expected return of:
+
+scalar_return = value(r' * x)
+
+# ## Multi-objective portfolio optimization
+
+# The previous model returned a single solution that minimized the variance,
+# ensuring that our expected return was at least \$50. In practice, we might
+# be willing to accept a slightly higher variance if it meant a much increased
+# expected return. To explore this problem space, we can instead formulate our
+# portfolio optimization problem with two objectives:
+#
+#  1. to minimize the variance
+#  2. to maximize the expected return
+#
+# The solution to this bi-objective problem is the
+# [efficient frontier](https://en.wikipedia.org/wiki/Efficient_frontier) of
+# modern portfolio theory, and each point in the solution is a point with the
+# best return for a fixed level of risk.
+
+model = Model(() -> MOA.Optimizer(Ipopt.Optimizer))
+set_silent(model)
+
+# We also need to choose a solution algorithm for `MOA`. For our problem, the
+# efficient frontier will have an infinite number of solutions. Since we cannot
+# find all of the solutions, we choose an approximation algorithm and limit the
+# number of solution points that are returned:
+
+set_optimizer_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
+set_optimizer_attribute(model, MOA.SolutionLimit(), 50)
+
+# Now we can define the rest of the model:
+
+@variable(model, x[1:3] >= 0)
+@constraint(model, sum(x) <= 1000)
+@expression(model, variance, x' * Q * x)
+@expression(model, expected_return, r' * x)
+## We want to minimize variance and maximize expected return, but we must pick
+## a single objective sense `Min`, and negate any `Max` objectives:
+@objective(model, Min, [variance, -expected_return])
+optimize!(model)
+solution_summary(model)
+
+# The algorithm found 50 different solutions. Let's plot them to see how they
+# differ:
+
+objective_space = Plots.hline(
+    [scalar_return];
+    label = "Single-objective solution",
+    linecolor = :red,
+)
+Plots.vline!(objective_space, [scalar_variance]; label = "", linecolor = :red)
+Plots.scatter!(
+    objective_space,
+    [value(variance; result = i) for i in 1:result_count(model)],
+    [value(expected_return; result = i) for i in 1:result_count(model)];
+    xlabel = "Variance",
+    ylabel = "Expected Return",
+    label = "",
+    title = "Objective space",
+    markercolor = "white",
+    markersize = 5,
+    legend = :bottomright,
+)
+for i in 1:result_count(model)
+    y = objective_value(model; result = i)
+    Plots.annotate!(objective_space, y[1], -y[2], (i, 3))
+end
+
+decision_space = StatsPlots.groupedbar(
+    vcat([value.(x; result = i)' for i in 1:result_count(model)]...);
+    bar_position = :stack,
+    label = ["IBM" "WMT" "SEHI"],
+    xlabel = "Solution #",
+    ylabel = "Investment (\$)",
+    title = "Decision space",
+)
+Plots.plot(objective_space, decision_space; layout = (2, 1), size = (600, 600))
+
+# Perhaps our trade-off wasn't so bad after all. Our original solution
+# corresponded to picking a solution #17. If we buy more SEHI, we can increase
+# the return, but the variance also increases. If we buy less SEHI, such as a
+# solution like #5 or #6, then we can achieve the corresponding return without
+# deploying all of our capital. We should also note that at no point should we
+# buy WMT.
diff --git a/previews/PR3525/tutorials/nonlinear/portfolio/319d794d.svg b/previews/PR3525/tutorials/nonlinear/portfolio/319d794d.svg
new file mode 100644
index 00000000000..e990c4d7b23
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/portfolio/319d794d.svg
@@ -0,0 +1,581 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="600" viewBox="0 0 2400 2400">
+<defs>
+  <clipPath id="clip510">
+    <rect x="0" y="0" width="2400" height="2400"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip510)" d="M0 2400 L2400 2400 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip511">
+    <rect x="480" y="240" width="1681" height="1681"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip510)" d="M286.311 964.927 L2352.76 964.927 L2352.76 153.712 L286.311 153.712  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip512">
+    <rect x="286" y="153" width="2067" height="812"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="344.796,964.927 344.796,153.712 "/>
+<polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="912.471,964.927 912.471,153.712 "/>
+<polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1480.15,964.927 1480.15,153.712 "/>
+<polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2047.82,964.927 2047.82,153.712 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,964.927 2352.76,964.927 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="344.796,964.927 344.796,946.03 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="912.471,964.927 912.471,946.03 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1480.15,964.927 1480.15,946.03 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2047.82,964.927 2047.82,946.03 "/>
+<path clip-path="url(#clip510)" d="M344.796 995.846 Q341.184 995.846 339.356 999.411 Q337.55 1002.95 337.55 1010.08 Q337.55 1017.19 339.356 1020.75 Q341.184 1024.3 344.796 1024.3 Q348.43 1024.3 350.235 1020.75 Q352.064 1017.19 352.064 1010.08 Q352.064 1002.95 350.235 999.411 Q348.43 995.846 344.796 995.846 M344.796 992.142 Q350.606 992.142 353.661 996.749 Q356.74 1001.33 356.74 1010.08 Q356.74 1018.81 353.661 1023.42 Q350.606 1028 344.796 1028 Q338.985 1028 335.907 1023.42 Q332.851 1018.81 332.851 1010.08 Q332.851 1001.33 335.907 996.749 Q338.985 992.142 344.796 992.142 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M823.455 1039.05 L839.775 1039.05 L839.775 1042.99 L817.83 1042.99 L817.83 1039.05 Q820.492 1036.3 825.076 1031.67 Q829.682 1027.02 830.863 1025.67 Q833.108 1023.15 833.988 1021.41 Q834.89 1019.65 834.89 1017.96 Q834.89 1015.21 832.946 1013.47 Q831.025 1011.74 827.923 1011.74 Q825.724 1011.74 823.27 1012.5 Q820.84 1013.27 818.062 1014.82 L818.062 1010.09 Q820.886 1008.96 823.34 1008.38 Q825.793 1007.8 827.83 1007.8 Q833.201 1007.8 836.395 1010.49 Q839.589 1013.17 839.589 1017.66 Q839.589 1019.79 838.779 1021.71 Q837.992 1023.61 835.886 1026.21 Q835.307 1026.88 832.205 1030.09 Q829.103 1033.29 823.455 1039.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M849.589 1037.11 L854.474 1037.11 L854.474 1042.99 L849.589 1042.99 L849.589 1037.11 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M874.659 1011.51 Q871.048 1011.51 869.219 1015.07 Q867.413 1018.61 867.413 1025.74 Q867.413 1032.85 869.219 1036.41 Q871.048 1039.96 874.659 1039.96 Q878.293 1039.96 880.098 1036.41 Q881.927 1032.85 881.927 1025.74 Q881.927 1018.61 880.098 1015.07 Q878.293 1011.51 874.659 1011.51 M874.659 1007.8 Q880.469 1007.8 883.524 1012.41 Q886.603 1016.99 886.603 1025.74 Q886.603 1034.47 883.524 1039.08 Q880.469 1043.66 874.659 1043.66 Q868.849 1043.66 865.77 1039.08 Q862.714 1034.47 862.714 1025.74 Q862.714 1016.99 865.77 1012.41 Q868.849 1007.8 874.659 1007.8 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M922.992 1017.52 L912.413 1028.15 L922.992 1038.73 L920.237 1041.53 L909.612 1030.9 L898.987 1041.53 L896.256 1038.73 L906.811 1028.15 L896.256 1017.52 L898.987 1014.72 L909.612 1025.35 L920.237 1014.72 L922.992 1017.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M935.353 1039.05 L942.992 1039.05 L942.992 1012.69 L934.682 1014.35 L934.682 1010.09 L942.945 1008.43 L947.621 1008.43 L947.621 1039.05 L955.26 1039.05 L955.26 1042.99 L935.353 1042.99 L935.353 1039.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M974.704 1011.51 Q971.093 1011.51 969.265 1015.07 Q967.459 1018.61 967.459 1025.74 Q967.459 1032.85 969.265 1036.41 Q971.093 1039.96 974.704 1039.96 Q978.339 1039.96 980.144 1036.41 Q981.973 1032.85 981.973 1025.74 Q981.973 1018.61 980.144 1015.07 Q978.339 1011.51 974.704 1011.51 M974.704 1007.8 Q980.515 1007.8 983.57 1012.41 Q986.649 1016.99 986.649 1025.74 Q986.649 1034.47 983.57 1039.08 Q980.515 1043.66 974.704 1043.66 Q968.894 1043.66 965.816 1039.08 Q962.76 1034.47 962.76 1025.74 Q962.76 1016.99 965.816 1012.41 Q968.894 1007.8 974.704 1007.8 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M999.325 990.807 L989.733 1005.8 L999.325 1005.8 L999.325 990.807 M998.328 987.497 L1003.11 987.497 L1003.11 1005.8 L1007.11 1005.8 L1007.11 1008.96 L1003.11 1008.96 L1003.11 1015.58 L999.325 1015.58 L999.325 1008.96 L986.649 1008.96 L986.649 1005.29 L998.328 987.497 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1400.53 1012.5 L1388.72 1030.95 L1400.53 1030.95 L1400.53 1012.5 M1399.3 1008.43 L1405.18 1008.43 L1405.18 1030.95 L1410.11 1030.95 L1410.11 1034.84 L1405.18 1034.84 L1405.18 1042.99 L1400.53 1042.99 L1400.53 1034.84 L1384.93 1034.84 L1384.93 1030.33 L1399.3 1008.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1417.84 1037.11 L1422.73 1037.11 L1422.73 1042.99 L1417.84 1042.99 L1417.84 1037.11 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1442.91 1011.51 Q1439.3 1011.51 1437.47 1015.07 Q1435.67 1018.61 1435.67 1025.74 Q1435.67 1032.85 1437.47 1036.41 Q1439.3 1039.96 1442.91 1039.96 Q1446.55 1039.96 1448.35 1036.41 Q1450.18 1032.85 1450.18 1025.74 Q1450.18 1018.61 1448.35 1015.07 Q1446.55 1011.51 1442.91 1011.51 M1442.91 1007.8 Q1448.72 1007.8 1451.78 1012.41 Q1454.86 1016.99 1454.86 1025.74 Q1454.86 1034.47 1451.78 1039.08 Q1448.72 1043.66 1442.91 1043.66 Q1437.1 1043.66 1434.02 1039.08 Q1430.97 1034.47 1430.97 1025.74 Q1430.97 1016.99 1434.02 1012.41 Q1437.1 1007.8 1442.91 1007.8 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1491.25 1017.52 L1480.67 1028.15 L1491.25 1038.73 L1488.49 1041.53 L1477.87 1030.9 L1467.24 1041.53 L1464.51 1038.73 L1475.07 1028.15 L1464.51 1017.52 L1467.24 1014.72 L1477.87 1025.35 L1488.49 1014.72 L1491.25 1017.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1503.61 1039.05 L1511.25 1039.05 L1511.25 1012.69 L1502.94 1014.35 L1502.94 1010.09 L1511.2 1008.43 L1515.88 1008.43 L1515.88 1039.05 L1523.51 1039.05 L1523.51 1042.99 L1503.61 1042.99 L1503.61 1039.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1542.96 1011.51 Q1539.35 1011.51 1537.52 1015.07 Q1535.71 1018.61 1535.71 1025.74 Q1535.71 1032.85 1537.52 1036.41 Q1539.35 1039.96 1542.96 1039.96 Q1546.59 1039.96 1548.4 1036.41 Q1550.23 1032.85 1550.23 1025.74 Q1550.23 1018.61 1548.4 1015.07 Q1546.59 1011.51 1542.96 1011.51 M1542.96 1007.8 Q1548.77 1007.8 1551.82 1012.41 Q1554.9 1016.99 1554.9 1025.74 Q1554.9 1034.47 1551.82 1039.08 Q1548.77 1043.66 1542.96 1043.66 Q1537.15 1043.66 1534.07 1039.08 Q1531.01 1034.47 1531.01 1025.74 Q1531.01 1016.99 1534.07 1012.41 Q1537.15 1007.8 1542.96 1007.8 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1567.58 990.807 L1557.99 1005.8 L1567.58 1005.8 L1567.58 990.807 M1566.58 987.497 L1571.36 987.497 L1571.36 1005.8 L1575.37 1005.8 L1575.37 1008.96 L1571.36 1008.96 L1571.36 1015.58 L1567.58 1015.58 L1567.58 1008.96 L1554.9 1008.96 L1554.9 1005.29 L1566.58 987.497 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1965.44 1023.84 Q1962.29 1023.84 1960.44 1026 Q1958.61 1028.15 1958.61 1031.9 Q1958.61 1035.63 1960.44 1037.8 Q1962.29 1039.96 1965.44 1039.96 Q1968.59 1039.96 1970.41 1037.8 Q1972.27 1035.63 1972.27 1031.9 Q1972.27 1028.15 1970.41 1026 Q1968.59 1023.84 1965.44 1023.84 M1974.72 1009.19 L1974.72 1013.45 Q1972.96 1012.62 1971.16 1012.18 Q1969.37 1011.74 1967.61 1011.74 Q1962.98 1011.74 1960.53 1014.86 Q1958.1 1017.99 1957.75 1024.31 Q1959.12 1022.29 1961.18 1021.23 Q1963.24 1020.14 1965.72 1020.14 Q1970.92 1020.14 1973.93 1023.31 Q1976.97 1026.46 1976.97 1031.9 Q1976.97 1037.22 1973.82 1040.44 Q1970.67 1043.66 1965.44 1043.66 Q1959.44 1043.66 1956.27 1039.08 Q1953.1 1034.47 1953.1 1025.74 Q1953.1 1017.55 1956.99 1012.69 Q1960.88 1007.8 1967.43 1007.8 Q1969.19 1007.8 1970.97 1008.15 Q1972.78 1008.5 1974.72 1009.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1985.02 1037.11 L1989.91 1037.11 L1989.91 1042.99 L1985.02 1042.99 L1985.02 1037.11 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2010.09 1011.51 Q2006.48 1011.51 2004.65 1015.07 Q2002.85 1018.61 2002.85 1025.74 Q2002.85 1032.85 2004.65 1036.41 Q2006.48 1039.96 2010.09 1039.96 Q2013.72 1039.96 2015.53 1036.41 Q2017.36 1032.85 2017.36 1025.74 Q2017.36 1018.61 2015.53 1015.07 Q2013.72 1011.51 2010.09 1011.51 M2010.09 1007.8 Q2015.9 1007.8 2018.96 1012.41 Q2022.04 1016.99 2022.04 1025.74 Q2022.04 1034.47 2018.96 1039.08 Q2015.9 1043.66 2010.09 1043.66 Q2004.28 1043.66 2001.2 1039.08 Q1998.15 1034.47 1998.15 1025.74 Q1998.15 1016.99 2001.2 1012.41 Q2004.28 1007.8 2010.09 1007.8 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2058.42 1017.52 L2047.85 1028.15 L2058.42 1038.73 L2055.67 1041.53 L2045.04 1030.9 L2034.42 1041.53 L2031.69 1038.73 L2042.24 1028.15 L2031.69 1017.52 L2034.42 1014.72 L2045.04 1025.35 L2055.67 1014.72 L2058.42 1017.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2070.78 1039.05 L2078.42 1039.05 L2078.42 1012.69 L2070.11 1014.35 L2070.11 1010.09 L2078.38 1008.43 L2083.05 1008.43 L2083.05 1039.05 L2090.69 1039.05 L2090.69 1042.99 L2070.78 1042.99 L2070.78 1039.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2110.14 1011.51 Q2106.53 1011.51 2104.7 1015.07 Q2102.89 1018.61 2102.89 1025.74 Q2102.89 1032.85 2104.7 1036.41 Q2106.53 1039.96 2110.14 1039.96 Q2113.77 1039.96 2115.58 1036.41 Q2117.4 1032.85 2117.4 1025.74 Q2117.4 1018.61 2115.58 1015.07 Q2113.77 1011.51 2110.14 1011.51 M2110.14 1007.8 Q2115.95 1007.8 2119 1012.41 Q2122.08 1016.99 2122.08 1025.74 Q2122.08 1034.47 2119 1039.08 Q2115.95 1043.66 2110.14 1043.66 Q2104.33 1043.66 2101.25 1039.08 Q2098.19 1034.47 2098.19 1025.74 Q2098.19 1016.99 2101.25 1012.41 Q2104.33 1007.8 2110.14 1007.8 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2134.76 990.807 L2125.17 1005.8 L2134.76 1005.8 L2134.76 990.807 M2133.76 987.497 L2138.54 987.497 L2138.54 1005.8 L2142.54 1005.8 L2142.54 1008.96 L2138.54 1008.96 L2138.54 1015.58 L2134.76 1015.58 L2134.76 1008.96 L2122.08 1008.96 L2122.08 1005.29 L2133.76 987.497 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1198.87 1128.22 L1180.73 1080.7 L1187.45 1080.7 L1202.5 1120.7 L1217.59 1080.7 L1224.27 1080.7 L1206.16 1128.22 L1198.87 1128.22 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1242.09 1110.3 Q1235 1110.3 1232.26 1111.92 Q1229.52 1113.54 1229.52 1117.46 Q1229.52 1120.58 1231.56 1122.42 Q1233.63 1124.24 1237.16 1124.24 Q1242.03 1124.24 1244.96 1120.8 Q1247.92 1117.33 1247.92 1111.6 L1247.92 1110.3 L1242.09 1110.3 M1253.78 1107.88 L1253.78 1128.22 L1247.92 1128.22 L1247.92 1122.81 Q1245.91 1126.05 1242.92 1127.61 Q1239.93 1129.14 1235.6 1129.14 Q1230.13 1129.14 1226.88 1126.08 Q1223.67 1123 1223.67 1117.84 Q1223.67 1111.82 1227.68 1108.77 Q1231.72 1105.71 1239.71 1105.71 L1247.92 1105.71 L1247.92 1105.14 Q1247.92 1101.1 1245.25 1098.9 Q1242.6 1096.67 1237.8 1096.67 Q1234.74 1096.67 1231.85 1097.41 Q1228.95 1098.14 1226.28 1099.6 L1226.28 1094.19 Q1229.49 1092.95 1232.51 1092.35 Q1235.54 1091.71 1238.4 1091.71 Q1246.14 1091.71 1249.96 1095.72 Q1253.78 1099.73 1253.78 1107.88 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1286.5 1098.04 Q1285.51 1097.47 1284.33 1097.22 Q1283.19 1096.93 1281.78 1096.93 Q1276.82 1096.93 1274.15 1100.18 Q1271.5 1103.39 1271.5 1109.44 L1271.5 1128.22 L1265.62 1128.22 L1265.62 1092.57 L1271.5 1092.57 L1271.5 1098.11 Q1273.35 1094.86 1276.31 1093.3 Q1279.27 1091.71 1283.5 1091.71 Q1284.11 1091.71 1284.84 1091.8 Q1285.57 1091.87 1286.46 1092.03 L1286.5 1098.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1292.64 1092.57 L1298.49 1092.57 L1298.49 1128.22 L1292.64 1128.22 L1292.64 1092.57 M1292.64 1078.69 L1298.49 1078.69 L1298.49 1086.11 L1292.64 1086.11 L1292.64 1078.69 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1326.95 1110.3 Q1319.85 1110.3 1317.11 1111.92 Q1314.38 1113.54 1314.38 1117.46 Q1314.38 1120.58 1316.41 1122.42 Q1318.48 1124.24 1322.02 1124.24 Q1326.89 1124.24 1329.81 1120.8 Q1332.77 1117.33 1332.77 1111.6 L1332.77 1110.3 L1326.95 1110.3 M1338.63 1107.88 L1338.63 1128.22 L1332.77 1128.22 L1332.77 1122.81 Q1330.77 1126.05 1327.78 1127.61 Q1324.79 1129.14 1320.46 1129.14 Q1314.98 1129.14 1311.74 1126.08 Q1308.52 1123 1308.52 1117.84 Q1308.52 1111.82 1312.53 1108.77 Q1316.57 1105.71 1324.56 1105.71 L1332.77 1105.71 L1332.77 1105.14 Q1332.77 1101.1 1330.1 1098.9 Q1327.46 1096.67 1322.65 1096.67 Q1319.6 1096.67 1316.7 1097.41 Q1313.8 1098.14 1311.13 1099.6 L1311.13 1094.19 Q1314.35 1092.95 1317.37 1092.35 Q1320.39 1091.71 1323.26 1091.71 Q1330.99 1091.71 1334.81 1095.72 Q1338.63 1099.73 1338.63 1107.88 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1380.33 1106.7 L1380.33 1128.22 L1374.47 1128.22 L1374.47 1106.89 Q1374.47 1101.83 1372.5 1099.32 Q1370.52 1096.8 1366.58 1096.8 Q1361.83 1096.8 1359.1 1099.83 Q1356.36 1102.85 1356.36 1108.07 L1356.36 1128.22 L1350.47 1128.22 L1350.47 1092.57 L1356.36 1092.57 L1356.36 1098.11 Q1358.46 1094.89 1361.29 1093.3 Q1364.16 1091.71 1367.88 1091.71 Q1374.02 1091.71 1377.18 1095.53 Q1380.33 1099.32 1380.33 1106.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1417.66 1093.94 L1417.66 1099.41 Q1415.18 1098.04 1412.66 1097.37 Q1410.18 1096.67 1407.63 1096.67 Q1401.94 1096.67 1398.79 1100.3 Q1395.64 1103.9 1395.64 1110.42 Q1395.64 1116.95 1398.79 1120.58 Q1401.94 1124.17 1407.63 1124.17 Q1410.18 1124.17 1412.66 1123.51 Q1415.18 1122.81 1417.66 1121.44 L1417.66 1126.85 Q1415.21 1127.99 1412.57 1128.57 Q1409.96 1129.14 1407 1129.14 Q1398.95 1129.14 1394.2 1124.08 Q1389.46 1119.02 1389.46 1110.42 Q1389.46 1101.7 1394.24 1096.71 Q1399.04 1091.71 1407.38 1091.71 Q1410.09 1091.71 1412.66 1092.28 Q1415.24 1092.82 1417.66 1093.94 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1458.34 1108.93 L1458.34 1111.79 L1431.41 1111.79 Q1431.79 1117.84 1435.04 1121.02 Q1438.32 1124.17 1444.14 1124.17 Q1447.52 1124.17 1450.67 1123.35 Q1453.85 1122.52 1456.97 1120.86 L1456.97 1126.4 Q1453.82 1127.74 1450.51 1128.44 Q1447.2 1129.14 1443.79 1129.14 Q1435.26 1129.14 1430.27 1124.17 Q1425.3 1119.21 1425.3 1110.74 Q1425.3 1101.99 1430.01 1096.87 Q1434.75 1091.71 1442.77 1091.71 Q1449.97 1091.71 1454.14 1096.36 Q1458.34 1100.97 1458.34 1108.93 M1452.48 1107.21 Q1452.42 1102.4 1449.78 1099.54 Q1447.17 1096.67 1442.84 1096.67 Q1437.94 1096.67 1434.98 1099.44 Q1432.05 1102.21 1431.6 1107.24 L1452.48 1107.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,941.969 2352.76,941.969 "/>
+<polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,734.335 2352.76,734.335 "/>
+<polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,526.7 2352.76,526.7 "/>
+<polyline clip-path="url(#clip512)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,319.066 2352.76,319.066 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,964.927 286.311,153.712 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,941.969 305.209,941.969 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,734.335 305.209,734.335 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,526.7 305.209,526.7 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,319.066 305.209,319.066 "/>
+<path clip-path="url(#clip510)" d="M238.367 927.768 Q234.756 927.768 232.927 931.333 Q231.122 934.874 231.122 942.004 Q231.122 949.11 232.927 952.675 Q234.756 956.217 238.367 956.217 Q242.001 956.217 243.807 952.675 Q245.635 949.11 245.635 942.004 Q245.635 934.874 243.807 931.333 Q242.001 927.768 238.367 927.768 M238.367 924.064 Q244.177 924.064 247.233 928.671 Q250.311 933.254 250.311 942.004 Q250.311 950.731 247.233 955.337 Q244.177 959.92 238.367 959.92 Q232.557 959.92 229.478 955.337 Q226.422 950.731 226.422 942.004 Q226.422 933.254 229.478 928.671 Q232.557 924.064 238.367 924.064 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M202.233 747.679 L218.552 747.679 L218.552 751.615 L196.608 751.615 L196.608 747.679 Q199.27 744.925 203.853 740.295 Q208.46 735.643 209.64 734.3 Q211.886 731.777 212.765 730.041 Q213.668 728.281 213.668 726.592 Q213.668 723.837 211.723 722.101 Q209.802 720.365 206.7 720.365 Q204.501 720.365 202.048 721.129 Q199.617 721.893 196.839 723.444 L196.839 718.721 Q199.663 717.587 202.117 717.008 Q204.571 716.43 206.608 716.43 Q211.978 716.43 215.173 719.115 Q218.367 721.8 218.367 726.291 Q218.367 728.42 217.557 730.342 Q216.77 732.24 214.663 734.832 Q214.085 735.504 210.983 738.721 Q207.881 741.916 202.233 747.679 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M238.367 720.133 Q234.756 720.133 232.927 723.698 Q231.122 727.24 231.122 734.369 Q231.122 741.476 232.927 745.041 Q234.756 748.582 238.367 748.582 Q242.001 748.582 243.807 745.041 Q245.635 741.476 245.635 734.369 Q245.635 727.24 243.807 723.698 Q242.001 720.133 238.367 720.133 M238.367 716.43 Q244.177 716.43 247.233 721.036 Q250.311 725.619 250.311 734.369 Q250.311 743.096 247.233 747.703 Q244.177 752.286 238.367 752.286 Q232.557 752.286 229.478 747.703 Q226.422 743.096 226.422 734.369 Q226.422 725.619 229.478 721.036 Q232.557 716.43 238.367 716.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M211.052 513.494 L199.247 531.943 L211.052 531.943 L211.052 513.494 M209.825 509.42 L215.705 509.42 L215.705 531.943 L220.635 531.943 L220.635 535.832 L215.705 535.832 L215.705 543.98 L211.052 543.98 L211.052 535.832 L195.45 535.832 L195.45 531.318 L209.825 509.42 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M238.367 512.499 Q234.756 512.499 232.927 516.064 Q231.122 519.605 231.122 526.735 Q231.122 533.841 232.927 537.406 Q234.756 540.948 238.367 540.948 Q242.001 540.948 243.807 537.406 Q245.635 533.841 245.635 526.735 Q245.635 519.605 243.807 516.064 Q242.001 512.499 238.367 512.499 M238.367 508.795 Q244.177 508.795 247.233 513.402 Q250.311 517.985 250.311 526.735 Q250.311 535.462 247.233 540.068 Q244.177 544.652 238.367 544.652 Q232.557 544.652 229.478 540.068 Q226.422 535.462 226.422 526.735 Q226.422 517.985 229.478 513.402 Q232.557 508.795 238.367 508.795 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M208.784 317.202 Q205.636 317.202 203.784 319.355 Q201.955 321.508 201.955 325.258 Q201.955 328.985 203.784 331.161 Q205.636 333.313 208.784 333.313 Q211.932 333.313 213.76 331.161 Q215.612 328.985 215.612 325.258 Q215.612 321.508 213.76 319.355 Q211.932 317.202 208.784 317.202 M218.066 302.55 L218.066 306.809 Q216.307 305.976 214.501 305.536 Q212.719 305.096 210.96 305.096 Q206.33 305.096 203.876 308.221 Q201.446 311.346 201.099 317.665 Q202.464 315.651 204.524 314.587 Q206.585 313.499 209.061 313.499 Q214.27 313.499 217.279 316.67 Q220.311 319.818 220.311 325.258 Q220.311 330.582 217.163 333.8 Q214.015 337.017 208.784 337.017 Q202.788 337.017 199.617 332.434 Q196.446 327.827 196.446 319.101 Q196.446 310.906 200.335 306.045 Q204.224 301.161 210.774 301.161 Q212.534 301.161 214.316 301.508 Q216.122 301.855 218.066 302.55 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M238.367 304.864 Q234.756 304.864 232.927 308.429 Q231.122 311.971 231.122 319.101 Q231.122 326.207 232.927 329.772 Q234.756 333.313 238.367 333.313 Q242.001 333.313 243.807 329.772 Q245.635 326.207 245.635 319.101 Q245.635 311.971 243.807 308.429 Q242.001 304.864 238.367 304.864 M238.367 301.161 Q244.177 301.161 247.233 305.767 Q250.311 310.351 250.311 319.101 Q250.311 327.827 247.233 332.434 Q244.177 337.017 238.367 337.017 Q232.557 337.017 229.478 332.434 Q226.422 327.827 226.422 319.101 Q226.422 310.351 229.478 305.767 Q232.557 301.161 238.367 301.161 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M97.6744 823.926 L97.6744 793.88 L103.085 793.88 L103.085 817.497 L117.153 817.497 L117.153 794.867 L122.564 794.867 L122.564 817.497 L139.784 817.497 L139.784 793.307 L145.194 793.307 L145.194 823.926 L97.6744 823.926 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M109.546 753.362 L126.893 766.253 L145.194 752.694 L145.194 759.601 L131.19 769.977 L145.194 780.353 L145.194 787.26 L126.543 773.414 L109.546 786.082 L109.546 779.175 L122.246 769.722 L109.546 760.269 L109.546 753.362 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M139.847 738.753 L158.753 738.753 L158.753 744.641 L109.546 744.641 L109.546 738.753 L114.957 738.753 Q111.774 736.907 110.247 734.106 Q108.687 731.273 108.687 727.359 Q108.687 720.866 113.843 716.823 Q119 712.749 127.402 712.749 Q135.805 712.749 140.961 716.823 Q146.117 720.866 146.117 727.359 Q146.117 731.273 144.59 734.106 Q143.03 736.907 139.847 738.753 M127.402 718.828 Q120.941 718.828 117.281 721.502 Q113.589 724.144 113.589 728.791 Q113.589 733.438 117.281 736.111 Q120.941 738.753 127.402 738.753 Q133.863 738.753 137.556 736.111 Q141.216 733.438 141.216 728.791 Q141.216 724.144 137.556 721.502 Q133.863 718.828 127.402 718.828 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M125.906 672.55 L128.771 672.55 L128.771 699.477 Q134.818 699.095 138.001 695.848 Q141.152 692.57 141.152 686.745 Q141.152 683.371 140.325 680.22 Q139.497 677.038 137.842 673.918 L143.38 673.918 Q144.717 677.069 145.417 680.38 Q146.117 683.69 146.117 687.095 Q146.117 695.625 141.152 700.623 Q136.187 705.588 127.721 705.588 Q118.968 705.588 113.843 700.877 Q108.687 696.135 108.687 688.114 Q108.687 680.921 113.334 676.751 Q117.949 672.55 125.906 672.55 M124.188 678.406 Q119.381 678.47 116.517 681.112 Q113.652 683.722 113.652 688.05 Q113.652 692.952 116.421 695.912 Q119.19 698.84 124.219 699.286 L124.188 678.406 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M110.915 637.284 L116.39 637.284 Q115.021 639.766 114.353 642.281 Q113.652 644.763 113.652 647.31 Q113.652 653.007 117.281 656.158 Q120.877 659.309 127.402 659.309 Q133.927 659.309 137.556 656.158 Q141.152 653.007 141.152 647.31 Q141.152 644.763 140.484 642.281 Q139.784 639.766 138.415 637.284 L143.826 637.284 Q144.972 639.735 145.545 642.376 Q146.117 644.986 146.117 647.946 Q146.117 655.999 141.057 660.741 Q135.996 665.484 127.402 665.484 Q118.681 665.484 113.684 660.71 Q108.687 655.903 108.687 647.564 Q108.687 644.859 109.26 642.281 Q109.801 639.703 110.915 637.284 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M99.425 621.306 L109.546 621.306 L109.546 609.243 L114.098 609.243 L114.098 621.306 L133.45 621.306 Q137.81 621.306 139.051 620.128 Q140.293 618.919 140.293 615.258 L140.293 609.243 L145.194 609.243 L145.194 615.258 Q145.194 622.038 142.68 624.616 Q140.134 627.194 133.45 627.194 L114.098 627.194 L114.098 631.491 L109.546 631.491 L109.546 627.194 L99.425 627.194 L99.425 621.306 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M125.906 571.049 L128.771 571.049 L128.771 597.976 Q134.818 597.594 138.001 594.347 Q141.152 591.069 141.152 585.244 Q141.152 581.87 140.325 578.719 Q139.497 575.536 137.842 572.417 L143.38 572.417 Q144.717 575.568 145.417 578.878 Q146.117 582.189 146.117 585.594 Q146.117 594.124 141.152 599.121 Q136.187 604.087 127.721 604.087 Q118.968 604.087 113.843 599.376 Q108.687 594.634 108.687 586.613 Q108.687 579.42 113.334 575.25 Q117.949 571.049 125.906 571.049 M124.188 576.905 Q119.381 576.969 116.517 579.61 Q113.652 582.22 113.652 586.549 Q113.652 591.451 116.421 594.411 Q119.19 597.339 124.219 597.785 L124.188 576.905 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M114.957 537.979 L95.6692 537.979 L95.6692 532.122 L145.194 532.122 L145.194 537.979 L139.847 537.979 Q143.03 539.825 144.59 542.658 Q146.117 545.458 146.117 549.405 Q146.117 555.866 140.961 559.94 Q135.805 563.983 127.402 563.983 Q119 563.983 113.843 559.94 Q108.687 555.866 108.687 549.405 Q108.687 545.458 110.247 542.658 Q111.774 539.825 114.957 537.979 M127.402 557.935 Q133.863 557.935 137.556 555.293 Q141.216 552.62 141.216 547.973 Q141.216 543.326 137.556 540.652 Q133.863 537.979 127.402 537.979 Q120.941 537.979 117.281 540.652 Q113.589 543.326 113.589 547.973 Q113.589 552.62 117.281 555.293 Q120.941 557.935 127.402 557.935 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M122.914 476.55 Q123.615 474.481 125.906 472.539 Q128.198 470.566 132.208 468.593 L145.194 462.068 L145.194 468.974 L133.004 475.054 Q128.23 477.409 126.67 479.637 Q125.111 481.833 125.111 485.653 L125.111 492.655 L145.194 492.655 L145.194 499.084 L97.6744 499.084 L97.6744 484.57 Q97.6744 476.422 101.08 472.412 Q104.486 468.402 111.361 468.402 Q115.848 468.402 118.809 470.502 Q121.769 472.571 122.914 476.55 M102.958 492.655 L119.827 492.655 L119.827 484.57 Q119.827 479.924 117.695 477.568 Q115.53 475.181 111.361 475.181 Q107.191 475.181 105.09 477.568 Q102.958 479.924 102.958 484.57 L102.958 492.655 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M125.906 426.483 L128.771 426.483 L128.771 453.41 Q134.818 453.028 138.001 449.782 Q141.152 446.504 141.152 440.679 Q141.152 437.305 140.325 434.154 Q139.497 430.971 137.842 427.852 L143.38 427.852 Q144.717 431.003 145.417 434.313 Q146.117 437.623 146.117 441.029 Q146.117 449.559 141.152 454.556 Q136.187 459.521 127.721 459.521 Q118.968 459.521 113.843 454.811 Q108.687 450.068 108.687 442.048 Q108.687 434.854 113.334 430.685 Q117.949 426.483 125.906 426.483 M124.188 432.34 Q119.381 432.404 116.517 435.045 Q113.652 437.655 113.652 441.984 Q113.652 446.885 116.421 449.846 Q119.19 452.774 124.219 453.219 L124.188 432.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M99.425 411.078 L109.546 411.078 L109.546 399.015 L114.098 399.015 L114.098 411.078 L133.45 411.078 Q137.81 411.078 139.051 409.901 Q140.293 408.691 140.293 405.031 L140.293 399.015 L145.194 399.015 L145.194 405.031 Q145.194 411.81 142.68 414.389 Q140.134 416.967 133.45 416.967 L114.098 416.967 L114.098 421.264 L109.546 421.264 L109.546 416.967 L99.425 416.967 L99.425 411.078 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M131.126 391.918 L109.546 391.918 L109.546 386.061 L130.903 386.061 Q135.964 386.061 138.51 384.088 Q141.025 382.114 141.025 378.168 Q141.025 373.425 138.001 370.688 Q134.977 367.919 129.758 367.919 L109.546 367.919 L109.546 362.062 L145.194 362.062 L145.194 367.919 L139.72 367.919 Q142.966 370.051 144.558 372.884 Q146.117 375.685 146.117 379.409 Q146.117 385.552 142.298 388.735 Q138.479 391.918 131.126 391.918 M108.687 377.181 L108.687 377.181 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M115.021 329.343 Q114.448 330.329 114.193 331.507 Q113.907 332.653 113.907 334.053 Q113.907 339.019 117.153 341.692 Q120.368 344.334 126.416 344.334 L145.194 344.334 L145.194 350.222 L109.546 350.222 L109.546 344.334 L115.085 344.334 Q111.838 342.488 110.279 339.528 Q108.687 336.568 108.687 332.335 Q108.687 331.73 108.783 330.998 Q108.846 330.266 109.005 329.375 L115.021 329.343 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M123.678 294.713 L145.194 294.713 L145.194 300.57 L123.869 300.57 Q118.809 300.57 116.294 302.543 Q113.78 304.516 113.78 308.463 Q113.78 313.206 116.803 315.943 Q119.827 318.68 125.047 318.68 L145.194 318.68 L145.194 324.568 L109.546 324.568 L109.546 318.68 L115.085 318.68 Q111.87 316.579 110.279 313.747 Q108.687 310.882 108.687 307.158 Q108.687 301.015 112.507 297.864 Q116.294 294.713 123.678 294.713 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1020.54 17.6457 Q1011.62 17.6457 1006.36 24.2892 Q1001.13 30.9327 1001.13 42.3968 Q1001.13 53.8203 1006.36 60.4638 Q1011.62 67.1073 1020.54 67.1073 Q1029.45 67.1073 1034.63 60.4638 Q1039.86 53.8203 1039.86 42.3968 Q1039.86 30.9327 1034.63 24.2892 Q1029.45 17.6457 1020.54 17.6457 M1020.54 11.0023 Q1033.26 11.0023 1040.87 19.5497 Q1048.49 28.0566 1048.49 42.3968 Q1048.49 56.6965 1040.87 65.2439 Q1033.26 73.7508 1020.54 73.7508 Q1007.78 73.7508 1000.12 65.2439 Q992.504 56.737 992.504 42.3968 Q992.504 28.0566 1000.12 19.5497 Q1007.78 11.0023 1020.54 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1093.53 49.9314 Q1093.53 41.7081 1090.13 37.0496 Q1086.77 32.3505 1080.85 32.3505 Q1074.94 32.3505 1071.54 37.0496 Q1068.17 41.7081 1068.17 49.9314 Q1068.17 58.1548 1071.54 62.8538 Q1074.94 67.5124 1080.85 67.5124 Q1086.77 67.5124 1090.13 62.8538 Q1093.53 58.1548 1093.53 49.9314 M1068.17 34.0924 Q1070.52 30.0415 1074.09 28.0971 Q1077.69 26.1121 1082.68 26.1121 Q1090.94 26.1121 1096.09 32.6746 Q1101.27 39.2371 1101.27 49.9314 Q1101.27 60.6258 1096.09 67.1883 Q1090.94 73.7508 1082.68 73.7508 Q1077.69 73.7508 1074.09 71.8063 Q1070.52 69.8214 1068.17 65.7705 L1068.17 72.576 L1060.68 72.576 L1060.68 9.54393 L1068.17 9.54393 L1068.17 34.0924 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1113.63 27.2059 L1121.08 27.2059 L1121.08 73.3862 Q1121.08 82.0551 1117.76 85.944 Q1114.48 89.8329 1107.14 89.8329 L1104.31 89.8329 L1104.31 83.5134 L1106.29 83.5134 Q1110.55 83.5134 1112.09 81.5285 Q1113.63 79.5841 1113.63 73.3862 L1113.63 27.2059 M1113.63 9.54393 L1121.08 9.54393 L1121.08 18.9825 L1113.63 18.9825 L1113.63 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1175.48 48.0275 L1175.48 51.6733 L1141.21 51.6733 Q1141.7 59.3701 1145.83 63.421 Q1150 67.4314 1157.42 67.4314 Q1161.71 67.4314 1165.72 66.3781 Q1169.77 65.3249 1173.74 63.2184 L1173.74 70.267 Q1169.73 71.9684 1165.52 72.8596 Q1161.31 73.7508 1156.97 73.7508 Q1146.11 73.7508 1139.75 67.4314 Q1133.44 61.1119 1133.44 50.3365 Q1133.44 39.1965 1139.43 32.6746 Q1145.47 26.1121 1155.67 26.1121 Q1164.83 26.1121 1170.14 32.0264 Q1175.48 37.9003 1175.48 48.0275 M1168.03 45.84 Q1167.95 39.7232 1164.59 36.0774 Q1161.26 32.4315 1155.76 32.4315 Q1149.52 32.4315 1145.75 35.9558 Q1142.02 39.4801 1141.46 45.8805 L1168.03 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1220.37 28.9478 L1220.37 35.9153 Q1217.21 34.1734 1214.01 33.3227 Q1210.85 32.4315 1207.61 32.4315 Q1200.36 32.4315 1196.35 37.0496 Q1192.34 41.6271 1192.34 49.9314 Q1192.34 58.2358 1196.35 62.8538 Q1200.36 67.4314 1207.61 67.4314 Q1210.85 67.4314 1214.01 66.5807 Q1217.21 65.6895 1220.37 63.9476 L1220.37 70.8341 Q1217.25 72.2924 1213.89 73.0216 Q1210.56 73.7508 1206.8 73.7508 Q1196.55 73.7508 1190.51 67.3098 Q1184.48 60.8689 1184.48 49.9314 Q1184.48 38.832 1190.55 32.472 Q1196.67 26.1121 1207.28 26.1121 Q1210.73 26.1121 1214.01 26.8413 Q1217.29 27.5299 1220.37 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1240.7 14.324 L1240.7 27.2059 L1256.06 27.2059 L1256.06 32.9987 L1240.7 32.9987 L1240.7 57.6282 Q1240.7 63.1779 1242.2 64.7578 Q1243.74 66.3376 1248.4 66.3376 L1256.06 66.3376 L1256.06 72.576 L1248.4 72.576 Q1239.77 72.576 1236.49 69.3758 Q1233.21 66.1351 1233.21 57.6282 L1233.21 32.9987 L1227.74 32.9987 L1227.74 27.2059 L1233.21 27.2059 L1233.21 14.324 L1240.7 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1265.86 27.2059 L1273.31 27.2059 L1273.31 72.576 L1265.86 72.576 L1265.86 27.2059 M1265.86 9.54393 L1273.31 9.54393 L1273.31 18.9825 L1265.86 18.9825 L1265.86 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1283.56 27.2059 L1291.46 27.2059 L1305.64 65.2844 L1319.82 27.2059 L1327.72 27.2059 L1310.7 72.576 L1300.58 72.576 L1283.56 27.2059 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1376.81 48.0275 L1376.81 51.6733 L1342.54 51.6733 Q1343.03 59.3701 1347.16 63.421 Q1351.33 67.4314 1358.75 67.4314 Q1363.04 67.4314 1367.05 66.3781 Q1371.1 65.3249 1375.07 63.2184 L1375.07 70.267 Q1371.06 71.9684 1366.85 72.8596 Q1362.64 73.7508 1358.3 73.7508 Q1347.44 73.7508 1341.08 67.4314 Q1334.76 61.1119 1334.76 50.3365 Q1334.76 39.1965 1340.76 32.6746 Q1346.8 26.1121 1357 26.1121 Q1366.16 26.1121 1371.47 32.0264 Q1376.81 37.9003 1376.81 48.0275 M1369.36 45.84 Q1369.28 39.7232 1365.92 36.0774 Q1362.59 32.4315 1357.09 32.4315 Q1350.85 32.4315 1347.08 35.9558 Q1343.35 39.4801 1342.79 45.8805 L1369.36 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1444.34 28.5427 L1444.34 35.5912 Q1441.18 33.9709 1437.78 33.1607 Q1434.38 32.3505 1430.73 32.3505 Q1425.18 32.3505 1422.39 34.0519 Q1419.63 35.7533 1419.63 39.156 Q1419.63 41.7486 1421.62 43.2475 Q1423.6 44.7058 1429.6 46.0426 L1432.15 46.6097 Q1440.09 48.3111 1443.41 51.4303 Q1446.77 54.509 1446.77 60.0587 Q1446.77 66.3781 1441.75 70.0644 Q1436.77 73.7508 1428.02 73.7508 Q1424.37 73.7508 1420.4 73.0216 Q1416.47 72.3329 1412.1 70.9151 L1412.1 63.2184 Q1416.23 65.3654 1420.24 66.4591 Q1424.25 67.5124 1428.18 67.5124 Q1433.44 67.5124 1436.28 65.73 Q1439.12 63.9071 1439.12 60.6258 Q1439.12 57.5877 1437.05 55.9673 Q1435.02 54.3469 1428.1 52.8481 L1425.51 52.2405 Q1418.58 50.7821 1415.5 47.7845 Q1412.42 44.7463 1412.42 39.4801 Q1412.42 33.0797 1416.96 29.5959 Q1421.49 26.1121 1429.84 26.1121 Q1433.97 26.1121 1437.62 26.7198 Q1441.26 27.3274 1444.34 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1465.85 65.7705 L1465.85 89.8329 L1458.36 89.8329 L1458.36 27.2059 L1465.85 27.2059 L1465.85 34.0924 Q1468.2 30.0415 1471.77 28.0971 Q1475.37 26.1121 1480.35 26.1121 Q1488.62 26.1121 1493.76 32.6746 Q1498.95 39.2371 1498.95 49.9314 Q1498.95 60.6258 1493.76 67.1883 Q1488.62 73.7508 1480.35 73.7508 Q1475.37 73.7508 1471.77 71.8063 Q1468.2 69.8214 1465.85 65.7705 M1491.21 49.9314 Q1491.21 41.7081 1487.81 37.0496 Q1484.45 32.3505 1478.53 32.3505 Q1472.62 32.3505 1469.21 37.0496 Q1465.85 41.7081 1465.85 49.9314 Q1465.85 58.1548 1469.21 62.8538 Q1472.62 67.5124 1478.53 67.5124 Q1484.45 67.5124 1487.81 62.8538 Q1491.21 58.1548 1491.21 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1531.92 49.7694 Q1522.89 49.7694 1519.41 51.8354 Q1515.92 53.9013 1515.92 58.8839 Q1515.92 62.8538 1518.51 65.2034 Q1521.15 67.5124 1525.64 67.5124 Q1531.84 67.5124 1535.57 63.1374 Q1539.34 58.7219 1539.34 51.4303 L1539.34 49.7694 L1531.92 49.7694 M1546.79 46.6907 L1546.79 72.576 L1539.34 72.576 L1539.34 65.6895 Q1536.78 69.8214 1532.98 71.8063 Q1529.17 73.7508 1523.66 73.7508 Q1516.69 73.7508 1512.56 69.8619 Q1508.47 65.9325 1508.47 59.3701 Q1508.47 51.7138 1513.57 47.825 Q1518.72 43.9361 1528.88 43.9361 L1539.34 43.9361 L1539.34 43.2069 Q1539.34 38.0623 1535.93 35.2672 Q1532.57 32.4315 1526.45 32.4315 Q1522.56 32.4315 1518.88 33.3632 Q1515.19 34.295 1511.79 36.1584 L1511.79 29.2718 Q1515.88 27.692 1519.73 26.9223 Q1523.58 26.1121 1527.22 26.1121 Q1537.07 26.1121 1541.93 31.2163 Q1546.79 36.3204 1546.79 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1594.79 28.9478 L1594.79 35.9153 Q1591.63 34.1734 1588.43 33.3227 Q1585.27 32.4315 1582.03 32.4315 Q1574.78 32.4315 1570.77 37.0496 Q1566.76 41.6271 1566.76 49.9314 Q1566.76 58.2358 1570.77 62.8538 Q1574.78 67.4314 1582.03 67.4314 Q1585.27 67.4314 1588.43 66.5807 Q1591.63 65.6895 1594.79 63.9476 L1594.79 70.8341 Q1591.67 72.2924 1588.31 73.0216 Q1584.99 73.7508 1581.22 73.7508 Q1570.97 73.7508 1564.94 67.3098 Q1558.9 60.8689 1558.9 49.9314 Q1558.9 38.832 1564.98 32.472 Q1571.09 26.1121 1581.71 26.1121 Q1585.15 26.1121 1588.43 26.8413 Q1591.71 27.5299 1594.79 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1646.56 48.0275 L1646.56 51.6733 L1612.29 51.6733 Q1612.78 59.3701 1616.91 63.421 Q1621.08 67.4314 1628.5 67.4314 Q1632.79 67.4314 1636.8 66.3781 Q1640.85 65.3249 1644.82 63.2184 L1644.82 70.267 Q1640.81 71.9684 1636.6 72.8596 Q1632.38 73.7508 1628.05 73.7508 Q1617.19 73.7508 1610.83 67.4314 Q1604.51 61.1119 1604.51 50.3365 Q1604.51 39.1965 1610.51 32.6746 Q1616.55 26.1121 1626.75 26.1121 Q1635.91 26.1121 1641.22 32.0264 Q1646.56 37.9003 1646.56 48.0275 M1639.11 45.84 Q1639.03 39.7232 1635.67 36.0774 Q1632.34 32.4315 1626.84 32.4315 Q1620.6 32.4315 1616.83 35.9558 Q1613.1 39.4801 1612.54 45.8805 L1639.11 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip512)" style="stroke:#ff0000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="-1780.13,422.883 4419.2,422.883 "/>
+<polyline clip-path="url(#clip512)" style="stroke:#ff0000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="987.246,1776.14 987.246,-657.503 "/>
+<circle clip-path="url(#clip512)" cx="344.796" cy="941.969" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="384.581" cy="812.114" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="424.366" cy="758.327" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="464.151" cy="717.054" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="503.936" cy="682.259" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="543.722" cy="651.605" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="583.507" cy="623.891" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="623.292" cy="598.406" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="663.077" cy="574.684" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="702.863" cy="552.405" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="742.648" cy="531.332" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="782.433" cy="511.289" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="822.218" cy="492.139" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="862.003" cy="473.771" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="901.789" cy="456.097" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="941.574" cy="439.599" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="981.359" cy="424.93" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1021.14" cy="411.608" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1060.93" cy="399.319" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1100.71" cy="387.854" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1140.5" cy="377.067" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1180.29" cy="366.85" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1220.07" cy="357.121" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1259.86" cy="347.817" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1299.64" cy="338.886" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1339.43" cy="330.286" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1379.21" cy="321.984" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1419" cy="313.95" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1458.78" cy="306.16" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1498.57" cy="298.592" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1538.35" cy="291.23" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1578.14" cy="284.057" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1617.92" cy="277.059" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1657.71" cy="270.224" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1697.49" cy="263.541" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1737.28" cy="257" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1777.06" cy="250.593" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1816.85" cy="244.312" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1856.63" cy="238.149" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1896.42" cy="232.098" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1936.2" cy="226.154" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="1975.99" cy="220.311" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2015.78" cy="214.563" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2055.56" cy="208.907" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2095.35" cy="203.338" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2135.13" cy="197.852" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2174.92" cy="192.445" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2214.7" cy="187.115" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2254.49" cy="181.858" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<circle clip-path="url(#clip512)" cx="2294.27" cy="176.671" r="18" fill="#ffffff" fill-rule="evenodd" fill-opacity="1" stroke="#000000" stroke-opacity="1" stroke-width="4.8"/>
+<path clip-path="url(#clip510)" d="M1501.38 937.887 L2283.87 937.887 L2283.87 834.207 L1501.38 834.207  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1501.38,937.887 2283.87,937.887 2283.87,834.207 1501.38,834.207 1501.38,937.887 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#ff0000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1524.34,886.047 1662.1,886.047 "/>
+<path clip-path="url(#clip510)" d="M1707.31 869.901 L1707.31 874.461 Q1704.64 873.188 1702.28 872.563 Q1699.92 871.938 1697.72 871.938 Q1693.9 871.938 1691.82 873.42 Q1689.76 874.901 1689.76 877.633 Q1689.76 879.924 1691.13 881.105 Q1692.51 882.262 1696.36 882.98 L1699.18 883.559 Q1704.41 884.554 1706.89 887.077 Q1709.39 889.577 1709.39 893.79 Q1709.39 898.813 1706.01 901.406 Q1702.65 903.998 1696.15 903.998 Q1693.69 903.998 1690.92 903.443 Q1688.16 902.887 1685.2 901.799 L1685.2 896.984 Q1688.05 898.582 1690.78 899.392 Q1693.51 900.202 1696.15 900.202 Q1700.15 900.202 1702.33 898.628 Q1704.5 897.054 1704.5 894.137 Q1704.5 891.591 1702.93 890.156 Q1701.38 888.721 1697.82 888.003 L1694.97 887.447 Q1689.74 886.406 1687.4 884.184 Q1685.06 881.961 1685.06 878.003 Q1685.06 873.42 1688.28 870.781 Q1691.52 868.142 1697.19 868.142 Q1699.62 868.142 1702.14 868.582 Q1704.67 869.022 1707.31 869.901 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1716.5 877.401 L1720.75 877.401 L1720.75 903.327 L1716.5 903.327 L1716.5 877.401 M1716.5 867.309 L1720.75 867.309 L1720.75 872.702 L1716.5 872.702 L1716.5 867.309 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1751.22 887.679 L1751.22 903.327 L1746.96 903.327 L1746.96 887.818 Q1746.96 884.137 1745.52 882.309 Q1744.09 880.48 1741.22 880.48 Q1737.77 880.48 1735.78 882.679 Q1733.79 884.878 1733.79 888.674 L1733.79 903.327 L1729.5 903.327 L1729.5 877.401 L1733.79 877.401 L1733.79 881.429 Q1735.31 879.091 1737.38 877.934 Q1739.46 876.776 1742.17 876.776 Q1746.63 876.776 1748.93 879.554 Q1751.22 882.309 1751.22 887.679 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1776.77 890.063 Q1776.77 885.434 1774.85 882.887 Q1772.95 880.341 1769.5 880.341 Q1766.08 880.341 1764.16 882.887 Q1762.26 885.434 1762.26 890.063 Q1762.26 894.67 1764.16 897.216 Q1766.08 899.762 1769.5 899.762 Q1772.95 899.762 1774.85 897.216 Q1776.77 894.67 1776.77 890.063 M1781.03 900.109 Q1781.03 906.73 1778.09 909.947 Q1775.15 913.188 1769.09 913.188 Q1766.84 913.188 1764.85 912.841 Q1762.86 912.517 1760.99 911.822 L1760.99 907.679 Q1762.86 908.697 1764.69 909.183 Q1766.52 909.669 1768.42 909.669 Q1772.61 909.669 1774.69 907.47 Q1776.77 905.295 1776.77 900.873 L1776.77 898.767 Q1775.45 901.058 1773.39 902.193 Q1771.33 903.327 1768.46 903.327 Q1763.69 903.327 1760.78 899.693 Q1757.86 896.058 1757.86 890.063 Q1757.86 884.045 1760.78 880.41 Q1763.69 876.776 1768.46 876.776 Q1771.33 876.776 1773.39 877.91 Q1775.45 879.045 1776.77 881.336 L1776.77 877.401 L1781.03 877.401 L1781.03 900.109 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1789.81 867.309 L1794.06 867.309 L1794.06 903.327 L1789.81 903.327 L1789.81 867.309 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1825.15 889.299 L1825.15 891.383 L1805.57 891.383 Q1805.85 895.781 1808.21 898.095 Q1810.59 900.387 1814.83 900.387 Q1817.28 900.387 1819.57 899.785 Q1821.89 899.183 1824.16 897.98 L1824.16 902.007 Q1821.87 902.98 1819.46 903.489 Q1817.05 903.998 1814.57 903.998 Q1808.37 903.998 1804.74 900.387 Q1801.12 896.776 1801.12 890.619 Q1801.12 884.253 1804.55 880.526 Q1808 876.776 1813.83 876.776 Q1819.06 876.776 1822.1 880.156 Q1825.15 883.512 1825.15 889.299 M1820.89 888.049 Q1820.85 884.554 1818.93 882.471 Q1817.03 880.387 1813.88 880.387 Q1810.31 880.387 1808.16 882.401 Q1806.03 884.415 1805.71 888.072 L1820.89 888.049 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1829.99 888.443 L1842.47 888.443 L1842.47 892.239 L1829.99 892.239 L1829.99 888.443 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1860.18 880.387 Q1856.75 880.387 1854.76 883.072 Q1852.77 885.734 1852.77 890.387 Q1852.77 895.04 1854.74 897.725 Q1856.73 900.387 1860.18 900.387 Q1863.58 900.387 1865.57 897.702 Q1867.56 895.017 1867.56 890.387 Q1867.56 885.781 1865.57 883.096 Q1863.58 880.387 1860.18 880.387 M1860.18 876.776 Q1865.73 876.776 1868.9 880.387 Q1872.07 883.998 1872.07 890.387 Q1872.07 896.753 1868.9 900.387 Q1865.73 903.998 1860.18 903.998 Q1854.6 903.998 1851.43 900.387 Q1848.28 896.753 1848.28 890.387 Q1848.28 883.998 1851.43 880.387 Q1854.6 876.776 1860.18 876.776 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1897.74 890.387 Q1897.74 885.688 1895.8 883.026 Q1893.88 880.341 1890.5 880.341 Q1887.12 880.341 1885.18 883.026 Q1883.25 885.688 1883.25 890.387 Q1883.25 895.086 1885.18 897.771 Q1887.12 900.433 1890.5 900.433 Q1893.88 900.433 1895.8 897.771 Q1897.74 895.086 1897.74 890.387 M1883.25 881.336 Q1884.6 879.022 1886.63 877.91 Q1888.69 876.776 1891.54 876.776 Q1896.26 876.776 1899.2 880.526 Q1902.17 884.276 1902.17 890.387 Q1902.17 896.498 1899.2 900.248 Q1896.26 903.998 1891.54 903.998 Q1888.69 903.998 1886.63 902.887 Q1884.6 901.753 1883.25 899.438 L1883.25 903.327 L1878.97 903.327 L1878.97 867.309 L1883.25 867.309 L1883.25 881.336 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1909.23 877.401 L1913.49 877.401 L1913.49 903.79 Q1913.49 908.744 1911.59 910.966 Q1909.71 913.188 1905.52 913.188 L1903.9 913.188 L1903.9 909.577 L1905.04 909.577 Q1907.47 909.577 1908.35 908.443 Q1909.23 907.332 1909.23 903.79 L1909.23 877.401 M1909.23 867.309 L1913.49 867.309 L1913.49 872.702 L1909.23 872.702 L1909.23 867.309 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1944.57 889.299 L1944.57 891.383 L1924.99 891.383 Q1925.27 895.781 1927.63 898.095 Q1930.01 900.387 1934.25 900.387 Q1936.7 900.387 1938.99 899.785 Q1941.31 899.183 1943.58 897.98 L1943.58 902.007 Q1941.29 902.98 1938.88 903.489 Q1936.47 903.998 1933.99 903.998 Q1927.79 903.998 1924.16 900.387 Q1920.55 896.776 1920.55 890.619 Q1920.55 884.253 1923.97 880.526 Q1927.42 876.776 1933.25 876.776 Q1938.49 876.776 1941.52 880.156 Q1944.57 883.512 1944.57 889.299 M1940.31 888.049 Q1940.27 884.554 1938.35 882.471 Q1936.45 880.387 1933.3 880.387 Q1929.74 880.387 1927.58 882.401 Q1925.45 884.415 1925.13 888.072 L1940.31 888.049 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1970.22 878.397 L1970.22 882.378 Q1968.42 881.383 1966.59 880.896 Q1964.78 880.387 1962.93 880.387 Q1958.79 880.387 1956.49 883.026 Q1954.2 885.642 1954.2 890.387 Q1954.2 895.133 1956.49 897.771 Q1958.79 900.387 1962.93 900.387 Q1964.78 900.387 1966.59 899.901 Q1968.42 899.392 1970.22 898.396 L1970.22 902.332 Q1968.44 903.165 1966.52 903.582 Q1964.62 903.998 1962.47 903.998 Q1956.61 903.998 1953.16 900.318 Q1949.71 896.637 1949.71 890.387 Q1949.71 884.045 1953.18 880.41 Q1956.68 876.776 1962.74 876.776 Q1964.71 876.776 1966.59 877.193 Q1968.46 877.586 1970.22 878.397 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1981.84 870.04 L1981.84 877.401 L1990.61 877.401 L1990.61 880.711 L1981.84 880.711 L1981.84 894.785 Q1981.84 897.957 1982.7 898.859 Q1983.58 899.762 1986.24 899.762 L1990.61 899.762 L1990.61 903.327 L1986.24 903.327 Q1981.31 903.327 1979.43 901.498 Q1977.56 899.646 1977.56 894.785 L1977.56 880.711 L1974.43 880.711 L1974.43 877.401 L1977.56 877.401 L1977.56 870.04 L1981.84 870.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1996.22 877.401 L2000.48 877.401 L2000.48 903.327 L1996.22 903.327 L1996.22 877.401 M1996.22 867.309 L2000.48 867.309 L2000.48 872.702 L1996.22 872.702 L1996.22 867.309 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2006.33 877.401 L2010.85 877.401 L2018.95 899.16 L2027.05 877.401 L2031.56 877.401 L2021.84 903.327 L2016.05 903.327 L2006.33 877.401 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2059.62 889.299 L2059.62 891.383 L2040.04 891.383 Q2040.31 895.781 2042.67 898.095 Q2045.06 900.387 2049.29 900.387 Q2051.75 900.387 2054.04 899.785 Q2056.35 899.183 2058.62 897.98 L2058.62 902.007 Q2056.33 902.98 2053.92 903.489 Q2051.52 903.998 2049.04 903.998 Q2042.84 903.998 2039.2 900.387 Q2035.59 896.776 2035.59 890.619 Q2035.59 884.253 2039.02 880.526 Q2042.47 876.776 2048.3 876.776 Q2053.53 876.776 2056.56 880.156 Q2059.62 883.512 2059.62 889.299 M2055.36 888.049 Q2055.31 884.554 2053.39 882.471 Q2051.49 880.387 2048.35 880.387 Q2044.78 880.387 2042.63 882.401 Q2040.5 884.415 2040.17 888.072 L2055.36 888.049 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2098.21 878.165 L2098.21 882.193 Q2096.4 881.267 2094.46 880.804 Q2092.51 880.341 2090.43 880.341 Q2087.26 880.341 2085.66 881.313 Q2084.09 882.285 2084.09 884.23 Q2084.09 885.711 2085.22 886.568 Q2086.35 887.401 2089.78 888.165 L2091.24 888.489 Q2095.78 889.461 2097.67 891.244 Q2099.6 893.003 2099.6 896.174 Q2099.6 899.785 2096.72 901.892 Q2093.88 903.998 2088.88 903.998 Q2086.79 903.998 2084.53 903.582 Q2082.28 903.188 2079.78 902.378 L2079.78 897.98 Q2082.14 899.207 2084.43 899.832 Q2086.73 900.433 2088.97 900.433 Q2091.98 900.433 2093.6 899.415 Q2095.22 898.373 2095.22 896.498 Q2095.22 894.762 2094.04 893.836 Q2092.88 892.91 2088.92 892.054 L2087.44 891.707 Q2083.48 890.873 2081.73 889.16 Q2079.97 887.424 2079.97 884.415 Q2079.97 880.758 2082.56 878.767 Q2085.15 876.776 2089.92 876.776 Q2092.28 876.776 2094.36 877.123 Q2096.45 877.471 2098.21 878.165 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2116.42 880.387 Q2113 880.387 2111.01 883.072 Q2109.02 885.734 2109.02 890.387 Q2109.02 895.04 2110.98 897.725 Q2112.97 900.387 2116.42 900.387 Q2119.83 900.387 2121.82 897.702 Q2123.81 895.017 2123.81 890.387 Q2123.81 885.781 2121.82 883.096 Q2119.83 880.387 2116.42 880.387 M2116.42 876.776 Q2121.98 876.776 2125.15 880.387 Q2128.32 883.998 2128.32 890.387 Q2128.32 896.753 2125.15 900.387 Q2121.98 903.998 2116.42 903.998 Q2110.85 903.998 2107.67 900.387 Q2104.53 896.753 2104.53 890.387 Q2104.53 883.998 2107.67 880.387 Q2110.85 876.776 2116.42 876.776 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2135.38 867.309 L2139.64 867.309 L2139.64 903.327 L2135.38 903.327 L2135.38 867.309 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2148.11 893.096 L2148.11 877.401 L2152.37 877.401 L2152.37 892.933 Q2152.37 896.614 2153.81 898.466 Q2155.24 900.295 2158.11 900.295 Q2161.56 900.295 2163.55 898.095 Q2165.57 895.896 2165.57 892.1 L2165.57 877.401 L2169.83 877.401 L2169.83 903.327 L2165.57 903.327 L2165.57 899.345 Q2164.02 901.707 2161.96 902.864 Q2159.92 903.998 2157.21 903.998 Q2152.74 903.998 2150.43 901.22 Q2148.11 898.443 2148.11 893.096 M2158.83 876.776 L2158.83 876.776 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2182.81 870.04 L2182.81 877.401 L2191.59 877.401 L2191.59 880.711 L2182.81 880.711 L2182.81 894.785 Q2182.81 897.957 2183.67 898.859 Q2184.55 899.762 2187.21 899.762 L2191.59 899.762 L2191.59 903.327 L2187.21 903.327 Q2182.28 903.327 2180.41 901.498 Q2178.53 899.646 2178.53 894.785 L2178.53 880.711 L2175.41 880.711 L2175.41 877.401 L2178.53 877.401 L2178.53 870.04 L2182.81 870.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2197.19 877.401 L2201.45 877.401 L2201.45 903.327 L2197.19 903.327 L2197.19 877.401 M2197.19 867.309 L2201.45 867.309 L2201.45 872.702 L2197.19 872.702 L2197.19 867.309 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2220.4 880.387 Q2216.98 880.387 2214.99 883.072 Q2213 885.734 2213 890.387 Q2213 895.04 2214.97 897.725 Q2216.96 900.387 2220.4 900.387 Q2223.81 900.387 2225.8 897.702 Q2227.79 895.017 2227.79 890.387 Q2227.79 885.781 2225.8 883.096 Q2223.81 880.387 2220.4 880.387 M2220.4 876.776 Q2225.96 876.776 2229.13 880.387 Q2232.3 883.998 2232.3 890.387 Q2232.3 896.753 2229.13 900.387 Q2225.96 903.998 2220.4 903.998 Q2214.83 903.998 2211.65 900.387 Q2208.51 896.753 2208.51 890.387 Q2208.51 883.998 2211.65 880.387 Q2214.83 876.776 2220.4 876.776 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2260.91 887.679 L2260.91 903.327 L2256.65 903.327 L2256.65 887.818 Q2256.65 884.137 2255.22 882.309 Q2253.78 880.48 2250.91 880.48 Q2247.46 880.48 2245.47 882.679 Q2243.48 884.878 2243.48 888.674 L2243.48 903.327 L2239.2 903.327 L2239.2 877.401 L2243.48 877.401 L2243.48 881.429 Q2245.01 879.091 2247.07 877.934 Q2249.15 876.776 2251.86 876.776 Q2256.33 876.776 2258.62 879.554 Q2260.91 882.309 2260.91 887.679 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M341.189 946.973 L344.053 946.973 L344.053 937.086 L340.937 937.711 L340.937 936.114 L344.036 935.489 L345.789 935.489 L345.789 946.973 L348.654 946.973 L348.654 948.449 L341.189 948.449 L341.189 946.973 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M382.576 817.119 L388.695 817.119 L388.695 818.594 L380.466 818.594 L380.466 817.119 Q381.464 816.086 383.183 814.349 Q384.911 812.605 385.353 812.101 Q386.195 811.155 386.525 810.504 Q386.864 809.844 386.864 809.211 Q386.864 808.178 386.135 807.527 Q385.414 806.876 384.251 806.876 Q383.426 806.876 382.506 807.162 Q381.595 807.448 380.553 808.03 L380.553 806.259 Q381.612 805.834 382.532 805.617 Q383.452 805.4 384.216 805.4 Q386.23 805.4 387.428 806.407 Q388.626 807.414 388.626 809.098 Q388.626 809.896 388.322 810.617 Q388.027 811.329 387.237 812.301 Q387.02 812.553 385.857 813.759 Q384.694 814.957 382.576 817.119 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M425.959 757.819 Q427.218 758.088 427.921 758.939 Q428.632 759.789 428.632 761.039 Q428.632 762.958 427.313 764.008 Q425.994 765.058 423.563 765.058 Q422.747 765.058 421.879 764.893 Q421.02 764.737 420.099 764.416 L420.099 762.723 Q420.829 763.149 421.697 763.366 Q422.565 763.583 423.511 763.583 Q425.16 763.583 426.02 762.932 Q426.888 762.281 426.888 761.039 Q426.888 759.893 426.08 759.251 Q425.282 758.6 423.849 758.6 L422.339 758.6 L422.339 757.159 L423.919 757.159 Q425.212 757.159 425.898 756.647 Q426.584 756.126 426.584 755.154 Q426.584 754.156 425.872 753.626 Q425.169 753.088 423.849 753.088 Q423.129 753.088 422.304 753.244 Q421.48 753.4 420.49 753.73 L420.49 752.168 Q421.488 751.89 422.356 751.751 Q423.233 751.612 424.006 751.612 Q426.002 751.612 427.165 752.524 Q428.329 753.426 428.329 754.972 Q428.329 756.048 427.712 756.795 Q427.096 757.532 425.959 757.819 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M465.28 712.102 L460.853 719.02 L465.28 719.02 L465.28 712.102 M464.82 710.574 L467.024 710.574 L467.024 719.02 L468.873 719.02 L468.873 720.478 L467.024 720.478 L467.024 723.534 L465.28 723.534 L465.28 720.478 L459.429 720.478 L459.429 718.786 L464.82 710.574 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M500.291 675.779 L507.174 675.779 L507.174 677.255 L501.896 677.255 L501.896 680.432 Q502.278 680.302 502.66 680.241 Q503.042 680.172 503.424 680.172 Q505.594 680.172 506.862 681.361 Q508.129 682.55 508.129 684.581 Q508.129 686.673 506.827 687.837 Q505.525 688.991 503.155 688.991 Q502.339 688.991 501.489 688.852 Q500.647 688.713 499.744 688.436 L499.744 686.673 Q500.525 687.099 501.358 687.307 Q502.192 687.515 503.12 687.515 Q504.622 687.515 505.499 686.726 Q506.376 685.936 506.376 684.581 Q506.376 683.227 505.499 682.437 Q504.622 681.647 503.12 681.647 Q502.417 681.647 501.714 681.804 Q501.02 681.96 500.291 682.29 L500.291 675.779 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M543.874 650.906 Q542.693 650.906 541.999 651.713 Q541.313 652.521 541.313 653.927 Q541.313 655.324 541.999 656.14 Q542.693 656.948 543.874 656.948 Q545.054 656.948 545.74 656.14 Q546.434 655.324 546.434 653.927 Q546.434 652.521 545.74 651.713 Q545.054 650.906 543.874 650.906 M547.354 645.411 L547.354 647.009 Q546.695 646.696 546.018 646.531 Q545.349 646.366 544.69 646.366 Q542.953 646.366 542.033 647.538 Q541.122 648.71 540.992 651.08 Q541.504 650.324 542.276 649.925 Q543.049 649.517 543.978 649.517 Q545.931 649.517 547.059 650.706 Q548.196 651.887 548.196 653.927 Q548.196 655.923 547.016 657.13 Q545.835 658.337 543.874 658.337 Q541.625 658.337 540.436 656.618 Q539.247 654.89 539.247 651.618 Q539.247 648.545 540.705 646.722 Q542.163 644.89 544.62 644.89 Q545.28 644.89 545.948 645.021 Q546.625 645.151 547.354 645.411 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M579.34 617.411 L587.674 617.411 L587.674 618.158 L582.969 630.371 L581.137 630.371 L585.564 618.887 L579.34 618.887 L579.34 617.411 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M623.292 598.731 Q622.042 598.731 621.322 599.399 Q620.61 600.068 620.61 601.24 Q620.61 602.412 621.322 603.08 Q622.042 603.748 623.292 603.748 Q624.542 603.748 625.263 603.08 Q625.983 602.403 625.983 601.24 Q625.983 600.068 625.263 599.399 Q624.551 598.731 623.292 598.731 M621.539 597.985 Q620.41 597.707 619.776 596.934 Q619.151 596.162 619.151 595.051 Q619.151 593.497 620.254 592.594 Q621.365 591.691 623.292 591.691 Q625.228 591.691 626.33 592.594 Q627.433 593.497 627.433 595.051 Q627.433 596.162 626.799 596.934 Q626.174 597.707 625.054 597.985 Q626.322 598.28 627.025 599.139 Q627.737 599.998 627.737 601.24 Q627.737 603.123 626.582 604.13 Q625.436 605.137 623.292 605.137 Q621.148 605.137 619.994 604.13 Q618.848 603.123 618.848 601.24 Q618.848 599.998 619.559 599.139 Q620.271 598.28 621.539 597.985 M620.896 595.215 Q620.896 596.222 621.521 596.787 Q622.155 597.351 623.292 597.351 Q624.421 597.351 625.054 596.787 Q625.697 596.222 625.697 595.215 Q625.697 594.208 625.054 593.644 Q624.421 593.08 623.292 593.08 Q622.155 593.08 621.521 593.644 Q620.896 594.208 620.896 595.215 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M659.44 580.895 L659.44 579.298 Q660.1 579.61 660.777 579.775 Q661.454 579.94 662.105 579.94 Q663.841 579.94 664.753 578.777 Q665.673 577.605 665.803 575.227 Q665.3 575.973 664.527 576.373 Q663.754 576.772 662.817 576.772 Q660.872 576.772 659.735 575.6 Q658.607 574.419 658.607 572.38 Q658.607 570.383 659.787 569.176 Q660.968 567.97 662.93 567.97 Q665.178 567.97 666.359 569.697 Q667.548 571.416 667.548 574.697 Q667.548 577.761 666.089 579.593 Q664.64 581.416 662.183 581.416 Q661.524 581.416 660.846 581.286 Q660.169 581.156 659.44 580.895 M662.93 575.4 Q664.11 575.4 664.796 574.593 Q665.491 573.786 665.491 572.38 Q665.491 570.982 664.796 570.175 Q664.11 569.359 662.93 569.359 Q661.749 569.359 661.055 570.175 Q660.369 570.982 660.369 572.38 Q660.369 573.786 661.055 574.593 Q661.749 575.4 662.93 575.4 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M693.37 557.409 L696.235 557.409 L696.235 547.522 L693.119 548.147 L693.119 546.55 L696.218 545.925 L697.971 545.925 L697.971 557.409 L700.836 557.409 L700.836 558.885 L693.37 558.885 L693.37 557.409 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M708.127 547.079 Q706.773 547.079 706.087 548.416 Q705.41 549.744 705.41 552.418 Q705.41 555.083 706.087 556.419 Q706.773 557.747 708.127 557.747 Q709.49 557.747 710.167 556.419 Q710.853 555.083 710.853 552.418 Q710.853 549.744 710.167 548.416 Q709.49 547.079 708.127 547.079 M708.127 545.69 Q710.306 545.69 711.452 547.418 Q712.606 549.136 712.606 552.418 Q712.606 555.69 711.452 557.418 Q710.306 559.136 708.127 559.136 Q705.948 559.136 704.794 557.418 Q703.648 555.69 703.648 552.418 Q703.648 549.136 704.794 547.418 Q705.948 545.69 708.127 545.69 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M733.386 536.336 L736.25 536.336 L736.25 526.449 L733.134 527.074 L733.134 525.477 L736.233 524.852 L737.986 524.852 L737.986 536.336 L740.851 536.336 L740.851 537.812 L733.386 537.812 L733.386 536.336 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M744.696 536.336 L747.561 536.336 L747.561 526.449 L744.445 527.074 L744.445 525.477 L747.544 524.852 L749.297 524.852 L749.297 536.336 L752.162 536.336 L752.162 537.812 L744.696 537.812 L744.696 536.336 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M773.24 516.294 L776.105 516.294 L776.105 506.407 L772.989 507.032 L772.989 505.434 L776.088 504.809 L777.841 504.809 L777.841 516.294 L780.706 516.294 L780.706 517.769 L773.24 517.769 L773.24 516.294 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M785.758 516.294 L791.877 516.294 L791.877 517.769 L783.648 517.769 L783.648 516.294 Q784.647 515.261 786.365 513.525 Q788.093 511.78 788.535 511.276 Q789.377 510.33 789.707 509.679 Q790.046 509.019 790.046 508.386 Q790.046 507.353 789.317 506.702 Q788.596 506.051 787.433 506.051 Q786.608 506.051 785.688 506.337 Q784.777 506.624 783.735 507.205 L783.735 505.434 Q784.794 505.009 785.714 504.792 Q786.634 504.575 787.398 504.575 Q789.412 504.575 790.61 505.582 Q791.808 506.589 791.808 508.273 Q791.808 509.071 791.504 509.792 Q791.209 510.504 790.419 511.476 Q790.202 511.728 789.039 512.934 Q787.876 514.132 785.758 516.294 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M812.848 497.143 L815.712 497.143 L815.712 487.256 L812.596 487.881 L812.596 486.284 L815.695 485.659 L817.448 485.659 L817.448 497.143 L820.313 497.143 L820.313 498.619 L812.848 498.619 L812.848 497.143 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M829.167 491.631 Q830.426 491.9 831.129 492.751 Q831.841 493.601 831.841 494.851 Q831.841 496.77 830.521 497.82 Q829.202 498.871 826.771 498.871 Q825.955 498.871 825.087 498.706 Q824.228 498.549 823.308 498.228 L823.308 496.535 Q824.037 496.961 824.905 497.178 Q825.773 497.395 826.719 497.395 Q828.368 497.395 829.228 496.744 Q830.096 496.093 830.096 494.851 Q830.096 493.706 829.289 493.063 Q828.49 492.412 827.058 492.412 L825.547 492.412 L825.547 490.971 L827.127 490.971 Q828.42 490.971 829.106 490.459 Q829.792 489.938 829.792 488.966 Q829.792 487.968 829.08 487.438 Q828.377 486.9 827.058 486.9 Q826.337 486.9 825.512 487.056 Q824.688 487.213 823.698 487.542 L823.698 485.98 Q824.697 485.702 825.565 485.563 Q826.441 485.424 827.214 485.424 Q829.21 485.424 830.374 486.336 Q831.537 487.239 831.537 488.784 Q831.537 489.86 830.92 490.607 Q830.304 491.345 829.167 491.631 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M852.42 478.775 L855.285 478.775 L855.285 468.888 L852.168 469.513 L852.168 467.916 L855.267 467.291 L857.021 467.291 L857.021 478.775 L859.885 478.775 L859.885 480.251 L852.42 480.251 L852.42 478.775 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M868.245 468.819 L863.818 475.737 L868.245 475.737 L868.245 468.819 M867.785 467.291 L869.99 467.291 L869.99 475.737 L871.838 475.737 L871.838 477.195 L869.99 477.195 L869.99 480.251 L868.245 480.251 L868.245 477.195 L862.394 477.195 L862.394 475.503 L867.785 467.291 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M892.483 461.101 L895.348 461.101 L895.348 451.214 L892.231 451.839 L892.231 450.242 L895.33 449.617 L897.084 449.617 L897.084 461.101 L899.948 461.101 L899.948 462.577 L892.483 462.577 L892.483 461.101 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M903.507 449.617 L910.391 449.617 L910.391 451.093 L905.113 451.093 L905.113 454.27 Q905.495 454.139 905.877 454.079 Q906.259 454.009 906.641 454.009 Q908.811 454.009 910.079 455.198 Q911.346 456.388 911.346 458.419 Q911.346 460.511 910.044 461.674 Q908.742 462.829 906.372 462.829 Q905.556 462.829 904.705 462.69 Q903.863 462.551 902.961 462.273 L902.961 460.511 Q903.742 460.936 904.575 461.145 Q905.408 461.353 906.337 461.353 Q907.839 461.353 908.716 460.563 Q909.592 459.773 909.592 458.419 Q909.592 457.065 908.716 456.275 Q907.839 455.485 906.337 455.485 Q905.634 455.485 904.931 455.641 Q904.237 455.797 903.507 456.127 L903.507 449.617 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M932.051 444.603 L934.916 444.603 L934.916 434.716 L931.8 435.341 L931.8 433.744 L934.899 433.119 L936.652 433.119 L936.652 444.603 L939.517 444.603 L939.517 446.079 L932.051 446.079 L932.051 444.603 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M947.025 438.9 Q945.845 438.9 945.15 439.707 Q944.465 440.514 944.465 441.921 Q944.465 443.318 945.15 444.134 Q945.845 444.941 947.025 444.941 Q948.206 444.941 948.892 444.134 Q949.586 443.318 949.586 441.921 Q949.586 440.514 948.892 439.707 Q948.206 438.9 947.025 438.9 M950.506 433.405 L950.506 435.002 Q949.846 434.69 949.169 434.525 Q948.501 434.36 947.841 434.36 Q946.105 434.36 945.185 435.532 Q944.274 436.704 944.143 439.073 Q944.655 438.318 945.428 437.919 Q946.201 437.511 947.129 437.511 Q949.083 437.511 950.211 438.7 Q951.348 439.881 951.348 441.921 Q951.348 443.917 950.168 445.124 Q948.987 446.33 947.025 446.33 Q944.777 446.33 943.588 444.612 Q942.399 442.884 942.399 439.612 Q942.399 436.539 943.857 434.716 Q945.315 432.884 947.772 432.884 Q948.432 432.884 949.1 433.014 Q949.777 433.145 950.506 433.405 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M972.036 429.934 L974.901 429.934 L974.901 420.047 L971.785 420.672 L971.785 419.075 L974.883 418.45 L976.637 418.45 L976.637 429.934 L979.502 429.934 L979.502 431.41 L972.036 431.41 L972.036 429.934 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M982.6 418.45 L990.934 418.45 L990.934 419.196 L986.229 431.41 L984.397 431.41 L988.824 419.925 L982.6 419.925 L982.6 418.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1011.67 416.612 L1014.53 416.612 L1014.53 406.725 L1011.42 407.35 L1011.42 405.753 L1014.52 405.128 L1016.27 405.128 L1016.27 416.612 L1019.13 416.612 L1019.13 418.088 L1011.67 418.088 L1011.67 416.612 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1026.43 411.934 Q1025.18 411.934 1024.46 412.602 Q1023.74 413.27 1023.74 414.442 Q1023.74 415.614 1024.46 416.283 Q1025.18 416.951 1026.43 416.951 Q1027.68 416.951 1028.4 416.283 Q1029.12 415.606 1029.12 414.442 Q1029.12 413.27 1028.4 412.602 Q1027.69 411.934 1026.43 411.934 M1024.67 411.187 Q1023.54 410.909 1022.91 410.137 Q1022.29 409.364 1022.29 408.253 Q1022.29 406.699 1023.39 405.797 Q1024.5 404.894 1026.43 404.894 Q1028.36 404.894 1029.46 405.797 Q1030.57 406.699 1030.57 408.253 Q1030.57 409.364 1029.93 410.137 Q1029.31 410.909 1028.19 411.187 Q1029.46 411.482 1030.16 412.342 Q1030.87 413.201 1030.87 414.442 Q1030.87 416.326 1029.72 417.333 Q1028.57 418.34 1026.43 418.34 Q1024.28 418.34 1023.13 417.333 Q1021.98 416.326 1021.98 414.442 Q1021.98 413.201 1022.69 412.342 Q1023.41 411.482 1024.67 411.187 M1024.03 408.418 Q1024.03 409.425 1024.66 409.989 Q1025.29 410.553 1026.43 410.553 Q1027.55 410.553 1028.19 409.989 Q1028.83 409.425 1028.83 408.418 Q1028.83 407.411 1028.19 406.847 Q1027.55 406.283 1026.43 406.283 Q1025.29 406.283 1024.66 406.847 Q1024.03 407.411 1024.03 408.418 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1051.47 404.323 L1054.34 404.323 L1054.34 394.436 L1051.22 395.061 L1051.22 393.464 L1054.32 392.839 L1056.07 392.839 L1056.07 404.323 L1058.94 404.323 L1058.94 405.799 L1051.47 405.799 L1051.47 404.323 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1062.53 405.53 L1062.53 403.933 Q1063.19 404.245 1063.87 404.41 Q1064.55 404.575 1065.2 404.575 Q1066.93 404.575 1067.84 403.412 Q1068.76 402.24 1068.89 399.862 Q1068.39 400.608 1067.62 401.007 Q1066.85 401.407 1065.91 401.407 Q1063.96 401.407 1062.83 400.235 Q1061.7 399.054 1061.7 397.014 Q1061.7 395.018 1062.88 393.811 Q1064.06 392.605 1066.02 392.605 Q1068.27 392.605 1069.45 394.332 Q1070.64 396.051 1070.64 399.332 Q1070.64 402.396 1069.18 404.228 Q1067.73 406.051 1065.27 406.051 Q1064.61 406.051 1063.94 405.921 Q1063.26 405.79 1062.53 405.53 M1066.02 400.035 Q1067.2 400.035 1067.89 399.228 Q1068.58 398.421 1068.58 397.014 Q1068.58 395.617 1067.89 394.81 Q1067.2 393.994 1066.02 393.994 Q1064.84 393.994 1064.15 394.81 Q1063.46 395.617 1063.46 397.014 Q1063.46 398.421 1064.15 399.228 Q1064.84 400.035 1066.02 400.035 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1092.75 392.859 L1098.87 392.859 L1098.87 394.334 L1090.65 394.334 L1090.65 392.859 Q1091.64 391.826 1093.36 390.089 Q1095.09 388.345 1095.53 387.841 Q1096.37 386.895 1096.7 386.244 Q1097.04 385.584 1097.04 384.951 Q1097.04 383.918 1096.31 383.267 Q1095.59 382.616 1094.43 382.616 Q1093.61 382.616 1092.69 382.902 Q1091.77 383.188 1090.73 383.77 L1090.73 381.999 Q1091.79 381.574 1092.71 381.357 Q1093.63 381.14 1094.4 381.14 Q1096.41 381.14 1097.61 382.147 Q1098.81 383.154 1098.81 384.838 Q1098.81 385.636 1098.5 386.357 Q1098.21 387.069 1097.42 388.041 Q1097.2 388.293 1096.04 389.499 Q1094.87 390.697 1092.75 392.859 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1106.31 382.529 Q1104.95 382.529 1104.27 383.866 Q1103.59 385.194 1103.59 387.867 Q1103.59 390.532 1104.27 391.869 Q1104.95 393.197 1106.31 393.197 Q1107.67 393.197 1108.34 391.869 Q1109.03 390.532 1109.03 387.867 Q1109.03 385.194 1108.34 383.866 Q1107.67 382.529 1106.31 382.529 M1106.31 381.14 Q1108.48 381.14 1109.63 382.867 Q1110.78 384.586 1110.78 387.867 Q1110.78 391.14 1109.63 392.867 Q1108.48 394.586 1106.31 394.586 Q1104.13 394.586 1102.97 392.867 Q1101.83 391.14 1101.83 387.867 Q1101.83 384.586 1102.97 382.867 Q1104.13 381.14 1106.31 381.14 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1132.77 382.071 L1138.89 382.071 L1138.89 383.547 L1130.66 383.547 L1130.66 382.071 Q1131.66 381.038 1133.38 379.302 Q1135.11 377.558 1135.55 377.054 Q1136.39 376.108 1136.72 375.457 Q1137.06 374.797 1137.06 374.163 Q1137.06 373.131 1136.33 372.479 Q1135.61 371.828 1134.45 371.828 Q1133.62 371.828 1132.7 372.115 Q1131.79 372.401 1130.75 372.983 L1130.75 371.212 Q1131.81 370.787 1132.73 370.57 Q1133.65 370.353 1134.41 370.353 Q1136.42 370.353 1137.62 371.36 Q1138.82 372.367 1138.82 374.051 Q1138.82 374.849 1138.52 375.57 Q1138.22 376.282 1137.43 377.254 Q1137.21 377.505 1136.05 378.712 Q1134.89 379.91 1132.77 382.071 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1142.87 382.071 L1145.74 382.071 L1145.74 372.184 L1142.62 372.809 L1142.62 371.212 L1145.72 370.587 L1147.47 370.587 L1147.47 382.071 L1150.34 382.071 L1150.34 383.547 L1142.87 383.547 L1142.87 382.071 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1172.62 371.854 L1178.74 371.854 L1178.74 373.33 L1170.52 373.33 L1170.52 371.854 Q1171.51 370.821 1173.23 369.085 Q1174.96 367.341 1175.4 366.837 Q1176.24 365.891 1176.57 365.24 Q1176.91 364.58 1176.91 363.947 Q1176.91 362.914 1176.18 362.262 Q1175.46 361.611 1174.3 361.611 Q1173.48 361.611 1172.56 361.898 Q1171.64 362.184 1170.6 362.766 L1170.6 360.995 Q1171.66 360.57 1172.58 360.353 Q1173.5 360.136 1174.27 360.136 Q1176.28 360.136 1177.48 361.143 Q1178.68 362.15 1178.68 363.834 Q1178.68 364.632 1178.37 365.353 Q1178.08 366.065 1177.29 367.037 Q1177.07 367.289 1175.91 368.495 Q1174.74 369.693 1172.62 371.854 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1183.94 371.854 L1190.06 371.854 L1190.06 373.33 L1181.83 373.33 L1181.83 371.854 Q1182.82 370.821 1184.54 369.085 Q1186.27 367.341 1186.71 366.837 Q1187.56 365.891 1187.89 365.24 Q1188.22 364.58 1188.22 363.947 Q1188.22 362.914 1187.49 362.262 Q1186.77 361.611 1185.61 361.611 Q1184.79 361.611 1183.87 361.898 Q1182.95 362.184 1181.91 362.766 L1181.91 360.995 Q1182.97 360.57 1183.89 360.353 Q1184.81 360.136 1185.58 360.136 Q1187.59 360.136 1188.79 361.143 Q1189.99 362.15 1189.99 363.834 Q1189.99 364.632 1189.68 365.353 Q1189.39 366.065 1188.6 367.037 Q1188.38 367.289 1187.22 368.495 Q1186.05 369.693 1183.94 371.854 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1212.23 362.126 L1218.35 362.126 L1218.35 363.601 L1210.12 363.601 L1210.12 362.126 Q1211.12 361.093 1212.84 359.357 Q1214.57 357.612 1215.01 357.108 Q1215.85 356.162 1216.18 355.511 Q1216.52 354.851 1216.52 354.218 Q1216.52 353.185 1215.79 352.534 Q1215.07 351.883 1213.91 351.883 Q1213.08 351.883 1212.16 352.169 Q1211.25 352.456 1210.21 353.037 L1210.21 351.266 Q1211.27 350.841 1212.19 350.624 Q1213.11 350.407 1213.87 350.407 Q1215.89 350.407 1217.08 351.414 Q1218.28 352.421 1218.28 354.105 Q1218.28 354.904 1217.98 355.624 Q1217.68 356.336 1216.89 357.308 Q1216.68 357.56 1215.51 358.766 Q1214.35 359.964 1212.23 362.126 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1227.34 356.614 Q1228.6 356.883 1229.31 357.733 Q1230.02 358.584 1230.02 359.834 Q1230.02 361.752 1228.7 362.803 Q1227.38 363.853 1224.95 363.853 Q1224.13 363.853 1223.26 363.688 Q1222.41 363.532 1221.49 363.211 L1221.49 361.518 Q1222.21 361.943 1223.08 362.16 Q1223.95 362.377 1224.9 362.377 Q1226.55 362.377 1227.41 361.726 Q1228.27 361.075 1228.27 359.834 Q1228.27 358.688 1227.47 358.046 Q1226.67 357.395 1225.24 357.395 L1223.72 357.395 L1223.72 355.954 L1225.3 355.954 Q1226.6 355.954 1227.28 355.442 Q1227.97 354.921 1227.97 353.949 Q1227.97 352.95 1227.26 352.421 Q1226.55 351.883 1225.24 351.883 Q1224.51 351.883 1223.69 352.039 Q1222.87 352.195 1221.88 352.525 L1221.88 350.963 Q1222.87 350.685 1223.74 350.546 Q1224.62 350.407 1225.39 350.407 Q1227.39 350.407 1228.55 351.319 Q1229.71 352.221 1229.71 353.766 Q1229.71 354.843 1229.1 355.589 Q1228.48 356.327 1227.34 356.614 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1251.8 352.821 L1257.92 352.821 L1257.92 354.297 L1249.7 354.297 L1249.7 352.821 Q1250.69 351.788 1252.41 350.052 Q1254.14 348.307 1254.58 347.804 Q1255.42 346.858 1255.75 346.207 Q1256.09 345.547 1256.09 344.913 Q1256.09 343.88 1255.36 343.229 Q1254.64 342.578 1253.48 342.578 Q1252.66 342.578 1251.74 342.865 Q1250.82 343.151 1249.78 343.733 L1249.78 341.962 Q1250.84 341.537 1251.76 341.32 Q1252.68 341.103 1253.45 341.103 Q1255.46 341.103 1256.66 342.11 Q1257.85 343.117 1257.85 344.801 Q1257.85 345.599 1257.55 346.32 Q1257.26 347.031 1256.47 348.004 Q1256.25 348.255 1255.09 349.462 Q1253.92 350.66 1251.8 352.821 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1266.42 342.865 L1262 349.783 L1266.42 349.783 L1266.42 342.865 M1265.96 341.337 L1268.17 341.337 L1268.17 349.783 L1270.02 349.783 L1270.02 351.241 L1268.17 351.241 L1268.17 354.297 L1266.42 354.297 L1266.42 351.241 L1260.57 351.241 L1260.57 349.549 L1265.96 341.337 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1291.87 343.89 L1297.99 343.89 L1297.99 345.366 L1289.76 345.366 L1289.76 343.89 Q1290.76 342.857 1292.48 341.121 Q1294.2 339.376 1294.65 338.873 Q1295.49 337.927 1295.82 337.276 Q1296.16 336.616 1296.16 335.982 Q1296.16 334.949 1295.43 334.298 Q1294.71 333.647 1293.54 333.647 Q1292.72 333.647 1291.8 333.934 Q1290.89 334.22 1289.84 334.802 L1289.84 333.031 Q1290.9 332.605 1291.82 332.388 Q1292.74 332.171 1293.51 332.171 Q1295.52 332.171 1296.72 333.178 Q1297.92 334.185 1297.92 335.869 Q1297.92 336.668 1297.61 337.388 Q1297.32 338.1 1296.53 339.072 Q1296.31 339.324 1295.15 340.531 Q1293.99 341.729 1291.87 343.89 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1301.69 332.406 L1308.57 332.406 L1308.57 333.882 L1303.29 333.882 L1303.29 337.059 Q1303.67 336.928 1304.05 336.868 Q1304.44 336.798 1304.82 336.798 Q1306.99 336.798 1308.26 337.987 Q1309.52 339.177 1309.52 341.208 Q1309.52 343.3 1308.22 344.463 Q1306.92 345.618 1304.55 345.618 Q1303.73 345.618 1302.88 345.479 Q1302.04 345.34 1301.14 345.062 L1301.14 343.3 Q1301.92 343.725 1302.75 343.934 Q1303.59 344.142 1304.52 344.142 Q1306.02 344.142 1306.89 343.352 Q1307.77 342.562 1307.77 341.208 Q1307.77 339.854 1306.89 339.064 Q1306.02 338.274 1304.52 338.274 Q1303.81 338.274 1303.11 338.43 Q1302.41 338.586 1301.69 338.916 L1301.69 332.406 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1331.44 335.291 L1337.56 335.291 L1337.56 336.766 L1329.33 336.766 L1329.33 335.291 Q1330.32 334.258 1332.04 332.521 Q1333.77 330.777 1334.21 330.273 Q1335.06 329.327 1335.39 328.676 Q1335.72 328.016 1335.72 327.383 Q1335.72 326.35 1334.99 325.699 Q1334.27 325.048 1333.11 325.048 Q1332.29 325.048 1331.37 325.334 Q1330.45 325.62 1329.41 326.202 L1329.41 324.431 Q1330.47 324.006 1331.39 323.789 Q1332.31 323.572 1333.08 323.572 Q1335.09 323.572 1336.29 324.579 Q1337.49 325.586 1337.49 327.27 Q1337.49 328.068 1337.18 328.789 Q1336.89 329.501 1336.1 330.473 Q1335.88 330.725 1334.72 331.931 Q1333.55 333.129 1331.44 335.291 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1345.2 329.587 Q1344.02 329.587 1343.33 330.395 Q1342.64 331.202 1342.64 332.608 Q1342.64 334.006 1343.33 334.822 Q1344.02 335.629 1345.2 335.629 Q1346.38 335.629 1347.07 334.822 Q1347.76 334.006 1347.76 332.608 Q1347.76 331.202 1347.07 330.395 Q1346.38 329.587 1345.2 329.587 M1348.68 324.093 L1348.68 325.69 Q1348.02 325.377 1347.35 325.212 Q1346.68 325.048 1346.02 325.048 Q1344.28 325.048 1343.36 326.219 Q1342.45 327.391 1342.32 329.761 Q1342.83 329.006 1343.61 328.607 Q1344.38 328.199 1345.31 328.199 Q1347.26 328.199 1348.39 329.388 Q1349.53 330.568 1349.53 332.608 Q1349.53 334.605 1348.35 335.811 Q1347.16 337.018 1345.2 337.018 Q1342.95 337.018 1341.77 335.299 Q1340.58 333.572 1340.58 330.299 Q1340.58 327.226 1342.03 325.403 Q1343.49 323.572 1345.95 323.572 Q1346.61 323.572 1347.28 323.702 Q1347.95 323.832 1348.68 324.093 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1371.42 326.988 L1377.54 326.988 L1377.54 328.464 L1369.31 328.464 L1369.31 326.988 Q1370.31 325.955 1372.03 324.219 Q1373.76 322.474 1374.2 321.971 Q1375.04 321.025 1375.37 320.374 Q1375.71 319.714 1375.71 319.08 Q1375.71 318.047 1374.98 317.396 Q1374.26 316.745 1373.1 316.745 Q1372.27 316.745 1371.35 317.032 Q1370.44 317.318 1369.4 317.9 L1369.4 316.129 Q1370.46 315.703 1371.38 315.486 Q1372.3 315.269 1373.06 315.269 Q1375.08 315.269 1376.27 316.276 Q1377.47 317.283 1377.47 318.967 Q1377.47 319.766 1377.17 320.486 Q1376.87 321.198 1376.08 322.17 Q1375.87 322.422 1374.7 323.629 Q1373.54 324.827 1371.42 326.988 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1380.78 315.504 L1389.11 315.504 L1389.11 316.25 L1384.41 328.464 L1382.58 328.464 L1387 316.979 L1380.78 316.979 L1380.78 315.504 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1411.05 318.954 L1417.17 318.954 L1417.17 320.43 L1408.94 320.43 L1408.94 318.954 Q1409.94 317.921 1411.66 316.185 Q1413.39 314.44 1413.83 313.937 Q1414.67 312.991 1415 312.34 Q1415.34 311.68 1415.34 311.046 Q1415.34 310.013 1414.61 309.362 Q1413.89 308.711 1412.73 308.711 Q1411.9 308.711 1410.98 308.998 Q1410.07 309.284 1409.03 309.866 L1409.03 308.095 Q1410.09 307.669 1411.01 307.452 Q1411.93 307.235 1412.69 307.235 Q1414.71 307.235 1415.91 308.242 Q1417.1 309.249 1417.1 310.933 Q1417.1 311.732 1416.8 312.452 Q1416.51 313.164 1415.72 314.136 Q1415.5 314.388 1414.34 315.595 Q1413.17 316.793 1411.05 318.954 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1424.6 314.275 Q1423.35 314.275 1422.63 314.944 Q1421.92 315.612 1421.92 316.784 Q1421.92 317.956 1422.63 318.624 Q1423.35 319.293 1424.6 319.293 Q1425.85 319.293 1426.57 318.624 Q1427.3 317.947 1427.3 316.784 Q1427.3 315.612 1426.57 314.944 Q1425.86 314.275 1424.6 314.275 M1422.85 313.529 Q1421.72 313.251 1421.09 312.478 Q1420.46 311.706 1420.46 310.595 Q1420.46 309.041 1421.57 308.138 Q1422.68 307.235 1424.6 307.235 Q1426.54 307.235 1427.64 308.138 Q1428.74 309.041 1428.74 310.595 Q1428.74 311.706 1428.11 312.478 Q1427.49 313.251 1426.37 313.529 Q1427.63 313.824 1428.34 314.683 Q1429.05 315.543 1429.05 316.784 Q1429.05 318.668 1427.89 319.675 Q1426.75 320.682 1424.6 320.682 Q1422.46 320.682 1421.31 319.675 Q1420.16 318.668 1420.16 316.784 Q1420.16 315.543 1420.87 314.683 Q1421.58 313.824 1422.85 313.529 M1422.21 310.76 Q1422.21 311.767 1422.83 312.331 Q1423.47 312.895 1424.6 312.895 Q1425.73 312.895 1426.37 312.331 Q1427.01 311.767 1427.01 310.76 Q1427.01 309.753 1426.37 309.189 Q1425.73 308.624 1424.6 308.624 Q1423.47 308.624 1422.83 309.189 Q1422.21 309.753 1422.21 310.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1450.86 311.164 L1456.98 311.164 L1456.98 312.64 L1448.75 312.64 L1448.75 311.164 Q1449.75 310.131 1451.46 308.395 Q1453.19 306.65 1453.63 306.147 Q1454.48 305.2 1454.81 304.549 Q1455.14 303.89 1455.14 303.256 Q1455.14 302.223 1454.42 301.572 Q1453.7 300.921 1452.53 300.921 Q1451.71 300.921 1450.79 301.207 Q1449.88 301.494 1448.83 302.075 L1448.83 300.305 Q1449.89 299.879 1450.81 299.662 Q1451.73 299.445 1452.5 299.445 Q1454.51 299.445 1455.71 300.452 Q1456.91 301.459 1456.91 303.143 Q1456.91 303.942 1456.6 304.662 Q1456.31 305.374 1455.52 306.346 Q1455.3 306.598 1454.14 307.805 Q1452.97 309.003 1450.86 311.164 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1460.71 312.371 L1460.71 310.773 Q1461.37 311.086 1462.05 311.251 Q1462.72 311.416 1463.37 311.416 Q1465.11 311.416 1466.02 310.253 Q1466.94 309.081 1467.07 306.702 Q1466.57 307.449 1465.8 307.848 Q1465.02 308.247 1464.09 308.247 Q1462.14 308.247 1461 307.075 Q1459.88 305.895 1459.88 303.855 Q1459.88 301.858 1461.06 300.652 Q1462.24 299.445 1464.2 299.445 Q1466.45 299.445 1467.63 301.173 Q1468.82 302.891 1468.82 306.173 Q1468.82 309.237 1467.36 311.068 Q1465.91 312.891 1463.45 312.891 Q1462.79 312.891 1462.12 312.761 Q1461.44 312.631 1460.71 312.371 M1464.2 306.876 Q1465.38 306.876 1466.06 306.068 Q1466.76 305.261 1466.76 303.855 Q1466.76 302.457 1466.06 301.65 Q1465.38 300.834 1464.2 300.834 Q1463.02 300.834 1462.32 301.65 Q1461.64 302.457 1461.64 303.855 Q1461.64 305.261 1462.32 306.068 Q1463.02 306.876 1464.2 306.876 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1494.38 298.085 Q1495.64 298.354 1496.34 299.204 Q1497.06 300.055 1497.06 301.305 Q1497.06 303.224 1495.74 304.274 Q1494.42 305.324 1491.99 305.324 Q1491.17 305.324 1490.3 305.159 Q1489.44 305.003 1488.52 304.682 L1488.52 302.989 Q1489.25 303.414 1490.12 303.632 Q1490.99 303.849 1491.94 303.849 Q1493.58 303.849 1494.44 303.197 Q1495.31 302.546 1495.31 301.305 Q1495.31 300.159 1494.5 299.517 Q1493.71 298.866 1492.27 298.866 L1490.76 298.866 L1490.76 297.425 L1492.34 297.425 Q1493.64 297.425 1494.32 296.913 Q1495.01 296.392 1495.01 295.42 Q1495.01 294.421 1494.3 293.892 Q1493.59 293.354 1492.27 293.354 Q1491.55 293.354 1490.73 293.51 Q1489.9 293.666 1488.91 293.996 L1488.91 292.434 Q1489.91 292.156 1490.78 292.017 Q1491.66 291.878 1492.43 291.878 Q1494.43 291.878 1495.59 292.79 Q1496.75 293.692 1496.75 295.237 Q1496.75 296.314 1496.14 297.06 Q1495.52 297.798 1494.38 298.085 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1504.13 293.267 Q1502.78 293.267 1502.09 294.604 Q1501.41 295.932 1501.41 298.605 Q1501.41 301.27 1502.09 302.607 Q1502.78 303.935 1504.13 303.935 Q1505.49 303.935 1506.17 302.607 Q1506.86 301.27 1506.86 298.605 Q1506.86 295.932 1506.17 294.604 Q1505.49 293.267 1504.13 293.267 M1504.13 291.878 Q1506.31 291.878 1507.46 293.606 Q1508.61 295.324 1508.61 298.605 Q1508.61 301.878 1507.46 303.605 Q1506.31 305.324 1504.13 305.324 Q1501.95 305.324 1500.8 303.605 Q1499.65 301.878 1499.65 298.605 Q1499.65 295.324 1500.8 293.606 Q1501.95 291.878 1504.13 291.878 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1534.4 290.722 Q1535.66 290.991 1536.36 291.842 Q1537.07 292.693 1537.07 293.943 Q1537.07 295.861 1535.75 296.912 Q1534.43 297.962 1532 297.962 Q1531.19 297.962 1530.32 297.797 Q1529.46 297.641 1528.54 297.32 L1528.54 295.627 Q1529.27 296.052 1530.14 296.269 Q1531 296.486 1531.95 296.486 Q1533.6 296.486 1534.46 295.835 Q1535.33 295.184 1535.33 293.943 Q1535.33 292.797 1534.52 292.155 Q1533.72 291.504 1532.29 291.504 L1530.78 291.504 L1530.78 290.063 L1532.36 290.063 Q1533.65 290.063 1534.34 289.55 Q1535.02 289.03 1535.02 288.057 Q1535.02 287.059 1534.31 286.53 Q1533.61 285.991 1532.29 285.991 Q1531.57 285.991 1530.74 286.148 Q1529.92 286.304 1528.93 286.634 L1528.93 285.071 Q1529.93 284.794 1530.8 284.655 Q1531.67 284.516 1532.45 284.516 Q1534.44 284.516 1535.6 285.427 Q1536.77 286.33 1536.77 287.875 Q1536.77 288.952 1536.15 289.698 Q1535.54 290.436 1534.4 290.722 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1540.7 296.234 L1543.56 296.234 L1543.56 286.347 L1540.45 286.972 L1540.45 285.375 L1543.55 284.75 L1545.3 284.75 L1545.3 296.234 L1548.17 296.234 L1548.17 297.71 L1540.7 297.71 L1540.7 296.234 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1574.25 283.549 Q1575.51 283.818 1576.21 284.669 Q1576.93 285.52 1576.93 286.77 Q1576.93 288.688 1575.61 289.738 Q1574.29 290.789 1571.86 290.789 Q1571.04 290.789 1570.17 290.624 Q1569.31 290.467 1568.39 290.146 L1568.39 288.454 Q1569.12 288.879 1569.99 289.096 Q1570.86 289.313 1571.81 289.313 Q1573.45 289.313 1574.31 288.662 Q1575.18 288.011 1575.18 286.77 Q1575.18 285.624 1574.37 284.981 Q1573.58 284.33 1572.14 284.33 L1570.63 284.33 L1570.63 282.889 L1572.21 282.889 Q1573.51 282.889 1574.19 282.377 Q1574.88 281.856 1574.88 280.884 Q1574.88 279.886 1574.17 279.356 Q1573.46 278.818 1572.14 278.818 Q1571.42 278.818 1570.6 278.974 Q1569.77 279.131 1568.78 279.461 L1568.78 277.898 Q1569.78 277.62 1570.65 277.481 Q1571.53 277.342 1572.3 277.342 Q1574.3 277.342 1575.46 278.254 Q1576.62 279.157 1576.62 280.702 Q1576.62 281.778 1576.01 282.525 Q1575.39 283.263 1574.25 283.549 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1581.76 289.061 L1587.88 289.061 L1587.88 290.537 L1579.65 290.537 L1579.65 289.061 Q1580.65 288.028 1582.37 286.292 Q1584.1 284.547 1584.54 284.044 Q1585.38 283.098 1585.71 282.447 Q1586.05 281.787 1586.05 281.153 Q1586.05 280.12 1585.32 279.469 Q1584.6 278.818 1583.44 278.818 Q1582.61 278.818 1581.69 279.105 Q1580.78 279.391 1579.74 279.973 L1579.74 278.202 Q1580.8 277.777 1581.72 277.56 Q1582.64 277.342 1583.4 277.342 Q1585.42 277.342 1586.61 278.349 Q1587.81 279.356 1587.81 281.04 Q1587.81 281.839 1587.51 282.559 Q1587.21 283.271 1586.42 284.243 Q1586.21 284.495 1585.04 285.702 Q1583.88 286.9 1581.76 289.061 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1613.86 276.551 Q1615.12 276.82 1615.82 277.671 Q1616.53 278.521 1616.53 279.771 Q1616.53 281.69 1615.21 282.74 Q1613.9 283.791 1611.46 283.791 Q1610.65 283.791 1609.78 283.626 Q1608.92 283.469 1608 283.148 L1608 281.455 Q1608.73 281.881 1609.6 282.098 Q1610.47 282.315 1611.41 282.315 Q1613.06 282.315 1613.92 281.664 Q1614.79 281.013 1614.79 279.771 Q1614.79 278.626 1613.98 277.983 Q1613.18 277.332 1611.75 277.332 L1610.24 277.332 L1610.24 275.891 L1611.82 275.891 Q1613.11 275.891 1613.8 275.379 Q1614.49 274.858 1614.49 273.886 Q1614.49 272.888 1613.77 272.358 Q1613.07 271.82 1611.75 271.82 Q1611.03 271.82 1610.21 271.976 Q1609.38 272.133 1608.39 272.462 L1608.39 270.9 Q1609.39 270.622 1610.26 270.483 Q1611.13 270.344 1611.91 270.344 Q1613.9 270.344 1615.07 271.256 Q1616.23 272.159 1616.23 273.704 Q1616.23 274.78 1615.61 275.527 Q1615 276.265 1613.86 276.551 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1625.17 276.551 Q1626.43 276.82 1627.13 277.671 Q1627.84 278.521 1627.84 279.771 Q1627.84 281.69 1626.53 282.74 Q1625.21 283.791 1622.78 283.791 Q1621.96 283.791 1621.09 283.626 Q1620.23 283.469 1619.31 283.148 L1619.31 281.455 Q1620.04 281.881 1620.91 282.098 Q1621.78 282.315 1622.72 282.315 Q1624.37 282.315 1625.23 281.664 Q1626.1 281.013 1626.1 279.771 Q1626.1 278.626 1625.29 277.983 Q1624.49 277.332 1623.06 277.332 L1621.55 277.332 L1621.55 275.891 L1623.13 275.891 Q1624.42 275.891 1625.11 275.379 Q1625.8 274.858 1625.8 273.886 Q1625.8 272.888 1625.08 272.358 Q1624.38 271.82 1623.06 271.82 Q1622.34 271.82 1621.52 271.976 Q1620.69 272.133 1619.7 272.462 L1619.7 270.9 Q1620.7 270.622 1621.57 270.483 Q1622.45 270.344 1623.22 270.344 Q1625.21 270.344 1626.38 271.256 Q1627.54 272.159 1627.54 273.704 Q1627.54 274.78 1626.92 275.527 Q1626.31 276.265 1625.17 276.551 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1653.43 269.716 Q1654.69 269.985 1655.39 270.836 Q1656.11 271.686 1656.11 272.936 Q1656.11 274.855 1654.79 275.905 Q1653.47 276.955 1651.04 276.955 Q1650.22 276.955 1649.35 276.79 Q1648.49 276.634 1647.57 276.313 L1647.57 274.62 Q1648.3 275.046 1649.17 275.263 Q1650.04 275.48 1650.98 275.48 Q1652.63 275.48 1653.49 274.829 Q1654.36 274.178 1654.36 272.936 Q1654.36 271.791 1653.55 271.148 Q1652.76 270.497 1651.32 270.497 L1649.81 270.497 L1649.81 269.056 L1651.39 269.056 Q1652.69 269.056 1653.37 268.544 Q1654.06 268.023 1654.06 267.051 Q1654.06 266.053 1653.35 265.523 Q1652.64 264.985 1651.32 264.985 Q1650.6 264.985 1649.78 265.141 Q1648.95 265.297 1647.96 265.627 L1647.96 264.065 Q1648.96 263.787 1649.83 263.648 Q1650.71 263.509 1651.48 263.509 Q1653.48 263.509 1654.64 264.421 Q1655.8 265.324 1655.8 266.869 Q1655.8 267.945 1655.19 268.692 Q1654.57 269.429 1653.43 269.716 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1664.25 265.271 L1659.82 272.19 L1664.25 272.19 L1664.25 265.271 M1663.79 263.744 L1665.99 263.744 L1665.99 272.19 L1667.84 272.19 L1667.84 273.648 L1665.99 273.648 L1665.99 276.704 L1664.25 276.704 L1664.25 273.648 L1658.4 273.648 L1658.4 271.955 L1663.79 263.744 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1693.5 263.033 Q1694.75 263.302 1695.46 264.153 Q1696.17 265.003 1696.17 266.253 Q1696.17 268.172 1694.85 269.222 Q1693.53 270.272 1691.1 270.272 Q1690.28 270.272 1689.42 270.107 Q1688.56 269.951 1687.64 269.63 L1687.64 267.937 Q1688.37 268.363 1689.23 268.58 Q1690.1 268.797 1691.05 268.797 Q1692.7 268.797 1693.56 268.146 Q1694.42 267.495 1694.42 266.253 Q1694.42 265.107 1693.62 264.465 Q1692.82 263.814 1691.39 263.814 L1689.88 263.814 L1689.88 262.373 L1691.46 262.373 Q1692.75 262.373 1693.44 261.861 Q1694.12 261.34 1694.12 260.368 Q1694.12 259.37 1693.41 258.84 Q1692.71 258.302 1691.39 258.302 Q1690.67 258.302 1689.84 258.458 Q1689.02 258.614 1688.03 258.944 L1688.03 257.382 Q1689.03 257.104 1689.89 256.965 Q1690.77 256.826 1691.54 256.826 Q1693.54 256.826 1694.7 257.738 Q1695.87 258.64 1695.87 260.186 Q1695.87 261.262 1695.25 262.009 Q1694.63 262.746 1693.5 263.033 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1699.51 257.061 L1706.4 257.061 L1706.4 258.536 L1701.12 258.536 L1701.12 261.713 Q1701.5 261.583 1701.88 261.522 Q1702.26 261.453 1702.65 261.453 Q1704.82 261.453 1706.08 262.642 Q1707.35 263.831 1707.35 265.863 Q1707.35 267.955 1706.05 269.118 Q1704.75 270.272 1702.38 270.272 Q1701.56 270.272 1700.71 270.133 Q1699.87 269.995 1698.96 269.717 L1698.96 267.955 Q1699.75 268.38 1700.58 268.588 Q1701.41 268.797 1702.34 268.797 Q1703.84 268.797 1704.72 268.007 Q1705.6 267.217 1705.6 265.863 Q1705.6 264.509 1704.72 263.719 Q1703.84 262.929 1702.34 262.929 Q1701.64 262.929 1700.94 263.085 Q1700.24 263.241 1699.51 263.571 L1699.51 257.061 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1733.06 256.492 Q1734.32 256.761 1735.03 257.612 Q1735.74 258.463 1735.74 259.713 Q1735.74 261.631 1734.42 262.681 Q1733.1 263.732 1730.67 263.732 Q1729.85 263.732 1728.98 263.567 Q1728.12 263.411 1727.2 263.089 L1727.2 261.397 Q1727.93 261.822 1728.8 262.039 Q1729.67 262.256 1730.62 262.256 Q1732.27 262.256 1733.12 261.605 Q1733.99 260.954 1733.99 259.713 Q1733.99 258.567 1733.19 257.924 Q1732.39 257.273 1730.95 257.273 L1729.44 257.273 L1729.44 255.832 L1731.02 255.832 Q1732.32 255.832 1733 255.32 Q1733.69 254.799 1733.69 253.827 Q1733.69 252.829 1732.98 252.299 Q1732.27 251.761 1730.95 251.761 Q1730.23 251.761 1729.41 251.918 Q1728.58 252.074 1727.6 252.404 L1727.6 250.841 Q1728.59 250.563 1729.46 250.424 Q1730.34 250.286 1731.11 250.286 Q1733.11 250.286 1734.27 251.197 Q1735.43 252.1 1735.43 253.645 Q1735.43 254.721 1734.82 255.468 Q1734.2 256.206 1733.06 256.492 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1743.03 256.301 Q1741.85 256.301 1741.15 257.108 Q1740.47 257.916 1740.47 259.322 Q1740.47 260.72 1741.15 261.536 Q1741.85 262.343 1743.03 262.343 Q1744.21 262.343 1744.9 261.536 Q1745.59 260.72 1745.59 259.322 Q1745.59 257.916 1744.9 257.108 Q1744.21 256.301 1743.03 256.301 M1746.51 250.806 L1746.51 252.404 Q1745.85 252.091 1745.17 251.926 Q1744.5 251.761 1743.85 251.761 Q1742.11 251.761 1741.19 252.933 Q1740.28 254.105 1740.15 256.475 Q1740.66 255.72 1741.43 255.32 Q1742.2 254.912 1743.13 254.912 Q1745.09 254.912 1746.22 256.102 Q1747.35 257.282 1747.35 259.322 Q1747.35 261.319 1746.17 262.525 Q1744.99 263.732 1743.03 263.732 Q1740.78 263.732 1739.59 262.013 Q1738.4 260.286 1738.4 257.013 Q1738.4 253.94 1739.86 252.117 Q1741.32 250.286 1743.78 250.286 Q1744.44 250.286 1745.1 250.416 Q1745.78 250.546 1746.51 250.806 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1773.05 250.085 Q1774.31 250.354 1775.01 251.205 Q1775.72 252.056 1775.72 253.306 Q1775.72 255.224 1774.4 256.274 Q1773.08 257.325 1770.65 257.325 Q1769.84 257.325 1768.97 257.16 Q1768.11 257.003 1767.19 256.682 L1767.19 254.99 Q1767.92 255.415 1768.79 255.632 Q1769.65 255.849 1770.6 255.849 Q1772.25 255.849 1773.11 255.198 Q1773.98 254.547 1773.98 253.306 Q1773.98 252.16 1773.17 251.517 Q1772.37 250.866 1770.94 250.866 L1769.43 250.866 L1769.43 249.425 L1771.01 249.425 Q1772.3 249.425 1772.99 248.913 Q1773.67 248.392 1773.67 247.42 Q1773.67 246.422 1772.96 245.892 Q1772.26 245.354 1770.94 245.354 Q1770.22 245.354 1769.39 245.51 Q1768.57 245.667 1767.58 245.997 L1767.58 244.434 Q1768.58 244.156 1769.45 244.017 Q1770.32 243.879 1771.1 243.879 Q1773.09 243.879 1774.26 244.79 Q1775.42 245.693 1775.42 247.238 Q1775.42 248.314 1774.8 249.061 Q1774.19 249.799 1773.05 250.085 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1778.6 244.113 L1786.94 244.113 L1786.94 244.859 L1782.23 257.073 L1780.4 257.073 L1784.83 245.589 L1778.6 245.589 L1778.6 244.113 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1812.68 243.804 Q1813.94 244.073 1814.64 244.924 Q1815.36 245.774 1815.36 247.024 Q1815.36 248.943 1814.04 249.993 Q1812.72 251.043 1810.29 251.043 Q1809.47 251.043 1808.6 250.878 Q1807.74 250.722 1806.82 250.401 L1806.82 248.708 Q1807.55 249.134 1808.42 249.351 Q1809.29 249.568 1810.23 249.568 Q1811.88 249.568 1812.74 248.917 Q1813.61 248.266 1813.61 247.024 Q1813.61 245.878 1812.8 245.236 Q1812.01 244.585 1810.57 244.585 L1809.06 244.585 L1809.06 243.144 L1810.64 243.144 Q1811.94 243.144 1812.62 242.632 Q1813.31 242.111 1813.31 241.139 Q1813.31 240.141 1812.6 239.611 Q1811.89 239.073 1810.57 239.073 Q1809.85 239.073 1809.03 239.229 Q1808.2 239.385 1807.21 239.715 L1807.21 238.153 Q1808.21 237.875 1809.08 237.736 Q1809.96 237.597 1810.73 237.597 Q1812.73 237.597 1813.89 238.509 Q1815.05 239.411 1815.05 240.957 Q1815.05 242.033 1814.44 242.78 Q1813.82 243.517 1812.68 243.804 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1822.43 244.637 Q1821.18 244.637 1820.46 245.306 Q1819.75 245.974 1819.75 247.146 Q1819.75 248.318 1820.46 248.986 Q1821.18 249.654 1822.43 249.654 Q1823.68 249.654 1824.4 248.986 Q1825.12 248.309 1825.12 247.146 Q1825.12 245.974 1824.4 245.306 Q1823.69 244.637 1822.43 244.637 M1820.68 243.891 Q1819.55 243.613 1818.91 242.84 Q1818.29 242.068 1818.29 240.957 Q1818.29 239.403 1819.39 238.5 Q1820.5 237.597 1822.43 237.597 Q1824.37 237.597 1825.47 238.5 Q1826.57 239.403 1826.57 240.957 Q1826.57 242.068 1825.94 242.84 Q1825.31 243.613 1824.19 243.891 Q1825.46 244.186 1826.16 245.045 Q1826.87 245.904 1826.87 247.146 Q1826.87 249.029 1825.72 250.036 Q1824.57 251.043 1822.43 251.043 Q1820.29 251.043 1819.13 250.036 Q1817.99 249.029 1817.99 247.146 Q1817.99 245.904 1818.7 245.045 Q1819.41 244.186 1820.68 243.891 M1820.03 241.122 Q1820.03 242.128 1820.66 242.693 Q1821.29 243.257 1822.43 243.257 Q1823.56 243.257 1824.19 242.693 Q1824.83 242.128 1824.83 241.122 Q1824.83 240.115 1824.19 239.55 Q1823.56 238.986 1822.43 238.986 Q1821.29 238.986 1820.66 239.55 Q1820.03 240.115 1820.03 241.122 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1852.48 237.641 Q1853.74 237.91 1854.45 238.761 Q1855.16 239.612 1855.16 240.862 Q1855.16 242.78 1853.84 243.83 Q1852.52 244.881 1850.09 244.881 Q1849.27 244.881 1848.41 244.716 Q1847.55 244.56 1846.63 244.238 L1846.63 242.546 Q1847.35 242.971 1848.22 243.188 Q1849.09 243.405 1850.04 243.405 Q1851.69 243.405 1852.55 242.754 Q1853.41 242.103 1853.41 240.862 Q1853.41 239.716 1852.61 239.073 Q1851.81 238.422 1850.38 238.422 L1848.87 238.422 L1848.87 236.981 L1850.44 236.981 Q1851.74 236.981 1852.42 236.469 Q1853.11 235.948 1853.11 234.976 Q1853.11 233.978 1852.4 233.448 Q1851.69 232.91 1850.38 232.91 Q1849.65 232.91 1848.83 233.067 Q1848.01 233.223 1847.02 233.553 L1847.02 231.99 Q1848.01 231.712 1848.88 231.573 Q1849.76 231.435 1850.53 231.435 Q1852.53 231.435 1853.69 232.346 Q1854.85 233.249 1854.85 234.794 Q1854.85 235.87 1854.24 236.617 Q1853.62 237.355 1852.48 237.641 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1858.54 244.36 L1858.54 242.763 Q1859.19 243.075 1859.87 243.24 Q1860.55 243.405 1861.2 243.405 Q1862.94 243.405 1863.85 242.242 Q1864.77 241.07 1864.9 238.691 Q1864.39 239.438 1863.62 239.837 Q1862.85 240.237 1861.91 240.237 Q1859.97 240.237 1858.83 239.065 Q1857.7 237.884 1857.7 235.844 Q1857.7 233.848 1858.88 232.641 Q1860.06 231.435 1862.02 231.435 Q1864.27 231.435 1865.45 233.162 Q1866.64 234.881 1866.64 238.162 Q1866.64 241.226 1865.18 243.058 Q1863.73 244.881 1861.28 244.881 Q1860.62 244.881 1859.94 244.75 Q1859.26 244.62 1858.54 244.36 M1862.02 238.865 Q1863.21 238.865 1863.89 238.058 Q1864.59 237.251 1864.59 235.844 Q1864.59 234.447 1863.89 233.639 Q1863.21 232.823 1862.02 232.823 Q1860.84 232.823 1860.15 233.639 Q1859.46 234.447 1859.46 235.844 Q1859.46 237.251 1860.15 238.058 Q1860.84 238.865 1862.02 238.865 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1891.98 227.146 L1887.56 234.065 L1891.98 234.065 L1891.98 227.146 M1891.52 225.618 L1893.73 225.618 L1893.73 234.065 L1895.58 234.065 L1895.58 235.523 L1893.73 235.523 L1893.73 238.578 L1891.98 238.578 L1891.98 235.523 L1886.13 235.523 L1886.13 233.83 L1891.52 225.618 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1902.23 226.773 Q1900.87 226.773 1900.19 228.11 Q1899.51 229.438 1899.51 232.111 Q1899.51 234.776 1900.19 236.113 Q1900.87 237.441 1902.23 237.441 Q1903.59 237.441 1904.27 236.113 Q1904.95 234.776 1904.95 232.111 Q1904.95 229.438 1904.27 228.11 Q1903.59 226.773 1902.23 226.773 M1902.23 225.384 Q1904.41 225.384 1905.55 227.111 Q1906.71 228.83 1906.71 232.111 Q1906.71 235.384 1905.55 237.111 Q1904.41 238.83 1902.23 238.83 Q1900.05 238.83 1898.89 237.111 Q1897.75 235.384 1897.75 232.111 Q1897.75 228.83 1898.89 227.111 Q1900.05 225.384 1902.23 225.384 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1932 221.202 L1927.57 228.12 L1932 228.12 L1932 221.202 M1931.54 219.674 L1933.74 219.674 L1933.74 228.12 L1935.59 228.12 L1935.59 229.579 L1933.74 229.579 L1933.74 232.634 L1932 232.634 L1932 229.579 L1926.15 229.579 L1926.15 227.886 L1931.54 219.674 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1938.8 231.159 L1941.66 231.159 L1941.66 221.271 L1938.54 221.896 L1938.54 220.299 L1941.64 219.674 L1943.4 219.674 L1943.4 231.159 L1946.26 231.159 L1946.26 232.634 L1938.8 232.634 L1938.8 231.159 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1971.85 215.359 L1967.43 222.277 L1971.85 222.277 L1971.85 215.359 M1971.39 213.831 L1973.6 213.831 L1973.6 222.277 L1975.45 222.277 L1975.45 223.735 L1973.6 223.735 L1973.6 226.791 L1971.85 226.791 L1971.85 223.735 L1966 223.735 L1966 222.043 L1971.39 213.831 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1979.86 225.315 L1985.98 225.315 L1985.98 226.791 L1977.75 226.791 L1977.75 225.315 Q1978.75 224.282 1980.46 222.546 Q1982.19 220.801 1982.63 220.298 Q1983.48 219.352 1983.81 218.701 Q1984.15 218.041 1984.15 217.407 Q1984.15 216.374 1983.42 215.723 Q1982.7 215.072 1981.53 215.072 Q1980.71 215.072 1979.79 215.359 Q1978.88 215.645 1977.83 216.227 L1977.83 214.456 Q1978.89 214.03 1979.81 213.813 Q1980.73 213.596 1981.5 213.596 Q1983.51 213.596 1984.71 214.603 Q1985.91 215.61 1985.91 217.294 Q1985.91 218.093 1985.6 218.813 Q1985.31 219.525 1984.52 220.497 Q1984.3 220.749 1983.14 221.956 Q1981.98 223.154 1979.86 225.315 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2011.46 209.611 L2007.03 216.529 L2011.46 216.529 L2011.46 209.611 M2011 208.083 L2013.21 208.083 L2013.21 216.529 L2015.05 216.529 L2015.05 217.988 L2013.21 217.988 L2013.21 221.043 L2011.46 221.043 L2011.46 217.988 L2005.61 217.988 L2005.61 216.295 L2011 208.083 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2023.27 214.055 Q2024.52 214.325 2025.23 215.175 Q2025.94 216.026 2025.94 217.276 Q2025.94 219.194 2024.62 220.245 Q2023.3 221.295 2020.87 221.295 Q2020.05 221.295 2019.19 221.13 Q2018.33 220.974 2017.41 220.653 L2017.41 218.96 Q2018.14 219.385 2019 219.602 Q2019.87 219.819 2020.82 219.819 Q2022.47 219.819 2023.33 219.168 Q2024.2 218.517 2024.2 217.276 Q2024.2 216.13 2023.39 215.488 Q2022.59 214.837 2021.16 214.837 L2019.65 214.837 L2019.65 213.396 L2021.23 213.396 Q2022.52 213.396 2023.21 212.884 Q2023.89 212.363 2023.89 211.39 Q2023.89 210.392 2023.18 209.863 Q2022.48 209.325 2021.16 209.325 Q2020.44 209.325 2019.61 209.481 Q2018.79 209.637 2017.8 209.967 L2017.8 208.404 Q2018.8 208.127 2019.66 207.988 Q2020.54 207.849 2021.31 207.849 Q2023.31 207.849 2024.47 208.76 Q2025.64 209.663 2025.64 211.208 Q2025.64 212.285 2025.02 213.031 Q2024.4 213.769 2023.27 214.055 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2051.03 203.955 L2046.61 210.873 L2051.03 210.873 L2051.03 203.955 M2050.57 202.427 L2052.78 202.427 L2052.78 210.873 L2054.63 210.873 L2054.63 212.331 L2052.78 212.331 L2052.78 215.387 L2051.03 215.387 L2051.03 212.331 L2045.18 212.331 L2045.18 210.639 L2050.57 202.427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2062.34 203.955 L2057.92 210.873 L2062.34 210.873 L2062.34 203.955 M2061.88 202.427 L2064.09 202.427 L2064.09 210.873 L2065.94 210.873 L2065.94 212.331 L2064.09 212.331 L2064.09 215.387 L2062.34 215.387 L2062.34 212.331 L2056.49 212.331 L2056.49 210.639 L2061.88 202.427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2091.1 198.386 L2086.67 205.304 L2091.1 205.304 L2091.1 198.386 M2090.64 196.858 L2092.84 196.858 L2092.84 205.304 L2094.69 205.304 L2094.69 206.762 L2092.84 206.762 L2092.84 209.818 L2091.1 209.818 L2091.1 206.762 L2085.25 206.762 L2085.25 205.07 L2090.64 196.858 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2097.61 196.858 L2104.49 196.858 L2104.49 198.333 L2099.21 198.333 L2099.21 201.511 Q2099.59 201.38 2099.98 201.32 Q2100.36 201.25 2100.74 201.25 Q2102.91 201.25 2104.18 202.439 Q2105.45 203.629 2105.45 205.66 Q2105.45 207.752 2104.14 208.915 Q2102.84 210.07 2100.47 210.07 Q2099.66 210.07 2098.8 209.931 Q2097.96 209.792 2097.06 209.514 L2097.06 207.752 Q2097.84 208.177 2098.67 208.386 Q2099.51 208.594 2100.44 208.594 Q2101.94 208.594 2102.82 207.804 Q2103.69 207.014 2103.69 205.66 Q2103.69 204.306 2102.82 203.516 Q2101.94 202.726 2100.44 202.726 Q2099.73 202.726 2099.03 202.882 Q2098.34 203.038 2097.61 203.368 L2097.61 196.858 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2130.66 192.9 L2126.24 199.818 L2130.66 199.818 L2130.66 192.9 M2130.2 191.372 L2132.41 191.372 L2132.41 199.818 L2134.26 199.818 L2134.26 201.276 L2132.41 201.276 L2132.41 204.332 L2130.66 204.332 L2130.66 201.276 L2124.81 201.276 L2124.81 199.584 L2130.2 191.372 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2141.12 197.153 Q2139.94 197.153 2139.25 197.96 Q2138.56 198.768 2138.56 200.174 Q2138.56 201.571 2139.25 202.387 Q2139.94 203.195 2141.12 203.195 Q2142.31 203.195 2142.99 202.387 Q2143.69 201.571 2143.69 200.174 Q2143.69 198.768 2142.99 197.96 Q2142.31 197.153 2141.12 197.153 M2144.61 191.658 L2144.61 193.256 Q2143.95 192.943 2143.27 192.778 Q2142.6 192.613 2141.94 192.613 Q2140.2 192.613 2139.28 193.785 Q2138.37 194.957 2138.24 197.327 Q2138.75 196.571 2139.53 196.172 Q2140.3 195.764 2141.23 195.764 Q2143.18 195.764 2144.31 196.953 Q2145.45 198.134 2145.45 200.174 Q2145.45 202.17 2144.27 203.377 Q2143.09 204.584 2141.12 204.584 Q2138.88 204.584 2137.69 202.865 Q2136.5 201.137 2136.5 197.865 Q2136.5 194.792 2137.96 192.969 Q2139.41 191.137 2141.87 191.137 Q2142.53 191.137 2143.2 191.268 Q2143.88 191.398 2144.61 191.658 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2170.65 187.493 L2166.22 194.412 L2170.65 194.412 L2170.65 187.493 M2170.19 185.965 L2172.39 185.965 L2172.39 194.412 L2174.24 194.412 L2174.24 195.87 L2172.39 195.87 L2172.39 198.925 L2170.65 198.925 L2170.65 195.87 L2164.8 195.87 L2164.8 194.177 L2170.19 185.965 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2176.7 185.965 L2185.03 185.965 L2185.03 186.712 L2180.33 198.925 L2178.5 198.925 L2182.92 187.441 L2176.7 187.441 L2176.7 185.965 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2210.28 182.163 L2205.86 189.081 L2210.28 189.081 L2210.28 182.163 M2209.82 180.635 L2212.03 180.635 L2212.03 189.081 L2213.88 189.081 L2213.88 190.54 L2212.03 190.54 L2212.03 193.595 L2210.28 193.595 L2210.28 190.54 L2204.43 190.54 L2204.43 188.847 L2209.82 180.635 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2220.53 187.441 Q2219.28 187.441 2218.56 188.109 Q2217.84 188.778 2217.84 189.949 Q2217.84 191.121 2218.56 191.79 Q2219.28 192.458 2220.53 192.458 Q2221.78 192.458 2222.5 191.79 Q2223.22 191.113 2223.22 189.949 Q2223.22 188.778 2222.5 188.109 Q2221.78 187.441 2220.53 187.441 M2218.77 186.694 Q2217.64 186.416 2217.01 185.644 Q2216.39 184.871 2216.39 183.76 Q2216.39 182.206 2217.49 181.304 Q2218.6 180.401 2220.53 180.401 Q2222.46 180.401 2223.56 181.304 Q2224.67 182.206 2224.67 183.76 Q2224.67 184.871 2224.03 185.644 Q2223.41 186.416 2222.29 186.694 Q2223.56 186.989 2224.26 187.849 Q2224.97 188.708 2224.97 189.949 Q2224.97 191.833 2223.82 192.84 Q2222.67 193.847 2220.53 193.847 Q2218.38 193.847 2217.23 192.84 Q2216.08 191.833 2216.08 189.949 Q2216.08 188.708 2216.79 187.849 Q2217.5 186.989 2218.77 186.694 M2218.13 183.925 Q2218.13 184.932 2218.75 185.496 Q2219.39 186.061 2220.53 186.061 Q2221.65 186.061 2222.29 185.496 Q2222.93 184.932 2222.93 183.925 Q2222.93 182.918 2222.29 182.354 Q2221.65 181.79 2220.53 181.79 Q2219.39 181.79 2218.75 182.354 Q2218.13 182.918 2218.13 183.925 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2250.09 176.906 L2245.66 183.824 L2250.09 183.824 L2250.09 176.906 M2249.63 175.378 L2251.83 175.378 L2251.83 183.824 L2253.68 183.824 L2253.68 185.283 L2251.83 185.283 L2251.83 188.338 L2250.09 188.338 L2250.09 185.283 L2244.23 185.283 L2244.23 183.59 L2249.63 175.378 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2256.63 188.069 L2256.63 186.472 Q2257.29 186.784 2257.97 186.949 Q2258.64 187.114 2259.3 187.114 Q2261.03 187.114 2261.94 185.951 Q2262.86 184.779 2262.99 182.401 Q2262.49 183.147 2261.72 183.546 Q2260.94 183.946 2260.01 183.946 Q2258.06 183.946 2256.93 182.774 Q2255.8 181.593 2255.8 179.553 Q2255.8 177.557 2256.98 176.35 Q2258.16 175.144 2260.12 175.144 Q2262.37 175.144 2263.55 176.871 Q2264.74 178.59 2264.74 181.871 Q2264.74 184.935 2263.28 186.767 Q2261.83 188.59 2259.37 188.59 Q2258.71 188.59 2258.04 188.46 Q2257.36 188.329 2256.63 188.069 M2260.12 182.574 Q2261.3 182.574 2261.99 181.767 Q2262.68 180.96 2262.68 179.553 Q2262.68 178.156 2261.99 177.349 Q2261.3 176.533 2260.12 176.533 Q2258.94 176.533 2258.25 177.349 Q2257.56 178.156 2257.56 179.553 Q2257.56 180.96 2258.25 181.767 Q2258.94 182.574 2260.12 182.574 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2284.78 170.191 L2291.67 170.191 L2291.67 171.667 L2286.39 171.667 L2286.39 174.844 Q2286.77 174.714 2287.15 174.653 Q2287.54 174.583 2287.92 174.583 Q2290.09 174.583 2291.35 175.773 Q2292.62 176.962 2292.62 178.993 Q2292.62 181.085 2291.32 182.248 Q2290.02 183.403 2287.65 183.403 Q2286.83 183.403 2285.98 183.264 Q2285.14 183.125 2284.24 182.847 L2284.24 181.085 Q2285.02 181.51 2285.85 181.719 Q2286.68 181.927 2287.61 181.927 Q2289.12 181.927 2289.99 181.137 Q2290.87 180.347 2290.87 178.993 Q2290.87 177.639 2289.99 176.849 Q2289.12 176.059 2287.61 176.059 Q2286.91 176.059 2286.21 176.215 Q2285.51 176.372 2284.78 176.701 L2284.78 170.191 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2299.83 171.346 Q2298.47 171.346 2297.79 172.682 Q2297.11 174.01 2297.11 176.684 Q2297.11 179.349 2297.79 180.686 Q2298.47 182.014 2299.83 182.014 Q2301.19 182.014 2301.87 180.686 Q2302.55 179.349 2302.55 176.684 Q2302.55 174.01 2301.87 172.682 Q2301.19 171.346 2299.83 171.346 M2299.83 169.957 Q2302.01 169.957 2303.15 171.684 Q2304.31 173.403 2304.31 176.684 Q2304.31 179.957 2303.15 181.684 Q2302.01 183.403 2299.83 183.403 Q2297.65 183.403 2296.49 181.684 Q2295.35 179.957 2295.35 176.684 Q2295.35 173.403 2296.49 171.684 Q2297.65 169.957 2299.83 169.957 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M286.311 2174.14 L2352.76 2174.14 L2352.76 1362.92 L286.311 1362.92  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip513">
+    <rect x="286" y="1362" width="2067" height="812"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="377.811,2174.14 377.811,1362.92 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="747.114,2174.14 747.114,1362.92 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1116.42,2174.14 1116.42,1362.92 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1485.72,2174.14 1485.72,1362.92 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1855.02,2174.14 1855.02,1362.92 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2224.33,2174.14 2224.33,1362.92 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,2174.14 2352.76,2174.14 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="377.811,2174.14 377.811,2155.24 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="747.114,2174.14 747.114,2155.24 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1116.42,2174.14 1116.42,2155.24 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1485.72,2174.14 1485.72,2155.24 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1855.02,2174.14 1855.02,2155.24 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2224.33,2174.14 2224.33,2155.24 "/>
+<path clip-path="url(#clip510)" d="M377.811 2205.06 Q374.2 2205.06 372.371 2208.62 Q370.566 2212.16 370.566 2219.29 Q370.566 2226.4 372.371 2229.97 Q374.2 2233.51 377.811 2233.51 Q381.445 2233.51 383.251 2229.97 Q385.08 2226.4 385.08 2219.29 Q385.08 2212.16 383.251 2208.62 Q381.445 2205.06 377.811 2205.06 M377.811 2201.35 Q383.621 2201.35 386.677 2205.96 Q389.756 2210.54 389.756 2219.29 Q389.756 2228.02 386.677 2232.63 Q383.621 2237.21 377.811 2237.21 Q372.001 2237.21 368.922 2232.63 Q365.867 2228.02 365.867 2219.29 Q365.867 2210.54 368.922 2205.96 Q372.001 2201.35 377.811 2201.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M721.802 2232.6 L729.441 2232.6 L729.441 2206.24 L721.13 2207.91 L721.13 2203.65 L729.394 2201.98 L734.07 2201.98 L734.07 2232.6 L741.709 2232.6 L741.709 2236.54 L721.802 2236.54 L721.802 2232.6 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M761.153 2205.06 Q757.542 2205.06 755.714 2208.62 Q753.908 2212.16 753.908 2219.29 Q753.908 2226.4 755.714 2229.97 Q757.542 2233.51 761.153 2233.51 Q764.788 2233.51 766.593 2229.97 Q768.422 2226.4 768.422 2219.29 Q768.422 2212.16 766.593 2208.62 Q764.788 2205.06 761.153 2205.06 M761.153 2201.35 Q766.964 2201.35 770.019 2205.96 Q773.098 2210.54 773.098 2219.29 Q773.098 2228.02 770.019 2232.63 Q766.964 2237.21 761.153 2237.21 Q755.343 2237.21 752.265 2232.63 Q749.209 2228.02 749.209 2219.29 Q749.209 2210.54 752.265 2205.96 Q755.343 2201.35 761.153 2201.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1095.19 2232.6 L1111.51 2232.6 L1111.51 2236.54 L1089.57 2236.54 L1089.57 2232.6 Q1092.23 2229.85 1096.81 2225.22 Q1101.42 2220.57 1102.6 2219.23 Q1104.84 2216.7 1105.72 2214.97 Q1106.63 2213.21 1106.63 2211.52 Q1106.63 2208.76 1104.68 2207.03 Q1102.76 2205.29 1099.66 2205.29 Q1097.46 2205.29 1095.01 2206.05 Q1092.57 2206.82 1089.8 2208.37 L1089.8 2203.65 Q1092.62 2202.51 1095.07 2201.93 Q1097.53 2201.35 1099.57 2201.35 Q1104.94 2201.35 1108.13 2204.04 Q1111.32 2206.73 1111.32 2211.22 Q1111.32 2213.35 1110.51 2215.27 Q1109.73 2217.16 1107.62 2219.76 Q1107.04 2220.43 1103.94 2223.65 Q1100.84 2226.84 1095.19 2232.6 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1131.32 2205.06 Q1127.71 2205.06 1125.88 2208.62 Q1124.08 2212.16 1124.08 2219.29 Q1124.08 2226.4 1125.88 2229.97 Q1127.71 2233.51 1131.32 2233.51 Q1134.96 2233.51 1136.76 2229.97 Q1138.59 2226.4 1138.59 2219.29 Q1138.59 2212.16 1136.76 2208.62 Q1134.96 2205.06 1131.32 2205.06 M1131.32 2201.35 Q1137.13 2201.35 1140.19 2205.96 Q1143.27 2210.54 1143.27 2219.29 Q1143.27 2228.02 1140.19 2232.63 Q1137.13 2237.21 1131.32 2237.21 Q1125.51 2237.21 1122.44 2232.63 Q1119.38 2228.02 1119.38 2219.29 Q1119.38 2210.54 1122.44 2205.96 Q1125.51 2201.35 1131.32 2201.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1474.56 2217.91 Q1477.92 2218.62 1479.79 2220.89 Q1481.69 2223.16 1481.69 2226.49 Q1481.69 2231.61 1478.17 2234.41 Q1474.66 2237.21 1468.17 2237.21 Q1466 2237.21 1463.68 2236.77 Q1461.39 2236.35 1458.94 2235.5 L1458.94 2230.98 Q1460.88 2232.12 1463.2 2232.7 Q1465.51 2233.28 1468.03 2233.28 Q1472.43 2233.28 1474.72 2231.54 Q1477.04 2229.8 1477.04 2226.49 Q1477.04 2223.44 1474.89 2221.73 Q1472.76 2219.99 1468.94 2219.99 L1464.91 2219.99 L1464.91 2216.15 L1469.12 2216.15 Q1472.57 2216.15 1474.4 2214.78 Q1476.23 2213.39 1476.23 2210.8 Q1476.23 2208.14 1474.33 2206.73 Q1472.46 2205.29 1468.94 2205.29 Q1467.02 2205.29 1464.82 2205.71 Q1462.62 2206.12 1459.98 2207 L1459.98 2202.84 Q1462.64 2202.1 1464.96 2201.73 Q1467.29 2201.35 1469.35 2201.35 Q1474.68 2201.35 1477.78 2203.79 Q1480.88 2206.19 1480.88 2210.31 Q1480.88 2213.18 1479.24 2215.17 Q1477.59 2217.14 1474.56 2217.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1500.56 2205.06 Q1496.95 2205.06 1495.12 2208.62 Q1493.31 2212.16 1493.31 2219.29 Q1493.31 2226.4 1495.12 2229.97 Q1496.95 2233.51 1500.56 2233.51 Q1504.19 2233.51 1506 2229.97 Q1507.83 2226.4 1507.83 2219.29 Q1507.83 2212.16 1506 2208.62 Q1504.19 2205.06 1500.56 2205.06 M1500.56 2201.35 Q1506.37 2201.35 1509.42 2205.96 Q1512.5 2210.54 1512.5 2219.29 Q1512.5 2228.02 1509.42 2232.63 Q1506.37 2237.21 1500.56 2237.21 Q1494.75 2237.21 1491.67 2232.63 Q1488.61 2228.02 1488.61 2219.29 Q1488.61 2210.54 1491.67 2205.96 Q1494.75 2201.35 1500.56 2201.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1843.19 2206.05 L1831.39 2224.5 L1843.19 2224.5 L1843.19 2206.05 M1841.97 2201.98 L1847.85 2201.98 L1847.85 2224.5 L1852.78 2224.5 L1852.78 2228.39 L1847.85 2228.39 L1847.85 2236.54 L1843.19 2236.54 L1843.19 2228.39 L1827.59 2228.39 L1827.59 2223.88 L1841.97 2201.98 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1870.51 2205.06 Q1866.9 2205.06 1865.07 2208.62 Q1863.26 2212.16 1863.26 2219.29 Q1863.26 2226.4 1865.07 2229.97 Q1866.9 2233.51 1870.51 2233.51 Q1874.14 2233.51 1875.95 2229.97 Q1877.78 2226.4 1877.78 2219.29 Q1877.78 2212.16 1875.95 2208.62 Q1874.14 2205.06 1870.51 2205.06 M1870.51 2201.35 Q1876.32 2201.35 1879.37 2205.96 Q1882.45 2210.54 1882.45 2219.29 Q1882.45 2228.02 1879.37 2232.63 Q1876.32 2237.21 1870.51 2237.21 Q1864.7 2237.21 1861.62 2232.63 Q1858.56 2228.02 1858.56 2219.29 Q1858.56 2210.54 1861.62 2205.96 Q1864.7 2201.35 1870.51 2201.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2199.02 2201.98 L2217.38 2201.98 L2217.38 2205.92 L2203.31 2205.92 L2203.31 2214.39 Q2204.33 2214.04 2205.34 2213.88 Q2206.36 2213.69 2207.38 2213.69 Q2213.17 2213.69 2216.55 2216.86 Q2219.93 2220.04 2219.93 2225.45 Q2219.93 2231.03 2216.46 2234.13 Q2212.98 2237.21 2206.66 2237.21 Q2204.49 2237.21 2202.22 2236.84 Q2199.97 2236.47 2197.57 2235.73 L2197.57 2231.03 Q2199.65 2232.16 2201.87 2232.72 Q2204.09 2233.28 2206.57 2233.28 Q2210.58 2233.28 2212.91 2231.17 Q2215.25 2229.06 2215.25 2225.45 Q2215.25 2221.84 2212.91 2219.73 Q2210.58 2217.63 2206.57 2217.63 Q2204.7 2217.63 2202.82 2218.04 Q2200.97 2218.46 2199.02 2219.34 L2199.02 2201.98 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M2239.14 2205.06 Q2235.53 2205.06 2233.7 2208.62 Q2231.9 2212.16 2231.9 2219.29 Q2231.9 2226.4 2233.7 2229.97 Q2235.53 2233.51 2239.14 2233.51 Q2242.77 2233.51 2244.58 2229.97 Q2246.41 2226.4 2246.41 2219.29 Q2246.41 2212.16 2244.58 2208.62 Q2242.77 2205.06 2239.14 2205.06 M2239.14 2201.35 Q2244.95 2201.35 2248.01 2205.96 Q2251.08 2210.54 2251.08 2219.29 Q2251.08 2228.02 2248.01 2232.63 Q2244.95 2237.21 2239.14 2237.21 Q2233.33 2237.21 2230.25 2232.63 Q2227.2 2228.02 2227.2 2219.29 Q2227.2 2210.54 2230.25 2205.96 Q2233.33 2201.35 2239.14 2201.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1184.33 2273.04 L1184.33 2279.31 Q1180.67 2277.56 1177.42 2276.7 Q1174.17 2275.84 1171.15 2275.84 Q1165.9 2275.84 1163.03 2277.88 Q1160.2 2279.92 1160.2 2283.67 Q1160.2 2286.83 1162.08 2288.45 Q1163.99 2290.04 1169.27 2291.03 L1173.15 2291.82 Q1180.35 2293.19 1183.75 2296.66 Q1187.19 2300.1 1187.19 2305.89 Q1187.19 2312.8 1182.54 2316.36 Q1177.93 2319.93 1168.98 2319.93 Q1165.61 2319.93 1161.79 2319.16 Q1158 2318.4 1153.93 2316.9 L1153.93 2310.28 Q1157.84 2312.48 1161.6 2313.59 Q1165.36 2314.71 1168.98 2314.71 Q1174.49 2314.71 1177.48 2312.54 Q1180.47 2310.38 1180.47 2306.37 Q1180.47 2302.87 1178.31 2300.89 Q1176.18 2298.92 1171.28 2297.93 L1167.36 2297.17 Q1160.17 2295.74 1156.95 2292.68 Q1153.74 2289.63 1153.74 2284.18 Q1153.74 2277.88 1158.16 2274.25 Q1162.62 2270.62 1170.42 2270.62 Q1173.76 2270.62 1177.23 2271.23 Q1180.7 2271.83 1184.33 2273.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1210.78 2287.46 Q1206.06 2287.46 1203.33 2291.15 Q1200.59 2294.81 1200.59 2301.21 Q1200.59 2307.61 1203.3 2311.3 Q1206.03 2314.96 1210.78 2314.96 Q1215.45 2314.96 1218.19 2311.27 Q1220.93 2307.58 1220.93 2301.21 Q1220.93 2294.88 1218.19 2291.19 Q1215.45 2287.46 1210.78 2287.46 M1210.78 2282.5 Q1218.41 2282.5 1222.77 2287.46 Q1227.14 2292.43 1227.14 2301.21 Q1227.14 2309.96 1222.77 2314.96 Q1218.41 2319.93 1210.78 2319.93 Q1203.1 2319.93 1198.74 2314.96 Q1194.42 2309.96 1194.42 2301.21 Q1194.42 2292.43 1198.74 2287.46 Q1203.1 2282.5 1210.78 2282.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1236.84 2269.48 L1242.7 2269.48 L1242.7 2319 L1236.84 2319 L1236.84 2269.48 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1254.35 2304.94 L1254.35 2283.36 L1260.21 2283.36 L1260.21 2304.71 Q1260.21 2309.77 1262.18 2312.32 Q1264.15 2314.83 1268.1 2314.83 Q1272.84 2314.83 1275.58 2311.81 Q1278.35 2308.79 1278.35 2303.57 L1278.35 2283.36 L1284.2 2283.36 L1284.2 2319 L1278.35 2319 L1278.35 2313.53 Q1276.21 2316.78 1273.38 2318.37 Q1270.58 2319.93 1266.86 2319.93 Q1260.71 2319.93 1257.53 2316.11 Q1254.35 2312.29 1254.35 2304.94 M1269.09 2282.5 L1269.09 2282.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1302.06 2273.23 L1302.06 2283.36 L1314.12 2283.36 L1314.12 2287.91 L1302.06 2287.91 L1302.06 2307.26 Q1302.06 2311.62 1303.24 2312.86 Q1304.45 2314.1 1308.11 2314.1 L1314.12 2314.1 L1314.12 2319 L1308.11 2319 Q1301.33 2319 1298.75 2316.49 Q1296.17 2313.94 1296.17 2307.26 L1296.17 2287.91 L1291.87 2287.91 L1291.87 2283.36 L1296.17 2283.36 L1296.17 2273.23 L1302.06 2273.23 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1321.83 2283.36 L1327.68 2283.36 L1327.68 2319 L1321.83 2319 L1321.83 2283.36 M1321.83 2269.48 L1327.68 2269.48 L1327.68 2276.89 L1321.83 2276.89 L1321.83 2269.48 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1353.75 2287.46 Q1349.04 2287.46 1346.3 2291.15 Q1343.56 2294.81 1343.56 2301.21 Q1343.56 2307.61 1346.27 2311.3 Q1349.01 2314.96 1353.75 2314.96 Q1358.43 2314.96 1361.17 2311.27 Q1363.9 2307.58 1363.9 2301.21 Q1363.9 2294.88 1361.17 2291.19 Q1358.43 2287.46 1353.75 2287.46 M1353.75 2282.5 Q1361.39 2282.5 1365.75 2287.46 Q1370.11 2292.43 1370.11 2301.21 Q1370.11 2309.96 1365.75 2314.96 Q1361.39 2319.93 1353.75 2319.93 Q1346.08 2319.93 1341.72 2314.96 Q1337.39 2309.96 1337.39 2301.21 Q1337.39 2292.43 1341.72 2287.46 Q1346.08 2282.5 1353.75 2282.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1409.45 2297.49 L1409.45 2319 L1403.59 2319 L1403.59 2297.68 Q1403.59 2292.62 1401.62 2290.1 Q1399.65 2287.59 1395.7 2287.59 Q1390.96 2287.59 1388.22 2290.61 Q1385.48 2293.64 1385.48 2298.86 L1385.48 2319 L1379.59 2319 L1379.59 2283.36 L1385.48 2283.36 L1385.48 2288.89 Q1387.58 2285.68 1390.42 2284.09 Q1393.28 2282.5 1397 2282.5 Q1403.15 2282.5 1406.3 2286.32 Q1409.45 2290.1 1409.45 2297.49 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1469.03 2290.33 L1459.77 2290.33 L1457.1 2300.96 L1466.42 2300.96 L1469.03 2290.33 M1464.26 2272.22 L1460.95 2285.42 L1470.24 2285.42 L1473.58 2272.22 L1478.68 2272.22 L1475.4 2285.42 L1485.33 2285.42 L1485.33 2290.33 L1474.16 2290.33 L1471.55 2300.96 L1481.67 2300.96 L1481.67 2305.83 L1470.31 2305.83 L1467 2319 L1461.9 2319 L1465.18 2305.83 L1455.86 2305.83 L1452.58 2319 L1447.45 2319 L1450.76 2305.83 L1440.74 2305.83 L1440.74 2300.96 L1451.94 2300.96 L1454.61 2290.33 L1444.37 2290.33 L1444.37 2285.42 L1455.86 2285.42 L1459.1 2272.22 L1464.26 2272.22 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,2174.14 2352.76,2174.14 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,1971.34 2352.76,1971.34 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,1768.53 2352.76,1768.53 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,1565.73 2352.76,1565.73 "/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="286.311,1362.92 2352.76,1362.92 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,2174.14 286.311,1362.92 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,2174.14 305.209,2174.14 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,1971.34 305.209,1971.34 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,1768.53 305.209,1768.53 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,1565.73 305.209,1565.73 "/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="286.311,1362.92 305.209,1362.92 "/>
+<path clip-path="url(#clip510)" d="M238.367 2159.94 Q234.756 2159.94 232.927 2163.5 Q231.122 2167.04 231.122 2174.17 Q231.122 2181.28 232.927 2184.85 Q234.756 2188.39 238.367 2188.39 Q242.001 2188.39 243.807 2184.85 Q245.635 2181.28 245.635 2174.17 Q245.635 2167.04 243.807 2163.5 Q242.001 2159.94 238.367 2159.94 M238.367 2156.23 Q244.177 2156.23 247.233 2160.84 Q250.311 2165.42 250.311 2174.17 Q250.311 2182.9 247.233 2187.51 Q244.177 2192.09 238.367 2192.09 Q232.557 2192.09 229.478 2187.51 Q226.422 2182.9 226.422 2174.17 Q226.422 2165.42 229.478 2160.84 Q232.557 2156.23 238.367 2156.23 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M172.071 1984.68 L188.39 1984.68 L188.39 1988.62 L166.446 1988.62 L166.446 1984.68 Q169.108 1981.93 173.691 1977.3 Q178.298 1972.64 179.478 1971.3 Q181.724 1968.78 182.603 1967.04 Q183.506 1965.28 183.506 1963.59 Q183.506 1960.84 181.562 1959.1 Q179.64 1957.37 176.538 1957.37 Q174.339 1957.37 171.886 1958.13 Q169.455 1958.89 166.677 1960.44 L166.677 1955.72 Q169.501 1954.59 171.955 1954.01 Q174.409 1953.43 176.446 1953.43 Q181.816 1953.43 185.011 1956.12 Q188.205 1958.8 188.205 1963.29 Q188.205 1965.42 187.395 1967.34 Q186.608 1969.24 184.501 1971.83 Q183.923 1972.5 180.821 1975.72 Q177.719 1978.92 172.071 1984.68 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M198.251 1954.06 L216.608 1954.06 L216.608 1957.99 L202.534 1957.99 L202.534 1966.46 Q203.552 1966.12 204.571 1965.95 Q205.589 1965.77 206.608 1965.77 Q212.395 1965.77 215.774 1968.94 Q219.154 1972.11 219.154 1977.53 Q219.154 1983.11 215.682 1986.21 Q212.21 1989.29 205.89 1989.29 Q203.714 1989.29 201.446 1988.92 Q199.2 1988.55 196.793 1987.81 L196.793 1983.11 Q198.876 1984.24 201.099 1984.8 Q203.321 1985.35 205.798 1985.35 Q209.802 1985.35 212.14 1983.25 Q214.478 1981.14 214.478 1977.53 Q214.478 1973.92 212.14 1971.81 Q209.802 1969.7 205.798 1969.7 Q203.923 1969.7 202.048 1970.12 Q200.196 1970.54 198.251 1971.42 L198.251 1954.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M238.367 1957.13 Q234.756 1957.13 232.927 1960.7 Q231.122 1964.24 231.122 1971.37 Q231.122 1978.48 232.927 1982.04 Q234.756 1985.58 238.367 1985.58 Q242.001 1985.58 243.807 1982.04 Q245.635 1978.48 245.635 1971.37 Q245.635 1964.24 243.807 1960.7 Q242.001 1957.13 238.367 1957.13 M238.367 1953.43 Q244.177 1953.43 247.233 1958.04 Q250.311 1962.62 250.311 1971.37 Q250.311 1980.1 247.233 1984.7 Q244.177 1989.29 238.367 1989.29 Q232.557 1989.29 229.478 1984.7 Q226.422 1980.1 226.422 1971.37 Q226.422 1962.62 229.478 1958.04 Q232.557 1953.43 238.367 1953.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M168.089 1751.25 L186.446 1751.25 L186.446 1755.19 L172.372 1755.19 L172.372 1763.66 Q173.39 1763.31 174.409 1763.15 Q175.427 1762.97 176.446 1762.97 Q182.233 1762.97 185.612 1766.14 Q188.992 1769.31 188.992 1774.72 Q188.992 1780.3 185.52 1783.4 Q182.048 1786.48 175.728 1786.48 Q173.552 1786.48 171.284 1786.11 Q169.039 1785.74 166.631 1785 L166.631 1780.3 Q168.714 1781.44 170.937 1781.99 Q173.159 1782.55 175.636 1782.55 Q179.64 1782.55 181.978 1780.44 Q184.316 1778.34 184.316 1774.72 Q184.316 1771.11 181.978 1769.01 Q179.64 1766.9 175.636 1766.9 Q173.761 1766.9 171.886 1767.32 Q170.034 1767.73 168.089 1768.61 L168.089 1751.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M208.205 1754.33 Q204.594 1754.33 202.765 1757.9 Q200.96 1761.44 200.96 1768.57 Q200.96 1775.67 202.765 1779.24 Q204.594 1782.78 208.205 1782.78 Q211.839 1782.78 213.645 1779.24 Q215.473 1775.67 215.473 1768.57 Q215.473 1761.44 213.645 1757.9 Q211.839 1754.33 208.205 1754.33 M208.205 1750.63 Q214.015 1750.63 217.071 1755.23 Q220.149 1759.82 220.149 1768.57 Q220.149 1777.29 217.071 1781.9 Q214.015 1786.48 208.205 1786.48 Q202.395 1786.48 199.316 1781.9 Q196.261 1777.29 196.261 1768.57 Q196.261 1759.82 199.316 1755.23 Q202.395 1750.63 208.205 1750.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M238.367 1754.33 Q234.756 1754.33 232.927 1757.9 Q231.122 1761.44 231.122 1768.57 Q231.122 1775.67 232.927 1779.24 Q234.756 1782.78 238.367 1782.78 Q242.001 1782.78 243.807 1779.24 Q245.635 1775.67 245.635 1768.57 Q245.635 1761.44 243.807 1757.9 Q242.001 1754.33 238.367 1754.33 M238.367 1750.63 Q244.177 1750.63 247.233 1755.23 Q250.311 1759.82 250.311 1768.57 Q250.311 1777.29 247.233 1781.9 Q244.177 1786.48 238.367 1786.48 Q232.557 1786.48 229.478 1781.9 Q226.422 1777.29 226.422 1768.57 Q226.422 1759.82 229.478 1755.23 Q232.557 1750.63 238.367 1750.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M166.863 1548.45 L189.085 1548.45 L189.085 1550.44 L176.538 1583.01 L171.654 1583.01 L183.46 1552.38 L166.863 1552.38 L166.863 1548.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M198.251 1548.45 L216.608 1548.45 L216.608 1552.38 L202.534 1552.38 L202.534 1560.86 Q203.552 1560.51 204.571 1560.35 Q205.589 1560.16 206.608 1560.16 Q212.395 1560.16 215.774 1563.33 Q219.154 1566.5 219.154 1571.92 Q219.154 1577.5 215.682 1580.6 Q212.21 1583.68 205.89 1583.68 Q203.714 1583.68 201.446 1583.31 Q199.2 1582.94 196.793 1582.2 L196.793 1577.5 Q198.876 1578.63 201.099 1579.19 Q203.321 1579.74 205.798 1579.74 Q209.802 1579.74 212.14 1577.64 Q214.478 1575.53 214.478 1571.92 Q214.478 1568.31 212.14 1566.2 Q209.802 1564.1 205.798 1564.1 Q203.923 1564.1 202.048 1564.51 Q200.196 1564.93 198.251 1565.81 L198.251 1548.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M238.367 1551.53 Q234.756 1551.53 232.927 1555.09 Q231.122 1558.63 231.122 1565.76 Q231.122 1572.87 232.927 1576.43 Q234.756 1579.98 238.367 1579.98 Q242.001 1579.98 243.807 1576.43 Q245.635 1572.87 245.635 1565.76 Q245.635 1558.63 243.807 1555.09 Q242.001 1551.53 238.367 1551.53 M238.367 1547.82 Q244.177 1547.82 247.233 1552.43 Q250.311 1557.01 250.311 1565.76 Q250.311 1574.49 247.233 1579.1 Q244.177 1583.68 238.367 1583.68 Q232.557 1583.68 229.478 1579.1 Q226.422 1574.49 226.422 1565.76 Q226.422 1557.01 229.478 1552.43 Q232.557 1547.82 238.367 1547.82 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M138.691 1376.27 L146.33 1376.27 L146.33 1349.9 L138.02 1351.57 L138.02 1347.31 L146.284 1345.64 L150.96 1345.64 L150.96 1376.27 L158.599 1376.27 L158.599 1380.2 L138.691 1380.2 L138.691 1376.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M178.043 1348.72 Q174.432 1348.72 172.603 1352.29 Q170.798 1355.83 170.798 1362.96 Q170.798 1370.07 172.603 1373.63 Q174.432 1377.17 178.043 1377.17 Q181.677 1377.17 183.483 1373.63 Q185.312 1370.07 185.312 1362.96 Q185.312 1355.83 183.483 1352.29 Q181.677 1348.72 178.043 1348.72 M178.043 1345.02 Q183.853 1345.02 186.909 1349.63 Q189.987 1354.21 189.987 1362.96 Q189.987 1371.69 186.909 1376.29 Q183.853 1380.88 178.043 1380.88 Q172.233 1380.88 169.154 1376.29 Q166.099 1371.69 166.099 1362.96 Q166.099 1354.21 169.154 1349.63 Q172.233 1345.02 178.043 1345.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M208.205 1348.72 Q204.594 1348.72 202.765 1352.29 Q200.96 1355.83 200.96 1362.96 Q200.96 1370.07 202.765 1373.63 Q204.594 1377.17 208.205 1377.17 Q211.839 1377.17 213.645 1373.63 Q215.473 1370.07 215.473 1362.96 Q215.473 1355.83 213.645 1352.29 Q211.839 1348.72 208.205 1348.72 M208.205 1345.02 Q214.015 1345.02 217.071 1349.63 Q220.149 1354.21 220.149 1362.96 Q220.149 1371.69 217.071 1376.29 Q214.015 1380.88 208.205 1380.88 Q202.395 1380.88 199.316 1376.29 Q196.261 1371.69 196.261 1362.96 Q196.261 1354.21 199.316 1349.63 Q202.395 1345.02 208.205 1345.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M238.367 1348.72 Q234.756 1348.72 232.927 1352.29 Q231.122 1355.83 231.122 1362.96 Q231.122 1370.07 232.927 1373.63 Q234.756 1377.17 238.367 1377.17 Q242.001 1377.17 243.807 1373.63 Q245.635 1370.07 245.635 1362.96 Q245.635 1355.83 243.807 1352.29 Q242.001 1348.72 238.367 1348.72 M238.367 1345.02 Q244.177 1345.02 247.233 1349.63 Q250.311 1354.21 250.311 1362.96 Q250.311 1371.69 247.233 1376.29 Q244.177 1380.88 238.367 1380.88 Q232.557 1380.88 229.478 1376.29 Q226.422 1371.69 226.422 1362.96 Q226.422 1354.21 229.478 1349.63 Q232.557 1345.02 238.367 1345.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M40.2442 2003.67 L40.2442 1997.24 L87.7642 1997.24 L87.7642 2003.67 L40.2442 2003.67 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M66.2481 1955.06 L87.7642 1955.06 L87.7642 1960.92 L66.439 1960.92 Q61.3783 1960.92 58.8638 1962.89 Q56.3494 1964.87 56.3494 1968.81 Q56.3494 1973.56 59.3731 1976.29 Q62.3968 1979.03 67.6167 1979.03 L87.7642 1979.03 L87.7642 1984.92 L52.1162 1984.92 L52.1162 1979.03 L57.6544 1979.03 Q54.4397 1976.93 52.8483 1974.1 Q51.2568 1971.23 51.2568 1967.51 Q51.2568 1961.37 55.0763 1958.21 Q58.8638 1955.06 66.2481 1955.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M52.1162 1947.58 L52.1162 1941.38 L82.035 1930.24 L52.1162 1919.1 L52.1162 1912.89 L87.7642 1926.26 L87.7642 1934.22 L52.1162 1947.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M68.4761 1874.31 L71.3406 1874.31 L71.3406 1901.24 Q77.3881 1900.86 80.5709 1897.61 Q83.7219 1894.33 83.7219 1888.51 Q83.7219 1885.14 82.8944 1881.98 Q82.0669 1878.8 80.4118 1875.68 L85.9499 1875.68 Q87.2867 1878.83 87.987 1882.14 Q88.6872 1885.45 88.6872 1888.86 Q88.6872 1897.39 83.7219 1902.39 Q78.7567 1907.35 70.2903 1907.35 Q61.5374 1907.35 56.4131 1902.64 Q51.2568 1897.9 51.2568 1889.88 Q51.2568 1882.69 55.9038 1878.52 Q60.5189 1874.31 68.4761 1874.31 M66.7573 1880.17 Q61.9512 1880.23 59.0866 1882.88 Q56.2221 1885.49 56.2221 1889.81 Q56.2221 1894.72 58.9912 1897.68 Q61.7602 1900.6 66.7892 1901.05 L66.7573 1880.17 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M53.1665 1841.98 L58.7047 1841.98 Q57.4316 1844.46 56.795 1847.13 Q56.1584 1849.81 56.1584 1852.67 Q56.1584 1857.03 57.4952 1859.23 Q58.832 1861.39 61.5056 1861.39 Q63.5426 1861.39 64.7203 1859.83 Q65.8661 1858.27 66.9165 1853.56 L67.3621 1851.56 Q68.6989 1845.32 71.1497 1842.71 Q73.5686 1840.07 77.9291 1840.07 Q82.8944 1840.07 85.7908 1844.01 Q88.6872 1847.93 88.6872 1854.8 Q88.6872 1857.67 88.1143 1860.79 Q87.5732 1863.87 86.4592 1867.31 L80.4118 1867.31 Q82.0987 1864.07 82.958 1860.91 Q83.7856 1857.76 83.7856 1854.68 Q83.7856 1850.54 82.3851 1848.31 Q80.9529 1846.08 78.3747 1846.08 Q75.9876 1846.08 74.7145 1847.71 Q73.4413 1849.3 72.2637 1854.74 L71.7862 1856.78 Q70.6404 1862.22 68.2851 1864.64 Q65.898 1867.06 61.7602 1867.06 Q56.7313 1867.06 53.9941 1863.49 Q51.2568 1859.93 51.2568 1853.37 Q51.2568 1850.12 51.7343 1847.26 Q52.2117 1844.4 53.1665 1841.98 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M41.9947 1824.95 L52.1162 1824.95 L52.1162 1812.89 L56.6677 1812.89 L56.6677 1824.95 L76.0194 1824.95 Q80.3799 1824.95 81.6213 1823.77 Q82.8626 1822.56 82.8626 1818.9 L82.8626 1812.89 L87.7642 1812.89 L87.7642 1818.9 Q87.7642 1825.68 85.2497 1828.26 Q82.7034 1830.84 76.0194 1830.84 L56.6677 1830.84 L56.6677 1835.13 L52.1162 1835.13 L52.1162 1830.84 L41.9947 1830.84 L41.9947 1824.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M58.9593 1777.43 Q55.0126 1775.23 53.1347 1772.18 Q51.2568 1769.12 51.2568 1764.98 Q51.2568 1759.41 55.1717 1756.39 Q59.0548 1753.37 66.2481 1753.37 L87.7642 1753.37 L87.7642 1759.25 L66.439 1759.25 Q61.3146 1759.25 58.832 1761.07 Q56.3494 1762.88 56.3494 1766.61 Q56.3494 1771.16 59.3731 1773.8 Q62.3968 1776.44 67.6167 1776.44 L87.7642 1776.44 L87.7642 1782.33 L66.439 1782.33 Q61.2828 1782.33 58.832 1784.14 Q56.3494 1785.96 56.3494 1789.75 Q56.3494 1794.23 59.4049 1796.88 Q62.4286 1799.52 67.6167 1799.52 L87.7642 1799.52 L87.7642 1805.41 L52.1162 1805.41 L52.1162 1799.52 L57.6544 1799.52 Q54.376 1797.51 52.8164 1794.71 Q51.2568 1791.91 51.2568 1788.06 Q51.2568 1784.18 53.2302 1781.47 Q55.2036 1778.73 58.9593 1777.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M68.4761 1711.19 L71.3406 1711.19 L71.3406 1738.12 Q77.3881 1737.74 80.5709 1734.49 Q83.7219 1731.21 83.7219 1725.39 Q83.7219 1722.01 82.8944 1718.86 Q82.0669 1715.68 80.4118 1712.56 L85.9499 1712.56 Q87.2867 1715.71 87.987 1719.02 Q88.6872 1722.33 88.6872 1725.74 Q88.6872 1734.27 83.7219 1739.27 Q78.7567 1744.23 70.2903 1744.23 Q61.5374 1744.23 56.4131 1739.52 Q51.2568 1734.78 51.2568 1726.76 Q51.2568 1719.56 55.9038 1715.39 Q60.5189 1711.19 68.4761 1711.19 M66.7573 1717.05 Q61.9512 1717.11 59.0866 1719.75 Q56.2221 1722.36 56.2221 1726.69 Q56.2221 1731.6 58.9912 1734.56 Q61.7602 1737.48 66.7892 1737.93 L66.7573 1717.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M66.2481 1671.95 L87.7642 1671.95 L87.7642 1677.8 L66.439 1677.8 Q61.3783 1677.8 58.8638 1679.78 Q56.3494 1681.75 56.3494 1685.7 Q56.3494 1690.44 59.3731 1693.18 Q62.3968 1695.92 67.6167 1695.92 L87.7642 1695.92 L87.7642 1701.8 L52.1162 1701.8 L52.1162 1695.92 L57.6544 1695.92 Q54.4397 1693.81 52.8483 1690.98 Q51.2568 1688.12 51.2568 1684.39 Q51.2568 1678.25 55.0763 1675.1 Q58.8638 1671.95 66.2481 1671.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M41.9947 1654.47 L52.1162 1654.47 L52.1162 1642.41 L56.6677 1642.41 L56.6677 1654.47 L76.0194 1654.47 Q80.3799 1654.47 81.6213 1653.3 Q82.8626 1652.09 82.8626 1648.43 L82.8626 1642.41 L87.7642 1642.41 L87.7642 1648.43 Q87.7642 1655.21 85.2497 1657.78 Q82.7034 1660.36 76.0194 1660.36 L56.6677 1660.36 L56.6677 1664.66 L52.1162 1664.66 L52.1162 1660.36 L41.9947 1660.36 L41.9947 1654.47 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M38.3026 1599.92 Q45.6232 1604.19 52.7846 1606.25 Q59.946 1608.32 67.2984 1608.32 Q74.6508 1608.32 81.8759 1606.25 Q89.0691 1604.15 96.3579 1599.92 L96.3579 1605.01 Q88.8782 1609.79 81.6531 1612.17 Q74.428 1614.53 67.2984 1614.53 Q60.2006 1614.53 53.0074 1612.17 Q45.8142 1609.82 38.3026 1605.01 L38.3026 1599.92 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M97.3445 1572.68 L97.3445 1575.86 L87.7642 1575.89 Q87.7005 1579.23 87.0003 1582.57 Q86.2682 1585.92 84.8359 1589.29 L79.1068 1589.29 Q81.1438 1586.04 82.1942 1582.73 Q83.2127 1579.39 83.2445 1575.86 L68.7307 1575.86 Q67.5849 1582.89 64.8476 1586.11 Q62.1104 1589.29 57.3361 1589.29 Q52.148 1589.29 49.1561 1585.82 Q46.1643 1582.35 45.7187 1575.86 L38.239 1575.86 L38.239 1572.68 L45.6232 1572.68 Q45.7505 1569.72 46.2598 1566.95 Q46.7372 1564.18 47.5965 1561.54 L53.1665 1561.54 Q51.8297 1564.18 51.0977 1566.98 Q50.3656 1569.75 50.2383 1572.68 L63.8291 1572.68 Q64.9431 1565.45 67.8077 1562.04 Q70.6722 1558.64 75.6375 1558.64 Q81.0165 1558.64 84.1357 1562.27 Q87.2231 1565.86 87.7005 1572.68 L97.3445 1572.68 M63.2562 1575.86 L50.2065 1575.86 Q50.6203 1579.55 52.3072 1581.49 Q53.9941 1583.43 56.795 1583.43 Q59.5322 1583.43 61.06 1581.65 Q62.5878 1579.84 63.2562 1575.86 M69.3673 1572.68 L83.149 1572.68 Q82.6079 1568.63 80.8574 1566.6 Q79.1068 1564.53 76.2422 1564.53 Q73.4413 1564.53 71.7862 1566.5 Q70.1312 1568.44 69.3673 1572.68 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M38.3026 1548.01 L38.3026 1542.92 Q45.8142 1538.14 53.0074 1535.79 Q60.2006 1533.4 67.2984 1533.4 Q74.428 1533.4 81.6531 1535.79 Q88.8782 1538.14 96.3579 1542.92 L96.3579 1548.01 Q89.0691 1543.77 81.8759 1541.71 Q74.6508 1539.61 67.2984 1539.61 Q59.946 1539.61 52.7846 1541.71 Q45.6232 1543.77 38.3026 1548.01 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1022.52 1228.03 L1022.52 1275.06 L1032.41 1275.06 Q1044.92 1275.06 1050.72 1269.39 Q1056.55 1263.72 1056.55 1251.49 Q1056.55 1239.33 1050.72 1233.7 Q1044.92 1228.03 1032.41 1228.03 L1022.52 1228.03 M1014.34 1221.31 L1031.15 1221.31 Q1048.73 1221.31 1056.95 1228.64 Q1065.18 1235.93 1065.18 1251.49 Q1065.18 1267.12 1056.91 1274.46 Q1048.65 1281.79 1031.15 1281.79 L1014.34 1281.79 L1014.34 1221.31 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1116.7 1257.24 L1116.7 1260.89 L1082.43 1260.89 Q1082.92 1268.58 1087.05 1272.63 Q1091.22 1276.64 1098.64 1276.64 Q1102.93 1276.64 1106.94 1275.59 Q1110.99 1274.54 1114.96 1272.43 L1114.96 1279.48 Q1110.95 1281.18 1106.74 1282.07 Q1102.53 1282.96 1098.19 1282.96 Q1087.34 1282.96 1080.98 1276.64 Q1074.66 1270.32 1074.66 1259.55 Q1074.66 1248.41 1080.65 1241.89 Q1086.69 1235.32 1096.9 1235.32 Q1106.05 1235.32 1111.36 1241.24 Q1116.7 1247.11 1116.7 1257.24 M1109.25 1255.05 Q1109.17 1248.94 1105.81 1245.29 Q1102.49 1241.64 1096.98 1241.64 Q1090.74 1241.64 1086.97 1245.17 Q1083.24 1248.69 1082.68 1255.09 L1109.25 1255.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1161.59 1238.16 L1161.59 1245.13 Q1158.43 1243.39 1155.23 1242.54 Q1152.07 1241.64 1148.83 1241.64 Q1141.58 1241.64 1137.57 1246.26 Q1133.56 1250.84 1133.56 1259.14 Q1133.56 1267.45 1137.57 1272.07 Q1141.58 1276.64 1148.83 1276.64 Q1152.07 1276.64 1155.23 1275.79 Q1158.43 1274.9 1161.59 1273.16 L1161.59 1280.05 Q1158.47 1281.5 1155.11 1282.23 Q1151.79 1282.96 1148.02 1282.96 Q1137.77 1282.96 1131.73 1276.52 Q1125.7 1270.08 1125.7 1259.14 Q1125.7 1248.04 1131.77 1241.68 Q1137.89 1235.32 1148.5 1235.32 Q1151.95 1235.32 1155.23 1236.05 Q1158.51 1236.74 1161.59 1238.16 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1174.55 1236.42 L1182.01 1236.42 L1182.01 1281.79 L1174.55 1281.79 L1174.55 1236.42 M1174.55 1218.76 L1182.01 1218.76 L1182.01 1228.19 L1174.55 1228.19 L1174.55 1218.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1226.52 1237.76 L1226.52 1244.8 Q1223.37 1243.18 1219.96 1242.37 Q1216.56 1241.56 1212.91 1241.56 Q1207.36 1241.56 1204.57 1243.26 Q1201.81 1244.97 1201.81 1248.37 Q1201.81 1250.96 1203.8 1252.46 Q1205.78 1253.92 1211.78 1255.25 L1214.33 1255.82 Q1222.27 1257.52 1225.59 1260.64 Q1228.96 1263.72 1228.96 1269.27 Q1228.96 1275.59 1223.93 1279.28 Q1218.95 1282.96 1210.2 1282.96 Q1206.55 1282.96 1202.58 1282.23 Q1198.65 1281.55 1194.28 1280.13 L1194.28 1272.43 Q1198.41 1274.58 1202.42 1275.67 Q1206.43 1276.72 1210.36 1276.72 Q1215.63 1276.72 1218.46 1274.94 Q1221.3 1273.12 1221.3 1269.84 Q1221.3 1266.8 1219.23 1265.18 Q1217.21 1263.56 1210.28 1262.06 L1207.69 1261.45 Q1200.76 1259.99 1197.68 1257 Q1194.6 1253.96 1194.6 1248.69 Q1194.6 1242.29 1199.14 1238.81 Q1203.68 1235.32 1212.02 1235.32 Q1216.15 1235.32 1219.8 1235.93 Q1223.45 1236.54 1226.52 1237.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1240.82 1236.42 L1248.28 1236.42 L1248.28 1281.79 L1240.82 1281.79 L1240.82 1236.42 M1240.82 1218.76 L1248.28 1218.76 L1248.28 1228.19 L1240.82 1228.19 L1240.82 1218.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1281.46 1241.64 Q1275.46 1241.64 1271.98 1246.34 Q1268.49 1251 1268.49 1259.14 Q1268.49 1267.29 1271.94 1271.99 Q1275.42 1276.64 1281.46 1276.64 Q1287.41 1276.64 1290.89 1271.94 Q1294.38 1267.25 1294.38 1259.14 Q1294.38 1251.08 1290.89 1246.38 Q1287.41 1241.64 1281.46 1241.64 M1281.46 1235.32 Q1291.18 1235.32 1296.73 1241.64 Q1302.28 1247.96 1302.28 1259.14 Q1302.28 1270.28 1296.73 1276.64 Q1291.18 1282.96 1281.46 1282.96 Q1271.69 1282.96 1266.14 1276.64 Q1260.63 1270.28 1260.63 1259.14 Q1260.63 1247.96 1266.14 1241.64 Q1271.69 1235.32 1281.46 1235.32 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1352.35 1254.4 L1352.35 1281.79 L1344.89 1281.79 L1344.89 1254.65 Q1344.89 1248.21 1342.38 1245.01 Q1339.87 1241.81 1334.85 1241.81 Q1328.81 1241.81 1325.33 1245.65 Q1321.84 1249.5 1321.84 1256.15 L1321.84 1281.79 L1314.35 1281.79 L1314.35 1236.42 L1321.84 1236.42 L1321.84 1243.47 Q1324.52 1239.38 1328.12 1237.35 Q1331.77 1235.32 1336.51 1235.32 Q1344.33 1235.32 1348.34 1240.19 Q1352.35 1245.01 1352.35 1254.4 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1422.51 1237.76 L1422.51 1244.8 Q1419.35 1243.18 1415.95 1242.37 Q1412.54 1241.56 1408.9 1241.56 Q1403.35 1241.56 1400.55 1243.26 Q1397.8 1244.97 1397.8 1248.37 Q1397.8 1250.96 1399.78 1252.46 Q1401.77 1253.92 1407.76 1255.25 L1410.31 1255.82 Q1418.25 1257.52 1421.58 1260.64 Q1424.94 1263.72 1424.94 1269.27 Q1424.94 1275.59 1419.91 1279.28 Q1414.93 1282.96 1406.18 1282.96 Q1402.54 1282.96 1398.57 1282.23 Q1394.64 1281.55 1390.26 1280.13 L1390.26 1272.43 Q1394.39 1274.58 1398.4 1275.67 Q1402.42 1276.72 1406.34 1276.72 Q1411.61 1276.72 1414.45 1274.94 Q1417.28 1273.12 1417.28 1269.84 Q1417.28 1266.8 1415.22 1265.18 Q1413.19 1263.56 1406.26 1262.06 L1403.67 1261.45 Q1396.74 1259.99 1393.67 1257 Q1390.59 1253.96 1390.59 1248.69 Q1390.59 1242.29 1395.12 1238.81 Q1399.66 1235.32 1408.01 1235.32 Q1412.14 1235.32 1415.78 1235.93 Q1419.43 1236.54 1422.51 1237.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1444.02 1274.98 L1444.02 1299.05 L1436.52 1299.05 L1436.52 1236.42 L1444.02 1236.42 L1444.02 1243.3 Q1446.37 1239.25 1449.93 1237.31 Q1453.54 1235.32 1458.52 1235.32 Q1466.78 1235.32 1471.93 1241.89 Q1477.11 1248.45 1477.11 1259.14 Q1477.11 1269.84 1471.93 1276.4 Q1466.78 1282.96 1458.52 1282.96 Q1453.54 1282.96 1449.93 1281.02 Q1446.37 1279.03 1444.02 1274.98 M1469.38 1259.14 Q1469.38 1250.92 1465.97 1246.26 Q1462.61 1241.56 1456.7 1241.56 Q1450.78 1241.56 1447.38 1246.26 Q1444.02 1250.92 1444.02 1259.14 Q1444.02 1267.37 1447.38 1272.07 Q1450.78 1276.72 1456.7 1276.72 Q1462.61 1276.72 1465.97 1272.07 Q1469.38 1267.37 1469.38 1259.14 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1510.09 1258.98 Q1501.05 1258.98 1497.57 1261.05 Q1494.09 1263.11 1494.09 1268.1 Q1494.09 1272.07 1496.68 1274.42 Q1499.31 1276.72 1503.81 1276.72 Q1510.01 1276.72 1513.73 1272.35 Q1517.5 1267.93 1517.5 1260.64 L1517.5 1258.98 L1510.09 1258.98 M1524.95 1255.9 L1524.95 1281.79 L1517.5 1281.79 L1517.5 1274.9 Q1514.95 1279.03 1511.14 1281.02 Q1507.33 1282.96 1501.82 1282.96 Q1494.86 1282.96 1490.72 1279.07 Q1486.63 1275.14 1486.63 1268.58 Q1486.63 1260.93 1491.74 1257.04 Q1496.88 1253.15 1507.05 1253.15 L1517.5 1253.15 L1517.5 1252.42 Q1517.5 1247.27 1514.1 1244.48 Q1510.74 1241.64 1504.62 1241.64 Q1500.73 1241.64 1497.04 1242.58 Q1493.36 1243.51 1489.96 1245.37 L1489.96 1238.48 Q1494.05 1236.9 1497.89 1236.13 Q1501.74 1235.32 1505.39 1235.32 Q1515.23 1235.32 1520.09 1240.43 Q1524.95 1245.53 1524.95 1255.9 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1572.96 1238.16 L1572.96 1245.13 Q1569.8 1243.39 1566.6 1242.54 Q1563.44 1241.64 1560.2 1241.64 Q1552.95 1241.64 1548.94 1246.26 Q1544.93 1250.84 1544.93 1259.14 Q1544.93 1267.45 1548.94 1272.07 Q1552.95 1276.64 1560.2 1276.64 Q1563.44 1276.64 1566.6 1275.79 Q1569.8 1274.9 1572.96 1273.16 L1572.96 1280.05 Q1569.84 1281.5 1566.48 1282.23 Q1563.15 1282.96 1559.39 1282.96 Q1549.14 1282.96 1543.1 1276.52 Q1537.07 1270.08 1537.07 1259.14 Q1537.07 1248.04 1543.14 1241.68 Q1549.26 1235.32 1559.87 1235.32 Q1563.32 1235.32 1566.6 1236.05 Q1569.88 1236.74 1572.96 1238.16 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M1624.73 1257.24 L1624.73 1260.89 L1590.46 1260.89 Q1590.94 1268.58 1595.08 1272.63 Q1599.25 1276.64 1606.66 1276.64 Q1610.96 1276.64 1614.97 1275.59 Q1619.02 1274.54 1622.99 1272.43 L1622.99 1279.48 Q1618.98 1281.18 1614.76 1282.07 Q1610.55 1282.96 1606.22 1282.96 Q1595.36 1282.96 1589 1276.64 Q1582.68 1270.32 1582.68 1259.55 Q1582.68 1248.41 1588.68 1241.89 Q1594.71 1235.32 1604.92 1235.32 Q1614.07 1235.32 1619.38 1241.24 Q1624.73 1247.11 1624.73 1257.24 M1617.28 1255.05 Q1617.19 1248.94 1613.83 1245.29 Q1610.51 1241.64 1605 1241.64 Q1598.76 1241.64 1595 1245.17 Q1591.27 1248.69 1590.7 1255.09 L1617.28 1255.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip513)" d="M399.969 2174.14 L399.969 2174.14 L429.514 2174.14 L429.514 2174.14 L399.969 2174.14 L399.969 2174.14  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="399.969,2174.14 399.969,2174.14 429.514,2174.14 429.514,2174.14 399.969,2174.14 "/>
+<path clip-path="url(#clip513)" d="M436.9 1958.75 L436.9 2078.86 L466.444 2078.86 L466.444 1958.75 L436.9 1958.75 L436.9 1958.75  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="436.9,1958.75 436.9,2078.86 466.444,2078.86 466.444,1958.75 436.9,1958.75 "/>
+<path clip-path="url(#clip513)" d="M473.83 1869.53 L473.83 2039.4 L503.374 2039.4 L503.374 1869.53 L473.83 1869.53 L473.83 1869.53  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="473.83,1869.53 473.83,2039.4 503.374,2039.4 503.374,1869.53 473.83,1869.53 "/>
+<path clip-path="url(#clip513)" d="M510.76 1801.07 L510.76 2009.11 L540.304 2009.11 L540.304 1801.07 L510.76 1801.07 L510.76 1801.07  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="510.76,1801.07 510.76,2009.11 540.304,2009.11 540.304,1801.07 510.76,1801.07 "/>
+<path clip-path="url(#clip513)" d="M547.691 1743.36 L547.691 1983.58 L577.235 1983.58 L577.235 1743.36 L547.691 1743.36 L547.691 1743.36  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="547.691,1743.36 547.691,1983.58 577.235,1983.58 577.235,1743.36 547.691,1743.36 "/>
+<path clip-path="url(#clip513)" d="M584.621 1692.51 L584.621 1961.09 L614.165 1961.09 L614.165 1692.51 L584.621 1692.51 L584.621 1692.51  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="584.621,1692.51 584.621,1961.09 614.165,1961.09 614.165,1692.51 584.621,1692.51 "/>
+<path clip-path="url(#clip513)" d="M621.551 1646.54 L621.551 1940.76 L651.095 1940.76 L651.095 1646.54 L621.551 1646.54 L621.551 1646.54  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="621.551,1646.54 621.551,1940.76 651.095,1940.76 651.095,1646.54 621.551,1646.54 "/>
+<path clip-path="url(#clip513)" d="M658.481 1604.27 L658.481 1922.06 L688.026 1922.06 L688.026 1604.27 L658.481 1604.27 L658.481 1604.27  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="658.481,1604.27 658.481,1922.06 688.026,1922.06 688.026,1604.27 658.481,1604.27 "/>
+<path clip-path="url(#clip513)" d="M695.412 1564.93 L695.412 1904.65 L724.956 1904.65 L724.956 1564.93 L695.412 1564.93 L695.412 1564.93  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="695.412,1564.93 695.412,1904.65 724.956,1904.65 724.956,1564.93 695.412,1564.93 "/>
+<path clip-path="url(#clip513)" d="M732.342 1527.97 L732.342 1888.3 L761.886 1888.3 L761.886 1527.97 L732.342 1527.97 L732.342 1527.97  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="732.342,1527.97 732.342,1888.3 761.886,1888.3 761.886,1527.97 732.342,1527.97 "/>
+<path clip-path="url(#clip513)" d="M769.272 1493.02 L769.272 1872.84 L798.816 1872.84 L798.816 1493.02 L769.272 1493.02 L769.272 1493.02  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="769.272,1493.02 769.272,1872.84 798.816,1872.84 798.816,1493.02 769.272,1493.02 "/>
+<path clip-path="url(#clip513)" d="M806.203 1459.77 L806.203 1858.14 L835.747 1858.14 L835.747 1459.77 L806.203 1459.77 L806.203 1459.77  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="806.203,1459.77 806.203,1858.14 835.747,1858.14 835.747,1459.77 806.203,1459.77 "/>
+<path clip-path="url(#clip513)" d="M843.133 1428.01 L843.133 1844.08 L872.677 1844.08 L872.677 1428.01 L843.133 1428.01 L843.133 1428.01  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="843.133,1428.01 843.133,1844.08 872.677,1844.08 872.677,1428.01 843.133,1428.01 "/>
+<path clip-path="url(#clip513)" d="M880.063 1397.54 L880.063 1830.61 L909.607 1830.61 L909.607 1397.54 L880.063 1397.54 L880.063 1397.54  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="880.063,1397.54 880.063,1830.61 909.607,1830.61 909.607,1397.54 880.063,1397.54 "/>
+<path clip-path="url(#clip513)" d="M916.993 1368.23 L916.993 1817.64 L946.538 1817.64 L946.538 1368.23 L916.993 1368.23 L916.993 1368.23  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="916.993,1368.23 916.993,1817.64 946.538,1817.64 946.538,1368.23 916.993,1368.23 "/>
+<path clip-path="url(#clip513)" d="M953.924 1362.92 L953.924 1793.51 L983.468 1793.51 L983.468 1362.92 L953.924 1362.92 L953.924 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="953.924,1362.92 953.924,1793.51 983.468,1793.51 983.468,1362.92 953.924,1362.92 "/>
+<path clip-path="url(#clip513)" d="M990.854 1362.92 L990.854 1769.49 L1020.4 1769.49 L1020.4 1362.92 L990.854 1362.92 L990.854 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="990.854,1362.92 990.854,1769.49 1020.4,1769.49 1020.4,1362.92 990.854,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1027.78 1362.92 L1027.78 1747.67 L1057.33 1747.67 L1057.33 1362.92 L1027.78 1362.92 L1027.78 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1027.78,1362.92 1027.78,1747.67 1057.33,1747.67 1057.33,1362.92 1027.78,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1064.71 1362.92 L1064.71 1727.55 L1094.26 1727.55 L1094.26 1362.92 L1064.71 1362.92 L1064.71 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1064.71,1362.92 1064.71,1727.55 1094.26,1727.55 1094.26,1362.92 1064.71,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1101.64 1362.92 L1101.64 1708.77 L1131.19 1708.77 L1131.19 1362.92 L1101.64 1362.92 L1101.64 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1101.64,1362.92 1101.64,1708.77 1131.19,1708.77 1131.19,1362.92 1101.64,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1138.58 1362.92 L1138.58 1691.1 L1168.12 1691.1 L1168.12 1362.92 L1138.58 1362.92 L1138.58 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1138.58,1362.92 1138.58,1691.1 1168.12,1691.1 1168.12,1362.92 1138.58,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1175.51 1362.92 L1175.51 1674.37 L1205.05 1674.37 L1205.05 1362.92 L1175.51 1362.92 L1175.51 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1175.51,1362.92 1175.51,1674.37 1205.05,1674.37 1205.05,1362.92 1175.51,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1212.44 1362.92 L1212.44 1658.44 L1241.98 1658.44 L1241.98 1362.92 L1212.44 1362.92 L1212.44 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1212.44,1362.92 1212.44,1658.44 1241.98,1658.44 1241.98,1362.92 1212.44,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1249.37 1362.92 L1249.37 1643.2 L1278.91 1643.2 L1278.91 1362.92 L1249.37 1362.92 L1249.37 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1249.37,1362.92 1249.37,1643.2 1278.91,1643.2 1278.91,1362.92 1249.37,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1286.3 1362.92 L1286.3 1628.58 L1315.84 1628.58 L1315.84 1362.92 L1286.3 1362.92 L1286.3 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1286.3,1362.92 1286.3,1628.58 1315.84,1628.58 1315.84,1362.92 1286.3,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1323.23 1362.92 L1323.23 1614.49 L1352.77 1614.49 L1352.77 1362.92 L1323.23 1362.92 L1323.23 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1323.23,1362.92 1323.23,1614.49 1352.77,1614.49 1352.77,1362.92 1323.23,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1360.16 1362.92 L1360.16 1600.9 L1389.7 1600.9 L1389.7 1362.92 L1360.16 1362.92 L1360.16 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1360.16,1362.92 1360.16,1600.9 1389.7,1600.9 1389.7,1362.92 1360.16,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1397.09 1362.92 L1397.09 1587.74 L1426.63 1587.74 L1426.63 1362.92 L1397.09 1362.92 L1397.09 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1397.09,1362.92 1397.09,1587.74 1426.63,1587.74 1426.63,1362.92 1397.09,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1434.02 1362.92 L1434.02 1574.98 L1463.56 1574.98 L1463.56 1362.92 L1434.02 1362.92 L1434.02 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1434.02,1362.92 1434.02,1574.98 1463.56,1574.98 1463.56,1362.92 1434.02,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1470.95 1362.92 L1470.95 1562.59 L1500.49 1562.59 L1500.49 1362.92 L1470.95 1362.92 L1470.95 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1470.95,1362.92 1470.95,1562.59 1500.49,1562.59 1500.49,1362.92 1470.95,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1507.88 1362.92 L1507.88 1550.53 L1537.42 1550.53 L1537.42 1362.92 L1507.88 1362.92 L1507.88 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1507.88,1362.92 1507.88,1550.53 1537.42,1550.53 1537.42,1362.92 1507.88,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1544.81 1362.92 L1544.81 1538.79 L1574.35 1538.79 L1574.35 1362.92 L1544.81 1362.92 L1544.81 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1544.81,1362.92 1544.81,1538.79 1574.35,1538.79 1574.35,1362.92 1544.81,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1581.74 1362.92 L1581.74 1527.33 L1611.28 1527.33 L1611.28 1362.92 L1581.74 1362.92 L1581.74 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1581.74,1362.92 1581.74,1527.33 1611.28,1527.33 1611.28,1362.92 1581.74,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1618.67 1362.92 L1618.67 1516.13 L1648.21 1516.13 L1648.21 1362.92 L1618.67 1362.92 L1618.67 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1618.67,1362.92 1618.67,1516.13 1648.21,1516.13 1648.21,1362.92 1618.67,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1655.6 1362.92 L1655.6 1505.19 L1685.14 1505.19 L1685.14 1362.92 L1655.6 1362.92 L1655.6 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1655.6,1362.92 1655.6,1505.19 1685.14,1505.19 1685.14,1362.92 1655.6,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1692.53 1362.92 L1692.53 1494.48 L1722.07 1494.48 L1722.07 1362.92 L1692.53 1362.92 L1692.53 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1692.53,1362.92 1692.53,1494.48 1722.07,1494.48 1722.07,1362.92 1692.53,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1729.46 1362.92 L1729.46 1483.98 L1759 1483.98 L1759 1362.92 L1729.46 1362.92 L1729.46 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1729.46,1362.92 1729.46,1483.98 1759,1483.98 1759,1362.92 1729.46,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1766.39 1362.92 L1766.39 1473.7 L1795.93 1473.7 L1795.93 1362.92 L1766.39 1362.92 L1766.39 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1766.39,1362.92 1766.39,1473.7 1795.93,1473.7 1795.93,1362.92 1766.39,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1803.32 1362.92 L1803.32 1463.6 L1832.86 1463.6 L1832.86 1362.92 L1803.32 1362.92 L1803.32 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1803.32,1362.92 1803.32,1463.6 1832.86,1463.6 1832.86,1362.92 1803.32,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1840.25 1362.92 L1840.25 1453.7 L1869.79 1453.7 L1869.79 1362.92 L1840.25 1362.92 L1840.25 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1840.25,1362.92 1840.25,1453.7 1869.79,1453.7 1869.79,1362.92 1840.25,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1877.18 1362.92 L1877.18 1443.96 L1906.73 1443.96 L1906.73 1362.92 L1877.18 1362.92 L1877.18 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1877.18,1362.92 1877.18,1443.96 1906.73,1443.96 1906.73,1362.92 1877.18,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1914.11 1362.92 L1914.11 1434.39 L1943.66 1434.39 L1943.66 1362.92 L1914.11 1362.92 L1914.11 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1914.11,1362.92 1914.11,1434.39 1943.66,1434.39 1943.66,1362.92 1914.11,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1951.04 1362.92 L1951.04 1424.98 L1980.59 1424.98 L1980.59 1362.92 L1951.04 1362.92 L1951.04 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1951.04,1362.92 1951.04,1424.98 1980.59,1424.98 1980.59,1362.92 1951.04,1362.92 "/>
+<path clip-path="url(#clip513)" d="M1987.97 1362.92 L1987.97 1415.72 L2017.52 1415.72 L2017.52 1362.92 L1987.97 1362.92 L1987.97 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1987.97,1362.92 1987.97,1415.72 2017.52,1415.72 2017.52,1362.92 1987.97,1362.92 "/>
+<path clip-path="url(#clip513)" d="M2024.9 1362.92 L2024.9 1406.6 L2054.45 1406.6 L2054.45 1362.92 L2024.9 1362.92 L2024.9 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2024.9,1362.92 2024.9,1406.6 2054.45,1406.6 2054.45,1362.92 2024.9,1362.92 "/>
+<path clip-path="url(#clip513)" d="M2061.83 1362.92 L2061.83 1397.61 L2091.38 1397.61 L2091.38 1362.92 L2061.83 1362.92 L2061.83 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2061.83,1362.92 2061.83,1397.61 2091.38,1397.61 2091.38,1362.92 2061.83,1362.92 "/>
+<path clip-path="url(#clip513)" d="M2098.76 1362.92 L2098.76 1388.76 L2128.31 1388.76 L2128.31 1362.92 L2098.76 1362.92 L2098.76 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2098.76,1362.92 2098.76,1388.76 2128.31,1388.76 2128.31,1362.92 2098.76,1362.92 "/>
+<path clip-path="url(#clip513)" d="M2135.69 1362.92 L2135.69 1380.03 L2165.24 1380.03 L2165.24 1362.92 L2135.69 1362.92 L2135.69 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2135.69,1362.92 2135.69,1380.03 2165.24,1380.03 2165.24,1362.92 2135.69,1362.92 "/>
+<path clip-path="url(#clip513)" d="M2172.62 1362.92 L2172.62 1371.42 L2202.17 1371.42 L2202.17 1362.92 L2172.62 1362.92 L2172.62 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2172.62,1362.92 2172.62,1371.42 2202.17,1371.42 2202.17,1362.92 2172.62,1362.92 "/>
+<path clip-path="url(#clip513)" d="M2209.55 1362.92 L2209.55 1362.92 L2239.1 1362.92 L2239.1 1362.92 L2209.55 1362.92 L2209.55 1362.92  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2209.55,1362.92 2209.55,1362.92 2239.1,1362.92 2239.1,1362.92 2209.55,1362.92 "/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="414.741" cy="2174.14" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="451.672" cy="1958.75" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="488.602" cy="1869.53" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="525.532" cy="1801.07" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="562.463" cy="1743.36" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="599.393" cy="1692.51" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="636.323" cy="1646.54" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="673.253" cy="1604.27" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="710.184" cy="1564.93" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="747.114" cy="1527.97" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="784.044" cy="1493.02" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="820.975" cy="1459.77" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="857.905" cy="1428.01" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="894.835" cy="1397.54" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="931.766" cy="1368.23" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="968.696" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1005.63" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1042.56" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1079.49" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1116.42" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1153.35" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1190.28" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1227.21" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1264.14" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1301.07" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1338" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1374.93" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1411.86" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1448.79" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1485.72" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1522.65" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1559.58" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1596.51" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1633.44" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1670.37" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1707.3" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1744.23" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1781.16" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1818.09" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1855.02" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1891.95" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1928.88" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="1965.81" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2002.74" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2039.67" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2076.6" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2113.53" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2150.47" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2187.4" cy="1362.92" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#009af9; stroke:none; fill-opacity:0" cx="2224.33" cy="1362.92" r="2"/>
+<path clip-path="url(#clip513)" d="M399.969 2174.14 L399.969 2174.14 L429.514 2174.14 L429.514 2174.14 L399.969 2174.14 L399.969 2174.14  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="399.969,2174.14 399.969,2174.14 429.514,2174.14 429.514,2174.14 399.969,2174.14 "/>
+<path clip-path="url(#clip513)" d="M436.9 2078.86 L436.9 2078.86 L466.444 2078.86 L466.444 2078.86 L436.9 2078.86 L436.9 2078.86  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="436.9,2078.86 436.9,2078.86 466.444,2078.86 466.444,2078.86 436.9,2078.86 "/>
+<path clip-path="url(#clip513)" d="M473.83 2039.4 L473.83 2039.4 L503.374 2039.4 L503.374 2039.4 L473.83 2039.4 L473.83 2039.4  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="473.83,2039.4 473.83,2039.4 503.374,2039.4 503.374,2039.4 473.83,2039.4 "/>
+<path clip-path="url(#clip513)" d="M510.76 2009.11 L510.76 2009.11 L540.304 2009.11 L540.304 2009.11 L510.76 2009.11 L510.76 2009.11  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="510.76,2009.11 510.76,2009.11 540.304,2009.11 540.304,2009.11 510.76,2009.11 "/>
+<path clip-path="url(#clip513)" d="M547.691 1983.58 L547.691 1983.58 L577.235 1983.58 L577.235 1983.58 L547.691 1983.58 L547.691 1983.58  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="547.691,1983.58 547.691,1983.58 577.235,1983.58 577.235,1983.58 547.691,1983.58 "/>
+<path clip-path="url(#clip513)" d="M584.621 1961.09 L584.621 1961.09 L614.165 1961.09 L614.165 1961.09 L584.621 1961.09 L584.621 1961.09  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="584.621,1961.09 584.621,1961.09 614.165,1961.09 614.165,1961.09 584.621,1961.09 "/>
+<path clip-path="url(#clip513)" d="M621.551 1940.76 L621.551 1940.76 L651.095 1940.76 L651.095 1940.76 L621.551 1940.76 L621.551 1940.76  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="621.551,1940.76 621.551,1940.76 651.095,1940.76 651.095,1940.76 621.551,1940.76 "/>
+<path clip-path="url(#clip513)" d="M658.481 1922.06 L658.481 1922.06 L688.026 1922.06 L688.026 1922.06 L658.481 1922.06 L658.481 1922.06  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="658.481,1922.06 658.481,1922.06 688.026,1922.06 688.026,1922.06 658.481,1922.06 "/>
+<path clip-path="url(#clip513)" d="M695.412 1904.65 L695.412 1904.65 L724.956 1904.65 L724.956 1904.65 L695.412 1904.65 L695.412 1904.65  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="695.412,1904.65 695.412,1904.65 724.956,1904.65 724.956,1904.65 695.412,1904.65 "/>
+<path clip-path="url(#clip513)" d="M732.342 1888.3 L732.342 1888.3 L761.886 1888.3 L761.886 1888.3 L732.342 1888.3 L732.342 1888.3  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="732.342,1888.3 732.342,1888.3 761.886,1888.3 761.886,1888.3 732.342,1888.3 "/>
+<path clip-path="url(#clip513)" d="M769.272 1872.84 L769.272 1872.84 L798.816 1872.84 L798.816 1872.84 L769.272 1872.84 L769.272 1872.84  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="769.272,1872.84 769.272,1872.84 798.816,1872.84 798.816,1872.84 769.272,1872.84 "/>
+<path clip-path="url(#clip513)" d="M806.203 1858.14 L806.203 1858.14 L835.747 1858.14 L835.747 1858.14 L806.203 1858.14 L806.203 1858.14  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="806.203,1858.14 806.203,1858.14 835.747,1858.14 835.747,1858.14 806.203,1858.14 "/>
+<path clip-path="url(#clip513)" d="M843.133 1844.08 L843.133 1844.08 L872.677 1844.08 L872.677 1844.08 L843.133 1844.08 L843.133 1844.08  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="843.133,1844.08 843.133,1844.08 872.677,1844.08 872.677,1844.08 843.133,1844.08 "/>
+<path clip-path="url(#clip513)" d="M880.063 1830.61 L880.063 1830.61 L909.607 1830.61 L909.607 1830.61 L880.063 1830.61 L880.063 1830.61  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="880.063,1830.61 880.063,1830.61 909.607,1830.61 909.607,1830.61 880.063,1830.61 "/>
+<path clip-path="url(#clip513)" d="M916.993 1817.64 L916.993 1817.64 L946.538 1817.64 L946.538 1817.64 L916.993 1817.64 L916.993 1817.64  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="916.993,1817.64 916.993,1817.64 946.538,1817.64 946.538,1817.64 916.993,1817.64 "/>
+<path clip-path="url(#clip513)" d="M953.924 1793.51 L953.924 1793.51 L983.468 1793.51 L983.468 1793.51 L953.924 1793.51 L953.924 1793.51  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="953.924,1793.51 953.924,1793.51 983.468,1793.51 983.468,1793.51 953.924,1793.51 "/>
+<path clip-path="url(#clip513)" d="M990.854 1769.49 L990.854 1769.49 L1020.4 1769.49 L1020.4 1769.49 L990.854 1769.49 L990.854 1769.49  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="990.854,1769.49 990.854,1769.49 1020.4,1769.49 1020.4,1769.49 990.854,1769.49 "/>
+<path clip-path="url(#clip513)" d="M1027.78 1747.67 L1027.78 1747.67 L1057.33 1747.67 L1057.33 1747.67 L1027.78 1747.67 L1027.78 1747.67  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1027.78,1747.67 1027.78,1747.67 1057.33,1747.67 1057.33,1747.67 1027.78,1747.67 "/>
+<path clip-path="url(#clip513)" d="M1064.71 1727.55 L1064.71 1727.55 L1094.26 1727.55 L1094.26 1727.55 L1064.71 1727.55 L1064.71 1727.55  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1064.71,1727.55 1064.71,1727.55 1094.26,1727.55 1094.26,1727.55 1064.71,1727.55 "/>
+<path clip-path="url(#clip513)" d="M1101.64 1708.77 L1101.64 1708.77 L1131.19 1708.77 L1131.19 1708.77 L1101.64 1708.77 L1101.64 1708.77  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1101.64,1708.77 1101.64,1708.77 1131.19,1708.77 1131.19,1708.77 1101.64,1708.77 "/>
+<path clip-path="url(#clip513)" d="M1138.58 1691.1 L1138.58 1691.1 L1168.12 1691.1 L1168.12 1691.1 L1138.58 1691.1 L1138.58 1691.1  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1138.58,1691.1 1138.58,1691.1 1168.12,1691.1 1168.12,1691.1 1138.58,1691.1 "/>
+<path clip-path="url(#clip513)" d="M1175.51 1674.37 L1175.51 1674.37 L1205.05 1674.37 L1205.05 1674.37 L1175.51 1674.37 L1175.51 1674.37  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1175.51,1674.37 1175.51,1674.37 1205.05,1674.37 1205.05,1674.37 1175.51,1674.37 "/>
+<path clip-path="url(#clip513)" d="M1212.44 1658.44 L1212.44 1658.44 L1241.98 1658.44 L1241.98 1658.44 L1212.44 1658.44 L1212.44 1658.44  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1212.44,1658.44 1212.44,1658.44 1241.98,1658.44 1241.98,1658.44 1212.44,1658.44 "/>
+<path clip-path="url(#clip513)" d="M1249.37 1643.2 L1249.37 1643.2 L1278.91 1643.2 L1278.91 1643.2 L1249.37 1643.2 L1249.37 1643.2  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1249.37,1643.2 1249.37,1643.2 1278.91,1643.2 1278.91,1643.2 1249.37,1643.2 "/>
+<path clip-path="url(#clip513)" d="M1286.3 1628.58 L1286.3 1628.58 L1315.84 1628.58 L1315.84 1628.58 L1286.3 1628.58 L1286.3 1628.58  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1286.3,1628.58 1286.3,1628.58 1315.84,1628.58 1315.84,1628.58 1286.3,1628.58 "/>
+<path clip-path="url(#clip513)" d="M1323.23 1614.49 L1323.23 1614.49 L1352.77 1614.49 L1352.77 1614.49 L1323.23 1614.49 L1323.23 1614.49  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1323.23,1614.49 1323.23,1614.49 1352.77,1614.49 1352.77,1614.49 1323.23,1614.49 "/>
+<path clip-path="url(#clip513)" d="M1360.16 1600.9 L1360.16 1600.9 L1389.7 1600.9 L1389.7 1600.9 L1360.16 1600.9 L1360.16 1600.9  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1360.16,1600.9 1360.16,1600.9 1389.7,1600.9 1389.7,1600.9 1360.16,1600.9 "/>
+<path clip-path="url(#clip513)" d="M1397.09 1587.74 L1397.09 1587.74 L1426.63 1587.74 L1426.63 1587.74 L1397.09 1587.74 L1397.09 1587.74  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1397.09,1587.74 1397.09,1587.74 1426.63,1587.74 1426.63,1587.74 1397.09,1587.74 "/>
+<path clip-path="url(#clip513)" d="M1434.02 1574.98 L1434.02 1574.98 L1463.56 1574.98 L1463.56 1574.98 L1434.02 1574.98 L1434.02 1574.98  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1434.02,1574.98 1434.02,1574.98 1463.56,1574.98 1463.56,1574.98 1434.02,1574.98 "/>
+<path clip-path="url(#clip513)" d="M1470.95 1562.59 L1470.95 1562.59 L1500.49 1562.59 L1500.49 1562.59 L1470.95 1562.59 L1470.95 1562.59  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1470.95,1562.59 1470.95,1562.59 1500.49,1562.59 1500.49,1562.59 1470.95,1562.59 "/>
+<path clip-path="url(#clip513)" d="M1507.88 1550.53 L1507.88 1550.53 L1537.42 1550.53 L1537.42 1550.53 L1507.88 1550.53 L1507.88 1550.53  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1507.88,1550.53 1507.88,1550.53 1537.42,1550.53 1537.42,1550.53 1507.88,1550.53 "/>
+<path clip-path="url(#clip513)" d="M1544.81 1538.79 L1544.81 1538.79 L1574.35 1538.79 L1574.35 1538.79 L1544.81 1538.79 L1544.81 1538.79  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1544.81,1538.79 1544.81,1538.79 1574.35,1538.79 1574.35,1538.79 1544.81,1538.79 "/>
+<path clip-path="url(#clip513)" d="M1581.74 1527.33 L1581.74 1527.33 L1611.28 1527.33 L1611.28 1527.33 L1581.74 1527.33 L1581.74 1527.33  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1581.74,1527.33 1581.74,1527.33 1611.28,1527.33 1611.28,1527.33 1581.74,1527.33 "/>
+<path clip-path="url(#clip513)" d="M1618.67 1516.13 L1618.67 1516.13 L1648.21 1516.13 L1648.21 1516.13 L1618.67 1516.13 L1618.67 1516.13  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1618.67,1516.13 1618.67,1516.13 1648.21,1516.13 1648.21,1516.13 1618.67,1516.13 "/>
+<path clip-path="url(#clip513)" d="M1655.6 1505.19 L1655.6 1505.19 L1685.14 1505.19 L1685.14 1505.19 L1655.6 1505.19 L1655.6 1505.19  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1655.6,1505.19 1655.6,1505.19 1685.14,1505.19 1685.14,1505.19 1655.6,1505.19 "/>
+<path clip-path="url(#clip513)" d="M1692.53 1494.48 L1692.53 1494.48 L1722.07 1494.48 L1722.07 1494.48 L1692.53 1494.48 L1692.53 1494.48  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1692.53,1494.48 1692.53,1494.48 1722.07,1494.48 1722.07,1494.48 1692.53,1494.48 "/>
+<path clip-path="url(#clip513)" d="M1729.46 1483.98 L1729.46 1483.98 L1759 1483.98 L1759 1483.98 L1729.46 1483.98 L1729.46 1483.98  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1729.46,1483.98 1729.46,1483.98 1759,1483.98 1759,1483.98 1729.46,1483.98 "/>
+<path clip-path="url(#clip513)" d="M1766.39 1473.7 L1766.39 1473.7 L1795.93 1473.7 L1795.93 1473.7 L1766.39 1473.7 L1766.39 1473.7  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1766.39,1473.7 1766.39,1473.7 1795.93,1473.7 1795.93,1473.7 1766.39,1473.7 "/>
+<path clip-path="url(#clip513)" d="M1803.32 1463.6 L1803.32 1463.6 L1832.86 1463.6 L1832.86 1463.6 L1803.32 1463.6 L1803.32 1463.6  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1803.32,1463.6 1803.32,1463.6 1832.86,1463.6 1832.86,1463.6 1803.32,1463.6 "/>
+<path clip-path="url(#clip513)" d="M1840.25 1453.7 L1840.25 1453.7 L1869.79 1453.7 L1869.79 1453.7 L1840.25 1453.7 L1840.25 1453.7  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1840.25,1453.7 1840.25,1453.7 1869.79,1453.7 1869.79,1453.7 1840.25,1453.7 "/>
+<path clip-path="url(#clip513)" d="M1877.18 1443.96 L1877.18 1443.96 L1906.73 1443.96 L1906.73 1443.96 L1877.18 1443.96 L1877.18 1443.96  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1877.18,1443.96 1877.18,1443.96 1906.73,1443.96 1906.73,1443.96 1877.18,1443.96 "/>
+<path clip-path="url(#clip513)" d="M1914.11 1434.39 L1914.11 1434.39 L1943.66 1434.39 L1943.66 1434.39 L1914.11 1434.39 L1914.11 1434.39  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1914.11,1434.39 1914.11,1434.39 1943.66,1434.39 1943.66,1434.39 1914.11,1434.39 "/>
+<path clip-path="url(#clip513)" d="M1951.04 1424.98 L1951.04 1424.98 L1980.59 1424.98 L1980.59 1424.98 L1951.04 1424.98 L1951.04 1424.98  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1951.04,1424.98 1951.04,1424.98 1980.59,1424.98 1980.59,1424.98 1951.04,1424.98 "/>
+<path clip-path="url(#clip513)" d="M1987.97 1415.72 L1987.97 1415.72 L2017.52 1415.72 L2017.52 1415.72 L1987.97 1415.72 L1987.97 1415.72  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1987.97,1415.72 1987.97,1415.72 2017.52,1415.72 2017.52,1415.72 1987.97,1415.72 "/>
+<path clip-path="url(#clip513)" d="M2024.9 1406.6 L2024.9 1406.6 L2054.45 1406.6 L2054.45 1406.6 L2024.9 1406.6 L2024.9 1406.6  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2024.9,1406.6 2024.9,1406.6 2054.45,1406.6 2054.45,1406.6 2024.9,1406.6 "/>
+<path clip-path="url(#clip513)" d="M2061.83 1397.61 L2061.83 1397.61 L2091.38 1397.61 L2091.38 1397.61 L2061.83 1397.61 L2061.83 1397.61  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2061.83,1397.61 2061.83,1397.61 2091.38,1397.61 2091.38,1397.61 2061.83,1397.61 "/>
+<path clip-path="url(#clip513)" d="M2098.76 1388.76 L2098.76 1388.76 L2128.31 1388.76 L2128.31 1388.76 L2098.76 1388.76 L2098.76 1388.76  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2098.76,1388.76 2098.76,1388.76 2128.31,1388.76 2128.31,1388.76 2098.76,1388.76 "/>
+<path clip-path="url(#clip513)" d="M2135.69 1380.03 L2135.69 1380.03 L2165.24 1380.03 L2165.24 1380.03 L2135.69 1380.03 L2135.69 1380.03  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2135.69,1380.03 2135.69,1380.03 2165.24,1380.03 2165.24,1380.03 2135.69,1380.03 "/>
+<path clip-path="url(#clip513)" d="M2172.62 1371.42 L2172.62 1371.42 L2202.17 1371.42 L2202.17 1371.42 L2172.62 1371.42 L2172.62 1371.42  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2172.62,1371.42 2172.62,1371.42 2202.17,1371.42 2202.17,1371.42 2172.62,1371.42 "/>
+<path clip-path="url(#clip513)" d="M2209.55 1362.92 L2209.55 1362.92 L2239.1 1362.92 L2239.1 1362.92 L2209.55 1362.92 L2209.55 1362.92  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2209.55,1362.92 2209.55,1362.92 2239.1,1362.92 2239.1,1362.92 2209.55,1362.92 "/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="414.741" cy="2174.14" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="451.672" cy="2078.86" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="488.602" cy="2039.4" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="525.532" cy="2009.11" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="562.463" cy="1983.58" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="599.393" cy="1961.09" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="636.323" cy="1940.76" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="673.253" cy="1922.06" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="710.184" cy="1904.65" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="747.114" cy="1888.3" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="784.044" cy="1872.84" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="820.975" cy="1858.14" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="857.905" cy="1844.08" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="894.835" cy="1830.61" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="931.766" cy="1817.64" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="968.696" cy="1793.51" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1005.63" cy="1769.49" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1042.56" cy="1747.67" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1079.49" cy="1727.55" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1116.42" cy="1708.77" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1153.35" cy="1691.1" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1190.28" cy="1674.37" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1227.21" cy="1658.44" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1264.14" cy="1643.2" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1301.07" cy="1628.58" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1338" cy="1614.49" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1374.93" cy="1600.9" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1411.86" cy="1587.74" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1448.79" cy="1574.98" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1485.72" cy="1562.59" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1522.65" cy="1550.53" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1559.58" cy="1538.79" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1596.51" cy="1527.33" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1633.44" cy="1516.13" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1670.37" cy="1505.19" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1707.3" cy="1494.48" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1744.23" cy="1483.98" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1781.16" cy="1473.7" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1818.09" cy="1463.6" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1855.02" cy="1453.7" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1891.95" cy="1443.96" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1928.88" cy="1434.39" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="1965.81" cy="1424.98" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2002.74" cy="1415.72" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2039.67" cy="1406.6" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2076.6" cy="1397.61" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2113.53" cy="1388.76" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2150.47" cy="1380.03" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2187.4" cy="1371.42" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#e26f46; stroke:none; fill-opacity:0" cx="2224.33" cy="1362.92" r="2"/>
+<path clip-path="url(#clip513)" d="M399.969 2174.14 L399.969 2174.14 L429.514 2174.14 L429.514 2174.14 L399.969 2174.14 L399.969 2174.14  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="399.969,2174.14 399.969,2174.14 429.514,2174.14 429.514,2174.14 399.969,2174.14 "/>
+<path clip-path="url(#clip513)" d="M436.9 2078.86 L436.9 2174.14 L466.444 2174.14 L466.444 2078.86 L436.9 2078.86 L436.9 2078.86  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="436.9,2078.86 436.9,2174.14 466.444,2174.14 466.444,2078.86 436.9,2078.86 "/>
+<path clip-path="url(#clip513)" d="M473.83 2039.4 L473.83 2174.14 L503.374 2174.14 L503.374 2039.4 L473.83 2039.4 L473.83 2039.4  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="473.83,2039.4 473.83,2174.14 503.374,2174.14 503.374,2039.4 473.83,2039.4 "/>
+<path clip-path="url(#clip513)" d="M510.76 2009.11 L510.76 2174.14 L540.304 2174.14 L540.304 2009.11 L510.76 2009.11 L510.76 2009.11  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="510.76,2009.11 510.76,2174.14 540.304,2174.14 540.304,2009.11 510.76,2009.11 "/>
+<path clip-path="url(#clip513)" d="M547.691 1983.58 L547.691 2174.14 L577.235 2174.14 L577.235 1983.58 L547.691 1983.58 L547.691 1983.58  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="547.691,1983.58 547.691,2174.14 577.235,2174.14 577.235,1983.58 547.691,1983.58 "/>
+<path clip-path="url(#clip513)" d="M584.621 1961.09 L584.621 2174.14 L614.165 2174.14 L614.165 1961.09 L584.621 1961.09 L584.621 1961.09  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="584.621,1961.09 584.621,2174.14 614.165,2174.14 614.165,1961.09 584.621,1961.09 "/>
+<path clip-path="url(#clip513)" d="M621.551 1940.76 L621.551 2174.14 L651.095 2174.14 L651.095 1940.76 L621.551 1940.76 L621.551 1940.76  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="621.551,1940.76 621.551,2174.14 651.095,2174.14 651.095,1940.76 621.551,1940.76 "/>
+<path clip-path="url(#clip513)" d="M658.481 1922.06 L658.481 2174.14 L688.026 2174.14 L688.026 1922.06 L658.481 1922.06 L658.481 1922.06  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="658.481,1922.06 658.481,2174.14 688.026,2174.14 688.026,1922.06 658.481,1922.06 "/>
+<path clip-path="url(#clip513)" d="M695.412 1904.65 L695.412 2174.14 L724.956 2174.14 L724.956 1904.65 L695.412 1904.65 L695.412 1904.65  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="695.412,1904.65 695.412,2174.14 724.956,2174.14 724.956,1904.65 695.412,1904.65 "/>
+<path clip-path="url(#clip513)" d="M732.342 1888.3 L732.342 2174.14 L761.886 2174.14 L761.886 1888.3 L732.342 1888.3 L732.342 1888.3  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="732.342,1888.3 732.342,2174.14 761.886,2174.14 761.886,1888.3 732.342,1888.3 "/>
+<path clip-path="url(#clip513)" d="M769.272 1872.84 L769.272 2174.14 L798.816 2174.14 L798.816 1872.84 L769.272 1872.84 L769.272 1872.84  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="769.272,1872.84 769.272,2174.14 798.816,2174.14 798.816,1872.84 769.272,1872.84 "/>
+<path clip-path="url(#clip513)" d="M806.203 1858.14 L806.203 2174.14 L835.747 2174.14 L835.747 1858.14 L806.203 1858.14 L806.203 1858.14  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="806.203,1858.14 806.203,2174.14 835.747,2174.14 835.747,1858.14 806.203,1858.14 "/>
+<path clip-path="url(#clip513)" d="M843.133 1844.08 L843.133 2174.14 L872.677 2174.14 L872.677 1844.08 L843.133 1844.08 L843.133 1844.08  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="843.133,1844.08 843.133,2174.14 872.677,2174.14 872.677,1844.08 843.133,1844.08 "/>
+<path clip-path="url(#clip513)" d="M880.063 1830.61 L880.063 2174.14 L909.607 2174.14 L909.607 1830.61 L880.063 1830.61 L880.063 1830.61  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="880.063,1830.61 880.063,2174.14 909.607,2174.14 909.607,1830.61 880.063,1830.61 "/>
+<path clip-path="url(#clip513)" d="M916.993 1817.64 L916.993 2174.14 L946.538 2174.14 L946.538 1817.64 L916.993 1817.64 L916.993 1817.64  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="916.993,1817.64 916.993,2174.14 946.538,2174.14 946.538,1817.64 916.993,1817.64 "/>
+<path clip-path="url(#clip513)" d="M953.924 1793.51 L953.924 2174.14 L983.468 2174.14 L983.468 1793.51 L953.924 1793.51 L953.924 1793.51  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="953.924,1793.51 953.924,2174.14 983.468,2174.14 983.468,1793.51 953.924,1793.51 "/>
+<path clip-path="url(#clip513)" d="M990.854 1769.49 L990.854 2174.14 L1020.4 2174.14 L1020.4 1769.49 L990.854 1769.49 L990.854 1769.49  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="990.854,1769.49 990.854,2174.14 1020.4,2174.14 1020.4,1769.49 990.854,1769.49 "/>
+<path clip-path="url(#clip513)" d="M1027.78 1747.67 L1027.78 2174.14 L1057.33 2174.14 L1057.33 1747.67 L1027.78 1747.67 L1027.78 1747.67  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1027.78,1747.67 1027.78,2174.14 1057.33,2174.14 1057.33,1747.67 1027.78,1747.67 "/>
+<path clip-path="url(#clip513)" d="M1064.71 1727.55 L1064.71 2174.14 L1094.26 2174.14 L1094.26 1727.55 L1064.71 1727.55 L1064.71 1727.55  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1064.71,1727.55 1064.71,2174.14 1094.26,2174.14 1094.26,1727.55 1064.71,1727.55 "/>
+<path clip-path="url(#clip513)" d="M1101.64 1708.77 L1101.64 2174.14 L1131.19 2174.14 L1131.19 1708.77 L1101.64 1708.77 L1101.64 1708.77  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1101.64,1708.77 1101.64,2174.14 1131.19,2174.14 1131.19,1708.77 1101.64,1708.77 "/>
+<path clip-path="url(#clip513)" d="M1138.58 1691.1 L1138.58 2174.14 L1168.12 2174.14 L1168.12 1691.1 L1138.58 1691.1 L1138.58 1691.1  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1138.58,1691.1 1138.58,2174.14 1168.12,2174.14 1168.12,1691.1 1138.58,1691.1 "/>
+<path clip-path="url(#clip513)" d="M1175.51 1674.37 L1175.51 2174.14 L1205.05 2174.14 L1205.05 1674.37 L1175.51 1674.37 L1175.51 1674.37  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1175.51,1674.37 1175.51,2174.14 1205.05,2174.14 1205.05,1674.37 1175.51,1674.37 "/>
+<path clip-path="url(#clip513)" d="M1212.44 1658.44 L1212.44 2174.14 L1241.98 2174.14 L1241.98 1658.44 L1212.44 1658.44 L1212.44 1658.44  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1212.44,1658.44 1212.44,2174.14 1241.98,2174.14 1241.98,1658.44 1212.44,1658.44 "/>
+<path clip-path="url(#clip513)" d="M1249.37 1643.2 L1249.37 2174.14 L1278.91 2174.14 L1278.91 1643.2 L1249.37 1643.2 L1249.37 1643.2  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1249.37,1643.2 1249.37,2174.14 1278.91,2174.14 1278.91,1643.2 1249.37,1643.2 "/>
+<path clip-path="url(#clip513)" d="M1286.3 1628.58 L1286.3 2174.14 L1315.84 2174.14 L1315.84 1628.58 L1286.3 1628.58 L1286.3 1628.58  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1286.3,1628.58 1286.3,2174.14 1315.84,2174.14 1315.84,1628.58 1286.3,1628.58 "/>
+<path clip-path="url(#clip513)" d="M1323.23 1614.49 L1323.23 2174.14 L1352.77 2174.14 L1352.77 1614.49 L1323.23 1614.49 L1323.23 1614.49  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1323.23,1614.49 1323.23,2174.14 1352.77,2174.14 1352.77,1614.49 1323.23,1614.49 "/>
+<path clip-path="url(#clip513)" d="M1360.16 1600.9 L1360.16 2174.14 L1389.7 2174.14 L1389.7 1600.9 L1360.16 1600.9 L1360.16 1600.9  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1360.16,1600.9 1360.16,2174.14 1389.7,2174.14 1389.7,1600.9 1360.16,1600.9 "/>
+<path clip-path="url(#clip513)" d="M1397.09 1587.74 L1397.09 2174.14 L1426.63 2174.14 L1426.63 1587.74 L1397.09 1587.74 L1397.09 1587.74  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1397.09,1587.74 1397.09,2174.14 1426.63,2174.14 1426.63,1587.74 1397.09,1587.74 "/>
+<path clip-path="url(#clip513)" d="M1434.02 1574.98 L1434.02 2174.14 L1463.56 2174.14 L1463.56 1574.98 L1434.02 1574.98 L1434.02 1574.98  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1434.02,1574.98 1434.02,2174.14 1463.56,2174.14 1463.56,1574.98 1434.02,1574.98 "/>
+<path clip-path="url(#clip513)" d="M1470.95 1562.59 L1470.95 2174.14 L1500.49 2174.14 L1500.49 1562.59 L1470.95 1562.59 L1470.95 1562.59  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1470.95,1562.59 1470.95,2174.14 1500.49,2174.14 1500.49,1562.59 1470.95,1562.59 "/>
+<path clip-path="url(#clip513)" d="M1507.88 1550.53 L1507.88 2174.14 L1537.42 2174.14 L1537.42 1550.53 L1507.88 1550.53 L1507.88 1550.53  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1507.88,1550.53 1507.88,2174.14 1537.42,2174.14 1537.42,1550.53 1507.88,1550.53 "/>
+<path clip-path="url(#clip513)" d="M1544.81 1538.79 L1544.81 2174.14 L1574.35 2174.14 L1574.35 1538.79 L1544.81 1538.79 L1544.81 1538.79  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1544.81,1538.79 1544.81,2174.14 1574.35,2174.14 1574.35,1538.79 1544.81,1538.79 "/>
+<path clip-path="url(#clip513)" d="M1581.74 1527.33 L1581.74 2174.14 L1611.28 2174.14 L1611.28 1527.33 L1581.74 1527.33 L1581.74 1527.33  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1581.74,1527.33 1581.74,2174.14 1611.28,2174.14 1611.28,1527.33 1581.74,1527.33 "/>
+<path clip-path="url(#clip513)" d="M1618.67 1516.13 L1618.67 2174.14 L1648.21 2174.14 L1648.21 1516.13 L1618.67 1516.13 L1618.67 1516.13  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1618.67,1516.13 1618.67,2174.14 1648.21,2174.14 1648.21,1516.13 1618.67,1516.13 "/>
+<path clip-path="url(#clip513)" d="M1655.6 1505.19 L1655.6 2174.14 L1685.14 2174.14 L1685.14 1505.19 L1655.6 1505.19 L1655.6 1505.19  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1655.6,1505.19 1655.6,2174.14 1685.14,2174.14 1685.14,1505.19 1655.6,1505.19 "/>
+<path clip-path="url(#clip513)" d="M1692.53 1494.48 L1692.53 2174.14 L1722.07 2174.14 L1722.07 1494.48 L1692.53 1494.48 L1692.53 1494.48  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1692.53,1494.48 1692.53,2174.14 1722.07,2174.14 1722.07,1494.48 1692.53,1494.48 "/>
+<path clip-path="url(#clip513)" d="M1729.46 1483.98 L1729.46 2174.14 L1759 2174.14 L1759 1483.98 L1729.46 1483.98 L1729.46 1483.98  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1729.46,1483.98 1729.46,2174.14 1759,2174.14 1759,1483.98 1729.46,1483.98 "/>
+<path clip-path="url(#clip513)" d="M1766.39 1473.7 L1766.39 2174.14 L1795.93 2174.14 L1795.93 1473.7 L1766.39 1473.7 L1766.39 1473.7  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1766.39,1473.7 1766.39,2174.14 1795.93,2174.14 1795.93,1473.7 1766.39,1473.7 "/>
+<path clip-path="url(#clip513)" d="M1803.32 1463.6 L1803.32 2174.14 L1832.86 2174.14 L1832.86 1463.6 L1803.32 1463.6 L1803.32 1463.6  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1803.32,1463.6 1803.32,2174.14 1832.86,2174.14 1832.86,1463.6 1803.32,1463.6 "/>
+<path clip-path="url(#clip513)" d="M1840.25 1453.7 L1840.25 2174.14 L1869.79 2174.14 L1869.79 1453.7 L1840.25 1453.7 L1840.25 1453.7  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1840.25,1453.7 1840.25,2174.14 1869.79,2174.14 1869.79,1453.7 1840.25,1453.7 "/>
+<path clip-path="url(#clip513)" d="M1877.18 1443.96 L1877.18 2174.14 L1906.73 2174.14 L1906.73 1443.96 L1877.18 1443.96 L1877.18 1443.96  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1877.18,1443.96 1877.18,2174.14 1906.73,2174.14 1906.73,1443.96 1877.18,1443.96 "/>
+<path clip-path="url(#clip513)" d="M1914.11 1434.39 L1914.11 2174.14 L1943.66 2174.14 L1943.66 1434.39 L1914.11 1434.39 L1914.11 1434.39  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1914.11,1434.39 1914.11,2174.14 1943.66,2174.14 1943.66,1434.39 1914.11,1434.39 "/>
+<path clip-path="url(#clip513)" d="M1951.04 1424.98 L1951.04 2174.14 L1980.59 2174.14 L1980.59 1424.98 L1951.04 1424.98 L1951.04 1424.98  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1951.04,1424.98 1951.04,2174.14 1980.59,2174.14 1980.59,1424.98 1951.04,1424.98 "/>
+<path clip-path="url(#clip513)" d="M1987.97 1415.72 L1987.97 2174.14 L2017.52 2174.14 L2017.52 1415.72 L1987.97 1415.72 L1987.97 1415.72  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1987.97,1415.72 1987.97,2174.14 2017.52,2174.14 2017.52,1415.72 1987.97,1415.72 "/>
+<path clip-path="url(#clip513)" d="M2024.9 1406.6 L2024.9 2174.14 L2054.45 2174.14 L2054.45 1406.6 L2024.9 1406.6 L2024.9 1406.6  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2024.9,1406.6 2024.9,2174.14 2054.45,2174.14 2054.45,1406.6 2024.9,1406.6 "/>
+<path clip-path="url(#clip513)" d="M2061.83 1397.61 L2061.83 2174.14 L2091.38 2174.14 L2091.38 1397.61 L2061.83 1397.61 L2061.83 1397.61  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2061.83,1397.61 2061.83,2174.14 2091.38,2174.14 2091.38,1397.61 2061.83,1397.61 "/>
+<path clip-path="url(#clip513)" d="M2098.76 1388.76 L2098.76 2174.14 L2128.31 2174.14 L2128.31 1388.76 L2098.76 1388.76 L2098.76 1388.76  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2098.76,1388.76 2098.76,2174.14 2128.31,2174.14 2128.31,1388.76 2098.76,1388.76 "/>
+<path clip-path="url(#clip513)" d="M2135.69 1380.03 L2135.69 2174.14 L2165.24 2174.14 L2165.24 1380.03 L2135.69 1380.03 L2135.69 1380.03  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2135.69,1380.03 2135.69,2174.14 2165.24,2174.14 2165.24,1380.03 2135.69,1380.03 "/>
+<path clip-path="url(#clip513)" d="M2172.62 1371.42 L2172.62 2174.14 L2202.17 2174.14 L2202.17 1371.42 L2172.62 1371.42 L2172.62 1371.42  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2172.62,1371.42 2172.62,2174.14 2202.17,2174.14 2202.17,1371.42 2172.62,1371.42 "/>
+<path clip-path="url(#clip513)" d="M2209.55 1362.92 L2209.55 2174.14 L2239.1 2174.14 L2239.1 1362.92 L2209.55 1362.92 L2209.55 1362.92  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip513)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2209.55,1362.92 2209.55,2174.14 2239.1,2174.14 2239.1,1362.92 2209.55,1362.92 "/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="414.741" cy="2174.14" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="451.672" cy="2078.86" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="488.602" cy="2039.4" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="525.532" cy="2009.11" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="562.463" cy="1983.58" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="599.393" cy="1961.09" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="636.323" cy="1940.76" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="673.253" cy="1922.06" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="710.184" cy="1904.65" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="747.114" cy="1888.3" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="784.044" cy="1872.84" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="820.975" cy="1858.14" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="857.905" cy="1844.08" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="894.835" cy="1830.61" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="931.766" cy="1817.64" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="968.696" cy="1793.51" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1005.63" cy="1769.49" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1042.56" cy="1747.67" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1079.49" cy="1727.55" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1116.42" cy="1708.77" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1153.35" cy="1691.1" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1190.28" cy="1674.37" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1227.21" cy="1658.44" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1264.14" cy="1643.2" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1301.07" cy="1628.58" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1338" cy="1614.49" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1374.93" cy="1600.9" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1411.86" cy="1587.74" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1448.79" cy="1574.98" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1485.72" cy="1562.59" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1522.65" cy="1550.53" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1559.58" cy="1538.79" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1596.51" cy="1527.33" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1633.44" cy="1516.13" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1670.37" cy="1505.19" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1707.3" cy="1494.48" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1744.23" cy="1483.98" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1781.16" cy="1473.7" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1818.09" cy="1463.6" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1855.02" cy="1453.7" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1891.95" cy="1443.96" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1928.88" cy="1434.39" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="1965.81" cy="1424.98" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="2002.74" cy="1415.72" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="2039.67" cy="1406.6" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="2076.6" cy="1397.61" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="2113.53" cy="1388.76" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="2150.47" cy="1380.03" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="2187.4" cy="1371.42" r="2"/>
+<circle clip-path="url(#clip513)" style="fill:#3da44d; stroke:none; fill-opacity:0" cx="2224.33" cy="1362.92" r="2"/>
+<path clip-path="url(#clip510)" d="M355.193 1597.33 L677.137 1597.33 L677.137 1389.97 L355.193 1389.97  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="355.193,1597.33 677.137,1597.33 677.137,1389.97 355.193,1389.97 355.193,1597.33 "/>
+<path clip-path="url(#clip510)" d="M378.153 1462.54 L515.916 1462.54 L515.916 1421.07 L378.153 1421.07 L378.153 1462.54  Z" fill="#009af9" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="378.153,1462.54 515.916,1462.54 515.916,1421.07 378.153,1421.07 378.153,1462.54 "/>
+<path clip-path="url(#clip510)" d="M538.877 1424.53 L543.553 1424.53 L543.553 1459.09 L538.877 1459.09 L538.877 1424.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M557.534 1442.58 L557.534 1455.24 L565.034 1455.24 Q568.807 1455.24 570.613 1453.69 Q572.441 1452.12 572.441 1448.9 Q572.441 1445.66 570.613 1444.13 Q568.807 1442.58 565.034 1442.58 L557.534 1442.58 M557.534 1428.37 L557.534 1438.78 L564.455 1438.78 Q567.881 1438.78 569.548 1437.51 Q571.238 1436.21 571.238 1433.58 Q571.238 1430.96 569.548 1429.66 Q567.881 1428.37 564.455 1428.37 L557.534 1428.37 M552.858 1424.53 L564.802 1424.53 Q570.15 1424.53 573.043 1426.75 Q575.937 1428.97 575.937 1433.07 Q575.937 1436.24 574.455 1438.11 Q572.974 1439.99 570.103 1440.45 Q573.552 1441.19 575.451 1443.55 Q577.372 1445.89 577.372 1449.41 Q577.372 1454.04 574.224 1456.56 Q571.076 1459.09 565.265 1459.09 L552.858 1459.09 L552.858 1424.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M585.381 1424.53 L592.349 1424.53 L601.168 1448.04 L610.034 1424.53 L617.001 1424.53 L617.001 1459.09 L612.441 1459.09 L612.441 1428.74 L603.529 1452.44 L598.83 1452.44 L589.918 1428.74 L589.918 1459.09 L585.381 1459.09 L585.381 1424.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M378.153 1514.38 L515.916 1514.38 L515.916 1472.91 L378.153 1472.91 L378.153 1514.38  Z" fill="#e26f46" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="378.153,1514.38 515.916,1514.38 515.916,1472.91 378.153,1472.91 378.153,1514.38 "/>
+<path clip-path="url(#clip510)" d="M538.877 1476.37 L543.599 1476.37 L550.867 1505.58 L558.113 1476.37 L563.367 1476.37 L570.636 1505.58 L577.881 1476.37 L582.626 1476.37 L573.946 1510.93 L568.066 1510.93 L560.775 1480.93 L553.414 1510.93 L547.534 1510.93 L538.877 1476.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M588.83 1476.37 L595.798 1476.37 L604.617 1499.88 L613.483 1476.37 L620.45 1476.37 L620.45 1510.93 L615.89 1510.93 L615.89 1480.58 L606.978 1504.28 L602.279 1504.28 L593.367 1480.58 L593.367 1510.93 L588.83 1510.93 L588.83 1476.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M624.941 1476.37 L654.177 1476.37 L654.177 1480.3 L641.909 1480.3 L641.909 1510.93 L637.209 1510.93 L637.209 1480.3 L624.941 1480.3 L624.941 1476.37 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M378.153 1566.22 L515.916 1566.22 L515.916 1524.75 L378.153 1524.75 L378.153 1566.22  Z" fill="#3da44d" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip510)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="378.153,1566.22 515.916,1566.22 515.916,1524.75 378.153,1524.75 378.153,1566.22 "/>
+<path clip-path="url(#clip510)" d="M561.122 1529.34 L561.122 1533.9 Q558.46 1532.63 556.099 1532 Q553.738 1531.38 551.539 1531.38 Q547.719 1531.38 545.636 1532.86 Q543.576 1534.34 543.576 1537.07 Q543.576 1539.36 544.941 1540.54 Q546.33 1541.7 550.173 1542.42 L552.997 1543 Q558.228 1543.99 560.705 1546.52 Q563.205 1549.02 563.205 1553.23 Q563.205 1558.25 559.826 1560.84 Q556.469 1563.44 549.965 1563.44 Q547.511 1563.44 544.733 1562.88 Q541.979 1562.33 539.016 1561.24 L539.016 1556.42 Q541.863 1558.02 544.594 1558.83 Q547.326 1559.64 549.965 1559.64 Q553.969 1559.64 556.145 1558.07 Q558.321 1556.49 558.321 1553.58 Q558.321 1551.03 556.747 1549.59 Q555.196 1548.16 551.631 1547.44 L548.784 1546.89 Q543.553 1545.84 541.215 1543.62 Q538.877 1541.4 538.877 1537.44 Q538.877 1532.86 542.094 1530.22 Q545.335 1527.58 551.006 1527.58 Q553.437 1527.58 555.96 1528.02 Q558.483 1528.46 561.122 1529.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M570.497 1528.21 L592.349 1528.21 L592.349 1532.14 L575.173 1532.14 L575.173 1542.37 L591.631 1542.37 L591.631 1546.31 L575.173 1546.31 L575.173 1558.83 L592.765 1558.83 L592.765 1562.77 L570.497 1562.77 L570.497 1528.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M600.45 1528.21 L605.126 1528.21 L605.126 1542.37 L622.117 1542.37 L622.117 1528.21 L626.793 1528.21 L626.793 1562.77 L622.117 1562.77 L622.117 1546.31 L605.126 1546.31 L605.126 1562.77 L600.45 1562.77 L600.45 1528.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip510)" d="M636.098 1528.21 L640.774 1528.21 L640.774 1562.77 L636.098 1562.77 L636.098 1528.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /></svg>
diff --git a/previews/PR3525/tutorials/nonlinear/portfolio/index.html b/previews/PR3525/tutorials/nonlinear/portfolio/index.html
new file mode 100644
index 00000000000..24a0713ba32
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/portfolio/index.html
@@ -0,0 +1,136 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Portfolio optimization · JuMP</title><meta name="title" content="Portfolio optimization · JuMP"/><meta property="og:title" content="Portfolio optimization · JuMP"/><meta property="twitter:title" content="Portfolio optimization · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li class="is-active"><a class="tocitem" href>Portfolio optimization</a><ul class="internal"><li><a class="tocitem" href="#Required-packages"><span>Required packages</span></a></li><li><a class="tocitem" href="#Formulation"><span>Formulation</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Multi-objective-portfolio-optimization"><span>Multi-objective portfolio optimization</span></a></li></ul></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Portfolio optimization</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Portfolio optimization</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/portfolio.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Portfolio-optimization"><a class="docs-heading-anchor" href="#Portfolio-optimization">Portfolio optimization</a><a id="Portfolio-optimization-1"></a><a class="docs-heading-anchor-permalink" href="#Portfolio-optimization" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../portfolio.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Arpit Bhatia.</strong></p><p>Optimization models play an increasingly important role in financial decisions. Many computational finance problems can be solved efficiently using modern optimization techniques.</p><p>This tutorial solves the famous Markowitz Portfolio Optimization problem with data from <a href="https://www2.isye.gatech.edu/~sahmed/isye6669/">lecture notes from a course taught at Georgia Tech by Shabbir Ahmed</a>.</p><h2 id="Required-packages"><a class="docs-heading-anchor" href="#Required-packages">Required packages</a><a id="Required-packages-1"></a><a class="docs-heading-anchor-permalink" href="#Required-packages" title="Permalink"></a></h2><p>This tutorial uses the following packages</p><pre><code class="language-julia hljs">using JuMP
+import DataFrames
+import Ipopt
+import MultiObjectiveAlgorithms as MOA
+import Plots
+import Statistics
+import StatsPlots</code></pre><h2 id="Formulation"><a class="docs-heading-anchor" href="#Formulation">Formulation</a><a id="Formulation-1"></a><a class="docs-heading-anchor-permalink" href="#Formulation" title="Permalink"></a></h2><p>Suppose we are considering investing 1000 dollars in three non-dividend paying stocks, IBM (IBM), Walmart (WMT), and Southern Electric (SEHI), for a one-month period.</p><p>We will use the initial money to buy shares of the three stocks at the current market prices, hold these for one month, and sell the shares off at the prevailing market prices at the end of the month.</p><p>As a rational investor, we hope to make some profit out of this endeavor, that is, the return on our investment should be positive.</p><p>Suppose we bought a stock at <span>$p$</span> dollars per share in the beginning of the month, and sold it off at <span>$s$</span> dollars per share at the end of the month. Then the one-month return on a share of the stock is <span>$\frac{s-p}{p}$</span>.</p><p>Since the stock prices are quite uncertain, so is the end-of-month return on our investment. Our goal is to invest in such a way that the expected end-of-month return is at least <span>$</span>50 or 5%. Furthermore, we want to make sure that the “risk” of not achieving our desired return is minimum.</p><p>Note that we are solving the problem under the following assumptions:</p><ol><li>We can trade any continuum of shares.</li><li>No short-selling is allowed.</li><li>There are no transaction costs.</li></ol><p>We model this problem by taking decision variables <span>$x_{i}, i=1,2,3,$</span> denoting the dollars invested in each of the 3 stocks.</p><p>Let us denote by <span>$\tilde{r}_{i}$</span> the random variable corresponding to the monthly return (increase in the stock price) per dollar for stock <span>$i$</span>.</p><p>Then, the return (or profit) on <span>$x_{i}$</span> dollars invested in stock <span>$i$</span> is <span>$\tilde{r}_{i} x_{i},$</span> and the total (random) return on our investment is <span>$\sum_{i=1}^{3} \tilde{r}_{i} x_{i}.$</span> The expected return on our investment is then <span>$\mathbb{E}\left[\sum_{i=1}^{3} \tilde{r}_{i} x_{i}\right]=\sum_{i=1}^{3} \overline{r}_{i} x_{i},$</span> where <span>$\overline{r}_{i}$</span> is the expected value of the <span>$\tilde{r}_{i}.$</span></p><p>Now we need to quantify the notion of “risk” in our investment.</p><p>Markowitz, in his Nobel prize winning work, showed that a rational investor’s notion of minimizing risk can be closely approximated by minimizing the variance of the return of the investment portfolio. This variance is given by:</p><p class="math-container">\[\operatorname{Var}\left[\sum_{i=1}^{3} \tilde{r}_{i} x_{i}\right] = \sum_{i=1}^{3} \sum_{j=1}^{3} x_{i} x_{j} \sigma_{i j}\]</p><p>where <span>$\sigma_{i j}$</span> is the covariance of the return of stock <span>$i$</span> with stock <span>$j$</span>.</p><p>Note that the right hand side of the equation is the most reduced form of the expression and we have not shown the intermediate steps involved in getting to this form. We can also write this equation as:</p><p class="math-container">\[\operatorname{Var}\left[\sum_{i=1}^{3} \tilde{r}_{i} x_{i}\right] =x^\top Q x\]</p><p>Where <span>$Q$</span> is the covariance matrix for the random vector <span>$\tilde{r}$</span>.</p><p>Finally, we can write the model as:</p><p class="math-container">\[\begin{aligned}
+\min x^\top Q x \\
+\text { s.t. } \sum_{i=1}^{3} x_{i} \leq 1000 \\
+\overline{r}^\top x \geq 50 \\
+x \geq 0
+\end{aligned}\]</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>For the data in our problem, we use the stock prices given below, in monthly values from November 2000, through November 2001.</p><pre><code class="language-julia hljs">df = DataFrames.DataFrame(
+    [
+        93.043 51.826 1.063
+        84.585 52.823 0.938
+        111.453 56.477 1.000
+        99.525 49.805 0.938
+        95.819 50.287 1.438
+        114.708 51.521 1.700
+        111.515 51.531 2.540
+        113.211 48.664 2.390
+        104.942 55.744 3.120
+        99.827 47.916 2.980
+        91.607 49.438 1.900
+        107.937 51.336 1.750
+        115.590 55.081 1.800
+    ],
+    [:IBM, :WMT, :SEHI],
+)</code></pre><div><div style = "float: left;"><span>13×3 DataFrame</span></div><div style = "clear: both;"></div></div><div class = "data-frame" style = "overflow-x: scroll;"><table class = "data-frame" style = "margin-bottom: 6px;"><thead><tr class = "header"><th class = "rowNumber" style = "font-weight: bold; text-align: right;">Row</th><th style = "text-align: left;">IBM</th><th style = "text-align: left;">WMT</th><th style = "text-align: left;">SEHI</th></tr><tr class = "subheader headerLastRow"><th class = "rowNumber" style = "font-weight: bold; text-align: right;"></th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th><th title = "Float64" style = "text-align: left;">Float64</th></tr></thead><tbody><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">1</td><td style = "text-align: right;">93.043</td><td style = "text-align: right;">51.826</td><td style = "text-align: right;">1.063</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">2</td><td style = "text-align: right;">84.585</td><td style = "text-align: right;">52.823</td><td style = "text-align: right;">0.938</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">3</td><td style = "text-align: right;">111.453</td><td style = "text-align: right;">56.477</td><td style = "text-align: right;">1.0</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">4</td><td style = "text-align: right;">99.525</td><td style = "text-align: right;">49.805</td><td style = "text-align: right;">0.938</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">5</td><td style = "text-align: right;">95.819</td><td style = "text-align: right;">50.287</td><td style = "text-align: right;">1.438</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">6</td><td style = "text-align: right;">114.708</td><td style = "text-align: right;">51.521</td><td style = "text-align: right;">1.7</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">7</td><td style = "text-align: right;">111.515</td><td style = "text-align: right;">51.531</td><td style = "text-align: right;">2.54</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">8</td><td style = "text-align: right;">113.211</td><td style = "text-align: right;">48.664</td><td style = "text-align: right;">2.39</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">9</td><td style = "text-align: right;">104.942</td><td style = "text-align: right;">55.744</td><td style = "text-align: right;">3.12</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">10</td><td style = "text-align: right;">99.827</td><td style = "text-align: right;">47.916</td><td style = "text-align: right;">2.98</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">11</td><td style = "text-align: right;">91.607</td><td style = "text-align: right;">49.438</td><td style = "text-align: right;">1.9</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">12</td><td style = "text-align: right;">107.937</td><td style = "text-align: right;">51.336</td><td style = "text-align: right;">1.75</td></tr><tr><td class = "rowNumber" style = "font-weight: bold; text-align: right;">13</td><td style = "text-align: right;">115.59</td><td style = "text-align: right;">55.081</td><td style = "text-align: right;">1.8</td></tr></tbody></table></div><p>Next, we compute the percentage return for the stock in each month:</p><pre><code class="language-julia hljs">returns = diff(Matrix(df); dims = 1) ./ Matrix(df[1:end-1, :])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">12×3 Matrix{Float64}:
+ -0.0909042   0.0192374    -0.117592
+  0.317645    0.0691744     0.0660981
+ -0.107023   -0.118137     -0.062
+ -0.0372369   0.00967774    0.533049
+  0.197132    0.0245391     0.182197
+ -0.0278359   0.000194096   0.494118
+  0.0152087  -0.0556364    -0.0590551
+ -0.0730406   0.145487      0.305439
+ -0.0487412  -0.140428     -0.0448718
+ -0.0823425   0.0317639    -0.362416
+  0.178261    0.0383915    -0.0789474
+  0.0709025   0.0729508     0.0285714</code></pre><p>The expected monthly return is:</p><pre><code class="language-julia hljs">r = vec(Statistics.mean(returns; dims = 1))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{Float64}:
+ 0.026002150277777348
+ 0.008101316405671459
+ 0.07371590949198982</code></pre><p>and the covariance matrix is:</p><pre><code class="language-julia hljs">Q = Statistics.cov(returns)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3×3 Matrix{Float64}:
+ 0.018641    0.00359853  0.00130976
+ 0.00359853  0.00643694  0.00488727
+ 0.00130976  0.00488727  0.0686828</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[1:3] &gt;= 0)
+@objective(model, Min, x&#39; * Q * x)
+@constraint(model, sum(x) &lt;= 1000)
+@constraint(model, r&#39; * x &gt;= 50)
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Ipopt
+
+* Status
+  Result count       : 1
+  Termination status : LOCALLY_SOLVED
+  Message from the solver:
+  &quot;Solve_Succeeded&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 2.26344e+04
+  Dual objective value : 4.52688e+04
+
+* Work counters
+  Solve time (sec)   : 5.27906e-03
+</code></pre><p>The optimal allocation of our assets is:</p><pre><code class="language-julia hljs">value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">3-element Vector{Float64}:
+ 497.04552984986407
+  -9.670578501816873e-9
+ 502.9544801594808</code></pre><p>So we spend <span>$</span>497 on IBM, and <span>$</span>503 on SEHI. This results in a variance of:</p><pre><code class="language-julia hljs">scalar_variance = value(x&#39; * Q * x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">22634.41784988414</code></pre><p>and an expected return of:</p><pre><code class="language-julia hljs">scalar_return = value(r&#39; * x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">49.99999950000236</code></pre><h2 id="Multi-objective-portfolio-optimization"><a class="docs-heading-anchor" href="#Multi-objective-portfolio-optimization">Multi-objective portfolio optimization</a><a id="Multi-objective-portfolio-optimization-1"></a><a class="docs-heading-anchor-permalink" href="#Multi-objective-portfolio-optimization" title="Permalink"></a></h2><p>The previous model returned a single solution that minimized the variance, ensuring that our expected return was at least <span>$</span>50. In practice, we might be willing to accept a slightly higher variance if it meant a much increased expected return. To explore this problem space, we can instead formulate our portfolio optimization problem with two objectives:</p><ol><li>to minimize the variance</li><li>to maximize the expected return</li></ol><p>The solution to this bi-objective problem is the <a href="https://en.wikipedia.org/wiki/Efficient_frontier">efficient frontier</a> of modern portfolio theory, and each point in the solution is a point with the best return for a fixed level of risk.</p><pre><code class="language-julia hljs">model = Model(() -&gt; MOA.Optimizer(Ipopt.Optimizer))
+set_silent(model)</code></pre><p>We also need to choose a solution algorithm for <code>MOA</code>. For our problem, the efficient frontier will have an infinite number of solutions. Since we cannot find all of the solutions, we choose an approximation algorithm and limit the number of solution points that are returned:</p><pre><code class="language-julia hljs">set_optimizer_attribute(model, MOA.Algorithm(), MOA.EpsilonConstraint())
+set_optimizer_attribute(model, MOA.SolutionLimit(), 50)</code></pre><p>Now we can define the rest of the model:</p><pre><code class="language-julia hljs">@variable(model, x[1:3] &gt;= 0)
+@constraint(model, sum(x) &lt;= 1000)
+@expression(model, variance, x&#39; * Q * x)
+@expression(model, expected_return, r&#39; * x)
+# We want to minimize variance and maximize expected return, but we must pick
+# a single objective sense `Min`, and negate any `Max` objectives:
+@objective(model, Min, [variance, -expected_return])
+optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : MOA[algorithm=MultiObjectiveAlgorithms.EpsilonConstraint, optimizer=Ipopt]
+
+* Status
+  Result count       : 50
+  Termination status : OPTIMAL
+  Message from the solver:
+  &quot;Solve complete. Found 50 solution(s)&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : NO_SOLUTION
+  Objective value    : [2.58170e-08,-5.36777e-05]
+  Objective bound    : [5.78303e-09,-7.37159e+01]
+  Dual objective value : -7.37159e+01
+
+* Work counters
+  Solve time (sec)   : 3.98179e-01
+</code></pre><p>The algorithm found 50 different solutions. Let&#39;s plot them to see how they differ:</p><pre><code class="language-julia hljs">objective_space = Plots.hline(
+    [scalar_return];
+    label = &quot;Single-objective solution&quot;,
+    linecolor = :red,
+)
+Plots.vline!(objective_space, [scalar_variance]; label = &quot;&quot;, linecolor = :red)
+Plots.scatter!(
+    objective_space,
+    [value(variance; result = i) for i in 1:result_count(model)],
+    [value(expected_return; result = i) for i in 1:result_count(model)];
+    xlabel = &quot;Variance&quot;,
+    ylabel = &quot;Expected Return&quot;,
+    label = &quot;&quot;,
+    title = &quot;Objective space&quot;,
+    markercolor = &quot;white&quot;,
+    markersize = 5,
+    legend = :bottomright,
+)
+for i in 1:result_count(model)
+    y = objective_value(model; result = i)
+    Plots.annotate!(objective_space, y[1], -y[2], (i, 3))
+end
+
+decision_space = StatsPlots.groupedbar(
+    vcat([value.(x; result = i)&#39; for i in 1:result_count(model)]...);
+    bar_position = :stack,
+    label = [&quot;IBM&quot; &quot;WMT&quot; &quot;SEHI&quot;],
+    xlabel = &quot;Solution #&quot;,
+    ylabel = &quot;Investment (\$)&quot;,
+    title = &quot;Decision space&quot;,
+)
+Plots.plot(objective_space, decision_space; layout = (2, 1), size = (600, 600))</code></pre><img src="319d794d.svg" alt="Example block output"/><p>Perhaps our trade-off wasn&#39;t so bad after all. Our original solution corresponded to picking a solution #17. If we buy more SEHI, we can increase the return, but the variance also increases. If we buy less SEHI, such as a solution like #5 or #6, then we can achieve the corresponding return without deploying all of our capital. We should also note that at no point should we buy WMT.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../simple_examples/">« Simple examples</a><a class="docs-footer-nextpage" href="../rocket_control/">Rocket Control »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/querying_hessians.jl b/previews/PR3525/tutorials/nonlinear/querying_hessians.jl
new file mode 100644
index 00000000000..d4a44ee72cc
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/querying_hessians.jl
@@ -0,0 +1,257 @@
+# Copyright (c) 2022 Oscar Dowson and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # Computing Hessians
+
+# The purpose of this tutorial is to demonstrate how to compute the Hessian of
+# the Lagrangian of a nonlinear program.
+
+# !!! warning
+#     This is an advanced tutorial that interacts with the low-level nonlinear
+#     interface of MathOptInterface.
+#
+#     By default, JuMP exports the `MOI` symbol as an alias for the
+#     MathOptInterface.jl package. We recommend making this more explicit in
+#     your code by adding the following lines:
+#     ```julia
+#     import MathOptInterface as MOI
+#     ```
+
+# Given a nonlinear program:
+# ```math
+# \begin{align}
+# & \min_{x \in \mathbb{R}^n} & f(x) \\
+# & \;\;\text{s.t.} & l \le g_i(x) \le u
+# \end{align}
+# ```
+# the Hessian of the Lagrangian is computed as:
+# ```math
+# H(x, \sigma, \mu) = \sigma\nabla^2 f(x) + \sum_{i=1}^m \mu_i \nabla^2 g_i(x)
+# ```
+# where ``x`` is a primal point, ``\sigma`` is a scalar (typically ``1``), and
+# ``\mu`` is a vector of weights corresponding to the Lagrangian dual of the
+# constraints.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import Ipopt
+import LinearAlgebra
+import Random
+import SparseArrays
+
+# ## The basic model
+
+# To demonstrate how to interact with the lower-level nonlinear interface, we
+# need an example model. The exact model isn't important; we use the model from
+# [The Rosenbrock function](@ref) tutorial, with some additional constraints to
+# demonstrate various features of the lower-level interface.
+
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[i = 1:2], start = -i)
+@constraint(model, g_1, x[1]^2 <= 1)
+@constraint(model, g_2, (x[1] + x[2])^2 <= 2)
+@objective(model, Min, (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2)
+optimize!(model)
+
+# ## The analytic solution
+
+# With a little work, it is possible to analytically derive the correct hessian:
+
+function analytic_hessian(x, σ, μ)
+    g_1_H = [2.0 0.0; 0.0 0.0]
+    g_2_H = [2.0 2.0; 2.0 2.0]
+    f_H = zeros(2, 2)
+    f_H[1, 1] = 2.0 + 1200.0 * x[1]^2 - 400.0 * x[2]
+    f_H[1, 2] = f_H[2, 1] = -400.0 * x[1]
+    f_H[2, 2] = 200.0
+    return σ * f_H + μ' * [g_1_H, g_2_H]
+end
+
+# Here are various points:
+
+analytic_hessian([1, 1], 0, [0, 0])
+
+#-
+
+analytic_hessian([1, 1], 0, [1, 0])
+
+#-
+
+analytic_hessian([1, 1], 0, [0, 1])
+
+#-
+
+analytic_hessian([1, 1], 1, [0, 0])
+
+# ## Create a nonlinear model
+
+# JuMP delegates automatic differentiation to the `MOI.Nonlinear` submodule.
+# Therefore, to compute the Hessian of the Lagrangian, we need to create a
+# [`MOI.Nonlinear.Model`](@ref) object:
+
+rows = Any[]
+nlp = MOI.Nonlinear.Model()
+for (F, S) in list_of_constraint_types(model)
+    if F <: VariableRef
+        continue  # Skip variable bounds
+    end
+    for ci in all_constraints(model, F, S)
+        push!(rows, ci)
+        object = constraint_object(ci)
+        MOI.Nonlinear.add_constraint(nlp, object.func, object.set)
+    end
+end
+MOI.Nonlinear.set_objective(nlp, objective_function(model))
+nlp
+
+# It is important that we save the constraint indices in a vector `rows`, so
+# that we know the order of the constraints in the nonlinear model.
+
+# Next, we need to convert our model into an [`MOI.Nonlinear.Evaluator`](@ref),
+# specifying an automatic differentiation backend. In this case, we use
+# [`MOI.Nonlinear.SparseReverseMode`](@ref):
+
+evaluator = MOI.Nonlinear.Evaluator(
+    nlp,
+    MOI.Nonlinear.SparseReverseMode(),
+    index.(all_variables(model)),
+)
+
+# Before computing anything with the evaluator, we need to initialize it.
+# Use [`MOI.features_available`](@ref) to see what we can query:
+
+MOI.features_available(evaluator)
+
+# Consult the MOI documentation for specifics, but to obtain the Hessian matrix,
+# we need to initialize `:Hess`:
+
+MOI.initialize(evaluator, [:Hess])
+
+# MOI represents the Hessian as a sparse matrix. Get the sparsity pattern as
+# follows:
+
+hessian_sparsity = MOI.hessian_lagrangian_structure(evaluator)
+
+# The sparsity pattern has a few properties of interest:
+# * Each element `(i, j)` indicates a structural non-zero in row `i` and column
+#   `j`
+# * The list may contain duplicates, in which case we should add the values
+#   together
+# * The list does not need to be sorted
+# * The list may contain any mix of lower- or upper-triangular indices
+# This format matches Julia's sparse-triplet form of a SparseArray, so we can
+# convert from the sparse Hessian representation to a Julia SparseArray as
+# follows:
+
+I = [i for (i, _) in hessian_sparsity]
+J = [j for (_, j) in hessian_sparsity]
+V = zeros(length(hessian_sparsity))
+n = num_variables(model)
+H = SparseArrays.sparse(I, J, V, n, n)
+
+# Of course, knowing where the zeros are isn't very interesting. We really want
+# to compute the value of the Hessian matrix at a point.
+
+MOI.eval_hessian_lagrangian(evaluator, V, ones(n), 1.0, ones(length(rows)))
+H = SparseArrays.sparse(I, J, V, n, n)
+
+# In practice, we often want to compute the value of the hessian at the optimal
+# solution.
+
+# First, we compute the primal solution. To do so, we need a vector of the
+# variables in the order that they were passed to the solver:
+
+x = all_variables(model)
+
+# Here `x[1]` is the variable that corresponds to column 1, and so on. Here's
+# the optimal primal solution:
+
+x_optimal = value.(x)
+
+# Next, we need the optimal dual solution associated with the nonlinear
+# constraints (this is where it is important to record the order of the
+# constraints as we added them to `nlp`):
+
+y_optimal = dual.(rows)
+
+# Now we can compute the Hessian at the optimal primal-dual point:
+
+MOI.eval_hessian_lagrangian(evaluator, V, x_optimal, 1.0, y_optimal)
+H = SparseArrays.sparse(I, J, V, n, n)
+
+# However, this Hessian isn't quite right because it isn't symmetric. We can fix
+# this by filling in the appropriate off-diagonal terms:
+
+function fill_off_diagonal(H)
+    ret = H + H'
+    row_vals = SparseArrays.rowvals(ret)
+    non_zeros = SparseArrays.nonzeros(ret)
+    for col in 1:size(ret, 2)
+        for i in SparseArrays.nzrange(ret, col)
+            if col == row_vals[i]
+                non_zeros[i] /= 2
+            end
+        end
+    end
+    return ret
+end
+
+fill_off_diagonal(H)
+
+# Putting everything together:
+
+function compute_optimal_hessian(model::Model)
+    rows = Any[]
+    nlp = MOI.Nonlinear.Model()
+    for (F, S) in list_of_constraint_types(model)
+        for ci in all_constraints(model, F, S)
+            push!(rows, ci)
+            object = constraint_object(ci)
+            MOI.Nonlinear.add_constraint(nlp, object.func, object.set)
+        end
+    end
+    MOI.Nonlinear.set_objective(nlp, objective_function(model))
+    x = all_variables(model)
+    backend = MOI.Nonlinear.SparseReverseMode()
+    evaluator = MOI.Nonlinear.Evaluator(nlp, backend, index.(x))
+    MOI.initialize(evaluator, [:Hess])
+    hessian_sparsity = MOI.hessian_lagrangian_structure(evaluator)
+    I = [i for (i, _) in hessian_sparsity]
+    J = [j for (_, j) in hessian_sparsity]
+    V = zeros(length(hessian_sparsity))
+    MOI.eval_hessian_lagrangian(evaluator, V, value.(x), 1.0, dual.(rows))
+    H = SparseArrays.sparse(I, J, V, length(x), length(x))
+    return Matrix(fill_off_diagonal(H))
+end
+
+H_star = compute_optimal_hessian(model)
+
+# If we compare our solution against the analytical solution:
+
+analytic_hessian(value.(x), 1.0, dual.([g_1, g_2]))
+
+# If we look at the eigenvalues of the Hessian:
+
+LinearAlgebra.eigvals(H_star)
+
+# we see that they are all positive. Therefore, the Hessian is positive
+# definite, and so the solution found by Ipopt is a local minimizer.
diff --git a/previews/PR3525/tutorials/nonlinear/querying_hessians/index.html b/previews/PR3525/tutorials/nonlinear/querying_hessians/index.html
new file mode 100644
index 00000000000..5aaf2a3cb50
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/querying_hessians/index.html
@@ -0,0 +1,139 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Computing Hessians · JuMP</title><meta name="title" content="Computing Hessians · JuMP"/><meta property="og:title" content="Computing Hessians · JuMP"/><meta property="twitter:title" content="Computing Hessians · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li class="is-active"><a class="tocitem" href>Computing Hessians</a><ul class="internal"><li><a class="tocitem" href="#The-basic-model"><span>The basic model</span></a></li><li><a class="tocitem" href="#The-analytic-solution"><span>The analytic solution</span></a></li><li><a class="tocitem" href="#Create-a-nonlinear-model"><span>Create a nonlinear model</span></a></li></ul></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Computing Hessians</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Computing Hessians</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/querying_hessians.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Computing-Hessians"><a class="docs-heading-anchor" href="#Computing-Hessians">Computing Hessians</a><a id="Computing-Hessians-1"></a><a class="docs-heading-anchor-permalink" href="#Computing-Hessians" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../querying_hessians.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>The purpose of this tutorial is to demonstrate how to compute the Hessian of the Lagrangian of a nonlinear program.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>This is an advanced tutorial that interacts with the low-level nonlinear interface of MathOptInterface.</p><p>By default, JuMP exports the <code>MOI</code> symbol as an alias for the MathOptInterface.jl package. We recommend making this more explicit in your code by adding the following lines:</p><pre><code class="language-julia hljs">import MathOptInterface as MOI</code></pre></div></div><p>Given a nonlinear program:</p><p class="math-container">\[\begin{align}
+&amp; \min_{x \in \mathbb{R}^n} &amp; f(x) \\
+&amp; \;\;\text{s.t.} &amp; l \le g_i(x) \le u
+\end{align}\]</p><p>the Hessian of the Lagrangian is computed as:</p><p class="math-container">\[H(x, \sigma, \mu) = \sigma\nabla^2 f(x) + \sum_{i=1}^m \mu_i \nabla^2 g_i(x)\]</p><p>where <span>$x$</span> is a primal point, <span>$\sigma$</span> is a scalar (typically <span>$1$</span>), and <span>$\mu$</span> is a vector of weights corresponding to the Lagrangian dual of the constraints.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import Ipopt
+import LinearAlgebra
+import Random
+import SparseArrays</code></pre><h2 id="The-basic-model"><a class="docs-heading-anchor" href="#The-basic-model">The basic model</a><a id="The-basic-model-1"></a><a class="docs-heading-anchor-permalink" href="#The-basic-model" title="Permalink"></a></h2><p>To demonstrate how to interact with the lower-level nonlinear interface, we need an example model. The exact model isn&#39;t important; we use the model from <a href="../simple_examples/#The-Rosenbrock-function">The Rosenbrock function</a> tutorial, with some additional constraints to demonstrate various features of the lower-level interface.</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[i = 1:2], start = -i)
+@constraint(model, g_1, x[1]^2 &lt;= 1)
+@constraint(model, g_2, (x[1] + x[2])^2 &lt;= 2)
+@objective(model, Min, (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2)
+optimize!(model)</code></pre><h2 id="The-analytic-solution"><a class="docs-heading-anchor" href="#The-analytic-solution">The analytic solution</a><a id="The-analytic-solution-1"></a><a class="docs-heading-anchor-permalink" href="#The-analytic-solution" title="Permalink"></a></h2><p>With a little work, it is possible to analytically derive the correct hessian:</p><pre><code class="language-julia hljs">function analytic_hessian(x, σ, μ)
+    g_1_H = [2.0 0.0; 0.0 0.0]
+    g_2_H = [2.0 2.0; 2.0 2.0]
+    f_H = zeros(2, 2)
+    f_H[1, 1] = 2.0 + 1200.0 * x[1]^2 - 400.0 * x[2]
+    f_H[1, 2] = f_H[2, 1] = -400.0 * x[1]
+    f_H[2, 2] = 200.0
+    return σ * f_H + μ&#39; * [g_1_H, g_2_H]
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">analytic_hessian (generic function with 1 method)</code></pre><p>Here are various points:</p><pre><code class="language-julia hljs">analytic_hessian([1, 1], 0, [0, 0])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+ 0.0  0.0
+ 0.0  0.0</code></pre><pre><code class="language-julia hljs">analytic_hessian([1, 1], 0, [1, 0])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+ 2.0  0.0
+ 0.0  0.0</code></pre><pre><code class="language-julia hljs">analytic_hessian([1, 1], 0, [0, 1])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+ 2.0  2.0
+ 2.0  2.0</code></pre><pre><code class="language-julia hljs">analytic_hessian([1, 1], 1, [0, 0])</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  802.0  -400.0
+ -400.0   200.0</code></pre><h2 id="Create-a-nonlinear-model"><a class="docs-heading-anchor" href="#Create-a-nonlinear-model">Create a nonlinear model</a><a id="Create-a-nonlinear-model-1"></a><a class="docs-heading-anchor-permalink" href="#Create-a-nonlinear-model" title="Permalink"></a></h2><p>JuMP delegates automatic differentiation to the <code>MOI.Nonlinear</code> submodule. Therefore, to compute the Hessian of the Lagrangian, we need to create a <a href="../../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Model"><code>MOI.Nonlinear.Model</code></a> object:</p><pre><code class="language-julia hljs">rows = Any[]
+nlp = MOI.Nonlinear.Model()
+for (F, S) in list_of_constraint_types(model)
+    if F &lt;: VariableRef
+        continue  # Skip variable bounds
+    end
+    for ci in all_constraints(model, F, S)
+        push!(rows, ci)
+        object = constraint_object(ci)
+        MOI.Nonlinear.add_constraint(nlp, object.func, object.set)
+    end
+end
+MOI.Nonlinear.set_objective(nlp, objective_function(model))
+nlp</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">A Nonlinear.Model with:
+ 1 objective
+ 0 parameters
+ 0 expressions
+ 2 constraints</code></pre><p>It is important that we save the constraint indices in a vector <code>rows</code>, so that we know the order of the constraints in the nonlinear model.</p><p>Next, we need to convert our model into an <a href="../../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.Evaluator"><code>MOI.Nonlinear.Evaluator</code></a>, specifying an automatic differentiation backend. In this case, we use <a href="../../../moi/submodules/Nonlinear/reference/#MathOptInterface.Nonlinear.SparseReverseMode"><code>MOI.Nonlinear.SparseReverseMode</code></a>:</p><pre><code class="language-julia hljs">evaluator = MOI.Nonlinear.Evaluator(
+    nlp,
+    MOI.Nonlinear.SparseReverseMode(),
+    index.(all_variables(model)),
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Nonlinear.Evaluator with available features:
+  * :Grad
+  * :Jac
+  * :JacVec
+  * :Hess
+  * :HessVec
+  * :ExprGraph</code></pre><p>Before computing anything with the evaluator, we need to initialize it. Use <a href="../../../moi/reference/nonlinear/#MathOptInterface.features_available"><code>MOI.features_available</code></a> to see what we can query:</p><pre><code class="language-julia hljs">MOI.features_available(evaluator)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">6-element Vector{Symbol}:
+ :Grad
+ :Jac
+ :JacVec
+ :Hess
+ :HessVec
+ :ExprGraph</code></pre><p>Consult the MOI documentation for specifics, but to obtain the Hessian matrix, we need to initialize <code>:Hess</code>:</p><pre><code class="language-julia hljs">MOI.initialize(evaluator, [:Hess])</code></pre><p>MOI represents the Hessian as a sparse matrix. Get the sparsity pattern as follows:</p><pre><code class="language-julia hljs">hessian_sparsity = MOI.hessian_lagrangian_structure(evaluator)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">7-element Vector{Tuple{Int64, Int64}}:
+ (1, 1)
+ (2, 2)
+ (2, 1)
+ (1, 1)
+ (1, 1)
+ (2, 2)
+ (2, 1)</code></pre><p>The sparsity pattern has a few properties of interest:</p><ul><li>Each element <code>(i, j)</code> indicates a structural non-zero in row <code>i</code> and column <code>j</code></li><li>The list may contain duplicates, in which case we should add the values together</li><li>The list does not need to be sorted</li><li>The list may contain any mix of lower- or upper-triangular indices</li></ul><p>This format matches Julia&#39;s sparse-triplet form of a SparseArray, so we can convert from the sparse Hessian representation to a Julia SparseArray as follows:</p><pre><code class="language-julia hljs">I = [i for (i, _) in hessian_sparsity]
+J = [j for (_, j) in hessian_sparsity]
+V = zeros(length(hessian_sparsity))
+n = num_variables(model)
+H = SparseArrays.sparse(I, J, V, n, n)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 SparseArrays.SparseMatrixCSC{Float64, Int64} with 3 stored entries:
+ 0.0   ⋅ 
+ 0.0  0.0</code></pre><p>Of course, knowing where the zeros are isn&#39;t very interesting. We really want to compute the value of the Hessian matrix at a point.</p><pre><code class="language-julia hljs">MOI.eval_hessian_lagrangian(evaluator, V, ones(n), 1.0, ones(length(rows)))
+H = SparseArrays.sparse(I, J, V, n, n)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 SparseArrays.SparseMatrixCSC{Float64, Int64} with 3 stored entries:
+  806.0     ⋅ 
+ -398.0  202.0</code></pre><p>In practice, we often want to compute the value of the hessian at the optimal solution.</p><p>First, we compute the primal solution. To do so, we need a vector of the variables in the order that they were passed to the solver:</p><pre><code class="language-julia hljs">x = all_variables(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{VariableRef}:
+ x[1]
+ x[2]</code></pre><p>Here <code>x[1]</code> is the variable that corresponds to column 1, and so on. Here&#39;s the optimal primal solution:</p><pre><code class="language-julia hljs">x_optimal = value.(x)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ 0.7903587565231842
+ 0.6238546272155127</code></pre><p>Next, we need the optimal dual solution associated with the nonlinear constraints (this is where it is important to record the order of the constraints as we added them to <code>nlp</code>):</p><pre><code class="language-julia hljs">y_optimal = dual.(rows)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+ -8.038451738599348e-8
+ -0.05744089305771262</code></pre><p>Now we can compute the Hessian at the optimal primal-dual point:</p><pre><code class="language-julia hljs">MOI.eval_hessian_lagrangian(evaluator, V, x_optimal, 1.0, y_optimal)
+H = SparseArrays.sparse(I, J, V, n, n)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 SparseArrays.SparseMatrixCSC{Float64, Int64} with 3 stored entries:
+  501.944     ⋅ 
+ -316.258  199.885</code></pre><p>However, this Hessian isn&#39;t quite right because it isn&#39;t symmetric. We can fix this by filling in the appropriate off-diagonal terms:</p><pre><code class="language-julia hljs">function fill_off_diagonal(H)
+    ret = H + H&#39;
+    row_vals = SparseArrays.rowvals(ret)
+    non_zeros = SparseArrays.nonzeros(ret)
+    for col in 1:size(ret, 2)
+        for i in SparseArrays.nzrange(ret, col)
+            if col == row_vals[i]
+                non_zeros[i] /= 2
+            end
+        end
+    end
+    return ret
+end
+
+fill_off_diagonal(H)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 SparseArrays.SparseMatrixCSC{Float64, Int64} with 4 stored entries:
+  501.944  -316.258
+ -316.258   199.885</code></pre><p>Putting everything together:</p><pre><code class="language-julia hljs">function compute_optimal_hessian(model::Model)
+    rows = Any[]
+    nlp = MOI.Nonlinear.Model()
+    for (F, S) in list_of_constraint_types(model)
+        for ci in all_constraints(model, F, S)
+            push!(rows, ci)
+            object = constraint_object(ci)
+            MOI.Nonlinear.add_constraint(nlp, object.func, object.set)
+        end
+    end
+    MOI.Nonlinear.set_objective(nlp, objective_function(model))
+    x = all_variables(model)
+    backend = MOI.Nonlinear.SparseReverseMode()
+    evaluator = MOI.Nonlinear.Evaluator(nlp, backend, index.(x))
+    MOI.initialize(evaluator, [:Hess])
+    hessian_sparsity = MOI.hessian_lagrangian_structure(evaluator)
+    I = [i for (i, _) in hessian_sparsity]
+    J = [j for (_, j) in hessian_sparsity]
+    V = zeros(length(hessian_sparsity))
+    MOI.eval_hessian_lagrangian(evaluator, V, value.(x), 1.0, dual.(rows))
+    H = SparseArrays.sparse(I, J, V, length(x), length(x))
+    return Matrix(fill_off_diagonal(H))
+end
+
+H_star = compute_optimal_hessian(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  501.944  -316.258
+ -316.258   199.885</code></pre><p>If we compare our solution against the analytical solution:</p><pre><code class="language-julia hljs">analytic_hessian(value.(x), 1.0, dual.([g_1, g_2]))</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2×2 Matrix{Float64}:
+  501.944  -316.258
+ -316.258   199.885</code></pre><p>If we look at the eigenvalues of the Hessian:</p><pre><code class="language-julia hljs">LinearAlgebra.eigvals(H_star)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Float64}:
+   0.4443995924983142
+ 701.3843426037456</code></pre><p>we see that they are all positive. Therefore, the Hessian is positive definite, and so the solution found by Ipopt is a local minimizer.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../nested_problems/">« Nested optimization problems</a><a class="docs-footer-nextpage" href="../complementarity/">Mixed complementarity problems »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/rocket_control.jl b/previews/PR3525/tutorials/nonlinear/rocket_control.jl
new file mode 100644
index 00000000000..0a2c5127998
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/rocket_control.jl
@@ -0,0 +1,155 @@
+# Copyright (c) 2019 Iain Dunning, and contributors                             #src
+#                                                                               #src
+# This work is licensed under a Creative Commons Attribution-ShareAlike 4.0     #src
+# International License. URL: http://creativecommons.org/licenses/by-sa/4.0.    #src
+
+# # Rocket Control
+
+# **This tutorial was originally contributed by Iain Dunning.**
+
+# The purpose of this tutorial is to demonstrate how to setup and solve a
+# nonlinear optimization problem.
+
+# The example is an optimal control problem of a nonlinear rocket.
+
+# This tutorial uses the following packages:
+
+using JuMP
+import Ipopt
+import Plots
+
+# ## Overview
+
+# Our goal is to maximize the final altitude of a vertically launched rocket.
+
+# We can control the thrust of the rocket, and must take account of the rocket
+# mass, fuel consumption rate, gravity, and aerodynamic drag.
+
+# Let us consider the basic description of the model (for the full description,
+# including parameters for the rocket, see [COPS3](https://www.mcs.anl.gov/~more/cops/cops3.pdf)).
+
+# There are three state variables in our model:
+
+# * Velocity: $x_v(t)$
+# * Altitude: $x_h(t)$
+# * Mass of rocket and remaining fuel, $x_m(t)$
+
+# and a single control variable:
+
+# * Thrust: $u_t(t)$.
+
+# There are three equations that control the dynamics of the rocket:
+
+#  * Rate of ascent: $$\frac{d x_h}{dt} = x_v$$
+#  * Acceleration: $$\frac{d x_v}{dt} = \frac{u_t - D(x_h, x_v)}{x_m} - g(x_h)$$
+#  * Rate of mass loss: $$\frac{d x_m}{dt} = -\frac{u_t}{c}$$
+
+# where drag $D(x_h, x_v)$ is a function of altitude and velocity, gravity
+# $g(x_h)$ is a function of altitude, and $c$ is a constant.
+
+# These forces are defined as:
+#
+# $$D(x_h, x_v) = D_c \cdot x_v^2 \cdot e^{-h_c \left( \frac{x_h-x_h(0)}{x_h(0)} \right)}$$
+# and
+# $$g(x_h) = g_0 \cdot \left( \frac{x_h(0)}{x_h} \right)^2$$
+
+# We use a discretized model of time, with a fixed number of time steps, $T$.
+
+# Our goal is thus to maximize $x_h(T)$.
+
+# ## Data
+
+# All parameters in this model have been normalized to be dimensionless, and
+# they are taken from [COPS3](https://www.mcs.anl.gov/~more/cops/cops3.pdf).
+
+h_0 = 1                      # Initial height
+v_0 = 0                      # Initial velocity
+m_0 = 1.0                    # Initial mass
+m_T = 0.6                    # Final mass
+g_0 = 1                      # Gravity at the surface
+h_c = 500                    # Used for drag
+c = 0.5 * sqrt(g_0 * h_0)    # Thrust-to-fuel mass
+D_c = 0.5 * 620 * m_0 / g_0  # Drag scaling
+u_t_max = 3.5 * g_0 * m_0    # Maximum thrust
+T_max = 0.2                  # Number of seconds
+T = 1_000                    # Number of time steps
+Δt = 0.2 / T;                # Time per discretized step
+
+# ## JuMP formulation
+
+# First, we create a model and choose an optimizer. Since this is a nonlinear
+# program, we need to use a nonlinear solver like Ipopt. We cannot use a linear
+# solver like HiGHS.
+
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+
+# Next, we create our state and control variables, which are each indexed by
+# `t`. It is good practice for nonlinear programs to always provide a starting
+# solution for each variable.
+
+@variable(model, x_v[1:T] >= 0, start = v_0)           # Velocity
+@variable(model, x_h[1:T] >= 0, start = h_0)           # Height
+@variable(model, x_m[1:T] >= m_T, start = m_0)         # Mass
+@variable(model, 0 <= u_t[1:T] <= u_t_max, start = 0); # Thrust
+
+# We implement boundary conditions by fixing variables to values.
+
+fix(x_v[1], v_0; force = true)
+fix(x_h[1], h_0; force = true)
+fix(x_m[1], m_0; force = true)
+fix(u_t[T], 0.0; force = true)
+
+# The objective is to maximize altitude at end of time of flight.
+
+@objective(model, Max, x_h[T])
+
+# Forces are defined as functions:
+
+D(x_h, x_v) = D_c * x_v^2 * exp(-h_c * (x_h - h_0) / h_0)
+g(x_h) = g_0 * (h_0 / x_h)^2
+
+# The dynamical equations are implemented as constraints.
+
+ddt(x::Vector, t::Int) = (x[t] - x[t-1]) / Δt
+@constraint(model, [t in 2:T], ddt(x_h, t) == x_v[t-1])
+@constraint(
+    model,
+    [t in 2:T],
+    ddt(x_v, t) == (u_t[t-1] - D(x_h[t-1], x_v[t-1])) / x_m[t-1] - g(x_h[t-1]),
+)
+@constraint(model, [t in 2:T], ddt(x_m, t) == -u_t[t-1] / c);
+
+# Now we optimize the model and check that we found a solution:
+
+optimize!(model)
+solution_summary(model)
+
+# Finally, we plot the solution:
+
+function plot_trajectory(y; kwargs...)
+    return Plots.plot(
+        (1:T) * Δt,
+        value.(y);
+        xlabel = "Time (s)",
+        legend = false,
+        kwargs...,
+    )
+end
+
+Plots.plot(
+    plot_trajectory(x_h; ylabel = "Altitude"),
+    plot_trajectory(x_m; ylabel = "Mass"),
+    plot_trajectory(x_v; ylabel = "Velocity"),
+    plot_trajectory(u_t; ylabel = "Thrust");
+    layout = (2, 2),
+)
+
+# ## Next steps
+
+# * Experiment with different values for the constants. How does the solution
+#   change? In particular, what happens if you change `T_max`?
+# * The dynamical equations use rectangular integration for the right-hand side
+#   terms. Modify the equations to use the [Trapezoidal rule](https://en.wikipedia.org/wiki/Trapezoidal_rule_(differential_equations))
+#   instead. (As an example, `x_v[t-1]` would become
+#   `0.5 * (x_v[t-1] + x_v[t])`.) Is there a difference?
diff --git a/previews/PR3525/tutorials/nonlinear/rocket_control/2e6d6b3d.svg b/previews/PR3525/tutorials/nonlinear/rocket_control/2e6d6b3d.svg
new file mode 100644
index 00000000000..e36df148e90
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/rocket_control/2e6d6b3d.svg
@@ -0,0 +1,130 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip570">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip570)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip571">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip570)" d="M277.528 623.18 L1181.75 623.18 L1181.75 47.2441 L277.528 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip572">
+    <rect x="277" y="47" width="905" height="577"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="302.265,623.18 302.265,47.2441 "/>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="515.738,623.18 515.738,47.2441 "/>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="729.211,623.18 729.211,47.2441 "/>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="942.685,623.18 942.685,47.2441 "/>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1156.16,623.18 1156.16,47.2441 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,623.18 1181.75,623.18 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="302.265,623.18 302.265,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="515.738,623.18 515.738,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="729.211,623.18 729.211,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="942.685,623.18 942.685,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1156.16,623.18 1156.16,604.282 "/>
+<path clip-path="url(#clip570)" d="M264.569 654.099 Q260.958 654.099 259.129 657.663 Q257.323 661.205 257.323 668.335 Q257.323 675.441 259.129 679.006 Q260.958 682.547 264.569 682.547 Q268.203 682.547 270.008 679.006 Q271.837 675.441 271.837 668.335 Q271.837 661.205 270.008 657.663 Q268.203 654.099 264.569 654.099 M264.569 650.395 Q270.379 650.395 273.434 655.001 Q276.513 659.585 276.513 668.335 Q276.513 677.061 273.434 681.668 Q270.379 686.251 264.569 686.251 Q258.759 686.251 255.68 681.668 Q252.624 677.061 252.624 668.335 Q252.624 659.585 255.68 655.001 Q258.759 650.395 264.569 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M284.731 679.7 L289.615 679.7 L289.615 685.58 L284.731 685.58 L284.731 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M309.8 654.099 Q306.189 654.099 304.36 657.663 Q302.555 661.205 302.555 668.335 Q302.555 675.441 304.36 679.006 Q306.189 682.547 309.8 682.547 Q313.434 682.547 315.24 679.006 Q317.068 675.441 317.068 668.335 Q317.068 661.205 315.24 657.663 Q313.434 654.099 309.8 654.099 M309.8 650.395 Q315.61 650.395 318.666 655.001 Q321.744 659.585 321.744 668.335 Q321.744 677.061 318.666 681.668 Q315.61 686.251 309.8 686.251 Q303.99 686.251 300.911 681.668 Q297.856 677.061 297.856 668.335 Q297.856 659.585 300.911 655.001 Q303.99 650.395 309.8 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M339.962 654.099 Q336.351 654.099 334.522 657.663 Q332.716 661.205 332.716 668.335 Q332.716 675.441 334.522 679.006 Q336.351 682.547 339.962 682.547 Q343.596 682.547 345.402 679.006 Q347.23 675.441 347.23 668.335 Q347.23 661.205 345.402 657.663 Q343.596 654.099 339.962 654.099 M339.962 650.395 Q345.772 650.395 348.828 655.001 Q351.906 659.585 351.906 668.335 Q351.906 677.061 348.828 681.668 Q345.772 686.251 339.962 686.251 Q334.152 686.251 331.073 681.668 Q328.017 677.061 328.017 668.335 Q328.017 659.585 331.073 655.001 Q334.152 650.395 339.962 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M478.539 654.099 Q474.928 654.099 473.1 657.663 Q471.294 661.205 471.294 668.335 Q471.294 675.441 473.1 679.006 Q474.928 682.547 478.539 682.547 Q482.174 682.547 483.979 679.006 Q485.808 675.441 485.808 668.335 Q485.808 661.205 483.979 657.663 Q482.174 654.099 478.539 654.099 M478.539 650.395 Q484.35 650.395 487.405 655.001 Q490.484 659.585 490.484 668.335 Q490.484 677.061 487.405 681.668 Q484.35 686.251 478.539 686.251 Q472.729 686.251 469.651 681.668 Q466.595 677.061 466.595 668.335 Q466.595 659.585 469.651 655.001 Q472.729 650.395 478.539 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M498.701 679.7 L503.586 679.7 L503.586 685.58 L498.701 685.58 L498.701 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M523.771 654.099 Q520.16 654.099 518.331 657.663 Q516.525 661.205 516.525 668.335 Q516.525 675.441 518.331 679.006 Q520.16 682.547 523.771 682.547 Q527.405 682.547 529.211 679.006 Q531.039 675.441 531.039 668.335 Q531.039 661.205 529.211 657.663 Q527.405 654.099 523.771 654.099 M523.771 650.395 Q529.581 650.395 532.636 655.001 Q535.715 659.585 535.715 668.335 Q535.715 677.061 532.636 681.668 Q529.581 686.251 523.771 686.251 Q517.961 686.251 514.882 681.668 Q511.826 677.061 511.826 668.335 Q511.826 659.585 514.882 655.001 Q517.961 650.395 523.771 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M543.979 651.02 L562.335 651.02 L562.335 654.955 L548.261 654.955 L548.261 663.427 Q549.28 663.08 550.298 662.918 Q551.317 662.733 552.335 662.733 Q558.122 662.733 561.502 665.904 Q564.882 669.075 564.882 674.492 Q564.882 680.071 561.409 683.172 Q557.937 686.251 551.618 686.251 Q549.442 686.251 547.173 685.881 Q544.928 685.51 542.521 684.77 L542.521 680.071 Q544.604 681.205 546.826 681.76 Q549.048 682.316 551.525 682.316 Q555.53 682.316 557.868 680.21 Q560.206 678.103 560.206 674.492 Q560.206 670.881 557.868 668.774 Q555.53 666.668 551.525 666.668 Q549.65 666.668 547.775 667.085 Q545.923 667.501 543.979 668.381 L543.979 651.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M691.515 654.099 Q687.904 654.099 686.075 657.663 Q684.27 661.205 684.27 668.335 Q684.27 675.441 686.075 679.006 Q687.904 682.547 691.515 682.547 Q695.149 682.547 696.955 679.006 Q698.783 675.441 698.783 668.335 Q698.783 661.205 696.955 657.663 Q695.149 654.099 691.515 654.099 M691.515 650.395 Q697.325 650.395 700.381 655.001 Q703.459 659.585 703.459 668.335 Q703.459 677.061 700.381 681.668 Q697.325 686.251 691.515 686.251 Q685.705 686.251 682.626 681.668 Q679.571 677.061 679.571 668.335 Q679.571 659.585 682.626 655.001 Q685.705 650.395 691.515 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M711.677 679.7 L716.561 679.7 L716.561 685.58 L711.677 685.58 L711.677 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M727.556 681.645 L735.195 681.645 L735.195 655.279 L726.885 656.946 L726.885 652.687 L735.149 651.02 L739.825 651.02 L739.825 681.645 L747.464 681.645 L747.464 685.58 L727.556 685.58 L727.556 681.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M766.908 654.099 Q763.297 654.099 761.468 657.663 Q759.663 661.205 759.663 668.335 Q759.663 675.441 761.468 679.006 Q763.297 682.547 766.908 682.547 Q770.542 682.547 772.348 679.006 Q774.176 675.441 774.176 668.335 Q774.176 661.205 772.348 657.663 Q770.542 654.099 766.908 654.099 M766.908 650.395 Q772.718 650.395 775.774 655.001 Q778.852 659.585 778.852 668.335 Q778.852 677.061 775.774 681.668 Q772.718 686.251 766.908 686.251 Q761.098 686.251 758.019 681.668 Q754.964 677.061 754.964 668.335 Q754.964 659.585 758.019 655.001 Q761.098 650.395 766.908 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M905.486 654.099 Q901.875 654.099 900.046 657.663 Q898.24 661.205 898.24 668.335 Q898.24 675.441 900.046 679.006 Q901.875 682.547 905.486 682.547 Q909.12 682.547 910.925 679.006 Q912.754 675.441 912.754 668.335 Q912.754 661.205 910.925 657.663 Q909.12 654.099 905.486 654.099 M905.486 650.395 Q911.296 650.395 914.351 655.001 Q917.43 659.585 917.43 668.335 Q917.43 677.061 914.351 681.668 Q911.296 686.251 905.486 686.251 Q899.676 686.251 896.597 681.668 Q893.541 677.061 893.541 668.335 Q893.541 659.585 896.597 655.001 Q899.676 650.395 905.486 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M925.648 679.7 L930.532 679.7 L930.532 685.58 L925.648 685.58 L925.648 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M941.527 681.645 L949.166 681.645 L949.166 655.279 L940.856 656.946 L940.856 652.687 L949.12 651.02 L953.796 651.02 L953.796 681.645 L961.434 681.645 L961.434 685.58 L941.527 685.58 L941.527 681.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M970.925 651.02 L989.282 651.02 L989.282 654.955 L975.208 654.955 L975.208 663.427 Q976.226 663.08 977.245 662.918 Q978.263 662.733 979.282 662.733 Q985.069 662.733 988.448 665.904 Q991.828 669.075 991.828 674.492 Q991.828 680.071 988.356 683.172 Q984.883 686.251 978.564 686.251 Q976.388 686.251 974.12 685.881 Q971.874 685.51 969.467 684.77 L969.467 680.071 Q971.55 681.205 973.772 681.76 Q975.995 682.316 978.471 682.316 Q982.476 682.316 984.814 680.21 Q987.152 678.103 987.152 674.492 Q987.152 670.881 984.814 668.774 Q982.476 666.668 978.471 666.668 Q976.596 666.668 974.721 667.085 Q972.87 667.501 970.925 668.381 L970.925 651.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1118.46 654.099 Q1114.85 654.099 1113.02 657.663 Q1111.22 661.205 1111.22 668.335 Q1111.22 675.441 1113.02 679.006 Q1114.85 682.547 1118.46 682.547 Q1122.1 682.547 1123.9 679.006 Q1125.73 675.441 1125.73 668.335 Q1125.73 661.205 1123.9 657.663 Q1122.1 654.099 1118.46 654.099 M1118.46 650.395 Q1124.27 650.395 1127.33 655.001 Q1130.41 659.585 1130.41 668.335 Q1130.41 677.061 1127.33 681.668 Q1124.27 686.251 1118.46 686.251 Q1112.65 686.251 1109.57 681.668 Q1106.52 677.061 1106.52 668.335 Q1106.52 659.585 1109.57 655.001 Q1112.65 650.395 1118.46 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1138.62 679.7 L1143.51 679.7 L1143.51 685.58 L1138.62 685.58 L1138.62 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1157.72 681.645 L1174.04 681.645 L1174.04 685.58 L1152.1 685.58 L1152.1 681.645 Q1154.76 678.89 1159.34 674.26 Q1163.95 669.608 1165.13 668.265 Q1167.37 665.742 1168.25 664.006 Q1169.16 662.247 1169.16 660.557 Q1169.16 657.802 1167.21 656.066 Q1165.29 654.33 1162.19 654.33 Q1159.99 654.33 1157.53 655.094 Q1155.1 655.858 1152.33 657.409 L1152.33 652.687 Q1155.15 651.552 1157.6 650.974 Q1160.06 650.395 1162.1 650.395 Q1167.47 650.395 1170.66 653.08 Q1173.85 655.765 1173.85 660.256 Q1173.85 662.386 1173.04 664.307 Q1172.26 666.205 1170.15 668.798 Q1169.57 669.469 1166.47 672.686 Q1163.37 675.881 1157.72 681.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1193.85 654.099 Q1190.24 654.099 1188.41 657.663 Q1186.61 661.205 1186.61 668.335 Q1186.61 675.441 1188.41 679.006 Q1190.24 682.547 1193.85 682.547 Q1197.49 682.547 1199.29 679.006 Q1201.12 675.441 1201.12 668.335 Q1201.12 661.205 1199.29 657.663 Q1197.49 654.099 1193.85 654.099 M1193.85 650.395 Q1199.66 650.395 1202.72 655.001 Q1205.8 659.585 1205.8 668.335 Q1205.8 677.061 1202.72 681.668 Q1199.66 686.251 1193.85 686.251 Q1188.04 686.251 1184.97 681.668 Q1181.91 677.061 1181.91 668.335 Q1181.91 659.585 1184.97 655.001 Q1188.04 650.395 1193.85 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M599.81 720.524 L640.009 720.524 L640.009 725.935 L623.14 725.935 L623.14 768.044 L616.679 768.044 L616.679 725.935 L599.81 725.935 L599.81 720.524 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M643.956 732.396 L649.812 732.396 L649.812 768.044 L643.956 768.044 L643.956 732.396 M643.956 718.519 L649.812 718.519 L649.812 725.935 L643.956 725.935 L643.956 718.519 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M689.821 739.239 Q692.017 735.292 695.073 733.414 Q698.128 731.537 702.266 731.537 Q707.836 731.537 710.86 735.451 Q713.883 739.335 713.883 746.528 L713.883 768.044 L707.995 768.044 L707.995 746.719 Q707.995 741.594 706.181 739.112 Q704.367 736.629 700.643 736.629 Q696.091 736.629 693.449 739.653 Q690.808 742.677 690.808 747.896 L690.808 768.044 L684.919 768.044 L684.919 746.719 Q684.919 741.563 683.105 739.112 Q681.291 736.629 677.503 736.629 Q673.015 736.629 670.374 739.685 Q667.732 742.708 667.732 747.896 L667.732 768.044 L661.844 768.044 L661.844 732.396 L667.732 732.396 L667.732 737.934 Q669.737 734.656 672.538 733.096 Q675.339 731.537 679.19 731.537 Q683.073 731.537 685.779 733.51 Q688.516 735.483 689.821 739.239 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M756.056 748.756 L756.056 751.62 L729.129 751.62 Q729.511 757.668 732.758 760.851 Q736.036 764.002 741.861 764.002 Q745.234 764.002 748.385 763.174 Q751.568 762.347 754.687 760.691 L754.687 766.23 Q751.536 767.566 748.226 768.267 Q744.916 768.967 741.51 768.967 Q732.98 768.967 727.983 764.002 Q723.018 759.036 723.018 750.57 Q723.018 741.817 727.729 736.693 Q732.471 731.537 740.492 731.537 Q747.685 731.537 751.855 736.183 Q756.056 740.799 756.056 748.756 M750.2 747.037 Q750.136 742.231 747.494 739.366 Q744.884 736.502 740.556 736.502 Q735.654 736.502 732.694 739.271 Q729.766 742.04 729.32 747.069 L750.2 747.037 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M800.457 718.582 Q796.192 725.903 794.123 733.064 Q792.054 740.226 792.054 747.578 Q792.054 754.93 794.123 762.156 Q796.224 769.349 800.457 776.638 L795.364 776.638 Q790.59 769.158 788.203 761.933 Q785.848 754.708 785.848 747.578 Q785.848 740.48 788.203 733.287 Q790.558 726.094 795.364 718.582 L800.457 718.582 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M834.545 733.446 L834.545 738.984 Q832.063 737.711 829.389 737.075 Q826.715 736.438 823.851 736.438 Q819.49 736.438 817.294 737.775 Q815.13 739.112 815.13 741.785 Q815.13 743.822 816.689 745 Q818.249 746.146 822.96 747.196 L824.965 747.642 Q831.203 748.979 833.813 751.429 Q836.455 753.848 836.455 758.209 Q836.455 763.174 832.508 766.07 Q828.593 768.967 821.718 768.967 Q818.854 768.967 815.735 768.394 Q812.647 767.853 809.21 766.739 L809.21 760.691 Q812.456 762.378 815.607 763.238 Q818.758 764.065 821.846 764.065 Q825.983 764.065 828.211 762.665 Q830.439 761.233 830.439 758.654 Q830.439 756.267 828.816 754.994 Q827.225 753.721 821.782 752.543 L819.745 752.066 Q814.302 750.92 811.883 748.565 Q809.464 746.178 809.464 742.04 Q809.464 737.011 813.029 734.274 Q816.594 731.537 823.151 731.537 Q826.397 731.537 829.262 732.014 Q832.126 732.491 834.545 733.446 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M844.858 718.582 L849.95 718.582 Q854.725 726.094 857.08 733.287 Q859.467 740.48 859.467 747.578 Q859.467 754.708 857.08 761.933 Q854.725 769.158 849.95 776.638 L844.858 776.638 Q849.091 769.349 851.16 762.156 Q853.26 754.93 853.26 747.578 Q853.26 740.226 851.16 733.064 Q849.091 725.903 844.858 718.582 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,521.847 1181.75,521.847 "/>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,394.297 1181.75,394.297 "/>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,266.747 1181.75,266.747 "/>
+<polyline clip-path="url(#clip572)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,139.198 1181.75,139.198 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,623.18 277.528,47.2441 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,521.847 296.426,521.847 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,394.297 296.426,394.297 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,266.747 296.426,266.747 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,139.198 296.426,139.198 "/>
+<path clip-path="url(#clip570)" d="M116.436 535.192 L124.075 535.192 L124.075 508.826 L115.765 510.493 L115.765 506.233 L124.029 504.567 L128.705 504.567 L128.705 535.192 L136.343 535.192 L136.343 539.127 L116.436 539.127 L116.436 535.192 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M145.788 533.247 L150.672 533.247 L150.672 539.127 L145.788 539.127 L145.788 533.247 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M170.857 507.645 Q167.246 507.645 165.417 511.21 Q163.612 514.752 163.612 521.881 Q163.612 528.988 165.417 532.553 Q167.246 536.094 170.857 536.094 Q174.491 536.094 176.297 532.553 Q178.126 528.988 178.126 521.881 Q178.126 514.752 176.297 511.21 Q174.491 507.645 170.857 507.645 M170.857 503.942 Q176.667 503.942 179.723 508.548 Q182.801 513.131 182.801 521.881 Q182.801 530.608 179.723 535.215 Q176.667 539.798 170.857 539.798 Q165.047 539.798 161.968 535.215 Q158.913 530.608 158.913 521.881 Q158.913 513.131 161.968 508.548 Q165.047 503.942 170.857 503.942 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M201.019 507.645 Q197.408 507.645 195.579 511.21 Q193.774 514.752 193.774 521.881 Q193.774 528.988 195.579 532.553 Q197.408 536.094 201.019 536.094 Q204.653 536.094 206.459 532.553 Q208.287 528.988 208.287 521.881 Q208.287 514.752 206.459 511.21 Q204.653 507.645 201.019 507.645 M201.019 503.942 Q206.829 503.942 209.885 508.548 Q212.963 513.131 212.963 521.881 Q212.963 530.608 209.885 535.215 Q206.829 539.798 201.019 539.798 Q195.209 539.798 192.13 535.215 Q189.075 530.608 189.075 521.881 Q189.075 513.131 192.13 508.548 Q195.209 503.942 201.019 503.942 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M225.209 535.192 L241.528 535.192 L241.528 539.127 L219.584 539.127 L219.584 535.192 Q222.246 532.437 226.829 527.807 Q231.435 523.155 232.616 521.812 Q234.861 519.289 235.741 517.553 Q236.644 515.793 236.644 514.104 Q236.644 511.349 234.699 509.613 Q232.778 507.877 229.676 507.877 Q227.477 507.877 225.023 508.641 Q222.593 509.405 219.815 510.956 L219.815 506.233 Q222.639 505.099 225.093 504.52 Q227.547 503.942 229.584 503.942 Q234.954 503.942 238.148 506.627 Q241.343 509.312 241.343 513.803 Q241.343 515.932 240.533 517.854 Q239.746 519.752 237.639 522.344 Q237.06 523.016 233.959 526.233 Q230.857 529.428 225.209 535.192 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M115.834 407.642 L123.473 407.642 L123.473 381.276 L115.163 382.943 L115.163 378.684 L123.427 377.017 L128.103 377.017 L128.103 407.642 L135.742 407.642 L135.742 411.577 L115.834 411.577 L115.834 407.642 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M145.186 405.697 L150.07 405.697 L150.07 411.577 L145.186 411.577 L145.186 405.697 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M170.255 380.096 Q166.644 380.096 164.815 383.661 Q163.01 387.202 163.01 394.332 Q163.01 401.438 164.815 405.003 Q166.644 408.545 170.255 408.545 Q173.889 408.545 175.695 405.003 Q177.524 401.438 177.524 394.332 Q177.524 387.202 175.695 383.661 Q173.889 380.096 170.255 380.096 M170.255 376.392 Q176.065 376.392 179.121 380.999 Q182.2 385.582 182.2 394.332 Q182.2 403.059 179.121 407.665 Q176.065 412.248 170.255 412.248 Q164.445 412.248 161.366 407.665 Q158.311 403.059 158.311 394.332 Q158.311 385.582 161.366 380.999 Q164.445 376.392 170.255 376.392 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M200.417 380.096 Q196.806 380.096 194.977 383.661 Q193.172 387.202 193.172 394.332 Q193.172 401.438 194.977 405.003 Q196.806 408.545 200.417 408.545 Q204.051 408.545 205.857 405.003 Q207.686 401.438 207.686 394.332 Q207.686 387.202 205.857 383.661 Q204.051 380.096 200.417 380.096 M200.417 376.392 Q206.227 376.392 209.283 380.999 Q212.361 385.582 212.361 394.332 Q212.361 403.059 209.283 407.665 Q206.227 412.248 200.417 412.248 Q194.607 412.248 191.528 407.665 Q188.473 403.059 188.473 394.332 Q188.473 385.582 191.528 380.999 Q194.607 376.392 200.417 376.392 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M220.625 377.017 L238.982 377.017 L238.982 380.952 L224.908 380.952 L224.908 389.424 Q225.926 389.077 226.945 388.915 Q227.963 388.73 228.982 388.73 Q234.769 388.73 238.148 391.901 Q241.528 395.073 241.528 400.489 Q241.528 406.068 238.056 409.17 Q234.584 412.248 228.264 412.248 Q226.088 412.248 223.82 411.878 Q221.574 411.508 219.167 410.767 L219.167 406.068 Q221.25 407.202 223.473 407.758 Q225.695 408.313 228.172 408.313 Q232.176 408.313 234.514 406.207 Q236.852 404.1 236.852 400.489 Q236.852 396.878 234.514 394.772 Q232.176 392.665 228.172 392.665 Q226.297 392.665 224.422 393.082 Q222.57 393.498 220.625 394.378 L220.625 377.017 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M114.931 280.092 L122.57 280.092 L122.57 253.727 L114.26 255.393 L114.26 251.134 L122.524 249.467 L127.2 249.467 L127.2 280.092 L134.839 280.092 L134.839 284.027 L114.931 284.027 L114.931 280.092 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M144.283 278.148 L149.167 278.148 L149.167 284.027 L144.283 284.027 L144.283 278.148 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M169.352 252.546 Q165.741 252.546 163.913 256.111 Q162.107 259.653 162.107 266.782 Q162.107 273.889 163.913 277.453 Q165.741 280.995 169.352 280.995 Q172.987 280.995 174.792 277.453 Q176.621 273.889 176.621 266.782 Q176.621 259.653 174.792 256.111 Q172.987 252.546 169.352 252.546 M169.352 248.842 Q175.163 248.842 178.218 253.449 Q181.297 258.032 181.297 266.782 Q181.297 275.509 178.218 280.115 Q175.163 284.699 169.352 284.699 Q163.542 284.699 160.464 280.115 Q157.408 275.509 157.408 266.782 Q157.408 258.032 160.464 253.449 Q163.542 248.842 169.352 248.842 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M199.514 252.546 Q195.903 252.546 194.075 256.111 Q192.269 259.653 192.269 266.782 Q192.269 273.889 194.075 277.453 Q195.903 280.995 199.514 280.995 Q203.149 280.995 204.954 277.453 Q206.783 273.889 206.783 266.782 Q206.783 259.653 204.954 256.111 Q203.149 252.546 199.514 252.546 M199.514 248.842 Q205.324 248.842 208.38 253.449 Q211.459 258.032 211.459 266.782 Q211.459 275.509 208.38 280.115 Q205.324 284.699 199.514 284.699 Q193.704 284.699 190.625 280.115 Q187.57 275.509 187.57 266.782 Q187.57 258.032 190.625 253.449 Q193.704 248.842 199.514 248.842 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M229.676 267.615 Q226.343 267.615 224.422 269.398 Q222.523 271.18 222.523 274.305 Q222.523 277.43 224.422 279.213 Q226.343 280.995 229.676 280.995 Q233.01 280.995 234.931 279.213 Q236.852 277.407 236.852 274.305 Q236.852 271.18 234.931 269.398 Q233.033 267.615 229.676 267.615 M225 265.625 Q221.991 264.884 220.301 262.824 Q218.635 260.764 218.635 257.801 Q218.635 253.657 221.574 251.25 Q224.537 248.842 229.676 248.842 Q234.838 248.842 237.778 251.25 Q240.718 253.657 240.718 257.801 Q240.718 260.764 239.028 262.824 Q237.361 264.884 234.375 265.625 Q237.755 266.412 239.63 268.703 Q241.528 270.995 241.528 274.305 Q241.528 279.328 238.449 282.014 Q235.394 284.699 229.676 284.699 Q223.959 284.699 220.88 282.014 Q217.824 279.328 217.824 274.305 Q217.824 270.995 219.723 268.703 Q221.621 266.412 225 265.625 M223.287 258.241 Q223.287 260.926 224.954 262.43 Q226.644 263.935 229.676 263.935 Q232.685 263.935 234.375 262.43 Q236.088 260.926 236.088 258.241 Q236.088 255.555 234.375 254.051 Q232.685 252.546 229.676 252.546 Q226.644 252.546 224.954 254.051 Q223.287 255.555 223.287 258.241 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M116.066 152.543 L123.705 152.543 L123.705 126.177 L115.394 127.844 L115.394 123.584 L123.658 121.918 L128.334 121.918 L128.334 152.543 L135.973 152.543 L135.973 156.478 L116.066 156.478 L116.066 152.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M145.417 150.598 L150.302 150.598 L150.302 156.478 L145.417 156.478 L145.417 150.598 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M170.487 124.996 Q166.876 124.996 165.047 128.561 Q163.241 132.103 163.241 139.232 Q163.241 146.339 165.047 149.904 Q166.876 153.445 170.487 153.445 Q174.121 153.445 175.926 149.904 Q177.755 146.339 177.755 139.232 Q177.755 132.103 175.926 128.561 Q174.121 124.996 170.487 124.996 M170.487 121.293 Q176.297 121.293 179.352 125.899 Q182.431 130.483 182.431 139.232 Q182.431 147.959 179.352 152.566 Q176.297 157.149 170.487 157.149 Q164.677 157.149 161.598 152.566 Q158.542 147.959 158.542 139.232 Q158.542 130.483 161.598 125.899 Q164.677 121.293 170.487 121.293 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M191.459 152.543 L199.098 152.543 L199.098 126.177 L190.788 127.844 L190.788 123.584 L199.051 121.918 L203.727 121.918 L203.727 152.543 L211.366 152.543 L211.366 156.478 L191.459 156.478 L191.459 152.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M221.621 152.543 L229.26 152.543 L229.26 126.177 L220.949 127.844 L220.949 123.584 L229.213 121.918 L233.889 121.918 L233.889 152.543 L241.528 152.543 L241.528 156.478 L221.621 156.478 L221.621 152.543 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M22.818 438.814 L46.4666 447.535 L46.4666 430.061 L22.818 438.814 M16.4842 442.442 L16.4842 435.154 L64.0042 417.043 L64.0042 423.727 L51.8138 428.056 L51.8138 449.476 L64.0042 453.805 L64.0042 460.585 L16.4842 442.442 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M14.479 410.359 L14.479 404.503 L64.0042 404.503 L64.0042 410.359 L14.479 410.359 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M18.2347 386.456 L28.3562 386.456 L28.3562 374.393 L32.9077 374.393 L32.9077 386.456 L52.2594 386.456 Q56.6199 386.456 57.8613 385.278 Q59.1026 384.069 59.1026 380.408 L59.1026 374.393 L64.0042 374.393 L64.0042 380.408 Q64.0042 387.188 61.4897 389.766 Q58.9434 392.344 52.2594 392.344 L32.9077 392.344 L32.9077 396.641 L28.3562 396.641 L28.3562 392.344 L18.2347 392.344 L18.2347 386.456 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M28.3562 366.69 L28.3562 360.834 L64.0042 360.834 L64.0042 366.69 L28.3562 366.69 M14.479 366.69 L14.479 360.834 L21.895 360.834 L21.895 366.69 L14.479 366.69 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M18.2347 342.787 L28.3562 342.787 L28.3562 330.724 L32.9077 330.724 L32.9077 342.787 L52.2594 342.787 Q56.6199 342.787 57.8613 341.61 Q59.1026 340.4 59.1026 336.74 L59.1026 330.724 L64.0042 330.724 L64.0042 336.74 Q64.0042 343.519 61.4897 346.097 Q58.9434 348.675 52.2594 348.675 L32.9077 348.675 L32.9077 352.972 L28.3562 352.972 L28.3562 348.675 L18.2347 348.675 L18.2347 342.787 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M49.9359 323.626 L28.3562 323.626 L28.3562 317.77 L49.7131 317.77 Q54.7739 317.77 57.3202 315.797 Q59.8346 313.823 59.8346 309.876 Q59.8346 305.134 56.8109 302.397 Q53.7872 299.628 48.5673 299.628 L28.3562 299.628 L28.3562 293.771 L64.0042 293.771 L64.0042 299.628 L58.5296 299.628 Q61.7762 301.76 63.3676 304.593 Q64.9272 307.394 64.9272 311.118 Q64.9272 317.261 61.1078 320.444 Q57.2883 323.626 49.9359 323.626 M27.4968 308.89 L27.4968 308.89 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M33.7671 258.251 L14.479 258.251 L14.479 252.394 L64.0042 252.394 L64.0042 258.251 L58.657 258.251 Q61.8398 260.097 63.3994 262.929 Q64.9272 265.73 64.9272 269.677 Q64.9272 276.138 59.771 280.212 Q54.6147 284.254 46.212 284.254 Q37.8093 284.254 32.6531 280.212 Q27.4968 276.138 27.4968 269.677 Q27.4968 265.73 29.0564 262.929 Q30.5842 260.097 33.7671 258.251 M46.212 278.207 Q52.6732 278.207 56.3653 275.565 Q60.0256 272.892 60.0256 268.245 Q60.0256 263.598 56.3653 260.924 Q52.6732 258.251 46.212 258.251 Q39.7508 258.251 36.0905 260.924 Q32.3984 263.598 32.3984 268.245 Q32.3984 272.892 36.0905 275.565 Q39.7508 278.207 46.212 278.207 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M44.7161 209.839 L47.5806 209.839 L47.5806 236.766 Q53.6281 236.384 56.8109 233.138 Q59.9619 229.86 59.9619 224.035 Q59.9619 220.661 59.1344 217.51 Q58.3069 214.327 56.6518 211.208 L62.1899 211.208 Q63.5267 214.359 64.227 217.669 Q64.9272 220.979 64.9272 224.385 Q64.9272 232.915 59.9619 237.912 Q54.9967 242.877 46.5303 242.877 Q37.7774 242.877 32.6531 238.167 Q27.4968 233.424 27.4968 225.404 Q27.4968 218.21 32.1438 214.041 Q36.7589 209.839 44.7161 209.839 M42.9973 215.696 Q38.1912 215.759 35.3266 218.401 Q32.4621 221.011 32.4621 225.34 Q32.4621 230.241 35.2312 233.202 Q38.0002 236.13 43.0292 236.575 L42.9973 215.696 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip572)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="303.119,606.88 303.973,606.88 304.827,606.876 305.681,606.867 306.535,606.854 307.389,606.837 308.243,606.816 309.096,606.79 309.95,606.76 310.804,606.726 311.658,606.688 312.512,606.645 313.366,606.598 314.22,606.546 315.074,606.49 315.928,606.43 316.781,606.365 317.635,606.297 318.489,606.223 319.343,606.146 320.197,606.064 321.051,605.978 321.905,605.887 322.759,605.792 323.613,605.692 324.466,605.589 325.32,605.481 326.174,605.368 327.028,605.251 327.882,605.13 328.736,605.004 329.59,604.874 330.444,604.74 331.298,604.601 332.151,604.457 333.005,604.31 333.859,604.158 334.713,604.001 335.567,603.841 336.421,603.676 337.275,603.506 338.129,603.332 338.983,603.154 339.837,602.971 340.69,602.784 341.544,602.593 342.398,602.397 343.252,602.197 344.106,601.992 344.96,601.783 345.814,601.57 346.668,601.352 347.522,601.13 348.375,600.904 349.229,600.673 350.083,600.438 350.937,600.199 351.791,599.956 352.645,599.708 353.499,599.455 354.353,599.199 355.207,598.938 356.06,598.673 356.914,598.403 357.768,598.129 358.622,597.851 359.476,597.569 360.33,597.282 361.184,596.992 362.038,596.696 362.892,596.397 363.746,596.093 364.599,595.786 365.453,595.473 366.307,595.157 367.161,594.837 368.015,594.512 368.869,594.183 369.723,593.85 370.577,593.512 371.431,593.171 372.284,592.825 373.138,592.476 373.992,592.122 374.846,591.763 375.7,591.401 376.554,591.035 377.408,590.664 378.262,590.29 379.116,589.911 379.969,589.528 380.823,589.141 381.677,588.75 382.531,588.355 383.385,587.956 384.239,587.553 385.093,587.146 385.947,586.734 386.801,586.319 387.654,585.9 388.508,585.477 389.362,585.049 390.216,584.618 391.07,584.183 391.924,583.744 392.778,583.301 393.632,582.854 394.486,582.403 395.34,581.948 396.193,581.49 397.047,581.028 397.901,580.562 398.755,580.093 399.609,579.622 400.463,579.148 401.317,578.671 402.171,578.193 403.025,577.713 403.878,577.231 404.732,576.747 405.586,576.262 406.44,575.776 407.294,575.288 408.148,574.799 409.002,574.309 409.856,573.818 410.71,573.325 411.563,572.832 412.417,572.337 413.271,571.84 414.125,571.343 414.979,570.845 415.833,570.345 416.687,569.844 417.541,569.342 418.395,568.839 419.249,568.335 420.102,567.83 420.956,567.323 421.81,566.815 422.664,566.306 423.518,565.796 424.372,565.285 425.226,564.772 426.08,564.258 426.934,563.743 427.787,563.227 428.641,562.71 429.495,562.191 430.349,561.671 431.203,561.15 432.057,560.627 432.911,560.104 433.765,559.579 434.619,559.053 435.472,558.525 436.326,557.996 437.18,557.466 438.034,556.935 438.888,556.402 439.742,555.869 440.596,555.333 441.45,554.797 442.304,554.259 443.158,553.72 444.011,553.179 444.865,552.637 445.719,552.094 446.573,551.549 447.427,551.003 448.281,550.456 449.135,549.907 449.989,549.357 450.843,548.805 451.696,548.253 452.55,547.698 453.404,547.142 454.258,546.585 455.112,546.027 455.966,545.467 456.82,544.905 457.674,544.342 458.528,543.778 459.381,543.212 460.235,542.644 461.089,542.076 461.943,541.505 462.797,540.933 463.651,540.36 464.505,539.785 465.359,539.209 466.213,538.631 467.066,538.051 467.92,537.47 468.774,536.888 469.628,536.303 470.482,535.718 471.336,535.13 472.19,534.541 473.044,533.951 473.898,533.359 474.752,532.765 475.605,532.17 476.459,531.573 477.313,530.974 478.167,530.374 479.021,529.772 479.875,529.168 480.729,528.563 481.583,527.956 482.437,527.347 483.29,526.736 484.144,526.124 484.998,525.51 485.852,524.895 486.706,524.277 487.56,523.658 488.414,523.037 489.268,522.414 490.122,521.79 490.975,521.164 491.829,520.535 492.683,519.905 493.537,519.274 494.391,518.64 495.245,518.005 496.099,517.367 496.953,516.728 497.807,516.087 498.661,515.444 499.514,514.799 500.368,514.152 501.222,513.503 502.076,512.852 502.93,512.2 503.784,511.545 504.638,510.888 505.492,510.229 506.346,509.569 507.199,508.906 508.053,508.241 508.907,507.575 509.761,506.906 510.615,506.235 511.469,505.562 512.323,504.887 513.177,504.21 514.031,503.53 514.884,502.849 515.738,502.165 516.592,501.479 517.446,500.791 518.3,500.101 519.154,499.409 520.008,498.714 520.862,498.018 521.716,497.319 522.569,496.617 523.423,495.914 524.277,495.208 525.131,494.5 525.985,493.789 526.839,493.077 527.693,492.361 528.547,491.644 529.401,490.924 530.255,490.202 531.108,489.477 531.962,488.75 532.816,488.02 533.67,487.288 534.524,486.554 535.378,485.817 536.232,485.077 537.086,484.335 537.94,483.591 538.793,482.843 539.647,482.094 540.501,481.341 541.355,480.586 542.209,479.829 543.063,479.069 543.917,478.306 544.771,477.54 545.625,476.772 546.478,476.001 547.332,475.227 548.186,474.451 549.04,473.671 549.894,472.889 550.748,472.104 551.602,471.317 552.456,470.526 553.31,469.732 554.164,468.936 555.017,468.137 555.871,467.334 556.725,466.529 557.579,465.721 558.433,464.91 559.287,464.095 560.141,463.278 560.995,462.458 561.849,461.634 562.702,460.808 563.556,459.978 564.41,459.145 565.264,458.309 566.118,457.469 566.972,456.627 567.826,455.781 568.68,454.932 569.534,454.079 570.387,453.223 571.241,452.364 572.095,451.502 572.949,450.636 573.803,449.766 574.657,448.893 575.511,448.017 576.365,447.137 577.219,446.254 578.072,445.367 578.926,444.476 579.78,443.582 580.634,442.684 581.488,441.782 582.342,440.877 583.196,439.968 584.05,439.055 584.904,438.138 585.758,437.218 586.611,436.293 587.465,435.365 588.319,434.433 589.173,433.497 590.027,432.557 590.881,431.613 591.735,430.665 592.589,429.713 593.443,428.757 594.296,427.796 595.15,426.832 596.004,425.863 596.858,424.89 597.712,423.913 598.566,422.932 599.42,421.947 600.274,420.957 601.128,419.963 601.981,418.965 602.835,417.962 603.689,416.955 604.543,415.944 605.397,414.928 606.251,413.909 607.105,412.884 607.959,411.856 608.813,410.823 609.667,409.786 610.52,408.745 611.374,407.699 612.228,406.649 613.082,405.596 613.936,404.538 614.79,403.476 615.644,402.41 616.498,401.341 617.352,400.268 618.205,399.192 619.059,398.112 619.913,397.029 620.767,395.944 621.621,394.856 622.475,393.765 623.329,392.673 624.183,391.578 625.037,390.483 625.89,389.387 626.744,388.29 627.598,387.194 628.452,386.098 629.306,385.004 630.16,383.911 631.014,382.821 631.868,381.733 632.722,380.647 633.576,379.563 634.429,378.482 635.283,377.404 636.137,376.328 636.991,375.255 637.845,374.184 638.699,373.116 639.553,372.05 640.407,370.987 641.261,369.926 642.114,368.868 642.968,367.812 643.822,366.759 644.676,365.708 645.53,364.66 646.384,363.614 647.238,362.57 648.092,361.529 648.946,360.49 649.799,359.454 650.653,358.42 651.507,357.388 652.361,356.359 653.215,355.332 654.069,354.308 654.923,353.285 655.777,352.265 656.631,351.248 657.484,350.233 658.338,349.22 659.192,348.209 660.046,347.2 660.9,346.194 661.754,345.19 662.608,344.188 663.462,343.189 664.316,342.192 665.17,341.196 666.023,340.204 666.877,339.213 667.731,338.224 668.585,337.238 669.439,336.254 670.293,335.272 671.147,334.292 672.001,333.315 672.855,332.339 673.708,331.366 674.562,330.394 675.416,329.425 676.27,328.458 677.124,327.493 677.978,326.53 678.832,325.57 679.686,324.611 680.54,323.655 681.393,322.7 682.247,321.748 683.101,320.797 683.955,319.849 684.809,318.903 685.663,317.959 686.517,317.016 687.371,316.076 688.225,315.138 689.079,314.202 689.932,313.268 690.786,312.336 691.64,311.406 692.494,310.478 693.348,309.552 694.202,308.628 695.056,307.706 695.91,306.786 696.764,305.868 697.617,304.952 698.471,304.038 699.325,303.125 700.179,302.215 701.033,301.307 701.887,300.401 702.741,299.496 703.595,298.594 704.449,297.693 705.302,296.795 706.156,295.898 707.01,295.003 707.864,294.111 708.718,293.22 709.572,292.331 710.426,291.444 711.28,290.559 712.134,289.676 712.987,288.794 713.841,287.915 714.695,287.037 715.549,286.162 716.403,285.288 717.257,284.416 718.111,283.546 718.965,282.678 719.819,281.812 720.673,280.948 721.526,280.085 722.38,279.225 723.234,278.366 724.088,277.509 724.942,276.654 725.796,275.801 726.65,274.95 727.504,274.1 728.358,273.253 729.211,272.407 730.065,271.563 730.919,270.721 731.773,269.881 732.627,269.042 733.481,268.206 734.335,267.371 735.189,266.538 736.043,265.707 736.896,264.878 737.75,264.051 738.604,263.225 739.458,262.401 740.312,261.579 741.166,260.759 742.02,259.941 742.874,259.124 743.728,258.31 744.582,257.497 745.435,256.686 746.289,255.876 747.143,255.069 747.997,254.263 748.851,253.459 749.705,252.657 750.559,251.857 751.413,251.058 752.267,250.262 753.12,249.467 753.974,248.674 754.828,247.882 755.682,247.093 756.536,246.305 757.39,245.519 758.244,244.735 759.098,243.952 759.952,243.171 760.805,242.393 761.659,241.615 762.513,240.84 763.367,240.066 764.221,239.295 765.075,238.525 765.929,237.756 766.783,236.99 767.637,236.225 768.491,235.462 769.344,234.701 770.198,233.941 771.052,233.184 771.906,232.428 772.76,231.673 773.614,230.921 774.468,230.17 775.322,229.421 776.176,228.674 777.029,227.929 777.883,227.185 778.737,226.443 779.591,225.703 780.445,224.964 781.299,224.228 782.153,223.493 783.007,222.759 783.861,222.028 784.714,221.298 785.568,220.57 786.422,219.844 787.276,219.119 788.13,218.396 788.984,217.675 789.838,216.956 790.692,216.238 791.546,215.523 792.399,214.808 793.253,214.096 794.107,213.385 794.961,212.676 795.815,211.969 796.669,211.264 797.523,210.56 798.377,209.858 799.231,209.158 800.085,208.459 800.938,207.762 801.792,207.067 802.646,206.374 803.5,205.682 804.354,204.992 805.208,204.304 806.062,203.617 806.916,202.932 807.77,202.249 808.623,201.568 809.477,200.888 810.331,200.21 811.185,199.534 812.039,198.859 812.893,198.187 813.747,197.515 814.601,196.846 815.455,196.178 816.308,195.512 817.162,194.848 818.016,194.186 818.87,193.525 819.724,192.866 820.578,192.208 821.432,191.552 822.286,190.898 823.14,190.246 823.994,189.596 824.847,188.947 825.701,188.299 826.555,187.654 827.409,187.01 828.263,186.368 829.117,185.728 829.971,185.089 830.825,184.452 831.679,183.817 832.532,183.183 833.386,182.551 834.24,181.921 835.094,181.293 835.948,180.666 836.802,180.041 837.656,179.418 838.51,178.796 839.364,178.176 840.217,177.558 841.071,176.941 841.925,176.326 842.779,175.713 843.633,175.102 844.487,174.492 845.341,173.884 846.195,173.277 847.049,172.672 847.902,172.069 848.756,171.468 849.61,170.868 850.464,170.27 851.318,169.674 852.172,169.08 853.026,168.487 853.88,167.896 854.734,167.306 855.588,166.718 856.441,166.132 857.295,165.548 858.149,164.965 859.003,164.384 859.857,163.804 860.711,163.227 861.565,162.651 862.419,162.076 863.273,161.504 864.126,160.933 864.98,160.364 865.834,159.796 866.688,159.23 867.542,158.666 868.396,158.104 869.25,157.543 870.104,156.984 870.958,156.426 871.811,155.87 872.665,155.316 873.519,154.764 874.373,154.213 875.227,153.664 876.081,153.117 876.935,152.571 877.789,152.027 878.643,151.485 879.497,150.944 880.35,150.405 881.204,149.868 882.058,149.332 882.912,148.798 883.766,148.266 884.62,147.736 885.474,147.207 886.328,146.679 887.182,146.154 888.035,145.63 888.889,145.108 889.743,144.587 890.597,144.069 891.451,143.552 892.305,143.036 893.159,142.522 894.013,142.01 894.867,141.5 895.72,140.991 896.574,140.484 897.428,139.979 898.282,139.475 899.136,138.973 899.99,138.473 900.844,137.974 901.698,137.477 902.552,136.981 903.406,136.488 904.259,135.996 905.113,135.506 905.967,135.017 906.821,134.53 907.675,134.045 908.529,133.561 909.383,133.079 910.237,132.599 911.091,132.12 911.944,131.643 912.798,131.168 913.652,130.695 914.506,130.223 915.36,129.752 916.214,129.284 917.068,128.817 917.922,128.352 918.776,127.888 919.629,127.426 920.483,126.966 921.337,126.508 922.191,126.051 923.045,125.596 923.899,125.142 924.753,124.69 925.607,124.24 926.461,123.792 927.314,123.345 928.168,122.9 929.022,122.456 929.876,122.014 930.73,121.574 931.584,121.136 932.438,120.699 933.292,120.264 934.146,119.83 935,119.398 935.853,118.968 936.707,118.54 937.561,118.113 938.415,117.688 939.269,117.265 940.123,116.843 940.977,116.423 941.831,116.004 942.685,115.587 943.538,115.172 944.392,114.759 945.246,114.347 946.1,113.937 946.954,113.528 947.808,113.122 948.662,112.717 949.516,112.313 950.37,111.911 951.223,111.511 952.077,111.113 952.931,110.716 953.785,110.321 954.639,109.928 955.493,109.536 956.347,109.146 957.201,108.757 958.055,108.37 958.909,107.985 959.762,107.602 960.616,107.22 961.47,106.84 962.324,106.462 963.178,106.085 964.032,105.71 964.886,105.336 965.74,104.964 966.594,104.594 967.447,104.226 968.301,103.859 969.155,103.494 970.009,103.131 970.863,102.769 971.717,102.409 972.571,102.05 973.425,101.693 974.279,101.338 975.132,100.985 975.986,100.633 976.84,100.283 977.694,99.9343 978.548,99.5875 979.402,99.2424 980.256,98.8989 981.11,98.5571 981.964,98.2169 982.817,97.8784 983.671,97.5416 984.525,97.2064 985.379,96.8729 986.233,96.5411 987.087,96.2109 987.941,95.8824 988.795,95.5556 989.649,95.2304 990.503,94.9069 991.356,94.585 992.21,94.2648 993.064,93.9463 993.918,93.6294 994.772,93.3142 995.626,93.0007 996.48,92.6888 997.334,92.3786 998.188,92.07 999.041,91.7631 999.895,91.4579 1000.75,91.1543 1001.6,90.8524 1002.46,90.5522 1003.31,90.2536 1004.16,89.9567 1005.02,89.6614 1005.87,89.3678 1006.73,89.0759 1007.58,88.7856 1008.43,88.497 1009.29,88.21 1010.14,87.9248 1011,87.6411 1011.85,87.3592 1012.7,87.0789 1013.56,86.8002 1014.41,86.5232 1015.27,86.2479 1016.12,85.9743 1016.97,85.7023 1017.83,85.4319 1018.68,85.1633 1019.53,84.8962 1020.39,84.6309 1021.24,84.3672 1022.1,84.1052 1022.95,83.8448 1023.8,83.5861 1024.66,83.329 1025.51,83.0737 1026.37,82.8199 1027.22,82.5679 1028.07,82.3175 1028.93,82.0687 1029.78,81.8216 1030.64,81.5762 1031.49,81.3324 1032.34,81.0903 1033.2,80.8499 1034.05,80.6111 1034.9,80.374 1035.76,80.1385 1036.61,79.9047 1037.47,79.6726 1038.32,79.4421 1039.17,79.2133 1040.03,78.9861 1040.88,78.7606 1041.74,78.5368 1042.59,78.3146 1043.44,78.0941 1044.3,77.8752 1045.15,77.658 1046.01,77.4425 1046.86,77.2286 1047.71,77.0164 1048.57,76.8058 1049.42,76.5969 1050.27,76.3897 1051.13,76.1841 1051.98,75.9801 1052.84,75.7779 1053.69,75.5773 1054.54,75.3783 1055.4,75.181 1056.25,74.9854 1057.11,74.7914 1057.96,74.5991 1058.81,74.4085 1059.67,74.2195 1060.52,74.0322 1061.38,73.8465 1062.23,73.6625 1063.08,73.4801 1063.94,73.2994 1064.79,73.1204 1065.65,72.943 1066.5,72.7673 1067.35,72.5932 1068.21,72.4208 1069.06,72.2501 1069.91,72.081 1070.77,71.9136 1071.62,71.7478 1072.48,71.5837 1073.33,71.4212 1074.18,71.2605 1075.04,71.1013 1075.89,70.9439 1076.75,70.788 1077.6,70.6339 1078.45,70.4814 1079.31,70.3305 1080.16,70.1814 1081.02,70.0338 1081.87,69.888 1082.72,69.7438 1083.58,69.6012 1084.43,69.4603 1085.28,69.3211 1086.14,69.1835 1086.99,69.0476 1087.85,68.9134 1088.7,68.7808 1089.55,68.6498 1090.41,68.5206 1091.26,68.3929 1092.12,68.267 1092.97,68.1427 1093.82,68.02 1094.68,67.899 1095.53,67.7797 1096.39,67.662 1097.24,67.546 1098.09,67.4317 1098.95,67.319 1099.8,67.2079 1100.65,67.0985 1101.51,66.9908 1102.36,66.8847 1103.22,66.7803 1104.07,66.6776 1104.92,66.5765 1105.78,66.4771 1106.63,66.3793 1107.49,66.2832 1108.34,66.1887 1109.19,66.0959 1110.05,66.0047 1110.9,65.9153 1111.76,65.8274 1112.61,65.7413 1113.46,65.6567 1114.32,65.5739 1115.17,65.4927 1116.02,65.4131 1116.88,65.3352 1117.73,65.259 1118.59,65.1845 1119.44,65.1115 1120.29,65.0403 1121.15,64.9707 1122,64.9028 1122.86,64.8365 1123.71,64.7718 1124.56,64.7089 1125.42,64.6476 1126.27,64.5879 1127.13,64.5299 1127.98,64.4736 1128.83,64.4189 1129.69,64.3659 1130.54,64.3145 1131.39,64.2648 1132.25,64.2168 1133.1,64.1704 1133.96,64.1256 1134.81,64.0826 1135.66,64.0411 1136.52,64.0014 1137.37,63.9633 1138.23,63.9268 1139.08,63.892 1139.93,63.8589 1140.79,63.8274 1141.64,63.7976 1142.5,63.7694 1143.35,63.7429 1144.2,63.7181 1145.06,63.6949 1145.91,63.6734 1146.76,63.6535 1147.62,63.6353 1148.47,63.6187 1149.33,63.6038 1150.18,63.5905 1151.03,63.5789 1151.89,63.569 1152.74,63.5607 1153.6,63.5541 1154.45,63.5491 1155.3,63.5458 1156.16,63.5442 "/>
+<path clip-path="url(#clip570)" d="M1448.54 623.18 L2352.76 623.18 L2352.76 47.2441 L1448.54 47.2441  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip573">
+    <rect x="1448" y="47" width="905" height="577"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1473.27,623.18 1473.27,47.2441 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1686.75,623.18 1686.75,47.2441 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1900.22,623.18 1900.22,47.2441 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2113.69,623.18 2113.69,47.2441 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2327.16,623.18 2327.16,47.2441 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,623.18 2352.76,623.18 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1473.27,623.18 1473.27,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1686.75,623.18 1686.75,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1900.22,623.18 1900.22,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2113.69,623.18 2113.69,604.282 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2327.16,623.18 2327.16,604.282 "/>
+<path clip-path="url(#clip570)" d="M1435.58 654.099 Q1431.96 654.099 1430.14 657.663 Q1428.33 661.205 1428.33 668.335 Q1428.33 675.441 1430.14 679.006 Q1431.96 682.547 1435.58 682.547 Q1439.21 682.547 1441.02 679.006 Q1442.84 675.441 1442.84 668.335 Q1442.84 661.205 1441.02 657.663 Q1439.21 654.099 1435.58 654.099 M1435.58 650.395 Q1441.39 650.395 1444.44 655.001 Q1447.52 659.585 1447.52 668.335 Q1447.52 677.061 1444.44 681.668 Q1441.39 686.251 1435.58 686.251 Q1429.77 686.251 1426.69 681.668 Q1423.63 677.061 1423.63 668.335 Q1423.63 659.585 1426.69 655.001 Q1429.77 650.395 1435.58 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1455.74 679.7 L1460.62 679.7 L1460.62 685.58 L1455.74 685.58 L1455.74 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1480.81 654.099 Q1477.2 654.099 1475.37 657.663 Q1473.56 661.205 1473.56 668.335 Q1473.56 675.441 1475.37 679.006 Q1477.2 682.547 1480.81 682.547 Q1484.44 682.547 1486.25 679.006 Q1488.08 675.441 1488.08 668.335 Q1488.08 661.205 1486.25 657.663 Q1484.44 654.099 1480.81 654.099 M1480.81 650.395 Q1486.62 650.395 1489.67 655.001 Q1492.75 659.585 1492.75 668.335 Q1492.75 677.061 1489.67 681.668 Q1486.62 686.251 1480.81 686.251 Q1475 686.251 1471.92 681.668 Q1468.86 677.061 1468.86 668.335 Q1468.86 659.585 1471.92 655.001 Q1475 650.395 1480.81 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1510.97 654.099 Q1507.36 654.099 1505.53 657.663 Q1503.72 661.205 1503.72 668.335 Q1503.72 675.441 1505.53 679.006 Q1507.36 682.547 1510.97 682.547 Q1514.6 682.547 1516.41 679.006 Q1518.24 675.441 1518.24 668.335 Q1518.24 661.205 1516.41 657.663 Q1514.6 654.099 1510.97 654.099 M1510.97 650.395 Q1516.78 650.395 1519.83 655.001 Q1522.91 659.585 1522.91 668.335 Q1522.91 677.061 1519.83 681.668 Q1516.78 686.251 1510.97 686.251 Q1505.16 686.251 1502.08 681.668 Q1499.02 677.061 1499.02 668.335 Q1499.02 659.585 1502.08 655.001 Q1505.16 650.395 1510.97 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1649.55 654.099 Q1645.94 654.099 1644.11 657.663 Q1642.3 661.205 1642.3 668.335 Q1642.3 675.441 1644.11 679.006 Q1645.94 682.547 1649.55 682.547 Q1653.18 682.547 1654.99 679.006 Q1656.82 675.441 1656.82 668.335 Q1656.82 661.205 1654.99 657.663 Q1653.18 654.099 1649.55 654.099 M1649.55 650.395 Q1655.36 650.395 1658.41 655.001 Q1661.49 659.585 1661.49 668.335 Q1661.49 677.061 1658.41 681.668 Q1655.36 686.251 1649.55 686.251 Q1643.74 686.251 1640.66 681.668 Q1637.6 677.061 1637.6 668.335 Q1637.6 659.585 1640.66 655.001 Q1643.74 650.395 1649.55 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1669.71 679.7 L1674.59 679.7 L1674.59 685.58 L1669.71 685.58 L1669.71 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1694.78 654.099 Q1691.17 654.099 1689.34 657.663 Q1687.53 661.205 1687.53 668.335 Q1687.53 675.441 1689.34 679.006 Q1691.17 682.547 1694.78 682.547 Q1698.41 682.547 1700.22 679.006 Q1702.05 675.441 1702.05 668.335 Q1702.05 661.205 1700.22 657.663 Q1698.41 654.099 1694.78 654.099 M1694.78 650.395 Q1700.59 650.395 1703.64 655.001 Q1706.72 659.585 1706.72 668.335 Q1706.72 677.061 1703.64 681.668 Q1700.59 686.251 1694.78 686.251 Q1688.97 686.251 1685.89 681.668 Q1682.83 677.061 1682.83 668.335 Q1682.83 659.585 1685.89 655.001 Q1688.97 650.395 1694.78 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1714.99 651.02 L1733.34 651.02 L1733.34 654.955 L1719.27 654.955 L1719.27 663.427 Q1720.29 663.08 1721.31 662.918 Q1722.32 662.733 1723.34 662.733 Q1729.13 662.733 1732.51 665.904 Q1735.89 669.075 1735.89 674.492 Q1735.89 680.071 1732.42 683.172 Q1728.94 686.251 1722.62 686.251 Q1720.45 686.251 1718.18 685.881 Q1715.94 685.51 1713.53 684.77 L1713.53 680.071 Q1715.61 681.205 1717.83 681.76 Q1720.06 682.316 1722.53 682.316 Q1726.54 682.316 1728.87 680.21 Q1731.21 678.103 1731.21 674.492 Q1731.21 670.881 1728.87 668.774 Q1726.54 666.668 1722.53 666.668 Q1720.66 666.668 1718.78 667.085 Q1716.93 667.501 1714.99 668.381 L1714.99 651.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1862.52 654.099 Q1858.91 654.099 1857.08 657.663 Q1855.28 661.205 1855.28 668.335 Q1855.28 675.441 1857.08 679.006 Q1858.91 682.547 1862.52 682.547 Q1866.16 682.547 1867.96 679.006 Q1869.79 675.441 1869.79 668.335 Q1869.79 661.205 1867.96 657.663 Q1866.16 654.099 1862.52 654.099 M1862.52 650.395 Q1868.33 650.395 1871.39 655.001 Q1874.47 659.585 1874.47 668.335 Q1874.47 677.061 1871.39 681.668 Q1868.33 686.251 1862.52 686.251 Q1856.71 686.251 1853.63 681.668 Q1850.58 677.061 1850.58 668.335 Q1850.58 659.585 1853.63 655.001 Q1856.71 650.395 1862.52 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1882.68 679.7 L1887.57 679.7 L1887.57 685.58 L1882.68 685.58 L1882.68 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1898.56 681.645 L1906.2 681.645 L1906.2 655.279 L1897.89 656.946 L1897.89 652.687 L1906.16 651.02 L1910.83 651.02 L1910.83 681.645 L1918.47 681.645 L1918.47 685.58 L1898.56 685.58 L1898.56 681.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1937.92 654.099 Q1934.3 654.099 1932.48 657.663 Q1930.67 661.205 1930.67 668.335 Q1930.67 675.441 1932.48 679.006 Q1934.3 682.547 1937.92 682.547 Q1941.55 682.547 1943.35 679.006 Q1945.18 675.441 1945.18 668.335 Q1945.18 661.205 1943.35 657.663 Q1941.55 654.099 1937.92 654.099 M1937.92 650.395 Q1943.73 650.395 1946.78 655.001 Q1949.86 659.585 1949.86 668.335 Q1949.86 677.061 1946.78 681.668 Q1943.73 686.251 1937.92 686.251 Q1932.1 686.251 1929.03 681.668 Q1925.97 677.061 1925.97 668.335 Q1925.97 659.585 1929.03 655.001 Q1932.1 650.395 1937.92 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2076.49 654.099 Q2072.88 654.099 2071.05 657.663 Q2069.25 661.205 2069.25 668.335 Q2069.25 675.441 2071.05 679.006 Q2072.88 682.547 2076.49 682.547 Q2080.13 682.547 2081.93 679.006 Q2083.76 675.441 2083.76 668.335 Q2083.76 661.205 2081.93 657.663 Q2080.13 654.099 2076.49 654.099 M2076.49 650.395 Q2082.3 650.395 2085.36 655.001 Q2088.44 659.585 2088.44 668.335 Q2088.44 677.061 2085.36 681.668 Q2082.3 686.251 2076.49 686.251 Q2070.68 686.251 2067.6 681.668 Q2064.55 677.061 2064.55 668.335 Q2064.55 659.585 2067.6 655.001 Q2070.68 650.395 2076.49 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2096.65 679.7 L2101.54 679.7 L2101.54 685.58 L2096.65 685.58 L2096.65 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2112.53 681.645 L2120.17 681.645 L2120.17 655.279 L2111.86 656.946 L2111.86 652.687 L2120.13 651.02 L2124.8 651.02 L2124.8 681.645 L2132.44 681.645 L2132.44 685.58 L2112.53 685.58 L2112.53 681.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2141.93 651.02 L2160.29 651.02 L2160.29 654.955 L2146.21 654.955 L2146.21 663.427 Q2147.23 663.08 2148.25 662.918 Q2149.27 662.733 2150.29 662.733 Q2156.08 662.733 2159.46 665.904 Q2162.83 669.075 2162.83 674.492 Q2162.83 680.071 2159.36 683.172 Q2155.89 686.251 2149.57 686.251 Q2147.4 686.251 2145.13 685.881 Q2142.88 685.51 2140.47 684.77 L2140.47 680.071 Q2142.56 681.205 2144.78 681.76 Q2147 682.316 2149.48 682.316 Q2153.48 682.316 2155.82 680.21 Q2158.16 678.103 2158.16 674.492 Q2158.16 670.881 2155.82 668.774 Q2153.48 666.668 2149.48 666.668 Q2147.6 666.668 2145.73 667.085 Q2143.88 667.501 2141.93 668.381 L2141.93 651.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2289.47 654.099 Q2285.86 654.099 2284.03 657.663 Q2282.22 661.205 2282.22 668.335 Q2282.22 675.441 2284.03 679.006 Q2285.86 682.547 2289.47 682.547 Q2293.1 682.547 2294.91 679.006 Q2296.74 675.441 2296.74 668.335 Q2296.74 661.205 2294.91 657.663 Q2293.1 654.099 2289.47 654.099 M2289.47 650.395 Q2295.28 650.395 2298.33 655.001 Q2301.41 659.585 2301.41 668.335 Q2301.41 677.061 2298.33 681.668 Q2295.28 686.251 2289.47 686.251 Q2283.66 686.251 2280.58 681.668 Q2277.52 677.061 2277.52 668.335 Q2277.52 659.585 2280.58 655.001 Q2283.66 650.395 2289.47 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2309.63 679.7 L2314.51 679.7 L2314.51 685.58 L2309.63 685.58 L2309.63 679.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2328.73 681.645 L2345.05 681.645 L2345.05 685.58 L2323.1 685.58 L2323.1 681.645 Q2325.76 678.89 2330.35 674.26 Q2334.95 669.608 2336.13 668.265 Q2338.38 665.742 2339.26 664.006 Q2340.16 662.247 2340.16 660.557 Q2340.16 657.802 2338.22 656.066 Q2336.3 654.33 2333.19 654.33 Q2331 654.33 2328.54 655.094 Q2326.11 655.858 2323.33 657.409 L2323.33 652.687 Q2326.16 651.552 2328.61 650.974 Q2331.07 650.395 2333.1 650.395 Q2338.47 650.395 2341.67 653.08 Q2344.86 655.765 2344.86 660.256 Q2344.86 662.386 2344.05 664.307 Q2343.26 666.205 2341.16 668.798 Q2340.58 669.469 2337.48 672.686 Q2334.38 675.881 2328.73 681.645 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2364.86 654.099 Q2361.25 654.099 2359.42 657.663 Q2357.62 661.205 2357.62 668.335 Q2357.62 675.441 2359.42 679.006 Q2361.25 682.547 2364.86 682.547 Q2368.5 682.547 2370.3 679.006 Q2372.13 675.441 2372.13 668.335 Q2372.13 661.205 2370.3 657.663 Q2368.5 654.099 2364.86 654.099 M2364.86 650.395 Q2370.67 650.395 2373.73 655.001 Q2376.81 659.585 2376.81 668.335 Q2376.81 677.061 2373.73 681.668 Q2370.67 686.251 2364.86 686.251 Q2359.05 686.251 2355.97 681.668 Q2352.92 677.061 2352.92 668.335 Q2352.92 659.585 2355.97 655.001 Q2359.05 650.395 2364.86 650.395 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1770.82 720.524 L1811.02 720.524 L1811.02 725.935 L1794.15 725.935 L1794.15 768.044 L1787.69 768.044 L1787.69 725.935 L1770.82 725.935 L1770.82 720.524 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1814.96 732.396 L1820.82 732.396 L1820.82 768.044 L1814.96 768.044 L1814.96 732.396 M1814.96 718.519 L1820.82 718.519 L1820.82 725.935 L1814.96 725.935 L1814.96 718.519 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1860.83 739.239 Q1863.02 735.292 1866.08 733.414 Q1869.14 731.537 1873.27 731.537 Q1878.84 731.537 1881.87 735.451 Q1884.89 739.335 1884.89 746.528 L1884.89 768.044 L1879 768.044 L1879 746.719 Q1879 741.594 1877.19 739.112 Q1875.37 736.629 1871.65 736.629 Q1867.1 736.629 1864.46 739.653 Q1861.81 742.677 1861.81 747.896 L1861.81 768.044 L1855.93 768.044 L1855.93 746.719 Q1855.93 741.563 1854.11 739.112 Q1852.3 736.629 1848.51 736.629 Q1844.02 736.629 1841.38 739.685 Q1838.74 742.708 1838.74 747.896 L1838.74 768.044 L1832.85 768.044 L1832.85 732.396 L1838.74 732.396 L1838.74 737.934 Q1840.74 734.656 1843.55 733.096 Q1846.35 731.537 1850.2 731.537 Q1854.08 731.537 1856.79 733.51 Q1859.52 735.483 1860.83 739.239 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1927.06 748.756 L1927.06 751.62 L1900.14 751.62 Q1900.52 757.668 1903.76 760.851 Q1907.04 764.002 1912.87 764.002 Q1916.24 764.002 1919.39 763.174 Q1922.58 762.347 1925.69 760.691 L1925.69 766.23 Q1922.54 767.566 1919.23 768.267 Q1915.92 768.967 1912.52 768.967 Q1903.99 768.967 1898.99 764.002 Q1894.03 759.036 1894.03 750.57 Q1894.03 741.817 1898.74 736.693 Q1903.48 731.537 1911.5 731.537 Q1918.69 731.537 1922.86 736.183 Q1927.06 740.799 1927.06 748.756 M1921.21 747.037 Q1921.14 742.231 1918.5 739.366 Q1915.89 736.502 1911.56 736.502 Q1906.66 736.502 1903.7 739.271 Q1900.77 742.04 1900.33 747.069 L1921.21 747.037 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1971.46 718.582 Q1967.2 725.903 1965.13 733.064 Q1963.06 740.226 1963.06 747.578 Q1963.06 754.93 1965.13 762.156 Q1967.23 769.349 1971.46 776.638 L1966.37 776.638 Q1961.6 769.158 1959.21 761.933 Q1956.85 754.708 1956.85 747.578 Q1956.85 740.48 1959.21 733.287 Q1961.57 726.094 1966.37 718.582 L1971.46 718.582 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2005.55 733.446 L2005.55 738.984 Q2003.07 737.711 2000.4 737.075 Q1997.72 736.438 1994.86 736.438 Q1990.5 736.438 1988.3 737.775 Q1986.14 739.112 1986.14 741.785 Q1986.14 743.822 1987.7 745 Q1989.26 746.146 1993.97 747.196 L1995.97 747.642 Q2002.21 748.979 2004.82 751.429 Q2007.46 753.848 2007.46 758.209 Q2007.46 763.174 2003.52 766.07 Q1999.6 768.967 1992.73 768.967 Q1989.86 768.967 1986.74 768.394 Q1983.65 767.853 1980.22 766.739 L1980.22 760.691 Q1983.46 762.378 1986.61 763.238 Q1989.77 764.065 1992.85 764.065 Q1996.99 764.065 1999.22 762.665 Q2001.45 761.233 2001.45 758.654 Q2001.45 756.267 1999.82 754.994 Q1998.23 753.721 1992.79 752.543 L1990.75 752.066 Q1985.31 750.92 1982.89 748.565 Q1980.47 746.178 1980.47 742.04 Q1980.47 737.011 1984.04 734.274 Q1987.6 731.537 1994.16 731.537 Q1997.4 731.537 2000.27 732.014 Q2003.13 732.491 2005.55 733.446 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2015.86 718.582 L2020.96 718.582 Q2025.73 726.094 2028.09 733.287 Q2030.47 740.48 2030.47 747.578 Q2030.47 754.708 2028.09 761.933 Q2025.73 769.158 2020.96 776.638 L2015.86 776.638 Q2020.1 769.349 2022.17 762.156 Q2024.27 754.93 2024.27 747.578 Q2024.27 740.226 2022.17 733.064 Q2020.1 725.903 2015.86 718.582 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,606.88 2352.76,606.88 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,471.046 2352.76,471.046 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,335.212 2352.76,335.212 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,199.378 2352.76,199.378 "/>
+<polyline clip-path="url(#clip573)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,63.5442 2352.76,63.5442 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,623.18 1448.54,47.2441 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,606.88 1467.43,606.88 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,471.046 1467.43,471.046 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,335.212 1467.43,335.212 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,199.378 1467.43,199.378 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,63.5442 1467.43,63.5442 "/>
+<path clip-path="url(#clip570)" d="M1355.2 592.679 Q1351.59 592.679 1349.76 596.244 Q1347.95 599.785 1347.95 606.915 Q1347.95 614.021 1349.76 617.586 Q1351.59 621.128 1355.2 621.128 Q1358.83 621.128 1360.64 617.586 Q1362.47 614.021 1362.47 606.915 Q1362.47 599.785 1360.64 596.244 Q1358.83 592.679 1355.2 592.679 M1355.2 588.975 Q1361.01 588.975 1364.06 593.582 Q1367.14 598.165 1367.14 606.915 Q1367.14 615.642 1364.06 620.248 Q1361.01 624.831 1355.2 624.831 Q1349.39 624.831 1346.31 620.248 Q1343.25 615.642 1343.25 606.915 Q1343.25 598.165 1346.31 593.582 Q1349.39 588.975 1355.2 588.975 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1375.36 618.281 L1380.24 618.281 L1380.24 624.16 L1375.36 624.16 L1375.36 618.281 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1401.01 605.017 Q1397.86 605.017 1396.01 607.169 Q1394.18 609.322 1394.18 613.072 Q1394.18 616.799 1396.01 618.975 Q1397.86 621.128 1401.01 621.128 Q1404.16 621.128 1405.98 618.975 Q1407.84 616.799 1407.84 613.072 Q1407.84 609.322 1405.98 607.169 Q1404.16 605.017 1401.01 605.017 M1410.29 590.364 L1410.29 594.623 Q1408.53 593.79 1406.72 593.35 Q1404.94 592.91 1403.18 592.91 Q1398.55 592.91 1396.1 596.035 Q1393.67 599.16 1393.32 605.48 Q1394.69 603.466 1396.75 602.401 Q1398.81 601.313 1401.29 601.313 Q1406.49 601.313 1409.5 604.484 Q1412.54 607.632 1412.54 613.072 Q1412.54 618.396 1409.39 621.614 Q1406.24 624.831 1401.01 624.831 Q1395.01 624.831 1391.84 620.248 Q1388.67 615.642 1388.67 606.915 Q1388.67 598.72 1392.56 593.859 Q1396.45 588.975 1403 588.975 Q1404.76 588.975 1406.54 589.322 Q1408.35 589.67 1410.29 590.364 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1356.26 456.845 Q1352.65 456.845 1350.82 460.41 Q1349.02 463.951 1349.02 471.081 Q1349.02 478.187 1350.82 481.752 Q1352.65 485.294 1356.26 485.294 Q1359.9 485.294 1361.7 481.752 Q1363.53 478.187 1363.53 471.081 Q1363.53 463.951 1361.7 460.41 Q1359.9 456.845 1356.26 456.845 M1356.26 453.141 Q1362.07 453.141 1365.13 457.748 Q1368.21 462.331 1368.21 471.081 Q1368.21 479.808 1365.13 484.414 Q1362.07 488.997 1356.26 488.997 Q1350.45 488.997 1347.37 484.414 Q1344.32 479.808 1344.32 471.081 Q1344.32 462.331 1347.37 457.748 Q1350.45 453.141 1356.26 453.141 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1376.42 482.447 L1381.31 482.447 L1381.31 488.326 L1376.42 488.326 L1376.42 482.447 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1390.31 453.766 L1412.54 453.766 L1412.54 455.757 L1399.99 488.326 L1395.1 488.326 L1406.91 457.701 L1390.31 457.701 L1390.31 453.766 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1355.45 321.011 Q1351.84 321.011 1350.01 324.576 Q1348.21 328.117 1348.21 335.247 Q1348.21 342.353 1350.01 345.918 Q1351.84 349.46 1355.45 349.46 Q1359.09 349.46 1360.89 345.918 Q1362.72 342.353 1362.72 335.247 Q1362.72 328.117 1360.89 324.576 Q1359.09 321.011 1355.45 321.011 M1355.45 317.307 Q1361.26 317.307 1364.32 321.914 Q1367.4 326.497 1367.4 335.247 Q1367.4 343.974 1364.32 348.58 Q1361.26 353.163 1355.45 353.163 Q1349.64 353.163 1346.56 348.58 Q1343.51 343.974 1343.51 335.247 Q1343.51 326.497 1346.56 321.914 Q1349.64 317.307 1355.45 317.307 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1375.61 346.613 L1380.5 346.613 L1380.5 352.492 L1375.61 352.492 L1375.61 346.613 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1400.68 336.08 Q1397.35 336.08 1395.43 337.863 Q1393.53 339.645 1393.53 342.77 Q1393.53 345.895 1395.43 347.677 Q1397.35 349.46 1400.68 349.46 Q1404.02 349.46 1405.94 347.677 Q1407.86 345.872 1407.86 342.77 Q1407.86 339.645 1405.94 337.863 Q1404.04 336.08 1400.68 336.08 M1396.01 334.089 Q1393 333.349 1391.31 331.289 Q1389.64 329.228 1389.64 326.265 Q1389.64 322.122 1392.58 319.715 Q1395.54 317.307 1400.68 317.307 Q1405.85 317.307 1408.79 319.715 Q1411.72 322.122 1411.72 326.265 Q1411.72 329.228 1410.04 331.289 Q1408.37 333.349 1405.38 334.089 Q1408.76 334.877 1410.64 337.168 Q1412.54 339.46 1412.54 342.77 Q1412.54 347.793 1409.46 350.478 Q1406.4 353.163 1400.68 353.163 Q1394.97 353.163 1391.89 350.478 Q1388.83 347.793 1388.83 342.77 Q1388.83 339.46 1390.73 337.168 Q1392.63 334.877 1396.01 334.089 M1394.29 326.705 Q1394.29 329.39 1395.96 330.895 Q1397.65 332.4 1400.68 332.4 Q1403.69 332.4 1405.38 330.895 Q1407.1 329.39 1407.1 326.705 Q1407.1 324.02 1405.38 322.515 Q1403.69 321.011 1400.68 321.011 Q1397.65 321.011 1395.96 322.515 Q1394.29 324.02 1394.29 326.705 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1355.54 185.177 Q1351.93 185.177 1350.1 188.742 Q1348.3 192.283 1348.3 199.413 Q1348.3 206.519 1350.1 210.084 Q1351.93 213.626 1355.54 213.626 Q1359.18 213.626 1360.98 210.084 Q1362.81 206.519 1362.81 199.413 Q1362.81 192.283 1360.98 188.742 Q1359.18 185.177 1355.54 185.177 M1355.54 181.473 Q1361.35 181.473 1364.41 186.08 Q1367.49 190.663 1367.49 199.413 Q1367.49 208.14 1364.41 212.746 Q1361.35 217.329 1355.54 217.329 Q1349.73 217.329 1346.66 212.746 Q1343.6 208.14 1343.6 199.413 Q1343.6 190.663 1346.66 186.08 Q1349.73 181.473 1355.54 181.473 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1375.71 210.779 L1380.59 210.779 L1380.59 216.658 L1375.71 216.658 L1375.71 210.779 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1390.91 215.941 L1390.91 211.681 Q1392.67 212.515 1394.48 212.954 Q1396.29 213.394 1398.02 213.394 Q1402.65 213.394 1405.08 210.292 Q1407.54 207.167 1407.88 200.825 Q1406.54 202.816 1404.48 203.88 Q1402.42 204.945 1399.92 204.945 Q1394.73 204.945 1391.7 201.82 Q1388.69 198.672 1388.69 193.232 Q1388.69 187.908 1391.84 184.691 Q1394.99 181.473 1400.22 181.473 Q1406.22 181.473 1409.36 186.08 Q1412.54 190.663 1412.54 199.413 Q1412.54 207.584 1408.65 212.468 Q1404.78 217.329 1398.23 217.329 Q1396.47 217.329 1394.66 216.982 Q1392.86 216.635 1390.91 215.941 M1400.22 201.288 Q1403.37 201.288 1405.2 199.135 Q1407.05 196.982 1407.05 193.232 Q1407.05 189.506 1405.2 187.353 Q1403.37 185.177 1400.22 185.177 Q1397.07 185.177 1395.22 187.353 Q1393.39 189.506 1393.39 193.232 Q1393.39 196.982 1395.22 199.135 Q1397.07 201.288 1400.22 201.288 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1346.17 76.889 L1353.81 76.889 L1353.81 50.5234 L1345.5 52.1901 L1345.5 47.9308 L1353.76 46.2642 L1358.44 46.2642 L1358.44 76.889 L1366.08 76.889 L1366.08 80.8242 L1346.17 80.8242 L1346.17 76.889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1375.52 74.9446 L1380.41 74.9446 L1380.41 80.8242 L1375.52 80.8242 L1375.52 74.9446 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1400.59 49.3429 Q1396.98 49.3429 1395.15 52.9076 Q1393.35 56.4493 1393.35 63.5789 Q1393.35 70.6853 1395.15 74.2501 Q1396.98 77.7918 1400.59 77.7918 Q1404.22 77.7918 1406.03 74.2501 Q1407.86 70.6853 1407.86 63.5789 Q1407.86 56.4493 1406.03 52.9076 Q1404.22 49.3429 1400.59 49.3429 M1400.59 45.6392 Q1406.4 45.6392 1409.46 50.2456 Q1412.54 54.8289 1412.54 63.5789 Q1412.54 72.3057 1409.46 76.9121 Q1406.4 81.4955 1400.59 81.4955 Q1394.78 81.4955 1391.7 76.9121 Q1388.65 72.3057 1388.65 63.5789 Q1388.65 54.8289 1391.7 50.2456 Q1394.78 45.6392 1400.59 45.6392 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1245.48 412.476 L1245.48 402.895 L1277.81 390.769 L1245.48 378.578 L1245.48 368.998 L1293 368.998 L1293 375.268 L1251.27 375.268 L1283.86 387.522 L1283.86 393.983 L1251.27 406.237 L1293 406.237 L1293 412.476 L1245.48 412.476 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1275.08 340.289 Q1275.08 347.386 1276.7 350.124 Q1278.32 352.861 1282.24 352.861 Q1285.36 352.861 1287.2 350.824 Q1289.02 348.755 1289.02 345.222 Q1289.02 340.352 1285.58 337.424 Q1282.11 334.464 1276.38 334.464 L1275.08 334.464 L1275.08 340.289 M1272.66 328.608 L1293 328.608 L1293 334.464 L1287.59 334.464 Q1290.83 336.469 1292.39 339.461 Q1293.92 342.453 1293.92 346.782 Q1293.92 352.256 1290.86 355.503 Q1287.78 358.717 1282.62 358.717 Q1276.61 358.717 1273.55 354.707 Q1270.49 350.665 1270.49 342.676 L1270.49 334.464 L1269.92 334.464 Q1265.88 334.464 1263.68 337.138 Q1261.45 339.779 1261.45 344.585 Q1261.45 347.641 1262.19 350.537 Q1262.92 353.434 1264.38 356.107 L1258.97 356.107 Q1257.73 352.893 1257.13 349.869 Q1256.49 346.845 1256.49 343.981 Q1256.49 336.246 1260.5 332.427 Q1264.51 328.608 1272.66 328.608 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1258.4 293.819 L1263.94 293.819 Q1262.66 296.302 1262.03 298.975 Q1261.39 301.649 1261.39 304.513 Q1261.39 308.874 1262.73 311.07 Q1264.06 313.234 1266.74 313.234 Q1268.78 313.234 1269.95 311.675 Q1271.1 310.115 1272.15 305.405 L1272.59 303.399 Q1273.93 297.161 1276.38 294.551 Q1278.8 291.909 1283.16 291.909 Q1288.13 291.909 1291.02 295.856 Q1293.92 299.771 1293.92 306.646 Q1293.92 309.51 1293.35 312.63 Q1292.81 315.717 1291.69 319.154 L1285.64 319.154 Q1287.33 315.908 1288.19 312.757 Q1289.02 309.606 1289.02 306.519 Q1289.02 302.381 1287.62 300.153 Q1286.19 297.925 1283.61 297.925 Q1281.22 297.925 1279.95 299.548 Q1278.67 301.14 1277.5 306.582 L1277.02 308.619 Q1275.87 314.062 1273.52 316.481 Q1271.13 318.9 1266.99 318.9 Q1261.96 318.9 1259.23 315.335 Q1256.49 311.77 1256.49 305.214 Q1256.49 301.967 1256.97 299.102 Q1257.44 296.238 1258.4 293.819 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1258.4 259.858 L1263.94 259.858 Q1262.66 262.341 1262.03 265.014 Q1261.39 267.688 1261.39 270.552 Q1261.39 274.913 1262.73 277.109 Q1264.06 279.273 1266.74 279.273 Q1268.78 279.273 1269.95 277.714 Q1271.1 276.154 1272.15 271.444 L1272.59 269.438 Q1273.93 263.2 1276.38 260.59 Q1278.8 257.948 1283.16 257.948 Q1288.13 257.948 1291.02 261.895 Q1293.92 265.81 1293.92 272.685 Q1293.92 275.549 1293.35 278.669 Q1292.81 281.756 1291.69 285.193 L1285.64 285.193 Q1287.33 281.947 1288.19 278.796 Q1289.02 275.645 1289.02 272.558 Q1289.02 268.42 1287.62 266.192 Q1286.19 263.964 1283.61 263.964 Q1281.22 263.964 1279.95 265.587 Q1278.67 267.178 1277.5 272.621 L1277.02 274.658 Q1275.87 280.101 1273.52 282.52 Q1271.13 284.939 1266.99 284.939 Q1261.96 284.939 1259.23 281.374 Q1256.49 277.809 1256.49 271.253 Q1256.49 268.006 1256.97 265.141 Q1257.44 262.277 1258.4 259.858 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip573)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1474.13,63.5442 1474.98,65.4457 1475.83,67.3472 1476.69,69.2488 1477.54,71.1503 1478.4,73.0518 1479.25,74.9533 1480.1,76.8549 1480.96,78.7564 1481.81,80.6579 1482.67,82.5594 1483.52,84.4609 1484.37,86.3624 1485.23,88.2639 1486.08,90.1654 1486.93,92.0669 1487.79,93.9684 1488.64,95.8699 1489.5,97.7713 1490.35,99.6728 1491.2,101.574 1492.06,103.476 1492.91,105.377 1493.77,107.279 1494.62,109.18 1495.47,111.082 1496.33,112.983 1497.18,114.884 1498.04,116.786 1498.89,118.687 1499.74,120.589 1500.6,122.49 1501.45,124.392 1502.3,126.293 1503.16,128.194 1504.01,130.096 1504.87,131.997 1505.72,133.898 1506.57,135.8 1507.43,137.701 1508.28,139.602 1509.14,141.504 1509.99,143.405 1510.84,145.306 1511.7,147.208 1512.55,149.109 1513.41,151.01 1514.26,152.912 1515.11,154.813 1515.97,156.714 1516.82,158.615 1517.67,160.516 1518.53,162.418 1519.38,164.319 1520.24,166.22 1521.09,168.121 1521.94,170.022 1522.8,171.923 1523.65,173.824 1524.51,175.725 1525.36,177.626 1526.21,179.527 1527.07,181.428 1527.92,183.329 1528.78,185.23 1529.63,187.131 1530.48,189.032 1531.34,190.933 1532.19,192.833 1533.04,194.734 1533.9,196.635 1534.75,198.536 1535.61,200.436 1536.46,202.337 1537.31,204.237 1538.17,206.138 1539.02,208.038 1539.88,209.938 1540.73,211.838 1541.58,213.738 1542.44,215.638 1543.29,217.538 1544.15,219.438 1545,221.338 1545.85,223.237 1546.71,225.136 1547.56,227.036 1548.41,228.935 1549.27,230.833 1550.12,232.732 1550.98,234.63 1551.83,236.528 1552.68,238.426 1553.54,240.323 1554.39,242.22 1555.25,244.116 1556.1,246.011 1556.95,247.906 1557.81,249.8 1558.66,251.692 1559.52,253.584 1560.37,255.473 1561.22,257.361 1562.08,259.246 1562.93,261.127 1563.78,263.003 1564.64,264.871 1565.49,266.729 1566.35,268.57 1567.2,270.384 1568.05,272.151 1568.91,273.849 1569.76,275.461 1570.62,276.985 1571.47,278.433 1572.32,279.816 1573.18,281.146 1574.03,282.435 1574.89,283.689 1575.74,284.918 1576.59,286.126 1577.45,287.317 1578.3,288.495 1579.16,289.663 1580.01,290.824 1580.86,291.978 1581.72,293.127 1582.57,294.273 1583.42,295.415 1584.28,296.556 1585.13,297.694 1585.99,298.832 1586.84,299.968 1587.69,301.103 1588.55,302.238 1589.4,303.373 1590.26,304.507 1591.11,305.641 1591.96,306.775 1592.82,307.909 1593.67,309.043 1594.53,310.177 1595.38,311.311 1596.23,312.445 1597.09,313.58 1597.94,314.715 1598.79,315.849 1599.65,316.985 1600.5,318.12 1601.36,319.256 1602.21,320.392 1603.06,321.528 1603.92,322.664 1604.77,323.801 1605.63,324.938 1606.48,326.076 1607.33,327.214 1608.19,328.352 1609.04,329.49 1609.9,330.629 1610.75,331.768 1611.6,332.908 1612.46,334.048 1613.31,335.188 1614.16,336.329 1615.02,337.47 1615.87,338.612 1616.73,339.754 1617.58,340.896 1618.43,342.039 1619.29,343.182 1620.14,344.326 1621,345.471 1621.85,346.615 1622.7,347.761 1623.56,348.906 1624.41,350.053 1625.27,351.199 1626.12,352.347 1626.97,353.494 1627.83,354.643 1628.68,355.792 1629.53,356.941 1630.39,358.091 1631.24,359.242 1632.1,360.393 1632.95,361.545 1633.8,362.698 1634.66,363.851 1635.51,365.005 1636.37,366.159 1637.22,367.314 1638.07,368.47 1638.93,369.626 1639.78,370.783 1640.64,371.941 1641.49,373.1 1642.34,374.259 1643.2,375.419 1644.05,376.58 1644.9,377.741 1645.76,378.903 1646.61,380.066 1647.47,381.23 1648.32,382.395 1649.17,383.56 1650.03,384.726 1650.88,385.894 1651.74,387.062 1652.59,388.23 1653.44,389.4 1654.3,390.571 1655.15,391.742 1656.01,392.915 1656.86,394.088 1657.71,395.262 1658.57,396.437 1659.42,397.614 1660.27,398.791 1661.13,399.969 1661.98,401.148 1662.84,402.329 1663.69,403.51 1664.54,404.692 1665.4,405.876 1666.25,407.06 1667.11,408.246 1667.96,409.432 1668.81,410.62 1669.67,411.809 1670.52,412.999 1671.38,414.191 1672.23,415.383 1673.08,416.577 1673.94,417.772 1674.79,418.968 1675.64,420.165 1676.5,421.364 1677.35,422.564 1678.21,423.765 1679.06,424.968 1679.91,426.171 1680.77,427.377 1681.62,428.583 1682.48,429.791 1683.33,431.001 1684.18,432.212 1685.04,433.424 1685.89,434.638 1686.75,435.853 1687.6,437.07 1688.45,438.288 1689.31,439.508 1690.16,440.729 1691.01,441.952 1691.87,443.176 1692.72,444.402 1693.58,445.63 1694.43,446.86 1695.28,448.091 1696.14,449.323 1696.99,450.558 1697.85,451.794 1698.7,453.032 1699.55,454.272 1700.41,455.514 1701.26,456.757 1702.12,458.002 1702.97,459.249 1703.82,460.498 1704.68,461.749 1705.53,463.002 1706.38,464.257 1707.24,465.514 1708.09,466.773 1708.95,468.034 1709.8,469.297 1710.65,470.562 1711.51,471.829 1712.36,473.098 1713.22,474.369 1714.07,475.643 1714.92,476.919 1715.78,478.197 1716.63,479.477 1717.49,480.76 1718.34,482.045 1719.19,483.332 1720.05,484.622 1720.9,485.914 1721.76,487.209 1722.61,488.506 1723.46,489.805 1724.32,491.108 1725.17,492.412 1726.02,493.719 1726.88,495.029 1727.73,496.342 1728.59,497.657 1729.44,498.975 1730.29,500.295 1731.15,501.619 1732,502.945 1732.86,504.274 1733.71,505.606 1734.56,506.941 1735.42,508.279 1736.27,509.619 1737.13,510.963 1737.98,512.31 1738.83,513.66 1739.69,515.012 1740.54,516.368 1741.39,517.728 1742.25,519.09 1743.1,520.456 1743.96,521.825 1744.81,523.197 1745.66,524.572 1746.52,525.951 1747.37,527.333 1748.23,528.718 1749.08,530.107 1749.93,531.499 1750.79,532.895 1751.64,534.294 1752.5,535.697 1753.35,537.103 1754.2,538.512 1755.06,539.925 1755.91,541.341 1756.76,542.761 1757.62,544.184 1758.47,545.61 1759.33,547.039 1760.18,548.472 1761.03,549.908 1761.89,551.347 1762.74,552.789 1763.6,554.233 1764.45,555.681 1765.3,557.131 1766.16,558.583 1767.01,560.037 1767.87,561.493 1768.72,562.951 1769.57,564.41 1770.43,565.87 1771.28,567.331 1772.13,568.792 1772.99,570.252 1773.84,571.711 1774.7,573.169 1775.55,574.624 1776.4,576.076 1777.26,577.524 1778.11,578.967 1778.97,580.404 1779.82,581.833 1780.67,583.253 1781.53,584.663 1782.38,586.061 1783.24,587.444 1784.09,588.81 1784.94,590.158 1785.8,591.483 1786.65,592.784 1787.5,594.056 1788.36,595.296 1789.21,596.499 1790.07,597.66 1790.92,598.775 1791.77,599.838 1792.63,600.843 1793.48,601.782 1794.34,602.649 1795.19,603.437 1796.04,604.137 1796.9,604.742 1797.75,605.246 1798.61,605.643 1799.46,605.936 1800.31,606.136 1801.17,606.264 1802.02,606.349 1802.87,606.408 1803.73,606.452 1804.58,606.486 1805.44,606.513 1806.29,606.536 1807.14,606.555 1808,606.572 1808.85,606.587 1809.71,606.599 1810.56,606.611 1811.41,606.621 1812.27,606.63 1813.12,606.639 1813.98,606.646 1814.83,606.653 1815.68,606.66 1816.54,606.666 1817.39,606.671 1818.24,606.677 1819.1,606.681 1819.95,606.686 1820.81,606.69 1821.66,606.694 1822.51,606.698 1823.37,606.702 1824.22,606.705 1825.08,606.708 1825.93,606.711 1826.78,606.714 1827.64,606.717 1828.49,606.72 1829.35,606.722 1830.2,606.725 1831.05,606.727 1831.91,606.729 1832.76,606.731 1833.61,606.734 1834.47,606.736 1835.32,606.737 1836.18,606.739 1837.03,606.741 1837.88,606.743 1838.74,606.744 1839.59,606.746 1840.45,606.748 1841.3,606.749 1842.15,606.751 1843.01,606.752 1843.86,606.753 1844.72,606.755 1845.57,606.756 1846.42,606.757 1847.28,606.759 1848.13,606.76 1848.99,606.761 1849.84,606.762 1850.69,606.763 1851.55,606.764 1852.4,606.765 1853.25,606.766 1854.11,606.767 1854.96,606.768 1855.82,606.769 1856.67,606.77 1857.52,606.771 1858.38,606.772 1859.23,606.773 1860.09,606.774 1860.94,606.774 1861.79,606.775 1862.65,606.776 1863.5,606.777 1864.36,606.778 1865.21,606.778 1866.06,606.779 1866.92,606.78 1867.77,606.781 1868.62,606.781 1869.48,606.782 1870.33,606.783 1871.19,606.783 1872.04,606.784 1872.89,606.785 1873.75,606.785 1874.6,606.786 1875.46,606.787 1876.31,606.787 1877.16,606.788 1878.02,606.788 1878.87,606.789 1879.73,606.789 1880.58,606.79 1881.43,606.791 1882.29,606.791 1883.14,606.792 1883.99,606.792 1884.85,606.793 1885.7,606.793 1886.56,606.794 1887.41,606.794 1888.26,606.795 1889.12,606.795 1889.97,606.796 1890.83,606.796 1891.68,606.797 1892.53,606.797 1893.39,606.797 1894.24,606.798 1895.1,606.798 1895.95,606.799 1896.8,606.799 1897.66,606.8 1898.51,606.8 1899.36,606.8 1900.22,606.801 1901.07,606.801 1901.93,606.802 1902.78,606.802 1903.63,606.802 1904.49,606.803 1905.34,606.803 1906.2,606.804 1907.05,606.804 1907.9,606.804 1908.76,606.805 1909.61,606.805 1910.47,606.805 1911.32,606.806 1912.17,606.806 1913.03,606.806 1913.88,606.807 1914.73,606.807 1915.59,606.807 1916.44,606.808 1917.3,606.808 1918.15,606.808 1919,606.809 1919.86,606.809 1920.71,606.809 1921.57,606.81 1922.42,606.81 1923.27,606.81 1924.13,606.811 1924.98,606.811 1925.84,606.811 1926.69,606.812 1927.54,606.812 1928.4,606.812 1929.25,606.812 1930.1,606.813 1930.96,606.813 1931.81,606.813 1932.67,606.814 1933.52,606.814 1934.37,606.814 1935.23,606.814 1936.08,606.815 1936.94,606.815 1937.79,606.815 1938.64,606.815 1939.5,606.816 1940.35,606.816 1941.21,606.816 1942.06,606.817 1942.91,606.817 1943.77,606.817 1944.62,606.817 1945.47,606.818 1946.33,606.818 1947.18,606.818 1948.04,606.818 1948.89,606.819 1949.74,606.819 1950.6,606.819 1951.45,606.819 1952.31,606.819 1953.16,606.82 1954.01,606.82 1954.87,606.82 1955.72,606.82 1956.58,606.821 1957.43,606.821 1958.28,606.821 1959.14,606.821 1959.99,606.822 1960.84,606.822 1961.7,606.822 1962.55,606.822 1963.41,606.822 1964.26,606.823 1965.11,606.823 1965.97,606.823 1966.82,606.823 1967.68,606.823 1968.53,606.824 1969.38,606.824 1970.24,606.824 1971.09,606.824 1971.95,606.824 1972.8,606.825 1973.65,606.825 1974.51,606.825 1975.36,606.825 1976.21,606.825 1977.07,606.826 1977.92,606.826 1978.78,606.826 1979.63,606.826 1980.48,606.826 1981.34,606.827 1982.19,606.827 1983.05,606.827 1983.9,606.827 1984.75,606.827 1985.61,606.828 1986.46,606.828 1987.32,606.828 1988.17,606.828 1989.02,606.828 1989.88,606.829 1990.73,606.829 1991.59,606.829 1992.44,606.829 1993.29,606.829 1994.15,606.829 1995,606.83 1995.85,606.83 1996.71,606.83 1997.56,606.83 1998.42,606.83 1999.27,606.83 2000.12,606.831 2000.98,606.831 2001.83,606.831 2002.69,606.831 2003.54,606.831 2004.39,606.831 2005.25,606.832 2006.1,606.832 2006.96,606.832 2007.81,606.832 2008.66,606.832 2009.52,606.832 2010.37,606.833 2011.22,606.833 2012.08,606.833 2012.93,606.833 2013.79,606.833 2014.64,606.833 2015.49,606.834 2016.35,606.834 2017.2,606.834 2018.06,606.834 2018.91,606.834 2019.76,606.834 2020.62,606.835 2021.47,606.835 2022.33,606.835 2023.18,606.835 2024.03,606.835 2024.89,606.835 2025.74,606.835 2026.59,606.836 2027.45,606.836 2028.3,606.836 2029.16,606.836 2030.01,606.836 2030.86,606.836 2031.72,606.836 2032.57,606.837 2033.43,606.837 2034.28,606.837 2035.13,606.837 2035.99,606.837 2036.84,606.837 2037.7,606.837 2038.55,606.838 2039.4,606.838 2040.26,606.838 2041.11,606.838 2041.96,606.838 2042.82,606.838 2043.67,606.838 2044.53,606.839 2045.38,606.839 2046.23,606.839 2047.09,606.839 2047.94,606.839 2048.8,606.839 2049.65,606.839 2050.5,606.84 2051.36,606.84 2052.21,606.84 2053.07,606.84 2053.92,606.84 2054.77,606.84 2055.63,606.84 2056.48,606.841 2057.33,606.841 2058.19,606.841 2059.04,606.841 2059.9,606.841 2060.75,606.841 2061.6,606.841 2062.46,606.841 2063.31,606.842 2064.17,606.842 2065.02,606.842 2065.87,606.842 2066.73,606.842 2067.58,606.842 2068.44,606.842 2069.29,606.842 2070.14,606.843 2071,606.843 2071.85,606.843 2072.7,606.843 2073.56,606.843 2074.41,606.843 2075.27,606.843 2076.12,606.843 2076.97,606.844 2077.83,606.844 2078.68,606.844 2079.54,606.844 2080.39,606.844 2081.24,606.844 2082.1,606.844 2082.95,606.844 2083.81,606.845 2084.66,606.845 2085.51,606.845 2086.37,606.845 2087.22,606.845 2088.07,606.845 2088.93,606.845 2089.78,606.845 2090.64,606.846 2091.49,606.846 2092.34,606.846 2093.2,606.846 2094.05,606.846 2094.91,606.846 2095.76,606.846 2096.61,606.846 2097.47,606.846 2098.32,606.847 2099.18,606.847 2100.03,606.847 2100.88,606.847 2101.74,606.847 2102.59,606.847 2103.44,606.847 2104.3,606.847 2105.15,606.848 2106.01,606.848 2106.86,606.848 2107.71,606.848 2108.57,606.848 2109.42,606.848 2110.28,606.848 2111.13,606.848 2111.98,606.848 2112.84,606.849 2113.69,606.849 2114.55,606.849 2115.4,606.849 2116.25,606.849 2117.11,606.849 2117.96,606.849 2118.82,606.849 2119.67,606.849 2120.52,606.85 2121.38,606.85 2122.23,606.85 2123.08,606.85 2123.94,606.85 2124.79,606.85 2125.65,606.85 2126.5,606.85 2127.35,606.85 2128.21,606.851 2129.06,606.851 2129.92,606.851 2130.77,606.851 2131.62,606.851 2132.48,606.851 2133.33,606.851 2134.19,606.851 2135.04,606.851 2135.89,606.852 2136.75,606.852 2137.6,606.852 2138.45,606.852 2139.31,606.852 2140.16,606.852 2141.02,606.852 2141.87,606.852 2142.72,606.852 2143.58,606.853 2144.43,606.853 2145.29,606.853 2146.14,606.853 2146.99,606.853 2147.85,606.853 2148.7,606.853 2149.56,606.853 2150.41,606.853 2151.26,606.853 2152.12,606.854 2152.97,606.854 2153.82,606.854 2154.68,606.854 2155.53,606.854 2156.39,606.854 2157.24,606.854 2158.09,606.854 2158.95,606.854 2159.8,606.855 2160.66,606.855 2161.51,606.855 2162.36,606.855 2163.22,606.855 2164.07,606.855 2164.93,606.855 2165.78,606.855 2166.63,606.855 2167.49,606.855 2168.34,606.856 2169.19,606.856 2170.05,606.856 2170.9,606.856 2171.76,606.856 2172.61,606.856 2173.46,606.856 2174.32,606.856 2175.17,606.856 2176.03,606.856 2176.88,606.857 2177.73,606.857 2178.59,606.857 2179.44,606.857 2180.3,606.857 2181.15,606.857 2182,606.857 2182.86,606.857 2183.71,606.857 2184.56,606.857 2185.42,606.858 2186.27,606.858 2187.13,606.858 2187.98,606.858 2188.83,606.858 2189.69,606.858 2190.54,606.858 2191.4,606.858 2192.25,606.858 2193.1,606.859 2193.96,606.859 2194.81,606.859 2195.67,606.859 2196.52,606.859 2197.37,606.859 2198.23,606.859 2199.08,606.859 2199.93,606.859 2200.79,606.859 2201.64,606.86 2202.5,606.86 2203.35,606.86 2204.2,606.86 2205.06,606.86 2205.91,606.86 2206.77,606.86 2207.62,606.86 2208.47,606.86 2209.33,606.86 2210.18,606.861 2211.04,606.861 2211.89,606.861 2212.74,606.861 2213.6,606.861 2214.45,606.861 2215.3,606.861 2216.16,606.861 2217.01,606.861 2217.87,606.861 2218.72,606.862 2219.57,606.862 2220.43,606.862 2221.28,606.862 2222.14,606.862 2222.99,606.862 2223.84,606.862 2224.7,606.862 2225.55,606.862 2226.41,606.862 2227.26,606.863 2228.11,606.863 2228.97,606.863 2229.82,606.863 2230.67,606.863 2231.53,606.863 2232.38,606.863 2233.24,606.863 2234.09,606.863 2234.94,606.863 2235.8,606.864 2236.65,606.864 2237.51,606.864 2238.36,606.864 2239.21,606.864 2240.07,606.864 2240.92,606.864 2241.78,606.864 2242.63,606.864 2243.48,606.865 2244.34,606.865 2245.19,606.865 2246.04,606.865 2246.9,606.865 2247.75,606.865 2248.61,606.865 2249.46,606.865 2250.31,606.865 2251.17,606.865 2252.02,606.866 2252.88,606.866 2253.73,606.866 2254.58,606.866 2255.44,606.866 2256.29,606.866 2257.15,606.866 2258,606.866 2258.85,606.866 2259.71,606.867 2260.56,606.867 2261.42,606.867 2262.27,606.867 2263.12,606.867 2263.98,606.867 2264.83,606.867 2265.68,606.867 2266.54,606.867 2267.39,606.868 2268.25,606.868 2269.1,606.868 2269.95,606.868 2270.81,606.868 2271.66,606.868 2272.52,606.868 2273.37,606.868 2274.22,606.868 2275.08,606.869 2275.93,606.869 2276.79,606.869 2277.64,606.869 2278.49,606.869 2279.35,606.869 2280.2,606.869 2281.05,606.869 2281.91,606.87 2282.76,606.87 2283.62,606.87 2284.47,606.87 2285.32,606.87 2286.18,606.87 2287.03,606.87 2287.89,606.87 2288.74,606.871 2289.59,606.871 2290.45,606.871 2291.3,606.871 2292.16,606.871 2293.01,606.871 2293.86,606.871 2294.72,606.871 2295.57,606.872 2296.42,606.872 2297.28,606.872 2298.13,606.872 2298.99,606.872 2299.84,606.872 2300.69,606.872 2301.55,606.873 2302.4,606.873 2303.26,606.873 2304.11,606.873 2304.96,606.873 2305.82,606.873 2306.67,606.873 2307.53,606.874 2308.38,606.874 2309.23,606.874 2310.09,606.874 2310.94,606.874 2311.79,606.874 2312.65,606.875 2313.5,606.875 2314.36,606.875 2315.21,606.875 2316.06,606.875 2316.92,606.876 2317.77,606.876 2318.63,606.876 2319.48,606.876 2320.33,606.876 2321.19,606.877 2322.04,606.877 2322.9,606.877 2323.75,606.878 2324.6,606.878 2325.46,606.878 2326.31,606.879 2327.16,606.88 "/>
+<path clip-path="url(#clip570)" d="M277.528 1423.18 L1181.75 1423.18 L1181.75 847.244 L277.528 847.244  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip574">
+    <rect x="277" y="847" width="905" height="577"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="302.265,1423.18 302.265,847.244 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="515.738,1423.18 515.738,847.244 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="729.211,1423.18 729.211,847.244 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="942.685,1423.18 942.685,847.244 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1156.16,1423.18 1156.16,847.244 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,1423.18 1181.75,1423.18 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="302.265,1423.18 302.265,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="515.738,1423.18 515.738,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="729.211,1423.18 729.211,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="942.685,1423.18 942.685,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1156.16,1423.18 1156.16,1404.28 "/>
+<path clip-path="url(#clip570)" d="M264.569 1454.1 Q260.958 1454.1 259.129 1457.66 Q257.323 1461.2 257.323 1468.33 Q257.323 1475.44 259.129 1479.01 Q260.958 1482.55 264.569 1482.55 Q268.203 1482.55 270.008 1479.01 Q271.837 1475.44 271.837 1468.33 Q271.837 1461.2 270.008 1457.66 Q268.203 1454.1 264.569 1454.1 M264.569 1450.39 Q270.379 1450.39 273.434 1455 Q276.513 1459.58 276.513 1468.33 Q276.513 1477.06 273.434 1481.67 Q270.379 1486.25 264.569 1486.25 Q258.759 1486.25 255.68 1481.67 Q252.624 1477.06 252.624 1468.33 Q252.624 1459.58 255.68 1455 Q258.759 1450.39 264.569 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M284.731 1479.7 L289.615 1479.7 L289.615 1485.58 L284.731 1485.58 L284.731 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M309.8 1454.1 Q306.189 1454.1 304.36 1457.66 Q302.555 1461.2 302.555 1468.33 Q302.555 1475.44 304.36 1479.01 Q306.189 1482.55 309.8 1482.55 Q313.434 1482.55 315.24 1479.01 Q317.068 1475.44 317.068 1468.33 Q317.068 1461.2 315.24 1457.66 Q313.434 1454.1 309.8 1454.1 M309.8 1450.39 Q315.61 1450.39 318.666 1455 Q321.744 1459.58 321.744 1468.33 Q321.744 1477.06 318.666 1481.67 Q315.61 1486.25 309.8 1486.25 Q303.99 1486.25 300.911 1481.67 Q297.856 1477.06 297.856 1468.33 Q297.856 1459.58 300.911 1455 Q303.99 1450.39 309.8 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M339.962 1454.1 Q336.351 1454.1 334.522 1457.66 Q332.716 1461.2 332.716 1468.33 Q332.716 1475.44 334.522 1479.01 Q336.351 1482.55 339.962 1482.55 Q343.596 1482.55 345.402 1479.01 Q347.23 1475.44 347.23 1468.33 Q347.23 1461.2 345.402 1457.66 Q343.596 1454.1 339.962 1454.1 M339.962 1450.39 Q345.772 1450.39 348.828 1455 Q351.906 1459.58 351.906 1468.33 Q351.906 1477.06 348.828 1481.67 Q345.772 1486.25 339.962 1486.25 Q334.152 1486.25 331.073 1481.67 Q328.017 1477.06 328.017 1468.33 Q328.017 1459.58 331.073 1455 Q334.152 1450.39 339.962 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M478.539 1454.1 Q474.928 1454.1 473.1 1457.66 Q471.294 1461.2 471.294 1468.33 Q471.294 1475.44 473.1 1479.01 Q474.928 1482.55 478.539 1482.55 Q482.174 1482.55 483.979 1479.01 Q485.808 1475.44 485.808 1468.33 Q485.808 1461.2 483.979 1457.66 Q482.174 1454.1 478.539 1454.1 M478.539 1450.39 Q484.35 1450.39 487.405 1455 Q490.484 1459.58 490.484 1468.33 Q490.484 1477.06 487.405 1481.67 Q484.35 1486.25 478.539 1486.25 Q472.729 1486.25 469.651 1481.67 Q466.595 1477.06 466.595 1468.33 Q466.595 1459.58 469.651 1455 Q472.729 1450.39 478.539 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M498.701 1479.7 L503.586 1479.7 L503.586 1485.58 L498.701 1485.58 L498.701 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M523.771 1454.1 Q520.16 1454.1 518.331 1457.66 Q516.525 1461.2 516.525 1468.33 Q516.525 1475.44 518.331 1479.01 Q520.16 1482.55 523.771 1482.55 Q527.405 1482.55 529.211 1479.01 Q531.039 1475.44 531.039 1468.33 Q531.039 1461.2 529.211 1457.66 Q527.405 1454.1 523.771 1454.1 M523.771 1450.39 Q529.581 1450.39 532.636 1455 Q535.715 1459.58 535.715 1468.33 Q535.715 1477.06 532.636 1481.67 Q529.581 1486.25 523.771 1486.25 Q517.961 1486.25 514.882 1481.67 Q511.826 1477.06 511.826 1468.33 Q511.826 1459.58 514.882 1455 Q517.961 1450.39 523.771 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M543.979 1451.02 L562.335 1451.02 L562.335 1454.96 L548.261 1454.96 L548.261 1463.43 Q549.28 1463.08 550.298 1462.92 Q551.317 1462.73 552.335 1462.73 Q558.122 1462.73 561.502 1465.9 Q564.882 1469.08 564.882 1474.49 Q564.882 1480.07 561.409 1483.17 Q557.937 1486.25 551.618 1486.25 Q549.442 1486.25 547.173 1485.88 Q544.928 1485.51 542.521 1484.77 L542.521 1480.07 Q544.604 1481.2 546.826 1481.76 Q549.048 1482.32 551.525 1482.32 Q555.53 1482.32 557.868 1480.21 Q560.206 1478.1 560.206 1474.49 Q560.206 1470.88 557.868 1468.77 Q555.53 1466.67 551.525 1466.67 Q549.65 1466.67 547.775 1467.08 Q545.923 1467.5 543.979 1468.38 L543.979 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M691.515 1454.1 Q687.904 1454.1 686.075 1457.66 Q684.27 1461.2 684.27 1468.33 Q684.27 1475.44 686.075 1479.01 Q687.904 1482.55 691.515 1482.55 Q695.149 1482.55 696.955 1479.01 Q698.783 1475.44 698.783 1468.33 Q698.783 1461.2 696.955 1457.66 Q695.149 1454.1 691.515 1454.1 M691.515 1450.39 Q697.325 1450.39 700.381 1455 Q703.459 1459.58 703.459 1468.33 Q703.459 1477.06 700.381 1481.67 Q697.325 1486.25 691.515 1486.25 Q685.705 1486.25 682.626 1481.67 Q679.571 1477.06 679.571 1468.33 Q679.571 1459.58 682.626 1455 Q685.705 1450.39 691.515 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M711.677 1479.7 L716.561 1479.7 L716.561 1485.58 L711.677 1485.58 L711.677 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M727.556 1481.64 L735.195 1481.64 L735.195 1455.28 L726.885 1456.95 L726.885 1452.69 L735.149 1451.02 L739.825 1451.02 L739.825 1481.64 L747.464 1481.64 L747.464 1485.58 L727.556 1485.58 L727.556 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M766.908 1454.1 Q763.297 1454.1 761.468 1457.66 Q759.663 1461.2 759.663 1468.33 Q759.663 1475.44 761.468 1479.01 Q763.297 1482.55 766.908 1482.55 Q770.542 1482.55 772.348 1479.01 Q774.176 1475.44 774.176 1468.33 Q774.176 1461.2 772.348 1457.66 Q770.542 1454.1 766.908 1454.1 M766.908 1450.39 Q772.718 1450.39 775.774 1455 Q778.852 1459.58 778.852 1468.33 Q778.852 1477.06 775.774 1481.67 Q772.718 1486.25 766.908 1486.25 Q761.098 1486.25 758.019 1481.67 Q754.964 1477.06 754.964 1468.33 Q754.964 1459.58 758.019 1455 Q761.098 1450.39 766.908 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M905.486 1454.1 Q901.875 1454.1 900.046 1457.66 Q898.24 1461.2 898.24 1468.33 Q898.24 1475.44 900.046 1479.01 Q901.875 1482.55 905.486 1482.55 Q909.12 1482.55 910.925 1479.01 Q912.754 1475.44 912.754 1468.33 Q912.754 1461.2 910.925 1457.66 Q909.12 1454.1 905.486 1454.1 M905.486 1450.39 Q911.296 1450.39 914.351 1455 Q917.43 1459.58 917.43 1468.33 Q917.43 1477.06 914.351 1481.67 Q911.296 1486.25 905.486 1486.25 Q899.676 1486.25 896.597 1481.67 Q893.541 1477.06 893.541 1468.33 Q893.541 1459.58 896.597 1455 Q899.676 1450.39 905.486 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M925.648 1479.7 L930.532 1479.7 L930.532 1485.58 L925.648 1485.58 L925.648 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M941.527 1481.64 L949.166 1481.64 L949.166 1455.28 L940.856 1456.95 L940.856 1452.69 L949.12 1451.02 L953.796 1451.02 L953.796 1481.64 L961.434 1481.64 L961.434 1485.58 L941.527 1485.58 L941.527 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M970.925 1451.02 L989.282 1451.02 L989.282 1454.96 L975.208 1454.96 L975.208 1463.43 Q976.226 1463.08 977.245 1462.92 Q978.263 1462.73 979.282 1462.73 Q985.069 1462.73 988.448 1465.9 Q991.828 1469.08 991.828 1474.49 Q991.828 1480.07 988.356 1483.17 Q984.883 1486.25 978.564 1486.25 Q976.388 1486.25 974.12 1485.88 Q971.874 1485.51 969.467 1484.77 L969.467 1480.07 Q971.55 1481.2 973.772 1481.76 Q975.995 1482.32 978.471 1482.32 Q982.476 1482.32 984.814 1480.21 Q987.152 1478.1 987.152 1474.49 Q987.152 1470.88 984.814 1468.77 Q982.476 1466.67 978.471 1466.67 Q976.596 1466.67 974.721 1467.08 Q972.87 1467.5 970.925 1468.38 L970.925 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1118.46 1454.1 Q1114.85 1454.1 1113.02 1457.66 Q1111.22 1461.2 1111.22 1468.33 Q1111.22 1475.44 1113.02 1479.01 Q1114.85 1482.55 1118.46 1482.55 Q1122.1 1482.55 1123.9 1479.01 Q1125.73 1475.44 1125.73 1468.33 Q1125.73 1461.2 1123.9 1457.66 Q1122.1 1454.1 1118.46 1454.1 M1118.46 1450.39 Q1124.27 1450.39 1127.33 1455 Q1130.41 1459.58 1130.41 1468.33 Q1130.41 1477.06 1127.33 1481.67 Q1124.27 1486.25 1118.46 1486.25 Q1112.65 1486.25 1109.57 1481.67 Q1106.52 1477.06 1106.52 1468.33 Q1106.52 1459.58 1109.57 1455 Q1112.65 1450.39 1118.46 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1138.62 1479.7 L1143.51 1479.7 L1143.51 1485.58 L1138.62 1485.58 L1138.62 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1157.72 1481.64 L1174.04 1481.64 L1174.04 1485.58 L1152.1 1485.58 L1152.1 1481.64 Q1154.76 1478.89 1159.34 1474.26 Q1163.95 1469.61 1165.13 1468.27 Q1167.37 1465.74 1168.25 1464.01 Q1169.16 1462.25 1169.16 1460.56 Q1169.16 1457.8 1167.21 1456.07 Q1165.29 1454.33 1162.19 1454.33 Q1159.99 1454.33 1157.53 1455.09 Q1155.1 1455.86 1152.33 1457.41 L1152.33 1452.69 Q1155.15 1451.55 1157.6 1450.97 Q1160.06 1450.39 1162.1 1450.39 Q1167.47 1450.39 1170.66 1453.08 Q1173.85 1455.77 1173.85 1460.26 Q1173.85 1462.39 1173.04 1464.31 Q1172.26 1466.2 1170.15 1468.8 Q1169.57 1469.47 1166.47 1472.69 Q1163.37 1475.88 1157.72 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1193.85 1454.1 Q1190.24 1454.1 1188.41 1457.66 Q1186.61 1461.2 1186.61 1468.33 Q1186.61 1475.44 1188.41 1479.01 Q1190.24 1482.55 1193.85 1482.55 Q1197.49 1482.55 1199.29 1479.01 Q1201.12 1475.44 1201.12 1468.33 Q1201.12 1461.2 1199.29 1457.66 Q1197.49 1454.1 1193.85 1454.1 M1193.85 1450.39 Q1199.66 1450.39 1202.72 1455 Q1205.8 1459.58 1205.8 1468.33 Q1205.8 1477.06 1202.72 1481.67 Q1199.66 1486.25 1193.85 1486.25 Q1188.04 1486.25 1184.97 1481.67 Q1181.91 1477.06 1181.91 1468.33 Q1181.91 1459.58 1184.97 1455 Q1188.04 1450.39 1193.85 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M599.81 1520.52 L640.009 1520.52 L640.009 1525.93 L623.14 1525.93 L623.14 1568.04 L616.679 1568.04 L616.679 1525.93 L599.81 1525.93 L599.81 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M643.956 1532.4 L649.812 1532.4 L649.812 1568.04 L643.956 1568.04 L643.956 1532.4 M643.956 1518.52 L649.812 1518.52 L649.812 1525.93 L643.956 1525.93 L643.956 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M689.821 1539.24 Q692.017 1535.29 695.073 1533.41 Q698.128 1531.54 702.266 1531.54 Q707.836 1531.54 710.86 1535.45 Q713.883 1539.33 713.883 1546.53 L713.883 1568.04 L707.995 1568.04 L707.995 1546.72 Q707.995 1541.59 706.181 1539.11 Q704.367 1536.63 700.643 1536.63 Q696.091 1536.63 693.449 1539.65 Q690.808 1542.68 690.808 1547.9 L690.808 1568.04 L684.919 1568.04 L684.919 1546.72 Q684.919 1541.56 683.105 1539.11 Q681.291 1536.63 677.503 1536.63 Q673.015 1536.63 670.374 1539.68 Q667.732 1542.71 667.732 1547.9 L667.732 1568.04 L661.844 1568.04 L661.844 1532.4 L667.732 1532.4 L667.732 1537.93 Q669.737 1534.66 672.538 1533.1 Q675.339 1531.54 679.19 1531.54 Q683.073 1531.54 685.779 1533.51 Q688.516 1535.48 689.821 1539.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M756.056 1548.76 L756.056 1551.62 L729.129 1551.62 Q729.511 1557.67 732.758 1560.85 Q736.036 1564 741.861 1564 Q745.234 1564 748.385 1563.17 Q751.568 1562.35 754.687 1560.69 L754.687 1566.23 Q751.536 1567.57 748.226 1568.27 Q744.916 1568.97 741.51 1568.97 Q732.98 1568.97 727.983 1564 Q723.018 1559.04 723.018 1550.57 Q723.018 1541.82 727.729 1536.69 Q732.471 1531.54 740.492 1531.54 Q747.685 1531.54 751.855 1536.18 Q756.056 1540.8 756.056 1548.76 M750.2 1547.04 Q750.136 1542.23 747.494 1539.37 Q744.884 1536.5 740.556 1536.5 Q735.654 1536.5 732.694 1539.27 Q729.766 1542.04 729.32 1547.07 L750.2 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M800.457 1518.58 Q796.192 1525.9 794.123 1533.06 Q792.054 1540.23 792.054 1547.58 Q792.054 1554.93 794.123 1562.16 Q796.224 1569.35 800.457 1576.64 L795.364 1576.64 Q790.59 1569.16 788.203 1561.93 Q785.848 1554.71 785.848 1547.58 Q785.848 1540.48 788.203 1533.29 Q790.558 1526.09 795.364 1518.58 L800.457 1518.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M834.545 1533.45 L834.545 1538.98 Q832.063 1537.71 829.389 1537.07 Q826.715 1536.44 823.851 1536.44 Q819.49 1536.44 817.294 1537.77 Q815.13 1539.11 815.13 1541.79 Q815.13 1543.82 816.689 1545 Q818.249 1546.15 822.96 1547.2 L824.965 1547.64 Q831.203 1548.98 833.813 1551.43 Q836.455 1553.85 836.455 1558.21 Q836.455 1563.17 832.508 1566.07 Q828.593 1568.97 821.718 1568.97 Q818.854 1568.97 815.735 1568.39 Q812.647 1567.85 809.21 1566.74 L809.21 1560.69 Q812.456 1562.38 815.607 1563.24 Q818.758 1564.07 821.846 1564.07 Q825.983 1564.07 828.211 1562.66 Q830.439 1561.23 830.439 1558.65 Q830.439 1556.27 828.816 1554.99 Q827.225 1553.72 821.782 1552.54 L819.745 1552.07 Q814.302 1550.92 811.883 1548.56 Q809.464 1546.18 809.464 1542.04 Q809.464 1537.01 813.029 1534.27 Q816.594 1531.54 823.151 1531.54 Q826.397 1531.54 829.262 1532.01 Q832.126 1532.49 834.545 1533.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M844.858 1518.58 L849.95 1518.58 Q854.725 1526.09 857.08 1533.29 Q859.467 1540.48 859.467 1547.58 Q859.467 1554.71 857.08 1561.93 Q854.725 1569.16 849.95 1576.64 L844.858 1576.64 Q849.091 1569.35 851.16 1562.16 Q853.26 1554.93 853.26 1547.58 Q853.26 1540.23 851.16 1533.06 Q849.091 1525.9 844.858 1518.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,1406.88 1181.75,1406.88 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,1322.61 1181.75,1322.61 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,1238.34 1181.75,1238.34 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,1154.07 1181.75,1154.07 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,1069.8 1181.75,1069.8 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,985.533 1181.75,985.533 "/>
+<polyline clip-path="url(#clip574)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="277.528,901.264 1181.75,901.264 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,1423.18 277.528,847.244 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,1406.88 296.426,1406.88 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,1322.61 296.426,1322.61 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,1238.34 296.426,1238.34 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,1154.07 296.426,1154.07 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,1069.8 296.426,1069.8 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,985.533 296.426,985.533 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="277.528,901.264 296.426,901.264 "/>
+<path clip-path="url(#clip570)" d="M154.19 1392.68 Q150.579 1392.68 148.751 1396.24 Q146.945 1399.78 146.945 1406.91 Q146.945 1414.02 148.751 1417.59 Q150.579 1421.13 154.19 1421.13 Q157.825 1421.13 159.63 1417.59 Q161.459 1414.02 161.459 1406.91 Q161.459 1399.78 159.63 1396.24 Q157.825 1392.68 154.19 1392.68 M154.19 1388.97 Q160.001 1388.97 163.056 1393.58 Q166.135 1398.16 166.135 1406.91 Q166.135 1415.64 163.056 1420.25 Q160.001 1424.83 154.19 1424.83 Q148.38 1424.83 145.302 1420.25 Q142.246 1415.64 142.246 1406.91 Q142.246 1398.16 145.302 1393.58 Q148.38 1388.97 154.19 1388.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M174.352 1418.28 L179.237 1418.28 L179.237 1424.16 L174.352 1424.16 L174.352 1418.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M199.422 1392.68 Q195.811 1392.68 193.982 1396.24 Q192.176 1399.78 192.176 1406.91 Q192.176 1414.02 193.982 1417.59 Q195.811 1421.13 199.422 1421.13 Q203.056 1421.13 204.862 1417.59 Q206.69 1414.02 206.69 1406.91 Q206.69 1399.78 204.862 1396.24 Q203.056 1392.68 199.422 1392.68 M199.422 1388.97 Q205.232 1388.97 208.287 1393.58 Q211.366 1398.16 211.366 1406.91 Q211.366 1415.64 208.287 1420.25 Q205.232 1424.83 199.422 1424.83 Q193.612 1424.83 190.533 1420.25 Q187.477 1415.64 187.477 1406.91 Q187.477 1398.16 190.533 1393.58 Q193.612 1388.97 199.422 1388.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M229.584 1392.68 Q225.973 1392.68 224.144 1396.24 Q222.338 1399.78 222.338 1406.91 Q222.338 1414.02 224.144 1417.59 Q225.973 1421.13 229.584 1421.13 Q233.218 1421.13 235.023 1417.59 Q236.852 1414.02 236.852 1406.91 Q236.852 1399.78 235.023 1396.24 Q233.218 1392.68 229.584 1392.68 M229.584 1388.97 Q235.394 1388.97 238.449 1393.58 Q241.528 1398.16 241.528 1406.91 Q241.528 1415.64 238.449 1420.25 Q235.394 1424.83 229.584 1424.83 Q223.773 1424.83 220.695 1420.25 Q217.639 1415.64 217.639 1406.91 Q217.639 1398.16 220.695 1393.58 Q223.773 1388.97 229.584 1388.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M155.788 1308.41 Q152.177 1308.41 150.348 1311.97 Q148.542 1315.52 148.542 1322.65 Q148.542 1329.75 150.348 1333.32 Q152.177 1336.86 155.788 1336.86 Q159.422 1336.86 161.227 1333.32 Q163.056 1329.75 163.056 1322.65 Q163.056 1315.52 161.227 1311.97 Q159.422 1308.41 155.788 1308.41 M155.788 1304.71 Q161.598 1304.71 164.653 1309.31 Q167.732 1313.9 167.732 1322.65 Q167.732 1331.37 164.653 1335.98 Q161.598 1340.56 155.788 1340.56 Q149.978 1340.56 146.899 1335.98 Q143.843 1331.37 143.843 1322.65 Q143.843 1313.9 146.899 1309.31 Q149.978 1304.71 155.788 1304.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M175.95 1334.01 L180.834 1334.01 L180.834 1339.89 L175.95 1339.89 L175.95 1334.01 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M201.019 1308.41 Q197.408 1308.41 195.579 1311.97 Q193.774 1315.52 193.774 1322.65 Q193.774 1329.75 195.579 1333.32 Q197.408 1336.86 201.019 1336.86 Q204.653 1336.86 206.459 1333.32 Q208.287 1329.75 208.287 1322.65 Q208.287 1315.52 206.459 1311.97 Q204.653 1308.41 201.019 1308.41 M201.019 1304.71 Q206.829 1304.71 209.885 1309.31 Q212.963 1313.9 212.963 1322.65 Q212.963 1331.37 209.885 1335.98 Q206.829 1340.56 201.019 1340.56 Q195.209 1340.56 192.13 1335.98 Q189.075 1331.37 189.075 1322.65 Q189.075 1313.9 192.13 1309.31 Q195.209 1304.71 201.019 1304.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M225.209 1335.96 L241.528 1335.96 L241.528 1339.89 L219.584 1339.89 L219.584 1335.96 Q222.246 1333.2 226.829 1328.57 Q231.435 1323.92 232.616 1322.58 Q234.861 1320.05 235.741 1318.32 Q236.644 1316.56 236.644 1314.87 Q236.644 1312.11 234.699 1310.38 Q232.778 1308.64 229.676 1308.64 Q227.477 1308.64 225.023 1309.4 Q222.593 1310.17 219.815 1311.72 L219.815 1307 Q222.639 1305.86 225.093 1305.28 Q227.547 1304.71 229.584 1304.71 Q234.954 1304.71 238.148 1307.39 Q241.343 1310.08 241.343 1314.57 Q241.343 1316.7 240.533 1318.62 Q239.746 1320.52 237.639 1323.11 Q237.06 1323.78 233.959 1327 Q230.857 1330.19 225.209 1335.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M153.704 1224.14 Q150.093 1224.14 148.265 1227.7 Q146.459 1231.25 146.459 1238.38 Q146.459 1245.48 148.265 1249.05 Q150.093 1252.59 153.704 1252.59 Q157.339 1252.59 159.144 1249.05 Q160.973 1245.48 160.973 1238.38 Q160.973 1231.25 159.144 1227.7 Q157.339 1224.14 153.704 1224.14 M153.704 1220.44 Q159.515 1220.44 162.57 1225.04 Q165.649 1229.63 165.649 1238.38 Q165.649 1247.1 162.57 1251.71 Q159.515 1256.29 153.704 1256.29 Q147.894 1256.29 144.816 1251.71 Q141.76 1247.1 141.76 1238.38 Q141.76 1229.63 144.816 1225.04 Q147.894 1220.44 153.704 1220.44 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M173.866 1249.74 L178.751 1249.74 L178.751 1255.62 L173.866 1255.62 L173.866 1249.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M198.936 1224.14 Q195.325 1224.14 193.496 1227.7 Q191.69 1231.25 191.69 1238.38 Q191.69 1245.48 193.496 1249.05 Q195.325 1252.59 198.936 1252.59 Q202.57 1252.59 204.375 1249.05 Q206.204 1245.48 206.204 1238.38 Q206.204 1231.25 204.375 1227.7 Q202.57 1224.14 198.936 1224.14 M198.936 1220.44 Q204.746 1220.44 207.801 1225.04 Q210.88 1229.63 210.88 1238.38 Q210.88 1247.1 207.801 1251.71 Q204.746 1256.29 198.936 1256.29 Q193.125 1256.29 190.047 1251.71 Q186.991 1247.1 186.991 1238.38 Q186.991 1229.63 190.047 1225.04 Q193.125 1220.44 198.936 1220.44 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M231.945 1225.14 L220.139 1243.58 L231.945 1243.58 L231.945 1225.14 M230.718 1221.06 L236.597 1221.06 L236.597 1243.58 L241.528 1243.58 L241.528 1247.47 L236.597 1247.47 L236.597 1255.62 L231.945 1255.62 L231.945 1247.47 L216.343 1247.47 L216.343 1242.96 L230.718 1221.06 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M154.028 1139.87 Q150.417 1139.87 148.589 1143.44 Q146.783 1146.98 146.783 1154.11 Q146.783 1161.21 148.589 1164.78 Q150.417 1168.32 154.028 1168.32 Q157.663 1168.32 159.468 1164.78 Q161.297 1161.21 161.297 1154.11 Q161.297 1146.98 159.468 1143.44 Q157.663 1139.87 154.028 1139.87 M154.028 1136.17 Q159.839 1136.17 162.894 1140.77 Q165.973 1145.36 165.973 1154.11 Q165.973 1162.83 162.894 1167.44 Q159.839 1172.02 154.028 1172.02 Q148.218 1172.02 145.14 1167.44 Q142.084 1162.83 142.084 1154.11 Q142.084 1145.36 145.14 1140.77 Q148.218 1136.17 154.028 1136.17 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M174.19 1165.47 L179.075 1165.47 L179.075 1171.35 L174.19 1171.35 L174.19 1165.47 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M199.26 1139.87 Q195.649 1139.87 193.82 1143.44 Q192.014 1146.98 192.014 1154.11 Q192.014 1161.21 193.82 1164.78 Q195.649 1168.32 199.26 1168.32 Q202.894 1168.32 204.699 1164.78 Q206.528 1161.21 206.528 1154.11 Q206.528 1146.98 204.699 1143.44 Q202.894 1139.87 199.26 1139.87 M199.26 1136.17 Q205.07 1136.17 208.125 1140.77 Q211.204 1145.36 211.204 1154.11 Q211.204 1162.83 208.125 1167.44 Q205.07 1172.02 199.26 1172.02 Q193.45 1172.02 190.371 1167.44 Q187.315 1162.83 187.315 1154.11 Q187.315 1145.36 190.371 1140.77 Q193.45 1136.17 199.26 1136.17 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M230 1152.21 Q226.852 1152.21 225 1154.36 Q223.172 1156.51 223.172 1160.26 Q223.172 1163.99 225 1166.17 Q226.852 1168.32 230 1168.32 Q233.148 1168.32 234.977 1166.17 Q236.829 1163.99 236.829 1160.26 Q236.829 1156.51 234.977 1154.36 Q233.148 1152.21 230 1152.21 M239.283 1137.56 L239.283 1141.82 Q237.523 1140.98 235.718 1140.54 Q233.935 1140.1 232.176 1140.1 Q227.547 1140.1 225.093 1143.23 Q222.662 1146.35 222.315 1152.67 Q223.681 1150.66 225.741 1149.59 Q227.801 1148.5 230.278 1148.5 Q235.486 1148.5 238.496 1151.68 Q241.528 1154.82 241.528 1160.26 Q241.528 1165.59 238.38 1168.81 Q235.232 1172.02 230 1172.02 Q224.005 1172.02 220.834 1167.44 Q217.662 1162.83 217.662 1154.11 Q217.662 1145.91 221.551 1141.05 Q225.44 1136.17 231.991 1136.17 Q233.75 1136.17 235.533 1136.51 Q237.338 1136.86 239.283 1137.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M154.283 1055.6 Q150.672 1055.6 148.843 1059.17 Q147.038 1062.71 147.038 1069.84 Q147.038 1076.94 148.843 1080.51 Q150.672 1084.05 154.283 1084.05 Q157.917 1084.05 159.723 1080.51 Q161.552 1076.94 161.552 1069.84 Q161.552 1062.71 159.723 1059.17 Q157.917 1055.6 154.283 1055.6 M154.283 1051.9 Q160.093 1051.9 163.149 1056.5 Q166.227 1061.09 166.227 1069.84 Q166.227 1078.56 163.149 1083.17 Q160.093 1087.75 154.283 1087.75 Q148.473 1087.75 145.394 1083.17 Q142.339 1078.56 142.339 1069.84 Q142.339 1061.09 145.394 1056.5 Q148.473 1051.9 154.283 1051.9 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M174.445 1081.2 L179.329 1081.2 L179.329 1087.08 L174.445 1087.08 L174.445 1081.2 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M199.514 1055.6 Q195.903 1055.6 194.075 1059.17 Q192.269 1062.71 192.269 1069.84 Q192.269 1076.94 194.075 1080.51 Q195.903 1084.05 199.514 1084.05 Q203.149 1084.05 204.954 1080.51 Q206.783 1076.94 206.783 1069.84 Q206.783 1062.71 204.954 1059.17 Q203.149 1055.6 199.514 1055.6 M199.514 1051.9 Q205.324 1051.9 208.38 1056.5 Q211.459 1061.09 211.459 1069.84 Q211.459 1078.56 208.38 1083.17 Q205.324 1087.75 199.514 1087.75 Q193.704 1087.75 190.625 1083.17 Q187.57 1078.56 187.57 1069.84 Q187.57 1061.09 190.625 1056.5 Q193.704 1051.9 199.514 1051.9 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M229.676 1070.67 Q226.343 1070.67 224.422 1072.45 Q222.523 1074.24 222.523 1077.36 Q222.523 1080.49 224.422 1082.27 Q226.343 1084.05 229.676 1084.05 Q233.01 1084.05 234.931 1082.27 Q236.852 1080.46 236.852 1077.36 Q236.852 1074.24 234.931 1072.45 Q233.033 1070.67 229.676 1070.67 M225 1068.68 Q221.991 1067.94 220.301 1065.88 Q218.635 1063.82 218.635 1060.86 Q218.635 1056.71 221.574 1054.31 Q224.537 1051.9 229.676 1051.9 Q234.838 1051.9 237.778 1054.31 Q240.718 1056.71 240.718 1060.86 Q240.718 1063.82 239.028 1065.88 Q237.361 1067.94 234.375 1068.68 Q237.755 1069.47 239.63 1071.76 Q241.528 1074.05 241.528 1077.36 Q241.528 1082.38 238.449 1085.07 Q235.394 1087.75 229.676 1087.75 Q223.959 1087.75 220.88 1085.07 Q217.824 1082.38 217.824 1077.36 Q217.824 1074.05 219.723 1071.76 Q221.621 1069.47 225 1068.68 M223.287 1061.3 Q223.287 1063.98 224.954 1065.49 Q226.644 1066.99 229.676 1066.99 Q232.685 1066.99 234.375 1065.49 Q236.088 1063.98 236.088 1061.3 Q236.088 1058.61 234.375 1057.11 Q232.685 1055.6 229.676 1055.6 Q226.644 1055.6 224.954 1057.11 Q223.287 1058.61 223.287 1061.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M154.19 971.332 Q150.579 971.332 148.751 974.897 Q146.945 978.439 146.945 985.568 Q146.945 992.675 148.751 996.239 Q150.579 999.781 154.19 999.781 Q157.825 999.781 159.63 996.239 Q161.459 992.675 161.459 985.568 Q161.459 978.439 159.63 974.897 Q157.825 971.332 154.19 971.332 M154.19 967.628 Q160.001 967.628 163.056 972.235 Q166.135 976.818 166.135 985.568 Q166.135 994.295 163.056 998.901 Q160.001 1003.48 154.19 1003.48 Q148.38 1003.48 145.302 998.901 Q142.246 994.295 142.246 985.568 Q142.246 976.818 145.302 972.235 Q148.38 967.628 154.19 967.628 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M174.352 996.934 L179.237 996.934 L179.237 1002.81 L174.352 1002.81 L174.352 996.934 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M190.232 998.878 L197.871 998.878 L197.871 972.513 L189.561 974.179 L189.561 969.92 L197.825 968.253 L202.5 968.253 L202.5 998.878 L210.139 998.878 L210.139 1002.81 L190.232 1002.81 L190.232 998.878 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M229.584 971.332 Q225.973 971.332 224.144 974.897 Q222.338 978.439 222.338 985.568 Q222.338 992.675 224.144 996.239 Q225.973 999.781 229.584 999.781 Q233.218 999.781 235.023 996.239 Q236.852 992.675 236.852 985.568 Q236.852 978.439 235.023 974.897 Q233.218 971.332 229.584 971.332 M229.584 967.628 Q235.394 967.628 238.449 972.235 Q241.528 976.818 241.528 985.568 Q241.528 994.295 238.449 998.901 Q235.394 1003.48 229.584 1003.48 Q223.773 1003.48 220.695 998.901 Q217.639 994.295 217.639 985.568 Q217.639 976.818 220.695 972.235 Q223.773 967.628 229.584 967.628 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M155.788 887.063 Q152.177 887.063 150.348 890.628 Q148.542 894.169 148.542 901.299 Q148.542 908.405 150.348 911.97 Q152.177 915.512 155.788 915.512 Q159.422 915.512 161.227 911.97 Q163.056 908.405 163.056 901.299 Q163.056 894.169 161.227 890.628 Q159.422 887.063 155.788 887.063 M155.788 883.359 Q161.598 883.359 164.653 887.966 Q167.732 892.549 167.732 901.299 Q167.732 910.026 164.653 914.632 Q161.598 919.215 155.788 919.215 Q149.978 919.215 146.899 914.632 Q143.843 910.026 143.843 901.299 Q143.843 892.549 146.899 887.966 Q149.978 883.359 155.788 883.359 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M175.95 912.665 L180.834 912.665 L180.834 918.544 L175.95 918.544 L175.95 912.665 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M191.829 914.609 L199.468 914.609 L199.468 888.243 L191.158 889.91 L191.158 885.651 L199.422 883.984 L204.098 883.984 L204.098 914.609 L211.736 914.609 L211.736 918.544 L191.829 918.544 L191.829 914.609 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M225.209 914.609 L241.528 914.609 L241.528 918.544 L219.584 918.544 L219.584 914.609 Q222.246 911.854 226.829 907.225 Q231.435 902.572 232.616 901.229 Q234.861 898.706 235.741 896.97 Q236.644 895.211 236.644 893.521 Q236.644 890.767 234.699 889.03 Q232.778 887.294 229.676 887.294 Q227.477 887.294 225.023 888.058 Q222.593 888.822 219.815 890.373 L219.815 885.651 Q222.639 884.517 225.093 883.938 Q227.547 883.359 229.584 883.359 Q234.954 883.359 238.148 886.044 Q241.343 888.73 241.343 893.22 Q241.343 895.35 240.533 897.271 Q239.746 899.169 237.639 901.762 Q237.06 902.433 233.959 905.651 Q230.857 908.845 225.209 914.609 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M91.504 1243.7 L43.984 1261.84 L43.984 1255.13 L83.9925 1240.07 L43.984 1224.98 L43.984 1218.3 L91.504 1236.41 L91.504 1243.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M72.2159 1186.19 L75.0805 1186.19 L75.0805 1213.11 Q81.1279 1212.73 84.3108 1209.48 Q87.4618 1206.21 87.4618 1200.38 Q87.4618 1197.01 86.6342 1193.86 Q85.8067 1190.67 84.1516 1187.55 L89.6898 1187.55 Q91.0266 1190.71 91.7268 1194.02 Q92.427 1197.33 92.427 1200.73 Q92.427 1209.26 87.4618 1214.26 Q82.4965 1219.22 74.0301 1219.22 Q65.2773 1219.22 60.1529 1214.51 Q54.9967 1209.77 54.9967 1201.75 Q54.9967 1194.56 59.6436 1190.39 Q64.2588 1186.19 72.2159 1186.19 M70.4972 1192.04 Q65.6911 1192.11 62.8265 1194.75 Q59.9619 1197.36 59.9619 1201.69 Q59.9619 1206.59 62.731 1209.55 Q65.5001 1212.48 70.529 1212.92 L70.4972 1192.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M41.9788 1176.57 L41.9788 1170.72 L91.504 1170.72 L91.504 1176.57 L41.9788 1176.57 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M59.9619 1144.65 Q59.9619 1149.36 63.654 1152.1 Q67.3143 1154.83 73.7119 1154.83 Q80.1094 1154.83 83.8015 1152.13 Q87.4618 1149.39 87.4618 1144.65 Q87.4618 1139.97 83.7697 1137.23 Q80.0776 1134.5 73.7119 1134.5 Q67.378 1134.5 63.6859 1137.23 Q59.9619 1139.97 59.9619 1144.65 M54.9967 1144.65 Q54.9967 1137.01 59.9619 1132.65 Q64.9272 1128.29 73.7119 1128.29 Q82.4647 1128.29 87.4618 1132.65 Q92.427 1137.01 92.427 1144.65 Q92.427 1152.32 87.4618 1156.68 Q82.4647 1161.01 73.7119 1161.01 Q64.9272 1161.01 59.9619 1156.68 Q54.9967 1152.32 54.9967 1144.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M57.2247 1092.93 L62.6992 1092.93 Q61.3306 1095.41 60.6622 1097.92 Q59.9619 1100.41 59.9619 1102.95 Q59.9619 1108.65 63.5904 1111.8 Q67.187 1114.95 73.7119 1114.95 Q80.2367 1114.95 83.8652 1111.8 Q87.4618 1108.65 87.4618 1102.95 Q87.4618 1100.41 86.7934 1097.92 Q86.0932 1095.41 84.7245 1092.93 L90.1354 1092.93 Q91.2812 1095.38 91.8541 1098.02 Q92.427 1100.63 92.427 1103.59 Q92.427 1111.64 87.3663 1116.39 Q82.3056 1121.13 73.7119 1121.13 Q64.9908 1121.13 59.9938 1116.35 Q54.9967 1111.55 54.9967 1103.21 Q54.9967 1100.5 55.5696 1097.92 Q56.1107 1095.35 57.2247 1092.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M55.8561 1082.74 L55.8561 1076.89 L91.504 1076.89 L91.504 1082.74 L55.8561 1082.74 M41.9788 1082.74 L41.9788 1076.89 L49.3949 1076.89 L49.3949 1082.74 L41.9788 1082.74 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M45.7346 1058.84 L55.8561 1058.84 L55.8561 1046.78 L60.4075 1046.78 L60.4075 1058.84 L79.7593 1058.84 Q84.1198 1058.84 85.3611 1057.66 Q86.6024 1056.45 86.6024 1052.79 L86.6024 1046.78 L91.504 1046.78 L91.504 1052.79 Q91.504 1059.57 88.9896 1062.15 Q86.4433 1064.73 79.7593 1064.73 L60.4075 1064.73 L60.4075 1069.02 L55.8561 1069.02 L55.8561 1064.73 L45.7346 1064.73 L45.7346 1058.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M94.8142 1024.24 Q101.18 1026.72 103.121 1029.08 Q105.063 1031.44 105.063 1035.38 L105.063 1040.06 L100.161 1040.06 L100.161 1036.62 Q100.161 1034.2 99.0155 1032.87 Q97.8697 1031.53 93.6047 1029.91 L90.9311 1028.86 L55.8561 1043.28 L55.8561 1037.07 L83.7378 1025.93 L55.8561 1014.79 L55.8561 1008.58 L94.8142 1024.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip574)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="303.119,1406.88 303.973,1404.77 304.827,1402.66 305.681,1400.55 306.535,1398.43 307.389,1396.31 308.243,1394.18 309.096,1392.05 309.95,1389.92 310.804,1387.79 311.658,1385.65 312.512,1383.5 313.366,1381.36 314.22,1379.21 315.074,1377.06 315.928,1374.91 316.781,1372.76 317.635,1370.6 318.489,1368.44 319.343,1366.28 320.197,1364.12 321.051,1361.95 321.905,1359.79 322.759,1357.62 323.613,1355.45 324.466,1353.29 325.32,1351.11 326.174,1348.94 327.028,1346.77 327.882,1344.6 328.736,1342.43 329.59,1340.25 330.444,1338.08 331.298,1335.91 332.151,1333.73 333.005,1331.56 333.859,1329.39 334.713,1327.21 335.567,1325.04 336.421,1322.87 337.275,1320.7 338.129,1318.53 338.983,1316.36 339.837,1314.19 340.69,1312.03 341.544,1309.86 342.398,1307.7 343.252,1305.53 344.106,1303.37 344.96,1301.21 345.814,1299.06 346.668,1296.9 347.522,1294.75 348.375,1292.6 349.229,1290.45 350.083,1288.3 350.937,1286.16 351.791,1284.02 352.645,1281.88 353.499,1279.74 354.353,1277.61 355.207,1275.47 356.06,1273.35 356.914,1271.22 357.768,1269.1 358.622,1266.98 359.476,1264.86 360.33,1262.75 361.184,1260.64 362.038,1258.54 362.892,1256.43 363.746,1254.33 364.599,1252.24 365.453,1250.15 366.307,1248.06 367.161,1245.97 368.015,1243.89 368.869,1241.81 369.723,1239.74 370.577,1237.67 371.431,1235.6 372.284,1233.54 373.138,1231.48 373.992,1229.43 374.846,1227.38 375.7,1225.33 376.554,1223.29 377.408,1221.25 378.262,1219.22 379.116,1217.19 379.969,1215.16 380.823,1213.14 381.677,1211.12 382.531,1209.11 383.385,1207.1 384.239,1205.1 385.093,1203.1 385.947,1201.11 386.801,1199.12 387.654,1197.14 388.508,1195.16 389.362,1193.19 390.216,1191.23 391.07,1189.28 391.924,1187.33 392.778,1185.4 393.632,1183.49 394.486,1181.6 395.34,1179.74 396.193,1177.94 397.047,1176.22 397.901,1174.63 398.755,1173.21 399.609,1171.94 400.463,1170.82 401.317,1169.81 402.171,1168.91 403.025,1168.08 403.878,1167.31 404.732,1166.59 405.586,1165.9 406.44,1165.25 407.294,1164.62 408.148,1164 409.002,1163.4 409.856,1162.8 410.71,1162.21 411.563,1161.63 412.417,1161.05 413.271,1160.48 414.125,1159.9 414.979,1159.32 415.833,1158.75 416.687,1158.17 417.541,1157.59 418.395,1157.01 419.249,1156.43 420.102,1155.85 420.956,1155.26 421.81,1154.67 422.664,1154.08 423.518,1153.49 424.372,1152.89 425.226,1152.29 426.08,1151.69 426.934,1151.09 427.787,1150.48 428.641,1149.87 429.495,1149.26 430.349,1148.64 431.203,1148.02 432.057,1147.4 432.911,1146.77 433.765,1146.14 434.619,1145.51 435.472,1144.88 436.326,1144.24 437.18,1143.6 438.034,1142.96 438.888,1142.31 439.742,1141.66 440.596,1141.01 441.45,1140.35 442.304,1139.69 443.158,1139.02 444.011,1138.36 444.865,1137.69 445.719,1137.01 446.573,1136.34 447.427,1135.65 448.281,1134.97 449.135,1134.28 449.989,1133.59 450.843,1132.9 451.696,1132.2 452.55,1131.49 453.404,1130.79 454.258,1130.08 455.112,1129.37 455.966,1128.65 456.82,1127.93 457.674,1127.2 458.528,1126.47 459.381,1125.74 460.235,1125.01 461.089,1124.27 461.943,1123.52 462.797,1122.77 463.651,1122.02 464.505,1121.27 465.359,1120.51 466.213,1119.74 467.066,1118.97 467.92,1118.2 468.774,1117.42 469.628,1116.64 470.482,1115.86 471.336,1115.07 472.19,1114.27 473.044,1113.47 473.898,1112.67 474.752,1111.86 475.605,1111.05 476.459,1110.24 477.313,1109.41 478.167,1108.59 479.021,1107.76 479.875,1106.92 480.729,1106.09 481.583,1105.24 482.437,1104.39 483.29,1103.54 484.144,1102.68 484.998,1101.82 485.852,1100.95 486.706,1100.07 487.56,1099.2 488.414,1098.31 489.268,1097.42 490.122,1096.53 490.975,1095.63 491.829,1094.73 492.683,1093.82 493.537,1092.9 494.391,1091.98 495.245,1091.05 496.099,1090.12 496.953,1089.19 497.807,1088.24 498.661,1087.3 499.514,1086.34 500.368,1085.38 501.222,1084.42 502.076,1083.45 502.93,1082.47 503.784,1081.49 504.638,1080.5 505.492,1079.5 506.346,1078.5 507.199,1077.5 508.053,1076.48 508.907,1075.47 509.761,1074.44 510.615,1073.41 511.469,1072.37 512.323,1071.32 513.177,1070.27 514.031,1069.22 514.884,1068.15 515.738,1067.08 516.592,1066 517.446,1064.92 518.3,1063.83 519.154,1062.73 520.008,1061.62 520.862,1060.51 521.716,1059.39 522.569,1058.26 523.423,1057.13 524.277,1055.99 525.131,1054.84 525.985,1053.68 526.839,1052.52 527.693,1051.34 528.547,1050.16 529.401,1048.98 530.255,1047.78 531.108,1046.58 531.962,1045.37 532.816,1044.15 533.67,1042.92 534.524,1041.69 535.378,1040.44 536.232,1039.19 537.086,1037.93 537.94,1036.66 538.793,1035.38 539.647,1034.1 540.501,1032.8 541.355,1031.5 542.209,1030.18 543.063,1028.86 543.917,1027.53 544.771,1026.19 545.625,1024.84 546.478,1023.48 547.332,1022.11 548.186,1020.73 549.04,1019.34 549.894,1017.95 550.748,1016.54 551.602,1015.12 552.456,1013.69 553.31,1012.25 554.164,1010.8 555.017,1009.34 555.871,1007.87 556.725,1006.39 557.579,1004.9 558.433,1003.4 559.287,1001.88 560.141,1000.36 560.995,998.823 561.849,997.276 562.702,995.717 563.556,994.147 564.41,992.565 565.264,990.971 566.118,989.365 566.972,987.747 567.826,986.117 568.68,984.475 569.534,982.82 570.387,981.152 571.241,979.472 572.095,977.779 572.949,976.074 573.803,974.355 574.657,972.623 575.511,970.877 576.365,969.119 577.219,967.346 578.072,965.561 578.926,963.761 579.78,961.948 580.634,960.121 581.488,958.28 582.342,956.425 583.196,954.556 584.05,952.673 584.904,950.776 585.758,948.865 586.611,946.939 587.465,945 588.319,943.046 589.173,941.079 590.027,939.098 590.881,937.103 591.735,935.094 592.589,933.072 593.443,931.038 594.296,928.99 595.15,926.931 596.004,924.859 596.858,922.776 597.712,920.682 598.566,918.579 599.42,916.466 600.274,914.345 601.128,912.217 601.981,910.082 602.835,907.943 603.689,905.801 604.543,903.657 605.397,901.514 606.251,899.374 607.105,897.239 607.959,895.112 608.813,892.997 609.667,890.896 610.52,888.814 611.374,886.756 612.228,884.726 613.082,882.731 613.936,880.776 614.79,878.869 615.644,877.018 616.498,875.232 617.352,873.52 618.205,871.894 619.059,870.367 619.913,868.95 620.767,867.66 621.621,866.511 622.475,865.522 623.329,864.712 624.183,864.098 625.037,863.703 625.89,863.544 626.744,863.64 627.598,864 628.452,864.622 629.306,865.479 630.16,866.509 631.014,867.645 631.868,868.837 632.722,870.06 633.576,871.3 634.429,872.548 635.283,873.799 636.137,875.052 636.991,876.303 637.845,877.552 638.699,878.798 639.553,880.039 640.407,881.276 641.261,882.509 642.114,883.736 642.968,884.958 643.822,886.174 644.676,887.385 645.53,888.591 646.384,889.791 647.238,890.986 648.092,892.176 648.946,893.36 649.799,894.538 650.653,895.711 651.507,896.879 652.361,898.042 653.215,899.2 654.069,900.352 654.923,901.5 655.777,902.642 656.631,903.78 657.484,904.913 658.338,906.041 659.192,907.164 660.046,908.283 660.9,909.397 661.754,910.507 662.608,911.613 663.462,912.714 664.316,913.811 665.17,914.903 666.023,915.992 666.877,917.076 667.731,918.157 668.585,919.233 669.439,920.306 670.293,921.375 671.147,922.44 672.001,923.502 672.855,924.56 673.708,925.614 674.562,926.665 675.416,927.713 676.27,928.757 677.124,929.798 677.978,930.835 678.832,931.87 679.686,932.901 680.54,933.929 681.393,934.954 682.247,935.976 683.101,936.995 683.955,938.011 684.809,939.025 685.663,940.035 686.517,941.043 687.371,942.048 688.225,943.05 689.079,944.05 689.932,945.047 690.786,946.042 691.64,947.034 692.494,948.023 693.348,949.011 694.202,949.996 695.056,950.978 695.91,951.958 696.764,952.936 697.617,953.912 698.471,954.885 699.325,955.857 700.179,956.826 701.033,957.793 701.887,958.758 702.741,959.721 703.595,960.682 704.449,961.641 705.302,962.598 706.156,963.553 707.01,964.507 707.864,965.458 708.718,966.408 709.572,967.356 710.426,968.302 711.28,969.247 712.134,970.189 712.987,971.13 713.841,972.07 714.695,973.008 715.549,973.944 716.403,974.878 717.257,975.811 718.111,976.743 718.965,977.673 719.819,978.602 720.673,979.529 721.526,980.454 722.38,981.379 723.234,982.302 724.088,983.223 724.942,984.143 725.796,985.062 726.65,985.98 727.504,986.896 728.358,987.811 729.211,988.724 730.065,989.637 730.919,990.548 731.773,991.458 732.627,992.367 733.481,993.275 734.335,994.182 735.189,995.087 736.043,995.991 736.896,996.895 737.75,997.797 738.604,998.698 739.458,999.598 740.312,1000.5 741.166,1001.4 742.02,1002.29 742.874,1003.19 743.728,1004.08 744.582,1004.98 745.435,1005.87 746.289,1006.76 747.143,1007.65 747.997,1008.55 748.851,1009.43 749.705,1010.32 750.559,1011.21 751.413,1012.1 752.267,1012.98 753.12,1013.87 753.974,1014.75 754.828,1015.64 755.682,1016.52 756.536,1017.4 757.39,1018.28 758.244,1019.17 759.098,1020.05 759.952,1020.92 760.805,1021.8 761.659,1022.68 762.513,1023.56 763.367,1024.43 764.221,1025.31 765.075,1026.18 765.929,1027.06 766.783,1027.93 767.637,1028.81 768.491,1029.68 769.344,1030.55 770.198,1031.42 771.052,1032.29 771.906,1033.16 772.76,1034.03 773.614,1034.9 774.468,1035.77 775.322,1036.63 776.176,1037.5 777.029,1038.37 777.883,1039.23 778.737,1040.1 779.591,1040.96 780.445,1041.83 781.299,1042.69 782.153,1043.56 783.007,1044.42 783.861,1045.28 784.714,1046.14 785.568,1047 786.422,1047.86 787.276,1048.72 788.13,1049.58 788.984,1050.44 789.838,1051.3 790.692,1052.16 791.546,1053.02 792.399,1053.88 793.253,1054.73 794.107,1055.59 794.961,1056.45 795.815,1057.3 796.669,1058.16 797.523,1059.01 798.377,1059.87 799.231,1060.72 800.085,1061.58 800.938,1062.43 801.792,1063.28 802.646,1064.14 803.5,1064.99 804.354,1065.84 805.208,1066.69 806.062,1067.54 806.916,1068.39 807.77,1069.25 808.623,1070.1 809.477,1070.95 810.331,1071.8 811.185,1072.65 812.039,1073.49 812.893,1074.34 813.747,1075.19 814.601,1076.04 815.455,1076.89 816.308,1077.73 817.162,1078.58 818.016,1079.43 818.87,1080.28 819.724,1081.12 820.578,1081.97 821.432,1082.81 822.286,1083.66 823.14,1084.5 823.994,1085.35 824.847,1086.19 825.701,1087.04 826.555,1087.88 827.409,1088.73 828.263,1089.57 829.117,1090.41 829.971,1091.26 830.825,1092.1 831.679,1092.94 832.532,1093.78 833.386,1094.63 834.24,1095.47 835.094,1096.31 835.948,1097.15 836.802,1097.99 837.656,1098.83 838.51,1099.68 839.364,1100.52 840.217,1101.36 841.071,1102.2 841.925,1103.04 842.779,1103.88 843.633,1104.72 844.487,1105.56 845.341,1106.39 846.195,1107.23 847.049,1108.07 847.902,1108.91 848.756,1109.75 849.61,1110.59 850.464,1111.43 851.318,1112.26 852.172,1113.1 853.026,1113.94 853.88,1114.78 854.734,1115.61 855.588,1116.45 856.441,1117.29 857.295,1118.12 858.149,1118.96 859.003,1119.8 859.857,1120.63 860.711,1121.47 861.565,1122.3 862.419,1123.14 863.273,1123.98 864.126,1124.81 864.98,1125.65 865.834,1126.48 866.688,1127.32 867.542,1128.15 868.396,1128.99 869.25,1129.82 870.104,1130.65 870.958,1131.49 871.811,1132.32 872.665,1133.16 873.519,1133.99 874.373,1134.82 875.227,1135.66 876.081,1136.49 876.935,1137.32 877.789,1138.16 878.643,1138.99 879.497,1139.82 880.35,1140.66 881.204,1141.49 882.058,1142.32 882.912,1143.15 883.766,1143.99 884.62,1144.82 885.474,1145.65 886.328,1146.48 887.182,1147.31 888.035,1148.15 888.889,1148.98 889.743,1149.81 890.597,1150.64 891.451,1151.47 892.305,1152.3 893.159,1153.13 894.013,1153.96 894.867,1154.79 895.72,1155.63 896.574,1156.46 897.428,1157.29 898.282,1158.12 899.136,1158.95 899.99,1159.78 900.844,1160.61 901.698,1161.44 902.552,1162.27 903.406,1163.1 904.259,1163.93 905.113,1164.76 905.967,1165.59 906.821,1166.42 907.675,1167.25 908.529,1168.08 909.383,1168.9 910.237,1169.73 911.091,1170.56 911.944,1171.39 912.798,1172.22 913.652,1173.05 914.506,1173.88 915.36,1174.71 916.214,1175.54 917.068,1176.36 917.922,1177.19 918.776,1178.02 919.629,1178.85 920.483,1179.68 921.337,1180.51 922.191,1181.33 923.045,1182.16 923.899,1182.99 924.753,1183.82 925.607,1184.65 926.461,1185.47 927.314,1186.3 928.168,1187.13 929.022,1187.96 929.876,1188.78 930.73,1189.61 931.584,1190.44 932.438,1191.27 933.292,1192.09 934.146,1192.92 935,1193.75 935.853,1194.57 936.707,1195.4 937.561,1196.23 938.415,1197.05 939.269,1197.88 940.123,1198.71 940.977,1199.53 941.831,1200.36 942.685,1201.19 943.538,1202.01 944.392,1202.84 945.246,1203.67 946.1,1204.49 946.954,1205.32 947.808,1206.15 948.662,1206.97 949.516,1207.8 950.37,1208.62 951.223,1209.45 952.077,1210.28 952.931,1211.1 953.785,1211.93 954.639,1212.75 955.493,1213.58 956.347,1214.4 957.201,1215.23 958.055,1216.06 958.909,1216.88 959.762,1217.71 960.616,1218.53 961.47,1219.36 962.324,1220.18 963.178,1221.01 964.032,1221.83 964.886,1222.66 965.74,1223.48 966.594,1224.31 967.447,1225.13 968.301,1225.96 969.155,1226.78 970.009,1227.61 970.863,1228.43 971.717,1229.26 972.571,1230.08 973.425,1230.91 974.279,1231.73 975.132,1232.56 975.986,1233.38 976.84,1234.21 977.694,1235.03 978.548,1235.86 979.402,1236.68 980.256,1237.51 981.11,1238.33 981.964,1239.15 982.817,1239.98 983.671,1240.8 984.525,1241.63 985.379,1242.45 986.233,1243.28 987.087,1244.1 987.941,1244.92 988.795,1245.75 989.649,1246.57 990.503,1247.4 991.356,1248.22 992.21,1249.04 993.064,1249.87 993.918,1250.69 994.772,1251.52 995.626,1252.34 996.48,1253.16 997.334,1253.99 998.188,1254.81 999.041,1255.63 999.895,1256.46 1000.75,1257.28 1001.6,1258.11 1002.46,1258.93 1003.31,1259.75 1004.16,1260.58 1005.02,1261.4 1005.87,1262.22 1006.73,1263.05 1007.58,1263.87 1008.43,1264.69 1009.29,1265.52 1010.14,1266.34 1011,1267.16 1011.85,1267.99 1012.7,1268.81 1013.56,1269.63 1014.41,1270.46 1015.27,1271.28 1016.12,1272.1 1016.97,1272.93 1017.83,1273.75 1018.68,1274.57 1019.53,1275.4 1020.39,1276.22 1021.24,1277.04 1022.1,1277.86 1022.95,1278.69 1023.8,1279.51 1024.66,1280.33 1025.51,1281.16 1026.37,1281.98 1027.22,1282.8 1028.07,1283.62 1028.93,1284.45 1029.78,1285.27 1030.64,1286.09 1031.49,1286.92 1032.34,1287.74 1033.2,1288.56 1034.05,1289.38 1034.9,1290.21 1035.76,1291.03 1036.61,1291.85 1037.47,1292.67 1038.32,1293.5 1039.17,1294.32 1040.03,1295.14 1040.88,1295.96 1041.74,1296.79 1042.59,1297.61 1043.44,1298.43 1044.3,1299.25 1045.15,1300.08 1046.01,1300.9 1046.86,1301.72 1047.71,1302.54 1048.57,1303.37 1049.42,1304.19 1050.27,1305.01 1051.13,1305.83 1051.98,1306.66 1052.84,1307.48 1053.69,1308.3 1054.54,1309.12 1055.4,1309.94 1056.25,1310.77 1057.11,1311.59 1057.96,1312.41 1058.81,1313.23 1059.67,1314.05 1060.52,1314.88 1061.38,1315.7 1062.23,1316.52 1063.08,1317.34 1063.94,1318.17 1064.79,1318.99 1065.65,1319.81 1066.5,1320.63 1067.35,1321.45 1068.21,1322.28 1069.06,1323.1 1069.91,1323.92 1070.77,1324.74 1071.62,1325.56 1072.48,1326.38 1073.33,1327.21 1074.18,1328.03 1075.04,1328.85 1075.89,1329.67 1076.75,1330.49 1077.6,1331.32 1078.45,1332.14 1079.31,1332.96 1080.16,1333.78 1081.02,1334.6 1081.87,1335.43 1082.72,1336.25 1083.58,1337.07 1084.43,1337.89 1085.28,1338.71 1086.14,1339.53 1086.99,1340.36 1087.85,1341.18 1088.7,1342 1089.55,1342.82 1090.41,1343.64 1091.26,1344.46 1092.12,1345.29 1092.97,1346.11 1093.82,1346.93 1094.68,1347.75 1095.53,1348.57 1096.39,1349.39 1097.24,1350.21 1098.09,1351.04 1098.95,1351.86 1099.8,1352.68 1100.65,1353.5 1101.51,1354.32 1102.36,1355.14 1103.22,1355.97 1104.07,1356.79 1104.92,1357.61 1105.78,1358.43 1106.63,1359.25 1107.49,1360.07 1108.34,1360.89 1109.19,1361.72 1110.05,1362.54 1110.9,1363.36 1111.76,1364.18 1112.61,1365 1113.46,1365.82 1114.32,1366.64 1115.17,1367.47 1116.02,1368.29 1116.88,1369.11 1117.73,1369.93 1118.59,1370.75 1119.44,1371.57 1120.29,1372.39 1121.15,1373.22 1122,1374.04 1122.86,1374.86 1123.71,1375.68 1124.56,1376.5 1125.42,1377.32 1126.27,1378.14 1127.13,1378.96 1127.98,1379.79 1128.83,1380.61 1129.69,1381.43 1130.54,1382.25 1131.39,1383.07 1132.25,1383.89 1133.1,1384.71 1133.96,1385.53 1134.81,1386.36 1135.66,1387.18 1136.52,1388 1137.37,1388.82 1138.23,1389.64 1139.08,1390.46 1139.93,1391.28 1140.79,1392.1 1141.64,1392.92 1142.5,1393.75 1143.35,1394.57 1144.2,1395.39 1145.06,1396.21 1145.91,1397.03 1146.76,1397.85 1147.62,1398.67 1148.47,1399.49 1149.33,1400.31 1150.18,1401.14 1151.03,1401.96 1151.89,1402.78 1152.74,1403.6 1153.6,1404.42 1154.45,1405.24 1155.3,1406.06 1156.16,1406.88 "/>
+<path clip-path="url(#clip570)" d="M1448.54 1423.18 L2352.76 1423.18 L2352.76 847.244 L1448.54 847.244  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip575">
+    <rect x="1448" y="847" width="905" height="577"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1473.27,1423.18 1473.27,847.244 "/>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1686.75,1423.18 1686.75,847.244 "/>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1900.22,1423.18 1900.22,847.244 "/>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2113.69,1423.18 2113.69,847.244 "/>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2327.16,1423.18 2327.16,847.244 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,1423.18 2352.76,1423.18 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1473.27,1423.18 1473.27,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1686.75,1423.18 1686.75,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1900.22,1423.18 1900.22,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2113.69,1423.18 2113.69,1404.28 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2327.16,1423.18 2327.16,1404.28 "/>
+<path clip-path="url(#clip570)" d="M1435.58 1454.1 Q1431.96 1454.1 1430.14 1457.66 Q1428.33 1461.2 1428.33 1468.33 Q1428.33 1475.44 1430.14 1479.01 Q1431.96 1482.55 1435.58 1482.55 Q1439.21 1482.55 1441.02 1479.01 Q1442.84 1475.44 1442.84 1468.33 Q1442.84 1461.2 1441.02 1457.66 Q1439.21 1454.1 1435.58 1454.1 M1435.58 1450.39 Q1441.39 1450.39 1444.44 1455 Q1447.52 1459.58 1447.52 1468.33 Q1447.52 1477.06 1444.44 1481.67 Q1441.39 1486.25 1435.58 1486.25 Q1429.77 1486.25 1426.69 1481.67 Q1423.63 1477.06 1423.63 1468.33 Q1423.63 1459.58 1426.69 1455 Q1429.77 1450.39 1435.58 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1455.74 1479.7 L1460.62 1479.7 L1460.62 1485.58 L1455.74 1485.58 L1455.74 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1480.81 1454.1 Q1477.2 1454.1 1475.37 1457.66 Q1473.56 1461.2 1473.56 1468.33 Q1473.56 1475.44 1475.37 1479.01 Q1477.2 1482.55 1480.81 1482.55 Q1484.44 1482.55 1486.25 1479.01 Q1488.08 1475.44 1488.08 1468.33 Q1488.08 1461.2 1486.25 1457.66 Q1484.44 1454.1 1480.81 1454.1 M1480.81 1450.39 Q1486.62 1450.39 1489.67 1455 Q1492.75 1459.58 1492.75 1468.33 Q1492.75 1477.06 1489.67 1481.67 Q1486.62 1486.25 1480.81 1486.25 Q1475 1486.25 1471.92 1481.67 Q1468.86 1477.06 1468.86 1468.33 Q1468.86 1459.58 1471.92 1455 Q1475 1450.39 1480.81 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1510.97 1454.1 Q1507.36 1454.1 1505.53 1457.66 Q1503.72 1461.2 1503.72 1468.33 Q1503.72 1475.44 1505.53 1479.01 Q1507.36 1482.55 1510.97 1482.55 Q1514.6 1482.55 1516.41 1479.01 Q1518.24 1475.44 1518.24 1468.33 Q1518.24 1461.2 1516.41 1457.66 Q1514.6 1454.1 1510.97 1454.1 M1510.97 1450.39 Q1516.78 1450.39 1519.83 1455 Q1522.91 1459.58 1522.91 1468.33 Q1522.91 1477.06 1519.83 1481.67 Q1516.78 1486.25 1510.97 1486.25 Q1505.16 1486.25 1502.08 1481.67 Q1499.02 1477.06 1499.02 1468.33 Q1499.02 1459.58 1502.08 1455 Q1505.16 1450.39 1510.97 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1649.55 1454.1 Q1645.94 1454.1 1644.11 1457.66 Q1642.3 1461.2 1642.3 1468.33 Q1642.3 1475.44 1644.11 1479.01 Q1645.94 1482.55 1649.55 1482.55 Q1653.18 1482.55 1654.99 1479.01 Q1656.82 1475.44 1656.82 1468.33 Q1656.82 1461.2 1654.99 1457.66 Q1653.18 1454.1 1649.55 1454.1 M1649.55 1450.39 Q1655.36 1450.39 1658.41 1455 Q1661.49 1459.58 1661.49 1468.33 Q1661.49 1477.06 1658.41 1481.67 Q1655.36 1486.25 1649.55 1486.25 Q1643.74 1486.25 1640.66 1481.67 Q1637.6 1477.06 1637.6 1468.33 Q1637.6 1459.58 1640.66 1455 Q1643.74 1450.39 1649.55 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1669.71 1479.7 L1674.59 1479.7 L1674.59 1485.58 L1669.71 1485.58 L1669.71 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1694.78 1454.1 Q1691.17 1454.1 1689.34 1457.66 Q1687.53 1461.2 1687.53 1468.33 Q1687.53 1475.44 1689.34 1479.01 Q1691.17 1482.55 1694.78 1482.55 Q1698.41 1482.55 1700.22 1479.01 Q1702.05 1475.44 1702.05 1468.33 Q1702.05 1461.2 1700.22 1457.66 Q1698.41 1454.1 1694.78 1454.1 M1694.78 1450.39 Q1700.59 1450.39 1703.64 1455 Q1706.72 1459.58 1706.72 1468.33 Q1706.72 1477.06 1703.64 1481.67 Q1700.59 1486.25 1694.78 1486.25 Q1688.97 1486.25 1685.89 1481.67 Q1682.83 1477.06 1682.83 1468.33 Q1682.83 1459.58 1685.89 1455 Q1688.97 1450.39 1694.78 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1714.99 1451.02 L1733.34 1451.02 L1733.34 1454.96 L1719.27 1454.96 L1719.27 1463.43 Q1720.29 1463.08 1721.31 1462.92 Q1722.32 1462.73 1723.34 1462.73 Q1729.13 1462.73 1732.51 1465.9 Q1735.89 1469.08 1735.89 1474.49 Q1735.89 1480.07 1732.42 1483.17 Q1728.94 1486.25 1722.62 1486.25 Q1720.45 1486.25 1718.18 1485.88 Q1715.94 1485.51 1713.53 1484.77 L1713.53 1480.07 Q1715.61 1481.2 1717.83 1481.76 Q1720.06 1482.32 1722.53 1482.32 Q1726.54 1482.32 1728.87 1480.21 Q1731.21 1478.1 1731.21 1474.49 Q1731.21 1470.88 1728.87 1468.77 Q1726.54 1466.67 1722.53 1466.67 Q1720.66 1466.67 1718.78 1467.08 Q1716.93 1467.5 1714.99 1468.38 L1714.99 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1862.52 1454.1 Q1858.91 1454.1 1857.08 1457.66 Q1855.28 1461.2 1855.28 1468.33 Q1855.28 1475.44 1857.08 1479.01 Q1858.91 1482.55 1862.52 1482.55 Q1866.16 1482.55 1867.96 1479.01 Q1869.79 1475.44 1869.79 1468.33 Q1869.79 1461.2 1867.96 1457.66 Q1866.16 1454.1 1862.52 1454.1 M1862.52 1450.39 Q1868.33 1450.39 1871.39 1455 Q1874.47 1459.58 1874.47 1468.33 Q1874.47 1477.06 1871.39 1481.67 Q1868.33 1486.25 1862.52 1486.25 Q1856.71 1486.25 1853.63 1481.67 Q1850.58 1477.06 1850.58 1468.33 Q1850.58 1459.58 1853.63 1455 Q1856.71 1450.39 1862.52 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1882.68 1479.7 L1887.57 1479.7 L1887.57 1485.58 L1882.68 1485.58 L1882.68 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1898.56 1481.64 L1906.2 1481.64 L1906.2 1455.28 L1897.89 1456.95 L1897.89 1452.69 L1906.16 1451.02 L1910.83 1451.02 L1910.83 1481.64 L1918.47 1481.64 L1918.47 1485.58 L1898.56 1485.58 L1898.56 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1937.92 1454.1 Q1934.3 1454.1 1932.48 1457.66 Q1930.67 1461.2 1930.67 1468.33 Q1930.67 1475.44 1932.48 1479.01 Q1934.3 1482.55 1937.92 1482.55 Q1941.55 1482.55 1943.35 1479.01 Q1945.18 1475.44 1945.18 1468.33 Q1945.18 1461.2 1943.35 1457.66 Q1941.55 1454.1 1937.92 1454.1 M1937.92 1450.39 Q1943.73 1450.39 1946.78 1455 Q1949.86 1459.58 1949.86 1468.33 Q1949.86 1477.06 1946.78 1481.67 Q1943.73 1486.25 1937.92 1486.25 Q1932.1 1486.25 1929.03 1481.67 Q1925.97 1477.06 1925.97 1468.33 Q1925.97 1459.58 1929.03 1455 Q1932.1 1450.39 1937.92 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2076.49 1454.1 Q2072.88 1454.1 2071.05 1457.66 Q2069.25 1461.2 2069.25 1468.33 Q2069.25 1475.44 2071.05 1479.01 Q2072.88 1482.55 2076.49 1482.55 Q2080.13 1482.55 2081.93 1479.01 Q2083.76 1475.44 2083.76 1468.33 Q2083.76 1461.2 2081.93 1457.66 Q2080.13 1454.1 2076.49 1454.1 M2076.49 1450.39 Q2082.3 1450.39 2085.36 1455 Q2088.44 1459.58 2088.44 1468.33 Q2088.44 1477.06 2085.36 1481.67 Q2082.3 1486.25 2076.49 1486.25 Q2070.68 1486.25 2067.6 1481.67 Q2064.55 1477.06 2064.55 1468.33 Q2064.55 1459.58 2067.6 1455 Q2070.68 1450.39 2076.49 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2096.65 1479.7 L2101.54 1479.7 L2101.54 1485.58 L2096.65 1485.58 L2096.65 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2112.53 1481.64 L2120.17 1481.64 L2120.17 1455.28 L2111.86 1456.95 L2111.86 1452.69 L2120.13 1451.02 L2124.8 1451.02 L2124.8 1481.64 L2132.44 1481.64 L2132.44 1485.58 L2112.53 1485.58 L2112.53 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2141.93 1451.02 L2160.29 1451.02 L2160.29 1454.96 L2146.21 1454.96 L2146.21 1463.43 Q2147.23 1463.08 2148.25 1462.92 Q2149.27 1462.73 2150.29 1462.73 Q2156.08 1462.73 2159.46 1465.9 Q2162.83 1469.08 2162.83 1474.49 Q2162.83 1480.07 2159.36 1483.17 Q2155.89 1486.25 2149.57 1486.25 Q2147.4 1486.25 2145.13 1485.88 Q2142.88 1485.51 2140.47 1484.77 L2140.47 1480.07 Q2142.56 1481.2 2144.78 1481.76 Q2147 1482.32 2149.48 1482.32 Q2153.48 1482.32 2155.82 1480.21 Q2158.16 1478.1 2158.16 1474.49 Q2158.16 1470.88 2155.82 1468.77 Q2153.48 1466.67 2149.48 1466.67 Q2147.6 1466.67 2145.73 1467.08 Q2143.88 1467.5 2141.93 1468.38 L2141.93 1451.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2289.47 1454.1 Q2285.86 1454.1 2284.03 1457.66 Q2282.22 1461.2 2282.22 1468.33 Q2282.22 1475.44 2284.03 1479.01 Q2285.86 1482.55 2289.47 1482.55 Q2293.1 1482.55 2294.91 1479.01 Q2296.74 1475.44 2296.74 1468.33 Q2296.74 1461.2 2294.91 1457.66 Q2293.1 1454.1 2289.47 1454.1 M2289.47 1450.39 Q2295.28 1450.39 2298.33 1455 Q2301.41 1459.58 2301.41 1468.33 Q2301.41 1477.06 2298.33 1481.67 Q2295.28 1486.25 2289.47 1486.25 Q2283.66 1486.25 2280.58 1481.67 Q2277.52 1477.06 2277.52 1468.33 Q2277.52 1459.58 2280.58 1455 Q2283.66 1450.39 2289.47 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2309.63 1479.7 L2314.51 1479.7 L2314.51 1485.58 L2309.63 1485.58 L2309.63 1479.7 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2328.73 1481.64 L2345.05 1481.64 L2345.05 1485.58 L2323.1 1485.58 L2323.1 1481.64 Q2325.76 1478.89 2330.35 1474.26 Q2334.95 1469.61 2336.13 1468.27 Q2338.38 1465.74 2339.26 1464.01 Q2340.16 1462.25 2340.16 1460.56 Q2340.16 1457.8 2338.22 1456.07 Q2336.3 1454.33 2333.19 1454.33 Q2331 1454.33 2328.54 1455.09 Q2326.11 1455.86 2323.33 1457.41 L2323.33 1452.69 Q2326.16 1451.55 2328.61 1450.97 Q2331.07 1450.39 2333.1 1450.39 Q2338.47 1450.39 2341.67 1453.08 Q2344.86 1455.77 2344.86 1460.26 Q2344.86 1462.39 2344.05 1464.31 Q2343.26 1466.2 2341.16 1468.8 Q2340.58 1469.47 2337.48 1472.69 Q2334.38 1475.88 2328.73 1481.64 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2364.86 1454.1 Q2361.25 1454.1 2359.42 1457.66 Q2357.62 1461.2 2357.62 1468.33 Q2357.62 1475.44 2359.42 1479.01 Q2361.25 1482.55 2364.86 1482.55 Q2368.5 1482.55 2370.3 1479.01 Q2372.13 1475.44 2372.13 1468.33 Q2372.13 1461.2 2370.3 1457.66 Q2368.5 1454.1 2364.86 1454.1 M2364.86 1450.39 Q2370.67 1450.39 2373.73 1455 Q2376.81 1459.58 2376.81 1468.33 Q2376.81 1477.06 2373.73 1481.67 Q2370.67 1486.25 2364.86 1486.25 Q2359.05 1486.25 2355.97 1481.67 Q2352.92 1477.06 2352.92 1468.33 Q2352.92 1459.58 2355.97 1455 Q2359.05 1450.39 2364.86 1450.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1770.82 1520.52 L1811.02 1520.52 L1811.02 1525.93 L1794.15 1525.93 L1794.15 1568.04 L1787.69 1568.04 L1787.69 1525.93 L1770.82 1525.93 L1770.82 1520.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1814.96 1532.4 L1820.82 1532.4 L1820.82 1568.04 L1814.96 1568.04 L1814.96 1532.4 M1814.96 1518.52 L1820.82 1518.52 L1820.82 1525.93 L1814.96 1525.93 L1814.96 1518.52 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1860.83 1539.24 Q1863.02 1535.29 1866.08 1533.41 Q1869.14 1531.54 1873.27 1531.54 Q1878.84 1531.54 1881.87 1535.45 Q1884.89 1539.33 1884.89 1546.53 L1884.89 1568.04 L1879 1568.04 L1879 1546.72 Q1879 1541.59 1877.19 1539.11 Q1875.37 1536.63 1871.65 1536.63 Q1867.1 1536.63 1864.46 1539.65 Q1861.81 1542.68 1861.81 1547.9 L1861.81 1568.04 L1855.93 1568.04 L1855.93 1546.72 Q1855.93 1541.56 1854.11 1539.11 Q1852.3 1536.63 1848.51 1536.63 Q1844.02 1536.63 1841.38 1539.68 Q1838.74 1542.71 1838.74 1547.9 L1838.74 1568.04 L1832.85 1568.04 L1832.85 1532.4 L1838.74 1532.4 L1838.74 1537.93 Q1840.74 1534.66 1843.55 1533.1 Q1846.35 1531.54 1850.2 1531.54 Q1854.08 1531.54 1856.79 1533.51 Q1859.52 1535.48 1860.83 1539.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1927.06 1548.76 L1927.06 1551.62 L1900.14 1551.62 Q1900.52 1557.67 1903.76 1560.85 Q1907.04 1564 1912.87 1564 Q1916.24 1564 1919.39 1563.17 Q1922.58 1562.35 1925.69 1560.69 L1925.69 1566.23 Q1922.54 1567.57 1919.23 1568.27 Q1915.92 1568.97 1912.52 1568.97 Q1903.99 1568.97 1898.99 1564 Q1894.03 1559.04 1894.03 1550.57 Q1894.03 1541.82 1898.74 1536.69 Q1903.48 1531.54 1911.5 1531.54 Q1918.69 1531.54 1922.86 1536.18 Q1927.06 1540.8 1927.06 1548.76 M1921.21 1547.04 Q1921.14 1542.23 1918.5 1539.37 Q1915.89 1536.5 1911.56 1536.5 Q1906.66 1536.5 1903.7 1539.27 Q1900.77 1542.04 1900.33 1547.07 L1921.21 1547.04 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1971.46 1518.58 Q1967.2 1525.9 1965.13 1533.06 Q1963.06 1540.23 1963.06 1547.58 Q1963.06 1554.93 1965.13 1562.16 Q1967.23 1569.35 1971.46 1576.64 L1966.37 1576.64 Q1961.6 1569.16 1959.21 1561.93 Q1956.85 1554.71 1956.85 1547.58 Q1956.85 1540.48 1959.21 1533.29 Q1961.57 1526.09 1966.37 1518.58 L1971.46 1518.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2005.55 1533.45 L2005.55 1538.98 Q2003.07 1537.71 2000.4 1537.07 Q1997.72 1536.44 1994.86 1536.44 Q1990.5 1536.44 1988.3 1537.77 Q1986.14 1539.11 1986.14 1541.79 Q1986.14 1543.82 1987.7 1545 Q1989.26 1546.15 1993.97 1547.2 L1995.97 1547.64 Q2002.21 1548.98 2004.82 1551.43 Q2007.46 1553.85 2007.46 1558.21 Q2007.46 1563.17 2003.52 1566.07 Q1999.6 1568.97 1992.73 1568.97 Q1989.86 1568.97 1986.74 1568.39 Q1983.65 1567.85 1980.22 1566.74 L1980.22 1560.69 Q1983.46 1562.38 1986.61 1563.24 Q1989.77 1564.07 1992.85 1564.07 Q1996.99 1564.07 1999.22 1562.66 Q2001.45 1561.23 2001.45 1558.65 Q2001.45 1556.27 1999.82 1554.99 Q1998.23 1553.72 1992.79 1552.54 L1990.75 1552.07 Q1985.31 1550.92 1982.89 1548.56 Q1980.47 1546.18 1980.47 1542.04 Q1980.47 1537.01 1984.04 1534.27 Q1987.6 1531.54 1994.16 1531.54 Q1997.4 1531.54 2000.27 1532.01 Q2003.13 1532.49 2005.55 1533.45 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M2015.86 1518.58 L2020.96 1518.58 Q2025.73 1526.09 2028.09 1533.29 Q2030.47 1540.48 2030.47 1547.58 Q2030.47 1554.71 2028.09 1561.93 Q2025.73 1569.16 2020.96 1576.64 L2015.86 1576.64 Q2020.1 1569.35 2022.17 1562.16 Q2024.27 1554.93 2024.27 1547.58 Q2024.27 1540.23 2022.17 1533.06 Q2020.1 1525.9 2015.86 1518.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,1406.88 2352.76,1406.88 "/>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,1251.63 2352.76,1251.63 "/>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,1096.38 2352.76,1096.38 "/>
+<polyline clip-path="url(#clip575)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1448.54,941.129 2352.76,941.129 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,1423.18 1448.54,847.244 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,1406.88 1467.43,1406.88 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,1251.63 1467.43,1251.63 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,1096.38 1467.43,1096.38 "/>
+<polyline clip-path="url(#clip570)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1448.54,941.129 1467.43,941.129 "/>
+<path clip-path="url(#clip570)" d="M1400.59 1392.68 Q1396.98 1392.68 1395.15 1396.24 Q1393.35 1399.78 1393.35 1406.91 Q1393.35 1414.02 1395.15 1417.59 Q1396.98 1421.13 1400.59 1421.13 Q1404.22 1421.13 1406.03 1417.59 Q1407.86 1414.02 1407.86 1406.91 Q1407.86 1399.78 1406.03 1396.24 Q1404.22 1392.68 1400.59 1392.68 M1400.59 1388.97 Q1406.4 1388.97 1409.46 1393.58 Q1412.54 1398.16 1412.54 1406.91 Q1412.54 1415.64 1409.46 1420.25 Q1406.4 1424.83 1400.59 1424.83 Q1394.78 1424.83 1391.7 1420.25 Q1388.65 1415.64 1388.65 1406.91 Q1388.65 1398.16 1391.7 1393.58 Q1394.78 1388.97 1400.59 1388.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1392.63 1264.97 L1400.27 1264.97 L1400.27 1238.61 L1391.96 1240.28 L1391.96 1236.02 L1400.22 1234.35 L1404.9 1234.35 L1404.9 1264.97 L1412.54 1264.97 L1412.54 1268.91 L1392.63 1268.91 L1392.63 1264.97 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1396.22 1109.72 L1412.54 1109.72 L1412.54 1113.66 L1390.59 1113.66 L1390.59 1109.72 Q1393.25 1106.97 1397.84 1102.34 Q1402.44 1097.69 1403.62 1096.34 Q1405.87 1093.82 1406.75 1092.09 Q1407.65 1090.33 1407.65 1088.64 Q1407.65 1085.88 1405.71 1084.15 Q1403.79 1082.41 1400.68 1082.41 Q1398.48 1082.41 1396.03 1083.17 Q1393.6 1083.94 1390.82 1085.49 L1390.82 1080.77 Q1393.65 1079.63 1396.1 1079.05 Q1398.55 1078.47 1400.59 1078.47 Q1405.96 1078.47 1409.16 1081.16 Q1412.35 1083.84 1412.35 1088.34 Q1412.35 1090.47 1411.54 1092.39 Q1410.75 1094.28 1408.65 1096.88 Q1408.07 1097.55 1404.97 1100.77 Q1401.86 1103.96 1396.22 1109.72 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1405.41 939.775 Q1408.76 940.493 1410.64 942.761 Q1412.54 945.03 1412.54 948.363 Q1412.54 953.479 1409.02 956.28 Q1405.5 959.081 1399.02 959.081 Q1396.84 959.081 1394.53 958.641 Q1392.23 958.224 1389.78 957.368 L1389.78 952.854 Q1391.73 953.988 1394.04 954.567 Q1396.35 955.146 1398.88 955.146 Q1403.28 955.146 1405.57 953.41 Q1407.88 951.673 1407.88 948.363 Q1407.88 945.308 1405.73 943.595 Q1403.6 941.859 1399.78 941.859 L1395.75 941.859 L1395.75 938.016 L1399.97 938.016 Q1403.41 938.016 1405.24 936.65 Q1407.07 935.261 1407.07 932.669 Q1407.07 930.007 1405.17 928.595 Q1403.3 927.16 1399.78 927.16 Q1397.86 927.16 1395.66 927.576 Q1393.46 927.993 1390.82 928.873 L1390.82 924.706 Q1393.48 923.965 1395.8 923.595 Q1398.14 923.225 1400.2 923.225 Q1405.52 923.225 1408.62 925.655 Q1411.72 928.062 1411.72 932.183 Q1411.72 935.053 1410.08 937.044 Q1408.44 939.011 1405.41 939.775 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1290.87 1238.91 L1290.87 1198.71 L1296.28 1198.71 L1296.28 1215.58 L1338.39 1215.58 L1338.39 1222.04 L1296.28 1222.04 L1296.28 1238.91 L1290.87 1238.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1316.87 1163.13 L1338.39 1163.13 L1338.39 1168.98 L1317.07 1168.98 Q1312 1168.98 1309.49 1170.96 Q1306.98 1172.93 1306.98 1176.88 Q1306.98 1181.62 1310 1184.36 Q1313.02 1187.09 1318.24 1187.09 L1338.39 1187.09 L1338.39 1192.98 L1288.87 1192.98 L1288.87 1187.09 L1308.28 1187.09 Q1305.07 1184.99 1303.47 1182.16 Q1301.88 1179.29 1301.88 1175.57 Q1301.88 1169.43 1305.7 1166.28 Q1309.49 1163.13 1316.87 1163.13 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1308.22 1130.79 Q1307.64 1131.77 1307.39 1132.95 Q1307.1 1134.1 1307.1 1135.5 Q1307.1 1140.46 1310.35 1143.14 Q1313.56 1145.78 1319.61 1145.78 L1338.39 1145.78 L1338.39 1151.67 L1302.74 1151.67 L1302.74 1145.78 L1308.28 1145.78 Q1305.03 1143.93 1303.47 1140.97 Q1301.88 1138.01 1301.88 1133.78 Q1301.88 1133.17 1301.98 1132.44 Q1302.04 1131.71 1302.2 1130.82 L1308.22 1130.79 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1324.32 1125.25 L1302.74 1125.25 L1302.74 1119.39 L1324.1 1119.39 Q1329.16 1119.39 1331.71 1117.42 Q1334.22 1115.45 1334.22 1111.5 Q1334.22 1106.76 1331.2 1104.02 Q1328.17 1101.25 1322.95 1101.25 L1302.74 1101.25 L1302.74 1095.39 L1338.39 1095.39 L1338.39 1101.25 L1332.92 1101.25 Q1336.16 1103.38 1337.75 1106.22 Q1339.31 1109.02 1339.31 1112.74 Q1339.31 1118.88 1335.49 1122.07 Q1331.67 1125.25 1324.32 1125.25 M1301.88 1110.51 L1301.88 1110.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1303.79 1060.61 L1309.33 1060.61 Q1308.06 1063.09 1307.42 1065.76 Q1306.78 1068.44 1306.78 1071.3 Q1306.78 1075.66 1308.12 1077.86 Q1309.46 1080.02 1312.13 1080.02 Q1314.17 1080.02 1315.35 1078.46 Q1316.49 1076.9 1317.54 1072.19 L1317.99 1070.19 Q1319.33 1063.95 1321.78 1061.34 Q1324.19 1058.7 1328.56 1058.7 Q1333.52 1058.7 1336.42 1062.64 Q1339.31 1066.56 1339.31 1073.43 Q1339.31 1076.3 1338.74 1079.42 Q1338.2 1082.5 1337.09 1085.94 L1331.04 1085.94 Q1332.72 1082.69 1333.58 1079.54 Q1334.41 1076.39 1334.41 1073.31 Q1334.41 1069.17 1333.01 1066.94 Q1331.58 1064.71 1329 1064.71 Q1326.61 1064.71 1325.34 1066.34 Q1324.07 1067.93 1322.89 1073.37 L1322.41 1075.41 Q1321.27 1080.85 1318.91 1083.27 Q1316.52 1085.69 1312.39 1085.69 Q1307.36 1085.69 1304.62 1082.12 Q1301.88 1078.56 1301.88 1072 Q1301.88 1068.75 1302.36 1065.89 Q1302.84 1063.02 1303.79 1060.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip570)" d="M1292.62 1043.58 L1302.74 1043.58 L1302.74 1031.51 L1307.29 1031.51 L1307.29 1043.58 L1326.65 1043.58 Q1331.01 1043.58 1332.25 1042.4 Q1333.49 1041.19 1333.49 1037.53 L1333.49 1031.51 L1338.39 1031.51 L1338.39 1037.53 Q1338.39 1044.31 1335.88 1046.89 Q1333.33 1049.47 1326.65 1049.47 L1307.29 1049.47 L1307.29 1053.76 L1302.74 1053.76 L1302.74 1049.47 L1292.62 1049.47 L1292.62 1043.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip575)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1474.13,863.544 1474.98,863.545 1475.83,863.546 1476.69,863.547 1477.54,863.547 1478.4,863.548 1479.25,863.549 1480.1,863.55 1480.96,863.551 1481.81,863.552 1482.67,863.553 1483.52,863.554 1484.37,863.555 1485.23,863.556 1486.08,863.557 1486.93,863.558 1487.79,863.559 1488.64,863.561 1489.5,863.562 1490.35,863.563 1491.2,863.565 1492.06,863.566 1492.91,863.567 1493.77,863.569 1494.62,863.57 1495.47,863.572 1496.33,863.574 1497.18,863.575 1498.04,863.577 1498.89,863.579 1499.74,863.581 1500.6,863.583 1501.45,863.585 1502.3,863.587 1503.16,863.589 1504.01,863.591 1504.87,863.594 1505.72,863.596 1506.57,863.599 1507.43,863.602 1508.28,863.604 1509.14,863.607 1509.99,863.61 1510.84,863.613 1511.7,863.617 1512.55,863.62 1513.41,863.624 1514.26,863.628 1515.11,863.632 1515.97,863.636 1516.82,863.64 1517.67,863.645 1518.53,863.65 1519.38,863.655 1520.24,863.66 1521.09,863.666 1521.94,863.671 1522.8,863.678 1523.65,863.684 1524.51,863.691 1525.36,863.698 1526.21,863.706 1527.07,863.714 1527.92,863.723 1528.78,863.732 1529.63,863.742 1530.48,863.753 1531.34,863.764 1532.19,863.776 1533.04,863.789 1533.9,863.803 1534.75,863.817 1535.61,863.833 1536.46,863.85 1537.31,863.869 1538.17,863.888 1539.02,863.91 1539.88,863.933 1540.73,863.959 1541.58,863.986 1542.44,864.017 1543.29,864.05 1544.15,864.086 1545,864.127 1545.85,864.171 1546.71,864.221 1547.56,864.276 1548.41,864.337 1549.27,864.407 1550.12,864.485 1550.98,864.574 1551.83,864.675 1552.68,864.792 1553.54,864.926 1554.39,865.083 1555.25,865.267 1556.1,865.486 1556.95,865.748 1557.81,866.065 1558.66,866.454 1559.52,866.939 1560.37,867.556 1561.22,868.354 1562.08,869.416 1562.93,870.871 1563.78,872.937 1564.64,876.006 1565.49,880.805 1566.35,888.692 1567.2,901.827 1568.05,921.785 1568.91,946.426 1569.76,971.216 1570.62,993.263 1571.47,1011.72 1572.32,1026.73 1573.18,1038.76 1574.03,1048.31 1574.89,1055.86 1575.74,1061.82 1576.59,1066.5 1577.45,1070.18 1578.3,1073.06 1579.16,1075.32 1580.01,1077.1 1580.86,1078.48 1581.72,1079.56 1582.57,1080.4 1583.42,1081.05 1584.28,1081.55 1585.13,1081.94 1585.99,1082.23 1586.84,1082.45 1587.69,1082.61 1588.55,1082.72 1589.4,1082.8 1590.26,1082.84 1591.11,1082.87 1591.96,1082.87 1592.82,1082.86 1593.67,1082.84 1594.53,1082.8 1595.38,1082.76 1596.23,1082.71 1597.09,1082.65 1597.94,1082.59 1598.79,1082.52 1599.65,1082.45 1600.5,1082.38 1601.36,1082.3 1602.21,1082.22 1603.06,1082.14 1603.92,1082.05 1604.77,1081.96 1605.63,1081.87 1606.48,1081.77 1607.33,1081.68 1608.19,1081.58 1609.04,1081.47 1609.9,1081.37 1610.75,1081.26 1611.6,1081.15 1612.46,1081.04 1613.31,1080.92 1614.16,1080.81 1615.02,1080.69 1615.87,1080.57 1616.73,1080.44 1617.58,1080.31 1618.43,1080.18 1619.29,1080.05 1620.14,1079.92 1621,1079.78 1621.85,1079.64 1622.7,1079.5 1623.56,1079.35 1624.41,1079.2 1625.27,1079.05 1626.12,1078.9 1626.97,1078.74 1627.83,1078.59 1628.68,1078.42 1629.53,1078.26 1630.39,1078.09 1631.24,1077.92 1632.1,1077.75 1632.95,1077.57 1633.8,1077.39 1634.66,1077.21 1635.51,1077.03 1636.37,1076.84 1637.22,1076.65 1638.07,1076.45 1638.93,1076.26 1639.78,1076.06 1640.64,1075.85 1641.49,1075.65 1642.34,1075.44 1643.2,1075.22 1644.05,1075.01 1644.9,1074.79 1645.76,1074.56 1646.61,1074.34 1647.47,1074.11 1648.32,1073.87 1649.17,1073.64 1650.03,1073.4 1650.88,1073.15 1651.74,1072.9 1652.59,1072.65 1653.44,1072.4 1654.3,1072.14 1655.15,1071.88 1656.01,1071.61 1656.86,1071.34 1657.71,1071.07 1658.57,1070.79 1659.42,1070.51 1660.27,1070.23 1661.13,1069.94 1661.98,1069.64 1662.84,1069.35 1663.69,1069.05 1664.54,1068.74 1665.4,1068.43 1666.25,1068.12 1667.11,1067.8 1667.96,1067.48 1668.81,1067.15 1669.67,1066.82 1670.52,1066.49 1671.38,1066.15 1672.23,1065.8 1673.08,1065.45 1673.94,1065.1 1674.79,1064.74 1675.64,1064.38 1676.5,1064.01 1677.35,1063.64 1678.21,1063.26 1679.06,1062.88 1679.91,1062.5 1680.77,1062.1 1681.62,1061.71 1682.48,1061.31 1683.33,1060.9 1684.18,1060.49 1685.04,1060.07 1685.89,1059.65 1686.75,1059.22 1687.6,1058.79 1688.45,1058.35 1689.31,1057.9 1690.16,1057.45 1691.01,1057 1691.87,1056.54 1692.72,1056.07 1693.58,1055.6 1694.43,1055.12 1695.28,1054.63 1696.14,1054.14 1696.99,1053.64 1697.85,1053.14 1698.7,1052.63 1699.55,1052.12 1700.41,1051.6 1701.26,1051.07 1702.12,1050.53 1702.97,1049.99 1703.82,1049.44 1704.68,1048.89 1705.53,1048.33 1706.38,1047.76 1707.24,1047.18 1708.09,1046.6 1708.95,1046.01 1709.8,1045.42 1710.65,1044.81 1711.51,1044.2 1712.36,1043.58 1713.22,1042.96 1714.07,1042.32 1714.92,1041.68 1715.78,1041.03 1716.63,1040.38 1717.49,1039.71 1718.34,1039.04 1719.19,1038.36 1720.05,1037.67 1720.9,1036.97 1721.76,1036.27 1722.61,1035.55 1723.46,1034.83 1724.32,1034.1 1725.17,1033.36 1726.02,1032.61 1726.88,1031.86 1727.73,1031.09 1728.59,1030.31 1729.44,1029.53 1730.29,1028.74 1731.15,1027.94 1732,1027.12 1732.86,1026.3 1733.71,1025.47 1734.56,1024.64 1735.42,1023.79 1736.27,1022.93 1737.13,1022.06 1737.98,1021.19 1738.83,1020.3 1739.69,1019.41 1740.54,1018.51 1741.39,1017.6 1742.25,1016.68 1743.1,1015.75 1743.96,1014.82 1744.81,1013.87 1745.66,1012.93 1746.52,1011.97 1747.37,1011.01 1748.23,1010.04 1749.08,1009.07 1749.93,1008.09 1750.79,1007.11 1751.64,1006.13 1752.5,1005.15 1753.35,1004.17 1754.2,1003.19 1755.06,1002.22 1755.91,1001.25 1756.76,1000.29 1757.62,999.347 1758.47,998.414 1759.33,997.499 1760.18,996.604 1761.03,995.734 1761.89,994.894 1762.74,994.088 1763.6,993.322 1764.45,992.602 1765.3,991.935 1766.16,991.329 1767.01,990.792 1767.87,990.333 1768.72,989.964 1769.57,989.696 1770.43,989.542 1771.28,989.517 1772.13,989.636 1772.99,989.919 1773.84,990.384 1774.7,991.055 1775.55,991.957 1776.4,993.118 1777.26,994.57 1778.11,996.347 1778.97,998.489 1779.82,1001.04 1780.67,1004.05 1781.53,1007.57 1782.38,1011.67 1783.24,1016.41 1784.09,1021.87 1784.94,1028.13 1785.8,1035.28 1786.65,1043.42 1787.5,1052.67 1788.36,1063.14 1789.21,1074.95 1790.07,1088.25 1790.92,1103.17 1791.77,1119.85 1792.63,1138.45 1793.48,1159.09 1794.34,1181.88 1795.19,1206.86 1796.04,1233.99 1796.9,1263 1797.75,1293.2 1798.61,1323.12 1799.46,1349.95 1800.31,1370.16 1801.17,1382.76 1802.02,1390.02 1802.87,1394.36 1803.73,1397.12 1804.58,1399 1805.44,1400.35 1806.29,1401.35 1807.14,1402.12 1808,1402.73 1808.85,1403.22 1809.71,1403.62 1810.56,1403.96 1811.41,1404.24 1812.27,1404.48 1813.12,1404.69 1813.98,1404.87 1814.83,1405.03 1815.68,1405.16 1816.54,1405.29 1817.39,1405.39 1818.24,1405.49 1819.1,1405.58 1819.95,1405.66 1820.81,1405.73 1821.66,1405.79 1822.51,1405.85 1823.37,1405.91 1824.22,1405.96 1825.08,1406 1825.93,1406.04 1826.78,1406.08 1827.64,1406.12 1828.49,1406.15 1829.35,1406.18 1830.2,1406.21 1831.05,1406.24 1831.91,1406.27 1832.76,1406.29 1833.61,1406.31 1834.47,1406.33 1835.32,1406.35 1836.18,1406.37 1837.03,1406.39 1837.88,1406.4 1838.74,1406.42 1839.59,1406.43 1840.45,1406.45 1841.3,1406.46 1842.15,1406.47 1843.01,1406.49 1843.86,1406.5 1844.72,1406.51 1845.57,1406.52 1846.42,1406.53 1847.28,1406.54 1848.13,1406.55 1848.99,1406.56 1849.84,1406.56 1850.69,1406.57 1851.55,1406.58 1852.4,1406.59 1853.25,1406.59 1854.11,1406.6 1854.96,1406.61 1855.82,1406.61 1856.67,1406.62 1857.52,1406.62 1858.38,1406.63 1859.23,1406.64 1860.09,1406.64 1860.94,1406.65 1861.79,1406.65 1862.65,1406.65 1863.5,1406.66 1864.36,1406.66 1865.21,1406.67 1866.06,1406.67 1866.92,1406.68 1867.77,1406.68 1868.62,1406.68 1869.48,1406.69 1870.33,1406.69 1871.19,1406.69 1872.04,1406.7 1872.89,1406.7 1873.75,1406.7 1874.6,1406.71 1875.46,1406.71 1876.31,1406.71 1877.16,1406.71 1878.02,1406.72 1878.87,1406.72 1879.73,1406.72 1880.58,1406.72 1881.43,1406.73 1882.29,1406.73 1883.14,1406.73 1883.99,1406.73 1884.85,1406.73 1885.7,1406.74 1886.56,1406.74 1887.41,1406.74 1888.26,1406.74 1889.12,1406.74 1889.97,1406.75 1890.83,1406.75 1891.68,1406.75 1892.53,1406.75 1893.39,1406.75 1894.24,1406.75 1895.1,1406.76 1895.95,1406.76 1896.8,1406.76 1897.66,1406.76 1898.51,1406.76 1899.36,1406.76 1900.22,1406.76 1901.07,1406.77 1901.93,1406.77 1902.78,1406.77 1903.63,1406.77 1904.49,1406.77 1905.34,1406.77 1906.2,1406.77 1907.05,1406.77 1907.9,1406.78 1908.76,1406.78 1909.61,1406.78 1910.47,1406.78 1911.32,1406.78 1912.17,1406.78 1913.03,1406.78 1913.88,1406.78 1914.73,1406.78 1915.59,1406.78 1916.44,1406.79 1917.3,1406.79 1918.15,1406.79 1919,1406.79 1919.86,1406.79 1920.71,1406.79 1921.57,1406.79 1922.42,1406.79 1923.27,1406.79 1924.13,1406.79 1924.98,1406.79 1925.84,1406.79 1926.69,1406.8 1927.54,1406.8 1928.4,1406.8 1929.25,1406.8 1930.1,1406.8 1930.96,1406.8 1931.81,1406.8 1932.67,1406.8 1933.52,1406.8 1934.37,1406.8 1935.23,1406.8 1936.08,1406.8 1936.94,1406.8 1937.79,1406.8 1938.64,1406.8 1939.5,1406.81 1940.35,1406.81 1941.21,1406.81 1942.06,1406.81 1942.91,1406.81 1943.77,1406.81 1944.62,1406.81 1945.47,1406.81 1946.33,1406.81 1947.18,1406.81 1948.04,1406.81 1948.89,1406.81 1949.74,1406.81 1950.6,1406.81 1951.45,1406.81 1952.31,1406.81 1953.16,1406.81 1954.01,1406.81 1954.87,1406.81 1955.72,1406.81 1956.58,1406.82 1957.43,1406.82 1958.28,1406.82 1959.14,1406.82 1959.99,1406.82 1960.84,1406.82 1961.7,1406.82 1962.55,1406.82 1963.41,1406.82 1964.26,1406.82 1965.11,1406.82 1965.97,1406.82 1966.82,1406.82 1967.68,1406.82 1968.53,1406.82 1969.38,1406.82 1970.24,1406.82 1971.09,1406.82 1971.95,1406.82 1972.8,1406.82 1973.65,1406.82 1974.51,1406.82 1975.36,1406.82 1976.21,1406.82 1977.07,1406.82 1977.92,1406.82 1978.78,1406.82 1979.63,1406.82 1980.48,1406.83 1981.34,1406.83 1982.19,1406.83 1983.05,1406.83 1983.9,1406.83 1984.75,1406.83 1985.61,1406.83 1986.46,1406.83 1987.32,1406.83 1988.17,1406.83 1989.02,1406.83 1989.88,1406.83 1990.73,1406.83 1991.59,1406.83 1992.44,1406.83 1993.29,1406.83 1994.15,1406.83 1995,1406.83 1995.85,1406.83 1996.71,1406.83 1997.56,1406.83 1998.42,1406.83 1999.27,1406.83 2000.12,1406.83 2000.98,1406.83 2001.83,1406.83 2002.69,1406.83 2003.54,1406.83 2004.39,1406.83 2005.25,1406.83 2006.1,1406.83 2006.96,1406.83 2007.81,1406.83 2008.66,1406.83 2009.52,1406.83 2010.37,1406.83 2011.22,1406.83 2012.08,1406.83 2012.93,1406.83 2013.79,1406.83 2014.64,1406.83 2015.49,1406.83 2016.35,1406.83 2017.2,1406.84 2018.06,1406.84 2018.91,1406.84 2019.76,1406.84 2020.62,1406.84 2021.47,1406.84 2022.33,1406.84 2023.18,1406.84 2024.03,1406.84 2024.89,1406.84 2025.74,1406.84 2026.59,1406.84 2027.45,1406.84 2028.3,1406.84 2029.16,1406.84 2030.01,1406.84 2030.86,1406.84 2031.72,1406.84 2032.57,1406.84 2033.43,1406.84 2034.28,1406.84 2035.13,1406.84 2035.99,1406.84 2036.84,1406.84 2037.7,1406.84 2038.55,1406.84 2039.4,1406.84 2040.26,1406.84 2041.11,1406.84 2041.96,1406.84 2042.82,1406.84 2043.67,1406.84 2044.53,1406.84 2045.38,1406.84 2046.23,1406.84 2047.09,1406.84 2047.94,1406.84 2048.8,1406.84 2049.65,1406.84 2050.5,1406.84 2051.36,1406.84 2052.21,1406.84 2053.07,1406.84 2053.92,1406.84 2054.77,1406.84 2055.63,1406.84 2056.48,1406.84 2057.33,1406.84 2058.19,1406.84 2059.04,1406.84 2059.9,1406.84 2060.75,1406.84 2061.6,1406.84 2062.46,1406.84 2063.31,1406.84 2064.17,1406.84 2065.02,1406.84 2065.87,1406.84 2066.73,1406.84 2067.58,1406.84 2068.44,1406.84 2069.29,1406.84 2070.14,1406.84 2071,1406.84 2071.85,1406.84 2072.7,1406.84 2073.56,1406.84 2074.41,1406.84 2075.27,1406.84 2076.12,1406.84 2076.97,1406.84 2077.83,1406.84 2078.68,1406.84 2079.54,1406.84 2080.39,1406.84 2081.24,1406.84 2082.1,1406.84 2082.95,1406.84 2083.81,1406.85 2084.66,1406.85 2085.51,1406.85 2086.37,1406.85 2087.22,1406.85 2088.07,1406.85 2088.93,1406.85 2089.78,1406.85 2090.64,1406.85 2091.49,1406.85 2092.34,1406.85 2093.2,1406.85 2094.05,1406.85 2094.91,1406.85 2095.76,1406.85 2096.61,1406.85 2097.47,1406.85 2098.32,1406.85 2099.18,1406.85 2100.03,1406.85 2100.88,1406.85 2101.74,1406.85 2102.59,1406.85 2103.44,1406.85 2104.3,1406.85 2105.15,1406.85 2106.01,1406.85 2106.86,1406.85 2107.71,1406.85 2108.57,1406.85 2109.42,1406.85 2110.28,1406.85 2111.13,1406.85 2111.98,1406.85 2112.84,1406.85 2113.69,1406.85 2114.55,1406.85 2115.4,1406.85 2116.25,1406.85 2117.11,1406.85 2117.96,1406.85 2118.82,1406.85 2119.67,1406.85 2120.52,1406.85 2121.38,1406.85 2122.23,1406.85 2123.08,1406.85 2123.94,1406.85 2124.79,1406.85 2125.65,1406.85 2126.5,1406.85 2127.35,1406.85 2128.21,1406.85 2129.06,1406.85 2129.92,1406.85 2130.77,1406.85 2131.62,1406.85 2132.48,1406.85 2133.33,1406.85 2134.19,1406.85 2135.04,1406.85 2135.89,1406.85 2136.75,1406.85 2137.6,1406.85 2138.45,1406.85 2139.31,1406.85 2140.16,1406.85 2141.02,1406.85 2141.87,1406.85 2142.72,1406.85 2143.58,1406.85 2144.43,1406.85 2145.29,1406.85 2146.14,1406.85 2146.99,1406.85 2147.85,1406.85 2148.7,1406.85 2149.56,1406.85 2150.41,1406.85 2151.26,1406.85 2152.12,1406.85 2152.97,1406.85 2153.82,1406.85 2154.68,1406.85 2155.53,1406.85 2156.39,1406.85 2157.24,1406.85 2158.09,1406.85 2158.95,1406.85 2159.8,1406.85 2160.66,1406.85 2161.51,1406.85 2162.36,1406.85 2163.22,1406.85 2164.07,1406.85 2164.93,1406.85 2165.78,1406.85 2166.63,1406.85 2167.49,1406.85 2168.34,1406.85 2169.19,1406.85 2170.05,1406.85 2170.9,1406.85 2171.76,1406.85 2172.61,1406.85 2173.46,1406.85 2174.32,1406.85 2175.17,1406.85 2176.03,1406.85 2176.88,1406.85 2177.73,1406.85 2178.59,1406.85 2179.44,1406.85 2180.3,1406.85 2181.15,1406.85 2182,1406.85 2182.86,1406.85 2183.71,1406.85 2184.56,1406.85 2185.42,1406.85 2186.27,1406.85 2187.13,1406.85 2187.98,1406.85 2188.83,1406.85 2189.69,1406.85 2190.54,1406.85 2191.4,1406.85 2192.25,1406.85 2193.1,1406.85 2193.96,1406.85 2194.81,1406.85 2195.67,1406.85 2196.52,1406.85 2197.37,1406.85 2198.23,1406.85 2199.08,1406.85 2199.93,1406.85 2200.79,1406.85 2201.64,1406.85 2202.5,1406.85 2203.35,1406.85 2204.2,1406.85 2205.06,1406.85 2205.91,1406.85 2206.77,1406.85 2207.62,1406.85 2208.47,1406.85 2209.33,1406.85 2210.18,1406.85 2211.04,1406.85 2211.89,1406.85 2212.74,1406.85 2213.6,1406.85 2214.45,1406.85 2215.3,1406.85 2216.16,1406.85 2217.01,1406.85 2217.87,1406.85 2218.72,1406.85 2219.57,1406.85 2220.43,1406.85 2221.28,1406.85 2222.14,1406.85 2222.99,1406.85 2223.84,1406.85 2224.7,1406.85 2225.55,1406.85 2226.41,1406.85 2227.26,1406.85 2228.11,1406.85 2228.97,1406.85 2229.82,1406.85 2230.67,1406.85 2231.53,1406.85 2232.38,1406.85 2233.24,1406.85 2234.09,1406.85 2234.94,1406.85 2235.8,1406.85 2236.65,1406.85 2237.51,1406.85 2238.36,1406.85 2239.21,1406.85 2240.07,1406.85 2240.92,1406.85 2241.78,1406.85 2242.63,1406.85 2243.48,1406.85 2244.34,1406.85 2245.19,1406.85 2246.04,1406.85 2246.9,1406.85 2247.75,1406.85 2248.61,1406.85 2249.46,1406.85 2250.31,1406.85 2251.17,1406.85 2252.02,1406.85 2252.88,1406.85 2253.73,1406.85 2254.58,1406.85 2255.44,1406.85 2256.29,1406.85 2257.15,1406.85 2258,1406.85 2258.85,1406.85 2259.71,1406.85 2260.56,1406.85 2261.42,1406.85 2262.27,1406.85 2263.12,1406.85 2263.98,1406.85 2264.83,1406.85 2265.68,1406.85 2266.54,1406.85 2267.39,1406.85 2268.25,1406.85 2269.1,1406.85 2269.95,1406.85 2270.81,1406.85 2271.66,1406.85 2272.52,1406.85 2273.37,1406.85 2274.22,1406.85 2275.08,1406.85 2275.93,1406.85 2276.79,1406.85 2277.64,1406.85 2278.49,1406.85 2279.35,1406.85 2280.2,1406.85 2281.05,1406.85 2281.91,1406.85 2282.76,1406.85 2283.62,1406.84 2284.47,1406.84 2285.32,1406.84 2286.18,1406.84 2287.03,1406.84 2287.89,1406.84 2288.74,1406.84 2289.59,1406.84 2290.45,1406.84 2291.3,1406.84 2292.16,1406.84 2293.01,1406.84 2293.86,1406.84 2294.72,1406.84 2295.57,1406.84 2296.42,1406.84 2297.28,1406.84 2298.13,1406.84 2298.99,1406.84 2299.84,1406.84 2300.69,1406.84 2301.55,1406.84 2302.4,1406.84 2303.26,1406.84 2304.11,1406.84 2304.96,1406.84 2305.82,1406.84 2306.67,1406.83 2307.53,1406.83 2308.38,1406.83 2309.23,1406.83 2310.09,1406.83 2310.94,1406.83 2311.79,1406.83 2312.65,1406.83 2313.5,1406.83 2314.36,1406.83 2315.21,1406.82 2316.06,1406.82 2316.92,1406.82 2317.77,1406.82 2318.63,1406.81 2319.48,1406.81 2320.33,1406.81 2321.19,1406.8 2322.04,1406.8 2322.9,1406.79 2323.75,1406.78 2324.6,1406.76 2325.46,1406.73 2326.31,1406.63 2327.16,1406.88 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/nonlinear/rocket_control/index.html b/previews/PR3525/tutorials/nonlinear/rocket_control/index.html
new file mode 100644
index 00000000000..ff2dc6f5977
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/rocket_control/index.html
@@ -0,0 +1,67 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Rocket Control · JuMP</title><meta name="title" content="Rocket Control · JuMP"/><meta property="og:title" content="Rocket Control · JuMP"/><meta property="twitter:title" content="Rocket Control · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li class="is-active"><a class="tocitem" href>Rocket Control</a><ul class="internal"><li><a class="tocitem" href="#Overview"><span>Overview</span></a></li><li><a class="tocitem" href="#Data"><span>Data</span></a></li><li><a class="tocitem" href="#JuMP-formulation"><span>JuMP formulation</span></a></li><li><a class="tocitem" href="#Next-steps"><span>Next steps</span></a></li></ul></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Rocket Control</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Rocket Control</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/rocket_control.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Rocket-Control"><a class="docs-heading-anchor" href="#Rocket-Control">Rocket Control</a><a id="Rocket-Control-1"></a><a class="docs-heading-anchor-permalink" href="#Rocket-Control" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../rocket_control.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Iain Dunning.</strong></p><p>The purpose of this tutorial is to demonstrate how to setup and solve a nonlinear optimization problem.</p><p>The example is an optimal control problem of a nonlinear rocket.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import Ipopt
+import Plots</code></pre><h2 id="Overview"><a class="docs-heading-anchor" href="#Overview">Overview</a><a id="Overview-1"></a><a class="docs-heading-anchor-permalink" href="#Overview" title="Permalink"></a></h2><p>Our goal is to maximize the final altitude of a vertically launched rocket.</p><p>We can control the thrust of the rocket, and must take account of the rocket mass, fuel consumption rate, gravity, and aerodynamic drag.</p><p>Let us consider the basic description of the model (for the full description, including parameters for the rocket, see <a href="https://www.mcs.anl.gov/~more/cops/cops3.pdf">COPS3</a>).</p><p>There are three state variables in our model:</p><ul><li>Velocity: <span>$x_v(t)$</span></li><li>Altitude: <span>$x_h(t)$</span></li><li>Mass of rocket and remaining fuel, <span>$x_m(t)$</span></li></ul><p>and a single control variable:</p><ul><li>Thrust: <span>$u_t(t)$</span>.</li></ul><p>There are three equations that control the dynamics of the rocket:</p><ul><li>Rate of ascent: <span>$\frac{d x_h}{dt} = x_v$</span></li><li>Acceleration: <span>$\frac{d x_v}{dt} = \frac{u_t - D(x_h, x_v)}{x_m} - g(x_h)$</span></li><li>Rate of mass loss: <span>$\frac{d x_m}{dt} = -\frac{u_t}{c}$</span></li></ul><p>where drag <span>$D(x_h, x_v)$</span> is a function of altitude and velocity, gravity <span>$g(x_h)$</span> is a function of altitude, and <span>$c$</span> is a constant.</p><p>These forces are defined as:</p><p class="math-container">\[D(x_h, x_v) = D_c \cdot x_v^2 \cdot e^{-h_c \left( \frac{x_h-x_h(0)}{x_h(0)} \right)}\]</p><p>and <span>$g(x_h) = g_0 \cdot \left( \frac{x_h(0)}{x_h} \right)^2$</span></p><p>We use a discretized model of time, with a fixed number of time steps, <span>$T$</span>.</p><p>Our goal is thus to maximize <span>$x_h(T)$</span>.</p><h2 id="Data"><a class="docs-heading-anchor" href="#Data">Data</a><a id="Data-1"></a><a class="docs-heading-anchor-permalink" href="#Data" title="Permalink"></a></h2><p>All parameters in this model have been normalized to be dimensionless, and they are taken from <a href="https://www.mcs.anl.gov/~more/cops/cops3.pdf">COPS3</a>.</p><pre><code class="language-julia hljs">h_0 = 1                      # Initial height
+v_0 = 0                      # Initial velocity
+m_0 = 1.0                    # Initial mass
+m_T = 0.6                    # Final mass
+g_0 = 1                      # Gravity at the surface
+h_c = 500                    # Used for drag
+c = 0.5 * sqrt(g_0 * h_0)    # Thrust-to-fuel mass
+D_c = 0.5 * 620 * m_0 / g_0  # Drag scaling
+u_t_max = 3.5 * g_0 * m_0    # Maximum thrust
+T_max = 0.2                  # Number of seconds
+T = 1_000                    # Number of time steps
+Δt = 0.2 / T;                # Time per discretized step</code></pre><h2 id="JuMP-formulation"><a class="docs-heading-anchor" href="#JuMP-formulation">JuMP formulation</a><a id="JuMP-formulation-1"></a><a class="docs-heading-anchor-permalink" href="#JuMP-formulation" title="Permalink"></a></h2><p>First, we create a model and choose an optimizer. Since this is a nonlinear program, we need to use a nonlinear solver like Ipopt. We cannot use a linear solver like HiGHS.</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+set_silent(model)</code></pre><p>Next, we create our state and control variables, which are each indexed by <code>t</code>. It is good practice for nonlinear programs to always provide a starting solution for each variable.</p><pre><code class="language-julia hljs">@variable(model, x_v[1:T] &gt;= 0, start = v_0)           # Velocity
+@variable(model, x_h[1:T] &gt;= 0, start = h_0)           # Height
+@variable(model, x_m[1:T] &gt;= m_T, start = m_0)         # Mass
+@variable(model, 0 &lt;= u_t[1:T] &lt;= u_t_max, start = 0); # Thrust</code></pre><p>We implement boundary conditions by fixing variables to values.</p><pre><code class="language-julia hljs">fix(x_v[1], v_0; force = true)
+fix(x_h[1], h_0; force = true)
+fix(x_m[1], m_0; force = true)
+fix(u_t[T], 0.0; force = true)</code></pre><p>The objective is to maximize altitude at end of time of flight.</p><pre><code class="language-julia hljs">@objective(model, Max, x_h[T])</code></pre><p>$ x_h_{1000} $</p><p>Forces are defined as functions:</p><pre><code class="language-julia hljs">D(x_h, x_v) = D_c * x_v^2 * exp(-h_c * (x_h - h_0) / h_0)
+g(x_h) = g_0 * (h_0 / x_h)^2</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">g (generic function with 1 method)</code></pre><p>The dynamical equations are implemented as constraints.</p><pre><code class="language-julia hljs">ddt(x::Vector, t::Int) = (x[t] - x[t-1]) / Δt
+@constraint(model, [t in 2:T], ddt(x_h, t) == x_v[t-1])
+@constraint(
+    model,
+    [t in 2:T],
+    ddt(x_v, t) == (u_t[t-1] - D(x_h[t-1], x_v[t-1])) / x_m[t-1] - g(x_h[t-1]),
+)
+@constraint(model, [t in 2:T], ddt(x_m, t) == -u_t[t-1] / c);</code></pre><p>Now we optimize the model and check that we found a solution:</p><pre><code class="language-julia hljs">optimize!(model)
+solution_summary(model)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Ipopt
+
+* Status
+  Result count       : 1
+  Termination status : LOCALLY_SOLVED
+  Message from the solver:
+  &quot;Solve_Succeeded&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 1.01278e+00
+  Dual objective value : 4.66547e+00
+
+* Work counters
+  Solve time (sec)   : 2.05758e-01
+</code></pre><p>Finally, we plot the solution:</p><pre><code class="language-julia hljs">function plot_trajectory(y; kwargs...)
+    return Plots.plot(
+        (1:T) * Δt,
+        value.(y);
+        xlabel = &quot;Time (s)&quot;,
+        legend = false,
+        kwargs...,
+    )
+end
+
+Plots.plot(
+    plot_trajectory(x_h; ylabel = &quot;Altitude&quot;),
+    plot_trajectory(x_m; ylabel = &quot;Mass&quot;),
+    plot_trajectory(x_v; ylabel = &quot;Velocity&quot;),
+    plot_trajectory(u_t; ylabel = &quot;Thrust&quot;);
+    layout = (2, 2),
+)</code></pre><img src="2e6d6b3d.svg" alt="Example block output"/><h2 id="Next-steps"><a class="docs-heading-anchor" href="#Next-steps">Next steps</a><a id="Next-steps-1"></a><a class="docs-heading-anchor-permalink" href="#Next-steps" title="Permalink"></a></h2><ul><li>Experiment with different values for the constants. How does the solution change? In particular, what happens if you change <code>T_max</code>?</li><li>The dynamical equations use rectangular integration for the right-hand side terms. Modify the equations to use the <a href="https://en.wikipedia.org/wiki/Trapezoidal_rule_(differential_equations)">Trapezoidal rule</a> instead. (As an example, <code>x_v[t-1]</code> would become <code>0.5 * (x_v[t-1] + x_v[t])</code>.) Is there a difference?</li></ul></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../portfolio/">« Portfolio optimization</a><a class="docs-footer-nextpage" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/simple_examples.jl b/previews/PR3525/tutorials/nonlinear/simple_examples.jl
new file mode 100644
index 00000000000..3a8ae666cba
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/simple_examples.jl
@@ -0,0 +1,168 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # Simple examples
+
+# This tutorial is a collection of examples of small nonlinear programs. It uses
+# the following packages:
+
+using JuMP
+import Ipopt
+import Random
+import Statistics
+import Test
+
+# ## The Rosenbrock function
+
+# A nonlinear example of the classical [Rosenbrock function](https://en.wikipedia.org/wiki/Rosenbrock_function).
+
+function example_rosenbrock()
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, x)
+    @variable(model, y)
+    @objective(model, Min, (1 - x)^2 + 100 * (y - x^2)^2)
+    optimize!(model)
+    Test.@test termination_status(model) == LOCALLY_SOLVED
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ 0.0 atol = 1e-10
+    Test.@test value(x) ≈ 1.0
+    Test.@test value(y) ≈ 1.0
+    return
+end
+
+example_rosenbrock()
+
+# ## The clnlbeam problem
+
+# Based on an AMPL model by Hande Y. Benson
+#
+# Copyright (C) 2001 Princeton University
+# All Rights Reserved
+
+# Permission to use, copy, modify, and distribute this software and its
+# documentation for any purpose and without fee is hereby granted, provided that
+# the above copyright notice appear in all copies and that the copyright notice
+# and this permission notice appear in all supporting documentation.
+#
+# Source:
+#
+# H. Maurer and H.D. Mittelman, "The non-linear beam via optimal control
+# with bound state variables," Optimal Control Applications and Methods 12, pp.
+# 19-31, 1991.
+
+function example_clnlbeam()
+    N = 1000
+    h = 1 / N
+    alpha = 350
+    model = Model(Ipopt.Optimizer)
+    @variables(model, begin
+        -1 <= t[1:(N+1)] <= 1
+        -0.05 <= x[1:(N+1)] <= 0.05
+        u[1:(N+1)]
+    end)
+    @objective(
+        model,
+        Min,
+        sum(
+            0.5 * h * (u[i+1]^2 + u[i]^2) +
+            0.5 * alpha * h * (cos(t[i+1]) + cos(t[i])) for i in 1:N
+        ),
+    )
+    @constraint(
+        model,
+        [i = 1:N],
+        x[i+1] - x[i] - 0.5 * h * (sin(t[i+1]) + sin(t[i])) == 0,
+    )
+    @constraint(
+        model,
+        [i = 1:N],
+        t[i+1] - t[i] - 0.5 * h * u[i+1] - 0.5 * h * u[i] == 0,
+    )
+    optimize!(model)
+    println("""
+    termination_status = $(termination_status(model))
+    primal_status      = $(primal_status(model))
+    objective_value    = $(objective_value(model))
+    """)
+    Test.@test termination_status(model) == LOCALLY_SOLVED  #src
+    Test.@test primal_status(model) == FEASIBLE_POINT  #src
+    Test.@test objective_value(model) ≈ 350.0  #src
+    return
+end
+
+example_clnlbeam()
+
+# ## Maximum likelihood estimation
+
+# This example uses nonlinear optimization to compute the maximum likelihood
+# estimate (MLE) of the parameters of a normal distribution, a.k.a., the sample
+# mean and variance.
+
+function example_mle()
+    n = 1_000
+    Random.seed!(1234)
+    data = randn(n)
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, μ, start = 0.0)
+    @variable(model, σ >= 0.0, start = 1.0)
+    @objective(
+        model,
+        Max,
+        n / 2 * log(1 / (2 * π * σ^2)) -
+        sum((data[i] - μ)^2 for i in 1:n) / (2 * σ^2)
+    )
+    optimize!(model)
+    println("μ             = ", value(μ))
+    println("mean(data)    = ", Statistics.mean(data))
+    println("σ^2           = ", value(σ)^2)
+    println("var(data)     = ", Statistics.var(data))
+    println("MLE objective = ", objective_value(model))
+    Test.@test value(μ) ≈ Statistics.mean(data) atol = 1e-3
+    Test.@test value(σ)^2 ≈ Statistics.var(data) atol = 1e-2
+    ## You can even do constrained MLE!
+    @constraint(model, μ == σ^2)
+    optimize!(model)
+    Test.@test value(μ) ≈ value(σ)^2
+    println()
+    println("With constraint μ == σ^2:")
+    println("μ                         = ", value(μ))
+    println("σ^2                       = ", value(σ)^2)
+    println("Constrained MLE objective = ", objective_value(model))
+    return
+end
+
+example_mle()
+
+# ## Quadratically constrained programs
+
+# A simple quadratically constrained program based on an
+# [example from Gurobi](https://www.gurobi.com/documentation/9.0/examples/qcp_c_c.html).
+
+function example_qcp()
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, x)
+    @variable(model, y >= 0)
+    @variable(model, z >= 0)
+    @objective(model, Max, x)
+    @constraint(model, x + y + z == 1)
+    @constraint(model, x * x + y * y - z * z <= 0)
+    @constraint(model, x * x - y * z <= 0)
+    optimize!(model)
+    print(model)
+    println("Objective value: ", objective_value(model))
+    println("x = ", value(x))
+    println("y = ", value(y))
+    Test.@test termination_status(model) == LOCALLY_SOLVED
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ 0.32699 atol = 1e-5
+    Test.@test value(x) ≈ 0.32699 atol = 1e-5
+    Test.@test value(y) ≈ 0.25707 atol = 1e-5
+    return
+end
+
+example_qcp()
diff --git a/previews/PR3525/tutorials/nonlinear/simple_examples/index.html b/previews/PR3525/tutorials/nonlinear/simple_examples/index.html
new file mode 100644
index 00000000000..bdd75a9016f
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/simple_examples/index.html
@@ -0,0 +1,183 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Simple examples · JuMP</title><meta name="title" content="Simple examples · JuMP"/><meta property="og:title" content="Simple examples · JuMP"/><meta property="twitter:title" content="Simple examples · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li class="is-active"><a class="tocitem" href>Simple examples</a><ul class="internal"><li><a class="tocitem" href="#The-Rosenbrock-function"><span>The Rosenbrock function</span></a></li><li><a class="tocitem" href="#The-clnlbeam-problem"><span>The clnlbeam problem</span></a></li><li><a class="tocitem" href="#Maximum-likelihood-estimation"><span>Maximum likelihood estimation</span></a></li><li><a class="tocitem" href="#Quadratically-constrained-programs"><span>Quadratically constrained programs</span></a></li></ul></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Simple examples</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Simple examples</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/simple_examples.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Simple-examples"><a class="docs-heading-anchor" href="#Simple-examples">Simple examples</a><a id="Simple-examples-1"></a><a class="docs-heading-anchor-permalink" href="#Simple-examples" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../simple_examples.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This tutorial is a collection of examples of small nonlinear programs. It uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import Ipopt
+import Random
+import Statistics
+import Test</code></pre><h2 id="The-Rosenbrock-function"><a class="docs-heading-anchor" href="#The-Rosenbrock-function">The Rosenbrock function</a><a id="The-Rosenbrock-function-1"></a><a class="docs-heading-anchor-permalink" href="#The-Rosenbrock-function" title="Permalink"></a></h2><p>A nonlinear example of the classical <a href="https://en.wikipedia.org/wiki/Rosenbrock_function">Rosenbrock function</a>.</p><pre><code class="language-julia hljs">function example_rosenbrock()
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, x)
+    @variable(model, y)
+    @objective(model, Min, (1 - x)^2 + 100 * (y - x^2)^2)
+    optimize!(model)
+    Test.@test termination_status(model) == LOCALLY_SOLVED
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ 0.0 atol = 1e-10
+    Test.@test value(x) ≈ 1.0
+    Test.@test value(y) ≈ 1.0
+    return
+end
+
+example_rosenbrock()</code></pre><h2 id="The-clnlbeam-problem"><a class="docs-heading-anchor" href="#The-clnlbeam-problem">The clnlbeam problem</a><a id="The-clnlbeam-problem-1"></a><a class="docs-heading-anchor-permalink" href="#The-clnlbeam-problem" title="Permalink"></a></h2><p>Based on an AMPL model by Hande Y. Benson</p><p>Copyright (C) 2001 Princeton University All Rights Reserved</p><p>Permission to use, copy, modify, and distribute this software and its documentation for any purpose and without fee is hereby granted, provided that the above copyright notice appear in all copies and that the copyright notice and this permission notice appear in all supporting documentation.</p><p>Source:</p><p>H. Maurer and H.D. Mittelman, &quot;The non-linear beam via optimal control with bound state variables,&quot; Optimal Control Applications and Methods 12, pp. 19-31, 1991.</p><pre><code class="language-julia hljs">function example_clnlbeam()
+    N = 1000
+    h = 1 / N
+    alpha = 350
+    model = Model(Ipopt.Optimizer)
+    @variables(model, begin
+        -1 &lt;= t[1:(N+1)] &lt;= 1
+        -0.05 &lt;= x[1:(N+1)] &lt;= 0.05
+        u[1:(N+1)]
+    end)
+    @objective(
+        model,
+        Min,
+        sum(
+            0.5 * h * (u[i+1]^2 + u[i]^2) +
+            0.5 * alpha * h * (cos(t[i+1]) + cos(t[i])) for i in 1:N
+        ),
+    )
+    @constraint(
+        model,
+        [i = 1:N],
+        x[i+1] - x[i] - 0.5 * h * (sin(t[i+1]) + sin(t[i])) == 0,
+    )
+    @constraint(
+        model,
+        [i = 1:N],
+        t[i+1] - t[i] - 0.5 * h * u[i+1] - 0.5 * h * u[i] == 0,
+    )
+    optimize!(model)
+    println(&quot;&quot;&quot;
+    termination_status = $(termination_status(model))
+    primal_status      = $(primal_status(model))
+    objective_value    = $(objective_value(model))
+    &quot;&quot;&quot;)
+    return
+end
+
+example_clnlbeam()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">This is Ipopt version 3.14.13, running with linear solver MUMPS 5.6.1.
+
+Number of nonzeros in equality constraint Jacobian...:     8000
+Number of nonzeros in inequality constraint Jacobian.:        0
+Number of nonzeros in Lagrangian Hessian.............:     4002
+
+Total number of variables............................:     3003
+                     variables with only lower bounds:        0
+                variables with lower and upper bounds:     2002
+                     variables with only upper bounds:        0
+Total number of equality constraints.................:     2000
+Total number of inequality constraints...............:        0
+        inequality constraints with only lower bounds:        0
+   inequality constraints with lower and upper bounds:        0
+        inequality constraints with only upper bounds:        0
+
+iter    objective    inf_pr   inf_du lg(mu)  ||d||  lg(rg) alpha_du alpha_pr  ls
+   0  3.5000000e+02 0.00e+00 0.00e+00  -1.0 0.00e+00    -  0.00e+00 0.00e+00   0
+   1  3.5000000e+02 0.00e+00 0.00e+00  -1.7 0.00e+00    -  1.00e+00 1.00e+00   0
+   2  3.5000000e+02 0.00e+00 0.00e+00  -3.8 0.00e+00  -2.0 1.00e+00 1.00e+00T  0
+   3  3.5000000e+02 0.00e+00 0.00e+00  -5.7 0.00e+00   0.2 1.00e+00 1.00e+00T  0
+   4  3.5000000e+02 0.00e+00 0.00e+00  -8.6 0.00e+00  -0.2 1.00e+00 1.00e+00T  0
+
+Number of Iterations....: 4
+
+                                   (scaled)                 (unscaled)
+Objective...............:   3.5000000000000318e+02    3.5000000000000318e+02
+Dual infeasibility......:   0.0000000000000000e+00    0.0000000000000000e+00
+Constraint violation....:   0.0000000000000000e+00    0.0000000000000000e+00
+Variable bound violation:   0.0000000000000000e+00    0.0000000000000000e+00
+Complementarity.........:   2.5059035596802450e-09    2.5059035596802450e-09
+Overall NLP error.......:   2.5059035596802450e-09    2.5059035596802450e-09
+
+
+Number of objective function evaluations             = 5
+Number of objective gradient evaluations             = 5
+Number of equality constraint evaluations            = 5
+Number of inequality constraint evaluations          = 0
+Number of equality constraint Jacobian evaluations   = 5
+Number of inequality constraint Jacobian evaluations = 0
+Number of Lagrangian Hessian evaluations             = 4
+Total seconds in IPOPT                               = 0.037
+
+EXIT: Optimal Solution Found.
+termination_status = LOCALLY_SOLVED
+primal_status      = FEASIBLE_POINT
+objective_value    = 350.0000000000032</code></pre><h2 id="Maximum-likelihood-estimation"><a class="docs-heading-anchor" href="#Maximum-likelihood-estimation">Maximum likelihood estimation</a><a id="Maximum-likelihood-estimation-1"></a><a class="docs-heading-anchor-permalink" href="#Maximum-likelihood-estimation" title="Permalink"></a></h2><p>This example uses nonlinear optimization to compute the maximum likelihood estimate (MLE) of the parameters of a normal distribution, a.k.a., the sample mean and variance.</p><pre><code class="language-julia hljs">function example_mle()
+    n = 1_000
+    Random.seed!(1234)
+    data = randn(n)
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, μ, start = 0.0)
+    @variable(model, σ &gt;= 0.0, start = 1.0)
+    @objective(
+        model,
+        Max,
+        n / 2 * log(1 / (2 * π * σ^2)) -
+        sum((data[i] - μ)^2 for i in 1:n) / (2 * σ^2)
+    )
+    optimize!(model)
+    println(&quot;μ             = &quot;, value(μ))
+    println(&quot;mean(data)    = &quot;, Statistics.mean(data))
+    println(&quot;σ^2           = &quot;, value(σ)^2)
+    println(&quot;var(data)     = &quot;, Statistics.var(data))
+    println(&quot;MLE objective = &quot;, objective_value(model))
+    Test.@test value(μ) ≈ Statistics.mean(data) atol = 1e-3
+    Test.@test value(σ)^2 ≈ Statistics.var(data) atol = 1e-2
+    # You can even do constrained MLE!
+    @constraint(model, μ == σ^2)
+    optimize!(model)
+    Test.@test value(μ) ≈ value(σ)^2
+    println()
+    println(&quot;With constraint μ == σ^2:&quot;)
+    println(&quot;μ                         = &quot;, value(μ))
+    println(&quot;σ^2                       = &quot;, value(σ)^2)
+    println(&quot;Constrained MLE objective = &quot;, objective_value(model))
+    return
+end
+
+example_mle()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">μ             = -0.0215521734290741
+mean(data)    = -0.021552173429074114
+σ^2           = 1.100101397871862
+var(data)     = 1.1012026004695599
+MLE objective = -1466.6397109231782
+
+With constraint μ == σ^2:
+μ                         = 0.6621385003734601
+σ^2                       = 0.66213850037346
+Constrained MLE objective = -1896.4889420749978</code></pre><h2 id="Quadratically-constrained-programs"><a class="docs-heading-anchor" href="#Quadratically-constrained-programs">Quadratically constrained programs</a><a id="Quadratically-constrained-programs-1"></a><a class="docs-heading-anchor-permalink" href="#Quadratically-constrained-programs" title="Permalink"></a></h2><p>A simple quadratically constrained program based on an <a href="https://www.gurobi.com/documentation/9.0/examples/qcp_c_c.html">example from Gurobi</a>.</p><pre><code class="language-julia hljs">function example_qcp()
+    model = Model(Ipopt.Optimizer)
+    set_silent(model)
+    @variable(model, x)
+    @variable(model, y &gt;= 0)
+    @variable(model, z &gt;= 0)
+    @objective(model, Max, x)
+    @constraint(model, x + y + z == 1)
+    @constraint(model, x * x + y * y - z * z &lt;= 0)
+    @constraint(model, x * x - y * z &lt;= 0)
+    optimize!(model)
+    print(model)
+    println(&quot;Objective value: &quot;, objective_value(model))
+    println(&quot;x = &quot;, value(x))
+    println(&quot;y = &quot;, value(y))
+    Test.@test termination_status(model) == LOCALLY_SOLVED
+    Test.@test primal_status(model) == FEASIBLE_POINT
+    Test.@test objective_value(model) ≈ 0.32699 atol = 1e-5
+    Test.@test value(x) ≈ 0.32699 atol = 1e-5
+    Test.@test value(y) ≈ 0.25707 atol = 1e-5
+    return
+end
+
+example_qcp()</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Max x
+Subject to
+ x + y + z = 1
+ x² + y² - z² ≤ 0
+ x² - y*z ≤ 0
+ y ≥ 0
+ z ≥ 0
+Objective value: 0.32699283491387243
+x = 0.32699283491387243
+y = 0.2570658388068964</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../introduction/">« Introduction</a><a class="docs-footer-nextpage" href="../portfolio/">Portfolio optimization »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory.jl b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory.jl
new file mode 100644
index 00000000000..638c7e1d8f0
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory.jl
@@ -0,0 +1,402 @@
+# MIT License                                                                     #src
+#                                                                                 #src
+# Copyright (c) 2020 Henrique Ferrolho                                            #src
+#                                                                                 #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy    #src
+# of this software and associated documentation files (the "Software"), to deal   #src
+# in the Software without restriction, including without limitation the rights    #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell       #src
+# copies of the Software, and to permit persons to whom the Software is           #src
+# furnished to do so, subject to the following conditions:                        #src
+#                                                                                 #src
+# The above copyright notice and this permission notice shall be included in all  #src
+# copies or substantial portions of the Software.                                 #src
+#                                                                                 #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR      #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,        #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE     #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER          #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,   #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE   #src
+# SOFTWARE.                                                                       #src
+
+# # Optimal control for a Space Shuttle reentry trajectory
+
+# **This tutorial was originally contributed by Henrique Ferrolho.**
+
+# This tutorial demonstrates how to compute a reentry trajectory for the
+# [Space Shuttle](https://en.wikipedia.org/wiki/Space_Shuttle),
+# by formulating and solving a nonlinear programming problem.
+# The problem was drawn from Chapter 6 of
+# ["Practical Methods for Optimal Control and Estimation Using Nonlinear Programming"](https://epubs.siam.org/doi/book/10.1137/1.9780898718577),
+# by John T. Betts.
+
+# !!! tip
+#     This tutorial is a more-complicated version of the [Rocket Control](@ref) example.
+#     If you are new to solving nonlinear programs in JuMP, you may want to start there instead.
+
+# The motion of the vehicle is defined by the following set of DAEs:
+#
+# ```math
+# \begin{aligned}
+# \dot{h} & = v \sin \gamma , \\
+# \dot{\phi} & = \frac{v}{r} \cos \gamma \sin \psi / \cos \theta , \\
+# \dot{\theta} & = \frac{v}{r} \cos \gamma \cos \psi , \\
+# \dot{v} & = -\frac{D}{m} - g \sin \gamma , \\
+# \dot{\gamma} & = \frac{L}{m v} \cos(\beta) + \cos \gamma \left ( \frac{v}{r} - \frac{g}{v} \right ) , \\
+# \dot{\psi} & = \frac{1}{m v \cos \gamma} L \sin(\beta) + \frac{v}{r \cos \theta} \cos \gamma \sin \psi \sin \theta , \\
+# q & \le q_U , \\
+# \end{aligned}
+# ```
+
+# where the aerodynamic heating on the vehicle wing leading edge is $q = q_a q_r$
+# and the dynamic variables are
+#
+# ```math
+# \begin{aligned}
+#      h & \quad \text{altitude (ft)},     \qquad & & \gamma \quad \text{flight path angle (rad)}, \\
+#   \phi & \quad \text{longitude (rad)},   \qquad & & \psi   \quad \text{azimuth (rad)},           \\
+# \theta & \quad \text{latitude (rad)},    \qquad & & \alpha \quad \text{angle of attack (rad)},   \\
+#      v & \quad \text{velocity (ft/sec)}, \qquad & & \beta  \quad \text{bank angle (rad)}.
+# \end{aligned}
+# ```
+
+# The aerodynamic and atmospheric forces on the vehicle are
+# specified by the following quantities (English units):
+#
+# ```math
+# \begin{aligned}
+#            D & = \frac{1}{2} c_D S \rho v^2,                  \qquad & a_0 & = -0.20704, \\
+#            L & = \frac{1}{2} c_L S \rho v^2,                  \qquad & a_1 & =  0.029244, \\
+#            g & = \mu / r^2,                                   \qquad & \mu & =  0.14076539 \times 10^{17}, \\
+#            r & = R_e + h,                                     \qquad & b_0 & =  0.07854, \\
+#         \rho & = \rho_0 \exp[-h/h_r],                         \qquad & b_1 & = -0.61592  \times 10^{-2}, \\
+#       \rho_0 & = 0.002378,                                    \qquad & b_2 & =  0.621408 \times 10^{-3}, \\
+#          h_r & = 23800,                                       \qquad & q_r & =  17700 \sqrt{\rho} (0.0001 v)^{3.07}, \\
+#          c_L & = a_0 + a_1 \hat{\alpha},                      \qquad & q_a & =  c_0 + c_1 \hat{\alpha} + c_2 \hat{\alpha}^2 + c_3 \hat{\alpha}^3, \\
+#          c_D & = b_0 + b_1 \hat{\alpha} + b_2 \hat{\alpha}^2, \qquad & c_0 & =  1.0672181, \\
+# \hat{\alpha} & = 180 \alpha / \pi,                            \qquad & c_1 & = -0.19213774 \times 10^{-1}, \\
+#          R_e & = 20902900,                                    \qquad & c_2 & =  0.21286289 \times 10^{-3}, \\
+#            S & = 2690,                                        \qquad & c_3 & = -0.10117249 \times 10^{-5}.
+# \end{aligned}
+# ```
+
+# The reentry trajectory begins at an altitude where the aerodynamic forces are quite small
+# with a weight of $w = 203000$ (lb) and mass $m = w / g_0$ (slug), where $g_0 = 32.174$ (ft/sec$^2$).
+# The initial conditions are as follows:
+#
+# ```math
+# \begin{aligned}
+#      h & = 260000 \text{ ft},  \qquad & v      & = 25600 \text{ ft/sec}, \\
+#   \phi & = 0      \text{ deg}, \qquad & \gamma & = -1    \text{ deg},    \\
+# \theta & = 0      \text{ deg}, \qquad & \psi   & = 90    \text{ deg}.
+# \end{aligned}
+# ```
+
+# The final point on the reentry trajectory occurs at the unknown (free) time $t_F$,
+# at the so-called terminal area energy management (TAEM) interface, which is defined by the conditions
+#
+# ```math
+# h = 80000 \text{ ft}, \qquad v = 2500 \text{ ft/sec}, \qquad \gamma = -5 \text{ deg}.
+# ```
+
+# As explained in the book, our goal is to maximize the final cross-range,
+# which is equivalent to maximizing the final latitude of the vehicle,
+# that is, $J = \theta(t_F)$.
+
+# ![Max cross-range shuttle reentry](../../assets/space_shuttle.png)
+
+# ### Approach
+# We will use a discretized model of time, with a fixed number of discretized points, $n$.
+# The decision variables at each point are going to be the state of the vehicle and the controls commanded to it.
+# In addition, we will also make each time step size $\Delta t$ a decision variable;
+# that way, we can either fix the time step size easily,
+# or allow the solver to fine-tune the duration between each adjacent pair of points.
+# Finally, in order to approximate the derivatives of the problem dynamics, we will use either rectangular
+# or [trapezoidal](http://en.wikipedia.org/wiki/Trapezoidal_rule) integration.
+
+# !!! warning
+#     Do not try to actually land a Space Shuttle using this notebook. There's
+#     no mesh refinement going on, which can lead to unrealistic trajectories
+#     having position and velocity errors with orders of magnitude $10^4$ ft and
+#     $10^2$ ft/sec, respectively.
+
+using JuMP
+import Interpolations
+import Ipopt
+
+## Global variables
+const w = 203000.0  # weight (lb)
+const g₀ = 32.174    # acceleration (ft/sec^2)
+const m = w / g₀    # mass (slug)
+
+## Aerodynamic and atmospheric forces on the vehicle
+const ρ₀ = 0.002378
+const hᵣ = 23800.0
+const Rₑ = 20902900.0
+const μ = 0.14076539e17
+const S = 2690.0
+const a₀ = -0.20704
+const a₁ = 0.029244
+const b₀ = 0.07854
+const b₁ = -0.61592e-2
+const b₂ = 0.621408e-3
+const c₀ = 1.0672181
+const c₁ = -0.19213774e-1
+const c₂ = 0.21286289e-3
+const c₃ = -0.10117249e-5
+
+## Initial conditions
+const h_s = 2.6          # altitude (ft) / 1e5
+const ϕ_s = deg2rad(0)   # longitude (rad)
+const θ_s = deg2rad(0)   # latitude (rad)
+const v_s = 2.56         # velocity (ft/sec) / 1e4
+const γ_s = deg2rad(-1)  # flight path angle (rad)
+const ψ_s = deg2rad(90)  # azimuth (rad)
+const α_s = deg2rad(0)   # angle of attack (rad)
+const β_s = deg2rad(0)   # bank angle (rad)
+const t_s = 1.00         # time step (sec)
+
+## Final conditions, the so-called Terminal Area Energy Management (TAEM)
+const h_t = 0.8          # altitude (ft) / 1e5
+const v_t = 0.25         # velocity (ft/sec) / 1e4
+const γ_t = deg2rad(-5)  # flight path angle (rad)
+
+## Number of mesh points (knots) to be used
+const n = 503
+
+## Integration scheme to be used for the dynamics
+const integration_rule = "rectangular"
+nothing #hide
+
+#-
+
+# !!! tip "Choose a good linear solver"
+#     Picking a good linear solver is **extremely important**
+#     to maximize the performance of nonlinear solvers.
+#     For the best results, it is advised to experiment different linear solvers.
+#
+#     For example, the linear solver `MA27` is outdated and can be quite slow.
+#     `MA57` is a much better alternative, especially for highly-sparse problems
+#     (such as trajectory optimization problems).
+
+## Uncomment the lines below to pass user options to the solver
+user_options = (
+## "mu_strategy" => "monotone",
+## "linear_solver" => "ma27",
+)
+
+## Create JuMP model, using Ipopt as the solver
+model = Model(optimizer_with_attributes(Ipopt.Optimizer, user_options...))
+
+@variables(model, begin
+    0 ≤ scaled_h[1:n]                # altitude (ft) / 1e5
+    ϕ[1:n]                # longitude (rad)
+    deg2rad(-89) ≤ θ[1:n] ≤ deg2rad(89)  # latitude (rad)
+    1e-4 ≤ scaled_v[1:n]                # velocity (ft/sec) / 1e4
+    deg2rad(-89) ≤ γ[1:n] ≤ deg2rad(89)  # flight path angle (rad)
+    ψ[1:n]                # azimuth (rad)
+    deg2rad(-90) ≤ α[1:n] ≤ deg2rad(90)  # angle of attack (rad)
+    deg2rad(-89) ≤ β[1:n] ≤ deg2rad(1)  # bank angle (rad)
+    ##        3.5 ≤       Δt[1:n] ≤ 4.5          # time step (sec)
+    Δt[1:n] == 4.0         # time step (sec)
+end);
+
+# !!! info
+#     Above you can find two alternatives for the `Δt` variables.
+#
+#     The first one, `3.5 ≤ Δt[1:n] ≤ 4.5` (currently commented), allows some wiggle room
+#     for the solver to adjust the time step size between pairs of mesh points. This is neat because
+#     it allows the solver to figure out which parts of the flight require more dense discretization than others.
+#     (Remember, the number of discretized points is fixed, and this example does not implement mesh refinement.)
+#     However, this makes the problem more complex to solve, and therefore leads to a longer computation time.
+#
+#     The second line, `Δt[1:n] == 4.0`, fixes the duration of every time step to exactly 4.0 seconds.
+#     This allows the problem to be solved faster. However, to do this we need to know beforehand that the
+#     close-to-optimal total duration of the flight is ~2009 seconds. Therefore, if we split the total duration
+#     in slices of 4.0 seconds, we know that we require `n = 503` knots to discretize the whole trajectory.
+
+## Fix initial conditions
+fix(scaled_h[1], h_s; force = true)
+fix(ϕ[1], ϕ_s; force = true)
+fix(θ[1], θ_s; force = true)
+fix(scaled_v[1], v_s; force = true)
+fix(γ[1], γ_s; force = true)
+fix(ψ[1], ψ_s; force = true)
+
+## Fix final conditions
+fix(scaled_h[n], h_t; force = true)
+fix(scaled_v[n], v_t; force = true)
+fix(γ[n], γ_t; force = true)
+
+## Initial guess: linear interpolation between boundary conditions
+x_s = [h_s, ϕ_s, θ_s, v_s, γ_s, ψ_s, α_s, β_s, t_s]
+x_t = [h_t, ϕ_s, θ_s, v_t, γ_t, ψ_s, α_s, β_s, t_s]
+interp_linear = Interpolations.LinearInterpolation([1, n], [x_s, x_t])
+initial_guess = mapreduce(transpose, vcat, interp_linear.(1:n))
+set_start_value.(all_variables(model), vec(initial_guess))
+
+## Functions to restore `h` and `v` to their true scale
+@expression(model, h[j = 1:n], scaled_h[j] * 1e5)
+@expression(model, v[j = 1:n], scaled_v[j] * 1e4)
+
+## Helper functions
+@expression(model, c_L[j = 1:n], a₀ + a₁ * rad2deg(α[j]))
+@expression(model, c_D[j = 1:n], b₀ + b₁ * rad2deg(α[j]) + b₂ * rad2deg(α[j])^2)
+@expression(model, ρ[j = 1:n], ρ₀ * exp(-h[j] / hᵣ))
+@expression(model, D[j = 1:n], 0.5 * c_D[j] * S * ρ[j] * v[j]^2)
+@expression(model, L[j = 1:n], 0.5 * c_L[j] * S * ρ[j] * v[j]^2)
+@expression(model, r[j = 1:n], Rₑ + h[j])
+@expression(model, g[j = 1:n], μ / r[j]^2)
+
+## Motion of the vehicle as a differential-algebraic system of equations (DAEs)
+@expression(model, δh[j = 1:n], v[j] * sin(γ[j]))
+@expression(
+    model,
+    δϕ[j = 1:n],
+    (v[j] / r[j]) * cos(γ[j]) * sin(ψ[j]) / cos(θ[j])
+)
+@expression(model, δθ[j = 1:n], (v[j] / r[j]) * cos(γ[j]) * cos(ψ[j]))
+@expression(model, δv[j = 1:n], -(D[j] / m) - g[j] * sin(γ[j]))
+@expression(
+    model,
+    δγ[j = 1:n],
+    (L[j] / (m * v[j])) * cos(β[j]) +
+    cos(γ[j]) * ((v[j] / r[j]) - (g[j] / v[j]))
+)
+@expression(
+    model,
+    δψ[j = 1:n],
+    (1 / (m * v[j] * cos(γ[j]))) * L[j] * sin(β[j]) +
+    (v[j] / (r[j] * cos(θ[j]))) * cos(γ[j]) * sin(ψ[j]) * sin(θ[j])
+)
+
+## System dynamics
+for j in 2:n
+    i = j - 1  # index of previous knot
+
+    if integration_rule == "rectangular"
+        ## Rectangular integration
+        @constraint(model, h[j] == h[i] + Δt[i] * δh[i])
+        @constraint(model, ϕ[j] == ϕ[i] + Δt[i] * δϕ[i])
+        @constraint(model, θ[j] == θ[i] + Δt[i] * δθ[i])
+        @constraint(model, v[j] == v[i] + Δt[i] * δv[i])
+        @constraint(model, γ[j] == γ[i] + Δt[i] * δγ[i])
+        @constraint(model, ψ[j] == ψ[i] + Δt[i] * δψ[i])
+    elseif integration_rule == "trapezoidal"
+        ## Trapezoidal integration
+        @constraint(model, h[j] == h[i] + 0.5 * Δt[i] * (δh[j] + δh[i]))
+        @constraint(model, ϕ[j] == ϕ[i] + 0.5 * Δt[i] * (δϕ[j] + δϕ[i]))
+        @constraint(model, θ[j] == θ[i] + 0.5 * Δt[i] * (δθ[j] + δθ[i]))
+        @constraint(model, v[j] == v[i] + 0.5 * Δt[i] * (δv[j] + δv[i]))
+        @constraint(model, γ[j] == γ[i] + 0.5 * Δt[i] * (δγ[j] + δγ[i]))
+        @constraint(model, ψ[j] == ψ[i] + 0.5 * Δt[i] * (δψ[j] + δψ[i]))
+    else
+        @error "Unexpected integration rule '$(integration_rule)'"
+    end
+end
+
+## Objective: Maximize cross-range
+@objective(model, Max, θ[n])
+
+set_silent(model)  # Hide solver's verbose output
+optimize!(model)  # Solve for the control and state
+@assert termination_status(model) == LOCALLY_SOLVED
+
+## Show final cross-range of the solution
+println(
+    "Final latitude θ = ",
+    round(objective_value(model) |> rad2deg; digits = 2),
+    "°",
+)
+
+# ### Plotting the results
+
+# Let's plot the results to visualize the optimal trajectory:
+
+using Plots
+ts = cumsum([0; value.(Δt)])[1:end-1]
+plt_altitude = plot(
+    ts,
+    value.(scaled_h);
+    legend = nothing,
+    title = "Altitude (100,000 ft)",
+)
+plt_longitude =
+    plot(ts, rad2deg.(value.(ϕ)); legend = nothing, title = "Longitude (deg)")
+plt_latitude =
+    plot(ts, rad2deg.(value.(θ)); legend = nothing, title = "Latitude (deg)")
+plt_velocity = plot(
+    ts,
+    value.(scaled_v);
+    legend = nothing,
+    title = "Velocity (1000 ft/sec)",
+)
+plt_flight_path =
+    plot(ts, rad2deg.(value.(γ)); legend = nothing, title = "Flight Path (deg)")
+plt_azimuth =
+    plot(ts, rad2deg.(value.(ψ)); legend = nothing, title = "Azimuth (deg)")
+
+plot(
+    plt_altitude,
+    plt_velocity,
+    plt_longitude,
+    plt_flight_path,
+    plt_latitude,
+    plt_azimuth;
+    layout = grid(3, 2),
+    linewidth = 2,
+    size = (700, 700),
+)
+
+#-
+
+function q(h, v, a)
+    ρ(h) = ρ₀ * exp(-h / hᵣ)
+    qᵣ(h, v) = 17700 * √ρ(h) * (0.0001 * v)^3.07
+    qₐ(a) = c₀ + c₁ * rad2deg(a) + c₂ * rad2deg(a)^2 + c₃ * rad2deg(a)^3
+    ## Aerodynamic heating on the vehicle wing leading edge
+    return qₐ(a) * qᵣ(h, v)
+end
+
+plt_attack_angle = plot(
+    ts[1:end-1],
+    rad2deg.(value.(α)[1:end-1]);
+    legend = nothing,
+    title = "Angle of Attack (deg)",
+)
+plt_bank_angle = plot(
+    ts[1:end-1],
+    rad2deg.(value.(β)[1:end-1]);
+    legend = nothing,
+    title = "Bank Angle (deg)",
+)
+plt_heating = plot(
+    ts,
+    q.(value.(scaled_h) * 1e5, value.(scaled_v) * 1e4, value.(α));
+    legend = nothing,
+    title = "Heating (BTU/ft/ft/sec)",
+)
+
+plot(
+    plt_attack_angle,
+    plt_bank_angle,
+    plt_heating;
+    layout = grid(3, 1),
+    linewidth = 2,
+    size = (700, 700),
+)
+
+#-
+
+plot(
+    rad2deg.(value.(ϕ)),
+    rad2deg.(value.(θ)),
+    value.(scaled_h);
+    linewidth = 2,
+    legend = nothing,
+    title = "Space Shuttle Reentry Trajectory",
+    xlabel = "Longitude (deg)",
+    ylabel = "Latitude (deg)",
+    zlabel = "Altitude (100,000 ft)",
+)
diff --git a/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/0d7a8519.svg b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/0d7a8519.svg
new file mode 100644
index 00000000000..cf38d8b6b43
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/0d7a8519.svg
@@ -0,0 +1,48 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="600" height="400" viewBox="0 0 2400 1600">
+<defs>
+  <clipPath id="clip750">
+    <rect x="0" y="0" width="2400" height="1600"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip750)" d="M0 1600 L2400 1600 L2400 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip751">
+    <rect x="480" y="0" width="1681" height="1600"/>
+  </clipPath>
+</defs>
+<defs>
+  <clipPath id="clip752">
+    <rect x="219" y="123" width="2135" height="1301"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip752)" d="M266.588 1408.12 L266.588 1372.25 L275.983 1354.31 L292.255 1364.67 L292.255 1400.54 L282.86 1418.48 L266.588 1408.12  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="791.763,1172.23 1166.96,847.3 1166.96,197.447 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="954.483,1219.2 1329.68,894.274 1329.68,244.42 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1117.2,1266.17 1492.4,941.247 1492.4,291.393 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1279.92,1313.15 1655.12,988.22 1655.12,338.367 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="773.371,1166.92 1423.22,1354.51 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="791.763,1172.23 796.265,1168.33 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="954.483,1219.2 958.986,1215.3 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1117.2,1266.17 1121.71,1262.27 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1279.92,1313.15 1284.43,1309.25 "/>
+<path clip-path="url(#clip750)" d="M764.355 1182.32 Q760.744 1182.32 758.916 1185.89 Q757.11 1189.43 757.11 1196.56 Q757.11 1203.67 758.916 1207.23 Q760.744 1210.77 764.355 1210.77 Q767.99 1210.77 769.795 1207.23 Q771.624 1203.67 771.624 1196.56 Q771.624 1189.43 769.795 1185.89 Q767.99 1182.32 764.355 1182.32 M764.355 1178.62 Q770.166 1178.62 773.221 1183.23 Q776.3 1187.81 776.3 1196.56 Q776.3 1205.29 773.221 1209.89 Q770.166 1214.48 764.355 1214.48 Q758.545 1214.48 755.467 1209.89 Q752.411 1205.29 752.411 1196.56 Q752.411 1187.81 755.467 1183.23 Q758.545 1178.62 764.355 1178.62 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M890.942 1256.84 L907.261 1256.84 L907.261 1260.78 L885.317 1260.78 L885.317 1256.84 Q887.979 1254.09 892.562 1249.46 Q897.169 1244.81 898.349 1243.46 Q900.594 1240.94 901.474 1239.2 Q902.377 1237.44 902.377 1235.75 Q902.377 1233 900.432 1231.26 Q898.511 1229.53 895.409 1229.53 Q893.21 1229.53 890.757 1230.29 Q888.326 1231.06 885.548 1232.61 L885.548 1227.88 Q888.372 1226.75 890.826 1226.17 Q893.28 1225.59 895.317 1225.59 Q900.687 1225.59 903.881 1228.28 Q907.076 1230.96 907.076 1235.45 Q907.076 1237.58 906.266 1239.5 Q905.479 1241.4 903.372 1244 Q902.793 1244.67 899.692 1247.88 Q896.59 1251.08 890.942 1256.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M927.076 1229.3 Q923.465 1229.3 921.636 1232.86 Q919.83 1236.4 919.83 1243.53 Q919.83 1250.64 921.636 1254.2 Q923.465 1257.75 927.076 1257.75 Q930.71 1257.75 932.516 1254.2 Q934.344 1250.64 934.344 1243.53 Q934.344 1236.4 932.516 1232.86 Q930.71 1229.3 927.076 1229.3 M927.076 1225.59 Q932.886 1225.59 935.941 1230.2 Q939.02 1234.78 939.02 1243.53 Q939.02 1252.26 935.941 1256.87 Q932.886 1261.45 927.076 1261.45 Q921.266 1261.45 918.187 1256.87 Q915.131 1252.26 915.131 1243.53 Q915.131 1234.78 918.187 1230.2 Q921.266 1225.59 927.076 1225.59 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1062.48 1277.26 L1050.68 1295.71 L1062.48 1295.71 L1062.48 1277.26 M1061.25 1273.19 L1067.13 1273.19 L1067.13 1295.71 L1072.06 1295.71 L1072.06 1299.6 L1067.13 1299.6 L1067.13 1307.75 L1062.48 1307.75 L1062.48 1299.6 L1046.88 1299.6 L1046.88 1295.09 L1061.25 1273.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1089.8 1276.27 Q1086.18 1276.27 1084.36 1279.83 Q1082.55 1283.38 1082.55 1290.51 Q1082.55 1297.61 1084.36 1301.18 Q1086.18 1304.72 1089.8 1304.72 Q1093.43 1304.72 1095.24 1301.18 Q1097.06 1297.61 1097.06 1290.51 Q1097.06 1283.38 1095.24 1279.83 Q1093.43 1276.27 1089.8 1276.27 M1089.8 1272.57 Q1095.61 1272.57 1098.66 1277.17 Q1101.74 1281.76 1101.74 1290.51 Q1101.74 1299.23 1098.66 1303.84 Q1095.61 1308.42 1089.8 1308.42 Q1083.99 1308.42 1080.91 1303.84 Q1077.85 1299.23 1077.85 1290.51 Q1077.85 1281.76 1080.91 1277.17 Q1083.99 1272.57 1089.8 1272.57 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1222.93 1335.58 Q1219.79 1335.58 1217.93 1337.73 Q1216.1 1339.89 1216.1 1343.64 Q1216.1 1347.36 1217.93 1349.54 Q1219.79 1351.69 1222.93 1351.69 Q1226.08 1351.69 1227.91 1349.54 Q1229.76 1347.36 1229.76 1343.64 Q1229.76 1339.89 1227.91 1337.73 Q1226.08 1335.58 1222.93 1335.58 M1232.22 1320.93 L1232.22 1325.19 Q1230.46 1324.35 1228.65 1323.91 Q1226.87 1323.47 1225.11 1323.47 Q1220.48 1323.47 1218.03 1326.6 Q1215.6 1329.72 1215.25 1336.04 Q1216.61 1334.03 1218.67 1332.96 Q1220.73 1331.88 1223.21 1331.88 Q1228.42 1331.88 1231.43 1335.05 Q1234.46 1338.2 1234.46 1343.64 Q1234.46 1348.96 1231.31 1352.18 Q1228.16 1355.4 1222.93 1355.4 Q1216.94 1355.4 1213.77 1350.81 Q1210.6 1346.21 1210.6 1337.48 Q1210.6 1329.28 1214.48 1324.42 Q1218.37 1319.54 1224.92 1319.54 Q1226.68 1319.54 1228.47 1319.89 Q1230.27 1320.23 1232.22 1320.93 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1252.52 1323.24 Q1248.91 1323.24 1247.08 1326.81 Q1245.27 1330.35 1245.27 1337.48 Q1245.27 1344.59 1247.08 1348.15 Q1248.91 1351.69 1252.52 1351.69 Q1256.15 1351.69 1257.96 1348.15 Q1259.78 1344.59 1259.78 1337.48 Q1259.78 1330.35 1257.96 1326.81 Q1256.15 1323.24 1252.52 1323.24 M1252.52 1319.54 Q1258.33 1319.54 1261.38 1324.15 Q1264.46 1328.73 1264.46 1337.48 Q1264.46 1346.21 1261.38 1350.81 Q1258.33 1355.4 1252.52 1355.4 Q1246.71 1355.4 1243.63 1350.81 Q1240.57 1346.21 1240.57 1337.48 Q1240.57 1328.73 1243.63 1324.15 Q1246.71 1319.54 1252.52 1319.54 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M721.854 1393.34 L728.284 1393.34 L728.284 1435.45 L751.423 1435.45 L751.423 1440.86 L721.854 1440.86 L721.854 1393.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M770.584 1409.32 Q765.873 1409.32 763.136 1413.01 Q760.399 1416.67 760.399 1423.07 Q760.399 1429.47 763.104 1433.16 Q765.842 1436.82 770.584 1436.82 Q775.263 1436.82 778 1433.13 Q780.737 1429.43 780.737 1423.07 Q780.737 1416.73 778 1413.04 Q775.263 1409.32 770.584 1409.32 M770.584 1404.35 Q778.223 1404.35 782.583 1409.32 Q786.944 1414.28 786.944 1423.07 Q786.944 1431.82 782.583 1436.82 Q778.223 1441.78 770.584 1441.78 Q762.913 1441.78 758.553 1436.82 Q754.224 1431.82 754.224 1423.07 Q754.224 1414.28 758.553 1409.32 Q762.913 1404.35 770.584 1404.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M826.284 1419.34 L826.284 1440.86 L820.427 1440.86 L820.427 1419.54 Q820.427 1414.47 818.454 1411.96 Q816.481 1409.45 812.534 1409.45 Q807.792 1409.45 805.054 1412.47 Q802.317 1415.49 802.317 1420.71 L802.317 1440.86 L796.429 1440.86 L796.429 1405.21 L802.317 1405.21 L802.317 1410.75 Q804.418 1407.54 807.25 1405.94 Q810.115 1404.35 813.839 1404.35 Q819.982 1404.35 823.133 1408.17 Q826.284 1411.96 826.284 1419.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M861.423 1422.62 Q861.423 1416.26 858.781 1412.76 Q856.171 1409.25 851.428 1409.25 Q846.718 1409.25 844.076 1412.76 Q841.466 1416.26 841.466 1422.62 Q841.466 1428.96 844.076 1432.46 Q846.718 1435.96 851.428 1435.96 Q856.171 1435.96 858.781 1432.46 Q861.423 1428.96 861.423 1422.62 M867.279 1436.44 Q867.279 1445.54 863.237 1449.96 Q859.195 1454.42 850.856 1454.42 Q847.768 1454.42 845.031 1453.94 Q842.294 1453.5 839.716 1452.54 L839.716 1446.84 Q842.294 1448.24 844.808 1448.91 Q847.323 1449.58 849.932 1449.58 Q855.693 1449.58 858.558 1446.56 Q861.423 1443.57 861.423 1437.49 L861.423 1434.59 Q859.608 1437.74 856.776 1439.3 Q853.943 1440.86 849.996 1440.86 Q843.439 1440.86 839.429 1435.86 Q835.419 1430.87 835.419 1422.62 Q835.419 1414.35 839.429 1409.35 Q843.439 1404.35 849.996 1404.35 Q853.943 1404.35 856.776 1405.91 Q859.608 1407.47 861.423 1410.62 L861.423 1405.21 L867.279 1405.21 L867.279 1436.44 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M879.342 1405.21 L885.199 1405.21 L885.199 1440.86 L879.342 1440.86 L879.342 1405.21 M879.342 1391.34 L885.199 1391.34 L885.199 1398.75 L879.342 1398.75 L879.342 1391.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M903.245 1395.09 L903.245 1405.21 L915.308 1405.21 L915.308 1409.76 L903.245 1409.76 L903.245 1429.12 Q903.245 1433.48 904.423 1434.72 Q905.632 1435.96 909.293 1435.96 L915.308 1435.96 L915.308 1440.86 L909.293 1440.86 Q902.513 1440.86 899.935 1438.35 Q897.357 1435.8 897.357 1429.12 L897.357 1409.76 L893.06 1409.76 L893.06 1405.21 L897.357 1405.21 L897.357 1395.09 L903.245 1395.09 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M922.406 1426.79 L922.406 1405.21 L928.263 1405.21 L928.263 1426.57 Q928.263 1431.63 930.236 1434.18 Q932.209 1436.69 936.156 1436.69 Q940.898 1436.69 943.636 1433.67 Q946.405 1430.64 946.405 1425.42 L946.405 1405.21 L952.261 1405.21 L952.261 1440.86 L946.405 1440.86 L946.405 1435.39 Q944.272 1438.63 941.44 1440.22 Q938.639 1441.78 934.915 1441.78 Q928.772 1441.78 925.589 1437.96 Q922.406 1434.14 922.406 1426.79 M937.143 1404.35 L937.143 1404.35 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M987.782 1410.62 L987.782 1391.34 L993.638 1391.34 L993.638 1440.86 L987.782 1440.86 L987.782 1435.51 Q985.936 1438.7 983.103 1440.26 Q980.302 1441.78 976.355 1441.78 Q969.894 1441.78 965.82 1436.63 Q961.778 1431.47 961.778 1423.07 Q961.778 1414.67 965.82 1409.51 Q969.894 1404.35 976.355 1404.35 Q980.302 1404.35 983.103 1405.91 Q985.936 1407.44 987.782 1410.62 M967.825 1423.07 Q967.825 1429.53 970.467 1433.22 Q973.141 1436.88 977.788 1436.88 Q982.435 1436.88 985.108 1433.22 Q987.782 1429.53 987.782 1423.07 Q987.782 1416.61 985.108 1412.95 Q982.435 1409.25 977.788 1409.25 Q973.141 1409.25 970.467 1412.95 Q967.825 1416.61 967.825 1423.07 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1036.19 1421.57 L1036.19 1424.44 L1009.27 1424.44 Q1009.65 1430.48 1012.89 1433.67 Q1016.17 1436.82 1022 1436.82 Q1025.37 1436.82 1028.52 1435.99 Q1031.71 1435.16 1034.82 1433.51 L1034.82 1439.05 Q1031.67 1440.38 1028.36 1441.08 Q1025.05 1441.78 1021.65 1441.78 Q1013.12 1441.78 1008.12 1436.82 Q1003.16 1431.85 1003.16 1423.39 Q1003.16 1414.63 1007.87 1409.51 Q1012.61 1404.35 1020.63 1404.35 Q1027.82 1404.35 1031.99 1409 Q1036.19 1413.62 1036.19 1421.57 M1030.34 1419.85 Q1030.27 1415.05 1027.63 1412.18 Q1025.02 1409.32 1020.69 1409.32 Q1015.79 1409.32 1012.83 1412.09 Q1009.9 1414.86 1009.46 1419.89 L1030.34 1419.85 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1080.59 1391.4 Q1076.33 1398.72 1074.26 1405.88 Q1072.19 1413.04 1072.19 1420.39 Q1072.19 1427.75 1074.26 1434.97 Q1076.36 1442.17 1080.59 1449.45 L1075.5 1449.45 Q1070.73 1441.97 1068.34 1434.75 Q1065.98 1427.52 1065.98 1420.39 Q1065.98 1413.3 1068.34 1406.1 Q1070.7 1398.91 1075.5 1391.4 L1080.59 1391.4 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1115.41 1410.62 L1115.41 1391.34 L1121.27 1391.34 L1121.27 1440.86 L1115.41 1440.86 L1115.41 1435.51 Q1113.57 1438.7 1110.74 1440.26 Q1107.93 1441.78 1103.99 1441.78 Q1097.53 1441.78 1093.45 1436.63 Q1089.41 1431.47 1089.41 1423.07 Q1089.41 1414.67 1093.45 1409.51 Q1097.53 1404.35 1103.99 1404.35 Q1107.93 1404.35 1110.74 1405.91 Q1113.57 1407.44 1115.41 1410.62 M1095.46 1423.07 Q1095.46 1429.53 1098.1 1433.22 Q1100.77 1436.88 1105.42 1436.88 Q1110.07 1436.88 1112.74 1433.22 Q1115.41 1429.53 1115.41 1423.07 Q1115.41 1416.61 1112.74 1412.95 Q1110.07 1409.25 1105.42 1409.25 Q1100.77 1409.25 1098.1 1412.95 Q1095.46 1416.61 1095.46 1423.07 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1163.83 1421.57 L1163.83 1424.44 L1136.9 1424.44 Q1137.28 1430.48 1140.53 1433.67 Q1143.81 1436.82 1149.63 1436.82 Q1153 1436.82 1156.15 1435.99 Q1159.34 1435.16 1162.46 1433.51 L1162.46 1439.05 Q1159.31 1440.38 1156 1441.08 Q1152.69 1441.78 1149.28 1441.78 Q1140.75 1441.78 1135.75 1436.82 Q1130.79 1431.85 1130.79 1423.39 Q1130.79 1414.63 1135.5 1409.51 Q1140.24 1404.35 1148.26 1404.35 Q1155.45 1404.35 1159.62 1409 Q1163.83 1413.62 1163.83 1421.57 M1157.97 1419.85 Q1157.91 1415.05 1155.26 1412.18 Q1152.65 1409.32 1148.32 1409.32 Q1143.42 1409.32 1140.46 1412.09 Q1137.54 1414.86 1137.09 1419.89 L1157.97 1419.85 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1196.9 1422.62 Q1196.9 1416.26 1194.25 1412.76 Q1191.64 1409.25 1186.9 1409.25 Q1182.19 1409.25 1179.55 1412.76 Q1176.94 1416.26 1176.94 1422.62 Q1176.94 1428.96 1179.55 1432.46 Q1182.19 1435.96 1186.9 1435.96 Q1191.64 1435.96 1194.25 1432.46 Q1196.9 1428.96 1196.9 1422.62 M1202.75 1436.44 Q1202.75 1445.54 1198.71 1449.96 Q1194.67 1454.42 1186.33 1454.42 Q1183.24 1454.42 1180.5 1453.94 Q1177.77 1453.5 1175.19 1452.54 L1175.19 1446.84 Q1177.77 1448.24 1180.28 1448.91 Q1182.8 1449.58 1185.41 1449.58 Q1191.17 1449.58 1194.03 1446.56 Q1196.9 1443.57 1196.9 1437.49 L1196.9 1434.59 Q1195.08 1437.74 1192.25 1439.3 Q1189.42 1440.86 1185.47 1440.86 Q1178.91 1440.86 1174.9 1435.86 Q1170.89 1430.87 1170.89 1422.62 Q1170.89 1414.35 1174.9 1409.35 Q1178.91 1404.35 1185.47 1404.35 Q1189.42 1404.35 1192.25 1405.91 Q1195.08 1407.47 1196.9 1410.62 L1196.9 1405.21 L1202.75 1405.21 L1202.75 1436.44 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1213.89 1391.4 L1218.98 1391.4 Q1223.76 1398.91 1226.11 1406.1 Q1228.5 1413.3 1228.5 1420.39 Q1228.5 1427.52 1226.11 1434.75 Q1223.76 1441.97 1218.98 1449.45 L1213.89 1449.45 Q1218.12 1442.17 1220.19 1434.97 Q1222.29 1427.75 1222.29 1420.39 Q1222.29 1413.04 1220.19 1405.88 Q1218.12 1398.72 1213.89 1391.4 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1433.84,1345.32 783.99,1157.72 783.99,507.868 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1537.4,1255.63 887.549,1068.04 887.549,418.183 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1640.96,1165.95 991.108,978.352 991.108,328.498 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1744.52,1076.26 1094.67,888.667 1094.67,238.813 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1423.22,1354.51 1798.42,1029.59 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1433.84,1345.32 1426.05,1343.07 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1537.4,1255.63 1529.6,1253.38 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1640.96,1165.95 1633.16,1163.7 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1744.52,1076.26 1736.72,1074.01 "/>
+<path clip-path="url(#clip750)" d="M1471.3 1344.48 Q1467.69 1344.48 1465.86 1348.04 Q1464.05 1351.59 1464.05 1358.72 Q1464.05 1365.82 1465.86 1369.39 Q1467.69 1372.93 1471.3 1372.93 Q1474.93 1372.93 1476.74 1369.39 Q1478.57 1365.82 1478.57 1358.72 Q1478.57 1351.59 1476.74 1348.04 Q1474.93 1344.48 1471.3 1344.48 M1471.3 1340.78 Q1477.11 1340.78 1480.17 1345.38 Q1483.24 1349.97 1483.24 1358.72 Q1483.24 1367.44 1480.17 1372.05 Q1477.11 1376.63 1471.3 1376.63 Q1465.49 1376.63 1462.41 1372.05 Q1459.36 1367.44 1459.36 1358.72 Q1459.36 1349.97 1462.41 1345.38 Q1465.49 1340.78 1471.3 1340.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1563.59 1282.34 L1571.23 1282.34 L1571.23 1255.98 L1562.92 1257.64 L1562.92 1253.38 L1571.18 1251.72 L1575.85 1251.72 L1575.85 1282.34 L1583.49 1282.34 L1583.49 1286.28 L1563.59 1286.28 L1563.59 1282.34 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1602.94 1254.79 Q1599.33 1254.79 1597.5 1258.36 Q1595.69 1261.9 1595.69 1269.03 Q1595.69 1276.14 1597.5 1279.7 Q1599.33 1283.24 1602.94 1283.24 Q1606.57 1283.24 1608.38 1279.7 Q1610.21 1276.14 1610.21 1269.03 Q1610.21 1261.9 1608.38 1258.36 Q1606.57 1254.79 1602.94 1254.79 M1602.94 1251.09 Q1608.75 1251.09 1611.8 1255.7 Q1614.88 1260.28 1614.88 1269.03 Q1614.88 1277.76 1611.8 1282.36 Q1608.75 1286.95 1602.94 1286.95 Q1597.13 1286.95 1594.05 1282.36 Q1590.99 1277.76 1590.99 1269.03 Q1590.99 1260.28 1594.05 1255.7 Q1597.13 1251.09 1602.94 1251.09 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1672.1 1192.66 L1688.42 1192.66 L1688.42 1196.59 L1666.47 1196.59 L1666.47 1192.66 Q1669.14 1189.9 1673.72 1185.27 Q1678.33 1180.62 1679.51 1179.28 Q1681.75 1176.75 1682.63 1175.02 Q1683.53 1173.26 1683.53 1171.57 Q1683.53 1168.81 1681.59 1167.08 Q1679.67 1165.34 1676.57 1165.34 Q1674.37 1165.34 1671.91 1166.11 Q1669.48 1166.87 1666.71 1168.42 L1666.71 1163.7 Q1669.53 1162.56 1671.98 1161.98 Q1674.44 1161.41 1676.47 1161.41 Q1681.84 1161.41 1685.04 1164.09 Q1688.23 1166.78 1688.23 1171.27 Q1688.23 1173.4 1687.42 1175.32 Q1686.64 1177.22 1684.53 1179.81 Q1683.95 1180.48 1680.85 1183.7 Q1677.75 1186.89 1672.1 1192.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1708.23 1165.11 Q1704.62 1165.11 1702.79 1168.67 Q1700.99 1172.22 1700.99 1179.35 Q1700.99 1186.45 1702.79 1190.02 Q1704.62 1193.56 1708.23 1193.56 Q1711.87 1193.56 1713.67 1190.02 Q1715.5 1186.45 1715.5 1179.35 Q1715.5 1172.22 1713.67 1168.67 Q1711.87 1165.11 1708.23 1165.11 M1708.23 1161.41 Q1714.04 1161.41 1717.1 1166.01 Q1720.18 1170.6 1720.18 1179.35 Q1720.18 1188.07 1717.1 1192.68 Q1714.04 1197.26 1708.23 1197.26 Q1702.42 1197.26 1699.34 1192.68 Q1696.29 1188.07 1696.29 1179.35 Q1696.29 1170.6 1699.34 1166.01 Q1702.42 1161.41 1708.23 1161.41 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1785.66 1088.27 Q1789.02 1088.99 1790.89 1091.26 Q1792.79 1093.53 1792.79 1096.86 Q1792.79 1101.98 1789.27 1104.78 Q1785.75 1107.58 1779.27 1107.58 Q1777.09 1107.58 1774.78 1107.14 Q1772.49 1106.72 1770.03 1105.86 L1770.03 1101.35 Q1771.98 1102.48 1774.29 1103.06 Q1776.61 1103.64 1779.13 1103.64 Q1783.53 1103.64 1785.82 1101.91 Q1788.14 1100.17 1788.14 1096.86 Q1788.14 1093.8 1785.98 1092.09 Q1783.85 1090.36 1780.03 1090.36 L1776.01 1090.36 L1776.01 1086.51 L1780.22 1086.51 Q1783.67 1086.51 1785.5 1085.15 Q1787.33 1083.76 1787.33 1081.17 Q1787.33 1078.5 1785.43 1077.09 Q1783.55 1075.66 1780.03 1075.66 Q1778.11 1075.66 1775.91 1076.07 Q1773.71 1076.49 1771.08 1077.37 L1771.08 1073.2 Q1773.74 1072.46 1776.05 1072.09 Q1778.39 1071.72 1780.45 1071.72 Q1785.77 1071.72 1788.88 1074.15 Q1791.98 1076.56 1791.98 1080.68 Q1791.98 1083.55 1790.33 1085.54 Q1788.69 1087.51 1785.66 1088.27 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1811.65 1075.42 Q1808.04 1075.42 1806.21 1078.99 Q1804.41 1082.53 1804.41 1089.66 Q1804.41 1096.77 1806.21 1100.33 Q1808.04 1103.87 1811.65 1103.87 Q1815.29 1103.87 1817.09 1100.33 Q1818.92 1096.77 1818.92 1089.66 Q1818.92 1082.53 1817.09 1078.99 Q1815.29 1075.42 1811.65 1075.42 M1811.65 1071.72 Q1817.46 1071.72 1820.52 1076.33 Q1823.6 1080.91 1823.6 1089.66 Q1823.6 1098.39 1820.52 1102.99 Q1817.46 1107.58 1811.65 1107.58 Q1805.84 1107.58 1802.76 1102.99 Q1799.71 1098.39 1799.71 1089.66 Q1799.71 1080.91 1802.76 1076.33 Q1805.84 1071.72 1811.65 1071.72 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1508.58 1324.68 L1515.01 1324.68 L1515.01 1366.78 L1538.15 1366.78 L1538.15 1372.2 L1508.58 1372.2 L1508.58 1324.68 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1560.84 1354.28 Q1553.74 1354.28 1551.01 1355.9 Q1548.27 1357.52 1548.27 1361.44 Q1548.27 1364.56 1550.31 1366.4 Q1552.38 1368.22 1555.91 1368.22 Q1560.78 1368.22 1563.71 1364.78 Q1566.67 1361.31 1566.67 1355.58 L1566.67 1354.28 L1560.84 1354.28 M1572.52 1351.86 L1572.52 1372.2 L1566.67 1372.2 L1566.67 1366.78 Q1564.66 1370.03 1561.67 1371.59 Q1558.68 1373.12 1554.35 1373.12 Q1548.87 1373.12 1545.63 1370.06 Q1542.41 1366.98 1542.41 1361.82 Q1542.41 1355.8 1546.42 1352.75 Q1550.47 1349.69 1558.45 1349.69 L1566.67 1349.69 L1566.67 1349.12 Q1566.67 1345.08 1563.99 1342.88 Q1561.35 1340.65 1556.55 1340.65 Q1553.49 1340.65 1550.59 1341.39 Q1547.7 1342.12 1545.02 1343.58 L1545.02 1338.17 Q1548.24 1336.93 1551.26 1336.32 Q1554.29 1335.69 1557.15 1335.69 Q1564.88 1335.69 1568.7 1339.7 Q1572.52 1343.71 1572.52 1351.86 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1590.38 1326.43 L1590.38 1336.55 L1602.44 1336.55 L1602.44 1341.1 L1590.38 1341.1 L1590.38 1360.45 Q1590.38 1364.81 1591.56 1366.05 Q1592.77 1367.29 1596.43 1367.29 L1602.44 1367.29 L1602.44 1372.2 L1596.43 1372.2 Q1589.65 1372.2 1587.07 1369.68 Q1584.49 1367.13 1584.49 1360.45 L1584.49 1341.1 L1580.19 1341.1 L1580.19 1336.55 L1584.49 1336.55 L1584.49 1326.43 L1590.38 1326.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1610.14 1336.55 L1616 1336.55 L1616 1372.2 L1610.14 1372.2 L1610.14 1336.55 M1610.14 1322.67 L1616 1322.67 L1616 1330.09 L1610.14 1330.09 L1610.14 1322.67 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1634.05 1326.43 L1634.05 1336.55 L1646.11 1336.55 L1646.11 1341.1 L1634.05 1341.1 L1634.05 1360.45 Q1634.05 1364.81 1635.23 1366.05 Q1636.43 1367.29 1640.09 1367.29 L1646.11 1367.29 L1646.11 1372.2 L1640.09 1372.2 Q1633.32 1372.2 1630.74 1369.68 Q1628.16 1367.13 1628.16 1360.45 L1628.16 1341.1 L1623.86 1341.1 L1623.86 1336.55 L1628.16 1336.55 L1628.16 1326.43 L1634.05 1326.43 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1653.21 1358.13 L1653.21 1336.55 L1659.06 1336.55 L1659.06 1357.9 Q1659.06 1362.96 1661.04 1365.51 Q1663.01 1368.03 1666.96 1368.03 Q1671.7 1368.03 1674.44 1365 Q1677.21 1361.98 1677.21 1356.76 L1677.21 1336.55 L1683.06 1336.55 L1683.06 1372.2 L1677.21 1372.2 L1677.21 1366.72 Q1675.07 1369.97 1672.24 1371.56 Q1669.44 1373.12 1665.72 1373.12 Q1659.57 1373.12 1656.39 1369.3 Q1653.21 1365.48 1653.21 1358.13 M1667.94 1335.69 L1667.94 1335.69 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1718.58 1341.96 L1718.58 1322.67 L1724.44 1322.67 L1724.44 1372.2 L1718.58 1372.2 L1718.58 1366.85 Q1716.74 1370.03 1713.91 1371.59 Q1711.1 1373.12 1707.16 1373.12 Q1700.7 1373.12 1696.62 1367.96 Q1692.58 1362.81 1692.58 1354.4 Q1692.58 1346 1696.62 1340.84 Q1700.7 1335.69 1707.16 1335.69 Q1711.1 1335.69 1713.91 1337.25 Q1716.74 1338.78 1718.58 1341.96 M1698.63 1354.4 Q1698.63 1360.86 1701.27 1364.56 Q1703.94 1368.22 1708.59 1368.22 Q1713.24 1368.22 1715.91 1364.56 Q1718.58 1360.86 1718.58 1354.4 Q1718.58 1347.94 1715.91 1344.28 Q1713.24 1340.59 1708.59 1340.59 Q1703.94 1340.59 1701.27 1344.28 Q1698.63 1347.94 1698.63 1354.4 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1767 1352.91 L1767 1355.77 L1740.07 1355.77 Q1740.45 1361.82 1743.7 1365 Q1746.98 1368.15 1752.8 1368.15 Q1756.17 1368.15 1759.32 1367.33 Q1762.51 1366.5 1765.63 1364.84 L1765.63 1370.38 Q1762.48 1371.72 1759.17 1372.42 Q1755.86 1373.12 1752.45 1373.12 Q1743.92 1373.12 1738.92 1368.15 Q1733.96 1363.19 1733.96 1354.72 Q1733.96 1345.97 1738.67 1340.84 Q1743.41 1335.69 1751.43 1335.69 Q1758.62 1335.69 1762.79 1340.33 Q1767 1344.95 1767 1352.91 M1761.14 1351.19 Q1761.08 1346.38 1758.43 1343.52 Q1755.82 1340.65 1751.49 1340.65 Q1746.59 1340.65 1743.63 1343.42 Q1740.7 1346.19 1740.26 1351.22 L1761.14 1351.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1811.4 1322.73 Q1807.13 1330.05 1805.06 1337.22 Q1802.99 1344.38 1802.99 1351.73 Q1802.99 1359.08 1805.06 1366.31 Q1807.16 1373.5 1811.4 1380.79 L1806.3 1380.79 Q1801.53 1373.31 1799.14 1366.08 Q1796.79 1358.86 1796.79 1351.73 Q1796.79 1344.63 1799.14 1337.44 Q1801.5 1330.25 1806.3 1322.73 L1811.4 1322.73 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1846.22 1341.96 L1846.22 1322.67 L1852.07 1322.67 L1852.07 1372.2 L1846.22 1372.2 L1846.22 1366.85 Q1844.37 1370.03 1841.54 1371.59 Q1838.74 1373.12 1834.79 1373.12 Q1828.33 1373.12 1824.25 1367.96 Q1820.21 1362.81 1820.21 1354.4 Q1820.21 1346 1824.25 1340.84 Q1828.33 1335.69 1834.79 1335.69 Q1838.74 1335.69 1841.54 1337.25 Q1844.37 1338.78 1846.22 1341.96 M1826.26 1354.4 Q1826.26 1360.86 1828.9 1364.56 Q1831.58 1368.22 1836.22 1368.22 Q1840.87 1368.22 1843.54 1364.56 Q1846.22 1360.86 1846.22 1354.4 Q1846.22 1347.94 1843.54 1344.28 Q1840.87 1340.59 1836.22 1340.59 Q1831.58 1340.59 1828.9 1344.28 Q1826.26 1347.94 1826.26 1354.4 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1894.63 1352.91 L1894.63 1355.77 L1867.7 1355.77 Q1868.08 1361.82 1871.33 1365 Q1874.61 1368.15 1880.43 1368.15 Q1883.81 1368.15 1886.96 1367.33 Q1890.14 1366.5 1893.26 1364.84 L1893.26 1370.38 Q1890.11 1371.72 1886.8 1372.42 Q1883.49 1373.12 1880.08 1373.12 Q1871.55 1373.12 1866.55 1368.15 Q1861.59 1363.19 1861.59 1354.72 Q1861.59 1345.97 1866.3 1340.84 Q1871.04 1335.69 1879.06 1335.69 Q1886.26 1335.69 1890.43 1340.33 Q1894.63 1344.95 1894.63 1352.91 M1888.77 1351.19 Q1888.71 1346.38 1886.07 1343.52 Q1883.46 1340.65 1879.13 1340.65 Q1874.23 1340.65 1871.27 1343.42 Q1868.34 1346.19 1867.89 1351.22 L1888.77 1351.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1927.7 1353.96 Q1927.7 1347.59 1925.06 1344.09 Q1922.45 1340.59 1917.7 1340.59 Q1912.99 1340.59 1910.35 1344.09 Q1907.74 1347.59 1907.74 1353.96 Q1907.74 1360.29 1910.35 1363.79 Q1912.99 1367.29 1917.7 1367.29 Q1922.45 1367.29 1925.06 1363.79 Q1927.7 1360.29 1927.7 1353.96 M1933.55 1367.77 Q1933.55 1376.87 1929.51 1381.3 Q1925.47 1385.75 1917.13 1385.75 Q1914.04 1385.75 1911.31 1385.28 Q1908.57 1384.83 1905.99 1383.88 L1905.99 1378.18 Q1908.57 1379.58 1911.08 1380.25 Q1913.6 1380.92 1916.21 1380.92 Q1921.97 1380.92 1924.83 1377.89 Q1927.7 1374.9 1927.7 1368.82 L1927.7 1365.93 Q1925.88 1369.08 1923.05 1370.64 Q1920.22 1372.2 1916.27 1372.2 Q1909.71 1372.2 1905.7 1367.2 Q1901.69 1362.2 1901.69 1353.96 Q1901.69 1345.68 1905.7 1340.68 Q1909.71 1335.69 1916.27 1335.69 Q1920.22 1335.69 1923.05 1337.25 Q1925.88 1338.81 1927.7 1341.96 L1927.7 1336.55 L1933.55 1336.55 L1933.55 1367.77 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1944.69 1322.73 L1949.79 1322.73 Q1954.56 1330.25 1956.92 1337.44 Q1959.3 1344.63 1959.3 1351.73 Q1959.3 1358.86 1956.92 1366.08 Q1954.56 1373.31 1949.79 1380.79 L1944.69 1380.79 Q1948.93 1373.5 1951 1366.31 Q1953.1 1359.08 1953.1 1351.73 Q1953.1 1344.38 1951 1337.22 Q1948.93 1330.05 1944.69 1322.73 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="773.371,1080.41 1148.56,755.48 1798.42,943.077 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="773.371,910.11 1148.56,585.183 1798.42,772.78 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="773.371,739.813 1148.56,414.886 1798.42,602.483 "/>
+<polyline clip-path="url(#clip752)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="773.371,569.516 1148.56,244.589 1798.42,432.185 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="773.371,1166.92 773.371,517.064 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="773.371,1080.41 777.873,1076.51 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="773.371,910.11 777.873,906.211 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="773.371,739.813 777.873,735.914 "/>
+<polyline clip-path="url(#clip750)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="773.371,569.516 777.873,565.617 "/>
+<path clip-path="url(#clip750)" d="M678.205 1093.75 L685.844 1093.75 L685.844 1067.39 L677.534 1069.05 L677.534 1064.79 L685.798 1063.13 L690.474 1063.13 L690.474 1093.75 L698.113 1093.75 L698.113 1097.69 L678.205 1097.69 L678.205 1093.75 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M707.557 1091.81 L712.441 1091.81 L712.441 1097.69 L707.557 1097.69 L707.557 1091.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M732.626 1066.21 Q729.015 1066.21 727.187 1069.77 Q725.381 1073.31 725.381 1080.44 Q725.381 1087.55 727.187 1091.11 Q729.015 1094.65 732.626 1094.65 Q736.261 1094.65 738.066 1091.11 Q739.895 1087.55 739.895 1080.44 Q739.895 1073.31 738.066 1069.77 Q736.261 1066.21 732.626 1066.21 M732.626 1062.5 Q738.437 1062.5 741.492 1067.11 Q744.571 1071.69 744.571 1080.44 Q744.571 1089.17 741.492 1093.78 Q738.437 1098.36 732.626 1098.36 Q726.816 1098.36 723.738 1093.78 Q720.682 1089.17 720.682 1080.44 Q720.682 1071.69 723.738 1067.11 Q726.816 1062.5 732.626 1062.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M679.201 923.455 L686.84 923.455 L686.84 897.089 L678.53 898.756 L678.53 894.497 L686.793 892.83 L691.469 892.83 L691.469 923.455 L699.108 923.455 L699.108 927.39 L679.201 927.39 L679.201 923.455 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M708.553 921.51 L713.437 921.51 L713.437 927.39 L708.553 927.39 L708.553 921.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M723.668 892.83 L742.025 892.83 L742.025 896.765 L727.951 896.765 L727.951 905.237 Q728.969 904.89 729.988 904.728 Q731.006 904.543 732.025 904.543 Q737.812 904.543 741.191 907.714 Q744.571 910.885 744.571 916.302 Q744.571 921.881 741.099 924.983 Q737.626 928.061 731.307 928.061 Q729.131 928.061 726.863 927.691 Q724.617 927.321 722.21 926.58 L722.21 921.881 Q724.293 923.015 726.515 923.571 Q728.738 924.126 731.214 924.126 Q735.219 924.126 737.557 922.02 Q739.895 919.913 739.895 916.302 Q739.895 912.691 737.557 910.585 Q735.219 908.478 731.214 908.478 Q729.339 908.478 727.464 908.895 Q725.613 909.311 723.668 910.191 L723.668 892.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M681.423 753.158 L697.742 753.158 L697.742 757.093 L675.798 757.093 L675.798 753.158 Q678.46 750.403 683.043 745.773 Q687.65 741.121 688.83 739.778 Q691.076 737.255 691.955 735.519 Q692.858 733.76 692.858 732.07 Q692.858 729.315 690.914 727.579 Q688.992 725.843 685.891 725.843 Q683.692 725.843 681.238 726.607 Q678.807 727.371 676.03 728.922 L676.03 724.2 Q678.854 723.065 681.307 722.487 Q683.761 721.908 685.798 721.908 Q691.168 721.908 694.363 724.593 Q697.557 727.278 697.557 731.769 Q697.557 733.899 696.747 735.82 Q695.96 737.718 693.854 740.311 Q693.275 740.982 690.173 744.199 Q687.071 747.394 681.423 753.158 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M707.557 751.213 L712.441 751.213 L712.441 757.093 L707.557 757.093 L707.557 751.213 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M732.626 725.612 Q729.015 725.612 727.187 729.176 Q725.381 732.718 725.381 739.848 Q725.381 746.954 727.187 750.519 Q729.015 754.06 732.626 754.06 Q736.261 754.06 738.066 750.519 Q739.895 746.954 739.895 739.848 Q739.895 732.718 738.066 729.176 Q736.261 725.612 732.626 725.612 M732.626 721.908 Q738.437 721.908 741.492 726.514 Q744.571 731.098 744.571 739.848 Q744.571 748.574 741.492 753.181 Q738.437 757.764 732.626 757.764 Q726.816 757.764 723.738 753.181 Q720.682 748.574 720.682 739.848 Q720.682 731.098 723.738 726.514 Q726.816 721.908 732.626 721.908 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M682.418 582.861 L698.738 582.861 L698.738 586.796 L676.793 586.796 L676.793 582.861 Q679.455 580.106 684.039 575.476 Q688.645 570.824 689.826 569.481 Q692.071 566.958 692.951 565.222 Q693.854 563.463 693.854 561.773 Q693.854 559.018 691.909 557.282 Q689.988 555.546 686.886 555.546 Q684.687 555.546 682.233 556.31 Q679.803 557.074 677.025 558.625 L677.025 553.902 Q679.849 552.768 682.303 552.189 Q684.756 551.611 686.793 551.611 Q692.164 551.611 695.358 554.296 Q698.553 556.981 698.553 561.472 Q698.553 563.601 697.742 565.523 Q696.955 567.421 694.849 570.013 Q694.27 570.685 691.168 573.902 Q688.067 577.097 682.418 582.861 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M708.553 580.916 L713.437 580.916 L713.437 586.796 L708.553 586.796 L708.553 580.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M723.668 552.236 L742.025 552.236 L742.025 556.171 L727.951 556.171 L727.951 564.643 Q728.969 564.296 729.988 564.134 Q731.006 563.949 732.025 563.949 Q737.812 563.949 741.191 567.12 Q744.571 570.291 744.571 575.708 Q744.571 581.286 741.099 584.388 Q737.626 587.467 731.307 587.467 Q729.131 587.467 726.863 587.097 Q724.617 586.726 722.21 585.986 L722.21 581.286 Q724.293 582.421 726.515 582.976 Q728.738 583.532 731.214 583.532 Q735.219 583.532 737.557 581.425 Q739.895 579.319 739.895 575.708 Q739.895 572.097 737.557 569.99 Q735.219 567.884 731.214 567.884 Q729.339 567.884 727.464 568.3 Q725.613 568.717 723.668 569.597 L723.668 552.236 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M577.156 1149.14 L600.805 1157.86 L600.805 1140.38 L577.156 1149.14 M570.822 1152.76 L570.822 1145.48 L618.342 1127.37 L618.342 1134.05 L606.152 1138.38 L606.152 1159.8 L618.342 1164.13 L618.342 1170.91 L570.822 1152.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M568.817 1120.68 L568.817 1114.83 L618.342 1114.83 L618.342 1120.68 L568.817 1120.68 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M572.573 1096.78 L582.694 1096.78 L582.694 1084.72 L587.246 1084.72 L587.246 1096.78 L606.597 1096.78 Q610.958 1096.78 612.199 1095.6 Q613.44 1094.39 613.44 1090.73 L613.44 1084.72 L618.342 1084.72 L618.342 1090.73 Q618.342 1097.51 615.828 1100.09 Q613.281 1102.67 606.597 1102.67 L587.246 1102.67 L587.246 1106.96 L582.694 1106.96 L582.694 1102.67 L572.573 1102.67 L572.573 1096.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M582.694 1077.01 L582.694 1071.16 L618.342 1071.16 L618.342 1077.01 L582.694 1077.01 M568.817 1077.01 L568.817 1071.16 L576.233 1071.16 L576.233 1077.01 L568.817 1077.01 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M572.573 1053.11 L582.694 1053.11 L582.694 1041.05 L587.246 1041.05 L587.246 1053.11 L606.597 1053.11 Q610.958 1053.11 612.199 1051.93 Q613.44 1050.72 613.44 1047.06 L613.44 1041.05 L618.342 1041.05 L618.342 1047.06 Q618.342 1053.84 615.828 1056.42 Q613.281 1059 606.597 1059 L587.246 1059 L587.246 1063.29 L582.694 1063.29 L582.694 1059 L572.573 1059 L572.573 1053.11 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M604.274 1033.95 L582.694 1033.95 L582.694 1028.09 L604.051 1028.09 Q609.112 1028.09 611.658 1026.12 Q614.173 1024.15 614.173 1020.2 Q614.173 1015.46 611.149 1012.72 Q608.125 1009.95 602.905 1009.95 L582.694 1009.95 L582.694 1004.09 L618.342 1004.09 L618.342 1009.95 L612.868 1009.95 Q616.114 1012.08 617.705 1014.92 Q619.265 1017.72 619.265 1021.44 Q619.265 1027.58 615.446 1030.77 Q611.626 1033.95 604.274 1033.95 M581.835 1019.21 L581.835 1019.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M588.105 968.573 L568.817 968.573 L568.817 962.717 L618.342 962.717 L618.342 968.573 L612.995 968.573 Q616.178 970.419 617.737 973.252 Q619.265 976.053 619.265 980 Q619.265 986.461 614.109 990.535 Q608.953 994.577 600.55 994.577 Q592.147 994.577 586.991 990.535 Q581.835 986.461 581.835 980 Q581.835 976.053 583.394 973.252 Q584.922 970.419 588.105 968.573 M600.55 988.53 Q607.011 988.53 610.703 985.888 Q614.363 983.214 614.363 978.567 Q614.363 973.92 610.703 971.247 Q607.011 968.573 600.55 968.573 Q594.089 968.573 590.428 971.247 Q586.736 973.92 586.736 978.567 Q586.736 983.214 590.428 985.888 Q594.089 988.53 600.55 988.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M599.054 920.162 L601.919 920.162 L601.919 947.089 Q607.966 946.707 611.149 943.46 Q614.3 940.182 614.3 934.358 Q614.3 930.984 613.472 927.833 Q612.645 924.65 610.99 921.531 L616.528 921.531 Q617.865 924.682 618.565 927.992 Q619.265 931.302 619.265 934.708 Q619.265 943.238 614.3 948.235 Q609.335 953.2 600.868 953.2 Q592.115 953.2 586.991 948.489 Q581.835 943.747 581.835 935.726 Q581.835 928.533 586.482 924.363 Q591.097 920.162 599.054 920.162 M597.335 926.018 Q592.529 926.082 589.665 928.724 Q586.8 931.334 586.8 935.662 Q586.8 940.564 589.569 943.524 Q592.338 946.452 597.367 946.898 L597.335 926.018 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M568.881 875.761 Q576.201 880.026 583.363 882.095 Q590.524 884.164 597.876 884.164 Q605.229 884.164 612.454 882.095 Q619.647 879.994 626.936 875.761 L626.936 880.854 Q619.456 885.628 612.231 888.015 Q605.006 890.37 597.876 890.37 Q590.779 890.37 583.585 888.015 Q576.392 885.66 568.881 880.854 L568.881 875.761 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M612.931 862.457 L612.931 851.953 L576.679 851.953 L578.97 863.38 L573.114 863.38 L570.822 852.017 L570.822 845.588 L612.931 845.588 L612.931 835.084 L618.342 835.084 L618.342 862.457 L612.931 862.457 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M575.055 808.348 Q575.055 813.314 579.957 815.828 Q584.827 818.311 594.63 818.311 Q604.401 818.311 609.303 815.828 Q614.173 813.314 614.173 808.348 Q614.173 803.351 609.303 800.869 Q604.401 798.354 594.63 798.354 Q584.827 798.354 579.957 800.869 Q575.055 803.351 575.055 808.348 M569.963 808.348 Q569.963 800.359 576.297 796.158 Q582.599 791.925 594.63 791.925 Q606.629 791.925 612.963 796.158 Q619.265 800.359 619.265 808.348 Q619.265 816.337 612.963 820.571 Q606.629 824.772 594.63 824.772 Q582.599 824.772 576.297 820.571 Q569.963 816.337 569.963 808.348 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M575.055 766.876 Q575.055 771.841 579.957 774.355 Q584.827 776.838 594.63 776.838 Q604.401 776.838 609.303 774.355 Q614.173 771.841 614.173 766.876 Q614.173 761.879 609.303 759.396 Q604.401 756.882 594.63 756.882 Q584.827 756.882 579.957 759.396 Q575.055 761.879 575.055 766.876 M569.963 766.876 Q569.963 758.887 576.297 754.685 Q582.599 750.452 594.63 750.452 Q606.629 750.452 612.963 754.685 Q619.265 758.887 619.265 766.876 Q619.265 774.865 612.963 779.098 Q606.629 783.299 594.63 783.299 Q582.599 783.299 576.297 779.098 Q569.963 774.865 569.963 766.876 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M610.258 738.485 L610.258 731.769 L615.732 731.769 L625.917 736.989 L625.917 741.095 L615.732 738.485 L610.258 738.485 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M575.055 704.683 Q575.055 709.648 579.957 712.163 Q584.827 714.645 594.63 714.645 Q604.401 714.645 609.303 712.163 Q614.173 709.648 614.173 704.683 Q614.173 699.686 609.303 697.203 Q604.401 694.689 594.63 694.689 Q584.827 694.689 579.957 697.203 Q575.055 699.686 575.055 704.683 M569.963 704.683 Q569.963 696.694 576.297 692.493 Q582.599 688.259 594.63 688.259 Q606.629 688.259 612.963 692.493 Q619.265 696.694 619.265 704.683 Q619.265 712.672 612.963 716.905 Q606.629 721.106 594.63 721.106 Q582.599 721.106 576.297 716.905 Q569.963 712.672 569.963 704.683 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M575.055 663.21 Q575.055 668.176 579.957 670.69 Q584.827 673.173 594.63 673.173 Q604.401 673.173 609.303 670.69 Q614.173 668.176 614.173 663.21 Q614.173 658.213 609.303 655.731 Q604.401 653.216 594.63 653.216 Q584.827 653.216 579.957 655.731 Q575.055 658.213 575.055 663.21 M569.963 663.21 Q569.963 655.221 576.297 651.02 Q582.599 646.787 594.63 646.787 Q606.629 646.787 612.963 651.02 Q619.265 655.221 619.265 663.21 Q619.265 671.199 612.963 675.432 Q606.629 679.634 594.63 679.634 Q582.599 679.634 576.297 675.432 Q569.963 671.199 569.963 663.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M575.055 621.738 Q575.055 626.703 579.957 629.217 Q584.827 631.7 594.63 631.7 Q604.401 631.7 609.303 629.217 Q614.173 626.703 614.173 621.738 Q614.173 616.741 609.303 614.258 Q604.401 611.744 594.63 611.744 Q584.827 611.744 579.957 614.258 Q575.055 616.741 575.055 621.738 M569.963 621.738 Q569.963 613.749 576.297 609.547 Q582.599 605.314 594.63 605.314 Q606.629 605.314 612.963 609.547 Q619.265 613.749 619.265 621.738 Q619.265 629.727 612.963 633.96 Q606.629 638.161 594.63 638.161 Q582.599 638.161 576.297 633.96 Q569.963 629.727 569.963 621.738 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M568.817 556.075 L573.687 556.075 L573.687 561.677 Q573.687 564.828 574.96 566.07 Q576.233 567.279 579.543 567.279 L582.694 567.279 L582.694 557.635 L587.246 557.635 L587.246 567.279 L618.342 567.279 L618.342 573.167 L587.246 573.167 L587.246 578.769 L582.694 578.769 L582.694 573.167 L580.211 573.167 Q574.26 573.167 571.554 570.398 Q568.817 567.629 568.817 561.614 L568.817 556.075 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M572.573 546.527 L582.694 546.527 L582.694 534.464 L587.246 534.464 L587.246 546.527 L606.597 546.527 Q610.958 546.527 612.199 545.349 Q613.44 544.14 613.44 540.479 L613.44 534.464 L618.342 534.464 L618.342 540.479 Q618.342 547.259 615.828 549.837 Q613.281 552.415 606.597 552.415 L587.246 552.415 L587.246 556.712 L582.694 556.712 L582.694 552.415 L572.573 552.415 L572.573 546.527 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M568.881 527.684 L568.881 522.592 Q576.392 517.818 583.585 515.462 Q590.779 513.075 597.876 513.075 Q605.006 513.075 612.231 515.462 Q619.456 517.818 626.936 522.592 L626.936 527.684 Q619.647 523.451 612.454 521.382 Q605.229 519.282 597.876 519.282 Q590.524 519.282 583.363 521.382 Q576.201 523.451 568.881 527.684 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M647.371 14.0809 L647.371 22.0612 Q642.712 19.8332 638.58 18.7395 Q634.448 17.6457 630.6 17.6457 Q623.916 17.6457 620.27 20.2383 Q616.665 22.8309 616.665 27.611 Q616.665 31.6214 619.055 33.6873 Q621.485 35.7128 628.21 36.9686 L633.152 37.9813 Q642.307 39.7232 646.641 44.1387 Q651.016 48.5136 651.016 55.8863 Q651.016 64.6767 645.102 69.2137 Q639.228 73.7508 627.845 73.7508 Q623.551 73.7508 618.69 72.7785 Q613.87 71.8063 608.685 69.9024 L608.685 61.4765 Q613.667 64.2716 618.447 65.6895 Q623.227 67.1073 627.845 67.1073 Q634.853 67.1073 638.661 64.3527 Q642.469 61.598 642.469 56.4939 Q642.469 52.0379 639.714 49.5264 Q637 47.0148 630.762 45.759 L625.779 44.7868 Q616.624 42.9639 612.533 39.075 Q608.441 35.1862 608.441 28.2591 Q608.441 20.2383 614.072 15.6203 Q619.743 11.0023 629.668 11.0023 Q633.922 11.0023 638.337 11.7719 Q642.753 12.5416 647.371 14.0809 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M670.663 65.7705 L670.663 89.8329 L663.169 89.8329 L663.169 27.2059 L670.663 27.2059 L670.663 34.0924 Q673.013 30.0415 676.578 28.0971 Q680.183 26.1121 685.166 26.1121 Q693.429 26.1121 698.574 32.6746 Q703.759 39.2371 703.759 49.9314 Q703.759 60.6258 698.574 67.1883 Q693.429 73.7508 685.166 73.7508 Q680.183 73.7508 676.578 71.8063 Q673.013 69.8214 670.663 65.7705 M696.022 49.9314 Q696.022 41.7081 692.619 37.0496 Q689.257 32.3505 683.343 32.3505 Q677.428 32.3505 674.026 37.0496 Q670.663 41.7081 670.663 49.9314 Q670.663 58.1548 674.026 62.8538 Q677.428 67.5124 683.343 67.5124 Q689.257 67.5124 692.619 62.8538 Q696.022 58.1548 696.022 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M736.734 49.7694 Q727.7 49.7694 724.216 51.8354 Q720.733 53.9013 720.733 58.8839 Q720.733 62.8538 723.325 65.2034 Q725.958 67.5124 730.455 67.5124 Q736.653 67.5124 740.379 63.1374 Q744.147 58.7219 744.147 51.4303 L744.147 49.7694 L736.734 49.7694 M751.6 46.6907 L751.6 72.576 L744.147 72.576 L744.147 65.6895 Q741.595 69.8214 737.787 71.8063 Q733.979 73.7508 728.47 73.7508 Q721.502 73.7508 717.37 69.8619 Q713.279 65.9325 713.279 59.3701 Q713.279 51.7138 718.383 47.825 Q723.528 43.9361 733.695 43.9361 L744.147 43.9361 L744.147 43.2069 Q744.147 38.0623 740.744 35.2672 Q737.382 32.4315 731.265 32.4315 Q727.376 32.4315 723.69 33.3632 Q720.003 34.295 716.601 36.1584 L716.601 29.2718 Q720.692 27.692 724.54 26.9223 Q728.389 26.1121 732.035 26.1121 Q741.878 26.1121 746.739 31.2163 Q751.6 36.3204 751.6 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M799.604 28.9478 L799.604 35.9153 Q796.444 34.1734 793.244 33.3227 Q790.084 32.4315 786.843 32.4315 Q779.592 32.4315 775.582 37.0496 Q771.571 41.6271 771.571 49.9314 Q771.571 58.2358 775.582 62.8538 Q779.592 67.4314 786.843 67.4314 Q790.084 67.4314 793.244 66.5807 Q796.444 65.6895 799.604 63.9476 L799.604 70.8341 Q796.484 72.2924 793.122 73.0216 Q789.8 73.7508 786.033 73.7508 Q775.784 73.7508 769.748 67.3098 Q763.713 60.8689 763.713 49.9314 Q763.713 38.832 769.789 32.472 Q775.906 26.1121 786.519 26.1121 Q789.962 26.1121 793.244 26.8413 Q796.525 27.5299 799.604 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M851.374 48.0275 L851.374 51.6733 L817.104 51.6733 Q817.59 59.3701 821.722 63.421 Q825.894 67.4314 833.307 67.4314 Q837.601 67.4314 841.612 66.3781 Q845.662 65.3249 849.632 63.2184 L849.632 70.267 Q845.622 71.9684 841.409 72.8596 Q837.196 73.7508 832.862 73.7508 Q822.005 73.7508 815.645 67.4314 Q809.326 61.1119 809.326 50.3365 Q809.326 39.1965 815.321 32.6746 Q821.357 26.1121 831.565 26.1121 Q840.72 26.1121 846.027 32.0264 Q851.374 37.9003 851.374 48.0275 M843.921 45.84 Q843.839 39.7232 840.477 36.0774 Q837.156 32.4315 831.646 32.4315 Q825.408 32.4315 821.641 35.9558 Q817.914 39.4801 817.347 45.8805 L843.921 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M926.559 14.0809 L926.559 22.0612 Q921.9 19.8332 917.769 18.7395 Q913.637 17.6457 909.788 17.6457 Q903.104 17.6457 899.458 20.2383 Q895.853 22.8309 895.853 27.611 Q895.853 31.6214 898.243 33.6873 Q900.674 35.7128 907.398 36.9686 L912.34 37.9813 Q921.495 39.7232 925.83 44.1387 Q930.205 48.5136 930.205 55.8863 Q930.205 64.6767 924.29 69.2137 Q918.417 73.7508 907.034 73.7508 Q902.74 73.7508 897.879 72.7785 Q893.058 71.8063 887.873 69.9024 L887.873 61.4765 Q892.855 64.2716 897.636 65.6895 Q902.416 67.1073 907.034 67.1073 Q914.042 67.1073 917.85 64.3527 Q921.657 61.598 921.657 56.4939 Q921.657 52.0379 918.903 49.5264 Q916.189 47.0148 909.95 45.759 L904.968 44.7868 Q895.813 42.9639 891.721 39.075 Q887.63 35.1862 887.63 28.2591 Q887.63 20.2383 893.261 15.6203 Q898.932 11.0023 908.857 11.0023 Q913.11 11.0023 917.525 11.7719 Q921.941 12.5416 926.559 14.0809 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M980.355 45.1919 L980.355 72.576 L972.901 72.576 L972.901 45.4349 Q972.901 38.994 970.39 35.7938 Q967.878 32.5936 962.855 32.5936 Q956.819 32.5936 953.335 36.4419 Q949.852 40.2903 949.852 46.9338 L949.852 72.576 L942.357 72.576 L942.357 9.54393 L949.852 9.54393 L949.852 34.2544 Q952.525 30.163 956.131 28.1376 Q959.776 26.1121 964.516 26.1121 Q972.334 26.1121 976.345 30.9732 Q980.355 35.7938 980.355 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M994.452 54.671 L994.452 27.2059 L1001.91 27.2059 L1001.91 54.3874 Q1001.91 60.8284 1004.42 64.0691 Q1006.93 67.2693 1011.95 67.2693 Q1017.99 67.2693 1021.47 63.421 Q1025 59.5726 1025 52.9291 L1025 27.2059 L1032.45 27.2059 L1032.45 72.576 L1025 72.576 L1025 65.6084 Q1022.28 69.7404 1018.68 71.7658 Q1015.11 73.7508 1010.37 73.7508 Q1002.55 73.7508 998.503 68.8897 Q994.452 64.0286 994.452 54.671 M1013.21 26.1121 L1013.21 26.1121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1055.18 14.324 L1055.18 27.2059 L1070.53 27.2059 L1070.53 32.9987 L1055.18 32.9987 L1055.18 57.6282 Q1055.18 63.1779 1056.67 64.7578 Q1058.21 66.3376 1062.87 66.3376 L1070.53 66.3376 L1070.53 72.576 L1062.87 72.576 Q1054.24 72.576 1050.96 69.3758 Q1047.68 66.1351 1047.68 57.6282 L1047.68 32.9987 L1042.21 32.9987 L1042.21 27.2059 L1047.68 27.2059 L1047.68 14.324 L1055.18 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1087.7 14.324 L1087.7 27.2059 L1103.06 27.2059 L1103.06 32.9987 L1087.7 32.9987 L1087.7 57.6282 Q1087.7 63.1779 1089.2 64.7578 Q1090.74 66.3376 1095.4 66.3376 L1103.06 66.3376 L1103.06 72.576 L1095.4 72.576 Q1086.77 72.576 1083.49 69.3758 Q1080.21 66.1351 1080.21 57.6282 L1080.21 32.9987 L1074.74 32.9987 L1074.74 27.2059 L1080.21 27.2059 L1080.21 14.324 L1087.7 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1112.86 9.54393 L1120.31 9.54393 L1120.31 72.576 L1112.86 72.576 L1112.86 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1174.72 48.0275 L1174.72 51.6733 L1140.45 51.6733 Q1140.93 59.3701 1145.06 63.421 Q1149.24 67.4314 1156.65 67.4314 Q1160.94 67.4314 1164.95 66.3781 Q1169.01 65.3249 1172.98 63.2184 L1172.98 70.267 Q1168.97 71.9684 1164.75 72.8596 Q1160.54 73.7508 1156.2 73.7508 Q1145.35 73.7508 1138.99 67.4314 Q1132.67 61.1119 1132.67 50.3365 Q1132.67 39.1965 1138.66 32.6746 Q1144.7 26.1121 1154.91 26.1121 Q1164.06 26.1121 1169.37 32.0264 Q1174.72 37.9003 1174.72 48.0275 M1167.26 45.84 Q1167.18 39.7232 1163.82 36.0774 Q1160.5 32.4315 1154.99 32.4315 Q1148.75 32.4315 1144.98 35.9558 Q1141.26 39.4801 1140.69 45.8805 L1167.26 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1242.33 44.2197 Q1244.96 45.1109 1247.43 48.0275 Q1249.94 50.9442 1252.45 56.0483 L1260.76 72.576 L1251.97 72.576 L1244.23 57.061 Q1241.23 50.9847 1238.4 48.9997 Q1235.6 47.0148 1230.74 47.0148 L1221.83 47.0148 L1221.83 72.576 L1213.65 72.576 L1213.65 12.096 L1232.12 12.096 Q1242.49 12.096 1247.59 16.4305 Q1252.7 20.7649 1252.7 29.5149 Q1252.7 35.2267 1250.02 38.994 Q1247.39 42.7613 1242.33 44.2197 M1221.83 18.8205 L1221.83 40.2903 L1232.12 40.2903 Q1238.03 40.2903 1241.03 37.5762 Q1244.07 34.8216 1244.07 29.5149 Q1244.07 24.2082 1241.03 21.5346 Q1238.03 18.8205 1232.12 18.8205 L1221.83 18.8205 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1306.05 48.0275 L1306.05 51.6733 L1271.78 51.6733 Q1272.26 59.3701 1276.4 63.421 Q1280.57 67.4314 1287.98 67.4314 Q1292.27 67.4314 1296.28 66.3781 Q1300.34 65.3249 1304.31 63.2184 L1304.31 70.267 Q1300.3 71.9684 1296.08 72.8596 Q1291.87 73.7508 1287.54 73.7508 Q1276.68 73.7508 1270.32 67.4314 Q1264 61.1119 1264 50.3365 Q1264 39.1965 1269.99 32.6746 Q1276.03 26.1121 1286.24 26.1121 Q1295.39 26.1121 1300.7 32.0264 Q1306.05 37.9003 1306.05 48.0275 M1298.59 45.84 Q1298.51 39.7232 1295.15 36.0774 Q1291.83 32.4315 1286.32 32.4315 Q1280.08 32.4315 1276.31 35.9558 Q1272.59 39.4801 1272.02 45.8805 L1298.59 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1357.09 48.0275 L1357.09 51.6733 L1322.82 51.6733 Q1323.3 59.3701 1327.44 63.421 Q1331.61 67.4314 1339.02 67.4314 Q1343.32 67.4314 1347.33 66.3781 Q1351.38 65.3249 1355.35 63.2184 L1355.35 70.267 Q1351.34 71.9684 1347.12 72.8596 Q1342.91 73.7508 1338.58 73.7508 Q1327.72 73.7508 1321.36 67.4314 Q1315.04 61.1119 1315.04 50.3365 Q1315.04 39.1965 1321.04 32.6746 Q1327.07 26.1121 1337.28 26.1121 Q1346.44 26.1121 1351.74 32.0264 Q1357.09 37.9003 1357.09 48.0275 M1349.64 45.84 Q1349.55 39.7232 1346.19 36.0774 Q1342.87 32.4315 1337.36 32.4315 Q1331.12 32.4315 1327.36 35.9558 Q1323.63 39.4801 1323.06 45.8805 L1349.64 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1407.04 45.1919 L1407.04 72.576 L1399.58 72.576 L1399.58 45.4349 Q1399.58 38.994 1397.07 35.7938 Q1394.56 32.5936 1389.54 32.5936 Q1383.5 32.5936 1380.02 36.4419 Q1376.53 40.2903 1376.53 46.9338 L1376.53 72.576 L1369.04 72.576 L1369.04 27.2059 L1376.53 27.2059 L1376.53 34.2544 Q1379.21 30.163 1382.81 28.1376 Q1386.46 26.1121 1391.2 26.1121 Q1399.02 26.1121 1403.03 30.9732 Q1407.04 35.7938 1407.04 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1429.28 14.324 L1429.28 27.2059 L1444.63 27.2059 L1444.63 32.9987 L1429.28 32.9987 L1429.28 57.6282 Q1429.28 63.1779 1430.78 64.7578 Q1432.31 66.3376 1436.97 66.3376 L1444.63 66.3376 L1444.63 72.576 L1436.97 72.576 Q1428.34 72.576 1425.06 69.3758 Q1421.78 66.1351 1421.78 57.6282 L1421.78 32.9987 L1416.31 32.9987 L1416.31 27.2059 L1421.78 27.2059 L1421.78 14.324 L1429.28 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1480.72 34.1734 Q1479.47 33.4443 1477.97 33.1202 Q1476.51 32.7556 1474.73 32.7556 Q1468.41 32.7556 1465.01 36.8875 Q1461.64 40.9789 1461.64 48.6757 L1461.64 72.576 L1454.15 72.576 L1454.15 27.2059 L1461.64 27.2059 L1461.64 34.2544 Q1463.99 30.1225 1467.76 28.1376 Q1471.53 26.1121 1476.91 26.1121 Q1477.68 26.1121 1478.62 26.2337 Q1479.55 26.3147 1480.68 26.5172 L1480.72 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1507.42 76.7889 Q1504.26 84.8907 1501.26 87.3618 Q1498.26 89.8329 1493.24 89.8329 L1487.29 89.8329 L1487.29 83.5945 L1491.66 83.5945 Q1494.74 83.5945 1496.44 82.1361 Q1498.14 80.6778 1500.21 75.2496 L1501.54 71.8468 L1483.19 27.2059 L1491.09 27.2059 L1505.27 62.6918 L1519.45 27.2059 L1527.35 27.2059 L1507.42 76.7889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1555.95 12.096 L1607.11 12.096 L1607.11 18.9825 L1585.64 18.9825 L1585.64 72.576 L1577.42 72.576 L1577.42 18.9825 L1555.95 18.9825 L1555.95 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1628.78 34.1734 Q1627.53 33.4443 1626.03 33.1202 Q1624.57 32.7556 1622.79 32.7556 Q1616.47 32.7556 1613.07 36.8875 Q1609.7 40.9789 1609.7 48.6757 L1609.7 72.576 L1602.21 72.576 L1602.21 27.2059 L1609.7 27.2059 L1609.7 34.2544 Q1612.05 30.1225 1615.82 28.1376 Q1619.59 26.1121 1624.98 26.1121 Q1625.75 26.1121 1626.68 26.2337 Q1627.61 26.3147 1628.74 26.5172 L1628.78 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1657.22 49.7694 Q1648.19 49.7694 1644.7 51.8354 Q1641.22 53.9013 1641.22 58.8839 Q1641.22 62.8538 1643.81 65.2034 Q1646.45 67.5124 1650.94 67.5124 Q1657.14 67.5124 1660.87 63.1374 Q1664.63 58.7219 1664.63 51.4303 L1664.63 49.7694 L1657.22 49.7694 M1672.09 46.6907 L1672.09 72.576 L1664.63 72.576 L1664.63 65.6895 Q1662.08 69.8214 1658.27 71.8063 Q1654.47 73.7508 1648.96 73.7508 Q1641.99 73.7508 1637.86 69.8619 Q1633.77 65.9325 1633.77 59.3701 Q1633.77 51.7138 1638.87 47.825 Q1644.01 43.9361 1654.18 43.9361 L1664.63 43.9361 L1664.63 43.2069 Q1664.63 38.0623 1661.23 35.2672 Q1657.87 32.4315 1651.75 32.4315 Q1647.86 32.4315 1644.18 33.3632 Q1640.49 34.295 1637.09 36.1584 L1637.09 29.2718 Q1641.18 27.692 1645.03 26.9223 Q1648.88 26.1121 1652.52 26.1121 Q1662.37 26.1121 1667.23 31.2163 Q1672.09 36.3204 1672.09 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1687.44 27.2059 L1694.89 27.2059 L1694.89 73.3862 Q1694.89 82.0551 1691.57 85.944 Q1688.29 89.8329 1680.96 89.8329 L1678.12 89.8329 L1678.12 83.5134 L1680.11 83.5134 Q1684.36 83.5134 1685.9 81.5285 Q1687.44 79.5841 1687.44 73.3862 L1687.44 27.2059 M1687.44 9.54393 L1694.89 9.54393 L1694.89 18.9825 L1687.44 18.9825 L1687.44 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1749.3 48.0275 L1749.3 51.6733 L1715.03 51.6733 Q1715.51 59.3701 1719.64 63.421 Q1723.82 67.4314 1731.23 67.4314 Q1735.52 67.4314 1739.53 66.3781 Q1743.59 65.3249 1747.56 63.2184 L1747.56 70.267 Q1743.55 71.9684 1739.33 72.8596 Q1735.12 73.7508 1730.78 73.7508 Q1719.93 73.7508 1713.57 67.4314 Q1707.25 61.1119 1707.25 50.3365 Q1707.25 39.1965 1713.24 32.6746 Q1719.28 26.1121 1729.49 26.1121 Q1738.64 26.1121 1743.95 32.0264 Q1749.3 37.9003 1749.3 48.0275 M1741.84 45.84 Q1741.76 39.7232 1738.4 36.0774 Q1735.08 32.4315 1729.57 32.4315 Q1723.33 32.4315 1719.56 35.9558 Q1715.84 39.4801 1715.27 45.8805 L1741.84 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1794.18 28.9478 L1794.18 35.9153 Q1791.02 34.1734 1787.82 33.3227 Q1784.66 32.4315 1781.42 32.4315 Q1774.17 32.4315 1770.16 37.0496 Q1766.15 41.6271 1766.15 49.9314 Q1766.15 58.2358 1770.16 62.8538 Q1774.17 67.4314 1781.42 67.4314 Q1784.66 67.4314 1787.82 66.5807 Q1791.02 65.6895 1794.18 63.9476 L1794.18 70.8341 Q1791.06 72.2924 1787.7 73.0216 Q1784.38 73.7508 1780.61 73.7508 Q1770.36 73.7508 1764.33 67.3098 Q1758.29 60.8689 1758.29 49.9314 Q1758.29 38.832 1764.37 32.472 Q1770.48 26.1121 1781.1 26.1121 Q1784.54 26.1121 1787.82 26.8413 Q1791.1 27.5299 1794.18 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1814.52 14.324 L1814.52 27.2059 L1829.87 27.2059 L1829.87 32.9987 L1814.52 32.9987 L1814.52 57.6282 Q1814.52 63.1779 1816.02 64.7578 Q1817.56 66.3376 1822.21 66.3376 L1829.87 66.3376 L1829.87 72.576 L1822.21 72.576 Q1813.59 72.576 1810.3 69.3758 Q1807.02 66.1351 1807.02 57.6282 L1807.02 32.9987 L1801.55 32.9987 L1801.55 27.2059 L1807.02 27.2059 L1807.02 14.324 L1814.52 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1857.25 32.4315 Q1851.26 32.4315 1847.78 37.1306 Q1844.29 41.7891 1844.29 49.9314 Q1844.29 58.0738 1847.73 62.7728 Q1851.22 67.4314 1857.25 67.4314 Q1863.21 67.4314 1866.69 62.7323 Q1870.18 58.0333 1870.18 49.9314 Q1870.18 41.8701 1866.69 37.1711 Q1863.21 32.4315 1857.25 32.4315 M1857.25 26.1121 Q1866.98 26.1121 1872.53 32.4315 Q1878.08 38.7509 1878.08 49.9314 Q1878.08 61.0714 1872.53 67.4314 Q1866.98 73.7508 1857.25 73.7508 Q1847.49 73.7508 1841.94 67.4314 Q1836.43 61.0714 1836.43 49.9314 Q1836.43 38.7509 1841.94 32.4315 Q1847.49 26.1121 1857.25 26.1121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1916.72 34.1734 Q1915.47 33.4443 1913.97 33.1202 Q1912.51 32.7556 1910.73 32.7556 Q1904.41 32.7556 1901 36.8875 Q1897.64 40.9789 1897.64 48.6757 L1897.64 72.576 L1890.15 72.576 L1890.15 27.2059 L1897.64 27.2059 L1897.64 34.2544 Q1899.99 30.1225 1903.76 28.1376 Q1907.53 26.1121 1912.91 26.1121 Q1913.68 26.1121 1914.62 26.2337 Q1915.55 26.3147 1916.68 26.5172 L1916.72 34.1734 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip750)" d="M1943.42 76.7889 Q1940.26 84.8907 1937.26 87.3618 Q1934.26 89.8329 1929.24 89.8329 L1923.28 89.8329 L1923.28 83.5945 L1927.66 83.5945 Q1930.74 83.5945 1932.44 82.1361 Q1934.14 80.6778 1936.21 75.2496 L1937.54 71.8468 L1919.19 27.2059 L1927.09 27.2059 L1941.27 62.6918 L1955.45 27.2059 L1963.35 27.2059 L1943.42 76.7889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip752)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="802.382,531.57 804.637,538.307 806.893,545.043 809.15,551.773 811.408,558.496 813.666,565.21 815.926,571.911 818.187,578.598 820.449,585.267 822.712,591.915 824.976,598.539 827.242,605.135 829.508,611.698 831.776,618.226 834.046,624.712 836.317,631.152 838.59,637.542 840.864,643.875 843.14,650.146 845.418,656.349 847.698,662.476 849.98,668.521 852.264,674.477 854.551,680.336 856.84,686.09 859.131,691.731 861.425,697.251 863.722,702.639 866.022,707.889 868.324,712.989 870.63,717.931 872.939,722.706 875.252,727.305 877.568,731.717 879.888,735.935 882.211,739.949 884.539,743.752 886.87,747.335 889.206,750.691 891.546,753.815 893.89,756.7 896.238,759.342 898.591,761.737 900.948,763.883 903.31,765.777 905.676,767.421 908.046,768.814 910.421,769.959 912.8,770.859 915.184,771.519 917.571,771.943 919.963,772.14 922.359,772.116 924.759,771.879 927.163,771.441 929.571,770.81 931.983,769.997 934.398,769.014 936.816,767.873 939.238,766.586 941.663,765.166 944.09,763.624 946.521,761.974 948.955,760.229 951.391,758.401 953.83,756.503 956.271,754.547 958.714,752.545 961.159,750.509 963.607,748.451 966.056,746.382 968.508,744.313 970.961,742.254 973.416,740.215 975.872,738.207 978.33,736.238 980.789,734.317 983.25,732.452 985.712,730.652 988.175,728.925 990.639,727.276 993.105,725.714 995.571,724.245 998.039,722.873 1000.51,721.606 1002.98,720.448 1005.45,719.404 1007.92,718.478 1010.39,717.674 1012.86,716.997 1015.34,716.449 1017.81,716.032 1020.29,715.751 1022.76,715.607 1025.24,715.601 1027.72,715.735 1030.2,716.011 1032.68,716.429 1035.16,716.989 1037.64,717.692 1040.12,718.537 1042.6,719.522 1045.08,720.648 1047.56,721.913 1050.05,723.313 1052.53,724.848 1055.02,726.513 1057.51,728.307 1059.99,730.224 1062.48,732.261 1064.97,734.413 1067.46,736.675 1069.95,739.04 1072.44,741.502 1074.93,744.055 1077.42,746.692 1079.92,749.403 1082.41,752.181 1084.91,755.016 1087.4,757.899 1089.9,760.819 1092.4,763.767 1094.9,766.73 1097.4,769.697 1099.9,772.656 1102.4,775.594 1104.9,778.499 1107.4,781.358 1109.91,784.158 1112.42,786.885 1114.92,789.527 1117.43,792.071 1119.94,794.505 1122.45,796.816 1124.96,798.993 1127.47,801.026 1129.98,802.905 1132.5,804.621 1135.01,806.167 1137.53,807.536 1140.04,808.725 1142.56,809.729 1145.07,810.547 1147.59,811.178 1150.11,811.623 1152.63,811.887 1155.15,811.972 1157.67,811.884 1160.19,811.631 1162.71,811.222 1165.22,810.665 1167.74,809.971 1170.26,809.153 1172.78,808.221 1175.3,807.19 1177.82,806.072 1180.34,804.882 1182.86,803.632 1185.38,802.338 1187.9,801.013 1190.42,799.671 1192.93,798.325 1195.45,796.989 1197.97,795.675 1200.48,794.397 1203,793.164 1205.51,791.989 1208.03,790.883 1210.54,789.854 1213.05,788.912 1215.56,788.064 1218.07,787.32 1220.58,786.684 1223.09,786.163 1225.6,785.762 1228.11,785.485 1230.62,785.335 1233.12,785.314 1235.63,785.424 1238.13,785.665 1240.64,786.038 1243.14,786.54 1245.64,787.17 1248.14,787.925 1250.64,788.8 1253.14,789.791 1255.64,790.891 1258.14,792.096 1260.63,793.396 1263.13,794.783 1265.62,796.249 1268.12,797.784 1270.61,799.377 1273.1,801.017 1275.59,802.692 1278.07,804.39 1280.56,806.097 1283.04,807.803 1285.53,809.492 1288.01,811.152 1290.49,812.769 1292.97,814.332 1295.44,815.827 1297.92,817.242 1300.39,818.567 1302.86,819.79 1305.33,820.903 1307.8,821.897 1310.26,822.766 1312.72,823.503 1315.18,824.105 1317.64,824.57 1320.09,824.896 1322.54,825.083 1324.99,825.136 1327.44,825.056 1329.88,824.849 1332.32,824.523 1334.76,824.084 1337.19,823.543 1339.63,822.908 1342.05,822.191 1344.48,821.404 1346.9,820.559 1349.32,819.668 1351.74,818.743 1354.15,817.799 1356.56,816.847 1358.97,815.899 1361.37,814.969 1363.78,814.068 1366.17,813.206 1368.57,812.395 1370.96,811.643 1373.35,810.961 1375.74,810.355 1378.12,809.833 1380.5,809.401 1382.87,809.063 1385.25,808.825 1387.62,808.688 1389.98,808.655 1392.35,808.725 1394.71,808.9 1397.07,809.177 1399.42,809.553 1401.77,810.025 1404.12,810.588 1406.46,811.236 1408.8,811.963 1411.14,812.762 1413.47,813.623 1415.8,814.538 1418.13,815.496 1420.45,816.488 1422.77,817.503 1425.08,818.529 1427.39,819.556 1429.7,820.571 1432,821.565 1434.3,822.525 1436.59,823.441 1438.88,824.304 1441.16,825.103 1443.44,825.832 1445.72,826.481 1447.99,827.045 1450.25,827.52 1452.51,827.9 1454.77,828.185 1457.02,828.372 1459.26,828.462 1461.5,828.457 1463.73,828.361 1465.96,828.177 1468.18,827.911 1470.4,827.57 1472.61,827.161 1474.81,826.693 1477.01,826.175 1479.21,825.617 1481.4,825.029 1483.58,824.42 1485.76,823.802 1487.93,823.184 1490.09,822.577 1492.26,821.99 1494.41,821.432 1496.56,820.913 1498.7,820.439 1500.84,820.018 1502.98,819.656 1505.1,819.359 1507.23,819.132 1509.34,818.977 1511.45,818.897 1513.56,818.894 1515.66,818.967 1517.75,819.117 1519.84,819.341 1521.92,819.637 1524,820 1526.07,820.428 1528.13,820.913 1530.19,821.449 1532.24,822.031 1534.29,822.65 1536.33,823.299 1538.36,823.969 1540.39,824.651 1542.41,825.338 1544.43,826.019 1546.44,826.688 1548.44,827.335 1550.43,827.954 1552.42,828.536 1554.4,829.076 1556.37,829.567 1558.34,830.004 1560.29,830.385 1562.25,830.706 1564.19,830.965 1566.13,831.162 1568.05,831.297 1569.98,831.372 1571.89,831.388 1573.79,831.351 1575.69,831.263 1577.58,831.131 1579.46,830.959 1581.34,830.755 1583.2,830.526 1585.06,830.278 1586.91,830.02 1588.76,829.758 1590.59,829.5 1592.42,829.253 1594.24,829.024 1596.05,828.821 1597.86,828.648 1599.65,828.512 1601.44,828.418 1603.22,828.368 1605,828.368 1606.76,828.42 1608.52,828.524 1610.27,828.684 1612.01,828.897 1613.74,829.164 1615.47,829.484 1617.19,829.853 1618.9,830.269 1620.6,830.728 1622.29,831.227 1623.98,831.759 1625.66,832.321 1627.32,832.906 1628.98,833.509 1630.64,834.123 1632.28,834.744 1633.91,835.364 1635.54,835.979 1637.16,836.582 1638.77,837.17 1640.37,837.737 1641.96,838.279 1643.54,838.794 1645.11,839.278 1646.67,839.73 1648.23,840.148 1649.77,840.532 1651.31,840.883 1652.83,841.202 1654.35,841.489 1655.86,841.749 1657.35,841.984 1658.84,842.198 1660.32,842.394 1661.79,842.578 1663.25,842.753 1664.7,842.926 1666.14,843.101 1667.57,843.282 1669,843.476 1670.41,843.686 1671.81,843.917 1673.21,844.174 1674.59,844.459 1675.97,844.776 1677.33,845.128 1678.69,845.518 1680.03,845.946 1681.37,846.413 1682.7,846.921 1684.02,847.469 1685.33,848.056 1686.63,848.681 1687.92,849.343 1689.2,850.038 1690.47,850.765 1691.73,851.52 1692.98,852.3 1694.22,853.102 1695.46,853.921 1696.68,854.755 1697.89,855.598 1699.09,856.449 1700.29,857.302 1701.47,858.155 1702.64,859.005 1703.8,859.849 1704.96,860.685 1706.1,861.511 1707.23,862.325 1708.36,863.127 1709.47,863.917 1710.57,864.695 1711.66,865.461 1712.74,866.217 1713.82,866.965 1714.88,867.706 1715.93,868.443 1716.97,869.179 1718,869.917 1719.02,870.66 1720.04,871.413 1721.04,872.177 1722.03,872.958 1723.01,873.759 1723.98,874.582 1724.94,875.431 1725.89,876.309 1726.83,877.22 1727.76,878.164 1728.69,879.145 1729.6,880.164 1730.5,881.221 1731.39,882.319 1732.27,883.456 1733.14,884.634 1734.01,885.851 1734.86,887.106 1735.7,888.399 1736.54,889.727 1737.36,891.088 1738.17,892.48 1738.98,893.9 1739.77,895.345 1740.56,896.813 1741.33,898.299 1742.1,899.802 1742.85,901.317 1743.6,902.843 1744.33,904.376 1745.06,905.913 1745.77,907.453 1746.48,908.993 1747.18,910.532 1747.87,912.068 1748.54,913.602 1749.21,915.134 1749.87,916.663 1750.52,918.191 1751.16,919.719 1751.79,921.25 1752.41,922.787 1753.02,924.333 1753.62,925.892 1754.21,927.468 1754.8,929.066 1755.37,930.692 1755.94,932.35 1756.49,934.047 1757.04,935.789 1757.58,937.581 1758.11,939.429 1758.63,941.339 1759.14,943.317 1759.64,945.368 1760.14,947.497 1760.63,949.709 1761.11,952.007 1761.58,954.394 1762.04,956.873 1762.49,959.445 1762.94,962.11 1763.38,964.868 1763.81,967.716 1764.24,970.652 1764.65,973.67 1765.06,976.763 1765.46,979.925 1765.86,983.144 1766.25,986.408 1766.63,989.706 1767,993.02 1767.36,996.334 1767.72,999.629 1768.07,1002.88 1768.42,1006.08 1768.75,1009.19 1769.08,1012.2 1769.41,1015.08 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/78651b37.svg b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/78651b37.svg
new file mode 100644
index 00000000000..38f200b35d6
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/78651b37.svg
@@ -0,0 +1,182 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="700" height="700" viewBox="0 0 2800 2800">
+<defs>
+  <clipPath id="clip630">
+    <rect x="0" y="0" width="2800" height="2800"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip630)" d="M0 2800 L2800 2800 L2800 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip631">
+    <rect x="560" y="280" width="1961" height="1961"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip630)" d="M159.765 790.501 L1352.76 790.501 L1352.76 153.712 L159.765 153.712  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip632">
+    <rect x="159" y="153" width="1194" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="193.529,790.501 193.529,153.712 "/>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="473.774,790.501 473.774,153.712 "/>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="754.018,790.501 754.018,153.712 "/>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1034.26,790.501 1034.26,153.712 "/>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1314.51,790.501 1314.51,153.712 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,790.501 1352.76,790.501 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="193.529,790.501 193.529,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="473.774,790.501 473.774,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="754.018,790.501 754.018,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1034.26,790.501 1034.26,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1314.51,790.501 1314.51,771.604 "/>
+<path clip-path="url(#clip630)" d="M193.529 824.62 Q189.918 824.62 188.089 828.185 Q186.283 831.726 186.283 838.856 Q186.283 845.962 188.089 849.527 Q189.918 853.069 193.529 853.069 Q197.163 853.069 198.969 849.527 Q200.797 845.962 200.797 838.856 Q200.797 831.726 198.969 828.185 Q197.163 824.62 193.529 824.62 M193.529 820.916 Q199.339 820.916 202.394 825.523 Q205.473 830.106 205.473 838.856 Q205.473 847.583 202.394 852.189 Q199.339 856.773 193.529 856.773 Q187.719 856.773 184.64 852.189 Q181.584 847.583 181.584 838.856 Q181.584 830.106 184.64 825.523 Q187.719 820.916 193.529 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M433.392 821.541 L451.748 821.541 L451.748 825.476 L437.674 825.476 L437.674 833.949 Q438.693 833.601 439.711 833.439 Q440.73 833.254 441.748 833.254 Q447.535 833.254 450.915 836.425 Q454.295 839.597 454.295 845.013 Q454.295 850.592 450.822 853.694 Q447.35 856.773 441.031 856.773 Q438.855 856.773 436.586 856.402 Q434.341 856.032 431.934 855.291 L431.934 850.592 Q434.017 851.726 436.239 852.282 Q438.461 852.837 440.938 852.837 Q444.943 852.837 447.281 850.731 Q449.619 848.624 449.619 845.013 Q449.619 841.402 447.281 839.296 Q444.943 837.189 440.938 837.189 Q439.063 837.189 437.188 837.606 Q435.336 838.023 433.392 838.902 L433.392 821.541 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M473.507 824.62 Q469.896 824.62 468.068 828.185 Q466.262 831.726 466.262 838.856 Q466.262 845.962 468.068 849.527 Q469.896 853.069 473.507 853.069 Q477.142 853.069 478.947 849.527 Q480.776 845.962 480.776 838.856 Q480.776 831.726 478.947 828.185 Q477.142 824.62 473.507 824.62 M473.507 820.916 Q479.318 820.916 482.373 825.523 Q485.452 830.106 485.452 838.856 Q485.452 847.583 482.373 852.189 Q479.318 856.773 473.507 856.773 Q467.697 856.773 464.619 852.189 Q461.563 847.583 461.563 838.856 Q461.563 830.106 464.619 825.523 Q467.697 820.916 473.507 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M503.669 824.62 Q500.058 824.62 498.23 828.185 Q496.424 831.726 496.424 838.856 Q496.424 845.962 498.23 849.527 Q500.058 853.069 503.669 853.069 Q507.304 853.069 509.109 849.527 Q510.938 845.962 510.938 838.856 Q510.938 831.726 509.109 828.185 Q507.304 824.62 503.669 824.62 M503.669 820.916 Q509.479 820.916 512.535 825.523 Q515.614 830.106 515.614 838.856 Q515.614 847.583 512.535 852.189 Q509.479 856.773 503.669 856.773 Q497.859 856.773 494.78 852.189 Q491.725 847.583 491.725 838.856 Q491.725 830.106 494.78 825.523 Q497.859 820.916 503.669 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M698.544 852.166 L706.183 852.166 L706.183 825.8 L697.873 827.467 L697.873 823.208 L706.137 821.541 L710.813 821.541 L710.813 852.166 L718.452 852.166 L718.452 856.101 L698.544 856.101 L698.544 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M737.896 824.62 Q734.285 824.62 732.456 828.185 Q730.651 831.726 730.651 838.856 Q730.651 845.962 732.456 849.527 Q734.285 853.069 737.896 853.069 Q741.53 853.069 743.336 849.527 Q745.164 845.962 745.164 838.856 Q745.164 831.726 743.336 828.185 Q741.53 824.62 737.896 824.62 M737.896 820.916 Q743.706 820.916 746.762 825.523 Q749.84 830.106 749.84 838.856 Q749.84 847.583 746.762 852.189 Q743.706 856.773 737.896 856.773 Q732.086 856.773 729.007 852.189 Q725.951 847.583 725.951 838.856 Q725.951 830.106 729.007 825.523 Q732.086 820.916 737.896 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M768.058 824.62 Q764.447 824.62 762.618 828.185 Q760.812 831.726 760.812 838.856 Q760.812 845.962 762.618 849.527 Q764.447 853.069 768.058 853.069 Q771.692 853.069 773.498 849.527 Q775.326 845.962 775.326 838.856 Q775.326 831.726 773.498 828.185 Q771.692 824.62 768.058 824.62 M768.058 820.916 Q773.868 820.916 776.923 825.523 Q780.002 830.106 780.002 838.856 Q780.002 847.583 776.923 852.189 Q773.868 856.773 768.058 856.773 Q762.248 856.773 759.169 852.189 Q756.113 847.583 756.113 838.856 Q756.113 830.106 759.169 825.523 Q762.248 820.916 768.058 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M798.22 824.62 Q794.609 824.62 792.78 828.185 Q790.974 831.726 790.974 838.856 Q790.974 845.962 792.78 849.527 Q794.609 853.069 798.22 853.069 Q801.854 853.069 803.659 849.527 Q805.488 845.962 805.488 838.856 Q805.488 831.726 803.659 828.185 Q801.854 824.62 798.22 824.62 M798.22 820.916 Q804.03 820.916 807.085 825.523 Q810.164 830.106 810.164 838.856 Q810.164 847.583 807.085 852.189 Q804.03 856.773 798.22 856.773 Q792.409 856.773 789.331 852.189 Q786.275 847.583 786.275 838.856 Q786.275 830.106 789.331 825.523 Q792.409 820.916 798.22 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M978.789 852.166 L986.428 852.166 L986.428 825.8 L978.118 827.467 L978.118 823.208 L986.382 821.541 L991.057 821.541 L991.057 852.166 L998.696 852.166 L998.696 856.101 L978.789 856.101 L978.789 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1008.19 821.541 L1026.54 821.541 L1026.54 825.476 L1012.47 825.476 L1012.47 833.949 Q1013.49 833.601 1014.51 833.439 Q1015.52 833.254 1016.54 833.254 Q1022.33 833.254 1025.71 836.425 Q1029.09 839.597 1029.09 845.013 Q1029.09 850.592 1025.62 853.694 Q1022.15 856.773 1015.83 856.773 Q1013.65 856.773 1011.38 856.402 Q1009.14 856.032 1006.73 855.291 L1006.73 850.592 Q1008.81 851.726 1011.03 852.282 Q1013.26 852.837 1015.73 852.837 Q1019.74 852.837 1022.08 850.731 Q1024.41 848.624 1024.41 845.013 Q1024.41 841.402 1022.08 839.296 Q1019.74 837.189 1015.73 837.189 Q1013.86 837.189 1011.98 837.606 Q1010.13 838.023 1008.19 838.902 L1008.19 821.541 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1048.3 824.62 Q1044.69 824.62 1042.86 828.185 Q1041.06 831.726 1041.06 838.856 Q1041.06 845.962 1042.86 849.527 Q1044.69 853.069 1048.3 853.069 Q1051.94 853.069 1053.74 849.527 Q1055.57 845.962 1055.57 838.856 Q1055.57 831.726 1053.74 828.185 Q1051.94 824.62 1048.3 824.62 M1048.3 820.916 Q1054.11 820.916 1057.17 825.523 Q1060.25 830.106 1060.25 838.856 Q1060.25 847.583 1057.17 852.189 Q1054.11 856.773 1048.3 856.773 Q1042.49 856.773 1039.41 852.189 Q1036.36 847.583 1036.36 838.856 Q1036.36 830.106 1039.41 825.523 Q1042.49 820.916 1048.3 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1078.46 824.62 Q1074.85 824.62 1073.02 828.185 Q1071.22 831.726 1071.22 838.856 Q1071.22 845.962 1073.02 849.527 Q1074.85 853.069 1078.46 853.069 Q1082.1 853.069 1083.9 849.527 Q1085.73 845.962 1085.73 838.856 Q1085.73 831.726 1083.9 828.185 Q1082.1 824.62 1078.46 824.62 M1078.46 820.916 Q1084.27 820.916 1087.33 825.523 Q1090.41 830.106 1090.41 838.856 Q1090.41 847.583 1087.33 852.189 Q1084.27 856.773 1078.46 856.773 Q1072.65 856.773 1069.58 852.189 Q1066.52 847.583 1066.52 838.856 Q1066.52 830.106 1069.58 825.523 Q1072.65 820.916 1078.46 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1263.12 852.166 L1279.44 852.166 L1279.44 856.101 L1257.49 856.101 L1257.49 852.166 Q1260.16 849.411 1264.74 844.782 Q1269.35 840.129 1270.53 838.786 Q1272.77 836.263 1273.65 834.527 Q1274.55 832.768 1274.55 831.078 Q1274.55 828.324 1272.61 826.587 Q1270.69 824.851 1267.59 824.851 Q1265.39 824.851 1262.93 825.615 Q1260.5 826.379 1257.73 827.93 L1257.73 823.208 Q1260.55 822.074 1263 821.495 Q1265.46 820.916 1267.49 820.916 Q1272.86 820.916 1276.06 823.601 Q1279.25 826.287 1279.25 830.777 Q1279.25 832.907 1278.44 834.828 Q1277.66 836.726 1275.55 839.319 Q1274.97 839.99 1271.87 843.208 Q1268.77 846.402 1263.12 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1299.25 824.62 Q1295.64 824.62 1293.81 828.185 Q1292.01 831.726 1292.01 838.856 Q1292.01 845.962 1293.81 849.527 Q1295.64 853.069 1299.25 853.069 Q1302.89 853.069 1304.69 849.527 Q1306.52 845.962 1306.52 838.856 Q1306.52 831.726 1304.69 828.185 Q1302.89 824.62 1299.25 824.62 M1299.25 820.916 Q1305.06 820.916 1308.12 825.523 Q1311.2 830.106 1311.2 838.856 Q1311.2 847.583 1308.12 852.189 Q1305.06 856.773 1299.25 856.773 Q1293.44 856.773 1290.36 852.189 Q1287.31 847.583 1287.31 838.856 Q1287.31 830.106 1290.36 825.523 Q1293.44 820.916 1299.25 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1329.42 824.62 Q1325.8 824.62 1323.98 828.185 Q1322.17 831.726 1322.17 838.856 Q1322.17 845.962 1323.98 849.527 Q1325.8 853.069 1329.42 853.069 Q1333.05 853.069 1334.86 849.527 Q1336.68 845.962 1336.68 838.856 Q1336.68 831.726 1334.86 828.185 Q1333.05 824.62 1329.42 824.62 M1329.42 820.916 Q1335.23 820.916 1338.28 825.523 Q1341.36 830.106 1341.36 838.856 Q1341.36 847.583 1338.28 852.189 Q1335.23 856.773 1329.42 856.773 Q1323.61 856.773 1320.53 852.189 Q1317.47 847.583 1317.47 838.856 Q1317.47 830.106 1320.53 825.523 Q1323.61 820.916 1329.42 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1359.58 824.62 Q1355.97 824.62 1354.14 828.185 Q1352.33 831.726 1352.33 838.856 Q1352.33 845.962 1354.14 849.527 Q1355.97 853.069 1359.58 853.069 Q1363.21 853.069 1365.02 849.527 Q1366.85 845.962 1366.85 838.856 Q1366.85 831.726 1365.02 828.185 Q1363.21 824.62 1359.58 824.62 M1359.58 820.916 Q1365.39 820.916 1368.44 825.523 Q1371.52 830.106 1371.52 838.856 Q1371.52 847.583 1368.44 852.189 Q1365.39 856.773 1359.58 856.773 Q1353.77 856.773 1350.69 852.189 Q1347.63 847.583 1347.63 838.856 Q1347.63 830.106 1350.69 825.523 Q1353.77 820.916 1359.58 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,705.729 1352.76,705.729 "/>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,538.856 1352.76,538.856 "/>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,371.983 1352.76,371.983 "/>
+<polyline clip-path="url(#clip632)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,205.109 1352.76,205.109 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,790.501 159.765,153.712 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,705.729 178.663,705.729 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,538.856 178.663,538.856 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,371.983 178.663,371.983 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,205.109 178.663,205.109 "/>
+<path clip-path="url(#clip630)" d="M51.3995 719.074 L59.0384 719.074 L59.0384 692.709 L50.7282 694.375 L50.7282 690.116 L58.9921 688.449 L63.668 688.449 L63.668 719.074 L71.3068 719.074 L71.3068 723.009 L51.3995 723.009 L51.3995 719.074 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M80.7512 717.13 L85.6354 717.13 L85.6354 723.009 L80.7512 723.009 L80.7512 717.13 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 691.528 Q102.209 691.528 100.381 695.093 Q98.5752 698.635 98.5752 705.764 Q98.5752 712.871 100.381 716.435 Q102.209 719.977 105.821 719.977 Q109.455 719.977 111.26 716.435 Q113.089 712.871 113.089 705.764 Q113.089 698.635 111.26 695.093 Q109.455 691.528 105.821 691.528 M105.821 687.824 Q111.631 687.824 114.686 692.431 Q117.765 697.014 117.765 705.764 Q117.765 714.491 114.686 719.097 Q111.631 723.681 105.821 723.681 Q100.01 723.681 96.9317 719.097 Q93.8761 714.491 93.8761 705.764 Q93.8761 697.014 96.9317 692.431 Q100.01 687.824 105.821 687.824 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M52.3949 552.201 L60.0337 552.201 L60.0337 525.835 L51.7236 527.502 L51.7236 523.243 L59.9874 521.576 L64.6633 521.576 L64.6633 552.201 L72.3022 552.201 L72.3022 556.136 L52.3949 556.136 L52.3949 552.201 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M81.7466 550.256 L86.6308 550.256 L86.6308 556.136 L81.7466 556.136 L81.7466 550.256 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M96.8622 521.576 L115.219 521.576 L115.219 525.511 L101.145 525.511 L101.145 533.983 Q102.163 533.636 103.182 533.474 Q104.2 533.289 105.219 533.289 Q111.006 533.289 114.385 536.46 Q117.765 539.632 117.765 545.048 Q117.765 550.627 114.293 553.729 Q110.82 556.807 104.501 556.807 Q102.325 556.807 100.057 556.437 Q97.8113 556.067 95.4039 555.326 L95.4039 550.627 Q97.4872 551.761 99.7094 552.317 Q101.932 552.872 104.408 552.872 Q108.413 552.872 110.751 550.766 Q113.089 548.659 113.089 545.048 Q113.089 541.437 110.751 539.331 Q108.413 537.224 104.408 537.224 Q102.534 537.224 100.659 537.641 Q98.8067 538.057 96.8622 538.937 L96.8622 521.576 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M54.6171 385.327 L70.9365 385.327 L70.9365 389.263 L48.9921 389.263 L48.9921 385.327 Q51.6541 382.573 56.2375 377.943 Q60.8439 373.29 62.0245 371.948 Q64.2698 369.425 65.1494 367.689 Q66.0522 365.929 66.0522 364.24 Q66.0522 361.485 64.1078 359.749 Q62.1865 358.013 59.0847 358.013 Q56.8856 358.013 54.4319 358.777 Q52.0014 359.541 49.2236 361.091 L49.2236 356.369 Q52.0477 355.235 54.5014 354.656 Q56.955 354.078 58.9921 354.078 Q64.3624 354.078 67.5568 356.763 Q70.7513 359.448 70.7513 363.939 Q70.7513 366.068 69.9411 367.99 Q69.1541 369.888 67.0476 372.48 Q66.4689 373.152 63.367 376.369 Q60.2652 379.564 54.6171 385.327 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M80.7512 383.383 L85.6354 383.383 L85.6354 389.263 L80.7512 389.263 L80.7512 383.383 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 357.781 Q102.209 357.781 100.381 361.346 Q98.5752 364.888 98.5752 372.017 Q98.5752 379.124 100.381 382.689 Q102.209 386.23 105.821 386.23 Q109.455 386.23 111.26 382.689 Q113.089 379.124 113.089 372.017 Q113.089 364.888 111.26 361.346 Q109.455 357.781 105.821 357.781 M105.821 354.078 Q111.631 354.078 114.686 358.684 Q117.765 363.267 117.765 372.017 Q117.765 380.744 114.686 385.351 Q111.631 389.934 105.821 389.934 Q100.01 389.934 96.9317 385.351 Q93.8761 380.744 93.8761 372.017 Q93.8761 363.267 96.9317 358.684 Q100.01 354.078 105.821 354.078 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M55.6125 218.454 L71.9318 218.454 L71.9318 222.389 L49.9875 222.389 L49.9875 218.454 Q52.6495 215.699 57.2328 211.07 Q61.8393 206.417 63.0198 205.074 Q65.2652 202.551 66.1448 200.815 Q67.0476 199.056 67.0476 197.366 Q67.0476 194.612 65.1032 192.875 Q63.1819 191.139 60.08 191.139 Q57.881 191.139 55.4273 191.903 Q52.9967 192.667 50.219 194.218 L50.219 189.496 Q53.043 188.362 55.4967 187.783 Q57.9504 187.204 59.9874 187.204 Q65.3578 187.204 68.5522 189.889 Q71.7466 192.574 71.7466 197.065 Q71.7466 199.195 70.9365 201.116 Q70.1494 203.014 68.0429 205.607 Q67.4642 206.278 64.3624 209.496 Q61.2606 212.69 55.6125 218.454 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M81.7466 216.51 L86.6308 216.51 L86.6308 222.389 L81.7466 222.389 L81.7466 216.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M96.8622 187.829 L115.219 187.829 L115.219 191.764 L101.145 191.764 L101.145 200.236 Q102.163 199.889 103.182 199.727 Q104.2 199.542 105.219 199.542 Q111.006 199.542 114.385 202.713 Q117.765 205.885 117.765 211.301 Q117.765 216.88 114.293 219.982 Q110.82 223.06 104.501 223.06 Q102.325 223.06 100.057 222.69 Q97.8113 222.32 95.4039 221.579 L95.4039 216.88 Q97.4872 218.014 99.7094 218.57 Q101.932 219.125 104.408 219.125 Q108.413 219.125 110.751 217.019 Q113.089 214.912 113.089 211.301 Q113.089 207.69 110.751 205.584 Q108.413 203.477 104.408 203.477 Q102.534 203.477 100.659 203.894 Q98.8067 204.311 96.8622 205.19 L96.8622 187.829 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M365.348 20.1573 L354.249 50.2555 L376.488 50.2555 L365.348 20.1573 M360.73 12.096 L370.007 12.096 L393.056 72.576 L384.549 72.576 L379.04 57.061 L351.778 57.061 L346.268 72.576 L337.64 72.576 L360.73 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M401.563 9.54393 L409.017 9.54393 L409.017 72.576 L401.563 72.576 L401.563 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M431.986 14.324 L431.986 27.2059 L447.338 27.2059 L447.338 32.9987 L431.986 32.9987 L431.986 57.6282 Q431.986 63.1779 433.484 64.7578 Q435.024 66.3376 439.682 66.3376 L447.338 66.3376 L447.338 72.576 L439.682 72.576 Q431.054 72.576 427.773 69.3758 Q424.491 66.1351 424.491 57.6282 L424.491 32.9987 L419.023 32.9987 L419.023 27.2059 L424.491 27.2059 L424.491 14.324 L431.986 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M457.142 27.2059 L464.595 27.2059 L464.595 72.576 L457.142 72.576 L457.142 27.2059 M457.142 9.54393 L464.595 9.54393 L464.595 18.9825 L457.142 18.9825 L457.142 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M487.564 14.324 L487.564 27.2059 L502.917 27.2059 L502.917 32.9987 L487.564 32.9987 L487.564 57.6282 Q487.564 63.1779 489.063 64.7578 Q490.602 66.3376 495.261 66.3376 L502.917 66.3376 L502.917 72.576 L495.261 72.576 Q486.632 72.576 483.351 69.3758 Q480.07 66.1351 480.07 57.6282 L480.07 32.9987 L474.601 32.9987 L474.601 27.2059 L480.07 27.2059 L480.07 14.324 L487.564 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M511.95 54.671 L511.95 27.2059 L519.404 27.2059 L519.404 54.3874 Q519.404 60.8284 521.916 64.0691 Q524.427 67.2693 529.45 67.2693 Q535.486 67.2693 538.97 63.421 Q542.494 59.5726 542.494 52.9291 L542.494 27.2059 L549.948 27.2059 L549.948 72.576 L542.494 72.576 L542.494 65.6084 Q539.78 69.7404 536.175 71.7658 Q532.61 73.7508 527.87 73.7508 Q520.052 73.7508 516.001 68.8897 Q511.95 64.0286 511.95 54.671 M530.706 26.1121 L530.706 26.1121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M595.156 34.0924 L595.156 9.54393 L602.61 9.54393 L602.61 72.576 L595.156 72.576 L595.156 65.7705 Q592.806 69.8214 589.201 71.8063 Q585.636 73.7508 580.613 73.7508 Q572.39 73.7508 567.205 67.1883 Q562.06 60.6258 562.06 49.9314 Q562.06 39.2371 567.205 32.6746 Q572.39 26.1121 580.613 26.1121 Q585.636 26.1121 589.201 28.0971 Q592.806 30.0415 595.156 34.0924 M569.757 49.9314 Q569.757 58.1548 573.119 62.8538 Q576.522 67.5124 582.436 67.5124 Q588.35 67.5124 591.753 62.8538 Q595.156 58.1548 595.156 49.9314 Q595.156 41.7081 591.753 37.0496 Q588.35 32.3505 582.436 32.3505 Q576.522 32.3505 573.119 37.0496 Q569.757 41.7081 569.757 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M656.77 48.0275 L656.77 51.6733 L622.5 51.6733 Q622.986 59.3701 627.118 63.421 Q631.29 67.4314 638.703 67.4314 Q642.997 67.4314 647.008 66.3781 Q651.058 65.3249 655.028 63.2184 L655.028 70.267 Q651.018 71.9684 646.805 72.8596 Q642.592 73.7508 638.258 73.7508 Q627.401 73.7508 621.041 67.4314 Q614.722 61.1119 614.722 50.3365 Q614.722 39.1965 620.717 32.6746 Q626.753 26.1121 636.961 26.1121 Q646.116 26.1121 651.423 32.0264 Q656.77 37.9003 656.77 48.0275 M649.317 45.84 Q649.236 39.7232 645.873 36.0774 Q642.552 32.4315 637.042 32.4315 Q630.804 32.4315 627.037 35.9558 Q623.31 39.4801 622.743 45.8805 L649.317 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M713.28 9.62495 Q707.852 18.942 705.219 28.0566 Q702.586 37.1711 702.586 46.5287 Q702.586 55.8863 705.219 65.0818 Q707.893 74.2369 713.28 83.5134 L706.799 83.5134 Q700.723 73.9938 697.684 64.7983 Q694.687 55.6027 694.687 46.5287 Q694.687 37.4952 697.684 28.3401 Q700.682 19.1851 706.799 9.62495 L713.28 9.62495 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M730.213 65.6895 L743.581 65.6895 L743.581 19.5497 L729.038 22.4663 L729.038 15.0127 L743.5 12.096 L751.683 12.096 L751.683 65.6895 L765.051 65.6895 L765.051 72.576 L730.213 72.576 L730.213 65.6895 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M799.078 17.4837 Q792.759 17.4837 789.559 23.7221 Q786.399 29.92 786.399 42.3968 Q786.399 54.833 789.559 61.0714 Q792.759 67.2693 799.078 67.2693 Q805.438 67.2693 808.598 61.0714 Q811.798 54.833 811.798 42.3968 Q811.798 29.92 808.598 23.7221 Q805.438 17.4837 799.078 17.4837 M799.078 11.0023 Q809.246 11.0023 814.593 19.0636 Q819.981 27.0843 819.981 42.3968 Q819.981 57.6687 814.593 65.73 Q809.246 73.7508 799.078 73.7508 Q788.911 73.7508 783.523 65.73 Q778.176 57.6687 778.176 42.3968 Q778.176 27.0843 783.523 19.0636 Q788.911 11.0023 799.078 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M851.862 17.4837 Q845.542 17.4837 842.342 23.7221 Q839.182 29.92 839.182 42.3968 Q839.182 54.833 842.342 61.0714 Q845.542 67.2693 851.862 67.2693 Q858.222 67.2693 861.381 61.0714 Q864.582 54.833 864.582 42.3968 Q864.582 29.92 861.381 23.7221 Q858.222 17.4837 851.862 17.4837 M851.862 11.0023 Q862.03 11.0023 867.377 19.0636 Q872.764 27.0843 872.764 42.3968 Q872.764 57.6687 867.377 65.73 Q862.03 73.7508 851.862 73.7508 Q841.694 73.7508 836.306 65.73 Q830.959 57.6687 830.959 42.3968 Q830.959 27.0843 836.306 19.0636 Q841.694 11.0023 851.862 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M887.996 62.2867 L896.543 62.2867 L896.543 69.2543 L889.9 82.2172 L884.674 82.2172 L887.996 69.2543 L887.996 62.2867 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M931.016 17.4837 Q924.697 17.4837 921.497 23.7221 Q918.337 29.92 918.337 42.3968 Q918.337 54.833 921.497 61.0714 Q924.697 67.2693 931.016 67.2693 Q937.376 67.2693 940.536 61.0714 Q943.736 54.833 943.736 42.3968 Q943.736 29.92 940.536 23.7221 Q937.376 17.4837 931.016 17.4837 M931.016 11.0023 Q941.184 11.0023 946.531 19.0636 Q951.919 27.0843 951.919 42.3968 Q951.919 57.6687 946.531 65.73 Q941.184 73.7508 931.016 73.7508 Q920.849 73.7508 915.461 65.73 Q910.114 57.6687 910.114 42.3968 Q910.114 27.0843 915.461 19.0636 Q920.849 11.0023 931.016 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M983.8 17.4837 Q977.48 17.4837 974.28 23.7221 Q971.12 29.92 971.12 42.3968 Q971.12 54.833 974.28 61.0714 Q977.48 67.2693 983.8 67.2693 Q990.16 67.2693 993.319 61.0714 Q996.52 54.833 996.52 42.3968 Q996.52 29.92 993.319 23.7221 Q990.16 17.4837 983.8 17.4837 M983.8 11.0023 Q993.967 11.0023 999.315 19.0636 Q1004.7 27.0843 1004.7 42.3968 Q1004.7 57.6687 999.315 65.73 Q993.967 73.7508 983.8 73.7508 Q973.632 73.7508 968.244 65.73 Q962.897 57.6687 962.897 42.3968 Q962.897 27.0843 968.244 19.0636 Q973.632 11.0023 983.8 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1036.58 17.4837 Q1030.26 17.4837 1027.06 23.7221 Q1023.9 29.92 1023.9 42.3968 Q1023.9 54.833 1027.06 61.0714 Q1030.26 67.2693 1036.58 67.2693 Q1042.94 67.2693 1046.1 61.0714 Q1049.3 54.833 1049.3 42.3968 Q1049.3 29.92 1046.1 23.7221 Q1042.94 17.4837 1036.58 17.4837 M1036.58 11.0023 Q1046.75 11.0023 1052.1 19.0636 Q1057.49 27.0843 1057.49 42.3968 Q1057.49 57.6687 1052.1 65.73 Q1046.75 73.7508 1036.58 73.7508 Q1026.42 73.7508 1021.03 65.73 Q1015.68 57.6687 1015.68 42.3968 Q1015.68 27.0843 1021.03 19.0636 Q1026.42 11.0023 1036.58 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1120.15 9.54393 L1120.15 15.7418 L1113.02 15.7418 Q1109.01 15.7418 1107.43 17.3622 Q1105.89 18.9825 1105.89 23.1955 L1105.89 27.2059 L1118.17 27.2059 L1118.17 32.9987 L1105.89 32.9987 L1105.89 72.576 L1098.4 72.576 L1098.4 32.9987 L1091.27 32.9987 L1091.27 27.2059 L1098.4 27.2059 L1098.4 24.0462 Q1098.4 16.471 1101.92 13.0277 Q1105.45 9.54393 1113.1 9.54393 L1120.15 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1132.31 14.324 L1132.31 27.2059 L1147.66 27.2059 L1147.66 32.9987 L1132.31 32.9987 L1132.31 57.6282 Q1132.31 63.1779 1133.8 64.7578 Q1135.34 66.3376 1140 66.3376 L1147.66 66.3376 L1147.66 72.576 L1140 72.576 Q1131.37 72.576 1128.09 69.3758 Q1124.81 66.1351 1124.81 57.6282 L1124.81 32.9987 L1119.34 32.9987 L1119.34 27.2059 L1124.81 27.2059 L1124.81 14.324 L1132.31 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1156.29 9.62495 L1162.77 9.62495 Q1168.85 19.1851 1171.84 28.3401 Q1174.88 37.4952 1174.88 46.5287 Q1174.88 55.6027 1171.84 64.7983 Q1168.85 73.9938 1162.77 83.5134 L1156.29 83.5134 Q1161.67 74.2369 1164.31 65.0818 Q1166.98 55.8863 1166.98 46.5287 Q1166.98 37.1711 1164.31 28.0566 Q1161.67 18.942 1156.29 9.62495 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip632)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="193.529,171.734 195.771,177.699 198.013,183.661 200.255,189.62 202.497,195.572 204.739,201.516 206.981,207.449 209.223,213.368 211.464,219.271 213.706,225.155 215.948,231.016 218.19,236.851 220.432,242.656 222.674,248.427 224.916,254.159 227.158,259.849 229.4,265.491 231.642,271.08 233.884,276.61 236.126,282.075 238.368,287.47 240.61,292.787 242.852,298.02 245.094,303.161 247.336,308.203 249.578,313.137 251.82,317.957 254.062,322.653 256.304,327.217 258.546,331.64 260.788,335.913 263.03,340.027 265.271,343.974 267.513,347.745 269.755,351.332 271.997,354.725 274.239,357.918 276.481,360.904 278.723,363.674 280.965,366.224 283.207,368.548 285.449,370.642 287.691,372.502 289.933,374.127 292.175,375.514 294.417,376.663 296.659,377.577 298.901,378.256 301.143,378.704 303.385,378.925 305.627,378.925 307.869,378.71 310.111,378.289 312.353,377.668 314.595,376.858 316.837,375.867 319.078,374.708 321.32,373.39 323.562,371.925 325.804,370.325 328.046,368.602 330.288,366.768 332.53,364.835 334.772,362.816 337.014,360.723 339.256,358.567 341.498,356.361 343.74,354.117 345.982,351.846 348.224,349.558 350.466,347.266 352.708,344.979 354.95,342.707 357.192,340.46 359.434,338.248 361.676,336.078 363.918,333.961 366.16,331.903 368.402,329.913 370.644,327.998 372.885,326.165 375.127,324.42 377.369,322.769 379.611,321.219 381.853,319.774 384.095,318.439 386.337,317.219 388.579,316.119 390.821,315.142 393.063,314.291 395.305,313.57 397.547,312.981 399.789,312.528 402.031,312.212 404.273,312.034 406.515,311.997 408.757,312.101 410.999,312.347 413.241,312.736 415.483,313.267 417.725,313.94 419.967,314.754 422.209,315.708 424.451,316.8 426.692,318.029 428.934,319.392 431.176,320.886 433.418,322.509 435.66,324.256 437.902,326.123 440.144,328.106 442.386,330.2 444.628,332.398 446.87,334.695 449.112,337.084 451.354,339.558 453.596,342.109 455.838,344.729 458.08,347.41 460.322,350.14 462.564,352.912 464.806,355.714 467.048,358.536 469.29,361.366 471.532,364.193 473.774,367.004 476.016,369.788 478.258,372.532 480.5,375.222 482.741,377.848 484.983,380.395 487.225,382.851 489.467,385.206 491.709,387.446 493.951,389.561 496.193,391.541 498.435,393.377 500.677,395.06 502.919,396.582 505.161,397.939 507.403,399.125 509.645,400.137 511.887,400.973 514.129,401.633 516.371,402.119 518.613,402.433 520.855,402.579 523.097,402.562 525.339,402.391 527.581,402.073 529.823,401.617 532.065,401.033 534.307,400.334 536.548,399.53 538.79,398.634 541.032,397.66 543.274,396.621 545.516,395.529 547.758,394.399 550,393.244 552.242,392.078 554.484,390.914 556.726,389.764 558.968,388.641 561.21,387.557 563.452,386.523 565.694,385.55 567.936,384.648 570.178,383.826 572.42,383.094 574.662,382.459 576.904,381.928 579.146,381.508 581.388,381.204 583.63,381.022 585.872,380.964 588.114,381.035 590.355,381.236 592.597,381.569 594.839,382.034 597.081,382.632 599.323,383.36 601.565,384.216 603.807,385.198 606.049,386.302 608.291,387.523 610.533,388.855 612.775,390.292 615.017,391.827 617.259,393.451 619.501,395.156 621.743,396.932 623.985,398.769 626.227,400.655 628.469,402.58 630.711,404.531 632.953,406.497 635.195,408.464 637.437,410.419 639.679,412.351 641.921,414.245 644.162,416.09 646.404,417.873 648.646,419.583 650.888,421.208 653.13,422.739 655.372,424.167 657.614,425.483 659.856,426.681 662.098,427.756 664.34,428.703 666.582,429.52 668.824,430.205 671.066,430.761 673.308,431.189 675.55,431.492 677.792,431.676 680.034,431.747 682.276,431.713 684.518,431.582 686.76,431.365 689.002,431.072 691.244,430.714 693.486,430.304 695.728,429.853 697.969,429.373 700.211,428.878 702.453,428.379 704.695,427.888 706.937,427.418 709.179,426.98 711.421,426.585 713.663,426.242 715.905,425.961 718.147,425.751 720.389,425.619 722.631,425.573 724.873,425.618 727.115,425.759 729.357,425.999 731.599,426.343 733.841,426.79 736.083,427.343 738.325,428 740.567,428.76 742.809,429.62 745.051,430.577 747.293,431.626 749.535,432.762 751.776,433.977 754.018,435.265 756.26,436.617 758.502,438.025 760.744,439.479 762.986,440.968 765.228,442.482 767.47,444.01 769.712,445.542 771.954,447.066 774.196,448.571 776.438,450.046 778.68,451.482 780.922,452.868 783.164,454.195 785.406,455.455 787.648,456.641 789.89,457.747 792.132,458.767 794.374,459.699 796.616,460.539 798.858,461.286 801.1,461.942 803.342,462.508 805.583,462.986 807.825,463.382 810.067,463.7 812.309,463.947 814.551,464.131 816.793,464.259 819.035,464.341 821.277,464.385 823.519,464.403 825.761,464.403 828.003,464.395 830.245,464.391 832.487,464.398 834.729,464.428 836.971,464.489 839.213,464.588 841.455,464.735 843.697,464.935 845.939,465.195 848.181,465.521 850.423,465.916 852.665,466.384 854.907,466.927 857.149,467.547 859.391,468.243 861.632,469.016 863.874,469.863 866.116,470.782 868.358,471.769 870.6,472.819 872.842,473.928 875.084,475.089 877.326,476.295 879.568,477.539 881.81,478.814 884.052,480.11 886.294,481.421 888.536,482.736 890.778,484.048 893.02,485.349 895.262,486.63 897.504,487.884 899.746,489.104 901.988,490.282 904.23,491.415 906.472,492.497 908.714,493.523 910.956,494.492 913.198,495.402 915.439,496.252 917.681,497.041 919.923,497.773 922.165,498.449 924.407,499.072 926.649,499.647 928.891,500.179 931.133,500.673 933.375,501.137 935.617,501.575 937.859,501.997 940.101,502.408 942.343,502.816 944.585,503.229 946.827,503.653 949.069,504.096 951.311,504.563 953.553,505.061 955.795,505.594 958.037,506.169 960.279,506.788 962.521,507.455 964.763,508.173 967.005,508.943 969.246,509.767 971.488,510.643 973.73,511.572 975.972,512.552 978.214,513.581 980.456,514.656 982.698,515.773 984.94,516.928 987.182,518.116 989.424,519.332 991.666,520.57 993.908,521.826 996.15,523.093 998.392,524.365 1000.63,525.636 1002.88,526.902 1005.12,528.156 1007.36,529.393 1009.6,530.61 1011.84,531.802 1014.09,532.966 1016.33,534.099 1018.57,535.2 1020.81,536.267 1023.05,537.3 1025.3,538.299 1027.54,539.265 1029.78,540.2 1032.02,541.107 1034.26,541.988 1036.51,542.848 1038.75,543.689 1040.99,544.516 1043.23,545.335 1045.47,546.149 1047.72,546.963 1049.96,547.783 1052.2,548.614 1054.44,549.459 1056.68,550.323 1058.92,551.21 1061.17,552.125 1063.41,553.068 1065.65,554.045 1067.89,555.056 1070.13,556.103 1072.38,557.188 1074.62,558.31 1076.86,559.469 1079.1,560.666 1081.34,561.897 1083.59,563.162 1085.83,564.459 1088.07,565.784 1090.31,567.135 1092.55,568.508 1094.8,569.9 1097.04,571.308 1099.28,572.727 1101.52,574.154 1103.76,575.585 1106.01,577.016 1108.25,578.445 1110.49,579.869 1112.73,581.284 1114.97,582.688 1117.22,584.08 1119.46,585.459 1121.7,586.823 1123.94,588.172 1126.18,589.506 1128.43,590.827 1130.67,592.134 1132.91,593.43 1135.15,594.718 1137.39,595.998 1139.64,597.274 1141.88,598.55 1144.12,599.827 1146.36,601.111 1148.6,602.403 1150.85,603.708 1153.09,605.029 1155.33,606.369 1157.57,607.731 1159.81,609.119 1162.05,610.535 1164.3,611.981 1166.54,613.459 1168.78,614.971 1171.02,616.518 1173.26,618.1 1175.51,619.719 1177.75,621.373 1179.99,623.062 1182.23,624.785 1184.47,626.541 1186.72,628.329 1188.96,630.145 1191.2,631.987 1193.44,633.854 1195.68,635.741 1197.93,637.646 1200.17,639.567 1202.41,641.499 1204.65,643.44 1206.89,645.387 1209.14,647.337 1211.38,649.287 1213.62,651.236 1215.86,653.181 1218.1,655.121 1220.35,657.054 1222.59,658.981 1224.83,660.9 1227.07,662.814 1229.31,664.721 1231.56,666.626 1233.8,668.528 1236.04,670.432 1238.28,672.34 1240.52,674.256 1242.77,676.185 1245.01,678.131 1247.25,680.1 1249.49,682.096 1251.73,684.126 1253.98,686.194 1256.22,688.308 1258.46,690.472 1260.7,692.693 1262.94,694.976 1265.19,697.326 1267.43,699.748 1269.67,702.247 1271.91,704.825 1274.15,707.487 1276.39,710.234 1278.64,713.068 1280.88,715.989 1283.12,718.997 1285.36,722.089 1287.6,725.262 1289.85,728.511 1292.09,731.831 1294.33,735.212 1296.57,738.645 1298.81,742.119 1301.06,745.62 1303.3,749.133 1305.54,752.642 1307.78,756.127 1310.02,759.568 1312.27,762.946 1314.51,766.238 1316.75,769.422 1318.99,772.479 "/>
+<path clip-path="url(#clip630)" d="M1559.76 790.501 L2752.76 790.501 L2752.76 153.712 L1559.76 153.712  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip633">
+    <rect x="1559" y="153" width="1194" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1593.53,790.501 1593.53,153.712 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1873.77,790.501 1873.77,153.712 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2154.02,790.501 2154.02,153.712 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2434.26,790.501 2434.26,153.712 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2714.51,790.501 2714.51,153.712 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,790.501 2752.76,790.501 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1593.53,790.501 1593.53,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1873.77,790.501 1873.77,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2154.02,790.501 2154.02,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2434.26,790.501 2434.26,771.604 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2714.51,790.501 2714.51,771.604 "/>
+<path clip-path="url(#clip630)" d="M1593.53 824.62 Q1589.92 824.62 1588.09 828.185 Q1586.28 831.726 1586.28 838.856 Q1586.28 845.962 1588.09 849.527 Q1589.92 853.069 1593.53 853.069 Q1597.16 853.069 1598.97 849.527 Q1600.8 845.962 1600.8 838.856 Q1600.8 831.726 1598.97 828.185 Q1597.16 824.62 1593.53 824.62 M1593.53 820.916 Q1599.34 820.916 1602.39 825.523 Q1605.47 830.106 1605.47 838.856 Q1605.47 847.583 1602.39 852.189 Q1599.34 856.773 1593.53 856.773 Q1587.72 856.773 1584.64 852.189 Q1581.58 847.583 1581.58 838.856 Q1581.58 830.106 1584.64 825.523 Q1587.72 820.916 1593.53 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1833.39 821.541 L1851.75 821.541 L1851.75 825.476 L1837.67 825.476 L1837.67 833.949 Q1838.69 833.601 1839.71 833.439 Q1840.73 833.254 1841.75 833.254 Q1847.54 833.254 1850.91 836.425 Q1854.29 839.597 1854.29 845.013 Q1854.29 850.592 1850.82 853.694 Q1847.35 856.773 1841.03 856.773 Q1838.85 856.773 1836.59 856.402 Q1834.34 856.032 1831.93 855.291 L1831.93 850.592 Q1834.02 851.726 1836.24 852.282 Q1838.46 852.837 1840.94 852.837 Q1844.94 852.837 1847.28 850.731 Q1849.62 848.624 1849.62 845.013 Q1849.62 841.402 1847.28 839.296 Q1844.94 837.189 1840.94 837.189 Q1839.06 837.189 1837.19 837.606 Q1835.34 838.023 1833.39 838.902 L1833.39 821.541 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1873.51 824.62 Q1869.9 824.62 1868.07 828.185 Q1866.26 831.726 1866.26 838.856 Q1866.26 845.962 1868.07 849.527 Q1869.9 853.069 1873.51 853.069 Q1877.14 853.069 1878.95 849.527 Q1880.78 845.962 1880.78 838.856 Q1880.78 831.726 1878.95 828.185 Q1877.14 824.62 1873.51 824.62 M1873.51 820.916 Q1879.32 820.916 1882.37 825.523 Q1885.45 830.106 1885.45 838.856 Q1885.45 847.583 1882.37 852.189 Q1879.32 856.773 1873.51 856.773 Q1867.7 856.773 1864.62 852.189 Q1861.56 847.583 1861.56 838.856 Q1861.56 830.106 1864.62 825.523 Q1867.7 820.916 1873.51 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1903.67 824.62 Q1900.06 824.62 1898.23 828.185 Q1896.42 831.726 1896.42 838.856 Q1896.42 845.962 1898.23 849.527 Q1900.06 853.069 1903.67 853.069 Q1907.3 853.069 1909.11 849.527 Q1910.94 845.962 1910.94 838.856 Q1910.94 831.726 1909.11 828.185 Q1907.3 824.62 1903.67 824.62 M1903.67 820.916 Q1909.48 820.916 1912.53 825.523 Q1915.61 830.106 1915.61 838.856 Q1915.61 847.583 1912.53 852.189 Q1909.48 856.773 1903.67 856.773 Q1897.86 856.773 1894.78 852.189 Q1891.72 847.583 1891.72 838.856 Q1891.72 830.106 1894.78 825.523 Q1897.86 820.916 1903.67 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2098.54 852.166 L2106.18 852.166 L2106.18 825.8 L2097.87 827.467 L2097.87 823.208 L2106.14 821.541 L2110.81 821.541 L2110.81 852.166 L2118.45 852.166 L2118.45 856.101 L2098.54 856.101 L2098.54 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2137.9 824.62 Q2134.28 824.62 2132.46 828.185 Q2130.65 831.726 2130.65 838.856 Q2130.65 845.962 2132.46 849.527 Q2134.28 853.069 2137.9 853.069 Q2141.53 853.069 2143.34 849.527 Q2145.16 845.962 2145.16 838.856 Q2145.16 831.726 2143.34 828.185 Q2141.53 824.62 2137.9 824.62 M2137.9 820.916 Q2143.71 820.916 2146.76 825.523 Q2149.84 830.106 2149.84 838.856 Q2149.84 847.583 2146.76 852.189 Q2143.71 856.773 2137.9 856.773 Q2132.09 856.773 2129.01 852.189 Q2125.95 847.583 2125.95 838.856 Q2125.95 830.106 2129.01 825.523 Q2132.09 820.916 2137.9 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2168.06 824.62 Q2164.45 824.62 2162.62 828.185 Q2160.81 831.726 2160.81 838.856 Q2160.81 845.962 2162.62 849.527 Q2164.45 853.069 2168.06 853.069 Q2171.69 853.069 2173.5 849.527 Q2175.33 845.962 2175.33 838.856 Q2175.33 831.726 2173.5 828.185 Q2171.69 824.62 2168.06 824.62 M2168.06 820.916 Q2173.87 820.916 2176.92 825.523 Q2180 830.106 2180 838.856 Q2180 847.583 2176.92 852.189 Q2173.87 856.773 2168.06 856.773 Q2162.25 856.773 2159.17 852.189 Q2156.11 847.583 2156.11 838.856 Q2156.11 830.106 2159.17 825.523 Q2162.25 820.916 2168.06 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2198.22 824.62 Q2194.61 824.62 2192.78 828.185 Q2190.97 831.726 2190.97 838.856 Q2190.97 845.962 2192.78 849.527 Q2194.61 853.069 2198.22 853.069 Q2201.85 853.069 2203.66 849.527 Q2205.49 845.962 2205.49 838.856 Q2205.49 831.726 2203.66 828.185 Q2201.85 824.62 2198.22 824.62 M2198.22 820.916 Q2204.03 820.916 2207.09 825.523 Q2210.16 830.106 2210.16 838.856 Q2210.16 847.583 2207.09 852.189 Q2204.03 856.773 2198.22 856.773 Q2192.41 856.773 2189.33 852.189 Q2186.28 847.583 2186.28 838.856 Q2186.28 830.106 2189.33 825.523 Q2192.41 820.916 2198.22 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2378.79 852.166 L2386.43 852.166 L2386.43 825.8 L2378.12 827.467 L2378.12 823.208 L2386.38 821.541 L2391.06 821.541 L2391.06 852.166 L2398.7 852.166 L2398.7 856.101 L2378.79 856.101 L2378.79 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2408.19 821.541 L2426.54 821.541 L2426.54 825.476 L2412.47 825.476 L2412.47 833.949 Q2413.49 833.601 2414.51 833.439 Q2415.52 833.254 2416.54 833.254 Q2422.33 833.254 2425.71 836.425 Q2429.09 839.597 2429.09 845.013 Q2429.09 850.592 2425.62 853.694 Q2422.15 856.773 2415.83 856.773 Q2413.65 856.773 2411.38 856.402 Q2409.14 856.032 2406.73 855.291 L2406.73 850.592 Q2408.81 851.726 2411.03 852.282 Q2413.26 852.837 2415.73 852.837 Q2419.74 852.837 2422.08 850.731 Q2424.41 848.624 2424.41 845.013 Q2424.41 841.402 2422.08 839.296 Q2419.74 837.189 2415.73 837.189 Q2413.86 837.189 2411.98 837.606 Q2410.13 838.023 2408.19 838.902 L2408.19 821.541 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2448.3 824.62 Q2444.69 824.62 2442.86 828.185 Q2441.06 831.726 2441.06 838.856 Q2441.06 845.962 2442.86 849.527 Q2444.69 853.069 2448.3 853.069 Q2451.94 853.069 2453.74 849.527 Q2455.57 845.962 2455.57 838.856 Q2455.57 831.726 2453.74 828.185 Q2451.94 824.62 2448.3 824.62 M2448.3 820.916 Q2454.11 820.916 2457.17 825.523 Q2460.25 830.106 2460.25 838.856 Q2460.25 847.583 2457.17 852.189 Q2454.11 856.773 2448.3 856.773 Q2442.49 856.773 2439.41 852.189 Q2436.36 847.583 2436.36 838.856 Q2436.36 830.106 2439.41 825.523 Q2442.49 820.916 2448.3 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2478.46 824.62 Q2474.85 824.62 2473.02 828.185 Q2471.22 831.726 2471.22 838.856 Q2471.22 845.962 2473.02 849.527 Q2474.85 853.069 2478.46 853.069 Q2482.1 853.069 2483.9 849.527 Q2485.73 845.962 2485.73 838.856 Q2485.73 831.726 2483.9 828.185 Q2482.1 824.62 2478.46 824.62 M2478.46 820.916 Q2484.27 820.916 2487.33 825.523 Q2490.41 830.106 2490.41 838.856 Q2490.41 847.583 2487.33 852.189 Q2484.27 856.773 2478.46 856.773 Q2472.65 856.773 2469.58 852.189 Q2466.52 847.583 2466.52 838.856 Q2466.52 830.106 2469.58 825.523 Q2472.65 820.916 2478.46 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2663.12 852.166 L2679.44 852.166 L2679.44 856.101 L2657.49 856.101 L2657.49 852.166 Q2660.16 849.411 2664.74 844.782 Q2669.35 840.129 2670.53 838.786 Q2672.77 836.263 2673.65 834.527 Q2674.55 832.768 2674.55 831.078 Q2674.55 828.324 2672.61 826.587 Q2670.69 824.851 2667.59 824.851 Q2665.39 824.851 2662.93 825.615 Q2660.5 826.379 2657.73 827.93 L2657.73 823.208 Q2660.55 822.074 2663 821.495 Q2665.46 820.916 2667.49 820.916 Q2672.86 820.916 2676.06 823.601 Q2679.25 826.287 2679.25 830.777 Q2679.25 832.907 2678.44 834.828 Q2677.66 836.726 2675.55 839.319 Q2674.97 839.99 2671.87 843.208 Q2668.77 846.402 2663.12 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2699.25 824.62 Q2695.64 824.62 2693.81 828.185 Q2692.01 831.726 2692.01 838.856 Q2692.01 845.962 2693.81 849.527 Q2695.64 853.069 2699.25 853.069 Q2702.89 853.069 2704.69 849.527 Q2706.52 845.962 2706.52 838.856 Q2706.52 831.726 2704.69 828.185 Q2702.89 824.62 2699.25 824.62 M2699.25 820.916 Q2705.06 820.916 2708.12 825.523 Q2711.2 830.106 2711.2 838.856 Q2711.2 847.583 2708.12 852.189 Q2705.06 856.773 2699.25 856.773 Q2693.44 856.773 2690.36 852.189 Q2687.31 847.583 2687.31 838.856 Q2687.31 830.106 2690.36 825.523 Q2693.44 820.916 2699.25 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2729.42 824.62 Q2725.8 824.62 2723.98 828.185 Q2722.17 831.726 2722.17 838.856 Q2722.17 845.962 2723.98 849.527 Q2725.8 853.069 2729.42 853.069 Q2733.05 853.069 2734.86 849.527 Q2736.68 845.962 2736.68 838.856 Q2736.68 831.726 2734.86 828.185 Q2733.05 824.62 2729.42 824.62 M2729.42 820.916 Q2735.23 820.916 2738.28 825.523 Q2741.36 830.106 2741.36 838.856 Q2741.36 847.583 2738.28 852.189 Q2735.23 856.773 2729.42 856.773 Q2723.61 856.773 2720.53 852.189 Q2717.47 847.583 2717.47 838.856 Q2717.47 830.106 2720.53 825.523 Q2723.61 820.916 2729.42 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2759.58 824.62 Q2755.97 824.62 2754.14 828.185 Q2752.33 831.726 2752.33 838.856 Q2752.33 845.962 2754.14 849.527 Q2755.97 853.069 2759.58 853.069 Q2763.21 853.069 2765.02 849.527 Q2766.85 845.962 2766.85 838.856 Q2766.85 831.726 2765.02 828.185 Q2763.21 824.62 2759.58 824.62 M2759.58 820.916 Q2765.39 820.916 2768.44 825.523 Q2771.52 830.106 2771.52 838.856 Q2771.52 847.583 2768.44 852.189 Q2765.39 856.773 2759.58 856.773 Q2753.77 856.773 2750.69 852.189 Q2747.63 847.583 2747.63 838.856 Q2747.63 830.106 2750.69 825.523 Q2753.77 820.916 2759.58 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,707.463 2752.76,707.463 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,577.432 2752.76,577.432 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,447.401 2752.76,447.401 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,317.369 2752.76,317.369 "/>
+<polyline clip-path="url(#clip633)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,187.338 2752.76,187.338 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,790.501 1559.76,153.712 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,707.463 1578.66,707.463 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,577.432 1578.66,577.432 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,447.401 1578.66,447.401 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,317.369 1578.66,317.369 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,187.338 1578.66,187.338 "/>
+<path clip-path="url(#clip630)" d="M1461.58 693.262 Q1457.97 693.262 1456.14 696.827 Q1454.34 700.368 1454.34 707.498 Q1454.34 714.604 1456.14 718.169 Q1457.97 721.711 1461.58 721.711 Q1465.22 721.711 1467.02 718.169 Q1468.85 714.604 1468.85 707.498 Q1468.85 700.368 1467.02 696.827 Q1465.22 693.262 1461.58 693.262 M1461.58 689.558 Q1467.39 689.558 1470.45 694.165 Q1473.53 698.748 1473.53 707.498 Q1473.53 716.225 1470.45 720.831 Q1467.39 725.415 1461.58 725.415 Q1455.77 725.415 1452.7 720.831 Q1449.64 716.225 1449.64 707.498 Q1449.64 698.748 1452.7 694.165 Q1455.77 689.558 1461.58 689.558 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1481.75 718.864 L1486.63 718.864 L1486.63 724.743 L1481.75 724.743 L1481.75 718.864 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1496.86 690.183 L1515.22 690.183 L1515.22 694.118 L1501.14 694.118 L1501.14 702.591 Q1502.16 702.243 1503.18 702.081 Q1504.2 701.896 1505.22 701.896 Q1511.01 701.896 1514.39 705.067 Q1517.76 708.239 1517.76 713.655 Q1517.76 719.234 1514.29 722.336 Q1510.82 725.415 1504.5 725.415 Q1502.33 725.415 1500.06 725.044 Q1497.81 724.674 1495.4 723.933 L1495.4 719.234 Q1497.49 720.368 1499.71 720.924 Q1501.93 721.479 1504.41 721.479 Q1508.41 721.479 1510.75 719.373 Q1513.09 717.266 1513.09 713.655 Q1513.09 710.044 1510.75 707.938 Q1508.41 705.831 1504.41 705.831 Q1502.53 705.831 1500.66 706.248 Q1498.81 706.665 1496.86 707.544 L1496.86 690.183 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1451.4 590.777 L1459.04 590.777 L1459.04 564.411 L1450.73 566.078 L1450.73 561.819 L1458.99 560.152 L1463.67 560.152 L1463.67 590.777 L1471.31 590.777 L1471.31 594.712 L1451.4 594.712 L1451.4 590.777 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1480.75 588.832 L1485.64 588.832 L1485.64 594.712 L1480.75 594.712 L1480.75 588.832 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1505.82 563.231 Q1502.21 563.231 1500.38 566.795 Q1498.58 570.337 1498.58 577.467 Q1498.58 584.573 1500.38 588.138 Q1502.21 591.68 1505.82 591.68 Q1509.45 591.68 1511.26 588.138 Q1513.09 584.573 1513.09 577.467 Q1513.09 570.337 1511.26 566.795 Q1509.45 563.231 1505.82 563.231 M1505.82 559.527 Q1511.63 559.527 1514.69 564.133 Q1517.76 568.717 1517.76 577.467 Q1517.76 586.194 1514.69 590.8 Q1511.63 595.383 1505.82 595.383 Q1500.01 595.383 1496.93 590.8 Q1493.88 586.194 1493.88 577.467 Q1493.88 568.717 1496.93 564.133 Q1500.01 559.527 1505.82 559.527 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1452.39 460.746 L1460.03 460.746 L1460.03 434.38 L1451.72 436.047 L1451.72 431.787 L1459.99 430.121 L1464.66 430.121 L1464.66 460.746 L1472.3 460.746 L1472.3 464.681 L1452.39 464.681 L1452.39 460.746 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1481.75 458.801 L1486.63 458.801 L1486.63 464.681 L1481.75 464.681 L1481.75 458.801 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1496.86 430.121 L1515.22 430.121 L1515.22 434.056 L1501.14 434.056 L1501.14 442.528 Q1502.16 442.181 1503.18 442.019 Q1504.2 441.834 1505.22 441.834 Q1511.01 441.834 1514.39 445.005 Q1517.76 448.176 1517.76 453.593 Q1517.76 459.171 1514.29 462.273 Q1510.82 465.352 1504.5 465.352 Q1502.33 465.352 1500.06 464.982 Q1497.81 464.611 1495.4 463.871 L1495.4 459.171 Q1497.49 460.306 1499.71 460.861 Q1501.93 461.417 1504.41 461.417 Q1508.41 461.417 1510.75 459.31 Q1513.09 457.204 1513.09 453.593 Q1513.09 449.982 1510.75 447.875 Q1508.41 445.769 1504.41 445.769 Q1502.53 445.769 1500.66 446.185 Q1498.81 446.602 1496.86 447.482 L1496.86 430.121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1454.62 330.714 L1470.94 330.714 L1470.94 334.649 L1448.99 334.649 L1448.99 330.714 Q1451.65 327.96 1456.24 323.33 Q1460.84 318.677 1462.02 317.335 Q1464.27 314.812 1465.15 313.076 Q1466.05 311.316 1466.05 309.626 Q1466.05 306.872 1464.11 305.136 Q1462.19 303.4 1459.08 303.4 Q1456.89 303.4 1454.43 304.164 Q1452 304.927 1449.22 306.478 L1449.22 301.756 Q1452.05 300.622 1454.5 300.043 Q1456.96 299.464 1458.99 299.464 Q1464.36 299.464 1467.56 302.15 Q1470.75 304.835 1470.75 309.326 Q1470.75 311.455 1469.94 313.376 Q1469.15 315.275 1467.05 317.867 Q1466.47 318.538 1463.37 321.756 Q1460.27 324.95 1454.62 330.714 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1480.75 328.77 L1485.64 328.77 L1485.64 334.649 L1480.75 334.649 L1480.75 328.77 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1505.82 303.168 Q1502.21 303.168 1500.38 306.733 Q1498.58 310.275 1498.58 317.404 Q1498.58 324.511 1500.38 328.075 Q1502.21 331.617 1505.82 331.617 Q1509.45 331.617 1511.26 328.075 Q1513.09 324.511 1513.09 317.404 Q1513.09 310.275 1511.26 306.733 Q1509.45 303.168 1505.82 303.168 M1505.82 299.464 Q1511.63 299.464 1514.69 304.071 Q1517.76 308.654 1517.76 317.404 Q1517.76 326.131 1514.69 330.737 Q1511.63 335.321 1505.82 335.321 Q1500.01 335.321 1496.93 330.737 Q1493.88 326.131 1493.88 317.404 Q1493.88 308.654 1496.93 304.071 Q1500.01 299.464 1505.82 299.464 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1455.61 200.683 L1471.93 200.683 L1471.93 204.618 L1449.99 204.618 L1449.99 200.683 Q1452.65 197.928 1457.23 193.299 Q1461.84 188.646 1463.02 187.303 Q1465.27 184.78 1466.14 183.044 Q1467.05 181.285 1467.05 179.595 Q1467.05 176.841 1465.1 175.104 Q1463.18 173.368 1460.08 173.368 Q1457.88 173.368 1455.43 174.132 Q1453 174.896 1450.22 176.447 L1450.22 171.725 Q1453.04 170.591 1455.5 170.012 Q1457.95 169.433 1459.99 169.433 Q1465.36 169.433 1468.55 172.118 Q1471.75 174.804 1471.75 179.294 Q1471.75 181.424 1470.94 183.345 Q1470.15 185.243 1468.04 187.836 Q1467.46 188.507 1464.36 191.725 Q1461.26 194.919 1455.61 200.683 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1481.75 198.739 L1486.63 198.739 L1486.63 204.618 L1481.75 204.618 L1481.75 198.739 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1496.86 170.058 L1515.22 170.058 L1515.22 173.993 L1501.14 173.993 L1501.14 182.466 Q1502.16 182.118 1503.18 181.956 Q1504.2 181.771 1505.22 181.771 Q1511.01 181.771 1514.39 184.942 Q1517.76 188.114 1517.76 193.53 Q1517.76 199.109 1514.29 202.211 Q1510.82 205.29 1504.5 205.29 Q1502.33 205.29 1500.06 204.919 Q1497.81 204.549 1495.4 203.808 L1495.4 199.109 Q1497.49 200.243 1499.71 200.799 Q1501.93 201.354 1504.41 201.354 Q1508.41 201.354 1510.75 199.248 Q1513.09 197.141 1513.09 193.53 Q1513.09 189.919 1510.75 187.813 Q1508.41 185.706 1504.41 185.706 Q1502.53 185.706 1500.66 186.123 Q1498.81 186.54 1496.86 187.419 L1496.86 170.058 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1742.16 72.576 L1719.07 12.096 L1727.61 12.096 L1746.77 63.0159 L1765.98 12.096 L1774.48 12.096 L1751.43 72.576 L1742.16 72.576 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1815.36 48.0275 L1815.36 51.6733 L1781.09 51.6733 Q1781.57 59.3701 1785.7 63.421 Q1789.88 67.4314 1797.29 67.4314 Q1801.58 67.4314 1805.59 66.3781 Q1809.64 65.3249 1813.61 63.2184 L1813.61 70.267 Q1809.6 71.9684 1805.39 72.8596 Q1801.18 73.7508 1796.84 73.7508 Q1785.99 73.7508 1779.63 67.4314 Q1773.31 61.1119 1773.31 50.3365 Q1773.31 39.1965 1779.3 32.6746 Q1785.34 26.1121 1795.55 26.1121 Q1804.7 26.1121 1810.01 32.0264 Q1815.36 37.9003 1815.36 48.0275 M1807.9 45.84 Q1807.82 39.7232 1804.46 36.0774 Q1801.14 32.4315 1795.63 32.4315 Q1789.39 32.4315 1785.62 35.9558 Q1781.9 39.4801 1781.33 45.8805 L1807.9 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1827.59 9.54393 L1835.04 9.54393 L1835.04 72.576 L1827.59 72.576 L1827.59 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1868.22 32.4315 Q1862.23 32.4315 1858.74 37.1306 Q1855.26 41.7891 1855.26 49.9314 Q1855.26 58.0738 1858.7 62.7728 Q1862.19 67.4314 1868.22 67.4314 Q1874.18 67.4314 1877.66 62.7323 Q1881.14 58.0333 1881.14 49.9314 Q1881.14 41.8701 1877.66 37.1711 Q1874.18 32.4315 1868.22 32.4315 M1868.22 26.1121 Q1877.94 26.1121 1883.49 32.4315 Q1889.04 38.7509 1889.04 49.9314 Q1889.04 61.0714 1883.49 67.4314 Q1877.94 73.7508 1868.22 73.7508 Q1858.46 73.7508 1852.91 67.4314 Q1847.4 61.0714 1847.4 49.9314 Q1847.4 38.7509 1852.91 32.4315 Q1858.46 26.1121 1868.22 26.1121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1934.05 28.9478 L1934.05 35.9153 Q1930.89 34.1734 1927.69 33.3227 Q1924.53 32.4315 1921.29 32.4315 Q1914.04 32.4315 1910.03 37.0496 Q1906.02 41.6271 1906.02 49.9314 Q1906.02 58.2358 1910.03 62.8538 Q1914.04 67.4314 1921.29 67.4314 Q1924.53 67.4314 1927.69 66.5807 Q1930.89 65.6895 1934.05 63.9476 L1934.05 70.8341 Q1930.93 72.2924 1927.57 73.0216 Q1924.24 73.7508 1920.48 73.7508 Q1910.23 73.7508 1904.19 67.3098 Q1898.16 60.8689 1898.16 49.9314 Q1898.16 38.832 1904.23 32.472 Q1910.35 26.1121 1920.96 26.1121 Q1924.41 26.1121 1927.69 26.8413 Q1930.97 27.5299 1934.05 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1947.01 27.2059 L1954.46 27.2059 L1954.46 72.576 L1947.01 72.576 L1947.01 27.2059 M1947.01 9.54393 L1954.46 9.54393 L1954.46 18.9825 L1947.01 18.9825 L1947.01 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1977.43 14.324 L1977.43 27.2059 L1992.79 27.2059 L1992.79 32.9987 L1977.43 32.9987 L1977.43 57.6282 Q1977.43 63.1779 1978.93 64.7578 Q1980.47 66.3376 1985.13 66.3376 L1992.79 66.3376 L1992.79 72.576 L1985.13 72.576 Q1976.5 72.576 1973.22 69.3758 Q1969.94 66.1351 1969.94 57.6282 L1969.94 32.9987 L1964.47 32.9987 L1964.47 27.2059 L1969.94 27.2059 L1969.94 14.324 L1977.43 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2021.47 76.7889 Q2018.31 84.8907 2015.31 87.3618 Q2012.31 89.8329 2007.29 89.8329 L2001.33 89.8329 L2001.33 83.5945 L2005.71 83.5945 Q2008.79 83.5945 2010.49 82.1361 Q2012.19 80.6778 2014.26 75.2496 L2015.59 71.8468 L1997.24 27.2059 L2005.14 27.2059 L2019.32 62.6918 L2033.5 27.2059 L2041.4 27.2059 L2021.47 76.7889 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2095.96 9.62495 Q2090.53 18.942 2087.9 28.0566 Q2085.27 37.1711 2085.27 46.5287 Q2085.27 55.8863 2087.9 65.0818 Q2090.57 74.2369 2095.96 83.5134 L2089.48 83.5134 Q2083.4 73.9938 2080.37 64.7983 Q2077.37 55.6027 2077.37 46.5287 Q2077.37 37.4952 2080.37 28.3401 Q2083.36 19.1851 2089.48 9.62495 L2095.96 9.62495 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2112.9 65.6895 L2126.26 65.6895 L2126.26 19.5497 L2111.72 22.4663 L2111.72 15.0127 L2126.18 12.096 L2134.37 12.096 L2134.37 65.6895 L2147.73 65.6895 L2147.73 72.576 L2112.9 72.576 L2112.9 65.6895 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2181.76 17.4837 Q2175.44 17.4837 2172.24 23.7221 Q2169.08 29.92 2169.08 42.3968 Q2169.08 54.833 2172.24 61.0714 Q2175.44 67.2693 2181.76 67.2693 Q2188.12 67.2693 2191.28 61.0714 Q2194.48 54.833 2194.48 42.3968 Q2194.48 29.92 2191.28 23.7221 Q2188.12 17.4837 2181.76 17.4837 M2181.76 11.0023 Q2191.93 11.0023 2197.28 19.0636 Q2202.66 27.0843 2202.66 42.3968 Q2202.66 57.6687 2197.28 65.73 Q2191.93 73.7508 2181.76 73.7508 Q2171.59 73.7508 2166.21 65.73 Q2160.86 57.6687 2160.86 42.3968 Q2160.86 27.0843 2166.21 19.0636 Q2171.59 11.0023 2181.76 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2234.54 17.4837 Q2228.22 17.4837 2225.02 23.7221 Q2221.86 29.92 2221.86 42.3968 Q2221.86 54.833 2225.02 61.0714 Q2228.22 67.2693 2234.54 67.2693 Q2240.9 67.2693 2244.06 61.0714 Q2247.26 54.833 2247.26 42.3968 Q2247.26 29.92 2244.06 23.7221 Q2240.9 17.4837 2234.54 17.4837 M2234.54 11.0023 Q2244.71 11.0023 2250.06 19.0636 Q2255.45 27.0843 2255.45 42.3968 Q2255.45 57.6687 2250.06 65.73 Q2244.71 73.7508 2234.54 73.7508 Q2224.38 73.7508 2218.99 65.73 Q2213.64 57.6687 2213.64 42.3968 Q2213.64 27.0843 2218.99 19.0636 Q2224.38 11.0023 2234.54 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2287.33 17.4837 Q2281.01 17.4837 2277.81 23.7221 Q2274.65 29.92 2274.65 42.3968 Q2274.65 54.833 2277.81 61.0714 Q2281.01 67.2693 2287.33 67.2693 Q2293.69 67.2693 2296.85 61.0714 Q2300.05 54.833 2300.05 42.3968 Q2300.05 29.92 2296.85 23.7221 Q2293.69 17.4837 2287.33 17.4837 M2287.33 11.0023 Q2297.5 11.0023 2302.84 19.0636 Q2308.23 27.0843 2308.23 42.3968 Q2308.23 57.6687 2302.84 65.73 Q2297.5 73.7508 2287.33 73.7508 Q2277.16 73.7508 2271.77 65.73 Q2266.42 57.6687 2266.42 42.3968 Q2266.42 27.0843 2271.77 19.0636 Q2277.16 11.0023 2287.33 11.0023 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2370.9 9.54393 L2370.9 15.7418 L2363.77 15.7418 Q2359.76 15.7418 2358.18 17.3622 Q2356.64 18.9825 2356.64 23.1955 L2356.64 27.2059 L2368.91 27.2059 L2368.91 32.9987 L2356.64 32.9987 L2356.64 72.576 L2349.14 72.576 L2349.14 32.9987 L2342.01 32.9987 L2342.01 27.2059 L2349.14 27.2059 L2349.14 24.0462 Q2349.14 16.471 2352.67 13.0277 Q2356.19 9.54393 2363.85 9.54393 L2370.9 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2383.05 14.324 L2383.05 27.2059 L2398.4 27.2059 L2398.4 32.9987 L2383.05 32.9987 L2383.05 57.6282 Q2383.05 63.1779 2384.55 64.7578 Q2386.09 66.3376 2390.75 66.3376 L2398.4 66.3376 L2398.4 72.576 L2390.75 72.576 Q2382.12 72.576 2378.84 69.3758 Q2375.56 66.1351 2375.56 57.6282 L2375.56 32.9987 L2370.09 32.9987 L2370.09 27.2059 L2375.56 27.2059 L2375.56 14.324 L2383.05 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2421.45 12.096 L2428.34 12.096 L2407.27 80.2727 L2400.39 80.2727 L2421.45 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2465.08 28.5427 L2465.08 35.5912 Q2461.92 33.9709 2458.52 33.1607 Q2455.12 32.3505 2451.47 32.3505 Q2445.92 32.3505 2443.13 34.0519 Q2440.37 35.7533 2440.37 39.156 Q2440.37 41.7486 2442.36 43.2475 Q2444.34 44.7058 2450.34 46.0426 L2452.89 46.6097 Q2460.83 48.3111 2464.15 51.4303 Q2467.51 54.509 2467.51 60.0587 Q2467.51 66.3781 2462.49 70.0644 Q2457.51 73.7508 2448.76 73.7508 Q2445.11 73.7508 2441.14 73.0216 Q2437.21 72.3329 2432.84 70.9151 L2432.84 63.2184 Q2436.97 65.3654 2440.98 66.4591 Q2444.99 67.5124 2448.92 67.5124 Q2454.18 67.5124 2457.02 65.73 Q2459.86 63.9071 2459.86 60.6258 Q2459.86 57.5877 2457.79 55.9673 Q2455.76 54.3469 2448.84 52.8481 L2446.24 52.2405 Q2439.32 50.7821 2436.24 47.7845 Q2433.16 44.7463 2433.16 39.4801 Q2433.16 33.0797 2437.7 29.5959 Q2442.23 26.1121 2450.58 26.1121 Q2454.71 26.1121 2458.36 26.7198 Q2462 27.3274 2465.08 28.5427 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2518.19 48.0275 L2518.19 51.6733 L2483.92 51.6733 Q2484.4 59.3701 2488.54 63.421 Q2492.71 67.4314 2500.12 67.4314 Q2504.42 67.4314 2508.43 66.3781 Q2512.48 65.3249 2516.45 63.2184 L2516.45 70.267 Q2512.44 71.9684 2508.22 72.8596 Q2504.01 73.7508 2499.68 73.7508 Q2488.82 73.7508 2482.46 67.4314 Q2476.14 61.1119 2476.14 50.3365 Q2476.14 39.1965 2482.14 32.6746 Q2488.17 26.1121 2498.38 26.1121 Q2507.53 26.1121 2512.84 32.0264 Q2518.19 37.9003 2518.19 48.0275 M2510.73 45.84 Q2510.65 39.7232 2507.29 36.0774 Q2503.97 32.4315 2498.46 32.4315 Q2492.22 32.4315 2488.45 35.9558 Q2484.73 39.4801 2484.16 45.8805 L2510.73 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2563.07 28.9478 L2563.07 35.9153 Q2559.91 34.1734 2556.71 33.3227 Q2553.55 32.4315 2550.31 32.4315 Q2543.06 32.4315 2539.05 37.0496 Q2535.04 41.6271 2535.04 49.9314 Q2535.04 58.2358 2539.05 62.8538 Q2543.06 67.4314 2550.31 67.4314 Q2553.55 67.4314 2556.71 66.5807 Q2559.91 65.6895 2563.07 63.9476 L2563.07 70.8341 Q2559.95 72.2924 2556.59 73.0216 Q2553.27 73.7508 2549.5 73.7508 Q2539.25 73.7508 2533.22 67.3098 Q2527.18 60.8689 2527.18 49.9314 Q2527.18 38.832 2533.26 32.472 Q2539.37 26.1121 2549.99 26.1121 Q2553.43 26.1121 2556.71 26.8413 Q2559.99 27.5299 2563.07 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2574.86 9.62495 L2581.34 9.62495 Q2587.42 19.1851 2590.42 28.3401 Q2593.45 37.4952 2593.45 46.5287 Q2593.45 55.6027 2590.42 64.7983 Q2587.42 73.9938 2581.34 83.5134 L2574.86 83.5134 Q2580.25 74.2369 2582.88 65.0818 Q2585.55 55.8863 2585.55 46.5287 Q2585.55 37.1711 2582.88 28.0566 Q2580.25 18.942 2574.86 9.62495 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip633)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="1593.53,171.734 1595.77,171.777 1598.01,171.828 1600.25,171.887 1602.5,171.956 1604.74,172.034 1606.98,172.123 1609.22,172.223 1611.46,172.335 1613.71,172.461 1615.95,172.601 1618.19,172.756 1620.43,172.927 1622.67,173.116 1624.92,173.324 1627.16,173.551 1629.4,173.8 1631.64,174.072 1633.88,174.367 1636.13,174.688 1638.37,175.035 1640.61,175.411 1642.85,175.817 1645.09,176.254 1647.34,176.724 1649.58,177.228 1651.82,177.768 1654.06,178.344 1656.3,178.958 1658.55,179.61 1660.79,180.303 1663.03,181.035 1665.27,181.808 1667.51,182.621 1669.76,183.475 1672,184.369 1674.24,185.302 1676.48,186.273 1678.72,187.281 1680.97,188.323 1683.21,189.399 1685.45,190.504 1687.69,191.637 1689.93,192.795 1692.17,193.974 1694.42,195.171 1696.66,196.383 1698.9,197.605 1701.14,198.835 1703.38,200.068 1705.63,201.301 1707.87,202.531 1710.11,203.754 1712.35,204.967 1714.59,206.167 1716.84,207.352 1719.08,208.519 1721.32,209.666 1723.56,210.791 1725.8,211.893 1728.05,212.97 1730.29,214.022 1732.53,215.048 1734.77,216.047 1737.01,217.018 1739.26,217.963 1741.5,218.88 1743.74,219.771 1745.98,220.635 1748.22,221.474 1750.47,222.287 1752.71,223.076 1754.95,223.841 1757.19,224.583 1759.43,225.304 1761.68,226.003 1763.92,226.682 1766.16,227.342 1768.4,227.983 1770.64,228.608 1772.89,229.216 1775.13,229.809 1777.37,230.387 1779.61,230.953 1781.85,231.505 1784.1,232.046 1786.34,232.577 1788.58,233.098 1790.82,233.61 1793.06,234.113 1795.31,234.61 1797.55,235.1 1799.79,235.585 1802.03,236.065 1804.27,236.542 1806.51,237.015 1808.76,237.487 1811,237.957 1813.24,238.426 1815.48,238.896 1817.72,239.368 1819.97,239.841 1822.21,240.318 1824.45,240.798 1826.69,241.283 1828.93,241.775 1831.18,242.273 1833.42,242.779 1835.66,243.294 1837.9,243.819 1840.14,244.355 1842.39,244.903 1844.63,245.465 1846.87,246.042 1849.11,246.634 1851.35,247.243 1853.6,247.87 1855.84,248.517 1858.08,249.185 1860.32,249.874 1862.56,250.587 1864.81,251.324 1867.05,252.086 1869.29,252.875 1871.53,253.692 1873.77,254.536 1876.02,255.41 1878.26,256.314 1880.5,257.247 1882.74,258.211 1884.98,259.205 1887.23,260.229 1889.47,261.283 1891.71,262.365 1893.95,263.476 1896.19,264.612 1898.44,265.774 1900.68,266.96 1902.92,268.166 1905.16,269.391 1907.4,270.633 1909.64,271.888 1911.89,273.154 1914.13,274.429 1916.37,275.709 1918.61,276.991 1920.85,278.273 1923.1,279.552 1925.34,280.826 1927.58,282.091 1929.82,283.346 1932.06,284.589 1934.31,285.817 1936.55,287.03 1938.79,288.226 1941.03,289.403 1943.27,290.562 1945.52,291.701 1947.76,292.821 1950,293.92 1952.24,295 1954.48,296.06 1956.73,297.1 1958.97,298.122 1961.21,299.126 1963.45,300.113 1965.69,301.083 1967.94,302.038 1970.18,302.978 1972.42,303.905 1974.66,304.82 1976.9,305.724 1979.15,306.619 1981.39,307.505 1983.63,308.385 1985.87,309.259 1988.11,310.128 1990.36,310.995 1992.6,311.861 1994.84,312.726 1997.08,313.594 1999.32,314.464 2001.57,315.339 2003.81,316.219 2006.05,317.108 2008.29,318.005 2010.53,318.912 2012.78,319.832 2015.02,320.764 2017.26,321.712 2019.5,322.675 2021.74,323.656 2023.98,324.655 2026.23,325.674 2028.47,326.713 2030.71,327.774 2032.95,328.857 2035.19,329.963 2037.44,331.093 2039.68,332.246 2041.92,333.422 2044.16,334.621 2046.4,335.843 2048.65,337.088 2050.89,338.353 2053.13,339.638 2055.37,340.942 2057.61,342.263 2059.86,343.599 2062.1,344.948 2064.34,346.307 2066.58,347.676 2068.82,349.051 2071.07,350.43 2073.31,351.811 2075.55,353.192 2077.79,354.57 2080.03,355.943 2082.28,357.31 2084.52,358.668 2086.76,360.017 2089,361.355 2091.24,362.68 2093.49,363.992 2095.73,365.29 2097.97,366.574 2100.21,367.844 2102.45,369.099 2104.7,370.339 2106.94,371.566 2109.18,372.779 2111.42,373.979 2113.66,375.166 2115.91,376.343 2118.15,377.51 2120.39,378.667 2122.63,379.816 2124.87,380.959 2127.11,382.097 2129.36,383.231 2131.6,384.362 2133.84,385.492 2136.08,386.623 2138.32,387.756 2140.57,388.892 2142.81,390.033 2145.05,391.181 2147.29,392.336 2149.53,393.501 2151.78,394.676 2154.02,395.863 2156.26,397.063 2158.5,398.277 2160.74,399.506 2162.99,400.75 2165.23,402.012 2167.47,403.29 2169.71,404.586 2171.95,405.899 2174.2,407.229 2176.44,408.577 2178.68,409.942 2180.92,411.322 2183.16,412.718 2185.41,414.128 2187.65,415.551 2189.89,416.986 2192.13,418.43 2194.37,419.883 2196.62,421.343 2198.86,422.808 2201.1,424.276 2203.34,425.746 2205.58,427.215 2207.83,428.683 2210.07,430.147 2212.31,431.606 2214.55,433.059 2216.79,434.505 2219.04,435.943 2221.28,437.372 2223.52,438.792 2225.76,440.202 2228,441.602 2230.25,442.993 2232.49,444.373 2234.73,445.744 2236.97,447.106 2239.21,448.459 2241.45,449.805 2243.7,451.144 2245.94,452.477 2248.18,453.806 2250.42,455.13 2252.66,456.452 2254.91,457.772 2257.15,459.092 2259.39,460.414 2261.63,461.737 2263.87,463.064 2266.12,464.396 2268.36,465.734 2270.6,467.078 2272.84,468.431 2275.08,469.793 2277.33,471.164 2279.57,472.546 2281.81,473.94 2284.05,475.345 2286.29,476.763 2288.54,478.192 2290.78,479.634 2293.02,481.088 2295.26,482.555 2297.5,484.032 2299.75,485.521 2301.99,487.02 2304.23,488.528 2306.47,490.044 2308.71,491.568 2310.96,493.097 2313.2,494.632 2315.44,496.17 2317.68,497.71 2319.92,499.252 2322.17,500.793 2324.41,502.334 2326.65,503.872 2328.89,505.406 2331.13,506.937 2333.38,508.463 2335.62,509.984 2337.86,511.498 2340.1,513.007 2342.34,514.509 2344.58,516.005 2346.83,517.495 2349.07,518.979 2351.31,520.457 2353.55,521.93 2355.79,523.399 2358.04,524.863 2360.28,526.324 2362.52,527.783 2364.76,529.24 2367,530.697 2369.25,532.154 2371.49,533.611 2373.73,535.071 2375.97,536.534 2378.21,538 2380.46,539.471 2382.7,540.947 2384.94,542.429 2387.18,543.918 2389.42,545.413 2391.67,546.917 2393.91,548.428 2396.15,549.947 2398.39,551.474 2400.63,553.009 2402.88,554.552 2405.12,556.103 2407.36,557.661 2409.6,559.226 2411.84,560.797 2414.09,562.373 2416.33,563.955 2418.57,565.54 2420.81,567.129 2423.05,568.721 2425.3,570.314 2427.54,571.908 2429.78,573.501 2432.02,575.095 2434.26,576.686 2436.51,578.276 2438.75,579.864 2440.99,581.448 2443.23,583.029 2445.47,584.606 2447.72,586.18 2449.96,587.749 2452.2,589.315 2454.44,590.877 2456.68,592.435 2458.92,593.991 2461.17,595.543 2463.41,597.093 2465.65,598.64 2467.89,600.187 2470.13,601.732 2472.38,603.276 2474.62,604.821 2476.86,606.367 2479.1,607.914 2481.34,609.463 2483.59,611.014 2485.83,612.569 2488.07,614.126 2490.31,615.688 2492.55,617.253 2494.8,618.823 2497.04,620.398 2499.28,621.977 2501.52,623.561 2503.76,625.15 2506.01,626.743 2508.25,628.341 2510.49,629.942 2512.73,631.548 2514.97,633.157 2517.22,634.769 2519.46,636.384 2521.7,638 2523.94,639.618 2526.18,641.237 2528.43,642.856 2530.67,644.474 2532.91,646.093 2535.15,647.709 2537.39,649.324 2539.64,650.937 2541.88,652.548 2544.12,654.155 2546.36,655.76 2548.6,657.361 2550.85,658.959 2553.09,660.553 2555.33,662.144 2557.57,663.732 2559.81,665.316 2562.05,666.897 2564.3,668.475 2566.54,670.051 2568.78,671.624 2571.02,673.195 2573.26,674.765 2575.51,676.334 2577.75,677.901 2579.99,679.468 2582.23,681.036 2584.47,682.603 2586.72,684.172 2588.96,685.741 2591.2,687.312 2593.44,688.885 2595.68,690.459 2597.93,692.036 2600.17,693.615 2602.41,695.197 2604.65,696.781 2606.89,698.367 2609.14,699.956 2611.38,701.546 2613.62,703.138 2615.86,704.732 2618.1,706.326 2620.35,707.921 2622.59,709.516 2624.83,711.111 2627.07,712.704 2629.31,714.295 2631.56,715.883 2633.8,717.468 2636.04,719.049 2638.28,720.624 2640.52,722.194 2642.77,723.757 2645.01,725.312 2647.25,726.859 2649.49,728.397 2651.73,729.926 2653.98,731.444 2656.22,732.952 2658.46,734.448 2660.7,735.932 2662.94,737.404 2665.19,738.865 2667.43,740.313 2669.67,741.748 2671.91,743.172 2674.15,744.584 2676.39,745.984 2678.64,747.374 2680.88,748.755 2683.12,750.126 2685.36,751.49 2687.6,752.848 2689.85,754.202 2692.09,755.553 2694.33,756.904 2696.57,758.256 2698.81,759.613 2701.06,760.977 2703.3,762.351 2705.54,763.738 2707.78,765.14 2710.02,766.561 2712.27,768.004 2714.51,769.469 2716.75,770.961 2718.99,772.479 "/>
+<path clip-path="url(#clip630)" d="M159.765 1723.83 L1352.76 1723.83 L1352.76 1087.05 L159.765 1087.05  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip634">
+    <rect x="159" y="1087" width="1194" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="193.529,1723.83 193.529,1087.05 "/>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="473.774,1723.83 473.774,1087.05 "/>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="754.018,1723.83 754.018,1087.05 "/>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1034.26,1723.83 1034.26,1087.05 "/>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1314.51,1723.83 1314.51,1087.05 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,1723.83 1352.76,1723.83 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="193.529,1723.83 193.529,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="473.774,1723.83 473.774,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="754.018,1723.83 754.018,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1034.26,1723.83 1034.26,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1314.51,1723.83 1314.51,1704.94 "/>
+<path clip-path="url(#clip630)" d="M193.529 1757.95 Q189.918 1757.95 188.089 1761.52 Q186.283 1765.06 186.283 1772.19 Q186.283 1779.3 188.089 1782.86 Q189.918 1786.4 193.529 1786.4 Q197.163 1786.4 198.969 1782.86 Q200.797 1779.3 200.797 1772.19 Q200.797 1765.06 198.969 1761.52 Q197.163 1757.95 193.529 1757.95 M193.529 1754.25 Q199.339 1754.25 202.394 1758.86 Q205.473 1763.44 205.473 1772.19 Q205.473 1780.92 202.394 1785.52 Q199.339 1790.11 193.529 1790.11 Q187.719 1790.11 184.64 1785.52 Q181.584 1780.92 181.584 1772.19 Q181.584 1763.44 184.64 1758.86 Q187.719 1754.25 193.529 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M433.392 1754.87 L451.748 1754.87 L451.748 1758.81 L437.674 1758.81 L437.674 1767.28 Q438.693 1766.93 439.711 1766.77 Q440.73 1766.59 441.748 1766.59 Q447.535 1766.59 450.915 1769.76 Q454.295 1772.93 454.295 1778.35 Q454.295 1783.93 450.822 1787.03 Q447.35 1790.11 441.031 1790.11 Q438.855 1790.11 436.586 1789.74 Q434.341 1789.37 431.934 1788.62 L431.934 1783.93 Q434.017 1785.06 436.239 1785.62 Q438.461 1786.17 440.938 1786.17 Q444.943 1786.17 447.281 1784.06 Q449.619 1781.96 449.619 1778.35 Q449.619 1774.74 447.281 1772.63 Q444.943 1770.52 440.938 1770.52 Q439.063 1770.52 437.188 1770.94 Q435.336 1771.36 433.392 1772.24 L433.392 1754.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M473.507 1757.95 Q469.896 1757.95 468.068 1761.52 Q466.262 1765.06 466.262 1772.19 Q466.262 1779.3 468.068 1782.86 Q469.896 1786.4 473.507 1786.4 Q477.142 1786.4 478.947 1782.86 Q480.776 1779.3 480.776 1772.19 Q480.776 1765.06 478.947 1761.52 Q477.142 1757.95 473.507 1757.95 M473.507 1754.25 Q479.318 1754.25 482.373 1758.86 Q485.452 1763.44 485.452 1772.19 Q485.452 1780.92 482.373 1785.52 Q479.318 1790.11 473.507 1790.11 Q467.697 1790.11 464.619 1785.52 Q461.563 1780.92 461.563 1772.19 Q461.563 1763.44 464.619 1758.86 Q467.697 1754.25 473.507 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M503.669 1757.95 Q500.058 1757.95 498.23 1761.52 Q496.424 1765.06 496.424 1772.19 Q496.424 1779.3 498.23 1782.86 Q500.058 1786.4 503.669 1786.4 Q507.304 1786.4 509.109 1782.86 Q510.938 1779.3 510.938 1772.19 Q510.938 1765.06 509.109 1761.52 Q507.304 1757.95 503.669 1757.95 M503.669 1754.25 Q509.479 1754.25 512.535 1758.86 Q515.614 1763.44 515.614 1772.19 Q515.614 1780.92 512.535 1785.52 Q509.479 1790.11 503.669 1790.11 Q497.859 1790.11 494.78 1785.52 Q491.725 1780.92 491.725 1772.19 Q491.725 1763.44 494.78 1758.86 Q497.859 1754.25 503.669 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M698.544 1785.5 L706.183 1785.5 L706.183 1759.13 L697.873 1760.8 L697.873 1756.54 L706.137 1754.87 L710.813 1754.87 L710.813 1785.5 L718.452 1785.5 L718.452 1789.43 L698.544 1789.43 L698.544 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M737.896 1757.95 Q734.285 1757.95 732.456 1761.52 Q730.651 1765.06 730.651 1772.19 Q730.651 1779.3 732.456 1782.86 Q734.285 1786.4 737.896 1786.4 Q741.53 1786.4 743.336 1782.86 Q745.164 1779.3 745.164 1772.19 Q745.164 1765.06 743.336 1761.52 Q741.53 1757.95 737.896 1757.95 M737.896 1754.25 Q743.706 1754.25 746.762 1758.86 Q749.84 1763.44 749.84 1772.19 Q749.84 1780.92 746.762 1785.52 Q743.706 1790.11 737.896 1790.11 Q732.086 1790.11 729.007 1785.52 Q725.951 1780.92 725.951 1772.19 Q725.951 1763.44 729.007 1758.86 Q732.086 1754.25 737.896 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M768.058 1757.95 Q764.447 1757.95 762.618 1761.52 Q760.812 1765.06 760.812 1772.19 Q760.812 1779.3 762.618 1782.86 Q764.447 1786.4 768.058 1786.4 Q771.692 1786.4 773.498 1782.86 Q775.326 1779.3 775.326 1772.19 Q775.326 1765.06 773.498 1761.52 Q771.692 1757.95 768.058 1757.95 M768.058 1754.25 Q773.868 1754.25 776.923 1758.86 Q780.002 1763.44 780.002 1772.19 Q780.002 1780.92 776.923 1785.52 Q773.868 1790.11 768.058 1790.11 Q762.248 1790.11 759.169 1785.52 Q756.113 1780.92 756.113 1772.19 Q756.113 1763.44 759.169 1758.86 Q762.248 1754.25 768.058 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M798.22 1757.95 Q794.609 1757.95 792.78 1761.52 Q790.974 1765.06 790.974 1772.19 Q790.974 1779.3 792.78 1782.86 Q794.609 1786.4 798.22 1786.4 Q801.854 1786.4 803.659 1782.86 Q805.488 1779.3 805.488 1772.19 Q805.488 1765.06 803.659 1761.52 Q801.854 1757.95 798.22 1757.95 M798.22 1754.25 Q804.03 1754.25 807.085 1758.86 Q810.164 1763.44 810.164 1772.19 Q810.164 1780.92 807.085 1785.52 Q804.03 1790.11 798.22 1790.11 Q792.409 1790.11 789.331 1785.52 Q786.275 1780.92 786.275 1772.19 Q786.275 1763.44 789.331 1758.86 Q792.409 1754.25 798.22 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M978.789 1785.5 L986.428 1785.5 L986.428 1759.13 L978.118 1760.8 L978.118 1756.54 L986.382 1754.87 L991.057 1754.87 L991.057 1785.5 L998.696 1785.5 L998.696 1789.43 L978.789 1789.43 L978.789 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1008.19 1754.87 L1026.54 1754.87 L1026.54 1758.81 L1012.47 1758.81 L1012.47 1767.28 Q1013.49 1766.93 1014.51 1766.77 Q1015.52 1766.59 1016.54 1766.59 Q1022.33 1766.59 1025.71 1769.76 Q1029.09 1772.93 1029.09 1778.35 Q1029.09 1783.93 1025.62 1787.03 Q1022.15 1790.11 1015.83 1790.11 Q1013.65 1790.11 1011.38 1789.74 Q1009.14 1789.37 1006.73 1788.62 L1006.73 1783.93 Q1008.81 1785.06 1011.03 1785.62 Q1013.26 1786.17 1015.73 1786.17 Q1019.74 1786.17 1022.08 1784.06 Q1024.41 1781.96 1024.41 1778.35 Q1024.41 1774.74 1022.08 1772.63 Q1019.74 1770.52 1015.73 1770.52 Q1013.86 1770.52 1011.98 1770.94 Q1010.13 1771.36 1008.19 1772.24 L1008.19 1754.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1048.3 1757.95 Q1044.69 1757.95 1042.86 1761.52 Q1041.06 1765.06 1041.06 1772.19 Q1041.06 1779.3 1042.86 1782.86 Q1044.69 1786.4 1048.3 1786.4 Q1051.94 1786.4 1053.74 1782.86 Q1055.57 1779.3 1055.57 1772.19 Q1055.57 1765.06 1053.74 1761.52 Q1051.94 1757.95 1048.3 1757.95 M1048.3 1754.25 Q1054.11 1754.25 1057.17 1758.86 Q1060.25 1763.44 1060.25 1772.19 Q1060.25 1780.92 1057.17 1785.52 Q1054.11 1790.11 1048.3 1790.11 Q1042.49 1790.11 1039.41 1785.52 Q1036.36 1780.92 1036.36 1772.19 Q1036.36 1763.44 1039.41 1758.86 Q1042.49 1754.25 1048.3 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1078.46 1757.95 Q1074.85 1757.95 1073.02 1761.52 Q1071.22 1765.06 1071.22 1772.19 Q1071.22 1779.3 1073.02 1782.86 Q1074.85 1786.4 1078.46 1786.4 Q1082.1 1786.4 1083.9 1782.86 Q1085.73 1779.3 1085.73 1772.19 Q1085.73 1765.06 1083.9 1761.52 Q1082.1 1757.95 1078.46 1757.95 M1078.46 1754.25 Q1084.27 1754.25 1087.33 1758.86 Q1090.41 1763.44 1090.41 1772.19 Q1090.41 1780.92 1087.33 1785.52 Q1084.27 1790.11 1078.46 1790.11 Q1072.65 1790.11 1069.58 1785.52 Q1066.52 1780.92 1066.52 1772.19 Q1066.52 1763.44 1069.58 1758.86 Q1072.65 1754.25 1078.46 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1263.12 1785.5 L1279.44 1785.5 L1279.44 1789.43 L1257.49 1789.43 L1257.49 1785.5 Q1260.16 1782.74 1264.74 1778.12 Q1269.35 1773.46 1270.53 1772.12 Q1272.77 1769.6 1273.65 1767.86 Q1274.55 1766.1 1274.55 1764.41 Q1274.55 1761.66 1272.61 1759.92 Q1270.69 1758.18 1267.59 1758.18 Q1265.39 1758.18 1262.93 1758.95 Q1260.5 1759.71 1257.73 1761.26 L1257.73 1756.54 Q1260.55 1755.41 1263 1754.83 Q1265.46 1754.25 1267.49 1754.25 Q1272.86 1754.25 1276.06 1756.93 Q1279.25 1759.62 1279.25 1764.11 Q1279.25 1766.24 1278.44 1768.16 Q1277.66 1770.06 1275.55 1772.65 Q1274.97 1773.32 1271.87 1776.54 Q1268.77 1779.74 1263.12 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1299.25 1757.95 Q1295.64 1757.95 1293.81 1761.52 Q1292.01 1765.06 1292.01 1772.19 Q1292.01 1779.3 1293.81 1782.86 Q1295.64 1786.4 1299.25 1786.4 Q1302.89 1786.4 1304.69 1782.86 Q1306.52 1779.3 1306.52 1772.19 Q1306.52 1765.06 1304.69 1761.52 Q1302.89 1757.95 1299.25 1757.95 M1299.25 1754.25 Q1305.06 1754.25 1308.12 1758.86 Q1311.2 1763.44 1311.2 1772.19 Q1311.2 1780.92 1308.12 1785.52 Q1305.06 1790.11 1299.25 1790.11 Q1293.44 1790.11 1290.36 1785.52 Q1287.31 1780.92 1287.31 1772.19 Q1287.31 1763.44 1290.36 1758.86 Q1293.44 1754.25 1299.25 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1329.42 1757.95 Q1325.8 1757.95 1323.98 1761.52 Q1322.17 1765.06 1322.17 1772.19 Q1322.17 1779.3 1323.98 1782.86 Q1325.8 1786.4 1329.42 1786.4 Q1333.05 1786.4 1334.86 1782.86 Q1336.68 1779.3 1336.68 1772.19 Q1336.68 1765.06 1334.86 1761.52 Q1333.05 1757.95 1329.42 1757.95 M1329.42 1754.25 Q1335.23 1754.25 1338.28 1758.86 Q1341.36 1763.44 1341.36 1772.19 Q1341.36 1780.92 1338.28 1785.52 Q1335.23 1790.11 1329.42 1790.11 Q1323.61 1790.11 1320.53 1785.52 Q1317.47 1780.92 1317.47 1772.19 Q1317.47 1763.44 1320.53 1758.86 Q1323.61 1754.25 1329.42 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1359.58 1757.95 Q1355.97 1757.95 1354.14 1761.52 Q1352.33 1765.06 1352.33 1772.19 Q1352.33 1779.3 1354.14 1782.86 Q1355.97 1786.4 1359.58 1786.4 Q1363.21 1786.4 1365.02 1782.86 Q1366.85 1779.3 1366.85 1772.19 Q1366.85 1765.06 1365.02 1761.52 Q1363.21 1757.95 1359.58 1757.95 M1359.58 1754.25 Q1365.39 1754.25 1368.44 1758.86 Q1371.52 1763.44 1371.52 1772.19 Q1371.52 1780.92 1368.44 1785.52 Q1365.39 1790.11 1359.58 1790.11 Q1353.77 1790.11 1350.69 1785.52 Q1347.63 1780.92 1347.63 1772.19 Q1347.63 1763.44 1350.69 1758.86 Q1353.77 1754.25 1359.58 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,1705.81 1352.76,1705.81 "/>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,1546.36 1352.76,1546.36 "/>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,1386.91 1352.76,1386.91 "/>
+<polyline clip-path="url(#clip634)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,1227.47 1352.76,1227.47 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,1723.83 159.765,1087.05 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,1705.81 178.663,1705.81 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,1546.36 178.663,1546.36 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,1386.91 178.663,1386.91 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,1227.47 178.663,1227.47 "/>
+<path clip-path="url(#clip630)" d="M105.821 1691.61 Q102.209 1691.61 100.381 1695.18 Q98.5752 1698.72 98.5752 1705.85 Q98.5752 1712.95 100.381 1716.52 Q102.209 1720.06 105.821 1720.06 Q109.455 1720.06 111.26 1716.52 Q113.089 1712.95 113.089 1705.85 Q113.089 1698.72 111.26 1695.18 Q109.455 1691.61 105.821 1691.61 M105.821 1687.91 Q111.631 1687.91 114.686 1692.51 Q117.765 1697.1 117.765 1705.85 Q117.765 1714.57 114.686 1719.18 Q111.631 1723.76 105.821 1723.76 Q100.01 1723.76 96.9317 1719.18 Q93.8761 1714.57 93.8761 1705.85 Q93.8761 1697.1 96.9317 1692.51 Q100.01 1687.91 105.821 1687.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M69.6865 1559.71 L86.0058 1559.71 L86.0058 1563.64 L64.0615 1563.64 L64.0615 1559.71 Q66.7235 1556.95 71.3068 1552.32 Q75.9133 1547.67 77.0938 1546.33 Q79.3392 1543.81 80.2188 1542.07 Q81.1216 1540.31 81.1216 1538.62 Q81.1216 1535.87 79.1771 1534.13 Q77.2559 1532.39 74.154 1532.39 Q71.955 1532.39 69.5013 1533.16 Q67.0707 1533.92 64.293 1535.47 L64.293 1530.75 Q67.117 1529.62 69.5707 1529.04 Q72.0244 1528.46 74.0614 1528.46 Q79.4318 1528.46 82.6262 1531.14 Q85.8206 1533.83 85.8206 1538.32 Q85.8206 1540.45 85.0105 1542.37 Q84.2234 1544.27 82.1169 1546.86 Q81.5382 1547.53 78.4364 1550.75 Q75.3346 1553.94 69.6865 1559.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 1532.16 Q102.209 1532.16 100.381 1535.73 Q98.5752 1539.27 98.5752 1546.4 Q98.5752 1553.5 100.381 1557.07 Q102.209 1560.61 105.821 1560.61 Q109.455 1560.61 111.26 1557.07 Q113.089 1553.5 113.089 1546.4 Q113.089 1539.27 111.26 1535.73 Q109.455 1532.16 105.821 1532.16 M105.821 1528.46 Q111.631 1528.46 114.686 1533.06 Q117.765 1537.65 117.765 1546.4 Q117.765 1555.12 114.686 1559.73 Q111.631 1564.31 105.821 1564.31 Q100.01 1564.31 96.9317 1559.73 Q93.8761 1555.12 93.8761 1546.4 Q93.8761 1537.65 96.9317 1533.06 Q100.01 1528.46 105.821 1528.46 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M78.5059 1373.71 L66.7004 1392.16 L78.5059 1392.16 L78.5059 1373.71 M77.279 1369.63 L83.1586 1369.63 L83.1586 1392.16 L88.0891 1392.16 L88.0891 1396.05 L83.1586 1396.05 L83.1586 1404.19 L78.5059 1404.19 L78.5059 1396.05 L62.9041 1396.05 L62.9041 1391.53 L77.279 1369.63 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 1372.71 Q102.209 1372.71 100.381 1376.28 Q98.5752 1379.82 98.5752 1386.95 Q98.5752 1394.06 100.381 1397.62 Q102.209 1401.16 105.821 1401.16 Q109.455 1401.16 111.26 1397.62 Q113.089 1394.06 113.089 1386.95 Q113.089 1379.82 111.26 1376.28 Q109.455 1372.71 105.821 1372.71 M105.821 1369.01 Q111.631 1369.01 114.686 1373.62 Q117.765 1378.2 117.765 1386.95 Q117.765 1395.68 114.686 1400.28 Q111.631 1404.87 105.821 1404.87 Q100.01 1404.87 96.9317 1400.28 Q93.8761 1395.68 93.8761 1386.95 Q93.8761 1378.2 96.9317 1373.62 Q100.01 1369.01 105.821 1369.01 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M76.2374 1225.6 Q73.0892 1225.6 71.2374 1227.75 Q69.4087 1229.91 69.4087 1233.66 Q69.4087 1237.38 71.2374 1239.56 Q73.0892 1241.71 76.2374 1241.71 Q79.3855 1241.71 81.2142 1239.56 Q83.066 1237.38 83.066 1233.66 Q83.066 1229.91 81.2142 1227.75 Q79.3855 1225.6 76.2374 1225.6 M85.5197 1210.95 L85.5197 1215.21 Q83.7605 1214.38 81.9549 1213.94 Q80.1725 1213.5 78.4133 1213.5 Q73.7837 1213.5 71.33 1216.62 Q68.8994 1219.75 68.5522 1226.06 Q69.9179 1224.05 71.9781 1222.99 Q74.0383 1221.9 76.5151 1221.9 Q81.7234 1221.9 84.7327 1225.07 Q87.7651 1228.22 87.7651 1233.66 Q87.7651 1238.98 84.6169 1242.2 Q81.4688 1245.42 76.2374 1245.42 Q70.242 1245.42 67.0707 1240.83 Q63.8995 1236.23 63.8995 1227.5 Q63.8995 1219.31 67.7883 1214.44 Q71.6772 1209.56 78.2281 1209.56 Q79.9873 1209.56 81.7697 1209.91 Q83.5753 1210.25 85.5197 1210.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 1213.26 Q102.209 1213.26 100.381 1216.83 Q98.5752 1220.37 98.5752 1227.5 Q98.5752 1234.61 100.381 1238.17 Q102.209 1241.71 105.821 1241.71 Q109.455 1241.71 111.26 1238.17 Q113.089 1234.61 113.089 1227.5 Q113.089 1220.37 111.26 1216.83 Q109.455 1213.26 105.821 1213.26 M105.821 1209.56 Q111.631 1209.56 114.686 1214.17 Q117.765 1218.75 117.765 1227.5 Q117.765 1236.23 114.686 1240.83 Q111.631 1245.42 105.821 1245.42 Q100.01 1245.42 96.9317 1240.83 Q93.8761 1236.23 93.8761 1227.5 Q93.8761 1218.75 96.9317 1214.17 Q100.01 1209.56 105.821 1209.56 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M433.849 945.429 L442.032 945.429 L442.032 999.023 L471.482 999.023 L471.482 1005.91 L433.849 1005.91 L433.849 945.429 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M495.868 965.765 Q489.873 965.765 486.389 970.464 Q482.905 975.122 482.905 983.265 Q482.905 991.407 486.349 996.106 Q489.832 1000.76 495.868 1000.76 Q501.823 1000.76 505.307 996.066 Q508.791 991.367 508.791 983.265 Q508.791 975.203 505.307 970.504 Q501.823 965.765 495.868 965.765 M495.868 959.445 Q505.59 959.445 511.14 965.765 Q516.69 972.084 516.69 983.265 Q516.69 994.405 511.14 1000.76 Q505.59 1007.08 495.868 1007.08 Q486.106 1007.08 480.556 1000.76 Q475.047 994.405 475.047 983.265 Q475.047 972.084 480.556 965.765 Q486.106 959.445 495.868 959.445 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M566.759 978.525 L566.759 1005.91 L559.305 1005.91 L559.305 978.768 Q559.305 972.327 556.794 969.127 Q554.282 965.927 549.259 965.927 Q543.223 965.927 539.74 969.775 Q536.256 973.624 536.256 980.267 L536.256 1005.91 L528.762 1005.91 L528.762 960.539 L536.256 960.539 L536.256 967.588 Q538.929 963.496 542.535 961.471 Q546.181 959.445 550.92 959.445 Q558.738 959.445 562.749 964.307 Q566.759 969.127 566.759 978.525 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M611.481 982.698 Q611.481 974.596 608.119 970.14 Q604.797 965.684 598.761 965.684 Q592.766 965.684 589.404 970.14 Q586.082 974.596 586.082 982.698 Q586.082 990.759 589.404 995.215 Q592.766 999.671 598.761 999.671 Q604.797 999.671 608.119 995.215 Q611.481 990.759 611.481 982.698 M618.935 1000.28 Q618.935 1011.86 613.79 1017.49 Q608.645 1023.17 598.032 1023.17 Q594.103 1023.17 590.619 1022.56 Q587.135 1021.99 583.854 1020.78 L583.854 1013.53 Q587.135 1015.31 590.335 1016.16 Q593.536 1017.01 596.857 1017.01 Q604.189 1017.01 607.835 1013.16 Q611.481 1009.35 611.481 1001.62 L611.481 997.929 Q609.172 1001.94 605.567 1003.92 Q601.961 1005.91 596.938 1005.91 Q588.593 1005.91 583.489 999.549 Q578.385 993.189 578.385 982.698 Q578.385 972.165 583.489 965.805 Q588.593 959.445 596.938 959.445 Q601.961 959.445 605.567 961.43 Q609.172 963.415 611.481 967.426 L611.481 960.539 L618.935 960.539 L618.935 1000.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M634.288 960.539 L641.741 960.539 L641.741 1005.91 L634.288 1005.91 L634.288 960.539 M634.288 942.877 L641.741 942.877 L641.741 952.316 L634.288 952.316 L634.288 942.877 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M664.71 947.657 L664.71 960.539 L680.063 960.539 L680.063 966.332 L664.71 966.332 L664.71 990.961 Q664.71 996.511 666.209 998.091 Q667.748 999.671 672.407 999.671 L680.063 999.671 L680.063 1005.91 L672.407 1005.91 Q663.778 1005.91 660.497 1002.71 Q657.216 999.468 657.216 990.961 L657.216 966.332 L651.747 966.332 L651.747 960.539 L657.216 960.539 L657.216 947.657 L664.71 947.657 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M689.096 988.004 L689.096 960.539 L696.55 960.539 L696.55 987.721 Q696.55 994.162 699.062 997.402 Q701.573 1000.6 706.596 1000.6 Q712.632 1000.6 716.116 996.754 Q719.64 992.906 719.64 986.262 L719.64 960.539 L727.094 960.539 L727.094 1005.91 L719.64 1005.91 L719.64 998.942 Q716.926 1003.07 713.321 1005.1 Q709.756 1007.08 705.016 1007.08 Q697.198 1007.08 693.147 1002.22 Q689.096 997.362 689.096 988.004 M707.852 959.445 L707.852 959.445 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M772.302 967.426 L772.302 942.877 L779.756 942.877 L779.756 1005.91 L772.302 1005.91 L772.302 999.104 Q769.952 1003.15 766.347 1005.14 Q762.782 1007.08 757.759 1007.08 Q749.536 1007.08 744.351 1000.52 Q739.206 993.959 739.206 983.265 Q739.206 972.57 744.351 966.008 Q749.536 959.445 757.759 959.445 Q762.782 959.445 766.347 961.43 Q769.952 963.375 772.302 967.426 M746.903 983.265 Q746.903 991.488 750.265 996.187 Q753.668 1000.85 759.582 1000.85 Q765.496 1000.85 768.899 996.187 Q772.302 991.488 772.302 983.265 Q772.302 975.041 768.899 970.383 Q765.496 965.684 759.582 965.684 Q753.668 965.684 750.265 970.383 Q746.903 975.041 746.903 983.265 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M833.916 981.361 L833.916 985.007 L799.646 985.007 Q800.132 992.703 804.264 996.754 Q808.436 1000.76 815.849 1000.76 Q820.143 1000.76 824.154 999.711 Q828.204 998.658 832.174 996.552 L832.174 1003.6 Q828.164 1005.3 823.951 1006.19 Q819.738 1007.08 815.404 1007.08 Q804.547 1007.08 798.187 1000.76 Q791.868 994.445 791.868 983.67 Q791.868 972.53 797.863 966.008 Q803.899 959.445 814.107 959.445 Q823.262 959.445 828.569 965.36 Q833.916 971.234 833.916 981.361 M826.463 979.173 Q826.382 973.057 823.019 969.411 Q819.698 965.765 814.188 965.765 Q807.95 965.765 804.183 969.289 Q800.456 972.813 799.889 979.214 L826.463 979.173 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M890.426 942.958 Q884.998 952.275 882.365 961.39 Q879.732 970.504 879.732 979.862 Q879.732 989.22 882.365 998.415 Q885.039 1007.57 890.426 1016.85 L883.945 1016.85 Q877.869 1007.33 874.83 998.132 Q871.833 988.936 871.833 979.862 Q871.833 970.829 874.83 961.673 Q877.828 952.518 883.945 942.958 L890.426 942.958 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M934.743 967.426 L934.743 942.877 L942.197 942.877 L942.197 1005.91 L934.743 1005.91 L934.743 999.104 Q932.394 1003.15 928.788 1005.14 Q925.224 1007.08 920.2 1007.08 Q911.977 1007.08 906.792 1000.52 Q901.647 993.959 901.647 983.265 Q901.647 972.57 906.792 966.008 Q911.977 959.445 920.2 959.445 Q925.224 959.445 928.788 961.43 Q932.394 963.375 934.743 967.426 M909.344 983.265 Q909.344 991.488 912.706 996.187 Q916.109 1000.85 922.023 1000.85 Q927.938 1000.85 931.34 996.187 Q934.743 991.488 934.743 983.265 Q934.743 975.041 931.34 970.383 Q927.938 965.684 922.023 965.684 Q916.109 965.684 912.706 970.383 Q909.344 975.041 909.344 983.265 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M996.357 981.361 L996.357 985.007 L962.087 985.007 Q962.573 992.703 966.705 996.754 Q970.877 1000.76 978.29 1000.76 Q982.584 1000.76 986.595 999.711 Q990.646 998.658 994.616 996.552 L994.616 1003.6 Q990.605 1005.3 986.392 1006.19 Q982.179 1007.08 977.845 1007.08 Q966.988 1007.08 960.629 1000.76 Q954.309 994.445 954.309 983.67 Q954.309 972.53 960.304 966.008 Q966.34 959.445 976.549 959.445 Q985.704 959.445 991.01 965.36 Q996.357 971.234 996.357 981.361 M988.904 979.173 Q988.823 973.057 985.461 969.411 Q982.139 965.765 976.63 965.765 Q970.391 965.765 966.624 969.289 Q962.897 972.813 962.33 979.214 L988.904 979.173 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1038.45 982.698 Q1038.45 974.596 1035.08 970.14 Q1031.76 965.684 1025.73 965.684 Q1019.73 965.684 1016.37 970.14 Q1013.05 974.596 1013.05 982.698 Q1013.05 990.759 1016.37 995.215 Q1019.73 999.671 1025.73 999.671 Q1031.76 999.671 1035.08 995.215 Q1038.45 990.759 1038.45 982.698 M1045.9 1000.28 Q1045.9 1011.86 1040.76 1017.49 Q1035.61 1023.17 1025 1023.17 Q1021.07 1023.17 1017.58 1022.56 Q1014.1 1021.99 1010.82 1020.78 L1010.82 1013.53 Q1014.1 1015.31 1017.3 1016.16 Q1020.5 1017.01 1023.82 1017.01 Q1031.15 1017.01 1034.8 1013.16 Q1038.45 1009.35 1038.45 1001.62 L1038.45 997.929 Q1036.14 1001.94 1032.53 1003.92 Q1028.93 1005.91 1023.9 1005.91 Q1015.56 1005.91 1010.45 999.549 Q1005.35 993.189 1005.35 982.698 Q1005.35 972.165 1010.45 965.805 Q1015.56 959.445 1023.9 959.445 Q1028.93 959.445 1032.53 961.43 Q1036.14 963.415 1038.45 967.426 L1038.45 960.539 L1045.9 960.539 L1045.9 1000.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1060.08 942.958 L1066.56 942.958 Q1072.64 952.518 1075.63 961.673 Q1078.67 970.829 1078.67 979.862 Q1078.67 988.936 1075.63 998.132 Q1072.64 1007.33 1066.56 1016.85 L1060.08 1016.85 Q1065.47 1007.57 1068.1 998.415 Q1070.77 989.22 1070.77 979.862 Q1070.77 970.504 1068.1 961.39 Q1065.47 952.275 1060.08 942.958 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip634)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="193.529,1705.81 195.771,1703.6 198.013,1701.39 200.255,1699.18 202.497,1696.97 204.739,1694.76 206.981,1692.55 209.223,1690.34 211.464,1688.13 213.706,1685.92 215.948,1683.71 218.19,1681.51 220.432,1679.3 222.674,1677.09 224.916,1674.88 227.158,1672.67 229.4,1670.47 231.642,1668.26 233.884,1666.06 236.126,1663.85 238.368,1661.65 240.61,1659.45 242.852,1657.24 245.094,1655.04 247.336,1652.85 249.578,1650.65 251.82,1648.45 254.062,1646.26 256.304,1644.07 258.546,1641.88 260.788,1639.69 263.03,1637.5 265.271,1635.32 267.513,1633.14 269.755,1630.96 271.997,1628.78 274.239,1626.61 276.481,1624.44 278.723,1622.28 280.965,1620.12 283.207,1617.96 285.449,1615.8 287.691,1613.65 289.933,1611.51 292.175,1609.36 294.417,1607.23 296.659,1605.09 298.901,1602.96 301.143,1600.84 303.385,1598.72 305.627,1596.6 307.869,1594.49 310.111,1592.39 312.353,1590.29 314.595,1588.19 316.837,1586.1 319.078,1584.01 321.32,1581.93 323.562,1579.85 325.804,1577.78 328.046,1575.71 330.288,1573.65 332.53,1571.59 334.772,1569.53 337.014,1567.48 339.256,1565.43 341.498,1563.39 343.74,1561.35 345.982,1559.31 348.224,1557.28 350.466,1555.25 352.708,1553.22 354.95,1551.2 357.192,1549.18 359.434,1547.16 361.676,1545.15 363.918,1543.13 366.16,1541.13 368.402,1539.12 370.644,1537.12 372.885,1535.12 375.127,1533.12 377.369,1531.13 379.611,1529.13 381.853,1527.14 384.095,1525.16 386.337,1523.17 388.579,1521.19 390.821,1519.2 393.063,1517.23 395.305,1515.25 397.547,1513.27 399.789,1511.3 402.031,1509.33 404.273,1507.36 406.515,1505.39 408.757,1503.43 410.999,1501.46 413.241,1499.5 415.483,1497.54 417.725,1495.59 419.967,1493.63 422.209,1491.68 424.451,1489.72 426.692,1487.77 428.934,1485.83 431.176,1483.88 433.418,1481.94 435.66,1479.99 437.902,1478.05 440.144,1476.12 442.386,1474.18 444.628,1472.25 446.87,1470.32 449.112,1468.39 451.354,1466.46 453.596,1464.54 455.838,1462.62 458.08,1460.7 460.322,1458.78 462.564,1456.87 464.806,1454.96 467.048,1453.05 469.29,1451.15 471.532,1449.25 473.774,1447.35 476.016,1445.46 478.258,1443.57 480.5,1441.68 482.741,1439.8 484.983,1437.92 487.225,1436.04 489.467,1434.17 491.709,1432.31 493.951,1430.45 496.193,1428.59 498.435,1426.74 500.677,1424.89 502.919,1423.05 505.161,1421.22 507.403,1419.39 509.645,1417.56 511.887,1415.74 514.129,1413.93 516.371,1412.12 518.613,1410.32 520.855,1408.52 523.097,1406.73 525.339,1404.94 527.581,1403.16 529.823,1401.39 532.065,1399.62 534.307,1397.86 536.548,1396.1 538.79,1394.35 541.032,1392.6 543.274,1390.86 545.516,1389.12 547.758,1387.39 550,1385.67 552.242,1383.95 554.484,1382.23 556.726,1380.52 558.968,1378.81 561.21,1377.11 563.452,1375.41 565.694,1373.72 567.936,1372.03 570.178,1370.34 572.42,1368.66 574.662,1366.98 576.904,1365.31 579.146,1363.64 581.388,1361.98 583.63,1360.31 585.872,1358.66 588.114,1357 590.355,1355.35 592.597,1353.71 594.839,1352.06 597.081,1350.43 599.323,1348.79 601.565,1347.16 603.807,1345.53 606.049,1343.91 608.291,1342.29 610.533,1340.67 612.775,1339.06 615.017,1337.45 617.259,1335.84 619.501,1334.24 621.743,1332.65 623.985,1331.05 626.227,1329.47 628.469,1327.88 630.711,1326.3 632.953,1324.73 635.195,1323.16 637.437,1321.6 639.679,1320.04 641.921,1318.48 644.162,1316.93 646.404,1315.39 648.646,1313.85 650.888,1312.32 653.13,1310.79 655.372,1309.27 657.614,1307.75 659.856,1306.24 662.098,1304.73 664.34,1303.24 666.582,1301.74 668.824,1300.26 671.066,1298.78 673.308,1297.3 675.55,1295.84 677.792,1294.38 680.034,1292.92 682.276,1291.47 684.518,1290.03 686.76,1288.59 689.002,1287.16 691.244,1285.74 693.486,1284.32 695.728,1282.91 697.969,1281.5 700.211,1280.1 702.453,1278.7 704.695,1277.31 706.937,1275.93 709.179,1274.55 711.421,1273.17 713.663,1271.81 715.905,1270.44 718.147,1269.08 720.389,1267.73 722.631,1266.38 724.873,1265.04 727.115,1263.7 729.357,1262.37 731.599,1261.04 733.841,1259.72 736.083,1258.4 738.325,1257.08 740.567,1255.78 742.809,1254.47 745.051,1253.17 747.293,1251.88 749.535,1250.59 751.776,1249.3 754.018,1248.02 756.26,1246.75 758.502,1245.48 760.744,1244.22 762.986,1242.96 765.228,1241.7 767.47,1240.46 769.712,1239.21 771.954,1237.98 774.196,1236.74 776.438,1235.52 778.68,1234.3 780.922,1233.09 783.164,1231.88 785.406,1230.68 787.648,1229.48 789.89,1228.29 792.132,1227.11 794.374,1225.93 796.616,1224.76 798.858,1223.6 801.1,1222.44 803.342,1221.29 805.583,1220.14 807.825,1219 810.067,1217.87 812.309,1216.75 814.551,1215.63 816.793,1214.52 819.035,1213.41 821.277,1212.31 823.519,1211.21 825.761,1210.13 828.003,1209.04 830.245,1207.97 832.487,1206.9 834.729,1205.83 836.971,1204.78 839.213,1203.72 841.455,1202.68 843.697,1201.64 845.939,1200.6 848.181,1199.57 850.423,1198.55 852.665,1197.53 854.907,1196.51 857.149,1195.51 859.391,1194.5 861.632,1193.51 863.874,1192.52 866.116,1191.53 868.358,1190.55 870.6,1189.58 872.842,1188.61 875.084,1187.65 877.326,1186.69 879.568,1185.74 881.81,1184.79 884.052,1183.85 886.294,1182.92 888.536,1181.99 890.778,1181.06 893.02,1180.15 895.262,1179.24 897.504,1178.33 899.746,1177.43 901.988,1176.54 904.23,1175.66 906.472,1174.78 908.714,1173.9 910.956,1173.04 913.198,1172.18 915.439,1171.32 917.681,1170.47 919.923,1169.63 922.165,1168.8 924.407,1167.97 926.649,1167.15 928.891,1166.33 931.133,1165.52 933.375,1164.72 935.617,1163.92 937.859,1163.13 940.101,1162.34 942.343,1161.57 944.585,1160.79 946.827,1160.03 949.069,1159.27 951.311,1158.51 953.553,1157.76 955.795,1157.02 958.037,1156.28 960.279,1155.55 962.521,1154.83 964.763,1154.11 967.005,1153.4 969.246,1152.69 971.488,1151.99 973.73,1151.29 975.972,1150.6 978.214,1149.91 980.456,1149.23 982.698,1148.56 984.94,1147.89 987.182,1147.23 989.424,1146.58 991.666,1145.93 993.908,1145.28 996.15,1144.64 998.392,1144.01 1000.63,1143.38 1002.88,1142.76 1005.12,1142.15 1007.36,1141.54 1009.6,1140.94 1011.84,1140.34 1014.09,1139.75 1016.33,1139.16 1018.57,1138.59 1020.81,1138.01 1023.05,1137.45 1025.3,1136.88 1027.54,1136.33 1029.78,1135.78 1032.02,1135.24 1034.26,1134.7 1036.51,1134.17 1038.75,1133.65 1040.99,1133.13 1043.23,1132.61 1045.47,1132.11 1047.72,1131.6 1049.96,1131.11 1052.2,1130.62 1054.44,1130.13 1056.68,1129.65 1058.92,1129.18 1061.17,1128.71 1063.41,1128.25 1065.65,1127.79 1067.89,1127.34 1070.13,1126.9 1072.38,1126.45 1074.62,1126.02 1076.86,1125.59 1079.1,1125.17 1081.34,1124.75 1083.59,1124.33 1085.83,1123.92 1088.07,1123.52 1090.31,1123.12 1092.55,1122.73 1094.8,1122.34 1097.04,1121.96 1099.28,1121.58 1101.52,1121.21 1103.76,1120.85 1106.01,1120.49 1108.25,1120.13 1110.49,1119.78 1112.73,1119.43 1114.97,1119.09 1117.22,1118.76 1119.46,1118.43 1121.7,1118.1 1123.94,1117.78 1126.18,1117.47 1128.43,1117.16 1130.67,1116.86 1132.91,1116.56 1135.15,1116.26 1137.39,1115.97 1139.64,1115.69 1141.88,1115.41 1144.12,1115.13 1146.36,1114.86 1148.6,1114.6 1150.85,1114.34 1153.09,1114.08 1155.33,1113.83 1157.57,1113.58 1159.81,1113.34 1162.05,1113.1 1164.3,1112.87 1166.54,1112.64 1168.78,1112.41 1171.02,1112.19 1173.26,1111.98 1175.51,1111.76 1177.75,1111.56 1179.99,1111.35 1182.23,1111.15 1184.47,1110.96 1186.72,1110.77 1188.96,1110.58 1191.2,1110.4 1193.44,1110.22 1195.68,1110.04 1197.93,1109.87 1200.17,1109.7 1202.41,1109.54 1204.65,1109.38 1206.89,1109.22 1209.14,1109.07 1211.38,1108.92 1213.62,1108.77 1215.86,1108.63 1218.1,1108.49 1220.35,1108.36 1222.59,1108.23 1224.83,1108.1 1227.07,1107.97 1229.31,1107.85 1231.56,1107.73 1233.8,1107.62 1236.04,1107.51 1238.28,1107.4 1240.52,1107.29 1242.77,1107.19 1245.01,1107.09 1247.25,1107 1249.49,1106.9 1251.73,1106.81 1253.98,1106.72 1256.22,1106.64 1258.46,1106.55 1260.7,1106.47 1262.94,1106.4 1265.19,1106.32 1267.43,1106.25 1269.67,1106.17 1271.91,1106.11 1274.15,1106.04 1276.39,1105.97 1278.64,1105.91 1280.88,1105.85 1283.12,1105.79 1285.36,1105.74 1287.6,1105.68 1289.85,1105.63 1292.09,1105.58 1294.33,1105.53 1296.57,1105.48 1298.81,1105.43 1301.06,1105.39 1303.3,1105.34 1305.54,1105.3 1307.78,1105.26 1310.02,1105.22 1312.27,1105.18 1314.51,1105.14 1316.75,1105.1 1318.99,1105.07 "/>
+<path clip-path="url(#clip630)" d="M1559.76 1723.83 L2752.76 1723.83 L2752.76 1087.05 L1559.76 1087.05  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip635">
+    <rect x="1559" y="1087" width="1194" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1593.53,1723.83 1593.53,1087.05 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1873.77,1723.83 1873.77,1087.05 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2154.02,1723.83 2154.02,1087.05 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2434.26,1723.83 2434.26,1087.05 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2714.51,1723.83 2714.51,1087.05 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1723.83 2752.76,1723.83 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1593.53,1723.83 1593.53,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1873.77,1723.83 1873.77,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2154.02,1723.83 2154.02,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2434.26,1723.83 2434.26,1704.94 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2714.51,1723.83 2714.51,1704.94 "/>
+<path clip-path="url(#clip630)" d="M1593.53 1757.95 Q1589.92 1757.95 1588.09 1761.52 Q1586.28 1765.06 1586.28 1772.19 Q1586.28 1779.3 1588.09 1782.86 Q1589.92 1786.4 1593.53 1786.4 Q1597.16 1786.4 1598.97 1782.86 Q1600.8 1779.3 1600.8 1772.19 Q1600.8 1765.06 1598.97 1761.52 Q1597.16 1757.95 1593.53 1757.95 M1593.53 1754.25 Q1599.34 1754.25 1602.39 1758.86 Q1605.47 1763.44 1605.47 1772.19 Q1605.47 1780.92 1602.39 1785.52 Q1599.34 1790.11 1593.53 1790.11 Q1587.72 1790.11 1584.64 1785.52 Q1581.58 1780.92 1581.58 1772.19 Q1581.58 1763.44 1584.64 1758.86 Q1587.72 1754.25 1593.53 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1833.39 1754.87 L1851.75 1754.87 L1851.75 1758.81 L1837.67 1758.81 L1837.67 1767.28 Q1838.69 1766.93 1839.71 1766.77 Q1840.73 1766.59 1841.75 1766.59 Q1847.54 1766.59 1850.91 1769.76 Q1854.29 1772.93 1854.29 1778.35 Q1854.29 1783.93 1850.82 1787.03 Q1847.35 1790.11 1841.03 1790.11 Q1838.85 1790.11 1836.59 1789.74 Q1834.34 1789.37 1831.93 1788.62 L1831.93 1783.93 Q1834.02 1785.06 1836.24 1785.62 Q1838.46 1786.17 1840.94 1786.17 Q1844.94 1786.17 1847.28 1784.06 Q1849.62 1781.96 1849.62 1778.35 Q1849.62 1774.74 1847.28 1772.63 Q1844.94 1770.52 1840.94 1770.52 Q1839.06 1770.52 1837.19 1770.94 Q1835.34 1771.36 1833.39 1772.24 L1833.39 1754.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1873.51 1757.95 Q1869.9 1757.95 1868.07 1761.52 Q1866.26 1765.06 1866.26 1772.19 Q1866.26 1779.3 1868.07 1782.86 Q1869.9 1786.4 1873.51 1786.4 Q1877.14 1786.4 1878.95 1782.86 Q1880.78 1779.3 1880.78 1772.19 Q1880.78 1765.06 1878.95 1761.52 Q1877.14 1757.95 1873.51 1757.95 M1873.51 1754.25 Q1879.32 1754.25 1882.37 1758.86 Q1885.45 1763.44 1885.45 1772.19 Q1885.45 1780.92 1882.37 1785.52 Q1879.32 1790.11 1873.51 1790.11 Q1867.7 1790.11 1864.62 1785.52 Q1861.56 1780.92 1861.56 1772.19 Q1861.56 1763.44 1864.62 1758.86 Q1867.7 1754.25 1873.51 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1903.67 1757.95 Q1900.06 1757.95 1898.23 1761.52 Q1896.42 1765.06 1896.42 1772.19 Q1896.42 1779.3 1898.23 1782.86 Q1900.06 1786.4 1903.67 1786.4 Q1907.3 1786.4 1909.11 1782.86 Q1910.94 1779.3 1910.94 1772.19 Q1910.94 1765.06 1909.11 1761.52 Q1907.3 1757.95 1903.67 1757.95 M1903.67 1754.25 Q1909.48 1754.25 1912.53 1758.86 Q1915.61 1763.44 1915.61 1772.19 Q1915.61 1780.92 1912.53 1785.52 Q1909.48 1790.11 1903.67 1790.11 Q1897.86 1790.11 1894.78 1785.52 Q1891.72 1780.92 1891.72 1772.19 Q1891.72 1763.44 1894.78 1758.86 Q1897.86 1754.25 1903.67 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2098.54 1785.5 L2106.18 1785.5 L2106.18 1759.13 L2097.87 1760.8 L2097.87 1756.54 L2106.14 1754.87 L2110.81 1754.87 L2110.81 1785.5 L2118.45 1785.5 L2118.45 1789.43 L2098.54 1789.43 L2098.54 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2137.9 1757.95 Q2134.28 1757.95 2132.46 1761.52 Q2130.65 1765.06 2130.65 1772.19 Q2130.65 1779.3 2132.46 1782.86 Q2134.28 1786.4 2137.9 1786.4 Q2141.53 1786.4 2143.34 1782.86 Q2145.16 1779.3 2145.16 1772.19 Q2145.16 1765.06 2143.34 1761.52 Q2141.53 1757.95 2137.9 1757.95 M2137.9 1754.25 Q2143.71 1754.25 2146.76 1758.86 Q2149.84 1763.44 2149.84 1772.19 Q2149.84 1780.92 2146.76 1785.52 Q2143.71 1790.11 2137.9 1790.11 Q2132.09 1790.11 2129.01 1785.52 Q2125.95 1780.92 2125.95 1772.19 Q2125.95 1763.44 2129.01 1758.86 Q2132.09 1754.25 2137.9 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2168.06 1757.95 Q2164.45 1757.95 2162.62 1761.52 Q2160.81 1765.06 2160.81 1772.19 Q2160.81 1779.3 2162.62 1782.86 Q2164.45 1786.4 2168.06 1786.4 Q2171.69 1786.4 2173.5 1782.86 Q2175.33 1779.3 2175.33 1772.19 Q2175.33 1765.06 2173.5 1761.52 Q2171.69 1757.95 2168.06 1757.95 M2168.06 1754.25 Q2173.87 1754.25 2176.92 1758.86 Q2180 1763.44 2180 1772.19 Q2180 1780.92 2176.92 1785.52 Q2173.87 1790.11 2168.06 1790.11 Q2162.25 1790.11 2159.17 1785.52 Q2156.11 1780.92 2156.11 1772.19 Q2156.11 1763.44 2159.17 1758.86 Q2162.25 1754.25 2168.06 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2198.22 1757.95 Q2194.61 1757.95 2192.78 1761.52 Q2190.97 1765.06 2190.97 1772.19 Q2190.97 1779.3 2192.78 1782.86 Q2194.61 1786.4 2198.22 1786.4 Q2201.85 1786.4 2203.66 1782.86 Q2205.49 1779.3 2205.49 1772.19 Q2205.49 1765.06 2203.66 1761.52 Q2201.85 1757.95 2198.22 1757.95 M2198.22 1754.25 Q2204.03 1754.25 2207.09 1758.86 Q2210.16 1763.44 2210.16 1772.19 Q2210.16 1780.92 2207.09 1785.52 Q2204.03 1790.11 2198.22 1790.11 Q2192.41 1790.11 2189.33 1785.52 Q2186.28 1780.92 2186.28 1772.19 Q2186.28 1763.44 2189.33 1758.86 Q2192.41 1754.25 2198.22 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2378.79 1785.5 L2386.43 1785.5 L2386.43 1759.13 L2378.12 1760.8 L2378.12 1756.54 L2386.38 1754.87 L2391.06 1754.87 L2391.06 1785.5 L2398.7 1785.5 L2398.7 1789.43 L2378.79 1789.43 L2378.79 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2408.19 1754.87 L2426.54 1754.87 L2426.54 1758.81 L2412.47 1758.81 L2412.47 1767.28 Q2413.49 1766.93 2414.51 1766.77 Q2415.52 1766.59 2416.54 1766.59 Q2422.33 1766.59 2425.71 1769.76 Q2429.09 1772.93 2429.09 1778.35 Q2429.09 1783.93 2425.62 1787.03 Q2422.15 1790.11 2415.83 1790.11 Q2413.65 1790.11 2411.38 1789.74 Q2409.14 1789.37 2406.73 1788.62 L2406.73 1783.93 Q2408.81 1785.06 2411.03 1785.62 Q2413.26 1786.17 2415.73 1786.17 Q2419.74 1786.17 2422.08 1784.06 Q2424.41 1781.96 2424.41 1778.35 Q2424.41 1774.74 2422.08 1772.63 Q2419.74 1770.52 2415.73 1770.52 Q2413.86 1770.52 2411.98 1770.94 Q2410.13 1771.36 2408.19 1772.24 L2408.19 1754.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2448.3 1757.95 Q2444.69 1757.95 2442.86 1761.52 Q2441.06 1765.06 2441.06 1772.19 Q2441.06 1779.3 2442.86 1782.86 Q2444.69 1786.4 2448.3 1786.4 Q2451.94 1786.4 2453.74 1782.86 Q2455.57 1779.3 2455.57 1772.19 Q2455.57 1765.06 2453.74 1761.52 Q2451.94 1757.95 2448.3 1757.95 M2448.3 1754.25 Q2454.11 1754.25 2457.17 1758.86 Q2460.25 1763.44 2460.25 1772.19 Q2460.25 1780.92 2457.17 1785.52 Q2454.11 1790.11 2448.3 1790.11 Q2442.49 1790.11 2439.41 1785.52 Q2436.36 1780.92 2436.36 1772.19 Q2436.36 1763.44 2439.41 1758.86 Q2442.49 1754.25 2448.3 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2478.46 1757.95 Q2474.85 1757.95 2473.02 1761.52 Q2471.22 1765.06 2471.22 1772.19 Q2471.22 1779.3 2473.02 1782.86 Q2474.85 1786.4 2478.46 1786.4 Q2482.1 1786.4 2483.9 1782.86 Q2485.73 1779.3 2485.73 1772.19 Q2485.73 1765.06 2483.9 1761.52 Q2482.1 1757.95 2478.46 1757.95 M2478.46 1754.25 Q2484.27 1754.25 2487.33 1758.86 Q2490.41 1763.44 2490.41 1772.19 Q2490.41 1780.92 2487.33 1785.52 Q2484.27 1790.11 2478.46 1790.11 Q2472.65 1790.11 2469.58 1785.52 Q2466.52 1780.92 2466.52 1772.19 Q2466.52 1763.44 2469.58 1758.86 Q2472.65 1754.25 2478.46 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2663.12 1785.5 L2679.44 1785.5 L2679.44 1789.43 L2657.49 1789.43 L2657.49 1785.5 Q2660.16 1782.74 2664.74 1778.12 Q2669.35 1773.46 2670.53 1772.12 Q2672.77 1769.6 2673.65 1767.86 Q2674.55 1766.1 2674.55 1764.41 Q2674.55 1761.66 2672.61 1759.92 Q2670.69 1758.18 2667.59 1758.18 Q2665.39 1758.18 2662.93 1758.95 Q2660.5 1759.71 2657.73 1761.26 L2657.73 1756.54 Q2660.55 1755.41 2663 1754.83 Q2665.46 1754.25 2667.49 1754.25 Q2672.86 1754.25 2676.06 1756.93 Q2679.25 1759.62 2679.25 1764.11 Q2679.25 1766.24 2678.44 1768.16 Q2677.66 1770.06 2675.55 1772.65 Q2674.97 1773.32 2671.87 1776.54 Q2668.77 1779.74 2663.12 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2699.25 1757.95 Q2695.64 1757.95 2693.81 1761.52 Q2692.01 1765.06 2692.01 1772.19 Q2692.01 1779.3 2693.81 1782.86 Q2695.64 1786.4 2699.25 1786.4 Q2702.89 1786.4 2704.69 1782.86 Q2706.52 1779.3 2706.52 1772.19 Q2706.52 1765.06 2704.69 1761.52 Q2702.89 1757.95 2699.25 1757.95 M2699.25 1754.25 Q2705.06 1754.25 2708.12 1758.86 Q2711.2 1763.44 2711.2 1772.19 Q2711.2 1780.92 2708.12 1785.52 Q2705.06 1790.11 2699.25 1790.11 Q2693.44 1790.11 2690.36 1785.52 Q2687.31 1780.92 2687.31 1772.19 Q2687.31 1763.44 2690.36 1758.86 Q2693.44 1754.25 2699.25 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2729.42 1757.95 Q2725.8 1757.95 2723.98 1761.52 Q2722.17 1765.06 2722.17 1772.19 Q2722.17 1779.3 2723.98 1782.86 Q2725.8 1786.4 2729.42 1786.4 Q2733.05 1786.4 2734.86 1782.86 Q2736.68 1779.3 2736.68 1772.19 Q2736.68 1765.06 2734.86 1761.52 Q2733.05 1757.95 2729.42 1757.95 M2729.42 1754.25 Q2735.23 1754.25 2738.28 1758.86 Q2741.36 1763.44 2741.36 1772.19 Q2741.36 1780.92 2738.28 1785.52 Q2735.23 1790.11 2729.42 1790.11 Q2723.61 1790.11 2720.53 1785.52 Q2717.47 1780.92 2717.47 1772.19 Q2717.47 1763.44 2720.53 1758.86 Q2723.61 1754.25 2729.42 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2759.58 1757.95 Q2755.97 1757.95 2754.14 1761.52 Q2752.33 1765.06 2752.33 1772.19 Q2752.33 1779.3 2754.14 1782.86 Q2755.97 1786.4 2759.58 1786.4 Q2763.21 1786.4 2765.02 1782.86 Q2766.85 1779.3 2766.85 1772.19 Q2766.85 1765.06 2765.02 1761.52 Q2763.21 1757.95 2759.58 1757.95 M2759.58 1754.25 Q2765.39 1754.25 2768.44 1758.86 Q2771.52 1763.44 2771.52 1772.19 Q2771.52 1780.92 2768.44 1785.52 Q2765.39 1790.11 2759.58 1790.11 Q2753.77 1790.11 2750.69 1785.52 Q2747.63 1780.92 2747.63 1772.19 Q2747.63 1763.44 2750.69 1758.86 Q2753.77 1754.25 2759.58 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,1672.06 2752.76,1672.06 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,1567.36 2752.76,1567.36 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,1462.66 2752.76,1462.66 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,1357.96 2752.76,1357.96 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,1253.26 2752.76,1253.26 "/>
+<polyline clip-path="url(#clip635)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,1148.56 2752.76,1148.56 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1723.83 1559.76,1087.05 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1672.06 1578.66,1672.06 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1567.36 1578.66,1567.36 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1462.66 1578.66,1462.66 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1357.96 1578.66,1357.96 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1253.26 1578.66,1253.26 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,1148.56 1578.66,1148.56 "/>
+<path clip-path="url(#clip630)" d="M1457.05 1672.51 L1486.72 1672.51 L1486.72 1676.44 L1457.05 1676.44 L1457.05 1672.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1496.86 1654.78 L1515.22 1654.78 L1515.22 1658.71 L1501.14 1658.71 L1501.14 1667.18 Q1502.16 1666.84 1503.18 1666.68 Q1504.2 1666.49 1505.22 1666.49 Q1511.01 1666.49 1514.39 1669.66 Q1517.76 1672.83 1517.76 1678.25 Q1517.76 1683.83 1514.29 1686.93 Q1510.82 1690.01 1504.5 1690.01 Q1502.33 1690.01 1500.06 1689.64 Q1497.81 1689.27 1495.4 1688.53 L1495.4 1683.83 Q1497.49 1684.96 1499.71 1685.52 Q1501.93 1686.07 1504.41 1686.07 Q1508.41 1686.07 1510.75 1683.97 Q1513.09 1681.86 1513.09 1678.25 Q1513.09 1674.64 1510.75 1672.53 Q1508.41 1670.43 1504.41 1670.43 Q1502.53 1670.43 1500.66 1670.84 Q1498.81 1671.26 1496.86 1672.14 L1496.86 1654.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1455.57 1567.81 L1485.24 1567.81 L1485.24 1571.74 L1455.57 1571.74 L1455.57 1567.81 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1508.18 1554.15 L1496.38 1572.6 L1508.18 1572.6 L1508.18 1554.15 M1506.95 1550.08 L1512.83 1550.08 L1512.83 1572.6 L1517.76 1572.6 L1517.76 1576.49 L1512.83 1576.49 L1512.83 1584.64 L1508.18 1584.64 L1508.18 1576.49 L1492.58 1576.49 L1492.58 1571.98 L1506.95 1550.08 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1456.7 1463.11 L1486.38 1463.11 L1486.38 1467.04 L1456.7 1467.04 L1456.7 1463.11 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1510.64 1461.3 Q1513.99 1462.02 1515.87 1464.29 Q1517.76 1466.56 1517.76 1469.89 Q1517.76 1475.01 1514.25 1477.81 Q1510.73 1480.61 1504.25 1480.61 Q1502.07 1480.61 1499.76 1480.17 Q1497.46 1479.75 1495.01 1478.9 L1495.01 1474.38 Q1496.95 1475.52 1499.27 1476.09 Q1501.58 1476.67 1504.11 1476.67 Q1508.51 1476.67 1510.8 1474.94 Q1513.11 1473.2 1513.11 1469.89 Q1513.11 1466.84 1510.96 1465.12 Q1508.83 1463.39 1505.01 1463.39 L1500.98 1463.39 L1500.98 1459.54 L1505.2 1459.54 Q1508.64 1459.54 1510.47 1458.18 Q1512.3 1456.79 1512.3 1454.2 Q1512.3 1451.53 1510.4 1450.12 Q1508.53 1448.69 1505.01 1448.69 Q1503.09 1448.69 1500.89 1449.1 Q1498.69 1449.52 1496.05 1450.4 L1496.05 1446.23 Q1498.71 1445.49 1501.03 1445.12 Q1503.37 1444.75 1505.43 1444.75 Q1510.75 1444.75 1513.85 1447.18 Q1516.95 1449.59 1516.95 1453.71 Q1516.95 1456.58 1515.31 1458.57 Q1513.67 1460.54 1510.64 1461.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1457.65 1358.41 L1487.33 1358.41 L1487.33 1362.34 L1457.65 1362.34 L1457.65 1358.41 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1501.45 1371.3 L1517.76 1371.3 L1517.76 1375.24 L1495.82 1375.24 L1495.82 1371.3 Q1498.48 1368.55 1503.07 1363.92 Q1507.67 1359.27 1508.85 1357.92 Q1511.1 1355.4 1511.98 1353.66 Q1512.88 1351.9 1512.88 1350.21 Q1512.88 1347.46 1510.94 1345.72 Q1509.01 1343.99 1505.91 1343.99 Q1503.71 1343.99 1501.26 1344.75 Q1498.83 1345.52 1496.05 1347.07 L1496.05 1342.34 Q1498.88 1341.21 1501.33 1340.63 Q1503.78 1340.05 1505.82 1340.05 Q1511.19 1340.05 1514.39 1342.74 Q1517.58 1345.42 1517.58 1349.91 Q1517.58 1352.04 1516.77 1353.96 Q1515.98 1355.86 1513.88 1358.46 Q1513.3 1359.13 1510.2 1362.34 Q1507.09 1365.54 1501.45 1371.3 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1457.28 1253.71 L1486.95 1253.71 L1486.95 1257.64 L1457.28 1257.64 L1457.28 1253.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1497.86 1266.6 L1505.5 1266.6 L1505.5 1240.24 L1497.19 1241.9 L1497.19 1237.64 L1505.45 1235.98 L1510.13 1235.98 L1510.13 1266.6 L1517.76 1266.6 L1517.76 1270.54 L1497.86 1270.54 L1497.86 1266.6 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1505.82 1134.36 Q1502.21 1134.36 1500.38 1137.92 Q1498.58 1141.46 1498.58 1148.59 Q1498.58 1155.7 1500.38 1159.26 Q1502.21 1162.81 1505.82 1162.81 Q1509.45 1162.81 1511.26 1159.26 Q1513.09 1155.7 1513.09 1148.59 Q1513.09 1141.46 1511.26 1137.92 Q1509.45 1134.36 1505.82 1134.36 M1505.82 1130.65 Q1511.63 1130.65 1514.69 1135.26 Q1517.76 1139.84 1517.76 1148.59 Q1517.76 1157.32 1514.69 1161.93 Q1511.63 1166.51 1505.82 1166.51 Q1500.01 1166.51 1496.93 1161.93 Q1493.88 1157.32 1493.88 1148.59 Q1493.88 1139.84 1496.93 1135.26 Q1500.01 1130.65 1505.82 1130.65 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1820.06 945.429 L1854.81 945.429 L1854.81 952.316 L1828.24 952.316 L1828.24 970.14 L1852.22 970.14 L1852.22 977.026 L1828.24 977.026 L1828.24 1005.91 L1820.06 1005.91 L1820.06 945.429 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1867.45 942.877 L1874.9 942.877 L1874.9 1005.91 L1867.45 1005.91 L1867.45 942.877 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1890.5 960.539 L1897.95 960.539 L1897.95 1005.91 L1890.5 1005.91 L1890.5 960.539 M1890.5 942.877 L1897.95 942.877 L1897.95 952.316 L1890.5 952.316 L1890.5 942.877 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1943.41 982.698 Q1943.41 974.596 1940.04 970.14 Q1936.72 965.684 1930.69 965.684 Q1924.69 965.684 1921.33 970.14 Q1918.01 974.596 1918.01 982.698 Q1918.01 990.759 1921.33 995.215 Q1924.69 999.671 1930.69 999.671 Q1936.72 999.671 1940.04 995.215 Q1943.41 990.759 1943.41 982.698 M1950.86 1000.28 Q1950.86 1011.86 1945.71 1017.49 Q1940.57 1023.17 1929.96 1023.17 Q1926.03 1023.17 1922.54 1022.56 Q1919.06 1021.99 1915.78 1020.78 L1915.78 1013.53 Q1919.06 1015.31 1922.26 1016.16 Q1925.46 1017.01 1928.78 1017.01 Q1936.11 1017.01 1939.76 1013.16 Q1943.41 1009.35 1943.41 1001.62 L1943.41 997.929 Q1941.1 1001.94 1937.49 1003.92 Q1933.89 1005.91 1928.86 1005.91 Q1920.52 1005.91 1915.41 999.549 Q1910.31 993.189 1910.31 982.698 Q1910.31 972.165 1915.41 965.805 Q1920.52 959.445 1928.86 959.445 Q1933.89 959.445 1937.49 961.43 Q1941.1 963.415 1943.41 967.426 L1943.41 960.539 L1950.86 960.539 L1950.86 1000.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2003.93 978.525 L2003.93 1005.91 L1996.47 1005.91 L1996.47 978.768 Q1996.47 972.327 1993.96 969.127 Q1991.45 965.927 1986.43 965.927 Q1980.39 965.927 1976.91 969.775 Q1973.42 973.624 1973.42 980.267 L1973.42 1005.91 L1965.93 1005.91 L1965.93 942.877 L1973.42 942.877 L1973.42 967.588 Q1976.1 963.496 1979.7 961.471 Q1983.35 959.445 1988.09 959.445 Q1995.91 959.445 1999.92 964.307 Q2003.93 969.127 2003.93 978.525 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2026.17 947.657 L2026.17 960.539 L2041.52 960.539 L2041.52 966.332 L2026.17 966.332 L2026.17 990.961 Q2026.17 996.511 2027.66 998.091 Q2029.2 999.671 2033.86 999.671 L2041.52 999.671 L2041.52 1005.91 L2033.86 1005.91 Q2025.23 1005.91 2021.95 1002.71 Q2018.67 999.468 2018.67 990.961 L2018.67 966.332 L2013.2 966.332 L2013.2 960.539 L2018.67 960.539 L2018.67 947.657 L2026.17 947.657 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2086.2 952.154 L2086.2 974.879 L2096.49 974.879 Q2102.2 974.879 2105.32 971.922 Q2108.44 968.965 2108.44 963.496 Q2108.44 958.068 2105.32 955.111 Q2102.2 952.154 2096.49 952.154 L2086.2 952.154 M2078.02 945.429 L2096.49 945.429 Q2106.66 945.429 2111.84 950.047 Q2117.07 954.625 2117.07 963.496 Q2117.07 972.449 2111.84 977.026 Q2106.66 981.604 2096.49 981.604 L2086.2 981.604 L2086.2 1005.91 L2078.02 1005.91 L2078.02 945.429 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2144.61 983.103 Q2135.58 983.103 2132.1 985.169 Q2128.61 987.235 2128.61 992.217 Q2128.61 996.187 2131.21 998.537 Q2133.84 1000.85 2138.34 1000.85 Q2144.53 1000.85 2148.26 996.471 Q2152.03 992.055 2152.03 984.764 L2152.03 983.103 L2144.61 983.103 M2159.48 980.024 L2159.48 1005.91 L2152.03 1005.91 L2152.03 999.023 Q2149.48 1003.15 2145.67 1005.14 Q2141.86 1007.08 2136.35 1007.08 Q2129.38 1007.08 2125.25 1003.2 Q2121.16 999.266 2121.16 992.703 Q2121.16 985.047 2126.26 981.158 Q2131.41 977.269 2141.58 977.269 L2152.03 977.269 L2152.03 976.54 Q2152.03 971.396 2148.62 968.601 Q2145.26 965.765 2139.15 965.765 Q2135.26 965.765 2131.57 966.697 Q2127.88 967.628 2124.48 969.492 L2124.48 962.605 Q2128.57 961.025 2132.42 960.256 Q2136.27 959.445 2139.92 959.445 Q2149.76 959.445 2154.62 964.55 Q2159.48 969.654 2159.48 980.024 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2182.21 947.657 L2182.21 960.539 L2197.56 960.539 L2197.56 966.332 L2182.21 966.332 L2182.21 990.961 Q2182.21 996.511 2183.71 998.091 Q2185.24 999.671 2189.9 999.671 L2197.56 999.671 L2197.56 1005.91 L2189.9 1005.91 Q2181.27 1005.91 2177.99 1002.71 Q2174.71 999.468 2174.71 990.961 L2174.71 966.332 L2169.24 966.332 L2169.24 960.539 L2174.71 960.539 L2174.71 947.657 L2182.21 947.657 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2245.08 978.525 L2245.08 1005.91 L2237.62 1005.91 L2237.62 978.768 Q2237.62 972.327 2235.11 969.127 Q2232.6 965.927 2227.58 965.927 Q2221.54 965.927 2218.06 969.775 Q2214.57 973.624 2214.57 980.267 L2214.57 1005.91 L2207.08 1005.91 L2207.08 942.877 L2214.57 942.877 L2214.57 967.588 Q2217.25 963.496 2220.85 961.471 Q2224.5 959.445 2229.24 959.445 Q2237.06 959.445 2241.07 964.307 Q2245.08 969.127 2245.08 978.525 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2304.22 942.958 Q2298.79 952.275 2296.16 961.39 Q2293.53 970.504 2293.53 979.862 Q2293.53 989.22 2296.16 998.415 Q2298.83 1007.57 2304.22 1016.85 L2297.74 1016.85 Q2291.66 1007.33 2288.62 998.132 Q2285.63 988.936 2285.63 979.862 Q2285.63 970.829 2288.62 961.673 Q2291.62 952.518 2297.74 942.958 L2304.22 942.958 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2348.54 967.426 L2348.54 942.877 L2355.99 942.877 L2355.99 1005.91 L2348.54 1005.91 L2348.54 999.104 Q2346.19 1003.15 2342.58 1005.14 Q2339.02 1007.08 2333.99 1007.08 Q2325.77 1007.08 2320.59 1000.52 Q2315.44 993.959 2315.44 983.265 Q2315.44 972.57 2320.59 966.008 Q2325.77 959.445 2333.99 959.445 Q2339.02 959.445 2342.58 961.43 Q2346.19 963.375 2348.54 967.426 M2323.14 983.265 Q2323.14 991.488 2326.5 996.187 Q2329.9 1000.85 2335.82 1000.85 Q2341.73 1000.85 2345.13 996.187 Q2348.54 991.488 2348.54 983.265 Q2348.54 975.041 2345.13 970.383 Q2341.73 965.684 2335.82 965.684 Q2329.9 965.684 2326.5 970.383 Q2323.14 975.041 2323.14 983.265 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2410.15 981.361 L2410.15 985.007 L2375.88 985.007 Q2376.37 992.703 2380.5 996.754 Q2384.67 1000.76 2392.08 1000.76 Q2396.38 1000.76 2400.39 999.711 Q2404.44 998.658 2408.41 996.552 L2408.41 1003.6 Q2404.4 1005.3 2400.19 1006.19 Q2395.97 1007.08 2391.64 1007.08 Q2380.78 1007.08 2374.42 1000.76 Q2368.1 994.445 2368.1 983.67 Q2368.1 972.53 2374.1 966.008 Q2380.13 959.445 2390.34 959.445 Q2399.5 959.445 2404.8 965.36 Q2410.15 971.234 2410.15 981.361 M2402.7 979.173 Q2402.62 973.057 2399.25 969.411 Q2395.93 965.765 2390.42 965.765 Q2384.18 965.765 2380.42 969.289 Q2376.69 972.813 2376.12 979.214 L2402.7 979.173 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2452.24 982.698 Q2452.24 974.596 2448.88 970.14 Q2445.56 965.684 2439.52 965.684 Q2433.52 965.684 2430.16 970.14 Q2426.84 974.596 2426.84 982.698 Q2426.84 990.759 2430.16 995.215 Q2433.52 999.671 2439.52 999.671 Q2445.56 999.671 2448.88 995.215 Q2452.24 990.759 2452.24 982.698 M2459.69 1000.28 Q2459.69 1011.86 2454.55 1017.49 Q2449.4 1023.17 2438.79 1023.17 Q2434.86 1023.17 2431.38 1022.56 Q2427.89 1021.99 2424.61 1020.78 L2424.61 1013.53 Q2427.89 1015.31 2431.09 1016.16 Q2434.29 1017.01 2437.62 1017.01 Q2444.95 1017.01 2448.59 1013.16 Q2452.24 1009.35 2452.24 1001.62 L2452.24 997.929 Q2449.93 1001.94 2446.33 1003.92 Q2442.72 1005.91 2437.7 1005.91 Q2429.35 1005.91 2424.25 999.549 Q2419.14 993.189 2419.14 982.698 Q2419.14 972.165 2424.25 965.805 Q2429.35 959.445 2437.7 959.445 Q2442.72 959.445 2446.33 961.43 Q2449.93 963.415 2452.24 967.426 L2452.24 960.539 L2459.69 960.539 L2459.69 1000.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2473.87 942.958 L2480.35 942.958 Q2486.43 952.518 2489.43 961.673 Q2492.47 970.829 2492.47 979.862 Q2492.47 988.936 2489.43 998.132 Q2486.43 1007.33 2480.35 1016.85 L2473.87 1016.85 Q2479.26 1007.57 2481.89 998.415 Q2484.57 989.22 2484.57 979.862 Q2484.57 970.504 2481.89 961.39 Q2479.26 952.275 2473.87 942.958 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip635)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="1593.53,1253.26 1595.77,1253.23 1598.01,1253.16 1600.25,1253.07 1602.5,1252.93 1604.74,1252.75 1606.98,1252.53 1609.22,1252.26 1611.46,1251.93 1613.71,1251.55 1615.95,1251.12 1618.19,1250.61 1620.43,1250.04 1622.67,1249.4 1624.92,1248.67 1627.16,1247.86 1629.4,1246.97 1631.64,1245.97 1633.88,1244.88 1636.13,1243.67 1638.37,1242.36 1640.61,1240.92 1642.85,1239.36 1645.09,1237.67 1647.34,1235.84 1649.58,1233.86 1651.82,1231.74 1654.06,1229.47 1656.3,1227.05 1658.55,1224.46 1660.79,1221.72 1663.03,1218.82 1665.27,1215.77 1667.51,1212.56 1669.76,1209.2 1672,1205.69 1674.24,1202.05 1676.48,1198.27 1678.72,1194.39 1680.97,1190.39 1683.21,1186.31 1685.45,1182.16 1687.69,1177.95 1689.93,1173.7 1692.17,1169.44 1694.42,1165.17 1696.66,1160.93 1698.9,1156.74 1701.14,1152.61 1703.38,1148.56 1705.63,1144.62 1707.87,1140.79 1710.11,1137.11 1712.35,1133.59 1714.59,1130.23 1716.84,1127.05 1719.08,1124.07 1721.32,1121.29 1723.56,1118.72 1725.8,1116.37 1728.05,1114.24 1730.29,1112.33 1732.53,1110.65 1734.77,1109.19 1737.01,1107.96 1739.26,1106.95 1741.5,1106.16 1743.74,1105.59 1745.98,1105.23 1748.22,1105.07 1750.47,1105.11 1752.71,1105.35 1754.95,1105.77 1757.19,1106.37 1759.43,1107.15 1761.68,1108.09 1763.92,1109.19 1766.16,1110.44 1768.4,1111.84 1770.64,1113.37 1772.89,1115.03 1775.13,1116.82 1777.37,1118.71 1779.61,1120.72 1781.85,1122.82 1784.1,1125.02 1786.34,1127.3 1788.58,1129.66 1790.82,1132.09 1793.06,1134.59 1795.31,1137.15 1797.55,1139.76 1799.79,1142.42 1802.03,1145.11 1804.27,1147.83 1806.51,1150.58 1808.76,1153.35 1811,1156.13 1813.24,1158.91 1815.48,1161.69 1817.72,1164.46 1819.97,1167.21 1822.21,1169.94 1824.45,1172.63 1826.69,1175.28 1828.93,1177.87 1831.18,1180.41 1833.42,1182.89 1835.66,1185.28 1837.9,1187.59 1840.14,1189.8 1842.39,1191.91 1844.63,1193.9 1846.87,1195.76 1849.11,1197.49 1851.35,1199.07 1853.6,1200.49 1855.84,1201.74 1858.08,1202.8 1860.32,1203.68 1862.56,1204.35 1864.81,1204.82 1867.05,1205.05 1869.29,1205.06 1871.53,1204.84 1873.77,1204.36 1876.02,1203.64 1878.26,1202.66 1880.5,1201.43 1882.74,1199.94 1884.98,1198.2 1887.23,1196.22 1889.47,1193.99 1891.71,1191.54 1893.95,1188.87 1896.19,1186 1898.44,1182.95 1900.68,1179.75 1902.92,1176.4 1905.16,1172.95 1907.4,1169.42 1909.64,1165.84 1911.89,1162.23 1914.13,1158.64 1916.37,1155.08 1918.61,1151.6 1920.85,1148.22 1923.1,1144.97 1925.34,1141.87 1927.58,1138.96 1929.82,1136.26 1932.06,1133.77 1934.31,1131.53 1936.55,1129.55 1938.79,1127.83 1941.03,1126.39 1943.27,1125.24 1945.52,1124.37 1947.76,1123.78 1950,1123.49 1952.24,1123.47 1954.48,1123.74 1956.73,1124.27 1958.97,1125.07 1961.21,1126.11 1963.45,1127.4 1965.69,1128.91 1967.94,1130.63 1970.18,1132.55 1972.42,1134.65 1974.66,1136.91 1976.9,1139.33 1979.15,1141.87 1981.39,1144.53 1983.63,1147.29 1985.87,1150.12 1988.11,1153.02 1990.36,1155.95 1992.6,1158.9 1994.84,1161.86 1997.08,1164.8 1999.32,1167.7 2001.57,1170.54 2003.81,1173.3 2006.05,1175.97 2008.29,1178.53 2010.53,1180.94 2012.78,1183.2 2015.02,1185.29 2017.26,1187.18 2019.5,1188.87 2021.74,1190.33 2023.98,1191.55 2026.23,1192.51 2028.47,1193.2 2030.71,1193.62 2032.95,1193.75 2035.19,1193.59 2037.44,1193.13 2039.68,1192.37 2041.92,1191.33 2044.16,1190 2046.4,1188.39 2048.65,1186.52 2050.89,1184.41 2053.13,1182.07 2055.37,1179.53 2057.61,1176.83 2059.86,1173.98 2062.1,1171.02 2064.34,1167.99 2066.58,1164.92 2068.82,1161.85 2071.07,1158.82 2073.31,1155.85 2075.55,1152.99 2077.79,1150.28 2080.03,1147.73 2082.28,1145.38 2084.52,1143.26 2086.76,1141.39 2089,1139.78 2091.24,1138.45 2093.49,1137.42 2095.73,1136.69 2097.97,1136.26 2100.21,1136.14 2102.45,1136.32 2104.7,1136.8 2106.94,1137.57 2109.18,1138.61 2111.42,1139.91 2113.66,1141.46 2115.91,1143.23 2118.15,1145.21 2120.39,1147.38 2122.63,1149.7 2124.87,1152.16 2127.11,1154.73 2129.36,1157.39 2131.6,1160.1 2133.84,1162.84 2136.08,1165.59 2138.32,1168.3 2140.57,1170.97 2142.81,1173.55 2145.05,1176.03 2147.29,1178.37 2149.53,1180.55 2151.78,1182.55 2154.02,1184.34 2156.26,1185.91 2158.5,1187.23 2160.74,1188.29 2162.99,1189.08 2165.23,1189.58 2167.47,1189.78 2169.71,1189.69 2171.95,1189.31 2174.2,1188.63 2176.44,1187.67 2178.68,1186.44 2180.92,1184.95 2183.16,1183.23 2185.41,1181.29 2187.65,1179.18 2189.89,1176.92 2192.13,1174.53 2194.37,1172.07 2196.62,1169.56 2198.86,1167.04 2201.1,1164.56 2203.34,1162.14 2205.58,1159.82 2207.83,1157.65 2210.07,1155.65 2212.31,1153.84 2214.55,1152.26 2216.79,1150.93 2219.04,1149.86 2221.28,1149.06 2223.52,1148.56 2225.76,1148.34 2228,1148.42 2230.25,1148.79 2232.49,1149.44 2234.73,1150.36 2236.97,1151.54 2239.21,1152.96 2241.45,1154.6 2243.7,1156.43 2245.94,1158.44 2248.18,1160.59 2250.42,1162.86 2252.66,1165.22 2254.91,1167.63 2257.15,1170.07 2259.39,1172.51 2261.63,1174.9 2263.87,1177.24 2266.12,1179.47 2268.36,1181.58 2270.6,1183.54 2272.84,1185.32 2275.08,1186.89 2277.33,1188.25 2279.57,1189.37 2281.81,1190.24 2284.05,1190.84 2286.29,1191.17 2288.54,1191.24 2290.78,1191.03 2293.02,1190.56 2295.26,1189.84 2297.5,1188.88 2299.75,1187.7 2301.99,1186.32 2304.23,1184.78 2306.47,1183.09 2308.71,1181.29 2310.96,1179.42 2313.2,1177.51 2315.44,1175.6 2317.68,1173.72 2319.92,1171.9 2322.17,1170.19 2324.41,1168.61 2326.65,1167.19 2328.89,1165.96 2331.13,1164.94 2333.38,1164.14 2335.62,1163.59 2337.86,1163.3 2340.1,1163.26 2342.34,1163.49 2344.58,1163.98 2346.83,1164.72 2349.07,1165.7 2351.31,1166.9 2353.55,1168.32 2355.79,1169.93 2358.04,1171.7 2360.28,1173.62 2362.52,1175.64 2364.76,1177.75 2367,1179.92 2369.25,1182.11 2371.49,1184.29 2373.73,1186.43 2375.97,1188.51 2378.21,1190.49 2380.46,1192.35 2382.7,1194.06 2384.94,1195.61 2387.18,1196.97 2389.42,1198.12 2391.67,1199.06 2393.91,1199.77 2396.15,1200.25 2398.39,1200.51 2400.63,1200.54 2402.88,1200.35 2405.12,1199.95 2407.36,1199.37 2409.6,1198.62 2411.84,1197.72 2414.09,1196.7 2416.33,1195.58 2418.57,1194.41 2420.81,1193.2 2423.05,1192 2425.3,1190.82 2427.54,1189.71 2429.78,1188.7 2432.02,1187.8 2434.26,1187.06 2436.51,1186.48 2438.75,1186.09 2440.99,1185.91 2443.23,1185.94 2445.47,1186.2 2447.72,1186.69 2449.96,1187.41 2452.2,1188.35 2454.44,1189.51 2456.68,1190.87 2458.92,1192.43 2461.17,1194.15 2463.41,1196.03 2465.65,1198.03 2467.89,1200.13 2470.13,1202.31 2472.38,1204.54 2474.62,1206.79 2476.86,1209.04 2479.1,1211.24 2481.34,1213.39 2483.59,1215.46 2485.83,1217.41 2488.07,1219.24 2490.31,1220.91 2492.55,1222.43 2494.8,1223.78 2497.04,1224.95 2499.28,1225.93 2501.52,1226.73 2503.76,1227.35 2506.01,1227.8 2508.25,1228.09 2510.49,1228.24 2512.73,1228.26 2514.97,1228.18 2517.22,1228.02 2519.46,1227.82 2521.7,1227.59 2523.94,1227.36 2526.18,1227.18 2528.43,1227.06 2530.67,1227.04 2532.91,1227.14 2535.15,1227.4 2537.39,1227.83 2539.64,1228.45 2541.88,1229.29 2544.12,1230.36 2546.36,1231.66 2548.6,1233.22 2550.85,1235.02 2553.09,1237.07 2555.33,1239.37 2557.57,1241.9 2559.81,1244.65 2562.05,1247.61 2564.3,1250.76 2566.54,1254.08 2568.78,1257.54 2571.02,1261.12 2573.26,1264.78 2575.51,1268.51 2577.75,1272.26 2579.99,1276.01 2582.23,1279.74 2584.47,1283.4 2586.72,1286.99 2588.96,1290.46 2591.2,1293.8 2593.44,1296.99 2595.68,1300.01 2597.93,1302.86 2600.17,1305.51 2602.41,1307.98 2604.65,1310.26 2606.89,1312.36 2609.14,1314.3 2611.38,1316.09 2613.62,1317.75 2615.86,1319.32 2618.1,1320.84 2620.35,1322.34 2622.59,1323.88 2624.83,1325.49 2627.07,1327.24 2629.31,1329.19 2631.56,1331.4 2633.8,1333.93 2636.04,1336.84 2638.28,1340.21 2640.52,1344.1 2642.77,1348.56 2645.01,1353.66 2647.25,1359.46 2649.49,1366.01 2651.73,1373.35 2653.98,1381.53 2656.22,1390.56 2658.46,1400.49 2660.7,1411.31 2662.94,1423.03 2665.19,1435.64 2667.43,1449.12 2669.67,1463.43 2671.91,1478.5 2674.15,1494.29 2676.39,1510.69 2678.64,1527.61 2680.88,1544.92 2683.12,1562.46 2685.36,1580.08 2687.6,1597.57 2689.85,1614.72 2692.09,1631.27 2694.33,1646.97 2696.57,1661.51 2698.81,1674.57 2701.06,1685.83 2703.3,1694.93 2705.54,1701.51 2707.78,1705.25 2710.02,1705.81 2712.27,1702.94 2714.51,1696.41 2716.75,1686.11 2718.99,1672.06 "/>
+<path clip-path="url(#clip630)" d="M159.765 2657.17 L1352.76 2657.17 L1352.76 2020.38 L159.765 2020.38  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip636">
+    <rect x="159" y="2020" width="1194" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="193.529,2657.17 193.529,2020.38 "/>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="473.774,2657.17 473.774,2020.38 "/>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="754.018,2657.17 754.018,2020.38 "/>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1034.26,2657.17 1034.26,2020.38 "/>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1314.51,2657.17 1314.51,2020.38 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,2657.17 1352.76,2657.17 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="193.529,2657.17 193.529,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="473.774,2657.17 473.774,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="754.018,2657.17 754.018,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1034.26,2657.17 1034.26,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1314.51,2657.17 1314.51,2638.27 "/>
+<path clip-path="url(#clip630)" d="M193.529 2691.29 Q189.918 2691.29 188.089 2694.85 Q186.283 2698.39 186.283 2705.52 Q186.283 2712.63 188.089 2716.19 Q189.918 2719.74 193.529 2719.74 Q197.163 2719.74 198.969 2716.19 Q200.797 2712.63 200.797 2705.52 Q200.797 2698.39 198.969 2694.85 Q197.163 2691.29 193.529 2691.29 M193.529 2687.58 Q199.339 2687.58 202.394 2692.19 Q205.473 2696.77 205.473 2705.52 Q205.473 2714.25 202.394 2718.86 Q199.339 2723.44 193.529 2723.44 Q187.719 2723.44 184.64 2718.86 Q181.584 2714.25 181.584 2705.52 Q181.584 2696.77 184.64 2692.19 Q187.719 2687.58 193.529 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M433.392 2688.21 L451.748 2688.21 L451.748 2692.14 L437.674 2692.14 L437.674 2700.62 Q438.693 2700.27 439.711 2700.11 Q440.73 2699.92 441.748 2699.92 Q447.535 2699.92 450.915 2703.09 Q454.295 2706.26 454.295 2711.68 Q454.295 2717.26 450.822 2720.36 Q447.35 2723.44 441.031 2723.44 Q438.855 2723.44 436.586 2723.07 Q434.341 2722.7 431.934 2721.96 L431.934 2717.26 Q434.017 2718.39 436.239 2718.95 Q438.461 2719.5 440.938 2719.5 Q444.943 2719.5 447.281 2717.4 Q449.619 2715.29 449.619 2711.68 Q449.619 2708.07 447.281 2705.96 Q444.943 2703.86 440.938 2703.86 Q439.063 2703.86 437.188 2704.27 Q435.336 2704.69 433.392 2705.57 L433.392 2688.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M473.507 2691.29 Q469.896 2691.29 468.068 2694.85 Q466.262 2698.39 466.262 2705.52 Q466.262 2712.63 468.068 2716.19 Q469.896 2719.74 473.507 2719.74 Q477.142 2719.74 478.947 2716.19 Q480.776 2712.63 480.776 2705.52 Q480.776 2698.39 478.947 2694.85 Q477.142 2691.29 473.507 2691.29 M473.507 2687.58 Q479.318 2687.58 482.373 2692.19 Q485.452 2696.77 485.452 2705.52 Q485.452 2714.25 482.373 2718.86 Q479.318 2723.44 473.507 2723.44 Q467.697 2723.44 464.619 2718.86 Q461.563 2714.25 461.563 2705.52 Q461.563 2696.77 464.619 2692.19 Q467.697 2687.58 473.507 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M503.669 2691.29 Q500.058 2691.29 498.23 2694.85 Q496.424 2698.39 496.424 2705.52 Q496.424 2712.63 498.23 2716.19 Q500.058 2719.74 503.669 2719.74 Q507.304 2719.74 509.109 2716.19 Q510.938 2712.63 510.938 2705.52 Q510.938 2698.39 509.109 2694.85 Q507.304 2691.29 503.669 2691.29 M503.669 2687.58 Q509.479 2687.58 512.535 2692.19 Q515.614 2696.77 515.614 2705.52 Q515.614 2714.25 512.535 2718.86 Q509.479 2723.44 503.669 2723.44 Q497.859 2723.44 494.78 2718.86 Q491.725 2714.25 491.725 2705.52 Q491.725 2696.77 494.78 2692.19 Q497.859 2687.58 503.669 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M698.544 2718.83 L706.183 2718.83 L706.183 2692.47 L697.873 2694.13 L697.873 2689.87 L706.137 2688.21 L710.813 2688.21 L710.813 2718.83 L718.452 2718.83 L718.452 2722.77 L698.544 2722.77 L698.544 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M737.896 2691.29 Q734.285 2691.29 732.456 2694.85 Q730.651 2698.39 730.651 2705.52 Q730.651 2712.63 732.456 2716.19 Q734.285 2719.74 737.896 2719.74 Q741.53 2719.74 743.336 2716.19 Q745.164 2712.63 745.164 2705.52 Q745.164 2698.39 743.336 2694.85 Q741.53 2691.29 737.896 2691.29 M737.896 2687.58 Q743.706 2687.58 746.762 2692.19 Q749.84 2696.77 749.84 2705.52 Q749.84 2714.25 746.762 2718.86 Q743.706 2723.44 737.896 2723.44 Q732.086 2723.44 729.007 2718.86 Q725.951 2714.25 725.951 2705.52 Q725.951 2696.77 729.007 2692.19 Q732.086 2687.58 737.896 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M768.058 2691.29 Q764.447 2691.29 762.618 2694.85 Q760.812 2698.39 760.812 2705.52 Q760.812 2712.63 762.618 2716.19 Q764.447 2719.74 768.058 2719.74 Q771.692 2719.74 773.498 2716.19 Q775.326 2712.63 775.326 2705.52 Q775.326 2698.39 773.498 2694.85 Q771.692 2691.29 768.058 2691.29 M768.058 2687.58 Q773.868 2687.58 776.923 2692.19 Q780.002 2696.77 780.002 2705.52 Q780.002 2714.25 776.923 2718.86 Q773.868 2723.44 768.058 2723.44 Q762.248 2723.44 759.169 2718.86 Q756.113 2714.25 756.113 2705.52 Q756.113 2696.77 759.169 2692.19 Q762.248 2687.58 768.058 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M798.22 2691.29 Q794.609 2691.29 792.78 2694.85 Q790.974 2698.39 790.974 2705.52 Q790.974 2712.63 792.78 2716.19 Q794.609 2719.74 798.22 2719.74 Q801.854 2719.74 803.659 2716.19 Q805.488 2712.63 805.488 2705.52 Q805.488 2698.39 803.659 2694.85 Q801.854 2691.29 798.22 2691.29 M798.22 2687.58 Q804.03 2687.58 807.085 2692.19 Q810.164 2696.77 810.164 2705.52 Q810.164 2714.25 807.085 2718.86 Q804.03 2723.44 798.22 2723.44 Q792.409 2723.44 789.331 2718.86 Q786.275 2714.25 786.275 2705.52 Q786.275 2696.77 789.331 2692.19 Q792.409 2687.58 798.22 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M978.789 2718.83 L986.428 2718.83 L986.428 2692.47 L978.118 2694.13 L978.118 2689.87 L986.382 2688.21 L991.057 2688.21 L991.057 2718.83 L998.696 2718.83 L998.696 2722.77 L978.789 2722.77 L978.789 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1008.19 2688.21 L1026.54 2688.21 L1026.54 2692.14 L1012.47 2692.14 L1012.47 2700.62 Q1013.49 2700.27 1014.51 2700.11 Q1015.52 2699.92 1016.54 2699.92 Q1022.33 2699.92 1025.71 2703.09 Q1029.09 2706.26 1029.09 2711.68 Q1029.09 2717.26 1025.62 2720.36 Q1022.15 2723.44 1015.83 2723.44 Q1013.65 2723.44 1011.38 2723.07 Q1009.14 2722.7 1006.73 2721.96 L1006.73 2717.26 Q1008.81 2718.39 1011.03 2718.95 Q1013.26 2719.5 1015.73 2719.5 Q1019.74 2719.5 1022.08 2717.4 Q1024.41 2715.29 1024.41 2711.68 Q1024.41 2708.07 1022.08 2705.96 Q1019.74 2703.86 1015.73 2703.86 Q1013.86 2703.86 1011.98 2704.27 Q1010.13 2704.69 1008.19 2705.57 L1008.19 2688.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1048.3 2691.29 Q1044.69 2691.29 1042.86 2694.85 Q1041.06 2698.39 1041.06 2705.52 Q1041.06 2712.63 1042.86 2716.19 Q1044.69 2719.74 1048.3 2719.74 Q1051.94 2719.74 1053.74 2716.19 Q1055.57 2712.63 1055.57 2705.52 Q1055.57 2698.39 1053.74 2694.85 Q1051.94 2691.29 1048.3 2691.29 M1048.3 2687.58 Q1054.11 2687.58 1057.17 2692.19 Q1060.25 2696.77 1060.25 2705.52 Q1060.25 2714.25 1057.17 2718.86 Q1054.11 2723.44 1048.3 2723.44 Q1042.49 2723.44 1039.41 2718.86 Q1036.36 2714.25 1036.36 2705.52 Q1036.36 2696.77 1039.41 2692.19 Q1042.49 2687.58 1048.3 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1078.46 2691.29 Q1074.85 2691.29 1073.02 2694.85 Q1071.22 2698.39 1071.22 2705.52 Q1071.22 2712.63 1073.02 2716.19 Q1074.85 2719.74 1078.46 2719.74 Q1082.1 2719.74 1083.9 2716.19 Q1085.73 2712.63 1085.73 2705.52 Q1085.73 2698.39 1083.9 2694.85 Q1082.1 2691.29 1078.46 2691.29 M1078.46 2687.58 Q1084.27 2687.58 1087.33 2692.19 Q1090.41 2696.77 1090.41 2705.52 Q1090.41 2714.25 1087.33 2718.86 Q1084.27 2723.44 1078.46 2723.44 Q1072.65 2723.44 1069.58 2718.86 Q1066.52 2714.25 1066.52 2705.52 Q1066.52 2696.77 1069.58 2692.19 Q1072.65 2687.58 1078.46 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1263.12 2718.83 L1279.44 2718.83 L1279.44 2722.77 L1257.49 2722.77 L1257.49 2718.83 Q1260.16 2716.08 1264.74 2711.45 Q1269.35 2706.8 1270.53 2705.45 Q1272.77 2702.93 1273.65 2701.19 Q1274.55 2699.43 1274.55 2697.74 Q1274.55 2694.99 1272.61 2693.25 Q1270.69 2691.52 1267.59 2691.52 Q1265.39 2691.52 1262.93 2692.28 Q1260.5 2693.05 1257.73 2694.6 L1257.73 2689.87 Q1260.55 2688.74 1263 2688.16 Q1265.46 2687.58 1267.49 2687.58 Q1272.86 2687.58 1276.06 2690.27 Q1279.25 2692.95 1279.25 2697.44 Q1279.25 2699.57 1278.44 2701.49 Q1277.66 2703.39 1275.55 2705.99 Q1274.97 2706.66 1271.87 2709.87 Q1268.77 2713.07 1263.12 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1299.25 2691.29 Q1295.64 2691.29 1293.81 2694.85 Q1292.01 2698.39 1292.01 2705.52 Q1292.01 2712.63 1293.81 2716.19 Q1295.64 2719.74 1299.25 2719.74 Q1302.89 2719.74 1304.69 2716.19 Q1306.52 2712.63 1306.52 2705.52 Q1306.52 2698.39 1304.69 2694.85 Q1302.89 2691.29 1299.25 2691.29 M1299.25 2687.58 Q1305.06 2687.58 1308.12 2692.19 Q1311.2 2696.77 1311.2 2705.52 Q1311.2 2714.25 1308.12 2718.86 Q1305.06 2723.44 1299.25 2723.44 Q1293.44 2723.44 1290.36 2718.86 Q1287.31 2714.25 1287.31 2705.52 Q1287.31 2696.77 1290.36 2692.19 Q1293.44 2687.58 1299.25 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1329.42 2691.29 Q1325.8 2691.29 1323.98 2694.85 Q1322.17 2698.39 1322.17 2705.52 Q1322.17 2712.63 1323.98 2716.19 Q1325.8 2719.74 1329.42 2719.74 Q1333.05 2719.74 1334.86 2716.19 Q1336.68 2712.63 1336.68 2705.52 Q1336.68 2698.39 1334.86 2694.85 Q1333.05 2691.29 1329.42 2691.29 M1329.42 2687.58 Q1335.23 2687.58 1338.28 2692.19 Q1341.36 2696.77 1341.36 2705.52 Q1341.36 2714.25 1338.28 2718.86 Q1335.23 2723.44 1329.42 2723.44 Q1323.61 2723.44 1320.53 2718.86 Q1317.47 2714.25 1317.47 2705.52 Q1317.47 2696.77 1320.53 2692.19 Q1323.61 2687.58 1329.42 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1359.58 2691.29 Q1355.97 2691.29 1354.14 2694.85 Q1352.33 2698.39 1352.33 2705.52 Q1352.33 2712.63 1354.14 2716.19 Q1355.97 2719.74 1359.58 2719.74 Q1363.21 2719.74 1365.02 2716.19 Q1366.85 2712.63 1366.85 2705.52 Q1366.85 2698.39 1365.02 2694.85 Q1363.21 2691.29 1359.58 2691.29 M1359.58 2687.58 Q1365.39 2687.58 1368.44 2692.19 Q1371.52 2696.77 1371.52 2705.52 Q1371.52 2714.25 1368.44 2718.86 Q1365.39 2723.44 1359.58 2723.44 Q1353.77 2723.44 1350.69 2718.86 Q1347.63 2714.25 1347.63 2705.52 Q1347.63 2696.77 1350.69 2692.19 Q1353.77 2687.58 1359.58 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,2639.15 1352.76,2639.15 "/>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,2463.38 1352.76,2463.38 "/>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,2287.62 1352.76,2287.62 "/>
+<polyline clip-path="url(#clip636)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="159.765,2111.85 1352.76,2111.85 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,2657.17 159.765,2020.38 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,2639.15 178.663,2639.15 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,2463.38 178.663,2463.38 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,2287.62 178.663,2287.62 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="159.765,2111.85 178.663,2111.85 "/>
+<path clip-path="url(#clip630)" d="M105.821 2624.94 Q102.209 2624.94 100.381 2628.51 Q98.5752 2632.05 98.5752 2639.18 Q98.5752 2646.29 100.381 2649.85 Q102.209 2653.39 105.821 2653.39 Q109.455 2653.39 111.26 2649.85 Q113.089 2646.29 113.089 2639.18 Q113.089 2632.05 111.26 2628.51 Q109.455 2624.94 105.821 2624.94 M105.821 2621.24 Q111.631 2621.24 114.686 2625.85 Q117.765 2630.43 117.765 2639.18 Q117.765 2647.91 114.686 2652.51 Q111.631 2657.1 105.821 2657.1 Q100.01 2657.1 96.9317 2652.51 Q93.8761 2647.91 93.8761 2639.18 Q93.8761 2630.43 96.9317 2625.85 Q100.01 2621.24 105.821 2621.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M66.4689 2476.73 L74.1077 2476.73 L74.1077 2450.36 L65.7976 2452.03 L65.7976 2447.77 L74.0614 2446.1 L78.7373 2446.1 L78.7373 2476.73 L86.3762 2476.73 L86.3762 2480.66 L66.4689 2480.66 L66.4689 2476.73 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 2449.18 Q102.209 2449.18 100.381 2452.75 Q98.5752 2456.29 98.5752 2463.42 Q98.5752 2470.52 100.381 2474.09 Q102.209 2477.63 105.821 2477.63 Q109.455 2477.63 111.26 2474.09 Q113.089 2470.52 113.089 2463.42 Q113.089 2456.29 111.26 2452.75 Q109.455 2449.18 105.821 2449.18 M105.821 2445.48 Q111.631 2445.48 114.686 2450.08 Q117.765 2454.67 117.765 2463.42 Q117.765 2472.14 114.686 2476.75 Q111.631 2481.33 105.821 2481.33 Q100.01 2481.33 96.9317 2476.75 Q93.8761 2472.14 93.8761 2463.42 Q93.8761 2454.67 96.9317 2450.08 Q100.01 2445.48 105.821 2445.48 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M69.6865 2300.96 L86.0058 2300.96 L86.0058 2304.9 L64.0615 2304.9 L64.0615 2300.96 Q66.7235 2298.21 71.3068 2293.58 Q75.9133 2288.93 77.0938 2287.58 Q79.3392 2285.06 80.2188 2283.32 Q81.1216 2281.56 81.1216 2279.87 Q81.1216 2277.12 79.1771 2275.38 Q77.2559 2273.65 74.154 2273.65 Q71.955 2273.65 69.5013 2274.41 Q67.0707 2275.18 64.293 2276.73 L64.293 2272 Q67.117 2270.87 69.5707 2270.29 Q72.0244 2269.71 74.0614 2269.71 Q79.4318 2269.71 82.6262 2272.4 Q85.8206 2275.08 85.8206 2279.57 Q85.8206 2281.7 85.0105 2283.62 Q84.2234 2285.52 82.1169 2288.12 Q81.5382 2288.79 78.4364 2292 Q75.3346 2295.2 69.6865 2300.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 2273.42 Q102.209 2273.42 100.381 2276.98 Q98.5752 2280.52 98.5752 2287.65 Q98.5752 2294.76 100.381 2298.32 Q102.209 2301.87 105.821 2301.87 Q109.455 2301.87 111.26 2298.32 Q113.089 2294.76 113.089 2287.65 Q113.089 2280.52 111.26 2276.98 Q109.455 2273.42 105.821 2273.42 M105.821 2269.71 Q111.631 2269.71 114.686 2274.32 Q117.765 2278.9 117.765 2287.65 Q117.765 2296.38 114.686 2300.99 Q111.631 2305.57 105.821 2305.57 Q100.01 2305.57 96.9317 2300.99 Q93.8761 2296.38 93.8761 2287.65 Q93.8761 2278.9 96.9317 2274.32 Q100.01 2269.71 105.821 2269.71 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M79.8253 2110.5 Q83.1818 2111.22 85.0567 2113.49 Q86.9549 2115.75 86.9549 2119.09 Q86.9549 2124.2 83.4364 2127 Q79.9179 2129.81 73.4364 2129.81 Q71.2605 2129.81 68.9457 2129.37 Q66.6541 2128.95 64.2004 2128.09 L64.2004 2123.58 Q66.1448 2124.71 68.4596 2125.29 Q70.7744 2125.87 73.2976 2125.87 Q77.6957 2125.87 79.9873 2124.13 Q82.3021 2122.4 82.3021 2119.09 Q82.3021 2116.03 80.1494 2114.32 Q78.0197 2112.58 74.2003 2112.58 L70.1726 2112.58 L70.1726 2108.74 L74.3855 2108.74 Q77.8346 2108.74 79.6633 2107.37 Q81.492 2105.99 81.492 2103.39 Q81.492 2100.73 79.5938 2099.32 Q77.7188 2097.88 74.2003 2097.88 Q72.279 2097.88 70.08 2098.3 Q67.8809 2098.72 65.242 2099.6 L65.242 2095.43 Q67.9041 2094.69 70.2189 2094.32 Q72.5568 2093.95 74.617 2093.95 Q79.941 2093.95 83.0429 2096.38 Q86.1447 2098.79 86.1447 2102.91 Q86.1447 2105.78 84.5012 2107.77 Q82.8577 2109.74 79.8253 2110.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M105.821 2097.65 Q102.209 2097.65 100.381 2101.22 Q98.5752 2104.76 98.5752 2111.89 Q98.5752 2119 100.381 2122.56 Q102.209 2126.1 105.821 2126.1 Q109.455 2126.1 111.26 2122.56 Q113.089 2119 113.089 2111.89 Q113.089 2104.76 111.26 2101.22 Q109.455 2097.65 105.821 2097.65 M105.821 2093.95 Q111.631 2093.95 114.686 2098.56 Q117.765 2103.14 117.765 2111.89 Q117.765 2120.62 114.686 2125.22 Q111.631 2129.81 105.821 2129.81 Q100.01 2129.81 96.9317 2125.22 Q93.8761 2120.62 93.8761 2111.89 Q93.8761 2103.14 96.9317 2098.56 Q100.01 2093.95 105.821 2093.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M469.436 1878.76 L477.619 1878.76 L477.619 1932.36 L507.069 1932.36 L507.069 1939.24 L469.436 1939.24 L469.436 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M535.952 1916.44 Q526.918 1916.44 523.435 1918.5 Q519.951 1920.57 519.951 1925.55 Q519.951 1929.52 522.543 1931.87 Q525.177 1934.18 529.673 1934.18 Q535.871 1934.18 539.598 1929.8 Q543.365 1925.39 543.365 1918.1 L543.365 1916.44 L535.952 1916.44 M550.819 1913.36 L550.819 1939.24 L543.365 1939.24 L543.365 1932.36 Q540.813 1936.49 537.005 1938.47 Q533.197 1940.42 527.688 1940.42 Q520.721 1940.42 516.589 1936.53 Q512.497 1932.6 512.497 1926.04 Q512.497 1918.38 517.601 1914.49 Q522.746 1910.6 532.914 1910.6 L543.365 1910.6 L543.365 1909.87 Q543.365 1904.73 539.962 1901.93 Q536.6 1899.1 530.483 1899.1 Q526.594 1899.1 522.908 1900.03 Q519.222 1900.96 515.819 1902.83 L515.819 1895.94 Q519.91 1894.36 523.759 1893.59 Q527.607 1892.78 531.253 1892.78 Q541.097 1892.78 545.958 1897.88 Q550.819 1902.99 550.819 1913.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M573.544 1880.99 L573.544 1893.87 L588.897 1893.87 L588.897 1899.67 L573.544 1899.67 L573.544 1924.29 Q573.544 1929.84 575.043 1931.42 Q576.583 1933 581.241 1933 L588.897 1933 L588.897 1939.24 L581.241 1939.24 Q572.613 1939.24 569.331 1936.04 Q566.05 1932.8 566.05 1924.29 L566.05 1899.67 L560.581 1899.67 L560.581 1893.87 L566.05 1893.87 L566.05 1880.99 L573.544 1880.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M598.7 1893.87 L606.154 1893.87 L606.154 1939.24 L598.7 1939.24 L598.7 1893.87 M598.7 1876.21 L606.154 1876.21 L606.154 1885.65 L598.7 1885.65 L598.7 1876.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M629.123 1880.99 L629.123 1893.87 L644.476 1893.87 L644.476 1899.67 L629.123 1899.67 L629.123 1924.29 Q629.123 1929.84 630.622 1931.42 Q632.161 1933 636.819 1933 L644.476 1933 L644.476 1939.24 L636.819 1939.24 Q628.191 1939.24 624.91 1936.04 Q621.629 1932.8 621.629 1924.29 L621.629 1899.67 L616.16 1899.67 L616.16 1893.87 L621.629 1893.87 L621.629 1880.99 L629.123 1880.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M653.509 1921.34 L653.509 1893.87 L660.963 1893.87 L660.963 1921.05 Q660.963 1927.5 663.474 1930.74 Q665.986 1933.94 671.009 1933.94 Q677.045 1933.94 680.529 1930.09 Q684.053 1926.24 684.053 1919.6 L684.053 1893.87 L691.507 1893.87 L691.507 1939.24 L684.053 1939.24 L684.053 1932.28 Q681.339 1936.41 677.734 1938.43 Q674.169 1940.42 669.429 1940.42 Q661.611 1940.42 657.56 1935.56 Q653.509 1930.7 653.509 1921.34 M672.265 1892.78 L672.265 1892.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M736.715 1900.76 L736.715 1876.21 L744.168 1876.21 L744.168 1939.24 L736.715 1939.24 L736.715 1932.44 Q734.365 1936.49 730.76 1938.47 Q727.195 1940.42 722.172 1940.42 Q713.949 1940.42 708.764 1933.85 Q703.619 1927.29 703.619 1916.6 Q703.619 1905.9 708.764 1899.34 Q713.949 1892.78 722.172 1892.78 Q727.195 1892.78 730.76 1894.76 Q734.365 1896.71 736.715 1900.76 M711.316 1916.6 Q711.316 1924.82 714.678 1929.52 Q718.081 1934.18 723.995 1934.18 Q729.909 1934.18 733.312 1929.52 Q736.715 1924.82 736.715 1916.6 Q736.715 1908.37 733.312 1903.72 Q729.909 1899.02 723.995 1899.02 Q718.081 1899.02 714.678 1903.72 Q711.316 1908.37 711.316 1916.6 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M798.329 1914.69 L798.329 1918.34 L764.058 1918.34 Q764.545 1926.04 768.676 1930.09 Q772.849 1934.1 780.262 1934.1 Q784.556 1934.1 788.566 1933.04 Q792.617 1931.99 796.587 1929.89 L796.587 1936.93 Q792.577 1938.64 788.364 1939.53 Q784.151 1940.42 779.816 1940.42 Q768.96 1940.42 762.6 1934.1 Q756.281 1927.78 756.281 1917 Q756.281 1905.86 762.276 1899.34 Q768.312 1892.78 778.52 1892.78 Q787.675 1892.78 792.982 1898.69 Q798.329 1904.57 798.329 1914.69 M790.875 1912.51 Q790.794 1906.39 787.432 1902.74 Q784.11 1899.1 778.601 1899.1 Q772.363 1899.1 768.595 1902.62 Q764.869 1906.15 764.301 1912.55 L790.875 1912.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M854.839 1876.29 Q849.411 1885.61 846.778 1894.72 Q844.145 1903.84 844.145 1913.2 Q844.145 1922.55 846.778 1931.75 Q849.451 1940.9 854.839 1950.18 L848.358 1950.18 Q842.281 1940.66 839.243 1931.46 Q836.246 1922.27 836.246 1913.2 Q836.246 1904.16 839.243 1895.01 Q842.241 1885.85 848.358 1876.29 L854.839 1876.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M899.156 1900.76 L899.156 1876.21 L906.61 1876.21 L906.61 1939.24 L899.156 1939.24 L899.156 1932.44 Q896.807 1936.49 893.201 1938.47 Q889.636 1940.42 884.613 1940.42 Q876.39 1940.42 871.205 1933.85 Q866.06 1927.29 866.06 1916.6 Q866.06 1905.9 871.205 1899.34 Q876.39 1892.78 884.613 1892.78 Q889.636 1892.78 893.201 1894.76 Q896.807 1896.71 899.156 1900.76 M873.757 1916.6 Q873.757 1924.82 877.119 1929.52 Q880.522 1934.18 886.436 1934.18 Q892.351 1934.18 895.753 1929.52 Q899.156 1924.82 899.156 1916.6 Q899.156 1908.37 895.753 1903.72 Q892.351 1899.02 886.436 1899.02 Q880.522 1899.02 877.119 1903.72 Q873.757 1908.37 873.757 1916.6 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M960.77 1914.69 L960.77 1918.34 L926.5 1918.34 Q926.986 1926.04 931.118 1930.09 Q935.29 1934.1 942.703 1934.1 Q946.997 1934.1 951.008 1933.04 Q955.059 1931.99 959.028 1929.89 L959.028 1936.93 Q955.018 1938.64 950.805 1939.53 Q946.592 1940.42 942.258 1940.42 Q931.401 1940.42 925.041 1934.1 Q918.722 1927.78 918.722 1917 Q918.722 1905.86 924.717 1899.34 Q930.753 1892.78 940.961 1892.78 Q950.116 1892.78 955.423 1898.69 Q960.77 1904.57 960.77 1914.69 M953.317 1912.51 Q953.236 1906.39 949.873 1902.74 Q946.552 1899.1 941.042 1899.1 Q934.804 1899.1 931.037 1902.62 Q927.31 1906.15 926.743 1912.55 L953.317 1912.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1002.86 1916.03 Q1002.86 1907.93 999.497 1903.47 Q996.175 1899.02 990.139 1899.02 Q984.144 1899.02 980.782 1903.47 Q977.46 1907.93 977.46 1916.03 Q977.46 1924.09 980.782 1928.55 Q984.144 1933 990.139 1933 Q996.175 1933 999.497 1928.55 Q1002.86 1924.09 1002.86 1916.03 M1010.31 1933.61 Q1010.31 1945.2 1005.17 1950.83 Q1000.02 1956.5 989.41 1956.5 Q985.481 1956.5 981.997 1955.89 Q978.513 1955.32 975.232 1954.11 L975.232 1946.86 Q978.513 1948.64 981.713 1949.49 Q984.914 1950.34 988.235 1950.34 Q995.568 1950.34 999.213 1946.49 Q1002.86 1942.69 1002.86 1934.95 L1002.86 1931.26 Q1000.55 1935.27 996.945 1937.26 Q993.34 1939.24 988.316 1939.24 Q979.972 1939.24 974.867 1932.88 Q969.763 1926.52 969.763 1916.03 Q969.763 1905.5 974.867 1899.14 Q979.972 1892.78 988.316 1892.78 Q993.34 1892.78 996.945 1894.76 Q1000.55 1896.75 1002.86 1900.76 L1002.86 1893.87 L1010.31 1893.87 L1010.31 1933.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1024.49 1876.29 L1030.97 1876.29 Q1037.05 1885.85 1040.05 1895.01 Q1043.08 1904.16 1043.08 1913.2 Q1043.08 1922.27 1040.05 1931.46 Q1037.05 1940.66 1030.97 1950.18 L1024.49 1950.18 Q1029.88 1940.9 1032.51 1931.75 Q1035.19 1922.55 1035.19 1913.2 Q1035.19 1903.84 1032.51 1894.72 Q1029.88 1885.61 1024.49 1876.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip636)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="193.529,2639.15 195.771,2639.15 198.013,2639.14 200.255,2639.14 202.497,2639.14 204.739,2639.13 206.981,2639.12 209.223,2639.11 211.464,2639.1 213.706,2639.09 215.948,2639.07 218.19,2639.05 220.432,2639.03 222.674,2639.01 224.916,2638.98 227.158,2638.95 229.4,2638.91 231.642,2638.87 233.884,2638.83 236.126,2638.78 238.368,2638.73 240.61,2638.67 242.852,2638.6 245.094,2638.53 247.336,2638.46 249.578,2638.37 251.82,2638.28 254.062,2638.19 256.304,2638.08 258.546,2637.97 260.788,2637.84 263.03,2637.71 265.271,2637.57 267.513,2637.41 269.755,2637.25 271.997,2637.07 274.239,2636.88 276.481,2636.68 278.723,2636.47 280.965,2636.24 283.207,2636 285.449,2635.75 287.691,2635.48 289.933,2635.2 292.175,2634.9 294.417,2634.59 296.659,2634.26 298.901,2633.92 301.143,2633.56 303.385,2633.19 305.627,2632.8 307.869,2632.39 310.111,2631.97 312.353,2631.54 314.595,2631.09 316.837,2630.62 319.078,2630.15 321.32,2629.65 323.562,2629.15 325.804,2628.63 328.046,2628.1 330.288,2627.55 332.53,2626.99 334.772,2626.43 337.014,2625.84 339.256,2625.25 341.498,2624.65 343.74,2624.04 345.982,2623.42 348.224,2622.78 350.466,2622.14 352.708,2621.49 354.95,2620.83 357.192,2620.16 359.434,2619.49 361.676,2618.8 363.918,2618.11 366.16,2617.41 368.402,2616.71 370.644,2616 372.885,2615.28 375.127,2614.55 377.369,2613.82 379.611,2613.09 381.853,2612.34 384.095,2611.6 386.337,2610.84 388.579,2610.08 390.821,2609.32 393.063,2608.55 395.305,2607.78 397.547,2607 399.789,2606.21 402.031,2605.43 404.273,2604.63 406.515,2603.83 408.757,2603.03 410.999,2602.23 413.241,2601.41 415.483,2600.6 417.725,2599.78 419.967,2598.95 422.209,2598.12 424.451,2597.29 426.692,2596.45 428.934,2595.61 431.176,2594.76 433.418,2593.91 435.66,2593.05 437.902,2592.19 440.144,2591.32 442.386,2590.45 444.628,2589.57 446.87,2588.69 449.112,2587.8 451.354,2586.9 453.596,2586 455.838,2585.1 458.08,2584.19 460.322,2583.27 462.564,2582.34 464.806,2581.41 467.048,2580.48 469.29,2579.53 471.532,2578.58 473.774,2577.62 476.016,2576.65 478.258,2575.68 480.5,2574.69 482.741,2573.7 484.983,2572.7 487.225,2571.69 489.467,2570.67 491.709,2569.64 493.951,2568.6 496.193,2567.56 498.435,2566.5 500.677,2565.43 502.919,2564.35 505.161,2563.26 507.403,2562.16 509.645,2561.05 511.887,2559.93 514.129,2558.8 516.371,2557.66 518.613,2556.51 520.855,2555.35 523.097,2554.17 525.339,2552.99 527.581,2551.8 529.823,2550.59 532.065,2549.38 534.307,2548.16 536.548,2546.92 538.79,2545.68 541.032,2544.43 543.274,2543.17 545.516,2541.9 547.758,2540.63 550,2539.34 552.242,2538.05 554.484,2536.75 556.726,2535.44 558.968,2534.13 561.21,2532.81 563.452,2531.48 565.694,2530.15 567.936,2528.81 570.178,2527.46 572.42,2526.11 574.662,2524.75 576.904,2523.39 579.146,2522.02 581.388,2520.64 583.63,2519.26 585.872,2517.88 588.114,2516.49 590.355,2515.09 592.597,2513.69 594.839,2512.28 597.081,2510.87 599.323,2509.46 601.565,2508.04 603.807,2506.61 606.049,2505.18 608.291,2503.75 610.533,2502.31 612.775,2500.86 615.017,2499.41 617.259,2497.96 619.501,2496.49 621.743,2495.03 623.985,2493.56 626.227,2492.08 628.469,2490.6 630.711,2489.11 632.953,2487.62 635.195,2486.12 637.437,2484.62 639.679,2483.11 641.921,2481.59 644.162,2480.07 646.404,2478.54 648.646,2477.01 650.888,2475.47 653.13,2473.92 655.372,2472.37 657.614,2470.81 659.856,2469.24 662.098,2467.67 664.34,2466.09 666.582,2464.51 668.824,2462.91 671.066,2461.32 673.308,2459.71 675.55,2458.1 677.792,2456.49 680.034,2454.87 682.276,2453.24 684.518,2451.6 686.76,2449.97 689.002,2448.32 691.244,2446.67 693.486,2445.02 695.728,2443.36 697.969,2441.69 700.211,2440.02 702.453,2438.35 704.695,2436.67 706.937,2434.99 709.179,2433.3 711.421,2431.61 713.663,2429.92 715.905,2428.22 718.147,2426.52 720.389,2424.81 722.631,2423.1 724.873,2421.39 727.115,2419.67 729.357,2417.96 731.599,2416.23 733.841,2414.51 736.083,2412.78 738.325,2411.05 740.567,2409.32 742.809,2407.58 745.051,2405.84 747.293,2404.1 749.535,2402.36 751.776,2400.61 754.018,2398.86 756.26,2397.11 758.502,2395.35 760.744,2393.59 762.986,2391.83 765.228,2390.07 767.47,2388.3 769.712,2386.53 771.954,2384.76 774.196,2382.99 776.438,2381.21 778.68,2379.43 780.922,2377.65 783.164,2375.86 785.406,2374.08 787.648,2372.29 789.89,2370.5 792.132,2368.7 794.374,2366.9 796.616,2365.11 798.858,2363.3 801.1,2361.5 803.342,2359.7 805.583,2357.89 807.825,2356.08 810.067,2354.27 812.309,2352.46 814.551,2350.64 816.793,2348.83 819.035,2347.01 821.277,2345.19 823.519,2343.37 825.761,2341.55 828.003,2339.73 830.245,2337.91 832.487,2336.08 834.729,2334.26 836.971,2332.43 839.213,2330.61 841.455,2328.78 843.697,2326.96 845.939,2325.13 848.181,2323.3 850.423,2321.47 852.665,2319.65 854.907,2317.82 857.149,2315.99 859.391,2314.16 861.632,2312.34 863.874,2310.51 866.116,2308.68 868.358,2306.86 870.6,2305.03 872.842,2303.2 875.084,2301.38 877.326,2299.55 879.568,2297.73 881.81,2295.9 884.052,2294.08 886.294,2292.26 888.536,2290.44 890.778,2288.61 893.02,2286.79 895.262,2284.98 897.504,2283.16 899.746,2281.34 901.988,2279.52 904.23,2277.71 906.472,2275.9 908.714,2274.09 910.956,2272.28 913.198,2270.47 915.439,2268.66 917.681,2266.86 919.923,2265.05 922.165,2263.25 924.407,2261.46 926.649,2259.66 928.891,2257.86 931.133,2256.07 933.375,2254.28 935.617,2252.49 937.859,2250.71 940.101,2248.93 942.343,2247.15 944.585,2245.37 946.827,2243.6 949.069,2241.82 951.311,2240.06 953.553,2238.29 955.795,2236.53 958.037,2234.77 960.279,2233.01 962.521,2231.26 964.763,2229.51 967.005,2227.76 969.246,2226.02 971.488,2224.28 973.73,2222.54 975.972,2220.81 978.214,2219.08 980.456,2217.35 982.698,2215.63 984.94,2213.91 987.182,2212.2 989.424,2210.49 991.666,2208.78 993.908,2207.08 996.15,2205.38 998.392,2203.69 1000.63,2202 1002.88,2200.31 1005.12,2198.63 1007.36,2196.96 1009.6,2195.29 1011.84,2193.62 1014.09,2191.96 1016.33,2190.31 1018.57,2188.66 1020.81,2187.01 1023.05,2185.37 1025.3,2183.74 1027.54,2182.11 1029.78,2180.48 1032.02,2178.87 1034.26,2177.25 1036.51,2175.65 1038.75,2174.05 1040.99,2172.45 1043.23,2170.86 1045.47,2169.28 1047.72,2167.7 1049.96,2166.13 1052.2,2164.57 1054.44,2163.01 1056.68,2161.46 1058.92,2159.91 1061.17,2158.37 1063.41,2156.84 1065.65,2155.31 1067.89,2153.79 1070.13,2152.28 1072.38,2150.77 1074.62,2149.27 1076.86,2147.78 1079.1,2146.29 1081.34,2144.81 1083.59,2143.34 1085.83,2141.88 1088.07,2140.42 1090.31,2138.97 1092.55,2137.52 1094.8,2136.08 1097.04,2134.65 1099.28,2133.23 1101.52,2131.82 1103.76,2130.41 1106.01,2129.01 1108.25,2127.62 1110.49,2126.24 1112.73,2124.86 1114.97,2123.49 1117.22,2122.13 1119.46,2120.78 1121.7,2119.44 1123.94,2118.11 1126.18,2116.78 1128.43,2115.46 1130.67,2114.15 1132.91,2112.85 1135.15,2111.56 1137.39,2110.28 1139.64,2109 1141.88,2107.74 1144.12,2106.48 1146.36,2105.23 1148.6,2103.99 1150.85,2102.76 1153.09,2101.54 1155.33,2100.33 1157.57,2099.12 1159.81,2097.93 1162.05,2096.75 1164.3,2095.57 1166.54,2094.4 1168.78,2093.25 1171.02,2092.1 1173.26,2090.96 1175.51,2089.83 1177.75,2088.71 1179.99,2087.6 1182.23,2086.5 1184.47,2085.4 1186.72,2084.32 1188.96,2083.25 1191.2,2082.18 1193.44,2081.13 1195.68,2080.09 1197.93,2079.05 1200.17,2078.03 1202.41,2077.02 1204.65,2076.01 1206.89,2075.02 1209.14,2074.03 1211.38,2073.06 1213.62,2072.09 1215.86,2071.14 1218.1,2070.2 1220.35,2069.26 1222.59,2068.34 1224.83,2067.43 1227.07,2066.53 1229.31,2065.64 1231.56,2064.76 1233.8,2063.88 1236.04,2063.02 1238.28,2062.18 1240.52,2061.34 1242.77,2060.51 1245.01,2059.69 1247.25,2058.88 1249.49,2058.09 1251.73,2057.3 1253.98,2056.53 1256.22,2055.76 1258.46,2055.01 1260.7,2054.26 1262.94,2053.53 1265.19,2052.8 1267.43,2052.09 1269.67,2051.39 1271.91,2050.69 1274.15,2050.01 1276.39,2049.34 1278.64,2048.67 1280.88,2048.02 1283.12,2047.37 1285.36,2046.74 1287.6,2046.11 1289.85,2045.5 1292.09,2044.89 1294.33,2044.3 1296.57,2043.71 1298.81,2043.14 1301.06,2042.57 1303.3,2042.01 1305.54,2041.47 1307.78,2040.93 1310.02,2040.4 1312.27,2039.89 1314.51,2039.38 1316.75,2038.89 1318.99,2038.4 "/>
+<path clip-path="url(#clip630)" d="M1559.76 2657.17 L2752.76 2657.17 L2752.76 2020.38 L1559.76 2020.38  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip637">
+    <rect x="1559" y="2020" width="1194" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1593.53,2657.17 1593.53,2020.38 "/>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1873.77,2657.17 1873.77,2020.38 "/>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2154.02,2657.17 2154.02,2020.38 "/>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2434.26,2657.17 2434.26,2020.38 "/>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2714.51,2657.17 2714.51,2020.38 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,2657.17 2752.76,2657.17 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1593.53,2657.17 1593.53,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1873.77,2657.17 1873.77,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2154.02,2657.17 2154.02,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2434.26,2657.17 2434.26,2638.27 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2714.51,2657.17 2714.51,2638.27 "/>
+<path clip-path="url(#clip630)" d="M1593.53 2691.29 Q1589.92 2691.29 1588.09 2694.85 Q1586.28 2698.39 1586.28 2705.52 Q1586.28 2712.63 1588.09 2716.19 Q1589.92 2719.74 1593.53 2719.74 Q1597.16 2719.74 1598.97 2716.19 Q1600.8 2712.63 1600.8 2705.52 Q1600.8 2698.39 1598.97 2694.85 Q1597.16 2691.29 1593.53 2691.29 M1593.53 2687.58 Q1599.34 2687.58 1602.39 2692.19 Q1605.47 2696.77 1605.47 2705.52 Q1605.47 2714.25 1602.39 2718.86 Q1599.34 2723.44 1593.53 2723.44 Q1587.72 2723.44 1584.64 2718.86 Q1581.58 2714.25 1581.58 2705.52 Q1581.58 2696.77 1584.64 2692.19 Q1587.72 2687.58 1593.53 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1833.39 2688.21 L1851.75 2688.21 L1851.75 2692.14 L1837.67 2692.14 L1837.67 2700.62 Q1838.69 2700.27 1839.71 2700.11 Q1840.73 2699.92 1841.75 2699.92 Q1847.54 2699.92 1850.91 2703.09 Q1854.29 2706.26 1854.29 2711.68 Q1854.29 2717.26 1850.82 2720.36 Q1847.35 2723.44 1841.03 2723.44 Q1838.85 2723.44 1836.59 2723.07 Q1834.34 2722.7 1831.93 2721.96 L1831.93 2717.26 Q1834.02 2718.39 1836.24 2718.95 Q1838.46 2719.5 1840.94 2719.5 Q1844.94 2719.5 1847.28 2717.4 Q1849.62 2715.29 1849.62 2711.68 Q1849.62 2708.07 1847.28 2705.96 Q1844.94 2703.86 1840.94 2703.86 Q1839.06 2703.86 1837.19 2704.27 Q1835.34 2704.69 1833.39 2705.57 L1833.39 2688.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1873.51 2691.29 Q1869.9 2691.29 1868.07 2694.85 Q1866.26 2698.39 1866.26 2705.52 Q1866.26 2712.63 1868.07 2716.19 Q1869.9 2719.74 1873.51 2719.74 Q1877.14 2719.74 1878.95 2716.19 Q1880.78 2712.63 1880.78 2705.52 Q1880.78 2698.39 1878.95 2694.85 Q1877.14 2691.29 1873.51 2691.29 M1873.51 2687.58 Q1879.32 2687.58 1882.37 2692.19 Q1885.45 2696.77 1885.45 2705.52 Q1885.45 2714.25 1882.37 2718.86 Q1879.32 2723.44 1873.51 2723.44 Q1867.7 2723.44 1864.62 2718.86 Q1861.56 2714.25 1861.56 2705.52 Q1861.56 2696.77 1864.62 2692.19 Q1867.7 2687.58 1873.51 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1903.67 2691.29 Q1900.06 2691.29 1898.23 2694.85 Q1896.42 2698.39 1896.42 2705.52 Q1896.42 2712.63 1898.23 2716.19 Q1900.06 2719.74 1903.67 2719.74 Q1907.3 2719.74 1909.11 2716.19 Q1910.94 2712.63 1910.94 2705.52 Q1910.94 2698.39 1909.11 2694.85 Q1907.3 2691.29 1903.67 2691.29 M1903.67 2687.58 Q1909.48 2687.58 1912.53 2692.19 Q1915.61 2696.77 1915.61 2705.52 Q1915.61 2714.25 1912.53 2718.86 Q1909.48 2723.44 1903.67 2723.44 Q1897.86 2723.44 1894.78 2718.86 Q1891.72 2714.25 1891.72 2705.52 Q1891.72 2696.77 1894.78 2692.19 Q1897.86 2687.58 1903.67 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2098.54 2718.83 L2106.18 2718.83 L2106.18 2692.47 L2097.87 2694.13 L2097.87 2689.87 L2106.14 2688.21 L2110.81 2688.21 L2110.81 2718.83 L2118.45 2718.83 L2118.45 2722.77 L2098.54 2722.77 L2098.54 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2137.9 2691.29 Q2134.28 2691.29 2132.46 2694.85 Q2130.65 2698.39 2130.65 2705.52 Q2130.65 2712.63 2132.46 2716.19 Q2134.28 2719.74 2137.9 2719.74 Q2141.53 2719.74 2143.34 2716.19 Q2145.16 2712.63 2145.16 2705.52 Q2145.16 2698.39 2143.34 2694.85 Q2141.53 2691.29 2137.9 2691.29 M2137.9 2687.58 Q2143.71 2687.58 2146.76 2692.19 Q2149.84 2696.77 2149.84 2705.52 Q2149.84 2714.25 2146.76 2718.86 Q2143.71 2723.44 2137.9 2723.44 Q2132.09 2723.44 2129.01 2718.86 Q2125.95 2714.25 2125.95 2705.52 Q2125.95 2696.77 2129.01 2692.19 Q2132.09 2687.58 2137.9 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2168.06 2691.29 Q2164.45 2691.29 2162.62 2694.85 Q2160.81 2698.39 2160.81 2705.52 Q2160.81 2712.63 2162.62 2716.19 Q2164.45 2719.74 2168.06 2719.74 Q2171.69 2719.74 2173.5 2716.19 Q2175.33 2712.63 2175.33 2705.52 Q2175.33 2698.39 2173.5 2694.85 Q2171.69 2691.29 2168.06 2691.29 M2168.06 2687.58 Q2173.87 2687.58 2176.92 2692.19 Q2180 2696.77 2180 2705.52 Q2180 2714.25 2176.92 2718.86 Q2173.87 2723.44 2168.06 2723.44 Q2162.25 2723.44 2159.17 2718.86 Q2156.11 2714.25 2156.11 2705.52 Q2156.11 2696.77 2159.17 2692.19 Q2162.25 2687.58 2168.06 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2198.22 2691.29 Q2194.61 2691.29 2192.78 2694.85 Q2190.97 2698.39 2190.97 2705.52 Q2190.97 2712.63 2192.78 2716.19 Q2194.61 2719.74 2198.22 2719.74 Q2201.85 2719.74 2203.66 2716.19 Q2205.49 2712.63 2205.49 2705.52 Q2205.49 2698.39 2203.66 2694.85 Q2201.85 2691.29 2198.22 2691.29 M2198.22 2687.58 Q2204.03 2687.58 2207.09 2692.19 Q2210.16 2696.77 2210.16 2705.52 Q2210.16 2714.25 2207.09 2718.86 Q2204.03 2723.44 2198.22 2723.44 Q2192.41 2723.44 2189.33 2718.86 Q2186.28 2714.25 2186.28 2705.52 Q2186.28 2696.77 2189.33 2692.19 Q2192.41 2687.58 2198.22 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2378.79 2718.83 L2386.43 2718.83 L2386.43 2692.47 L2378.12 2694.13 L2378.12 2689.87 L2386.38 2688.21 L2391.06 2688.21 L2391.06 2718.83 L2398.7 2718.83 L2398.7 2722.77 L2378.79 2722.77 L2378.79 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2408.19 2688.21 L2426.54 2688.21 L2426.54 2692.14 L2412.47 2692.14 L2412.47 2700.62 Q2413.49 2700.27 2414.51 2700.11 Q2415.52 2699.92 2416.54 2699.92 Q2422.33 2699.92 2425.71 2703.09 Q2429.09 2706.26 2429.09 2711.68 Q2429.09 2717.26 2425.62 2720.36 Q2422.15 2723.44 2415.83 2723.44 Q2413.65 2723.44 2411.38 2723.07 Q2409.14 2722.7 2406.73 2721.96 L2406.73 2717.26 Q2408.81 2718.39 2411.03 2718.95 Q2413.26 2719.5 2415.73 2719.5 Q2419.74 2719.5 2422.08 2717.4 Q2424.41 2715.29 2424.41 2711.68 Q2424.41 2708.07 2422.08 2705.96 Q2419.74 2703.86 2415.73 2703.86 Q2413.86 2703.86 2411.98 2704.27 Q2410.13 2704.69 2408.19 2705.57 L2408.19 2688.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2448.3 2691.29 Q2444.69 2691.29 2442.86 2694.85 Q2441.06 2698.39 2441.06 2705.52 Q2441.06 2712.63 2442.86 2716.19 Q2444.69 2719.74 2448.3 2719.74 Q2451.94 2719.74 2453.74 2716.19 Q2455.57 2712.63 2455.57 2705.52 Q2455.57 2698.39 2453.74 2694.85 Q2451.94 2691.29 2448.3 2691.29 M2448.3 2687.58 Q2454.11 2687.58 2457.17 2692.19 Q2460.25 2696.77 2460.25 2705.52 Q2460.25 2714.25 2457.17 2718.86 Q2454.11 2723.44 2448.3 2723.44 Q2442.49 2723.44 2439.41 2718.86 Q2436.36 2714.25 2436.36 2705.52 Q2436.36 2696.77 2439.41 2692.19 Q2442.49 2687.58 2448.3 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2478.46 2691.29 Q2474.85 2691.29 2473.02 2694.85 Q2471.22 2698.39 2471.22 2705.52 Q2471.22 2712.63 2473.02 2716.19 Q2474.85 2719.74 2478.46 2719.74 Q2482.1 2719.74 2483.9 2716.19 Q2485.73 2712.63 2485.73 2705.52 Q2485.73 2698.39 2483.9 2694.85 Q2482.1 2691.29 2478.46 2691.29 M2478.46 2687.58 Q2484.27 2687.58 2487.33 2692.19 Q2490.41 2696.77 2490.41 2705.52 Q2490.41 2714.25 2487.33 2718.86 Q2484.27 2723.44 2478.46 2723.44 Q2472.65 2723.44 2469.58 2718.86 Q2466.52 2714.25 2466.52 2705.52 Q2466.52 2696.77 2469.58 2692.19 Q2472.65 2687.58 2478.46 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2663.12 2718.83 L2679.44 2718.83 L2679.44 2722.77 L2657.49 2722.77 L2657.49 2718.83 Q2660.16 2716.08 2664.74 2711.45 Q2669.35 2706.8 2670.53 2705.45 Q2672.77 2702.93 2673.65 2701.19 Q2674.55 2699.43 2674.55 2697.74 Q2674.55 2694.99 2672.61 2693.25 Q2670.69 2691.52 2667.59 2691.52 Q2665.39 2691.52 2662.93 2692.28 Q2660.5 2693.05 2657.73 2694.6 L2657.73 2689.87 Q2660.55 2688.74 2663 2688.16 Q2665.46 2687.58 2667.49 2687.58 Q2672.86 2687.58 2676.06 2690.27 Q2679.25 2692.95 2679.25 2697.44 Q2679.25 2699.57 2678.44 2701.49 Q2677.66 2703.39 2675.55 2705.99 Q2674.97 2706.66 2671.87 2709.87 Q2668.77 2713.07 2663.12 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2699.25 2691.29 Q2695.64 2691.29 2693.81 2694.85 Q2692.01 2698.39 2692.01 2705.52 Q2692.01 2712.63 2693.81 2716.19 Q2695.64 2719.74 2699.25 2719.74 Q2702.89 2719.74 2704.69 2716.19 Q2706.52 2712.63 2706.52 2705.52 Q2706.52 2698.39 2704.69 2694.85 Q2702.89 2691.29 2699.25 2691.29 M2699.25 2687.58 Q2705.06 2687.58 2708.12 2692.19 Q2711.2 2696.77 2711.2 2705.52 Q2711.2 2714.25 2708.12 2718.86 Q2705.06 2723.44 2699.25 2723.44 Q2693.44 2723.44 2690.36 2718.86 Q2687.31 2714.25 2687.31 2705.52 Q2687.31 2696.77 2690.36 2692.19 Q2693.44 2687.58 2699.25 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2729.42 2691.29 Q2725.8 2691.29 2723.98 2694.85 Q2722.17 2698.39 2722.17 2705.52 Q2722.17 2712.63 2723.98 2716.19 Q2725.8 2719.74 2729.42 2719.74 Q2733.05 2719.74 2734.86 2716.19 Q2736.68 2712.63 2736.68 2705.52 Q2736.68 2698.39 2734.86 2694.85 Q2733.05 2691.29 2729.42 2691.29 M2729.42 2687.58 Q2735.23 2687.58 2738.28 2692.19 Q2741.36 2696.77 2741.36 2705.52 Q2741.36 2714.25 2738.28 2718.86 Q2735.23 2723.44 2729.42 2723.44 Q2723.61 2723.44 2720.53 2718.86 Q2717.47 2714.25 2717.47 2705.52 Q2717.47 2696.77 2720.53 2692.19 Q2723.61 2687.58 2729.42 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2759.58 2691.29 Q2755.97 2691.29 2754.14 2694.85 Q2752.33 2698.39 2752.33 2705.52 Q2752.33 2712.63 2754.14 2716.19 Q2755.97 2719.74 2759.58 2719.74 Q2763.21 2719.74 2765.02 2716.19 Q2766.85 2712.63 2766.85 2705.52 Q2766.85 2698.39 2765.02 2694.85 Q2763.21 2691.29 2759.58 2691.29 M2759.58 2687.58 Q2765.39 2687.58 2768.44 2692.19 Q2771.52 2696.77 2771.52 2705.52 Q2771.52 2714.25 2768.44 2718.86 Q2765.39 2723.44 2759.58 2723.44 Q2753.77 2723.44 2750.69 2718.86 Q2747.63 2714.25 2747.63 2705.52 Q2747.63 2696.77 2750.69 2692.19 Q2753.77 2687.58 2759.58 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,2549.77 2752.76,2549.77 "/>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,2403.67 2752.76,2403.67 "/>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,2257.56 2752.76,2257.56 "/>
+<polyline clip-path="url(#clip637)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1559.76,2111.45 2752.76,2111.45 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,2657.17 1559.76,2020.38 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,2549.77 1578.66,2549.77 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,2403.67 1578.66,2403.67 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,2257.56 1578.66,2257.56 "/>
+<polyline clip-path="url(#clip630)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1559.76,2111.45 1578.66,2111.45 "/>
+<path clip-path="url(#clip630)" d="M1469.69 2563.12 L1486.01 2563.12 L1486.01 2567.05 L1464.06 2567.05 L1464.06 2563.12 Q1466.72 2560.36 1471.31 2555.74 Q1475.91 2551.08 1477.09 2549.74 Q1479.34 2547.22 1480.22 2545.48 Q1481.12 2543.72 1481.12 2542.03 Q1481.12 2539.28 1479.18 2537.54 Q1477.26 2535.8 1474.15 2535.8 Q1471.95 2535.8 1469.5 2536.57 Q1467.07 2537.33 1464.29 2538.88 L1464.29 2534.16 Q1467.12 2533.03 1469.57 2532.45 Q1472.02 2531.87 1474.06 2531.87 Q1479.43 2531.87 1482.63 2534.55 Q1485.82 2537.24 1485.82 2541.73 Q1485.82 2543.86 1485.01 2545.78 Q1484.22 2547.68 1482.12 2550.27 Q1481.54 2550.94 1478.44 2554.16 Q1475.33 2557.36 1469.69 2563.12 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1505.82 2535.57 Q1502.21 2535.57 1500.38 2539.14 Q1498.58 2542.68 1498.58 2549.81 Q1498.58 2556.92 1500.38 2560.48 Q1502.21 2564.02 1505.82 2564.02 Q1509.45 2564.02 1511.26 2560.48 Q1513.09 2556.92 1513.09 2549.81 Q1513.09 2542.68 1511.26 2539.14 Q1509.45 2535.57 1505.82 2535.57 M1505.82 2531.87 Q1511.63 2531.87 1514.69 2536.48 Q1517.76 2541.06 1517.76 2549.81 Q1517.76 2558.54 1514.69 2563.14 Q1511.63 2567.73 1505.82 2567.73 Q1500.01 2567.73 1496.93 2563.14 Q1493.88 2558.54 1493.88 2549.81 Q1493.88 2541.06 1496.93 2536.48 Q1500.01 2531.87 1505.82 2531.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1478.51 2390.46 L1466.7 2408.91 L1478.51 2408.91 L1478.51 2390.46 M1477.28 2386.39 L1483.16 2386.39 L1483.16 2408.91 L1488.09 2408.91 L1488.09 2412.8 L1483.16 2412.8 L1483.16 2420.95 L1478.51 2420.95 L1478.51 2412.8 L1462.9 2412.8 L1462.9 2408.29 L1477.28 2386.39 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1505.82 2389.47 Q1502.21 2389.47 1500.38 2393.03 Q1498.58 2396.57 1498.58 2403.7 Q1498.58 2410.81 1500.38 2414.37 Q1502.21 2417.92 1505.82 2417.92 Q1509.45 2417.92 1511.26 2414.37 Q1513.09 2410.81 1513.09 2403.7 Q1513.09 2396.57 1511.26 2393.03 Q1509.45 2389.47 1505.82 2389.47 M1505.82 2385.76 Q1511.63 2385.76 1514.69 2390.37 Q1517.76 2394.95 1517.76 2403.7 Q1517.76 2412.43 1514.69 2417.04 Q1511.63 2421.62 1505.82 2421.62 Q1500.01 2421.62 1496.93 2417.04 Q1493.88 2412.43 1493.88 2403.7 Q1493.88 2394.95 1496.93 2390.37 Q1500.01 2385.76 1505.82 2385.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1476.24 2255.7 Q1473.09 2255.7 1471.24 2257.85 Q1469.41 2260 1469.41 2263.75 Q1469.41 2267.48 1471.24 2269.66 Q1473.09 2271.81 1476.24 2271.81 Q1479.39 2271.81 1481.21 2269.66 Q1483.07 2267.48 1483.07 2263.75 Q1483.07 2260 1481.21 2257.85 Q1479.39 2255.7 1476.24 2255.7 M1485.52 2241.05 L1485.52 2245.3 Q1483.76 2244.47 1481.95 2244.03 Q1480.17 2243.59 1478.41 2243.59 Q1473.78 2243.59 1471.33 2246.72 Q1468.9 2249.84 1468.55 2256.16 Q1469.92 2254.15 1471.98 2253.08 Q1474.04 2251.99 1476.52 2251.99 Q1481.72 2251.99 1484.73 2255.17 Q1487.77 2258.31 1487.77 2263.75 Q1487.77 2269.08 1484.62 2272.29 Q1481.47 2275.51 1476.24 2275.51 Q1470.24 2275.51 1467.07 2270.93 Q1463.9 2266.32 1463.9 2257.6 Q1463.9 2249.4 1467.79 2244.54 Q1471.68 2239.66 1478.23 2239.66 Q1479.99 2239.66 1481.77 2240 Q1483.58 2240.35 1485.52 2241.05 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1505.82 2243.36 Q1502.21 2243.36 1500.38 2246.92 Q1498.58 2250.47 1498.58 2257.6 Q1498.58 2264.7 1500.38 2268.27 Q1502.21 2271.81 1505.82 2271.81 Q1509.45 2271.81 1511.26 2268.27 Q1513.09 2264.7 1513.09 2257.6 Q1513.09 2250.47 1511.26 2246.92 Q1509.45 2243.36 1505.82 2243.36 M1505.82 2239.66 Q1511.63 2239.66 1514.69 2244.26 Q1517.76 2248.85 1517.76 2257.6 Q1517.76 2266.32 1514.69 2270.93 Q1511.63 2275.51 1505.82 2275.51 Q1500.01 2275.51 1496.93 2270.93 Q1493.88 2266.32 1493.88 2257.6 Q1493.88 2248.85 1496.93 2244.26 Q1500.01 2239.66 1505.82 2239.66 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1475.66 2112.32 Q1472.33 2112.32 1470.4 2114.1 Q1468.51 2115.89 1468.51 2119.01 Q1468.51 2122.14 1470.4 2123.92 Q1472.33 2125.7 1475.66 2125.7 Q1478.99 2125.7 1480.91 2123.92 Q1482.83 2122.11 1482.83 2119.01 Q1482.83 2115.89 1480.91 2114.1 Q1479.02 2112.32 1475.66 2112.32 M1470.98 2110.33 Q1467.97 2109.59 1466.28 2107.53 Q1464.62 2105.47 1464.62 2102.51 Q1464.62 2098.36 1467.56 2095.96 Q1470.52 2093.55 1475.66 2093.55 Q1480.82 2093.55 1483.76 2095.96 Q1486.7 2098.36 1486.7 2102.51 Q1486.7 2105.47 1485.01 2107.53 Q1483.34 2109.59 1480.36 2110.33 Q1483.74 2111.12 1485.61 2113.41 Q1487.51 2115.7 1487.51 2119.01 Q1487.51 2124.04 1484.43 2126.72 Q1481.38 2129.41 1475.66 2129.41 Q1469.94 2129.41 1466.86 2126.72 Q1463.81 2124.04 1463.81 2119.01 Q1463.81 2115.7 1465.7 2113.41 Q1467.6 2111.12 1470.98 2110.33 M1469.27 2102.95 Q1469.27 2105.63 1470.94 2107.14 Q1472.63 2108.64 1475.66 2108.64 Q1478.67 2108.64 1480.36 2107.14 Q1482.07 2105.63 1482.07 2102.95 Q1482.07 2100.26 1480.36 2098.76 Q1478.67 2097.25 1475.66 2097.25 Q1472.63 2097.25 1470.94 2098.76 Q1469.27 2100.26 1469.27 2102.95 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1505.82 2097.25 Q1502.21 2097.25 1500.38 2100.82 Q1498.58 2104.36 1498.58 2111.49 Q1498.58 2118.6 1500.38 2122.16 Q1502.21 2125.7 1505.82 2125.7 Q1509.45 2125.7 1511.26 2122.16 Q1513.09 2118.6 1513.09 2111.49 Q1513.09 2104.36 1511.26 2100.82 Q1509.45 2097.25 1505.82 2097.25 M1505.82 2093.55 Q1511.63 2093.55 1514.69 2098.16 Q1517.76 2102.74 1517.76 2111.49 Q1517.76 2120.22 1514.69 2124.82 Q1511.63 2129.41 1505.82 2129.41 Q1500.01 2129.41 1496.93 2124.82 Q1493.88 2120.22 1493.88 2111.49 Q1493.88 2102.74 1496.93 2098.16 Q1500.01 2093.55 1505.82 2093.55 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1893.19 1886.82 L1882.1 1916.92 L1904.33 1916.92 L1893.19 1886.82 M1888.58 1878.76 L1897.85 1878.76 L1920.9 1939.24 L1912.4 1939.24 L1906.89 1923.73 L1879.62 1923.73 L1874.11 1939.24 L1865.49 1939.24 L1888.58 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1926.17 1893.87 L1961.57 1893.87 L1961.57 1900.68 L1933.54 1933.29 L1961.57 1933.29 L1961.57 1939.24 L1925.16 1939.24 L1925.16 1932.44 L1953.19 1899.83 L1926.17 1899.83 L1926.17 1893.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M1972.96 1893.87 L1980.41 1893.87 L1980.41 1939.24 L1972.96 1939.24 L1972.96 1893.87 M1972.96 1876.21 L1980.41 1876.21 L1980.41 1885.65 L1972.96 1885.65 L1972.96 1876.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2031.33 1902.58 Q2034.13 1897.56 2038.01 1895.17 Q2041.9 1892.78 2047.17 1892.78 Q2054.26 1892.78 2058.11 1897.76 Q2061.96 1902.7 2061.96 1911.86 L2061.96 1939.24 L2054.46 1939.24 L2054.46 1912.1 Q2054.46 1905.58 2052.15 1902.42 Q2049.84 1899.26 2045.1 1899.26 Q2039.31 1899.26 2035.95 1903.11 Q2032.59 1906.96 2032.59 1913.6 L2032.59 1939.24 L2025.09 1939.24 L2025.09 1912.1 Q2025.09 1905.54 2022.78 1902.42 Q2020.47 1899.26 2015.65 1899.26 Q2009.94 1899.26 2006.58 1903.15 Q2003.22 1907 2003.22 1913.6 L2003.22 1939.24 L1995.72 1939.24 L1995.72 1893.87 L2003.22 1893.87 L2003.22 1900.92 Q2005.77 1896.75 2009.33 1894.76 Q2012.9 1892.78 2017.8 1892.78 Q2022.74 1892.78 2026.19 1895.29 Q2029.67 1897.8 2031.33 1902.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2076.05 1921.34 L2076.05 1893.87 L2083.51 1893.87 L2083.51 1921.05 Q2083.51 1927.5 2086.02 1930.74 Q2088.53 1933.94 2093.55 1933.94 Q2099.59 1933.94 2103.07 1930.09 Q2106.6 1926.24 2106.6 1919.6 L2106.6 1893.87 L2114.05 1893.87 L2114.05 1939.24 L2106.6 1939.24 L2106.6 1932.28 Q2103.88 1936.41 2100.28 1938.43 Q2096.71 1940.42 2091.97 1940.42 Q2084.15 1940.42 2080.1 1935.56 Q2076.05 1930.7 2076.05 1921.34 M2094.81 1892.78 L2094.81 1892.78 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2136.78 1880.99 L2136.78 1893.87 L2152.13 1893.87 L2152.13 1899.67 L2136.78 1899.67 L2136.78 1924.29 Q2136.78 1929.84 2138.27 1931.42 Q2139.81 1933 2144.47 1933 L2152.13 1933 L2152.13 1939.24 L2144.47 1939.24 Q2135.84 1939.24 2132.56 1936.04 Q2129.28 1932.8 2129.28 1924.29 L2129.28 1899.67 L2123.81 1899.67 L2123.81 1893.87 L2129.28 1893.87 L2129.28 1880.99 L2136.78 1880.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2199.65 1911.86 L2199.65 1939.24 L2192.19 1939.24 L2192.19 1912.1 Q2192.19 1905.66 2189.68 1902.46 Q2187.17 1899.26 2182.15 1899.26 Q2176.11 1899.26 2172.63 1903.11 Q2169.14 1906.96 2169.14 1913.6 L2169.14 1939.24 L2161.65 1939.24 L2161.65 1876.21 L2169.14 1876.21 L2169.14 1900.92 Q2171.82 1896.83 2175.42 1894.8 Q2179.07 1892.78 2183.81 1892.78 Q2191.62 1892.78 2195.64 1897.64 Q2199.65 1902.46 2199.65 1911.86 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2258.79 1876.29 Q2253.36 1885.61 2250.73 1894.72 Q2248.09 1903.84 2248.09 1913.2 Q2248.09 1922.55 2250.73 1931.75 Q2253.4 1940.9 2258.79 1950.18 L2252.31 1950.18 Q2246.23 1940.66 2243.19 1931.46 Q2240.2 1922.27 2240.2 1913.2 Q2240.2 1904.16 2243.19 1895.01 Q2246.19 1885.85 2252.31 1876.29 L2258.79 1876.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2303.11 1900.76 L2303.11 1876.21 L2310.56 1876.21 L2310.56 1939.24 L2303.11 1939.24 L2303.11 1932.44 Q2300.76 1936.49 2297.15 1938.47 Q2293.59 1940.42 2288.56 1940.42 Q2280.34 1940.42 2275.15 1933.85 Q2270.01 1927.29 2270.01 1916.6 Q2270.01 1905.9 2275.15 1899.34 Q2280.34 1892.78 2288.56 1892.78 Q2293.59 1892.78 2297.15 1894.76 Q2300.76 1896.71 2303.11 1900.76 M2277.71 1916.6 Q2277.71 1924.82 2281.07 1929.52 Q2284.47 1934.18 2290.39 1934.18 Q2296.3 1934.18 2299.7 1929.52 Q2303.11 1924.82 2303.11 1916.6 Q2303.11 1908.37 2299.7 1903.72 Q2296.3 1899.02 2290.39 1899.02 Q2284.47 1899.02 2281.07 1903.72 Q2277.71 1908.37 2277.71 1916.6 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2364.72 1914.69 L2364.72 1918.34 L2330.45 1918.34 Q2330.94 1926.04 2335.07 1930.09 Q2339.24 1934.1 2346.65 1934.1 Q2350.95 1934.1 2354.96 1933.04 Q2359.01 1931.99 2362.98 1929.89 L2362.98 1936.93 Q2358.97 1938.64 2354.75 1939.53 Q2350.54 1940.42 2346.21 1940.42 Q2335.35 1940.42 2328.99 1934.1 Q2322.67 1927.78 2322.67 1917 Q2322.67 1905.86 2328.67 1899.34 Q2334.7 1892.78 2344.91 1892.78 Q2354.07 1892.78 2359.37 1898.69 Q2364.72 1904.57 2364.72 1914.69 M2357.27 1912.51 Q2357.19 1906.39 2353.82 1902.74 Q2350.5 1899.1 2344.99 1899.1 Q2338.75 1899.1 2334.99 1902.62 Q2331.26 1906.15 2330.69 1912.55 L2357.27 1912.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2406.81 1916.03 Q2406.81 1907.93 2403.45 1903.47 Q2400.12 1899.02 2394.09 1899.02 Q2388.09 1899.02 2384.73 1903.47 Q2381.41 1907.93 2381.41 1916.03 Q2381.41 1924.09 2384.73 1928.55 Q2388.09 1933 2394.09 1933 Q2400.12 1933 2403.45 1928.55 Q2406.81 1924.09 2406.81 1916.03 M2414.26 1933.61 Q2414.26 1945.2 2409.12 1950.83 Q2403.97 1956.5 2393.36 1956.5 Q2389.43 1956.5 2385.95 1955.89 Q2382.46 1955.32 2379.18 1954.11 L2379.18 1946.86 Q2382.46 1948.64 2385.66 1949.49 Q2388.86 1950.34 2392.19 1950.34 Q2399.52 1950.34 2403.16 1946.49 Q2406.81 1942.69 2406.81 1934.95 L2406.81 1931.26 Q2404.5 1935.27 2400.89 1937.26 Q2397.29 1939.24 2392.27 1939.24 Q2383.92 1939.24 2378.82 1932.88 Q2373.71 1926.52 2373.71 1916.03 Q2373.71 1905.5 2378.82 1899.14 Q2383.92 1892.78 2392.27 1892.78 Q2397.29 1892.78 2400.89 1894.76 Q2404.5 1896.75 2406.81 1900.76 L2406.81 1893.87 L2414.26 1893.87 L2414.26 1933.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip630)" d="M2428.44 1876.29 L2434.92 1876.29 Q2441 1885.85 2444 1895.01 Q2447.03 1904.16 2447.03 1913.2 Q2447.03 1922.27 2444 1931.46 Q2441 1940.66 2434.92 1950.18 L2428.44 1950.18 Q2433.83 1940.9 2436.46 1931.75 Q2439.14 1922.55 2439.14 1913.2 Q2439.14 1903.84 2436.46 1894.72 Q2433.83 1885.61 2428.44 1876.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip637)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="1593.53,2038.4 1595.77,2038.52 1598.01,2038.64 1600.25,2038.77 1602.5,2038.92 1604.74,2039.07 1606.98,2039.24 1609.22,2039.42 1611.46,2039.61 1613.71,2039.82 1615.95,2040.04 1618.19,2040.28 1620.43,2040.54 1622.67,2040.82 1624.92,2041.12 1627.16,2041.44 1629.4,2041.78 1631.64,2042.15 1633.88,2042.55 1636.13,2042.97 1638.37,2043.43 1640.61,2043.92 1642.85,2044.43 1645.09,2044.99 1647.34,2045.58 1649.58,2046.21 1651.82,2046.87 1654.06,2047.58 1656.3,2048.33 1658.55,2049.13 1660.79,2049.96 1663.03,2050.84 1665.27,2051.77 1667.51,2052.74 1669.76,2053.76 1672,2054.82 1674.24,2055.93 1676.48,2057.08 1678.72,2058.27 1680.97,2059.49 1683.21,2060.76 1685.45,2062.06 1687.69,2063.39 1689.93,2064.75 1692.17,2066.13 1694.42,2067.53 1696.66,2068.95 1698.9,2070.38 1701.14,2071.82 1703.38,2073.26 1705.63,2074.7 1707.87,2076.13 1710.11,2077.56 1712.35,2078.97 1714.59,2080.37 1716.84,2081.75 1719.08,2083.11 1721.32,2084.44 1723.56,2085.75 1725.8,2087.03 1728.05,2088.27 1730.29,2089.49 1732.53,2090.67 1734.77,2091.82 1737.01,2092.94 1739.26,2094.03 1741.5,2095.08 1743.74,2096.09 1745.98,2097.08 1748.22,2098.03 1750.47,2098.95 1752.71,2099.85 1754.95,2100.71 1757.19,2101.54 1759.43,2102.35 1761.68,2103.13 1763.92,2103.89 1766.16,2104.62 1768.4,2105.33 1770.64,2106.02 1772.89,2106.69 1775.13,2107.34 1777.37,2107.97 1779.61,2108.59 1781.85,2109.19 1784.1,2109.77 1786.34,2110.35 1788.58,2110.91 1790.82,2111.46 1793.06,2112 1795.31,2112.53 1797.55,2113.06 1799.79,2113.57 1802.03,2114.09 1804.27,2114.59 1806.51,2115.1 1808.76,2115.6 1811,2116.1 1813.24,2116.6 1815.48,2117.09 1817.72,2117.59 1819.97,2118.1 1822.21,2118.6 1824.45,2119.11 1826.69,2119.63 1828.93,2120.15 1831.18,2120.68 1833.42,2121.22 1835.66,2121.77 1837.9,2122.33 1840.14,2122.9 1842.39,2123.49 1844.63,2124.09 1846.87,2124.71 1849.11,2125.35 1851.35,2126 1853.6,2126.67 1855.84,2127.37 1858.08,2128.09 1860.32,2128.83 1862.56,2129.6 1864.81,2130.4 1867.05,2131.22 1869.29,2132.08 1871.53,2132.96 1873.77,2133.88 1876.02,2134.82 1878.26,2135.8 1880.5,2136.82 1882.74,2137.86 1884.98,2138.94 1887.23,2140.05 1889.47,2141.2 1891.71,2142.38 1893.95,2143.59 1896.19,2144.82 1898.44,2146.09 1900.68,2147.38 1902.92,2148.69 1905.16,2150.02 1907.4,2151.37 1909.64,2152.74 1911.89,2154.12 1914.13,2155.5 1916.37,2156.89 1918.61,2158.28 1920.85,2159.67 1923.1,2161.06 1925.34,2162.43 1927.58,2163.8 1929.82,2165.15 1932.06,2166.49 1934.31,2167.8 1936.55,2169.1 1938.79,2170.37 1941.03,2171.63 1943.27,2172.85 1945.52,2174.06 1947.76,2175.23 1950,2176.38 1952.24,2177.51 1954.48,2178.61 1956.73,2179.69 1958.97,2180.74 1961.21,2181.77 1963.45,2182.78 1965.69,2183.77 1967.94,2184.73 1970.18,2185.68 1972.42,2186.61 1974.66,2187.53 1976.9,2188.43 1979.15,2189.32 1981.39,2190.2 1983.63,2191.07 1985.87,2191.93 1988.11,2192.78 1990.36,2193.63 1992.6,2194.48 1994.84,2195.33 1997.08,2196.18 1999.32,2197.03 2001.57,2197.88 2003.81,2198.74 2006.05,2199.61 2008.29,2200.48 2010.53,2201.37 2012.78,2202.27 2015.02,2203.19 2017.26,2204.12 2019.5,2205.06 2021.74,2206.03 2023.98,2207.02 2026.23,2208.02 2028.47,2209.05 2030.71,2210.11 2032.95,2211.18 2035.19,2212.29 2037.44,2213.41 2039.68,2214.57 2041.92,2215.75 2044.16,2216.95 2046.4,2218.18 2048.65,2219.43 2050.89,2220.71 2053.13,2222.01 2055.37,2223.33 2057.61,2224.66 2059.86,2226.02 2062.1,2227.39 2064.34,2228.77 2066.58,2230.15 2068.82,2231.55 2071.07,2232.95 2073.31,2234.35 2075.55,2235.75 2077.79,2237.14 2080.03,2238.53 2082.28,2239.91 2084.52,2241.28 2086.76,2242.63 2089,2243.97 2091.24,2245.3 2093.49,2246.61 2095.73,2247.9 2097.97,2249.17 2100.21,2250.43 2102.45,2251.67 2104.7,2252.88 2106.94,2254.09 2109.18,2255.27 2111.42,2256.44 2113.66,2257.59 2115.91,2258.72 2118.15,2259.85 2120.39,2260.96 2122.63,2262.06 2124.87,2263.15 2127.11,2264.24 2129.36,2265.32 2131.6,2266.39 2133.84,2267.46 2136.08,2268.54 2138.32,2269.61 2140.57,2270.69 2142.81,2271.77 2145.05,2272.86 2147.29,2273.95 2149.53,2275.06 2151.78,2276.17 2154.02,2277.3 2156.26,2278.44 2158.5,2279.6 2160.74,2280.78 2162.99,2281.97 2165.23,2283.18 2167.47,2284.41 2169.71,2285.65 2171.95,2286.92 2174.2,2288.21 2176.44,2289.51 2178.68,2290.83 2180.92,2292.18 2183.16,2293.53 2185.41,2294.91 2187.65,2296.3 2189.89,2297.7 2192.13,2299.11 2194.37,2300.53 2196.62,2301.96 2198.86,2303.4 2201.1,2304.84 2203.34,2306.28 2205.58,2307.72 2207.83,2309.16 2210.07,2310.59 2212.31,2312.02 2214.55,2313.44 2216.79,2314.85 2219.04,2316.25 2221.28,2317.64 2223.52,2319.02 2225.76,2320.39 2228,2321.75 2230.25,2323.09 2232.49,2324.43 2234.73,2325.75 2236.97,2327.06 2239.21,2328.36 2241.45,2329.65 2243.7,2330.93 2245.94,2332.21 2248.18,2333.47 2250.42,2334.74 2252.66,2336 2254.91,2337.26 2257.15,2338.51 2259.39,2339.77 2261.63,2341.03 2263.87,2342.3 2266.12,2343.57 2268.36,2344.84 2270.6,2346.13 2272.84,2347.42 2275.08,2348.72 2277.33,2350.03 2279.57,2351.36 2281.81,2352.7 2284.05,2354.05 2286.29,2355.41 2288.54,2356.79 2290.78,2358.18 2293.02,2359.59 2295.26,2361.01 2297.5,2362.44 2299.75,2363.89 2301.99,2365.35 2304.23,2366.82 2306.47,2368.3 2308.71,2369.78 2310.96,2371.28 2313.2,2372.78 2315.44,2374.29 2317.68,2375.79 2319.92,2377.3 2322.17,2378.82 2324.41,2380.33 2326.65,2381.83 2328.89,2383.34 2331.13,2384.84 2333.38,2386.34 2335.62,2387.83 2337.86,2389.31 2340.1,2390.79 2342.34,2392.26 2344.58,2393.72 2346.83,2395.18 2349.07,2396.63 2351.31,2398.08 2353.55,2399.52 2355.79,2400.95 2358.04,2402.38 2360.28,2403.81 2362.52,2405.24 2364.76,2406.66 2367,2408.08 2369.25,2409.51 2371.49,2410.93 2373.73,2412.36 2375.97,2413.8 2378.21,2415.23 2380.46,2416.67 2382.7,2418.12 2384.94,2419.58 2387.18,2421.04 2389.42,2422.51 2391.67,2424 2393.91,2425.49 2396.15,2426.99 2398.39,2428.5 2400.63,2430.02 2402.88,2431.54 2405.12,2433.08 2407.36,2434.63 2409.6,2436.18 2411.84,2437.75 2414.09,2439.32 2416.33,2440.89 2418.57,2442.48 2420.81,2444.06 2423.05,2445.65 2425.3,2447.24 2427.54,2448.84 2429.78,2450.43 2432.02,2452.03 2434.26,2453.62 2436.51,2455.22 2438.75,2456.81 2440.99,2458.4 2443.23,2459.99 2445.47,2461.57 2447.72,2463.15 2449.96,2464.73 2452.2,2466.3 2454.44,2467.87 2456.68,2469.44 2458.92,2471.01 2461.17,2472.57 2463.41,2474.13 2465.65,2475.69 2467.89,2477.25 2470.13,2478.81 2472.38,2480.37 2474.62,2481.93 2476.86,2483.5 2479.1,2485.06 2481.34,2486.63 2483.59,2488.2 2485.83,2489.78 2488.07,2491.36 2490.31,2492.95 2492.55,2494.54 2494.8,2496.14 2497.04,2497.74 2499.28,2499.35 2501.52,2500.96 2503.76,2502.58 2506.01,2504.21 2508.25,2505.84 2510.49,2507.47 2512.73,2509.11 2514.97,2510.76 2517.22,2512.4 2519.46,2514.05 2521.7,2515.71 2523.94,2517.36 2526.18,2519.02 2528.43,2520.67 2530.67,2522.33 2532.91,2523.99 2535.15,2525.64 2537.39,2527.3 2539.64,2528.95 2541.88,2530.6 2544.12,2532.25 2546.36,2533.89 2548.6,2535.53 2550.85,2537.17 2553.09,2538.81 2555.33,2540.44 2557.57,2542.07 2559.81,2543.7 2562.05,2545.33 2564.3,2546.95 2566.54,2548.57 2568.78,2550.2 2571.02,2551.81 2573.26,2553.43 2575.51,2555.05 2577.75,2556.67 2579.99,2558.29 2582.23,2559.91 2584.47,2561.52 2586.72,2563.14 2588.96,2564.76 2591.2,2566.39 2593.44,2568.01 2595.68,2569.63 2597.93,2571.26 2600.17,2572.88 2602.41,2574.51 2604.65,2576.13 2606.89,2577.76 2609.14,2579.38 2611.38,2581.01 2613.62,2582.63 2615.86,2584.25 2618.1,2585.86 2620.35,2587.47 2622.59,2589.08 2624.83,2590.68 2627.07,2592.27 2629.31,2593.86 2631.56,2595.44 2633.8,2597 2636.04,2598.56 2638.28,2600.11 2640.52,2601.65 2642.77,2603.17 2645.01,2604.68 2647.25,2606.18 2649.49,2607.67 2651.73,2609.14 2653.98,2610.59 2656.22,2612.03 2658.46,2613.45 2660.7,2614.86 2662.94,2616.25 2665.19,2617.62 2667.43,2618.97 2669.67,2620.3 2671.91,2621.61 2674.15,2622.9 2676.39,2624.16 2678.64,2625.4 2680.88,2626.61 2683.12,2627.79 2685.36,2628.95 2687.6,2630.06 2689.85,2631.15 2692.09,2632.18 2694.33,2633.18 2696.57,2634.12 2698.81,2635.01 2701.06,2635.83 2703.3,2636.59 2705.54,2637.27 2707.78,2637.86 2710.02,2638.36 2712.27,2638.74 2714.51,2639.01 2716.75,2639.15 2718.99,2639.13 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/fa7e5a76.svg b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/fa7e5a76.svg
new file mode 100644
index 00000000000..daa9b9e13dd
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/fa7e5a76.svg
@@ -0,0 +1,97 @@
+<?xml version="1.0" encoding="utf-8"?>
+<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" width="700" height="700" viewBox="0 0 2800 2800">
+<defs>
+  <clipPath id="clip690">
+    <rect x="0" y="0" width="2800" height="2800"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip690)" d="M0 2800 L2800 2800 L2800 0 L0 0  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip691">
+    <rect x="560" y="280" width="1961" height="1961"/>
+  </clipPath>
+</defs>
+<path clip-path="url(#clip690)" d="M188.677 790.501 L2752.76 790.501 L2752.76 153.712 L188.677 153.712  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip692">
+    <rect x="188" y="153" width="2565" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="261.245,790.501 261.245,153.712 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="864.774,790.501 864.774,153.712 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1468.3,790.501 1468.3,153.712 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2071.83,790.501 2071.83,153.712 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2675.36,790.501 2675.36,153.712 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,790.501 2752.76,790.501 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="261.245,790.501 261.245,771.604 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="864.774,790.501 864.774,771.604 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1468.3,790.501 1468.3,771.604 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2071.83,790.501 2071.83,771.604 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2675.36,790.501 2675.36,771.604 "/>
+<path clip-path="url(#clip690)" d="M261.245 824.62 Q257.634 824.62 255.805 828.185 Q254 831.726 254 838.856 Q254 845.962 255.805 849.527 Q257.634 853.069 261.245 853.069 Q264.879 853.069 266.685 849.527 Q268.514 845.962 268.514 838.856 Q268.514 831.726 266.685 828.185 Q264.879 824.62 261.245 824.62 M261.245 820.916 Q267.055 820.916 270.111 825.523 Q273.189 830.106 273.189 838.856 Q273.189 847.583 270.111 852.189 Q267.055 856.773 261.245 856.773 Q255.435 856.773 252.356 852.189 Q249.301 847.583 249.301 838.856 Q249.301 830.106 252.356 825.523 Q255.435 820.916 261.245 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M824.392 821.541 L842.748 821.541 L842.748 825.476 L828.674 825.476 L828.674 833.949 Q829.693 833.601 830.711 833.439 Q831.73 833.254 832.748 833.254 Q838.535 833.254 841.915 836.425 Q845.295 839.597 845.295 845.013 Q845.295 850.592 841.822 853.694 Q838.35 856.773 832.031 856.773 Q829.855 856.773 827.586 856.402 Q825.341 856.032 822.934 855.291 L822.934 850.592 Q825.017 851.726 827.239 852.282 Q829.461 852.837 831.938 852.837 Q835.943 852.837 838.281 850.731 Q840.619 848.624 840.619 845.013 Q840.619 841.402 838.281 839.296 Q835.943 837.189 831.938 837.189 Q830.063 837.189 828.188 837.606 Q826.336 838.023 824.392 838.902 L824.392 821.541 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M864.507 824.62 Q860.896 824.62 859.068 828.185 Q857.262 831.726 857.262 838.856 Q857.262 845.962 859.068 849.527 Q860.896 853.069 864.507 853.069 Q868.142 853.069 869.947 849.527 Q871.776 845.962 871.776 838.856 Q871.776 831.726 869.947 828.185 Q868.142 824.62 864.507 824.62 M864.507 820.916 Q870.318 820.916 873.373 825.523 Q876.452 830.106 876.452 838.856 Q876.452 847.583 873.373 852.189 Q870.318 856.773 864.507 856.773 Q858.697 856.773 855.619 852.189 Q852.563 847.583 852.563 838.856 Q852.563 830.106 855.619 825.523 Q858.697 820.916 864.507 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M894.669 824.62 Q891.058 824.62 889.23 828.185 Q887.424 831.726 887.424 838.856 Q887.424 845.962 889.23 849.527 Q891.058 853.069 894.669 853.069 Q898.304 853.069 900.109 849.527 Q901.938 845.962 901.938 838.856 Q901.938 831.726 900.109 828.185 Q898.304 824.62 894.669 824.62 M894.669 820.916 Q900.479 820.916 903.535 825.523 Q906.614 830.106 906.614 838.856 Q906.614 847.583 903.535 852.189 Q900.479 856.773 894.669 856.773 Q888.859 856.773 885.78 852.189 Q882.725 847.583 882.725 838.856 Q882.725 830.106 885.78 825.523 Q888.859 820.916 894.669 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1412.83 852.166 L1420.47 852.166 L1420.47 825.8 L1412.16 827.467 L1412.16 823.208 L1420.42 821.541 L1425.1 821.541 L1425.1 852.166 L1432.74 852.166 L1432.74 856.101 L1412.83 856.101 L1412.83 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1452.18 824.62 Q1448.57 824.62 1446.74 828.185 Q1444.93 831.726 1444.93 838.856 Q1444.93 845.962 1446.74 849.527 Q1448.57 853.069 1452.18 853.069 Q1455.81 853.069 1457.62 849.527 Q1459.45 845.962 1459.45 838.856 Q1459.45 831.726 1457.62 828.185 Q1455.81 824.62 1452.18 824.62 M1452.18 820.916 Q1457.99 820.916 1461.05 825.523 Q1464.12 830.106 1464.12 838.856 Q1464.12 847.583 1461.05 852.189 Q1457.99 856.773 1452.18 856.773 Q1446.37 856.773 1443.29 852.189 Q1440.24 847.583 1440.24 838.856 Q1440.24 830.106 1443.29 825.523 Q1446.37 820.916 1452.18 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1482.34 824.62 Q1478.73 824.62 1476.9 828.185 Q1475.1 831.726 1475.1 838.856 Q1475.1 845.962 1476.9 849.527 Q1478.73 853.069 1482.34 853.069 Q1485.98 853.069 1487.78 849.527 Q1489.61 845.962 1489.61 838.856 Q1489.61 831.726 1487.78 828.185 Q1485.98 824.62 1482.34 824.62 M1482.34 820.916 Q1488.15 820.916 1491.21 825.523 Q1494.29 830.106 1494.29 838.856 Q1494.29 847.583 1491.21 852.189 Q1488.15 856.773 1482.34 856.773 Q1476.53 856.773 1473.45 852.189 Q1470.4 847.583 1470.4 838.856 Q1470.4 830.106 1473.45 825.523 Q1476.53 820.916 1482.34 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1512.5 824.62 Q1508.89 824.62 1507.06 828.185 Q1505.26 831.726 1505.26 838.856 Q1505.26 845.962 1507.06 849.527 Q1508.89 853.069 1512.5 853.069 Q1516.14 853.069 1517.94 849.527 Q1519.77 845.962 1519.77 838.856 Q1519.77 831.726 1517.94 828.185 Q1516.14 824.62 1512.5 824.62 M1512.5 820.916 Q1518.31 820.916 1521.37 825.523 Q1524.45 830.106 1524.45 838.856 Q1524.45 847.583 1521.37 852.189 Q1518.31 856.773 1512.5 856.773 Q1506.69 856.773 1503.61 852.189 Q1500.56 847.583 1500.56 838.856 Q1500.56 830.106 1503.61 825.523 Q1506.69 820.916 1512.5 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2016.36 852.166 L2024 852.166 L2024 825.8 L2015.69 827.467 L2015.69 823.208 L2023.95 821.541 L2028.63 821.541 L2028.63 852.166 L2036.26 852.166 L2036.26 856.101 L2016.36 856.101 L2016.36 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2045.75 821.541 L2064.11 821.541 L2064.11 825.476 L2050.04 825.476 L2050.04 833.949 Q2051.06 833.601 2052.07 833.439 Q2053.09 833.254 2054.11 833.254 Q2059.9 833.254 2063.28 836.425 Q2066.66 839.597 2066.66 845.013 Q2066.66 850.592 2063.19 853.694 Q2059.71 856.773 2053.39 856.773 Q2051.22 856.773 2048.95 856.402 Q2046.7 856.032 2044.3 855.291 L2044.3 850.592 Q2046.38 851.726 2048.6 852.282 Q2050.82 852.837 2053.3 852.837 Q2057.31 852.837 2059.64 850.731 Q2061.98 848.624 2061.98 845.013 Q2061.98 841.402 2059.64 839.296 Q2057.31 837.189 2053.3 837.189 Q2051.43 837.189 2049.55 837.606 Q2047.7 838.023 2045.75 838.902 L2045.75 821.541 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2085.87 824.62 Q2082.26 824.62 2080.43 828.185 Q2078.62 831.726 2078.62 838.856 Q2078.62 845.962 2080.43 849.527 Q2082.26 853.069 2085.87 853.069 Q2089.5 853.069 2091.31 849.527 Q2093.14 845.962 2093.14 838.856 Q2093.14 831.726 2091.31 828.185 Q2089.5 824.62 2085.87 824.62 M2085.87 820.916 Q2091.68 820.916 2094.74 825.523 Q2097.81 830.106 2097.81 838.856 Q2097.81 847.583 2094.74 852.189 Q2091.68 856.773 2085.87 856.773 Q2080.06 856.773 2076.98 852.189 Q2073.93 847.583 2073.93 838.856 Q2073.93 830.106 2076.98 825.523 Q2080.06 820.916 2085.87 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2116.03 824.62 Q2112.42 824.62 2110.59 828.185 Q2108.79 831.726 2108.79 838.856 Q2108.79 845.962 2110.59 849.527 Q2112.42 853.069 2116.03 853.069 Q2119.67 853.069 2121.47 849.527 Q2123.3 845.962 2123.3 838.856 Q2123.3 831.726 2121.47 828.185 Q2119.67 824.62 2116.03 824.62 M2116.03 820.916 Q2121.84 820.916 2124.9 825.523 Q2127.98 830.106 2127.98 838.856 Q2127.98 847.583 2124.9 852.189 Q2121.84 856.773 2116.03 856.773 Q2110.22 856.773 2107.14 852.189 Q2104.09 847.583 2104.09 838.856 Q2104.09 830.106 2107.14 825.523 Q2110.22 820.916 2116.03 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2623.97 852.166 L2640.29 852.166 L2640.29 856.101 L2618.35 856.101 L2618.35 852.166 Q2621.01 849.411 2625.59 844.782 Q2630.2 840.129 2631.38 838.786 Q2633.62 836.263 2634.5 834.527 Q2635.41 832.768 2635.41 831.078 Q2635.41 828.324 2633.46 826.587 Q2631.54 824.851 2628.44 824.851 Q2626.24 824.851 2623.79 825.615 Q2621.36 826.379 2618.58 827.93 L2618.58 823.208 Q2621.4 822.074 2623.86 821.495 Q2626.31 820.916 2628.35 820.916 Q2633.72 820.916 2636.91 823.601 Q2640.1 826.287 2640.1 830.777 Q2640.1 832.907 2639.29 834.828 Q2638.51 836.726 2636.4 839.319 Q2635.82 839.99 2632.72 843.208 Q2629.62 846.402 2623.97 852.166 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2660.1 824.62 Q2656.49 824.62 2654.67 828.185 Q2652.86 831.726 2652.86 838.856 Q2652.86 845.962 2654.67 849.527 Q2656.49 853.069 2660.1 853.069 Q2663.74 853.069 2665.54 849.527 Q2667.37 845.962 2667.37 838.856 Q2667.37 831.726 2665.54 828.185 Q2663.74 824.62 2660.1 824.62 M2660.1 820.916 Q2665.92 820.916 2668.97 825.523 Q2672.05 830.106 2672.05 838.856 Q2672.05 847.583 2668.97 852.189 Q2665.92 856.773 2660.1 856.773 Q2654.29 856.773 2651.22 852.189 Q2648.16 847.583 2648.16 838.856 Q2648.16 830.106 2651.22 825.523 Q2654.29 820.916 2660.1 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2690.27 824.62 Q2686.66 824.62 2684.83 828.185 Q2683.02 831.726 2683.02 838.856 Q2683.02 845.962 2684.83 849.527 Q2686.66 853.069 2690.27 853.069 Q2693.9 853.069 2695.71 849.527 Q2697.54 845.962 2697.54 838.856 Q2697.54 831.726 2695.71 828.185 Q2693.9 824.62 2690.27 824.62 M2690.27 820.916 Q2696.08 820.916 2699.13 825.523 Q2702.21 830.106 2702.21 838.856 Q2702.21 847.583 2699.13 852.189 Q2696.08 856.773 2690.27 856.773 Q2684.46 856.773 2681.38 852.189 Q2678.32 847.583 2678.32 838.856 Q2678.32 830.106 2681.38 825.523 Q2684.46 820.916 2690.27 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2720.43 824.62 Q2716.82 824.62 2714.99 828.185 Q2713.18 831.726 2713.18 838.856 Q2713.18 845.962 2714.99 849.527 Q2716.82 853.069 2720.43 853.069 Q2724.06 853.069 2725.87 849.527 Q2727.7 845.962 2727.7 838.856 Q2727.7 831.726 2725.87 828.185 Q2724.06 824.62 2720.43 824.62 M2720.43 820.916 Q2726.24 820.916 2729.29 825.523 Q2732.37 830.106 2732.37 838.856 Q2732.37 847.583 2729.29 852.189 Q2726.24 856.773 2720.43 856.773 Q2714.62 856.773 2711.54 852.189 Q2708.48 847.583 2708.48 838.856 Q2708.48 830.106 2711.54 825.523 Q2714.62 820.916 2720.43 820.916 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,786.533 2752.76,786.533 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,659.948 2752.76,659.948 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,533.363 2752.76,533.363 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,406.778 2752.76,406.778 "/>
+<polyline clip-path="url(#clip692)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,280.193 2752.76,280.193 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,790.501 188.677,153.712 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,786.533 207.574,786.533 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,659.948 207.574,659.948 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,533.363 207.574,533.363 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,406.778 207.574,406.778 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,280.193 207.574,280.193 "/>
+<path clip-path="url(#clip690)" d="M49.9875 799.878 L57.6263 799.878 L57.6263 773.513 L49.3162 775.179 L49.3162 770.92 L57.58 769.253 L62.2559 769.253 L62.2559 799.878 L69.8948 799.878 L69.8948 803.813 L49.9875 803.813 L49.9875 799.878 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M89.9178 784.67 Q86.7697 784.67 84.9179 786.823 Q83.0892 788.976 83.0892 792.726 Q83.0892 796.452 84.9179 798.628 Q86.7697 800.781 89.9178 800.781 Q93.066 800.781 94.8947 798.628 Q96.7465 796.452 96.7465 792.726 Q96.7465 788.976 94.8947 786.823 Q93.066 784.67 89.9178 784.67 M99.2002 770.017 L99.2002 774.277 Q97.4409 773.443 95.6354 773.003 Q93.853 772.564 92.0937 772.564 Q87.4641 772.564 85.0105 775.689 Q82.5799 778.814 82.2327 785.133 Q83.5984 783.119 85.6586 782.054 Q87.7188 780.966 90.1956 780.966 Q95.4039 780.966 98.4132 784.138 Q101.446 787.286 101.446 792.726 Q101.446 798.05 98.2974 801.267 Q95.1493 804.485 89.9178 804.485 Q83.9225 804.485 80.7512 799.901 Q77.5799 795.295 77.5799 786.568 Q77.5799 778.374 81.4688 773.513 Q85.3577 768.628 91.9086 768.628 Q93.6678 768.628 95.4502 768.976 Q97.2558 769.323 99.2002 770.017 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M109.501 797.934 L114.385 797.934 L114.385 803.813 L109.501 803.813 L109.501 797.934 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M135.149 784.67 Q132.001 784.67 130.149 786.823 Q128.32 788.976 128.32 792.726 Q128.32 796.452 130.149 798.628 Q132.001 800.781 135.149 800.781 Q138.297 800.781 140.126 798.628 Q141.978 796.452 141.978 792.726 Q141.978 788.976 140.126 786.823 Q138.297 784.67 135.149 784.67 M144.431 770.017 L144.431 774.277 Q142.672 773.443 140.867 773.003 Q139.084 772.564 137.325 772.564 Q132.695 772.564 130.242 775.689 Q127.811 778.814 127.464 785.133 Q128.83 783.119 130.89 782.054 Q132.95 780.966 135.427 780.966 Q140.635 780.966 143.644 784.138 Q146.677 787.286 146.677 792.726 Q146.677 798.05 143.529 801.267 Q140.381 804.485 135.149 804.485 Q129.154 804.485 125.982 799.901 Q122.811 795.295 122.811 786.568 Q122.811 778.374 126.7 773.513 Q130.589 768.628 137.14 768.628 Q138.899 768.628 140.681 768.976 Q142.487 769.323 144.431 770.017 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M50.2421 673.293 L57.881 673.293 L57.881 646.928 L49.5708 648.594 L49.5708 644.335 L57.8347 642.668 L62.5106 642.668 L62.5106 673.293 L70.1494 673.293 L70.1494 677.228 L50.2421 677.228 L50.2421 673.293 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M90.1725 658.085 Q87.0243 658.085 85.1725 660.238 Q83.3438 662.39 83.3438 666.14 Q83.3438 669.867 85.1725 672.043 Q87.0243 674.196 90.1725 674.196 Q93.3206 674.196 95.1493 672.043 Q97.0011 669.867 97.0011 666.14 Q97.0011 662.39 95.1493 660.238 Q93.3206 658.085 90.1725 658.085 M99.4548 643.432 L99.4548 647.691 Q97.6956 646.858 95.89 646.418 Q94.1076 645.978 92.3484 645.978 Q87.7188 645.978 85.2651 649.103 Q82.8345 652.228 82.4873 658.548 Q83.8531 656.534 85.9132 655.469 Q87.9734 654.381 90.4502 654.381 Q95.6585 654.381 98.6678 657.553 Q101.7 660.701 101.7 666.14 Q101.7 671.464 98.552 674.682 Q95.4039 677.9 90.1725 677.9 Q84.1771 677.9 81.0058 673.316 Q77.8346 668.71 77.8346 659.983 Q77.8346 651.789 81.7234 646.928 Q85.6123 642.043 92.1632 642.043 Q93.9224 642.043 95.7048 642.391 Q97.5104 642.738 99.4548 643.432 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M109.756 671.349 L114.64 671.349 L114.64 677.228 L109.756 677.228 L109.756 671.349 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.825 660.816 Q131.492 660.816 129.57 662.599 Q127.672 664.381 127.672 667.506 Q127.672 670.631 129.57 672.414 Q131.492 674.196 134.825 674.196 Q138.158 674.196 140.08 672.414 Q142.001 670.608 142.001 667.506 Q142.001 664.381 140.08 662.599 Q138.181 660.816 134.825 660.816 M130.149 658.826 Q127.14 658.085 125.45 656.025 Q123.783 653.965 123.783 651.002 Q123.783 646.858 126.723 644.451 Q129.686 642.043 134.825 642.043 Q139.987 642.043 142.927 644.451 Q145.867 646.858 145.867 651.002 Q145.867 653.965 144.177 656.025 Q142.51 658.085 139.524 658.826 Q142.904 659.613 144.779 661.904 Q146.677 664.196 146.677 667.506 Q146.677 672.529 143.598 675.214 Q140.543 677.9 134.825 677.9 Q129.107 677.9 126.029 675.214 Q122.973 672.529 122.973 667.506 Q122.973 664.196 124.871 661.904 Q126.769 659.613 130.149 658.826 M128.436 651.441 Q128.436 654.127 130.103 655.631 Q131.793 657.136 134.825 657.136 Q137.834 657.136 139.524 655.631 Q141.237 654.127 141.237 651.441 Q141.237 648.756 139.524 647.252 Q137.834 645.747 134.825 645.747 Q131.793 645.747 130.103 647.252 Q128.436 648.756 128.436 651.441 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M50.1495 546.708 L57.7884 546.708 L57.7884 520.342 L49.4782 522.009 L49.4782 517.75 L57.7421 516.083 L62.418 516.083 L62.418 546.708 L70.0568 546.708 L70.0568 550.643 L50.1495 550.643 L50.1495 546.708 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M78.3207 516.083 L100.543 516.083 L100.543 518.074 L87.9965 550.643 L83.1123 550.643 L94.9178 520.018 L78.3207 520.018 L78.3207 516.083 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M109.663 544.764 L114.547 544.764 L114.547 550.643 L109.663 550.643 L109.663 544.764 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 519.162 Q131.121 519.162 129.293 522.727 Q127.487 526.268 127.487 533.398 Q127.487 540.504 129.293 544.069 Q131.121 547.611 134.732 547.611 Q138.367 547.611 140.172 544.069 Q142.001 540.504 142.001 533.398 Q142.001 526.268 140.172 522.727 Q138.367 519.162 134.732 519.162 M134.732 515.458 Q140.543 515.458 143.598 520.065 Q146.677 524.648 146.677 533.398 Q146.677 542.125 143.598 546.731 Q140.543 551.315 134.732 551.315 Q128.922 551.315 125.844 546.731 Q122.788 542.125 122.788 533.398 Q122.788 524.648 125.844 520.065 Q128.922 515.458 134.732 515.458 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M51.7467 420.123 L59.3856 420.123 L59.3856 393.757 L51.0754 395.424 L51.0754 391.165 L59.3393 389.498 L64.0152 389.498 L64.0152 420.123 L71.654 420.123 L71.654 424.058 L51.7467 424.058 L51.7467 420.123 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M79.9179 389.498 L102.14 389.498 L102.14 391.489 L89.5938 424.058 L84.7095 424.058 L96.515 393.433 L79.9179 393.433 L79.9179 389.498 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M111.26 418.179 L116.145 418.179 L116.145 424.058 L111.26 424.058 L111.26 418.179 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M130.357 420.123 L146.677 420.123 L146.677 424.058 L124.732 424.058 L124.732 420.123 Q127.394 417.368 131.978 412.739 Q136.584 408.086 137.765 406.743 Q140.01 404.22 140.89 402.484 Q141.793 400.725 141.793 399.035 Q141.793 396.28 139.848 394.544 Q137.927 392.808 134.825 392.808 Q132.626 392.808 130.172 393.572 Q127.742 394.336 124.964 395.887 L124.964 391.165 Q127.788 390.031 130.242 389.452 Q132.695 388.873 134.732 388.873 Q140.103 388.873 143.297 391.558 Q146.492 394.243 146.492 398.734 Q146.492 400.864 145.681 402.785 Q144.894 404.683 142.788 407.276 Q142.209 407.947 139.107 411.165 Q136.006 414.359 130.357 420.123 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M49.6634 293.538 L57.3023 293.538 L57.3023 267.172 L48.9921 268.839 L48.9921 264.58 L57.256 262.913 L61.9319 262.913 L61.9319 293.538 L69.5707 293.538 L69.5707 297.473 L49.6634 297.473 L49.6634 293.538 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M77.8346 262.913 L100.057 262.913 L100.057 264.904 L87.5104 297.473 L82.6262 297.473 L94.4317 266.848 L77.8346 266.848 L77.8346 262.913 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M109.177 291.593 L114.061 291.593 L114.061 297.473 L109.177 297.473 L109.177 291.593 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M137.094 266.987 L125.288 285.436 L137.094 285.436 L137.094 266.987 M135.867 262.913 L141.746 262.913 L141.746 285.436 L146.677 285.436 L146.677 289.325 L141.746 289.325 L141.746 297.473 L137.094 297.473 L137.094 289.325 L121.492 289.325 L121.492 284.811 L135.867 262.913 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1061.76 20.1573 L1050.66 50.2555 L1072.9 50.2555 L1061.76 20.1573 M1057.14 12.096 L1066.42 12.096 L1089.47 72.576 L1080.96 72.576 L1075.45 57.061 L1048.19 57.061 L1042.68 72.576 L1034.05 72.576 L1057.14 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1135.69 45.1919 L1135.69 72.576 L1128.23 72.576 L1128.23 45.4349 Q1128.23 38.994 1125.72 35.7938 Q1123.21 32.5936 1118.19 32.5936 Q1112.15 32.5936 1108.67 36.4419 Q1105.18 40.2903 1105.18 46.9338 L1105.18 72.576 L1097.69 72.576 L1097.69 27.2059 L1105.18 27.2059 L1105.18 34.2544 Q1107.86 30.163 1111.46 28.1376 Q1115.11 26.1121 1119.85 26.1121 Q1127.67 26.1121 1131.68 30.9732 Q1135.69 35.7938 1135.69 45.1919 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1180.41 49.3643 Q1180.41 41.2625 1177.05 36.8065 Q1173.72 32.3505 1167.69 32.3505 Q1161.69 32.3505 1158.33 36.8065 Q1155.01 41.2625 1155.01 49.3643 Q1155.01 57.4256 1158.33 61.8816 Q1161.69 66.3376 1167.69 66.3376 Q1173.72 66.3376 1177.05 61.8816 Q1180.41 57.4256 1180.41 49.3643 M1187.86 66.9452 Q1187.86 78.5308 1182.72 84.1616 Q1177.57 89.8329 1166.96 89.8329 Q1163.03 89.8329 1159.55 89.2252 Q1156.06 88.6581 1152.78 87.4428 L1152.78 80.1917 Q1156.06 81.9741 1159.26 82.8248 Q1162.46 83.6755 1165.78 83.6755 Q1173.12 83.6755 1176.76 79.8271 Q1180.41 76.0193 1180.41 68.282 L1180.41 64.5957 Q1178.1 68.6061 1174.49 70.5911 Q1170.89 72.576 1165.87 72.576 Q1157.52 72.576 1152.42 66.2161 Q1147.31 59.8562 1147.31 49.3643 Q1147.31 38.832 1152.42 32.472 Q1157.52 26.1121 1165.87 26.1121 Q1170.89 26.1121 1174.49 28.0971 Q1178.1 30.082 1180.41 34.0924 L1180.41 27.2059 L1187.86 27.2059 L1187.86 66.9452 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1203.21 9.54393 L1210.67 9.54393 L1210.67 72.576 L1203.21 72.576 L1203.21 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1265.07 48.0275 L1265.07 51.6733 L1230.8 51.6733 Q1231.29 59.3701 1235.42 63.421 Q1239.59 67.4314 1247.01 67.4314 Q1251.3 67.4314 1255.31 66.3781 Q1259.36 65.3249 1263.33 63.2184 L1263.33 70.267 Q1259.32 71.9684 1255.11 72.8596 Q1250.89 73.7508 1246.56 73.7508 Q1235.7 73.7508 1229.34 67.4314 Q1223.02 61.1119 1223.02 50.3365 Q1223.02 39.1965 1229.02 32.6746 Q1235.05 26.1121 1245.26 26.1121 Q1254.42 26.1121 1259.73 32.0264 Q1265.07 37.9003 1265.07 48.0275 M1257.62 45.84 Q1257.54 39.7232 1254.18 36.0774 Q1250.85 32.4315 1245.34 32.4315 Q1239.11 32.4315 1235.34 35.9558 Q1231.61 39.4801 1231.04 45.8805 L1257.62 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1321.26 32.4315 Q1315.26 32.4315 1311.78 37.1306 Q1308.3 41.7891 1308.3 49.9314 Q1308.3 58.0738 1311.74 62.7728 Q1315.22 67.4314 1321.26 67.4314 Q1327.21 67.4314 1330.7 62.7323 Q1334.18 58.0333 1334.18 49.9314 Q1334.18 41.8701 1330.7 37.1711 Q1327.21 32.4315 1321.26 32.4315 M1321.26 26.1121 Q1330.98 26.1121 1336.53 32.4315 Q1342.08 38.7509 1342.08 49.9314 Q1342.08 61.0714 1336.53 67.4314 Q1330.98 73.7508 1321.26 73.7508 Q1311.5 73.7508 1305.95 67.4314 Q1300.44 61.0714 1300.44 49.9314 Q1300.44 38.7509 1305.95 32.4315 Q1311.5 26.1121 1321.26 26.1121 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1377.4 9.54393 L1377.4 15.7418 L1370.27 15.7418 Q1366.26 15.7418 1364.68 17.3622 Q1363.14 18.9825 1363.14 23.1955 L1363.14 27.2059 L1375.42 27.2059 L1375.42 32.9987 L1363.14 32.9987 L1363.14 72.576 L1355.65 72.576 L1355.65 32.9987 L1348.52 32.9987 L1348.52 27.2059 L1355.65 27.2059 L1355.65 24.0462 Q1355.65 16.471 1359.17 13.0277 Q1362.7 9.54393 1370.36 9.54393 L1377.4 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1430.55 20.1573 L1419.45 50.2555 L1441.69 50.2555 L1430.55 20.1573 M1425.93 12.096 L1435.21 12.096 L1458.26 72.576 L1449.75 72.576 L1444.24 57.061 L1416.98 57.061 L1411.47 72.576 L1402.84 72.576 L1425.93 12.096 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1472.68 14.324 L1472.68 27.2059 L1488.03 27.2059 L1488.03 32.9987 L1472.68 32.9987 L1472.68 57.6282 Q1472.68 63.1779 1474.18 64.7578 Q1475.72 66.3376 1480.38 66.3376 L1488.03 66.3376 L1488.03 72.576 L1480.38 72.576 Q1471.75 72.576 1468.47 69.3758 Q1465.19 66.1351 1465.19 57.6282 L1465.19 32.9987 L1459.72 32.9987 L1459.72 27.2059 L1465.19 27.2059 L1465.19 14.324 L1472.68 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1505.21 14.324 L1505.21 27.2059 L1520.56 27.2059 L1520.56 32.9987 L1505.21 32.9987 L1505.21 57.6282 Q1505.21 63.1779 1506.71 64.7578 Q1508.25 66.3376 1512.91 66.3376 L1520.56 66.3376 L1520.56 72.576 L1512.91 72.576 Q1504.28 72.576 1501 69.3758 Q1497.72 66.1351 1497.72 57.6282 L1497.72 32.9987 L1492.25 32.9987 L1492.25 27.2059 L1497.72 27.2059 L1497.72 14.324 L1505.21 14.324 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1550.99 49.7694 Q1541.95 49.7694 1538.47 51.8354 Q1534.98 53.9013 1534.98 58.8839 Q1534.98 62.8538 1537.58 65.2034 Q1540.21 67.5124 1544.71 67.5124 Q1550.9 67.5124 1554.63 63.1374 Q1558.4 58.7219 1558.4 51.4303 L1558.4 49.7694 L1550.99 49.7694 M1565.85 46.6907 L1565.85 72.576 L1558.4 72.576 L1558.4 65.6895 Q1555.85 69.8214 1552.04 71.8063 Q1548.23 73.7508 1542.72 73.7508 Q1535.75 73.7508 1531.62 69.8619 Q1527.53 65.9325 1527.53 59.3701 Q1527.53 51.7138 1532.63 47.825 Q1537.78 43.9361 1547.95 43.9361 L1558.4 43.9361 L1558.4 43.2069 Q1558.4 38.0623 1555 35.2672 Q1551.63 32.4315 1545.52 32.4315 Q1541.63 32.4315 1537.94 33.3632 Q1534.25 34.295 1530.85 36.1584 L1530.85 29.2718 Q1534.94 27.692 1538.79 26.9223 Q1542.64 26.1121 1546.29 26.1121 Q1556.13 26.1121 1560.99 31.2163 Q1565.85 36.3204 1565.85 46.6907 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1613.86 28.9478 L1613.86 35.9153 Q1610.7 34.1734 1607.5 33.3227 Q1604.34 32.4315 1601.09 32.4315 Q1593.84 32.4315 1589.83 37.0496 Q1585.82 41.6271 1585.82 49.9314 Q1585.82 58.2358 1589.83 62.8538 Q1593.84 67.4314 1601.09 67.4314 Q1604.34 67.4314 1607.5 66.5807 Q1610.7 65.6895 1613.86 63.9476 L1613.86 70.8341 Q1610.74 72.2924 1607.37 73.0216 Q1604.05 73.7508 1600.28 73.7508 Q1590.04 73.7508 1584 67.3098 Q1577.96 60.8689 1577.96 49.9314 Q1577.96 38.832 1584.04 32.472 Q1590.16 26.1121 1600.77 26.1121 Q1604.21 26.1121 1607.5 26.8413 Q1610.78 27.5299 1613.86 28.9478 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1626.53 9.54393 L1634.03 9.54393 L1634.03 46.7717 L1656.27 27.2059 L1665.79 27.2059 L1641.73 48.4326 L1666.8 72.576 L1657.08 72.576 L1634.03 50.4176 L1634.03 72.576 L1626.53 72.576 L1626.53 9.54393 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1719.14 9.62495 Q1713.71 18.942 1711.08 28.0566 Q1708.44 37.1711 1708.44 46.5287 Q1708.44 55.8863 1711.08 65.0818 Q1713.75 74.2369 1719.14 83.5134 L1712.66 83.5134 Q1706.58 73.9938 1703.54 64.7983 Q1700.54 55.6027 1700.54 46.5287 Q1700.54 37.4952 1703.54 28.3401 Q1706.54 19.1851 1712.66 9.62495 L1719.14 9.62495 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1763.45 34.0924 L1763.45 9.54393 L1770.91 9.54393 L1770.91 72.576 L1763.45 72.576 L1763.45 65.7705 Q1761.11 69.8214 1757.5 71.8063 Q1753.94 73.7508 1748.91 73.7508 Q1740.69 73.7508 1735.5 67.1883 Q1730.36 60.6258 1730.36 49.9314 Q1730.36 39.2371 1735.5 32.6746 Q1740.69 26.1121 1748.91 26.1121 Q1753.94 26.1121 1757.5 28.0971 Q1761.11 30.0415 1763.45 34.0924 M1738.06 49.9314 Q1738.06 58.1548 1741.42 62.8538 Q1744.82 67.5124 1750.74 67.5124 Q1756.65 67.5124 1760.05 62.8538 Q1763.45 58.1548 1763.45 49.9314 Q1763.45 41.7081 1760.05 37.0496 Q1756.65 32.3505 1750.74 32.3505 Q1744.82 32.3505 1741.42 37.0496 Q1738.06 41.7081 1738.06 49.9314 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1825.07 48.0275 L1825.07 51.6733 L1790.8 51.6733 Q1791.28 59.3701 1795.42 63.421 Q1799.59 67.4314 1807 67.4314 Q1811.3 67.4314 1815.31 66.3781 Q1819.36 65.3249 1823.33 63.2184 L1823.33 70.267 Q1819.32 71.9684 1815.1 72.8596 Q1810.89 73.7508 1806.56 73.7508 Q1795.7 73.7508 1789.34 67.4314 Q1783.02 61.1119 1783.02 50.3365 Q1783.02 39.1965 1789.02 32.6746 Q1795.05 26.1121 1805.26 26.1121 Q1814.42 26.1121 1819.72 32.0264 Q1825.07 37.9003 1825.07 48.0275 M1817.62 45.84 Q1817.53 39.7232 1814.17 36.0774 Q1810.85 32.4315 1805.34 32.4315 Q1799.1 32.4315 1795.34 35.9558 Q1791.61 39.4801 1791.04 45.8805 L1817.62 45.84 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1867.16 49.3643 Q1867.16 41.2625 1863.8 36.8065 Q1860.47 32.3505 1854.44 32.3505 Q1848.44 32.3505 1845.08 36.8065 Q1841.76 41.2625 1841.76 49.3643 Q1841.76 57.4256 1845.08 61.8816 Q1848.44 66.3376 1854.44 66.3376 Q1860.47 66.3376 1863.8 61.8816 Q1867.16 57.4256 1867.16 49.3643 M1874.61 66.9452 Q1874.61 78.5308 1869.47 84.1616 Q1864.32 89.8329 1853.71 89.8329 Q1849.78 89.8329 1846.3 89.2252 Q1842.81 88.6581 1839.53 87.4428 L1839.53 80.1917 Q1842.81 81.9741 1846.01 82.8248 Q1849.21 83.6755 1852.53 83.6755 Q1859.87 83.6755 1863.51 79.8271 Q1867.16 76.0193 1867.16 68.282 L1867.16 64.5957 Q1864.85 68.6061 1861.24 70.5911 Q1857.64 72.576 1852.62 72.576 Q1844.27 72.576 1839.17 66.2161 Q1834.06 59.8562 1834.06 49.3643 Q1834.06 38.832 1839.17 32.472 Q1844.27 26.1121 1852.62 26.1121 Q1857.64 26.1121 1861.24 28.0971 Q1864.85 30.082 1867.16 34.0924 L1867.16 27.2059 L1874.61 27.2059 L1874.61 66.9452 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1888.79 9.62495 L1895.27 9.62495 Q1901.35 19.1851 1904.35 28.3401 Q1907.38 37.4952 1907.38 46.5287 Q1907.38 55.6027 1904.35 64.7983 Q1901.35 73.9938 1895.27 83.5134 L1888.79 83.5134 Q1894.18 74.2369 1896.81 65.0818 Q1899.48 55.8863 1899.48 46.5287 Q1899.48 37.1711 1896.81 28.0566 Q1894.18 18.942 1888.79 9.62495 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip692)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="261.245,237.832 266.073,234.59 270.902,231.464 275.73,228.457 280.558,225.572 285.386,222.813 290.214,220.184 295.043,217.689 299.871,215.334 304.699,213.124 309.527,211.063 314.356,209.159 319.184,207.418 324.012,205.846 328.84,204.451 333.668,203.24 338.497,202.221 343.325,201.402 348.153,200.792 352.981,200.398 357.81,200.23 362.638,200.295 367.466,200.603 372.294,201.16 377.123,201.973 381.951,203.05 386.779,204.396 391.607,206.015 396.435,207.91 401.264,210.082 406.092,212.531 410.92,215.255 415.748,218.248 420.577,221.504 425.405,225.012 430.233,228.762 435.061,232.738 439.89,236.924 444.718,241.298 449.546,245.841 454.374,250.529 459.202,255.335 464.031,260.234 468.859,265.197 473.687,270.197 478.515,275.206 483.344,280.195 488.172,285.138 493,290.007 497.828,294.777 502.656,299.426 507.485,303.93 512.313,308.271 517.141,312.43 521.969,316.391 526.798,320.14 531.626,323.665 536.454,326.957 541.282,330.006 546.111,332.806 550.939,335.352 555.767,337.641 560.595,339.669 565.423,341.435 570.252,342.94 575.08,344.183 579.908,345.166 584.736,345.89 589.565,346.358 594.393,346.573 599.221,346.537 604.049,346.255 608.878,345.73 613.706,344.967 618.534,343.968 623.362,342.74 628.19,341.286 633.019,339.612 637.847,337.721 642.675,335.619 647.503,333.311 652.332,330.802 657.16,328.098 661.988,325.204 666.816,322.126 671.645,318.87 676.473,315.442 681.301,311.849 686.129,308.096 690.957,304.192 695.786,300.143 700.614,295.958 705.442,291.644 710.27,287.21 715.099,282.665 719.927,278.018 724.755,273.281 729.583,268.464 734.411,263.578 739.24,258.636 744.068,253.651 748.896,248.637 753.724,243.608 758.553,238.581 763.381,233.572 768.209,228.599 773.037,223.682 777.866,218.839 782.694,214.093 787.522,209.465 792.35,204.978 797.178,200.658 802.007,196.53 806.835,192.62 811.663,188.956 816.491,185.566 821.32,182.478 826.148,179.721 830.976,177.325 835.804,175.319 840.633,173.73 845.461,172.586 850.289,171.913 855.117,171.734 859.945,172.072 864.774,172.944 869.602,174.365 874.43,176.346 879.258,178.891 884.087,182.002 888.915,185.672 893.743,189.889 898.571,194.636 903.399,199.888 908.228,205.612 913.056,211.771 917.884,218.319 922.712,225.208 927.541,232.381 932.369,239.779 937.197,247.339 942.025,254.996 946.854,262.682 951.682,270.331 956.51,277.875 961.338,285.25 966.166,292.394 970.995,299.247 975.823,305.754 980.651,311.865 985.479,317.535 990.308,322.723 995.136,327.395 999.964,331.522 1004.79,335.08 1009.62,338.05 1014.45,340.42 1019.28,342.181 1024.11,343.329 1028.93,343.865 1033.76,343.795 1038.59,343.125 1043.42,341.868 1048.25,340.039 1053.07,337.656 1057.9,334.739 1062.73,331.312 1067.56,327.399 1072.39,323.028 1077.22,318.23 1082.04,313.035 1086.87,307.476 1091.7,301.59 1096.53,295.414 1101.36,288.986 1106.19,282.346 1111.01,275.537 1115.84,268.604 1120.67,261.591 1125.5,254.545 1130.33,247.515 1135.15,240.552 1139.98,233.706 1144.81,227.029 1149.64,220.575 1154.47,214.398 1159.3,208.55 1164.12,203.087 1168.95,198.06 1173.78,193.523 1178.61,189.524 1183.44,186.112 1188.26,183.331 1193.09,181.221 1197.92,179.818 1202.75,179.152 1207.58,179.245 1212.41,180.113 1217.23,181.763 1222.06,184.194 1226.89,187.395 1231.72,191.343 1236.55,196.007 1241.38,201.346 1246.2,207.308 1251.03,213.831 1255.86,220.845 1260.69,228.271 1265.52,236.026 1270.34,244.018 1275.17,252.154 1280,260.337 1284.83,268.471 1289.66,276.461 1294.49,284.212 1299.31,291.637 1304.14,298.652 1308.97,305.18 1313.8,311.151 1318.63,316.505 1323.46,321.189 1328.28,325.159 1333.11,328.382 1337.94,330.833 1342.77,332.496 1347.6,333.364 1352.42,333.44 1357.25,332.731 1362.08,331.257 1366.91,329.042 1371.74,326.116 1376.57,322.517 1381.39,318.289 1386.22,313.48 1391.05,308.142 1395.88,302.334 1400.71,296.118 1405.54,289.557 1410.36,282.72 1415.19,275.678 1420.02,268.504 1424.85,261.273 1429.68,254.061 1434.5,246.946 1439.33,240.003 1444.16,233.311 1448.99,226.944 1453.82,220.978 1458.65,215.482 1463.47,210.524 1468.3,206.169 1473.13,202.473 1477.96,199.488 1482.79,197.258 1487.62,195.818 1492.44,195.193 1497.27,195.401 1502.1,196.444 1506.93,198.316 1511.76,200.999 1516.58,204.46 1521.41,208.657 1526.24,213.534 1531.07,219.026 1535.9,225.055 1540.73,231.537 1545.55,238.379 1550.38,245.481 1555.21,252.741 1560.04,260.053 1564.87,267.312 1569.7,274.414 1574.52,281.258 1579.35,287.748 1584.18,293.797 1589.01,299.323 1593.84,304.256 1598.66,308.532 1603.49,312.103 1608.32,314.929 1613.15,316.981 1617.98,318.244 1622.81,318.711 1627.63,318.388 1632.46,317.291 1637.29,315.447 1642.12,312.89 1646.95,309.665 1651.77,305.824 1656.6,301.424 1661.43,296.533 1666.26,291.221 1671.09,285.563 1675.92,279.639 1680.74,273.531 1685.57,267.325 1690.4,261.106 1695.23,254.96 1700.06,248.973 1704.89,243.227 1709.71,237.804 1714.54,232.78 1719.37,228.227 1724.2,224.21 1729.03,220.79 1733.85,218.015 1738.68,215.928 1743.51,214.559 1748.34,213.929 1753.17,214.048 1758,214.911 1762.82,216.505 1767.65,218.803 1772.48,221.765 1777.31,225.341 1782.14,229.471 1786.97,234.084 1791.79,239.102 1796.62,244.439 1801.45,250.005 1806.28,255.705 1811.11,261.445 1815.93,267.129 1820.76,272.664 1825.59,277.96 1830.42,282.935 1835.25,287.511 1840.08,291.621 1844.9,295.203 1849.73,298.211 1854.56,300.605 1859.39,302.357 1864.22,303.452 1869.05,303.885 1873.87,303.661 1878.7,302.796 1883.53,301.318 1888.36,299.262 1893.19,296.671 1898.01,293.598 1902.84,290.101 1907.67,286.244 1912.5,282.096 1917.33,277.73 1922.16,273.221 1926.98,268.645 1931.81,264.08 1936.64,259.599 1941.47,255.279 1946.3,251.188 1951.13,247.394 1955.95,243.958 1960.78,240.933 1965.61,238.369 1970.44,236.303 1975.27,234.767 1980.09,233.781 1984.92,233.357 1989.75,233.496 1994.58,234.19 1999.41,235.418 2004.24,237.152 2009.06,239.356 2013.89,241.982 2018.72,244.977 2023.55,248.282 2028.38,251.831 2033.2,255.558 2038.03,259.391 2042.86,263.259 2047.69,267.093 2052.52,270.824 2057.35,274.389 2062.17,277.727 2067,280.785 2071.83,283.517 2076.66,285.884 2081.49,287.854 2086.32,289.405 2091.14,290.523 2095.97,291.203 2100.8,291.449 2105.63,291.272 2110.46,290.69 2115.28,289.73 2120.11,288.425 2124.94,286.811 2129.77,284.932 2134.6,282.832 2139.43,280.562 2144.25,278.172 2149.08,275.711 2153.91,273.233 2158.74,270.785 2163.57,268.415 2168.4,266.169 2173.22,264.086 2178.05,262.202 2182.88,260.549 2187.71,259.151 2192.54,258.028 2197.36,257.193 2202.19,256.652 2207.02,256.405 2211.85,256.446 2216.68,256.763 2221.51,257.339 2226.33,258.152 2231.16,259.175 2235.99,260.379 2240.82,261.732 2245.65,263.201 2250.48,264.752 2255.3,266.354 2260.13,267.973 2264.96,269.581 2269.79,271.152 2274.62,272.665 2279.44,274.1 2284.27,275.446 2289.1,276.694 2293.93,277.841 2298.76,278.89 2303.59,279.848 2308.41,280.725 2313.24,281.537 2318.07,282.302 2322.9,283.041 2327.73,283.775 2332.56,284.528 2337.38,285.321 2342.21,286.177 2347.04,287.113 2351.87,288.145 2356.7,289.285 2361.52,290.538 2366.35,291.907 2371.18,293.386 2376.01,294.964 2380.84,296.621 2385.67,298.333 2390.49,300.069 2395.32,301.789 2400.15,303.452 2404.98,305.009 2409.81,306.409 2414.64,307.597 2419.46,308.52 2424.29,309.124 2429.12,309.359 2433.95,309.177 2438.78,308.542 2443.6,307.422 2448.43,305.799 2453.26,303.666 2458.09,301.032 2462.92,297.923 2467.75,294.38 2472.57,290.467 2477.4,286.264 2482.23,281.873 2487.06,277.417 2491.89,273.036 2496.71,268.893 2501.54,265.164 2506.37,262.042 2511.2,259.735 2516.03,258.459 2520.86,258.437 2525.68,259.896 2530.51,263.059 2535.34,268.144 2540.17,275.355 2545,284.877 2549.83,296.873 2554.65,311.468 2559.48,328.752 2564.31,348.763 2569.14,371.485 2573.97,396.837 2578.79,424.666 2583.62,454.736 2588.45,486.725 2593.28,520.214 2598.11,554.686 2602.94,589.516 2607.76,623.975 2612.59,657.227 2617.42,688.332 2622.25,716.25 2627.08,739.855 2631.91,757.944 2636.73,769.253 2641.56,772.479 2646.39,766.3 2651.22,749.403 2656.05,720.513 2660.87,678.419 2665.7,622.005 2670.53,550.282 2675.36,462.415 2680.19,357.751 "/>
+<path clip-path="url(#clip690)" d="M188.677 1723.83 L2752.76 1723.83 L2752.76 1087.05 L188.677 1087.05  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip693">
+    <rect x="188" y="1087" width="2565" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="261.245,1723.83 261.245,1087.05 "/>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="864.774,1723.83 864.774,1087.05 "/>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1468.3,1723.83 1468.3,1087.05 "/>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2071.83,1723.83 2071.83,1087.05 "/>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2675.36,1723.83 2675.36,1087.05 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,1723.83 2752.76,1723.83 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="261.245,1723.83 261.245,1704.94 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="864.774,1723.83 864.774,1704.94 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1468.3,1723.83 1468.3,1704.94 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2071.83,1723.83 2071.83,1704.94 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2675.36,1723.83 2675.36,1704.94 "/>
+<path clip-path="url(#clip690)" d="M261.245 1757.95 Q257.634 1757.95 255.805 1761.52 Q254 1765.06 254 1772.19 Q254 1779.3 255.805 1782.86 Q257.634 1786.4 261.245 1786.4 Q264.879 1786.4 266.685 1782.86 Q268.514 1779.3 268.514 1772.19 Q268.514 1765.06 266.685 1761.52 Q264.879 1757.95 261.245 1757.95 M261.245 1754.25 Q267.055 1754.25 270.111 1758.86 Q273.189 1763.44 273.189 1772.19 Q273.189 1780.92 270.111 1785.52 Q267.055 1790.11 261.245 1790.11 Q255.435 1790.11 252.356 1785.52 Q249.301 1780.92 249.301 1772.19 Q249.301 1763.44 252.356 1758.86 Q255.435 1754.25 261.245 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M824.392 1754.87 L842.748 1754.87 L842.748 1758.81 L828.674 1758.81 L828.674 1767.28 Q829.693 1766.93 830.711 1766.77 Q831.73 1766.59 832.748 1766.59 Q838.535 1766.59 841.915 1769.76 Q845.295 1772.93 845.295 1778.35 Q845.295 1783.93 841.822 1787.03 Q838.35 1790.11 832.031 1790.11 Q829.855 1790.11 827.586 1789.74 Q825.341 1789.37 822.934 1788.62 L822.934 1783.93 Q825.017 1785.06 827.239 1785.62 Q829.461 1786.17 831.938 1786.17 Q835.943 1786.17 838.281 1784.06 Q840.619 1781.96 840.619 1778.35 Q840.619 1774.74 838.281 1772.63 Q835.943 1770.52 831.938 1770.52 Q830.063 1770.52 828.188 1770.94 Q826.336 1771.36 824.392 1772.24 L824.392 1754.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M864.507 1757.95 Q860.896 1757.95 859.068 1761.52 Q857.262 1765.06 857.262 1772.19 Q857.262 1779.3 859.068 1782.86 Q860.896 1786.4 864.507 1786.4 Q868.142 1786.4 869.947 1782.86 Q871.776 1779.3 871.776 1772.19 Q871.776 1765.06 869.947 1761.52 Q868.142 1757.95 864.507 1757.95 M864.507 1754.25 Q870.318 1754.25 873.373 1758.86 Q876.452 1763.44 876.452 1772.19 Q876.452 1780.92 873.373 1785.52 Q870.318 1790.11 864.507 1790.11 Q858.697 1790.11 855.619 1785.52 Q852.563 1780.92 852.563 1772.19 Q852.563 1763.44 855.619 1758.86 Q858.697 1754.25 864.507 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M894.669 1757.95 Q891.058 1757.95 889.23 1761.52 Q887.424 1765.06 887.424 1772.19 Q887.424 1779.3 889.23 1782.86 Q891.058 1786.4 894.669 1786.4 Q898.304 1786.4 900.109 1782.86 Q901.938 1779.3 901.938 1772.19 Q901.938 1765.06 900.109 1761.52 Q898.304 1757.95 894.669 1757.95 M894.669 1754.25 Q900.479 1754.25 903.535 1758.86 Q906.614 1763.44 906.614 1772.19 Q906.614 1780.92 903.535 1785.52 Q900.479 1790.11 894.669 1790.11 Q888.859 1790.11 885.78 1785.52 Q882.725 1780.92 882.725 1772.19 Q882.725 1763.44 885.78 1758.86 Q888.859 1754.25 894.669 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1412.83 1785.5 L1420.47 1785.5 L1420.47 1759.13 L1412.16 1760.8 L1412.16 1756.54 L1420.42 1754.87 L1425.1 1754.87 L1425.1 1785.5 L1432.74 1785.5 L1432.74 1789.43 L1412.83 1789.43 L1412.83 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1452.18 1757.95 Q1448.57 1757.95 1446.74 1761.52 Q1444.93 1765.06 1444.93 1772.19 Q1444.93 1779.3 1446.74 1782.86 Q1448.57 1786.4 1452.18 1786.4 Q1455.81 1786.4 1457.62 1782.86 Q1459.45 1779.3 1459.45 1772.19 Q1459.45 1765.06 1457.62 1761.52 Q1455.81 1757.95 1452.18 1757.95 M1452.18 1754.25 Q1457.99 1754.25 1461.05 1758.86 Q1464.12 1763.44 1464.12 1772.19 Q1464.12 1780.92 1461.05 1785.52 Q1457.99 1790.11 1452.18 1790.11 Q1446.37 1790.11 1443.29 1785.52 Q1440.24 1780.92 1440.24 1772.19 Q1440.24 1763.44 1443.29 1758.86 Q1446.37 1754.25 1452.18 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1482.34 1757.95 Q1478.73 1757.95 1476.9 1761.52 Q1475.1 1765.06 1475.1 1772.19 Q1475.1 1779.3 1476.9 1782.86 Q1478.73 1786.4 1482.34 1786.4 Q1485.98 1786.4 1487.78 1782.86 Q1489.61 1779.3 1489.61 1772.19 Q1489.61 1765.06 1487.78 1761.52 Q1485.98 1757.95 1482.34 1757.95 M1482.34 1754.25 Q1488.15 1754.25 1491.21 1758.86 Q1494.29 1763.44 1494.29 1772.19 Q1494.29 1780.92 1491.21 1785.52 Q1488.15 1790.11 1482.34 1790.11 Q1476.53 1790.11 1473.45 1785.52 Q1470.4 1780.92 1470.4 1772.19 Q1470.4 1763.44 1473.45 1758.86 Q1476.53 1754.25 1482.34 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1512.5 1757.95 Q1508.89 1757.95 1507.06 1761.52 Q1505.26 1765.06 1505.26 1772.19 Q1505.26 1779.3 1507.06 1782.86 Q1508.89 1786.4 1512.5 1786.4 Q1516.14 1786.4 1517.94 1782.86 Q1519.77 1779.3 1519.77 1772.19 Q1519.77 1765.06 1517.94 1761.52 Q1516.14 1757.95 1512.5 1757.95 M1512.5 1754.25 Q1518.31 1754.25 1521.37 1758.86 Q1524.45 1763.44 1524.45 1772.19 Q1524.45 1780.92 1521.37 1785.52 Q1518.31 1790.11 1512.5 1790.11 Q1506.69 1790.11 1503.61 1785.52 Q1500.56 1780.92 1500.56 1772.19 Q1500.56 1763.44 1503.61 1758.86 Q1506.69 1754.25 1512.5 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2016.36 1785.5 L2024 1785.5 L2024 1759.13 L2015.69 1760.8 L2015.69 1756.54 L2023.95 1754.87 L2028.63 1754.87 L2028.63 1785.5 L2036.26 1785.5 L2036.26 1789.43 L2016.36 1789.43 L2016.36 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2045.75 1754.87 L2064.11 1754.87 L2064.11 1758.81 L2050.04 1758.81 L2050.04 1767.28 Q2051.06 1766.93 2052.07 1766.77 Q2053.09 1766.59 2054.11 1766.59 Q2059.9 1766.59 2063.28 1769.76 Q2066.66 1772.93 2066.66 1778.35 Q2066.66 1783.93 2063.19 1787.03 Q2059.71 1790.11 2053.39 1790.11 Q2051.22 1790.11 2048.95 1789.74 Q2046.7 1789.37 2044.3 1788.62 L2044.3 1783.93 Q2046.38 1785.06 2048.6 1785.62 Q2050.82 1786.17 2053.3 1786.17 Q2057.31 1786.17 2059.64 1784.06 Q2061.98 1781.96 2061.98 1778.35 Q2061.98 1774.74 2059.64 1772.63 Q2057.31 1770.52 2053.3 1770.52 Q2051.43 1770.52 2049.55 1770.94 Q2047.7 1771.36 2045.75 1772.24 L2045.75 1754.87 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2085.87 1757.95 Q2082.26 1757.95 2080.43 1761.52 Q2078.62 1765.06 2078.62 1772.19 Q2078.62 1779.3 2080.43 1782.86 Q2082.26 1786.4 2085.87 1786.4 Q2089.5 1786.4 2091.31 1782.86 Q2093.14 1779.3 2093.14 1772.19 Q2093.14 1765.06 2091.31 1761.52 Q2089.5 1757.95 2085.87 1757.95 M2085.87 1754.25 Q2091.68 1754.25 2094.74 1758.86 Q2097.81 1763.44 2097.81 1772.19 Q2097.81 1780.92 2094.74 1785.52 Q2091.68 1790.11 2085.87 1790.11 Q2080.06 1790.11 2076.98 1785.52 Q2073.93 1780.92 2073.93 1772.19 Q2073.93 1763.44 2076.98 1758.86 Q2080.06 1754.25 2085.87 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2116.03 1757.95 Q2112.42 1757.95 2110.59 1761.52 Q2108.79 1765.06 2108.79 1772.19 Q2108.79 1779.3 2110.59 1782.86 Q2112.42 1786.4 2116.03 1786.4 Q2119.67 1786.4 2121.47 1782.86 Q2123.3 1779.3 2123.3 1772.19 Q2123.3 1765.06 2121.47 1761.52 Q2119.67 1757.95 2116.03 1757.95 M2116.03 1754.25 Q2121.84 1754.25 2124.9 1758.86 Q2127.98 1763.44 2127.98 1772.19 Q2127.98 1780.92 2124.9 1785.52 Q2121.84 1790.11 2116.03 1790.11 Q2110.22 1790.11 2107.14 1785.52 Q2104.09 1780.92 2104.09 1772.19 Q2104.09 1763.44 2107.14 1758.86 Q2110.22 1754.25 2116.03 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2623.97 1785.5 L2640.29 1785.5 L2640.29 1789.43 L2618.35 1789.43 L2618.35 1785.5 Q2621.01 1782.74 2625.59 1778.12 Q2630.2 1773.46 2631.38 1772.12 Q2633.62 1769.6 2634.5 1767.86 Q2635.41 1766.1 2635.41 1764.41 Q2635.41 1761.66 2633.46 1759.92 Q2631.54 1758.18 2628.44 1758.18 Q2626.24 1758.18 2623.79 1758.95 Q2621.36 1759.71 2618.58 1761.26 L2618.58 1756.54 Q2621.4 1755.41 2623.86 1754.83 Q2626.31 1754.25 2628.35 1754.25 Q2633.72 1754.25 2636.91 1756.93 Q2640.1 1759.62 2640.1 1764.11 Q2640.1 1766.24 2639.29 1768.16 Q2638.51 1770.06 2636.4 1772.65 Q2635.82 1773.32 2632.72 1776.54 Q2629.62 1779.74 2623.97 1785.5 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2660.1 1757.95 Q2656.49 1757.95 2654.67 1761.52 Q2652.86 1765.06 2652.86 1772.19 Q2652.86 1779.3 2654.67 1782.86 Q2656.49 1786.4 2660.1 1786.4 Q2663.74 1786.4 2665.54 1782.86 Q2667.37 1779.3 2667.37 1772.19 Q2667.37 1765.06 2665.54 1761.52 Q2663.74 1757.95 2660.1 1757.95 M2660.1 1754.25 Q2665.92 1754.25 2668.97 1758.86 Q2672.05 1763.44 2672.05 1772.19 Q2672.05 1780.92 2668.97 1785.52 Q2665.92 1790.11 2660.1 1790.11 Q2654.29 1790.11 2651.22 1785.52 Q2648.16 1780.92 2648.16 1772.19 Q2648.16 1763.44 2651.22 1758.86 Q2654.29 1754.25 2660.1 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2690.27 1757.95 Q2686.66 1757.95 2684.83 1761.52 Q2683.02 1765.06 2683.02 1772.19 Q2683.02 1779.3 2684.83 1782.86 Q2686.66 1786.4 2690.27 1786.4 Q2693.9 1786.4 2695.71 1782.86 Q2697.54 1779.3 2697.54 1772.19 Q2697.54 1765.06 2695.71 1761.52 Q2693.9 1757.95 2690.27 1757.95 M2690.27 1754.25 Q2696.08 1754.25 2699.13 1758.86 Q2702.21 1763.44 2702.21 1772.19 Q2702.21 1780.92 2699.13 1785.52 Q2696.08 1790.11 2690.27 1790.11 Q2684.46 1790.11 2681.38 1785.52 Q2678.32 1780.92 2678.32 1772.19 Q2678.32 1763.44 2681.38 1758.86 Q2684.46 1754.25 2690.27 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2720.43 1757.95 Q2716.82 1757.95 2714.99 1761.52 Q2713.18 1765.06 2713.18 1772.19 Q2713.18 1779.3 2714.99 1782.86 Q2716.82 1786.4 2720.43 1786.4 Q2724.06 1786.4 2725.87 1782.86 Q2727.7 1779.3 2727.7 1772.19 Q2727.7 1765.06 2725.87 1761.52 Q2724.06 1757.95 2720.43 1757.95 M2720.43 1754.25 Q2726.24 1754.25 2729.29 1758.86 Q2732.37 1763.44 2732.37 1772.19 Q2732.37 1780.92 2729.29 1785.52 Q2726.24 1790.11 2720.43 1790.11 Q2714.62 1790.11 2711.54 1785.52 Q2708.48 1780.92 2708.48 1772.19 Q2708.48 1763.44 2711.54 1758.86 Q2714.62 1754.25 2720.43 1754.25 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,1590.75 2752.76,1590.75 "/>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,1428.81 2752.76,1428.81 "/>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,1266.87 2752.76,1266.87 "/>
+<polyline clip-path="url(#clip693)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,1104.93 2752.76,1104.93 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,1723.83 188.677,1087.05 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,1590.75 207.574,1590.75 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,1428.81 207.574,1428.81 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,1266.87 207.574,1266.87 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,1104.93 207.574,1104.93 "/>
+<path clip-path="url(#clip690)" d="M54.8023 1591.21 L84.478 1591.21 L84.478 1595.14 L54.8023 1595.14 L54.8023 1591.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M105.149 1588.89 Q102.001 1588.89 100.149 1591.04 Q98.3206 1593.2 98.3206 1596.95 Q98.3206 1600.67 100.149 1602.85 Q102.001 1605 105.149 1605 Q108.297 1605 110.126 1602.85 Q111.978 1600.67 111.978 1596.95 Q111.978 1593.2 110.126 1591.04 Q108.297 1588.89 105.149 1588.89 M114.432 1574.24 L114.432 1578.5 Q112.672 1577.66 110.867 1577.22 Q109.084 1576.78 107.325 1576.78 Q102.696 1576.78 100.242 1579.91 Q97.8113 1583.03 97.4641 1589.35 Q98.8298 1587.34 100.89 1586.27 Q102.95 1585.19 105.427 1585.19 Q110.635 1585.19 113.645 1588.36 Q116.677 1591.51 116.677 1596.95 Q116.677 1602.27 113.529 1605.49 Q110.381 1608.71 105.149 1608.71 Q99.1539 1608.71 95.9826 1604.12 Q92.8113 1599.52 92.8113 1590.79 Q92.8113 1582.59 96.7002 1577.73 Q100.589 1572.85 107.14 1572.85 Q108.899 1572.85 110.682 1573.2 Q112.487 1573.54 114.432 1574.24 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 1576.55 Q131.121 1576.55 129.293 1580.12 Q127.487 1583.66 127.487 1590.79 Q127.487 1597.89 129.293 1601.46 Q131.121 1605 134.732 1605 Q138.367 1605 140.172 1601.46 Q142.001 1597.89 142.001 1590.79 Q142.001 1583.66 140.172 1580.12 Q138.367 1576.55 134.732 1576.55 M134.732 1572.85 Q140.543 1572.85 143.598 1577.46 Q146.677 1582.04 146.677 1590.79 Q146.677 1599.52 143.598 1604.12 Q140.543 1608.71 134.732 1608.71 Q128.922 1608.71 125.844 1604.12 Q122.788 1599.52 122.788 1590.79 Q122.788 1582.04 125.844 1577.46 Q128.922 1572.85 134.732 1572.85 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M54.8023 1429.26 L84.478 1429.26 L84.478 1433.2 L54.8023 1433.2 L54.8023 1429.26 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M107.418 1415.61 L95.6122 1434.05 L107.418 1434.05 L107.418 1415.61 M106.191 1411.53 L112.07 1411.53 L112.07 1434.05 L117.001 1434.05 L117.001 1437.94 L112.07 1437.94 L112.07 1446.09 L107.418 1446.09 L107.418 1437.94 L91.816 1437.94 L91.816 1433.43 L106.191 1411.53 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 1414.61 Q131.121 1414.61 129.293 1418.18 Q127.487 1421.72 127.487 1428.85 Q127.487 1435.95 129.293 1439.52 Q131.121 1443.06 134.732 1443.06 Q138.367 1443.06 140.172 1439.52 Q142.001 1435.95 142.001 1428.85 Q142.001 1421.72 140.172 1418.18 Q138.367 1414.61 134.732 1414.61 M134.732 1410.91 Q140.543 1410.91 143.598 1415.51 Q146.677 1420.1 146.677 1428.85 Q146.677 1437.57 143.598 1442.18 Q140.543 1446.76 134.732 1446.76 Q128.922 1446.76 125.844 1442.18 Q122.788 1437.57 122.788 1428.85 Q122.788 1420.1 125.844 1415.51 Q128.922 1410.91 134.732 1410.91 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M54.8023 1267.32 L84.478 1267.32 L84.478 1271.26 L54.8023 1271.26 L54.8023 1267.32 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M98.5983 1280.21 L114.918 1280.21 L114.918 1284.15 L92.9734 1284.15 L92.9734 1280.21 Q95.6354 1277.46 100.219 1272.83 Q104.825 1268.18 106.006 1266.83 Q108.251 1264.31 109.131 1262.58 Q110.033 1260.82 110.033 1259.13 Q110.033 1256.37 108.089 1254.64 Q106.168 1252.9 103.066 1252.9 Q100.867 1252.9 98.4132 1253.66 Q95.9826 1254.43 93.2049 1255.98 L93.2049 1251.26 Q96.0289 1250.12 98.4826 1249.54 Q100.936 1248.96 102.973 1248.96 Q108.344 1248.96 111.538 1251.65 Q114.733 1254.33 114.733 1258.83 Q114.733 1260.96 113.922 1262.88 Q113.135 1264.77 111.029 1267.37 Q110.45 1268.04 107.348 1271.26 Q104.246 1274.45 98.5983 1280.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 1252.67 Q131.121 1252.67 129.293 1256.23 Q127.487 1259.77 127.487 1266.9 Q127.487 1274.01 129.293 1277.58 Q131.121 1281.12 134.732 1281.12 Q138.367 1281.12 140.172 1277.58 Q142.001 1274.01 142.001 1266.9 Q142.001 1259.77 140.172 1256.23 Q138.367 1252.67 134.732 1252.67 M134.732 1248.96 Q140.543 1248.96 143.598 1253.57 Q146.677 1258.15 146.677 1266.9 Q146.677 1275.63 143.598 1280.24 Q140.543 1284.82 134.732 1284.82 Q128.922 1284.82 125.844 1280.24 Q122.788 1275.63 122.788 1266.9 Q122.788 1258.15 125.844 1253.57 Q128.922 1248.96 134.732 1248.96 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 1090.73 Q131.121 1090.73 129.293 1094.29 Q127.487 1097.83 127.487 1104.96 Q127.487 1112.07 129.293 1115.63 Q131.121 1119.17 134.732 1119.17 Q138.367 1119.17 140.172 1115.63 Q142.001 1112.07 142.001 1104.96 Q142.001 1097.83 140.172 1094.29 Q138.367 1090.73 134.732 1090.73 M134.732 1087.02 Q140.543 1087.02 143.598 1091.63 Q146.677 1096.21 146.677 1104.96 Q146.677 1113.69 143.598 1118.3 Q140.543 1122.88 134.732 1122.88 Q128.922 1122.88 125.844 1118.3 Q122.788 1113.69 122.788 1104.96 Q122.788 1096.21 125.844 1091.63 Q128.922 1087.02 134.732 1087.02 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1127.38 977.026 L1127.38 999.185 L1140.51 999.185 Q1147.11 999.185 1150.27 996.471 Q1153.47 993.716 1153.47 988.085 Q1153.47 982.414 1150.27 979.74 Q1147.11 977.026 1140.51 977.026 L1127.38 977.026 M1127.38 952.154 L1127.38 970.383 L1139.49 970.383 Q1145.49 970.383 1148.41 968.155 Q1151.36 965.886 1151.36 961.268 Q1151.36 956.691 1148.41 954.422 Q1145.49 952.154 1139.49 952.154 L1127.38 952.154 M1119.2 945.429 L1140.1 945.429 Q1149.46 945.429 1154.52 949.318 Q1159.59 953.207 1159.59 960.377 Q1159.59 965.927 1156.99 969.208 Q1154.4 972.489 1149.38 973.3 Q1155.41 974.596 1158.74 978.728 Q1162.1 982.819 1162.1 988.977 Q1162.1 997.078 1156.59 1001.49 Q1151.08 1005.91 1140.91 1005.91 L1119.2 1005.91 L1119.2 945.429 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1196.41 983.103 Q1187.38 983.103 1183.89 985.169 Q1180.41 987.235 1180.41 992.217 Q1180.41 996.187 1183 998.537 Q1185.63 1000.85 1190.13 1000.85 Q1196.33 1000.85 1200.06 996.471 Q1203.82 992.055 1203.82 984.764 L1203.82 983.103 L1196.41 983.103 M1211.28 980.024 L1211.28 1005.91 L1203.82 1005.91 L1203.82 999.023 Q1201.27 1003.15 1197.46 1005.14 Q1193.65 1007.08 1188.15 1007.08 Q1181.18 1007.08 1177.05 1003.2 Q1172.95 999.266 1172.95 992.703 Q1172.95 985.047 1178.06 981.158 Q1183.2 977.269 1193.37 977.269 L1203.82 977.269 L1203.82 976.54 Q1203.82 971.396 1200.42 968.601 Q1197.06 965.765 1190.94 965.765 Q1187.05 965.765 1183.37 966.697 Q1179.68 967.628 1176.28 969.492 L1176.28 962.605 Q1180.37 961.025 1184.22 960.256 Q1188.06 959.445 1191.71 959.445 Q1201.55 959.445 1206.42 964.55 Q1211.28 969.654 1211.28 980.024 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1264.34 978.525 L1264.34 1005.91 L1256.89 1005.91 L1256.89 978.768 Q1256.89 972.327 1254.38 969.127 Q1251.87 965.927 1246.84 965.927 Q1240.81 965.927 1237.32 969.775 Q1233.84 973.624 1233.84 980.267 L1233.84 1005.91 L1226.35 1005.91 L1226.35 960.539 L1233.84 960.539 L1233.84 967.588 Q1236.51 963.496 1240.12 961.471 Q1243.76 959.445 1248.5 959.445 Q1256.32 959.445 1260.33 964.307 Q1264.34 969.127 1264.34 978.525 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1278.93 942.877 L1286.42 942.877 L1286.42 980.105 L1308.66 960.539 L1318.18 960.539 L1294.12 981.766 L1319.19 1005.91 L1309.47 1005.91 L1286.42 983.751 L1286.42 1005.91 L1278.93 1005.91 L1278.93 942.877 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1374.16 953.491 L1363.06 983.589 L1385.3 983.589 L1374.16 953.491 M1369.55 945.429 L1378.82 945.429 L1401.87 1005.91 L1393.36 1005.91 L1387.86 990.394 L1360.59 990.394 L1355.08 1005.91 L1346.45 1005.91 L1369.55 945.429 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1448.09 978.525 L1448.09 1005.91 L1440.64 1005.91 L1440.64 978.768 Q1440.64 972.327 1438.13 969.127 Q1435.62 965.927 1430.59 965.927 Q1424.56 965.927 1421.07 969.775 Q1417.59 973.624 1417.59 980.267 L1417.59 1005.91 L1410.09 1005.91 L1410.09 960.539 L1417.59 960.539 L1417.59 967.588 Q1420.26 963.496 1423.87 961.471 Q1427.51 959.445 1432.25 959.445 Q1440.07 959.445 1444.08 964.307 Q1448.09 969.127 1448.09 978.525 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1492.81 982.698 Q1492.81 974.596 1489.45 970.14 Q1486.13 965.684 1480.09 965.684 Q1474.1 965.684 1470.74 970.14 Q1467.41 974.596 1467.41 982.698 Q1467.41 990.759 1470.74 995.215 Q1474.1 999.671 1480.09 999.671 Q1486.13 999.671 1489.45 995.215 Q1492.81 990.759 1492.81 982.698 M1500.27 1000.28 Q1500.27 1011.86 1495.12 1017.49 Q1489.98 1023.17 1479.37 1023.17 Q1475.44 1023.17 1471.95 1022.56 Q1468.47 1021.99 1465.19 1020.78 L1465.19 1013.53 Q1468.47 1015.31 1471.67 1016.16 Q1474.87 1017.01 1478.19 1017.01 Q1485.52 1017.01 1489.17 1013.16 Q1492.81 1009.35 1492.81 1001.62 L1492.81 997.929 Q1490.51 1001.94 1486.9 1003.92 Q1483.29 1005.91 1478.27 1005.91 Q1469.93 1005.91 1464.82 999.549 Q1459.72 993.189 1459.72 982.698 Q1459.72 972.165 1464.82 965.805 Q1469.93 959.445 1478.27 959.445 Q1483.29 959.445 1486.9 961.43 Q1490.51 963.415 1492.81 967.426 L1492.81 960.539 L1500.27 960.539 L1500.27 1000.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1515.62 942.877 L1523.07 942.877 L1523.07 1005.91 L1515.62 1005.91 L1515.62 942.877 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1577.48 981.361 L1577.48 985.007 L1543.21 985.007 Q1543.69 992.703 1547.83 996.754 Q1552 1000.76 1559.41 1000.76 Q1563.7 1000.76 1567.72 999.711 Q1571.77 998.658 1575.74 996.552 L1575.74 1003.6 Q1571.73 1005.3 1567.51 1006.19 Q1563.3 1007.08 1558.97 1007.08 Q1548.11 1007.08 1541.75 1000.76 Q1535.43 994.445 1535.43 983.67 Q1535.43 972.53 1541.42 966.008 Q1547.46 959.445 1557.67 959.445 Q1566.82 959.445 1572.13 965.36 Q1577.48 971.234 1577.48 981.361 M1570.02 979.173 Q1569.94 973.057 1566.58 969.411 Q1563.26 965.765 1557.75 965.765 Q1551.51 965.765 1547.74 969.289 Q1544.02 972.813 1543.45 979.214 L1570.02 979.173 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1633.99 942.958 Q1628.56 952.275 1625.93 961.39 Q1623.29 970.504 1623.29 979.862 Q1623.29 989.22 1625.93 998.415 Q1628.6 1007.57 1633.99 1016.85 L1627.51 1016.85 Q1621.43 1007.33 1618.39 998.132 Q1615.39 988.936 1615.39 979.862 Q1615.39 970.829 1618.39 961.673 Q1621.39 952.518 1627.51 942.958 L1633.99 942.958 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1678.3 967.426 L1678.3 942.877 L1685.76 942.877 L1685.76 1005.91 L1678.3 1005.91 L1678.3 999.104 Q1675.96 1003.15 1672.35 1005.14 Q1668.79 1007.08 1663.76 1007.08 Q1655.54 1007.08 1650.35 1000.52 Q1645.21 993.959 1645.21 983.265 Q1645.21 972.57 1650.35 966.008 Q1655.54 959.445 1663.76 959.445 Q1668.79 959.445 1672.35 961.43 Q1675.96 963.375 1678.3 967.426 M1652.91 983.265 Q1652.91 991.488 1656.27 996.187 Q1659.67 1000.85 1665.59 1000.85 Q1671.5 1000.85 1674.9 996.187 Q1678.3 991.488 1678.3 983.265 Q1678.3 975.041 1674.9 970.383 Q1671.5 965.684 1665.59 965.684 Q1659.67 965.684 1656.27 970.383 Q1652.91 975.041 1652.91 983.265 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1739.92 981.361 L1739.92 985.007 L1705.65 985.007 Q1706.13 992.703 1710.27 996.754 Q1714.44 1000.76 1721.85 1000.76 Q1726.15 1000.76 1730.16 999.711 Q1734.21 998.658 1738.18 996.552 L1738.18 1003.6 Q1734.17 1005.3 1729.95 1006.19 Q1725.74 1007.08 1721.41 1007.08 Q1710.55 1007.08 1704.19 1000.76 Q1697.87 994.445 1697.87 983.67 Q1697.87 972.53 1703.87 966.008 Q1709.9 959.445 1720.11 959.445 Q1729.27 959.445 1734.57 965.36 Q1739.92 971.234 1739.92 981.361 M1732.47 979.173 Q1732.38 973.057 1729.02 969.411 Q1725.7 965.765 1720.19 965.765 Q1713.95 965.765 1710.19 969.289 Q1706.46 972.813 1705.89 979.214 L1732.47 979.173 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1782.01 982.698 Q1782.01 974.596 1778.65 970.14 Q1775.32 965.684 1769.29 965.684 Q1763.29 965.684 1759.93 970.14 Q1756.61 974.596 1756.61 982.698 Q1756.61 990.759 1759.93 995.215 Q1763.29 999.671 1769.29 999.671 Q1775.32 999.671 1778.65 995.215 Q1782.01 990.759 1782.01 982.698 M1789.46 1000.28 Q1789.46 1011.86 1784.32 1017.49 Q1779.17 1023.17 1768.56 1023.17 Q1764.63 1023.17 1761.15 1022.56 Q1757.66 1021.99 1754.38 1020.78 L1754.38 1013.53 Q1757.66 1015.31 1760.86 1016.16 Q1764.06 1017.01 1767.38 1017.01 Q1774.72 1017.01 1778.36 1013.16 Q1782.01 1009.35 1782.01 1001.62 L1782.01 997.929 Q1779.7 1001.94 1776.09 1003.92 Q1772.49 1005.91 1767.47 1005.91 Q1759.12 1005.91 1754.02 999.549 Q1748.91 993.189 1748.91 982.698 Q1748.91 972.165 1754.02 965.805 Q1759.12 959.445 1767.47 959.445 Q1772.49 959.445 1776.09 961.43 Q1779.7 963.415 1782.01 967.426 L1782.01 960.539 L1789.46 960.539 L1789.46 1000.28 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1803.64 942.958 L1810.12 942.958 Q1816.2 952.518 1819.2 961.673 Q1822.23 970.829 1822.23 979.862 Q1822.23 988.936 1819.2 998.132 Q1816.2 1007.33 1810.12 1016.85 L1803.64 1016.85 Q1809.03 1007.57 1811.66 998.415 Q1814.33 989.22 1814.33 979.862 Q1814.33 970.504 1811.66 961.39 Q1809.03 952.275 1803.64 942.958 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip693)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="261.245,1705.81 266.073,1705.04 270.902,1704.26 275.73,1703.49 280.558,1702.72 285.386,1701.96 290.214,1701.2 295.043,1700.44 299.871,1699.68 304.699,1698.94 309.527,1698.19 314.356,1697.46 319.184,1696.74 324.012,1696.02 328.84,1695.31 333.668,1694.62 338.497,1693.94 343.325,1693.27 348.153,1692.62 352.981,1691.98 357.81,1691.37 362.638,1690.77 367.466,1690.19 372.294,1689.63 377.123,1689.1 381.951,1688.59 386.779,1688.11 391.607,1687.65 396.435,1687.22 401.264,1686.81 406.092,1686.44 410.92,1686.09 415.748,1685.77 420.577,1685.47 425.405,1685.21 430.233,1684.96 435.061,1684.74 439.89,1684.54 444.718,1684.36 449.546,1684.2 454.374,1684.05 459.202,1683.91 464.031,1683.78 468.859,1683.65 473.687,1683.51 478.515,1683.37 483.344,1683.22 488.172,1683.06 493,1682.87 497.828,1682.67 502.656,1682.43 507.485,1682.17 512.313,1681.87 517.141,1681.54 521.969,1681.16 526.798,1680.75 531.626,1680.28 536.454,1679.77 541.282,1679.22 546.111,1678.61 550.939,1677.95 555.767,1677.25 560.595,1676.49 565.423,1675.68 570.252,1674.82 575.08,1673.92 579.908,1672.96 584.736,1671.96 589.565,1670.9 594.393,1669.81 599.221,1668.67 604.049,1667.48 608.878,1666.26 613.706,1664.99 618.534,1663.68 623.362,1662.34 628.19,1660.97 633.019,1659.55 637.847,1658.11 642.675,1656.64 647.503,1655.14 652.332,1653.61 657.16,1652.05 661.988,1650.47 666.816,1648.87 671.645,1647.24 676.473,1645.6 681.301,1643.94 686.129,1642.26 690.957,1640.56 695.786,1638.85 700.614,1637.13 705.442,1635.4 710.27,1633.66 715.099,1631.91 719.927,1630.15 724.755,1628.39 729.583,1626.62 734.411,1624.86 739.24,1623.09 744.068,1621.32 748.896,1619.55 753.724,1617.79 758.553,1616.03 763.381,1614.28 768.209,1612.54 773.037,1610.81 777.866,1609.1 782.694,1607.4 787.522,1605.71 792.35,1604.04 797.178,1602.39 802.007,1600.77 806.835,1599.17 811.663,1597.6 816.491,1596.05 821.32,1594.54 826.148,1593.06 830.976,1591.62 835.804,1590.21 840.633,1588.85 845.461,1587.52 850.289,1586.24 855.117,1585 859.945,1583.81 864.774,1582.67 869.602,1581.58 874.43,1580.54 879.258,1579.54 884.087,1578.59 888.915,1577.7 893.743,1576.85 898.571,1576.04 903.399,1575.28 908.228,1574.56 913.056,1573.87 917.884,1573.22 922.712,1572.6 927.541,1572 932.369,1571.42 937.197,1570.85 942.025,1570.29 946.854,1569.74 951.682,1569.18 956.51,1568.61 961.338,1568.02 966.166,1567.41 970.995,1566.78 975.823,1566.11 980.651,1565.41 985.479,1564.66 990.308,1563.88 995.136,1563.05 999.964,1562.16 1004.79,1561.23 1009.62,1560.24 1014.45,1559.2 1019.28,1558.11 1024.11,1556.96 1028.93,1555.76 1033.76,1554.51 1038.59,1553.2 1043.42,1551.85 1048.25,1550.45 1053.07,1549 1057.9,1547.51 1062.73,1545.98 1067.56,1544.42 1072.39,1542.82 1077.22,1541.18 1082.04,1539.52 1086.87,1537.84 1091.7,1536.13 1096.53,1534.4 1101.36,1532.66 1106.19,1530.91 1111.01,1529.15 1115.84,1527.38 1120.67,1525.62 1125.5,1523.85 1130.33,1522.1 1135.15,1520.35 1139.98,1518.61 1144.81,1516.89 1149.64,1515.19 1154.47,1513.51 1159.3,1511.86 1164.12,1510.23 1168.95,1508.64 1173.78,1507.09 1178.61,1505.57 1183.44,1504.09 1188.26,1502.65 1193.09,1501.26 1197.92,1499.91 1202.75,1498.61 1207.58,1497.36 1212.41,1496.16 1217.23,1495.01 1222.06,1493.91 1226.89,1492.85 1231.72,1491.84 1236.55,1490.87 1241.38,1489.94 1246.2,1489.05 1251.03,1488.2 1255.86,1487.37 1260.69,1486.57 1265.52,1485.79 1270.34,1485.03 1275.17,1484.27 1280,1483.51 1284.83,1482.75 1289.66,1481.99 1294.49,1481.2 1299.31,1480.4 1304.14,1479.57 1308.97,1478.72 1313.8,1477.83 1318.63,1476.9 1323.46,1475.93 1328.28,1474.92 1333.11,1473.86 1337.94,1472.76 1342.77,1471.61 1347.6,1470.42 1352.42,1469.17 1357.25,1467.89 1362.08,1466.55 1366.91,1465.18 1371.74,1463.77 1376.57,1462.32 1381.39,1460.83 1386.22,1459.32 1391.05,1457.78 1395.88,1456.21 1400.71,1454.62 1405.54,1453.02 1410.36,1451.4 1415.19,1449.78 1420.02,1448.15 1424.85,1446.52 1429.68,1444.9 1434.5,1443.28 1439.33,1441.68 1444.16,1440.08 1448.99,1438.51 1453.82,1436.96 1458.65,1435.44 1463.47,1433.94 1468.3,1432.47 1473.13,1431.04 1477.96,1429.64 1482.79,1428.27 1487.62,1426.95 1492.44,1425.66 1497.27,1424.41 1502.1,1423.2 1506.93,1422.03 1511.76,1420.9 1516.58,1419.8 1521.41,1418.74 1526.24,1417.7 1531.07,1416.7 1535.9,1415.72 1540.73,1414.75 1545.55,1413.81 1550.38,1412.87 1555.21,1411.95 1560.04,1411.02 1564.87,1410.1 1569.7,1409.16 1574.52,1408.22 1579.35,1407.26 1584.18,1406.28 1589.01,1405.28 1593.84,1404.25 1598.66,1403.2 1603.49,1402.11 1608.32,1400.99 1613.15,1399.84 1617.98,1398.66 1622.81,1397.44 1627.63,1396.19 1632.46,1394.9 1637.29,1393.59 1642.12,1392.25 1646.95,1390.88 1651.77,1389.48 1656.6,1388.06 1661.43,1386.63 1666.26,1385.18 1671.09,1383.71 1675.92,1382.24 1680.74,1380.76 1685.57,1379.28 1690.4,1377.8 1695.23,1376.32 1700.06,1374.85 1704.89,1373.39 1709.71,1371.94 1714.54,1370.51 1719.37,1369.1 1724.2,1367.7 1729.03,1366.33 1733.85,1364.98 1738.68,1363.66 1743.51,1362.36 1748.34,1361.09 1753.17,1359.84 1758,1358.62 1762.82,1357.43 1767.65,1356.26 1772.48,1355.11 1777.31,1353.98 1782.14,1352.88 1786.97,1351.78 1791.79,1350.71 1796.62,1349.64 1801.45,1348.58 1806.28,1347.53 1811.11,1346.48 1815.93,1345.43 1820.76,1344.37 1825.59,1343.31 1830.42,1342.23 1835.25,1341.15 1840.08,1340.05 1844.9,1338.93 1849.73,1337.8 1854.56,1336.65 1859.39,1335.47 1864.22,1334.28 1869.05,1333.07 1873.87,1331.84 1878.7,1330.58 1883.53,1329.32 1888.36,1328.03 1893.19,1326.73 1898.01,1325.41 1902.84,1324.08 1907.67,1322.74 1912.5,1321.39 1917.33,1320.04 1922.16,1318.68 1926.98,1317.33 1931.81,1315.97 1936.64,1314.61 1941.47,1313.26 1946.3,1311.92 1951.13,1310.58 1955.95,1309.25 1960.78,1307.94 1965.61,1306.63 1970.44,1305.35 1975.27,1304.07 1980.09,1302.81 1984.92,1301.56 1989.75,1300.33 1994.58,1299.11 1999.41,1297.91 2004.24,1296.72 2009.06,1295.54 2013.89,1294.37 2018.72,1293.21 2023.55,1292.06 2028.38,1290.92 2033.2,1289.78 2038.03,1288.65 2042.86,1287.51 2047.69,1286.38 2052.52,1285.25 2057.35,1284.11 2062.17,1282.97 2067,1281.83 2071.83,1280.68 2076.66,1279.52 2081.49,1278.35 2086.32,1277.18 2091.14,1275.99 2095.97,1274.8 2100.8,1273.6 2105.63,1272.39 2110.46,1271.17 2115.28,1269.94 2120.11,1268.7 2124.94,1267.46 2129.77,1266.22 2134.6,1264.97 2139.43,1263.71 2144.25,1262.45 2149.08,1261.19 2153.91,1259.93 2158.74,1258.67 2163.57,1257.41 2168.4,1256.16 2173.22,1254.9 2178.05,1253.65 2182.88,1252.4 2187.71,1251.16 2192.54,1249.92 2197.36,1248.69 2202.19,1247.46 2207.02,1246.24 2211.85,1245.02 2216.68,1243.81 2221.51,1242.6 2226.33,1241.4 2231.16,1240.21 2235.99,1239.01 2240.82,1237.82 2245.65,1236.64 2250.48,1235.46 2255.3,1234.28 2260.13,1233.1 2264.96,1231.92 2269.79,1230.75 2274.62,1229.58 2279.44,1228.4 2284.27,1227.23 2289.1,1226.06 2293.93,1224.89 2298.76,1223.71 2303.59,1222.54 2308.41,1221.37 2313.24,1220.2 2318.07,1219.02 2322.9,1217.85 2327.73,1216.67 2332.56,1215.5 2337.38,1214.32 2342.21,1213.15 2347.04,1211.97 2351.87,1210.79 2356.7,1209.61 2361.52,1208.43 2366.35,1207.25 2371.18,1206.06 2376.01,1204.87 2380.84,1203.68 2385.67,1202.49 2390.49,1201.29 2395.32,1200.09 2400.15,1198.89 2404.98,1197.68 2409.81,1196.46 2414.64,1195.24 2419.46,1194.02 2424.29,1192.79 2429.12,1191.55 2433.95,1190.31 2438.78,1189.06 2443.6,1187.81 2448.43,1186.55 2453.26,1185.28 2458.09,1184.02 2462.92,1182.75 2467.75,1181.47 2472.57,1180.2 2477.4,1178.93 2482.23,1177.65 2487.06,1176.38 2491.89,1175.1 2496.71,1173.83 2501.54,1172.57 2506.37,1171.3 2511.2,1170.04 2516.03,1168.79 2520.86,1167.53 2525.68,1166.28 2530.51,1165.03 2535.34,1163.79 2540.17,1162.54 2545,1161.28 2549.83,1160.02 2554.65,1158.76 2559.48,1157.48 2564.31,1156.18 2569.14,1154.86 2573.97,1153.52 2578.79,1152.15 2583.62,1150.74 2588.45,1149.28 2593.28,1147.77 2598.11,1146.2 2602.94,1144.57 2607.76,1142.87 2612.59,1141.08 2617.42,1139.2 2622.25,1137.23 2627.08,1135.16 2631.91,1132.98 2636.73,1130.69 2641.56,1128.29 2646.39,1125.77 2651.22,1123.14 2656.05,1120.39 2660.87,1117.54 2665.7,1114.57 2670.53,1111.5 2675.36,1108.33 2680.19,1105.07 "/>
+<path clip-path="url(#clip690)" d="M188.677 2657.17 L2752.76 2657.17 L2752.76 2020.38 L188.677 2020.38  Z" fill="#ffffff" fill-rule="evenodd" fill-opacity="1"/>
+<defs>
+  <clipPath id="clip694">
+    <rect x="188" y="2020" width="2565" height="638"/>
+  </clipPath>
+</defs>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="261.245,2657.17 261.245,2020.38 "/>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="863.571,2657.17 863.571,2020.38 "/>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="1465.9,2657.17 1465.9,2020.38 "/>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2068.22,2657.17 2068.22,2020.38 "/>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="2670.55,2657.17 2670.55,2020.38 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,2657.17 2752.76,2657.17 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="261.245,2657.17 261.245,2638.27 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="863.571,2657.17 863.571,2638.27 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="1465.9,2657.17 1465.9,2638.27 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2068.22,2657.17 2068.22,2638.27 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="2670.55,2657.17 2670.55,2638.27 "/>
+<path clip-path="url(#clip690)" d="M261.245 2691.29 Q257.634 2691.29 255.805 2694.85 Q254 2698.39 254 2705.52 Q254 2712.63 255.805 2716.19 Q257.634 2719.74 261.245 2719.74 Q264.879 2719.74 266.685 2716.19 Q268.514 2712.63 268.514 2705.52 Q268.514 2698.39 266.685 2694.85 Q264.879 2691.29 261.245 2691.29 M261.245 2687.58 Q267.055 2687.58 270.111 2692.19 Q273.189 2696.77 273.189 2705.52 Q273.189 2714.25 270.111 2718.86 Q267.055 2723.44 261.245 2723.44 Q255.435 2723.44 252.356 2718.86 Q249.301 2714.25 249.301 2705.52 Q249.301 2696.77 252.356 2692.19 Q255.435 2687.58 261.245 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M823.19 2688.21 L841.546 2688.21 L841.546 2692.14 L827.472 2692.14 L827.472 2700.62 Q828.491 2700.27 829.509 2700.11 Q830.528 2699.92 831.546 2699.92 Q837.333 2699.92 840.713 2703.09 Q844.092 2706.26 844.092 2711.68 Q844.092 2717.26 840.62 2720.36 Q837.148 2723.44 830.829 2723.44 Q828.653 2723.44 826.384 2723.07 Q824.139 2722.7 821.731 2721.96 L821.731 2717.26 Q823.815 2718.39 826.037 2718.95 Q828.259 2719.5 830.736 2719.5 Q834.741 2719.5 837.078 2717.4 Q839.416 2715.29 839.416 2711.68 Q839.416 2708.07 837.078 2705.96 Q834.741 2703.86 830.736 2703.86 Q828.861 2703.86 826.986 2704.27 Q825.134 2704.69 823.19 2705.57 L823.19 2688.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M863.305 2691.29 Q859.694 2691.29 857.865 2694.85 Q856.06 2698.39 856.06 2705.52 Q856.06 2712.63 857.865 2716.19 Q859.694 2719.74 863.305 2719.74 Q866.939 2719.74 868.745 2716.19 Q870.574 2712.63 870.574 2705.52 Q870.574 2698.39 868.745 2694.85 Q866.939 2691.29 863.305 2691.29 M863.305 2687.58 Q869.115 2687.58 872.171 2692.19 Q875.25 2696.77 875.25 2705.52 Q875.25 2714.25 872.171 2718.86 Q869.115 2723.44 863.305 2723.44 Q857.495 2723.44 854.416 2718.86 Q851.361 2714.25 851.361 2705.52 Q851.361 2696.77 854.416 2692.19 Q857.495 2687.58 863.305 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M893.467 2691.29 Q889.856 2691.29 888.027 2694.85 Q886.222 2698.39 886.222 2705.52 Q886.222 2712.63 888.027 2716.19 Q889.856 2719.74 893.467 2719.74 Q897.101 2719.74 898.907 2716.19 Q900.736 2712.63 900.736 2705.52 Q900.736 2698.39 898.907 2694.85 Q897.101 2691.29 893.467 2691.29 M893.467 2687.58 Q899.277 2687.58 902.333 2692.19 Q905.411 2696.77 905.411 2705.52 Q905.411 2714.25 902.333 2718.86 Q899.277 2723.44 893.467 2723.44 Q887.657 2723.44 884.578 2718.86 Q881.523 2714.25 881.523 2705.52 Q881.523 2696.77 884.578 2692.19 Q887.657 2687.58 893.467 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1410.42 2718.83 L1418.06 2718.83 L1418.06 2692.47 L1409.75 2694.13 L1409.75 2689.87 L1418.02 2688.21 L1422.69 2688.21 L1422.69 2718.83 L1430.33 2718.83 L1430.33 2722.77 L1410.42 2722.77 L1410.42 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1449.78 2691.29 Q1446.16 2691.29 1444.34 2694.85 Q1442.53 2698.39 1442.53 2705.52 Q1442.53 2712.63 1444.34 2716.19 Q1446.16 2719.74 1449.78 2719.74 Q1453.41 2719.74 1455.21 2716.19 Q1457.04 2712.63 1457.04 2705.52 Q1457.04 2698.39 1455.21 2694.85 Q1453.41 2691.29 1449.78 2691.29 M1449.78 2687.58 Q1455.59 2687.58 1458.64 2692.19 Q1461.72 2696.77 1461.72 2705.52 Q1461.72 2714.25 1458.64 2718.86 Q1455.59 2723.44 1449.78 2723.44 Q1443.96 2723.44 1440.89 2718.86 Q1437.83 2714.25 1437.83 2705.52 Q1437.83 2696.77 1440.89 2692.19 Q1443.96 2687.58 1449.78 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1479.94 2691.29 Q1476.33 2691.29 1474.5 2694.85 Q1472.69 2698.39 1472.69 2705.52 Q1472.69 2712.63 1474.5 2716.19 Q1476.33 2719.74 1479.94 2719.74 Q1483.57 2719.74 1485.38 2716.19 Q1487.21 2712.63 1487.21 2705.52 Q1487.21 2698.39 1485.38 2694.85 Q1483.57 2691.29 1479.94 2691.29 M1479.94 2687.58 Q1485.75 2687.58 1488.8 2692.19 Q1491.88 2696.77 1491.88 2705.52 Q1491.88 2714.25 1488.8 2718.86 Q1485.75 2723.44 1479.94 2723.44 Q1474.13 2723.44 1471.05 2718.86 Q1467.99 2714.25 1467.99 2705.52 Q1467.99 2696.77 1471.05 2692.19 Q1474.13 2687.58 1479.94 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1510.1 2691.29 Q1506.49 2691.29 1504.66 2694.85 Q1502.85 2698.39 1502.85 2705.52 Q1502.85 2712.63 1504.66 2716.19 Q1506.49 2719.74 1510.1 2719.74 Q1513.73 2719.74 1515.54 2716.19 Q1517.37 2712.63 1517.37 2705.52 Q1517.37 2698.39 1515.54 2694.85 Q1513.73 2691.29 1510.1 2691.29 M1510.1 2687.58 Q1515.91 2687.58 1518.96 2692.19 Q1522.04 2696.77 1522.04 2705.52 Q1522.04 2714.25 1518.96 2718.86 Q1515.91 2723.44 1510.1 2723.44 Q1504.29 2723.44 1501.21 2718.86 Q1498.15 2714.25 1498.15 2705.52 Q1498.15 2696.77 1501.21 2692.19 Q1504.29 2687.58 1510.1 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2012.75 2718.83 L2020.39 2718.83 L2020.39 2692.47 L2012.08 2694.13 L2012.08 2689.87 L2020.34 2688.21 L2025.02 2688.21 L2025.02 2718.83 L2032.66 2718.83 L2032.66 2722.77 L2012.75 2722.77 L2012.75 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2042.15 2688.21 L2060.5 2688.21 L2060.5 2692.14 L2046.43 2692.14 L2046.43 2700.62 Q2047.45 2700.27 2048.47 2700.11 Q2049.49 2699.92 2050.5 2699.92 Q2056.29 2699.92 2059.67 2703.09 Q2063.05 2706.26 2063.05 2711.68 Q2063.05 2717.26 2059.58 2720.36 Q2056.11 2723.44 2049.79 2723.44 Q2047.61 2723.44 2045.34 2723.07 Q2043.1 2722.7 2040.69 2721.96 L2040.69 2717.26 Q2042.77 2718.39 2045 2718.95 Q2047.22 2719.5 2049.69 2719.5 Q2053.7 2719.5 2056.04 2717.4 Q2058.37 2715.29 2058.37 2711.68 Q2058.37 2708.07 2056.04 2705.96 Q2053.7 2703.86 2049.69 2703.86 Q2047.82 2703.86 2045.94 2704.27 Q2044.09 2704.69 2042.15 2705.57 L2042.15 2688.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2082.26 2691.29 Q2078.65 2691.29 2076.82 2694.85 Q2075.02 2698.39 2075.02 2705.52 Q2075.02 2712.63 2076.82 2716.19 Q2078.65 2719.74 2082.26 2719.74 Q2085.9 2719.74 2087.7 2716.19 Q2089.53 2712.63 2089.53 2705.52 Q2089.53 2698.39 2087.7 2694.85 Q2085.9 2691.29 2082.26 2691.29 M2082.26 2687.58 Q2088.07 2687.58 2091.13 2692.19 Q2094.21 2696.77 2094.21 2705.52 Q2094.21 2714.25 2091.13 2718.86 Q2088.07 2723.44 2082.26 2723.44 Q2076.45 2723.44 2073.37 2718.86 Q2070.32 2714.25 2070.32 2705.52 Q2070.32 2696.77 2073.37 2692.19 Q2076.45 2687.58 2082.26 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2112.43 2691.29 Q2108.81 2691.29 2106.99 2694.85 Q2105.18 2698.39 2105.18 2705.52 Q2105.18 2712.63 2106.99 2716.19 Q2108.81 2719.74 2112.43 2719.74 Q2116.06 2719.74 2117.87 2716.19 Q2119.69 2712.63 2119.69 2705.52 Q2119.69 2698.39 2117.87 2694.85 Q2116.06 2691.29 2112.43 2691.29 M2112.43 2687.58 Q2118.24 2687.58 2121.29 2692.19 Q2124.37 2696.77 2124.37 2705.52 Q2124.37 2714.25 2121.29 2718.86 Q2118.24 2723.44 2112.43 2723.44 Q2106.62 2723.44 2103.54 2718.86 Q2100.48 2714.25 2100.48 2705.52 Q2100.48 2696.77 2103.54 2692.19 Q2106.62 2687.58 2112.43 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2619.16 2718.83 L2635.48 2718.83 L2635.48 2722.77 L2613.54 2722.77 L2613.54 2718.83 Q2616.2 2716.08 2620.78 2711.45 Q2625.39 2706.8 2626.57 2705.45 Q2628.81 2702.93 2629.69 2701.19 Q2630.6 2699.43 2630.6 2697.74 Q2630.6 2694.99 2628.65 2693.25 Q2626.73 2691.52 2623.63 2691.52 Q2621.43 2691.52 2618.98 2692.28 Q2616.55 2693.05 2613.77 2694.6 L2613.77 2689.87 Q2616.59 2688.74 2619.05 2688.16 Q2621.5 2687.58 2623.54 2687.58 Q2628.91 2687.58 2632.1 2690.27 Q2635.3 2692.95 2635.3 2697.44 Q2635.3 2699.57 2634.49 2701.49 Q2633.7 2703.39 2631.59 2705.99 Q2631.01 2706.66 2627.91 2709.87 Q2624.81 2713.07 2619.16 2718.83 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2655.3 2691.29 Q2651.68 2691.29 2649.86 2694.85 Q2648.05 2698.39 2648.05 2705.52 Q2648.05 2712.63 2649.86 2716.19 Q2651.68 2719.74 2655.3 2719.74 Q2658.93 2719.74 2660.74 2716.19 Q2662.56 2712.63 2662.56 2705.52 Q2662.56 2698.39 2660.74 2694.85 Q2658.93 2691.29 2655.3 2691.29 M2655.3 2687.58 Q2661.11 2687.58 2664.16 2692.19 Q2667.24 2696.77 2667.24 2705.52 Q2667.24 2714.25 2664.16 2718.86 Q2661.11 2723.44 2655.3 2723.44 Q2649.49 2723.44 2646.41 2718.86 Q2643.35 2714.25 2643.35 2705.52 Q2643.35 2696.77 2646.41 2692.19 Q2649.49 2687.58 2655.3 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2685.46 2691.29 Q2681.85 2691.29 2680.02 2694.85 Q2678.21 2698.39 2678.21 2705.52 Q2678.21 2712.63 2680.02 2716.19 Q2681.85 2719.74 2685.46 2719.74 Q2689.09 2719.74 2690.9 2716.19 Q2692.73 2712.63 2692.73 2705.52 Q2692.73 2698.39 2690.9 2694.85 Q2689.09 2691.29 2685.46 2691.29 M2685.46 2687.58 Q2691.27 2687.58 2694.32 2692.19 Q2697.4 2696.77 2697.4 2705.52 Q2697.4 2714.25 2694.32 2718.86 Q2691.27 2723.44 2685.46 2723.44 Q2679.65 2723.44 2676.57 2718.86 Q2673.51 2714.25 2673.51 2705.52 Q2673.51 2696.77 2676.57 2692.19 Q2679.65 2687.58 2685.46 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M2715.62 2691.29 Q2712.01 2691.29 2710.18 2694.85 Q2708.37 2698.39 2708.37 2705.52 Q2708.37 2712.63 2710.18 2716.19 Q2712.01 2719.74 2715.62 2719.74 Q2719.25 2719.74 2721.06 2716.19 Q2722.89 2712.63 2722.89 2705.52 Q2722.89 2698.39 2721.06 2694.85 Q2719.25 2691.29 2715.62 2691.29 M2715.62 2687.58 Q2721.43 2687.58 2724.49 2692.19 Q2727.56 2696.77 2727.56 2705.52 Q2727.56 2714.25 2724.49 2718.86 Q2721.43 2723.44 2715.62 2723.44 Q2709.81 2723.44 2706.73 2718.86 Q2703.68 2714.25 2703.68 2705.52 Q2703.68 2696.77 2706.73 2692.19 Q2709.81 2687.58 2715.62 2687.58 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,2645.99 2752.76,2645.99 "/>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,2466.41 2752.76,2466.41 "/>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,2286.84 2752.76,2286.84 "/>
+<polyline clip-path="url(#clip694)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:2; stroke-opacity:0.1; fill:none" points="188.677,2107.27 2752.76,2107.27 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,2657.17 188.677,2020.38 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,2645.99 207.574,2645.99 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,2466.41 207.574,2466.41 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,2286.84 207.574,2286.84 "/>
+<polyline clip-path="url(#clip690)" style="stroke:#000000; stroke-linecap:round; stroke-linejoin:round; stroke-width:4; stroke-opacity:1; fill:none" points="188.677,2107.27 207.574,2107.27 "/>
+<path clip-path="url(#clip690)" d="M134.732 2631.79 Q131.121 2631.79 129.293 2635.35 Q127.487 2638.89 127.487 2646.02 Q127.487 2653.13 129.293 2656.69 Q131.121 2660.23 134.732 2660.23 Q138.367 2660.23 140.172 2656.69 Q142.001 2653.13 142.001 2646.02 Q142.001 2638.89 140.172 2635.35 Q138.367 2631.79 134.732 2631.79 M134.732 2628.08 Q140.543 2628.08 143.598 2632.69 Q146.677 2637.27 146.677 2646.02 Q146.677 2654.75 143.598 2659.35 Q140.543 2663.94 134.732 2663.94 Q128.922 2663.94 125.844 2659.35 Q122.788 2654.75 122.788 2646.02 Q122.788 2637.27 125.844 2632.69 Q128.922 2628.08 134.732 2628.08 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M94.6169 2449.13 L112.973 2449.13 L112.973 2453.07 L98.8993 2453.07 L98.8993 2461.54 Q99.9178 2461.19 100.936 2461.03 Q101.955 2460.85 102.973 2460.85 Q108.76 2460.85 112.14 2464.02 Q115.52 2467.19 115.52 2472.61 Q115.52 2478.18 112.047 2481.29 Q108.575 2484.36 102.256 2484.36 Q100.08 2484.36 97.8113 2483.99 Q95.566 2483.62 93.1586 2482.88 L93.1586 2478.18 Q95.2419 2479.32 97.4641 2479.87 Q99.6863 2480.43 102.163 2480.43 Q106.168 2480.43 108.506 2478.32 Q110.844 2476.22 110.844 2472.61 Q110.844 2468.99 108.506 2466.89 Q106.168 2464.78 102.163 2464.78 Q100.288 2464.78 98.4132 2465.2 Q96.5613 2465.62 94.6169 2466.49 L94.6169 2449.13 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 2452.21 Q131.121 2452.21 129.293 2455.78 Q127.487 2459.32 127.487 2466.45 Q127.487 2473.55 129.293 2477.12 Q131.121 2480.66 134.732 2480.66 Q138.367 2480.66 140.172 2477.12 Q142.001 2473.55 142.001 2466.45 Q142.001 2459.32 140.172 2455.78 Q138.367 2452.21 134.732 2452.21 M134.732 2448.51 Q140.543 2448.51 143.598 2453.12 Q146.677 2457.7 146.677 2466.45 Q146.677 2475.18 143.598 2479.78 Q140.543 2484.36 134.732 2484.36 Q128.922 2484.36 125.844 2479.78 Q122.788 2475.18 122.788 2466.45 Q122.788 2457.7 125.844 2453.12 Q128.922 2448.51 134.732 2448.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M65.2189 2300.19 L72.8577 2300.19 L72.8577 2273.82 L64.5476 2275.49 L64.5476 2271.23 L72.8114 2269.56 L77.4873 2269.56 L77.4873 2300.19 L85.1262 2300.19 L85.1262 2304.12 L65.2189 2304.12 L65.2189 2300.19 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M104.571 2272.64 Q100.959 2272.64 99.1307 2276.2 Q97.3252 2279.75 97.3252 2286.88 Q97.3252 2293.98 99.1307 2297.55 Q100.959 2301.09 104.571 2301.09 Q108.205 2301.09 110.01 2297.55 Q111.839 2293.98 111.839 2286.88 Q111.839 2279.75 110.01 2276.2 Q108.205 2272.64 104.571 2272.64 M104.571 2268.94 Q110.381 2268.94 113.436 2273.54 Q116.515 2278.13 116.515 2286.88 Q116.515 2295.6 113.436 2300.21 Q110.381 2304.79 104.571 2304.79 Q98.7604 2304.79 95.6817 2300.21 Q92.6262 2295.6 92.6262 2286.88 Q92.6262 2278.13 95.6817 2273.54 Q98.7604 2268.94 104.571 2268.94 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 2272.64 Q131.121 2272.64 129.293 2276.2 Q127.487 2279.75 127.487 2286.88 Q127.487 2293.98 129.293 2297.55 Q131.121 2301.09 134.732 2301.09 Q138.367 2301.09 140.172 2297.55 Q142.001 2293.98 142.001 2286.88 Q142.001 2279.75 140.172 2276.2 Q138.367 2272.64 134.732 2272.64 M134.732 2268.94 Q140.543 2268.94 143.598 2273.54 Q146.677 2278.13 146.677 2286.88 Q146.677 2295.6 143.598 2300.21 Q140.543 2304.79 134.732 2304.79 Q128.922 2304.79 125.844 2300.21 Q122.788 2295.6 122.788 2286.88 Q122.788 2278.13 125.844 2273.54 Q128.922 2268.94 134.732 2268.94 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M65.2189 2120.61 L72.8577 2120.61 L72.8577 2094.25 L64.5476 2095.91 L64.5476 2091.65 L72.8114 2089.99 L77.4873 2089.99 L77.4873 2120.61 L85.1262 2120.61 L85.1262 2124.55 L65.2189 2124.55 L65.2189 2120.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M94.6169 2089.99 L112.973 2089.99 L112.973 2093.92 L98.8993 2093.92 L98.8993 2102.39 Q99.9178 2102.05 100.936 2101.89 Q101.955 2101.7 102.973 2101.7 Q108.76 2101.7 112.14 2104.87 Q115.52 2108.04 115.52 2113.46 Q115.52 2119.04 112.047 2122.14 Q108.575 2125.22 102.256 2125.22 Q100.08 2125.22 97.8113 2124.85 Q95.566 2124.48 93.1586 2123.74 L93.1586 2119.04 Q95.2419 2120.17 97.4641 2120.73 Q99.6863 2121.28 102.163 2121.28 Q106.168 2121.28 108.506 2119.18 Q110.844 2117.07 110.844 2113.46 Q110.844 2109.85 108.506 2107.74 Q106.168 2105.64 102.163 2105.64 Q100.288 2105.64 98.4132 2106.05 Q96.5613 2106.47 94.6169 2107.35 L94.6169 2089.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M134.732 2093.07 Q131.121 2093.07 129.293 2096.63 Q127.487 2100.17 127.487 2107.3 Q127.487 2114.41 129.293 2117.97 Q131.121 2121.52 134.732 2121.52 Q138.367 2121.52 140.172 2117.97 Q142.001 2114.41 142.001 2107.3 Q142.001 2100.17 140.172 2096.63 Q138.367 2093.07 134.732 2093.07 M134.732 2089.36 Q140.543 2089.36 143.598 2093.97 Q146.677 2098.55 146.677 2107.3 Q146.677 2116.03 143.598 2120.64 Q140.543 2125.22 134.732 2125.22 Q128.922 2125.22 125.844 2120.64 Q122.788 2116.03 122.788 2107.3 Q122.788 2098.55 125.844 2093.97 Q128.922 2089.36 134.732 2089.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1013.96 1878.76 L1022.14 1878.76 L1022.14 1903.55 L1051.87 1903.55 L1051.87 1878.76 L1060.06 1878.76 L1060.06 1939.24 L1051.87 1939.24 L1051.87 1910.44 L1022.14 1910.44 L1022.14 1939.24 L1013.96 1939.24 L1013.96 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1114.82 1914.69 L1114.82 1918.34 L1080.55 1918.34 Q1081.04 1926.04 1085.17 1930.09 Q1089.34 1934.1 1096.76 1934.1 Q1101.05 1934.1 1105.06 1933.04 Q1109.11 1931.99 1113.08 1929.89 L1113.08 1936.93 Q1109.07 1938.64 1104.86 1939.53 Q1100.65 1940.42 1096.31 1940.42 Q1085.46 1940.42 1079.1 1934.1 Q1072.78 1927.78 1072.78 1917 Q1072.78 1905.86 1078.77 1899.34 Q1084.81 1892.78 1095.02 1892.78 Q1104.17 1892.78 1109.48 1898.69 Q1114.82 1904.57 1114.82 1914.69 M1107.37 1912.51 Q1107.29 1906.39 1103.93 1902.74 Q1100.61 1899.1 1095.1 1899.1 Q1088.86 1899.1 1085.09 1902.62 Q1081.36 1906.15 1080.8 1912.55 L1107.37 1912.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1147.68 1916.44 Q1138.64 1916.44 1135.16 1918.5 Q1131.68 1920.57 1131.68 1925.55 Q1131.68 1929.52 1134.27 1931.87 Q1136.9 1934.18 1141.4 1934.18 Q1147.6 1934.18 1151.32 1929.8 Q1155.09 1925.39 1155.09 1918.1 L1155.09 1916.44 L1147.68 1916.44 M1162.54 1913.36 L1162.54 1939.24 L1155.09 1939.24 L1155.09 1932.36 Q1152.54 1936.49 1148.73 1938.47 Q1144.92 1940.42 1139.41 1940.42 Q1132.45 1940.42 1128.31 1936.53 Q1124.22 1932.6 1124.22 1926.04 Q1124.22 1918.38 1129.33 1914.49 Q1134.47 1910.6 1144.64 1910.6 L1155.09 1910.6 L1155.09 1909.87 Q1155.09 1904.73 1151.69 1901.93 Q1148.33 1899.1 1142.21 1899.1 Q1138.32 1899.1 1134.63 1900.03 Q1130.95 1900.96 1127.54 1902.83 L1127.54 1895.94 Q1131.64 1894.36 1135.48 1893.59 Q1139.33 1892.78 1142.98 1892.78 Q1152.82 1892.78 1157.68 1897.88 Q1162.54 1902.99 1162.54 1913.36 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1185.27 1880.99 L1185.27 1893.87 L1200.62 1893.87 L1200.62 1899.67 L1185.27 1899.67 L1185.27 1924.29 Q1185.27 1929.84 1186.77 1931.42 Q1188.31 1933 1192.97 1933 L1200.62 1933 L1200.62 1939.24 L1192.97 1939.24 Q1184.34 1939.24 1181.06 1936.04 Q1177.78 1932.8 1177.78 1924.29 L1177.78 1899.67 L1172.31 1899.67 L1172.31 1893.87 L1177.78 1893.87 L1177.78 1880.99 L1185.27 1880.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1210.43 1893.87 L1217.88 1893.87 L1217.88 1939.24 L1210.43 1939.24 L1210.43 1893.87 M1210.43 1876.21 L1217.88 1876.21 L1217.88 1885.65 L1210.43 1885.65 L1210.43 1876.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1271.19 1911.86 L1271.19 1939.24 L1263.74 1939.24 L1263.74 1912.1 Q1263.74 1905.66 1261.22 1902.46 Q1258.71 1899.26 1253.69 1899.26 Q1247.65 1899.26 1244.17 1903.11 Q1240.69 1906.96 1240.69 1913.6 L1240.69 1939.24 L1233.19 1939.24 L1233.19 1893.87 L1240.69 1893.87 L1240.69 1900.92 Q1243.36 1896.83 1246.96 1894.8 Q1250.61 1892.78 1255.35 1892.78 Q1263.17 1892.78 1267.18 1897.64 Q1271.19 1902.46 1271.19 1911.86 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1315.91 1916.03 Q1315.91 1907.93 1312.55 1903.47 Q1309.23 1899.02 1303.19 1899.02 Q1297.2 1899.02 1293.83 1903.47 Q1290.51 1907.93 1290.51 1916.03 Q1290.51 1924.09 1293.83 1928.55 Q1297.2 1933 1303.19 1933 Q1309.23 1933 1312.55 1928.55 Q1315.91 1924.09 1315.91 1916.03 M1323.36 1933.61 Q1323.36 1945.2 1318.22 1950.83 Q1313.08 1956.5 1302.46 1956.5 Q1298.53 1956.5 1295.05 1955.89 Q1291.57 1955.32 1288.28 1954.11 L1288.28 1946.86 Q1291.57 1948.64 1294.77 1949.49 Q1297.97 1950.34 1301.29 1950.34 Q1308.62 1950.34 1312.27 1946.49 Q1315.91 1942.69 1315.91 1934.95 L1315.91 1931.26 Q1313.6 1935.27 1310 1937.26 Q1306.39 1939.24 1301.37 1939.24 Q1293.02 1939.24 1287.92 1932.88 Q1282.82 1926.52 1282.82 1916.03 Q1282.82 1905.5 1287.92 1899.14 Q1293.02 1892.78 1301.37 1892.78 Q1306.39 1892.78 1310 1894.76 Q1313.6 1896.75 1315.91 1900.76 L1315.91 1893.87 L1323.36 1893.87 L1323.36 1933.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1382.99 1876.29 Q1377.57 1885.61 1374.93 1894.72 Q1372.3 1903.84 1372.3 1913.2 Q1372.3 1922.55 1374.93 1931.75 Q1377.61 1940.9 1382.99 1950.18 L1376.51 1950.18 Q1370.44 1940.66 1367.4 1931.46 Q1364.4 1922.27 1364.4 1913.2 Q1364.4 1904.16 1367.4 1895.01 Q1370.4 1885.85 1376.51 1876.29 L1382.99 1876.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1405.96 1910.36 L1405.96 1932.52 L1419.09 1932.52 Q1425.69 1932.52 1428.85 1929.8 Q1432.05 1927.05 1432.05 1921.42 Q1432.05 1915.75 1428.85 1913.07 Q1425.69 1910.36 1419.09 1910.36 L1405.96 1910.36 M1405.96 1885.49 L1405.96 1903.72 L1418.07 1903.72 Q1424.07 1903.72 1426.99 1901.49 Q1429.94 1899.22 1429.94 1894.6 Q1429.94 1890.02 1426.99 1887.76 Q1424.07 1885.49 1418.07 1885.49 L1405.96 1885.49 M1397.78 1878.76 L1418.68 1878.76 Q1428.04 1878.76 1433.1 1882.65 Q1438.17 1886.54 1438.17 1893.71 Q1438.17 1899.26 1435.57 1902.54 Q1432.98 1905.82 1427.96 1906.63 Q1433.99 1907.93 1437.32 1912.06 Q1440.68 1916.15 1440.68 1922.31 Q1440.68 1930.41 1435.17 1934.83 Q1429.66 1939.24 1419.49 1939.24 L1397.78 1939.24 L1397.78 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1446.31 1878.76 L1497.47 1878.76 L1497.47 1885.65 L1476 1885.65 L1476 1939.24 L1467.78 1939.24 L1467.78 1885.65 L1446.31 1885.65 L1446.31 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1504.44 1878.76 L1512.66 1878.76 L1512.66 1915.5 Q1512.66 1925.23 1516.19 1929.52 Q1519.71 1933.77 1527.61 1933.77 Q1535.47 1933.77 1538.99 1929.52 Q1542.52 1925.23 1542.52 1915.5 L1542.52 1878.76 L1550.74 1878.76 L1550.74 1916.52 Q1550.74 1928.35 1544.87 1934.38 Q1539.03 1940.42 1527.61 1940.42 Q1516.15 1940.42 1510.27 1934.38 Q1504.44 1928.35 1504.44 1916.52 L1504.44 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1579.02 1878.76 L1585.9 1878.76 L1564.84 1946.94 L1557.95 1946.94 L1579.02 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1616.69 1876.21 L1616.69 1882.41 L1609.56 1882.41 Q1605.55 1882.41 1603.97 1884.03 Q1602.43 1885.65 1602.43 1889.86 L1602.43 1893.87 L1614.71 1893.87 L1614.71 1899.67 L1602.43 1899.67 L1602.43 1939.24 L1594.94 1939.24 L1594.94 1899.67 L1587.81 1899.67 L1587.81 1893.87 L1594.94 1893.87 L1594.94 1890.71 Q1594.94 1883.14 1598.46 1879.69 Q1601.99 1876.21 1609.64 1876.21 L1616.69 1876.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1628.84 1880.99 L1628.84 1893.87 L1644.2 1893.87 L1644.2 1899.67 L1628.84 1899.67 L1628.84 1924.29 Q1628.84 1929.84 1630.34 1931.42 Q1631.88 1933 1636.54 1933 L1644.2 1933 L1644.2 1939.24 L1636.54 1939.24 Q1627.91 1939.24 1624.63 1936.04 Q1621.35 1932.8 1621.35 1924.29 L1621.35 1899.67 L1615.88 1899.67 L1615.88 1893.87 L1621.35 1893.87 L1621.35 1880.99 L1628.84 1880.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1667.25 1878.76 L1674.13 1878.76 L1653.07 1946.94 L1646.18 1946.94 L1667.25 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1704.92 1876.21 L1704.92 1882.41 L1697.79 1882.41 Q1693.78 1882.41 1692.2 1884.03 Q1690.66 1885.65 1690.66 1889.86 L1690.66 1893.87 L1702.93 1893.87 L1702.93 1899.67 L1690.66 1899.67 L1690.66 1939.24 L1683.17 1939.24 L1683.17 1899.67 L1676.04 1899.67 L1676.04 1893.87 L1683.17 1893.87 L1683.17 1890.71 Q1683.17 1883.14 1686.69 1879.69 Q1690.21 1876.21 1697.87 1876.21 L1704.92 1876.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1717.07 1880.99 L1717.07 1893.87 L1732.43 1893.87 L1732.43 1899.67 L1717.07 1899.67 L1717.07 1924.29 Q1717.07 1929.84 1718.57 1931.42 Q1720.11 1933 1724.77 1933 L1732.43 1933 L1732.43 1939.24 L1724.77 1939.24 Q1716.14 1939.24 1712.86 1936.04 Q1709.58 1932.8 1709.58 1924.29 L1709.58 1899.67 L1704.11 1899.67 L1704.11 1893.87 L1709.58 1893.87 L1709.58 1880.99 L1717.07 1880.99 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1755.47 1878.76 L1762.36 1878.76 L1741.3 1946.94 L1734.41 1946.94 L1755.47 1878.76 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1799.1 1895.21 L1799.1 1902.26 Q1795.94 1900.64 1792.54 1899.83 Q1789.14 1899.02 1785.49 1899.02 Q1779.94 1899.02 1777.15 1900.72 Q1774.39 1902.42 1774.39 1905.82 Q1774.39 1908.42 1776.38 1909.91 Q1778.36 1911.37 1784.36 1912.71 L1786.91 1913.28 Q1794.85 1914.98 1798.17 1918.1 Q1801.53 1921.18 1801.53 1926.73 Q1801.53 1933.04 1796.51 1936.73 Q1791.53 1940.42 1782.78 1940.42 Q1779.13 1940.42 1775.16 1939.69 Q1771.23 1939 1766.86 1937.58 L1766.86 1929.89 Q1770.99 1932.03 1775 1933.13 Q1779.01 1934.18 1782.94 1934.18 Q1788.21 1934.18 1791.04 1932.4 Q1793.88 1930.57 1793.88 1927.29 Q1793.88 1924.25 1791.81 1922.63 Q1789.79 1921.01 1782.86 1919.51 L1780.27 1918.91 Q1773.34 1917.45 1770.26 1914.45 Q1767.18 1911.41 1767.18 1906.15 Q1767.18 1899.75 1771.72 1896.26 Q1776.26 1892.78 1784.6 1892.78 Q1788.73 1892.78 1792.38 1893.39 Q1796.02 1893.99 1799.1 1895.21 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1852.21 1914.69 L1852.21 1918.34 L1817.94 1918.34 Q1818.43 1926.04 1822.56 1930.09 Q1826.73 1934.1 1834.14 1934.1 Q1838.44 1934.1 1842.45 1933.04 Q1846.5 1931.99 1850.47 1929.89 L1850.47 1936.93 Q1846.46 1938.64 1842.25 1939.53 Q1838.03 1940.42 1833.7 1940.42 Q1822.84 1940.42 1816.48 1934.1 Q1810.16 1927.78 1810.16 1917 Q1810.16 1905.86 1816.16 1899.34 Q1822.19 1892.78 1832.4 1892.78 Q1841.56 1892.78 1846.86 1898.69 Q1852.21 1904.57 1852.21 1914.69 M1844.76 1912.51 Q1844.68 1906.39 1841.31 1902.74 Q1837.99 1899.1 1832.48 1899.1 Q1826.24 1899.1 1822.48 1902.62 Q1818.75 1906.15 1818.18 1912.55 L1844.76 1912.51 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1897.09 1895.61 L1897.09 1902.58 Q1893.93 1900.84 1890.73 1899.99 Q1887.57 1899.1 1884.33 1899.1 Q1877.08 1899.1 1873.07 1903.72 Q1869.06 1908.29 1869.06 1916.6 Q1869.06 1924.9 1873.07 1929.52 Q1877.08 1934.1 1884.33 1934.1 Q1887.57 1934.1 1890.73 1933.25 Q1893.93 1932.36 1897.09 1930.61 L1897.09 1937.5 Q1893.98 1938.96 1890.61 1939.69 Q1887.29 1940.42 1883.52 1940.42 Q1873.27 1940.42 1867.24 1933.98 Q1861.2 1927.54 1861.2 1916.6 Q1861.2 1905.5 1867.28 1899.14 Q1873.4 1892.78 1884.01 1892.78 Q1887.45 1892.78 1890.73 1893.51 Q1894.02 1894.2 1897.09 1895.61 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><path clip-path="url(#clip690)" d="M1908.88 1876.29 L1915.36 1876.29 Q1921.44 1885.85 1924.44 1895.01 Q1927.48 1904.16 1927.48 1913.2 Q1927.48 1922.27 1924.44 1931.46 Q1921.44 1940.66 1915.36 1950.18 L1908.88 1950.18 Q1914.27 1940.9 1916.9 1931.75 Q1919.58 1922.55 1919.58 1913.2 Q1919.58 1903.84 1916.9 1894.72 Q1914.27 1885.61 1908.88 1876.29 Z" fill="#000000" fill-rule="nonzero" fill-opacity="1" /><polyline clip-path="url(#clip694)" style="stroke:#009af9; stroke-linecap:round; stroke-linejoin:round; stroke-width:8; stroke-opacity:1; fill:none" points="261.245,2459.47 266.064,2452.39 270.882,2445.05 275.701,2437.44 280.52,2429.56 285.338,2421.41 290.157,2412.97 294.975,2404.26 299.794,2395.25 304.613,2385.95 309.431,2376.37 314.25,2366.5 319.068,2356.34 323.887,2345.9 328.706,2335.18 333.524,2324.19 338.343,2312.95 343.161,2301.47 347.98,2289.76 352.799,2277.84 357.617,2265.74 362.436,2253.48 367.254,2241.1 372.073,2228.62 376.892,2216.09 381.71,2203.55 386.529,2191.05 391.348,2178.64 396.166,2166.37 400.985,2154.3 405.803,2142.5 410.622,2131.01 415.441,2119.93 420.259,2109.29 425.078,2099.19 429.896,2089.67 434.715,2080.8 439.534,2072.65 444.352,2065.26 449.171,2058.7 453.989,2053 458.808,2048.2 463.627,2044.33 468.445,2041.41 473.264,2039.45 478.083,2038.45 482.901,2038.4 487.72,2039.29 492.538,2041.08 497.357,2043.75 502.176,2047.25 506.994,2051.53 511.813,2056.54 516.631,2062.22 521.45,2068.51 526.269,2075.34 531.087,2082.66 535.906,2090.39 540.724,2098.47 545.543,2106.83 550.362,2115.43 555.18,2124.2 559.999,2133.08 564.818,2142.03 569.636,2150.99 574.455,2159.92 579.273,2168.79 584.092,2177.55 588.911,2186.18 593.729,2194.63 598.548,2202.9 603.366,2210.96 608.185,2218.78 613.004,2226.35 617.822,2233.65 622.641,2240.68 627.459,2247.42 632.278,2253.87 637.097,2260.02 641.915,2265.86 646.734,2271.4 651.553,2276.62 656.371,2281.53 661.19,2286.13 666.008,2290.41 670.827,2294.38 675.646,2298.04 680.464,2301.38 685.283,2304.42 690.101,2307.15 694.92,2309.57 699.739,2311.68 704.557,2313.5 709.376,2315.01 714.194,2316.22 719.013,2317.13 723.832,2317.75 728.65,2318.07 733.469,2318.09 738.288,2317.83 743.106,2317.27 747.925,2316.43 752.743,2315.3 757.562,2313.88 762.381,2312.18 767.199,2310.2 772.018,2307.93 776.836,2305.4 781.655,2302.59 786.474,2299.51 791.292,2296.17 796.111,2292.57 800.929,2288.73 805.748,2284.64 810.567,2280.33 815.385,2275.79 820.204,2271.06 825.023,2266.13 829.841,2261.03 834.66,2255.78 839.478,2250.41 844.297,2244.93 849.116,2239.38 853.934,2233.78 858.753,2228.18 863.571,2222.59 868.39,2217.07 873.209,2211.66 878.027,2206.38 882.846,2201.3 887.664,2196.44 892.483,2191.87 897.302,2187.61 902.12,2183.72 906.939,2180.22 911.758,2177.17 916.576,2174.59 921.395,2172.51 926.213,2170.96 931.032,2169.96 935.851,2169.51 940.669,2169.63 945.488,2170.3 950.306,2171.53 955.125,2173.3 959.944,2175.59 964.762,2178.37 969.581,2181.62 974.399,2185.28 979.218,2189.34 984.037,2193.73 988.855,2198.42 993.674,2203.37 998.492,2208.52 1003.31,2213.84 1008.13,2219.27 1012.95,2224.77 1017.77,2230.31 1022.59,2235.83 1027.4,2241.32 1032.22,2246.72 1037.04,2252.02 1041.86,2257.17 1046.68,2262.17 1051.5,2266.98 1056.32,2271.58 1061.13,2275.95 1065.95,2280.09 1070.77,2283.98 1075.59,2287.6 1080.41,2290.95 1085.23,2294.03 1090.05,2296.82 1094.86,2299.33 1099.68,2301.54 1104.5,2303.47 1109.32,2305.1 1114.14,2306.45 1118.96,2307.51 1123.78,2308.29 1128.59,2308.8 1133.41,2309.03 1138.23,2309.01 1143.05,2308.73 1147.87,2308.2 1152.69,2307.45 1157.51,2306.48 1162.33,2305.3 1167.14,2303.93 1171.96,2302.4 1176.78,2300.72 1181.6,2298.9 1186.42,2296.98 1191.24,2294.98 1196.06,2292.93 1200.87,2290.84 1205.69,2288.76 1210.51,2286.71 1215.33,2284.73 1220.15,2282.83 1224.97,2281.07 1229.79,2279.46 1234.6,2278.04 1239.42,2276.83 1244.24,2275.87 1249.06,2275.17 1253.88,2274.75 1258.7,2274.64 1263.52,2274.85 1268.33,2275.38 1273.15,2276.24 1277.97,2277.43 1282.79,2278.94 1287.61,2280.76 1292.43,2282.88 1297.25,2285.29 1302.06,2287.95 1306.88,2290.85 1311.7,2293.96 1316.52,2297.24 1321.34,2300.68 1326.16,2304.24 1330.98,2307.89 1335.8,2311.59 1340.61,2315.32 1345.43,2319.06 1350.25,2322.76 1355.07,2326.42 1359.89,2330 1364.71,2333.48 1369.53,2336.85 1374.34,2340.08 1379.16,2343.17 1383.98,2346.1 1388.8,2348.85 1393.62,2351.43 1398.44,2353.82 1403.26,2356.02 1408.07,2358.03 1412.89,2359.84 1417.71,2361.46 1422.53,2362.89 1427.35,2364.13 1432.17,2365.19 1436.99,2366.08 1441.8,2366.8 1446.62,2367.36 1451.44,2367.78 1456.26,2368.07 1461.08,2368.24 1465.9,2368.31 1470.72,2368.29 1475.53,2368.2 1480.35,2368.06 1485.17,2367.9 1489.99,2367.72 1494.81,2367.55 1499.63,2367.42 1504.45,2367.33 1509.27,2367.31 1514.08,2367.38 1518.9,2367.56 1523.72,2367.86 1528.54,2368.3 1533.36,2368.89 1538.18,2369.64 1543,2370.55 1547.81,2371.64 1552.63,2372.91 1557.45,2374.34 1562.27,2375.95 1567.09,2377.72 1571.91,2379.65 1576.73,2381.72 1581.54,2383.92 1586.36,2386.23 1591.18,2388.64 1596,2391.14 1600.82,2393.7 1605.64,2396.31 1610.46,2398.94 1615.27,2401.59 1620.09,2404.23 1624.91,2406.85 1629.73,2409.43 1634.55,2411.96 1639.37,2414.43 1644.19,2416.83 1649,2419.14 1653.82,2421.36 1658.64,2423.48 1663.46,2425.49 1668.28,2427.4 1673.1,2429.21 1677.92,2430.9 1682.74,2432.48 1687.55,2433.96 1692.37,2435.34 1697.19,2436.62 1702.01,2437.8 1706.83,2438.9 1711.65,2439.92 1716.47,2440.87 1721.28,2441.76 1726.1,2442.6 1730.92,2443.41 1735.74,2444.18 1740.56,2444.94 1745.38,2445.69 1750.2,2446.44 1755.01,2447.22 1759.83,2448.02 1764.65,2448.86 1769.47,2449.75 1774.29,2450.7 1779.11,2451.7 1783.93,2452.78 1788.74,2453.93 1793.56,2455.16 1798.38,2456.47 1803.2,2457.86 1808.02,2459.33 1812.84,2460.88 1817.66,2462.49 1822.47,2464.18 1827.29,2465.92 1832.11,2467.72 1836.93,2469.56 1841.75,2471.45 1846.57,2473.35 1851.39,2475.28 1856.21,2477.22 1861.02,2479.16 1865.84,2481.09 1870.66,2483 1875.48,2484.89 1880.3,2486.76 1885.12,2488.58 1889.94,2490.37 1894.75,2492.11 1899.57,2493.8 1904.39,2495.44 1909.21,2497.02 1914.03,2498.56 1918.85,2500.03 1923.67,2501.46 1928.48,2502.83 1933.3,2504.15 1938.12,2505.43 1942.94,2506.67 1947.76,2507.86 1952.58,2509.02 1957.4,2510.15 1962.21,2511.26 1967.03,2512.34 1971.85,2513.42 1976.67,2514.48 1981.49,2515.55 1986.31,2516.61 1991.13,2517.68 1995.94,2518.77 2000.76,2519.87 2005.58,2520.99 2010.4,2522.13 2015.22,2523.3 2020.04,2524.5 2024.86,2525.72 2029.68,2526.97 2034.49,2528.25 2039.31,2529.56 2044.13,2530.9 2048.95,2532.25 2053.77,2533.63 2058.59,2535.03 2063.41,2536.44 2068.22,2537.86 2073.04,2539.29 2077.86,2540.72 2082.68,2542.15 2087.5,2543.58 2092.32,2545 2097.14,2546.41 2101.95,2547.8 2106.77,2549.17 2111.59,2550.53 2116.41,2551.86 2121.23,2553.17 2126.05,2554.45 2130.87,2555.71 2135.68,2556.94 2140.5,2558.15 2145.32,2559.33 2150.14,2560.48 2154.96,2561.61 2159.78,2562.72 2164.6,2563.81 2169.41,2564.88 2174.23,2565.93 2179.05,2566.96 2183.87,2567.99 2188.69,2569 2193.51,2570 2198.33,2571 2203.15,2571.99 2207.96,2572.98 2212.78,2573.97 2217.6,2574.96 2222.42,2575.96 2227.24,2576.96 2232.06,2577.96 2236.88,2578.96 2241.69,2579.97 2246.51,2580.99 2251.33,2582.01 2256.15,2583.04 2260.97,2584.06 2265.79,2585.09 2270.61,2586.13 2275.42,2587.16 2280.24,2588.19 2285.06,2589.21 2289.88,2590.23 2294.7,2591.25 2299.52,2592.25 2304.34,2593.25 2309.15,2594.24 2313.97,2595.21 2318.79,2596.17 2323.61,2597.12 2328.43,2598.05 2333.25,2598.97 2338.07,2599.87 2342.88,2600.76 2347.7,2601.63 2352.52,2602.48 2357.34,2603.32 2362.16,2604.14 2366.98,2604.95 2371.8,2605.74 2376.62,2606.52 2381.43,2607.29 2386.25,2608.04 2391.07,2608.79 2395.89,2609.52 2400.71,2610.25 2405.53,2610.97 2410.35,2611.68 2415.16,2612.38 2419.98,2613.09 2424.8,2613.78 2429.62,2614.48 2434.44,2615.17 2439.26,2615.85 2444.08,2616.54 2448.89,2617.22 2453.71,2617.91 2458.53,2618.58 2463.35,2619.26 2468.17,2619.93 2472.99,2620.6 2477.81,2621.27 2482.62,2621.92 2487.44,2622.58 2492.26,2623.22 2497.08,2623.86 2501.9,2624.49 2506.72,2625.11 2511.54,2625.71 2516.35,2626.31 2521.17,2626.89 2525.99,2627.46 2530.81,2628.01 2535.63,2628.55 2540.45,2629.07 2545.27,2629.57 2550.09,2630.06 2554.9,2630.53 2559.72,2630.99 2564.54,2631.43 2569.36,2631.85 2574.18,2632.26 2579,2632.65 2583.82,2633.03 2588.63,2633.39 2593.45,2633.75 2598.27,2634.09 2603.09,2634.41 2607.91,2634.73 2612.73,2635.05 2617.55,2635.35 2622.36,2635.65 2627.18,2635.95 2632,2636.25 2636.82,2636.55 2641.64,2636.85 2646.46,2637.15 2651.28,2637.46 2656.09,2637.78 2660.91,2638.11 2665.73,2638.44 2670.55,2638.79 2675.37,2639.15 2680.19,2637.25 "/>
+</svg>
diff --git a/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/index.html b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/index.html
new file mode 100644
index 00000000000..1af1792dece
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/space_shuttle_reentry_trajectory/index.html
@@ -0,0 +1,270 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Optimal control for a Space Shuttle reentry trajectory · JuMP</title><meta name="title" content="Optimal control for a Space Shuttle reentry trajectory · JuMP"/><meta property="og:title" content="Optimal control for a Space Shuttle reentry trajectory · JuMP"/><meta property="twitter:title" content="Optimal control for a Space Shuttle reentry trajectory · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li class="is-active"><a class="tocitem" href>Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Optimal control for a Space Shuttle reentry trajectory</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Optimal control for a Space Shuttle reentry trajectory</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/space_shuttle_reentry_trajectory.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="Optimal-control-for-a-Space-Shuttle-reentry-trajectory"><a class="docs-heading-anchor" href="#Optimal-control-for-a-Space-Shuttle-reentry-trajectory">Optimal control for a Space Shuttle reentry trajectory</a><a id="Optimal-control-for-a-Space-Shuttle-reentry-trajectory-1"></a><a class="docs-heading-anchor-permalink" href="#Optimal-control-for-a-Space-Shuttle-reentry-trajectory" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../space_shuttle_reentry_trajectory.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p><strong>This tutorial was originally contributed by Henrique Ferrolho.</strong></p><p>This tutorial demonstrates how to compute a reentry trajectory for the <a href="https://en.wikipedia.org/wiki/Space_Shuttle">Space Shuttle</a>, by formulating and solving a nonlinear programming problem. The problem was drawn from Chapter 6 of <a href="https://epubs.siam.org/doi/book/10.1137/1.9780898718577">&quot;Practical Methods for Optimal Control and Estimation Using Nonlinear Programming&quot;</a>, by John T. Betts.</p><div class="admonition is-success"><header class="admonition-header">Tip</header><div class="admonition-body"><p>This tutorial is a more-complicated version of the <a href="../rocket_control/#Rocket-Control">Rocket Control</a> example. If you are new to solving nonlinear programs in JuMP, you may want to start there instead.</p></div></div><p>The motion of the vehicle is defined by the following set of DAEs:</p><p class="math-container">\[\begin{aligned}
+\dot{h} &amp; = v \sin \gamma , \\
+\dot{\phi} &amp; = \frac{v}{r} \cos \gamma \sin \psi / \cos \theta , \\
+\dot{\theta} &amp; = \frac{v}{r} \cos \gamma \cos \psi , \\
+\dot{v} &amp; = -\frac{D}{m} - g \sin \gamma , \\
+\dot{\gamma} &amp; = \frac{L}{m v} \cos(\beta) + \cos \gamma \left ( \frac{v}{r} - \frac{g}{v} \right ) , \\
+\dot{\psi} &amp; = \frac{1}{m v \cos \gamma} L \sin(\beta) + \frac{v}{r \cos \theta} \cos \gamma \sin \psi \sin \theta , \\
+q &amp; \le q_U , \\
+\end{aligned}\]</p><p>where the aerodynamic heating on the vehicle wing leading edge is <span>$q = q_a q_r$</span> and the dynamic variables are</p><p class="math-container">\[\begin{aligned}
+     h &amp; \quad \text{altitude (ft)},     \qquad &amp; &amp; \gamma \quad \text{flight path angle (rad)}, \\
+  \phi &amp; \quad \text{longitude (rad)},   \qquad &amp; &amp; \psi   \quad \text{azimuth (rad)},           \\
+\theta &amp; \quad \text{latitude (rad)},    \qquad &amp; &amp; \alpha \quad \text{angle of attack (rad)},   \\
+     v &amp; \quad \text{velocity (ft/sec)}, \qquad &amp; &amp; \beta  \quad \text{bank angle (rad)}.
+\end{aligned}\]</p><p>The aerodynamic and atmospheric forces on the vehicle are specified by the following quantities (English units):</p><p class="math-container">\[\begin{aligned}
+           D &amp; = \frac{1}{2} c_D S \rho v^2,                  \qquad &amp; a_0 &amp; = -0.20704, \\
+           L &amp; = \frac{1}{2} c_L S \rho v^2,                  \qquad &amp; a_1 &amp; =  0.029244, \\
+           g &amp; = \mu / r^2,                                   \qquad &amp; \mu &amp; =  0.14076539 \times 10^{17}, \\
+           r &amp; = R_e + h,                                     \qquad &amp; b_0 &amp; =  0.07854, \\
+        \rho &amp; = \rho_0 \exp[-h/h_r],                         \qquad &amp; b_1 &amp; = -0.61592  \times 10^{-2}, \\
+      \rho_0 &amp; = 0.002378,                                    \qquad &amp; b_2 &amp; =  0.621408 \times 10^{-3}, \\
+         h_r &amp; = 23800,                                       \qquad &amp; q_r &amp; =  17700 \sqrt{\rho} (0.0001 v)^{3.07}, \\
+         c_L &amp; = a_0 + a_1 \hat{\alpha},                      \qquad &amp; q_a &amp; =  c_0 + c_1 \hat{\alpha} + c_2 \hat{\alpha}^2 + c_3 \hat{\alpha}^3, \\
+         c_D &amp; = b_0 + b_1 \hat{\alpha} + b_2 \hat{\alpha}^2, \qquad &amp; c_0 &amp; =  1.0672181, \\
+\hat{\alpha} &amp; = 180 \alpha / \pi,                            \qquad &amp; c_1 &amp; = -0.19213774 \times 10^{-1}, \\
+         R_e &amp; = 20902900,                                    \qquad &amp; c_2 &amp; =  0.21286289 \times 10^{-3}, \\
+           S &amp; = 2690,                                        \qquad &amp; c_3 &amp; = -0.10117249 \times 10^{-5}.
+\end{aligned}\]</p><p>The reentry trajectory begins at an altitude where the aerodynamic forces are quite small with a weight of <span>$w = 203000$</span> (lb) and mass <span>$m = w / g_0$</span> (slug), where <span>$g_0 = 32.174$</span> (ft/sec<span>$^2$</span>). The initial conditions are as follows:</p><p class="math-container">\[\begin{aligned}
+     h &amp; = 260000 \text{ ft},  \qquad &amp; v      &amp; = 25600 \text{ ft/sec}, \\
+  \phi &amp; = 0      \text{ deg}, \qquad &amp; \gamma &amp; = -1    \text{ deg},    \\
+\theta &amp; = 0      \text{ deg}, \qquad &amp; \psi   &amp; = 90    \text{ deg}.
+\end{aligned}\]</p><p>The final point on the reentry trajectory occurs at the unknown (free) time <span>$t_F$</span>, at the so-called terminal area energy management (TAEM) interface, which is defined by the conditions</p><p class="math-container">\[h = 80000 \text{ ft}, \qquad v = 2500 \text{ ft/sec}, \qquad \gamma = -5 \text{ deg}.\]</p><p>As explained in the book, our goal is to maximize the final cross-range, which is equivalent to maximizing the final latitude of the vehicle, that is, <span>$J = \theta(t_F)$</span>.</p><p><img src="../../../assets/space_shuttle.png" alt="Max cross-range shuttle reentry"/></p><h3 id="Approach"><a class="docs-heading-anchor" href="#Approach">Approach</a><a id="Approach-1"></a><a class="docs-heading-anchor-permalink" href="#Approach" title="Permalink"></a></h3><p>We will use a discretized model of time, with a fixed number of discretized points, <span>$n$</span>. The decision variables at each point are going to be the state of the vehicle and the controls commanded to it. In addition, we will also make each time step size <span>$\Delta t$</span> a decision variable; that way, we can either fix the time step size easily, or allow the solver to fine-tune the duration between each adjacent pair of points. Finally, in order to approximate the derivatives of the problem dynamics, we will use either rectangular or <a href="http://en.wikipedia.org/wiki/Trapezoidal_rule">trapezoidal</a> integration.</p><div class="admonition is-warning"><header class="admonition-header">Warning</header><div class="admonition-body"><p>Do not try to actually land a Space Shuttle using this notebook. There&#39;s no mesh refinement going on, which can lead to unrealistic trajectories having position and velocity errors with orders of magnitude <span>$10^4$</span> ft and <span>$10^2$</span> ft/sec, respectively.</p></div></div><pre><code class="language-julia hljs">using JuMP
+import Interpolations
+import Ipopt
+
+# Global variables
+const w = 203000.0  # weight (lb)
+const g₀ = 32.174    # acceleration (ft/sec^2)
+const m = w / g₀    # mass (slug)
+
+# Aerodynamic and atmospheric forces on the vehicle
+const ρ₀ = 0.002378
+const hᵣ = 23800.0
+const Rₑ = 20902900.0
+const μ = 0.14076539e17
+const S = 2690.0
+const a₀ = -0.20704
+const a₁ = 0.029244
+const b₀ = 0.07854
+const b₁ = -0.61592e-2
+const b₂ = 0.621408e-3
+const c₀ = 1.0672181
+const c₁ = -0.19213774e-1
+const c₂ = 0.21286289e-3
+const c₃ = -0.10117249e-5
+
+# Initial conditions
+const h_s = 2.6          # altitude (ft) / 1e5
+const ϕ_s = deg2rad(0)   # longitude (rad)
+const θ_s = deg2rad(0)   # latitude (rad)
+const v_s = 2.56         # velocity (ft/sec) / 1e4
+const γ_s = deg2rad(-1)  # flight path angle (rad)
+const ψ_s = deg2rad(90)  # azimuth (rad)
+const α_s = deg2rad(0)   # angle of attack (rad)
+const β_s = deg2rad(0)   # bank angle (rad)
+const t_s = 1.00         # time step (sec)
+
+# Final conditions, the so-called Terminal Area Energy Management (TAEM)
+const h_t = 0.8          # altitude (ft) / 1e5
+const v_t = 0.25         # velocity (ft/sec) / 1e4
+const γ_t = deg2rad(-5)  # flight path angle (rad)
+
+# Number of mesh points (knots) to be used
+const n = 503
+
+# Integration scheme to be used for the dynamics
+const integration_rule = &quot;rectangular&quot;</code></pre><div class="admonition is-success"><header class="admonition-header">Choose a good linear solver</header><div class="admonition-body"><p>Picking a good linear solver is <strong>extremely important</strong> to maximize the performance of nonlinear solvers. For the best results, it is advised to experiment different linear solvers.</p><p>For example, the linear solver <code>MA27</code> is outdated and can be quite slow. <code>MA57</code> is a much better alternative, especially for highly-sparse problems (such as trajectory optimization problems).</p></div></div><pre><code class="language-julia hljs"># Uncomment the lines below to pass user options to the solver
+user_options = (
+# &quot;mu_strategy&quot; =&gt; &quot;monotone&quot;,
+# &quot;linear_solver&quot; =&gt; &quot;ma27&quot;,
+)
+
+# Create JuMP model, using Ipopt as the solver
+model = Model(optimizer_with_attributes(Ipopt.Optimizer, user_options...))
+
+@variables(model, begin
+    0 ≤ scaled_h[1:n]                # altitude (ft) / 1e5
+    ϕ[1:n]                # longitude (rad)
+    deg2rad(-89) ≤ θ[1:n] ≤ deg2rad(89)  # latitude (rad)
+    1e-4 ≤ scaled_v[1:n]                # velocity (ft/sec) / 1e4
+    deg2rad(-89) ≤ γ[1:n] ≤ deg2rad(89)  # flight path angle (rad)
+    ψ[1:n]                # azimuth (rad)
+    deg2rad(-90) ≤ α[1:n] ≤ deg2rad(90)  # angle of attack (rad)
+    deg2rad(-89) ≤ β[1:n] ≤ deg2rad(1)  # bank angle (rad)
+    #        3.5 ≤       Δt[1:n] ≤ 4.5          # time step (sec)
+    Δt[1:n] == 4.0         # time step (sec)
+end);</code></pre><div class="admonition is-info"><header class="admonition-header">Info</header><div class="admonition-body"><p>Above you can find two alternatives for the <code>Δt</code> variables.</p><p>The first one, <code>3.5 ≤ Δt[1:n] ≤ 4.5</code> (currently commented), allows some wiggle room for the solver to adjust the time step size between pairs of mesh points. This is neat because it allows the solver to figure out which parts of the flight require more dense discretization than others. (Remember, the number of discretized points is fixed, and this example does not implement mesh refinement.) However, this makes the problem more complex to solve, and therefore leads to a longer computation time.</p><p>The second line, <code>Δt[1:n] == 4.0</code>, fixes the duration of every time step to exactly 4.0 seconds. This allows the problem to be solved faster. However, to do this we need to know beforehand that the close-to-optimal total duration of the flight is ~2009 seconds. Therefore, if we split the total duration in slices of 4.0 seconds, we know that we require <code>n = 503</code> knots to discretize the whole trajectory.</p></div></div><pre><code class="language-julia hljs"># Fix initial conditions
+fix(scaled_h[1], h_s; force = true)
+fix(ϕ[1], ϕ_s; force = true)
+fix(θ[1], θ_s; force = true)
+fix(scaled_v[1], v_s; force = true)
+fix(γ[1], γ_s; force = true)
+fix(ψ[1], ψ_s; force = true)
+
+# Fix final conditions
+fix(scaled_h[n], h_t; force = true)
+fix(scaled_v[n], v_t; force = true)
+fix(γ[n], γ_t; force = true)
+
+# Initial guess: linear interpolation between boundary conditions
+x_s = [h_s, ϕ_s, θ_s, v_s, γ_s, ψ_s, α_s, β_s, t_s]
+x_t = [h_t, ϕ_s, θ_s, v_t, γ_t, ψ_s, α_s, β_s, t_s]
+interp_linear = Interpolations.LinearInterpolation([1, n], [x_s, x_t])
+initial_guess = mapreduce(transpose, vcat, interp_linear.(1:n))
+set_start_value.(all_variables(model), vec(initial_guess))
+
+# Functions to restore `h` and `v` to their true scale
+@expression(model, h[j = 1:n], scaled_h[j] * 1e5)
+@expression(model, v[j = 1:n], scaled_v[j] * 1e4)
+
+# Helper functions
+@expression(model, c_L[j = 1:n], a₀ + a₁ * rad2deg(α[j]))
+@expression(model, c_D[j = 1:n], b₀ + b₁ * rad2deg(α[j]) + b₂ * rad2deg(α[j])^2)
+@expression(model, ρ[j = 1:n], ρ₀ * exp(-h[j] / hᵣ))
+@expression(model, D[j = 1:n], 0.5 * c_D[j] * S * ρ[j] * v[j]^2)
+@expression(model, L[j = 1:n], 0.5 * c_L[j] * S * ρ[j] * v[j]^2)
+@expression(model, r[j = 1:n], Rₑ + h[j])
+@expression(model, g[j = 1:n], μ / r[j]^2)
+
+# Motion of the vehicle as a differential-algebraic system of equations (DAEs)
+@expression(model, δh[j = 1:n], v[j] * sin(γ[j]))
+@expression(
+    model,
+    δϕ[j = 1:n],
+    (v[j] / r[j]) * cos(γ[j]) * sin(ψ[j]) / cos(θ[j])
+)
+@expression(model, δθ[j = 1:n], (v[j] / r[j]) * cos(γ[j]) * cos(ψ[j]))
+@expression(model, δv[j = 1:n], -(D[j] / m) - g[j] * sin(γ[j]))
+@expression(
+    model,
+    δγ[j = 1:n],
+    (L[j] / (m * v[j])) * cos(β[j]) +
+    cos(γ[j]) * ((v[j] / r[j]) - (g[j] / v[j]))
+)
+@expression(
+    model,
+    δψ[j = 1:n],
+    (1 / (m * v[j] * cos(γ[j]))) * L[j] * sin(β[j]) +
+    (v[j] / (r[j] * cos(θ[j]))) * cos(γ[j]) * sin(ψ[j]) * sin(θ[j])
+)
+
+# System dynamics
+for j in 2:n
+    i = j - 1  # index of previous knot
+
+    if integration_rule == &quot;rectangular&quot;
+        # Rectangular integration
+        @constraint(model, h[j] == h[i] + Δt[i] * δh[i])
+        @constraint(model, ϕ[j] == ϕ[i] + Δt[i] * δϕ[i])
+        @constraint(model, θ[j] == θ[i] + Δt[i] * δθ[i])
+        @constraint(model, v[j] == v[i] + Δt[i] * δv[i])
+        @constraint(model, γ[j] == γ[i] + Δt[i] * δγ[i])
+        @constraint(model, ψ[j] == ψ[i] + Δt[i] * δψ[i])
+    elseif integration_rule == &quot;trapezoidal&quot;
+        # Trapezoidal integration
+        @constraint(model, h[j] == h[i] + 0.5 * Δt[i] * (δh[j] + δh[i]))
+        @constraint(model, ϕ[j] == ϕ[i] + 0.5 * Δt[i] * (δϕ[j] + δϕ[i]))
+        @constraint(model, θ[j] == θ[i] + 0.5 * Δt[i] * (δθ[j] + δθ[i]))
+        @constraint(model, v[j] == v[i] + 0.5 * Δt[i] * (δv[j] + δv[i]))
+        @constraint(model, γ[j] == γ[i] + 0.5 * Δt[i] * (δγ[j] + δγ[i]))
+        @constraint(model, ψ[j] == ψ[i] + 0.5 * Δt[i] * (δψ[j] + δψ[i]))
+    else
+        @error &quot;Unexpected integration rule &#39;$(integration_rule)&#39;&quot;
+    end
+end
+
+# Objective: Maximize cross-range
+@objective(model, Max, θ[n])
+
+set_silent(model)  # Hide solver&#39;s verbose output
+optimize!(model)  # Solve for the control and state
+@assert termination_status(model) == LOCALLY_SOLVED
+
+# Show final cross-range of the solution
+println(
+    &quot;Final latitude θ = &quot;,
+    round(objective_value(model) |&gt; rad2deg; digits = 2),
+    &quot;°&quot;,
+)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Final latitude θ = 34.18°</code></pre><h3 id="Plotting-the-results"><a class="docs-heading-anchor" href="#Plotting-the-results">Plotting the results</a><a id="Plotting-the-results-1"></a><a class="docs-heading-anchor-permalink" href="#Plotting-the-results" title="Permalink"></a></h3><p>Let&#39;s plot the results to visualize the optimal trajectory:</p><pre><code class="language-julia hljs">using Plots
+ts = cumsum([0; value.(Δt)])[1:end-1]
+plt_altitude = plot(
+    ts,
+    value.(scaled_h);
+    legend = nothing,
+    title = &quot;Altitude (100,000 ft)&quot;,
+)
+plt_longitude =
+    plot(ts, rad2deg.(value.(ϕ)); legend = nothing, title = &quot;Longitude (deg)&quot;)
+plt_latitude =
+    plot(ts, rad2deg.(value.(θ)); legend = nothing, title = &quot;Latitude (deg)&quot;)
+plt_velocity = plot(
+    ts,
+    value.(scaled_v);
+    legend = nothing,
+    title = &quot;Velocity (1000 ft/sec)&quot;,
+)
+plt_flight_path =
+    plot(ts, rad2deg.(value.(γ)); legend = nothing, title = &quot;Flight Path (deg)&quot;)
+plt_azimuth =
+    plot(ts, rad2deg.(value.(ψ)); legend = nothing, title = &quot;Azimuth (deg)&quot;)
+
+plot(
+    plt_altitude,
+    plt_velocity,
+    plt_longitude,
+    plt_flight_path,
+    plt_latitude,
+    plt_azimuth;
+    layout = grid(3, 2),
+    linewidth = 2,
+    size = (700, 700),
+)</code></pre><img src="78651b37.svg" alt="Example block output"/><pre><code class="language-julia hljs">function q(h, v, a)
+    ρ(h) = ρ₀ * exp(-h / hᵣ)
+    qᵣ(h, v) = 17700 * √ρ(h) * (0.0001 * v)^3.07
+    qₐ(a) = c₀ + c₁ * rad2deg(a) + c₂ * rad2deg(a)^2 + c₃ * rad2deg(a)^3
+    # Aerodynamic heating on the vehicle wing leading edge
+    return qₐ(a) * qᵣ(h, v)
+end
+
+plt_attack_angle = plot(
+    ts[1:end-1],
+    rad2deg.(value.(α)[1:end-1]);
+    legend = nothing,
+    title = &quot;Angle of Attack (deg)&quot;,
+)
+plt_bank_angle = plot(
+    ts[1:end-1],
+    rad2deg.(value.(β)[1:end-1]);
+    legend = nothing,
+    title = &quot;Bank Angle (deg)&quot;,
+)
+plt_heating = plot(
+    ts,
+    q.(value.(scaled_h) * 1e5, value.(scaled_v) * 1e4, value.(α));
+    legend = nothing,
+    title = &quot;Heating (BTU/ft/ft/sec)&quot;,
+)
+
+plot(
+    plt_attack_angle,
+    plt_bank_angle,
+    plt_heating;
+    layout = grid(3, 1),
+    linewidth = 2,
+    size = (700, 700),
+)</code></pre><img src="fa7e5a76.svg" alt="Example block output"/><pre><code class="language-julia hljs">plot(
+    rad2deg.(value.(ϕ)),
+    rad2deg.(value.(θ)),
+    value.(scaled_h);
+    linewidth = 2,
+    legend = nothing,
+    title = &quot;Space Shuttle Reentry Trajectory&quot;,
+    xlabel = &quot;Longitude (deg)&quot;,
+    ylabel = &quot;Latitude (deg)&quot;,
+    zlabel = &quot;Altitude (100,000 ft)&quot;,
+)</code></pre><img src="0d7a8519.svg" alt="Example block output"/></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../rocket_control/">« Rocket Control</a><a class="docs-footer-nextpage" href="../tips_and_tricks/">Tips and tricks »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/tips_and_tricks.jl b/previews/PR3525/tutorials/nonlinear/tips_and_tricks.jl
new file mode 100644
index 00000000000..26f96d0d97f
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/tips_and_tricks.jl
@@ -0,0 +1,129 @@
+# Copyright 2017, Iain Dunning, Joey Huchette, Miles Lubin, and contributors    #src
+# This Source Code Form is subject to the terms of the Mozilla Public License   #src
+# v.2.0. If a copy of the MPL was not distributed with this file, You can       #src
+# obtain one at https://mozilla.org/MPL/2.0/.                                   #src
+
+# # [Tips and tricks](@id nonlinear_tips_and_tricks)
+
+# This example collates some tips and tricks you can use when formulating
+# nonlinear programs. It uses the following packages:
+
+using JuMP
+import Ipopt
+import Test
+
+# ## User-defined operators with vector outputs
+
+# A common situation is to have a user-defined operator like the following that
+# returns multiple outputs (we define `function_calls` to keep track of how
+# many times we call this method):
+
+function_calls = 0
+function foo(x, y)
+    global function_calls += 1
+    common_term = x^2 + y^2
+    term_1 = sqrt(1 + common_term)
+    term_2 = common_term
+    return term_1, term_2
+end
+
+# For example, the first term might be used in the objective, and the second
+# term might be used in a constraint, and often they share work that is
+# expensive to evaluate.
+
+# This is a problem for JuMP, because it requires user-defined operators to
+# return a single number. One option is to define two separate functions, the
+# first returning the first argument, and the second returning the second
+# argument.
+
+foo_1(x, y) = foo(x, y)[1]
+foo_2(x, y) = foo(x, y)[2]
+
+# However, if the common term is expensive to compute, this approach is wasteful
+# because it will evaluate the expensive term twice. Let's have a look at how
+# many times we evaluate `x^2 + y^2` during a solve:
+
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[1:2] >= 0, start = 0.1)
+@operator(model, op_foo_1, 2, foo_1)
+@operator(model, op_foo_2, 2, foo_2)
+@objective(model, Max, op_foo_1(x[1], x[2]))
+@constraint(model, op_foo_2(x[1], x[2]) <= 2)
+function_calls = 0
+optimize!(model)
+Test.@test objective_value(model) ≈ √3 atol = 1e-4
+Test.@test value.(x) ≈ [1.0, 1.0] atol = 1e-4
+println("Naive approach: function calls = $(function_calls)")
+naive_approach = function_calls  #src
+
+# An alternative approach is to use _memoization_, which uses a cache to store
+# the result of function evaluations. We can write a memoization function as
+# follows:
+
+"""
+    memoize(foo::Function, n_outputs::Int)
+
+Take a function `foo` and return a vector of length `n_outputs`, where element
+`i` is a function that returns the equivalent of `foo(x...)[i]`.
+
+To avoid duplication of work, cache the most-recent evaluations of `foo`.
+Because `foo_i` is auto-differentiated with ForwardDiff, our cache needs to
+work when `x` is a `Float64` and a `ForwardDiff.Dual`.
+"""
+function memoize(foo::Function, n_outputs::Int)
+    last_x, last_f = nothing, nothing
+    last_dx, last_dfdx = nothing, nothing
+    function foo_i(i, x::T...) where {T<:Real}
+        if T == Float64
+            if x !== last_x
+                last_x, last_f = x, foo(x...)
+            end
+            return last_f[i]::T
+        else
+            if x !== last_dx
+                last_dx, last_dfdx = x, foo(x...)
+            end
+            return last_dfdx[i]::T
+        end
+    end
+    return [(x...) -> foo_i(i, x...) for i in 1:n_outputs]
+end
+
+# Let's see how it works. First, construct the memoized versions of `foo`:
+
+memoized_foo = memoize(foo, 2)
+
+# Now try evaluating the first element of `memoized_foo`.
+
+function_calls = 0
+memoized_foo[1](1.0, 1.0)
+Test.@test function_calls == 1  #src
+println("function_calls = ", function_calls)
+
+# As expected, this evaluated the function once. However, if we call the
+# function again, we hit the cache instead of needing to re-compute `foo` and
+# `function_calls` is still `1`!
+
+memoized_foo[1](1.0, 1.0)
+Test.@test function_calls == 1  #src
+println("function_calls = ", function_calls)
+
+# Now let's see how this works during a real solve:
+
+model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[1:2] >= 0, start = 0.1)
+@operator(model, op_foo_1, 2, memoized_foo[1])
+@operator(model, op_foo_2, 2, memoized_foo[2])
+@objective(model, Max, op_foo_1(x[1], x[2]))
+@constraint(model, op_foo_2(x[1], x[2]) <= 2)
+function_calls = 0
+optimize!(model)
+Test.@test objective_value(model) ≈ √3 atol = 1e-4
+Test.@test value.(x) ≈ [1.0, 1.0] atol = 1e-4
+println("Memoized approach: function_calls = $(function_calls)")
+Test.@test function_calls <= naive_approach / 2 + 1  #src
+
+# Compared to the naive approach, the memoized approach requires half as many
+# function evaluations.
diff --git a/previews/PR3525/tutorials/nonlinear/tips_and_tricks/index.html b/previews/PR3525/tutorials/nonlinear/tips_and_tricks/index.html
new file mode 100644
index 00000000000..b4ffa465468
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/tips_and_tricks/index.html
@@ -0,0 +1,70 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>Tips and tricks · JuMP</title><meta name="title" content="Tips and tricks · JuMP"/><meta property="og:title" content="Tips and tricks · JuMP"/><meta property="twitter:title" content="Tips and tricks · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li class="is-active"><a class="tocitem" href>Tips and tricks</a><ul class="internal"><li><a class="tocitem" href="#User-defined-operators-with-vector-outputs"><span>User-defined operators with vector outputs</span></a></li></ul></li><li><a class="tocitem" href="../user_defined_hessians/">User-defined Hessians</a></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>Tips and tricks</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>Tips and tricks</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/tips_and_tricks.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="nonlinear_tips_and_tricks"><a class="docs-heading-anchor" href="#nonlinear_tips_and_tricks">Tips and tricks</a><a id="nonlinear_tips_and_tricks-1"></a><a class="docs-heading-anchor-permalink" href="#nonlinear_tips_and_tricks" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../tips_and_tricks.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>This example collates some tips and tricks you can use when formulating nonlinear programs. It uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import Ipopt
+import Test</code></pre><h2 id="User-defined-operators-with-vector-outputs"><a class="docs-heading-anchor" href="#User-defined-operators-with-vector-outputs">User-defined operators with vector outputs</a><a id="User-defined-operators-with-vector-outputs-1"></a><a class="docs-heading-anchor-permalink" href="#User-defined-operators-with-vector-outputs" title="Permalink"></a></h2><p>A common situation is to have a user-defined operator like the following that returns multiple outputs (we define <code>function_calls</code> to keep track of how many times we call this method):</p><pre><code class="language-julia hljs">function_calls = 0
+function foo(x, y)
+    global function_calls += 1
+    common_term = x^2 + y^2
+    term_1 = sqrt(1 + common_term)
+    term_2 = common_term
+    return term_1, term_2
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">foo (generic function with 1 method)</code></pre><p>For example, the first term might be used in the objective, and the second term might be used in a constraint, and often they share work that is expensive to evaluate.</p><p>This is a problem for JuMP, because it requires user-defined operators to return a single number. One option is to define two separate functions, the first returning the first argument, and the second returning the second argument.</p><pre><code class="language-julia hljs">foo_1(x, y) = foo(x, y)[1]
+foo_2(x, y) = foo(x, y)[2]</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">foo_2 (generic function with 1 method)</code></pre><p>However, if the common term is expensive to compute, this approach is wasteful because it will evaluate the expensive term twice. Let&#39;s have a look at how many times we evaluate <code>x^2 + y^2</code> during a solve:</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[1:2] &gt;= 0, start = 0.1)
+@operator(model, op_foo_1, 2, foo_1)
+@operator(model, op_foo_2, 2, foo_2)
+@objective(model, Max, op_foo_1(x[1], x[2]))
+@constraint(model, op_foo_2(x[1], x[2]) &lt;= 2)
+function_calls = 0
+optimize!(model)
+Test.@test objective_value(model) ≈ √3 atol = 1e-4
+Test.@test value.(x) ≈ [1.0, 1.0] atol = 1e-4
+println(&quot;Naive approach: function calls = $(function_calls)&quot;)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Naive approach: function calls = 44</code></pre><p>An alternative approach is to use <em>memoization</em>, which uses a cache to store the result of function evaluations. We can write a memoization function as follows:</p><pre><code class="language-julia hljs">&quot;&quot;&quot;
+    memoize(foo::Function, n_outputs::Int)
+
+Take a function `foo` and return a vector of length `n_outputs`, where element
+`i` is a function that returns the equivalent of `foo(x...)[i]`.
+
+To avoid duplication of work, cache the most-recent evaluations of `foo`.
+Because `foo_i` is auto-differentiated with ForwardDiff, our cache needs to
+work when `x` is a `Float64` and a `ForwardDiff.Dual`.
+&quot;&quot;&quot;
+function memoize(foo::Function, n_outputs::Int)
+    last_x, last_f = nothing, nothing
+    last_dx, last_dfdx = nothing, nothing
+    function foo_i(i, x::T...) where {T&lt;:Real}
+        if T == Float64
+            if x !== last_x
+                last_x, last_f = x, foo(x...)
+            end
+            return last_f[i]::T
+        else
+            if x !== last_dx
+                last_dx, last_dfdx = x, foo(x...)
+            end
+            return last_dfdx[i]::T
+        end
+    end
+    return [(x...) -&gt; foo_i(i, x...) for i in 1:n_outputs]
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Main.memoize</code></pre><p>Let&#39;s see how it works. First, construct the memoized versions of <code>foo</code>:</p><pre><code class="language-julia hljs">memoized_foo = memoize(foo, 2)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">2-element Vector{Main.var&quot;#4#7&quot;{Int64, Main.var&quot;#foo_i#5&quot;{typeof(Main.foo)}}}:
+ #4 (generic function with 1 method)
+ #4 (generic function with 1 method)</code></pre><p>Now try evaluating the first element of <code>memoized_foo</code>.</p><pre><code class="language-julia hljs">function_calls = 0
+memoized_foo[1](1.0, 1.0)
+println(&quot;function_calls = &quot;, function_calls)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">function_calls = 1</code></pre><p>As expected, this evaluated the function once. However, if we call the function again, we hit the cache instead of needing to re-compute <code>foo</code> and <code>function_calls</code> is still <code>1</code>!</p><pre><code class="language-julia hljs">memoized_foo[1](1.0, 1.0)
+println(&quot;function_calls = &quot;, function_calls)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">function_calls = 1</code></pre><p>Now let&#39;s see how this works during a real solve:</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+set_silent(model)
+@variable(model, x[1:2] &gt;= 0, start = 0.1)
+@operator(model, op_foo_1, 2, memoized_foo[1])
+@operator(model, op_foo_2, 2, memoized_foo[2])
+@objective(model, Max, op_foo_1(x[1], x[2]))
+@constraint(model, op_foo_2(x[1], x[2]) &lt;= 2)
+function_calls = 0
+optimize!(model)
+Test.@test objective_value(model) ≈ √3 atol = 1e-4
+Test.@test value.(x) ≈ [1.0, 1.0] atol = 1e-4
+println(&quot;Memoized approach: function_calls = $(function_calls)&quot;)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">Memoized approach: function_calls = 22</code></pre><p>Compared to the naive approach, the memoized approach requires half as many function evaluations.</p></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../space_shuttle_reentry_trajectory/">« Optimal control for a Space Shuttle reentry trajectory</a><a class="docs-footer-nextpage" href="../user_defined_hessians/">User-defined Hessians »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
diff --git a/previews/PR3525/tutorials/nonlinear/user_defined_hessians.jl b/previews/PR3525/tutorials/nonlinear/user_defined_hessians.jl
new file mode 100644
index 00000000000..b7f75e42d20
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/user_defined_hessians.jl
@@ -0,0 +1,75 @@
+# Copyright (c) 2022 Oscar Dowson and contributors                               #src
+#                                                                                #src
+# Permission is hereby granted, free of charge, to any person obtaining a copy   #src
+# of this software and associated documentation files (the "Software"), to deal  #src
+# in the Software without restriction, including without limitation the rights   #src
+# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell      #src
+# copies of the Software, and to permit persons to whom the Software is          #src
+# furnished to do so, subject to the following conditions:                       #src
+#                                                                                #src
+# The above copyright notice and this permission notice shall be included in all #src
+# copies or substantial portions of the Software.                                #src
+#                                                                                #src
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR     #src
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,       #src
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE    #src
+# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER         #src
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,  #src
+# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE  #src
+# SOFTWARE.                                                                      #src
+
+# # User-defined Hessians
+
+# In this tutorial, we explain how to write a user-defined operator (see
+# [User-defined operators](@ref jump_user_defined_operators)) with a Hessian
+# matrix explicitly provided by the user.
+#
+# For a more advanced example, see [Nested optimization problems](@ref).
+
+# This tutorial uses the following packages:
+
+using JuMP
+import Ipopt
+
+# ## Rosenbrock example
+
+# As a simple example, we consider the Rosenbrock function:
+
+rosenbrock(x...) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2
+
+# which has the gradient vector:
+
+function ∇rosenbrock(g::AbstractVector, x...)
+    g[1] = 400 * x[1]^3 - 400 * x[1] * x[2] + 2 * x[1] - 2
+    g[2] = 200 * (x[2] - x[1]^2)
+    return
+end
+
+# and the Hessian matrix:
+
+function ∇²rosenbrock(H::AbstractMatrix, x...)
+    H[1, 1] = 1200 * x[1]^2 - 400 * x[2] + 2
+    ## H[1, 2] = -400 * x[1] <-- not needed because Hessian is symmetric
+    H[2, 1] = -400 * x[1]
+    H[2, 2] = 200.0
+    return
+end
+
+# You may assume the Hessian matrix `H` is initialized with zeros, and
+# because it is symmetric you need only to fill in the non-zero of the
+# lower-triangular terms.
+
+# The matrix type passed in as `H` depends on the automatic differentiation
+# system, so make sure the first argument to the Hessian function supports an
+# `AbstractMatrix` (it may be something other than `Matrix{Float64}`). However,
+# you may assume only that `H` supports `size(H)` and `setindex!`.
+
+# Now that we have the function, its gradient, and its Hessian, we can construct
+# a JuMP model, add the operator, and use it in a macro:
+
+model = Model(Ipopt.Optimizer)
+@variable(model, x[1:2])
+@operator(model, op_rosenbrock, 2, rosenbrock, ∇rosenbrock, ∇²rosenbrock)
+@objective(model, Min, op_rosenbrock(x[1], x[2]))
+optimize!(model)
+solution_summary(model; verbose = true)
diff --git a/previews/PR3525/tutorials/nonlinear/user_defined_hessians/index.html b/previews/PR3525/tutorials/nonlinear/user_defined_hessians/index.html
new file mode 100644
index 00000000000..30b6e6c4a4c
--- /dev/null
+++ b/previews/PR3525/tutorials/nonlinear/user_defined_hessians/index.html
@@ -0,0 +1,42 @@
+<!DOCTYPE html>
+<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>User-defined Hessians · JuMP</title><meta name="title" content="User-defined Hessians · JuMP"/><meta property="og:title" content="User-defined Hessians · JuMP"/><meta property="twitter:title" content="User-defined Hessians · JuMP"/><meta name="description" content="Documentation for JuMP."/><meta property="og:description" content="Documentation for JuMP."/><meta property="twitter:description" content="Documentation for JuMP."/><script async src="https://www.googletagmanager.com/gtag/js?id=UA-44252521-1"></script><script>  window.dataLayer = window.dataLayer || [];
+  function gtag(){dataLayer.push(arguments);}
+  gtag('js', new Date());
+  gtag('config', 'UA-44252521-1', {'page_path': location.pathname + location.search + location.hash});
+</script><script data-outdated-warner src="../../../assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="../../.."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="../../../assets/documenter.js"></script><script src="../../../search_index.js"></script><script src="../../../siteinfo.js"></script><script src="../../../../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="../../../assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="../../../assets/themeswap.js"></script><link href="../../../assets/extra_styles.css" rel="stylesheet" type="text/css"/></head><body><div id="documenter"><nav class="docs-sidebar"><a class="docs-logo" href="../../../"><img class="docs-light-only" src="../../../assets/logo.svg" alt="JuMP logo"/><img class="docs-dark-only" src="../../../assets/logo-dark.svg" alt="JuMP logo"/></a><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li><input class="collapse-toggle" id="menuitem-1" type="checkbox"/><label class="tocitem" for="menuitem-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../">Introduction</a></li><li><a class="tocitem" href="../../../should_i_use/">Should you use JuMP?</a></li><li><a class="tocitem" href="../../../installation/">Installation Guide</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2" type="checkbox" checked/><label class="tocitem" for="menuitem-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-2-1" type="checkbox"/><label class="tocitem" for="menuitem-2-1"><span class="docs-label">Getting started</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../getting_started/introduction/">Introduction</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_julia/">Getting started with Julia</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_JuMP/">Getting started with JuMP</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_sets_and_indexing/">Getting started with sets and indexing</a></li><li><a class="tocitem" href="../../getting_started/getting_started_with_data_and_plotting/">Getting started with data and plotting</a></li><li><a class="tocitem" href="../../getting_started/debugging/">Debugging</a></li><li><a class="tocitem" href="../../getting_started/design_patterns_for_larger_models/">Design patterns for larger models</a></li><li><a class="tocitem" href="../../getting_started/performance_tips/">Performance tips</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-2" type="checkbox"/><label class="tocitem" for="menuitem-2-2"><span class="docs-label">Linear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../linear/introduction/">Introduction</a></li><li><a class="tocitem" href="../../linear/knapsack/">The knapsack problem example</a></li><li><a class="tocitem" href="../../linear/diet/">The diet problem</a></li><li><a class="tocitem" href="../../linear/cannery/">The cannery problem</a></li><li><a class="tocitem" href="../../linear/factory_schedule/">The factory schedule example</a></li><li><a class="tocitem" href="../../linear/multi/">The multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/multi_commodity_network/">The network multi-commodity flow problem</a></li><li><a class="tocitem" href="../../linear/tips_and_tricks/">Tips and tricks</a></li><li><a class="tocitem" href="../../linear/facility_location/">The facility location problem</a></li><li><a class="tocitem" href="../../linear/finance/">Financial modeling problems</a></li><li><a class="tocitem" href="../../linear/geographic_clustering/">Geographical clustering</a></li><li><a class="tocitem" href="../../linear/network_flows/">Network flow problems</a></li><li><a class="tocitem" href="../../linear/transp/">The transportation problem</a></li><li><a class="tocitem" href="../../linear/multi_objective_knapsack/">Multi-objective knapsack</a></li><li><a class="tocitem" href="../../linear/multi_objective_examples/">Simple multi-objective examples</a></li><li><a class="tocitem" href="../../linear/sudoku/">Sudoku</a></li><li><a class="tocitem" href="../../linear/n-queens/">N-Queens</a></li><li><a class="tocitem" href="../../linear/constraint_programming/">Constraint programming</a></li><li><a class="tocitem" href="../../linear/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../linear/lp_sensitivity/">Sensitivity analysis of a linear program</a></li><li><a class="tocitem" href="../../linear/mip_duality/">Computing the duals of a mixed-integer program</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-3" type="checkbox" checked/><label class="tocitem" for="menuitem-2-3"><span class="docs-label">Nonlinear programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../introduction/">Introduction</a></li><li><a class="tocitem" href="../simple_examples/">Simple examples</a></li><li><a class="tocitem" href="../portfolio/">Portfolio optimization</a></li><li><a class="tocitem" href="../rocket_control/">Rocket Control</a></li><li><a class="tocitem" href="../space_shuttle_reentry_trajectory/">Optimal control for a Space Shuttle reentry trajectory</a></li><li><a class="tocitem" href="../tips_and_tricks/">Tips and tricks</a></li><li class="is-active"><a class="tocitem" href>User-defined Hessians</a><ul class="internal"><li><a class="tocitem" href="#Rosenbrock-example"><span>Rosenbrock example</span></a></li></ul></li><li><a class="tocitem" href="../nested_problems/">Nested optimization problems</a></li><li><a class="tocitem" href="../querying_hessians/">Computing Hessians</a></li><li><a class="tocitem" href="../complementarity/">Mixed complementarity problems</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-4" type="checkbox"/><label class="tocitem" for="menuitem-2-4"><span class="docs-label">Conic programs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../conic/introduction/">Introduction</a></li><li><a class="tocitem" href="../../conic/start_values/">Primal and dual warm-starts</a></li><li><a class="tocitem" href="../../conic/tips_and_tricks/">Tips and Tricks</a></li><li><a class="tocitem" href="../../conic/simple_examples/">Simple semidefinite programming examples</a></li><li><a class="tocitem" href="../../conic/dualization/">Dualization</a></li><li><a class="tocitem" href="../../conic/arbitrary_precision/">Arbitrary precision arithmetic</a></li><li><a class="tocitem" href="../../conic/logistic_regression/">Logistic regression</a></li><li><a class="tocitem" href="../../conic/experiment_design/">Experiment design</a></li><li><a class="tocitem" href="../../conic/min_ellipse/">Minimal ellipses</a></li><li><a class="tocitem" href="../../conic/ellipse_approx/">Ellipsoid approximation</a></li><li><a class="tocitem" href="../../conic/quantum_discrimination/">Quantum state discrimination</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-5" type="checkbox"/><label class="tocitem" for="menuitem-2-5"><span class="docs-label">Algorithms</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../algorithms/benders_decomposition/">Benders decomposition</a></li><li><a class="tocitem" href="../../algorithms/cutting_stock_column_generation/">Column generation</a></li><li><a class="tocitem" href="../../algorithms/tsp_lazy_constraints/">Traveling Salesperson Problem</a></li><li><a class="tocitem" href="../../algorithms/parallelism/">Parallelism</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-2-6" type="checkbox"/><label class="tocitem" for="menuitem-2-6"><span class="docs-label">Applications</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../applications/power_systems/">Power Systems</a></li><li><a class="tocitem" href="../../applications/optimal_power_flow/">Optimal power flow</a></li><li><a class="tocitem" href="../../applications/web_app/">Serving web apps</a></li><li><a class="tocitem" href="../../applications/two_stage_stochastic/">Two-stage stochastic programs</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-3" type="checkbox"/><label class="tocitem" for="menuitem-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../manual/models/">Models</a></li><li><a class="tocitem" href="../../../manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../manual/expressions/">Expressions</a></li><li><a class="tocitem" href="../../../manual/objective/">Objectives</a></li><li><a class="tocitem" href="../../../manual/containers/">Containers</a></li><li><a class="tocitem" href="../../../manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../manual/callbacks/">Solver-independent Callbacks</a></li><li><a class="tocitem" href="../../../manual/complex/">Complex number support</a></li><li><a class="tocitem" href="../../../manual/nonlinear/">Nonlinear Modeling</a></li><li><a class="tocitem" href="../../../manual/nlp/">Nonlinear Modeling (Legacy)</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-4" type="checkbox"/><label class="tocitem" for="menuitem-4"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../api/JuMP/">JuMP</a></li><li><a class="tocitem" href="../../../api/JuMP.Containers/">JuMP.Containers</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-5" type="checkbox"/><label class="tocitem" for="menuitem-5"><span class="docs-label">Background Information</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../background/algebraic_modeling_languages/">Algebraic modeling languages</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-6" type="checkbox"/><label class="tocitem" for="menuitem-6"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../developers/contributing/">Contributing</a></li><li><a class="tocitem" href="../../../developers/extensions/">Extensions</a></li><li><a class="tocitem" href="../../../developers/custom_solver_binaries/">Custom binaries</a></li><li><a class="tocitem" href="../../../developers/style/">Style Guide</a></li><li><a class="tocitem" href="../../../developers/roadmap/">Roadmap</a></li><li><a class="tocitem" href="../../../developers/checklists/">Checklists</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-7" type="checkbox"/><label class="tocitem" for="menuitem-7"><span class="docs-label">Solvers</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../packages/solvers/">Introduction</a></li><li><a class="tocitem" href="../../../packages/AmplNLWriter/">jump-dev/AmplNLWriter.jl</a></li><li><a class="tocitem" href="../../../packages/BARON/">jump-dev/BARON.jl</a></li><li><a class="tocitem" href="../../../packages/CPLEX/">jump-dev/CPLEX.jl</a></li><li><a class="tocitem" href="../../../packages/CSDP/">jump-dev/CSDP.jl</a></li><li><a class="tocitem" href="../../../packages/Cbc/">jump-dev/Cbc.jl</a></li><li><a class="tocitem" href="../../../packages/Clp/">jump-dev/Clp.jl</a></li><li><a class="tocitem" href="../../../packages/Dualization/">jump-dev/Dualization.jl</a></li><li><a class="tocitem" href="../../../packages/ECOS/">jump-dev/ECOS.jl</a></li><li><a class="tocitem" href="../../../packages/GLPK/">jump-dev/GLPK.jl</a></li><li><a class="tocitem" href="../../../packages/Gurobi/">jump-dev/Gurobi.jl</a></li><li><a class="tocitem" href="../../../packages/HiGHS/">jump-dev/HiGHS.jl</a></li><li><a class="tocitem" href="../../../packages/Ipopt/">jump-dev/Ipopt.jl</a></li><li><a class="tocitem" href="../../../packages/KNITRO/">jump-dev/KNITRO.jl</a></li><li><a class="tocitem" href="../../../packages/MiniZinc/">jump-dev/MiniZinc.jl</a></li><li><a class="tocitem" href="../../../packages/MosekTools/">jump-dev/MosekTools.jl</a></li><li><a class="tocitem" href="../../../packages/MultiObjectiveAlgorithms/">jump-dev/MultiObjectiveAlgorithms.jl</a></li><li><a class="tocitem" href="../../../packages/Pajarito/">jump-dev/Pajarito.jl</a></li><li><a class="tocitem" href="../../../packages/Pavito/">jump-dev/Pavito.jl</a></li><li><a class="tocitem" href="../../../packages/SCS/">jump-dev/SCS.jl</a></li><li><a class="tocitem" href="../../../packages/SDPA/">jump-dev/SDPA.jl</a></li><li><a class="tocitem" href="../../../packages/SDPNAL/">jump-dev/SDPNAL.jl</a></li><li><a class="tocitem" href="../../../packages/SDPT3/">jump-dev/SDPT3.jl</a></li><li><a class="tocitem" href="../../../packages/SeDuMi/">jump-dev/SeDuMi.jl</a></li><li><a class="tocitem" href="../../../packages/Xpress/">jump-dev/Xpress.jl</a></li><li><a class="tocitem" href="../../../packages/COPT/">COPT-Public/COPT.jl</a></li><li><a class="tocitem" href="../../../packages/GAMS/">GAMS-dev/GAMS.jl</a></li><li><a class="tocitem" href="../../../packages/NLopt/">JuliaOpt/NLopt.jl</a></li><li><a class="tocitem" href="../../../packages/CDDLib/">JuliaPolyhedra/CDDLib.jl</a></li><li><a class="tocitem" href="../../../packages/MadNLP/">MadNLP/MadNLP.jl</a></li><li><a class="tocitem" href="../../../packages/PATHSolver/">chkwon/PATHSolver.jl</a></li><li><a class="tocitem" href="../../../packages/Hypatia/">chriscoey/Hypatia.jl</a></li><li><a class="tocitem" href="../../../packages/DAQP/">darnstrom/DAQP.jl</a></li><li><a class="tocitem" href="../../../packages/Tulip/">ds4dm/Tulip.jl</a></li><li><a class="tocitem" href="../../../packages/Loraine/">kocvara/Loraine.jl</a></li><li><a class="tocitem" href="../../../packages/Alpine/">lanl-ansi/Alpine.jl</a></li><li><a class="tocitem" href="../../../packages/Juniper/">lanl-ansi/Juniper.jl</a></li><li><a class="tocitem" href="../../../packages/ProxSDP/">mariohsouto/ProxSDP.jl</a></li><li><a class="tocitem" href="../../../packages/NEOSServer/">odow/NEOSServer.jl</a></li><li><a class="tocitem" href="../../../packages/OSQP/">osqp/OSQP.jl</a></li><li><a class="tocitem" href="../../../packages/CDCS/">oxfordcontrol/CDCS.jl</a></li><li><a class="tocitem" href="../../../packages/COSMO/">oxfordcontrol/COSMO.jl</a></li><li><a class="tocitem" href="../../../packages/Clarabel/">oxfordcontrol/Clarabel.jl</a></li><li><a class="tocitem" href="../../../packages/SCIP/">scipopt/SCIP.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-8" type="checkbox"/><label class="tocitem" for="menuitem-8"><span class="docs-label">Extensions</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../extensions/introduction/">Introduction</a></li><li><a class="tocitem" href="../../../packages/DiffOpt/">jump-dev/DiffOpt.jl</a></li><li><a class="tocitem" href="../../../packages/ParametricOptInterface/">jump-dev/ParametricOptInterface.jl</a></li><li><a class="tocitem" href="../../../packages/PolyJuMP/">jump-dev/PolyJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SumOfSquares/">jump-dev/SumOfSquares.jl</a></li><li><a class="tocitem" href="../../../packages/InfiniteOpt/">infiniteopt/InfiniteOpt.jl</a></li><li><a class="tocitem" href="../../../packages/BilevelJuMP/">joaquimg/BilevelJuMP.jl</a></li><li><a class="tocitem" href="../../../packages/SDDP/">odow/SDDP.jl</a></li><li><a class="tocitem" href="../../../packages/Plasmo/">plasmo-dev/Plasmo.jl</a></li><li><a class="tocitem" href="../../../extensions/DimensionalData/">rafaqz/DimensionalData.jl</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9" type="checkbox"/><label class="tocitem" for="menuitem-9"><span class="docs-label">MathOptInterface</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-1" type="checkbox"/><label class="tocitem" for="menuitem-9-1"><span class="docs-label">Introduction</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/">Introduction</a></li><li><a class="tocitem" href="../../../moi/background/motivation/">Motivation</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-2" type="checkbox"/><label class="tocitem" for="menuitem-9-2"><span class="docs-label">Tutorials</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/tutorials/example/">Solving a problem using MathOptInterface</a></li><li><a class="tocitem" href="../../../moi/tutorials/implementing/">Implementing a solver interface</a></li><li><a class="tocitem" href="../../../moi/tutorials/mathprogbase/">Transitioning from MathProgBase</a></li><li><a class="tocitem" href="../../../moi/tutorials/bridging_constraint/">Implementing a constraint bridge</a></li><li><a class="tocitem" href="../../../moi/tutorials/manipulating_expressions/">Manipulating expressions</a></li><li><a class="tocitem" href="../../../moi/tutorials/latency/">Latency</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-3" type="checkbox"/><label class="tocitem" for="menuitem-9-3"><span class="docs-label">Manual</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/manual/standard_form/">Standard form problem</a></li><li><a class="tocitem" href="../../../moi/manual/models/">Models</a></li><li><a class="tocitem" href="../../../moi/manual/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/manual/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/manual/solutions/">Solutions</a></li><li><a class="tocitem" href="../../../moi/manual/modification/">Problem modification</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-4" type="checkbox"/><label class="tocitem" for="menuitem-9-4"><span class="docs-label">Background</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/background/duality/">Duality</a></li><li><a class="tocitem" href="../../../moi/background/infeasibility_certificates/">Infeasibility certificates</a></li><li><a class="tocitem" href="../../../moi/background/naming_conventions/">Naming conventions</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-5" type="checkbox"/><label class="tocitem" for="menuitem-9-5"><span class="docs-label">API Reference</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/reference/standard_form/">Standard form</a></li><li><a class="tocitem" href="../../../moi/reference/models/">Models</a></li><li><a class="tocitem" href="../../../moi/reference/variables/">Variables</a></li><li><a class="tocitem" href="../../../moi/reference/constraints/">Constraints</a></li><li><a class="tocitem" href="../../../moi/reference/modification/">Modifications</a></li><li><a class="tocitem" href="../../../moi/reference/nonlinear/">Nonlinear programming</a></li><li><a class="tocitem" href="../../../moi/reference/callbacks/">Callbacks</a></li><li><a class="tocitem" href="../../../moi/reference/errors/">Errors</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6"><span class="docs-label">Submodules</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><input class="collapse-toggle" id="menuitem-9-6-1" type="checkbox"/><label class="tocitem" for="menuitem-9-6-1"><span class="docs-label">Benchmarks</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Benchmarks/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-2" type="checkbox"/><label class="tocitem" for="menuitem-9-6-2"><span class="docs-label">Bridges</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Bridges/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/list_of_bridges/">List of bridges</a></li><li><a class="tocitem" href="../../../moi/submodules/Bridges/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-3" type="checkbox"/><label class="tocitem" for="menuitem-9-6-3"><span class="docs-label">FileFormats</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/FileFormats/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/FileFormats/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-4" type="checkbox"/><label class="tocitem" for="menuitem-9-6-4"><span class="docs-label">Nonlinear</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Nonlinear/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-5" type="checkbox"/><label class="tocitem" for="menuitem-9-6-5"><span class="docs-label">Utilities</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Utilities/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Utilities/reference/">API Reference</a></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-6-6" type="checkbox"/><label class="tocitem" for="menuitem-9-6-6"><span class="docs-label">Test</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/submodules/Test/overview/">Overview</a></li><li><a class="tocitem" href="../../../moi/submodules/Test/reference/">API Reference</a></li></ul></li></ul></li><li><input class="collapse-toggle" id="menuitem-9-7" type="checkbox"/><label class="tocitem" for="menuitem-9-7"><span class="docs-label">Developer Docs</span><i class="docs-chevron"></i></label><ul class="collapsed"><li><a class="tocitem" href="../../../moi/developer/checklists/">Checklists</a></li></ul></li><li><a class="tocitem" href="../../../moi/release_notes/">Release notes</a></li></ul></li><li><a class="tocitem" href="../../../release_notes/">Release notes</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li><a class="is-disabled">Tutorials</a></li><li><a class="is-disabled">Nonlinear programs</a></li><li class="is-active"><a href>User-defined Hessians</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>User-defined Hessians</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/jump-dev/JuMP.jl/blob/master/docs/src/tutorials/nonlinear/user_defined_hessians.jl" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="User-defined-Hessians"><a class="docs-heading-anchor" href="#User-defined-Hessians">User-defined Hessians</a><a id="User-defined-Hessians-1"></a><a class="docs-heading-anchor-permalink" href="#User-defined-Hessians" title="Permalink"></a></h1><p><em>This tutorial was generated using <a href="https://github.com/fredrikekre/Literate.jl">Literate.jl</a>.</em> <a href="../user_defined_hessians.jl"><em>Download the source as a <code>.jl</code> file</em></a>.</p><p>In this tutorial, we explain how to write a user-defined operator (see <a href="../../../manual/nonlinear/#jump_user_defined_operators">User-defined operators</a>) with a Hessian matrix explicitly provided by the user.</p><p>For a more advanced example, see <a href="../nested_problems/#Nested-optimization-problems">Nested optimization problems</a>.</p><p>This tutorial uses the following packages:</p><pre><code class="language-julia hljs">using JuMP
+import Ipopt</code></pre><h2 id="Rosenbrock-example"><a class="docs-heading-anchor" href="#Rosenbrock-example">Rosenbrock example</a><a id="Rosenbrock-example-1"></a><a class="docs-heading-anchor-permalink" href="#Rosenbrock-example" title="Permalink"></a></h2><p>As a simple example, we consider the Rosenbrock function:</p><pre><code class="language-julia hljs">rosenbrock(x...) = (1 - x[1])^2 + 100 * (x[2] - x[1]^2)^2</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">rosenbrock (generic function with 1 method)</code></pre><p>which has the gradient vector:</p><pre><code class="language-julia hljs">function ∇rosenbrock(g::AbstractVector, x...)
+    g[1] = 400 * x[1]^3 - 400 * x[1] * x[2] + 2 * x[1] - 2
+    g[2] = 200 * (x[2] - x[1]^2)
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">∇rosenbrock (generic function with 1 method)</code></pre><p>and the Hessian matrix:</p><pre><code class="language-julia hljs">function ∇²rosenbrock(H::AbstractMatrix, x...)
+    H[1, 1] = 1200 * x[1]^2 - 400 * x[2] + 2
+    # H[1, 2] = -400 * x[1] &lt;-- not needed because Hessian is symmetric
+    H[2, 1] = -400 * x[1]
+    H[2, 2] = 200.0
+    return
+end</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">∇²rosenbrock (generic function with 1 method)</code></pre><p>You may assume the Hessian matrix <code>H</code> is initialized with zeros, and because it is symmetric you need only to fill in the non-zero of the lower-triangular terms.</p><p>The matrix type passed in as <code>H</code> depends on the automatic differentiation system, so make sure the first argument to the Hessian function supports an <code>AbstractMatrix</code> (it may be something other than <code>Matrix{Float64}</code>). However, you may assume only that <code>H</code> supports <code>size(H)</code> and <code>setindex!</code>.</p><p>Now that we have the function, its gradient, and its Hessian, we can construct a JuMP model, add the operator, and use it in a macro:</p><pre><code class="language-julia hljs">model = Model(Ipopt.Optimizer)
+@variable(model, x[1:2])
+@operator(model, op_rosenbrock, 2, rosenbrock, ∇rosenbrock, ∇²rosenbrock)
+@objective(model, Min, op_rosenbrock(x[1], x[2]))
+optimize!(model)
+solution_summary(model; verbose = true)</code></pre><pre class="documenter-example-output"><code class="nohighlight hljs ansi">* Solver : Ipopt
+
+* Status
+  Result count       : 1
+  Termination status : LOCALLY_SOLVED
+  Message from the solver:
+  &quot;Solve_Succeeded&quot;
+
+* Candidate solution (result #1)
+  Primal status      : FEASIBLE_POINT
+  Dual status        : FEASIBLE_POINT
+  Objective value    : 8.14943e-28
+  Dual objective value : 0.00000e+00
+  Primal solution :
+    x[1] : 1.00000e+00
+    x[2] : 1.00000e+00
+  Dual solution :
+
+* Work counters
+  Solve time (sec)   : 4.97761e-02
+</code></pre></article><nav class="docs-footer"><a class="docs-footer-prevpage" href="../tips_and_tricks/">« Tips and tricks</a><a class="docs-footer-nextpage" href="../nested_problems/">Nested optimization problems »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option><option value="auto">Automatic (OS)</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.0.0 on <span class="colophon-date" title="Saturday 23 September 2023 04:49">Saturday 23 September 2023</span>. Using Julia version 1.9.3.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>